{ "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": 0.6761029958724976, "learning_rate": 2.25e-07, "loss": 1.1815, "step": 10 }, { "grad_norm": 0.7563536763191223, "learning_rate": 4.75e-07, "loss": 1.1666, "step": 20 }, { "grad_norm": 0.5853390097618103, "learning_rate": 7.25e-07, "loss": 1.1662, "step": 30 }, { "grad_norm": 0.5181789398193359, "learning_rate": 9.75e-07, "loss": 1.1668, "step": 40 }, { "grad_norm": 0.5811195373535156, "learning_rate": 1.2250000000000001e-06, "loss": 1.165, "step": 50 }, { "grad_norm": 0.6445841193199158, "learning_rate": 1.475e-06, "loss": 1.1408, "step": 60 }, { "grad_norm": 0.8516707420349121, "learning_rate": 1.7250000000000002e-06, "loss": 1.0966, "step": 70 }, { "grad_norm": 0.8835493922233582, "learning_rate": 1.975e-06, "loss": 1.1136, "step": 80 }, { "grad_norm": 1.5233590602874756, "learning_rate": 2.225e-06, "loss": 1.0919, "step": 90 }, { "grad_norm": 0.5479657053947449, "learning_rate": 2.4750000000000004e-06, "loss": 1.0819, "step": 100 }, { "grad_norm": 0.697577178478241, "learning_rate": 2.725e-06, "loss": 1.082, "step": 110 }, { "grad_norm": 0.698092520236969, "learning_rate": 2.975e-06, "loss": 1.0778, "step": 120 }, { "grad_norm": 0.7650100588798523, "learning_rate": 3.225e-06, "loss": 1.1299, "step": 130 }, { "grad_norm": 1.0576108694076538, "learning_rate": 3.4750000000000006e-06, "loss": 1.087, "step": 140 }, { "grad_norm": 1.2974655628204346, "learning_rate": 3.725e-06, "loss": 1.0752, "step": 150 }, { "grad_norm": 0.7555553913116455, "learning_rate": 3.975e-06, "loss": 1.104, "step": 160 }, { "grad_norm": 0.9139035940170288, "learning_rate": 4.225e-06, "loss": 1.087, "step": 170 }, { "grad_norm": 3.3658993244171143, "learning_rate": 4.475e-06, "loss": 1.1038, "step": 180 }, { "grad_norm": 1.1735931634902954, "learning_rate": 4.7250000000000005e-06, "loss": 1.0619, "step": 190 }, { "grad_norm": 1.0521894693374634, "learning_rate": 4.975000000000001e-06, "loss": 1.0843, "step": 200 }, { "grad_norm": 1.0023242235183716, "learning_rate": 5.225e-06, "loss": 1.0605, "step": 210 }, { "grad_norm": 0.8745774030685425, "learning_rate": 5.475e-06, "loss": 1.0601, "step": 220 }, { "grad_norm": 0.9043939709663391, "learning_rate": 5.725e-06, "loss": 1.0847, "step": 230 }, { "grad_norm": 0.8872116208076477, "learning_rate": 5.975e-06, "loss": 1.0827, "step": 240 }, { "grad_norm": 1.0163224935531616, "learning_rate": 6.2250000000000005e-06, "loss": 1.0603, "step": 250 }, { "grad_norm": 0.8510384559631348, "learning_rate": 6.475000000000001e-06, "loss": 1.0693, "step": 260 }, { "grad_norm": 1.0313646793365479, "learning_rate": 6.725000000000001e-06, "loss": 1.0626, "step": 270 }, { "grad_norm": 1.0553487539291382, "learning_rate": 6.975000000000001e-06, "loss": 1.0503, "step": 280 }, { "grad_norm": 1.0697721242904663, "learning_rate": 7.2249999999999994e-06, "loss": 1.0699, "step": 290 }, { "grad_norm": 1.2373771667480469, "learning_rate": 7.4750000000000004e-06, "loss": 1.0562, "step": 300 }, { "grad_norm": 1.1066043376922607, "learning_rate": 7.725e-06, "loss": 1.0482, "step": 310 }, { "grad_norm": 1.1379404067993164, "learning_rate": 7.975e-06, "loss": 1.0286, "step": 320 }, { "grad_norm": 1.427206039428711, "learning_rate": 8.225e-06, "loss": 0.9964, "step": 330 }, { "grad_norm": 1.6374868154525757, "learning_rate": 8.475000000000001e-06, "loss": 0.9749, "step": 340 }, { "grad_norm": 2.043039560317993, "learning_rate": 8.725e-06, "loss": 0.9378, "step": 350 }, { "grad_norm": 1.6409122943878174, "learning_rate": 8.975e-06, "loss": 0.8952, "step": 360 }, { "grad_norm": 2.0171873569488525, "learning_rate": 9.225e-06, "loss": 0.9053, "step": 370 }, { "grad_norm": 2.136888265609741, "learning_rate": 9.475e-06, "loss": 0.8617, "step": 380 }, { "grad_norm": 2.130859851837158, "learning_rate": 9.725000000000001e-06, "loss": 0.8403, "step": 390 }, { "grad_norm": 2.319349527359009, "learning_rate": 9.975e-06, "loss": 0.8064, "step": 400 }, { "grad_norm": 1.9087624549865723, "learning_rate": 1.0225e-05, "loss": 0.7846, "step": 410 }, { "grad_norm": 2.4495067596435547, "learning_rate": 1.0475e-05, "loss": 0.7572, "step": 420 }, { "grad_norm": 2.5376803874969482, "learning_rate": 1.0725e-05, "loss": 0.7383, "step": 430 }, { "grad_norm": 2.5507090091705322, "learning_rate": 1.0975e-05, "loss": 0.6996, "step": 440 }, { "grad_norm": 2.0685322284698486, "learning_rate": 1.1225e-05, "loss": 0.6421, "step": 450 }, { "grad_norm": 2.6918599605560303, "learning_rate": 1.1475000000000001e-05, "loss": 0.6492, "step": 460 }, { "grad_norm": 2.6962332725524902, "learning_rate": 1.1725e-05, "loss": 0.6115, "step": 470 }, { "grad_norm": 2.609485626220703, "learning_rate": 1.1975e-05, "loss": 0.5801, "step": 480 }, { "grad_norm": 2.912489175796509, "learning_rate": 1.2225e-05, "loss": 0.5448, "step": 490 }, { "grad_norm": 3.552344799041748, "learning_rate": 1.2475e-05, "loss": 0.5077, "step": 500 }, { "grad_norm": 2.7159171104431152, "learning_rate": 1.2725000000000001e-05, "loss": 0.5062, "step": 510 }, { "grad_norm": 3.0200300216674805, "learning_rate": 1.2975e-05, "loss": 0.5033, "step": 520 }, { "grad_norm": 3.329540491104126, "learning_rate": 1.3225000000000001e-05, "loss": 0.484, "step": 530 }, { "grad_norm": 2.6327760219573975, "learning_rate": 1.3475000000000002e-05, "loss": 0.4329, "step": 540 }, { "grad_norm": 3.2794344425201416, "learning_rate": 1.3725000000000002e-05, "loss": 0.4525, "step": 550 }, { "grad_norm": 3.022820234298706, "learning_rate": 1.3975000000000003e-05, "loss": 0.4539, "step": 560 }, { "grad_norm": 2.8281803131103516, "learning_rate": 1.4225e-05, "loss": 0.4053, "step": 570 }, { "grad_norm": 2.546422004699707, "learning_rate": 1.4475e-05, "loss": 0.3898, "step": 580 }, { "grad_norm": 2.8810317516326904, "learning_rate": 1.4725e-05, "loss": 0.3714, "step": 590 }, { "grad_norm": 2.8545873165130615, "learning_rate": 1.4975e-05, "loss": 0.3619, "step": 600 }, { "grad_norm": 3.0319111347198486, "learning_rate": 1.5225e-05, "loss": 0.3342, "step": 610 }, { "grad_norm": 2.834226369857788, "learning_rate": 1.5475e-05, "loss": 0.3234, "step": 620 }, { "grad_norm": 2.561368227005005, "learning_rate": 1.5725e-05, "loss": 0.3099, "step": 630 }, { "grad_norm": 3.991291046142578, "learning_rate": 1.5975000000000002e-05, "loss": 0.2767, "step": 640 }, { "grad_norm": 3.3727052211761475, "learning_rate": 1.6225e-05, "loss": 0.2543, "step": 650 }, { "grad_norm": 3.6452219486236572, "learning_rate": 1.6475e-05, "loss": 0.2749, "step": 660 }, { "grad_norm": 2.226874828338623, "learning_rate": 1.6725000000000003e-05, "loss": 0.2218, "step": 670 }, { "grad_norm": 3.247556447982788, "learning_rate": 1.6975000000000003e-05, "loss": 0.2281, "step": 680 }, { "grad_norm": 3.4518654346466064, "learning_rate": 1.7225e-05, "loss": 0.2321, "step": 690 }, { "grad_norm": 3.7096617221832275, "learning_rate": 1.7475e-05, "loss": 0.2463, "step": 700 }, { "grad_norm": 2.5420761108398438, "learning_rate": 1.7725e-05, "loss": 0.2067, "step": 710 }, { "grad_norm": 3.68207049369812, "learning_rate": 1.7975e-05, "loss": 0.2314, "step": 720 }, { "grad_norm": 2.539124011993408, "learning_rate": 1.8225e-05, "loss": 0.2122, "step": 730 }, { "grad_norm": 1.9244205951690674, "learning_rate": 1.8475000000000002e-05, "loss": 0.1693, "step": 740 }, { "grad_norm": 1.8760128021240234, "learning_rate": 1.8725e-05, "loss": 0.1922, "step": 750 }, { "grad_norm": 2.021172046661377, "learning_rate": 1.8975e-05, "loss": 0.1874, "step": 760 }, { "grad_norm": 2.423238515853882, "learning_rate": 1.9225e-05, "loss": 0.1965, "step": 770 }, { "grad_norm": 2.385610580444336, "learning_rate": 1.9475000000000002e-05, "loss": 0.2303, "step": 780 }, { "grad_norm": 1.8096500635147095, "learning_rate": 1.9725000000000002e-05, "loss": 0.1679, "step": 790 }, { "grad_norm": 3.0623903274536133, "learning_rate": 1.9975e-05, "loss": 0.1734, "step": 800 }, { "grad_norm": 2.449864625930786, "learning_rate": 2.0225000000000004e-05, "loss": 0.1907, "step": 810 }, { "grad_norm": 2.1408116817474365, "learning_rate": 2.0475e-05, "loss": 0.1872, "step": 820 }, { "grad_norm": 2.5102152824401855, "learning_rate": 2.0725e-05, "loss": 0.2015, "step": 830 }, { "grad_norm": 1.5780187845230103, "learning_rate": 2.0975e-05, "loss": 0.1771, "step": 840 }, { "grad_norm": 2.1477725505828857, "learning_rate": 2.1225e-05, "loss": 0.1875, "step": 850 }, { "grad_norm": 2.2428781986236572, "learning_rate": 2.1475e-05, "loss": 0.1733, "step": 860 }, { "grad_norm": 2.2376439571380615, "learning_rate": 2.1725e-05, "loss": 0.1988, "step": 870 }, { "grad_norm": 2.6404902935028076, "learning_rate": 2.1975000000000002e-05, "loss": 0.1664, "step": 880 }, { "grad_norm": 2.1725926399230957, "learning_rate": 2.2225e-05, "loss": 0.1763, "step": 890 }, { "grad_norm": 2.062162160873413, "learning_rate": 2.2475e-05, "loss": 0.1779, "step": 900 }, { "grad_norm": 1.9669569730758667, "learning_rate": 2.2725000000000003e-05, "loss": 0.1683, "step": 910 }, { "grad_norm": 1.773060917854309, "learning_rate": 2.2975000000000003e-05, "loss": 0.1483, "step": 920 }, { "grad_norm": 2.4532370567321777, "learning_rate": 2.3225000000000002e-05, "loss": 0.1585, "step": 930 }, { "grad_norm": 1.9468446969985962, "learning_rate": 2.3475e-05, "loss": 0.1683, "step": 940 }, { "grad_norm": 2.307331085205078, "learning_rate": 2.3725e-05, "loss": 0.1481, "step": 950 }, { "grad_norm": 1.7834405899047852, "learning_rate": 2.3975e-05, "loss": 0.1569, "step": 960 }, { "grad_norm": 2.166104555130005, "learning_rate": 2.4225e-05, "loss": 0.1795, "step": 970 }, { "grad_norm": 2.2634499073028564, "learning_rate": 2.4475000000000002e-05, "loss": 0.1385, "step": 980 }, { "grad_norm": 2.1755588054656982, "learning_rate": 2.4725e-05, "loss": 0.1396, "step": 990 }, { "grad_norm": 1.646173119544983, "learning_rate": 2.4975e-05, "loss": 0.1378, "step": 1000 }, { "grad_norm": 2.145331859588623, "learning_rate": 2.5225e-05, "loss": 0.1369, "step": 1010 }, { "grad_norm": 2.3671674728393555, "learning_rate": 2.5475e-05, "loss": 0.1682, "step": 1020 }, { "grad_norm": 1.9531300067901611, "learning_rate": 2.5725e-05, "loss": 0.1461, "step": 1030 }, { "grad_norm": 2.27108097076416, "learning_rate": 2.5974999999999998e-05, "loss": 0.1488, "step": 1040 }, { "grad_norm": 2.0536067485809326, "learning_rate": 2.6225e-05, "loss": 0.1675, "step": 1050 }, { "grad_norm": 1.8817025423049927, "learning_rate": 2.6475e-05, "loss": 0.1741, "step": 1060 }, { "grad_norm": 2.0515267848968506, "learning_rate": 2.6725e-05, "loss": 0.179, "step": 1070 }, { "grad_norm": 2.2118711471557617, "learning_rate": 2.6975000000000002e-05, "loss": 0.1828, "step": 1080 }, { "grad_norm": 1.8559057712554932, "learning_rate": 2.7225e-05, "loss": 0.153, "step": 1090 }, { "grad_norm": 2.3655967712402344, "learning_rate": 2.7475e-05, "loss": 0.1334, "step": 1100 }, { "grad_norm": 2.276686906814575, "learning_rate": 2.7725e-05, "loss": 0.1786, "step": 1110 }, { "grad_norm": 1.726273775100708, "learning_rate": 2.7975000000000002e-05, "loss": 0.1338, "step": 1120 }, { "grad_norm": 1.9896104335784912, "learning_rate": 2.8225e-05, "loss": 0.1608, "step": 1130 }, { "grad_norm": 2.365017890930176, "learning_rate": 2.8475e-05, "loss": 0.1812, "step": 1140 }, { "grad_norm": 1.846018671989441, "learning_rate": 2.8725e-05, "loss": 0.1731, "step": 1150 }, { "grad_norm": 1.5463918447494507, "learning_rate": 2.8975000000000003e-05, "loss": 0.1503, "step": 1160 }, { "grad_norm": 1.473053216934204, "learning_rate": 2.9225000000000002e-05, "loss": 0.1547, "step": 1170 }, { "grad_norm": 1.8464571237564087, "learning_rate": 2.9475e-05, "loss": 0.1357, "step": 1180 }, { "grad_norm": 1.782782793045044, "learning_rate": 2.9725000000000004e-05, "loss": 0.1498, "step": 1190 }, { "grad_norm": 2.8830974102020264, "learning_rate": 2.9975000000000004e-05, "loss": 0.1545, "step": 1200 }, { "grad_norm": 3.099349021911621, "learning_rate": 3.0225000000000003e-05, "loss": 0.1708, "step": 1210 }, { "grad_norm": 1.7821590900421143, "learning_rate": 3.0475000000000002e-05, "loss": 0.1436, "step": 1220 }, { "grad_norm": 2.111889362335205, "learning_rate": 3.0725e-05, "loss": 0.1597, "step": 1230 }, { "grad_norm": 1.8504425287246704, "learning_rate": 3.0975e-05, "loss": 0.164, "step": 1240 }, { "grad_norm": 1.7522603273391724, "learning_rate": 3.122500000000001e-05, "loss": 0.138, "step": 1250 }, { "grad_norm": 2.072736978530884, "learning_rate": 3.1475e-05, "loss": 0.1478, "step": 1260 }, { "grad_norm": 1.3050060272216797, "learning_rate": 3.1725e-05, "loss": 0.1267, "step": 1270 }, { "grad_norm": 1.2972536087036133, "learning_rate": 3.1975e-05, "loss": 0.1539, "step": 1280 }, { "grad_norm": 1.4086273908615112, "learning_rate": 3.2225e-05, "loss": 0.1421, "step": 1290 }, { "grad_norm": 1.8989828824996948, "learning_rate": 3.2474999999999997e-05, "loss": 0.1437, "step": 1300 }, { "grad_norm": 1.6552454233169556, "learning_rate": 3.2725e-05, "loss": 0.1371, "step": 1310 }, { "grad_norm": 1.7883408069610596, "learning_rate": 3.2975e-05, "loss": 0.1538, "step": 1320 }, { "grad_norm": 1.7611762285232544, "learning_rate": 3.3225e-05, "loss": 0.1205, "step": 1330 }, { "grad_norm": 1.4610449075698853, "learning_rate": 3.3475e-05, "loss": 0.1181, "step": 1340 }, { "grad_norm": 2.0274479389190674, "learning_rate": 3.3725e-05, "loss": 0.1589, "step": 1350 }, { "grad_norm": 2.25285005569458, "learning_rate": 3.3975e-05, "loss": 0.1503, "step": 1360 }, { "grad_norm": 1.4241001605987549, "learning_rate": 3.4225e-05, "loss": 0.1251, "step": 1370 }, { "grad_norm": 1.7866268157958984, "learning_rate": 3.4475000000000005e-05, "loss": 0.1203, "step": 1380 }, { "grad_norm": 1.5070596933364868, "learning_rate": 3.4725000000000004e-05, "loss": 0.1386, "step": 1390 }, { "grad_norm": 1.8578696250915527, "learning_rate": 3.4975e-05, "loss": 0.1187, "step": 1400 }, { "grad_norm": 1.0750019550323486, "learning_rate": 3.5225e-05, "loss": 0.1328, "step": 1410 }, { "grad_norm": 1.9186291694641113, "learning_rate": 3.5475e-05, "loss": 0.138, "step": 1420 }, { "grad_norm": 1.4942432641983032, "learning_rate": 3.5725e-05, "loss": 0.1225, "step": 1430 }, { "grad_norm": 1.4335705041885376, "learning_rate": 3.5975e-05, "loss": 0.1391, "step": 1440 }, { "grad_norm": 1.3445247411727905, "learning_rate": 3.6225000000000006e-05, "loss": 0.1498, "step": 1450 }, { "grad_norm": 1.356865644454956, "learning_rate": 3.6475000000000006e-05, "loss": 0.1404, "step": 1460 }, { "grad_norm": 1.2415904998779297, "learning_rate": 3.6725000000000005e-05, "loss": 0.1422, "step": 1470 }, { "grad_norm": 1.5398808717727661, "learning_rate": 3.6975000000000004e-05, "loss": 0.1204, "step": 1480 }, { "grad_norm": 0.9805710911750793, "learning_rate": 3.7225000000000004e-05, "loss": 0.1158, "step": 1490 }, { "grad_norm": 0.9968165159225464, "learning_rate": 3.7475e-05, "loss": 0.1373, "step": 1500 }, { "grad_norm": 1.4018433094024658, "learning_rate": 3.7725e-05, "loss": 0.1305, "step": 1510 }, { "grad_norm": 1.6201788187026978, "learning_rate": 3.7975e-05, "loss": 0.1349, "step": 1520 }, { "grad_norm": 1.224007248878479, "learning_rate": 3.8225e-05, "loss": 0.1329, "step": 1530 }, { "grad_norm": 1.5622652769088745, "learning_rate": 3.8475e-05, "loss": 0.1217, "step": 1540 }, { "grad_norm": 1.3765416145324707, "learning_rate": 3.8725e-05, "loss": 0.1382, "step": 1550 }, { "grad_norm": 1.1185956001281738, "learning_rate": 3.8975e-05, "loss": 0.1539, "step": 1560 }, { "grad_norm": 1.3464394807815552, "learning_rate": 3.9225e-05, "loss": 0.1139, "step": 1570 }, { "grad_norm": 1.298974871635437, "learning_rate": 3.9475000000000004e-05, "loss": 0.1301, "step": 1580 }, { "grad_norm": 1.2972770929336548, "learning_rate": 3.9725e-05, "loss": 0.1361, "step": 1590 }, { "grad_norm": 1.173863410949707, "learning_rate": 3.9975e-05, "loss": 0.1263, "step": 1600 }, { "grad_norm": 1.2208551168441772, "learning_rate": 4.0225e-05, "loss": 0.1103, "step": 1610 }, { "grad_norm": 1.0531644821166992, "learning_rate": 4.0475e-05, "loss": 0.1235, "step": 1620 }, { "grad_norm": 1.6145387887954712, "learning_rate": 4.0725e-05, "loss": 0.1623, "step": 1630 }, { "grad_norm": 1.279681921005249, "learning_rate": 4.0975e-05, "loss": 0.1332, "step": 1640 }, { "grad_norm": 1.2884405851364136, "learning_rate": 4.1225e-05, "loss": 0.128, "step": 1650 }, { "grad_norm": 1.9087835550308228, "learning_rate": 4.1475000000000005e-05, "loss": 0.1402, "step": 1660 }, { "grad_norm": 1.0130082368850708, "learning_rate": 4.1725000000000005e-05, "loss": 0.1145, "step": 1670 }, { "grad_norm": 1.3137768507003784, "learning_rate": 4.1975000000000004e-05, "loss": 0.1238, "step": 1680 }, { "grad_norm": 1.232548713684082, "learning_rate": 4.2225e-05, "loss": 0.1286, "step": 1690 }, { "grad_norm": 1.1484763622283936, "learning_rate": 4.2475e-05, "loss": 0.1335, "step": 1700 }, { "grad_norm": 1.7057782411575317, "learning_rate": 4.2725e-05, "loss": 0.1488, "step": 1710 }, { "grad_norm": 1.213960886001587, "learning_rate": 4.2975e-05, "loss": 0.1307, "step": 1720 }, { "grad_norm": 1.9664653539657593, "learning_rate": 4.322500000000001e-05, "loss": 0.1357, "step": 1730 }, { "grad_norm": 1.126333236694336, "learning_rate": 4.3475000000000006e-05, "loss": 0.1283, "step": 1740 }, { "grad_norm": 1.1565032005310059, "learning_rate": 4.3725000000000006e-05, "loss": 0.1241, "step": 1750 }, { "grad_norm": 1.663559913635254, "learning_rate": 4.3975e-05, "loss": 0.1301, "step": 1760 }, { "grad_norm": 1.26711106300354, "learning_rate": 4.4225e-05, "loss": 0.1189, "step": 1770 }, { "grad_norm": 0.7977402806282043, "learning_rate": 4.4475e-05, "loss": 0.1397, "step": 1780 }, { "grad_norm": 1.537605881690979, "learning_rate": 4.4725e-05, "loss": 0.1173, "step": 1790 }, { "grad_norm": 1.3866472244262695, "learning_rate": 4.4975e-05, "loss": 0.124, "step": 1800 }, { "grad_norm": 1.0080815553665161, "learning_rate": 4.5225e-05, "loss": 0.122, "step": 1810 }, { "grad_norm": 1.582551121711731, "learning_rate": 4.5475e-05, "loss": 0.137, "step": 1820 }, { "grad_norm": 1.613423228263855, "learning_rate": 4.5725e-05, "loss": 0.1409, "step": 1830 }, { "grad_norm": 1.205538034439087, "learning_rate": 4.5975e-05, "loss": 0.1392, "step": 1840 }, { "grad_norm": 0.9034583568572998, "learning_rate": 4.6225e-05, "loss": 0.153, "step": 1850 }, { "grad_norm": 1.082283854484558, "learning_rate": 4.6475000000000005e-05, "loss": 0.1458, "step": 1860 }, { "grad_norm": 0.9472025036811829, "learning_rate": 4.6725000000000004e-05, "loss": 0.1237, "step": 1870 }, { "grad_norm": 0.9682486057281494, "learning_rate": 4.6975000000000003e-05, "loss": 0.0993, "step": 1880 }, { "grad_norm": 1.13823664188385, "learning_rate": 4.7225e-05, "loss": 0.1208, "step": 1890 }, { "grad_norm": 1.181975245475769, "learning_rate": 4.7475e-05, "loss": 0.1109, "step": 1900 }, { "grad_norm": 1.3218460083007812, "learning_rate": 4.7725e-05, "loss": 0.1467, "step": 1910 }, { "grad_norm": 0.9129258394241333, "learning_rate": 4.7975e-05, "loss": 0.12, "step": 1920 }, { "grad_norm": 1.154158592224121, "learning_rate": 4.822500000000001e-05, "loss": 0.1059, "step": 1930 }, { "grad_norm": 1.2182955741882324, "learning_rate": 4.8475000000000006e-05, "loss": 0.1195, "step": 1940 }, { "grad_norm": 0.9422802329063416, "learning_rate": 4.8725000000000005e-05, "loss": 0.1057, "step": 1950 }, { "grad_norm": 1.3931019306182861, "learning_rate": 4.8975000000000005e-05, "loss": 0.1298, "step": 1960 }, { "grad_norm": 1.1495671272277832, "learning_rate": 4.9225000000000004e-05, "loss": 0.1053, "step": 1970 }, { "grad_norm": 1.6453421115875244, "learning_rate": 4.9475e-05, "loss": 0.1316, "step": 1980 }, { "grad_norm": 1.1271600723266602, "learning_rate": 4.9725e-05, "loss": 0.1069, "step": 1990 }, { "grad_norm": 0.8730885982513428, "learning_rate": 4.9975e-05, "loss": 0.1111, "step": 2000 }, { "grad_norm": 1.2023570537567139, "learning_rate": 5.0225e-05, "loss": 0.1565, "step": 2010 }, { "grad_norm": 1.1006542444229126, "learning_rate": 5.047500000000001e-05, "loss": 0.1171, "step": 2020 }, { "grad_norm": 1.220909595489502, "learning_rate": 5.0725e-05, "loss": 0.1092, "step": 2030 }, { "grad_norm": 1.0662195682525635, "learning_rate": 5.0975000000000006e-05, "loss": 0.1258, "step": 2040 }, { "grad_norm": 1.1347318887710571, "learning_rate": 5.1225e-05, "loss": 0.121, "step": 2050 }, { "grad_norm": 1.2810487747192383, "learning_rate": 5.1475000000000004e-05, "loss": 0.1113, "step": 2060 }, { "grad_norm": 1.0043649673461914, "learning_rate": 5.1725000000000004e-05, "loss": 0.1114, "step": 2070 }, { "grad_norm": 1.2492817640304565, "learning_rate": 5.197500000000001e-05, "loss": 0.1194, "step": 2080 }, { "grad_norm": 0.8887073993682861, "learning_rate": 5.2225e-05, "loss": 0.1143, "step": 2090 }, { "grad_norm": 1.0805801153182983, "learning_rate": 5.247500000000001e-05, "loss": 0.1122, "step": 2100 }, { "grad_norm": 0.9702354073524475, "learning_rate": 5.2725e-05, "loss": 0.1328, "step": 2110 }, { "grad_norm": 0.9734779000282288, "learning_rate": 5.297500000000001e-05, "loss": 0.1125, "step": 2120 }, { "grad_norm": 1.0616555213928223, "learning_rate": 5.3225e-05, "loss": 0.1162, "step": 2130 }, { "grad_norm": 0.8958854675292969, "learning_rate": 5.3475e-05, "loss": 0.1052, "step": 2140 }, { "grad_norm": 0.9093354344367981, "learning_rate": 5.3725000000000005e-05, "loss": 0.121, "step": 2150 }, { "grad_norm": 0.9452520608901978, "learning_rate": 5.3975e-05, "loss": 0.1037, "step": 2160 }, { "grad_norm": 0.920436680316925, "learning_rate": 5.4225000000000003e-05, "loss": 0.1046, "step": 2170 }, { "grad_norm": 0.9094617962837219, "learning_rate": 5.4474999999999996e-05, "loss": 0.1331, "step": 2180 }, { "grad_norm": 0.9136719107627869, "learning_rate": 5.4725e-05, "loss": 0.1145, "step": 2190 }, { "grad_norm": 0.954717755317688, "learning_rate": 5.4975e-05, "loss": 0.1057, "step": 2200 }, { "grad_norm": 0.9337154030799866, "learning_rate": 5.522500000000001e-05, "loss": 0.1041, "step": 2210 }, { "grad_norm": 1.16348397731781, "learning_rate": 5.5475e-05, "loss": 0.1094, "step": 2220 }, { "grad_norm": 0.9119200110435486, "learning_rate": 5.5725000000000006e-05, "loss": 0.1046, "step": 2230 }, { "grad_norm": 0.8683666586875916, "learning_rate": 5.5975e-05, "loss": 0.0995, "step": 2240 }, { "grad_norm": 1.027845859527588, "learning_rate": 5.6225000000000005e-05, "loss": 0.1247, "step": 2250 }, { "grad_norm": 0.8807549476623535, "learning_rate": 5.6475e-05, "loss": 0.1031, "step": 2260 }, { "grad_norm": 0.9062234163284302, "learning_rate": 5.6725e-05, "loss": 0.102, "step": 2270 }, { "grad_norm": 1.0883545875549316, "learning_rate": 5.6975e-05, "loss": 0.0993, "step": 2280 }, { "grad_norm": 1.167423963546753, "learning_rate": 5.722500000000001e-05, "loss": 0.1452, "step": 2290 }, { "grad_norm": 1.275164008140564, "learning_rate": 5.7475e-05, "loss": 0.1116, "step": 2300 }, { "grad_norm": 0.9930912256240845, "learning_rate": 5.772500000000001e-05, "loss": 0.1011, "step": 2310 }, { "grad_norm": 0.9516489505767822, "learning_rate": 5.7975e-05, "loss": 0.0912, "step": 2320 }, { "grad_norm": 0.8570199608802795, "learning_rate": 5.8225000000000006e-05, "loss": 0.1353, "step": 2330 }, { "grad_norm": 1.015928864479065, "learning_rate": 5.8475000000000005e-05, "loss": 0.1117, "step": 2340 }, { "grad_norm": 0.8145533204078674, "learning_rate": 5.8725000000000004e-05, "loss": 0.1059, "step": 2350 }, { "grad_norm": 0.989684522151947, "learning_rate": 5.8975000000000004e-05, "loss": 0.1063, "step": 2360 }, { "grad_norm": 0.8462041616439819, "learning_rate": 5.922500000000001e-05, "loss": 0.1101, "step": 2370 }, { "grad_norm": 0.9715782999992371, "learning_rate": 5.9475e-05, "loss": 0.1221, "step": 2380 }, { "grad_norm": 0.787759006023407, "learning_rate": 5.9724999999999995e-05, "loss": 0.1261, "step": 2390 }, { "grad_norm": 0.8051850199699402, "learning_rate": 5.9975e-05, "loss": 0.1102, "step": 2400 }, { "grad_norm": 0.7450733780860901, "learning_rate": 6.0225e-05, "loss": 0.1136, "step": 2410 }, { "grad_norm": 0.9344789981842041, "learning_rate": 6.0475000000000006e-05, "loss": 0.1524, "step": 2420 }, { "grad_norm": 0.8917458057403564, "learning_rate": 6.0725e-05, "loss": 0.118, "step": 2430 }, { "grad_norm": 1.3216769695281982, "learning_rate": 6.0975000000000005e-05, "loss": 0.1267, "step": 2440 }, { "grad_norm": 0.8598557114601135, "learning_rate": 6.1225e-05, "loss": 0.1117, "step": 2450 }, { "grad_norm": 1.091797947883606, "learning_rate": 6.1475e-05, "loss": 0.1041, "step": 2460 }, { "grad_norm": 0.9134456515312195, "learning_rate": 6.1725e-05, "loss": 0.1048, "step": 2470 }, { "grad_norm": 0.9386423826217651, "learning_rate": 6.1975e-05, "loss": 0.1241, "step": 2480 }, { "grad_norm": 0.8325107097625732, "learning_rate": 6.2225e-05, "loss": 0.1014, "step": 2490 }, { "grad_norm": 0.7299706339836121, "learning_rate": 6.2475e-05, "loss": 0.0875, "step": 2500 }, { "grad_norm": 1.0623804330825806, "learning_rate": 6.2725e-05, "loss": 0.0964, "step": 2510 }, { "grad_norm": 0.7739264965057373, "learning_rate": 6.297500000000001e-05, "loss": 0.1012, "step": 2520 }, { "grad_norm": 0.7497016787528992, "learning_rate": 6.3225e-05, "loss": 0.0899, "step": 2530 }, { "grad_norm": 0.8457396626472473, "learning_rate": 6.347500000000001e-05, "loss": 0.0994, "step": 2540 }, { "grad_norm": 0.8682423233985901, "learning_rate": 6.3725e-05, "loss": 0.1065, "step": 2550 }, { "grad_norm": 0.957253634929657, "learning_rate": 6.397500000000001e-05, "loss": 0.1148, "step": 2560 }, { "grad_norm": 1.0101720094680786, "learning_rate": 6.4225e-05, "loss": 0.1005, "step": 2570 }, { "grad_norm": 0.8969477415084839, "learning_rate": 6.447500000000001e-05, "loss": 0.1009, "step": 2580 }, { "grad_norm": 0.6215335726737976, "learning_rate": 6.4725e-05, "loss": 0.0924, "step": 2590 }, { "grad_norm": 0.8298277258872986, "learning_rate": 6.497500000000001e-05, "loss": 0.1121, "step": 2600 }, { "grad_norm": 0.8357805609703064, "learning_rate": 6.5225e-05, "loss": 0.1053, "step": 2610 }, { "grad_norm": 0.9589259028434753, "learning_rate": 6.5475e-05, "loss": 0.085, "step": 2620 }, { "grad_norm": 0.949306309223175, "learning_rate": 6.5725e-05, "loss": 0.095, "step": 2630 }, { "grad_norm": 0.902950644493103, "learning_rate": 6.5975e-05, "loss": 0.1133, "step": 2640 }, { "grad_norm": 0.7899484634399414, "learning_rate": 6.6225e-05, "loss": 0.1189, "step": 2650 }, { "grad_norm": 0.8533879518508911, "learning_rate": 6.6475e-05, "loss": 0.115, "step": 2660 }, { "grad_norm": 0.7568981051445007, "learning_rate": 6.672500000000001e-05, "loss": 0.1146, "step": 2670 }, { "grad_norm": 0.8769376873970032, "learning_rate": 6.6975e-05, "loss": 0.1334, "step": 2680 }, { "grad_norm": 1.0533318519592285, "learning_rate": 6.722500000000001e-05, "loss": 0.1118, "step": 2690 }, { "grad_norm": 0.7932769060134888, "learning_rate": 6.7475e-05, "loss": 0.096, "step": 2700 }, { "grad_norm": 0.986504077911377, "learning_rate": 6.7725e-05, "loss": 0.1081, "step": 2710 }, { "grad_norm": 1.0495063066482544, "learning_rate": 6.7975e-05, "loss": 0.1285, "step": 2720 }, { "grad_norm": 0.9881197810173035, "learning_rate": 6.8225e-05, "loss": 0.1158, "step": 2730 }, { "grad_norm": 0.7855017185211182, "learning_rate": 6.8475e-05, "loss": 0.1081, "step": 2740 }, { "grad_norm": 0.8344347476959229, "learning_rate": 6.8725e-05, "loss": 0.1232, "step": 2750 }, { "grad_norm": 0.8609263300895691, "learning_rate": 6.8975e-05, "loss": 0.0995, "step": 2760 }, { "grad_norm": 0.687235414981842, "learning_rate": 6.9225e-05, "loss": 0.101, "step": 2770 }, { "grad_norm": 0.9225020408630371, "learning_rate": 6.9475e-05, "loss": 0.1072, "step": 2780 }, { "grad_norm": 0.8237553834915161, "learning_rate": 6.9725e-05, "loss": 0.1026, "step": 2790 }, { "grad_norm": 1.0421725511550903, "learning_rate": 6.997500000000001e-05, "loss": 0.0946, "step": 2800 }, { "grad_norm": 0.8532367944717407, "learning_rate": 7.022500000000001e-05, "loss": 0.1291, "step": 2810 }, { "grad_norm": 1.0447393655776978, "learning_rate": 7.0475e-05, "loss": 0.1076, "step": 2820 }, { "grad_norm": 0.6408739686012268, "learning_rate": 7.072500000000001e-05, "loss": 0.0941, "step": 2830 }, { "grad_norm": 0.6376145482063293, "learning_rate": 7.0975e-05, "loss": 0.1164, "step": 2840 }, { "grad_norm": 0.7246596813201904, "learning_rate": 7.122500000000001e-05, "loss": 0.1085, "step": 2850 }, { "grad_norm": 0.7269245982170105, "learning_rate": 7.1475e-05, "loss": 0.1023, "step": 2860 }, { "grad_norm": 0.6763855814933777, "learning_rate": 7.172500000000001e-05, "loss": 0.0884, "step": 2870 }, { "grad_norm": 0.7523444890975952, "learning_rate": 7.1975e-05, "loss": 0.0936, "step": 2880 }, { "grad_norm": 0.7801661491394043, "learning_rate": 7.2225e-05, "loss": 0.0999, "step": 2890 }, { "grad_norm": 0.6233508586883545, "learning_rate": 7.2475e-05, "loss": 0.102, "step": 2900 }, { "grad_norm": 0.8353660702705383, "learning_rate": 7.272499999999999e-05, "loss": 0.0838, "step": 2910 }, { "grad_norm": 0.7469881176948547, "learning_rate": 7.2975e-05, "loss": 0.0966, "step": 2920 }, { "grad_norm": 0.696208655834198, "learning_rate": 7.3225e-05, "loss": 0.0841, "step": 2930 }, { "grad_norm": 0.7753432989120483, "learning_rate": 7.347500000000001e-05, "loss": 0.0872, "step": 2940 }, { "grad_norm": 0.7968539595603943, "learning_rate": 7.3725e-05, "loss": 0.084, "step": 2950 }, { "grad_norm": 0.7662535309791565, "learning_rate": 7.397500000000001e-05, "loss": 0.0953, "step": 2960 }, { "grad_norm": 0.926124095916748, "learning_rate": 7.4225e-05, "loss": 0.089, "step": 2970 }, { "grad_norm": 0.739693284034729, "learning_rate": 7.447500000000001e-05, "loss": 0.0912, "step": 2980 }, { "grad_norm": 0.9120153188705444, "learning_rate": 7.4725e-05, "loss": 0.1163, "step": 2990 }, { "grad_norm": 0.7063814401626587, "learning_rate": 7.4975e-05, "loss": 0.1001, "step": 3000 }, { "grad_norm": 0.7749399542808533, "learning_rate": 7.5225e-05, "loss": 0.0803, "step": 3010 }, { "grad_norm": 0.5911969542503357, "learning_rate": 7.5475e-05, "loss": 0.0846, "step": 3020 }, { "grad_norm": 0.6140455007553101, "learning_rate": 7.5725e-05, "loss": 0.089, "step": 3030 }, { "grad_norm": 0.7764491438865662, "learning_rate": 7.5975e-05, "loss": 0.0958, "step": 3040 }, { "grad_norm": 0.8187344670295715, "learning_rate": 7.6225e-05, "loss": 0.0922, "step": 3050 }, { "grad_norm": 0.6083786487579346, "learning_rate": 7.6475e-05, "loss": 0.106, "step": 3060 }, { "grad_norm": 0.5438709855079651, "learning_rate": 7.672500000000001e-05, "loss": 0.0724, "step": 3070 }, { "grad_norm": 0.9673739075660706, "learning_rate": 7.697500000000001e-05, "loss": 0.0925, "step": 3080 }, { "grad_norm": 1.1821873188018799, "learning_rate": 7.722500000000001e-05, "loss": 0.1172, "step": 3090 }, { "grad_norm": 0.7934598326683044, "learning_rate": 7.747500000000001e-05, "loss": 0.095, "step": 3100 }, { "grad_norm": 0.7798365354537964, "learning_rate": 7.7725e-05, "loss": 0.0833, "step": 3110 }, { "grad_norm": 0.6985222101211548, "learning_rate": 7.797500000000001e-05, "loss": 0.0886, "step": 3120 }, { "grad_norm": 0.8220073580741882, "learning_rate": 7.8225e-05, "loss": 0.0872, "step": 3130 }, { "grad_norm": 0.7756733894348145, "learning_rate": 7.8475e-05, "loss": 0.0936, "step": 3140 }, { "grad_norm": 0.6070908308029175, "learning_rate": 7.8725e-05, "loss": 0.0875, "step": 3150 }, { "grad_norm": 0.6493401527404785, "learning_rate": 7.8975e-05, "loss": 0.1059, "step": 3160 }, { "grad_norm": 0.8217419385910034, "learning_rate": 7.9225e-05, "loss": 0.092, "step": 3170 }, { "grad_norm": 0.6711463332176208, "learning_rate": 7.9475e-05, "loss": 0.1029, "step": 3180 }, { "grad_norm": 0.7399564981460571, "learning_rate": 7.9725e-05, "loss": 0.1006, "step": 3190 }, { "grad_norm": 0.7110775709152222, "learning_rate": 7.9975e-05, "loss": 0.086, "step": 3200 }, { "grad_norm": 0.6737504005432129, "learning_rate": 8.022500000000001e-05, "loss": 0.0907, "step": 3210 }, { "grad_norm": 0.8783057332038879, "learning_rate": 8.0475e-05, "loss": 0.1184, "step": 3220 }, { "grad_norm": 1.0593303442001343, "learning_rate": 8.072500000000001e-05, "loss": 0.1108, "step": 3230 }, { "grad_norm": 0.66653972864151, "learning_rate": 8.0975e-05, "loss": 0.0905, "step": 3240 }, { "grad_norm": 0.5497848391532898, "learning_rate": 8.122500000000001e-05, "loss": 0.0981, "step": 3250 }, { "grad_norm": 0.7563647627830505, "learning_rate": 8.1475e-05, "loss": 0.0965, "step": 3260 }, { "grad_norm": 0.937836229801178, "learning_rate": 8.172500000000001e-05, "loss": 0.0929, "step": 3270 }, { "grad_norm": 0.7271844148635864, "learning_rate": 8.1975e-05, "loss": 0.0862, "step": 3280 }, { "grad_norm": 0.5945541858673096, "learning_rate": 8.2225e-05, "loss": 0.0864, "step": 3290 }, { "grad_norm": 0.6838584542274475, "learning_rate": 8.2475e-05, "loss": 0.0817, "step": 3300 }, { "grad_norm": 0.5694160461425781, "learning_rate": 8.2725e-05, "loss": 0.0925, "step": 3310 }, { "grad_norm": 0.6169196367263794, "learning_rate": 8.2975e-05, "loss": 0.1093, "step": 3320 }, { "grad_norm": 0.6839786171913147, "learning_rate": 8.3225e-05, "loss": 0.0899, "step": 3330 }, { "grad_norm": 0.8115282654762268, "learning_rate": 8.347500000000001e-05, "loss": 0.0916, "step": 3340 }, { "grad_norm": 0.6347768902778625, "learning_rate": 8.3725e-05, "loss": 0.0887, "step": 3350 }, { "grad_norm": 0.8171128630638123, "learning_rate": 8.397500000000001e-05, "loss": 0.0869, "step": 3360 }, { "grad_norm": 0.695966362953186, "learning_rate": 8.422500000000001e-05, "loss": 0.0912, "step": 3370 }, { "grad_norm": 0.5397314429283142, "learning_rate": 8.447500000000001e-05, "loss": 0.13, "step": 3380 }, { "grad_norm": 0.7052793502807617, "learning_rate": 8.4725e-05, "loss": 0.0873, "step": 3390 }, { "grad_norm": 0.6719943284988403, "learning_rate": 8.4975e-05, "loss": 0.0934, "step": 3400 }, { "grad_norm": 0.7729360461235046, "learning_rate": 8.5225e-05, "loss": 0.0963, "step": 3410 }, { "grad_norm": 0.6340463161468506, "learning_rate": 8.5475e-05, "loss": 0.0797, "step": 3420 }, { "grad_norm": 0.7421686053276062, "learning_rate": 8.5725e-05, "loss": 0.0881, "step": 3430 }, { "grad_norm": 0.6552782654762268, "learning_rate": 8.5975e-05, "loss": 0.1018, "step": 3440 }, { "grad_norm": 0.7354780435562134, "learning_rate": 8.6225e-05, "loss": 0.0869, "step": 3450 }, { "grad_norm": 0.7000070810317993, "learning_rate": 8.6475e-05, "loss": 0.0795, "step": 3460 }, { "grad_norm": 0.7293923497200012, "learning_rate": 8.672500000000001e-05, "loss": 0.096, "step": 3470 }, { "grad_norm": 0.6346569657325745, "learning_rate": 8.6975e-05, "loss": 0.0945, "step": 3480 }, { "grad_norm": 0.7220290303230286, "learning_rate": 8.7225e-05, "loss": 0.0907, "step": 3490 }, { "grad_norm": 0.5624504685401917, "learning_rate": 8.747500000000001e-05, "loss": 0.077, "step": 3500 }, { "grad_norm": 0.680640459060669, "learning_rate": 8.7725e-05, "loss": 0.0883, "step": 3510 }, { "grad_norm": 0.8767821788787842, "learning_rate": 8.797500000000001e-05, "loss": 0.0854, "step": 3520 }, { "grad_norm": 0.825100839138031, "learning_rate": 8.8225e-05, "loss": 0.092, "step": 3530 }, { "grad_norm": 0.6357647776603699, "learning_rate": 8.847500000000001e-05, "loss": 0.0969, "step": 3540 }, { "grad_norm": 0.684821605682373, "learning_rate": 8.8725e-05, "loss": 0.0893, "step": 3550 }, { "grad_norm": 0.9425753355026245, "learning_rate": 8.897500000000001e-05, "loss": 0.0778, "step": 3560 }, { "grad_norm": 0.5827871561050415, "learning_rate": 8.9225e-05, "loss": 0.0922, "step": 3570 }, { "grad_norm": 0.655940592288971, "learning_rate": 8.9475e-05, "loss": 0.1097, "step": 3580 }, { "grad_norm": 0.8615493774414062, "learning_rate": 8.9725e-05, "loss": 0.1089, "step": 3590 }, { "grad_norm": 0.7456936240196228, "learning_rate": 8.9975e-05, "loss": 0.1002, "step": 3600 }, { "grad_norm": 0.6297682523727417, "learning_rate": 9.0225e-05, "loss": 0.0925, "step": 3610 }, { "grad_norm": 0.7372392416000366, "learning_rate": 9.0475e-05, "loss": 0.0995, "step": 3620 }, { "grad_norm": 0.5858998894691467, "learning_rate": 9.072500000000001e-05, "loss": 0.0892, "step": 3630 }, { "grad_norm": 0.5557467937469482, "learning_rate": 9.0975e-05, "loss": 0.0947, "step": 3640 }, { "grad_norm": 0.7630837559700012, "learning_rate": 9.122500000000001e-05, "loss": 0.0918, "step": 3650 }, { "grad_norm": 0.6135874390602112, "learning_rate": 9.1475e-05, "loss": 0.0902, "step": 3660 }, { "grad_norm": 0.6430949568748474, "learning_rate": 9.172500000000001e-05, "loss": 0.0936, "step": 3670 }, { "grad_norm": 0.5987229347229004, "learning_rate": 9.1975e-05, "loss": 0.0874, "step": 3680 }, { "grad_norm": 0.7609283328056335, "learning_rate": 9.2225e-05, "loss": 0.0945, "step": 3690 }, { "grad_norm": 0.644307017326355, "learning_rate": 9.2475e-05, "loss": 0.0844, "step": 3700 }, { "grad_norm": 0.5809186697006226, "learning_rate": 9.2725e-05, "loss": 0.0864, "step": 3710 }, { "grad_norm": 0.6777516603469849, "learning_rate": 9.2975e-05, "loss": 0.0892, "step": 3720 }, { "grad_norm": 0.7385057210922241, "learning_rate": 9.3225e-05, "loss": 0.0801, "step": 3730 }, { "grad_norm": 0.719260036945343, "learning_rate": 9.3475e-05, "loss": 0.0788, "step": 3740 }, { "grad_norm": 0.6442648768424988, "learning_rate": 9.3725e-05, "loss": 0.0984, "step": 3750 }, { "grad_norm": 0.5414933562278748, "learning_rate": 9.397500000000001e-05, "loss": 0.0867, "step": 3760 }, { "grad_norm": 0.7927324771881104, "learning_rate": 9.422500000000001e-05, "loss": 0.1007, "step": 3770 }, { "grad_norm": 0.7975754141807556, "learning_rate": 9.4475e-05, "loss": 0.1, "step": 3780 }, { "grad_norm": 0.7105162739753723, "learning_rate": 9.472500000000001e-05, "loss": 0.0943, "step": 3790 }, { "grad_norm": 0.6313347816467285, "learning_rate": 9.4975e-05, "loss": 0.1138, "step": 3800 }, { "grad_norm": 0.6835134625434875, "learning_rate": 9.522500000000001e-05, "loss": 0.0847, "step": 3810 }, { "grad_norm": 0.7880350947380066, "learning_rate": 9.5475e-05, "loss": 0.0984, "step": 3820 }, { "grad_norm": 0.6833621859550476, "learning_rate": 9.572500000000001e-05, "loss": 0.0818, "step": 3830 }, { "grad_norm": 0.6672348380088806, "learning_rate": 9.5975e-05, "loss": 0.0788, "step": 3840 }, { "grad_norm": 0.8173919916152954, "learning_rate": 9.622500000000001e-05, "loss": 0.112, "step": 3850 }, { "grad_norm": 0.7613169550895691, "learning_rate": 9.6475e-05, "loss": 0.0896, "step": 3860 }, { "grad_norm": 0.6094372868537903, "learning_rate": 9.6725e-05, "loss": 0.085, "step": 3870 }, { "grad_norm": 0.5990262031555176, "learning_rate": 9.6975e-05, "loss": 0.0839, "step": 3880 }, { "grad_norm": 0.5258170962333679, "learning_rate": 9.7225e-05, "loss": 0.0906, "step": 3890 }, { "grad_norm": 0.7673389315605164, "learning_rate": 9.747500000000001e-05, "loss": 0.0938, "step": 3900 }, { "grad_norm": 0.5469895005226135, "learning_rate": 9.7725e-05, "loss": 0.0942, "step": 3910 }, { "grad_norm": 0.9264959096908569, "learning_rate": 9.797500000000001e-05, "loss": 0.0951, "step": 3920 }, { "grad_norm": 0.7334533929824829, "learning_rate": 9.8225e-05, "loss": 0.0892, "step": 3930 }, { "grad_norm": 0.7770402431488037, "learning_rate": 9.847500000000001e-05, "loss": 0.1074, "step": 3940 }, { "grad_norm": 0.6447131037712097, "learning_rate": 9.8725e-05, "loss": 0.0894, "step": 3950 }, { "grad_norm": 0.7296046614646912, "learning_rate": 9.897500000000001e-05, "loss": 0.1039, "step": 3960 }, { "grad_norm": 0.7497442960739136, "learning_rate": 9.9225e-05, "loss": 0.0912, "step": 3970 }, { "grad_norm": 0.5707175731658936, "learning_rate": 9.9475e-05, "loss": 0.1029, "step": 3980 }, { "grad_norm": 0.6703144907951355, "learning_rate": 9.9725e-05, "loss": 0.0976, "step": 3990 }, { "grad_norm": 0.6869527697563171, "learning_rate": 9.9975e-05, "loss": 0.0908, "step": 4000 }, { "grad_norm": 0.7922571897506714, "learning_rate": 9.999999653982884e-05, "loss": 0.0835, "step": 4010 }, { "grad_norm": 0.4596622586250305, "learning_rate": 9.999998457874392e-05, "loss": 0.0873, "step": 4020 }, { "grad_norm": 0.6612597107887268, "learning_rate": 9.999996407402913e-05, "loss": 0.0925, "step": 4030 }, { "grad_norm": 0.7540565729141235, "learning_rate": 9.999993502568801e-05, "loss": 0.0732, "step": 4040 }, { "grad_norm": 0.7125549912452698, "learning_rate": 9.999989743372548e-05, "loss": 0.09, "step": 4050 }, { "grad_norm": 0.5988572239875793, "learning_rate": 9.999985129814798e-05, "loss": 0.0866, "step": 4060 }, { "grad_norm": 0.5813486576080322, "learning_rate": 9.99997966189634e-05, "loss": 0.0838, "step": 4070 }, { "grad_norm": 0.5938986539840698, "learning_rate": 9.999973339618107e-05, "loss": 0.0894, "step": 4080 }, { "grad_norm": 0.5952826142311096, "learning_rate": 9.999966162981179e-05, "loss": 0.0901, "step": 4090 }, { "grad_norm": 0.694084644317627, "learning_rate": 9.999958131986784e-05, "loss": 0.0791, "step": 4100 }, { "grad_norm": 0.585374653339386, "learning_rate": 9.999949246636293e-05, "loss": 0.0874, "step": 4110 }, { "grad_norm": 0.647630512714386, "learning_rate": 9.999939506931224e-05, "loss": 0.0949, "step": 4120 }, { "grad_norm": 0.7726472020149231, "learning_rate": 9.999928912873243e-05, "loss": 0.0932, "step": 4130 }, { "grad_norm": 0.6887000799179077, "learning_rate": 9.999917464464159e-05, "loss": 0.0929, "step": 4140 }, { "grad_norm": 0.6294090151786804, "learning_rate": 9.999905161705929e-05, "loss": 0.1007, "step": 4150 }, { "grad_norm": 0.6160502433776855, "learning_rate": 9.999892004600653e-05, "loss": 0.0847, "step": 4160 }, { "grad_norm": 0.46545442938804626, "learning_rate": 9.999877993150581e-05, "loss": 0.0887, "step": 4170 }, { "grad_norm": 0.7072447538375854, "learning_rate": 9.999863127358108e-05, "loss": 0.0848, "step": 4180 }, { "grad_norm": 0.87849360704422, "learning_rate": 9.999847407225773e-05, "loss": 0.0961, "step": 4190 }, { "grad_norm": 0.6127546429634094, "learning_rate": 9.999830832756262e-05, "loss": 0.086, "step": 4200 }, { "grad_norm": 0.6082271337509155, "learning_rate": 9.999813403952407e-05, "loss": 0.0917, "step": 4210 }, { "grad_norm": 0.5903285145759583, "learning_rate": 9.999795120817187e-05, "loss": 0.0993, "step": 4220 }, { "grad_norm": 0.7160294651985168, "learning_rate": 9.999775983353725e-05, "loss": 0.1159, "step": 4230 }, { "grad_norm": 0.7454285621643066, "learning_rate": 9.999755991565292e-05, "loss": 0.109, "step": 4240 }, { "grad_norm": 0.5859376192092896, "learning_rate": 9.999735145455303e-05, "loss": 0.1006, "step": 4250 }, { "grad_norm": 0.6352154016494751, "learning_rate": 9.99971344502732e-05, "loss": 0.1062, "step": 4260 }, { "grad_norm": 0.6472386121749878, "learning_rate": 9.999690890285053e-05, "loss": 0.1005, "step": 4270 }, { "grad_norm": 0.7759812474250793, "learning_rate": 9.999667481232356e-05, "loss": 0.0855, "step": 4280 }, { "grad_norm": 0.6223403811454773, "learning_rate": 9.999643217873225e-05, "loss": 0.0945, "step": 4290 }, { "grad_norm": 0.7257487773895264, "learning_rate": 9.999618100211809e-05, "loss": 0.0822, "step": 4300 }, { "grad_norm": 0.7074180841445923, "learning_rate": 9.999592128252402e-05, "loss": 0.0893, "step": 4310 }, { "grad_norm": 0.6273301839828491, "learning_rate": 9.999565301999437e-05, "loss": 0.1094, "step": 4320 }, { "grad_norm": 0.4955883026123047, "learning_rate": 9.999537621457502e-05, "loss": 0.0878, "step": 4330 }, { "grad_norm": 0.7243594527244568, "learning_rate": 9.999509086631323e-05, "loss": 0.0834, "step": 4340 }, { "grad_norm": 0.7856554985046387, "learning_rate": 9.99947969752578e-05, "loss": 0.072, "step": 4350 }, { "grad_norm": 0.624135434627533, "learning_rate": 9.999449454145891e-05, "loss": 0.0759, "step": 4360 }, { "grad_norm": 0.8243559002876282, "learning_rate": 9.999418356496827e-05, "loss": 0.088, "step": 4370 }, { "grad_norm": 0.644287645816803, "learning_rate": 9.999386404583899e-05, "loss": 0.0827, "step": 4380 }, { "grad_norm": 0.5592523217201233, "learning_rate": 9.999353598412568e-05, "loss": 0.0809, "step": 4390 }, { "grad_norm": 0.44779136776924133, "learning_rate": 9.999319937988442e-05, "loss": 0.0935, "step": 4400 }, { "grad_norm": 0.5721815824508667, "learning_rate": 9.999285423317268e-05, "loss": 0.0842, "step": 4410 }, { "grad_norm": 0.5345107316970825, "learning_rate": 9.999250054404947e-05, "loss": 0.0881, "step": 4420 }, { "grad_norm": 0.6738518476486206, "learning_rate": 9.99921383125752e-05, "loss": 0.0839, "step": 4430 }, { "grad_norm": 0.7395676374435425, "learning_rate": 9.99917675388118e-05, "loss": 0.0842, "step": 4440 }, { "grad_norm": 0.6641296148300171, "learning_rate": 9.99913882228226e-05, "loss": 0.1022, "step": 4450 }, { "grad_norm": 0.503906786441803, "learning_rate": 9.999100036467242e-05, "loss": 0.092, "step": 4460 }, { "grad_norm": 0.5745255351066589, "learning_rate": 9.999060396442753e-05, "loss": 0.0852, "step": 4470 }, { "grad_norm": 0.6688005924224854, "learning_rate": 9.999019902215566e-05, "loss": 0.0918, "step": 4480 }, { "grad_norm": 0.5806884765625, "learning_rate": 9.998978553792602e-05, "loss": 0.0893, "step": 4490 }, { "grad_norm": 0.514080822467804, "learning_rate": 9.998936351180926e-05, "loss": 0.0883, "step": 4500 }, { "grad_norm": 0.5436116456985474, "learning_rate": 9.998893294387747e-05, "loss": 0.1089, "step": 4510 }, { "grad_norm": 0.4456198215484619, "learning_rate": 9.998849383420426e-05, "loss": 0.0873, "step": 4520 }, { "grad_norm": 0.5701560974121094, "learning_rate": 9.998804618286465e-05, "loss": 0.0854, "step": 4530 }, { "grad_norm": 0.7784899473190308, "learning_rate": 9.99875899899351e-05, "loss": 0.1036, "step": 4540 }, { "grad_norm": 0.5711070895195007, "learning_rate": 9.99871252554936e-05, "loss": 0.095, "step": 4550 }, { "grad_norm": 0.7075477838516235, "learning_rate": 9.998665197961955e-05, "loss": 0.0918, "step": 4560 }, { "grad_norm": 0.5052989721298218, "learning_rate": 9.998617016239379e-05, "loss": 0.0995, "step": 4570 }, { "grad_norm": 0.7311478853225708, "learning_rate": 9.998567980389869e-05, "loss": 0.0836, "step": 4580 }, { "grad_norm": 0.5660061836242676, "learning_rate": 9.998518090421802e-05, "loss": 0.0806, "step": 4590 }, { "grad_norm": 0.412896990776062, "learning_rate": 9.998467346343703e-05, "loss": 0.0749, "step": 4600 }, { "grad_norm": 0.5979690551757812, "learning_rate": 9.998415748164243e-05, "loss": 0.0858, "step": 4610 }, { "grad_norm": 0.4640246629714966, "learning_rate": 9.998363295892238e-05, "loss": 0.0718, "step": 4620 }, { "grad_norm": 0.7836756706237793, "learning_rate": 9.998309989536652e-05, "loss": 0.0893, "step": 4630 }, { "grad_norm": 0.6672424077987671, "learning_rate": 9.998255829106593e-05, "loss": 0.0806, "step": 4640 }, { "grad_norm": 0.5762547850608826, "learning_rate": 9.998200814611316e-05, "loss": 0.0709, "step": 4650 }, { "grad_norm": 0.613883376121521, "learning_rate": 9.998144946060219e-05, "loss": 0.0818, "step": 4660 }, { "grad_norm": 0.7439338564872742, "learning_rate": 9.998088223462852e-05, "loss": 0.0729, "step": 4670 }, { "grad_norm": 0.6300364136695862, "learning_rate": 9.998030646828905e-05, "loss": 0.0818, "step": 4680 }, { "grad_norm": 0.6701124310493469, "learning_rate": 9.997972216168217e-05, "loss": 0.0866, "step": 4690 }, { "grad_norm": 0.6878014206886292, "learning_rate": 9.997912931490771e-05, "loss": 0.0883, "step": 4700 }, { "grad_norm": 0.571118175983429, "learning_rate": 9.9978527928067e-05, "loss": 0.0793, "step": 4710 }, { "grad_norm": 0.5451750755310059, "learning_rate": 9.997791800126277e-05, "loss": 0.072, "step": 4720 }, { "grad_norm": 0.49134472012519836, "learning_rate": 9.997729953459927e-05, "loss": 0.0883, "step": 4730 }, { "grad_norm": 0.5274432301521301, "learning_rate": 9.997667252818214e-05, "loss": 0.0774, "step": 4740 }, { "grad_norm": 0.6017180681228638, "learning_rate": 9.997603698211855e-05, "loss": 0.0704, "step": 4750 }, { "grad_norm": 0.5904856324195862, "learning_rate": 9.99753928965171e-05, "loss": 0.0799, "step": 4760 }, { "grad_norm": 0.6767410039901733, "learning_rate": 9.997474027148781e-05, "loss": 0.0892, "step": 4770 }, { "grad_norm": 0.7341177463531494, "learning_rate": 9.997407910714223e-05, "loss": 0.0968, "step": 4780 }, { "grad_norm": 0.6868203282356262, "learning_rate": 9.997340940359332e-05, "loss": 0.0743, "step": 4790 }, { "grad_norm": 0.4444654583930969, "learning_rate": 9.997273116095552e-05, "loss": 0.0629, "step": 4800 }, { "grad_norm": 0.5083044171333313, "learning_rate": 9.997204437934473e-05, "loss": 0.0789, "step": 4810 }, { "grad_norm": 0.6148148775100708, "learning_rate": 9.997134905887829e-05, "loss": 0.0844, "step": 4820 }, { "grad_norm": 0.43725401163101196, "learning_rate": 9.997064519967501e-05, "loss": 0.0705, "step": 4830 }, { "grad_norm": 0.49173521995544434, "learning_rate": 9.996993280185517e-05, "loss": 0.072, "step": 4840 }, { "grad_norm": 0.685658872127533, "learning_rate": 9.99692118655405e-05, "loss": 0.0756, "step": 4850 }, { "grad_norm": 0.5965655446052551, "learning_rate": 9.996848239085417e-05, "loss": 0.0668, "step": 4860 }, { "grad_norm": 0.6387121677398682, "learning_rate": 9.996774437792085e-05, "loss": 0.072, "step": 4870 }, { "grad_norm": 0.5074065327644348, "learning_rate": 9.996699782686664e-05, "loss": 0.0746, "step": 4880 }, { "grad_norm": 0.7313151955604553, "learning_rate": 9.996624273781909e-05, "loss": 0.0929, "step": 4890 }, { "grad_norm": 0.670464813709259, "learning_rate": 9.996547911090725e-05, "loss": 0.0622, "step": 4900 }, { "grad_norm": 0.710430920124054, "learning_rate": 9.996470694626157e-05, "loss": 0.0824, "step": 4910 }, { "grad_norm": 0.4416649639606476, "learning_rate": 9.996392624401403e-05, "loss": 0.0786, "step": 4920 }, { "grad_norm": 0.5147579312324524, "learning_rate": 9.996313700429801e-05, "loss": 0.083, "step": 4930 }, { "grad_norm": 0.5165359377861023, "learning_rate": 9.996233922724836e-05, "loss": 0.0647, "step": 4940 }, { "grad_norm": 0.5489733815193176, "learning_rate": 9.996153291300141e-05, "loss": 0.0696, "step": 4950 }, { "grad_norm": 0.49455639719963074, "learning_rate": 9.996071806169494e-05, "loss": 0.0699, "step": 4960 }, { "grad_norm": 0.510345995426178, "learning_rate": 9.995989467346817e-05, "loss": 0.0834, "step": 4970 }, { "grad_norm": 0.6204002499580383, "learning_rate": 9.995906274846183e-05, "loss": 0.0755, "step": 4980 }, { "grad_norm": 0.5430346727371216, "learning_rate": 9.995822228681803e-05, "loss": 0.0796, "step": 4990 }, { "grad_norm": 0.6124932169914246, "learning_rate": 9.99573732886804e-05, "loss": 0.0916, "step": 5000 }, { "grad_norm": 0.5360018610954285, "learning_rate": 9.995651575419402e-05, "loss": 0.0724, "step": 5010 }, { "grad_norm": 0.7668290734291077, "learning_rate": 9.995564968350541e-05, "loss": 0.0754, "step": 5020 }, { "grad_norm": 0.7208867073059082, "learning_rate": 9.995477507676256e-05, "loss": 0.0663, "step": 5030 }, { "grad_norm": 0.5542147159576416, "learning_rate": 9.995389193411493e-05, "loss": 0.065, "step": 5040 }, { "grad_norm": 0.5439578890800476, "learning_rate": 9.995300025571339e-05, "loss": 0.0697, "step": 5050 }, { "grad_norm": 0.42148104310035706, "learning_rate": 9.995210004171034e-05, "loss": 0.0518, "step": 5060 }, { "grad_norm": 0.5144997239112854, "learning_rate": 9.995119129225956e-05, "loss": 0.0888, "step": 5070 }, { "grad_norm": 0.36481842398643494, "learning_rate": 9.995027400751637e-05, "loss": 0.0827, "step": 5080 }, { "grad_norm": 0.5539650917053223, "learning_rate": 9.994934818763751e-05, "loss": 0.0631, "step": 5090 }, { "grad_norm": 0.5973619818687439, "learning_rate": 9.994841383278115e-05, "loss": 0.0769, "step": 5100 }, { "grad_norm": 0.5902206301689148, "learning_rate": 9.994747094310695e-05, "loss": 0.0812, "step": 5110 }, { "grad_norm": 0.49874693155288696, "learning_rate": 9.994651951877604e-05, "loss": 0.0861, "step": 5120 }, { "grad_norm": 0.5823637843132019, "learning_rate": 9.994555955995099e-05, "loss": 0.0871, "step": 5130 }, { "grad_norm": 0.6945248246192932, "learning_rate": 9.994459106679581e-05, "loss": 0.0703, "step": 5140 }, { "grad_norm": 0.5863952040672302, "learning_rate": 9.994361403947603e-05, "loss": 0.0612, "step": 5150 }, { "grad_norm": 0.43351665139198303, "learning_rate": 9.994262847815854e-05, "loss": 0.0892, "step": 5160 }, { "grad_norm": 0.6038205623626709, "learning_rate": 9.99416343830118e-05, "loss": 0.0845, "step": 5170 }, { "grad_norm": 0.589590847492218, "learning_rate": 9.994063175420565e-05, "loss": 0.0719, "step": 5180 }, { "grad_norm": 0.6400354504585266, "learning_rate": 9.99396205919114e-05, "loss": 0.1005, "step": 5190 }, { "grad_norm": 0.6751314997673035, "learning_rate": 9.993860089630185e-05, "loss": 0.0685, "step": 5200 }, { "grad_norm": 0.5899213552474976, "learning_rate": 9.993757266755123e-05, "loss": 0.0768, "step": 5210 }, { "grad_norm": 0.5585102438926697, "learning_rate": 9.993653590583522e-05, "loss": 0.0809, "step": 5220 }, { "grad_norm": 0.459936261177063, "learning_rate": 9.993549061133102e-05, "loss": 0.0707, "step": 5230 }, { "grad_norm": 0.4345394968986511, "learning_rate": 9.993443678421719e-05, "loss": 0.0732, "step": 5240 }, { "grad_norm": 0.5344228744506836, "learning_rate": 9.993337442467384e-05, "loss": 0.0912, "step": 5250 }, { "grad_norm": 0.5962293148040771, "learning_rate": 9.993230353288248e-05, "loss": 0.0702, "step": 5260 }, { "grad_norm": 0.6955822110176086, "learning_rate": 9.993122410902608e-05, "loss": 0.0753, "step": 5270 }, { "grad_norm": 0.7007014155387878, "learning_rate": 9.993013615328912e-05, "loss": 0.0984, "step": 5280 }, { "grad_norm": 0.4660937488079071, "learning_rate": 9.992903966585747e-05, "loss": 0.0793, "step": 5290 }, { "grad_norm": 0.43879473209381104, "learning_rate": 9.992793464691852e-05, "loss": 0.0655, "step": 5300 }, { "grad_norm": 0.5337172150611877, "learning_rate": 9.992682109666105e-05, "loss": 0.0686, "step": 5310 }, { "grad_norm": 0.4357560873031616, "learning_rate": 9.992569901527538e-05, "loss": 0.0645, "step": 5320 }, { "grad_norm": 0.6230254769325256, "learning_rate": 9.99245684029532e-05, "loss": 0.0847, "step": 5330 }, { "grad_norm": 0.6597492098808289, "learning_rate": 9.992342925988774e-05, "loss": 0.0753, "step": 5340 }, { "grad_norm": 0.4314482808113098, "learning_rate": 9.992228158627361e-05, "loss": 0.0619, "step": 5350 }, { "grad_norm": 0.5893598198890686, "learning_rate": 9.992112538230693e-05, "loss": 0.0681, "step": 5360 }, { "grad_norm": 0.6665131449699402, "learning_rate": 9.991996064818527e-05, "loss": 0.0633, "step": 5370 }, { "grad_norm": 0.4917565584182739, "learning_rate": 9.991878738410768e-05, "loss": 0.0651, "step": 5380 }, { "grad_norm": 0.5501397848129272, "learning_rate": 9.991760559027457e-05, "loss": 0.0627, "step": 5390 }, { "grad_norm": 0.5942494869232178, "learning_rate": 9.991641526688793e-05, "loss": 0.0749, "step": 5400 }, { "grad_norm": 0.5910245180130005, "learning_rate": 9.991521641415113e-05, "loss": 0.0671, "step": 5410 }, { "grad_norm": 0.5512598752975464, "learning_rate": 9.991400903226904e-05, "loss": 0.0711, "step": 5420 }, { "grad_norm": 0.5433836579322815, "learning_rate": 9.991279312144794e-05, "loss": 0.0697, "step": 5430 }, { "grad_norm": 0.6601438522338867, "learning_rate": 9.991156868189564e-05, "loss": 0.0883, "step": 5440 }, { "grad_norm": 0.518059492111206, "learning_rate": 9.991033571382131e-05, "loss": 0.0557, "step": 5450 }, { "grad_norm": 0.5181022882461548, "learning_rate": 9.990909421743569e-05, "loss": 0.0631, "step": 5460 }, { "grad_norm": 0.4280373454093933, "learning_rate": 9.990784419295085e-05, "loss": 0.0653, "step": 5470 }, { "grad_norm": 0.39804643392562866, "learning_rate": 9.990658564058044e-05, "loss": 0.0648, "step": 5480 }, { "grad_norm": 0.44536906480789185, "learning_rate": 9.990531856053948e-05, "loss": 0.0621, "step": 5490 }, { "grad_norm": 0.5840182304382324, "learning_rate": 9.99040429530445e-05, "loss": 0.0727, "step": 5500 }, { "grad_norm": 0.5432393550872803, "learning_rate": 9.990275881831346e-05, "loss": 0.075, "step": 5510 }, { "grad_norm": 0.5221742391586304, "learning_rate": 9.990146615656577e-05, "loss": 0.0608, "step": 5520 }, { "grad_norm": 0.5557933449745178, "learning_rate": 9.990016496802233e-05, "loss": 0.0691, "step": 5530 }, { "grad_norm": 0.45769888162612915, "learning_rate": 9.989885525290548e-05, "loss": 0.0577, "step": 5540 }, { "grad_norm": 0.5839444398880005, "learning_rate": 9.989753701143897e-05, "loss": 0.071, "step": 5550 }, { "grad_norm": 0.5236352682113647, "learning_rate": 9.989621024384812e-05, "loss": 0.062, "step": 5560 }, { "grad_norm": 0.5250220894813538, "learning_rate": 9.989487495035959e-05, "loss": 0.0697, "step": 5570 }, { "grad_norm": 0.7494485378265381, "learning_rate": 9.989353113120156e-05, "loss": 0.0822, "step": 5580 }, { "grad_norm": 0.565776526927948, "learning_rate": 9.989217878660366e-05, "loss": 0.0671, "step": 5590 }, { "grad_norm": 0.43329837918281555, "learning_rate": 9.989081791679695e-05, "loss": 0.0735, "step": 5600 }, { "grad_norm": 0.43007132411003113, "learning_rate": 9.988944852201397e-05, "loss": 0.0632, "step": 5610 }, { "grad_norm": 0.5941780805587769, "learning_rate": 9.988807060248873e-05, "loss": 0.0689, "step": 5620 }, { "grad_norm": 0.41451340913772583, "learning_rate": 9.988668415845665e-05, "loss": 0.0634, "step": 5630 }, { "grad_norm": 0.5798977613449097, "learning_rate": 9.988528919015466e-05, "loss": 0.0727, "step": 5640 }, { "grad_norm": 0.4443800449371338, "learning_rate": 9.988388569782112e-05, "loss": 0.0768, "step": 5650 }, { "grad_norm": 0.5046193599700928, "learning_rate": 9.988247368169583e-05, "loss": 0.0686, "step": 5660 }, { "grad_norm": 0.6700305938720703, "learning_rate": 9.988105314202007e-05, "loss": 0.0673, "step": 5670 }, { "grad_norm": 0.605789303779602, "learning_rate": 9.987962407903659e-05, "loss": 0.0867, "step": 5680 }, { "grad_norm": 0.589546263217926, "learning_rate": 9.987818649298957e-05, "loss": 0.0707, "step": 5690 }, { "grad_norm": 0.4860541522502899, "learning_rate": 9.987674038412465e-05, "loss": 0.0687, "step": 5700 }, { "grad_norm": 0.38682985305786133, "learning_rate": 9.987528575268891e-05, "loss": 0.0794, "step": 5710 }, { "grad_norm": 0.5290650725364685, "learning_rate": 9.987382259893095e-05, "loss": 0.0774, "step": 5720 }, { "grad_norm": 0.5679202675819397, "learning_rate": 9.987235092310074e-05, "loss": 0.079, "step": 5730 }, { "grad_norm": 0.40512609481811523, "learning_rate": 9.987087072544978e-05, "loss": 0.0847, "step": 5740 }, { "grad_norm": 0.5167580842971802, "learning_rate": 9.9869382006231e-05, "loss": 0.0729, "step": 5750 }, { "grad_norm": 0.5534264445304871, "learning_rate": 9.986788476569875e-05, "loss": 0.0778, "step": 5760 }, { "grad_norm": 0.6568450927734375, "learning_rate": 9.986637900410887e-05, "loss": 0.0754, "step": 5770 }, { "grad_norm": 0.6243554353713989, "learning_rate": 9.986486472171869e-05, "loss": 0.0836, "step": 5780 }, { "grad_norm": 0.572994589805603, "learning_rate": 9.986334191878692e-05, "loss": 0.0625, "step": 5790 }, { "grad_norm": 0.5128185749053955, "learning_rate": 9.986181059557378e-05, "loss": 0.0747, "step": 5800 }, { "grad_norm": 0.6092382073402405, "learning_rate": 9.986027075234094e-05, "loss": 0.0744, "step": 5810 }, { "grad_norm": 0.4885047376155853, "learning_rate": 9.985872238935152e-05, "loss": 0.0633, "step": 5820 }, { "grad_norm": 0.48529869318008423, "learning_rate": 9.985716550687008e-05, "loss": 0.0837, "step": 5830 }, { "grad_norm": 0.3994753658771515, "learning_rate": 9.985560010516264e-05, "loss": 0.0753, "step": 5840 }, { "grad_norm": 0.6280574202537537, "learning_rate": 9.985402618449668e-05, "loss": 0.0717, "step": 5850 }, { "grad_norm": 0.6276044249534607, "learning_rate": 9.985244374514118e-05, "loss": 0.0804, "step": 5860 }, { "grad_norm": 0.5145954489707947, "learning_rate": 9.985085278736651e-05, "loss": 0.0787, "step": 5870 }, { "grad_norm": 0.6301870346069336, "learning_rate": 9.984925331144452e-05, "loss": 0.0789, "step": 5880 }, { "grad_norm": 0.4724043607711792, "learning_rate": 9.984764531764851e-05, "loss": 0.0688, "step": 5890 }, { "grad_norm": 0.4268043339252472, "learning_rate": 9.984602880625326e-05, "loss": 0.0919, "step": 5900 }, { "grad_norm": 0.4951004087924957, "learning_rate": 9.9844403777535e-05, "loss": 0.0782, "step": 5910 }, { "grad_norm": 0.6100689172744751, "learning_rate": 9.984277023177135e-05, "loss": 0.0687, "step": 5920 }, { "grad_norm": 0.5842722058296204, "learning_rate": 9.984112816924148e-05, "loss": 0.0621, "step": 5930 }, { "grad_norm": 0.5501399040222168, "learning_rate": 9.983947759022596e-05, "loss": 0.0731, "step": 5940 }, { "grad_norm": 0.5284683704376221, "learning_rate": 9.983781849500682e-05, "loss": 0.0661, "step": 5950 }, { "grad_norm": 0.6251770257949829, "learning_rate": 9.98361508838676e-05, "loss": 0.0611, "step": 5960 }, { "grad_norm": 0.917611300945282, "learning_rate": 9.98344747570932e-05, "loss": 0.0782, "step": 5970 }, { "grad_norm": 0.340389221906662, "learning_rate": 9.983279011497004e-05, "loss": 0.0707, "step": 5980 }, { "grad_norm": 0.6553482413291931, "learning_rate": 9.983109695778596e-05, "loss": 0.0859, "step": 5990 }, { "grad_norm": 0.4129652678966522, "learning_rate": 9.982939528583032e-05, "loss": 0.0913, "step": 6000 }, { "grad_norm": 0.48813319206237793, "learning_rate": 9.982768509939385e-05, "loss": 0.0724, "step": 6010 }, { "grad_norm": 0.4837656319141388, "learning_rate": 9.982596639876879e-05, "loss": 0.0706, "step": 6020 }, { "grad_norm": 0.4291953146457672, "learning_rate": 9.982423918424881e-05, "loss": 0.0759, "step": 6030 }, { "grad_norm": 0.553473949432373, "learning_rate": 9.982250345612908e-05, "loss": 0.0693, "step": 6040 }, { "grad_norm": 0.6651439070701599, "learning_rate": 9.982075921470611e-05, "loss": 0.0832, "step": 6050 }, { "grad_norm": 0.5682922005653381, "learning_rate": 9.981900646027802e-05, "loss": 0.064, "step": 6060 }, { "grad_norm": 0.6850480437278748, "learning_rate": 9.981724519314425e-05, "loss": 0.0689, "step": 6070 }, { "grad_norm": 0.6340208053588867, "learning_rate": 9.981547541360581e-05, "loss": 0.0806, "step": 6080 }, { "grad_norm": 0.5057369470596313, "learning_rate": 9.981369712196508e-05, "loss": 0.0759, "step": 6090 }, { "grad_norm": 0.45176857709884644, "learning_rate": 9.981191031852592e-05, "loss": 0.0837, "step": 6100 }, { "grad_norm": 0.4483061134815216, "learning_rate": 9.981011500359362e-05, "loss": 0.0603, "step": 6110 }, { "grad_norm": 0.6229919195175171, "learning_rate": 9.9808311177475e-05, "loss": 0.0713, "step": 6120 }, { "grad_norm": 0.6018860340118408, "learning_rate": 9.980649884047826e-05, "loss": 0.0627, "step": 6130 }, { "grad_norm": 0.4409308135509491, "learning_rate": 9.980467799291307e-05, "loss": 0.0676, "step": 6140 }, { "grad_norm": 0.4250388741493225, "learning_rate": 9.980284863509058e-05, "loss": 0.0948, "step": 6150 }, { "grad_norm": 0.6210755705833435, "learning_rate": 9.980101076732334e-05, "loss": 0.0956, "step": 6160 }, { "grad_norm": 0.5588876605033875, "learning_rate": 9.979916438992544e-05, "loss": 0.0833, "step": 6170 }, { "grad_norm": 0.3771438002586365, "learning_rate": 9.979730950321237e-05, "loss": 0.068, "step": 6180 }, { "grad_norm": 0.43541061878204346, "learning_rate": 9.979544610750104e-05, "loss": 0.0843, "step": 6190 }, { "grad_norm": 0.37360885739326477, "learning_rate": 9.97935742031099e-05, "loss": 0.0671, "step": 6200 }, { "grad_norm": 0.5165831446647644, "learning_rate": 9.979169379035878e-05, "loss": 0.0785, "step": 6210 }, { "grad_norm": 0.5206578969955444, "learning_rate": 9.978980486956899e-05, "loss": 0.0688, "step": 6220 }, { "grad_norm": 0.47757357358932495, "learning_rate": 9.978790744106332e-05, "loss": 0.0698, "step": 6230 }, { "grad_norm": 0.38382279872894287, "learning_rate": 9.978600150516594e-05, "loss": 0.0653, "step": 6240 }, { "grad_norm": 0.6557047963142395, "learning_rate": 9.978408706220259e-05, "loss": 0.0774, "step": 6250 }, { "grad_norm": 0.6153005957603455, "learning_rate": 9.978216411250032e-05, "loss": 0.0643, "step": 6260 }, { "grad_norm": 0.33512213826179504, "learning_rate": 9.978023265638778e-05, "loss": 0.0651, "step": 6270 }, { "grad_norm": 0.48282137513160706, "learning_rate": 9.977829269419495e-05, "loss": 0.0689, "step": 6280 }, { "grad_norm": 0.46981507539749146, "learning_rate": 9.977634422625335e-05, "loss": 0.0632, "step": 6290 }, { "grad_norm": 0.5184574723243713, "learning_rate": 9.97743872528959e-05, "loss": 0.0643, "step": 6300 }, { "grad_norm": 0.41853150725364685, "learning_rate": 9.9772421774457e-05, "loss": 0.0606, "step": 6310 }, { "grad_norm": 0.4675877094268799, "learning_rate": 9.977044779127252e-05, "loss": 0.0672, "step": 6320 }, { "grad_norm": 0.5638267993927002, "learning_rate": 9.976846530367971e-05, "loss": 0.0697, "step": 6330 }, { "grad_norm": 0.6309231519699097, "learning_rate": 9.976647431201735e-05, "loss": 0.077, "step": 6340 }, { "grad_norm": 0.4996522068977356, "learning_rate": 9.976447481662568e-05, "loss": 0.0695, "step": 6350 }, { "grad_norm": 0.5326684713363647, "learning_rate": 9.976246681784629e-05, "loss": 0.0608, "step": 6360 }, { "grad_norm": 0.4683757424354553, "learning_rate": 9.976045031602234e-05, "loss": 0.0579, "step": 6370 }, { "grad_norm": 0.5087814927101135, "learning_rate": 9.975842531149837e-05, "loss": 0.0639, "step": 6380 }, { "grad_norm": 0.5223796963691711, "learning_rate": 9.975639180462043e-05, "loss": 0.0665, "step": 6390 }, { "grad_norm": 0.6121158599853516, "learning_rate": 9.975434979573596e-05, "loss": 0.0749, "step": 6400 }, { "grad_norm": 0.6001796722412109, "learning_rate": 9.97522992851939e-05, "loss": 0.0556, "step": 6410 }, { "grad_norm": 0.3696732819080353, "learning_rate": 9.975024027334461e-05, "loss": 0.0665, "step": 6420 }, { "grad_norm": 0.4969823956489563, "learning_rate": 9.974817276053993e-05, "loss": 0.06, "step": 6430 }, { "grad_norm": 0.5263616442680359, "learning_rate": 9.974609674713315e-05, "loss": 0.064, "step": 6440 }, { "grad_norm": 0.5208451151847839, "learning_rate": 9.9744012233479e-05, "loss": 0.057, "step": 6450 }, { "grad_norm": 0.49089622497558594, "learning_rate": 9.974191921993366e-05, "loss": 0.0612, "step": 6460 }, { "grad_norm": 0.42745640873908997, "learning_rate": 9.973981770685474e-05, "loss": 0.0599, "step": 6470 }, { "grad_norm": 0.4036257565021515, "learning_rate": 9.97377076946014e-05, "loss": 0.056, "step": 6480 }, { "grad_norm": 0.4832909107208252, "learning_rate": 9.973558918353412e-05, "loss": 0.0695, "step": 6490 }, { "grad_norm": 0.42404434084892273, "learning_rate": 9.973346217401494e-05, "loss": 0.0553, "step": 6500 }, { "grad_norm": 0.5073220133781433, "learning_rate": 9.973132666640726e-05, "loss": 0.0549, "step": 6510 }, { "grad_norm": 0.5929568409919739, "learning_rate": 9.972918266107602e-05, "loss": 0.0608, "step": 6520 }, { "grad_norm": 0.44907253980636597, "learning_rate": 9.972703015838756e-05, "loss": 0.0576, "step": 6530 }, { "grad_norm": 0.5596585273742676, "learning_rate": 9.97248691587097e-05, "loss": 0.0754, "step": 6540 }, { "grad_norm": 0.6843361258506775, "learning_rate": 9.972269966241166e-05, "loss": 0.0589, "step": 6550 }, { "grad_norm": 0.6214447021484375, "learning_rate": 9.972052166986417e-05, "loss": 0.0788, "step": 6560 }, { "grad_norm": 0.6511563062667847, "learning_rate": 9.971833518143938e-05, "loss": 0.0786, "step": 6570 }, { "grad_norm": 0.381474107503891, "learning_rate": 9.971614019751093e-05, "loss": 0.0633, "step": 6580 }, { "grad_norm": 0.5049062371253967, "learning_rate": 9.971393671845383e-05, "loss": 0.073, "step": 6590 }, { "grad_norm": 0.5644552111625671, "learning_rate": 9.971172474464464e-05, "loss": 0.0792, "step": 6600 }, { "grad_norm": 0.6642492413520813, "learning_rate": 9.97095042764613e-05, "loss": 0.058, "step": 6610 }, { "grad_norm": 0.5352599024772644, "learning_rate": 9.970727531428324e-05, "loss": 0.0619, "step": 6620 }, { "grad_norm": 0.5707711577415466, "learning_rate": 9.970503785849132e-05, "loss": 0.0635, "step": 6630 }, { "grad_norm": 0.4083103835582733, "learning_rate": 9.970279190946788e-05, "loss": 0.0591, "step": 6640 }, { "grad_norm": 0.5111121535301208, "learning_rate": 9.970053746759667e-05, "loss": 0.0771, "step": 6650 }, { "grad_norm": 0.37323197722435, "learning_rate": 9.969827453326292e-05, "loss": 0.0562, "step": 6660 }, { "grad_norm": 0.48695480823516846, "learning_rate": 9.969600310685332e-05, "loss": 0.0538, "step": 6670 }, { "grad_norm": 0.527205765247345, "learning_rate": 9.969372318875596e-05, "loss": 0.0819, "step": 6680 }, { "grad_norm": 0.4591539204120636, "learning_rate": 9.969143477936043e-05, "loss": 0.0717, "step": 6690 }, { "grad_norm": 0.5441296100616455, "learning_rate": 9.968913787905775e-05, "loss": 0.0689, "step": 6700 }, { "grad_norm": 0.5800323486328125, "learning_rate": 9.968683248824045e-05, "loss": 0.0629, "step": 6710 }, { "grad_norm": 0.4312031865119934, "learning_rate": 9.968451860730238e-05, "loss": 0.0692, "step": 6720 }, { "grad_norm": 0.6016779541969299, "learning_rate": 9.968219623663896e-05, "loss": 0.0734, "step": 6730 }, { "grad_norm": 0.442147433757782, "learning_rate": 9.967986537664702e-05, "loss": 0.091, "step": 6740 }, { "grad_norm": 0.6016287207603455, "learning_rate": 9.967752602772483e-05, "loss": 0.0744, "step": 6750 }, { "grad_norm": 0.4495202302932739, "learning_rate": 9.967517819027212e-05, "loss": 0.0612, "step": 6760 }, { "grad_norm": 0.5369052290916443, "learning_rate": 9.967282186469009e-05, "loss": 0.0678, "step": 6770 }, { "grad_norm": 0.4877650737762451, "learning_rate": 9.967045705138135e-05, "loss": 0.0772, "step": 6780 }, { "grad_norm": 0.6003624200820923, "learning_rate": 9.966808375074998e-05, "loss": 0.0604, "step": 6790 }, { "grad_norm": 0.563779890537262, "learning_rate": 9.966570196320154e-05, "loss": 0.0701, "step": 6800 }, { "grad_norm": 0.4964142441749573, "learning_rate": 9.966331168914299e-05, "loss": 0.0583, "step": 6810 }, { "grad_norm": 0.5814568400382996, "learning_rate": 9.966091292898277e-05, "loss": 0.0767, "step": 6820 }, { "grad_norm": 0.5543912053108215, "learning_rate": 9.965850568313076e-05, "loss": 0.0683, "step": 6830 }, { "grad_norm": 0.4617127478122711, "learning_rate": 9.965608995199827e-05, "loss": 0.0765, "step": 6840 }, { "grad_norm": 0.4692162275314331, "learning_rate": 9.965366573599812e-05, "loss": 0.065, "step": 6850 }, { "grad_norm": 0.4988427460193634, "learning_rate": 9.965123303554453e-05, "loss": 0.0687, "step": 6860 }, { "grad_norm": 0.6298457384109497, "learning_rate": 9.964879185105317e-05, "loss": 0.0811, "step": 6870 }, { "grad_norm": 0.4696967303752899, "learning_rate": 9.964634218294119e-05, "loss": 0.077, "step": 6880 }, { "grad_norm": 0.5070906281471252, "learning_rate": 9.964388403162714e-05, "loss": 0.0745, "step": 6890 }, { "grad_norm": 0.635201632976532, "learning_rate": 9.96414173975311e-05, "loss": 0.0754, "step": 6900 }, { "grad_norm": 0.6624345183372498, "learning_rate": 9.963894228107451e-05, "loss": 0.0706, "step": 6910 }, { "grad_norm": 0.38734129071235657, "learning_rate": 9.963645868268032e-05, "loss": 0.0784, "step": 6920 }, { "grad_norm": 0.46802932024002075, "learning_rate": 9.963396660277289e-05, "loss": 0.076, "step": 6930 }, { "grad_norm": 0.5243999361991882, "learning_rate": 9.963146604177807e-05, "loss": 0.0552, "step": 6940 }, { "grad_norm": 0.546851634979248, "learning_rate": 9.962895700012311e-05, "loss": 0.0676, "step": 6950 }, { "grad_norm": 0.6749495267868042, "learning_rate": 9.962643947823677e-05, "loss": 0.0643, "step": 6960 }, { "grad_norm": 0.4293058514595032, "learning_rate": 9.962391347654921e-05, "loss": 0.0732, "step": 6970 }, { "grad_norm": 0.5060395002365112, "learning_rate": 9.962137899549204e-05, "loss": 0.0661, "step": 6980 }, { "grad_norm": 0.670983612537384, "learning_rate": 9.961883603549835e-05, "loss": 0.0651, "step": 6990 }, { "grad_norm": 0.5163580775260925, "learning_rate": 9.961628459700267e-05, "loss": 0.0687, "step": 7000 }, { "grad_norm": 0.45854124426841736, "learning_rate": 9.961372468044095e-05, "loss": 0.0755, "step": 7010 }, { "grad_norm": 0.6142167448997498, "learning_rate": 9.961115628625062e-05, "loss": 0.0704, "step": 7020 }, { "grad_norm": 0.5922306776046753, "learning_rate": 9.960857941487056e-05, "loss": 0.0851, "step": 7030 }, { "grad_norm": 0.4486408233642578, "learning_rate": 9.960599406674106e-05, "loss": 0.0637, "step": 7040 }, { "grad_norm": 0.5354169011116028, "learning_rate": 9.960340024230393e-05, "loss": 0.0609, "step": 7050 }, { "grad_norm": 0.5517086982727051, "learning_rate": 9.960079794200232e-05, "loss": 0.0607, "step": 7060 }, { "grad_norm": 0.4629434049129486, "learning_rate": 9.959818716628096e-05, "loss": 0.0688, "step": 7070 }, { "grad_norm": 0.4609384834766388, "learning_rate": 9.95955679155859e-05, "loss": 0.0782, "step": 7080 }, { "grad_norm": 0.5497509837150574, "learning_rate": 9.959294019036472e-05, "loss": 0.068, "step": 7090 }, { "grad_norm": 0.5816410779953003, "learning_rate": 9.959030399106646e-05, "loss": 0.0607, "step": 7100 }, { "grad_norm": 0.551622748374939, "learning_rate": 9.958765931814153e-05, "loss": 0.0728, "step": 7110 }, { "grad_norm": 0.5836251974105835, "learning_rate": 9.958500617204184e-05, "loss": 0.0691, "step": 7120 }, { "grad_norm": 0.5904495716094971, "learning_rate": 9.958234455322075e-05, "loss": 0.0697, "step": 7130 }, { "grad_norm": 0.5385616421699524, "learning_rate": 9.957967446213308e-05, "loss": 0.0945, "step": 7140 }, { "grad_norm": 0.4840993583202362, "learning_rate": 9.957699589923501e-05, "loss": 0.0884, "step": 7150 }, { "grad_norm": 0.7249341011047363, "learning_rate": 9.957430886498431e-05, "loss": 0.077, "step": 7160 }, { "grad_norm": 0.46973973512649536, "learning_rate": 9.957161335984008e-05, "loss": 0.0682, "step": 7170 }, { "grad_norm": 0.6283811926841736, "learning_rate": 9.956890938426291e-05, "loss": 0.0731, "step": 7180 }, { "grad_norm": 0.4361761510372162, "learning_rate": 9.956619693871482e-05, "loss": 0.0722, "step": 7190 }, { "grad_norm": 0.8731459379196167, "learning_rate": 9.956347602365934e-05, "loss": 0.0731, "step": 7200 }, { "grad_norm": 0.5074436068534851, "learning_rate": 9.956074663956135e-05, "loss": 0.0795, "step": 7210 }, { "grad_norm": 0.6189024448394775, "learning_rate": 9.955800878688726e-05, "loss": 0.0754, "step": 7220 }, { "grad_norm": 0.6042683124542236, "learning_rate": 9.955526246610489e-05, "loss": 0.0648, "step": 7230 }, { "grad_norm": 0.6283889412879944, "learning_rate": 9.955250767768349e-05, "loss": 0.0655, "step": 7240 }, { "grad_norm": 0.4982430040836334, "learning_rate": 9.95497444220938e-05, "loss": 0.0719, "step": 7250 }, { "grad_norm": 0.47474780678749084, "learning_rate": 9.954697269980797e-05, "loss": 0.0665, "step": 7260 }, { "grad_norm": 0.5328575968742371, "learning_rate": 9.954419251129962e-05, "loss": 0.0671, "step": 7270 }, { "grad_norm": 0.4238245487213135, "learning_rate": 9.95414038570438e-05, "loss": 0.0635, "step": 7280 }, { "grad_norm": 0.35859522223472595, "learning_rate": 9.953860673751703e-05, "loss": 0.0746, "step": 7290 }, { "grad_norm": 0.42004865407943726, "learning_rate": 9.953580115319725e-05, "loss": 0.0702, "step": 7300 }, { "grad_norm": 0.48875242471694946, "learning_rate": 9.953298710456387e-05, "loss": 0.0721, "step": 7310 }, { "grad_norm": 0.7424805760383606, "learning_rate": 9.953016459209771e-05, "loss": 0.0704, "step": 7320 }, { "grad_norm": 0.5641757845878601, "learning_rate": 9.952733361628108e-05, "loss": 0.0751, "step": 7330 }, { "grad_norm": 0.5122624039649963, "learning_rate": 9.952449417759772e-05, "loss": 0.0725, "step": 7340 }, { "grad_norm": 0.4795132279396057, "learning_rate": 9.952164627653279e-05, "loss": 0.0617, "step": 7350 }, { "grad_norm": 0.6270226836204529, "learning_rate": 9.951878991357292e-05, "loss": 0.0695, "step": 7360 }, { "grad_norm": 0.5537957549095154, "learning_rate": 9.951592508920622e-05, "loss": 0.0739, "step": 7370 }, { "grad_norm": 0.5048555135726929, "learning_rate": 9.951305180392219e-05, "loss": 0.0664, "step": 7380 }, { "grad_norm": 0.4507305324077606, "learning_rate": 9.951017005821178e-05, "loss": 0.0639, "step": 7390 }, { "grad_norm": 0.3755262792110443, "learning_rate": 9.95072798525674e-05, "loss": 0.0628, "step": 7400 }, { "grad_norm": 0.4105563163757324, "learning_rate": 9.950438118748293e-05, "loss": 0.0679, "step": 7410 }, { "grad_norm": 0.6010271906852722, "learning_rate": 9.950147406345366e-05, "loss": 0.0821, "step": 7420 }, { "grad_norm": 0.4674542248249054, "learning_rate": 9.949855848097635e-05, "loss": 0.0763, "step": 7430 }, { "grad_norm": 0.4510331153869629, "learning_rate": 9.949563444054916e-05, "loss": 0.0595, "step": 7440 }, { "grad_norm": 0.39099085330963135, "learning_rate": 9.949270194267178e-05, "loss": 0.0646, "step": 7450 }, { "grad_norm": 0.4098275899887085, "learning_rate": 9.948976098784526e-05, "loss": 0.0666, "step": 7460 }, { "grad_norm": 0.5646865367889404, "learning_rate": 9.948681157657213e-05, "loss": 0.0731, "step": 7470 }, { "grad_norm": 0.578012228012085, "learning_rate": 9.948385370935638e-05, "loss": 0.0756, "step": 7480 }, { "grad_norm": 0.5717450976371765, "learning_rate": 9.94808873867034e-05, "loss": 0.072, "step": 7490 }, { "grad_norm": 0.4814620614051819, "learning_rate": 9.947791260912009e-05, "loss": 0.063, "step": 7500 }, { "grad_norm": 0.3758082985877991, "learning_rate": 9.947492937711474e-05, "loss": 0.0583, "step": 7510 }, { "grad_norm": 0.4597966969013214, "learning_rate": 9.947193769119707e-05, "loss": 0.0592, "step": 7520 }, { "grad_norm": 0.38302165269851685, "learning_rate": 9.946893755187834e-05, "loss": 0.0533, "step": 7530 }, { "grad_norm": 0.5751453042030334, "learning_rate": 9.946592895967115e-05, "loss": 0.0628, "step": 7540 }, { "grad_norm": 0.5188515186309814, "learning_rate": 9.94629119150896e-05, "loss": 0.0612, "step": 7550 }, { "grad_norm": 0.46274009346961975, "learning_rate": 9.94598864186492e-05, "loss": 0.065, "step": 7560 }, { "grad_norm": 0.5588036179542542, "learning_rate": 9.945685247086696e-05, "loss": 0.0805, "step": 7570 }, { "grad_norm": 0.5267701745033264, "learning_rate": 9.945381007226129e-05, "loss": 0.0716, "step": 7580 }, { "grad_norm": 0.5407025218009949, "learning_rate": 9.945075922335203e-05, "loss": 0.074, "step": 7590 }, { "grad_norm": 0.5358482599258423, "learning_rate": 9.944769992466049e-05, "loss": 0.0841, "step": 7600 }, { "grad_norm": 0.6815295815467834, "learning_rate": 9.944463217670945e-05, "loss": 0.0714, "step": 7610 }, { "grad_norm": 0.6438639163970947, "learning_rate": 9.944155598002307e-05, "loss": 0.0685, "step": 7620 }, { "grad_norm": 0.43018969893455505, "learning_rate": 9.943847133512701e-05, "loss": 0.0701, "step": 7630 }, { "grad_norm": 0.44768181443214417, "learning_rate": 9.943537824254834e-05, "loss": 0.0657, "step": 7640 }, { "grad_norm": 0.5029955506324768, "learning_rate": 9.943227670281559e-05, "loss": 0.064, "step": 7650 }, { "grad_norm": 0.4008708596229553, "learning_rate": 9.942916671645873e-05, "loss": 0.0603, "step": 7660 }, { "grad_norm": 0.4451042115688324, "learning_rate": 9.942604828400916e-05, "loss": 0.0774, "step": 7670 }, { "grad_norm": 0.4590539038181305, "learning_rate": 9.942292140599975e-05, "loss": 0.0692, "step": 7680 }, { "grad_norm": 0.6522374153137207, "learning_rate": 9.94197860829648e-05, "loss": 0.0687, "step": 7690 }, { "grad_norm": 0.6158431172370911, "learning_rate": 9.941664231544004e-05, "loss": 0.0562, "step": 7700 }, { "grad_norm": 0.5502973794937134, "learning_rate": 9.941349010396264e-05, "loss": 0.084, "step": 7710 }, { "grad_norm": 0.534764289855957, "learning_rate": 9.941032944907125e-05, "loss": 0.0804, "step": 7720 }, { "grad_norm": 0.3911565840244293, "learning_rate": 9.940716035130596e-05, "loss": 0.0703, "step": 7730 }, { "grad_norm": 0.5235242247581482, "learning_rate": 9.940398281120821e-05, "loss": 0.0725, "step": 7740 }, { "grad_norm": 0.421417236328125, "learning_rate": 9.940079682932102e-05, "loss": 0.0584, "step": 7750 }, { "grad_norm": 0.37505316734313965, "learning_rate": 9.939760240618877e-05, "loss": 0.072, "step": 7760 }, { "grad_norm": 0.48741206526756287, "learning_rate": 9.939439954235729e-05, "loss": 0.0646, "step": 7770 }, { "grad_norm": 0.5279900431632996, "learning_rate": 9.939118823837387e-05, "loss": 0.0748, "step": 7780 }, { "grad_norm": 0.48059365153312683, "learning_rate": 9.938796849478725e-05, "loss": 0.0729, "step": 7790 }, { "grad_norm": 0.5172619819641113, "learning_rate": 9.938474031214755e-05, "loss": 0.0653, "step": 7800 }, { "grad_norm": 0.4354655146598816, "learning_rate": 9.938150369100643e-05, "loss": 0.0712, "step": 7810 }, { "grad_norm": 0.43515875935554504, "learning_rate": 9.93782586319169e-05, "loss": 0.0493, "step": 7820 }, { "grad_norm": 0.5339779257774353, "learning_rate": 9.937500513543348e-05, "loss": 0.0605, "step": 7830 }, { "grad_norm": 0.6047636866569519, "learning_rate": 9.937174320211207e-05, "loss": 0.061, "step": 7840 }, { "grad_norm": 0.4340154528617859, "learning_rate": 9.936847283251009e-05, "loss": 0.0628, "step": 7850 }, { "grad_norm": 0.48803460597991943, "learning_rate": 9.936519402718632e-05, "loss": 0.0714, "step": 7860 }, { "grad_norm": 0.4228019118309021, "learning_rate": 9.936190678670102e-05, "loss": 0.0732, "step": 7870 }, { "grad_norm": 0.5470113754272461, "learning_rate": 9.935861111161593e-05, "loss": 0.0669, "step": 7880 }, { "grad_norm": 0.48369860649108887, "learning_rate": 9.935530700249416e-05, "loss": 0.0611, "step": 7890 }, { "grad_norm": 0.49406149983406067, "learning_rate": 9.935199445990028e-05, "loss": 0.0766, "step": 7900 }, { "grad_norm": 0.46070238947868347, "learning_rate": 9.934867348440033e-05, "loss": 0.0701, "step": 7910 }, { "grad_norm": 0.5278700590133667, "learning_rate": 9.934534407656176e-05, "loss": 0.0897, "step": 7920 }, { "grad_norm": 0.5314106345176697, "learning_rate": 9.93420062369535e-05, "loss": 0.0644, "step": 7930 }, { "grad_norm": 0.5266605019569397, "learning_rate": 9.933865996614589e-05, "loss": 0.0629, "step": 7940 }, { "grad_norm": 0.5834699869155884, "learning_rate": 9.933530526471068e-05, "loss": 0.0619, "step": 7950 }, { "grad_norm": 0.42864227294921875, "learning_rate": 9.933194213322114e-05, "loss": 0.0598, "step": 7960 }, { "grad_norm": 0.38566312193870544, "learning_rate": 9.932857057225192e-05, "loss": 0.0604, "step": 7970 }, { "grad_norm": 0.45390307903289795, "learning_rate": 9.932519058237912e-05, "loss": 0.056, "step": 7980 }, { "grad_norm": 0.4154333770275116, "learning_rate": 9.932180216418032e-05, "loss": 0.0558, "step": 7990 }, { "grad_norm": 0.5145527124404907, "learning_rate": 9.931840531823446e-05, "loss": 0.0572, "step": 8000 }, { "grad_norm": 0.5134528875350952, "learning_rate": 9.9315000045122e-05, "loss": 0.0491, "step": 8010 }, { "grad_norm": 0.36085838079452515, "learning_rate": 9.931158634542481e-05, "loss": 0.0591, "step": 8020 }, { "grad_norm": 0.5780758261680603, "learning_rate": 9.930816421972617e-05, "loss": 0.0557, "step": 8030 }, { "grad_norm": 0.4679228663444519, "learning_rate": 9.930473366861086e-05, "loss": 0.0584, "step": 8040 }, { "grad_norm": 0.3450517952442169, "learning_rate": 9.930129469266505e-05, "loss": 0.0668, "step": 8050 }, { "grad_norm": 0.4972163140773773, "learning_rate": 9.929784729247638e-05, "loss": 0.0616, "step": 8060 }, { "grad_norm": 0.4642854928970337, "learning_rate": 9.929439146863389e-05, "loss": 0.0636, "step": 8070 }, { "grad_norm": 0.5426614284515381, "learning_rate": 9.92909272217281e-05, "loss": 0.0794, "step": 8080 }, { "grad_norm": 0.5176125168800354, "learning_rate": 9.928745455235097e-05, "loss": 0.066, "step": 8090 }, { "grad_norm": 0.48984748125076294, "learning_rate": 9.928397346109588e-05, "loss": 0.068, "step": 8100 }, { "grad_norm": 0.658992350101471, "learning_rate": 9.928048394855762e-05, "loss": 0.0728, "step": 8110 }, { "grad_norm": 0.40063366293907166, "learning_rate": 9.92769860153325e-05, "loss": 0.0568, "step": 8120 }, { "grad_norm": 0.4496573507785797, "learning_rate": 9.927347966201819e-05, "loss": 0.0812, "step": 8130 }, { "grad_norm": 0.5002442598342896, "learning_rate": 9.926996488921383e-05, "loss": 0.0699, "step": 8140 }, { "grad_norm": 0.4831896424293518, "learning_rate": 9.926644169752001e-05, "loss": 0.0594, "step": 8150 }, { "grad_norm": 0.6507985591888428, "learning_rate": 9.926291008753875e-05, "loss": 0.0759, "step": 8160 }, { "grad_norm": 0.40655791759490967, "learning_rate": 9.92593700598735e-05, "loss": 0.0686, "step": 8170 }, { "grad_norm": 0.4329284131526947, "learning_rate": 9.925582161512915e-05, "loss": 0.0626, "step": 8180 }, { "grad_norm": 0.4074851870536804, "learning_rate": 9.925226475391205e-05, "loss": 0.0663, "step": 8190 }, { "grad_norm": 0.5021207928657532, "learning_rate": 9.924869947682993e-05, "loss": 0.0637, "step": 8200 }, { "grad_norm": 0.5115489959716797, "learning_rate": 9.924512578449204e-05, "loss": 0.0519, "step": 8210 }, { "grad_norm": 0.47963109612464905, "learning_rate": 9.924154367750901e-05, "loss": 0.0612, "step": 8220 }, { "grad_norm": 0.4317528307437897, "learning_rate": 9.923795315649293e-05, "loss": 0.0634, "step": 8230 }, { "grad_norm": 0.46913963556289673, "learning_rate": 9.92343542220573e-05, "loss": 0.0562, "step": 8240 }, { "grad_norm": 0.41140568256378174, "learning_rate": 9.92307468748171e-05, "loss": 0.0589, "step": 8250 }, { "grad_norm": 0.4147430956363678, "learning_rate": 9.922713111538873e-05, "loss": 0.0493, "step": 8260 }, { "grad_norm": 0.5207788944244385, "learning_rate": 9.922350694439003e-05, "loss": 0.0611, "step": 8270 }, { "grad_norm": 0.46464428305625916, "learning_rate": 9.921987436244024e-05, "loss": 0.0686, "step": 8280 }, { "grad_norm": 0.3517715036869049, "learning_rate": 9.921623337016008e-05, "loss": 0.0494, "step": 8290 }, { "grad_norm": 0.40213337540626526, "learning_rate": 9.921258396817172e-05, "loss": 0.0581, "step": 8300 }, { "grad_norm": 0.5705678462982178, "learning_rate": 9.920892615709874e-05, "loss": 0.053, "step": 8310 }, { "grad_norm": 0.45341503620147705, "learning_rate": 9.920525993756612e-05, "loss": 0.0554, "step": 8320 }, { "grad_norm": 0.37895533442497253, "learning_rate": 9.920158531020036e-05, "loss": 0.0564, "step": 8330 }, { "grad_norm": 0.41272205114364624, "learning_rate": 9.919790227562933e-05, "loss": 0.0486, "step": 8340 }, { "grad_norm": 0.48362997174263, "learning_rate": 9.919421083448237e-05, "loss": 0.0686, "step": 8350 }, { "grad_norm": 0.5972383618354797, "learning_rate": 9.919051098739022e-05, "loss": 0.0513, "step": 8360 }, { "grad_norm": 0.445295512676239, "learning_rate": 9.918680273498514e-05, "loss": 0.0568, "step": 8370 }, { "grad_norm": 0.4838412404060364, "learning_rate": 9.918308607790072e-05, "loss": 0.0648, "step": 8380 }, { "grad_norm": 0.5608775615692139, "learning_rate": 9.917936101677205e-05, "loss": 0.0477, "step": 8390 }, { "grad_norm": 0.5712345242500305, "learning_rate": 9.917562755223564e-05, "loss": 0.0519, "step": 8400 }, { "grad_norm": 0.5544970035552979, "learning_rate": 9.917188568492944e-05, "loss": 0.0641, "step": 8410 }, { "grad_norm": 0.30817434191703796, "learning_rate": 9.916813541549283e-05, "loss": 0.0504, "step": 8420 }, { "grad_norm": 0.4768874943256378, "learning_rate": 9.916437674456663e-05, "loss": 0.0574, "step": 8430 }, { "grad_norm": 0.38819000124931335, "learning_rate": 9.916060967279308e-05, "loss": 0.0575, "step": 8440 }, { "grad_norm": 0.47178661823272705, "learning_rate": 9.91568342008159e-05, "loss": 0.0514, "step": 8450 }, { "grad_norm": 0.4042569696903229, "learning_rate": 9.915305032928019e-05, "loss": 0.0521, "step": 8460 }, { "grad_norm": 0.4750474691390991, "learning_rate": 9.914925805883253e-05, "loss": 0.052, "step": 8470 }, { "grad_norm": 0.4352496862411499, "learning_rate": 9.914545739012088e-05, "loss": 0.0457, "step": 8480 }, { "grad_norm": 0.5269179344177246, "learning_rate": 9.91416483237947e-05, "loss": 0.0619, "step": 8490 }, { "grad_norm": 0.5110798478126526, "learning_rate": 9.913783086050485e-05, "loss": 0.0643, "step": 8500 }, { "grad_norm": 0.4994186758995056, "learning_rate": 9.913400500090364e-05, "loss": 0.0574, "step": 8510 }, { "grad_norm": 0.4593810439109802, "learning_rate": 9.913017074564479e-05, "loss": 0.0472, "step": 8520 }, { "grad_norm": 0.5426880121231079, "learning_rate": 9.912632809538348e-05, "loss": 0.0658, "step": 8530 }, { "grad_norm": 0.3970870077610016, "learning_rate": 9.912247705077629e-05, "loss": 0.0541, "step": 8540 }, { "grad_norm": 0.45283493399620056, "learning_rate": 9.911861761248127e-05, "loss": 0.0563, "step": 8550 }, { "grad_norm": 0.4398593008518219, "learning_rate": 9.91147497811579e-05, "loss": 0.0606, "step": 8560 }, { "grad_norm": 0.560821533203125, "learning_rate": 9.911087355746709e-05, "loss": 0.0949, "step": 8570 }, { "grad_norm": 0.5040959119796753, "learning_rate": 9.910698894207117e-05, "loss": 0.0738, "step": 8580 }, { "grad_norm": 0.5236146450042725, "learning_rate": 9.910309593563392e-05, "loss": 0.0658, "step": 8590 }, { "grad_norm": 0.4230406880378723, "learning_rate": 9.909919453882057e-05, "loss": 0.0512, "step": 8600 }, { "grad_norm": 0.45447272062301636, "learning_rate": 9.90952847522977e-05, "loss": 0.0604, "step": 8610 }, { "grad_norm": 0.43524789810180664, "learning_rate": 9.909136657673346e-05, "loss": 0.0726, "step": 8620 }, { "grad_norm": 0.4029393494129181, "learning_rate": 9.908744001279731e-05, "loss": 0.0583, "step": 8630 }, { "grad_norm": 0.5720531344413757, "learning_rate": 9.90835050611602e-05, "loss": 0.0703, "step": 8640 }, { "grad_norm": 0.4635620713233948, "learning_rate": 9.90795617224945e-05, "loss": 0.0668, "step": 8650 }, { "grad_norm": 0.5083674192428589, "learning_rate": 9.907560999747405e-05, "loss": 0.0687, "step": 8660 }, { "grad_norm": 0.39934784173965454, "learning_rate": 9.907164988677408e-05, "loss": 0.0569, "step": 8670 }, { "grad_norm": 0.426264226436615, "learning_rate": 9.906768139107124e-05, "loss": 0.0565, "step": 8680 }, { "grad_norm": 0.4083506464958191, "learning_rate": 9.906370451104367e-05, "loss": 0.0564, "step": 8690 }, { "grad_norm": 0.5251712203025818, "learning_rate": 9.905971924737088e-05, "loss": 0.0606, "step": 8700 }, { "grad_norm": 0.4491998851299286, "learning_rate": 9.905572560073387e-05, "loss": 0.0655, "step": 8710 }, { "grad_norm": 0.5979124903678894, "learning_rate": 9.905172357181501e-05, "loss": 0.083, "step": 8720 }, { "grad_norm": 0.4582366943359375, "learning_rate": 9.904771316129817e-05, "loss": 0.0597, "step": 8730 }, { "grad_norm": 0.6281919479370117, "learning_rate": 9.904369436986862e-05, "loss": 0.0581, "step": 8740 }, { "grad_norm": 0.48306596279144287, "learning_rate": 9.903966719821303e-05, "loss": 0.0783, "step": 8750 }, { "grad_norm": 0.5081343054771423, "learning_rate": 9.903563164701956e-05, "loss": 0.0562, "step": 8760 }, { "grad_norm": 0.5898358821868896, "learning_rate": 9.903158771697778e-05, "loss": 0.0702, "step": 8770 }, { "grad_norm": 0.5742965340614319, "learning_rate": 9.902753540877867e-05, "loss": 0.0717, "step": 8780 }, { "grad_norm": 0.43891242146492004, "learning_rate": 9.902347472311466e-05, "loss": 0.0589, "step": 8790 }, { "grad_norm": 0.5116381645202637, "learning_rate": 9.901940566067962e-05, "loss": 0.0521, "step": 8800 }, { "grad_norm": 0.604548990726471, "learning_rate": 9.901532822216883e-05, "loss": 0.0662, "step": 8810 }, { "grad_norm": 0.5128077864646912, "learning_rate": 9.901124240827904e-05, "loss": 0.0652, "step": 8820 }, { "grad_norm": 0.3789030909538269, "learning_rate": 9.900714821970835e-05, "loss": 0.0628, "step": 8830 }, { "grad_norm": 0.5139439105987549, "learning_rate": 9.900304565715641e-05, "loss": 0.0619, "step": 8840 }, { "grad_norm": 0.4353529214859009, "learning_rate": 9.899893472132419e-05, "loss": 0.0578, "step": 8850 }, { "grad_norm": 0.6824318766593933, "learning_rate": 9.899481541291415e-05, "loss": 0.0664, "step": 8860 }, { "grad_norm": 0.4788649380207062, "learning_rate": 9.899068773263016e-05, "loss": 0.0664, "step": 8870 }, { "grad_norm": 0.35767924785614014, "learning_rate": 9.898655168117754e-05, "loss": 0.0667, "step": 8880 }, { "grad_norm": 0.5351145267486572, "learning_rate": 9.898240725926302e-05, "loss": 0.0664, "step": 8890 }, { "grad_norm": 0.596908688545227, "learning_rate": 9.897825446759478e-05, "loss": 0.0669, "step": 8900 }, { "grad_norm": 0.40790003538131714, "learning_rate": 9.897409330688241e-05, "loss": 0.0599, "step": 8910 }, { "grad_norm": 0.6705203652381897, "learning_rate": 9.896992377783692e-05, "loss": 0.0837, "step": 8920 }, { "grad_norm": 0.4882623255252838, "learning_rate": 9.89657458811708e-05, "loss": 0.0572, "step": 8930 }, { "grad_norm": 0.330678254365921, "learning_rate": 9.896155961759792e-05, "loss": 0.056, "step": 8940 }, { "grad_norm": 0.38340359926223755, "learning_rate": 9.895736498783361e-05, "loss": 0.0553, "step": 8950 }, { "grad_norm": 0.5479548573493958, "learning_rate": 9.895316199259462e-05, "loss": 0.0524, "step": 8960 }, { "grad_norm": 0.396541953086853, "learning_rate": 9.894895063259909e-05, "loss": 0.0472, "step": 8970 }, { "grad_norm": 0.3088059723377228, "learning_rate": 9.894473090856667e-05, "loss": 0.0521, "step": 8980 }, { "grad_norm": 0.553148627281189, "learning_rate": 9.894050282121839e-05, "loss": 0.0882, "step": 8990 }, { "grad_norm": 0.3659014105796814, "learning_rate": 9.893626637127668e-05, "loss": 0.0625, "step": 9000 }, { "grad_norm": 0.3885219097137451, "learning_rate": 9.893202155946546e-05, "loss": 0.0623, "step": 9010 }, { "grad_norm": 0.5201266407966614, "learning_rate": 9.892776838651006e-05, "loss": 0.0584, "step": 9020 }, { "grad_norm": 0.4573838412761688, "learning_rate": 9.892350685313722e-05, "loss": 0.0733, "step": 9030 }, { "grad_norm": 0.40488407015800476, "learning_rate": 9.891923696007513e-05, "loss": 0.0693, "step": 9040 }, { "grad_norm": 0.39656195044517517, "learning_rate": 9.891495870805336e-05, "loss": 0.0554, "step": 9050 }, { "grad_norm": 0.454537957906723, "learning_rate": 9.891067209780298e-05, "loss": 0.0618, "step": 9060 }, { "grad_norm": 0.3404441177845001, "learning_rate": 9.890637713005646e-05, "loss": 0.0529, "step": 9070 }, { "grad_norm": 0.5432325005531311, "learning_rate": 9.890207380554767e-05, "loss": 0.0602, "step": 9080 }, { "grad_norm": 0.45556360483169556, "learning_rate": 9.889776212501196e-05, "loss": 0.0661, "step": 9090 }, { "grad_norm": 0.44622308015823364, "learning_rate": 9.889344208918605e-05, "loss": 0.0663, "step": 9100 }, { "grad_norm": 0.4856739044189453, "learning_rate": 9.888911369880812e-05, "loss": 0.0615, "step": 9110 }, { "grad_norm": 0.4817464053630829, "learning_rate": 9.888477695461777e-05, "loss": 0.0636, "step": 9120 }, { "grad_norm": 0.4267953932285309, "learning_rate": 9.888043185735607e-05, "loss": 0.0617, "step": 9130 }, { "grad_norm": 0.4349353611469269, "learning_rate": 9.887607840776542e-05, "loss": 0.0479, "step": 9140 }, { "grad_norm": 0.40090593695640564, "learning_rate": 9.887171660658975e-05, "loss": 0.0661, "step": 9150 }, { "grad_norm": 0.4729679226875305, "learning_rate": 9.886734645457435e-05, "loss": 0.052, "step": 9160 }, { "grad_norm": 0.42229101061820984, "learning_rate": 9.886296795246597e-05, "loss": 0.0556, "step": 9170 }, { "grad_norm": 0.4325605630874634, "learning_rate": 9.885858110101276e-05, "loss": 0.0546, "step": 9180 }, { "grad_norm": 0.40665385127067566, "learning_rate": 9.885418590096434e-05, "loss": 0.0486, "step": 9190 }, { "grad_norm": 0.46506062150001526, "learning_rate": 9.88497823530717e-05, "loss": 0.0604, "step": 9200 }, { "grad_norm": 0.5361029505729675, "learning_rate": 9.884537045808732e-05, "loss": 0.0683, "step": 9210 }, { "grad_norm": 0.3800247311592102, "learning_rate": 9.884095021676502e-05, "loss": 0.067, "step": 9220 }, { "grad_norm": 0.4250519275665283, "learning_rate": 9.883652162986017e-05, "loss": 0.0537, "step": 9230 }, { "grad_norm": 0.4387487769126892, "learning_rate": 9.883208469812943e-05, "loss": 0.0616, "step": 9240 }, { "grad_norm": 0.35138627886772156, "learning_rate": 9.882763942233098e-05, "loss": 0.0481, "step": 9250 }, { "grad_norm": 0.3925662338733673, "learning_rate": 9.882318580322441e-05, "loss": 0.0474, "step": 9260 }, { "grad_norm": 0.5544969439506531, "learning_rate": 9.881872384157067e-05, "loss": 0.0576, "step": 9270 }, { "grad_norm": 0.5182788372039795, "learning_rate": 9.881425353813225e-05, "loss": 0.0552, "step": 9280 }, { "grad_norm": 0.4960181713104248, "learning_rate": 9.880977489367296e-05, "loss": 0.0629, "step": 9290 }, { "grad_norm": 0.32602325081825256, "learning_rate": 9.88052879089581e-05, "loss": 0.0701, "step": 9300 }, { "grad_norm": 0.5217883586883545, "learning_rate": 9.880079258475434e-05, "loss": 0.0559, "step": 9310 }, { "grad_norm": 0.4490637481212616, "learning_rate": 9.879628892182985e-05, "loss": 0.0761, "step": 9320 }, { "grad_norm": 0.4241343140602112, "learning_rate": 9.879177692095416e-05, "loss": 0.0552, "step": 9330 }, { "grad_norm": 0.3716433346271515, "learning_rate": 9.878725658289825e-05, "loss": 0.0633, "step": 9340 }, { "grad_norm": 0.5137599110603333, "learning_rate": 9.878272790843454e-05, "loss": 0.0585, "step": 9350 }, { "grad_norm": 0.44407278299331665, "learning_rate": 9.877819089833682e-05, "loss": 0.0516, "step": 9360 }, { "grad_norm": 0.3401010036468506, "learning_rate": 9.877364555338038e-05, "loss": 0.0685, "step": 9370 }, { "grad_norm": 0.4613490700721741, "learning_rate": 9.876909187434186e-05, "loss": 0.055, "step": 9380 }, { "grad_norm": 0.5075790286064148, "learning_rate": 9.876452986199939e-05, "loss": 0.0586, "step": 9390 }, { "grad_norm": 0.39198777079582214, "learning_rate": 9.875995951713248e-05, "loss": 0.0665, "step": 9400 }, { "grad_norm": 0.47865691781044006, "learning_rate": 9.875538084052207e-05, "loss": 0.0621, "step": 9410 }, { "grad_norm": 0.49071067571640015, "learning_rate": 9.875079383295053e-05, "loss": 0.0612, "step": 9420 }, { "grad_norm": 0.391577810049057, "learning_rate": 9.874619849520167e-05, "loss": 0.0503, "step": 9430 }, { "grad_norm": 0.3682577610015869, "learning_rate": 9.874159482806069e-05, "loss": 0.0522, "step": 9440 }, { "grad_norm": 0.45112356543540955, "learning_rate": 9.873698283231426e-05, "loss": 0.0472, "step": 9450 }, { "grad_norm": 0.4336397349834442, "learning_rate": 9.87323625087504e-05, "loss": 0.0534, "step": 9460 }, { "grad_norm": 0.43640100955963135, "learning_rate": 9.872773385815863e-05, "loss": 0.0701, "step": 9470 }, { "grad_norm": 0.4128406047821045, "learning_rate": 9.872309688132986e-05, "loss": 0.0551, "step": 9480 }, { "grad_norm": 0.30556362867355347, "learning_rate": 9.871845157905639e-05, "loss": 0.0519, "step": 9490 }, { "grad_norm": 0.43085187673568726, "learning_rate": 9.871379795213201e-05, "loss": 0.0509, "step": 9500 }, { "grad_norm": 0.42648836970329285, "learning_rate": 9.87091360013519e-05, "loss": 0.0534, "step": 9510 }, { "grad_norm": 0.4473173916339874, "learning_rate": 9.870446572751262e-05, "loss": 0.0589, "step": 9520 }, { "grad_norm": 0.2974000871181488, "learning_rate": 9.869978713141224e-05, "loss": 0.0546, "step": 9530 }, { "grad_norm": 0.7192185521125793, "learning_rate": 9.869510021385016e-05, "loss": 0.0677, "step": 9540 }, { "grad_norm": 0.3184703588485718, "learning_rate": 9.869040497562727e-05, "loss": 0.0553, "step": 9550 }, { "grad_norm": 0.2838827073574066, "learning_rate": 9.868570141754587e-05, "loss": 0.0584, "step": 9560 }, { "grad_norm": 0.32816004753112793, "learning_rate": 9.868098954040965e-05, "loss": 0.0644, "step": 9570 }, { "grad_norm": 0.48796218633651733, "learning_rate": 9.867626934502374e-05, "loss": 0.0658, "step": 9580 }, { "grad_norm": 0.45754116773605347, "learning_rate": 9.86715408321947e-05, "loss": 0.0536, "step": 9590 }, { "grad_norm": 0.5245179533958435, "learning_rate": 9.86668040027305e-05, "loss": 0.0523, "step": 9600 }, { "grad_norm": 0.42810186743736267, "learning_rate": 9.866205885744053e-05, "loss": 0.0531, "step": 9610 }, { "grad_norm": 0.4342571794986725, "learning_rate": 9.865730539713563e-05, "loss": 0.0606, "step": 9620 }, { "grad_norm": 0.39896559715270996, "learning_rate": 9.8652543622628e-05, "loss": 0.0472, "step": 9630 }, { "grad_norm": 0.3972679376602173, "learning_rate": 9.864777353473132e-05, "loss": 0.0498, "step": 9640 }, { "grad_norm": 0.28592944145202637, "learning_rate": 9.864299513426068e-05, "loss": 0.0602, "step": 9650 }, { "grad_norm": 0.42367616295814514, "learning_rate": 9.863820842203254e-05, "loss": 0.0637, "step": 9660 }, { "grad_norm": 0.6054775714874268, "learning_rate": 9.863341339886483e-05, "loss": 0.0652, "step": 9670 }, { "grad_norm": 0.37150052189826965, "learning_rate": 9.86286100655769e-05, "loss": 0.0507, "step": 9680 }, { "grad_norm": 0.665690541267395, "learning_rate": 9.862379842298953e-05, "loss": 0.0597, "step": 9690 }, { "grad_norm": 0.5936024188995361, "learning_rate": 9.861897847192485e-05, "loss": 0.058, "step": 9700 }, { "grad_norm": 0.5027899742126465, "learning_rate": 9.86141502132065e-05, "loss": 0.0522, "step": 9710 }, { "grad_norm": 0.47375962138175964, "learning_rate": 9.860931364765946e-05, "loss": 0.0599, "step": 9720 }, { "grad_norm": 0.5365285277366638, "learning_rate": 9.860446877611021e-05, "loss": 0.069, "step": 9730 }, { "grad_norm": 0.4371667802333832, "learning_rate": 9.859961559938655e-05, "loss": 0.0522, "step": 9740 }, { "grad_norm": 0.43319669365882874, "learning_rate": 9.85947541183178e-05, "loss": 0.0559, "step": 9750 }, { "grad_norm": 0.6399970650672913, "learning_rate": 9.858988433373463e-05, "loss": 0.0594, "step": 9760 }, { "grad_norm": 0.5436923503875732, "learning_rate": 9.858500624646918e-05, "loss": 0.0571, "step": 9770 }, { "grad_norm": 0.45646122097969055, "learning_rate": 9.858011985735497e-05, "loss": 0.0542, "step": 9780 }, { "grad_norm": 0.4701903760433197, "learning_rate": 9.857522516722693e-05, "loss": 0.0562, "step": 9790 }, { "grad_norm": 0.4031171500682831, "learning_rate": 9.857032217692145e-05, "loss": 0.0501, "step": 9800 }, { "grad_norm": 0.6182960867881775, "learning_rate": 9.856541088727631e-05, "loss": 0.0596, "step": 9810 }, { "grad_norm": 0.3839779496192932, "learning_rate": 9.856049129913072e-05, "loss": 0.0583, "step": 9820 }, { "grad_norm": 0.5714422464370728, "learning_rate": 9.85555634133253e-05, "loss": 0.0662, "step": 9830 }, { "grad_norm": 0.30358070135116577, "learning_rate": 9.855062723070208e-05, "loss": 0.0713, "step": 9840 }, { "grad_norm": 0.4603899121284485, "learning_rate": 9.854568275210454e-05, "loss": 0.0641, "step": 9850 }, { "grad_norm": 0.3353356719017029, "learning_rate": 9.854072997837754e-05, "loss": 0.0495, "step": 9860 }, { "grad_norm": 0.46183839440345764, "learning_rate": 9.853576891036737e-05, "loss": 0.0505, "step": 9870 }, { "grad_norm": 0.35489097237586975, "learning_rate": 9.853079954892177e-05, "loss": 0.0541, "step": 9880 }, { "grad_norm": 0.4294109344482422, "learning_rate": 9.852582189488983e-05, "loss": 0.0478, "step": 9890 }, { "grad_norm": 0.42998623847961426, "learning_rate": 9.852083594912212e-05, "loss": 0.0576, "step": 9900 }, { "grad_norm": 0.4606285095214844, "learning_rate": 9.851584171247058e-05, "loss": 0.0596, "step": 9910 }, { "grad_norm": 0.46461987495422363, "learning_rate": 9.851083918578863e-05, "loss": 0.0638, "step": 9920 }, { "grad_norm": 0.5371437072753906, "learning_rate": 9.850582836993103e-05, "loss": 0.0626, "step": 9930 }, { "grad_norm": 0.39495211839675903, "learning_rate": 9.850080926575397e-05, "loss": 0.0596, "step": 9940 }, { "grad_norm": 0.46348708868026733, "learning_rate": 9.849578187411515e-05, "loss": 0.053, "step": 9950 }, { "grad_norm": 0.4862635135650635, "learning_rate": 9.849074619587354e-05, "loss": 0.0509, "step": 9960 }, { "grad_norm": 0.44562292098999023, "learning_rate": 9.848570223188964e-05, "loss": 0.0603, "step": 9970 }, { "grad_norm": 0.3885752558708191, "learning_rate": 9.848064998302531e-05, "loss": 0.0608, "step": 9980 }, { "grad_norm": 0.37116748094558716, "learning_rate": 9.847558945014386e-05, "loss": 0.0673, "step": 9990 }, { "grad_norm": 0.5641176104545593, "learning_rate": 9.847052063410996e-05, "loss": 0.0652, "step": 10000 }, { "grad_norm": 0.4751187860965729, "learning_rate": 9.846544353578977e-05, "loss": 0.0572, "step": 10010 }, { "grad_norm": 0.4770708978176117, "learning_rate": 9.846035815605081e-05, "loss": 0.0549, "step": 10020 }, { "grad_norm": 0.5084720253944397, "learning_rate": 9.845526449576204e-05, "loss": 0.0479, "step": 10030 }, { "grad_norm": 0.44815853238105774, "learning_rate": 9.845016255579383e-05, "loss": 0.065, "step": 10040 }, { "grad_norm": 0.39578181505203247, "learning_rate": 9.844505233701794e-05, "loss": 0.0595, "step": 10050 }, { "grad_norm": 0.3861648738384247, "learning_rate": 9.843993384030757e-05, "loss": 0.0616, "step": 10060 }, { "grad_norm": 0.4313238859176636, "learning_rate": 9.843480706653737e-05, "loss": 0.044, "step": 10070 }, { "grad_norm": 0.402051717042923, "learning_rate": 9.84296720165833e-05, "loss": 0.0542, "step": 10080 }, { "grad_norm": 0.31146085262298584, "learning_rate": 9.842452869132286e-05, "loss": 0.0538, "step": 10090 }, { "grad_norm": 0.5426959991455078, "learning_rate": 9.841937709163489e-05, "loss": 0.0552, "step": 10100 }, { "grad_norm": 0.32620906829833984, "learning_rate": 9.841421721839962e-05, "loss": 0.0408, "step": 10110 }, { "grad_norm": 0.3584690988063812, "learning_rate": 9.840904907249879e-05, "loss": 0.0484, "step": 10120 }, { "grad_norm": 0.5380610227584839, "learning_rate": 9.840387265481545e-05, "loss": 0.0638, "step": 10130 }, { "grad_norm": 0.45872828364372253, "learning_rate": 9.839868796623411e-05, "loss": 0.0534, "step": 10140 }, { "grad_norm": 0.5014376044273376, "learning_rate": 9.839349500764072e-05, "loss": 0.0655, "step": 10150 }, { "grad_norm": 0.505397617816925, "learning_rate": 9.83882937799226e-05, "loss": 0.0633, "step": 10160 }, { "grad_norm": 0.4063566327095032, "learning_rate": 9.838308428396849e-05, "loss": 0.0642, "step": 10170 }, { "grad_norm": 0.3756226599216461, "learning_rate": 9.837786652066854e-05, "loss": 0.0677, "step": 10180 }, { "grad_norm": 0.5517645478248596, "learning_rate": 9.837264049091437e-05, "loss": 0.086, "step": 10190 }, { "grad_norm": 0.41189447045326233, "learning_rate": 9.836740619559893e-05, "loss": 0.0594, "step": 10200 }, { "grad_norm": 0.4002956449985504, "learning_rate": 9.836216363561659e-05, "loss": 0.0583, "step": 10210 }, { "grad_norm": 0.48634690046310425, "learning_rate": 9.835691281186322e-05, "loss": 0.0593, "step": 10220 }, { "grad_norm": 0.2887404263019562, "learning_rate": 9.8351653725236e-05, "loss": 0.0644, "step": 10230 }, { "grad_norm": 0.39379122853279114, "learning_rate": 9.83463863766336e-05, "loss": 0.0551, "step": 10240 }, { "grad_norm": 0.36479777097702026, "learning_rate": 9.834111076695602e-05, "loss": 0.0615, "step": 10250 }, { "grad_norm": 0.48342975974082947, "learning_rate": 9.833582689710477e-05, "loss": 0.0503, "step": 10260 }, { "grad_norm": 0.5260495543479919, "learning_rate": 9.833053476798268e-05, "loss": 0.067, "step": 10270 }, { "grad_norm": 0.43544256687164307, "learning_rate": 9.832523438049404e-05, "loss": 0.0595, "step": 10280 }, { "grad_norm": 0.45477429032325745, "learning_rate": 9.831992573554454e-05, "loss": 0.0642, "step": 10290 }, { "grad_norm": 0.5157197117805481, "learning_rate": 9.831460883404128e-05, "loss": 0.0548, "step": 10300 }, { "grad_norm": 0.4957706928253174, "learning_rate": 9.830928367689278e-05, "loss": 0.0499, "step": 10310 }, { "grad_norm": 0.5947393774986267, "learning_rate": 9.830395026500896e-05, "loss": 0.0629, "step": 10320 }, { "grad_norm": 0.46407824754714966, "learning_rate": 9.829860859930115e-05, "loss": 0.0532, "step": 10330 }, { "grad_norm": 0.49252256751060486, "learning_rate": 9.829325868068212e-05, "loss": 0.0629, "step": 10340 }, { "grad_norm": 0.47916579246520996, "learning_rate": 9.8287900510066e-05, "loss": 0.0705, "step": 10350 }, { "grad_norm": 0.36580899357795715, "learning_rate": 9.828253408836834e-05, "loss": 0.0523, "step": 10360 }, { "grad_norm": 0.4259977638721466, "learning_rate": 9.827715941650615e-05, "loss": 0.0535, "step": 10370 }, { "grad_norm": 0.42342692613601685, "learning_rate": 9.82717764953978e-05, "loss": 0.0509, "step": 10380 }, { "grad_norm": 0.36418184638023376, "learning_rate": 9.826638532596308e-05, "loss": 0.059, "step": 10390 }, { "grad_norm": 0.49507808685302734, "learning_rate": 9.82609859091232e-05, "loss": 0.0594, "step": 10400 }, { "grad_norm": 0.3593519330024719, "learning_rate": 9.825557824580076e-05, "loss": 0.0525, "step": 10410 }, { "grad_norm": 0.29485583305358887, "learning_rate": 9.82501623369198e-05, "loss": 0.0601, "step": 10420 }, { "grad_norm": 0.44923654198646545, "learning_rate": 9.824473818340574e-05, "loss": 0.0521, "step": 10430 }, { "grad_norm": 0.39667847752571106, "learning_rate": 9.823930578618541e-05, "loss": 0.0516, "step": 10440 }, { "grad_norm": 0.547106146812439, "learning_rate": 9.823386514618709e-05, "loss": 0.0641, "step": 10450 }, { "grad_norm": 0.3933587372303009, "learning_rate": 9.82284162643404e-05, "loss": 0.0509, "step": 10460 }, { "grad_norm": 0.3514941334724426, "learning_rate": 9.822295914157642e-05, "loss": 0.0508, "step": 10470 }, { "grad_norm": 0.4827096164226532, "learning_rate": 9.821749377882763e-05, "loss": 0.0536, "step": 10480 }, { "grad_norm": 0.35368067026138306, "learning_rate": 9.821202017702791e-05, "loss": 0.0499, "step": 10490 }, { "grad_norm": 0.48578619956970215, "learning_rate": 9.820653833711253e-05, "loss": 0.0479, "step": 10500 }, { "grad_norm": 0.5305749177932739, "learning_rate": 9.820104826001822e-05, "loss": 0.0658, "step": 10510 }, { "grad_norm": 0.3814452886581421, "learning_rate": 9.819554994668305e-05, "loss": 0.0478, "step": 10520 }, { "grad_norm": 0.3398468792438507, "learning_rate": 9.819004339804654e-05, "loss": 0.0545, "step": 10530 }, { "grad_norm": 0.3820374608039856, "learning_rate": 9.818452861504961e-05, "loss": 0.0504, "step": 10540 }, { "grad_norm": 0.45631274580955505, "learning_rate": 9.81790055986346e-05, "loss": 0.0489, "step": 10550 }, { "grad_norm": 0.4260309040546417, "learning_rate": 9.817347434974523e-05, "loss": 0.0499, "step": 10560 }, { "grad_norm": 0.49264460802078247, "learning_rate": 9.816793486932664e-05, "loss": 0.0508, "step": 10570 }, { "grad_norm": 0.45166561007499695, "learning_rate": 9.816238715832538e-05, "loss": 0.0568, "step": 10580 }, { "grad_norm": 0.6123629212379456, "learning_rate": 9.815683121768939e-05, "loss": 0.0608, "step": 10590 }, { "grad_norm": 0.4158664643764496, "learning_rate": 9.815126704836804e-05, "loss": 0.0674, "step": 10600 }, { "grad_norm": 0.3846834599971771, "learning_rate": 9.81456946513121e-05, "loss": 0.0711, "step": 10610 }, { "grad_norm": 0.6076033115386963, "learning_rate": 9.814011402747373e-05, "loss": 0.0727, "step": 10620 }, { "grad_norm": 0.3267817795276642, "learning_rate": 9.813452517780651e-05, "loss": 0.051, "step": 10630 }, { "grad_norm": 0.3991529047489166, "learning_rate": 9.81289281032654e-05, "loss": 0.066, "step": 10640 }, { "grad_norm": 0.4951857626438141, "learning_rate": 9.812332280480683e-05, "loss": 0.0596, "step": 10650 }, { "grad_norm": 0.5243456363677979, "learning_rate": 9.811770928338854e-05, "loss": 0.0635, "step": 10660 }, { "grad_norm": 0.4668377637863159, "learning_rate": 9.811208753996979e-05, "loss": 0.0587, "step": 10670 }, { "grad_norm": 0.36280956864356995, "learning_rate": 9.810645757551113e-05, "loss": 0.0432, "step": 10680 }, { "grad_norm": 0.31785961985588074, "learning_rate": 9.810081939097459e-05, "loss": 0.0497, "step": 10690 }, { "grad_norm": 0.5240728855133057, "learning_rate": 9.809517298732356e-05, "loss": 0.0667, "step": 10700 }, { "grad_norm": 0.3580385446548462, "learning_rate": 9.80895183655229e-05, "loss": 0.0527, "step": 10710 }, { "grad_norm": 0.37145039439201355, "learning_rate": 9.808385552653877e-05, "loss": 0.0535, "step": 10720 }, { "grad_norm": 0.4058728814125061, "learning_rate": 9.807818447133886e-05, "loss": 0.0447, "step": 10730 }, { "grad_norm": 0.4557889699935913, "learning_rate": 9.807250520089215e-05, "loss": 0.0495, "step": 10740 }, { "grad_norm": 0.4143455922603607, "learning_rate": 9.806681771616908e-05, "loss": 0.0558, "step": 10750 }, { "grad_norm": 0.3459220230579376, "learning_rate": 9.80611220181415e-05, "loss": 0.0561, "step": 10760 }, { "grad_norm": 0.517833411693573, "learning_rate": 9.805541810778264e-05, "loss": 0.0575, "step": 10770 }, { "grad_norm": 0.625338613986969, "learning_rate": 9.804970598606716e-05, "loss": 0.0557, "step": 10780 }, { "grad_norm": 0.43566834926605225, "learning_rate": 9.804398565397106e-05, "loss": 0.0424, "step": 10790 }, { "grad_norm": 0.38331151008605957, "learning_rate": 9.803825711247183e-05, "loss": 0.0613, "step": 10800 }, { "grad_norm": 0.5012357831001282, "learning_rate": 9.803252036254831e-05, "loss": 0.0482, "step": 10810 }, { "grad_norm": 0.41896000504493713, "learning_rate": 9.802677540518076e-05, "loss": 0.0622, "step": 10820 }, { "grad_norm": 0.29345741868019104, "learning_rate": 9.802102224135081e-05, "loss": 0.0534, "step": 10830 }, { "grad_norm": 0.4123251140117645, "learning_rate": 9.801526087204155e-05, "loss": 0.0577, "step": 10840 }, { "grad_norm": 0.5099521279335022, "learning_rate": 9.800949129823743e-05, "loss": 0.0566, "step": 10850 }, { "grad_norm": 0.3923605680465698, "learning_rate": 9.80037135209243e-05, "loss": 0.0487, "step": 10860 }, { "grad_norm": 0.3620224893093109, "learning_rate": 9.799792754108946e-05, "loss": 0.0492, "step": 10870 }, { "grad_norm": 0.402283251285553, "learning_rate": 9.799213335972152e-05, "loss": 0.054, "step": 10880 }, { "grad_norm": 0.3544284701347351, "learning_rate": 9.798633097781058e-05, "loss": 0.0556, "step": 10890 }, { "grad_norm": 0.4319859445095062, "learning_rate": 9.79805203963481e-05, "loss": 0.0554, "step": 10900 }, { "grad_norm": 0.3889198303222656, "learning_rate": 9.797470161632697e-05, "loss": 0.0567, "step": 10910 }, { "grad_norm": 0.5046876072883606, "learning_rate": 9.796887463874145e-05, "loss": 0.0575, "step": 10920 }, { "grad_norm": 0.3703787326812744, "learning_rate": 9.796303946458718e-05, "loss": 0.0698, "step": 10930 }, { "grad_norm": 0.3974722623825073, "learning_rate": 9.795719609486127e-05, "loss": 0.0616, "step": 10940 }, { "grad_norm": 0.47873780131340027, "learning_rate": 9.795134453056219e-05, "loss": 0.0565, "step": 10950 }, { "grad_norm": 0.4170961081981659, "learning_rate": 9.794548477268979e-05, "loss": 0.0494, "step": 10960 }, { "grad_norm": 0.4438859820365906, "learning_rate": 9.793961682224537e-05, "loss": 0.0484, "step": 10970 }, { "grad_norm": 0.42010512948036194, "learning_rate": 9.793374068023156e-05, "loss": 0.0525, "step": 10980 }, { "grad_norm": 0.4260512888431549, "learning_rate": 9.792785634765247e-05, "loss": 0.0566, "step": 10990 }, { "grad_norm": 0.4151779115200043, "learning_rate": 9.792196382551357e-05, "loss": 0.0536, "step": 11000 }, { "grad_norm": 0.39805591106414795, "learning_rate": 9.791606311482171e-05, "loss": 0.0603, "step": 11010 }, { "grad_norm": 0.3375794291496277, "learning_rate": 9.791015421658518e-05, "loss": 0.0534, "step": 11020 }, { "grad_norm": 0.4492846429347992, "learning_rate": 9.790423713181362e-05, "loss": 0.0569, "step": 11030 }, { "grad_norm": 0.41649293899536133, "learning_rate": 9.789831186151814e-05, "loss": 0.0514, "step": 11040 }, { "grad_norm": 0.4356219172477722, "learning_rate": 9.789237840671118e-05, "loss": 0.0444, "step": 11050 }, { "grad_norm": 0.368782103061676, "learning_rate": 9.78864367684066e-05, "loss": 0.061, "step": 11060 }, { "grad_norm": 0.45823583006858826, "learning_rate": 9.788048694761968e-05, "loss": 0.0523, "step": 11070 }, { "grad_norm": 0.43339163064956665, "learning_rate": 9.787452894536709e-05, "loss": 0.0518, "step": 11080 }, { "grad_norm": 0.44091081619262695, "learning_rate": 9.786856276266685e-05, "loss": 0.0521, "step": 11090 }, { "grad_norm": 0.5203498601913452, "learning_rate": 9.786258840053845e-05, "loss": 0.0554, "step": 11100 }, { "grad_norm": 0.43340960144996643, "learning_rate": 9.785660586000273e-05, "loss": 0.0592, "step": 11110 }, { "grad_norm": 0.2581512928009033, "learning_rate": 9.785061514208196e-05, "loss": 0.0507, "step": 11120 }, { "grad_norm": 0.3394492268562317, "learning_rate": 9.784461624779977e-05, "loss": 0.058, "step": 11130 }, { "grad_norm": 0.435917466878891, "learning_rate": 9.783860917818123e-05, "loss": 0.0547, "step": 11140 }, { "grad_norm": 0.38876137137413025, "learning_rate": 9.783259393425277e-05, "loss": 0.0467, "step": 11150 }, { "grad_norm": 0.5615711212158203, "learning_rate": 9.782657051704221e-05, "loss": 0.0524, "step": 11160 }, { "grad_norm": 0.4207378625869751, "learning_rate": 9.782053892757883e-05, "loss": 0.0509, "step": 11170 }, { "grad_norm": 0.4981745481491089, "learning_rate": 9.781449916689324e-05, "loss": 0.0556, "step": 11180 }, { "grad_norm": 0.47185641527175903, "learning_rate": 9.780845123601746e-05, "loss": 0.0632, "step": 11190 }, { "grad_norm": 0.4129422605037689, "learning_rate": 9.780239513598492e-05, "loss": 0.0591, "step": 11200 }, { "grad_norm": 0.4809286594390869, "learning_rate": 9.779633086783047e-05, "loss": 0.0585, "step": 11210 }, { "grad_norm": 0.2983725666999817, "learning_rate": 9.779025843259031e-05, "loss": 0.0567, "step": 11220 }, { "grad_norm": 0.4982427656650543, "learning_rate": 9.778417783130204e-05, "loss": 0.0693, "step": 11230 }, { "grad_norm": 0.4564981758594513, "learning_rate": 9.777808906500468e-05, "loss": 0.0613, "step": 11240 }, { "grad_norm": 0.44986167550086975, "learning_rate": 9.777199213473862e-05, "loss": 0.0471, "step": 11250 }, { "grad_norm": 0.41396594047546387, "learning_rate": 9.77658870415457e-05, "loss": 0.0471, "step": 11260 }, { "grad_norm": 0.3870123326778412, "learning_rate": 9.775977378646906e-05, "loss": 0.0533, "step": 11270 }, { "grad_norm": 0.5347506403923035, "learning_rate": 9.775365237055331e-05, "loss": 0.0673, "step": 11280 }, { "grad_norm": 0.4133090376853943, "learning_rate": 9.774752279484445e-05, "loss": 0.0609, "step": 11290 }, { "grad_norm": 0.36520135402679443, "learning_rate": 9.774138506038984e-05, "loss": 0.0528, "step": 11300 }, { "grad_norm": 0.4887140989303589, "learning_rate": 9.773523916823826e-05, "loss": 0.0563, "step": 11310 }, { "grad_norm": 0.4354543685913086, "learning_rate": 9.772908511943986e-05, "loss": 0.06, "step": 11320 }, { "grad_norm": 0.4253315329551697, "learning_rate": 9.77229229150462e-05, "loss": 0.0515, "step": 11330 }, { "grad_norm": 0.5269914269447327, "learning_rate": 9.771675255611024e-05, "loss": 0.0583, "step": 11340 }, { "grad_norm": 0.3863091766834259, "learning_rate": 9.771057404368632e-05, "loss": 0.0723, "step": 11350 }, { "grad_norm": 0.38139981031417847, "learning_rate": 9.770438737883018e-05, "loss": 0.0493, "step": 11360 }, { "grad_norm": 0.382118284702301, "learning_rate": 9.769819256259898e-05, "loss": 0.0436, "step": 11370 }, { "grad_norm": 0.4704022705554962, "learning_rate": 9.769198959605119e-05, "loss": 0.0502, "step": 11380 }, { "grad_norm": 0.4023139476776123, "learning_rate": 9.768577848024678e-05, "loss": 0.0606, "step": 11390 }, { "grad_norm": 0.32199645042419434, "learning_rate": 9.767955921624702e-05, "loss": 0.061, "step": 11400 }, { "grad_norm": 0.5573130249977112, "learning_rate": 9.767333180511465e-05, "loss": 0.0509, "step": 11410 }, { "grad_norm": 0.41511887311935425, "learning_rate": 9.766709624791373e-05, "loss": 0.0737, "step": 11420 }, { "grad_norm": 0.338253378868103, "learning_rate": 9.766085254570975e-05, "loss": 0.0528, "step": 11430 }, { "grad_norm": 0.33362752199172974, "learning_rate": 9.76546006995696e-05, "loss": 0.0511, "step": 11440 }, { "grad_norm": 0.5915864109992981, "learning_rate": 9.764834071056155e-05, "loss": 0.0541, "step": 11450 }, { "grad_norm": 0.3982852101325989, "learning_rate": 9.764207257975526e-05, "loss": 0.0581, "step": 11460 }, { "grad_norm": 0.6179614067077637, "learning_rate": 9.763579630822179e-05, "loss": 0.051, "step": 11470 }, { "grad_norm": 0.5472491979598999, "learning_rate": 9.762951189703356e-05, "loss": 0.0555, "step": 11480 }, { "grad_norm": 0.4021734893321991, "learning_rate": 9.762321934726442e-05, "loss": 0.0553, "step": 11490 }, { "grad_norm": 0.41894689202308655, "learning_rate": 9.761691865998959e-05, "loss": 0.0525, "step": 11500 }, { "grad_norm": 0.44521433115005493, "learning_rate": 9.76106098362857e-05, "loss": 0.0471, "step": 11510 }, { "grad_norm": 0.48150479793548584, "learning_rate": 9.760429287723072e-05, "loss": 0.0497, "step": 11520 }, { "grad_norm": 0.4337771534919739, "learning_rate": 9.759796778390406e-05, "loss": 0.0509, "step": 11530 }, { "grad_norm": 0.41955989599227905, "learning_rate": 9.759163455738653e-05, "loss": 0.0579, "step": 11540 }, { "grad_norm": 0.38667476177215576, "learning_rate": 9.75852931987603e-05, "loss": 0.0623, "step": 11550 }, { "grad_norm": 0.45622333884239197, "learning_rate": 9.757894370910891e-05, "loss": 0.0504, "step": 11560 }, { "grad_norm": 0.4149587154388428, "learning_rate": 9.757258608951733e-05, "loss": 0.0452, "step": 11570 }, { "grad_norm": 0.5297603607177734, "learning_rate": 9.75662203410719e-05, "loss": 0.062, "step": 11580 }, { "grad_norm": 0.4180716574192047, "learning_rate": 9.755984646486034e-05, "loss": 0.0547, "step": 11590 }, { "grad_norm": 0.43595725297927856, "learning_rate": 9.75534644619718e-05, "loss": 0.0483, "step": 11600 }, { "grad_norm": 0.504188597202301, "learning_rate": 9.754707433349676e-05, "loss": 0.0612, "step": 11610 }, { "grad_norm": 0.5628002882003784, "learning_rate": 9.754067608052715e-05, "loss": 0.0664, "step": 11620 }, { "grad_norm": 0.5041863322257996, "learning_rate": 9.753426970415622e-05, "loss": 0.0725, "step": 11630 }, { "grad_norm": 0.40071845054626465, "learning_rate": 9.752785520547868e-05, "loss": 0.048, "step": 11640 }, { "grad_norm": 0.39181390404701233, "learning_rate": 9.752143258559056e-05, "loss": 0.0522, "step": 11650 }, { "grad_norm": 0.4500504732131958, "learning_rate": 9.751500184558933e-05, "loss": 0.049, "step": 11660 }, { "grad_norm": 0.358595073223114, "learning_rate": 9.750856298657383e-05, "loss": 0.0499, "step": 11670 }, { "grad_norm": 0.42780405282974243, "learning_rate": 9.750211600964428e-05, "loss": 0.0612, "step": 11680 }, { "grad_norm": 0.526384174823761, "learning_rate": 9.749566091590226e-05, "loss": 0.0493, "step": 11690 }, { "grad_norm": 0.49467167258262634, "learning_rate": 9.748919770645083e-05, "loss": 0.0587, "step": 11700 }, { "grad_norm": 0.4926881194114685, "learning_rate": 9.748272638239432e-05, "loss": 0.0561, "step": 11710 }, { "grad_norm": 0.5607455372810364, "learning_rate": 9.747624694483855e-05, "loss": 0.0597, "step": 11720 }, { "grad_norm": 0.3787783682346344, "learning_rate": 9.746975939489065e-05, "loss": 0.0479, "step": 11730 }, { "grad_norm": 0.41074538230895996, "learning_rate": 9.746326373365918e-05, "loss": 0.0494, "step": 11740 }, { "grad_norm": 0.5132369995117188, "learning_rate": 9.745675996225403e-05, "loss": 0.0593, "step": 11750 }, { "grad_norm": 0.3881279528141022, "learning_rate": 9.745024808178657e-05, "loss": 0.0466, "step": 11760 }, { "grad_norm": 0.36282071471214294, "learning_rate": 9.744372809336947e-05, "loss": 0.0574, "step": 11770 }, { "grad_norm": 0.5786557793617249, "learning_rate": 9.743719999811682e-05, "loss": 0.0536, "step": 11780 }, { "grad_norm": 0.4061337411403656, "learning_rate": 9.743066379714412e-05, "loss": 0.0553, "step": 11790 }, { "grad_norm": 0.3574906885623932, "learning_rate": 9.74241194915682e-05, "loss": 0.0677, "step": 11800 }, { "grad_norm": 0.5040960907936096, "learning_rate": 9.741756708250731e-05, "loss": 0.0714, "step": 11810 }, { "grad_norm": 0.5296720266342163, "learning_rate": 9.741100657108109e-05, "loss": 0.0594, "step": 11820 }, { "grad_norm": 0.46749138832092285, "learning_rate": 9.740443795841054e-05, "loss": 0.0677, "step": 11830 }, { "grad_norm": 0.33205875754356384, "learning_rate": 9.739786124561805e-05, "loss": 0.0496, "step": 11840 }, { "grad_norm": 0.3939608633518219, "learning_rate": 9.73912764338274e-05, "loss": 0.0551, "step": 11850 }, { "grad_norm": 0.43862563371658325, "learning_rate": 9.738468352416377e-05, "loss": 0.0561, "step": 11860 }, { "grad_norm": 0.4175872504711151, "learning_rate": 9.737808251775369e-05, "loss": 0.0501, "step": 11870 }, { "grad_norm": 0.2807920277118683, "learning_rate": 9.737147341572512e-05, "loss": 0.0661, "step": 11880 }, { "grad_norm": 0.44273754954338074, "learning_rate": 9.736485621920735e-05, "loss": 0.0526, "step": 11890 }, { "grad_norm": 0.5374293327331543, "learning_rate": 9.735823092933108e-05, "loss": 0.0553, "step": 11900 }, { "grad_norm": 0.37572231888771057, "learning_rate": 9.735159754722838e-05, "loss": 0.0563, "step": 11910 }, { "grad_norm": 0.3700706660747528, "learning_rate": 9.734495607403275e-05, "loss": 0.0457, "step": 11920 }, { "grad_norm": 0.39291462302207947, "learning_rate": 9.733830651087901e-05, "loss": 0.0624, "step": 11930 }, { "grad_norm": 0.37224680185317993, "learning_rate": 9.733164885890338e-05, "loss": 0.0464, "step": 11940 }, { "grad_norm": 0.4431273937225342, "learning_rate": 9.732498311924349e-05, "loss": 0.0448, "step": 11950 }, { "grad_norm": 0.6007319092750549, "learning_rate": 9.731830929303833e-05, "loss": 0.0504, "step": 11960 }, { "grad_norm": 0.4150267243385315, "learning_rate": 9.731162738142827e-05, "loss": 0.0492, "step": 11970 }, { "grad_norm": 0.550639271736145, "learning_rate": 9.730493738555506e-05, "loss": 0.0565, "step": 11980 }, { "grad_norm": 0.3644391596317291, "learning_rate": 9.729823930656186e-05, "loss": 0.0599, "step": 11990 }, { "grad_norm": 0.3797757923603058, "learning_rate": 9.729153314559316e-05, "loss": 0.0653, "step": 12000 }, { "grad_norm": 0.5106303095817566, "learning_rate": 9.728481890379486e-05, "loss": 0.0628, "step": 12010 }, { "grad_norm": 0.3625558912754059, "learning_rate": 9.727809658231428e-05, "loss": 0.0577, "step": 12020 }, { "grad_norm": 0.3309819996356964, "learning_rate": 9.727136618230003e-05, "loss": 0.0467, "step": 12030 }, { "grad_norm": 0.6955719590187073, "learning_rate": 9.726462770490219e-05, "loss": 0.0568, "step": 12040 }, { "grad_norm": 0.35353800654411316, "learning_rate": 9.725788115127214e-05, "loss": 0.0578, "step": 12050 }, { "grad_norm": 0.45401135087013245, "learning_rate": 9.725112652256274e-05, "loss": 0.0606, "step": 12060 }, { "grad_norm": 0.5995681881904602, "learning_rate": 9.724436381992812e-05, "loss": 0.0766, "step": 12070 }, { "grad_norm": 0.5340020656585693, "learning_rate": 9.723759304452387e-05, "loss": 0.0601, "step": 12080 }, { "grad_norm": 0.4388681948184967, "learning_rate": 9.72308141975069e-05, "loss": 0.0466, "step": 12090 }, { "grad_norm": 0.5645894408226013, "learning_rate": 9.722402728003557e-05, "loss": 0.0657, "step": 12100 }, { "grad_norm": 0.539197564125061, "learning_rate": 9.721723229326953e-05, "loss": 0.0509, "step": 12110 }, { "grad_norm": 0.5114811658859253, "learning_rate": 9.721042923836992e-05, "loss": 0.0486, "step": 12120 }, { "grad_norm": 0.3472784757614136, "learning_rate": 9.720361811649914e-05, "loss": 0.0542, "step": 12130 }, { "grad_norm": 0.6010377407073975, "learning_rate": 9.719679892882106e-05, "loss": 0.0543, "step": 12140 }, { "grad_norm": 0.42782267928123474, "learning_rate": 9.718997167650085e-05, "loss": 0.0502, "step": 12150 }, { "grad_norm": 0.3958961069583893, "learning_rate": 9.718313636070515e-05, "loss": 0.0444, "step": 12160 }, { "grad_norm": 0.3391648232936859, "learning_rate": 9.717629298260192e-05, "loss": 0.0478, "step": 12170 }, { "grad_norm": 0.5416630506515503, "learning_rate": 9.716944154336047e-05, "loss": 0.0627, "step": 12180 }, { "grad_norm": 0.6059003472328186, "learning_rate": 9.716258204415157e-05, "loss": 0.0501, "step": 12190 }, { "grad_norm": 0.43234023451805115, "learning_rate": 9.715571448614728e-05, "loss": 0.058, "step": 12200 }, { "grad_norm": 0.37056875228881836, "learning_rate": 9.71488388705211e-05, "loss": 0.0695, "step": 12210 }, { "grad_norm": 0.4082079231739044, "learning_rate": 9.714195519844788e-05, "loss": 0.051, "step": 12220 }, { "grad_norm": 0.4176998436450958, "learning_rate": 9.713506347110386e-05, "loss": 0.0566, "step": 12230 }, { "grad_norm": 0.3377353549003601, "learning_rate": 9.712816368966662e-05, "loss": 0.0557, "step": 12240 }, { "grad_norm": 0.3956582844257355, "learning_rate": 9.712125585531517e-05, "loss": 0.0663, "step": 12250 }, { "grad_norm": 0.47129589319229126, "learning_rate": 9.711433996922988e-05, "loss": 0.0666, "step": 12260 }, { "grad_norm": 0.4851278066635132, "learning_rate": 9.710741603259245e-05, "loss": 0.051, "step": 12270 }, { "grad_norm": 0.4341801702976227, "learning_rate": 9.710048404658603e-05, "loss": 0.053, "step": 12280 }, { "grad_norm": 0.32304689288139343, "learning_rate": 9.709354401239508e-05, "loss": 0.0549, "step": 12290 }, { "grad_norm": 0.5723078846931458, "learning_rate": 9.708659593120546e-05, "loss": 0.0478, "step": 12300 }, { "grad_norm": 0.5800574421882629, "learning_rate": 9.707963980420443e-05, "loss": 0.0554, "step": 12310 }, { "grad_norm": 0.5850123763084412, "learning_rate": 9.707267563258058e-05, "loss": 0.0524, "step": 12320 }, { "grad_norm": 0.4016854465007782, "learning_rate": 9.70657034175239e-05, "loss": 0.0494, "step": 12330 }, { "grad_norm": 0.3682669997215271, "learning_rate": 9.705872316022577e-05, "loss": 0.0585, "step": 12340 }, { "grad_norm": 0.33294740319252014, "learning_rate": 9.705173486187891e-05, "loss": 0.0545, "step": 12350 }, { "grad_norm": 0.3587924838066101, "learning_rate": 9.704473852367741e-05, "loss": 0.0569, "step": 12360 }, { "grad_norm": 0.3734003007411957, "learning_rate": 9.70377341468168e-05, "loss": 0.0513, "step": 12370 }, { "grad_norm": 0.39112064242362976, "learning_rate": 9.703072173249389e-05, "loss": 0.0482, "step": 12380 }, { "grad_norm": 0.38721996545791626, "learning_rate": 9.702370128190693e-05, "loss": 0.0594, "step": 12390 }, { "grad_norm": 0.4615304470062256, "learning_rate": 9.701667279625552e-05, "loss": 0.0548, "step": 12400 }, { "grad_norm": 0.43418386578559875, "learning_rate": 9.700963627674065e-05, "loss": 0.0561, "step": 12410 }, { "grad_norm": 0.3895587921142578, "learning_rate": 9.700259172456466e-05, "loss": 0.0481, "step": 12420 }, { "grad_norm": 0.5697787404060364, "learning_rate": 9.699553914093124e-05, "loss": 0.0654, "step": 12430 }, { "grad_norm": 0.3490683138370514, "learning_rate": 9.698847852704553e-05, "loss": 0.059, "step": 12440 }, { "grad_norm": 0.37184515595436096, "learning_rate": 9.6981409884114e-05, "loss": 0.044, "step": 12450 }, { "grad_norm": 0.37314918637275696, "learning_rate": 9.697433321334443e-05, "loss": 0.0613, "step": 12460 }, { "grad_norm": 0.3511878252029419, "learning_rate": 9.696724851594607e-05, "loss": 0.0469, "step": 12470 }, { "grad_norm": 0.4524681568145752, "learning_rate": 9.696015579312952e-05, "loss": 0.0396, "step": 12480 }, { "grad_norm": 0.486029714345932, "learning_rate": 9.695305504610668e-05, "loss": 0.0559, "step": 12490 }, { "grad_norm": 0.3373386561870575, "learning_rate": 9.694594627609092e-05, "loss": 0.0513, "step": 12500 }, { "grad_norm": 0.4608883261680603, "learning_rate": 9.693882948429691e-05, "loss": 0.0589, "step": 12510 }, { "grad_norm": 0.5590483546257019, "learning_rate": 9.693170467194071e-05, "loss": 0.0502, "step": 12520 }, { "grad_norm": 0.33597615361213684, "learning_rate": 9.692457184023977e-05, "loss": 0.0504, "step": 12530 }, { "grad_norm": 0.44897031784057617, "learning_rate": 9.691743099041291e-05, "loss": 0.0595, "step": 12540 }, { "grad_norm": 0.47105565667152405, "learning_rate": 9.691028212368027e-05, "loss": 0.0529, "step": 12550 }, { "grad_norm": 0.5152426958084106, "learning_rate": 9.690312524126342e-05, "loss": 0.0459, "step": 12560 }, { "grad_norm": 0.3278726041316986, "learning_rate": 9.689596034438527e-05, "loss": 0.0479, "step": 12570 }, { "grad_norm": 0.2648175060749054, "learning_rate": 9.688878743427012e-05, "loss": 0.0426, "step": 12580 }, { "grad_norm": 0.4250586926937103, "learning_rate": 9.688160651214359e-05, "loss": 0.0588, "step": 12590 }, { "grad_norm": 0.41975754499435425, "learning_rate": 9.687441757923273e-05, "loss": 0.0548, "step": 12600 }, { "grad_norm": 0.4461625814437866, "learning_rate": 9.68672206367659e-05, "loss": 0.0509, "step": 12610 }, { "grad_norm": 1.02378511428833, "learning_rate": 9.686001568597291e-05, "loss": 0.0503, "step": 12620 }, { "grad_norm": 0.4283071458339691, "learning_rate": 9.685280272808486e-05, "loss": 0.0451, "step": 12630 }, { "grad_norm": 0.4863217771053314, "learning_rate": 9.684558176433424e-05, "loss": 0.0622, "step": 12640 }, { "grad_norm": 0.32946816086769104, "learning_rate": 9.683835279595495e-05, "loss": 0.0528, "step": 12650 }, { "grad_norm": 0.3540911376476288, "learning_rate": 9.683111582418216e-05, "loss": 0.0547, "step": 12660 }, { "grad_norm": 0.45451727509498596, "learning_rate": 9.682387085025254e-05, "loss": 0.0483, "step": 12670 }, { "grad_norm": 0.3120499551296234, "learning_rate": 9.681661787540401e-05, "loss": 0.0543, "step": 12680 }, { "grad_norm": 0.3868289291858673, "learning_rate": 9.680935690087593e-05, "loss": 0.0524, "step": 12690 }, { "grad_norm": 0.45504817366600037, "learning_rate": 9.680208792790901e-05, "loss": 0.0472, "step": 12700 }, { "grad_norm": 0.41042348742485046, "learning_rate": 9.679481095774529e-05, "loss": 0.0514, "step": 12710 }, { "grad_norm": 0.4944710433483124, "learning_rate": 9.678752599162822e-05, "loss": 0.0494, "step": 12720 }, { "grad_norm": 0.4920808672904968, "learning_rate": 9.678023303080259e-05, "loss": 0.0498, "step": 12730 }, { "grad_norm": 0.4703144431114197, "learning_rate": 9.677293207651459e-05, "loss": 0.054, "step": 12740 }, { "grad_norm": 0.4269450008869171, "learning_rate": 9.676562313001173e-05, "loss": 0.0423, "step": 12750 }, { "grad_norm": 0.5220159888267517, "learning_rate": 9.675830619254293e-05, "loss": 0.0655, "step": 12760 }, { "grad_norm": 0.4033772051334381, "learning_rate": 9.675098126535843e-05, "loss": 0.0605, "step": 12770 }, { "grad_norm": 0.7506333589553833, "learning_rate": 9.674364834970988e-05, "loss": 0.0581, "step": 12780 }, { "grad_norm": 0.4681108891963959, "learning_rate": 9.673630744685028e-05, "loss": 0.0616, "step": 12790 }, { "grad_norm": 0.45256727933883667, "learning_rate": 9.672895855803397e-05, "loss": 0.0585, "step": 12800 }, { "grad_norm": 0.36609089374542236, "learning_rate": 9.672160168451667e-05, "loss": 0.0571, "step": 12810 }, { "grad_norm": 0.3918741047382355, "learning_rate": 9.671423682755549e-05, "loss": 0.0688, "step": 12820 }, { "grad_norm": 0.46430838108062744, "learning_rate": 9.670686398840888e-05, "loss": 0.0506, "step": 12830 }, { "grad_norm": 0.4042271673679352, "learning_rate": 9.669948316833664e-05, "loss": 0.0505, "step": 12840 }, { "grad_norm": 0.535817563533783, "learning_rate": 9.669209436859997e-05, "loss": 0.046, "step": 12850 }, { "grad_norm": 0.5485854744911194, "learning_rate": 9.66846975904614e-05, "loss": 0.0515, "step": 12860 }, { "grad_norm": 0.5250694751739502, "learning_rate": 9.667729283518483e-05, "loss": 0.0614, "step": 12870 }, { "grad_norm": 0.47662341594696045, "learning_rate": 9.666988010403557e-05, "loss": 0.0469, "step": 12880 }, { "grad_norm": 0.4304984211921692, "learning_rate": 9.66624593982802e-05, "loss": 0.0542, "step": 12890 }, { "grad_norm": 0.45833736658096313, "learning_rate": 9.665503071918675e-05, "loss": 0.0602, "step": 12900 }, { "grad_norm": 0.34906119108200073, "learning_rate": 9.664759406802456e-05, "loss": 0.044, "step": 12910 }, { "grad_norm": 0.3858863413333893, "learning_rate": 9.664014944606437e-05, "loss": 0.0414, "step": 12920 }, { "grad_norm": 0.406013160943985, "learning_rate": 9.663269685457822e-05, "loss": 0.0467, "step": 12930 }, { "grad_norm": 0.4640014171600342, "learning_rate": 9.662523629483962e-05, "loss": 0.0531, "step": 12940 }, { "grad_norm": 0.36297667026519775, "learning_rate": 9.661776776812333e-05, "loss": 0.043, "step": 12950 }, { "grad_norm": 0.5010892152786255, "learning_rate": 9.661029127570553e-05, "loss": 0.0456, "step": 12960 }, { "grad_norm": 0.47536858916282654, "learning_rate": 9.660280681886373e-05, "loss": 0.0466, "step": 12970 }, { "grad_norm": 0.3523949980735779, "learning_rate": 9.659531439887685e-05, "loss": 0.0447, "step": 12980 }, { "grad_norm": 0.3876064419746399, "learning_rate": 9.658781401702511e-05, "loss": 0.0468, "step": 12990 }, { "grad_norm": 0.43853360414505005, "learning_rate": 9.658030567459015e-05, "loss": 0.0493, "step": 13000 }, { "grad_norm": 0.40964168310165405, "learning_rate": 9.65727893728549e-05, "loss": 0.0648, "step": 13010 }, { "grad_norm": 0.49323898553848267, "learning_rate": 9.656526511310375e-05, "loss": 0.0582, "step": 13020 }, { "grad_norm": 0.4800907373428345, "learning_rate": 9.655773289662233e-05, "loss": 0.062, "step": 13030 }, { "grad_norm": 0.3703405261039734, "learning_rate": 9.655019272469772e-05, "loss": 0.0422, "step": 13040 }, { "grad_norm": 0.41318279504776, "learning_rate": 9.654264459861832e-05, "loss": 0.0477, "step": 13050 }, { "grad_norm": 0.43725356459617615, "learning_rate": 9.653508851967391e-05, "loss": 0.0458, "step": 13060 }, { "grad_norm": 0.4591667950153351, "learning_rate": 9.65275244891556e-05, "loss": 0.0636, "step": 13070 }, { "grad_norm": 0.49103495478630066, "learning_rate": 9.651995250835591e-05, "loss": 0.0473, "step": 13080 }, { "grad_norm": 0.43925589323043823, "learning_rate": 9.651237257856862e-05, "loss": 0.0506, "step": 13090 }, { "grad_norm": 0.4364607036113739, "learning_rate": 9.6504784701089e-05, "loss": 0.0514, "step": 13100 }, { "grad_norm": 0.39672595262527466, "learning_rate": 9.649718887721357e-05, "loss": 0.0422, "step": 13110 }, { "grad_norm": 0.40537571907043457, "learning_rate": 9.648958510824028e-05, "loss": 0.0469, "step": 13120 }, { "grad_norm": 0.39295223355293274, "learning_rate": 9.648197339546837e-05, "loss": 0.0631, "step": 13130 }, { "grad_norm": 0.49244382977485657, "learning_rate": 9.647435374019851e-05, "loss": 0.0569, "step": 13140 }, { "grad_norm": 0.44925445318222046, "learning_rate": 9.646672614373266e-05, "loss": 0.0553, "step": 13150 }, { "grad_norm": 0.368744820356369, "learning_rate": 9.645909060737418e-05, "loss": 0.0409, "step": 13160 }, { "grad_norm": 0.4982262849807739, "learning_rate": 9.645144713242778e-05, "loss": 0.0493, "step": 13170 }, { "grad_norm": 0.36018484830856323, "learning_rate": 9.64437957201995e-05, "loss": 0.0493, "step": 13180 }, { "grad_norm": 0.31433358788490295, "learning_rate": 9.643613637199678e-05, "loss": 0.04, "step": 13190 }, { "grad_norm": 0.42270970344543457, "learning_rate": 9.642846908912839e-05, "loss": 0.0514, "step": 13200 }, { "grad_norm": 0.5491811633110046, "learning_rate": 9.642079387290444e-05, "loss": 0.0383, "step": 13210 }, { "grad_norm": 0.3911082148551941, "learning_rate": 9.641311072463644e-05, "loss": 0.0449, "step": 13220 }, { "grad_norm": 0.5788602828979492, "learning_rate": 9.640541964563722e-05, "loss": 0.0524, "step": 13230 }, { "grad_norm": 0.4877532124519348, "learning_rate": 9.639772063722096e-05, "loss": 0.0667, "step": 13240 }, { "grad_norm": 0.4742819368839264, "learning_rate": 9.639001370070324e-05, "loss": 0.0513, "step": 13250 }, { "grad_norm": 0.48730432987213135, "learning_rate": 9.638229883740095e-05, "loss": 0.052, "step": 13260 }, { "grad_norm": 0.42229464650154114, "learning_rate": 9.637457604863233e-05, "loss": 0.0466, "step": 13270 }, { "grad_norm": 0.4238821566104889, "learning_rate": 9.636684533571703e-05, "loss": 0.0501, "step": 13280 }, { "grad_norm": 0.4863249957561493, "learning_rate": 9.635910669997599e-05, "loss": 0.051, "step": 13290 }, { "grad_norm": 0.4387257397174835, "learning_rate": 9.635136014273154e-05, "loss": 0.0443, "step": 13300 }, { "grad_norm": 0.35371389985084534, "learning_rate": 9.634360566530735e-05, "loss": 0.0515, "step": 13310 }, { "grad_norm": 0.5465670824050903, "learning_rate": 9.633584326902845e-05, "loss": 0.0617, "step": 13320 }, { "grad_norm": 0.5684611797332764, "learning_rate": 9.632807295522124e-05, "loss": 0.0555, "step": 13330 }, { "grad_norm": 0.44190114736557007, "learning_rate": 9.632029472521342e-05, "loss": 0.0553, "step": 13340 }, { "grad_norm": 0.3808106482028961, "learning_rate": 9.631250858033409e-05, "loss": 0.0491, "step": 13350 }, { "grad_norm": 0.5018193125724792, "learning_rate": 9.630471452191371e-05, "loss": 0.0579, "step": 13360 }, { "grad_norm": 0.48012006282806396, "learning_rate": 9.629691255128405e-05, "loss": 0.0597, "step": 13370 }, { "grad_norm": 0.40859538316726685, "learning_rate": 9.628910266977825e-05, "loss": 0.0494, "step": 13380 }, { "grad_norm": 0.3586263060569763, "learning_rate": 9.628128487873083e-05, "loss": 0.0487, "step": 13390 }, { "grad_norm": 0.35043865442276, "learning_rate": 9.627345917947761e-05, "loss": 0.0525, "step": 13400 }, { "grad_norm": 0.4736713171005249, "learning_rate": 9.626562557335579e-05, "loss": 0.0498, "step": 13410 }, { "grad_norm": 0.43245580792427063, "learning_rate": 9.625778406170393e-05, "loss": 0.0538, "step": 13420 }, { "grad_norm": 0.4239475727081299, "learning_rate": 9.624993464586193e-05, "loss": 0.0504, "step": 13430 }, { "grad_norm": 0.48460060358047485, "learning_rate": 9.624207732717105e-05, "loss": 0.0534, "step": 13440 }, { "grad_norm": 0.45920565724372864, "learning_rate": 9.623421210697386e-05, "loss": 0.0647, "step": 13450 }, { "grad_norm": 0.3592129349708557, "learning_rate": 9.622633898661434e-05, "loss": 0.064, "step": 13460 }, { "grad_norm": 0.36792293190956116, "learning_rate": 9.621845796743778e-05, "loss": 0.0648, "step": 13470 }, { "grad_norm": 0.3306911885738373, "learning_rate": 9.621056905079082e-05, "loss": 0.0537, "step": 13480 }, { "grad_norm": 0.3974781334400177, "learning_rate": 9.620267223802149e-05, "loss": 0.0567, "step": 13490 }, { "grad_norm": 0.3938208818435669, "learning_rate": 9.619476753047911e-05, "loss": 0.0416, "step": 13500 }, { "grad_norm": 0.47415992617607117, "learning_rate": 9.618685492951438e-05, "loss": 0.0544, "step": 13510 }, { "grad_norm": 0.3124696612358093, "learning_rate": 9.617893443647938e-05, "loss": 0.0593, "step": 13520 }, { "grad_norm": 0.37564000487327576, "learning_rate": 9.617100605272746e-05, "loss": 0.0607, "step": 13530 }, { "grad_norm": 0.5579952597618103, "learning_rate": 9.616306977961338e-05, "loss": 0.0554, "step": 13540 }, { "grad_norm": 0.3738827407360077, "learning_rate": 9.615512561849326e-05, "loss": 0.0499, "step": 13550 }, { "grad_norm": 0.5069516897201538, "learning_rate": 9.61471735707245e-05, "loss": 0.0602, "step": 13560 }, { "grad_norm": 0.33480140566825867, "learning_rate": 9.613921363766592e-05, "loss": 0.0436, "step": 13570 }, { "grad_norm": 0.4419878125190735, "learning_rate": 9.613124582067763e-05, "loss": 0.0515, "step": 13580 }, { "grad_norm": 0.3923986852169037, "learning_rate": 9.612327012112112e-05, "loss": 0.0558, "step": 13590 }, { "grad_norm": 0.3670264184474945, "learning_rate": 9.611528654035921e-05, "loss": 0.0667, "step": 13600 }, { "grad_norm": 0.3800758123397827, "learning_rate": 9.610729507975611e-05, "loss": 0.0482, "step": 13610 }, { "grad_norm": 0.3686220645904541, "learning_rate": 9.609929574067731e-05, "loss": 0.0578, "step": 13620 }, { "grad_norm": 0.4565120339393616, "learning_rate": 9.609128852448967e-05, "loss": 0.0497, "step": 13630 }, { "grad_norm": 0.49069541692733765, "learning_rate": 9.608327343256143e-05, "loss": 0.0495, "step": 13640 }, { "grad_norm": 0.41290831565856934, "learning_rate": 9.607525046626216e-05, "loss": 0.0549, "step": 13650 }, { "grad_norm": 0.4971753656864166, "learning_rate": 9.606721962696272e-05, "loss": 0.0558, "step": 13660 }, { "grad_norm": 0.36033356189727783, "learning_rate": 9.60591809160354e-05, "loss": 0.0494, "step": 13670 }, { "grad_norm": 0.3267273008823395, "learning_rate": 9.605113433485378e-05, "loss": 0.0455, "step": 13680 }, { "grad_norm": 0.5122120380401611, "learning_rate": 9.604307988479279e-05, "loss": 0.051, "step": 13690 }, { "grad_norm": 0.3366595208644867, "learning_rate": 9.603501756722876e-05, "loss": 0.0471, "step": 13700 }, { "grad_norm": 0.5737008452415466, "learning_rate": 9.602694738353927e-05, "loss": 0.043, "step": 13710 }, { "grad_norm": 0.3774555027484894, "learning_rate": 9.601886933510331e-05, "loss": 0.0416, "step": 13720 }, { "grad_norm": 0.5089319944381714, "learning_rate": 9.60107834233012e-05, "loss": 0.046, "step": 13730 }, { "grad_norm": 0.44284263253211975, "learning_rate": 9.60026896495146e-05, "loss": 0.0411, "step": 13740 }, { "grad_norm": 0.41548094153404236, "learning_rate": 9.599458801512652e-05, "loss": 0.0477, "step": 13750 }, { "grad_norm": 0.45940569043159485, "learning_rate": 9.598647852152129e-05, "loss": 0.0691, "step": 13760 }, { "grad_norm": 0.3099132776260376, "learning_rate": 9.597836117008462e-05, "loss": 0.0466, "step": 13770 }, { "grad_norm": 0.4400736391544342, "learning_rate": 9.597023596220356e-05, "loss": 0.045, "step": 13780 }, { "grad_norm": 0.34535908699035645, "learning_rate": 9.596210289926643e-05, "loss": 0.048, "step": 13790 }, { "grad_norm": 0.3673900365829468, "learning_rate": 9.5953961982663e-05, "loss": 0.0402, "step": 13800 }, { "grad_norm": 0.4364774823188782, "learning_rate": 9.594581321378431e-05, "loss": 0.0565, "step": 13810 }, { "grad_norm": 0.3994610905647278, "learning_rate": 9.593765659402276e-05, "loss": 0.0509, "step": 13820 }, { "grad_norm": 0.4173692464828491, "learning_rate": 9.59294921247721e-05, "loss": 0.0467, "step": 13830 }, { "grad_norm": 0.5575636029243469, "learning_rate": 9.59213198074274e-05, "loss": 0.0552, "step": 13840 }, { "grad_norm": 0.5074074864387512, "learning_rate": 9.59131396433851e-05, "loss": 0.0439, "step": 13850 }, { "grad_norm": 0.32489362359046936, "learning_rate": 9.590495163404297e-05, "loss": 0.041, "step": 13860 }, { "grad_norm": 0.25676652789115906, "learning_rate": 9.589675578080009e-05, "loss": 0.0533, "step": 13870 }, { "grad_norm": 0.39172300696372986, "learning_rate": 9.588855208505694e-05, "loss": 0.0396, "step": 13880 }, { "grad_norm": 0.41817981004714966, "learning_rate": 9.588034054821529e-05, "loss": 0.0515, "step": 13890 }, { "grad_norm": 0.4096446931362152, "learning_rate": 9.587212117167826e-05, "loss": 0.0398, "step": 13900 }, { "grad_norm": 0.3786260485649109, "learning_rate": 9.586389395685033e-05, "loss": 0.0432, "step": 13910 }, { "grad_norm": 0.4116535484790802, "learning_rate": 9.585565890513733e-05, "loss": 0.0512, "step": 13920 }, { "grad_norm": 0.33770298957824707, "learning_rate": 9.584741601794636e-05, "loss": 0.0438, "step": 13930 }, { "grad_norm": 0.4570930004119873, "learning_rate": 9.58391652966859e-05, "loss": 0.0432, "step": 13940 }, { "grad_norm": 0.3238999843597412, "learning_rate": 9.583090674276583e-05, "loss": 0.0504, "step": 13950 }, { "grad_norm": 0.29662466049194336, "learning_rate": 9.582264035759726e-05, "loss": 0.0457, "step": 13960 }, { "grad_norm": 0.331054151058197, "learning_rate": 9.58143661425927e-05, "loss": 0.0694, "step": 13970 }, { "grad_norm": 0.33007314801216125, "learning_rate": 9.580608409916601e-05, "loss": 0.0549, "step": 13980 }, { "grad_norm": 0.360419362783432, "learning_rate": 9.579779422873233e-05, "loss": 0.0518, "step": 13990 }, { "grad_norm": 0.3795151710510254, "learning_rate": 9.578949653270819e-05, "loss": 0.057, "step": 14000 }, { "grad_norm": 0.36661216616630554, "learning_rate": 9.578119101251144e-05, "loss": 0.0455, "step": 14010 }, { "grad_norm": 0.440208375453949, "learning_rate": 9.577287766956127e-05, "loss": 0.0593, "step": 14020 }, { "grad_norm": 0.48664698004722595, "learning_rate": 9.57645565052782e-05, "loss": 0.0551, "step": 14030 }, { "grad_norm": 0.3731134533882141, "learning_rate": 9.575622752108407e-05, "loss": 0.0429, "step": 14040 }, { "grad_norm": 0.39964932203292847, "learning_rate": 9.57478907184021e-05, "loss": 0.0453, "step": 14050 }, { "grad_norm": 0.3326372802257538, "learning_rate": 9.573954609865681e-05, "loss": 0.0497, "step": 14060 }, { "grad_norm": 0.4095710515975952, "learning_rate": 9.573119366327408e-05, "loss": 0.0505, "step": 14070 }, { "grad_norm": 0.43315377831459045, "learning_rate": 9.57228334136811e-05, "loss": 0.046, "step": 14080 }, { "grad_norm": 0.31943729519844055, "learning_rate": 9.571446535130641e-05, "loss": 0.0416, "step": 14090 }, { "grad_norm": 0.4142836928367615, "learning_rate": 9.570608947757988e-05, "loss": 0.0619, "step": 14100 }, { "grad_norm": 0.48247185349464417, "learning_rate": 9.569770579393274e-05, "loss": 0.055, "step": 14110 }, { "grad_norm": 0.42115458846092224, "learning_rate": 9.56893143017975e-05, "loss": 0.043, "step": 14120 }, { "grad_norm": 0.4776102900505066, "learning_rate": 9.568091500260806e-05, "loss": 0.0509, "step": 14130 }, { "grad_norm": 0.4058060348033905, "learning_rate": 9.567250789779961e-05, "loss": 0.0513, "step": 14140 }, { "grad_norm": 0.4384530782699585, "learning_rate": 9.566409298880872e-05, "loss": 0.0488, "step": 14150 }, { "grad_norm": 0.4400002360343933, "learning_rate": 9.565567027707326e-05, "loss": 0.0649, "step": 14160 }, { "grad_norm": 0.4317375123500824, "learning_rate": 9.56472397640324e-05, "loss": 0.0492, "step": 14170 }, { "grad_norm": 0.31067875027656555, "learning_rate": 9.563880145112675e-05, "loss": 0.0492, "step": 14180 }, { "grad_norm": 0.5590316653251648, "learning_rate": 9.563035533979814e-05, "loss": 0.0491, "step": 14190 }, { "grad_norm": 0.3807728886604309, "learning_rate": 9.562190143148981e-05, "loss": 0.0453, "step": 14200 }, { "grad_norm": 0.3210044801235199, "learning_rate": 9.561343972764627e-05, "loss": 0.0522, "step": 14210 }, { "grad_norm": 0.4194645583629608, "learning_rate": 9.560497022971343e-05, "loss": 0.056, "step": 14220 }, { "grad_norm": 0.3650674819946289, "learning_rate": 9.559649293913847e-05, "loss": 0.0513, "step": 14230 }, { "grad_norm": 0.39155083894729614, "learning_rate": 9.558800785736993e-05, "loss": 0.0401, "step": 14240 }, { "grad_norm": 0.34746092557907104, "learning_rate": 9.557951498585767e-05, "loss": 0.0532, "step": 14250 }, { "grad_norm": 0.37726861238479614, "learning_rate": 9.557101432605293e-05, "loss": 0.0498, "step": 14260 }, { "grad_norm": 0.4621816575527191, "learning_rate": 9.556250587940818e-05, "loss": 0.0578, "step": 14270 }, { "grad_norm": 0.5448092222213745, "learning_rate": 9.555398964737734e-05, "loss": 0.0579, "step": 14280 }, { "grad_norm": 0.48102977871894836, "learning_rate": 9.554546563141555e-05, "loss": 0.0429, "step": 14290 }, { "grad_norm": 0.38829922676086426, "learning_rate": 9.553693383297937e-05, "loss": 0.0549, "step": 14300 }, { "grad_norm": 0.4639061391353607, "learning_rate": 9.552839425352663e-05, "loss": 0.0458, "step": 14310 }, { "grad_norm": 0.42079389095306396, "learning_rate": 9.551984689451652e-05, "loss": 0.0589, "step": 14320 }, { "grad_norm": 0.5671975016593933, "learning_rate": 9.551129175740953e-05, "loss": 0.0478, "step": 14330 }, { "grad_norm": 0.41350021958351135, "learning_rate": 9.550272884366754e-05, "loss": 0.0507, "step": 14340 }, { "grad_norm": 0.6176038980484009, "learning_rate": 9.549415815475369e-05, "loss": 0.054, "step": 14350 }, { "grad_norm": 0.38574671745300293, "learning_rate": 9.548557969213247e-05, "loss": 0.0482, "step": 14360 }, { "grad_norm": 0.47915202379226685, "learning_rate": 9.547699345726972e-05, "loss": 0.0517, "step": 14370 }, { "grad_norm": 0.4768649935722351, "learning_rate": 9.546839945163257e-05, "loss": 0.0496, "step": 14380 }, { "grad_norm": 0.394803524017334, "learning_rate": 9.545979767668953e-05, "loss": 0.0572, "step": 14390 }, { "grad_norm": 0.3038513958454132, "learning_rate": 9.54511881339104e-05, "loss": 0.0502, "step": 14400 }, { "grad_norm": 0.38212859630584717, "learning_rate": 9.54425708247663e-05, "loss": 0.0455, "step": 14410 }, { "grad_norm": 0.4567480683326721, "learning_rate": 9.543394575072972e-05, "loss": 0.0437, "step": 14420 }, { "grad_norm": 0.30338889360427856, "learning_rate": 9.542531291327441e-05, "loss": 0.0557, "step": 14430 }, { "grad_norm": 0.35160401463508606, "learning_rate": 9.541667231387552e-05, "loss": 0.0398, "step": 14440 }, { "grad_norm": 0.4692779779434204, "learning_rate": 9.540802395400949e-05, "loss": 0.0531, "step": 14450 }, { "grad_norm": 0.4204387366771698, "learning_rate": 9.539936783515406e-05, "loss": 0.0488, "step": 14460 }, { "grad_norm": 0.45448464155197144, "learning_rate": 9.539070395878835e-05, "loss": 0.0522, "step": 14470 }, { "grad_norm": 0.5011891722679138, "learning_rate": 9.538203232639277e-05, "loss": 0.047, "step": 14480 }, { "grad_norm": 0.5898475050926208, "learning_rate": 9.537335293944907e-05, "loss": 0.0497, "step": 14490 }, { "grad_norm": 0.31516993045806885, "learning_rate": 9.536466579944032e-05, "loss": 0.0537, "step": 14500 }, { "grad_norm": 0.3265690505504608, "learning_rate": 9.535597090785091e-05, "loss": 0.0516, "step": 14510 }, { "grad_norm": 0.42489388585090637, "learning_rate": 9.534726826616656e-05, "loss": 0.0417, "step": 14520 }, { "grad_norm": 0.5068410634994507, "learning_rate": 9.53385578758743e-05, "loss": 0.0712, "step": 14530 }, { "grad_norm": 0.41895627975463867, "learning_rate": 9.532983973846252e-05, "loss": 0.0383, "step": 14540 }, { "grad_norm": 0.4666752815246582, "learning_rate": 9.53211138554209e-05, "loss": 0.0414, "step": 14550 }, { "grad_norm": 0.3855378329753876, "learning_rate": 9.531238022824047e-05, "loss": 0.0467, "step": 14560 }, { "grad_norm": 0.3786771893501282, "learning_rate": 9.530363885841355e-05, "loss": 0.0392, "step": 14570 }, { "grad_norm": 0.53058260679245, "learning_rate": 9.52948897474338e-05, "loss": 0.0547, "step": 14580 }, { "grad_norm": 0.4891042411327362, "learning_rate": 9.528613289679622e-05, "loss": 0.0711, "step": 14590 }, { "grad_norm": 0.517229437828064, "learning_rate": 9.52773683079971e-05, "loss": 0.0631, "step": 14600 }, { "grad_norm": 0.3413327634334564, "learning_rate": 9.526859598253407e-05, "loss": 0.0449, "step": 14610 }, { "grad_norm": 0.29882726073265076, "learning_rate": 9.525981592190609e-05, "loss": 0.0475, "step": 14620 }, { "grad_norm": 0.3783096373081207, "learning_rate": 9.525102812761342e-05, "loss": 0.0449, "step": 14630 }, { "grad_norm": 0.35433995723724365, "learning_rate": 9.524223260115768e-05, "loss": 0.0459, "step": 14640 }, { "grad_norm": 0.4358241856098175, "learning_rate": 9.523342934404175e-05, "loss": 0.0496, "step": 14650 }, { "grad_norm": 0.3696601390838623, "learning_rate": 9.522461835776989e-05, "loss": 0.0502, "step": 14660 }, { "grad_norm": 0.6944381594657898, "learning_rate": 9.521579964384764e-05, "loss": 0.0621, "step": 14670 }, { "grad_norm": 0.5446110367774963, "learning_rate": 9.52069732037819e-05, "loss": 0.0498, "step": 14680 }, { "grad_norm": 0.3564583361148834, "learning_rate": 9.519813903908083e-05, "loss": 0.053, "step": 14690 }, { "grad_norm": 0.30427059531211853, "learning_rate": 9.5189297151254e-05, "loss": 0.0549, "step": 14700 }, { "grad_norm": 0.584262490272522, "learning_rate": 9.518044754181218e-05, "loss": 0.0588, "step": 14710 }, { "grad_norm": 0.38490501046180725, "learning_rate": 9.51715902122676e-05, "loss": 0.0558, "step": 14720 }, { "grad_norm": 0.42347288131713867, "learning_rate": 9.516272516413368e-05, "loss": 0.0536, "step": 14730 }, { "grad_norm": 0.3323464095592499, "learning_rate": 9.515385239892525e-05, "loss": 0.0596, "step": 14740 }, { "grad_norm": 0.5366449952125549, "learning_rate": 9.514497191815839e-05, "loss": 0.0615, "step": 14750 }, { "grad_norm": 0.42886072397232056, "learning_rate": 9.513608372335055e-05, "loss": 0.0607, "step": 14760 }, { "grad_norm": 0.524726152420044, "learning_rate": 9.512718781602045e-05, "loss": 0.0558, "step": 14770 }, { "grad_norm": 0.5502650141716003, "learning_rate": 9.511828419768823e-05, "loss": 0.0496, "step": 14780 }, { "grad_norm": 0.40363621711730957, "learning_rate": 9.510937286987521e-05, "loss": 0.0442, "step": 14790 }, { "grad_norm": 1.2850364446640015, "learning_rate": 9.510045383410408e-05, "loss": 0.0548, "step": 14800 }, { "grad_norm": 0.3417937755584717, "learning_rate": 9.509152709189892e-05, "loss": 0.0448, "step": 14810 }, { "grad_norm": 0.294826477766037, "learning_rate": 9.508259264478504e-05, "loss": 0.0513, "step": 14820 }, { "grad_norm": 0.46790018677711487, "learning_rate": 9.507365049428909e-05, "loss": 0.0643, "step": 14830 }, { "grad_norm": 0.5265920162200928, "learning_rate": 9.506470064193902e-05, "loss": 0.0708, "step": 14840 }, { "grad_norm": 0.32199153304100037, "learning_rate": 9.505574308926414e-05, "loss": 0.0672, "step": 14850 }, { "grad_norm": 0.4631146490573883, "learning_rate": 9.504677783779505e-05, "loss": 0.0507, "step": 14860 }, { "grad_norm": 0.4620554447174072, "learning_rate": 9.503780488906365e-05, "loss": 0.0488, "step": 14870 }, { "grad_norm": 0.4582701027393341, "learning_rate": 9.502882424460319e-05, "loss": 0.0567, "step": 14880 }, { "grad_norm": 0.388089120388031, "learning_rate": 9.501983590594821e-05, "loss": 0.051, "step": 14890 }, { "grad_norm": 0.5068809390068054, "learning_rate": 9.501083987463455e-05, "loss": 0.0481, "step": 14900 }, { "grad_norm": 0.5229238867759705, "learning_rate": 9.500183615219942e-05, "loss": 0.0561, "step": 14910 }, { "grad_norm": 0.4877883195877075, "learning_rate": 9.49928247401813e-05, "loss": 0.0505, "step": 14920 }, { "grad_norm": 0.5255085229873657, "learning_rate": 9.498380564011997e-05, "loss": 0.0466, "step": 14930 }, { "grad_norm": 0.4226263165473938, "learning_rate": 9.497477885355656e-05, "loss": 0.0629, "step": 14940 }, { "grad_norm": 0.4635895788669586, "learning_rate": 9.496574438203353e-05, "loss": 0.0688, "step": 14950 }, { "grad_norm": 0.3292471170425415, "learning_rate": 9.495670222709459e-05, "loss": 0.0493, "step": 14960 }, { "grad_norm": 0.38932323455810547, "learning_rate": 9.494765239028483e-05, "loss": 0.0436, "step": 14970 }, { "grad_norm": 0.4457554519176483, "learning_rate": 9.493859487315057e-05, "loss": 0.048, "step": 14980 }, { "grad_norm": 0.6447538137435913, "learning_rate": 9.492952967723953e-05, "loss": 0.0551, "step": 14990 }, { "grad_norm": 0.48476147651672363, "learning_rate": 9.492045680410068e-05, "loss": 0.0578, "step": 15000 }, { "grad_norm": 0.41299766302108765, "learning_rate": 9.491137625528436e-05, "loss": 0.0469, "step": 15010 }, { "grad_norm": 0.4888252019882202, "learning_rate": 9.490228803234215e-05, "loss": 0.0517, "step": 15020 }, { "grad_norm": 0.42953062057495117, "learning_rate": 9.489319213682701e-05, "loss": 0.0462, "step": 15030 }, { "grad_norm": 0.39810898900032043, "learning_rate": 9.488408857029316e-05, "loss": 0.0621, "step": 15040 }, { "grad_norm": 0.39696118235588074, "learning_rate": 9.487497733429616e-05, "loss": 0.0646, "step": 15050 }, { "grad_norm": 0.45226624608039856, "learning_rate": 9.486585843039286e-05, "loss": 0.0601, "step": 15060 }, { "grad_norm": 0.3571203947067261, "learning_rate": 9.485673186014143e-05, "loss": 0.0515, "step": 15070 }, { "grad_norm": 0.45444974303245544, "learning_rate": 9.484759762510137e-05, "loss": 0.051, "step": 15080 }, { "grad_norm": 0.5148141384124756, "learning_rate": 9.483845572683346e-05, "loss": 0.0622, "step": 15090 }, { "grad_norm": 0.43601322174072266, "learning_rate": 9.48293061668998e-05, "loss": 0.0527, "step": 15100 }, { "grad_norm": 0.4823493957519531, "learning_rate": 9.48201489468638e-05, "loss": 0.0442, "step": 15110 }, { "grad_norm": 0.4572224020957947, "learning_rate": 9.481098406829016e-05, "loss": 0.0422, "step": 15120 }, { "grad_norm": 0.32653188705444336, "learning_rate": 9.480181153274495e-05, "loss": 0.0437, "step": 15130 }, { "grad_norm": 0.3230467736721039, "learning_rate": 9.479263134179548e-05, "loss": 0.0337, "step": 15140 }, { "grad_norm": 0.3946402668952942, "learning_rate": 9.478344349701039e-05, "loss": 0.0553, "step": 15150 }, { "grad_norm": 0.6574546098709106, "learning_rate": 9.477424799995964e-05, "loss": 0.0541, "step": 15160 }, { "grad_norm": 0.5054164528846741, "learning_rate": 9.476504485221448e-05, "loss": 0.0579, "step": 15170 }, { "grad_norm": 0.3685353696346283, "learning_rate": 9.475583405534748e-05, "loss": 0.0528, "step": 15180 }, { "grad_norm": 0.4246561527252197, "learning_rate": 9.474661561093251e-05, "loss": 0.0463, "step": 15190 }, { "grad_norm": 0.4507405459880829, "learning_rate": 9.473738952054478e-05, "loss": 0.0551, "step": 15200 }, { "grad_norm": 0.40283361077308655, "learning_rate": 9.472815578576073e-05, "loss": 0.0432, "step": 15210 }, { "grad_norm": 0.4525303244590759, "learning_rate": 9.471891440815817e-05, "loss": 0.0451, "step": 15220 }, { "grad_norm": 0.37030744552612305, "learning_rate": 9.470966538931621e-05, "loss": 0.0367, "step": 15230 }, { "grad_norm": 0.42380577325820923, "learning_rate": 9.470040873081525e-05, "loss": 0.0451, "step": 15240 }, { "grad_norm": 0.4200498163700104, "learning_rate": 9.469114443423698e-05, "loss": 0.0612, "step": 15250 }, { "grad_norm": 0.4669121205806732, "learning_rate": 9.468187250116445e-05, "loss": 0.0553, "step": 15260 }, { "grad_norm": 0.3305835425853729, "learning_rate": 9.467259293318197e-05, "loss": 0.0484, "step": 15270 }, { "grad_norm": 0.36870723962783813, "learning_rate": 9.466330573187514e-05, "loss": 0.048, "step": 15280 }, { "grad_norm": 0.5179764032363892, "learning_rate": 9.46540108988309e-05, "loss": 0.0567, "step": 15290 }, { "grad_norm": 0.3303847014904022, "learning_rate": 9.46447084356375e-05, "loss": 0.0476, "step": 15300 }, { "grad_norm": 0.29788798093795776, "learning_rate": 9.463539834388447e-05, "loss": 0.0592, "step": 15310 }, { "grad_norm": 0.3917042911052704, "learning_rate": 9.462608062516263e-05, "loss": 0.0402, "step": 15320 }, { "grad_norm": 0.5237878561019897, "learning_rate": 9.461675528106413e-05, "loss": 0.0542, "step": 15330 }, { "grad_norm": 0.4257214367389679, "learning_rate": 9.460742231318244e-05, "loss": 0.053, "step": 15340 }, { "grad_norm": 0.34464818239212036, "learning_rate": 9.459808172311229e-05, "loss": 0.0493, "step": 15350 }, { "grad_norm": 0.4172993302345276, "learning_rate": 9.458873351244972e-05, "loss": 0.0467, "step": 15360 }, { "grad_norm": 0.4048592746257782, "learning_rate": 9.457937768279211e-05, "loss": 0.0454, "step": 15370 }, { "grad_norm": 0.36036667227745056, "learning_rate": 9.45700142357381e-05, "loss": 0.0501, "step": 15380 }, { "grad_norm": 0.3749224841594696, "learning_rate": 9.456064317288765e-05, "loss": 0.044, "step": 15390 }, { "grad_norm": 0.31668007373809814, "learning_rate": 9.455126449584201e-05, "loss": 0.0481, "step": 15400 }, { "grad_norm": 0.4623856842517853, "learning_rate": 9.454187820620375e-05, "loss": 0.0409, "step": 15410 }, { "grad_norm": 0.45110252499580383, "learning_rate": 9.453248430557673e-05, "loss": 0.053, "step": 15420 }, { "grad_norm": 0.3828640580177307, "learning_rate": 9.452308279556611e-05, "loss": 0.0718, "step": 15430 }, { "grad_norm": 0.4405403435230255, "learning_rate": 9.451367367777835e-05, "loss": 0.0453, "step": 15440 }, { "grad_norm": 0.4564746916294098, "learning_rate": 9.450425695382122e-05, "loss": 0.0605, "step": 15450 }, { "grad_norm": 0.36025211215019226, "learning_rate": 9.449483262530375e-05, "loss": 0.0472, "step": 15460 }, { "grad_norm": 0.3153211176395416, "learning_rate": 9.448540069383633e-05, "loss": 0.04, "step": 15470 }, { "grad_norm": 0.5100336670875549, "learning_rate": 9.447596116103061e-05, "loss": 0.0586, "step": 15480 }, { "grad_norm": 0.4389280378818512, "learning_rate": 9.446651402849955e-05, "loss": 0.0551, "step": 15490 }, { "grad_norm": 0.41152244806289673, "learning_rate": 9.44570592978574e-05, "loss": 0.0519, "step": 15500 }, { "grad_norm": 0.48146718740463257, "learning_rate": 9.444759697071972e-05, "loss": 0.0519, "step": 15510 }, { "grad_norm": 0.48547276854515076, "learning_rate": 9.443812704870336e-05, "loss": 0.042, "step": 15520 }, { "grad_norm": 0.30935242772102356, "learning_rate": 9.442864953342649e-05, "loss": 0.0423, "step": 15530 }, { "grad_norm": 0.40783968567848206, "learning_rate": 9.441916442650852e-05, "loss": 0.0479, "step": 15540 }, { "grad_norm": 0.3471452593803406, "learning_rate": 9.440967172957023e-05, "loss": 0.0475, "step": 15550 }, { "grad_norm": 0.4507414698600769, "learning_rate": 9.440017144423364e-05, "loss": 0.0553, "step": 15560 }, { "grad_norm": 0.4128723740577698, "learning_rate": 9.439066357212209e-05, "loss": 0.0601, "step": 15570 }, { "grad_norm": 0.3955446183681488, "learning_rate": 9.438114811486022e-05, "loss": 0.0596, "step": 15580 }, { "grad_norm": 0.3410930931568146, "learning_rate": 9.4371625074074e-05, "loss": 0.0426, "step": 15590 }, { "grad_norm": 0.40432360768318176, "learning_rate": 9.436209445139059e-05, "loss": 0.0478, "step": 15600 }, { "grad_norm": 0.45004796981811523, "learning_rate": 9.435255624843855e-05, "loss": 0.0562, "step": 15610 }, { "grad_norm": 0.4552745819091797, "learning_rate": 9.43430104668477e-05, "loss": 0.0467, "step": 15620 }, { "grad_norm": 0.6279423832893372, "learning_rate": 9.433345710824914e-05, "loss": 0.0512, "step": 15630 }, { "grad_norm": 0.36806240677833557, "learning_rate": 9.432389617427529e-05, "loss": 0.0509, "step": 15640 }, { "grad_norm": 0.2883579432964325, "learning_rate": 9.431432766655984e-05, "loss": 0.0562, "step": 15650 }, { "grad_norm": 0.40366771817207336, "learning_rate": 9.430475158673778e-05, "loss": 0.0525, "step": 15660 }, { "grad_norm": 0.4013756215572357, "learning_rate": 9.429516793644542e-05, "loss": 0.0422, "step": 15670 }, { "grad_norm": 0.3459937274456024, "learning_rate": 9.428557671732034e-05, "loss": 0.0397, "step": 15680 }, { "grad_norm": 0.4271726608276367, "learning_rate": 9.42759779310014e-05, "loss": 0.0513, "step": 15690 }, { "grad_norm": 0.35739678144454956, "learning_rate": 9.426637157912879e-05, "loss": 0.0419, "step": 15700 }, { "grad_norm": 0.5320121645927429, "learning_rate": 9.425675766334397e-05, "loss": 0.0579, "step": 15710 }, { "grad_norm": 0.3765610158443451, "learning_rate": 9.424713618528968e-05, "loss": 0.0396, "step": 15720 }, { "grad_norm": 0.45914870500564575, "learning_rate": 9.423750714661e-05, "loss": 0.0592, "step": 15730 }, { "grad_norm": 0.36211714148521423, "learning_rate": 9.422787054895022e-05, "loss": 0.0443, "step": 15740 }, { "grad_norm": 0.485203355550766, "learning_rate": 9.4218226393957e-05, "loss": 0.0546, "step": 15750 }, { "grad_norm": 0.5170741081237793, "learning_rate": 9.420857468327828e-05, "loss": 0.0475, "step": 15760 }, { "grad_norm": 0.4184960126876831, "learning_rate": 9.419891541856323e-05, "loss": 0.0489, "step": 15770 }, { "grad_norm": 0.3796289265155792, "learning_rate": 9.41892486014624e-05, "loss": 0.0458, "step": 15780 }, { "grad_norm": 0.402134507894516, "learning_rate": 9.417957423362756e-05, "loss": 0.0507, "step": 15790 }, { "grad_norm": 0.35367879271507263, "learning_rate": 9.416989231671178e-05, "loss": 0.0488, "step": 15800 }, { "grad_norm": 0.4789433777332306, "learning_rate": 9.416020285236946e-05, "loss": 0.0441, "step": 15810 }, { "grad_norm": 0.344201922416687, "learning_rate": 9.415050584225626e-05, "loss": 0.0476, "step": 15820 }, { "grad_norm": 0.3578861355781555, "learning_rate": 9.414080128802914e-05, "loss": 0.0467, "step": 15830 }, { "grad_norm": 0.4975089430809021, "learning_rate": 9.413108919134632e-05, "loss": 0.0421, "step": 15840 }, { "grad_norm": 0.3628584146499634, "learning_rate": 9.412136955386734e-05, "loss": 0.0426, "step": 15850 }, { "grad_norm": 0.3021527826786041, "learning_rate": 9.411164237725303e-05, "loss": 0.0481, "step": 15860 }, { "grad_norm": 0.32320529222488403, "learning_rate": 9.41019076631655e-05, "loss": 0.0484, "step": 15870 }, { "grad_norm": 0.36792007088661194, "learning_rate": 9.409216541326815e-05, "loss": 0.0511, "step": 15880 }, { "grad_norm": 0.33881375193595886, "learning_rate": 9.408241562922564e-05, "loss": 0.0475, "step": 15890 }, { "grad_norm": 0.3707522451877594, "learning_rate": 9.407265831270395e-05, "loss": 0.0367, "step": 15900 }, { "grad_norm": 0.38283729553222656, "learning_rate": 9.406289346537035e-05, "loss": 0.0515, "step": 15910 }, { "grad_norm": 0.29525628685951233, "learning_rate": 9.405312108889339e-05, "loss": 0.0478, "step": 15920 }, { "grad_norm": 0.45787256956100464, "learning_rate": 9.404334118494288e-05, "loss": 0.0495, "step": 15930 }, { "grad_norm": 0.2682856619358063, "learning_rate": 9.403355375518995e-05, "loss": 0.0422, "step": 15940 }, { "grad_norm": 0.4111100435256958, "learning_rate": 9.4023758801307e-05, "loss": 0.0413, "step": 15950 }, { "grad_norm": 0.3633157014846802, "learning_rate": 9.401395632496774e-05, "loss": 0.0469, "step": 15960 }, { "grad_norm": 0.4817461669445038, "learning_rate": 9.400414632784711e-05, "loss": 0.0536, "step": 15970 }, { "grad_norm": 0.46225178241729736, "learning_rate": 9.39943288116214e-05, "loss": 0.0383, "step": 15980 }, { "grad_norm": 0.2876233458518982, "learning_rate": 9.398450377796815e-05, "loss": 0.0429, "step": 15990 }, { "grad_norm": 0.4556947946548462, "learning_rate": 9.397467122856616e-05, "loss": 0.0433, "step": 16000 }, { "grad_norm": 0.3629319369792938, "learning_rate": 9.396483116509558e-05, "loss": 0.0426, "step": 16010 }, { "grad_norm": 0.4058677554130554, "learning_rate": 9.39549835892378e-05, "loss": 0.0469, "step": 16020 }, { "grad_norm": 0.40097180008888245, "learning_rate": 9.39451285026755e-05, "loss": 0.0437, "step": 16030 }, { "grad_norm": 0.465803861618042, "learning_rate": 9.393526590709262e-05, "loss": 0.0444, "step": 16040 }, { "grad_norm": 0.435294508934021, "learning_rate": 9.392539580417444e-05, "loss": 0.0378, "step": 16050 }, { "grad_norm": 0.3122302293777466, "learning_rate": 9.391551819560747e-05, "loss": 0.0384, "step": 16060 }, { "grad_norm": 0.4216817617416382, "learning_rate": 9.390563308307955e-05, "loss": 0.0427, "step": 16070 }, { "grad_norm": 0.3735422194004059, "learning_rate": 9.389574046827974e-05, "loss": 0.044, "step": 16080 }, { "grad_norm": 0.5247023105621338, "learning_rate": 9.388584035289845e-05, "loss": 0.0453, "step": 16090 }, { "grad_norm": 0.3488292694091797, "learning_rate": 9.387593273862732e-05, "loss": 0.0486, "step": 16100 }, { "grad_norm": 0.34498563408851624, "learning_rate": 9.386601762715929e-05, "loss": 0.0528, "step": 16110 }, { "grad_norm": 0.5049160718917847, "learning_rate": 9.38560950201886e-05, "loss": 0.0549, "step": 16120 }, { "grad_norm": 0.5394055843353271, "learning_rate": 9.384616491941071e-05, "loss": 0.0544, "step": 16130 }, { "grad_norm": 0.4184073209762573, "learning_rate": 9.383622732652245e-05, "loss": 0.0407, "step": 16140 }, { "grad_norm": 0.2506311237812042, "learning_rate": 9.382628224322187e-05, "loss": 0.0493, "step": 16150 }, { "grad_norm": 0.39066964387893677, "learning_rate": 9.381632967120829e-05, "loss": 0.0462, "step": 16160 }, { "grad_norm": 0.3739645481109619, "learning_rate": 9.380636961218235e-05, "loss": 0.0442, "step": 16170 }, { "grad_norm": 0.3919001817703247, "learning_rate": 9.379640206784597e-05, "loss": 0.039, "step": 16180 }, { "grad_norm": 0.45810648798942566, "learning_rate": 9.378642703990229e-05, "loss": 0.0413, "step": 16190 }, { "grad_norm": 0.3835254907608032, "learning_rate": 9.37764445300558e-05, "loss": 0.043, "step": 16200 }, { "grad_norm": 0.39895468950271606, "learning_rate": 9.376645454001222e-05, "loss": 0.0608, "step": 16210 }, { "grad_norm": 0.3542245030403137, "learning_rate": 9.375645707147858e-05, "loss": 0.0395, "step": 16220 }, { "grad_norm": 0.311703622341156, "learning_rate": 9.374645212616316e-05, "loss": 0.0535, "step": 16230 }, { "grad_norm": 0.3619893789291382, "learning_rate": 9.373643970577555e-05, "loss": 0.0541, "step": 16240 }, { "grad_norm": 0.42417067289352417, "learning_rate": 9.372641981202659e-05, "loss": 0.0457, "step": 16250 }, { "grad_norm": 0.4493752717971802, "learning_rate": 9.37163924466284e-05, "loss": 0.0427, "step": 16260 }, { "grad_norm": 0.43539753556251526, "learning_rate": 9.370635761129438e-05, "loss": 0.0486, "step": 16270 }, { "grad_norm": 0.3239724934101105, "learning_rate": 9.36963153077392e-05, "loss": 0.0402, "step": 16280 }, { "grad_norm": 0.26273036003112793, "learning_rate": 9.368626553767888e-05, "loss": 0.0445, "step": 16290 }, { "grad_norm": 0.28807926177978516, "learning_rate": 9.367620830283057e-05, "loss": 0.0344, "step": 16300 }, { "grad_norm": 0.41244831681251526, "learning_rate": 9.366614360491281e-05, "loss": 0.0497, "step": 16310 }, { "grad_norm": 0.42351648211479187, "learning_rate": 9.365607144564539e-05, "loss": 0.0402, "step": 16320 }, { "grad_norm": 0.3081357777118683, "learning_rate": 9.364599182674934e-05, "loss": 0.0523, "step": 16330 }, { "grad_norm": 0.34125956892967224, "learning_rate": 9.3635904749947e-05, "loss": 0.0388, "step": 16340 }, { "grad_norm": 0.3014134168624878, "learning_rate": 9.362581021696202e-05, "loss": 0.0369, "step": 16350 }, { "grad_norm": 0.406343936920166, "learning_rate": 9.361570822951921e-05, "loss": 0.0502, "step": 16360 }, { "grad_norm": 0.36394941806793213, "learning_rate": 9.360559878934476e-05, "loss": 0.0521, "step": 16370 }, { "grad_norm": 0.2905309498310089, "learning_rate": 9.359548189816611e-05, "loss": 0.0458, "step": 16380 }, { "grad_norm": 0.4542585015296936, "learning_rate": 9.358535755771193e-05, "loss": 0.0558, "step": 16390 }, { "grad_norm": 0.42960962653160095, "learning_rate": 9.357522576971221e-05, "loss": 0.0476, "step": 16400 }, { "grad_norm": 0.44195595383644104, "learning_rate": 9.356508653589819e-05, "loss": 0.0465, "step": 16410 }, { "grad_norm": 0.6270227432250977, "learning_rate": 9.355493985800237e-05, "loss": 0.0467, "step": 16420 }, { "grad_norm": 0.4317963421344757, "learning_rate": 9.354478573775857e-05, "loss": 0.0554, "step": 16430 }, { "grad_norm": 0.3556862771511078, "learning_rate": 9.353462417690186e-05, "loss": 0.0499, "step": 16440 }, { "grad_norm": 0.3149041533470154, "learning_rate": 9.352445517716853e-05, "loss": 0.0417, "step": 16450 }, { "grad_norm": 0.37002524733543396, "learning_rate": 9.351427874029621e-05, "loss": 0.044, "step": 16460 }, { "grad_norm": 0.6977484226226807, "learning_rate": 9.350409486802379e-05, "loss": 0.0492, "step": 16470 }, { "grad_norm": 0.429383784532547, "learning_rate": 9.349390356209138e-05, "loss": 0.0568, "step": 16480 }, { "grad_norm": 0.5485024452209473, "learning_rate": 9.348370482424042e-05, "loss": 0.0633, "step": 16490 }, { "grad_norm": 0.5575824975967407, "learning_rate": 9.347349865621357e-05, "loss": 0.0493, "step": 16500 }, { "grad_norm": 0.45760786533355713, "learning_rate": 9.346328505975481e-05, "loss": 0.0494, "step": 16510 }, { "grad_norm": 0.2458934485912323, "learning_rate": 9.345306403660936e-05, "loss": 0.0479, "step": 16520 }, { "grad_norm": 0.4567202925682068, "learning_rate": 9.344283558852371e-05, "loss": 0.0513, "step": 16530 }, { "grad_norm": 0.3749772012233734, "learning_rate": 9.343259971724563e-05, "loss": 0.0458, "step": 16540 }, { "grad_norm": 0.5054233074188232, "learning_rate": 9.342235642452413e-05, "loss": 0.047, "step": 16550 }, { "grad_norm": 0.5033112168312073, "learning_rate": 9.341210571210954e-05, "loss": 0.0538, "step": 16560 }, { "grad_norm": 0.4105849862098694, "learning_rate": 9.340184758175338e-05, "loss": 0.046, "step": 16570 }, { "grad_norm": 0.3119151294231415, "learning_rate": 9.339158203520854e-05, "loss": 0.0562, "step": 16580 }, { "grad_norm": 0.36816850304603577, "learning_rate": 9.338130907422908e-05, "loss": 0.0401, "step": 16590 }, { "grad_norm": 0.3755655884742737, "learning_rate": 9.337102870057037e-05, "loss": 0.0423, "step": 16600 }, { "grad_norm": 0.3725506067276001, "learning_rate": 9.336074091598907e-05, "loss": 0.0456, "step": 16610 }, { "grad_norm": 0.29709142446517944, "learning_rate": 9.335044572224306e-05, "loss": 0.0449, "step": 16620 }, { "grad_norm": 0.4599640667438507, "learning_rate": 9.334014312109151e-05, "loss": 0.0441, "step": 16630 }, { "grad_norm": 0.4381673336029053, "learning_rate": 9.332983311429486e-05, "loss": 0.0507, "step": 16640 }, { "grad_norm": 0.42602235078811646, "learning_rate": 9.33195157036148e-05, "loss": 0.0485, "step": 16650 }, { "grad_norm": 0.36848264932632446, "learning_rate": 9.330919089081432e-05, "loss": 0.0427, "step": 16660 }, { "grad_norm": 0.5218507051467896, "learning_rate": 9.32988586776576e-05, "loss": 0.0548, "step": 16670 }, { "grad_norm": 0.4297339618206024, "learning_rate": 9.328851906591016e-05, "loss": 0.0417, "step": 16680 }, { "grad_norm": 0.4371405839920044, "learning_rate": 9.327817205733875e-05, "loss": 0.048, "step": 16690 }, { "grad_norm": 0.33267611265182495, "learning_rate": 9.326781765371142e-05, "loss": 0.0403, "step": 16700 }, { "grad_norm": 0.3325687348842621, "learning_rate": 9.325745585679741e-05, "loss": 0.0422, "step": 16710 }, { "grad_norm": 0.47313085198402405, "learning_rate": 9.32470866683673e-05, "loss": 0.0593, "step": 16720 }, { "grad_norm": 0.38448894023895264, "learning_rate": 9.323671009019288e-05, "loss": 0.0472, "step": 16730 }, { "grad_norm": 0.3576223850250244, "learning_rate": 9.322632612404725e-05, "loss": 0.0464, "step": 16740 }, { "grad_norm": 0.480765700340271, "learning_rate": 9.321593477170471e-05, "loss": 0.0513, "step": 16750 }, { "grad_norm": 0.43290647864341736, "learning_rate": 9.320553603494088e-05, "loss": 0.04, "step": 16760 }, { "grad_norm": 0.349958598613739, "learning_rate": 9.319512991553261e-05, "loss": 0.0473, "step": 16770 }, { "grad_norm": 0.40382057428359985, "learning_rate": 9.318471641525803e-05, "loss": 0.0456, "step": 16780 }, { "grad_norm": 0.4867876470088959, "learning_rate": 9.317429553589652e-05, "loss": 0.0421, "step": 16790 }, { "grad_norm": 0.4283430278301239, "learning_rate": 9.316386727922873e-05, "loss": 0.0472, "step": 16800 }, { "grad_norm": 0.40420493483543396, "learning_rate": 9.315343164703656e-05, "loss": 0.0386, "step": 16810 }, { "grad_norm": 0.7976725101470947, "learning_rate": 9.314298864110316e-05, "loss": 0.0428, "step": 16820 }, { "grad_norm": 0.35119864344596863, "learning_rate": 9.313253826321295e-05, "loss": 0.0391, "step": 16830 }, { "grad_norm": 0.3885014057159424, "learning_rate": 9.312208051515165e-05, "loss": 0.0462, "step": 16840 }, { "grad_norm": 0.44930800795555115, "learning_rate": 9.311161539870618e-05, "loss": 0.0464, "step": 16850 }, { "grad_norm": 0.409423291683197, "learning_rate": 9.310114291566474e-05, "loss": 0.0474, "step": 16860 }, { "grad_norm": 0.4722048342227936, "learning_rate": 9.309066306781679e-05, "loss": 0.0451, "step": 16870 }, { "grad_norm": 0.3385743200778961, "learning_rate": 9.308017585695306e-05, "loss": 0.0424, "step": 16880 }, { "grad_norm": 0.4321271777153015, "learning_rate": 9.306968128486552e-05, "loss": 0.0462, "step": 16890 }, { "grad_norm": 0.4545794725418091, "learning_rate": 9.30591793533474e-05, "loss": 0.0404, "step": 16900 }, { "grad_norm": 0.3380703628063202, "learning_rate": 9.304867006419321e-05, "loss": 0.0357, "step": 16910 }, { "grad_norm": 0.3898671865463257, "learning_rate": 9.303815341919868e-05, "loss": 0.0456, "step": 16920 }, { "grad_norm": 0.5399901866912842, "learning_rate": 9.302762942016084e-05, "loss": 0.0549, "step": 16930 }, { "grad_norm": 0.32030537724494934, "learning_rate": 9.301709806887792e-05, "loss": 0.0466, "step": 16940 }, { "grad_norm": 0.5294259786605835, "learning_rate": 9.300655936714948e-05, "loss": 0.0575, "step": 16950 }, { "grad_norm": 0.3014405071735382, "learning_rate": 9.299601331677627e-05, "loss": 0.0421, "step": 16960 }, { "grad_norm": 0.3840538263320923, "learning_rate": 9.298545991956033e-05, "loss": 0.0496, "step": 16970 }, { "grad_norm": 0.41702505946159363, "learning_rate": 9.297489917730493e-05, "loss": 0.0453, "step": 16980 }, { "grad_norm": 0.43218371272087097, "learning_rate": 9.296433109181464e-05, "loss": 0.0561, "step": 16990 }, { "grad_norm": 0.44265615940093994, "learning_rate": 9.295375566489523e-05, "loss": 0.0487, "step": 17000 }, { "grad_norm": 0.43241962790489197, "learning_rate": 9.294317289835379e-05, "loss": 0.054, "step": 17010 }, { "grad_norm": 0.4174678921699524, "learning_rate": 9.293258279399859e-05, "loss": 0.0547, "step": 17020 }, { "grad_norm": 0.3761466443538666, "learning_rate": 9.292198535363919e-05, "loss": 0.0421, "step": 17030 }, { "grad_norm": 0.39206749200820923, "learning_rate": 9.291138057908641e-05, "loss": 0.0362, "step": 17040 }, { "grad_norm": 0.43873801827430725, "learning_rate": 9.290076847215234e-05, "loss": 0.0426, "step": 17050 }, { "grad_norm": 0.6198487877845764, "learning_rate": 9.289014903465025e-05, "loss": 0.0606, "step": 17060 }, { "grad_norm": 0.3943873941898346, "learning_rate": 9.287952226839475e-05, "loss": 0.0431, "step": 17070 }, { "grad_norm": 0.49714696407318115, "learning_rate": 9.286888817520164e-05, "loss": 0.0405, "step": 17080 }, { "grad_norm": 0.5080433487892151, "learning_rate": 9.285824675688803e-05, "loss": 0.0456, "step": 17090 }, { "grad_norm": 0.32719486951828003, "learning_rate": 9.28475980152722e-05, "loss": 0.0452, "step": 17100 }, { "grad_norm": 0.425748735666275, "learning_rate": 9.283694195217379e-05, "loss": 0.0468, "step": 17110 }, { "grad_norm": 0.4639224410057068, "learning_rate": 9.282627856941356e-05, "loss": 0.0425, "step": 17120 }, { "grad_norm": 0.4149128198623657, "learning_rate": 9.281560786881363e-05, "loss": 0.0408, "step": 17130 }, { "grad_norm": 0.4927929937839508, "learning_rate": 9.280492985219733e-05, "loss": 0.0371, "step": 17140 }, { "grad_norm": 0.4534832835197449, "learning_rate": 9.279424452138924e-05, "loss": 0.0502, "step": 17150 }, { "grad_norm": 0.4746643304824829, "learning_rate": 9.278355187821517e-05, "loss": 0.0455, "step": 17160 }, { "grad_norm": 0.314298540353775, "learning_rate": 9.277285192450224e-05, "loss": 0.052, "step": 17170 }, { "grad_norm": 0.33604174852371216, "learning_rate": 9.276214466207875e-05, "loss": 0.0451, "step": 17180 }, { "grad_norm": 0.3651675879955292, "learning_rate": 9.275143009277427e-05, "loss": 0.0414, "step": 17190 }, { "grad_norm": 0.47247475385665894, "learning_rate": 9.274070821841964e-05, "loss": 0.0452, "step": 17200 }, { "grad_norm": 0.5929951667785645, "learning_rate": 9.272997904084696e-05, "loss": 0.0443, "step": 17210 }, { "grad_norm": 0.525265097618103, "learning_rate": 9.271924256188951e-05, "loss": 0.0437, "step": 17220 }, { "grad_norm": 0.41702139377593994, "learning_rate": 9.270849878338189e-05, "loss": 0.0433, "step": 17230 }, { "grad_norm": 0.47086599469184875, "learning_rate": 9.269774770715991e-05, "loss": 0.0496, "step": 17240 }, { "grad_norm": 0.3984179198741913, "learning_rate": 9.268698933506061e-05, "loss": 0.0551, "step": 17250 }, { "grad_norm": 0.3754305839538574, "learning_rate": 9.267622366892235e-05, "loss": 0.0396, "step": 17260 }, { "grad_norm": 0.43526363372802734, "learning_rate": 9.266545071058465e-05, "loss": 0.0427, "step": 17270 }, { "grad_norm": 0.3591809570789337, "learning_rate": 9.265467046188833e-05, "loss": 0.0395, "step": 17280 }, { "grad_norm": 0.46703261137008667, "learning_rate": 9.264388292467543e-05, "loss": 0.0521, "step": 17290 }, { "grad_norm": 0.2936357855796814, "learning_rate": 9.263308810078926e-05, "loss": 0.0419, "step": 17300 }, { "grad_norm": 0.33434489369392395, "learning_rate": 9.262228599207434e-05, "loss": 0.0465, "step": 17310 }, { "grad_norm": 0.39960870146751404, "learning_rate": 9.261147660037647e-05, "loss": 0.0313, "step": 17320 }, { "grad_norm": 0.28160327672958374, "learning_rate": 9.26006599275427e-05, "loss": 0.0387, "step": 17330 }, { "grad_norm": 0.4366476535797119, "learning_rate": 9.258983597542124e-05, "loss": 0.0442, "step": 17340 }, { "grad_norm": 0.45082083344459534, "learning_rate": 9.257900474586167e-05, "loss": 0.0435, "step": 17350 }, { "grad_norm": 0.42390620708465576, "learning_rate": 9.256816624071471e-05, "loss": 0.0489, "step": 17360 }, { "grad_norm": 0.5121392607688904, "learning_rate": 9.25573204618324e-05, "loss": 0.0472, "step": 17370 }, { "grad_norm": 0.5527360439300537, "learning_rate": 9.254646741106796e-05, "loss": 0.0482, "step": 17380 }, { "grad_norm": 0.27402645349502563, "learning_rate": 9.253560709027589e-05, "loss": 0.0429, "step": 17390 }, { "grad_norm": 0.4829094707965851, "learning_rate": 9.252473950131192e-05, "loss": 0.0503, "step": 17400 }, { "grad_norm": 0.4193277359008789, "learning_rate": 9.251386464603302e-05, "loss": 0.0417, "step": 17410 }, { "grad_norm": 0.5107531547546387, "learning_rate": 9.250298252629741e-05, "loss": 0.0515, "step": 17420 }, { "grad_norm": 0.5589893460273743, "learning_rate": 9.249209314396454e-05, "loss": 0.049, "step": 17430 }, { "grad_norm": 0.5346506834030151, "learning_rate": 9.248119650089513e-05, "loss": 0.048, "step": 17440 }, { "grad_norm": 0.4266258180141449, "learning_rate": 9.247029259895108e-05, "loss": 0.0471, "step": 17450 }, { "grad_norm": 0.34965404868125916, "learning_rate": 9.24593814399956e-05, "loss": 0.0449, "step": 17460 }, { "grad_norm": 0.40211114287376404, "learning_rate": 9.244846302589309e-05, "loss": 0.039, "step": 17470 }, { "grad_norm": 0.6297250390052795, "learning_rate": 9.243753735850923e-05, "loss": 0.0485, "step": 17480 }, { "grad_norm": 0.43853557109832764, "learning_rate": 9.24266044397109e-05, "loss": 0.0412, "step": 17490 }, { "grad_norm": 0.43931421637535095, "learning_rate": 9.241566427136624e-05, "loss": 0.0411, "step": 17500 }, { "grad_norm": 0.39488616585731506, "learning_rate": 9.240471685534463e-05, "loss": 0.0464, "step": 17510 }, { "grad_norm": 0.3664679229259491, "learning_rate": 9.239376219351667e-05, "loss": 0.0497, "step": 17520 }, { "grad_norm": 0.48180365562438965, "learning_rate": 9.238280028775425e-05, "loss": 0.0485, "step": 17530 }, { "grad_norm": 0.4713331460952759, "learning_rate": 9.237183113993041e-05, "loss": 0.0519, "step": 17540 }, { "grad_norm": 0.32249706983566284, "learning_rate": 9.236085475191952e-05, "loss": 0.0453, "step": 17550 }, { "grad_norm": 0.4545963406562805, "learning_rate": 9.234987112559709e-05, "loss": 0.0607, "step": 17560 }, { "grad_norm": 0.3661666512489319, "learning_rate": 9.233888026283999e-05, "loss": 0.0574, "step": 17570 }, { "grad_norm": 0.4459022581577301, "learning_rate": 9.232788216552619e-05, "loss": 0.0497, "step": 17580 }, { "grad_norm": 0.3776527941226959, "learning_rate": 9.231687683553502e-05, "loss": 0.0505, "step": 17590 }, { "grad_norm": 0.346103310585022, "learning_rate": 9.230586427474698e-05, "loss": 0.0597, "step": 17600 }, { "grad_norm": 0.3353917896747589, "learning_rate": 9.229484448504379e-05, "loss": 0.0534, "step": 17610 }, { "grad_norm": 0.41808539628982544, "learning_rate": 9.228381746830843e-05, "loss": 0.0484, "step": 17620 }, { "grad_norm": 0.4168696403503418, "learning_rate": 9.227278322642514e-05, "loss": 0.0419, "step": 17630 }, { "grad_norm": 0.35938361287117004, "learning_rate": 9.226174176127937e-05, "loss": 0.0473, "step": 17640 }, { "grad_norm": 0.38253864645957947, "learning_rate": 9.22506930747578e-05, "loss": 0.0471, "step": 17650 }, { "grad_norm": 0.4006974399089813, "learning_rate": 9.223963716874831e-05, "loss": 0.0639, "step": 17660 }, { "grad_norm": 0.5745075941085815, "learning_rate": 9.222857404514012e-05, "loss": 0.0536, "step": 17670 }, { "grad_norm": 0.4076533615589142, "learning_rate": 9.221750370582355e-05, "loss": 0.0484, "step": 17680 }, { "grad_norm": 0.37384435534477234, "learning_rate": 9.220642615269028e-05, "loss": 0.0457, "step": 17690 }, { "grad_norm": 0.4498443603515625, "learning_rate": 9.219534138763311e-05, "loss": 0.0534, "step": 17700 }, { "grad_norm": 0.5934979319572449, "learning_rate": 9.218424941254613e-05, "loss": 0.0506, "step": 17710 }, { "grad_norm": 0.4469815492630005, "learning_rate": 9.217315022932468e-05, "loss": 0.0544, "step": 17720 }, { "grad_norm": 0.3906957507133484, "learning_rate": 9.216204383986528e-05, "loss": 0.0446, "step": 17730 }, { "grad_norm": 0.35756608843803406, "learning_rate": 9.215093024606574e-05, "loss": 0.0528, "step": 17740 }, { "grad_norm": 0.3261098265647888, "learning_rate": 9.213980944982506e-05, "loss": 0.0652, "step": 17750 }, { "grad_norm": 0.5021231770515442, "learning_rate": 9.212868145304346e-05, "loss": 0.0492, "step": 17760 }, { "grad_norm": 0.41222208738327026, "learning_rate": 9.211754625762241e-05, "loss": 0.0413, "step": 17770 }, { "grad_norm": 0.3968563973903656, "learning_rate": 9.210640386546463e-05, "loss": 0.056, "step": 17780 }, { "grad_norm": 0.3694923520088196, "learning_rate": 9.209525427847405e-05, "loss": 0.0354, "step": 17790 }, { "grad_norm": 0.4841952323913574, "learning_rate": 9.208409749855583e-05, "loss": 0.044, "step": 17800 }, { "grad_norm": 0.577039897441864, "learning_rate": 9.207293352761633e-05, "loss": 0.0519, "step": 17810 }, { "grad_norm": 0.4591573476791382, "learning_rate": 9.206176236756319e-05, "loss": 0.0454, "step": 17820 }, { "grad_norm": 0.5120747685432434, "learning_rate": 9.205058402030525e-05, "loss": 0.0447, "step": 17830 }, { "grad_norm": 0.37144720554351807, "learning_rate": 9.203939848775259e-05, "loss": 0.0475, "step": 17840 }, { "grad_norm": 0.39717400074005127, "learning_rate": 9.202820577181652e-05, "loss": 0.0386, "step": 17850 }, { "grad_norm": 0.47220924496650696, "learning_rate": 9.201700587440953e-05, "loss": 0.0478, "step": 17860 }, { "grad_norm": 0.3904956579208374, "learning_rate": 9.200579879744544e-05, "loss": 0.0486, "step": 17870 }, { "grad_norm": 0.3612119257450104, "learning_rate": 9.199458454283918e-05, "loss": 0.0428, "step": 17880 }, { "grad_norm": 0.4318479895591736, "learning_rate": 9.198336311250697e-05, "loss": 0.054, "step": 17890 }, { "grad_norm": 0.47453224658966064, "learning_rate": 9.197213450836626e-05, "loss": 0.0418, "step": 17900 }, { "grad_norm": 0.48484647274017334, "learning_rate": 9.19608987323357e-05, "loss": 0.054, "step": 17910 }, { "grad_norm": 0.2858984172344208, "learning_rate": 9.194965578633517e-05, "loss": 0.0438, "step": 17920 }, { "grad_norm": 0.3555147647857666, "learning_rate": 9.193840567228582e-05, "loss": 0.0439, "step": 17930 }, { "grad_norm": 0.46921679377555847, "learning_rate": 9.192714839210994e-05, "loss": 0.046, "step": 17940 }, { "grad_norm": 0.34727177023887634, "learning_rate": 9.19158839477311e-05, "loss": 0.051, "step": 17950 }, { "grad_norm": 0.3017183542251587, "learning_rate": 9.190461234107411e-05, "loss": 0.0476, "step": 17960 }, { "grad_norm": 0.3307497799396515, "learning_rate": 9.189333357406496e-05, "loss": 0.038, "step": 17970 }, { "grad_norm": 0.4358021914958954, "learning_rate": 9.188204764863089e-05, "loss": 0.0618, "step": 17980 }, { "grad_norm": 0.5257874727249146, "learning_rate": 9.187075456670033e-05, "loss": 0.0399, "step": 17990 }, { "grad_norm": 0.3544803261756897, "learning_rate": 9.1859454330203e-05, "loss": 0.0456, "step": 18000 }, { "grad_norm": 0.3849443793296814, "learning_rate": 9.18481469410698e-05, "loss": 0.0568, "step": 18010 }, { "grad_norm": 0.4470137655735016, "learning_rate": 9.183683240123281e-05, "loss": 0.043, "step": 18020 }, { "grad_norm": 0.40113896131515503, "learning_rate": 9.182551071262541e-05, "loss": 0.0531, "step": 18030 }, { "grad_norm": 0.36523938179016113, "learning_rate": 9.181418187718218e-05, "loss": 0.0461, "step": 18040 }, { "grad_norm": 0.4203440546989441, "learning_rate": 9.180284589683888e-05, "loss": 0.0422, "step": 18050 }, { "grad_norm": 0.42607852816581726, "learning_rate": 9.17915027735325e-05, "loss": 0.0453, "step": 18060 }, { "grad_norm": 0.45590680837631226, "learning_rate": 9.178015250920133e-05, "loss": 0.044, "step": 18070 }, { "grad_norm": 0.4263000190258026, "learning_rate": 9.176879510578477e-05, "loss": 0.0445, "step": 18080 }, { "grad_norm": 0.31647491455078125, "learning_rate": 9.17574305652235e-05, "loss": 0.0465, "step": 18090 }, { "grad_norm": 0.3906029164791107, "learning_rate": 9.174605888945942e-05, "loss": 0.0541, "step": 18100 }, { "grad_norm": 0.3948797583580017, "learning_rate": 9.173468008043564e-05, "loss": 0.0427, "step": 18110 }, { "grad_norm": 0.416551798582077, "learning_rate": 9.172329414009648e-05, "loss": 0.0425, "step": 18120 }, { "grad_norm": 0.4006410241127014, "learning_rate": 9.171190107038747e-05, "loss": 0.0453, "step": 18130 }, { "grad_norm": 0.2680487334728241, "learning_rate": 9.170050087325541e-05, "loss": 0.035, "step": 18140 }, { "grad_norm": 0.5589796304702759, "learning_rate": 9.168909355064824e-05, "loss": 0.0427, "step": 18150 }, { "grad_norm": 0.2735098898410797, "learning_rate": 9.167767910451519e-05, "loss": 0.0405, "step": 18160 }, { "grad_norm": 0.37234580516815186, "learning_rate": 9.166625753680669e-05, "loss": 0.0421, "step": 18170 }, { "grad_norm": 0.37030887603759766, "learning_rate": 9.165482884947431e-05, "loss": 0.0377, "step": 18180 }, { "grad_norm": 0.3537011742591858, "learning_rate": 9.164339304447098e-05, "loss": 0.0468, "step": 18190 }, { "grad_norm": 0.32214877009391785, "learning_rate": 9.163195012375072e-05, "loss": 0.0388, "step": 18200 }, { "grad_norm": 0.6039819121360779, "learning_rate": 9.16205000892688e-05, "loss": 0.047, "step": 18210 }, { "grad_norm": 0.35538432002067566, "learning_rate": 9.160904294298175e-05, "loss": 0.0442, "step": 18220 }, { "grad_norm": 0.29548946022987366, "learning_rate": 9.159757868684727e-05, "loss": 0.0419, "step": 18230 }, { "grad_norm": 0.46034613251686096, "learning_rate": 9.15861073228243e-05, "loss": 0.0403, "step": 18240 }, { "grad_norm": 0.38978222012519836, "learning_rate": 9.157462885287296e-05, "loss": 0.0545, "step": 18250 }, { "grad_norm": 0.5844812393188477, "learning_rate": 9.156314327895461e-05, "loss": 0.046, "step": 18260 }, { "grad_norm": 0.4234762489795685, "learning_rate": 9.155165060303185e-05, "loss": 0.0516, "step": 18270 }, { "grad_norm": 0.32366377115249634, "learning_rate": 9.154015082706841e-05, "loss": 0.0387, "step": 18280 }, { "grad_norm": 0.48008978366851807, "learning_rate": 9.152864395302936e-05, "loss": 0.0479, "step": 18290 }, { "grad_norm": 0.5725422501564026, "learning_rate": 9.151712998288085e-05, "loss": 0.0464, "step": 18300 }, { "grad_norm": 0.3635484576225281, "learning_rate": 9.150560891859031e-05, "loss": 0.0421, "step": 18310 }, { "grad_norm": 0.4042969048023224, "learning_rate": 9.14940807621264e-05, "loss": 0.0462, "step": 18320 }, { "grad_norm": 0.44490331411361694, "learning_rate": 9.148254551545894e-05, "loss": 0.0501, "step": 18330 }, { "grad_norm": 0.3666524887084961, "learning_rate": 9.147100318055901e-05, "loss": 0.047, "step": 18340 }, { "grad_norm": 0.46952831745147705, "learning_rate": 9.145945375939888e-05, "loss": 0.048, "step": 18350 }, { "grad_norm": 0.5350469350814819, "learning_rate": 9.144789725395203e-05, "loss": 0.0441, "step": 18360 }, { "grad_norm": 0.3776175081729889, "learning_rate": 9.14363336661931e-05, "loss": 0.0427, "step": 18370 }, { "grad_norm": 0.38360580801963806, "learning_rate": 9.142476299809806e-05, "loss": 0.0448, "step": 18380 }, { "grad_norm": 0.3208144009113312, "learning_rate": 9.1413185251644e-05, "loss": 0.0358, "step": 18390 }, { "grad_norm": 0.31478366255760193, "learning_rate": 9.140160042880923e-05, "loss": 0.0519, "step": 18400 }, { "grad_norm": 0.3654134273529053, "learning_rate": 9.139000853157327e-05, "loss": 0.0557, "step": 18410 }, { "grad_norm": 0.39947330951690674, "learning_rate": 9.137840956191688e-05, "loss": 0.0577, "step": 18420 }, { "grad_norm": 0.293052077293396, "learning_rate": 9.136680352182199e-05, "loss": 0.0402, "step": 18430 }, { "grad_norm": 0.4447372257709503, "learning_rate": 9.135519041327177e-05, "loss": 0.041, "step": 18440 }, { "grad_norm": 0.3369760513305664, "learning_rate": 9.134357023825058e-05, "loss": 0.0521, "step": 18450 }, { "grad_norm": 0.4902631342411041, "learning_rate": 9.133194299874398e-05, "loss": 0.044, "step": 18460 }, { "grad_norm": 0.4107988178730011, "learning_rate": 9.132030869673876e-05, "loss": 0.0457, "step": 18470 }, { "grad_norm": 0.2528683543205261, "learning_rate": 9.130866733422288e-05, "loss": 0.0418, "step": 18480 }, { "grad_norm": 0.4020969867706299, "learning_rate": 9.129701891318556e-05, "loss": 0.0359, "step": 18490 }, { "grad_norm": 0.4140454828739166, "learning_rate": 9.128536343561718e-05, "loss": 0.0485, "step": 18500 }, { "grad_norm": 0.5321866869926453, "learning_rate": 9.127370090350934e-05, "loss": 0.0433, "step": 18510 }, { "grad_norm": 0.43026381731033325, "learning_rate": 9.126203131885487e-05, "loss": 0.0451, "step": 18520 }, { "grad_norm": 0.30244073271751404, "learning_rate": 9.125035468364775e-05, "loss": 0.0422, "step": 18530 }, { "grad_norm": 0.39508056640625, "learning_rate": 9.123867099988322e-05, "loss": 0.0368, "step": 18540 }, { "grad_norm": 0.381181538105011, "learning_rate": 9.122698026955769e-05, "loss": 0.0401, "step": 18550 }, { "grad_norm": 0.27848219871520996, "learning_rate": 9.12152824946688e-05, "loss": 0.0523, "step": 18560 }, { "grad_norm": 0.4633197784423828, "learning_rate": 9.120357767721538e-05, "loss": 0.0456, "step": 18570 }, { "grad_norm": 0.4981534481048584, "learning_rate": 9.119186581919745e-05, "loss": 0.0439, "step": 18580 }, { "grad_norm": 0.5147528648376465, "learning_rate": 9.118014692261624e-05, "loss": 0.0451, "step": 18590 }, { "grad_norm": 0.47986769676208496, "learning_rate": 9.116842098947422e-05, "loss": 0.0541, "step": 18600 }, { "grad_norm": 0.32791537046432495, "learning_rate": 9.115668802177499e-05, "loss": 0.0358, "step": 18610 }, { "grad_norm": 0.3344408869743347, "learning_rate": 9.114494802152342e-05, "loss": 0.0561, "step": 18620 }, { "grad_norm": 0.5441393256187439, "learning_rate": 9.113320099072555e-05, "loss": 0.0697, "step": 18630 }, { "grad_norm": 0.4057316184043884, "learning_rate": 9.112144693138864e-05, "loss": 0.0457, "step": 18640 }, { "grad_norm": 0.33202430605888367, "learning_rate": 9.110968584552111e-05, "loss": 0.0419, "step": 18650 }, { "grad_norm": 0.37603721022605896, "learning_rate": 9.109791773513264e-05, "loss": 0.045, "step": 18660 }, { "grad_norm": 0.3573337495326996, "learning_rate": 9.108614260223403e-05, "loss": 0.0469, "step": 18670 }, { "grad_norm": 0.41120338439941406, "learning_rate": 9.107436044883738e-05, "loss": 0.0453, "step": 18680 }, { "grad_norm": 0.3838746249675751, "learning_rate": 9.10625712769559e-05, "loss": 0.0425, "step": 18690 }, { "grad_norm": 0.27205732464790344, "learning_rate": 9.105077508860406e-05, "loss": 0.0316, "step": 18700 }, { "grad_norm": 0.3402130901813507, "learning_rate": 9.103897188579751e-05, "loss": 0.0365, "step": 18710 }, { "grad_norm": 0.2528117597103119, "learning_rate": 9.102716167055308e-05, "loss": 0.0321, "step": 18720 }, { "grad_norm": 0.44199350476264954, "learning_rate": 9.10153444448888e-05, "loss": 0.0676, "step": 18730 }, { "grad_norm": 0.37308377027511597, "learning_rate": 9.100352021082393e-05, "loss": 0.039, "step": 18740 }, { "grad_norm": 0.5521243810653687, "learning_rate": 9.099168897037891e-05, "loss": 0.035, "step": 18750 }, { "grad_norm": 0.44428330659866333, "learning_rate": 9.097985072557538e-05, "loss": 0.044, "step": 18760 }, { "grad_norm": 0.38120511174201965, "learning_rate": 9.096800547843615e-05, "loss": 0.0579, "step": 18770 }, { "grad_norm": 0.4320257604122162, "learning_rate": 9.095615323098526e-05, "loss": 0.0497, "step": 18780 }, { "grad_norm": 0.4016210436820984, "learning_rate": 9.094429398524795e-05, "loss": 0.0428, "step": 18790 }, { "grad_norm": 0.3907445967197418, "learning_rate": 9.093242774325061e-05, "loss": 0.0438, "step": 18800 }, { "grad_norm": 0.4039296805858612, "learning_rate": 9.092055450702088e-05, "loss": 0.0424, "step": 18810 }, { "grad_norm": 0.3910869061946869, "learning_rate": 9.090867427858756e-05, "loss": 0.0465, "step": 18820 }, { "grad_norm": 0.44339442253112793, "learning_rate": 9.089678705998066e-05, "loss": 0.0404, "step": 18830 }, { "grad_norm": 0.5187246799468994, "learning_rate": 9.088489285323139e-05, "loss": 0.0401, "step": 18840 }, { "grad_norm": 0.3473981022834778, "learning_rate": 9.087299166037212e-05, "loss": 0.0444, "step": 18850 }, { "grad_norm": 0.3581177890300751, "learning_rate": 9.086108348343647e-05, "loss": 0.045, "step": 18860 }, { "grad_norm": 0.3726074993610382, "learning_rate": 9.08491683244592e-05, "loss": 0.0428, "step": 18870 }, { "grad_norm": 0.2888478934764862, "learning_rate": 9.08372461854763e-05, "loss": 0.0369, "step": 18880 }, { "grad_norm": 0.3754815459251404, "learning_rate": 9.082531706852492e-05, "loss": 0.0384, "step": 18890 }, { "grad_norm": 0.29851868748664856, "learning_rate": 9.081338097564342e-05, "loss": 0.049, "step": 18900 }, { "grad_norm": 0.426614910364151, "learning_rate": 9.080143790887137e-05, "loss": 0.0379, "step": 18910 }, { "grad_norm": 0.3969997465610504, "learning_rate": 9.07894878702495e-05, "loss": 0.0343, "step": 18920 }, { "grad_norm": 0.45425185561180115, "learning_rate": 9.077753086181974e-05, "loss": 0.0468, "step": 18930 }, { "grad_norm": 0.40578898787498474, "learning_rate": 9.076556688562524e-05, "loss": 0.0354, "step": 18940 }, { "grad_norm": 0.3376113474369049, "learning_rate": 9.075359594371029e-05, "loss": 0.0328, "step": 18950 }, { "grad_norm": 0.40915289521217346, "learning_rate": 9.07416180381204e-05, "loss": 0.0457, "step": 18960 }, { "grad_norm": 0.3925960659980774, "learning_rate": 9.072963317090228e-05, "loss": 0.047, "step": 18970 }, { "grad_norm": 0.3116712272167206, "learning_rate": 9.071764134410382e-05, "loss": 0.0405, "step": 18980 }, { "grad_norm": 0.3816083073616028, "learning_rate": 9.070564255977407e-05, "loss": 0.0393, "step": 18990 }, { "grad_norm": 0.3613416850566864, "learning_rate": 9.06936368199633e-05, "loss": 0.046, "step": 19000 }, { "grad_norm": 0.42124244570732117, "learning_rate": 9.0681624126723e-05, "loss": 0.0429, "step": 19010 }, { "grad_norm": 0.4253132939338684, "learning_rate": 9.066960448210576e-05, "loss": 0.0359, "step": 19020 }, { "grad_norm": 0.6186023354530334, "learning_rate": 9.065757788816543e-05, "loss": 0.0393, "step": 19030 }, { "grad_norm": 0.452333003282547, "learning_rate": 9.064554434695705e-05, "loss": 0.0519, "step": 19040 }, { "grad_norm": 0.32338300347328186, "learning_rate": 9.063350386053677e-05, "loss": 0.0432, "step": 19050 }, { "grad_norm": 0.4271892309188843, "learning_rate": 9.062145643096202e-05, "loss": 0.043, "step": 19060 }, { "grad_norm": 0.4847141206264496, "learning_rate": 9.060940206029136e-05, "loss": 0.0483, "step": 19070 }, { "grad_norm": 0.41598525643348694, "learning_rate": 9.059734075058457e-05, "loss": 0.0507, "step": 19080 }, { "grad_norm": 0.3369375765323639, "learning_rate": 9.058527250390257e-05, "loss": 0.0456, "step": 19090 }, { "grad_norm": 0.45386847853660583, "learning_rate": 9.057319732230752e-05, "loss": 0.0493, "step": 19100 }, { "grad_norm": 0.30137506127357483, "learning_rate": 9.056111520786273e-05, "loss": 0.0408, "step": 19110 }, { "grad_norm": 0.22991997003555298, "learning_rate": 9.054902616263268e-05, "loss": 0.0358, "step": 19120 }, { "grad_norm": 0.3608854413032532, "learning_rate": 9.05369301886831e-05, "loss": 0.056, "step": 19130 }, { "grad_norm": 0.3725128769874573, "learning_rate": 9.052482728808083e-05, "loss": 0.0416, "step": 19140 }, { "grad_norm": 0.37393832206726074, "learning_rate": 9.051271746289391e-05, "loss": 0.0334, "step": 19150 }, { "grad_norm": 0.33893412351608276, "learning_rate": 9.050060071519162e-05, "loss": 0.0391, "step": 19160 }, { "grad_norm": 0.24287879467010498, "learning_rate": 9.048847704704437e-05, "loss": 0.0347, "step": 19170 }, { "grad_norm": 0.41802307963371277, "learning_rate": 9.047634646052376e-05, "loss": 0.0439, "step": 19180 }, { "grad_norm": 0.44936078786849976, "learning_rate": 9.046420895770256e-05, "loss": 0.0637, "step": 19190 }, { "grad_norm": 0.2718231678009033, "learning_rate": 9.045206454065473e-05, "loss": 0.0403, "step": 19200 }, { "grad_norm": 0.43449002504348755, "learning_rate": 9.043991321145546e-05, "loss": 0.0404, "step": 19210 }, { "grad_norm": 0.3844178020954132, "learning_rate": 9.042775497218105e-05, "loss": 0.0381, "step": 19220 }, { "grad_norm": 0.47673070430755615, "learning_rate": 9.041558982490901e-05, "loss": 0.0433, "step": 19230 }, { "grad_norm": 0.36020800471305847, "learning_rate": 9.040341777171805e-05, "loss": 0.0472, "step": 19240 }, { "grad_norm": 0.37268969416618347, "learning_rate": 9.039123881468802e-05, "loss": 0.0442, "step": 19250 }, { "grad_norm": 0.4230155050754547, "learning_rate": 9.037905295589998e-05, "loss": 0.0472, "step": 19260 }, { "grad_norm": 0.28076499700546265, "learning_rate": 9.036686019743617e-05, "loss": 0.0345, "step": 19270 }, { "grad_norm": 0.4091346859931946, "learning_rate": 9.035466054137997e-05, "loss": 0.0375, "step": 19280 }, { "grad_norm": 0.38286155462265015, "learning_rate": 9.0342453989816e-05, "loss": 0.0461, "step": 19290 }, { "grad_norm": 0.3727771043777466, "learning_rate": 9.033024054483e-05, "loss": 0.0438, "step": 19300 }, { "grad_norm": 0.42486366629600525, "learning_rate": 9.031802020850894e-05, "loss": 0.0416, "step": 19310 }, { "grad_norm": 0.4798310399055481, "learning_rate": 9.030579298294092e-05, "loss": 0.0539, "step": 19320 }, { "grad_norm": 0.4331081211566925, "learning_rate": 9.029355887021524e-05, "loss": 0.0554, "step": 19330 }, { "grad_norm": 0.4676852226257324, "learning_rate": 9.028131787242238e-05, "loss": 0.0601, "step": 19340 }, { "grad_norm": 0.3316482901573181, "learning_rate": 9.026906999165399e-05, "loss": 0.0523, "step": 19350 }, { "grad_norm": 0.3035166561603546, "learning_rate": 9.025681523000291e-05, "loss": 0.041, "step": 19360 }, { "grad_norm": 0.3647500276565552, "learning_rate": 9.024455358956315e-05, "loss": 0.0407, "step": 19370 }, { "grad_norm": 0.3767302930355072, "learning_rate": 9.023228507242984e-05, "loss": 0.0447, "step": 19380 }, { "grad_norm": 0.3961118757724762, "learning_rate": 9.022000968069937e-05, "loss": 0.0427, "step": 19390 }, { "grad_norm": 0.24218441545963287, "learning_rate": 9.020772741646928e-05, "loss": 0.0374, "step": 19400 }, { "grad_norm": 0.3090212643146515, "learning_rate": 9.019543828183826e-05, "loss": 0.0318, "step": 19410 }, { "grad_norm": 0.4355372190475464, "learning_rate": 9.018314227890616e-05, "loss": 0.0352, "step": 19420 }, { "grad_norm": 0.5045985579490662, "learning_rate": 9.017083940977408e-05, "loss": 0.0515, "step": 19430 }, { "grad_norm": 0.4036380648612976, "learning_rate": 9.015852967654422e-05, "loss": 0.0414, "step": 19440 }, { "grad_norm": 0.5341901183128357, "learning_rate": 9.014621308131996e-05, "loss": 0.0399, "step": 19450 }, { "grad_norm": 0.31643757224082947, "learning_rate": 9.01338896262059e-05, "loss": 0.0399, "step": 19460 }, { "grad_norm": 0.39851996302604675, "learning_rate": 9.012155931330777e-05, "loss": 0.0473, "step": 19470 }, { "grad_norm": 0.39220842719078064, "learning_rate": 9.010922214473246e-05, "loss": 0.0393, "step": 19480 }, { "grad_norm": 0.4328375458717346, "learning_rate": 9.009687812258808e-05, "loss": 0.0499, "step": 19490 }, { "grad_norm": 0.416694313287735, "learning_rate": 9.00845272489839e-05, "loss": 0.0379, "step": 19500 }, { "grad_norm": 0.5160975456237793, "learning_rate": 9.007216952603031e-05, "loss": 0.0526, "step": 19510 }, { "grad_norm": 0.3423212468624115, "learning_rate": 9.005980495583894e-05, "loss": 0.0366, "step": 19520 }, { "grad_norm": 0.3859868049621582, "learning_rate": 9.004743354052252e-05, "loss": 0.0345, "step": 19530 }, { "grad_norm": 0.38834983110427856, "learning_rate": 9.003505528219503e-05, "loss": 0.0403, "step": 19540 }, { "grad_norm": 0.43343114852905273, "learning_rate": 9.002267018297154e-05, "loss": 0.0427, "step": 19550 }, { "grad_norm": 0.28561484813690186, "learning_rate": 9.001027824496834e-05, "loss": 0.0468, "step": 19560 }, { "grad_norm": 0.24498526751995087, "learning_rate": 8.999787947030287e-05, "loss": 0.0369, "step": 19570 }, { "grad_norm": 0.3048973083496094, "learning_rate": 8.998547386109376e-05, "loss": 0.0467, "step": 19580 }, { "grad_norm": 0.36560919880867004, "learning_rate": 8.997306141946073e-05, "loss": 0.0383, "step": 19590 }, { "grad_norm": 0.4780881404876709, "learning_rate": 8.996064214752481e-05, "loss": 0.0484, "step": 19600 }, { "grad_norm": 0.427433043718338, "learning_rate": 8.994821604740806e-05, "loss": 0.0405, "step": 19610 }, { "grad_norm": 0.36430981755256653, "learning_rate": 8.993578312123377e-05, "loss": 0.0401, "step": 19620 }, { "grad_norm": 0.2986157536506653, "learning_rate": 8.992334337112639e-05, "loss": 0.0449, "step": 19630 }, { "grad_norm": 0.29766207933425903, "learning_rate": 8.991089679921154e-05, "loss": 0.0384, "step": 19640 }, { "grad_norm": 0.44817987084388733, "learning_rate": 8.989844340761599e-05, "loss": 0.046, "step": 19650 }, { "grad_norm": 0.38498401641845703, "learning_rate": 8.988598319846768e-05, "loss": 0.0409, "step": 19660 }, { "grad_norm": 0.42861494421958923, "learning_rate": 8.987351617389574e-05, "loss": 0.0404, "step": 19670 }, { "grad_norm": 0.26756441593170166, "learning_rate": 8.98610423360304e-05, "loss": 0.0367, "step": 19680 }, { "grad_norm": 0.3594866991043091, "learning_rate": 8.984856168700317e-05, "loss": 0.0519, "step": 19690 }, { "grad_norm": 0.30769267678260803, "learning_rate": 8.983607422894658e-05, "loss": 0.0441, "step": 19700 }, { "grad_norm": 0.3295173943042755, "learning_rate": 8.982357996399442e-05, "loss": 0.0451, "step": 19710 }, { "grad_norm": 0.4125346839427948, "learning_rate": 8.981107889428164e-05, "loss": 0.0429, "step": 19720 }, { "grad_norm": 0.387704461812973, "learning_rate": 8.979857102194428e-05, "loss": 0.0357, "step": 19730 }, { "grad_norm": 0.3428466022014618, "learning_rate": 8.978605634911968e-05, "loss": 0.0411, "step": 19740 }, { "grad_norm": 0.37864014506340027, "learning_rate": 8.977353487794616e-05, "loss": 0.0378, "step": 19750 }, { "grad_norm": 0.38317444920539856, "learning_rate": 8.976100661056334e-05, "loss": 0.0494, "step": 19760 }, { "grad_norm": 0.4220070242881775, "learning_rate": 8.974847154911197e-05, "loss": 0.0383, "step": 19770 }, { "grad_norm": 0.3511526584625244, "learning_rate": 8.973592969573393e-05, "loss": 0.0431, "step": 19780 }, { "grad_norm": 0.3907250761985779, "learning_rate": 8.972338105257228e-05, "loss": 0.0485, "step": 19790 }, { "grad_norm": 0.32624751329421997, "learning_rate": 8.971082562177125e-05, "loss": 0.035, "step": 19800 }, { "grad_norm": 0.34733739495277405, "learning_rate": 8.96982634054762e-05, "loss": 0.0388, "step": 19810 }, { "grad_norm": 0.2854984998703003, "learning_rate": 8.96856944058337e-05, "loss": 0.0402, "step": 19820 }, { "grad_norm": 0.4180918037891388, "learning_rate": 8.967311862499144e-05, "loss": 0.0337, "step": 19830 }, { "grad_norm": 0.3809911906719208, "learning_rate": 8.966053606509825e-05, "loss": 0.0385, "step": 19840 }, { "grad_norm": 0.5960827469825745, "learning_rate": 8.964794672830417e-05, "loss": 0.0565, "step": 19850 }, { "grad_norm": 0.45330703258514404, "learning_rate": 8.963535061676038e-05, "loss": 0.0545, "step": 19860 }, { "grad_norm": 0.4510144889354706, "learning_rate": 8.962274773261918e-05, "loss": 0.0436, "step": 19870 }, { "grad_norm": 0.3827895224094391, "learning_rate": 8.961013807803409e-05, "loss": 0.0513, "step": 19880 }, { "grad_norm": 0.26040416955947876, "learning_rate": 8.959752165515973e-05, "loss": 0.0471, "step": 19890 }, { "grad_norm": 0.3907918334007263, "learning_rate": 8.958489846615193e-05, "loss": 0.0622, "step": 19900 }, { "grad_norm": 0.523786723613739, "learning_rate": 8.957226851316762e-05, "loss": 0.0458, "step": 19910 }, { "grad_norm": 0.3377697467803955, "learning_rate": 8.955963179836493e-05, "loss": 0.042, "step": 19920 }, { "grad_norm": 0.47487035393714905, "learning_rate": 8.954698832390312e-05, "loss": 0.0496, "step": 19930 }, { "grad_norm": 0.4586860239505768, "learning_rate": 8.953433809194263e-05, "loss": 0.0444, "step": 19940 }, { "grad_norm": 0.42220598459243774, "learning_rate": 8.9521681104645e-05, "loss": 0.043, "step": 19950 }, { "grad_norm": 0.41198256611824036, "learning_rate": 8.9509017364173e-05, "loss": 0.0441, "step": 19960 }, { "grad_norm": 0.38917484879493713, "learning_rate": 8.949634687269052e-05, "loss": 0.0437, "step": 19970 }, { "grad_norm": 0.3824921250343323, "learning_rate": 8.948366963236259e-05, "loss": 0.0518, "step": 19980 }, { "grad_norm": 0.34154802560806274, "learning_rate": 8.947098564535538e-05, "loss": 0.0424, "step": 19990 }, { "grad_norm": 0.400522381067276, "learning_rate": 8.945829491383627e-05, "loss": 0.0445, "step": 20000 }, { "grad_norm": 0.42089319229125977, "learning_rate": 8.944559743997374e-05, "loss": 0.0424, "step": 20010 }, { "grad_norm": 0.4583476781845093, "learning_rate": 8.943289322593746e-05, "loss": 0.0411, "step": 20020 }, { "grad_norm": 0.4554859399795532, "learning_rate": 8.942018227389821e-05, "loss": 0.0403, "step": 20030 }, { "grad_norm": 0.4052519202232361, "learning_rate": 8.940746458602795e-05, "loss": 0.0601, "step": 20040 }, { "grad_norm": 0.3584510087966919, "learning_rate": 8.939474016449979e-05, "loss": 0.0477, "step": 20050 }, { "grad_norm": 0.30960696935653687, "learning_rate": 8.938200901148799e-05, "loss": 0.0395, "step": 20060 }, { "grad_norm": 0.5141444802284241, "learning_rate": 8.936927112916795e-05, "loss": 0.0485, "step": 20070 }, { "grad_norm": 0.3938404619693756, "learning_rate": 8.935652651971622e-05, "loss": 0.041, "step": 20080 }, { "grad_norm": 0.43486854434013367, "learning_rate": 8.934377518531052e-05, "loss": 0.0535, "step": 20090 }, { "grad_norm": 0.35364341735839844, "learning_rate": 8.933101712812967e-05, "loss": 0.0364, "step": 20100 }, { "grad_norm": 0.3806506097316742, "learning_rate": 8.931825235035374e-05, "loss": 0.045, "step": 20110 }, { "grad_norm": 0.468478798866272, "learning_rate": 8.930548085416382e-05, "loss": 0.0464, "step": 20120 }, { "grad_norm": 0.3677324056625366, "learning_rate": 8.92927026417422e-05, "loss": 0.0441, "step": 20130 }, { "grad_norm": 0.31933945417404175, "learning_rate": 8.92799177152724e-05, "loss": 0.036, "step": 20140 }, { "grad_norm": 0.3275756239891052, "learning_rate": 8.926712607693895e-05, "loss": 0.041, "step": 20150 }, { "grad_norm": 0.255050927400589, "learning_rate": 8.925432772892762e-05, "loss": 0.041, "step": 20160 }, { "grad_norm": 0.5525949597358704, "learning_rate": 8.924152267342529e-05, "loss": 0.0519, "step": 20170 }, { "grad_norm": 0.49652761220932007, "learning_rate": 8.922871091261998e-05, "loss": 0.0464, "step": 20180 }, { "grad_norm": 0.36945199966430664, "learning_rate": 8.92158924487009e-05, "loss": 0.0469, "step": 20190 }, { "grad_norm": 0.3766356110572815, "learning_rate": 8.920306728385834e-05, "loss": 0.0381, "step": 20200 }, { "grad_norm": 0.3643183410167694, "learning_rate": 8.919023542028379e-05, "loss": 0.0414, "step": 20210 }, { "grad_norm": 0.3533569872379303, "learning_rate": 8.917739686016988e-05, "loss": 0.0398, "step": 20220 }, { "grad_norm": 0.40675315260887146, "learning_rate": 8.916455160571033e-05, "loss": 0.0555, "step": 20230 }, { "grad_norm": 0.4598707854747772, "learning_rate": 8.915169965910008e-05, "loss": 0.0475, "step": 20240 }, { "grad_norm": 0.5491113066673279, "learning_rate": 8.913884102253514e-05, "loss": 0.0424, "step": 20250 }, { "grad_norm": 0.29689711332321167, "learning_rate": 8.912597569821273e-05, "loss": 0.0439, "step": 20260 }, { "grad_norm": 0.3093806207180023, "learning_rate": 8.911310368833118e-05, "loss": 0.0401, "step": 20270 }, { "grad_norm": 0.280629426240921, "learning_rate": 8.910022499508994e-05, "loss": 0.0354, "step": 20280 }, { "grad_norm": 0.2981836497783661, "learning_rate": 8.908733962068965e-05, "loss": 0.047, "step": 20290 }, { "grad_norm": 0.27566447854042053, "learning_rate": 8.907444756733207e-05, "loss": 0.042, "step": 20300 }, { "grad_norm": 0.4141961634159088, "learning_rate": 8.906154883722006e-05, "loss": 0.05, "step": 20310 }, { "grad_norm": 0.34922319650650024, "learning_rate": 8.904864343255773e-05, "loss": 0.0456, "step": 20320 }, { "grad_norm": 0.4518245756626129, "learning_rate": 8.90357313555502e-05, "loss": 0.0408, "step": 20330 }, { "grad_norm": 0.4281523525714874, "learning_rate": 8.90228126084038e-05, "loss": 0.051, "step": 20340 }, { "grad_norm": 0.4067462384700775, "learning_rate": 8.900988719332601e-05, "loss": 0.0481, "step": 20350 }, { "grad_norm": 0.2793036997318268, "learning_rate": 8.899695511252542e-05, "loss": 0.0449, "step": 20360 }, { "grad_norm": 0.2554127275943756, "learning_rate": 8.898401636821176e-05, "loss": 0.0489, "step": 20370 }, { "grad_norm": 0.3153645992279053, "learning_rate": 8.897107096259593e-05, "loss": 0.0584, "step": 20380 }, { "grad_norm": 0.3920586109161377, "learning_rate": 8.895811889788994e-05, "loss": 0.0429, "step": 20390 }, { "grad_norm": 0.22751308977603912, "learning_rate": 8.894516017630692e-05, "loss": 0.0362, "step": 20400 }, { "grad_norm": 0.644084632396698, "learning_rate": 8.893219480006118e-05, "loss": 0.059, "step": 20410 }, { "grad_norm": 0.3315582573413849, "learning_rate": 8.891922277136817e-05, "loss": 0.0414, "step": 20420 }, { "grad_norm": 0.49869224429130554, "learning_rate": 8.890624409244441e-05, "loss": 0.0383, "step": 20430 }, { "grad_norm": 0.38611891865730286, "learning_rate": 8.889325876550763e-05, "loss": 0.0464, "step": 20440 }, { "grad_norm": 0.349631130695343, "learning_rate": 8.888026679277666e-05, "loss": 0.0479, "step": 20450 }, { "grad_norm": 0.3932313621044159, "learning_rate": 8.886726817647147e-05, "loss": 0.0603, "step": 20460 }, { "grad_norm": 0.2654241621494293, "learning_rate": 8.885426291881319e-05, "loss": 0.0548, "step": 20470 }, { "grad_norm": 0.4541614055633545, "learning_rate": 8.884125102202401e-05, "loss": 0.0496, "step": 20480 }, { "grad_norm": 0.45806583762168884, "learning_rate": 8.882823248832736e-05, "loss": 0.0457, "step": 20490 }, { "grad_norm": 0.30538409948349, "learning_rate": 8.881520731994772e-05, "loss": 0.041, "step": 20500 }, { "grad_norm": 0.37396806478500366, "learning_rate": 8.880217551911077e-05, "loss": 0.041, "step": 20510 }, { "grad_norm": 0.3051307797431946, "learning_rate": 8.878913708804323e-05, "loss": 0.042, "step": 20520 }, { "grad_norm": 0.2981354296207428, "learning_rate": 8.877609202897308e-05, "loss": 0.0514, "step": 20530 }, { "grad_norm": 0.3865205943584442, "learning_rate": 8.876304034412933e-05, "loss": 0.0437, "step": 20540 }, { "grad_norm": 0.3953356444835663, "learning_rate": 8.874998203574214e-05, "loss": 0.0521, "step": 20550 }, { "grad_norm": 0.5168335437774658, "learning_rate": 8.873691710604284e-05, "loss": 0.0474, "step": 20560 }, { "grad_norm": 0.5399808287620544, "learning_rate": 8.872384555726387e-05, "loss": 0.0441, "step": 20570 }, { "grad_norm": 0.3928869068622589, "learning_rate": 8.871076739163878e-05, "loss": 0.0404, "step": 20580 }, { "grad_norm": 0.35954761505126953, "learning_rate": 8.86976826114023e-05, "loss": 0.0424, "step": 20590 }, { "grad_norm": 0.3302852511405945, "learning_rate": 8.868459121879023e-05, "loss": 0.0516, "step": 20600 }, { "grad_norm": 0.4456954002380371, "learning_rate": 8.867149321603956e-05, "loss": 0.0558, "step": 20610 }, { "grad_norm": 0.43692341446876526, "learning_rate": 8.865838860538835e-05, "loss": 0.0462, "step": 20620 }, { "grad_norm": 0.34937888383865356, "learning_rate": 8.864527738907585e-05, "loss": 0.0419, "step": 20630 }, { "grad_norm": 0.41840243339538574, "learning_rate": 8.863215956934239e-05, "loss": 0.0497, "step": 20640 }, { "grad_norm": 0.28806430101394653, "learning_rate": 8.861903514842947e-05, "loss": 0.0307, "step": 20650 }, { "grad_norm": 0.31348147988319397, "learning_rate": 8.860590412857966e-05, "loss": 0.0341, "step": 20660 }, { "grad_norm": 0.42511752247810364, "learning_rate": 8.85927665120367e-05, "loss": 0.0406, "step": 20670 }, { "grad_norm": 0.4101243019104004, "learning_rate": 8.857962230104546e-05, "loss": 0.0476, "step": 20680 }, { "grad_norm": 0.5143110752105713, "learning_rate": 8.856647149785193e-05, "loss": 0.0461, "step": 20690 }, { "grad_norm": 0.31499186158180237, "learning_rate": 8.855331410470322e-05, "loss": 0.0464, "step": 20700 }, { "grad_norm": 0.4482618570327759, "learning_rate": 8.854015012384756e-05, "loss": 0.0369, "step": 20710 }, { "grad_norm": 0.39029839634895325, "learning_rate": 8.852697955753433e-05, "loss": 0.0417, "step": 20720 }, { "grad_norm": 0.3368028402328491, "learning_rate": 8.851380240801399e-05, "loss": 0.0486, "step": 20730 }, { "grad_norm": 0.30617523193359375, "learning_rate": 8.850061867753818e-05, "loss": 0.0355, "step": 20740 }, { "grad_norm": 0.3456473648548126, "learning_rate": 8.848742836835963e-05, "loss": 0.0588, "step": 20750 }, { "grad_norm": 0.45062053203582764, "learning_rate": 8.847423148273221e-05, "loss": 0.0431, "step": 20760 }, { "grad_norm": 0.4744965136051178, "learning_rate": 8.846102802291092e-05, "loss": 0.0582, "step": 20770 }, { "grad_norm": 0.3918915092945099, "learning_rate": 8.844781799115183e-05, "loss": 0.0516, "step": 20780 }, { "grad_norm": 0.3925683796405792, "learning_rate": 8.84346013897122e-05, "loss": 0.0364, "step": 20790 }, { "grad_norm": 0.3703937232494354, "learning_rate": 8.842137822085038e-05, "loss": 0.0524, "step": 20800 }, { "grad_norm": 0.31951749324798584, "learning_rate": 8.840814848682585e-05, "loss": 0.0353, "step": 20810 }, { "grad_norm": 0.4079970419406891, "learning_rate": 8.83949121898992e-05, "loss": 0.0535, "step": 20820 }, { "grad_norm": 0.3418506681919098, "learning_rate": 8.838166933233217e-05, "loss": 0.0413, "step": 20830 }, { "grad_norm": 0.31880834698677063, "learning_rate": 8.83684199163876e-05, "loss": 0.0436, "step": 20840 }, { "grad_norm": 0.4193747341632843, "learning_rate": 8.835516394432943e-05, "loss": 0.046, "step": 20850 }, { "grad_norm": 0.4736286997795105, "learning_rate": 8.834190141842276e-05, "loss": 0.0462, "step": 20860 }, { "grad_norm": 0.4005904197692871, "learning_rate": 8.83286323409338e-05, "loss": 0.0432, "step": 20870 }, { "grad_norm": 0.3923864960670471, "learning_rate": 8.831535671412986e-05, "loss": 0.0453, "step": 20880 }, { "grad_norm": 0.3008451759815216, "learning_rate": 8.830207454027938e-05, "loss": 0.0499, "step": 20890 }, { "grad_norm": 0.420369029045105, "learning_rate": 8.828878582165192e-05, "loss": 0.0386, "step": 20900 }, { "grad_norm": 0.43533632159233093, "learning_rate": 8.827549056051818e-05, "loss": 0.042, "step": 20910 }, { "grad_norm": 0.4246492385864258, "learning_rate": 8.826218875914993e-05, "loss": 0.0442, "step": 20920 }, { "grad_norm": 0.3505193591117859, "learning_rate": 8.82488804198201e-05, "loss": 0.0416, "step": 20930 }, { "grad_norm": 0.29105156660079956, "learning_rate": 8.82355655448027e-05, "loss": 0.0413, "step": 20940 }, { "grad_norm": 0.48125025629997253, "learning_rate": 8.822224413637293e-05, "loss": 0.0476, "step": 20950 }, { "grad_norm": 0.3808507025241852, "learning_rate": 8.820891619680697e-05, "loss": 0.0399, "step": 20960 }, { "grad_norm": 0.3213857412338257, "learning_rate": 8.819558172838227e-05, "loss": 0.0407, "step": 20970 }, { "grad_norm": 0.34955185651779175, "learning_rate": 8.818224073337731e-05, "loss": 0.0379, "step": 20980 }, { "grad_norm": 0.4470818042755127, "learning_rate": 8.816889321407169e-05, "loss": 0.041, "step": 20990 }, { "grad_norm": 0.3247339427471161, "learning_rate": 8.815553917274615e-05, "loss": 0.0449, "step": 21000 }, { "grad_norm": 0.34057772159576416, "learning_rate": 8.81421786116825e-05, "loss": 0.0471, "step": 21010 }, { "grad_norm": 0.4255092740058899, "learning_rate": 8.812881153316373e-05, "loss": 0.0405, "step": 21020 }, { "grad_norm": 0.3734200596809387, "learning_rate": 8.81154379394739e-05, "loss": 0.0452, "step": 21030 }, { "grad_norm": 0.4055563807487488, "learning_rate": 8.810205783289818e-05, "loss": 0.0484, "step": 21040 }, { "grad_norm": 0.31297409534454346, "learning_rate": 8.808867121572286e-05, "loss": 0.0432, "step": 21050 }, { "grad_norm": 0.34992390871047974, "learning_rate": 8.807527809023537e-05, "loss": 0.0539, "step": 21060 }, { "grad_norm": 0.3509614169597626, "learning_rate": 8.80618784587242e-05, "loss": 0.0398, "step": 21070 }, { "grad_norm": 0.38835057616233826, "learning_rate": 8.804847232347902e-05, "loss": 0.04, "step": 21080 }, { "grad_norm": 0.3832983374595642, "learning_rate": 8.803505968679054e-05, "loss": 0.0465, "step": 21090 }, { "grad_norm": 0.3430365025997162, "learning_rate": 8.802164055095061e-05, "loss": 0.0517, "step": 21100 }, { "grad_norm": 0.4143311381340027, "learning_rate": 8.80082149182522e-05, "loss": 0.0379, "step": 21110 }, { "grad_norm": 0.37317395210266113, "learning_rate": 8.79947827909894e-05, "loss": 0.0382, "step": 21120 }, { "grad_norm": 0.3410966694355011, "learning_rate": 8.798134417145738e-05, "loss": 0.0512, "step": 21130 }, { "grad_norm": 0.38124364614486694, "learning_rate": 8.796789906195243e-05, "loss": 0.0426, "step": 21140 }, { "grad_norm": 0.31873053312301636, "learning_rate": 8.795444746477195e-05, "loss": 0.039, "step": 21150 }, { "grad_norm": 0.4050602316856384, "learning_rate": 8.794098938221446e-05, "loss": 0.041, "step": 21160 }, { "grad_norm": 0.3586251139640808, "learning_rate": 8.792752481657957e-05, "loss": 0.0382, "step": 21170 }, { "grad_norm": 0.4076809585094452, "learning_rate": 8.791405377016802e-05, "loss": 0.0375, "step": 21180 }, { "grad_norm": 0.4166826009750366, "learning_rate": 8.790057624528163e-05, "loss": 0.041, "step": 21190 }, { "grad_norm": 0.40960222482681274, "learning_rate": 8.788709224422333e-05, "loss": 0.0475, "step": 21200 }, { "grad_norm": 0.33329784870147705, "learning_rate": 8.787360176929717e-05, "loss": 0.0369, "step": 21210 }, { "grad_norm": 0.6371012330055237, "learning_rate": 8.786010482280834e-05, "loss": 0.0428, "step": 21220 }, { "grad_norm": 0.2961939573287964, "learning_rate": 8.784660140706306e-05, "loss": 0.0468, "step": 21230 }, { "grad_norm": 0.26791974902153015, "learning_rate": 8.783309152436872e-05, "loss": 0.0387, "step": 21240 }, { "grad_norm": 0.25324487686157227, "learning_rate": 8.781957517703375e-05, "loss": 0.0356, "step": 21250 }, { "grad_norm": 0.3960321843624115, "learning_rate": 8.780605236736776e-05, "loss": 0.0469, "step": 21260 }, { "grad_norm": 0.2949391305446625, "learning_rate": 8.779252309768142e-05, "loss": 0.0508, "step": 21270 }, { "grad_norm": 0.34569036960601807, "learning_rate": 8.777898737028652e-05, "loss": 0.0483, "step": 21280 }, { "grad_norm": 0.47781771421432495, "learning_rate": 8.776544518749591e-05, "loss": 0.0413, "step": 21290 }, { "grad_norm": 0.40496665239334106, "learning_rate": 8.775189655162364e-05, "loss": 0.0422, "step": 21300 }, { "grad_norm": 0.48264461755752563, "learning_rate": 8.773834146498474e-05, "loss": 0.0413, "step": 21310 }, { "grad_norm": 0.39800506830215454, "learning_rate": 8.772477992989545e-05, "loss": 0.035, "step": 21320 }, { "grad_norm": 0.518764853477478, "learning_rate": 8.771121194867304e-05, "loss": 0.0496, "step": 21330 }, { "grad_norm": 0.31574663519859314, "learning_rate": 8.769763752363589e-05, "loss": 0.037, "step": 21340 }, { "grad_norm": 0.3175544738769531, "learning_rate": 8.768405665710352e-05, "loss": 0.0396, "step": 21350 }, { "grad_norm": 0.4163910746574402, "learning_rate": 8.767046935139655e-05, "loss": 0.0425, "step": 21360 }, { "grad_norm": 0.30809471011161804, "learning_rate": 8.765687560883666e-05, "loss": 0.04, "step": 21370 }, { "grad_norm": 0.558050274848938, "learning_rate": 8.764327543174664e-05, "loss": 0.0389, "step": 21380 }, { "grad_norm": 0.47053563594818115, "learning_rate": 8.762966882245038e-05, "loss": 0.0471, "step": 21390 }, { "grad_norm": 0.4496358036994934, "learning_rate": 8.761605578327291e-05, "loss": 0.0496, "step": 21400 }, { "grad_norm": 0.34919261932373047, "learning_rate": 8.76024363165403e-05, "loss": 0.0522, "step": 21410 }, { "grad_norm": 0.3304552733898163, "learning_rate": 8.758881042457976e-05, "loss": 0.0428, "step": 21420 }, { "grad_norm": 0.36418119072914124, "learning_rate": 8.757517810971957e-05, "loss": 0.0391, "step": 21430 }, { "grad_norm": 0.31919944286346436, "learning_rate": 8.756153937428913e-05, "loss": 0.0423, "step": 21440 }, { "grad_norm": 0.2733667492866516, "learning_rate": 8.754789422061889e-05, "loss": 0.0482, "step": 21450 }, { "grad_norm": 0.39597684144973755, "learning_rate": 8.753424265104052e-05, "loss": 0.0415, "step": 21460 }, { "grad_norm": 0.45799776911735535, "learning_rate": 8.752058466788659e-05, "loss": 0.0483, "step": 21470 }, { "grad_norm": 0.3721894323825836, "learning_rate": 8.750692027349097e-05, "loss": 0.0381, "step": 21480 }, { "grad_norm": 0.47384440898895264, "learning_rate": 8.749324947018847e-05, "loss": 0.049, "step": 21490 }, { "grad_norm": 0.3436046838760376, "learning_rate": 8.747957226031507e-05, "loss": 0.0425, "step": 21500 }, { "grad_norm": 0.42051997780799866, "learning_rate": 8.746588864620787e-05, "loss": 0.0454, "step": 21510 }, { "grad_norm": 0.39613842964172363, "learning_rate": 8.745219863020498e-05, "loss": 0.0412, "step": 21520 }, { "grad_norm": 0.4400048851966858, "learning_rate": 8.743850221464564e-05, "loss": 0.0422, "step": 21530 }, { "grad_norm": 0.44960999488830566, "learning_rate": 8.742479940187026e-05, "loss": 0.038, "step": 21540 }, { "grad_norm": 0.37469175457954407, "learning_rate": 8.74110901942202e-05, "loss": 0.0408, "step": 21550 }, { "grad_norm": 0.29895710945129395, "learning_rate": 8.739737459403803e-05, "loss": 0.0408, "step": 21560 }, { "grad_norm": 0.33055418729782104, "learning_rate": 8.738365260366737e-05, "loss": 0.0382, "step": 21570 }, { "grad_norm": 0.32193732261657715, "learning_rate": 8.736992422545292e-05, "loss": 0.0377, "step": 21580 }, { "grad_norm": 0.3136952221393585, "learning_rate": 8.73561894617405e-05, "loss": 0.0508, "step": 21590 }, { "grad_norm": 0.2966558635234833, "learning_rate": 8.734244831487697e-05, "loss": 0.0413, "step": 21600 }, { "grad_norm": 0.39792412519454956, "learning_rate": 8.732870078721035e-05, "loss": 0.0517, "step": 21610 }, { "grad_norm": 0.4109293818473816, "learning_rate": 8.731494688108972e-05, "loss": 0.0439, "step": 21620 }, { "grad_norm": 0.367432564496994, "learning_rate": 8.730118659886523e-05, "loss": 0.0416, "step": 21630 }, { "grad_norm": 0.37633174657821655, "learning_rate": 8.728741994288814e-05, "loss": 0.0401, "step": 21640 }, { "grad_norm": 0.46089741587638855, "learning_rate": 8.727364691551079e-05, "loss": 0.0371, "step": 21650 }, { "grad_norm": 0.3592952489852905, "learning_rate": 8.725986751908661e-05, "loss": 0.045, "step": 21660 }, { "grad_norm": 0.3154095411300659, "learning_rate": 8.724608175597016e-05, "loss": 0.0422, "step": 21670 }, { "grad_norm": 0.38313189148902893, "learning_rate": 8.723228962851699e-05, "loss": 0.0361, "step": 21680 }, { "grad_norm": 0.44766736030578613, "learning_rate": 8.721849113908385e-05, "loss": 0.0456, "step": 21690 }, { "grad_norm": 0.4576180875301361, "learning_rate": 8.720468629002848e-05, "loss": 0.0395, "step": 21700 }, { "grad_norm": 0.3592998683452606, "learning_rate": 8.719087508370978e-05, "loss": 0.043, "step": 21710 }, { "grad_norm": 0.5045570731163025, "learning_rate": 8.717705752248772e-05, "loss": 0.0528, "step": 21720 }, { "grad_norm": 0.39559173583984375, "learning_rate": 8.71632336087233e-05, "loss": 0.0445, "step": 21730 }, { "grad_norm": 0.38040032982826233, "learning_rate": 8.71494033447787e-05, "loss": 0.044, "step": 21740 }, { "grad_norm": 0.36753854155540466, "learning_rate": 8.713556673301708e-05, "loss": 0.0366, "step": 21750 }, { "grad_norm": 0.3345470726490021, "learning_rate": 8.712172377580278e-05, "loss": 0.0561, "step": 21760 }, { "grad_norm": 0.3832482099533081, "learning_rate": 8.710787447550114e-05, "loss": 0.0524, "step": 21770 }, { "grad_norm": 0.2852039337158203, "learning_rate": 8.70940188344787e-05, "loss": 0.057, "step": 21780 }, { "grad_norm": 0.406660258769989, "learning_rate": 8.708015685510293e-05, "loss": 0.0431, "step": 21790 }, { "grad_norm": 0.48122528195381165, "learning_rate": 8.706628853974252e-05, "loss": 0.0527, "step": 21800 }, { "grad_norm": 0.37612074613571167, "learning_rate": 8.705241389076715e-05, "loss": 0.0366, "step": 21810 }, { "grad_norm": 0.2610594630241394, "learning_rate": 8.703853291054764e-05, "loss": 0.0356, "step": 21820 }, { "grad_norm": 0.5062795877456665, "learning_rate": 8.702464560145587e-05, "loss": 0.0481, "step": 21830 }, { "grad_norm": 0.3348817229270935, "learning_rate": 8.701075196586476e-05, "loss": 0.0373, "step": 21840 }, { "grad_norm": 0.341550350189209, "learning_rate": 8.699685200614842e-05, "loss": 0.0384, "step": 21850 }, { "grad_norm": 0.3907760977745056, "learning_rate": 8.698294572468193e-05, "loss": 0.0505, "step": 21860 }, { "grad_norm": 0.2589126229286194, "learning_rate": 8.696903312384148e-05, "loss": 0.0427, "step": 21870 }, { "grad_norm": 0.3681630492210388, "learning_rate": 8.695511420600439e-05, "loss": 0.042, "step": 21880 }, { "grad_norm": 0.3733985126018524, "learning_rate": 8.694118897354901e-05, "loss": 0.0371, "step": 21890 }, { "grad_norm": 0.3642220199108124, "learning_rate": 8.692725742885478e-05, "loss": 0.0372, "step": 21900 }, { "grad_norm": 0.43001943826675415, "learning_rate": 8.691331957430221e-05, "loss": 0.0405, "step": 21910 }, { "grad_norm": 0.440024733543396, "learning_rate": 8.68993754122729e-05, "loss": 0.0395, "step": 21920 }, { "grad_norm": 0.4081821143627167, "learning_rate": 8.688542494514955e-05, "loss": 0.0392, "step": 21930 }, { "grad_norm": 0.474967360496521, "learning_rate": 8.68714681753159e-05, "loss": 0.0406, "step": 21940 }, { "grad_norm": 0.5901462435722351, "learning_rate": 8.685750510515676e-05, "loss": 0.0365, "step": 21950 }, { "grad_norm": 0.45236632227897644, "learning_rate": 8.684353573705805e-05, "loss": 0.0471, "step": 21960 }, { "grad_norm": 0.4270511865615845, "learning_rate": 8.682956007340677e-05, "loss": 0.0433, "step": 21970 }, { "grad_norm": 0.5176544189453125, "learning_rate": 8.681557811659095e-05, "loss": 0.0396, "step": 21980 }, { "grad_norm": 0.349273681640625, "learning_rate": 8.680158986899974e-05, "loss": 0.0356, "step": 21990 }, { "grad_norm": 0.38764578104019165, "learning_rate": 8.678759533302335e-05, "loss": 0.0393, "step": 22000 }, { "grad_norm": 0.31418541073799133, "learning_rate": 8.677359451105308e-05, "loss": 0.0361, "step": 22010 }, { "grad_norm": 0.38115939497947693, "learning_rate": 8.675958740548123e-05, "loss": 0.0456, "step": 22020 }, { "grad_norm": 0.4455114006996155, "learning_rate": 8.674557401870129e-05, "loss": 0.0518, "step": 22030 }, { "grad_norm": 0.333416610956192, "learning_rate": 8.673155435310775e-05, "loss": 0.0435, "step": 22040 }, { "grad_norm": 0.3142000436782837, "learning_rate": 8.671752841109617e-05, "loss": 0.0385, "step": 22050 }, { "grad_norm": 0.41842135787010193, "learning_rate": 8.670349619506321e-05, "loss": 0.0365, "step": 22060 }, { "grad_norm": 0.3800038695335388, "learning_rate": 8.66894577074066e-05, "loss": 0.0343, "step": 22070 }, { "grad_norm": 0.3686750829219818, "learning_rate": 8.667541295052513e-05, "loss": 0.0368, "step": 22080 }, { "grad_norm": 0.3050124943256378, "learning_rate": 8.666136192681865e-05, "loss": 0.0485, "step": 22090 }, { "grad_norm": 0.4265190064907074, "learning_rate": 8.664730463868811e-05, "loss": 0.041, "step": 22100 }, { "grad_norm": 0.4303155243396759, "learning_rate": 8.663324108853552e-05, "loss": 0.0458, "step": 22110 }, { "grad_norm": 0.3570331335067749, "learning_rate": 8.661917127876395e-05, "loss": 0.052, "step": 22120 }, { "grad_norm": 0.36905020475387573, "learning_rate": 8.660509521177754e-05, "loss": 0.0461, "step": 22130 }, { "grad_norm": 0.2869006395339966, "learning_rate": 8.65910128899815e-05, "loss": 0.0434, "step": 22140 }, { "grad_norm": 0.35252127051353455, "learning_rate": 8.657692431578214e-05, "loss": 0.0409, "step": 22150 }, { "grad_norm": 0.3613854944705963, "learning_rate": 8.656282949158679e-05, "loss": 0.0345, "step": 22160 }, { "grad_norm": 0.3555740714073181, "learning_rate": 8.654872841980388e-05, "loss": 0.0342, "step": 22170 }, { "grad_norm": 0.4200621247291565, "learning_rate": 8.653462110284289e-05, "loss": 0.041, "step": 22180 }, { "grad_norm": 0.3773014545440674, "learning_rate": 8.652050754311437e-05, "loss": 0.0422, "step": 22190 }, { "grad_norm": 0.2679567039012909, "learning_rate": 8.650638774302995e-05, "loss": 0.039, "step": 22200 }, { "grad_norm": 0.4530584216117859, "learning_rate": 8.649226170500233e-05, "loss": 0.0397, "step": 22210 }, { "grad_norm": 0.39609286189079285, "learning_rate": 8.647812943144524e-05, "loss": 0.0384, "step": 22220 }, { "grad_norm": 0.3236865997314453, "learning_rate": 8.646399092477351e-05, "loss": 0.0356, "step": 22230 }, { "grad_norm": 0.538063108921051, "learning_rate": 8.644984618740301e-05, "loss": 0.0471, "step": 22240 }, { "grad_norm": 0.34312504529953003, "learning_rate": 8.643569522175073e-05, "loss": 0.0314, "step": 22250 }, { "grad_norm": 0.4421459138393402, "learning_rate": 8.642153803023463e-05, "loss": 0.0418, "step": 22260 }, { "grad_norm": 0.34835365414619446, "learning_rate": 8.640737461527383e-05, "loss": 0.0351, "step": 22270 }, { "grad_norm": 0.37220489978790283, "learning_rate": 8.639320497928845e-05, "loss": 0.0436, "step": 22280 }, { "grad_norm": 0.4148053824901581, "learning_rate": 8.637902912469969e-05, "loss": 0.0441, "step": 22290 }, { "grad_norm": 0.4836260974407196, "learning_rate": 8.636484705392982e-05, "loss": 0.0415, "step": 22300 }, { "grad_norm": 0.2892773151397705, "learning_rate": 8.635065876940216e-05, "loss": 0.04, "step": 22310 }, { "grad_norm": 0.4370537996292114, "learning_rate": 8.633646427354112e-05, "loss": 0.0524, "step": 22320 }, { "grad_norm": 0.263750284910202, "learning_rate": 8.632226356877213e-05, "loss": 0.0344, "step": 22330 }, { "grad_norm": 0.3663138747215271, "learning_rate": 8.630805665752173e-05, "loss": 0.0478, "step": 22340 }, { "grad_norm": 0.31544771790504456, "learning_rate": 8.629384354221748e-05, "loss": 0.0475, "step": 22350 }, { "grad_norm": 0.28082889318466187, "learning_rate": 8.627962422528797e-05, "loss": 0.0384, "step": 22360 }, { "grad_norm": 0.24584394693374634, "learning_rate": 8.626539870916296e-05, "loss": 0.0546, "step": 22370 }, { "grad_norm": 0.23329463601112366, "learning_rate": 8.625116699627317e-05, "loss": 0.0407, "step": 22380 }, { "grad_norm": 0.3067748248577118, "learning_rate": 8.623692908905041e-05, "loss": 0.0335, "step": 22390 }, { "grad_norm": 0.4134661555290222, "learning_rate": 8.622268498992755e-05, "loss": 0.0422, "step": 22400 }, { "grad_norm": 0.23823022842407227, "learning_rate": 8.620843470133851e-05, "loss": 0.0566, "step": 22410 }, { "grad_norm": 0.4276304841041565, "learning_rate": 8.619417822571829e-05, "loss": 0.045, "step": 22420 }, { "grad_norm": 0.3850962221622467, "learning_rate": 8.617991556550292e-05, "loss": 0.0365, "step": 22430 }, { "grad_norm": 0.37281689047813416, "learning_rate": 8.616564672312952e-05, "loss": 0.0392, "step": 22440 }, { "grad_norm": 0.5011849999427795, "learning_rate": 8.61513717010362e-05, "loss": 0.0429, "step": 22450 }, { "grad_norm": 0.3807845413684845, "learning_rate": 8.613709050166221e-05, "loss": 0.0418, "step": 22460 }, { "grad_norm": 0.41479942202568054, "learning_rate": 8.61228031274478e-05, "loss": 0.0384, "step": 22470 }, { "grad_norm": 0.48810163140296936, "learning_rate": 8.610850958083431e-05, "loss": 0.0402, "step": 22480 }, { "grad_norm": 0.46404823660850525, "learning_rate": 8.609420986426409e-05, "loss": 0.033, "step": 22490 }, { "grad_norm": 0.30605193972587585, "learning_rate": 8.60799039801806e-05, "loss": 0.0383, "step": 22500 }, { "grad_norm": 0.3244999647140503, "learning_rate": 8.606559193102828e-05, "loss": 0.0312, "step": 22510 }, { "grad_norm": 0.31936630606651306, "learning_rate": 8.605127371925273e-05, "loss": 0.0397, "step": 22520 }, { "grad_norm": 0.3821476995944977, "learning_rate": 8.603694934730047e-05, "loss": 0.0434, "step": 22530 }, { "grad_norm": 0.36246493458747864, "learning_rate": 8.602261881761919e-05, "loss": 0.0317, "step": 22540 }, { "grad_norm": 0.21070095896720886, "learning_rate": 8.600828213265759e-05, "loss": 0.045, "step": 22550 }, { "grad_norm": 0.4448404312133789, "learning_rate": 8.599393929486539e-05, "loss": 0.0464, "step": 22560 }, { "grad_norm": 0.31255823373794556, "learning_rate": 8.59795903066934e-05, "loss": 0.0386, "step": 22570 }, { "grad_norm": 0.3509812355041504, "learning_rate": 8.596523517059347e-05, "loss": 0.0327, "step": 22580 }, { "grad_norm": 0.5167020559310913, "learning_rate": 8.59508738890185e-05, "loss": 0.0338, "step": 22590 }, { "grad_norm": 0.38254138827323914, "learning_rate": 8.593650646442246e-05, "loss": 0.0355, "step": 22600 }, { "grad_norm": 0.3734796941280365, "learning_rate": 8.59221328992603e-05, "loss": 0.0386, "step": 22610 }, { "grad_norm": 0.23933658003807068, "learning_rate": 8.590775319598813e-05, "loss": 0.0341, "step": 22620 }, { "grad_norm": 0.26682794094085693, "learning_rate": 8.589336735706301e-05, "loss": 0.0364, "step": 22630 }, { "grad_norm": 0.40212202072143555, "learning_rate": 8.587897538494307e-05, "loss": 0.0493, "step": 22640 }, { "grad_norm": 0.5005204081535339, "learning_rate": 8.586457728208756e-05, "loss": 0.0401, "step": 22650 }, { "grad_norm": 0.5092453360557556, "learning_rate": 8.585017305095667e-05, "loss": 0.0401, "step": 22660 }, { "grad_norm": 0.35865893959999084, "learning_rate": 8.583576269401173e-05, "loss": 0.0358, "step": 22670 }, { "grad_norm": 0.3802188038825989, "learning_rate": 8.582134621371504e-05, "loss": 0.0386, "step": 22680 }, { "grad_norm": 0.39219585061073303, "learning_rate": 8.580692361253e-05, "loss": 0.0361, "step": 22690 }, { "grad_norm": 0.4148102104663849, "learning_rate": 8.579249489292104e-05, "loss": 0.0318, "step": 22700 }, { "grad_norm": 0.4279390871524811, "learning_rate": 8.577806005735363e-05, "loss": 0.0423, "step": 22710 }, { "grad_norm": 0.4426058232784271, "learning_rate": 8.576361910829429e-05, "loss": 0.0344, "step": 22720 }, { "grad_norm": 0.3587065637111664, "learning_rate": 8.574917204821057e-05, "loss": 0.0377, "step": 22730 }, { "grad_norm": 0.8575193285942078, "learning_rate": 8.57347188795711e-05, "loss": 0.0405, "step": 22740 }, { "grad_norm": 0.42764079570770264, "learning_rate": 8.572025960484551e-05, "loss": 0.0495, "step": 22750 }, { "grad_norm": 0.3938884735107422, "learning_rate": 8.57057942265045e-05, "loss": 0.0512, "step": 22760 }, { "grad_norm": 0.498761385679245, "learning_rate": 8.569132274701984e-05, "loss": 0.0392, "step": 22770 }, { "grad_norm": 0.4283505380153656, "learning_rate": 8.567684516886427e-05, "loss": 0.0382, "step": 22780 }, { "grad_norm": 0.40680810809135437, "learning_rate": 8.56623614945116e-05, "loss": 0.0375, "step": 22790 }, { "grad_norm": 0.2916584610939026, "learning_rate": 8.564787172643675e-05, "loss": 0.027, "step": 22800 }, { "grad_norm": 0.3912285566329956, "learning_rate": 8.563337586711559e-05, "loss": 0.0426, "step": 22810 }, { "grad_norm": 0.29218655824661255, "learning_rate": 8.561887391902506e-05, "loss": 0.0464, "step": 22820 }, { "grad_norm": 0.3488316237926483, "learning_rate": 8.560436588464316e-05, "loss": 0.0333, "step": 22830 }, { "grad_norm": 0.3841594457626343, "learning_rate": 8.55898517664489e-05, "loss": 0.0416, "step": 22840 }, { "grad_norm": 0.4847552478313446, "learning_rate": 8.557533156692236e-05, "loss": 0.0576, "step": 22850 }, { "grad_norm": 0.5436500310897827, "learning_rate": 8.556080528854467e-05, "loss": 0.0366, "step": 22860 }, { "grad_norm": 0.3897629976272583, "learning_rate": 8.554627293379791e-05, "loss": 0.0427, "step": 22870 }, { "grad_norm": 0.4483153223991394, "learning_rate": 8.553173450516531e-05, "loss": 0.049, "step": 22880 }, { "grad_norm": 0.39499321579933167, "learning_rate": 8.551719000513108e-05, "loss": 0.0366, "step": 22890 }, { "grad_norm": 0.43053191900253296, "learning_rate": 8.550263943618049e-05, "loss": 0.0562, "step": 22900 }, { "grad_norm": 0.4140032231807709, "learning_rate": 8.54880828007998e-05, "loss": 0.0454, "step": 22910 }, { "grad_norm": 0.36741915345191956, "learning_rate": 8.547352010147637e-05, "loss": 0.0452, "step": 22920 }, { "grad_norm": 0.43070998787879944, "learning_rate": 8.545895134069855e-05, "loss": 0.0419, "step": 22930 }, { "grad_norm": 0.33491790294647217, "learning_rate": 8.544437652095576e-05, "loss": 0.0422, "step": 22940 }, { "grad_norm": 0.36579862236976624, "learning_rate": 8.542979564473843e-05, "loss": 0.0488, "step": 22950 }, { "grad_norm": 0.37147650122642517, "learning_rate": 8.541520871453802e-05, "loss": 0.0358, "step": 22960 }, { "grad_norm": 0.4220205247402191, "learning_rate": 8.540061573284705e-05, "loss": 0.0399, "step": 22970 }, { "grad_norm": 0.28371089696884155, "learning_rate": 8.538601670215906e-05, "loss": 0.0432, "step": 22980 }, { "grad_norm": 0.41187942028045654, "learning_rate": 8.537141162496864e-05, "loss": 0.0434, "step": 22990 }, { "grad_norm": 0.3334291875362396, "learning_rate": 8.535680050377137e-05, "loss": 0.043, "step": 23000 }, { "grad_norm": 0.22704418003559113, "learning_rate": 8.534218334106391e-05, "loss": 0.0379, "step": 23010 }, { "grad_norm": 0.45907261967658997, "learning_rate": 8.532756013934393e-05, "loss": 0.0386, "step": 23020 }, { "grad_norm": 0.4918723404407501, "learning_rate": 8.531293090111012e-05, "loss": 0.0422, "step": 23030 }, { "grad_norm": 0.3397078812122345, "learning_rate": 8.529829562886225e-05, "loss": 0.0399, "step": 23040 }, { "grad_norm": 0.39604079723358154, "learning_rate": 8.528365432510105e-05, "loss": 0.0522, "step": 23050 }, { "grad_norm": 0.42443132400512695, "learning_rate": 8.526900699232833e-05, "loss": 0.0407, "step": 23060 }, { "grad_norm": 0.3962762951850891, "learning_rate": 8.525435363304695e-05, "loss": 0.0385, "step": 23070 }, { "grad_norm": 0.4457318186759949, "learning_rate": 8.523969424976072e-05, "loss": 0.041, "step": 23080 }, { "grad_norm": 0.3031093180179596, "learning_rate": 8.522502884497457e-05, "loss": 0.0357, "step": 23090 }, { "grad_norm": 0.4402967691421509, "learning_rate": 8.521035742119437e-05, "loss": 0.0408, "step": 23100 }, { "grad_norm": 0.34582045674324036, "learning_rate": 8.519567998092712e-05, "loss": 0.0464, "step": 23110 }, { "grad_norm": 0.33856064081192017, "learning_rate": 8.518099652668075e-05, "loss": 0.0411, "step": 23120 }, { "grad_norm": 0.35588476061820984, "learning_rate": 8.516630706096429e-05, "loss": 0.0384, "step": 23130 }, { "grad_norm": 0.40583810210227966, "learning_rate": 8.515161158628773e-05, "loss": 0.0328, "step": 23140 }, { "grad_norm": 0.4152519106864929, "learning_rate": 8.513691010516216e-05, "loss": 0.0592, "step": 23150 }, { "grad_norm": 0.2911645770072937, "learning_rate": 8.512220262009966e-05, "loss": 0.0324, "step": 23160 }, { "grad_norm": 0.32568591833114624, "learning_rate": 8.510748913361332e-05, "loss": 0.0396, "step": 23170 }, { "grad_norm": 0.41318342089653015, "learning_rate": 8.509276964821726e-05, "loss": 0.0389, "step": 23180 }, { "grad_norm": 0.44507354497909546, "learning_rate": 8.507804416642669e-05, "loss": 0.0427, "step": 23190 }, { "grad_norm": 0.48454269766807556, "learning_rate": 8.506331269075774e-05, "loss": 0.039, "step": 23200 }, { "grad_norm": 0.4072667062282562, "learning_rate": 8.504857522372765e-05, "loss": 0.0372, "step": 23210 }, { "grad_norm": 0.46987900137901306, "learning_rate": 8.503383176785461e-05, "loss": 0.034, "step": 23220 }, { "grad_norm": 0.4234916865825653, "learning_rate": 8.501908232565792e-05, "loss": 0.0415, "step": 23230 }, { "grad_norm": 0.41354188323020935, "learning_rate": 8.50043268996578e-05, "loss": 0.0389, "step": 23240 }, { "grad_norm": 0.35893651843070984, "learning_rate": 8.498956549237562e-05, "loss": 0.0417, "step": 23250 }, { "grad_norm": 0.27212977409362793, "learning_rate": 8.497479810633366e-05, "loss": 0.0336, "step": 23260 }, { "grad_norm": 0.4540194272994995, "learning_rate": 8.496002474405525e-05, "loss": 0.0383, "step": 23270 }, { "grad_norm": 0.30823206901550293, "learning_rate": 8.494524540806478e-05, "loss": 0.0381, "step": 23280 }, { "grad_norm": 0.37241417169570923, "learning_rate": 8.493046010088761e-05, "loss": 0.0477, "step": 23290 }, { "grad_norm": 0.4115442633628845, "learning_rate": 8.491566882505018e-05, "loss": 0.0369, "step": 23300 }, { "grad_norm": 0.5394027233123779, "learning_rate": 8.490087158307988e-05, "loss": 0.0432, "step": 23310 }, { "grad_norm": 0.6023816466331482, "learning_rate": 8.488606837750518e-05, "loss": 0.0435, "step": 23320 }, { "grad_norm": 0.42321014404296875, "learning_rate": 8.487125921085552e-05, "loss": 0.0407, "step": 23330 }, { "grad_norm": 0.28299933671951294, "learning_rate": 8.485644408566141e-05, "loss": 0.0386, "step": 23340 }, { "grad_norm": 0.4084049463272095, "learning_rate": 8.484162300445431e-05, "loss": 0.0361, "step": 23350 }, { "grad_norm": 0.37535980343818665, "learning_rate": 8.482679596976676e-05, "loss": 0.0356, "step": 23360 }, { "grad_norm": 0.4310929775238037, "learning_rate": 8.48119629841323e-05, "loss": 0.0414, "step": 23370 }, { "grad_norm": 0.40994179248809814, "learning_rate": 8.479712405008547e-05, "loss": 0.046, "step": 23380 }, { "grad_norm": 0.4271068274974823, "learning_rate": 8.478227917016184e-05, "loss": 0.0513, "step": 23390 }, { "grad_norm": 0.2996543347835541, "learning_rate": 8.476742834689801e-05, "loss": 0.0488, "step": 23400 }, { "grad_norm": 0.4483526647090912, "learning_rate": 8.475257158283157e-05, "loss": 0.0456, "step": 23410 }, { "grad_norm": 0.45919039845466614, "learning_rate": 8.473770888050112e-05, "loss": 0.0352, "step": 23420 }, { "grad_norm": 0.47748345136642456, "learning_rate": 8.47228402424463e-05, "loss": 0.0392, "step": 23430 }, { "grad_norm": 0.4802006781101227, "learning_rate": 8.470796567120775e-05, "loss": 0.0381, "step": 23440 }, { "grad_norm": 0.4022848904132843, "learning_rate": 8.469308516932714e-05, "loss": 0.0384, "step": 23450 }, { "grad_norm": 0.42933160066604614, "learning_rate": 8.467819873934714e-05, "loss": 0.0376, "step": 23460 }, { "grad_norm": 0.424484521150589, "learning_rate": 8.466330638381143e-05, "loss": 0.0417, "step": 23470 }, { "grad_norm": 0.4195660948753357, "learning_rate": 8.464840810526469e-05, "loss": 0.0357, "step": 23480 }, { "grad_norm": 0.4760362207889557, "learning_rate": 8.463350390625264e-05, "loss": 0.0435, "step": 23490 }, { "grad_norm": 0.3515620231628418, "learning_rate": 8.4618593789322e-05, "loss": 0.0501, "step": 23500 }, { "grad_norm": 0.38782167434692383, "learning_rate": 8.46036777570205e-05, "loss": 0.0422, "step": 23510 }, { "grad_norm": 0.4131300747394562, "learning_rate": 8.458875581189688e-05, "loss": 0.0567, "step": 23520 }, { "grad_norm": 0.3049772083759308, "learning_rate": 8.457382795650092e-05, "loss": 0.0388, "step": 23530 }, { "grad_norm": 0.34532737731933594, "learning_rate": 8.455889419338335e-05, "loss": 0.0418, "step": 23540 }, { "grad_norm": 0.4148641526699066, "learning_rate": 8.454395452509593e-05, "loss": 0.0365, "step": 23550 }, { "grad_norm": 0.34379449486732483, "learning_rate": 8.452900895419146e-05, "loss": 0.0418, "step": 23560 }, { "grad_norm": 0.2767189145088196, "learning_rate": 8.451405748322376e-05, "loss": 0.0309, "step": 23570 }, { "grad_norm": 0.3619690239429474, "learning_rate": 8.449910011474759e-05, "loss": 0.0354, "step": 23580 }, { "grad_norm": 0.3809093236923218, "learning_rate": 8.448413685131876e-05, "loss": 0.044, "step": 23590 }, { "grad_norm": 0.5122804641723633, "learning_rate": 8.446916769549407e-05, "loss": 0.0428, "step": 23600 }, { "grad_norm": 0.30292370915412903, "learning_rate": 8.445419264983136e-05, "loss": 0.0411, "step": 23610 }, { "grad_norm": 0.2744395434856415, "learning_rate": 8.443921171688947e-05, "loss": 0.0343, "step": 23620 }, { "grad_norm": 0.25474071502685547, "learning_rate": 8.442422489922819e-05, "loss": 0.0325, "step": 23630 }, { "grad_norm": 0.3177626132965088, "learning_rate": 8.440923219940838e-05, "loss": 0.0438, "step": 23640 }, { "grad_norm": 0.36669886112213135, "learning_rate": 8.439423361999189e-05, "loss": 0.0372, "step": 23650 }, { "grad_norm": 0.41075390577316284, "learning_rate": 8.437922916354155e-05, "loss": 0.0377, "step": 23660 }, { "grad_norm": 0.3035646677017212, "learning_rate": 8.436421883262123e-05, "loss": 0.0389, "step": 23670 }, { "grad_norm": 0.3558953106403351, "learning_rate": 8.434920262979577e-05, "loss": 0.0504, "step": 23680 }, { "grad_norm": 0.3870665729045868, "learning_rate": 8.433418055763104e-05, "loss": 0.0398, "step": 23690 }, { "grad_norm": 0.3611515760421753, "learning_rate": 8.431915261869389e-05, "loss": 0.0406, "step": 23700 }, { "grad_norm": 0.2739030420780182, "learning_rate": 8.43041188155522e-05, "loss": 0.0347, "step": 23710 }, { "grad_norm": 0.3271558880805969, "learning_rate": 8.428907915077481e-05, "loss": 0.0378, "step": 23720 }, { "grad_norm": 0.3377957046031952, "learning_rate": 8.42740336269316e-05, "loss": 0.0407, "step": 23730 }, { "grad_norm": 0.27283692359924316, "learning_rate": 8.425898224659345e-05, "loss": 0.0303, "step": 23740 }, { "grad_norm": 0.5483874082565308, "learning_rate": 8.42439250123322e-05, "loss": 0.0327, "step": 23750 }, { "grad_norm": 0.32592543959617615, "learning_rate": 8.422886192672076e-05, "loss": 0.035, "step": 23760 }, { "grad_norm": 0.38730180263519287, "learning_rate": 8.421379299233297e-05, "loss": 0.0306, "step": 23770 }, { "grad_norm": 0.5284054279327393, "learning_rate": 8.419871821174371e-05, "loss": 0.0363, "step": 23780 }, { "grad_norm": 0.3201657831668854, "learning_rate": 8.418363758752884e-05, "loss": 0.0346, "step": 23790 }, { "grad_norm": 0.30794161558151245, "learning_rate": 8.416855112226523e-05, "loss": 0.0403, "step": 23800 }, { "grad_norm": 0.3457799553871155, "learning_rate": 8.415345881853075e-05, "loss": 0.0443, "step": 23810 }, { "grad_norm": 0.42790302634239197, "learning_rate": 8.413836067890426e-05, "loss": 0.042, "step": 23820 }, { "grad_norm": 0.5086534023284912, "learning_rate": 8.41232567059656e-05, "loss": 0.0339, "step": 23830 }, { "grad_norm": 0.3118613362312317, "learning_rate": 8.410814690229565e-05, "loss": 0.0385, "step": 23840 }, { "grad_norm": 0.2413056641817093, "learning_rate": 8.409303127047626e-05, "loss": 0.0391, "step": 23850 }, { "grad_norm": 0.3206638693809509, "learning_rate": 8.407790981309028e-05, "loss": 0.0359, "step": 23860 }, { "grad_norm": 0.404121071100235, "learning_rate": 8.406278253272153e-05, "loss": 0.0439, "step": 23870 }, { "grad_norm": 0.4268483519554138, "learning_rate": 8.404764943195487e-05, "loss": 0.0398, "step": 23880 }, { "grad_norm": 0.3956283628940582, "learning_rate": 8.403251051337613e-05, "loss": 0.0417, "step": 23890 }, { "grad_norm": 0.4352128207683563, "learning_rate": 8.401736577957214e-05, "loss": 0.0374, "step": 23900 }, { "grad_norm": 0.41245144605636597, "learning_rate": 8.40022152331307e-05, "loss": 0.0356, "step": 23910 }, { "grad_norm": 0.35506582260131836, "learning_rate": 8.398705887664064e-05, "loss": 0.0462, "step": 23920 }, { "grad_norm": 0.26331689953804016, "learning_rate": 8.397189671269177e-05, "loss": 0.0362, "step": 23930 }, { "grad_norm": 0.48084762692451477, "learning_rate": 8.395672874387488e-05, "loss": 0.0374, "step": 23940 }, { "grad_norm": 0.35344016551971436, "learning_rate": 8.394155497278177e-05, "loss": 0.0307, "step": 23950 }, { "grad_norm": 0.5332676768302917, "learning_rate": 8.392637540200523e-05, "loss": 0.0522, "step": 23960 }, { "grad_norm": 0.336294561624527, "learning_rate": 8.391119003413902e-05, "loss": 0.0371, "step": 23970 }, { "grad_norm": 0.3540698289871216, "learning_rate": 8.38959988717779e-05, "loss": 0.0366, "step": 23980 }, { "grad_norm": 0.24453212320804596, "learning_rate": 8.388080191751764e-05, "loss": 0.0425, "step": 23990 }, { "grad_norm": 0.4369630217552185, "learning_rate": 8.386559917395496e-05, "loss": 0.0408, "step": 24000 }, { "grad_norm": 0.2878793776035309, "learning_rate": 8.385039064368761e-05, "loss": 0.0351, "step": 24010 }, { "grad_norm": 0.38767677545547485, "learning_rate": 8.383517632931431e-05, "loss": 0.0355, "step": 24020 }, { "grad_norm": 0.3352391719818115, "learning_rate": 8.381995623343477e-05, "loss": 0.0393, "step": 24030 }, { "grad_norm": 0.43177375197410583, "learning_rate": 8.380473035864968e-05, "loss": 0.0387, "step": 24040 }, { "grad_norm": 0.34035804867744446, "learning_rate": 8.378949870756076e-05, "loss": 0.0374, "step": 24050 }, { "grad_norm": 0.33925414085388184, "learning_rate": 8.377426128277063e-05, "loss": 0.0399, "step": 24060 }, { "grad_norm": 0.3824838697910309, "learning_rate": 8.375901808688298e-05, "loss": 0.0366, "step": 24070 }, { "grad_norm": 0.32667380571365356, "learning_rate": 8.374376912250246e-05, "loss": 0.0466, "step": 24080 }, { "grad_norm": 0.3394438624382019, "learning_rate": 8.372851439223468e-05, "loss": 0.0357, "step": 24090 }, { "grad_norm": 0.3306507170200348, "learning_rate": 8.371325389868627e-05, "loss": 0.0396, "step": 24100 }, { "grad_norm": 0.2547563314437866, "learning_rate": 8.369798764446482e-05, "loss": 0.034, "step": 24110 }, { "grad_norm": 0.4680992364883423, "learning_rate": 8.368271563217893e-05, "loss": 0.0357, "step": 24120 }, { "grad_norm": 0.39404335618019104, "learning_rate": 8.366743786443817e-05, "loss": 0.0451, "step": 24130 }, { "grad_norm": 0.3825374245643616, "learning_rate": 8.365215434385309e-05, "loss": 0.0414, "step": 24140 }, { "grad_norm": 0.4209728538990021, "learning_rate": 8.36368650730352e-05, "loss": 0.0311, "step": 24150 }, { "grad_norm": 0.1764087975025177, "learning_rate": 8.362157005459705e-05, "loss": 0.0399, "step": 24160 }, { "grad_norm": 0.40005213022232056, "learning_rate": 8.360626929115213e-05, "loss": 0.0355, "step": 24170 }, { "grad_norm": 0.5462839603424072, "learning_rate": 8.359096278531492e-05, "loss": 0.0439, "step": 24180 }, { "grad_norm": 0.5288591384887695, "learning_rate": 8.357565053970088e-05, "loss": 0.0384, "step": 24190 }, { "grad_norm": 0.42704474925994873, "learning_rate": 8.356033255692647e-05, "loss": 0.039, "step": 24200 }, { "grad_norm": 0.3554462790489197, "learning_rate": 8.354500883960911e-05, "loss": 0.0345, "step": 24210 }, { "grad_norm": 0.3273358643054962, "learning_rate": 8.352967939036717e-05, "loss": 0.0395, "step": 24220 }, { "grad_norm": 0.45380279421806335, "learning_rate": 8.35143442118201e-05, "loss": 0.046, "step": 24230 }, { "grad_norm": 0.286253422498703, "learning_rate": 8.349900330658819e-05, "loss": 0.0363, "step": 24240 }, { "grad_norm": 0.35668477416038513, "learning_rate": 8.348365667729284e-05, "loss": 0.0395, "step": 24250 }, { "grad_norm": 0.3696689009666443, "learning_rate": 8.346830432655633e-05, "loss": 0.0406, "step": 24260 }, { "grad_norm": 0.4457920491695404, "learning_rate": 8.345294625700195e-05, "loss": 0.0354, "step": 24270 }, { "grad_norm": 0.5073724389076233, "learning_rate": 8.343758247125402e-05, "loss": 0.0385, "step": 24280 }, { "grad_norm": 0.36965033411979675, "learning_rate": 8.342221297193776e-05, "loss": 0.0461, "step": 24290 }, { "grad_norm": 0.38563424348831177, "learning_rate": 8.34068377616794e-05, "loss": 0.0489, "step": 24300 }, { "grad_norm": 0.4709669351577759, "learning_rate": 8.339145684310615e-05, "loss": 0.0411, "step": 24310 }, { "grad_norm": 0.5124526619911194, "learning_rate": 8.337607021884618e-05, "loss": 0.0405, "step": 24320 }, { "grad_norm": 0.6664032340049744, "learning_rate": 8.336067789152867e-05, "loss": 0.0337, "step": 24330 }, { "grad_norm": 0.27801313996315, "learning_rate": 8.334527986378369e-05, "loss": 0.0383, "step": 24340 }, { "grad_norm": 0.3391468822956085, "learning_rate": 8.332987613824239e-05, "loss": 0.0438, "step": 24350 }, { "grad_norm": 0.3527829349040985, "learning_rate": 8.331446671753685e-05, "loss": 0.0404, "step": 24360 }, { "grad_norm": 0.3530299663543701, "learning_rate": 8.329905160430007e-05, "loss": 0.0353, "step": 24370 }, { "grad_norm": 0.31352078914642334, "learning_rate": 8.328363080116611e-05, "loss": 0.0304, "step": 24380 }, { "grad_norm": 0.36333486437797546, "learning_rate": 8.326820431076997e-05, "loss": 0.0354, "step": 24390 }, { "grad_norm": 0.3797644078731537, "learning_rate": 8.325277213574759e-05, "loss": 0.041, "step": 24400 }, { "grad_norm": 0.3361836075782776, "learning_rate": 8.32373342787359e-05, "loss": 0.0345, "step": 24410 }, { "grad_norm": 0.35425370931625366, "learning_rate": 8.322189074237285e-05, "loss": 0.04, "step": 24420 }, { "grad_norm": 0.4441620409488678, "learning_rate": 8.32064415292973e-05, "loss": 0.0405, "step": 24430 }, { "grad_norm": 0.2873696982860565, "learning_rate": 8.319098664214907e-05, "loss": 0.0285, "step": 24440 }, { "grad_norm": 0.326540470123291, "learning_rate": 8.3175526083569e-05, "loss": 0.036, "step": 24450 }, { "grad_norm": 0.38394445180892944, "learning_rate": 8.316005985619889e-05, "loss": 0.0371, "step": 24460 }, { "grad_norm": 0.3449840843677521, "learning_rate": 8.314458796268147e-05, "loss": 0.0394, "step": 24470 }, { "grad_norm": 0.2449468970298767, "learning_rate": 8.312911040566047e-05, "loss": 0.034, "step": 24480 }, { "grad_norm": 0.41786935925483704, "learning_rate": 8.31136271877806e-05, "loss": 0.0344, "step": 24490 }, { "grad_norm": 0.43175387382507324, "learning_rate": 8.309813831168748e-05, "loss": 0.0412, "step": 24500 }, { "grad_norm": 0.35912227630615234, "learning_rate": 8.308264378002777e-05, "loss": 0.0344, "step": 24510 }, { "grad_norm": 0.2551943361759186, "learning_rate": 8.306714359544906e-05, "loss": 0.048, "step": 24520 }, { "grad_norm": 0.406084269285202, "learning_rate": 8.30516377605999e-05, "loss": 0.0356, "step": 24530 }, { "grad_norm": 0.309723824262619, "learning_rate": 8.30361262781298e-05, "loss": 0.0367, "step": 24540 }, { "grad_norm": 0.4117676317691803, "learning_rate": 8.302060915068924e-05, "loss": 0.0431, "step": 24550 }, { "grad_norm": 0.2334555983543396, "learning_rate": 8.300508638092972e-05, "loss": 0.032, "step": 24560 }, { "grad_norm": 0.28578856587409973, "learning_rate": 8.298955797150361e-05, "loss": 0.0379, "step": 24570 }, { "grad_norm": 0.5012202262878418, "learning_rate": 8.297402392506433e-05, "loss": 0.0356, "step": 24580 }, { "grad_norm": 0.3160671591758728, "learning_rate": 8.295848424426617e-05, "loss": 0.031, "step": 24590 }, { "grad_norm": 0.29379576444625854, "learning_rate": 8.29429389317645e-05, "loss": 0.0326, "step": 24600 }, { "grad_norm": 0.4294414222240448, "learning_rate": 8.292738799021556e-05, "loss": 0.0413, "step": 24610 }, { "grad_norm": 0.21364232897758484, "learning_rate": 8.291183142227656e-05, "loss": 0.0346, "step": 24620 }, { "grad_norm": 0.22700537741184235, "learning_rate": 8.289626923060572e-05, "loss": 0.0484, "step": 24630 }, { "grad_norm": 0.39942535758018494, "learning_rate": 8.288070141786218e-05, "loss": 0.042, "step": 24640 }, { "grad_norm": 0.32788562774658203, "learning_rate": 8.286512798670605e-05, "loss": 0.0361, "step": 24650 }, { "grad_norm": 0.3216712474822998, "learning_rate": 8.284954893979842e-05, "loss": 0.0377, "step": 24660 }, { "grad_norm": 0.3630431592464447, "learning_rate": 8.283396427980131e-05, "loss": 0.0419, "step": 24670 }, { "grad_norm": 0.35481172800064087, "learning_rate": 8.281837400937771e-05, "loss": 0.0379, "step": 24680 }, { "grad_norm": 0.33102402091026306, "learning_rate": 8.28027781311916e-05, "loss": 0.0365, "step": 24690 }, { "grad_norm": 0.4007476568222046, "learning_rate": 8.278717664790785e-05, "loss": 0.0392, "step": 24700 }, { "grad_norm": 0.4112757742404938, "learning_rate": 8.277156956219234e-05, "loss": 0.0371, "step": 24710 }, { "grad_norm": 0.44775301218032837, "learning_rate": 8.275595687671189e-05, "loss": 0.0371, "step": 24720 }, { "grad_norm": 0.3086736798286438, "learning_rate": 8.27403385941343e-05, "loss": 0.0357, "step": 24730 }, { "grad_norm": 0.3905102014541626, "learning_rate": 8.272471471712828e-05, "loss": 0.0387, "step": 24740 }, { "grad_norm": 0.39998337626457214, "learning_rate": 8.270908524836355e-05, "loss": 0.0409, "step": 24750 }, { "grad_norm": 0.3242371082305908, "learning_rate": 8.269345019051074e-05, "loss": 0.0439, "step": 24760 }, { "grad_norm": 0.3844413459300995, "learning_rate": 8.267780954624147e-05, "loss": 0.0376, "step": 24770 }, { "grad_norm": 0.3872860372066498, "learning_rate": 8.266216331822827e-05, "loss": 0.0456, "step": 24780 }, { "grad_norm": 0.3160785138607025, "learning_rate": 8.264651150914469e-05, "loss": 0.0337, "step": 24790 }, { "grad_norm": 0.2832431197166443, "learning_rate": 8.263085412166517e-05, "loss": 0.0294, "step": 24800 }, { "grad_norm": 0.38999292254447937, "learning_rate": 8.261519115846514e-05, "loss": 0.036, "step": 24810 }, { "grad_norm": 0.40533342957496643, "learning_rate": 8.259952262222096e-05, "loss": 0.0324, "step": 24820 }, { "grad_norm": 0.29180586338043213, "learning_rate": 8.258384851560997e-05, "loss": 0.0365, "step": 24830 }, { "grad_norm": 0.4456496238708496, "learning_rate": 8.256816884131044e-05, "loss": 0.0423, "step": 24840 }, { "grad_norm": 0.2761116921901703, "learning_rate": 8.255248360200159e-05, "loss": 0.0345, "step": 24850 }, { "grad_norm": 0.3139166831970215, "learning_rate": 8.253679280036359e-05, "loss": 0.032, "step": 24860 }, { "grad_norm": 0.7535373568534851, "learning_rate": 8.252109643907762e-05, "loss": 0.0472, "step": 24870 }, { "grad_norm": 0.3448292315006256, "learning_rate": 8.250539452082569e-05, "loss": 0.0328, "step": 24880 }, { "grad_norm": 0.404787540435791, "learning_rate": 8.248968704829087e-05, "loss": 0.0358, "step": 24890 }, { "grad_norm": 0.4616161286830902, "learning_rate": 8.247397402415714e-05, "loss": 0.0421, "step": 24900 }, { "grad_norm": 0.5396891832351685, "learning_rate": 8.24582554511094e-05, "loss": 0.0323, "step": 24910 }, { "grad_norm": 0.3733033835887909, "learning_rate": 8.244253133183355e-05, "loss": 0.0336, "step": 24920 }, { "grad_norm": 0.33110669255256653, "learning_rate": 8.24268016690164e-05, "loss": 0.0381, "step": 24930 }, { "grad_norm": 0.40224409103393555, "learning_rate": 8.241106646534571e-05, "loss": 0.0443, "step": 24940 }, { "grad_norm": 0.4289349913597107, "learning_rate": 8.23953257235102e-05, "loss": 0.0387, "step": 24950 }, { "grad_norm": 0.3414551019668579, "learning_rate": 8.237957944619956e-05, "loss": 0.0345, "step": 24960 }, { "grad_norm": 0.35124412178993225, "learning_rate": 8.236382763610437e-05, "loss": 0.0349, "step": 24970 }, { "grad_norm": 0.3861720860004425, "learning_rate": 8.234807029591619e-05, "loss": 0.0379, "step": 24980 }, { "grad_norm": 0.3230857253074646, "learning_rate": 8.233230742832752e-05, "loss": 0.0501, "step": 24990 }, { "grad_norm": 0.44061756134033203, "learning_rate": 8.231653903603178e-05, "loss": 0.038, "step": 25000 }, { "grad_norm": 0.3359307646751404, "learning_rate": 8.23007651217234e-05, "loss": 0.047, "step": 25010 }, { "grad_norm": 0.38233229517936707, "learning_rate": 8.228498568809769e-05, "loss": 0.0376, "step": 25020 }, { "grad_norm": 0.45167237520217896, "learning_rate": 8.22692007378509e-05, "loss": 0.043, "step": 25030 }, { "grad_norm": 0.3277878165245056, "learning_rate": 8.225341027368028e-05, "loss": 0.0341, "step": 25040 }, { "grad_norm": 0.3347816467285156, "learning_rate": 8.223761429828399e-05, "loss": 0.0384, "step": 25050 }, { "grad_norm": 0.3868264853954315, "learning_rate": 8.22218128143611e-05, "loss": 0.0342, "step": 25060 }, { "grad_norm": 0.3303772211074829, "learning_rate": 8.220600582461166e-05, "loss": 0.0373, "step": 25070 }, { "grad_norm": 0.39521005749702454, "learning_rate": 8.219019333173668e-05, "loss": 0.0361, "step": 25080 }, { "grad_norm": 0.3503231108188629, "learning_rate": 8.217437533843805e-05, "loss": 0.0347, "step": 25090 }, { "grad_norm": 0.41100743412971497, "learning_rate": 8.215855184741867e-05, "loss": 0.0356, "step": 25100 }, { "grad_norm": 0.35115280747413635, "learning_rate": 8.21427228613823e-05, "loss": 0.0356, "step": 25110 }, { "grad_norm": 0.43417951464653015, "learning_rate": 8.21268883830337e-05, "loss": 0.0386, "step": 25120 }, { "grad_norm": 0.3175574541091919, "learning_rate": 8.211104841507855e-05, "loss": 0.0478, "step": 25130 }, { "grad_norm": 0.32249751687049866, "learning_rate": 8.209520296022346e-05, "loss": 0.0404, "step": 25140 }, { "grad_norm": 0.36947301030158997, "learning_rate": 8.207935202117599e-05, "loss": 0.0361, "step": 25150 }, { "grad_norm": 0.3328196406364441, "learning_rate": 8.206349560064463e-05, "loss": 0.0308, "step": 25160 }, { "grad_norm": 0.5253269076347351, "learning_rate": 8.204763370133881e-05, "loss": 0.039, "step": 25170 }, { "grad_norm": 0.38155224919319153, "learning_rate": 8.203176632596892e-05, "loss": 0.0315, "step": 25180 }, { "grad_norm": 0.3366939127445221, "learning_rate": 8.20158934772462e-05, "loss": 0.0373, "step": 25190 }, { "grad_norm": 0.2700023651123047, "learning_rate": 8.200001515788294e-05, "loss": 0.0284, "step": 25200 }, { "grad_norm": 0.3761123716831207, "learning_rate": 8.198413137059228e-05, "loss": 0.0355, "step": 25210 }, { "grad_norm": 0.4062768220901489, "learning_rate": 8.196824211808835e-05, "loss": 0.0314, "step": 25220 }, { "grad_norm": 0.3375073969364166, "learning_rate": 8.195234740308617e-05, "loss": 0.0426, "step": 25230 }, { "grad_norm": 0.3641826808452606, "learning_rate": 8.193644722830171e-05, "loss": 0.0294, "step": 25240 }, { "grad_norm": 0.2904910445213318, "learning_rate": 8.19205415964519e-05, "loss": 0.0318, "step": 25250 }, { "grad_norm": 0.32483744621276855, "learning_rate": 8.190463051025456e-05, "loss": 0.0327, "step": 25260 }, { "grad_norm": 0.3073263466358185, "learning_rate": 8.188871397242843e-05, "loss": 0.035, "step": 25270 }, { "grad_norm": 0.2996107041835785, "learning_rate": 8.187279198569326e-05, "loss": 0.0353, "step": 25280 }, { "grad_norm": 0.2913076877593994, "learning_rate": 8.185686455276966e-05, "loss": 0.0331, "step": 25290 }, { "grad_norm": 0.360664039850235, "learning_rate": 8.184093167637921e-05, "loss": 0.0368, "step": 25300 }, { "grad_norm": 0.32249799370765686, "learning_rate": 8.182499335924437e-05, "loss": 0.0361, "step": 25310 }, { "grad_norm": 0.40871861577033997, "learning_rate": 8.18090496040886e-05, "loss": 0.0332, "step": 25320 }, { "grad_norm": 0.47021591663360596, "learning_rate": 8.179310041363621e-05, "loss": 0.0528, "step": 25330 }, { "grad_norm": 0.2854859232902527, "learning_rate": 8.17771457906125e-05, "loss": 0.0352, "step": 25340 }, { "grad_norm": 0.43502217531204224, "learning_rate": 8.176118573774371e-05, "loss": 0.0333, "step": 25350 }, { "grad_norm": 0.4522343575954437, "learning_rate": 8.174522025775692e-05, "loss": 0.0301, "step": 25360 }, { "grad_norm": 0.43395885825157166, "learning_rate": 8.172924935338022e-05, "loss": 0.037, "step": 25370 }, { "grad_norm": 0.2441299855709076, "learning_rate": 8.171327302734262e-05, "loss": 0.0306, "step": 25380 }, { "grad_norm": 0.26225823163986206, "learning_rate": 8.169729128237401e-05, "loss": 0.0311, "step": 25390 }, { "grad_norm": 0.38448673486709595, "learning_rate": 8.168130412120525e-05, "loss": 0.0353, "step": 25400 }, { "grad_norm": 0.4891740083694458, "learning_rate": 8.16653115465681e-05, "loss": 0.0359, "step": 25410 }, { "grad_norm": 0.5516822338104248, "learning_rate": 8.164931356119526e-05, "loss": 0.0327, "step": 25420 }, { "grad_norm": 0.4772128760814667, "learning_rate": 8.163331016782032e-05, "loss": 0.0454, "step": 25430 }, { "grad_norm": 0.30457308888435364, "learning_rate": 8.161730136917785e-05, "loss": 0.0341, "step": 25440 }, { "grad_norm": 0.2929803133010864, "learning_rate": 8.160128716800333e-05, "loss": 0.036, "step": 25450 }, { "grad_norm": 0.46573197841644287, "learning_rate": 8.158526756703313e-05, "loss": 0.0332, "step": 25460 }, { "grad_norm": 0.4263439178466797, "learning_rate": 8.156924256900455e-05, "loss": 0.0403, "step": 25470 }, { "grad_norm": 0.3237394094467163, "learning_rate": 8.155321217665584e-05, "loss": 0.0358, "step": 25480 }, { "grad_norm": 0.29006215929985046, "learning_rate": 8.153717639272614e-05, "loss": 0.0307, "step": 25490 }, { "grad_norm": 0.4373793601989746, "learning_rate": 8.152113521995555e-05, "loss": 0.038, "step": 25500 }, { "grad_norm": 0.4048101007938385, "learning_rate": 8.150508866108505e-05, "loss": 0.0337, "step": 25510 }, { "grad_norm": 0.41570544242858887, "learning_rate": 8.148903671885657e-05, "loss": 0.0346, "step": 25520 }, { "grad_norm": 0.3663279116153717, "learning_rate": 8.147297939601292e-05, "loss": 0.0437, "step": 25530 }, { "grad_norm": 0.39687755703926086, "learning_rate": 8.145691669529792e-05, "loss": 0.034, "step": 25540 }, { "grad_norm": 0.40612027049064636, "learning_rate": 8.144084861945618e-05, "loss": 0.0377, "step": 25550 }, { "grad_norm": 0.3716074228286743, "learning_rate": 8.142477517123333e-05, "loss": 0.0399, "step": 25560 }, { "grad_norm": 0.3596218228340149, "learning_rate": 8.140869635337586e-05, "loss": 0.0398, "step": 25570 }, { "grad_norm": 0.2900316119194031, "learning_rate": 8.139261216863123e-05, "loss": 0.0358, "step": 25580 }, { "grad_norm": 0.28936344385147095, "learning_rate": 8.137652261974776e-05, "loss": 0.034, "step": 25590 }, { "grad_norm": 0.3946276903152466, "learning_rate": 8.136042770947472e-05, "loss": 0.0372, "step": 25600 }, { "grad_norm": 0.25171881914138794, "learning_rate": 8.134432744056228e-05, "loss": 0.036, "step": 25610 }, { "grad_norm": 0.4506589472293854, "learning_rate": 8.132822181576158e-05, "loss": 0.041, "step": 25620 }, { "grad_norm": 0.33447587490081787, "learning_rate": 8.131211083782459e-05, "loss": 0.0402, "step": 25630 }, { "grad_norm": 0.3888431489467621, "learning_rate": 8.129599450950424e-05, "loss": 0.0384, "step": 25640 }, { "grad_norm": 0.4518430233001709, "learning_rate": 8.127987283355438e-05, "loss": 0.0438, "step": 25650 }, { "grad_norm": 0.318880170583725, "learning_rate": 8.126374581272976e-05, "loss": 0.0341, "step": 25660 }, { "grad_norm": 0.34955281019210815, "learning_rate": 8.124761344978605e-05, "loss": 0.0537, "step": 25670 }, { "grad_norm": 0.3909468352794647, "learning_rate": 8.12314757474798e-05, "loss": 0.0376, "step": 25680 }, { "grad_norm": 0.7420759201049805, "learning_rate": 8.121533270856856e-05, "loss": 0.0401, "step": 25690 }, { "grad_norm": 0.40249404311180115, "learning_rate": 8.119918433581069e-05, "loss": 0.0359, "step": 25700 }, { "grad_norm": 0.45905154943466187, "learning_rate": 8.118303063196551e-05, "loss": 0.0383, "step": 25710 }, { "grad_norm": 0.3683827519416809, "learning_rate": 8.116687159979326e-05, "loss": 0.0339, "step": 25720 }, { "grad_norm": 0.5205880999565125, "learning_rate": 8.115070724205508e-05, "loss": 0.0453, "step": 25730 }, { "grad_norm": 0.49238601326942444, "learning_rate": 8.113453756151296e-05, "loss": 0.042, "step": 25740 }, { "grad_norm": 0.37877926230430603, "learning_rate": 8.111836256092995e-05, "loss": 0.0396, "step": 25750 }, { "grad_norm": 0.4203492999076843, "learning_rate": 8.110218224306985e-05, "loss": 0.0384, "step": 25760 }, { "grad_norm": 0.3819907307624817, "learning_rate": 8.108599661069745e-05, "loss": 0.0367, "step": 25770 }, { "grad_norm": 0.3560817837715149, "learning_rate": 8.106980566657845e-05, "loss": 0.0302, "step": 25780 }, { "grad_norm": 0.30645355582237244, "learning_rate": 8.10536094134794e-05, "loss": 0.0396, "step": 25790 }, { "grad_norm": 0.3767828941345215, "learning_rate": 8.103740785416783e-05, "loss": 0.0306, "step": 25800 }, { "grad_norm": 0.3980296850204468, "learning_rate": 8.102120099141212e-05, "loss": 0.0405, "step": 25810 }, { "grad_norm": 0.2875564694404602, "learning_rate": 8.100498882798163e-05, "loss": 0.031, "step": 25820 }, { "grad_norm": 0.37688159942626953, "learning_rate": 8.09887713666465e-05, "loss": 0.0351, "step": 25830 }, { "grad_norm": 0.4503801465034485, "learning_rate": 8.09725486101779e-05, "loss": 0.047, "step": 25840 }, { "grad_norm": 0.3758561611175537, "learning_rate": 8.095632056134784e-05, "loss": 0.0362, "step": 25850 }, { "grad_norm": 0.40596824884414673, "learning_rate": 8.094008722292925e-05, "loss": 0.038, "step": 25860 }, { "grad_norm": 0.3892761468887329, "learning_rate": 8.092384859769598e-05, "loss": 0.0416, "step": 25870 }, { "grad_norm": 0.37360602617263794, "learning_rate": 8.090760468842275e-05, "loss": 0.048, "step": 25880 }, { "grad_norm": 0.30138635635375977, "learning_rate": 8.089135549788521e-05, "loss": 0.0399, "step": 25890 }, { "grad_norm": 0.4746328294277191, "learning_rate": 8.087510102885987e-05, "loss": 0.0375, "step": 25900 }, { "grad_norm": 0.3291516900062561, "learning_rate": 8.085884128412422e-05, "loss": 0.0404, "step": 25910 }, { "grad_norm": 0.4964464008808136, "learning_rate": 8.084257626645659e-05, "loss": 0.042, "step": 25920 }, { "grad_norm": 0.2583368718624115, "learning_rate": 8.08263059786362e-05, "loss": 0.0336, "step": 25930 }, { "grad_norm": 0.35212764143943787, "learning_rate": 8.081003042344325e-05, "loss": 0.0323, "step": 25940 }, { "grad_norm": 0.3423161506652832, "learning_rate": 8.079374960365872e-05, "loss": 0.0397, "step": 25950 }, { "grad_norm": 0.28621289134025574, "learning_rate": 8.077746352206463e-05, "loss": 0.0448, "step": 25960 }, { "grad_norm": 0.3876400291919708, "learning_rate": 8.076117218144377e-05, "loss": 0.0385, "step": 25970 }, { "grad_norm": 0.42656201124191284, "learning_rate": 8.074487558457991e-05, "loss": 0.0336, "step": 25980 }, { "grad_norm": 0.26247721910476685, "learning_rate": 8.072857373425768e-05, "loss": 0.0347, "step": 25990 }, { "grad_norm": 0.38890403509140015, "learning_rate": 8.071226663326264e-05, "loss": 0.0355, "step": 26000 }, { "grad_norm": 0.5182070136070251, "learning_rate": 8.069595428438121e-05, "loss": 0.0441, "step": 26010 }, { "grad_norm": 0.34127771854400635, "learning_rate": 8.067963669040072e-05, "loss": 0.0346, "step": 26020 }, { "grad_norm": 0.46000194549560547, "learning_rate": 8.066331385410942e-05, "loss": 0.0415, "step": 26030 }, { "grad_norm": 0.38254252076148987, "learning_rate": 8.064698577829641e-05, "loss": 0.0429, "step": 26040 }, { "grad_norm": 0.299446165561676, "learning_rate": 8.063065246575175e-05, "loss": 0.0405, "step": 26050 }, { "grad_norm": 0.32849544286727905, "learning_rate": 8.061431391926631e-05, "loss": 0.0313, "step": 26060 }, { "grad_norm": 0.3796185255050659, "learning_rate": 8.059797014163195e-05, "loss": 0.0385, "step": 26070 }, { "grad_norm": 0.5133435130119324, "learning_rate": 8.058162113564133e-05, "loss": 0.0438, "step": 26080 }, { "grad_norm": 0.4854804575443268, "learning_rate": 8.056526690408806e-05, "loss": 0.0436, "step": 26090 }, { "grad_norm": 0.35503727197647095, "learning_rate": 8.054890744976666e-05, "loss": 0.0408, "step": 26100 }, { "grad_norm": 0.3246578872203827, "learning_rate": 8.053254277547248e-05, "loss": 0.032, "step": 26110 }, { "grad_norm": 0.35740596055984497, "learning_rate": 8.051617288400182e-05, "loss": 0.0301, "step": 26120 }, { "grad_norm": 0.31831032037734985, "learning_rate": 8.049979777815182e-05, "loss": 0.0353, "step": 26130 }, { "grad_norm": 0.2991023361682892, "learning_rate": 8.048341746072054e-05, "loss": 0.0349, "step": 26140 }, { "grad_norm": 0.4571765661239624, "learning_rate": 8.046703193450696e-05, "loss": 0.0437, "step": 26150 }, { "grad_norm": 0.21096958220005035, "learning_rate": 8.04506412023109e-05, "loss": 0.0278, "step": 26160 }, { "grad_norm": 0.3082798421382904, "learning_rate": 8.043424526693306e-05, "loss": 0.0386, "step": 26170 }, { "grad_norm": 0.37864869832992554, "learning_rate": 8.04178441311751e-05, "loss": 0.0334, "step": 26180 }, { "grad_norm": 0.39674848318099976, "learning_rate": 8.04014377978395e-05, "loss": 0.038, "step": 26190 }, { "grad_norm": 0.34516820311546326, "learning_rate": 8.038502626972967e-05, "loss": 0.0372, "step": 26200 }, { "grad_norm": 0.6828068494796753, "learning_rate": 8.036860954964989e-05, "loss": 0.0411, "step": 26210 }, { "grad_norm": 0.44646766781806946, "learning_rate": 8.035218764040531e-05, "loss": 0.0343, "step": 26220 }, { "grad_norm": 0.5038284659385681, "learning_rate": 8.033576054480199e-05, "loss": 0.0409, "step": 26230 }, { "grad_norm": 0.3691393733024597, "learning_rate": 8.031932826564688e-05, "loss": 0.0299, "step": 26240 }, { "grad_norm": 0.3955191373825073, "learning_rate": 8.030289080574782e-05, "loss": 0.0375, "step": 26250 }, { "grad_norm": 0.4298481345176697, "learning_rate": 8.028644816791349e-05, "loss": 0.0435, "step": 26260 }, { "grad_norm": 0.33288517594337463, "learning_rate": 8.027000035495351e-05, "loss": 0.0326, "step": 26270 }, { "grad_norm": 0.29604625701904297, "learning_rate": 8.025354736967836e-05, "loss": 0.0518, "step": 26280 }, { "grad_norm": 0.34544017910957336, "learning_rate": 8.023708921489941e-05, "loss": 0.0449, "step": 26290 }, { "grad_norm": 0.36060136556625366, "learning_rate": 8.022062589342887e-05, "loss": 0.045, "step": 26300 }, { "grad_norm": 0.3582453429698944, "learning_rate": 8.020415740807993e-05, "loss": 0.0509, "step": 26310 }, { "grad_norm": 0.38886675238609314, "learning_rate": 8.018768376166656e-05, "loss": 0.037, "step": 26320 }, { "grad_norm": 0.28893569111824036, "learning_rate": 8.017120495700368e-05, "loss": 0.0352, "step": 26330 }, { "grad_norm": 0.5204185247421265, "learning_rate": 8.015472099690704e-05, "loss": 0.0389, "step": 26340 }, { "grad_norm": 0.3708176910877228, "learning_rate": 8.013823188419332e-05, "loss": 0.0459, "step": 26350 }, { "grad_norm": 0.40386122465133667, "learning_rate": 8.012173762168006e-05, "loss": 0.0435, "step": 26360 }, { "grad_norm": 0.32825154066085815, "learning_rate": 8.010523821218567e-05, "loss": 0.0295, "step": 26370 }, { "grad_norm": 0.2984057664871216, "learning_rate": 8.008873365852945e-05, "loss": 0.0359, "step": 26380 }, { "grad_norm": 0.3428570330142975, "learning_rate": 8.007222396353157e-05, "loss": 0.0377, "step": 26390 }, { "grad_norm": 0.4285345673561096, "learning_rate": 8.00557091300131e-05, "loss": 0.0355, "step": 26400 }, { "grad_norm": 0.27066510915756226, "learning_rate": 8.003918916079597e-05, "loss": 0.0339, "step": 26410 }, { "grad_norm": 0.2925100028514862, "learning_rate": 8.002266405870298e-05, "loss": 0.0305, "step": 26420 }, { "grad_norm": 0.32015976309776306, "learning_rate": 8.000613382655782e-05, "loss": 0.0395, "step": 26430 }, { "grad_norm": 0.3475983142852783, "learning_rate": 7.998959846718505e-05, "loss": 0.035, "step": 26440 }, { "grad_norm": 0.2828799784183502, "learning_rate": 7.997305798341012e-05, "loss": 0.0402, "step": 26450 }, { "grad_norm": 0.3076075315475464, "learning_rate": 7.995651237805937e-05, "loss": 0.0364, "step": 26460 }, { "grad_norm": 0.27346357703208923, "learning_rate": 7.993996165395996e-05, "loss": 0.0315, "step": 26470 }, { "grad_norm": 0.3401721715927124, "learning_rate": 7.992340581393996e-05, "loss": 0.0368, "step": 26480 }, { "grad_norm": 0.3702329993247986, "learning_rate": 7.990684486082831e-05, "loss": 0.039, "step": 26490 }, { "grad_norm": 0.38054656982421875, "learning_rate": 7.989027879745482e-05, "loss": 0.0428, "step": 26500 }, { "grad_norm": 0.4392836391925812, "learning_rate": 7.98737076266502e-05, "loss": 0.0577, "step": 26510 }, { "grad_norm": 0.3554844558238983, "learning_rate": 7.985713135124598e-05, "loss": 0.0364, "step": 26520 }, { "grad_norm": 0.3335460126399994, "learning_rate": 7.98405499740746e-05, "loss": 0.037, "step": 26530 }, { "grad_norm": 0.5356020331382751, "learning_rate": 7.98239634979694e-05, "loss": 0.0395, "step": 26540 }, { "grad_norm": 0.5868046283721924, "learning_rate": 7.98073719257645e-05, "loss": 0.0396, "step": 26550 }, { "grad_norm": 0.4528317451477051, "learning_rate": 7.979077526029499e-05, "loss": 0.0421, "step": 26560 }, { "grad_norm": 0.3055046796798706, "learning_rate": 7.977417350439675e-05, "loss": 0.0314, "step": 26570 }, { "grad_norm": 0.34636417031288147, "learning_rate": 7.97575666609066e-05, "loss": 0.0349, "step": 26580 }, { "grad_norm": 0.37135419249534607, "learning_rate": 7.974095473266216e-05, "loss": 0.0408, "step": 26590 }, { "grad_norm": 0.39142096042633057, "learning_rate": 7.972433772250198e-05, "loss": 0.036, "step": 26600 }, { "grad_norm": 0.3480176627635956, "learning_rate": 7.970771563326544e-05, "loss": 0.0321, "step": 26610 }, { "grad_norm": 0.5041651725769043, "learning_rate": 7.96910884677928e-05, "loss": 0.0455, "step": 26620 }, { "grad_norm": 0.34483039379119873, "learning_rate": 7.967445622892523e-05, "loss": 0.0414, "step": 26630 }, { "grad_norm": 0.39462459087371826, "learning_rate": 7.965781891950465e-05, "loss": 0.0383, "step": 26640 }, { "grad_norm": 0.42720064520835876, "learning_rate": 7.964117654237397e-05, "loss": 0.0333, "step": 26650 }, { "grad_norm": 0.6198845505714417, "learning_rate": 7.962452910037692e-05, "loss": 0.0436, "step": 26660 }, { "grad_norm": 0.3953778147697449, "learning_rate": 7.96078765963581e-05, "loss": 0.0402, "step": 26670 }, { "grad_norm": 0.44110167026519775, "learning_rate": 7.95912190331629e-05, "loss": 0.0413, "step": 26680 }, { "grad_norm": 0.3897399306297302, "learning_rate": 7.957455641363772e-05, "loss": 0.0376, "step": 26690 }, { "grad_norm": 0.42895370721817017, "learning_rate": 7.955788874062968e-05, "loss": 0.0475, "step": 26700 }, { "grad_norm": 0.4643186330795288, "learning_rate": 7.95412160169869e-05, "loss": 0.0458, "step": 26710 }, { "grad_norm": 0.3607482612133026, "learning_rate": 7.952453824555824e-05, "loss": 0.0335, "step": 26720 }, { "grad_norm": 0.46873506903648376, "learning_rate": 7.95078554291935e-05, "loss": 0.0353, "step": 26730 }, { "grad_norm": 0.43965718150138855, "learning_rate": 7.94911675707433e-05, "loss": 0.0406, "step": 26740 }, { "grad_norm": 0.3609529435634613, "learning_rate": 7.947447467305915e-05, "loss": 0.0427, "step": 26750 }, { "grad_norm": 0.46472692489624023, "learning_rate": 7.94577767389934e-05, "loss": 0.0365, "step": 26760 }, { "grad_norm": 0.3092304468154907, "learning_rate": 7.944107377139928e-05, "loss": 0.0328, "step": 26770 }, { "grad_norm": 0.2796841859817505, "learning_rate": 7.942436577313088e-05, "loss": 0.0421, "step": 26780 }, { "grad_norm": 0.3119579255580902, "learning_rate": 7.940765274704312e-05, "loss": 0.0339, "step": 26790 }, { "grad_norm": 0.3343827426433563, "learning_rate": 7.939093469599181e-05, "loss": 0.0367, "step": 26800 }, { "grad_norm": 0.32798364758491516, "learning_rate": 7.93742116228336e-05, "loss": 0.0314, "step": 26810 }, { "grad_norm": 0.42234787344932556, "learning_rate": 7.935748353042602e-05, "loss": 0.0369, "step": 26820 }, { "grad_norm": 0.5260567665100098, "learning_rate": 7.934075042162744e-05, "loss": 0.0409, "step": 26830 }, { "grad_norm": 0.39778295159339905, "learning_rate": 7.932401229929705e-05, "loss": 0.0484, "step": 26840 }, { "grad_norm": 0.5160802602767944, "learning_rate": 7.9307269166295e-05, "loss": 0.0408, "step": 26850 }, { "grad_norm": 0.24159912765026093, "learning_rate": 7.92905210254822e-05, "loss": 0.0334, "step": 26860 }, { "grad_norm": 0.46134883165359497, "learning_rate": 7.927376787972045e-05, "loss": 0.0345, "step": 26870 }, { "grad_norm": 0.4036721885204315, "learning_rate": 7.92570097318724e-05, "loss": 0.0375, "step": 26880 }, { "grad_norm": 0.18445070087909698, "learning_rate": 7.924024658480158e-05, "loss": 0.0335, "step": 26890 }, { "grad_norm": 0.329331636428833, "learning_rate": 7.922347844137233e-05, "loss": 0.0327, "step": 26900 }, { "grad_norm": 0.21598315238952637, "learning_rate": 7.92067053044499e-05, "loss": 0.0445, "step": 26910 }, { "grad_norm": 0.46565866470336914, "learning_rate": 7.918992717690031e-05, "loss": 0.0453, "step": 26920 }, { "grad_norm": 0.4159870147705078, "learning_rate": 7.917314406159053e-05, "loss": 0.0366, "step": 26930 }, { "grad_norm": 0.6237631440162659, "learning_rate": 7.915635596138832e-05, "loss": 0.0482, "step": 26940 }, { "grad_norm": 0.31939491629600525, "learning_rate": 7.913956287916228e-05, "loss": 0.0388, "step": 26950 }, { "grad_norm": 0.3397940397262573, "learning_rate": 7.912276481778193e-05, "loss": 0.0348, "step": 26960 }, { "grad_norm": 0.3334890902042389, "learning_rate": 7.910596178011759e-05, "loss": 0.0489, "step": 26970 }, { "grad_norm": 0.37364527583122253, "learning_rate": 7.908915376904043e-05, "loss": 0.0298, "step": 26980 }, { "grad_norm": 0.369153767824173, "learning_rate": 7.907234078742247e-05, "loss": 0.0323, "step": 26990 }, { "grad_norm": 0.4861246347427368, "learning_rate": 7.90555228381366e-05, "loss": 0.0342, "step": 27000 }, { "grad_norm": 0.3709961175918579, "learning_rate": 7.903869992405656e-05, "loss": 0.026, "step": 27010 }, { "grad_norm": 0.4358808994293213, "learning_rate": 7.902187204805691e-05, "loss": 0.0361, "step": 27020 }, { "grad_norm": 0.2538052797317505, "learning_rate": 7.900503921301308e-05, "loss": 0.0283, "step": 27030 }, { "grad_norm": 0.2760520875453949, "learning_rate": 7.898820142180133e-05, "loss": 0.0433, "step": 27040 }, { "grad_norm": 0.3614853322505951, "learning_rate": 7.897135867729879e-05, "loss": 0.0289, "step": 27050 }, { "grad_norm": 0.29958754777908325, "learning_rate": 7.89545109823834e-05, "loss": 0.0288, "step": 27060 }, { "grad_norm": 0.2412579208612442, "learning_rate": 7.8937658339934e-05, "loss": 0.0299, "step": 27070 }, { "grad_norm": 0.2944524884223938, "learning_rate": 7.892080075283026e-05, "loss": 0.0343, "step": 27080 }, { "grad_norm": 0.1997598260641098, "learning_rate": 7.890393822395263e-05, "loss": 0.0272, "step": 27090 }, { "grad_norm": 0.3850151598453522, "learning_rate": 7.88870707561825e-05, "loss": 0.0289, "step": 27100 }, { "grad_norm": 0.42234164476394653, "learning_rate": 7.887019835240203e-05, "loss": 0.0318, "step": 27110 }, { "grad_norm": 0.3960827589035034, "learning_rate": 7.885332101549427e-05, "loss": 0.0307, "step": 27120 }, { "grad_norm": 0.23325732350349426, "learning_rate": 7.883643874834308e-05, "loss": 0.0247, "step": 27130 }, { "grad_norm": 0.32328563928604126, "learning_rate": 7.881955155383321e-05, "loss": 0.0384, "step": 27140 }, { "grad_norm": 0.3102768063545227, "learning_rate": 7.880265943485017e-05, "loss": 0.0282, "step": 27150 }, { "grad_norm": 0.34575513005256653, "learning_rate": 7.878576239428038e-05, "loss": 0.0328, "step": 27160 }, { "grad_norm": 0.35904282331466675, "learning_rate": 7.87688604350111e-05, "loss": 0.039, "step": 27170 }, { "grad_norm": 0.5473823547363281, "learning_rate": 7.875195355993042e-05, "loss": 0.0386, "step": 27180 }, { "grad_norm": 0.27213194966316223, "learning_rate": 7.873504177192724e-05, "loss": 0.0379, "step": 27190 }, { "grad_norm": 0.4472668468952179, "learning_rate": 7.87181250738913e-05, "loss": 0.0479, "step": 27200 }, { "grad_norm": 0.3564784526824951, "learning_rate": 7.870120346871324e-05, "loss": 0.0361, "step": 27210 }, { "grad_norm": 0.4639066755771637, "learning_rate": 7.86842769592845e-05, "loss": 0.0348, "step": 27220 }, { "grad_norm": 0.3347332775592804, "learning_rate": 7.866734554849732e-05, "loss": 0.0323, "step": 27230 }, { "grad_norm": 0.31300589442253113, "learning_rate": 7.865040923924486e-05, "loss": 0.0323, "step": 27240 }, { "grad_norm": 0.4995079040527344, "learning_rate": 7.863346803442104e-05, "loss": 0.041, "step": 27250 }, { "grad_norm": 0.30001577734947205, "learning_rate": 7.861652193692067e-05, "loss": 0.0271, "step": 27260 }, { "grad_norm": 0.2252751588821411, "learning_rate": 7.859957094963937e-05, "loss": 0.0245, "step": 27270 }, { "grad_norm": 0.3922477662563324, "learning_rate": 7.858261507547357e-05, "loss": 0.032, "step": 27280 }, { "grad_norm": 0.2888537347316742, "learning_rate": 7.856565431732061e-05, "loss": 0.0283, "step": 27290 }, { "grad_norm": 0.34654146432876587, "learning_rate": 7.854868867807859e-05, "loss": 0.0281, "step": 27300 }, { "grad_norm": 0.39918768405914307, "learning_rate": 7.85317181606465e-05, "loss": 0.0356, "step": 27310 }, { "grad_norm": 0.5283766388893127, "learning_rate": 7.85147427679241e-05, "loss": 0.0318, "step": 27320 }, { "grad_norm": 0.2625671327114105, "learning_rate": 7.849776250281205e-05, "loss": 0.0241, "step": 27330 }, { "grad_norm": 0.29065313935279846, "learning_rate": 7.84807773682118e-05, "loss": 0.0368, "step": 27340 }, { "grad_norm": 0.24147652089595795, "learning_rate": 7.846378736702565e-05, "loss": 0.0384, "step": 27350 }, { "grad_norm": 0.25481006503105164, "learning_rate": 7.844679250215671e-05, "loss": 0.0328, "step": 27360 }, { "grad_norm": 0.3348134458065033, "learning_rate": 7.842979277650898e-05, "loss": 0.0292, "step": 27370 }, { "grad_norm": 0.3773537874221802, "learning_rate": 7.84127881929872e-05, "loss": 0.033, "step": 27380 }, { "grad_norm": 0.38449805974960327, "learning_rate": 7.839577875449704e-05, "loss": 0.0339, "step": 27390 }, { "grad_norm": 0.4374493956565857, "learning_rate": 7.837876446394489e-05, "loss": 0.0336, "step": 27400 }, { "grad_norm": 0.3020811080932617, "learning_rate": 7.836174532423805e-05, "loss": 0.0427, "step": 27410 }, { "grad_norm": 0.4208977222442627, "learning_rate": 7.834472133828466e-05, "loss": 0.0351, "step": 27420 }, { "grad_norm": 0.4899100065231323, "learning_rate": 7.832769250899359e-05, "loss": 0.0416, "step": 27430 }, { "grad_norm": 0.5405052304267883, "learning_rate": 7.831065883927464e-05, "loss": 0.0441, "step": 27440 }, { "grad_norm": 0.3263664245605469, "learning_rate": 7.829362033203841e-05, "loss": 0.0389, "step": 27450 }, { "grad_norm": 0.4233110547065735, "learning_rate": 7.827657699019628e-05, "loss": 0.0352, "step": 27460 }, { "grad_norm": 0.37127527594566345, "learning_rate": 7.825952881666052e-05, "loss": 0.05, "step": 27470 }, { "grad_norm": 0.46998587250709534, "learning_rate": 7.824247581434418e-05, "loss": 0.038, "step": 27480 }, { "grad_norm": 0.45544058084487915, "learning_rate": 7.822541798616116e-05, "loss": 0.0366, "step": 27490 }, { "grad_norm": 0.404319703578949, "learning_rate": 7.820835533502617e-05, "loss": 0.0495, "step": 27500 }, { "grad_norm": 0.3993402421474457, "learning_rate": 7.819128786385475e-05, "loss": 0.0441, "step": 27510 }, { "grad_norm": 0.364109605550766, "learning_rate": 7.817421557556329e-05, "loss": 0.0462, "step": 27520 }, { "grad_norm": 0.42181316018104553, "learning_rate": 7.815713847306893e-05, "loss": 0.0426, "step": 27530 }, { "grad_norm": 0.3194626271724701, "learning_rate": 7.81400565592897e-05, "loss": 0.0389, "step": 27540 }, { "grad_norm": 0.4364367127418518, "learning_rate": 7.812296983714444e-05, "loss": 0.0431, "step": 27550 }, { "grad_norm": 0.3321273922920227, "learning_rate": 7.810587830955281e-05, "loss": 0.034, "step": 27560 }, { "grad_norm": 0.4120270311832428, "learning_rate": 7.808878197943528e-05, "loss": 0.0424, "step": 27570 }, { "grad_norm": 0.4474543333053589, "learning_rate": 7.807168084971312e-05, "loss": 0.0321, "step": 27580 }, { "grad_norm": 0.30533382296562195, "learning_rate": 7.805457492330849e-05, "loss": 0.0309, "step": 27590 }, { "grad_norm": 0.35155850648880005, "learning_rate": 7.803746420314428e-05, "loss": 0.0348, "step": 27600 }, { "grad_norm": 0.49057674407958984, "learning_rate": 7.802034869214428e-05, "loss": 0.0436, "step": 27610 }, { "grad_norm": 0.37842628359794617, "learning_rate": 7.800322839323303e-05, "loss": 0.0405, "step": 27620 }, { "grad_norm": 0.3014879524707794, "learning_rate": 7.798610330933593e-05, "loss": 0.0315, "step": 27630 }, { "grad_norm": 0.3553333878517151, "learning_rate": 7.796897344337922e-05, "loss": 0.0332, "step": 27640 }, { "grad_norm": 0.30381855368614197, "learning_rate": 7.795183879828989e-05, "loss": 0.0414, "step": 27650 }, { "grad_norm": 0.2969249188899994, "learning_rate": 7.793469937699579e-05, "loss": 0.047, "step": 27660 }, { "grad_norm": 0.3252469599246979, "learning_rate": 7.791755518242558e-05, "loss": 0.0473, "step": 27670 }, { "grad_norm": 0.34360969066619873, "learning_rate": 7.790040621750876e-05, "loss": 0.0452, "step": 27680 }, { "grad_norm": 0.3362528085708618, "learning_rate": 7.788325248517558e-05, "loss": 0.0377, "step": 27690 }, { "grad_norm": 0.40340563654899597, "learning_rate": 7.786609398835715e-05, "loss": 0.0428, "step": 27700 }, { "grad_norm": 0.39036667346954346, "learning_rate": 7.784893072998541e-05, "loss": 0.0286, "step": 27710 }, { "grad_norm": 0.3052334785461426, "learning_rate": 7.783176271299306e-05, "loss": 0.0344, "step": 27720 }, { "grad_norm": 0.34968432784080505, "learning_rate": 7.781458994031368e-05, "loss": 0.0349, "step": 27730 }, { "grad_norm": 0.3870021402835846, "learning_rate": 7.779741241488161e-05, "loss": 0.0347, "step": 27740 }, { "grad_norm": 0.4005075693130493, "learning_rate": 7.7780230139632e-05, "loss": 0.0297, "step": 27750 }, { "grad_norm": 0.31876620650291443, "learning_rate": 7.776304311750087e-05, "loss": 0.0347, "step": 27760 }, { "grad_norm": 0.35238540172576904, "learning_rate": 7.7745851351425e-05, "loss": 0.0439, "step": 27770 }, { "grad_norm": 0.36001667380332947, "learning_rate": 7.772865484434197e-05, "loss": 0.0352, "step": 27780 }, { "grad_norm": 0.5033126473426819, "learning_rate": 7.77114535991902e-05, "loss": 0.0394, "step": 27790 }, { "grad_norm": 0.3363870084285736, "learning_rate": 7.769424761890893e-05, "loss": 0.0368, "step": 27800 }, { "grad_norm": 0.4962339401245117, "learning_rate": 7.767703690643817e-05, "loss": 0.0381, "step": 27810 }, { "grad_norm": 0.3566552698612213, "learning_rate": 7.76598214647188e-05, "loss": 0.0336, "step": 27820 }, { "grad_norm": 0.2858937680721283, "learning_rate": 7.764260129669241e-05, "loss": 0.0324, "step": 27830 }, { "grad_norm": 0.34467554092407227, "learning_rate": 7.76253764053015e-05, "loss": 0.0377, "step": 27840 }, { "grad_norm": 0.3275582790374756, "learning_rate": 7.760814679348932e-05, "loss": 0.0284, "step": 27850 }, { "grad_norm": 0.33883795142173767, "learning_rate": 7.759091246419992e-05, "loss": 0.0309, "step": 27860 }, { "grad_norm": 0.3502163290977478, "learning_rate": 7.75736734203782e-05, "loss": 0.0281, "step": 27870 }, { "grad_norm": 0.4938773214817047, "learning_rate": 7.755642966496985e-05, "loss": 0.0481, "step": 27880 }, { "grad_norm": 0.4082662761211395, "learning_rate": 7.753918120092132e-05, "loss": 0.0355, "step": 27890 }, { "grad_norm": 0.30317023396492004, "learning_rate": 7.752192803117993e-05, "loss": 0.0333, "step": 27900 }, { "grad_norm": 0.35569247603416443, "learning_rate": 7.750467015869377e-05, "loss": 0.0357, "step": 27910 }, { "grad_norm": 0.27570146322250366, "learning_rate": 7.748740758641174e-05, "loss": 0.0354, "step": 27920 }, { "grad_norm": 0.5266336798667908, "learning_rate": 7.74701403172835e-05, "loss": 0.0434, "step": 27930 }, { "grad_norm": 0.4414750039577484, "learning_rate": 7.745286835425962e-05, "loss": 0.0412, "step": 27940 }, { "grad_norm": 0.4926278591156006, "learning_rate": 7.743559170029138e-05, "loss": 0.0409, "step": 27950 }, { "grad_norm": 0.3318462073802948, "learning_rate": 7.741831035833087e-05, "loss": 0.045, "step": 27960 }, { "grad_norm": 0.36425113677978516, "learning_rate": 7.740102433133102e-05, "loss": 0.0287, "step": 27970 }, { "grad_norm": 0.3761444091796875, "learning_rate": 7.738373362224553e-05, "loss": 0.0294, "step": 27980 }, { "grad_norm": 0.3408065140247345, "learning_rate": 7.73664382340289e-05, "loss": 0.0286, "step": 27990 }, { "grad_norm": 0.3931618630886078, "learning_rate": 7.734913816963647e-05, "loss": 0.0426, "step": 28000 }, { "grad_norm": 0.33868929743766785, "learning_rate": 7.73318334320243e-05, "loss": 0.033, "step": 28010 }, { "grad_norm": 0.420854777097702, "learning_rate": 7.731452402414934e-05, "loss": 0.0416, "step": 28020 }, { "grad_norm": 0.40709730982780457, "learning_rate": 7.729720994896928e-05, "loss": 0.035, "step": 28030 }, { "grad_norm": 0.30630889534950256, "learning_rate": 7.727989120944262e-05, "loss": 0.0344, "step": 28040 }, { "grad_norm": 0.27232328057289124, "learning_rate": 7.726256780852865e-05, "loss": 0.0319, "step": 28050 }, { "grad_norm": 0.35328763723373413, "learning_rate": 7.724523974918749e-05, "loss": 0.0333, "step": 28060 }, { "grad_norm": 0.4321204125881195, "learning_rate": 7.722790703438002e-05, "loss": 0.0299, "step": 28070 }, { "grad_norm": 0.3730962574481964, "learning_rate": 7.72105696670679e-05, "loss": 0.031, "step": 28080 }, { "grad_norm": 0.3928508758544922, "learning_rate": 7.719322765021364e-05, "loss": 0.035, "step": 28090 }, { "grad_norm": 0.38700392842292786, "learning_rate": 7.717588098678051e-05, "loss": 0.0299, "step": 28100 }, { "grad_norm": 0.30108553171157837, "learning_rate": 7.715852967973258e-05, "loss": 0.0335, "step": 28110 }, { "grad_norm": 0.4458332061767578, "learning_rate": 7.714117373203474e-05, "loss": 0.033, "step": 28120 }, { "grad_norm": 0.3975520730018616, "learning_rate": 7.712381314665259e-05, "loss": 0.0327, "step": 28130 }, { "grad_norm": 0.37704870104789734, "learning_rate": 7.710644792655261e-05, "loss": 0.0325, "step": 28140 }, { "grad_norm": 0.5432448983192444, "learning_rate": 7.708907807470207e-05, "loss": 0.0455, "step": 28150 }, { "grad_norm": 0.3820757269859314, "learning_rate": 7.707170359406896e-05, "loss": 0.0357, "step": 28160 }, { "grad_norm": 0.3365892171859741, "learning_rate": 7.705432448762213e-05, "loss": 0.0324, "step": 28170 }, { "grad_norm": 0.48138195276260376, "learning_rate": 7.703694075833117e-05, "loss": 0.0383, "step": 28180 }, { "grad_norm": 0.4664008915424347, "learning_rate": 7.70195524091665e-05, "loss": 0.05, "step": 28190 }, { "grad_norm": 0.42994311451911926, "learning_rate": 7.70021594430993e-05, "loss": 0.0395, "step": 28200 }, { "grad_norm": 0.26750901341438293, "learning_rate": 7.698476186310157e-05, "loss": 0.0355, "step": 28210 }, { "grad_norm": 0.3192513585090637, "learning_rate": 7.696735967214608e-05, "loss": 0.0315, "step": 28220 }, { "grad_norm": 0.40777868032455444, "learning_rate": 7.694995287320636e-05, "loss": 0.036, "step": 28230 }, { "grad_norm": 0.33064591884613037, "learning_rate": 7.693254146925679e-05, "loss": 0.0332, "step": 28240 }, { "grad_norm": 0.41388240456581116, "learning_rate": 7.691512546327251e-05, "loss": 0.0448, "step": 28250 }, { "grad_norm": 0.3798489570617676, "learning_rate": 7.689770485822939e-05, "loss": 0.0383, "step": 28260 }, { "grad_norm": 0.3758605718612671, "learning_rate": 7.688027965710416e-05, "loss": 0.0253, "step": 28270 }, { "grad_norm": 0.33611002564430237, "learning_rate": 7.686284986287433e-05, "loss": 0.034, "step": 28280 }, { "grad_norm": 0.2564827799797058, "learning_rate": 7.684541547851817e-05, "loss": 0.0265, "step": 28290 }, { "grad_norm": 0.2643558979034424, "learning_rate": 7.68279765070147e-05, "loss": 0.0304, "step": 28300 }, { "grad_norm": 0.3952539563179016, "learning_rate": 7.68105329513438e-05, "loss": 0.0323, "step": 28310 }, { "grad_norm": 0.33427584171295166, "learning_rate": 7.67930848144861e-05, "loss": 0.033, "step": 28320 }, { "grad_norm": 0.3190000057220459, "learning_rate": 7.6775632099423e-05, "loss": 0.0378, "step": 28330 }, { "grad_norm": 0.29688894748687744, "learning_rate": 7.675817480913667e-05, "loss": 0.0331, "step": 28340 }, { "grad_norm": 0.33674854040145874, "learning_rate": 7.674071294661011e-05, "loss": 0.0336, "step": 28350 }, { "grad_norm": 0.39096227288246155, "learning_rate": 7.672324651482707e-05, "loss": 0.0293, "step": 28360 }, { "grad_norm": 0.3788200616836548, "learning_rate": 7.670577551677209e-05, "loss": 0.037, "step": 28370 }, { "grad_norm": 0.5654116868972778, "learning_rate": 7.668829995543047e-05, "loss": 0.0331, "step": 28380 }, { "grad_norm": 0.307299941778183, "learning_rate": 7.667081983378832e-05, "loss": 0.0425, "step": 28390 }, { "grad_norm": 0.33009660243988037, "learning_rate": 7.66533351548325e-05, "loss": 0.0329, "step": 28400 }, { "grad_norm": 0.3920241594314575, "learning_rate": 7.663584592155069e-05, "loss": 0.0364, "step": 28410 }, { "grad_norm": 0.44816985726356506, "learning_rate": 7.661835213693129e-05, "loss": 0.0421, "step": 28420 }, { "grad_norm": 0.3834410011768341, "learning_rate": 7.660085380396353e-05, "loss": 0.0371, "step": 28430 }, { "grad_norm": 0.21140503883361816, "learning_rate": 7.658335092563738e-05, "loss": 0.0336, "step": 28440 }, { "grad_norm": 0.3162246644496918, "learning_rate": 7.656584350494362e-05, "loss": 0.0326, "step": 28450 }, { "grad_norm": 0.31559526920318604, "learning_rate": 7.654833154487378e-05, "loss": 0.0332, "step": 28460 }, { "grad_norm": 0.38430753350257874, "learning_rate": 7.653081504842017e-05, "loss": 0.0344, "step": 28470 }, { "grad_norm": 0.39389318227767944, "learning_rate": 7.65132940185759e-05, "loss": 0.0422, "step": 28480 }, { "grad_norm": 0.2913570702075958, "learning_rate": 7.649576845833481e-05, "loss": 0.0292, "step": 28490 }, { "grad_norm": 0.27591291069984436, "learning_rate": 7.647823837069156e-05, "loss": 0.0322, "step": 28500 }, { "grad_norm": 0.2923194468021393, "learning_rate": 7.646070375864156e-05, "loss": 0.0342, "step": 28510 }, { "grad_norm": 0.3920915126800537, "learning_rate": 7.644316462518099e-05, "loss": 0.038, "step": 28520 }, { "grad_norm": 0.4243364632129669, "learning_rate": 7.642562097330679e-05, "loss": 0.0367, "step": 28530 }, { "grad_norm": 0.36690255999565125, "learning_rate": 7.640807280601671e-05, "loss": 0.0294, "step": 28540 }, { "grad_norm": 0.2953391969203949, "learning_rate": 7.639052012630927e-05, "loss": 0.0382, "step": 28550 }, { "grad_norm": 0.3355949819087982, "learning_rate": 7.63729629371837e-05, "loss": 0.0354, "step": 28560 }, { "grad_norm": 0.27649199962615967, "learning_rate": 7.635540124164009e-05, "loss": 0.0384, "step": 28570 }, { "grad_norm": 0.37006378173828125, "learning_rate": 7.633783504267922e-05, "loss": 0.0359, "step": 28580 }, { "grad_norm": 0.4283670485019684, "learning_rate": 7.632026434330269e-05, "loss": 0.0334, "step": 28590 }, { "grad_norm": 0.3557332158088684, "learning_rate": 7.630268914651282e-05, "loss": 0.0286, "step": 28600 }, { "grad_norm": 0.31453171372413635, "learning_rate": 7.628510945531278e-05, "loss": 0.05, "step": 28610 }, { "grad_norm": 0.32995739579200745, "learning_rate": 7.626752527270641e-05, "loss": 0.0385, "step": 28620 }, { "grad_norm": 0.30690160393714905, "learning_rate": 7.62499366016984e-05, "loss": 0.029, "step": 28630 }, { "grad_norm": 0.2550498843193054, "learning_rate": 7.623234344529416e-05, "loss": 0.0338, "step": 28640 }, { "grad_norm": 0.32686957716941833, "learning_rate": 7.62147458064999e-05, "loss": 0.0385, "step": 28650 }, { "grad_norm": 0.4058102071285248, "learning_rate": 7.619714368832254e-05, "loss": 0.0417, "step": 28660 }, { "grad_norm": 0.30510416626930237, "learning_rate": 7.61795370937698e-05, "loss": 0.0294, "step": 28670 }, { "grad_norm": 0.3339969217777252, "learning_rate": 7.61619260258502e-05, "loss": 0.0343, "step": 28680 }, { "grad_norm": 0.45802581310272217, "learning_rate": 7.614431048757298e-05, "loss": 0.0367, "step": 28690 }, { "grad_norm": 0.21961526572704315, "learning_rate": 7.612669048194814e-05, "loss": 0.0316, "step": 28700 }, { "grad_norm": 0.38173916935920715, "learning_rate": 7.610906601198646e-05, "loss": 0.0333, "step": 28710 }, { "grad_norm": 0.33181652426719666, "learning_rate": 7.60914370806995e-05, "loss": 0.0354, "step": 28720 }, { "grad_norm": 0.44616129994392395, "learning_rate": 7.607380369109953e-05, "loss": 0.0527, "step": 28730 }, { "grad_norm": 0.3499505817890167, "learning_rate": 7.605616584619961e-05, "loss": 0.0276, "step": 28740 }, { "grad_norm": 0.400657594203949, "learning_rate": 7.603852354901362e-05, "loss": 0.0349, "step": 28750 }, { "grad_norm": 0.25779181718826294, "learning_rate": 7.602087680255609e-05, "loss": 0.04, "step": 28760 }, { "grad_norm": 0.3141298294067383, "learning_rate": 7.600322560984238e-05, "loss": 0.028, "step": 28770 }, { "grad_norm": 0.3493082821369171, "learning_rate": 7.598556997388863e-05, "loss": 0.0312, "step": 28780 }, { "grad_norm": 0.31864693760871887, "learning_rate": 7.596790989771166e-05, "loss": 0.0346, "step": 28790 }, { "grad_norm": 0.331810861825943, "learning_rate": 7.595024538432914e-05, "loss": 0.0312, "step": 28800 }, { "grad_norm": 0.3089234530925751, "learning_rate": 7.59325764367594e-05, "loss": 0.0234, "step": 28810 }, { "grad_norm": 0.3891953229904175, "learning_rate": 7.59149030580216e-05, "loss": 0.0328, "step": 28820 }, { "grad_norm": 0.37453126907348633, "learning_rate": 7.589722525113562e-05, "loss": 0.0312, "step": 28830 }, { "grad_norm": 0.32016730308532715, "learning_rate": 7.587954301912216e-05, "loss": 0.0252, "step": 28840 }, { "grad_norm": 0.5131394267082214, "learning_rate": 7.586185636500263e-05, "loss": 0.0354, "step": 28850 }, { "grad_norm": 0.30896949768066406, "learning_rate": 7.584416529179914e-05, "loss": 0.0367, "step": 28860 }, { "grad_norm": 0.36465591192245483, "learning_rate": 7.582646980253465e-05, "loss": 0.0367, "step": 28870 }, { "grad_norm": 0.27505260705947876, "learning_rate": 7.580876990023282e-05, "loss": 0.0431, "step": 28880 }, { "grad_norm": 0.37258702516555786, "learning_rate": 7.579106558791809e-05, "loss": 0.0297, "step": 28890 }, { "grad_norm": 0.2438904494047165, "learning_rate": 7.577335686861565e-05, "loss": 0.0328, "step": 28900 }, { "grad_norm": 0.38579660654067993, "learning_rate": 7.575564374535141e-05, "loss": 0.0324, "step": 28910 }, { "grad_norm": 0.2918814420700073, "learning_rate": 7.573792622115207e-05, "loss": 0.0291, "step": 28920 }, { "grad_norm": 0.3705998659133911, "learning_rate": 7.572020429904507e-05, "loss": 0.0263, "step": 28930 }, { "grad_norm": 0.3904705047607422, "learning_rate": 7.570247798205861e-05, "loss": 0.0312, "step": 28940 }, { "grad_norm": 0.2505742013454437, "learning_rate": 7.568474727322164e-05, "loss": 0.0374, "step": 28950 }, { "grad_norm": 0.33579015731811523, "learning_rate": 7.566701217556384e-05, "loss": 0.0331, "step": 28960 }, { "grad_norm": 0.3150647282600403, "learning_rate": 7.564927269211564e-05, "loss": 0.0347, "step": 28970 }, { "grad_norm": 0.25225749611854553, "learning_rate": 7.563152882590824e-05, "loss": 0.0293, "step": 28980 }, { "grad_norm": 0.3730027973651886, "learning_rate": 7.56137805799736e-05, "loss": 0.0285, "step": 28990 }, { "grad_norm": 0.35498788952827454, "learning_rate": 7.559602795734439e-05, "loss": 0.0296, "step": 29000 }, { "grad_norm": 0.3170672357082367, "learning_rate": 7.557827096105408e-05, "loss": 0.0342, "step": 29010 }, { "grad_norm": 0.37514883279800415, "learning_rate": 7.55605095941368e-05, "loss": 0.0324, "step": 29020 }, { "grad_norm": 0.3750271499156952, "learning_rate": 7.554274385962752e-05, "loss": 0.0342, "step": 29030 }, { "grad_norm": 0.33672577142715454, "learning_rate": 7.552497376056191e-05, "loss": 0.0336, "step": 29040 }, { "grad_norm": 0.46722108125686646, "learning_rate": 7.550719929997639e-05, "loss": 0.0355, "step": 29050 }, { "grad_norm": 0.32697802782058716, "learning_rate": 7.548942048090813e-05, "loss": 0.0362, "step": 29060 }, { "grad_norm": 0.35832470655441284, "learning_rate": 7.547163730639506e-05, "loss": 0.0324, "step": 29070 }, { "grad_norm": 0.25295230746269226, "learning_rate": 7.545384977947583e-05, "loss": 0.0296, "step": 29080 }, { "grad_norm": 0.3542795777320862, "learning_rate": 7.543605790318981e-05, "loss": 0.0303, "step": 29090 }, { "grad_norm": 0.38729536533355713, "learning_rate": 7.54182616805772e-05, "loss": 0.0346, "step": 29100 }, { "grad_norm": 0.3727435767650604, "learning_rate": 7.540046111467885e-05, "loss": 0.0317, "step": 29110 }, { "grad_norm": 0.3449874520301819, "learning_rate": 7.53826562085364e-05, "loss": 0.0357, "step": 29120 }, { "grad_norm": 0.2984464168548584, "learning_rate": 7.536484696519221e-05, "loss": 0.0288, "step": 29130 }, { "grad_norm": 0.29881665110588074, "learning_rate": 7.534703338768942e-05, "loss": 0.0316, "step": 29140 }, { "grad_norm": 0.2272832840681076, "learning_rate": 7.532921547907185e-05, "loss": 0.0309, "step": 29150 }, { "grad_norm": 0.3407251834869385, "learning_rate": 7.531139324238412e-05, "loss": 0.0373, "step": 29160 }, { "grad_norm": 0.3549704849720001, "learning_rate": 7.529356668067157e-05, "loss": 0.045, "step": 29170 }, { "grad_norm": 0.3870837688446045, "learning_rate": 7.527573579698023e-05, "loss": 0.0362, "step": 29180 }, { "grad_norm": 0.28299564123153687, "learning_rate": 7.525790059435693e-05, "loss": 0.0375, "step": 29190 }, { "grad_norm": 0.3485655188560486, "learning_rate": 7.524006107584926e-05, "loss": 0.0332, "step": 29200 }, { "grad_norm": 0.3662683367729187, "learning_rate": 7.522221724450544e-05, "loss": 0.0337, "step": 29210 }, { "grad_norm": 0.3208119869232178, "learning_rate": 7.520436910337451e-05, "loss": 0.0466, "step": 29220 }, { "grad_norm": 0.2418733388185501, "learning_rate": 7.518651665550627e-05, "loss": 0.0327, "step": 29230 }, { "grad_norm": 0.37058112025260925, "learning_rate": 7.516865990395117e-05, "loss": 0.0479, "step": 29240 }, { "grad_norm": 0.3605184257030487, "learning_rate": 7.515079885176047e-05, "loss": 0.0256, "step": 29250 }, { "grad_norm": 0.2642790377140045, "learning_rate": 7.513293350198612e-05, "loss": 0.0384, "step": 29260 }, { "grad_norm": 0.34947824478149414, "learning_rate": 7.511506385768081e-05, "loss": 0.0336, "step": 29270 }, { "grad_norm": 0.2839927077293396, "learning_rate": 7.509718992189801e-05, "loss": 0.0397, "step": 29280 }, { "grad_norm": 0.44148555397987366, "learning_rate": 7.507931169769182e-05, "loss": 0.0349, "step": 29290 }, { "grad_norm": 0.38079339265823364, "learning_rate": 7.506142918811722e-05, "loss": 0.0363, "step": 29300 }, { "grad_norm": 0.579110860824585, "learning_rate": 7.504354239622978e-05, "loss": 0.046, "step": 29310 }, { "grad_norm": 0.3369601368904114, "learning_rate": 7.50256513250859e-05, "loss": 0.033, "step": 29320 }, { "grad_norm": 0.38278958201408386, "learning_rate": 7.500775597774265e-05, "loss": 0.0341, "step": 29330 }, { "grad_norm": 0.4332749843597412, "learning_rate": 7.498985635725788e-05, "loss": 0.0301, "step": 29340 }, { "grad_norm": 0.45364445447921753, "learning_rate": 7.497195246669012e-05, "loss": 0.0413, "step": 29350 }, { "grad_norm": 0.38476455211639404, "learning_rate": 7.495404430909868e-05, "loss": 0.0399, "step": 29360 }, { "grad_norm": 0.46290522813796997, "learning_rate": 7.493613188754356e-05, "loss": 0.0367, "step": 29370 }, { "grad_norm": 0.4317973256111145, "learning_rate": 7.49182152050855e-05, "loss": 0.0402, "step": 29380 }, { "grad_norm": 0.31891366839408875, "learning_rate": 7.490029426478598e-05, "loss": 0.0293, "step": 29390 }, { "grad_norm": 0.2852075397968292, "learning_rate": 7.488236906970719e-05, "loss": 0.0349, "step": 29400 }, { "grad_norm": 0.42748042941093445, "learning_rate": 7.486443962291207e-05, "loss": 0.0328, "step": 29410 }, { "grad_norm": 0.27846699953079224, "learning_rate": 7.484650592746424e-05, "loss": 0.0291, "step": 29420 }, { "grad_norm": 0.35918399691581726, "learning_rate": 7.482856798642811e-05, "loss": 0.0363, "step": 29430 }, { "grad_norm": 0.2991722524166107, "learning_rate": 7.481062580286878e-05, "loss": 0.0375, "step": 29440 }, { "grad_norm": 0.31425347924232483, "learning_rate": 7.479267937985208e-05, "loss": 0.0375, "step": 29450 }, { "grad_norm": 0.2798384428024292, "learning_rate": 7.477472872044456e-05, "loss": 0.0319, "step": 29460 }, { "grad_norm": 0.3820919692516327, "learning_rate": 7.475677382771347e-05, "loss": 0.0359, "step": 29470 }, { "grad_norm": 0.7827063798904419, "learning_rate": 7.473881470472683e-05, "loss": 0.0412, "step": 29480 }, { "grad_norm": 0.3940792381763458, "learning_rate": 7.472085135455337e-05, "loss": 0.0386, "step": 29490 }, { "grad_norm": 0.3569975793361664, "learning_rate": 7.470288378026256e-05, "loss": 0.0426, "step": 29500 }, { "grad_norm": 0.49167659878730774, "learning_rate": 7.468491198492451e-05, "loss": 0.0349, "step": 29510 }, { "grad_norm": 0.2630719244480133, "learning_rate": 7.466693597161013e-05, "loss": 0.034, "step": 29520 }, { "grad_norm": 0.3420684039592743, "learning_rate": 7.464895574339104e-05, "loss": 0.0384, "step": 29530 }, { "grad_norm": 0.22406761348247528, "learning_rate": 7.463097130333958e-05, "loss": 0.0472, "step": 29540 }, { "grad_norm": 0.31031322479248047, "learning_rate": 7.461298265452876e-05, "loss": 0.036, "step": 29550 }, { "grad_norm": 0.3280699849128723, "learning_rate": 7.459498980003239e-05, "loss": 0.0302, "step": 29560 }, { "grad_norm": 0.5173920392990112, "learning_rate": 7.457699274292493e-05, "loss": 0.0385, "step": 29570 }, { "grad_norm": 0.2614956200122833, "learning_rate": 7.455899148628159e-05, "loss": 0.0376, "step": 29580 }, { "grad_norm": 0.43635398149490356, "learning_rate": 7.45409860331783e-05, "loss": 0.032, "step": 29590 }, { "grad_norm": 0.32740217447280884, "learning_rate": 7.452297638669169e-05, "loss": 0.0312, "step": 29600 }, { "grad_norm": 0.3544849455356598, "learning_rate": 7.450496254989911e-05, "loss": 0.0281, "step": 29610 }, { "grad_norm": 0.3802172839641571, "learning_rate": 7.448694452587866e-05, "loss": 0.032, "step": 29620 }, { "grad_norm": 0.6424501538276672, "learning_rate": 7.44689223177091e-05, "loss": 0.0428, "step": 29630 }, { "grad_norm": 0.5133510828018188, "learning_rate": 7.445089592846994e-05, "loss": 0.0308, "step": 29640 }, { "grad_norm": 0.3184981048107147, "learning_rate": 7.443286536124141e-05, "loss": 0.0362, "step": 29650 }, { "grad_norm": 0.3121246099472046, "learning_rate": 7.441483061910443e-05, "loss": 0.0418, "step": 29660 }, { "grad_norm": 0.40527528524398804, "learning_rate": 7.439679170514064e-05, "loss": 0.0374, "step": 29670 }, { "grad_norm": 0.36677443981170654, "learning_rate": 7.43787486224324e-05, "loss": 0.0364, "step": 29680 }, { "grad_norm": 0.306583970785141, "learning_rate": 7.436070137406276e-05, "loss": 0.0363, "step": 29690 }, { "grad_norm": 0.2668543756008148, "learning_rate": 7.434264996311556e-05, "loss": 0.0254, "step": 29700 }, { "grad_norm": 0.34624990820884705, "learning_rate": 7.432459439267525e-05, "loss": 0.0334, "step": 29710 }, { "grad_norm": 0.42884373664855957, "learning_rate": 7.430653466582701e-05, "loss": 0.0361, "step": 29720 }, { "grad_norm": 0.372234970331192, "learning_rate": 7.42884707856568e-05, "loss": 0.0439, "step": 29730 }, { "grad_norm": 0.2820330262184143, "learning_rate": 7.427040275525122e-05, "loss": 0.0291, "step": 29740 }, { "grad_norm": 0.3524520695209503, "learning_rate": 7.42523305776976e-05, "loss": 0.0379, "step": 29750 }, { "grad_norm": 0.32432878017425537, "learning_rate": 7.4234254256084e-05, "loss": 0.0261, "step": 29760 }, { "grad_norm": 0.2550288736820221, "learning_rate": 7.421617379349915e-05, "loss": 0.0338, "step": 29770 }, { "grad_norm": 0.20530468225479126, "learning_rate": 7.41980891930325e-05, "loss": 0.0333, "step": 29780 }, { "grad_norm": 0.27922922372817993, "learning_rate": 7.418000045777425e-05, "loss": 0.0251, "step": 29790 }, { "grad_norm": 0.35306140780448914, "learning_rate": 7.416190759081523e-05, "loss": 0.0313, "step": 29800 }, { "grad_norm": 0.3079155683517456, "learning_rate": 7.414381059524704e-05, "loss": 0.0239, "step": 29810 }, { "grad_norm": 0.34018394351005554, "learning_rate": 7.412570947416195e-05, "loss": 0.0257, "step": 29820 }, { "grad_norm": 0.32261601090431213, "learning_rate": 7.410760423065295e-05, "loss": 0.0408, "step": 29830 }, { "grad_norm": 0.28448984026908875, "learning_rate": 7.408949486781372e-05, "loss": 0.0366, "step": 29840 }, { "grad_norm": 0.477862685918808, "learning_rate": 7.407138138873868e-05, "loss": 0.0396, "step": 29850 }, { "grad_norm": 0.5320006012916565, "learning_rate": 7.405326379652292e-05, "loss": 0.0359, "step": 29860 }, { "grad_norm": 0.316231906414032, "learning_rate": 7.403514209426222e-05, "loss": 0.0332, "step": 29870 }, { "grad_norm": 0.27635905146598816, "learning_rate": 7.40170162850531e-05, "loss": 0.033, "step": 29880 }, { "grad_norm": 0.24353782832622528, "learning_rate": 7.399888637199278e-05, "loss": 0.0293, "step": 29890 }, { "grad_norm": 0.6122894883155823, "learning_rate": 7.398075235817914e-05, "loss": 0.0446, "step": 29900 }, { "grad_norm": 0.3783051073551178, "learning_rate": 7.39626142467108e-05, "loss": 0.0336, "step": 29910 }, { "grad_norm": 0.40504464507102966, "learning_rate": 7.394447204068706e-05, "loss": 0.0321, "step": 29920 }, { "grad_norm": 0.5348434448242188, "learning_rate": 7.392632574320793e-05, "loss": 0.0454, "step": 29930 }, { "grad_norm": 0.38196229934692383, "learning_rate": 7.390817535737411e-05, "loss": 0.0327, "step": 29940 }, { "grad_norm": 0.49396103620529175, "learning_rate": 7.389002088628703e-05, "loss": 0.0347, "step": 29950 }, { "grad_norm": 0.3509999215602875, "learning_rate": 7.387186233304877e-05, "loss": 0.0336, "step": 29960 }, { "grad_norm": 0.2792578339576721, "learning_rate": 7.385369970076212e-05, "loss": 0.0336, "step": 29970 }, { "grad_norm": 0.34264639019966125, "learning_rate": 7.38355329925306e-05, "loss": 0.0397, "step": 29980 }, { "grad_norm": 0.3580147922039032, "learning_rate": 7.381736221145838e-05, "loss": 0.0356, "step": 29990 }, { "grad_norm": 0.31287825107574463, "learning_rate": 7.37991873606504e-05, "loss": 0.0308, "step": 30000 }, { "grad_norm": 0.29401326179504395, "learning_rate": 7.378100844321218e-05, "loss": 0.0307, "step": 30010 }, { "grad_norm": 0.2652488648891449, "learning_rate": 7.376282546225004e-05, "loss": 0.0275, "step": 30020 }, { "grad_norm": 0.4415093958377838, "learning_rate": 7.374463842087094e-05, "loss": 0.0302, "step": 30030 }, { "grad_norm": 0.3499053418636322, "learning_rate": 7.372644732218254e-05, "loss": 0.0305, "step": 30040 }, { "grad_norm": 0.390694260597229, "learning_rate": 7.370825216929322e-05, "loss": 0.0414, "step": 30050 }, { "grad_norm": 0.24039839208126068, "learning_rate": 7.369005296531205e-05, "loss": 0.0306, "step": 30060 }, { "grad_norm": 0.36107635498046875, "learning_rate": 7.367184971334873e-05, "loss": 0.0358, "step": 30070 }, { "grad_norm": 0.2778252065181732, "learning_rate": 7.365364241651371e-05, "loss": 0.0364, "step": 30080 }, { "grad_norm": 0.4024977385997772, "learning_rate": 7.363543107791815e-05, "loss": 0.0286, "step": 30090 }, { "grad_norm": 0.3762066960334778, "learning_rate": 7.361721570067384e-05, "loss": 0.039, "step": 30100 }, { "grad_norm": 0.4157502055168152, "learning_rate": 7.359899628789331e-05, "loss": 0.0454, "step": 30110 }, { "grad_norm": 0.34642767906188965, "learning_rate": 7.358077284268974e-05, "loss": 0.036, "step": 30120 }, { "grad_norm": 0.4243360459804535, "learning_rate": 7.356254536817702e-05, "loss": 0.036, "step": 30130 }, { "grad_norm": 0.2871382534503937, "learning_rate": 7.354431386746973e-05, "loss": 0.0334, "step": 30140 }, { "grad_norm": 0.41643768548965454, "learning_rate": 7.352607834368316e-05, "loss": 0.0315, "step": 30150 }, { "grad_norm": 0.3985518515110016, "learning_rate": 7.350783879993324e-05, "loss": 0.0304, "step": 30160 }, { "grad_norm": 0.4312460422515869, "learning_rate": 7.348959523933658e-05, "loss": 0.0377, "step": 30170 }, { "grad_norm": 0.37257853150367737, "learning_rate": 7.347134766501057e-05, "loss": 0.0294, "step": 30180 }, { "grad_norm": 0.31765928864479065, "learning_rate": 7.345309608007315e-05, "loss": 0.0296, "step": 30190 }, { "grad_norm": 0.3824405074119568, "learning_rate": 7.343484048764308e-05, "loss": 0.0318, "step": 30200 }, { "grad_norm": 0.4744684100151062, "learning_rate": 7.341658089083972e-05, "loss": 0.0346, "step": 30210 }, { "grad_norm": 0.21255740523338318, "learning_rate": 7.339831729278313e-05, "loss": 0.027, "step": 30220 }, { "grad_norm": 0.4007621109485626, "learning_rate": 7.338004969659404e-05, "loss": 0.0325, "step": 30230 }, { "grad_norm": 0.38080912828445435, "learning_rate": 7.336177810539391e-05, "loss": 0.0381, "step": 30240 }, { "grad_norm": 0.27487912774086, "learning_rate": 7.334350252230485e-05, "loss": 0.0285, "step": 30250 }, { "grad_norm": 0.3170788884162903, "learning_rate": 7.332522295044965e-05, "loss": 0.0285, "step": 30260 }, { "grad_norm": 0.3423978090286255, "learning_rate": 7.33069393929518e-05, "loss": 0.0349, "step": 30270 }, { "grad_norm": 0.35587552189826965, "learning_rate": 7.328865185293545e-05, "loss": 0.0426, "step": 30280 }, { "grad_norm": 0.4096185564994812, "learning_rate": 7.327036033352546e-05, "loss": 0.0346, "step": 30290 }, { "grad_norm": 0.3841792941093445, "learning_rate": 7.325206483784733e-05, "loss": 0.0361, "step": 30300 }, { "grad_norm": 0.37720420956611633, "learning_rate": 7.323376536902724e-05, "loss": 0.047, "step": 30310 }, { "grad_norm": 0.34969261288642883, "learning_rate": 7.321546193019213e-05, "loss": 0.0402, "step": 30320 }, { "grad_norm": 0.34905311465263367, "learning_rate": 7.31971545244695e-05, "loss": 0.035, "step": 30330 }, { "grad_norm": 0.31442713737487793, "learning_rate": 7.31788431549876e-05, "loss": 0.0343, "step": 30340 }, { "grad_norm": 0.3891836404800415, "learning_rate": 7.316052782487534e-05, "loss": 0.0314, "step": 30350 }, { "grad_norm": 0.3087092936038971, "learning_rate": 7.314220853726234e-05, "loss": 0.0389, "step": 30360 }, { "grad_norm": 0.34460318088531494, "learning_rate": 7.312388529527884e-05, "loss": 0.0364, "step": 30370 }, { "grad_norm": 0.3135760724544525, "learning_rate": 7.310555810205577e-05, "loss": 0.0354, "step": 30380 }, { "grad_norm": 0.42431870102882385, "learning_rate": 7.308722696072476e-05, "loss": 0.039, "step": 30390 }, { "grad_norm": 0.31189626455307007, "learning_rate": 7.306889187441811e-05, "loss": 0.0365, "step": 30400 }, { "grad_norm": 0.2478368580341339, "learning_rate": 7.305055284626876e-05, "loss": 0.0305, "step": 30410 }, { "grad_norm": 0.43222472071647644, "learning_rate": 7.303220987941037e-05, "loss": 0.0369, "step": 30420 }, { "grad_norm": 0.34886500239372253, "learning_rate": 7.301386297697726e-05, "loss": 0.0368, "step": 30430 }, { "grad_norm": 0.3777206838130951, "learning_rate": 7.299551214210438e-05, "loss": 0.0335, "step": 30440 }, { "grad_norm": 0.24788518249988556, "learning_rate": 7.297715737792738e-05, "loss": 0.0299, "step": 30450 }, { "grad_norm": 0.2961154580116272, "learning_rate": 7.295879868758265e-05, "loss": 0.032, "step": 30460 }, { "grad_norm": 0.3917672634124756, "learning_rate": 7.294043607420713e-05, "loss": 0.0331, "step": 30470 }, { "grad_norm": 0.29437389969825745, "learning_rate": 7.292206954093852e-05, "loss": 0.0299, "step": 30480 }, { "grad_norm": 0.48777127265930176, "learning_rate": 7.290369909091515e-05, "loss": 0.0312, "step": 30490 }, { "grad_norm": 0.4394455552101135, "learning_rate": 7.2885324727276e-05, "loss": 0.0442, "step": 30500 }, { "grad_norm": 0.28011471033096313, "learning_rate": 7.286694645316076e-05, "loss": 0.0293, "step": 30510 }, { "grad_norm": 0.3919658958911896, "learning_rate": 7.284856427170982e-05, "loss": 0.0333, "step": 30520 }, { "grad_norm": 0.3118146061897278, "learning_rate": 7.283017818606414e-05, "loss": 0.0389, "step": 30530 }, { "grad_norm": 0.36421695351600647, "learning_rate": 7.28117881993654e-05, "loss": 0.0388, "step": 30540 }, { "grad_norm": 0.2729530334472656, "learning_rate": 7.279339431475598e-05, "loss": 0.029, "step": 30550 }, { "grad_norm": 0.2649135887622833, "learning_rate": 7.277499653537887e-05, "loss": 0.0315, "step": 30560 }, { "grad_norm": 0.26595649123191833, "learning_rate": 7.275659486437776e-05, "loss": 0.0345, "step": 30570 }, { "grad_norm": 0.3687240481376648, "learning_rate": 7.273818930489695e-05, "loss": 0.0454, "step": 30580 }, { "grad_norm": 0.2884783446788788, "learning_rate": 7.271977986008151e-05, "loss": 0.0356, "step": 30590 }, { "grad_norm": 0.3243173062801361, "learning_rate": 7.270136653307705e-05, "loss": 0.0391, "step": 30600 }, { "grad_norm": 0.42477166652679443, "learning_rate": 7.268294932702994e-05, "loss": 0.0347, "step": 30610 }, { "grad_norm": 0.3003362715244293, "learning_rate": 7.266452824508719e-05, "loss": 0.0382, "step": 30620 }, { "grad_norm": 0.36809685826301575, "learning_rate": 7.264610329039643e-05, "loss": 0.0366, "step": 30630 }, { "grad_norm": 0.30296775698661804, "learning_rate": 7.262767446610599e-05, "loss": 0.0318, "step": 30640 }, { "grad_norm": 0.3357521593570709, "learning_rate": 7.260924177536485e-05, "loss": 0.038, "step": 30650 }, { "grad_norm": 0.4093165099620819, "learning_rate": 7.259080522132265e-05, "loss": 0.0383, "step": 30660 }, { "grad_norm": 0.45553988218307495, "learning_rate": 7.257236480712972e-05, "loss": 0.0358, "step": 30670 }, { "grad_norm": 0.4391424357891083, "learning_rate": 7.255392053593697e-05, "loss": 0.0358, "step": 30680 }, { "grad_norm": 0.33004388213157654, "learning_rate": 7.253547241089607e-05, "loss": 0.0346, "step": 30690 }, { "grad_norm": 0.43670305609703064, "learning_rate": 7.251702043515927e-05, "loss": 0.0385, "step": 30700 }, { "grad_norm": 0.36576953530311584, "learning_rate": 7.249856461187952e-05, "loss": 0.0295, "step": 30710 }, { "grad_norm": 0.5077559351921082, "learning_rate": 7.248010494421042e-05, "loss": 0.0342, "step": 30720 }, { "grad_norm": 0.2855348289012909, "learning_rate": 7.246164143530622e-05, "loss": 0.0428, "step": 30730 }, { "grad_norm": 0.5057206749916077, "learning_rate": 7.244317408832181e-05, "loss": 0.032, "step": 30740 }, { "grad_norm": 0.3424486219882965, "learning_rate": 7.242470290641279e-05, "loss": 0.0316, "step": 30750 }, { "grad_norm": 0.3859381675720215, "learning_rate": 7.240622789273536e-05, "loss": 0.0357, "step": 30760 }, { "grad_norm": 0.2878436744213104, "learning_rate": 7.238774905044638e-05, "loss": 0.0317, "step": 30770 }, { "grad_norm": 0.4304049015045166, "learning_rate": 7.236926638270341e-05, "loss": 0.0365, "step": 30780 }, { "grad_norm": 0.39890676736831665, "learning_rate": 7.23507798926646e-05, "loss": 0.0369, "step": 30790 }, { "grad_norm": 0.3225370943546295, "learning_rate": 7.23322895834888e-05, "loss": 0.0298, "step": 30800 }, { "grad_norm": 0.2422320693731308, "learning_rate": 7.231379545833552e-05, "loss": 0.0353, "step": 30810 }, { "grad_norm": 0.3055320084095001, "learning_rate": 7.229529752036487e-05, "loss": 0.0351, "step": 30820 }, { "grad_norm": 0.3394361734390259, "learning_rate": 7.227679577273765e-05, "loss": 0.0334, "step": 30830 }, { "grad_norm": 0.3178844749927521, "learning_rate": 7.225829021861529e-05, "loss": 0.0325, "step": 30840 }, { "grad_norm": 0.321135938167572, "learning_rate": 7.223978086115992e-05, "loss": 0.0303, "step": 30850 }, { "grad_norm": 0.35148540139198303, "learning_rate": 7.222126770353425e-05, "loss": 0.0363, "step": 30860 }, { "grad_norm": 0.3563328981399536, "learning_rate": 7.22027507489017e-05, "loss": 0.0321, "step": 30870 }, { "grad_norm": 0.3000914454460144, "learning_rate": 7.218423000042627e-05, "loss": 0.0334, "step": 30880 }, { "grad_norm": 0.3550443649291992, "learning_rate": 7.216570546127268e-05, "loss": 0.0359, "step": 30890 }, { "grad_norm": 0.38305673003196716, "learning_rate": 7.214717713460626e-05, "loss": 0.0384, "step": 30900 }, { "grad_norm": 0.3312513530254364, "learning_rate": 7.2128645023593e-05, "loss": 0.0381, "step": 30910 }, { "grad_norm": 0.47307151556015015, "learning_rate": 7.211010913139951e-05, "loss": 0.0387, "step": 30920 }, { "grad_norm": 0.38194698095321655, "learning_rate": 7.209156946119308e-05, "loss": 0.0243, "step": 30930 }, { "grad_norm": 0.3986644148826599, "learning_rate": 7.207302601614166e-05, "loss": 0.0291, "step": 30940 }, { "grad_norm": 0.24950721859931946, "learning_rate": 7.205447879941378e-05, "loss": 0.0338, "step": 30950 }, { "grad_norm": 0.249501571059227, "learning_rate": 7.203592781417866e-05, "loss": 0.0299, "step": 30960 }, { "grad_norm": 0.43348929286003113, "learning_rate": 7.201737306360617e-05, "loss": 0.0341, "step": 30970 }, { "grad_norm": 0.31982341408729553, "learning_rate": 7.19988145508668e-05, "loss": 0.034, "step": 30980 }, { "grad_norm": 0.37614452838897705, "learning_rate": 7.198025227913168e-05, "loss": 0.0311, "step": 30990 }, { "grad_norm": 0.2868342101573944, "learning_rate": 7.196168625157261e-05, "loss": 0.0341, "step": 31000 }, { "grad_norm": 0.4188278913497925, "learning_rate": 7.194311647136201e-05, "loss": 0.0303, "step": 31010 }, { "grad_norm": 0.349845826625824, "learning_rate": 7.192454294167297e-05, "loss": 0.0276, "step": 31020 }, { "grad_norm": 0.3660240173339844, "learning_rate": 7.190596566567917e-05, "loss": 0.0323, "step": 31030 }, { "grad_norm": 0.3585551381111145, "learning_rate": 7.188738464655496e-05, "loss": 0.0337, "step": 31040 }, { "grad_norm": 0.24803079664707184, "learning_rate": 7.186879988747533e-05, "loss": 0.0299, "step": 31050 }, { "grad_norm": 0.3097092807292938, "learning_rate": 7.185021139161592e-05, "loss": 0.0248, "step": 31060 }, { "grad_norm": 0.330502986907959, "learning_rate": 7.1831619162153e-05, "loss": 0.0323, "step": 31070 }, { "grad_norm": 0.3445526957511902, "learning_rate": 7.181302320226345e-05, "loss": 0.0289, "step": 31080 }, { "grad_norm": 0.3111013174057007, "learning_rate": 7.179442351512482e-05, "loss": 0.0402, "step": 31090 }, { "grad_norm": 0.3895997405052185, "learning_rate": 7.177582010391528e-05, "loss": 0.0311, "step": 31100 }, { "grad_norm": 0.4312616288661957, "learning_rate": 7.175721297181366e-05, "loss": 0.0337, "step": 31110 }, { "grad_norm": 0.4023899734020233, "learning_rate": 7.173860212199942e-05, "loss": 0.0333, "step": 31120 }, { "grad_norm": 0.38059350848197937, "learning_rate": 7.171998755765263e-05, "loss": 0.0326, "step": 31130 }, { "grad_norm": 0.5335258841514587, "learning_rate": 7.170136928195398e-05, "loss": 0.0335, "step": 31140 }, { "grad_norm": 0.27144065499305725, "learning_rate": 7.168274729808489e-05, "loss": 0.032, "step": 31150 }, { "grad_norm": 0.4899453818798065, "learning_rate": 7.166412160922728e-05, "loss": 0.0343, "step": 31160 }, { "grad_norm": 0.31835824251174927, "learning_rate": 7.164549221856382e-05, "loss": 0.031, "step": 31170 }, { "grad_norm": 0.3914514482021332, "learning_rate": 7.162685912927775e-05, "loss": 0.0415, "step": 31180 }, { "grad_norm": 0.2878391444683075, "learning_rate": 7.160822234455294e-05, "loss": 0.0294, "step": 31190 }, { "grad_norm": 0.2910045385360718, "learning_rate": 7.158958186757391e-05, "loss": 0.0241, "step": 31200 }, { "grad_norm": 0.36100098490715027, "learning_rate": 7.157093770152582e-05, "loss": 0.0364, "step": 31210 }, { "grad_norm": 0.46048834919929504, "learning_rate": 7.155228984959446e-05, "loss": 0.0327, "step": 31220 }, { "grad_norm": 0.32164719700813293, "learning_rate": 7.153363831496621e-05, "loss": 0.0372, "step": 31230 }, { "grad_norm": 0.35258352756500244, "learning_rate": 7.151498310082811e-05, "loss": 0.0417, "step": 31240 }, { "grad_norm": 0.39342761039733887, "learning_rate": 7.149632421036784e-05, "loss": 0.0403, "step": 31250 }, { "grad_norm": 0.37272122502326965, "learning_rate": 7.147766164677369e-05, "loss": 0.0565, "step": 31260 }, { "grad_norm": 0.24750667810440063, "learning_rate": 7.145899541323459e-05, "loss": 0.0262, "step": 31270 }, { "grad_norm": 0.34300094842910767, "learning_rate": 7.144032551294007e-05, "loss": 0.044, "step": 31280 }, { "grad_norm": 0.3448483645915985, "learning_rate": 7.14216519490803e-05, "loss": 0.0427, "step": 31290 }, { "grad_norm": 0.4726579189300537, "learning_rate": 7.140297472484609e-05, "loss": 0.0442, "step": 31300 }, { "grad_norm": 0.3580295741558075, "learning_rate": 7.138429384342891e-05, "loss": 0.0576, "step": 31310 }, { "grad_norm": 0.3693157434463501, "learning_rate": 7.136560930802074e-05, "loss": 0.0361, "step": 31320 }, { "grad_norm": 0.49580132961273193, "learning_rate": 7.134692112181431e-05, "loss": 0.0347, "step": 31330 }, { "grad_norm": 0.35250794887542725, "learning_rate": 7.13282292880029e-05, "loss": 0.0291, "step": 31340 }, { "grad_norm": 0.2764410674571991, "learning_rate": 7.130953380978043e-05, "loss": 0.0331, "step": 31350 }, { "grad_norm": 0.3125995099544525, "learning_rate": 7.129083469034144e-05, "loss": 0.0349, "step": 31360 }, { "grad_norm": 0.362079381942749, "learning_rate": 7.127213193288112e-05, "loss": 0.0376, "step": 31370 }, { "grad_norm": 0.28960636258125305, "learning_rate": 7.125342554059522e-05, "loss": 0.032, "step": 31380 }, { "grad_norm": 0.39852023124694824, "learning_rate": 7.12347155166802e-05, "loss": 0.0345, "step": 31390 }, { "grad_norm": 0.2636469602584839, "learning_rate": 7.121600186433306e-05, "loss": 0.0311, "step": 31400 }, { "grad_norm": 0.19974780082702637, "learning_rate": 7.119728458675148e-05, "loss": 0.0327, "step": 31410 }, { "grad_norm": 0.4074191451072693, "learning_rate": 7.117856368713369e-05, "loss": 0.0287, "step": 31420 }, { "grad_norm": 0.380524605512619, "learning_rate": 7.115983916867861e-05, "loss": 0.0337, "step": 31430 }, { "grad_norm": 0.3846590220928192, "learning_rate": 7.114111103458574e-05, "loss": 0.0321, "step": 31440 }, { "grad_norm": 0.32720088958740234, "learning_rate": 7.11223792880552e-05, "loss": 0.0315, "step": 31450 }, { "grad_norm": 0.4223559498786926, "learning_rate": 7.110364393228773e-05, "loss": 0.0331, "step": 31460 }, { "grad_norm": 0.2774523198604584, "learning_rate": 7.108490497048471e-05, "loss": 0.0319, "step": 31470 }, { "grad_norm": 0.40254977345466614, "learning_rate": 7.10661624058481e-05, "loss": 0.0409, "step": 31480 }, { "grad_norm": 0.3661767840385437, "learning_rate": 7.10474162415805e-05, "loss": 0.0415, "step": 31490 }, { "grad_norm": 0.20422662794589996, "learning_rate": 7.102866648088511e-05, "loss": 0.0383, "step": 31500 }, { "grad_norm": 0.29256927967071533, "learning_rate": 7.100991312696576e-05, "loss": 0.0322, "step": 31510 }, { "grad_norm": 0.4394405484199524, "learning_rate": 7.099115618302686e-05, "loss": 0.0363, "step": 31520 }, { "grad_norm": 0.3760278522968292, "learning_rate": 7.097239565227349e-05, "loss": 0.0428, "step": 31530 }, { "grad_norm": 0.3319401741027832, "learning_rate": 7.09536315379113e-05, "loss": 0.0315, "step": 31540 }, { "grad_norm": 0.31192833185195923, "learning_rate": 7.093486384314656e-05, "loss": 0.0369, "step": 31550 }, { "grad_norm": 0.38439157605171204, "learning_rate": 7.091609257118616e-05, "loss": 0.0362, "step": 31560 }, { "grad_norm": 0.417256623506546, "learning_rate": 7.08973177252376e-05, "loss": 0.0425, "step": 31570 }, { "grad_norm": 0.3587152063846588, "learning_rate": 7.087853930850898e-05, "loss": 0.0296, "step": 31580 }, { "grad_norm": 0.485907644033432, "learning_rate": 7.085975732420903e-05, "loss": 0.0329, "step": 31590 }, { "grad_norm": 0.32358860969543457, "learning_rate": 7.084097177554706e-05, "loss": 0.0241, "step": 31600 }, { "grad_norm": 0.38578325510025024, "learning_rate": 7.082218266573301e-05, "loss": 0.0365, "step": 31610 }, { "grad_norm": 0.3428724408149719, "learning_rate": 7.080338999797743e-05, "loss": 0.0356, "step": 31620 }, { "grad_norm": 0.28559401631355286, "learning_rate": 7.07845937754915e-05, "loss": 0.0344, "step": 31630 }, { "grad_norm": 0.3407251238822937, "learning_rate": 7.076579400148693e-05, "loss": 0.0347, "step": 31640 }, { "grad_norm": 0.4067568778991699, "learning_rate": 7.074699067917611e-05, "loss": 0.0342, "step": 31650 }, { "grad_norm": 0.3329307734966278, "learning_rate": 7.072818381177201e-05, "loss": 0.0308, "step": 31660 }, { "grad_norm": 0.3974453806877136, "learning_rate": 7.070937340248823e-05, "loss": 0.0356, "step": 31670 }, { "grad_norm": 0.35249611735343933, "learning_rate": 7.069055945453893e-05, "loss": 0.0336, "step": 31680 }, { "grad_norm": 0.471356600522995, "learning_rate": 7.067174197113892e-05, "loss": 0.0309, "step": 31690 }, { "grad_norm": 0.31072407960891724, "learning_rate": 7.065292095550355e-05, "loss": 0.0351, "step": 31700 }, { "grad_norm": 0.2705758213996887, "learning_rate": 7.063409641084887e-05, "loss": 0.0324, "step": 31710 }, { "grad_norm": 0.36589112877845764, "learning_rate": 7.061526834039145e-05, "loss": 0.0365, "step": 31720 }, { "grad_norm": 0.34519049525260925, "learning_rate": 7.05964367473485e-05, "loss": 0.0275, "step": 31730 }, { "grad_norm": 0.322584331035614, "learning_rate": 7.057760163493783e-05, "loss": 0.0398, "step": 31740 }, { "grad_norm": 0.340356707572937, "learning_rate": 7.055876300637783e-05, "loss": 0.0303, "step": 31750 }, { "grad_norm": 0.3875391483306885, "learning_rate": 7.053992086488753e-05, "loss": 0.0336, "step": 31760 }, { "grad_norm": 0.2705978751182556, "learning_rate": 7.052107521368651e-05, "loss": 0.0392, "step": 31770 }, { "grad_norm": 0.2999366819858551, "learning_rate": 7.0502226055995e-05, "loss": 0.0379, "step": 31780 }, { "grad_norm": 0.4192316234111786, "learning_rate": 7.048337339503379e-05, "loss": 0.0458, "step": 31790 }, { "grad_norm": 0.37892234325408936, "learning_rate": 7.046451723402427e-05, "loss": 0.0354, "step": 31800 }, { "grad_norm": 0.251666396856308, "learning_rate": 7.044565757618848e-05, "loss": 0.0395, "step": 31810 }, { "grad_norm": 0.4888108968734741, "learning_rate": 7.042679442474899e-05, "loss": 0.0339, "step": 31820 }, { "grad_norm": 0.2640402019023895, "learning_rate": 7.040792778292902e-05, "loss": 0.0445, "step": 31830 }, { "grad_norm": 0.3624168336391449, "learning_rate": 7.038905765395234e-05, "loss": 0.0489, "step": 31840 }, { "grad_norm": 0.44101786613464355, "learning_rate": 7.037018404104334e-05, "loss": 0.0375, "step": 31850 }, { "grad_norm": 0.3941471576690674, "learning_rate": 7.035130694742702e-05, "loss": 0.0332, "step": 31860 }, { "grad_norm": 0.4214556813240051, "learning_rate": 7.033242637632897e-05, "loss": 0.039, "step": 31870 }, { "grad_norm": 0.4649091064929962, "learning_rate": 7.031354233097534e-05, "loss": 0.0353, "step": 31880 }, { "grad_norm": 0.4209668040275574, "learning_rate": 7.029465481459289e-05, "loss": 0.0503, "step": 31890 }, { "grad_norm": 0.45193055272102356, "learning_rate": 7.027576383040898e-05, "loss": 0.03, "step": 31900 }, { "grad_norm": 0.2778977155685425, "learning_rate": 7.025686938165159e-05, "loss": 0.0419, "step": 31910 }, { "grad_norm": 0.4421692192554474, "learning_rate": 7.023797147154924e-05, "loss": 0.052, "step": 31920 }, { "grad_norm": 0.2963405251502991, "learning_rate": 7.021907010333111e-05, "loss": 0.0299, "step": 31930 }, { "grad_norm": 0.3037007451057434, "learning_rate": 7.020016528022685e-05, "loss": 0.0348, "step": 31940 }, { "grad_norm": 0.3447698950767517, "learning_rate": 7.018125700546683e-05, "loss": 0.0374, "step": 31950 }, { "grad_norm": 0.37651950120925903, "learning_rate": 7.016234528228196e-05, "loss": 0.0379, "step": 31960 }, { "grad_norm": 0.33178767561912537, "learning_rate": 7.014343011390372e-05, "loss": 0.0314, "step": 31970 }, { "grad_norm": 0.3245031237602234, "learning_rate": 7.01245115035642e-05, "loss": 0.0254, "step": 31980 }, { "grad_norm": 0.29566672444343567, "learning_rate": 7.010558945449606e-05, "loss": 0.0382, "step": 31990 }, { "grad_norm": 0.30193445086479187, "learning_rate": 7.008666396993258e-05, "loss": 0.0289, "step": 32000 }, { "grad_norm": 0.3854001760482788, "learning_rate": 7.006773505310759e-05, "loss": 0.0363, "step": 32010 }, { "grad_norm": 0.42294448614120483, "learning_rate": 7.004880270725553e-05, "loss": 0.0342, "step": 32020 }, { "grad_norm": 0.37380659580230713, "learning_rate": 7.002986693561144e-05, "loss": 0.0294, "step": 32030 }, { "grad_norm": 0.3652181029319763, "learning_rate": 7.001092774141089e-05, "loss": 0.032, "step": 32040 }, { "grad_norm": 0.3773382604122162, "learning_rate": 6.999198512789009e-05, "loss": 0.0378, "step": 32050 }, { "grad_norm": 0.3701460659503937, "learning_rate": 6.997303909828584e-05, "loss": 0.0379, "step": 32060 }, { "grad_norm": 0.3669898509979248, "learning_rate": 6.995408965583544e-05, "loss": 0.0388, "step": 32070 }, { "grad_norm": 0.20858655869960785, "learning_rate": 6.993513680377688e-05, "loss": 0.0251, "step": 32080 }, { "grad_norm": 0.2936285734176636, "learning_rate": 6.991618054534868e-05, "loss": 0.0369, "step": 32090 }, { "grad_norm": 0.3601170778274536, "learning_rate": 6.989722088378991e-05, "loss": 0.0358, "step": 32100 }, { "grad_norm": 0.2873420715332031, "learning_rate": 6.987825782234027e-05, "loss": 0.0333, "step": 32110 }, { "grad_norm": 0.296253502368927, "learning_rate": 6.985929136424006e-05, "loss": 0.0333, "step": 32120 }, { "grad_norm": 0.4093761146068573, "learning_rate": 6.984032151273012e-05, "loss": 0.0391, "step": 32130 }, { "grad_norm": 0.3616648018360138, "learning_rate": 6.982134827105186e-05, "loss": 0.0328, "step": 32140 }, { "grad_norm": 0.40710514783859253, "learning_rate": 6.980237164244729e-05, "loss": 0.0303, "step": 32150 }, { "grad_norm": 0.3429698050022125, "learning_rate": 6.9783391630159e-05, "loss": 0.0257, "step": 32160 }, { "grad_norm": 0.40756598114967346, "learning_rate": 6.976440823743015e-05, "loss": 0.0395, "step": 32170 }, { "grad_norm": 0.25720930099487305, "learning_rate": 6.974542146750451e-05, "loss": 0.0261, "step": 32180 }, { "grad_norm": 0.2818390130996704, "learning_rate": 6.972643132362637e-05, "loss": 0.026, "step": 32190 }, { "grad_norm": 0.34017425775527954, "learning_rate": 6.970743780904064e-05, "loss": 0.0314, "step": 32200 }, { "grad_norm": 0.3486344516277313, "learning_rate": 6.968844092699277e-05, "loss": 0.0371, "step": 32210 }, { "grad_norm": 0.3573107421398163, "learning_rate": 6.966944068072883e-05, "loss": 0.0348, "step": 32220 }, { "grad_norm": 0.47524771094322205, "learning_rate": 6.965043707349545e-05, "loss": 0.0339, "step": 32230 }, { "grad_norm": 0.321214884519577, "learning_rate": 6.963143010853982e-05, "loss": 0.0354, "step": 32240 }, { "grad_norm": 0.4015645980834961, "learning_rate": 6.961241978910971e-05, "loss": 0.0394, "step": 32250 }, { "grad_norm": 0.4016427993774414, "learning_rate": 6.959340611845344e-05, "loss": 0.0316, "step": 32260 }, { "grad_norm": 0.34013935923576355, "learning_rate": 6.957438909981995e-05, "loss": 0.0282, "step": 32270 }, { "grad_norm": 0.31314200162887573, "learning_rate": 6.955536873645872e-05, "loss": 0.0241, "step": 32280 }, { "grad_norm": 0.32320284843444824, "learning_rate": 6.953634503161982e-05, "loss": 0.0345, "step": 32290 }, { "grad_norm": 0.34529298543930054, "learning_rate": 6.951731798855387e-05, "loss": 0.0307, "step": 32300 }, { "grad_norm": 0.24415238201618195, "learning_rate": 6.949828761051208e-05, "loss": 0.0394, "step": 32310 }, { "grad_norm": 0.38104039430618286, "learning_rate": 6.947925390074622e-05, "loss": 0.0272, "step": 32320 }, { "grad_norm": 0.3162935972213745, "learning_rate": 6.946021686250863e-05, "loss": 0.0272, "step": 32330 }, { "grad_norm": 0.2892204225063324, "learning_rate": 6.94411764990522e-05, "loss": 0.037, "step": 32340 }, { "grad_norm": 0.3483217656612396, "learning_rate": 6.942213281363044e-05, "loss": 0.0296, "step": 32350 }, { "grad_norm": 0.27951478958129883, "learning_rate": 6.940308580949737e-05, "loss": 0.03, "step": 32360 }, { "grad_norm": 0.34793978929519653, "learning_rate": 6.93840354899076e-05, "loss": 0.0294, "step": 32370 }, { "grad_norm": 0.33439987897872925, "learning_rate": 6.936498185811633e-05, "loss": 0.037, "step": 32380 }, { "grad_norm": 0.3082813322544098, "learning_rate": 6.93459249173793e-05, "loss": 0.0264, "step": 32390 }, { "grad_norm": 0.5222180485725403, "learning_rate": 6.932686467095279e-05, "loss": 0.0336, "step": 32400 }, { "grad_norm": 0.411638081073761, "learning_rate": 6.930780112209373e-05, "loss": 0.0356, "step": 32410 }, { "grad_norm": 0.4170653223991394, "learning_rate": 6.92887342740595e-05, "loss": 0.033, "step": 32420 }, { "grad_norm": 0.4221501648426056, "learning_rate": 6.926966413010816e-05, "loss": 0.0381, "step": 32430 }, { "grad_norm": 0.33482423424720764, "learning_rate": 6.925059069349824e-05, "loss": 0.026, "step": 32440 }, { "grad_norm": 0.44047248363494873, "learning_rate": 6.923151396748886e-05, "loss": 0.028, "step": 32450 }, { "grad_norm": 0.30128127336502075, "learning_rate": 6.921243395533974e-05, "loss": 0.0276, "step": 32460 }, { "grad_norm": 0.2712172269821167, "learning_rate": 6.919335066031109e-05, "loss": 0.0317, "step": 32470 }, { "grad_norm": 0.28761279582977295, "learning_rate": 6.917426408566379e-05, "loss": 0.0352, "step": 32480 }, { "grad_norm": 0.45394474267959595, "learning_rate": 6.915517423465916e-05, "loss": 0.0309, "step": 32490 }, { "grad_norm": 0.21915625035762787, "learning_rate": 6.913608111055914e-05, "loss": 0.0311, "step": 32500 }, { "grad_norm": 0.3464447259902954, "learning_rate": 6.911698471662623e-05, "loss": 0.0265, "step": 32510 }, { "grad_norm": 0.32205140590667725, "learning_rate": 6.90978850561235e-05, "loss": 0.0282, "step": 32520 }, { "grad_norm": 0.3660857677459717, "learning_rate": 6.907878213231454e-05, "loss": 0.0287, "step": 32530 }, { "grad_norm": 0.4474292993545532, "learning_rate": 6.90596759484635e-05, "loss": 0.0335, "step": 32540 }, { "grad_norm": 0.3250960111618042, "learning_rate": 6.904056650783514e-05, "loss": 0.0429, "step": 32550 }, { "grad_norm": 0.2821216583251953, "learning_rate": 6.902145381369471e-05, "loss": 0.0337, "step": 32560 }, { "grad_norm": 0.2759222388267517, "learning_rate": 6.900233786930808e-05, "loss": 0.0343, "step": 32570 }, { "grad_norm": 0.44416284561157227, "learning_rate": 6.898321867794161e-05, "loss": 0.0366, "step": 32580 }, { "grad_norm": 0.25989270210266113, "learning_rate": 6.896409624286226e-05, "loss": 0.0331, "step": 32590 }, { "grad_norm": 0.32897284626960754, "learning_rate": 6.894497056733754e-05, "loss": 0.0415, "step": 32600 }, { "grad_norm": 0.32119008898735046, "learning_rate": 6.89258416546355e-05, "loss": 0.0306, "step": 32610 }, { "grad_norm": 0.3650382161140442, "learning_rate": 6.890670950802474e-05, "loss": 0.037, "step": 32620 }, { "grad_norm": 0.4225190281867981, "learning_rate": 6.88875741307744e-05, "loss": 0.0302, "step": 32630 }, { "grad_norm": 0.35721904039382935, "learning_rate": 6.886843552615425e-05, "loss": 0.03, "step": 32640 }, { "grad_norm": 0.2983289957046509, "learning_rate": 6.884929369743451e-05, "loss": 0.0308, "step": 32650 }, { "grad_norm": 0.41014814376831055, "learning_rate": 6.8830148647886e-05, "loss": 0.0311, "step": 32660 }, { "grad_norm": 0.32783326506614685, "learning_rate": 6.88110003807801e-05, "loss": 0.0366, "step": 32670 }, { "grad_norm": 0.40114930272102356, "learning_rate": 6.87918488993887e-05, "loss": 0.0288, "step": 32680 }, { "grad_norm": 0.4239172637462616, "learning_rate": 6.877269420698431e-05, "loss": 0.0372, "step": 32690 }, { "grad_norm": 0.3991296887397766, "learning_rate": 6.875353630683989e-05, "loss": 0.0347, "step": 32700 }, { "grad_norm": 0.42622220516204834, "learning_rate": 6.873437520222905e-05, "loss": 0.0419, "step": 32710 }, { "grad_norm": 0.2825235426425934, "learning_rate": 6.871521089642585e-05, "loss": 0.03, "step": 32720 }, { "grad_norm": 0.2813358008861542, "learning_rate": 6.869604339270498e-05, "loss": 0.0397, "step": 32730 }, { "grad_norm": 0.36665064096450806, "learning_rate": 6.867687269434164e-05, "loss": 0.0394, "step": 32740 }, { "grad_norm": 0.2882465720176697, "learning_rate": 6.865769880461156e-05, "loss": 0.0279, "step": 32750 }, { "grad_norm": 0.23881036043167114, "learning_rate": 6.863852172679104e-05, "loss": 0.0325, "step": 32760 }, { "grad_norm": 0.47116899490356445, "learning_rate": 6.861934146415693e-05, "loss": 0.0315, "step": 32770 }, { "grad_norm": 0.3510734736919403, "learning_rate": 6.86001580199866e-05, "loss": 0.0392, "step": 32780 }, { "grad_norm": 0.2921178340911865, "learning_rate": 6.858097139755798e-05, "loss": 0.0281, "step": 32790 }, { "grad_norm": 0.40627509355545044, "learning_rate": 6.856178160014955e-05, "loss": 0.0297, "step": 32800 }, { "grad_norm": 0.4378393888473511, "learning_rate": 6.85425886310403e-05, "loss": 0.0322, "step": 32810 }, { "grad_norm": 0.34023648500442505, "learning_rate": 6.852339249350979e-05, "loss": 0.0347, "step": 32820 }, { "grad_norm": 0.32180577516555786, "learning_rate": 6.850419319083812e-05, "loss": 0.0398, "step": 32830 }, { "grad_norm": 0.26052430272102356, "learning_rate": 6.848499072630592e-05, "loss": 0.0243, "step": 32840 }, { "grad_norm": 0.3464442491531372, "learning_rate": 6.846578510319439e-05, "loss": 0.0318, "step": 32850 }, { "grad_norm": 0.27677223086357117, "learning_rate": 6.844657632478519e-05, "loss": 0.0356, "step": 32860 }, { "grad_norm": 0.25377124547958374, "learning_rate": 6.842736439436063e-05, "loss": 0.0335, "step": 32870 }, { "grad_norm": 0.283037006855011, "learning_rate": 6.84081493152035e-05, "loss": 0.0324, "step": 32880 }, { "grad_norm": 0.2954995632171631, "learning_rate": 6.83889310905971e-05, "loss": 0.0264, "step": 32890 }, { "grad_norm": 0.34121426939964294, "learning_rate": 6.836970972382533e-05, "loss": 0.0369, "step": 32900 }, { "grad_norm": 0.4198858439922333, "learning_rate": 6.835048521817257e-05, "loss": 0.0281, "step": 32910 }, { "grad_norm": 0.39270415902137756, "learning_rate": 6.833125757692379e-05, "loss": 0.0353, "step": 32920 }, { "grad_norm": 0.27924299240112305, "learning_rate": 6.831202680336441e-05, "loss": 0.0306, "step": 32930 }, { "grad_norm": 0.21626971662044525, "learning_rate": 6.829279290078052e-05, "loss": 0.0257, "step": 32940 }, { "grad_norm": 0.318778932094574, "learning_rate": 6.827355587245863e-05, "loss": 0.0288, "step": 32950 }, { "grad_norm": 0.25727027654647827, "learning_rate": 6.82543157216858e-05, "loss": 0.0319, "step": 32960 }, { "grad_norm": 0.3195636570453644, "learning_rate": 6.823507245174969e-05, "loss": 0.0348, "step": 32970 }, { "grad_norm": 0.32618454098701477, "learning_rate": 6.821582606593841e-05, "loss": 0.0281, "step": 32980 }, { "grad_norm": 0.3496266007423401, "learning_rate": 6.81965765675407e-05, "loss": 0.0319, "step": 32990 }, { "grad_norm": 0.48112615942955017, "learning_rate": 6.81773239598457e-05, "loss": 0.0333, "step": 33000 }, { "grad_norm": 0.27832847833633423, "learning_rate": 6.815806824614319e-05, "loss": 0.033, "step": 33010 }, { "grad_norm": 0.33404436707496643, "learning_rate": 6.813880942972343e-05, "loss": 0.0269, "step": 33020 }, { "grad_norm": 0.2564575970172882, "learning_rate": 6.811954751387726e-05, "loss": 0.0308, "step": 33030 }, { "grad_norm": 0.30799955129623413, "learning_rate": 6.810028250189598e-05, "loss": 0.0275, "step": 33040 }, { "grad_norm": 0.49501049518585205, "learning_rate": 6.808101439707147e-05, "loss": 0.0385, "step": 33050 }, { "grad_norm": 0.3179355561733246, "learning_rate": 6.806174320269609e-05, "loss": 0.0414, "step": 33060 }, { "grad_norm": 0.2675655782222748, "learning_rate": 6.804246892206281e-05, "loss": 0.0337, "step": 33070 }, { "grad_norm": 0.7826859354972839, "learning_rate": 6.802319155846506e-05, "loss": 0.0499, "step": 33080 }, { "grad_norm": 0.4959794580936432, "learning_rate": 6.800391111519679e-05, "loss": 0.0307, "step": 33090 }, { "grad_norm": 0.3507099747657776, "learning_rate": 6.798462759555253e-05, "loss": 0.0433, "step": 33100 }, { "grad_norm": 0.3013898730278015, "learning_rate": 6.79653410028273e-05, "loss": 0.0293, "step": 33110 }, { "grad_norm": 0.22632314264774323, "learning_rate": 6.794605134031663e-05, "loss": 0.0302, "step": 33120 }, { "grad_norm": 0.3924561142921448, "learning_rate": 6.792675861131661e-05, "loss": 0.0362, "step": 33130 }, { "grad_norm": 0.3685589134693146, "learning_rate": 6.790746281912386e-05, "loss": 0.0405, "step": 33140 }, { "grad_norm": 0.3584800064563751, "learning_rate": 6.788816396703546e-05, "loss": 0.0266, "step": 33150 }, { "grad_norm": 0.3981892168521881, "learning_rate": 6.78688620583491e-05, "loss": 0.031, "step": 33160 }, { "grad_norm": 0.2817741632461548, "learning_rate": 6.784955709636292e-05, "loss": 0.031, "step": 33170 }, { "grad_norm": 0.4053814113140106, "learning_rate": 6.783024908437564e-05, "loss": 0.0296, "step": 33180 }, { "grad_norm": 0.27680647373199463, "learning_rate": 6.781093802568641e-05, "loss": 0.0248, "step": 33190 }, { "grad_norm": 0.3184238374233246, "learning_rate": 6.779162392359504e-05, "loss": 0.0278, "step": 33200 }, { "grad_norm": 0.32090842723846436, "learning_rate": 6.777230678140172e-05, "loss": 0.0323, "step": 33210 }, { "grad_norm": 0.374491810798645, "learning_rate": 6.775298660240726e-05, "loss": 0.033, "step": 33220 }, { "grad_norm": 0.42561715841293335, "learning_rate": 6.773366338991292e-05, "loss": 0.0483, "step": 33230 }, { "grad_norm": 0.3247292637825012, "learning_rate": 6.771433714722052e-05, "loss": 0.0316, "step": 33240 }, { "grad_norm": 0.2423112988471985, "learning_rate": 6.769500787763239e-05, "loss": 0.0332, "step": 33250 }, { "grad_norm": 0.2981983423233032, "learning_rate": 6.76756755844514e-05, "loss": 0.0335, "step": 33260 }, { "grad_norm": 0.407898485660553, "learning_rate": 6.765634027098087e-05, "loss": 0.0278, "step": 33270 }, { "grad_norm": 0.46930545568466187, "learning_rate": 6.763700194052468e-05, "loss": 0.0385, "step": 33280 }, { "grad_norm": 0.2693851888179779, "learning_rate": 6.761766059638723e-05, "loss": 0.0347, "step": 33290 }, { "grad_norm": 0.34948521852493286, "learning_rate": 6.759831624187345e-05, "loss": 0.0293, "step": 33300 }, { "grad_norm": 0.5101813673973083, "learning_rate": 6.757896888028871e-05, "loss": 0.0398, "step": 33310 }, { "grad_norm": 0.3511612117290497, "learning_rate": 6.7559618514939e-05, "loss": 0.0317, "step": 33320 }, { "grad_norm": 0.3712605834007263, "learning_rate": 6.754026514913073e-05, "loss": 0.0314, "step": 33330 }, { "grad_norm": 0.20845310389995575, "learning_rate": 6.752090878617087e-05, "loss": 0.0434, "step": 33340 }, { "grad_norm": 0.5104928612709045, "learning_rate": 6.75015494293669e-05, "loss": 0.0297, "step": 33350 }, { "grad_norm": 0.38849133253097534, "learning_rate": 6.74821870820268e-05, "loss": 0.0296, "step": 33360 }, { "grad_norm": 0.23884133994579315, "learning_rate": 6.746282174745907e-05, "loss": 0.036, "step": 33370 }, { "grad_norm": 0.41885271668434143, "learning_rate": 6.744345342897271e-05, "loss": 0.0354, "step": 33380 }, { "grad_norm": 0.3742160499095917, "learning_rate": 6.742408212987724e-05, "loss": 0.032, "step": 33390 }, { "grad_norm": 0.31664925813674927, "learning_rate": 6.740470785348269e-05, "loss": 0.0252, "step": 33400 }, { "grad_norm": 0.6284233927726746, "learning_rate": 6.738533060309958e-05, "loss": 0.03, "step": 33410 }, { "grad_norm": 0.4803586006164551, "learning_rate": 6.736595038203894e-05, "loss": 0.029, "step": 33420 }, { "grad_norm": 0.31877759099006653, "learning_rate": 6.734656719361236e-05, "loss": 0.0323, "step": 33430 }, { "grad_norm": 0.31819948554039, "learning_rate": 6.732718104113189e-05, "loss": 0.0351, "step": 33440 }, { "grad_norm": 0.3635142743587494, "learning_rate": 6.730779192791006e-05, "loss": 0.0286, "step": 33450 }, { "grad_norm": 0.3298574686050415, "learning_rate": 6.728839985725997e-05, "loss": 0.0325, "step": 33460 }, { "grad_norm": 0.46698427200317383, "learning_rate": 6.726900483249517e-05, "loss": 0.0328, "step": 33470 }, { "grad_norm": 0.293472558259964, "learning_rate": 6.724960685692976e-05, "loss": 0.0317, "step": 33480 }, { "grad_norm": 0.36672335863113403, "learning_rate": 6.723020593387833e-05, "loss": 0.0245, "step": 33490 }, { "grad_norm": 0.5602622032165527, "learning_rate": 6.721080206665593e-05, "loss": 0.0367, "step": 33500 }, { "grad_norm": 0.22372280061244965, "learning_rate": 6.719139525857819e-05, "loss": 0.0388, "step": 33510 }, { "grad_norm": 0.3309312164783478, "learning_rate": 6.717198551296117e-05, "loss": 0.0314, "step": 33520 }, { "grad_norm": 0.3444872200489044, "learning_rate": 6.715257283312148e-05, "loss": 0.0325, "step": 33530 }, { "grad_norm": 0.24479568004608154, "learning_rate": 6.713315722237623e-05, "loss": 0.0332, "step": 33540 }, { "grad_norm": 0.3334200382232666, "learning_rate": 6.7113738684043e-05, "loss": 0.0344, "step": 33550 }, { "grad_norm": 0.2684912085533142, "learning_rate": 6.709431722143989e-05, "loss": 0.0321, "step": 33560 }, { "grad_norm": 0.3439849317073822, "learning_rate": 6.70748928378855e-05, "loss": 0.0374, "step": 33570 }, { "grad_norm": 0.30711326003074646, "learning_rate": 6.705546553669891e-05, "loss": 0.0227, "step": 33580 }, { "grad_norm": 0.28137171268463135, "learning_rate": 6.703603532119974e-05, "loss": 0.0252, "step": 33590 }, { "grad_norm": 0.38521698117256165, "learning_rate": 6.701660219470808e-05, "loss": 0.0267, "step": 33600 }, { "grad_norm": 0.32725781202316284, "learning_rate": 6.69971661605445e-05, "loss": 0.0266, "step": 33610 }, { "grad_norm": 0.34714454412460327, "learning_rate": 6.697772722203008e-05, "loss": 0.0263, "step": 33620 }, { "grad_norm": 0.298020601272583, "learning_rate": 6.695828538248643e-05, "loss": 0.0286, "step": 33630 }, { "grad_norm": 0.3296477198600769, "learning_rate": 6.693884064523563e-05, "loss": 0.0297, "step": 33640 }, { "grad_norm": 0.24345311522483826, "learning_rate": 6.691939301360023e-05, "loss": 0.0326, "step": 33650 }, { "grad_norm": 0.367960125207901, "learning_rate": 6.689994249090333e-05, "loss": 0.0264, "step": 33660 }, { "grad_norm": 0.34055835008621216, "learning_rate": 6.688048908046845e-05, "loss": 0.0341, "step": 33670 }, { "grad_norm": 0.2551913857460022, "learning_rate": 6.686103278561969e-05, "loss": 0.0239, "step": 33680 }, { "grad_norm": 0.3376066982746124, "learning_rate": 6.684157360968156e-05, "loss": 0.0305, "step": 33690 }, { "grad_norm": 0.24858558177947998, "learning_rate": 6.682211155597911e-05, "loss": 0.0232, "step": 33700 }, { "grad_norm": 0.3544309437274933, "learning_rate": 6.680264662783789e-05, "loss": 0.0326, "step": 33710 }, { "grad_norm": 0.3632745146751404, "learning_rate": 6.678317882858391e-05, "loss": 0.0326, "step": 33720 }, { "grad_norm": 0.17982500791549683, "learning_rate": 6.67637081615437e-05, "loss": 0.0276, "step": 33730 }, { "grad_norm": 0.26074469089508057, "learning_rate": 6.674423463004427e-05, "loss": 0.0231, "step": 33740 }, { "grad_norm": 0.3468298017978668, "learning_rate": 6.672475823741308e-05, "loss": 0.0283, "step": 33750 }, { "grad_norm": 0.4096843898296356, "learning_rate": 6.670527898697811e-05, "loss": 0.0361, "step": 33760 }, { "grad_norm": 0.3143300414085388, "learning_rate": 6.668579688206788e-05, "loss": 0.033, "step": 33770 }, { "grad_norm": 0.3473871052265167, "learning_rate": 6.666631192601131e-05, "loss": 0.0268, "step": 33780 }, { "grad_norm": 0.23020489513874054, "learning_rate": 6.664682412213785e-05, "loss": 0.031, "step": 33790 }, { "grad_norm": 0.3754231631755829, "learning_rate": 6.662733347377745e-05, "loss": 0.0369, "step": 33800 }, { "grad_norm": 0.24642910063266754, "learning_rate": 6.660783998426051e-05, "loss": 0.0251, "step": 33810 }, { "grad_norm": 0.36285245418548584, "learning_rate": 6.658834365691794e-05, "loss": 0.0264, "step": 33820 }, { "grad_norm": 0.33037513494491577, "learning_rate": 6.656884449508115e-05, "loss": 0.0287, "step": 33830 }, { "grad_norm": 0.3286421298980713, "learning_rate": 6.654934250208198e-05, "loss": 0.0304, "step": 33840 }, { "grad_norm": 0.29689979553222656, "learning_rate": 6.65298376812528e-05, "loss": 0.0266, "step": 33850 }, { "grad_norm": 0.3952597677707672, "learning_rate": 6.651033003592646e-05, "loss": 0.0322, "step": 33860 }, { "grad_norm": 0.2881523072719574, "learning_rate": 6.649081956943626e-05, "loss": 0.0313, "step": 33870 }, { "grad_norm": 0.34775739908218384, "learning_rate": 6.647130628511604e-05, "loss": 0.034, "step": 33880 }, { "grad_norm": 0.3705708384513855, "learning_rate": 6.645179018630005e-05, "loss": 0.0283, "step": 33890 }, { "grad_norm": 0.21263962984085083, "learning_rate": 6.643227127632309e-05, "loss": 0.0329, "step": 33900 }, { "grad_norm": 0.3931506276130676, "learning_rate": 6.641274955852038e-05, "loss": 0.0339, "step": 33910 }, { "grad_norm": 0.41484716534614563, "learning_rate": 6.639322503622768e-05, "loss": 0.0499, "step": 33920 }, { "grad_norm": 0.42615261673927307, "learning_rate": 6.637369771278116e-05, "loss": 0.03, "step": 33930 }, { "grad_norm": 0.5076171159744263, "learning_rate": 6.635416759151751e-05, "loss": 0.0306, "step": 33940 }, { "grad_norm": 0.34290045499801636, "learning_rate": 6.633463467577394e-05, "loss": 0.0443, "step": 33950 }, { "grad_norm": 0.39934396743774414, "learning_rate": 6.631509896888803e-05, "loss": 0.0317, "step": 33960 }, { "grad_norm": 0.3072216212749481, "learning_rate": 6.629556047419794e-05, "loss": 0.0291, "step": 33970 }, { "grad_norm": 0.302854984998703, "learning_rate": 6.627601919504223e-05, "loss": 0.0269, "step": 33980 }, { "grad_norm": 0.5065611004829407, "learning_rate": 6.625647513476001e-05, "loss": 0.0249, "step": 33990 }, { "grad_norm": 0.3450503349304199, "learning_rate": 6.62369282966908e-05, "loss": 0.0321, "step": 34000 }, { "grad_norm": 0.2077847272157669, "learning_rate": 6.621737868417464e-05, "loss": 0.0299, "step": 34010 }, { "grad_norm": 0.2567243278026581, "learning_rate": 6.619782630055198e-05, "loss": 0.0307, "step": 34020 }, { "grad_norm": 0.375530481338501, "learning_rate": 6.617827114916382e-05, "loss": 0.0327, "step": 34030 }, { "grad_norm": 0.2698637843132019, "learning_rate": 6.615871323335161e-05, "loss": 0.0341, "step": 34040 }, { "grad_norm": 0.2987326681613922, "learning_rate": 6.613915255645725e-05, "loss": 0.0336, "step": 34050 }, { "grad_norm": 0.47651976346969604, "learning_rate": 6.611958912182312e-05, "loss": 0.0338, "step": 34060 }, { "grad_norm": 0.24815359711647034, "learning_rate": 6.610002293279207e-05, "loss": 0.0318, "step": 34070 }, { "grad_norm": 0.2750488221645355, "learning_rate": 6.608045399270746e-05, "loss": 0.0291, "step": 34080 }, { "grad_norm": 0.3003648817539215, "learning_rate": 6.606088230491304e-05, "loss": 0.0252, "step": 34090 }, { "grad_norm": 0.33099281787872314, "learning_rate": 6.604130787275312e-05, "loss": 0.0354, "step": 34100 }, { "grad_norm": 0.4268539845943451, "learning_rate": 6.602173069957242e-05, "loss": 0.0307, "step": 34110 }, { "grad_norm": 0.45899468660354614, "learning_rate": 6.600215078871612e-05, "loss": 0.0298, "step": 34120 }, { "grad_norm": 0.2859494090080261, "learning_rate": 6.598256814352992e-05, "loss": 0.0311, "step": 34130 }, { "grad_norm": 0.28786590695381165, "learning_rate": 6.596298276735995e-05, "loss": 0.0232, "step": 34140 }, { "grad_norm": 0.6843684315681458, "learning_rate": 6.594339466355282e-05, "loss": 0.032, "step": 34150 }, { "grad_norm": 0.29149219393730164, "learning_rate": 6.592380383545558e-05, "loss": 0.0278, "step": 34160 }, { "grad_norm": 0.43925270438194275, "learning_rate": 6.590421028641577e-05, "loss": 0.0369, "step": 34170 }, { "grad_norm": 0.30552563071250916, "learning_rate": 6.588461401978143e-05, "loss": 0.0324, "step": 34180 }, { "grad_norm": 0.3091585338115692, "learning_rate": 6.586501503890099e-05, "loss": 0.0277, "step": 34190 }, { "grad_norm": 0.22587950527668, "learning_rate": 6.584541334712338e-05, "loss": 0.0233, "step": 34200 }, { "grad_norm": 0.33110976219177246, "learning_rate": 6.582580894779802e-05, "loss": 0.0304, "step": 34210 }, { "grad_norm": 0.3293417990207672, "learning_rate": 6.580620184427473e-05, "loss": 0.0253, "step": 34220 }, { "grad_norm": 0.21688708662986755, "learning_rate": 6.578659203990385e-05, "loss": 0.0264, "step": 34230 }, { "grad_norm": 0.45413705706596375, "learning_rate": 6.576697953803615e-05, "loss": 0.0254, "step": 34240 }, { "grad_norm": 0.31626221537590027, "learning_rate": 6.57473643420229e-05, "loss": 0.0306, "step": 34250 }, { "grad_norm": 0.2625456750392914, "learning_rate": 6.572774645521574e-05, "loss": 0.0357, "step": 34260 }, { "grad_norm": 0.38699638843536377, "learning_rate": 6.570812588096688e-05, "loss": 0.0362, "step": 34270 }, { "grad_norm": 0.22561591863632202, "learning_rate": 6.568850262262893e-05, "loss": 0.031, "step": 34280 }, { "grad_norm": 0.3653604984283447, "learning_rate": 6.566887668355497e-05, "loss": 0.0295, "step": 34290 }, { "grad_norm": 0.30038967728614807, "learning_rate": 6.564924806709851e-05, "loss": 0.0199, "step": 34300 }, { "grad_norm": 0.4233432710170746, "learning_rate": 6.562961677661359e-05, "loss": 0.0346, "step": 34310 }, { "grad_norm": 0.35048389434814453, "learning_rate": 6.56099828154546e-05, "loss": 0.0252, "step": 34320 }, { "grad_norm": 0.3105415105819702, "learning_rate": 6.55903461869765e-05, "loss": 0.0272, "step": 34330 }, { "grad_norm": 0.2435675859451294, "learning_rate": 6.557070689453465e-05, "loss": 0.0207, "step": 34340 }, { "grad_norm": 0.31002920866012573, "learning_rate": 6.555106494148482e-05, "loss": 0.0332, "step": 34350 }, { "grad_norm": 0.4081987738609314, "learning_rate": 6.553142033118333e-05, "loss": 0.0312, "step": 34360 }, { "grad_norm": 0.3820144832134247, "learning_rate": 6.551177306698688e-05, "loss": 0.0382, "step": 34370 }, { "grad_norm": 0.29856234788894653, "learning_rate": 6.549212315225267e-05, "loss": 0.0305, "step": 34380 }, { "grad_norm": 0.2875464856624603, "learning_rate": 6.547247059033833e-05, "loss": 0.0319, "step": 34390 }, { "grad_norm": 0.31402620673179626, "learning_rate": 6.545281538460193e-05, "loss": 0.0313, "step": 34400 }, { "grad_norm": 0.29383474588394165, "learning_rate": 6.543315753840202e-05, "loss": 0.0262, "step": 34410 }, { "grad_norm": 0.27547207474708557, "learning_rate": 6.541349705509758e-05, "loss": 0.0298, "step": 34420 }, { "grad_norm": 0.3891315460205078, "learning_rate": 6.539383393804805e-05, "loss": 0.0323, "step": 34430 }, { "grad_norm": 0.28181934356689453, "learning_rate": 6.537416819061333e-05, "loss": 0.0353, "step": 34440 }, { "grad_norm": 0.24873791635036469, "learning_rate": 6.535449981615375e-05, "loss": 0.0317, "step": 34450 }, { "grad_norm": 0.2678792178630829, "learning_rate": 6.53348288180301e-05, "loss": 0.0256, "step": 34460 }, { "grad_norm": 0.34954407811164856, "learning_rate": 6.531515519960361e-05, "loss": 0.0394, "step": 34470 }, { "grad_norm": 0.3181208670139313, "learning_rate": 6.529547896423597e-05, "loss": 0.027, "step": 34480 }, { "grad_norm": 0.4825807511806488, "learning_rate": 6.52758001152893e-05, "loss": 0.0322, "step": 34490 }, { "grad_norm": 0.25105980038642883, "learning_rate": 6.525611865612618e-05, "loss": 0.0324, "step": 34500 }, { "grad_norm": 0.3764382302761078, "learning_rate": 6.523643459010966e-05, "loss": 0.0331, "step": 34510 }, { "grad_norm": 0.40570706129074097, "learning_rate": 6.521674792060317e-05, "loss": 0.0276, "step": 34520 }, { "grad_norm": 0.40729087591171265, "learning_rate": 6.519705865097063e-05, "loss": 0.0391, "step": 34530 }, { "grad_norm": 0.34416884183883667, "learning_rate": 6.517736678457641e-05, "loss": 0.0269, "step": 34540 }, { "grad_norm": 0.2959001064300537, "learning_rate": 6.515767232478534e-05, "loss": 0.029, "step": 34550 }, { "grad_norm": 0.4066186249256134, "learning_rate": 6.51379752749626e-05, "loss": 0.0311, "step": 34560 }, { "grad_norm": 0.3216560184955597, "learning_rate": 6.511827563847393e-05, "loss": 0.0317, "step": 34570 }, { "grad_norm": 0.2930659055709839, "learning_rate": 6.509857341868542e-05, "loss": 0.0337, "step": 34580 }, { "grad_norm": 0.3011499345302582, "learning_rate": 6.507886861896367e-05, "loss": 0.029, "step": 34590 }, { "grad_norm": 0.45084530115127563, "learning_rate": 6.505916124267567e-05, "loss": 0.0435, "step": 34600 }, { "grad_norm": 0.3718794584274292, "learning_rate": 6.503945129318891e-05, "loss": 0.0333, "step": 34610 }, { "grad_norm": 0.4349821209907532, "learning_rate": 6.501973877387122e-05, "loss": 0.0279, "step": 34620 }, { "grad_norm": 0.4114886522293091, "learning_rate": 6.500002368809098e-05, "loss": 0.0295, "step": 34630 }, { "grad_norm": 0.40220266580581665, "learning_rate": 6.498030603921694e-05, "loss": 0.0346, "step": 34640 }, { "grad_norm": 0.39880645275115967, "learning_rate": 6.496058583061832e-05, "loss": 0.0347, "step": 34650 }, { "grad_norm": 0.29983121156692505, "learning_rate": 6.494086306566475e-05, "loss": 0.031, "step": 34660 }, { "grad_norm": 0.43941935896873474, "learning_rate": 6.492113774772632e-05, "loss": 0.0301, "step": 34670 }, { "grad_norm": 0.29904016852378845, "learning_rate": 6.490140988017354e-05, "loss": 0.0328, "step": 34680 }, { "grad_norm": 0.3821343183517456, "learning_rate": 6.488167946637736e-05, "loss": 0.0325, "step": 34690 }, { "grad_norm": 0.3078303635120392, "learning_rate": 6.486194650970915e-05, "loss": 0.0364, "step": 34700 }, { "grad_norm": 0.3488632142543793, "learning_rate": 6.48422110135408e-05, "loss": 0.0371, "step": 34710 }, { "grad_norm": 0.39931172132492065, "learning_rate": 6.482247298124451e-05, "loss": 0.0444, "step": 34720 }, { "grad_norm": 0.3984919786453247, "learning_rate": 6.480273241619297e-05, "loss": 0.0292, "step": 34730 }, { "grad_norm": 0.35439926385879517, "learning_rate": 6.478298932175933e-05, "loss": 0.0378, "step": 34740 }, { "grad_norm": 0.3603019118309021, "learning_rate": 6.476324370131712e-05, "loss": 0.0341, "step": 34750 }, { "grad_norm": 0.4267524778842926, "learning_rate": 6.474349555824036e-05, "loss": 0.0369, "step": 34760 }, { "grad_norm": 0.37786927819252014, "learning_rate": 6.472374489590342e-05, "loss": 0.0285, "step": 34770 }, { "grad_norm": 0.23539698123931885, "learning_rate": 6.470399171768118e-05, "loss": 0.0422, "step": 34780 }, { "grad_norm": 0.36325305700302124, "learning_rate": 6.468423602694891e-05, "loss": 0.0345, "step": 34790 }, { "grad_norm": 0.4816164970397949, "learning_rate": 6.466447782708232e-05, "loss": 0.0286, "step": 34800 }, { "grad_norm": 0.24106675386428833, "learning_rate": 6.464471712145754e-05, "loss": 0.0268, "step": 34810 }, { "grad_norm": 0.30489885807037354, "learning_rate": 6.462495391345114e-05, "loss": 0.0312, "step": 34820 }, { "grad_norm": 0.2859043478965759, "learning_rate": 6.46051882064401e-05, "loss": 0.0272, "step": 34830 }, { "grad_norm": 0.4147610366344452, "learning_rate": 6.458542000380186e-05, "loss": 0.0266, "step": 34840 }, { "grad_norm": 0.27081337571144104, "learning_rate": 6.456564930891424e-05, "loss": 0.0264, "step": 34850 }, { "grad_norm": 0.40630847215652466, "learning_rate": 6.454587612515555e-05, "loss": 0.0334, "step": 34860 }, { "grad_norm": 0.3485332131385803, "learning_rate": 6.452610045590444e-05, "loss": 0.0346, "step": 34870 }, { "grad_norm": 0.25861525535583496, "learning_rate": 6.450632230454005e-05, "loss": 0.0291, "step": 34880 }, { "grad_norm": 0.39389896392822266, "learning_rate": 6.448654167444195e-05, "loss": 0.0369, "step": 34890 }, { "grad_norm": 0.39950457215309143, "learning_rate": 6.446675856899005e-05, "loss": 0.032, "step": 34900 }, { "grad_norm": 0.3213989734649658, "learning_rate": 6.444697299156481e-05, "loss": 0.0283, "step": 34910 }, { "grad_norm": 0.3581341803073883, "learning_rate": 6.442718494554701e-05, "loss": 0.0288, "step": 34920 }, { "grad_norm": 0.30938857793807983, "learning_rate": 6.440739443431787e-05, "loss": 0.0301, "step": 34930 }, { "grad_norm": 0.3765062987804413, "learning_rate": 6.438760146125906e-05, "loss": 0.03, "step": 34940 }, { "grad_norm": 0.30646762251853943, "learning_rate": 6.436780602975267e-05, "loss": 0.0331, "step": 34950 }, { "grad_norm": 0.31834876537323, "learning_rate": 6.43480081431812e-05, "loss": 0.0304, "step": 34960 }, { "grad_norm": 0.354956716299057, "learning_rate": 6.432820780492756e-05, "loss": 0.0335, "step": 34970 }, { "grad_norm": 0.19084897637367249, "learning_rate": 6.430840501837506e-05, "loss": 0.0374, "step": 34980 }, { "grad_norm": 0.41553306579589844, "learning_rate": 6.428859978690748e-05, "loss": 0.0273, "step": 34990 }, { "grad_norm": 0.45410576462745667, "learning_rate": 6.426879211390901e-05, "loss": 0.0276, "step": 35000 }, { "grad_norm": 0.38153547048568726, "learning_rate": 6.424898200276422e-05, "loss": 0.0333, "step": 35010 }, { "grad_norm": 0.2956709861755371, "learning_rate": 6.42291694568581e-05, "loss": 0.0289, "step": 35020 }, { "grad_norm": 0.3565311133861542, "learning_rate": 6.42093544795761e-05, "loss": 0.0363, "step": 35030 }, { "grad_norm": 0.24428191781044006, "learning_rate": 6.418953707430403e-05, "loss": 0.0358, "step": 35040 }, { "grad_norm": 0.37385332584381104, "learning_rate": 6.416971724442819e-05, "loss": 0.0326, "step": 35050 }, { "grad_norm": 0.2259884625673294, "learning_rate": 6.414989499333519e-05, "loss": 0.0295, "step": 35060 }, { "grad_norm": 0.498883455991745, "learning_rate": 6.413007032441214e-05, "loss": 0.0329, "step": 35070 }, { "grad_norm": 0.5384232997894287, "learning_rate": 6.411024324104653e-05, "loss": 0.0378, "step": 35080 }, { "grad_norm": 0.37817156314849854, "learning_rate": 6.409041374662628e-05, "loss": 0.0286, "step": 35090 }, { "grad_norm": 0.3952972888946533, "learning_rate": 6.407058184453967e-05, "loss": 0.041, "step": 35100 }, { "grad_norm": 0.33034029603004456, "learning_rate": 6.405074753817548e-05, "loss": 0.0308, "step": 35110 }, { "grad_norm": 0.3890833556652069, "learning_rate": 6.40309108309228e-05, "loss": 0.0294, "step": 35120 }, { "grad_norm": 0.47982022166252136, "learning_rate": 6.401107172617122e-05, "loss": 0.0395, "step": 35130 }, { "grad_norm": 0.3092453181743622, "learning_rate": 6.399123022731068e-05, "loss": 0.0442, "step": 35140 }, { "grad_norm": 0.3122251629829407, "learning_rate": 6.397138633773157e-05, "loss": 0.0281, "step": 35150 }, { "grad_norm": 0.33381378650665283, "learning_rate": 6.395154006082463e-05, "loss": 0.0293, "step": 35160 }, { "grad_norm": 0.3462754487991333, "learning_rate": 6.393169139998109e-05, "loss": 0.0258, "step": 35170 }, { "grad_norm": 0.23761755228042603, "learning_rate": 6.39118403585925e-05, "loss": 0.0328, "step": 35180 }, { "grad_norm": 0.4119192957878113, "learning_rate": 6.38919869400509e-05, "loss": 0.0315, "step": 35190 }, { "grad_norm": 0.3333044648170471, "learning_rate": 6.387213114774865e-05, "loss": 0.0301, "step": 35200 }, { "grad_norm": 0.3768611252307892, "learning_rate": 6.385227298507863e-05, "loss": 0.0505, "step": 35210 }, { "grad_norm": 0.3467606008052826, "learning_rate": 6.3832412455434e-05, "loss": 0.0239, "step": 35220 }, { "grad_norm": 0.34310516715049744, "learning_rate": 6.381254956220841e-05, "loss": 0.0279, "step": 35230 }, { "grad_norm": 0.20647920668125153, "learning_rate": 6.379268430879586e-05, "loss": 0.0283, "step": 35240 }, { "grad_norm": 0.29533815383911133, "learning_rate": 6.37728166985908e-05, "loss": 0.0295, "step": 35250 }, { "grad_norm": 0.33052942156791687, "learning_rate": 6.375294673498804e-05, "loss": 0.0258, "step": 35260 }, { "grad_norm": 0.46348336338996887, "learning_rate": 6.373307442138284e-05, "loss": 0.0338, "step": 35270 }, { "grad_norm": 0.23805145919322968, "learning_rate": 6.371319976117081e-05, "loss": 0.0447, "step": 35280 }, { "grad_norm": 0.41405943036079407, "learning_rate": 6.3693322757748e-05, "loss": 0.0257, "step": 35290 }, { "grad_norm": 0.4260944724082947, "learning_rate": 6.367344341451086e-05, "loss": 0.0302, "step": 35300 }, { "grad_norm": 0.3387469947338104, "learning_rate": 6.36535617348562e-05, "loss": 0.0302, "step": 35310 }, { "grad_norm": 0.38218453526496887, "learning_rate": 6.363367772218128e-05, "loss": 0.0261, "step": 35320 }, { "grad_norm": 0.31038331985473633, "learning_rate": 6.36137913798837e-05, "loss": 0.032, "step": 35330 }, { "grad_norm": 0.33271950483322144, "learning_rate": 6.359390271136151e-05, "loss": 0.0225, "step": 35340 }, { "grad_norm": 0.32816970348358154, "learning_rate": 6.357401172001314e-05, "loss": 0.0299, "step": 35350 }, { "grad_norm": 0.30582883954048157, "learning_rate": 6.355411840923742e-05, "loss": 0.0302, "step": 35360 }, { "grad_norm": 0.40859469771385193, "learning_rate": 6.353422278243358e-05, "loss": 0.0316, "step": 35370 }, { "grad_norm": 0.38067370653152466, "learning_rate": 6.351432484300121e-05, "loss": 0.032, "step": 35380 }, { "grad_norm": 0.35437989234924316, "learning_rate": 6.349442459434036e-05, "loss": 0.0327, "step": 35390 }, { "grad_norm": 0.3385222256183624, "learning_rate": 6.34745220398514e-05, "loss": 0.03, "step": 35400 }, { "grad_norm": 0.2885946035385132, "learning_rate": 6.345461718293518e-05, "loss": 0.0299, "step": 35410 }, { "grad_norm": 0.3354690670967102, "learning_rate": 6.343471002699286e-05, "loss": 0.029, "step": 35420 }, { "grad_norm": 0.38339513540267944, "learning_rate": 6.341480057542602e-05, "loss": 0.0344, "step": 35430 }, { "grad_norm": 0.2625017762184143, "learning_rate": 6.339488883163667e-05, "loss": 0.031, "step": 35440 }, { "grad_norm": 0.2859392762184143, "learning_rate": 6.337497479902716e-05, "loss": 0.0277, "step": 35450 }, { "grad_norm": 0.29913344979286194, "learning_rate": 6.335505848100027e-05, "loss": 0.0289, "step": 35460 }, { "grad_norm": 0.3004017770290375, "learning_rate": 6.333513988095915e-05, "loss": 0.0251, "step": 35470 }, { "grad_norm": 0.2505719065666199, "learning_rate": 6.331521900230735e-05, "loss": 0.0283, "step": 35480 }, { "grad_norm": 0.20506446063518524, "learning_rate": 6.329529584844878e-05, "loss": 0.0325, "step": 35490 }, { "grad_norm": 0.47782838344573975, "learning_rate": 6.327537042278777e-05, "loss": 0.0313, "step": 35500 }, { "grad_norm": 0.34496235847473145, "learning_rate": 6.325544272872905e-05, "loss": 0.0305, "step": 35510 }, { "grad_norm": 0.2707523703575134, "learning_rate": 6.323551276967771e-05, "loss": 0.0347, "step": 35520 }, { "grad_norm": 0.29279160499572754, "learning_rate": 6.321558054903922e-05, "loss": 0.0266, "step": 35530 }, { "grad_norm": 0.25805607438087463, "learning_rate": 6.319564607021947e-05, "loss": 0.0305, "step": 35540 }, { "grad_norm": 0.3026716411113739, "learning_rate": 6.31757093366247e-05, "loss": 0.0301, "step": 35550 }, { "grad_norm": 0.4499499201774597, "learning_rate": 6.315577035166154e-05, "loss": 0.0252, "step": 35560 }, { "grad_norm": 0.37255752086639404, "learning_rate": 6.313582911873708e-05, "loss": 0.0296, "step": 35570 }, { "grad_norm": 0.33941084146499634, "learning_rate": 6.311588564125865e-05, "loss": 0.0265, "step": 35580 }, { "grad_norm": 0.36209750175476074, "learning_rate": 6.30959399226341e-05, "loss": 0.0319, "step": 35590 }, { "grad_norm": 0.26244330406188965, "learning_rate": 6.30759919662716e-05, "loss": 0.0425, "step": 35600 }, { "grad_norm": 0.2232251614332199, "learning_rate": 6.30560417755797e-05, "loss": 0.0276, "step": 35610 }, { "grad_norm": 0.22750307619571686, "learning_rate": 6.303608935396735e-05, "loss": 0.0222, "step": 35620 }, { "grad_norm": 0.30780845880508423, "learning_rate": 6.301613470484386e-05, "loss": 0.0259, "step": 35630 }, { "grad_norm": 0.3843134939670563, "learning_rate": 6.299617783161893e-05, "loss": 0.0353, "step": 35640 }, { "grad_norm": 0.2543470859527588, "learning_rate": 6.297621873770266e-05, "loss": 0.0286, "step": 35650 }, { "grad_norm": 0.27408987283706665, "learning_rate": 6.29562574265055e-05, "loss": 0.0341, "step": 35660 }, { "grad_norm": 0.37824103236198425, "learning_rate": 6.293629390143834e-05, "loss": 0.0282, "step": 35670 }, { "grad_norm": 0.3792819380760193, "learning_rate": 6.291632816591232e-05, "loss": 0.028, "step": 35680 }, { "grad_norm": 0.318488746881485, "learning_rate": 6.28963602233391e-05, "loss": 0.0271, "step": 35690 }, { "grad_norm": 0.3470093607902527, "learning_rate": 6.287639007713062e-05, "loss": 0.03, "step": 35700 }, { "grad_norm": 0.2083543986082077, "learning_rate": 6.285641773069926e-05, "loss": 0.0284, "step": 35710 }, { "grad_norm": 0.25223836302757263, "learning_rate": 6.283644318745773e-05, "loss": 0.0238, "step": 35720 }, { "grad_norm": 0.25446054339408875, "learning_rate": 6.281646645081912e-05, "loss": 0.0298, "step": 35730 }, { "grad_norm": 0.3603477478027344, "learning_rate": 6.279648752419693e-05, "loss": 0.0276, "step": 35740 }, { "grad_norm": 0.2951990067958832, "learning_rate": 6.2776506411005e-05, "loss": 0.0337, "step": 35750 }, { "grad_norm": 0.27934175729751587, "learning_rate": 6.275652311465758e-05, "loss": 0.0304, "step": 35760 }, { "grad_norm": 0.2649315297603607, "learning_rate": 6.273653763856926e-05, "loss": 0.0319, "step": 35770 }, { "grad_norm": 0.4377989172935486, "learning_rate": 6.271654998615501e-05, "loss": 0.0368, "step": 35780 }, { "grad_norm": 0.418061226606369, "learning_rate": 6.269656016083013e-05, "loss": 0.0313, "step": 35790 }, { "grad_norm": 0.5763787031173706, "learning_rate": 6.267656816601038e-05, "loss": 0.0387, "step": 35800 }, { "grad_norm": 0.33447888493537903, "learning_rate": 6.265657400511185e-05, "loss": 0.0368, "step": 35810 }, { "grad_norm": 0.38612571358680725, "learning_rate": 6.263657768155098e-05, "loss": 0.031, "step": 35820 }, { "grad_norm": 0.3644116520881653, "learning_rate": 6.261657919874457e-05, "loss": 0.0301, "step": 35830 }, { "grad_norm": 0.30849388241767883, "learning_rate": 6.259657856010986e-05, "loss": 0.0356, "step": 35840 }, { "grad_norm": 0.3424804210662842, "learning_rate": 6.257657576906439e-05, "loss": 0.0302, "step": 35850 }, { "grad_norm": 0.38196516036987305, "learning_rate": 6.255657082902609e-05, "loss": 0.025, "step": 35860 }, { "grad_norm": 0.34331896901130676, "learning_rate": 6.253656374341325e-05, "loss": 0.0243, "step": 35870 }, { "grad_norm": 0.39463022351264954, "learning_rate": 6.251655451564457e-05, "loss": 0.03, "step": 35880 }, { "grad_norm": 0.301117479801178, "learning_rate": 6.249654314913902e-05, "loss": 0.0306, "step": 35890 }, { "grad_norm": 0.2523041367530823, "learning_rate": 6.247652964731604e-05, "loss": 0.0246, "step": 35900 }, { "grad_norm": 0.3596455454826355, "learning_rate": 6.245651401359537e-05, "loss": 0.0427, "step": 35910 }, { "grad_norm": 0.38726964592933655, "learning_rate": 6.243649625139715e-05, "loss": 0.0362, "step": 35920 }, { "grad_norm": 0.3308843672275543, "learning_rate": 6.241647636414185e-05, "loss": 0.029, "step": 35930 }, { "grad_norm": 0.4218362271785736, "learning_rate": 6.239645435525034e-05, "loss": 0.0395, "step": 35940 }, { "grad_norm": 0.22692479193210602, "learning_rate": 6.237643022814381e-05, "loss": 0.033, "step": 35950 }, { "grad_norm": 0.2858905792236328, "learning_rate": 6.235640398624386e-05, "loss": 0.0273, "step": 35960 }, { "grad_norm": 0.4042072892189026, "learning_rate": 6.233637563297243e-05, "loss": 0.0452, "step": 35970 }, { "grad_norm": 0.37597009539604187, "learning_rate": 6.23163451717518e-05, "loss": 0.0339, "step": 35980 }, { "grad_norm": 0.31130945682525635, "learning_rate": 6.229631260600463e-05, "loss": 0.0327, "step": 35990 }, { "grad_norm": 0.3667357265949249, "learning_rate": 6.227627793915392e-05, "loss": 0.037, "step": 36000 }, { "grad_norm": 0.272391140460968, "learning_rate": 6.225624117462309e-05, "loss": 0.0266, "step": 36010 }, { "grad_norm": 0.3393958508968353, "learning_rate": 6.223620231583586e-05, "loss": 0.0274, "step": 36020 }, { "grad_norm": 0.22076882421970367, "learning_rate": 6.221616136621629e-05, "loss": 0.0238, "step": 36030 }, { "grad_norm": 0.3978841006755829, "learning_rate": 6.219611832918887e-05, "loss": 0.0297, "step": 36040 }, { "grad_norm": 0.267187237739563, "learning_rate": 6.217607320817838e-05, "loss": 0.0264, "step": 36050 }, { "grad_norm": 0.44387009739875793, "learning_rate": 6.215602600661001e-05, "loss": 0.0289, "step": 36060 }, { "grad_norm": 0.3652515411376953, "learning_rate": 6.213597672790925e-05, "loss": 0.0286, "step": 36070 }, { "grad_norm": 0.2558199465274811, "learning_rate": 6.2115925375502e-05, "loss": 0.0326, "step": 36080 }, { "grad_norm": 0.234083354473114, "learning_rate": 6.209587195281447e-05, "loss": 0.032, "step": 36090 }, { "grad_norm": 0.35305044054985046, "learning_rate": 6.207581646327324e-05, "loss": 0.0352, "step": 36100 }, { "grad_norm": 0.23192167282104492, "learning_rate": 6.205575891030526e-05, "loss": 0.0215, "step": 36110 }, { "grad_norm": 0.28234440088272095, "learning_rate": 6.203569929733781e-05, "loss": 0.0238, "step": 36120 }, { "grad_norm": 0.4871624708175659, "learning_rate": 6.201563762779852e-05, "loss": 0.0295, "step": 36130 }, { "grad_norm": 0.23497548699378967, "learning_rate": 6.199557390511538e-05, "loss": 0.0276, "step": 36140 }, { "grad_norm": 0.36319833993911743, "learning_rate": 6.197550813271675e-05, "loss": 0.0289, "step": 36150 }, { "grad_norm": 0.32072925567626953, "learning_rate": 6.195544031403131e-05, "loss": 0.0291, "step": 36160 }, { "grad_norm": 0.30324169993400574, "learning_rate": 6.19353704524881e-05, "loss": 0.0259, "step": 36170 }, { "grad_norm": 0.2784166932106018, "learning_rate": 6.191529855151652e-05, "loss": 0.0258, "step": 36180 }, { "grad_norm": 0.27476832270622253, "learning_rate": 6.189522461454629e-05, "loss": 0.0305, "step": 36190 }, { "grad_norm": 0.36195966601371765, "learning_rate": 6.187514864500752e-05, "loss": 0.0291, "step": 36200 }, { "grad_norm": 0.35367605090141296, "learning_rate": 6.185507064633062e-05, "loss": 0.0312, "step": 36210 }, { "grad_norm": 0.41222846508026123, "learning_rate": 6.18349906219464e-05, "loss": 0.0366, "step": 36220 }, { "grad_norm": 0.4521612226963043, "learning_rate": 6.181490857528596e-05, "loss": 0.0308, "step": 36230 }, { "grad_norm": 0.2261437028646469, "learning_rate": 6.179482450978077e-05, "loss": 0.0217, "step": 36240 }, { "grad_norm": 0.2419871836900711, "learning_rate": 6.177473842886269e-05, "loss": 0.0277, "step": 36250 }, { "grad_norm": 0.25027406215667725, "learning_rate": 6.175465033596382e-05, "loss": 0.0337, "step": 36260 }, { "grad_norm": 0.3627731204032898, "learning_rate": 6.173456023451671e-05, "loss": 0.032, "step": 36270 }, { "grad_norm": 0.46260684728622437, "learning_rate": 6.171446812795422e-05, "loss": 0.0381, "step": 36280 }, { "grad_norm": 0.316913902759552, "learning_rate": 6.169437401970949e-05, "loss": 0.0438, "step": 36290 }, { "grad_norm": 0.2208050787448883, "learning_rate": 6.16742779132161e-05, "loss": 0.0199, "step": 36300 }, { "grad_norm": 0.27062809467315674, "learning_rate": 6.165417981190789e-05, "loss": 0.0248, "step": 36310 }, { "grad_norm": 0.3509986996650696, "learning_rate": 6.16340797192191e-05, "loss": 0.026, "step": 36320 }, { "grad_norm": 0.30186283588409424, "learning_rate": 6.161397763858427e-05, "loss": 0.0288, "step": 36330 }, { "grad_norm": 0.3009902834892273, "learning_rate": 6.159387357343834e-05, "loss": 0.0247, "step": 36340 }, { "grad_norm": 0.36380574107170105, "learning_rate": 6.157376752721648e-05, "loss": 0.0344, "step": 36350 }, { "grad_norm": 0.23550072312355042, "learning_rate": 6.155365950335428e-05, "loss": 0.0233, "step": 36360 }, { "grad_norm": 0.5785002708435059, "learning_rate": 6.153354950528768e-05, "loss": 0.0237, "step": 36370 }, { "grad_norm": 0.396829217672348, "learning_rate": 6.151343753645293e-05, "loss": 0.0306, "step": 36380 }, { "grad_norm": 0.3860087990760803, "learning_rate": 6.149332360028657e-05, "loss": 0.0347, "step": 36390 }, { "grad_norm": 0.35231655836105347, "learning_rate": 6.147320770022555e-05, "loss": 0.029, "step": 36400 }, { "grad_norm": 0.325264573097229, "learning_rate": 6.145308983970715e-05, "loss": 0.0357, "step": 36410 }, { "grad_norm": 0.45381397008895874, "learning_rate": 6.143297002216892e-05, "loss": 0.0258, "step": 36420 }, { "grad_norm": 0.3658424913883209, "learning_rate": 6.141284825104882e-05, "loss": 0.0317, "step": 36430 }, { "grad_norm": 0.16283011436462402, "learning_rate": 6.13927245297851e-05, "loss": 0.0391, "step": 36440 }, { "grad_norm": 0.26432004570961, "learning_rate": 6.137259886181633e-05, "loss": 0.0311, "step": 36450 }, { "grad_norm": 0.327710896730423, "learning_rate": 6.135247125058145e-05, "loss": 0.0383, "step": 36460 }, { "grad_norm": 0.18290583789348602, "learning_rate": 6.133234169951974e-05, "loss": 0.0248, "step": 36470 }, { "grad_norm": 0.4640633463859558, "learning_rate": 6.131221021207078e-05, "loss": 0.0388, "step": 36480 }, { "grad_norm": 0.22078892588615417, "learning_rate": 6.129207679167448e-05, "loss": 0.0235, "step": 36490 }, { "grad_norm": 0.23080085217952728, "learning_rate": 6.127194144177109e-05, "loss": 0.0279, "step": 36500 }, { "grad_norm": 0.2914152145385742, "learning_rate": 6.125180416580118e-05, "loss": 0.0278, "step": 36510 }, { "grad_norm": 0.36813947558403015, "learning_rate": 6.123166496720571e-05, "loss": 0.0241, "step": 36520 }, { "grad_norm": 0.3742811977863312, "learning_rate": 6.121152384942588e-05, "loss": 0.0305, "step": 36530 }, { "grad_norm": 0.4253886044025421, "learning_rate": 6.119138081590324e-05, "loss": 0.0268, "step": 36540 }, { "grad_norm": 0.5228016972541809, "learning_rate": 6.117123587007971e-05, "loss": 0.0303, "step": 36550 }, { "grad_norm": 0.2832942008972168, "learning_rate": 6.11510890153975e-05, "loss": 0.0336, "step": 36560 }, { "grad_norm": 0.3723990321159363, "learning_rate": 6.113094025529916e-05, "loss": 0.0234, "step": 36570 }, { "grad_norm": 0.43385905027389526, "learning_rate": 6.111078959322757e-05, "loss": 0.0399, "step": 36580 }, { "grad_norm": 0.33770185708999634, "learning_rate": 6.109063703262592e-05, "loss": 0.0324, "step": 36590 }, { "grad_norm": 0.39058446884155273, "learning_rate": 6.107048257693772e-05, "loss": 0.0321, "step": 36600 }, { "grad_norm": 0.3270834982395172, "learning_rate": 6.105032622960683e-05, "loss": 0.0276, "step": 36610 }, { "grad_norm": 0.3812582194805145, "learning_rate": 6.103016799407743e-05, "loss": 0.029, "step": 36620 }, { "grad_norm": 0.2493305206298828, "learning_rate": 6.1010007873793984e-05, "loss": 0.0276, "step": 36630 }, { "grad_norm": 0.3013969361782074, "learning_rate": 6.098984587220131e-05, "loss": 0.0252, "step": 36640 }, { "grad_norm": 0.3089326322078705, "learning_rate": 6.096968199274456e-05, "loss": 0.0291, "step": 36650 }, { "grad_norm": 0.2996092438697815, "learning_rate": 6.0949516238869166e-05, "loss": 0.0253, "step": 36660 }, { "grad_norm": 0.3877412676811218, "learning_rate": 6.092934861402092e-05, "loss": 0.0333, "step": 36670 }, { "grad_norm": 0.3571794927120209, "learning_rate": 6.0909179121645924e-05, "loss": 0.0409, "step": 36680 }, { "grad_norm": 0.4096467196941376, "learning_rate": 6.0889007765190576e-05, "loss": 0.0319, "step": 36690 }, { "grad_norm": 0.21174763143062592, "learning_rate": 6.086883454810162e-05, "loss": 0.0253, "step": 36700 }, { "grad_norm": 0.3079909682273865, "learning_rate": 6.0848659473826084e-05, "loss": 0.0357, "step": 36710 }, { "grad_norm": 0.5205111503601074, "learning_rate": 6.082848254581138e-05, "loss": 0.0319, "step": 36720 }, { "grad_norm": 0.4216163456439972, "learning_rate": 6.080830376750517e-05, "loss": 0.029, "step": 36730 }, { "grad_norm": 0.2117813229560852, "learning_rate": 6.0788123142355445e-05, "loss": 0.0231, "step": 36740 }, { "grad_norm": 0.38924679160118103, "learning_rate": 6.076794067381052e-05, "loss": 0.0272, "step": 36750 }, { "grad_norm": 0.3168506920337677, "learning_rate": 6.074775636531905e-05, "loss": 0.0308, "step": 36760 }, { "grad_norm": 0.41020333766937256, "learning_rate": 6.072757022032997e-05, "loss": 0.0322, "step": 36770 }, { "grad_norm": 0.27145567536354065, "learning_rate": 6.070738224229253e-05, "loss": 0.0415, "step": 36780 }, { "grad_norm": 0.2782915532588959, "learning_rate": 6.0687192434656314e-05, "loss": 0.0283, "step": 36790 }, { "grad_norm": 0.2758910357952118, "learning_rate": 6.066700080087121e-05, "loss": 0.0284, "step": 36800 }, { "grad_norm": 0.45325055718421936, "learning_rate": 6.0646807344387424e-05, "loss": 0.0276, "step": 36810 }, { "grad_norm": 0.4146229922771454, "learning_rate": 6.062661206865543e-05, "loss": 0.0355, "step": 36820 }, { "grad_norm": 0.36530208587646484, "learning_rate": 6.06064149771261e-05, "loss": 0.0327, "step": 36830 }, { "grad_norm": 0.3250839412212372, "learning_rate": 6.058621607325051e-05, "loss": 0.025, "step": 36840 }, { "grad_norm": 0.47416818141937256, "learning_rate": 6.056601536048014e-05, "loss": 0.0379, "step": 36850 }, { "grad_norm": 0.39684659242630005, "learning_rate": 6.0545812842266725e-05, "loss": 0.0407, "step": 36860 }, { "grad_norm": 0.41087692975997925, "learning_rate": 6.052560852206232e-05, "loss": 0.0294, "step": 36870 }, { "grad_norm": 0.26388126611709595, "learning_rate": 6.05054024033193e-05, "loss": 0.0324, "step": 36880 }, { "grad_norm": 0.37005728483200073, "learning_rate": 6.048519448949032e-05, "loss": 0.0273, "step": 36890 }, { "grad_norm": 0.2569239139556885, "learning_rate": 6.046498478402839e-05, "loss": 0.0329, "step": 36900 }, { "grad_norm": 0.3469601273536682, "learning_rate": 6.044477329038677e-05, "loss": 0.0285, "step": 36910 }, { "grad_norm": 0.3977715075016022, "learning_rate": 6.042456001201906e-05, "loss": 0.03, "step": 36920 }, { "grad_norm": 0.2864101231098175, "learning_rate": 6.040434495237917e-05, "loss": 0.0326, "step": 36930 }, { "grad_norm": 0.5013142228126526, "learning_rate": 6.0384128114921256e-05, "loss": 0.0294, "step": 36940 }, { "grad_norm": 0.32638686895370483, "learning_rate": 6.036390950309987e-05, "loss": 0.02, "step": 36950 }, { "grad_norm": 0.25990909337997437, "learning_rate": 6.0343689120369805e-05, "loss": 0.0293, "step": 36960 }, { "grad_norm": 0.3432592749595642, "learning_rate": 6.032346697018616e-05, "loss": 0.0251, "step": 36970 }, { "grad_norm": 0.4071638882160187, "learning_rate": 6.0303243056004375e-05, "loss": 0.0429, "step": 36980 }, { "grad_norm": 0.29994189739227295, "learning_rate": 6.0283017381280136e-05, "loss": 0.0258, "step": 36990 }, { "grad_norm": 0.34714117646217346, "learning_rate": 6.026278994946948e-05, "loss": 0.0283, "step": 37000 }, { "grad_norm": 0.18971660733222961, "learning_rate": 6.02425607640287e-05, "loss": 0.0291, "step": 37010 }, { "grad_norm": 0.34647929668426514, "learning_rate": 6.022232982841441e-05, "loss": 0.0324, "step": 37020 }, { "grad_norm": 0.2823461890220642, "learning_rate": 6.020209714608355e-05, "loss": 0.0292, "step": 37030 }, { "grad_norm": 0.35393279790878296, "learning_rate": 6.018186272049331e-05, "loss": 0.0282, "step": 37040 }, { "grad_norm": 0.25276288390159607, "learning_rate": 6.01616265551012e-05, "loss": 0.0354, "step": 37050 }, { "grad_norm": 0.32464489340782166, "learning_rate": 6.014138865336503e-05, "loss": 0.0285, "step": 37060 }, { "grad_norm": 0.2943052053451538, "learning_rate": 6.0121149018742905e-05, "loss": 0.0326, "step": 37070 }, { "grad_norm": 0.2607434093952179, "learning_rate": 6.010090765469325e-05, "loss": 0.0253, "step": 37080 }, { "grad_norm": 0.42675694823265076, "learning_rate": 6.008066456467473e-05, "loss": 0.0292, "step": 37090 }, { "grad_norm": 0.3586258888244629, "learning_rate": 6.0060419752146335e-05, "loss": 0.0295, "step": 37100 }, { "grad_norm": 0.37444794178009033, "learning_rate": 6.0040173220567353e-05, "loss": 0.031, "step": 37110 }, { "grad_norm": 0.2702488303184509, "learning_rate": 6.001992497339737e-05, "loss": 0.0277, "step": 37120 }, { "grad_norm": 0.2261854112148285, "learning_rate": 5.999967501409626e-05, "loss": 0.0321, "step": 37130 }, { "grad_norm": 0.3645779490470886, "learning_rate": 5.997942334612418e-05, "loss": 0.0267, "step": 37140 }, { "grad_norm": 0.343575656414032, "learning_rate": 5.995916997294158e-05, "loss": 0.0254, "step": 37150 }, { "grad_norm": 0.3030606508255005, "learning_rate": 5.9938914898009214e-05, "loss": 0.0284, "step": 37160 }, { "grad_norm": 0.35497429966926575, "learning_rate": 5.991865812478813e-05, "loss": 0.0223, "step": 37170 }, { "grad_norm": 0.3474418520927429, "learning_rate": 5.989839965673964e-05, "loss": 0.0419, "step": 37180 }, { "grad_norm": 0.2981249690055847, "learning_rate": 5.987813949732539e-05, "loss": 0.032, "step": 37190 }, { "grad_norm": 0.3991018235683441, "learning_rate": 5.9857877650007255e-05, "loss": 0.0301, "step": 37200 }, { "grad_norm": 0.2873897850513458, "learning_rate": 5.983761411824744e-05, "loss": 0.0251, "step": 37210 }, { "grad_norm": 0.40076911449432373, "learning_rate": 5.981734890550844e-05, "loss": 0.034, "step": 37220 }, { "grad_norm": 0.27578094601631165, "learning_rate": 5.979708201525301e-05, "loss": 0.0381, "step": 37230 }, { "grad_norm": 0.43257343769073486, "learning_rate": 5.977681345094422e-05, "loss": 0.0276, "step": 37240 }, { "grad_norm": 0.3325927257537842, "learning_rate": 5.97565432160454e-05, "loss": 0.0285, "step": 37250 }, { "grad_norm": 0.30094683170318604, "learning_rate": 5.9736271314020186e-05, "loss": 0.0344, "step": 37260 }, { "grad_norm": 0.3384847640991211, "learning_rate": 5.971599774833251e-05, "loss": 0.0305, "step": 37270 }, { "grad_norm": 0.24770879745483398, "learning_rate": 5.9695722522446525e-05, "loss": 0.0315, "step": 37280 }, { "grad_norm": 0.3690321743488312, "learning_rate": 5.9675445639826765e-05, "loss": 0.0309, "step": 37290 }, { "grad_norm": 0.2480725795030594, "learning_rate": 5.965516710393796e-05, "loss": 0.0244, "step": 37300 }, { "grad_norm": 0.37007641792297363, "learning_rate": 5.963488691824516e-05, "loss": 0.0359, "step": 37310 }, { "grad_norm": 0.35824158787727356, "learning_rate": 5.96146050862137e-05, "loss": 0.0398, "step": 37320 }, { "grad_norm": 0.3807729482650757, "learning_rate": 5.959432161130919e-05, "loss": 0.0269, "step": 37330 }, { "grad_norm": 0.4524393379688263, "learning_rate": 5.9574036496997545e-05, "loss": 0.0325, "step": 37340 }, { "grad_norm": 0.40004608035087585, "learning_rate": 5.955374974674488e-05, "loss": 0.0336, "step": 37350 }, { "grad_norm": 0.20567241311073303, "learning_rate": 5.9533461364017696e-05, "loss": 0.0281, "step": 37360 }, { "grad_norm": 0.38064903020858765, "learning_rate": 5.9513171352282716e-05, "loss": 0.03, "step": 37370 }, { "grad_norm": 0.33080533146858215, "learning_rate": 5.949287971500692e-05, "loss": 0.0249, "step": 37380 }, { "grad_norm": 0.2151990532875061, "learning_rate": 5.947258645565762e-05, "loss": 0.0292, "step": 37390 }, { "grad_norm": 0.2943710684776306, "learning_rate": 5.945229157770237e-05, "loss": 0.0274, "step": 37400 }, { "grad_norm": 0.3838779032230377, "learning_rate": 5.9431995084609006e-05, "loss": 0.0296, "step": 37410 }, { "grad_norm": 0.37061643600463867, "learning_rate": 5.941169697984564e-05, "loss": 0.0279, "step": 37420 }, { "grad_norm": 0.5125941634178162, "learning_rate": 5.9391397266880675e-05, "loss": 0.0307, "step": 37430 }, { "grad_norm": 0.3324512541294098, "learning_rate": 5.937109594918279e-05, "loss": 0.0402, "step": 37440 }, { "grad_norm": 0.30351322889328003, "learning_rate": 5.9350793030220884e-05, "loss": 0.0364, "step": 37450 }, { "grad_norm": 0.33103886246681213, "learning_rate": 5.933048851346421e-05, "loss": 0.0263, "step": 37460 }, { "grad_norm": 0.35483765602111816, "learning_rate": 5.931018240238224e-05, "loss": 0.0487, "step": 37470 }, { "grad_norm": 0.29057765007019043, "learning_rate": 5.928987470044471e-05, "loss": 0.0248, "step": 37480 }, { "grad_norm": 0.42389991879463196, "learning_rate": 5.9269565411121695e-05, "loss": 0.0301, "step": 37490 }, { "grad_norm": 0.34925565123558044, "learning_rate": 5.924925453788347e-05, "loss": 0.0307, "step": 37500 }, { "grad_norm": 0.18408691883087158, "learning_rate": 5.92289420842006e-05, "loss": 0.0312, "step": 37510 }, { "grad_norm": 0.3475601077079773, "learning_rate": 5.9208628053543945e-05, "loss": 0.0392, "step": 37520 }, { "grad_norm": 0.7216134667396545, "learning_rate": 5.918831244938462e-05, "loss": 0.0331, "step": 37530 }, { "grad_norm": 0.37183523178100586, "learning_rate": 5.9167995275194e-05, "loss": 0.029, "step": 37540 }, { "grad_norm": 0.3470090627670288, "learning_rate": 5.914767653444373e-05, "loss": 0.0312, "step": 37550 }, { "grad_norm": 0.3760302662849426, "learning_rate": 5.912735623060572e-05, "loss": 0.0303, "step": 37560 }, { "grad_norm": 0.27488356828689575, "learning_rate": 5.910703436715217e-05, "loss": 0.0306, "step": 37570 }, { "grad_norm": 0.3057427406311035, "learning_rate": 5.908671094755552e-05, "loss": 0.0278, "step": 37580 }, { "grad_norm": 0.30461686849594116, "learning_rate": 5.906638597528851e-05, "loss": 0.0274, "step": 37590 }, { "grad_norm": 0.3560223877429962, "learning_rate": 5.9046059453824076e-05, "loss": 0.0315, "step": 37600 }, { "grad_norm": 0.41276463866233826, "learning_rate": 5.9025731386635505e-05, "loss": 0.0265, "step": 37610 }, { "grad_norm": 0.2848760485649109, "learning_rate": 5.900540177719629e-05, "loss": 0.028, "step": 37620 }, { "grad_norm": 0.3175349235534668, "learning_rate": 5.898507062898021e-05, "loss": 0.0349, "step": 37630 }, { "grad_norm": 0.5526381731033325, "learning_rate": 5.8964737945461316e-05, "loss": 0.0304, "step": 37640 }, { "grad_norm": 0.26491081714630127, "learning_rate": 5.8944403730113885e-05, "loss": 0.0304, "step": 37650 }, { "grad_norm": 0.3739095628261566, "learning_rate": 5.892406798641248e-05, "loss": 0.0395, "step": 37660 }, { "grad_norm": 0.3880654275417328, "learning_rate": 5.890373071783193e-05, "loss": 0.0275, "step": 37670 }, { "grad_norm": 0.3423152267932892, "learning_rate": 5.888339192784732e-05, "loss": 0.0288, "step": 37680 }, { "grad_norm": 0.30671295523643494, "learning_rate": 5.8863051619934003e-05, "loss": 0.0306, "step": 37690 }, { "grad_norm": 0.3088410198688507, "learning_rate": 5.8842709797567554e-05, "loss": 0.0292, "step": 37700 }, { "grad_norm": 0.39664050936698914, "learning_rate": 5.8822366464223855e-05, "loss": 0.0227, "step": 37710 }, { "grad_norm": 0.25691795349121094, "learning_rate": 5.880202162337901e-05, "loss": 0.03, "step": 37720 }, { "grad_norm": 0.3827754557132721, "learning_rate": 5.8781675278509405e-05, "loss": 0.0264, "step": 37730 }, { "grad_norm": 0.24233271181583405, "learning_rate": 5.8761327433091696e-05, "loss": 0.0268, "step": 37740 }, { "grad_norm": 0.5200873613357544, "learning_rate": 5.874097809060275e-05, "loss": 0.0389, "step": 37750 }, { "grad_norm": 0.4294643998146057, "learning_rate": 5.87206272545197e-05, "loss": 0.0409, "step": 37760 }, { "grad_norm": 0.29161399602890015, "learning_rate": 5.8700274928319955e-05, "loss": 0.0348, "step": 37770 }, { "grad_norm": 0.40343454480171204, "learning_rate": 5.867992111548118e-05, "loss": 0.0287, "step": 37780 }, { "grad_norm": 0.4240621328353882, "learning_rate": 5.865956581948131e-05, "loss": 0.0299, "step": 37790 }, { "grad_norm": 0.2346186339855194, "learning_rate": 5.863920904379845e-05, "loss": 0.0258, "step": 37800 }, { "grad_norm": 0.3758551776409149, "learning_rate": 5.8618850791911064e-05, "loss": 0.0327, "step": 37810 }, { "grad_norm": 0.32813653349876404, "learning_rate": 5.859849106729779e-05, "loss": 0.031, "step": 37820 }, { "grad_norm": 0.44888633489608765, "learning_rate": 5.857812987343758e-05, "loss": 0.0305, "step": 37830 }, { "grad_norm": 0.2534797787666321, "learning_rate": 5.855776721380957e-05, "loss": 0.032, "step": 37840 }, { "grad_norm": 0.3633183538913727, "learning_rate": 5.8537403091893217e-05, "loss": 0.0274, "step": 37850 }, { "grad_norm": 0.30427441000938416, "learning_rate": 5.851703751116816e-05, "loss": 0.0322, "step": 37860 }, { "grad_norm": 0.227594256401062, "learning_rate": 5.8496670475114336e-05, "loss": 0.0241, "step": 37870 }, { "grad_norm": 0.379237562417984, "learning_rate": 5.84763019872119e-05, "loss": 0.0401, "step": 37880 }, { "grad_norm": 0.3004677891731262, "learning_rate": 5.845593205094131e-05, "loss": 0.0264, "step": 37890 }, { "grad_norm": 0.36432182788848877, "learning_rate": 5.843556066978318e-05, "loss": 0.0251, "step": 37900 }, { "grad_norm": 0.20327645540237427, "learning_rate": 5.8415187847218455e-05, "loss": 0.0295, "step": 37910 }, { "grad_norm": 0.1852848380804062, "learning_rate": 5.839481358672827e-05, "loss": 0.0251, "step": 37920 }, { "grad_norm": 0.36320972442626953, "learning_rate": 5.837443789179407e-05, "loss": 0.0288, "step": 37930 }, { "grad_norm": 0.2779415547847748, "learning_rate": 5.8354060765897445e-05, "loss": 0.0299, "step": 37940 }, { "grad_norm": 0.2743995487689972, "learning_rate": 5.8333682212520334e-05, "loss": 0.0398, "step": 37950 }, { "grad_norm": 0.38788744807243347, "learning_rate": 5.831330223514486e-05, "loss": 0.0293, "step": 37960 }, { "grad_norm": 0.3318988084793091, "learning_rate": 5.8292920837253396e-05, "loss": 0.0297, "step": 37970 }, { "grad_norm": 0.358808308839798, "learning_rate": 5.827253802232857e-05, "loss": 0.026, "step": 37980 }, { "grad_norm": 0.3885212242603302, "learning_rate": 5.825215379385327e-05, "loss": 0.024, "step": 37990 }, { "grad_norm": 0.29052069783210754, "learning_rate": 5.823176815531057e-05, "loss": 0.0416, "step": 38000 }, { "grad_norm": 0.1875964105129242, "learning_rate": 5.8211381110183826e-05, "loss": 0.0355, "step": 38010 }, { "grad_norm": 0.3943345546722412, "learning_rate": 5.8190992661956645e-05, "loss": 0.0216, "step": 38020 }, { "grad_norm": 0.4633764326572418, "learning_rate": 5.817060281411284e-05, "loss": 0.0382, "step": 38030 }, { "grad_norm": 0.258151113986969, "learning_rate": 5.815021157013647e-05, "loss": 0.0303, "step": 38040 }, { "grad_norm": 0.355292409658432, "learning_rate": 5.8129818933511856e-05, "loss": 0.042, "step": 38050 }, { "grad_norm": 0.28255730867385864, "learning_rate": 5.8109424907723544e-05, "loss": 0.0325, "step": 38060 }, { "grad_norm": 0.24229998886585236, "learning_rate": 5.80890294962563e-05, "loss": 0.0254, "step": 38070 }, { "grad_norm": 0.4450189769268036, "learning_rate": 5.806863270259515e-05, "loss": 0.0383, "step": 38080 }, { "grad_norm": 0.3045405447483063, "learning_rate": 5.804823453022536e-05, "loss": 0.0295, "step": 38090 }, { "grad_norm": 0.29822954535484314, "learning_rate": 5.80278349826324e-05, "loss": 0.0324, "step": 38100 }, { "grad_norm": 0.25328463315963745, "learning_rate": 5.8007434063302014e-05, "loss": 0.0269, "step": 38110 }, { "grad_norm": 0.30063173174858093, "learning_rate": 5.7987031775720136e-05, "loss": 0.0342, "step": 38120 }, { "grad_norm": 0.4101058840751648, "learning_rate": 5.7966628123372976e-05, "loss": 0.0411, "step": 38130 }, { "grad_norm": 0.450713574886322, "learning_rate": 5.7946223109746956e-05, "loss": 0.0393, "step": 38140 }, { "grad_norm": 0.32993683218955994, "learning_rate": 5.7925816738328754e-05, "loss": 0.028, "step": 38150 }, { "grad_norm": 0.30033740401268005, "learning_rate": 5.790540901260521e-05, "loss": 0.0292, "step": 38160 }, { "grad_norm": 0.2214137315750122, "learning_rate": 5.788499993606351e-05, "loss": 0.0342, "step": 38170 }, { "grad_norm": 0.38458141684532166, "learning_rate": 5.786458951219096e-05, "loss": 0.0362, "step": 38180 }, { "grad_norm": 0.316832572221756, "learning_rate": 5.784417774447517e-05, "loss": 0.0344, "step": 38190 }, { "grad_norm": 0.3238622844219208, "learning_rate": 5.782376463640393e-05, "loss": 0.0281, "step": 38200 }, { "grad_norm": 0.28722143173217773, "learning_rate": 5.780335019146531e-05, "loss": 0.0289, "step": 38210 }, { "grad_norm": 0.319123238325119, "learning_rate": 5.778293441314755e-05, "loss": 0.0276, "step": 38220 }, { "grad_norm": 0.4002346098423004, "learning_rate": 5.776251730493917e-05, "loss": 0.0355, "step": 38230 }, { "grad_norm": 0.26390933990478516, "learning_rate": 5.774209887032887e-05, "loss": 0.0302, "step": 38240 }, { "grad_norm": 0.2745715379714966, "learning_rate": 5.772167911280565e-05, "loss": 0.0282, "step": 38250 }, { "grad_norm": 0.5398614406585693, "learning_rate": 5.770125803585864e-05, "loss": 0.0287, "step": 38260 }, { "grad_norm": 0.3811841607093811, "learning_rate": 5.768083564297726e-05, "loss": 0.0385, "step": 38270 }, { "grad_norm": 0.362582802772522, "learning_rate": 5.766041193765114e-05, "loss": 0.0308, "step": 38280 }, { "grad_norm": 0.25432708859443665, "learning_rate": 5.763998692337015e-05, "loss": 0.0263, "step": 38290 }, { "grad_norm": 0.3509345054626465, "learning_rate": 5.761956060362433e-05, "loss": 0.0307, "step": 38300 }, { "grad_norm": 0.2853885591030121, "learning_rate": 5.7599132981904e-05, "loss": 0.0291, "step": 38310 }, { "grad_norm": 0.367134690284729, "learning_rate": 5.75787040616997e-05, "loss": 0.0347, "step": 38320 }, { "grad_norm": 0.39799702167510986, "learning_rate": 5.755827384650212e-05, "loss": 0.0284, "step": 38330 }, { "grad_norm": 0.27904176712036133, "learning_rate": 5.753784233980228e-05, "loss": 0.0276, "step": 38340 }, { "grad_norm": 0.228213369846344, "learning_rate": 5.751740954509135e-05, "loss": 0.0183, "step": 38350 }, { "grad_norm": 0.234116330742836, "learning_rate": 5.7496975465860715e-05, "loss": 0.0191, "step": 38360 }, { "grad_norm": 0.20440953969955444, "learning_rate": 5.747654010560202e-05, "loss": 0.0248, "step": 38370 }, { "grad_norm": 0.2522279918193817, "learning_rate": 5.7456103467807097e-05, "loss": 0.0341, "step": 38380 }, { "grad_norm": 0.5098846554756165, "learning_rate": 5.7435665555968046e-05, "loss": 0.0304, "step": 38390 }, { "grad_norm": 0.31540870666503906, "learning_rate": 5.74152263735771e-05, "loss": 0.0335, "step": 38400 }, { "grad_norm": 0.37625691294670105, "learning_rate": 5.739478592412677e-05, "loss": 0.0263, "step": 38410 }, { "grad_norm": 0.5110689401626587, "learning_rate": 5.7374344211109766e-05, "loss": 0.0362, "step": 38420 }, { "grad_norm": 0.32982325553894043, "learning_rate": 5.735390123801904e-05, "loss": 0.0271, "step": 38430 }, { "grad_norm": 0.3015527129173279, "learning_rate": 5.7333457008347704e-05, "loss": 0.0355, "step": 38440 }, { "grad_norm": 0.3003924489021301, "learning_rate": 5.7313011525589156e-05, "loss": 0.0301, "step": 38450 }, { "grad_norm": 0.18305502831935883, "learning_rate": 5.729256479323694e-05, "loss": 0.0276, "step": 38460 }, { "grad_norm": 0.19333778321743011, "learning_rate": 5.727211681478485e-05, "loss": 0.0315, "step": 38470 }, { "grad_norm": 0.3140905797481537, "learning_rate": 5.7251667593726886e-05, "loss": 0.0293, "step": 38480 }, { "grad_norm": 0.334370493888855, "learning_rate": 5.723121713355728e-05, "loss": 0.0378, "step": 38490 }, { "grad_norm": 0.30536651611328125, "learning_rate": 5.721076543777044e-05, "loss": 0.0308, "step": 38500 }, { "grad_norm": 0.25201743841171265, "learning_rate": 5.7190312509860986e-05, "loss": 0.0222, "step": 38510 }, { "grad_norm": 0.32103827595710754, "learning_rate": 5.716985835332379e-05, "loss": 0.0301, "step": 38520 }, { "grad_norm": 0.23907960951328278, "learning_rate": 5.714940297165389e-05, "loss": 0.0276, "step": 38530 }, { "grad_norm": 0.2514108717441559, "learning_rate": 5.712894636834656e-05, "loss": 0.0276, "step": 38540 }, { "grad_norm": 0.39192062616348267, "learning_rate": 5.7108488546897276e-05, "loss": 0.019, "step": 38550 }, { "grad_norm": 0.25076860189437866, "learning_rate": 5.708802951080172e-05, "loss": 0.0247, "step": 38560 }, { "grad_norm": 0.2750754952430725, "learning_rate": 5.706756926355576e-05, "loss": 0.0278, "step": 38570 }, { "grad_norm": 0.40422508120536804, "learning_rate": 5.704710780865554e-05, "loss": 0.0296, "step": 38580 }, { "grad_norm": 0.31738507747650146, "learning_rate": 5.7026645149597325e-05, "loss": 0.0236, "step": 38590 }, { "grad_norm": 0.40895992517471313, "learning_rate": 5.700618128987764e-05, "loss": 0.0328, "step": 38600 }, { "grad_norm": 0.3103794455528259, "learning_rate": 5.698571623299317e-05, "loss": 0.0273, "step": 38610 }, { "grad_norm": 0.2137165069580078, "learning_rate": 5.696524998244086e-05, "loss": 0.026, "step": 38620 }, { "grad_norm": 0.41965967416763306, "learning_rate": 5.6944782541717836e-05, "loss": 0.0325, "step": 38630 }, { "grad_norm": 0.24281790852546692, "learning_rate": 5.69243139143214e-05, "loss": 0.0234, "step": 38640 }, { "grad_norm": 0.2720239758491516, "learning_rate": 5.6903844103749125e-05, "loss": 0.0331, "step": 38650 }, { "grad_norm": 0.3132447302341461, "learning_rate": 5.688337311349869e-05, "loss": 0.0225, "step": 38660 }, { "grad_norm": 0.23257693648338318, "learning_rate": 5.6862900947068074e-05, "loss": 0.0269, "step": 38670 }, { "grad_norm": 0.3107931315898895, "learning_rate": 5.6842427607955374e-05, "loss": 0.0267, "step": 38680 }, { "grad_norm": 0.15824508666992188, "learning_rate": 5.682195309965893e-05, "loss": 0.023, "step": 38690 }, { "grad_norm": 0.31630533933639526, "learning_rate": 5.6801477425677294e-05, "loss": 0.0281, "step": 38700 }, { "grad_norm": 0.4989127218723297, "learning_rate": 5.678100058950917e-05, "loss": 0.0293, "step": 38710 }, { "grad_norm": 0.4808846116065979, "learning_rate": 5.676052259465352e-05, "loss": 0.0309, "step": 38720 }, { "grad_norm": 0.29068106412887573, "learning_rate": 5.674004344460945e-05, "loss": 0.0213, "step": 38730 }, { "grad_norm": 0.3056376278400421, "learning_rate": 5.6719563142876295e-05, "loss": 0.0217, "step": 38740 }, { "grad_norm": 0.2505461573600769, "learning_rate": 5.669908169295359e-05, "loss": 0.0236, "step": 38750 }, { "grad_norm": 0.4367232322692871, "learning_rate": 5.667859909834105e-05, "loss": 0.0283, "step": 38760 }, { "grad_norm": 0.21221014857292175, "learning_rate": 5.6658115362538565e-05, "loss": 0.0293, "step": 38770 }, { "grad_norm": 0.41661301255226135, "learning_rate": 5.663763048904628e-05, "loss": 0.0279, "step": 38780 }, { "grad_norm": 0.3815287947654724, "learning_rate": 5.661714448136447e-05, "loss": 0.0248, "step": 38790 }, { "grad_norm": 0.3069358170032501, "learning_rate": 5.659665734299366e-05, "loss": 0.025, "step": 38800 }, { "grad_norm": 0.292583703994751, "learning_rate": 5.6576169077434516e-05, "loss": 0.0304, "step": 38810 }, { "grad_norm": 0.2845503091812134, "learning_rate": 5.6555679688187944e-05, "loss": 0.0284, "step": 38820 }, { "grad_norm": 0.39070361852645874, "learning_rate": 5.6535189178755e-05, "loss": 0.0301, "step": 38830 }, { "grad_norm": 0.413616418838501, "learning_rate": 5.651469755263698e-05, "loss": 0.0277, "step": 38840 }, { "grad_norm": 0.19122813642024994, "learning_rate": 5.6494204813335316e-05, "loss": 0.0317, "step": 38850 }, { "grad_norm": 0.36019983887672424, "learning_rate": 5.647371096435168e-05, "loss": 0.0243, "step": 38860 }, { "grad_norm": 0.26727086305618286, "learning_rate": 5.645321600918788e-05, "loss": 0.0238, "step": 38870 }, { "grad_norm": 0.3677837550640106, "learning_rate": 5.643271995134597e-05, "loss": 0.0306, "step": 38880 }, { "grad_norm": 0.24486300349235535, "learning_rate": 5.641222279432814e-05, "loss": 0.0229, "step": 38890 }, { "grad_norm": 0.42959320545196533, "learning_rate": 5.6391724541636834e-05, "loss": 0.026, "step": 38900 }, { "grad_norm": 0.3245719075202942, "learning_rate": 5.6371225196774605e-05, "loss": 0.0297, "step": 38910 }, { "grad_norm": 0.27453553676605225, "learning_rate": 5.635072476324423e-05, "loss": 0.0297, "step": 38920 }, { "grad_norm": 0.3444298207759857, "learning_rate": 5.63302232445487e-05, "loss": 0.0296, "step": 38930 }, { "grad_norm": 0.251188188791275, "learning_rate": 5.6309720644191144e-05, "loss": 0.0299, "step": 38940 }, { "grad_norm": 0.368061900138855, "learning_rate": 5.628921696567491e-05, "loss": 0.0239, "step": 38950 }, { "grad_norm": 0.19973738491535187, "learning_rate": 5.62687122125035e-05, "loss": 0.0264, "step": 38960 }, { "grad_norm": 0.3512950837612152, "learning_rate": 5.624820638818062e-05, "loss": 0.0203, "step": 38970 }, { "grad_norm": 0.350694477558136, "learning_rate": 5.6227699496210164e-05, "loss": 0.0331, "step": 38980 }, { "grad_norm": 0.22771872580051422, "learning_rate": 5.6207191540096195e-05, "loss": 0.0273, "step": 38990 }, { "grad_norm": 0.30925145745277405, "learning_rate": 5.618668252334296e-05, "loss": 0.0254, "step": 39000 }, { "grad_norm": 0.3685552179813385, "learning_rate": 5.616617244945488e-05, "loss": 0.0273, "step": 39010 }, { "grad_norm": 0.32708799839019775, "learning_rate": 5.614566132193656e-05, "loss": 0.0264, "step": 39020 }, { "grad_norm": 0.37043628096580505, "learning_rate": 5.612514914429282e-05, "loss": 0.0309, "step": 39030 }, { "grad_norm": 0.3181270360946655, "learning_rate": 5.610463592002863e-05, "loss": 0.0282, "step": 39040 }, { "grad_norm": 0.30719801783561707, "learning_rate": 5.608412165264909e-05, "loss": 0.0243, "step": 39050 }, { "grad_norm": 0.3603699803352356, "learning_rate": 5.606360634565959e-05, "loss": 0.0235, "step": 39060 }, { "grad_norm": 0.16301876306533813, "learning_rate": 5.604309000256559e-05, "loss": 0.0317, "step": 39070 }, { "grad_norm": 0.32629844546318054, "learning_rate": 5.6022572626872785e-05, "loss": 0.0273, "step": 39080 }, { "grad_norm": 0.36305734515190125, "learning_rate": 5.600205422208704e-05, "loss": 0.0312, "step": 39090 }, { "grad_norm": 0.3047879636287689, "learning_rate": 5.59815347917144e-05, "loss": 0.0321, "step": 39100 }, { "grad_norm": 0.290285587310791, "learning_rate": 5.596101433926103e-05, "loss": 0.0284, "step": 39110 }, { "grad_norm": 0.38185203075408936, "learning_rate": 5.5940492868233364e-05, "loss": 0.0231, "step": 39120 }, { "grad_norm": 0.340901255607605, "learning_rate": 5.591997038213793e-05, "loss": 0.0319, "step": 39130 }, { "grad_norm": 0.28316473960876465, "learning_rate": 5.5899446884481475e-05, "loss": 0.0289, "step": 39140 }, { "grad_norm": 0.29894116520881653, "learning_rate": 5.5878922378770906e-05, "loss": 0.0406, "step": 39150 }, { "grad_norm": 0.3582388162612915, "learning_rate": 5.5858396868513285e-05, "loss": 0.0255, "step": 39160 }, { "grad_norm": 0.2677235007286072, "learning_rate": 5.583787035721586e-05, "loss": 0.0227, "step": 39170 }, { "grad_norm": 0.3738596737384796, "learning_rate": 5.581734284838606e-05, "loss": 0.0245, "step": 39180 }, { "grad_norm": 0.34539994597435, "learning_rate": 5.579681434553147e-05, "loss": 0.026, "step": 39190 }, { "grad_norm": 0.24838773906230927, "learning_rate": 5.5776284852159854e-05, "loss": 0.0252, "step": 39200 }, { "grad_norm": 0.3654478192329407, "learning_rate": 5.575575437177913e-05, "loss": 0.0299, "step": 39210 }, { "grad_norm": 0.34805619716644287, "learning_rate": 5.573522290789742e-05, "loss": 0.0248, "step": 39220 }, { "grad_norm": 0.25580912828445435, "learning_rate": 5.571469046402298e-05, "loss": 0.0255, "step": 39230 }, { "grad_norm": 0.30127987265586853, "learning_rate": 5.5694157043664205e-05, "loss": 0.0334, "step": 39240 }, { "grad_norm": 0.458554208278656, "learning_rate": 5.567362265032975e-05, "loss": 0.0266, "step": 39250 }, { "grad_norm": 0.30537471175193787, "learning_rate": 5.565308728752836e-05, "loss": 0.0238, "step": 39260 }, { "grad_norm": 0.24881480634212494, "learning_rate": 5.5632550958768956e-05, "loss": 0.0223, "step": 39270 }, { "grad_norm": 0.40057504177093506, "learning_rate": 5.5612013667560636e-05, "loss": 0.0289, "step": 39280 }, { "grad_norm": 0.40313127636909485, "learning_rate": 5.5591475417412676e-05, "loss": 0.0447, "step": 39290 }, { "grad_norm": 0.3292504847049713, "learning_rate": 5.557093621183451e-05, "loss": 0.0303, "step": 39300 }, { "grad_norm": 0.3924351632595062, "learning_rate": 5.55503960543357e-05, "loss": 0.0264, "step": 39310 }, { "grad_norm": 0.2725837528705597, "learning_rate": 5.552985494842601e-05, "loss": 0.0284, "step": 39320 }, { "grad_norm": 0.2561286389827728, "learning_rate": 5.550931289761534e-05, "loss": 0.0253, "step": 39330 }, { "grad_norm": 0.3880566358566284, "learning_rate": 5.548876990541378e-05, "loss": 0.0274, "step": 39340 }, { "grad_norm": 0.3768443763256073, "learning_rate": 5.5468225975331564e-05, "loss": 0.0337, "step": 39350 }, { "grad_norm": 0.2911495566368103, "learning_rate": 5.544768111087909e-05, "loss": 0.028, "step": 39360 }, { "grad_norm": 0.3053523302078247, "learning_rate": 5.54271353155669e-05, "loss": 0.0341, "step": 39370 }, { "grad_norm": 0.38366684317588806, "learning_rate": 5.5406588592905715e-05, "loss": 0.0265, "step": 39380 }, { "grad_norm": 0.26197168231010437, "learning_rate": 5.5386040946406416e-05, "loss": 0.0324, "step": 39390 }, { "grad_norm": 0.4041553735733032, "learning_rate": 5.536549237958004e-05, "loss": 0.0367, "step": 39400 }, { "grad_norm": 0.332260400056839, "learning_rate": 5.5344942895937744e-05, "loss": 0.0374, "step": 39410 }, { "grad_norm": 0.6285987496376038, "learning_rate": 5.5324392498990904e-05, "loss": 0.0253, "step": 39420 }, { "grad_norm": 0.42487001419067383, "learning_rate": 5.5303841192251e-05, "loss": 0.0297, "step": 39430 }, { "grad_norm": 0.3321819603443146, "learning_rate": 5.52832889792297e-05, "loss": 0.03, "step": 39440 }, { "grad_norm": 0.3423032760620117, "learning_rate": 5.526273586343881e-05, "loss": 0.0262, "step": 39450 }, { "grad_norm": 0.28423354029655457, "learning_rate": 5.5242181848390306e-05, "loss": 0.0299, "step": 39460 }, { "grad_norm": 0.3198718726634979, "learning_rate": 5.5221626937596285e-05, "loss": 0.0216, "step": 39470 }, { "grad_norm": 0.24399109184741974, "learning_rate": 5.520107113456903e-05, "loss": 0.0274, "step": 39480 }, { "grad_norm": 0.2828325033187866, "learning_rate": 5.5180514442820974e-05, "loss": 0.0364, "step": 39490 }, { "grad_norm": 0.329592227935791, "learning_rate": 5.515995686586469e-05, "loss": 0.0265, "step": 39500 }, { "grad_norm": 0.4455339014530182, "learning_rate": 5.5139398407212916e-05, "loss": 0.0204, "step": 39510 }, { "grad_norm": 0.3508807420730591, "learning_rate": 5.511883907037849e-05, "loss": 0.029, "step": 39520 }, { "grad_norm": 0.4496566355228424, "learning_rate": 5.509827885887449e-05, "loss": 0.0233, "step": 39530 }, { "grad_norm": 0.4632536470890045, "learning_rate": 5.507771777621406e-05, "loss": 0.0324, "step": 39540 }, { "grad_norm": 0.3689822554588318, "learning_rate": 5.505715582591052e-05, "loss": 0.0365, "step": 39550 }, { "grad_norm": 0.3901641368865967, "learning_rate": 5.50365930114774e-05, "loss": 0.0334, "step": 39560 }, { "grad_norm": 0.25260597467422485, "learning_rate": 5.5016029336428255e-05, "loss": 0.0192, "step": 39570 }, { "grad_norm": 0.2899860143661499, "learning_rate": 5.49954648042769e-05, "loss": 0.0304, "step": 39580 }, { "grad_norm": 0.39074084162712097, "learning_rate": 5.4974899418537226e-05, "loss": 0.0374, "step": 39590 }, { "grad_norm": 0.5460547804832458, "learning_rate": 5.4954333182723316e-05, "loss": 0.0287, "step": 39600 }, { "grad_norm": 0.301165908575058, "learning_rate": 5.493376610034937e-05, "loss": 0.0218, "step": 39610 }, { "grad_norm": 0.3788565993309021, "learning_rate": 5.4913198174929735e-05, "loss": 0.0258, "step": 39620 }, { "grad_norm": 0.23136374354362488, "learning_rate": 5.489262940997889e-05, "loss": 0.0281, "step": 39630 }, { "grad_norm": 0.24306495487689972, "learning_rate": 5.487205980901151e-05, "loss": 0.0335, "step": 39640 }, { "grad_norm": 0.28800418972969055, "learning_rate": 5.485148937554234e-05, "loss": 0.0464, "step": 39650 }, { "grad_norm": 0.2717556357383728, "learning_rate": 5.483091811308635e-05, "loss": 0.0217, "step": 39660 }, { "grad_norm": 0.3205527067184448, "learning_rate": 5.4810346025158564e-05, "loss": 0.0234, "step": 39670 }, { "grad_norm": 0.288568377494812, "learning_rate": 5.478977311527421e-05, "loss": 0.024, "step": 39680 }, { "grad_norm": 0.30298691987991333, "learning_rate": 5.476919938694863e-05, "loss": 0.0236, "step": 39690 }, { "grad_norm": 0.38649770617485046, "learning_rate": 5.474862484369733e-05, "loss": 0.0204, "step": 39700 }, { "grad_norm": 0.3468031883239746, "learning_rate": 5.472804948903589e-05, "loss": 0.0278, "step": 39710 }, { "grad_norm": 0.38410764932632446, "learning_rate": 5.470747332648013e-05, "loss": 0.0268, "step": 39720 }, { "grad_norm": 0.24840883910655975, "learning_rate": 5.468689635954591e-05, "loss": 0.0421, "step": 39730 }, { "grad_norm": 0.21962475776672363, "learning_rate": 5.46663185917493e-05, "loss": 0.0289, "step": 39740 }, { "grad_norm": 0.3293627202510834, "learning_rate": 5.464574002660645e-05, "loss": 0.0282, "step": 39750 }, { "grad_norm": 0.1943988800048828, "learning_rate": 5.4625160667633724e-05, "loss": 0.0336, "step": 39760 }, { "grad_norm": 0.38377881050109863, "learning_rate": 5.4604580518347505e-05, "loss": 0.0286, "step": 39770 }, { "grad_norm": 0.39059901237487793, "learning_rate": 5.458399958226442e-05, "loss": 0.03, "step": 39780 }, { "grad_norm": 0.31635695695877075, "learning_rate": 5.456341786290119e-05, "loss": 0.0279, "step": 39790 }, { "grad_norm": 0.39490288496017456, "learning_rate": 5.454283536377465e-05, "loss": 0.0363, "step": 39800 }, { "grad_norm": 0.22676829993724823, "learning_rate": 5.452225208840179e-05, "loss": 0.029, "step": 39810 }, { "grad_norm": 0.24325025081634521, "learning_rate": 5.450166804029975e-05, "loss": 0.0323, "step": 39820 }, { "grad_norm": 0.3273036777973175, "learning_rate": 5.448108322298574e-05, "loss": 0.0272, "step": 39830 }, { "grad_norm": 0.3565960228443146, "learning_rate": 5.446049763997717e-05, "loss": 0.0309, "step": 39840 }, { "grad_norm": 0.37649935483932495, "learning_rate": 5.4439911294791546e-05, "loss": 0.0241, "step": 39850 }, { "grad_norm": 0.33429667353630066, "learning_rate": 5.441932419094652e-05, "loss": 0.0263, "step": 39860 }, { "grad_norm": 0.3283766806125641, "learning_rate": 5.439873633195985e-05, "loss": 0.0249, "step": 39870 }, { "grad_norm": 0.2579801678657532, "learning_rate": 5.437814772134947e-05, "loss": 0.0313, "step": 39880 }, { "grad_norm": 0.3443048298358917, "learning_rate": 5.4357558362633366e-05, "loss": 0.028, "step": 39890 }, { "grad_norm": 0.24329161643981934, "learning_rate": 5.4336968259329726e-05, "loss": 0.0216, "step": 39900 }, { "grad_norm": 0.2744307219982147, "learning_rate": 5.431637741495681e-05, "loss": 0.0292, "step": 39910 }, { "grad_norm": 0.4230690598487854, "learning_rate": 5.429578583303307e-05, "loss": 0.0318, "step": 39920 }, { "grad_norm": 0.20571129024028778, "learning_rate": 5.427519351707701e-05, "loss": 0.0373, "step": 39930 }, { "grad_norm": 0.26585546135902405, "learning_rate": 5.42546004706073e-05, "loss": 0.0279, "step": 39940 }, { "grad_norm": 0.2707172632217407, "learning_rate": 5.4234006697142735e-05, "loss": 0.0249, "step": 39950 }, { "grad_norm": 0.32131755352020264, "learning_rate": 5.421341220020224e-05, "loss": 0.0261, "step": 39960 }, { "grad_norm": 0.20609712600708008, "learning_rate": 5.419281698330482e-05, "loss": 0.0305, "step": 39970 }, { "grad_norm": 0.1970214694738388, "learning_rate": 5.4172221049969665e-05, "loss": 0.0282, "step": 39980 }, { "grad_norm": 0.44425472617149353, "learning_rate": 5.415162440371604e-05, "loss": 0.0265, "step": 39990 }, { "grad_norm": 0.3140880763530731, "learning_rate": 5.413102704806334e-05, "loss": 0.0295, "step": 40000 }, { "grad_norm": 0.37148067355155945, "learning_rate": 5.41104289865311e-05, "loss": 0.0292, "step": 40010 }, { "grad_norm": 0.30676886439323425, "learning_rate": 5.408983022263898e-05, "loss": 0.0258, "step": 40020 }, { "grad_norm": 0.362233430147171, "learning_rate": 5.406923075990673e-05, "loss": 0.0332, "step": 40030 }, { "grad_norm": 0.43133795261383057, "learning_rate": 5.404863060185423e-05, "loss": 0.0291, "step": 40040 }, { "grad_norm": 0.38706421852111816, "learning_rate": 5.40280297520015e-05, "loss": 0.0297, "step": 40050 }, { "grad_norm": 0.307841032743454, "learning_rate": 5.400742821386865e-05, "loss": 0.0284, "step": 40060 }, { "grad_norm": 0.34652772545814514, "learning_rate": 5.398682599097592e-05, "loss": 0.0297, "step": 40070 }, { "grad_norm": 0.22666803002357483, "learning_rate": 5.396622308684367e-05, "loss": 0.022, "step": 40080 }, { "grad_norm": 0.411469042301178, "learning_rate": 5.394561950499236e-05, "loss": 0.0309, "step": 40090 }, { "grad_norm": 0.26579177379608154, "learning_rate": 5.39250152489426e-05, "loss": 0.0281, "step": 40100 }, { "grad_norm": 0.33952826261520386, "learning_rate": 5.390441032221507e-05, "loss": 0.0232, "step": 40110 }, { "grad_norm": 0.31425487995147705, "learning_rate": 5.388380472833062e-05, "loss": 0.0256, "step": 40120 }, { "grad_norm": 0.2223336398601532, "learning_rate": 5.386319847081016e-05, "loss": 0.026, "step": 40130 }, { "grad_norm": 0.3563458323478699, "learning_rate": 5.384259155317473e-05, "loss": 0.0274, "step": 40140 }, { "grad_norm": 0.3748290240764618, "learning_rate": 5.38219839789455e-05, "loss": 0.0306, "step": 40150 }, { "grad_norm": 0.2578279674053192, "learning_rate": 5.380137575164376e-05, "loss": 0.0306, "step": 40160 }, { "grad_norm": 0.2932383120059967, "learning_rate": 5.378076687479085e-05, "loss": 0.0273, "step": 40170 }, { "grad_norm": 0.2145087867975235, "learning_rate": 5.3760157351908305e-05, "loss": 0.027, "step": 40180 }, { "grad_norm": 0.22968128323554993, "learning_rate": 5.373954718651768e-05, "loss": 0.028, "step": 40190 }, { "grad_norm": 0.2595188617706299, "learning_rate": 5.371893638214074e-05, "loss": 0.0217, "step": 40200 }, { "grad_norm": 0.38622617721557617, "learning_rate": 5.369832494229927e-05, "loss": 0.0359, "step": 40210 }, { "grad_norm": 0.3642527461051941, "learning_rate": 5.367771287051524e-05, "loss": 0.0349, "step": 40220 }, { "grad_norm": 0.2865864038467407, "learning_rate": 5.365710017031065e-05, "loss": 0.0249, "step": 40230 }, { "grad_norm": 0.3310123085975647, "learning_rate": 5.3636486845207654e-05, "loss": 0.0298, "step": 40240 }, { "grad_norm": 0.7912841439247131, "learning_rate": 5.3615872898728526e-05, "loss": 0.0301, "step": 40250 }, { "grad_norm": 0.4682861864566803, "learning_rate": 5.3595258334395614e-05, "loss": 0.0322, "step": 40260 }, { "grad_norm": 0.3180302381515503, "learning_rate": 5.3574643155731396e-05, "loss": 0.0389, "step": 40270 }, { "grad_norm": 0.2950079143047333, "learning_rate": 5.3554027366258405e-05, "loss": 0.0243, "step": 40280 }, { "grad_norm": 0.535807728767395, "learning_rate": 5.3533410969499355e-05, "loss": 0.0312, "step": 40290 }, { "grad_norm": 0.19918526709079742, "learning_rate": 5.3512793968977e-05, "loss": 0.024, "step": 40300 }, { "grad_norm": 0.265551894903183, "learning_rate": 5.349217636821423e-05, "loss": 0.0331, "step": 40310 }, { "grad_norm": 0.24739503860473633, "learning_rate": 5.347155817073405e-05, "loss": 0.0257, "step": 40320 }, { "grad_norm": 0.3457552194595337, "learning_rate": 5.34509393800595e-05, "loss": 0.0281, "step": 40330 }, { "grad_norm": 0.294408917427063, "learning_rate": 5.343031999971381e-05, "loss": 0.0252, "step": 40340 }, { "grad_norm": 0.295585036277771, "learning_rate": 5.340970003322026e-05, "loss": 0.0291, "step": 40350 }, { "grad_norm": 0.3786352574825287, "learning_rate": 5.33890794841022e-05, "loss": 0.0289, "step": 40360 }, { "grad_norm": 0.28718850016593933, "learning_rate": 5.336845835588318e-05, "loss": 0.0289, "step": 40370 }, { "grad_norm": 0.32028254866600037, "learning_rate": 5.334783665208674e-05, "loss": 0.0259, "step": 40380 }, { "grad_norm": 0.27566781640052795, "learning_rate": 5.332721437623657e-05, "loss": 0.0216, "step": 40390 }, { "grad_norm": 0.3514571487903595, "learning_rate": 5.3306591531856464e-05, "loss": 0.0298, "step": 40400 }, { "grad_norm": 0.29699617624282837, "learning_rate": 5.3285968122470295e-05, "loss": 0.0288, "step": 40410 }, { "grad_norm": 0.2645794451236725, "learning_rate": 5.326534415160207e-05, "loss": 0.0275, "step": 40420 }, { "grad_norm": 0.2912450134754181, "learning_rate": 5.324471962277582e-05, "loss": 0.0328, "step": 40430 }, { "grad_norm": 0.553871214389801, "learning_rate": 5.3224094539515746e-05, "loss": 0.0285, "step": 40440 }, { "grad_norm": 0.34018972516059875, "learning_rate": 5.3203468905346076e-05, "loss": 0.0281, "step": 40450 }, { "grad_norm": 0.29417017102241516, "learning_rate": 5.3182842723791195e-05, "loss": 0.0278, "step": 40460 }, { "grad_norm": 0.22847053408622742, "learning_rate": 5.316221599837554e-05, "loss": 0.0276, "step": 40470 }, { "grad_norm": 0.29635676741600037, "learning_rate": 5.314158873262366e-05, "loss": 0.0357, "step": 40480 }, { "grad_norm": 0.2696135342121124, "learning_rate": 5.312096093006018e-05, "loss": 0.0253, "step": 40490 }, { "grad_norm": 0.274080753326416, "learning_rate": 5.3100332594209846e-05, "loss": 0.0265, "step": 40500 }, { "grad_norm": 0.2754381597042084, "learning_rate": 5.3079703728597454e-05, "loss": 0.0279, "step": 40510 }, { "grad_norm": 0.29976686835289, "learning_rate": 5.305907433674794e-05, "loss": 0.0272, "step": 40520 }, { "grad_norm": 0.3660055100917816, "learning_rate": 5.3038444422186284e-05, "loss": 0.0243, "step": 40530 }, { "grad_norm": 0.28712135553359985, "learning_rate": 5.301781398843757e-05, "loss": 0.0215, "step": 40540 }, { "grad_norm": 0.27114126086235046, "learning_rate": 5.299718303902699e-05, "loss": 0.0254, "step": 40550 }, { "grad_norm": 0.2297104001045227, "learning_rate": 5.297655157747979e-05, "loss": 0.0302, "step": 40560 }, { "grad_norm": 0.2936817407608032, "learning_rate": 5.295591960732136e-05, "loss": 0.0237, "step": 40570 }, { "grad_norm": 0.34442684054374695, "learning_rate": 5.293528713207708e-05, "loss": 0.0268, "step": 40580 }, { "grad_norm": 0.21224328875541687, "learning_rate": 5.291465415527253e-05, "loss": 0.0305, "step": 40590 }, { "grad_norm": 0.2792039215564728, "learning_rate": 5.2894020680433296e-05, "loss": 0.026, "step": 40600 }, { "grad_norm": 0.3194691836833954, "learning_rate": 5.287338671108507e-05, "loss": 0.0293, "step": 40610 }, { "grad_norm": 0.3441384732723236, "learning_rate": 5.285275225075367e-05, "loss": 0.0245, "step": 40620 }, { "grad_norm": 0.47020548582077026, "learning_rate": 5.283211730296492e-05, "loss": 0.0216, "step": 40630 }, { "grad_norm": 0.3473193347454071, "learning_rate": 5.281148187124477e-05, "loss": 0.0278, "step": 40640 }, { "grad_norm": 0.3236335217952728, "learning_rate": 5.279084595911927e-05, "loss": 0.0267, "step": 40650 }, { "grad_norm": 0.2368704229593277, "learning_rate": 5.277020957011453e-05, "loss": 0.0344, "step": 40660 }, { "grad_norm": 0.3249897360801697, "learning_rate": 5.274957270775673e-05, "loss": 0.0331, "step": 40670 }, { "grad_norm": 0.23025356233119965, "learning_rate": 5.2728935375572164e-05, "loss": 0.0182, "step": 40680 }, { "grad_norm": 0.2047051191329956, "learning_rate": 5.2708297577087176e-05, "loss": 0.0225, "step": 40690 }, { "grad_norm": 0.3331270217895508, "learning_rate": 5.2687659315828196e-05, "loss": 0.024, "step": 40700 }, { "grad_norm": 0.32851776480674744, "learning_rate": 5.266702059532175e-05, "loss": 0.0233, "step": 40710 }, { "grad_norm": 0.3253847062587738, "learning_rate": 5.264638141909444e-05, "loss": 0.0301, "step": 40720 }, { "grad_norm": 0.41495996713638306, "learning_rate": 5.2625741790672925e-05, "loss": 0.0308, "step": 40730 }, { "grad_norm": 0.24448262155056, "learning_rate": 5.2605101713583935e-05, "loss": 0.0239, "step": 40740 }, { "grad_norm": 0.4057137072086334, "learning_rate": 5.2584461191354315e-05, "loss": 0.0351, "step": 40750 }, { "grad_norm": 0.27585670351982117, "learning_rate": 5.2563820227510964e-05, "loss": 0.0246, "step": 40760 }, { "grad_norm": 0.3033878207206726, "learning_rate": 5.2543178825580864e-05, "loss": 0.024, "step": 40770 }, { "grad_norm": 0.5482703447341919, "learning_rate": 5.2522536989091054e-05, "loss": 0.0402, "step": 40780 }, { "grad_norm": 0.386498361825943, "learning_rate": 5.2501894721568655e-05, "loss": 0.0255, "step": 40790 }, { "grad_norm": 0.31273600459098816, "learning_rate": 5.248125202654089e-05, "loss": 0.0201, "step": 40800 }, { "grad_norm": 0.33372271060943604, "learning_rate": 5.246060890753501e-05, "loss": 0.0344, "step": 40810 }, { "grad_norm": 0.26731595396995544, "learning_rate": 5.243996536807837e-05, "loss": 0.0301, "step": 40820 }, { "grad_norm": 0.29579222202301025, "learning_rate": 5.2419321411698384e-05, "loss": 0.0248, "step": 40830 }, { "grad_norm": 0.2675968110561371, "learning_rate": 5.239867704192253e-05, "loss": 0.0287, "step": 40840 }, { "grad_norm": 0.29448947310447693, "learning_rate": 5.237803226227838e-05, "loss": 0.0273, "step": 40850 }, { "grad_norm": 0.3763067126274109, "learning_rate": 5.235738707629354e-05, "loss": 0.0293, "step": 40860 }, { "grad_norm": 0.2621137797832489, "learning_rate": 5.233674148749575e-05, "loss": 0.0281, "step": 40870 }, { "grad_norm": 0.4009001553058624, "learning_rate": 5.231609549941272e-05, "loss": 0.0313, "step": 40880 }, { "grad_norm": 0.32504892349243164, "learning_rate": 5.229544911557233e-05, "loss": 0.0322, "step": 40890 }, { "grad_norm": 0.29126986861228943, "learning_rate": 5.227480233950246e-05, "loss": 0.0282, "step": 40900 }, { "grad_norm": 0.3503286838531494, "learning_rate": 5.22541551747311e-05, "loss": 0.028, "step": 40910 }, { "grad_norm": 0.3503158986568451, "learning_rate": 5.223350762478626e-05, "loss": 0.0245, "step": 40920 }, { "grad_norm": 0.20451214909553528, "learning_rate": 5.221285969319606e-05, "loss": 0.0219, "step": 40930 }, { "grad_norm": 0.3641137182712555, "learning_rate": 5.219221138348865e-05, "loss": 0.0253, "step": 40940 }, { "grad_norm": 0.400177538394928, "learning_rate": 5.217156269919228e-05, "loss": 0.0282, "step": 40950 }, { "grad_norm": 0.18711473047733307, "learning_rate": 5.215091364383523e-05, "loss": 0.0244, "step": 40960 }, { "grad_norm": 0.3649730384349823, "learning_rate": 5.213026422094588e-05, "loss": 0.0288, "step": 40970 }, { "grad_norm": 0.22772598266601562, "learning_rate": 5.210961443405262e-05, "loss": 0.0263, "step": 40980 }, { "grad_norm": 0.21382640302181244, "learning_rate": 5.208896428668396e-05, "loss": 0.0195, "step": 40990 }, { "grad_norm": 0.41075146198272705, "learning_rate": 5.206831378236845e-05, "loss": 0.026, "step": 41000 }, { "grad_norm": 0.3246753513813019, "learning_rate": 5.2047662924634666e-05, "loss": 0.0229, "step": 41010 }, { "grad_norm": 0.29579296708106995, "learning_rate": 5.20270117170113e-05, "loss": 0.0193, "step": 41020 }, { "grad_norm": 0.47893115878105164, "learning_rate": 5.200636016302707e-05, "loss": 0.0351, "step": 41030 }, { "grad_norm": 0.36364519596099854, "learning_rate": 5.1985708266210754e-05, "loss": 0.029, "step": 41040 }, { "grad_norm": 0.39704668521881104, "learning_rate": 5.1965056030091206e-05, "loss": 0.0302, "step": 41050 }, { "grad_norm": 0.22258351743221283, "learning_rate": 5.194440345819732e-05, "loss": 0.0226, "step": 41060 }, { "grad_norm": 0.21899226307868958, "learning_rate": 5.1923750554058084e-05, "loss": 0.0301, "step": 41070 }, { "grad_norm": 0.2581762671470642, "learning_rate": 5.1903097321202476e-05, "loss": 0.0239, "step": 41080 }, { "grad_norm": 0.3996482789516449, "learning_rate": 5.18824437631596e-05, "loss": 0.0374, "step": 41090 }, { "grad_norm": 0.2499598264694214, "learning_rate": 5.186178988345856e-05, "loss": 0.0311, "step": 41100 }, { "grad_norm": 0.40230777859687805, "learning_rate": 5.184113568562855e-05, "loss": 0.0325, "step": 41110 }, { "grad_norm": 0.3672657608985901, "learning_rate": 5.18204811731988e-05, "loss": 0.0326, "step": 41120 }, { "grad_norm": 0.3893367648124695, "learning_rate": 5.179982634969862e-05, "loss": 0.0269, "step": 41130 }, { "grad_norm": 0.3941788077354431, "learning_rate": 5.177917121865734e-05, "loss": 0.0335, "step": 41140 }, { "grad_norm": 0.32222384214401245, "learning_rate": 5.1758515783604346e-05, "loss": 0.0219, "step": 41150 }, { "grad_norm": 0.40800872445106506, "learning_rate": 5.173786004806911e-05, "loss": 0.0307, "step": 41160 }, { "grad_norm": 0.42298710346221924, "learning_rate": 5.1717204015581135e-05, "loss": 0.0262, "step": 41170 }, { "grad_norm": 0.3415125906467438, "learning_rate": 5.169654768966994e-05, "loss": 0.0268, "step": 41180 }, { "grad_norm": 0.4278954267501831, "learning_rate": 5.1675891073865156e-05, "loss": 0.0337, "step": 41190 }, { "grad_norm": 0.3873600959777832, "learning_rate": 5.1655234171696424e-05, "loss": 0.022, "step": 41200 }, { "grad_norm": 0.3823595345020294, "learning_rate": 5.163457698669343e-05, "loss": 0.0255, "step": 41210 }, { "grad_norm": 0.2716676890850067, "learning_rate": 5.1613919522385936e-05, "loss": 0.0228, "step": 41220 }, { "grad_norm": 0.3747277557849884, "learning_rate": 5.1593261782303746e-05, "loss": 0.0355, "step": 41230 }, { "grad_norm": 0.34067007899284363, "learning_rate": 5.157260376997669e-05, "loss": 0.0215, "step": 41240 }, { "grad_norm": 0.3624221384525299, "learning_rate": 5.155194548893466e-05, "loss": 0.0275, "step": 41250 }, { "grad_norm": 0.20809932053089142, "learning_rate": 5.153128694270759e-05, "loss": 0.0309, "step": 41260 }, { "grad_norm": 0.3814058303833008, "learning_rate": 5.151062813482548e-05, "loss": 0.0363, "step": 41270 }, { "grad_norm": 0.25879451632499695, "learning_rate": 5.148996906881832e-05, "loss": 0.0196, "step": 41280 }, { "grad_norm": 0.5438963174819946, "learning_rate": 5.1469309748216196e-05, "loss": 0.0241, "step": 41290 }, { "grad_norm": 0.25319305062294006, "learning_rate": 5.144865017654923e-05, "loss": 0.0219, "step": 41300 }, { "grad_norm": 0.28541502356529236, "learning_rate": 5.1427990357347556e-05, "loss": 0.0303, "step": 41310 }, { "grad_norm": 0.26448488235473633, "learning_rate": 5.140733029414139e-05, "loss": 0.0293, "step": 41320 }, { "grad_norm": 0.35156771540641785, "learning_rate": 5.138666999046098e-05, "loss": 0.0222, "step": 41330 }, { "grad_norm": 0.37918686866760254, "learning_rate": 5.136600944983658e-05, "loss": 0.0258, "step": 41340 }, { "grad_norm": 0.32933664321899414, "learning_rate": 5.134534867579853e-05, "loss": 0.0257, "step": 41350 }, { "grad_norm": 0.38826867938041687, "learning_rate": 5.1324687671877184e-05, "loss": 0.0212, "step": 41360 }, { "grad_norm": 0.34945666790008545, "learning_rate": 5.130402644160296e-05, "loss": 0.0353, "step": 41370 }, { "grad_norm": 0.25896313786506653, "learning_rate": 5.128336498850628e-05, "loss": 0.0276, "step": 41380 }, { "grad_norm": 0.30559271574020386, "learning_rate": 5.126270331611761e-05, "loss": 0.0254, "step": 41390 }, { "grad_norm": 0.3508128523826599, "learning_rate": 5.124204142796748e-05, "loss": 0.025, "step": 41400 }, { "grad_norm": 0.26511821150779724, "learning_rate": 5.122137932758644e-05, "loss": 0.0231, "step": 41410 }, { "grad_norm": 0.2786664366722107, "learning_rate": 5.1200717018505074e-05, "loss": 0.0266, "step": 41420 }, { "grad_norm": 0.24434006214141846, "learning_rate": 5.118005450425403e-05, "loss": 0.0256, "step": 41430 }, { "grad_norm": 0.29376548528671265, "learning_rate": 5.115939178836392e-05, "loss": 0.0209, "step": 41440 }, { "grad_norm": 0.399898499250412, "learning_rate": 5.113872887436547e-05, "loss": 0.0318, "step": 41450 }, { "grad_norm": 0.17286264896392822, "learning_rate": 5.11180657657894e-05, "loss": 0.0304, "step": 41460 }, { "grad_norm": 0.4005465507507324, "learning_rate": 5.109740246616649e-05, "loss": 0.0227, "step": 41470 }, { "grad_norm": 0.4275234043598175, "learning_rate": 5.10767389790275e-05, "loss": 0.0376, "step": 41480 }, { "grad_norm": 0.31608307361602783, "learning_rate": 5.1056075307903265e-05, "loss": 0.0263, "step": 41490 }, { "grad_norm": 0.3107813000679016, "learning_rate": 5.1035411456324634e-05, "loss": 0.0266, "step": 41500 }, { "grad_norm": 0.2809270918369293, "learning_rate": 5.1014747427822504e-05, "loss": 0.0296, "step": 41510 }, { "grad_norm": 0.4315887987613678, "learning_rate": 5.0994083225927804e-05, "loss": 0.0288, "step": 41520 }, { "grad_norm": 0.24360565841197968, "learning_rate": 5.0973418854171475e-05, "loss": 0.0283, "step": 41530 }, { "grad_norm": 0.22777099907398224, "learning_rate": 5.095275431608447e-05, "loss": 0.021, "step": 41540 }, { "grad_norm": 0.3152449429035187, "learning_rate": 5.093208961519782e-05, "loss": 0.0236, "step": 41550 }, { "grad_norm": 0.36699846386909485, "learning_rate": 5.091142475504255e-05, "loss": 0.0232, "step": 41560 }, { "grad_norm": 0.41960713267326355, "learning_rate": 5.089075973914971e-05, "loss": 0.0307, "step": 41570 }, { "grad_norm": 0.32013264298439026, "learning_rate": 5.08700945710504e-05, "loss": 0.0341, "step": 41580 }, { "grad_norm": 0.46276795864105225, "learning_rate": 5.0849429254275714e-05, "loss": 0.0255, "step": 41590 }, { "grad_norm": 0.21287396550178528, "learning_rate": 5.0828763792356804e-05, "loss": 0.0247, "step": 41600 }, { "grad_norm": 0.3479083776473999, "learning_rate": 5.080809818882483e-05, "loss": 0.0291, "step": 41610 }, { "grad_norm": 0.3874790370464325, "learning_rate": 5.078743244721097e-05, "loss": 0.0238, "step": 41620 }, { "grad_norm": 0.32151561975479126, "learning_rate": 5.0766766571046455e-05, "loss": 0.027, "step": 41630 }, { "grad_norm": 0.2423357516527176, "learning_rate": 5.07461005638625e-05, "loss": 0.0208, "step": 41640 }, { "grad_norm": 0.29442837834358215, "learning_rate": 5.072543442919037e-05, "loss": 0.0231, "step": 41650 }, { "grad_norm": 0.30717721581459045, "learning_rate": 5.070476817056132e-05, "loss": 0.033, "step": 41660 }, { "grad_norm": 0.2530716061592102, "learning_rate": 5.068410179150668e-05, "loss": 0.03, "step": 41670 }, { "grad_norm": 0.24381814897060394, "learning_rate": 5.066343529555775e-05, "loss": 0.0245, "step": 41680 }, { "grad_norm": 0.21075493097305298, "learning_rate": 5.064276868624587e-05, "loss": 0.0273, "step": 41690 }, { "grad_norm": 0.26675257086753845, "learning_rate": 5.0622101967102396e-05, "loss": 0.0238, "step": 41700 }, { "grad_norm": 0.4215582311153412, "learning_rate": 5.0601435141658714e-05, "loss": 0.028, "step": 41710 }, { "grad_norm": 0.6508362293243408, "learning_rate": 5.058076821344622e-05, "loss": 0.0323, "step": 41720 }, { "grad_norm": 0.4194391369819641, "learning_rate": 5.056010118599632e-05, "loss": 0.0308, "step": 41730 }, { "grad_norm": 0.3493707478046417, "learning_rate": 5.053943406284044e-05, "loss": 0.024, "step": 41740 }, { "grad_norm": 0.3153296411037445, "learning_rate": 5.051876684751006e-05, "loss": 0.0287, "step": 41750 }, { "grad_norm": 0.18082863092422485, "learning_rate": 5.0498099543536584e-05, "loss": 0.0244, "step": 41760 }, { "grad_norm": 0.302306592464447, "learning_rate": 5.047743215445152e-05, "loss": 0.0232, "step": 41770 }, { "grad_norm": 0.32733121514320374, "learning_rate": 5.045676468378637e-05, "loss": 0.0252, "step": 41780 }, { "grad_norm": 0.31038516759872437, "learning_rate": 5.0436097135072626e-05, "loss": 0.0252, "step": 41790 }, { "grad_norm": 0.3211532235145569, "learning_rate": 5.041542951184181e-05, "loss": 0.0292, "step": 41800 }, { "grad_norm": 0.2862666845321655, "learning_rate": 5.039476181762545e-05, "loss": 0.0254, "step": 41810 }, { "grad_norm": 0.17726625502109528, "learning_rate": 5.037409405595508e-05, "loss": 0.0194, "step": 41820 }, { "grad_norm": 0.45986947417259216, "learning_rate": 5.035342623036229e-05, "loss": 0.0281, "step": 41830 }, { "grad_norm": 0.384502112865448, "learning_rate": 5.033275834437862e-05, "loss": 0.0303, "step": 41840 }, { "grad_norm": 0.35973894596099854, "learning_rate": 5.031209040153564e-05, "loss": 0.0264, "step": 41850 }, { "grad_norm": 0.3137392997741699, "learning_rate": 5.0291422405364955e-05, "loss": 0.0346, "step": 41860 }, { "grad_norm": 0.32233405113220215, "learning_rate": 5.0270754359398133e-05, "loss": 0.0252, "step": 41870 }, { "grad_norm": 0.34476354718208313, "learning_rate": 5.025008626716682e-05, "loss": 0.0235, "step": 41880 }, { "grad_norm": 0.3239130973815918, "learning_rate": 5.0229418132202585e-05, "loss": 0.0245, "step": 41890 }, { "grad_norm": 0.3771968185901642, "learning_rate": 5.020874995803707e-05, "loss": 0.0257, "step": 41900 }, { "grad_norm": 0.3034530282020569, "learning_rate": 5.01880817482019e-05, "loss": 0.0284, "step": 41910 }, { "grad_norm": 0.3721100389957428, "learning_rate": 5.01674135062287e-05, "loss": 0.0222, "step": 41920 }, { "grad_norm": 0.49684882164001465, "learning_rate": 5.0146745235649115e-05, "loss": 0.0289, "step": 41930 }, { "grad_norm": 0.37644341588020325, "learning_rate": 5.0126076939994795e-05, "loss": 0.0235, "step": 41940 }, { "grad_norm": 0.2031085044145584, "learning_rate": 5.010540862279736e-05, "loss": 0.0315, "step": 41950 }, { "grad_norm": 0.2959617078304291, "learning_rate": 5.008474028758846e-05, "loss": 0.0317, "step": 41960 }, { "grad_norm": 0.23636344075202942, "learning_rate": 5.0064071937899784e-05, "loss": 0.023, "step": 41970 }, { "grad_norm": 0.35336601734161377, "learning_rate": 5.004340357726296e-05, "loss": 0.025, "step": 41980 }, { "grad_norm": 0.34601661562919617, "learning_rate": 5.002273520920965e-05, "loss": 0.0273, "step": 41990 }, { "grad_norm": 0.31703507900238037, "learning_rate": 5.000206683727151e-05, "loss": 0.0256, "step": 42000 }, { "grad_norm": 0.21060369908809662, "learning_rate": 4.998139846498021e-05, "loss": 0.0218, "step": 42010 }, { "grad_norm": 0.33722707629203796, "learning_rate": 4.9960730095867405e-05, "loss": 0.0316, "step": 42020 }, { "grad_norm": 0.29739269614219666, "learning_rate": 4.9940061733464755e-05, "loss": 0.031, "step": 42030 }, { "grad_norm": 0.33965322375297546, "learning_rate": 4.991939338130392e-05, "loss": 0.0372, "step": 42040 }, { "grad_norm": 0.38241925835609436, "learning_rate": 4.989872504291653e-05, "loss": 0.0333, "step": 42050 }, { "grad_norm": 0.2763594388961792, "learning_rate": 4.9878056721834273e-05, "loss": 0.0235, "step": 42060 }, { "grad_norm": 0.200200617313385, "learning_rate": 4.98573884215888e-05, "loss": 0.0238, "step": 42070 }, { "grad_norm": 0.4093233048915863, "learning_rate": 4.9836720145711715e-05, "loss": 0.0317, "step": 42080 }, { "grad_norm": 0.3025432229042053, "learning_rate": 4.9816051897734725e-05, "loss": 0.0329, "step": 42090 }, { "grad_norm": 0.4188012182712555, "learning_rate": 4.979538368118941e-05, "loss": 0.0336, "step": 42100 }, { "grad_norm": 0.22938260436058044, "learning_rate": 4.9774715499607446e-05, "loss": 0.0287, "step": 42110 }, { "grad_norm": 0.24022801220417023, "learning_rate": 4.9754047356520416e-05, "loss": 0.0306, "step": 42120 }, { "grad_norm": 0.21665354073047638, "learning_rate": 4.973337925545999e-05, "loss": 0.0289, "step": 42130 }, { "grad_norm": 0.3529248535633087, "learning_rate": 4.9712711199957746e-05, "loss": 0.0255, "step": 42140 }, { "grad_norm": 0.3158392906188965, "learning_rate": 4.969204319354531e-05, "loss": 0.0299, "step": 42150 }, { "grad_norm": 0.3073175847530365, "learning_rate": 4.9671375239754267e-05, "loss": 0.0251, "step": 42160 }, { "grad_norm": 0.22251389920711517, "learning_rate": 4.9650707342116195e-05, "loss": 0.0249, "step": 42170 }, { "grad_norm": 0.269316166639328, "learning_rate": 4.96300395041627e-05, "loss": 0.0213, "step": 42180 }, { "grad_norm": 0.2874390482902527, "learning_rate": 4.960937172942532e-05, "loss": 0.029, "step": 42190 }, { "grad_norm": 0.2908846139907837, "learning_rate": 4.958870402143565e-05, "loss": 0.0248, "step": 42200 }, { "grad_norm": 0.18826036155223846, "learning_rate": 4.9568036383725186e-05, "loss": 0.0253, "step": 42210 }, { "grad_norm": 0.2297188937664032, "learning_rate": 4.9547368819825516e-05, "loss": 0.0258, "step": 42220 }, { "grad_norm": 0.22761385142803192, "learning_rate": 4.952670133326812e-05, "loss": 0.0243, "step": 42230 }, { "grad_norm": 0.33502236008644104, "learning_rate": 4.950603392758453e-05, "loss": 0.0261, "step": 42240 }, { "grad_norm": 0.2912108302116394, "learning_rate": 4.948536660630621e-05, "loss": 0.027, "step": 42250 }, { "grad_norm": 0.3089187443256378, "learning_rate": 4.9464699372964685e-05, "loss": 0.0285, "step": 42260 }, { "grad_norm": 0.3017914891242981, "learning_rate": 4.9444032231091395e-05, "loss": 0.025, "step": 42270 }, { "grad_norm": 0.3646116256713867, "learning_rate": 4.9423365184217776e-05, "loss": 0.0291, "step": 42280 }, { "grad_norm": 0.3992080092430115, "learning_rate": 4.940269823587529e-05, "loss": 0.0284, "step": 42290 }, { "grad_norm": 0.17994844913482666, "learning_rate": 4.938203138959533e-05, "loss": 0.0174, "step": 42300 }, { "grad_norm": 0.32942649722099304, "learning_rate": 4.9361364648909325e-05, "loss": 0.0318, "step": 42310 }, { "grad_norm": 0.3030155301094055, "learning_rate": 4.9340698017348605e-05, "loss": 0.0206, "step": 42320 }, { "grad_norm": 0.3998437523841858, "learning_rate": 4.932003149844458e-05, "loss": 0.0241, "step": 42330 }, { "grad_norm": 0.3856890797615051, "learning_rate": 4.929936509572857e-05, "loss": 0.0348, "step": 42340 }, { "grad_norm": 0.3979509174823761, "learning_rate": 4.927869881273191e-05, "loss": 0.0307, "step": 42350 }, { "grad_norm": 0.2737308442592621, "learning_rate": 4.9258032652985894e-05, "loss": 0.0256, "step": 42360 }, { "grad_norm": 0.30434414744377136, "learning_rate": 4.9237366620021786e-05, "loss": 0.0312, "step": 42370 }, { "grad_norm": 0.2711087465286255, "learning_rate": 4.921670071737089e-05, "loss": 0.0236, "step": 42380 }, { "grad_norm": 0.4929925203323364, "learning_rate": 4.91960349485644e-05, "loss": 0.0315, "step": 42390 }, { "grad_norm": 0.3100684881210327, "learning_rate": 4.9175369317133566e-05, "loss": 0.0236, "step": 42400 }, { "grad_norm": 0.2865561544895172, "learning_rate": 4.915470382660953e-05, "loss": 0.0268, "step": 42410 }, { "grad_norm": 0.3048727512359619, "learning_rate": 4.9134038480523524e-05, "loss": 0.026, "step": 42420 }, { "grad_norm": 0.24459117650985718, "learning_rate": 4.911337328240664e-05, "loss": 0.0274, "step": 42430 }, { "grad_norm": 0.2579374611377716, "learning_rate": 4.909270823579003e-05, "loss": 0.0319, "step": 42440 }, { "grad_norm": 0.38562530279159546, "learning_rate": 4.907204334420476e-05, "loss": 0.0249, "step": 42450 }, { "grad_norm": 0.2863835394382477, "learning_rate": 4.9051378611181884e-05, "loss": 0.0298, "step": 42460 }, { "grad_norm": 0.1916125863790512, "learning_rate": 4.9030714040252486e-05, "loss": 0.0291, "step": 42470 }, { "grad_norm": 0.20947903394699097, "learning_rate": 4.901004963494752e-05, "loss": 0.0234, "step": 42480 }, { "grad_norm": 0.4231193959712982, "learning_rate": 4.898938539879802e-05, "loss": 0.0299, "step": 42490 }, { "grad_norm": 0.28924426436424255, "learning_rate": 4.8968721335334904e-05, "loss": 0.0249, "step": 42500 }, { "grad_norm": 0.28912341594696045, "learning_rate": 4.894805744808912e-05, "loss": 0.03, "step": 42510 }, { "grad_norm": 0.2658681869506836, "learning_rate": 4.8927393740591524e-05, "loss": 0.0281, "step": 42520 }, { "grad_norm": 0.2975062131881714, "learning_rate": 4.890673021637302e-05, "loss": 0.0281, "step": 42530 }, { "grad_norm": 0.2338578850030899, "learning_rate": 4.888606687896442e-05, "loss": 0.0299, "step": 42540 }, { "grad_norm": 0.32980602979660034, "learning_rate": 4.886540373189652e-05, "loss": 0.0283, "step": 42550 }, { "grad_norm": 0.23512287437915802, "learning_rate": 4.88447407787001e-05, "loss": 0.0278, "step": 42560 }, { "grad_norm": 0.3255782425403595, "learning_rate": 4.8824078022905864e-05, "loss": 0.0281, "step": 42570 }, { "grad_norm": 0.3617420494556427, "learning_rate": 4.880341546804455e-05, "loss": 0.0228, "step": 42580 }, { "grad_norm": 0.3435925245285034, "learning_rate": 4.878275311764679e-05, "loss": 0.0286, "step": 42590 }, { "grad_norm": 0.3037911355495453, "learning_rate": 4.8762090975243254e-05, "loss": 0.0234, "step": 42600 }, { "grad_norm": 0.23865796625614166, "learning_rate": 4.874142904436448e-05, "loss": 0.0236, "step": 42610 }, { "grad_norm": 0.2575293779373169, "learning_rate": 4.8720767328541094e-05, "loss": 0.0306, "step": 42620 }, { "grad_norm": 0.3381730616092682, "learning_rate": 4.870010583130358e-05, "loss": 0.0278, "step": 42630 }, { "grad_norm": 0.2310461401939392, "learning_rate": 4.867944455618241e-05, "loss": 0.0242, "step": 42640 }, { "grad_norm": 0.3430148959159851, "learning_rate": 4.865878350670807e-05, "loss": 0.0318, "step": 42650 }, { "grad_norm": 0.405351847410202, "learning_rate": 4.8638122686410914e-05, "loss": 0.0232, "step": 42660 }, { "grad_norm": 0.26711443066596985, "learning_rate": 4.861746209882137e-05, "loss": 0.0217, "step": 42670 }, { "grad_norm": 0.3105545938014984, "learning_rate": 4.859680174746972e-05, "loss": 0.0273, "step": 42680 }, { "grad_norm": 0.41647833585739136, "learning_rate": 4.857614163588629e-05, "loss": 0.0371, "step": 42690 }, { "grad_norm": 0.2589680850505829, "learning_rate": 4.855548176760131e-05, "loss": 0.0267, "step": 42700 }, { "grad_norm": 0.3206012547016144, "learning_rate": 4.8534822146144986e-05, "loss": 0.0311, "step": 42710 }, { "grad_norm": 0.3447114825248718, "learning_rate": 4.851416277504749e-05, "loss": 0.0361, "step": 42720 }, { "grad_norm": 0.28948265314102173, "learning_rate": 4.8493503657838923e-05, "loss": 0.0201, "step": 42730 }, { "grad_norm": 0.166915163397789, "learning_rate": 4.84728447980494e-05, "loss": 0.0314, "step": 42740 }, { "grad_norm": 0.32826802134513855, "learning_rate": 4.8452186199208914e-05, "loss": 0.0317, "step": 42750 }, { "grad_norm": 0.339011013507843, "learning_rate": 4.843152786484749e-05, "loss": 0.0272, "step": 42760 }, { "grad_norm": 0.3309607207775116, "learning_rate": 4.8410869798495035e-05, "loss": 0.0232, "step": 42770 }, { "grad_norm": 0.25454869866371155, "learning_rate": 4.8390212003681486e-05, "loss": 0.0288, "step": 42780 }, { "grad_norm": 0.3782138228416443, "learning_rate": 4.836955448393667e-05, "loss": 0.028, "step": 42790 }, { "grad_norm": 0.2808328866958618, "learning_rate": 4.8348897242790406e-05, "loss": 0.0251, "step": 42800 }, { "grad_norm": 0.2719181180000305, "learning_rate": 4.832824028377243e-05, "loss": 0.0263, "step": 42810 }, { "grad_norm": 0.3392316997051239, "learning_rate": 4.830758361041249e-05, "loss": 0.0272, "step": 42820 }, { "grad_norm": 0.41504085063934326, "learning_rate": 4.828692722624022e-05, "loss": 0.0309, "step": 42830 }, { "grad_norm": 0.1999688744544983, "learning_rate": 4.826627113478522e-05, "loss": 0.0401, "step": 42840 }, { "grad_norm": 0.3644067049026489, "learning_rate": 4.824561533957708e-05, "loss": 0.0212, "step": 42850 }, { "grad_norm": 0.34271764755249023, "learning_rate": 4.822495984414527e-05, "loss": 0.026, "step": 42860 }, { "grad_norm": 0.3497777283191681, "learning_rate": 4.8204304652019304e-05, "loss": 0.023, "step": 42870 }, { "grad_norm": 0.2726843059062958, "learning_rate": 4.8183649766728537e-05, "loss": 0.0235, "step": 42880 }, { "grad_norm": 0.47242093086242676, "learning_rate": 4.816299519180237e-05, "loss": 0.0313, "step": 42890 }, { "grad_norm": 0.27528616786003113, "learning_rate": 4.8142340930770066e-05, "loss": 0.0293, "step": 42900 }, { "grad_norm": 0.29185163974761963, "learning_rate": 4.81216869871609e-05, "loss": 0.0244, "step": 42910 }, { "grad_norm": 0.26044490933418274, "learning_rate": 4.8101033364504064e-05, "loss": 0.0253, "step": 42920 }, { "grad_norm": 0.2532626688480377, "learning_rate": 4.808038006632866e-05, "loss": 0.0286, "step": 42930 }, { "grad_norm": 0.35790058970451355, "learning_rate": 4.805972709616383e-05, "loss": 0.0249, "step": 42940 }, { "grad_norm": 0.37486252188682556, "learning_rate": 4.8039074457538556e-05, "loss": 0.0244, "step": 42950 }, { "grad_norm": 0.2104116678237915, "learning_rate": 4.801842215398184e-05, "loss": 0.0294, "step": 42960 }, { "grad_norm": 0.39049798250198364, "learning_rate": 4.799777018902256e-05, "loss": 0.0239, "step": 42970 }, { "grad_norm": 0.25516510009765625, "learning_rate": 4.797711856618961e-05, "loss": 0.0209, "step": 42980 }, { "grad_norm": 0.35240381956100464, "learning_rate": 4.7956467289011765e-05, "loss": 0.0312, "step": 42990 }, { "grad_norm": 0.25880858302116394, "learning_rate": 4.793581636101778e-05, "loss": 0.0245, "step": 43000 }, { "grad_norm": 0.30970561504364014, "learning_rate": 4.7915165785736326e-05, "loss": 0.0274, "step": 43010 }, { "grad_norm": 0.3313915729522705, "learning_rate": 4.789451556669599e-05, "loss": 0.0253, "step": 43020 }, { "grad_norm": 0.33177700638771057, "learning_rate": 4.7873865707425384e-05, "loss": 0.0306, "step": 43030 }, { "grad_norm": 0.41357314586639404, "learning_rate": 4.7853216211452975e-05, "loss": 0.0267, "step": 43040 }, { "grad_norm": 0.40847116708755493, "learning_rate": 4.78325670823072e-05, "loss": 0.0257, "step": 43050 }, { "grad_norm": 0.2763146758079529, "learning_rate": 4.781191832351641e-05, "loss": 0.0259, "step": 43060 }, { "grad_norm": 0.3863812983036041, "learning_rate": 4.7791269938608965e-05, "loss": 0.0267, "step": 43070 }, { "grad_norm": 0.3069241940975189, "learning_rate": 4.777062193111305e-05, "loss": 0.0277, "step": 43080 }, { "grad_norm": 0.30790573358535767, "learning_rate": 4.77499743045569e-05, "loss": 0.0283, "step": 43090 }, { "grad_norm": 0.3469662368297577, "learning_rate": 4.7729327062468596e-05, "loss": 0.025, "step": 43100 }, { "grad_norm": 0.2589978575706482, "learning_rate": 4.770868020837619e-05, "loss": 0.0297, "step": 43110 }, { "grad_norm": 0.23964950442314148, "learning_rate": 4.768803374580768e-05, "loss": 0.0185, "step": 43120 }, { "grad_norm": 0.3469233214855194, "learning_rate": 4.766738767829094e-05, "loss": 0.0259, "step": 43130 }, { "grad_norm": 0.2832576632499695, "learning_rate": 4.764674200935388e-05, "loss": 0.0275, "step": 43140 }, { "grad_norm": 0.1947699338197708, "learning_rate": 4.762609674252424e-05, "loss": 0.0218, "step": 43150 }, { "grad_norm": 0.2561270296573639, "learning_rate": 4.760545188132974e-05, "loss": 0.0364, "step": 43160 }, { "grad_norm": 0.49580931663513184, "learning_rate": 4.758480742929801e-05, "loss": 0.0251, "step": 43170 }, { "grad_norm": 0.23876309394836426, "learning_rate": 4.756416338995664e-05, "loss": 0.0267, "step": 43180 }, { "grad_norm": 0.2545771300792694, "learning_rate": 4.7543519766833145e-05, "loss": 0.0194, "step": 43190 }, { "grad_norm": 0.17966583371162415, "learning_rate": 4.752287656345492e-05, "loss": 0.0258, "step": 43200 }, { "grad_norm": 0.2997973561286926, "learning_rate": 4.7502233783349355e-05, "loss": 0.0208, "step": 43210 }, { "grad_norm": 0.3560543656349182, "learning_rate": 4.7481591430043694e-05, "loss": 0.0234, "step": 43220 }, { "grad_norm": 0.29411938786506653, "learning_rate": 4.746094950706521e-05, "loss": 0.0287, "step": 43230 }, { "grad_norm": 0.3385934829711914, "learning_rate": 4.744030801794099e-05, "loss": 0.0262, "step": 43240 }, { "grad_norm": 0.21520888805389404, "learning_rate": 4.741966696619813e-05, "loss": 0.0325, "step": 43250 }, { "grad_norm": 0.3024258315563202, "learning_rate": 4.739902635536361e-05, "loss": 0.0273, "step": 43260 }, { "grad_norm": 0.23343151807785034, "learning_rate": 4.737838618896436e-05, "loss": 0.0245, "step": 43270 }, { "grad_norm": 0.1936550736427307, "learning_rate": 4.7357746470527203e-05, "loss": 0.0267, "step": 43280 }, { "grad_norm": 0.24413877725601196, "learning_rate": 4.73371072035789e-05, "loss": 0.0221, "step": 43290 }, { "grad_norm": 0.39997971057891846, "learning_rate": 4.731646839164616e-05, "loss": 0.0232, "step": 43300 }, { "grad_norm": 0.4588563144207001, "learning_rate": 4.729583003825556e-05, "loss": 0.0234, "step": 43310 }, { "grad_norm": 0.21478614211082458, "learning_rate": 4.727519214693367e-05, "loss": 0.0269, "step": 43320 }, { "grad_norm": 0.28315871953964233, "learning_rate": 4.725455472120689e-05, "loss": 0.024, "step": 43330 }, { "grad_norm": 0.34865063428878784, "learning_rate": 4.723391776460164e-05, "loss": 0.0272, "step": 43340 }, { "grad_norm": 0.26821234822273254, "learning_rate": 4.7213281280644186e-05, "loss": 0.0231, "step": 43350 }, { "grad_norm": 0.23513010144233704, "learning_rate": 4.719264527286075e-05, "loss": 0.0231, "step": 43360 }, { "grad_norm": 0.2221122682094574, "learning_rate": 4.717200974477744e-05, "loss": 0.0244, "step": 43370 }, { "grad_norm": 0.3799761235713959, "learning_rate": 4.715137469992034e-05, "loss": 0.0272, "step": 43380 }, { "grad_norm": 0.2739192247390747, "learning_rate": 4.713074014181539e-05, "loss": 0.0207, "step": 43390 }, { "grad_norm": 0.3347055912017822, "learning_rate": 4.711010607398846e-05, "loss": 0.0236, "step": 43400 }, { "grad_norm": 0.2936241030693054, "learning_rate": 4.708947249996538e-05, "loss": 0.021, "step": 43410 }, { "grad_norm": 0.3281097114086151, "learning_rate": 4.706883942327183e-05, "loss": 0.0237, "step": 43420 }, { "grad_norm": 0.31421327590942383, "learning_rate": 4.704820684743347e-05, "loss": 0.0257, "step": 43430 }, { "grad_norm": 0.29983848333358765, "learning_rate": 4.702757477597581e-05, "loss": 0.0276, "step": 43440 }, { "grad_norm": 0.2751658856868744, "learning_rate": 4.700694321242435e-05, "loss": 0.0247, "step": 43450 }, { "grad_norm": 0.36071309447288513, "learning_rate": 4.698631216030441e-05, "loss": 0.0255, "step": 43460 }, { "grad_norm": 0.17148002982139587, "learning_rate": 4.6965681623141314e-05, "loss": 0.023, "step": 43470 }, { "grad_norm": 0.23936940729618073, "learning_rate": 4.694505160446024e-05, "loss": 0.0237, "step": 43480 }, { "grad_norm": 0.28011542558670044, "learning_rate": 4.692442210778628e-05, "loss": 0.0252, "step": 43490 }, { "grad_norm": 0.426998496055603, "learning_rate": 4.690379313664448e-05, "loss": 0.0247, "step": 43500 }, { "grad_norm": 0.3375594913959503, "learning_rate": 4.688316469455973e-05, "loss": 0.0239, "step": 43510 }, { "grad_norm": 0.21717527508735657, "learning_rate": 4.6862536785056914e-05, "loss": 0.0233, "step": 43520 }, { "grad_norm": 0.21452657878398895, "learning_rate": 4.684190941166073e-05, "loss": 0.0254, "step": 43530 }, { "grad_norm": 0.2520662546157837, "learning_rate": 4.682128257789587e-05, "loss": 0.0237, "step": 43540 }, { "grad_norm": 0.34061700105667114, "learning_rate": 4.680065628728687e-05, "loss": 0.0233, "step": 43550 }, { "grad_norm": 0.32727715373039246, "learning_rate": 4.678003054335822e-05, "loss": 0.0282, "step": 43560 }, { "grad_norm": 0.4509339928627014, "learning_rate": 4.675940534963428e-05, "loss": 0.0282, "step": 43570 }, { "grad_norm": 0.3979341685771942, "learning_rate": 4.673878070963931e-05, "loss": 0.0271, "step": 43580 }, { "grad_norm": 0.19807995855808258, "learning_rate": 4.671815662689756e-05, "loss": 0.0156, "step": 43590 }, { "grad_norm": 0.325743168592453, "learning_rate": 4.669753310493306e-05, "loss": 0.0248, "step": 43600 }, { "grad_norm": 0.2405092418193817, "learning_rate": 4.6676910147269845e-05, "loss": 0.0262, "step": 43610 }, { "grad_norm": 0.4185771644115448, "learning_rate": 4.665628775743177e-05, "loss": 0.0282, "step": 43620 }, { "grad_norm": 0.20976507663726807, "learning_rate": 4.66356659389427e-05, "loss": 0.0209, "step": 43630 }, { "grad_norm": 0.25410711765289307, "learning_rate": 4.661504469532629e-05, "loss": 0.0202, "step": 43640 }, { "grad_norm": 0.2267414629459381, "learning_rate": 4.659442403010617e-05, "loss": 0.0225, "step": 43650 }, { "grad_norm": 0.34941181540489197, "learning_rate": 4.6573803946805845e-05, "loss": 0.0206, "step": 43660 }, { "grad_norm": 0.23137693107128143, "learning_rate": 4.6553184448948695e-05, "loss": 0.0271, "step": 43670 }, { "grad_norm": 0.272182434797287, "learning_rate": 4.653256554005807e-05, "loss": 0.0202, "step": 43680 }, { "grad_norm": 0.2776859998703003, "learning_rate": 4.6511947223657145e-05, "loss": 0.0318, "step": 43690 }, { "grad_norm": 0.314035564661026, "learning_rate": 4.649132950326906e-05, "loss": 0.0301, "step": 43700 }, { "grad_norm": 0.36126214265823364, "learning_rate": 4.647071238241679e-05, "loss": 0.0283, "step": 43710 }, { "grad_norm": 0.23467814922332764, "learning_rate": 4.6450095864623264e-05, "loss": 0.0231, "step": 43720 }, { "grad_norm": 0.3739486634731293, "learning_rate": 4.642947995341123e-05, "loss": 0.0231, "step": 43730 }, { "grad_norm": 0.28906241059303284, "learning_rate": 4.640886465230345e-05, "loss": 0.0291, "step": 43740 }, { "grad_norm": 0.3254108428955078, "learning_rate": 4.6388249964822485e-05, "loss": 0.0243, "step": 43750 }, { "grad_norm": 0.29781845211982727, "learning_rate": 4.63676358944908e-05, "loss": 0.022, "step": 43760 }, { "grad_norm": 0.5679448246955872, "learning_rate": 4.634702244483081e-05, "loss": 0.0291, "step": 43770 }, { "grad_norm": 0.5098491311073303, "learning_rate": 4.632640961936475e-05, "loss": 0.0263, "step": 43780 }, { "grad_norm": 0.385100781917572, "learning_rate": 4.6305797421614835e-05, "loss": 0.0231, "step": 43790 }, { "grad_norm": 0.22941671311855316, "learning_rate": 4.6285185855103105e-05, "loss": 0.0239, "step": 43800 }, { "grad_norm": 0.4871102273464203, "learning_rate": 4.626457492335151e-05, "loss": 0.0224, "step": 43810 }, { "grad_norm": 0.2142333686351776, "learning_rate": 4.624396462988188e-05, "loss": 0.02, "step": 43820 }, { "grad_norm": 0.2641635239124298, "learning_rate": 4.6223354978215985e-05, "loss": 0.0229, "step": 43830 }, { "grad_norm": 0.24134889245033264, "learning_rate": 4.620274597187544e-05, "loss": 0.0257, "step": 43840 }, { "grad_norm": 0.40622833371162415, "learning_rate": 4.6182137614381736e-05, "loss": 0.0253, "step": 43850 }, { "grad_norm": 0.38804563879966736, "learning_rate": 4.616152990925631e-05, "loss": 0.0221, "step": 43860 }, { "grad_norm": 0.3359515070915222, "learning_rate": 4.614092286002041e-05, "loss": 0.0255, "step": 43870 }, { "grad_norm": 0.23329368233680725, "learning_rate": 4.612031647019528e-05, "loss": 0.0264, "step": 43880 }, { "grad_norm": 0.2402563840150833, "learning_rate": 4.609971074330193e-05, "loss": 0.0251, "step": 43890 }, { "grad_norm": 0.3800068497657776, "learning_rate": 4.607910568286136e-05, "loss": 0.032, "step": 43900 }, { "grad_norm": 0.2874143719673157, "learning_rate": 4.605850129239438e-05, "loss": 0.0234, "step": 43910 }, { "grad_norm": 0.3339273929595947, "learning_rate": 4.603789757542174e-05, "loss": 0.0253, "step": 43920 }, { "grad_norm": 0.45912808179855347, "learning_rate": 4.601729453546404e-05, "loss": 0.0311, "step": 43930 }, { "grad_norm": 0.34894540905952454, "learning_rate": 4.599669217604177e-05, "loss": 0.0239, "step": 43940 }, { "grad_norm": 0.25462543964385986, "learning_rate": 4.597609050067532e-05, "loss": 0.0285, "step": 43950 }, { "grad_norm": 0.21048521995544434, "learning_rate": 4.595548951288495e-05, "loss": 0.027, "step": 43960 }, { "grad_norm": 0.4793723523616791, "learning_rate": 4.593488921619081e-05, "loss": 0.0302, "step": 43970 }, { "grad_norm": 0.3670549988746643, "learning_rate": 4.591428961411289e-05, "loss": 0.0284, "step": 43980 }, { "grad_norm": 0.4229987859725952, "learning_rate": 4.589369071017117e-05, "loss": 0.0346, "step": 43990 }, { "grad_norm": 0.2283364087343216, "learning_rate": 4.587309250788538e-05, "loss": 0.0309, "step": 44000 }, { "grad_norm": 0.31789377331733704, "learning_rate": 4.585249501077522e-05, "loss": 0.0221, "step": 44010 }, { "grad_norm": 0.2800077795982361, "learning_rate": 4.583189822236021e-05, "loss": 0.0273, "step": 44020 }, { "grad_norm": 0.3678238093852997, "learning_rate": 4.5811302146159816e-05, "loss": 0.027, "step": 44030 }, { "grad_norm": 0.27219557762145996, "learning_rate": 4.579070678569332e-05, "loss": 0.0252, "step": 44040 }, { "grad_norm": 0.3808377683162689, "learning_rate": 4.5770112144479904e-05, "loss": 0.0255, "step": 44050 }, { "grad_norm": 0.313752144575119, "learning_rate": 4.5749518226038645e-05, "loss": 0.0215, "step": 44060 }, { "grad_norm": 0.40472638607025146, "learning_rate": 4.572892503388845e-05, "loss": 0.0242, "step": 44070 }, { "grad_norm": 0.2843494415283203, "learning_rate": 4.570833257154817e-05, "loss": 0.0244, "step": 44080 }, { "grad_norm": 0.3684968650341034, "learning_rate": 4.568774084253646e-05, "loss": 0.0321, "step": 44090 }, { "grad_norm": 0.3493039309978485, "learning_rate": 4.566714985037191e-05, "loss": 0.0404, "step": 44100 }, { "grad_norm": 0.4908412992954254, "learning_rate": 4.564655959857295e-05, "loss": 0.0279, "step": 44110 }, { "grad_norm": 0.3175119459629059, "learning_rate": 4.5625970090657893e-05, "loss": 0.0246, "step": 44120 }, { "grad_norm": 0.31109821796417236, "learning_rate": 4.560538133014491e-05, "loss": 0.0236, "step": 44130 }, { "grad_norm": 0.3247232735157013, "learning_rate": 4.5584793320552055e-05, "loss": 0.0225, "step": 44140 }, { "grad_norm": 0.26521560549736023, "learning_rate": 4.556420606539728e-05, "loss": 0.024, "step": 44150 }, { "grad_norm": 0.2785126268863678, "learning_rate": 4.554361956819836e-05, "loss": 0.0223, "step": 44160 }, { "grad_norm": 0.24118202924728394, "learning_rate": 4.552303383247299e-05, "loss": 0.0218, "step": 44170 }, { "grad_norm": 0.2405754029750824, "learning_rate": 4.5502448861738664e-05, "loss": 0.0218, "step": 44180 }, { "grad_norm": 0.328885555267334, "learning_rate": 4.548186465951284e-05, "loss": 0.0284, "step": 44190 }, { "grad_norm": 0.26997843384742737, "learning_rate": 4.546128122931277e-05, "loss": 0.0251, "step": 44200 }, { "grad_norm": 0.29499706625938416, "learning_rate": 4.544069857465561e-05, "loss": 0.0232, "step": 44210 }, { "grad_norm": 0.34520143270492554, "learning_rate": 4.542011669905837e-05, "loss": 0.0208, "step": 44220 }, { "grad_norm": 0.31775540113449097, "learning_rate": 4.53995356060379e-05, "loss": 0.0188, "step": 44230 }, { "grad_norm": 0.295165091753006, "learning_rate": 4.5378955299110994e-05, "loss": 0.0194, "step": 44240 }, { "grad_norm": 0.3167988955974579, "learning_rate": 4.5358375781794233e-05, "loss": 0.0286, "step": 44250 }, { "grad_norm": 0.3913153409957886, "learning_rate": 4.533779705760411e-05, "loss": 0.0262, "step": 44260 }, { "grad_norm": 0.33885839581489563, "learning_rate": 4.5317219130056934e-05, "loss": 0.027, "step": 44270 }, { "grad_norm": 0.2102726250886917, "learning_rate": 4.5296642002668964e-05, "loss": 0.0295, "step": 44280 }, { "grad_norm": 0.30527693033218384, "learning_rate": 4.5276065678956216e-05, "loss": 0.0312, "step": 44290 }, { "grad_norm": 0.29008209705352783, "learning_rate": 4.525549016243466e-05, "loss": 0.0247, "step": 44300 }, { "grad_norm": 0.3848123550415039, "learning_rate": 4.523491545662008e-05, "loss": 0.0226, "step": 44310 }, { "grad_norm": 0.31892988085746765, "learning_rate": 4.5214341565028104e-05, "loss": 0.0282, "step": 44320 }, { "grad_norm": 0.2921539545059204, "learning_rate": 4.519376849117428e-05, "loss": 0.0249, "step": 44330 }, { "grad_norm": 0.21745634078979492, "learning_rate": 4.517319623857395e-05, "loss": 0.025, "step": 44340 }, { "grad_norm": 0.31576383113861084, "learning_rate": 4.5152624810742394e-05, "loss": 0.0267, "step": 44350 }, { "grad_norm": 0.1937670111656189, "learning_rate": 4.5132054211194664e-05, "loss": 0.0256, "step": 44360 }, { "grad_norm": 0.26496219635009766, "learning_rate": 4.511148444344574e-05, "loss": 0.0225, "step": 44370 }, { "grad_norm": 0.3786691427230835, "learning_rate": 4.509091551101041e-05, "loss": 0.0287, "step": 44380 }, { "grad_norm": 0.26358193159103394, "learning_rate": 4.507034741740338e-05, "loss": 0.024, "step": 44390 }, { "grad_norm": 0.3671339750289917, "learning_rate": 4.5049780166139145e-05, "loss": 0.0309, "step": 44400 }, { "grad_norm": 0.3700360953807831, "learning_rate": 4.5029213760732075e-05, "loss": 0.0398, "step": 44410 }, { "grad_norm": 0.21968580782413483, "learning_rate": 4.5008648204696434e-05, "loss": 0.0271, "step": 44420 }, { "grad_norm": 0.2519817352294922, "learning_rate": 4.4988083501546284e-05, "loss": 0.0216, "step": 44430 }, { "grad_norm": 0.3932206332683563, "learning_rate": 4.4967519654795606e-05, "loss": 0.024, "step": 44440 }, { "grad_norm": 0.317618191242218, "learning_rate": 4.494695666795816e-05, "loss": 0.0257, "step": 44450 }, { "grad_norm": 0.4540092945098877, "learning_rate": 4.4926394544547644e-05, "loss": 0.0254, "step": 44460 }, { "grad_norm": 0.3005790710449219, "learning_rate": 4.490583328807752e-05, "loss": 0.0304, "step": 44470 }, { "grad_norm": 0.22888635098934174, "learning_rate": 4.488527290206117e-05, "loss": 0.0254, "step": 44480 }, { "grad_norm": 0.25024694204330444, "learning_rate": 4.48647133900118e-05, "loss": 0.0205, "step": 44490 }, { "grad_norm": 0.21766290068626404, "learning_rate": 4.484415475544244e-05, "loss": 0.0277, "step": 44500 }, { "grad_norm": 0.2901817560195923, "learning_rate": 4.4823597001866035e-05, "loss": 0.0269, "step": 44510 }, { "grad_norm": 0.32047200202941895, "learning_rate": 4.480304013279532e-05, "loss": 0.0251, "step": 44520 }, { "grad_norm": 0.33207982778549194, "learning_rate": 4.478248415174292e-05, "loss": 0.0252, "step": 44530 }, { "grad_norm": 0.3029417097568512, "learning_rate": 4.476192906222126e-05, "loss": 0.0287, "step": 44540 }, { "grad_norm": 0.3694290220737457, "learning_rate": 4.474137486774268e-05, "loss": 0.0363, "step": 44550 }, { "grad_norm": 0.2928085923194885, "learning_rate": 4.4720821571819296e-05, "loss": 0.0248, "step": 44560 }, { "grad_norm": 0.43926751613616943, "learning_rate": 4.470026917796314e-05, "loss": 0.0351, "step": 44570 }, { "grad_norm": 0.36648836731910706, "learning_rate": 4.4679717689686005e-05, "loss": 0.0453, "step": 44580 }, { "grad_norm": 0.2738758623600006, "learning_rate": 4.465916711049963e-05, "loss": 0.026, "step": 44590 }, { "grad_norm": 0.43063464760780334, "learning_rate": 4.4638617443915524e-05, "loss": 0.026, "step": 44600 }, { "grad_norm": 0.43008580803871155, "learning_rate": 4.4618068693445055e-05, "loss": 0.0298, "step": 44610 }, { "grad_norm": 0.39966779947280884, "learning_rate": 4.459752086259946e-05, "loss": 0.0236, "step": 44620 }, { "grad_norm": 0.47977009415626526, "learning_rate": 4.457697395488977e-05, "loss": 0.0353, "step": 44630 }, { "grad_norm": 0.272236704826355, "learning_rate": 4.455642797382693e-05, "loss": 0.0273, "step": 44640 }, { "grad_norm": 0.33910420536994934, "learning_rate": 4.4535882922921643e-05, "loss": 0.0301, "step": 44650 }, { "grad_norm": 0.40408268570899963, "learning_rate": 4.451533880568455e-05, "loss": 0.0228, "step": 44660 }, { "grad_norm": 0.23396600782871246, "learning_rate": 4.449479562562603e-05, "loss": 0.0282, "step": 44670 }, { "grad_norm": 0.33557605743408203, "learning_rate": 4.447425338625639e-05, "loss": 0.0336, "step": 44680 }, { "grad_norm": 0.32639279961586, "learning_rate": 4.4453712091085705e-05, "loss": 0.0245, "step": 44690 }, { "grad_norm": 0.30517688393592834, "learning_rate": 4.443317174362392e-05, "loss": 0.0291, "step": 44700 }, { "grad_norm": 0.281877338886261, "learning_rate": 4.441263234738085e-05, "loss": 0.0216, "step": 44710 }, { "grad_norm": 0.19527214765548706, "learning_rate": 4.4392093905866086e-05, "loss": 0.0181, "step": 44720 }, { "grad_norm": 0.34514281153678894, "learning_rate": 4.437155642258911e-05, "loss": 0.023, "step": 44730 }, { "grad_norm": 0.22232429683208466, "learning_rate": 4.4351019901059177e-05, "loss": 0.0293, "step": 44740 }, { "grad_norm": 0.30729052424430847, "learning_rate": 4.433048434478547e-05, "loss": 0.0327, "step": 44750 }, { "grad_norm": 0.29776057600975037, "learning_rate": 4.4309949757276906e-05, "loss": 0.0333, "step": 44760 }, { "grad_norm": 0.2627825438976288, "learning_rate": 4.428941614204233e-05, "loss": 0.0306, "step": 44770 }, { "grad_norm": 0.2834220826625824, "learning_rate": 4.426888350259034e-05, "loss": 0.0281, "step": 44780 }, { "grad_norm": 0.2634991705417633, "learning_rate": 4.424835184242939e-05, "loss": 0.0394, "step": 44790 }, { "grad_norm": 0.28900906443595886, "learning_rate": 4.422782116506784e-05, "loss": 0.0224, "step": 44800 }, { "grad_norm": 0.4659944474697113, "learning_rate": 4.4207291474013754e-05, "loss": 0.0334, "step": 44810 }, { "grad_norm": 0.31797122955322266, "learning_rate": 4.418676277277514e-05, "loss": 0.0309, "step": 44820 }, { "grad_norm": 0.25521910190582275, "learning_rate": 4.416623506485976e-05, "loss": 0.0292, "step": 44830 }, { "grad_norm": 0.326088547706604, "learning_rate": 4.414570835377527e-05, "loss": 0.0233, "step": 44840 }, { "grad_norm": 0.3211456835269928, "learning_rate": 4.412518264302909e-05, "loss": 0.0242, "step": 44850 }, { "grad_norm": 0.3089480698108673, "learning_rate": 4.4104657936128535e-05, "loss": 0.0206, "step": 44860 }, { "grad_norm": 0.3771248757839203, "learning_rate": 4.408413423658071e-05, "loss": 0.0284, "step": 44870 }, { "grad_norm": 0.2681575417518616, "learning_rate": 4.406361154789253e-05, "loss": 0.0255, "step": 44880 }, { "grad_norm": 0.4068813621997833, "learning_rate": 4.4043089873570776e-05, "loss": 0.0262, "step": 44890 }, { "grad_norm": 0.3005160391330719, "learning_rate": 4.4022569217122025e-05, "loss": 0.0203, "step": 44900 }, { "grad_norm": 0.25361397862434387, "learning_rate": 4.400204958205274e-05, "loss": 0.0235, "step": 44910 }, { "grad_norm": 0.34936633706092834, "learning_rate": 4.3981530971869125e-05, "loss": 0.0339, "step": 44920 }, { "grad_norm": 0.23684464395046234, "learning_rate": 4.396101339007727e-05, "loss": 0.0261, "step": 44930 }, { "grad_norm": 0.25274842977523804, "learning_rate": 4.394049684018304e-05, "loss": 0.0208, "step": 44940 }, { "grad_norm": 0.3332567811012268, "learning_rate": 4.3919981325692186e-05, "loss": 0.0286, "step": 44950 }, { "grad_norm": 0.3469844460487366, "learning_rate": 4.389946685011024e-05, "loss": 0.0216, "step": 44960 }, { "grad_norm": 0.3677496016025543, "learning_rate": 4.387895341694255e-05, "loss": 0.0353, "step": 44970 }, { "grad_norm": 0.1911594420671463, "learning_rate": 4.3858441029694324e-05, "loss": 0.0298, "step": 44980 }, { "grad_norm": 0.2685864567756653, "learning_rate": 4.3837929691870527e-05, "loss": 0.0217, "step": 44990 }, { "grad_norm": 0.34948575496673584, "learning_rate": 4.381741940697604e-05, "loss": 0.0271, "step": 45000 }, { "grad_norm": 0.3439924120903015, "learning_rate": 4.379691017851547e-05, "loss": 0.0226, "step": 45010 }, { "grad_norm": 0.296705961227417, "learning_rate": 4.3776402009993304e-05, "loss": 0.0277, "step": 45020 }, { "grad_norm": 0.3051050901412964, "learning_rate": 4.3755894904913794e-05, "loss": 0.0226, "step": 45030 }, { "grad_norm": 0.30100974440574646, "learning_rate": 4.373538886678109e-05, "loss": 0.0231, "step": 45040 }, { "grad_norm": 0.35412487387657166, "learning_rate": 4.371488389909909e-05, "loss": 0.0209, "step": 45050 }, { "grad_norm": 0.27510473132133484, "learning_rate": 4.3694380005371515e-05, "loss": 0.0255, "step": 45060 }, { "grad_norm": 0.3331506848335266, "learning_rate": 4.367387718910196e-05, "loss": 0.0227, "step": 45070 }, { "grad_norm": 0.29006317257881165, "learning_rate": 4.3653375453793764e-05, "loss": 0.0271, "step": 45080 }, { "grad_norm": 0.3298453688621521, "learning_rate": 4.3632874802950136e-05, "loss": 0.0246, "step": 45090 }, { "grad_norm": 0.26931020617485046, "learning_rate": 4.3612375240074034e-05, "loss": 0.0253, "step": 45100 }, { "grad_norm": 0.28700003027915955, "learning_rate": 4.3591876768668325e-05, "loss": 0.0248, "step": 45110 }, { "grad_norm": 0.30975815653800964, "learning_rate": 4.3571379392235605e-05, "loss": 0.0283, "step": 45120 }, { "grad_norm": 0.3888063132762909, "learning_rate": 4.3550883114278335e-05, "loss": 0.028, "step": 45130 }, { "grad_norm": 0.2487630397081375, "learning_rate": 4.353038793829876e-05, "loss": 0.0247, "step": 45140 }, { "grad_norm": 0.32761627435684204, "learning_rate": 4.350989386779891e-05, "loss": 0.019, "step": 45150 }, { "grad_norm": 0.2457244098186493, "learning_rate": 4.3489400906280724e-05, "loss": 0.0323, "step": 45160 }, { "grad_norm": 0.4048705995082855, "learning_rate": 4.3468909057245845e-05, "loss": 0.024, "step": 45170 }, { "grad_norm": 0.22376546263694763, "learning_rate": 4.3448418324195794e-05, "loss": 0.0247, "step": 45180 }, { "grad_norm": 0.27211305499076843, "learning_rate": 4.342792871063184e-05, "loss": 0.025, "step": 45190 }, { "grad_norm": 0.22852762043476105, "learning_rate": 4.3407440220055145e-05, "loss": 0.0224, "step": 45200 }, { "grad_norm": 0.37260115146636963, "learning_rate": 4.33869528559666e-05, "loss": 0.025, "step": 45210 }, { "grad_norm": 0.3060579299926758, "learning_rate": 4.336646662186696e-05, "loss": 0.0187, "step": 45220 }, { "grad_norm": 0.24001148343086243, "learning_rate": 4.334598152125672e-05, "loss": 0.0239, "step": 45230 }, { "grad_norm": 0.3264702558517456, "learning_rate": 4.3325497557636276e-05, "loss": 0.0262, "step": 45240 }, { "grad_norm": 0.19878318905830383, "learning_rate": 4.330501473450574e-05, "loss": 0.0233, "step": 45250 }, { "grad_norm": 0.3701659142971039, "learning_rate": 4.328453305536507e-05, "loss": 0.021, "step": 45260 }, { "grad_norm": 0.28429752588272095, "learning_rate": 4.326405252371404e-05, "loss": 0.0231, "step": 45270 }, { "grad_norm": 0.288117378950119, "learning_rate": 4.324357314305221e-05, "loss": 0.0237, "step": 45280 }, { "grad_norm": 0.1920369565486908, "learning_rate": 4.3223094916878945e-05, "loss": 0.0214, "step": 45290 }, { "grad_norm": 0.2693002223968506, "learning_rate": 4.320261784869338e-05, "loss": 0.0282, "step": 45300 }, { "grad_norm": 0.29342716932296753, "learning_rate": 4.318214194199455e-05, "loss": 0.0253, "step": 45310 }, { "grad_norm": 0.24948996305465698, "learning_rate": 4.316166720028118e-05, "loss": 0.0195, "step": 45320 }, { "grad_norm": 0.4893255829811096, "learning_rate": 4.3141193627051864e-05, "loss": 0.0213, "step": 45330 }, { "grad_norm": 0.4208785891532898, "learning_rate": 4.312072122580496e-05, "loss": 0.0273, "step": 45340 }, { "grad_norm": 0.15521332621574402, "learning_rate": 4.3100250000038646e-05, "loss": 0.0241, "step": 45350 }, { "grad_norm": 0.35409167408943176, "learning_rate": 4.307977995325091e-05, "loss": 0.0324, "step": 45360 }, { "grad_norm": 0.3076947331428528, "learning_rate": 4.30593110889395e-05, "loss": 0.025, "step": 45370 }, { "grad_norm": 0.2820110023021698, "learning_rate": 4.3038843410602016e-05, "loss": 0.0202, "step": 45380 }, { "grad_norm": 0.20695021748542786, "learning_rate": 4.3018376921735774e-05, "loss": 0.0237, "step": 45390 }, { "grad_norm": 0.15835118293762207, "learning_rate": 4.299791162583799e-05, "loss": 0.0205, "step": 45400 }, { "grad_norm": 0.2530699074268341, "learning_rate": 4.29774475264056e-05, "loss": 0.0241, "step": 45410 }, { "grad_norm": 0.3531148135662079, "learning_rate": 4.2956984626935365e-05, "loss": 0.0198, "step": 45420 }, { "grad_norm": 0.2922425866127014, "learning_rate": 4.293652293092383e-05, "loss": 0.0226, "step": 45430 }, { "grad_norm": 0.35900574922561646, "learning_rate": 4.2916062441867324e-05, "loss": 0.0256, "step": 45440 }, { "grad_norm": 0.31971779465675354, "learning_rate": 4.289560316326201e-05, "loss": 0.0297, "step": 45450 }, { "grad_norm": 0.41828683018684387, "learning_rate": 4.28751450986038e-05, "loss": 0.0273, "step": 45460 }, { "grad_norm": 0.35646626353263855, "learning_rate": 4.2854688251388444e-05, "loss": 0.0255, "step": 45470 }, { "grad_norm": 0.23679235577583313, "learning_rate": 4.2834232625111425e-05, "loss": 0.0254, "step": 45480 }, { "grad_norm": 0.300664484500885, "learning_rate": 4.2813778223268086e-05, "loss": 0.0208, "step": 45490 }, { "grad_norm": 0.33869311213493347, "learning_rate": 4.2793325049353477e-05, "loss": 0.0267, "step": 45500 }, { "grad_norm": 0.30471256375312805, "learning_rate": 4.2772873106862535e-05, "loss": 0.0252, "step": 45510 }, { "grad_norm": 0.1788402646780014, "learning_rate": 4.275242239928993e-05, "loss": 0.0218, "step": 45520 }, { "grad_norm": 0.2271624505519867, "learning_rate": 4.273197293013009e-05, "loss": 0.0197, "step": 45530 }, { "grad_norm": 0.21330872178077698, "learning_rate": 4.271152470287731e-05, "loss": 0.0228, "step": 45540 }, { "grad_norm": 0.20243962109088898, "learning_rate": 4.2691077721025594e-05, "loss": 0.0236, "step": 45550 }, { "grad_norm": 0.22019201517105103, "learning_rate": 4.267063198806883e-05, "loss": 0.0286, "step": 45560 }, { "grad_norm": 0.31544041633605957, "learning_rate": 4.2650187507500574e-05, "loss": 0.023, "step": 45570 }, { "grad_norm": 0.4218234717845917, "learning_rate": 4.2629744282814275e-05, "loss": 0.0251, "step": 45580 }, { "grad_norm": 0.23974516987800598, "learning_rate": 4.2609302317503074e-05, "loss": 0.0191, "step": 45590 }, { "grad_norm": 0.3059488534927368, "learning_rate": 4.258886161505999e-05, "loss": 0.024, "step": 45600 }, { "grad_norm": 0.37886562943458557, "learning_rate": 4.2568422178977775e-05, "loss": 0.0266, "step": 45610 }, { "grad_norm": 0.3135715126991272, "learning_rate": 4.254798401274894e-05, "loss": 0.0304, "step": 45620 }, { "grad_norm": 0.3859933614730835, "learning_rate": 4.252754711986583e-05, "loss": 0.0245, "step": 45630 }, { "grad_norm": 0.24783499538898468, "learning_rate": 4.250711150382052e-05, "loss": 0.0233, "step": 45640 }, { "grad_norm": 0.3508247137069702, "learning_rate": 4.248667716810495e-05, "loss": 0.0238, "step": 45650 }, { "grad_norm": 0.21389062702655792, "learning_rate": 4.246624411621074e-05, "loss": 0.0201, "step": 45660 }, { "grad_norm": 0.36523008346557617, "learning_rate": 4.244581235162938e-05, "loss": 0.0267, "step": 45670 }, { "grad_norm": 0.2122746706008911, "learning_rate": 4.2425381877852075e-05, "loss": 0.0339, "step": 45680 }, { "grad_norm": 0.35891908407211304, "learning_rate": 4.2404952698369856e-05, "loss": 0.0256, "step": 45690 }, { "grad_norm": 0.28702449798583984, "learning_rate": 4.238452481667349e-05, "loss": 0.0247, "step": 45700 }, { "grad_norm": 0.22457139194011688, "learning_rate": 4.2364098236253526e-05, "loss": 0.0265, "step": 45710 }, { "grad_norm": 0.3721160292625427, "learning_rate": 4.2343672960600356e-05, "loss": 0.0238, "step": 45720 }, { "grad_norm": 0.2546898424625397, "learning_rate": 4.232324899320406e-05, "loss": 0.0232, "step": 45730 }, { "grad_norm": 0.29357796907424927, "learning_rate": 4.230282633755457e-05, "loss": 0.0254, "step": 45740 }, { "grad_norm": 0.24235741794109344, "learning_rate": 4.2282404997141515e-05, "loss": 0.0255, "step": 45750 }, { "grad_norm": 0.3069970905780792, "learning_rate": 4.2261984975454397e-05, "loss": 0.0287, "step": 45760 }, { "grad_norm": 0.3963543772697449, "learning_rate": 4.224156627598239e-05, "loss": 0.023, "step": 45770 }, { "grad_norm": 0.252528578042984, "learning_rate": 4.222114890221453e-05, "loss": 0.0222, "step": 45780 }, { "grad_norm": 0.37599462270736694, "learning_rate": 4.2200732857639546e-05, "loss": 0.0401, "step": 45790 }, { "grad_norm": 0.1702679693698883, "learning_rate": 4.2180318145746035e-05, "loss": 0.023, "step": 45800 }, { "grad_norm": 0.45095095038414, "learning_rate": 4.215990477002227e-05, "loss": 0.0327, "step": 45810 }, { "grad_norm": 0.3531598150730133, "learning_rate": 4.2139492733956356e-05, "loss": 0.0222, "step": 45820 }, { "grad_norm": 0.3167467713356018, "learning_rate": 4.211908204103615e-05, "loss": 0.0354, "step": 45830 }, { "grad_norm": 0.25951048731803894, "learning_rate": 4.2098672694749265e-05, "loss": 0.0356, "step": 45840 }, { "grad_norm": 0.4006842076778412, "learning_rate": 4.2078264698583133e-05, "loss": 0.0322, "step": 45850 }, { "grad_norm": 0.23418790102005005, "learning_rate": 4.205785805602488e-05, "loss": 0.0244, "step": 45860 }, { "grad_norm": 0.29125943779945374, "learning_rate": 4.203745277056149e-05, "loss": 0.0279, "step": 45870 }, { "grad_norm": 0.2872177064418793, "learning_rate": 4.201704884567964e-05, "loss": 0.0232, "step": 45880 }, { "grad_norm": 0.28904739022254944, "learning_rate": 4.1996646284865816e-05, "loss": 0.0325, "step": 45890 }, { "grad_norm": 0.29732850193977356, "learning_rate": 4.197624509160625e-05, "loss": 0.0212, "step": 45900 }, { "grad_norm": 0.4031480550765991, "learning_rate": 4.195584526938692e-05, "loss": 0.0229, "step": 45910 }, { "grad_norm": 0.24755842983722687, "learning_rate": 4.193544682169365e-05, "loss": 0.0216, "step": 45920 }, { "grad_norm": 0.43409600853919983, "learning_rate": 4.1915049752011946e-05, "loss": 0.0243, "step": 45930 }, { "grad_norm": 0.20357394218444824, "learning_rate": 4.189465406382712e-05, "loss": 0.0192, "step": 45940 }, { "grad_norm": 0.2604990005493164, "learning_rate": 4.187425976062422e-05, "loss": 0.026, "step": 45950 }, { "grad_norm": 0.262361615896225, "learning_rate": 4.18538668458881e-05, "loss": 0.0234, "step": 45960 }, { "grad_norm": 0.43027549982070923, "learning_rate": 4.183347532310333e-05, "loss": 0.0213, "step": 45970 }, { "grad_norm": 0.27107271552085876, "learning_rate": 4.181308519575429e-05, "loss": 0.0223, "step": 45980 }, { "grad_norm": 0.2614259421825409, "learning_rate": 4.179269646732507e-05, "loss": 0.0305, "step": 45990 }, { "grad_norm": 0.22082842886447906, "learning_rate": 4.177230914129954e-05, "loss": 0.0215, "step": 46000 }, { "grad_norm": 0.2973494529724121, "learning_rate": 4.175192322116136e-05, "loss": 0.023, "step": 46010 }, { "grad_norm": 0.28665831685066223, "learning_rate": 4.173153871039391e-05, "loss": 0.02, "step": 46020 }, { "grad_norm": 0.3678448498249054, "learning_rate": 4.171115561248036e-05, "loss": 0.0294, "step": 46030 }, { "grad_norm": 0.23745089769363403, "learning_rate": 4.16907739309036e-05, "loss": 0.0255, "step": 46040 }, { "grad_norm": 0.2573113441467285, "learning_rate": 4.167039366914633e-05, "loss": 0.02, "step": 46050 }, { "grad_norm": 0.2904004156589508, "learning_rate": 4.165001483069096e-05, "loss": 0.025, "step": 46060 }, { "grad_norm": 0.32303646206855774, "learning_rate": 4.16296374190197e-05, "loss": 0.0303, "step": 46070 }, { "grad_norm": 0.4620932638645172, "learning_rate": 4.1609261437614464e-05, "loss": 0.0313, "step": 46080 }, { "grad_norm": 0.4405098855495453, "learning_rate": 4.158888688995696e-05, "loss": 0.0285, "step": 46090 }, { "grad_norm": 0.32334673404693604, "learning_rate": 4.1568513779528645e-05, "loss": 0.0221, "step": 46100 }, { "grad_norm": 0.253244012594223, "learning_rate": 4.1548142109810704e-05, "loss": 0.0287, "step": 46110 }, { "grad_norm": 0.26475951075553894, "learning_rate": 4.152777188428414e-05, "loss": 0.0213, "step": 46120 }, { "grad_norm": 0.4328058660030365, "learning_rate": 4.1507403106429646e-05, "loss": 0.0228, "step": 46130 }, { "grad_norm": 0.264797180891037, "learning_rate": 4.148703577972768e-05, "loss": 0.0202, "step": 46140 }, { "grad_norm": 0.23695579171180725, "learning_rate": 4.146666990765846e-05, "loss": 0.0219, "step": 46150 }, { "grad_norm": 0.39827582240104675, "learning_rate": 4.1446305493701986e-05, "loss": 0.0227, "step": 46160 }, { "grad_norm": 0.2359229475259781, "learning_rate": 4.142594254133796e-05, "loss": 0.0273, "step": 46170 }, { "grad_norm": 0.31059059500694275, "learning_rate": 4.1405581054045835e-05, "loss": 0.0291, "step": 46180 }, { "grad_norm": 0.31923407316207886, "learning_rate": 4.1385221035304864e-05, "loss": 0.021, "step": 46190 }, { "grad_norm": 0.29144734144210815, "learning_rate": 4.136486248859398e-05, "loss": 0.0248, "step": 46200 }, { "grad_norm": 0.19397668540477753, "learning_rate": 4.1344505417391955e-05, "loss": 0.0257, "step": 46210 }, { "grad_norm": 0.26740762591362, "learning_rate": 4.132414982517721e-05, "loss": 0.0206, "step": 46220 }, { "grad_norm": 0.24467410147190094, "learning_rate": 4.130379571542798e-05, "loss": 0.0232, "step": 46230 }, { "grad_norm": 0.19235728681087494, "learning_rate": 4.12834430916222e-05, "loss": 0.0227, "step": 46240 }, { "grad_norm": 0.3316344618797302, "learning_rate": 4.126309195723763e-05, "loss": 0.0238, "step": 46250 }, { "grad_norm": 0.36015695333480835, "learning_rate": 4.124274231575168e-05, "loss": 0.0289, "step": 46260 }, { "grad_norm": 0.2750757932662964, "learning_rate": 4.122239417064154e-05, "loss": 0.0212, "step": 46270 }, { "grad_norm": 0.3603149950504303, "learning_rate": 4.1202047525384184e-05, "loss": 0.0229, "step": 46280 }, { "grad_norm": 0.2277197390794754, "learning_rate": 4.118170238345627e-05, "loss": 0.0195, "step": 46290 }, { "grad_norm": 0.19433332979679108, "learning_rate": 4.1161358748334256e-05, "loss": 0.0197, "step": 46300 }, { "grad_norm": 0.33856838941574097, "learning_rate": 4.1141016623494266e-05, "loss": 0.0197, "step": 46310 }, { "grad_norm": 0.31430238485336304, "learning_rate": 4.112067601241227e-05, "loss": 0.0199, "step": 46320 }, { "grad_norm": 0.27550825476646423, "learning_rate": 4.110033691856387e-05, "loss": 0.0212, "step": 46330 }, { "grad_norm": 0.3216664791107178, "learning_rate": 4.107999934542451e-05, "loss": 0.0264, "step": 46340 }, { "grad_norm": 0.2584909498691559, "learning_rate": 4.105966329646928e-05, "loss": 0.0234, "step": 46350 }, { "grad_norm": 0.19541442394256592, "learning_rate": 4.103932877517308e-05, "loss": 0.0208, "step": 46360 }, { "grad_norm": 0.3635447919368744, "learning_rate": 4.101899578501052e-05, "loss": 0.0188, "step": 46370 }, { "grad_norm": 0.2366589456796646, "learning_rate": 4.099866432945595e-05, "loss": 0.0215, "step": 46380 }, { "grad_norm": 0.4254608452320099, "learning_rate": 4.097833441198346e-05, "loss": 0.0242, "step": 46390 }, { "grad_norm": 0.16474510729312897, "learning_rate": 4.095800603606685e-05, "loss": 0.0182, "step": 46400 }, { "grad_norm": 0.3486086130142212, "learning_rate": 4.093767920517975e-05, "loss": 0.0263, "step": 46410 }, { "grad_norm": 0.26800402998924255, "learning_rate": 4.091735392279539e-05, "loss": 0.0224, "step": 46420 }, { "grad_norm": 0.16151639819145203, "learning_rate": 4.089703019238685e-05, "loss": 0.0191, "step": 46430 }, { "grad_norm": 0.32913681864738464, "learning_rate": 4.0876708017426866e-05, "loss": 0.0293, "step": 46440 }, { "grad_norm": 0.37912771105766296, "learning_rate": 4.085638740138798e-05, "loss": 0.0246, "step": 46450 }, { "grad_norm": 0.39029452204704285, "learning_rate": 4.08360683477424e-05, "loss": 0.0279, "step": 46460 }, { "grad_norm": 0.2710576057434082, "learning_rate": 4.0815750859962085e-05, "loss": 0.0281, "step": 46470 }, { "grad_norm": 0.28308966755867004, "learning_rate": 4.079543494151879e-05, "loss": 0.027, "step": 46480 }, { "grad_norm": 0.2966724932193756, "learning_rate": 4.07751205958839e-05, "loss": 0.0265, "step": 46490 }, { "grad_norm": 0.16805225610733032, "learning_rate": 4.0754807826528615e-05, "loss": 0.0237, "step": 46500 }, { "grad_norm": 0.3619931936264038, "learning_rate": 4.073449663692379e-05, "loss": 0.0273, "step": 46510 }, { "grad_norm": 0.2110588103532791, "learning_rate": 4.07141870305401e-05, "loss": 0.0207, "step": 46520 }, { "grad_norm": 0.31307080388069153, "learning_rate": 4.0693879010847866e-05, "loss": 0.0449, "step": 46530 }, { "grad_norm": 0.38216158747673035, "learning_rate": 4.06735725813172e-05, "loss": 0.0248, "step": 46540 }, { "grad_norm": 0.27152490615844727, "learning_rate": 4.0653267745417903e-05, "loss": 0.0173, "step": 46550 }, { "grad_norm": 0.25320175290107727, "learning_rate": 4.063296450661949e-05, "loss": 0.0187, "step": 46560 }, { "grad_norm": 0.29732546210289, "learning_rate": 4.061266286839128e-05, "loss": 0.0242, "step": 46570 }, { "grad_norm": 0.3208789527416229, "learning_rate": 4.0592362834202225e-05, "loss": 0.0339, "step": 46580 }, { "grad_norm": 0.2878996431827545, "learning_rate": 4.057206440752107e-05, "loss": 0.0182, "step": 46590 }, { "grad_norm": 0.27551236748695374, "learning_rate": 4.0551767591816245e-05, "loss": 0.0211, "step": 46600 }, { "grad_norm": 0.31940945982933044, "learning_rate": 4.0531472390555935e-05, "loss": 0.0256, "step": 46610 }, { "grad_norm": 0.3149506151676178, "learning_rate": 4.051117880720802e-05, "loss": 0.0254, "step": 46620 }, { "grad_norm": 0.2426106482744217, "learning_rate": 4.049088684524015e-05, "loss": 0.0192, "step": 46630 }, { "grad_norm": 0.2820647954940796, "learning_rate": 4.0470596508119636e-05, "loss": 0.0205, "step": 46640 }, { "grad_norm": 0.3115662634372711, "learning_rate": 4.0450307799313524e-05, "loss": 0.0279, "step": 46650 }, { "grad_norm": 0.37810078263282776, "learning_rate": 4.0430020722288656e-05, "loss": 0.0246, "step": 46660 }, { "grad_norm": 0.2792154550552368, "learning_rate": 4.040973528051148e-05, "loss": 0.0284, "step": 46670 }, { "grad_norm": 0.6870496273040771, "learning_rate": 4.038945147744827e-05, "loss": 0.0228, "step": 46680 }, { "grad_norm": 0.1936219036579132, "learning_rate": 4.0369169316564945e-05, "loss": 0.0198, "step": 46690 }, { "grad_norm": 0.3831855058670044, "learning_rate": 4.034888880132718e-05, "loss": 0.0243, "step": 46700 }, { "grad_norm": 0.4087212383747101, "learning_rate": 4.032860993520035e-05, "loss": 0.0227, "step": 46710 }, { "grad_norm": 0.2516622543334961, "learning_rate": 4.030833272164959e-05, "loss": 0.0174, "step": 46720 }, { "grad_norm": 0.603061854839325, "learning_rate": 4.0288057164139705e-05, "loss": 0.0332, "step": 46730 }, { "grad_norm": 0.25427159667015076, "learning_rate": 4.0267783266135205e-05, "loss": 0.0202, "step": 46740 }, { "grad_norm": 0.24526725709438324, "learning_rate": 4.024751103110039e-05, "loss": 0.0243, "step": 46750 }, { "grad_norm": 0.3042866885662079, "learning_rate": 4.0227240462499176e-05, "loss": 0.0265, "step": 46760 }, { "grad_norm": 0.1823091208934784, "learning_rate": 4.020697156379531e-05, "loss": 0.0281, "step": 46770 }, { "grad_norm": 0.34550848603248596, "learning_rate": 4.018670433845215e-05, "loss": 0.0278, "step": 46780 }, { "grad_norm": 0.24418224394321442, "learning_rate": 4.016643878993284e-05, "loss": 0.0184, "step": 46790 }, { "grad_norm": 0.2981255054473877, "learning_rate": 4.014617492170017e-05, "loss": 0.0302, "step": 46800 }, { "grad_norm": 0.2777955234050751, "learning_rate": 4.0125912737216726e-05, "loss": 0.0433, "step": 46810 }, { "grad_norm": 0.4495181441307068, "learning_rate": 4.0105652239944735e-05, "loss": 0.0266, "step": 46820 }, { "grad_norm": 0.35379937291145325, "learning_rate": 4.0085393433346144e-05, "loss": 0.0295, "step": 46830 }, { "grad_norm": 0.23831678926944733, "learning_rate": 4.006513632088268e-05, "loss": 0.0174, "step": 46840 }, { "grad_norm": 0.1711750328540802, "learning_rate": 4.004488090601567e-05, "loss": 0.0235, "step": 46850 }, { "grad_norm": 0.28778985142707825, "learning_rate": 4.002462719220626e-05, "loss": 0.0222, "step": 46860 }, { "grad_norm": 0.25204160809516907, "learning_rate": 4.000437518291522e-05, "loss": 0.0242, "step": 46870 }, { "grad_norm": 0.2859007716178894, "learning_rate": 3.9984124881603094e-05, "loss": 0.0243, "step": 46880 }, { "grad_norm": 0.18235918879508972, "learning_rate": 3.9963876291730086e-05, "loss": 0.0208, "step": 46890 }, { "grad_norm": 0.30621281266212463, "learning_rate": 3.994362941675614e-05, "loss": 0.0186, "step": 46900 }, { "grad_norm": 0.2111380249261856, "learning_rate": 3.992338426014088e-05, "loss": 0.0234, "step": 46910 }, { "grad_norm": 0.4059106707572937, "learning_rate": 3.9903140825343636e-05, "loss": 0.0257, "step": 46920 }, { "grad_norm": 0.3200792074203491, "learning_rate": 3.98828991158235e-05, "loss": 0.0213, "step": 46930 }, { "grad_norm": 0.25819045305252075, "learning_rate": 3.9862659135039185e-05, "loss": 0.0297, "step": 46940 }, { "grad_norm": 0.23403654992580414, "learning_rate": 3.984242088644918e-05, "loss": 0.0203, "step": 46950 }, { "grad_norm": 0.360723078250885, "learning_rate": 3.9822184373511615e-05, "loss": 0.0237, "step": 46960 }, { "grad_norm": 0.2820136845111847, "learning_rate": 3.98019495996844e-05, "loss": 0.026, "step": 46970 }, { "grad_norm": 0.26897892355918884, "learning_rate": 3.978171656842507e-05, "loss": 0.0222, "step": 46980 }, { "grad_norm": 0.2288092076778412, "learning_rate": 3.976148528319091e-05, "loss": 0.0204, "step": 46990 }, { "grad_norm": 0.2855401635169983, "learning_rate": 3.974125574743888e-05, "loss": 0.027, "step": 47000 }, { "grad_norm": 0.25720664858818054, "learning_rate": 3.9721027964625686e-05, "loss": 0.023, "step": 47010 }, { "grad_norm": 0.27024132013320923, "learning_rate": 3.9700801938207676e-05, "loss": 0.0241, "step": 47020 }, { "grad_norm": 0.30294033885002136, "learning_rate": 3.9680577671640916e-05, "loss": 0.0203, "step": 47030 }, { "grad_norm": 0.4499610960483551, "learning_rate": 3.966035516838121e-05, "loss": 0.031, "step": 47040 }, { "grad_norm": 0.27755433320999146, "learning_rate": 3.9640134431884014e-05, "loss": 0.0182, "step": 47050 }, { "grad_norm": 0.2846256196498871, "learning_rate": 3.961991546560451e-05, "loss": 0.0185, "step": 47060 }, { "grad_norm": 0.34044525027275085, "learning_rate": 3.959969827299753e-05, "loss": 0.0248, "step": 47070 }, { "grad_norm": 0.2792038023471832, "learning_rate": 3.9579482857517684e-05, "loss": 0.0259, "step": 47080 }, { "grad_norm": 0.2949799597263336, "learning_rate": 3.955926922261921e-05, "loss": 0.0234, "step": 47090 }, { "grad_norm": 0.19214527308940887, "learning_rate": 3.9539057371756084e-05, "loss": 0.0256, "step": 47100 }, { "grad_norm": 0.46027928590774536, "learning_rate": 3.951884730838195e-05, "loss": 0.0261, "step": 47110 }, { "grad_norm": 0.32386815547943115, "learning_rate": 3.949863903595012e-05, "loss": 0.0211, "step": 47120 }, { "grad_norm": 0.2356725037097931, "learning_rate": 3.947843255791369e-05, "loss": 0.0229, "step": 47130 }, { "grad_norm": 0.32086145877838135, "learning_rate": 3.9458227877725364e-05, "loss": 0.0265, "step": 47140 }, { "grad_norm": 0.4177538752555847, "learning_rate": 3.943802499883758e-05, "loss": 0.0253, "step": 47150 }, { "grad_norm": 0.3250780403614044, "learning_rate": 3.9417823924702437e-05, "loss": 0.0255, "step": 47160 }, { "grad_norm": 0.29231390357017517, "learning_rate": 3.939762465877178e-05, "loss": 0.0239, "step": 47170 }, { "grad_norm": 0.27611514925956726, "learning_rate": 3.937742720449708e-05, "loss": 0.0261, "step": 47180 }, { "grad_norm": 0.2902250289916992, "learning_rate": 3.9357231565329563e-05, "loss": 0.0221, "step": 47190 }, { "grad_norm": 0.32005420327186584, "learning_rate": 3.933703774472008e-05, "loss": 0.0212, "step": 47200 }, { "grad_norm": 0.32354307174682617, "learning_rate": 3.93168457461192e-05, "loss": 0.0184, "step": 47210 }, { "grad_norm": 0.2805355191230774, "learning_rate": 3.9296655572977216e-05, "loss": 0.0239, "step": 47220 }, { "grad_norm": 0.4478228986263275, "learning_rate": 3.927646722874404e-05, "loss": 0.0256, "step": 47230 }, { "grad_norm": 0.28884580731391907, "learning_rate": 3.925628071686934e-05, "loss": 0.0251, "step": 47240 }, { "grad_norm": 0.24222524464130402, "learning_rate": 3.9236096040802415e-05, "loss": 0.0235, "step": 47250 }, { "grad_norm": 0.41511037945747375, "learning_rate": 3.9215913203992294e-05, "loss": 0.0348, "step": 47260 }, { "grad_norm": 0.2333238422870636, "learning_rate": 3.9195732209887645e-05, "loss": 0.0204, "step": 47270 }, { "grad_norm": 0.25154685974121094, "learning_rate": 3.9175553061936875e-05, "loss": 0.0232, "step": 47280 }, { "grad_norm": 0.4015050232410431, "learning_rate": 3.9155375763588045e-05, "loss": 0.0289, "step": 47290 }, { "grad_norm": 0.289604514837265, "learning_rate": 3.913520031828889e-05, "loss": 0.0195, "step": 47300 }, { "grad_norm": 0.4471123218536377, "learning_rate": 3.911502672948685e-05, "loss": 0.0236, "step": 47310 }, { "grad_norm": 0.44820526242256165, "learning_rate": 3.9094855000629014e-05, "loss": 0.0235, "step": 47320 }, { "grad_norm": 0.4192374646663666, "learning_rate": 3.907468513516223e-05, "loss": 0.019, "step": 47330 }, { "grad_norm": 0.27201956510543823, "learning_rate": 3.905451713653294e-05, "loss": 0.0201, "step": 47340 }, { "grad_norm": 0.17869700491428375, "learning_rate": 3.903435100818731e-05, "loss": 0.0229, "step": 47350 }, { "grad_norm": 0.3520878255367279, "learning_rate": 3.901418675357117e-05, "loss": 0.0225, "step": 47360 }, { "grad_norm": 0.2516184151172638, "learning_rate": 3.8994024376130075e-05, "loss": 0.0337, "step": 47370 }, { "grad_norm": 0.28813135623931885, "learning_rate": 3.8973863879309194e-05, "loss": 0.021, "step": 47380 }, { "grad_norm": 0.26693105697631836, "learning_rate": 3.8953705266553394e-05, "loss": 0.0232, "step": 47390 }, { "grad_norm": 0.293021559715271, "learning_rate": 3.893354854130727e-05, "loss": 0.0247, "step": 47400 }, { "grad_norm": 0.324576199054718, "learning_rate": 3.8913393707015006e-05, "loss": 0.0216, "step": 47410 }, { "grad_norm": 0.35046514868736267, "learning_rate": 3.889324076712056e-05, "loss": 0.0308, "step": 47420 }, { "grad_norm": 0.3114398419857025, "learning_rate": 3.8873089725067476e-05, "loss": 0.0242, "step": 47430 }, { "grad_norm": 0.42658042907714844, "learning_rate": 3.885294058429905e-05, "loss": 0.0375, "step": 47440 }, { "grad_norm": 0.18129093945026398, "learning_rate": 3.8832793348258206e-05, "loss": 0.0216, "step": 47450 }, { "grad_norm": 0.2554892301559448, "learning_rate": 3.881264802038756e-05, "loss": 0.0402, "step": 47460 }, { "grad_norm": 0.4327127933502197, "learning_rate": 3.879250460412939e-05, "loss": 0.0305, "step": 47470 }, { "grad_norm": 0.3112451732158661, "learning_rate": 3.8772363102925644e-05, "loss": 0.0226, "step": 47480 }, { "grad_norm": 0.2645559012889862, "learning_rate": 3.875222352021798e-05, "loss": 0.0194, "step": 47490 }, { "grad_norm": 0.24572154879570007, "learning_rate": 3.8732085859447686e-05, "loss": 0.0244, "step": 47500 }, { "grad_norm": 0.2878628969192505, "learning_rate": 3.871195012405575e-05, "loss": 0.0209, "step": 47510 }, { "grad_norm": 0.34504637122154236, "learning_rate": 3.869181631748278e-05, "loss": 0.0246, "step": 47520 }, { "grad_norm": 0.36100584268569946, "learning_rate": 3.867168444316915e-05, "loss": 0.0217, "step": 47530 }, { "grad_norm": 0.1581612229347229, "learning_rate": 3.865155450455481e-05, "loss": 0.0208, "step": 47540 }, { "grad_norm": 0.3401126265525818, "learning_rate": 3.8631426505079426e-05, "loss": 0.0289, "step": 47550 }, { "grad_norm": 0.30159467458724976, "learning_rate": 3.8611300448182304e-05, "loss": 0.0349, "step": 47560 }, { "grad_norm": 0.3279382288455963, "learning_rate": 3.859117633730248e-05, "loss": 0.0247, "step": 47570 }, { "grad_norm": 0.3454087972640991, "learning_rate": 3.857105417587858e-05, "loss": 0.0267, "step": 47580 }, { "grad_norm": 0.2804475426673889, "learning_rate": 3.855093396734894e-05, "loss": 0.0208, "step": 47590 }, { "grad_norm": 0.2130003422498703, "learning_rate": 3.8530815715151545e-05, "loss": 0.0361, "step": 47600 }, { "grad_norm": 0.27666306495666504, "learning_rate": 3.851069942272405e-05, "loss": 0.0214, "step": 47610 }, { "grad_norm": 0.34866562485694885, "learning_rate": 3.849058509350382e-05, "loss": 0.0309, "step": 47620 }, { "grad_norm": 0.36038145422935486, "learning_rate": 3.8470472730927783e-05, "loss": 0.0316, "step": 47630 }, { "grad_norm": 0.19593441486358643, "learning_rate": 3.845036233843264e-05, "loss": 0.0211, "step": 47640 }, { "grad_norm": 0.34522438049316406, "learning_rate": 3.843025391945469e-05, "loss": 0.0247, "step": 47650 }, { "grad_norm": 0.34359365701675415, "learning_rate": 3.841014747742992e-05, "loss": 0.0236, "step": 47660 }, { "grad_norm": 0.34454765915870667, "learning_rate": 3.839004301579397e-05, "loss": 0.0203, "step": 47670 }, { "grad_norm": 0.3214361369609833, "learning_rate": 3.8369940537982097e-05, "loss": 0.0266, "step": 47680 }, { "grad_norm": 0.22720851004123688, "learning_rate": 3.834984004742933e-05, "loss": 0.0188, "step": 47690 }, { "grad_norm": 0.621728777885437, "learning_rate": 3.832974154757026e-05, "loss": 0.0184, "step": 47700 }, { "grad_norm": 0.44129058718681335, "learning_rate": 3.830964504183919e-05, "loss": 0.0203, "step": 47710 }, { "grad_norm": 0.2780434787273407, "learning_rate": 3.828955053367003e-05, "loss": 0.0278, "step": 47720 }, { "grad_norm": 0.2589358985424042, "learning_rate": 3.826945802649642e-05, "loss": 0.0237, "step": 47730 }, { "grad_norm": 0.2553653120994568, "learning_rate": 3.824936752375159e-05, "loss": 0.0287, "step": 47740 }, { "grad_norm": 0.3919108808040619, "learning_rate": 3.822927902886848e-05, "loss": 0.021, "step": 47750 }, { "grad_norm": 0.28431957960128784, "learning_rate": 3.8209192545279653e-05, "loss": 0.0205, "step": 47760 }, { "grad_norm": 0.22877784073352814, "learning_rate": 3.8189108076417326e-05, "loss": 0.0223, "step": 47770 }, { "grad_norm": 0.34667059779167175, "learning_rate": 3.816902562571342e-05, "loss": 0.0296, "step": 47780 }, { "grad_norm": 0.28025805950164795, "learning_rate": 3.814894519659944e-05, "loss": 0.0202, "step": 47790 }, { "grad_norm": 0.3098633885383606, "learning_rate": 3.812886679250661e-05, "loss": 0.0186, "step": 47800 }, { "grad_norm": 0.1408066302537918, "learning_rate": 3.810879041686575e-05, "loss": 0.018, "step": 47810 }, { "grad_norm": 0.38756850361824036, "learning_rate": 3.808871607310741e-05, "loss": 0.0292, "step": 47820 }, { "grad_norm": 0.360871285200119, "learning_rate": 3.806864376466169e-05, "loss": 0.026, "step": 47830 }, { "grad_norm": 0.26908695697784424, "learning_rate": 3.804857349495845e-05, "loss": 0.0207, "step": 47840 }, { "grad_norm": 0.3450258672237396, "learning_rate": 3.802850526742713e-05, "loss": 0.0427, "step": 47850 }, { "grad_norm": 0.24796725809574127, "learning_rate": 3.800843908549683e-05, "loss": 0.0209, "step": 47860 }, { "grad_norm": 0.25566017627716064, "learning_rate": 3.7988374952596325e-05, "loss": 0.0196, "step": 47870 }, { "grad_norm": 0.2901777923107147, "learning_rate": 3.7968312872154e-05, "loss": 0.0229, "step": 47880 }, { "grad_norm": 0.29651767015457153, "learning_rate": 3.7948252847597965e-05, "loss": 0.0271, "step": 47890 }, { "grad_norm": 0.22225825488567352, "learning_rate": 3.7928194882355885e-05, "loss": 0.0255, "step": 47900 }, { "grad_norm": 0.33674904704093933, "learning_rate": 3.790813897985515e-05, "loss": 0.0189, "step": 47910 }, { "grad_norm": 0.33057504892349243, "learning_rate": 3.7888085143522726e-05, "loss": 0.0318, "step": 47920 }, { "grad_norm": 0.27092376351356506, "learning_rate": 3.7868033376785314e-05, "loss": 0.0255, "step": 47930 }, { "grad_norm": 0.3240043818950653, "learning_rate": 3.784798368306919e-05, "loss": 0.0188, "step": 47940 }, { "grad_norm": 0.26793143153190613, "learning_rate": 3.782793606580029e-05, "loss": 0.023, "step": 47950 }, { "grad_norm": 0.320686399936676, "learning_rate": 3.7807890528404205e-05, "loss": 0.0306, "step": 47960 }, { "grad_norm": 0.3285270631313324, "learning_rate": 3.778784707430616e-05, "loss": 0.0266, "step": 47970 }, { "grad_norm": 0.35424473881721497, "learning_rate": 3.776780570693107e-05, "loss": 0.0207, "step": 47980 }, { "grad_norm": 0.19859811663627625, "learning_rate": 3.774776642970342e-05, "loss": 0.0297, "step": 47990 }, { "grad_norm": 0.23540645837783813, "learning_rate": 3.77277292460474e-05, "loss": 0.022, "step": 48000 }, { "grad_norm": 0.3437831401824951, "learning_rate": 3.770769415938678e-05, "loss": 0.0255, "step": 48010 }, { "grad_norm": 0.5829012989997864, "learning_rate": 3.768766117314506e-05, "loss": 0.0342, "step": 48020 }, { "grad_norm": 0.2860701084136963, "learning_rate": 3.76676302907453e-05, "loss": 0.0245, "step": 48030 }, { "grad_norm": 0.2792800962924957, "learning_rate": 3.764760151561021e-05, "loss": 0.026, "step": 48040 }, { "grad_norm": 0.3123692572116852, "learning_rate": 3.76275748511622e-05, "loss": 0.0222, "step": 48050 }, { "grad_norm": 0.31344369053840637, "learning_rate": 3.7607550300823255e-05, "loss": 0.0308, "step": 48060 }, { "grad_norm": 0.3917751610279083, "learning_rate": 3.7587527868015044e-05, "loss": 0.0341, "step": 48070 }, { "grad_norm": 0.28217992186546326, "learning_rate": 3.756750755615881e-05, "loss": 0.0246, "step": 48080 }, { "grad_norm": 0.20946045219898224, "learning_rate": 3.754748936867553e-05, "loss": 0.0259, "step": 48090 }, { "grad_norm": 0.3030908703804016, "learning_rate": 3.752747330898573e-05, "loss": 0.0251, "step": 48100 }, { "grad_norm": 0.2830674350261688, "learning_rate": 3.750745938050962e-05, "loss": 0.0281, "step": 48110 }, { "grad_norm": 0.32200026512145996, "learning_rate": 3.7487447586667025e-05, "loss": 0.0245, "step": 48120 }, { "grad_norm": 0.27185729146003723, "learning_rate": 3.7467437930877426e-05, "loss": 0.0225, "step": 48130 }, { "grad_norm": 0.2852543294429779, "learning_rate": 3.744743041655992e-05, "loss": 0.0234, "step": 48140 }, { "grad_norm": 0.3128868639469147, "learning_rate": 3.742742504713324e-05, "loss": 0.0232, "step": 48150 }, { "grad_norm": 0.2609853446483612, "learning_rate": 3.740742182601576e-05, "loss": 0.0233, "step": 48160 }, { "grad_norm": 0.4007067382335663, "learning_rate": 3.7387420756625464e-05, "loss": 0.0318, "step": 48170 }, { "grad_norm": 0.4094886779785156, "learning_rate": 3.736742184238002e-05, "loss": 0.0251, "step": 48180 }, { "grad_norm": 0.18940956890583038, "learning_rate": 3.7347425086696684e-05, "loss": 0.0259, "step": 48190 }, { "grad_norm": 0.34898945689201355, "learning_rate": 3.732743049299235e-05, "loss": 0.0273, "step": 48200 }, { "grad_norm": 0.33446797728538513, "learning_rate": 3.730743806468354e-05, "loss": 0.0228, "step": 48210 }, { "grad_norm": 0.31978824734687805, "learning_rate": 3.7287447805186436e-05, "loss": 0.0234, "step": 48220 }, { "grad_norm": 0.2578546702861786, "learning_rate": 3.726745971791682e-05, "loss": 0.0308, "step": 48230 }, { "grad_norm": 0.43733540177345276, "learning_rate": 3.724747380629008e-05, "loss": 0.0358, "step": 48240 }, { "grad_norm": 0.2615794539451599, "learning_rate": 3.72274900737213e-05, "loss": 0.0274, "step": 48250 }, { "grad_norm": 0.37345951795578003, "learning_rate": 3.7207508523625123e-05, "loss": 0.0258, "step": 48260 }, { "grad_norm": 0.24256311357021332, "learning_rate": 3.718752915941588e-05, "loss": 0.0213, "step": 48270 }, { "grad_norm": 0.3803572654724121, "learning_rate": 3.7167551984507464e-05, "loss": 0.024, "step": 48280 }, { "grad_norm": 0.29927778244018555, "learning_rate": 3.714757700231346e-05, "loss": 0.0211, "step": 48290 }, { "grad_norm": 0.3016252815723419, "learning_rate": 3.712760421624703e-05, "loss": 0.0211, "step": 48300 }, { "grad_norm": 0.293801486492157, "learning_rate": 3.710763362972099e-05, "loss": 0.026, "step": 48310 }, { "grad_norm": 0.43274229764938354, "learning_rate": 3.708766524614774e-05, "loss": 0.0247, "step": 48320 }, { "grad_norm": 0.21095111966133118, "learning_rate": 3.7067699068939335e-05, "loss": 0.0174, "step": 48330 }, { "grad_norm": 0.21484017372131348, "learning_rate": 3.704773510150748e-05, "loss": 0.0208, "step": 48340 }, { "grad_norm": 0.31976422667503357, "learning_rate": 3.702777334726344e-05, "loss": 0.0271, "step": 48350 }, { "grad_norm": 0.20234771072864532, "learning_rate": 3.7007813809618164e-05, "loss": 0.0172, "step": 48360 }, { "grad_norm": 0.27481982111930847, "learning_rate": 3.6987856491982145e-05, "loss": 0.0297, "step": 48370 }, { "grad_norm": 0.25408947467803955, "learning_rate": 3.696790139776558e-05, "loss": 0.0252, "step": 48380 }, { "grad_norm": 0.28169286251068115, "learning_rate": 3.6947948530378235e-05, "loss": 0.0238, "step": 48390 }, { "grad_norm": 0.3819150924682617, "learning_rate": 3.6927997893229516e-05, "loss": 0.0211, "step": 48400 }, { "grad_norm": 0.27112045884132385, "learning_rate": 3.690804948972845e-05, "loss": 0.0185, "step": 48410 }, { "grad_norm": 0.2115633636713028, "learning_rate": 3.688810332328363e-05, "loss": 0.0233, "step": 48420 }, { "grad_norm": 0.33931905031204224, "learning_rate": 3.686815939730336e-05, "loss": 0.0247, "step": 48430 }, { "grad_norm": 0.3319424092769623, "learning_rate": 3.684821771519548e-05, "loss": 0.0213, "step": 48440 }, { "grad_norm": 0.36225825548171997, "learning_rate": 3.68282782803675e-05, "loss": 0.0226, "step": 48450 }, { "grad_norm": 0.2568676173686981, "learning_rate": 3.6808341096226504e-05, "loss": 0.0271, "step": 48460 }, { "grad_norm": 0.3106599748134613, "learning_rate": 3.678840616617924e-05, "loss": 0.026, "step": 48470 }, { "grad_norm": 0.3357921242713928, "learning_rate": 3.6768473493632e-05, "loss": 0.0219, "step": 48480 }, { "grad_norm": 0.37809237837791443, "learning_rate": 3.6748543081990783e-05, "loss": 0.0229, "step": 48490 }, { "grad_norm": 0.31792375445365906, "learning_rate": 3.672861493466112e-05, "loss": 0.0224, "step": 48500 }, { "grad_norm": 0.344628244638443, "learning_rate": 3.670868905504818e-05, "loss": 0.0256, "step": 48510 }, { "grad_norm": 0.25681400299072266, "learning_rate": 3.6688765446556784e-05, "loss": 0.0216, "step": 48520 }, { "grad_norm": 0.2344694435596466, "learning_rate": 3.6668844112591276e-05, "loss": 0.0254, "step": 48530 }, { "grad_norm": 0.35666337609291077, "learning_rate": 3.664892505655573e-05, "loss": 0.0202, "step": 48540 }, { "grad_norm": 0.35512444376945496, "learning_rate": 3.662900828185373e-05, "loss": 0.026, "step": 48550 }, { "grad_norm": 0.23886728286743164, "learning_rate": 3.6609093791888516e-05, "loss": 0.0211, "step": 48560 }, { "grad_norm": 0.34450846910476685, "learning_rate": 3.658918159006292e-05, "loss": 0.0305, "step": 48570 }, { "grad_norm": 0.2944753170013428, "learning_rate": 3.656927167977942e-05, "loss": 0.0283, "step": 48580 }, { "grad_norm": 0.31921276450157166, "learning_rate": 3.654936406444006e-05, "loss": 0.027, "step": 48590 }, { "grad_norm": 0.2567547559738159, "learning_rate": 3.65294587474465e-05, "loss": 0.0207, "step": 48600 }, { "grad_norm": 0.26508867740631104, "learning_rate": 3.650955573220002e-05, "loss": 0.0485, "step": 48610 }, { "grad_norm": 0.34060341119766235, "learning_rate": 3.648965502210149e-05, "loss": 0.029, "step": 48620 }, { "grad_norm": 0.3326394557952881, "learning_rate": 3.646975662055142e-05, "loss": 0.0246, "step": 48630 }, { "grad_norm": 0.29724234342575073, "learning_rate": 3.644986053094987e-05, "loss": 0.024, "step": 48640 }, { "grad_norm": 0.25652897357940674, "learning_rate": 3.642996675669659e-05, "loss": 0.0224, "step": 48650 }, { "grad_norm": 0.21779242157936096, "learning_rate": 3.641007530119083e-05, "loss": 0.0203, "step": 48660 }, { "grad_norm": 0.27537956833839417, "learning_rate": 3.639018616783153e-05, "loss": 0.0218, "step": 48670 }, { "grad_norm": 0.26952511072158813, "learning_rate": 3.637029936001719e-05, "loss": 0.018, "step": 48680 }, { "grad_norm": 0.3341737389564514, "learning_rate": 3.6350414881145886e-05, "loss": 0.0285, "step": 48690 }, { "grad_norm": 0.2985546290874481, "learning_rate": 3.6330532734615386e-05, "loss": 0.0214, "step": 48700 }, { "grad_norm": 0.2261364758014679, "learning_rate": 3.6310652923822975e-05, "loss": 0.0275, "step": 48710 }, { "grad_norm": 0.31400343775749207, "learning_rate": 3.629077545216558e-05, "loss": 0.0244, "step": 48720 }, { "grad_norm": 0.3024771809577942, "learning_rate": 3.627090032303969e-05, "loss": 0.0278, "step": 48730 }, { "grad_norm": 0.18215546011924744, "learning_rate": 3.625102753984146e-05, "loss": 0.0294, "step": 48740 }, { "grad_norm": 0.2509731352329254, "learning_rate": 3.623115710596659e-05, "loss": 0.0219, "step": 48750 }, { "grad_norm": 0.2858276963233948, "learning_rate": 3.6211289024810395e-05, "loss": 0.0305, "step": 48760 }, { "grad_norm": 0.17937037348747253, "learning_rate": 3.619142329976777e-05, "loss": 0.0216, "step": 48770 }, { "grad_norm": 1.689718246459961, "learning_rate": 3.6171559934233247e-05, "loss": 0.0208, "step": 48780 }, { "grad_norm": 0.3700413703918457, "learning_rate": 3.615169893160093e-05, "loss": 0.0223, "step": 48790 }, { "grad_norm": 0.2912406325340271, "learning_rate": 3.61318402952645e-05, "loss": 0.0226, "step": 48800 }, { "grad_norm": 0.2861112952232361, "learning_rate": 3.6111984028617285e-05, "loss": 0.0229, "step": 48810 }, { "grad_norm": 0.2704501748085022, "learning_rate": 3.6092130135052134e-05, "loss": 0.0215, "step": 48820 }, { "grad_norm": 0.23458628356456757, "learning_rate": 3.6072278617961584e-05, "loss": 0.0262, "step": 48830 }, { "grad_norm": 0.34679460525512695, "learning_rate": 3.605242948073767e-05, "loss": 0.0263, "step": 48840 }, { "grad_norm": 0.24661332368850708, "learning_rate": 3.603258272677212e-05, "loss": 0.0243, "step": 48850 }, { "grad_norm": 0.27189937233924866, "learning_rate": 3.601273835945616e-05, "loss": 0.0254, "step": 48860 }, { "grad_norm": 0.3100932538509369, "learning_rate": 3.5992896382180664e-05, "loss": 0.0213, "step": 48870 }, { "grad_norm": 0.3731097877025604, "learning_rate": 3.597305679833609e-05, "loss": 0.0257, "step": 48880 }, { "grad_norm": 0.3031611740589142, "learning_rate": 3.595321961131245e-05, "loss": 0.021, "step": 48890 }, { "grad_norm": 0.3338579833507538, "learning_rate": 3.593338482449942e-05, "loss": 0.0306, "step": 48900 }, { "grad_norm": 0.27649185061454773, "learning_rate": 3.591355244128618e-05, "loss": 0.0333, "step": 48910 }, { "grad_norm": 0.37998202443122864, "learning_rate": 3.589372246506158e-05, "loss": 0.0244, "step": 48920 }, { "grad_norm": 0.18710799515247345, "learning_rate": 3.5873894899213984e-05, "loss": 0.022, "step": 48930 }, { "grad_norm": 0.2713533639907837, "learning_rate": 3.5854069747131416e-05, "loss": 0.0249, "step": 48940 }, { "grad_norm": 0.28441670536994934, "learning_rate": 3.583424701220143e-05, "loss": 0.0264, "step": 48950 }, { "grad_norm": 0.2553033232688904, "learning_rate": 3.581442669781121e-05, "loss": 0.0262, "step": 48960 }, { "grad_norm": 0.18992571532726288, "learning_rate": 3.579460880734749e-05, "loss": 0.0255, "step": 48970 }, { "grad_norm": 0.2592583894729614, "learning_rate": 3.577479334419657e-05, "loss": 0.0266, "step": 48980 }, { "grad_norm": 0.2263365089893341, "learning_rate": 3.575498031174444e-05, "loss": 0.0207, "step": 48990 }, { "grad_norm": 0.24984736740589142, "learning_rate": 3.573516971337657e-05, "loss": 0.024, "step": 49000 }, { "grad_norm": 0.39688053727149963, "learning_rate": 3.5715361552478046e-05, "loss": 0.0287, "step": 49010 }, { "grad_norm": 0.35526415705680847, "learning_rate": 3.5695555832433536e-05, "loss": 0.0223, "step": 49020 }, { "grad_norm": 0.49714627861976624, "learning_rate": 3.5675752556627325e-05, "loss": 0.0245, "step": 49030 }, { "grad_norm": 0.4163139760494232, "learning_rate": 3.565595172844322e-05, "loss": 0.0238, "step": 49040 }, { "grad_norm": 0.3151470124721527, "learning_rate": 3.5636153351264666e-05, "loss": 0.0256, "step": 49050 }, { "grad_norm": 0.3420717418193817, "learning_rate": 3.5616357428474655e-05, "loss": 0.0226, "step": 49060 }, { "grad_norm": 0.2560413181781769, "learning_rate": 3.559656396345575e-05, "loss": 0.0256, "step": 49070 }, { "grad_norm": 0.2742632329463959, "learning_rate": 3.5576772959590146e-05, "loss": 0.0252, "step": 49080 }, { "grad_norm": 0.394894540309906, "learning_rate": 3.5556984420259545e-05, "loss": 0.0287, "step": 49090 }, { "grad_norm": 0.23682454228401184, "learning_rate": 3.5537198348845305e-05, "loss": 0.0218, "step": 49100 }, { "grad_norm": 0.2476414442062378, "learning_rate": 3.551741474872831e-05, "loss": 0.029, "step": 49110 }, { "grad_norm": 0.3183618187904358, "learning_rate": 3.549763362328903e-05, "loss": 0.0279, "step": 49120 }, { "grad_norm": 0.3177550733089447, "learning_rate": 3.5477854975907515e-05, "loss": 0.0217, "step": 49130 }, { "grad_norm": 0.3468798100948334, "learning_rate": 3.5458078809963416e-05, "loss": 0.0267, "step": 49140 }, { "grad_norm": 0.255823016166687, "learning_rate": 3.543830512883594e-05, "loss": 0.0188, "step": 49150 }, { "grad_norm": 0.28846415877342224, "learning_rate": 3.5418533935903824e-05, "loss": 0.0233, "step": 49160 }, { "grad_norm": 0.4195103645324707, "learning_rate": 3.539876523454547e-05, "loss": 0.0221, "step": 49170 }, { "grad_norm": 0.24586355686187744, "learning_rate": 3.537899902813878e-05, "loss": 0.0213, "step": 49180 }, { "grad_norm": 0.22027094662189484, "learning_rate": 3.5359235320061293e-05, "loss": 0.025, "step": 49190 }, { "grad_norm": 0.21475133299827576, "learning_rate": 3.533947411369003e-05, "loss": 0.0232, "step": 49200 }, { "grad_norm": 0.2450375109910965, "learning_rate": 3.53197154124017e-05, "loss": 0.0222, "step": 49210 }, { "grad_norm": 0.3308071792125702, "learning_rate": 3.52999592195725e-05, "loss": 0.0228, "step": 49220 }, { "grad_norm": 0.35465577244758606, "learning_rate": 3.5280205538578224e-05, "loss": 0.021, "step": 49230 }, { "grad_norm": 0.39256930351257324, "learning_rate": 3.5260454372794236e-05, "loss": 0.0272, "step": 49240 }, { "grad_norm": 0.2453056424856186, "learning_rate": 3.524070572559545e-05, "loss": 0.0243, "step": 49250 }, { "grad_norm": 0.34758907556533813, "learning_rate": 3.5220959600356395e-05, "loss": 0.0231, "step": 49260 }, { "grad_norm": 0.20060931146144867, "learning_rate": 3.5201216000451145e-05, "loss": 0.019, "step": 49270 }, { "grad_norm": 0.2104896456003189, "learning_rate": 3.5181474929253335e-05, "loss": 0.0246, "step": 49280 }, { "grad_norm": 0.34011587500572205, "learning_rate": 3.516173639013615e-05, "loss": 0.0222, "step": 49290 }, { "grad_norm": 0.2598845660686493, "learning_rate": 3.5142000386472406e-05, "loss": 0.0219, "step": 49300 }, { "grad_norm": 0.19782337546348572, "learning_rate": 3.5122266921634427e-05, "loss": 0.0205, "step": 49310 }, { "grad_norm": 0.42507871985435486, "learning_rate": 3.510253599899413e-05, "loss": 0.0237, "step": 49320 }, { "grad_norm": 0.2605026662349701, "learning_rate": 3.5082807621922965e-05, "loss": 0.0199, "step": 49330 }, { "grad_norm": 0.43943434953689575, "learning_rate": 3.506308179379201e-05, "loss": 0.0269, "step": 49340 }, { "grad_norm": 0.2587142884731293, "learning_rate": 3.5043358517971844e-05, "loss": 0.0231, "step": 49350 }, { "grad_norm": 0.21060240268707275, "learning_rate": 3.502363779783264e-05, "loss": 0.0203, "step": 49360 }, { "grad_norm": 0.2962685227394104, "learning_rate": 3.500391963674415e-05, "loss": 0.0169, "step": 49370 }, { "grad_norm": 0.29016226530075073, "learning_rate": 3.4984204038075615e-05, "loss": 0.0256, "step": 49380 }, { "grad_norm": 0.18591052293777466, "learning_rate": 3.496449100519595e-05, "loss": 0.0222, "step": 49390 }, { "grad_norm": 0.27910247445106506, "learning_rate": 3.494478054147354e-05, "loss": 0.0248, "step": 49400 }, { "grad_norm": 0.31037333607673645, "learning_rate": 3.4925072650276395e-05, "loss": 0.0263, "step": 49410 }, { "grad_norm": 0.24366776645183563, "learning_rate": 3.4905367334972016e-05, "loss": 0.0222, "step": 49420 }, { "grad_norm": 0.3126356899738312, "learning_rate": 3.488566459892752e-05, "loss": 0.0185, "step": 49430 }, { "grad_norm": 0.3208194077014923, "learning_rate": 3.4865964445509585e-05, "loss": 0.0234, "step": 49440 }, { "grad_norm": 0.2517258822917938, "learning_rate": 3.484626687808438e-05, "loss": 0.0226, "step": 49450 }, { "grad_norm": 0.23991751670837402, "learning_rate": 3.4826571900017735e-05, "loss": 0.0174, "step": 49460 }, { "grad_norm": 0.2736490070819855, "learning_rate": 3.480687951467495e-05, "loss": 0.0172, "step": 49470 }, { "grad_norm": 0.42687004804611206, "learning_rate": 3.4787189725420925e-05, "loss": 0.0226, "step": 49480 }, { "grad_norm": 0.22753942012786865, "learning_rate": 3.4767502535620086e-05, "loss": 0.0189, "step": 49490 }, { "grad_norm": 0.5292224884033203, "learning_rate": 3.474781794863648e-05, "loss": 0.0365, "step": 49500 }, { "grad_norm": 0.27988865971565247, "learning_rate": 3.472813596783363e-05, "loss": 0.028, "step": 49510 }, { "grad_norm": 0.2775477468967438, "learning_rate": 3.470845659657466e-05, "loss": 0.0232, "step": 49520 }, { "grad_norm": 0.31065118312835693, "learning_rate": 3.468877983822223e-05, "loss": 0.0265, "step": 49530 }, { "grad_norm": 0.25561290979385376, "learning_rate": 3.466910569613855e-05, "loss": 0.0226, "step": 49540 }, { "grad_norm": 0.3114805221557617, "learning_rate": 3.464943417368542e-05, "loss": 0.028, "step": 49550 }, { "grad_norm": 0.2881017029285431, "learning_rate": 3.462976527422415e-05, "loss": 0.0176, "step": 49560 }, { "grad_norm": 0.3042401671409607, "learning_rate": 3.461009900111562e-05, "loss": 0.0225, "step": 49570 }, { "grad_norm": 0.4054849147796631, "learning_rate": 3.459043535772023e-05, "loss": 0.0281, "step": 49580 }, { "grad_norm": 0.3439994752407074, "learning_rate": 3.4570774347398014e-05, "loss": 0.025, "step": 49590 }, { "grad_norm": 0.3881033658981323, "learning_rate": 3.4551115973508454e-05, "loss": 0.0276, "step": 49600 }, { "grad_norm": 0.47282493114471436, "learning_rate": 3.453146023941066e-05, "loss": 0.0289, "step": 49610 }, { "grad_norm": 0.23326560854911804, "learning_rate": 3.451180714846325e-05, "loss": 0.0211, "step": 49620 }, { "grad_norm": 0.32472795248031616, "learning_rate": 3.449215670402438e-05, "loss": 0.0244, "step": 49630 }, { "grad_norm": 0.2834755778312683, "learning_rate": 3.447250890945181e-05, "loss": 0.022, "step": 49640 }, { "grad_norm": 0.23987354338169098, "learning_rate": 3.4452863768102754e-05, "loss": 0.0214, "step": 49650 }, { "grad_norm": 0.3872559070587158, "learning_rate": 3.443322128333409e-05, "loss": 0.0268, "step": 49660 }, { "grad_norm": 0.4698694944381714, "learning_rate": 3.441358145850215e-05, "loss": 0.026, "step": 49670 }, { "grad_norm": 0.2707425355911255, "learning_rate": 3.439394429696286e-05, "loss": 0.0216, "step": 49680 }, { "grad_norm": 0.23754802346229553, "learning_rate": 3.4374309802071644e-05, "loss": 0.0271, "step": 49690 }, { "grad_norm": 0.18491503596305847, "learning_rate": 3.435467797718353e-05, "loss": 0.0219, "step": 49700 }, { "grad_norm": 0.257535845041275, "learning_rate": 3.433504882565306e-05, "loss": 0.0209, "step": 49710 }, { "grad_norm": 0.3118817210197449, "learning_rate": 3.43154223508343e-05, "loss": 0.0265, "step": 49720 }, { "grad_norm": 0.22786293923854828, "learning_rate": 3.429579855608089e-05, "loss": 0.0199, "step": 49730 }, { "grad_norm": 0.24206507205963135, "learning_rate": 3.427617744474597e-05, "loss": 0.0217, "step": 49740 }, { "grad_norm": 0.28834307193756104, "learning_rate": 3.425655902018231e-05, "loss": 0.0269, "step": 49750 }, { "grad_norm": 0.20712946355342865, "learning_rate": 3.423694328574211e-05, "loss": 0.0187, "step": 49760 }, { "grad_norm": 0.20000804960727692, "learning_rate": 3.42173302447772e-05, "loss": 0.0201, "step": 49770 }, { "grad_norm": 0.23150086402893066, "learning_rate": 3.419771990063886e-05, "loss": 0.0189, "step": 49780 }, { "grad_norm": 0.2452390193939209, "learning_rate": 3.417811225667803e-05, "loss": 0.0193, "step": 49790 }, { "grad_norm": 0.3110176622867584, "learning_rate": 3.415850731624508e-05, "loss": 0.0213, "step": 49800 }, { "grad_norm": 0.2528907060623169, "learning_rate": 3.4138905082689945e-05, "loss": 0.0216, "step": 49810 }, { "grad_norm": 0.2620627284049988, "learning_rate": 3.4119305559362145e-05, "loss": 0.0239, "step": 49820 }, { "grad_norm": 0.28054696321487427, "learning_rate": 3.4099708749610684e-05, "loss": 0.019, "step": 49830 }, { "grad_norm": 0.32490846514701843, "learning_rate": 3.408011465678413e-05, "loss": 0.0261, "step": 49840 }, { "grad_norm": 0.36669066548347473, "learning_rate": 3.406052328423055e-05, "loss": 0.028, "step": 49850 }, { "grad_norm": 0.24983906745910645, "learning_rate": 3.4040934635297615e-05, "loss": 0.0228, "step": 49860 }, { "grad_norm": 0.27988457679748535, "learning_rate": 3.4021348713332466e-05, "loss": 0.0195, "step": 49870 }, { "grad_norm": 0.3658175766468048, "learning_rate": 3.4001765521681807e-05, "loss": 0.0239, "step": 49880 }, { "grad_norm": 0.21537911891937256, "learning_rate": 3.398218506369188e-05, "loss": 0.032, "step": 49890 }, { "grad_norm": 0.263365775346756, "learning_rate": 3.3962607342708404e-05, "loss": 0.0232, "step": 49900 }, { "grad_norm": 0.41478562355041504, "learning_rate": 3.394303236207673e-05, "loss": 0.0206, "step": 49910 }, { "grad_norm": 0.3124955892562866, "learning_rate": 3.392346012514166e-05, "loss": 0.0254, "step": 49920 }, { "grad_norm": 0.30996188521385193, "learning_rate": 3.390389063524757e-05, "loss": 0.0215, "step": 49930 }, { "grad_norm": 0.1767340451478958, "learning_rate": 3.3884323895738324e-05, "loss": 0.0228, "step": 49940 }, { "grad_norm": 0.37738102674484253, "learning_rate": 3.386475990995738e-05, "loss": 0.032, "step": 49950 }, { "grad_norm": 0.3414837121963501, "learning_rate": 3.384519868124765e-05, "loss": 0.0227, "step": 49960 }, { "grad_norm": 0.24742253124713898, "learning_rate": 3.3825640212951645e-05, "loss": 0.021, "step": 49970 }, { "grad_norm": 0.3251328766345978, "learning_rate": 3.380608450841134e-05, "loss": 0.0209, "step": 49980 }, { "grad_norm": 0.4447783827781677, "learning_rate": 3.3786531570968305e-05, "loss": 0.0223, "step": 49990 }, { "grad_norm": 0.2870200276374817, "learning_rate": 3.3766981403963584e-05, "loss": 0.0293, "step": 50000 }, { "grad_norm": 0.23829899728298187, "learning_rate": 3.374743401073775e-05, "loss": 0.0251, "step": 50010 }, { "grad_norm": 0.2971671521663666, "learning_rate": 3.372788939463095e-05, "loss": 0.0187, "step": 50020 }, { "grad_norm": 0.2540227770805359, "learning_rate": 3.370834755898281e-05, "loss": 0.0227, "step": 50030 }, { "grad_norm": 0.4053068459033966, "learning_rate": 3.3688808507132493e-05, "loss": 0.0246, "step": 50040 }, { "grad_norm": 0.25722524523735046, "learning_rate": 3.3669272242418685e-05, "loss": 0.0241, "step": 50050 }, { "grad_norm": 0.258108526468277, "learning_rate": 3.364973876817961e-05, "loss": 0.0201, "step": 50060 }, { "grad_norm": 0.2658497989177704, "learning_rate": 3.363020808775299e-05, "loss": 0.0182, "step": 50070 }, { "grad_norm": 0.1581139713525772, "learning_rate": 3.361068020447611e-05, "loss": 0.023, "step": 50080 }, { "grad_norm": 0.3933049142360687, "learning_rate": 3.3591155121685724e-05, "loss": 0.0221, "step": 50090 }, { "grad_norm": 0.328186959028244, "learning_rate": 3.357163284271814e-05, "loss": 0.0244, "step": 50100 }, { "grad_norm": 0.3311421871185303, "learning_rate": 3.355211337090919e-05, "loss": 0.0188, "step": 50110 }, { "grad_norm": 0.20223630964756012, "learning_rate": 3.353259670959421e-05, "loss": 0.0277, "step": 50120 }, { "grad_norm": 0.31816747784614563, "learning_rate": 3.351308286210808e-05, "loss": 0.0249, "step": 50130 }, { "grad_norm": 0.30647003650665283, "learning_rate": 3.3493571831785156e-05, "loss": 0.0232, "step": 50140 }, { "grad_norm": 0.35677942633628845, "learning_rate": 3.347406362195936e-05, "loss": 0.0225, "step": 50150 }, { "grad_norm": 0.2347257137298584, "learning_rate": 3.345455823596411e-05, "loss": 0.0225, "step": 50160 }, { "grad_norm": 0.19886821508407593, "learning_rate": 3.3435055677132346e-05, "loss": 0.0207, "step": 50170 }, { "grad_norm": 0.3641270101070404, "learning_rate": 3.3415555948796505e-05, "loss": 0.0227, "step": 50180 }, { "grad_norm": 0.3097756505012512, "learning_rate": 3.3396059054288556e-05, "loss": 0.0238, "step": 50190 }, { "grad_norm": 0.4374328851699829, "learning_rate": 3.3376564996940015e-05, "loss": 0.031, "step": 50200 }, { "grad_norm": 0.2015402615070343, "learning_rate": 3.3357073780081836e-05, "loss": 0.0211, "step": 50210 }, { "grad_norm": 0.23088648915290833, "learning_rate": 3.333758540704459e-05, "loss": 0.0204, "step": 50220 }, { "grad_norm": 0.2922285497188568, "learning_rate": 3.331809988115827e-05, "loss": 0.0182, "step": 50230 }, { "grad_norm": 0.27641403675079346, "learning_rate": 3.329861720575244e-05, "loss": 0.0204, "step": 50240 }, { "grad_norm": 0.2152974009513855, "learning_rate": 3.3279137384156126e-05, "loss": 0.0195, "step": 50250 }, { "grad_norm": 0.305027037858963, "learning_rate": 3.3259660419697934e-05, "loss": 0.0258, "step": 50260 }, { "grad_norm": 0.3589897155761719, "learning_rate": 3.3240186315705926e-05, "loss": 0.0257, "step": 50270 }, { "grad_norm": 0.37262219190597534, "learning_rate": 3.322071507550769e-05, "loss": 0.0418, "step": 50280 }, { "grad_norm": 0.23141781985759735, "learning_rate": 3.320124670243033e-05, "loss": 0.0242, "step": 50290 }, { "grad_norm": 0.16839388012886047, "learning_rate": 3.318178119980045e-05, "loss": 0.0198, "step": 50300 }, { "grad_norm": 0.3565589487552643, "learning_rate": 3.316231857094421e-05, "loss": 0.0236, "step": 50310 }, { "grad_norm": 0.39477697014808655, "learning_rate": 3.31428588191872e-05, "loss": 0.0237, "step": 50320 }, { "grad_norm": 0.37813809514045715, "learning_rate": 3.312340194785458e-05, "loss": 0.0265, "step": 50330 }, { "grad_norm": 0.28331297636032104, "learning_rate": 3.3103947960270975e-05, "loss": 0.0253, "step": 50340 }, { "grad_norm": 0.3180115222930908, "learning_rate": 3.308449685976058e-05, "loss": 0.0201, "step": 50350 }, { "grad_norm": 0.47336387634277344, "learning_rate": 3.3065048649647024e-05, "loss": 0.0198, "step": 50360 }, { "grad_norm": 0.32513830065727234, "learning_rate": 3.304560333325348e-05, "loss": 0.0195, "step": 50370 }, { "grad_norm": 0.40246453881263733, "learning_rate": 3.3026160913902624e-05, "loss": 0.025, "step": 50380 }, { "grad_norm": 0.43418681621551514, "learning_rate": 3.300672139491662e-05, "loss": 0.0331, "step": 50390 }, { "grad_norm": 0.3315282464027405, "learning_rate": 3.298728477961717e-05, "loss": 0.024, "step": 50400 }, { "grad_norm": 0.19408541917800903, "learning_rate": 3.296785107132544e-05, "loss": 0.0261, "step": 50410 }, { "grad_norm": 0.35181888937950134, "learning_rate": 3.2948420273362156e-05, "loss": 0.0165, "step": 50420 }, { "grad_norm": 0.2822467088699341, "learning_rate": 3.292899238904747e-05, "loss": 0.0247, "step": 50430 }, { "grad_norm": 0.16296406090259552, "learning_rate": 3.29095674217011e-05, "loss": 0.021, "step": 50440 }, { "grad_norm": 0.2287563979625702, "learning_rate": 3.289014537464224e-05, "loss": 0.0244, "step": 50450 }, { "grad_norm": 0.29123952984809875, "learning_rate": 3.287072625118955e-05, "loss": 0.0217, "step": 50460 }, { "grad_norm": 0.27279818058013916, "learning_rate": 3.285131005466129e-05, "loss": 0.0211, "step": 50470 }, { "grad_norm": 0.22288332879543304, "learning_rate": 3.2831896788375105e-05, "loss": 0.0154, "step": 50480 }, { "grad_norm": 0.22669167816638947, "learning_rate": 3.281248645564822e-05, "loss": 0.021, "step": 50490 }, { "grad_norm": 0.334644079208374, "learning_rate": 3.2793079059797306e-05, "loss": 0.0266, "step": 50500 }, { "grad_norm": 0.23065316677093506, "learning_rate": 3.277367460413859e-05, "loss": 0.0217, "step": 50510 }, { "grad_norm": 0.26435375213623047, "learning_rate": 3.275427309198773e-05, "loss": 0.028, "step": 50520 }, { "grad_norm": 0.2340727299451828, "learning_rate": 3.273487452665993e-05, "loss": 0.026, "step": 50530 }, { "grad_norm": 0.2518666386604309, "learning_rate": 3.271547891146986e-05, "loss": 0.0237, "step": 50540 }, { "grad_norm": 0.36178112030029297, "learning_rate": 3.269608624973173e-05, "loss": 0.0191, "step": 50550 }, { "grad_norm": 0.21673910319805145, "learning_rate": 3.26766965447592e-05, "loss": 0.0189, "step": 50560 }, { "grad_norm": 0.3032207190990448, "learning_rate": 3.2657309799865424e-05, "loss": 0.0257, "step": 50570 }, { "grad_norm": 0.24450628459453583, "learning_rate": 3.2637926018363084e-05, "loss": 0.0198, "step": 50580 }, { "grad_norm": 0.30319127440452576, "learning_rate": 3.261854520356432e-05, "loss": 0.0242, "step": 50590 }, { "grad_norm": 0.23698759078979492, "learning_rate": 3.2599167358780816e-05, "loss": 0.0188, "step": 50600 }, { "grad_norm": 0.18863654136657715, "learning_rate": 3.2579792487323676e-05, "loss": 0.0173, "step": 50610 }, { "grad_norm": 0.3427007496356964, "learning_rate": 3.256042059250358e-05, "loss": 0.0299, "step": 50620 }, { "grad_norm": 0.43859702348709106, "learning_rate": 3.254105167763062e-05, "loss": 0.0307, "step": 50630 }, { "grad_norm": 0.25258973240852356, "learning_rate": 3.252168574601443e-05, "loss": 0.0221, "step": 50640 }, { "grad_norm": 0.22563913464546204, "learning_rate": 3.2502322800964115e-05, "loss": 0.0182, "step": 50650 }, { "grad_norm": 0.45416170358657837, "learning_rate": 3.248296284578825e-05, "loss": 0.0219, "step": 50660 }, { "grad_norm": 0.31049856543540955, "learning_rate": 3.246360588379497e-05, "loss": 0.0242, "step": 50670 }, { "grad_norm": 0.27120211720466614, "learning_rate": 3.24442519182918e-05, "loss": 0.0277, "step": 50680 }, { "grad_norm": 0.3081126809120178, "learning_rate": 3.2424900952585845e-05, "loss": 0.0204, "step": 50690 }, { "grad_norm": 0.18495789170265198, "learning_rate": 3.240555298998362e-05, "loss": 0.0171, "step": 50700 }, { "grad_norm": 0.2555736303329468, "learning_rate": 3.2386208033791194e-05, "loss": 0.0181, "step": 50710 }, { "grad_norm": 0.34666404128074646, "learning_rate": 3.2366866087314064e-05, "loss": 0.0182, "step": 50720 }, { "grad_norm": 0.24039199948310852, "learning_rate": 3.234752715385727e-05, "loss": 0.0249, "step": 50730 }, { "grad_norm": 0.30349141359329224, "learning_rate": 3.232819123672529e-05, "loss": 0.0279, "step": 50740 }, { "grad_norm": 0.33770760893821716, "learning_rate": 3.2308858339222076e-05, "loss": 0.0294, "step": 50750 }, { "grad_norm": 0.2559765577316284, "learning_rate": 3.2289528464651144e-05, "loss": 0.0262, "step": 50760 }, { "grad_norm": 0.24187397956848145, "learning_rate": 3.22702016163154e-05, "loss": 0.0216, "step": 50770 }, { "grad_norm": 0.28182974457740784, "learning_rate": 3.2250877797517295e-05, "loss": 0.0222, "step": 50780 }, { "grad_norm": 0.21610800921916962, "learning_rate": 3.223155701155872e-05, "loss": 0.0233, "step": 50790 }, { "grad_norm": 0.3663833439350128, "learning_rate": 3.2212239261741105e-05, "loss": 0.0313, "step": 50800 }, { "grad_norm": 0.16906589269638062, "learning_rate": 3.219292455136528e-05, "loss": 0.0179, "step": 50810 }, { "grad_norm": 0.2104618102312088, "learning_rate": 3.217361288373165e-05, "loss": 0.0198, "step": 50820 }, { "grad_norm": 0.24052175879478455, "learning_rate": 3.215430426214002e-05, "loss": 0.0195, "step": 50830 }, { "grad_norm": 0.2586689293384552, "learning_rate": 3.2134998689889705e-05, "loss": 0.019, "step": 50840 }, { "grad_norm": 0.4760999083518982, "learning_rate": 3.211569617027952e-05, "loss": 0.0236, "step": 50850 }, { "grad_norm": 0.17826321721076965, "learning_rate": 3.209639670660769e-05, "loss": 0.0243, "step": 50860 }, { "grad_norm": 0.35667434334754944, "learning_rate": 3.207710030217202e-05, "loss": 0.0245, "step": 50870 }, { "grad_norm": 0.48142480850219727, "learning_rate": 3.205780696026972e-05, "loss": 0.025, "step": 50880 }, { "grad_norm": 0.3015507161617279, "learning_rate": 3.203851668419749e-05, "loss": 0.0213, "step": 50890 }, { "grad_norm": 0.25707998871803284, "learning_rate": 3.201922947725149e-05, "loss": 0.0232, "step": 50900 }, { "grad_norm": 0.2790391147136688, "learning_rate": 3.199994534272742e-05, "loss": 0.021, "step": 50910 }, { "grad_norm": 0.3216175436973572, "learning_rate": 3.1980664283920394e-05, "loss": 0.0207, "step": 50920 }, { "grad_norm": 0.36952969431877136, "learning_rate": 3.196138630412499e-05, "loss": 0.0223, "step": 50930 }, { "grad_norm": 0.2233770340681076, "learning_rate": 3.194211140663532e-05, "loss": 0.0169, "step": 50940 }, { "grad_norm": 0.32424214482307434, "learning_rate": 3.1922839594744914e-05, "loss": 0.0291, "step": 50950 }, { "grad_norm": 0.28256383538246155, "learning_rate": 3.190357087174683e-05, "loss": 0.0232, "step": 50960 }, { "grad_norm": 0.21287879347801208, "learning_rate": 3.188430524093353e-05, "loss": 0.0186, "step": 50970 }, { "grad_norm": 0.2605641186237335, "learning_rate": 3.1865042705597014e-05, "loss": 0.0226, "step": 50980 }, { "grad_norm": 0.30308690667152405, "learning_rate": 3.184578326902868e-05, "loss": 0.0205, "step": 50990 }, { "grad_norm": 0.24872365593910217, "learning_rate": 3.182652693451949e-05, "loss": 0.0248, "step": 51000 }, { "grad_norm": 0.3345320522785187, "learning_rate": 3.180727370535981e-05, "loss": 0.0208, "step": 51010 }, { "grad_norm": 0.40074440836906433, "learning_rate": 3.178802358483945e-05, "loss": 0.0221, "step": 51020 }, { "grad_norm": 0.28171756863594055, "learning_rate": 3.17687765762478e-05, "loss": 0.0228, "step": 51030 }, { "grad_norm": 0.27018505334854126, "learning_rate": 3.174953268287357e-05, "loss": 0.0295, "step": 51040 }, { "grad_norm": 0.4210221469402313, "learning_rate": 3.173029190800508e-05, "loss": 0.021, "step": 51050 }, { "grad_norm": 0.20023022592067719, "learning_rate": 3.171105425493e-05, "loss": 0.0272, "step": 51060 }, { "grad_norm": 0.25748416781425476, "learning_rate": 3.169181972693557e-05, "loss": 0.017, "step": 51070 }, { "grad_norm": 0.22575713694095612, "learning_rate": 3.1672588327308404e-05, "loss": 0.0197, "step": 51080 }, { "grad_norm": 0.37162017822265625, "learning_rate": 3.1653360059334646e-05, "loss": 0.0266, "step": 51090 }, { "grad_norm": 0.5259240865707397, "learning_rate": 3.163413492629985e-05, "loss": 0.0387, "step": 51100 }, { "grad_norm": 0.26434266567230225, "learning_rate": 3.161491293148912e-05, "loss": 0.0192, "step": 51110 }, { "grad_norm": 0.25001049041748047, "learning_rate": 3.1595694078186925e-05, "loss": 0.021, "step": 51120 }, { "grad_norm": 0.3989798128604889, "learning_rate": 3.157647836967724e-05, "loss": 0.0246, "step": 51130 }, { "grad_norm": 0.3675088584423065, "learning_rate": 3.155726580924354e-05, "loss": 0.0216, "step": 51140 }, { "grad_norm": 0.4026659429073334, "learning_rate": 3.1538056400168676e-05, "loss": 0.0303, "step": 51150 }, { "grad_norm": 0.17537111043930054, "learning_rate": 3.151885014573506e-05, "loss": 0.025, "step": 51160 }, { "grad_norm": 0.28063759207725525, "learning_rate": 3.149964704922448e-05, "loss": 0.0286, "step": 51170 }, { "grad_norm": 0.4575224816799164, "learning_rate": 3.148044711391825e-05, "loss": 0.0247, "step": 51180 }, { "grad_norm": 0.24716320633888245, "learning_rate": 3.146125034309707e-05, "loss": 0.0218, "step": 51190 }, { "grad_norm": 0.22794388234615326, "learning_rate": 3.1442056740041195e-05, "loss": 0.0256, "step": 51200 }, { "grad_norm": 0.3148103654384613, "learning_rate": 3.142286630803026e-05, "loss": 0.0349, "step": 51210 }, { "grad_norm": 0.26046687364578247, "learning_rate": 3.140367905034337e-05, "loss": 0.0237, "step": 51220 }, { "grad_norm": 0.1485769897699356, "learning_rate": 3.138449497025914e-05, "loss": 0.0211, "step": 51230 }, { "grad_norm": 0.2332075834274292, "learning_rate": 3.136531407105557e-05, "loss": 0.0222, "step": 51240 }, { "grad_norm": 0.2885095477104187, "learning_rate": 3.1346136356010184e-05, "loss": 0.0287, "step": 51250 }, { "grad_norm": 0.26952239871025085, "learning_rate": 3.132696182839988e-05, "loss": 0.0206, "step": 51260 }, { "grad_norm": 0.3359386622905731, "learning_rate": 3.130779049150111e-05, "loss": 0.0258, "step": 51270 }, { "grad_norm": 0.4670616090297699, "learning_rate": 3.128862234858971e-05, "loss": 0.0246, "step": 51280 }, { "grad_norm": 0.21080465614795685, "learning_rate": 3.1269457402941005e-05, "loss": 0.021, "step": 51290 }, { "grad_norm": 0.27561211585998535, "learning_rate": 3.125029565782974e-05, "loss": 0.0222, "step": 51300 }, { "grad_norm": 0.44605669379234314, "learning_rate": 3.123113711653013e-05, "loss": 0.0256, "step": 51310 }, { "grad_norm": 0.257636159658432, "learning_rate": 3.121198178231587e-05, "loss": 0.027, "step": 51320 }, { "grad_norm": 0.3340184688568115, "learning_rate": 3.119282965846006e-05, "loss": 0.0245, "step": 51330 }, { "grad_norm": 0.42774060368537903, "learning_rate": 3.1173680748235303e-05, "loss": 0.0285, "step": 51340 }, { "grad_norm": 0.1800614446401596, "learning_rate": 3.115453505491358e-05, "loss": 0.0234, "step": 51350 }, { "grad_norm": 0.14978191256523132, "learning_rate": 3.1135392581766404e-05, "loss": 0.0202, "step": 51360 }, { "grad_norm": 0.30525633692741394, "learning_rate": 3.111625333206467e-05, "loss": 0.0217, "step": 51370 }, { "grad_norm": 0.31780219078063965, "learning_rate": 3.109711730907881e-05, "loss": 0.0204, "step": 51380 }, { "grad_norm": 0.2678334712982178, "learning_rate": 3.107798451607859e-05, "loss": 0.0224, "step": 51390 }, { "grad_norm": 0.21670089662075043, "learning_rate": 3.105885495633328e-05, "loss": 0.0191, "step": 51400 }, { "grad_norm": 0.6844123005867004, "learning_rate": 3.103972863311163e-05, "loss": 0.0188, "step": 51410 }, { "grad_norm": 0.5090069770812988, "learning_rate": 3.102060554968178e-05, "loss": 0.0234, "step": 51420 }, { "grad_norm": 0.2653824985027313, "learning_rate": 3.100148570931137e-05, "loss": 0.0224, "step": 51430 }, { "grad_norm": 0.39338013529777527, "learning_rate": 3.098236911526744e-05, "loss": 0.0241, "step": 51440 }, { "grad_norm": 0.3100659251213074, "learning_rate": 3.0963255770816495e-05, "loss": 0.0184, "step": 51450 }, { "grad_norm": 0.24388548731803894, "learning_rate": 3.094414567922447e-05, "loss": 0.0195, "step": 51460 }, { "grad_norm": 0.23768199980258942, "learning_rate": 3.092503884375678e-05, "loss": 0.0235, "step": 51470 }, { "grad_norm": 0.3962836265563965, "learning_rate": 3.0905935267678255e-05, "loss": 0.0235, "step": 51480 }, { "grad_norm": 0.23898276686668396, "learning_rate": 3.0886834954253166e-05, "loss": 0.0204, "step": 51490 }, { "grad_norm": 0.3719974458217621, "learning_rate": 3.086773790674524e-05, "loss": 0.0242, "step": 51500 }, { "grad_norm": 0.23844848573207855, "learning_rate": 3.0848644128417617e-05, "loss": 0.0214, "step": 51510 }, { "grad_norm": 0.2962380647659302, "learning_rate": 3.082955362253294e-05, "loss": 0.0212, "step": 51520 }, { "grad_norm": 0.2527254819869995, "learning_rate": 3.081046639235323e-05, "loss": 0.0194, "step": 51530 }, { "grad_norm": 0.30306917428970337, "learning_rate": 3.0791382441139986e-05, "loss": 0.0161, "step": 51540 }, { "grad_norm": 0.2155856192111969, "learning_rate": 3.077230177215411e-05, "loss": 0.0287, "step": 51550 }, { "grad_norm": 0.1485523134469986, "learning_rate": 3.0753224388656e-05, "loss": 0.0168, "step": 51560 }, { "grad_norm": 0.2229122668504715, "learning_rate": 3.073415029390544e-05, "loss": 0.0251, "step": 51570 }, { "grad_norm": 0.2865997850894928, "learning_rate": 3.071507949116166e-05, "loss": 0.0187, "step": 51580 }, { "grad_norm": 0.24968193471431732, "learning_rate": 3.069601198368337e-05, "loss": 0.0201, "step": 51590 }, { "grad_norm": 0.23697233200073242, "learning_rate": 3.067694777472864e-05, "loss": 0.0198, "step": 51600 }, { "grad_norm": 0.24887807667255402, "learning_rate": 3.065788686755508e-05, "loss": 0.0193, "step": 51610 }, { "grad_norm": 0.36362630128860474, "learning_rate": 3.063882926541961e-05, "loss": 0.0224, "step": 51620 }, { "grad_norm": 0.2680646777153015, "learning_rate": 3.061977497157872e-05, "loss": 0.021, "step": 51630 }, { "grad_norm": 0.26864349842071533, "learning_rate": 3.0600723989288224e-05, "loss": 0.0187, "step": 51640 }, { "grad_norm": 0.29909634590148926, "learning_rate": 3.058167632180343e-05, "loss": 0.0267, "step": 51650 }, { "grad_norm": 0.27314579486846924, "learning_rate": 3.0562631972379074e-05, "loss": 0.0245, "step": 51660 }, { "grad_norm": 0.29358142614364624, "learning_rate": 3.0543590944269276e-05, "loss": 0.0224, "step": 51670 }, { "grad_norm": 0.40923428535461426, "learning_rate": 3.052455324072766e-05, "loss": 0.0322, "step": 51680 }, { "grad_norm": 0.15211810171604156, "learning_rate": 3.0505518865007244e-05, "loss": 0.0246, "step": 51690 }, { "grad_norm": 0.36326634883880615, "learning_rate": 3.048648782036048e-05, "loss": 0.02, "step": 51700 }, { "grad_norm": 0.2299046665430069, "learning_rate": 3.0467460110039236e-05, "loss": 0.02, "step": 51710 }, { "grad_norm": 0.2527828514575958, "learning_rate": 3.0448435737294866e-05, "loss": 0.0227, "step": 51720 }, { "grad_norm": 0.2955164909362793, "learning_rate": 3.042941470537808e-05, "loss": 0.0183, "step": 51730 }, { "grad_norm": 0.33227792382240295, "learning_rate": 3.0410397017539072e-05, "loss": 0.0211, "step": 51740 }, { "grad_norm": 0.32056868076324463, "learning_rate": 3.039138267702742e-05, "loss": 0.0192, "step": 51750 }, { "grad_norm": 0.23186714947223663, "learning_rate": 3.0372371687092193e-05, "loss": 0.0218, "step": 51760 }, { "grad_norm": 0.35201868414878845, "learning_rate": 3.0353364050981823e-05, "loss": 0.0217, "step": 51770 }, { "grad_norm": 0.23978883028030396, "learning_rate": 3.033435977194418e-05, "loss": 0.0247, "step": 51780 }, { "grad_norm": 0.27615752816200256, "learning_rate": 3.0315358853226618e-05, "loss": 0.0194, "step": 51790 }, { "grad_norm": 0.2104773074388504, "learning_rate": 3.0296361298075844e-05, "loss": 0.0188, "step": 51800 }, { "grad_norm": 0.26354533433914185, "learning_rate": 3.027736710973803e-05, "loss": 0.028, "step": 51810 }, { "grad_norm": 0.19725313782691956, "learning_rate": 3.025837629145875e-05, "loss": 0.0172, "step": 51820 }, { "grad_norm": 0.2802758514881134, "learning_rate": 3.0239388846483048e-05, "loss": 0.0176, "step": 51830 }, { "grad_norm": 0.1682496815919876, "learning_rate": 3.022040477805532e-05, "loss": 0.0237, "step": 51840 }, { "grad_norm": 0.35110044479370117, "learning_rate": 3.020142408941946e-05, "loss": 0.0179, "step": 51850 }, { "grad_norm": 0.27254170179367065, "learning_rate": 3.018244678381873e-05, "loss": 0.0215, "step": 51860 }, { "grad_norm": 0.3134009838104248, "learning_rate": 3.0163472864495812e-05, "loss": 0.0235, "step": 51870 }, { "grad_norm": 0.2628602385520935, "learning_rate": 3.014450233469287e-05, "loss": 0.0268, "step": 51880 }, { "grad_norm": 0.407168447971344, "learning_rate": 3.0125535197651422e-05, "loss": 0.0275, "step": 51890 }, { "grad_norm": 0.37462109327316284, "learning_rate": 3.0106571456612448e-05, "loss": 0.017, "step": 51900 }, { "grad_norm": 0.2863660752773285, "learning_rate": 3.0087611114816305e-05, "loss": 0.0293, "step": 51910 }, { "grad_norm": 0.2265452891588211, "learning_rate": 3.006865417550284e-05, "loss": 0.0212, "step": 51920 }, { "grad_norm": 0.3207246959209442, "learning_rate": 3.0049700641911242e-05, "loss": 0.0207, "step": 51930 }, { "grad_norm": 0.29001128673553467, "learning_rate": 3.0030750517280175e-05, "loss": 0.0231, "step": 51940 }, { "grad_norm": 0.4123289883136749, "learning_rate": 3.0011803804847682e-05, "loss": 0.0275, "step": 51950 }, { "grad_norm": 0.41102081537246704, "learning_rate": 2.9992860507851224e-05, "loss": 0.024, "step": 51960 }, { "grad_norm": 0.3623693287372589, "learning_rate": 2.9973920629527723e-05, "loss": 0.0186, "step": 51970 }, { "grad_norm": 0.21030092239379883, "learning_rate": 2.9954984173113453e-05, "loss": 0.0205, "step": 51980 }, { "grad_norm": 0.42235997319221497, "learning_rate": 2.993605114184418e-05, "loss": 0.0259, "step": 51990 }, { "grad_norm": 0.27850377559661865, "learning_rate": 2.9917121538955005e-05, "loss": 0.0258, "step": 52000 }, { "grad_norm": 0.40086349844932556, "learning_rate": 2.98981953676805e-05, "loss": 0.0336, "step": 52010 }, { "grad_norm": 0.20025193691253662, "learning_rate": 2.9879272631254594e-05, "loss": 0.0202, "step": 52020 }, { "grad_norm": 0.2789117097854614, "learning_rate": 2.986035333291073e-05, "loss": 0.023, "step": 52030 }, { "grad_norm": 0.2641284465789795, "learning_rate": 2.9841437475881646e-05, "loss": 0.0289, "step": 52040 }, { "grad_norm": 0.3521985411643982, "learning_rate": 2.9822525063399555e-05, "loss": 0.0216, "step": 52050 }, { "grad_norm": 0.24073202908039093, "learning_rate": 2.9803616098696087e-05, "loss": 0.0226, "step": 52060 }, { "grad_norm": 0.2933636009693146, "learning_rate": 2.9784710585002228e-05, "loss": 0.0299, "step": 52070 }, { "grad_norm": 0.4885231554508209, "learning_rate": 2.9765808525548467e-05, "loss": 0.0223, "step": 52080 }, { "grad_norm": 0.24598008394241333, "learning_rate": 2.974690992356461e-05, "loss": 0.0263, "step": 52090 }, { "grad_norm": 0.15074601769447327, "learning_rate": 2.9728014782279934e-05, "loss": 0.0269, "step": 52100 }, { "grad_norm": 0.27842533588409424, "learning_rate": 2.970912310492307e-05, "loss": 0.0241, "step": 52110 }, { "grad_norm": 0.22500857710838318, "learning_rate": 2.9690234894722134e-05, "loss": 0.0241, "step": 52120 }, { "grad_norm": 0.1955108493566513, "learning_rate": 2.9671350154904577e-05, "loss": 0.0307, "step": 52130 }, { "grad_norm": 0.32437047362327576, "learning_rate": 2.9652468888697282e-05, "loss": 0.0275, "step": 52140 }, { "grad_norm": 0.2837466895580292, "learning_rate": 2.9633591099326562e-05, "loss": 0.0214, "step": 52150 }, { "grad_norm": 0.4648064076900482, "learning_rate": 2.9614716790018072e-05, "loss": 0.0204, "step": 52160 }, { "grad_norm": 0.22383587062358856, "learning_rate": 2.959584596399697e-05, "loss": 0.0183, "step": 52170 }, { "grad_norm": 0.3232772648334503, "learning_rate": 2.9576978624487717e-05, "loss": 0.0315, "step": 52180 }, { "grad_norm": 0.5211901068687439, "learning_rate": 2.9558114774714263e-05, "loss": 0.0251, "step": 52190 }, { "grad_norm": 0.2699965834617615, "learning_rate": 2.9539254417899897e-05, "loss": 0.0261, "step": 52200 }, { "grad_norm": 0.2495141327381134, "learning_rate": 2.9520397557267365e-05, "loss": 0.0212, "step": 52210 }, { "grad_norm": 0.3220495581626892, "learning_rate": 2.9501544196038765e-05, "loss": 0.0243, "step": 52220 }, { "grad_norm": 0.1753779500722885, "learning_rate": 2.9482694337435614e-05, "loss": 0.0302, "step": 52230 }, { "grad_norm": 0.24284346401691437, "learning_rate": 2.946384798467887e-05, "loss": 0.024, "step": 52240 }, { "grad_norm": 0.2516273558139801, "learning_rate": 2.9445005140988824e-05, "loss": 0.0137, "step": 52250 }, { "grad_norm": 0.2949960231781006, "learning_rate": 2.942616580958524e-05, "loss": 0.0199, "step": 52260 }, { "grad_norm": 0.2787572145462036, "learning_rate": 2.9407329993687193e-05, "loss": 0.0198, "step": 52270 }, { "grad_norm": 0.29498037695884705, "learning_rate": 2.938849769651326e-05, "loss": 0.0187, "step": 52280 }, { "grad_norm": 0.299285352230072, "learning_rate": 2.9369668921281345e-05, "loss": 0.0232, "step": 52290 }, { "grad_norm": 0.2866487503051758, "learning_rate": 2.9350843671208773e-05, "loss": 0.0201, "step": 52300 }, { "grad_norm": 0.33028239011764526, "learning_rate": 2.9332021949512245e-05, "loss": 0.0221, "step": 52310 }, { "grad_norm": 0.5361647009849548, "learning_rate": 2.9313203759407908e-05, "loss": 0.0253, "step": 52320 }, { "grad_norm": 0.28123611211776733, "learning_rate": 2.929438910411127e-05, "loss": 0.0352, "step": 52330 }, { "grad_norm": 0.36752015352249146, "learning_rate": 2.927557798683722e-05, "loss": 0.0268, "step": 52340 }, { "grad_norm": 0.29126787185668945, "learning_rate": 2.9256770410800095e-05, "loss": 0.0203, "step": 52350 }, { "grad_norm": 0.2518174350261688, "learning_rate": 2.9237966379213554e-05, "loss": 0.026, "step": 52360 }, { "grad_norm": 0.2478351593017578, "learning_rate": 2.9219165895290736e-05, "loss": 0.0251, "step": 52370 }, { "grad_norm": 0.3136651813983917, "learning_rate": 2.9200368962244083e-05, "loss": 0.0248, "step": 52380 }, { "grad_norm": 0.36509233713150024, "learning_rate": 2.9181575583285536e-05, "loss": 0.0208, "step": 52390 }, { "grad_norm": 0.3826802968978882, "learning_rate": 2.916278576162632e-05, "loss": 0.0238, "step": 52400 }, { "grad_norm": 0.2487131655216217, "learning_rate": 2.9143999500477116e-05, "loss": 0.0201, "step": 52410 }, { "grad_norm": 0.24744918942451477, "learning_rate": 2.9125216803048004e-05, "loss": 0.0252, "step": 52420 }, { "grad_norm": 0.2551926076412201, "learning_rate": 2.9106437672548403e-05, "loss": 0.0216, "step": 52430 }, { "grad_norm": 0.18572860956192017, "learning_rate": 2.9087662112187154e-05, "loss": 0.0181, "step": 52440 }, { "grad_norm": 0.2475799173116684, "learning_rate": 2.9068890125172498e-05, "loss": 0.0186, "step": 52450 }, { "grad_norm": 0.22334955632686615, "learning_rate": 2.9050121714712058e-05, "loss": 0.0292, "step": 52460 }, { "grad_norm": 0.465059757232666, "learning_rate": 2.9031356884012835e-05, "loss": 0.0266, "step": 52470 }, { "grad_norm": 0.3627249002456665, "learning_rate": 2.9012595636281247e-05, "loss": 0.0237, "step": 52480 }, { "grad_norm": 0.21153800189495087, "learning_rate": 2.8993837974723016e-05, "loss": 0.0384, "step": 52490 }, { "grad_norm": 0.22309763729572296, "learning_rate": 2.8975083902543394e-05, "loss": 0.0195, "step": 52500 }, { "grad_norm": 0.23651038110256195, "learning_rate": 2.895633342294688e-05, "loss": 0.0242, "step": 52510 }, { "grad_norm": 0.2578047215938568, "learning_rate": 2.893758653913744e-05, "loss": 0.0185, "step": 52520 }, { "grad_norm": 0.2624593675136566, "learning_rate": 2.8918843254318423e-05, "loss": 0.0206, "step": 52530 }, { "grad_norm": 0.36563563346862793, "learning_rate": 2.8900103571692483e-05, "loss": 0.0205, "step": 52540 }, { "grad_norm": 0.25496363639831543, "learning_rate": 2.8881367494461796e-05, "loss": 0.0185, "step": 52550 }, { "grad_norm": 0.25084131956100464, "learning_rate": 2.886263502582779e-05, "loss": 0.022, "step": 52560 }, { "grad_norm": 0.2645547389984131, "learning_rate": 2.8843906168991353e-05, "loss": 0.0212, "step": 52570 }, { "grad_norm": 0.2633552551269531, "learning_rate": 2.8825180927152728e-05, "loss": 0.0225, "step": 52580 }, { "grad_norm": 0.2683145999908447, "learning_rate": 2.8806459303511547e-05, "loss": 0.017, "step": 52590 }, { "grad_norm": 0.3491232991218567, "learning_rate": 2.8787741301266852e-05, "loss": 0.0196, "step": 52600 }, { "grad_norm": 0.28433018922805786, "learning_rate": 2.876902692361697e-05, "loss": 0.0181, "step": 52610 }, { "grad_norm": 0.30070042610168457, "learning_rate": 2.875031617375976e-05, "loss": 0.0175, "step": 52620 }, { "grad_norm": 0.2550864517688751, "learning_rate": 2.8731609054892318e-05, "loss": 0.016, "step": 52630 }, { "grad_norm": 0.23751211166381836, "learning_rate": 2.871290557021119e-05, "loss": 0.0174, "step": 52640 }, { "grad_norm": 0.1688743382692337, "learning_rate": 2.8694205722912298e-05, "loss": 0.0235, "step": 52650 }, { "grad_norm": 0.2756563425064087, "learning_rate": 2.8675509516190936e-05, "loss": 0.0227, "step": 52660 }, { "grad_norm": 0.23226523399353027, "learning_rate": 2.8656816953241772e-05, "loss": 0.0231, "step": 52670 }, { "grad_norm": 0.3148583471775055, "learning_rate": 2.863812803725887e-05, "loss": 0.0214, "step": 52680 }, { "grad_norm": 0.2641846537590027, "learning_rate": 2.8619442771435623e-05, "loss": 0.0166, "step": 52690 }, { "grad_norm": 0.2701467275619507, "learning_rate": 2.860076115896484e-05, "loss": 0.0269, "step": 52700 }, { "grad_norm": 0.30042147636413574, "learning_rate": 2.8582083203038712e-05, "loss": 0.0235, "step": 52710 }, { "grad_norm": 0.290576308965683, "learning_rate": 2.8563408906848778e-05, "loss": 0.0179, "step": 52720 }, { "grad_norm": 0.3033144474029541, "learning_rate": 2.8544738273585993e-05, "loss": 0.0183, "step": 52730 }, { "grad_norm": 0.2779764235019684, "learning_rate": 2.8526071306440595e-05, "loss": 0.0217, "step": 52740 }, { "grad_norm": 0.44571372866630554, "learning_rate": 2.850740800860233e-05, "loss": 0.0169, "step": 52750 }, { "grad_norm": 0.29587337374687195, "learning_rate": 2.8488748383260178e-05, "loss": 0.0203, "step": 52760 }, { "grad_norm": 0.21172894537448883, "learning_rate": 2.8470092433602635e-05, "loss": 0.0246, "step": 52770 }, { "grad_norm": 0.2139103263616562, "learning_rate": 2.845144016281745e-05, "loss": 0.0185, "step": 52780 }, { "grad_norm": 0.25725334882736206, "learning_rate": 2.8432791574091744e-05, "loss": 0.0287, "step": 52790 }, { "grad_norm": 0.42072594165802, "learning_rate": 2.8414146670612134e-05, "loss": 0.0216, "step": 52800 }, { "grad_norm": 0.2148822396993637, "learning_rate": 2.8395505455564446e-05, "loss": 0.0212, "step": 52810 }, { "grad_norm": 0.43245166540145874, "learning_rate": 2.837686793213403e-05, "loss": 0.0209, "step": 52820 }, { "grad_norm": 0.3461530804634094, "learning_rate": 2.8358234103505477e-05, "loss": 0.0211, "step": 52830 }, { "grad_norm": 0.20084404945373535, "learning_rate": 2.8339603972862806e-05, "loss": 0.0195, "step": 52840 }, { "grad_norm": 0.4008887708187103, "learning_rate": 2.83209775433894e-05, "loss": 0.0211, "step": 52850 }, { "grad_norm": 0.19102880358695984, "learning_rate": 2.8302354818268013e-05, "loss": 0.0202, "step": 52860 }, { "grad_norm": 0.43746358156204224, "learning_rate": 2.8283735800680754e-05, "loss": 0.0244, "step": 52870 }, { "grad_norm": 0.2726457715034485, "learning_rate": 2.826512049380913e-05, "loss": 0.0211, "step": 52880 }, { "grad_norm": 0.1958026885986328, "learning_rate": 2.8246508900833935e-05, "loss": 0.0204, "step": 52890 }, { "grad_norm": 0.21308813989162445, "learning_rate": 2.822790102493541e-05, "loss": 0.0241, "step": 52900 }, { "grad_norm": 0.16371001303195953, "learning_rate": 2.820929686929314e-05, "loss": 0.0223, "step": 52910 }, { "grad_norm": 0.29472842812538147, "learning_rate": 2.819069643708605e-05, "loss": 0.0203, "step": 52920 }, { "grad_norm": 0.19799987971782684, "learning_rate": 2.8172099731492484e-05, "loss": 0.0226, "step": 52930 }, { "grad_norm": 0.21523986756801605, "learning_rate": 2.815350675569003e-05, "loss": 0.0306, "step": 52940 }, { "grad_norm": 0.2093271017074585, "learning_rate": 2.8134917512855807e-05, "loss": 0.0278, "step": 52950 }, { "grad_norm": 0.1594429463148117, "learning_rate": 2.8116332006166134e-05, "loss": 0.0144, "step": 52960 }, { "grad_norm": 0.30906862020492554, "learning_rate": 2.809775023879685e-05, "loss": 0.0277, "step": 52970 }, { "grad_norm": 0.3609706461429596, "learning_rate": 2.807917221392299e-05, "loss": 0.0227, "step": 52980 }, { "grad_norm": 0.31501203775405884, "learning_rate": 2.8060597934719067e-05, "loss": 0.0223, "step": 52990 }, { "grad_norm": 0.2716636061668396, "learning_rate": 2.8042027404358927e-05, "loss": 0.0224, "step": 53000 }, { "grad_norm": 0.32955917716026306, "learning_rate": 2.802346062601572e-05, "loss": 0.0196, "step": 53010 }, { "grad_norm": 0.30463945865631104, "learning_rate": 2.800489760286207e-05, "loss": 0.0251, "step": 53020 }, { "grad_norm": 0.3501180112361908, "learning_rate": 2.7986338338069834e-05, "loss": 0.0243, "step": 53030 }, { "grad_norm": 0.3236221671104431, "learning_rate": 2.7967782834810297e-05, "loss": 0.0301, "step": 53040 }, { "grad_norm": 0.250508189201355, "learning_rate": 2.794923109625409e-05, "loss": 0.0217, "step": 53050 }, { "grad_norm": 0.3045683205127716, "learning_rate": 2.7930683125571205e-05, "loss": 0.0194, "step": 53060 }, { "grad_norm": 0.2571888267993927, "learning_rate": 2.7912138925930997e-05, "loss": 0.0224, "step": 53070 }, { "grad_norm": 0.14949552714824677, "learning_rate": 2.7893598500502117e-05, "loss": 0.0346, "step": 53080 }, { "grad_norm": 0.27427464723587036, "learning_rate": 2.7875061852452644e-05, "loss": 0.0223, "step": 53090 }, { "grad_norm": 0.3404780328273773, "learning_rate": 2.7856528984949982e-05, "loss": 0.0223, "step": 53100 }, { "grad_norm": 0.2898141145706177, "learning_rate": 2.7837999901160888e-05, "loss": 0.0234, "step": 53110 }, { "grad_norm": 0.2375764399766922, "learning_rate": 2.7819474604251484e-05, "loss": 0.0231, "step": 53120 }, { "grad_norm": 0.297902375459671, "learning_rate": 2.780095309738725e-05, "loss": 0.0273, "step": 53130 }, { "grad_norm": 0.3076430857181549, "learning_rate": 2.778243538373294e-05, "loss": 0.0205, "step": 53140 }, { "grad_norm": 0.22227996587753296, "learning_rate": 2.7763921466452826e-05, "loss": 0.0232, "step": 53150 }, { "grad_norm": 0.2623712122440338, "learning_rate": 2.7745411348710336e-05, "loss": 0.0338, "step": 53160 }, { "grad_norm": 0.36785081028938293, "learning_rate": 2.7726905033668395e-05, "loss": 0.0255, "step": 53170 }, { "grad_norm": 0.35857900977134705, "learning_rate": 2.7708402524489214e-05, "loss": 0.0309, "step": 53180 }, { "grad_norm": 0.49953117966651917, "learning_rate": 2.7689903824334364e-05, "loss": 0.0244, "step": 53190 }, { "grad_norm": 0.30346694588661194, "learning_rate": 2.7671408936364785e-05, "loss": 0.024, "step": 53200 }, { "grad_norm": 0.20324958860874176, "learning_rate": 2.765291786374069e-05, "loss": 0.0205, "step": 53210 }, { "grad_norm": 0.2962481379508972, "learning_rate": 2.7634430609621775e-05, "loss": 0.0227, "step": 53220 }, { "grad_norm": 0.28733187913894653, "learning_rate": 2.7615947177166956e-05, "loss": 0.0229, "step": 53230 }, { "grad_norm": 0.29135167598724365, "learning_rate": 2.7597467569534553e-05, "loss": 0.0163, "step": 53240 }, { "grad_norm": 0.25453513860702515, "learning_rate": 2.757899178988226e-05, "loss": 0.0209, "step": 53250 }, { "grad_norm": 0.26137831807136536, "learning_rate": 2.7560519841367005e-05, "loss": 0.0199, "step": 53260 }, { "grad_norm": 0.3109023869037628, "learning_rate": 2.7542051727145237e-05, "loss": 0.0239, "step": 53270 }, { "grad_norm": 0.28888988494873047, "learning_rate": 2.7523587450372578e-05, "loss": 0.0275, "step": 53280 }, { "grad_norm": 0.2900976836681366, "learning_rate": 2.750512701420409e-05, "loss": 0.0187, "step": 53290 }, { "grad_norm": 0.3148730397224426, "learning_rate": 2.7486670421794158e-05, "loss": 0.0221, "step": 53300 }, { "grad_norm": 0.2766130864620209, "learning_rate": 2.7468217676296515e-05, "loss": 0.0335, "step": 53310 }, { "grad_norm": 0.4270761013031006, "learning_rate": 2.7449768780864226e-05, "loss": 0.0241, "step": 53320 }, { "grad_norm": 0.25023892521858215, "learning_rate": 2.7431323738649724e-05, "loss": 0.0187, "step": 53330 }, { "grad_norm": 0.39738789200782776, "learning_rate": 2.7412882552804713e-05, "loss": 0.0279, "step": 53340 }, { "grad_norm": 0.25051653385162354, "learning_rate": 2.739444522648032e-05, "loss": 0.0175, "step": 53350 }, { "grad_norm": 0.40930721163749695, "learning_rate": 2.737601176282697e-05, "loss": 0.0269, "step": 53360 }, { "grad_norm": 0.47864168882369995, "learning_rate": 2.735758216499445e-05, "loss": 0.0246, "step": 53370 }, { "grad_norm": 0.38522791862487793, "learning_rate": 2.7339156436131864e-05, "loss": 0.0241, "step": 53380 }, { "grad_norm": 0.34272611141204834, "learning_rate": 2.7320734579387663e-05, "loss": 0.0214, "step": 53390 }, { "grad_norm": 0.2650604546070099, "learning_rate": 2.730231659790966e-05, "loss": 0.0187, "step": 53400 }, { "grad_norm": 0.2260386347770691, "learning_rate": 2.728390249484494e-05, "loss": 0.0263, "step": 53410 }, { "grad_norm": 0.32910943031311035, "learning_rate": 2.7265492273340032e-05, "loss": 0.02, "step": 53420 }, { "grad_norm": 0.1706121563911438, "learning_rate": 2.7247085936540678e-05, "loss": 0.02, "step": 53430 }, { "grad_norm": 0.5017240047454834, "learning_rate": 2.722868348759205e-05, "loss": 0.0217, "step": 53440 }, { "grad_norm": 0.2479250729084015, "learning_rate": 2.7210284929638635e-05, "loss": 0.0205, "step": 53450 }, { "grad_norm": 0.2341696321964264, "learning_rate": 2.7191890265824183e-05, "loss": 0.0244, "step": 53460 }, { "grad_norm": 0.2649107873439789, "learning_rate": 2.7173499499291926e-05, "loss": 0.0177, "step": 53470 }, { "grad_norm": 0.325374037027359, "learning_rate": 2.7155112633184277e-05, "loss": 0.0198, "step": 53480 }, { "grad_norm": 0.4198813736438751, "learning_rate": 2.713672967064307e-05, "loss": 0.0224, "step": 53490 }, { "grad_norm": 0.2997257709503174, "learning_rate": 2.711835061480945e-05, "loss": 0.0193, "step": 53500 }, { "grad_norm": 0.22654937207698822, "learning_rate": 2.7099975468823896e-05, "loss": 0.0237, "step": 53510 }, { "grad_norm": 0.18821029365062714, "learning_rate": 2.708160423582622e-05, "loss": 0.0181, "step": 53520 }, { "grad_norm": 0.4984179437160492, "learning_rate": 2.706323691895557e-05, "loss": 0.0302, "step": 53530 }, { "grad_norm": 0.2800101637840271, "learning_rate": 2.70448735213504e-05, "loss": 0.0236, "step": 53540 }, { "grad_norm": 0.30458760261535645, "learning_rate": 2.702651404614852e-05, "loss": 0.0213, "step": 53550 }, { "grad_norm": 0.23399637639522552, "learning_rate": 2.7008158496487056e-05, "loss": 0.0258, "step": 53560 }, { "grad_norm": 0.2209867686033249, "learning_rate": 2.6989806875502487e-05, "loss": 0.0181, "step": 53570 }, { "grad_norm": 0.22993232309818268, "learning_rate": 2.6971459186330584e-05, "loss": 0.0156, "step": 53580 }, { "grad_norm": 0.3290483057498932, "learning_rate": 2.695311543210648e-05, "loss": 0.0194, "step": 53590 }, { "grad_norm": 0.2957577705383301, "learning_rate": 2.693477561596463e-05, "loss": 0.0226, "step": 53600 }, { "grad_norm": 0.2555469870567322, "learning_rate": 2.6916439741038756e-05, "loss": 0.0198, "step": 53610 }, { "grad_norm": 0.42791748046875, "learning_rate": 2.689810781046203e-05, "loss": 0.0201, "step": 53620 }, { "grad_norm": 0.24878773093223572, "learning_rate": 2.6879779827366823e-05, "loss": 0.0204, "step": 53630 }, { "grad_norm": 0.2392246574163437, "learning_rate": 2.6861455794884904e-05, "loss": 0.0173, "step": 53640 }, { "grad_norm": 0.219930037856102, "learning_rate": 2.6843135716147373e-05, "loss": 0.0231, "step": 53650 }, { "grad_norm": 0.2427489161491394, "learning_rate": 2.6824819594284556e-05, "loss": 0.0189, "step": 53660 }, { "grad_norm": 0.274906188249588, "learning_rate": 2.6806507432426275e-05, "loss": 0.0221, "step": 53670 }, { "grad_norm": 0.24159806966781616, "learning_rate": 2.6788199233701512e-05, "loss": 0.0235, "step": 53680 }, { "grad_norm": 0.31141456961631775, "learning_rate": 2.6769895001238652e-05, "loss": 0.0227, "step": 53690 }, { "grad_norm": 0.3953535854816437, "learning_rate": 2.67515947381654e-05, "loss": 0.0196, "step": 53700 }, { "grad_norm": 0.2251686453819275, "learning_rate": 2.6733298447608745e-05, "loss": 0.0209, "step": 53710 }, { "grad_norm": 0.29532939195632935, "learning_rate": 2.6715006132695074e-05, "loss": 0.0223, "step": 53720 }, { "grad_norm": 0.3097621202468872, "learning_rate": 2.6696717796549985e-05, "loss": 0.0267, "step": 53730 }, { "grad_norm": 0.41395023465156555, "learning_rate": 2.667843344229848e-05, "loss": 0.0173, "step": 53740 }, { "grad_norm": 0.39401549100875854, "learning_rate": 2.6660153073064846e-05, "loss": 0.0222, "step": 53750 }, { "grad_norm": 0.27429062128067017, "learning_rate": 2.664187669197271e-05, "loss": 0.0272, "step": 53760 }, { "grad_norm": 0.27023231983184814, "learning_rate": 2.6623604302145005e-05, "loss": 0.0167, "step": 53770 }, { "grad_norm": 0.3723195493221283, "learning_rate": 2.6605335906703975e-05, "loss": 0.0209, "step": 53780 }, { "grad_norm": 0.27033936977386475, "learning_rate": 2.65870715087712e-05, "loss": 0.0256, "step": 53790 }, { "grad_norm": 0.3979285657405853, "learning_rate": 2.6568811111467573e-05, "loss": 0.0185, "step": 53800 }, { "grad_norm": 0.26843878626823425, "learning_rate": 2.6550554717913258e-05, "loss": 0.0238, "step": 53810 }, { "grad_norm": 0.2345675826072693, "learning_rate": 2.6532302331227805e-05, "loss": 0.0237, "step": 53820 }, { "grad_norm": 0.2823543846607208, "learning_rate": 2.651405395453004e-05, "loss": 0.017, "step": 53830 }, { "grad_norm": 0.23798289895057678, "learning_rate": 2.6495809590938115e-05, "loss": 0.0267, "step": 53840 }, { "grad_norm": 0.21721844375133514, "learning_rate": 2.64775692435695e-05, "loss": 0.0257, "step": 53850 }, { "grad_norm": 0.22293460369110107, "learning_rate": 2.6459332915540928e-05, "loss": 0.0227, "step": 53860 }, { "grad_norm": 0.3104263246059418, "learning_rate": 2.644110060996856e-05, "loss": 0.0307, "step": 53870 }, { "grad_norm": 0.2545885145664215, "learning_rate": 2.642287232996774e-05, "loss": 0.0193, "step": 53880 }, { "grad_norm": 0.25961020588874817, "learning_rate": 2.6404648078653205e-05, "loss": 0.0194, "step": 53890 }, { "grad_norm": 0.33804312348365784, "learning_rate": 2.6386427859139002e-05, "loss": 0.0186, "step": 53900 }, { "grad_norm": 0.4857775568962097, "learning_rate": 2.63682116745384e-05, "loss": 0.027, "step": 53910 }, { "grad_norm": 0.3475382924079895, "learning_rate": 2.6349999527964138e-05, "loss": 0.0184, "step": 53920 }, { "grad_norm": 0.2574484944343567, "learning_rate": 2.63317914225281e-05, "loss": 0.0305, "step": 53930 }, { "grad_norm": 0.31835973262786865, "learning_rate": 2.6313587361341585e-05, "loss": 0.0227, "step": 53940 }, { "grad_norm": 0.553512692451477, "learning_rate": 2.6295387347515165e-05, "loss": 0.0318, "step": 53950 }, { "grad_norm": 0.2701524496078491, "learning_rate": 2.6277191384158727e-05, "loss": 0.0269, "step": 53960 }, { "grad_norm": 0.26496264338493347, "learning_rate": 2.6258999474381457e-05, "loss": 0.0243, "step": 53970 }, { "grad_norm": 0.4833148419857025, "learning_rate": 2.624081162129186e-05, "loss": 0.0164, "step": 53980 }, { "grad_norm": 0.20182718336582184, "learning_rate": 2.6222627827997765e-05, "loss": 0.0289, "step": 53990 }, { "grad_norm": 0.36739322543144226, "learning_rate": 2.6204448097606236e-05, "loss": 0.0166, "step": 54000 }, { "grad_norm": 0.27404820919036865, "learning_rate": 2.6186272433223726e-05, "loss": 0.0264, "step": 54010 }, { "grad_norm": 0.2336975485086441, "learning_rate": 2.6168100837955943e-05, "loss": 0.0177, "step": 54020 }, { "grad_norm": 0.2290973663330078, "learning_rate": 2.6149933314907926e-05, "loss": 0.0282, "step": 54030 }, { "grad_norm": 0.3115488886833191, "learning_rate": 2.6131769867184e-05, "loss": 0.0252, "step": 54040 }, { "grad_norm": 0.32569119334220886, "learning_rate": 2.611361049788783e-05, "loss": 0.0179, "step": 54050 }, { "grad_norm": 0.307586133480072, "learning_rate": 2.6095455210122292e-05, "loss": 0.0198, "step": 54060 }, { "grad_norm": 0.1509462296962738, "learning_rate": 2.6077304006989712e-05, "loss": 0.0161, "step": 54070 }, { "grad_norm": 0.4587663412094116, "learning_rate": 2.6059156891591562e-05, "loss": 0.0209, "step": 54080 }, { "grad_norm": 0.2249906212091446, "learning_rate": 2.6041013867028718e-05, "loss": 0.0234, "step": 54090 }, { "grad_norm": 0.2670636475086212, "learning_rate": 2.6022874936401347e-05, "loss": 0.0173, "step": 54100 }, { "grad_norm": 0.25228413939476013, "learning_rate": 2.6004740102808832e-05, "loss": 0.0232, "step": 54110 }, { "grad_norm": 0.18302284181118011, "learning_rate": 2.598660936935e-05, "loss": 0.0165, "step": 54120 }, { "grad_norm": 0.3632902204990387, "learning_rate": 2.5968482739122845e-05, "loss": 0.0245, "step": 54130 }, { "grad_norm": 0.35868337750434875, "learning_rate": 2.595036021522472e-05, "loss": 0.0207, "step": 54140 }, { "grad_norm": 0.16948416829109192, "learning_rate": 2.5932241800752278e-05, "loss": 0.0197, "step": 54150 }, { "grad_norm": 0.3750610947608948, "learning_rate": 2.5914127498801453e-05, "loss": 0.0348, "step": 54160 }, { "grad_norm": 0.33963873982429504, "learning_rate": 2.5896017312467497e-05, "loss": 0.0187, "step": 54170 }, { "grad_norm": 0.29451584815979004, "learning_rate": 2.587791124484493e-05, "loss": 0.018, "step": 54180 }, { "grad_norm": 0.17828412353992462, "learning_rate": 2.5859809299027615e-05, "loss": 0.0229, "step": 54190 }, { "grad_norm": 0.5083809494972229, "learning_rate": 2.5841711478108632e-05, "loss": 0.0331, "step": 54200 }, { "grad_norm": 0.2599181532859802, "learning_rate": 2.582361778518043e-05, "loss": 0.0198, "step": 54210 }, { "grad_norm": 0.267986923456192, "learning_rate": 2.580552822333472e-05, "loss": 0.024, "step": 54220 }, { "grad_norm": 0.21616490185260773, "learning_rate": 2.578744279566252e-05, "loss": 0.0205, "step": 54230 }, { "grad_norm": 0.21275384724140167, "learning_rate": 2.576936150525413e-05, "loss": 0.0335, "step": 54240 }, { "grad_norm": 0.297934889793396, "learning_rate": 2.5751284355199168e-05, "loss": 0.0167, "step": 54250 }, { "grad_norm": 0.3383019268512726, "learning_rate": 2.573321134858646e-05, "loss": 0.019, "step": 54260 }, { "grad_norm": 0.20977731049060822, "learning_rate": 2.5715142488504286e-05, "loss": 0.0177, "step": 54270 }, { "grad_norm": 0.30297520756721497, "learning_rate": 2.5697077778040042e-05, "loss": 0.0228, "step": 54280 }, { "grad_norm": 0.19375832378864288, "learning_rate": 2.5679017220280522e-05, "loss": 0.0315, "step": 54290 }, { "grad_norm": 0.20685088634490967, "learning_rate": 2.5660960818311796e-05, "loss": 0.0193, "step": 54300 }, { "grad_norm": 0.2957373857498169, "learning_rate": 2.564290857521915e-05, "loss": 0.0245, "step": 54310 }, { "grad_norm": 0.35275936126708984, "learning_rate": 2.5624860494087298e-05, "loss": 0.022, "step": 54320 }, { "grad_norm": 0.2929324209690094, "learning_rate": 2.5606816578000115e-05, "loss": 0.0261, "step": 54330 }, { "grad_norm": 0.2786456048488617, "learning_rate": 2.558877683004082e-05, "loss": 0.0235, "step": 54340 }, { "grad_norm": 0.27315759658813477, "learning_rate": 2.557074125329192e-05, "loss": 0.0185, "step": 54350 }, { "grad_norm": 0.3174957036972046, "learning_rate": 2.5552709850835195e-05, "loss": 0.0208, "step": 54360 }, { "grad_norm": 0.21155494451522827, "learning_rate": 2.5534682625751738e-05, "loss": 0.016, "step": 54370 }, { "grad_norm": 0.23284457623958588, "learning_rate": 2.551665958112186e-05, "loss": 0.0284, "step": 54380 }, { "grad_norm": 0.3648332953453064, "learning_rate": 2.549864072002527e-05, "loss": 0.0179, "step": 54390 }, { "grad_norm": 0.3631495237350464, "learning_rate": 2.5480626045540858e-05, "loss": 0.0251, "step": 54400 }, { "grad_norm": 0.23622381687164307, "learning_rate": 2.546261556074684e-05, "loss": 0.0182, "step": 54410 }, { "grad_norm": 0.3805238902568817, "learning_rate": 2.5444609268720726e-05, "loss": 0.0251, "step": 54420 }, { "grad_norm": 0.28518158197402954, "learning_rate": 2.5426607172539297e-05, "loss": 0.0191, "step": 54430 }, { "grad_norm": 0.45397835969924927, "learning_rate": 2.5408609275278617e-05, "loss": 0.0288, "step": 54440 }, { "grad_norm": 0.2494930922985077, "learning_rate": 2.5390615580014055e-05, "loss": 0.0235, "step": 54450 }, { "grad_norm": 0.19577234983444214, "learning_rate": 2.5372626089820207e-05, "loss": 0.0244, "step": 54460 }, { "grad_norm": 0.31497761607170105, "learning_rate": 2.5354640807770997e-05, "loss": 0.0184, "step": 54470 }, { "grad_norm": 0.2842310070991516, "learning_rate": 2.5336659736939622e-05, "loss": 0.0197, "step": 54480 }, { "grad_norm": 0.28803959488868713, "learning_rate": 2.5318682880398554e-05, "loss": 0.0254, "step": 54490 }, { "grad_norm": 0.27626460790634155, "learning_rate": 2.530071024121956e-05, "loss": 0.0271, "step": 54500 }, { "grad_norm": 0.5418146848678589, "learning_rate": 2.5282741822473627e-05, "loss": 0.0192, "step": 54510 }, { "grad_norm": 0.30908188223838806, "learning_rate": 2.526477762723114e-05, "loss": 0.0262, "step": 54520 }, { "grad_norm": 0.24163246154785156, "learning_rate": 2.5246817658561618e-05, "loss": 0.0196, "step": 54530 }, { "grad_norm": 0.21141575276851654, "learning_rate": 2.5228861919533965e-05, "loss": 0.018, "step": 54540 }, { "grad_norm": 0.23891089856624603, "learning_rate": 2.5210910413216326e-05, "loss": 0.0354, "step": 54550 }, { "grad_norm": 0.3702000379562378, "learning_rate": 2.5192963142676086e-05, "loss": 0.0195, "step": 54560 }, { "grad_norm": 0.27960190176963806, "learning_rate": 2.517502011098001e-05, "loss": 0.0187, "step": 54570 }, { "grad_norm": 0.2104295790195465, "learning_rate": 2.5157081321193987e-05, "loss": 0.0253, "step": 54580 }, { "grad_norm": 0.2628540098667145, "learning_rate": 2.5139146776383356e-05, "loss": 0.0211, "step": 54590 }, { "grad_norm": 0.24825036525726318, "learning_rate": 2.5121216479612575e-05, "loss": 0.0198, "step": 54600 }, { "grad_norm": 0.21393682062625885, "learning_rate": 2.510329043394546e-05, "loss": 0.0265, "step": 54610 }, { "grad_norm": 0.3836612105369568, "learning_rate": 2.508536864244508e-05, "loss": 0.0224, "step": 54620 }, { "grad_norm": 0.3059808611869812, "learning_rate": 2.5067451108173778e-05, "loss": 0.0201, "step": 54630 }, { "grad_norm": 0.2592499554157257, "learning_rate": 2.5049537834193204e-05, "loss": 0.0187, "step": 54640 }, { "grad_norm": 0.25171157717704773, "learning_rate": 2.5031628823564194e-05, "loss": 0.0233, "step": 54650 }, { "grad_norm": 0.2440548837184906, "learning_rate": 2.5013724079346933e-05, "loss": 0.0245, "step": 54660 }, { "grad_norm": 0.3018760085105896, "learning_rate": 2.4995823604600854e-05, "loss": 0.0174, "step": 54670 }, { "grad_norm": 0.38770994544029236, "learning_rate": 2.497792740238465e-05, "loss": 0.0236, "step": 54680 }, { "grad_norm": 0.24448612332344055, "learning_rate": 2.49600354757563e-05, "loss": 0.0293, "step": 54690 }, { "grad_norm": 0.3261798322200775, "learning_rate": 2.494214782777306e-05, "loss": 0.0216, "step": 54700 }, { "grad_norm": 0.2348642200231552, "learning_rate": 2.4924264461491386e-05, "loss": 0.0195, "step": 54710 }, { "grad_norm": 0.42119431495666504, "learning_rate": 2.4906385379967133e-05, "loss": 0.0241, "step": 54720 }, { "grad_norm": 0.36883634328842163, "learning_rate": 2.4888510586255285e-05, "loss": 0.0192, "step": 54730 }, { "grad_norm": 0.2244393676519394, "learning_rate": 2.487064008341018e-05, "loss": 0.0235, "step": 54740 }, { "grad_norm": 0.37828972935676575, "learning_rate": 2.4852773874485407e-05, "loss": 0.0195, "step": 54750 }, { "grad_norm": 0.36310237646102905, "learning_rate": 2.483491196253377e-05, "loss": 0.0206, "step": 54760 }, { "grad_norm": 0.3727313280105591, "learning_rate": 2.4817054350607443e-05, "loss": 0.0253, "step": 54770 }, { "grad_norm": 0.31966909766197205, "learning_rate": 2.4799201041757742e-05, "loss": 0.0179, "step": 54780 }, { "grad_norm": 0.277908593416214, "learning_rate": 2.4781352039035373e-05, "loss": 0.0268, "step": 54790 }, { "grad_norm": 0.24865396320819855, "learning_rate": 2.4763507345490194e-05, "loss": 0.0184, "step": 54800 }, { "grad_norm": 0.26708388328552246, "learning_rate": 2.4745666964171386e-05, "loss": 0.0223, "step": 54810 }, { "grad_norm": 0.30594438314437866, "learning_rate": 2.4727830898127387e-05, "loss": 0.0182, "step": 54820 }, { "grad_norm": 0.3038524389266968, "learning_rate": 2.4709999150405895e-05, "loss": 0.0241, "step": 54830 }, { "grad_norm": 0.45349031686782837, "learning_rate": 2.469217172405388e-05, "loss": 0.0238, "step": 54840 }, { "grad_norm": 0.2735951840877533, "learning_rate": 2.4674348622117527e-05, "loss": 0.0168, "step": 54850 }, { "grad_norm": 0.28689947724342346, "learning_rate": 2.465652984764234e-05, "loss": 0.0203, "step": 54860 }, { "grad_norm": 0.22499097883701324, "learning_rate": 2.4638715403673056e-05, "loss": 0.0174, "step": 54870 }, { "grad_norm": 0.398593544960022, "learning_rate": 2.4620905293253676e-05, "loss": 0.0216, "step": 54880 }, { "grad_norm": 0.3670974671840668, "learning_rate": 2.4603099519427463e-05, "loss": 0.0239, "step": 54890 }, { "grad_norm": 0.31919845938682556, "learning_rate": 2.458529808523695e-05, "loss": 0.0207, "step": 54900 }, { "grad_norm": 0.2610417604446411, "learning_rate": 2.456750099372387e-05, "loss": 0.0194, "step": 54910 }, { "grad_norm": 0.3604589104652405, "learning_rate": 2.4549708247929327e-05, "loss": 0.0281, "step": 54920 }, { "grad_norm": 0.2586463987827301, "learning_rate": 2.4531919850893554e-05, "loss": 0.0204, "step": 54930 }, { "grad_norm": 0.3002761900424957, "learning_rate": 2.4514135805656125e-05, "loss": 0.0175, "step": 54940 }, { "grad_norm": 0.5488714575767517, "learning_rate": 2.449635611525587e-05, "loss": 0.023, "step": 54950 }, { "grad_norm": 0.2697174549102783, "learning_rate": 2.447858078273079e-05, "loss": 0.0215, "step": 54960 }, { "grad_norm": 0.3102734088897705, "learning_rate": 2.4460809811118275e-05, "loss": 0.0274, "step": 54970 }, { "grad_norm": 0.2543473541736603, "learning_rate": 2.444304320345483e-05, "loss": 0.0187, "step": 54980 }, { "grad_norm": 0.5040074586868286, "learning_rate": 2.4425280962776354e-05, "loss": 0.0248, "step": 54990 }, { "grad_norm": 0.19326600432395935, "learning_rate": 2.4407523092117875e-05, "loss": 0.0207, "step": 55000 }, { "grad_norm": 0.2526605725288391, "learning_rate": 2.438976959451374e-05, "loss": 0.0172, "step": 55010 }, { "grad_norm": 0.29454755783081055, "learning_rate": 2.4372020472997565e-05, "loss": 0.0199, "step": 55020 }, { "grad_norm": 0.2832791805267334, "learning_rate": 2.4354275730602122e-05, "loss": 0.0217, "step": 55030 }, { "grad_norm": 0.22052156925201416, "learning_rate": 2.433653537035958e-05, "loss": 0.0193, "step": 55040 }, { "grad_norm": 0.24124526977539062, "learning_rate": 2.431879939530123e-05, "loss": 0.0221, "step": 55050 }, { "grad_norm": 0.28355178236961365, "learning_rate": 2.4301067808457684e-05, "loss": 0.0188, "step": 55060 }, { "grad_norm": 0.29363298416137695, "learning_rate": 2.428334061285878e-05, "loss": 0.0213, "step": 55070 }, { "grad_norm": 0.4717520773410797, "learning_rate": 2.426561781153361e-05, "loss": 0.0182, "step": 55080 }, { "grad_norm": 0.26417842507362366, "learning_rate": 2.424789940751052e-05, "loss": 0.0163, "step": 55090 }, { "grad_norm": 0.2578654885292053, "learning_rate": 2.423018540381712e-05, "loss": 0.0235, "step": 55100 }, { "grad_norm": 0.23384493589401245, "learning_rate": 2.421247580348021e-05, "loss": 0.0208, "step": 55110 }, { "grad_norm": 0.22927771508693695, "learning_rate": 2.4194770609525897e-05, "loss": 0.0175, "step": 55120 }, { "grad_norm": 0.34098806977272034, "learning_rate": 2.417706982497951e-05, "loss": 0.0265, "step": 55130 }, { "grad_norm": 0.3511075973510742, "learning_rate": 2.415937345286563e-05, "loss": 0.0269, "step": 55140 }, { "grad_norm": 0.39078599214553833, "learning_rate": 2.4141681496208087e-05, "loss": 0.0221, "step": 55150 }, { "grad_norm": 0.3278678059577942, "learning_rate": 2.4123993958029946e-05, "loss": 0.0155, "step": 55160 }, { "grad_norm": 0.3326279819011688, "learning_rate": 2.4106310841353548e-05, "loss": 0.019, "step": 55170 }, { "grad_norm": 0.2628838121891022, "learning_rate": 2.4088632149200398e-05, "loss": 0.0317, "step": 55180 }, { "grad_norm": 0.3208889067173004, "learning_rate": 2.4070957884591367e-05, "loss": 0.0203, "step": 55190 }, { "grad_norm": 0.4123530387878418, "learning_rate": 2.4053288050546464e-05, "loss": 0.0292, "step": 55200 }, { "grad_norm": 0.3444884121417999, "learning_rate": 2.403562265008498e-05, "loss": 0.0235, "step": 55210 }, { "grad_norm": 0.32687488198280334, "learning_rate": 2.4017961686225483e-05, "loss": 0.0284, "step": 55220 }, { "grad_norm": 0.2592519223690033, "learning_rate": 2.400030516198568e-05, "loss": 0.0291, "step": 55230 }, { "grad_norm": 0.26218390464782715, "learning_rate": 2.3982653080382673e-05, "loss": 0.0199, "step": 55240 }, { "grad_norm": 0.29601484537124634, "learning_rate": 2.396500544443266e-05, "loss": 0.023, "step": 55250 }, { "grad_norm": 0.580190896987915, "learning_rate": 2.3947362257151156e-05, "loss": 0.0232, "step": 55260 }, { "grad_norm": 0.3554472327232361, "learning_rate": 2.39297235215529e-05, "loss": 0.0249, "step": 55270 }, { "grad_norm": 0.37281766533851624, "learning_rate": 2.3912089240651873e-05, "loss": 0.0227, "step": 55280 }, { "grad_norm": 0.20633794367313385, "learning_rate": 2.389445941746129e-05, "loss": 0.0263, "step": 55290 }, { "grad_norm": 0.26536881923675537, "learning_rate": 2.3876834054993625e-05, "loss": 0.0216, "step": 55300 }, { "grad_norm": 0.17517827451229095, "learning_rate": 2.3859213156260522e-05, "loss": 0.0176, "step": 55310 }, { "grad_norm": 0.2186461091041565, "learning_rate": 2.3841596724272948e-05, "loss": 0.0206, "step": 55320 }, { "grad_norm": 0.2435792237520218, "learning_rate": 2.382398476204106e-05, "loss": 0.0223, "step": 55330 }, { "grad_norm": 0.2126416116952896, "learning_rate": 2.3806377272574254e-05, "loss": 0.0214, "step": 55340 }, { "grad_norm": 0.26457464694976807, "learning_rate": 2.3788774258881174e-05, "loss": 0.0194, "step": 55350 }, { "grad_norm": 0.24580730497837067, "learning_rate": 2.37711757239697e-05, "loss": 0.0184, "step": 55360 }, { "grad_norm": 0.31060585379600525, "learning_rate": 2.3753581670846954e-05, "loss": 0.0191, "step": 55370 }, { "grad_norm": 0.22675615549087524, "learning_rate": 2.3735992102519216e-05, "loss": 0.0218, "step": 55380 }, { "grad_norm": 0.3417225778102875, "learning_rate": 2.371840702199215e-05, "loss": 0.023, "step": 55390 }, { "grad_norm": 0.36862924695014954, "learning_rate": 2.3700826432270494e-05, "loss": 0.0226, "step": 55400 }, { "grad_norm": 0.2511278986930847, "learning_rate": 2.3683250336358326e-05, "loss": 0.0285, "step": 55410 }, { "grad_norm": 0.2195347398519516, "learning_rate": 2.3665678737258923e-05, "loss": 0.0241, "step": 55420 }, { "grad_norm": 0.10522628575563431, "learning_rate": 2.3648111637974745e-05, "loss": 0.0172, "step": 55430 }, { "grad_norm": 0.2289244681596756, "learning_rate": 2.36305490415076e-05, "loss": 0.0277, "step": 55440 }, { "grad_norm": 0.5193437933921814, "learning_rate": 2.3612990950858392e-05, "loss": 0.0196, "step": 55450 }, { "grad_norm": 0.21326173841953278, "learning_rate": 2.359543736902735e-05, "loss": 0.0227, "step": 55460 }, { "grad_norm": 0.2515532374382019, "learning_rate": 2.3577888299013896e-05, "loss": 0.0199, "step": 55470 }, { "grad_norm": 0.2661832869052887, "learning_rate": 2.3560343743816683e-05, "loss": 0.017, "step": 55480 }, { "grad_norm": 0.3073561489582062, "learning_rate": 2.354280370643362e-05, "loss": 0.0285, "step": 55490 }, { "grad_norm": 0.30670905113220215, "learning_rate": 2.3525268189861777e-05, "loss": 0.0198, "step": 55500 }, { "grad_norm": 0.30819830298423767, "learning_rate": 2.3507737197097513e-05, "loss": 0.017, "step": 55510 }, { "grad_norm": 0.3441518247127533, "learning_rate": 2.34902107311364e-05, "loss": 0.0209, "step": 55520 }, { "grad_norm": 0.3313232362270355, "learning_rate": 2.3472688794973225e-05, "loss": 0.0226, "step": 55530 }, { "grad_norm": 0.3037632405757904, "learning_rate": 2.3455171391602016e-05, "loss": 0.024, "step": 55540 }, { "grad_norm": 0.34075042605400085, "learning_rate": 2.343765852401601e-05, "loss": 0.0231, "step": 55550 }, { "grad_norm": 0.5211811661720276, "learning_rate": 2.342015019520768e-05, "loss": 0.0258, "step": 55560 }, { "grad_norm": 0.2676220238208771, "learning_rate": 2.3402646408168742e-05, "loss": 0.0171, "step": 55570 }, { "grad_norm": 0.3301401138305664, "learning_rate": 2.3385147165890074e-05, "loss": 0.022, "step": 55580 }, { "grad_norm": 0.28201526403427124, "learning_rate": 2.336765247136184e-05, "loss": 0.0209, "step": 55590 }, { "grad_norm": 0.31905215978622437, "learning_rate": 2.33501623275734e-05, "loss": 0.0283, "step": 55600 }, { "grad_norm": 0.19237057864665985, "learning_rate": 2.333267673751334e-05, "loss": 0.0148, "step": 55610 }, { "grad_norm": 0.2331121563911438, "learning_rate": 2.33151957041695e-05, "loss": 0.0194, "step": 55620 }, { "grad_norm": 0.5061233043670654, "learning_rate": 2.329771923052884e-05, "loss": 0.0243, "step": 55630 }, { "grad_norm": 0.4790826737880707, "learning_rate": 2.3280247319577697e-05, "loss": 0.0203, "step": 55640 }, { "grad_norm": 0.2600797414779663, "learning_rate": 2.3262779974301473e-05, "loss": 0.0235, "step": 55650 }, { "grad_norm": 0.2418091893196106, "learning_rate": 2.3245317197684895e-05, "loss": 0.0187, "step": 55660 }, { "grad_norm": 0.17051304876804352, "learning_rate": 2.322785899271188e-05, "loss": 0.0194, "step": 55670 }, { "grad_norm": 0.16377021372318268, "learning_rate": 2.3210405362365507e-05, "loss": 0.028, "step": 55680 }, { "grad_norm": 0.30083754658699036, "learning_rate": 2.31929563096282e-05, "loss": 0.0217, "step": 55690 }, { "grad_norm": 0.29237228631973267, "learning_rate": 2.317551183748146e-05, "loss": 0.0165, "step": 55700 }, { "grad_norm": 0.1525973230600357, "learning_rate": 2.3158071948906103e-05, "loss": 0.0172, "step": 55710 }, { "grad_norm": 0.3874724507331848, "learning_rate": 2.314063664688212e-05, "loss": 0.0158, "step": 55720 }, { "grad_norm": 0.2980857491493225, "learning_rate": 2.3123205934388725e-05, "loss": 0.0162, "step": 55730 }, { "grad_norm": 0.24091419577598572, "learning_rate": 2.310577981440436e-05, "loss": 0.0301, "step": 55740 }, { "grad_norm": 0.2026970535516739, "learning_rate": 2.3088358289906665e-05, "loss": 0.0194, "step": 55750 }, { "grad_norm": 0.40737584233283997, "learning_rate": 2.307094136387252e-05, "loss": 0.0187, "step": 55760 }, { "grad_norm": 0.2923775613307953, "learning_rate": 2.305352903927796e-05, "loss": 0.0293, "step": 55770 }, { "grad_norm": 0.23082886636257172, "learning_rate": 2.303612131909831e-05, "loss": 0.0173, "step": 55780 }, { "grad_norm": 0.2795584201812744, "learning_rate": 2.3018718206308054e-05, "loss": 0.0247, "step": 55790 }, { "grad_norm": 0.3219940960407257, "learning_rate": 2.3001319703880925e-05, "loss": 0.0201, "step": 55800 }, { "grad_norm": 0.3776647448539734, "learning_rate": 2.2983925814789835e-05, "loss": 0.0167, "step": 55810 }, { "grad_norm": 0.30613768100738525, "learning_rate": 2.2966536542006957e-05, "loss": 0.0183, "step": 55820 }, { "grad_norm": 0.22061985731124878, "learning_rate": 2.294915188850358e-05, "loss": 0.0163, "step": 55830 }, { "grad_norm": 0.3052194118499756, "learning_rate": 2.2931771857250333e-05, "loss": 0.015, "step": 55840 }, { "grad_norm": 0.1801384836435318, "learning_rate": 2.2914396451216946e-05, "loss": 0.0166, "step": 55850 }, { "grad_norm": 0.4012281894683838, "learning_rate": 2.2897025673372412e-05, "loss": 0.0187, "step": 55860 }, { "grad_norm": 0.243429496884346, "learning_rate": 2.287965952668494e-05, "loss": 0.0197, "step": 55870 }, { "grad_norm": 0.24747729301452637, "learning_rate": 2.2862298014121873e-05, "loss": 0.0196, "step": 55880 }, { "grad_norm": 0.3213168680667877, "learning_rate": 2.28449411386499e-05, "loss": 0.0211, "step": 55890 }, { "grad_norm": 0.3091227412223816, "learning_rate": 2.282758890323477e-05, "loss": 0.0233, "step": 55900 }, { "grad_norm": 0.2067926824092865, "learning_rate": 2.2810241310841528e-05, "loss": 0.0209, "step": 55910 }, { "grad_norm": 0.24098652601242065, "learning_rate": 2.2792898364434412e-05, "loss": 0.0174, "step": 55920 }, { "grad_norm": 0.22998301684856415, "learning_rate": 2.2775560066976843e-05, "loss": 0.0207, "step": 55930 }, { "grad_norm": 0.19612008333206177, "learning_rate": 2.275822642143147e-05, "loss": 0.0196, "step": 55940 }, { "grad_norm": 0.21758770942687988, "learning_rate": 2.274089743076014e-05, "loss": 0.0161, "step": 55950 }, { "grad_norm": 0.41526028513908386, "learning_rate": 2.2723573097923917e-05, "loss": 0.0199, "step": 55960 }, { "grad_norm": 0.211795374751091, "learning_rate": 2.2706253425883024e-05, "loss": 0.0167, "step": 55970 }, { "grad_norm": 0.4606720209121704, "learning_rate": 2.2688938417596933e-05, "loss": 0.0227, "step": 55980 }, { "grad_norm": 0.23382899165153503, "learning_rate": 2.2671628076024305e-05, "loss": 0.0231, "step": 55990 }, { "grad_norm": 0.29126766324043274, "learning_rate": 2.2654322404123008e-05, "loss": 0.0141, "step": 56000 }, { "grad_norm": 0.16671347618103027, "learning_rate": 2.2637021404850105e-05, "loss": 0.0151, "step": 56010 }, { "grad_norm": 0.29955407977104187, "learning_rate": 2.2619725081161885e-05, "loss": 0.0206, "step": 56020 }, { "grad_norm": 0.17811550199985504, "learning_rate": 2.260243343601376e-05, "loss": 0.0194, "step": 56030 }, { "grad_norm": 0.23111838102340698, "learning_rate": 2.2585146472360473e-05, "loss": 0.0215, "step": 56040 }, { "grad_norm": 0.3005419373512268, "learning_rate": 2.2567864193155834e-05, "loss": 0.0197, "step": 56050 }, { "grad_norm": 0.27738437056541443, "learning_rate": 2.255058660135294e-05, "loss": 0.0221, "step": 56060 }, { "grad_norm": 0.5012184977531433, "learning_rate": 2.253331369990407e-05, "loss": 0.022, "step": 56070 }, { "grad_norm": 0.1525881141424179, "learning_rate": 2.251604549176063e-05, "loss": 0.0212, "step": 56080 }, { "grad_norm": 0.1335427314043045, "learning_rate": 2.249878197987337e-05, "loss": 0.0178, "step": 56090 }, { "grad_norm": 0.17725148797035217, "learning_rate": 2.2481523167192087e-05, "loss": 0.0188, "step": 56100 }, { "grad_norm": 0.28297150135040283, "learning_rate": 2.2464269056665867e-05, "loss": 0.0186, "step": 56110 }, { "grad_norm": 0.36523789167404175, "learning_rate": 2.2447019651242958e-05, "loss": 0.0246, "step": 56120 }, { "grad_norm": 0.3616890609264374, "learning_rate": 2.2429774953870818e-05, "loss": 0.0222, "step": 56130 }, { "grad_norm": 0.37530699372291565, "learning_rate": 2.241253496749611e-05, "loss": 0.0254, "step": 56140 }, { "grad_norm": 0.27625343203544617, "learning_rate": 2.2395299695064614e-05, "loss": 0.0177, "step": 56150 }, { "grad_norm": 0.29714730381965637, "learning_rate": 2.237806913952145e-05, "loss": 0.0214, "step": 56160 }, { "grad_norm": 0.22726137936115265, "learning_rate": 2.2360843303810798e-05, "loss": 0.0148, "step": 56170 }, { "grad_norm": 0.3614886701107025, "learning_rate": 2.2343622190876084e-05, "loss": 0.0214, "step": 56180 }, { "grad_norm": 0.27106204628944397, "learning_rate": 2.2326405803659935e-05, "loss": 0.0176, "step": 56190 }, { "grad_norm": 0.2657414674758911, "learning_rate": 2.230919414510416e-05, "loss": 0.0191, "step": 56200 }, { "grad_norm": 0.4252246916294098, "learning_rate": 2.229198721814976e-05, "loss": 0.0246, "step": 56210 }, { "grad_norm": 0.3361543118953705, "learning_rate": 2.2274785025736948e-05, "loss": 0.0171, "step": 56220 }, { "grad_norm": 0.241159126162529, "learning_rate": 2.225758757080507e-05, "loss": 0.0232, "step": 56230 }, { "grad_norm": 0.4955720007419586, "learning_rate": 2.2240394856292723e-05, "loss": 0.0195, "step": 56240 }, { "grad_norm": 0.2927810251712799, "learning_rate": 2.2223206885137664e-05, "loss": 0.0183, "step": 56250 }, { "grad_norm": 0.44001996517181396, "learning_rate": 2.2206023660276853e-05, "loss": 0.0292, "step": 56260 }, { "grad_norm": 0.2359338104724884, "learning_rate": 2.218884518464645e-05, "loss": 0.0255, "step": 56270 }, { "grad_norm": 0.3620072603225708, "learning_rate": 2.2171671461181732e-05, "loss": 0.0216, "step": 56280 }, { "grad_norm": 0.3035140037536621, "learning_rate": 2.2154502492817292e-05, "loss": 0.0261, "step": 56290 }, { "grad_norm": 0.2614527642726898, "learning_rate": 2.2137338282486782e-05, "loss": 0.019, "step": 56300 }, { "grad_norm": 0.4365866780281067, "learning_rate": 2.2120178833123113e-05, "loss": 0.0213, "step": 56310 }, { "grad_norm": 0.2904083728790283, "learning_rate": 2.210302414765838e-05, "loss": 0.023, "step": 56320 }, { "grad_norm": 0.20598115026950836, "learning_rate": 2.2085874229023808e-05, "loss": 0.0181, "step": 56330 }, { "grad_norm": 0.19319936633110046, "learning_rate": 2.2068729080149907e-05, "loss": 0.0198, "step": 56340 }, { "grad_norm": 0.29661068320274353, "learning_rate": 2.205158870396625e-05, "loss": 0.0234, "step": 56350 }, { "grad_norm": 0.3352031409740448, "learning_rate": 2.2034453103401732e-05, "loss": 0.0226, "step": 56360 }, { "grad_norm": 0.20702317357063293, "learning_rate": 2.20173222813843e-05, "loss": 0.0142, "step": 56370 }, { "grad_norm": 0.18753351271152496, "learning_rate": 2.200019624084116e-05, "loss": 0.0221, "step": 56380 }, { "grad_norm": 0.3008200526237488, "learning_rate": 2.1983074984698687e-05, "loss": 0.0182, "step": 56390 }, { "grad_norm": 0.3322056829929352, "learning_rate": 2.1965958515882433e-05, "loss": 0.0268, "step": 56400 }, { "grad_norm": 0.2331034541130066, "learning_rate": 2.1948846837317162e-05, "loss": 0.0227, "step": 56410 }, { "grad_norm": 0.22703085839748383, "learning_rate": 2.1931739951926738e-05, "loss": 0.0148, "step": 56420 }, { "grad_norm": 0.270680695772171, "learning_rate": 2.191463786263429e-05, "loss": 0.0186, "step": 56430 }, { "grad_norm": 0.22675640881061554, "learning_rate": 2.1897540572362095e-05, "loss": 0.0222, "step": 56440 }, { "grad_norm": 0.1575874537229538, "learning_rate": 2.1880448084031614e-05, "loss": 0.0187, "step": 56450 }, { "grad_norm": 0.2326916605234146, "learning_rate": 2.1863360400563482e-05, "loss": 0.0184, "step": 56460 }, { "grad_norm": 0.4656137228012085, "learning_rate": 2.184627752487754e-05, "loss": 0.0255, "step": 56470 }, { "grad_norm": 0.2382478415966034, "learning_rate": 2.1829199459892725e-05, "loss": 0.0202, "step": 56480 }, { "grad_norm": 0.2668207287788391, "learning_rate": 2.1812126208527282e-05, "loss": 0.0195, "step": 56490 }, { "grad_norm": 0.25915688276290894, "learning_rate": 2.179505777369852e-05, "loss": 0.0247, "step": 56500 }, { "grad_norm": 0.25532862544059753, "learning_rate": 2.1777994158322974e-05, "loss": 0.0168, "step": 56510 }, { "grad_norm": 0.29350557923316956, "learning_rate": 2.176093536531637e-05, "loss": 0.0251, "step": 56520 }, { "grad_norm": 0.30688416957855225, "learning_rate": 2.1743881397593537e-05, "loss": 0.0193, "step": 56530 }, { "grad_norm": 0.4193277060985565, "learning_rate": 2.1726832258068595e-05, "loss": 0.0261, "step": 56540 }, { "grad_norm": 0.37551847100257874, "learning_rate": 2.170978794965472e-05, "loss": 0.0242, "step": 56550 }, { "grad_norm": 0.18926995992660522, "learning_rate": 2.169274847526438e-05, "loss": 0.0174, "step": 56560 }, { "grad_norm": 0.35665687918663025, "learning_rate": 2.1675713837809103e-05, "loss": 0.0197, "step": 56570 }, { "grad_norm": 0.2967902421951294, "learning_rate": 2.1658684040199655e-05, "loss": 0.0176, "step": 56580 }, { "grad_norm": 0.3975285589694977, "learning_rate": 2.1641659085345974e-05, "loss": 0.0204, "step": 56590 }, { "grad_norm": 0.38147974014282227, "learning_rate": 2.1624638976157154e-05, "loss": 0.0236, "step": 56600 }, { "grad_norm": 0.31761324405670166, "learning_rate": 2.1607623715541476e-05, "loss": 0.0285, "step": 56610 }, { "grad_norm": 0.2850551903247833, "learning_rate": 2.159061330640636e-05, "loss": 0.0184, "step": 56620 }, { "grad_norm": 0.20010386407375336, "learning_rate": 2.1573607751658425e-05, "loss": 0.0127, "step": 56630 }, { "grad_norm": 0.20882780849933624, "learning_rate": 2.155660705420347e-05, "loss": 0.0215, "step": 56640 }, { "grad_norm": 0.21722300350666046, "learning_rate": 2.153961121694644e-05, "loss": 0.0185, "step": 56650 }, { "grad_norm": 0.2544526159763336, "learning_rate": 2.152262024279145e-05, "loss": 0.0273, "step": 56660 }, { "grad_norm": 0.4592951536178589, "learning_rate": 2.150563413464183e-05, "loss": 0.0241, "step": 56670 }, { "grad_norm": 0.25330454111099243, "learning_rate": 2.1488652895399974e-05, "loss": 0.0181, "step": 56680 }, { "grad_norm": 0.2485608458518982, "learning_rate": 2.1471676527967587e-05, "loss": 0.0182, "step": 56690 }, { "grad_norm": 0.22487454116344452, "learning_rate": 2.1454705035245414e-05, "loss": 0.0276, "step": 56700 }, { "grad_norm": 0.12542356550693512, "learning_rate": 2.143773842013343e-05, "loss": 0.0144, "step": 56710 }, { "grad_norm": 0.2640679180622101, "learning_rate": 2.1420776685530796e-05, "loss": 0.0221, "step": 56720 }, { "grad_norm": 0.31304705142974854, "learning_rate": 2.1403819834335742e-05, "loss": 0.0144, "step": 56730 }, { "grad_norm": 0.18929104506969452, "learning_rate": 2.1386867869445808e-05, "loss": 0.0235, "step": 56740 }, { "grad_norm": 0.16429004073143005, "learning_rate": 2.1369920793757548e-05, "loss": 0.0246, "step": 56750 }, { "grad_norm": 0.2186392843723297, "learning_rate": 2.1352978610166828e-05, "loss": 0.0176, "step": 56760 }, { "grad_norm": 0.4477491080760956, "learning_rate": 2.1336041321568546e-05, "loss": 0.0199, "step": 56770 }, { "grad_norm": 0.41468074917793274, "learning_rate": 2.131910893085684e-05, "loss": 0.019, "step": 56780 }, { "grad_norm": 0.25260382890701294, "learning_rate": 2.1302181440925012e-05, "loss": 0.0226, "step": 56790 }, { "grad_norm": 0.1810789257287979, "learning_rate": 2.128525885466546e-05, "loss": 0.0178, "step": 56800 }, { "grad_norm": 0.45567062497138977, "learning_rate": 2.1268341174969847e-05, "loss": 0.0205, "step": 56810 }, { "grad_norm": 0.2745405435562134, "learning_rate": 2.1251428404728894e-05, "loss": 0.0164, "step": 56820 }, { "grad_norm": 0.2305057793855667, "learning_rate": 2.123452054683256e-05, "loss": 0.0174, "step": 56830 }, { "grad_norm": 0.31520795822143555, "learning_rate": 2.1217617604169927e-05, "loss": 0.0218, "step": 56840 }, { "grad_norm": 0.44780829548835754, "learning_rate": 2.120071957962924e-05, "loss": 0.0219, "step": 56850 }, { "grad_norm": 0.25850412249565125, "learning_rate": 2.1183826476097917e-05, "loss": 0.0285, "step": 56860 }, { "grad_norm": 0.24719905853271484, "learning_rate": 2.116693829646254e-05, "loss": 0.0169, "step": 56870 }, { "grad_norm": 0.45027074217796326, "learning_rate": 2.1150055043608806e-05, "loss": 0.018, "step": 56880 }, { "grad_norm": 0.33791083097457886, "learning_rate": 2.113317672042162e-05, "loss": 0.0235, "step": 56890 }, { "grad_norm": 0.26458945870399475, "learning_rate": 2.1116303329785025e-05, "loss": 0.0178, "step": 56900 }, { "grad_norm": 0.4332471489906311, "learning_rate": 2.109943487458222e-05, "loss": 0.025, "step": 56910 }, { "grad_norm": 0.18276606500148773, "learning_rate": 2.1082571357695574e-05, "loss": 0.0184, "step": 56920 }, { "grad_norm": 0.2412535697221756, "learning_rate": 2.1065712782006557e-05, "loss": 0.0208, "step": 56930 }, { "grad_norm": 0.2952820658683777, "learning_rate": 2.104885915039591e-05, "loss": 0.0177, "step": 56940 }, { "grad_norm": 0.33397233486175537, "learning_rate": 2.103201046574338e-05, "loss": 0.0197, "step": 56950 }, { "grad_norm": 0.35885849595069885, "learning_rate": 2.1015166730928022e-05, "loss": 0.0172, "step": 56960 }, { "grad_norm": 0.26745864748954773, "learning_rate": 2.0998327948827912e-05, "loss": 0.0185, "step": 56970 }, { "grad_norm": 0.2602320611476898, "learning_rate": 2.0981494122320363e-05, "loss": 0.0199, "step": 56980 }, { "grad_norm": 0.2755548655986786, "learning_rate": 2.0964665254281822e-05, "loss": 0.022, "step": 56990 }, { "grad_norm": 0.29027578234672546, "learning_rate": 2.094784134758784e-05, "loss": 0.0213, "step": 57000 }, { "grad_norm": 0.26825442910194397, "learning_rate": 2.0931022405113226e-05, "loss": 0.0297, "step": 57010 }, { "grad_norm": 0.22393959760665894, "learning_rate": 2.091420842973183e-05, "loss": 0.026, "step": 57020 }, { "grad_norm": 0.18536017835140228, "learning_rate": 2.0897399424316715e-05, "loss": 0.0173, "step": 57030 }, { "grad_norm": 0.21496260166168213, "learning_rate": 2.0880595391740078e-05, "loss": 0.0221, "step": 57040 }, { "grad_norm": 0.21231715381145477, "learning_rate": 2.0863796334873277e-05, "loss": 0.0237, "step": 57050 }, { "grad_norm": 0.23818548023700714, "learning_rate": 2.08470022565868e-05, "loss": 0.0129, "step": 57060 }, { "grad_norm": 0.2347785383462906, "learning_rate": 2.0830213159750317e-05, "loss": 0.0156, "step": 57070 }, { "grad_norm": 0.2709754407405853, "learning_rate": 2.0813429047232596e-05, "loss": 0.018, "step": 57080 }, { "grad_norm": 0.2080024778842926, "learning_rate": 2.0796649921901594e-05, "loss": 0.0205, "step": 57090 }, { "grad_norm": 0.3014460802078247, "learning_rate": 2.077987578662441e-05, "loss": 0.0204, "step": 57100 }, { "grad_norm": 0.29359811544418335, "learning_rate": 2.0763106644267277e-05, "loss": 0.0247, "step": 57110 }, { "grad_norm": 0.2905443608760834, "learning_rate": 2.0746342497695607e-05, "loss": 0.0179, "step": 57120 }, { "grad_norm": 0.16148801147937775, "learning_rate": 2.0729583349773886e-05, "loss": 0.0131, "step": 57130 }, { "grad_norm": 0.2778623700141907, "learning_rate": 2.0712829203365853e-05, "loss": 0.0182, "step": 57140 }, { "grad_norm": 0.17139846086502075, "learning_rate": 2.0696080061334267e-05, "loss": 0.0179, "step": 57150 }, { "grad_norm": 0.22349268198013306, "learning_rate": 2.067933592654117e-05, "loss": 0.017, "step": 57160 }, { "grad_norm": 0.2902613580226898, "learning_rate": 2.066259680184763e-05, "loss": 0.0172, "step": 57170 }, { "grad_norm": 0.26715773344039917, "learning_rate": 2.0645862690113908e-05, "loss": 0.0217, "step": 57180 }, { "grad_norm": 0.29514655470848083, "learning_rate": 2.0629133594199436e-05, "loss": 0.0237, "step": 57190 }, { "grad_norm": 0.312926709651947, "learning_rate": 2.0612409516962704e-05, "loss": 0.0237, "step": 57200 }, { "grad_norm": 0.3494541049003601, "learning_rate": 2.0595690461261467e-05, "loss": 0.0192, "step": 57210 }, { "grad_norm": 0.37269145250320435, "learning_rate": 2.0578976429952503e-05, "loss": 0.0158, "step": 57220 }, { "grad_norm": 0.3671823740005493, "learning_rate": 2.0562267425891802e-05, "loss": 0.0211, "step": 57230 }, { "grad_norm": 0.27391424775123596, "learning_rate": 2.0545563451934467e-05, "loss": 0.0163, "step": 57240 }, { "grad_norm": 0.2194031924009323, "learning_rate": 2.0528864510934764e-05, "loss": 0.022, "step": 57250 }, { "grad_norm": 0.27352187037467957, "learning_rate": 2.0512170605746096e-05, "loss": 0.0286, "step": 57260 }, { "grad_norm": 0.28261786699295044, "learning_rate": 2.0495481739220963e-05, "loss": 0.0221, "step": 57270 }, { "grad_norm": 0.34532099962234497, "learning_rate": 2.0478797914211045e-05, "loss": 0.0229, "step": 57280 }, { "grad_norm": 0.33728528022766113, "learning_rate": 2.046211913356716e-05, "loss": 0.0216, "step": 57290 }, { "grad_norm": 0.17538094520568848, "learning_rate": 2.0445445400139247e-05, "loss": 0.0278, "step": 57300 }, { "grad_norm": 0.39846494793891907, "learning_rate": 2.0428776716776405e-05, "loss": 0.0179, "step": 57310 }, { "grad_norm": 0.3293435871601105, "learning_rate": 2.0412113086326856e-05, "loss": 0.0228, "step": 57320 }, { "grad_norm": 0.22896412014961243, "learning_rate": 2.0395454511637918e-05, "loss": 0.018, "step": 57330 }, { "grad_norm": 0.35125645995140076, "learning_rate": 2.037880099555616e-05, "loss": 0.0239, "step": 57340 }, { "grad_norm": 0.21810375154018402, "learning_rate": 2.0362152540927144e-05, "loss": 0.0182, "step": 57350 }, { "grad_norm": 0.298893541097641, "learning_rate": 2.0345509150595666e-05, "loss": 0.0281, "step": 57360 }, { "grad_norm": 0.21178841590881348, "learning_rate": 2.0328870827405617e-05, "loss": 0.0223, "step": 57370 }, { "grad_norm": 0.39912623167037964, "learning_rate": 2.0312237574200043e-05, "loss": 0.0216, "step": 57380 }, { "grad_norm": 0.265929251909256, "learning_rate": 2.029560939382112e-05, "loss": 0.02, "step": 57390 }, { "grad_norm": 0.20898570120334625, "learning_rate": 2.0278986289110097e-05, "loss": 0.0195, "step": 57400 }, { "grad_norm": 0.2410685271024704, "learning_rate": 2.0262368262907484e-05, "loss": 0.0201, "step": 57410 }, { "grad_norm": 0.34987807273864746, "learning_rate": 2.024575531805279e-05, "loss": 0.0241, "step": 57420 }, { "grad_norm": 0.21549451351165771, "learning_rate": 2.0229147457384735e-05, "loss": 0.0189, "step": 57430 }, { "grad_norm": 0.3018372058868408, "learning_rate": 2.0212544683741157e-05, "loss": 0.0397, "step": 57440 }, { "grad_norm": 0.35442954301834106, "learning_rate": 2.0195946999958976e-05, "loss": 0.0314, "step": 57450 }, { "grad_norm": 0.21555998921394348, "learning_rate": 2.017935440887434e-05, "loss": 0.0219, "step": 57460 }, { "grad_norm": 0.2507862150669098, "learning_rate": 2.0162766913322423e-05, "loss": 0.0196, "step": 57470 }, { "grad_norm": 0.23834754526615143, "learning_rate": 2.0146184516137588e-05, "loss": 0.0208, "step": 57480 }, { "grad_norm": 0.24280637502670288, "learning_rate": 2.012960722015332e-05, "loss": 0.0233, "step": 57490 }, { "grad_norm": 0.18876326084136963, "learning_rate": 2.0113035028202214e-05, "loss": 0.0205, "step": 57500 }, { "grad_norm": 0.25865378975868225, "learning_rate": 2.009646794311602e-05, "loss": 0.0224, "step": 57510 }, { "grad_norm": 0.18438972532749176, "learning_rate": 2.007990596772559e-05, "loss": 0.0189, "step": 57520 }, { "grad_norm": 0.3156030476093292, "learning_rate": 2.0063349104860923e-05, "loss": 0.0236, "step": 57530 }, { "grad_norm": 0.32265982031822205, "learning_rate": 2.0046797357351116e-05, "loss": 0.0306, "step": 57540 }, { "grad_norm": 0.1406739205121994, "learning_rate": 2.0030250728024412e-05, "loss": 0.0147, "step": 57550 }, { "grad_norm": 0.2253369688987732, "learning_rate": 2.001370921970819e-05, "loss": 0.0156, "step": 57560 }, { "grad_norm": 0.33639413118362427, "learning_rate": 1.9997172835228932e-05, "loss": 0.0217, "step": 57570 }, { "grad_norm": 0.3690180778503418, "learning_rate": 1.9980641577412262e-05, "loss": 0.02, "step": 57580 }, { "grad_norm": 0.28519406914711, "learning_rate": 1.9964115449082925e-05, "loss": 0.02, "step": 57590 }, { "grad_norm": 0.15045952796936035, "learning_rate": 1.9947594453064742e-05, "loss": 0.0277, "step": 57600 }, { "grad_norm": 0.1795876920223236, "learning_rate": 1.9931078592180774e-05, "loss": 0.02, "step": 57610 }, { "grad_norm": 0.3384062647819519, "learning_rate": 1.9914567869253065e-05, "loss": 0.02, "step": 57620 }, { "grad_norm": 0.21313445270061493, "learning_rate": 1.989806228710287e-05, "loss": 0.0248, "step": 57630 }, { "grad_norm": 0.3835389316082001, "learning_rate": 1.9881561848550555e-05, "loss": 0.0172, "step": 57640 }, { "grad_norm": 0.3077521026134491, "learning_rate": 1.9865066556415544e-05, "loss": 0.0219, "step": 57650 }, { "grad_norm": 0.3693576753139496, "learning_rate": 1.98485764135165e-05, "loss": 0.0198, "step": 57660 }, { "grad_norm": 0.7823641300201416, "learning_rate": 1.983209142267109e-05, "loss": 0.0173, "step": 57670 }, { "grad_norm": 0.23349347710609436, "learning_rate": 1.9815611586696165e-05, "loss": 0.0186, "step": 57680 }, { "grad_norm": 0.3392822742462158, "learning_rate": 1.9799136908407667e-05, "loss": 0.0209, "step": 57690 }, { "grad_norm": 0.18217985332012177, "learning_rate": 1.9782667390620678e-05, "loss": 0.0162, "step": 57700 }, { "grad_norm": 0.287094384431839, "learning_rate": 1.976620303614939e-05, "loss": 0.0179, "step": 57710 }, { "grad_norm": 0.35814571380615234, "learning_rate": 1.9749743847807108e-05, "loss": 0.0165, "step": 57720 }, { "grad_norm": 0.270842045545578, "learning_rate": 1.9733289828406272e-05, "loss": 0.0158, "step": 57730 }, { "grad_norm": 0.38901519775390625, "learning_rate": 1.9716840980758382e-05, "loss": 0.017, "step": 57740 }, { "grad_norm": 0.25726446509361267, "learning_rate": 1.9700397307674134e-05, "loss": 0.0173, "step": 57750 }, { "grad_norm": 0.319186806678772, "learning_rate": 1.968395881196328e-05, "loss": 0.0199, "step": 57760 }, { "grad_norm": 0.4496251046657562, "learning_rate": 1.966752549643473e-05, "loss": 0.0235, "step": 57770 }, { "grad_norm": 0.19496674835681915, "learning_rate": 1.965109736389647e-05, "loss": 0.0179, "step": 57780 }, { "grad_norm": 0.254999041557312, "learning_rate": 1.9634674417155645e-05, "loss": 0.0181, "step": 57790 }, { "grad_norm": 0.2557966113090515, "learning_rate": 1.9618256659018434e-05, "loss": 0.0339, "step": 57800 }, { "grad_norm": 0.38757747411727905, "learning_rate": 1.9601844092290257e-05, "loss": 0.0175, "step": 57810 }, { "grad_norm": 0.2262801229953766, "learning_rate": 1.9585436719775512e-05, "loss": 0.016, "step": 57820 }, { "grad_norm": 0.1541181057691574, "learning_rate": 1.9569034544277793e-05, "loss": 0.0186, "step": 57830 }, { "grad_norm": 0.23508498072624207, "learning_rate": 1.9552637568599798e-05, "loss": 0.0186, "step": 57840 }, { "grad_norm": 0.2917666435241699, "learning_rate": 1.953624579554327e-05, "loss": 0.0249, "step": 57850 }, { "grad_norm": 0.3433464765548706, "learning_rate": 1.951985922790918e-05, "loss": 0.0228, "step": 57860 }, { "grad_norm": 0.28017768263816833, "learning_rate": 1.9503477868497505e-05, "loss": 0.0158, "step": 57870 }, { "grad_norm": 0.38685956597328186, "learning_rate": 1.9487101720107375e-05, "loss": 0.0215, "step": 57880 }, { "grad_norm": 0.28735682368278503, "learning_rate": 1.9470730785537032e-05, "loss": 0.0186, "step": 57890 }, { "grad_norm": 0.31067341566085815, "learning_rate": 1.9454365067583823e-05, "loss": 0.0203, "step": 57900 }, { "grad_norm": 0.24404305219650269, "learning_rate": 1.9438004569044215e-05, "loss": 0.0153, "step": 57910 }, { "grad_norm": 0.20578202605247498, "learning_rate": 1.9421649292713724e-05, "loss": 0.037, "step": 57920 }, { "grad_norm": 0.28309065103530884, "learning_rate": 1.9405299241387076e-05, "loss": 0.0214, "step": 57930 }, { "grad_norm": 0.23788036406040192, "learning_rate": 1.9388954417858007e-05, "loss": 0.0206, "step": 57940 }, { "grad_norm": 0.49573925137519836, "learning_rate": 1.9372614824919417e-05, "loss": 0.0201, "step": 57950 }, { "grad_norm": 0.42312297224998474, "learning_rate": 1.9356280465363284e-05, "loss": 0.0193, "step": 57960 }, { "grad_norm": 0.2543971836566925, "learning_rate": 1.9339951341980723e-05, "loss": 0.0138, "step": 57970 }, { "grad_norm": 0.32876136898994446, "learning_rate": 1.9323627457561916e-05, "loss": 0.0159, "step": 57980 }, { "grad_norm": 0.2827005982398987, "learning_rate": 1.9307308814896198e-05, "loss": 0.0137, "step": 57990 }, { "grad_norm": 0.19929808378219604, "learning_rate": 1.9290995416771935e-05, "loss": 0.0184, "step": 58000 }, { "grad_norm": 0.36392688751220703, "learning_rate": 1.9274687265976665e-05, "loss": 0.0179, "step": 58010 }, { "grad_norm": 0.13912473618984222, "learning_rate": 1.9258384365297e-05, "loss": 0.0198, "step": 58020 }, { "grad_norm": 0.24551965296268463, "learning_rate": 1.924208671751866e-05, "loss": 0.0204, "step": 58030 }, { "grad_norm": 0.4172170162200928, "learning_rate": 1.9225794325426492e-05, "loss": 0.019, "step": 58040 }, { "grad_norm": 0.29276296496391296, "learning_rate": 1.920950719180436e-05, "loss": 0.0172, "step": 58050 }, { "grad_norm": 0.16248130798339844, "learning_rate": 1.919322531943536e-05, "loss": 0.0188, "step": 58060 }, { "grad_norm": 0.25500738620758057, "learning_rate": 1.917694871110157e-05, "loss": 0.0215, "step": 58070 }, { "grad_norm": 0.3445356786251068, "learning_rate": 1.9160677369584234e-05, "loss": 0.0156, "step": 58080 }, { "grad_norm": 0.2612844705581665, "learning_rate": 1.9144411297663694e-05, "loss": 0.0186, "step": 58090 }, { "grad_norm": 0.15104882419109344, "learning_rate": 1.9128150498119328e-05, "loss": 0.0209, "step": 58100 }, { "grad_norm": 0.24771355092525482, "learning_rate": 1.9111894973729726e-05, "loss": 0.014, "step": 58110 }, { "grad_norm": 0.35358139872550964, "learning_rate": 1.9095644727272454e-05, "loss": 0.0263, "step": 58120 }, { "grad_norm": 0.24560904502868652, "learning_rate": 1.907939976152429e-05, "loss": 0.0204, "step": 58130 }, { "grad_norm": 0.3745304048061371, "learning_rate": 1.906316007926101e-05, "loss": 0.0214, "step": 58140 }, { "grad_norm": 0.21165473759174347, "learning_rate": 1.904692568325755e-05, "loss": 0.0171, "step": 58150 }, { "grad_norm": 0.330952912569046, "learning_rate": 1.9030696576287925e-05, "loss": 0.0228, "step": 58160 }, { "grad_norm": 0.3414957523345947, "learning_rate": 1.9014472761125242e-05, "loss": 0.0159, "step": 58170 }, { "grad_norm": 0.1502486914396286, "learning_rate": 1.899825424054172e-05, "loss": 0.021, "step": 58180 }, { "grad_norm": 0.2093658596277237, "learning_rate": 1.898204101730863e-05, "loss": 0.018, "step": 58190 }, { "grad_norm": 0.22773748636245728, "learning_rate": 1.8965833094196394e-05, "loss": 0.0358, "step": 58200 }, { "grad_norm": 0.3980037271976471, "learning_rate": 1.8949630473974495e-05, "loss": 0.0171, "step": 58210 }, { "grad_norm": 0.22955362498760223, "learning_rate": 1.8933433159411517e-05, "loss": 0.0216, "step": 58220 }, { "grad_norm": 0.24660800397396088, "learning_rate": 1.891724115327514e-05, "loss": 0.0182, "step": 58230 }, { "grad_norm": 0.3606422543525696, "learning_rate": 1.8901054458332152e-05, "loss": 0.027, "step": 58240 }, { "grad_norm": 0.20834265649318695, "learning_rate": 1.8884873077348364e-05, "loss": 0.0218, "step": 58250 }, { "grad_norm": 0.2909124791622162, "learning_rate": 1.88686970130888e-05, "loss": 0.0213, "step": 58260 }, { "grad_norm": 0.24792081117630005, "learning_rate": 1.8852526268317456e-05, "loss": 0.0216, "step": 58270 }, { "grad_norm": 0.35827577114105225, "learning_rate": 1.883636084579749e-05, "loss": 0.0205, "step": 58280 }, { "grad_norm": 0.3671056926250458, "learning_rate": 1.8820200748291145e-05, "loss": 0.0222, "step": 58290 }, { "grad_norm": 0.1969919055700302, "learning_rate": 1.8804045978559686e-05, "loss": 0.0204, "step": 58300 }, { "grad_norm": 0.3233807682991028, "learning_rate": 1.8787896539363594e-05, "loss": 0.0186, "step": 58310 }, { "grad_norm": 0.2716097831726074, "learning_rate": 1.8771752433462298e-05, "loss": 0.0181, "step": 58320 }, { "grad_norm": 0.29378923773765564, "learning_rate": 1.8755613663614447e-05, "loss": 0.0191, "step": 58330 }, { "grad_norm": 0.36573347449302673, "learning_rate": 1.873948023257767e-05, "loss": 0.0178, "step": 58340 }, { "grad_norm": 0.26567429304122925, "learning_rate": 1.8723352143108747e-05, "loss": 0.0174, "step": 58350 }, { "grad_norm": 0.4659031629562378, "learning_rate": 1.870722939796352e-05, "loss": 0.0171, "step": 58360 }, { "grad_norm": 0.3033443093299866, "learning_rate": 1.869111199989693e-05, "loss": 0.0171, "step": 58370 }, { "grad_norm": 0.2530156970024109, "learning_rate": 1.8674999951663018e-05, "loss": 0.0185, "step": 58380 }, { "grad_norm": 0.23504379391670227, "learning_rate": 1.8658893256014857e-05, "loss": 0.0192, "step": 58390 }, { "grad_norm": 0.2448907345533371, "learning_rate": 1.8642791915704655e-05, "loss": 0.0169, "step": 58400 }, { "grad_norm": 0.26944220066070557, "learning_rate": 1.8626695933483697e-05, "loss": 0.0181, "step": 58410 }, { "grad_norm": 0.22711703181266785, "learning_rate": 1.8610605312102335e-05, "loss": 0.0234, "step": 58420 }, { "grad_norm": 0.2883249819278717, "learning_rate": 1.8594520054310032e-05, "loss": 0.0286, "step": 58430 }, { "grad_norm": 0.18531641364097595, "learning_rate": 1.8578440162855326e-05, "loss": 0.0166, "step": 58440 }, { "grad_norm": 0.1400091052055359, "learning_rate": 1.8562365640485774e-05, "loss": 0.0208, "step": 58450 }, { "grad_norm": 0.12389209866523743, "learning_rate": 1.854629648994815e-05, "loss": 0.0178, "step": 58460 }, { "grad_norm": 0.2482842653989792, "learning_rate": 1.8530232713988183e-05, "loss": 0.0175, "step": 58470 }, { "grad_norm": 0.40238699316978455, "learning_rate": 1.851417431535074e-05, "loss": 0.0172, "step": 58480 }, { "grad_norm": 0.28835758566856384, "learning_rate": 1.8498121296779785e-05, "loss": 0.0328, "step": 58490 }, { "grad_norm": 0.22627240419387817, "learning_rate": 1.848207366101829e-05, "loss": 0.024, "step": 58500 }, { "grad_norm": 0.23345191776752472, "learning_rate": 1.8466031410808422e-05, "loss": 0.0149, "step": 58510 }, { "grad_norm": 0.3413105905056, "learning_rate": 1.844999454889129e-05, "loss": 0.0235, "step": 58520 }, { "grad_norm": 0.4175876975059509, "learning_rate": 1.8433963078007237e-05, "loss": 0.0225, "step": 58530 }, { "grad_norm": 0.28431376814842224, "learning_rate": 1.841793700089554e-05, "loss": 0.0182, "step": 58540 }, { "grad_norm": 0.23826864361763, "learning_rate": 1.8401916320294644e-05, "loss": 0.0166, "step": 58550 }, { "grad_norm": 0.30853214859962463, "learning_rate": 1.838590103894205e-05, "loss": 0.0165, "step": 58560 }, { "grad_norm": 0.27111974358558655, "learning_rate": 1.8369891159574292e-05, "loss": 0.0189, "step": 58570 }, { "grad_norm": 0.252015084028244, "learning_rate": 1.835388668492708e-05, "loss": 0.0225, "step": 58580 }, { "grad_norm": 0.19275444746017456, "learning_rate": 1.8337887617735095e-05, "loss": 0.0169, "step": 58590 }, { "grad_norm": 0.21732576191425323, "learning_rate": 1.8321893960732157e-05, "loss": 0.0186, "step": 58600 }, { "grad_norm": 0.28616029024124146, "learning_rate": 1.8305905716651138e-05, "loss": 0.0162, "step": 58610 }, { "grad_norm": 0.43195047974586487, "learning_rate": 1.8289922888224e-05, "loss": 0.02, "step": 58620 }, { "grad_norm": 0.2937805652618408, "learning_rate": 1.8273945478181765e-05, "loss": 0.0213, "step": 58630 }, { "grad_norm": 0.2986987829208374, "learning_rate": 1.8257973489254558e-05, "loss": 0.0177, "step": 58640 }, { "grad_norm": 0.33707624673843384, "learning_rate": 1.824200692417152e-05, "loss": 0.0188, "step": 58650 }, { "grad_norm": 0.2845839560031891, "learning_rate": 1.8226045785660912e-05, "loss": 0.0177, "step": 58660 }, { "grad_norm": 0.3304685354232788, "learning_rate": 1.821009007645006e-05, "loss": 0.0168, "step": 58670 }, { "grad_norm": 0.2802976071834564, "learning_rate": 1.8194139799265357e-05, "loss": 0.017, "step": 58680 }, { "grad_norm": 0.4258672893047333, "learning_rate": 1.8178194956832295e-05, "loss": 0.0243, "step": 58690 }, { "grad_norm": 0.18069973587989807, "learning_rate": 1.8162255551875346e-05, "loss": 0.017, "step": 58700 }, { "grad_norm": 0.4358164072036743, "learning_rate": 1.8146321587118194e-05, "loss": 0.0255, "step": 58710 }, { "grad_norm": 0.22036853432655334, "learning_rate": 1.8130393065283448e-05, "loss": 0.0148, "step": 58720 }, { "grad_norm": 0.22803246974945068, "learning_rate": 1.8114469989092925e-05, "loss": 0.0171, "step": 58730 }, { "grad_norm": 0.22879646718502045, "learning_rate": 1.8098552361267397e-05, "loss": 0.0157, "step": 58740 }, { "grad_norm": 0.30488550662994385, "learning_rate": 1.8082640184526763e-05, "loss": 0.0159, "step": 58750 }, { "grad_norm": 0.317101389169693, "learning_rate": 1.806673346158999e-05, "loss": 0.0162, "step": 58760 }, { "grad_norm": 0.2966046631336212, "learning_rate": 1.8050832195175067e-05, "loss": 0.0219, "step": 58770 }, { "grad_norm": 0.31519240140914917, "learning_rate": 1.8034936387999136e-05, "loss": 0.0192, "step": 58780 }, { "grad_norm": 0.21722312271595, "learning_rate": 1.8019046042778315e-05, "loss": 0.0205, "step": 58790 }, { "grad_norm": 0.25206470489501953, "learning_rate": 1.800316116222785e-05, "loss": 0.017, "step": 58800 }, { "grad_norm": 0.1299126297235489, "learning_rate": 1.7987281749062018e-05, "loss": 0.0349, "step": 58810 }, { "grad_norm": 0.17017222940921783, "learning_rate": 1.7971407805994195e-05, "loss": 0.0294, "step": 58820 }, { "grad_norm": 0.12423186004161835, "learning_rate": 1.7955539335736787e-05, "loss": 0.0141, "step": 58830 }, { "grad_norm": 0.4076095521450043, "learning_rate": 1.7939676341001304e-05, "loss": 0.0223, "step": 58840 }, { "grad_norm": 0.18624964356422424, "learning_rate": 1.7923818824498275e-05, "loss": 0.0153, "step": 58850 }, { "grad_norm": 0.22409707307815552, "learning_rate": 1.7907966788937315e-05, "loss": 0.0214, "step": 58860 }, { "grad_norm": 0.19811707735061646, "learning_rate": 1.7892120237027116e-05, "loss": 0.0209, "step": 58870 }, { "grad_norm": 0.206318199634552, "learning_rate": 1.7876279171475413e-05, "loss": 0.0198, "step": 58880 }, { "grad_norm": 0.27743980288505554, "learning_rate": 1.7860443594989028e-05, "loss": 0.0174, "step": 58890 }, { "grad_norm": 0.319833904504776, "learning_rate": 1.784461351027379e-05, "loss": 0.0159, "step": 58900 }, { "grad_norm": 0.3422689139842987, "learning_rate": 1.7828788920034677e-05, "loss": 0.0203, "step": 58910 }, { "grad_norm": 0.3859159052371979, "learning_rate": 1.7812969826975623e-05, "loss": 0.021, "step": 58920 }, { "grad_norm": 0.23076902329921722, "learning_rate": 1.7797156233799738e-05, "loss": 0.0222, "step": 58930 }, { "grad_norm": 0.1700349748134613, "learning_rate": 1.7781348143209094e-05, "loss": 0.0187, "step": 58940 }, { "grad_norm": 0.2891952693462372, "learning_rate": 1.7765545557904862e-05, "loss": 0.0302, "step": 58950 }, { "grad_norm": 0.4048326909542084, "learning_rate": 1.7749748480587302e-05, "loss": 0.0197, "step": 58960 }, { "grad_norm": 0.23847752809524536, "learning_rate": 1.773395691395564e-05, "loss": 0.0197, "step": 58970 }, { "grad_norm": 0.20770767331123352, "learning_rate": 1.7718170860708305e-05, "loss": 0.0156, "step": 58980 }, { "grad_norm": 0.398637056350708, "learning_rate": 1.770239032354264e-05, "loss": 0.0161, "step": 58990 }, { "grad_norm": 0.3308981657028198, "learning_rate": 1.7686615305155124e-05, "loss": 0.0295, "step": 59000 }, { "grad_norm": 0.28201907873153687, "learning_rate": 1.767084580824128e-05, "loss": 0.0148, "step": 59010 }, { "grad_norm": 0.5500414967536926, "learning_rate": 1.765508183549569e-05, "loss": 0.0238, "step": 59020 }, { "grad_norm": 0.2793124318122864, "learning_rate": 1.763932338961199e-05, "loss": 0.0183, "step": 59030 }, { "grad_norm": 0.2106088548898697, "learning_rate": 1.762357047328284e-05, "loss": 0.0158, "step": 59040 }, { "grad_norm": 0.4212256371974945, "learning_rate": 1.7607823089199997e-05, "loss": 0.0248, "step": 59050 }, { "grad_norm": 0.27705463767051697, "learning_rate": 1.7592081240054265e-05, "loss": 0.0218, "step": 59060 }, { "grad_norm": 0.4218805432319641, "learning_rate": 1.7576344928535488e-05, "loss": 0.0194, "step": 59070 }, { "grad_norm": 0.3070235252380371, "learning_rate": 1.7560614157332573e-05, "loss": 0.0181, "step": 59080 }, { "grad_norm": 0.212284654378891, "learning_rate": 1.7544888929133495e-05, "loss": 0.0212, "step": 59090 }, { "grad_norm": 0.2898494601249695, "learning_rate": 1.752916924662522e-05, "loss": 0.0184, "step": 59100 }, { "grad_norm": 0.3183417320251465, "learning_rate": 1.751345511249387e-05, "loss": 0.018, "step": 59110 }, { "grad_norm": 0.21250976622104645, "learning_rate": 1.749774652942452e-05, "loss": 0.0176, "step": 59120 }, { "grad_norm": 0.26950719952583313, "learning_rate": 1.748204350010135e-05, "loss": 0.0211, "step": 59130 }, { "grad_norm": 0.2532559037208557, "learning_rate": 1.7466346027207574e-05, "loss": 0.017, "step": 59140 }, { "grad_norm": 0.44334501028060913, "learning_rate": 1.745065411342547e-05, "loss": 0.0189, "step": 59150 }, { "grad_norm": 0.3226482570171356, "learning_rate": 1.7434967761436366e-05, "loss": 0.0247, "step": 59160 }, { "grad_norm": 0.36995190382003784, "learning_rate": 1.741928697392058e-05, "loss": 0.018, "step": 59170 }, { "grad_norm": 0.3659702241420746, "learning_rate": 1.7403611753557597e-05, "loss": 0.0173, "step": 59180 }, { "grad_norm": 0.2975819706916809, "learning_rate": 1.738794210302584e-05, "loss": 0.0146, "step": 59190 }, { "grad_norm": 0.15851807594299316, "learning_rate": 1.7372278025002837e-05, "loss": 0.0164, "step": 59200 }, { "grad_norm": 0.30392980575561523, "learning_rate": 1.7356619522165164e-05, "loss": 0.0178, "step": 59210 }, { "grad_norm": 0.2423694133758545, "learning_rate": 1.7340966597188378e-05, "loss": 0.0157, "step": 59220 }, { "grad_norm": 0.17012134194374084, "learning_rate": 1.732531925274722e-05, "loss": 0.0189, "step": 59230 }, { "grad_norm": 0.20194603502750397, "learning_rate": 1.7309677491515318e-05, "loss": 0.018, "step": 59240 }, { "grad_norm": 0.27433615922927856, "learning_rate": 1.7294041316165455e-05, "loss": 0.0155, "step": 59250 }, { "grad_norm": 0.3623974323272705, "learning_rate": 1.727841072936942e-05, "loss": 0.0196, "step": 59260 }, { "grad_norm": 0.2578783631324768, "learning_rate": 1.7262785733798058e-05, "loss": 0.015, "step": 59270 }, { "grad_norm": 0.3961235284805298, "learning_rate": 1.7247166332121246e-05, "loss": 0.0223, "step": 59280 }, { "grad_norm": 0.4054250121116638, "learning_rate": 1.7231552527007933e-05, "loss": 0.0195, "step": 59290 }, { "grad_norm": 0.2540391981601715, "learning_rate": 1.721594432112606e-05, "loss": 0.0163, "step": 59300 }, { "grad_norm": 0.29574811458587646, "learning_rate": 1.7200341717142655e-05, "loss": 0.0243, "step": 59310 }, { "grad_norm": 0.2911272644996643, "learning_rate": 1.718474471772377e-05, "loss": 0.0211, "step": 59320 }, { "grad_norm": 0.23949509859085083, "learning_rate": 1.7169153325534528e-05, "loss": 0.0165, "step": 59330 }, { "grad_norm": 0.21558859944343567, "learning_rate": 1.7153567543239045e-05, "loss": 0.0156, "step": 59340 }, { "grad_norm": 0.3007288873195648, "learning_rate": 1.7137987373500525e-05, "loss": 0.0164, "step": 59350 }, { "grad_norm": 0.2464645802974701, "learning_rate": 1.7122412818981198e-05, "loss": 0.0189, "step": 59360 }, { "grad_norm": 0.4138014316558838, "learning_rate": 1.7106843882342283e-05, "loss": 0.0199, "step": 59370 }, { "grad_norm": 0.12487170845270157, "learning_rate": 1.709128056624415e-05, "loss": 0.0152, "step": 59380 }, { "grad_norm": 0.2780168950557709, "learning_rate": 1.70757228733461e-05, "loss": 0.0156, "step": 59390 }, { "grad_norm": 0.19745804369449615, "learning_rate": 1.706017080630653e-05, "loss": 0.0165, "step": 59400 }, { "grad_norm": 0.3772984743118286, "learning_rate": 1.7044624367782873e-05, "loss": 0.0191, "step": 59410 }, { "grad_norm": 0.15286649763584137, "learning_rate": 1.7029083560431553e-05, "loss": 0.0184, "step": 59420 }, { "grad_norm": 0.23267163336277008, "learning_rate": 1.701354838690813e-05, "loss": 0.0206, "step": 59430 }, { "grad_norm": 0.2241603285074234, "learning_rate": 1.6998018849867087e-05, "loss": 0.0179, "step": 59440 }, { "grad_norm": 0.1832296997308731, "learning_rate": 1.698249495196202e-05, "loss": 0.0149, "step": 59450 }, { "grad_norm": 0.2109879106283188, "learning_rate": 1.6966976695845528e-05, "loss": 0.019, "step": 59460 }, { "grad_norm": 0.24989613890647888, "learning_rate": 1.6951464084169268e-05, "loss": 0.0155, "step": 59470 }, { "grad_norm": 0.46159934997558594, "learning_rate": 1.6935957119583916e-05, "loss": 0.0215, "step": 59480 }, { "grad_norm": 0.1338990181684494, "learning_rate": 1.6920455804739205e-05, "loss": 0.015, "step": 59490 }, { "grad_norm": 0.25761860609054565, "learning_rate": 1.6904960142283855e-05, "loss": 0.0198, "step": 59500 }, { "grad_norm": 0.15555135905742645, "learning_rate": 1.6889470134865666e-05, "loss": 0.0165, "step": 59510 }, { "grad_norm": 0.21288371086120605, "learning_rate": 1.687398578513145e-05, "loss": 0.0174, "step": 59520 }, { "grad_norm": 0.16954153776168823, "learning_rate": 1.6858507095727066e-05, "loss": 0.0181, "step": 59530 }, { "grad_norm": 0.24640578031539917, "learning_rate": 1.6843034069297403e-05, "loss": 0.016, "step": 59540 }, { "grad_norm": 0.3730843961238861, "learning_rate": 1.682756670848637e-05, "loss": 0.0292, "step": 59550 }, { "grad_norm": 0.3251771628856659, "learning_rate": 1.6812105015936937e-05, "loss": 0.0161, "step": 59560 }, { "grad_norm": 0.2816565930843353, "learning_rate": 1.6796648994291027e-05, "loss": 0.0252, "step": 59570 }, { "grad_norm": 0.1345682591199875, "learning_rate": 1.678119864618973e-05, "loss": 0.0135, "step": 59580 }, { "grad_norm": 0.2853657305240631, "learning_rate": 1.6765753974273023e-05, "loss": 0.0109, "step": 59590 }, { "grad_norm": 0.15286360681056976, "learning_rate": 1.675031498118001e-05, "loss": 0.0142, "step": 59600 }, { "grad_norm": 0.43498262763023376, "learning_rate": 1.6734881669548804e-05, "loss": 0.0153, "step": 59610 }, { "grad_norm": 0.20211005210876465, "learning_rate": 1.671945404201648e-05, "loss": 0.0224, "step": 59620 }, { "grad_norm": 0.48475182056427, "learning_rate": 1.670403210121927e-05, "loss": 0.0191, "step": 59630 }, { "grad_norm": 0.2601301670074463, "learning_rate": 1.6688615849792312e-05, "loss": 0.0168, "step": 59640 }, { "grad_norm": 0.3670455515384674, "learning_rate": 1.6673205290369832e-05, "loss": 0.025, "step": 59650 }, { "grad_norm": 0.33290526270866394, "learning_rate": 1.6657800425585086e-05, "loss": 0.0216, "step": 59660 }, { "grad_norm": 0.35659539699554443, "learning_rate": 1.6642401258070327e-05, "loss": 0.0307, "step": 59670 }, { "grad_norm": 0.3185548782348633, "learning_rate": 1.6627007790456884e-05, "loss": 0.0236, "step": 59680 }, { "grad_norm": 0.1826833039522171, "learning_rate": 1.6611620025375037e-05, "loss": 0.0155, "step": 59690 }, { "grad_norm": 0.24005433917045593, "learning_rate": 1.6596237965454154e-05, "loss": 0.0216, "step": 59700 }, { "grad_norm": 0.40912336111068726, "learning_rate": 1.6580861613322607e-05, "loss": 0.0237, "step": 59710 }, { "grad_norm": 0.21835766732692719, "learning_rate": 1.6565490971607796e-05, "loss": 0.0181, "step": 59720 }, { "grad_norm": 0.23041804134845734, "learning_rate": 1.655012604293614e-05, "loss": 0.0171, "step": 59730 }, { "grad_norm": 0.17199446260929108, "learning_rate": 1.653476682993309e-05, "loss": 0.024, "step": 59740 }, { "grad_norm": 0.3257841467857361, "learning_rate": 1.651941333522311e-05, "loss": 0.02, "step": 59750 }, { "grad_norm": 0.3586983382701874, "learning_rate": 1.6504065561429715e-05, "loss": 0.0247, "step": 59760 }, { "grad_norm": 0.30988821387290955, "learning_rate": 1.6488723511175385e-05, "loss": 0.0227, "step": 59770 }, { "grad_norm": 0.27981653809547424, "learning_rate": 1.6473387187081668e-05, "loss": 0.0176, "step": 59780 }, { "grad_norm": 0.19169099628925323, "learning_rate": 1.6458056591769123e-05, "loss": 0.0173, "step": 59790 }, { "grad_norm": 0.36244556307792664, "learning_rate": 1.6442731727857335e-05, "loss": 0.0192, "step": 59800 }, { "grad_norm": 0.26448747515678406, "learning_rate": 1.642741259796492e-05, "loss": 0.0141, "step": 59810 }, { "grad_norm": 0.2034919112920761, "learning_rate": 1.641209920470944e-05, "loss": 0.0189, "step": 59820 }, { "grad_norm": 0.20843614637851715, "learning_rate": 1.639679155070762e-05, "loss": 0.0165, "step": 59830 }, { "grad_norm": 0.3957218825817108, "learning_rate": 1.6381489638575048e-05, "loss": 0.0177, "step": 59840 }, { "grad_norm": 0.2586445212364197, "learning_rate": 1.636619347092643e-05, "loss": 0.0176, "step": 59850 }, { "grad_norm": 0.4632699489593506, "learning_rate": 1.6350903050375476e-05, "loss": 0.0196, "step": 59860 }, { "grad_norm": 0.3258873522281647, "learning_rate": 1.6335618379534856e-05, "loss": 0.0209, "step": 59870 }, { "grad_norm": 0.24582374095916748, "learning_rate": 1.6320339461016364e-05, "loss": 0.0151, "step": 59880 }, { "grad_norm": 0.17532837390899658, "learning_rate": 1.6305066297430687e-05, "loss": 0.0177, "step": 59890 }, { "grad_norm": 0.4035719633102417, "learning_rate": 1.6289798891387654e-05, "loss": 0.0222, "step": 59900 }, { "grad_norm": 0.3377779424190521, "learning_rate": 1.6274537245495995e-05, "loss": 0.0152, "step": 59910 }, { "grad_norm": 0.20877934992313385, "learning_rate": 1.6259281362363527e-05, "loss": 0.017, "step": 59920 }, { "grad_norm": 0.2505120635032654, "learning_rate": 1.6244031244597068e-05, "loss": 0.0254, "step": 59930 }, { "grad_norm": 0.2785097062587738, "learning_rate": 1.622878689480244e-05, "loss": 0.0191, "step": 59940 }, { "grad_norm": 0.3631170988082886, "learning_rate": 1.6213548315584498e-05, "loss": 0.0242, "step": 59950 }, { "grad_norm": 0.3183939754962921, "learning_rate": 1.6198315509547074e-05, "loss": 0.0194, "step": 59960 }, { "grad_norm": 0.322325199842453, "learning_rate": 1.6183088479293056e-05, "loss": 0.0156, "step": 59970 }, { "grad_norm": 0.28899118304252625, "learning_rate": 1.6167867227424316e-05, "loss": 0.0173, "step": 59980 }, { "grad_norm": 0.1793215125799179, "learning_rate": 1.615265175654176e-05, "loss": 0.0229, "step": 59990 }, { "grad_norm": 0.2933169901371002, "learning_rate": 1.613744206924529e-05, "loss": 0.0184, "step": 60000 }, { "grad_norm": 0.2365458458662033, "learning_rate": 1.6122238168133845e-05, "loss": 0.0164, "step": 60010 }, { "grad_norm": 0.16610728204250336, "learning_rate": 1.6107040055805305e-05, "loss": 0.0139, "step": 60020 }, { "grad_norm": 0.3341626226902008, "learning_rate": 1.6091847734856673e-05, "loss": 0.0168, "step": 60030 }, { "grad_norm": 0.2762663662433624, "learning_rate": 1.6076661207883865e-05, "loss": 0.0172, "step": 60040 }, { "grad_norm": 0.20409590005874634, "learning_rate": 1.6061480477481848e-05, "loss": 0.0157, "step": 60050 }, { "grad_norm": 0.3685643672943115, "learning_rate": 1.604630554624461e-05, "loss": 0.0151, "step": 60060 }, { "grad_norm": 0.21101896464824677, "learning_rate": 1.603113641676509e-05, "loss": 0.0176, "step": 60070 }, { "grad_norm": 0.23108690977096558, "learning_rate": 1.6015973091635338e-05, "loss": 0.0143, "step": 60080 }, { "grad_norm": 0.1319497972726822, "learning_rate": 1.6000815573446288e-05, "loss": 0.0152, "step": 60090 }, { "grad_norm": 0.3501098155975342, "learning_rate": 1.5985663864788002e-05, "loss": 0.0252, "step": 60100 }, { "grad_norm": 0.299599289894104, "learning_rate": 1.597051796824946e-05, "loss": 0.0249, "step": 60110 }, { "grad_norm": 0.30804774165153503, "learning_rate": 1.5955377886418682e-05, "loss": 0.0194, "step": 60120 }, { "grad_norm": 0.2569340467453003, "learning_rate": 1.5940243621882704e-05, "loss": 0.018, "step": 60130 }, { "grad_norm": 0.22271735966205597, "learning_rate": 1.5925115177227555e-05, "loss": 0.0219, "step": 60140 }, { "grad_norm": 0.49567005038261414, "learning_rate": 1.5909992555038288e-05, "loss": 0.0207, "step": 60150 }, { "grad_norm": 0.3761115074157715, "learning_rate": 1.5894875757898912e-05, "loss": 0.0269, "step": 60160 }, { "grad_norm": 0.34998801350593567, "learning_rate": 1.58797647883925e-05, "loss": 0.0187, "step": 60170 }, { "grad_norm": 0.17842870950698853, "learning_rate": 1.586465964910109e-05, "loss": 0.0203, "step": 60180 }, { "grad_norm": 0.26497745513916016, "learning_rate": 1.5849560342605734e-05, "loss": 0.0224, "step": 60190 }, { "grad_norm": 0.4445621073246002, "learning_rate": 1.583446687148651e-05, "loss": 0.0147, "step": 60200 }, { "grad_norm": 0.23535290360450745, "learning_rate": 1.581937923832248e-05, "loss": 0.0177, "step": 60210 }, { "grad_norm": 0.15870165824890137, "learning_rate": 1.5804297445691663e-05, "loss": 0.0204, "step": 60220 }, { "grad_norm": 0.3516395390033722, "learning_rate": 1.578922149617119e-05, "loss": 0.022, "step": 60230 }, { "grad_norm": 0.26230332255363464, "learning_rate": 1.5774151392337084e-05, "loss": 0.0147, "step": 60240 }, { "grad_norm": 0.4372745156288147, "learning_rate": 1.575908713676442e-05, "loss": 0.0199, "step": 60250 }, { "grad_norm": 0.2841051518917084, "learning_rate": 1.5744028732027293e-05, "loss": 0.0219, "step": 60260 }, { "grad_norm": 0.3102153539657593, "learning_rate": 1.5728976180698723e-05, "loss": 0.0196, "step": 60270 }, { "grad_norm": 0.24634923040866852, "learning_rate": 1.5713929485350836e-05, "loss": 0.019, "step": 60280 }, { "grad_norm": 0.2585832476615906, "learning_rate": 1.569888864855465e-05, "loss": 0.0194, "step": 60290 }, { "grad_norm": 0.1637200564146042, "learning_rate": 1.568385367288028e-05, "loss": 0.0182, "step": 60300 }, { "grad_norm": 0.2852608561515808, "learning_rate": 1.5668824560896755e-05, "loss": 0.0172, "step": 60310 }, { "grad_norm": 0.21118396520614624, "learning_rate": 1.5653801315172156e-05, "loss": 0.0193, "step": 60320 }, { "grad_norm": 0.2586810290813446, "learning_rate": 1.5638783938273554e-05, "loss": 0.0189, "step": 60330 }, { "grad_norm": 0.2892961800098419, "learning_rate": 1.5623772432766966e-05, "loss": 0.0245, "step": 60340 }, { "grad_norm": 0.2589518427848816, "learning_rate": 1.5608766801217507e-05, "loss": 0.0169, "step": 60350 }, { "grad_norm": 0.21281969547271729, "learning_rate": 1.5593767046189184e-05, "loss": 0.0215, "step": 60360 }, { "grad_norm": 0.24498526751995087, "learning_rate": 1.5578773170245064e-05, "loss": 0.0145, "step": 60370 }, { "grad_norm": 0.2872294783592224, "learning_rate": 1.5563785175947182e-05, "loss": 0.0213, "step": 60380 }, { "grad_norm": 0.2644209563732147, "learning_rate": 1.5548803065856582e-05, "loss": 0.0184, "step": 60390 }, { "grad_norm": 0.42925527691841125, "learning_rate": 1.55338268425333e-05, "loss": 0.0195, "step": 60400 }, { "grad_norm": 0.20108842849731445, "learning_rate": 1.5518856508536373e-05, "loss": 0.0202, "step": 60410 }, { "grad_norm": 0.34154728055000305, "learning_rate": 1.5503892066423786e-05, "loss": 0.0227, "step": 60420 }, { "grad_norm": 0.3085213601589203, "learning_rate": 1.548893351875258e-05, "loss": 0.0293, "step": 60430 }, { "grad_norm": 0.28093549609184265, "learning_rate": 1.547398086807875e-05, "loss": 0.0222, "step": 60440 }, { "grad_norm": 0.4129672646522522, "learning_rate": 1.5459034116957305e-05, "loss": 0.0271, "step": 60450 }, { "grad_norm": 0.28606322407722473, "learning_rate": 1.544409326794225e-05, "loss": 0.0254, "step": 60460 }, { "grad_norm": 0.3400273323059082, "learning_rate": 1.542915832358651e-05, "loss": 0.022, "step": 60470 }, { "grad_norm": 0.31788647174835205, "learning_rate": 1.541422928644214e-05, "loss": 0.0164, "step": 60480 }, { "grad_norm": 0.39245080947875977, "learning_rate": 1.5399306159060024e-05, "loss": 0.024, "step": 60490 }, { "grad_norm": 0.17161010205745697, "learning_rate": 1.5384388943990185e-05, "loss": 0.0142, "step": 60500 }, { "grad_norm": 0.3770642578601837, "learning_rate": 1.5369477643781526e-05, "loss": 0.0194, "step": 60510 }, { "grad_norm": 0.23114396631717682, "learning_rate": 1.5354572260981985e-05, "loss": 0.0184, "step": 60520 }, { "grad_norm": 0.3430136740207672, "learning_rate": 1.533967279813851e-05, "loss": 0.0135, "step": 60530 }, { "grad_norm": 0.3602962791919708, "learning_rate": 1.5324779257796956e-05, "loss": 0.0218, "step": 60540 }, { "grad_norm": 0.31197744607925415, "learning_rate": 1.5309891642502293e-05, "loss": 0.0167, "step": 60550 }, { "grad_norm": 0.2611826956272125, "learning_rate": 1.5295009954798357e-05, "loss": 0.0135, "step": 60560 }, { "grad_norm": 0.31563758850097656, "learning_rate": 1.5280134197228036e-05, "loss": 0.021, "step": 60570 }, { "grad_norm": 0.1799880713224411, "learning_rate": 1.526526437233319e-05, "loss": 0.0159, "step": 60580 }, { "grad_norm": 0.23312298953533173, "learning_rate": 1.525040048265466e-05, "loss": 0.0179, "step": 60590 }, { "grad_norm": 0.2649540901184082, "learning_rate": 1.5235542530732288e-05, "loss": 0.0159, "step": 60600 }, { "grad_norm": 0.27566176652908325, "learning_rate": 1.5220690519104901e-05, "loss": 0.0183, "step": 60610 }, { "grad_norm": 0.27584606409072876, "learning_rate": 1.5205844450310275e-05, "loss": 0.0156, "step": 60620 }, { "grad_norm": 0.22948962450027466, "learning_rate": 1.519100432688521e-05, "loss": 0.0178, "step": 60630 }, { "grad_norm": 0.3995744585990906, "learning_rate": 1.5176170151365476e-05, "loss": 0.0157, "step": 60640 }, { "grad_norm": 0.32172515988349915, "learning_rate": 1.5161341926285833e-05, "loss": 0.0144, "step": 60650 }, { "grad_norm": 0.3690272867679596, "learning_rate": 1.5146519654180025e-05, "loss": 0.0155, "step": 60660 }, { "grad_norm": 0.2761736512184143, "learning_rate": 1.5131703337580739e-05, "loss": 0.0203, "step": 60670 }, { "grad_norm": 0.2594081461429596, "learning_rate": 1.5116892979019731e-05, "loss": 0.0236, "step": 60680 }, { "grad_norm": 0.21009068191051483, "learning_rate": 1.5102088581027623e-05, "loss": 0.0165, "step": 60690 }, { "grad_norm": 0.3457786440849304, "learning_rate": 1.5087290146134154e-05, "loss": 0.0186, "step": 60700 }, { "grad_norm": 0.42447951436042786, "learning_rate": 1.5072497676867915e-05, "loss": 0.0183, "step": 60710 }, { "grad_norm": 0.1586858481168747, "learning_rate": 1.505771117575655e-05, "loss": 0.024, "step": 60720 }, { "grad_norm": 0.1492522954940796, "learning_rate": 1.5042930645326691e-05, "loss": 0.0177, "step": 60730 }, { "grad_norm": 0.3243412673473358, "learning_rate": 1.5028156088103879e-05, "loss": 0.0225, "step": 60740 }, { "grad_norm": 0.3072807192802429, "learning_rate": 1.5013387506612735e-05, "loss": 0.0189, "step": 60750 }, { "grad_norm": 0.19253350794315338, "learning_rate": 1.4998624903376767e-05, "loss": 0.0212, "step": 60760 }, { "grad_norm": 0.4198589622974396, "learning_rate": 1.4983868280918517e-05, "loss": 0.0183, "step": 60770 }, { "grad_norm": 0.2900877594947815, "learning_rate": 1.4969117641759478e-05, "loss": 0.0256, "step": 60780 }, { "grad_norm": 0.28088900446891785, "learning_rate": 1.495437298842014e-05, "loss": 0.0179, "step": 60790 }, { "grad_norm": 0.395773708820343, "learning_rate": 1.4939634323419976e-05, "loss": 0.0205, "step": 60800 }, { "grad_norm": 0.2825687527656555, "learning_rate": 1.4924901649277384e-05, "loss": 0.0221, "step": 60810 }, { "grad_norm": 0.2017241269350052, "learning_rate": 1.4910174968509793e-05, "loss": 0.016, "step": 60820 }, { "grad_norm": 0.40299567580223083, "learning_rate": 1.4895454283633598e-05, "loss": 0.0184, "step": 60830 }, { "grad_norm": 0.3185761868953705, "learning_rate": 1.488073959716415e-05, "loss": 0.0183, "step": 60840 }, { "grad_norm": 0.29756227135658264, "learning_rate": 1.4866030911615791e-05, "loss": 0.0199, "step": 60850 }, { "grad_norm": 0.1699523776769638, "learning_rate": 1.4851328229501849e-05, "loss": 0.0147, "step": 60860 }, { "grad_norm": 0.21802528202533722, "learning_rate": 1.4836631553334562e-05, "loss": 0.0133, "step": 60870 }, { "grad_norm": 0.3287712037563324, "learning_rate": 1.4821940885625251e-05, "loss": 0.0182, "step": 60880 }, { "grad_norm": 0.3700808584690094, "learning_rate": 1.4807256228884109e-05, "loss": 0.016, "step": 60890 }, { "grad_norm": 0.14049549400806427, "learning_rate": 1.4792577585620353e-05, "loss": 0.0157, "step": 60900 }, { "grad_norm": 0.3582264482975006, "learning_rate": 1.4777904958342164e-05, "loss": 0.0231, "step": 60910 }, { "grad_norm": 0.23005342483520508, "learning_rate": 1.4763238349556691e-05, "loss": 0.0161, "step": 60920 }, { "grad_norm": 0.28705546259880066, "learning_rate": 1.4748577761770072e-05, "loss": 0.0203, "step": 60930 }, { "grad_norm": 0.25119495391845703, "learning_rate": 1.4733923197487354e-05, "loss": 0.0198, "step": 60940 }, { "grad_norm": 0.34415164589881897, "learning_rate": 1.4719274659212662e-05, "loss": 0.0213, "step": 60950 }, { "grad_norm": 0.22797423601150513, "learning_rate": 1.4704632149448983e-05, "loss": 0.017, "step": 60960 }, { "grad_norm": 0.2087285816669464, "learning_rate": 1.4689995670698343e-05, "loss": 0.0221, "step": 60970 }, { "grad_norm": 0.2897280156612396, "learning_rate": 1.4675365225461728e-05, "loss": 0.0234, "step": 60980 }, { "grad_norm": 0.3048265874385834, "learning_rate": 1.4660740816239032e-05, "loss": 0.0136, "step": 60990 }, { "grad_norm": 0.24426554143428802, "learning_rate": 1.464612244552923e-05, "loss": 0.0194, "step": 61000 }, { "grad_norm": 0.25088924169540405, "learning_rate": 1.4631510115830161e-05, "loss": 0.0183, "step": 61010 }, { "grad_norm": 0.283636212348938, "learning_rate": 1.4616903829638679e-05, "loss": 0.0163, "step": 61020 }, { "grad_norm": 0.23185957968235016, "learning_rate": 1.46023035894506e-05, "loss": 0.0182, "step": 61030 }, { "grad_norm": 0.4526485502719879, "learning_rate": 1.4587709397760713e-05, "loss": 0.0226, "step": 61040 }, { "grad_norm": 0.23012913763523102, "learning_rate": 1.4573121257062755e-05, "loss": 0.0182, "step": 61050 }, { "grad_norm": 0.4395601749420166, "learning_rate": 1.4558539169849472e-05, "loss": 0.025, "step": 61060 }, { "grad_norm": 0.2899342477321625, "learning_rate": 1.4543963138612499e-05, "loss": 0.0162, "step": 61070 }, { "grad_norm": 0.2829831540584564, "learning_rate": 1.4529393165842498e-05, "loss": 0.0196, "step": 61080 }, { "grad_norm": 0.31747663021087646, "learning_rate": 1.4514829254029084e-05, "loss": 0.0189, "step": 61090 }, { "grad_norm": 0.2308177649974823, "learning_rate": 1.450027140566083e-05, "loss": 0.0198, "step": 61100 }, { "grad_norm": 0.24726106226444244, "learning_rate": 1.4485719623225268e-05, "loss": 0.0168, "step": 61110 }, { "grad_norm": 0.2437392771244049, "learning_rate": 1.4471173909208912e-05, "loss": 0.0185, "step": 61120 }, { "grad_norm": 0.21736672520637512, "learning_rate": 1.4456634266097236e-05, "loss": 0.017, "step": 61130 }, { "grad_norm": 0.19251319766044617, "learning_rate": 1.444210069637461e-05, "loss": 0.0241, "step": 61140 }, { "grad_norm": 0.26834389567375183, "learning_rate": 1.4427573202524502e-05, "loss": 0.0131, "step": 61150 }, { "grad_norm": 0.1294165700674057, "learning_rate": 1.4413051787029208e-05, "loss": 0.0242, "step": 61160 }, { "grad_norm": 0.2029317021369934, "learning_rate": 1.439853645237006e-05, "loss": 0.0201, "step": 61170 }, { "grad_norm": 0.17263326048851013, "learning_rate": 1.4384027201027344e-05, "loss": 0.016, "step": 61180 }, { "grad_norm": 0.19823811948299408, "learning_rate": 1.4369524035480253e-05, "loss": 0.0195, "step": 61190 }, { "grad_norm": 0.1854691058397293, "learning_rate": 1.4355026958207035e-05, "loss": 0.0169, "step": 61200 }, { "grad_norm": 0.22184103727340698, "learning_rate": 1.434053597168481e-05, "loss": 0.019, "step": 61210 }, { "grad_norm": 0.2090413123369217, "learning_rate": 1.432605107838969e-05, "loss": 0.0187, "step": 61220 }, { "grad_norm": 0.5071840286254883, "learning_rate": 1.4311572280796759e-05, "loss": 0.0203, "step": 61230 }, { "grad_norm": 0.3273185193538666, "learning_rate": 1.4297099581380047e-05, "loss": 0.022, "step": 61240 }, { "grad_norm": 0.24172665178775787, "learning_rate": 1.4282632982612538e-05, "loss": 0.0211, "step": 61250 }, { "grad_norm": 0.23815098404884338, "learning_rate": 1.4268172486966197e-05, "loss": 0.0226, "step": 61260 }, { "grad_norm": 0.20149284601211548, "learning_rate": 1.4253718096911894e-05, "loss": 0.0187, "step": 61270 }, { "grad_norm": 0.373115211725235, "learning_rate": 1.4239269814919504e-05, "loss": 0.0248, "step": 61280 }, { "grad_norm": 0.401851624250412, "learning_rate": 1.4224827643457843e-05, "loss": 0.0263, "step": 61290 }, { "grad_norm": 0.36254557967185974, "learning_rate": 1.4210391584994686e-05, "loss": 0.0156, "step": 61300 }, { "grad_norm": 0.21458245813846588, "learning_rate": 1.4195961641996763e-05, "loss": 0.0207, "step": 61310 }, { "grad_norm": 0.25625181198120117, "learning_rate": 1.4181537816929751e-05, "loss": 0.0178, "step": 61320 }, { "grad_norm": 0.2591456174850464, "learning_rate": 1.4167120112258303e-05, "loss": 0.0206, "step": 61330 }, { "grad_norm": 0.2184354066848755, "learning_rate": 1.4152708530445963e-05, "loss": 0.0129, "step": 61340 }, { "grad_norm": 0.2840140163898468, "learning_rate": 1.4138303073955344e-05, "loss": 0.0184, "step": 61350 }, { "grad_norm": 0.41882815957069397, "learning_rate": 1.4123903745247897e-05, "loss": 0.0206, "step": 61360 }, { "grad_norm": 0.13779670000076294, "learning_rate": 1.4109510546784082e-05, "loss": 0.0121, "step": 61370 }, { "grad_norm": 0.13455650210380554, "learning_rate": 1.4095123481023325e-05, "loss": 0.0201, "step": 61380 }, { "grad_norm": 0.4116850793361664, "learning_rate": 1.4080742550423936e-05, "loss": 0.0141, "step": 61390 }, { "grad_norm": 0.26301461458206177, "learning_rate": 1.4066367757443278e-05, "loss": 0.0172, "step": 61400 }, { "grad_norm": 0.2475104033946991, "learning_rate": 1.4051999104537572e-05, "loss": 0.0139, "step": 61410 }, { "grad_norm": 0.3393418490886688, "learning_rate": 1.4037636594162034e-05, "loss": 0.0176, "step": 61420 }, { "grad_norm": 0.26114264130592346, "learning_rate": 1.402328022877083e-05, "loss": 0.0158, "step": 61430 }, { "grad_norm": 0.22169643640518188, "learning_rate": 1.4008930010817073e-05, "loss": 0.0201, "step": 61440 }, { "grad_norm": 0.40301480889320374, "learning_rate": 1.3994585942752831e-05, "loss": 0.0171, "step": 61450 }, { "grad_norm": 0.2985692024230957, "learning_rate": 1.398024802702909e-05, "loss": 0.0156, "step": 61460 }, { "grad_norm": 0.32558006048202515, "learning_rate": 1.3965916266095819e-05, "loss": 0.0201, "step": 61470 }, { "grad_norm": 0.34362515807151794, "learning_rate": 1.395159066240192e-05, "loss": 0.0135, "step": 61480 }, { "grad_norm": 0.30754464864730835, "learning_rate": 1.3937271218395248e-05, "loss": 0.0191, "step": 61490 }, { "grad_norm": 0.44654569029808044, "learning_rate": 1.3922957936522613e-05, "loss": 0.0185, "step": 61500 }, { "grad_norm": 0.3092978000640869, "learning_rate": 1.3908650819229758e-05, "loss": 0.0238, "step": 61510 }, { "grad_norm": 0.25142091512680054, "learning_rate": 1.389434986896137e-05, "loss": 0.0146, "step": 61520 }, { "grad_norm": 0.3688249886035919, "learning_rate": 1.388005508816112e-05, "loss": 0.0149, "step": 61530 }, { "grad_norm": 0.2497493326663971, "learning_rate": 1.3865766479271557e-05, "loss": 0.0218, "step": 61540 }, { "grad_norm": 0.20366351306438446, "learning_rate": 1.3851484044734225e-05, "loss": 0.03, "step": 61550 }, { "grad_norm": 0.27456745505332947, "learning_rate": 1.3837207786989608e-05, "loss": 0.0271, "step": 61560 }, { "grad_norm": 0.377637654542923, "learning_rate": 1.382293770847713e-05, "loss": 0.0198, "step": 61570 }, { "grad_norm": 0.2314268797636032, "learning_rate": 1.3808673811635159e-05, "loss": 0.0196, "step": 61580 }, { "grad_norm": 0.4452657401561737, "learning_rate": 1.3794416098900975e-05, "loss": 0.0248, "step": 61590 }, { "grad_norm": 0.17892597615718842, "learning_rate": 1.3780164572710886e-05, "loss": 0.0246, "step": 61600 }, { "grad_norm": 0.4596096873283386, "learning_rate": 1.3765919235500035e-05, "loss": 0.0257, "step": 61610 }, { "grad_norm": 0.13757963478565216, "learning_rate": 1.375168008970259e-05, "loss": 0.016, "step": 61620 }, { "grad_norm": 0.22948504984378815, "learning_rate": 1.3737447137751635e-05, "loss": 0.0153, "step": 61630 }, { "grad_norm": 0.29532134532928467, "learning_rate": 1.372322038207915e-05, "loss": 0.0163, "step": 61640 }, { "grad_norm": 0.23589681088924408, "learning_rate": 1.3708999825116159e-05, "loss": 0.0287, "step": 61650 }, { "grad_norm": 0.2605575621128082, "learning_rate": 1.3694785469292526e-05, "loss": 0.0239, "step": 61660 }, { "grad_norm": 0.3176431357860565, "learning_rate": 1.3680577317037101e-05, "loss": 0.019, "step": 61670 }, { "grad_norm": 0.2769848704338074, "learning_rate": 1.3666375370777678e-05, "loss": 0.0182, "step": 61680 }, { "grad_norm": 0.24291260540485382, "learning_rate": 1.3652179632940981e-05, "loss": 0.0173, "step": 61690 }, { "grad_norm": 0.2006673514842987, "learning_rate": 1.3637990105952664e-05, "loss": 0.017, "step": 61700 }, { "grad_norm": 0.18631312251091003, "learning_rate": 1.3623806792237337e-05, "loss": 0.0175, "step": 61710 }, { "grad_norm": 0.2692461311817169, "learning_rate": 1.3609629694218551e-05, "loss": 0.0165, "step": 61720 }, { "grad_norm": 0.26056110858917236, "learning_rate": 1.3595458814318762e-05, "loss": 0.0164, "step": 61730 }, { "grad_norm": 0.27112922072410583, "learning_rate": 1.3581294154959389e-05, "loss": 0.0204, "step": 61740 }, { "grad_norm": 0.23757120966911316, "learning_rate": 1.3567135718560792e-05, "loss": 0.0153, "step": 61750 }, { "grad_norm": 0.26792827248573303, "learning_rate": 1.3552983507542261e-05, "loss": 0.0196, "step": 61760 }, { "grad_norm": 0.2549617290496826, "learning_rate": 1.3538837524322023e-05, "loss": 0.0192, "step": 61770 }, { "grad_norm": 0.1514061838388443, "learning_rate": 1.3524697771317251e-05, "loss": 0.0246, "step": 61780 }, { "grad_norm": 0.22728998959064484, "learning_rate": 1.3510564250943997e-05, "loss": 0.0169, "step": 61790 }, { "grad_norm": 0.1865837424993515, "learning_rate": 1.3496436965617353e-05, "loss": 0.0164, "step": 61800 }, { "grad_norm": 0.194132000207901, "learning_rate": 1.3482315917751243e-05, "loss": 0.0226, "step": 61810 }, { "grad_norm": 0.174301877617836, "learning_rate": 1.3468201109758582e-05, "loss": 0.0129, "step": 61820 }, { "grad_norm": 0.2768377959728241, "learning_rate": 1.3454092544051222e-05, "loss": 0.0157, "step": 61830 }, { "grad_norm": 0.2139902561903, "learning_rate": 1.3439990223039878e-05, "loss": 0.0174, "step": 61840 }, { "grad_norm": 0.2392110377550125, "learning_rate": 1.3425894149134321e-05, "loss": 0.0157, "step": 61850 }, { "grad_norm": 0.39629337191581726, "learning_rate": 1.3411804324743132e-05, "loss": 0.0196, "step": 61860 }, { "grad_norm": 0.3786141276359558, "learning_rate": 1.3397720752273896e-05, "loss": 0.0213, "step": 61870 }, { "grad_norm": 0.21049584448337555, "learning_rate": 1.3383643434133108e-05, "loss": 0.0186, "step": 61880 }, { "grad_norm": 0.2482592910528183, "learning_rate": 1.3369572372726196e-05, "loss": 0.0181, "step": 61890 }, { "grad_norm": 0.2178439497947693, "learning_rate": 1.3355507570457525e-05, "loss": 0.0134, "step": 61900 }, { "grad_norm": 0.34533560276031494, "learning_rate": 1.334144902973038e-05, "loss": 0.0176, "step": 61910 }, { "grad_norm": 0.17615675926208496, "learning_rate": 1.3327396752946997e-05, "loss": 0.02, "step": 61920 }, { "grad_norm": 0.2718856632709503, "learning_rate": 1.3313350742508496e-05, "loss": 0.0195, "step": 61930 }, { "grad_norm": 0.26388785243034363, "learning_rate": 1.3299311000814973e-05, "loss": 0.0214, "step": 61940 }, { "grad_norm": 0.15709397196769714, "learning_rate": 1.3285277530265432e-05, "loss": 0.0196, "step": 61950 }, { "grad_norm": 0.41443124413490295, "learning_rate": 1.3271250333257813e-05, "loss": 0.0185, "step": 61960 }, { "grad_norm": 0.18430444598197937, "learning_rate": 1.3257229412188977e-05, "loss": 0.0154, "step": 61970 }, { "grad_norm": 0.24538514018058777, "learning_rate": 1.3243214769454726e-05, "loss": 0.0187, "step": 61980 }, { "grad_norm": 0.4496997594833374, "learning_rate": 1.3229206407449751e-05, "loss": 0.0169, "step": 61990 }, { "grad_norm": 0.23953396081924438, "learning_rate": 1.3215204328567742e-05, "loss": 0.0188, "step": 62000 }, { "grad_norm": 0.2187613993883133, "learning_rate": 1.3201208535201232e-05, "loss": 0.0155, "step": 62010 }, { "grad_norm": 0.19378313422203064, "learning_rate": 1.3187219029741737e-05, "loss": 0.0153, "step": 62020 }, { "grad_norm": 0.39403289556503296, "learning_rate": 1.3173235814579693e-05, "loss": 0.0186, "step": 62030 }, { "grad_norm": 0.30998584628105164, "learning_rate": 1.3159258892104398e-05, "loss": 0.0196, "step": 62040 }, { "grad_norm": 0.2148025929927826, "learning_rate": 1.3145288264704198e-05, "loss": 0.0208, "step": 62050 }, { "grad_norm": 0.451877236366272, "learning_rate": 1.3131323934766237e-05, "loss": 0.018, "step": 62060 }, { "grad_norm": 0.20979943871498108, "learning_rate": 1.311736590467666e-05, "loss": 0.0195, "step": 62070 }, { "grad_norm": 0.26349887251853943, "learning_rate": 1.3103414176820505e-05, "loss": 0.02, "step": 62080 }, { "grad_norm": 0.15881040692329407, "learning_rate": 1.308946875358174e-05, "loss": 0.0269, "step": 62090 }, { "grad_norm": 0.34114745259284973, "learning_rate": 1.3075529637343276e-05, "loss": 0.0292, "step": 62100 }, { "grad_norm": 0.27746936678886414, "learning_rate": 1.3061596830486883e-05, "loss": 0.0216, "step": 62110 }, { "grad_norm": 0.23011483252048492, "learning_rate": 1.3047670335393353e-05, "loss": 0.0157, "step": 62120 }, { "grad_norm": 0.31394562125205994, "learning_rate": 1.3033750154442298e-05, "loss": 0.0237, "step": 62130 }, { "grad_norm": 0.2466064989566803, "learning_rate": 1.3019836290012316e-05, "loss": 0.0199, "step": 62140 }, { "grad_norm": 0.2810741364955902, "learning_rate": 1.30059287444809e-05, "loss": 0.0192, "step": 62150 }, { "grad_norm": 0.2937858998775482, "learning_rate": 1.299202752022447e-05, "loss": 0.028, "step": 62160 }, { "grad_norm": 0.32269081473350525, "learning_rate": 1.2978132619618371e-05, "loss": 0.0156, "step": 62170 }, { "grad_norm": 0.33949315547943115, "learning_rate": 1.2964244045036866e-05, "loss": 0.0198, "step": 62180 }, { "grad_norm": 0.3154217302799225, "learning_rate": 1.295036179885311e-05, "loss": 0.0153, "step": 62190 }, { "grad_norm": 0.2266789972782135, "learning_rate": 1.293648588343922e-05, "loss": 0.0213, "step": 62200 }, { "grad_norm": 0.3874512314796448, "learning_rate": 1.2922616301166196e-05, "loss": 0.0201, "step": 62210 }, { "grad_norm": 0.3198745846748352, "learning_rate": 1.2908753054403976e-05, "loss": 0.0167, "step": 62220 }, { "grad_norm": 0.18880094587802887, "learning_rate": 1.2894896145521429e-05, "loss": 0.0156, "step": 62230 }, { "grad_norm": 0.2561801075935364, "learning_rate": 1.2881045576886275e-05, "loss": 0.0219, "step": 62240 }, { "grad_norm": 0.28589144349098206, "learning_rate": 1.2867201350865254e-05, "loss": 0.0182, "step": 62250 }, { "grad_norm": 0.23896558582782745, "learning_rate": 1.2853363469823914e-05, "loss": 0.0195, "step": 62260 }, { "grad_norm": 0.27996644377708435, "learning_rate": 1.283953193612682e-05, "loss": 0.0167, "step": 62270 }, { "grad_norm": 0.23506948351860046, "learning_rate": 1.2825706752137372e-05, "loss": 0.0171, "step": 62280 }, { "grad_norm": 0.33447715640068054, "learning_rate": 1.2811887920217896e-05, "loss": 0.0184, "step": 62290 }, { "grad_norm": 0.21621505916118622, "learning_rate": 1.2798075442729707e-05, "loss": 0.0142, "step": 62300 }, { "grad_norm": 0.346418559551239, "learning_rate": 1.2784269322032922e-05, "loss": 0.0167, "step": 62310 }, { "grad_norm": 0.19333595037460327, "learning_rate": 1.2770469560486686e-05, "loss": 0.0192, "step": 62320 }, { "grad_norm": 0.2074822187423706, "learning_rate": 1.2756676160448956e-05, "loss": 0.0215, "step": 62330 }, { "grad_norm": 0.4263104200363159, "learning_rate": 1.2742889124276663e-05, "loss": 0.0166, "step": 62340 }, { "grad_norm": 0.19985882937908173, "learning_rate": 1.2729108454325639e-05, "loss": 0.017, "step": 62350 }, { "grad_norm": 0.28367361426353455, "learning_rate": 1.2715334152950614e-05, "loss": 0.0185, "step": 62360 }, { "grad_norm": 0.23947258293628693, "learning_rate": 1.2701566222505246e-05, "loss": 0.0198, "step": 62370 }, { "grad_norm": 0.26672422885894775, "learning_rate": 1.2687804665342107e-05, "loss": 0.0154, "step": 62380 }, { "grad_norm": 0.32509151101112366, "learning_rate": 1.267404948381265e-05, "loss": 0.0168, "step": 62390 }, { "grad_norm": 0.2299453765153885, "learning_rate": 1.2660300680267267e-05, "loss": 0.0144, "step": 62400 }, { "grad_norm": 0.3943111002445221, "learning_rate": 1.2646558257055257e-05, "loss": 0.0131, "step": 62410 }, { "grad_norm": 0.2898985743522644, "learning_rate": 1.263282221652482e-05, "loss": 0.0167, "step": 62420 }, { "grad_norm": 0.3710041344165802, "learning_rate": 1.2619092561023088e-05, "loss": 0.0217, "step": 62430 }, { "grad_norm": 0.16534192860126495, "learning_rate": 1.2605369292896036e-05, "loss": 0.0181, "step": 62440 }, { "grad_norm": 0.244819775223732, "learning_rate": 1.2591652414488658e-05, "loss": 0.0191, "step": 62450 }, { "grad_norm": 0.2959974408149719, "learning_rate": 1.2577941928144732e-05, "loss": 0.0191, "step": 62460 }, { "grad_norm": 0.17623446881771088, "learning_rate": 1.256423783620706e-05, "loss": 0.0211, "step": 62470 }, { "grad_norm": 0.23594985902309418, "learning_rate": 1.2550540141017264e-05, "loss": 0.0169, "step": 62480 }, { "grad_norm": 0.30756908655166626, "learning_rate": 1.253684884491591e-05, "loss": 0.0182, "step": 62490 }, { "grad_norm": 0.6810060739517212, "learning_rate": 1.2523163950242483e-05, "loss": 0.0168, "step": 62500 }, { "grad_norm": 0.2198706567287445, "learning_rate": 1.2509485459335313e-05, "loss": 0.0146, "step": 62510 }, { "grad_norm": 0.33990904688835144, "learning_rate": 1.2495813374531739e-05, "loss": 0.0145, "step": 62520 }, { "grad_norm": 0.26179540157318115, "learning_rate": 1.2482147698167907e-05, "loss": 0.0144, "step": 62530 }, { "grad_norm": 0.23086147010326385, "learning_rate": 1.2468488432578911e-05, "loss": 0.0164, "step": 62540 }, { "grad_norm": 0.2087850719690323, "learning_rate": 1.2454835580098761e-05, "loss": 0.0185, "step": 62550 }, { "grad_norm": 0.2820693254470825, "learning_rate": 1.2441189143060338e-05, "loss": 0.0168, "step": 62560 }, { "grad_norm": 0.2454601228237152, "learning_rate": 1.242754912379548e-05, "loss": 0.0185, "step": 62570 }, { "grad_norm": 0.20011965930461884, "learning_rate": 1.2413915524634844e-05, "loss": 0.0314, "step": 62580 }, { "grad_norm": 0.3000512719154358, "learning_rate": 1.2400288347908073e-05, "loss": 0.0216, "step": 62590 }, { "grad_norm": 0.25992605090141296, "learning_rate": 1.2386667595943663e-05, "loss": 0.0139, "step": 62600 }, { "grad_norm": 0.3547814190387726, "learning_rate": 1.2373053271069035e-05, "loss": 0.0175, "step": 62610 }, { "grad_norm": 0.19513729214668274, "learning_rate": 1.2359445375610501e-05, "loss": 0.0194, "step": 62620 }, { "grad_norm": 0.25447648763656616, "learning_rate": 1.2345843911893301e-05, "loss": 0.0148, "step": 62630 }, { "grad_norm": 0.21068993210792542, "learning_rate": 1.2332248882241498e-05, "loss": 0.0281, "step": 62640 }, { "grad_norm": 0.23169472813606262, "learning_rate": 1.2318660288978178e-05, "loss": 0.016, "step": 62650 }, { "grad_norm": 0.3354966640472412, "learning_rate": 1.2305078134425213e-05, "loss": 0.0171, "step": 62660 }, { "grad_norm": 0.25169655680656433, "learning_rate": 1.2291502420903434e-05, "loss": 0.0173, "step": 62670 }, { "grad_norm": 0.26880109310150146, "learning_rate": 1.2277933150732567e-05, "loss": 0.0197, "step": 62680 }, { "grad_norm": 0.21259185671806335, "learning_rate": 1.2264370326231216e-05, "loss": 0.0166, "step": 62690 }, { "grad_norm": 0.3073618710041046, "learning_rate": 1.2250813949716927e-05, "loss": 0.0178, "step": 62700 }, { "grad_norm": 0.21359552443027496, "learning_rate": 1.2237264023506063e-05, "loss": 0.0164, "step": 62710 }, { "grad_norm": 0.2252417355775833, "learning_rate": 1.2223720549913987e-05, "loss": 0.0152, "step": 62720 }, { "grad_norm": 0.3207096755504608, "learning_rate": 1.221018353125487e-05, "loss": 0.0163, "step": 62730 }, { "grad_norm": 0.16753199696540833, "learning_rate": 1.2196652969841837e-05, "loss": 0.0195, "step": 62740 }, { "grad_norm": 0.374188631772995, "learning_rate": 1.2183128867986904e-05, "loss": 0.0183, "step": 62750 }, { "grad_norm": 0.29937824606895447, "learning_rate": 1.2169611228000927e-05, "loss": 0.0186, "step": 62760 }, { "grad_norm": 0.297758013010025, "learning_rate": 1.2156100052193752e-05, "loss": 0.0195, "step": 62770 }, { "grad_norm": 0.19095806777477264, "learning_rate": 1.214259534287403e-05, "loss": 0.0198, "step": 62780 }, { "grad_norm": 0.19176991283893585, "learning_rate": 1.2129097102349363e-05, "loss": 0.0135, "step": 62790 }, { "grad_norm": 0.20591308176517487, "learning_rate": 1.2115605332926227e-05, "loss": 0.0196, "step": 62800 }, { "grad_norm": 0.3158997595310211, "learning_rate": 1.2102120036910003e-05, "loss": 0.0264, "step": 62810 }, { "grad_norm": 0.33249467611312866, "learning_rate": 1.2088641216604956e-05, "loss": 0.0211, "step": 62820 }, { "grad_norm": 0.25931012630462646, "learning_rate": 1.2075168874314264e-05, "loss": 0.0162, "step": 62830 }, { "grad_norm": 0.18384334444999695, "learning_rate": 1.2061703012339942e-05, "loss": 0.016, "step": 62840 }, { "grad_norm": 0.39624878764152527, "learning_rate": 1.204824363298297e-05, "loss": 0.0203, "step": 62850 }, { "grad_norm": 0.27901360392570496, "learning_rate": 1.2034790738543173e-05, "loss": 0.0197, "step": 62860 }, { "grad_norm": 0.20613856613636017, "learning_rate": 1.2021344331319284e-05, "loss": 0.0198, "step": 62870 }, { "grad_norm": 0.1379621922969818, "learning_rate": 1.200790441360894e-05, "loss": 0.0193, "step": 62880 }, { "grad_norm": 0.17578452825546265, "learning_rate": 1.199447098770864e-05, "loss": 0.0154, "step": 62890 }, { "grad_norm": 0.3139868378639221, "learning_rate": 1.198104405591381e-05, "loss": 0.0189, "step": 62900 }, { "grad_norm": 0.21802818775177002, "learning_rate": 1.1967623620518697e-05, "loss": 0.0134, "step": 62910 }, { "grad_norm": 0.3225894868373871, "learning_rate": 1.1954209683816554e-05, "loss": 0.0174, "step": 62920 }, { "grad_norm": 0.3748307526111603, "learning_rate": 1.1940802248099402e-05, "loss": 0.0219, "step": 62930 }, { "grad_norm": 0.5350742340087891, "learning_rate": 1.1927401315658232e-05, "loss": 0.0197, "step": 62940 }, { "grad_norm": 0.3037809431552887, "learning_rate": 1.1914006888782898e-05, "loss": 0.0175, "step": 62950 }, { "grad_norm": 0.37043115496635437, "learning_rate": 1.1900618969762106e-05, "loss": 0.0205, "step": 62960 }, { "grad_norm": 0.23000358045101166, "learning_rate": 1.188723756088354e-05, "loss": 0.0179, "step": 62970 }, { "grad_norm": 0.29263490438461304, "learning_rate": 1.1873862664433672e-05, "loss": 0.0236, "step": 62980 }, { "grad_norm": 0.33945363759994507, "learning_rate": 1.1860494282697927e-05, "loss": 0.0187, "step": 62990 }, { "grad_norm": 0.41983717679977417, "learning_rate": 1.1847132417960588e-05, "loss": 0.0166, "step": 63000 }, { "grad_norm": 0.15376892685890198, "learning_rate": 1.1833777072504832e-05, "loss": 0.023, "step": 63010 }, { "grad_norm": 0.2136448174715042, "learning_rate": 1.1820428248612731e-05, "loss": 0.015, "step": 63020 }, { "grad_norm": 0.4515572786331177, "learning_rate": 1.1807085948565245e-05, "loss": 0.0192, "step": 63030 }, { "grad_norm": 0.17297238111495972, "learning_rate": 1.1793750174642172e-05, "loss": 0.0166, "step": 63040 }, { "grad_norm": 0.1064140573143959, "learning_rate": 1.1780420929122254e-05, "loss": 0.0198, "step": 63050 }, { "grad_norm": 0.3607650399208069, "learning_rate": 1.176709821428309e-05, "loss": 0.0172, "step": 63060 }, { "grad_norm": 0.3608740270137787, "learning_rate": 1.1753782032401173e-05, "loss": 0.0265, "step": 63070 }, { "grad_norm": 0.36475062370300293, "learning_rate": 1.1740472385751866e-05, "loss": 0.0207, "step": 63080 }, { "grad_norm": 0.22973372042179108, "learning_rate": 1.172716927660943e-05, "loss": 0.016, "step": 63090 }, { "grad_norm": 0.3276558518409729, "learning_rate": 1.1713872707247015e-05, "loss": 0.021, "step": 63100 }, { "grad_norm": 0.21066470444202423, "learning_rate": 1.1700582679936595e-05, "loss": 0.0178, "step": 63110 }, { "grad_norm": 0.1771051585674286, "learning_rate": 1.1687299196949136e-05, "loss": 0.025, "step": 63120 }, { "grad_norm": 0.4072054922580719, "learning_rate": 1.1674022260554374e-05, "loss": 0.0176, "step": 63130 }, { "grad_norm": 0.27914857864379883, "learning_rate": 1.1660751873020987e-05, "loss": 0.0189, "step": 63140 }, { "grad_norm": 0.2700081467628479, "learning_rate": 1.1647488036616538e-05, "loss": 0.0214, "step": 63150 }, { "grad_norm": 0.1953909695148468, "learning_rate": 1.1634230753607417e-05, "loss": 0.0224, "step": 63160 }, { "grad_norm": 0.3240007162094116, "learning_rate": 1.1620980026258982e-05, "loss": 0.0194, "step": 63170 }, { "grad_norm": 0.22296994924545288, "learning_rate": 1.1607735856835373e-05, "loss": 0.0143, "step": 63180 }, { "grad_norm": 0.20830903947353363, "learning_rate": 1.1594498247599677e-05, "loss": 0.018, "step": 63190 }, { "grad_norm": 0.2137666642665863, "learning_rate": 1.158126720081384e-05, "loss": 0.0274, "step": 63200 }, { "grad_norm": 0.29034528136253357, "learning_rate": 1.156804271873868e-05, "loss": 0.0134, "step": 63210 }, { "grad_norm": 0.22307585179805756, "learning_rate": 1.1554824803633924e-05, "loss": 0.0193, "step": 63220 }, { "grad_norm": 0.2941693961620331, "learning_rate": 1.154161345775811e-05, "loss": 0.015, "step": 63230 }, { "grad_norm": 0.34555938839912415, "learning_rate": 1.1528408683368724e-05, "loss": 0.0165, "step": 63240 }, { "grad_norm": 0.30854904651641846, "learning_rate": 1.1515210482722088e-05, "loss": 0.0182, "step": 63250 }, { "grad_norm": 0.2355850487947464, "learning_rate": 1.150201885807342e-05, "loss": 0.0215, "step": 63260 }, { "grad_norm": 0.11114408075809479, "learning_rate": 1.1488833811676807e-05, "loss": 0.0124, "step": 63270 }, { "grad_norm": 0.39230892062187195, "learning_rate": 1.1475655345785213e-05, "loss": 0.0189, "step": 63280 }, { "grad_norm": 0.3842412531375885, "learning_rate": 1.1462483462650481e-05, "loss": 0.0172, "step": 63290 }, { "grad_norm": 0.2710634469985962, "learning_rate": 1.1449318164523331e-05, "loss": 0.0161, "step": 63300 }, { "grad_norm": 0.37269070744514465, "learning_rate": 1.1436159453653334e-05, "loss": 0.0164, "step": 63310 }, { "grad_norm": 0.28933048248291016, "learning_rate": 1.1423007332288955e-05, "loss": 0.0189, "step": 63320 }, { "grad_norm": 0.4364277124404907, "learning_rate": 1.1409861802677546e-05, "loss": 0.0153, "step": 63330 }, { "grad_norm": 0.19960206747055054, "learning_rate": 1.1396722867065313e-05, "loss": 0.0159, "step": 63340 }, { "grad_norm": 0.38401806354522705, "learning_rate": 1.1383590527697352e-05, "loss": 0.0186, "step": 63350 }, { "grad_norm": 0.1403190940618515, "learning_rate": 1.137046478681758e-05, "loss": 0.0142, "step": 63360 }, { "grad_norm": 0.2414824217557907, "learning_rate": 1.1357345646668888e-05, "loss": 0.0342, "step": 63370 }, { "grad_norm": 0.16285325586795807, "learning_rate": 1.1344233109492924e-05, "loss": 0.0199, "step": 63380 }, { "grad_norm": 0.2290903776884079, "learning_rate": 1.1331127177530292e-05, "loss": 0.0133, "step": 63390 }, { "grad_norm": 0.2920635938644409, "learning_rate": 1.1318027853020441e-05, "loss": 0.0161, "step": 63400 }, { "grad_norm": 0.1772473007440567, "learning_rate": 1.1304935138201645e-05, "loss": 0.0127, "step": 63410 }, { "grad_norm": 0.27098092436790466, "learning_rate": 1.1291849035311153e-05, "loss": 0.0193, "step": 63420 }, { "grad_norm": 0.19893863797187805, "learning_rate": 1.1278769546584972e-05, "loss": 0.0161, "step": 63430 }, { "grad_norm": 0.1581411212682724, "learning_rate": 1.1265696674258052e-05, "loss": 0.0144, "step": 63440 }, { "grad_norm": 0.20354995131492615, "learning_rate": 1.1252630420564186e-05, "loss": 0.0188, "step": 63450 }, { "grad_norm": 0.2250606119632721, "learning_rate": 1.1239570787736036e-05, "loss": 0.0179, "step": 63460 }, { "grad_norm": 0.27678805589675903, "learning_rate": 1.1226517778005135e-05, "loss": 0.0174, "step": 63470 }, { "grad_norm": 0.16925957798957825, "learning_rate": 1.1213471393601893e-05, "loss": 0.0148, "step": 63480 }, { "grad_norm": 0.20003634691238403, "learning_rate": 1.1200431636755587e-05, "loss": 0.0223, "step": 63490 }, { "grad_norm": 0.4261694848537445, "learning_rate": 1.1187398509694336e-05, "loss": 0.0221, "step": 63500 }, { "grad_norm": 0.3470763862133026, "learning_rate": 1.1174372014645146e-05, "loss": 0.0143, "step": 63510 }, { "grad_norm": 0.22881463170051575, "learning_rate": 1.1161352153833899e-05, "loss": 0.0239, "step": 63520 }, { "grad_norm": 0.33726659417152405, "learning_rate": 1.1148338929485325e-05, "loss": 0.0204, "step": 63530 }, { "grad_norm": 0.22603845596313477, "learning_rate": 1.113533234382304e-05, "loss": 0.0137, "step": 63540 }, { "grad_norm": 0.30126434564590454, "learning_rate": 1.1122332399069513e-05, "loss": 0.0165, "step": 63550 }, { "grad_norm": 0.34906765818595886, "learning_rate": 1.1109339097446047e-05, "loss": 0.0201, "step": 63560 }, { "grad_norm": 0.3981902599334717, "learning_rate": 1.1096352441172897e-05, "loss": 0.0248, "step": 63570 }, { "grad_norm": 0.20551085472106934, "learning_rate": 1.1083372432469086e-05, "loss": 0.0129, "step": 63580 }, { "grad_norm": 0.12676569819450378, "learning_rate": 1.107039907355255e-05, "loss": 0.0189, "step": 63590 }, { "grad_norm": 0.4597172439098358, "learning_rate": 1.1057432366640103e-05, "loss": 0.0151, "step": 63600 }, { "grad_norm": 0.278947651386261, "learning_rate": 1.1044472313947352e-05, "loss": 0.0242, "step": 63610 }, { "grad_norm": 0.38134631514549255, "learning_rate": 1.1031518917688877e-05, "loss": 0.0154, "step": 63620 }, { "grad_norm": 0.261793851852417, "learning_rate": 1.1018572180078007e-05, "loss": 0.0275, "step": 63630 }, { "grad_norm": 0.12225132435560226, "learning_rate": 1.1005632103327018e-05, "loss": 0.0174, "step": 63640 }, { "grad_norm": 0.26787054538726807, "learning_rate": 1.0992698689646996e-05, "loss": 0.0167, "step": 63650 }, { "grad_norm": 0.3008846044540405, "learning_rate": 1.0979771941247919e-05, "loss": 0.018, "step": 63660 }, { "grad_norm": 0.3667605519294739, "learning_rate": 1.0966851860338611e-05, "loss": 0.0175, "step": 63670 }, { "grad_norm": 0.20518222451210022, "learning_rate": 1.0953938449126766e-05, "loss": 0.0159, "step": 63680 }, { "grad_norm": 0.4218579828739166, "learning_rate": 1.0941031709818933e-05, "loss": 0.028, "step": 63690 }, { "grad_norm": 0.2846311628818512, "learning_rate": 1.0928131644620509e-05, "loss": 0.0272, "step": 63700 }, { "grad_norm": 0.25568529963493347, "learning_rate": 1.0915238255735766e-05, "loss": 0.0218, "step": 63710 }, { "grad_norm": 0.17138192057609558, "learning_rate": 1.0902351545367833e-05, "loss": 0.0193, "step": 63720 }, { "grad_norm": 0.3498132526874542, "learning_rate": 1.0889471515718702e-05, "loss": 0.0185, "step": 63730 }, { "grad_norm": 0.1542021483182907, "learning_rate": 1.087659816898921e-05, "loss": 0.012, "step": 63740 }, { "grad_norm": 0.22989509999752045, "learning_rate": 1.0863731507379082e-05, "loss": 0.0164, "step": 63750 }, { "grad_norm": 0.2570978105068207, "learning_rate": 1.0850871533086827e-05, "loss": 0.0204, "step": 63760 }, { "grad_norm": 0.2987711727619171, "learning_rate": 1.0838018248309927e-05, "loss": 0.0145, "step": 63770 }, { "grad_norm": 0.2873609960079193, "learning_rate": 1.0825171655244615e-05, "loss": 0.025, "step": 63780 }, { "grad_norm": 0.28481295704841614, "learning_rate": 1.0812331756086025e-05, "loss": 0.0171, "step": 63790 }, { "grad_norm": 0.48719173669815063, "learning_rate": 1.079949855302817e-05, "loss": 0.0203, "step": 63800 }, { "grad_norm": 0.5216068029403687, "learning_rate": 1.0786672048263852e-05, "loss": 0.0227, "step": 63810 }, { "grad_norm": 0.333024799823761, "learning_rate": 1.0773852243984817e-05, "loss": 0.0189, "step": 63820 }, { "grad_norm": 0.2247474491596222, "learning_rate": 1.0761039142381586e-05, "loss": 0.0237, "step": 63830 }, { "grad_norm": 0.27088552713394165, "learning_rate": 1.0748232745643577e-05, "loss": 0.0218, "step": 63840 }, { "grad_norm": 0.3451622426509857, "learning_rate": 1.0735433055959055e-05, "loss": 0.0226, "step": 63850 }, { "grad_norm": 0.2391504943370819, "learning_rate": 1.0722640075515133e-05, "loss": 0.0171, "step": 63860 }, { "grad_norm": 0.2922590672969818, "learning_rate": 1.0709853806497795e-05, "loss": 0.0175, "step": 63870 }, { "grad_norm": 0.3402223289012909, "learning_rate": 1.0697074251091831e-05, "loss": 0.0201, "step": 63880 }, { "grad_norm": 0.19469474256038666, "learning_rate": 1.0684301411480962e-05, "loss": 0.0192, "step": 63890 }, { "grad_norm": 0.226236492395401, "learning_rate": 1.067153528984769e-05, "loss": 0.0135, "step": 63900 }, { "grad_norm": 0.22758527100086212, "learning_rate": 1.0658775888373395e-05, "loss": 0.0217, "step": 63910 }, { "grad_norm": 0.33834564685821533, "learning_rate": 1.0646023209238314e-05, "loss": 0.0235, "step": 63920 }, { "grad_norm": 0.24667632579803467, "learning_rate": 1.0633277254621537e-05, "loss": 0.0157, "step": 63930 }, { "grad_norm": 0.23185817897319794, "learning_rate": 1.0620538026700994e-05, "loss": 0.0137, "step": 63940 }, { "grad_norm": 0.2853374183177948, "learning_rate": 1.0607805527653486e-05, "loss": 0.0194, "step": 63950 }, { "grad_norm": 0.26028138399124146, "learning_rate": 1.059507975965462e-05, "loss": 0.0184, "step": 63960 }, { "grad_norm": 0.18079957365989685, "learning_rate": 1.0582360724878898e-05, "loss": 0.0161, "step": 63970 }, { "grad_norm": 0.3358848989009857, "learning_rate": 1.0569648425499651e-05, "loss": 0.0209, "step": 63980 }, { "grad_norm": 0.23365812003612518, "learning_rate": 1.0556942863689063e-05, "loss": 0.019, "step": 63990 }, { "grad_norm": 0.3679761290550232, "learning_rate": 1.054424404161819e-05, "loss": 0.0179, "step": 64000 }, { "grad_norm": 0.4403315484523773, "learning_rate": 1.053155196145686e-05, "loss": 0.017, "step": 64010 }, { "grad_norm": 0.2664390504360199, "learning_rate": 1.0518866625373863e-05, "loss": 0.0154, "step": 64020 }, { "grad_norm": 0.2304506003856659, "learning_rate": 1.0506188035536735e-05, "loss": 0.0178, "step": 64030 }, { "grad_norm": 0.31015849113464355, "learning_rate": 1.0493516194111919e-05, "loss": 0.0162, "step": 64040 }, { "grad_norm": 0.333156943321228, "learning_rate": 1.0480851103264688e-05, "loss": 0.0156, "step": 64050 }, { "grad_norm": 0.1867668330669403, "learning_rate": 1.046819276515913e-05, "loss": 0.016, "step": 64060 }, { "grad_norm": 0.274870365858078, "learning_rate": 1.0455541181958256e-05, "loss": 0.0165, "step": 64070 }, { "grad_norm": 0.26981809735298157, "learning_rate": 1.0442896355823822e-05, "loss": 0.0182, "step": 64080 }, { "grad_norm": 0.3498190641403198, "learning_rate": 1.0430258288916539e-05, "loss": 0.0157, "step": 64090 }, { "grad_norm": 0.20756134390830994, "learning_rate": 1.0417626983395868e-05, "loss": 0.0173, "step": 64100 }, { "grad_norm": 0.20247577130794525, "learning_rate": 1.0405002441420165e-05, "loss": 0.0145, "step": 64110 }, { "grad_norm": 0.248078852891922, "learning_rate": 1.039238466514662e-05, "loss": 0.0155, "step": 64120 }, { "grad_norm": 0.2056397646665573, "learning_rate": 1.0379773656731262e-05, "loss": 0.0219, "step": 64130 }, { "grad_norm": 0.40627458691596985, "learning_rate": 1.0367169418328986e-05, "loss": 0.0145, "step": 64140 }, { "grad_norm": 0.38759881258010864, "learning_rate": 1.0354571952093484e-05, "loss": 0.0229, "step": 64150 }, { "grad_norm": 0.3142791986465454, "learning_rate": 1.034198126017733e-05, "loss": 0.0173, "step": 64160 }, { "grad_norm": 0.2427975982427597, "learning_rate": 1.0329397344731928e-05, "loss": 0.018, "step": 64170 }, { "grad_norm": 0.3480152487754822, "learning_rate": 1.0316820207907524e-05, "loss": 0.0239, "step": 64180 }, { "grad_norm": 0.19282934069633484, "learning_rate": 1.0304249851853215e-05, "loss": 0.0177, "step": 64190 }, { "grad_norm": 0.20368333160877228, "learning_rate": 1.0291686278716933e-05, "loss": 0.0158, "step": 64200 }, { "grad_norm": 0.43045029044151306, "learning_rate": 1.0279129490645418e-05, "loss": 0.0168, "step": 64210 }, { "grad_norm": 0.34151026606559753, "learning_rate": 1.0266579489784328e-05, "loss": 0.0155, "step": 64220 }, { "grad_norm": 0.30090251564979553, "learning_rate": 1.0254036278278084e-05, "loss": 0.0174, "step": 64230 }, { "grad_norm": 0.20021432638168335, "learning_rate": 1.0241499858269982e-05, "loss": 0.0147, "step": 64240 }, { "grad_norm": 0.3306528627872467, "learning_rate": 1.0228970231902169e-05, "loss": 0.0176, "step": 64250 }, { "grad_norm": 0.2422294020652771, "learning_rate": 1.0216447401315582e-05, "loss": 0.0169, "step": 64260 }, { "grad_norm": 0.3355819880962372, "learning_rate": 1.020393136865007e-05, "loss": 0.015, "step": 64270 }, { "grad_norm": 0.3695368766784668, "learning_rate": 1.0191422136044242e-05, "loss": 0.0277, "step": 64280 }, { "grad_norm": 0.1476833075284958, "learning_rate": 1.017891970563563e-05, "loss": 0.0168, "step": 64290 }, { "grad_norm": 0.20699772238731384, "learning_rate": 1.0166424079560516e-05, "loss": 0.0259, "step": 64300 }, { "grad_norm": 0.21070849895477295, "learning_rate": 1.0153935259954078e-05, "loss": 0.0166, "step": 64310 }, { "grad_norm": 0.2653617858886719, "learning_rate": 1.0141453248950311e-05, "loss": 0.0174, "step": 64320 }, { "grad_norm": 0.24136018753051758, "learning_rate": 1.0128978048682054e-05, "loss": 0.0187, "step": 64330 }, { "grad_norm": 0.2722874879837036, "learning_rate": 1.0116509661280982e-05, "loss": 0.0192, "step": 64340 }, { "grad_norm": 0.24911482632160187, "learning_rate": 1.0104048088877576e-05, "loss": 0.0152, "step": 64350 }, { "grad_norm": 0.13108189404010773, "learning_rate": 1.0091593333601201e-05, "loss": 0.0175, "step": 64360 }, { "grad_norm": 0.22668394446372986, "learning_rate": 1.0079145397580031e-05, "loss": 0.0172, "step": 64370 }, { "grad_norm": 0.4392786920070648, "learning_rate": 1.006670428294107e-05, "loss": 0.0226, "step": 64380 }, { "grad_norm": 0.21149882674217224, "learning_rate": 1.0054269991810166e-05, "loss": 0.0144, "step": 64390 }, { "grad_norm": 0.2598309814929962, "learning_rate": 1.0041842526312024e-05, "loss": 0.0152, "step": 64400 }, { "grad_norm": 0.17156031727790833, "learning_rate": 1.0029421888570101e-05, "loss": 0.0121, "step": 64410 }, { "grad_norm": 0.22441796958446503, "learning_rate": 1.0017008080706813e-05, "loss": 0.0264, "step": 64420 }, { "grad_norm": 0.34980472922325134, "learning_rate": 1.0004601104843287e-05, "loss": 0.0158, "step": 64430 }, { "grad_norm": 0.2525722086429596, "learning_rate": 9.992200963099562e-06, "loss": 0.0188, "step": 64440 }, { "grad_norm": 0.2345365732908249, "learning_rate": 9.979807657594486e-06, "loss": 0.0176, "step": 64450 }, { "grad_norm": 0.2997077405452728, "learning_rate": 9.967421190445703e-06, "loss": 0.018, "step": 64460 }, { "grad_norm": 0.26481688022613525, "learning_rate": 9.955041563769769e-06, "loss": 0.0184, "step": 64470 }, { "grad_norm": 0.20529472827911377, "learning_rate": 9.942668779681974e-06, "loss": 0.0219, "step": 64480 }, { "grad_norm": 0.32331588864326477, "learning_rate": 9.930302840296541e-06, "loss": 0.0172, "step": 64490 }, { "grad_norm": 0.19126175343990326, "learning_rate": 9.917943747726426e-06, "loss": 0.016, "step": 64500 }, { "grad_norm": 0.1655167192220688, "learning_rate": 9.905591504083484e-06, "loss": 0.0155, "step": 64510 }, { "grad_norm": 0.21120594441890717, "learning_rate": 9.893246111478382e-06, "loss": 0.0144, "step": 64520 }, { "grad_norm": 0.33239293098449707, "learning_rate": 9.88090757202057e-06, "loss": 0.0178, "step": 64530 }, { "grad_norm": 0.25308623909950256, "learning_rate": 9.868575887818421e-06, "loss": 0.0216, "step": 64540 }, { "grad_norm": 0.268120139837265, "learning_rate": 9.856251060979044e-06, "loss": 0.019, "step": 64550 }, { "grad_norm": 0.25056707859039307, "learning_rate": 9.843933093608426e-06, "loss": 0.0215, "step": 64560 }, { "grad_norm": 0.27775466442108154, "learning_rate": 9.831621987811368e-06, "loss": 0.0211, "step": 64570 }, { "grad_norm": 0.4139273166656494, "learning_rate": 9.819317745691509e-06, "loss": 0.0168, "step": 64580 }, { "grad_norm": 0.35639894008636475, "learning_rate": 9.8070203693513e-06, "loss": 0.0153, "step": 64590 }, { "grad_norm": 0.33096638321876526, "learning_rate": 9.794729860892048e-06, "loss": 0.0202, "step": 64600 }, { "grad_norm": 0.3840719759464264, "learning_rate": 9.782446222413827e-06, "loss": 0.0215, "step": 64610 }, { "grad_norm": 0.3327469527721405, "learning_rate": 9.770169456015598e-06, "loss": 0.0205, "step": 64620 }, { "grad_norm": 0.1520814448595047, "learning_rate": 9.75789956379512e-06, "loss": 0.0195, "step": 64630 }, { "grad_norm": 0.290188729763031, "learning_rate": 9.74563654784898e-06, "loss": 0.017, "step": 64640 }, { "grad_norm": 0.1786966472864151, "learning_rate": 9.733380410272596e-06, "loss": 0.0162, "step": 64650 }, { "grad_norm": 0.27237892150878906, "learning_rate": 9.721131153160207e-06, "loss": 0.0153, "step": 64660 }, { "grad_norm": 0.2126149982213974, "learning_rate": 9.708888778604886e-06, "loss": 0.0193, "step": 64670 }, { "grad_norm": 0.25859615206718445, "learning_rate": 9.696653288698488e-06, "loss": 0.0186, "step": 64680 }, { "grad_norm": 0.3907659649848938, "learning_rate": 9.684424685531763e-06, "loss": 0.0219, "step": 64690 }, { "grad_norm": 0.1752796322107315, "learning_rate": 9.672202971194216e-06, "loss": 0.0201, "step": 64700 }, { "grad_norm": 0.20298214256763458, "learning_rate": 9.659988147774213e-06, "loss": 0.0177, "step": 64710 }, { "grad_norm": 0.2404669225215912, "learning_rate": 9.647780217358942e-06, "loss": 0.0188, "step": 64720 }, { "grad_norm": 0.2287594974040985, "learning_rate": 9.635579182034376e-06, "loss": 0.0174, "step": 64730 }, { "grad_norm": 0.17943228781223297, "learning_rate": 9.623385043885386e-06, "loss": 0.0231, "step": 64740 }, { "grad_norm": 0.3770908713340759, "learning_rate": 9.61119780499557e-06, "loss": 0.0165, "step": 64750 }, { "grad_norm": 0.19062519073486328, "learning_rate": 9.599017467447418e-06, "loss": 0.0161, "step": 64760 }, { "grad_norm": 0.19293132424354553, "learning_rate": 9.586844033322206e-06, "loss": 0.0187, "step": 64770 }, { "grad_norm": 0.30598509311676025, "learning_rate": 9.574677504700052e-06, "loss": 0.0189, "step": 64780 }, { "grad_norm": 0.2710326910018921, "learning_rate": 9.562517883659877e-06, "loss": 0.0316, "step": 64790 }, { "grad_norm": 0.21166273951530457, "learning_rate": 9.55036517227944e-06, "loss": 0.0159, "step": 64800 }, { "grad_norm": 0.34792283177375793, "learning_rate": 9.538219372635282e-06, "loss": 0.0199, "step": 64810 }, { "grad_norm": 0.21399907767772675, "learning_rate": 9.526080486802802e-06, "loss": 0.0141, "step": 64820 }, { "grad_norm": 0.21132230758666992, "learning_rate": 9.513948516856203e-06, "loss": 0.0181, "step": 64830 }, { "grad_norm": 0.16811510920524597, "learning_rate": 9.501823464868504e-06, "loss": 0.0144, "step": 64840 }, { "grad_norm": 0.3430436849594116, "learning_rate": 9.489705332911547e-06, "loss": 0.0157, "step": 64850 }, { "grad_norm": 0.4297305941581726, "learning_rate": 9.477594123055994e-06, "loss": 0.0202, "step": 64860 }, { "grad_norm": 0.2551192343235016, "learning_rate": 9.465489837371321e-06, "loss": 0.0216, "step": 64870 }, { "grad_norm": 0.32049837708473206, "learning_rate": 9.453392477925794e-06, "loss": 0.0167, "step": 64880 }, { "grad_norm": 0.16718789935112, "learning_rate": 9.441302046786566e-06, "loss": 0.0168, "step": 64890 }, { "grad_norm": 0.25972920656204224, "learning_rate": 9.429218546019519e-06, "loss": 0.0201, "step": 64900 }, { "grad_norm": 0.24879834055900574, "learning_rate": 9.41714197768941e-06, "loss": 0.0234, "step": 64910 }, { "grad_norm": 0.32284289598464966, "learning_rate": 9.405072343859805e-06, "loss": 0.0202, "step": 64920 }, { "grad_norm": 0.34189558029174805, "learning_rate": 9.393009646593043e-06, "loss": 0.0177, "step": 64930 }, { "grad_norm": 0.2924114763736725, "learning_rate": 9.38095388795035e-06, "loss": 0.0189, "step": 64940 }, { "grad_norm": 0.36749282479286194, "learning_rate": 9.36890506999169e-06, "loss": 0.0154, "step": 64950 }, { "grad_norm": 0.22246553003787994, "learning_rate": 9.356863194775894e-06, "loss": 0.0188, "step": 64960 }, { "grad_norm": 0.23965084552764893, "learning_rate": 9.344828264360583e-06, "loss": 0.0205, "step": 64970 }, { "grad_norm": 0.26184746623039246, "learning_rate": 9.332800280802201e-06, "loss": 0.0215, "step": 64980 }, { "grad_norm": 0.17091482877731323, "learning_rate": 9.32077924615602e-06, "loss": 0.0161, "step": 64990 }, { "grad_norm": 0.32556241750717163, "learning_rate": 9.308765162476063e-06, "loss": 0.0187, "step": 65000 }, { "grad_norm": 0.2289493829011917, "learning_rate": 9.296758031815239e-06, "loss": 0.0247, "step": 65010 }, { "grad_norm": 0.2775402367115021, "learning_rate": 9.284757856225229e-06, "loss": 0.0187, "step": 65020 }, { "grad_norm": 0.20179177820682526, "learning_rate": 9.272764637756538e-06, "loss": 0.0181, "step": 65030 }, { "grad_norm": 0.40414267778396606, "learning_rate": 9.260778378458479e-06, "loss": 0.0191, "step": 65040 }, { "grad_norm": 0.15654464066028595, "learning_rate": 9.248799080379172e-06, "loss": 0.0215, "step": 65050 }, { "grad_norm": 0.2538520097732544, "learning_rate": 9.236826745565558e-06, "loss": 0.0203, "step": 65060 }, { "grad_norm": 0.23462341725826263, "learning_rate": 9.224861376063388e-06, "loss": 0.0228, "step": 65070 }, { "grad_norm": 0.15390609204769135, "learning_rate": 9.212902973917192e-06, "loss": 0.0092, "step": 65080 }, { "grad_norm": 0.2011672407388687, "learning_rate": 9.20095154117035e-06, "loss": 0.0159, "step": 65090 }, { "grad_norm": 0.285806268453598, "learning_rate": 9.189007079865036e-06, "loss": 0.014, "step": 65100 }, { "grad_norm": 0.43744465708732605, "learning_rate": 9.177069592042226e-06, "loss": 0.0164, "step": 65110 }, { "grad_norm": 0.2648254334926605, "learning_rate": 9.165139079741724e-06, "loss": 0.0256, "step": 65120 }, { "grad_norm": 0.3898414671421051, "learning_rate": 9.153215545002098e-06, "loss": 0.0235, "step": 65130 }, { "grad_norm": 0.2393018901348114, "learning_rate": 9.141298989860798e-06, "loss": 0.0229, "step": 65140 }, { "grad_norm": 0.18414057791233063, "learning_rate": 9.129389416353994e-06, "loss": 0.0159, "step": 65150 }, { "grad_norm": 0.20945484936237335, "learning_rate": 9.11748682651673e-06, "loss": 0.0154, "step": 65160 }, { "grad_norm": 0.35457873344421387, "learning_rate": 9.105591222382837e-06, "loss": 0.0204, "step": 65170 }, { "grad_norm": 0.16795560717582703, "learning_rate": 9.093702605984915e-06, "loss": 0.0208, "step": 65180 }, { "grad_norm": 0.34667032957077026, "learning_rate": 9.081820979354455e-06, "loss": 0.0192, "step": 65190 }, { "grad_norm": 0.3428152799606323, "learning_rate": 9.069946344521663e-06, "loss": 0.0171, "step": 65200 }, { "grad_norm": 0.19433586299419403, "learning_rate": 9.058078703515598e-06, "loss": 0.0164, "step": 65210 }, { "grad_norm": 0.3068735599517822, "learning_rate": 9.046218058364125e-06, "loss": 0.0224, "step": 65220 }, { "grad_norm": 0.33880794048309326, "learning_rate": 9.034364411093893e-06, "loss": 0.0184, "step": 65230 }, { "grad_norm": 0.3233110308647156, "learning_rate": 9.022517763730371e-06, "loss": 0.0175, "step": 65240 }, { "grad_norm": 0.15631315112113953, "learning_rate": 9.010678118297827e-06, "loss": 0.0173, "step": 65250 }, { "grad_norm": 0.2645179033279419, "learning_rate": 8.998845476819345e-06, "loss": 0.0159, "step": 65260 }, { "grad_norm": 0.2119128704071045, "learning_rate": 8.987019841316773e-06, "loss": 0.0163, "step": 65270 }, { "grad_norm": 0.2923799753189087, "learning_rate": 8.975201213810802e-06, "loss": 0.0136, "step": 65280 }, { "grad_norm": 0.2188945859670639, "learning_rate": 8.963389596320915e-06, "loss": 0.0165, "step": 65290 }, { "grad_norm": 0.20281203091144562, "learning_rate": 8.951584990865391e-06, "loss": 0.0168, "step": 65300 }, { "grad_norm": 0.3000743091106415, "learning_rate": 8.939787399461319e-06, "loss": 0.0235, "step": 65310 }, { "grad_norm": 0.4493682384490967, "learning_rate": 8.927996824124591e-06, "loss": 0.0195, "step": 65320 }, { "grad_norm": 0.46345028281211853, "learning_rate": 8.916213266869854e-06, "loss": 0.0203, "step": 65330 }, { "grad_norm": 0.2146388441324234, "learning_rate": 8.904436729710658e-06, "loss": 0.0154, "step": 65340 }, { "grad_norm": 0.28455600142478943, "learning_rate": 8.892667214659245e-06, "loss": 0.0139, "step": 65350 }, { "grad_norm": 0.4556962847709656, "learning_rate": 8.880904723726713e-06, "loss": 0.016, "step": 65360 }, { "grad_norm": 0.229885995388031, "learning_rate": 8.869149258922971e-06, "loss": 0.013, "step": 65370 }, { "grad_norm": 0.24869471788406372, "learning_rate": 8.857400822256662e-06, "loss": 0.0187, "step": 65380 }, { "grad_norm": 0.29520687460899353, "learning_rate": 8.845659415735324e-06, "loss": 0.0203, "step": 65390 }, { "grad_norm": 0.25553157925605774, "learning_rate": 8.83392504136521e-06, "loss": 0.0172, "step": 65400 }, { "grad_norm": 0.19528616964817047, "learning_rate": 8.822197701151407e-06, "loss": 0.0213, "step": 65410 }, { "grad_norm": 0.2824881970882416, "learning_rate": 8.810477397097804e-06, "loss": 0.018, "step": 65420 }, { "grad_norm": 0.12194132059812546, "learning_rate": 8.798764131207077e-06, "loss": 0.0154, "step": 65430 }, { "grad_norm": 0.28284767270088196, "learning_rate": 8.787057905480706e-06, "loss": 0.015, "step": 65440 }, { "grad_norm": 0.2774244546890259, "learning_rate": 8.775358721918958e-06, "loss": 0.0204, "step": 65450 }, { "grad_norm": 0.24763457477092743, "learning_rate": 8.763666582520923e-06, "loss": 0.0184, "step": 65460 }, { "grad_norm": 0.2508252263069153, "learning_rate": 8.751981489284445e-06, "loss": 0.0157, "step": 65470 }, { "grad_norm": 0.23871645331382751, "learning_rate": 8.740303444206188e-06, "loss": 0.0181, "step": 65480 }, { "grad_norm": 0.17003552615642548, "learning_rate": 8.72863244928162e-06, "loss": 0.0129, "step": 65490 }, { "grad_norm": 0.2074890434741974, "learning_rate": 8.716968506504991e-06, "loss": 0.0205, "step": 65500 }, { "grad_norm": 0.3200225234031677, "learning_rate": 8.70531161786935e-06, "loss": 0.0119, "step": 65510 }, { "grad_norm": 0.2644946575164795, "learning_rate": 8.693661785366558e-06, "loss": 0.0237, "step": 65520 }, { "grad_norm": 0.20937645435333252, "learning_rate": 8.682019010987208e-06, "loss": 0.0203, "step": 65530 }, { "grad_norm": 0.3090267777442932, "learning_rate": 8.670383296720786e-06, "loss": 0.0184, "step": 65540 }, { "grad_norm": 0.354397177696228, "learning_rate": 8.658754644555478e-06, "loss": 0.0155, "step": 65550 }, { "grad_norm": 0.21898673474788666, "learning_rate": 8.647133056478313e-06, "loss": 0.0154, "step": 65560 }, { "grad_norm": 0.2437669038772583, "learning_rate": 8.635518534475123e-06, "loss": 0.0184, "step": 65570 }, { "grad_norm": 0.19081228971481323, "learning_rate": 8.623911080530467e-06, "loss": 0.0201, "step": 65580 }, { "grad_norm": 0.23383817076683044, "learning_rate": 8.61231069662779e-06, "loss": 0.0175, "step": 65590 }, { "grad_norm": 0.32488134503364563, "learning_rate": 8.600717384749252e-06, "loss": 0.0176, "step": 65600 }, { "grad_norm": 0.4128585159778595, "learning_rate": 8.58913114687584e-06, "loss": 0.0153, "step": 65610 }, { "grad_norm": 0.4328790307044983, "learning_rate": 8.57755198498732e-06, "loss": 0.0224, "step": 65620 }, { "grad_norm": 0.25498610734939575, "learning_rate": 8.565979901062265e-06, "loss": 0.0335, "step": 65630 }, { "grad_norm": 0.20918957889080048, "learning_rate": 8.554414897078033e-06, "loss": 0.015, "step": 65640 }, { "grad_norm": 0.20457753539085388, "learning_rate": 8.542856975010727e-06, "loss": 0.0137, "step": 65650 }, { "grad_norm": 0.17999228835105896, "learning_rate": 8.531306136835337e-06, "loss": 0.0133, "step": 65660 }, { "grad_norm": 0.35080233216285706, "learning_rate": 8.519762384525548e-06, "loss": 0.0159, "step": 65670 }, { "grad_norm": 0.27117207646369934, "learning_rate": 8.508225720053875e-06, "loss": 0.0173, "step": 65680 }, { "grad_norm": 0.27547505497932434, "learning_rate": 8.496696145391625e-06, "loss": 0.0163, "step": 65690 }, { "grad_norm": 0.26471614837646484, "learning_rate": 8.485173662508889e-06, "loss": 0.0148, "step": 65700 }, { "grad_norm": 0.26440712809562683, "learning_rate": 8.473658273374536e-06, "loss": 0.0167, "step": 65710 }, { "grad_norm": 0.22747166454792023, "learning_rate": 8.462149979956253e-06, "loss": 0.0159, "step": 65720 }, { "grad_norm": 0.16348114609718323, "learning_rate": 8.450648784220461e-06, "loss": 0.0176, "step": 65730 }, { "grad_norm": 0.2983046770095825, "learning_rate": 8.439154688132417e-06, "loss": 0.0163, "step": 65740 }, { "grad_norm": 0.4969117343425751, "learning_rate": 8.427667693656143e-06, "loss": 0.0181, "step": 65750 }, { "grad_norm": 0.276297003030777, "learning_rate": 8.416187802754454e-06, "loss": 0.0174, "step": 65760 }, { "grad_norm": 0.1949697732925415, "learning_rate": 8.404715017388965e-06, "loss": 0.0152, "step": 65770 }, { "grad_norm": 0.3424392342567444, "learning_rate": 8.393249339520015e-06, "loss": 0.0232, "step": 65780 }, { "grad_norm": 0.2650253474712372, "learning_rate": 8.381790771106834e-06, "loss": 0.0163, "step": 65790 }, { "grad_norm": 0.24623876810073853, "learning_rate": 8.370339314107339e-06, "loss": 0.015, "step": 65800 }, { "grad_norm": 0.32690706849098206, "learning_rate": 8.358894970478281e-06, "loss": 0.0149, "step": 65810 }, { "grad_norm": 0.2513197958469391, "learning_rate": 8.347457742175196e-06, "loss": 0.0143, "step": 65820 }, { "grad_norm": 0.34200742840766907, "learning_rate": 8.33602763115236e-06, "loss": 0.0129, "step": 65830 }, { "grad_norm": 0.2781871259212494, "learning_rate": 8.324604639362916e-06, "loss": 0.0229, "step": 65840 }, { "grad_norm": 0.16135212779045105, "learning_rate": 8.31318876875869e-06, "loss": 0.0164, "step": 65850 }, { "grad_norm": 0.36274680495262146, "learning_rate": 8.30178002129039e-06, "loss": 0.0163, "step": 65860 }, { "grad_norm": 0.2479373812675476, "learning_rate": 8.290378398907423e-06, "loss": 0.016, "step": 65870 }, { "grad_norm": 0.284271776676178, "learning_rate": 8.278983903558029e-06, "loss": 0.0237, "step": 65880 }, { "grad_norm": 0.250142365694046, "learning_rate": 8.26759653718921e-06, "loss": 0.0181, "step": 65890 }, { "grad_norm": 0.2686717212200165, "learning_rate": 8.25621630174676e-06, "loss": 0.0183, "step": 65900 }, { "grad_norm": 0.2828916311264038, "learning_rate": 8.244843199175261e-06, "loss": 0.027, "step": 65910 }, { "grad_norm": 0.2022179812192917, "learning_rate": 8.233477231418046e-06, "loss": 0.0177, "step": 65920 }, { "grad_norm": 0.2710687816143036, "learning_rate": 8.22211840041725e-06, "loss": 0.0196, "step": 65930 }, { "grad_norm": 0.3143119513988495, "learning_rate": 8.210766708113792e-06, "loss": 0.0224, "step": 65940 }, { "grad_norm": 0.3025912046432495, "learning_rate": 8.199422156447367e-06, "loss": 0.0205, "step": 65950 }, { "grad_norm": 0.31161797046661377, "learning_rate": 8.188084747356451e-06, "loss": 0.0233, "step": 65960 }, { "grad_norm": 0.2110697329044342, "learning_rate": 8.176754482778299e-06, "loss": 0.0174, "step": 65970 }, { "grad_norm": 0.40123412013053894, "learning_rate": 8.165431364648918e-06, "loss": 0.0209, "step": 65980 }, { "grad_norm": 0.20773063600063324, "learning_rate": 8.154115394903162e-06, "loss": 0.0154, "step": 65990 }, { "grad_norm": 0.2818543314933777, "learning_rate": 8.142806575474582e-06, "loss": 0.0167, "step": 66000 }, { "grad_norm": 0.22643901407718658, "learning_rate": 8.131504908295562e-06, "loss": 0.0149, "step": 66010 }, { "grad_norm": 0.23598985373973846, "learning_rate": 8.120210395297262e-06, "loss": 0.0139, "step": 66020 }, { "grad_norm": 0.18635423481464386, "learning_rate": 8.108923038409565e-06, "loss": 0.0197, "step": 66030 }, { "grad_norm": 0.3684239089488983, "learning_rate": 8.097642839561226e-06, "loss": 0.0244, "step": 66040 }, { "grad_norm": 0.29914391040802, "learning_rate": 8.08636980067966e-06, "loss": 0.0225, "step": 66050 }, { "grad_norm": 0.2878152132034302, "learning_rate": 8.075103923691186e-06, "loss": 0.0125, "step": 66060 }, { "grad_norm": 0.18854598701000214, "learning_rate": 8.063845210520793e-06, "loss": 0.0176, "step": 66070 }, { "grad_norm": 0.16202446818351746, "learning_rate": 8.052593663092295e-06, "loss": 0.0188, "step": 66080 }, { "grad_norm": 0.4233764410018921, "learning_rate": 8.04134928332827e-06, "loss": 0.0212, "step": 66090 }, { "grad_norm": 0.2669322192668915, "learning_rate": 8.030112073150086e-06, "loss": 0.0175, "step": 66100 }, { "grad_norm": 0.3069935142993927, "learning_rate": 8.018882034477881e-06, "loss": 0.0148, "step": 66110 }, { "grad_norm": 0.34544268250465393, "learning_rate": 8.007659169230541e-06, "loss": 0.0173, "step": 66120 }, { "grad_norm": 0.6333113312721252, "learning_rate": 7.996443479325755e-06, "loss": 0.0198, "step": 66130 }, { "grad_norm": 0.21503569185733795, "learning_rate": 7.985234966679977e-06, "loss": 0.0205, "step": 66140 }, { "grad_norm": 0.4213978946208954, "learning_rate": 7.974033633208438e-06, "loss": 0.0216, "step": 66150 }, { "grad_norm": 0.21531322598457336, "learning_rate": 7.962839480825135e-06, "loss": 0.0148, "step": 66160 }, { "grad_norm": 0.3099919855594635, "learning_rate": 7.951652511442858e-06, "loss": 0.0229, "step": 66170 }, { "grad_norm": 0.2923071086406708, "learning_rate": 7.940472726973125e-06, "loss": 0.0207, "step": 66180 }, { "grad_norm": 0.3573063313961029, "learning_rate": 7.929300129326289e-06, "loss": 0.0149, "step": 66190 }, { "grad_norm": 0.4327256679534912, "learning_rate": 7.91813472041142e-06, "loss": 0.022, "step": 66200 }, { "grad_norm": 0.3483238220214844, "learning_rate": 7.906976502136376e-06, "loss": 0.0185, "step": 66210 }, { "grad_norm": 0.2507501244544983, "learning_rate": 7.89582547640782e-06, "loss": 0.0198, "step": 66220 }, { "grad_norm": 0.165841206908226, "learning_rate": 7.884681645131115e-06, "loss": 0.019, "step": 66230 }, { "grad_norm": 0.3975257873535156, "learning_rate": 7.87354501021048e-06, "loss": 0.0179, "step": 66240 }, { "grad_norm": 0.29602959752082825, "learning_rate": 7.86241557354882e-06, "loss": 0.0162, "step": 66250 }, { "grad_norm": 0.38734084367752075, "learning_rate": 7.8512933370479e-06, "loss": 0.0205, "step": 66260 }, { "grad_norm": 0.26797229051589966, "learning_rate": 7.840178302608158e-06, "loss": 0.019, "step": 66270 }, { "grad_norm": 0.25927281379699707, "learning_rate": 7.82907047212888e-06, "loss": 0.0209, "step": 66280 }, { "grad_norm": 0.393329381942749, "learning_rate": 7.81796984750809e-06, "loss": 0.0196, "step": 66290 }, { "grad_norm": 0.21646004915237427, "learning_rate": 7.806876430642546e-06, "loss": 0.0232, "step": 66300 }, { "grad_norm": 0.4182911515235901, "learning_rate": 7.795790223427862e-06, "loss": 0.0197, "step": 66310 }, { "grad_norm": 0.29440364241600037, "learning_rate": 7.784711227758324e-06, "loss": 0.0217, "step": 66320 }, { "grad_norm": 0.33033254742622375, "learning_rate": 7.773639445527053e-06, "loss": 0.0192, "step": 66330 }, { "grad_norm": 0.21823042631149292, "learning_rate": 7.762574878625905e-06, "loss": 0.0201, "step": 66340 }, { "grad_norm": 0.1940721571445465, "learning_rate": 7.751517528945513e-06, "loss": 0.0143, "step": 66350 }, { "grad_norm": 0.26531192660331726, "learning_rate": 7.740467398375278e-06, "loss": 0.0188, "step": 66360 }, { "grad_norm": 0.29044783115386963, "learning_rate": 7.729424488803378e-06, "loss": 0.0166, "step": 66370 }, { "grad_norm": 0.2658028304576874, "learning_rate": 7.71838880211671e-06, "loss": 0.024, "step": 66380 }, { "grad_norm": 0.26390358805656433, "learning_rate": 7.707360340200997e-06, "loss": 0.0136, "step": 66390 }, { "grad_norm": 0.23339252173900604, "learning_rate": 7.696339104940697e-06, "loss": 0.0175, "step": 66400 }, { "grad_norm": 0.25403350591659546, "learning_rate": 7.685325098219038e-06, "loss": 0.0182, "step": 66410 }, { "grad_norm": 0.2586964964866638, "learning_rate": 7.674318321918017e-06, "loss": 0.021, "step": 66420 }, { "grad_norm": 0.3255021274089813, "learning_rate": 7.663318777918366e-06, "loss": 0.0208, "step": 66430 }, { "grad_norm": 0.208319753408432, "learning_rate": 7.652326468099647e-06, "loss": 0.0218, "step": 66440 }, { "grad_norm": 0.30993807315826416, "learning_rate": 7.641341394340096e-06, "loss": 0.0149, "step": 66450 }, { "grad_norm": 0.31200945377349854, "learning_rate": 7.630363558516818e-06, "loss": 0.0196, "step": 66460 }, { "grad_norm": 0.30585741996765137, "learning_rate": 7.619392962505578e-06, "loss": 0.0226, "step": 66470 }, { "grad_norm": 0.26244696974754333, "learning_rate": 7.6084296081809725e-06, "loss": 0.0201, "step": 66480 }, { "grad_norm": 0.44154492020606995, "learning_rate": 7.597473497416347e-06, "loss": 0.0219, "step": 66490 }, { "grad_norm": 0.17275336384773254, "learning_rate": 7.586524632083764e-06, "loss": 0.0172, "step": 66500 }, { "grad_norm": 0.39047396183013916, "learning_rate": 7.5755830140541326e-06, "loss": 0.0181, "step": 66510 }, { "grad_norm": 0.18846699595451355, "learning_rate": 7.5646486451970475e-06, "loss": 0.0165, "step": 66520 }, { "grad_norm": 0.2487601339817047, "learning_rate": 7.553721527380897e-06, "loss": 0.0248, "step": 66530 }, { "grad_norm": 0.21681340038776398, "learning_rate": 7.542801662472826e-06, "loss": 0.0141, "step": 66540 }, { "grad_norm": 0.1862841099500656, "learning_rate": 7.5318890523387475e-06, "loss": 0.0134, "step": 66550 }, { "grad_norm": 0.22284387052059174, "learning_rate": 7.52098369884332e-06, "loss": 0.0194, "step": 66560 }, { "grad_norm": 0.2936161160469055, "learning_rate": 7.510085603849992e-06, "loss": 0.0237, "step": 66570 }, { "grad_norm": 0.3275645673274994, "learning_rate": 7.499194769220918e-06, "loss": 0.0154, "step": 66580 }, { "grad_norm": 0.22850555181503296, "learning_rate": 7.48831119681706e-06, "loss": 0.0256, "step": 66590 }, { "grad_norm": 0.35712549090385437, "learning_rate": 7.477434888498119e-06, "loss": 0.0194, "step": 66600 }, { "grad_norm": 0.25481271743774414, "learning_rate": 7.466565846122564e-06, "loss": 0.0218, "step": 66610 }, { "grad_norm": 0.4195387363433838, "learning_rate": 7.455704071547626e-06, "loss": 0.0184, "step": 66620 }, { "grad_norm": 0.30372729897499084, "learning_rate": 7.444849566629247e-06, "loss": 0.0182, "step": 66630 }, { "grad_norm": 0.38176003098487854, "learning_rate": 7.434002333222212e-06, "loss": 0.015, "step": 66640 }, { "grad_norm": 0.2694861888885498, "learning_rate": 7.423162373179976e-06, "loss": 0.0202, "step": 66650 }, { "grad_norm": 0.3771134316921234, "learning_rate": 7.412329688354835e-06, "loss": 0.0212, "step": 66660 }, { "grad_norm": 0.24455171823501587, "learning_rate": 7.40150428059776e-06, "loss": 0.0179, "step": 66670 }, { "grad_norm": 0.3901103734970093, "learning_rate": 7.3906861517585354e-06, "loss": 0.0215, "step": 66680 }, { "grad_norm": 0.31908515095710754, "learning_rate": 7.37987530368569e-06, "loss": 0.0166, "step": 66690 }, { "grad_norm": 0.2137337327003479, "learning_rate": 7.369071738226474e-06, "loss": 0.0169, "step": 66700 }, { "grad_norm": 0.23716148734092712, "learning_rate": 7.358275457226954e-06, "loss": 0.0159, "step": 66710 }, { "grad_norm": 0.27068665623664856, "learning_rate": 7.347486462531899e-06, "loss": 0.0171, "step": 66720 }, { "grad_norm": 0.18403682112693787, "learning_rate": 7.336704755984858e-06, "loss": 0.0141, "step": 66730 }, { "grad_norm": 0.23390699923038483, "learning_rate": 7.325930339428133e-06, "loss": 0.0197, "step": 66740 }, { "grad_norm": 0.278536319732666, "learning_rate": 7.315163214702769e-06, "loss": 0.0175, "step": 66750 }, { "grad_norm": 0.3790436387062073, "learning_rate": 7.304403383648595e-06, "loss": 0.0196, "step": 66760 }, { "grad_norm": 0.1802259385585785, "learning_rate": 7.293650848104139e-06, "loss": 0.0151, "step": 66770 }, { "grad_norm": 0.38605713844299316, "learning_rate": 7.2829056099067374e-06, "loss": 0.022, "step": 66780 }, { "grad_norm": 0.2338934689760208, "learning_rate": 7.2721676708924494e-06, "loss": 0.0177, "step": 66790 }, { "grad_norm": 0.41278570890426636, "learning_rate": 7.261437032896096e-06, "loss": 0.0312, "step": 66800 }, { "grad_norm": 0.28571102023124695, "learning_rate": 7.250713697751255e-06, "loss": 0.0194, "step": 66810 }, { "grad_norm": 0.2399584949016571, "learning_rate": 7.239997667290255e-06, "loss": 0.0166, "step": 66820 }, { "grad_norm": 0.1632802039384842, "learning_rate": 7.2292889433441425e-06, "loss": 0.0161, "step": 66830 }, { "grad_norm": 0.23418691754341125, "learning_rate": 7.218587527742793e-06, "loss": 0.0145, "step": 66840 }, { "grad_norm": 0.7099646329879761, "learning_rate": 7.207893422314749e-06, "loss": 0.0165, "step": 66850 }, { "grad_norm": 0.22476166486740112, "learning_rate": 7.1972066288873545e-06, "loss": 0.0125, "step": 66860 }, { "grad_norm": 0.25811904668807983, "learning_rate": 7.186527149286687e-06, "loss": 0.0295, "step": 66870 }, { "grad_norm": 0.33137238025665283, "learning_rate": 7.175854985337576e-06, "loss": 0.021, "step": 66880 }, { "grad_norm": 0.2775493562221527, "learning_rate": 7.16519013886362e-06, "loss": 0.0204, "step": 66890 }, { "grad_norm": 0.20654205977916718, "learning_rate": 7.154532611687109e-06, "loss": 0.0167, "step": 66900 }, { "grad_norm": 0.29168573021888733, "learning_rate": 7.143882405629177e-06, "loss": 0.0154, "step": 66910 }, { "grad_norm": 0.1696626991033554, "learning_rate": 7.133239522509605e-06, "loss": 0.0152, "step": 66920 }, { "grad_norm": 0.3691703975200653, "learning_rate": 7.1226039641469956e-06, "loss": 0.0149, "step": 66930 }, { "grad_norm": 0.21113207936286926, "learning_rate": 7.111975732358678e-06, "loss": 0.0204, "step": 66940 }, { "grad_norm": 0.2609310746192932, "learning_rate": 7.101354828960693e-06, "loss": 0.0135, "step": 66950 }, { "grad_norm": 0.18335023522377014, "learning_rate": 7.090741255767918e-06, "loss": 0.0123, "step": 66960 }, { "grad_norm": 0.2522009015083313, "learning_rate": 7.080135014593875e-06, "loss": 0.0153, "step": 66970 }, { "grad_norm": 0.2952776551246643, "learning_rate": 7.069536107250896e-06, "loss": 0.0151, "step": 66980 }, { "grad_norm": 0.3436570465564728, "learning_rate": 7.058944535550049e-06, "loss": 0.0135, "step": 66990 }, { "grad_norm": 0.3492586016654968, "learning_rate": 7.048360301301138e-06, "loss": 0.019, "step": 67000 }, { "grad_norm": 0.21212637424468994, "learning_rate": 7.03778340631272e-06, "loss": 0.0171, "step": 67010 }, { "grad_norm": 0.27424558997154236, "learning_rate": 7.0272138523921e-06, "loss": 0.0135, "step": 67020 }, { "grad_norm": 0.21721015870571136, "learning_rate": 7.016651641345334e-06, "loss": 0.0209, "step": 67030 }, { "grad_norm": 0.3645990490913391, "learning_rate": 7.0060967749772e-06, "loss": 0.0157, "step": 67040 }, { "grad_norm": 0.3738721013069153, "learning_rate": 6.995549255091238e-06, "loss": 0.0316, "step": 67050 }, { "grad_norm": 0.23841752111911774, "learning_rate": 6.98500908348973e-06, "loss": 0.0222, "step": 67060 }, { "grad_norm": 0.3601408004760742, "learning_rate": 6.974476261973711e-06, "loss": 0.0183, "step": 67070 }, { "grad_norm": 0.16672340035438538, "learning_rate": 6.963950792342949e-06, "loss": 0.0128, "step": 67080 }, { "grad_norm": 0.22421860694885254, "learning_rate": 6.9534326763959715e-06, "loss": 0.0144, "step": 67090 }, { "grad_norm": 0.30868053436279297, "learning_rate": 6.942921915929995e-06, "loss": 0.0171, "step": 67100 }, { "grad_norm": 0.39771202206611633, "learning_rate": 6.9324185127410735e-06, "loss": 0.0164, "step": 67110 }, { "grad_norm": 0.3389632999897003, "learning_rate": 6.921922468623921e-06, "loss": 0.0187, "step": 67120 }, { "grad_norm": 0.25474271178245544, "learning_rate": 6.911433785372023e-06, "loss": 0.0192, "step": 67130 }, { "grad_norm": 0.1602192372083664, "learning_rate": 6.900952464777632e-06, "loss": 0.0206, "step": 67140 }, { "grad_norm": 0.35802018642425537, "learning_rate": 6.8904785086316815e-06, "loss": 0.0183, "step": 67150 }, { "grad_norm": 0.24868950247764587, "learning_rate": 6.880011918723927e-06, "loss": 0.0159, "step": 67160 }, { "grad_norm": 0.2517662048339844, "learning_rate": 6.86955269684279e-06, "loss": 0.0148, "step": 67170 }, { "grad_norm": 0.45491111278533936, "learning_rate": 6.859100844775473e-06, "loss": 0.0286, "step": 67180 }, { "grad_norm": 0.3628300726413727, "learning_rate": 6.84865636430792e-06, "loss": 0.0145, "step": 67190 }, { "grad_norm": 0.15618404746055603, "learning_rate": 6.838219257224804e-06, "loss": 0.0101, "step": 67200 }, { "grad_norm": 0.16180981695652008, "learning_rate": 6.827789525309536e-06, "loss": 0.0205, "step": 67210 }, { "grad_norm": 0.3387438952922821, "learning_rate": 6.81736717034428e-06, "loss": 0.0218, "step": 67220 }, { "grad_norm": 0.2438049167394638, "learning_rate": 6.806952194109933e-06, "loss": 0.0246, "step": 67230 }, { "grad_norm": 0.19845996797084808, "learning_rate": 6.796544598386112e-06, "loss": 0.0147, "step": 67240 }, { "grad_norm": 0.4070035517215729, "learning_rate": 6.786144384951204e-06, "loss": 0.0226, "step": 67250 }, { "grad_norm": 0.21765881776809692, "learning_rate": 6.775751555582322e-06, "loss": 0.0175, "step": 67260 }, { "grad_norm": 0.24242641031742096, "learning_rate": 6.76536611205531e-06, "loss": 0.017, "step": 67270 }, { "grad_norm": 0.165487602353096, "learning_rate": 6.754988056144762e-06, "loss": 0.0201, "step": 67280 }, { "grad_norm": 0.3093235492706299, "learning_rate": 6.744617389624014e-06, "loss": 0.017, "step": 67290 }, { "grad_norm": 0.389522910118103, "learning_rate": 6.734254114265087e-06, "loss": 0.0157, "step": 67300 }, { "grad_norm": 0.20521409809589386, "learning_rate": 6.723898231838843e-06, "loss": 0.0167, "step": 67310 }, { "grad_norm": 0.22603419423103333, "learning_rate": 6.713549744114766e-06, "loss": 0.0204, "step": 67320 }, { "grad_norm": 0.18112926185131073, "learning_rate": 6.703208652861159e-06, "loss": 0.0187, "step": 67330 }, { "grad_norm": 0.3036172688007355, "learning_rate": 6.69287495984503e-06, "loss": 0.0194, "step": 67340 }, { "grad_norm": 0.22572718560695648, "learning_rate": 6.6825486668321e-06, "loss": 0.0176, "step": 67350 }, { "grad_norm": 0.19858890771865845, "learning_rate": 6.672229775586886e-06, "loss": 0.019, "step": 67360 }, { "grad_norm": 0.16791000962257385, "learning_rate": 6.661918287872576e-06, "loss": 0.0145, "step": 67370 }, { "grad_norm": 0.18757638335227966, "learning_rate": 6.6516142054511346e-06, "loss": 0.0164, "step": 67380 }, { "grad_norm": 0.21367385983467102, "learning_rate": 6.641317530083241e-06, "loss": 0.0164, "step": 67390 }, { "grad_norm": 0.2785041928291321, "learning_rate": 6.631028263528322e-06, "loss": 0.0174, "step": 67400 }, { "grad_norm": 0.40402647852897644, "learning_rate": 6.620746407544537e-06, "loss": 0.0154, "step": 67410 }, { "grad_norm": 0.2798149585723877, "learning_rate": 6.610471963888742e-06, "loss": 0.0202, "step": 67420 }, { "grad_norm": 0.22628694772720337, "learning_rate": 6.600204934316606e-06, "loss": 0.0147, "step": 67430 }, { "grad_norm": 0.3544834852218628, "learning_rate": 6.589945320582452e-06, "loss": 0.0209, "step": 67440 }, { "grad_norm": 0.2815662622451782, "learning_rate": 6.579693124439374e-06, "loss": 0.0167, "step": 67450 }, { "grad_norm": 0.1828363984823227, "learning_rate": 6.56944834763919e-06, "loss": 0.016, "step": 67460 }, { "grad_norm": 0.3927805423736572, "learning_rate": 6.5592109919324575e-06, "loss": 0.0177, "step": 67470 }, { "grad_norm": 0.2566711902618408, "learning_rate": 6.54898105906846e-06, "loss": 0.0205, "step": 67480 }, { "grad_norm": 0.26309797167778015, "learning_rate": 6.5387585507952155e-06, "loss": 0.0254, "step": 67490 }, { "grad_norm": 0.30788829922676086, "learning_rate": 6.528543468859461e-06, "loss": 0.0185, "step": 67500 }, { "grad_norm": 0.20335769653320312, "learning_rate": 6.518335815006682e-06, "loss": 0.0248, "step": 67510 }, { "grad_norm": 0.43045851588249207, "learning_rate": 6.5081355909810845e-06, "loss": 0.0228, "step": 67520 }, { "grad_norm": 0.25432634353637695, "learning_rate": 6.497942798525608e-06, "loss": 0.0189, "step": 67530 }, { "grad_norm": 0.3309043049812317, "learning_rate": 6.487757439381936e-06, "loss": 0.022, "step": 67540 }, { "grad_norm": 0.329324334859848, "learning_rate": 6.477579515290433e-06, "loss": 0.0194, "step": 67550 }, { "grad_norm": 0.17949755489826202, "learning_rate": 6.467409027990273e-06, "loss": 0.0252, "step": 67560 }, { "grad_norm": 0.23966757953166962, "learning_rate": 6.457245979219279e-06, "loss": 0.0235, "step": 67570 }, { "grad_norm": 0.5001302361488342, "learning_rate": 6.447090370714054e-06, "loss": 0.0276, "step": 67580 }, { "grad_norm": 0.2180784046649933, "learning_rate": 6.436942204209917e-06, "loss": 0.0182, "step": 67590 }, { "grad_norm": 0.24725812673568726, "learning_rate": 6.4268014814408804e-06, "loss": 0.0257, "step": 67600 }, { "grad_norm": 0.2629917562007904, "learning_rate": 6.41666820413977e-06, "loss": 0.0187, "step": 67610 }, { "grad_norm": 0.2337781935930252, "learning_rate": 6.406542374038033e-06, "loss": 0.023, "step": 67620 }, { "grad_norm": 0.18078899383544922, "learning_rate": 6.396423992865935e-06, "loss": 0.0176, "step": 67630 }, { "grad_norm": 0.2791828513145447, "learning_rate": 6.386313062352412e-06, "loss": 0.0153, "step": 67640 }, { "grad_norm": 0.2490597367286682, "learning_rate": 6.376209584225152e-06, "loss": 0.0128, "step": 67650 }, { "grad_norm": 0.2782345414161682, "learning_rate": 6.366113560210557e-06, "loss": 0.0169, "step": 67660 }, { "grad_norm": 0.3985343873500824, "learning_rate": 6.356024992033766e-06, "loss": 0.0218, "step": 67670 }, { "grad_norm": 0.16362927854061127, "learning_rate": 6.345943881418648e-06, "loss": 0.0157, "step": 67680 }, { "grad_norm": 0.2258237898349762, "learning_rate": 6.33587023008777e-06, "loss": 0.0183, "step": 67690 }, { "grad_norm": 0.33019131422042847, "learning_rate": 6.32580403976245e-06, "loss": 0.0167, "step": 67700 }, { "grad_norm": 0.38782599568367004, "learning_rate": 6.31574531216273e-06, "loss": 0.0161, "step": 67710 }, { "grad_norm": 0.2882084846496582, "learning_rate": 6.305694049007371e-06, "loss": 0.0205, "step": 67720 }, { "grad_norm": 0.34673866629600525, "learning_rate": 6.2956502520138575e-06, "loss": 0.0179, "step": 67730 }, { "grad_norm": 0.4866564869880676, "learning_rate": 6.285613922898409e-06, "loss": 0.0162, "step": 67740 }, { "grad_norm": 0.26956209540367126, "learning_rate": 6.275585063375927e-06, "loss": 0.0188, "step": 67750 }, { "grad_norm": 0.23945170640945435, "learning_rate": 6.265563675160113e-06, "loss": 0.0199, "step": 67760 }, { "grad_norm": 0.23837541043758392, "learning_rate": 6.25554975996332e-06, "loss": 0.0174, "step": 67770 }, { "grad_norm": 0.25383418798446655, "learning_rate": 6.245543319496661e-06, "loss": 0.0137, "step": 67780 }, { "grad_norm": 0.27658209204673767, "learning_rate": 6.2355443554699685e-06, "loss": 0.0163, "step": 67790 }, { "grad_norm": 0.31048229336738586, "learning_rate": 6.225552869591766e-06, "loss": 0.017, "step": 67800 }, { "grad_norm": 0.2676808834075928, "learning_rate": 6.215568863569365e-06, "loss": 0.0187, "step": 67810 }, { "grad_norm": 0.38390904664993286, "learning_rate": 6.205592339108712e-06, "loss": 0.024, "step": 67820 }, { "grad_norm": 0.24285100400447845, "learning_rate": 6.195623297914577e-06, "loss": 0.0162, "step": 67830 }, { "grad_norm": 0.25300517678260803, "learning_rate": 6.185661741690357e-06, "loss": 0.0125, "step": 67840 }, { "grad_norm": 0.24598750472068787, "learning_rate": 6.1757076721382145e-06, "loss": 0.022, "step": 67850 }, { "grad_norm": 0.25099021196365356, "learning_rate": 6.165761090959038e-06, "loss": 0.0245, "step": 67860 }, { "grad_norm": 0.28118667006492615, "learning_rate": 6.155821999852424e-06, "loss": 0.0125, "step": 67870 }, { "grad_norm": 0.19157834351062775, "learning_rate": 6.145890400516696e-06, "loss": 0.017, "step": 67880 }, { "grad_norm": 0.1952914148569107, "learning_rate": 6.1359662946488816e-06, "loss": 0.0149, "step": 67890 }, { "grad_norm": 0.23988297581672668, "learning_rate": 6.12604968394474e-06, "loss": 0.0119, "step": 67900 }, { "grad_norm": 0.36938345432281494, "learning_rate": 6.11614057009875e-06, "loss": 0.0195, "step": 67910 }, { "grad_norm": 0.2526499330997467, "learning_rate": 6.106238954804111e-06, "loss": 0.0132, "step": 67920 }, { "grad_norm": 0.20559945702552795, "learning_rate": 6.096344839752738e-06, "loss": 0.0159, "step": 67930 }, { "grad_norm": 0.25582507252693176, "learning_rate": 6.086458226635278e-06, "loss": 0.0164, "step": 67940 }, { "grad_norm": 0.23708154261112213, "learning_rate": 6.076579117141046e-06, "loss": 0.0304, "step": 67950 }, { "grad_norm": 0.3651677072048187, "learning_rate": 6.066707512958153e-06, "loss": 0.0182, "step": 67960 }, { "grad_norm": 0.16487976908683777, "learning_rate": 6.056843415773361e-06, "loss": 0.0176, "step": 67970 }, { "grad_norm": 0.31039920449256897, "learning_rate": 6.0469868272721776e-06, "loss": 0.0181, "step": 67980 }, { "grad_norm": 0.2716139554977417, "learning_rate": 6.037137749138844e-06, "loss": 0.0226, "step": 67990 }, { "grad_norm": 0.39727067947387695, "learning_rate": 6.027296183056252e-06, "loss": 0.0223, "step": 68000 }, { "grad_norm": 0.3340345025062561, "learning_rate": 6.017462130706114e-06, "loss": 0.0179, "step": 68010 }, { "grad_norm": 0.2434191107749939, "learning_rate": 6.0076355937687516e-06, "loss": 0.0155, "step": 68020 }, { "grad_norm": 0.3003213405609131, "learning_rate": 5.9978165739232925e-06, "loss": 0.0167, "step": 68030 }, { "grad_norm": 0.20214757323265076, "learning_rate": 5.988005072847508e-06, "loss": 0.0164, "step": 68040 }, { "grad_norm": 0.34140104055404663, "learning_rate": 5.978201092217928e-06, "loss": 0.0223, "step": 68050 }, { "grad_norm": 0.17061281204223633, "learning_rate": 5.9684046337097895e-06, "loss": 0.0116, "step": 68060 }, { "grad_norm": 0.17809420824050903, "learning_rate": 5.958615698997016e-06, "loss": 0.0173, "step": 68070 }, { "grad_norm": 0.2519287168979645, "learning_rate": 5.948834289752303e-06, "loss": 0.0133, "step": 68080 }, { "grad_norm": 0.22733627259731293, "learning_rate": 5.939060407646996e-06, "loss": 0.0222, "step": 68090 }, { "grad_norm": 0.24131658673286438, "learning_rate": 5.929294054351198e-06, "loss": 0.0173, "step": 68100 }, { "grad_norm": 0.16600921750068665, "learning_rate": 5.919535231533707e-06, "loss": 0.0148, "step": 68110 }, { "grad_norm": 0.39407625794410706, "learning_rate": 5.909783940862046e-06, "loss": 0.016, "step": 68120 }, { "grad_norm": 0.2343468964099884, "learning_rate": 5.900040184002436e-06, "loss": 0.0165, "step": 68130 }, { "grad_norm": 0.37365734577178955, "learning_rate": 5.890303962619831e-06, "loss": 0.0161, "step": 68140 }, { "grad_norm": 0.36229702830314636, "learning_rate": 5.88057527837787e-06, "loss": 0.0157, "step": 68150 }, { "grad_norm": 0.18355508148670197, "learning_rate": 5.870854132938924e-06, "loss": 0.0189, "step": 68160 }, { "grad_norm": 0.20397968590259552, "learning_rate": 5.861140527964071e-06, "loss": 0.0181, "step": 68170 }, { "grad_norm": 0.1624712496995926, "learning_rate": 5.851434465113098e-06, "loss": 0.0167, "step": 68180 }, { "grad_norm": 0.2813211977481842, "learning_rate": 5.841735946044524e-06, "loss": 0.0169, "step": 68190 }, { "grad_norm": 0.15707245469093323, "learning_rate": 5.832044972415523e-06, "loss": 0.0159, "step": 68200 }, { "grad_norm": 0.16244125366210938, "learning_rate": 5.8223615458820655e-06, "loss": 0.0117, "step": 68210 }, { "grad_norm": 0.18723280727863312, "learning_rate": 5.812685668098733e-06, "loss": 0.0175, "step": 68220 }, { "grad_norm": 0.27854248881340027, "learning_rate": 5.803017340718913e-06, "loss": 0.0186, "step": 68230 }, { "grad_norm": 0.11438197642564774, "learning_rate": 5.793356565394636e-06, "loss": 0.0151, "step": 68240 }, { "grad_norm": 0.22583957016468048, "learning_rate": 5.783703343776658e-06, "loss": 0.0142, "step": 68250 }, { "grad_norm": 0.22674517333507538, "learning_rate": 5.774057677514477e-06, "loss": 0.0151, "step": 68260 }, { "grad_norm": 0.33382266759872437, "learning_rate": 5.764419568256235e-06, "loss": 0.0191, "step": 68270 }, { "grad_norm": 0.34908267855644226, "learning_rate": 5.754789017648865e-06, "loss": 0.0298, "step": 68280 }, { "grad_norm": 0.2021152526140213, "learning_rate": 5.745166027337934e-06, "loss": 0.0225, "step": 68290 }, { "grad_norm": 0.2239011824131012, "learning_rate": 5.735550598967754e-06, "loss": 0.0205, "step": 68300 }, { "grad_norm": 0.2768767178058624, "learning_rate": 5.725942734181339e-06, "loss": 0.0196, "step": 68310 }, { "grad_norm": 0.30414941906929016, "learning_rate": 5.7163424346204146e-06, "loss": 0.0146, "step": 68320 }, { "grad_norm": 0.2676033675670624, "learning_rate": 5.706749701925407e-06, "loss": 0.0169, "step": 68330 }, { "grad_norm": 0.2832874357700348, "learning_rate": 5.697164537735461e-06, "loss": 0.0157, "step": 68340 }, { "grad_norm": 0.17522242665290833, "learning_rate": 5.687586943688406e-06, "loss": 0.0115, "step": 68350 }, { "grad_norm": 0.2337973564863205, "learning_rate": 5.678016921420792e-06, "loss": 0.0157, "step": 68360 }, { "grad_norm": 0.3433060050010681, "learning_rate": 5.668454472567875e-06, "loss": 0.0191, "step": 68370 }, { "grad_norm": 0.3411816954612732, "learning_rate": 5.658899598763617e-06, "loss": 0.0216, "step": 68380 }, { "grad_norm": 0.23765110969543457, "learning_rate": 5.6493523016407035e-06, "loss": 0.0128, "step": 68390 }, { "grad_norm": 0.21328553557395935, "learning_rate": 5.6398125828304615e-06, "loss": 0.0154, "step": 68400 }, { "grad_norm": 0.3147386312484741, "learning_rate": 5.630280443963015e-06, "loss": 0.0152, "step": 68410 }, { "grad_norm": 0.2939440608024597, "learning_rate": 5.620755886667112e-06, "loss": 0.0172, "step": 68420 }, { "grad_norm": 0.2838704586029053, "learning_rate": 5.611238912570271e-06, "loss": 0.0201, "step": 68430 }, { "grad_norm": 0.3227742612361908, "learning_rate": 5.601729523298649e-06, "loss": 0.0203, "step": 68440 }, { "grad_norm": 0.3364826738834381, "learning_rate": 5.592227720477161e-06, "loss": 0.0152, "step": 68450 }, { "grad_norm": 0.2783088684082031, "learning_rate": 5.582733505729415e-06, "loss": 0.0181, "step": 68460 }, { "grad_norm": 0.3412209153175354, "learning_rate": 5.573246880677668e-06, "loss": 0.0176, "step": 68470 }, { "grad_norm": 0.3386336863040924, "learning_rate": 5.56376784694298e-06, "loss": 0.0182, "step": 68480 }, { "grad_norm": 0.4129336476325989, "learning_rate": 5.554296406145027e-06, "loss": 0.0233, "step": 68490 }, { "grad_norm": 0.25773799419403076, "learning_rate": 5.544832559902219e-06, "loss": 0.0161, "step": 68500 }, { "grad_norm": 0.24171900749206543, "learning_rate": 5.53537630983168e-06, "loss": 0.0229, "step": 68510 }, { "grad_norm": 0.1833696961402893, "learning_rate": 5.5259276575492125e-06, "loss": 0.0216, "step": 68520 }, { "grad_norm": 0.1965678185224533, "learning_rate": 5.516486604669357e-06, "loss": 0.0217, "step": 68530 }, { "grad_norm": 0.22600196301937103, "learning_rate": 5.507053152805303e-06, "loss": 0.0213, "step": 68540 }, { "grad_norm": 0.2780742943286896, "learning_rate": 5.497627303568975e-06, "loss": 0.0241, "step": 68550 }, { "grad_norm": 0.1369883269071579, "learning_rate": 5.488209058571003e-06, "loss": 0.013, "step": 68560 }, { "grad_norm": 0.35263946652412415, "learning_rate": 5.4787984194207054e-06, "loss": 0.0159, "step": 68570 }, { "grad_norm": 0.2607669234275818, "learning_rate": 5.469395387726095e-06, "loss": 0.0196, "step": 68580 }, { "grad_norm": 0.2257557064294815, "learning_rate": 5.45999996509391e-06, "loss": 0.0174, "step": 68590 }, { "grad_norm": 0.1625475287437439, "learning_rate": 5.450612153129536e-06, "loss": 0.0194, "step": 68600 }, { "grad_norm": 0.18622490763664246, "learning_rate": 5.4412319534371426e-06, "loss": 0.0188, "step": 68610 }, { "grad_norm": 0.2930493950843811, "learning_rate": 5.431859367619513e-06, "loss": 0.0134, "step": 68620 }, { "grad_norm": 0.3626245856285095, "learning_rate": 5.422494397278172e-06, "loss": 0.0149, "step": 68630 }, { "grad_norm": 0.30992841720581055, "learning_rate": 5.413137044013344e-06, "loss": 0.018, "step": 68640 }, { "grad_norm": 0.251536101102829, "learning_rate": 5.403787309423941e-06, "loss": 0.0182, "step": 68650 }, { "grad_norm": 0.16337871551513672, "learning_rate": 5.3944451951075835e-06, "loss": 0.0128, "step": 68660 }, { "grad_norm": 0.3086988925933838, "learning_rate": 5.385110702660562e-06, "loss": 0.0158, "step": 68670 }, { "grad_norm": 0.19869717955589294, "learning_rate": 5.375783833677922e-06, "loss": 0.0187, "step": 68680 }, { "grad_norm": 0.34726154804229736, "learning_rate": 5.36646458975334e-06, "loss": 0.0175, "step": 68690 }, { "grad_norm": 0.34788045287132263, "learning_rate": 5.3571529724792294e-06, "loss": 0.0161, "step": 68700 }, { "grad_norm": 0.42486101388931274, "learning_rate": 5.347848983446702e-06, "loss": 0.0212, "step": 68710 }, { "grad_norm": 0.22055506706237793, "learning_rate": 5.3385526242455185e-06, "loss": 0.0158, "step": 68720 }, { "grad_norm": 0.1070917621254921, "learning_rate": 5.329263896464226e-06, "loss": 0.0168, "step": 68730 }, { "grad_norm": 0.2169661968946457, "learning_rate": 5.3199828016899715e-06, "loss": 0.0172, "step": 68740 }, { "grad_norm": 0.3213416635990143, "learning_rate": 5.310709341508657e-06, "loss": 0.0223, "step": 68750 }, { "grad_norm": 0.2418212741613388, "learning_rate": 5.301443517504861e-06, "loss": 0.0146, "step": 68760 }, { "grad_norm": 0.2928101420402527, "learning_rate": 5.292185331261862e-06, "loss": 0.0111, "step": 68770 }, { "grad_norm": 0.24170194566249847, "learning_rate": 5.28293478436162e-06, "loss": 0.0147, "step": 68780 }, { "grad_norm": 0.7934021949768066, "learning_rate": 5.273691878384829e-06, "loss": 0.0238, "step": 68790 }, { "grad_norm": 0.18847347795963287, "learning_rate": 5.264456614910812e-06, "loss": 0.0159, "step": 68800 }, { "grad_norm": 0.33924049139022827, "learning_rate": 5.255228995517647e-06, "loss": 0.0255, "step": 68810 }, { "grad_norm": 0.22312024235725403, "learning_rate": 5.246009021782067e-06, "loss": 0.0181, "step": 68820 }, { "grad_norm": 0.18499448895454407, "learning_rate": 5.2367966952795225e-06, "loss": 0.0233, "step": 68830 }, { "grad_norm": 0.12731128931045532, "learning_rate": 5.2275920175841485e-06, "loss": 0.0141, "step": 68840 }, { "grad_norm": 0.21909436583518982, "learning_rate": 5.218394990268771e-06, "loss": 0.0154, "step": 68850 }, { "grad_norm": 0.3301892578601837, "learning_rate": 5.209205614904916e-06, "loss": 0.0172, "step": 68860 }, { "grad_norm": 0.27678462862968445, "learning_rate": 5.200023893062777e-06, "loss": 0.0176, "step": 68870 }, { "grad_norm": 0.15470027923583984, "learning_rate": 5.190849826311289e-06, "loss": 0.0123, "step": 68880 }, { "grad_norm": 0.3420887291431427, "learning_rate": 5.181683416218025e-06, "loss": 0.0186, "step": 68890 }, { "grad_norm": 0.3277873396873474, "learning_rate": 5.172524664349276e-06, "loss": 0.0121, "step": 68900 }, { "grad_norm": 0.2492445409297943, "learning_rate": 5.1633735722700416e-06, "loss": 0.012, "step": 68910 }, { "grad_norm": 0.171132430434227, "learning_rate": 5.154230141543958e-06, "loss": 0.0144, "step": 68920 }, { "grad_norm": 0.20151287317276, "learning_rate": 5.145094373733433e-06, "loss": 0.0161, "step": 68930 }, { "grad_norm": 0.19402453303337097, "learning_rate": 5.135966270399478e-06, "loss": 0.0133, "step": 68940 }, { "grad_norm": 0.28665152192115784, "learning_rate": 5.126845833101857e-06, "loss": 0.0209, "step": 68950 }, { "grad_norm": 0.3246762454509735, "learning_rate": 5.1177330633989994e-06, "loss": 0.0162, "step": 68960 }, { "grad_norm": 0.2984873950481415, "learning_rate": 5.108627962848033e-06, "loss": 0.0307, "step": 68970 }, { "grad_norm": 0.20417019724845886, "learning_rate": 5.099530533004759e-06, "loss": 0.0155, "step": 68980 }, { "grad_norm": 0.18644298613071442, "learning_rate": 5.090440775423699e-06, "loss": 0.0125, "step": 68990 }, { "grad_norm": 0.2576781213283539, "learning_rate": 5.081358691658022e-06, "loss": 0.0192, "step": 69000 }, { "grad_norm": 0.3301491141319275, "learning_rate": 5.072284283259621e-06, "loss": 0.0198, "step": 69010 }, { "grad_norm": 0.19787341356277466, "learning_rate": 5.063217551779053e-06, "loss": 0.0155, "step": 69020 }, { "grad_norm": 0.24769078195095062, "learning_rate": 5.054158498765588e-06, "loss": 0.0115, "step": 69030 }, { "grad_norm": 0.28046414256095886, "learning_rate": 5.0451071257671625e-06, "loss": 0.0178, "step": 69040 }, { "grad_norm": 0.4612308144569397, "learning_rate": 5.036063434330407e-06, "loss": 0.0202, "step": 69050 }, { "grad_norm": 0.35124099254608154, "learning_rate": 5.027027426000652e-06, "loss": 0.0158, "step": 69060 }, { "grad_norm": 0.12873171269893646, "learning_rate": 5.017999102321886e-06, "loss": 0.014, "step": 69070 }, { "grad_norm": 0.5483587384223938, "learning_rate": 5.0089784648368224e-06, "loss": 0.0191, "step": 69080 }, { "grad_norm": 0.20416156947612762, "learning_rate": 4.999965515086829e-06, "loss": 0.0152, "step": 69090 }, { "grad_norm": 0.19357246160507202, "learning_rate": 4.990960254611976e-06, "loss": 0.0173, "step": 69100 }, { "grad_norm": 0.47450295090675354, "learning_rate": 4.9819626849510194e-06, "loss": 0.0177, "step": 69110 }, { "grad_norm": 0.25199323892593384, "learning_rate": 4.97297280764138e-06, "loss": 0.0228, "step": 69120 }, { "grad_norm": 0.18655407428741455, "learning_rate": 4.963990624219211e-06, "loss": 0.0187, "step": 69130 }, { "grad_norm": 0.32354602217674255, "learning_rate": 4.955016136219298e-06, "loss": 0.0152, "step": 69140 }, { "grad_norm": 0.24087949097156525, "learning_rate": 4.946049345175146e-06, "loss": 0.0168, "step": 69150 }, { "grad_norm": 0.4242513179779053, "learning_rate": 4.937090252618937e-06, "loss": 0.0183, "step": 69160 }, { "grad_norm": 0.25000298023223877, "learning_rate": 4.928138860081521e-06, "loss": 0.0212, "step": 69170 }, { "grad_norm": 0.22185154259204865, "learning_rate": 4.919195169092472e-06, "loss": 0.0186, "step": 69180 }, { "grad_norm": 0.156826913356781, "learning_rate": 4.910259181179994e-06, "loss": 0.0162, "step": 69190 }, { "grad_norm": 0.33415135741233826, "learning_rate": 4.901330897871015e-06, "loss": 0.0192, "step": 69200 }, { "grad_norm": 0.13393506407737732, "learning_rate": 4.8924103206911375e-06, "loss": 0.0119, "step": 69210 }, { "grad_norm": 0.3669587969779968, "learning_rate": 4.883497451164637e-06, "loss": 0.0232, "step": 69220 }, { "grad_norm": 0.20965056121349335, "learning_rate": 4.874592290814484e-06, "loss": 0.0155, "step": 69230 }, { "grad_norm": 0.4098762571811676, "learning_rate": 4.865694841162327e-06, "loss": 0.0263, "step": 69240 }, { "grad_norm": 0.18264317512512207, "learning_rate": 4.856805103728496e-06, "loss": 0.0175, "step": 69250 }, { "grad_norm": 0.31164857745170593, "learning_rate": 4.847923080032007e-06, "loss": 0.0119, "step": 69260 }, { "grad_norm": 0.18686561286449432, "learning_rate": 4.8390487715905486e-06, "loss": 0.0177, "step": 69270 }, { "grad_norm": 0.14136385917663574, "learning_rate": 4.8301821799205e-06, "loss": 0.0142, "step": 69280 }, { "grad_norm": 0.25265052914619446, "learning_rate": 4.821323306536918e-06, "loss": 0.0144, "step": 69290 }, { "grad_norm": 0.31285566091537476, "learning_rate": 4.8124721529535455e-06, "loss": 0.0141, "step": 69300 }, { "grad_norm": 0.4428972601890564, "learning_rate": 4.803628720682807e-06, "loss": 0.0246, "step": 69310 }, { "grad_norm": 0.267375111579895, "learning_rate": 4.794793011235776e-06, "loss": 0.0195, "step": 69320 }, { "grad_norm": 0.2826465666294098, "learning_rate": 4.78596502612228e-06, "loss": 0.0206, "step": 69330 }, { "grad_norm": 0.24968598783016205, "learning_rate": 4.777144766850738e-06, "loss": 0.0208, "step": 69340 }, { "grad_norm": 0.528527021408081, "learning_rate": 4.768332234928313e-06, "loss": 0.0197, "step": 69350 }, { "grad_norm": 0.16437751054763794, "learning_rate": 4.759527431860828e-06, "loss": 0.0138, "step": 69360 }, { "grad_norm": 0.39765501022338867, "learning_rate": 4.750730359152755e-06, "loss": 0.0245, "step": 69370 }, { "grad_norm": 0.217641219496727, "learning_rate": 4.741941018307311e-06, "loss": 0.0168, "step": 69380 }, { "grad_norm": 0.4476323425769806, "learning_rate": 4.733159410826321e-06, "loss": 0.019, "step": 69390 }, { "grad_norm": 0.27291393280029297, "learning_rate": 4.724385538210357e-06, "loss": 0.0165, "step": 69400 }, { "grad_norm": 0.2537993788719177, "learning_rate": 4.715619401958599e-06, "loss": 0.0176, "step": 69410 }, { "grad_norm": 0.27312201261520386, "learning_rate": 4.706861003568958e-06, "loss": 0.0206, "step": 69420 }, { "grad_norm": 0.27349212765693665, "learning_rate": 4.698110344538003e-06, "loss": 0.0175, "step": 69430 }, { "grad_norm": 0.3771287202835083, "learning_rate": 4.689367426360975e-06, "loss": 0.0192, "step": 69440 }, { "grad_norm": 0.24315345287322998, "learning_rate": 4.680632250531819e-06, "loss": 0.0182, "step": 69450 }, { "grad_norm": 0.2745179533958435, "learning_rate": 4.671904818543115e-06, "loss": 0.0163, "step": 69460 }, { "grad_norm": 0.3607577085494995, "learning_rate": 4.66318513188615e-06, "loss": 0.0167, "step": 69470 }, { "grad_norm": 0.2288270890712738, "learning_rate": 4.654473192050884e-06, "loss": 0.0146, "step": 69480 }, { "grad_norm": 0.19381897151470184, "learning_rate": 4.64576900052594e-06, "loss": 0.0165, "step": 69490 }, { "grad_norm": 0.3958435654640198, "learning_rate": 4.637072558798638e-06, "loss": 0.0152, "step": 69500 }, { "grad_norm": 0.37853920459747314, "learning_rate": 4.628383868354969e-06, "loss": 0.0234, "step": 69510 }, { "grad_norm": 0.2204560786485672, "learning_rate": 4.6197029306795595e-06, "loss": 0.0141, "step": 69520 }, { "grad_norm": 0.2703377902507782, "learning_rate": 4.611029747255779e-06, "loss": 0.0164, "step": 69530 }, { "grad_norm": 0.4189227819442749, "learning_rate": 4.6023643195656164e-06, "loss": 0.0194, "step": 69540 }, { "grad_norm": 0.22198137640953064, "learning_rate": 4.593706649089768e-06, "loss": 0.015, "step": 69550 }, { "grad_norm": 0.3844919800758362, "learning_rate": 4.585056737307597e-06, "loss": 0.0167, "step": 69560 }, { "grad_norm": 0.3644072711467743, "learning_rate": 4.576414585697103e-06, "loss": 0.0119, "step": 69570 }, { "grad_norm": 0.46726784110069275, "learning_rate": 4.567780195735044e-06, "loss": 0.0232, "step": 69580 }, { "grad_norm": 0.2719601094722748, "learning_rate": 4.559153568896757e-06, "loss": 0.0138, "step": 69590 }, { "grad_norm": 0.37156355381011963, "learning_rate": 4.550534706656329e-06, "loss": 0.0181, "step": 69600 }, { "grad_norm": 0.11361686885356903, "learning_rate": 4.541923610486465e-06, "loss": 0.0174, "step": 69610 }, { "grad_norm": 0.37627944350242615, "learning_rate": 4.533320281858578e-06, "loss": 0.0186, "step": 69620 }, { "grad_norm": 0.19618721306324005, "learning_rate": 4.52472472224274e-06, "loss": 0.0205, "step": 69630 }, { "grad_norm": 0.28378987312316895, "learning_rate": 4.51613693310769e-06, "loss": 0.02, "step": 69640 }, { "grad_norm": 0.2637694180011749, "learning_rate": 4.507556915920868e-06, "loss": 0.0233, "step": 69650 }, { "grad_norm": 0.2542993128299713, "learning_rate": 4.498984672148332e-06, "loss": 0.0144, "step": 69660 }, { "grad_norm": 0.23326537013053894, "learning_rate": 4.490420203254864e-06, "loss": 0.019, "step": 69670 }, { "grad_norm": 0.4015498161315918, "learning_rate": 4.481863510703893e-06, "loss": 0.0237, "step": 69680 }, { "grad_norm": 0.1681022346019745, "learning_rate": 4.473314595957523e-06, "loss": 0.0214, "step": 69690 }, { "grad_norm": 0.4018820822238922, "learning_rate": 4.464773460476535e-06, "loss": 0.021, "step": 69700 }, { "grad_norm": 0.22786679863929749, "learning_rate": 4.456240105720372e-06, "loss": 0.0134, "step": 69710 }, { "grad_norm": 0.1913766711950302, "learning_rate": 4.4477145331471405e-06, "loss": 0.0299, "step": 69720 }, { "grad_norm": 0.25137534737586975, "learning_rate": 4.439196744213653e-06, "loss": 0.0191, "step": 69730 }, { "grad_norm": 0.31517869234085083, "learning_rate": 4.430686740375339e-06, "loss": 0.0181, "step": 69740 }, { "grad_norm": 0.2685234248638153, "learning_rate": 4.422184523086342e-06, "loss": 0.0165, "step": 69750 }, { "grad_norm": 0.20259754359722137, "learning_rate": 4.41369009379946e-06, "loss": 0.0129, "step": 69760 }, { "grad_norm": 0.2527030110359192, "learning_rate": 4.405203453966139e-06, "loss": 0.014, "step": 69770 }, { "grad_norm": 0.4077662527561188, "learning_rate": 4.396724605036539e-06, "loss": 0.0197, "step": 69780 }, { "grad_norm": 0.27667859196662903, "learning_rate": 4.388253548459437e-06, "loss": 0.0186, "step": 69790 }, { "grad_norm": 0.2230650633573532, "learning_rate": 4.3797902856823395e-06, "loss": 0.0148, "step": 69800 }, { "grad_norm": 0.20265966653823853, "learning_rate": 4.371334818151357e-06, "loss": 0.013, "step": 69810 }, { "grad_norm": 0.22219780087471008, "learning_rate": 4.362887147311306e-06, "loss": 0.0156, "step": 69820 }, { "grad_norm": 0.20594094693660736, "learning_rate": 4.354447274605672e-06, "loss": 0.0144, "step": 69830 }, { "grad_norm": 0.34561607241630554, "learning_rate": 4.346015201476572e-06, "loss": 0.0226, "step": 69840 }, { "grad_norm": 0.18973124027252197, "learning_rate": 4.337590929364855e-06, "loss": 0.0152, "step": 69850 }, { "grad_norm": 0.17831803858280182, "learning_rate": 4.329174459709973e-06, "loss": 0.0134, "step": 69860 }, { "grad_norm": 0.23230519890785217, "learning_rate": 4.320765793950071e-06, "loss": 0.0167, "step": 69870 }, { "grad_norm": 0.3707294762134552, "learning_rate": 4.312364933521962e-06, "loss": 0.0187, "step": 69880 }, { "grad_norm": 0.2886568605899811, "learning_rate": 4.303971879861129e-06, "loss": 0.0171, "step": 69890 }, { "grad_norm": 0.24801118671894073, "learning_rate": 4.295586634401716e-06, "loss": 0.0219, "step": 69900 }, { "grad_norm": 0.18441201746463776, "learning_rate": 4.287209198576536e-06, "loss": 0.0182, "step": 69910 }, { "grad_norm": 0.2518272399902344, "learning_rate": 4.278839573817045e-06, "loss": 0.018, "step": 69920 }, { "grad_norm": 0.4404922127723694, "learning_rate": 4.270477761553399e-06, "loss": 0.0209, "step": 69930 }, { "grad_norm": 0.3421786427497864, "learning_rate": 4.262123763214393e-06, "loss": 0.0149, "step": 69940 }, { "grad_norm": 0.2264496386051178, "learning_rate": 4.253777580227508e-06, "loss": 0.0161, "step": 69950 }, { "grad_norm": 0.2549991309642792, "learning_rate": 4.2454392140188755e-06, "loss": 0.0172, "step": 69960 }, { "grad_norm": 0.12201916426420212, "learning_rate": 4.2371086660132785e-06, "loss": 0.0194, "step": 69970 }, { "grad_norm": 0.20551122725009918, "learning_rate": 4.228785937634205e-06, "loss": 0.0128, "step": 69980 }, { "grad_norm": 0.1995553821325302, "learning_rate": 4.220471030303758e-06, "loss": 0.0107, "step": 69990 }, { "grad_norm": 0.25404930114746094, "learning_rate": 4.212163945442754e-06, "loss": 0.0124, "step": 70000 }, { "grad_norm": 0.22559644281864166, "learning_rate": 4.203864684470621e-06, "loss": 0.017, "step": 70010 }, { "grad_norm": 0.30904555320739746, "learning_rate": 4.19557324880549e-06, "loss": 0.0156, "step": 70020 }, { "grad_norm": 0.25021710991859436, "learning_rate": 4.187289639864145e-06, "loss": 0.0105, "step": 70030 }, { "grad_norm": 0.1469952017068863, "learning_rate": 4.1790138590619974e-06, "loss": 0.014, "step": 70040 }, { "grad_norm": 0.3118102550506592, "learning_rate": 4.170745907813195e-06, "loss": 0.0216, "step": 70050 }, { "grad_norm": 0.18755130469799042, "learning_rate": 4.162485787530479e-06, "loss": 0.0206, "step": 70060 }, { "grad_norm": 0.20320157706737518, "learning_rate": 4.154233499625282e-06, "loss": 0.0164, "step": 70070 }, { "grad_norm": 0.13700035214424133, "learning_rate": 4.145989045507692e-06, "loss": 0.0146, "step": 70080 }, { "grad_norm": 0.2177557647228241, "learning_rate": 4.1377524265864666e-06, "loss": 0.0186, "step": 70090 }, { "grad_norm": 0.3259153664112091, "learning_rate": 4.1295236442690175e-06, "loss": 0.0148, "step": 70100 }, { "grad_norm": 0.3559451997280121, "learning_rate": 4.121302699961421e-06, "loss": 0.0167, "step": 70110 }, { "grad_norm": 0.3509526550769806, "learning_rate": 4.113089595068403e-06, "loss": 0.0178, "step": 70120 }, { "grad_norm": 0.26099666953086853, "learning_rate": 4.104884330993364e-06, "loss": 0.0176, "step": 70130 }, { "grad_norm": 0.15837900340557098, "learning_rate": 4.0966869091383585e-06, "loss": 0.0256, "step": 70140 }, { "grad_norm": 0.2607680857181549, "learning_rate": 4.088497330904101e-06, "loss": 0.0176, "step": 70150 }, { "grad_norm": 0.3337295949459076, "learning_rate": 4.080315597689976e-06, "loss": 0.0153, "step": 70160 }, { "grad_norm": 0.17792163789272308, "learning_rate": 4.072141710893995e-06, "loss": 0.0242, "step": 70170 }, { "grad_norm": 0.3386412262916565, "learning_rate": 4.063975671912879e-06, "loss": 0.0219, "step": 70180 }, { "grad_norm": 0.3205508887767792, "learning_rate": 4.055817482141949e-06, "loss": 0.0116, "step": 70190 }, { "grad_norm": 0.2525264322757721, "learning_rate": 4.047667142975259e-06, "loss": 0.0159, "step": 70200 }, { "grad_norm": 0.15101179480552673, "learning_rate": 4.039524655805443e-06, "loss": 0.0182, "step": 70210 }, { "grad_norm": 0.276082307100296, "learning_rate": 4.03139002202384e-06, "loss": 0.0116, "step": 70220 }, { "grad_norm": 0.2791575491428375, "learning_rate": 4.023263243020447e-06, "loss": 0.0173, "step": 70230 }, { "grad_norm": 0.3666582703590393, "learning_rate": 4.015144320183884e-06, "loss": 0.0203, "step": 70240 }, { "grad_norm": 0.2271096259355545, "learning_rate": 4.007033254901482e-06, "loss": 0.0137, "step": 70250 }, { "grad_norm": 0.336552232503891, "learning_rate": 3.9989300485591795e-06, "loss": 0.0163, "step": 70260 }, { "grad_norm": 0.3106187582015991, "learning_rate": 3.990834702541601e-06, "loss": 0.021, "step": 70270 }, { "grad_norm": 0.22099782526493073, "learning_rate": 3.982747218232019e-06, "loss": 0.0137, "step": 70280 }, { "grad_norm": 0.21986839175224304, "learning_rate": 3.97466759701236e-06, "loss": 0.02, "step": 70290 }, { "grad_norm": 0.5258110761642456, "learning_rate": 3.966595840263226e-06, "loss": 0.0178, "step": 70300 }, { "grad_norm": 0.32134610414505005, "learning_rate": 3.958531949363836e-06, "loss": 0.0149, "step": 70310 }, { "grad_norm": 0.2713308036327362, "learning_rate": 3.950475925692098e-06, "loss": 0.0216, "step": 70320 }, { "grad_norm": 0.3124409019947052, "learning_rate": 3.9424277706245685e-06, "loss": 0.0149, "step": 70330 }, { "grad_norm": 0.2541128396987915, "learning_rate": 3.934387485536451e-06, "loss": 0.0154, "step": 70340 }, { "grad_norm": 0.19547845423221588, "learning_rate": 3.926355071801619e-06, "loss": 0.0155, "step": 70350 }, { "grad_norm": 0.2140783667564392, "learning_rate": 3.9183305307925965e-06, "loss": 0.0139, "step": 70360 }, { "grad_norm": 0.37067145109176636, "learning_rate": 3.910313863880533e-06, "loss": 0.0136, "step": 70370 }, { "grad_norm": 0.18785227835178375, "learning_rate": 3.902305072435292e-06, "loss": 0.0134, "step": 70380 }, { "grad_norm": 0.37499019503593445, "learning_rate": 3.894304157825329e-06, "loss": 0.0171, "step": 70390 }, { "grad_norm": 0.18704582750797272, "learning_rate": 3.886311121417791e-06, "loss": 0.0123, "step": 70400 }, { "grad_norm": 0.13269509375095367, "learning_rate": 3.878325964578472e-06, "loss": 0.0131, "step": 70410 }, { "grad_norm": 0.16692763566970825, "learning_rate": 3.870348688671815e-06, "loss": 0.011, "step": 70420 }, { "grad_norm": 0.3648436367511749, "learning_rate": 3.862379295060931e-06, "loss": 0.0123, "step": 70430 }, { "grad_norm": 0.2919115126132965, "learning_rate": 3.854417785107534e-06, "loss": 0.0228, "step": 70440 }, { "grad_norm": 0.13319051265716553, "learning_rate": 3.8464641601720755e-06, "loss": 0.0158, "step": 70450 }, { "grad_norm": 0.30338820815086365, "learning_rate": 3.838518421613579e-06, "loss": 0.0145, "step": 70460 }, { "grad_norm": 0.2614433467388153, "learning_rate": 3.830580570789766e-06, "loss": 0.0169, "step": 70470 }, { "grad_norm": 0.16016419231891632, "learning_rate": 3.822650609057005e-06, "loss": 0.0128, "step": 70480 }, { "grad_norm": 0.23332960903644562, "learning_rate": 3.814728537770285e-06, "loss": 0.0154, "step": 70490 }, { "grad_norm": 0.24422229826450348, "learning_rate": 3.806814358283306e-06, "loss": 0.017, "step": 70500 }, { "grad_norm": 0.422802209854126, "learning_rate": 3.7989080719483596e-06, "loss": 0.0192, "step": 70510 }, { "grad_norm": 0.13451500236988068, "learning_rate": 3.7910096801164143e-06, "loss": 0.0173, "step": 70520 }, { "grad_norm": 0.17856840789318085, "learning_rate": 3.7831191841371016e-06, "loss": 0.0184, "step": 70530 }, { "grad_norm": 0.3258407413959503, "learning_rate": 3.775236585358688e-06, "loss": 0.0198, "step": 70540 }, { "grad_norm": 0.28731825947761536, "learning_rate": 3.7673618851280843e-06, "loss": 0.019, "step": 70550 }, { "grad_norm": 0.35122063755989075, "learning_rate": 3.759495084790887e-06, "loss": 0.0185, "step": 70560 }, { "grad_norm": 0.3745070695877075, "learning_rate": 3.751636185691282e-06, "loss": 0.0142, "step": 70570 }, { "grad_norm": 0.24836598336696625, "learning_rate": 3.7437851891721607e-06, "loss": 0.017, "step": 70580 }, { "grad_norm": 0.2767041027545929, "learning_rate": 3.7359420965750404e-06, "loss": 0.0176, "step": 70590 }, { "grad_norm": 0.25151848793029785, "learning_rate": 3.7281069092400922e-06, "loss": 0.0148, "step": 70600 }, { "grad_norm": 0.27988675236701965, "learning_rate": 3.7202796285061348e-06, "loss": 0.0154, "step": 70610 }, { "grad_norm": 0.3623696267604828, "learning_rate": 3.712460255710637e-06, "loss": 0.0181, "step": 70620 }, { "grad_norm": 0.2210632860660553, "learning_rate": 3.704648792189719e-06, "loss": 0.0137, "step": 70630 }, { "grad_norm": 0.26006704568862915, "learning_rate": 3.696845239278124e-06, "loss": 0.0174, "step": 70640 }, { "grad_norm": 0.3150807023048401, "learning_rate": 3.689049598309302e-06, "loss": 0.0178, "step": 70650 }, { "grad_norm": 0.2525748014450073, "learning_rate": 3.681261870615288e-06, "loss": 0.0262, "step": 70660 }, { "grad_norm": 0.09699316322803497, "learning_rate": 3.6734820575268004e-06, "loss": 0.024, "step": 70670 }, { "grad_norm": 0.1254376620054245, "learning_rate": 3.665710160373209e-06, "loss": 0.0145, "step": 70680 }, { "grad_norm": 0.17478634417057037, "learning_rate": 3.65794618048248e-06, "loss": 0.0125, "step": 70690 }, { "grad_norm": 0.15530088543891907, "learning_rate": 3.6501901191813125e-06, "loss": 0.0132, "step": 70700 }, { "grad_norm": 0.3408016562461853, "learning_rate": 3.642441977794975e-06, "loss": 0.0192, "step": 70710 }, { "grad_norm": 0.4539959132671356, "learning_rate": 3.634701757647424e-06, "loss": 0.0157, "step": 70720 }, { "grad_norm": 0.23352104425430298, "learning_rate": 3.6269694600612468e-06, "loss": 0.0147, "step": 70730 }, { "grad_norm": 0.19591158628463745, "learning_rate": 3.619245086357681e-06, "loss": 0.0173, "step": 70740 }, { "grad_norm": 0.4346160888671875, "learning_rate": 3.611528637856615e-06, "loss": 0.0165, "step": 70750 }, { "grad_norm": 0.23303978145122528, "learning_rate": 3.6038201158765884e-06, "loss": 0.0139, "step": 70760 }, { "grad_norm": 0.5572485327720642, "learning_rate": 3.5961195217347534e-06, "loss": 0.0224, "step": 70770 }, { "grad_norm": 0.2027047574520111, "learning_rate": 3.588426856746946e-06, "loss": 0.0156, "step": 70780 }, { "grad_norm": 0.2868638038635254, "learning_rate": 3.5807421222276316e-06, "loss": 0.0248, "step": 70790 }, { "grad_norm": 0.2450616955757141, "learning_rate": 3.5730653194899156e-06, "loss": 0.0177, "step": 70800 }, { "grad_norm": 0.35576027631759644, "learning_rate": 3.565396449845554e-06, "loss": 0.0161, "step": 70810 }, { "grad_norm": 0.34719187021255493, "learning_rate": 3.557735514604954e-06, "loss": 0.0173, "step": 70820 }, { "grad_norm": 0.129586860537529, "learning_rate": 3.5500825150771633e-06, "loss": 0.0144, "step": 70830 }, { "grad_norm": 0.23509356379508972, "learning_rate": 3.5424374525698466e-06, "loss": 0.0179, "step": 70840 }, { "grad_norm": 0.18331590294837952, "learning_rate": 3.5348003283893704e-06, "loss": 0.0138, "step": 70850 }, { "grad_norm": 0.19558824598789215, "learning_rate": 3.527171143840685e-06, "loss": 0.0173, "step": 70860 }, { "grad_norm": 0.13859549164772034, "learning_rate": 3.519549900227409e-06, "loss": 0.0198, "step": 70870 }, { "grad_norm": 0.3546936810016632, "learning_rate": 3.511936598851828e-06, "loss": 0.0161, "step": 70880 }, { "grad_norm": 0.27748820185661316, "learning_rate": 3.504331241014813e-06, "loss": 0.0163, "step": 70890 }, { "grad_norm": 0.27345943450927734, "learning_rate": 3.4967338280159414e-06, "loss": 0.013, "step": 70900 }, { "grad_norm": 0.21899572014808655, "learning_rate": 3.4891443611533846e-06, "loss": 0.0214, "step": 70910 }, { "grad_norm": 0.20855437219142914, "learning_rate": 3.481562841723984e-06, "loss": 0.0224, "step": 70920 }, { "grad_norm": 0.2368185669183731, "learning_rate": 3.473989271023215e-06, "loss": 0.0177, "step": 70930 }, { "grad_norm": 0.32330086827278137, "learning_rate": 3.4664236503451854e-06, "loss": 0.0211, "step": 70940 }, { "grad_norm": 0.2537406384944916, "learning_rate": 3.4588659809826674e-06, "loss": 0.0181, "step": 70950 }, { "grad_norm": 0.2654414772987366, "learning_rate": 3.451316264227039e-06, "loss": 0.0235, "step": 70960 }, { "grad_norm": 0.4380921423435211, "learning_rate": 3.4437745013683574e-06, "loss": 0.0156, "step": 70970 }, { "grad_norm": 0.2524336576461792, "learning_rate": 3.4362406936952916e-06, "loss": 0.0153, "step": 70980 }, { "grad_norm": 0.3128397762775421, "learning_rate": 3.4287148424951786e-06, "loss": 0.017, "step": 70990 }, { "grad_norm": 0.33494308590888977, "learning_rate": 3.4211969490539617e-06, "loss": 0.0159, "step": 71000 }, { "grad_norm": 0.17923195660114288, "learning_rate": 3.4136870146562584e-06, "loss": 0.0161, "step": 71010 }, { "grad_norm": 0.2540709376335144, "learning_rate": 3.406185040585308e-06, "loss": 0.0159, "step": 71020 }, { "grad_norm": 0.29048484563827515, "learning_rate": 3.3986910281229966e-06, "loss": 0.0167, "step": 71030 }, { "grad_norm": 0.24701586365699768, "learning_rate": 3.391204978549828e-06, "loss": 0.0174, "step": 71040 }, { "grad_norm": 0.2759680151939392, "learning_rate": 3.3837268931449785e-06, "loss": 0.0248, "step": 71050 }, { "grad_norm": 0.36407122015953064, "learning_rate": 3.376256773186248e-06, "loss": 0.0206, "step": 71060 }, { "grad_norm": 0.30813899636268616, "learning_rate": 3.3687946199500664e-06, "loss": 0.0198, "step": 71070 }, { "grad_norm": 0.3142029643058777, "learning_rate": 3.3613404347115295e-06, "loss": 0.0172, "step": 71080 }, { "grad_norm": 0.14473587274551392, "learning_rate": 3.35389421874433e-06, "loss": 0.0128, "step": 71090 }, { "grad_norm": 0.12482167035341263, "learning_rate": 3.3464559733208446e-06, "loss": 0.0156, "step": 71100 }, { "grad_norm": 0.28623175621032715, "learning_rate": 3.3390256997120505e-06, "loss": 0.0213, "step": 71110 }, { "grad_norm": 0.30327939987182617, "learning_rate": 3.331603399187583e-06, "loss": 0.0203, "step": 71120 }, { "grad_norm": 0.19538676738739014, "learning_rate": 3.324189073015721e-06, "loss": 0.022, "step": 71130 }, { "grad_norm": 0.3055110275745392, "learning_rate": 3.3167827224633465e-06, "loss": 0.0153, "step": 71140 }, { "grad_norm": 0.21767857670783997, "learning_rate": 3.309384348796024e-06, "loss": 0.0135, "step": 71150 }, { "grad_norm": 0.19454139471054077, "learning_rate": 3.301993953277921e-06, "loss": 0.0161, "step": 71160 }, { "grad_norm": 0.3314671516418457, "learning_rate": 3.29461153717186e-06, "loss": 0.0144, "step": 71170 }, { "grad_norm": 0.23286540806293488, "learning_rate": 3.287237101739293e-06, "loss": 0.0175, "step": 71180 }, { "grad_norm": 0.29772982001304626, "learning_rate": 3.2798706482403075e-06, "loss": 0.0212, "step": 71190 }, { "grad_norm": 0.19591844081878662, "learning_rate": 3.2725121779336285e-06, "loss": 0.0179, "step": 71200 }, { "grad_norm": 0.29496875405311584, "learning_rate": 3.265161692076618e-06, "loss": 0.0173, "step": 71210 }, { "grad_norm": 0.13809329271316528, "learning_rate": 3.257819191925282e-06, "loss": 0.0119, "step": 71220 }, { "grad_norm": 0.2603239119052887, "learning_rate": 3.2504846787342392e-06, "loss": 0.0157, "step": 71230 }, { "grad_norm": 0.23766623437404633, "learning_rate": 3.243158153756759e-06, "loss": 0.02, "step": 71240 }, { "grad_norm": 0.20584367215633392, "learning_rate": 3.2358396182447514e-06, "loss": 0.0148, "step": 71250 }, { "grad_norm": 0.16233889758586884, "learning_rate": 3.2285290734487496e-06, "loss": 0.0134, "step": 71260 }, { "grad_norm": 0.290155291557312, "learning_rate": 3.221226520617926e-06, "loss": 0.0169, "step": 71270 }, { "grad_norm": 0.244461789727211, "learning_rate": 3.213931961000094e-06, "loss": 0.0196, "step": 71280 }, { "grad_norm": 0.23338976502418518, "learning_rate": 3.2066453958416733e-06, "loss": 0.0156, "step": 71290 }, { "grad_norm": 0.35382992029190063, "learning_rate": 3.199366826387773e-06, "loss": 0.0182, "step": 71300 }, { "grad_norm": 0.21547752618789673, "learning_rate": 3.192096253882071e-06, "loss": 0.0145, "step": 71310 }, { "grad_norm": 0.237313911318779, "learning_rate": 3.1848336795669177e-06, "loss": 0.0144, "step": 71320 }, { "grad_norm": 0.3031441271305084, "learning_rate": 3.1775791046833035e-06, "loss": 0.0208, "step": 71330 }, { "grad_norm": 0.3000488877296448, "learning_rate": 3.170332530470804e-06, "loss": 0.0148, "step": 71340 }, { "grad_norm": 0.27963343262672424, "learning_rate": 3.1630939581677012e-06, "loss": 0.0181, "step": 71350 }, { "grad_norm": 0.1798340529203415, "learning_rate": 3.155863389010838e-06, "loss": 0.0161, "step": 71360 }, { "grad_norm": 0.30227524042129517, "learning_rate": 3.1486408242357323e-06, "loss": 0.0199, "step": 71370 }, { "grad_norm": 0.4203377366065979, "learning_rate": 3.1414262650765248e-06, "loss": 0.0347, "step": 71380 }, { "grad_norm": 0.37993693351745605, "learning_rate": 3.134219712765979e-06, "loss": 0.0189, "step": 71390 }, { "grad_norm": 0.38353031873703003, "learning_rate": 3.1270211685355044e-06, "loss": 0.0192, "step": 71400 }, { "grad_norm": 0.17575247585773468, "learning_rate": 3.1198306336151338e-06, "loss": 0.0189, "step": 71410 }, { "grad_norm": 0.2285991758108139, "learning_rate": 3.11264810923354e-06, "loss": 0.0176, "step": 71420 }, { "grad_norm": 0.11586761474609375, "learning_rate": 3.105473596618008e-06, "loss": 0.013, "step": 71430 }, { "grad_norm": 0.2569783926010132, "learning_rate": 3.0983070969944683e-06, "loss": 0.0164, "step": 71440 }, { "grad_norm": 0.20340655744075775, "learning_rate": 3.09114861158748e-06, "loss": 0.0163, "step": 71450 }, { "grad_norm": 0.3400299847126007, "learning_rate": 3.0839981416202314e-06, "loss": 0.0177, "step": 71460 }, { "grad_norm": 0.30596593022346497, "learning_rate": 3.0768556883145507e-06, "loss": 0.0176, "step": 71470 }, { "grad_norm": 0.2678568661212921, "learning_rate": 3.0697212528908834e-06, "loss": 0.0155, "step": 71480 }, { "grad_norm": 0.26887044310569763, "learning_rate": 3.0625948365683e-06, "loss": 0.0165, "step": 71490 }, { "grad_norm": 0.28849083185195923, "learning_rate": 3.055476440564531e-06, "loss": 0.0188, "step": 71500 }, { "grad_norm": 0.16828663647174835, "learning_rate": 3.0483660660958924e-06, "loss": 0.0161, "step": 71510 }, { "grad_norm": 0.3142358660697937, "learning_rate": 3.0412637143773624e-06, "loss": 0.0213, "step": 71520 }, { "grad_norm": 0.30755332112312317, "learning_rate": 3.034169386622554e-06, "loss": 0.0152, "step": 71530 }, { "grad_norm": 0.2750614285469055, "learning_rate": 3.027083084043658e-06, "loss": 0.014, "step": 71540 }, { "grad_norm": 0.21752934157848358, "learning_rate": 3.0200048078515676e-06, "loss": 0.0176, "step": 71550 }, { "grad_norm": 0.268208384513855, "learning_rate": 3.012934559255742e-06, "loss": 0.0165, "step": 71560 }, { "grad_norm": 0.32569777965545654, "learning_rate": 3.0058723394642994e-06, "loss": 0.0183, "step": 71570 }, { "grad_norm": 0.2742486000061035, "learning_rate": 2.998818149683985e-06, "loss": 0.0256, "step": 71580 }, { "grad_norm": 0.21306689083576202, "learning_rate": 2.9917719911201627e-06, "loss": 0.0164, "step": 71590 }, { "grad_norm": 0.24539946019649506, "learning_rate": 2.9847338649768352e-06, "loss": 0.0139, "step": 71600 }, { "grad_norm": 0.3309750556945801, "learning_rate": 2.977703772456608e-06, "loss": 0.0134, "step": 71610 }, { "grad_norm": 0.23175808787345886, "learning_rate": 2.9706817147607535e-06, "loss": 0.0154, "step": 71620 }, { "grad_norm": 0.30368876457214355, "learning_rate": 2.9636676930891393e-06, "loss": 0.0149, "step": 71630 }, { "grad_norm": 0.2816413342952728, "learning_rate": 2.9566617086402625e-06, "loss": 0.0153, "step": 71640 }, { "grad_norm": 0.24059081077575684, "learning_rate": 2.9496637626112655e-06, "loss": 0.0146, "step": 71650 }, { "grad_norm": 0.3186641335487366, "learning_rate": 2.9426738561979027e-06, "loss": 0.0162, "step": 71660 }, { "grad_norm": 0.3691211938858032, "learning_rate": 2.9356919905945524e-06, "loss": 0.0355, "step": 71670 }, { "grad_norm": 0.2381126880645752, "learning_rate": 2.928718166994243e-06, "loss": 0.0172, "step": 71680 }, { "grad_norm": 0.3149474263191223, "learning_rate": 2.9217523865885833e-06, "loss": 0.0214, "step": 71690 }, { "grad_norm": 0.25325140357017517, "learning_rate": 2.9147946505678537e-06, "loss": 0.015, "step": 71700 }, { "grad_norm": 0.17644795775413513, "learning_rate": 2.9078449601209313e-06, "loss": 0.0146, "step": 71710 }, { "grad_norm": 0.2678646445274353, "learning_rate": 2.900903316435333e-06, "loss": 0.016, "step": 71720 }, { "grad_norm": 0.28465303778648376, "learning_rate": 2.8939697206971983e-06, "loss": 0.0215, "step": 71730 }, { "grad_norm": 0.342885822057724, "learning_rate": 2.8870441740912746e-06, "loss": 0.0139, "step": 71740 }, { "grad_norm": 0.2334807962179184, "learning_rate": 2.8801266778009762e-06, "loss": 0.0157, "step": 71750 }, { "grad_norm": 0.26986998319625854, "learning_rate": 2.8732172330082798e-06, "loss": 0.0217, "step": 71760 }, { "grad_norm": 0.2425841987133026, "learning_rate": 2.8663158408938517e-06, "loss": 0.0172, "step": 71770 }, { "grad_norm": 0.2168741226196289, "learning_rate": 2.859422502636938e-06, "loss": 0.0143, "step": 71780 }, { "grad_norm": 0.2251550704240799, "learning_rate": 2.8525372194154076e-06, "loss": 0.0229, "step": 71790 }, { "grad_norm": 0.3095115125179291, "learning_rate": 2.8456599924057913e-06, "loss": 0.0147, "step": 71800 }, { "grad_norm": 0.2530849277973175, "learning_rate": 2.8387908227831995e-06, "loss": 0.0211, "step": 71810 }, { "grad_norm": 0.2062912881374359, "learning_rate": 2.831929711721404e-06, "loss": 0.0171, "step": 71820 }, { "grad_norm": 0.20997662842273712, "learning_rate": 2.825076660392767e-06, "loss": 0.014, "step": 71830 }, { "grad_norm": 0.1949143409729004, "learning_rate": 2.8182316699682908e-06, "loss": 0.0186, "step": 71840 }, { "grad_norm": 0.2376009225845337, "learning_rate": 2.811394741617601e-06, "loss": 0.0145, "step": 71850 }, { "grad_norm": 0.3747197091579437, "learning_rate": 2.8045658765089356e-06, "loss": 0.0187, "step": 71860 }, { "grad_norm": 0.2666197121143341, "learning_rate": 2.7977450758091605e-06, "loss": 0.0262, "step": 71870 }, { "grad_norm": 0.2377639263868332, "learning_rate": 2.790932340683783e-06, "loss": 0.0194, "step": 71880 }, { "grad_norm": 0.19261103868484497, "learning_rate": 2.7841276722968823e-06, "loss": 0.0177, "step": 71890 }, { "grad_norm": 0.20585989952087402, "learning_rate": 2.7773310718112067e-06, "loss": 0.0128, "step": 71900 }, { "grad_norm": 0.3924744427204132, "learning_rate": 2.7705425403881102e-06, "loss": 0.0193, "step": 71910 }, { "grad_norm": 0.25684794783592224, "learning_rate": 2.763762079187565e-06, "loss": 0.0136, "step": 71920 }, { "grad_norm": 0.13918624818325043, "learning_rate": 2.756989689368178e-06, "loss": 0.0168, "step": 71930 }, { "grad_norm": 0.27037665247917175, "learning_rate": 2.750225372087134e-06, "loss": 0.0184, "step": 71940 }, { "grad_norm": 0.1668097972869873, "learning_rate": 2.7434691285003033e-06, "loss": 0.0158, "step": 71950 }, { "grad_norm": 0.19584369659423828, "learning_rate": 2.7367209597621178e-06, "loss": 0.0175, "step": 71960 }, { "grad_norm": 0.17972514033317566, "learning_rate": 2.729980867025683e-06, "loss": 0.0158, "step": 71970 }, { "grad_norm": 0.31404992938041687, "learning_rate": 2.7232488514426724e-06, "loss": 0.0176, "step": 71980 }, { "grad_norm": 0.21571987867355347, "learning_rate": 2.7165249141634097e-06, "loss": 0.0158, "step": 71990 }, { "grad_norm": 0.36925333738327026, "learning_rate": 2.709809056336837e-06, "loss": 0.0225, "step": 72000 }, { "grad_norm": 0.1750541627407074, "learning_rate": 2.703101279110498e-06, "loss": 0.016, "step": 72010 }, { "grad_norm": 0.3668135106563568, "learning_rate": 2.696401583630587e-06, "loss": 0.0215, "step": 72020 }, { "grad_norm": 0.26068583130836487, "learning_rate": 2.689709971041887e-06, "loss": 0.0119, "step": 72030 }, { "grad_norm": 0.283792108297348, "learning_rate": 2.6830264424878126e-06, "loss": 0.0164, "step": 72040 }, { "grad_norm": 0.3920533061027527, "learning_rate": 2.676350999110394e-06, "loss": 0.0283, "step": 72050 }, { "grad_norm": 0.24248728156089783, "learning_rate": 2.6696836420502856e-06, "loss": 0.0254, "step": 72060 }, { "grad_norm": 0.2600630819797516, "learning_rate": 2.663024372446765e-06, "loss": 0.0139, "step": 72070 }, { "grad_norm": 0.2966748774051666, "learning_rate": 2.6563731914377055e-06, "loss": 0.0161, "step": 72080 }, { "grad_norm": 0.17793267965316772, "learning_rate": 2.6497301001596087e-06, "loss": 0.0206, "step": 72090 }, { "grad_norm": 0.22966621816158295, "learning_rate": 2.643095099747611e-06, "loss": 0.0189, "step": 72100 }, { "grad_norm": 0.28533613681793213, "learning_rate": 2.636468191335445e-06, "loss": 0.0189, "step": 72110 }, { "grad_norm": 0.2967464327812195, "learning_rate": 2.629849376055471e-06, "loss": 0.015, "step": 72120 }, { "grad_norm": 0.22749026119709015, "learning_rate": 2.6232386550386678e-06, "loss": 0.0159, "step": 72130 }, { "grad_norm": 0.25285133719444275, "learning_rate": 2.6166360294146097e-06, "loss": 0.0153, "step": 72140 }, { "grad_norm": 0.14524982869625092, "learning_rate": 2.6100415003115275e-06, "loss": 0.0163, "step": 72150 }, { "grad_norm": 0.1968054324388504, "learning_rate": 2.603455068856225e-06, "loss": 0.0169, "step": 72160 }, { "grad_norm": 0.2499818056821823, "learning_rate": 2.5968767361741584e-06, "loss": 0.0182, "step": 72170 }, { "grad_norm": 0.14592798054218292, "learning_rate": 2.590306503389378e-06, "loss": 0.0187, "step": 72180 }, { "grad_norm": 0.17969255149364471, "learning_rate": 2.583744371624558e-06, "loss": 0.0226, "step": 72190 }, { "grad_norm": 0.2834809124469757, "learning_rate": 2.57719034200099e-06, "loss": 0.0117, "step": 72200 }, { "grad_norm": 0.19983090460300446, "learning_rate": 2.570644415638568e-06, "loss": 0.0143, "step": 72210 }, { "grad_norm": 0.30502036213874817, "learning_rate": 2.5641065936558295e-06, "loss": 0.0205, "step": 72220 }, { "grad_norm": 0.19208848476409912, "learning_rate": 2.557576877169898e-06, "loss": 0.0248, "step": 72230 }, { "grad_norm": 0.17832589149475098, "learning_rate": 2.5510552672965205e-06, "loss": 0.0132, "step": 72240 }, { "grad_norm": 0.32833272218704224, "learning_rate": 2.544541765150077e-06, "loss": 0.015, "step": 72250 }, { "grad_norm": 0.1611303687095642, "learning_rate": 2.5380363718435165e-06, "loss": 0.0194, "step": 72260 }, { "grad_norm": 0.3042205274105072, "learning_rate": 2.5315390884884714e-06, "loss": 0.0133, "step": 72270 }, { "grad_norm": 0.2197938710451126, "learning_rate": 2.5250499161951214e-06, "loss": 0.0183, "step": 72280 }, { "grad_norm": 0.23648011684417725, "learning_rate": 2.518568856072301e-06, "loss": 0.0139, "step": 72290 }, { "grad_norm": 0.3762761056423187, "learning_rate": 2.512095909227441e-06, "loss": 0.0169, "step": 72300 }, { "grad_norm": 0.23338183760643005, "learning_rate": 2.50563107676659e-06, "loss": 0.0192, "step": 72310 }, { "grad_norm": 0.22774937748908997, "learning_rate": 2.4991743597944152e-06, "loss": 0.0144, "step": 72320 }, { "grad_norm": 0.1451081484556198, "learning_rate": 2.492725759414205e-06, "loss": 0.018, "step": 72330 }, { "grad_norm": 0.1472988873720169, "learning_rate": 2.4862852767278234e-06, "loss": 0.012, "step": 72340 }, { "grad_norm": 0.12701693177223206, "learning_rate": 2.479852912835784e-06, "loss": 0.0103, "step": 72350 }, { "grad_norm": 0.37838053703308105, "learning_rate": 2.4734286688372075e-06, "loss": 0.0192, "step": 72360 }, { "grad_norm": 0.23041251301765442, "learning_rate": 2.467012545829811e-06, "loss": 0.0254, "step": 72370 }, { "grad_norm": 0.30511224269866943, "learning_rate": 2.460604544909945e-06, "loss": 0.0167, "step": 72380 }, { "grad_norm": 0.3365471661090851, "learning_rate": 2.454204667172555e-06, "loss": 0.0124, "step": 72390 }, { "grad_norm": 0.3084985017776489, "learning_rate": 2.447812913711217e-06, "loss": 0.0148, "step": 72400 }, { "grad_norm": 0.10985590517520905, "learning_rate": 2.44142928561808e-06, "loss": 0.0166, "step": 72410 }, { "grad_norm": 0.23800240457057953, "learning_rate": 2.435053783983965e-06, "loss": 0.0149, "step": 72420 }, { "grad_norm": 0.1322605162858963, "learning_rate": 2.4286864098982453e-06, "loss": 0.0167, "step": 72430 }, { "grad_norm": 0.12549009919166565, "learning_rate": 2.4223271644489397e-06, "loss": 0.0149, "step": 72440 }, { "grad_norm": 0.16230063140392303, "learning_rate": 2.415976048722679e-06, "loss": 0.0164, "step": 72450 }, { "grad_norm": 0.18670116364955902, "learning_rate": 2.4096330638046673e-06, "loss": 0.0151, "step": 72460 }, { "grad_norm": 0.214021697640419, "learning_rate": 2.4032982107787816e-06, "loss": 0.0201, "step": 72470 }, { "grad_norm": 0.1862841099500656, "learning_rate": 2.396971490727451e-06, "loss": 0.018, "step": 72480 }, { "grad_norm": 0.2656818926334381, "learning_rate": 2.3906529047317493e-06, "loss": 0.0121, "step": 72490 }, { "grad_norm": 0.11454588174819946, "learning_rate": 2.3843424538713465e-06, "loss": 0.0133, "step": 72500 }, { "grad_norm": 0.27681055665016174, "learning_rate": 2.3780401392245298e-06, "loss": 0.0151, "step": 72510 }, { "grad_norm": 0.25188523530960083, "learning_rate": 2.3717459618681883e-06, "loss": 0.0143, "step": 72520 }, { "grad_norm": 0.20268350839614868, "learning_rate": 2.3654599228778274e-06, "loss": 0.0148, "step": 72530 }, { "grad_norm": 0.3666434586048126, "learning_rate": 2.3591820233275607e-06, "loss": 0.019, "step": 72540 }, { "grad_norm": 0.14007611572742462, "learning_rate": 2.3529122642901024e-06, "loss": 0.0205, "step": 72550 }, { "grad_norm": 0.1679840385913849, "learning_rate": 2.3466506468367845e-06, "loss": 0.0272, "step": 72560 }, { "grad_norm": 0.14700938761234283, "learning_rate": 2.3403971720375506e-06, "loss": 0.0164, "step": 72570 }, { "grad_norm": 0.287370502948761, "learning_rate": 2.334151840960952e-06, "loss": 0.0133, "step": 72580 }, { "grad_norm": 0.1488703191280365, "learning_rate": 2.3279146546741347e-06, "loss": 0.0135, "step": 72590 }, { "grad_norm": 0.2927349805831909, "learning_rate": 2.32168561424288e-06, "loss": 0.0159, "step": 72600 }, { "grad_norm": 0.23841816186904907, "learning_rate": 2.3154647207315307e-06, "loss": 0.0147, "step": 72610 }, { "grad_norm": 0.23143163323402405, "learning_rate": 2.309251975203103e-06, "loss": 0.021, "step": 72620 }, { "grad_norm": 0.2672823369503021, "learning_rate": 2.303047378719159e-06, "loss": 0.0166, "step": 72630 }, { "grad_norm": 0.28058022260665894, "learning_rate": 2.2968509323399e-06, "loss": 0.0165, "step": 72640 }, { "grad_norm": 0.2870379686355591, "learning_rate": 2.290662637124147e-06, "loss": 0.0142, "step": 72650 }, { "grad_norm": 0.1875629872083664, "learning_rate": 2.2844824941292807e-06, "loss": 0.0104, "step": 72660 }, { "grad_norm": 0.37562644481658936, "learning_rate": 2.2783105044113406e-06, "loss": 0.0236, "step": 72670 }, { "grad_norm": 0.3176143765449524, "learning_rate": 2.272146669024944e-06, "loss": 0.0145, "step": 72680 }, { "grad_norm": 0.26300591230392456, "learning_rate": 2.26599098902332e-06, "loss": 0.0141, "step": 72690 }, { "grad_norm": 0.24277450144290924, "learning_rate": 2.2598434654583113e-06, "loss": 0.0136, "step": 72700 }, { "grad_norm": 0.3413017988204956, "learning_rate": 2.253704099380355e-06, "loss": 0.0221, "step": 72710 }, { "grad_norm": 0.20800402760505676, "learning_rate": 2.247572891838512e-06, "loss": 0.0197, "step": 72720 }, { "grad_norm": 0.3284994065761566, "learning_rate": 2.241449843880422e-06, "loss": 0.0185, "step": 72730 }, { "grad_norm": 0.22409139573574066, "learning_rate": 2.235334956552354e-06, "loss": 0.017, "step": 72740 }, { "grad_norm": 0.19979684054851532, "learning_rate": 2.2292282308991775e-06, "loss": 0.0181, "step": 72750 }, { "grad_norm": 0.250469446182251, "learning_rate": 2.223129667964363e-06, "loss": 0.0152, "step": 72760 }, { "grad_norm": 0.6143090724945068, "learning_rate": 2.2170392687899834e-06, "loss": 0.0147, "step": 72770 }, { "grad_norm": 0.1703893393278122, "learning_rate": 2.210957034416733e-06, "loss": 0.0131, "step": 72780 }, { "grad_norm": 0.31827229261398315, "learning_rate": 2.2048829658838867e-06, "loss": 0.017, "step": 72790 }, { "grad_norm": 0.17131054401397705, "learning_rate": 2.1988170642293525e-06, "loss": 0.012, "step": 72800 }, { "grad_norm": 0.11154773086309433, "learning_rate": 2.1927593304896075e-06, "loss": 0.013, "step": 72810 }, { "grad_norm": 0.16534942388534546, "learning_rate": 2.1867097656997626e-06, "loss": 0.0138, "step": 72820 }, { "grad_norm": 0.21697485446929932, "learning_rate": 2.180668370893524e-06, "loss": 0.0152, "step": 72830 }, { "grad_norm": 0.2555943429470062, "learning_rate": 2.174635147103199e-06, "loss": 0.0144, "step": 72840 }, { "grad_norm": 0.32596278190612793, "learning_rate": 2.1686100953597075e-06, "loss": 0.0137, "step": 72850 }, { "grad_norm": 0.25951477885246277, "learning_rate": 2.1625932166925433e-06, "loss": 0.0173, "step": 72860 }, { "grad_norm": 0.38038408756256104, "learning_rate": 2.1565845121298556e-06, "loss": 0.0176, "step": 72870 }, { "grad_norm": 0.32105207443237305, "learning_rate": 2.150583982698351e-06, "loss": 0.0193, "step": 72880 }, { "grad_norm": 0.26015031337738037, "learning_rate": 2.144591629423359e-06, "loss": 0.0115, "step": 72890 }, { "grad_norm": 0.28850412368774414, "learning_rate": 2.138607453328817e-06, "loss": 0.0177, "step": 72900 }, { "grad_norm": 0.25008347630500793, "learning_rate": 2.132631455437234e-06, "loss": 0.0164, "step": 72910 }, { "grad_norm": 0.25304606556892395, "learning_rate": 2.1266636367697714e-06, "loss": 0.0282, "step": 72920 }, { "grad_norm": 0.27970173954963684, "learning_rate": 2.120703998346152e-06, "loss": 0.0164, "step": 72930 }, { "grad_norm": 0.25665029883384705, "learning_rate": 2.1147525411847114e-06, "loss": 0.0193, "step": 72940 }, { "grad_norm": 0.20921556651592255, "learning_rate": 2.108809266302403e-06, "loss": 0.0137, "step": 72950 }, { "grad_norm": 0.32677942514419556, "learning_rate": 2.102874174714764e-06, "loss": 0.0195, "step": 72960 }, { "grad_norm": 0.254833459854126, "learning_rate": 2.0969472674359393e-06, "loss": 0.012, "step": 72970 }, { "grad_norm": 0.29005691409111023, "learning_rate": 2.091028545478668e-06, "loss": 0.0228, "step": 72980 }, { "grad_norm": 0.20735937356948853, "learning_rate": 2.08511800985432e-06, "loss": 0.0151, "step": 72990 }, { "grad_norm": 0.31692802906036377, "learning_rate": 2.07921566157282e-06, "loss": 0.0231, "step": 73000 }, { "grad_norm": 0.2964298725128174, "learning_rate": 2.073321501642728e-06, "loss": 0.0132, "step": 73010 }, { "grad_norm": 0.20011624693870544, "learning_rate": 2.0674355310711937e-06, "loss": 0.0147, "step": 73020 }, { "grad_norm": 0.23412951827049255, "learning_rate": 2.0615577508639682e-06, "loss": 0.0129, "step": 73030 }, { "grad_norm": 0.19831815361976624, "learning_rate": 2.055688162025399e-06, "loss": 0.0142, "step": 73040 }, { "grad_norm": 0.14089830219745636, "learning_rate": 2.0498267655584546e-06, "loss": 0.0219, "step": 73050 }, { "grad_norm": 0.248702734708786, "learning_rate": 2.0439735624646626e-06, "loss": 0.0163, "step": 73060 }, { "grad_norm": 0.4186318516731262, "learning_rate": 2.0381285537442008e-06, "loss": 0.0121, "step": 73070 }, { "grad_norm": 0.2724093794822693, "learning_rate": 2.032291740395803e-06, "loss": 0.0189, "step": 73080 }, { "grad_norm": 0.23451180756092072, "learning_rate": 2.0264631234168276e-06, "loss": 0.0166, "step": 73090 }, { "grad_norm": 0.24867787957191467, "learning_rate": 2.0206427038032326e-06, "loss": 0.0163, "step": 73100 }, { "grad_norm": 0.1600228250026703, "learning_rate": 2.0148304825495457e-06, "loss": 0.01, "step": 73110 }, { "grad_norm": 0.2154684215784073, "learning_rate": 2.0090264606489496e-06, "loss": 0.0126, "step": 73120 }, { "grad_norm": 0.2902691662311554, "learning_rate": 2.003230639093162e-06, "loss": 0.015, "step": 73130 }, { "grad_norm": 0.24096781015396118, "learning_rate": 1.9974430188725525e-06, "loss": 0.0188, "step": 73140 }, { "grad_norm": 0.3211272060871124, "learning_rate": 1.9916636009760513e-06, "loss": 0.0187, "step": 73150 }, { "grad_norm": 0.21238775551319122, "learning_rate": 1.985892386391208e-06, "loss": 0.0215, "step": 73160 }, { "grad_norm": 0.16600586473941803, "learning_rate": 1.980129376104173e-06, "loss": 0.0202, "step": 73170 }, { "grad_norm": 0.15419000387191772, "learning_rate": 1.9743745710996796e-06, "loss": 0.0167, "step": 73180 }, { "grad_norm": 0.16962073743343353, "learning_rate": 1.9686279723610757e-06, "loss": 0.0168, "step": 73190 }, { "grad_norm": 0.2813991904258728, "learning_rate": 1.962889580870281e-06, "loss": 0.0202, "step": 73200 }, { "grad_norm": 0.2354975789785385, "learning_rate": 1.9571593976078385e-06, "loss": 0.0192, "step": 73210 }, { "grad_norm": 0.21567925810813904, "learning_rate": 1.9514374235528876e-06, "loss": 0.0188, "step": 73220 }, { "grad_norm": 0.22200189530849457, "learning_rate": 1.9457236596831408e-06, "loss": 0.0152, "step": 73230 }, { "grad_norm": 0.23225367069244385, "learning_rate": 1.940018106974939e-06, "loss": 0.0202, "step": 73240 }, { "grad_norm": 0.19598564505577087, "learning_rate": 1.9343207664032025e-06, "loss": 0.014, "step": 73250 }, { "grad_norm": 0.20498931407928467, "learning_rate": 1.928631638941436e-06, "loss": 0.0123, "step": 73260 }, { "grad_norm": 0.23439927399158478, "learning_rate": 1.922950725561784e-06, "loss": 0.0141, "step": 73270 }, { "grad_norm": 0.22526195645332336, "learning_rate": 1.9172780272349312e-06, "loss": 0.0107, "step": 73280 }, { "grad_norm": 0.2720731198787689, "learning_rate": 1.911613544930202e-06, "loss": 0.0153, "step": 73290 }, { "grad_norm": 0.1951770931482315, "learning_rate": 1.9059572796155e-06, "loss": 0.0191, "step": 73300 }, { "grad_norm": 0.42760103940963745, "learning_rate": 1.9003092322573192e-06, "loss": 0.015, "step": 73310 }, { "grad_norm": 0.22362016141414642, "learning_rate": 1.8946694038207647e-06, "loss": 0.0235, "step": 73320 }, { "grad_norm": 0.49854934215545654, "learning_rate": 1.8890377952695215e-06, "loss": 0.0161, "step": 73330 }, { "grad_norm": 0.22775684297084808, "learning_rate": 1.8834144075658865e-06, "loss": 0.0128, "step": 73340 }, { "grad_norm": 0.3576570451259613, "learning_rate": 1.8777992416707357e-06, "loss": 0.0182, "step": 73350 }, { "grad_norm": 0.3996261656284332, "learning_rate": 1.8721922985435458e-06, "loss": 0.0194, "step": 73360 }, { "grad_norm": 0.1543685644865036, "learning_rate": 1.8665935791424006e-06, "loss": 0.0192, "step": 73370 }, { "grad_norm": 0.39775794744491577, "learning_rate": 1.8610030844239512e-06, "loss": 0.0157, "step": 73380 }, { "grad_norm": 0.2855686843395233, "learning_rate": 1.8554208153434838e-06, "loss": 0.0155, "step": 73390 }, { "grad_norm": 0.16864299774169922, "learning_rate": 1.8498467728548296e-06, "loss": 0.0114, "step": 73400 }, { "grad_norm": 0.19109556078910828, "learning_rate": 1.8442809579104547e-06, "loss": 0.0201, "step": 73410 }, { "grad_norm": 0.3114645481109619, "learning_rate": 1.8387233714614039e-06, "loss": 0.0178, "step": 73420 }, { "grad_norm": 0.2261662483215332, "learning_rate": 1.8331740144573174e-06, "loss": 0.0224, "step": 73430 }, { "grad_norm": 0.21569521725177765, "learning_rate": 1.8276328878464255e-06, "loss": 0.0141, "step": 73440 }, { "grad_norm": 0.25348174571990967, "learning_rate": 1.82209999257556e-06, "loss": 0.0158, "step": 73450 }, { "grad_norm": 0.15764376521110535, "learning_rate": 1.8165753295901311e-06, "loss": 0.014, "step": 73460 }, { "grad_norm": 0.1656404286623001, "learning_rate": 1.8110588998341616e-06, "loss": 0.0191, "step": 73470 }, { "grad_norm": 0.28103336691856384, "learning_rate": 1.805550704250253e-06, "loss": 0.0272, "step": 73480 }, { "grad_norm": 0.25565752387046814, "learning_rate": 1.8000507437796077e-06, "loss": 0.0157, "step": 73490 }, { "grad_norm": 0.3122187852859497, "learning_rate": 1.7945590193620242e-06, "loss": 0.0158, "step": 73500 }, { "grad_norm": 0.5174843072891235, "learning_rate": 1.7890755319358742e-06, "loss": 0.0163, "step": 73510 }, { "grad_norm": 0.3166279196739197, "learning_rate": 1.7836002824381525e-06, "loss": 0.0186, "step": 73520 }, { "grad_norm": 0.2881190776824951, "learning_rate": 1.7781332718044165e-06, "loss": 0.0139, "step": 73530 }, { "grad_norm": 0.4314686954021454, "learning_rate": 1.772674500968835e-06, "loss": 0.0208, "step": 73540 }, { "grad_norm": 0.3100545406341553, "learning_rate": 1.7672239708641624e-06, "loss": 0.0197, "step": 73550 }, { "grad_norm": 0.18029358983039856, "learning_rate": 1.7617816824217371e-06, "loss": 0.0137, "step": 73560 }, { "grad_norm": 0.2834247350692749, "learning_rate": 1.7563476365715148e-06, "loss": 0.0227, "step": 73570 }, { "grad_norm": 0.17871420085430145, "learning_rate": 1.750921834242003e-06, "loss": 0.0236, "step": 73580 }, { "grad_norm": 0.1684088557958603, "learning_rate": 1.7455042763603436e-06, "loss": 0.0139, "step": 73590 }, { "grad_norm": 0.2167806476354599, "learning_rate": 1.7400949638522345e-06, "loss": 0.0128, "step": 73600 }, { "grad_norm": 0.1841246634721756, "learning_rate": 1.734693897641987e-06, "loss": 0.0266, "step": 73610 }, { "grad_norm": 0.19805924594402313, "learning_rate": 1.7293010786524955e-06, "loss": 0.0117, "step": 73620 }, { "grad_norm": 0.3982715606689453, "learning_rate": 1.72391650780524e-06, "loss": 0.0203, "step": 73630 }, { "grad_norm": 0.30105552077293396, "learning_rate": 1.718540186020301e-06, "loss": 0.0162, "step": 73640 }, { "grad_norm": 0.219637930393219, "learning_rate": 1.7131721142163437e-06, "loss": 0.0183, "step": 73650 }, { "grad_norm": 0.3854406177997589, "learning_rate": 1.7078122933106233e-06, "loss": 0.0212, "step": 73660 }, { "grad_norm": 0.12150073796510696, "learning_rate": 1.7024607242189905e-06, "loss": 0.0187, "step": 73670 }, { "grad_norm": 0.1987123191356659, "learning_rate": 1.6971174078558749e-06, "loss": 0.0174, "step": 73680 }, { "grad_norm": 0.21762511134147644, "learning_rate": 1.6917823451343073e-06, "loss": 0.018, "step": 73690 }, { "grad_norm": 0.40538474917411804, "learning_rate": 1.6864555369659142e-06, "loss": 0.0165, "step": 73700 }, { "grad_norm": 0.35590335726737976, "learning_rate": 1.681136984260878e-06, "loss": 0.0178, "step": 73710 }, { "grad_norm": 0.22814416885375977, "learning_rate": 1.6758266879280172e-06, "loss": 0.0167, "step": 73720 }, { "grad_norm": 0.22651122510433197, "learning_rate": 1.6705246488746996e-06, "loss": 0.0145, "step": 73730 }, { "grad_norm": 0.15399862825870514, "learning_rate": 1.6652308680069062e-06, "loss": 0.0168, "step": 73740 }, { "grad_norm": 0.24078938364982605, "learning_rate": 1.6599453462292081e-06, "loss": 0.0174, "step": 73750 }, { "grad_norm": 0.3149976134300232, "learning_rate": 1.6546680844447326e-06, "loss": 0.0195, "step": 73760 }, { "grad_norm": 0.3025924861431122, "learning_rate": 1.6493990835552475e-06, "loss": 0.0219, "step": 73770 }, { "grad_norm": 0.27536219358444214, "learning_rate": 1.6441383444610547e-06, "loss": 0.0142, "step": 73780 }, { "grad_norm": 0.2682667374610901, "learning_rate": 1.6388858680610908e-06, "loss": 0.0183, "step": 73790 }, { "grad_norm": 0.3398154377937317, "learning_rate": 1.6336416552528543e-06, "loss": 0.0182, "step": 73800 }, { "grad_norm": 0.21235176920890808, "learning_rate": 1.6284057069324288e-06, "loss": 0.0221, "step": 73810 }, { "grad_norm": 0.27562659978866577, "learning_rate": 1.623178023994504e-06, "loss": 0.0201, "step": 73820 }, { "grad_norm": 0.17363780736923218, "learning_rate": 1.6179586073323483e-06, "loss": 0.0179, "step": 73830 }, { "grad_norm": 0.3732071816921234, "learning_rate": 1.612747457837821e-06, "loss": 0.0166, "step": 73840 }, { "grad_norm": 0.23545804619789124, "learning_rate": 1.607544576401354e-06, "loss": 0.015, "step": 73850 }, { "grad_norm": 0.21099533140659332, "learning_rate": 1.6023499639119754e-06, "loss": 0.0145, "step": 73860 }, { "grad_norm": 0.30234506726264954, "learning_rate": 1.5971636212573138e-06, "loss": 0.0155, "step": 73870 }, { "grad_norm": 0.2693105936050415, "learning_rate": 1.5919855493235714e-06, "loss": 0.0148, "step": 73880 }, { "grad_norm": 0.20417609810829163, "learning_rate": 1.586815748995535e-06, "loss": 0.01, "step": 73890 }, { "grad_norm": 0.21386343240737915, "learning_rate": 1.5816542211565866e-06, "loss": 0.0128, "step": 73900 }, { "grad_norm": 0.2510240077972412, "learning_rate": 1.5765009666886765e-06, "loss": 0.0132, "step": 73910 }, { "grad_norm": 0.25318458676338196, "learning_rate": 1.5713559864723781e-06, "loss": 0.0201, "step": 73920 }, { "grad_norm": 0.28556591272354126, "learning_rate": 1.5662192813868104e-06, "loss": 0.0138, "step": 73930 }, { "grad_norm": 0.16452234983444214, "learning_rate": 1.561090852309699e-06, "loss": 0.0206, "step": 73940 }, { "grad_norm": 0.175352081656456, "learning_rate": 1.5559707001173651e-06, "loss": 0.0119, "step": 73950 }, { "grad_norm": 0.46188119053840637, "learning_rate": 1.5508588256846757e-06, "loss": 0.0186, "step": 73960 }, { "grad_norm": 0.18517832458019257, "learning_rate": 1.5457552298851319e-06, "loss": 0.0144, "step": 73970 }, { "grad_norm": 0.20704782009124756, "learning_rate": 1.5406599135907918e-06, "loss": 0.0185, "step": 73980 }, { "grad_norm": 0.22993610799312592, "learning_rate": 1.5355728776723088e-06, "loss": 0.0183, "step": 73990 }, { "grad_norm": 0.18414178490638733, "learning_rate": 1.5304941229989155e-06, "loss": 0.0125, "step": 74000 }, { "grad_norm": 0.21183019876480103, "learning_rate": 1.5254236504384345e-06, "loss": 0.0164, "step": 74010 }, { "grad_norm": 0.16768725216388702, "learning_rate": 1.5203614608572726e-06, "loss": 0.0176, "step": 74020 }, { "grad_norm": 0.23021507263183594, "learning_rate": 1.5153075551204044e-06, "loss": 0.0142, "step": 74030 }, { "grad_norm": 0.3694782555103302, "learning_rate": 1.5102619340914225e-06, "loss": 0.0266, "step": 74040 }, { "grad_norm": 0.33162009716033936, "learning_rate": 1.505224598632482e-06, "loss": 0.0141, "step": 74050 }, { "grad_norm": 0.15292860567569733, "learning_rate": 1.5001955496043164e-06, "loss": 0.0196, "step": 74060 }, { "grad_norm": 0.3765750527381897, "learning_rate": 1.4951747878662604e-06, "loss": 0.0146, "step": 74070 }, { "grad_norm": 0.17165212333202362, "learning_rate": 1.4901623142762221e-06, "loss": 0.0149, "step": 74080 }, { "grad_norm": 0.13996146619319916, "learning_rate": 1.4851581296907001e-06, "loss": 0.0173, "step": 74090 }, { "grad_norm": 0.28226742148399353, "learning_rate": 1.4801622349647714e-06, "loss": 0.0142, "step": 74100 }, { "grad_norm": 0.33563220500946045, "learning_rate": 1.4751746309520976e-06, "loss": 0.0174, "step": 74110 }, { "grad_norm": 0.22659581899642944, "learning_rate": 1.4701953185049132e-06, "loss": 0.021, "step": 74120 }, { "grad_norm": 0.2734839916229248, "learning_rate": 1.4652242984740661e-06, "loss": 0.0184, "step": 74130 }, { "grad_norm": 0.16941331326961517, "learning_rate": 1.4602615717089484e-06, "loss": 0.0232, "step": 74140 }, { "grad_norm": 0.28226789832115173, "learning_rate": 1.4553071390575657e-06, "loss": 0.0203, "step": 74150 }, { "grad_norm": 0.2031104415655136, "learning_rate": 1.450361001366496e-06, "loss": 0.0257, "step": 74160 }, { "grad_norm": 0.3541474938392639, "learning_rate": 1.4454231594809021e-06, "loss": 0.0161, "step": 74170 }, { "grad_norm": 0.19208641350269318, "learning_rate": 1.4404936142445036e-06, "loss": 0.019, "step": 74180 }, { "grad_norm": 0.18805727362632751, "learning_rate": 1.4355723664996546e-06, "loss": 0.0153, "step": 74190 }, { "grad_norm": 0.3348205089569092, "learning_rate": 1.4306594170872433e-06, "loss": 0.0146, "step": 74200 }, { "grad_norm": 0.3672725558280945, "learning_rate": 1.4257547668467598e-06, "loss": 0.0257, "step": 74210 }, { "grad_norm": 0.28728359937667847, "learning_rate": 1.4208584166162886e-06, "loss": 0.0181, "step": 74220 }, { "grad_norm": 0.31810814142227173, "learning_rate": 1.4159703672324553e-06, "loss": 0.0205, "step": 74230 }, { "grad_norm": 0.26865559816360474, "learning_rate": 1.4110906195305252e-06, "loss": 0.0203, "step": 74240 }, { "grad_norm": 0.589221179485321, "learning_rate": 1.406219174344292e-06, "loss": 0.0189, "step": 74250 }, { "grad_norm": 0.3129373788833618, "learning_rate": 1.401356032506157e-06, "loss": 0.0151, "step": 74260 }, { "grad_norm": 0.5635836720466614, "learning_rate": 1.3965011948471051e-06, "loss": 0.0206, "step": 74270 }, { "grad_norm": 0.29368677735328674, "learning_rate": 1.3916546621966842e-06, "loss": 0.0155, "step": 74280 }, { "grad_norm": 0.3145950436592102, "learning_rate": 1.3868164353830482e-06, "loss": 0.0116, "step": 74290 }, { "grad_norm": 0.37132906913757324, "learning_rate": 1.3819865152329082e-06, "loss": 0.0137, "step": 74300 }, { "grad_norm": 0.20941904187202454, "learning_rate": 1.377164902571565e-06, "loss": 0.0177, "step": 74310 }, { "grad_norm": 0.19350971281528473, "learning_rate": 1.372351598222904e-06, "loss": 0.0178, "step": 74320 }, { "grad_norm": 0.23451952636241913, "learning_rate": 1.3675466030093843e-06, "loss": 0.0184, "step": 74330 }, { "grad_norm": 0.3095012903213501, "learning_rate": 1.3627499177520486e-06, "loss": 0.0173, "step": 74340 }, { "grad_norm": 0.13780994713306427, "learning_rate": 1.3579615432705196e-06, "loss": 0.0203, "step": 74350 }, { "grad_norm": 0.2731039524078369, "learning_rate": 1.3531814803829978e-06, "loss": 0.0198, "step": 74360 }, { "grad_norm": 0.12460819631814957, "learning_rate": 1.3484097299062803e-06, "loss": 0.0171, "step": 74370 }, { "grad_norm": 0.09978358447551727, "learning_rate": 1.343646292655698e-06, "loss": 0.0183, "step": 74380 }, { "grad_norm": 0.2552436590194702, "learning_rate": 1.3388911694452277e-06, "loss": 0.0158, "step": 74390 }, { "grad_norm": 0.3685879707336426, "learning_rate": 1.3341443610873638e-06, "loss": 0.0233, "step": 74400 }, { "grad_norm": 0.2478252649307251, "learning_rate": 1.329405868393213e-06, "loss": 0.0146, "step": 74410 }, { "grad_norm": 0.33159157633781433, "learning_rate": 1.3246756921724613e-06, "loss": 0.023, "step": 74420 }, { "grad_norm": 0.262236624956131, "learning_rate": 1.3199538332333506e-06, "loss": 0.012, "step": 74430 }, { "grad_norm": 0.31717222929000854, "learning_rate": 1.3152402923827411e-06, "loss": 0.0189, "step": 74440 }, { "grad_norm": 0.4015195369720459, "learning_rate": 1.3105350704260277e-06, "loss": 0.0142, "step": 74450 }, { "grad_norm": 0.3500507175922394, "learning_rate": 1.305838168167206e-06, "loss": 0.0128, "step": 74460 }, { "grad_norm": 0.16596904397010803, "learning_rate": 1.301149586408862e-06, "loss": 0.0139, "step": 74470 }, { "grad_norm": 0.2711133360862732, "learning_rate": 1.2964693259521322e-06, "loss": 0.0189, "step": 74480 }, { "grad_norm": 0.2687573730945587, "learning_rate": 1.2917973875967548e-06, "loss": 0.0146, "step": 74490 }, { "grad_norm": 0.31487223505973816, "learning_rate": 1.2871337721410249e-06, "loss": 0.0185, "step": 74500 }, { "grad_norm": 0.4732453227043152, "learning_rate": 1.2824784803818379e-06, "loss": 0.0168, "step": 74510 }, { "grad_norm": 0.29700329899787903, "learning_rate": 1.2778315131146467e-06, "loss": 0.0152, "step": 74520 }, { "grad_norm": 0.3681633770465851, "learning_rate": 1.2731928711334994e-06, "loss": 0.0182, "step": 74530 }, { "grad_norm": 0.3312181532382965, "learning_rate": 1.2685625552310122e-06, "loss": 0.0154, "step": 74540 }, { "grad_norm": 0.1524854600429535, "learning_rate": 1.263940566198374e-06, "loss": 0.0151, "step": 74550 }, { "grad_norm": 0.1736474335193634, "learning_rate": 1.259326904825353e-06, "loss": 0.0161, "step": 74560 }, { "grad_norm": 0.32202446460723877, "learning_rate": 1.2547215719003137e-06, "loss": 0.0209, "step": 74570 }, { "grad_norm": 0.16342516243457794, "learning_rate": 1.2501245682101703e-06, "loss": 0.0154, "step": 74580 }, { "grad_norm": 0.21373388171195984, "learning_rate": 1.2455358945404171e-06, "loss": 0.0167, "step": 74590 }, { "grad_norm": 0.24530096352100372, "learning_rate": 1.2409555516751493e-06, "loss": 0.0151, "step": 74600 }, { "grad_norm": 0.20746222138404846, "learning_rate": 1.2363835403970125e-06, "loss": 0.0149, "step": 74610 }, { "grad_norm": 0.37318500876426697, "learning_rate": 1.2318198614872489e-06, "loss": 0.015, "step": 74620 }, { "grad_norm": 0.3583592474460602, "learning_rate": 1.2272645157256457e-06, "loss": 0.0156, "step": 74630 }, { "grad_norm": 0.26848655939102173, "learning_rate": 1.222717503890608e-06, "loss": 0.0222, "step": 74640 }, { "grad_norm": 0.334648072719574, "learning_rate": 1.218178826759081e-06, "loss": 0.0217, "step": 74650 }, { "grad_norm": 0.4097161591053009, "learning_rate": 1.213648485106611e-06, "loss": 0.0199, "step": 74660 }, { "grad_norm": 0.1501341313123703, "learning_rate": 1.2091264797073066e-06, "loss": 0.0247, "step": 74670 }, { "grad_norm": 0.32192596793174744, "learning_rate": 1.2046128113338494e-06, "loss": 0.026, "step": 74680 }, { "grad_norm": 0.18419398367404938, "learning_rate": 1.200107480757512e-06, "loss": 0.0179, "step": 74690 }, { "grad_norm": 0.24138736724853516, "learning_rate": 1.1956104887481168e-06, "loss": 0.0224, "step": 74700 }, { "grad_norm": 0.3283846378326416, "learning_rate": 1.191121836074094e-06, "loss": 0.0131, "step": 74710 }, { "grad_norm": 0.35612863302230835, "learning_rate": 1.1866415235024186e-06, "loss": 0.0119, "step": 74720 }, { "grad_norm": 0.2779250741004944, "learning_rate": 1.182169551798662e-06, "loss": 0.0249, "step": 74730 }, { "grad_norm": 0.1879705786705017, "learning_rate": 1.177705921726957e-06, "loss": 0.0124, "step": 74740 }, { "grad_norm": 0.23412713408470154, "learning_rate": 1.173250634050016e-06, "loss": 0.0169, "step": 74750 }, { "grad_norm": 0.33226174116134644, "learning_rate": 1.1688036895291354e-06, "loss": 0.0169, "step": 74760 }, { "grad_norm": 0.2321154624223709, "learning_rate": 1.1643650889241574e-06, "loss": 0.0138, "step": 74770 }, { "grad_norm": 0.3570016324520111, "learning_rate": 1.1599348329935311e-06, "loss": 0.0168, "step": 74780 }, { "grad_norm": 0.22538842260837555, "learning_rate": 1.1555129224942673e-06, "loss": 0.024, "step": 74790 }, { "grad_norm": 0.11590112000703812, "learning_rate": 1.1510993581819396e-06, "loss": 0.0198, "step": 74800 }, { "grad_norm": 0.5223960280418396, "learning_rate": 1.1466941408107112e-06, "loss": 0.0222, "step": 74810 }, { "grad_norm": 0.15710341930389404, "learning_rate": 1.1422972711333247e-06, "loss": 0.0176, "step": 74820 }, { "grad_norm": 0.4105745553970337, "learning_rate": 1.1379087499010565e-06, "loss": 0.0168, "step": 74830 }, { "grad_norm": 0.2930414378643036, "learning_rate": 1.1335285778638127e-06, "loss": 0.0121, "step": 74840 }, { "grad_norm": 0.22655747830867767, "learning_rate": 1.1291567557700333e-06, "loss": 0.0158, "step": 74850 }, { "grad_norm": 0.29033344984054565, "learning_rate": 1.124793284366743e-06, "loss": 0.0161, "step": 74860 }, { "grad_norm": 0.12652960419654846, "learning_rate": 1.1204381643995455e-06, "loss": 0.014, "step": 74870 }, { "grad_norm": 0.13572391867637634, "learning_rate": 1.116091396612595e-06, "loss": 0.0172, "step": 74880 }, { "grad_norm": 0.2502043843269348, "learning_rate": 1.1117529817486538e-06, "loss": 0.0136, "step": 74890 }, { "grad_norm": 0.20559090375900269, "learning_rate": 1.107422920549034e-06, "loss": 0.0137, "step": 74900 }, { "grad_norm": 0.2820134460926056, "learning_rate": 1.1031012137536157e-06, "loss": 0.0163, "step": 74910 }, { "grad_norm": 0.3321278989315033, "learning_rate": 1.0987878621008695e-06, "loss": 0.0139, "step": 74920 }, { "grad_norm": 0.29164090752601624, "learning_rate": 1.0944828663278284e-06, "loss": 0.0169, "step": 74930 }, { "grad_norm": 0.24218064546585083, "learning_rate": 1.0901862271700925e-06, "loss": 0.015, "step": 74940 }, { "grad_norm": 0.1839224249124527, "learning_rate": 1.0858979453618467e-06, "loss": 0.0159, "step": 74950 }, { "grad_norm": 0.3120897710323334, "learning_rate": 1.0816180216358441e-06, "loss": 0.0167, "step": 74960 }, { "grad_norm": 0.21849024295806885, "learning_rate": 1.0773464567233937e-06, "loss": 0.013, "step": 74970 }, { "grad_norm": 0.20124758780002594, "learning_rate": 1.0730832513543953e-06, "loss": 0.0148, "step": 74980 }, { "grad_norm": 0.28408288955688477, "learning_rate": 1.0688284062573218e-06, "loss": 0.0173, "step": 74990 }, { "grad_norm": 0.303241103887558, "learning_rate": 1.0645819221591969e-06, "loss": 0.0105, "step": 75000 }, { "grad_norm": 0.38117220997810364, "learning_rate": 1.0603437997856347e-06, "loss": 0.0162, "step": 75010 }, { "grad_norm": 0.1826811134815216, "learning_rate": 1.0561140398608228e-06, "loss": 0.0158, "step": 75020 }, { "grad_norm": 0.22240418195724487, "learning_rate": 1.0518926431074937e-06, "loss": 0.0143, "step": 75030 }, { "grad_norm": 0.14453715085983276, "learning_rate": 1.0476796102469877e-06, "loss": 0.0159, "step": 75040 }, { "grad_norm": 0.3318973183631897, "learning_rate": 1.0434749419991786e-06, "loss": 0.0193, "step": 75050 }, { "grad_norm": 0.30087947845458984, "learning_rate": 1.0392786390825415e-06, "loss": 0.0175, "step": 75060 }, { "grad_norm": 0.36299365758895874, "learning_rate": 1.0350907022141087e-06, "loss": 0.0143, "step": 75070 }, { "grad_norm": 0.25102612376213074, "learning_rate": 1.0309111321094744e-06, "loss": 0.0161, "step": 75080 }, { "grad_norm": 0.17389553785324097, "learning_rate": 1.026739929482834e-06, "loss": 0.0153, "step": 75090 }, { "grad_norm": 0.3927614390850067, "learning_rate": 1.022577095046906e-06, "loss": 0.014, "step": 75100 }, { "grad_norm": 0.2317502349615097, "learning_rate": 1.0184226295130217e-06, "loss": 0.0132, "step": 75110 }, { "grad_norm": 0.36883309483528137, "learning_rate": 1.0142765335910575e-06, "loss": 0.0179, "step": 75120 }, { "grad_norm": 0.3639971613883972, "learning_rate": 1.0101388079894746e-06, "loss": 0.0157, "step": 75130 }, { "grad_norm": 0.2480919361114502, "learning_rate": 1.0060094534152908e-06, "loss": 0.0163, "step": 75140 }, { "grad_norm": 0.2904423773288727, "learning_rate": 1.0018884705741028e-06, "loss": 0.0157, "step": 75150 }, { "grad_norm": 0.2982962429523468, "learning_rate": 9.977758601700805e-07, "loss": 0.0221, "step": 75160 }, { "grad_norm": 0.20910318195819855, "learning_rate": 9.93671622905945e-07, "loss": 0.0264, "step": 75170 }, { "grad_norm": 0.3110122084617615, "learning_rate": 9.895757594830024e-07, "loss": 0.0164, "step": 75180 }, { "grad_norm": 0.2033739537000656, "learning_rate": 9.854882706011204e-07, "loss": 0.0133, "step": 75190 }, { "grad_norm": 0.34265977144241333, "learning_rate": 9.814091569587513e-07, "loss": 0.0172, "step": 75200 }, { "grad_norm": 0.5576249957084656, "learning_rate": 9.77338419252888e-07, "loss": 0.018, "step": 75210 }, { "grad_norm": 0.28581520915031433, "learning_rate": 9.732760581791289e-07, "loss": 0.0186, "step": 75220 }, { "grad_norm": 0.20675940811634064, "learning_rate": 9.692220744315972e-07, "loss": 0.017, "step": 75230 }, { "grad_norm": 0.1950102001428604, "learning_rate": 9.651764687030162e-07, "loss": 0.0154, "step": 75240 }, { "grad_norm": 0.1528272032737732, "learning_rate": 9.611392416846776e-07, "loss": 0.0152, "step": 75250 }, { "grad_norm": 0.2869224548339844, "learning_rate": 9.571103940664238e-07, "loss": 0.0191, "step": 75260 }, { "grad_norm": 0.22669433057308197, "learning_rate": 9.530899265366822e-07, "loss": 0.018, "step": 75270 }, { "grad_norm": 0.13284200429916382, "learning_rate": 9.490778397824307e-07, "loss": 0.0156, "step": 75280 }, { "grad_norm": 0.17685440182685852, "learning_rate": 9.450741344892322e-07, "loss": 0.015, "step": 75290 }, { "grad_norm": 0.4193316102027893, "learning_rate": 9.410788113412116e-07, "loss": 0.0162, "step": 75300 }, { "grad_norm": 0.2155950665473938, "learning_rate": 9.370918710210563e-07, "loss": 0.0172, "step": 75310 }, { "grad_norm": 0.2350604087114334, "learning_rate": 9.331133142100268e-07, "loss": 0.0173, "step": 75320 }, { "grad_norm": 0.6948925852775574, "learning_rate": 9.291431415879459e-07, "loss": 0.0164, "step": 75330 }, { "grad_norm": 0.26492059230804443, "learning_rate": 9.251813538332155e-07, "loss": 0.0222, "step": 75340 }, { "grad_norm": 0.1505436897277832, "learning_rate": 9.212279516227884e-07, "loss": 0.0138, "step": 75350 }, { "grad_norm": 0.29003384709358215, "learning_rate": 9.172829356322021e-07, "loss": 0.0172, "step": 75360 }, { "grad_norm": 0.26314184069633484, "learning_rate": 9.133463065355452e-07, "loss": 0.019, "step": 75370 }, { "grad_norm": 0.4487380087375641, "learning_rate": 9.094180650054796e-07, "loss": 0.0222, "step": 75380 }, { "grad_norm": 0.40510129928588867, "learning_rate": 9.054982117132404e-07, "loss": 0.0146, "step": 75390 }, { "grad_norm": 0.2133757770061493, "learning_rate": 9.015867473286143e-07, "loss": 0.0215, "step": 75400 }, { "grad_norm": 0.33535122871398926, "learning_rate": 8.976836725199778e-07, "loss": 0.016, "step": 75410 }, { "grad_norm": 0.2877900302410126, "learning_rate": 8.937889879542416e-07, "loss": 0.016, "step": 75420 }, { "grad_norm": 0.4122331738471985, "learning_rate": 8.899026942969068e-07, "loss": 0.0178, "step": 75430 }, { "grad_norm": 0.2139163464307785, "learning_rate": 8.860247922120424e-07, "loss": 0.018, "step": 75440 }, { "grad_norm": 0.3635210394859314, "learning_rate": 8.821552823622737e-07, "loss": 0.0135, "step": 75450 }, { "grad_norm": 0.25289255380630493, "learning_rate": 8.782941654087884e-07, "loss": 0.0161, "step": 75460 }, { "grad_norm": 0.4530225694179535, "learning_rate": 8.744414420113478e-07, "loss": 0.0193, "step": 75470 }, { "grad_norm": 0.14657655358314514, "learning_rate": 8.705971128282753e-07, "loss": 0.0151, "step": 75480 }, { "grad_norm": 0.38127556443214417, "learning_rate": 8.66761178516473e-07, "loss": 0.0152, "step": 75490 }, { "grad_norm": 0.24372771382331848, "learning_rate": 8.629336397313781e-07, "loss": 0.0168, "step": 75500 }, { "grad_norm": 0.2121659815311432, "learning_rate": 8.591144971270282e-07, "loss": 0.0222, "step": 75510 }, { "grad_norm": 0.27209699153900146, "learning_rate": 8.553037513560069e-07, "loss": 0.0385, "step": 75520 }, { "grad_norm": 0.2981838881969452, "learning_rate": 8.515014030694546e-07, "loss": 0.0168, "step": 75530 }, { "grad_norm": 0.2240973263978958, "learning_rate": 8.47707452917107e-07, "loss": 0.0163, "step": 75540 }, { "grad_norm": 0.284253865480423, "learning_rate": 8.439219015472343e-07, "loss": 0.0166, "step": 75550 }, { "grad_norm": 0.397988498210907, "learning_rate": 8.401447496066861e-07, "loss": 0.0112, "step": 75560 }, { "grad_norm": 0.43670016527175903, "learning_rate": 8.363759977408792e-07, "loss": 0.0135, "step": 75570 }, { "grad_norm": 0.2043374478816986, "learning_rate": 8.326156465937817e-07, "loss": 0.017, "step": 75580 }, { "grad_norm": 0.5134929418563843, "learning_rate": 8.28863696807941e-07, "loss": 0.0282, "step": 75590 }, { "grad_norm": 0.283855140209198, "learning_rate": 8.251201490244609e-07, "loss": 0.0183, "step": 75600 }, { "grad_norm": 0.13521994650363922, "learning_rate": 8.213850038830129e-07, "loss": 0.0166, "step": 75610 }, { "grad_norm": 0.29241475462913513, "learning_rate": 8.176582620218309e-07, "loss": 0.0163, "step": 75620 }, { "grad_norm": 0.25795117020606995, "learning_rate": 8.139399240777057e-07, "loss": 0.0137, "step": 75630 }, { "grad_norm": 0.29480305314064026, "learning_rate": 8.102299906860122e-07, "loss": 0.0174, "step": 75640 }, { "grad_norm": 0.6844567060470581, "learning_rate": 8.065284624806657e-07, "loss": 0.0198, "step": 75650 }, { "grad_norm": 0.28266051411628723, "learning_rate": 8.028353400941601e-07, "loss": 0.0259, "step": 75660 }, { "grad_norm": 0.21224310994148254, "learning_rate": 7.991506241575575e-07, "loss": 0.0197, "step": 75670 }, { "grad_norm": 0.4189591705799103, "learning_rate": 7.954743153004541e-07, "loss": 0.0179, "step": 75680 }, { "grad_norm": 0.1741294264793396, "learning_rate": 7.918064141510528e-07, "loss": 0.0194, "step": 75690 }, { "grad_norm": 0.31992965936660767, "learning_rate": 7.881469213360859e-07, "loss": 0.0245, "step": 75700 }, { "grad_norm": 0.20351696014404297, "learning_rate": 7.844958374808642e-07, "loss": 0.0123, "step": 75710 }, { "grad_norm": 0.14564432203769684, "learning_rate": 7.808531632092608e-07, "loss": 0.0135, "step": 75720 }, { "grad_norm": 0.2313598245382309, "learning_rate": 7.772188991436946e-07, "loss": 0.0172, "step": 75730 }, { "grad_norm": 0.14630350470542908, "learning_rate": 7.735930459051799e-07, "loss": 0.0168, "step": 75740 }, { "grad_norm": 0.24651873111724854, "learning_rate": 7.699756041132655e-07, "loss": 0.0211, "step": 75750 }, { "grad_norm": 0.1773328334093094, "learning_rate": 7.663665743860793e-07, "loss": 0.0122, "step": 75760 }, { "grad_norm": 0.22811256349086761, "learning_rate": 7.627659573403001e-07, "loss": 0.0163, "step": 75770 }, { "grad_norm": 0.19611626863479614, "learning_rate": 7.591737535911802e-07, "loss": 0.0163, "step": 75780 }, { "grad_norm": 0.16716088354587555, "learning_rate": 7.555899637525288e-07, "loss": 0.0122, "step": 75790 }, { "grad_norm": 0.26098424196243286, "learning_rate": 7.520145884367058e-07, "loss": 0.0154, "step": 75800 }, { "grad_norm": 0.19552752375602722, "learning_rate": 7.484476282546615e-07, "loss": 0.0146, "step": 75810 }, { "grad_norm": 0.34180229902267456, "learning_rate": 7.448890838158806e-07, "loss": 0.0229, "step": 75820 }, { "grad_norm": 0.1824348270893097, "learning_rate": 7.413389557284267e-07, "loss": 0.0205, "step": 75830 }, { "grad_norm": 0.2417364865541458, "learning_rate": 7.377972445989201e-07, "loss": 0.0184, "step": 75840 }, { "grad_norm": 0.363971084356308, "learning_rate": 7.342639510325377e-07, "loss": 0.0163, "step": 75850 }, { "grad_norm": 0.23682312667369843, "learning_rate": 7.307390756330246e-07, "loss": 0.0191, "step": 75860 }, { "grad_norm": 0.2887866795063019, "learning_rate": 7.272226190026876e-07, "loss": 0.0135, "step": 75870 }, { "grad_norm": 0.16018418967723846, "learning_rate": 7.237145817423907e-07, "loss": 0.0135, "step": 75880 }, { "grad_norm": 0.21539859473705292, "learning_rate": 7.202149644515654e-07, "loss": 0.0149, "step": 75890 }, { "grad_norm": 0.2880699932575226, "learning_rate": 7.167237677281946e-07, "loss": 0.0176, "step": 75900 }, { "grad_norm": 0.1629408448934555, "learning_rate": 7.132409921688288e-07, "loss": 0.019, "step": 75910 }, { "grad_norm": 0.22966326773166656, "learning_rate": 7.09766638368592e-07, "loss": 0.0105, "step": 75920 }, { "grad_norm": 0.31782761216163635, "learning_rate": 7.063007069211313e-07, "loss": 0.0143, "step": 75930 }, { "grad_norm": 0.3755432069301605, "learning_rate": 7.028431984187067e-07, "loss": 0.0131, "step": 75940 }, { "grad_norm": 0.14575815200805664, "learning_rate": 6.993941134520898e-07, "loss": 0.0111, "step": 75950 }, { "grad_norm": 0.2299565076828003, "learning_rate": 6.959534526106537e-07, "loss": 0.0146, "step": 75960 }, { "grad_norm": 0.23179113864898682, "learning_rate": 6.925212164822948e-07, "loss": 0.0144, "step": 75970 }, { "grad_norm": 0.4030802845954895, "learning_rate": 6.890974056535049e-07, "loss": 0.0207, "step": 75980 }, { "grad_norm": 0.23908260464668274, "learning_rate": 6.856820207093106e-07, "loss": 0.0119, "step": 75990 }, { "grad_norm": 0.18735451996326447, "learning_rate": 6.822750622333063e-07, "loss": 0.0131, "step": 76000 }, { "grad_norm": 0.4693894684314728, "learning_rate": 6.788765308076539e-07, "loss": 0.0215, "step": 76010 }, { "grad_norm": 0.35276681184768677, "learning_rate": 6.754864270130668e-07, "loss": 0.0148, "step": 76020 }, { "grad_norm": 0.1955677717924118, "learning_rate": 6.721047514288203e-07, "loss": 0.0139, "step": 76030 }, { "grad_norm": 0.3143593668937683, "learning_rate": 6.687315046327469e-07, "loss": 0.0181, "step": 76040 }, { "grad_norm": 0.2932458817958832, "learning_rate": 6.65366687201252e-07, "loss": 0.0152, "step": 76050 }, { "grad_norm": 0.2590087652206421, "learning_rate": 6.620102997092814e-07, "loss": 0.0181, "step": 76060 }, { "grad_norm": 0.19937913119792938, "learning_rate": 6.586623427303596e-07, "loss": 0.0116, "step": 76070 }, { "grad_norm": 0.20815223455429077, "learning_rate": 6.553228168365455e-07, "loss": 0.0193, "step": 76080 }, { "grad_norm": 0.24353426694869995, "learning_rate": 6.519917225984884e-07, "loss": 0.0187, "step": 76090 }, { "grad_norm": 1.0261237621307373, "learning_rate": 6.486690605853718e-07, "loss": 0.0206, "step": 76100 }, { "grad_norm": 0.36151382327079773, "learning_rate": 6.453548313649527e-07, "loss": 0.0139, "step": 76110 }, { "grad_norm": 0.4026942849159241, "learning_rate": 6.420490355035446e-07, "loss": 0.0193, "step": 76120 }, { "grad_norm": 0.15700963139533997, "learning_rate": 6.387516735660071e-07, "loss": 0.017, "step": 76130 }, { "grad_norm": 0.46701475977897644, "learning_rate": 6.35462746115778e-07, "loss": 0.0131, "step": 76140 }, { "grad_norm": 0.27725329995155334, "learning_rate": 6.321822537148358e-07, "loss": 0.0179, "step": 76150 }, { "grad_norm": 0.30389633774757385, "learning_rate": 6.289101969237432e-07, "loss": 0.0165, "step": 76160 }, { "grad_norm": 0.22925233840942383, "learning_rate": 6.256465763015918e-07, "loss": 0.0149, "step": 76170 }, { "grad_norm": 0.30316993594169617, "learning_rate": 6.22391392406052e-07, "loss": 0.0167, "step": 76180 }, { "grad_norm": 0.29565298557281494, "learning_rate": 6.191446457933403e-07, "loss": 0.0137, "step": 76190 }, { "grad_norm": 0.2746436297893524, "learning_rate": 6.159063370182406e-07, "loss": 0.0159, "step": 76200 }, { "grad_norm": 0.10651510953903198, "learning_rate": 6.126764666340879e-07, "loss": 0.0127, "step": 76210 }, { "grad_norm": 0.22838106751441956, "learning_rate": 6.094550351927852e-07, "loss": 0.0123, "step": 76220 }, { "grad_norm": 0.32535868883132935, "learning_rate": 6.062420432447757e-07, "loss": 0.0146, "step": 76230 }, { "grad_norm": 0.19811557233333588, "learning_rate": 6.030374913390813e-07, "loss": 0.0134, "step": 76240 }, { "grad_norm": 0.27481141686439514, "learning_rate": 5.998413800232694e-07, "loss": 0.0178, "step": 76250 }, { "grad_norm": 0.2499595433473587, "learning_rate": 5.966537098434755e-07, "loss": 0.0158, "step": 76260 }, { "grad_norm": 0.14588525891304016, "learning_rate": 5.934744813443694e-07, "loss": 0.0158, "step": 76270 }, { "grad_norm": 0.24630488455295563, "learning_rate": 5.903036950692054e-07, "loss": 0.0171, "step": 76280 }, { "grad_norm": 0.1908562183380127, "learning_rate": 5.871413515597835e-07, "loss": 0.0195, "step": 76290 }, { "grad_norm": 0.22830557823181152, "learning_rate": 5.839874513564547e-07, "loss": 0.0142, "step": 76300 }, { "grad_norm": 0.32672321796417236, "learning_rate": 5.808419949981436e-07, "loss": 0.0159, "step": 76310 }, { "grad_norm": 0.1261412352323532, "learning_rate": 5.777049830223257e-07, "loss": 0.0157, "step": 76320 }, { "grad_norm": 0.1518452763557434, "learning_rate": 5.745764159650114e-07, "loss": 0.0121, "step": 76330 }, { "grad_norm": 0.29470494389533997, "learning_rate": 5.71456294360806e-07, "loss": 0.0145, "step": 76340 }, { "grad_norm": 0.36531326174736023, "learning_rate": 5.683446187428443e-07, "loss": 0.0178, "step": 76350 }, { "grad_norm": 0.4145505428314209, "learning_rate": 5.652413896428288e-07, "loss": 0.0153, "step": 76360 }, { "grad_norm": 0.34932446479797363, "learning_rate": 5.621466075910131e-07, "loss": 0.0183, "step": 76370 }, { "grad_norm": 0.14137886464595795, "learning_rate": 5.590602731162187e-07, "loss": 0.0147, "step": 76380 }, { "grad_norm": 0.3464077115058899, "learning_rate": 5.559823867458125e-07, "loss": 0.0139, "step": 76390 }, { "grad_norm": 0.19243820011615753, "learning_rate": 5.529129490057128e-07, "loss": 0.0225, "step": 76400 }, { "grad_norm": 0.2847762405872345, "learning_rate": 5.498519604204167e-07, "loss": 0.0154, "step": 76410 }, { "grad_norm": 0.307015985250473, "learning_rate": 5.467994215129501e-07, "loss": 0.0152, "step": 76420 }, { "grad_norm": 0.2744212746620178, "learning_rate": 5.437553328049183e-07, "loss": 0.016, "step": 76430 }, { "grad_norm": 0.3888895511627197, "learning_rate": 5.407196948164661e-07, "loss": 0.0127, "step": 76440 }, { "grad_norm": 0.32557350397109985, "learning_rate": 5.376925080663009e-07, "loss": 0.016, "step": 76450 }, { "grad_norm": 0.23114006221294403, "learning_rate": 5.346737730716977e-07, "loss": 0.0173, "step": 76460 }, { "grad_norm": 0.3590814471244812, "learning_rate": 5.316634903484607e-07, "loss": 0.0154, "step": 76470 }, { "grad_norm": 0.31293538212776184, "learning_rate": 5.286616604109728e-07, "loss": 0.0141, "step": 76480 }, { "grad_norm": 0.16901829838752747, "learning_rate": 5.256682837721627e-07, "loss": 0.0214, "step": 76490 }, { "grad_norm": 0.28497257828712463, "learning_rate": 5.226833609435156e-07, "loss": 0.0178, "step": 76500 }, { "grad_norm": 0.3045322299003601, "learning_rate": 5.19706892435079e-07, "loss": 0.0163, "step": 76510 }, { "grad_norm": 0.3111637830734253, "learning_rate": 5.167388787554406e-07, "loss": 0.021, "step": 76520 }, { "grad_norm": 0.31871017813682556, "learning_rate": 5.137793204117614e-07, "loss": 0.0179, "step": 76530 }, { "grad_norm": 0.2813889682292938, "learning_rate": 5.108282179097479e-07, "loss": 0.0155, "step": 76540 }, { "grad_norm": 0.2567300796508789, "learning_rate": 5.078855717536523e-07, "loss": 0.013, "step": 76550 }, { "grad_norm": 0.19568084180355072, "learning_rate": 5.049513824463059e-07, "loss": 0.0138, "step": 76560 }, { "grad_norm": 0.2129015028476715, "learning_rate": 5.020256504890742e-07, "loss": 0.0143, "step": 76570 }, { "grad_norm": 0.1476200819015503, "learning_rate": 4.991083763818849e-07, "loss": 0.0168, "step": 76580 }, { "grad_norm": 0.43077462911605835, "learning_rate": 4.961995606232228e-07, "loss": 0.0223, "step": 76590 }, { "grad_norm": 0.3460274338722229, "learning_rate": 4.932992037101236e-07, "loss": 0.0159, "step": 76600 }, { "grad_norm": 0.4122338891029358, "learning_rate": 4.904073061381798e-07, "loss": 0.0201, "step": 76610 }, { "grad_norm": 0.26158714294433594, "learning_rate": 4.875238684015349e-07, "loss": 0.017, "step": 76620 }, { "grad_norm": 0.17000862956047058, "learning_rate": 4.846488909928948e-07, "loss": 0.0126, "step": 76630 }, { "grad_norm": 0.19051317870616913, "learning_rate": 4.81782374403511e-07, "loss": 0.0209, "step": 76640 }, { "grad_norm": 0.1935071051120758, "learning_rate": 4.789243191231918e-07, "loss": 0.0154, "step": 76650 }, { "grad_norm": 0.2518639862537384, "learning_rate": 4.76074725640302e-07, "loss": 0.0203, "step": 76660 }, { "grad_norm": 0.2651249170303345, "learning_rate": 4.732335944417632e-07, "loss": 0.0233, "step": 76670 }, { "grad_norm": 0.37875795364379883, "learning_rate": 4.704009260130371e-07, "loss": 0.019, "step": 76680 }, { "grad_norm": 0.1922609955072403, "learning_rate": 4.675767208381587e-07, "loss": 0.0172, "step": 76690 }, { "grad_norm": 0.4098288118839264, "learning_rate": 4.647609793997032e-07, "loss": 0.0164, "step": 76700 }, { "grad_norm": 0.212840273976326, "learning_rate": 4.619537021788023e-07, "loss": 0.0155, "step": 76710 }, { "grad_norm": 0.2272498458623886, "learning_rate": 4.5915488965515005e-07, "loss": 0.0152, "step": 76720 }, { "grad_norm": 0.25534430146217346, "learning_rate": 4.563645423069807e-07, "loss": 0.0153, "step": 76730 }, { "grad_norm": 0.26383721828460693, "learning_rate": 4.535826606110849e-07, "loss": 0.0165, "step": 76740 }, { "grad_norm": 0.3042580783367157, "learning_rate": 4.508092450428214e-07, "loss": 0.0174, "step": 76750 }, { "grad_norm": 0.24601727724075317, "learning_rate": 4.480442960760778e-07, "loss": 0.019, "step": 76760 }, { "grad_norm": 0.2019413262605667, "learning_rate": 4.4528781418332053e-07, "loss": 0.0145, "step": 76770 }, { "grad_norm": 0.44759321212768555, "learning_rate": 4.4253979983555073e-07, "loss": 0.0204, "step": 76780 }, { "grad_norm": 0.2567749321460724, "learning_rate": 4.3980025350233154e-07, "loss": 0.0141, "step": 76790 }, { "grad_norm": 0.2015332579612732, "learning_rate": 4.370691756517664e-07, "loss": 0.0163, "step": 76800 }, { "grad_norm": 0.16930127143859863, "learning_rate": 4.3434656675053753e-07, "loss": 0.0133, "step": 76810 }, { "grad_norm": 0.28624486923217773, "learning_rate": 4.3163242726385613e-07, "loss": 0.0136, "step": 76820 }, { "grad_norm": 0.2918396592140198, "learning_rate": 4.289267576554956e-07, "loss": 0.0165, "step": 76830 }, { "grad_norm": 0.1834639459848404, "learning_rate": 4.262295583877807e-07, "loss": 0.0129, "step": 76840 }, { "grad_norm": 0.17719130218029022, "learning_rate": 4.235408299215815e-07, "loss": 0.0188, "step": 76850 }, { "grad_norm": 0.20672188699245453, "learning_rate": 4.2086057271634173e-07, "loss": 0.0161, "step": 76860 }, { "grad_norm": 0.2605820894241333, "learning_rate": 4.181887872300394e-07, "loss": 0.0145, "step": 76870 }, { "grad_norm": 0.3007975220680237, "learning_rate": 4.155254739192038e-07, "loss": 0.0165, "step": 76880 }, { "grad_norm": 0.2710750102996826, "learning_rate": 4.128706332389265e-07, "loss": 0.0235, "step": 76890 }, { "grad_norm": 0.2019353210926056, "learning_rate": 4.1022426564284453e-07, "loss": 0.0178, "step": 76900 }, { "grad_norm": 0.19921839237213135, "learning_rate": 4.075863715831574e-07, "loss": 0.0192, "step": 76910 }, { "grad_norm": 0.20106086134910583, "learning_rate": 4.0495695151059887e-07, "loss": 0.0173, "step": 76920 }, { "grad_norm": 0.36065614223480225, "learning_rate": 4.023360058744763e-07, "loss": 0.0157, "step": 76930 }, { "grad_norm": 0.35690659284591675, "learning_rate": 3.997235351226314e-07, "loss": 0.0138, "step": 76940 }, { "grad_norm": 0.3293420374393463, "learning_rate": 3.971195397014571e-07, "loss": 0.0174, "step": 76950 }, { "grad_norm": 0.3136514127254486, "learning_rate": 3.9452402005591417e-07, "loss": 0.0216, "step": 76960 }, { "grad_norm": 0.30848196148872375, "learning_rate": 3.9193697662950334e-07, "loss": 0.0178, "step": 76970 }, { "grad_norm": 0.30635055899620056, "learning_rate": 3.893584098642822e-07, "loss": 0.0208, "step": 76980 }, { "grad_norm": 0.09390762448310852, "learning_rate": 3.867883202008538e-07, "loss": 0.0148, "step": 76990 }, { "grad_norm": 0.48012685775756836, "learning_rate": 3.842267080783779e-07, "loss": 0.0148, "step": 77000 }, { "grad_norm": 0.2025376260280609, "learning_rate": 3.816735739345656e-07, "loss": 0.0199, "step": 77010 }, { "grad_norm": 0.28926366567611694, "learning_rate": 3.791289182056679e-07, "loss": 0.0193, "step": 77020 }, { "grad_norm": 0.23517213761806488, "learning_rate": 3.7659274132650913e-07, "loss": 0.0176, "step": 77030 }, { "grad_norm": 0.22688014805316925, "learning_rate": 3.7406504373044826e-07, "loss": 0.0128, "step": 77040 }, { "grad_norm": 0.3499630093574524, "learning_rate": 3.7154582584939533e-07, "loss": 0.0167, "step": 77050 }, { "grad_norm": 0.300702840089798, "learning_rate": 3.690350881138227e-07, "loss": 0.0174, "step": 77060 }, { "grad_norm": 0.20587357878684998, "learning_rate": 3.665328309527427e-07, "loss": 0.0157, "step": 77070 }, { "grad_norm": 0.37815922498703003, "learning_rate": 3.640390547937245e-07, "loss": 0.0133, "step": 77080 }, { "grad_norm": 0.1641746610403061, "learning_rate": 3.615537600628882e-07, "loss": 0.0193, "step": 77090 }, { "grad_norm": 0.39512720704078674, "learning_rate": 3.5907694718489427e-07, "loss": 0.0174, "step": 77100 }, { "grad_norm": 0.2718081474304199, "learning_rate": 3.566086165829707e-07, "loss": 0.0182, "step": 77110 }, { "grad_norm": 0.4324350655078888, "learning_rate": 3.5414876867888024e-07, "loss": 0.0175, "step": 77120 }, { "grad_norm": 0.23759622871875763, "learning_rate": 3.5169740389295326e-07, "loss": 0.02, "step": 77130 }, { "grad_norm": 0.27249225974082947, "learning_rate": 3.4925452264405466e-07, "loss": 0.013, "step": 77140 }, { "grad_norm": 0.13666470348834991, "learning_rate": 3.468201253496062e-07, "loss": 0.0124, "step": 77150 }, { "grad_norm": 0.3768061101436615, "learning_rate": 3.443942124255861e-07, "loss": 0.0205, "step": 77160 }, { "grad_norm": 0.3359485864639282, "learning_rate": 3.4197678428650183e-07, "loss": 0.0175, "step": 77170 }, { "grad_norm": 0.3963509798049927, "learning_rate": 3.3956784134544504e-07, "loss": 0.0157, "step": 77180 }, { "grad_norm": 0.3300403356552124, "learning_rate": 3.3716738401402547e-07, "loss": 0.0129, "step": 77190 }, { "grad_norm": 0.2507683336734772, "learning_rate": 3.3477541270241495e-07, "loss": 0.0132, "step": 77200 }, { "grad_norm": 0.2437649965286255, "learning_rate": 3.3239192781934215e-07, "loss": 0.0146, "step": 77210 }, { "grad_norm": 0.20819492638111115, "learning_rate": 3.3001692977207563e-07, "loss": 0.0136, "step": 77220 }, { "grad_norm": 0.22300051152706146, "learning_rate": 3.276504189664409e-07, "loss": 0.0157, "step": 77230 }, { "grad_norm": 0.1907591074705124, "learning_rate": 3.252923958068088e-07, "loss": 0.0164, "step": 77240 }, { "grad_norm": 0.14679448306560516, "learning_rate": 3.2294286069609046e-07, "loss": 0.0147, "step": 77250 }, { "grad_norm": 0.1878879815340042, "learning_rate": 3.2060181403577583e-07, "loss": 0.0116, "step": 77260 }, { "grad_norm": 0.31444627046585083, "learning_rate": 3.1826925622587823e-07, "loss": 0.0133, "step": 77270 }, { "grad_norm": 0.21137183904647827, "learning_rate": 3.159451876649622e-07, "loss": 0.0146, "step": 77280 }, { "grad_norm": 0.30637088418006897, "learning_rate": 3.1362960875015444e-07, "loss": 0.0111, "step": 77290 }, { "grad_norm": 0.29697346687316895, "learning_rate": 3.1132251987711637e-07, "loss": 0.0182, "step": 77300 }, { "grad_norm": 0.2703283429145813, "learning_rate": 3.0902392144007695e-07, "loss": 0.0163, "step": 77310 }, { "grad_norm": 0.4134961664676666, "learning_rate": 3.0673381383179436e-07, "loss": 0.0148, "step": 77320 }, { "grad_norm": 0.23619477450847626, "learning_rate": 3.0445219744358875e-07, "loss": 0.0129, "step": 77330 }, { "grad_norm": 0.19956858456134796, "learning_rate": 3.021790726653262e-07, "loss": 0.0201, "step": 77340 }, { "grad_norm": 0.3830779492855072, "learning_rate": 2.9991443988542366e-07, "loss": 0.018, "step": 77350 }, { "grad_norm": 0.1733531802892685, "learning_rate": 2.976582994908439e-07, "loss": 0.0123, "step": 77360 }, { "grad_norm": 0.3347592055797577, "learning_rate": 2.954106518671007e-07, "loss": 0.0177, "step": 77370 }, { "grad_norm": 0.1661069393157959, "learning_rate": 2.9317149739825356e-07, "loss": 0.0122, "step": 77380 }, { "grad_norm": 0.3947153389453888, "learning_rate": 2.909408364669075e-07, "loss": 0.0212, "step": 77390 }, { "grad_norm": 0.4722351133823395, "learning_rate": 2.887186694542299e-07, "loss": 0.0156, "step": 77400 }, { "grad_norm": 0.1784670352935791, "learning_rate": 2.865049967399225e-07, "loss": 0.0135, "step": 77410 }, { "grad_norm": 0.17216211557388306, "learning_rate": 2.842998187022439e-07, "loss": 0.0136, "step": 77420 }, { "grad_norm": 0.3223273456096649, "learning_rate": 2.8210313571800374e-07, "loss": 0.0177, "step": 77430 }, { "grad_norm": 0.43822774291038513, "learning_rate": 2.7991494816255184e-07, "loss": 0.0146, "step": 77440 }, { "grad_norm": 0.3138616681098938, "learning_rate": 2.777352564097779e-07, "loss": 0.0161, "step": 77450 }, { "grad_norm": 0.25784721970558167, "learning_rate": 2.755640608321508e-07, "loss": 0.0145, "step": 77460 }, { "grad_norm": 0.2914261817932129, "learning_rate": 2.73401361800657e-07, "loss": 0.0164, "step": 77470 }, { "grad_norm": 0.3528920114040375, "learning_rate": 2.7124715968484537e-07, "loss": 0.0188, "step": 77480 }, { "grad_norm": 0.18906252086162567, "learning_rate": 2.691014548528104e-07, "loss": 0.0142, "step": 77490 }, { "grad_norm": 0.14240583777427673, "learning_rate": 2.669642476711864e-07, "loss": 0.0173, "step": 77500 }, { "grad_norm": 0.17577897012233734, "learning_rate": 2.648355385051815e-07, "loss": 0.015, "step": 77510 }, { "grad_norm": 0.3327033817768097, "learning_rate": 2.627153277185157e-07, "loss": 0.0188, "step": 77520 }, { "grad_norm": 0.2742299437522888, "learning_rate": 2.606036156734881e-07, "loss": 0.0145, "step": 77530 }, { "grad_norm": 0.26142483949661255, "learning_rate": 2.5850040273092127e-07, "loss": 0.0175, "step": 77540 }, { "grad_norm": 0.38295432925224304, "learning_rate": 2.5640568925020536e-07, "loss": 0.0231, "step": 77550 }, { "grad_norm": 0.1396450251340866, "learning_rate": 2.5431947558927083e-07, "loss": 0.0173, "step": 77560 }, { "grad_norm": 0.24684301018714905, "learning_rate": 2.522417621045825e-07, "loss": 0.0141, "step": 77570 }, { "grad_norm": 0.23203332722187042, "learning_rate": 2.501725491511786e-07, "loss": 0.0192, "step": 77580 }, { "grad_norm": 0.26832398772239685, "learning_rate": 2.481118370826263e-07, "loss": 0.0141, "step": 77590 }, { "grad_norm": 0.3100508451461792, "learning_rate": 2.4605962625104393e-07, "loss": 0.0155, "step": 77600 }, { "grad_norm": 0.289106547832489, "learning_rate": 2.440159170070955e-07, "loss": 0.0244, "step": 77610 }, { "grad_norm": 0.3733128011226654, "learning_rate": 2.41980709699996e-07, "loss": 0.014, "step": 77620 }, { "grad_norm": 0.2614688277244568, "learning_rate": 2.3995400467751174e-07, "loss": 0.0164, "step": 77630 }, { "grad_norm": 0.2574116885662079, "learning_rate": 2.3793580228594902e-07, "loss": 0.0219, "step": 77640 }, { "grad_norm": 0.3068291246891022, "learning_rate": 2.3592610287015982e-07, "loss": 0.016, "step": 77650 }, { "grad_norm": 0.4625388979911804, "learning_rate": 2.3392490677354718e-07, "loss": 0.0163, "step": 77660 }, { "grad_norm": 0.21700602769851685, "learning_rate": 2.3193221433806533e-07, "loss": 0.0135, "step": 77670 }, { "grad_norm": 0.2493399828672409, "learning_rate": 2.299480259042086e-07, "loss": 0.0176, "step": 77680 }, { "grad_norm": 0.16599693894386292, "learning_rate": 2.2797234181102244e-07, "loss": 0.014, "step": 77690 }, { "grad_norm": 0.19753949344158173, "learning_rate": 2.2600516239609238e-07, "loss": 0.0136, "step": 77700 }, { "grad_norm": 0.20567713677883148, "learning_rate": 2.240464879955606e-07, "loss": 0.0191, "step": 77710 }, { "grad_norm": 0.21695993840694427, "learning_rate": 2.2209631894410387e-07, "loss": 0.0175, "step": 77720 }, { "grad_norm": 0.32029610872268677, "learning_rate": 2.2015465557496117e-07, "loss": 0.0156, "step": 77730 }, { "grad_norm": 0.4406692385673523, "learning_rate": 2.1822149821990602e-07, "loss": 0.0184, "step": 77740 }, { "grad_norm": 0.3780105412006378, "learning_rate": 2.1629684720926303e-07, "loss": 0.0212, "step": 77750 }, { "grad_norm": 0.22353625297546387, "learning_rate": 2.1438070287189693e-07, "loss": 0.017, "step": 77760 }, { "grad_norm": 0.2784854471683502, "learning_rate": 2.1247306553523472e-07, "loss": 0.0169, "step": 77770 }, { "grad_norm": 0.32866573333740234, "learning_rate": 2.105739355252323e-07, "loss": 0.0192, "step": 77780 }, { "grad_norm": 0.1677897721529007, "learning_rate": 2.0868331316639678e-07, "loss": 0.0115, "step": 77790 }, { "grad_norm": 0.2475966215133667, "learning_rate": 2.0680119878179193e-07, "loss": 0.0137, "step": 77800 }, { "grad_norm": 0.3803560137748718, "learning_rate": 2.0492759269301054e-07, "loss": 0.0213, "step": 77810 }, { "grad_norm": 0.3079056143760681, "learning_rate": 2.0306249522021315e-07, "loss": 0.013, "step": 77820 }, { "grad_norm": 0.1934896558523178, "learning_rate": 2.0120590668207816e-07, "loss": 0.015, "step": 77830 }, { "grad_norm": 0.23894371092319489, "learning_rate": 1.993578273958574e-07, "loss": 0.0158, "step": 77840 }, { "grad_norm": 0.2135193645954132, "learning_rate": 1.975182576773371e-07, "loss": 0.0169, "step": 77850 }, { "grad_norm": 0.37617674469947815, "learning_rate": 1.9568719784083812e-07, "loss": 0.0183, "step": 77860 }, { "grad_norm": 0.22230303287506104, "learning_rate": 1.9386464819924898e-07, "loss": 0.019, "step": 77870 }, { "grad_norm": 0.2846488952636719, "learning_rate": 1.9205060906399285e-07, "loss": 0.0148, "step": 77880 }, { "grad_norm": 0.17706547677516937, "learning_rate": 1.9024508074503845e-07, "loss": 0.0195, "step": 77890 }, { "grad_norm": 0.21057352423667908, "learning_rate": 1.8844806355089452e-07, "loss": 0.0271, "step": 77900 }, { "grad_norm": 0.2009446769952774, "learning_rate": 1.8665955778863208e-07, "loss": 0.024, "step": 77910 }, { "grad_norm": 0.19536744058132172, "learning_rate": 1.848795637638512e-07, "loss": 0.0138, "step": 77920 }, { "grad_norm": 0.3889157176017761, "learning_rate": 1.831080817807085e-07, "loss": 0.0225, "step": 77930 }, { "grad_norm": 0.25851696729660034, "learning_rate": 1.813451121418952e-07, "loss": 0.0279, "step": 77940 }, { "grad_norm": 0.32750919461250305, "learning_rate": 1.795906551486648e-07, "loss": 0.0143, "step": 77950 }, { "grad_norm": 0.31018245220184326, "learning_rate": 1.7784471110079414e-07, "loss": 0.0172, "step": 77960 }, { "grad_norm": 0.26844486594200134, "learning_rate": 1.7610728029662793e-07, "loss": 0.0164, "step": 77970 }, { "grad_norm": 0.2122388780117035, "learning_rate": 1.7437836303303979e-07, "loss": 0.0132, "step": 77980 }, { "grad_norm": 0.222346231341362, "learning_rate": 1.726579596054545e-07, "loss": 0.0112, "step": 77990 }, { "grad_norm": 0.4127156734466553, "learning_rate": 1.7094607030784803e-07, "loss": 0.0225, "step": 78000 }, { "grad_norm": 0.2773822844028473, "learning_rate": 1.6924269543272532e-07, "loss": 0.0165, "step": 78010 }, { "grad_norm": 0.2004968822002411, "learning_rate": 1.6754783527115348e-07, "loss": 0.012, "step": 78020 }, { "grad_norm": 0.10728754848241806, "learning_rate": 1.6586149011273422e-07, "loss": 0.0146, "step": 78030 }, { "grad_norm": 0.17978216707706451, "learning_rate": 1.6418366024562038e-07, "loss": 0.0238, "step": 78040 }, { "grad_norm": 0.1923259049654007, "learning_rate": 1.6251434595651037e-07, "loss": 0.0175, "step": 78050 }, { "grad_norm": 0.19355958700180054, "learning_rate": 1.608535475306372e-07, "loss": 0.019, "step": 78060 }, { "grad_norm": 0.23274879157543182, "learning_rate": 1.5920126525179048e-07, "loss": 0.014, "step": 78070 }, { "grad_norm": 0.21527494490146637, "learning_rate": 1.5755749940229435e-07, "loss": 0.0146, "step": 78080 }, { "grad_norm": 0.13493476808071136, "learning_rate": 1.559222502630353e-07, "loss": 0.0137, "step": 78090 }, { "grad_norm": 0.2564123272895813, "learning_rate": 1.5429551811341757e-07, "loss": 0.02, "step": 78100 }, { "grad_norm": 0.27660197019577026, "learning_rate": 1.5267730323141882e-07, "loss": 0.0219, "step": 78110 }, { "grad_norm": 0.21654240787029266, "learning_rate": 1.5106760589353454e-07, "loss": 0.0134, "step": 78120 }, { "grad_norm": 0.3372347056865692, "learning_rate": 1.4946642637483355e-07, "loss": 0.0126, "step": 78130 }, { "grad_norm": 0.476730614900589, "learning_rate": 1.4787376494889703e-07, "loss": 0.0167, "step": 78140 }, { "grad_norm": 0.22028295695781708, "learning_rate": 1.4628962188787955e-07, "loss": 0.0146, "step": 78150 }, { "grad_norm": 0.31120505928993225, "learning_rate": 1.4471399746247006e-07, "loss": 0.0158, "step": 78160 }, { "grad_norm": 0.2398531287908554, "learning_rate": 1.4314689194188103e-07, "loss": 0.0132, "step": 78170 }, { "grad_norm": 0.44421109557151794, "learning_rate": 1.4158830559390933e-07, "loss": 0.0208, "step": 78180 }, { "grad_norm": 0.15755784511566162, "learning_rate": 1.4003823868486422e-07, "loss": 0.0311, "step": 78190 }, { "grad_norm": 0.2723582983016968, "learning_rate": 1.3849669147960598e-07, "loss": 0.0198, "step": 78200 }, { "grad_norm": 0.2541719377040863, "learning_rate": 1.3696366424155726e-07, "loss": 0.0145, "step": 78210 }, { "grad_norm": 0.4741217792034149, "learning_rate": 1.35439157232653e-07, "loss": 0.0239, "step": 78220 }, { "grad_norm": 0.24858951568603516, "learning_rate": 1.3392317071340144e-07, "loss": 0.0187, "step": 78230 }, { "grad_norm": 0.40596309304237366, "learning_rate": 1.3241570494283984e-07, "loss": 0.0156, "step": 78240 }, { "grad_norm": 0.3818765878677368, "learning_rate": 1.3091676017855103e-07, "loss": 0.0202, "step": 78250 }, { "grad_norm": 0.36101463437080383, "learning_rate": 1.2942633667666904e-07, "loss": 0.0179, "step": 78260 }, { "grad_norm": 0.21344012022018433, "learning_rate": 1.2794443469185679e-07, "loss": 0.0194, "step": 78270 }, { "grad_norm": 0.25909972190856934, "learning_rate": 1.2647105447734508e-07, "loss": 0.0144, "step": 78280 }, { "grad_norm": 0.18665634095668793, "learning_rate": 1.2500619628488254e-07, "loss": 0.0139, "step": 78290 }, { "grad_norm": 0.2433728575706482, "learning_rate": 1.2354986036477446e-07, "loss": 0.0202, "step": 78300 }, { "grad_norm": 0.3332197368144989, "learning_rate": 1.221020469658718e-07, "loss": 0.017, "step": 78310 }, { "grad_norm": 0.14160899817943573, "learning_rate": 1.2066275633556556e-07, "loss": 0.0144, "step": 78320 }, { "grad_norm": 0.30127763748168945, "learning_rate": 1.192319887197979e-07, "loss": 0.0167, "step": 78330 }, { "grad_norm": 0.23708564043045044, "learning_rate": 1.1780974436303438e-07, "loss": 0.0122, "step": 78340 }, { "grad_norm": 0.38797399401664734, "learning_rate": 1.1639602350830836e-07, "loss": 0.0203, "step": 78350 }, { "grad_norm": 0.29192882776260376, "learning_rate": 1.1499082639718217e-07, "loss": 0.0175, "step": 78360 }, { "grad_norm": 0.1894802302122116, "learning_rate": 1.1359415326976374e-07, "loss": 0.0117, "step": 78370 }, { "grad_norm": 0.24644558131694794, "learning_rate": 1.1220600436470663e-07, "loss": 0.0138, "step": 78380 }, { "grad_norm": 0.2749774158000946, "learning_rate": 1.1082637991920996e-07, "loss": 0.0216, "step": 78390 }, { "grad_norm": 0.2622465193271637, "learning_rate": 1.0945528016901851e-07, "loss": 0.0176, "step": 78400 }, { "grad_norm": 0.21997424960136414, "learning_rate": 1.0809270534840599e-07, "loss": 0.0153, "step": 78410 }, { "grad_norm": 0.18095150589942932, "learning_rate": 1.067386556902028e-07, "loss": 0.0143, "step": 78420 }, { "grad_norm": 0.16293862462043762, "learning_rate": 1.05393131425785e-07, "loss": 0.0121, "step": 78430 }, { "grad_norm": 0.33009976148605347, "learning_rate": 1.0405613278505199e-07, "loss": 0.0157, "step": 78440 }, { "grad_norm": 0.27784237265586853, "learning_rate": 1.027276599964766e-07, "loss": 0.0141, "step": 78450 }, { "grad_norm": 0.47461646795272827, "learning_rate": 1.0140771328704391e-07, "loss": 0.0144, "step": 78460 }, { "grad_norm": 0.21866023540496826, "learning_rate": 1.0009629288231237e-07, "loss": 0.0159, "step": 78470 }, { "grad_norm": 0.1497015804052353, "learning_rate": 9.879339900635276e-08, "loss": 0.0211, "step": 78480 }, { "grad_norm": 0.4029284119606018, "learning_rate": 9.74990318817981e-08, "loss": 0.022, "step": 78490 }, { "grad_norm": 0.33083873987197876, "learning_rate": 9.621319172982701e-08, "loss": 0.0176, "step": 78500 }, { "grad_norm": 0.20096002519130707, "learning_rate": 9.493587877015264e-08, "loss": 0.0152, "step": 78510 }, { "grad_norm": 0.3415110409259796, "learning_rate": 9.366709322102263e-08, "loss": 0.0136, "step": 78520 }, { "grad_norm": 0.3054371178150177, "learning_rate": 9.240683529924688e-08, "loss": 0.0181, "step": 78530 }, { "grad_norm": 0.17184144258499146, "learning_rate": 9.115510522016979e-08, "loss": 0.0133, "step": 78540 }, { "grad_norm": 0.28832823038101196, "learning_rate": 8.991190319767583e-08, "loss": 0.0133, "step": 78550 }, { "grad_norm": 0.09601807594299316, "learning_rate": 8.867722944419509e-08, "loss": 0.0133, "step": 78560 }, { "grad_norm": 0.33299699425697327, "learning_rate": 8.745108417069214e-08, "loss": 0.0129, "step": 78570 }, { "grad_norm": 0.1617569774389267, "learning_rate": 8.623346758669381e-08, "loss": 0.0137, "step": 78580 }, { "grad_norm": 0.22788670659065247, "learning_rate": 8.502437990025037e-08, "loss": 0.0164, "step": 78590 }, { "grad_norm": 0.4676435887813568, "learning_rate": 8.382382131795764e-08, "loss": 0.0144, "step": 78600 }, { "grad_norm": 0.3290773928165436, "learning_rate": 8.263179204496818e-08, "loss": 0.0248, "step": 78610 }, { "grad_norm": 0.1836480349302292, "learning_rate": 8.144829228496354e-08, "loss": 0.0152, "step": 78620 }, { "grad_norm": 0.2874489426612854, "learning_rate": 8.027332224016526e-08, "loss": 0.0179, "step": 78630 }, { "grad_norm": 0.17362073063850403, "learning_rate": 7.910688211135164e-08, "loss": 0.0175, "step": 78640 }, { "grad_norm": 0.23525509238243103, "learning_rate": 7.794897209783546e-08, "loss": 0.02, "step": 78650 }, { "grad_norm": 0.16933242976665497, "learning_rate": 7.679959239746403e-08, "loss": 0.0195, "step": 78660 }, { "grad_norm": 0.18156221508979797, "learning_rate": 7.565874320664135e-08, "loss": 0.0145, "step": 78670 }, { "grad_norm": 0.3354067802429199, "learning_rate": 7.45264247203059e-08, "loss": 0.0155, "step": 78680 }, { "grad_norm": 0.1854788064956665, "learning_rate": 7.340263713194184e-08, "loss": 0.0141, "step": 78690 }, { "grad_norm": 0.29155969619750977, "learning_rate": 7.228738063356777e-08, "loss": 0.0149, "step": 78700 }, { "grad_norm": 0.5217209458351135, "learning_rate": 7.118065541575903e-08, "loss": 0.0196, "step": 78710 }, { "grad_norm": 0.20333586633205414, "learning_rate": 7.008246166761435e-08, "loss": 0.0125, "step": 78720 }, { "grad_norm": 0.229765385389328, "learning_rate": 6.899279957679472e-08, "loss": 0.0161, "step": 78730 }, { "grad_norm": 0.3258533775806427, "learning_rate": 6.791166932949011e-08, "loss": 0.0167, "step": 78740 }, { "grad_norm": 0.20833338797092438, "learning_rate": 6.683907111043608e-08, "loss": 0.0158, "step": 78750 }, { "grad_norm": 0.15425653755664825, "learning_rate": 6.577500510290824e-08, "loss": 0.0149, "step": 78760 }, { "grad_norm": 0.2784014940261841, "learning_rate": 6.47194714887278e-08, "loss": 0.0299, "step": 78770 }, { "grad_norm": 0.2893034517765045, "learning_rate": 6.367247044825608e-08, "loss": 0.0169, "step": 78780 }, { "grad_norm": 0.29933834075927734, "learning_rate": 6.263400216039994e-08, "loss": 0.0154, "step": 78790 }, { "grad_norm": 0.37594202160835266, "learning_rate": 6.160406680260078e-08, "loss": 0.0345, "step": 78800 }, { "grad_norm": 0.22510294616222382, "learning_rate": 6.058266455084561e-08, "loss": 0.0162, "step": 78810 }, { "grad_norm": 0.351658433675766, "learning_rate": 5.956979557967257e-08, "loss": 0.014, "step": 78820 }, { "grad_norm": 0.32228589057922363, "learning_rate": 5.856546006214325e-08, "loss": 0.0203, "step": 78830 }, { "grad_norm": 0.30937451124191284, "learning_rate": 5.7569658169881466e-08, "loss": 0.0177, "step": 78840 }, { "grad_norm": 0.2008385956287384, "learning_rate": 5.658239007303445e-08, "loss": 0.016, "step": 78850 }, { "grad_norm": 0.23379649221897125, "learning_rate": 5.560365594030059e-08, "loss": 0.0154, "step": 78860 }, { "grad_norm": 0.2485516369342804, "learning_rate": 5.463345593891833e-08, "loss": 0.0142, "step": 78870 }, { "grad_norm": 0.15967904031276703, "learning_rate": 5.367179023467173e-08, "loss": 0.0163, "step": 78880 }, { "grad_norm": 0.21849006414413452, "learning_rate": 5.271865899187378e-08, "loss": 0.0132, "step": 78890 }, { "grad_norm": 0.34209775924682617, "learning_rate": 5.177406237340532e-08, "loss": 0.0263, "step": 78900 }, { "grad_norm": 0.17302757501602173, "learning_rate": 5.083800054065946e-08, "loss": 0.0155, "step": 78910 }, { "grad_norm": 0.4039648473262787, "learning_rate": 4.991047365358603e-08, "loss": 0.0149, "step": 78920 }, { "grad_norm": 0.3302052617073059, "learning_rate": 4.899148187067493e-08, "loss": 0.0238, "step": 78930 }, { "grad_norm": 0.1561989188194275, "learning_rate": 4.808102534895609e-08, "loss": 0.0141, "step": 78940 }, { "grad_norm": 0.24716341495513916, "learning_rate": 4.717910424400507e-08, "loss": 0.0148, "step": 78950 }, { "grad_norm": 0.3680934011936188, "learning_rate": 4.628571870993193e-08, "loss": 0.0181, "step": 78960 }, { "grad_norm": 0.4067935347557068, "learning_rate": 4.5400868899392324e-08, "loss": 0.016, "step": 78970 }, { "grad_norm": 0.31102073192596436, "learning_rate": 4.4524554963576436e-08, "loss": 0.0198, "step": 78980 }, { "grad_norm": 0.33236774802207947, "learning_rate": 4.36567770522367e-08, "loss": 0.0191, "step": 78990 }, { "grad_norm": 0.26949429512023926, "learning_rate": 4.279753531364339e-08, "loss": 0.0126, "step": 79000 }, { "grad_norm": 0.36219000816345215, "learning_rate": 4.1946829894617955e-08, "loss": 0.0206, "step": 79010 }, { "grad_norm": 0.29801642894744873, "learning_rate": 4.11046609405219e-08, "loss": 0.0216, "step": 79020 }, { "grad_norm": 0.1461934894323349, "learning_rate": 4.027102859526233e-08, "loss": 0.0105, "step": 79030 }, { "grad_norm": 0.3418191373348236, "learning_rate": 3.9445933001280856e-08, "loss": 0.0171, "step": 79040 }, { "grad_norm": 0.15484492480754852, "learning_rate": 3.8629374299564704e-08, "loss": 0.0168, "step": 79050 }, { "grad_norm": 0.21222157776355743, "learning_rate": 3.78213526296467e-08, "loss": 0.0221, "step": 79060 }, { "grad_norm": 0.21940068900585175, "learning_rate": 3.702186812958308e-08, "loss": 0.0167, "step": 79070 }, { "grad_norm": 0.32391542196273804, "learning_rate": 3.623092093599789e-08, "loss": 0.0158, "step": 79080 }, { "grad_norm": 0.4243459403514862, "learning_rate": 3.544851118403303e-08, "loss": 0.0147, "step": 79090 }, { "grad_norm": 0.36368632316589355, "learning_rate": 3.46746390073871e-08, "loss": 0.0148, "step": 79100 }, { "grad_norm": 0.2879585921764374, "learning_rate": 3.390930453829322e-08, "loss": 0.0205, "step": 79110 }, { "grad_norm": 0.25402992963790894, "learning_rate": 3.3152507907519005e-08, "loss": 0.0137, "step": 79120 }, { "grad_norm": 0.19237695634365082, "learning_rate": 3.240424924438323e-08, "loss": 0.0162, "step": 79130 }, { "grad_norm": 0.33294644951820374, "learning_rate": 3.1664528676750295e-08, "loss": 0.0148, "step": 79140 }, { "grad_norm": 0.24470897018909454, "learning_rate": 3.093334633100797e-08, "loss": 0.021, "step": 79150 }, { "grad_norm": 0.34802600741386414, "learning_rate": 3.021070233210077e-08, "loss": 0.0176, "step": 79160 }, { "grad_norm": 0.4046611785888672, "learning_rate": 2.9496596803507692e-08, "loss": 0.0142, "step": 79170 }, { "grad_norm": 0.26175397634506226, "learning_rate": 2.879102986725335e-08, "loss": 0.0175, "step": 79180 }, { "grad_norm": 0.1378336399793625, "learning_rate": 2.809400164389131e-08, "loss": 0.0133, "step": 79190 }, { "grad_norm": 0.2632203996181488, "learning_rate": 2.740551225253185e-08, "loss": 0.0154, "step": 79200 }, { "grad_norm": 0.2624439001083374, "learning_rate": 2.6725561810819754e-08, "loss": 0.0174, "step": 79210 }, { "grad_norm": 0.4259192943572998, "learning_rate": 2.605415043493431e-08, "loss": 0.0132, "step": 79220 }, { "grad_norm": 0.25712475180625916, "learning_rate": 2.5391278239605965e-08, "loss": 0.0158, "step": 79230 }, { "grad_norm": 0.401471883058548, "learning_rate": 2.4736945338094118e-08, "loss": 0.0169, "step": 79240 }, { "grad_norm": 0.20857858657836914, "learning_rate": 2.4091151842214887e-08, "loss": 0.0109, "step": 79250 }, { "grad_norm": 0.24941426515579224, "learning_rate": 2.3453897862318884e-08, "loss": 0.0133, "step": 79260 }, { "grad_norm": 0.1626323014497757, "learning_rate": 2.2825183507285686e-08, "loss": 0.017, "step": 79270 }, { "grad_norm": 0.1657794713973999, "learning_rate": 2.220500888455157e-08, "loss": 0.018, "step": 79280 }, { "grad_norm": 0.20393748581409454, "learning_rate": 2.1593374100081777e-08, "loss": 0.0197, "step": 79290 }, { "grad_norm": 0.2534448206424713, "learning_rate": 2.09902792583927e-08, "loss": 0.0219, "step": 79300 }, { "grad_norm": 0.23502720892429352, "learning_rate": 2.0395724462540788e-08, "loss": 0.0166, "step": 79310 }, { "grad_norm": 0.5971697568893433, "learning_rate": 1.9809709814111453e-08, "loss": 0.0157, "step": 79320 }, { "grad_norm": 0.394707590341568, "learning_rate": 1.923223541324126e-08, "loss": 0.0256, "step": 79330 }, { "grad_norm": 0.19164367020130157, "learning_rate": 1.8663301358606833e-08, "loss": 0.0202, "step": 79340 }, { "grad_norm": 0.14659716188907623, "learning_rate": 1.8102907747419295e-08, "loss": 0.0257, "step": 79350 }, { "grad_norm": 0.16700445115566254, "learning_rate": 1.7551054675435385e-08, "loss": 0.0171, "step": 79360 }, { "grad_norm": 0.1606929898262024, "learning_rate": 1.7007742236957447e-08, "loss": 0.0185, "step": 79370 }, { "grad_norm": 0.3780255615711212, "learning_rate": 1.647297052481678e-08, "loss": 0.0198, "step": 79380 }, { "grad_norm": 0.32303404808044434, "learning_rate": 1.5946739630390285e-08, "loss": 0.0216, "step": 79390 }, { "grad_norm": 0.2628069221973419, "learning_rate": 1.5429049643606032e-08, "loss": 0.0276, "step": 79400 }, { "grad_norm": 0.1473257690668106, "learning_rate": 1.4919900652909935e-08, "loss": 0.0183, "step": 79410 }, { "grad_norm": 0.1603219211101532, "learning_rate": 1.4419292745310175e-08, "loss": 0.0204, "step": 79420 }, { "grad_norm": 0.147111713886261, "learning_rate": 1.3927226006343885e-08, "loss": 0.0137, "step": 79430 }, { "grad_norm": 0.2906664311885834, "learning_rate": 1.3443700520093805e-08, "loss": 0.021, "step": 79440 }, { "grad_norm": 0.23740154504776, "learning_rate": 1.296871636917718e-08, "loss": 0.0142, "step": 79450 }, { "grad_norm": 0.48596224188804626, "learning_rate": 1.2502273634762419e-08, "loss": 0.017, "step": 79460 }, { "grad_norm": 0.46360963582992554, "learning_rate": 1.2044372396546876e-08, "loss": 0.0161, "step": 79470 }, { "grad_norm": 0.31101861596107483, "learning_rate": 1.1595012732773524e-08, "loss": 0.0185, "step": 79480 }, { "grad_norm": 0.1272638738155365, "learning_rate": 1.1154194720225386e-08, "loss": 0.0171, "step": 79490 }, { "grad_norm": 0.28491777181625366, "learning_rate": 1.0721918434231093e-08, "loss": 0.0135, "step": 79500 }, { "grad_norm": 0.2817208170890808, "learning_rate": 1.0298183948648232e-08, "loss": 0.0145, "step": 79510 }, { "grad_norm": 0.29850253462791443, "learning_rate": 9.88299133588555e-09, "loss": 0.0182, "step": 79520 }, { "grad_norm": 0.22690963745117188, "learning_rate": 9.476340666891847e-09, "loss": 0.0218, "step": 79530 }, { "grad_norm": 0.11509546637535095, "learning_rate": 9.078232011139332e-09, "loss": 0.0138, "step": 79540 }, { "grad_norm": 0.3423186242580414, "learning_rate": 8.68866543666802e-09, "loss": 0.0145, "step": 79550 }, { "grad_norm": 0.2232813537120819, "learning_rate": 8.307641010035782e-09, "loss": 0.0161, "step": 79560 }, { "grad_norm": 0.16768577694892883, "learning_rate": 7.93515879635165e-09, "loss": 0.0142, "step": 79570 }, { "grad_norm": 0.2111569195985794, "learning_rate": 7.571218859264706e-09, "loss": 0.0152, "step": 79580 }, { "grad_norm": 0.34918442368507385, "learning_rate": 7.215821260958544e-09, "loss": 0.0188, "step": 79590 }, { "grad_norm": 0.17950581014156342, "learning_rate": 6.868966062162363e-09, "loss": 0.016, "step": 79600 }, { "grad_norm": 0.17146535217761993, "learning_rate": 6.530653322145419e-09, "loss": 0.0145, "step": 79610 }, { "grad_norm": 0.2473909705877304, "learning_rate": 6.200883098717025e-09, "loss": 0.0099, "step": 79620 }, { "grad_norm": 0.36363378167152405, "learning_rate": 5.879655448226551e-09, "loss": 0.0242, "step": 79630 }, { "grad_norm": 0.33896854519844055, "learning_rate": 5.566970425557872e-09, "loss": 0.014, "step": 79640 }, { "grad_norm": 0.36923450231552124, "learning_rate": 5.26282808414047e-09, "loss": 0.012, "step": 79650 }, { "grad_norm": 0.1781793087720871, "learning_rate": 4.967228475949437e-09, "loss": 0.0183, "step": 79660 }, { "grad_norm": 0.3218682110309601, "learning_rate": 4.680171651494369e-09, "loss": 0.0146, "step": 79670 }, { "grad_norm": 0.26850762963294983, "learning_rate": 4.40165765981937e-09, "loss": 0.0147, "step": 79680 }, { "grad_norm": 0.10137514770030975, "learning_rate": 4.131686548519698e-09, "loss": 0.0121, "step": 79690 }, { "grad_norm": 0.1839645802974701, "learning_rate": 3.8702583637251214e-09, "loss": 0.0131, "step": 79700 }, { "grad_norm": 0.15573911368846893, "learning_rate": 3.617373150105463e-09, "loss": 0.0158, "step": 79710 }, { "grad_norm": 0.19181248545646667, "learning_rate": 3.3730309508706036e-09, "loss": 0.0145, "step": 79720 }, { "grad_norm": 0.28205743432044983, "learning_rate": 3.137231807781582e-09, "loss": 0.0127, "step": 79730 }, { "grad_norm": 0.10607615113258362, "learning_rate": 2.9099757611172894e-09, "loss": 0.0175, "step": 79740 }, { "grad_norm": 0.3072241246700287, "learning_rate": 2.6912628497133275e-09, "loss": 0.0151, "step": 79750 }, { "grad_norm": 0.6532378196716309, "learning_rate": 2.481093110945354e-09, "loss": 0.0173, "step": 79760 }, { "grad_norm": 0.2145971953868866, "learning_rate": 2.2794665807235327e-09, "loss": 0.0152, "step": 79770 }, { "grad_norm": 0.1950969099998474, "learning_rate": 2.0863832934980843e-09, "loss": 0.0128, "step": 79780 }, { "grad_norm": 0.31209269165992737, "learning_rate": 1.901843282264837e-09, "loss": 0.0212, "step": 79790 }, { "grad_norm": 0.15475918352603912, "learning_rate": 1.7258465785541245e-09, "loss": 0.0168, "step": 79800 }, { "grad_norm": 0.2457994818687439, "learning_rate": 1.5583932124418887e-09, "loss": 0.0138, "step": 79810 }, { "grad_norm": 0.2318040132522583, "learning_rate": 1.3994832125441282e-09, "loss": 0.0138, "step": 79820 }, { "grad_norm": 0.27512383460998535, "learning_rate": 1.2491166060057958e-09, "loss": 0.0166, "step": 79830 }, { "grad_norm": 0.29202356934547424, "learning_rate": 1.1072934185230032e-09, "loss": 0.0165, "step": 79840 }, { "grad_norm": 0.47649675607681274, "learning_rate": 9.740136743319195e-10, "loss": 0.0206, "step": 79850 }, { "grad_norm": 0.38062286376953125, "learning_rate": 8.492773962087696e-10, "loss": 0.0148, "step": 79860 }, { "grad_norm": 0.278562068939209, "learning_rate": 7.330846054587337e-10, "loss": 0.0179, "step": 79870 }, { "grad_norm": 0.18849030137062073, "learning_rate": 6.254353219492526e-10, "loss": 0.0153, "step": 79880 }, { "grad_norm": 0.4613894522190094, "learning_rate": 5.263295640600685e-10, "loss": 0.0118, "step": 79890 }, { "grad_norm": 0.16556547582149506, "learning_rate": 4.357673487387359e-10, "loss": 0.015, "step": 79900 }, { "grad_norm": 0.2511727213859558, "learning_rate": 3.537486914506616e-10, "loss": 0.014, "step": 79910 }, { "grad_norm": 0.17387272417545319, "learning_rate": 2.8027360621241117e-10, "loss": 0.0158, "step": 79920 }, { "grad_norm": 0.16988982260227203, "learning_rate": 2.153421055806071e-10, "loss": 0.0126, "step": 79930 }, { "grad_norm": 0.216630756855011, "learning_rate": 1.5895420064637733e-10, "loss": 0.0174, "step": 79940 }, { "grad_norm": 0.2052004337310791, "learning_rate": 1.1110990105200891e-10, "loss": 0.0139, "step": 79950 }, { "grad_norm": 0.41070348024368286, "learning_rate": 7.180921496874326e-11, "loss": 0.0222, "step": 79960 }, { "grad_norm": 0.36656156182289124, "learning_rate": 4.10521491134297e-11, "loss": 0.0125, "step": 79970 }, { "grad_norm": 0.15284347534179688, "learning_rate": 1.883870874297422e-11, "loss": 0.0158, "step": 79980 }, { "grad_norm": 0.31738725304603577, "learning_rate": 5.1688976432373584e-12, "loss": 0.0158, "step": 79990 }, { "grad_norm": 0.3925632834434509, "learning_rate": 4.271816234080461e-14, "loss": 0.0123, "step": 80000 } ], "logging_steps": 10, "max_steps": 80000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 80000, "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 }