{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 80000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 1.2053630352020264, "learning_rate": 2.25e-07, "loss": 1.4342, "step": 10 }, { "grad_norm": 0.9269222617149353, "learning_rate": 4.75e-07, "loss": 1.4202, "step": 20 }, { "grad_norm": 1.3775074481964111, "learning_rate": 7.25e-07, "loss": 1.4495, "step": 30 }, { "grad_norm": 1.1235616207122803, "learning_rate": 9.75e-07, "loss": 1.4157, "step": 40 }, { "grad_norm": 0.8801957368850708, "learning_rate": 1.2250000000000001e-06, "loss": 1.4119, "step": 50 }, { "grad_norm": 1.3744202852249146, "learning_rate": 1.475e-06, "loss": 1.3902, "step": 60 }, { "grad_norm": 1.7822085618972778, "learning_rate": 1.7250000000000002e-06, "loss": 1.247, "step": 70 }, { "grad_norm": 1.8199536800384521, "learning_rate": 1.975e-06, "loss": 1.2077, "step": 80 }, { "grad_norm": 1.6485875844955444, "learning_rate": 2.225e-06, "loss": 1.1986, "step": 90 }, { "grad_norm": 3.090141534805298, "learning_rate": 2.4750000000000004e-06, "loss": 1.2104, "step": 100 }, { "grad_norm": 1.4917453527450562, "learning_rate": 2.725e-06, "loss": 1.1712, "step": 110 }, { "grad_norm": 1.2559216022491455, "learning_rate": 2.975e-06, "loss": 1.1504, "step": 120 }, { "grad_norm": 1.308040976524353, "learning_rate": 3.225e-06, "loss": 1.1103, "step": 130 }, { "grad_norm": 1.5208508968353271, "learning_rate": 3.4750000000000006e-06, "loss": 1.113, "step": 140 }, { "grad_norm": 1.1622034311294556, "learning_rate": 3.725e-06, "loss": 1.0894, "step": 150 }, { "grad_norm": 1.497748851776123, "learning_rate": 3.975e-06, "loss": 1.1007, "step": 160 }, { "grad_norm": 2.3048512935638428, "learning_rate": 4.225e-06, "loss": 1.0965, "step": 170 }, { "grad_norm": 1.5084750652313232, "learning_rate": 4.475e-06, "loss": 1.0887, "step": 180 }, { "grad_norm": 1.6654006242752075, "learning_rate": 4.7250000000000005e-06, "loss": 1.0901, "step": 190 }, { "grad_norm": 1.7622994184494019, "learning_rate": 4.975000000000001e-06, "loss": 1.0842, "step": 200 }, { "grad_norm": 1.3067325353622437, "learning_rate": 5.225e-06, "loss": 1.0666, "step": 210 }, { "grad_norm": 1.4524353742599487, "learning_rate": 5.475e-06, "loss": 1.0836, "step": 220 }, { "grad_norm": 1.2625765800476074, "learning_rate": 5.725e-06, "loss": 1.0676, "step": 230 }, { "grad_norm": 1.540671944618225, "learning_rate": 5.975e-06, "loss": 1.0908, "step": 240 }, { "grad_norm": 1.5681145191192627, "learning_rate": 6.2250000000000005e-06, "loss": 1.0705, "step": 250 }, { "grad_norm": 1.108666181564331, "learning_rate": 6.475000000000001e-06, "loss": 1.054, "step": 260 }, { "grad_norm": 1.337371826171875, "learning_rate": 6.725000000000001e-06, "loss": 1.0743, "step": 270 }, { "grad_norm": 2.464503765106201, "learning_rate": 6.975000000000001e-06, "loss": 1.0694, "step": 280 }, { "grad_norm": 1.5221251249313354, "learning_rate": 7.2249999999999994e-06, "loss": 1.0629, "step": 290 }, { "grad_norm": 1.243592381477356, "learning_rate": 7.4750000000000004e-06, "loss": 1.0517, "step": 300 }, { "grad_norm": 1.164477825164795, "learning_rate": 7.725e-06, "loss": 1.0877, "step": 310 }, { "grad_norm": 1.3105690479278564, "learning_rate": 7.975e-06, "loss": 1.0379, "step": 320 }, { "grad_norm": 1.3562792539596558, "learning_rate": 8.225e-06, "loss": 1.0432, "step": 330 }, { "grad_norm": 1.5559680461883545, "learning_rate": 8.475000000000001e-06, "loss": 1.0397, "step": 340 }, { "grad_norm": 1.7620410919189453, "learning_rate": 8.725e-06, "loss": 1.0282, "step": 350 }, { "grad_norm": 1.8455649614334106, "learning_rate": 8.975e-06, "loss": 1.0015, "step": 360 }, { "grad_norm": 1.874539852142334, "learning_rate": 9.225e-06, "loss": 0.9944, "step": 370 }, { "grad_norm": 1.7814366817474365, "learning_rate": 9.475e-06, "loss": 0.9628, "step": 380 }, { "grad_norm": 1.8449420928955078, "learning_rate": 9.725000000000001e-06, "loss": 0.9551, "step": 390 }, { "grad_norm": 1.98610258102417, "learning_rate": 9.975e-06, "loss": 0.9286, "step": 400 }, { "grad_norm": 2.135633707046509, "learning_rate": 1.0225e-05, "loss": 0.8775, "step": 410 }, { "grad_norm": 2.3415558338165283, "learning_rate": 1.0475e-05, "loss": 0.8606, "step": 420 }, { "grad_norm": 2.247953176498413, "learning_rate": 1.0725e-05, "loss": 0.8536, "step": 430 }, { "grad_norm": 2.5676543712615967, "learning_rate": 1.0975e-05, "loss": 0.8198, "step": 440 }, { "grad_norm": 2.281613349914551, "learning_rate": 1.1225e-05, "loss": 0.7699, "step": 450 }, { "grad_norm": 2.3858494758605957, "learning_rate": 1.1475000000000001e-05, "loss": 0.79, "step": 460 }, { "grad_norm": 2.3041932582855225, "learning_rate": 1.1725e-05, "loss": 0.7727, "step": 470 }, { "grad_norm": 2.7523887157440186, "learning_rate": 1.1975e-05, "loss": 0.7298, "step": 480 }, { "grad_norm": 2.3698928356170654, "learning_rate": 1.2225e-05, "loss": 0.7137, "step": 490 }, { "grad_norm": 3.249725818634033, "learning_rate": 1.2475e-05, "loss": 0.6605, "step": 500 }, { "grad_norm": 2.4796881675720215, "learning_rate": 1.2725000000000001e-05, "loss": 0.6483, "step": 510 }, { "grad_norm": 2.644423484802246, "learning_rate": 1.2975e-05, "loss": 0.6244, "step": 520 }, { "grad_norm": 3.029061794281006, "learning_rate": 1.3225000000000001e-05, "loss": 0.6183, "step": 530 }, { "grad_norm": 2.4775869846343994, "learning_rate": 1.3475000000000002e-05, "loss": 0.5324, "step": 540 }, { "grad_norm": 3.343822479248047, "learning_rate": 1.3725000000000002e-05, "loss": 0.5894, "step": 550 }, { "grad_norm": 2.8227598667144775, "learning_rate": 1.3975000000000003e-05, "loss": 0.5712, "step": 560 }, { "grad_norm": 2.746751308441162, "learning_rate": 1.4225e-05, "loss": 0.5122, "step": 570 }, { "grad_norm": 2.3939247131347656, "learning_rate": 1.4475e-05, "loss": 0.4983, "step": 580 }, { "grad_norm": 2.4002280235290527, "learning_rate": 1.4725e-05, "loss": 0.496, "step": 590 }, { "grad_norm": 2.2184393405914307, "learning_rate": 1.4975e-05, "loss": 0.4799, "step": 600 }, { "grad_norm": 2.848628520965576, "learning_rate": 1.5225e-05, "loss": 0.4632, "step": 610 }, { "grad_norm": 3.476722240447998, "learning_rate": 1.5475e-05, "loss": 0.4304, "step": 620 }, { "grad_norm": 2.5262041091918945, "learning_rate": 1.5725e-05, "loss": 0.4025, "step": 630 }, { "grad_norm": 2.956139326095581, "learning_rate": 1.5975000000000002e-05, "loss": 0.394, "step": 640 }, { "grad_norm": 2.7580301761627197, "learning_rate": 1.6225e-05, "loss": 0.3661, "step": 650 }, { "grad_norm": 2.9885311126708984, "learning_rate": 1.6475e-05, "loss": 0.3726, "step": 660 }, { "grad_norm": 2.793079376220703, "learning_rate": 1.6725000000000003e-05, "loss": 0.3307, "step": 670 }, { "grad_norm": 3.8971872329711914, "learning_rate": 1.6975000000000003e-05, "loss": 0.3209, "step": 680 }, { "grad_norm": 3.6536030769348145, "learning_rate": 1.7225e-05, "loss": 0.317, "step": 690 }, { "grad_norm": 3.162339210510254, "learning_rate": 1.7475e-05, "loss": 0.3222, "step": 700 }, { "grad_norm": 3.199712038040161, "learning_rate": 1.7725e-05, "loss": 0.2558, "step": 710 }, { "grad_norm": 3.501288890838623, "learning_rate": 1.7975e-05, "loss": 0.2901, "step": 720 }, { "grad_norm": 2.3992342948913574, "learning_rate": 1.8225e-05, "loss": 0.2534, "step": 730 }, { "grad_norm": 2.3639609813690186, "learning_rate": 1.8475000000000002e-05, "loss": 0.2201, "step": 740 }, { "grad_norm": 2.8643500804901123, "learning_rate": 1.8725e-05, "loss": 0.2407, "step": 750 }, { "grad_norm": 2.215841770172119, "learning_rate": 1.8975e-05, "loss": 0.2221, "step": 760 }, { "grad_norm": 2.5744059085845947, "learning_rate": 1.9225e-05, "loss": 0.2279, "step": 770 }, { "grad_norm": 2.482105016708374, "learning_rate": 1.9475000000000002e-05, "loss": 0.2491, "step": 780 }, { "grad_norm": 2.6929218769073486, "learning_rate": 1.9725000000000002e-05, "loss": 0.196, "step": 790 }, { "grad_norm": 2.671656608581543, "learning_rate": 1.9975e-05, "loss": 0.2055, "step": 800 }, { "grad_norm": 2.664733648300171, "learning_rate": 2.0225000000000004e-05, "loss": 0.1959, "step": 810 }, { "grad_norm": 2.2152364253997803, "learning_rate": 2.0475e-05, "loss": 0.2008, "step": 820 }, { "grad_norm": 2.9868967533111572, "learning_rate": 2.0725e-05, "loss": 0.2235, "step": 830 }, { "grad_norm": 2.3160581588745117, "learning_rate": 2.0975e-05, "loss": 0.2061, "step": 840 }, { "grad_norm": 2.5026683807373047, "learning_rate": 2.1225e-05, "loss": 0.2166, "step": 850 }, { "grad_norm": 2.31593918800354, "learning_rate": 2.1475e-05, "loss": 0.1804, "step": 860 }, { "grad_norm": 2.2536234855651855, "learning_rate": 2.1725e-05, "loss": 0.2112, "step": 870 }, { "grad_norm": 3.1359593868255615, "learning_rate": 2.1975000000000002e-05, "loss": 0.1803, "step": 880 }, { "grad_norm": 2.642592668533325, "learning_rate": 2.2225e-05, "loss": 0.1959, "step": 890 }, { "grad_norm": 3.202789068222046, "learning_rate": 2.2475e-05, "loss": 0.1941, "step": 900 }, { "grad_norm": 1.487007975578308, "learning_rate": 2.2725000000000003e-05, "loss": 0.1806, "step": 910 }, { "grad_norm": 1.792157769203186, "learning_rate": 2.2975000000000003e-05, "loss": 0.1685, "step": 920 }, { "grad_norm": 2.424095392227173, "learning_rate": 2.3225000000000002e-05, "loss": 0.1834, "step": 930 }, { "grad_norm": 3.2668988704681396, "learning_rate": 2.3475e-05, "loss": 0.1847, "step": 940 }, { "grad_norm": 1.6416304111480713, "learning_rate": 2.3725e-05, "loss": 0.1518, "step": 950 }, { "grad_norm": 2.0032975673675537, "learning_rate": 2.3975e-05, "loss": 0.1766, "step": 960 }, { "grad_norm": 2.1476218700408936, "learning_rate": 2.4225e-05, "loss": 0.2108, "step": 970 }, { "grad_norm": 1.7695879936218262, "learning_rate": 2.4475000000000002e-05, "loss": 0.1672, "step": 980 }, { "grad_norm": 2.4151110649108887, "learning_rate": 2.4725e-05, "loss": 0.1608, "step": 990 }, { "grad_norm": 1.9638254642486572, "learning_rate": 2.4975e-05, "loss": 0.144, "step": 1000 }, { "grad_norm": 1.9342906475067139, "learning_rate": 2.5225e-05, "loss": 0.1615, "step": 1010 }, { "grad_norm": 2.854442834854126, "learning_rate": 2.5475e-05, "loss": 0.1898, "step": 1020 }, { "grad_norm": 2.0341577529907227, "learning_rate": 2.5725e-05, "loss": 0.1435, "step": 1030 }, { "grad_norm": 1.9076242446899414, "learning_rate": 2.5974999999999998e-05, "loss": 0.1614, "step": 1040 }, { "grad_norm": 2.5645337104797363, "learning_rate": 2.6225e-05, "loss": 0.1669, "step": 1050 }, { "grad_norm": 2.118049383163452, "learning_rate": 2.6475e-05, "loss": 0.1801, "step": 1060 }, { "grad_norm": 2.913080930709839, "learning_rate": 2.6725e-05, "loss": 0.1892, "step": 1070 }, { "grad_norm": 2.4111437797546387, "learning_rate": 2.6975000000000002e-05, "loss": 0.205, "step": 1080 }, { "grad_norm": 1.5605792999267578, "learning_rate": 2.7225e-05, "loss": 0.167, "step": 1090 }, { "grad_norm": 1.7601714134216309, "learning_rate": 2.7475e-05, "loss": 0.1374, "step": 1100 }, { "grad_norm": 2.472677230834961, "learning_rate": 2.7725e-05, "loss": 0.2081, "step": 1110 }, { "grad_norm": 1.785762906074524, "learning_rate": 2.7975000000000002e-05, "loss": 0.1314, "step": 1120 }, { "grad_norm": 1.9420827627182007, "learning_rate": 2.8225e-05, "loss": 0.1768, "step": 1130 }, { "grad_norm": 2.4474246501922607, "learning_rate": 2.8475e-05, "loss": 0.2022, "step": 1140 }, { "grad_norm": 1.9413552284240723, "learning_rate": 2.8725e-05, "loss": 0.1871, "step": 1150 }, { "grad_norm": 1.611482858657837, "learning_rate": 2.8975000000000003e-05, "loss": 0.1701, "step": 1160 }, { "grad_norm": 2.0752696990966797, "learning_rate": 2.9225000000000002e-05, "loss": 0.1653, "step": 1170 }, { "grad_norm": 1.7608940601348877, "learning_rate": 2.9475e-05, "loss": 0.1448, "step": 1180 }, { "grad_norm": 2.277385950088501, "learning_rate": 2.9725000000000004e-05, "loss": 0.154, "step": 1190 }, { "grad_norm": 2.94287371635437, "learning_rate": 2.9975000000000004e-05, "loss": 0.1583, "step": 1200 }, { "grad_norm": 3.120795488357544, "learning_rate": 3.0225000000000003e-05, "loss": 0.1881, "step": 1210 }, { "grad_norm": 1.5817204713821411, "learning_rate": 3.0475000000000002e-05, "loss": 0.1661, "step": 1220 }, { "grad_norm": 2.4430978298187256, "learning_rate": 3.0725e-05, "loss": 0.1701, "step": 1230 }, { "grad_norm": 1.5089224576950073, "learning_rate": 3.0975e-05, "loss": 0.1745, "step": 1240 }, { "grad_norm": 2.1234333515167236, "learning_rate": 3.122500000000001e-05, "loss": 0.1439, "step": 1250 }, { "grad_norm": 2.3545491695404053, "learning_rate": 3.1475e-05, "loss": 0.1687, "step": 1260 }, { "grad_norm": 1.6465274095535278, "learning_rate": 3.1725e-05, "loss": 0.1281, "step": 1270 }, { "grad_norm": 1.408788800239563, "learning_rate": 3.1975e-05, "loss": 0.1418, "step": 1280 }, { "grad_norm": 1.7399520874023438, "learning_rate": 3.2225e-05, "loss": 0.1558, "step": 1290 }, { "grad_norm": 2.0221333503723145, "learning_rate": 3.2474999999999997e-05, "loss": 0.147, "step": 1300 }, { "grad_norm": 1.7454241514205933, "learning_rate": 3.2725e-05, "loss": 0.1533, "step": 1310 }, { "grad_norm": 1.9043464660644531, "learning_rate": 3.2975e-05, "loss": 0.1499, "step": 1320 }, { "grad_norm": 2.1937477588653564, "learning_rate": 3.3225e-05, "loss": 0.1213, "step": 1330 }, { "grad_norm": 1.674835205078125, "learning_rate": 3.3475e-05, "loss": 0.1274, "step": 1340 }, { "grad_norm": 1.7342044115066528, "learning_rate": 3.3725e-05, "loss": 0.1598, "step": 1350 }, { "grad_norm": 2.2671303749084473, "learning_rate": 3.3975e-05, "loss": 0.1628, "step": 1360 }, { "grad_norm": 1.8875504732131958, "learning_rate": 3.4225e-05, "loss": 0.1326, "step": 1370 }, { "grad_norm": 2.0658180713653564, "learning_rate": 3.4475000000000005e-05, "loss": 0.1355, "step": 1380 }, { "grad_norm": 1.9204620122909546, "learning_rate": 3.4725000000000004e-05, "loss": 0.1377, "step": 1390 }, { "grad_norm": 1.7382718324661255, "learning_rate": 3.4975e-05, "loss": 0.1324, "step": 1400 }, { "grad_norm": 1.7024857997894287, "learning_rate": 3.5225e-05, "loss": 0.1478, "step": 1410 }, { "grad_norm": 1.9206066131591797, "learning_rate": 3.5475e-05, "loss": 0.1423, "step": 1420 }, { "grad_norm": 1.6727944612503052, "learning_rate": 3.5725e-05, "loss": 0.125, "step": 1430 }, { "grad_norm": 1.6063494682312012, "learning_rate": 3.5975e-05, "loss": 0.1557, "step": 1440 }, { "grad_norm": 1.9116986989974976, "learning_rate": 3.6225000000000006e-05, "loss": 0.1757, "step": 1450 }, { "grad_norm": 1.403788447380066, "learning_rate": 3.6475000000000006e-05, "loss": 0.1436, "step": 1460 }, { "grad_norm": 1.7817156314849854, "learning_rate": 3.6725000000000005e-05, "loss": 0.1462, "step": 1470 }, { "grad_norm": 1.806011438369751, "learning_rate": 3.6975000000000004e-05, "loss": 0.131, "step": 1480 }, { "grad_norm": 1.3438639640808105, "learning_rate": 3.7225000000000004e-05, "loss": 0.1312, "step": 1490 }, { "grad_norm": 1.2519056797027588, "learning_rate": 3.7475e-05, "loss": 0.1332, "step": 1500 }, { "grad_norm": 1.7497494220733643, "learning_rate": 3.7725e-05, "loss": 0.1404, "step": 1510 }, { "grad_norm": 1.805696725845337, "learning_rate": 3.7975e-05, "loss": 0.1405, "step": 1520 }, { "grad_norm": 1.3020778894424438, "learning_rate": 3.8225e-05, "loss": 0.1533, "step": 1530 }, { "grad_norm": 1.719855785369873, "learning_rate": 3.8475e-05, "loss": 0.1328, "step": 1540 }, { "grad_norm": 1.60231351852417, "learning_rate": 3.8725e-05, "loss": 0.136, "step": 1550 }, { "grad_norm": 1.6975375413894653, "learning_rate": 3.8975e-05, "loss": 0.1561, "step": 1560 }, { "grad_norm": 1.444487452507019, "learning_rate": 3.9225e-05, "loss": 0.1351, "step": 1570 }, { "grad_norm": 1.3843132257461548, "learning_rate": 3.9475000000000004e-05, "loss": 0.1418, "step": 1580 }, { "grad_norm": 1.3392664194107056, "learning_rate": 3.9725e-05, "loss": 0.1633, "step": 1590 }, { "grad_norm": 1.5310261249542236, "learning_rate": 3.9975e-05, "loss": 0.1442, "step": 1600 }, { "grad_norm": 1.4612329006195068, "learning_rate": 4.0225e-05, "loss": 0.1449, "step": 1610 }, { "grad_norm": 1.4247416257858276, "learning_rate": 4.0475e-05, "loss": 0.1494, "step": 1620 }, { "grad_norm": 1.1356186866760254, "learning_rate": 4.0725e-05, "loss": 0.1688, "step": 1630 }, { "grad_norm": 1.2113525867462158, "learning_rate": 4.0975e-05, "loss": 0.1306, "step": 1640 }, { "grad_norm": 1.1675111055374146, "learning_rate": 4.1225e-05, "loss": 0.1382, "step": 1650 }, { "grad_norm": 1.1555137634277344, "learning_rate": 4.1475000000000005e-05, "loss": 0.132, "step": 1660 }, { "grad_norm": 1.0827919244766235, "learning_rate": 4.1725000000000005e-05, "loss": 0.1164, "step": 1670 }, { "grad_norm": 1.5777450799942017, "learning_rate": 4.1975000000000004e-05, "loss": 0.1305, "step": 1680 }, { "grad_norm": 1.6959420442581177, "learning_rate": 4.2225e-05, "loss": 0.132, "step": 1690 }, { "grad_norm": 1.12321937084198, "learning_rate": 4.2475e-05, "loss": 0.1387, "step": 1700 }, { "grad_norm": 1.5011814832687378, "learning_rate": 4.2725e-05, "loss": 0.1445, "step": 1710 }, { "grad_norm": 1.3961702585220337, "learning_rate": 4.2975e-05, "loss": 0.1361, "step": 1720 }, { "grad_norm": 1.4111226797103882, "learning_rate": 4.322500000000001e-05, "loss": 0.1291, "step": 1730 }, { "grad_norm": 1.3344310522079468, "learning_rate": 4.3475000000000006e-05, "loss": 0.1218, "step": 1740 }, { "grad_norm": 1.4845929145812988, "learning_rate": 4.3725000000000006e-05, "loss": 0.1294, "step": 1750 }, { "grad_norm": 1.5573481321334839, "learning_rate": 4.3975e-05, "loss": 0.1388, "step": 1760 }, { "grad_norm": 1.5748850107192993, "learning_rate": 4.4225e-05, "loss": 0.1218, "step": 1770 }, { "grad_norm": 1.0115056037902832, "learning_rate": 4.4475e-05, "loss": 0.1482, "step": 1780 }, { "grad_norm": 1.3764228820800781, "learning_rate": 4.4725e-05, "loss": 0.1299, "step": 1790 }, { "grad_norm": 1.0131447315216064, "learning_rate": 4.4975e-05, "loss": 0.1336, "step": 1800 }, { "grad_norm": 1.5371311902999878, "learning_rate": 4.5225e-05, "loss": 0.1266, "step": 1810 }, { "grad_norm": 1.537814974784851, "learning_rate": 4.5475e-05, "loss": 0.1405, "step": 1820 }, { "grad_norm": 1.491924524307251, "learning_rate": 4.5725e-05, "loss": 0.1361, "step": 1830 }, { "grad_norm": 1.2180269956588745, "learning_rate": 4.5975e-05, "loss": 0.1364, "step": 1840 }, { "grad_norm": 1.4018622636795044, "learning_rate": 4.6225e-05, "loss": 0.1559, "step": 1850 }, { "grad_norm": 1.3367363214492798, "learning_rate": 4.6475000000000005e-05, "loss": 0.1483, "step": 1860 }, { "grad_norm": 0.9836007952690125, "learning_rate": 4.6725000000000004e-05, "loss": 0.1372, "step": 1870 }, { "grad_norm": 1.120926856994629, "learning_rate": 4.6975000000000003e-05, "loss": 0.1106, "step": 1880 }, { "grad_norm": 1.429448127746582, "learning_rate": 4.7225e-05, "loss": 0.1336, "step": 1890 }, { "grad_norm": 1.2175664901733398, "learning_rate": 4.7475e-05, "loss": 0.1415, "step": 1900 }, { "grad_norm": 1.3281458616256714, "learning_rate": 4.7725e-05, "loss": 0.1667, "step": 1910 }, { "grad_norm": 1.0261378288269043, "learning_rate": 4.7975e-05, "loss": 0.1487, "step": 1920 }, { "grad_norm": 0.904735267162323, "learning_rate": 4.822500000000001e-05, "loss": 0.1242, "step": 1930 }, { "grad_norm": 1.4760242700576782, "learning_rate": 4.8475000000000006e-05, "loss": 0.1398, "step": 1940 }, { "grad_norm": 1.2324155569076538, "learning_rate": 4.8725000000000005e-05, "loss": 0.1225, "step": 1950 }, { "grad_norm": 1.42280912399292, "learning_rate": 4.8975000000000005e-05, "loss": 0.1667, "step": 1960 }, { "grad_norm": 1.3768543004989624, "learning_rate": 4.9225000000000004e-05, "loss": 0.1191, "step": 1970 }, { "grad_norm": 1.27783203125, "learning_rate": 4.9475e-05, "loss": 0.1356, "step": 1980 }, { "grad_norm": 1.2203842401504517, "learning_rate": 4.9725e-05, "loss": 0.1118, "step": 1990 }, { "grad_norm": 1.2020679712295532, "learning_rate": 4.9975e-05, "loss": 0.138, "step": 2000 }, { "grad_norm": 0.9593155384063721, "learning_rate": 5.0225e-05, "loss": 0.1517, "step": 2010 }, { "grad_norm": 1.4347420930862427, "learning_rate": 5.047500000000001e-05, "loss": 0.1317, "step": 2020 }, { "grad_norm": 1.282668113708496, "learning_rate": 5.0725e-05, "loss": 0.1172, "step": 2030 }, { "grad_norm": 1.0012363195419312, "learning_rate": 5.0975000000000006e-05, "loss": 0.1323, "step": 2040 }, { "grad_norm": 1.146945834159851, "learning_rate": 5.1225e-05, "loss": 0.1374, "step": 2050 }, { "grad_norm": 1.308192491531372, "learning_rate": 5.1475000000000004e-05, "loss": 0.13, "step": 2060 }, { "grad_norm": 1.3196141719818115, "learning_rate": 5.1725000000000004e-05, "loss": 0.1199, "step": 2070 }, { "grad_norm": 1.219812273979187, "learning_rate": 5.197500000000001e-05, "loss": 0.1327, "step": 2080 }, { "grad_norm": 1.5042529106140137, "learning_rate": 5.2225e-05, "loss": 0.1436, "step": 2090 }, { "grad_norm": 1.1922507286071777, "learning_rate": 5.247500000000001e-05, "loss": 0.1304, "step": 2100 }, { "grad_norm": 1.393477439880371, "learning_rate": 5.2725e-05, "loss": 0.1542, "step": 2110 }, { "grad_norm": 1.1674258708953857, "learning_rate": 5.297500000000001e-05, "loss": 0.1317, "step": 2120 }, { "grad_norm": 0.8777985572814941, "learning_rate": 5.3225e-05, "loss": 0.123, "step": 2130 }, { "grad_norm": 1.094232201576233, "learning_rate": 5.3475e-05, "loss": 0.1101, "step": 2140 }, { "grad_norm": 1.5808614492416382, "learning_rate": 5.3725000000000005e-05, "loss": 0.1463, "step": 2150 }, { "grad_norm": 1.0420020818710327, "learning_rate": 5.3975e-05, "loss": 0.1284, "step": 2160 }, { "grad_norm": 1.0645002126693726, "learning_rate": 5.4225000000000003e-05, "loss": 0.1136, "step": 2170 }, { "grad_norm": 0.8317804932594299, "learning_rate": 5.4474999999999996e-05, "loss": 0.1331, "step": 2180 }, { "grad_norm": 0.9006568789482117, "learning_rate": 5.4725e-05, "loss": 0.1192, "step": 2190 }, { "grad_norm": 0.7460271120071411, "learning_rate": 5.4975e-05, "loss": 0.1092, "step": 2200 }, { "grad_norm": 0.9070411920547485, "learning_rate": 5.522500000000001e-05, "loss": 0.1167, "step": 2210 }, { "grad_norm": 1.0192139148712158, "learning_rate": 5.5475e-05, "loss": 0.1183, "step": 2220 }, { "grad_norm": 1.136453628540039, "learning_rate": 5.5725000000000006e-05, "loss": 0.1247, "step": 2230 }, { "grad_norm": 1.0337615013122559, "learning_rate": 5.5975e-05, "loss": 0.1253, "step": 2240 }, { "grad_norm": 1.0261378288269043, "learning_rate": 5.6225000000000005e-05, "loss": 0.1378, "step": 2250 }, { "grad_norm": 1.2173715829849243, "learning_rate": 5.6475e-05, "loss": 0.1087, "step": 2260 }, { "grad_norm": 0.973348081111908, "learning_rate": 5.6725e-05, "loss": 0.1103, "step": 2270 }, { "grad_norm": 1.0401755571365356, "learning_rate": 5.6975e-05, "loss": 0.1108, "step": 2280 }, { "grad_norm": 1.1816215515136719, "learning_rate": 5.722500000000001e-05, "loss": 0.1489, "step": 2290 }, { "grad_norm": 1.1433942317962646, "learning_rate": 5.7475e-05, "loss": 0.1113, "step": 2300 }, { "grad_norm": 0.8139522075653076, "learning_rate": 5.772500000000001e-05, "loss": 0.1129, "step": 2310 }, { "grad_norm": 0.8885473012924194, "learning_rate": 5.7975e-05, "loss": 0.1029, "step": 2320 }, { "grad_norm": 0.7962767481803894, "learning_rate": 5.8225000000000006e-05, "loss": 0.1357, "step": 2330 }, { "grad_norm": 1.1593412160873413, "learning_rate": 5.8475000000000005e-05, "loss": 0.1222, "step": 2340 }, { "grad_norm": 0.8327237963676453, "learning_rate": 5.8725000000000004e-05, "loss": 0.1223, "step": 2350 }, { "grad_norm": 1.03765070438385, "learning_rate": 5.8975000000000004e-05, "loss": 0.1263, "step": 2360 }, { "grad_norm": 0.8425435423851013, "learning_rate": 5.922500000000001e-05, "loss": 0.1152, "step": 2370 }, { "grad_norm": 0.94791579246521, "learning_rate": 5.9475e-05, "loss": 0.1197, "step": 2380 }, { "grad_norm": 0.9763002395629883, "learning_rate": 5.9724999999999995e-05, "loss": 0.133, "step": 2390 }, { "grad_norm": 0.6696591973304749, "learning_rate": 5.9975e-05, "loss": 0.1081, "step": 2400 }, { "grad_norm": 0.8269146680831909, "learning_rate": 6.0225e-05, "loss": 0.1159, "step": 2410 }, { "grad_norm": 1.1337480545043945, "learning_rate": 6.0475000000000006e-05, "loss": 0.1679, "step": 2420 }, { "grad_norm": 1.2941079139709473, "learning_rate": 6.0725e-05, "loss": 0.1302, "step": 2430 }, { "grad_norm": 1.309309720993042, "learning_rate": 6.0975000000000005e-05, "loss": 0.1425, "step": 2440 }, { "grad_norm": 1.1263314485549927, "learning_rate": 6.1225e-05, "loss": 0.1249, "step": 2450 }, { "grad_norm": 0.7029446959495544, "learning_rate": 6.1475e-05, "loss": 0.1159, "step": 2460 }, { "grad_norm": 1.05169677734375, "learning_rate": 6.1725e-05, "loss": 0.1155, "step": 2470 }, { "grad_norm": 0.6957229971885681, "learning_rate": 6.1975e-05, "loss": 0.1356, "step": 2480 }, { "grad_norm": 1.199585199356079, "learning_rate": 6.2225e-05, "loss": 0.1282, "step": 2490 }, { "grad_norm": 0.875529944896698, "learning_rate": 6.2475e-05, "loss": 0.1041, "step": 2500 }, { "grad_norm": 1.1835284233093262, "learning_rate": 6.2725e-05, "loss": 0.0953, "step": 2510 }, { "grad_norm": 1.2956169843673706, "learning_rate": 6.297500000000001e-05, "loss": 0.1481, "step": 2520 }, { "grad_norm": 1.1971521377563477, "learning_rate": 6.3225e-05, "loss": 0.1188, "step": 2530 }, { "grad_norm": 0.853269100189209, "learning_rate": 6.347500000000001e-05, "loss": 0.1211, "step": 2540 }, { "grad_norm": 1.195206642150879, "learning_rate": 6.3725e-05, "loss": 0.1388, "step": 2550 }, { "grad_norm": 0.8857545852661133, "learning_rate": 6.397500000000001e-05, "loss": 0.1333, "step": 2560 }, { "grad_norm": 0.8606415390968323, "learning_rate": 6.4225e-05, "loss": 0.1162, "step": 2570 }, { "grad_norm": 0.7992307543754578, "learning_rate": 6.447500000000001e-05, "loss": 0.122, "step": 2580 }, { "grad_norm": 0.8283895254135132, "learning_rate": 6.4725e-05, "loss": 0.1128, "step": 2590 }, { "grad_norm": 0.9576820135116577, "learning_rate": 6.497500000000001e-05, "loss": 0.1225, "step": 2600 }, { "grad_norm": 0.8319986462593079, "learning_rate": 6.5225e-05, "loss": 0.1255, "step": 2610 }, { "grad_norm": 0.8325812816619873, "learning_rate": 6.5475e-05, "loss": 0.0984, "step": 2620 }, { "grad_norm": 0.973759651184082, "learning_rate": 6.5725e-05, "loss": 0.1188, "step": 2630 }, { "grad_norm": 0.8219820857048035, "learning_rate": 6.5975e-05, "loss": 0.1357, "step": 2640 }, { "grad_norm": 0.9811965823173523, "learning_rate": 6.6225e-05, "loss": 0.1255, "step": 2650 }, { "grad_norm": 1.155971646308899, "learning_rate": 6.6475e-05, "loss": 0.1329, "step": 2660 }, { "grad_norm": 0.8589267730712891, "learning_rate": 6.672500000000001e-05, "loss": 0.1338, "step": 2670 }, { "grad_norm": 1.2097415924072266, "learning_rate": 6.6975e-05, "loss": 0.1493, "step": 2680 }, { "grad_norm": 0.8028292059898376, "learning_rate": 6.722500000000001e-05, "loss": 0.1159, "step": 2690 }, { "grad_norm": 1.0501439571380615, "learning_rate": 6.7475e-05, "loss": 0.1112, "step": 2700 }, { "grad_norm": 0.8657853007316589, "learning_rate": 6.7725e-05, "loss": 0.1195, "step": 2710 }, { "grad_norm": 0.7739560604095459, "learning_rate": 6.7975e-05, "loss": 0.1425, "step": 2720 }, { "grad_norm": 0.8187438249588013, "learning_rate": 6.8225e-05, "loss": 0.1077, "step": 2730 }, { "grad_norm": 0.6499968767166138, "learning_rate": 6.8475e-05, "loss": 0.1055, "step": 2740 }, { "grad_norm": 0.9387922286987305, "learning_rate": 6.8725e-05, "loss": 0.112, "step": 2750 }, { "grad_norm": 0.9827699661254883, "learning_rate": 6.8975e-05, "loss": 0.107, "step": 2760 }, { "grad_norm": 0.9723964333534241, "learning_rate": 6.9225e-05, "loss": 0.1095, "step": 2770 }, { "grad_norm": 0.6807293891906738, "learning_rate": 6.9475e-05, "loss": 0.1433, "step": 2780 }, { "grad_norm": 0.8857168555259705, "learning_rate": 6.9725e-05, "loss": 0.1158, "step": 2790 }, { "grad_norm": 0.8057045340538025, "learning_rate": 6.997500000000001e-05, "loss": 0.116, "step": 2800 }, { "grad_norm": 1.0817015171051025, "learning_rate": 7.022500000000001e-05, "loss": 0.1531, "step": 2810 }, { "grad_norm": 1.0477477312088013, "learning_rate": 7.0475e-05, "loss": 0.1338, "step": 2820 }, { "grad_norm": 0.7444818615913391, "learning_rate": 7.072500000000001e-05, "loss": 0.1009, "step": 2830 }, { "grad_norm": 0.7331891059875488, "learning_rate": 7.0975e-05, "loss": 0.1311, "step": 2840 }, { "grad_norm": 0.8939816355705261, "learning_rate": 7.122500000000001e-05, "loss": 0.1263, "step": 2850 }, { "grad_norm": 0.8404167294502258, "learning_rate": 7.1475e-05, "loss": 0.1191, "step": 2860 }, { "grad_norm": 0.7516897916793823, "learning_rate": 7.172500000000001e-05, "loss": 0.1076, "step": 2870 }, { "grad_norm": 0.8399802446365356, "learning_rate": 7.1975e-05, "loss": 0.0988, "step": 2880 }, { "grad_norm": 0.8216364979743958, "learning_rate": 7.2225e-05, "loss": 0.1136, "step": 2890 }, { "grad_norm": 0.6912533640861511, "learning_rate": 7.2475e-05, "loss": 0.1157, "step": 2900 }, { "grad_norm": 0.9111034870147705, "learning_rate": 7.272499999999999e-05, "loss": 0.0941, "step": 2910 }, { "grad_norm": 0.7684377431869507, "learning_rate": 7.2975e-05, "loss": 0.1105, "step": 2920 }, { "grad_norm": 0.9105249643325806, "learning_rate": 7.3225e-05, "loss": 0.0948, "step": 2930 }, { "grad_norm": 0.8023672103881836, "learning_rate": 7.347500000000001e-05, "loss": 0.0941, "step": 2940 }, { "grad_norm": 0.8010969758033752, "learning_rate": 7.3725e-05, "loss": 0.1019, "step": 2950 }, { "grad_norm": 0.6008653044700623, "learning_rate": 7.397500000000001e-05, "loss": 0.1023, "step": 2960 }, { "grad_norm": 0.7626001238822937, "learning_rate": 7.4225e-05, "loss": 0.1031, "step": 2970 }, { "grad_norm": 0.9728943705558777, "learning_rate": 7.447500000000001e-05, "loss": 0.1139, "step": 2980 }, { "grad_norm": 1.079416036605835, "learning_rate": 7.4725e-05, "loss": 0.1422, "step": 2990 }, { "grad_norm": 0.8510723114013672, "learning_rate": 7.4975e-05, "loss": 0.1267, "step": 3000 }, { "grad_norm": 0.8633529543876648, "learning_rate": 7.5225e-05, "loss": 0.1127, "step": 3010 }, { "grad_norm": 0.7210726141929626, "learning_rate": 7.5475e-05, "loss": 0.1087, "step": 3020 }, { "grad_norm": 0.8364072442054749, "learning_rate": 7.5725e-05, "loss": 0.11, "step": 3030 }, { "grad_norm": 0.7001913189888, "learning_rate": 7.5975e-05, "loss": 0.1102, "step": 3040 }, { "grad_norm": 0.7382663488388062, "learning_rate": 7.6225e-05, "loss": 0.1146, "step": 3050 }, { "grad_norm": 0.7547609210014343, "learning_rate": 7.6475e-05, "loss": 0.1167, "step": 3060 }, { "grad_norm": 0.7982203364372253, "learning_rate": 7.672500000000001e-05, "loss": 0.0965, "step": 3070 }, { "grad_norm": 0.687176525592804, "learning_rate": 7.697500000000001e-05, "loss": 0.1044, "step": 3080 }, { "grad_norm": 1.1279523372650146, "learning_rate": 7.722500000000001e-05, "loss": 0.1391, "step": 3090 }, { "grad_norm": 0.9666332602500916, "learning_rate": 7.747500000000001e-05, "loss": 0.1081, "step": 3100 }, { "grad_norm": 0.6977924108505249, "learning_rate": 7.7725e-05, "loss": 0.0954, "step": 3110 }, { "grad_norm": 0.8750542402267456, "learning_rate": 7.797500000000001e-05, "loss": 0.0954, "step": 3120 }, { "grad_norm": 0.9140886664390564, "learning_rate": 7.8225e-05, "loss": 0.0955, "step": 3130 }, { "grad_norm": 0.8891413807868958, "learning_rate": 7.8475e-05, "loss": 0.104, "step": 3140 }, { "grad_norm": 0.6941218972206116, "learning_rate": 7.8725e-05, "loss": 0.103, "step": 3150 }, { "grad_norm": 1.2845144271850586, "learning_rate": 7.8975e-05, "loss": 0.1204, "step": 3160 }, { "grad_norm": 0.6309693455696106, "learning_rate": 7.9225e-05, "loss": 0.1185, "step": 3170 }, { "grad_norm": 0.81076979637146, "learning_rate": 7.9475e-05, "loss": 0.1362, "step": 3180 }, { "grad_norm": 0.7089153528213501, "learning_rate": 7.9725e-05, "loss": 0.1269, "step": 3190 }, { "grad_norm": 0.9384571313858032, "learning_rate": 7.9975e-05, "loss": 0.1096, "step": 3200 }, { "grad_norm": 1.1528443098068237, "learning_rate": 8.022500000000001e-05, "loss": 0.1184, "step": 3210 }, { "grad_norm": 1.2716522216796875, "learning_rate": 8.0475e-05, "loss": 0.1513, "step": 3220 }, { "grad_norm": 0.732600212097168, "learning_rate": 8.072500000000001e-05, "loss": 0.1245, "step": 3230 }, { "grad_norm": 0.780157744884491, "learning_rate": 8.0975e-05, "loss": 0.107, "step": 3240 }, { "grad_norm": 0.9881311655044556, "learning_rate": 8.122500000000001e-05, "loss": 0.1235, "step": 3250 }, { "grad_norm": 0.900874674320221, "learning_rate": 8.1475e-05, "loss": 0.1164, "step": 3260 }, { "grad_norm": 1.605912446975708, "learning_rate": 8.172500000000001e-05, "loss": 0.1164, "step": 3270 }, { "grad_norm": 1.0107494592666626, "learning_rate": 8.1975e-05, "loss": 0.104, "step": 3280 }, { "grad_norm": 0.8771623969078064, "learning_rate": 8.2225e-05, "loss": 0.0997, "step": 3290 }, { "grad_norm": 0.6297816634178162, "learning_rate": 8.2475e-05, "loss": 0.1073, "step": 3300 }, { "grad_norm": 0.7569895386695862, "learning_rate": 8.2725e-05, "loss": 0.1078, "step": 3310 }, { "grad_norm": 0.6372128129005432, "learning_rate": 8.2975e-05, "loss": 0.1348, "step": 3320 }, { "grad_norm": 0.6992149949073792, "learning_rate": 8.3225e-05, "loss": 0.1008, "step": 3330 }, { "grad_norm": 0.7970927357673645, "learning_rate": 8.347500000000001e-05, "loss": 0.0981, "step": 3340 }, { "grad_norm": 0.7715363502502441, "learning_rate": 8.3725e-05, "loss": 0.0984, "step": 3350 }, { "grad_norm": 0.8130492568016052, "learning_rate": 8.397500000000001e-05, "loss": 0.1041, "step": 3360 }, { "grad_norm": 0.8436772227287292, "learning_rate": 8.422500000000001e-05, "loss": 0.105, "step": 3370 }, { "grad_norm": 0.8188045024871826, "learning_rate": 8.447500000000001e-05, "loss": 0.141, "step": 3380 }, { "grad_norm": 0.7041822671890259, "learning_rate": 8.4725e-05, "loss": 0.1085, "step": 3390 }, { "grad_norm": 0.9086775779724121, "learning_rate": 8.4975e-05, "loss": 0.1211, "step": 3400 }, { "grad_norm": 0.7645243406295776, "learning_rate": 8.5225e-05, "loss": 0.1116, "step": 3410 }, { "grad_norm": 0.883697509765625, "learning_rate": 8.5475e-05, "loss": 0.1082, "step": 3420 }, { "grad_norm": 0.8456624746322632, "learning_rate": 8.5725e-05, "loss": 0.1151, "step": 3430 }, { "grad_norm": 0.644473135471344, "learning_rate": 8.5975e-05, "loss": 0.1186, "step": 3440 }, { "grad_norm": 0.5890116095542908, "learning_rate": 8.6225e-05, "loss": 0.0981, "step": 3450 }, { "grad_norm": 0.656669557094574, "learning_rate": 8.6475e-05, "loss": 0.0828, "step": 3460 }, { "grad_norm": 0.7963256239891052, "learning_rate": 8.672500000000001e-05, "loss": 0.1173, "step": 3470 }, { "grad_norm": 0.790062427520752, "learning_rate": 8.6975e-05, "loss": 0.1225, "step": 3480 }, { "grad_norm": 1.0700428485870361, "learning_rate": 8.7225e-05, "loss": 0.1227, "step": 3490 }, { "grad_norm": 0.8014659285545349, "learning_rate": 8.747500000000001e-05, "loss": 0.1016, "step": 3500 }, { "grad_norm": 0.8617985248565674, "learning_rate": 8.7725e-05, "loss": 0.1053, "step": 3510 }, { "grad_norm": 0.8216811418533325, "learning_rate": 8.797500000000001e-05, "loss": 0.1008, "step": 3520 }, { "grad_norm": 0.9559829235076904, "learning_rate": 8.8225e-05, "loss": 0.1156, "step": 3530 }, { "grad_norm": 0.8086785078048706, "learning_rate": 8.847500000000001e-05, "loss": 0.1242, "step": 3540 }, { "grad_norm": 0.5376898050308228, "learning_rate": 8.8725e-05, "loss": 0.1164, "step": 3550 }, { "grad_norm": 0.9335516691207886, "learning_rate": 8.897500000000001e-05, "loss": 0.0946, "step": 3560 }, { "grad_norm": 0.8135546445846558, "learning_rate": 8.9225e-05, "loss": 0.1148, "step": 3570 }, { "grad_norm": 0.8387097716331482, "learning_rate": 8.9475e-05, "loss": 0.1249, "step": 3580 }, { "grad_norm": 0.7676780223846436, "learning_rate": 8.9725e-05, "loss": 0.1181, "step": 3590 }, { "grad_norm": 0.8485918045043945, "learning_rate": 8.9975e-05, "loss": 0.1014, "step": 3600 }, { "grad_norm": 0.8829543590545654, "learning_rate": 9.0225e-05, "loss": 0.1213, "step": 3610 }, { "grad_norm": 0.6433270573616028, "learning_rate": 9.0475e-05, "loss": 0.1041, "step": 3620 }, { "grad_norm": 0.7232736349105835, "learning_rate": 9.072500000000001e-05, "loss": 0.0882, "step": 3630 }, { "grad_norm": 0.5299816131591797, "learning_rate": 9.0975e-05, "loss": 0.0961, "step": 3640 }, { "grad_norm": 0.7980873584747314, "learning_rate": 9.122500000000001e-05, "loss": 0.1096, "step": 3650 }, { "grad_norm": 1.038492202758789, "learning_rate": 9.1475e-05, "loss": 0.1021, "step": 3660 }, { "grad_norm": 0.6758244633674622, "learning_rate": 9.172500000000001e-05, "loss": 0.1269, "step": 3670 }, { "grad_norm": 0.861500084400177, "learning_rate": 9.1975e-05, "loss": 0.1093, "step": 3680 }, { "grad_norm": 0.8153539299964905, "learning_rate": 9.2225e-05, "loss": 0.1, "step": 3690 }, { "grad_norm": 0.5889577865600586, "learning_rate": 9.2475e-05, "loss": 0.095, "step": 3700 }, { "grad_norm": 0.7692815661430359, "learning_rate": 9.2725e-05, "loss": 0.0944, "step": 3710 }, { "grad_norm": 0.6408378481864929, "learning_rate": 9.2975e-05, "loss": 0.0994, "step": 3720 }, { "grad_norm": 0.7504821419715881, "learning_rate": 9.3225e-05, "loss": 0.0998, "step": 3730 }, { "grad_norm": 0.7728430032730103, "learning_rate": 9.3475e-05, "loss": 0.0936, "step": 3740 }, { "grad_norm": 0.6226685047149658, "learning_rate": 9.3725e-05, "loss": 0.1162, "step": 3750 }, { "grad_norm": 0.6690667867660522, "learning_rate": 9.397500000000001e-05, "loss": 0.0927, "step": 3760 }, { "grad_norm": 0.8930069208145142, "learning_rate": 9.422500000000001e-05, "loss": 0.1158, "step": 3770 }, { "grad_norm": 0.8254082202911377, "learning_rate": 9.4475e-05, "loss": 0.1269, "step": 3780 }, { "grad_norm": 0.5827232599258423, "learning_rate": 9.472500000000001e-05, "loss": 0.0987, "step": 3790 }, { "grad_norm": 0.6908542513847351, "learning_rate": 9.4975e-05, "loss": 0.1276, "step": 3800 }, { "grad_norm": 0.8289167881011963, "learning_rate": 9.522500000000001e-05, "loss": 0.0949, "step": 3810 }, { "grad_norm": 0.8883662819862366, "learning_rate": 9.5475e-05, "loss": 0.1199, "step": 3820 }, { "grad_norm": 0.8748525381088257, "learning_rate": 9.572500000000001e-05, "loss": 0.0956, "step": 3830 }, { "grad_norm": 0.6845082640647888, "learning_rate": 9.5975e-05, "loss": 0.0936, "step": 3840 }, { "grad_norm": 0.7968460321426392, "learning_rate": 9.622500000000001e-05, "loss": 0.1164, "step": 3850 }, { "grad_norm": 0.979860782623291, "learning_rate": 9.6475e-05, "loss": 0.093, "step": 3860 }, { "grad_norm": 0.83113694190979, "learning_rate": 9.6725e-05, "loss": 0.0917, "step": 3870 }, { "grad_norm": 0.7219645977020264, "learning_rate": 9.6975e-05, "loss": 0.0995, "step": 3880 }, { "grad_norm": 1.1919983625411987, "learning_rate": 9.7225e-05, "loss": 0.1184, "step": 3890 }, { "grad_norm": 0.6192863583564758, "learning_rate": 9.747500000000001e-05, "loss": 0.1218, "step": 3900 }, { "grad_norm": 0.5816722512245178, "learning_rate": 9.7725e-05, "loss": 0.097, "step": 3910 }, { "grad_norm": 0.8529530167579651, "learning_rate": 9.797500000000001e-05, "loss": 0.1092, "step": 3920 }, { "grad_norm": 0.6242178678512573, "learning_rate": 9.8225e-05, "loss": 0.0934, "step": 3930 }, { "grad_norm": 0.7202402949333191, "learning_rate": 9.847500000000001e-05, "loss": 0.1179, "step": 3940 }, { "grad_norm": 0.797784149646759, "learning_rate": 9.8725e-05, "loss": 0.1197, "step": 3950 }, { "grad_norm": 0.8117775917053223, "learning_rate": 9.897500000000001e-05, "loss": 0.1226, "step": 3960 }, { "grad_norm": 0.6316511034965515, "learning_rate": 9.9225e-05, "loss": 0.0963, "step": 3970 }, { "grad_norm": 0.5852547287940979, "learning_rate": 9.9475e-05, "loss": 0.1069, "step": 3980 }, { "grad_norm": 0.841005802154541, "learning_rate": 9.9725e-05, "loss": 0.1205, "step": 3990 }, { "grad_norm": 0.722737729549408, "learning_rate": 9.9975e-05, "loss": 0.1122, "step": 4000 }, { "grad_norm": 0.8771932721138, "learning_rate": 9.999999653982884e-05, "loss": 0.1003, "step": 4010 }, { "grad_norm": 0.6893371343612671, "learning_rate": 9.999998457874392e-05, "loss": 0.128, "step": 4020 }, { "grad_norm": 0.7851032018661499, "learning_rate": 9.999996407402913e-05, "loss": 0.1111, "step": 4030 }, { "grad_norm": 0.531270444393158, "learning_rate": 9.999993502568801e-05, "loss": 0.088, "step": 4040 }, { "grad_norm": 0.870501697063446, "learning_rate": 9.999989743372548e-05, "loss": 0.1166, "step": 4050 }, { "grad_norm": 0.8013886213302612, "learning_rate": 9.999985129814798e-05, "loss": 0.1084, "step": 4060 }, { "grad_norm": 0.5944463014602661, "learning_rate": 9.99997966189634e-05, "loss": 0.09, "step": 4070 }, { "grad_norm": 0.695816159248352, "learning_rate": 9.999973339618107e-05, "loss": 0.1169, "step": 4080 }, { "grad_norm": 0.7032367587089539, "learning_rate": 9.999966162981179e-05, "loss": 0.1094, "step": 4090 }, { "grad_norm": 0.6343681812286377, "learning_rate": 9.999958131986784e-05, "loss": 0.0944, "step": 4100 }, { "grad_norm": 0.5676589012145996, "learning_rate": 9.999949246636293e-05, "loss": 0.1152, "step": 4110 }, { "grad_norm": 0.6929842829704285, "learning_rate": 9.999939506931224e-05, "loss": 0.1088, "step": 4120 }, { "grad_norm": 0.7446123361587524, "learning_rate": 9.999928912873243e-05, "loss": 0.0995, "step": 4130 }, { "grad_norm": 0.7113760709762573, "learning_rate": 9.999917464464159e-05, "loss": 0.106, "step": 4140 }, { "grad_norm": 0.7499163746833801, "learning_rate": 9.999905161705929e-05, "loss": 0.1124, "step": 4150 }, { "grad_norm": 0.6199526786804199, "learning_rate": 9.999892004600653e-05, "loss": 0.1032, "step": 4160 }, { "grad_norm": 0.5899577140808105, "learning_rate": 9.999877993150581e-05, "loss": 0.1056, "step": 4170 }, { "grad_norm": 0.7150478959083557, "learning_rate": 9.999863127358108e-05, "loss": 0.0906, "step": 4180 }, { "grad_norm": 0.6905643343925476, "learning_rate": 9.999847407225773e-05, "loss": 0.0924, "step": 4190 }, { "grad_norm": 0.661781907081604, "learning_rate": 9.999830832756262e-05, "loss": 0.0895, "step": 4200 }, { "grad_norm": 0.5257040858268738, "learning_rate": 9.999813403952407e-05, "loss": 0.1052, "step": 4210 }, { "grad_norm": 0.7297170162200928, "learning_rate": 9.999795120817187e-05, "loss": 0.1043, "step": 4220 }, { "grad_norm": 0.7200263738632202, "learning_rate": 9.999775983353725e-05, "loss": 0.1345, "step": 4230 }, { "grad_norm": 0.664472222328186, "learning_rate": 9.999755991565292e-05, "loss": 0.1123, "step": 4240 }, { "grad_norm": 0.6688179969787598, "learning_rate": 9.999735145455303e-05, "loss": 0.1177, "step": 4250 }, { "grad_norm": 0.6845320463180542, "learning_rate": 9.99971344502732e-05, "loss": 0.119, "step": 4260 }, { "grad_norm": 0.563615083694458, "learning_rate": 9.999690890285053e-05, "loss": 0.1047, "step": 4270 }, { "grad_norm": 0.9347758293151855, "learning_rate": 9.999667481232356e-05, "loss": 0.0977, "step": 4280 }, { "grad_norm": 0.9161531925201416, "learning_rate": 9.999643217873225e-05, "loss": 0.0987, "step": 4290 }, { "grad_norm": 0.8940795063972473, "learning_rate": 9.999618100211809e-05, "loss": 0.091, "step": 4300 }, { "grad_norm": 0.6546100974082947, "learning_rate": 9.999592128252402e-05, "loss": 0.1034, "step": 4310 }, { "grad_norm": 0.6724491119384766, "learning_rate": 9.999565301999437e-05, "loss": 0.1239, "step": 4320 }, { "grad_norm": 0.8360249400138855, "learning_rate": 9.999537621457502e-05, "loss": 0.1019, "step": 4330 }, { "grad_norm": 0.5764826536178589, "learning_rate": 9.999509086631323e-05, "loss": 0.0985, "step": 4340 }, { "grad_norm": 0.8100711107254028, "learning_rate": 9.99947969752578e-05, "loss": 0.0852, "step": 4350 }, { "grad_norm": 0.5834611654281616, "learning_rate": 9.999449454145891e-05, "loss": 0.0842, "step": 4360 }, { "grad_norm": 0.5566731095314026, "learning_rate": 9.999418356496827e-05, "loss": 0.0928, "step": 4370 }, { "grad_norm": 0.6908332705497742, "learning_rate": 9.999386404583899e-05, "loss": 0.0985, "step": 4380 }, { "grad_norm": 0.6797628402709961, "learning_rate": 9.999353598412568e-05, "loss": 0.0941, "step": 4390 }, { "grad_norm": 0.6660715341567993, "learning_rate": 9.999319937988442e-05, "loss": 0.1106, "step": 4400 }, { "grad_norm": 0.5183005332946777, "learning_rate": 9.999285423317268e-05, "loss": 0.1039, "step": 4410 }, { "grad_norm": 0.7903044819831848, "learning_rate": 9.999250054404947e-05, "loss": 0.1214, "step": 4420 }, { "grad_norm": 0.870417594909668, "learning_rate": 9.99921383125752e-05, "loss": 0.1068, "step": 4430 }, { "grad_norm": 0.6724756956100464, "learning_rate": 9.99917675388118e-05, "loss": 0.1155, "step": 4440 }, { "grad_norm": 0.6742640733718872, "learning_rate": 9.99913882228226e-05, "loss": 0.113, "step": 4450 }, { "grad_norm": 0.6911227107048035, "learning_rate": 9.999100036467242e-05, "loss": 0.1066, "step": 4460 }, { "grad_norm": 0.561302900314331, "learning_rate": 9.999060396442753e-05, "loss": 0.0915, "step": 4470 }, { "grad_norm": 0.6041096448898315, "learning_rate": 9.999019902215566e-05, "loss": 0.0982, "step": 4480 }, { "grad_norm": 0.7279731035232544, "learning_rate": 9.998978553792602e-05, "loss": 0.1125, "step": 4490 }, { "grad_norm": 0.9391347765922546, "learning_rate": 9.998936351180926e-05, "loss": 0.1128, "step": 4500 }, { "grad_norm": 0.6217988133430481, "learning_rate": 9.998893294387747e-05, "loss": 0.1139, "step": 4510 }, { "grad_norm": 0.8541237115859985, "learning_rate": 9.998849383420426e-05, "loss": 0.1075, "step": 4520 }, { "grad_norm": 0.5951516032218933, "learning_rate": 9.998804618286465e-05, "loss": 0.1117, "step": 4530 }, { "grad_norm": 0.7563536763191223, "learning_rate": 9.99875899899351e-05, "loss": 0.121, "step": 4540 }, { "grad_norm": 0.625174880027771, "learning_rate": 9.99871252554936e-05, "loss": 0.1086, "step": 4550 }, { "grad_norm": 0.6306079626083374, "learning_rate": 9.998665197961955e-05, "loss": 0.1014, "step": 4560 }, { "grad_norm": 0.7365911602973938, "learning_rate": 9.998617016239379e-05, "loss": 0.1079, "step": 4570 }, { "grad_norm": 1.1544067859649658, "learning_rate": 9.998567980389869e-05, "loss": 0.0954, "step": 4580 }, { "grad_norm": 0.762345016002655, "learning_rate": 9.998518090421802e-05, "loss": 0.1066, "step": 4590 }, { "grad_norm": 0.7564781904220581, "learning_rate": 9.998467346343703e-05, "loss": 0.0983, "step": 4600 }, { "grad_norm": 0.6269709467887878, "learning_rate": 9.998415748164243e-05, "loss": 0.0975, "step": 4610 }, { "grad_norm": 0.7456287741661072, "learning_rate": 9.998363295892238e-05, "loss": 0.093, "step": 4620 }, { "grad_norm": 0.724442720413208, "learning_rate": 9.998309989536652e-05, "loss": 0.1087, "step": 4630 }, { "grad_norm": 0.6254262328147888, "learning_rate": 9.998255829106593e-05, "loss": 0.105, "step": 4640 }, { "grad_norm": 0.7582572102546692, "learning_rate": 9.998200814611316e-05, "loss": 0.0892, "step": 4650 }, { "grad_norm": 0.7670857310295105, "learning_rate": 9.998144946060219e-05, "loss": 0.1114, "step": 4660 }, { "grad_norm": 0.8679442405700684, "learning_rate": 9.998088223462852e-05, "loss": 0.0956, "step": 4670 }, { "grad_norm": 0.7250534892082214, "learning_rate": 9.998030646828905e-05, "loss": 0.1, "step": 4680 }, { "grad_norm": 0.783789336681366, "learning_rate": 9.997972216168217e-05, "loss": 0.1024, "step": 4690 }, { "grad_norm": 0.5240048766136169, "learning_rate": 9.997912931490771e-05, "loss": 0.0893, "step": 4700 }, { "grad_norm": 0.624828040599823, "learning_rate": 9.9978527928067e-05, "loss": 0.0888, "step": 4710 }, { "grad_norm": 0.5406963229179382, "learning_rate": 9.997791800126277e-05, "loss": 0.0912, "step": 4720 }, { "grad_norm": 0.6690589785575867, "learning_rate": 9.997729953459927e-05, "loss": 0.1004, "step": 4730 }, { "grad_norm": 0.6017341613769531, "learning_rate": 9.997667252818214e-05, "loss": 0.1031, "step": 4740 }, { "grad_norm": 0.514665961265564, "learning_rate": 9.997603698211855e-05, "loss": 0.0867, "step": 4750 }, { "grad_norm": 0.812982976436615, "learning_rate": 9.99753928965171e-05, "loss": 0.0926, "step": 4760 }, { "grad_norm": 0.6571866869926453, "learning_rate": 9.997474027148781e-05, "loss": 0.1083, "step": 4770 }, { "grad_norm": 0.8686188459396362, "learning_rate": 9.997407910714223e-05, "loss": 0.1239, "step": 4780 }, { "grad_norm": 0.5965883731842041, "learning_rate": 9.997340940359332e-05, "loss": 0.0849, "step": 4790 }, { "grad_norm": 0.538587749004364, "learning_rate": 9.997273116095552e-05, "loss": 0.0873, "step": 4800 }, { "grad_norm": 0.539819061756134, "learning_rate": 9.997204437934473e-05, "loss": 0.094, "step": 4810 }, { "grad_norm": 0.6976571083068848, "learning_rate": 9.997134905887829e-05, "loss": 0.1032, "step": 4820 }, { "grad_norm": 0.5441659688949585, "learning_rate": 9.997064519967501e-05, "loss": 0.1038, "step": 4830 }, { "grad_norm": 0.7157810926437378, "learning_rate": 9.996993280185517e-05, "loss": 0.0945, "step": 4840 }, { "grad_norm": 0.6999992728233337, "learning_rate": 9.99692118655405e-05, "loss": 0.0857, "step": 4850 }, { "grad_norm": 0.5045496821403503, "learning_rate": 9.996848239085417e-05, "loss": 0.083, "step": 4860 }, { "grad_norm": 0.8596316576004028, "learning_rate": 9.996774437792085e-05, "loss": 0.0922, "step": 4870 }, { "grad_norm": 0.583432674407959, "learning_rate": 9.996699782686664e-05, "loss": 0.0908, "step": 4880 }, { "grad_norm": 0.863905131816864, "learning_rate": 9.996624273781909e-05, "loss": 0.1049, "step": 4890 }, { "grad_norm": 0.729310154914856, "learning_rate": 9.996547911090725e-05, "loss": 0.088, "step": 4900 }, { "grad_norm": 0.5281977653503418, "learning_rate": 9.996470694626157e-05, "loss": 0.0977, "step": 4910 }, { "grad_norm": 0.5110204219818115, "learning_rate": 9.996392624401403e-05, "loss": 0.1082, "step": 4920 }, { "grad_norm": 0.5242742896080017, "learning_rate": 9.996313700429801e-05, "loss": 0.0948, "step": 4930 }, { "grad_norm": 0.5709401965141296, "learning_rate": 9.996233922724836e-05, "loss": 0.0813, "step": 4940 }, { "grad_norm": 0.7085327506065369, "learning_rate": 9.996153291300141e-05, "loss": 0.0955, "step": 4950 }, { "grad_norm": 0.6172772645950317, "learning_rate": 9.996071806169494e-05, "loss": 0.101, "step": 4960 }, { "grad_norm": 0.6315256357192993, "learning_rate": 9.995989467346817e-05, "loss": 0.1182, "step": 4970 }, { "grad_norm": 0.597316324710846, "learning_rate": 9.995906274846183e-05, "loss": 0.1033, "step": 4980 }, { "grad_norm": 0.6045640707015991, "learning_rate": 9.995822228681803e-05, "loss": 0.1005, "step": 4990 }, { "grad_norm": 0.6864244341850281, "learning_rate": 9.99573732886804e-05, "loss": 0.0971, "step": 5000 }, { "grad_norm": 0.6878169178962708, "learning_rate": 9.995651575419402e-05, "loss": 0.1032, "step": 5010 }, { "grad_norm": 0.6954208016395569, "learning_rate": 9.995564968350541e-05, "loss": 0.0976, "step": 5020 }, { "grad_norm": 0.7564736604690552, "learning_rate": 9.995477507676256e-05, "loss": 0.088, "step": 5030 }, { "grad_norm": 0.5695114731788635, "learning_rate": 9.995389193411493e-05, "loss": 0.0874, "step": 5040 }, { "grad_norm": 0.494892418384552, "learning_rate": 9.995300025571339e-05, "loss": 0.0875, "step": 5050 }, { "grad_norm": 0.5463091135025024, "learning_rate": 9.995210004171034e-05, "loss": 0.071, "step": 5060 }, { "grad_norm": 0.6044389605522156, "learning_rate": 9.995119129225956e-05, "loss": 0.1333, "step": 5070 }, { "grad_norm": 0.5507385730743408, "learning_rate": 9.995027400751637e-05, "loss": 0.1128, "step": 5080 }, { "grad_norm": 0.6101757884025574, "learning_rate": 9.994934818763751e-05, "loss": 0.0827, "step": 5090 }, { "grad_norm": 0.6691555380821228, "learning_rate": 9.994841383278115e-05, "loss": 0.09, "step": 5100 }, { "grad_norm": 0.5066030621528625, "learning_rate": 9.994747094310695e-05, "loss": 0.097, "step": 5110 }, { "grad_norm": 0.5511738061904907, "learning_rate": 9.994651951877604e-05, "loss": 0.1167, "step": 5120 }, { "grad_norm": 0.8133506774902344, "learning_rate": 9.994555955995099e-05, "loss": 0.1104, "step": 5130 }, { "grad_norm": 0.7519100308418274, "learning_rate": 9.994459106679581e-05, "loss": 0.1009, "step": 5140 }, { "grad_norm": 0.6145743727684021, "learning_rate": 9.994361403947603e-05, "loss": 0.0919, "step": 5150 }, { "grad_norm": 0.7198712825775146, "learning_rate": 9.994262847815854e-05, "loss": 0.1194, "step": 5160 }, { "grad_norm": 0.6399878859519958, "learning_rate": 9.99416343830118e-05, "loss": 0.0923, "step": 5170 }, { "grad_norm": 0.6226606965065002, "learning_rate": 9.994063175420565e-05, "loss": 0.0848, "step": 5180 }, { "grad_norm": 0.5749218463897705, "learning_rate": 9.99396205919114e-05, "loss": 0.1136, "step": 5190 }, { "grad_norm": 0.5506098866462708, "learning_rate": 9.993860089630185e-05, "loss": 0.0777, "step": 5200 }, { "grad_norm": 0.7908716201782227, "learning_rate": 9.993757266755123e-05, "loss": 0.105, "step": 5210 }, { "grad_norm": 0.7023361921310425, "learning_rate": 9.993653590583522e-05, "loss": 0.0944, "step": 5220 }, { "grad_norm": 0.4785168766975403, "learning_rate": 9.993549061133102e-05, "loss": 0.0846, "step": 5230 }, { "grad_norm": 0.5928329825401306, "learning_rate": 9.993443678421719e-05, "loss": 0.0936, "step": 5240 }, { "grad_norm": 0.5552789568901062, "learning_rate": 9.993337442467384e-05, "loss": 0.1097, "step": 5250 }, { "grad_norm": 0.7582218050956726, "learning_rate": 9.993230353288248e-05, "loss": 0.0853, "step": 5260 }, { "grad_norm": 0.7792263627052307, "learning_rate": 9.993122410902608e-05, "loss": 0.1061, "step": 5270 }, { "grad_norm": 0.9738714098930359, "learning_rate": 9.993013615328912e-05, "loss": 0.1053, "step": 5280 }, { "grad_norm": 0.5658345818519592, "learning_rate": 9.992903966585747e-05, "loss": 0.0998, "step": 5290 }, { "grad_norm": 0.6523349285125732, "learning_rate": 9.992793464691852e-05, "loss": 0.0788, "step": 5300 }, { "grad_norm": 0.4318457245826721, "learning_rate": 9.992682109666105e-05, "loss": 0.0829, "step": 5310 }, { "grad_norm": 0.7916048765182495, "learning_rate": 9.992569901527538e-05, "loss": 0.094, "step": 5320 }, { "grad_norm": 0.6421202421188354, "learning_rate": 9.99245684029532e-05, "loss": 0.1025, "step": 5330 }, { "grad_norm": 0.5853082537651062, "learning_rate": 9.992342925988774e-05, "loss": 0.0908, "step": 5340 }, { "grad_norm": 0.5042808651924133, "learning_rate": 9.992228158627361e-05, "loss": 0.0763, "step": 5350 }, { "grad_norm": 0.6050172448158264, "learning_rate": 9.992112538230693e-05, "loss": 0.0908, "step": 5360 }, { "grad_norm": 0.5845838189125061, "learning_rate": 9.991996064818527e-05, "loss": 0.0725, "step": 5370 }, { "grad_norm": 0.4302046298980713, "learning_rate": 9.991878738410768e-05, "loss": 0.077, "step": 5380 }, { "grad_norm": 0.5815016031265259, "learning_rate": 9.991760559027457e-05, "loss": 0.0758, "step": 5390 }, { "grad_norm": 0.5468171238899231, "learning_rate": 9.991641526688793e-05, "loss": 0.0952, "step": 5400 }, { "grad_norm": 0.7248405814170837, "learning_rate": 9.991521641415113e-05, "loss": 0.0915, "step": 5410 }, { "grad_norm": 0.6060589551925659, "learning_rate": 9.991400903226904e-05, "loss": 0.0866, "step": 5420 }, { "grad_norm": 0.7981271743774414, "learning_rate": 9.991279312144794e-05, "loss": 0.0865, "step": 5430 }, { "grad_norm": 0.579706609249115, "learning_rate": 9.991156868189564e-05, "loss": 0.1041, "step": 5440 }, { "grad_norm": 0.5968762636184692, "learning_rate": 9.991033571382131e-05, "loss": 0.0785, "step": 5450 }, { "grad_norm": 0.7235113382339478, "learning_rate": 9.990909421743569e-05, "loss": 0.0779, "step": 5460 }, { "grad_norm": 0.544607937335968, "learning_rate": 9.990784419295085e-05, "loss": 0.0849, "step": 5470 }, { "grad_norm": 0.5355772376060486, "learning_rate": 9.990658564058044e-05, "loss": 0.0765, "step": 5480 }, { "grad_norm": 0.5730797052383423, "learning_rate": 9.990531856053948e-05, "loss": 0.089, "step": 5490 }, { "grad_norm": 0.6055258512496948, "learning_rate": 9.99040429530445e-05, "loss": 0.0864, "step": 5500 }, { "grad_norm": 0.6526451706886292, "learning_rate": 9.990275881831346e-05, "loss": 0.0861, "step": 5510 }, { "grad_norm": 0.4366335868835449, "learning_rate": 9.990146615656577e-05, "loss": 0.0661, "step": 5520 }, { "grad_norm": 0.4900181293487549, "learning_rate": 9.990016496802233e-05, "loss": 0.0788, "step": 5530 }, { "grad_norm": 0.6177137494087219, "learning_rate": 9.989885525290548e-05, "loss": 0.0803, "step": 5540 }, { "grad_norm": 0.7926065921783447, "learning_rate": 9.989753701143897e-05, "loss": 0.0975, "step": 5550 }, { "grad_norm": 0.40347418189048767, "learning_rate": 9.989621024384812e-05, "loss": 0.0889, "step": 5560 }, { "grad_norm": 0.6479153037071228, "learning_rate": 9.989487495035959e-05, "loss": 0.0876, "step": 5570 }, { "grad_norm": 0.7795159816741943, "learning_rate": 9.989353113120156e-05, "loss": 0.0984, "step": 5580 }, { "grad_norm": 0.5587451457977295, "learning_rate": 9.989217878660366e-05, "loss": 0.0853, "step": 5590 }, { "grad_norm": 0.754899799823761, "learning_rate": 9.989081791679695e-05, "loss": 0.1003, "step": 5600 }, { "grad_norm": 0.5546886324882507, "learning_rate": 9.988944852201397e-05, "loss": 0.0802, "step": 5610 }, { "grad_norm": 0.5835463404655457, "learning_rate": 9.988807060248873e-05, "loss": 0.0799, "step": 5620 }, { "grad_norm": 0.6488384008407593, "learning_rate": 9.988668415845665e-05, "loss": 0.0807, "step": 5630 }, { "grad_norm": 0.5939942598342896, "learning_rate": 9.988528919015466e-05, "loss": 0.0797, "step": 5640 }, { "grad_norm": 0.4799278974533081, "learning_rate": 9.988388569782112e-05, "loss": 0.0924, "step": 5650 }, { "grad_norm": 0.6716076135635376, "learning_rate": 9.988247368169583e-05, "loss": 0.0845, "step": 5660 }, { "grad_norm": 0.5582509636878967, "learning_rate": 9.988105314202007e-05, "loss": 0.0863, "step": 5670 }, { "grad_norm": 0.5180999636650085, "learning_rate": 9.987962407903659e-05, "loss": 0.1035, "step": 5680 }, { "grad_norm": 0.6384917497634888, "learning_rate": 9.987818649298957e-05, "loss": 0.086, "step": 5690 }, { "grad_norm": 0.6923719644546509, "learning_rate": 9.987674038412465e-05, "loss": 0.0883, "step": 5700 }, { "grad_norm": 0.5110870599746704, "learning_rate": 9.987528575268891e-05, "loss": 0.1088, "step": 5710 }, { "grad_norm": 0.7042368054389954, "learning_rate": 9.987382259893095e-05, "loss": 0.092, "step": 5720 }, { "grad_norm": 0.6341381669044495, "learning_rate": 9.987235092310074e-05, "loss": 0.0921, "step": 5730 }, { "grad_norm": 0.47959277033805847, "learning_rate": 9.987087072544978e-05, "loss": 0.0945, "step": 5740 }, { "grad_norm": 0.5953345894813538, "learning_rate": 9.9869382006231e-05, "loss": 0.0746, "step": 5750 }, { "grad_norm": 0.6727377772331238, "learning_rate": 9.986788476569875e-05, "loss": 0.0845, "step": 5760 }, { "grad_norm": 0.5924807786941528, "learning_rate": 9.986637900410887e-05, "loss": 0.0876, "step": 5770 }, { "grad_norm": 0.5657296180725098, "learning_rate": 9.986486472171869e-05, "loss": 0.1087, "step": 5780 }, { "grad_norm": 0.6226595044136047, "learning_rate": 9.986334191878692e-05, "loss": 0.0838, "step": 5790 }, { "grad_norm": 0.7165924310684204, "learning_rate": 9.986181059557378e-05, "loss": 0.0947, "step": 5800 }, { "grad_norm": 0.9753275513648987, "learning_rate": 9.986027075234094e-05, "loss": 0.0908, "step": 5810 }, { "grad_norm": 0.6576001644134521, "learning_rate": 9.985872238935152e-05, "loss": 0.0815, "step": 5820 }, { "grad_norm": 0.45263931155204773, "learning_rate": 9.985716550687008e-05, "loss": 0.0867, "step": 5830 }, { "grad_norm": 0.6188015937805176, "learning_rate": 9.985560010516264e-05, "loss": 0.0765, "step": 5840 }, { "grad_norm": 0.6847587823867798, "learning_rate": 9.985402618449668e-05, "loss": 0.0811, "step": 5850 }, { "grad_norm": 0.6032217741012573, "learning_rate": 9.985244374514118e-05, "loss": 0.1033, "step": 5860 }, { "grad_norm": 0.6631447672843933, "learning_rate": 9.985085278736651e-05, "loss": 0.0904, "step": 5870 }, { "grad_norm": 0.6037219166755676, "learning_rate": 9.984925331144452e-05, "loss": 0.0987, "step": 5880 }, { "grad_norm": 0.44121643900871277, "learning_rate": 9.984764531764851e-05, "loss": 0.0914, "step": 5890 }, { "grad_norm": 0.5306256413459778, "learning_rate": 9.984602880625326e-05, "loss": 0.1148, "step": 5900 }, { "grad_norm": 0.588921308517456, "learning_rate": 9.9844403777535e-05, "loss": 0.1009, "step": 5910 }, { "grad_norm": 0.861057698726654, "learning_rate": 9.984277023177135e-05, "loss": 0.0799, "step": 5920 }, { "grad_norm": 0.5513086915016174, "learning_rate": 9.984112816924148e-05, "loss": 0.071, "step": 5930 }, { "grad_norm": 0.5905616879463196, "learning_rate": 9.983947759022596e-05, "loss": 0.0953, "step": 5940 }, { "grad_norm": 0.5747668147087097, "learning_rate": 9.983781849500682e-05, "loss": 0.0833, "step": 5950 }, { "grad_norm": 0.4530848562717438, "learning_rate": 9.98361508838676e-05, "loss": 0.0753, "step": 5960 }, { "grad_norm": 0.7012318968772888, "learning_rate": 9.98344747570932e-05, "loss": 0.0908, "step": 5970 }, { "grad_norm": 0.5476173162460327, "learning_rate": 9.983279011497004e-05, "loss": 0.0958, "step": 5980 }, { "grad_norm": 0.6118975281715393, "learning_rate": 9.983109695778596e-05, "loss": 0.1092, "step": 5990 }, { "grad_norm": 0.7510342597961426, "learning_rate": 9.982939528583032e-05, "loss": 0.1241, "step": 6000 }, { "grad_norm": 0.6679460406303406, "learning_rate": 9.982768509939385e-05, "loss": 0.1057, "step": 6010 }, { "grad_norm": 0.5381629467010498, "learning_rate": 9.982596639876879e-05, "loss": 0.0981, "step": 6020 }, { "grad_norm": 0.5261770486831665, "learning_rate": 9.982423918424881e-05, "loss": 0.0886, "step": 6030 }, { "grad_norm": 0.6286594867706299, "learning_rate": 9.982250345612908e-05, "loss": 0.0898, "step": 6040 }, { "grad_norm": 0.4993520975112915, "learning_rate": 9.982075921470611e-05, "loss": 0.0962, "step": 6050 }, { "grad_norm": 0.6804307699203491, "learning_rate": 9.981900646027802e-05, "loss": 0.086, "step": 6060 }, { "grad_norm": 0.7932005524635315, "learning_rate": 9.981724519314425e-05, "loss": 0.0984, "step": 6070 }, { "grad_norm": 0.6358413696289062, "learning_rate": 9.981547541360581e-05, "loss": 0.1049, "step": 6080 }, { "grad_norm": 0.5939364433288574, "learning_rate": 9.981369712196508e-05, "loss": 0.0974, "step": 6090 }, { "grad_norm": 0.5588685274124146, "learning_rate": 9.981191031852592e-05, "loss": 0.0995, "step": 6100 }, { "grad_norm": 0.5709642767906189, "learning_rate": 9.981011500359362e-05, "loss": 0.0829, "step": 6110 }, { "grad_norm": 0.7823707461357117, "learning_rate": 9.9808311177475e-05, "loss": 0.0938, "step": 6120 }, { "grad_norm": 0.5275198221206665, "learning_rate": 9.980649884047826e-05, "loss": 0.0803, "step": 6130 }, { "grad_norm": 0.6380882859230042, "learning_rate": 9.980467799291307e-05, "loss": 0.0839, "step": 6140 }, { "grad_norm": 0.5099921226501465, "learning_rate": 9.980284863509058e-05, "loss": 0.1136, "step": 6150 }, { "grad_norm": 0.7005736827850342, "learning_rate": 9.980101076732334e-05, "loss": 0.1076, "step": 6160 }, { "grad_norm": 0.6694428324699402, "learning_rate": 9.979916438992544e-05, "loss": 0.0899, "step": 6170 }, { "grad_norm": 0.4303048551082611, "learning_rate": 9.979730950321237e-05, "loss": 0.0711, "step": 6180 }, { "grad_norm": 0.5969268679618835, "learning_rate": 9.979544610750104e-05, "loss": 0.0977, "step": 6190 }, { "grad_norm": 0.5547540783882141, "learning_rate": 9.97935742031099e-05, "loss": 0.092, "step": 6200 }, { "grad_norm": 0.988711416721344, "learning_rate": 9.979169379035878e-05, "loss": 0.097, "step": 6210 }, { "grad_norm": 0.5021083950996399, "learning_rate": 9.978980486956899e-05, "loss": 0.088, "step": 6220 }, { "grad_norm": 0.6447035074234009, "learning_rate": 9.978790744106332e-05, "loss": 0.0883, "step": 6230 }, { "grad_norm": 0.5419172048568726, "learning_rate": 9.978600150516594e-05, "loss": 0.0969, "step": 6240 }, { "grad_norm": 0.4582507312297821, "learning_rate": 9.978408706220259e-05, "loss": 0.0924, "step": 6250 }, { "grad_norm": 0.5601506233215332, "learning_rate": 9.978216411250032e-05, "loss": 0.0802, "step": 6260 }, { "grad_norm": 0.7054854035377502, "learning_rate": 9.978023265638778e-05, "loss": 0.0809, "step": 6270 }, { "grad_norm": 0.5286251306533813, "learning_rate": 9.977829269419495e-05, "loss": 0.087, "step": 6280 }, { "grad_norm": 0.5873297452926636, "learning_rate": 9.977634422625335e-05, "loss": 0.0786, "step": 6290 }, { "grad_norm": 0.5765334367752075, "learning_rate": 9.97743872528959e-05, "loss": 0.0842, "step": 6300 }, { "grad_norm": 0.6363047957420349, "learning_rate": 9.9772421774457e-05, "loss": 0.0716, "step": 6310 }, { "grad_norm": 0.5545669198036194, "learning_rate": 9.977044779127252e-05, "loss": 0.0825, "step": 6320 }, { "grad_norm": 0.806815505027771, "learning_rate": 9.976846530367971e-05, "loss": 0.1003, "step": 6330 }, { "grad_norm": 0.6564372777938843, "learning_rate": 9.976647431201735e-05, "loss": 0.0969, "step": 6340 }, { "grad_norm": 0.7408301830291748, "learning_rate": 9.976447481662568e-05, "loss": 0.0892, "step": 6350 }, { "grad_norm": 0.5188537240028381, "learning_rate": 9.976246681784629e-05, "loss": 0.0948, "step": 6360 }, { "grad_norm": 0.6625853180885315, "learning_rate": 9.976045031602234e-05, "loss": 0.0887, "step": 6370 }, { "grad_norm": 0.6388776302337646, "learning_rate": 9.975842531149837e-05, "loss": 0.0895, "step": 6380 }, { "grad_norm": 0.6781452298164368, "learning_rate": 9.975639180462043e-05, "loss": 0.0973, "step": 6390 }, { "grad_norm": 0.5789165496826172, "learning_rate": 9.975434979573596e-05, "loss": 0.0903, "step": 6400 }, { "grad_norm": 0.5901700854301453, "learning_rate": 9.97522992851939e-05, "loss": 0.0665, "step": 6410 }, { "grad_norm": 0.49265411496162415, "learning_rate": 9.975024027334461e-05, "loss": 0.0798, "step": 6420 }, { "grad_norm": 0.5130177736282349, "learning_rate": 9.974817276053993e-05, "loss": 0.0803, "step": 6430 }, { "grad_norm": 0.9185003042221069, "learning_rate": 9.974609674713315e-05, "loss": 0.0884, "step": 6440 }, { "grad_norm": 0.6996639966964722, "learning_rate": 9.9744012233479e-05, "loss": 0.0836, "step": 6450 }, { "grad_norm": 0.48580166697502136, "learning_rate": 9.974191921993366e-05, "loss": 0.0851, "step": 6460 }, { "grad_norm": 0.47902533411979675, "learning_rate": 9.973981770685474e-05, "loss": 0.0906, "step": 6470 }, { "grad_norm": 0.5141099691390991, "learning_rate": 9.97377076946014e-05, "loss": 0.0777, "step": 6480 }, { "grad_norm": 0.7393907904624939, "learning_rate": 9.973558918353412e-05, "loss": 0.0898, "step": 6490 }, { "grad_norm": 0.5084715485572815, "learning_rate": 9.973346217401494e-05, "loss": 0.0843, "step": 6500 }, { "grad_norm": 0.626142144203186, "learning_rate": 9.973132666640726e-05, "loss": 0.0764, "step": 6510 }, { "grad_norm": 0.6941993832588196, "learning_rate": 9.972918266107602e-05, "loss": 0.0897, "step": 6520 }, { "grad_norm": 0.5942189693450928, "learning_rate": 9.972703015838756e-05, "loss": 0.0758, "step": 6530 }, { "grad_norm": 0.4647875726222992, "learning_rate": 9.97248691587097e-05, "loss": 0.0967, "step": 6540 }, { "grad_norm": 0.6835445761680603, "learning_rate": 9.972269966241166e-05, "loss": 0.0747, "step": 6550 }, { "grad_norm": 0.600824773311615, "learning_rate": 9.972052166986417e-05, "loss": 0.0898, "step": 6560 }, { "grad_norm": 0.6691086292266846, "learning_rate": 9.971833518143938e-05, "loss": 0.0947, "step": 6570 }, { "grad_norm": 0.6514070630073547, "learning_rate": 9.971614019751093e-05, "loss": 0.0713, "step": 6580 }, { "grad_norm": 0.5951218008995056, "learning_rate": 9.971393671845383e-05, "loss": 0.1055, "step": 6590 }, { "grad_norm": 0.716000497341156, "learning_rate": 9.971172474464464e-05, "loss": 0.0872, "step": 6600 }, { "grad_norm": 0.6435450315475464, "learning_rate": 9.97095042764613e-05, "loss": 0.072, "step": 6610 }, { "grad_norm": 0.7254830598831177, "learning_rate": 9.970727531428324e-05, "loss": 0.0775, "step": 6620 }, { "grad_norm": 0.5940116047859192, "learning_rate": 9.970503785849132e-05, "loss": 0.0822, "step": 6630 }, { "grad_norm": 0.4756982922554016, "learning_rate": 9.970279190946788e-05, "loss": 0.0656, "step": 6640 }, { "grad_norm": 0.5552905201911926, "learning_rate": 9.970053746759667e-05, "loss": 0.093, "step": 6650 }, { "grad_norm": 0.4415166974067688, "learning_rate": 9.969827453326292e-05, "loss": 0.0772, "step": 6660 }, { "grad_norm": 0.5142070055007935, "learning_rate": 9.969600310685332e-05, "loss": 0.075, "step": 6670 }, { "grad_norm": 0.6408231258392334, "learning_rate": 9.969372318875596e-05, "loss": 0.0978, "step": 6680 }, { "grad_norm": 0.9072073698043823, "learning_rate": 9.969143477936043e-05, "loss": 0.0903, "step": 6690 }, { "grad_norm": 0.49139076471328735, "learning_rate": 9.968913787905775e-05, "loss": 0.0885, "step": 6700 }, { "grad_norm": 0.5735374093055725, "learning_rate": 9.968683248824045e-05, "loss": 0.0939, "step": 6710 }, { "grad_norm": 0.4863685965538025, "learning_rate": 9.968451860730238e-05, "loss": 0.099, "step": 6720 }, { "grad_norm": 0.5857629776000977, "learning_rate": 9.968219623663896e-05, "loss": 0.0783, "step": 6730 }, { "grad_norm": 0.44635188579559326, "learning_rate": 9.967986537664702e-05, "loss": 0.1137, "step": 6740 }, { "grad_norm": 0.5553225874900818, "learning_rate": 9.967752602772483e-05, "loss": 0.0823, "step": 6750 }, { "grad_norm": 0.684736967086792, "learning_rate": 9.967517819027212e-05, "loss": 0.091, "step": 6760 }, { "grad_norm": 0.5828658938407898, "learning_rate": 9.967282186469009e-05, "loss": 0.0778, "step": 6770 }, { "grad_norm": 0.5681456327438354, "learning_rate": 9.967045705138135e-05, "loss": 0.0911, "step": 6780 }, { "grad_norm": 0.6602146029472351, "learning_rate": 9.966808375074998e-05, "loss": 0.0951, "step": 6790 }, { "grad_norm": 0.7860845923423767, "learning_rate": 9.966570196320154e-05, "loss": 0.1075, "step": 6800 }, { "grad_norm": 0.5823718905448914, "learning_rate": 9.966331168914299e-05, "loss": 0.0711, "step": 6810 }, { "grad_norm": 0.553925096988678, "learning_rate": 9.966091292898277e-05, "loss": 0.0962, "step": 6820 }, { "grad_norm": 0.5356718897819519, "learning_rate": 9.965850568313076e-05, "loss": 0.0875, "step": 6830 }, { "grad_norm": 0.6481915712356567, "learning_rate": 9.965608995199827e-05, "loss": 0.0885, "step": 6840 }, { "grad_norm": 0.5665972828865051, "learning_rate": 9.965366573599812e-05, "loss": 0.0803, "step": 6850 }, { "grad_norm": 0.7178620100021362, "learning_rate": 9.965123303554453e-05, "loss": 0.0774, "step": 6860 }, { "grad_norm": 0.6748842597007751, "learning_rate": 9.964879185105317e-05, "loss": 0.1008, "step": 6870 }, { "grad_norm": 0.4321046769618988, "learning_rate": 9.964634218294119e-05, "loss": 0.0958, "step": 6880 }, { "grad_norm": 0.5556907653808594, "learning_rate": 9.964388403162714e-05, "loss": 0.0902, "step": 6890 }, { "grad_norm": 0.6041448712348938, "learning_rate": 9.96414173975311e-05, "loss": 0.0922, "step": 6900 }, { "grad_norm": 0.5850904583930969, "learning_rate": 9.963894228107451e-05, "loss": 0.0854, "step": 6910 }, { "grad_norm": 0.45640096068382263, "learning_rate": 9.963645868268032e-05, "loss": 0.0876, "step": 6920 }, { "grad_norm": 0.5964476466178894, "learning_rate": 9.963396660277289e-05, "loss": 0.1016, "step": 6930 }, { "grad_norm": 0.7682547569274902, "learning_rate": 9.963146604177807e-05, "loss": 0.0765, "step": 6940 }, { "grad_norm": 0.5879827737808228, "learning_rate": 9.962895700012311e-05, "loss": 0.0731, "step": 6950 }, { "grad_norm": 0.49877339601516724, "learning_rate": 9.962643947823677e-05, "loss": 0.0848, "step": 6960 }, { "grad_norm": 0.75579833984375, "learning_rate": 9.962391347654921e-05, "loss": 0.0798, "step": 6970 }, { "grad_norm": 0.7037603855133057, "learning_rate": 9.962137899549204e-05, "loss": 0.0945, "step": 6980 }, { "grad_norm": 0.6609319448471069, "learning_rate": 9.961883603549835e-05, "loss": 0.0901, "step": 6990 }, { "grad_norm": 0.60015869140625, "learning_rate": 9.961628459700267e-05, "loss": 0.0855, "step": 7000 }, { "grad_norm": 0.5790077447891235, "learning_rate": 9.961372468044095e-05, "loss": 0.0936, "step": 7010 }, { "grad_norm": 0.7288147807121277, "learning_rate": 9.961115628625062e-05, "loss": 0.0779, "step": 7020 }, { "grad_norm": 0.6946198344230652, "learning_rate": 9.960857941487056e-05, "loss": 0.109, "step": 7030 }, { "grad_norm": 0.6286454200744629, "learning_rate": 9.960599406674106e-05, "loss": 0.0835, "step": 7040 }, { "grad_norm": 0.5455347299575806, "learning_rate": 9.960340024230393e-05, "loss": 0.0774, "step": 7050 }, { "grad_norm": 0.5175648331642151, "learning_rate": 9.960079794200232e-05, "loss": 0.0818, "step": 7060 }, { "grad_norm": 0.5776930451393127, "learning_rate": 9.959818716628096e-05, "loss": 0.0823, "step": 7070 }, { "grad_norm": 0.638106107711792, "learning_rate": 9.95955679155859e-05, "loss": 0.1002, "step": 7080 }, { "grad_norm": 0.7558200359344482, "learning_rate": 9.959294019036472e-05, "loss": 0.089, "step": 7090 }, { "grad_norm": 0.6559443473815918, "learning_rate": 9.959030399106646e-05, "loss": 0.08, "step": 7100 }, { "grad_norm": 0.5221924781799316, "learning_rate": 9.958765931814153e-05, "loss": 0.092, "step": 7110 }, { "grad_norm": 0.5100626945495605, "learning_rate": 9.958500617204184e-05, "loss": 0.0761, "step": 7120 }, { "grad_norm": 0.4257447123527527, "learning_rate": 9.958234455322075e-05, "loss": 0.0783, "step": 7130 }, { "grad_norm": 0.5710707306861877, "learning_rate": 9.957967446213308e-05, "loss": 0.0925, "step": 7140 }, { "grad_norm": 0.49289223551750183, "learning_rate": 9.957699589923501e-05, "loss": 0.0897, "step": 7150 }, { "grad_norm": 1.0447601079940796, "learning_rate": 9.957430886498431e-05, "loss": 0.0971, "step": 7160 }, { "grad_norm": 0.56944739818573, "learning_rate": 9.957161335984008e-05, "loss": 0.0981, "step": 7170 }, { "grad_norm": 0.6015657186508179, "learning_rate": 9.956890938426291e-05, "loss": 0.0791, "step": 7180 }, { "grad_norm": 0.6126651167869568, "learning_rate": 9.956619693871482e-05, "loss": 0.0963, "step": 7190 }, { "grad_norm": 0.694652259349823, "learning_rate": 9.956347602365934e-05, "loss": 0.0815, "step": 7200 }, { "grad_norm": 0.6294214129447937, "learning_rate": 9.956074663956135e-05, "loss": 0.0825, "step": 7210 }, { "grad_norm": 0.6573867201805115, "learning_rate": 9.955800878688726e-05, "loss": 0.094, "step": 7220 }, { "grad_norm": 0.9771936535835266, "learning_rate": 9.955526246610489e-05, "loss": 0.0818, "step": 7230 }, { "grad_norm": 0.7391542196273804, "learning_rate": 9.955250767768349e-05, "loss": 0.0929, "step": 7240 }, { "grad_norm": 0.5137601494789124, "learning_rate": 9.95497444220938e-05, "loss": 0.0932, "step": 7250 }, { "grad_norm": 0.5528672933578491, "learning_rate": 9.954697269980797e-05, "loss": 0.0815, "step": 7260 }, { "grad_norm": 0.4785914123058319, "learning_rate": 9.954419251129962e-05, "loss": 0.0796, "step": 7270 }, { "grad_norm": 0.5979615449905396, "learning_rate": 9.95414038570438e-05, "loss": 0.0786, "step": 7280 }, { "grad_norm": 0.46930164098739624, "learning_rate": 9.953860673751703e-05, "loss": 0.0832, "step": 7290 }, { "grad_norm": 0.5151110291481018, "learning_rate": 9.953580115319725e-05, "loss": 0.0801, "step": 7300 }, { "grad_norm": 0.8184095025062561, "learning_rate": 9.953298710456387e-05, "loss": 0.0854, "step": 7310 }, { "grad_norm": 0.5467885732650757, "learning_rate": 9.953016459209771e-05, "loss": 0.0712, "step": 7320 }, { "grad_norm": 0.596036434173584, "learning_rate": 9.952733361628108e-05, "loss": 0.0794, "step": 7330 }, { "grad_norm": 0.6872668862342834, "learning_rate": 9.952449417759772e-05, "loss": 0.0858, "step": 7340 }, { "grad_norm": 0.6539696455001831, "learning_rate": 9.952164627653279e-05, "loss": 0.0808, "step": 7350 }, { "grad_norm": 0.5589662194252014, "learning_rate": 9.951878991357292e-05, "loss": 0.0925, "step": 7360 }, { "grad_norm": 0.4663572907447815, "learning_rate": 9.951592508920622e-05, "loss": 0.0823, "step": 7370 }, { "grad_norm": 0.5893650054931641, "learning_rate": 9.951305180392219e-05, "loss": 0.0785, "step": 7380 }, { "grad_norm": 0.6422627568244934, "learning_rate": 9.951017005821178e-05, "loss": 0.0829, "step": 7390 }, { "grad_norm": 0.6863510012626648, "learning_rate": 9.95072798525674e-05, "loss": 0.0903, "step": 7400 }, { "grad_norm": 0.849290668964386, "learning_rate": 9.950438118748293e-05, "loss": 0.0932, "step": 7410 }, { "grad_norm": 0.6894379258155823, "learning_rate": 9.950147406345366e-05, "loss": 0.1119, "step": 7420 }, { "grad_norm": 0.6132301092147827, "learning_rate": 9.949855848097635e-05, "loss": 0.0939, "step": 7430 }, { "grad_norm": 0.4760252833366394, "learning_rate": 9.949563444054916e-05, "loss": 0.0843, "step": 7440 }, { "grad_norm": 0.4669642746448517, "learning_rate": 9.949270194267178e-05, "loss": 0.0845, "step": 7450 }, { "grad_norm": 0.447228342294693, "learning_rate": 9.948976098784526e-05, "loss": 0.0891, "step": 7460 }, { "grad_norm": 0.5224108695983887, "learning_rate": 9.948681157657213e-05, "loss": 0.0868, "step": 7470 }, { "grad_norm": 0.5446855425834656, "learning_rate": 9.948385370935638e-05, "loss": 0.094, "step": 7480 }, { "grad_norm": 0.7080115079879761, "learning_rate": 9.94808873867034e-05, "loss": 0.0936, "step": 7490 }, { "grad_norm": 0.4475601613521576, "learning_rate": 9.947791260912009e-05, "loss": 0.0692, "step": 7500 }, { "grad_norm": 0.6576502323150635, "learning_rate": 9.947492937711474e-05, "loss": 0.0703, "step": 7510 }, { "grad_norm": 0.46500372886657715, "learning_rate": 9.947193769119707e-05, "loss": 0.0697, "step": 7520 }, { "grad_norm": 0.6065836548805237, "learning_rate": 9.946893755187834e-05, "loss": 0.0704, "step": 7530 }, { "grad_norm": 0.3981691598892212, "learning_rate": 9.946592895967115e-05, "loss": 0.0879, "step": 7540 }, { "grad_norm": 0.630163848400116, "learning_rate": 9.94629119150896e-05, "loss": 0.0818, "step": 7550 }, { "grad_norm": 0.6392139196395874, "learning_rate": 9.94598864186492e-05, "loss": 0.085, "step": 7560 }, { "grad_norm": 0.46114152669906616, "learning_rate": 9.945685247086696e-05, "loss": 0.0851, "step": 7570 }, { "grad_norm": 0.548022985458374, "learning_rate": 9.945381007226129e-05, "loss": 0.0844, "step": 7580 }, { "grad_norm": 0.7136789560317993, "learning_rate": 9.945075922335203e-05, "loss": 0.0919, "step": 7590 }, { "grad_norm": 0.6495311856269836, "learning_rate": 9.944769992466049e-05, "loss": 0.0963, "step": 7600 }, { "grad_norm": 0.5275823473930359, "learning_rate": 9.944463217670945e-05, "loss": 0.0952, "step": 7610 }, { "grad_norm": 0.4410739541053772, "learning_rate": 9.944155598002307e-05, "loss": 0.0882, "step": 7620 }, { "grad_norm": 0.5673771500587463, "learning_rate": 9.943847133512701e-05, "loss": 0.0869, "step": 7630 }, { "grad_norm": 0.8780397772789001, "learning_rate": 9.943537824254834e-05, "loss": 0.0932, "step": 7640 }, { "grad_norm": 0.590837299823761, "learning_rate": 9.943227670281559e-05, "loss": 0.0763, "step": 7650 }, { "grad_norm": 0.5749790072441101, "learning_rate": 9.942916671645873e-05, "loss": 0.0778, "step": 7660 }, { "grad_norm": 0.6229447722434998, "learning_rate": 9.942604828400916e-05, "loss": 0.0877, "step": 7670 }, { "grad_norm": 0.6113720536231995, "learning_rate": 9.942292140599975e-05, "loss": 0.0815, "step": 7680 }, { "grad_norm": 0.47927767038345337, "learning_rate": 9.94197860829648e-05, "loss": 0.0771, "step": 7690 }, { "grad_norm": 0.6479179263114929, "learning_rate": 9.941664231544004e-05, "loss": 0.0761, "step": 7700 }, { "grad_norm": 0.5829828977584839, "learning_rate": 9.941349010396264e-05, "loss": 0.0931, "step": 7710 }, { "grad_norm": 0.6902940273284912, "learning_rate": 9.941032944907125e-05, "loss": 0.0971, "step": 7720 }, { "grad_norm": 0.6796245574951172, "learning_rate": 9.940716035130596e-05, "loss": 0.0952, "step": 7730 }, { "grad_norm": 0.48314526677131653, "learning_rate": 9.940398281120821e-05, "loss": 0.0925, "step": 7740 }, { "grad_norm": 0.7225641012191772, "learning_rate": 9.940079682932102e-05, "loss": 0.0775, "step": 7750 }, { "grad_norm": 0.558650016784668, "learning_rate": 9.939760240618877e-05, "loss": 0.0864, "step": 7760 }, { "grad_norm": 0.4721904397010803, "learning_rate": 9.939439954235729e-05, "loss": 0.0801, "step": 7770 }, { "grad_norm": 0.4908062517642975, "learning_rate": 9.939118823837387e-05, "loss": 0.0881, "step": 7780 }, { "grad_norm": 0.7112247347831726, "learning_rate": 9.938796849478725e-05, "loss": 0.0816, "step": 7790 }, { "grad_norm": 0.4501150846481323, "learning_rate": 9.938474031214755e-05, "loss": 0.0748, "step": 7800 }, { "grad_norm": 0.5668996572494507, "learning_rate": 9.938150369100643e-05, "loss": 0.0805, "step": 7810 }, { "grad_norm": 0.40691065788269043, "learning_rate": 9.93782586319169e-05, "loss": 0.0575, "step": 7820 }, { "grad_norm": 0.5177756547927856, "learning_rate": 9.937500513543348e-05, "loss": 0.0675, "step": 7830 }, { "grad_norm": 0.660274088382721, "learning_rate": 9.937174320211207e-05, "loss": 0.0689, "step": 7840 }, { "grad_norm": 0.5574555397033691, "learning_rate": 9.936847283251009e-05, "loss": 0.08, "step": 7850 }, { "grad_norm": 0.5414860248565674, "learning_rate": 9.936519402718632e-05, "loss": 0.0832, "step": 7860 }, { "grad_norm": 0.6551302671432495, "learning_rate": 9.936190678670102e-05, "loss": 0.0865, "step": 7870 }, { "grad_norm": 0.4710102379322052, "learning_rate": 9.935861111161593e-05, "loss": 0.0791, "step": 7880 }, { "grad_norm": 0.4702812433242798, "learning_rate": 9.935530700249416e-05, "loss": 0.0692, "step": 7890 }, { "grad_norm": 0.5498860478401184, "learning_rate": 9.935199445990028e-05, "loss": 0.1019, "step": 7900 }, { "grad_norm": 0.6428347229957581, "learning_rate": 9.934867348440033e-05, "loss": 0.0907, "step": 7910 }, { "grad_norm": 0.5675632357597351, "learning_rate": 9.934534407656176e-05, "loss": 0.1143, "step": 7920 }, { "grad_norm": 0.7845016717910767, "learning_rate": 9.93420062369535e-05, "loss": 0.0867, "step": 7930 }, { "grad_norm": 0.4507503807544708, "learning_rate": 9.933865996614589e-05, "loss": 0.0763, "step": 7940 }, { "grad_norm": 0.6266567707061768, "learning_rate": 9.933530526471068e-05, "loss": 0.0681, "step": 7950 }, { "grad_norm": 0.5347049236297607, "learning_rate": 9.933194213322114e-05, "loss": 0.0744, "step": 7960 }, { "grad_norm": 0.5267623662948608, "learning_rate": 9.932857057225192e-05, "loss": 0.0795, "step": 7970 }, { "grad_norm": 0.6786030530929565, "learning_rate": 9.932519058237912e-05, "loss": 0.0821, "step": 7980 }, { "grad_norm": 0.560933530330658, "learning_rate": 9.932180216418032e-05, "loss": 0.0685, "step": 7990 }, { "grad_norm": 0.6568529605865479, "learning_rate": 9.931840531823446e-05, "loss": 0.083, "step": 8000 }, { "grad_norm": 0.5254671573638916, "learning_rate": 9.9315000045122e-05, "loss": 0.0698, "step": 8010 }, { "grad_norm": 0.4987260103225708, "learning_rate": 9.931158634542481e-05, "loss": 0.08, "step": 8020 }, { "grad_norm": 0.48918747901916504, "learning_rate": 9.930816421972617e-05, "loss": 0.0785, "step": 8030 }, { "grad_norm": 0.5406364798545837, "learning_rate": 9.930473366861086e-05, "loss": 0.0702, "step": 8040 }, { "grad_norm": 0.6343660354614258, "learning_rate": 9.930129469266505e-05, "loss": 0.0931, "step": 8050 }, { "grad_norm": 0.4606381058692932, "learning_rate": 9.929784729247638e-05, "loss": 0.0751, "step": 8060 }, { "grad_norm": 0.5130869150161743, "learning_rate": 9.929439146863389e-05, "loss": 0.0796, "step": 8070 }, { "grad_norm": 0.6414664387702942, "learning_rate": 9.92909272217281e-05, "loss": 0.1005, "step": 8080 }, { "grad_norm": 0.6709681749343872, "learning_rate": 9.928745455235097e-05, "loss": 0.0875, "step": 8090 }, { "grad_norm": 0.5680482983589172, "learning_rate": 9.928397346109588e-05, "loss": 0.0775, "step": 8100 }, { "grad_norm": 0.6708239912986755, "learning_rate": 9.928048394855762e-05, "loss": 0.0871, "step": 8110 }, { "grad_norm": 0.405321329832077, "learning_rate": 9.92769860153325e-05, "loss": 0.0752, "step": 8120 }, { "grad_norm": 0.48069751262664795, "learning_rate": 9.927347966201819e-05, "loss": 0.0957, "step": 8130 }, { "grad_norm": 0.6011830568313599, "learning_rate": 9.926996488921383e-05, "loss": 0.0856, "step": 8140 }, { "grad_norm": 0.5772767663002014, "learning_rate": 9.926644169752001e-05, "loss": 0.0735, "step": 8150 }, { "grad_norm": 0.6673007607460022, "learning_rate": 9.926291008753875e-05, "loss": 0.1045, "step": 8160 }, { "grad_norm": 0.541763961315155, "learning_rate": 9.92593700598735e-05, "loss": 0.0844, "step": 8170 }, { "grad_norm": 0.41507673263549805, "learning_rate": 9.925582161512915e-05, "loss": 0.0809, "step": 8180 }, { "grad_norm": 0.6382731199264526, "learning_rate": 9.925226475391205e-05, "loss": 0.0769, "step": 8190 }, { "grad_norm": 0.6581949591636658, "learning_rate": 9.924869947682993e-05, "loss": 0.0832, "step": 8200 }, { "grad_norm": 0.7050005793571472, "learning_rate": 9.924512578449204e-05, "loss": 0.0809, "step": 8210 }, { "grad_norm": 0.5828319787979126, "learning_rate": 9.924154367750901e-05, "loss": 0.081, "step": 8220 }, { "grad_norm": 0.5671278834342957, "learning_rate": 9.923795315649293e-05, "loss": 0.0823, "step": 8230 }, { "grad_norm": 0.48076727986335754, "learning_rate": 9.92343542220573e-05, "loss": 0.0794, "step": 8240 }, { "grad_norm": 0.4607784152030945, "learning_rate": 9.92307468748171e-05, "loss": 0.0726, "step": 8250 }, { "grad_norm": 0.5762013792991638, "learning_rate": 9.922713111538873e-05, "loss": 0.0626, "step": 8260 }, { "grad_norm": 0.49484574794769287, "learning_rate": 9.922350694439003e-05, "loss": 0.0679, "step": 8270 }, { "grad_norm": 0.7395768165588379, "learning_rate": 9.921987436244024e-05, "loss": 0.0929, "step": 8280 }, { "grad_norm": 0.6062762141227722, "learning_rate": 9.921623337016008e-05, "loss": 0.0697, "step": 8290 }, { "grad_norm": 0.4801269769668579, "learning_rate": 9.921258396817172e-05, "loss": 0.0879, "step": 8300 }, { "grad_norm": 0.6565815210342407, "learning_rate": 9.920892615709874e-05, "loss": 0.0734, "step": 8310 }, { "grad_norm": 0.5539771914482117, "learning_rate": 9.920525993756612e-05, "loss": 0.0658, "step": 8320 }, { "grad_norm": 0.4040979743003845, "learning_rate": 9.920158531020036e-05, "loss": 0.0685, "step": 8330 }, { "grad_norm": 0.9375623464584351, "learning_rate": 9.919790227562933e-05, "loss": 0.0692, "step": 8340 }, { "grad_norm": 0.5704438090324402, "learning_rate": 9.919421083448237e-05, "loss": 0.0855, "step": 8350 }, { "grad_norm": 0.5858630537986755, "learning_rate": 9.919051098739022e-05, "loss": 0.0773, "step": 8360 }, { "grad_norm": 0.4432525932788849, "learning_rate": 9.918680273498514e-05, "loss": 0.0738, "step": 8370 }, { "grad_norm": 0.5864804983139038, "learning_rate": 9.918308607790072e-05, "loss": 0.075, "step": 8380 }, { "grad_norm": 0.518041729927063, "learning_rate": 9.917936101677205e-05, "loss": 0.0749, "step": 8390 }, { "grad_norm": 0.581553041934967, "learning_rate": 9.917562755223564e-05, "loss": 0.0734, "step": 8400 }, { "grad_norm": 0.5372946262359619, "learning_rate": 9.917188568492944e-05, "loss": 0.0735, "step": 8410 }, { "grad_norm": 0.6127381324768066, "learning_rate": 9.916813541549283e-05, "loss": 0.0686, "step": 8420 }, { "grad_norm": 0.5905211567878723, "learning_rate": 9.916437674456663e-05, "loss": 0.0768, "step": 8430 }, { "grad_norm": 0.4481150507926941, "learning_rate": 9.916060967279308e-05, "loss": 0.0707, "step": 8440 }, { "grad_norm": 0.5405648350715637, "learning_rate": 9.91568342008159e-05, "loss": 0.0773, "step": 8450 }, { "grad_norm": 0.42580410838127136, "learning_rate": 9.915305032928019e-05, "loss": 0.0712, "step": 8460 }, { "grad_norm": 0.5743812322616577, "learning_rate": 9.914925805883253e-05, "loss": 0.0742, "step": 8470 }, { "grad_norm": 0.590469479560852, "learning_rate": 9.914545739012088e-05, "loss": 0.0805, "step": 8480 }, { "grad_norm": 0.9256635308265686, "learning_rate": 9.91416483237947e-05, "loss": 0.0847, "step": 8490 }, { "grad_norm": 0.6017370820045471, "learning_rate": 9.913783086050485e-05, "loss": 0.0947, "step": 8500 }, { "grad_norm": 0.5271223783493042, "learning_rate": 9.913400500090364e-05, "loss": 0.0801, "step": 8510 }, { "grad_norm": 0.47385579347610474, "learning_rate": 9.913017074564479e-05, "loss": 0.0695, "step": 8520 }, { "grad_norm": 0.4879683554172516, "learning_rate": 9.912632809538348e-05, "loss": 0.0932, "step": 8530 }, { "grad_norm": 0.7781684994697571, "learning_rate": 9.912247705077629e-05, "loss": 0.0833, "step": 8540 }, { "grad_norm": 0.6138082146644592, "learning_rate": 9.911861761248127e-05, "loss": 0.0932, "step": 8550 }, { "grad_norm": 0.979443371295929, "learning_rate": 9.91147497811579e-05, "loss": 0.0823, "step": 8560 }, { "grad_norm": 0.6702413558959961, "learning_rate": 9.911087355746709e-05, "loss": 0.1151, "step": 8570 }, { "grad_norm": 0.5899507999420166, "learning_rate": 9.910698894207117e-05, "loss": 0.0938, "step": 8580 }, { "grad_norm": 0.43661507964134216, "learning_rate": 9.910309593563392e-05, "loss": 0.0914, "step": 8590 }, { "grad_norm": 0.5376884341239929, "learning_rate": 9.909919453882057e-05, "loss": 0.0778, "step": 8600 }, { "grad_norm": 0.7419084310531616, "learning_rate": 9.90952847522977e-05, "loss": 0.0896, "step": 8610 }, { "grad_norm": 0.5395239591598511, "learning_rate": 9.909136657673346e-05, "loss": 0.0952, "step": 8620 }, { "grad_norm": 0.7367041110992432, "learning_rate": 9.908744001279731e-05, "loss": 0.0831, "step": 8630 }, { "grad_norm": 0.7893736362457275, "learning_rate": 9.90835050611602e-05, "loss": 0.0959, "step": 8640 }, { "grad_norm": 0.7373064756393433, "learning_rate": 9.90795617224945e-05, "loss": 0.0931, "step": 8650 }, { "grad_norm": 0.8787965774536133, "learning_rate": 9.907560999747405e-05, "loss": 0.0929, "step": 8660 }, { "grad_norm": 0.4952884316444397, "learning_rate": 9.907164988677408e-05, "loss": 0.0768, "step": 8670 }, { "grad_norm": 0.48620396852493286, "learning_rate": 9.906768139107124e-05, "loss": 0.0638, "step": 8680 }, { "grad_norm": 0.5874187350273132, "learning_rate": 9.906370451104367e-05, "loss": 0.0715, "step": 8690 }, { "grad_norm": 0.4344575107097626, "learning_rate": 9.905971924737088e-05, "loss": 0.0664, "step": 8700 }, { "grad_norm": 0.502485990524292, "learning_rate": 9.905572560073387e-05, "loss": 0.0793, "step": 8710 }, { "grad_norm": 0.6959590911865234, "learning_rate": 9.905172357181501e-05, "loss": 0.0944, "step": 8720 }, { "grad_norm": 0.5871922373771667, "learning_rate": 9.904771316129817e-05, "loss": 0.0784, "step": 8730 }, { "grad_norm": 0.607302188873291, "learning_rate": 9.904369436986862e-05, "loss": 0.082, "step": 8740 }, { "grad_norm": 0.4934439957141876, "learning_rate": 9.903966719821303e-05, "loss": 0.0861, "step": 8750 }, { "grad_norm": 0.4737924635410309, "learning_rate": 9.903563164701956e-05, "loss": 0.0825, "step": 8760 }, { "grad_norm": 0.6529693603515625, "learning_rate": 9.903158771697778e-05, "loss": 0.0914, "step": 8770 }, { "grad_norm": 0.6064938902854919, "learning_rate": 9.902753540877867e-05, "loss": 0.0893, "step": 8780 }, { "grad_norm": 0.5504766702651978, "learning_rate": 9.902347472311466e-05, "loss": 0.0814, "step": 8790 }, { "grad_norm": 0.615012526512146, "learning_rate": 9.901940566067962e-05, "loss": 0.0728, "step": 8800 }, { "grad_norm": 0.5324083566665649, "learning_rate": 9.901532822216883e-05, "loss": 0.0708, "step": 8810 }, { "grad_norm": 0.4996388852596283, "learning_rate": 9.901124240827904e-05, "loss": 0.0776, "step": 8820 }, { "grad_norm": 0.49456894397735596, "learning_rate": 9.900714821970835e-05, "loss": 0.0693, "step": 8830 }, { "grad_norm": 0.6475685238838196, "learning_rate": 9.900304565715641e-05, "loss": 0.0751, "step": 8840 }, { "grad_norm": 0.5752007365226746, "learning_rate": 9.899893472132419e-05, "loss": 0.0632, "step": 8850 }, { "grad_norm": 0.5340248346328735, "learning_rate": 9.899481541291415e-05, "loss": 0.0636, "step": 8860 }, { "grad_norm": 0.7737900018692017, "learning_rate": 9.899068773263016e-05, "loss": 0.077, "step": 8870 }, { "grad_norm": 0.37414830923080444, "learning_rate": 9.898655168117754e-05, "loss": 0.0783, "step": 8880 }, { "grad_norm": 0.5136921405792236, "learning_rate": 9.898240725926302e-05, "loss": 0.0817, "step": 8890 }, { "grad_norm": 0.567905604839325, "learning_rate": 9.897825446759478e-05, "loss": 0.0853, "step": 8900 }, { "grad_norm": 0.5663051009178162, "learning_rate": 9.897409330688241e-05, "loss": 0.0827, "step": 8910 }, { "grad_norm": 0.7741860151290894, "learning_rate": 9.896992377783692e-05, "loss": 0.1094, "step": 8920 }, { "grad_norm": 0.4944756329059601, "learning_rate": 9.89657458811708e-05, "loss": 0.0803, "step": 8930 }, { "grad_norm": 0.5111486315727234, "learning_rate": 9.896155961759792e-05, "loss": 0.0794, "step": 8940 }, { "grad_norm": 0.43211665749549866, "learning_rate": 9.895736498783361e-05, "loss": 0.0798, "step": 8950 }, { "grad_norm": 0.5259597301483154, "learning_rate": 9.895316199259462e-05, "loss": 0.079, "step": 8960 }, { "grad_norm": 0.49870339035987854, "learning_rate": 9.894895063259909e-05, "loss": 0.0742, "step": 8970 }, { "grad_norm": 0.5592668652534485, "learning_rate": 9.894473090856667e-05, "loss": 0.0817, "step": 8980 }, { "grad_norm": 0.5838103890419006, "learning_rate": 9.894050282121839e-05, "loss": 0.1061, "step": 8990 }, { "grad_norm": 0.5287079215049744, "learning_rate": 9.893626637127668e-05, "loss": 0.0758, "step": 9000 }, { "grad_norm": 0.4071723222732544, "learning_rate": 9.893202155946546e-05, "loss": 0.066, "step": 9010 }, { "grad_norm": 0.643903911113739, "learning_rate": 9.892776838651006e-05, "loss": 0.0781, "step": 9020 }, { "grad_norm": 0.7000106573104858, "learning_rate": 9.892350685313722e-05, "loss": 0.0899, "step": 9030 }, { "grad_norm": 0.6075334548950195, "learning_rate": 9.891923696007513e-05, "loss": 0.0837, "step": 9040 }, { "grad_norm": 0.6167601346969604, "learning_rate": 9.891495870805336e-05, "loss": 0.0826, "step": 9050 }, { "grad_norm": 0.6569895148277283, "learning_rate": 9.891067209780298e-05, "loss": 0.0881, "step": 9060 }, { "grad_norm": 0.535500705242157, "learning_rate": 9.890637713005646e-05, "loss": 0.0611, "step": 9070 }, { "grad_norm": 0.688590407371521, "learning_rate": 9.890207380554767e-05, "loss": 0.0679, "step": 9080 }, { "grad_norm": 0.43138429522514343, "learning_rate": 9.889776212501196e-05, "loss": 0.0829, "step": 9090 }, { "grad_norm": 0.43781688809394836, "learning_rate": 9.889344208918605e-05, "loss": 0.0762, "step": 9100 }, { "grad_norm": 0.5062068700790405, "learning_rate": 9.888911369880812e-05, "loss": 0.0771, "step": 9110 }, { "grad_norm": 0.5138459801673889, "learning_rate": 9.888477695461777e-05, "loss": 0.0818, "step": 9120 }, { "grad_norm": 0.5627698302268982, "learning_rate": 9.888043185735607e-05, "loss": 0.0845, "step": 9130 }, { "grad_norm": 0.4411916732788086, "learning_rate": 9.887607840776542e-05, "loss": 0.0648, "step": 9140 }, { "grad_norm": 0.4085550904273987, "learning_rate": 9.887171660658975e-05, "loss": 0.0879, "step": 9150 }, { "grad_norm": 0.5767067670822144, "learning_rate": 9.886734645457435e-05, "loss": 0.0699, "step": 9160 }, { "grad_norm": 0.7612881064414978, "learning_rate": 9.886296795246597e-05, "loss": 0.0804, "step": 9170 }, { "grad_norm": 0.41306087374687195, "learning_rate": 9.885858110101276e-05, "loss": 0.0723, "step": 9180 }, { "grad_norm": 0.3480064868927002, "learning_rate": 9.885418590096434e-05, "loss": 0.0581, "step": 9190 }, { "grad_norm": 0.5207046866416931, "learning_rate": 9.88497823530717e-05, "loss": 0.0764, "step": 9200 }, { "grad_norm": 0.6360418200492859, "learning_rate": 9.884537045808732e-05, "loss": 0.097, "step": 9210 }, { "grad_norm": 0.4946810305118561, "learning_rate": 9.884095021676502e-05, "loss": 0.0906, "step": 9220 }, { "grad_norm": 0.4844728410243988, "learning_rate": 9.883652162986017e-05, "loss": 0.0765, "step": 9230 }, { "grad_norm": 0.4989716112613678, "learning_rate": 9.883208469812943e-05, "loss": 0.0797, "step": 9240 }, { "grad_norm": 0.5509399175643921, "learning_rate": 9.882763942233098e-05, "loss": 0.0656, "step": 9250 }, { "grad_norm": 0.4249165952205658, "learning_rate": 9.882318580322441e-05, "loss": 0.0647, "step": 9260 }, { "grad_norm": 0.4830184876918793, "learning_rate": 9.881872384157067e-05, "loss": 0.0684, "step": 9270 }, { "grad_norm": 0.500627338886261, "learning_rate": 9.881425353813225e-05, "loss": 0.0722, "step": 9280 }, { "grad_norm": 0.711771547794342, "learning_rate": 9.880977489367296e-05, "loss": 0.0886, "step": 9290 }, { "grad_norm": 0.8191519379615784, "learning_rate": 9.88052879089581e-05, "loss": 0.0925, "step": 9300 }, { "grad_norm": 0.5459639430046082, "learning_rate": 9.880079258475434e-05, "loss": 0.075, "step": 9310 }, { "grad_norm": 0.4796655476093292, "learning_rate": 9.879628892182985e-05, "loss": 0.0993, "step": 9320 }, { "grad_norm": 0.5281363129615784, "learning_rate": 9.879177692095416e-05, "loss": 0.0883, "step": 9330 }, { "grad_norm": 0.5225226283073425, "learning_rate": 9.878725658289825e-05, "loss": 0.0889, "step": 9340 }, { "grad_norm": 0.5861853361129761, "learning_rate": 9.878272790843454e-05, "loss": 0.0698, "step": 9350 }, { "grad_norm": 0.6576318144798279, "learning_rate": 9.877819089833682e-05, "loss": 0.0696, "step": 9360 }, { "grad_norm": 0.3819389045238495, "learning_rate": 9.877364555338038e-05, "loss": 0.0794, "step": 9370 }, { "grad_norm": 0.5659712553024292, "learning_rate": 9.876909187434186e-05, "loss": 0.0856, "step": 9380 }, { "grad_norm": 0.6339531540870667, "learning_rate": 9.876452986199939e-05, "loss": 0.0706, "step": 9390 }, { "grad_norm": 0.6172010898590088, "learning_rate": 9.875995951713248e-05, "loss": 0.0916, "step": 9400 }, { "grad_norm": 0.5129136443138123, "learning_rate": 9.875538084052207e-05, "loss": 0.085, "step": 9410 }, { "grad_norm": 0.6326509118080139, "learning_rate": 9.875079383295053e-05, "loss": 0.076, "step": 9420 }, { "grad_norm": 0.4755033552646637, "learning_rate": 9.874619849520167e-05, "loss": 0.0665, "step": 9430 }, { "grad_norm": 0.5580241084098816, "learning_rate": 9.874159482806069e-05, "loss": 0.0686, "step": 9440 }, { "grad_norm": 0.38971179723739624, "learning_rate": 9.873698283231426e-05, "loss": 0.0683, "step": 9450 }, { "grad_norm": 0.547447919845581, "learning_rate": 9.87323625087504e-05, "loss": 0.0639, "step": 9460 }, { "grad_norm": 0.5784258842468262, "learning_rate": 9.872773385815863e-05, "loss": 0.0897, "step": 9470 }, { "grad_norm": 0.3843481242656708, "learning_rate": 9.872309688132986e-05, "loss": 0.0659, "step": 9480 }, { "grad_norm": 0.5482068657875061, "learning_rate": 9.871845157905639e-05, "loss": 0.0727, "step": 9490 }, { "grad_norm": 0.6478517651557922, "learning_rate": 9.871379795213201e-05, "loss": 0.0627, "step": 9500 }, { "grad_norm": 0.4976891577243805, "learning_rate": 9.87091360013519e-05, "loss": 0.0586, "step": 9510 }, { "grad_norm": 0.650375485420227, "learning_rate": 9.870446572751262e-05, "loss": 0.0763, "step": 9520 }, { "grad_norm": 0.46597355604171753, "learning_rate": 9.869978713141224e-05, "loss": 0.068, "step": 9530 }, { "grad_norm": 0.8787654042243958, "learning_rate": 9.869510021385016e-05, "loss": 0.0758, "step": 9540 }, { "grad_norm": 0.7162914276123047, "learning_rate": 9.869040497562727e-05, "loss": 0.0762, "step": 9550 }, { "grad_norm": 0.5390929579734802, "learning_rate": 9.868570141754587e-05, "loss": 0.0946, "step": 9560 }, { "grad_norm": 0.9277867674827576, "learning_rate": 9.868098954040965e-05, "loss": 0.1202, "step": 9570 }, { "grad_norm": 0.5127830505371094, "learning_rate": 9.867626934502374e-05, "loss": 0.091, "step": 9580 }, { "grad_norm": 0.5056758522987366, "learning_rate": 9.86715408321947e-05, "loss": 0.0758, "step": 9590 }, { "grad_norm": 0.5533879995346069, "learning_rate": 9.86668040027305e-05, "loss": 0.074, "step": 9600 }, { "grad_norm": 0.5922725796699524, "learning_rate": 9.866205885744053e-05, "loss": 0.0785, "step": 9610 }, { "grad_norm": 0.7745269536972046, "learning_rate": 9.865730539713563e-05, "loss": 0.0832, "step": 9620 }, { "grad_norm": 0.4456944763660431, "learning_rate": 9.8652543622628e-05, "loss": 0.071, "step": 9630 }, { "grad_norm": 0.5401076078414917, "learning_rate": 9.864777353473132e-05, "loss": 0.0651, "step": 9640 }, { "grad_norm": 0.555546224117279, "learning_rate": 9.864299513426068e-05, "loss": 0.0744, "step": 9650 }, { "grad_norm": 0.5066705346107483, "learning_rate": 9.863820842203254e-05, "loss": 0.0724, "step": 9660 }, { "grad_norm": 0.46909618377685547, "learning_rate": 9.863341339886483e-05, "loss": 0.0677, "step": 9670 }, { "grad_norm": 0.47798001766204834, "learning_rate": 9.86286100655769e-05, "loss": 0.0659, "step": 9680 }, { "grad_norm": 0.6591066718101501, "learning_rate": 9.862379842298953e-05, "loss": 0.0647, "step": 9690 }, { "grad_norm": 0.5044112205505371, "learning_rate": 9.861897847192485e-05, "loss": 0.0749, "step": 9700 }, { "grad_norm": 0.6145663857460022, "learning_rate": 9.86141502132065e-05, "loss": 0.0687, "step": 9710 }, { "grad_norm": 0.6314069032669067, "learning_rate": 9.860931364765946e-05, "loss": 0.0771, "step": 9720 }, { "grad_norm": 0.524142324924469, "learning_rate": 9.860446877611021e-05, "loss": 0.0763, "step": 9730 }, { "grad_norm": 0.4904966652393341, "learning_rate": 9.859961559938655e-05, "loss": 0.0707, "step": 9740 }, { "grad_norm": 0.5073102116584778, "learning_rate": 9.85947541183178e-05, "loss": 0.0671, "step": 9750 }, { "grad_norm": 0.5241628289222717, "learning_rate": 9.858988433373463e-05, "loss": 0.0722, "step": 9760 }, { "grad_norm": 0.7146803736686707, "learning_rate": 9.858500624646918e-05, "loss": 0.0677, "step": 9770 }, { "grad_norm": 0.7479626536369324, "learning_rate": 9.858011985735497e-05, "loss": 0.078, "step": 9780 }, { "grad_norm": 0.5281181335449219, "learning_rate": 9.857522516722693e-05, "loss": 0.0762, "step": 9790 }, { "grad_norm": 0.4875970482826233, "learning_rate": 9.857032217692145e-05, "loss": 0.062, "step": 9800 }, { "grad_norm": 0.7117987871170044, "learning_rate": 9.856541088727631e-05, "loss": 0.0778, "step": 9810 }, { "grad_norm": 0.4940958321094513, "learning_rate": 9.856049129913072e-05, "loss": 0.0825, "step": 9820 }, { "grad_norm": 0.49938327074050903, "learning_rate": 9.85555634133253e-05, "loss": 0.0852, "step": 9830 }, { "grad_norm": 0.5042890906333923, "learning_rate": 9.855062723070208e-05, "loss": 0.087, "step": 9840 }, { "grad_norm": 0.5094678401947021, "learning_rate": 9.854568275210454e-05, "loss": 0.0903, "step": 9850 }, { "grad_norm": 0.4095175266265869, "learning_rate": 9.854072997837754e-05, "loss": 0.0671, "step": 9860 }, { "grad_norm": 0.5062433481216431, "learning_rate": 9.853576891036737e-05, "loss": 0.0703, "step": 9870 }, { "grad_norm": 0.5823844075202942, "learning_rate": 9.853079954892177e-05, "loss": 0.0756, "step": 9880 }, { "grad_norm": 0.5807788968086243, "learning_rate": 9.852582189488983e-05, "loss": 0.0693, "step": 9890 }, { "grad_norm": 0.6407591104507446, "learning_rate": 9.852083594912212e-05, "loss": 0.0851, "step": 9900 }, { "grad_norm": 0.4696917235851288, "learning_rate": 9.851584171247058e-05, "loss": 0.0815, "step": 9910 }, { "grad_norm": 0.5102702379226685, "learning_rate": 9.851083918578863e-05, "loss": 0.079, "step": 9920 }, { "grad_norm": 0.468550443649292, "learning_rate": 9.850582836993103e-05, "loss": 0.0645, "step": 9930 }, { "grad_norm": 0.3980836868286133, "learning_rate": 9.850080926575397e-05, "loss": 0.0751, "step": 9940 }, { "grad_norm": 0.5484049320220947, "learning_rate": 9.849578187411515e-05, "loss": 0.0721, "step": 9950 }, { "grad_norm": 0.6695907115936279, "learning_rate": 9.849074619587354e-05, "loss": 0.0673, "step": 9960 }, { "grad_norm": 0.38077282905578613, "learning_rate": 9.848570223188964e-05, "loss": 0.0797, "step": 9970 }, { "grad_norm": 0.5326346158981323, "learning_rate": 9.848064998302531e-05, "loss": 0.0672, "step": 9980 }, { "grad_norm": 0.4479021728038788, "learning_rate": 9.847558945014386e-05, "loss": 0.0833, "step": 9990 }, { "grad_norm": 0.6351951956748962, "learning_rate": 9.847052063410996e-05, "loss": 0.0861, "step": 10000 }, { "grad_norm": 0.5696467757225037, "learning_rate": 9.846544353578977e-05, "loss": 0.0806, "step": 10010 }, { "grad_norm": 0.4318893551826477, "learning_rate": 9.846035815605081e-05, "loss": 0.0739, "step": 10020 }, { "grad_norm": 0.5342860817909241, "learning_rate": 9.845526449576204e-05, "loss": 0.0635, "step": 10030 }, { "grad_norm": 0.673427939414978, "learning_rate": 9.845016255579383e-05, "loss": 0.0778, "step": 10040 }, { "grad_norm": 0.7255218029022217, "learning_rate": 9.844505233701794e-05, "loss": 0.0782, "step": 10050 }, { "grad_norm": 0.6014148592948914, "learning_rate": 9.843993384030757e-05, "loss": 0.0669, "step": 10060 }, { "grad_norm": 0.5475618243217468, "learning_rate": 9.843480706653737e-05, "loss": 0.0664, "step": 10070 }, { "grad_norm": 0.5645968317985535, "learning_rate": 9.84296720165833e-05, "loss": 0.072, "step": 10080 }, { "grad_norm": 0.4759524166584015, "learning_rate": 9.842452869132286e-05, "loss": 0.0701, "step": 10090 }, { "grad_norm": 0.6006396412849426, "learning_rate": 9.841937709163489e-05, "loss": 0.0706, "step": 10100 }, { "grad_norm": 0.4686315655708313, "learning_rate": 9.841421721839962e-05, "loss": 0.0583, "step": 10110 }, { "grad_norm": 0.41415855288505554, "learning_rate": 9.840904907249879e-05, "loss": 0.0618, "step": 10120 }, { "grad_norm": 0.8192910552024841, "learning_rate": 9.840387265481545e-05, "loss": 0.0772, "step": 10130 }, { "grad_norm": 0.4077901840209961, "learning_rate": 9.839868796623411e-05, "loss": 0.067, "step": 10140 }, { "grad_norm": 0.4440906345844269, "learning_rate": 9.839349500764072e-05, "loss": 0.0853, "step": 10150 }, { "grad_norm": 0.911167562007904, "learning_rate": 9.83882937799226e-05, "loss": 0.0827, "step": 10160 }, { "grad_norm": 0.5382157564163208, "learning_rate": 9.838308428396849e-05, "loss": 0.0733, "step": 10170 }, { "grad_norm": 0.6804007887840271, "learning_rate": 9.837786652066854e-05, "loss": 0.1, "step": 10180 }, { "grad_norm": 0.6075385212898254, "learning_rate": 9.837264049091437e-05, "loss": 0.0993, "step": 10190 }, { "grad_norm": 0.5617505311965942, "learning_rate": 9.836740619559893e-05, "loss": 0.0859, "step": 10200 }, { "grad_norm": 0.4483393430709839, "learning_rate": 9.836216363561659e-05, "loss": 0.0774, "step": 10210 }, { "grad_norm": 0.48230212926864624, "learning_rate": 9.835691281186322e-05, "loss": 0.0686, "step": 10220 }, { "grad_norm": 0.4897559583187103, "learning_rate": 9.8351653725236e-05, "loss": 0.0738, "step": 10230 }, { "grad_norm": 0.5695568323135376, "learning_rate": 9.83463863766336e-05, "loss": 0.0729, "step": 10240 }, { "grad_norm": 0.5760354995727539, "learning_rate": 9.834111076695602e-05, "loss": 0.0776, "step": 10250 }, { "grad_norm": 0.35239318013191223, "learning_rate": 9.833582689710477e-05, "loss": 0.0648, "step": 10260 }, { "grad_norm": 0.4330148696899414, "learning_rate": 9.833053476798268e-05, "loss": 0.0723, "step": 10270 }, { "grad_norm": 0.6426020860671997, "learning_rate": 9.832523438049404e-05, "loss": 0.0818, "step": 10280 }, { "grad_norm": 0.5862906575202942, "learning_rate": 9.831992573554454e-05, "loss": 0.0755, "step": 10290 }, { "grad_norm": 0.6063821911811829, "learning_rate": 9.831460883404128e-05, "loss": 0.0671, "step": 10300 }, { "grad_norm": 0.7037986516952515, "learning_rate": 9.830928367689278e-05, "loss": 0.0598, "step": 10310 }, { "grad_norm": 0.5465759038925171, "learning_rate": 9.830395026500896e-05, "loss": 0.0694, "step": 10320 }, { "grad_norm": 0.43156737089157104, "learning_rate": 9.829860859930115e-05, "loss": 0.0796, "step": 10330 }, { "grad_norm": 0.47795847058296204, "learning_rate": 9.829325868068212e-05, "loss": 0.0791, "step": 10340 }, { "grad_norm": 1.0658137798309326, "learning_rate": 9.8287900510066e-05, "loss": 0.0919, "step": 10350 }, { "grad_norm": 0.6167594194412231, "learning_rate": 9.828253408836834e-05, "loss": 0.0737, "step": 10360 }, { "grad_norm": 0.4834536612033844, "learning_rate": 9.827715941650615e-05, "loss": 0.0816, "step": 10370 }, { "grad_norm": 0.5429937243461609, "learning_rate": 9.82717764953978e-05, "loss": 0.0794, "step": 10380 }, { "grad_norm": 0.5254470109939575, "learning_rate": 9.826638532596308e-05, "loss": 0.0729, "step": 10390 }, { "grad_norm": 0.41563504934310913, "learning_rate": 9.82609859091232e-05, "loss": 0.0714, "step": 10400 }, { "grad_norm": 0.5170457363128662, "learning_rate": 9.825557824580076e-05, "loss": 0.0646, "step": 10410 }, { "grad_norm": 0.4019351899623871, "learning_rate": 9.82501623369198e-05, "loss": 0.0801, "step": 10420 }, { "grad_norm": 0.8127867579460144, "learning_rate": 9.824473818340574e-05, "loss": 0.0645, "step": 10430 }, { "grad_norm": 0.7363020181655884, "learning_rate": 9.823930578618541e-05, "loss": 0.073, "step": 10440 }, { "grad_norm": 0.5438559651374817, "learning_rate": 9.823386514618709e-05, "loss": 0.0899, "step": 10450 }, { "grad_norm": 0.4696241021156311, "learning_rate": 9.82284162643404e-05, "loss": 0.0705, "step": 10460 }, { "grad_norm": 0.6001116037368774, "learning_rate": 9.822295914157642e-05, "loss": 0.0736, "step": 10470 }, { "grad_norm": 0.6082790493965149, "learning_rate": 9.821749377882763e-05, "loss": 0.0662, "step": 10480 }, { "grad_norm": 0.45247912406921387, "learning_rate": 9.821202017702791e-05, "loss": 0.0746, "step": 10490 }, { "grad_norm": 0.5969871282577515, "learning_rate": 9.820653833711253e-05, "loss": 0.0685, "step": 10500 }, { "grad_norm": 0.5251826643943787, "learning_rate": 9.820104826001822e-05, "loss": 0.081, "step": 10510 }, { "grad_norm": 0.4517503082752228, "learning_rate": 9.819554994668305e-05, "loss": 0.0634, "step": 10520 }, { "grad_norm": 0.46986058354377747, "learning_rate": 9.819004339804654e-05, "loss": 0.0842, "step": 10530 }, { "grad_norm": 0.44665783643722534, "learning_rate": 9.818452861504961e-05, "loss": 0.067, "step": 10540 }, { "grad_norm": 0.4912552833557129, "learning_rate": 9.81790055986346e-05, "loss": 0.065, "step": 10550 }, { "grad_norm": 0.41713839769363403, "learning_rate": 9.817347434974523e-05, "loss": 0.0603, "step": 10560 }, { "grad_norm": 0.43722209334373474, "learning_rate": 9.816793486932664e-05, "loss": 0.0635, "step": 10570 }, { "grad_norm": 0.49308961629867554, "learning_rate": 9.816238715832538e-05, "loss": 0.0682, "step": 10580 }, { "grad_norm": 0.5093157887458801, "learning_rate": 9.815683121768939e-05, "loss": 0.0798, "step": 10590 }, { "grad_norm": 0.5000339150428772, "learning_rate": 9.815126704836804e-05, "loss": 0.0766, "step": 10600 }, { "grad_norm": 0.5386168956756592, "learning_rate": 9.81456946513121e-05, "loss": 0.0911, "step": 10610 }, { "grad_norm": 0.5155866742134094, "learning_rate": 9.814011402747373e-05, "loss": 0.0854, "step": 10620 }, { "grad_norm": 0.4434163570404053, "learning_rate": 9.813452517780651e-05, "loss": 0.0649, "step": 10630 }, { "grad_norm": 0.4044584333896637, "learning_rate": 9.81289281032654e-05, "loss": 0.0743, "step": 10640 }, { "grad_norm": 0.4347461760044098, "learning_rate": 9.812332280480683e-05, "loss": 0.0722, "step": 10650 }, { "grad_norm": 0.5800251960754395, "learning_rate": 9.811770928338854e-05, "loss": 0.0741, "step": 10660 }, { "grad_norm": 0.5768042206764221, "learning_rate": 9.811208753996979e-05, "loss": 0.0763, "step": 10670 }, { "grad_norm": 0.486581951379776, "learning_rate": 9.810645757551113e-05, "loss": 0.0641, "step": 10680 }, { "grad_norm": 0.428398996591568, "learning_rate": 9.810081939097459e-05, "loss": 0.0568, "step": 10690 }, { "grad_norm": 0.46722647547721863, "learning_rate": 9.809517298732356e-05, "loss": 0.095, "step": 10700 }, { "grad_norm": 0.5371106266975403, "learning_rate": 9.80895183655229e-05, "loss": 0.0698, "step": 10710 }, { "grad_norm": 0.5203412771224976, "learning_rate": 9.808385552653877e-05, "loss": 0.0692, "step": 10720 }, { "grad_norm": 0.5944212675094604, "learning_rate": 9.807818447133886e-05, "loss": 0.0666, "step": 10730 }, { "grad_norm": 0.5216562747955322, "learning_rate": 9.807250520089215e-05, "loss": 0.0634, "step": 10740 }, { "grad_norm": 0.44715264439582825, "learning_rate": 9.806681771616908e-05, "loss": 0.0673, "step": 10750 }, { "grad_norm": 0.5659223198890686, "learning_rate": 9.80611220181415e-05, "loss": 0.0822, "step": 10760 }, { "grad_norm": 0.6624575257301331, "learning_rate": 9.805541810778264e-05, "loss": 0.0831, "step": 10770 }, { "grad_norm": 0.5560933947563171, "learning_rate": 9.804970598606716e-05, "loss": 0.0794, "step": 10780 }, { "grad_norm": 0.494097501039505, "learning_rate": 9.804398565397106e-05, "loss": 0.0715, "step": 10790 }, { "grad_norm": 0.5722383856773376, "learning_rate": 9.803825711247183e-05, "loss": 0.0817, "step": 10800 }, { "grad_norm": 0.4962082803249359, "learning_rate": 9.803252036254831e-05, "loss": 0.0671, "step": 10810 }, { "grad_norm": 0.6213139295578003, "learning_rate": 9.802677540518076e-05, "loss": 0.0709, "step": 10820 }, { "grad_norm": 0.4941056966781616, "learning_rate": 9.802102224135081e-05, "loss": 0.0845, "step": 10830 }, { "grad_norm": 0.40092533826828003, "learning_rate": 9.801526087204155e-05, "loss": 0.0921, "step": 10840 }, { "grad_norm": 0.6704264283180237, "learning_rate": 9.800949129823743e-05, "loss": 0.078, "step": 10850 }, { "grad_norm": 0.5484113097190857, "learning_rate": 9.80037135209243e-05, "loss": 0.0664, "step": 10860 }, { "grad_norm": 0.4514169692993164, "learning_rate": 9.799792754108946e-05, "loss": 0.0672, "step": 10870 }, { "grad_norm": 0.5143822431564331, "learning_rate": 9.799213335972152e-05, "loss": 0.0822, "step": 10880 }, { "grad_norm": 0.4949771463871002, "learning_rate": 9.798633097781058e-05, "loss": 0.0791, "step": 10890 }, { "grad_norm": 0.5748525857925415, "learning_rate": 9.79805203963481e-05, "loss": 0.0736, "step": 10900 }, { "grad_norm": 0.5098513960838318, "learning_rate": 9.797470161632697e-05, "loss": 0.0701, "step": 10910 }, { "grad_norm": 0.42941924929618835, "learning_rate": 9.796887463874145e-05, "loss": 0.0697, "step": 10920 }, { "grad_norm": 0.5374167561531067, "learning_rate": 9.796303946458718e-05, "loss": 0.0834, "step": 10930 }, { "grad_norm": 0.6455413103103638, "learning_rate": 9.795719609486127e-05, "loss": 0.0736, "step": 10940 }, { "grad_norm": 0.8693521022796631, "learning_rate": 9.795134453056219e-05, "loss": 0.0654, "step": 10950 }, { "grad_norm": 0.511764407157898, "learning_rate": 9.794548477268979e-05, "loss": 0.0606, "step": 10960 }, { "grad_norm": 0.43636012077331543, "learning_rate": 9.793961682224537e-05, "loss": 0.0658, "step": 10970 }, { "grad_norm": 0.40464913845062256, "learning_rate": 9.793374068023156e-05, "loss": 0.0728, "step": 10980 }, { "grad_norm": 0.5621328949928284, "learning_rate": 9.792785634765247e-05, "loss": 0.0694, "step": 10990 }, { "grad_norm": 0.6612173318862915, "learning_rate": 9.792196382551357e-05, "loss": 0.0738, "step": 11000 }, { "grad_norm": 0.32839319109916687, "learning_rate": 9.791606311482171e-05, "loss": 0.0682, "step": 11010 }, { "grad_norm": 0.47045114636421204, "learning_rate": 9.791015421658518e-05, "loss": 0.0741, "step": 11020 }, { "grad_norm": 0.6596391201019287, "learning_rate": 9.790423713181362e-05, "loss": 0.0732, "step": 11030 }, { "grad_norm": 0.4934157133102417, "learning_rate": 9.789831186151814e-05, "loss": 0.0714, "step": 11040 }, { "grad_norm": 0.8182904720306396, "learning_rate": 9.789237840671118e-05, "loss": 0.0682, "step": 11050 }, { "grad_norm": 0.5634306073188782, "learning_rate": 9.78864367684066e-05, "loss": 0.0853, "step": 11060 }, { "grad_norm": 0.5029646754264832, "learning_rate": 9.788048694761968e-05, "loss": 0.0707, "step": 11070 }, { "grad_norm": 0.47228795289993286, "learning_rate": 9.787452894536709e-05, "loss": 0.0693, "step": 11080 }, { "grad_norm": 0.4358139634132385, "learning_rate": 9.786856276266685e-05, "loss": 0.0711, "step": 11090 }, { "grad_norm": 0.5826097726821899, "learning_rate": 9.786258840053845e-05, "loss": 0.086, "step": 11100 }, { "grad_norm": 0.4975408911705017, "learning_rate": 9.785660586000273e-05, "loss": 0.0799, "step": 11110 }, { "grad_norm": 0.4100853204727173, "learning_rate": 9.785061514208196e-05, "loss": 0.0687, "step": 11120 }, { "grad_norm": 0.5571218132972717, "learning_rate": 9.784461624779977e-05, "loss": 0.0774, "step": 11130 }, { "grad_norm": 0.40655702352523804, "learning_rate": 9.783860917818123e-05, "loss": 0.0673, "step": 11140 }, { "grad_norm": 0.4935269057750702, "learning_rate": 9.783259393425277e-05, "loss": 0.0566, "step": 11150 }, { "grad_norm": 0.6237859725952148, "learning_rate": 9.782657051704221e-05, "loss": 0.074, "step": 11160 }, { "grad_norm": 0.4604838490486145, "learning_rate": 9.782053892757883e-05, "loss": 0.0615, "step": 11170 }, { "grad_norm": 0.5829905271530151, "learning_rate": 9.781449916689324e-05, "loss": 0.0788, "step": 11180 }, { "grad_norm": 0.4764639139175415, "learning_rate": 9.780845123601746e-05, "loss": 0.0835, "step": 11190 }, { "grad_norm": 0.4758394658565521, "learning_rate": 9.780239513598492e-05, "loss": 0.0756, "step": 11200 }, { "grad_norm": 0.7424488067626953, "learning_rate": 9.779633086783047e-05, "loss": 0.0751, "step": 11210 }, { "grad_norm": 0.47871366143226624, "learning_rate": 9.779025843259031e-05, "loss": 0.0652, "step": 11220 }, { "grad_norm": 0.45222610235214233, "learning_rate": 9.778417783130204e-05, "loss": 0.0861, "step": 11230 }, { "grad_norm": 0.5839142203330994, "learning_rate": 9.777808906500468e-05, "loss": 0.0802, "step": 11240 }, { "grad_norm": 0.44187086820602417, "learning_rate": 9.777199213473862e-05, "loss": 0.0606, "step": 11250 }, { "grad_norm": 0.6075965166091919, "learning_rate": 9.77658870415457e-05, "loss": 0.0648, "step": 11260 }, { "grad_norm": 0.5015285015106201, "learning_rate": 9.775977378646906e-05, "loss": 0.0641, "step": 11270 }, { "grad_norm": 0.6677071452140808, "learning_rate": 9.775365237055331e-05, "loss": 0.082, "step": 11280 }, { "grad_norm": 0.49968186020851135, "learning_rate": 9.774752279484445e-05, "loss": 0.0705, "step": 11290 }, { "grad_norm": 0.5253375172615051, "learning_rate": 9.774138506038984e-05, "loss": 0.0613, "step": 11300 }, { "grad_norm": 0.5682740211486816, "learning_rate": 9.773523916823826e-05, "loss": 0.0699, "step": 11310 }, { "grad_norm": 0.45025768876075745, "learning_rate": 9.772908511943986e-05, "loss": 0.0705, "step": 11320 }, { "grad_norm": 0.5605946183204651, "learning_rate": 9.77229229150462e-05, "loss": 0.0653, "step": 11330 }, { "grad_norm": 0.5137900114059448, "learning_rate": 9.771675255611024e-05, "loss": 0.0642, "step": 11340 }, { "grad_norm": 0.37184658646583557, "learning_rate": 9.771057404368632e-05, "loss": 0.0936, "step": 11350 }, { "grad_norm": 0.5087929964065552, "learning_rate": 9.770438737883018e-05, "loss": 0.0719, "step": 11360 }, { "grad_norm": 0.39544370770454407, "learning_rate": 9.769819256259898e-05, "loss": 0.0547, "step": 11370 }, { "grad_norm": 0.4365968108177185, "learning_rate": 9.769198959605119e-05, "loss": 0.075, "step": 11380 }, { "grad_norm": 0.44108158349990845, "learning_rate": 9.768577848024678e-05, "loss": 0.0782, "step": 11390 }, { "grad_norm": 0.566439688205719, "learning_rate": 9.767955921624702e-05, "loss": 0.0877, "step": 11400 }, { "grad_norm": 0.6236082315444946, "learning_rate": 9.767333180511465e-05, "loss": 0.0695, "step": 11410 }, { "grad_norm": 0.5956291556358337, "learning_rate": 9.766709624791373e-05, "loss": 0.0831, "step": 11420 }, { "grad_norm": 1.1096388101577759, "learning_rate": 9.766085254570975e-05, "loss": 0.0827, "step": 11430 }, { "grad_norm": 0.5564572215080261, "learning_rate": 9.76546006995696e-05, "loss": 0.0777, "step": 11440 }, { "grad_norm": 0.49741995334625244, "learning_rate": 9.764834071056155e-05, "loss": 0.0674, "step": 11450 }, { "grad_norm": 0.42839279770851135, "learning_rate": 9.764207257975526e-05, "loss": 0.0748, "step": 11460 }, { "grad_norm": 0.6109262108802795, "learning_rate": 9.763579630822179e-05, "loss": 0.0692, "step": 11470 }, { "grad_norm": 0.5252758860588074, "learning_rate": 9.762951189703356e-05, "loss": 0.0692, "step": 11480 }, { "grad_norm": 0.5073116421699524, "learning_rate": 9.762321934726442e-05, "loss": 0.0732, "step": 11490 }, { "grad_norm": 0.54023277759552, "learning_rate": 9.761691865998959e-05, "loss": 0.0692, "step": 11500 }, { "grad_norm": 0.4904443025588989, "learning_rate": 9.76106098362857e-05, "loss": 0.0583, "step": 11510 }, { "grad_norm": 0.3946409821510315, "learning_rate": 9.760429287723072e-05, "loss": 0.065, "step": 11520 }, { "grad_norm": 0.48899170756340027, "learning_rate": 9.759796778390406e-05, "loss": 0.0618, "step": 11530 }, { "grad_norm": 0.4522281885147095, "learning_rate": 9.759163455738653e-05, "loss": 0.0575, "step": 11540 }, { "grad_norm": 0.6133468747138977, "learning_rate": 9.75852931987603e-05, "loss": 0.0779, "step": 11550 }, { "grad_norm": 0.385006308555603, "learning_rate": 9.757894370910891e-05, "loss": 0.0689, "step": 11560 }, { "grad_norm": 0.5277513265609741, "learning_rate": 9.757258608951733e-05, "loss": 0.0665, "step": 11570 }, { "grad_norm": 0.4698334038257599, "learning_rate": 9.75662203410719e-05, "loss": 0.0783, "step": 11580 }, { "grad_norm": 0.52388995885849, "learning_rate": 9.755984646486034e-05, "loss": 0.0755, "step": 11590 }, { "grad_norm": 0.5272423028945923, "learning_rate": 9.75534644619718e-05, "loss": 0.0746, "step": 11600 }, { "grad_norm": 0.7911658883094788, "learning_rate": 9.754707433349676e-05, "loss": 0.091, "step": 11610 }, { "grad_norm": 0.8299140930175781, "learning_rate": 9.754067608052715e-05, "loss": 0.0927, "step": 11620 }, { "grad_norm": 0.5359364151954651, "learning_rate": 9.753426970415622e-05, "loss": 0.1, "step": 11630 }, { "grad_norm": 0.7308338284492493, "learning_rate": 9.752785520547868e-05, "loss": 0.0755, "step": 11640 }, { "grad_norm": 0.5782484412193298, "learning_rate": 9.752143258559056e-05, "loss": 0.0772, "step": 11650 }, { "grad_norm": 0.6581267714500427, "learning_rate": 9.751500184558933e-05, "loss": 0.0847, "step": 11660 }, { "grad_norm": 0.5454909801483154, "learning_rate": 9.750856298657383e-05, "loss": 0.0773, "step": 11670 }, { "grad_norm": 0.4134276509284973, "learning_rate": 9.750211600964428e-05, "loss": 0.0757, "step": 11680 }, { "grad_norm": 0.6887159943580627, "learning_rate": 9.749566091590226e-05, "loss": 0.0756, "step": 11690 }, { "grad_norm": 0.38918474316596985, "learning_rate": 9.748919770645083e-05, "loss": 0.0713, "step": 11700 }, { "grad_norm": 0.4668690860271454, "learning_rate": 9.748272638239432e-05, "loss": 0.0796, "step": 11710 }, { "grad_norm": 0.7167181372642517, "learning_rate": 9.747624694483855e-05, "loss": 0.0835, "step": 11720 }, { "grad_norm": 0.4739501178264618, "learning_rate": 9.746975939489065e-05, "loss": 0.0701, "step": 11730 }, { "grad_norm": 1.364719033241272, "learning_rate": 9.746326373365918e-05, "loss": 0.0684, "step": 11740 }, { "grad_norm": 0.5381312966346741, "learning_rate": 9.745675996225403e-05, "loss": 0.0708, "step": 11750 }, { "grad_norm": 0.4678381681442261, "learning_rate": 9.745024808178657e-05, "loss": 0.0714, "step": 11760 }, { "grad_norm": 0.5445761680603027, "learning_rate": 9.744372809336947e-05, "loss": 0.0703, "step": 11770 }, { "grad_norm": 0.6615184545516968, "learning_rate": 9.743719999811682e-05, "loss": 0.0785, "step": 11780 }, { "grad_norm": 0.6702180504798889, "learning_rate": 9.743066379714412e-05, "loss": 0.0678, "step": 11790 }, { "grad_norm": 0.5265853404998779, "learning_rate": 9.74241194915682e-05, "loss": 0.0791, "step": 11800 }, { "grad_norm": 0.49477532505989075, "learning_rate": 9.741756708250731e-05, "loss": 0.0876, "step": 11810 }, { "grad_norm": 0.5634013414382935, "learning_rate": 9.741100657108109e-05, "loss": 0.0701, "step": 11820 }, { "grad_norm": 0.5944783687591553, "learning_rate": 9.740443795841054e-05, "loss": 0.0836, "step": 11830 }, { "grad_norm": 0.5200526714324951, "learning_rate": 9.739786124561805e-05, "loss": 0.0654, "step": 11840 }, { "grad_norm": 0.39425966143608093, "learning_rate": 9.73912764338274e-05, "loss": 0.0683, "step": 11850 }, { "grad_norm": 0.4976321756839752, "learning_rate": 9.738468352416377e-05, "loss": 0.0767, "step": 11860 }, { "grad_norm": 0.47787609696388245, "learning_rate": 9.737808251775369e-05, "loss": 0.0742, "step": 11870 }, { "grad_norm": 0.47410646080970764, "learning_rate": 9.737147341572512e-05, "loss": 0.0788, "step": 11880 }, { "grad_norm": 0.5118182301521301, "learning_rate": 9.736485621920735e-05, "loss": 0.0639, "step": 11890 }, { "grad_norm": 0.4452165961265564, "learning_rate": 9.735823092933108e-05, "loss": 0.0753, "step": 11900 }, { "grad_norm": 0.448414146900177, "learning_rate": 9.735159754722838e-05, "loss": 0.0661, "step": 11910 }, { "grad_norm": 0.4335973560810089, "learning_rate": 9.734495607403275e-05, "loss": 0.061, "step": 11920 }, { "grad_norm": 0.46143803000450134, "learning_rate": 9.733830651087901e-05, "loss": 0.0729, "step": 11930 }, { "grad_norm": 0.4409673511981964, "learning_rate": 9.733164885890338e-05, "loss": 0.0635, "step": 11940 }, { "grad_norm": 0.45240700244903564, "learning_rate": 9.732498311924349e-05, "loss": 0.0589, "step": 11950 }, { "grad_norm": 0.6319700479507446, "learning_rate": 9.731830929303833e-05, "loss": 0.0698, "step": 11960 }, { "grad_norm": 0.4397014081478119, "learning_rate": 9.731162738142827e-05, "loss": 0.0588, "step": 11970 }, { "grad_norm": 0.6172637939453125, "learning_rate": 9.730493738555506e-05, "loss": 0.0675, "step": 11980 }, { "grad_norm": 0.8628009557723999, "learning_rate": 9.729823930656186e-05, "loss": 0.072, "step": 11990 }, { "grad_norm": 0.4729338586330414, "learning_rate": 9.729153314559316e-05, "loss": 0.0763, "step": 12000 }, { "grad_norm": 0.5150229334831238, "learning_rate": 9.728481890379486e-05, "loss": 0.0704, "step": 12010 }, { "grad_norm": 0.5475828051567078, "learning_rate": 9.727809658231428e-05, "loss": 0.0642, "step": 12020 }, { "grad_norm": 0.46736517548561096, "learning_rate": 9.727136618230003e-05, "loss": 0.0669, "step": 12030 }, { "grad_norm": 0.4753209054470062, "learning_rate": 9.726462770490219e-05, "loss": 0.0708, "step": 12040 }, { "grad_norm": 0.5074476599693298, "learning_rate": 9.725788115127214e-05, "loss": 0.0725, "step": 12050 }, { "grad_norm": 0.5337621569633484, "learning_rate": 9.725112652256274e-05, "loss": 0.0671, "step": 12060 }, { "grad_norm": 0.42688125371932983, "learning_rate": 9.724436381992812e-05, "loss": 0.0941, "step": 12070 }, { "grad_norm": 1.41287362575531, "learning_rate": 9.723759304452387e-05, "loss": 0.0807, "step": 12080 }, { "grad_norm": 0.5471555590629578, "learning_rate": 9.72308141975069e-05, "loss": 0.0625, "step": 12090 }, { "grad_norm": 0.49223870038986206, "learning_rate": 9.722402728003557e-05, "loss": 0.0759, "step": 12100 }, { "grad_norm": 0.4926984906196594, "learning_rate": 9.721723229326953e-05, "loss": 0.0806, "step": 12110 }, { "grad_norm": 0.4695297181606293, "learning_rate": 9.721042923836992e-05, "loss": 0.0618, "step": 12120 }, { "grad_norm": 0.5524077415466309, "learning_rate": 9.720361811649914e-05, "loss": 0.0675, "step": 12130 }, { "grad_norm": 0.48522084951400757, "learning_rate": 9.719679892882106e-05, "loss": 0.0704, "step": 12140 }, { "grad_norm": 0.553882360458374, "learning_rate": 9.718997167650085e-05, "loss": 0.0789, "step": 12150 }, { "grad_norm": 0.5192822217941284, "learning_rate": 9.718313636070515e-05, "loss": 0.0704, "step": 12160 }, { "grad_norm": 0.4105469584465027, "learning_rate": 9.717629298260192e-05, "loss": 0.0611, "step": 12170 }, { "grad_norm": 0.6587246656417847, "learning_rate": 9.716944154336047e-05, "loss": 0.0684, "step": 12180 }, { "grad_norm": 0.5574089288711548, "learning_rate": 9.716258204415157e-05, "loss": 0.0651, "step": 12190 }, { "grad_norm": 0.5409682989120483, "learning_rate": 9.715571448614728e-05, "loss": 0.0794, "step": 12200 }, { "grad_norm": 0.4388304054737091, "learning_rate": 9.71488388705211e-05, "loss": 0.0785, "step": 12210 }, { "grad_norm": 0.5468053221702576, "learning_rate": 9.714195519844788e-05, "loss": 0.0641, "step": 12220 }, { "grad_norm": 0.5171226263046265, "learning_rate": 9.713506347110386e-05, "loss": 0.0684, "step": 12230 }, { "grad_norm": 0.3501579761505127, "learning_rate": 9.712816368966662e-05, "loss": 0.0729, "step": 12240 }, { "grad_norm": 0.5142301917076111, "learning_rate": 9.712125585531517e-05, "loss": 0.0852, "step": 12250 }, { "grad_norm": 0.6512768864631653, "learning_rate": 9.711433996922988e-05, "loss": 0.0864, "step": 12260 }, { "grad_norm": 0.44142627716064453, "learning_rate": 9.710741603259245e-05, "loss": 0.077, "step": 12270 }, { "grad_norm": 0.5139098167419434, "learning_rate": 9.710048404658603e-05, "loss": 0.0721, "step": 12280 }, { "grad_norm": 0.4491320252418518, "learning_rate": 9.709354401239508e-05, "loss": 0.0779, "step": 12290 }, { "grad_norm": 0.5197832584381104, "learning_rate": 9.708659593120546e-05, "loss": 0.0566, "step": 12300 }, { "grad_norm": 0.37610703706741333, "learning_rate": 9.707963980420443e-05, "loss": 0.0629, "step": 12310 }, { "grad_norm": 0.5972456336021423, "learning_rate": 9.707267563258058e-05, "loss": 0.0754, "step": 12320 }, { "grad_norm": 0.4756535589694977, "learning_rate": 9.70657034175239e-05, "loss": 0.0623, "step": 12330 }, { "grad_norm": 0.47580909729003906, "learning_rate": 9.705872316022577e-05, "loss": 0.0764, "step": 12340 }, { "grad_norm": 0.5529713034629822, "learning_rate": 9.705173486187891e-05, "loss": 0.0701, "step": 12350 }, { "grad_norm": 0.41850462555885315, "learning_rate": 9.704473852367741e-05, "loss": 0.0714, "step": 12360 }, { "grad_norm": 0.44575896859169006, "learning_rate": 9.70377341468168e-05, "loss": 0.0638, "step": 12370 }, { "grad_norm": 0.5899453163146973, "learning_rate": 9.703072173249389e-05, "loss": 0.0614, "step": 12380 }, { "grad_norm": 0.5986238718032837, "learning_rate": 9.702370128190693e-05, "loss": 0.0701, "step": 12390 }, { "grad_norm": 0.47803622484207153, "learning_rate": 9.701667279625552e-05, "loss": 0.0762, "step": 12400 }, { "grad_norm": 0.5859593152999878, "learning_rate": 9.700963627674065e-05, "loss": 0.065, "step": 12410 }, { "grad_norm": 0.6580540537834167, "learning_rate": 9.700259172456466e-05, "loss": 0.0684, "step": 12420 }, { "grad_norm": 0.7720853090286255, "learning_rate": 9.699553914093124e-05, "loss": 0.0845, "step": 12430 }, { "grad_norm": 0.5211976766586304, "learning_rate": 9.698847852704553e-05, "loss": 0.0772, "step": 12440 }, { "grad_norm": 0.55735182762146, "learning_rate": 9.6981409884114e-05, "loss": 0.0746, "step": 12450 }, { "grad_norm": 0.459127813577652, "learning_rate": 9.697433321334443e-05, "loss": 0.0745, "step": 12460 }, { "grad_norm": 0.33813756704330444, "learning_rate": 9.696724851594607e-05, "loss": 0.0764, "step": 12470 }, { "grad_norm": 0.7896983623504639, "learning_rate": 9.696015579312952e-05, "loss": 0.0548, "step": 12480 }, { "grad_norm": 0.43227827548980713, "learning_rate": 9.695305504610668e-05, "loss": 0.0615, "step": 12490 }, { "grad_norm": 0.4911881387233734, "learning_rate": 9.694594627609092e-05, "loss": 0.0725, "step": 12500 }, { "grad_norm": 0.6200262308120728, "learning_rate": 9.693882948429691e-05, "loss": 0.0721, "step": 12510 }, { "grad_norm": 0.5444630980491638, "learning_rate": 9.693170467194071e-05, "loss": 0.0699, "step": 12520 }, { "grad_norm": 0.6918035745620728, "learning_rate": 9.692457184023977e-05, "loss": 0.0705, "step": 12530 }, { "grad_norm": 0.6012853384017944, "learning_rate": 9.691743099041291e-05, "loss": 0.0823, "step": 12540 }, { "grad_norm": 0.6323930025100708, "learning_rate": 9.691028212368027e-05, "loss": 0.0716, "step": 12550 }, { "grad_norm": 0.6641017198562622, "learning_rate": 9.690312524126342e-05, "loss": 0.0683, "step": 12560 }, { "grad_norm": 0.3848079741001129, "learning_rate": 9.689596034438527e-05, "loss": 0.0658, "step": 12570 }, { "grad_norm": 0.5244726538658142, "learning_rate": 9.688878743427012e-05, "loss": 0.0669, "step": 12580 }, { "grad_norm": 0.5121013522148132, "learning_rate": 9.688160651214359e-05, "loss": 0.0756, "step": 12590 }, { "grad_norm": 0.5181689262390137, "learning_rate": 9.687441757923273e-05, "loss": 0.0575, "step": 12600 }, { "grad_norm": 0.30349403619766235, "learning_rate": 9.68672206367659e-05, "loss": 0.0596, "step": 12610 }, { "grad_norm": 0.5458177924156189, "learning_rate": 9.686001568597291e-05, "loss": 0.0666, "step": 12620 }, { "grad_norm": 0.39036527276039124, "learning_rate": 9.685280272808486e-05, "loss": 0.0591, "step": 12630 }, { "grad_norm": 0.4898240268230438, "learning_rate": 9.684558176433424e-05, "loss": 0.0713, "step": 12640 }, { "grad_norm": 0.541029155254364, "learning_rate": 9.683835279595495e-05, "loss": 0.0687, "step": 12650 }, { "grad_norm": 0.49637168645858765, "learning_rate": 9.683111582418216e-05, "loss": 0.0677, "step": 12660 }, { "grad_norm": 0.4720176160335541, "learning_rate": 9.682387085025254e-05, "loss": 0.0665, "step": 12670 }, { "grad_norm": 0.6918014883995056, "learning_rate": 9.681661787540401e-05, "loss": 0.073, "step": 12680 }, { "grad_norm": 0.6071722507476807, "learning_rate": 9.680935690087593e-05, "loss": 0.0787, "step": 12690 }, { "grad_norm": 0.4909054934978485, "learning_rate": 9.680208792790901e-05, "loss": 0.0706, "step": 12700 }, { "grad_norm": 0.5297897458076477, "learning_rate": 9.679481095774529e-05, "loss": 0.0683, "step": 12710 }, { "grad_norm": 0.586498498916626, "learning_rate": 9.678752599162822e-05, "loss": 0.0667, "step": 12720 }, { "grad_norm": 0.41355350613594055, "learning_rate": 9.678023303080259e-05, "loss": 0.06, "step": 12730 }, { "grad_norm": 0.43144622445106506, "learning_rate": 9.677293207651459e-05, "loss": 0.0721, "step": 12740 }, { "grad_norm": 0.3842413127422333, "learning_rate": 9.676562313001173e-05, "loss": 0.0558, "step": 12750 }, { "grad_norm": 0.7279663681983948, "learning_rate": 9.675830619254293e-05, "loss": 0.0723, "step": 12760 }, { "grad_norm": 0.5314579606056213, "learning_rate": 9.675098126535843e-05, "loss": 0.08, "step": 12770 }, { "grad_norm": 0.6942381262779236, "learning_rate": 9.674364834970988e-05, "loss": 0.0697, "step": 12780 }, { "grad_norm": 0.5868197679519653, "learning_rate": 9.673630744685028e-05, "loss": 0.0779, "step": 12790 }, { "grad_norm": 0.47557395696640015, "learning_rate": 9.672895855803397e-05, "loss": 0.0745, "step": 12800 }, { "grad_norm": 0.5325714945793152, "learning_rate": 9.672160168451667e-05, "loss": 0.0689, "step": 12810 }, { "grad_norm": 0.4045470356941223, "learning_rate": 9.671423682755549e-05, "loss": 0.0658, "step": 12820 }, { "grad_norm": 0.6070201992988586, "learning_rate": 9.670686398840888e-05, "loss": 0.0656, "step": 12830 }, { "grad_norm": 0.46287012100219727, "learning_rate": 9.669948316833664e-05, "loss": 0.0637, "step": 12840 }, { "grad_norm": 0.7335346341133118, "learning_rate": 9.669209436859997e-05, "loss": 0.0598, "step": 12850 }, { "grad_norm": 0.5310336947441101, "learning_rate": 9.66846975904614e-05, "loss": 0.0732, "step": 12860 }, { "grad_norm": 0.455966055393219, "learning_rate": 9.667729283518483e-05, "loss": 0.0607, "step": 12870 }, { "grad_norm": 0.5177672505378723, "learning_rate": 9.666988010403557e-05, "loss": 0.0588, "step": 12880 }, { "grad_norm": 0.5165739059448242, "learning_rate": 9.66624593982802e-05, "loss": 0.0649, "step": 12890 }, { "grad_norm": 0.59503573179245, "learning_rate": 9.665503071918675e-05, "loss": 0.0891, "step": 12900 }, { "grad_norm": 0.5321347117424011, "learning_rate": 9.664759406802456e-05, "loss": 0.068, "step": 12910 }, { "grad_norm": 0.7382553219795227, "learning_rate": 9.664014944606437e-05, "loss": 0.0691, "step": 12920 }, { "grad_norm": 0.629702627658844, "learning_rate": 9.663269685457822e-05, "loss": 0.0705, "step": 12930 }, { "grad_norm": 0.5099784731864929, "learning_rate": 9.662523629483962e-05, "loss": 0.0634, "step": 12940 }, { "grad_norm": 0.5357096791267395, "learning_rate": 9.661776776812333e-05, "loss": 0.0605, "step": 12950 }, { "grad_norm": 0.45946601033210754, "learning_rate": 9.661029127570553e-05, "loss": 0.0567, "step": 12960 }, { "grad_norm": 0.46223580837249756, "learning_rate": 9.660280681886373e-05, "loss": 0.0583, "step": 12970 }, { "grad_norm": 0.44987237453460693, "learning_rate": 9.659531439887685e-05, "loss": 0.0567, "step": 12980 }, { "grad_norm": 0.44017326831817627, "learning_rate": 9.658781401702511e-05, "loss": 0.0587, "step": 12990 }, { "grad_norm": 0.46805351972579956, "learning_rate": 9.658030567459015e-05, "loss": 0.0661, "step": 13000 }, { "grad_norm": 0.583610475063324, "learning_rate": 9.65727893728549e-05, "loss": 0.0871, "step": 13010 }, { "grad_norm": 0.6627160310745239, "learning_rate": 9.656526511310375e-05, "loss": 0.0808, "step": 13020 }, { "grad_norm": 0.5809372067451477, "learning_rate": 9.655773289662233e-05, "loss": 0.0743, "step": 13030 }, { "grad_norm": 0.5028517246246338, "learning_rate": 9.655019272469772e-05, "loss": 0.0634, "step": 13040 }, { "grad_norm": 0.7129963636398315, "learning_rate": 9.654264459861832e-05, "loss": 0.0603, "step": 13050 }, { "grad_norm": 0.4750165045261383, "learning_rate": 9.653508851967391e-05, "loss": 0.0589, "step": 13060 }, { "grad_norm": 0.4426215887069702, "learning_rate": 9.65275244891556e-05, "loss": 0.0809, "step": 13070 }, { "grad_norm": 0.5911847352981567, "learning_rate": 9.651995250835591e-05, "loss": 0.0568, "step": 13080 }, { "grad_norm": 0.5024567246437073, "learning_rate": 9.651237257856862e-05, "loss": 0.0679, "step": 13090 }, { "grad_norm": 0.7512474656105042, "learning_rate": 9.6504784701089e-05, "loss": 0.072, "step": 13100 }, { "grad_norm": 0.5578538775444031, "learning_rate": 9.649718887721357e-05, "loss": 0.0694, "step": 13110 }, { "grad_norm": 0.6599379777908325, "learning_rate": 9.648958510824028e-05, "loss": 0.0591, "step": 13120 }, { "grad_norm": 0.46119460463523865, "learning_rate": 9.648197339546837e-05, "loss": 0.0777, "step": 13130 }, { "grad_norm": 0.6064987182617188, "learning_rate": 9.647435374019851e-05, "loss": 0.0715, "step": 13140 }, { "grad_norm": 0.5132782459259033, "learning_rate": 9.646672614373266e-05, "loss": 0.0705, "step": 13150 }, { "grad_norm": 0.43166783452033997, "learning_rate": 9.645909060737418e-05, "loss": 0.0519, "step": 13160 }, { "grad_norm": 0.4727652966976166, "learning_rate": 9.645144713242778e-05, "loss": 0.0635, "step": 13170 }, { "grad_norm": 0.4390740394592285, "learning_rate": 9.64437957201995e-05, "loss": 0.0709, "step": 13180 }, { "grad_norm": 0.43201467394828796, "learning_rate": 9.643613637199678e-05, "loss": 0.0488, "step": 13190 }, { "grad_norm": 0.49980124831199646, "learning_rate": 9.642846908912839e-05, "loss": 0.0601, "step": 13200 }, { "grad_norm": 0.42739060521125793, "learning_rate": 9.642079387290444e-05, "loss": 0.0511, "step": 13210 }, { "grad_norm": 0.43815878033638, "learning_rate": 9.641311072463644e-05, "loss": 0.0617, "step": 13220 }, { "grad_norm": 0.46281665563583374, "learning_rate": 9.640541964563722e-05, "loss": 0.0695, "step": 13230 }, { "grad_norm": 1.2241469621658325, "learning_rate": 9.639772063722096e-05, "loss": 0.0772, "step": 13240 }, { "grad_norm": 0.4418402314186096, "learning_rate": 9.639001370070324e-05, "loss": 0.0761, "step": 13250 }, { "grad_norm": 0.5697529315948486, "learning_rate": 9.638229883740095e-05, "loss": 0.07, "step": 13260 }, { "grad_norm": 0.4529099762439728, "learning_rate": 9.637457604863233e-05, "loss": 0.0628, "step": 13270 }, { "grad_norm": 0.6263208985328674, "learning_rate": 9.636684533571703e-05, "loss": 0.0592, "step": 13280 }, { "grad_norm": 0.5833448767662048, "learning_rate": 9.635910669997599e-05, "loss": 0.0667, "step": 13290 }, { "grad_norm": 0.49336129426956177, "learning_rate": 9.635136014273154e-05, "loss": 0.0595, "step": 13300 }, { "grad_norm": 0.4078526496887207, "learning_rate": 9.634360566530735e-05, "loss": 0.0746, "step": 13310 }, { "grad_norm": 0.6260297894477844, "learning_rate": 9.633584326902845e-05, "loss": 0.0824, "step": 13320 }, { "grad_norm": 0.5920906066894531, "learning_rate": 9.632807295522124e-05, "loss": 0.0664, "step": 13330 }, { "grad_norm": 0.5717576742172241, "learning_rate": 9.632029472521342e-05, "loss": 0.0777, "step": 13340 }, { "grad_norm": 0.4631733298301697, "learning_rate": 9.631250858033409e-05, "loss": 0.0568, "step": 13350 }, { "grad_norm": 0.40033048391342163, "learning_rate": 9.630471452191371e-05, "loss": 0.0759, "step": 13360 }, { "grad_norm": 0.4062081575393677, "learning_rate": 9.629691255128405e-05, "loss": 0.0866, "step": 13370 }, { "grad_norm": 0.7073371410369873, "learning_rate": 9.628910266977825e-05, "loss": 0.0681, "step": 13380 }, { "grad_norm": 0.4445957541465759, "learning_rate": 9.628128487873083e-05, "loss": 0.0708, "step": 13390 }, { "grad_norm": 0.4187319874763489, "learning_rate": 9.627345917947761e-05, "loss": 0.0806, "step": 13400 }, { "grad_norm": 0.4601149559020996, "learning_rate": 9.626562557335579e-05, "loss": 0.066, "step": 13410 }, { "grad_norm": 0.38216379284858704, "learning_rate": 9.625778406170393e-05, "loss": 0.0638, "step": 13420 }, { "grad_norm": 0.5466590523719788, "learning_rate": 9.624993464586193e-05, "loss": 0.0718, "step": 13430 }, { "grad_norm": 0.5218749046325684, "learning_rate": 9.624207732717105e-05, "loss": 0.0766, "step": 13440 }, { "grad_norm": 0.714627742767334, "learning_rate": 9.623421210697386e-05, "loss": 0.07, "step": 13450 }, { "grad_norm": 0.4092749357223511, "learning_rate": 9.622633898661434e-05, "loss": 0.0819, "step": 13460 }, { "grad_norm": 0.49629244208335876, "learning_rate": 9.621845796743778e-05, "loss": 0.0838, "step": 13470 }, { "grad_norm": 0.4324739873409271, "learning_rate": 9.621056905079082e-05, "loss": 0.0749, "step": 13480 }, { "grad_norm": 0.43576082587242126, "learning_rate": 9.620267223802149e-05, "loss": 0.0666, "step": 13490 }, { "grad_norm": 0.5109706521034241, "learning_rate": 9.619476753047911e-05, "loss": 0.0606, "step": 13500 }, { "grad_norm": 0.46637558937072754, "learning_rate": 9.618685492951438e-05, "loss": 0.0672, "step": 13510 }, { "grad_norm": 0.4376745820045471, "learning_rate": 9.617893443647938e-05, "loss": 0.0674, "step": 13520 }, { "grad_norm": 0.48372602462768555, "learning_rate": 9.617100605272746e-05, "loss": 0.0718, "step": 13530 }, { "grad_norm": 0.43598562479019165, "learning_rate": 9.616306977961338e-05, "loss": 0.0728, "step": 13540 }, { "grad_norm": 0.41948139667510986, "learning_rate": 9.615512561849326e-05, "loss": 0.0517, "step": 13550 }, { "grad_norm": 0.4470612108707428, "learning_rate": 9.61471735707245e-05, "loss": 0.0617, "step": 13560 }, { "grad_norm": 0.505235493183136, "learning_rate": 9.613921363766592e-05, "loss": 0.0639, "step": 13570 }, { "grad_norm": 0.5572078227996826, "learning_rate": 9.613124582067763e-05, "loss": 0.0693, "step": 13580 }, { "grad_norm": 0.592846155166626, "learning_rate": 9.612327012112112e-05, "loss": 0.0697, "step": 13590 }, { "grad_norm": 0.5719813108444214, "learning_rate": 9.611528654035921e-05, "loss": 0.0809, "step": 13600 }, { "grad_norm": 0.5089952945709229, "learning_rate": 9.610729507975611e-05, "loss": 0.0644, "step": 13610 }, { "grad_norm": 0.502956211566925, "learning_rate": 9.609929574067731e-05, "loss": 0.077, "step": 13620 }, { "grad_norm": 0.46921467781066895, "learning_rate": 9.609128852448967e-05, "loss": 0.0573, "step": 13630 }, { "grad_norm": 0.4989677369594574, "learning_rate": 9.608327343256143e-05, "loss": 0.061, "step": 13640 }, { "grad_norm": 0.4597410559654236, "learning_rate": 9.607525046626216e-05, "loss": 0.0659, "step": 13650 }, { "grad_norm": 0.6454246044158936, "learning_rate": 9.606721962696272e-05, "loss": 0.0699, "step": 13660 }, { "grad_norm": 0.4892130196094513, "learning_rate": 9.60591809160354e-05, "loss": 0.0653, "step": 13670 }, { "grad_norm": 0.38182902336120605, "learning_rate": 9.605113433485378e-05, "loss": 0.0537, "step": 13680 }, { "grad_norm": 0.6010304093360901, "learning_rate": 9.604307988479279e-05, "loss": 0.0591, "step": 13690 }, { "grad_norm": 0.4232689440250397, "learning_rate": 9.603501756722876e-05, "loss": 0.0628, "step": 13700 }, { "grad_norm": 0.46342718601226807, "learning_rate": 9.602694738353927e-05, "loss": 0.0597, "step": 13710 }, { "grad_norm": 0.35658493638038635, "learning_rate": 9.601886933510331e-05, "loss": 0.0661, "step": 13720 }, { "grad_norm": 0.4141998291015625, "learning_rate": 9.60107834233012e-05, "loss": 0.0642, "step": 13730 }, { "grad_norm": 0.5906035900115967, "learning_rate": 9.60026896495146e-05, "loss": 0.0644, "step": 13740 }, { "grad_norm": 0.8446847796440125, "learning_rate": 9.599458801512652e-05, "loss": 0.0693, "step": 13750 }, { "grad_norm": 0.5100727081298828, "learning_rate": 9.598647852152129e-05, "loss": 0.085, "step": 13760 }, { "grad_norm": 0.52426677942276, "learning_rate": 9.597836117008462e-05, "loss": 0.0614, "step": 13770 }, { "grad_norm": 0.42374324798583984, "learning_rate": 9.597023596220356e-05, "loss": 0.0583, "step": 13780 }, { "grad_norm": 1.2665348052978516, "learning_rate": 9.596210289926643e-05, "loss": 0.072, "step": 13790 }, { "grad_norm": 0.5634767413139343, "learning_rate": 9.5953961982663e-05, "loss": 0.0622, "step": 13800 }, { "grad_norm": 0.7088893055915833, "learning_rate": 9.594581321378431e-05, "loss": 0.079, "step": 13810 }, { "grad_norm": 0.42110151052474976, "learning_rate": 9.593765659402276e-05, "loss": 0.0744, "step": 13820 }, { "grad_norm": 0.6793797016143799, "learning_rate": 9.59294921247721e-05, "loss": 0.0697, "step": 13830 }, { "grad_norm": 0.5676230788230896, "learning_rate": 9.59213198074274e-05, "loss": 0.0756, "step": 13840 }, { "grad_norm": 0.43754056096076965, "learning_rate": 9.59131396433851e-05, "loss": 0.0537, "step": 13850 }, { "grad_norm": 0.3993590474128723, "learning_rate": 9.590495163404297e-05, "loss": 0.0607, "step": 13860 }, { "grad_norm": 0.456595778465271, "learning_rate": 9.589675578080009e-05, "loss": 0.0593, "step": 13870 }, { "grad_norm": 0.4655943214893341, "learning_rate": 9.588855208505694e-05, "loss": 0.0509, "step": 13880 }, { "grad_norm": 0.46786022186279297, "learning_rate": 9.588034054821529e-05, "loss": 0.0678, "step": 13890 }, { "grad_norm": 0.42947566509246826, "learning_rate": 9.587212117167826e-05, "loss": 0.0595, "step": 13900 }, { "grad_norm": 0.4759654402732849, "learning_rate": 9.586389395685033e-05, "loss": 0.0595, "step": 13910 }, { "grad_norm": 0.5599742531776428, "learning_rate": 9.585565890513733e-05, "loss": 0.0677, "step": 13920 }, { "grad_norm": 0.525069534778595, "learning_rate": 9.584741601794636e-05, "loss": 0.0648, "step": 13930 }, { "grad_norm": 0.4940713942050934, "learning_rate": 9.58391652966859e-05, "loss": 0.0616, "step": 13940 }, { "grad_norm": 0.42760440707206726, "learning_rate": 9.583090674276583e-05, "loss": 0.0527, "step": 13950 }, { "grad_norm": 0.5313862562179565, "learning_rate": 9.582264035759726e-05, "loss": 0.0688, "step": 13960 }, { "grad_norm": 0.4666474759578705, "learning_rate": 9.58143661425927e-05, "loss": 0.0946, "step": 13970 }, { "grad_norm": 0.5030765533447266, "learning_rate": 9.580608409916601e-05, "loss": 0.0741, "step": 13980 }, { "grad_norm": 0.4865291118621826, "learning_rate": 9.579779422873233e-05, "loss": 0.062, "step": 13990 }, { "grad_norm": 0.4550444781780243, "learning_rate": 9.578949653270819e-05, "loss": 0.0525, "step": 14000 }, { "grad_norm": 0.46030089259147644, "learning_rate": 9.578119101251144e-05, "loss": 0.0644, "step": 14010 }, { "grad_norm": 0.44403159618377686, "learning_rate": 9.577287766956127e-05, "loss": 0.0662, "step": 14020 }, { "grad_norm": 0.46070826053619385, "learning_rate": 9.57645565052782e-05, "loss": 0.0732, "step": 14030 }, { "grad_norm": 0.5882596373558044, "learning_rate": 9.575622752108407e-05, "loss": 0.0617, "step": 14040 }, { "grad_norm": 0.4353800117969513, "learning_rate": 9.57478907184021e-05, "loss": 0.0555, "step": 14050 }, { "grad_norm": 0.5321919918060303, "learning_rate": 9.573954609865681e-05, "loss": 0.0635, "step": 14060 }, { "grad_norm": 0.6407208442687988, "learning_rate": 9.573119366327408e-05, "loss": 0.0676, "step": 14070 }, { "grad_norm": 0.3554493486881256, "learning_rate": 9.57228334136811e-05, "loss": 0.06, "step": 14080 }, { "grad_norm": 0.37713468074798584, "learning_rate": 9.571446535130641e-05, "loss": 0.0667, "step": 14090 }, { "grad_norm": 0.469918429851532, "learning_rate": 9.570608947757988e-05, "loss": 0.0667, "step": 14100 }, { "grad_norm": 0.5037508606910706, "learning_rate": 9.569770579393274e-05, "loss": 0.0692, "step": 14110 }, { "grad_norm": 0.5194280743598938, "learning_rate": 9.56893143017975e-05, "loss": 0.0674, "step": 14120 }, { "grad_norm": 0.5188848972320557, "learning_rate": 9.568091500260806e-05, "loss": 0.0582, "step": 14130 }, { "grad_norm": 0.5067747235298157, "learning_rate": 9.567250789779961e-05, "loss": 0.0684, "step": 14140 }, { "grad_norm": 0.38894012570381165, "learning_rate": 9.566409298880872e-05, "loss": 0.0619, "step": 14150 }, { "grad_norm": 0.7821222543716431, "learning_rate": 9.565567027707326e-05, "loss": 0.0851, "step": 14160 }, { "grad_norm": 0.5528215765953064, "learning_rate": 9.56472397640324e-05, "loss": 0.0663, "step": 14170 }, { "grad_norm": 0.5752657651901245, "learning_rate": 9.563880145112675e-05, "loss": 0.0768, "step": 14180 }, { "grad_norm": 0.6070843935012817, "learning_rate": 9.563035533979814e-05, "loss": 0.072, "step": 14190 }, { "grad_norm": 0.4512389898300171, "learning_rate": 9.562190143148981e-05, "loss": 0.065, "step": 14200 }, { "grad_norm": 0.3500981628894806, "learning_rate": 9.561343972764627e-05, "loss": 0.0634, "step": 14210 }, { "grad_norm": 0.5131586790084839, "learning_rate": 9.560497022971343e-05, "loss": 0.0693, "step": 14220 }, { "grad_norm": 0.5896312594413757, "learning_rate": 9.559649293913847e-05, "loss": 0.0665, "step": 14230 }, { "grad_norm": 0.5618447065353394, "learning_rate": 9.558800785736993e-05, "loss": 0.074, "step": 14240 }, { "grad_norm": 0.43290871381759644, "learning_rate": 9.557951498585767e-05, "loss": 0.0707, "step": 14250 }, { "grad_norm": 0.46709153056144714, "learning_rate": 9.557101432605293e-05, "loss": 0.0649, "step": 14260 }, { "grad_norm": 0.7131912112236023, "learning_rate": 9.556250587940818e-05, "loss": 0.0788, "step": 14270 }, { "grad_norm": 0.6857901215553284, "learning_rate": 9.555398964737734e-05, "loss": 0.0682, "step": 14280 }, { "grad_norm": 0.4148392975330353, "learning_rate": 9.554546563141555e-05, "loss": 0.0581, "step": 14290 }, { "grad_norm": 0.4593449532985687, "learning_rate": 9.553693383297937e-05, "loss": 0.0656, "step": 14300 }, { "grad_norm": 0.5164635181427002, "learning_rate": 9.552839425352663e-05, "loss": 0.0596, "step": 14310 }, { "grad_norm": 0.5512014627456665, "learning_rate": 9.551984689451652e-05, "loss": 0.072, "step": 14320 }, { "grad_norm": 0.46850889921188354, "learning_rate": 9.551129175740953e-05, "loss": 0.0667, "step": 14330 }, { "grad_norm": 0.4253566563129425, "learning_rate": 9.550272884366754e-05, "loss": 0.0726, "step": 14340 }, { "grad_norm": 0.6134167909622192, "learning_rate": 9.549415815475369e-05, "loss": 0.0674, "step": 14350 }, { "grad_norm": 0.6227335929870605, "learning_rate": 9.548557969213247e-05, "loss": 0.0621, "step": 14360 }, { "grad_norm": 0.42364680767059326, "learning_rate": 9.547699345726972e-05, "loss": 0.0669, "step": 14370 }, { "grad_norm": 0.44078412652015686, "learning_rate": 9.546839945163257e-05, "loss": 0.0583, "step": 14380 }, { "grad_norm": 0.6047835350036621, "learning_rate": 9.545979767668953e-05, "loss": 0.0793, "step": 14390 }, { "grad_norm": 0.5372642874717712, "learning_rate": 9.54511881339104e-05, "loss": 0.0736, "step": 14400 }, { "grad_norm": 0.6143385171890259, "learning_rate": 9.54425708247663e-05, "loss": 0.0617, "step": 14410 }, { "grad_norm": 0.457565575838089, "learning_rate": 9.543394575072972e-05, "loss": 0.0655, "step": 14420 }, { "grad_norm": 0.4323413372039795, "learning_rate": 9.542531291327441e-05, "loss": 0.0664, "step": 14430 }, { "grad_norm": 0.5343565344810486, "learning_rate": 9.541667231387552e-05, "loss": 0.058, "step": 14440 }, { "grad_norm": 0.5047639012336731, "learning_rate": 9.540802395400949e-05, "loss": 0.0638, "step": 14450 }, { "grad_norm": 0.6144313812255859, "learning_rate": 9.539936783515406e-05, "loss": 0.0602, "step": 14460 }, { "grad_norm": 0.4427989721298218, "learning_rate": 9.539070395878835e-05, "loss": 0.0631, "step": 14470 }, { "grad_norm": 0.42528507113456726, "learning_rate": 9.538203232639277e-05, "loss": 0.0588, "step": 14480 }, { "grad_norm": 0.6187255382537842, "learning_rate": 9.537335293944907e-05, "loss": 0.0629, "step": 14490 }, { "grad_norm": 0.5377559065818787, "learning_rate": 9.536466579944032e-05, "loss": 0.0782, "step": 14500 }, { "grad_norm": 0.42575210332870483, "learning_rate": 9.535597090785091e-05, "loss": 0.067, "step": 14510 }, { "grad_norm": 0.5066750645637512, "learning_rate": 9.534726826616656e-05, "loss": 0.055, "step": 14520 }, { "grad_norm": 0.38841962814331055, "learning_rate": 9.53385578758743e-05, "loss": 0.0832, "step": 14530 }, { "grad_norm": 0.4399453401565552, "learning_rate": 9.532983973846252e-05, "loss": 0.0573, "step": 14540 }, { "grad_norm": 0.45209014415740967, "learning_rate": 9.53211138554209e-05, "loss": 0.057, "step": 14550 }, { "grad_norm": 0.4618039131164551, "learning_rate": 9.531238022824047e-05, "loss": 0.0595, "step": 14560 }, { "grad_norm": 0.5368629097938538, "learning_rate": 9.530363885841355e-05, "loss": 0.0552, "step": 14570 }, { "grad_norm": 0.5297271609306335, "learning_rate": 9.52948897474338e-05, "loss": 0.0634, "step": 14580 }, { "grad_norm": 0.6465821862220764, "learning_rate": 9.528613289679622e-05, "loss": 0.0772, "step": 14590 }, { "grad_norm": 0.4754258096218109, "learning_rate": 9.52773683079971e-05, "loss": 0.0676, "step": 14600 }, { "grad_norm": 0.47023913264274597, "learning_rate": 9.526859598253407e-05, "loss": 0.0603, "step": 14610 }, { "grad_norm": 0.5196773409843445, "learning_rate": 9.525981592190609e-05, "loss": 0.0612, "step": 14620 }, { "grad_norm": 0.5412208437919617, "learning_rate": 9.525102812761342e-05, "loss": 0.0568, "step": 14630 }, { "grad_norm": 0.39522358775138855, "learning_rate": 9.524223260115768e-05, "loss": 0.0493, "step": 14640 }, { "grad_norm": 0.45558539032936096, "learning_rate": 9.523342934404175e-05, "loss": 0.0595, "step": 14650 }, { "grad_norm": 0.4912519156932831, "learning_rate": 9.522461835776989e-05, "loss": 0.0626, "step": 14660 }, { "grad_norm": 0.8292794823646545, "learning_rate": 9.521579964384764e-05, "loss": 0.0772, "step": 14670 }, { "grad_norm": 0.5148866772651672, "learning_rate": 9.52069732037819e-05, "loss": 0.0611, "step": 14680 }, { "grad_norm": 0.4858807623386383, "learning_rate": 9.519813903908083e-05, "loss": 0.063, "step": 14690 }, { "grad_norm": 0.4394304156303406, "learning_rate": 9.5189297151254e-05, "loss": 0.0614, "step": 14700 }, { "grad_norm": 0.7162570953369141, "learning_rate": 9.518044754181218e-05, "loss": 0.0703, "step": 14710 }, { "grad_norm": 0.5640280842781067, "learning_rate": 9.51715902122676e-05, "loss": 0.0674, "step": 14720 }, { "grad_norm": 0.6195242404937744, "learning_rate": 9.516272516413368e-05, "loss": 0.0666, "step": 14730 }, { "grad_norm": 0.47283053398132324, "learning_rate": 9.515385239892525e-05, "loss": 0.0845, "step": 14740 }, { "grad_norm": 0.4446857273578644, "learning_rate": 9.514497191815839e-05, "loss": 0.0649, "step": 14750 }, { "grad_norm": 0.37392014265060425, "learning_rate": 9.513608372335055e-05, "loss": 0.0644, "step": 14760 }, { "grad_norm": 0.5323464870452881, "learning_rate": 9.512718781602045e-05, "loss": 0.0694, "step": 14770 }, { "grad_norm": 0.5024527311325073, "learning_rate": 9.511828419768823e-05, "loss": 0.0633, "step": 14780 }, { "grad_norm": 0.48092085123062134, "learning_rate": 9.510937286987521e-05, "loss": 0.0544, "step": 14790 }, { "grad_norm": 0.7296614050865173, "learning_rate": 9.510045383410408e-05, "loss": 0.0711, "step": 14800 }, { "grad_norm": 0.45427438616752625, "learning_rate": 9.509152709189892e-05, "loss": 0.053, "step": 14810 }, { "grad_norm": 0.4783822298049927, "learning_rate": 9.508259264478504e-05, "loss": 0.0718, "step": 14820 }, { "grad_norm": 0.5371935963630676, "learning_rate": 9.507365049428909e-05, "loss": 0.07, "step": 14830 }, { "grad_norm": 0.5308293700218201, "learning_rate": 9.506470064193902e-05, "loss": 0.0755, "step": 14840 }, { "grad_norm": 0.6376842856407166, "learning_rate": 9.505574308926414e-05, "loss": 0.0766, "step": 14850 }, { "grad_norm": 0.4896347224712372, "learning_rate": 9.504677783779505e-05, "loss": 0.0676, "step": 14860 }, { "grad_norm": 0.45003005862236023, "learning_rate": 9.503780488906365e-05, "loss": 0.0603, "step": 14870 }, { "grad_norm": 0.5541497468948364, "learning_rate": 9.502882424460319e-05, "loss": 0.06, "step": 14880 }, { "grad_norm": 0.5367873907089233, "learning_rate": 9.501983590594821e-05, "loss": 0.0705, "step": 14890 }, { "grad_norm": 0.5817331671714783, "learning_rate": 9.501083987463455e-05, "loss": 0.0679, "step": 14900 }, { "grad_norm": 0.5602604746818542, "learning_rate": 9.500183615219942e-05, "loss": 0.0651, "step": 14910 }, { "grad_norm": 0.5003082156181335, "learning_rate": 9.49928247401813e-05, "loss": 0.0521, "step": 14920 }, { "grad_norm": 0.5456083416938782, "learning_rate": 9.498380564011997e-05, "loss": 0.0586, "step": 14930 }, { "grad_norm": 0.5838845372200012, "learning_rate": 9.497477885355656e-05, "loss": 0.0777, "step": 14940 }, { "grad_norm": 0.5460173487663269, "learning_rate": 9.496574438203353e-05, "loss": 0.0766, "step": 14950 }, { "grad_norm": 0.48008400201797485, "learning_rate": 9.495670222709459e-05, "loss": 0.0663, "step": 14960 }, { "grad_norm": 0.40302956104278564, "learning_rate": 9.494765239028483e-05, "loss": 0.0522, "step": 14970 }, { "grad_norm": 0.6059410572052002, "learning_rate": 9.493859487315057e-05, "loss": 0.0606, "step": 14980 }, { "grad_norm": 0.44906291365623474, "learning_rate": 9.492952967723953e-05, "loss": 0.0591, "step": 14990 }, { "grad_norm": 0.5250199437141418, "learning_rate": 9.492045680410068e-05, "loss": 0.0703, "step": 15000 }, { "grad_norm": 0.5336766839027405, "learning_rate": 9.491137625528436e-05, "loss": 0.0619, "step": 15010 }, { "grad_norm": 0.48436376452445984, "learning_rate": 9.490228803234215e-05, "loss": 0.0661, "step": 15020 }, { "grad_norm": 0.7936002016067505, "learning_rate": 9.489319213682701e-05, "loss": 0.0688, "step": 15030 }, { "grad_norm": 0.3609626889228821, "learning_rate": 9.488408857029316e-05, "loss": 0.0682, "step": 15040 }, { "grad_norm": 0.4212193787097931, "learning_rate": 9.487497733429616e-05, "loss": 0.0801, "step": 15050 }, { "grad_norm": 0.485275000333786, "learning_rate": 9.486585843039286e-05, "loss": 0.0778, "step": 15060 }, { "grad_norm": 0.5388281941413879, "learning_rate": 9.485673186014143e-05, "loss": 0.0587, "step": 15070 }, { "grad_norm": 0.4555697441101074, "learning_rate": 9.484759762510137e-05, "loss": 0.0626, "step": 15080 }, { "grad_norm": 0.4726381003856659, "learning_rate": 9.483845572683346e-05, "loss": 0.0745, "step": 15090 }, { "grad_norm": 0.5239920020103455, "learning_rate": 9.48293061668998e-05, "loss": 0.055, "step": 15100 }, { "grad_norm": 0.4965386688709259, "learning_rate": 9.48201489468638e-05, "loss": 0.0677, "step": 15110 }, { "grad_norm": 0.5860244631767273, "learning_rate": 9.481098406829016e-05, "loss": 0.0633, "step": 15120 }, { "grad_norm": 0.5226569771766663, "learning_rate": 9.480181153274495e-05, "loss": 0.065, "step": 15130 }, { "grad_norm": 0.5421877503395081, "learning_rate": 9.479263134179548e-05, "loss": 0.0523, "step": 15140 }, { "grad_norm": 0.6064016819000244, "learning_rate": 9.478344349701039e-05, "loss": 0.0655, "step": 15150 }, { "grad_norm": 0.4677354395389557, "learning_rate": 9.477424799995964e-05, "loss": 0.066, "step": 15160 }, { "grad_norm": 0.44992202520370483, "learning_rate": 9.476504485221448e-05, "loss": 0.0678, "step": 15170 }, { "grad_norm": 0.4492607116699219, "learning_rate": 9.475583405534748e-05, "loss": 0.0567, "step": 15180 }, { "grad_norm": 0.6467487812042236, "learning_rate": 9.474661561093251e-05, "loss": 0.0596, "step": 15190 }, { "grad_norm": 0.5050837397575378, "learning_rate": 9.473738952054478e-05, "loss": 0.0625, "step": 15200 }, { "grad_norm": 0.41910600662231445, "learning_rate": 9.472815578576073e-05, "loss": 0.0621, "step": 15210 }, { "grad_norm": 0.5831831693649292, "learning_rate": 9.471891440815817e-05, "loss": 0.067, "step": 15220 }, { "grad_norm": 0.46198537945747375, "learning_rate": 9.470966538931621e-05, "loss": 0.0527, "step": 15230 }, { "grad_norm": 0.4151425063610077, "learning_rate": 9.470040873081525e-05, "loss": 0.0513, "step": 15240 }, { "grad_norm": 0.7539243698120117, "learning_rate": 9.469114443423698e-05, "loss": 0.0779, "step": 15250 }, { "grad_norm": 0.5392773151397705, "learning_rate": 9.468187250116445e-05, "loss": 0.0614, "step": 15260 }, { "grad_norm": 0.5989169478416443, "learning_rate": 9.467259293318197e-05, "loss": 0.0629, "step": 15270 }, { "grad_norm": 0.4141986072063446, "learning_rate": 9.466330573187514e-05, "loss": 0.0569, "step": 15280 }, { "grad_norm": 0.4631454348564148, "learning_rate": 9.46540108988309e-05, "loss": 0.0696, "step": 15290 }, { "grad_norm": 0.5661997199058533, "learning_rate": 9.46447084356375e-05, "loss": 0.0666, "step": 15300 }, { "grad_norm": 0.45535436272621155, "learning_rate": 9.463539834388447e-05, "loss": 0.0648, "step": 15310 }, { "grad_norm": 0.3616853654384613, "learning_rate": 9.462608062516263e-05, "loss": 0.0481, "step": 15320 }, { "grad_norm": 0.49670541286468506, "learning_rate": 9.461675528106413e-05, "loss": 0.0578, "step": 15330 }, { "grad_norm": 0.3948276937007904, "learning_rate": 9.460742231318244e-05, "loss": 0.057, "step": 15340 }, { "grad_norm": 0.7175767421722412, "learning_rate": 9.459808172311229e-05, "loss": 0.0696, "step": 15350 }, { "grad_norm": 0.4298359155654907, "learning_rate": 9.458873351244972e-05, "loss": 0.053, "step": 15360 }, { "grad_norm": 0.5673350691795349, "learning_rate": 9.457937768279211e-05, "loss": 0.0606, "step": 15370 }, { "grad_norm": 0.5679828524589539, "learning_rate": 9.45700142357381e-05, "loss": 0.0596, "step": 15380 }, { "grad_norm": 0.4382030665874481, "learning_rate": 9.456064317288765e-05, "loss": 0.0626, "step": 15390 }, { "grad_norm": 0.4954645037651062, "learning_rate": 9.455126449584201e-05, "loss": 0.0564, "step": 15400 }, { "grad_norm": 0.5111734867095947, "learning_rate": 9.454187820620375e-05, "loss": 0.055, "step": 15410 }, { "grad_norm": 0.41514071822166443, "learning_rate": 9.453248430557673e-05, "loss": 0.0541, "step": 15420 }, { "grad_norm": 0.5505982041358948, "learning_rate": 9.452308279556611e-05, "loss": 0.0834, "step": 15430 }, { "grad_norm": 0.4559839963912964, "learning_rate": 9.451367367777835e-05, "loss": 0.0541, "step": 15440 }, { "grad_norm": 0.43780091404914856, "learning_rate": 9.450425695382122e-05, "loss": 0.0623, "step": 15450 }, { "grad_norm": 0.4328004717826843, "learning_rate": 9.449483262530375e-05, "loss": 0.0551, "step": 15460 }, { "grad_norm": 0.5346675515174866, "learning_rate": 9.448540069383633e-05, "loss": 0.0548, "step": 15470 }, { "grad_norm": 0.5515182018280029, "learning_rate": 9.447596116103061e-05, "loss": 0.0604, "step": 15480 }, { "grad_norm": 0.47372350096702576, "learning_rate": 9.446651402849955e-05, "loss": 0.0634, "step": 15490 }, { "grad_norm": 0.3762182295322418, "learning_rate": 9.44570592978574e-05, "loss": 0.056, "step": 15500 }, { "grad_norm": 0.4448675513267517, "learning_rate": 9.444759697071972e-05, "loss": 0.0664, "step": 15510 }, { "grad_norm": 0.33103078603744507, "learning_rate": 9.443812704870336e-05, "loss": 0.0594, "step": 15520 }, { "grad_norm": 0.5099455118179321, "learning_rate": 9.442864953342649e-05, "loss": 0.0579, "step": 15530 }, { "grad_norm": 0.5295532941818237, "learning_rate": 9.441916442650852e-05, "loss": 0.0548, "step": 15540 }, { "grad_norm": 0.4555094540119171, "learning_rate": 9.440967172957023e-05, "loss": 0.0572, "step": 15550 }, { "grad_norm": 0.44758161902427673, "learning_rate": 9.440017144423364e-05, "loss": 0.0642, "step": 15560 }, { "grad_norm": 0.44421622157096863, "learning_rate": 9.439066357212209e-05, "loss": 0.065, "step": 15570 }, { "grad_norm": 0.4382454752922058, "learning_rate": 9.438114811486022e-05, "loss": 0.0743, "step": 15580 }, { "grad_norm": 0.3652874827384949, "learning_rate": 9.4371625074074e-05, "loss": 0.0605, "step": 15590 }, { "grad_norm": 0.37606281042099, "learning_rate": 9.436209445139059e-05, "loss": 0.0554, "step": 15600 }, { "grad_norm": 0.4193786084651947, "learning_rate": 9.435255624843855e-05, "loss": 0.0679, "step": 15610 }, { "grad_norm": 0.43372946977615356, "learning_rate": 9.43430104668477e-05, "loss": 0.0515, "step": 15620 }, { "grad_norm": 0.6009494662284851, "learning_rate": 9.433345710824914e-05, "loss": 0.0557, "step": 15630 }, { "grad_norm": 0.41870078444480896, "learning_rate": 9.432389617427529e-05, "loss": 0.0531, "step": 15640 }, { "grad_norm": 0.5964620113372803, "learning_rate": 9.431432766655984e-05, "loss": 0.0528, "step": 15650 }, { "grad_norm": 0.4538164734840393, "learning_rate": 9.430475158673778e-05, "loss": 0.0588, "step": 15660 }, { "grad_norm": 0.4989793598651886, "learning_rate": 9.429516793644542e-05, "loss": 0.0588, "step": 15670 }, { "grad_norm": 0.5303258299827576, "learning_rate": 9.428557671732034e-05, "loss": 0.0571, "step": 15680 }, { "grad_norm": 0.4449562132358551, "learning_rate": 9.42759779310014e-05, "loss": 0.049, "step": 15690 }, { "grad_norm": 0.558624267578125, "learning_rate": 9.426637157912879e-05, "loss": 0.0628, "step": 15700 }, { "grad_norm": 0.5549849271774292, "learning_rate": 9.425675766334397e-05, "loss": 0.0693, "step": 15710 }, { "grad_norm": 0.6356759071350098, "learning_rate": 9.424713618528968e-05, "loss": 0.0646, "step": 15720 }, { "grad_norm": 0.5164716243743896, "learning_rate": 9.423750714661e-05, "loss": 0.0733, "step": 15730 }, { "grad_norm": 0.6939764618873596, "learning_rate": 9.422787054895022e-05, "loss": 0.0645, "step": 15740 }, { "grad_norm": 0.4152311384677887, "learning_rate": 9.4218226393957e-05, "loss": 0.0626, "step": 15750 }, { "grad_norm": 0.5085019469261169, "learning_rate": 9.420857468327828e-05, "loss": 0.0599, "step": 15760 }, { "grad_norm": 0.5079953670501709, "learning_rate": 9.419891541856323e-05, "loss": 0.0633, "step": 15770 }, { "grad_norm": 0.4482229948043823, "learning_rate": 9.41892486014624e-05, "loss": 0.0585, "step": 15780 }, { "grad_norm": 0.519662082195282, "learning_rate": 9.417957423362756e-05, "loss": 0.0714, "step": 15790 }, { "grad_norm": 0.4432123303413391, "learning_rate": 9.416989231671178e-05, "loss": 0.0655, "step": 15800 }, { "grad_norm": 0.3876844644546509, "learning_rate": 9.416020285236946e-05, "loss": 0.0614, "step": 15810 }, { "grad_norm": 0.5640828609466553, "learning_rate": 9.415050584225626e-05, "loss": 0.0605, "step": 15820 }, { "grad_norm": 0.48297518491744995, "learning_rate": 9.414080128802914e-05, "loss": 0.0547, "step": 15830 }, { "grad_norm": 0.518843948841095, "learning_rate": 9.413108919134632e-05, "loss": 0.0614, "step": 15840 }, { "grad_norm": 0.4817582368850708, "learning_rate": 9.412136955386734e-05, "loss": 0.0631, "step": 15850 }, { "grad_norm": 0.3891797363758087, "learning_rate": 9.411164237725303e-05, "loss": 0.0606, "step": 15860 }, { "grad_norm": 0.4747012257575989, "learning_rate": 9.41019076631655e-05, "loss": 0.0693, "step": 15870 }, { "grad_norm": 0.5432597994804382, "learning_rate": 9.409216541326815e-05, "loss": 0.0618, "step": 15880 }, { "grad_norm": 0.39103296399116516, "learning_rate": 9.408241562922564e-05, "loss": 0.0648, "step": 15890 }, { "grad_norm": 0.4791688323020935, "learning_rate": 9.407265831270395e-05, "loss": 0.0585, "step": 15900 }, { "grad_norm": 0.3963146209716797, "learning_rate": 9.406289346537035e-05, "loss": 0.0592, "step": 15910 }, { "grad_norm": 0.4559868276119232, "learning_rate": 9.405312108889339e-05, "loss": 0.0677, "step": 15920 }, { "grad_norm": 0.5707365870475769, "learning_rate": 9.404334118494288e-05, "loss": 0.0698, "step": 15930 }, { "grad_norm": 0.4699433147907257, "learning_rate": 9.403355375518995e-05, "loss": 0.0637, "step": 15940 }, { "grad_norm": 0.5572870969772339, "learning_rate": 9.4023758801307e-05, "loss": 0.0574, "step": 15950 }, { "grad_norm": 0.45805174112319946, "learning_rate": 9.401395632496774e-05, "loss": 0.0576, "step": 15960 }, { "grad_norm": 0.6077924370765686, "learning_rate": 9.400414632784711e-05, "loss": 0.0697, "step": 15970 }, { "grad_norm": 0.7214168906211853, "learning_rate": 9.39943288116214e-05, "loss": 0.0721, "step": 15980 }, { "grad_norm": 0.5017766356468201, "learning_rate": 9.398450377796815e-05, "loss": 0.0624, "step": 15990 }, { "grad_norm": 0.48519477248191833, "learning_rate": 9.397467122856616e-05, "loss": 0.0611, "step": 16000 }, { "grad_norm": 0.6013720631599426, "learning_rate": 9.396483116509558e-05, "loss": 0.0708, "step": 16010 }, { "grad_norm": 0.3977709710597992, "learning_rate": 9.39549835892378e-05, "loss": 0.0604, "step": 16020 }, { "grad_norm": 0.5590380430221558, "learning_rate": 9.39451285026755e-05, "loss": 0.0561, "step": 16030 }, { "grad_norm": 0.42309895157814026, "learning_rate": 9.393526590709262e-05, "loss": 0.0572, "step": 16040 }, { "grad_norm": 0.5346372723579407, "learning_rate": 9.392539580417444e-05, "loss": 0.0473, "step": 16050 }, { "grad_norm": 0.5024663805961609, "learning_rate": 9.391551819560747e-05, "loss": 0.0642, "step": 16060 }, { "grad_norm": 0.5493946075439453, "learning_rate": 9.390563308307955e-05, "loss": 0.0582, "step": 16070 }, { "grad_norm": 0.39252936840057373, "learning_rate": 9.389574046827974e-05, "loss": 0.0582, "step": 16080 }, { "grad_norm": 0.566171407699585, "learning_rate": 9.388584035289845e-05, "loss": 0.0599, "step": 16090 }, { "grad_norm": 0.45794281363487244, "learning_rate": 9.387593273862732e-05, "loss": 0.065, "step": 16100 }, { "grad_norm": 0.32747918367385864, "learning_rate": 9.386601762715929e-05, "loss": 0.065, "step": 16110 }, { "grad_norm": 0.6759500503540039, "learning_rate": 9.38560950201886e-05, "loss": 0.0755, "step": 16120 }, { "grad_norm": 0.4168781638145447, "learning_rate": 9.384616491941071e-05, "loss": 0.0626, "step": 16130 }, { "grad_norm": 0.4554072916507721, "learning_rate": 9.383622732652245e-05, "loss": 0.0487, "step": 16140 }, { "grad_norm": 0.43456578254699707, "learning_rate": 9.382628224322187e-05, "loss": 0.0564, "step": 16150 }, { "grad_norm": 0.4549276828765869, "learning_rate": 9.381632967120829e-05, "loss": 0.0625, "step": 16160 }, { "grad_norm": 0.6024060249328613, "learning_rate": 9.380636961218235e-05, "loss": 0.0648, "step": 16170 }, { "grad_norm": 0.3500462472438812, "learning_rate": 9.379640206784597e-05, "loss": 0.0443, "step": 16180 }, { "grad_norm": 0.5890312194824219, "learning_rate": 9.378642703990229e-05, "loss": 0.0584, "step": 16190 }, { "grad_norm": 0.44781753420829773, "learning_rate": 9.37764445300558e-05, "loss": 0.059, "step": 16200 }, { "grad_norm": 0.39172273874282837, "learning_rate": 9.376645454001222e-05, "loss": 0.0687, "step": 16210 }, { "grad_norm": 0.4562806785106659, "learning_rate": 9.375645707147858e-05, "loss": 0.052, "step": 16220 }, { "grad_norm": 0.6239166855812073, "learning_rate": 9.374645212616316e-05, "loss": 0.0808, "step": 16230 }, { "grad_norm": 0.5177644491195679, "learning_rate": 9.373643970577555e-05, "loss": 0.0673, "step": 16240 }, { "grad_norm": 0.5159828066825867, "learning_rate": 9.372641981202659e-05, "loss": 0.0673, "step": 16250 }, { "grad_norm": 0.5805414319038391, "learning_rate": 9.37163924466284e-05, "loss": 0.0645, "step": 16260 }, { "grad_norm": 0.4945962131023407, "learning_rate": 9.370635761129438e-05, "loss": 0.0554, "step": 16270 }, { "grad_norm": 0.5192605257034302, "learning_rate": 9.36963153077392e-05, "loss": 0.0646, "step": 16280 }, { "grad_norm": 0.33394330739974976, "learning_rate": 9.368626553767888e-05, "loss": 0.0556, "step": 16290 }, { "grad_norm": 0.44750338792800903, "learning_rate": 9.367620830283057e-05, "loss": 0.0495, "step": 16300 }, { "grad_norm": 0.5787836909294128, "learning_rate": 9.366614360491281e-05, "loss": 0.0724, "step": 16310 }, { "grad_norm": 0.4896298944950104, "learning_rate": 9.365607144564539e-05, "loss": 0.0564, "step": 16320 }, { "grad_norm": 0.47996848821640015, "learning_rate": 9.364599182674934e-05, "loss": 0.0711, "step": 16330 }, { "grad_norm": 0.30449819564819336, "learning_rate": 9.3635904749947e-05, "loss": 0.0548, "step": 16340 }, { "grad_norm": 0.47159332036972046, "learning_rate": 9.362581021696202e-05, "loss": 0.0557, "step": 16350 }, { "grad_norm": 0.4702777564525604, "learning_rate": 9.361570822951921e-05, "loss": 0.0594, "step": 16360 }, { "grad_norm": 0.4793546199798584, "learning_rate": 9.360559878934476e-05, "loss": 0.0653, "step": 16370 }, { "grad_norm": 0.5430279970169067, "learning_rate": 9.359548189816611e-05, "loss": 0.0597, "step": 16380 }, { "grad_norm": 0.4489835202693939, "learning_rate": 9.358535755771193e-05, "loss": 0.0648, "step": 16390 }, { "grad_norm": 0.4306167662143707, "learning_rate": 9.357522576971221e-05, "loss": 0.0569, "step": 16400 }, { "grad_norm": 0.5380247235298157, "learning_rate": 9.356508653589819e-05, "loss": 0.0632, "step": 16410 }, { "grad_norm": 0.6335141658782959, "learning_rate": 9.355493985800237e-05, "loss": 0.0639, "step": 16420 }, { "grad_norm": 0.5168572068214417, "learning_rate": 9.354478573775857e-05, "loss": 0.0643, "step": 16430 }, { "grad_norm": 0.41899338364601135, "learning_rate": 9.353462417690186e-05, "loss": 0.0581, "step": 16440 }, { "grad_norm": 0.47973722219467163, "learning_rate": 9.352445517716853e-05, "loss": 0.0547, "step": 16450 }, { "grad_norm": 0.4517166316509247, "learning_rate": 9.351427874029621e-05, "loss": 0.0646, "step": 16460 }, { "grad_norm": 2.5260653495788574, "learning_rate": 9.350409486802379e-05, "loss": 0.0614, "step": 16470 }, { "grad_norm": 0.6471713781356812, "learning_rate": 9.349390356209138e-05, "loss": 0.0751, "step": 16480 }, { "grad_norm": 0.4392150342464447, "learning_rate": 9.348370482424042e-05, "loss": 0.0723, "step": 16490 }, { "grad_norm": 0.5120335221290588, "learning_rate": 9.347349865621357e-05, "loss": 0.0643, "step": 16500 }, { "grad_norm": 0.4169086515903473, "learning_rate": 9.346328505975481e-05, "loss": 0.0611, "step": 16510 }, { "grad_norm": 0.5503948926925659, "learning_rate": 9.345306403660936e-05, "loss": 0.0702, "step": 16520 }, { "grad_norm": 0.5113946795463562, "learning_rate": 9.344283558852371e-05, "loss": 0.0666, "step": 16530 }, { "grad_norm": 0.4749281704425812, "learning_rate": 9.343259971724563e-05, "loss": 0.0593, "step": 16540 }, { "grad_norm": 0.45343276858329773, "learning_rate": 9.342235642452413e-05, "loss": 0.0605, "step": 16550 }, { "grad_norm": 0.4611659646034241, "learning_rate": 9.341210571210954e-05, "loss": 0.0638, "step": 16560 }, { "grad_norm": 0.6672820448875427, "learning_rate": 9.340184758175338e-05, "loss": 0.0622, "step": 16570 }, { "grad_norm": 0.43796566128730774, "learning_rate": 9.339158203520854e-05, "loss": 0.0695, "step": 16580 }, { "grad_norm": 0.34677615761756897, "learning_rate": 9.338130907422908e-05, "loss": 0.0543, "step": 16590 }, { "grad_norm": 0.40485644340515137, "learning_rate": 9.337102870057037e-05, "loss": 0.0548, "step": 16600 }, { "grad_norm": 0.527022659778595, "learning_rate": 9.336074091598907e-05, "loss": 0.0613, "step": 16610 }, { "grad_norm": 0.5201993584632874, "learning_rate": 9.335044572224306e-05, "loss": 0.0518, "step": 16620 }, { "grad_norm": 0.3616187870502472, "learning_rate": 9.334014312109151e-05, "loss": 0.0504, "step": 16630 }, { "grad_norm": 0.5474782586097717, "learning_rate": 9.332983311429486e-05, "loss": 0.0613, "step": 16640 }, { "grad_norm": 0.4680752754211426, "learning_rate": 9.33195157036148e-05, "loss": 0.071, "step": 16650 }, { "grad_norm": 0.35190659761428833, "learning_rate": 9.330919089081432e-05, "loss": 0.0584, "step": 16660 }, { "grad_norm": 0.3766917586326599, "learning_rate": 9.32988586776576e-05, "loss": 0.0834, "step": 16670 }, { "grad_norm": 0.4790816605091095, "learning_rate": 9.328851906591016e-05, "loss": 0.0635, "step": 16680 }, { "grad_norm": 0.4139423966407776, "learning_rate": 9.327817205733875e-05, "loss": 0.0607, "step": 16690 }, { "grad_norm": 0.37538447976112366, "learning_rate": 9.326781765371142e-05, "loss": 0.0565, "step": 16700 }, { "grad_norm": 0.5016010403633118, "learning_rate": 9.325745585679741e-05, "loss": 0.0709, "step": 16710 }, { "grad_norm": 0.5998192429542542, "learning_rate": 9.32470866683673e-05, "loss": 0.0866, "step": 16720 }, { "grad_norm": 0.5319188833236694, "learning_rate": 9.323671009019288e-05, "loss": 0.074, "step": 16730 }, { "grad_norm": 0.4386845529079437, "learning_rate": 9.322632612404725e-05, "loss": 0.0542, "step": 16740 }, { "grad_norm": 0.4699106216430664, "learning_rate": 9.321593477170471e-05, "loss": 0.0678, "step": 16750 }, { "grad_norm": 0.4634452760219574, "learning_rate": 9.320553603494088e-05, "loss": 0.0623, "step": 16760 }, { "grad_norm": 0.34045830368995667, "learning_rate": 9.319512991553261e-05, "loss": 0.0586, "step": 16770 }, { "grad_norm": 0.4425995945930481, "learning_rate": 9.318471641525803e-05, "loss": 0.0585, "step": 16780 }, { "grad_norm": 0.4619419276714325, "learning_rate": 9.317429553589652e-05, "loss": 0.0603, "step": 16790 }, { "grad_norm": 0.37375694513320923, "learning_rate": 9.316386727922873e-05, "loss": 0.0613, "step": 16800 }, { "grad_norm": 0.480654239654541, "learning_rate": 9.315343164703656e-05, "loss": 0.063, "step": 16810 }, { "grad_norm": 0.49134114384651184, "learning_rate": 9.314298864110316e-05, "loss": 0.0562, "step": 16820 }, { "grad_norm": 0.49232447147369385, "learning_rate": 9.313253826321295e-05, "loss": 0.0604, "step": 16830 }, { "grad_norm": 0.42667582631111145, "learning_rate": 9.312208051515165e-05, "loss": 0.0571, "step": 16840 }, { "grad_norm": 0.47124162316322327, "learning_rate": 9.311161539870618e-05, "loss": 0.0565, "step": 16850 }, { "grad_norm": 0.4664103090763092, "learning_rate": 9.310114291566474e-05, "loss": 0.0763, "step": 16860 }, { "grad_norm": 0.49231213331222534, "learning_rate": 9.309066306781679e-05, "loss": 0.0536, "step": 16870 }, { "grad_norm": 0.4833139181137085, "learning_rate": 9.308017585695306e-05, "loss": 0.0511, "step": 16880 }, { "grad_norm": 0.45079705119132996, "learning_rate": 9.306968128486552e-05, "loss": 0.053, "step": 16890 }, { "grad_norm": 0.4521995484828949, "learning_rate": 9.30591793533474e-05, "loss": 0.057, "step": 16900 }, { "grad_norm": 0.43332868814468384, "learning_rate": 9.304867006419321e-05, "loss": 0.0464, "step": 16910 }, { "grad_norm": 0.3855275511741638, "learning_rate": 9.303815341919868e-05, "loss": 0.0496, "step": 16920 }, { "grad_norm": 0.5961402654647827, "learning_rate": 9.302762942016084e-05, "loss": 0.056, "step": 16930 }, { "grad_norm": 0.42664310336112976, "learning_rate": 9.301709806887792e-05, "loss": 0.0495, "step": 16940 }, { "grad_norm": 0.5001668334007263, "learning_rate": 9.300655936714948e-05, "loss": 0.0735, "step": 16950 }, { "grad_norm": 0.4717599153518677, "learning_rate": 9.299601331677627e-05, "loss": 0.0646, "step": 16960 }, { "grad_norm": 0.6102184653282166, "learning_rate": 9.298545991956033e-05, "loss": 0.0722, "step": 16970 }, { "grad_norm": 0.4675225019454956, "learning_rate": 9.297489917730493e-05, "loss": 0.073, "step": 16980 }, { "grad_norm": 0.46892234683036804, "learning_rate": 9.296433109181464e-05, "loss": 0.066, "step": 16990 }, { "grad_norm": 0.568635106086731, "learning_rate": 9.295375566489523e-05, "loss": 0.0626, "step": 17000 }, { "grad_norm": 0.5475248694419861, "learning_rate": 9.294317289835379e-05, "loss": 0.0723, "step": 17010 }, { "grad_norm": 0.493978887796402, "learning_rate": 9.293258279399859e-05, "loss": 0.0791, "step": 17020 }, { "grad_norm": 0.754314661026001, "learning_rate": 9.292198535363919e-05, "loss": 0.0548, "step": 17030 }, { "grad_norm": 0.5110141634941101, "learning_rate": 9.291138057908641e-05, "loss": 0.0478, "step": 17040 }, { "grad_norm": 0.7755782008171082, "learning_rate": 9.290076847215234e-05, "loss": 0.0601, "step": 17050 }, { "grad_norm": 0.6641294956207275, "learning_rate": 9.289014903465025e-05, "loss": 0.0771, "step": 17060 }, { "grad_norm": 0.46855369210243225, "learning_rate": 9.287952226839475e-05, "loss": 0.0679, "step": 17070 }, { "grad_norm": 0.45120611786842346, "learning_rate": 9.286888817520164e-05, "loss": 0.0681, "step": 17080 }, { "grad_norm": 0.3525122404098511, "learning_rate": 9.285824675688803e-05, "loss": 0.0548, "step": 17090 }, { "grad_norm": 0.47571879625320435, "learning_rate": 9.28475980152722e-05, "loss": 0.0525, "step": 17100 }, { "grad_norm": 0.6898687481880188, "learning_rate": 9.283694195217379e-05, "loss": 0.0592, "step": 17110 }, { "grad_norm": 0.6624576449394226, "learning_rate": 9.282627856941356e-05, "loss": 0.0561, "step": 17120 }, { "grad_norm": 0.4787940979003906, "learning_rate": 9.281560786881363e-05, "loss": 0.0524, "step": 17130 }, { "grad_norm": 0.6806244850158691, "learning_rate": 9.280492985219733e-05, "loss": 0.0478, "step": 17140 }, { "grad_norm": 0.4918190538883209, "learning_rate": 9.279424452138924e-05, "loss": 0.0642, "step": 17150 }, { "grad_norm": 0.48647576570510864, "learning_rate": 9.278355187821517e-05, "loss": 0.0538, "step": 17160 }, { "grad_norm": 0.38167718052864075, "learning_rate": 9.277285192450224e-05, "loss": 0.0583, "step": 17170 }, { "grad_norm": 0.44126951694488525, "learning_rate": 9.276214466207875e-05, "loss": 0.0563, "step": 17180 }, { "grad_norm": 0.4167722761631012, "learning_rate": 9.275143009277427e-05, "loss": 0.0506, "step": 17190 }, { "grad_norm": 0.5259160399436951, "learning_rate": 9.274070821841964e-05, "loss": 0.0615, "step": 17200 }, { "grad_norm": 0.4057341516017914, "learning_rate": 9.272997904084696e-05, "loss": 0.058, "step": 17210 }, { "grad_norm": 0.4543001353740692, "learning_rate": 9.271924256188951e-05, "loss": 0.0592, "step": 17220 }, { "grad_norm": 0.5559465289115906, "learning_rate": 9.270849878338189e-05, "loss": 0.0649, "step": 17230 }, { "grad_norm": 0.5345778465270996, "learning_rate": 9.269774770715991e-05, "loss": 0.0679, "step": 17240 }, { "grad_norm": 0.4127146601676941, "learning_rate": 9.268698933506061e-05, "loss": 0.0801, "step": 17250 }, { "grad_norm": 0.40572234988212585, "learning_rate": 9.267622366892235e-05, "loss": 0.0528, "step": 17260 }, { "grad_norm": 0.5233277082443237, "learning_rate": 9.266545071058465e-05, "loss": 0.0534, "step": 17270 }, { "grad_norm": 0.38023456931114197, "learning_rate": 9.265467046188833e-05, "loss": 0.0506, "step": 17280 }, { "grad_norm": 0.42138534784317017, "learning_rate": 9.264388292467543e-05, "loss": 0.0609, "step": 17290 }, { "grad_norm": 0.48383596539497375, "learning_rate": 9.263308810078926e-05, "loss": 0.0505, "step": 17300 }, { "grad_norm": 0.31681832671165466, "learning_rate": 9.262228599207434e-05, "loss": 0.0531, "step": 17310 }, { "grad_norm": 0.39751219749450684, "learning_rate": 9.261147660037647e-05, "loss": 0.0463, "step": 17320 }, { "grad_norm": 0.45648327469825745, "learning_rate": 9.26006599275427e-05, "loss": 0.056, "step": 17330 }, { "grad_norm": 0.4842958450317383, "learning_rate": 9.258983597542124e-05, "loss": 0.053, "step": 17340 }, { "grad_norm": 0.5681790113449097, "learning_rate": 9.257900474586167e-05, "loss": 0.0517, "step": 17350 }, { "grad_norm": 0.6164266467094421, "learning_rate": 9.256816624071471e-05, "loss": 0.0679, "step": 17360 }, { "grad_norm": 0.6167702674865723, "learning_rate": 9.25573204618324e-05, "loss": 0.0613, "step": 17370 }, { "grad_norm": 0.5438864231109619, "learning_rate": 9.254646741106796e-05, "loss": 0.0632, "step": 17380 }, { "grad_norm": 0.5017785429954529, "learning_rate": 9.253560709027589e-05, "loss": 0.0594, "step": 17390 }, { "grad_norm": 0.6832423806190491, "learning_rate": 9.252473950131192e-05, "loss": 0.0655, "step": 17400 }, { "grad_norm": 0.4172319769859314, "learning_rate": 9.251386464603302e-05, "loss": 0.0475, "step": 17410 }, { "grad_norm": 0.6350406408309937, "learning_rate": 9.250298252629741e-05, "loss": 0.0699, "step": 17420 }, { "grad_norm": 0.5214536786079407, "learning_rate": 9.249209314396454e-05, "loss": 0.0594, "step": 17430 }, { "grad_norm": 0.45430788397789, "learning_rate": 9.248119650089513e-05, "loss": 0.0623, "step": 17440 }, { "grad_norm": 0.5118427872657776, "learning_rate": 9.247029259895108e-05, "loss": 0.0642, "step": 17450 }, { "grad_norm": 0.43721243739128113, "learning_rate": 9.24593814399956e-05, "loss": 0.0543, "step": 17460 }, { "grad_norm": 0.4777287244796753, "learning_rate": 9.244846302589309e-05, "loss": 0.051, "step": 17470 }, { "grad_norm": 0.4039797782897949, "learning_rate": 9.243753735850923e-05, "loss": 0.0603, "step": 17480 }, { "grad_norm": 0.3349793553352356, "learning_rate": 9.24266044397109e-05, "loss": 0.0542, "step": 17490 }, { "grad_norm": 0.339129775762558, "learning_rate": 9.241566427136624e-05, "loss": 0.0491, "step": 17500 }, { "grad_norm": 0.4675064980983734, "learning_rate": 9.240471685534463e-05, "loss": 0.0594, "step": 17510 }, { "grad_norm": 0.47944995760917664, "learning_rate": 9.239376219351667e-05, "loss": 0.0601, "step": 17520 }, { "grad_norm": 0.5144142508506775, "learning_rate": 9.238280028775425e-05, "loss": 0.0555, "step": 17530 }, { "grad_norm": 0.6617809534072876, "learning_rate": 9.237183113993041e-05, "loss": 0.0628, "step": 17540 }, { "grad_norm": 0.5482438206672668, "learning_rate": 9.236085475191952e-05, "loss": 0.0611, "step": 17550 }, { "grad_norm": 0.5492287278175354, "learning_rate": 9.234987112559709e-05, "loss": 0.0796, "step": 17560 }, { "grad_norm": 0.5411219000816345, "learning_rate": 9.233888026283999e-05, "loss": 0.0701, "step": 17570 }, { "grad_norm": 0.50343918800354, "learning_rate": 9.232788216552619e-05, "loss": 0.0638, "step": 17580 }, { "grad_norm": 0.504294753074646, "learning_rate": 9.231687683553502e-05, "loss": 0.0751, "step": 17590 }, { "grad_norm": 0.6040167808532715, "learning_rate": 9.230586427474698e-05, "loss": 0.0798, "step": 17600 }, { "grad_norm": 0.4067048132419586, "learning_rate": 9.229484448504379e-05, "loss": 0.0665, "step": 17610 }, { "grad_norm": 0.3408883213996887, "learning_rate": 9.228381746830843e-05, "loss": 0.0626, "step": 17620 }, { "grad_norm": 0.34936094284057617, "learning_rate": 9.227278322642514e-05, "loss": 0.0638, "step": 17630 }, { "grad_norm": 0.42182013392448425, "learning_rate": 9.226174176127937e-05, "loss": 0.0642, "step": 17640 }, { "grad_norm": 0.46956875920295715, "learning_rate": 9.22506930747578e-05, "loss": 0.056, "step": 17650 }, { "grad_norm": 0.62131267786026, "learning_rate": 9.223963716874831e-05, "loss": 0.067, "step": 17660 }, { "grad_norm": 0.628571629524231, "learning_rate": 9.222857404514012e-05, "loss": 0.0627, "step": 17670 }, { "grad_norm": 0.7347908616065979, "learning_rate": 9.221750370582355e-05, "loss": 0.0637, "step": 17680 }, { "grad_norm": 0.47607874870300293, "learning_rate": 9.220642615269028e-05, "loss": 0.0622, "step": 17690 }, { "grad_norm": 0.4554952383041382, "learning_rate": 9.219534138763311e-05, "loss": 0.0567, "step": 17700 }, { "grad_norm": 0.6007072925567627, "learning_rate": 9.218424941254613e-05, "loss": 0.0619, "step": 17710 }, { "grad_norm": 0.516014575958252, "learning_rate": 9.217315022932468e-05, "loss": 0.0623, "step": 17720 }, { "grad_norm": 0.5176360607147217, "learning_rate": 9.216204383986528e-05, "loss": 0.0529, "step": 17730 }, { "grad_norm": 0.4749462902545929, "learning_rate": 9.215093024606574e-05, "loss": 0.0687, "step": 17740 }, { "grad_norm": 0.5475049614906311, "learning_rate": 9.213980944982506e-05, "loss": 0.0749, "step": 17750 }, { "grad_norm": 0.5475859642028809, "learning_rate": 9.212868145304346e-05, "loss": 0.0679, "step": 17760 }, { "grad_norm": 0.45934587717056274, "learning_rate": 9.211754625762241e-05, "loss": 0.0487, "step": 17770 }, { "grad_norm": 0.5467910766601562, "learning_rate": 9.210640386546463e-05, "loss": 0.0712, "step": 17780 }, { "grad_norm": 0.4945966303348541, "learning_rate": 9.209525427847405e-05, "loss": 0.0491, "step": 17790 }, { "grad_norm": 0.42486312985420227, "learning_rate": 9.208409749855583e-05, "loss": 0.0517, "step": 17800 }, { "grad_norm": 0.48245301842689514, "learning_rate": 9.207293352761633e-05, "loss": 0.0648, "step": 17810 }, { "grad_norm": 0.5737563967704773, "learning_rate": 9.206176236756319e-05, "loss": 0.0649, "step": 17820 }, { "grad_norm": 0.5313713550567627, "learning_rate": 9.205058402030525e-05, "loss": 0.0563, "step": 17830 }, { "grad_norm": 0.5656278729438782, "learning_rate": 9.203939848775259e-05, "loss": 0.0569, "step": 17840 }, { "grad_norm": 0.648314893245697, "learning_rate": 9.202820577181652e-05, "loss": 0.0585, "step": 17850 }, { "grad_norm": 0.6543805599212646, "learning_rate": 9.201700587440953e-05, "loss": 0.0653, "step": 17860 }, { "grad_norm": 0.5698879957199097, "learning_rate": 9.200579879744544e-05, "loss": 0.073, "step": 17870 }, { "grad_norm": 0.5386166572570801, "learning_rate": 9.199458454283918e-05, "loss": 0.0699, "step": 17880 }, { "grad_norm": 0.47109454870224, "learning_rate": 9.198336311250697e-05, "loss": 0.0655, "step": 17890 }, { "grad_norm": 0.48978087306022644, "learning_rate": 9.197213450836626e-05, "loss": 0.0575, "step": 17900 }, { "grad_norm": 0.5634490847587585, "learning_rate": 9.19608987323357e-05, "loss": 0.064, "step": 17910 }, { "grad_norm": 0.5524494051933289, "learning_rate": 9.194965578633517e-05, "loss": 0.0603, "step": 17920 }, { "grad_norm": 0.42589735984802246, "learning_rate": 9.193840567228582e-05, "loss": 0.0544, "step": 17930 }, { "grad_norm": 0.43831005692481995, "learning_rate": 9.192714839210994e-05, "loss": 0.0579, "step": 17940 }, { "grad_norm": 0.5629490613937378, "learning_rate": 9.19158839477311e-05, "loss": 0.0705, "step": 17950 }, { "grad_norm": 0.5831650495529175, "learning_rate": 9.190461234107411e-05, "loss": 0.0574, "step": 17960 }, { "grad_norm": 0.40399590134620667, "learning_rate": 9.189333357406496e-05, "loss": 0.0441, "step": 17970 }, { "grad_norm": 0.4796231687068939, "learning_rate": 9.188204764863089e-05, "loss": 0.0632, "step": 17980 }, { "grad_norm": 0.5432906150817871, "learning_rate": 9.187075456670033e-05, "loss": 0.0531, "step": 17990 }, { "grad_norm": 0.49187418818473816, "learning_rate": 9.1859454330203e-05, "loss": 0.058, "step": 18000 }, { "grad_norm": 0.5052868127822876, "learning_rate": 9.18481469410698e-05, "loss": 0.0748, "step": 18010 }, { "grad_norm": 0.4672457277774811, "learning_rate": 9.183683240123281e-05, "loss": 0.0639, "step": 18020 }, { "grad_norm": 0.4809999465942383, "learning_rate": 9.182551071262541e-05, "loss": 0.0666, "step": 18030 }, { "grad_norm": 0.45834454894065857, "learning_rate": 9.181418187718218e-05, "loss": 0.0614, "step": 18040 }, { "grad_norm": 0.4950123429298401, "learning_rate": 9.180284589683888e-05, "loss": 0.0604, "step": 18050 }, { "grad_norm": 0.4096226096153259, "learning_rate": 9.17915027735325e-05, "loss": 0.0464, "step": 18060 }, { "grad_norm": 0.47023317217826843, "learning_rate": 9.178015250920133e-05, "loss": 0.067, "step": 18070 }, { "grad_norm": 0.3153039515018463, "learning_rate": 9.176879510578477e-05, "loss": 0.0477, "step": 18080 }, { "grad_norm": 0.2905244827270508, "learning_rate": 9.17574305652235e-05, "loss": 0.0494, "step": 18090 }, { "grad_norm": 0.38857465982437134, "learning_rate": 9.174605888945942e-05, "loss": 0.0638, "step": 18100 }, { "grad_norm": 0.6374789476394653, "learning_rate": 9.173468008043564e-05, "loss": 0.0597, "step": 18110 }, { "grad_norm": 0.6782222390174866, "learning_rate": 9.172329414009648e-05, "loss": 0.0639, "step": 18120 }, { "grad_norm": 0.7534418106079102, "learning_rate": 9.171190107038747e-05, "loss": 0.0652, "step": 18130 }, { "grad_norm": 0.4202195107936859, "learning_rate": 9.170050087325541e-05, "loss": 0.0612, "step": 18140 }, { "grad_norm": 0.5995569825172424, "learning_rate": 9.168909355064824e-05, "loss": 0.0596, "step": 18150 }, { "grad_norm": 0.5557471513748169, "learning_rate": 9.167767910451519e-05, "loss": 0.0608, "step": 18160 }, { "grad_norm": 0.5201064348220825, "learning_rate": 9.166625753680669e-05, "loss": 0.068, "step": 18170 }, { "grad_norm": 0.7855533361434937, "learning_rate": 9.165482884947431e-05, "loss": 0.0505, "step": 18180 }, { "grad_norm": 0.46077537536621094, "learning_rate": 9.164339304447098e-05, "loss": 0.0586, "step": 18190 }, { "grad_norm": 0.2890339195728302, "learning_rate": 9.163195012375072e-05, "loss": 0.0507, "step": 18200 }, { "grad_norm": 0.6043679118156433, "learning_rate": 9.16205000892688e-05, "loss": 0.062, "step": 18210 }, { "grad_norm": 0.4353998303413391, "learning_rate": 9.160904294298175e-05, "loss": 0.0471, "step": 18220 }, { "grad_norm": 0.5830026865005493, "learning_rate": 9.159757868684727e-05, "loss": 0.0568, "step": 18230 }, { "grad_norm": 0.45585957169532776, "learning_rate": 9.15861073228243e-05, "loss": 0.0623, "step": 18240 }, { "grad_norm": 0.44286587834358215, "learning_rate": 9.157462885287296e-05, "loss": 0.0547, "step": 18250 }, { "grad_norm": 0.5214270949363708, "learning_rate": 9.156314327895461e-05, "loss": 0.0485, "step": 18260 }, { "grad_norm": 0.5171087980270386, "learning_rate": 9.155165060303185e-05, "loss": 0.0632, "step": 18270 }, { "grad_norm": 0.5567808151245117, "learning_rate": 9.154015082706841e-05, "loss": 0.0495, "step": 18280 }, { "grad_norm": 0.5508670210838318, "learning_rate": 9.152864395302936e-05, "loss": 0.0604, "step": 18290 }, { "grad_norm": 0.525783896446228, "learning_rate": 9.151712998288085e-05, "loss": 0.0589, "step": 18300 }, { "grad_norm": 0.44447869062423706, "learning_rate": 9.150560891859031e-05, "loss": 0.0448, "step": 18310 }, { "grad_norm": 0.4345409870147705, "learning_rate": 9.14940807621264e-05, "loss": 0.0507, "step": 18320 }, { "grad_norm": 0.3588390648365021, "learning_rate": 9.148254551545894e-05, "loss": 0.0598, "step": 18330 }, { "grad_norm": 0.4938879907131195, "learning_rate": 9.147100318055901e-05, "loss": 0.0539, "step": 18340 }, { "grad_norm": 0.4875383973121643, "learning_rate": 9.145945375939888e-05, "loss": 0.0583, "step": 18350 }, { "grad_norm": 0.6766542196273804, "learning_rate": 9.144789725395203e-05, "loss": 0.066, "step": 18360 }, { "grad_norm": 0.5939799547195435, "learning_rate": 9.14363336661931e-05, "loss": 0.0585, "step": 18370 }, { "grad_norm": 0.5775653123855591, "learning_rate": 9.142476299809806e-05, "loss": 0.0601, "step": 18380 }, { "grad_norm": 0.5849500298500061, "learning_rate": 9.1413185251644e-05, "loss": 0.0667, "step": 18390 }, { "grad_norm": 0.6319483518600464, "learning_rate": 9.140160042880923e-05, "loss": 0.0779, "step": 18400 }, { "grad_norm": 0.5169095993041992, "learning_rate": 9.139000853157327e-05, "loss": 0.0747, "step": 18410 }, { "grad_norm": 0.43618229031562805, "learning_rate": 9.137840956191688e-05, "loss": 0.066, "step": 18420 }, { "grad_norm": 0.39687636494636536, "learning_rate": 9.136680352182199e-05, "loss": 0.0532, "step": 18430 }, { "grad_norm": 0.4806501269340515, "learning_rate": 9.135519041327177e-05, "loss": 0.0528, "step": 18440 }, { "grad_norm": 0.40351542830467224, "learning_rate": 9.134357023825058e-05, "loss": 0.063, "step": 18450 }, { "grad_norm": 0.5812011957168579, "learning_rate": 9.133194299874398e-05, "loss": 0.073, "step": 18460 }, { "grad_norm": 0.5662934184074402, "learning_rate": 9.132030869673876e-05, "loss": 0.0727, "step": 18470 }, { "grad_norm": 0.504705011844635, "learning_rate": 9.130866733422288e-05, "loss": 0.0608, "step": 18480 }, { "grad_norm": 0.5005030632019043, "learning_rate": 9.129701891318556e-05, "loss": 0.0502, "step": 18490 }, { "grad_norm": 0.44742631912231445, "learning_rate": 9.128536343561718e-05, "loss": 0.0656, "step": 18500 }, { "grad_norm": 0.47324907779693604, "learning_rate": 9.127370090350934e-05, "loss": 0.0538, "step": 18510 }, { "grad_norm": 0.4334179162979126, "learning_rate": 9.126203131885487e-05, "loss": 0.0536, "step": 18520 }, { "grad_norm": 0.5398903489112854, "learning_rate": 9.125035468364775e-05, "loss": 0.0583, "step": 18530 }, { "grad_norm": 0.4064897298812866, "learning_rate": 9.123867099988322e-05, "loss": 0.0569, "step": 18540 }, { "grad_norm": 0.40881678462028503, "learning_rate": 9.122698026955769e-05, "loss": 0.0526, "step": 18550 }, { "grad_norm": 0.3965896666049957, "learning_rate": 9.12152824946688e-05, "loss": 0.067, "step": 18560 }, { "grad_norm": 0.5065609216690063, "learning_rate": 9.120357767721538e-05, "loss": 0.0658, "step": 18570 }, { "grad_norm": 0.6343811750411987, "learning_rate": 9.119186581919745e-05, "loss": 0.0548, "step": 18580 }, { "grad_norm": 0.5957531929016113, "learning_rate": 9.118014692261624e-05, "loss": 0.0619, "step": 18590 }, { "grad_norm": 0.6743603348731995, "learning_rate": 9.116842098947422e-05, "loss": 0.0707, "step": 18600 }, { "grad_norm": 0.5265337228775024, "learning_rate": 9.115668802177499e-05, "loss": 0.0565, "step": 18610 }, { "grad_norm": 0.48426082730293274, "learning_rate": 9.114494802152342e-05, "loss": 0.0606, "step": 18620 }, { "grad_norm": 0.6896503567695618, "learning_rate": 9.113320099072555e-05, "loss": 0.0871, "step": 18630 }, { "grad_norm": 0.47568202018737793, "learning_rate": 9.112144693138864e-05, "loss": 0.0529, "step": 18640 }, { "grad_norm": 0.4487495720386505, "learning_rate": 9.110968584552111e-05, "loss": 0.0613, "step": 18650 }, { "grad_norm": 0.4444642663002014, "learning_rate": 9.109791773513264e-05, "loss": 0.057, "step": 18660 }, { "grad_norm": 0.5269257426261902, "learning_rate": 9.108614260223403e-05, "loss": 0.0516, "step": 18670 }, { "grad_norm": 0.43984511494636536, "learning_rate": 9.107436044883738e-05, "loss": 0.0537, "step": 18680 }, { "grad_norm": 0.4126061201095581, "learning_rate": 9.10625712769559e-05, "loss": 0.0463, "step": 18690 }, { "grad_norm": 0.4543047249317169, "learning_rate": 9.105077508860406e-05, "loss": 0.0466, "step": 18700 }, { "grad_norm": 0.4182523787021637, "learning_rate": 9.103897188579751e-05, "loss": 0.0503, "step": 18710 }, { "grad_norm": 0.3419549763202667, "learning_rate": 9.102716167055308e-05, "loss": 0.0374, "step": 18720 }, { "grad_norm": 0.5685030817985535, "learning_rate": 9.10153444448888e-05, "loss": 0.0902, "step": 18730 }, { "grad_norm": 0.3629709482192993, "learning_rate": 9.100352021082393e-05, "loss": 0.0447, "step": 18740 }, { "grad_norm": 0.5477064251899719, "learning_rate": 9.099168897037891e-05, "loss": 0.0506, "step": 18750 }, { "grad_norm": 0.4857929050922394, "learning_rate": 9.097985072557538e-05, "loss": 0.0607, "step": 18760 }, { "grad_norm": 0.44234731793403625, "learning_rate": 9.096800547843615e-05, "loss": 0.0657, "step": 18770 }, { "grad_norm": 0.5189287066459656, "learning_rate": 9.095615323098526e-05, "loss": 0.0794, "step": 18780 }, { "grad_norm": 0.472918838262558, "learning_rate": 9.094429398524795e-05, "loss": 0.0522, "step": 18790 }, { "grad_norm": 0.43172407150268555, "learning_rate": 9.093242774325061e-05, "loss": 0.0705, "step": 18800 }, { "grad_norm": 0.42534106969833374, "learning_rate": 9.092055450702088e-05, "loss": 0.0691, "step": 18810 }, { "grad_norm": 0.7142289876937866, "learning_rate": 9.090867427858756e-05, "loss": 0.0538, "step": 18820 }, { "grad_norm": 0.5565353631973267, "learning_rate": 9.089678705998066e-05, "loss": 0.0603, "step": 18830 }, { "grad_norm": 0.43675288558006287, "learning_rate": 9.088489285323139e-05, "loss": 0.0628, "step": 18840 }, { "grad_norm": 0.5818544030189514, "learning_rate": 9.087299166037212e-05, "loss": 0.0548, "step": 18850 }, { "grad_norm": 0.3711598515510559, "learning_rate": 9.086108348343647e-05, "loss": 0.0719, "step": 18860 }, { "grad_norm": 0.5386465787887573, "learning_rate": 9.08491683244592e-05, "loss": 0.0634, "step": 18870 }, { "grad_norm": 0.5412456393241882, "learning_rate": 9.08372461854763e-05, "loss": 0.0518, "step": 18880 }, { "grad_norm": 0.4550391733646393, "learning_rate": 9.082531706852492e-05, "loss": 0.0555, "step": 18890 }, { "grad_norm": 0.43972766399383545, "learning_rate": 9.081338097564342e-05, "loss": 0.071, "step": 18900 }, { "grad_norm": 0.5744100213050842, "learning_rate": 9.080143790887137e-05, "loss": 0.0561, "step": 18910 }, { "grad_norm": 0.5303049087524414, "learning_rate": 9.07894878702495e-05, "loss": 0.0557, "step": 18920 }, { "grad_norm": 0.39748692512512207, "learning_rate": 9.077753086181974e-05, "loss": 0.065, "step": 18930 }, { "grad_norm": 0.3316650986671448, "learning_rate": 9.076556688562524e-05, "loss": 0.0457, "step": 18940 }, { "grad_norm": 0.3932066857814789, "learning_rate": 9.075359594371029e-05, "loss": 0.051, "step": 18950 }, { "grad_norm": 0.8090145587921143, "learning_rate": 9.07416180381204e-05, "loss": 0.0588, "step": 18960 }, { "grad_norm": 0.41861510276794434, "learning_rate": 9.072963317090228e-05, "loss": 0.0532, "step": 18970 }, { "grad_norm": 0.44364169239997864, "learning_rate": 9.071764134410382e-05, "loss": 0.0509, "step": 18980 }, { "grad_norm": 0.5170662999153137, "learning_rate": 9.070564255977407e-05, "loss": 0.0537, "step": 18990 }, { "grad_norm": 0.41457799077033997, "learning_rate": 9.06936368199633e-05, "loss": 0.0586, "step": 19000 }, { "grad_norm": 0.37512409687042236, "learning_rate": 9.0681624126723e-05, "loss": 0.052, "step": 19010 }, { "grad_norm": 0.5980738401412964, "learning_rate": 9.066960448210576e-05, "loss": 0.0518, "step": 19020 }, { "grad_norm": 0.4443334937095642, "learning_rate": 9.065757788816543e-05, "loss": 0.0409, "step": 19030 }, { "grad_norm": 0.32433056831359863, "learning_rate": 9.064554434695705e-05, "loss": 0.0578, "step": 19040 }, { "grad_norm": 0.32925549149513245, "learning_rate": 9.063350386053677e-05, "loss": 0.0653, "step": 19050 }, { "grad_norm": 0.37901103496551514, "learning_rate": 9.062145643096202e-05, "loss": 0.0526, "step": 19060 }, { "grad_norm": 0.5345590114593506, "learning_rate": 9.060940206029136e-05, "loss": 0.0659, "step": 19070 }, { "grad_norm": 0.3265751600265503, "learning_rate": 9.059734075058457e-05, "loss": 0.0587, "step": 19080 }, { "grad_norm": 0.47640475630760193, "learning_rate": 9.058527250390257e-05, "loss": 0.0687, "step": 19090 }, { "grad_norm": 0.4685288071632385, "learning_rate": 9.057319732230752e-05, "loss": 0.0754, "step": 19100 }, { "grad_norm": 0.4665108323097229, "learning_rate": 9.056111520786273e-05, "loss": 0.0509, "step": 19110 }, { "grad_norm": 0.3972438871860504, "learning_rate": 9.054902616263268e-05, "loss": 0.048, "step": 19120 }, { "grad_norm": 0.3623020052909851, "learning_rate": 9.05369301886831e-05, "loss": 0.0697, "step": 19130 }, { "grad_norm": 0.5432302951812744, "learning_rate": 9.052482728808083e-05, "loss": 0.0605, "step": 19140 }, { "grad_norm": 0.6102005839347839, "learning_rate": 9.051271746289391e-05, "loss": 0.0421, "step": 19150 }, { "grad_norm": 0.5171313285827637, "learning_rate": 9.050060071519162e-05, "loss": 0.0623, "step": 19160 }, { "grad_norm": 0.44670766592025757, "learning_rate": 9.048847704704437e-05, "loss": 0.0519, "step": 19170 }, { "grad_norm": 0.3955223858356476, "learning_rate": 9.047634646052376e-05, "loss": 0.0501, "step": 19180 }, { "grad_norm": 0.49705076217651367, "learning_rate": 9.046420895770256e-05, "loss": 0.066, "step": 19190 }, { "grad_norm": 0.450741708278656, "learning_rate": 9.045206454065473e-05, "loss": 0.0536, "step": 19200 }, { "grad_norm": 0.44259607791900635, "learning_rate": 9.043991321145546e-05, "loss": 0.0493, "step": 19210 }, { "grad_norm": 0.4946995675563812, "learning_rate": 9.042775497218105e-05, "loss": 0.0426, "step": 19220 }, { "grad_norm": 0.47719988226890564, "learning_rate": 9.041558982490901e-05, "loss": 0.0616, "step": 19230 }, { "grad_norm": 0.4165644347667694, "learning_rate": 9.040341777171805e-05, "loss": 0.0565, "step": 19240 }, { "grad_norm": 0.39894649386405945, "learning_rate": 9.039123881468802e-05, "loss": 0.0572, "step": 19250 }, { "grad_norm": 0.3115175664424896, "learning_rate": 9.037905295589998e-05, "loss": 0.0487, "step": 19260 }, { "grad_norm": 0.3820194602012634, "learning_rate": 9.036686019743617e-05, "loss": 0.0447, "step": 19270 }, { "grad_norm": 0.4397749900817871, "learning_rate": 9.035466054137997e-05, "loss": 0.0541, "step": 19280 }, { "grad_norm": 0.5270267724990845, "learning_rate": 9.0342453989816e-05, "loss": 0.0589, "step": 19290 }, { "grad_norm": 0.5196409821510315, "learning_rate": 9.033024054483e-05, "loss": 0.0507, "step": 19300 }, { "grad_norm": 0.4816286861896515, "learning_rate": 9.031802020850894e-05, "loss": 0.0615, "step": 19310 }, { "grad_norm": 0.5437389612197876, "learning_rate": 9.030579298294092e-05, "loss": 0.0664, "step": 19320 }, { "grad_norm": 0.3841761350631714, "learning_rate": 9.029355887021524e-05, "loss": 0.0715, "step": 19330 }, { "grad_norm": 0.5355914831161499, "learning_rate": 9.028131787242238e-05, "loss": 0.0774, "step": 19340 }, { "grad_norm": 0.45925965905189514, "learning_rate": 9.026906999165399e-05, "loss": 0.0697, "step": 19350 }, { "grad_norm": 0.7212886214256287, "learning_rate": 9.025681523000291e-05, "loss": 0.0522, "step": 19360 }, { "grad_norm": 0.6658646464347839, "learning_rate": 9.024455358956315e-05, "loss": 0.052, "step": 19370 }, { "grad_norm": 0.6733608245849609, "learning_rate": 9.023228507242984e-05, "loss": 0.0734, "step": 19380 }, { "grad_norm": 0.4477936029434204, "learning_rate": 9.022000968069937e-05, "loss": 0.0649, "step": 19390 }, { "grad_norm": 0.32198452949523926, "learning_rate": 9.020772741646928e-05, "loss": 0.0547, "step": 19400 }, { "grad_norm": 0.49080732464790344, "learning_rate": 9.019543828183826e-05, "loss": 0.0557, "step": 19410 }, { "grad_norm": 0.4467203617095947, "learning_rate": 9.018314227890616e-05, "loss": 0.0472, "step": 19420 }, { "grad_norm": 0.4970865547657013, "learning_rate": 9.017083940977408e-05, "loss": 0.0617, "step": 19430 }, { "grad_norm": 0.3876662254333496, "learning_rate": 9.015852967654422e-05, "loss": 0.0494, "step": 19440 }, { "grad_norm": 0.3884558081626892, "learning_rate": 9.014621308131996e-05, "loss": 0.0568, "step": 19450 }, { "grad_norm": 0.5742124319076538, "learning_rate": 9.01338896262059e-05, "loss": 0.056, "step": 19460 }, { "grad_norm": 0.41038277745246887, "learning_rate": 9.012155931330777e-05, "loss": 0.0558, "step": 19470 }, { "grad_norm": 0.354241281747818, "learning_rate": 9.010922214473246e-05, "loss": 0.0487, "step": 19480 }, { "grad_norm": 0.3407929837703705, "learning_rate": 9.009687812258808e-05, "loss": 0.0689, "step": 19490 }, { "grad_norm": 0.5101284384727478, "learning_rate": 9.00845272489839e-05, "loss": 0.0542, "step": 19500 }, { "grad_norm": 0.5098211169242859, "learning_rate": 9.007216952603031e-05, "loss": 0.0713, "step": 19510 }, { "grad_norm": 0.3895675837993622, "learning_rate": 9.005980495583894e-05, "loss": 0.0506, "step": 19520 }, { "grad_norm": 0.5471907258033752, "learning_rate": 9.004743354052252e-05, "loss": 0.054, "step": 19530 }, { "grad_norm": 0.42291560769081116, "learning_rate": 9.003505528219503e-05, "loss": 0.0505, "step": 19540 }, { "grad_norm": 0.43541765213012695, "learning_rate": 9.002267018297154e-05, "loss": 0.0532, "step": 19550 }, { "grad_norm": 0.4310440719127655, "learning_rate": 9.001027824496834e-05, "loss": 0.0586, "step": 19560 }, { "grad_norm": 0.5019768476486206, "learning_rate": 8.999787947030287e-05, "loss": 0.0476, "step": 19570 }, { "grad_norm": 0.46474140882492065, "learning_rate": 8.998547386109376e-05, "loss": 0.0606, "step": 19580 }, { "grad_norm": 0.5117955803871155, "learning_rate": 8.997306141946073e-05, "loss": 0.0555, "step": 19590 }, { "grad_norm": 0.6753983497619629, "learning_rate": 8.996064214752481e-05, "loss": 0.0648, "step": 19600 }, { "grad_norm": 0.5175697207450867, "learning_rate": 8.994821604740806e-05, "loss": 0.0531, "step": 19610 }, { "grad_norm": 0.42182740569114685, "learning_rate": 8.993578312123377e-05, "loss": 0.0556, "step": 19620 }, { "grad_norm": 0.44502848386764526, "learning_rate": 8.992334337112639e-05, "loss": 0.0581, "step": 19630 }, { "grad_norm": 0.38120150566101074, "learning_rate": 8.991089679921154e-05, "loss": 0.0506, "step": 19640 }, { "grad_norm": 0.5038617253303528, "learning_rate": 8.989844340761599e-05, "loss": 0.0603, "step": 19650 }, { "grad_norm": 0.4949837625026703, "learning_rate": 8.988598319846768e-05, "loss": 0.0456, "step": 19660 }, { "grad_norm": 0.49368607997894287, "learning_rate": 8.987351617389574e-05, "loss": 0.0427, "step": 19670 }, { "grad_norm": 0.34541550278663635, "learning_rate": 8.98610423360304e-05, "loss": 0.0496, "step": 19680 }, { "grad_norm": 0.34406375885009766, "learning_rate": 8.984856168700317e-05, "loss": 0.0653, "step": 19690 }, { "grad_norm": 0.4657233953475952, "learning_rate": 8.983607422894658e-05, "loss": 0.0457, "step": 19700 }, { "grad_norm": 0.3874724507331848, "learning_rate": 8.982357996399442e-05, "loss": 0.0566, "step": 19710 }, { "grad_norm": 0.5173290967941284, "learning_rate": 8.981107889428164e-05, "loss": 0.0529, "step": 19720 }, { "grad_norm": 0.33754390478134155, "learning_rate": 8.979857102194428e-05, "loss": 0.0467, "step": 19730 }, { "grad_norm": 0.4815927743911743, "learning_rate": 8.978605634911968e-05, "loss": 0.0504, "step": 19740 }, { "grad_norm": 0.4062243700027466, "learning_rate": 8.977353487794616e-05, "loss": 0.043, "step": 19750 }, { "grad_norm": 0.8775585889816284, "learning_rate": 8.976100661056334e-05, "loss": 0.0662, "step": 19760 }, { "grad_norm": 0.4922148585319519, "learning_rate": 8.974847154911197e-05, "loss": 0.0479, "step": 19770 }, { "grad_norm": 0.5128970742225647, "learning_rate": 8.973592969573393e-05, "loss": 0.057, "step": 19780 }, { "grad_norm": 0.5156457424163818, "learning_rate": 8.972338105257228e-05, "loss": 0.064, "step": 19790 }, { "grad_norm": 0.3603215515613556, "learning_rate": 8.971082562177125e-05, "loss": 0.0493, "step": 19800 }, { "grad_norm": 0.509071946144104, "learning_rate": 8.96982634054762e-05, "loss": 0.0524, "step": 19810 }, { "grad_norm": 0.4571978747844696, "learning_rate": 8.96856944058337e-05, "loss": 0.0574, "step": 19820 }, { "grad_norm": 0.36111360788345337, "learning_rate": 8.967311862499144e-05, "loss": 0.0396, "step": 19830 }, { "grad_norm": 0.45356881618499756, "learning_rate": 8.966053606509825e-05, "loss": 0.0565, "step": 19840 }, { "grad_norm": 0.6300064325332642, "learning_rate": 8.964794672830417e-05, "loss": 0.0691, "step": 19850 }, { "grad_norm": 0.37247610092163086, "learning_rate": 8.963535061676038e-05, "loss": 0.0641, "step": 19860 }, { "grad_norm": 0.4141060709953308, "learning_rate": 8.962274773261918e-05, "loss": 0.0497, "step": 19870 }, { "grad_norm": 0.397163063287735, "learning_rate": 8.961013807803409e-05, "loss": 0.0621, "step": 19880 }, { "grad_norm": 0.3611961007118225, "learning_rate": 8.959752165515973e-05, "loss": 0.0626, "step": 19890 }, { "grad_norm": 1.0399025678634644, "learning_rate": 8.958489846615193e-05, "loss": 0.0731, "step": 19900 }, { "grad_norm": 0.5891239643096924, "learning_rate": 8.957226851316762e-05, "loss": 0.0529, "step": 19910 }, { "grad_norm": 0.38724803924560547, "learning_rate": 8.955963179836493e-05, "loss": 0.0572, "step": 19920 }, { "grad_norm": 0.3947535753250122, "learning_rate": 8.954698832390312e-05, "loss": 0.0575, "step": 19930 }, { "grad_norm": 0.5295503735542297, "learning_rate": 8.953433809194263e-05, "loss": 0.0577, "step": 19940 }, { "grad_norm": 0.534307062625885, "learning_rate": 8.9521681104645e-05, "loss": 0.0455, "step": 19950 }, { "grad_norm": 0.4366675317287445, "learning_rate": 8.9509017364173e-05, "loss": 0.0518, "step": 19960 }, { "grad_norm": 0.46128350496292114, "learning_rate": 8.949634687269052e-05, "loss": 0.0489, "step": 19970 }, { "grad_norm": 0.4270622134208679, "learning_rate": 8.948366963236259e-05, "loss": 0.0494, "step": 19980 }, { "grad_norm": 0.515483021736145, "learning_rate": 8.947098564535538e-05, "loss": 0.0497, "step": 19990 }, { "grad_norm": 0.44137969613075256, "learning_rate": 8.945829491383627e-05, "loss": 0.0544, "step": 20000 }, { "grad_norm": 0.39026105403900146, "learning_rate": 8.944559743997374e-05, "loss": 0.0517, "step": 20010 }, { "grad_norm": 0.36006975173950195, "learning_rate": 8.943289322593746e-05, "loss": 0.0449, "step": 20020 }, { "grad_norm": 0.38809922337532043, "learning_rate": 8.942018227389821e-05, "loss": 0.0427, "step": 20030 }, { "grad_norm": 0.3220726251602173, "learning_rate": 8.940746458602795e-05, "loss": 0.0619, "step": 20040 }, { "grad_norm": 0.5484541058540344, "learning_rate": 8.939474016449979e-05, "loss": 0.0606, "step": 20050 }, { "grad_norm": 0.38770559430122375, "learning_rate": 8.938200901148799e-05, "loss": 0.0487, "step": 20060 }, { "grad_norm": 0.5610180497169495, "learning_rate": 8.936927112916795e-05, "loss": 0.0542, "step": 20070 }, { "grad_norm": 0.452405720949173, "learning_rate": 8.935652651971622e-05, "loss": 0.0513, "step": 20080 }, { "grad_norm": 0.3402520716190338, "learning_rate": 8.934377518531052e-05, "loss": 0.0515, "step": 20090 }, { "grad_norm": 0.4618561863899231, "learning_rate": 8.933101712812967e-05, "loss": 0.0591, "step": 20100 }, { "grad_norm": 0.4677424430847168, "learning_rate": 8.931825235035374e-05, "loss": 0.0495, "step": 20110 }, { "grad_norm": 0.46529415249824524, "learning_rate": 8.930548085416382e-05, "loss": 0.0491, "step": 20120 }, { "grad_norm": 0.605416476726532, "learning_rate": 8.92927026417422e-05, "loss": 0.0485, "step": 20130 }, { "grad_norm": 0.44228923320770264, "learning_rate": 8.92799177152724e-05, "loss": 0.0436, "step": 20140 }, { "grad_norm": 0.39852818846702576, "learning_rate": 8.926712607693895e-05, "loss": 0.0445, "step": 20150 }, { "grad_norm": 0.5788927674293518, "learning_rate": 8.925432772892762e-05, "loss": 0.0555, "step": 20160 }, { "grad_norm": 0.5010403394699097, "learning_rate": 8.924152267342529e-05, "loss": 0.0618, "step": 20170 }, { "grad_norm": 0.5367171764373779, "learning_rate": 8.922871091261998e-05, "loss": 0.0573, "step": 20180 }, { "grad_norm": 0.3400565981864929, "learning_rate": 8.92158924487009e-05, "loss": 0.0451, "step": 20190 }, { "grad_norm": 0.4784688949584961, "learning_rate": 8.920306728385834e-05, "loss": 0.0474, "step": 20200 }, { "grad_norm": 0.4261954426765442, "learning_rate": 8.919023542028379e-05, "loss": 0.0461, "step": 20210 }, { "grad_norm": 0.32660630345344543, "learning_rate": 8.917739686016988e-05, "loss": 0.0496, "step": 20220 }, { "grad_norm": 0.6281941533088684, "learning_rate": 8.916455160571033e-05, "loss": 0.05, "step": 20230 }, { "grad_norm": 0.41156670451164246, "learning_rate": 8.915169965910008e-05, "loss": 0.0483, "step": 20240 }, { "grad_norm": 0.49838316440582275, "learning_rate": 8.913884102253514e-05, "loss": 0.0434, "step": 20250 }, { "grad_norm": 0.3182540535926819, "learning_rate": 8.912597569821273e-05, "loss": 0.0461, "step": 20260 }, { "grad_norm": 0.3555995523929596, "learning_rate": 8.911310368833118e-05, "loss": 0.0506, "step": 20270 }, { "grad_norm": 0.3389626443386078, "learning_rate": 8.910022499508994e-05, "loss": 0.0418, "step": 20280 }, { "grad_norm": 0.3541262745857239, "learning_rate": 8.908733962068965e-05, "loss": 0.0564, "step": 20290 }, { "grad_norm": 0.45788249373435974, "learning_rate": 8.907444756733207e-05, "loss": 0.0448, "step": 20300 }, { "grad_norm": 0.46382299065589905, "learning_rate": 8.906154883722006e-05, "loss": 0.0528, "step": 20310 }, { "grad_norm": 0.44846561551094055, "learning_rate": 8.904864343255773e-05, "loss": 0.0494, "step": 20320 }, { "grad_norm": 0.4514502286911011, "learning_rate": 8.90357313555502e-05, "loss": 0.0496, "step": 20330 }, { "grad_norm": 0.43376708030700684, "learning_rate": 8.90228126084038e-05, "loss": 0.0497, "step": 20340 }, { "grad_norm": 0.4526825547218323, "learning_rate": 8.900988719332601e-05, "loss": 0.059, "step": 20350 }, { "grad_norm": 0.27132073044776917, "learning_rate": 8.899695511252542e-05, "loss": 0.0478, "step": 20360 }, { "grad_norm": 0.48647794127464294, "learning_rate": 8.898401636821176e-05, "loss": 0.0663, "step": 20370 }, { "grad_norm": 0.3816002309322357, "learning_rate": 8.897107096259593e-05, "loss": 0.0709, "step": 20380 }, { "grad_norm": 0.36909422278404236, "learning_rate": 8.895811889788994e-05, "loss": 0.0541, "step": 20390 }, { "grad_norm": 0.470339298248291, "learning_rate": 8.894516017630692e-05, "loss": 0.0492, "step": 20400 }, { "grad_norm": 0.44825097918510437, "learning_rate": 8.893219480006118e-05, "loss": 0.0512, "step": 20410 }, { "grad_norm": 0.4775484800338745, "learning_rate": 8.891922277136817e-05, "loss": 0.0445, "step": 20420 }, { "grad_norm": 0.4285160005092621, "learning_rate": 8.890624409244441e-05, "loss": 0.0471, "step": 20430 }, { "grad_norm": 0.3874906301498413, "learning_rate": 8.889325876550763e-05, "loss": 0.0548, "step": 20440 }, { "grad_norm": 0.44157177209854126, "learning_rate": 8.888026679277666e-05, "loss": 0.0508, "step": 20450 }, { "grad_norm": 0.4691239893436432, "learning_rate": 8.886726817647147e-05, "loss": 0.0716, "step": 20460 }, { "grad_norm": 0.4029920995235443, "learning_rate": 8.885426291881319e-05, "loss": 0.0626, "step": 20470 }, { "grad_norm": 0.4757447838783264, "learning_rate": 8.884125102202401e-05, "loss": 0.0574, "step": 20480 }, { "grad_norm": 0.5424618721008301, "learning_rate": 8.882823248832736e-05, "loss": 0.0403, "step": 20490 }, { "grad_norm": 0.47620105743408203, "learning_rate": 8.881520731994772e-05, "loss": 0.0406, "step": 20500 }, { "grad_norm": 0.49721646308898926, "learning_rate": 8.880217551911077e-05, "loss": 0.0538, "step": 20510 }, { "grad_norm": 0.3602999448776245, "learning_rate": 8.878913708804323e-05, "loss": 0.0526, "step": 20520 }, { "grad_norm": 0.41017967462539673, "learning_rate": 8.877609202897308e-05, "loss": 0.0497, "step": 20530 }, { "grad_norm": 0.388289213180542, "learning_rate": 8.876304034412933e-05, "loss": 0.0409, "step": 20540 }, { "grad_norm": 0.38175931572914124, "learning_rate": 8.874998203574214e-05, "loss": 0.0486, "step": 20550 }, { "grad_norm": 0.4789811372756958, "learning_rate": 8.873691710604284e-05, "loss": 0.0488, "step": 20560 }, { "grad_norm": 0.5643757581710815, "learning_rate": 8.872384555726387e-05, "loss": 0.0561, "step": 20570 }, { "grad_norm": 0.44682133197784424, "learning_rate": 8.871076739163878e-05, "loss": 0.0465, "step": 20580 }, { "grad_norm": 0.40012675523757935, "learning_rate": 8.86976826114023e-05, "loss": 0.0469, "step": 20590 }, { "grad_norm": 0.5057584643363953, "learning_rate": 8.868459121879023e-05, "loss": 0.0542, "step": 20600 }, { "grad_norm": 0.41566193103790283, "learning_rate": 8.867149321603956e-05, "loss": 0.0602, "step": 20610 }, { "grad_norm": 0.45000872015953064, "learning_rate": 8.865838860538835e-05, "loss": 0.0507, "step": 20620 }, { "grad_norm": 0.5248885750770569, "learning_rate": 8.864527738907585e-05, "loss": 0.0596, "step": 20630 }, { "grad_norm": 0.4857695698738098, "learning_rate": 8.863215956934239e-05, "loss": 0.0519, "step": 20640 }, { "grad_norm": 0.4319910407066345, "learning_rate": 8.861903514842947e-05, "loss": 0.0495, "step": 20650 }, { "grad_norm": 0.5426684021949768, "learning_rate": 8.860590412857966e-05, "loss": 0.0417, "step": 20660 }, { "grad_norm": 0.5357169508934021, "learning_rate": 8.85927665120367e-05, "loss": 0.0432, "step": 20670 }, { "grad_norm": 0.5071842670440674, "learning_rate": 8.857962230104546e-05, "loss": 0.0529, "step": 20680 }, { "grad_norm": 0.598289430141449, "learning_rate": 8.856647149785193e-05, "loss": 0.0486, "step": 20690 }, { "grad_norm": 0.5182011127471924, "learning_rate": 8.855331410470322e-05, "loss": 0.0511, "step": 20700 }, { "grad_norm": 0.40677353739738464, "learning_rate": 8.854015012384756e-05, "loss": 0.0444, "step": 20710 }, { "grad_norm": 0.42823415994644165, "learning_rate": 8.852697955753433e-05, "loss": 0.0631, "step": 20720 }, { "grad_norm": 0.3602295517921448, "learning_rate": 8.851380240801399e-05, "loss": 0.0525, "step": 20730 }, { "grad_norm": 0.4241332709789276, "learning_rate": 8.850061867753818e-05, "loss": 0.0392, "step": 20740 }, { "grad_norm": 0.5359804034233093, "learning_rate": 8.848742836835963e-05, "loss": 0.0597, "step": 20750 }, { "grad_norm": 0.7888877391815186, "learning_rate": 8.847423148273221e-05, "loss": 0.0489, "step": 20760 }, { "grad_norm": 0.43647801876068115, "learning_rate": 8.846102802291092e-05, "loss": 0.0571, "step": 20770 }, { "grad_norm": 0.4391912817955017, "learning_rate": 8.844781799115183e-05, "loss": 0.0473, "step": 20780 }, { "grad_norm": 0.3857673108577728, "learning_rate": 8.84346013897122e-05, "loss": 0.0425, "step": 20790 }, { "grad_norm": 0.4359808564186096, "learning_rate": 8.842137822085038e-05, "loss": 0.0642, "step": 20800 }, { "grad_norm": 0.44838181138038635, "learning_rate": 8.840814848682585e-05, "loss": 0.0468, "step": 20810 }, { "grad_norm": 0.4208584427833557, "learning_rate": 8.83949121898992e-05, "loss": 0.056, "step": 20820 }, { "grad_norm": 0.5259532332420349, "learning_rate": 8.838166933233217e-05, "loss": 0.0479, "step": 20830 }, { "grad_norm": 0.2766227424144745, "learning_rate": 8.83684199163876e-05, "loss": 0.0495, "step": 20840 }, { "grad_norm": 0.5862826108932495, "learning_rate": 8.835516394432943e-05, "loss": 0.0547, "step": 20850 }, { "grad_norm": 0.5907891392707825, "learning_rate": 8.834190141842276e-05, "loss": 0.059, "step": 20860 }, { "grad_norm": 0.551679790019989, "learning_rate": 8.83286323409338e-05, "loss": 0.0475, "step": 20870 }, { "grad_norm": 0.5769739151000977, "learning_rate": 8.831535671412986e-05, "loss": 0.0645, "step": 20880 }, { "grad_norm": 0.5202513933181763, "learning_rate": 8.830207454027938e-05, "loss": 0.0561, "step": 20890 }, { "grad_norm": 0.5336427688598633, "learning_rate": 8.828878582165192e-05, "loss": 0.0549, "step": 20900 }, { "grad_norm": 0.4067496955394745, "learning_rate": 8.827549056051818e-05, "loss": 0.0561, "step": 20910 }, { "grad_norm": 0.4299774467945099, "learning_rate": 8.826218875914993e-05, "loss": 0.0461, "step": 20920 }, { "grad_norm": 0.39135149121284485, "learning_rate": 8.82488804198201e-05, "loss": 0.0474, "step": 20930 }, { "grad_norm": 0.32737430930137634, "learning_rate": 8.82355655448027e-05, "loss": 0.053, "step": 20940 }, { "grad_norm": 0.4949188232421875, "learning_rate": 8.822224413637293e-05, "loss": 0.0495, "step": 20950 }, { "grad_norm": 0.48346179723739624, "learning_rate": 8.820891619680697e-05, "loss": 0.041, "step": 20960 }, { "grad_norm": 0.4979545772075653, "learning_rate": 8.819558172838227e-05, "loss": 0.0506, "step": 20970 }, { "grad_norm": 0.3736691176891327, "learning_rate": 8.818224073337731e-05, "loss": 0.0415, "step": 20980 }, { "grad_norm": 0.4146130681037903, "learning_rate": 8.816889321407169e-05, "loss": 0.0496, "step": 20990 }, { "grad_norm": 0.5926865339279175, "learning_rate": 8.815553917274615e-05, "loss": 0.0622, "step": 21000 }, { "grad_norm": 0.4271896183490753, "learning_rate": 8.81421786116825e-05, "loss": 0.0571, "step": 21010 }, { "grad_norm": 0.43903353810310364, "learning_rate": 8.812881153316373e-05, "loss": 0.0442, "step": 21020 }, { "grad_norm": 0.424229234457016, "learning_rate": 8.81154379394739e-05, "loss": 0.055, "step": 21030 }, { "grad_norm": 0.5703893899917603, "learning_rate": 8.810205783289818e-05, "loss": 0.0578, "step": 21040 }, { "grad_norm": 0.5078112483024597, "learning_rate": 8.808867121572286e-05, "loss": 0.0559, "step": 21050 }, { "grad_norm": 0.42451605200767517, "learning_rate": 8.807527809023537e-05, "loss": 0.067, "step": 21060 }, { "grad_norm": 0.40063178539276123, "learning_rate": 8.80618784587242e-05, "loss": 0.0539, "step": 21070 }, { "grad_norm": 0.4258522391319275, "learning_rate": 8.804847232347902e-05, "loss": 0.057, "step": 21080 }, { "grad_norm": 0.3831099569797516, "learning_rate": 8.803505968679054e-05, "loss": 0.0634, "step": 21090 }, { "grad_norm": 0.5071573257446289, "learning_rate": 8.802164055095061e-05, "loss": 0.0599, "step": 21100 }, { "grad_norm": 0.4149579107761383, "learning_rate": 8.80082149182522e-05, "loss": 0.0549, "step": 21110 }, { "grad_norm": 0.4436695873737335, "learning_rate": 8.79947827909894e-05, "loss": 0.0571, "step": 21120 }, { "grad_norm": 0.5852754712104797, "learning_rate": 8.798134417145738e-05, "loss": 0.063, "step": 21130 }, { "grad_norm": 0.46663719415664673, "learning_rate": 8.796789906195243e-05, "loss": 0.0472, "step": 21140 }, { "grad_norm": 0.4049902856349945, "learning_rate": 8.795444746477195e-05, "loss": 0.0478, "step": 21150 }, { "grad_norm": 0.5231156349182129, "learning_rate": 8.794098938221446e-05, "loss": 0.0526, "step": 21160 }, { "grad_norm": 0.396064817905426, "learning_rate": 8.792752481657957e-05, "loss": 0.0528, "step": 21170 }, { "grad_norm": 0.46249985694885254, "learning_rate": 8.791405377016802e-05, "loss": 0.0485, "step": 21180 }, { "grad_norm": 0.5191993713378906, "learning_rate": 8.790057624528163e-05, "loss": 0.0528, "step": 21190 }, { "grad_norm": 0.5748463273048401, "learning_rate": 8.788709224422333e-05, "loss": 0.0518, "step": 21200 }, { "grad_norm": 0.5829198956489563, "learning_rate": 8.787360176929717e-05, "loss": 0.061, "step": 21210 }, { "grad_norm": 0.30733823776245117, "learning_rate": 8.786010482280834e-05, "loss": 0.0574, "step": 21220 }, { "grad_norm": 0.48887068033218384, "learning_rate": 8.784660140706306e-05, "loss": 0.0517, "step": 21230 }, { "grad_norm": 0.4663970470428467, "learning_rate": 8.783309152436872e-05, "loss": 0.0425, "step": 21240 }, { "grad_norm": 0.348314106464386, "learning_rate": 8.781957517703375e-05, "loss": 0.0436, "step": 21250 }, { "grad_norm": 0.44801145792007446, "learning_rate": 8.780605236736776e-05, "loss": 0.0557, "step": 21260 }, { "grad_norm": 0.40143969655036926, "learning_rate": 8.779252309768142e-05, "loss": 0.0485, "step": 21270 }, { "grad_norm": 0.4110606908798218, "learning_rate": 8.777898737028652e-05, "loss": 0.0529, "step": 21280 }, { "grad_norm": 0.42004892230033875, "learning_rate": 8.776544518749591e-05, "loss": 0.0575, "step": 21290 }, { "grad_norm": 0.4459332227706909, "learning_rate": 8.775189655162364e-05, "loss": 0.0502, "step": 21300 }, { "grad_norm": 0.3278050124645233, "learning_rate": 8.773834146498474e-05, "loss": 0.0477, "step": 21310 }, { "grad_norm": 0.31756463646888733, "learning_rate": 8.772477992989545e-05, "loss": 0.0411, "step": 21320 }, { "grad_norm": 0.35584452748298645, "learning_rate": 8.771121194867304e-05, "loss": 0.0547, "step": 21330 }, { "grad_norm": 0.44468697905540466, "learning_rate": 8.769763752363589e-05, "loss": 0.0433, "step": 21340 }, { "grad_norm": 0.42723193764686584, "learning_rate": 8.768405665710352e-05, "loss": 0.0516, "step": 21350 }, { "grad_norm": 0.49863573908805847, "learning_rate": 8.767046935139655e-05, "loss": 0.0527, "step": 21360 }, { "grad_norm": 0.7053219676017761, "learning_rate": 8.765687560883666e-05, "loss": 0.0497, "step": 21370 }, { "grad_norm": 0.4649471342563629, "learning_rate": 8.764327543174664e-05, "loss": 0.0492, "step": 21380 }, { "grad_norm": 0.6420255899429321, "learning_rate": 8.762966882245038e-05, "loss": 0.0487, "step": 21390 }, { "grad_norm": 0.5020867586135864, "learning_rate": 8.761605578327291e-05, "loss": 0.0511, "step": 21400 }, { "grad_norm": 0.47011804580688477, "learning_rate": 8.76024363165403e-05, "loss": 0.0565, "step": 21410 }, { "grad_norm": 0.4428064227104187, "learning_rate": 8.758881042457976e-05, "loss": 0.0486, "step": 21420 }, { "grad_norm": 0.40517449378967285, "learning_rate": 8.757517810971957e-05, "loss": 0.0447, "step": 21430 }, { "grad_norm": 0.2825189530849457, "learning_rate": 8.756153937428913e-05, "loss": 0.0417, "step": 21440 }, { "grad_norm": 0.49543726444244385, "learning_rate": 8.754789422061889e-05, "loss": 0.063, "step": 21450 }, { "grad_norm": 0.3365825414657593, "learning_rate": 8.753424265104052e-05, "loss": 0.0496, "step": 21460 }, { "grad_norm": 0.571965754032135, "learning_rate": 8.752058466788659e-05, "loss": 0.0657, "step": 21470 }, { "grad_norm": 0.3389694392681122, "learning_rate": 8.750692027349097e-05, "loss": 0.047, "step": 21480 }, { "grad_norm": 0.539064347743988, "learning_rate": 8.749324947018847e-05, "loss": 0.049, "step": 21490 }, { "grad_norm": 0.4241516590118408, "learning_rate": 8.747957226031507e-05, "loss": 0.0469, "step": 21500 }, { "grad_norm": 0.426035076379776, "learning_rate": 8.746588864620787e-05, "loss": 0.0502, "step": 21510 }, { "grad_norm": 0.33460095524787903, "learning_rate": 8.745219863020498e-05, "loss": 0.0364, "step": 21520 }, { "grad_norm": 0.48605218529701233, "learning_rate": 8.743850221464564e-05, "loss": 0.0526, "step": 21530 }, { "grad_norm": 1.3572310209274292, "learning_rate": 8.742479940187026e-05, "loss": 0.0463, "step": 21540 }, { "grad_norm": 0.4568389356136322, "learning_rate": 8.74110901942202e-05, "loss": 0.0486, "step": 21550 }, { "grad_norm": 0.5301503539085388, "learning_rate": 8.739737459403803e-05, "loss": 0.046, "step": 21560 }, { "grad_norm": 0.4475809931755066, "learning_rate": 8.738365260366737e-05, "loss": 0.0465, "step": 21570 }, { "grad_norm": 0.37409332394599915, "learning_rate": 8.736992422545292e-05, "loss": 0.0374, "step": 21580 }, { "grad_norm": 0.4051114618778229, "learning_rate": 8.73561894617405e-05, "loss": 0.0549, "step": 21590 }, { "grad_norm": 0.4351801872253418, "learning_rate": 8.734244831487697e-05, "loss": 0.0487, "step": 21600 }, { "grad_norm": 0.4241715371608734, "learning_rate": 8.732870078721035e-05, "loss": 0.0562, "step": 21610 }, { "grad_norm": 0.34067460894584656, "learning_rate": 8.731494688108972e-05, "loss": 0.0515, "step": 21620 }, { "grad_norm": 0.5309017300605774, "learning_rate": 8.730118659886523e-05, "loss": 0.0514, "step": 21630 }, { "grad_norm": 0.44083672761917114, "learning_rate": 8.728741994288814e-05, "loss": 0.049, "step": 21640 }, { "grad_norm": 0.43010812997817993, "learning_rate": 8.727364691551079e-05, "loss": 0.0387, "step": 21650 }, { "grad_norm": 0.34887662529945374, "learning_rate": 8.725986751908661e-05, "loss": 0.0471, "step": 21660 }, { "grad_norm": 0.491019070148468, "learning_rate": 8.724608175597016e-05, "loss": 0.0512, "step": 21670 }, { "grad_norm": 0.43225711584091187, "learning_rate": 8.723228962851699e-05, "loss": 0.048, "step": 21680 }, { "grad_norm": 0.3576586842536926, "learning_rate": 8.721849113908385e-05, "loss": 0.0521, "step": 21690 }, { "grad_norm": 0.5150206685066223, "learning_rate": 8.720468629002848e-05, "loss": 0.0533, "step": 21700 }, { "grad_norm": 0.44534701108932495, "learning_rate": 8.719087508370978e-05, "loss": 0.0521, "step": 21710 }, { "grad_norm": 0.4893801510334015, "learning_rate": 8.717705752248772e-05, "loss": 0.0656, "step": 21720 }, { "grad_norm": 0.4499804377555847, "learning_rate": 8.71632336087233e-05, "loss": 0.0437, "step": 21730 }, { "grad_norm": 0.35747480392456055, "learning_rate": 8.71494033447787e-05, "loss": 0.0456, "step": 21740 }, { "grad_norm": 0.413719117641449, "learning_rate": 8.713556673301708e-05, "loss": 0.0471, "step": 21750 }, { "grad_norm": 0.35208556056022644, "learning_rate": 8.712172377580278e-05, "loss": 0.0627, "step": 21760 }, { "grad_norm": 0.42669522762298584, "learning_rate": 8.710787447550114e-05, "loss": 0.0528, "step": 21770 }, { "grad_norm": 0.330323189496994, "learning_rate": 8.70940188344787e-05, "loss": 0.0614, "step": 21780 }, { "grad_norm": 0.5014104843139648, "learning_rate": 8.708015685510293e-05, "loss": 0.041, "step": 21790 }, { "grad_norm": 0.4463307857513428, "learning_rate": 8.706628853974252e-05, "loss": 0.0583, "step": 21800 }, { "grad_norm": 0.4056156873703003, "learning_rate": 8.705241389076715e-05, "loss": 0.0409, "step": 21810 }, { "grad_norm": 0.4161947965621948, "learning_rate": 8.703853291054764e-05, "loss": 0.0442, "step": 21820 }, { "grad_norm": 0.582490086555481, "learning_rate": 8.702464560145587e-05, "loss": 0.0596, "step": 21830 }, { "grad_norm": 0.3105570673942566, "learning_rate": 8.701075196586476e-05, "loss": 0.0436, "step": 21840 }, { "grad_norm": 0.43746259808540344, "learning_rate": 8.699685200614842e-05, "loss": 0.0432, "step": 21850 }, { "grad_norm": 0.35583731532096863, "learning_rate": 8.698294572468193e-05, "loss": 0.0609, "step": 21860 }, { "grad_norm": 0.4591095745563507, "learning_rate": 8.696903312384148e-05, "loss": 0.0473, "step": 21870 }, { "grad_norm": 0.5040755867958069, "learning_rate": 8.695511420600439e-05, "loss": 0.0405, "step": 21880 }, { "grad_norm": 0.5670108795166016, "learning_rate": 8.694118897354901e-05, "loss": 0.0508, "step": 21890 }, { "grad_norm": 2.994516611099243, "learning_rate": 8.692725742885478e-05, "loss": 0.056, "step": 21900 }, { "grad_norm": 0.6622418761253357, "learning_rate": 8.691331957430221e-05, "loss": 0.0639, "step": 21910 }, { "grad_norm": 0.605545699596405, "learning_rate": 8.68993754122729e-05, "loss": 0.0579, "step": 21920 }, { "grad_norm": 0.46545350551605225, "learning_rate": 8.688542494514955e-05, "loss": 0.0534, "step": 21930 }, { "grad_norm": 0.6081550717353821, "learning_rate": 8.68714681753159e-05, "loss": 0.0608, "step": 21940 }, { "grad_norm": 0.4233928918838501, "learning_rate": 8.685750510515676e-05, "loss": 0.0438, "step": 21950 }, { "grad_norm": 0.6955615282058716, "learning_rate": 8.684353573705805e-05, "loss": 0.0567, "step": 21960 }, { "grad_norm": 0.43856847286224365, "learning_rate": 8.682956007340677e-05, "loss": 0.0548, "step": 21970 }, { "grad_norm": 0.5028722882270813, "learning_rate": 8.681557811659095e-05, "loss": 0.0497, "step": 21980 }, { "grad_norm": 0.45552143454551697, "learning_rate": 8.680158986899974e-05, "loss": 0.0463, "step": 21990 }, { "grad_norm": 0.39489954710006714, "learning_rate": 8.678759533302335e-05, "loss": 0.0459, "step": 22000 }, { "grad_norm": 0.48942580819129944, "learning_rate": 8.677359451105308e-05, "loss": 0.0446, "step": 22010 }, { "grad_norm": 0.5118188261985779, "learning_rate": 8.675958740548123e-05, "loss": 0.0609, "step": 22020 }, { "grad_norm": 0.5530210137367249, "learning_rate": 8.674557401870129e-05, "loss": 0.063, "step": 22030 }, { "grad_norm": 0.4567714035511017, "learning_rate": 8.673155435310775e-05, "loss": 0.0535, "step": 22040 }, { "grad_norm": 0.3954084813594818, "learning_rate": 8.671752841109617e-05, "loss": 0.0456, "step": 22050 }, { "grad_norm": 0.4116666316986084, "learning_rate": 8.670349619506321e-05, "loss": 0.0505, "step": 22060 }, { "grad_norm": 0.46054646372795105, "learning_rate": 8.66894577074066e-05, "loss": 0.0477, "step": 22070 }, { "grad_norm": 0.4588325619697571, "learning_rate": 8.667541295052513e-05, "loss": 0.044, "step": 22080 }, { "grad_norm": 0.4753906726837158, "learning_rate": 8.666136192681865e-05, "loss": 0.063, "step": 22090 }, { "grad_norm": 0.38339975476264954, "learning_rate": 8.664730463868811e-05, "loss": 0.0484, "step": 22100 }, { "grad_norm": 0.5974498987197876, "learning_rate": 8.663324108853552e-05, "loss": 0.0605, "step": 22110 }, { "grad_norm": 0.42582517862319946, "learning_rate": 8.661917127876395e-05, "loss": 0.0697, "step": 22120 }, { "grad_norm": 0.47493547201156616, "learning_rate": 8.660509521177754e-05, "loss": 0.0543, "step": 22130 }, { "grad_norm": 0.3874149024486542, "learning_rate": 8.65910128899815e-05, "loss": 0.048, "step": 22140 }, { "grad_norm": 0.501567006111145, "learning_rate": 8.657692431578214e-05, "loss": 0.0511, "step": 22150 }, { "grad_norm": 0.35595953464508057, "learning_rate": 8.656282949158679e-05, "loss": 0.052, "step": 22160 }, { "grad_norm": 0.35510116815567017, "learning_rate": 8.654872841980388e-05, "loss": 0.0398, "step": 22170 }, { "grad_norm": 0.4452972114086151, "learning_rate": 8.653462110284289e-05, "loss": 0.0491, "step": 22180 }, { "grad_norm": 0.4100826680660248, "learning_rate": 8.652050754311437e-05, "loss": 0.0563, "step": 22190 }, { "grad_norm": 0.41589415073394775, "learning_rate": 8.650638774302995e-05, "loss": 0.0582, "step": 22200 }, { "grad_norm": 0.4756990969181061, "learning_rate": 8.649226170500233e-05, "loss": 0.0511, "step": 22210 }, { "grad_norm": 0.43057453632354736, "learning_rate": 8.647812943144524e-05, "loss": 0.051, "step": 22220 }, { "grad_norm": 0.45281752943992615, "learning_rate": 8.646399092477351e-05, "loss": 0.0458, "step": 22230 }, { "grad_norm": 0.5379845499992371, "learning_rate": 8.644984618740301e-05, "loss": 0.0527, "step": 22240 }, { "grad_norm": 0.4347709119319916, "learning_rate": 8.643569522175073e-05, "loss": 0.0409, "step": 22250 }, { "grad_norm": 0.47421005368232727, "learning_rate": 8.642153803023463e-05, "loss": 0.0539, "step": 22260 }, { "grad_norm": 0.3322734534740448, "learning_rate": 8.640737461527383e-05, "loss": 0.0469, "step": 22270 }, { "grad_norm": 0.35321369767189026, "learning_rate": 8.639320497928845e-05, "loss": 0.0512, "step": 22280 }, { "grad_norm": 0.5008605718612671, "learning_rate": 8.637902912469969e-05, "loss": 0.0543, "step": 22290 }, { "grad_norm": 0.3985813856124878, "learning_rate": 8.636484705392982e-05, "loss": 0.0505, "step": 22300 }, { "grad_norm": 0.3045957088470459, "learning_rate": 8.635065876940216e-05, "loss": 0.0428, "step": 22310 }, { "grad_norm": 0.35886675119400024, "learning_rate": 8.633646427354112e-05, "loss": 0.0594, "step": 22320 }, { "grad_norm": 0.392053484916687, "learning_rate": 8.632226356877213e-05, "loss": 0.0439, "step": 22330 }, { "grad_norm": 0.2648507058620453, "learning_rate": 8.630805665752173e-05, "loss": 0.0539, "step": 22340 }, { "grad_norm": 0.39792945981025696, "learning_rate": 8.629384354221748e-05, "loss": 0.0479, "step": 22350 }, { "grad_norm": 0.4875844120979309, "learning_rate": 8.627962422528797e-05, "loss": 0.0434, "step": 22360 }, { "grad_norm": 0.6298582553863525, "learning_rate": 8.626539870916296e-05, "loss": 0.0651, "step": 22370 }, { "grad_norm": 0.3512748181819916, "learning_rate": 8.625116699627317e-05, "loss": 0.0473, "step": 22380 }, { "grad_norm": 0.37791964411735535, "learning_rate": 8.623692908905041e-05, "loss": 0.0397, "step": 22390 }, { "grad_norm": 0.5757318139076233, "learning_rate": 8.622268498992755e-05, "loss": 0.0603, "step": 22400 }, { "grad_norm": 0.3281555771827698, "learning_rate": 8.620843470133851e-05, "loss": 0.0596, "step": 22410 }, { "grad_norm": 0.5406818985939026, "learning_rate": 8.619417822571829e-05, "loss": 0.0494, "step": 22420 }, { "grad_norm": 0.489723265171051, "learning_rate": 8.617991556550292e-05, "loss": 0.0464, "step": 22430 }, { "grad_norm": 0.36802881956100464, "learning_rate": 8.616564672312952e-05, "loss": 0.0461, "step": 22440 }, { "grad_norm": 0.33596882224082947, "learning_rate": 8.61513717010362e-05, "loss": 0.0532, "step": 22450 }, { "grad_norm": 0.5109977722167969, "learning_rate": 8.613709050166221e-05, "loss": 0.0531, "step": 22460 }, { "grad_norm": 0.3962125778198242, "learning_rate": 8.61228031274478e-05, "loss": 0.0426, "step": 22470 }, { "grad_norm": 0.5039965510368347, "learning_rate": 8.610850958083431e-05, "loss": 0.059, "step": 22480 }, { "grad_norm": 0.3958951532840729, "learning_rate": 8.609420986426409e-05, "loss": 0.0466, "step": 22490 }, { "grad_norm": 0.4691601097583771, "learning_rate": 8.60799039801806e-05, "loss": 0.0449, "step": 22500 }, { "grad_norm": 0.3933141231536865, "learning_rate": 8.606559193102828e-05, "loss": 0.0453, "step": 22510 }, { "grad_norm": 0.43596747517585754, "learning_rate": 8.605127371925273e-05, "loss": 0.0448, "step": 22520 }, { "grad_norm": 0.5462401509284973, "learning_rate": 8.603694934730047e-05, "loss": 0.0585, "step": 22530 }, { "grad_norm": 0.3303402364253998, "learning_rate": 8.602261881761919e-05, "loss": 0.0402, "step": 22540 }, { "grad_norm": 0.402269184589386, "learning_rate": 8.600828213265759e-05, "loss": 0.0563, "step": 22550 }, { "grad_norm": 0.5505662560462952, "learning_rate": 8.599393929486539e-05, "loss": 0.0612, "step": 22560 }, { "grad_norm": 0.5308811068534851, "learning_rate": 8.59795903066934e-05, "loss": 0.042, "step": 22570 }, { "grad_norm": 0.324098140001297, "learning_rate": 8.596523517059347e-05, "loss": 0.0368, "step": 22580 }, { "grad_norm": 0.5125874876976013, "learning_rate": 8.59508738890185e-05, "loss": 0.0482, "step": 22590 }, { "grad_norm": 0.3213435411453247, "learning_rate": 8.593650646442246e-05, "loss": 0.043, "step": 22600 }, { "grad_norm": 0.5047370791435242, "learning_rate": 8.59221328992603e-05, "loss": 0.0418, "step": 22610 }, { "grad_norm": 0.4034327566623688, "learning_rate": 8.590775319598813e-05, "loss": 0.0532, "step": 22620 }, { "grad_norm": 0.3704429864883423, "learning_rate": 8.589336735706301e-05, "loss": 0.0473, "step": 22630 }, { "grad_norm": 0.5030616521835327, "learning_rate": 8.587897538494307e-05, "loss": 0.0654, "step": 22640 }, { "grad_norm": 0.5412161350250244, "learning_rate": 8.586457728208756e-05, "loss": 0.0479, "step": 22650 }, { "grad_norm": 0.48785701394081116, "learning_rate": 8.585017305095667e-05, "loss": 0.0487, "step": 22660 }, { "grad_norm": 0.37969890236854553, "learning_rate": 8.583576269401173e-05, "loss": 0.0487, "step": 22670 }, { "grad_norm": 0.4053172469139099, "learning_rate": 8.582134621371504e-05, "loss": 0.051, "step": 22680 }, { "grad_norm": 0.6398476362228394, "learning_rate": 8.580692361253e-05, "loss": 0.0519, "step": 22690 }, { "grad_norm": 0.49880605936050415, "learning_rate": 8.579249489292104e-05, "loss": 0.0461, "step": 22700 }, { "grad_norm": 0.4377114474773407, "learning_rate": 8.577806005735363e-05, "loss": 0.0541, "step": 22710 }, { "grad_norm": 0.4778149127960205, "learning_rate": 8.576361910829429e-05, "loss": 0.0441, "step": 22720 }, { "grad_norm": 0.5418125987052917, "learning_rate": 8.574917204821057e-05, "loss": 0.051, "step": 22730 }, { "grad_norm": 0.4083711504936218, "learning_rate": 8.57347188795711e-05, "loss": 0.0471, "step": 22740 }, { "grad_norm": 0.5700498819351196, "learning_rate": 8.572025960484551e-05, "loss": 0.0609, "step": 22750 }, { "grad_norm": 0.43246594071388245, "learning_rate": 8.57057942265045e-05, "loss": 0.0555, "step": 22760 }, { "grad_norm": 0.5156602263450623, "learning_rate": 8.569132274701984e-05, "loss": 0.0472, "step": 22770 }, { "grad_norm": 0.3462342321872711, "learning_rate": 8.567684516886427e-05, "loss": 0.045, "step": 22780 }, { "grad_norm": 0.40801700949668884, "learning_rate": 8.56623614945116e-05, "loss": 0.057, "step": 22790 }, { "grad_norm": 0.46404391527175903, "learning_rate": 8.564787172643675e-05, "loss": 0.0482, "step": 22800 }, { "grad_norm": 0.6036810874938965, "learning_rate": 8.563337586711559e-05, "loss": 0.0556, "step": 22810 }, { "grad_norm": 0.318336546421051, "learning_rate": 8.561887391902506e-05, "loss": 0.0628, "step": 22820 }, { "grad_norm": 0.5613479614257812, "learning_rate": 8.560436588464316e-05, "loss": 0.042, "step": 22830 }, { "grad_norm": 0.323117733001709, "learning_rate": 8.55898517664489e-05, "loss": 0.0466, "step": 22840 }, { "grad_norm": 0.5414700508117676, "learning_rate": 8.557533156692236e-05, "loss": 0.0611, "step": 22850 }, { "grad_norm": 0.2475014179944992, "learning_rate": 8.556080528854467e-05, "loss": 0.0407, "step": 22860 }, { "grad_norm": 0.4807136058807373, "learning_rate": 8.554627293379791e-05, "loss": 0.0464, "step": 22870 }, { "grad_norm": 0.5871314406394958, "learning_rate": 8.553173450516531e-05, "loss": 0.0522, "step": 22880 }, { "grad_norm": 0.4529235064983368, "learning_rate": 8.551719000513108e-05, "loss": 0.0437, "step": 22890 }, { "grad_norm": 0.3638545274734497, "learning_rate": 8.550263943618049e-05, "loss": 0.0638, "step": 22900 }, { "grad_norm": 0.35599684715270996, "learning_rate": 8.54880828007998e-05, "loss": 0.0562, "step": 22910 }, { "grad_norm": 0.4881564974784851, "learning_rate": 8.547352010147637e-05, "loss": 0.068, "step": 22920 }, { "grad_norm": 0.523106038570404, "learning_rate": 8.545895134069855e-05, "loss": 0.0507, "step": 22930 }, { "grad_norm": 0.36673855781555176, "learning_rate": 8.544437652095576e-05, "loss": 0.0442, "step": 22940 }, { "grad_norm": 0.8085877895355225, "learning_rate": 8.542979564473843e-05, "loss": 0.0649, "step": 22950 }, { "grad_norm": 0.35676997900009155, "learning_rate": 8.541520871453802e-05, "loss": 0.051, "step": 22960 }, { "grad_norm": 0.4252561926841736, "learning_rate": 8.540061573284705e-05, "loss": 0.0468, "step": 22970 }, { "grad_norm": 1.099100947380066, "learning_rate": 8.538601670215906e-05, "loss": 0.0467, "step": 22980 }, { "grad_norm": 0.5984564423561096, "learning_rate": 8.537141162496864e-05, "loss": 0.0561, "step": 22990 }, { "grad_norm": 0.47643744945526123, "learning_rate": 8.535680050377137e-05, "loss": 0.052, "step": 23000 }, { "grad_norm": 0.4069572389125824, "learning_rate": 8.534218334106391e-05, "loss": 0.0517, "step": 23010 }, { "grad_norm": 0.3468594253063202, "learning_rate": 8.532756013934393e-05, "loss": 0.0524, "step": 23020 }, { "grad_norm": 0.4831400215625763, "learning_rate": 8.531293090111012e-05, "loss": 0.0539, "step": 23030 }, { "grad_norm": 0.39108917117118835, "learning_rate": 8.529829562886225e-05, "loss": 0.0516, "step": 23040 }, { "grad_norm": 0.6687740087509155, "learning_rate": 8.528365432510105e-05, "loss": 0.0607, "step": 23050 }, { "grad_norm": 0.4431750774383545, "learning_rate": 8.526900699232833e-05, "loss": 0.0485, "step": 23060 }, { "grad_norm": 0.42656025290489197, "learning_rate": 8.525435363304695e-05, "loss": 0.0425, "step": 23070 }, { "grad_norm": 0.3610136806964874, "learning_rate": 8.523969424976072e-05, "loss": 0.0387, "step": 23080 }, { "grad_norm": 0.4783131778240204, "learning_rate": 8.522502884497457e-05, "loss": 0.0562, "step": 23090 }, { "grad_norm": 0.4186384379863739, "learning_rate": 8.521035742119437e-05, "loss": 0.0478, "step": 23100 }, { "grad_norm": 0.37053948640823364, "learning_rate": 8.519567998092712e-05, "loss": 0.0602, "step": 23110 }, { "grad_norm": 0.3725856840610504, "learning_rate": 8.518099652668075e-05, "loss": 0.0609, "step": 23120 }, { "grad_norm": 0.48931729793548584, "learning_rate": 8.516630706096429e-05, "loss": 0.0488, "step": 23130 }, { "grad_norm": 0.5132979154586792, "learning_rate": 8.515161158628773e-05, "loss": 0.0443, "step": 23140 }, { "grad_norm": 0.3688201606273651, "learning_rate": 8.513691010516216e-05, "loss": 0.0583, "step": 23150 }, { "grad_norm": 0.3798530101776123, "learning_rate": 8.512220262009966e-05, "loss": 0.0459, "step": 23160 }, { "grad_norm": 0.3740488886833191, "learning_rate": 8.510748913361332e-05, "loss": 0.052, "step": 23170 }, { "grad_norm": 0.34930089116096497, "learning_rate": 8.509276964821726e-05, "loss": 0.0424, "step": 23180 }, { "grad_norm": 0.43100622296333313, "learning_rate": 8.507804416642669e-05, "loss": 0.0529, "step": 23190 }, { "grad_norm": 0.40214356780052185, "learning_rate": 8.506331269075774e-05, "loss": 0.0496, "step": 23200 }, { "grad_norm": 0.382472962141037, "learning_rate": 8.504857522372765e-05, "loss": 0.052, "step": 23210 }, { "grad_norm": 0.41511648893356323, "learning_rate": 8.503383176785461e-05, "loss": 0.0465, "step": 23220 }, { "grad_norm": 0.6137897968292236, "learning_rate": 8.501908232565792e-05, "loss": 0.052, "step": 23230 }, { "grad_norm": 0.362815797328949, "learning_rate": 8.50043268996578e-05, "loss": 0.0534, "step": 23240 }, { "grad_norm": 0.38176071643829346, "learning_rate": 8.498956549237562e-05, "loss": 0.054, "step": 23250 }, { "grad_norm": 0.3112301826477051, "learning_rate": 8.497479810633366e-05, "loss": 0.0376, "step": 23260 }, { "grad_norm": 0.4531206488609314, "learning_rate": 8.496002474405525e-05, "loss": 0.0436, "step": 23270 }, { "grad_norm": 0.34634530544281006, "learning_rate": 8.494524540806478e-05, "loss": 0.0472, "step": 23280 }, { "grad_norm": 0.3106862008571625, "learning_rate": 8.493046010088761e-05, "loss": 0.0582, "step": 23290 }, { "grad_norm": 0.4038488566875458, "learning_rate": 8.491566882505018e-05, "loss": 0.0472, "step": 23300 }, { "grad_norm": 0.41617685556411743, "learning_rate": 8.490087158307988e-05, "loss": 0.0532, "step": 23310 }, { "grad_norm": 0.6685087084770203, "learning_rate": 8.488606837750518e-05, "loss": 0.0516, "step": 23320 }, { "grad_norm": 0.4782409369945526, "learning_rate": 8.487125921085552e-05, "loss": 0.0462, "step": 23330 }, { "grad_norm": 0.34531012177467346, "learning_rate": 8.485644408566141e-05, "loss": 0.047, "step": 23340 }, { "grad_norm": 0.3915881812572479, "learning_rate": 8.484162300445431e-05, "loss": 0.0451, "step": 23350 }, { "grad_norm": 0.40550026297569275, "learning_rate": 8.482679596976676e-05, "loss": 0.0522, "step": 23360 }, { "grad_norm": 0.4235950708389282, "learning_rate": 8.48119629841323e-05, "loss": 0.0529, "step": 23370 }, { "grad_norm": 0.3881789445877075, "learning_rate": 8.479712405008547e-05, "loss": 0.0579, "step": 23380 }, { "grad_norm": 0.3772604763507843, "learning_rate": 8.478227917016184e-05, "loss": 0.0564, "step": 23390 }, { "grad_norm": 0.28406834602355957, "learning_rate": 8.476742834689801e-05, "loss": 0.0485, "step": 23400 }, { "grad_norm": 0.3527803421020508, "learning_rate": 8.475257158283157e-05, "loss": 0.0522, "step": 23410 }, { "grad_norm": 0.4654757082462311, "learning_rate": 8.473770888050112e-05, "loss": 0.0391, "step": 23420 }, { "grad_norm": 0.4557593762874603, "learning_rate": 8.47228402424463e-05, "loss": 0.0426, "step": 23430 }, { "grad_norm": 0.4799598753452301, "learning_rate": 8.470796567120775e-05, "loss": 0.0537, "step": 23440 }, { "grad_norm": 0.4615231454372406, "learning_rate": 8.469308516932714e-05, "loss": 0.0522, "step": 23450 }, { "grad_norm": 0.34500402212142944, "learning_rate": 8.467819873934714e-05, "loss": 0.041, "step": 23460 }, { "grad_norm": 0.4953031539916992, "learning_rate": 8.466330638381143e-05, "loss": 0.0488, "step": 23470 }, { "grad_norm": 0.5763334631919861, "learning_rate": 8.464840810526469e-05, "loss": 0.048, "step": 23480 }, { "grad_norm": 0.35955217480659485, "learning_rate": 8.463350390625264e-05, "loss": 0.0454, "step": 23490 }, { "grad_norm": 0.34185510873794556, "learning_rate": 8.4618593789322e-05, "loss": 0.0683, "step": 23500 }, { "grad_norm": 0.2810869514942169, "learning_rate": 8.46036777570205e-05, "loss": 0.0489, "step": 23510 }, { "grad_norm": 0.5101418495178223, "learning_rate": 8.458875581189688e-05, "loss": 0.0728, "step": 23520 }, { "grad_norm": 0.3633774518966675, "learning_rate": 8.457382795650092e-05, "loss": 0.0524, "step": 23530 }, { "grad_norm": 0.40289682149887085, "learning_rate": 8.455889419338335e-05, "loss": 0.0573, "step": 23540 }, { "grad_norm": 0.34815892577171326, "learning_rate": 8.454395452509593e-05, "loss": 0.0526, "step": 23550 }, { "grad_norm": 0.45836859941482544, "learning_rate": 8.452900895419146e-05, "loss": 0.0479, "step": 23560 }, { "grad_norm": 0.8581242561340332, "learning_rate": 8.451405748322376e-05, "loss": 0.0483, "step": 23570 }, { "grad_norm": 0.37323710322380066, "learning_rate": 8.449910011474759e-05, "loss": 0.0503, "step": 23580 }, { "grad_norm": 0.3981158435344696, "learning_rate": 8.448413685131876e-05, "loss": 0.0506, "step": 23590 }, { "grad_norm": 0.54217529296875, "learning_rate": 8.446916769549407e-05, "loss": 0.0573, "step": 23600 }, { "grad_norm": 0.29206618666648865, "learning_rate": 8.445419264983136e-05, "loss": 0.0526, "step": 23610 }, { "grad_norm": 0.4246256351470947, "learning_rate": 8.443921171688947e-05, "loss": 0.0456, "step": 23620 }, { "grad_norm": 0.30671417713165283, "learning_rate": 8.442422489922819e-05, "loss": 0.0386, "step": 23630 }, { "grad_norm": 0.4665154218673706, "learning_rate": 8.440923219940838e-05, "loss": 0.0582, "step": 23640 }, { "grad_norm": 0.4126928150653839, "learning_rate": 8.439423361999189e-05, "loss": 0.0481, "step": 23650 }, { "grad_norm": 0.5136278867721558, "learning_rate": 8.437922916354155e-05, "loss": 0.0428, "step": 23660 }, { "grad_norm": 0.37525734305381775, "learning_rate": 8.436421883262123e-05, "loss": 0.0409, "step": 23670 }, { "grad_norm": 0.2921887934207916, "learning_rate": 8.434920262979577e-05, "loss": 0.0498, "step": 23680 }, { "grad_norm": 0.32408472895622253, "learning_rate": 8.433418055763104e-05, "loss": 0.043, "step": 23690 }, { "grad_norm": 0.4303460717201233, "learning_rate": 8.431915261869389e-05, "loss": 0.0534, "step": 23700 }, { "grad_norm": 0.4719163179397583, "learning_rate": 8.43041188155522e-05, "loss": 0.0395, "step": 23710 }, { "grad_norm": 0.48658373951911926, "learning_rate": 8.428907915077481e-05, "loss": 0.0529, "step": 23720 }, { "grad_norm": 0.4469674825668335, "learning_rate": 8.42740336269316e-05, "loss": 0.0541, "step": 23730 }, { "grad_norm": 0.7235364317893982, "learning_rate": 8.425898224659345e-05, "loss": 0.0437, "step": 23740 }, { "grad_norm": 0.33482903242111206, "learning_rate": 8.42439250123322e-05, "loss": 0.048, "step": 23750 }, { "grad_norm": 0.41178786754608154, "learning_rate": 8.422886192672076e-05, "loss": 0.0594, "step": 23760 }, { "grad_norm": 0.4865293800830841, "learning_rate": 8.421379299233297e-05, "loss": 0.0525, "step": 23770 }, { "grad_norm": 0.458116739988327, "learning_rate": 8.419871821174371e-05, "loss": 0.0534, "step": 23780 }, { "grad_norm": 0.3849198520183563, "learning_rate": 8.418363758752884e-05, "loss": 0.0418, "step": 23790 }, { "grad_norm": 0.3294745683670044, "learning_rate": 8.416855112226523e-05, "loss": 0.0434, "step": 23800 }, { "grad_norm": 0.35620197653770447, "learning_rate": 8.415345881853075e-05, "loss": 0.0466, "step": 23810 }, { "grad_norm": 0.48055699467658997, "learning_rate": 8.413836067890426e-05, "loss": 0.0415, "step": 23820 }, { "grad_norm": 0.445039838552475, "learning_rate": 8.41232567059656e-05, "loss": 0.0421, "step": 23830 }, { "grad_norm": 0.4945099353790283, "learning_rate": 8.410814690229565e-05, "loss": 0.047, "step": 23840 }, { "grad_norm": 0.290902316570282, "learning_rate": 8.409303127047626e-05, "loss": 0.0419, "step": 23850 }, { "grad_norm": 0.4061161279678345, "learning_rate": 8.407790981309028e-05, "loss": 0.0406, "step": 23860 }, { "grad_norm": 0.35482463240623474, "learning_rate": 8.406278253272153e-05, "loss": 0.0511, "step": 23870 }, { "grad_norm": 0.2745746076107025, "learning_rate": 8.404764943195487e-05, "loss": 0.0368, "step": 23880 }, { "grad_norm": 0.6149085760116577, "learning_rate": 8.403251051337613e-05, "loss": 0.0435, "step": 23890 }, { "grad_norm": 0.319644957780838, "learning_rate": 8.401736577957214e-05, "loss": 0.0408, "step": 23900 }, { "grad_norm": 0.3799138069152832, "learning_rate": 8.40022152331307e-05, "loss": 0.0474, "step": 23910 }, { "grad_norm": 0.4946913719177246, "learning_rate": 8.398705887664064e-05, "loss": 0.057, "step": 23920 }, { "grad_norm": 0.527604877948761, "learning_rate": 8.397189671269177e-05, "loss": 0.0406, "step": 23930 }, { "grad_norm": 0.35631000995635986, "learning_rate": 8.395672874387488e-05, "loss": 0.0421, "step": 23940 }, { "grad_norm": 0.44251665472984314, "learning_rate": 8.394155497278177e-05, "loss": 0.0424, "step": 23950 }, { "grad_norm": 0.6148049235343933, "learning_rate": 8.392637540200523e-05, "loss": 0.0645, "step": 23960 }, { "grad_norm": 0.39325186610221863, "learning_rate": 8.391119003413902e-05, "loss": 0.0476, "step": 23970 }, { "grad_norm": 0.3542497456073761, "learning_rate": 8.38959988717779e-05, "loss": 0.0434, "step": 23980 }, { "grad_norm": 0.3820861279964447, "learning_rate": 8.388080191751764e-05, "loss": 0.0597, "step": 23990 }, { "grad_norm": 0.47771507501602173, "learning_rate": 8.386559917395496e-05, "loss": 0.0487, "step": 24000 }, { "grad_norm": 0.3473193049430847, "learning_rate": 8.385039064368761e-05, "loss": 0.0403, "step": 24010 }, { "grad_norm": 0.46650436520576477, "learning_rate": 8.383517632931431e-05, "loss": 0.0425, "step": 24020 }, { "grad_norm": 0.40216320753097534, "learning_rate": 8.381995623343477e-05, "loss": 0.047, "step": 24030 }, { "grad_norm": 0.4067218005657196, "learning_rate": 8.380473035864968e-05, "loss": 0.04, "step": 24040 }, { "grad_norm": 0.3386552929878235, "learning_rate": 8.378949870756076e-05, "loss": 0.0445, "step": 24050 }, { "grad_norm": 0.3564370274543762, "learning_rate": 8.377426128277063e-05, "loss": 0.0517, "step": 24060 }, { "grad_norm": 0.3454565703868866, "learning_rate": 8.375901808688298e-05, "loss": 0.0447, "step": 24070 }, { "grad_norm": 0.5459020137786865, "learning_rate": 8.374376912250246e-05, "loss": 0.0494, "step": 24080 }, { "grad_norm": 0.4104040861129761, "learning_rate": 8.372851439223468e-05, "loss": 0.0354, "step": 24090 }, { "grad_norm": 0.333077996969223, "learning_rate": 8.371325389868627e-05, "loss": 0.0486, "step": 24100 }, { "grad_norm": 0.5328429341316223, "learning_rate": 8.369798764446482e-05, "loss": 0.0473, "step": 24110 }, { "grad_norm": 0.44869500398635864, "learning_rate": 8.368271563217893e-05, "loss": 0.0449, "step": 24120 }, { "grad_norm": 0.36349955201148987, "learning_rate": 8.366743786443817e-05, "loss": 0.0597, "step": 24130 }, { "grad_norm": 0.38908183574676514, "learning_rate": 8.365215434385309e-05, "loss": 0.0535, "step": 24140 }, { "grad_norm": 0.40752142667770386, "learning_rate": 8.36368650730352e-05, "loss": 0.0473, "step": 24150 }, { "grad_norm": 0.5249229073524475, "learning_rate": 8.362157005459705e-05, "loss": 0.0441, "step": 24160 }, { "grad_norm": 0.3923241198062897, "learning_rate": 8.360626929115213e-05, "loss": 0.0403, "step": 24170 }, { "grad_norm": 0.43253082036972046, "learning_rate": 8.359096278531492e-05, "loss": 0.044, "step": 24180 }, { "grad_norm": 0.4422493278980255, "learning_rate": 8.357565053970088e-05, "loss": 0.0422, "step": 24190 }, { "grad_norm": 0.3346785604953766, "learning_rate": 8.356033255692647e-05, "loss": 0.0433, "step": 24200 }, { "grad_norm": 1.0375055074691772, "learning_rate": 8.354500883960911e-05, "loss": 0.0472, "step": 24210 }, { "grad_norm": 0.5038827657699585, "learning_rate": 8.352967939036717e-05, "loss": 0.0454, "step": 24220 }, { "grad_norm": 0.4012564718723297, "learning_rate": 8.35143442118201e-05, "loss": 0.0554, "step": 24230 }, { "grad_norm": 0.3814355134963989, "learning_rate": 8.349900330658819e-05, "loss": 0.0468, "step": 24240 }, { "grad_norm": 0.7142460346221924, "learning_rate": 8.348365667729284e-05, "loss": 0.0607, "step": 24250 }, { "grad_norm": 0.5560631155967712, "learning_rate": 8.346830432655633e-05, "loss": 0.0548, "step": 24260 }, { "grad_norm": 0.47137489914894104, "learning_rate": 8.345294625700195e-05, "loss": 0.0543, "step": 24270 }, { "grad_norm": 0.3614949584007263, "learning_rate": 8.343758247125402e-05, "loss": 0.0492, "step": 24280 }, { "grad_norm": 0.40714550018310547, "learning_rate": 8.342221297193776e-05, "loss": 0.0598, "step": 24290 }, { "grad_norm": 0.39931249618530273, "learning_rate": 8.34068377616794e-05, "loss": 0.0552, "step": 24300 }, { "grad_norm": 0.5930362939834595, "learning_rate": 8.339145684310615e-05, "loss": 0.0498, "step": 24310 }, { "grad_norm": 0.46895530819892883, "learning_rate": 8.337607021884618e-05, "loss": 0.047, "step": 24320 }, { "grad_norm": 0.5696265697479248, "learning_rate": 8.336067789152867e-05, "loss": 0.0383, "step": 24330 }, { "grad_norm": 0.4125528037548065, "learning_rate": 8.334527986378369e-05, "loss": 0.0504, "step": 24340 }, { "grad_norm": 0.6866901516914368, "learning_rate": 8.332987613824239e-05, "loss": 0.0546, "step": 24350 }, { "grad_norm": 0.624443531036377, "learning_rate": 8.331446671753685e-05, "loss": 0.0423, "step": 24360 }, { "grad_norm": 0.4513109028339386, "learning_rate": 8.329905160430007e-05, "loss": 0.0441, "step": 24370 }, { "grad_norm": 0.44951826333999634, "learning_rate": 8.328363080116611e-05, "loss": 0.0423, "step": 24380 }, { "grad_norm": 0.6827129125595093, "learning_rate": 8.326820431076997e-05, "loss": 0.0391, "step": 24390 }, { "grad_norm": 0.39998653531074524, "learning_rate": 8.325277213574759e-05, "loss": 0.0529, "step": 24400 }, { "grad_norm": 0.27457115054130554, "learning_rate": 8.32373342787359e-05, "loss": 0.0372, "step": 24410 }, { "grad_norm": 0.3917638063430786, "learning_rate": 8.322189074237285e-05, "loss": 0.0473, "step": 24420 }, { "grad_norm": 0.46711087226867676, "learning_rate": 8.32064415292973e-05, "loss": 0.039, "step": 24430 }, { "grad_norm": 0.3324368894100189, "learning_rate": 8.319098664214907e-05, "loss": 0.0375, "step": 24440 }, { "grad_norm": 0.4320794641971588, "learning_rate": 8.3175526083569e-05, "loss": 0.048, "step": 24450 }, { "grad_norm": 0.4225679337978363, "learning_rate": 8.316005985619889e-05, "loss": 0.0391, "step": 24460 }, { "grad_norm": 0.4447201192378998, "learning_rate": 8.314458796268147e-05, "loss": 0.0499, "step": 24470 }, { "grad_norm": 0.4218451678752899, "learning_rate": 8.312911040566047e-05, "loss": 0.0424, "step": 24480 }, { "grad_norm": 0.3983577489852905, "learning_rate": 8.31136271877806e-05, "loss": 0.0332, "step": 24490 }, { "grad_norm": 0.35630372166633606, "learning_rate": 8.309813831168748e-05, "loss": 0.0531, "step": 24500 }, { "grad_norm": 0.45867645740509033, "learning_rate": 8.308264378002777e-05, "loss": 0.0322, "step": 24510 }, { "grad_norm": 0.44922521710395813, "learning_rate": 8.306714359544906e-05, "loss": 0.0523, "step": 24520 }, { "grad_norm": 0.3143346905708313, "learning_rate": 8.30516377605999e-05, "loss": 0.0383, "step": 24530 }, { "grad_norm": 0.3928501009941101, "learning_rate": 8.30361262781298e-05, "loss": 0.0387, "step": 24540 }, { "grad_norm": 0.42401552200317383, "learning_rate": 8.302060915068924e-05, "loss": 0.0485, "step": 24550 }, { "grad_norm": 0.29262420535087585, "learning_rate": 8.300508638092972e-05, "loss": 0.0356, "step": 24560 }, { "grad_norm": 0.3955881595611572, "learning_rate": 8.298955797150361e-05, "loss": 0.043, "step": 24570 }, { "grad_norm": 0.6022711992263794, "learning_rate": 8.297402392506433e-05, "loss": 0.0493, "step": 24580 }, { "grad_norm": 0.36351150274276733, "learning_rate": 8.295848424426617e-05, "loss": 0.0439, "step": 24590 }, { "grad_norm": 0.5078209042549133, "learning_rate": 8.29429389317645e-05, "loss": 0.0432, "step": 24600 }, { "grad_norm": 0.3746792674064636, "learning_rate": 8.292738799021556e-05, "loss": 0.0518, "step": 24610 }, { "grad_norm": 0.4495626389980316, "learning_rate": 8.291183142227656e-05, "loss": 0.0379, "step": 24620 }, { "grad_norm": 0.3444778323173523, "learning_rate": 8.289626923060572e-05, "loss": 0.0461, "step": 24630 }, { "grad_norm": 0.426509827375412, "learning_rate": 8.288070141786218e-05, "loss": 0.0533, "step": 24640 }, { "grad_norm": 0.4286760091781616, "learning_rate": 8.286512798670605e-05, "loss": 0.0434, "step": 24650 }, { "grad_norm": 0.3465385138988495, "learning_rate": 8.284954893979842e-05, "loss": 0.0399, "step": 24660 }, { "grad_norm": 0.23688384890556335, "learning_rate": 8.283396427980131e-05, "loss": 0.0468, "step": 24670 }, { "grad_norm": 0.40307170152664185, "learning_rate": 8.281837400937771e-05, "loss": 0.0435, "step": 24680 }, { "grad_norm": 0.449711412191391, "learning_rate": 8.28027781311916e-05, "loss": 0.0416, "step": 24690 }, { "grad_norm": 0.4746142029762268, "learning_rate": 8.278717664790785e-05, "loss": 0.0462, "step": 24700 }, { "grad_norm": 0.43509528040885925, "learning_rate": 8.277156956219234e-05, "loss": 0.0433, "step": 24710 }, { "grad_norm": 0.34044620394706726, "learning_rate": 8.275595687671189e-05, "loss": 0.0435, "step": 24720 }, { "grad_norm": 0.34475553035736084, "learning_rate": 8.27403385941343e-05, "loss": 0.0481, "step": 24730 }, { "grad_norm": 0.4779917299747467, "learning_rate": 8.272471471712828e-05, "loss": 0.0547, "step": 24740 }, { "grad_norm": 0.3332698345184326, "learning_rate": 8.270908524836355e-05, "loss": 0.0531, "step": 24750 }, { "grad_norm": 0.45173391699790955, "learning_rate": 8.269345019051074e-05, "loss": 0.0517, "step": 24760 }, { "grad_norm": 0.47815608978271484, "learning_rate": 8.267780954624147e-05, "loss": 0.05, "step": 24770 }, { "grad_norm": 0.41493701934814453, "learning_rate": 8.266216331822827e-05, "loss": 0.0507, "step": 24780 }, { "grad_norm": 0.3590889275074005, "learning_rate": 8.264651150914469e-05, "loss": 0.0456, "step": 24790 }, { "grad_norm": 0.32927295565605164, "learning_rate": 8.263085412166517e-05, "loss": 0.0358, "step": 24800 }, { "grad_norm": 0.5677127838134766, "learning_rate": 8.261519115846514e-05, "loss": 0.0427, "step": 24810 }, { "grad_norm": 0.32637202739715576, "learning_rate": 8.259952262222096e-05, "loss": 0.0479, "step": 24820 }, { "grad_norm": 0.3953128159046173, "learning_rate": 8.258384851560997e-05, "loss": 0.0487, "step": 24830 }, { "grad_norm": 0.5728833079338074, "learning_rate": 8.256816884131044e-05, "loss": 0.0414, "step": 24840 }, { "grad_norm": 0.508868396282196, "learning_rate": 8.255248360200159e-05, "loss": 0.0404, "step": 24850 }, { "grad_norm": 0.4068582355976105, "learning_rate": 8.253679280036359e-05, "loss": 0.0386, "step": 24860 }, { "grad_norm": 0.4653545916080475, "learning_rate": 8.252109643907762e-05, "loss": 0.0567, "step": 24870 }, { "grad_norm": 0.4028494656085968, "learning_rate": 8.250539452082569e-05, "loss": 0.0457, "step": 24880 }, { "grad_norm": 0.4276951551437378, "learning_rate": 8.248968704829087e-05, "loss": 0.0461, "step": 24890 }, { "grad_norm": 0.46083804965019226, "learning_rate": 8.247397402415714e-05, "loss": 0.0468, "step": 24900 }, { "grad_norm": 0.5104926228523254, "learning_rate": 8.24582554511094e-05, "loss": 0.0436, "step": 24910 }, { "grad_norm": 0.3728441298007965, "learning_rate": 8.244253133183355e-05, "loss": 0.0394, "step": 24920 }, { "grad_norm": 0.44234246015548706, "learning_rate": 8.24268016690164e-05, "loss": 0.0355, "step": 24930 }, { "grad_norm": 0.30904513597488403, "learning_rate": 8.241106646534571e-05, "loss": 0.0534, "step": 24940 }, { "grad_norm": 0.638610303401947, "learning_rate": 8.23953257235102e-05, "loss": 0.0482, "step": 24950 }, { "grad_norm": 0.47651296854019165, "learning_rate": 8.237957944619956e-05, "loss": 0.0469, "step": 24960 }, { "grad_norm": 0.4391637146472931, "learning_rate": 8.236382763610437e-05, "loss": 0.0443, "step": 24970 }, { "grad_norm": 0.4562581181526184, "learning_rate": 8.234807029591619e-05, "loss": 0.0428, "step": 24980 }, { "grad_norm": 0.41244643926620483, "learning_rate": 8.233230742832752e-05, "loss": 0.0583, "step": 24990 }, { "grad_norm": 0.4955912232398987, "learning_rate": 8.231653903603178e-05, "loss": 0.0473, "step": 25000 }, { "grad_norm": 0.33756646513938904, "learning_rate": 8.23007651217234e-05, "loss": 0.0529, "step": 25010 }, { "grad_norm": 0.42939475178718567, "learning_rate": 8.228498568809769e-05, "loss": 0.0456, "step": 25020 }, { "grad_norm": 0.5824031829833984, "learning_rate": 8.22692007378509e-05, "loss": 0.0529, "step": 25030 }, { "grad_norm": 0.6059701442718506, "learning_rate": 8.225341027368028e-05, "loss": 0.0455, "step": 25040 }, { "grad_norm": 0.44520220160484314, "learning_rate": 8.223761429828399e-05, "loss": 0.045, "step": 25050 }, { "grad_norm": 0.564903974533081, "learning_rate": 8.22218128143611e-05, "loss": 0.0467, "step": 25060 }, { "grad_norm": 0.39217278361320496, "learning_rate": 8.220600582461166e-05, "loss": 0.0483, "step": 25070 }, { "grad_norm": 1.7232410907745361, "learning_rate": 8.219019333173668e-05, "loss": 0.0544, "step": 25080 }, { "grad_norm": 0.5147327780723572, "learning_rate": 8.217437533843805e-05, "loss": 0.0499, "step": 25090 }, { "grad_norm": 0.3927862346172333, "learning_rate": 8.215855184741867e-05, "loss": 0.0429, "step": 25100 }, { "grad_norm": 0.5533998012542725, "learning_rate": 8.21427228613823e-05, "loss": 0.0502, "step": 25110 }, { "grad_norm": 0.5637776851654053, "learning_rate": 8.21268883830337e-05, "loss": 0.0486, "step": 25120 }, { "grad_norm": 0.527957022190094, "learning_rate": 8.211104841507855e-05, "loss": 0.0621, "step": 25130 }, { "grad_norm": 0.32288050651550293, "learning_rate": 8.209520296022346e-05, "loss": 0.0459, "step": 25140 }, { "grad_norm": 0.4973096251487732, "learning_rate": 8.207935202117599e-05, "loss": 0.0436, "step": 25150 }, { "grad_norm": 0.47082969546318054, "learning_rate": 8.206349560064463e-05, "loss": 0.0426, "step": 25160 }, { "grad_norm": 0.3316916525363922, "learning_rate": 8.204763370133881e-05, "loss": 0.0405, "step": 25170 }, { "grad_norm": 1.8481509685516357, "learning_rate": 8.203176632596892e-05, "loss": 0.059, "step": 25180 }, { "grad_norm": 0.5942482352256775, "learning_rate": 8.20158934772462e-05, "loss": 0.0553, "step": 25190 }, { "grad_norm": 0.38257479667663574, "learning_rate": 8.200001515788294e-05, "loss": 0.0388, "step": 25200 }, { "grad_norm": 0.43473020195961, "learning_rate": 8.198413137059228e-05, "loss": 0.0423, "step": 25210 }, { "grad_norm": 0.48924410343170166, "learning_rate": 8.196824211808835e-05, "loss": 0.0374, "step": 25220 }, { "grad_norm": 0.37256452441215515, "learning_rate": 8.195234740308617e-05, "loss": 0.054, "step": 25230 }, { "grad_norm": 0.4073905944824219, "learning_rate": 8.193644722830171e-05, "loss": 0.0364, "step": 25240 }, { "grad_norm": 0.3709266185760498, "learning_rate": 8.19205415964519e-05, "loss": 0.0433, "step": 25250 }, { "grad_norm": 0.5608689188957214, "learning_rate": 8.190463051025456e-05, "loss": 0.0408, "step": 25260 }, { "grad_norm": 0.2828281819820404, "learning_rate": 8.188871397242843e-05, "loss": 0.0383, "step": 25270 }, { "grad_norm": 0.453049898147583, "learning_rate": 8.187279198569326e-05, "loss": 0.0384, "step": 25280 }, { "grad_norm": 0.3376624286174774, "learning_rate": 8.185686455276966e-05, "loss": 0.0446, "step": 25290 }, { "grad_norm": 0.40665504336357117, "learning_rate": 8.184093167637921e-05, "loss": 0.0441, "step": 25300 }, { "grad_norm": 0.5521065592765808, "learning_rate": 8.182499335924437e-05, "loss": 0.0551, "step": 25310 }, { "grad_norm": 0.447621613740921, "learning_rate": 8.18090496040886e-05, "loss": 0.0354, "step": 25320 }, { "grad_norm": 0.3296108543872833, "learning_rate": 8.179310041363621e-05, "loss": 0.0547, "step": 25330 }, { "grad_norm": 0.41974014043807983, "learning_rate": 8.17771457906125e-05, "loss": 0.0375, "step": 25340 }, { "grad_norm": 0.443829745054245, "learning_rate": 8.176118573774371e-05, "loss": 0.0465, "step": 25350 }, { "grad_norm": 0.4233633279800415, "learning_rate": 8.174522025775692e-05, "loss": 0.0423, "step": 25360 }, { "grad_norm": 0.5851500034332275, "learning_rate": 8.172924935338022e-05, "loss": 0.0421, "step": 25370 }, { "grad_norm": 0.33485376834869385, "learning_rate": 8.171327302734262e-05, "loss": 0.0351, "step": 25380 }, { "grad_norm": 0.27571216225624084, "learning_rate": 8.169729128237401e-05, "loss": 0.0417, "step": 25390 }, { "grad_norm": 0.2711070775985718, "learning_rate": 8.168130412120525e-05, "loss": 0.0359, "step": 25400 }, { "grad_norm": 0.3814058005809784, "learning_rate": 8.16653115465681e-05, "loss": 0.0454, "step": 25410 }, { "grad_norm": 0.44824352860450745, "learning_rate": 8.164931356119526e-05, "loss": 0.0337, "step": 25420 }, { "grad_norm": 0.43667274713516235, "learning_rate": 8.163331016782032e-05, "loss": 0.0497, "step": 25430 }, { "grad_norm": 0.39893975853919983, "learning_rate": 8.161730136917785e-05, "loss": 0.0375, "step": 25440 }, { "grad_norm": 0.4473954439163208, "learning_rate": 8.160128716800333e-05, "loss": 0.0421, "step": 25450 }, { "grad_norm": 0.45794540643692017, "learning_rate": 8.158526756703313e-05, "loss": 0.0373, "step": 25460 }, { "grad_norm": 0.4625878930091858, "learning_rate": 8.156924256900455e-05, "loss": 0.0494, "step": 25470 }, { "grad_norm": 0.4105668067932129, "learning_rate": 8.155321217665584e-05, "loss": 0.0406, "step": 25480 }, { "grad_norm": 0.4491726756095886, "learning_rate": 8.153717639272614e-05, "loss": 0.0387, "step": 25490 }, { "grad_norm": 0.4199806749820709, "learning_rate": 8.152113521995555e-05, "loss": 0.0448, "step": 25500 }, { "grad_norm": 0.3340696096420288, "learning_rate": 8.150508866108505e-05, "loss": 0.0438, "step": 25510 }, { "grad_norm": 0.3993602395057678, "learning_rate": 8.148903671885657e-05, "loss": 0.044, "step": 25520 }, { "grad_norm": 0.6653848886489868, "learning_rate": 8.147297939601292e-05, "loss": 0.0528, "step": 25530 }, { "grad_norm": 0.4589952826499939, "learning_rate": 8.145691669529792e-05, "loss": 0.048, "step": 25540 }, { "grad_norm": 0.46121829748153687, "learning_rate": 8.144084861945618e-05, "loss": 0.0489, "step": 25550 }, { "grad_norm": 0.5802536606788635, "learning_rate": 8.142477517123333e-05, "loss": 0.0453, "step": 25560 }, { "grad_norm": 0.4604322910308838, "learning_rate": 8.140869635337586e-05, "loss": 0.0408, "step": 25570 }, { "grad_norm": 0.33784621953964233, "learning_rate": 8.139261216863123e-05, "loss": 0.0413, "step": 25580 }, { "grad_norm": 0.4863550662994385, "learning_rate": 8.137652261974776e-05, "loss": 0.0471, "step": 25590 }, { "grad_norm": 0.3104639947414398, "learning_rate": 8.136042770947472e-05, "loss": 0.0444, "step": 25600 }, { "grad_norm": 0.3375232219696045, "learning_rate": 8.134432744056228e-05, "loss": 0.0455, "step": 25610 }, { "grad_norm": 0.42852672934532166, "learning_rate": 8.132822181576158e-05, "loss": 0.0465, "step": 25620 }, { "grad_norm": 0.40180808305740356, "learning_rate": 8.131211083782459e-05, "loss": 0.046, "step": 25630 }, { "grad_norm": 1.0443493127822876, "learning_rate": 8.129599450950424e-05, "loss": 0.0424, "step": 25640 }, { "grad_norm": 0.46095529198646545, "learning_rate": 8.127987283355438e-05, "loss": 0.0556, "step": 25650 }, { "grad_norm": 0.3888355493545532, "learning_rate": 8.126374581272976e-05, "loss": 0.0435, "step": 25660 }, { "grad_norm": 0.6895122528076172, "learning_rate": 8.124761344978605e-05, "loss": 0.0594, "step": 25670 }, { "grad_norm": 0.42029625177383423, "learning_rate": 8.12314757474798e-05, "loss": 0.0478, "step": 25680 }, { "grad_norm": 0.30390727519989014, "learning_rate": 8.121533270856856e-05, "loss": 0.0431, "step": 25690 }, { "grad_norm": 0.34002310037612915, "learning_rate": 8.119918433581069e-05, "loss": 0.0368, "step": 25700 }, { "grad_norm": 0.413675457239151, "learning_rate": 8.118303063196551e-05, "loss": 0.0464, "step": 25710 }, { "grad_norm": 0.3485362231731415, "learning_rate": 8.116687159979326e-05, "loss": 0.0441, "step": 25720 }, { "grad_norm": 0.47929853200912476, "learning_rate": 8.115070724205508e-05, "loss": 0.0486, "step": 25730 }, { "grad_norm": 0.43092411756515503, "learning_rate": 8.113453756151296e-05, "loss": 0.0389, "step": 25740 }, { "grad_norm": 0.3494811952114105, "learning_rate": 8.111836256092995e-05, "loss": 0.0407, "step": 25750 }, { "grad_norm": 0.3224918246269226, "learning_rate": 8.110218224306985e-05, "loss": 0.0462, "step": 25760 }, { "grad_norm": 0.2650294303894043, "learning_rate": 8.108599661069745e-05, "loss": 0.0314, "step": 25770 }, { "grad_norm": 0.32532644271850586, "learning_rate": 8.106980566657845e-05, "loss": 0.0352, "step": 25780 }, { "grad_norm": 0.3268069326877594, "learning_rate": 8.10536094134794e-05, "loss": 0.0372, "step": 25790 }, { "grad_norm": 0.47673168778419495, "learning_rate": 8.103740785416783e-05, "loss": 0.0418, "step": 25800 }, { "grad_norm": 0.33009573817253113, "learning_rate": 8.102120099141212e-05, "loss": 0.0454, "step": 25810 }, { "grad_norm": 0.5860452651977539, "learning_rate": 8.100498882798163e-05, "loss": 0.0446, "step": 25820 }, { "grad_norm": 0.41805386543273926, "learning_rate": 8.09887713666465e-05, "loss": 0.0409, "step": 25830 }, { "grad_norm": 0.5390881299972534, "learning_rate": 8.09725486101779e-05, "loss": 0.054, "step": 25840 }, { "grad_norm": 0.45309510827064514, "learning_rate": 8.095632056134784e-05, "loss": 0.0413, "step": 25850 }, { "grad_norm": 0.3486673831939697, "learning_rate": 8.094008722292925e-05, "loss": 0.0429, "step": 25860 }, { "grad_norm": 0.36372482776641846, "learning_rate": 8.092384859769598e-05, "loss": 0.0481, "step": 25870 }, { "grad_norm": 0.49135473370552063, "learning_rate": 8.090760468842275e-05, "loss": 0.071, "step": 25880 }, { "grad_norm": 0.2892146706581116, "learning_rate": 8.089135549788521e-05, "loss": 0.0428, "step": 25890 }, { "grad_norm": 0.4071658253669739, "learning_rate": 8.087510102885987e-05, "loss": 0.0421, "step": 25900 }, { "grad_norm": 0.32909753918647766, "learning_rate": 8.085884128412422e-05, "loss": 0.0417, "step": 25910 }, { "grad_norm": 0.5393486618995667, "learning_rate": 8.084257626645659e-05, "loss": 0.0488, "step": 25920 }, { "grad_norm": 0.39570561051368713, "learning_rate": 8.08263059786362e-05, "loss": 0.0391, "step": 25930 }, { "grad_norm": 0.6386479139328003, "learning_rate": 8.081003042344325e-05, "loss": 0.0399, "step": 25940 }, { "grad_norm": 0.49663302302360535, "learning_rate": 8.079374960365872e-05, "loss": 0.0472, "step": 25950 }, { "grad_norm": 0.4018016755580902, "learning_rate": 8.077746352206463e-05, "loss": 0.0481, "step": 25960 }, { "grad_norm": 0.3568146824836731, "learning_rate": 8.076117218144377e-05, "loss": 0.0459, "step": 25970 }, { "grad_norm": 0.7933804988861084, "learning_rate": 8.074487558457991e-05, "loss": 0.0485, "step": 25980 }, { "grad_norm": 0.48007044196128845, "learning_rate": 8.072857373425768e-05, "loss": 0.0404, "step": 25990 }, { "grad_norm": 0.3918653428554535, "learning_rate": 8.071226663326264e-05, "loss": 0.0463, "step": 26000 }, { "grad_norm": 0.5068749785423279, "learning_rate": 8.069595428438121e-05, "loss": 0.044, "step": 26010 }, { "grad_norm": 0.3002578318119049, "learning_rate": 8.067963669040072e-05, "loss": 0.0407, "step": 26020 }, { "grad_norm": 0.4849611818790436, "learning_rate": 8.066331385410942e-05, "loss": 0.0463, "step": 26030 }, { "grad_norm": 0.36581405997276306, "learning_rate": 8.064698577829641e-05, "loss": 0.0416, "step": 26040 }, { "grad_norm": 0.36505088210105896, "learning_rate": 8.063065246575175e-05, "loss": 0.053, "step": 26050 }, { "grad_norm": 0.37941351532936096, "learning_rate": 8.061431391926631e-05, "loss": 0.0385, "step": 26060 }, { "grad_norm": 0.2561269700527191, "learning_rate": 8.059797014163195e-05, "loss": 0.0425, "step": 26070 }, { "grad_norm": 0.6118792295455933, "learning_rate": 8.058162113564133e-05, "loss": 0.0526, "step": 26080 }, { "grad_norm": 0.4233216345310211, "learning_rate": 8.056526690408806e-05, "loss": 0.0523, "step": 26090 }, { "grad_norm": 0.5693548917770386, "learning_rate": 8.054890744976666e-05, "loss": 0.0445, "step": 26100 }, { "grad_norm": 0.30014675855636597, "learning_rate": 8.053254277547248e-05, "loss": 0.0363, "step": 26110 }, { "grad_norm": 0.29987791180610657, "learning_rate": 8.051617288400182e-05, "loss": 0.0418, "step": 26120 }, { "grad_norm": 0.3521197438240051, "learning_rate": 8.049979777815182e-05, "loss": 0.0462, "step": 26130 }, { "grad_norm": 0.4544470012187958, "learning_rate": 8.048341746072054e-05, "loss": 0.0393, "step": 26140 }, { "grad_norm": 0.44968539476394653, "learning_rate": 8.046703193450696e-05, "loss": 0.0541, "step": 26150 }, { "grad_norm": 0.3115447759628296, "learning_rate": 8.04506412023109e-05, "loss": 0.0428, "step": 26160 }, { "grad_norm": 0.36931759119033813, "learning_rate": 8.043424526693306e-05, "loss": 0.0382, "step": 26170 }, { "grad_norm": 0.494671106338501, "learning_rate": 8.04178441311751e-05, "loss": 0.0427, "step": 26180 }, { "grad_norm": 0.49963486194610596, "learning_rate": 8.04014377978395e-05, "loss": 0.0485, "step": 26190 }, { "grad_norm": 0.30555158853530884, "learning_rate": 8.038502626972967e-05, "loss": 0.0465, "step": 26200 }, { "grad_norm": 0.5973759889602661, "learning_rate": 8.036860954964989e-05, "loss": 0.0509, "step": 26210 }, { "grad_norm": 0.35372570157051086, "learning_rate": 8.035218764040531e-05, "loss": 0.0417, "step": 26220 }, { "grad_norm": 0.4513118863105774, "learning_rate": 8.033576054480199e-05, "loss": 0.0445, "step": 26230 }, { "grad_norm": 0.2980908453464508, "learning_rate": 8.031932826564688e-05, "loss": 0.0458, "step": 26240 }, { "grad_norm": 0.3640444874763489, "learning_rate": 8.030289080574782e-05, "loss": 0.0506, "step": 26250 }, { "grad_norm": 0.38614875078201294, "learning_rate": 8.028644816791349e-05, "loss": 0.0498, "step": 26260 }, { "grad_norm": 0.3461349904537201, "learning_rate": 8.027000035495351e-05, "loss": 0.038, "step": 26270 }, { "grad_norm": 0.3889789879322052, "learning_rate": 8.025354736967836e-05, "loss": 0.0543, "step": 26280 }, { "grad_norm": 0.3935631215572357, "learning_rate": 8.023708921489941e-05, "loss": 0.0584, "step": 26290 }, { "grad_norm": 0.38191184401512146, "learning_rate": 8.022062589342887e-05, "loss": 0.0558, "step": 26300 }, { "grad_norm": 0.4984627664089203, "learning_rate": 8.020415740807993e-05, "loss": 0.059, "step": 26310 }, { "grad_norm": 0.641731858253479, "learning_rate": 8.018768376166656e-05, "loss": 0.0513, "step": 26320 }, { "grad_norm": 0.35234010219573975, "learning_rate": 8.017120495700368e-05, "loss": 0.045, "step": 26330 }, { "grad_norm": 0.553178071975708, "learning_rate": 8.015472099690704e-05, "loss": 0.0403, "step": 26340 }, { "grad_norm": 0.3168197274208069, "learning_rate": 8.013823188419332e-05, "loss": 0.0394, "step": 26350 }, { "grad_norm": 0.3104337751865387, "learning_rate": 8.012173762168006e-05, "loss": 0.0455, "step": 26360 }, { "grad_norm": 0.26156508922576904, "learning_rate": 8.010523821218567e-05, "loss": 0.0377, "step": 26370 }, { "grad_norm": 0.33832311630249023, "learning_rate": 8.008873365852945e-05, "loss": 0.0415, "step": 26380 }, { "grad_norm": 0.47381168603897095, "learning_rate": 8.007222396353157e-05, "loss": 0.0413, "step": 26390 }, { "grad_norm": 0.47621607780456543, "learning_rate": 8.00557091300131e-05, "loss": 0.0438, "step": 26400 }, { "grad_norm": 0.33281025290489197, "learning_rate": 8.003918916079597e-05, "loss": 0.0395, "step": 26410 }, { "grad_norm": 0.42099103331565857, "learning_rate": 8.002266405870298e-05, "loss": 0.0455, "step": 26420 }, { "grad_norm": 0.3561139702796936, "learning_rate": 8.000613382655782e-05, "loss": 0.047, "step": 26430 }, { "grad_norm": 0.3897320032119751, "learning_rate": 7.998959846718505e-05, "loss": 0.0423, "step": 26440 }, { "grad_norm": 0.3406539559364319, "learning_rate": 7.997305798341012e-05, "loss": 0.0429, "step": 26450 }, { "grad_norm": 0.2997303605079651, "learning_rate": 7.995651237805937e-05, "loss": 0.041, "step": 26460 }, { "grad_norm": 0.47039756178855896, "learning_rate": 7.993996165395996e-05, "loss": 0.0374, "step": 26470 }, { "grad_norm": 0.37447234988212585, "learning_rate": 7.992340581393996e-05, "loss": 0.0471, "step": 26480 }, { "grad_norm": 0.36221152544021606, "learning_rate": 7.990684486082831e-05, "loss": 0.0433, "step": 26490 }, { "grad_norm": 0.4842582941055298, "learning_rate": 7.989027879745482e-05, "loss": 0.0475, "step": 26500 }, { "grad_norm": 0.3153762221336365, "learning_rate": 7.98737076266502e-05, "loss": 0.0617, "step": 26510 }, { "grad_norm": 0.31081917881965637, "learning_rate": 7.985713135124598e-05, "loss": 0.0408, "step": 26520 }, { "grad_norm": 0.26500824093818665, "learning_rate": 7.98405499740746e-05, "loss": 0.0444, "step": 26530 }, { "grad_norm": 0.3627825677394867, "learning_rate": 7.98239634979694e-05, "loss": 0.042, "step": 26540 }, { "grad_norm": 0.5712037682533264, "learning_rate": 7.98073719257645e-05, "loss": 0.0419, "step": 26550 }, { "grad_norm": 0.4637783467769623, "learning_rate": 7.979077526029499e-05, "loss": 0.054, "step": 26560 }, { "grad_norm": 0.4266360104084015, "learning_rate": 7.977417350439675e-05, "loss": 0.042, "step": 26570 }, { "grad_norm": 0.45742344856262207, "learning_rate": 7.97575666609066e-05, "loss": 0.0399, "step": 26580 }, { "grad_norm": 0.44338154792785645, "learning_rate": 7.974095473266216e-05, "loss": 0.0391, "step": 26590 }, { "grad_norm": 0.5932219624519348, "learning_rate": 7.972433772250198e-05, "loss": 0.0385, "step": 26600 }, { "grad_norm": 0.29408302903175354, "learning_rate": 7.970771563326544e-05, "loss": 0.0442, "step": 26610 }, { "grad_norm": 0.355484277009964, "learning_rate": 7.96910884677928e-05, "loss": 0.0515, "step": 26620 }, { "grad_norm": 0.7626615762710571, "learning_rate": 7.967445622892523e-05, "loss": 0.0485, "step": 26630 }, { "grad_norm": 0.43483787775039673, "learning_rate": 7.965781891950465e-05, "loss": 0.0405, "step": 26640 }, { "grad_norm": 0.4171666204929352, "learning_rate": 7.964117654237397e-05, "loss": 0.0434, "step": 26650 }, { "grad_norm": 0.4364902079105377, "learning_rate": 7.962452910037692e-05, "loss": 0.0504, "step": 26660 }, { "grad_norm": 0.38849711418151855, "learning_rate": 7.96078765963581e-05, "loss": 0.0453, "step": 26670 }, { "grad_norm": 0.42287665605545044, "learning_rate": 7.95912190331629e-05, "loss": 0.0477, "step": 26680 }, { "grad_norm": 0.34724152088165283, "learning_rate": 7.957455641363772e-05, "loss": 0.0437, "step": 26690 }, { "grad_norm": 0.6364507079124451, "learning_rate": 7.955788874062968e-05, "loss": 0.0562, "step": 26700 }, { "grad_norm": 0.41866329312324524, "learning_rate": 7.95412160169869e-05, "loss": 0.0523, "step": 26710 }, { "grad_norm": 0.3924047648906708, "learning_rate": 7.952453824555824e-05, "loss": 0.0402, "step": 26720 }, { "grad_norm": 0.3156697452068329, "learning_rate": 7.95078554291935e-05, "loss": 0.0356, "step": 26730 }, { "grad_norm": 0.4383092522621155, "learning_rate": 7.94911675707433e-05, "loss": 0.0384, "step": 26740 }, { "grad_norm": 0.36498379707336426, "learning_rate": 7.947447467305915e-05, "loss": 0.0509, "step": 26750 }, { "grad_norm": 0.3938833475112915, "learning_rate": 7.94577767389934e-05, "loss": 0.0375, "step": 26760 }, { "grad_norm": 0.4598841667175293, "learning_rate": 7.944107377139928e-05, "loss": 0.042, "step": 26770 }, { "grad_norm": 0.46244820952415466, "learning_rate": 7.942436577313088e-05, "loss": 0.0431, "step": 26780 }, { "grad_norm": 0.321576327085495, "learning_rate": 7.940765274704312e-05, "loss": 0.0429, "step": 26790 }, { "grad_norm": 0.5479474663734436, "learning_rate": 7.939093469599181e-05, "loss": 0.0506, "step": 26800 }, { "grad_norm": 0.49310141801834106, "learning_rate": 7.93742116228336e-05, "loss": 0.0401, "step": 26810 }, { "grad_norm": 0.4446338713169098, "learning_rate": 7.935748353042602e-05, "loss": 0.0483, "step": 26820 }, { "grad_norm": 0.3221665620803833, "learning_rate": 7.934075042162744e-05, "loss": 0.0417, "step": 26830 }, { "grad_norm": 0.4433667063713074, "learning_rate": 7.932401229929705e-05, "loss": 0.044, "step": 26840 }, { "grad_norm": 0.3655078113079071, "learning_rate": 7.9307269166295e-05, "loss": 0.0385, "step": 26850 }, { "grad_norm": 0.38074690103530884, "learning_rate": 7.92905210254822e-05, "loss": 0.0387, "step": 26860 }, { "grad_norm": 0.35916954278945923, "learning_rate": 7.927376787972045e-05, "loss": 0.0386, "step": 26870 }, { "grad_norm": 0.48109006881713867, "learning_rate": 7.92570097318724e-05, "loss": 0.0404, "step": 26880 }, { "grad_norm": 0.4652099907398224, "learning_rate": 7.924024658480158e-05, "loss": 0.0392, "step": 26890 }, { "grad_norm": 0.4737197756767273, "learning_rate": 7.922347844137233e-05, "loss": 0.0449, "step": 26900 }, { "grad_norm": 0.5309536457061768, "learning_rate": 7.92067053044499e-05, "loss": 0.0461, "step": 26910 }, { "grad_norm": 0.4425123631954193, "learning_rate": 7.918992717690031e-05, "loss": 0.0483, "step": 26920 }, { "grad_norm": 0.3889213502407074, "learning_rate": 7.917314406159053e-05, "loss": 0.0403, "step": 26930 }, { "grad_norm": 0.7405775189399719, "learning_rate": 7.915635596138832e-05, "loss": 0.0706, "step": 26940 }, { "grad_norm": 0.6435184478759766, "learning_rate": 7.913956287916228e-05, "loss": 0.0484, "step": 26950 }, { "grad_norm": 0.5962311029434204, "learning_rate": 7.912276481778193e-05, "loss": 0.041, "step": 26960 }, { "grad_norm": 0.39521288871765137, "learning_rate": 7.910596178011759e-05, "loss": 0.0542, "step": 26970 }, { "grad_norm": 0.3213997781276703, "learning_rate": 7.908915376904043e-05, "loss": 0.0343, "step": 26980 }, { "grad_norm": 0.4201030731201172, "learning_rate": 7.907234078742247e-05, "loss": 0.049, "step": 26990 }, { "grad_norm": 0.4244071841239929, "learning_rate": 7.90555228381366e-05, "loss": 0.0418, "step": 27000 }, { "grad_norm": 0.36400875449180603, "learning_rate": 7.903869992405656e-05, "loss": 0.0295, "step": 27010 }, { "grad_norm": 0.4057026207447052, "learning_rate": 7.902187204805691e-05, "loss": 0.04, "step": 27020 }, { "grad_norm": 0.521324098110199, "learning_rate": 7.900503921301308e-05, "loss": 0.0387, "step": 27030 }, { "grad_norm": 0.4652025103569031, "learning_rate": 7.898820142180133e-05, "loss": 0.0532, "step": 27040 }, { "grad_norm": 0.40566667914390564, "learning_rate": 7.897135867729879e-05, "loss": 0.0433, "step": 27050 }, { "grad_norm": 0.4253018796443939, "learning_rate": 7.89545109823834e-05, "loss": 0.0439, "step": 27060 }, { "grad_norm": 0.4462663233280182, "learning_rate": 7.8937658339934e-05, "loss": 0.0445, "step": 27070 }, { "grad_norm": 0.5074495673179626, "learning_rate": 7.892080075283026e-05, "loss": 0.053, "step": 27080 }, { "grad_norm": 0.3637610375881195, "learning_rate": 7.890393822395263e-05, "loss": 0.0411, "step": 27090 }, { "grad_norm": 0.3919934928417206, "learning_rate": 7.88870707561825e-05, "loss": 0.0424, "step": 27100 }, { "grad_norm": 0.5485628247261047, "learning_rate": 7.887019835240203e-05, "loss": 0.0532, "step": 27110 }, { "grad_norm": 0.3479108214378357, "learning_rate": 7.885332101549427e-05, "loss": 0.0461, "step": 27120 }, { "grad_norm": 0.46514829993247986, "learning_rate": 7.883643874834308e-05, "loss": 0.0472, "step": 27130 }, { "grad_norm": 0.35578882694244385, "learning_rate": 7.881955155383321e-05, "loss": 0.0509, "step": 27140 }, { "grad_norm": 0.46632087230682373, "learning_rate": 7.880265943485017e-05, "loss": 0.041, "step": 27150 }, { "grad_norm": 0.37303856015205383, "learning_rate": 7.878576239428038e-05, "loss": 0.0379, "step": 27160 }, { "grad_norm": 0.3962392807006836, "learning_rate": 7.87688604350111e-05, "loss": 0.0465, "step": 27170 }, { "grad_norm": 0.5666502714157104, "learning_rate": 7.875195355993042e-05, "loss": 0.048, "step": 27180 }, { "grad_norm": 0.501179575920105, "learning_rate": 7.873504177192724e-05, "loss": 0.0521, "step": 27190 }, { "grad_norm": 0.4152958393096924, "learning_rate": 7.87181250738913e-05, "loss": 0.0592, "step": 27200 }, { "grad_norm": 0.2729153633117676, "learning_rate": 7.870120346871324e-05, "loss": 0.0396, "step": 27210 }, { "grad_norm": 0.4044935405254364, "learning_rate": 7.86842769592845e-05, "loss": 0.0439, "step": 27220 }, { "grad_norm": 0.3470717966556549, "learning_rate": 7.866734554849732e-05, "loss": 0.0378, "step": 27230 }, { "grad_norm": 0.28031694889068604, "learning_rate": 7.865040923924486e-05, "loss": 0.0383, "step": 27240 }, { "grad_norm": 0.38506898283958435, "learning_rate": 7.863346803442104e-05, "loss": 0.0485, "step": 27250 }, { "grad_norm": 0.3959350883960724, "learning_rate": 7.861652193692067e-05, "loss": 0.0408, "step": 27260 }, { "grad_norm": 0.33680328726768494, "learning_rate": 7.859957094963937e-05, "loss": 0.0329, "step": 27270 }, { "grad_norm": 0.4261421859264374, "learning_rate": 7.858261507547357e-05, "loss": 0.0384, "step": 27280 }, { "grad_norm": 0.49002501368522644, "learning_rate": 7.856565431732061e-05, "loss": 0.0382, "step": 27290 }, { "grad_norm": 0.3169425129890442, "learning_rate": 7.854868867807859e-05, "loss": 0.0289, "step": 27300 }, { "grad_norm": 0.5761030316352844, "learning_rate": 7.85317181606465e-05, "loss": 0.0481, "step": 27310 }, { "grad_norm": 0.38996532559394836, "learning_rate": 7.85147427679241e-05, "loss": 0.057, "step": 27320 }, { "grad_norm": 0.45963600277900696, "learning_rate": 7.849776250281205e-05, "loss": 0.0305, "step": 27330 }, { "grad_norm": 0.4149990677833557, "learning_rate": 7.84807773682118e-05, "loss": 0.0423, "step": 27340 }, { "grad_norm": 0.5223187804222107, "learning_rate": 7.846378736702565e-05, "loss": 0.0379, "step": 27350 }, { "grad_norm": 0.4530482590198517, "learning_rate": 7.844679250215671e-05, "loss": 0.0407, "step": 27360 }, { "grad_norm": 0.37521854043006897, "learning_rate": 7.842979277650898e-05, "loss": 0.038, "step": 27370 }, { "grad_norm": 0.41155821084976196, "learning_rate": 7.84127881929872e-05, "loss": 0.0371, "step": 27380 }, { "grad_norm": 0.59986811876297, "learning_rate": 7.839577875449704e-05, "loss": 0.0405, "step": 27390 }, { "grad_norm": 0.6846920251846313, "learning_rate": 7.837876446394489e-05, "loss": 0.0463, "step": 27400 }, { "grad_norm": 0.3527023196220398, "learning_rate": 7.836174532423805e-05, "loss": 0.0451, "step": 27410 }, { "grad_norm": 0.3746587038040161, "learning_rate": 7.834472133828466e-05, "loss": 0.0401, "step": 27420 }, { "grad_norm": 0.37821170687675476, "learning_rate": 7.832769250899359e-05, "loss": 0.0455, "step": 27430 }, { "grad_norm": 0.6144205927848816, "learning_rate": 7.831065883927464e-05, "loss": 0.0529, "step": 27440 }, { "grad_norm": 0.29919180274009705, "learning_rate": 7.829362033203841e-05, "loss": 0.0449, "step": 27450 }, { "grad_norm": 0.37507036328315735, "learning_rate": 7.827657699019628e-05, "loss": 0.0398, "step": 27460 }, { "grad_norm": 0.4389922022819519, "learning_rate": 7.825952881666052e-05, "loss": 0.0547, "step": 27470 }, { "grad_norm": 0.3641795516014099, "learning_rate": 7.824247581434418e-05, "loss": 0.045, "step": 27480 }, { "grad_norm": 0.5237253308296204, "learning_rate": 7.822541798616116e-05, "loss": 0.0406, "step": 27490 }, { "grad_norm": 0.37936824560165405, "learning_rate": 7.820835533502617e-05, "loss": 0.0494, "step": 27500 }, { "grad_norm": 0.6221336126327515, "learning_rate": 7.819128786385475e-05, "loss": 0.0485, "step": 27510 }, { "grad_norm": 0.5405499935150146, "learning_rate": 7.817421557556329e-05, "loss": 0.0456, "step": 27520 }, { "grad_norm": 0.3340403735637665, "learning_rate": 7.815713847306893e-05, "loss": 0.0526, "step": 27530 }, { "grad_norm": 0.30877313017845154, "learning_rate": 7.81400565592897e-05, "loss": 0.0416, "step": 27540 }, { "grad_norm": 0.3654496371746063, "learning_rate": 7.812296983714444e-05, "loss": 0.0441, "step": 27550 }, { "grad_norm": 0.3131622076034546, "learning_rate": 7.810587830955281e-05, "loss": 0.0305, "step": 27560 }, { "grad_norm": 0.4263973832130432, "learning_rate": 7.808878197943528e-05, "loss": 0.0347, "step": 27570 }, { "grad_norm": 0.4199911952018738, "learning_rate": 7.807168084971312e-05, "loss": 0.0362, "step": 27580 }, { "grad_norm": 0.34059587121009827, "learning_rate": 7.805457492330849e-05, "loss": 0.0394, "step": 27590 }, { "grad_norm": 0.4747062921524048, "learning_rate": 7.803746420314428e-05, "loss": 0.0432, "step": 27600 }, { "grad_norm": 0.37899044156074524, "learning_rate": 7.802034869214428e-05, "loss": 0.0526, "step": 27610 }, { "grad_norm": 0.3223215937614441, "learning_rate": 7.800322839323303e-05, "loss": 0.0353, "step": 27620 }, { "grad_norm": 0.23094071447849274, "learning_rate": 7.798610330933593e-05, "loss": 0.0324, "step": 27630 }, { "grad_norm": 0.3673175573348999, "learning_rate": 7.796897344337922e-05, "loss": 0.041, "step": 27640 }, { "grad_norm": 0.5740441679954529, "learning_rate": 7.795183879828989e-05, "loss": 0.0574, "step": 27650 }, { "grad_norm": 0.5016072392463684, "learning_rate": 7.793469937699579e-05, "loss": 0.0539, "step": 27660 }, { "grad_norm": 0.5247713923454285, "learning_rate": 7.791755518242558e-05, "loss": 0.0539, "step": 27670 }, { "grad_norm": 0.536815881729126, "learning_rate": 7.790040621750876e-05, "loss": 0.0443, "step": 27680 }, { "grad_norm": 0.3934064507484436, "learning_rate": 7.788325248517558e-05, "loss": 0.0333, "step": 27690 }, { "grad_norm": 0.4312394857406616, "learning_rate": 7.786609398835715e-05, "loss": 0.0463, "step": 27700 }, { "grad_norm": 0.2879452705383301, "learning_rate": 7.784893072998541e-05, "loss": 0.0415, "step": 27710 }, { "grad_norm": 0.2758459746837616, "learning_rate": 7.783176271299306e-05, "loss": 0.0345, "step": 27720 }, { "grad_norm": 0.42457276582717896, "learning_rate": 7.781458994031368e-05, "loss": 0.0355, "step": 27730 }, { "grad_norm": 0.4621751010417938, "learning_rate": 7.779741241488161e-05, "loss": 0.0393, "step": 27740 }, { "grad_norm": 0.30026817321777344, "learning_rate": 7.7780230139632e-05, "loss": 0.0423, "step": 27750 }, { "grad_norm": 0.28432270884513855, "learning_rate": 7.776304311750087e-05, "loss": 0.0406, "step": 27760 }, { "grad_norm": 0.4930708408355713, "learning_rate": 7.7745851351425e-05, "loss": 0.0605, "step": 27770 }, { "grad_norm": 0.4781675338745117, "learning_rate": 7.772865484434197e-05, "loss": 0.0452, "step": 27780 }, { "grad_norm": 0.5138699412345886, "learning_rate": 7.77114535991902e-05, "loss": 0.049, "step": 27790 }, { "grad_norm": 0.3412366509437561, "learning_rate": 7.769424761890893e-05, "loss": 0.0416, "step": 27800 }, { "grad_norm": 0.4690694212913513, "learning_rate": 7.767703690643817e-05, "loss": 0.0448, "step": 27810 }, { "grad_norm": 0.3011980652809143, "learning_rate": 7.76598214647188e-05, "loss": 0.04, "step": 27820 }, { "grad_norm": 0.2929426431655884, "learning_rate": 7.764260129669241e-05, "loss": 0.0399, "step": 27830 }, { "grad_norm": 0.3192470073699951, "learning_rate": 7.76253764053015e-05, "loss": 0.0462, "step": 27840 }, { "grad_norm": 0.3061908185482025, "learning_rate": 7.760814679348932e-05, "loss": 0.0386, "step": 27850 }, { "grad_norm": 0.33945196866989136, "learning_rate": 7.759091246419992e-05, "loss": 0.035, "step": 27860 }, { "grad_norm": 0.26244068145751953, "learning_rate": 7.75736734203782e-05, "loss": 0.034, "step": 27870 }, { "grad_norm": 0.47904735803604126, "learning_rate": 7.755642966496985e-05, "loss": 0.0561, "step": 27880 }, { "grad_norm": 0.3973439037799835, "learning_rate": 7.753918120092132e-05, "loss": 0.0421, "step": 27890 }, { "grad_norm": 0.35809051990509033, "learning_rate": 7.752192803117993e-05, "loss": 0.0362, "step": 27900 }, { "grad_norm": 0.4360904097557068, "learning_rate": 7.750467015869377e-05, "loss": 0.0441, "step": 27910 }, { "grad_norm": 0.3737451434135437, "learning_rate": 7.748740758641174e-05, "loss": 0.0479, "step": 27920 }, { "grad_norm": 0.403119295835495, "learning_rate": 7.74701403172835e-05, "loss": 0.0408, "step": 27930 }, { "grad_norm": 0.6595788598060608, "learning_rate": 7.745286835425962e-05, "loss": 0.0442, "step": 27940 }, { "grad_norm": 0.5225640535354614, "learning_rate": 7.743559170029138e-05, "loss": 0.0418, "step": 27950 }, { "grad_norm": 0.32908663153648376, "learning_rate": 7.741831035833087e-05, "loss": 0.0471, "step": 27960 }, { "grad_norm": 0.5833932161331177, "learning_rate": 7.740102433133102e-05, "loss": 0.0375, "step": 27970 }, { "grad_norm": 0.40432360768318176, "learning_rate": 7.738373362224553e-05, "loss": 0.0368, "step": 27980 }, { "grad_norm": 0.4828067123889923, "learning_rate": 7.73664382340289e-05, "loss": 0.0422, "step": 27990 }, { "grad_norm": 0.36361947655677795, "learning_rate": 7.734913816963647e-05, "loss": 0.0393, "step": 28000 }, { "grad_norm": 0.32647886872291565, "learning_rate": 7.73318334320243e-05, "loss": 0.0305, "step": 28010 }, { "grad_norm": 0.5811080932617188, "learning_rate": 7.731452402414934e-05, "loss": 0.0494, "step": 28020 }, { "grad_norm": 0.4285527467727661, "learning_rate": 7.729720994896928e-05, "loss": 0.0414, "step": 28030 }, { "grad_norm": 0.3096672296524048, "learning_rate": 7.727989120944262e-05, "loss": 0.0405, "step": 28040 }, { "grad_norm": 0.33184418082237244, "learning_rate": 7.726256780852865e-05, "loss": 0.0428, "step": 28050 }, { "grad_norm": 0.4910343885421753, "learning_rate": 7.724523974918749e-05, "loss": 0.0347, "step": 28060 }, { "grad_norm": 0.2491484433412552, "learning_rate": 7.722790703438002e-05, "loss": 0.0384, "step": 28070 }, { "grad_norm": 0.41282397508621216, "learning_rate": 7.72105696670679e-05, "loss": 0.0365, "step": 28080 }, { "grad_norm": 0.4557013213634491, "learning_rate": 7.719322765021364e-05, "loss": 0.0436, "step": 28090 }, { "grad_norm": 0.28697189688682556, "learning_rate": 7.717588098678051e-05, "loss": 0.0377, "step": 28100 }, { "grad_norm": 0.3909761309623718, "learning_rate": 7.715852967973258e-05, "loss": 0.0454, "step": 28110 }, { "grad_norm": 0.46204492449760437, "learning_rate": 7.714117373203474e-05, "loss": 0.045, "step": 28120 }, { "grad_norm": 0.3343733251094818, "learning_rate": 7.712381314665259e-05, "loss": 0.0355, "step": 28130 }, { "grad_norm": 0.4143180549144745, "learning_rate": 7.710644792655261e-05, "loss": 0.0434, "step": 28140 }, { "grad_norm": 0.6403763890266418, "learning_rate": 7.708907807470207e-05, "loss": 0.0553, "step": 28150 }, { "grad_norm": 0.2999241352081299, "learning_rate": 7.707170359406896e-05, "loss": 0.0493, "step": 28160 }, { "grad_norm": 0.426533579826355, "learning_rate": 7.705432448762213e-05, "loss": 0.0435, "step": 28170 }, { "grad_norm": 0.36758458614349365, "learning_rate": 7.703694075833117e-05, "loss": 0.049, "step": 28180 }, { "grad_norm": 0.4261493384838104, "learning_rate": 7.70195524091665e-05, "loss": 0.06, "step": 28190 }, { "grad_norm": 0.4765612781047821, "learning_rate": 7.70021594430993e-05, "loss": 0.0498, "step": 28200 }, { "grad_norm": 0.26595786213874817, "learning_rate": 7.698476186310157e-05, "loss": 0.0405, "step": 28210 }, { "grad_norm": 0.3932715356349945, "learning_rate": 7.696735967214608e-05, "loss": 0.0378, "step": 28220 }, { "grad_norm": 0.3210163116455078, "learning_rate": 7.694995287320636e-05, "loss": 0.0454, "step": 28230 }, { "grad_norm": 0.41854822635650635, "learning_rate": 7.693254146925679e-05, "loss": 0.0455, "step": 28240 }, { "grad_norm": 0.3940410614013672, "learning_rate": 7.691512546327251e-05, "loss": 0.0449, "step": 28250 }, { "grad_norm": 0.5562671422958374, "learning_rate": 7.689770485822939e-05, "loss": 0.0417, "step": 28260 }, { "grad_norm": 0.3517301380634308, "learning_rate": 7.688027965710416e-05, "loss": 0.0396, "step": 28270 }, { "grad_norm": 0.3444156348705292, "learning_rate": 7.686284986287433e-05, "loss": 0.0408, "step": 28280 }, { "grad_norm": 0.2448325753211975, "learning_rate": 7.684541547851817e-05, "loss": 0.0374, "step": 28290 }, { "grad_norm": 0.29115918278694153, "learning_rate": 7.68279765070147e-05, "loss": 0.0333, "step": 28300 }, { "grad_norm": 0.4082273244857788, "learning_rate": 7.68105329513438e-05, "loss": 0.03, "step": 28310 }, { "grad_norm": 0.2418806254863739, "learning_rate": 7.67930848144861e-05, "loss": 0.0355, "step": 28320 }, { "grad_norm": 0.42830798029899597, "learning_rate": 7.6775632099423e-05, "loss": 0.0417, "step": 28330 }, { "grad_norm": 0.4485929310321808, "learning_rate": 7.675817480913667e-05, "loss": 0.0399, "step": 28340 }, { "grad_norm": 0.47905126214027405, "learning_rate": 7.674071294661011e-05, "loss": 0.0376, "step": 28350 }, { "grad_norm": 0.3787349760532379, "learning_rate": 7.672324651482707e-05, "loss": 0.0362, "step": 28360 }, { "grad_norm": 0.35275357961654663, "learning_rate": 7.670577551677209e-05, "loss": 0.0496, "step": 28370 }, { "grad_norm": 0.42305439710617065, "learning_rate": 7.668829995543047e-05, "loss": 0.0424, "step": 28380 }, { "grad_norm": 0.3728469908237457, "learning_rate": 7.667081983378832e-05, "loss": 0.0439, "step": 28390 }, { "grad_norm": 0.538091778755188, "learning_rate": 7.66533351548325e-05, "loss": 0.0314, "step": 28400 }, { "grad_norm": 0.4781495928764343, "learning_rate": 7.663584592155069e-05, "loss": 0.0381, "step": 28410 }, { "grad_norm": 0.4488678574562073, "learning_rate": 7.661835213693129e-05, "loss": 0.0471, "step": 28420 }, { "grad_norm": 0.4179801344871521, "learning_rate": 7.660085380396353e-05, "loss": 0.0423, "step": 28430 }, { "grad_norm": 0.3882874846458435, "learning_rate": 7.658335092563738e-05, "loss": 0.0414, "step": 28440 }, { "grad_norm": 0.3492262661457062, "learning_rate": 7.656584350494362e-05, "loss": 0.0374, "step": 28450 }, { "grad_norm": 0.40309375524520874, "learning_rate": 7.654833154487378e-05, "loss": 0.0377, "step": 28460 }, { "grad_norm": 0.4568929970264435, "learning_rate": 7.653081504842017e-05, "loss": 0.0417, "step": 28470 }, { "grad_norm": 0.4714494049549103, "learning_rate": 7.65132940185759e-05, "loss": 0.0453, "step": 28480 }, { "grad_norm": 0.2832322120666504, "learning_rate": 7.649576845833481e-05, "loss": 0.0394, "step": 28490 }, { "grad_norm": 0.31470027565956116, "learning_rate": 7.647823837069156e-05, "loss": 0.0385, "step": 28500 }, { "grad_norm": 0.40301573276519775, "learning_rate": 7.646070375864156e-05, "loss": 0.0363, "step": 28510 }, { "grad_norm": 0.5694223642349243, "learning_rate": 7.644316462518099e-05, "loss": 0.0497, "step": 28520 }, { "grad_norm": 0.4789738655090332, "learning_rate": 7.642562097330679e-05, "loss": 0.0421, "step": 28530 }, { "grad_norm": 0.28033918142318726, "learning_rate": 7.640807280601671e-05, "loss": 0.0472, "step": 28540 }, { "grad_norm": 0.319541871547699, "learning_rate": 7.639052012630927e-05, "loss": 0.0483, "step": 28550 }, { "grad_norm": 0.4233451187610626, "learning_rate": 7.63729629371837e-05, "loss": 0.0451, "step": 28560 }, { "grad_norm": 0.45968547463417053, "learning_rate": 7.635540124164009e-05, "loss": 0.0473, "step": 28570 }, { "grad_norm": 0.3370971977710724, "learning_rate": 7.633783504267922e-05, "loss": 0.047, "step": 28580 }, { "grad_norm": 0.4213160276412964, "learning_rate": 7.632026434330269e-05, "loss": 0.0405, "step": 28590 }, { "grad_norm": 0.34906402230262756, "learning_rate": 7.630268914651282e-05, "loss": 0.0385, "step": 28600 }, { "grad_norm": 0.36204713582992554, "learning_rate": 7.628510945531278e-05, "loss": 0.0585, "step": 28610 }, { "grad_norm": 0.4308859705924988, "learning_rate": 7.626752527270641e-05, "loss": 0.043, "step": 28620 }, { "grad_norm": 0.27000388503074646, "learning_rate": 7.62499366016984e-05, "loss": 0.0368, "step": 28630 }, { "grad_norm": 0.3279731273651123, "learning_rate": 7.623234344529416e-05, "loss": 0.0398, "step": 28640 }, { "grad_norm": 0.3076024651527405, "learning_rate": 7.62147458064999e-05, "loss": 0.0457, "step": 28650 }, { "grad_norm": 0.5349287390708923, "learning_rate": 7.619714368832254e-05, "loss": 0.0499, "step": 28660 }, { "grad_norm": 0.31568479537963867, "learning_rate": 7.61795370937698e-05, "loss": 0.0381, "step": 28670 }, { "grad_norm": 0.3899810016155243, "learning_rate": 7.61619260258502e-05, "loss": 0.0452, "step": 28680 }, { "grad_norm": 0.3421291708946228, "learning_rate": 7.614431048757298e-05, "loss": 0.0418, "step": 28690 }, { "grad_norm": 0.2653948664665222, "learning_rate": 7.612669048194814e-05, "loss": 0.0459, "step": 28700 }, { "grad_norm": 0.4338412582874298, "learning_rate": 7.610906601198646e-05, "loss": 0.0405, "step": 28710 }, { "grad_norm": 0.32987290620803833, "learning_rate": 7.60914370806995e-05, "loss": 0.0431, "step": 28720 }, { "grad_norm": 0.38611915707588196, "learning_rate": 7.607380369109953e-05, "loss": 0.0573, "step": 28730 }, { "grad_norm": 0.24165256321430206, "learning_rate": 7.605616584619961e-05, "loss": 0.0361, "step": 28740 }, { "grad_norm": 0.3304843306541443, "learning_rate": 7.603852354901362e-05, "loss": 0.0376, "step": 28750 }, { "grad_norm": 0.3440503776073456, "learning_rate": 7.602087680255609e-05, "loss": 0.0563, "step": 28760 }, { "grad_norm": 0.5553044676780701, "learning_rate": 7.600322560984238e-05, "loss": 0.0317, "step": 28770 }, { "grad_norm": 0.38242414593696594, "learning_rate": 7.598556997388863e-05, "loss": 0.0441, "step": 28780 }, { "grad_norm": 0.5023542046546936, "learning_rate": 7.596790989771166e-05, "loss": 0.0518, "step": 28790 }, { "grad_norm": 0.3600582778453827, "learning_rate": 7.595024538432914e-05, "loss": 0.0378, "step": 28800 }, { "grad_norm": 0.40248537063598633, "learning_rate": 7.59325764367594e-05, "loss": 0.0348, "step": 28810 }, { "grad_norm": 0.3173551559448242, "learning_rate": 7.59149030580216e-05, "loss": 0.0383, "step": 28820 }, { "grad_norm": 0.5275803208351135, "learning_rate": 7.589722525113562e-05, "loss": 0.0408, "step": 28830 }, { "grad_norm": 0.34753528237342834, "learning_rate": 7.587954301912216e-05, "loss": 0.0353, "step": 28840 }, { "grad_norm": 0.6514681577682495, "learning_rate": 7.586185636500263e-05, "loss": 0.0477, "step": 28850 }, { "grad_norm": 0.4473678469657898, "learning_rate": 7.584416529179914e-05, "loss": 0.043, "step": 28860 }, { "grad_norm": 0.35315391421318054, "learning_rate": 7.582646980253465e-05, "loss": 0.0422, "step": 28870 }, { "grad_norm": 0.4749445915222168, "learning_rate": 7.580876990023282e-05, "loss": 0.0509, "step": 28880 }, { "grad_norm": 0.38109973073005676, "learning_rate": 7.579106558791809e-05, "loss": 0.0408, "step": 28890 }, { "grad_norm": 0.4461543560028076, "learning_rate": 7.577335686861565e-05, "loss": 0.0433, "step": 28900 }, { "grad_norm": 0.40982699394226074, "learning_rate": 7.575564374535141e-05, "loss": 0.0483, "step": 28910 }, { "grad_norm": 0.23552946746349335, "learning_rate": 7.573792622115207e-05, "loss": 0.0382, "step": 28920 }, { "grad_norm": 0.35186851024627686, "learning_rate": 7.572020429904507e-05, "loss": 0.0348, "step": 28930 }, { "grad_norm": 0.42397695779800415, "learning_rate": 7.570247798205861e-05, "loss": 0.0412, "step": 28940 }, { "grad_norm": 0.283224880695343, "learning_rate": 7.568474727322164e-05, "loss": 0.0407, "step": 28950 }, { "grad_norm": 0.3390374779701233, "learning_rate": 7.566701217556384e-05, "loss": 0.0366, "step": 28960 }, { "grad_norm": 0.45780041813850403, "learning_rate": 7.564927269211564e-05, "loss": 0.0433, "step": 28970 }, { "grad_norm": 0.37772849202156067, "learning_rate": 7.563152882590824e-05, "loss": 0.0397, "step": 28980 }, { "grad_norm": 0.36108162999153137, "learning_rate": 7.56137805799736e-05, "loss": 0.0388, "step": 28990 }, { "grad_norm": 0.39955639839172363, "learning_rate": 7.559602795734439e-05, "loss": 0.0378, "step": 29000 }, { "grad_norm": 0.3557141125202179, "learning_rate": 7.557827096105408e-05, "loss": 0.0359, "step": 29010 }, { "grad_norm": 0.36891278624534607, "learning_rate": 7.55605095941368e-05, "loss": 0.032, "step": 29020 }, { "grad_norm": 0.3893932104110718, "learning_rate": 7.554274385962752e-05, "loss": 0.0412, "step": 29030 }, { "grad_norm": 0.4118756651878357, "learning_rate": 7.552497376056191e-05, "loss": 0.0475, "step": 29040 }, { "grad_norm": 0.5609123706817627, "learning_rate": 7.550719929997639e-05, "loss": 0.0451, "step": 29050 }, { "grad_norm": 0.4859306812286377, "learning_rate": 7.548942048090813e-05, "loss": 0.0477, "step": 29060 }, { "grad_norm": 0.44424280524253845, "learning_rate": 7.547163730639506e-05, "loss": 0.0431, "step": 29070 }, { "grad_norm": 0.28975704312324524, "learning_rate": 7.545384977947583e-05, "loss": 0.0349, "step": 29080 }, { "grad_norm": 0.40354230999946594, "learning_rate": 7.543605790318981e-05, "loss": 0.0335, "step": 29090 }, { "grad_norm": 0.41531801223754883, "learning_rate": 7.54182616805772e-05, "loss": 0.0401, "step": 29100 }, { "grad_norm": 0.3329031467437744, "learning_rate": 7.540046111467885e-05, "loss": 0.0383, "step": 29110 }, { "grad_norm": 0.43346577882766724, "learning_rate": 7.53826562085364e-05, "loss": 0.0363, "step": 29120 }, { "grad_norm": 0.3449716866016388, "learning_rate": 7.536484696519221e-05, "loss": 0.0322, "step": 29130 }, { "grad_norm": 0.2827724516391754, "learning_rate": 7.534703338768942e-05, "loss": 0.0336, "step": 29140 }, { "grad_norm": 0.3271254003047943, "learning_rate": 7.532921547907185e-05, "loss": 0.0426, "step": 29150 }, { "grad_norm": 0.3453180491924286, "learning_rate": 7.531139324238412e-05, "loss": 0.0512, "step": 29160 }, { "grad_norm": 0.40045249462127686, "learning_rate": 7.529356668067157e-05, "loss": 0.0484, "step": 29170 }, { "grad_norm": 0.32384899258613586, "learning_rate": 7.527573579698023e-05, "loss": 0.0364, "step": 29180 }, { "grad_norm": 0.41694849729537964, "learning_rate": 7.525790059435693e-05, "loss": 0.0359, "step": 29190 }, { "grad_norm": 0.36786043643951416, "learning_rate": 7.524006107584926e-05, "loss": 0.041, "step": 29200 }, { "grad_norm": 0.507087230682373, "learning_rate": 7.522221724450544e-05, "loss": 0.0386, "step": 29210 }, { "grad_norm": 0.4101198613643646, "learning_rate": 7.520436910337451e-05, "loss": 0.0453, "step": 29220 }, { "grad_norm": 0.3416949212551117, "learning_rate": 7.518651665550627e-05, "loss": 0.0349, "step": 29230 }, { "grad_norm": 0.43765199184417725, "learning_rate": 7.516865990395117e-05, "loss": 0.0492, "step": 29240 }, { "grad_norm": 0.40132197737693787, "learning_rate": 7.515079885176047e-05, "loss": 0.0366, "step": 29250 }, { "grad_norm": 0.25837385654449463, "learning_rate": 7.513293350198612e-05, "loss": 0.04, "step": 29260 }, { "grad_norm": 0.3132605254650116, "learning_rate": 7.511506385768081e-05, "loss": 0.0442, "step": 29270 }, { "grad_norm": 0.30295318365097046, "learning_rate": 7.509718992189801e-05, "loss": 0.0433, "step": 29280 }, { "grad_norm": 0.34258419275283813, "learning_rate": 7.507931169769182e-05, "loss": 0.0411, "step": 29290 }, { "grad_norm": 0.3994949758052826, "learning_rate": 7.506142918811722e-05, "loss": 0.041, "step": 29300 }, { "grad_norm": 0.3947858512401581, "learning_rate": 7.504354239622978e-05, "loss": 0.0434, "step": 29310 }, { "grad_norm": 0.48675259947776794, "learning_rate": 7.50256513250859e-05, "loss": 0.0403, "step": 29320 }, { "grad_norm": 0.4619904160499573, "learning_rate": 7.500775597774265e-05, "loss": 0.0428, "step": 29330 }, { "grad_norm": 0.6688867807388306, "learning_rate": 7.498985635725788e-05, "loss": 0.0337, "step": 29340 }, { "grad_norm": 0.4710502326488495, "learning_rate": 7.497195246669012e-05, "loss": 0.0431, "step": 29350 }, { "grad_norm": 0.6249610185623169, "learning_rate": 7.495404430909868e-05, "loss": 0.0432, "step": 29360 }, { "grad_norm": 0.4495009183883667, "learning_rate": 7.493613188754356e-05, "loss": 0.0391, "step": 29370 }, { "grad_norm": 0.3036944270133972, "learning_rate": 7.49182152050855e-05, "loss": 0.0387, "step": 29380 }, { "grad_norm": 0.2636653482913971, "learning_rate": 7.490029426478598e-05, "loss": 0.0377, "step": 29390 }, { "grad_norm": 0.3095249831676483, "learning_rate": 7.488236906970719e-05, "loss": 0.0347, "step": 29400 }, { "grad_norm": 0.40117090940475464, "learning_rate": 7.486443962291207e-05, "loss": 0.0429, "step": 29410 }, { "grad_norm": 0.5427181124687195, "learning_rate": 7.484650592746424e-05, "loss": 0.0408, "step": 29420 }, { "grad_norm": 0.35234859585762024, "learning_rate": 7.482856798642811e-05, "loss": 0.0416, "step": 29430 }, { "grad_norm": 0.311495840549469, "learning_rate": 7.481062580286878e-05, "loss": 0.0416, "step": 29440 }, { "grad_norm": 0.29673612117767334, "learning_rate": 7.479267937985208e-05, "loss": 0.0379, "step": 29450 }, { "grad_norm": 0.4048323929309845, "learning_rate": 7.477472872044456e-05, "loss": 0.0335, "step": 29460 }, { "grad_norm": 0.389631450176239, "learning_rate": 7.475677382771347e-05, "loss": 0.0442, "step": 29470 }, { "grad_norm": 0.34066012501716614, "learning_rate": 7.473881470472683e-05, "loss": 0.047, "step": 29480 }, { "grad_norm": 0.44642001390457153, "learning_rate": 7.472085135455337e-05, "loss": 0.0371, "step": 29490 }, { "grad_norm": 0.2909397482872009, "learning_rate": 7.470288378026256e-05, "loss": 0.0494, "step": 29500 }, { "grad_norm": 0.4006388783454895, "learning_rate": 7.468491198492451e-05, "loss": 0.038, "step": 29510 }, { "grad_norm": 0.4275747835636139, "learning_rate": 7.466693597161013e-05, "loss": 0.0357, "step": 29520 }, { "grad_norm": 0.2906120717525482, "learning_rate": 7.464895574339104e-05, "loss": 0.0361, "step": 29530 }, { "grad_norm": 0.2494349479675293, "learning_rate": 7.463097130333958e-05, "loss": 0.0429, "step": 29540 }, { "grad_norm": 0.2802888751029968, "learning_rate": 7.461298265452876e-05, "loss": 0.045, "step": 29550 }, { "grad_norm": 0.6028896570205688, "learning_rate": 7.459498980003239e-05, "loss": 0.0315, "step": 29560 }, { "grad_norm": 0.5212306380271912, "learning_rate": 7.457699274292493e-05, "loss": 0.0573, "step": 29570 }, { "grad_norm": 0.415396511554718, "learning_rate": 7.455899148628159e-05, "loss": 0.0523, "step": 29580 }, { "grad_norm": 0.4729115068912506, "learning_rate": 7.45409860331783e-05, "loss": 0.0515, "step": 29590 }, { "grad_norm": 0.5107051730155945, "learning_rate": 7.452297638669169e-05, "loss": 0.04, "step": 29600 }, { "grad_norm": 0.26970723271369934, "learning_rate": 7.450496254989911e-05, "loss": 0.0298, "step": 29610 }, { "grad_norm": 0.49365243315696716, "learning_rate": 7.448694452587866e-05, "loss": 0.0352, "step": 29620 }, { "grad_norm": 0.505470335483551, "learning_rate": 7.44689223177091e-05, "loss": 0.0465, "step": 29630 }, { "grad_norm": 0.3909023106098175, "learning_rate": 7.445089592846994e-05, "loss": 0.0404, "step": 29640 }, { "grad_norm": 0.36452916264533997, "learning_rate": 7.443286536124141e-05, "loss": 0.0363, "step": 29650 }, { "grad_norm": 0.386944442987442, "learning_rate": 7.441483061910443e-05, "loss": 0.0418, "step": 29660 }, { "grad_norm": 0.3513111174106598, "learning_rate": 7.439679170514064e-05, "loss": 0.0369, "step": 29670 }, { "grad_norm": 0.3793705403804779, "learning_rate": 7.43787486224324e-05, "loss": 0.0433, "step": 29680 }, { "grad_norm": 0.25998595356941223, "learning_rate": 7.436070137406276e-05, "loss": 0.0338, "step": 29690 }, { "grad_norm": 0.412728488445282, "learning_rate": 7.434264996311556e-05, "loss": 0.0316, "step": 29700 }, { "grad_norm": 0.44448867440223694, "learning_rate": 7.432459439267525e-05, "loss": 0.0376, "step": 29710 }, { "grad_norm": 0.3361751139163971, "learning_rate": 7.430653466582701e-05, "loss": 0.0429, "step": 29720 }, { "grad_norm": 0.4603467285633087, "learning_rate": 7.42884707856568e-05, "loss": 0.0507, "step": 29730 }, { "grad_norm": 0.3202536106109619, "learning_rate": 7.427040275525122e-05, "loss": 0.0349, "step": 29740 }, { "grad_norm": 0.3036775588989258, "learning_rate": 7.42523305776976e-05, "loss": 0.0416, "step": 29750 }, { "grad_norm": 0.33258256316185, "learning_rate": 7.4234254256084e-05, "loss": 0.0381, "step": 29760 }, { "grad_norm": 0.5415392518043518, "learning_rate": 7.421617379349915e-05, "loss": 0.0483, "step": 29770 }, { "grad_norm": 0.31616589426994324, "learning_rate": 7.41980891930325e-05, "loss": 0.0381, "step": 29780 }, { "grad_norm": 0.4960305988788605, "learning_rate": 7.418000045777425e-05, "loss": 0.0315, "step": 29790 }, { "grad_norm": 0.31241118907928467, "learning_rate": 7.416190759081523e-05, "loss": 0.0342, "step": 29800 }, { "grad_norm": 0.36948975920677185, "learning_rate": 7.414381059524704e-05, "loss": 0.0411, "step": 29810 }, { "grad_norm": 0.5239475965499878, "learning_rate": 7.412570947416195e-05, "loss": 0.0309, "step": 29820 }, { "grad_norm": 0.608345091342926, "learning_rate": 7.410760423065295e-05, "loss": 0.0509, "step": 29830 }, { "grad_norm": 0.3842405378818512, "learning_rate": 7.408949486781372e-05, "loss": 0.0556, "step": 29840 }, { "grad_norm": 0.3680179715156555, "learning_rate": 7.407138138873868e-05, "loss": 0.0428, "step": 29850 }, { "grad_norm": 0.4242091476917267, "learning_rate": 7.405326379652292e-05, "loss": 0.0426, "step": 29860 }, { "grad_norm": 0.36455950140953064, "learning_rate": 7.403514209426222e-05, "loss": 0.0394, "step": 29870 }, { "grad_norm": 0.4437798857688904, "learning_rate": 7.40170162850531e-05, "loss": 0.0402, "step": 29880 }, { "grad_norm": 0.31920093297958374, "learning_rate": 7.399888637199278e-05, "loss": 0.0351, "step": 29890 }, { "grad_norm": 0.6063499450683594, "learning_rate": 7.398075235817914e-05, "loss": 0.0553, "step": 29900 }, { "grad_norm": 0.45597758889198303, "learning_rate": 7.39626142467108e-05, "loss": 0.0404, "step": 29910 }, { "grad_norm": 0.4241149127483368, "learning_rate": 7.394447204068706e-05, "loss": 0.0397, "step": 29920 }, { "grad_norm": 0.2949988543987274, "learning_rate": 7.392632574320793e-05, "loss": 0.0425, "step": 29930 }, { "grad_norm": 0.47669661045074463, "learning_rate": 7.390817535737411e-05, "loss": 0.0408, "step": 29940 }, { "grad_norm": 0.4402061402797699, "learning_rate": 7.389002088628703e-05, "loss": 0.0453, "step": 29950 }, { "grad_norm": 0.3085498511791229, "learning_rate": 7.387186233304877e-05, "loss": 0.0469, "step": 29960 }, { "grad_norm": 0.38491135835647583, "learning_rate": 7.385369970076212e-05, "loss": 0.0392, "step": 29970 }, { "grad_norm": 0.39261701703071594, "learning_rate": 7.38355329925306e-05, "loss": 0.0408, "step": 29980 }, { "grad_norm": 0.4498954117298126, "learning_rate": 7.381736221145838e-05, "loss": 0.0443, "step": 29990 }, { "grad_norm": 0.5229883790016174, "learning_rate": 7.37991873606504e-05, "loss": 0.0378, "step": 30000 }, { "grad_norm": 0.47697827219963074, "learning_rate": 7.378100844321218e-05, "loss": 0.0506, "step": 30010 }, { "grad_norm": 0.36108434200286865, "learning_rate": 7.376282546225004e-05, "loss": 0.0373, "step": 30020 }, { "grad_norm": 0.822694718837738, "learning_rate": 7.374463842087094e-05, "loss": 0.0317, "step": 30030 }, { "grad_norm": 0.3957127332687378, "learning_rate": 7.372644732218254e-05, "loss": 0.038, "step": 30040 }, { "grad_norm": 0.7130834460258484, "learning_rate": 7.370825216929322e-05, "loss": 0.0562, "step": 30050 }, { "grad_norm": 0.3458322286605835, "learning_rate": 7.369005296531205e-05, "loss": 0.0439, "step": 30060 }, { "grad_norm": 0.4313622713088989, "learning_rate": 7.367184971334873e-05, "loss": 0.0452, "step": 30070 }, { "grad_norm": 0.2871253192424774, "learning_rate": 7.365364241651371e-05, "loss": 0.0448, "step": 30080 }, { "grad_norm": 0.3819494843482971, "learning_rate": 7.363543107791815e-05, "loss": 0.0292, "step": 30090 }, { "grad_norm": 0.3657209277153015, "learning_rate": 7.361721570067384e-05, "loss": 0.0384, "step": 30100 }, { "grad_norm": 0.3728867173194885, "learning_rate": 7.359899628789331e-05, "loss": 0.0456, "step": 30110 }, { "grad_norm": 0.43867236375808716, "learning_rate": 7.358077284268974e-05, "loss": 0.0439, "step": 30120 }, { "grad_norm": 0.37760093808174133, "learning_rate": 7.356254536817702e-05, "loss": 0.0413, "step": 30130 }, { "grad_norm": 0.35309991240501404, "learning_rate": 7.354431386746973e-05, "loss": 0.0389, "step": 30140 }, { "grad_norm": 0.45194560289382935, "learning_rate": 7.352607834368316e-05, "loss": 0.0358, "step": 30150 }, { "grad_norm": 0.5147750973701477, "learning_rate": 7.350783879993324e-05, "loss": 0.0352, "step": 30160 }, { "grad_norm": 0.3538753390312195, "learning_rate": 7.348959523933658e-05, "loss": 0.0362, "step": 30170 }, { "grad_norm": 0.44009920954704285, "learning_rate": 7.347134766501057e-05, "loss": 0.0376, "step": 30180 }, { "grad_norm": 0.3574671447277069, "learning_rate": 7.345309608007315e-05, "loss": 0.0304, "step": 30190 }, { "grad_norm": 0.38922709226608276, "learning_rate": 7.343484048764308e-05, "loss": 0.0386, "step": 30200 }, { "grad_norm": 0.42078712582588196, "learning_rate": 7.341658089083972e-05, "loss": 0.0426, "step": 30210 }, { "grad_norm": 0.41991642117500305, "learning_rate": 7.339831729278313e-05, "loss": 0.0302, "step": 30220 }, { "grad_norm": 0.40675896406173706, "learning_rate": 7.338004969659404e-05, "loss": 0.0385, "step": 30230 }, { "grad_norm": 0.32117998600006104, "learning_rate": 7.336177810539391e-05, "loss": 0.0406, "step": 30240 }, { "grad_norm": 0.3436444401741028, "learning_rate": 7.334350252230485e-05, "loss": 0.0385, "step": 30250 }, { "grad_norm": 0.31102120876312256, "learning_rate": 7.332522295044965e-05, "loss": 0.0323, "step": 30260 }, { "grad_norm": 0.4228072762489319, "learning_rate": 7.33069393929518e-05, "loss": 0.0445, "step": 30270 }, { "grad_norm": 0.39836812019348145, "learning_rate": 7.328865185293545e-05, "loss": 0.0485, "step": 30280 }, { "grad_norm": 0.5268368124961853, "learning_rate": 7.327036033352546e-05, "loss": 0.0419, "step": 30290 }, { "grad_norm": 0.3873426914215088, "learning_rate": 7.325206483784733e-05, "loss": 0.0334, "step": 30300 }, { "grad_norm": 0.499878853559494, "learning_rate": 7.323376536902724e-05, "loss": 0.0505, "step": 30310 }, { "grad_norm": 0.40389126539230347, "learning_rate": 7.321546193019213e-05, "loss": 0.047, "step": 30320 }, { "grad_norm": 0.4478541910648346, "learning_rate": 7.31971545244695e-05, "loss": 0.0428, "step": 30330 }, { "grad_norm": 0.4566628634929657, "learning_rate": 7.31788431549876e-05, "loss": 0.0408, "step": 30340 }, { "grad_norm": 0.4176470637321472, "learning_rate": 7.316052782487534e-05, "loss": 0.0357, "step": 30350 }, { "grad_norm": 0.9210534691810608, "learning_rate": 7.314220853726234e-05, "loss": 0.0452, "step": 30360 }, { "grad_norm": 0.43832048773765564, "learning_rate": 7.312388529527884e-05, "loss": 0.0423, "step": 30370 }, { "grad_norm": 0.39308568835258484, "learning_rate": 7.310555810205577e-05, "loss": 0.0448, "step": 30380 }, { "grad_norm": 0.34652847051620483, "learning_rate": 7.308722696072476e-05, "loss": 0.051, "step": 30390 }, { "grad_norm": 0.3791120648384094, "learning_rate": 7.306889187441811e-05, "loss": 0.0402, "step": 30400 }, { "grad_norm": 0.2791789770126343, "learning_rate": 7.305055284626876e-05, "loss": 0.0349, "step": 30410 }, { "grad_norm": 0.31909990310668945, "learning_rate": 7.303220987941037e-05, "loss": 0.0445, "step": 30420 }, { "grad_norm": 0.35841384530067444, "learning_rate": 7.301386297697726e-05, "loss": 0.0399, "step": 30430 }, { "grad_norm": 0.3933398425579071, "learning_rate": 7.299551214210438e-05, "loss": 0.0368, "step": 30440 }, { "grad_norm": 0.2915961444377899, "learning_rate": 7.297715737792738e-05, "loss": 0.0362, "step": 30450 }, { "grad_norm": 0.30026280879974365, "learning_rate": 7.295879868758265e-05, "loss": 0.0336, "step": 30460 }, { "grad_norm": 0.5805015563964844, "learning_rate": 7.294043607420713e-05, "loss": 0.0422, "step": 30470 }, { "grad_norm": 0.2055945098400116, "learning_rate": 7.292206954093852e-05, "loss": 0.0345, "step": 30480 }, { "grad_norm": 0.41511785984039307, "learning_rate": 7.290369909091515e-05, "loss": 0.0358, "step": 30490 }, { "grad_norm": 0.3091619312763214, "learning_rate": 7.2885324727276e-05, "loss": 0.0408, "step": 30500 }, { "grad_norm": 0.3037759065628052, "learning_rate": 7.286694645316076e-05, "loss": 0.0377, "step": 30510 }, { "grad_norm": 0.3713877201080322, "learning_rate": 7.284856427170982e-05, "loss": 0.0364, "step": 30520 }, { "grad_norm": 0.3330424129962921, "learning_rate": 7.283017818606414e-05, "loss": 0.0393, "step": 30530 }, { "grad_norm": 0.3898671567440033, "learning_rate": 7.28117881993654e-05, "loss": 0.0351, "step": 30540 }, { "grad_norm": 0.2734774053096771, "learning_rate": 7.279339431475598e-05, "loss": 0.0331, "step": 30550 }, { "grad_norm": 0.25155434012413025, "learning_rate": 7.277499653537887e-05, "loss": 0.0322, "step": 30560 }, { "grad_norm": 0.33362409472465515, "learning_rate": 7.275659486437776e-05, "loss": 0.0428, "step": 30570 }, { "grad_norm": 0.2858913540840149, "learning_rate": 7.273818930489695e-05, "loss": 0.0468, "step": 30580 }, { "grad_norm": 0.2993853688240051, "learning_rate": 7.271977986008151e-05, "loss": 0.043, "step": 30590 }, { "grad_norm": 0.33437085151672363, "learning_rate": 7.270136653307705e-05, "loss": 0.0335, "step": 30600 }, { "grad_norm": 0.37277379631996155, "learning_rate": 7.268294932702994e-05, "loss": 0.0389, "step": 30610 }, { "grad_norm": 1.063772439956665, "learning_rate": 7.266452824508719e-05, "loss": 0.0432, "step": 30620 }, { "grad_norm": 0.3516625165939331, "learning_rate": 7.264610329039643e-05, "loss": 0.04, "step": 30630 }, { "grad_norm": 0.27955055236816406, "learning_rate": 7.262767446610599e-05, "loss": 0.0417, "step": 30640 }, { "grad_norm": 0.4510030448436737, "learning_rate": 7.260924177536485e-05, "loss": 0.0464, "step": 30650 }, { "grad_norm": 0.4952462911605835, "learning_rate": 7.259080522132265e-05, "loss": 0.0411, "step": 30660 }, { "grad_norm": 0.43045032024383545, "learning_rate": 7.257236480712972e-05, "loss": 0.039, "step": 30670 }, { "grad_norm": 0.35407769680023193, "learning_rate": 7.255392053593697e-05, "loss": 0.0347, "step": 30680 }, { "grad_norm": 0.3668380677700043, "learning_rate": 7.253547241089607e-05, "loss": 0.0403, "step": 30690 }, { "grad_norm": 0.3720569908618927, "learning_rate": 7.251702043515927e-05, "loss": 0.0382, "step": 30700 }, { "grad_norm": 0.2586430609226227, "learning_rate": 7.249856461187952e-05, "loss": 0.0363, "step": 30710 }, { "grad_norm": 0.43367746472358704, "learning_rate": 7.248010494421042e-05, "loss": 0.0442, "step": 30720 }, { "grad_norm": 0.3281877636909485, "learning_rate": 7.246164143530622e-05, "loss": 0.0433, "step": 30730 }, { "grad_norm": 0.49445194005966187, "learning_rate": 7.244317408832181e-05, "loss": 0.0382, "step": 30740 }, { "grad_norm": 0.39398354291915894, "learning_rate": 7.242470290641279e-05, "loss": 0.0364, "step": 30750 }, { "grad_norm": 0.26538756489753723, "learning_rate": 7.240622789273536e-05, "loss": 0.0387, "step": 30760 }, { "grad_norm": 0.40901219844818115, "learning_rate": 7.238774905044638e-05, "loss": 0.0358, "step": 30770 }, { "grad_norm": 0.3927267789840698, "learning_rate": 7.236926638270341e-05, "loss": 0.0401, "step": 30780 }, { "grad_norm": 0.34331247210502625, "learning_rate": 7.23507798926646e-05, "loss": 0.0317, "step": 30790 }, { "grad_norm": 0.642281711101532, "learning_rate": 7.23322895834888e-05, "loss": 0.0406, "step": 30800 }, { "grad_norm": 0.32781779766082764, "learning_rate": 7.231379545833552e-05, "loss": 0.0345, "step": 30810 }, { "grad_norm": 0.3144664168357849, "learning_rate": 7.229529752036487e-05, "loss": 0.0406, "step": 30820 }, { "grad_norm": 0.30724719166755676, "learning_rate": 7.227679577273765e-05, "loss": 0.0426, "step": 30830 }, { "grad_norm": 0.42650073766708374, "learning_rate": 7.225829021861529e-05, "loss": 0.0418, "step": 30840 }, { "grad_norm": 0.2798592448234558, "learning_rate": 7.223978086115992e-05, "loss": 0.0305, "step": 30850 }, { "grad_norm": 0.4433861970901489, "learning_rate": 7.222126770353425e-05, "loss": 0.047, "step": 30860 }, { "grad_norm": 0.34812673926353455, "learning_rate": 7.22027507489017e-05, "loss": 0.0348, "step": 30870 }, { "grad_norm": 0.34606122970581055, "learning_rate": 7.218423000042627e-05, "loss": 0.0399, "step": 30880 }, { "grad_norm": 0.30389517545700073, "learning_rate": 7.216570546127268e-05, "loss": 0.0401, "step": 30890 }, { "grad_norm": 0.459025502204895, "learning_rate": 7.214717713460626e-05, "loss": 0.0468, "step": 30900 }, { "grad_norm": 0.6530762910842896, "learning_rate": 7.2128645023593e-05, "loss": 0.0468, "step": 30910 }, { "grad_norm": 0.2986014485359192, "learning_rate": 7.211010913139951e-05, "loss": 0.0423, "step": 30920 }, { "grad_norm": 0.5235673189163208, "learning_rate": 7.209156946119308e-05, "loss": 0.0332, "step": 30930 }, { "grad_norm": 0.31240028142929077, "learning_rate": 7.207302601614166e-05, "loss": 0.0367, "step": 30940 }, { "grad_norm": 0.2778224050998688, "learning_rate": 7.205447879941378e-05, "loss": 0.0409, "step": 30950 }, { "grad_norm": 0.3599897027015686, "learning_rate": 7.203592781417866e-05, "loss": 0.0365, "step": 30960 }, { "grad_norm": 0.40806928277015686, "learning_rate": 7.201737306360617e-05, "loss": 0.039, "step": 30970 }, { "grad_norm": 0.3594401180744171, "learning_rate": 7.19988145508668e-05, "loss": 0.039, "step": 30980 }, { "grad_norm": 0.35286977887153625, "learning_rate": 7.198025227913168e-05, "loss": 0.0409, "step": 30990 }, { "grad_norm": 0.5012198090553284, "learning_rate": 7.196168625157261e-05, "loss": 0.0385, "step": 31000 }, { "grad_norm": 0.3404976427555084, "learning_rate": 7.194311647136201e-05, "loss": 0.0368, "step": 31010 }, { "grad_norm": 0.3309280574321747, "learning_rate": 7.192454294167297e-05, "loss": 0.0346, "step": 31020 }, { "grad_norm": 0.47411954402923584, "learning_rate": 7.190596566567917e-05, "loss": 0.037, "step": 31030 }, { "grad_norm": 0.5573288798332214, "learning_rate": 7.188738464655496e-05, "loss": 0.0367, "step": 31040 }, { "grad_norm": 0.42123842239379883, "learning_rate": 7.186879988747533e-05, "loss": 0.0284, "step": 31050 }, { "grad_norm": 0.2675393223762512, "learning_rate": 7.185021139161592e-05, "loss": 0.0244, "step": 31060 }, { "grad_norm": 0.2995927333831787, "learning_rate": 7.1831619162153e-05, "loss": 0.031, "step": 31070 }, { "grad_norm": 0.462922602891922, "learning_rate": 7.181302320226345e-05, "loss": 0.0361, "step": 31080 }, { "grad_norm": 0.2751423418521881, "learning_rate": 7.179442351512482e-05, "loss": 0.0472, "step": 31090 }, { "grad_norm": 0.33119866251945496, "learning_rate": 7.177582010391528e-05, "loss": 0.0366, "step": 31100 }, { "grad_norm": 0.46989795565605164, "learning_rate": 7.175721297181366e-05, "loss": 0.0428, "step": 31110 }, { "grad_norm": 0.47106459736824036, "learning_rate": 7.173860212199942e-05, "loss": 0.0365, "step": 31120 }, { "grad_norm": 0.38083258271217346, "learning_rate": 7.171998755765263e-05, "loss": 0.037, "step": 31130 }, { "grad_norm": 0.36752596497535706, "learning_rate": 7.170136928195398e-05, "loss": 0.0377, "step": 31140 }, { "grad_norm": 0.42133426666259766, "learning_rate": 7.168274729808489e-05, "loss": 0.039, "step": 31150 }, { "grad_norm": 0.3094574511051178, "learning_rate": 7.166412160922728e-05, "loss": 0.0399, "step": 31160 }, { "grad_norm": 0.54563307762146, "learning_rate": 7.164549221856382e-05, "loss": 0.0403, "step": 31170 }, { "grad_norm": 0.41468968987464905, "learning_rate": 7.162685912927775e-05, "loss": 0.0604, "step": 31180 }, { "grad_norm": 0.6454212069511414, "learning_rate": 7.160822234455294e-05, "loss": 0.0427, "step": 31190 }, { "grad_norm": 0.3706059157848358, "learning_rate": 7.158958186757391e-05, "loss": 0.0341, "step": 31200 }, { "grad_norm": 0.4547070562839508, "learning_rate": 7.157093770152582e-05, "loss": 0.0386, "step": 31210 }, { "grad_norm": 0.37033846974372864, "learning_rate": 7.155228984959446e-05, "loss": 0.0329, "step": 31220 }, { "grad_norm": 0.3293616473674774, "learning_rate": 7.153363831496621e-05, "loss": 0.0406, "step": 31230 }, { "grad_norm": 0.46528157591819763, "learning_rate": 7.151498310082811e-05, "loss": 0.0495, "step": 31240 }, { "grad_norm": 0.4070849120616913, "learning_rate": 7.149632421036784e-05, "loss": 0.0429, "step": 31250 }, { "grad_norm": 0.6138168573379517, "learning_rate": 7.147766164677369e-05, "loss": 0.0578, "step": 31260 }, { "grad_norm": 0.25162258744239807, "learning_rate": 7.145899541323459e-05, "loss": 0.0288, "step": 31270 }, { "grad_norm": 0.5309728384017944, "learning_rate": 7.144032551294007e-05, "loss": 0.0503, "step": 31280 }, { "grad_norm": 0.4348769783973694, "learning_rate": 7.14216519490803e-05, "loss": 0.0443, "step": 31290 }, { "grad_norm": 0.4870973229408264, "learning_rate": 7.140297472484609e-05, "loss": 0.0495, "step": 31300 }, { "grad_norm": 0.41867971420288086, "learning_rate": 7.138429384342891e-05, "loss": 0.053, "step": 31310 }, { "grad_norm": 0.2750757336616516, "learning_rate": 7.136560930802074e-05, "loss": 0.0376, "step": 31320 }, { "grad_norm": 0.4237576723098755, "learning_rate": 7.134692112181431e-05, "loss": 0.0325, "step": 31330 }, { "grad_norm": 0.4021366536617279, "learning_rate": 7.13282292880029e-05, "loss": 0.0364, "step": 31340 }, { "grad_norm": 0.21253930032253265, "learning_rate": 7.130953380978043e-05, "loss": 0.0351, "step": 31350 }, { "grad_norm": 0.7102525234222412, "learning_rate": 7.129083469034144e-05, "loss": 0.0567, "step": 31360 }, { "grad_norm": 0.47470974922180176, "learning_rate": 7.127213193288112e-05, "loss": 0.0493, "step": 31370 }, { "grad_norm": 0.402465283870697, "learning_rate": 7.125342554059522e-05, "loss": 0.0383, "step": 31380 }, { "grad_norm": 0.43541643023490906, "learning_rate": 7.12347155166802e-05, "loss": 0.0464, "step": 31390 }, { "grad_norm": 0.24474115669727325, "learning_rate": 7.121600186433306e-05, "loss": 0.0346, "step": 31400 }, { "grad_norm": 0.40275683999061584, "learning_rate": 7.119728458675148e-05, "loss": 0.0429, "step": 31410 }, { "grad_norm": 0.38207656145095825, "learning_rate": 7.117856368713369e-05, "loss": 0.0349, "step": 31420 }, { "grad_norm": 0.5730084776878357, "learning_rate": 7.115983916867861e-05, "loss": 0.0404, "step": 31430 }, { "grad_norm": 0.3434499204158783, "learning_rate": 7.114111103458574e-05, "loss": 0.0356, "step": 31440 }, { "grad_norm": 0.4180084764957428, "learning_rate": 7.11223792880552e-05, "loss": 0.0404, "step": 31450 }, { "grad_norm": 0.3090139329433441, "learning_rate": 7.110364393228773e-05, "loss": 0.037, "step": 31460 }, { "grad_norm": 0.41091230511665344, "learning_rate": 7.108490497048471e-05, "loss": 0.037, "step": 31470 }, { "grad_norm": 0.36829090118408203, "learning_rate": 7.10661624058481e-05, "loss": 0.0441, "step": 31480 }, { "grad_norm": 0.23653389513492584, "learning_rate": 7.10474162415805e-05, "loss": 0.0372, "step": 31490 }, { "grad_norm": 0.3568721413612366, "learning_rate": 7.102866648088511e-05, "loss": 0.0395, "step": 31500 }, { "grad_norm": 0.3689764440059662, "learning_rate": 7.100991312696576e-05, "loss": 0.0392, "step": 31510 }, { "grad_norm": 0.36415085196495056, "learning_rate": 7.099115618302686e-05, "loss": 0.0374, "step": 31520 }, { "grad_norm": 0.4962669014930725, "learning_rate": 7.097239565227349e-05, "loss": 0.0543, "step": 31530 }, { "grad_norm": 0.35356372594833374, "learning_rate": 7.09536315379113e-05, "loss": 0.0395, "step": 31540 }, { "grad_norm": 0.31781700253486633, "learning_rate": 7.093486384314656e-05, "loss": 0.0365, "step": 31550 }, { "grad_norm": 0.3736351430416107, "learning_rate": 7.091609257118616e-05, "loss": 0.0418, "step": 31560 }, { "grad_norm": 0.41627514362335205, "learning_rate": 7.08973177252376e-05, "loss": 0.0432, "step": 31570 }, { "grad_norm": 0.2827978730201721, "learning_rate": 7.087853930850898e-05, "loss": 0.0378, "step": 31580 }, { "grad_norm": 0.3262895941734314, "learning_rate": 7.085975732420903e-05, "loss": 0.0363, "step": 31590 }, { "grad_norm": 0.3555229902267456, "learning_rate": 7.084097177554706e-05, "loss": 0.0268, "step": 31600 }, { "grad_norm": 1.1012325286865234, "learning_rate": 7.082218266573301e-05, "loss": 0.0429, "step": 31610 }, { "grad_norm": 0.5203496217727661, "learning_rate": 7.080338999797743e-05, "loss": 0.0455, "step": 31620 }, { "grad_norm": 0.3969353139400482, "learning_rate": 7.07845937754915e-05, "loss": 0.0397, "step": 31630 }, { "grad_norm": 0.3387327492237091, "learning_rate": 7.076579400148693e-05, "loss": 0.0402, "step": 31640 }, { "grad_norm": 0.5363636612892151, "learning_rate": 7.074699067917611e-05, "loss": 0.0488, "step": 31650 }, { "grad_norm": 0.31525057554244995, "learning_rate": 7.072818381177201e-05, "loss": 0.0453, "step": 31660 }, { "grad_norm": 0.4885159730911255, "learning_rate": 7.070937340248823e-05, "loss": 0.0461, "step": 31670 }, { "grad_norm": 0.43016964197158813, "learning_rate": 7.069055945453893e-05, "loss": 0.0534, "step": 31680 }, { "grad_norm": 0.45866522192955017, "learning_rate": 7.067174197113892e-05, "loss": 0.0423, "step": 31690 }, { "grad_norm": 0.29000943899154663, "learning_rate": 7.065292095550355e-05, "loss": 0.0351, "step": 31700 }, { "grad_norm": 0.35033664107322693, "learning_rate": 7.063409641084887e-05, "loss": 0.0346, "step": 31710 }, { "grad_norm": 0.3945915400981903, "learning_rate": 7.061526834039145e-05, "loss": 0.0518, "step": 31720 }, { "grad_norm": 0.3244918882846832, "learning_rate": 7.05964367473485e-05, "loss": 0.0364, "step": 31730 }, { "grad_norm": 0.4738881587982178, "learning_rate": 7.057760163493783e-05, "loss": 0.0432, "step": 31740 }, { "grad_norm": 0.3403174579143524, "learning_rate": 7.055876300637783e-05, "loss": 0.0311, "step": 31750 }, { "grad_norm": 0.4599550664424896, "learning_rate": 7.053992086488753e-05, "loss": 0.047, "step": 31760 }, { "grad_norm": 0.23456673324108124, "learning_rate": 7.052107521368651e-05, "loss": 0.0472, "step": 31770 }, { "grad_norm": 0.35974499583244324, "learning_rate": 7.0502226055995e-05, "loss": 0.0418, "step": 31780 }, { "grad_norm": 0.2858923077583313, "learning_rate": 7.048337339503379e-05, "loss": 0.0514, "step": 31790 }, { "grad_norm": 0.7331211566925049, "learning_rate": 7.046451723402427e-05, "loss": 0.0469, "step": 31800 }, { "grad_norm": 0.38016682863235474, "learning_rate": 7.044565757618848e-05, "loss": 0.0511, "step": 31810 }, { "grad_norm": 0.355712354183197, "learning_rate": 7.042679442474899e-05, "loss": 0.0442, "step": 31820 }, { "grad_norm": 0.48870041966438293, "learning_rate": 7.040792778292902e-05, "loss": 0.0552, "step": 31830 }, { "grad_norm": 2.0208022594451904, "learning_rate": 7.038905765395234e-05, "loss": 0.0504, "step": 31840 }, { "grad_norm": 0.46737149357795715, "learning_rate": 7.037018404104334e-05, "loss": 0.0458, "step": 31850 }, { "grad_norm": 0.4424101710319519, "learning_rate": 7.035130694742702e-05, "loss": 0.0324, "step": 31860 }, { "grad_norm": 0.47655874490737915, "learning_rate": 7.033242637632897e-05, "loss": 0.0425, "step": 31870 }, { "grad_norm": 0.42147210240364075, "learning_rate": 7.031354233097534e-05, "loss": 0.0419, "step": 31880 }, { "grad_norm": 0.37526392936706543, "learning_rate": 7.029465481459289e-05, "loss": 0.0493, "step": 31890 }, { "grad_norm": 0.39129045605659485, "learning_rate": 7.027576383040898e-05, "loss": 0.0367, "step": 31900 }, { "grad_norm": 0.22589905560016632, "learning_rate": 7.025686938165159e-05, "loss": 0.0474, "step": 31910 }, { "grad_norm": 0.5676992535591125, "learning_rate": 7.023797147154924e-05, "loss": 0.062, "step": 31920 }, { "grad_norm": 0.4577309787273407, "learning_rate": 7.021907010333111e-05, "loss": 0.0374, "step": 31930 }, { "grad_norm": 0.4462897777557373, "learning_rate": 7.020016528022685e-05, "loss": 0.0421, "step": 31940 }, { "grad_norm": 0.43119651079177856, "learning_rate": 7.018125700546683e-05, "loss": 0.0433, "step": 31950 }, { "grad_norm": 0.4319450557231903, "learning_rate": 7.016234528228196e-05, "loss": 0.0457, "step": 31960 }, { "grad_norm": 0.3672531247138977, "learning_rate": 7.014343011390372e-05, "loss": 0.038, "step": 31970 }, { "grad_norm": 0.38375124335289, "learning_rate": 7.01245115035642e-05, "loss": 0.0334, "step": 31980 }, { "grad_norm": 0.2588622272014618, "learning_rate": 7.010558945449606e-05, "loss": 0.0484, "step": 31990 }, { "grad_norm": 0.3954734206199646, "learning_rate": 7.008666396993258e-05, "loss": 0.034, "step": 32000 }, { "grad_norm": 0.4380244314670563, "learning_rate": 7.006773505310759e-05, "loss": 0.0477, "step": 32010 }, { "grad_norm": 0.5150413513183594, "learning_rate": 7.004880270725553e-05, "loss": 0.0376, "step": 32020 }, { "grad_norm": 0.39612966775894165, "learning_rate": 7.002986693561144e-05, "loss": 0.0398, "step": 32030 }, { "grad_norm": 0.3690773844718933, "learning_rate": 7.001092774141089e-05, "loss": 0.0325, "step": 32040 }, { "grad_norm": 0.31439918279647827, "learning_rate": 6.999198512789009e-05, "loss": 0.0303, "step": 32050 }, { "grad_norm": 0.4036528468132019, "learning_rate": 6.997303909828584e-05, "loss": 0.0389, "step": 32060 }, { "grad_norm": 0.4763999581336975, "learning_rate": 6.995408965583544e-05, "loss": 0.0502, "step": 32070 }, { "grad_norm": 0.30422523617744446, "learning_rate": 6.993513680377688e-05, "loss": 0.0317, "step": 32080 }, { "grad_norm": 0.21629348397254944, "learning_rate": 6.991618054534868e-05, "loss": 0.0394, "step": 32090 }, { "grad_norm": 0.3435465395450592, "learning_rate": 6.989722088378991e-05, "loss": 0.0349, "step": 32100 }, { "grad_norm": 0.4410245716571808, "learning_rate": 6.987825782234027e-05, "loss": 0.0361, "step": 32110 }, { "grad_norm": 0.34679660201072693, "learning_rate": 6.985929136424006e-05, "loss": 0.0354, "step": 32120 }, { "grad_norm": 0.3450935184955597, "learning_rate": 6.984032151273012e-05, "loss": 0.0333, "step": 32130 }, { "grad_norm": 0.3071792423725128, "learning_rate": 6.982134827105186e-05, "loss": 0.041, "step": 32140 }, { "grad_norm": 0.4824727177619934, "learning_rate": 6.980237164244729e-05, "loss": 0.0357, "step": 32150 }, { "grad_norm": 0.4314482510089874, "learning_rate": 6.9783391630159e-05, "loss": 0.0343, "step": 32160 }, { "grad_norm": 0.2749071717262268, "learning_rate": 6.976440823743015e-05, "loss": 0.0352, "step": 32170 }, { "grad_norm": 0.48575645685195923, "learning_rate": 6.974542146750451e-05, "loss": 0.0335, "step": 32180 }, { "grad_norm": 0.3164687752723694, "learning_rate": 6.972643132362637e-05, "loss": 0.0391, "step": 32190 }, { "grad_norm": 0.3370152413845062, "learning_rate": 6.970743780904064e-05, "loss": 0.034, "step": 32200 }, { "grad_norm": 0.41407766938209534, "learning_rate": 6.968844092699277e-05, "loss": 0.0391, "step": 32210 }, { "grad_norm": 0.39851999282836914, "learning_rate": 6.966944068072883e-05, "loss": 0.0375, "step": 32220 }, { "grad_norm": 0.3520883321762085, "learning_rate": 6.965043707349545e-05, "loss": 0.0345, "step": 32230 }, { "grad_norm": 0.4748893082141876, "learning_rate": 6.963143010853982e-05, "loss": 0.041, "step": 32240 }, { "grad_norm": 0.47049421072006226, "learning_rate": 6.961241978910971e-05, "loss": 0.047, "step": 32250 }, { "grad_norm": 0.293062299489975, "learning_rate": 6.959340611845344e-05, "loss": 0.0395, "step": 32260 }, { "grad_norm": 0.4330907464027405, "learning_rate": 6.957438909981995e-05, "loss": 0.0325, "step": 32270 }, { "grad_norm": 0.3438727557659149, "learning_rate": 6.955536873645872e-05, "loss": 0.0299, "step": 32280 }, { "grad_norm": 0.36830899119377136, "learning_rate": 6.953634503161982e-05, "loss": 0.0474, "step": 32290 }, { "grad_norm": 0.39001449942588806, "learning_rate": 6.951731798855387e-05, "loss": 0.0351, "step": 32300 }, { "grad_norm": 0.2580878436565399, "learning_rate": 6.949828761051208e-05, "loss": 0.0393, "step": 32310 }, { "grad_norm": 0.3422410190105438, "learning_rate": 6.947925390074622e-05, "loss": 0.0339, "step": 32320 }, { "grad_norm": 0.48531410098075867, "learning_rate": 6.946021686250863e-05, "loss": 0.0386, "step": 32330 }, { "grad_norm": 0.7806640267372131, "learning_rate": 6.94411764990522e-05, "loss": 0.041, "step": 32340 }, { "grad_norm": 0.35767585039138794, "learning_rate": 6.942213281363044e-05, "loss": 0.0363, "step": 32350 }, { "grad_norm": 0.3631153404712677, "learning_rate": 6.940308580949737e-05, "loss": 0.0402, "step": 32360 }, { "grad_norm": 0.3991076946258545, "learning_rate": 6.93840354899076e-05, "loss": 0.0354, "step": 32370 }, { "grad_norm": 0.3762581944465637, "learning_rate": 6.936498185811633e-05, "loss": 0.0416, "step": 32380 }, { "grad_norm": 0.2386532723903656, "learning_rate": 6.93459249173793e-05, "loss": 0.0318, "step": 32390 }, { "grad_norm": 0.7004915475845337, "learning_rate": 6.932686467095279e-05, "loss": 0.0394, "step": 32400 }, { "grad_norm": 0.42834755778312683, "learning_rate": 6.930780112209373e-05, "loss": 0.0372, "step": 32410 }, { "grad_norm": 0.3264813721179962, "learning_rate": 6.92887342740595e-05, "loss": 0.0392, "step": 32420 }, { "grad_norm": 0.45966076850891113, "learning_rate": 6.926966413010816e-05, "loss": 0.049, "step": 32430 }, { "grad_norm": 0.24655377864837646, "learning_rate": 6.925059069349824e-05, "loss": 0.03, "step": 32440 }, { "grad_norm": 0.29012617468833923, "learning_rate": 6.923151396748886e-05, "loss": 0.045, "step": 32450 }, { "grad_norm": 0.28648367524147034, "learning_rate": 6.921243395533974e-05, "loss": 0.0403, "step": 32460 }, { "grad_norm": 0.5354576110839844, "learning_rate": 6.919335066031109e-05, "loss": 0.0425, "step": 32470 }, { "grad_norm": 0.42251232266426086, "learning_rate": 6.917426408566379e-05, "loss": 0.0464, "step": 32480 }, { "grad_norm": 0.43508559465408325, "learning_rate": 6.915517423465916e-05, "loss": 0.0388, "step": 32490 }, { "grad_norm": 0.3969968259334564, "learning_rate": 6.913608111055914e-05, "loss": 0.034, "step": 32500 }, { "grad_norm": 0.3323821425437927, "learning_rate": 6.911698471662623e-05, "loss": 0.0351, "step": 32510 }, { "grad_norm": 0.4404945969581604, "learning_rate": 6.90978850561235e-05, "loss": 0.0318, "step": 32520 }, { "grad_norm": 0.4615349769592285, "learning_rate": 6.907878213231454e-05, "loss": 0.0305, "step": 32530 }, { "grad_norm": 0.43097731471061707, "learning_rate": 6.90596759484635e-05, "loss": 0.0392, "step": 32540 }, { "grad_norm": 0.5045691728591919, "learning_rate": 6.904056650783514e-05, "loss": 0.0438, "step": 32550 }, { "grad_norm": 0.2751624584197998, "learning_rate": 6.902145381369471e-05, "loss": 0.0387, "step": 32560 }, { "grad_norm": 0.33800646662712097, "learning_rate": 6.900233786930808e-05, "loss": 0.035, "step": 32570 }, { "grad_norm": 0.40548521280288696, "learning_rate": 6.898321867794161e-05, "loss": 0.0348, "step": 32580 }, { "grad_norm": 0.3340889811515808, "learning_rate": 6.896409624286226e-05, "loss": 0.0404, "step": 32590 }, { "grad_norm": 0.46721580624580383, "learning_rate": 6.894497056733754e-05, "loss": 0.0403, "step": 32600 }, { "grad_norm": 0.35704120993614197, "learning_rate": 6.89258416546355e-05, "loss": 0.0403, "step": 32610 }, { "grad_norm": 0.40137508511543274, "learning_rate": 6.890670950802474e-05, "loss": 0.0464, "step": 32620 }, { "grad_norm": 0.53504478931427, "learning_rate": 6.88875741307744e-05, "loss": 0.0359, "step": 32630 }, { "grad_norm": 0.3356524407863617, "learning_rate": 6.886843552615425e-05, "loss": 0.0391, "step": 32640 }, { "grad_norm": 0.19990281760692596, "learning_rate": 6.884929369743451e-05, "loss": 0.0392, "step": 32650 }, { "grad_norm": 0.39891454577445984, "learning_rate": 6.8830148647886e-05, "loss": 0.0401, "step": 32660 }, { "grad_norm": 0.3989163637161255, "learning_rate": 6.88110003807801e-05, "loss": 0.0481, "step": 32670 }, { "grad_norm": 0.41572293639183044, "learning_rate": 6.87918488993887e-05, "loss": 0.0347, "step": 32680 }, { "grad_norm": 0.48333266377449036, "learning_rate": 6.877269420698431e-05, "loss": 0.0439, "step": 32690 }, { "grad_norm": 0.4988410174846649, "learning_rate": 6.875353630683989e-05, "loss": 0.035, "step": 32700 }, { "grad_norm": 0.5203794836997986, "learning_rate": 6.873437520222905e-05, "loss": 0.0438, "step": 32710 }, { "grad_norm": 0.43748709559440613, "learning_rate": 6.871521089642585e-05, "loss": 0.0357, "step": 32720 }, { "grad_norm": 0.3314973711967468, "learning_rate": 6.869604339270498e-05, "loss": 0.0327, "step": 32730 }, { "grad_norm": 0.4135560989379883, "learning_rate": 6.867687269434164e-05, "loss": 0.0462, "step": 32740 }, { "grad_norm": 0.3194199800491333, "learning_rate": 6.865769880461156e-05, "loss": 0.0373, "step": 32750 }, { "grad_norm": 0.3205300271511078, "learning_rate": 6.863852172679104e-05, "loss": 0.0289, "step": 32760 }, { "grad_norm": 0.43700605630874634, "learning_rate": 6.861934146415693e-05, "loss": 0.0344, "step": 32770 }, { "grad_norm": 0.46242305636405945, "learning_rate": 6.86001580199866e-05, "loss": 0.0394, "step": 32780 }, { "grad_norm": 0.315108984708786, "learning_rate": 6.858097139755798e-05, "loss": 0.0324, "step": 32790 }, { "grad_norm": 0.3850724995136261, "learning_rate": 6.856178160014955e-05, "loss": 0.0389, "step": 32800 }, { "grad_norm": 0.4151507019996643, "learning_rate": 6.85425886310403e-05, "loss": 0.0345, "step": 32810 }, { "grad_norm": 0.26415935158729553, "learning_rate": 6.852339249350979e-05, "loss": 0.0377, "step": 32820 }, { "grad_norm": 0.2476998269557953, "learning_rate": 6.850419319083812e-05, "loss": 0.0401, "step": 32830 }, { "grad_norm": 0.6928474307060242, "learning_rate": 6.848499072630592e-05, "loss": 0.035, "step": 32840 }, { "grad_norm": 0.3688086271286011, "learning_rate": 6.846578510319439e-05, "loss": 0.0433, "step": 32850 }, { "grad_norm": 0.43446964025497437, "learning_rate": 6.844657632478519e-05, "loss": 0.0437, "step": 32860 }, { "grad_norm": 0.46275219321250916, "learning_rate": 6.842736439436063e-05, "loss": 0.0303, "step": 32870 }, { "grad_norm": 0.23473010957241058, "learning_rate": 6.84081493152035e-05, "loss": 0.0397, "step": 32880 }, { "grad_norm": 0.38005727529525757, "learning_rate": 6.83889310905971e-05, "loss": 0.0443, "step": 32890 }, { "grad_norm": 0.3190629184246063, "learning_rate": 6.836970972382533e-05, "loss": 0.0417, "step": 32900 }, { "grad_norm": 0.3158527910709381, "learning_rate": 6.835048521817257e-05, "loss": 0.0409, "step": 32910 }, { "grad_norm": 0.3662504553794861, "learning_rate": 6.833125757692379e-05, "loss": 0.038, "step": 32920 }, { "grad_norm": 0.40988248586654663, "learning_rate": 6.831202680336441e-05, "loss": 0.0341, "step": 32930 }, { "grad_norm": 0.5632420182228088, "learning_rate": 6.829279290078052e-05, "loss": 0.0343, "step": 32940 }, { "grad_norm": 0.3831706941127777, "learning_rate": 6.827355587245863e-05, "loss": 0.0373, "step": 32950 }, { "grad_norm": 0.24560265243053436, "learning_rate": 6.82543157216858e-05, "loss": 0.0326, "step": 32960 }, { "grad_norm": 0.27151262760162354, "learning_rate": 6.823507245174969e-05, "loss": 0.0351, "step": 32970 }, { "grad_norm": 0.3996759355068207, "learning_rate": 6.821582606593841e-05, "loss": 0.0293, "step": 32980 }, { "grad_norm": 0.49082425236701965, "learning_rate": 6.81965765675407e-05, "loss": 0.0299, "step": 32990 }, { "grad_norm": 0.3655385375022888, "learning_rate": 6.81773239598457e-05, "loss": 0.0336, "step": 33000 }, { "grad_norm": 0.34236636757850647, "learning_rate": 6.815806824614319e-05, "loss": 0.0397, "step": 33010 }, { "grad_norm": 0.4864516854286194, "learning_rate": 6.813880942972343e-05, "loss": 0.0289, "step": 33020 }, { "grad_norm": 0.4157164394855499, "learning_rate": 6.811954751387726e-05, "loss": 0.0365, "step": 33030 }, { "grad_norm": 0.31957927346229553, "learning_rate": 6.810028250189598e-05, "loss": 0.0309, "step": 33040 }, { "grad_norm": 0.4760971665382385, "learning_rate": 6.808101439707147e-05, "loss": 0.0516, "step": 33050 }, { "grad_norm": 0.375587522983551, "learning_rate": 6.806174320269609e-05, "loss": 0.0344, "step": 33060 }, { "grad_norm": 0.8078973293304443, "learning_rate": 6.804246892206281e-05, "loss": 0.0408, "step": 33070 }, { "grad_norm": 0.34135860204696655, "learning_rate": 6.802319155846506e-05, "loss": 0.0507, "step": 33080 }, { "grad_norm": 0.34089675545692444, "learning_rate": 6.800391111519679e-05, "loss": 0.0372, "step": 33090 }, { "grad_norm": 0.3384679853916168, "learning_rate": 6.798462759555253e-05, "loss": 0.0434, "step": 33100 }, { "grad_norm": 0.3952086269855499, "learning_rate": 6.79653410028273e-05, "loss": 0.0345, "step": 33110 }, { "grad_norm": 0.20377448201179504, "learning_rate": 6.794605134031663e-05, "loss": 0.0381, "step": 33120 }, { "grad_norm": 0.3485511839389801, "learning_rate": 6.792675861131661e-05, "loss": 0.0341, "step": 33130 }, { "grad_norm": 0.4555666148662567, "learning_rate": 6.790746281912386e-05, "loss": 0.0446, "step": 33140 }, { "grad_norm": 0.23464615643024445, "learning_rate": 6.788816396703546e-05, "loss": 0.0281, "step": 33150 }, { "grad_norm": 0.44563111662864685, "learning_rate": 6.78688620583491e-05, "loss": 0.032, "step": 33160 }, { "grad_norm": 0.521282434463501, "learning_rate": 6.784955709636292e-05, "loss": 0.0312, "step": 33170 }, { "grad_norm": 0.5344993472099304, "learning_rate": 6.783024908437564e-05, "loss": 0.0391, "step": 33180 }, { "grad_norm": 0.27696719765663147, "learning_rate": 6.781093802568641e-05, "loss": 0.034, "step": 33190 }, { "grad_norm": 0.4062991440296173, "learning_rate": 6.779162392359504e-05, "loss": 0.0346, "step": 33200 }, { "grad_norm": 0.451267808675766, "learning_rate": 6.777230678140172e-05, "loss": 0.0412, "step": 33210 }, { "grad_norm": 0.4215492904186249, "learning_rate": 6.775298660240726e-05, "loss": 0.0416, "step": 33220 }, { "grad_norm": 0.3410007655620575, "learning_rate": 6.773366338991292e-05, "loss": 0.0497, "step": 33230 }, { "grad_norm": 0.6450453400611877, "learning_rate": 6.771433714722052e-05, "loss": 0.037, "step": 33240 }, { "grad_norm": 0.3910854756832123, "learning_rate": 6.769500787763239e-05, "loss": 0.0424, "step": 33250 }, { "grad_norm": 0.30908435583114624, "learning_rate": 6.76756755844514e-05, "loss": 0.0451, "step": 33260 }, { "grad_norm": 0.3762715756893158, "learning_rate": 6.765634027098087e-05, "loss": 0.0376, "step": 33270 }, { "grad_norm": 0.40570002794265747, "learning_rate": 6.763700194052468e-05, "loss": 0.0469, "step": 33280 }, { "grad_norm": 0.43942052125930786, "learning_rate": 6.761766059638723e-05, "loss": 0.0392, "step": 33290 }, { "grad_norm": 0.4121146500110626, "learning_rate": 6.759831624187345e-05, "loss": 0.0393, "step": 33300 }, { "grad_norm": 0.34930676221847534, "learning_rate": 6.757896888028871e-05, "loss": 0.0392, "step": 33310 }, { "grad_norm": 0.3175962567329407, "learning_rate": 6.7559618514939e-05, "loss": 0.0385, "step": 33320 }, { "grad_norm": 0.2347732037305832, "learning_rate": 6.754026514913073e-05, "loss": 0.036, "step": 33330 }, { "grad_norm": 0.36021557450294495, "learning_rate": 6.752090878617087e-05, "loss": 0.0428, "step": 33340 }, { "grad_norm": 0.42727142572402954, "learning_rate": 6.75015494293669e-05, "loss": 0.0342, "step": 33350 }, { "grad_norm": 0.4252704679965973, "learning_rate": 6.74821870820268e-05, "loss": 0.0304, "step": 33360 }, { "grad_norm": 0.3277760148048401, "learning_rate": 6.746282174745907e-05, "loss": 0.0443, "step": 33370 }, { "grad_norm": 0.40954843163490295, "learning_rate": 6.744345342897271e-05, "loss": 0.0366, "step": 33380 }, { "grad_norm": 0.8744706511497498, "learning_rate": 6.742408212987724e-05, "loss": 0.0375, "step": 33390 }, { "grad_norm": 0.5287116765975952, "learning_rate": 6.740470785348269e-05, "loss": 0.0388, "step": 33400 }, { "grad_norm": 0.25175705552101135, "learning_rate": 6.738533060309958e-05, "loss": 0.0396, "step": 33410 }, { "grad_norm": 0.28978604078292847, "learning_rate": 6.736595038203894e-05, "loss": 0.0303, "step": 33420 }, { "grad_norm": 0.2870142459869385, "learning_rate": 6.734656719361236e-05, "loss": 0.0402, "step": 33430 }, { "grad_norm": 0.3199765682220459, "learning_rate": 6.732718104113189e-05, "loss": 0.0332, "step": 33440 }, { "grad_norm": 0.36716508865356445, "learning_rate": 6.730779192791006e-05, "loss": 0.033, "step": 33450 }, { "grad_norm": 0.3392215967178345, "learning_rate": 6.728839985725997e-05, "loss": 0.0359, "step": 33460 }, { "grad_norm": 0.38952261209487915, "learning_rate": 6.726900483249517e-05, "loss": 0.033, "step": 33470 }, { "grad_norm": 0.42340087890625, "learning_rate": 6.724960685692976e-05, "loss": 0.0322, "step": 33480 }, { "grad_norm": 0.6024302244186401, "learning_rate": 6.723020593387833e-05, "loss": 0.0399, "step": 33490 }, { "grad_norm": 0.4627910852432251, "learning_rate": 6.721080206665593e-05, "loss": 0.0375, "step": 33500 }, { "grad_norm": 0.3657429814338684, "learning_rate": 6.719139525857819e-05, "loss": 0.0456, "step": 33510 }, { "grad_norm": 0.37105655670166016, "learning_rate": 6.717198551296117e-05, "loss": 0.0354, "step": 33520 }, { "grad_norm": 0.33065739274024963, "learning_rate": 6.715257283312148e-05, "loss": 0.0386, "step": 33530 }, { "grad_norm": 0.46932080388069153, "learning_rate": 6.713315722237623e-05, "loss": 0.0403, "step": 33540 }, { "grad_norm": 0.48487505316734314, "learning_rate": 6.7113738684043e-05, "loss": 0.0339, "step": 33550 }, { "grad_norm": 0.29140162467956543, "learning_rate": 6.709431722143989e-05, "loss": 0.0371, "step": 33560 }, { "grad_norm": 0.4053085446357727, "learning_rate": 6.70748928378855e-05, "loss": 0.0426, "step": 33570 }, { "grad_norm": 0.4138706624507904, "learning_rate": 6.705546553669891e-05, "loss": 0.0312, "step": 33580 }, { "grad_norm": 0.25818896293640137, "learning_rate": 6.703603532119974e-05, "loss": 0.0274, "step": 33590 }, { "grad_norm": 0.20649205148220062, "learning_rate": 6.701660219470808e-05, "loss": 0.0319, "step": 33600 }, { "grad_norm": 0.3715382218360901, "learning_rate": 6.69971661605445e-05, "loss": 0.0307, "step": 33610 }, { "grad_norm": 0.29524558782577515, "learning_rate": 6.697772722203008e-05, "loss": 0.0323, "step": 33620 }, { "grad_norm": 0.3290821313858032, "learning_rate": 6.695828538248643e-05, "loss": 0.0387, "step": 33630 }, { "grad_norm": 0.26442280411720276, "learning_rate": 6.693884064523563e-05, "loss": 0.0315, "step": 33640 }, { "grad_norm": 0.4548465609550476, "learning_rate": 6.691939301360023e-05, "loss": 0.0361, "step": 33650 }, { "grad_norm": 0.3455100953578949, "learning_rate": 6.689994249090333e-05, "loss": 0.0314, "step": 33660 }, { "grad_norm": 0.49459755420684814, "learning_rate": 6.688048908046845e-05, "loss": 0.0392, "step": 33670 }, { "grad_norm": 0.3008935749530792, "learning_rate": 6.686103278561969e-05, "loss": 0.0308, "step": 33680 }, { "grad_norm": 0.3188508450984955, "learning_rate": 6.684157360968156e-05, "loss": 0.0365, "step": 33690 }, { "grad_norm": 0.33981990814208984, "learning_rate": 6.682211155597911e-05, "loss": 0.0323, "step": 33700 }, { "grad_norm": 0.33237430453300476, "learning_rate": 6.680264662783789e-05, "loss": 0.0336, "step": 33710 }, { "grad_norm": 0.39792364835739136, "learning_rate": 6.678317882858391e-05, "loss": 0.0345, "step": 33720 }, { "grad_norm": 0.40285733342170715, "learning_rate": 6.67637081615437e-05, "loss": 0.034, "step": 33730 }, { "grad_norm": 0.2901296019554138, "learning_rate": 6.674423463004427e-05, "loss": 0.0288, "step": 33740 }, { "grad_norm": 0.3581145703792572, "learning_rate": 6.672475823741308e-05, "loss": 0.0344, "step": 33750 }, { "grad_norm": 0.7074571847915649, "learning_rate": 6.670527898697811e-05, "loss": 0.0372, "step": 33760 }, { "grad_norm": 0.30502474308013916, "learning_rate": 6.668579688206788e-05, "loss": 0.0372, "step": 33770 }, { "grad_norm": 0.44667014479637146, "learning_rate": 6.666631192601131e-05, "loss": 0.0403, "step": 33780 }, { "grad_norm": 0.3513295352458954, "learning_rate": 6.664682412213785e-05, "loss": 0.0431, "step": 33790 }, { "grad_norm": 0.44961118698120117, "learning_rate": 6.662733347377745e-05, "loss": 0.0429, "step": 33800 }, { "grad_norm": 0.3876330852508545, "learning_rate": 6.660783998426051e-05, "loss": 0.0315, "step": 33810 }, { "grad_norm": 0.7797275185585022, "learning_rate": 6.658834365691794e-05, "loss": 0.0458, "step": 33820 }, { "grad_norm": 0.48653265833854675, "learning_rate": 6.656884449508115e-05, "loss": 0.0398, "step": 33830 }, { "grad_norm": 0.574275553226471, "learning_rate": 6.654934250208198e-05, "loss": 0.0382, "step": 33840 }, { "grad_norm": 0.3648197054862976, "learning_rate": 6.65298376812528e-05, "loss": 0.04, "step": 33850 }, { "grad_norm": 0.8477703928947449, "learning_rate": 6.651033003592646e-05, "loss": 0.0392, "step": 33860 }, { "grad_norm": 0.30198779702186584, "learning_rate": 6.649081956943626e-05, "loss": 0.0315, "step": 33870 }, { "grad_norm": 0.710267186164856, "learning_rate": 6.647130628511604e-05, "loss": 0.0412, "step": 33880 }, { "grad_norm": 0.29167965054512024, "learning_rate": 6.645179018630005e-05, "loss": 0.0323, "step": 33890 }, { "grad_norm": 0.44506314396858215, "learning_rate": 6.643227127632309e-05, "loss": 0.0386, "step": 33900 }, { "grad_norm": 0.5101012587547302, "learning_rate": 6.641274955852038e-05, "loss": 0.0423, "step": 33910 }, { "grad_norm": 0.5106651782989502, "learning_rate": 6.639322503622768e-05, "loss": 0.0639, "step": 33920 }, { "grad_norm": 0.4049110412597656, "learning_rate": 6.637369771278116e-05, "loss": 0.0452, "step": 33930 }, { "grad_norm": 0.27299681305885315, "learning_rate": 6.635416759151751e-05, "loss": 0.033, "step": 33940 }, { "grad_norm": 1.3565828800201416, "learning_rate": 6.633463467577394e-05, "loss": 0.0597, "step": 33950 }, { "grad_norm": 0.45550209283828735, "learning_rate": 6.631509896888803e-05, "loss": 0.0366, "step": 33960 }, { "grad_norm": 0.3074426054954529, "learning_rate": 6.629556047419794e-05, "loss": 0.0326, "step": 33970 }, { "grad_norm": 0.4452645480632782, "learning_rate": 6.627601919504223e-05, "loss": 0.0439, "step": 33980 }, { "grad_norm": 0.38954398036003113, "learning_rate": 6.625647513476001e-05, "loss": 0.0318, "step": 33990 }, { "grad_norm": 0.3718114197254181, "learning_rate": 6.62369282966908e-05, "loss": 0.0494, "step": 34000 }, { "grad_norm": 0.2777450680732727, "learning_rate": 6.621737868417464e-05, "loss": 0.037, "step": 34010 }, { "grad_norm": 0.3762982189655304, "learning_rate": 6.619782630055198e-05, "loss": 0.0363, "step": 34020 }, { "grad_norm": 0.33422836661338806, "learning_rate": 6.617827114916382e-05, "loss": 0.0352, "step": 34030 }, { "grad_norm": 0.3250778913497925, "learning_rate": 6.615871323335161e-05, "loss": 0.0345, "step": 34040 }, { "grad_norm": 0.41024282574653625, "learning_rate": 6.613915255645725e-05, "loss": 0.0378, "step": 34050 }, { "grad_norm": 0.3266327977180481, "learning_rate": 6.611958912182312e-05, "loss": 0.0309, "step": 34060 }, { "grad_norm": 0.40312108397483826, "learning_rate": 6.610002293279207e-05, "loss": 0.0404, "step": 34070 }, { "grad_norm": 0.44591814279556274, "learning_rate": 6.608045399270746e-05, "loss": 0.0369, "step": 34080 }, { "grad_norm": 0.297992467880249, "learning_rate": 6.606088230491304e-05, "loss": 0.0267, "step": 34090 }, { "grad_norm": 0.3154335916042328, "learning_rate": 6.604130787275312e-05, "loss": 0.0369, "step": 34100 }, { "grad_norm": 0.41308021545410156, "learning_rate": 6.602173069957242e-05, "loss": 0.0309, "step": 34110 }, { "grad_norm": 0.42279475927352905, "learning_rate": 6.600215078871612e-05, "loss": 0.0289, "step": 34120 }, { "grad_norm": 0.2742043435573578, "learning_rate": 6.598256814352992e-05, "loss": 0.0267, "step": 34130 }, { "grad_norm": 0.31284260749816895, "learning_rate": 6.596298276735995e-05, "loss": 0.0263, "step": 34140 }, { "grad_norm": 0.40704992413520813, "learning_rate": 6.594339466355282e-05, "loss": 0.0333, "step": 34150 }, { "grad_norm": 0.35192981362342834, "learning_rate": 6.592380383545558e-05, "loss": 0.0333, "step": 34160 }, { "grad_norm": 0.3271801471710205, "learning_rate": 6.590421028641577e-05, "loss": 0.0438, "step": 34170 }, { "grad_norm": 0.39635539054870605, "learning_rate": 6.588461401978143e-05, "loss": 0.0336, "step": 34180 }, { "grad_norm": 0.37296128273010254, "learning_rate": 6.586501503890099e-05, "loss": 0.0326, "step": 34190 }, { "grad_norm": 0.30929574370384216, "learning_rate": 6.584541334712338e-05, "loss": 0.0359, "step": 34200 }, { "grad_norm": 0.3135336935520172, "learning_rate": 6.582580894779802e-05, "loss": 0.0374, "step": 34210 }, { "grad_norm": 0.23494742810726166, "learning_rate": 6.580620184427473e-05, "loss": 0.027, "step": 34220 }, { "grad_norm": 0.42121341824531555, "learning_rate": 6.578659203990385e-05, "loss": 0.0332, "step": 34230 }, { "grad_norm": 0.416575163602829, "learning_rate": 6.576697953803615e-05, "loss": 0.0309, "step": 34240 }, { "grad_norm": 0.3168685734272003, "learning_rate": 6.57473643420229e-05, "loss": 0.0275, "step": 34250 }, { "grad_norm": 0.33462074398994446, "learning_rate": 6.572774645521574e-05, "loss": 0.0369, "step": 34260 }, { "grad_norm": 0.5358186960220337, "learning_rate": 6.570812588096688e-05, "loss": 0.0425, "step": 34270 }, { "grad_norm": 0.3055015504360199, "learning_rate": 6.568850262262893e-05, "loss": 0.0285, "step": 34280 }, { "grad_norm": 0.3449562191963196, "learning_rate": 6.566887668355497e-05, "loss": 0.0369, "step": 34290 }, { "grad_norm": 0.2354498654603958, "learning_rate": 6.564924806709851e-05, "loss": 0.0285, "step": 34300 }, { "grad_norm": 0.41170549392700195, "learning_rate": 6.562961677661359e-05, "loss": 0.0346, "step": 34310 }, { "grad_norm": 0.5046586394309998, "learning_rate": 6.56099828154546e-05, "loss": 0.034, "step": 34320 }, { "grad_norm": 0.32015231251716614, "learning_rate": 6.55903461869765e-05, "loss": 0.0311, "step": 34330 }, { "grad_norm": 0.3032272160053253, "learning_rate": 6.557070689453465e-05, "loss": 0.0293, "step": 34340 }, { "grad_norm": 0.35525211691856384, "learning_rate": 6.555106494148482e-05, "loss": 0.04, "step": 34350 }, { "grad_norm": 0.3444287180900574, "learning_rate": 6.553142033118333e-05, "loss": 0.0366, "step": 34360 }, { "grad_norm": 0.41129070520401, "learning_rate": 6.551177306698688e-05, "loss": 0.0428, "step": 34370 }, { "grad_norm": 0.2558669149875641, "learning_rate": 6.549212315225267e-05, "loss": 0.0377, "step": 34380 }, { "grad_norm": 0.43134385347366333, "learning_rate": 6.547247059033833e-05, "loss": 0.0453, "step": 34390 }, { "grad_norm": 1.5732808113098145, "learning_rate": 6.545281538460193e-05, "loss": 0.0418, "step": 34400 }, { "grad_norm": 0.3435264527797699, "learning_rate": 6.543315753840202e-05, "loss": 0.0388, "step": 34410 }, { "grad_norm": 0.46152356266975403, "learning_rate": 6.541349705509758e-05, "loss": 0.0394, "step": 34420 }, { "grad_norm": 0.42438605427742004, "learning_rate": 6.539383393804805e-05, "loss": 0.0334, "step": 34430 }, { "grad_norm": 0.3846417963504791, "learning_rate": 6.537416819061333e-05, "loss": 0.0433, "step": 34440 }, { "grad_norm": 0.28305524587631226, "learning_rate": 6.535449981615375e-05, "loss": 0.0472, "step": 34450 }, { "grad_norm": 0.3522157371044159, "learning_rate": 6.53348288180301e-05, "loss": 0.0338, "step": 34460 }, { "grad_norm": 0.3839015066623688, "learning_rate": 6.531515519960361e-05, "loss": 0.0378, "step": 34470 }, { "grad_norm": 0.2414085417985916, "learning_rate": 6.529547896423597e-05, "loss": 0.0325, "step": 34480 }, { "grad_norm": 0.5532206296920776, "learning_rate": 6.52758001152893e-05, "loss": 0.0414, "step": 34490 }, { "grad_norm": 0.2768282890319824, "learning_rate": 6.525611865612618e-05, "loss": 0.0275, "step": 34500 }, { "grad_norm": 0.3098631799221039, "learning_rate": 6.523643459010966e-05, "loss": 0.0331, "step": 34510 }, { "grad_norm": 0.5123671293258667, "learning_rate": 6.521674792060317e-05, "loss": 0.032, "step": 34520 }, { "grad_norm": 0.36253711581230164, "learning_rate": 6.519705865097063e-05, "loss": 0.0394, "step": 34530 }, { "grad_norm": 0.45691338181495667, "learning_rate": 6.517736678457641e-05, "loss": 0.0335, "step": 34540 }, { "grad_norm": 0.5154988169670105, "learning_rate": 6.515767232478534e-05, "loss": 0.0372, "step": 34550 }, { "grad_norm": 0.2654683291912079, "learning_rate": 6.51379752749626e-05, "loss": 0.033, "step": 34560 }, { "grad_norm": 0.33716049790382385, "learning_rate": 6.511827563847393e-05, "loss": 0.0361, "step": 34570 }, { "grad_norm": 0.4835299849510193, "learning_rate": 6.509857341868542e-05, "loss": 0.0417, "step": 34580 }, { "grad_norm": 0.44779837131500244, "learning_rate": 6.507886861896367e-05, "loss": 0.0346, "step": 34590 }, { "grad_norm": 0.43351098895072937, "learning_rate": 6.505916124267567e-05, "loss": 0.0441, "step": 34600 }, { "grad_norm": 0.4297954738140106, "learning_rate": 6.503945129318891e-05, "loss": 0.0375, "step": 34610 }, { "grad_norm": 0.47383055090904236, "learning_rate": 6.501973877387122e-05, "loss": 0.0283, "step": 34620 }, { "grad_norm": 0.3953867554664612, "learning_rate": 6.500002368809098e-05, "loss": 0.0341, "step": 34630 }, { "grad_norm": 0.4641437530517578, "learning_rate": 6.498030603921694e-05, "loss": 0.0359, "step": 34640 }, { "grad_norm": 0.40813323855400085, "learning_rate": 6.496058583061832e-05, "loss": 0.0407, "step": 34650 }, { "grad_norm": 0.32358238101005554, "learning_rate": 6.494086306566475e-05, "loss": 0.0293, "step": 34660 }, { "grad_norm": 0.4489951431751251, "learning_rate": 6.492113774772632e-05, "loss": 0.0384, "step": 34670 }, { "grad_norm": 0.3393784463405609, "learning_rate": 6.490140988017354e-05, "loss": 0.042, "step": 34680 }, { "grad_norm": 0.45739924907684326, "learning_rate": 6.488167946637736e-05, "loss": 0.0338, "step": 34690 }, { "grad_norm": 0.4010557532310486, "learning_rate": 6.486194650970915e-05, "loss": 0.0373, "step": 34700 }, { "grad_norm": 0.31319624185562134, "learning_rate": 6.48422110135408e-05, "loss": 0.0412, "step": 34710 }, { "grad_norm": 0.32101649045944214, "learning_rate": 6.482247298124451e-05, "loss": 0.0471, "step": 34720 }, { "grad_norm": 0.3559264838695526, "learning_rate": 6.480273241619297e-05, "loss": 0.0323, "step": 34730 }, { "grad_norm": 0.26332324743270874, "learning_rate": 6.478298932175933e-05, "loss": 0.04, "step": 34740 }, { "grad_norm": 0.2148020714521408, "learning_rate": 6.476324370131712e-05, "loss": 0.0333, "step": 34750 }, { "grad_norm": 0.42284971475601196, "learning_rate": 6.474349555824036e-05, "loss": 0.0362, "step": 34760 }, { "grad_norm": 0.4130350351333618, "learning_rate": 6.472374489590342e-05, "loss": 0.0352, "step": 34770 }, { "grad_norm": 0.4169442355632782, "learning_rate": 6.470399171768118e-05, "loss": 0.0473, "step": 34780 }, { "grad_norm": 0.4874761998653412, "learning_rate": 6.468423602694891e-05, "loss": 0.0391, "step": 34790 }, { "grad_norm": 0.5974038243293762, "learning_rate": 6.466447782708232e-05, "loss": 0.0364, "step": 34800 }, { "grad_norm": 0.3558851182460785, "learning_rate": 6.464471712145754e-05, "loss": 0.0344, "step": 34810 }, { "grad_norm": 0.21651186048984528, "learning_rate": 6.462495391345114e-05, "loss": 0.0375, "step": 34820 }, { "grad_norm": 0.35218966007232666, "learning_rate": 6.46051882064401e-05, "loss": 0.0305, "step": 34830 }, { "grad_norm": 0.38029664754867554, "learning_rate": 6.458542000380186e-05, "loss": 0.0399, "step": 34840 }, { "grad_norm": 0.45729008316993713, "learning_rate": 6.456564930891424e-05, "loss": 0.0327, "step": 34850 }, { "grad_norm": 0.38175421953201294, "learning_rate": 6.454587612515555e-05, "loss": 0.0421, "step": 34860 }, { "grad_norm": 0.370307594537735, "learning_rate": 6.452610045590444e-05, "loss": 0.0404, "step": 34870 }, { "grad_norm": 0.39962759613990784, "learning_rate": 6.450632230454005e-05, "loss": 0.0356, "step": 34880 }, { "grad_norm": 0.5098312497138977, "learning_rate": 6.448654167444195e-05, "loss": 0.0403, "step": 34890 }, { "grad_norm": 0.3101332485675812, "learning_rate": 6.446675856899005e-05, "loss": 0.0419, "step": 34900 }, { "grad_norm": 0.36017727851867676, "learning_rate": 6.444697299156481e-05, "loss": 0.0357, "step": 34910 }, { "grad_norm": 0.3655643165111542, "learning_rate": 6.442718494554701e-05, "loss": 0.0366, "step": 34920 }, { "grad_norm": 0.38786235451698303, "learning_rate": 6.440739443431787e-05, "loss": 0.0339, "step": 34930 }, { "grad_norm": 0.3547041714191437, "learning_rate": 6.438760146125906e-05, "loss": 0.0293, "step": 34940 }, { "grad_norm": 0.3419681191444397, "learning_rate": 6.436780602975267e-05, "loss": 0.0366, "step": 34950 }, { "grad_norm": 0.30990853905677795, "learning_rate": 6.43480081431812e-05, "loss": 0.0385, "step": 34960 }, { "grad_norm": 0.3983279764652252, "learning_rate": 6.432820780492756e-05, "loss": 0.0383, "step": 34970 }, { "grad_norm": 0.3099328577518463, "learning_rate": 6.430840501837506e-05, "loss": 0.0413, "step": 34980 }, { "grad_norm": 0.3702574074268341, "learning_rate": 6.428859978690748e-05, "loss": 0.0376, "step": 34990 }, { "grad_norm": 0.37367260456085205, "learning_rate": 6.426879211390901e-05, "loss": 0.0368, "step": 35000 }, { "grad_norm": 0.4146895706653595, "learning_rate": 6.424898200276422e-05, "loss": 0.0359, "step": 35010 }, { "grad_norm": 0.2763304114341736, "learning_rate": 6.42291694568581e-05, "loss": 0.0305, "step": 35020 }, { "grad_norm": 0.2820489704608917, "learning_rate": 6.42093544795761e-05, "loss": 0.0378, "step": 35030 }, { "grad_norm": 0.36778226494789124, "learning_rate": 6.418953707430403e-05, "loss": 0.0353, "step": 35040 }, { "grad_norm": 0.3174384534358978, "learning_rate": 6.416971724442819e-05, "loss": 0.0334, "step": 35050 }, { "grad_norm": 0.3698807656764984, "learning_rate": 6.414989499333519e-05, "loss": 0.0328, "step": 35060 }, { "grad_norm": 0.4107685983181, "learning_rate": 6.413007032441214e-05, "loss": 0.0314, "step": 35070 }, { "grad_norm": 0.6483583450317383, "learning_rate": 6.411024324104653e-05, "loss": 0.0409, "step": 35080 }, { "grad_norm": 0.29009971022605896, "learning_rate": 6.409041374662628e-05, "loss": 0.0316, "step": 35090 }, { "grad_norm": 0.45542967319488525, "learning_rate": 6.407058184453967e-05, "loss": 0.047, "step": 35100 }, { "grad_norm": 0.38743335008621216, "learning_rate": 6.405074753817548e-05, "loss": 0.0409, "step": 35110 }, { "grad_norm": 0.3982580304145813, "learning_rate": 6.40309108309228e-05, "loss": 0.0319, "step": 35120 }, { "grad_norm": 0.39018693566322327, "learning_rate": 6.401107172617122e-05, "loss": 0.0336, "step": 35130 }, { "grad_norm": 0.288872629404068, "learning_rate": 6.399123022731068e-05, "loss": 0.0603, "step": 35140 }, { "grad_norm": 0.41573604941368103, "learning_rate": 6.397138633773157e-05, "loss": 0.0315, "step": 35150 }, { "grad_norm": 0.4514138400554657, "learning_rate": 6.395154006082463e-05, "loss": 0.0348, "step": 35160 }, { "grad_norm": 0.31663769483566284, "learning_rate": 6.393169139998109e-05, "loss": 0.0328, "step": 35170 }, { "grad_norm": 0.3988460600376129, "learning_rate": 6.39118403585925e-05, "loss": 0.0355, "step": 35180 }, { "grad_norm": 0.31321197748184204, "learning_rate": 6.38919869400509e-05, "loss": 0.0383, "step": 35190 }, { "grad_norm": 0.24613280594348907, "learning_rate": 6.387213114774865e-05, "loss": 0.0366, "step": 35200 }, { "grad_norm": 0.5521673560142517, "learning_rate": 6.385227298507863e-05, "loss": 0.0615, "step": 35210 }, { "grad_norm": 0.30619046092033386, "learning_rate": 6.3832412455434e-05, "loss": 0.0349, "step": 35220 }, { "grad_norm": 0.27076178789138794, "learning_rate": 6.381254956220841e-05, "loss": 0.0341, "step": 35230 }, { "grad_norm": 0.4249899685382843, "learning_rate": 6.379268430879586e-05, "loss": 0.0363, "step": 35240 }, { "grad_norm": 0.4583786427974701, "learning_rate": 6.37728166985908e-05, "loss": 0.0388, "step": 35250 }, { "grad_norm": 0.33009710907936096, "learning_rate": 6.375294673498804e-05, "loss": 0.041, "step": 35260 }, { "grad_norm": 0.3828693628311157, "learning_rate": 6.373307442138284e-05, "loss": 0.0399, "step": 35270 }, { "grad_norm": 0.35639387369155884, "learning_rate": 6.371319976117081e-05, "loss": 0.0443, "step": 35280 }, { "grad_norm": 0.39455756545066833, "learning_rate": 6.3693322757748e-05, "loss": 0.0313, "step": 35290 }, { "grad_norm": 0.25107744336128235, "learning_rate": 6.367344341451086e-05, "loss": 0.0343, "step": 35300 }, { "grad_norm": 0.41773781180381775, "learning_rate": 6.36535617348562e-05, "loss": 0.0388, "step": 35310 }, { "grad_norm": 0.31315070390701294, "learning_rate": 6.363367772218128e-05, "loss": 0.0363, "step": 35320 }, { "grad_norm": 0.28505265712738037, "learning_rate": 6.36137913798837e-05, "loss": 0.0315, "step": 35330 }, { "grad_norm": 0.3431493937969208, "learning_rate": 6.359390271136151e-05, "loss": 0.0295, "step": 35340 }, { "grad_norm": 0.3360559940338135, "learning_rate": 6.357401172001314e-05, "loss": 0.0275, "step": 35350 }, { "grad_norm": 0.2930402457714081, "learning_rate": 6.355411840923742e-05, "loss": 0.0288, "step": 35360 }, { "grad_norm": 0.48775213956832886, "learning_rate": 6.353422278243358e-05, "loss": 0.0352, "step": 35370 }, { "grad_norm": 0.408101350069046, "learning_rate": 6.351432484300121e-05, "loss": 0.0364, "step": 35380 }, { "grad_norm": 0.4059446156024933, "learning_rate": 6.349442459434036e-05, "loss": 0.0359, "step": 35390 }, { "grad_norm": 0.3122618496417999, "learning_rate": 6.34745220398514e-05, "loss": 0.0395, "step": 35400 }, { "grad_norm": 0.2552597224712372, "learning_rate": 6.345461718293518e-05, "loss": 0.0367, "step": 35410 }, { "grad_norm": 0.3538784980773926, "learning_rate": 6.343471002699286e-05, "loss": 0.0398, "step": 35420 }, { "grad_norm": 0.440025269985199, "learning_rate": 6.341480057542602e-05, "loss": 0.0375, "step": 35430 }, { "grad_norm": 0.3165484368801117, "learning_rate": 6.339488883163667e-05, "loss": 0.0332, "step": 35440 }, { "grad_norm": 0.23486848175525665, "learning_rate": 6.337497479902716e-05, "loss": 0.033, "step": 35450 }, { "grad_norm": 0.5312992334365845, "learning_rate": 6.335505848100027e-05, "loss": 0.0375, "step": 35460 }, { "grad_norm": 0.26420778036117554, "learning_rate": 6.333513988095915e-05, "loss": 0.0266, "step": 35470 }, { "grad_norm": 0.2741440236568451, "learning_rate": 6.331521900230735e-05, "loss": 0.0271, "step": 35480 }, { "grad_norm": 0.460483193397522, "learning_rate": 6.329529584844878e-05, "loss": 0.0365, "step": 35490 }, { "grad_norm": 0.3500629961490631, "learning_rate": 6.327537042278777e-05, "loss": 0.0373, "step": 35500 }, { "grad_norm": 0.37420952320098877, "learning_rate": 6.325544272872905e-05, "loss": 0.0371, "step": 35510 }, { "grad_norm": 0.29625123739242554, "learning_rate": 6.323551276967771e-05, "loss": 0.0418, "step": 35520 }, { "grad_norm": 0.4382452964782715, "learning_rate": 6.321558054903922e-05, "loss": 0.0301, "step": 35530 }, { "grad_norm": 0.35036736726760864, "learning_rate": 6.319564607021947e-05, "loss": 0.0407, "step": 35540 }, { "grad_norm": 0.3074062168598175, "learning_rate": 6.31757093366247e-05, "loss": 0.0282, "step": 35550 }, { "grad_norm": 0.402275413274765, "learning_rate": 6.315577035166154e-05, "loss": 0.0302, "step": 35560 }, { "grad_norm": 0.5117396712303162, "learning_rate": 6.313582911873708e-05, "loss": 0.0309, "step": 35570 }, { "grad_norm": 0.3865792453289032, "learning_rate": 6.311588564125865e-05, "loss": 0.031, "step": 35580 }, { "grad_norm": 0.42295581102371216, "learning_rate": 6.30959399226341e-05, "loss": 0.0383, "step": 35590 }, { "grad_norm": 0.3599883019924164, "learning_rate": 6.30759919662716e-05, "loss": 0.0462, "step": 35600 }, { "grad_norm": 0.3900999128818512, "learning_rate": 6.30560417755797e-05, "loss": 0.0346, "step": 35610 }, { "grad_norm": 0.4848412573337555, "learning_rate": 6.303608935396735e-05, "loss": 0.0259, "step": 35620 }, { "grad_norm": 0.3685753047466278, "learning_rate": 6.301613470484386e-05, "loss": 0.0295, "step": 35630 }, { "grad_norm": 0.43412795662879944, "learning_rate": 6.299617783161893e-05, "loss": 0.0341, "step": 35640 }, { "grad_norm": 0.3355984389781952, "learning_rate": 6.297621873770266e-05, "loss": 0.039, "step": 35650 }, { "grad_norm": 0.43858209252357483, "learning_rate": 6.29562574265055e-05, "loss": 0.0346, "step": 35660 }, { "grad_norm": 0.4915112555027008, "learning_rate": 6.293629390143834e-05, "loss": 0.0389, "step": 35670 }, { "grad_norm": 0.39846912026405334, "learning_rate": 6.291632816591232e-05, "loss": 0.0357, "step": 35680 }, { "grad_norm": 0.5466178059577942, "learning_rate": 6.28963602233391e-05, "loss": 0.0351, "step": 35690 }, { "grad_norm": 0.49589279294013977, "learning_rate": 6.287639007713062e-05, "loss": 0.0396, "step": 35700 }, { "grad_norm": 0.41033411026000977, "learning_rate": 6.285641773069926e-05, "loss": 0.0353, "step": 35710 }, { "grad_norm": 1.0568357706069946, "learning_rate": 6.283644318745773e-05, "loss": 0.0324, "step": 35720 }, { "grad_norm": 0.3484207093715668, "learning_rate": 6.281646645081912e-05, "loss": 0.0371, "step": 35730 }, { "grad_norm": 0.32397404313087463, "learning_rate": 6.279648752419693e-05, "loss": 0.0316, "step": 35740 }, { "grad_norm": 0.2790728211402893, "learning_rate": 6.2776506411005e-05, "loss": 0.0432, "step": 35750 }, { "grad_norm": 0.30700644850730896, "learning_rate": 6.275652311465758e-05, "loss": 0.0275, "step": 35760 }, { "grad_norm": 0.30628326535224915, "learning_rate": 6.273653763856926e-05, "loss": 0.0329, "step": 35770 }, { "grad_norm": 0.3452499508857727, "learning_rate": 6.271654998615501e-05, "loss": 0.0393, "step": 35780 }, { "grad_norm": 0.5299791097640991, "learning_rate": 6.269656016083013e-05, "loss": 0.0281, "step": 35790 }, { "grad_norm": 0.3576256334781647, "learning_rate": 6.267656816601038e-05, "loss": 0.0451, "step": 35800 }, { "grad_norm": 0.36563822627067566, "learning_rate": 6.265657400511185e-05, "loss": 0.0397, "step": 35810 }, { "grad_norm": 0.26926368474960327, "learning_rate": 6.263657768155098e-05, "loss": 0.0354, "step": 35820 }, { "grad_norm": 0.3153313994407654, "learning_rate": 6.261657919874457e-05, "loss": 0.0372, "step": 35830 }, { "grad_norm": 0.9977007508277893, "learning_rate": 6.259657856010986e-05, "loss": 0.0447, "step": 35840 }, { "grad_norm": 0.39519837498664856, "learning_rate": 6.257657576906439e-05, "loss": 0.0346, "step": 35850 }, { "grad_norm": 0.30170321464538574, "learning_rate": 6.255657082902609e-05, "loss": 0.0301, "step": 35860 }, { "grad_norm": 1.1096850633621216, "learning_rate": 6.253656374341325e-05, "loss": 0.032, "step": 35870 }, { "grad_norm": 0.4578339457511902, "learning_rate": 6.251655451564457e-05, "loss": 0.0382, "step": 35880 }, { "grad_norm": 0.40250879526138306, "learning_rate": 6.249654314913902e-05, "loss": 0.0323, "step": 35890 }, { "grad_norm": 0.29122671484947205, "learning_rate": 6.247652964731604e-05, "loss": 0.0301, "step": 35900 }, { "grad_norm": 0.26037684082984924, "learning_rate": 6.245651401359537e-05, "loss": 0.0456, "step": 35910 }, { "grad_norm": 0.31507110595703125, "learning_rate": 6.243649625139715e-05, "loss": 0.0423, "step": 35920 }, { "grad_norm": 0.28504589200019836, "learning_rate": 6.241647636414185e-05, "loss": 0.0287, "step": 35930 }, { "grad_norm": 0.4100695550441742, "learning_rate": 6.239645435525034e-05, "loss": 0.0427, "step": 35940 }, { "grad_norm": 0.3527255654335022, "learning_rate": 6.237643022814381e-05, "loss": 0.0373, "step": 35950 }, { "grad_norm": 0.42613187432289124, "learning_rate": 6.235640398624386e-05, "loss": 0.0287, "step": 35960 }, { "grad_norm": 0.39037495851516724, "learning_rate": 6.233637563297243e-05, "loss": 0.0359, "step": 35970 }, { "grad_norm": 0.3732101023197174, "learning_rate": 6.23163451717518e-05, "loss": 0.0456, "step": 35980 }, { "grad_norm": 0.37024784088134766, "learning_rate": 6.229631260600463e-05, "loss": 0.0343, "step": 35990 }, { "grad_norm": 0.4886071979999542, "learning_rate": 6.227627793915392e-05, "loss": 0.0405, "step": 36000 }, { "grad_norm": 0.49175795912742615, "learning_rate": 6.225624117462309e-05, "loss": 0.0339, "step": 36010 }, { "grad_norm": 0.25153225660324097, "learning_rate": 6.223620231583586e-05, "loss": 0.0338, "step": 36020 }, { "grad_norm": 0.27721041440963745, "learning_rate": 6.221616136621629e-05, "loss": 0.0355, "step": 36030 }, { "grad_norm": 0.31411224603652954, "learning_rate": 6.219611832918887e-05, "loss": 0.0308, "step": 36040 }, { "grad_norm": 0.3195299208164215, "learning_rate": 6.217607320817838e-05, "loss": 0.0277, "step": 36050 }, { "grad_norm": 0.23785947263240814, "learning_rate": 6.215602600661001e-05, "loss": 0.0281, "step": 36060 }, { "grad_norm": 0.2322513610124588, "learning_rate": 6.213597672790925e-05, "loss": 0.0279, "step": 36070 }, { "grad_norm": 0.3265317380428314, "learning_rate": 6.2115925375502e-05, "loss": 0.0304, "step": 36080 }, { "grad_norm": 0.48268505930900574, "learning_rate": 6.209587195281447e-05, "loss": 0.0318, "step": 36090 }, { "grad_norm": 0.320332407951355, "learning_rate": 6.207581646327324e-05, "loss": 0.0432, "step": 36100 }, { "grad_norm": 0.4618764817714691, "learning_rate": 6.205575891030526e-05, "loss": 0.0302, "step": 36110 }, { "grad_norm": 0.287662148475647, "learning_rate": 6.203569929733781e-05, "loss": 0.0324, "step": 36120 }, { "grad_norm": 0.4881218373775482, "learning_rate": 6.201563762779852e-05, "loss": 0.0317, "step": 36130 }, { "grad_norm": 0.2646121382713318, "learning_rate": 6.199557390511538e-05, "loss": 0.026, "step": 36140 }, { "grad_norm": 0.3384959399700165, "learning_rate": 6.197550813271675e-05, "loss": 0.0338, "step": 36150 }, { "grad_norm": 0.29432570934295654, "learning_rate": 6.195544031403131e-05, "loss": 0.0394, "step": 36160 }, { "grad_norm": 0.3267546594142914, "learning_rate": 6.19353704524881e-05, "loss": 0.0307, "step": 36170 }, { "grad_norm": 0.2683994472026825, "learning_rate": 6.191529855151652e-05, "loss": 0.0289, "step": 36180 }, { "grad_norm": 0.30635154247283936, "learning_rate": 6.189522461454629e-05, "loss": 0.0375, "step": 36190 }, { "grad_norm": 0.33015885949134827, "learning_rate": 6.187514864500752e-05, "loss": 0.0292, "step": 36200 }, { "grad_norm": 0.756668746471405, "learning_rate": 6.185507064633062e-05, "loss": 0.0414, "step": 36210 }, { "grad_norm": 0.5217661261558533, "learning_rate": 6.18349906219464e-05, "loss": 0.048, "step": 36220 }, { "grad_norm": 0.39272329211235046, "learning_rate": 6.181490857528596e-05, "loss": 0.0334, "step": 36230 }, { "grad_norm": 0.311147004365921, "learning_rate": 6.179482450978077e-05, "loss": 0.0347, "step": 36240 }, { "grad_norm": 0.32758763432502747, "learning_rate": 6.177473842886269e-05, "loss": 0.0342, "step": 36250 }, { "grad_norm": 0.4657890498638153, "learning_rate": 6.175465033596382e-05, "loss": 0.039, "step": 36260 }, { "grad_norm": 0.3977874219417572, "learning_rate": 6.173456023451671e-05, "loss": 0.0315, "step": 36270 }, { "grad_norm": 0.5417835116386414, "learning_rate": 6.171446812795422e-05, "loss": 0.0435, "step": 36280 }, { "grad_norm": 0.2343945950269699, "learning_rate": 6.169437401970949e-05, "loss": 0.0369, "step": 36290 }, { "grad_norm": 0.2669716775417328, "learning_rate": 6.16742779132161e-05, "loss": 0.0275, "step": 36300 }, { "grad_norm": 0.4683694839477539, "learning_rate": 6.165417981190789e-05, "loss": 0.0322, "step": 36310 }, { "grad_norm": 0.32416826486587524, "learning_rate": 6.16340797192191e-05, "loss": 0.0313, "step": 36320 }, { "grad_norm": 0.30849888920783997, "learning_rate": 6.161397763858427e-05, "loss": 0.0353, "step": 36330 }, { "grad_norm": 0.42188596725463867, "learning_rate": 6.159387357343834e-05, "loss": 0.0359, "step": 36340 }, { "grad_norm": 0.5261901617050171, "learning_rate": 6.157376752721648e-05, "loss": 0.033, "step": 36350 }, { "grad_norm": 0.8088988661766052, "learning_rate": 6.155365950335428e-05, "loss": 0.0316, "step": 36360 }, { "grad_norm": 0.47895118594169617, "learning_rate": 6.153354950528768e-05, "loss": 0.0299, "step": 36370 }, { "grad_norm": 0.3608957827091217, "learning_rate": 6.151343753645293e-05, "loss": 0.0349, "step": 36380 }, { "grad_norm": 0.3349943459033966, "learning_rate": 6.149332360028657e-05, "loss": 0.0421, "step": 36390 }, { "grad_norm": 0.3170572519302368, "learning_rate": 6.147320770022555e-05, "loss": 0.0374, "step": 36400 }, { "grad_norm": 0.41022032499313354, "learning_rate": 6.145308983970715e-05, "loss": 0.0371, "step": 36410 }, { "grad_norm": 0.3416978716850281, "learning_rate": 6.143297002216892e-05, "loss": 0.0404, "step": 36420 }, { "grad_norm": 0.3360217213630676, "learning_rate": 6.141284825104882e-05, "loss": 0.035, "step": 36430 }, { "grad_norm": 0.3252827227115631, "learning_rate": 6.13927245297851e-05, "loss": 0.0403, "step": 36440 }, { "grad_norm": 0.24997763335704803, "learning_rate": 6.137259886181633e-05, "loss": 0.0417, "step": 36450 }, { "grad_norm": 0.2602977156639099, "learning_rate": 6.135247125058145e-05, "loss": 0.0311, "step": 36460 }, { "grad_norm": 0.21695983409881592, "learning_rate": 6.133234169951974e-05, "loss": 0.0259, "step": 36470 }, { "grad_norm": 0.3727875351905823, "learning_rate": 6.131221021207078e-05, "loss": 0.036, "step": 36480 }, { "grad_norm": 0.2786376178264618, "learning_rate": 6.129207679167448e-05, "loss": 0.0287, "step": 36490 }, { "grad_norm": 0.2949736416339874, "learning_rate": 6.127194144177109e-05, "loss": 0.0256, "step": 36500 }, { "grad_norm": 0.25350984930992126, "learning_rate": 6.125180416580118e-05, "loss": 0.0284, "step": 36510 }, { "grad_norm": 0.22819910943508148, "learning_rate": 6.123166496720571e-05, "loss": 0.0263, "step": 36520 }, { "grad_norm": 0.36921072006225586, "learning_rate": 6.121152384942588e-05, "loss": 0.0346, "step": 36530 }, { "grad_norm": 0.28168684244155884, "learning_rate": 6.119138081590324e-05, "loss": 0.0297, "step": 36540 }, { "grad_norm": 0.5178931355476379, "learning_rate": 6.117123587007971e-05, "loss": 0.0363, "step": 36550 }, { "grad_norm": 0.46571096777915955, "learning_rate": 6.11510890153975e-05, "loss": 0.0347, "step": 36560 }, { "grad_norm": 0.3499256670475006, "learning_rate": 6.113094025529916e-05, "loss": 0.0283, "step": 36570 }, { "grad_norm": 0.41242292523384094, "learning_rate": 6.111078959322757e-05, "loss": 0.0552, "step": 36580 }, { "grad_norm": 0.4132468104362488, "learning_rate": 6.109063703262592e-05, "loss": 0.0384, "step": 36590 }, { "grad_norm": 0.27261826395988464, "learning_rate": 6.107048257693772e-05, "loss": 0.0391, "step": 36600 }, { "grad_norm": 0.3989141583442688, "learning_rate": 6.105032622960683e-05, "loss": 0.0372, "step": 36610 }, { "grad_norm": 0.3141096830368042, "learning_rate": 6.103016799407743e-05, "loss": 0.0379, "step": 36620 }, { "grad_norm": 0.36469730734825134, "learning_rate": 6.1010007873793984e-05, "loss": 0.0359, "step": 36630 }, { "grad_norm": 0.33665215969085693, "learning_rate": 6.098984587220131e-05, "loss": 0.0347, "step": 36640 }, { "grad_norm": 0.318759560585022, "learning_rate": 6.096968199274456e-05, "loss": 0.0324, "step": 36650 }, { "grad_norm": 0.28684571385383606, "learning_rate": 6.0949516238869166e-05, "loss": 0.0316, "step": 36660 }, { "grad_norm": 0.32496151328086853, "learning_rate": 6.092934861402092e-05, "loss": 0.0346, "step": 36670 }, { "grad_norm": 0.35519078373908997, "learning_rate": 6.0909179121645924e-05, "loss": 0.0356, "step": 36680 }, { "grad_norm": 0.3639073967933655, "learning_rate": 6.0889007765190576e-05, "loss": 0.0407, "step": 36690 }, { "grad_norm": 0.2729485034942627, "learning_rate": 6.086883454810162e-05, "loss": 0.0316, "step": 36700 }, { "grad_norm": 0.6521144509315491, "learning_rate": 6.0848659473826084e-05, "loss": 0.0414, "step": 36710 }, { "grad_norm": 0.3580501079559326, "learning_rate": 6.082848254581138e-05, "loss": 0.0366, "step": 36720 }, { "grad_norm": 0.30776795744895935, "learning_rate": 6.080830376750517e-05, "loss": 0.0327, "step": 36730 }, { "grad_norm": 0.31976377964019775, "learning_rate": 6.0788123142355445e-05, "loss": 0.0333, "step": 36740 }, { "grad_norm": 0.36663538217544556, "learning_rate": 6.076794067381052e-05, "loss": 0.0295, "step": 36750 }, { "grad_norm": 0.32124486565589905, "learning_rate": 6.074775636531905e-05, "loss": 0.036, "step": 36760 }, { "grad_norm": 0.5050252079963684, "learning_rate": 6.072757022032997e-05, "loss": 0.0334, "step": 36770 }, { "grad_norm": 0.48478952050209045, "learning_rate": 6.070738224229253e-05, "loss": 0.0413, "step": 36780 }, { "grad_norm": 0.33356624841690063, "learning_rate": 6.0687192434656314e-05, "loss": 0.0364, "step": 36790 }, { "grad_norm": 0.39732441306114197, "learning_rate": 6.066700080087121e-05, "loss": 0.0347, "step": 36800 }, { "grad_norm": 0.22675427794456482, "learning_rate": 6.0646807344387424e-05, "loss": 0.0366, "step": 36810 }, { "grad_norm": 0.36001506447792053, "learning_rate": 6.062661206865543e-05, "loss": 0.0433, "step": 36820 }, { "grad_norm": 0.573038637638092, "learning_rate": 6.06064149771261e-05, "loss": 0.04, "step": 36830 }, { "grad_norm": 0.41522684693336487, "learning_rate": 6.058621607325051e-05, "loss": 0.0441, "step": 36840 }, { "grad_norm": 0.31787022948265076, "learning_rate": 6.056601536048014e-05, "loss": 0.0375, "step": 36850 }, { "grad_norm": 0.36979159712791443, "learning_rate": 6.0545812842266725e-05, "loss": 0.0435, "step": 36860 }, { "grad_norm": 0.36761751770973206, "learning_rate": 6.052560852206232e-05, "loss": 0.0357, "step": 36870 }, { "grad_norm": 0.5044904947280884, "learning_rate": 6.05054024033193e-05, "loss": 0.0424, "step": 36880 }, { "grad_norm": 0.3638826906681061, "learning_rate": 6.048519448949032e-05, "loss": 0.0347, "step": 36890 }, { "grad_norm": 0.43500009179115295, "learning_rate": 6.046498478402839e-05, "loss": 0.0446, "step": 36900 }, { "grad_norm": 0.3370830714702606, "learning_rate": 6.044477329038677e-05, "loss": 0.0372, "step": 36910 }, { "grad_norm": 0.3800532817840576, "learning_rate": 6.042456001201906e-05, "loss": 0.0362, "step": 36920 }, { "grad_norm": 0.34446945786476135, "learning_rate": 6.040434495237917e-05, "loss": 0.0392, "step": 36930 }, { "grad_norm": 0.3821319341659546, "learning_rate": 6.0384128114921256e-05, "loss": 0.0403, "step": 36940 }, { "grad_norm": 0.3451218008995056, "learning_rate": 6.036390950309987e-05, "loss": 0.0272, "step": 36950 }, { "grad_norm": 0.294948935508728, "learning_rate": 6.0343689120369805e-05, "loss": 0.0365, "step": 36960 }, { "grad_norm": 0.27618345618247986, "learning_rate": 6.032346697018616e-05, "loss": 0.0277, "step": 36970 }, { "grad_norm": 0.4187869131565094, "learning_rate": 6.0303243056004375e-05, "loss": 0.0567, "step": 36980 }, { "grad_norm": 0.42622172832489014, "learning_rate": 6.0283017381280136e-05, "loss": 0.0383, "step": 36990 }, { "grad_norm": 0.2941015362739563, "learning_rate": 6.026278994946948e-05, "loss": 0.0297, "step": 37000 }, { "grad_norm": 0.3306584060192108, "learning_rate": 6.02425607640287e-05, "loss": 0.037, "step": 37010 }, { "grad_norm": 0.3094938397407532, "learning_rate": 6.022232982841441e-05, "loss": 0.0366, "step": 37020 }, { "grad_norm": 0.3264892101287842, "learning_rate": 6.020209714608355e-05, "loss": 0.034, "step": 37030 }, { "grad_norm": 0.4130893051624298, "learning_rate": 6.018186272049331e-05, "loss": 0.0348, "step": 37040 }, { "grad_norm": 0.3100169897079468, "learning_rate": 6.01616265551012e-05, "loss": 0.0379, "step": 37050 }, { "grad_norm": 0.3627271056175232, "learning_rate": 6.014138865336503e-05, "loss": 0.0369, "step": 37060 }, { "grad_norm": 0.47036412358283997, "learning_rate": 6.0121149018742905e-05, "loss": 0.037, "step": 37070 }, { "grad_norm": 0.20691023766994476, "learning_rate": 6.010090765469325e-05, "loss": 0.0291, "step": 37080 }, { "grad_norm": 0.39475786685943604, "learning_rate": 6.008066456467473e-05, "loss": 0.0304, "step": 37090 }, { "grad_norm": 0.26822930574417114, "learning_rate": 6.0060419752146335e-05, "loss": 0.0328, "step": 37100 }, { "grad_norm": 0.5084847211837769, "learning_rate": 6.0040173220567353e-05, "loss": 0.0384, "step": 37110 }, { "grad_norm": 0.29233717918395996, "learning_rate": 6.001992497339737e-05, "loss": 0.0332, "step": 37120 }, { "grad_norm": 0.31781619787216187, "learning_rate": 5.999967501409626e-05, "loss": 0.0401, "step": 37130 }, { "grad_norm": 0.2936908006668091, "learning_rate": 5.997942334612418e-05, "loss": 0.03, "step": 37140 }, { "grad_norm": 0.2841346561908722, "learning_rate": 5.995916997294158e-05, "loss": 0.0289, "step": 37150 }, { "grad_norm": 0.24672724306583405, "learning_rate": 5.9938914898009214e-05, "loss": 0.0357, "step": 37160 }, { "grad_norm": 0.2889630198478699, "learning_rate": 5.991865812478813e-05, "loss": 0.0323, "step": 37170 }, { "grad_norm": 0.4057155251502991, "learning_rate": 5.989839965673964e-05, "loss": 0.0477, "step": 37180 }, { "grad_norm": 0.4238637685775757, "learning_rate": 5.987813949732539e-05, "loss": 0.0433, "step": 37190 }, { "grad_norm": 0.30854466557502747, "learning_rate": 5.9857877650007255e-05, "loss": 0.0363, "step": 37200 }, { "grad_norm": 0.2495735138654709, "learning_rate": 5.983761411824744e-05, "loss": 0.0313, "step": 37210 }, { "grad_norm": 0.3783765137195587, "learning_rate": 5.981734890550844e-05, "loss": 0.0403, "step": 37220 }, { "grad_norm": 0.32070237398147583, "learning_rate": 5.979708201525301e-05, "loss": 0.035, "step": 37230 }, { "grad_norm": 0.47395873069763184, "learning_rate": 5.977681345094422e-05, "loss": 0.0393, "step": 37240 }, { "grad_norm": 0.2777870297431946, "learning_rate": 5.97565432160454e-05, "loss": 0.0357, "step": 37250 }, { "grad_norm": 0.4189743101596832, "learning_rate": 5.9736271314020186e-05, "loss": 0.0325, "step": 37260 }, { "grad_norm": 0.30369260907173157, "learning_rate": 5.971599774833251e-05, "loss": 0.0369, "step": 37270 }, { "grad_norm": 0.3842667043209076, "learning_rate": 5.9695722522446525e-05, "loss": 0.0436, "step": 37280 }, { "grad_norm": 0.3682945966720581, "learning_rate": 5.9675445639826765e-05, "loss": 0.0397, "step": 37290 }, { "grad_norm": 0.2342943549156189, "learning_rate": 5.965516710393796e-05, "loss": 0.0276, "step": 37300 }, { "grad_norm": 0.28762736916542053, "learning_rate": 5.963488691824516e-05, "loss": 0.0403, "step": 37310 }, { "grad_norm": 0.467693567276001, "learning_rate": 5.96146050862137e-05, "loss": 0.0449, "step": 37320 }, { "grad_norm": 0.34321534633636475, "learning_rate": 5.959432161130919e-05, "loss": 0.0331, "step": 37330 }, { "grad_norm": 0.31499791145324707, "learning_rate": 5.9574036496997545e-05, "loss": 0.0398, "step": 37340 }, { "grad_norm": 0.4077901840209961, "learning_rate": 5.955374974674488e-05, "loss": 0.0349, "step": 37350 }, { "grad_norm": 0.41331321001052856, "learning_rate": 5.9533461364017696e-05, "loss": 0.0391, "step": 37360 }, { "grad_norm": 0.3560965359210968, "learning_rate": 5.9513171352282716e-05, "loss": 0.0401, "step": 37370 }, { "grad_norm": 0.3462299406528473, "learning_rate": 5.949287971500692e-05, "loss": 0.0369, "step": 37380 }, { "grad_norm": 0.35917866230010986, "learning_rate": 5.947258645565762e-05, "loss": 0.0384, "step": 37390 }, { "grad_norm": 0.3025048077106476, "learning_rate": 5.945229157770237e-05, "loss": 0.0295, "step": 37400 }, { "grad_norm": 0.25973665714263916, "learning_rate": 5.9431995084609006e-05, "loss": 0.0352, "step": 37410 }, { "grad_norm": 0.5398517847061157, "learning_rate": 5.941169697984564e-05, "loss": 0.0343, "step": 37420 }, { "grad_norm": 0.30037349462509155, "learning_rate": 5.9391397266880675e-05, "loss": 0.0304, "step": 37430 }, { "grad_norm": 0.46512430906295776, "learning_rate": 5.937109594918279e-05, "loss": 0.054, "step": 37440 }, { "grad_norm": 0.3280571699142456, "learning_rate": 5.9350793030220884e-05, "loss": 0.0422, "step": 37450 }, { "grad_norm": 0.2084931582212448, "learning_rate": 5.933048851346421e-05, "loss": 0.0287, "step": 37460 }, { "grad_norm": 0.3579173684120178, "learning_rate": 5.931018240238224e-05, "loss": 0.0452, "step": 37470 }, { "grad_norm": 0.3805476725101471, "learning_rate": 5.928987470044471e-05, "loss": 0.03, "step": 37480 }, { "grad_norm": 0.4201737940311432, "learning_rate": 5.9269565411121695e-05, "loss": 0.0377, "step": 37490 }, { "grad_norm": 0.5169932246208191, "learning_rate": 5.924925453788347e-05, "loss": 0.0356, "step": 37500 }, { "grad_norm": 0.30655673146247864, "learning_rate": 5.92289420842006e-05, "loss": 0.0376, "step": 37510 }, { "grad_norm": 0.49497702717781067, "learning_rate": 5.9208628053543945e-05, "loss": 0.0431, "step": 37520 }, { "grad_norm": 0.464799165725708, "learning_rate": 5.918831244938462e-05, "loss": 0.0362, "step": 37530 }, { "grad_norm": 0.4460234045982361, "learning_rate": 5.9167995275194e-05, "loss": 0.0379, "step": 37540 }, { "grad_norm": 0.29964226484298706, "learning_rate": 5.914767653444373e-05, "loss": 0.0344, "step": 37550 }, { "grad_norm": 0.46053770184516907, "learning_rate": 5.912735623060572e-05, "loss": 0.0395, "step": 37560 }, { "grad_norm": 0.43255114555358887, "learning_rate": 5.910703436715217e-05, "loss": 0.029, "step": 37570 }, { "grad_norm": 0.3136543929576874, "learning_rate": 5.908671094755552e-05, "loss": 0.0275, "step": 37580 }, { "grad_norm": 0.24918121099472046, "learning_rate": 5.906638597528851e-05, "loss": 0.0287, "step": 37590 }, { "grad_norm": 0.39796650409698486, "learning_rate": 5.9046059453824076e-05, "loss": 0.0315, "step": 37600 }, { "grad_norm": 0.42364126443862915, "learning_rate": 5.9025731386635505e-05, "loss": 0.0351, "step": 37610 }, { "grad_norm": 0.3310207426548004, "learning_rate": 5.900540177719629e-05, "loss": 0.0295, "step": 37620 }, { "grad_norm": 0.3818235993385315, "learning_rate": 5.898507062898021e-05, "loss": 0.0261, "step": 37630 }, { "grad_norm": 0.39477649331092834, "learning_rate": 5.8964737945461316e-05, "loss": 0.0347, "step": 37640 }, { "grad_norm": 0.2651137411594391, "learning_rate": 5.8944403730113885e-05, "loss": 0.032, "step": 37650 }, { "grad_norm": 0.345803439617157, "learning_rate": 5.892406798641248e-05, "loss": 0.0416, "step": 37660 }, { "grad_norm": 0.4042108952999115, "learning_rate": 5.890373071783193e-05, "loss": 0.0257, "step": 37670 }, { "grad_norm": 0.2950100600719452, "learning_rate": 5.888339192784732e-05, "loss": 0.0307, "step": 37680 }, { "grad_norm": 0.4004287123680115, "learning_rate": 5.8863051619934003e-05, "loss": 0.0312, "step": 37690 }, { "grad_norm": 0.3095647394657135, "learning_rate": 5.8842709797567554e-05, "loss": 0.03, "step": 37700 }, { "grad_norm": 0.33425793051719666, "learning_rate": 5.8822366464223855e-05, "loss": 0.0317, "step": 37710 }, { "grad_norm": 0.19850729405879974, "learning_rate": 5.880202162337901e-05, "loss": 0.0302, "step": 37720 }, { "grad_norm": 0.31944695115089417, "learning_rate": 5.8781675278509405e-05, "loss": 0.0282, "step": 37730 }, { "grad_norm": 0.32231244444847107, "learning_rate": 5.8761327433091696e-05, "loss": 0.032, "step": 37740 }, { "grad_norm": 0.3211873173713684, "learning_rate": 5.874097809060275e-05, "loss": 0.0339, "step": 37750 }, { "grad_norm": 0.5324199795722961, "learning_rate": 5.87206272545197e-05, "loss": 0.0562, "step": 37760 }, { "grad_norm": 0.357657790184021, "learning_rate": 5.8700274928319955e-05, "loss": 0.0365, "step": 37770 }, { "grad_norm": 0.429069846868515, "learning_rate": 5.867992111548118e-05, "loss": 0.0342, "step": 37780 }, { "grad_norm": 0.2617952227592468, "learning_rate": 5.865956581948131e-05, "loss": 0.0256, "step": 37790 }, { "grad_norm": 0.33625125885009766, "learning_rate": 5.863920904379845e-05, "loss": 0.0279, "step": 37800 }, { "grad_norm": 0.2905385196208954, "learning_rate": 5.8618850791911064e-05, "loss": 0.0343, "step": 37810 }, { "grad_norm": 0.4277837574481964, "learning_rate": 5.859849106729779e-05, "loss": 0.0327, "step": 37820 }, { "grad_norm": 0.5173076391220093, "learning_rate": 5.857812987343758e-05, "loss": 0.0338, "step": 37830 }, { "grad_norm": 0.4078936278820038, "learning_rate": 5.855776721380957e-05, "loss": 0.0362, "step": 37840 }, { "grad_norm": 0.2813391089439392, "learning_rate": 5.8537403091893217e-05, "loss": 0.0288, "step": 37850 }, { "grad_norm": 0.4137971103191376, "learning_rate": 5.851703751116816e-05, "loss": 0.0291, "step": 37860 }, { "grad_norm": 0.3745587468147278, "learning_rate": 5.8496670475114336e-05, "loss": 0.036, "step": 37870 }, { "grad_norm": 0.4488217234611511, "learning_rate": 5.84763019872119e-05, "loss": 0.0375, "step": 37880 }, { "grad_norm": 0.31519681215286255, "learning_rate": 5.845593205094131e-05, "loss": 0.0269, "step": 37890 }, { "grad_norm": 0.2836901843547821, "learning_rate": 5.843556066978318e-05, "loss": 0.0369, "step": 37900 }, { "grad_norm": 0.3003327548503876, "learning_rate": 5.8415187847218455e-05, "loss": 0.0397, "step": 37910 }, { "grad_norm": 0.35059863328933716, "learning_rate": 5.839481358672827e-05, "loss": 0.0321, "step": 37920 }, { "grad_norm": 0.3311695456504822, "learning_rate": 5.837443789179407e-05, "loss": 0.0371, "step": 37930 }, { "grad_norm": 0.41631177067756653, "learning_rate": 5.8354060765897445e-05, "loss": 0.0328, "step": 37940 }, { "grad_norm": 0.42839741706848145, "learning_rate": 5.8333682212520334e-05, "loss": 0.0516, "step": 37950 }, { "grad_norm": 0.2444884032011032, "learning_rate": 5.831330223514486e-05, "loss": 0.0345, "step": 37960 }, { "grad_norm": 0.23818068206310272, "learning_rate": 5.8292920837253396e-05, "loss": 0.0396, "step": 37970 }, { "grad_norm": 0.29656723141670227, "learning_rate": 5.827253802232857e-05, "loss": 0.0324, "step": 37980 }, { "grad_norm": 0.6155910491943359, "learning_rate": 5.825215379385327e-05, "loss": 0.0397, "step": 37990 }, { "grad_norm": 0.47043099999427795, "learning_rate": 5.823176815531057e-05, "loss": 0.0499, "step": 38000 }, { "grad_norm": 0.21027012169361115, "learning_rate": 5.8211381110183826e-05, "loss": 0.0374, "step": 38010 }, { "grad_norm": 0.2780565023422241, "learning_rate": 5.8190992661956645e-05, "loss": 0.0255, "step": 38020 }, { "grad_norm": 0.29381924867630005, "learning_rate": 5.817060281411284e-05, "loss": 0.0357, "step": 38030 }, { "grad_norm": 0.3656598925590515, "learning_rate": 5.815021157013647e-05, "loss": 0.0384, "step": 38040 }, { "grad_norm": 0.35453927516937256, "learning_rate": 5.8129818933511856e-05, "loss": 0.0413, "step": 38050 }, { "grad_norm": 0.3541162312030792, "learning_rate": 5.8109424907723544e-05, "loss": 0.0445, "step": 38060 }, { "grad_norm": 0.2727401852607727, "learning_rate": 5.80890294962563e-05, "loss": 0.03, "step": 38070 }, { "grad_norm": 0.4663982391357422, "learning_rate": 5.806863270259515e-05, "loss": 0.045, "step": 38080 }, { "grad_norm": 0.3501894176006317, "learning_rate": 5.804823453022536e-05, "loss": 0.0375, "step": 38090 }, { "grad_norm": 0.34620314836502075, "learning_rate": 5.80278349826324e-05, "loss": 0.0413, "step": 38100 }, { "grad_norm": 0.32268470525741577, "learning_rate": 5.8007434063302014e-05, "loss": 0.038, "step": 38110 }, { "grad_norm": 0.2741183042526245, "learning_rate": 5.7987031775720136e-05, "loss": 0.0325, "step": 38120 }, { "grad_norm": 0.32209786772727966, "learning_rate": 5.7966628123372976e-05, "loss": 0.0456, "step": 38130 }, { "grad_norm": 0.532467246055603, "learning_rate": 5.7946223109746956e-05, "loss": 0.0403, "step": 38140 }, { "grad_norm": 0.37740710377693176, "learning_rate": 5.7925816738328754e-05, "loss": 0.0296, "step": 38150 }, { "grad_norm": 0.3337598741054535, "learning_rate": 5.790540901260521e-05, "loss": 0.035, "step": 38160 }, { "grad_norm": 0.24894890189170837, "learning_rate": 5.788499993606351e-05, "loss": 0.0288, "step": 38170 }, { "grad_norm": 0.6989966630935669, "learning_rate": 5.786458951219096e-05, "loss": 0.0387, "step": 38180 }, { "grad_norm": 0.24671570956707, "learning_rate": 5.784417774447517e-05, "loss": 0.0356, "step": 38190 }, { "grad_norm": 0.42605122923851013, "learning_rate": 5.782376463640393e-05, "loss": 0.035, "step": 38200 }, { "grad_norm": 0.4542469382286072, "learning_rate": 5.780335019146531e-05, "loss": 0.037, "step": 38210 }, { "grad_norm": 0.28732967376708984, "learning_rate": 5.778293441314755e-05, "loss": 0.0345, "step": 38220 }, { "grad_norm": 0.3263244330883026, "learning_rate": 5.776251730493917e-05, "loss": 0.0507, "step": 38230 }, { "grad_norm": 0.26268020272254944, "learning_rate": 5.774209887032887e-05, "loss": 0.0368, "step": 38240 }, { "grad_norm": 0.25676003098487854, "learning_rate": 5.772167911280565e-05, "loss": 0.0288, "step": 38250 }, { "grad_norm": 0.3066030740737915, "learning_rate": 5.770125803585864e-05, "loss": 0.0335, "step": 38260 }, { "grad_norm": 0.5286359786987305, "learning_rate": 5.768083564297726e-05, "loss": 0.056, "step": 38270 }, { "grad_norm": 0.3054659366607666, "learning_rate": 5.766041193765114e-05, "loss": 0.0365, "step": 38280 }, { "grad_norm": 0.36658990383148193, "learning_rate": 5.763998692337015e-05, "loss": 0.0328, "step": 38290 }, { "grad_norm": 0.5544977188110352, "learning_rate": 5.761956060362433e-05, "loss": 0.0318, "step": 38300 }, { "grad_norm": 0.30548304319381714, "learning_rate": 5.7599132981904e-05, "loss": 0.0346, "step": 38310 }, { "grad_norm": 0.3135985732078552, "learning_rate": 5.75787040616997e-05, "loss": 0.0355, "step": 38320 }, { "grad_norm": 0.2881709039211273, "learning_rate": 5.755827384650212e-05, "loss": 0.0353, "step": 38330 }, { "grad_norm": 0.3748873174190521, "learning_rate": 5.753784233980228e-05, "loss": 0.0377, "step": 38340 }, { "grad_norm": 0.41563642024993896, "learning_rate": 5.751740954509135e-05, "loss": 0.0265, "step": 38350 }, { "grad_norm": 0.2874599099159241, "learning_rate": 5.7496975465860715e-05, "loss": 0.0253, "step": 38360 }, { "grad_norm": 1.6051427125930786, "learning_rate": 5.747654010560202e-05, "loss": 0.0332, "step": 38370 }, { "grad_norm": 0.7840553522109985, "learning_rate": 5.7456103467807097e-05, "loss": 0.0445, "step": 38380 }, { "grad_norm": 0.4936456084251404, "learning_rate": 5.7435665555968046e-05, "loss": 0.0346, "step": 38390 }, { "grad_norm": 0.29553329944610596, "learning_rate": 5.74152263735771e-05, "loss": 0.0406, "step": 38400 }, { "grad_norm": 0.45666879415512085, "learning_rate": 5.739478592412677e-05, "loss": 0.0346, "step": 38410 }, { "grad_norm": 0.5454400181770325, "learning_rate": 5.7374344211109766e-05, "loss": 0.0423, "step": 38420 }, { "grad_norm": 0.3900587260723114, "learning_rate": 5.735390123801904e-05, "loss": 0.0389, "step": 38430 }, { "grad_norm": 0.4669129550457001, "learning_rate": 5.7333457008347704e-05, "loss": 0.0439, "step": 38440 }, { "grad_norm": 0.4936992824077606, "learning_rate": 5.7313011525589156e-05, "loss": 0.0377, "step": 38450 }, { "grad_norm": 0.32543981075286865, "learning_rate": 5.729256479323694e-05, "loss": 0.0338, "step": 38460 }, { "grad_norm": 0.36337390542030334, "learning_rate": 5.727211681478485e-05, "loss": 0.0408, "step": 38470 }, { "grad_norm": 0.43021276593208313, "learning_rate": 5.7251667593726886e-05, "loss": 0.0453, "step": 38480 }, { "grad_norm": 0.3162735104560852, "learning_rate": 5.723121713355728e-05, "loss": 0.0421, "step": 38490 }, { "grad_norm": 0.3775591254234314, "learning_rate": 5.721076543777044e-05, "loss": 0.0363, "step": 38500 }, { "grad_norm": 0.33459362387657166, "learning_rate": 5.7190312509860986e-05, "loss": 0.0316, "step": 38510 }, { "grad_norm": 0.5948604941368103, "learning_rate": 5.716985835332379e-05, "loss": 0.0384, "step": 38520 }, { "grad_norm": 0.32119593024253845, "learning_rate": 5.714940297165389e-05, "loss": 0.0365, "step": 38530 }, { "grad_norm": 0.38366809487342834, "learning_rate": 5.712894636834656e-05, "loss": 0.0328, "step": 38540 }, { "grad_norm": 0.3140579164028168, "learning_rate": 5.7108488546897276e-05, "loss": 0.0388, "step": 38550 }, { "grad_norm": 0.25659868121147156, "learning_rate": 5.708802951080172e-05, "loss": 0.0325, "step": 38560 }, { "grad_norm": 0.3248683512210846, "learning_rate": 5.706756926355576e-05, "loss": 0.0369, "step": 38570 }, { "grad_norm": 0.2656831443309784, "learning_rate": 5.704710780865554e-05, "loss": 0.03, "step": 38580 }, { "grad_norm": 0.3191818296909332, "learning_rate": 5.7026645149597325e-05, "loss": 0.0296, "step": 38590 }, { "grad_norm": 0.49703532457351685, "learning_rate": 5.700618128987764e-05, "loss": 0.0376, "step": 38600 }, { "grad_norm": 0.19366350769996643, "learning_rate": 5.698571623299317e-05, "loss": 0.0282, "step": 38610 }, { "grad_norm": 0.2880598306655884, "learning_rate": 5.696524998244086e-05, "loss": 0.0289, "step": 38620 }, { "grad_norm": 0.5127614140510559, "learning_rate": 5.6944782541717836e-05, "loss": 0.0458, "step": 38630 }, { "grad_norm": 0.383620023727417, "learning_rate": 5.69243139143214e-05, "loss": 0.0294, "step": 38640 }, { "grad_norm": 0.6170369386672974, "learning_rate": 5.6903844103749125e-05, "loss": 0.043, "step": 38650 }, { "grad_norm": 0.3553656041622162, "learning_rate": 5.688337311349869e-05, "loss": 0.0304, "step": 38660 }, { "grad_norm": 0.2775672376155853, "learning_rate": 5.6862900947068074e-05, "loss": 0.0327, "step": 38670 }, { "grad_norm": 0.4585508704185486, "learning_rate": 5.6842427607955374e-05, "loss": 0.035, "step": 38680 }, { "grad_norm": 0.42006099224090576, "learning_rate": 5.682195309965893e-05, "loss": 0.0262, "step": 38690 }, { "grad_norm": 0.4142446517944336, "learning_rate": 5.6801477425677294e-05, "loss": 0.0331, "step": 38700 }, { "grad_norm": 0.2846263349056244, "learning_rate": 5.678100058950917e-05, "loss": 0.042, "step": 38710 }, { "grad_norm": 0.4771672189235687, "learning_rate": 5.676052259465352e-05, "loss": 0.0372, "step": 38720 }, { "grad_norm": 0.20331275463104248, "learning_rate": 5.674004344460945e-05, "loss": 0.0295, "step": 38730 }, { "grad_norm": 0.34640222787857056, "learning_rate": 5.6719563142876295e-05, "loss": 0.0294, "step": 38740 }, { "grad_norm": 0.38950490951538086, "learning_rate": 5.669908169295359e-05, "loss": 0.0287, "step": 38750 }, { "grad_norm": 0.3922334611415863, "learning_rate": 5.667859909834105e-05, "loss": 0.031, "step": 38760 }, { "grad_norm": 0.41909873485565186, "learning_rate": 5.6658115362538565e-05, "loss": 0.0355, "step": 38770 }, { "grad_norm": 0.5298258066177368, "learning_rate": 5.663763048904628e-05, "loss": 0.0333, "step": 38780 }, { "grad_norm": 0.4149409532546997, "learning_rate": 5.661714448136447e-05, "loss": 0.0301, "step": 38790 }, { "grad_norm": 0.6743105053901672, "learning_rate": 5.659665734299366e-05, "loss": 0.0305, "step": 38800 }, { "grad_norm": 0.32440781593322754, "learning_rate": 5.6576169077434516e-05, "loss": 0.0357, "step": 38810 }, { "grad_norm": 0.18818879127502441, "learning_rate": 5.6555679688187944e-05, "loss": 0.0366, "step": 38820 }, { "grad_norm": 0.49419400095939636, "learning_rate": 5.6535189178755e-05, "loss": 0.0335, "step": 38830 }, { "grad_norm": 0.4487735629081726, "learning_rate": 5.651469755263698e-05, "loss": 0.0379, "step": 38840 }, { "grad_norm": 0.4451219141483307, "learning_rate": 5.6494204813335316e-05, "loss": 0.0389, "step": 38850 }, { "grad_norm": 0.30507224798202515, "learning_rate": 5.647371096435168e-05, "loss": 0.0291, "step": 38860 }, { "grad_norm": 0.41101589798927307, "learning_rate": 5.645321600918788e-05, "loss": 0.0281, "step": 38870 }, { "grad_norm": 0.42725083231925964, "learning_rate": 5.643271995134597e-05, "loss": 0.04, "step": 38880 }, { "grad_norm": 0.34966522455215454, "learning_rate": 5.641222279432814e-05, "loss": 0.0338, "step": 38890 }, { "grad_norm": 0.3880998194217682, "learning_rate": 5.6391724541636834e-05, "loss": 0.0304, "step": 38900 }, { "grad_norm": 0.4650860130786896, "learning_rate": 5.6371225196774605e-05, "loss": 0.0339, "step": 38910 }, { "grad_norm": 0.3557707965373993, "learning_rate": 5.635072476324423e-05, "loss": 0.0417, "step": 38920 }, { "grad_norm": 0.2581176161766052, "learning_rate": 5.63302232445487e-05, "loss": 0.028, "step": 38930 }, { "grad_norm": 0.2867545187473297, "learning_rate": 5.6309720644191144e-05, "loss": 0.0364, "step": 38940 }, { "grad_norm": 0.36922699213027954, "learning_rate": 5.628921696567491e-05, "loss": 0.0313, "step": 38950 }, { "grad_norm": 0.2141343355178833, "learning_rate": 5.62687122125035e-05, "loss": 0.0321, "step": 38960 }, { "grad_norm": 0.4325307011604309, "learning_rate": 5.624820638818062e-05, "loss": 0.0343, "step": 38970 }, { "grad_norm": 0.3206692636013031, "learning_rate": 5.6227699496210164e-05, "loss": 0.0396, "step": 38980 }, { "grad_norm": 0.36844825744628906, "learning_rate": 5.6207191540096195e-05, "loss": 0.0344, "step": 38990 }, { "grad_norm": 0.5713405609130859, "learning_rate": 5.618668252334296e-05, "loss": 0.0396, "step": 39000 }, { "grad_norm": 0.20688115060329437, "learning_rate": 5.616617244945488e-05, "loss": 0.0266, "step": 39010 }, { "grad_norm": 0.3069442808628082, "learning_rate": 5.614566132193656e-05, "loss": 0.0297, "step": 39020 }, { "grad_norm": 0.3217885494232178, "learning_rate": 5.612514914429282e-05, "loss": 0.0357, "step": 39030 }, { "grad_norm": 0.33663323521614075, "learning_rate": 5.610463592002863e-05, "loss": 0.0326, "step": 39040 }, { "grad_norm": 0.3713063895702362, "learning_rate": 5.608412165264909e-05, "loss": 0.0318, "step": 39050 }, { "grad_norm": 0.2904558479785919, "learning_rate": 5.606360634565959e-05, "loss": 0.0242, "step": 39060 }, { "grad_norm": 0.4080592095851898, "learning_rate": 5.604309000256559e-05, "loss": 0.0368, "step": 39070 }, { "grad_norm": 0.4403197765350342, "learning_rate": 5.6022572626872785e-05, "loss": 0.0293, "step": 39080 }, { "grad_norm": 0.38177934288978577, "learning_rate": 5.600205422208704e-05, "loss": 0.0348, "step": 39090 }, { "grad_norm": 0.3542807400226593, "learning_rate": 5.59815347917144e-05, "loss": 0.0311, "step": 39100 }, { "grad_norm": 0.33730992674827576, "learning_rate": 5.596101433926103e-05, "loss": 0.0314, "step": 39110 }, { "grad_norm": 0.39097630977630615, "learning_rate": 5.5940492868233364e-05, "loss": 0.0249, "step": 39120 }, { "grad_norm": 0.23644472658634186, "learning_rate": 5.591997038213793e-05, "loss": 0.0288, "step": 39130 }, { "grad_norm": 0.2210102081298828, "learning_rate": 5.5899446884481475e-05, "loss": 0.0242, "step": 39140 }, { "grad_norm": 0.31840574741363525, "learning_rate": 5.5878922378770906e-05, "loss": 0.049, "step": 39150 }, { "grad_norm": 0.24251841008663177, "learning_rate": 5.5858396868513285e-05, "loss": 0.0268, "step": 39160 }, { "grad_norm": 0.4785861670970917, "learning_rate": 5.583787035721586e-05, "loss": 0.0307, "step": 39170 }, { "grad_norm": 0.20945824682712555, "learning_rate": 5.581734284838606e-05, "loss": 0.029, "step": 39180 }, { "grad_norm": 0.4587981402873993, "learning_rate": 5.579681434553147e-05, "loss": 0.0382, "step": 39190 }, { "grad_norm": 0.32221201062202454, "learning_rate": 5.5776284852159854e-05, "loss": 0.0287, "step": 39200 }, { "grad_norm": 0.37911006808280945, "learning_rate": 5.575575437177913e-05, "loss": 0.0456, "step": 39210 }, { "grad_norm": 0.45002204179763794, "learning_rate": 5.573522290789742e-05, "loss": 0.0311, "step": 39220 }, { "grad_norm": 0.333015114068985, "learning_rate": 5.571469046402298e-05, "loss": 0.0366, "step": 39230 }, { "grad_norm": 0.2568771541118622, "learning_rate": 5.5694157043664205e-05, "loss": 0.0362, "step": 39240 }, { "grad_norm": 0.33882540464401245, "learning_rate": 5.567362265032975e-05, "loss": 0.0287, "step": 39250 }, { "grad_norm": 0.29734376072883606, "learning_rate": 5.565308728752836e-05, "loss": 0.0369, "step": 39260 }, { "grad_norm": 0.2799386978149414, "learning_rate": 5.5632550958768956e-05, "loss": 0.0294, "step": 39270 }, { "grad_norm": 0.34602370858192444, "learning_rate": 5.5612013667560636e-05, "loss": 0.0366, "step": 39280 }, { "grad_norm": 0.4738324284553528, "learning_rate": 5.5591475417412676e-05, "loss": 0.045, "step": 39290 }, { "grad_norm": 0.31296437978744507, "learning_rate": 5.557093621183451e-05, "loss": 0.0312, "step": 39300 }, { "grad_norm": 0.2627897560596466, "learning_rate": 5.55503960543357e-05, "loss": 0.0277, "step": 39310 }, { "grad_norm": 0.3314532935619354, "learning_rate": 5.552985494842601e-05, "loss": 0.0312, "step": 39320 }, { "grad_norm": 0.25172632932662964, "learning_rate": 5.550931289761534e-05, "loss": 0.035, "step": 39330 }, { "grad_norm": 0.3669332265853882, "learning_rate": 5.548876990541378e-05, "loss": 0.0321, "step": 39340 }, { "grad_norm": 0.4523698687553406, "learning_rate": 5.5468225975331564e-05, "loss": 0.0382, "step": 39350 }, { "grad_norm": 0.4516647756099701, "learning_rate": 5.544768111087909e-05, "loss": 0.0351, "step": 39360 }, { "grad_norm": 0.3244909644126892, "learning_rate": 5.54271353155669e-05, "loss": 0.0399, "step": 39370 }, { "grad_norm": 0.2515568137168884, "learning_rate": 5.5406588592905715e-05, "loss": 0.0291, "step": 39380 }, { "grad_norm": 0.2889227867126465, "learning_rate": 5.5386040946406416e-05, "loss": 0.0345, "step": 39390 }, { "grad_norm": 0.3047686219215393, "learning_rate": 5.536549237958004e-05, "loss": 0.0381, "step": 39400 }, { "grad_norm": 0.4690896272659302, "learning_rate": 5.5344942895937744e-05, "loss": 0.0436, "step": 39410 }, { "grad_norm": 0.2854244112968445, "learning_rate": 5.5324392498990904e-05, "loss": 0.0306, "step": 39420 }, { "grad_norm": 0.3968675434589386, "learning_rate": 5.5303841192251e-05, "loss": 0.0265, "step": 39430 }, { "grad_norm": 0.3448735773563385, "learning_rate": 5.52832889792297e-05, "loss": 0.0296, "step": 39440 }, { "grad_norm": 0.3004032075405121, "learning_rate": 5.526273586343881e-05, "loss": 0.0266, "step": 39450 }, { "grad_norm": 0.4442402422428131, "learning_rate": 5.5242181848390306e-05, "loss": 0.0305, "step": 39460 }, { "grad_norm": 0.3599540889263153, "learning_rate": 5.5221626937596285e-05, "loss": 0.0299, "step": 39470 }, { "grad_norm": 0.2407539188861847, "learning_rate": 5.520107113456903e-05, "loss": 0.0259, "step": 39480 }, { "grad_norm": 0.3794853687286377, "learning_rate": 5.5180514442820974e-05, "loss": 0.04, "step": 39490 }, { "grad_norm": 0.27469244599342346, "learning_rate": 5.515995686586469e-05, "loss": 0.0236, "step": 39500 }, { "grad_norm": 0.28405511379241943, "learning_rate": 5.5139398407212916e-05, "loss": 0.0261, "step": 39510 }, { "grad_norm": 0.276443749666214, "learning_rate": 5.511883907037849e-05, "loss": 0.0306, "step": 39520 }, { "grad_norm": 0.3831934928894043, "learning_rate": 5.509827885887449e-05, "loss": 0.035, "step": 39530 }, { "grad_norm": 0.35739096999168396, "learning_rate": 5.507771777621406e-05, "loss": 0.031, "step": 39540 }, { "grad_norm": 0.42074528336524963, "learning_rate": 5.505715582591052e-05, "loss": 0.0475, "step": 39550 }, { "grad_norm": 0.444074809551239, "learning_rate": 5.50365930114774e-05, "loss": 0.0393, "step": 39560 }, { "grad_norm": 0.17405687272548676, "learning_rate": 5.5016029336428255e-05, "loss": 0.0187, "step": 39570 }, { "grad_norm": 0.2927856743335724, "learning_rate": 5.49954648042769e-05, "loss": 0.0292, "step": 39580 }, { "grad_norm": 0.26972484588623047, "learning_rate": 5.4974899418537226e-05, "loss": 0.0335, "step": 39590 }, { "grad_norm": 0.33170682191848755, "learning_rate": 5.4954333182723316e-05, "loss": 0.0303, "step": 39600 }, { "grad_norm": 0.6568697094917297, "learning_rate": 5.493376610034937e-05, "loss": 0.0276, "step": 39610 }, { "grad_norm": 0.3988064229488373, "learning_rate": 5.4913198174929735e-05, "loss": 0.0274, "step": 39620 }, { "grad_norm": 0.28339481353759766, "learning_rate": 5.489262940997889e-05, "loss": 0.0349, "step": 39630 }, { "grad_norm": 0.33475854992866516, "learning_rate": 5.487205980901151e-05, "loss": 0.0464, "step": 39640 }, { "grad_norm": 1.0824029445648193, "learning_rate": 5.485148937554234e-05, "loss": 0.0485, "step": 39650 }, { "grad_norm": 0.20446228981018066, "learning_rate": 5.483091811308635e-05, "loss": 0.0273, "step": 39660 }, { "grad_norm": 0.3580886125564575, "learning_rate": 5.4810346025158564e-05, "loss": 0.0349, "step": 39670 }, { "grad_norm": 0.3955991268157959, "learning_rate": 5.478977311527421e-05, "loss": 0.0261, "step": 39680 }, { "grad_norm": 0.4817716181278229, "learning_rate": 5.476919938694863e-05, "loss": 0.0276, "step": 39690 }, { "grad_norm": 0.30168020725250244, "learning_rate": 5.474862484369733e-05, "loss": 0.0282, "step": 39700 }, { "grad_norm": 0.3785935640335083, "learning_rate": 5.472804948903589e-05, "loss": 0.0315, "step": 39710 }, { "grad_norm": 0.31107431650161743, "learning_rate": 5.470747332648013e-05, "loss": 0.0266, "step": 39720 }, { "grad_norm": 0.28169894218444824, "learning_rate": 5.468689635954591e-05, "loss": 0.0465, "step": 39730 }, { "grad_norm": 0.31367576122283936, "learning_rate": 5.46663185917493e-05, "loss": 0.0381, "step": 39740 }, { "grad_norm": 0.3506675362586975, "learning_rate": 5.464574002660645e-05, "loss": 0.033, "step": 39750 }, { "grad_norm": 0.22630541026592255, "learning_rate": 5.4625160667633724e-05, "loss": 0.0316, "step": 39760 }, { "grad_norm": 0.29482561349868774, "learning_rate": 5.4604580518347505e-05, "loss": 0.0302, "step": 39770 }, { "grad_norm": 0.481052428483963, "learning_rate": 5.458399958226442e-05, "loss": 0.0449, "step": 39780 }, { "grad_norm": 0.4695395529270172, "learning_rate": 5.456341786290119e-05, "loss": 0.0314, "step": 39790 }, { "grad_norm": 0.3720542788505554, "learning_rate": 5.454283536377465e-05, "loss": 0.0351, "step": 39800 }, { "grad_norm": 0.3427899181842804, "learning_rate": 5.452225208840179e-05, "loss": 0.0372, "step": 39810 }, { "grad_norm": 0.25580447912216187, "learning_rate": 5.450166804029975e-05, "loss": 0.0403, "step": 39820 }, { "grad_norm": 0.3865721821784973, "learning_rate": 5.448108322298574e-05, "loss": 0.0301, "step": 39830 }, { "grad_norm": 0.36900973320007324, "learning_rate": 5.446049763997717e-05, "loss": 0.0301, "step": 39840 }, { "grad_norm": 0.47057247161865234, "learning_rate": 5.4439911294791546e-05, "loss": 0.0273, "step": 39850 }, { "grad_norm": 0.42795974016189575, "learning_rate": 5.441932419094652e-05, "loss": 0.0277, "step": 39860 }, { "grad_norm": 0.24890805780887604, "learning_rate": 5.439873633195985e-05, "loss": 0.0305, "step": 39870 }, { "grad_norm": 0.4441549479961395, "learning_rate": 5.437814772134947e-05, "loss": 0.0418, "step": 39880 }, { "grad_norm": 0.31821879744529724, "learning_rate": 5.4357558362633366e-05, "loss": 0.0274, "step": 39890 }, { "grad_norm": 0.28148597478866577, "learning_rate": 5.4336968259329726e-05, "loss": 0.0219, "step": 39900 }, { "grad_norm": 0.3121783435344696, "learning_rate": 5.431637741495681e-05, "loss": 0.0355, "step": 39910 }, { "grad_norm": 0.6635969877243042, "learning_rate": 5.429578583303307e-05, "loss": 0.0326, "step": 39920 }, { "grad_norm": 0.26996150612831116, "learning_rate": 5.427519351707701e-05, "loss": 0.0401, "step": 39930 }, { "grad_norm": 0.36687731742858887, "learning_rate": 5.42546004706073e-05, "loss": 0.0351, "step": 39940 }, { "grad_norm": 0.4283522069454193, "learning_rate": 5.4234006697142735e-05, "loss": 0.0319, "step": 39950 }, { "grad_norm": 0.28473979234695435, "learning_rate": 5.421341220020224e-05, "loss": 0.0275, "step": 39960 }, { "grad_norm": 0.19746792316436768, "learning_rate": 5.419281698330482e-05, "loss": 0.0466, "step": 39970 }, { "grad_norm": 0.29907703399658203, "learning_rate": 5.4172221049969665e-05, "loss": 0.0306, "step": 39980 }, { "grad_norm": 0.32268938422203064, "learning_rate": 5.415162440371604e-05, "loss": 0.031, "step": 39990 }, { "grad_norm": 0.3762607276439667, "learning_rate": 5.413102704806334e-05, "loss": 0.0387, "step": 40000 }, { "grad_norm": 0.377129465341568, "learning_rate": 5.41104289865311e-05, "loss": 0.0313, "step": 40010 }, { "grad_norm": 0.3301573395729065, "learning_rate": 5.408983022263898e-05, "loss": 0.0298, "step": 40020 }, { "grad_norm": 0.37831762433052063, "learning_rate": 5.406923075990673e-05, "loss": 0.0347, "step": 40030 }, { "grad_norm": 0.3968008756637573, "learning_rate": 5.404863060185423e-05, "loss": 0.0265, "step": 40040 }, { "grad_norm": 0.5083783268928528, "learning_rate": 5.40280297520015e-05, "loss": 0.0378, "step": 40050 }, { "grad_norm": 0.3050060272216797, "learning_rate": 5.400742821386865e-05, "loss": 0.0307, "step": 40060 }, { "grad_norm": 0.2729138731956482, "learning_rate": 5.398682599097592e-05, "loss": 0.0342, "step": 40070 }, { "grad_norm": 0.4279729127883911, "learning_rate": 5.396622308684367e-05, "loss": 0.0357, "step": 40080 }, { "grad_norm": 0.4285860061645508, "learning_rate": 5.394561950499236e-05, "loss": 0.034, "step": 40090 }, { "grad_norm": 0.21397773921489716, "learning_rate": 5.39250152489426e-05, "loss": 0.0297, "step": 40100 }, { "grad_norm": 0.3118903338909149, "learning_rate": 5.390441032221507e-05, "loss": 0.0289, "step": 40110 }, { "grad_norm": 0.3961353898048401, "learning_rate": 5.388380472833062e-05, "loss": 0.0386, "step": 40120 }, { "grad_norm": 0.3492627441883087, "learning_rate": 5.386319847081016e-05, "loss": 0.0334, "step": 40130 }, { "grad_norm": 0.4787634015083313, "learning_rate": 5.384259155317473e-05, "loss": 0.0429, "step": 40140 }, { "grad_norm": 0.4272725284099579, "learning_rate": 5.38219839789455e-05, "loss": 0.0394, "step": 40150 }, { "grad_norm": 0.2554149627685547, "learning_rate": 5.380137575164376e-05, "loss": 0.0363, "step": 40160 }, { "grad_norm": 0.3445258140563965, "learning_rate": 5.378076687479085e-05, "loss": 0.0315, "step": 40170 }, { "grad_norm": 0.3428848087787628, "learning_rate": 5.3760157351908305e-05, "loss": 0.0345, "step": 40180 }, { "grad_norm": 0.45091718435287476, "learning_rate": 5.373954718651768e-05, "loss": 0.036, "step": 40190 }, { "grad_norm": 0.3806847035884857, "learning_rate": 5.371893638214074e-05, "loss": 0.0295, "step": 40200 }, { "grad_norm": 0.42308124899864197, "learning_rate": 5.369832494229927e-05, "loss": 0.0342, "step": 40210 }, { "grad_norm": 0.44990572333335876, "learning_rate": 5.367771287051524e-05, "loss": 0.0424, "step": 40220 }, { "grad_norm": 0.24067053198814392, "learning_rate": 5.365710017031065e-05, "loss": 0.0243, "step": 40230 }, { "grad_norm": 0.34335362911224365, "learning_rate": 5.3636486845207654e-05, "loss": 0.0345, "step": 40240 }, { "grad_norm": 0.36622947454452515, "learning_rate": 5.3615872898728526e-05, "loss": 0.029, "step": 40250 }, { "grad_norm": 0.4881168007850647, "learning_rate": 5.3595258334395614e-05, "loss": 0.0323, "step": 40260 }, { "grad_norm": 0.3083716630935669, "learning_rate": 5.3574643155731396e-05, "loss": 0.0384, "step": 40270 }, { "grad_norm": 0.2629639506340027, "learning_rate": 5.3554027366258405e-05, "loss": 0.0337, "step": 40280 }, { "grad_norm": 0.2658032476902008, "learning_rate": 5.3533410969499355e-05, "loss": 0.0282, "step": 40290 }, { "grad_norm": 0.6094432473182678, "learning_rate": 5.3512793968977e-05, "loss": 0.031, "step": 40300 }, { "grad_norm": 0.4322122633457184, "learning_rate": 5.349217636821423e-05, "loss": 0.0324, "step": 40310 }, { "grad_norm": 0.33209434151649475, "learning_rate": 5.347155817073405e-05, "loss": 0.0281, "step": 40320 }, { "grad_norm": 0.27011430263519287, "learning_rate": 5.34509393800595e-05, "loss": 0.0308, "step": 40330 }, { "grad_norm": 0.3055100440979004, "learning_rate": 5.343031999971381e-05, "loss": 0.0266, "step": 40340 }, { "grad_norm": 0.3219704031944275, "learning_rate": 5.340970003322026e-05, "loss": 0.0312, "step": 40350 }, { "grad_norm": 0.2193479835987091, "learning_rate": 5.33890794841022e-05, "loss": 0.0294, "step": 40360 }, { "grad_norm": 0.2546006739139557, "learning_rate": 5.336845835588318e-05, "loss": 0.0289, "step": 40370 }, { "grad_norm": 0.45870229601860046, "learning_rate": 5.334783665208674e-05, "loss": 0.0291, "step": 40380 }, { "grad_norm": 0.3777989447116852, "learning_rate": 5.332721437623657e-05, "loss": 0.028, "step": 40390 }, { "grad_norm": 0.243757426738739, "learning_rate": 5.3306591531856464e-05, "loss": 0.0312, "step": 40400 }, { "grad_norm": 0.21236884593963623, "learning_rate": 5.3285968122470295e-05, "loss": 0.0298, "step": 40410 }, { "grad_norm": 0.27906838059425354, "learning_rate": 5.326534415160207e-05, "loss": 0.0436, "step": 40420 }, { "grad_norm": 0.31436771154403687, "learning_rate": 5.324471962277582e-05, "loss": 0.034, "step": 40430 }, { "grad_norm": 0.34910982847213745, "learning_rate": 5.3224094539515746e-05, "loss": 0.0307, "step": 40440 }, { "grad_norm": 0.31494876742362976, "learning_rate": 5.3203468905346076e-05, "loss": 0.0287, "step": 40450 }, { "grad_norm": 0.21733878552913666, "learning_rate": 5.3182842723791195e-05, "loss": 0.0327, "step": 40460 }, { "grad_norm": 0.4139343798160553, "learning_rate": 5.316221599837554e-05, "loss": 0.0269, "step": 40470 }, { "grad_norm": 0.3454950153827667, "learning_rate": 5.314158873262366e-05, "loss": 0.0309, "step": 40480 }, { "grad_norm": 0.3055659234523773, "learning_rate": 5.312096093006018e-05, "loss": 0.0218, "step": 40490 }, { "grad_norm": 0.27375665307044983, "learning_rate": 5.3100332594209846e-05, "loss": 0.0321, "step": 40500 }, { "grad_norm": 0.2621246874332428, "learning_rate": 5.3079703728597454e-05, "loss": 0.0314, "step": 40510 }, { "grad_norm": 0.5094656348228455, "learning_rate": 5.305907433674794e-05, "loss": 0.0303, "step": 40520 }, { "grad_norm": 0.3245670795440674, "learning_rate": 5.3038444422186284e-05, "loss": 0.0275, "step": 40530 }, { "grad_norm": 0.33794862031936646, "learning_rate": 5.301781398843757e-05, "loss": 0.0263, "step": 40540 }, { "grad_norm": 0.30060014128685, "learning_rate": 5.299718303902699e-05, "loss": 0.0326, "step": 40550 }, { "grad_norm": 0.2729952335357666, "learning_rate": 5.297655157747979e-05, "loss": 0.0312, "step": 40560 }, { "grad_norm": 0.30148738622665405, "learning_rate": 5.295591960732136e-05, "loss": 0.0281, "step": 40570 }, { "grad_norm": 0.31718310713768005, "learning_rate": 5.293528713207708e-05, "loss": 0.0333, "step": 40580 }, { "grad_norm": 0.3212392032146454, "learning_rate": 5.291465415527253e-05, "loss": 0.0296, "step": 40590 }, { "grad_norm": 0.4359895586967468, "learning_rate": 5.2894020680433296e-05, "loss": 0.0323, "step": 40600 }, { "grad_norm": 1.0322129726409912, "learning_rate": 5.287338671108507e-05, "loss": 0.028, "step": 40610 }, { "grad_norm": 0.3758457601070404, "learning_rate": 5.285275225075367e-05, "loss": 0.0306, "step": 40620 }, { "grad_norm": 0.3315991759300232, "learning_rate": 5.283211730296492e-05, "loss": 0.0259, "step": 40630 }, { "grad_norm": 0.34110620617866516, "learning_rate": 5.281148187124477e-05, "loss": 0.0314, "step": 40640 }, { "grad_norm": 0.3066267669200897, "learning_rate": 5.279084595911927e-05, "loss": 0.0274, "step": 40650 }, { "grad_norm": 0.27189040184020996, "learning_rate": 5.277020957011453e-05, "loss": 0.0418, "step": 40660 }, { "grad_norm": 0.3147447407245636, "learning_rate": 5.274957270775673e-05, "loss": 0.0401, "step": 40670 }, { "grad_norm": 0.3088659942150116, "learning_rate": 5.2728935375572164e-05, "loss": 0.0251, "step": 40680 }, { "grad_norm": 0.30623504519462585, "learning_rate": 5.2708297577087176e-05, "loss": 0.0299, "step": 40690 }, { "grad_norm": 0.41523826122283936, "learning_rate": 5.2687659315828196e-05, "loss": 0.0327, "step": 40700 }, { "grad_norm": 0.34172630310058594, "learning_rate": 5.266702059532175e-05, "loss": 0.0303, "step": 40710 }, { "grad_norm": 0.4370494782924652, "learning_rate": 5.264638141909444e-05, "loss": 0.0367, "step": 40720 }, { "grad_norm": 0.24113360047340393, "learning_rate": 5.2625741790672925e-05, "loss": 0.0379, "step": 40730 }, { "grad_norm": 0.4423922598361969, "learning_rate": 5.2605101713583935e-05, "loss": 0.0311, "step": 40740 }, { "grad_norm": 0.26356926560401917, "learning_rate": 5.2584461191354315e-05, "loss": 0.0346, "step": 40750 }, { "grad_norm": 0.46690666675567627, "learning_rate": 5.2563820227510964e-05, "loss": 0.0298, "step": 40760 }, { "grad_norm": 0.33731958270072937, "learning_rate": 5.2543178825580864e-05, "loss": 0.0297, "step": 40770 }, { "grad_norm": 0.42221179604530334, "learning_rate": 5.2522536989091054e-05, "loss": 0.0379, "step": 40780 }, { "grad_norm": 0.47102323174476624, "learning_rate": 5.2501894721568655e-05, "loss": 0.0311, "step": 40790 }, { "grad_norm": 0.3693348467350006, "learning_rate": 5.248125202654089e-05, "loss": 0.0295, "step": 40800 }, { "grad_norm": 0.3245201110839844, "learning_rate": 5.246060890753501e-05, "loss": 0.0344, "step": 40810 }, { "grad_norm": 0.31204187870025635, "learning_rate": 5.243996536807837e-05, "loss": 0.0312, "step": 40820 }, { "grad_norm": 0.37519359588623047, "learning_rate": 5.2419321411698384e-05, "loss": 0.0264, "step": 40830 }, { "grad_norm": 0.2687559425830841, "learning_rate": 5.239867704192253e-05, "loss": 0.0352, "step": 40840 }, { "grad_norm": 0.27680453658103943, "learning_rate": 5.237803226227838e-05, "loss": 0.0263, "step": 40850 }, { "grad_norm": 0.4832965135574341, "learning_rate": 5.235738707629354e-05, "loss": 0.0325, "step": 40860 }, { "grad_norm": 0.2818724811077118, "learning_rate": 5.233674148749575e-05, "loss": 0.0389, "step": 40870 }, { "grad_norm": 0.4409089982509613, "learning_rate": 5.231609549941272e-05, "loss": 0.0366, "step": 40880 }, { "grad_norm": 0.22074437141418457, "learning_rate": 5.229544911557233e-05, "loss": 0.0341, "step": 40890 }, { "grad_norm": 0.3191276490688324, "learning_rate": 5.227480233950246e-05, "loss": 0.03, "step": 40900 }, { "grad_norm": 0.3409595191478729, "learning_rate": 5.22541551747311e-05, "loss": 0.0349, "step": 40910 }, { "grad_norm": 0.2711296081542969, "learning_rate": 5.223350762478626e-05, "loss": 0.0211, "step": 40920 }, { "grad_norm": 0.32798174023628235, "learning_rate": 5.221285969319606e-05, "loss": 0.0326, "step": 40930 }, { "grad_norm": 0.22688508033752441, "learning_rate": 5.219221138348865e-05, "loss": 0.0283, "step": 40940 }, { "grad_norm": 0.2485329955816269, "learning_rate": 5.217156269919228e-05, "loss": 0.0314, "step": 40950 }, { "grad_norm": 0.21769753098487854, "learning_rate": 5.215091364383523e-05, "loss": 0.0287, "step": 40960 }, { "grad_norm": 0.39657124876976013, "learning_rate": 5.213026422094588e-05, "loss": 0.0372, "step": 40970 }, { "grad_norm": 0.23704253137111664, "learning_rate": 5.210961443405262e-05, "loss": 0.0342, "step": 40980 }, { "grad_norm": 0.3585151135921478, "learning_rate": 5.208896428668396e-05, "loss": 0.0281, "step": 40990 }, { "grad_norm": 0.46037787199020386, "learning_rate": 5.206831378236845e-05, "loss": 0.0414, "step": 41000 }, { "grad_norm": 0.38390371203422546, "learning_rate": 5.2047662924634666e-05, "loss": 0.0274, "step": 41010 }, { "grad_norm": 0.3274611830711365, "learning_rate": 5.20270117170113e-05, "loss": 0.0346, "step": 41020 }, { "grad_norm": 0.7508341670036316, "learning_rate": 5.200636016302707e-05, "loss": 0.0436, "step": 41030 }, { "grad_norm": 0.3284476101398468, "learning_rate": 5.1985708266210754e-05, "loss": 0.0412, "step": 41040 }, { "grad_norm": 0.3791100084781647, "learning_rate": 5.1965056030091206e-05, "loss": 0.0358, "step": 41050 }, { "grad_norm": 0.3324459493160248, "learning_rate": 5.194440345819732e-05, "loss": 0.0263, "step": 41060 }, { "grad_norm": 0.35557642579078674, "learning_rate": 5.1923750554058084e-05, "loss": 0.036, "step": 41070 }, { "grad_norm": 0.1965755969285965, "learning_rate": 5.1903097321202476e-05, "loss": 0.0265, "step": 41080 }, { "grad_norm": 0.34696146845817566, "learning_rate": 5.18824437631596e-05, "loss": 0.0457, "step": 41090 }, { "grad_norm": 0.2406734973192215, "learning_rate": 5.186178988345856e-05, "loss": 0.0349, "step": 41100 }, { "grad_norm": 0.39577150344848633, "learning_rate": 5.184113568562855e-05, "loss": 0.036, "step": 41110 }, { "grad_norm": 0.3325609564781189, "learning_rate": 5.18204811731988e-05, "loss": 0.0291, "step": 41120 }, { "grad_norm": 0.35515862703323364, "learning_rate": 5.179982634969862e-05, "loss": 0.0375, "step": 41130 }, { "grad_norm": 0.36739158630371094, "learning_rate": 5.177917121865734e-05, "loss": 0.0352, "step": 41140 }, { "grad_norm": 0.28405532240867615, "learning_rate": 5.1758515783604346e-05, "loss": 0.0259, "step": 41150 }, { "grad_norm": 0.46739983558654785, "learning_rate": 5.173786004806911e-05, "loss": 0.0301, "step": 41160 }, { "grad_norm": 0.4010964035987854, "learning_rate": 5.1717204015581135e-05, "loss": 0.0347, "step": 41170 }, { "grad_norm": 0.3443734049797058, "learning_rate": 5.169654768966994e-05, "loss": 0.0318, "step": 41180 }, { "grad_norm": 0.3184385895729065, "learning_rate": 5.1675891073865156e-05, "loss": 0.03, "step": 41190 }, { "grad_norm": 0.19758693873882294, "learning_rate": 5.1655234171696424e-05, "loss": 0.0257, "step": 41200 }, { "grad_norm": 0.3013400435447693, "learning_rate": 5.163457698669343e-05, "loss": 0.0276, "step": 41210 }, { "grad_norm": 0.29260891675949097, "learning_rate": 5.1613919522385936e-05, "loss": 0.0324, "step": 41220 }, { "grad_norm": 0.6009972095489502, "learning_rate": 5.1593261782303746e-05, "loss": 0.0472, "step": 41230 }, { "grad_norm": 0.23069562017917633, "learning_rate": 5.157260376997669e-05, "loss": 0.0294, "step": 41240 }, { "grad_norm": 0.36816462874412537, "learning_rate": 5.155194548893466e-05, "loss": 0.0355, "step": 41250 }, { "grad_norm": 0.3132797181606293, "learning_rate": 5.153128694270759e-05, "loss": 0.0333, "step": 41260 }, { "grad_norm": 0.23600460588932037, "learning_rate": 5.151062813482548e-05, "loss": 0.0389, "step": 41270 }, { "grad_norm": 0.46082237362861633, "learning_rate": 5.148996906881832e-05, "loss": 0.0285, "step": 41280 }, { "grad_norm": 0.3361528515815735, "learning_rate": 5.1469309748216196e-05, "loss": 0.0254, "step": 41290 }, { "grad_norm": 0.36395522952079773, "learning_rate": 5.144865017654923e-05, "loss": 0.0265, "step": 41300 }, { "grad_norm": 0.4212530553340912, "learning_rate": 5.1427990357347556e-05, "loss": 0.0388, "step": 41310 }, { "grad_norm": 0.3766266703605652, "learning_rate": 5.140733029414139e-05, "loss": 0.0257, "step": 41320 }, { "grad_norm": 0.3019654452800751, "learning_rate": 5.138666999046098e-05, "loss": 0.0262, "step": 41330 }, { "grad_norm": 0.3207542598247528, "learning_rate": 5.136600944983658e-05, "loss": 0.0244, "step": 41340 }, { "grad_norm": 0.30540382862091064, "learning_rate": 5.134534867579853e-05, "loss": 0.0344, "step": 41350 }, { "grad_norm": 0.482019305229187, "learning_rate": 5.1324687671877184e-05, "loss": 0.0288, "step": 41360 }, { "grad_norm": 0.41045278310775757, "learning_rate": 5.130402644160296e-05, "loss": 0.0337, "step": 41370 }, { "grad_norm": 0.3446139693260193, "learning_rate": 5.128336498850628e-05, "loss": 0.0266, "step": 41380 }, { "grad_norm": 0.31066054105758667, "learning_rate": 5.126270331611761e-05, "loss": 0.0276, "step": 41390 }, { "grad_norm": 0.38154298067092896, "learning_rate": 5.124204142796748e-05, "loss": 0.0252, "step": 41400 }, { "grad_norm": 0.3189740777015686, "learning_rate": 5.122137932758644e-05, "loss": 0.0381, "step": 41410 }, { "grad_norm": 0.30371665954589844, "learning_rate": 5.1200717018505074e-05, "loss": 0.0285, "step": 41420 }, { "grad_norm": 0.4739474356174469, "learning_rate": 5.118005450425403e-05, "loss": 0.0314, "step": 41430 }, { "grad_norm": 0.48510733246803284, "learning_rate": 5.115939178836392e-05, "loss": 0.0291, "step": 41440 }, { "grad_norm": 0.24669741094112396, "learning_rate": 5.113872887436547e-05, "loss": 0.0318, "step": 41450 }, { "grad_norm": 0.3468456566333771, "learning_rate": 5.11180657657894e-05, "loss": 0.0299, "step": 41460 }, { "grad_norm": 0.2573186755180359, "learning_rate": 5.109740246616649e-05, "loss": 0.0265, "step": 41470 }, { "grad_norm": 0.5589383840560913, "learning_rate": 5.10767389790275e-05, "loss": 0.0339, "step": 41480 }, { "grad_norm": 0.5178821682929993, "learning_rate": 5.1056075307903265e-05, "loss": 0.0341, "step": 41490 }, { "grad_norm": 0.42176562547683716, "learning_rate": 5.1035411456324634e-05, "loss": 0.0295, "step": 41500 }, { "grad_norm": 0.4803510904312134, "learning_rate": 5.1014747427822504e-05, "loss": 0.0347, "step": 41510 }, { "grad_norm": 0.44469165802001953, "learning_rate": 5.0994083225927804e-05, "loss": 0.0317, "step": 41520 }, { "grad_norm": 0.2936970889568329, "learning_rate": 5.0973418854171475e-05, "loss": 0.0332, "step": 41530 }, { "grad_norm": 0.2923780083656311, "learning_rate": 5.095275431608447e-05, "loss": 0.0308, "step": 41540 }, { "grad_norm": 0.22835052013397217, "learning_rate": 5.093208961519782e-05, "loss": 0.0294, "step": 41550 }, { "grad_norm": 0.23647643625736237, "learning_rate": 5.091142475504255e-05, "loss": 0.0292, "step": 41560 }, { "grad_norm": 0.4250863194465637, "learning_rate": 5.089075973914971e-05, "loss": 0.0314, "step": 41570 }, { "grad_norm": 0.6061997413635254, "learning_rate": 5.08700945710504e-05, "loss": 0.043, "step": 41580 }, { "grad_norm": 0.34273064136505127, "learning_rate": 5.0849429254275714e-05, "loss": 0.0277, "step": 41590 }, { "grad_norm": 0.932310163974762, "learning_rate": 5.0828763792356804e-05, "loss": 0.0265, "step": 41600 }, { "grad_norm": 0.22607697546482086, "learning_rate": 5.080809818882483e-05, "loss": 0.0343, "step": 41610 }, { "grad_norm": 0.44229966402053833, "learning_rate": 5.078743244721097e-05, "loss": 0.03, "step": 41620 }, { "grad_norm": 0.39367035031318665, "learning_rate": 5.0766766571046455e-05, "loss": 0.0323, "step": 41630 }, { "grad_norm": 0.2757779359817505, "learning_rate": 5.07461005638625e-05, "loss": 0.0232, "step": 41640 }, { "grad_norm": 0.35079047083854675, "learning_rate": 5.072543442919037e-05, "loss": 0.0368, "step": 41650 }, { "grad_norm": 0.25638723373413086, "learning_rate": 5.070476817056132e-05, "loss": 0.0358, "step": 41660 }, { "grad_norm": 0.28537487983703613, "learning_rate": 5.068410179150668e-05, "loss": 0.0237, "step": 41670 }, { "grad_norm": 0.27763795852661133, "learning_rate": 5.066343529555775e-05, "loss": 0.0285, "step": 41680 }, { "grad_norm": 0.32174617052078247, "learning_rate": 5.064276868624587e-05, "loss": 0.033, "step": 41690 }, { "grad_norm": 0.2929624915122986, "learning_rate": 5.0622101967102396e-05, "loss": 0.029, "step": 41700 }, { "grad_norm": 0.4006405472755432, "learning_rate": 5.0601435141658714e-05, "loss": 0.025, "step": 41710 }, { "grad_norm": 0.3837173581123352, "learning_rate": 5.058076821344622e-05, "loss": 0.0343, "step": 41720 }, { "grad_norm": 0.32343849539756775, "learning_rate": 5.056010118599632e-05, "loss": 0.0311, "step": 41730 }, { "grad_norm": 0.26642850041389465, "learning_rate": 5.053943406284044e-05, "loss": 0.0292, "step": 41740 }, { "grad_norm": 0.4481806755065918, "learning_rate": 5.051876684751006e-05, "loss": 0.0368, "step": 41750 }, { "grad_norm": 0.29391738772392273, "learning_rate": 5.0498099543536584e-05, "loss": 0.0211, "step": 41760 }, { "grad_norm": 0.334979772567749, "learning_rate": 5.047743215445152e-05, "loss": 0.0303, "step": 41770 }, { "grad_norm": 0.5234194397926331, "learning_rate": 5.045676468378637e-05, "loss": 0.0307, "step": 41780 }, { "grad_norm": 0.41283005475997925, "learning_rate": 5.0436097135072626e-05, "loss": 0.0364, "step": 41790 }, { "grad_norm": 0.40058743953704834, "learning_rate": 5.041542951184181e-05, "loss": 0.0315, "step": 41800 }, { "grad_norm": 0.20846164226531982, "learning_rate": 5.039476181762545e-05, "loss": 0.0236, "step": 41810 }, { "grad_norm": 0.2209630310535431, "learning_rate": 5.037409405595508e-05, "loss": 0.0225, "step": 41820 }, { "grad_norm": 0.3115362823009491, "learning_rate": 5.035342623036229e-05, "loss": 0.0295, "step": 41830 }, { "grad_norm": 0.3783484101295471, "learning_rate": 5.033275834437862e-05, "loss": 0.0297, "step": 41840 }, { "grad_norm": 0.2921448051929474, "learning_rate": 5.031209040153564e-05, "loss": 0.0229, "step": 41850 }, { "grad_norm": 0.3775777518749237, "learning_rate": 5.0291422405364955e-05, "loss": 0.0457, "step": 41860 }, { "grad_norm": 0.27881497144699097, "learning_rate": 5.0270754359398133e-05, "loss": 0.0288, "step": 41870 }, { "grad_norm": 0.29564785957336426, "learning_rate": 5.025008626716682e-05, "loss": 0.029, "step": 41880 }, { "grad_norm": 0.29137545824050903, "learning_rate": 5.0229418132202585e-05, "loss": 0.0311, "step": 41890 }, { "grad_norm": 0.4581378102302551, "learning_rate": 5.020874995803707e-05, "loss": 0.0373, "step": 41900 }, { "grad_norm": 0.3101798892021179, "learning_rate": 5.01880817482019e-05, "loss": 0.0308, "step": 41910 }, { "grad_norm": 0.2140248715877533, "learning_rate": 5.01674135062287e-05, "loss": 0.0294, "step": 41920 }, { "grad_norm": 0.308072030544281, "learning_rate": 5.0146745235649115e-05, "loss": 0.035, "step": 41930 }, { "grad_norm": 0.32817602157592773, "learning_rate": 5.0126076939994795e-05, "loss": 0.0307, "step": 41940 }, { "grad_norm": 0.33126991987228394, "learning_rate": 5.010540862279736e-05, "loss": 0.031, "step": 41950 }, { "grad_norm": 0.25165385007858276, "learning_rate": 5.008474028758846e-05, "loss": 0.0281, "step": 41960 }, { "grad_norm": 0.3116292655467987, "learning_rate": 5.0064071937899784e-05, "loss": 0.0333, "step": 41970 }, { "grad_norm": 0.41106662154197693, "learning_rate": 5.004340357726296e-05, "loss": 0.0257, "step": 41980 }, { "grad_norm": 0.5440400242805481, "learning_rate": 5.002273520920965e-05, "loss": 0.0353, "step": 41990 }, { "grad_norm": 0.3332263231277466, "learning_rate": 5.000206683727151e-05, "loss": 0.0293, "step": 42000 }, { "grad_norm": 0.287736713886261, "learning_rate": 4.998139846498021e-05, "loss": 0.0272, "step": 42010 }, { "grad_norm": 0.341852068901062, "learning_rate": 4.9960730095867405e-05, "loss": 0.0322, "step": 42020 }, { "grad_norm": 0.2643021047115326, "learning_rate": 4.9940061733464755e-05, "loss": 0.0407, "step": 42030 }, { "grad_norm": 0.41398438811302185, "learning_rate": 4.991939338130392e-05, "loss": 0.035, "step": 42040 }, { "grad_norm": 0.5368446111679077, "learning_rate": 4.989872504291653e-05, "loss": 0.0428, "step": 42050 }, { "grad_norm": 0.28105610609054565, "learning_rate": 4.9878056721834273e-05, "loss": 0.0293, "step": 42060 }, { "grad_norm": 0.2839337885379791, "learning_rate": 4.98573884215888e-05, "loss": 0.0289, "step": 42070 }, { "grad_norm": 0.39809322357177734, "learning_rate": 4.9836720145711715e-05, "loss": 0.0327, "step": 42080 }, { "grad_norm": 0.29926809668540955, "learning_rate": 4.9816051897734725e-05, "loss": 0.0364, "step": 42090 }, { "grad_norm": 0.35407325625419617, "learning_rate": 4.979538368118941e-05, "loss": 0.0319, "step": 42100 }, { "grad_norm": 0.19998611509799957, "learning_rate": 4.9774715499607446e-05, "loss": 0.0283, "step": 42110 }, { "grad_norm": 0.3796684145927429, "learning_rate": 4.9754047356520416e-05, "loss": 0.0328, "step": 42120 }, { "grad_norm": 0.37497538328170776, "learning_rate": 4.973337925545999e-05, "loss": 0.0322, "step": 42130 }, { "grad_norm": 0.22844168543815613, "learning_rate": 4.9712711199957746e-05, "loss": 0.0271, "step": 42140 }, { "grad_norm": 0.3930107355117798, "learning_rate": 4.969204319354531e-05, "loss": 0.0374, "step": 42150 }, { "grad_norm": 0.3894137740135193, "learning_rate": 4.9671375239754267e-05, "loss": 0.0311, "step": 42160 }, { "grad_norm": 0.5502642393112183, "learning_rate": 4.9650707342116195e-05, "loss": 0.0331, "step": 42170 }, { "grad_norm": 0.3802609443664551, "learning_rate": 4.96300395041627e-05, "loss": 0.0258, "step": 42180 }, { "grad_norm": 0.3258596658706665, "learning_rate": 4.960937172942532e-05, "loss": 0.0257, "step": 42190 }, { "grad_norm": 0.27992257475852966, "learning_rate": 4.958870402143565e-05, "loss": 0.0228, "step": 42200 }, { "grad_norm": 0.2962858974933624, "learning_rate": 4.9568036383725186e-05, "loss": 0.0346, "step": 42210 }, { "grad_norm": 0.3901124894618988, "learning_rate": 4.9547368819825516e-05, "loss": 0.0285, "step": 42220 }, { "grad_norm": 0.4133591949939728, "learning_rate": 4.952670133326812e-05, "loss": 0.032, "step": 42230 }, { "grad_norm": 0.4138253927230835, "learning_rate": 4.950603392758453e-05, "loss": 0.0283, "step": 42240 }, { "grad_norm": 0.35022231936454773, "learning_rate": 4.948536660630621e-05, "loss": 0.0252, "step": 42250 }, { "grad_norm": 0.34825435280799866, "learning_rate": 4.9464699372964685e-05, "loss": 0.0347, "step": 42260 }, { "grad_norm": 0.36615073680877686, "learning_rate": 4.9444032231091395e-05, "loss": 0.0313, "step": 42270 }, { "grad_norm": 0.3489153981208801, "learning_rate": 4.9423365184217776e-05, "loss": 0.0345, "step": 42280 }, { "grad_norm": 0.30872857570648193, "learning_rate": 4.940269823587529e-05, "loss": 0.0255, "step": 42290 }, { "grad_norm": 0.2963182032108307, "learning_rate": 4.938203138959533e-05, "loss": 0.0265, "step": 42300 }, { "grad_norm": 0.39348796010017395, "learning_rate": 4.9361364648909325e-05, "loss": 0.0299, "step": 42310 }, { "grad_norm": 0.3340744376182556, "learning_rate": 4.9340698017348605e-05, "loss": 0.0328, "step": 42320 }, { "grad_norm": 0.4586162269115448, "learning_rate": 4.932003149844458e-05, "loss": 0.0355, "step": 42330 }, { "grad_norm": 0.3841073513031006, "learning_rate": 4.929936509572857e-05, "loss": 0.0442, "step": 42340 }, { "grad_norm": 0.3153359591960907, "learning_rate": 4.927869881273191e-05, "loss": 0.0323, "step": 42350 }, { "grad_norm": 0.5687932372093201, "learning_rate": 4.9258032652985894e-05, "loss": 0.0305, "step": 42360 }, { "grad_norm": 0.2947957515716553, "learning_rate": 4.9237366620021786e-05, "loss": 0.0345, "step": 42370 }, { "grad_norm": 0.2850123941898346, "learning_rate": 4.921670071737089e-05, "loss": 0.0291, "step": 42380 }, { "grad_norm": 0.358061820268631, "learning_rate": 4.91960349485644e-05, "loss": 0.0349, "step": 42390 }, { "grad_norm": 0.2835993766784668, "learning_rate": 4.9175369317133566e-05, "loss": 0.0297, "step": 42400 }, { "grad_norm": 0.31788793206214905, "learning_rate": 4.915470382660953e-05, "loss": 0.0326, "step": 42410 }, { "grad_norm": 0.3474521338939667, "learning_rate": 4.9134038480523524e-05, "loss": 0.0284, "step": 42420 }, { "grad_norm": 0.48974600434303284, "learning_rate": 4.911337328240664e-05, "loss": 0.0393, "step": 42430 }, { "grad_norm": 0.4073431193828583, "learning_rate": 4.909270823579003e-05, "loss": 0.0334, "step": 42440 }, { "grad_norm": 0.3813727796077728, "learning_rate": 4.907204334420476e-05, "loss": 0.0328, "step": 42450 }, { "grad_norm": 0.31856316328048706, "learning_rate": 4.9051378611181884e-05, "loss": 0.0318, "step": 42460 }, { "grad_norm": 0.33298155665397644, "learning_rate": 4.9030714040252486e-05, "loss": 0.0297, "step": 42470 }, { "grad_norm": 0.29421719908714294, "learning_rate": 4.901004963494752e-05, "loss": 0.024, "step": 42480 }, { "grad_norm": 0.4078287184238434, "learning_rate": 4.898938539879802e-05, "loss": 0.0381, "step": 42490 }, { "grad_norm": 0.3390599489212036, "learning_rate": 4.8968721335334904e-05, "loss": 0.0298, "step": 42500 }, { "grad_norm": 0.2166406661272049, "learning_rate": 4.894805744808912e-05, "loss": 0.0362, "step": 42510 }, { "grad_norm": 0.23960983753204346, "learning_rate": 4.8927393740591524e-05, "loss": 0.0297, "step": 42520 }, { "grad_norm": 0.3302842378616333, "learning_rate": 4.890673021637302e-05, "loss": 0.0331, "step": 42530 }, { "grad_norm": 0.3393148183822632, "learning_rate": 4.888606687896442e-05, "loss": 0.0312, "step": 42540 }, { "grad_norm": 0.28771576285362244, "learning_rate": 4.886540373189652e-05, "loss": 0.0317, "step": 42550 }, { "grad_norm": 0.35722047090530396, "learning_rate": 4.88447407787001e-05, "loss": 0.0287, "step": 42560 }, { "grad_norm": 0.3798920214176178, "learning_rate": 4.8824078022905864e-05, "loss": 0.0304, "step": 42570 }, { "grad_norm": 0.2770136594772339, "learning_rate": 4.880341546804455e-05, "loss": 0.0342, "step": 42580 }, { "grad_norm": 0.31200847029685974, "learning_rate": 4.878275311764679e-05, "loss": 0.0272, "step": 42590 }, { "grad_norm": 0.3373633921146393, "learning_rate": 4.8762090975243254e-05, "loss": 0.0292, "step": 42600 }, { "grad_norm": 0.3275454640388489, "learning_rate": 4.874142904436448e-05, "loss": 0.0269, "step": 42610 }, { "grad_norm": 0.1743728667497635, "learning_rate": 4.8720767328541094e-05, "loss": 0.0416, "step": 42620 }, { "grad_norm": 0.4316449761390686, "learning_rate": 4.870010583130358e-05, "loss": 0.0269, "step": 42630 }, { "grad_norm": 0.3334869146347046, "learning_rate": 4.867944455618241e-05, "loss": 0.0259, "step": 42640 }, { "grad_norm": 0.3002561330795288, "learning_rate": 4.865878350670807e-05, "loss": 0.0325, "step": 42650 }, { "grad_norm": 0.43142935633659363, "learning_rate": 4.8638122686410914e-05, "loss": 0.0323, "step": 42660 }, { "grad_norm": 0.2513692378997803, "learning_rate": 4.861746209882137e-05, "loss": 0.0268, "step": 42670 }, { "grad_norm": 0.2760622799396515, "learning_rate": 4.859680174746972e-05, "loss": 0.0367, "step": 42680 }, { "grad_norm": 0.3496687114238739, "learning_rate": 4.857614163588629e-05, "loss": 0.0417, "step": 42690 }, { "grad_norm": 0.33157652616500854, "learning_rate": 4.855548176760131e-05, "loss": 0.0302, "step": 42700 }, { "grad_norm": 0.2626250386238098, "learning_rate": 4.8534822146144986e-05, "loss": 0.0279, "step": 42710 }, { "grad_norm": 0.34687644243240356, "learning_rate": 4.851416277504749e-05, "loss": 0.0381, "step": 42720 }, { "grad_norm": 0.36534079909324646, "learning_rate": 4.8493503657838923e-05, "loss": 0.0237, "step": 42730 }, { "grad_norm": 0.37919044494628906, "learning_rate": 4.84728447980494e-05, "loss": 0.0284, "step": 42740 }, { "grad_norm": 0.23677659034729004, "learning_rate": 4.8452186199208914e-05, "loss": 0.0287, "step": 42750 }, { "grad_norm": 0.33847707509994507, "learning_rate": 4.843152786484749e-05, "loss": 0.0292, "step": 42760 }, { "grad_norm": 0.269763708114624, "learning_rate": 4.8410869798495035e-05, "loss": 0.027, "step": 42770 }, { "grad_norm": 0.22173747420310974, "learning_rate": 4.8390212003681486e-05, "loss": 0.0276, "step": 42780 }, { "grad_norm": 0.36171284317970276, "learning_rate": 4.836955448393667e-05, "loss": 0.0341, "step": 42790 }, { "grad_norm": 0.2540743947029114, "learning_rate": 4.8348897242790406e-05, "loss": 0.0279, "step": 42800 }, { "grad_norm": 0.195899099111557, "learning_rate": 4.832824028377243e-05, "loss": 0.0262, "step": 42810 }, { "grad_norm": 0.35311126708984375, "learning_rate": 4.830758361041249e-05, "loss": 0.0283, "step": 42820 }, { "grad_norm": 0.38184717297554016, "learning_rate": 4.828692722624022e-05, "loss": 0.0306, "step": 42830 }, { "grad_norm": 0.20580202341079712, "learning_rate": 4.826627113478522e-05, "loss": 0.0442, "step": 42840 }, { "grad_norm": 0.3615662753582001, "learning_rate": 4.824561533957708e-05, "loss": 0.0241, "step": 42850 }, { "grad_norm": 0.35421642661094666, "learning_rate": 4.822495984414527e-05, "loss": 0.0293, "step": 42860 }, { "grad_norm": 0.39073699712753296, "learning_rate": 4.8204304652019304e-05, "loss": 0.0296, "step": 42870 }, { "grad_norm": 0.26333653926849365, "learning_rate": 4.8183649766728537e-05, "loss": 0.0269, "step": 42880 }, { "grad_norm": 0.5940724015235901, "learning_rate": 4.816299519180237e-05, "loss": 0.0322, "step": 42890 }, { "grad_norm": 0.32504570484161377, "learning_rate": 4.8142340930770066e-05, "loss": 0.0295, "step": 42900 }, { "grad_norm": 0.19449026882648468, "learning_rate": 4.81216869871609e-05, "loss": 0.0257, "step": 42910 }, { "grad_norm": 0.2310492843389511, "learning_rate": 4.8101033364504064e-05, "loss": 0.0292, "step": 42920 }, { "grad_norm": 0.26650211215019226, "learning_rate": 4.808038006632866e-05, "loss": 0.0318, "step": 42930 }, { "grad_norm": 0.40684670209884644, "learning_rate": 4.805972709616383e-05, "loss": 0.0333, "step": 42940 }, { "grad_norm": 0.22864219546318054, "learning_rate": 4.8039074457538556e-05, "loss": 0.0243, "step": 42950 }, { "grad_norm": 0.23347008228302002, "learning_rate": 4.801842215398184e-05, "loss": 0.0337, "step": 42960 }, { "grad_norm": 0.3403799533843994, "learning_rate": 4.799777018902256e-05, "loss": 0.0322, "step": 42970 }, { "grad_norm": 0.5089182257652283, "learning_rate": 4.797711856618961e-05, "loss": 0.0305, "step": 42980 }, { "grad_norm": 0.3766322135925293, "learning_rate": 4.7956467289011765e-05, "loss": 0.0294, "step": 42990 }, { "grad_norm": 0.23764847218990326, "learning_rate": 4.793581636101778e-05, "loss": 0.0285, "step": 43000 }, { "grad_norm": 0.2690874934196472, "learning_rate": 4.7915165785736326e-05, "loss": 0.0268, "step": 43010 }, { "grad_norm": 0.4103753864765167, "learning_rate": 4.789451556669599e-05, "loss": 0.028, "step": 43020 }, { "grad_norm": 0.25576627254486084, "learning_rate": 4.7873865707425384e-05, "loss": 0.0372, "step": 43030 }, { "grad_norm": 0.3434242308139801, "learning_rate": 4.7853216211452975e-05, "loss": 0.0348, "step": 43040 }, { "grad_norm": 0.3843148946762085, "learning_rate": 4.78325670823072e-05, "loss": 0.0311, "step": 43050 }, { "grad_norm": 0.1921498328447342, "learning_rate": 4.781191832351641e-05, "loss": 0.0296, "step": 43060 }, { "grad_norm": 0.3163248300552368, "learning_rate": 4.7791269938608965e-05, "loss": 0.033, "step": 43070 }, { "grad_norm": 0.22530721127986908, "learning_rate": 4.777062193111305e-05, "loss": 0.028, "step": 43080 }, { "grad_norm": 0.2848086655139923, "learning_rate": 4.77499743045569e-05, "loss": 0.0322, "step": 43090 }, { "grad_norm": 0.2873125374317169, "learning_rate": 4.7729327062468596e-05, "loss": 0.0296, "step": 43100 }, { "grad_norm": 0.34738853573799133, "learning_rate": 4.770868020837619e-05, "loss": 0.0355, "step": 43110 }, { "grad_norm": 0.28047099709510803, "learning_rate": 4.768803374580768e-05, "loss": 0.027, "step": 43120 }, { "grad_norm": 0.3599770665168762, "learning_rate": 4.766738767829094e-05, "loss": 0.0308, "step": 43130 }, { "grad_norm": 0.26989641785621643, "learning_rate": 4.764674200935388e-05, "loss": 0.0325, "step": 43140 }, { "grad_norm": 0.377315491437912, "learning_rate": 4.762609674252424e-05, "loss": 0.026, "step": 43150 }, { "grad_norm": 0.23346658051013947, "learning_rate": 4.760545188132974e-05, "loss": 0.0408, "step": 43160 }, { "grad_norm": 0.31100672483444214, "learning_rate": 4.758480742929801e-05, "loss": 0.0369, "step": 43170 }, { "grad_norm": 0.2900516986846924, "learning_rate": 4.756416338995664e-05, "loss": 0.0335, "step": 43180 }, { "grad_norm": 0.22703580558300018, "learning_rate": 4.7543519766833145e-05, "loss": 0.0205, "step": 43190 }, { "grad_norm": 0.24127478897571564, "learning_rate": 4.752287656345492e-05, "loss": 0.0342, "step": 43200 }, { "grad_norm": 0.21517634391784668, "learning_rate": 4.7502233783349355e-05, "loss": 0.035, "step": 43210 }, { "grad_norm": 0.3223497271537781, "learning_rate": 4.7481591430043694e-05, "loss": 0.0272, "step": 43220 }, { "grad_norm": 0.3152965307235718, "learning_rate": 4.746094950706521e-05, "loss": 0.0311, "step": 43230 }, { "grad_norm": 0.3217834532260895, "learning_rate": 4.744030801794099e-05, "loss": 0.027, "step": 43240 }, { "grad_norm": 0.2258588671684265, "learning_rate": 4.741966696619813e-05, "loss": 0.0339, "step": 43250 }, { "grad_norm": 0.34543129801750183, "learning_rate": 4.739902635536361e-05, "loss": 0.0328, "step": 43260 }, { "grad_norm": 0.29038164019584656, "learning_rate": 4.737838618896436e-05, "loss": 0.0287, "step": 43270 }, { "grad_norm": 0.32539102435112, "learning_rate": 4.7357746470527203e-05, "loss": 0.0341, "step": 43280 }, { "grad_norm": 0.3049773573875427, "learning_rate": 4.73371072035789e-05, "loss": 0.0343, "step": 43290 }, { "grad_norm": 0.33420732617378235, "learning_rate": 4.731646839164616e-05, "loss": 0.0292, "step": 43300 }, { "grad_norm": 0.3767375349998474, "learning_rate": 4.729583003825556e-05, "loss": 0.0276, "step": 43310 }, { "grad_norm": 0.3858088552951813, "learning_rate": 4.727519214693367e-05, "loss": 0.0254, "step": 43320 }, { "grad_norm": 0.42304614186286926, "learning_rate": 4.725455472120689e-05, "loss": 0.0272, "step": 43330 }, { "grad_norm": 0.2060118466615677, "learning_rate": 4.723391776460164e-05, "loss": 0.0278, "step": 43340 }, { "grad_norm": 0.37200692296028137, "learning_rate": 4.7213281280644186e-05, "loss": 0.0301, "step": 43350 }, { "grad_norm": 0.4230763018131256, "learning_rate": 4.719264527286075e-05, "loss": 0.0303, "step": 43360 }, { "grad_norm": 0.3709808588027954, "learning_rate": 4.717200974477744e-05, "loss": 0.0299, "step": 43370 }, { "grad_norm": 0.43785688281059265, "learning_rate": 4.715137469992034e-05, "loss": 0.0348, "step": 43380 }, { "grad_norm": 0.3650277853012085, "learning_rate": 4.713074014181539e-05, "loss": 0.0292, "step": 43390 }, { "grad_norm": 0.35230085253715515, "learning_rate": 4.711010607398846e-05, "loss": 0.0241, "step": 43400 }, { "grad_norm": 0.2800483703613281, "learning_rate": 4.708947249996538e-05, "loss": 0.027, "step": 43410 }, { "grad_norm": 0.2581128180027008, "learning_rate": 4.706883942327183e-05, "loss": 0.0291, "step": 43420 }, { "grad_norm": 0.3050054609775543, "learning_rate": 4.704820684743347e-05, "loss": 0.0258, "step": 43430 }, { "grad_norm": 0.3319941759109497, "learning_rate": 4.702757477597581e-05, "loss": 0.0311, "step": 43440 }, { "grad_norm": 0.3002064824104309, "learning_rate": 4.700694321242435e-05, "loss": 0.0306, "step": 43450 }, { "grad_norm": 0.339303582906723, "learning_rate": 4.698631216030441e-05, "loss": 0.0295, "step": 43460 }, { "grad_norm": 0.3141556978225708, "learning_rate": 4.6965681623141314e-05, "loss": 0.03, "step": 43470 }, { "grad_norm": 0.24786745011806488, "learning_rate": 4.694505160446024e-05, "loss": 0.0247, "step": 43480 }, { "grad_norm": 0.2021576464176178, "learning_rate": 4.692442210778628e-05, "loss": 0.0233, "step": 43490 }, { "grad_norm": 0.4380667209625244, "learning_rate": 4.690379313664448e-05, "loss": 0.0349, "step": 43500 }, { "grad_norm": 0.47237518429756165, "learning_rate": 4.688316469455973e-05, "loss": 0.0259, "step": 43510 }, { "grad_norm": 0.2976691722869873, "learning_rate": 4.6862536785056914e-05, "loss": 0.026, "step": 43520 }, { "grad_norm": 0.7835997343063354, "learning_rate": 4.684190941166073e-05, "loss": 0.0318, "step": 43530 }, { "grad_norm": 0.3139994740486145, "learning_rate": 4.682128257789587e-05, "loss": 0.0222, "step": 43540 }, { "grad_norm": 0.4129171371459961, "learning_rate": 4.680065628728687e-05, "loss": 0.0291, "step": 43550 }, { "grad_norm": 0.5334774255752563, "learning_rate": 4.678003054335822e-05, "loss": 0.0306, "step": 43560 }, { "grad_norm": 0.36046117544174194, "learning_rate": 4.675940534963428e-05, "loss": 0.0354, "step": 43570 }, { "grad_norm": 0.2907547950744629, "learning_rate": 4.673878070963931e-05, "loss": 0.0289, "step": 43580 }, { "grad_norm": 0.3560117483139038, "learning_rate": 4.671815662689756e-05, "loss": 0.0257, "step": 43590 }, { "grad_norm": 0.2662395238876343, "learning_rate": 4.669753310493306e-05, "loss": 0.0291, "step": 43600 }, { "grad_norm": 0.27633941173553467, "learning_rate": 4.6676910147269845e-05, "loss": 0.0335, "step": 43610 }, { "grad_norm": 0.2979801297187805, "learning_rate": 4.665628775743177e-05, "loss": 0.0259, "step": 43620 }, { "grad_norm": 0.3498460650444031, "learning_rate": 4.66356659389427e-05, "loss": 0.0299, "step": 43630 }, { "grad_norm": 0.28138965368270874, "learning_rate": 4.661504469532629e-05, "loss": 0.0283, "step": 43640 }, { "grad_norm": 0.31603097915649414, "learning_rate": 4.659442403010617e-05, "loss": 0.0293, "step": 43650 }, { "grad_norm": 0.29437193274497986, "learning_rate": 4.6573803946805845e-05, "loss": 0.0244, "step": 43660 }, { "grad_norm": 0.46382936835289, "learning_rate": 4.6553184448948695e-05, "loss": 0.0348, "step": 43670 }, { "grad_norm": 0.32693567872047424, "learning_rate": 4.653256554005807e-05, "loss": 0.0237, "step": 43680 }, { "grad_norm": 0.18559247255325317, "learning_rate": 4.6511947223657145e-05, "loss": 0.0356, "step": 43690 }, { "grad_norm": 0.3254639804363251, "learning_rate": 4.649132950326906e-05, "loss": 0.0313, "step": 43700 }, { "grad_norm": 0.29967617988586426, "learning_rate": 4.647071238241679e-05, "loss": 0.0285, "step": 43710 }, { "grad_norm": 0.2836104929447174, "learning_rate": 4.6450095864623264e-05, "loss": 0.031, "step": 43720 }, { "grad_norm": 0.25272729992866516, "learning_rate": 4.642947995341123e-05, "loss": 0.0179, "step": 43730 }, { "grad_norm": 0.2952108085155487, "learning_rate": 4.640886465230345e-05, "loss": 0.0268, "step": 43740 }, { "grad_norm": 0.4691588282585144, "learning_rate": 4.6388249964822485e-05, "loss": 0.0326, "step": 43750 }, { "grad_norm": 0.33944085240364075, "learning_rate": 4.63676358944908e-05, "loss": 0.0321, "step": 43760 }, { "grad_norm": 0.567405641078949, "learning_rate": 4.634702244483081e-05, "loss": 0.0394, "step": 43770 }, { "grad_norm": 0.45442456007003784, "learning_rate": 4.632640961936475e-05, "loss": 0.039, "step": 43780 }, { "grad_norm": 0.35308438539505005, "learning_rate": 4.6305797421614835e-05, "loss": 0.0238, "step": 43790 }, { "grad_norm": 0.3705441653728485, "learning_rate": 4.6285185855103105e-05, "loss": 0.0313, "step": 43800 }, { "grad_norm": 0.2918452024459839, "learning_rate": 4.626457492335151e-05, "loss": 0.0286, "step": 43810 }, { "grad_norm": 0.2915481626987457, "learning_rate": 4.624396462988188e-05, "loss": 0.0239, "step": 43820 }, { "grad_norm": 0.3032636046409607, "learning_rate": 4.6223354978215985e-05, "loss": 0.0217, "step": 43830 }, { "grad_norm": 0.195792555809021, "learning_rate": 4.620274597187544e-05, "loss": 0.0254, "step": 43840 }, { "grad_norm": 0.29320186376571655, "learning_rate": 4.6182137614381736e-05, "loss": 0.0295, "step": 43850 }, { "grad_norm": 0.40679481625556946, "learning_rate": 4.616152990925631e-05, "loss": 0.0261, "step": 43860 }, { "grad_norm": 0.3247382342815399, "learning_rate": 4.614092286002041e-05, "loss": 0.0289, "step": 43870 }, { "grad_norm": 0.3491905927658081, "learning_rate": 4.612031647019528e-05, "loss": 0.0265, "step": 43880 }, { "grad_norm": 0.2670280337333679, "learning_rate": 4.609971074330193e-05, "loss": 0.0328, "step": 43890 }, { "grad_norm": 0.3622512221336365, "learning_rate": 4.607910568286136e-05, "loss": 0.0366, "step": 43900 }, { "grad_norm": 0.36064812541007996, "learning_rate": 4.605850129239438e-05, "loss": 0.0282, "step": 43910 }, { "grad_norm": 0.28650951385498047, "learning_rate": 4.603789757542174e-05, "loss": 0.0265, "step": 43920 }, { "grad_norm": 0.35220208764076233, "learning_rate": 4.601729453546404e-05, "loss": 0.034, "step": 43930 }, { "grad_norm": 0.3391779661178589, "learning_rate": 4.599669217604177e-05, "loss": 0.0269, "step": 43940 }, { "grad_norm": 0.43798932433128357, "learning_rate": 4.597609050067532e-05, "loss": 0.0325, "step": 43950 }, { "grad_norm": 0.4309149384498596, "learning_rate": 4.595548951288495e-05, "loss": 0.0372, "step": 43960 }, { "grad_norm": 0.37276026606559753, "learning_rate": 4.593488921619081e-05, "loss": 0.0288, "step": 43970 }, { "grad_norm": 0.36712995171546936, "learning_rate": 4.591428961411289e-05, "loss": 0.032, "step": 43980 }, { "grad_norm": 0.3607511818408966, "learning_rate": 4.589369071017117e-05, "loss": 0.0377, "step": 43990 }, { "grad_norm": 0.3790132701396942, "learning_rate": 4.587309250788538e-05, "loss": 0.0331, "step": 44000 }, { "grad_norm": 0.3396655321121216, "learning_rate": 4.585249501077522e-05, "loss": 0.0297, "step": 44010 }, { "grad_norm": 0.4158136248588562, "learning_rate": 4.583189822236021e-05, "loss": 0.0314, "step": 44020 }, { "grad_norm": 0.34314101934432983, "learning_rate": 4.5811302146159816e-05, "loss": 0.0295, "step": 44030 }, { "grad_norm": 0.2965499758720398, "learning_rate": 4.579070678569332e-05, "loss": 0.0271, "step": 44040 }, { "grad_norm": 0.4257897138595581, "learning_rate": 4.5770112144479904e-05, "loss": 0.0355, "step": 44050 }, { "grad_norm": 0.2930309474468231, "learning_rate": 4.5749518226038645e-05, "loss": 0.023, "step": 44060 }, { "grad_norm": 0.3936893343925476, "learning_rate": 4.572892503388845e-05, "loss": 0.0322, "step": 44070 }, { "grad_norm": 0.35365021228790283, "learning_rate": 4.570833257154817e-05, "loss": 0.0296, "step": 44080 }, { "grad_norm": 0.22552664577960968, "learning_rate": 4.568774084253646e-05, "loss": 0.0296, "step": 44090 }, { "grad_norm": 0.3340725302696228, "learning_rate": 4.566714985037191e-05, "loss": 0.0343, "step": 44100 }, { "grad_norm": 0.39502331614494324, "learning_rate": 4.564655959857295e-05, "loss": 0.0285, "step": 44110 }, { "grad_norm": 0.35847851634025574, "learning_rate": 4.5625970090657893e-05, "loss": 0.0281, "step": 44120 }, { "grad_norm": 0.3133408725261688, "learning_rate": 4.560538133014491e-05, "loss": 0.0257, "step": 44130 }, { "grad_norm": 0.4257599711418152, "learning_rate": 4.5584793320552055e-05, "loss": 0.0302, "step": 44140 }, { "grad_norm": 0.26875609159469604, "learning_rate": 4.556420606539728e-05, "loss": 0.0285, "step": 44150 }, { "grad_norm": 0.39001110196113586, "learning_rate": 4.554361956819836e-05, "loss": 0.021, "step": 44160 }, { "grad_norm": 0.19313615560531616, "learning_rate": 4.552303383247299e-05, "loss": 0.0224, "step": 44170 }, { "grad_norm": 0.25288838148117065, "learning_rate": 4.5502448861738664e-05, "loss": 0.0229, "step": 44180 }, { "grad_norm": 0.29825401306152344, "learning_rate": 4.548186465951284e-05, "loss": 0.0395, "step": 44190 }, { "grad_norm": 0.35442718863487244, "learning_rate": 4.546128122931277e-05, "loss": 0.0301, "step": 44200 }, { "grad_norm": 0.2768581211566925, "learning_rate": 4.544069857465561e-05, "loss": 0.0256, "step": 44210 }, { "grad_norm": 0.3336041271686554, "learning_rate": 4.542011669905837e-05, "loss": 0.0276, "step": 44220 }, { "grad_norm": 0.3687874674797058, "learning_rate": 4.53995356060379e-05, "loss": 0.029, "step": 44230 }, { "grad_norm": 0.17095649242401123, "learning_rate": 4.5378955299110994e-05, "loss": 0.0291, "step": 44240 }, { "grad_norm": 0.5149233341217041, "learning_rate": 4.5358375781794233e-05, "loss": 0.0356, "step": 44250 }, { "grad_norm": 0.29647526144981384, "learning_rate": 4.533779705760411e-05, "loss": 0.0366, "step": 44260 }, { "grad_norm": 0.3346736431121826, "learning_rate": 4.5317219130056934e-05, "loss": 0.0326, "step": 44270 }, { "grad_norm": 0.23897488415241241, "learning_rate": 4.5296642002668964e-05, "loss": 0.0321, "step": 44280 }, { "grad_norm": 0.32031309604644775, "learning_rate": 4.5276065678956216e-05, "loss": 0.0299, "step": 44290 }, { "grad_norm": 0.269980251789093, "learning_rate": 4.525549016243466e-05, "loss": 0.0288, "step": 44300 }, { "grad_norm": 0.35078006982803345, "learning_rate": 4.523491545662008e-05, "loss": 0.0234, "step": 44310 }, { "grad_norm": 0.37547236680984497, "learning_rate": 4.5214341565028104e-05, "loss": 0.0389, "step": 44320 }, { "grad_norm": 0.38299280405044556, "learning_rate": 4.519376849117428e-05, "loss": 0.0304, "step": 44330 }, { "grad_norm": 0.22639310359954834, "learning_rate": 4.517319623857395e-05, "loss": 0.0401, "step": 44340 }, { "grad_norm": 0.3491508662700653, "learning_rate": 4.5152624810742394e-05, "loss": 0.0292, "step": 44350 }, { "grad_norm": 0.2663043439388275, "learning_rate": 4.5132054211194664e-05, "loss": 0.0264, "step": 44360 }, { "grad_norm": 0.3757513165473938, "learning_rate": 4.511148444344574e-05, "loss": 0.0261, "step": 44370 }, { "grad_norm": 0.3506980538368225, "learning_rate": 4.509091551101041e-05, "loss": 0.0352, "step": 44380 }, { "grad_norm": 0.3162216246128082, "learning_rate": 4.507034741740338e-05, "loss": 0.0282, "step": 44390 }, { "grad_norm": 0.5814462900161743, "learning_rate": 4.5049780166139145e-05, "loss": 0.0346, "step": 44400 }, { "grad_norm": 0.512417733669281, "learning_rate": 4.5029213760732075e-05, "loss": 0.05, "step": 44410 }, { "grad_norm": 0.313705176115036, "learning_rate": 4.5008648204696434e-05, "loss": 0.0291, "step": 44420 }, { "grad_norm": 0.2667359411716461, "learning_rate": 4.4988083501546284e-05, "loss": 0.0284, "step": 44430 }, { "grad_norm": 0.2544246017932892, "learning_rate": 4.4967519654795606e-05, "loss": 0.0276, "step": 44440 }, { "grad_norm": 0.26840755343437195, "learning_rate": 4.494695666795816e-05, "loss": 0.0325, "step": 44450 }, { "grad_norm": 0.329637736082077, "learning_rate": 4.4926394544547644e-05, "loss": 0.0311, "step": 44460 }, { "grad_norm": 0.3767392933368683, "learning_rate": 4.490583328807752e-05, "loss": 0.034, "step": 44470 }, { "grad_norm": 0.5673972964286804, "learning_rate": 4.488527290206117e-05, "loss": 0.0415, "step": 44480 }, { "grad_norm": 0.3321633040904999, "learning_rate": 4.48647133900118e-05, "loss": 0.0331, "step": 44490 }, { "grad_norm": 0.2710108757019043, "learning_rate": 4.484415475544244e-05, "loss": 0.0318, "step": 44500 }, { "grad_norm": 0.2767534554004669, "learning_rate": 4.4823597001866035e-05, "loss": 0.0297, "step": 44510 }, { "grad_norm": 0.295246422290802, "learning_rate": 4.480304013279532e-05, "loss": 0.0272, "step": 44520 }, { "grad_norm": 0.2945733368396759, "learning_rate": 4.478248415174292e-05, "loss": 0.0235, "step": 44530 }, { "grad_norm": 0.305042028427124, "learning_rate": 4.476192906222126e-05, "loss": 0.0311, "step": 44540 }, { "grad_norm": 0.4362780749797821, "learning_rate": 4.474137486774268e-05, "loss": 0.0378, "step": 44550 }, { "grad_norm": 0.3891456127166748, "learning_rate": 4.4720821571819296e-05, "loss": 0.0298, "step": 44560 }, { "grad_norm": 0.32452842593193054, "learning_rate": 4.470026917796314e-05, "loss": 0.0352, "step": 44570 }, { "grad_norm": 0.3766691982746124, "learning_rate": 4.4679717689686005e-05, "loss": 0.0532, "step": 44580 }, { "grad_norm": 0.3759106397628784, "learning_rate": 4.465916711049963e-05, "loss": 0.0303, "step": 44590 }, { "grad_norm": 0.3580482006072998, "learning_rate": 4.4638617443915524e-05, "loss": 0.024, "step": 44600 }, { "grad_norm": 0.4303860068321228, "learning_rate": 4.4618068693445055e-05, "loss": 0.0225, "step": 44610 }, { "grad_norm": 0.33907341957092285, "learning_rate": 4.459752086259946e-05, "loss": 0.0357, "step": 44620 }, { "grad_norm": 0.5109567046165466, "learning_rate": 4.457697395488977e-05, "loss": 0.0355, "step": 44630 }, { "grad_norm": 0.2894116938114166, "learning_rate": 4.455642797382693e-05, "loss": 0.031, "step": 44640 }, { "grad_norm": 0.3209804892539978, "learning_rate": 4.4535882922921643e-05, "loss": 0.0345, "step": 44650 }, { "grad_norm": 0.2580784857273102, "learning_rate": 4.451533880568455e-05, "loss": 0.0271, "step": 44660 }, { "grad_norm": 0.23439280688762665, "learning_rate": 4.449479562562603e-05, "loss": 0.0316, "step": 44670 }, { "grad_norm": 0.20791757106781006, "learning_rate": 4.447425338625639e-05, "loss": 0.0295, "step": 44680 }, { "grad_norm": 0.3278716802597046, "learning_rate": 4.4453712091085705e-05, "loss": 0.0287, "step": 44690 }, { "grad_norm": 0.3316086232662201, "learning_rate": 4.443317174362392e-05, "loss": 0.04, "step": 44700 }, { "grad_norm": 0.2529977262020111, "learning_rate": 4.441263234738085e-05, "loss": 0.0281, "step": 44710 }, { "grad_norm": 0.2269420623779297, "learning_rate": 4.4392093905866086e-05, "loss": 0.0276, "step": 44720 }, { "grad_norm": 0.32554370164871216, "learning_rate": 4.437155642258911e-05, "loss": 0.0281, "step": 44730 }, { "grad_norm": 0.1931874305009842, "learning_rate": 4.4351019901059177e-05, "loss": 0.0314, "step": 44740 }, { "grad_norm": 0.3539837896823883, "learning_rate": 4.433048434478547e-05, "loss": 0.0335, "step": 44750 }, { "grad_norm": 0.2879921793937683, "learning_rate": 4.4309949757276906e-05, "loss": 0.0286, "step": 44760 }, { "grad_norm": 0.26850926876068115, "learning_rate": 4.428941614204233e-05, "loss": 0.0325, "step": 44770 }, { "grad_norm": 0.23016402125358582, "learning_rate": 4.426888350259034e-05, "loss": 0.0347, "step": 44780 }, { "grad_norm": 0.2837170362472534, "learning_rate": 4.424835184242939e-05, "loss": 0.0342, "step": 44790 }, { "grad_norm": 0.28358352184295654, "learning_rate": 4.422782116506784e-05, "loss": 0.0286, "step": 44800 }, { "grad_norm": 0.45515644550323486, "learning_rate": 4.4207291474013754e-05, "loss": 0.0309, "step": 44810 }, { "grad_norm": 0.46303683519363403, "learning_rate": 4.418676277277514e-05, "loss": 0.0428, "step": 44820 }, { "grad_norm": 0.19289714097976685, "learning_rate": 4.416623506485976e-05, "loss": 0.0329, "step": 44830 }, { "grad_norm": 0.39645564556121826, "learning_rate": 4.414570835377527e-05, "loss": 0.0265, "step": 44840 }, { "grad_norm": 0.363725870847702, "learning_rate": 4.412518264302909e-05, "loss": 0.0261, "step": 44850 }, { "grad_norm": 0.4452974796295166, "learning_rate": 4.4104657936128535e-05, "loss": 0.0327, "step": 44860 }, { "grad_norm": 0.361424058675766, "learning_rate": 4.408413423658071e-05, "loss": 0.0321, "step": 44870 }, { "grad_norm": 0.3182947635650635, "learning_rate": 4.406361154789253e-05, "loss": 0.0293, "step": 44880 }, { "grad_norm": 0.40459927916526794, "learning_rate": 4.4043089873570776e-05, "loss": 0.0266, "step": 44890 }, { "grad_norm": 0.3906290531158447, "learning_rate": 4.4022569217122025e-05, "loss": 0.0318, "step": 44900 }, { "grad_norm": 0.29762792587280273, "learning_rate": 4.400204958205274e-05, "loss": 0.029, "step": 44910 }, { "grad_norm": 0.5626985430717468, "learning_rate": 4.3981530971869125e-05, "loss": 0.0409, "step": 44920 }, { "grad_norm": 0.29038697481155396, "learning_rate": 4.396101339007727e-05, "loss": 0.0254, "step": 44930 }, { "grad_norm": 0.35178855061531067, "learning_rate": 4.394049684018304e-05, "loss": 0.027, "step": 44940 }, { "grad_norm": 0.2751390039920807, "learning_rate": 4.3919981325692186e-05, "loss": 0.0327, "step": 44950 }, { "grad_norm": 0.25586631894111633, "learning_rate": 4.389946685011024e-05, "loss": 0.0233, "step": 44960 }, { "grad_norm": 0.304569810628891, "learning_rate": 4.387895341694255e-05, "loss": 0.0327, "step": 44970 }, { "grad_norm": 0.280373215675354, "learning_rate": 4.3858441029694324e-05, "loss": 0.0357, "step": 44980 }, { "grad_norm": 0.3466290533542633, "learning_rate": 4.3837929691870527e-05, "loss": 0.0211, "step": 44990 }, { "grad_norm": 0.33275365829467773, "learning_rate": 4.381741940697604e-05, "loss": 0.0309, "step": 45000 }, { "grad_norm": 0.2310163825750351, "learning_rate": 4.379691017851547e-05, "loss": 0.0228, "step": 45010 }, { "grad_norm": 0.2403174340724945, "learning_rate": 4.3776402009993304e-05, "loss": 0.0298, "step": 45020 }, { "grad_norm": 0.27077874541282654, "learning_rate": 4.3755894904913794e-05, "loss": 0.0268, "step": 45030 }, { "grad_norm": 0.31501394510269165, "learning_rate": 4.373538886678109e-05, "loss": 0.0323, "step": 45040 }, { "grad_norm": 0.35718467831611633, "learning_rate": 4.371488389909909e-05, "loss": 0.0278, "step": 45050 }, { "grad_norm": 0.2752644121646881, "learning_rate": 4.3694380005371515e-05, "loss": 0.025, "step": 45060 }, { "grad_norm": 0.2648972272872925, "learning_rate": 4.367387718910196e-05, "loss": 0.0278, "step": 45070 }, { "grad_norm": 0.45281660556793213, "learning_rate": 4.3653375453793764e-05, "loss": 0.0292, "step": 45080 }, { "grad_norm": 0.2888779938220978, "learning_rate": 4.3632874802950136e-05, "loss": 0.0256, "step": 45090 }, { "grad_norm": 0.3279014825820923, "learning_rate": 4.3612375240074034e-05, "loss": 0.03, "step": 45100 }, { "grad_norm": 0.5267463326454163, "learning_rate": 4.3591876768668325e-05, "loss": 0.0323, "step": 45110 }, { "grad_norm": 0.35782867670059204, "learning_rate": 4.3571379392235605e-05, "loss": 0.0276, "step": 45120 }, { "grad_norm": 0.4062255024909973, "learning_rate": 4.3550883114278335e-05, "loss": 0.0306, "step": 45130 }, { "grad_norm": 0.2840602695941925, "learning_rate": 4.353038793829876e-05, "loss": 0.0259, "step": 45140 }, { "grad_norm": 0.22338806092739105, "learning_rate": 4.350989386779891e-05, "loss": 0.0269, "step": 45150 }, { "grad_norm": 0.32176414132118225, "learning_rate": 4.3489400906280724e-05, "loss": 0.0333, "step": 45160 }, { "grad_norm": 0.40094849467277527, "learning_rate": 4.3468909057245845e-05, "loss": 0.0278, "step": 45170 }, { "grad_norm": 0.25807109475135803, "learning_rate": 4.3448418324195794e-05, "loss": 0.0339, "step": 45180 }, { "grad_norm": 0.30635979771614075, "learning_rate": 4.342792871063184e-05, "loss": 0.0322, "step": 45190 }, { "grad_norm": 0.38057103753089905, "learning_rate": 4.3407440220055145e-05, "loss": 0.0252, "step": 45200 }, { "grad_norm": 0.3158365488052368, "learning_rate": 4.33869528559666e-05, "loss": 0.0252, "step": 45210 }, { "grad_norm": 0.22863365709781647, "learning_rate": 4.336646662186696e-05, "loss": 0.0276, "step": 45220 }, { "grad_norm": 0.2764401137828827, "learning_rate": 4.334598152125672e-05, "loss": 0.0329, "step": 45230 }, { "grad_norm": 0.3258334696292877, "learning_rate": 4.3325497557636276e-05, "loss": 0.0344, "step": 45240 }, { "grad_norm": 0.2765081822872162, "learning_rate": 4.330501473450574e-05, "loss": 0.0277, "step": 45250 }, { "grad_norm": 0.31254467368125916, "learning_rate": 4.328453305536507e-05, "loss": 0.0217, "step": 45260 }, { "grad_norm": 0.24205182492733002, "learning_rate": 4.326405252371404e-05, "loss": 0.0232, "step": 45270 }, { "grad_norm": 0.43465179204940796, "learning_rate": 4.324357314305221e-05, "loss": 0.0246, "step": 45280 }, { "grad_norm": 0.25122347474098206, "learning_rate": 4.3223094916878945e-05, "loss": 0.0275, "step": 45290 }, { "grad_norm": 0.21251361072063446, "learning_rate": 4.320261784869338e-05, "loss": 0.0336, "step": 45300 }, { "grad_norm": 0.2863582968711853, "learning_rate": 4.318214194199455e-05, "loss": 0.0364, "step": 45310 }, { "grad_norm": 0.24522094428539276, "learning_rate": 4.316166720028118e-05, "loss": 0.0241, "step": 45320 }, { "grad_norm": 0.249123215675354, "learning_rate": 4.3141193627051864e-05, "loss": 0.0292, "step": 45330 }, { "grad_norm": 0.34045976400375366, "learning_rate": 4.312072122580496e-05, "loss": 0.0269, "step": 45340 }, { "grad_norm": 0.20528727769851685, "learning_rate": 4.3100250000038646e-05, "loss": 0.036, "step": 45350 }, { "grad_norm": 0.4081381857395172, "learning_rate": 4.307977995325091e-05, "loss": 0.0374, "step": 45360 }, { "grad_norm": 0.3690406084060669, "learning_rate": 4.30593110889395e-05, "loss": 0.0324, "step": 45370 }, { "grad_norm": 0.2874111235141754, "learning_rate": 4.3038843410602016e-05, "loss": 0.0266, "step": 45380 }, { "grad_norm": 0.24401675164699554, "learning_rate": 4.3018376921735774e-05, "loss": 0.0301, "step": 45390 }, { "grad_norm": 0.30649232864379883, "learning_rate": 4.299791162583799e-05, "loss": 0.0269, "step": 45400 }, { "grad_norm": 0.31628450751304626, "learning_rate": 4.29774475264056e-05, "loss": 0.026, "step": 45410 }, { "grad_norm": 0.23301494121551514, "learning_rate": 4.2956984626935365e-05, "loss": 0.0229, "step": 45420 }, { "grad_norm": 0.601489245891571, "learning_rate": 4.293652293092383e-05, "loss": 0.0307, "step": 45430 }, { "grad_norm": 0.28542643785476685, "learning_rate": 4.2916062441867324e-05, "loss": 0.0213, "step": 45440 }, { "grad_norm": 0.3051929473876953, "learning_rate": 4.289560316326201e-05, "loss": 0.0314, "step": 45450 }, { "grad_norm": 0.6683725118637085, "learning_rate": 4.28751450986038e-05, "loss": 0.0397, "step": 45460 }, { "grad_norm": 0.36703890562057495, "learning_rate": 4.2854688251388444e-05, "loss": 0.0368, "step": 45470 }, { "grad_norm": 0.35711097717285156, "learning_rate": 4.2834232625111425e-05, "loss": 0.0335, "step": 45480 }, { "grad_norm": 0.38023641705513, "learning_rate": 4.2813778223268086e-05, "loss": 0.0275, "step": 45490 }, { "grad_norm": 0.3492583930492401, "learning_rate": 4.2793325049353477e-05, "loss": 0.0276, "step": 45500 }, { "grad_norm": 0.24461619555950165, "learning_rate": 4.2772873106862535e-05, "loss": 0.0246, "step": 45510 }, { "grad_norm": 0.32730716466903687, "learning_rate": 4.275242239928993e-05, "loss": 0.026, "step": 45520 }, { "grad_norm": 0.2344401776790619, "learning_rate": 4.273197293013009e-05, "loss": 0.0217, "step": 45530 }, { "grad_norm": 0.19823332130908966, "learning_rate": 4.271152470287731e-05, "loss": 0.0205, "step": 45540 }, { "grad_norm": 0.2282036691904068, "learning_rate": 4.2691077721025594e-05, "loss": 0.0288, "step": 45550 }, { "grad_norm": 0.4222443103790283, "learning_rate": 4.267063198806883e-05, "loss": 0.0305, "step": 45560 }, { "grad_norm": 0.31529897451400757, "learning_rate": 4.2650187507500574e-05, "loss": 0.0242, "step": 45570 }, { "grad_norm": 1.0766233205795288, "learning_rate": 4.2629744282814275e-05, "loss": 0.0313, "step": 45580 }, { "grad_norm": 0.2940675914287567, "learning_rate": 4.2609302317503074e-05, "loss": 0.0239, "step": 45590 }, { "grad_norm": 0.25227120518684387, "learning_rate": 4.258886161505999e-05, "loss": 0.0201, "step": 45600 }, { "grad_norm": 0.31608065962791443, "learning_rate": 4.2568422178977775e-05, "loss": 0.0298, "step": 45610 }, { "grad_norm": 0.35899871587753296, "learning_rate": 4.254798401274894e-05, "loss": 0.032, "step": 45620 }, { "grad_norm": 0.39202553033828735, "learning_rate": 4.252754711986583e-05, "loss": 0.0263, "step": 45630 }, { "grad_norm": 0.30002158880233765, "learning_rate": 4.250711150382052e-05, "loss": 0.0288, "step": 45640 }, { "grad_norm": 0.2709210515022278, "learning_rate": 4.248667716810495e-05, "loss": 0.0222, "step": 45650 }, { "grad_norm": 0.38872602581977844, "learning_rate": 4.246624411621074e-05, "loss": 0.0243, "step": 45660 }, { "grad_norm": 0.36436569690704346, "learning_rate": 4.244581235162938e-05, "loss": 0.0314, "step": 45670 }, { "grad_norm": 0.27507007122039795, "learning_rate": 4.2425381877852075e-05, "loss": 0.0371, "step": 45680 }, { "grad_norm": 0.3454704284667969, "learning_rate": 4.2404952698369856e-05, "loss": 0.0325, "step": 45690 }, { "grad_norm": 0.3160475790500641, "learning_rate": 4.238452481667349e-05, "loss": 0.0255, "step": 45700 }, { "grad_norm": 0.3840092122554779, "learning_rate": 4.2364098236253526e-05, "loss": 0.0376, "step": 45710 }, { "grad_norm": 0.2886694669723511, "learning_rate": 4.2343672960600356e-05, "loss": 0.0315, "step": 45720 }, { "grad_norm": 0.38507533073425293, "learning_rate": 4.232324899320406e-05, "loss": 0.0249, "step": 45730 }, { "grad_norm": 0.32078760862350464, "learning_rate": 4.230282633755457e-05, "loss": 0.0278, "step": 45740 }, { "grad_norm": 0.3223148584365845, "learning_rate": 4.2282404997141515e-05, "loss": 0.025, "step": 45750 }, { "grad_norm": 0.26345133781433105, "learning_rate": 4.2261984975454397e-05, "loss": 0.0333, "step": 45760 }, { "grad_norm": 0.3404091000556946, "learning_rate": 4.224156627598239e-05, "loss": 0.0237, "step": 45770 }, { "grad_norm": 0.2829555571079254, "learning_rate": 4.222114890221453e-05, "loss": 0.027, "step": 45780 }, { "grad_norm": 0.4208187758922577, "learning_rate": 4.2200732857639546e-05, "loss": 0.0385, "step": 45790 }, { "grad_norm": 0.3410298228263855, "learning_rate": 4.2180318145746035e-05, "loss": 0.0305, "step": 45800 }, { "grad_norm": 0.5119121670722961, "learning_rate": 4.215990477002227e-05, "loss": 0.0383, "step": 45810 }, { "grad_norm": 0.3068057894706726, "learning_rate": 4.2139492733956356e-05, "loss": 0.0282, "step": 45820 }, { "grad_norm": 0.36299100518226624, "learning_rate": 4.211908204103615e-05, "loss": 0.0333, "step": 45830 }, { "grad_norm": 0.351956307888031, "learning_rate": 4.2098672694749265e-05, "loss": 0.0317, "step": 45840 }, { "grad_norm": 0.29388585686683655, "learning_rate": 4.2078264698583133e-05, "loss": 0.0302, "step": 45850 }, { "grad_norm": 0.27682313323020935, "learning_rate": 4.205785805602488e-05, "loss": 0.0292, "step": 45860 }, { "grad_norm": 0.3254208266735077, "learning_rate": 4.203745277056149e-05, "loss": 0.0276, "step": 45870 }, { "grad_norm": 0.3178623914718628, "learning_rate": 4.201704884567964e-05, "loss": 0.0275, "step": 45880 }, { "grad_norm": 0.4442288875579834, "learning_rate": 4.1996646284865816e-05, "loss": 0.035, "step": 45890 }, { "grad_norm": 0.2831260859966278, "learning_rate": 4.197624509160625e-05, "loss": 0.0231, "step": 45900 }, { "grad_norm": 0.24348784983158112, "learning_rate": 4.195584526938692e-05, "loss": 0.0261, "step": 45910 }, { "grad_norm": 0.21868811547756195, "learning_rate": 4.193544682169365e-05, "loss": 0.0282, "step": 45920 }, { "grad_norm": 0.419485867023468, "learning_rate": 4.1915049752011946e-05, "loss": 0.0322, "step": 45930 }, { "grad_norm": 0.3099985122680664, "learning_rate": 4.189465406382712e-05, "loss": 0.0226, "step": 45940 }, { "grad_norm": 0.38333284854888916, "learning_rate": 4.187425976062422e-05, "loss": 0.0315, "step": 45950 }, { "grad_norm": 0.3144848048686981, "learning_rate": 4.18538668458881e-05, "loss": 0.026, "step": 45960 }, { "grad_norm": 0.4408167600631714, "learning_rate": 4.183347532310333e-05, "loss": 0.0312, "step": 45970 }, { "grad_norm": 0.19821900129318237, "learning_rate": 4.181308519575429e-05, "loss": 0.0242, "step": 45980 }, { "grad_norm": 0.36842113733291626, "learning_rate": 4.179269646732507e-05, "loss": 0.0442, "step": 45990 }, { "grad_norm": 1.8681771755218506, "learning_rate": 4.177230914129954e-05, "loss": 0.0307, "step": 46000 }, { "grad_norm": 0.4371603727340698, "learning_rate": 4.175192322116136e-05, "loss": 0.0276, "step": 46010 }, { "grad_norm": 0.45537540316581726, "learning_rate": 4.173153871039391e-05, "loss": 0.0269, "step": 46020 }, { "grad_norm": 0.32393255829811096, "learning_rate": 4.171115561248036e-05, "loss": 0.0229, "step": 46030 }, { "grad_norm": 0.3130713105201721, "learning_rate": 4.16907739309036e-05, "loss": 0.0249, "step": 46040 }, { "grad_norm": 0.36163604259490967, "learning_rate": 4.167039366914633e-05, "loss": 0.0238, "step": 46050 }, { "grad_norm": 0.34745022654533386, "learning_rate": 4.165001483069096e-05, "loss": 0.026, "step": 46060 }, { "grad_norm": 0.41039517521858215, "learning_rate": 4.16296374190197e-05, "loss": 0.0339, "step": 46070 }, { "grad_norm": 0.21769839525222778, "learning_rate": 4.1609261437614464e-05, "loss": 0.0274, "step": 46080 }, { "grad_norm": 0.5209725499153137, "learning_rate": 4.158888688995696e-05, "loss": 0.0288, "step": 46090 }, { "grad_norm": 0.3778620958328247, "learning_rate": 4.1568513779528645e-05, "loss": 0.0242, "step": 46100 }, { "grad_norm": 0.4750834107398987, "learning_rate": 4.1548142109810704e-05, "loss": 0.0399, "step": 46110 }, { "grad_norm": 0.3399733901023865, "learning_rate": 4.152777188428414e-05, "loss": 0.0328, "step": 46120 }, { "grad_norm": 0.28787967562675476, "learning_rate": 4.1507403106429646e-05, "loss": 0.0293, "step": 46130 }, { "grad_norm": 0.3345862329006195, "learning_rate": 4.148703577972768e-05, "loss": 0.0278, "step": 46140 }, { "grad_norm": 0.3708057999610901, "learning_rate": 4.146666990765846e-05, "loss": 0.0283, "step": 46150 }, { "grad_norm": 0.2246943861246109, "learning_rate": 4.1446305493701986e-05, "loss": 0.0258, "step": 46160 }, { "grad_norm": 0.3511385917663574, "learning_rate": 4.142594254133796e-05, "loss": 0.0302, "step": 46170 }, { "grad_norm": 0.3988099992275238, "learning_rate": 4.1405581054045835e-05, "loss": 0.0348, "step": 46180 }, { "grad_norm": 0.26834872364997864, "learning_rate": 4.1385221035304864e-05, "loss": 0.0226, "step": 46190 }, { "grad_norm": 0.2904064953327179, "learning_rate": 4.136486248859398e-05, "loss": 0.0335, "step": 46200 }, { "grad_norm": 0.24010789394378662, "learning_rate": 4.1344505417391955e-05, "loss": 0.0325, "step": 46210 }, { "grad_norm": 0.33896884322166443, "learning_rate": 4.132414982517721e-05, "loss": 0.0306, "step": 46220 }, { "grad_norm": 0.27848559617996216, "learning_rate": 4.130379571542798e-05, "loss": 0.0266, "step": 46230 }, { "grad_norm": 0.2659015655517578, "learning_rate": 4.12834430916222e-05, "loss": 0.0308, "step": 46240 }, { "grad_norm": 0.5384422540664673, "learning_rate": 4.126309195723763e-05, "loss": 0.0296, "step": 46250 }, { "grad_norm": 0.2918144762516022, "learning_rate": 4.124274231575168e-05, "loss": 0.0316, "step": 46260 }, { "grad_norm": 0.2658092677593231, "learning_rate": 4.122239417064154e-05, "loss": 0.0229, "step": 46270 }, { "grad_norm": 0.3025457561016083, "learning_rate": 4.1202047525384184e-05, "loss": 0.0264, "step": 46280 }, { "grad_norm": 0.37909600138664246, "learning_rate": 4.118170238345627e-05, "loss": 0.0287, "step": 46290 }, { "grad_norm": 0.2573390603065491, "learning_rate": 4.1161358748334256e-05, "loss": 0.0232, "step": 46300 }, { "grad_norm": 0.2808193862438202, "learning_rate": 4.1141016623494266e-05, "loss": 0.0321, "step": 46310 }, { "grad_norm": 0.38292837142944336, "learning_rate": 4.112067601241227e-05, "loss": 0.0237, "step": 46320 }, { "grad_norm": 0.2848583161830902, "learning_rate": 4.110033691856387e-05, "loss": 0.0264, "step": 46330 }, { "grad_norm": 0.44706419110298157, "learning_rate": 4.107999934542451e-05, "loss": 0.0307, "step": 46340 }, { "grad_norm": 0.3563629388809204, "learning_rate": 4.105966329646928e-05, "loss": 0.0277, "step": 46350 }, { "grad_norm": 0.38719305396080017, "learning_rate": 4.103932877517308e-05, "loss": 0.0264, "step": 46360 }, { "grad_norm": 0.38082531094551086, "learning_rate": 4.101899578501052e-05, "loss": 0.0303, "step": 46370 }, { "grad_norm": 0.5552971959114075, "learning_rate": 4.099866432945595e-05, "loss": 0.0308, "step": 46380 }, { "grad_norm": 0.4515254497528076, "learning_rate": 4.097833441198346e-05, "loss": 0.0229, "step": 46390 }, { "grad_norm": 1.044371247291565, "learning_rate": 4.095800603606685e-05, "loss": 0.0263, "step": 46400 }, { "grad_norm": 0.3000558912754059, "learning_rate": 4.093767920517975e-05, "loss": 0.0238, "step": 46410 }, { "grad_norm": 0.49306219816207886, "learning_rate": 4.091735392279539e-05, "loss": 0.0252, "step": 46420 }, { "grad_norm": 0.2245464026927948, "learning_rate": 4.089703019238685e-05, "loss": 0.0239, "step": 46430 }, { "grad_norm": 0.3768848478794098, "learning_rate": 4.0876708017426866e-05, "loss": 0.0295, "step": 46440 }, { "grad_norm": 0.2404879778623581, "learning_rate": 4.085638740138798e-05, "loss": 0.0264, "step": 46450 }, { "grad_norm": 0.3156888484954834, "learning_rate": 4.08360683477424e-05, "loss": 0.0284, "step": 46460 }, { "grad_norm": 0.30731064081192017, "learning_rate": 4.0815750859962085e-05, "loss": 0.0292, "step": 46470 }, { "grad_norm": 0.2593454420566559, "learning_rate": 4.079543494151879e-05, "loss": 0.0268, "step": 46480 }, { "grad_norm": 0.20382623374462128, "learning_rate": 4.07751205958839e-05, "loss": 0.0254, "step": 46490 }, { "grad_norm": 0.3260827362537384, "learning_rate": 4.0754807826528615e-05, "loss": 0.024, "step": 46500 }, { "grad_norm": 0.25850996375083923, "learning_rate": 4.073449663692379e-05, "loss": 0.0346, "step": 46510 }, { "grad_norm": 0.2155696600675583, "learning_rate": 4.07141870305401e-05, "loss": 0.0257, "step": 46520 }, { "grad_norm": 0.39113086462020874, "learning_rate": 4.0693879010847866e-05, "loss": 0.0447, "step": 46530 }, { "grad_norm": 0.3500964343547821, "learning_rate": 4.06735725813172e-05, "loss": 0.0275, "step": 46540 }, { "grad_norm": 0.30327218770980835, "learning_rate": 4.0653267745417903e-05, "loss": 0.0278, "step": 46550 }, { "grad_norm": 0.32602179050445557, "learning_rate": 4.063296450661949e-05, "loss": 0.0253, "step": 46560 }, { "grad_norm": 0.3982005715370178, "learning_rate": 4.061266286839128e-05, "loss": 0.0293, "step": 46570 }, { "grad_norm": 0.5854814648628235, "learning_rate": 4.0592362834202225e-05, "loss": 0.0305, "step": 46580 }, { "grad_norm": 0.2244933545589447, "learning_rate": 4.057206440752107e-05, "loss": 0.0253, "step": 46590 }, { "grad_norm": 0.2845698297023773, "learning_rate": 4.0551767591816245e-05, "loss": 0.028, "step": 46600 }, { "grad_norm": 0.23458905518054962, "learning_rate": 4.0531472390555935e-05, "loss": 0.0277, "step": 46610 }, { "grad_norm": 0.31181788444519043, "learning_rate": 4.051117880720802e-05, "loss": 0.0257, "step": 46620 }, { "grad_norm": 0.34296953678131104, "learning_rate": 4.049088684524015e-05, "loss": 0.0278, "step": 46630 }, { "grad_norm": 0.31305158138275146, "learning_rate": 4.0470596508119636e-05, "loss": 0.0234, "step": 46640 }, { "grad_norm": 0.28424593806266785, "learning_rate": 4.0450307799313524e-05, "loss": 0.0283, "step": 46650 }, { "grad_norm": 0.3251582682132721, "learning_rate": 4.0430020722288656e-05, "loss": 0.031, "step": 46660 }, { "grad_norm": 0.3097296357154846, "learning_rate": 4.040973528051148e-05, "loss": 0.0338, "step": 46670 }, { "grad_norm": 0.289883553981781, "learning_rate": 4.038945147744827e-05, "loss": 0.0271, "step": 46680 }, { "grad_norm": 0.2773623764514923, "learning_rate": 4.0369169316564945e-05, "loss": 0.0279, "step": 46690 }, { "grad_norm": 0.31666624546051025, "learning_rate": 4.034888880132718e-05, "loss": 0.029, "step": 46700 }, { "grad_norm": 0.33274537324905396, "learning_rate": 4.032860993520035e-05, "loss": 0.0268, "step": 46710 }, { "grad_norm": 0.22150108218193054, "learning_rate": 4.030833272164959e-05, "loss": 0.0297, "step": 46720 }, { "grad_norm": 0.1661393642425537, "learning_rate": 4.0288057164139705e-05, "loss": 0.031, "step": 46730 }, { "grad_norm": 0.2653530538082123, "learning_rate": 4.0267783266135205e-05, "loss": 0.0312, "step": 46740 }, { "grad_norm": 0.9208967089653015, "learning_rate": 4.024751103110039e-05, "loss": 0.03, "step": 46750 }, { "grad_norm": 0.3086594343185425, "learning_rate": 4.0227240462499176e-05, "loss": 0.0352, "step": 46760 }, { "grad_norm": 0.3658120334148407, "learning_rate": 4.020697156379531e-05, "loss": 0.0362, "step": 46770 }, { "grad_norm": 0.4104883074760437, "learning_rate": 4.018670433845215e-05, "loss": 0.035, "step": 46780 }, { "grad_norm": 0.33253464102745056, "learning_rate": 4.016643878993284e-05, "loss": 0.0278, "step": 46790 }, { "grad_norm": 0.328774094581604, "learning_rate": 4.014617492170017e-05, "loss": 0.0267, "step": 46800 }, { "grad_norm": 0.3559597432613373, "learning_rate": 4.0125912737216726e-05, "loss": 0.0343, "step": 46810 }, { "grad_norm": 0.36034053564071655, "learning_rate": 4.0105652239944735e-05, "loss": 0.0306, "step": 46820 }, { "grad_norm": 0.4124458432197571, "learning_rate": 4.0085393433346144e-05, "loss": 0.0299, "step": 46830 }, { "grad_norm": 0.24719303846359253, "learning_rate": 4.006513632088268e-05, "loss": 0.0171, "step": 46840 }, { "grad_norm": 0.22238929569721222, "learning_rate": 4.004488090601567e-05, "loss": 0.0302, "step": 46850 }, { "grad_norm": 0.22459359467029572, "learning_rate": 4.002462719220626e-05, "loss": 0.0228, "step": 46860 }, { "grad_norm": 0.21206408739089966, "learning_rate": 4.000437518291522e-05, "loss": 0.0286, "step": 46870 }, { "grad_norm": 0.2950855493545532, "learning_rate": 3.9984124881603094e-05, "loss": 0.0242, "step": 46880 }, { "grad_norm": 0.2564554810523987, "learning_rate": 3.9963876291730086e-05, "loss": 0.0206, "step": 46890 }, { "grad_norm": 0.17774678766727448, "learning_rate": 3.994362941675614e-05, "loss": 0.0251, "step": 46900 }, { "grad_norm": 0.2733098268508911, "learning_rate": 3.992338426014088e-05, "loss": 0.0282, "step": 46910 }, { "grad_norm": 0.2943713665008545, "learning_rate": 3.9903140825343636e-05, "loss": 0.0301, "step": 46920 }, { "grad_norm": 0.22364622354507446, "learning_rate": 3.98828991158235e-05, "loss": 0.025, "step": 46930 }, { "grad_norm": 0.3145638406276703, "learning_rate": 3.9862659135039185e-05, "loss": 0.0308, "step": 46940 }, { "grad_norm": 0.26467859745025635, "learning_rate": 3.984242088644918e-05, "loss": 0.0241, "step": 46950 }, { "grad_norm": 0.3178854286670685, "learning_rate": 3.9822184373511615e-05, "loss": 0.0244, "step": 46960 }, { "grad_norm": 0.376301109790802, "learning_rate": 3.98019495996844e-05, "loss": 0.0275, "step": 46970 }, { "grad_norm": 0.1776588261127472, "learning_rate": 3.978171656842507e-05, "loss": 0.0286, "step": 46980 }, { "grad_norm": 0.40390318632125854, "learning_rate": 3.976148528319091e-05, "loss": 0.031, "step": 46990 }, { "grad_norm": 0.20241349935531616, "learning_rate": 3.974125574743888e-05, "loss": 0.024, "step": 47000 }, { "grad_norm": 0.35895872116088867, "learning_rate": 3.9721027964625686e-05, "loss": 0.0252, "step": 47010 }, { "grad_norm": 0.2521379292011261, "learning_rate": 3.9700801938207676e-05, "loss": 0.0212, "step": 47020 }, { "grad_norm": 0.38996827602386475, "learning_rate": 3.9680577671640916e-05, "loss": 0.0237, "step": 47030 }, { "grad_norm": 0.26953279972076416, "learning_rate": 3.966035516838121e-05, "loss": 0.034, "step": 47040 }, { "grad_norm": 0.21092316508293152, "learning_rate": 3.9640134431884014e-05, "loss": 0.0227, "step": 47050 }, { "grad_norm": 0.37828823924064636, "learning_rate": 3.961991546560451e-05, "loss": 0.0259, "step": 47060 }, { "grad_norm": 0.2912762463092804, "learning_rate": 3.959969827299753e-05, "loss": 0.027, "step": 47070 }, { "grad_norm": 0.39027780294418335, "learning_rate": 3.9579482857517684e-05, "loss": 0.0299, "step": 47080 }, { "grad_norm": 0.3926183879375458, "learning_rate": 3.955926922261921e-05, "loss": 0.0285, "step": 47090 }, { "grad_norm": 0.2526826858520508, "learning_rate": 3.9539057371756084e-05, "loss": 0.0257, "step": 47100 }, { "grad_norm": 0.3576968014240265, "learning_rate": 3.951884730838195e-05, "loss": 0.032, "step": 47110 }, { "grad_norm": 0.3328922986984253, "learning_rate": 3.949863903595012e-05, "loss": 0.0276, "step": 47120 }, { "grad_norm": 0.3177435100078583, "learning_rate": 3.947843255791369e-05, "loss": 0.0262, "step": 47130 }, { "grad_norm": 0.23778200149536133, "learning_rate": 3.9458227877725364e-05, "loss": 0.0376, "step": 47140 }, { "grad_norm": 0.26770463585853577, "learning_rate": 3.943802499883758e-05, "loss": 0.0277, "step": 47150 }, { "grad_norm": 0.27279382944107056, "learning_rate": 3.9417823924702437e-05, "loss": 0.0265, "step": 47160 }, { "grad_norm": 0.29477542638778687, "learning_rate": 3.939762465877178e-05, "loss": 0.0313, "step": 47170 }, { "grad_norm": 0.3236062824726105, "learning_rate": 3.937742720449708e-05, "loss": 0.0346, "step": 47180 }, { "grad_norm": 0.3654320538043976, "learning_rate": 3.9357231565329563e-05, "loss": 0.0259, "step": 47190 }, { "grad_norm": 0.3033146858215332, "learning_rate": 3.933703774472008e-05, "loss": 0.0222, "step": 47200 }, { "grad_norm": 0.19182604551315308, "learning_rate": 3.93168457461192e-05, "loss": 0.0208, "step": 47210 }, { "grad_norm": 0.35912084579467773, "learning_rate": 3.9296655572977216e-05, "loss": 0.03, "step": 47220 }, { "grad_norm": 0.2855452597141266, "learning_rate": 3.927646722874404e-05, "loss": 0.0328, "step": 47230 }, { "grad_norm": 0.3353494703769684, "learning_rate": 3.925628071686934e-05, "loss": 0.0291, "step": 47240 }, { "grad_norm": 0.401933878660202, "learning_rate": 3.9236096040802415e-05, "loss": 0.0299, "step": 47250 }, { "grad_norm": 0.45283040404319763, "learning_rate": 3.9215913203992294e-05, "loss": 0.0444, "step": 47260 }, { "grad_norm": 0.2707918882369995, "learning_rate": 3.9195732209887645e-05, "loss": 0.0301, "step": 47270 }, { "grad_norm": 0.2241167426109314, "learning_rate": 3.9175553061936875e-05, "loss": 0.0308, "step": 47280 }, { "grad_norm": 0.24660678207874298, "learning_rate": 3.9155375763588045e-05, "loss": 0.0328, "step": 47290 }, { "grad_norm": 0.3185593783855438, "learning_rate": 3.913520031828889e-05, "loss": 0.0282, "step": 47300 }, { "grad_norm": 0.20014053583145142, "learning_rate": 3.911502672948685e-05, "loss": 0.0306, "step": 47310 }, { "grad_norm": 0.2985779345035553, "learning_rate": 3.9094855000629014e-05, "loss": 0.0292, "step": 47320 }, { "grad_norm": 0.5509023070335388, "learning_rate": 3.907468513516223e-05, "loss": 0.0268, "step": 47330 }, { "grad_norm": 0.31702280044555664, "learning_rate": 3.905451713653294e-05, "loss": 0.0261, "step": 47340 }, { "grad_norm": 0.3109776973724365, "learning_rate": 3.903435100818731e-05, "loss": 0.0261, "step": 47350 }, { "grad_norm": 0.37491700053215027, "learning_rate": 3.901418675357117e-05, "loss": 0.035, "step": 47360 }, { "grad_norm": 0.17865653336048126, "learning_rate": 3.8994024376130075e-05, "loss": 0.035, "step": 47370 }, { "grad_norm": 0.29176008701324463, "learning_rate": 3.8973863879309194e-05, "loss": 0.0301, "step": 47380 }, { "grad_norm": 0.2629626989364624, "learning_rate": 3.8953705266553394e-05, "loss": 0.0286, "step": 47390 }, { "grad_norm": 0.3549432158470154, "learning_rate": 3.893354854130727e-05, "loss": 0.0278, "step": 47400 }, { "grad_norm": 0.2946341335773468, "learning_rate": 3.8913393707015006e-05, "loss": 0.0272, "step": 47410 }, { "grad_norm": 0.32060694694519043, "learning_rate": 3.889324076712056e-05, "loss": 0.0325, "step": 47420 }, { "grad_norm": 0.35620447993278503, "learning_rate": 3.8873089725067476e-05, "loss": 0.0257, "step": 47430 }, { "grad_norm": 0.3123556971549988, "learning_rate": 3.885294058429905e-05, "loss": 0.0407, "step": 47440 }, { "grad_norm": 0.2634073793888092, "learning_rate": 3.8832793348258206e-05, "loss": 0.0326, "step": 47450 }, { "grad_norm": 0.5697506070137024, "learning_rate": 3.881264802038756e-05, "loss": 0.0452, "step": 47460 }, { "grad_norm": 0.3299696147441864, "learning_rate": 3.879250460412939e-05, "loss": 0.0308, "step": 47470 }, { "grad_norm": 0.21743644773960114, "learning_rate": 3.8772363102925644e-05, "loss": 0.0316, "step": 47480 }, { "grad_norm": 0.4192817807197571, "learning_rate": 3.875222352021798e-05, "loss": 0.0278, "step": 47490 }, { "grad_norm": 0.3536454737186432, "learning_rate": 3.8732085859447686e-05, "loss": 0.0346, "step": 47500 }, { "grad_norm": 0.3557846248149872, "learning_rate": 3.871195012405575e-05, "loss": 0.0309, "step": 47510 }, { "grad_norm": 0.2943195402622223, "learning_rate": 3.869181631748278e-05, "loss": 0.0288, "step": 47520 }, { "grad_norm": 0.4225221574306488, "learning_rate": 3.867168444316915e-05, "loss": 0.0269, "step": 47530 }, { "grad_norm": 0.3526115119457245, "learning_rate": 3.865155450455481e-05, "loss": 0.0306, "step": 47540 }, { "grad_norm": 0.3007601201534271, "learning_rate": 3.8631426505079426e-05, "loss": 0.0324, "step": 47550 }, { "grad_norm": 0.37388014793395996, "learning_rate": 3.8611300448182304e-05, "loss": 0.0418, "step": 47560 }, { "grad_norm": 0.3242442011833191, "learning_rate": 3.859117633730248e-05, "loss": 0.0265, "step": 47570 }, { "grad_norm": 0.3610241115093231, "learning_rate": 3.857105417587858e-05, "loss": 0.032, "step": 47580 }, { "grad_norm": 0.3024025559425354, "learning_rate": 3.855093396734894e-05, "loss": 0.0246, "step": 47590 }, { "grad_norm": 0.3491925299167633, "learning_rate": 3.8530815715151545e-05, "loss": 0.0409, "step": 47600 }, { "grad_norm": 0.2838631272315979, "learning_rate": 3.851069942272405e-05, "loss": 0.0308, "step": 47610 }, { "grad_norm": 0.2760331630706787, "learning_rate": 3.849058509350382e-05, "loss": 0.0272, "step": 47620 }, { "grad_norm": 0.20720450580120087, "learning_rate": 3.8470472730927783e-05, "loss": 0.0371, "step": 47630 }, { "grad_norm": 0.3362441658973694, "learning_rate": 3.845036233843264e-05, "loss": 0.0229, "step": 47640 }, { "grad_norm": 0.4596294164657593, "learning_rate": 3.843025391945469e-05, "loss": 0.0336, "step": 47650 }, { "grad_norm": 0.26580414175987244, "learning_rate": 3.841014747742992e-05, "loss": 0.0279, "step": 47660 }, { "grad_norm": 1.2270954847335815, "learning_rate": 3.839004301579397e-05, "loss": 0.0252, "step": 47670 }, { "grad_norm": 0.32452860474586487, "learning_rate": 3.8369940537982097e-05, "loss": 0.0307, "step": 47680 }, { "grad_norm": 0.3292032778263092, "learning_rate": 3.834984004742933e-05, "loss": 0.0284, "step": 47690 }, { "grad_norm": 0.2690221667289734, "learning_rate": 3.832974154757026e-05, "loss": 0.0276, "step": 47700 }, { "grad_norm": 0.44978588819503784, "learning_rate": 3.830964504183919e-05, "loss": 0.0298, "step": 47710 }, { "grad_norm": 0.2996146082878113, "learning_rate": 3.828955053367003e-05, "loss": 0.0301, "step": 47720 }, { "grad_norm": 0.36938610672950745, "learning_rate": 3.826945802649642e-05, "loss": 0.0318, "step": 47730 }, { "grad_norm": 0.29526060819625854, "learning_rate": 3.824936752375159e-05, "loss": 0.0322, "step": 47740 }, { "grad_norm": 0.43261027336120605, "learning_rate": 3.822927902886848e-05, "loss": 0.0277, "step": 47750 }, { "grad_norm": 0.4203372001647949, "learning_rate": 3.8209192545279653e-05, "loss": 0.0284, "step": 47760 }, { "grad_norm": 0.3920585513114929, "learning_rate": 3.8189108076417326e-05, "loss": 0.024, "step": 47770 }, { "grad_norm": 0.28341907262802124, "learning_rate": 3.816902562571342e-05, "loss": 0.0261, "step": 47780 }, { "grad_norm": 0.40418440103530884, "learning_rate": 3.814894519659944e-05, "loss": 0.0232, "step": 47790 }, { "grad_norm": 0.2716235816478729, "learning_rate": 3.812886679250661e-05, "loss": 0.0253, "step": 47800 }, { "grad_norm": 0.21140675246715546, "learning_rate": 3.810879041686575e-05, "loss": 0.0241, "step": 47810 }, { "grad_norm": 0.47795775532722473, "learning_rate": 3.808871607310741e-05, "loss": 0.0318, "step": 47820 }, { "grad_norm": 0.416391521692276, "learning_rate": 3.806864376466169e-05, "loss": 0.0299, "step": 47830 }, { "grad_norm": 0.3899977505207062, "learning_rate": 3.804857349495845e-05, "loss": 0.0272, "step": 47840 }, { "grad_norm": 0.5411096811294556, "learning_rate": 3.802850526742713e-05, "loss": 0.0503, "step": 47850 }, { "grad_norm": 0.24844114482402802, "learning_rate": 3.800843908549683e-05, "loss": 0.0263, "step": 47860 }, { "grad_norm": 0.3324877619743347, "learning_rate": 3.7988374952596325e-05, "loss": 0.0221, "step": 47870 }, { "grad_norm": 0.32661300897598267, "learning_rate": 3.7968312872154e-05, "loss": 0.0229, "step": 47880 }, { "grad_norm": 0.35199272632598877, "learning_rate": 3.7948252847597965e-05, "loss": 0.0273, "step": 47890 }, { "grad_norm": 0.2262992113828659, "learning_rate": 3.7928194882355885e-05, "loss": 0.0221, "step": 47900 }, { "grad_norm": 0.3478792607784271, "learning_rate": 3.790813897985515e-05, "loss": 0.0217, "step": 47910 }, { "grad_norm": 0.315985769033432, "learning_rate": 3.7888085143522726e-05, "loss": 0.0276, "step": 47920 }, { "grad_norm": 0.295158326625824, "learning_rate": 3.7868033376785314e-05, "loss": 0.0264, "step": 47930 }, { "grad_norm": 0.2661047577857971, "learning_rate": 3.784798368306919e-05, "loss": 0.0262, "step": 47940 }, { "grad_norm": 0.3017040193080902, "learning_rate": 3.782793606580029e-05, "loss": 0.0263, "step": 47950 }, { "grad_norm": 0.3441462814807892, "learning_rate": 3.7807890528404205e-05, "loss": 0.0307, "step": 47960 }, { "grad_norm": 0.25303301215171814, "learning_rate": 3.778784707430616e-05, "loss": 0.0245, "step": 47970 }, { "grad_norm": 0.3902812600135803, "learning_rate": 3.776780570693107e-05, "loss": 0.0329, "step": 47980 }, { "grad_norm": 0.21373382210731506, "learning_rate": 3.774776642970342e-05, "loss": 0.0292, "step": 47990 }, { "grad_norm": 0.22393830120563507, "learning_rate": 3.77277292460474e-05, "loss": 0.0211, "step": 48000 }, { "grad_norm": 0.42706170678138733, "learning_rate": 3.770769415938678e-05, "loss": 0.0274, "step": 48010 }, { "grad_norm": 0.32682767510414124, "learning_rate": 3.768766117314506e-05, "loss": 0.0376, "step": 48020 }, { "grad_norm": 0.3446238040924072, "learning_rate": 3.76676302907453e-05, "loss": 0.0307, "step": 48030 }, { "grad_norm": 0.38379865884780884, "learning_rate": 3.764760151561021e-05, "loss": 0.0291, "step": 48040 }, { "grad_norm": 0.24008464813232422, "learning_rate": 3.76275748511622e-05, "loss": 0.0233, "step": 48050 }, { "grad_norm": 0.2851758897304535, "learning_rate": 3.7607550300823255e-05, "loss": 0.0231, "step": 48060 }, { "grad_norm": 0.47847262024879456, "learning_rate": 3.7587527868015044e-05, "loss": 0.0405, "step": 48070 }, { "grad_norm": 0.4937220513820648, "learning_rate": 3.756750755615881e-05, "loss": 0.0263, "step": 48080 }, { "grad_norm": 0.21488437056541443, "learning_rate": 3.754748936867553e-05, "loss": 0.0261, "step": 48090 }, { "grad_norm": 0.2506716251373291, "learning_rate": 3.752747330898573e-05, "loss": 0.0318, "step": 48100 }, { "grad_norm": 0.5097799897193909, "learning_rate": 3.750745938050962e-05, "loss": 0.0284, "step": 48110 }, { "grad_norm": 0.25794243812561035, "learning_rate": 3.7487447586667025e-05, "loss": 0.0244, "step": 48120 }, { "grad_norm": 0.21330364048480988, "learning_rate": 3.7467437930877426e-05, "loss": 0.0215, "step": 48130 }, { "grad_norm": 0.34173154830932617, "learning_rate": 3.744743041655992e-05, "loss": 0.0242, "step": 48140 }, { "grad_norm": 0.2893482446670532, "learning_rate": 3.742742504713324e-05, "loss": 0.0248, "step": 48150 }, { "grad_norm": 0.2003277987241745, "learning_rate": 3.740742182601576e-05, "loss": 0.0308, "step": 48160 }, { "grad_norm": 0.2862209975719452, "learning_rate": 3.7387420756625464e-05, "loss": 0.0298, "step": 48170 }, { "grad_norm": 0.29473569989204407, "learning_rate": 3.736742184238002e-05, "loss": 0.024, "step": 48180 }, { "grad_norm": 0.4028685986995697, "learning_rate": 3.7347425086696684e-05, "loss": 0.0275, "step": 48190 }, { "grad_norm": 0.410108745098114, "learning_rate": 3.732743049299235e-05, "loss": 0.0325, "step": 48200 }, { "grad_norm": 0.2783535122871399, "learning_rate": 3.730743806468354e-05, "loss": 0.0225, "step": 48210 }, { "grad_norm": 0.26562854647636414, "learning_rate": 3.7287447805186436e-05, "loss": 0.0233, "step": 48220 }, { "grad_norm": 0.4218313992023468, "learning_rate": 3.726745971791682e-05, "loss": 0.0339, "step": 48230 }, { "grad_norm": 0.4405825734138489, "learning_rate": 3.724747380629008e-05, "loss": 0.0426, "step": 48240 }, { "grad_norm": 0.24536627531051636, "learning_rate": 3.72274900737213e-05, "loss": 0.0297, "step": 48250 }, { "grad_norm": 0.2860066592693329, "learning_rate": 3.7207508523625123e-05, "loss": 0.0251, "step": 48260 }, { "grad_norm": 0.45370447635650635, "learning_rate": 3.718752915941588e-05, "loss": 0.024, "step": 48270 }, { "grad_norm": 0.237461656332016, "learning_rate": 3.7167551984507464e-05, "loss": 0.0258, "step": 48280 }, { "grad_norm": 0.27103137969970703, "learning_rate": 3.714757700231346e-05, "loss": 0.0272, "step": 48290 }, { "grad_norm": 0.2951757609844208, "learning_rate": 3.712760421624703e-05, "loss": 0.0237, "step": 48300 }, { "grad_norm": 0.3675907552242279, "learning_rate": 3.710763362972099e-05, "loss": 0.026, "step": 48310 }, { "grad_norm": 0.27711474895477295, "learning_rate": 3.708766524614774e-05, "loss": 0.0244, "step": 48320 }, { "grad_norm": 0.34653371572494507, "learning_rate": 3.7067699068939335e-05, "loss": 0.0192, "step": 48330 }, { "grad_norm": 0.25353872776031494, "learning_rate": 3.704773510150748e-05, "loss": 0.0245, "step": 48340 }, { "grad_norm": 0.36530470848083496, "learning_rate": 3.702777334726344e-05, "loss": 0.0285, "step": 48350 }, { "grad_norm": 0.25863027572631836, "learning_rate": 3.7007813809618164e-05, "loss": 0.0216, "step": 48360 }, { "grad_norm": 0.2497633546590805, "learning_rate": 3.6987856491982145e-05, "loss": 0.0307, "step": 48370 }, { "grad_norm": 0.8471742868423462, "learning_rate": 3.696790139776558e-05, "loss": 0.0285, "step": 48380 }, { "grad_norm": 0.22302567958831787, "learning_rate": 3.6947948530378235e-05, "loss": 0.0264, "step": 48390 }, { "grad_norm": 0.4554558992385864, "learning_rate": 3.6927997893229516e-05, "loss": 0.0258, "step": 48400 }, { "grad_norm": 0.18091338872909546, "learning_rate": 3.690804948972845e-05, "loss": 0.0207, "step": 48410 }, { "grad_norm": 0.3251163959503174, "learning_rate": 3.688810332328363e-05, "loss": 0.0279, "step": 48420 }, { "grad_norm": 0.24582266807556152, "learning_rate": 3.686815939730336e-05, "loss": 0.0298, "step": 48430 }, { "grad_norm": 0.29646000266075134, "learning_rate": 3.684821771519548e-05, "loss": 0.0302, "step": 48440 }, { "grad_norm": 0.30500003695487976, "learning_rate": 3.68282782803675e-05, "loss": 0.0271, "step": 48450 }, { "grad_norm": 0.32887667417526245, "learning_rate": 3.6808341096226504e-05, "loss": 0.0339, "step": 48460 }, { "grad_norm": 0.34532564878463745, "learning_rate": 3.678840616617924e-05, "loss": 0.0267, "step": 48470 }, { "grad_norm": 0.2950853109359741, "learning_rate": 3.6768473493632e-05, "loss": 0.0407, "step": 48480 }, { "grad_norm": 0.2753916382789612, "learning_rate": 3.6748543081990783e-05, "loss": 0.0238, "step": 48490 }, { "grad_norm": 0.2220233976840973, "learning_rate": 3.672861493466112e-05, "loss": 0.0229, "step": 48500 }, { "grad_norm": 0.2264292687177658, "learning_rate": 3.670868905504818e-05, "loss": 0.024, "step": 48510 }, { "grad_norm": 0.4414746165275574, "learning_rate": 3.6688765446556784e-05, "loss": 0.0259, "step": 48520 }, { "grad_norm": 0.345072478055954, "learning_rate": 3.6668844112591276e-05, "loss": 0.0299, "step": 48530 }, { "grad_norm": 0.2411363571882248, "learning_rate": 3.664892505655573e-05, "loss": 0.0238, "step": 48540 }, { "grad_norm": 0.3355760872364044, "learning_rate": 3.662900828185373e-05, "loss": 0.0311, "step": 48550 }, { "grad_norm": 0.21856296062469482, "learning_rate": 3.6609093791888516e-05, "loss": 0.0253, "step": 48560 }, { "grad_norm": 0.36397644877433777, "learning_rate": 3.658918159006292e-05, "loss": 0.0352, "step": 48570 }, { "grad_norm": 0.7630717754364014, "learning_rate": 3.656927167977942e-05, "loss": 0.0291, "step": 48580 }, { "grad_norm": 0.41778674721717834, "learning_rate": 3.654936406444006e-05, "loss": 0.0308, "step": 48590 }, { "grad_norm": 0.38441118597984314, "learning_rate": 3.65294587474465e-05, "loss": 0.0264, "step": 48600 }, { "grad_norm": 0.39109259843826294, "learning_rate": 3.650955573220002e-05, "loss": 0.0546, "step": 48610 }, { "grad_norm": 0.3136315643787384, "learning_rate": 3.648965502210149e-05, "loss": 0.0285, "step": 48620 }, { "grad_norm": 0.2558601200580597, "learning_rate": 3.646975662055142e-05, "loss": 0.0266, "step": 48630 }, { "grad_norm": 0.23731087148189545, "learning_rate": 3.644986053094987e-05, "loss": 0.0285, "step": 48640 }, { "grad_norm": 0.3295361399650574, "learning_rate": 3.642996675669659e-05, "loss": 0.0266, "step": 48650 }, { "grad_norm": 0.34680235385894775, "learning_rate": 3.641007530119083e-05, "loss": 0.0237, "step": 48660 }, { "grad_norm": 0.3526586592197418, "learning_rate": 3.639018616783153e-05, "loss": 0.0276, "step": 48670 }, { "grad_norm": 0.18979255855083466, "learning_rate": 3.637029936001719e-05, "loss": 0.0243, "step": 48680 }, { "grad_norm": 0.43930378556251526, "learning_rate": 3.6350414881145886e-05, "loss": 0.0298, "step": 48690 }, { "grad_norm": 0.6061629056930542, "learning_rate": 3.6330532734615386e-05, "loss": 0.02, "step": 48700 }, { "grad_norm": 0.3473208248615265, "learning_rate": 3.6310652923822975e-05, "loss": 0.0303, "step": 48710 }, { "grad_norm": 0.27596694231033325, "learning_rate": 3.629077545216558e-05, "loss": 0.0298, "step": 48720 }, { "grad_norm": 0.3076344430446625, "learning_rate": 3.627090032303969e-05, "loss": 0.0313, "step": 48730 }, { "grad_norm": 0.3430240750312805, "learning_rate": 3.625102753984146e-05, "loss": 0.0341, "step": 48740 }, { "grad_norm": 0.3784986734390259, "learning_rate": 3.623115710596659e-05, "loss": 0.0276, "step": 48750 }, { "grad_norm": 0.5643737316131592, "learning_rate": 3.6211289024810395e-05, "loss": 0.032, "step": 48760 }, { "grad_norm": 0.35653308033943176, "learning_rate": 3.619142329976777e-05, "loss": 0.0269, "step": 48770 }, { "grad_norm": 0.2932227849960327, "learning_rate": 3.6171559934233247e-05, "loss": 0.0209, "step": 48780 }, { "grad_norm": 0.2563788592815399, "learning_rate": 3.615169893160093e-05, "loss": 0.0236, "step": 48790 }, { "grad_norm": 0.3883267045021057, "learning_rate": 3.61318402952645e-05, "loss": 0.0291, "step": 48800 }, { "grad_norm": 0.43621277809143066, "learning_rate": 3.6111984028617285e-05, "loss": 0.0328, "step": 48810 }, { "grad_norm": 0.25798743963241577, "learning_rate": 3.6092130135052134e-05, "loss": 0.0249, "step": 48820 }, { "grad_norm": 0.32144835591316223, "learning_rate": 3.6072278617961584e-05, "loss": 0.0322, "step": 48830 }, { "grad_norm": 0.4369761049747467, "learning_rate": 3.605242948073767e-05, "loss": 0.0347, "step": 48840 }, { "grad_norm": 0.45225924253463745, "learning_rate": 3.603258272677212e-05, "loss": 0.0248, "step": 48850 }, { "grad_norm": 0.3053380846977234, "learning_rate": 3.601273835945616e-05, "loss": 0.0337, "step": 48860 }, { "grad_norm": 0.2526508867740631, "learning_rate": 3.5992896382180664e-05, "loss": 0.0297, "step": 48870 }, { "grad_norm": 0.47149723768234253, "learning_rate": 3.597305679833609e-05, "loss": 0.0351, "step": 48880 }, { "grad_norm": 0.3141786754131317, "learning_rate": 3.595321961131245e-05, "loss": 0.0278, "step": 48890 }, { "grad_norm": 0.38470759987831116, "learning_rate": 3.593338482449942e-05, "loss": 0.0271, "step": 48900 }, { "grad_norm": 0.39210841059684753, "learning_rate": 3.591355244128618e-05, "loss": 0.0353, "step": 48910 }, { "grad_norm": 0.33842724561691284, "learning_rate": 3.589372246506158e-05, "loss": 0.0297, "step": 48920 }, { "grad_norm": 0.2637436091899872, "learning_rate": 3.5873894899213984e-05, "loss": 0.0247, "step": 48930 }, { "grad_norm": 0.36833465099334717, "learning_rate": 3.5854069747131416e-05, "loss": 0.0319, "step": 48940 }, { "grad_norm": 0.26607170701026917, "learning_rate": 3.583424701220143e-05, "loss": 0.0291, "step": 48950 }, { "grad_norm": 0.48030784726142883, "learning_rate": 3.581442669781121e-05, "loss": 0.0269, "step": 48960 }, { "grad_norm": 0.2922275960445404, "learning_rate": 3.579460880734749e-05, "loss": 0.0247, "step": 48970 }, { "grad_norm": 0.2363666445016861, "learning_rate": 3.577479334419657e-05, "loss": 0.0242, "step": 48980 }, { "grad_norm": 0.28801417350769043, "learning_rate": 3.575498031174444e-05, "loss": 0.0214, "step": 48990 }, { "grad_norm": 0.41997548937797546, "learning_rate": 3.573516971337657e-05, "loss": 0.0261, "step": 49000 }, { "grad_norm": 0.3913714289665222, "learning_rate": 3.5715361552478046e-05, "loss": 0.0299, "step": 49010 }, { "grad_norm": 0.2803252339363098, "learning_rate": 3.5695555832433536e-05, "loss": 0.0272, "step": 49020 }, { "grad_norm": 0.2724943161010742, "learning_rate": 3.5675752556627325e-05, "loss": 0.0241, "step": 49030 }, { "grad_norm": 0.358425498008728, "learning_rate": 3.565595172844322e-05, "loss": 0.032, "step": 49040 }, { "grad_norm": 0.2933673560619354, "learning_rate": 3.5636153351264666e-05, "loss": 0.0302, "step": 49050 }, { "grad_norm": 0.25223371386528015, "learning_rate": 3.5616357428474655e-05, "loss": 0.0291, "step": 49060 }, { "grad_norm": 0.24176755547523499, "learning_rate": 3.559656396345575e-05, "loss": 0.0289, "step": 49070 }, { "grad_norm": 0.37704965472221375, "learning_rate": 3.5576772959590146e-05, "loss": 0.0253, "step": 49080 }, { "grad_norm": 0.29905980825424194, "learning_rate": 3.5556984420259545e-05, "loss": 0.0253, "step": 49090 }, { "grad_norm": 0.21997404098510742, "learning_rate": 3.5537198348845305e-05, "loss": 0.0228, "step": 49100 }, { "grad_norm": 0.23586106300354004, "learning_rate": 3.551741474872831e-05, "loss": 0.0258, "step": 49110 }, { "grad_norm": 0.6898719072341919, "learning_rate": 3.549763362328903e-05, "loss": 0.0275, "step": 49120 }, { "grad_norm": 0.2966266870498657, "learning_rate": 3.5477854975907515e-05, "loss": 0.0259, "step": 49130 }, { "grad_norm": 0.30270177125930786, "learning_rate": 3.5458078809963416e-05, "loss": 0.0241, "step": 49140 }, { "grad_norm": 0.24779631197452545, "learning_rate": 3.543830512883594e-05, "loss": 0.0243, "step": 49150 }, { "grad_norm": 0.48094722628593445, "learning_rate": 3.5418533935903824e-05, "loss": 0.0314, "step": 49160 }, { "grad_norm": 0.40183404088020325, "learning_rate": 3.539876523454547e-05, "loss": 0.0234, "step": 49170 }, { "grad_norm": 0.3270345628261566, "learning_rate": 3.537899902813878e-05, "loss": 0.0208, "step": 49180 }, { "grad_norm": 0.4396396577358246, "learning_rate": 3.5359235320061293e-05, "loss": 0.0236, "step": 49190 }, { "grad_norm": 0.26670143008232117, "learning_rate": 3.533947411369003e-05, "loss": 0.0241, "step": 49200 }, { "grad_norm": 0.24952442944049835, "learning_rate": 3.53197154124017e-05, "loss": 0.0182, "step": 49210 }, { "grad_norm": 0.2977648377418518, "learning_rate": 3.52999592195725e-05, "loss": 0.0265, "step": 49220 }, { "grad_norm": 0.3835899233818054, "learning_rate": 3.5280205538578224e-05, "loss": 0.0229, "step": 49230 }, { "grad_norm": 0.3760789930820465, "learning_rate": 3.5260454372794236e-05, "loss": 0.024, "step": 49240 }, { "grad_norm": 0.18844331800937653, "learning_rate": 3.524070572559545e-05, "loss": 0.0232, "step": 49250 }, { "grad_norm": 0.2242782860994339, "learning_rate": 3.5220959600356395e-05, "loss": 0.0223, "step": 49260 }, { "grad_norm": 0.28269538283348083, "learning_rate": 3.5201216000451145e-05, "loss": 0.0205, "step": 49270 }, { "grad_norm": 0.3284122049808502, "learning_rate": 3.5181474929253335e-05, "loss": 0.0246, "step": 49280 }, { "grad_norm": 0.3466275930404663, "learning_rate": 3.516173639013615e-05, "loss": 0.0226, "step": 49290 }, { "grad_norm": 0.21626801788806915, "learning_rate": 3.5142000386472406e-05, "loss": 0.0284, "step": 49300 }, { "grad_norm": 0.7098300457000732, "learning_rate": 3.5122266921634427e-05, "loss": 0.0259, "step": 49310 }, { "grad_norm": 0.34332576394081116, "learning_rate": 3.510253599899413e-05, "loss": 0.0257, "step": 49320 }, { "grad_norm": 0.24407848715782166, "learning_rate": 3.5082807621922965e-05, "loss": 0.022, "step": 49330 }, { "grad_norm": 0.34730347990989685, "learning_rate": 3.506308179379201e-05, "loss": 0.0276, "step": 49340 }, { "grad_norm": 0.35806843638420105, "learning_rate": 3.5043358517971844e-05, "loss": 0.0253, "step": 49350 }, { "grad_norm": 0.2937357723712921, "learning_rate": 3.502363779783264e-05, "loss": 0.0251, "step": 49360 }, { "grad_norm": 0.2990967035293579, "learning_rate": 3.500391963674415e-05, "loss": 0.0214, "step": 49370 }, { "grad_norm": 0.340166836977005, "learning_rate": 3.4984204038075615e-05, "loss": 0.0279, "step": 49380 }, { "grad_norm": 0.26615673303604126, "learning_rate": 3.496449100519595e-05, "loss": 0.0311, "step": 49390 }, { "grad_norm": 0.3147943317890167, "learning_rate": 3.494478054147354e-05, "loss": 0.0256, "step": 49400 }, { "grad_norm": 0.36668550968170166, "learning_rate": 3.4925072650276395e-05, "loss": 0.0252, "step": 49410 }, { "grad_norm": 0.3546971380710602, "learning_rate": 3.4905367334972016e-05, "loss": 0.0262, "step": 49420 }, { "grad_norm": 0.3221641480922699, "learning_rate": 3.488566459892752e-05, "loss": 0.0241, "step": 49430 }, { "grad_norm": 0.22279685735702515, "learning_rate": 3.4865964445509585e-05, "loss": 0.0226, "step": 49440 }, { "grad_norm": 0.25079092383384705, "learning_rate": 3.484626687808438e-05, "loss": 0.0256, "step": 49450 }, { "grad_norm": 0.3386121690273285, "learning_rate": 3.4826571900017735e-05, "loss": 0.0258, "step": 49460 }, { "grad_norm": 0.2935529053211212, "learning_rate": 3.480687951467495e-05, "loss": 0.0241, "step": 49470 }, { "grad_norm": 0.39419302344322205, "learning_rate": 3.4787189725420925e-05, "loss": 0.0291, "step": 49480 }, { "grad_norm": 0.23677098751068115, "learning_rate": 3.4767502535620086e-05, "loss": 0.0196, "step": 49490 }, { "grad_norm": 0.29739880561828613, "learning_rate": 3.474781794863648e-05, "loss": 0.034, "step": 49500 }, { "grad_norm": 0.3578672409057617, "learning_rate": 3.472813596783363e-05, "loss": 0.0288, "step": 49510 }, { "grad_norm": 0.2830444574356079, "learning_rate": 3.470845659657466e-05, "loss": 0.0287, "step": 49520 }, { "grad_norm": 0.6593328714370728, "learning_rate": 3.468877983822223e-05, "loss": 0.0255, "step": 49530 }, { "grad_norm": 0.18474207818508148, "learning_rate": 3.466910569613855e-05, "loss": 0.0216, "step": 49540 }, { "grad_norm": 0.3248511552810669, "learning_rate": 3.464943417368542e-05, "loss": 0.0385, "step": 49550 }, { "grad_norm": 0.2802344262599945, "learning_rate": 3.462976527422415e-05, "loss": 0.0221, "step": 49560 }, { "grad_norm": 0.4479866325855255, "learning_rate": 3.461009900111562e-05, "loss": 0.0272, "step": 49570 }, { "grad_norm": 0.29422834515571594, "learning_rate": 3.459043535772023e-05, "loss": 0.0257, "step": 49580 }, { "grad_norm": 0.30449724197387695, "learning_rate": 3.4570774347398014e-05, "loss": 0.0266, "step": 49590 }, { "grad_norm": 0.2731662392616272, "learning_rate": 3.4551115973508454e-05, "loss": 0.0284, "step": 49600 }, { "grad_norm": 0.4074258804321289, "learning_rate": 3.453146023941066e-05, "loss": 0.0295, "step": 49610 }, { "grad_norm": 0.2531382739543915, "learning_rate": 3.451180714846325e-05, "loss": 0.0256, "step": 49620 }, { "grad_norm": 0.23659966886043549, "learning_rate": 3.449215670402438e-05, "loss": 0.03, "step": 49630 }, { "grad_norm": 0.6867993474006653, "learning_rate": 3.447250890945181e-05, "loss": 0.0275, "step": 49640 }, { "grad_norm": 0.31755331158638, "learning_rate": 3.4452863768102754e-05, "loss": 0.0219, "step": 49650 }, { "grad_norm": 0.23244144022464752, "learning_rate": 3.443322128333409e-05, "loss": 0.0261, "step": 49660 }, { "grad_norm": 0.36761990189552307, "learning_rate": 3.441358145850215e-05, "loss": 0.0279, "step": 49670 }, { "grad_norm": 0.34697866439819336, "learning_rate": 3.439394429696286e-05, "loss": 0.0276, "step": 49680 }, { "grad_norm": 0.4018297791481018, "learning_rate": 3.4374309802071644e-05, "loss": 0.0258, "step": 49690 }, { "grad_norm": 0.2682487964630127, "learning_rate": 3.435467797718353e-05, "loss": 0.0273, "step": 49700 }, { "grad_norm": 0.288592666387558, "learning_rate": 3.433504882565306e-05, "loss": 0.0236, "step": 49710 }, { "grad_norm": 0.20388957858085632, "learning_rate": 3.43154223508343e-05, "loss": 0.0288, "step": 49720 }, { "grad_norm": 0.23121587932109833, "learning_rate": 3.429579855608089e-05, "loss": 0.0232, "step": 49730 }, { "grad_norm": 0.30692026019096375, "learning_rate": 3.427617744474597e-05, "loss": 0.0241, "step": 49740 }, { "grad_norm": 0.29577434062957764, "learning_rate": 3.425655902018231e-05, "loss": 0.0229, "step": 49750 }, { "grad_norm": 0.28320470452308655, "learning_rate": 3.423694328574211e-05, "loss": 0.0276, "step": 49760 }, { "grad_norm": 0.29956114292144775, "learning_rate": 3.42173302447772e-05, "loss": 0.0267, "step": 49770 }, { "grad_norm": 0.25782445073127747, "learning_rate": 3.419771990063886e-05, "loss": 0.0197, "step": 49780 }, { "grad_norm": 0.4272925853729248, "learning_rate": 3.417811225667803e-05, "loss": 0.0196, "step": 49790 }, { "grad_norm": 0.1955159604549408, "learning_rate": 3.415850731624508e-05, "loss": 0.0213, "step": 49800 }, { "grad_norm": 0.22611111402511597, "learning_rate": 3.4138905082689945e-05, "loss": 0.0285, "step": 49810 }, { "grad_norm": 0.2827783226966858, "learning_rate": 3.4119305559362145e-05, "loss": 0.0328, "step": 49820 }, { "grad_norm": 0.34999319911003113, "learning_rate": 3.4099708749610684e-05, "loss": 0.0267, "step": 49830 }, { "grad_norm": 0.5287339687347412, "learning_rate": 3.408011465678413e-05, "loss": 0.031, "step": 49840 }, { "grad_norm": 0.20671580731868744, "learning_rate": 3.406052328423055e-05, "loss": 0.0266, "step": 49850 }, { "grad_norm": 0.3505605161190033, "learning_rate": 3.4040934635297615e-05, "loss": 0.0245, "step": 49860 }, { "grad_norm": 0.4145299792289734, "learning_rate": 3.4021348713332466e-05, "loss": 0.022, "step": 49870 }, { "grad_norm": 0.49741673469543457, "learning_rate": 3.4001765521681807e-05, "loss": 0.0274, "step": 49880 }, { "grad_norm": 0.2788117825984955, "learning_rate": 3.398218506369188e-05, "loss": 0.0329, "step": 49890 }, { "grad_norm": 0.48873695731163025, "learning_rate": 3.3962607342708404e-05, "loss": 0.0265, "step": 49900 }, { "grad_norm": 0.3616291880607605, "learning_rate": 3.394303236207673e-05, "loss": 0.0266, "step": 49910 }, { "grad_norm": 0.2697364389896393, "learning_rate": 3.392346012514166e-05, "loss": 0.0261, "step": 49920 }, { "grad_norm": 0.4628520607948303, "learning_rate": 3.390389063524757e-05, "loss": 0.0221, "step": 49930 }, { "grad_norm": 0.17878121137619019, "learning_rate": 3.3884323895738324e-05, "loss": 0.025, "step": 49940 }, { "grad_norm": 0.3842317759990692, "learning_rate": 3.386475990995738e-05, "loss": 0.0257, "step": 49950 }, { "grad_norm": 0.3208770751953125, "learning_rate": 3.384519868124765e-05, "loss": 0.0225, "step": 49960 }, { "grad_norm": 0.36625128984451294, "learning_rate": 3.3825640212951645e-05, "loss": 0.0238, "step": 49970 }, { "grad_norm": 0.27916422486305237, "learning_rate": 3.380608450841134e-05, "loss": 0.0222, "step": 49980 }, { "grad_norm": 0.3773685097694397, "learning_rate": 3.3786531570968305e-05, "loss": 0.0236, "step": 49990 }, { "grad_norm": 0.3097712993621826, "learning_rate": 3.3766981403963584e-05, "loss": 0.0286, "step": 50000 }, { "grad_norm": 0.2901213765144348, "learning_rate": 3.374743401073775e-05, "loss": 0.0227, "step": 50010 }, { "grad_norm": 0.3819594979286194, "learning_rate": 3.372788939463095e-05, "loss": 0.0209, "step": 50020 }, { "grad_norm": 0.3408072292804718, "learning_rate": 3.370834755898281e-05, "loss": 0.0201, "step": 50030 }, { "grad_norm": 0.265224426984787, "learning_rate": 3.3688808507132493e-05, "loss": 0.0269, "step": 50040 }, { "grad_norm": 0.163824662566185, "learning_rate": 3.3669272242418685e-05, "loss": 0.0285, "step": 50050 }, { "grad_norm": 0.2848840057849884, "learning_rate": 3.364973876817961e-05, "loss": 0.0314, "step": 50060 }, { "grad_norm": 0.3839367628097534, "learning_rate": 3.363020808775299e-05, "loss": 0.0243, "step": 50070 }, { "grad_norm": 0.24585673213005066, "learning_rate": 3.361068020447611e-05, "loss": 0.0294, "step": 50080 }, { "grad_norm": 0.23527373373508453, "learning_rate": 3.3591155121685724e-05, "loss": 0.031, "step": 50090 }, { "grad_norm": 0.3157847225666046, "learning_rate": 3.357163284271814e-05, "loss": 0.0204, "step": 50100 }, { "grad_norm": 0.34025838971138, "learning_rate": 3.355211337090919e-05, "loss": 0.0214, "step": 50110 }, { "grad_norm": 0.1914052814245224, "learning_rate": 3.353259670959421e-05, "loss": 0.0282, "step": 50120 }, { "grad_norm": 0.24017344415187836, "learning_rate": 3.351308286210808e-05, "loss": 0.022, "step": 50130 }, { "grad_norm": 0.2031400352716446, "learning_rate": 3.3493571831785156e-05, "loss": 0.0221, "step": 50140 }, { "grad_norm": 0.2691650092601776, "learning_rate": 3.347406362195936e-05, "loss": 0.0246, "step": 50150 }, { "grad_norm": 0.39617347717285156, "learning_rate": 3.345455823596411e-05, "loss": 0.0244, "step": 50160 }, { "grad_norm": 0.253402441740036, "learning_rate": 3.3435055677132346e-05, "loss": 0.0245, "step": 50170 }, { "grad_norm": 0.4167455732822418, "learning_rate": 3.3415555948796505e-05, "loss": 0.0308, "step": 50180 }, { "grad_norm": 0.40133097767829895, "learning_rate": 3.3396059054288556e-05, "loss": 0.0246, "step": 50190 }, { "grad_norm": 0.6910465359687805, "learning_rate": 3.3376564996940015e-05, "loss": 0.0383, "step": 50200 }, { "grad_norm": 0.36642298102378845, "learning_rate": 3.3357073780081836e-05, "loss": 0.0251, "step": 50210 }, { "grad_norm": 0.2507147490978241, "learning_rate": 3.333758540704459e-05, "loss": 0.0263, "step": 50220 }, { "grad_norm": 0.23031583428382874, "learning_rate": 3.331809988115827e-05, "loss": 0.0195, "step": 50230 }, { "grad_norm": 0.2863946557044983, "learning_rate": 3.329861720575244e-05, "loss": 0.0299, "step": 50240 }, { "grad_norm": 0.31760627031326294, "learning_rate": 3.3279137384156126e-05, "loss": 0.0215, "step": 50250 }, { "grad_norm": 0.3434286117553711, "learning_rate": 3.3259660419697934e-05, "loss": 0.0296, "step": 50260 }, { "grad_norm": 0.315247505903244, "learning_rate": 3.3240186315705926e-05, "loss": 0.0314, "step": 50270 }, { "grad_norm": 0.28059694170951843, "learning_rate": 3.322071507550769e-05, "loss": 0.0309, "step": 50280 }, { "grad_norm": 0.2943766713142395, "learning_rate": 3.320124670243033e-05, "loss": 0.0261, "step": 50290 }, { "grad_norm": 0.23309649527072906, "learning_rate": 3.318178119980045e-05, "loss": 0.0238, "step": 50300 }, { "grad_norm": 0.26936575770378113, "learning_rate": 3.316231857094421e-05, "loss": 0.0213, "step": 50310 }, { "grad_norm": 0.3229908347129822, "learning_rate": 3.31428588191872e-05, "loss": 0.0248, "step": 50320 }, { "grad_norm": 0.30244940519332886, "learning_rate": 3.312340194785458e-05, "loss": 0.0274, "step": 50330 }, { "grad_norm": 0.2891885042190552, "learning_rate": 3.3103947960270975e-05, "loss": 0.0232, "step": 50340 }, { "grad_norm": 0.3242493271827698, "learning_rate": 3.308449685976058e-05, "loss": 0.022, "step": 50350 }, { "grad_norm": 0.21134421229362488, "learning_rate": 3.3065048649647024e-05, "loss": 0.0258, "step": 50360 }, { "grad_norm": 0.29463183879852295, "learning_rate": 3.304560333325348e-05, "loss": 0.0217, "step": 50370 }, { "grad_norm": 0.31881454586982727, "learning_rate": 3.3026160913902624e-05, "loss": 0.026, "step": 50380 }, { "grad_norm": 0.4639522135257721, "learning_rate": 3.300672139491662e-05, "loss": 0.0321, "step": 50390 }, { "grad_norm": 0.33543771505355835, "learning_rate": 3.298728477961717e-05, "loss": 0.0263, "step": 50400 }, { "grad_norm": 0.21216818690299988, "learning_rate": 3.296785107132544e-05, "loss": 0.031, "step": 50410 }, { "grad_norm": 0.29293128848075867, "learning_rate": 3.2948420273362156e-05, "loss": 0.0214, "step": 50420 }, { "grad_norm": 0.289567232131958, "learning_rate": 3.292899238904747e-05, "loss": 0.0258, "step": 50430 }, { "grad_norm": 0.27880123257637024, "learning_rate": 3.29095674217011e-05, "loss": 0.0235, "step": 50440 }, { "grad_norm": 0.32566627860069275, "learning_rate": 3.289014537464224e-05, "loss": 0.0283, "step": 50450 }, { "grad_norm": 0.24274127185344696, "learning_rate": 3.287072625118955e-05, "loss": 0.0256, "step": 50460 }, { "grad_norm": 0.26816830039024353, "learning_rate": 3.285131005466129e-05, "loss": 0.0198, "step": 50470 }, { "grad_norm": 0.2914634943008423, "learning_rate": 3.2831896788375105e-05, "loss": 0.0224, "step": 50480 }, { "grad_norm": 0.2374742478132248, "learning_rate": 3.281248645564822e-05, "loss": 0.0221, "step": 50490 }, { "grad_norm": 0.21404726803302765, "learning_rate": 3.2793079059797306e-05, "loss": 0.0211, "step": 50500 }, { "grad_norm": 0.33060526847839355, "learning_rate": 3.277367460413859e-05, "loss": 0.025, "step": 50510 }, { "grad_norm": 0.28235507011413574, "learning_rate": 3.275427309198773e-05, "loss": 0.0317, "step": 50520 }, { "grad_norm": 0.211080864071846, "learning_rate": 3.273487452665993e-05, "loss": 0.024, "step": 50530 }, { "grad_norm": 0.5892410278320312, "learning_rate": 3.271547891146986e-05, "loss": 0.026, "step": 50540 }, { "grad_norm": 0.2754521071910858, "learning_rate": 3.269608624973173e-05, "loss": 0.0231, "step": 50550 }, { "grad_norm": 0.4811086654663086, "learning_rate": 3.26766965447592e-05, "loss": 0.0269, "step": 50560 }, { "grad_norm": 0.2865823805332184, "learning_rate": 3.2657309799865424e-05, "loss": 0.0234, "step": 50570 }, { "grad_norm": 0.3058965802192688, "learning_rate": 3.2637926018363084e-05, "loss": 0.0274, "step": 50580 }, { "grad_norm": 0.42972853779792786, "learning_rate": 3.261854520356432e-05, "loss": 0.0269, "step": 50590 }, { "grad_norm": 0.32284998893737793, "learning_rate": 3.2599167358780816e-05, "loss": 0.0214, "step": 50600 }, { "grad_norm": 0.2995269298553467, "learning_rate": 3.2579792487323676e-05, "loss": 0.0227, "step": 50610 }, { "grad_norm": 0.3226497173309326, "learning_rate": 3.256042059250358e-05, "loss": 0.0308, "step": 50620 }, { "grad_norm": 0.5417720079421997, "learning_rate": 3.254105167763062e-05, "loss": 0.0328, "step": 50630 }, { "grad_norm": 0.3986850082874298, "learning_rate": 3.252168574601443e-05, "loss": 0.0299, "step": 50640 }, { "grad_norm": 0.2698827087879181, "learning_rate": 3.2502322800964115e-05, "loss": 0.0209, "step": 50650 }, { "grad_norm": 0.3920242488384247, "learning_rate": 3.248296284578825e-05, "loss": 0.0259, "step": 50660 }, { "grad_norm": 0.23100902140140533, "learning_rate": 3.246360588379497e-05, "loss": 0.0264, "step": 50670 }, { "grad_norm": 0.2873522639274597, "learning_rate": 3.24442519182918e-05, "loss": 0.0299, "step": 50680 }, { "grad_norm": 0.29326769709587097, "learning_rate": 3.2424900952585845e-05, "loss": 0.0251, "step": 50690 }, { "grad_norm": 0.25564903020858765, "learning_rate": 3.240555298998362e-05, "loss": 0.0239, "step": 50700 }, { "grad_norm": 0.3186485171318054, "learning_rate": 3.2386208033791194e-05, "loss": 0.0226, "step": 50710 }, { "grad_norm": 0.2800309360027313, "learning_rate": 3.2366866087314064e-05, "loss": 0.0265, "step": 50720 }, { "grad_norm": 0.31218788027763367, "learning_rate": 3.234752715385727e-05, "loss": 0.0246, "step": 50730 }, { "grad_norm": 0.3528038561344147, "learning_rate": 3.232819123672529e-05, "loss": 0.0319, "step": 50740 }, { "grad_norm": 0.28137069940567017, "learning_rate": 3.2308858339222076e-05, "loss": 0.0288, "step": 50750 }, { "grad_norm": 0.3322960138320923, "learning_rate": 3.2289528464651144e-05, "loss": 0.0296, "step": 50760 }, { "grad_norm": 0.3745540976524353, "learning_rate": 3.22702016163154e-05, "loss": 0.0294, "step": 50770 }, { "grad_norm": 0.30979278683662415, "learning_rate": 3.2250877797517295e-05, "loss": 0.0319, "step": 50780 }, { "grad_norm": 0.18951566517353058, "learning_rate": 3.223155701155872e-05, "loss": 0.0262, "step": 50790 }, { "grad_norm": 0.31685787439346313, "learning_rate": 3.2212239261741105e-05, "loss": 0.0373, "step": 50800 }, { "grad_norm": 0.83212810754776, "learning_rate": 3.219292455136528e-05, "loss": 0.027, "step": 50810 }, { "grad_norm": 0.22453396022319794, "learning_rate": 3.217361288373165e-05, "loss": 0.0214, "step": 50820 }, { "grad_norm": 0.2718587815761566, "learning_rate": 3.215430426214002e-05, "loss": 0.0241, "step": 50830 }, { "grad_norm": 0.1899358034133911, "learning_rate": 3.2134998689889705e-05, "loss": 0.0185, "step": 50840 }, { "grad_norm": 0.3110361695289612, "learning_rate": 3.211569617027952e-05, "loss": 0.0297, "step": 50850 }, { "grad_norm": 0.20090986788272858, "learning_rate": 3.209639670660769e-05, "loss": 0.0233, "step": 50860 }, { "grad_norm": 0.4735032618045807, "learning_rate": 3.207710030217202e-05, "loss": 0.0302, "step": 50870 }, { "grad_norm": 0.40189671516418457, "learning_rate": 3.205780696026972e-05, "loss": 0.0232, "step": 50880 }, { "grad_norm": 0.9672529101371765, "learning_rate": 3.203851668419749e-05, "loss": 0.0234, "step": 50890 }, { "grad_norm": 0.20457297563552856, "learning_rate": 3.201922947725149e-05, "loss": 0.024, "step": 50900 }, { "grad_norm": 0.26900383830070496, "learning_rate": 3.199994534272742e-05, "loss": 0.0232, "step": 50910 }, { "grad_norm": 0.20649047195911407, "learning_rate": 3.1980664283920394e-05, "loss": 0.0275, "step": 50920 }, { "grad_norm": 0.3702187240123749, "learning_rate": 3.196138630412499e-05, "loss": 0.0254, "step": 50930 }, { "grad_norm": 0.3226373493671417, "learning_rate": 3.194211140663532e-05, "loss": 0.0228, "step": 50940 }, { "grad_norm": 0.3482334315776825, "learning_rate": 3.1922839594744914e-05, "loss": 0.0297, "step": 50950 }, { "grad_norm": 0.5151451230049133, "learning_rate": 3.190357087174683e-05, "loss": 0.0332, "step": 50960 }, { "grad_norm": 0.2090667188167572, "learning_rate": 3.188430524093353e-05, "loss": 0.0229, "step": 50970 }, { "grad_norm": 0.36529654264450073, "learning_rate": 3.1865042705597014e-05, "loss": 0.0274, "step": 50980 }, { "grad_norm": 0.47921621799468994, "learning_rate": 3.184578326902868e-05, "loss": 0.0227, "step": 50990 }, { "grad_norm": 0.45738381147384644, "learning_rate": 3.182652693451949e-05, "loss": 0.0318, "step": 51000 }, { "grad_norm": 0.4048961400985718, "learning_rate": 3.180727370535981e-05, "loss": 0.025, "step": 51010 }, { "grad_norm": 0.2733074128627777, "learning_rate": 3.178802358483945e-05, "loss": 0.0245, "step": 51020 }, { "grad_norm": 0.30762797594070435, "learning_rate": 3.17687765762478e-05, "loss": 0.025, "step": 51030 }, { "grad_norm": 0.22030523419380188, "learning_rate": 3.174953268287357e-05, "loss": 0.0332, "step": 51040 }, { "grad_norm": 0.36797505617141724, "learning_rate": 3.173029190800508e-05, "loss": 0.0266, "step": 51050 }, { "grad_norm": 0.3435105085372925, "learning_rate": 3.171105425493e-05, "loss": 0.0307, "step": 51060 }, { "grad_norm": 0.3324095606803894, "learning_rate": 3.169181972693557e-05, "loss": 0.0235, "step": 51070 }, { "grad_norm": 0.31163904070854187, "learning_rate": 3.1672588327308404e-05, "loss": 0.0233, "step": 51080 }, { "grad_norm": 0.35617300868034363, "learning_rate": 3.1653360059334646e-05, "loss": 0.0279, "step": 51090 }, { "grad_norm": 0.5980144739151001, "learning_rate": 3.163413492629985e-05, "loss": 0.0312, "step": 51100 }, { "grad_norm": 0.3388665020465851, "learning_rate": 3.161491293148912e-05, "loss": 0.0264, "step": 51110 }, { "grad_norm": 0.35952991247177124, "learning_rate": 3.1595694078186925e-05, "loss": 0.025, "step": 51120 }, { "grad_norm": 0.3278998136520386, "learning_rate": 3.157647836967724e-05, "loss": 0.0238, "step": 51130 }, { "grad_norm": 0.24868373572826385, "learning_rate": 3.155726580924354e-05, "loss": 0.0258, "step": 51140 }, { "grad_norm": 0.42419105768203735, "learning_rate": 3.1538056400168676e-05, "loss": 0.0363, "step": 51150 }, { "grad_norm": 0.458481103181839, "learning_rate": 3.151885014573506e-05, "loss": 0.0313, "step": 51160 }, { "grad_norm": 0.2567797303199768, "learning_rate": 3.149964704922448e-05, "loss": 0.0386, "step": 51170 }, { "grad_norm": 0.24917387962341309, "learning_rate": 3.148044711391825e-05, "loss": 0.0261, "step": 51180 }, { "grad_norm": 0.6269537210464478, "learning_rate": 3.146125034309707e-05, "loss": 0.0243, "step": 51190 }, { "grad_norm": 0.3240121304988861, "learning_rate": 3.1442056740041195e-05, "loss": 0.0287, "step": 51200 }, { "grad_norm": 0.30182361602783203, "learning_rate": 3.142286630803026e-05, "loss": 0.0299, "step": 51210 }, { "grad_norm": 0.3347014784812927, "learning_rate": 3.140367905034337e-05, "loss": 0.0258, "step": 51220 }, { "grad_norm": 0.2547009587287903, "learning_rate": 3.138449497025914e-05, "loss": 0.0269, "step": 51230 }, { "grad_norm": 0.3908693194389343, "learning_rate": 3.136531407105557e-05, "loss": 0.0262, "step": 51240 }, { "grad_norm": 0.2753146290779114, "learning_rate": 3.1346136356010184e-05, "loss": 0.0301, "step": 51250 }, { "grad_norm": 0.3066171407699585, "learning_rate": 3.132696182839988e-05, "loss": 0.0246, "step": 51260 }, { "grad_norm": 0.3984329104423523, "learning_rate": 3.130779049150111e-05, "loss": 0.0327, "step": 51270 }, { "grad_norm": 0.4372691214084625, "learning_rate": 3.128862234858971e-05, "loss": 0.026, "step": 51280 }, { "grad_norm": 0.2564268112182617, "learning_rate": 3.1269457402941005e-05, "loss": 0.0242, "step": 51290 }, { "grad_norm": 0.44173383712768555, "learning_rate": 3.125029565782974e-05, "loss": 0.0287, "step": 51300 }, { "grad_norm": 0.44308754801750183, "learning_rate": 3.123113711653013e-05, "loss": 0.0323, "step": 51310 }, { "grad_norm": 0.21498793363571167, "learning_rate": 3.121198178231587e-05, "loss": 0.0297, "step": 51320 }, { "grad_norm": 0.3187587261199951, "learning_rate": 3.119282965846006e-05, "loss": 0.0261, "step": 51330 }, { "grad_norm": 0.3302195072174072, "learning_rate": 3.1173680748235303e-05, "loss": 0.0274, "step": 51340 }, { "grad_norm": 0.32790112495422363, "learning_rate": 3.115453505491358e-05, "loss": 0.0268, "step": 51350 }, { "grad_norm": 0.1307666152715683, "learning_rate": 3.1135392581766404e-05, "loss": 0.0258, "step": 51360 }, { "grad_norm": 0.323391854763031, "learning_rate": 3.111625333206467e-05, "loss": 0.0243, "step": 51370 }, { "grad_norm": 0.2462749481201172, "learning_rate": 3.109711730907881e-05, "loss": 0.0205, "step": 51380 }, { "grad_norm": 0.2730616629123688, "learning_rate": 3.107798451607859e-05, "loss": 0.02, "step": 51390 }, { "grad_norm": 0.4605347812175751, "learning_rate": 3.105885495633328e-05, "loss": 0.0261, "step": 51400 }, { "grad_norm": 0.25904884934425354, "learning_rate": 3.103972863311163e-05, "loss": 0.0241, "step": 51410 }, { "grad_norm": 0.36110252141952515, "learning_rate": 3.102060554968178e-05, "loss": 0.0286, "step": 51420 }, { "grad_norm": 0.24559162557125092, "learning_rate": 3.100148570931137e-05, "loss": 0.025, "step": 51430 }, { "grad_norm": 0.28420358896255493, "learning_rate": 3.098236911526744e-05, "loss": 0.0267, "step": 51440 }, { "grad_norm": 0.2721305191516876, "learning_rate": 3.0963255770816495e-05, "loss": 0.022, "step": 51450 }, { "grad_norm": 0.3095899522304535, "learning_rate": 3.094414567922447e-05, "loss": 0.0236, "step": 51460 }, { "grad_norm": 0.30281203985214233, "learning_rate": 3.092503884375678e-05, "loss": 0.0268, "step": 51470 }, { "grad_norm": 0.39397692680358887, "learning_rate": 3.0905935267678255e-05, "loss": 0.028, "step": 51480 }, { "grad_norm": 0.39357611536979675, "learning_rate": 3.0886834954253166e-05, "loss": 0.0257, "step": 51490 }, { "grad_norm": 0.42051130533218384, "learning_rate": 3.086773790674524e-05, "loss": 0.0266, "step": 51500 }, { "grad_norm": 0.4961516559123993, "learning_rate": 3.0848644128417617e-05, "loss": 0.0299, "step": 51510 }, { "grad_norm": 0.3602179288864136, "learning_rate": 3.082955362253294e-05, "loss": 0.0315, "step": 51520 }, { "grad_norm": 0.5169492363929749, "learning_rate": 3.081046639235323e-05, "loss": 0.0335, "step": 51530 }, { "grad_norm": 0.25537195801734924, "learning_rate": 3.0791382441139986e-05, "loss": 0.023, "step": 51540 }, { "grad_norm": 0.5837985873222351, "learning_rate": 3.077230177215411e-05, "loss": 0.0305, "step": 51550 }, { "grad_norm": 0.2059871107339859, "learning_rate": 3.0753224388656e-05, "loss": 0.0221, "step": 51560 }, { "grad_norm": 0.31810951232910156, "learning_rate": 3.073415029390544e-05, "loss": 0.0274, "step": 51570 }, { "grad_norm": 0.33599916100502014, "learning_rate": 3.071507949116166e-05, "loss": 0.0247, "step": 51580 }, { "grad_norm": 0.21872110664844513, "learning_rate": 3.069601198368337e-05, "loss": 0.0163, "step": 51590 }, { "grad_norm": 0.42173412442207336, "learning_rate": 3.067694777472864e-05, "loss": 0.0199, "step": 51600 }, { "grad_norm": 0.3480243682861328, "learning_rate": 3.065788686755508e-05, "loss": 0.0262, "step": 51610 }, { "grad_norm": 0.20382705330848694, "learning_rate": 3.063882926541961e-05, "loss": 0.0217, "step": 51620 }, { "grad_norm": 0.2831219434738159, "learning_rate": 3.061977497157872e-05, "loss": 0.0256, "step": 51630 }, { "grad_norm": 0.2966301143169403, "learning_rate": 3.0600723989288224e-05, "loss": 0.0224, "step": 51640 }, { "grad_norm": 0.32087811827659607, "learning_rate": 3.058167632180343e-05, "loss": 0.0217, "step": 51650 }, { "grad_norm": 0.2959129810333252, "learning_rate": 3.0562631972379074e-05, "loss": 0.0218, "step": 51660 }, { "grad_norm": 0.27700772881507874, "learning_rate": 3.0543590944269276e-05, "loss": 0.0291, "step": 51670 }, { "grad_norm": 0.34619131684303284, "learning_rate": 3.052455324072766e-05, "loss": 0.0355, "step": 51680 }, { "grad_norm": 0.26193830370903015, "learning_rate": 3.0505518865007244e-05, "loss": 0.0238, "step": 51690 }, { "grad_norm": 0.3573153614997864, "learning_rate": 3.048648782036048e-05, "loss": 0.0294, "step": 51700 }, { "grad_norm": 0.23130853474140167, "learning_rate": 3.0467460110039236e-05, "loss": 0.022, "step": 51710 }, { "grad_norm": 0.27633851766586304, "learning_rate": 3.0448435737294866e-05, "loss": 0.028, "step": 51720 }, { "grad_norm": 0.3598272204399109, "learning_rate": 3.042941470537808e-05, "loss": 0.0238, "step": 51730 }, { "grad_norm": 0.23230022192001343, "learning_rate": 3.0410397017539072e-05, "loss": 0.0235, "step": 51740 }, { "grad_norm": 0.4572983682155609, "learning_rate": 3.039138267702742e-05, "loss": 0.0213, "step": 51750 }, { "grad_norm": 0.20211225748062134, "learning_rate": 3.0372371687092193e-05, "loss": 0.0213, "step": 51760 }, { "grad_norm": 0.2703433036804199, "learning_rate": 3.0353364050981823e-05, "loss": 0.0276, "step": 51770 }, { "grad_norm": 0.2254716455936432, "learning_rate": 3.033435977194418e-05, "loss": 0.0224, "step": 51780 }, { "grad_norm": 0.37998732924461365, "learning_rate": 3.0315358853226618e-05, "loss": 0.0258, "step": 51790 }, { "grad_norm": 0.4039699137210846, "learning_rate": 3.0296361298075844e-05, "loss": 0.0221, "step": 51800 }, { "grad_norm": 0.3206237554550171, "learning_rate": 3.027736710973803e-05, "loss": 0.0294, "step": 51810 }, { "grad_norm": 0.20163436233997345, "learning_rate": 3.025837629145875e-05, "loss": 0.0233, "step": 51820 }, { "grad_norm": 0.2791247069835663, "learning_rate": 3.0239388846483048e-05, "loss": 0.0251, "step": 51830 }, { "grad_norm": 0.24992766976356506, "learning_rate": 3.022040477805532e-05, "loss": 0.0192, "step": 51840 }, { "grad_norm": 0.3642243444919586, "learning_rate": 3.020142408941946e-05, "loss": 0.0224, "step": 51850 }, { "grad_norm": 0.30369725823402405, "learning_rate": 3.018244678381873e-05, "loss": 0.0307, "step": 51860 }, { "grad_norm": 0.2663653790950775, "learning_rate": 3.0163472864495812e-05, "loss": 0.0233, "step": 51870 }, { "grad_norm": 0.29820579290390015, "learning_rate": 3.014450233469287e-05, "loss": 0.0306, "step": 51880 }, { "grad_norm": 0.2852649986743927, "learning_rate": 3.0125535197651422e-05, "loss": 0.0314, "step": 51890 }, { "grad_norm": 0.2874735891819, "learning_rate": 3.0106571456612448e-05, "loss": 0.0233, "step": 51900 }, { "grad_norm": 0.37392768263816833, "learning_rate": 3.0087611114816305e-05, "loss": 0.0346, "step": 51910 }, { "grad_norm": 0.6055276989936829, "learning_rate": 3.006865417550284e-05, "loss": 0.0376, "step": 51920 }, { "grad_norm": 0.31513717770576477, "learning_rate": 3.0049700641911242e-05, "loss": 0.0305, "step": 51930 }, { "grad_norm": 0.3246115744113922, "learning_rate": 3.0030750517280175e-05, "loss": 0.0255, "step": 51940 }, { "grad_norm": 0.331459641456604, "learning_rate": 3.0011803804847682e-05, "loss": 0.0344, "step": 51950 }, { "grad_norm": 0.40476056933403015, "learning_rate": 2.9992860507851224e-05, "loss": 0.0337, "step": 51960 }, { "grad_norm": 0.32946932315826416, "learning_rate": 2.9973920629527723e-05, "loss": 0.0259, "step": 51970 }, { "grad_norm": 0.17158393561840057, "learning_rate": 2.9954984173113453e-05, "loss": 0.0218, "step": 51980 }, { "grad_norm": 0.3243962824344635, "learning_rate": 2.993605114184418e-05, "loss": 0.0269, "step": 51990 }, { "grad_norm": 0.5033209323883057, "learning_rate": 2.9917121538955005e-05, "loss": 0.0293, "step": 52000 }, { "grad_norm": 0.2870010733604431, "learning_rate": 2.98981953676805e-05, "loss": 0.0336, "step": 52010 }, { "grad_norm": 0.2535390257835388, "learning_rate": 2.9879272631254594e-05, "loss": 0.0276, "step": 52020 }, { "grad_norm": 0.16167178750038147, "learning_rate": 2.986035333291073e-05, "loss": 0.0257, "step": 52030 }, { "grad_norm": 0.1454952210187912, "learning_rate": 2.9841437475881646e-05, "loss": 0.0342, "step": 52040 }, { "grad_norm": 0.4481346011161804, "learning_rate": 2.9822525063399555e-05, "loss": 0.0261, "step": 52050 }, { "grad_norm": 0.25729793310165405, "learning_rate": 2.9803616098696087e-05, "loss": 0.0306, "step": 52060 }, { "grad_norm": 0.2532678246498108, "learning_rate": 2.9784710585002228e-05, "loss": 0.0278, "step": 52070 }, { "grad_norm": 0.3578082025051117, "learning_rate": 2.9765808525548467e-05, "loss": 0.0261, "step": 52080 }, { "grad_norm": 0.781118631362915, "learning_rate": 2.974690992356461e-05, "loss": 0.0257, "step": 52090 }, { "grad_norm": 0.2840929627418518, "learning_rate": 2.9728014782279934e-05, "loss": 0.0329, "step": 52100 }, { "grad_norm": 0.3221554458141327, "learning_rate": 2.970912310492307e-05, "loss": 0.0287, "step": 52110 }, { "grad_norm": 0.22002339363098145, "learning_rate": 2.9690234894722134e-05, "loss": 0.0281, "step": 52120 }, { "grad_norm": 0.22922591865062714, "learning_rate": 2.9671350154904577e-05, "loss": 0.0306, "step": 52130 }, { "grad_norm": 0.3593374192714691, "learning_rate": 2.9652468888697282e-05, "loss": 0.0298, "step": 52140 }, { "grad_norm": 0.2191937416791916, "learning_rate": 2.9633591099326562e-05, "loss": 0.024, "step": 52150 }, { "grad_norm": 0.22788693010807037, "learning_rate": 2.9614716790018072e-05, "loss": 0.0248, "step": 52160 }, { "grad_norm": 0.2464110404253006, "learning_rate": 2.959584596399697e-05, "loss": 0.0196, "step": 52170 }, { "grad_norm": 0.28401243686676025, "learning_rate": 2.9576978624487717e-05, "loss": 0.0228, "step": 52180 }, { "grad_norm": 0.1788586527109146, "learning_rate": 2.9558114774714263e-05, "loss": 0.024, "step": 52190 }, { "grad_norm": 0.21741512417793274, "learning_rate": 2.9539254417899897e-05, "loss": 0.0355, "step": 52200 }, { "grad_norm": 0.27676519751548767, "learning_rate": 2.9520397557267365e-05, "loss": 0.022, "step": 52210 }, { "grad_norm": 0.29850366711616516, "learning_rate": 2.9501544196038765e-05, "loss": 0.0277, "step": 52220 }, { "grad_norm": 0.685477077960968, "learning_rate": 2.9482694337435614e-05, "loss": 0.0368, "step": 52230 }, { "grad_norm": 0.39603444933891296, "learning_rate": 2.946384798467887e-05, "loss": 0.0303, "step": 52240 }, { "grad_norm": 0.23719704151153564, "learning_rate": 2.9445005140988824e-05, "loss": 0.019, "step": 52250 }, { "grad_norm": 0.3424656093120575, "learning_rate": 2.942616580958524e-05, "loss": 0.0254, "step": 52260 }, { "grad_norm": 0.2469351440668106, "learning_rate": 2.9407329993687193e-05, "loss": 0.0243, "step": 52270 }, { "grad_norm": 0.3575243651866913, "learning_rate": 2.938849769651326e-05, "loss": 0.0211, "step": 52280 }, { "grad_norm": 0.33849862217903137, "learning_rate": 2.9369668921281345e-05, "loss": 0.0213, "step": 52290 }, { "grad_norm": 0.18702909350395203, "learning_rate": 2.9350843671208773e-05, "loss": 0.0244, "step": 52300 }, { "grad_norm": 0.3383807837963104, "learning_rate": 2.9332021949512245e-05, "loss": 0.0228, "step": 52310 }, { "grad_norm": 0.2886316478252411, "learning_rate": 2.9313203759407908e-05, "loss": 0.022, "step": 52320 }, { "grad_norm": 0.2817305624485016, "learning_rate": 2.929438910411127e-05, "loss": 0.037, "step": 52330 }, { "grad_norm": 0.579754650592804, "learning_rate": 2.927557798683722e-05, "loss": 0.0327, "step": 52340 }, { "grad_norm": 0.7014976143836975, "learning_rate": 2.9256770410800095e-05, "loss": 0.0229, "step": 52350 }, { "grad_norm": 0.163089781999588, "learning_rate": 2.9237966379213554e-05, "loss": 0.0211, "step": 52360 }, { "grad_norm": 0.3644043803215027, "learning_rate": 2.9219165895290736e-05, "loss": 0.0292, "step": 52370 }, { "grad_norm": 0.3144599497318268, "learning_rate": 2.9200368962244083e-05, "loss": 0.0288, "step": 52380 }, { "grad_norm": 0.26837316155433655, "learning_rate": 2.9181575583285536e-05, "loss": 0.0249, "step": 52390 }, { "grad_norm": 0.2736000120639801, "learning_rate": 2.916278576162632e-05, "loss": 0.0245, "step": 52400 }, { "grad_norm": 0.22532618045806885, "learning_rate": 2.9143999500477116e-05, "loss": 0.0227, "step": 52410 }, { "grad_norm": 0.2747610807418823, "learning_rate": 2.9125216803048004e-05, "loss": 0.0273, "step": 52420 }, { "grad_norm": 0.2936910092830658, "learning_rate": 2.9106437672548403e-05, "loss": 0.0224, "step": 52430 }, { "grad_norm": 0.2751618027687073, "learning_rate": 2.9087662112187154e-05, "loss": 0.0301, "step": 52440 }, { "grad_norm": 0.31671857833862305, "learning_rate": 2.9068890125172498e-05, "loss": 0.0232, "step": 52450 }, { "grad_norm": 0.4322492182254791, "learning_rate": 2.9050121714712058e-05, "loss": 0.0306, "step": 52460 }, { "grad_norm": 0.30543485283851624, "learning_rate": 2.9031356884012835e-05, "loss": 0.0241, "step": 52470 }, { "grad_norm": 0.29255053400993347, "learning_rate": 2.9012595636281247e-05, "loss": 0.0282, "step": 52480 }, { "grad_norm": 0.4140423536300659, "learning_rate": 2.8993837974723016e-05, "loss": 0.0391, "step": 52490 }, { "grad_norm": 0.24707894027233124, "learning_rate": 2.8975083902543394e-05, "loss": 0.0252, "step": 52500 }, { "grad_norm": 0.21723929047584534, "learning_rate": 2.895633342294688e-05, "loss": 0.0192, "step": 52510 }, { "grad_norm": 0.2620295286178589, "learning_rate": 2.893758653913744e-05, "loss": 0.026, "step": 52520 }, { "grad_norm": 0.2801995873451233, "learning_rate": 2.8918843254318423e-05, "loss": 0.0228, "step": 52530 }, { "grad_norm": 0.3735124468803406, "learning_rate": 2.8900103571692483e-05, "loss": 0.0287, "step": 52540 }, { "grad_norm": 0.25837868452072144, "learning_rate": 2.8881367494461796e-05, "loss": 0.0248, "step": 52550 }, { "grad_norm": 0.4143223166465759, "learning_rate": 2.886263502582779e-05, "loss": 0.0283, "step": 52560 }, { "grad_norm": 0.3329590857028961, "learning_rate": 2.8843906168991353e-05, "loss": 0.0207, "step": 52570 }, { "grad_norm": 0.2410111278295517, "learning_rate": 2.8825180927152728e-05, "loss": 0.0254, "step": 52580 }, { "grad_norm": 0.24278776347637177, "learning_rate": 2.8806459303511547e-05, "loss": 0.0226, "step": 52590 }, { "grad_norm": 0.2685908079147339, "learning_rate": 2.8787741301266852e-05, "loss": 0.0252, "step": 52600 }, { "grad_norm": 0.3355857729911804, "learning_rate": 2.876902692361697e-05, "loss": 0.0274, "step": 52610 }, { "grad_norm": 0.2884525656700134, "learning_rate": 2.875031617375976e-05, "loss": 0.021, "step": 52620 }, { "grad_norm": 0.23980645835399628, "learning_rate": 2.8731609054892318e-05, "loss": 0.0195, "step": 52630 }, { "grad_norm": 0.26127275824546814, "learning_rate": 2.871290557021119e-05, "loss": 0.0266, "step": 52640 }, { "grad_norm": 0.22123804688453674, "learning_rate": 2.8694205722912298e-05, "loss": 0.0225, "step": 52650 }, { "grad_norm": 0.22238661348819733, "learning_rate": 2.8675509516190936e-05, "loss": 0.0294, "step": 52660 }, { "grad_norm": 0.2763286828994751, "learning_rate": 2.8656816953241772e-05, "loss": 0.032, "step": 52670 }, { "grad_norm": 0.3156989514827728, "learning_rate": 2.863812803725887e-05, "loss": 0.0248, "step": 52680 }, { "grad_norm": 0.23623813688755035, "learning_rate": 2.8619442771435623e-05, "loss": 0.0242, "step": 52690 }, { "grad_norm": 0.27849629521369934, "learning_rate": 2.860076115896484e-05, "loss": 0.0351, "step": 52700 }, { "grad_norm": 0.2788524031639099, "learning_rate": 2.8582083203038712e-05, "loss": 0.023, "step": 52710 }, { "grad_norm": 0.2508413791656494, "learning_rate": 2.8563408906848778e-05, "loss": 0.025, "step": 52720 }, { "grad_norm": 0.2569248080253601, "learning_rate": 2.8544738273585993e-05, "loss": 0.023, "step": 52730 }, { "grad_norm": 0.2661169171333313, "learning_rate": 2.8526071306440595e-05, "loss": 0.0214, "step": 52740 }, { "grad_norm": 0.24416999518871307, "learning_rate": 2.850740800860233e-05, "loss": 0.0221, "step": 52750 }, { "grad_norm": 0.3786260485649109, "learning_rate": 2.8488748383260178e-05, "loss": 0.0207, "step": 52760 }, { "grad_norm": 0.36202090978622437, "learning_rate": 2.8470092433602635e-05, "loss": 0.0365, "step": 52770 }, { "grad_norm": 0.2748601734638214, "learning_rate": 2.845144016281745e-05, "loss": 0.0223, "step": 52780 }, { "grad_norm": 0.22725802659988403, "learning_rate": 2.8432791574091744e-05, "loss": 0.0403, "step": 52790 }, { "grad_norm": 0.43767866492271423, "learning_rate": 2.8414146670612134e-05, "loss": 0.0306, "step": 52800 }, { "grad_norm": 0.23631714284420013, "learning_rate": 2.8395505455564446e-05, "loss": 0.0234, "step": 52810 }, { "grad_norm": 0.31819605827331543, "learning_rate": 2.837686793213403e-05, "loss": 0.0267, "step": 52820 }, { "grad_norm": 0.2649964988231659, "learning_rate": 2.8358234103505477e-05, "loss": 0.0275, "step": 52830 }, { "grad_norm": 0.17775124311447144, "learning_rate": 2.8339603972862806e-05, "loss": 0.0222, "step": 52840 }, { "grad_norm": 0.23271894454956055, "learning_rate": 2.83209775433894e-05, "loss": 0.0265, "step": 52850 }, { "grad_norm": 0.32652273774147034, "learning_rate": 2.8302354818268013e-05, "loss": 0.0235, "step": 52860 }, { "grad_norm": 0.27402377128601074, "learning_rate": 2.8283735800680754e-05, "loss": 0.027, "step": 52870 }, { "grad_norm": 0.2634587585926056, "learning_rate": 2.826512049380913e-05, "loss": 0.0249, "step": 52880 }, { "grad_norm": 0.3375643789768219, "learning_rate": 2.8246508900833935e-05, "loss": 0.0294, "step": 52890 }, { "grad_norm": 0.30042240023612976, "learning_rate": 2.822790102493541e-05, "loss": 0.0237, "step": 52900 }, { "grad_norm": 0.2843535244464874, "learning_rate": 2.820929686929314e-05, "loss": 0.0221, "step": 52910 }, { "grad_norm": 0.26112866401672363, "learning_rate": 2.819069643708605e-05, "loss": 0.0267, "step": 52920 }, { "grad_norm": 0.43632224202156067, "learning_rate": 2.8172099731492484e-05, "loss": 0.0241, "step": 52930 }, { "grad_norm": 0.36301228404045105, "learning_rate": 2.815350675569003e-05, "loss": 0.0331, "step": 52940 }, { "grad_norm": 0.2576262950897217, "learning_rate": 2.8134917512855807e-05, "loss": 0.0314, "step": 52950 }, { "grad_norm": 0.2791222631931305, "learning_rate": 2.8116332006166134e-05, "loss": 0.0159, "step": 52960 }, { "grad_norm": 0.5415496826171875, "learning_rate": 2.809775023879685e-05, "loss": 0.036, "step": 52970 }, { "grad_norm": 0.26344776153564453, "learning_rate": 2.807917221392299e-05, "loss": 0.0239, "step": 52980 }, { "grad_norm": 0.26813432574272156, "learning_rate": 2.8060597934719067e-05, "loss": 0.0219, "step": 52990 }, { "grad_norm": 0.3548946678638458, "learning_rate": 2.8042027404358927e-05, "loss": 0.0205, "step": 53000 }, { "grad_norm": 0.32274454832077026, "learning_rate": 2.802346062601572e-05, "loss": 0.0241, "step": 53010 }, { "grad_norm": 0.5872946381568909, "learning_rate": 2.800489760286207e-05, "loss": 0.0357, "step": 53020 }, { "grad_norm": 0.29974842071533203, "learning_rate": 2.7986338338069834e-05, "loss": 0.0259, "step": 53030 }, { "grad_norm": 0.3936728537082672, "learning_rate": 2.7967782834810297e-05, "loss": 0.0293, "step": 53040 }, { "grad_norm": 0.5744096636772156, "learning_rate": 2.794923109625409e-05, "loss": 0.0279, "step": 53050 }, { "grad_norm": 0.2652173340320587, "learning_rate": 2.7930683125571205e-05, "loss": 0.0245, "step": 53060 }, { "grad_norm": 0.25915560126304626, "learning_rate": 2.7912138925930997e-05, "loss": 0.0252, "step": 53070 }, { "grad_norm": 0.20127664506435394, "learning_rate": 2.7893598500502117e-05, "loss": 0.0435, "step": 53080 }, { "grad_norm": 0.3718765079975128, "learning_rate": 2.7875061852452644e-05, "loss": 0.0274, "step": 53090 }, { "grad_norm": 0.5991924405097961, "learning_rate": 2.7856528984949982e-05, "loss": 0.0352, "step": 53100 }, { "grad_norm": 0.307047039270401, "learning_rate": 2.7837999901160888e-05, "loss": 0.0222, "step": 53110 }, { "grad_norm": 0.44630691409111023, "learning_rate": 2.7819474604251484e-05, "loss": 0.026, "step": 53120 }, { "grad_norm": 0.37243834137916565, "learning_rate": 2.780095309738725e-05, "loss": 0.0264, "step": 53130 }, { "grad_norm": 0.2892020344734192, "learning_rate": 2.778243538373294e-05, "loss": 0.0217, "step": 53140 }, { "grad_norm": 0.14220814406871796, "learning_rate": 2.7763921466452826e-05, "loss": 0.0227, "step": 53150 }, { "grad_norm": 0.28570041060447693, "learning_rate": 2.7745411348710336e-05, "loss": 0.0323, "step": 53160 }, { "grad_norm": 0.38153645396232605, "learning_rate": 2.7726905033668395e-05, "loss": 0.0252, "step": 53170 }, { "grad_norm": 0.30639639496803284, "learning_rate": 2.7708402524489214e-05, "loss": 0.0285, "step": 53180 }, { "grad_norm": 0.3054954707622528, "learning_rate": 2.7689903824334364e-05, "loss": 0.0311, "step": 53190 }, { "grad_norm": 0.3172019124031067, "learning_rate": 2.7671408936364785e-05, "loss": 0.025, "step": 53200 }, { "grad_norm": 0.31456419825553894, "learning_rate": 2.765291786374069e-05, "loss": 0.0235, "step": 53210 }, { "grad_norm": 0.2836350202560425, "learning_rate": 2.7634430609621775e-05, "loss": 0.021, "step": 53220 }, { "grad_norm": 0.3478929102420807, "learning_rate": 2.7615947177166956e-05, "loss": 0.0242, "step": 53230 }, { "grad_norm": 0.18989717960357666, "learning_rate": 2.7597467569534553e-05, "loss": 0.0197, "step": 53240 }, { "grad_norm": 0.2928919494152069, "learning_rate": 2.757899178988226e-05, "loss": 0.0238, "step": 53250 }, { "grad_norm": 0.30214688181877136, "learning_rate": 2.7560519841367005e-05, "loss": 0.0245, "step": 53260 }, { "grad_norm": 2.0576348304748535, "learning_rate": 2.7542051727145237e-05, "loss": 0.0207, "step": 53270 }, { "grad_norm": 0.30671370029449463, "learning_rate": 2.7523587450372578e-05, "loss": 0.0306, "step": 53280 }, { "grad_norm": 0.5145731568336487, "learning_rate": 2.750512701420409e-05, "loss": 0.0286, "step": 53290 }, { "grad_norm": 0.46394577622413635, "learning_rate": 2.7486670421794158e-05, "loss": 0.0251, "step": 53300 }, { "grad_norm": 0.20215055346488953, "learning_rate": 2.7468217676296515e-05, "loss": 0.0301, "step": 53310 }, { "grad_norm": 0.296084463596344, "learning_rate": 2.7449768780864226e-05, "loss": 0.0255, "step": 53320 }, { "grad_norm": 0.2803071439266205, "learning_rate": 2.7431323738649724e-05, "loss": 0.0205, "step": 53330 }, { "grad_norm": 0.4495072364807129, "learning_rate": 2.7412882552804713e-05, "loss": 0.0314, "step": 53340 }, { "grad_norm": 0.1548687219619751, "learning_rate": 2.739444522648032e-05, "loss": 0.0168, "step": 53350 }, { "grad_norm": 0.18049892783164978, "learning_rate": 2.737601176282697e-05, "loss": 0.0188, "step": 53360 }, { "grad_norm": 0.34805557131767273, "learning_rate": 2.735758216499445e-05, "loss": 0.0259, "step": 53370 }, { "grad_norm": 0.16811944544315338, "learning_rate": 2.7339156436131864e-05, "loss": 0.0255, "step": 53380 }, { "grad_norm": 0.22335216403007507, "learning_rate": 2.7320734579387663e-05, "loss": 0.0217, "step": 53390 }, { "grad_norm": 0.45192819833755493, "learning_rate": 2.730231659790966e-05, "loss": 0.0195, "step": 53400 }, { "grad_norm": 0.30212029814720154, "learning_rate": 2.728390249484494e-05, "loss": 0.0286, "step": 53410 }, { "grad_norm": 0.36962753534317017, "learning_rate": 2.7265492273340032e-05, "loss": 0.0254, "step": 53420 }, { "grad_norm": 0.3485543727874756, "learning_rate": 2.7247085936540678e-05, "loss": 0.0299, "step": 53430 }, { "grad_norm": 0.43768197298049927, "learning_rate": 2.722868348759205e-05, "loss": 0.0269, "step": 53440 }, { "grad_norm": 0.2152346670627594, "learning_rate": 2.7210284929638635e-05, "loss": 0.0274, "step": 53450 }, { "grad_norm": 0.3255520164966583, "learning_rate": 2.7191890265824183e-05, "loss": 0.022, "step": 53460 }, { "grad_norm": 0.16709452867507935, "learning_rate": 2.7173499499291926e-05, "loss": 0.0168, "step": 53470 }, { "grad_norm": 0.37385591864585876, "learning_rate": 2.7155112633184277e-05, "loss": 0.0258, "step": 53480 }, { "grad_norm": 0.3541612923145294, "learning_rate": 2.713672967064307e-05, "loss": 0.0281, "step": 53490 }, { "grad_norm": 0.5790172219276428, "learning_rate": 2.711835061480945e-05, "loss": 0.0221, "step": 53500 }, { "grad_norm": 0.30375635623931885, "learning_rate": 2.7099975468823896e-05, "loss": 0.0282, "step": 53510 }, { "grad_norm": 0.23359976708889008, "learning_rate": 2.708160423582622e-05, "loss": 0.0218, "step": 53520 }, { "grad_norm": 0.49379774928092957, "learning_rate": 2.706323691895557e-05, "loss": 0.0312, "step": 53530 }, { "grad_norm": 0.283949077129364, "learning_rate": 2.70448735213504e-05, "loss": 0.0222, "step": 53540 }, { "grad_norm": 0.27952319383621216, "learning_rate": 2.702651404614852e-05, "loss": 0.023, "step": 53550 }, { "grad_norm": 0.5772614479064941, "learning_rate": 2.7008158496487056e-05, "loss": 0.0269, "step": 53560 }, { "grad_norm": 0.36468034982681274, "learning_rate": 2.6989806875502487e-05, "loss": 0.0188, "step": 53570 }, { "grad_norm": 0.28525182604789734, "learning_rate": 2.6971459186330584e-05, "loss": 0.0213, "step": 53580 }, { "grad_norm": 0.37800973653793335, "learning_rate": 2.695311543210648e-05, "loss": 0.0189, "step": 53590 }, { "grad_norm": 0.45690611004829407, "learning_rate": 2.693477561596463e-05, "loss": 0.0191, "step": 53600 }, { "grad_norm": 0.30672264099121094, "learning_rate": 2.6916439741038756e-05, "loss": 0.0248, "step": 53610 }, { "grad_norm": 0.3280640244483948, "learning_rate": 2.689810781046203e-05, "loss": 0.0247, "step": 53620 }, { "grad_norm": 0.40851864218711853, "learning_rate": 2.6879779827366823e-05, "loss": 0.025, "step": 53630 }, { "grad_norm": 0.3450163006782532, "learning_rate": 2.6861455794884904e-05, "loss": 0.0237, "step": 53640 }, { "grad_norm": 0.2522551119327545, "learning_rate": 2.6843135716147373e-05, "loss": 0.021, "step": 53650 }, { "grad_norm": 0.5005213618278503, "learning_rate": 2.6824819594284556e-05, "loss": 0.0249, "step": 53660 }, { "grad_norm": 0.20280355215072632, "learning_rate": 2.6806507432426275e-05, "loss": 0.0275, "step": 53670 }, { "grad_norm": 0.26097020506858826, "learning_rate": 2.6788199233701512e-05, "loss": 0.0273, "step": 53680 }, { "grad_norm": 0.21130885183811188, "learning_rate": 2.6769895001238652e-05, "loss": 0.0294, "step": 53690 }, { "grad_norm": 0.28593045473098755, "learning_rate": 2.67515947381654e-05, "loss": 0.0211, "step": 53700 }, { "grad_norm": 0.4222937524318695, "learning_rate": 2.6733298447608745e-05, "loss": 0.0205, "step": 53710 }, { "grad_norm": 0.2616530954837799, "learning_rate": 2.6715006132695074e-05, "loss": 0.0279, "step": 53720 }, { "grad_norm": 0.2836628258228302, "learning_rate": 2.6696717796549985e-05, "loss": 0.0303, "step": 53730 }, { "grad_norm": 0.239694744348526, "learning_rate": 2.667843344229848e-05, "loss": 0.0185, "step": 53740 }, { "grad_norm": 0.4816327691078186, "learning_rate": 2.6660153073064846e-05, "loss": 0.0257, "step": 53750 }, { "grad_norm": 0.3031069040298462, "learning_rate": 2.664187669197271e-05, "loss": 0.0264, "step": 53760 }, { "grad_norm": 0.4826335608959198, "learning_rate": 2.6623604302145005e-05, "loss": 0.0288, "step": 53770 }, { "grad_norm": 0.4296121895313263, "learning_rate": 2.6605335906703975e-05, "loss": 0.026, "step": 53780 }, { "grad_norm": 0.2560221552848816, "learning_rate": 2.65870715087712e-05, "loss": 0.0291, "step": 53790 }, { "grad_norm": 0.2737094759941101, "learning_rate": 2.6568811111467573e-05, "loss": 0.0247, "step": 53800 }, { "grad_norm": 0.23528151214122772, "learning_rate": 2.6550554717913258e-05, "loss": 0.0238, "step": 53810 }, { "grad_norm": 0.36033010482788086, "learning_rate": 2.6532302331227805e-05, "loss": 0.027, "step": 53820 }, { "grad_norm": 0.4145735204219818, "learning_rate": 2.651405395453004e-05, "loss": 0.0211, "step": 53830 }, { "grad_norm": 0.33191022276878357, "learning_rate": 2.6495809590938115e-05, "loss": 0.0298, "step": 53840 }, { "grad_norm": 0.2344287484884262, "learning_rate": 2.64775692435695e-05, "loss": 0.0296, "step": 53850 }, { "grad_norm": 0.22728289663791656, "learning_rate": 2.6459332915540928e-05, "loss": 0.0233, "step": 53860 }, { "grad_norm": 0.22035282850265503, "learning_rate": 2.644110060996856e-05, "loss": 0.0368, "step": 53870 }, { "grad_norm": 0.4322686195373535, "learning_rate": 2.642287232996774e-05, "loss": 0.022, "step": 53880 }, { "grad_norm": 0.2381691187620163, "learning_rate": 2.6404648078653205e-05, "loss": 0.0196, "step": 53890 }, { "grad_norm": 0.2636928856372833, "learning_rate": 2.6386427859139002e-05, "loss": 0.023, "step": 53900 }, { "grad_norm": 0.3857135474681854, "learning_rate": 2.63682116745384e-05, "loss": 0.0223, "step": 53910 }, { "grad_norm": 0.3731079399585724, "learning_rate": 2.6349999527964138e-05, "loss": 0.0208, "step": 53920 }, { "grad_norm": 0.24082300066947937, "learning_rate": 2.63317914225281e-05, "loss": 0.0349, "step": 53930 }, { "grad_norm": 0.48523104190826416, "learning_rate": 2.6313587361341585e-05, "loss": 0.0305, "step": 53940 }, { "grad_norm": 0.4496244192123413, "learning_rate": 2.6295387347515165e-05, "loss": 0.0348, "step": 53950 }, { "grad_norm": 1.2900785207748413, "learning_rate": 2.6277191384158727e-05, "loss": 0.0268, "step": 53960 }, { "grad_norm": 1.9089350700378418, "learning_rate": 2.6258999474381457e-05, "loss": 0.034, "step": 53970 }, { "grad_norm": 0.2892269194126129, "learning_rate": 2.624081162129186e-05, "loss": 0.0252, "step": 53980 }, { "grad_norm": 0.45140281319618225, "learning_rate": 2.6222627827997765e-05, "loss": 0.0309, "step": 53990 }, { "grad_norm": 0.25804081559181213, "learning_rate": 2.6204448097606236e-05, "loss": 0.0199, "step": 54000 }, { "grad_norm": 0.26047804951667786, "learning_rate": 2.6186272433223726e-05, "loss": 0.0258, "step": 54010 }, { "grad_norm": 1.0658663511276245, "learning_rate": 2.6168100837955943e-05, "loss": 0.0267, "step": 54020 }, { "grad_norm": 0.2170770764350891, "learning_rate": 2.6149933314907926e-05, "loss": 0.0311, "step": 54030 }, { "grad_norm": 0.3450786769390106, "learning_rate": 2.6131769867184e-05, "loss": 0.0287, "step": 54040 }, { "grad_norm": 0.2916197180747986, "learning_rate": 2.611361049788783e-05, "loss": 0.0233, "step": 54050 }, { "grad_norm": 0.3802040219306946, "learning_rate": 2.6095455210122292e-05, "loss": 0.0248, "step": 54060 }, { "grad_norm": 0.24825869500637054, "learning_rate": 2.6077304006989712e-05, "loss": 0.0173, "step": 54070 }, { "grad_norm": 0.4242515563964844, "learning_rate": 2.6059156891591562e-05, "loss": 0.0253, "step": 54080 }, { "grad_norm": 0.2531021237373352, "learning_rate": 2.6041013867028718e-05, "loss": 0.0238, "step": 54090 }, { "grad_norm": 0.2584965229034424, "learning_rate": 2.6022874936401347e-05, "loss": 0.0222, "step": 54100 }, { "grad_norm": 0.8524219393730164, "learning_rate": 2.6004740102808832e-05, "loss": 0.0351, "step": 54110 }, { "grad_norm": 0.21234627068042755, "learning_rate": 2.598660936935e-05, "loss": 0.0184, "step": 54120 }, { "grad_norm": 0.3387731909751892, "learning_rate": 2.5968482739122845e-05, "loss": 0.0322, "step": 54130 }, { "grad_norm": 0.3266259431838989, "learning_rate": 2.595036021522472e-05, "loss": 0.0221, "step": 54140 }, { "grad_norm": 0.30001014471054077, "learning_rate": 2.5932241800752278e-05, "loss": 0.0232, "step": 54150 }, { "grad_norm": 0.42290550470352173, "learning_rate": 2.5914127498801453e-05, "loss": 0.0334, "step": 54160 }, { "grad_norm": 0.2878262996673584, "learning_rate": 2.5896017312467497e-05, "loss": 0.0231, "step": 54170 }, { "grad_norm": 0.34798961877822876, "learning_rate": 2.587791124484493e-05, "loss": 0.0235, "step": 54180 }, { "grad_norm": 0.2091403305530548, "learning_rate": 2.5859809299027615e-05, "loss": 0.023, "step": 54190 }, { "grad_norm": 0.25187602639198303, "learning_rate": 2.5841711478108632e-05, "loss": 0.0315, "step": 54200 }, { "grad_norm": 0.30438393354415894, "learning_rate": 2.582361778518043e-05, "loss": 0.0266, "step": 54210 }, { "grad_norm": 0.3726152777671814, "learning_rate": 2.580552822333472e-05, "loss": 0.0286, "step": 54220 }, { "grad_norm": 0.42748284339904785, "learning_rate": 2.578744279566252e-05, "loss": 0.0254, "step": 54230 }, { "grad_norm": 0.24402223527431488, "learning_rate": 2.576936150525413e-05, "loss": 0.0308, "step": 54240 }, { "grad_norm": 0.17642314732074738, "learning_rate": 2.5751284355199168e-05, "loss": 0.017, "step": 54250 }, { "grad_norm": 0.2615188658237457, "learning_rate": 2.573321134858646e-05, "loss": 0.0212, "step": 54260 }, { "grad_norm": 0.3143431544303894, "learning_rate": 2.5715142488504286e-05, "loss": 0.0166, "step": 54270 }, { "grad_norm": 0.34387674927711487, "learning_rate": 2.5697077778040042e-05, "loss": 0.0246, "step": 54280 }, { "grad_norm": 0.2976512312889099, "learning_rate": 2.5679017220280522e-05, "loss": 0.03, "step": 54290 }, { "grad_norm": 0.30180326104164124, "learning_rate": 2.5660960818311796e-05, "loss": 0.0245, "step": 54300 }, { "grad_norm": 0.25123655796051025, "learning_rate": 2.564290857521915e-05, "loss": 0.0249, "step": 54310 }, { "grad_norm": 0.21262900531291962, "learning_rate": 2.5624860494087298e-05, "loss": 0.0218, "step": 54320 }, { "grad_norm": 0.23920901119709015, "learning_rate": 2.5606816578000115e-05, "loss": 0.0223, "step": 54330 }, { "grad_norm": 0.34263262152671814, "learning_rate": 2.558877683004082e-05, "loss": 0.0207, "step": 54340 }, { "grad_norm": 0.33229386806488037, "learning_rate": 2.557074125329192e-05, "loss": 0.0241, "step": 54350 }, { "grad_norm": 0.3298364579677582, "learning_rate": 2.5552709850835195e-05, "loss": 0.0196, "step": 54360 }, { "grad_norm": 0.4817213714122772, "learning_rate": 2.5534682625751738e-05, "loss": 0.0212, "step": 54370 }, { "grad_norm": 0.30303481221199036, "learning_rate": 2.551665958112186e-05, "loss": 0.0269, "step": 54380 }, { "grad_norm": 0.3451858460903168, "learning_rate": 2.549864072002527e-05, "loss": 0.0187, "step": 54390 }, { "grad_norm": 0.2885010242462158, "learning_rate": 2.5480626045540858e-05, "loss": 0.0264, "step": 54400 }, { "grad_norm": 0.20041018724441528, "learning_rate": 2.546261556074684e-05, "loss": 0.021, "step": 54410 }, { "grad_norm": 1.1128129959106445, "learning_rate": 2.5444609268720726e-05, "loss": 0.0209, "step": 54420 }, { "grad_norm": 0.17984099686145782, "learning_rate": 2.5426607172539297e-05, "loss": 0.0224, "step": 54430 }, { "grad_norm": 0.14793740212917328, "learning_rate": 2.5408609275278617e-05, "loss": 0.0337, "step": 54440 }, { "grad_norm": 0.36917075514793396, "learning_rate": 2.5390615580014055e-05, "loss": 0.0295, "step": 54450 }, { "grad_norm": 0.2679440975189209, "learning_rate": 2.5372626089820207e-05, "loss": 0.0285, "step": 54460 }, { "grad_norm": 0.20072397589683533, "learning_rate": 2.5354640807770997e-05, "loss": 0.0185, "step": 54470 }, { "grad_norm": 0.3893570303916931, "learning_rate": 2.5336659736939622e-05, "loss": 0.0293, "step": 54480 }, { "grad_norm": 0.41496574878692627, "learning_rate": 2.5318682880398554e-05, "loss": 0.0291, "step": 54490 }, { "grad_norm": 0.3270072937011719, "learning_rate": 2.530071024121956e-05, "loss": 0.028, "step": 54500 }, { "grad_norm": 0.29667896032333374, "learning_rate": 2.5282741822473627e-05, "loss": 0.0213, "step": 54510 }, { "grad_norm": 0.27602410316467285, "learning_rate": 2.526477762723114e-05, "loss": 0.0233, "step": 54520 }, { "grad_norm": 0.3246791660785675, "learning_rate": 2.5246817658561618e-05, "loss": 0.0308, "step": 54530 }, { "grad_norm": 0.37724846601486206, "learning_rate": 2.5228861919533965e-05, "loss": 0.0228, "step": 54540 }, { "grad_norm": 0.5499162673950195, "learning_rate": 2.5210910413216326e-05, "loss": 0.0248, "step": 54550 }, { "grad_norm": 0.2658417522907257, "learning_rate": 2.5192963142676086e-05, "loss": 0.0225, "step": 54560 }, { "grad_norm": 0.22406557202339172, "learning_rate": 2.517502011098001e-05, "loss": 0.0202, "step": 54570 }, { "grad_norm": 0.34431490302085876, "learning_rate": 2.5157081321193987e-05, "loss": 0.0287, "step": 54580 }, { "grad_norm": 0.2363564521074295, "learning_rate": 2.5139146776383356e-05, "loss": 0.0155, "step": 54590 }, { "grad_norm": 0.4497396647930145, "learning_rate": 2.5121216479612575e-05, "loss": 0.02, "step": 54600 }, { "grad_norm": 0.2931268513202667, "learning_rate": 2.510329043394546e-05, "loss": 0.0288, "step": 54610 }, { "grad_norm": 0.48961159586906433, "learning_rate": 2.508536864244508e-05, "loss": 0.0319, "step": 54620 }, { "grad_norm": 0.31475281715393066, "learning_rate": 2.5067451108173778e-05, "loss": 0.021, "step": 54630 }, { "grad_norm": 1.0633777379989624, "learning_rate": 2.5049537834193204e-05, "loss": 0.0197, "step": 54640 }, { "grad_norm": 0.14352113008499146, "learning_rate": 2.5031628823564194e-05, "loss": 0.0201, "step": 54650 }, { "grad_norm": 0.20986776053905487, "learning_rate": 2.5013724079346933e-05, "loss": 0.0235, "step": 54660 }, { "grad_norm": 0.3088906705379486, "learning_rate": 2.4995823604600854e-05, "loss": 0.0245, "step": 54670 }, { "grad_norm": 0.24658937752246857, "learning_rate": 2.497792740238465e-05, "loss": 0.0239, "step": 54680 }, { "grad_norm": 0.3259439170360565, "learning_rate": 2.49600354757563e-05, "loss": 0.0242, "step": 54690 }, { "grad_norm": 0.7953764200210571, "learning_rate": 2.494214782777306e-05, "loss": 0.0226, "step": 54700 }, { "grad_norm": 0.3762132227420807, "learning_rate": 2.4924264461491386e-05, "loss": 0.0232, "step": 54710 }, { "grad_norm": 0.3731154203414917, "learning_rate": 2.4906385379967133e-05, "loss": 0.0336, "step": 54720 }, { "grad_norm": 0.49672505259513855, "learning_rate": 2.4888510586255285e-05, "loss": 0.0253, "step": 54730 }, { "grad_norm": 0.2637103497982025, "learning_rate": 2.487064008341018e-05, "loss": 0.0262, "step": 54740 }, { "grad_norm": 0.1952059417963028, "learning_rate": 2.4852773874485407e-05, "loss": 0.0204, "step": 54750 }, { "grad_norm": 0.24074898660182953, "learning_rate": 2.483491196253377e-05, "loss": 0.0205, "step": 54760 }, { "grad_norm": 0.5095449090003967, "learning_rate": 2.4817054350607443e-05, "loss": 0.0321, "step": 54770 }, { "grad_norm": 0.1552075743675232, "learning_rate": 2.4799201041757742e-05, "loss": 0.0276, "step": 54780 }, { "grad_norm": 0.428828626871109, "learning_rate": 2.4781352039035373e-05, "loss": 0.0278, "step": 54790 }, { "grad_norm": 0.12515945732593536, "learning_rate": 2.4763507345490194e-05, "loss": 0.0199, "step": 54800 }, { "grad_norm": 0.2712872624397278, "learning_rate": 2.4745666964171386e-05, "loss": 0.0204, "step": 54810 }, { "grad_norm": 0.45624834299087524, "learning_rate": 2.4727830898127387e-05, "loss": 0.0236, "step": 54820 }, { "grad_norm": 0.26198631525039673, "learning_rate": 2.4709999150405895e-05, "loss": 0.0267, "step": 54830 }, { "grad_norm": 0.34234872460365295, "learning_rate": 2.469217172405388e-05, "loss": 0.0237, "step": 54840 }, { "grad_norm": 0.2541612684726715, "learning_rate": 2.4674348622117527e-05, "loss": 0.0174, "step": 54850 }, { "grad_norm": 0.25027117133140564, "learning_rate": 2.465652984764234e-05, "loss": 0.0199, "step": 54860 }, { "grad_norm": 0.2936442196369171, "learning_rate": 2.4638715403673056e-05, "loss": 0.0268, "step": 54870 }, { "grad_norm": 0.366182416677475, "learning_rate": 2.4620905293253676e-05, "loss": 0.023, "step": 54880 }, { "grad_norm": 0.338479220867157, "learning_rate": 2.4603099519427463e-05, "loss": 0.0245, "step": 54890 }, { "grad_norm": 0.371425062417984, "learning_rate": 2.458529808523695e-05, "loss": 0.0273, "step": 54900 }, { "grad_norm": 0.16385070979595184, "learning_rate": 2.456750099372387e-05, "loss": 0.0211, "step": 54910 }, { "grad_norm": 0.29080304503440857, "learning_rate": 2.4549708247929327e-05, "loss": 0.0302, "step": 54920 }, { "grad_norm": 0.21331605315208435, "learning_rate": 2.4531919850893554e-05, "loss": 0.0218, "step": 54930 }, { "grad_norm": 0.24516178667545319, "learning_rate": 2.4514135805656125e-05, "loss": 0.0205, "step": 54940 }, { "grad_norm": 0.4604477286338806, "learning_rate": 2.449635611525587e-05, "loss": 0.0241, "step": 54950 }, { "grad_norm": 0.2103290557861328, "learning_rate": 2.447858078273079e-05, "loss": 0.0302, "step": 54960 }, { "grad_norm": 0.29516345262527466, "learning_rate": 2.4460809811118275e-05, "loss": 0.0271, "step": 54970 }, { "grad_norm": 0.4411737024784088, "learning_rate": 2.444304320345483e-05, "loss": 0.024, "step": 54980 }, { "grad_norm": 0.2225779891014099, "learning_rate": 2.4425280962776354e-05, "loss": 0.0341, "step": 54990 }, { "grad_norm": 0.32982420921325684, "learning_rate": 2.4407523092117875e-05, "loss": 0.0258, "step": 55000 }, { "grad_norm": 0.3696056306362152, "learning_rate": 2.438976959451374e-05, "loss": 0.0252, "step": 55010 }, { "grad_norm": 0.22161558270454407, "learning_rate": 2.4372020472997565e-05, "loss": 0.0254, "step": 55020 }, { "grad_norm": 0.23810547590255737, "learning_rate": 2.4354275730602122e-05, "loss": 0.02, "step": 55030 }, { "grad_norm": 0.2236335724592209, "learning_rate": 2.433653537035958e-05, "loss": 0.0239, "step": 55040 }, { "grad_norm": 0.3316064178943634, "learning_rate": 2.431879939530123e-05, "loss": 0.0257, "step": 55050 }, { "grad_norm": 0.33532753586769104, "learning_rate": 2.4301067808457684e-05, "loss": 0.022, "step": 55060 }, { "grad_norm": 0.27676406502723694, "learning_rate": 2.428334061285878e-05, "loss": 0.0251, "step": 55070 }, { "grad_norm": 0.3036685883998871, "learning_rate": 2.426561781153361e-05, "loss": 0.0234, "step": 55080 }, { "grad_norm": 0.6317281723022461, "learning_rate": 2.424789940751052e-05, "loss": 0.0241, "step": 55090 }, { "grad_norm": 0.3736001253128052, "learning_rate": 2.423018540381712e-05, "loss": 0.0263, "step": 55100 }, { "grad_norm": 0.5009517669677734, "learning_rate": 2.421247580348021e-05, "loss": 0.0222, "step": 55110 }, { "grad_norm": 0.1897704303264618, "learning_rate": 2.4194770609525897e-05, "loss": 0.026, "step": 55120 }, { "grad_norm": 0.404052734375, "learning_rate": 2.417706982497951e-05, "loss": 0.0249, "step": 55130 }, { "grad_norm": 0.4961673617362976, "learning_rate": 2.415937345286563e-05, "loss": 0.0288, "step": 55140 }, { "grad_norm": 0.4810545742511749, "learning_rate": 2.4141681496208087e-05, "loss": 0.0236, "step": 55150 }, { "grad_norm": 0.22199951112270355, "learning_rate": 2.4123993958029946e-05, "loss": 0.0245, "step": 55160 }, { "grad_norm": 0.33598095178604126, "learning_rate": 2.4106310841353548e-05, "loss": 0.0217, "step": 55170 }, { "grad_norm": 0.23044990003108978, "learning_rate": 2.4088632149200398e-05, "loss": 0.0258, "step": 55180 }, { "grad_norm": 0.22573456168174744, "learning_rate": 2.4070957884591367e-05, "loss": 0.024, "step": 55190 }, { "grad_norm": 0.5309553742408752, "learning_rate": 2.4053288050546464e-05, "loss": 0.0307, "step": 55200 }, { "grad_norm": 0.3479788303375244, "learning_rate": 2.403562265008498e-05, "loss": 0.0197, "step": 55210 }, { "grad_norm": 0.43785807490348816, "learning_rate": 2.4017961686225483e-05, "loss": 0.0308, "step": 55220 }, { "grad_norm": 0.23737077414989471, "learning_rate": 2.400030516198568e-05, "loss": 0.0314, "step": 55230 }, { "grad_norm": 0.34759607911109924, "learning_rate": 2.3982653080382673e-05, "loss": 0.0233, "step": 55240 }, { "grad_norm": 0.2909890413284302, "learning_rate": 2.396500544443266e-05, "loss": 0.0217, "step": 55250 }, { "grad_norm": 0.26028189063072205, "learning_rate": 2.3947362257151156e-05, "loss": 0.0161, "step": 55260 }, { "grad_norm": 0.3777708411216736, "learning_rate": 2.39297235215529e-05, "loss": 0.0224, "step": 55270 }, { "grad_norm": 0.2440500408411026, "learning_rate": 2.3912089240651873e-05, "loss": 0.0229, "step": 55280 }, { "grad_norm": 0.28037869930267334, "learning_rate": 2.389445941746129e-05, "loss": 0.0314, "step": 55290 }, { "grad_norm": 0.39153948426246643, "learning_rate": 2.3876834054993625e-05, "loss": 0.0236, "step": 55300 }, { "grad_norm": 0.2287539541721344, "learning_rate": 2.3859213156260522e-05, "loss": 0.0263, "step": 55310 }, { "grad_norm": 0.41224339604377747, "learning_rate": 2.3841596724272948e-05, "loss": 0.0253, "step": 55320 }, { "grad_norm": 0.6060338616371155, "learning_rate": 2.382398476204106e-05, "loss": 0.0276, "step": 55330 }, { "grad_norm": 0.4091012477874756, "learning_rate": 2.3806377272574254e-05, "loss": 0.0276, "step": 55340 }, { "grad_norm": 0.35084807872772217, "learning_rate": 2.3788774258881174e-05, "loss": 0.0227, "step": 55350 }, { "grad_norm": 0.36100855469703674, "learning_rate": 2.37711757239697e-05, "loss": 0.0266, "step": 55360 }, { "grad_norm": 0.3765948712825775, "learning_rate": 2.3753581670846954e-05, "loss": 0.0233, "step": 55370 }, { "grad_norm": 0.21298038959503174, "learning_rate": 2.3735992102519216e-05, "loss": 0.0237, "step": 55380 }, { "grad_norm": 0.23480328917503357, "learning_rate": 2.371840702199215e-05, "loss": 0.0221, "step": 55390 }, { "grad_norm": 0.16861167550086975, "learning_rate": 2.3700826432270494e-05, "loss": 0.0179, "step": 55400 }, { "grad_norm": 0.3179710805416107, "learning_rate": 2.3683250336358326e-05, "loss": 0.0341, "step": 55410 }, { "grad_norm": 0.2250511646270752, "learning_rate": 2.3665678737258923e-05, "loss": 0.031, "step": 55420 }, { "grad_norm": 0.19226352870464325, "learning_rate": 2.3648111637974745e-05, "loss": 0.0214, "step": 55430 }, { "grad_norm": 0.22353816032409668, "learning_rate": 2.36305490415076e-05, "loss": 0.0231, "step": 55440 }, { "grad_norm": 0.41556409001350403, "learning_rate": 2.3612990950858392e-05, "loss": 0.0265, "step": 55450 }, { "grad_norm": 0.27700719237327576, "learning_rate": 2.359543736902735e-05, "loss": 0.0225, "step": 55460 }, { "grad_norm": 0.30529096722602844, "learning_rate": 2.3577888299013896e-05, "loss": 0.0234, "step": 55470 }, { "grad_norm": 0.3377828001976013, "learning_rate": 2.3560343743816683e-05, "loss": 0.0209, "step": 55480 }, { "grad_norm": 0.20564508438110352, "learning_rate": 2.354280370643362e-05, "loss": 0.0189, "step": 55490 }, { "grad_norm": 0.2097746729850769, "learning_rate": 2.3525268189861777e-05, "loss": 0.0188, "step": 55500 }, { "grad_norm": 0.37630146741867065, "learning_rate": 2.3507737197097513e-05, "loss": 0.028, "step": 55510 }, { "grad_norm": 0.3197227120399475, "learning_rate": 2.34902107311364e-05, "loss": 0.0228, "step": 55520 }, { "grad_norm": 0.25811517238616943, "learning_rate": 2.3472688794973225e-05, "loss": 0.024, "step": 55530 }, { "grad_norm": 0.28416597843170166, "learning_rate": 2.3455171391602016e-05, "loss": 0.0294, "step": 55540 }, { "grad_norm": 0.6565330624580383, "learning_rate": 2.343765852401601e-05, "loss": 0.0162, "step": 55550 }, { "grad_norm": 0.13653066754341125, "learning_rate": 2.342015019520768e-05, "loss": 0.0266, "step": 55560 }, { "grad_norm": 0.25513404607772827, "learning_rate": 2.3402646408168742e-05, "loss": 0.0211, "step": 55570 }, { "grad_norm": 0.6478719711303711, "learning_rate": 2.3385147165890074e-05, "loss": 0.0264, "step": 55580 }, { "grad_norm": 0.5274732112884521, "learning_rate": 2.336765247136184e-05, "loss": 0.0284, "step": 55590 }, { "grad_norm": 0.5067312121391296, "learning_rate": 2.33501623275734e-05, "loss": 0.0293, "step": 55600 }, { "grad_norm": 0.36886313557624817, "learning_rate": 2.333267673751334e-05, "loss": 0.0239, "step": 55610 }, { "grad_norm": 0.2632676661014557, "learning_rate": 2.33151957041695e-05, "loss": 0.0217, "step": 55620 }, { "grad_norm": 0.3965230882167816, "learning_rate": 2.329771923052884e-05, "loss": 0.0241, "step": 55630 }, { "grad_norm": 0.28203266859054565, "learning_rate": 2.3280247319577697e-05, "loss": 0.0256, "step": 55640 }, { "grad_norm": 0.29277336597442627, "learning_rate": 2.3262779974301473e-05, "loss": 0.0219, "step": 55650 }, { "grad_norm": 0.2847592234611511, "learning_rate": 2.3245317197684895e-05, "loss": 0.0234, "step": 55660 }, { "grad_norm": 0.21946394443511963, "learning_rate": 2.322785899271188e-05, "loss": 0.0206, "step": 55670 }, { "grad_norm": 0.2748124897480011, "learning_rate": 2.3210405362365507e-05, "loss": 0.0285, "step": 55680 }, { "grad_norm": 0.3067530691623688, "learning_rate": 2.31929563096282e-05, "loss": 0.0261, "step": 55690 }, { "grad_norm": 0.4341123402118683, "learning_rate": 2.317551183748146e-05, "loss": 0.0194, "step": 55700 }, { "grad_norm": 0.24072304368019104, "learning_rate": 2.3158071948906103e-05, "loss": 0.025, "step": 55710 }, { "grad_norm": 0.14218257367610931, "learning_rate": 2.314063664688212e-05, "loss": 0.0185, "step": 55720 }, { "grad_norm": 0.24896422028541565, "learning_rate": 2.3123205934388725e-05, "loss": 0.0224, "step": 55730 }, { "grad_norm": 0.24545910954475403, "learning_rate": 2.310577981440436e-05, "loss": 0.0244, "step": 55740 }, { "grad_norm": 0.2647637724876404, "learning_rate": 2.3088358289906665e-05, "loss": 0.0232, "step": 55750 }, { "grad_norm": 0.3818734586238861, "learning_rate": 2.307094136387252e-05, "loss": 0.0208, "step": 55760 }, { "grad_norm": 0.3359818756580353, "learning_rate": 2.305352903927796e-05, "loss": 0.0282, "step": 55770 }, { "grad_norm": 0.5151082277297974, "learning_rate": 2.303612131909831e-05, "loss": 0.0167, "step": 55780 }, { "grad_norm": 0.2890724539756775, "learning_rate": 2.3018718206308054e-05, "loss": 0.026, "step": 55790 }, { "grad_norm": 0.34358495473861694, "learning_rate": 2.3001319703880925e-05, "loss": 0.0246, "step": 55800 }, { "grad_norm": 0.3386407196521759, "learning_rate": 2.2983925814789835e-05, "loss": 0.0201, "step": 55810 }, { "grad_norm": 0.2826140820980072, "learning_rate": 2.2966536542006957e-05, "loss": 0.0182, "step": 55820 }, { "grad_norm": 0.49855709075927734, "learning_rate": 2.294915188850358e-05, "loss": 0.0275, "step": 55830 }, { "grad_norm": 0.3449437916278839, "learning_rate": 2.2931771857250333e-05, "loss": 0.0304, "step": 55840 }, { "grad_norm": 0.36838802695274353, "learning_rate": 2.2914396451216946e-05, "loss": 0.0224, "step": 55850 }, { "grad_norm": 0.2915659248828888, "learning_rate": 2.2897025673372412e-05, "loss": 0.0231, "step": 55860 }, { "grad_norm": 0.2995680272579193, "learning_rate": 2.287965952668494e-05, "loss": 0.0225, "step": 55870 }, { "grad_norm": 0.2608845829963684, "learning_rate": 2.2862298014121873e-05, "loss": 0.0241, "step": 55880 }, { "grad_norm": 0.22033268213272095, "learning_rate": 2.28449411386499e-05, "loss": 0.0242, "step": 55890 }, { "grad_norm": 0.22591011226177216, "learning_rate": 2.282758890323477e-05, "loss": 0.0228, "step": 55900 }, { "grad_norm": 0.18409176170825958, "learning_rate": 2.2810241310841528e-05, "loss": 0.0229, "step": 55910 }, { "grad_norm": 0.30944618582725525, "learning_rate": 2.2792898364434412e-05, "loss": 0.0239, "step": 55920 }, { "grad_norm": 0.248247891664505, "learning_rate": 2.2775560066976843e-05, "loss": 0.0212, "step": 55930 }, { "grad_norm": 0.28226524591445923, "learning_rate": 2.275822642143147e-05, "loss": 0.0294, "step": 55940 }, { "grad_norm": 0.3299994170665741, "learning_rate": 2.274089743076014e-05, "loss": 0.0259, "step": 55950 }, { "grad_norm": 0.3812748193740845, "learning_rate": 2.2723573097923917e-05, "loss": 0.0256, "step": 55960 }, { "grad_norm": 0.2993490695953369, "learning_rate": 2.2706253425883024e-05, "loss": 0.0148, "step": 55970 }, { "grad_norm": 0.36315426230430603, "learning_rate": 2.2688938417596933e-05, "loss": 0.0243, "step": 55980 }, { "grad_norm": 0.330693781375885, "learning_rate": 2.2671628076024305e-05, "loss": 0.0263, "step": 55990 }, { "grad_norm": 0.2153942584991455, "learning_rate": 2.2654322404123008e-05, "loss": 0.0164, "step": 56000 }, { "grad_norm": 0.1839608997106552, "learning_rate": 2.2637021404850105e-05, "loss": 0.02, "step": 56010 }, { "grad_norm": 0.2569739818572998, "learning_rate": 2.2619725081161885e-05, "loss": 0.0215, "step": 56020 }, { "grad_norm": 0.9842431545257568, "learning_rate": 2.260243343601376e-05, "loss": 0.0232, "step": 56030 }, { "grad_norm": 0.25822246074676514, "learning_rate": 2.2585146472360473e-05, "loss": 0.0246, "step": 56040 }, { "grad_norm": 0.23196497559547424, "learning_rate": 2.2567864193155834e-05, "loss": 0.0194, "step": 56050 }, { "grad_norm": 0.29587218165397644, "learning_rate": 2.255058660135294e-05, "loss": 0.0218, "step": 56060 }, { "grad_norm": 0.4114300310611725, "learning_rate": 2.253331369990407e-05, "loss": 0.0217, "step": 56070 }, { "grad_norm": 0.20999839901924133, "learning_rate": 2.251604549176063e-05, "loss": 0.0252, "step": 56080 }, { "grad_norm": 0.17957554757595062, "learning_rate": 2.249878197987337e-05, "loss": 0.0214, "step": 56090 }, { "grad_norm": 0.22281445562839508, "learning_rate": 2.2481523167192087e-05, "loss": 0.0206, "step": 56100 }, { "grad_norm": 0.21557137370109558, "learning_rate": 2.2464269056665867e-05, "loss": 0.0242, "step": 56110 }, { "grad_norm": 0.3256968855857849, "learning_rate": 2.2447019651242958e-05, "loss": 0.026, "step": 56120 }, { "grad_norm": 0.293916255235672, "learning_rate": 2.2429774953870818e-05, "loss": 0.0212, "step": 56130 }, { "grad_norm": 0.46411168575286865, "learning_rate": 2.241253496749611e-05, "loss": 0.0237, "step": 56140 }, { "grad_norm": 0.2107139229774475, "learning_rate": 2.2395299695064614e-05, "loss": 0.017, "step": 56150 }, { "grad_norm": 0.27863961458206177, "learning_rate": 2.237806913952145e-05, "loss": 0.0284, "step": 56160 }, { "grad_norm": 0.4914872646331787, "learning_rate": 2.2360843303810798e-05, "loss": 0.018, "step": 56170 }, { "grad_norm": 0.27534106373786926, "learning_rate": 2.2343622190876084e-05, "loss": 0.0182, "step": 56180 }, { "grad_norm": 0.29144299030303955, "learning_rate": 2.2326405803659935e-05, "loss": 0.0237, "step": 56190 }, { "grad_norm": 0.19439615309238434, "learning_rate": 2.230919414510416e-05, "loss": 0.0191, "step": 56200 }, { "grad_norm": 0.3467516005039215, "learning_rate": 2.229198721814976e-05, "loss": 0.0218, "step": 56210 }, { "grad_norm": 0.3718661069869995, "learning_rate": 2.2274785025736948e-05, "loss": 0.0249, "step": 56220 }, { "grad_norm": 0.14045703411102295, "learning_rate": 2.225758757080507e-05, "loss": 0.0214, "step": 56230 }, { "grad_norm": 0.4485246241092682, "learning_rate": 2.2240394856292723e-05, "loss": 0.0234, "step": 56240 }, { "grad_norm": 0.4598727822303772, "learning_rate": 2.2223206885137664e-05, "loss": 0.0214, "step": 56250 }, { "grad_norm": 0.47986358404159546, "learning_rate": 2.2206023660276853e-05, "loss": 0.0252, "step": 56260 }, { "grad_norm": 0.2114441990852356, "learning_rate": 2.218884518464645e-05, "loss": 0.0241, "step": 56270 }, { "grad_norm": 0.3712207078933716, "learning_rate": 2.2171671461181732e-05, "loss": 0.0215, "step": 56280 }, { "grad_norm": 0.17181257903575897, "learning_rate": 2.2154502492817292e-05, "loss": 0.0208, "step": 56290 }, { "grad_norm": 0.31055793166160583, "learning_rate": 2.2137338282486782e-05, "loss": 0.0191, "step": 56300 }, { "grad_norm": 0.325827419757843, "learning_rate": 2.2120178833123113e-05, "loss": 0.0194, "step": 56310 }, { "grad_norm": 0.30452361702919006, "learning_rate": 2.210302414765838e-05, "loss": 0.0197, "step": 56320 }, { "grad_norm": 0.24546484649181366, "learning_rate": 2.2085874229023808e-05, "loss": 0.0159, "step": 56330 }, { "grad_norm": 0.2630923092365265, "learning_rate": 2.2068729080149907e-05, "loss": 0.0217, "step": 56340 }, { "grad_norm": 0.35616356134414673, "learning_rate": 2.205158870396625e-05, "loss": 0.0252, "step": 56350 }, { "grad_norm": 0.22666959464550018, "learning_rate": 2.2034453103401732e-05, "loss": 0.0199, "step": 56360 }, { "grad_norm": 0.11779752373695374, "learning_rate": 2.20173222813843e-05, "loss": 0.0198, "step": 56370 }, { "grad_norm": 0.36398768424987793, "learning_rate": 2.200019624084116e-05, "loss": 0.022, "step": 56380 }, { "grad_norm": 0.3548026382923126, "learning_rate": 2.1983074984698687e-05, "loss": 0.0207, "step": 56390 }, { "grad_norm": 0.31833595037460327, "learning_rate": 2.1965958515882433e-05, "loss": 0.0216, "step": 56400 }, { "grad_norm": 0.24413204193115234, "learning_rate": 2.1948846837317162e-05, "loss": 0.0248, "step": 56410 }, { "grad_norm": 0.3270242512226105, "learning_rate": 2.1931739951926738e-05, "loss": 0.0232, "step": 56420 }, { "grad_norm": 0.47467976808547974, "learning_rate": 2.191463786263429e-05, "loss": 0.0201, "step": 56430 }, { "grad_norm": 0.24091379344463348, "learning_rate": 2.1897540572362095e-05, "loss": 0.0234, "step": 56440 }, { "grad_norm": 0.45816752314567566, "learning_rate": 2.1880448084031614e-05, "loss": 0.0178, "step": 56450 }, { "grad_norm": 0.23507152497768402, "learning_rate": 2.1863360400563482e-05, "loss": 0.0215, "step": 56460 }, { "grad_norm": 0.19520603120326996, "learning_rate": 2.184627752487754e-05, "loss": 0.0259, "step": 56470 }, { "grad_norm": 0.3158094882965088, "learning_rate": 2.1829199459892725e-05, "loss": 0.0247, "step": 56480 }, { "grad_norm": 0.24773474037647247, "learning_rate": 2.1812126208527282e-05, "loss": 0.0226, "step": 56490 }, { "grad_norm": 0.25185245275497437, "learning_rate": 2.179505777369852e-05, "loss": 0.0263, "step": 56500 }, { "grad_norm": 0.22268417477607727, "learning_rate": 2.1777994158322974e-05, "loss": 0.0203, "step": 56510 }, { "grad_norm": 0.22729744017124176, "learning_rate": 2.176093536531637e-05, "loss": 0.0241, "step": 56520 }, { "grad_norm": 0.36633914709091187, "learning_rate": 2.1743881397593537e-05, "loss": 0.0242, "step": 56530 }, { "grad_norm": 0.45590832829475403, "learning_rate": 2.1726832258068595e-05, "loss": 0.0354, "step": 56540 }, { "grad_norm": 0.3823604881763458, "learning_rate": 2.170978794965472e-05, "loss": 0.0265, "step": 56550 }, { "grad_norm": 0.3961062431335449, "learning_rate": 2.169274847526438e-05, "loss": 0.0232, "step": 56560 }, { "grad_norm": 0.4857843220233917, "learning_rate": 2.1675713837809103e-05, "loss": 0.0248, "step": 56570 }, { "grad_norm": 0.21233075857162476, "learning_rate": 2.1658684040199655e-05, "loss": 0.0214, "step": 56580 }, { "grad_norm": 0.22500815987586975, "learning_rate": 2.1641659085345974e-05, "loss": 0.0219, "step": 56590 }, { "grad_norm": 0.24975068867206573, "learning_rate": 2.1624638976157154e-05, "loss": 0.0205, "step": 56600 }, { "grad_norm": 0.34358537197113037, "learning_rate": 2.1607623715541476e-05, "loss": 0.0264, "step": 56610 }, { "grad_norm": 0.5835800766944885, "learning_rate": 2.159061330640636e-05, "loss": 0.0226, "step": 56620 }, { "grad_norm": 0.1901322305202484, "learning_rate": 2.1573607751658425e-05, "loss": 0.0205, "step": 56630 }, { "grad_norm": 0.43900248408317566, "learning_rate": 2.155660705420347e-05, "loss": 0.0245, "step": 56640 }, { "grad_norm": 0.12464737892150879, "learning_rate": 2.153961121694644e-05, "loss": 0.0162, "step": 56650 }, { "grad_norm": 0.31795021891593933, "learning_rate": 2.152262024279145e-05, "loss": 0.0305, "step": 56660 }, { "grad_norm": 0.6281510591506958, "learning_rate": 2.150563413464183e-05, "loss": 0.0269, "step": 56670 }, { "grad_norm": 0.20498982071876526, "learning_rate": 2.1488652895399974e-05, "loss": 0.0153, "step": 56680 }, { "grad_norm": 0.28844889998435974, "learning_rate": 2.1471676527967587e-05, "loss": 0.0259, "step": 56690 }, { "grad_norm": 0.25409021973609924, "learning_rate": 2.1454705035245414e-05, "loss": 0.0291, "step": 56700 }, { "grad_norm": 0.2759891748428345, "learning_rate": 2.143773842013343e-05, "loss": 0.0142, "step": 56710 }, { "grad_norm": 0.42701300978660583, "learning_rate": 2.1420776685530796e-05, "loss": 0.0263, "step": 56720 }, { "grad_norm": 0.4144575297832489, "learning_rate": 2.1403819834335742e-05, "loss": 0.0218, "step": 56730 }, { "grad_norm": 0.23791617155075073, "learning_rate": 2.1386867869445808e-05, "loss": 0.0256, "step": 56740 }, { "grad_norm": 0.23268967866897583, "learning_rate": 2.1369920793757548e-05, "loss": 0.0285, "step": 56750 }, { "grad_norm": 0.36169424653053284, "learning_rate": 2.1352978610166828e-05, "loss": 0.0207, "step": 56760 }, { "grad_norm": 0.31992462277412415, "learning_rate": 2.1336041321568546e-05, "loss": 0.025, "step": 56770 }, { "grad_norm": 0.30725306272506714, "learning_rate": 2.131910893085684e-05, "loss": 0.0195, "step": 56780 }, { "grad_norm": 0.3966454863548279, "learning_rate": 2.1302181440925012e-05, "loss": 0.0328, "step": 56790 }, { "grad_norm": 0.24939295649528503, "learning_rate": 2.128525885466546e-05, "loss": 0.0213, "step": 56800 }, { "grad_norm": 0.3438483476638794, "learning_rate": 2.1268341174969847e-05, "loss": 0.0302, "step": 56810 }, { "grad_norm": 0.3125218152999878, "learning_rate": 2.1251428404728894e-05, "loss": 0.0255, "step": 56820 }, { "grad_norm": 0.31323087215423584, "learning_rate": 2.123452054683256e-05, "loss": 0.0182, "step": 56830 }, { "grad_norm": 0.30543938279151917, "learning_rate": 2.1217617604169927e-05, "loss": 0.0249, "step": 56840 }, { "grad_norm": 0.36244168877601624, "learning_rate": 2.120071957962924e-05, "loss": 0.0218, "step": 56850 }, { "grad_norm": 0.33143118023872375, "learning_rate": 2.1183826476097917e-05, "loss": 0.0332, "step": 56860 }, { "grad_norm": 0.18745210766792297, "learning_rate": 2.116693829646254e-05, "loss": 0.0188, "step": 56870 }, { "grad_norm": 0.5067057013511658, "learning_rate": 2.1150055043608806e-05, "loss": 0.0244, "step": 56880 }, { "grad_norm": 0.348483145236969, "learning_rate": 2.113317672042162e-05, "loss": 0.0267, "step": 56890 }, { "grad_norm": 0.25203701853752136, "learning_rate": 2.1116303329785025e-05, "loss": 0.0153, "step": 56900 }, { "grad_norm": 0.3839127719402313, "learning_rate": 2.109943487458222e-05, "loss": 0.0206, "step": 56910 }, { "grad_norm": 0.2212899923324585, "learning_rate": 2.1082571357695574e-05, "loss": 0.0167, "step": 56920 }, { "grad_norm": 0.2768297493457794, "learning_rate": 2.1065712782006557e-05, "loss": 0.0166, "step": 56930 }, { "grad_norm": 0.4495992660522461, "learning_rate": 2.104885915039591e-05, "loss": 0.0217, "step": 56940 }, { "grad_norm": 0.2741677463054657, "learning_rate": 2.103201046574338e-05, "loss": 0.0219, "step": 56950 }, { "grad_norm": 0.28390511870384216, "learning_rate": 2.1015166730928022e-05, "loss": 0.0221, "step": 56960 }, { "grad_norm": 0.6165077090263367, "learning_rate": 2.0998327948827912e-05, "loss": 0.0246, "step": 56970 }, { "grad_norm": 0.5655339360237122, "learning_rate": 2.0981494122320363e-05, "loss": 0.0266, "step": 56980 }, { "grad_norm": 0.3564375042915344, "learning_rate": 2.0964665254281822e-05, "loss": 0.0296, "step": 56990 }, { "grad_norm": 0.4449792504310608, "learning_rate": 2.094784134758784e-05, "loss": 0.0268, "step": 57000 }, { "grad_norm": 0.3863370418548584, "learning_rate": 2.0931022405113226e-05, "loss": 0.035, "step": 57010 }, { "grad_norm": 0.3540549874305725, "learning_rate": 2.091420842973183e-05, "loss": 0.0259, "step": 57020 }, { "grad_norm": 0.27535995841026306, "learning_rate": 2.0897399424316715e-05, "loss": 0.0146, "step": 57030 }, { "grad_norm": 0.3983425796031952, "learning_rate": 2.0880595391740078e-05, "loss": 0.025, "step": 57040 }, { "grad_norm": 0.20079512894153595, "learning_rate": 2.0863796334873277e-05, "loss": 0.0251, "step": 57050 }, { "grad_norm": 0.3142465054988861, "learning_rate": 2.08470022565868e-05, "loss": 0.0216, "step": 57060 }, { "grad_norm": 0.31793901324272156, "learning_rate": 2.0830213159750317e-05, "loss": 0.0196, "step": 57070 }, { "grad_norm": 0.22294504940509796, "learning_rate": 2.0813429047232596e-05, "loss": 0.0239, "step": 57080 }, { "grad_norm": 0.43454959988594055, "learning_rate": 2.0796649921901594e-05, "loss": 0.023, "step": 57090 }, { "grad_norm": 0.29009753465652466, "learning_rate": 2.077987578662441e-05, "loss": 0.0204, "step": 57100 }, { "grad_norm": 0.37024980783462524, "learning_rate": 2.0763106644267277e-05, "loss": 0.0274, "step": 57110 }, { "grad_norm": 0.2207527458667755, "learning_rate": 2.0746342497695607e-05, "loss": 0.0176, "step": 57120 }, { "grad_norm": 0.39032435417175293, "learning_rate": 2.0729583349773886e-05, "loss": 0.0203, "step": 57130 }, { "grad_norm": 0.345583438873291, "learning_rate": 2.0712829203365853e-05, "loss": 0.0225, "step": 57140 }, { "grad_norm": 0.2086036652326584, "learning_rate": 2.0696080061334267e-05, "loss": 0.0168, "step": 57150 }, { "grad_norm": 0.3477088212966919, "learning_rate": 2.067933592654117e-05, "loss": 0.0213, "step": 57160 }, { "grad_norm": 0.28459179401397705, "learning_rate": 2.066259680184763e-05, "loss": 0.0277, "step": 57170 }, { "grad_norm": 0.20712432265281677, "learning_rate": 2.0645862690113908e-05, "loss": 0.0217, "step": 57180 }, { "grad_norm": 0.29840508103370667, "learning_rate": 2.0629133594199436e-05, "loss": 0.0322, "step": 57190 }, { "grad_norm": 0.17039933800697327, "learning_rate": 2.0612409516962704e-05, "loss": 0.0186, "step": 57200 }, { "grad_norm": 0.3058634400367737, "learning_rate": 2.0595690461261467e-05, "loss": 0.024, "step": 57210 }, { "grad_norm": 0.3685203492641449, "learning_rate": 2.0578976429952503e-05, "loss": 0.0215, "step": 57220 }, { "grad_norm": 0.28455424308776855, "learning_rate": 2.0562267425891802e-05, "loss": 0.0224, "step": 57230 }, { "grad_norm": 0.2604030668735504, "learning_rate": 2.0545563451934467e-05, "loss": 0.0274, "step": 57240 }, { "grad_norm": 0.35518887639045715, "learning_rate": 2.0528864510934764e-05, "loss": 0.0261, "step": 57250 }, { "grad_norm": 0.17928186058998108, "learning_rate": 2.0512170605746096e-05, "loss": 0.028, "step": 57260 }, { "grad_norm": 0.5591768026351929, "learning_rate": 2.0495481739220963e-05, "loss": 0.0256, "step": 57270 }, { "grad_norm": 0.3631732761859894, "learning_rate": 2.0478797914211045e-05, "loss": 0.0231, "step": 57280 }, { "grad_norm": 0.3671826124191284, "learning_rate": 2.046211913356716e-05, "loss": 0.0273, "step": 57290 }, { "grad_norm": 0.3292042016983032, "learning_rate": 2.0445445400139247e-05, "loss": 0.0239, "step": 57300 }, { "grad_norm": 0.47564107179641724, "learning_rate": 2.0428776716776405e-05, "loss": 0.0237, "step": 57310 }, { "grad_norm": 0.30474478006362915, "learning_rate": 2.0412113086326856e-05, "loss": 0.0265, "step": 57320 }, { "grad_norm": 0.24296462535858154, "learning_rate": 2.0395454511637918e-05, "loss": 0.0221, "step": 57330 }, { "grad_norm": 0.2851669490337372, "learning_rate": 2.037880099555616e-05, "loss": 0.0157, "step": 57340 }, { "grad_norm": 0.33800408244132996, "learning_rate": 2.0362152540927144e-05, "loss": 0.0196, "step": 57350 }, { "grad_norm": 0.28007814288139343, "learning_rate": 2.0345509150595666e-05, "loss": 0.0366, "step": 57360 }, { "grad_norm": 0.25503894686698914, "learning_rate": 2.0328870827405617e-05, "loss": 0.0272, "step": 57370 }, { "grad_norm": 0.3168717622756958, "learning_rate": 2.0312237574200043e-05, "loss": 0.0244, "step": 57380 }, { "grad_norm": 0.2516375780105591, "learning_rate": 2.029560939382112e-05, "loss": 0.0252, "step": 57390 }, { "grad_norm": 0.28878697752952576, "learning_rate": 2.0278986289110097e-05, "loss": 0.0193, "step": 57400 }, { "grad_norm": 0.35629740357398987, "learning_rate": 2.0262368262907484e-05, "loss": 0.0262, "step": 57410 }, { "grad_norm": 0.42045605182647705, "learning_rate": 2.024575531805279e-05, "loss": 0.0283, "step": 57420 }, { "grad_norm": 0.43429458141326904, "learning_rate": 2.0229147457384735e-05, "loss": 0.0232, "step": 57430 }, { "grad_norm": 0.31075865030288696, "learning_rate": 2.0212544683741157e-05, "loss": 0.0325, "step": 57440 }, { "grad_norm": 0.25118207931518555, "learning_rate": 2.0195946999958976e-05, "loss": 0.0295, "step": 57450 }, { "grad_norm": 0.2615946829319, "learning_rate": 2.017935440887434e-05, "loss": 0.0272, "step": 57460 }, { "grad_norm": 0.3438457250595093, "learning_rate": 2.0162766913322423e-05, "loss": 0.0232, "step": 57470 }, { "grad_norm": 0.27966776490211487, "learning_rate": 2.0146184516137588e-05, "loss": 0.0269, "step": 57480 }, { "grad_norm": 0.17211613059043884, "learning_rate": 2.012960722015332e-05, "loss": 0.0259, "step": 57490 }, { "grad_norm": 0.45963186025619507, "learning_rate": 2.0113035028202214e-05, "loss": 0.0242, "step": 57500 }, { "grad_norm": 0.2793557345867157, "learning_rate": 2.009646794311602e-05, "loss": 0.0295, "step": 57510 }, { "grad_norm": 0.1920567899942398, "learning_rate": 2.007990596772559e-05, "loss": 0.0263, "step": 57520 }, { "grad_norm": 0.2849411070346832, "learning_rate": 2.0063349104860923e-05, "loss": 0.0262, "step": 57530 }, { "grad_norm": 0.27085554599761963, "learning_rate": 2.0046797357351116e-05, "loss": 0.0285, "step": 57540 }, { "grad_norm": 0.7047070860862732, "learning_rate": 2.0030250728024412e-05, "loss": 0.0198, "step": 57550 }, { "grad_norm": 0.22885239124298096, "learning_rate": 2.001370921970819e-05, "loss": 0.02, "step": 57560 }, { "grad_norm": 0.2870716154575348, "learning_rate": 1.9997172835228932e-05, "loss": 0.0164, "step": 57570 }, { "grad_norm": 0.3593797981739044, "learning_rate": 1.9980641577412262e-05, "loss": 0.02, "step": 57580 }, { "grad_norm": 0.2477998435497284, "learning_rate": 1.9964115449082925e-05, "loss": 0.0236, "step": 57590 }, { "grad_norm": 0.791777491569519, "learning_rate": 1.9947594453064742e-05, "loss": 0.0281, "step": 57600 }, { "grad_norm": 0.2586710751056671, "learning_rate": 1.9931078592180774e-05, "loss": 0.0232, "step": 57610 }, { "grad_norm": 0.40289878845214844, "learning_rate": 1.9914567869253065e-05, "loss": 0.0239, "step": 57620 }, { "grad_norm": 0.3191557228565216, "learning_rate": 1.989806228710287e-05, "loss": 0.0315, "step": 57630 }, { "grad_norm": 0.38854995369911194, "learning_rate": 1.9881561848550555e-05, "loss": 0.02, "step": 57640 }, { "grad_norm": 0.25063514709472656, "learning_rate": 1.9865066556415544e-05, "loss": 0.0256, "step": 57650 }, { "grad_norm": 0.3530242443084717, "learning_rate": 1.98485764135165e-05, "loss": 0.0249, "step": 57660 }, { "grad_norm": 0.2004314661026001, "learning_rate": 1.983209142267109e-05, "loss": 0.0207, "step": 57670 }, { "grad_norm": 0.4508689343929291, "learning_rate": 1.9815611586696165e-05, "loss": 0.023, "step": 57680 }, { "grad_norm": 0.3715018928050995, "learning_rate": 1.9799136908407667e-05, "loss": 0.0297, "step": 57690 }, { "grad_norm": 0.31772398948669434, "learning_rate": 1.9782667390620678e-05, "loss": 0.0212, "step": 57700 }, { "grad_norm": 0.32126277685165405, "learning_rate": 1.976620303614939e-05, "loss": 0.0233, "step": 57710 }, { "grad_norm": 0.23013564944267273, "learning_rate": 1.9749743847807108e-05, "loss": 0.0195, "step": 57720 }, { "grad_norm": 0.37210676074028015, "learning_rate": 1.9733289828406272e-05, "loss": 0.0193, "step": 57730 }, { "grad_norm": 0.35416361689567566, "learning_rate": 1.9716840980758382e-05, "loss": 0.0226, "step": 57740 }, { "grad_norm": 0.3070623576641083, "learning_rate": 1.9700397307674134e-05, "loss": 0.0261, "step": 57750 }, { "grad_norm": 0.36237832903862, "learning_rate": 1.968395881196328e-05, "loss": 0.0219, "step": 57760 }, { "grad_norm": 0.49220773577690125, "learning_rate": 1.966752549643473e-05, "loss": 0.0299, "step": 57770 }, { "grad_norm": 0.4149192273616791, "learning_rate": 1.965109736389647e-05, "loss": 0.0202, "step": 57780 }, { "grad_norm": 0.3783203065395355, "learning_rate": 1.9634674417155645e-05, "loss": 0.0194, "step": 57790 }, { "grad_norm": 0.2203470766544342, "learning_rate": 1.9618256659018434e-05, "loss": 0.0405, "step": 57800 }, { "grad_norm": 0.3318914473056793, "learning_rate": 1.9601844092290257e-05, "loss": 0.0257, "step": 57810 }, { "grad_norm": 0.32421422004699707, "learning_rate": 1.9585436719775512e-05, "loss": 0.0227, "step": 57820 }, { "grad_norm": 0.259194552898407, "learning_rate": 1.9569034544277793e-05, "loss": 0.0208, "step": 57830 }, { "grad_norm": 0.28431153297424316, "learning_rate": 1.9552637568599798e-05, "loss": 0.018, "step": 57840 }, { "grad_norm": 0.3073532283306122, "learning_rate": 1.953624579554327e-05, "loss": 0.0203, "step": 57850 }, { "grad_norm": 0.47262269258499146, "learning_rate": 1.951985922790918e-05, "loss": 0.0303, "step": 57860 }, { "grad_norm": 0.20651724934577942, "learning_rate": 1.9503477868497505e-05, "loss": 0.0179, "step": 57870 }, { "grad_norm": 0.2949320673942566, "learning_rate": 1.9487101720107375e-05, "loss": 0.0273, "step": 57880 }, { "grad_norm": 0.3324459195137024, "learning_rate": 1.9470730785537032e-05, "loss": 0.024, "step": 57890 }, { "grad_norm": 0.3667524456977844, "learning_rate": 1.9454365067583823e-05, "loss": 0.0237, "step": 57900 }, { "grad_norm": 0.2909926176071167, "learning_rate": 1.9438004569044215e-05, "loss": 0.0203, "step": 57910 }, { "grad_norm": 0.18367543816566467, "learning_rate": 1.9421649292713724e-05, "loss": 0.0346, "step": 57920 }, { "grad_norm": 0.31986984610557556, "learning_rate": 1.9405299241387076e-05, "loss": 0.0303, "step": 57930 }, { "grad_norm": 0.2450931817293167, "learning_rate": 1.9388954417858007e-05, "loss": 0.0221, "step": 57940 }, { "grad_norm": 0.3248755931854248, "learning_rate": 1.9372614824919417e-05, "loss": 0.0197, "step": 57950 }, { "grad_norm": 0.4162290394306183, "learning_rate": 1.9356280465363284e-05, "loss": 0.0237, "step": 57960 }, { "grad_norm": 0.2735581398010254, "learning_rate": 1.9339951341980723e-05, "loss": 0.0186, "step": 57970 }, { "grad_norm": 0.31403103470802307, "learning_rate": 1.9323627457561916e-05, "loss": 0.0245, "step": 57980 }, { "grad_norm": 0.6322119235992432, "learning_rate": 1.9307308814896198e-05, "loss": 0.0171, "step": 57990 }, { "grad_norm": 0.34325265884399414, "learning_rate": 1.9290995416771935e-05, "loss": 0.022, "step": 58000 }, { "grad_norm": 0.3136552572250366, "learning_rate": 1.9274687265976665e-05, "loss": 0.0205, "step": 58010 }, { "grad_norm": 0.30705970525741577, "learning_rate": 1.9258384365297e-05, "loss": 0.0246, "step": 58020 }, { "grad_norm": 0.2240619659423828, "learning_rate": 1.924208671751866e-05, "loss": 0.0188, "step": 58030 }, { "grad_norm": 0.27677589654922485, "learning_rate": 1.9225794325426492e-05, "loss": 0.0259, "step": 58040 }, { "grad_norm": 0.2025829702615738, "learning_rate": 1.920950719180436e-05, "loss": 0.0237, "step": 58050 }, { "grad_norm": 0.1876946985721588, "learning_rate": 1.919322531943536e-05, "loss": 0.018, "step": 58060 }, { "grad_norm": 0.7559908032417297, "learning_rate": 1.917694871110157e-05, "loss": 0.0253, "step": 58070 }, { "grad_norm": 0.3666146993637085, "learning_rate": 1.9160677369584234e-05, "loss": 0.0217, "step": 58080 }, { "grad_norm": 0.41947850584983826, "learning_rate": 1.9144411297663694e-05, "loss": 0.0201, "step": 58090 }, { "grad_norm": 0.20527486503124237, "learning_rate": 1.9128150498119328e-05, "loss": 0.0218, "step": 58100 }, { "grad_norm": 0.4453205168247223, "learning_rate": 1.9111894973729726e-05, "loss": 0.0225, "step": 58110 }, { "grad_norm": 0.5193272829055786, "learning_rate": 1.9095644727272454e-05, "loss": 0.0251, "step": 58120 }, { "grad_norm": 0.2942920923233032, "learning_rate": 1.907939976152429e-05, "loss": 0.0251, "step": 58130 }, { "grad_norm": 0.28420743346214294, "learning_rate": 1.906316007926101e-05, "loss": 0.0273, "step": 58140 }, { "grad_norm": 0.26653003692626953, "learning_rate": 1.904692568325755e-05, "loss": 0.0198, "step": 58150 }, { "grad_norm": 0.312002956867218, "learning_rate": 1.9030696576287925e-05, "loss": 0.0218, "step": 58160 }, { "grad_norm": 0.37589678168296814, "learning_rate": 1.9014472761125242e-05, "loss": 0.0176, "step": 58170 }, { "grad_norm": 0.21382251381874084, "learning_rate": 1.899825424054172e-05, "loss": 0.0183, "step": 58180 }, { "grad_norm": 0.41111740469932556, "learning_rate": 1.898204101730863e-05, "loss": 0.0218, "step": 58190 }, { "grad_norm": 0.20835815370082855, "learning_rate": 1.8965833094196394e-05, "loss": 0.0305, "step": 58200 }, { "grad_norm": 0.17250129580497742, "learning_rate": 1.8949630473974495e-05, "loss": 0.0193, "step": 58210 }, { "grad_norm": 0.23639339208602905, "learning_rate": 1.8933433159411517e-05, "loss": 0.0235, "step": 58220 }, { "grad_norm": 0.2135934829711914, "learning_rate": 1.891724115327514e-05, "loss": 0.02, "step": 58230 }, { "grad_norm": 0.2599526047706604, "learning_rate": 1.8901054458332152e-05, "loss": 0.018, "step": 58240 }, { "grad_norm": 0.2951146960258484, "learning_rate": 1.8884873077348364e-05, "loss": 0.0217, "step": 58250 }, { "grad_norm": 0.22656549513339996, "learning_rate": 1.88686970130888e-05, "loss": 0.0227, "step": 58260 }, { "grad_norm": 0.2236468493938446, "learning_rate": 1.8852526268317456e-05, "loss": 0.0252, "step": 58270 }, { "grad_norm": 0.2114478498697281, "learning_rate": 1.883636084579749e-05, "loss": 0.0252, "step": 58280 }, { "grad_norm": 0.3583441972732544, "learning_rate": 1.8820200748291145e-05, "loss": 0.0232, "step": 58290 }, { "grad_norm": 0.3268442153930664, "learning_rate": 1.8804045978559686e-05, "loss": 0.0198, "step": 58300 }, { "grad_norm": 0.24488408863544464, "learning_rate": 1.8787896539363594e-05, "loss": 0.0229, "step": 58310 }, { "grad_norm": 0.2486395239830017, "learning_rate": 1.8771752433462298e-05, "loss": 0.0217, "step": 58320 }, { "grad_norm": 0.4380449652671814, "learning_rate": 1.8755613663614447e-05, "loss": 0.0267, "step": 58330 }, { "grad_norm": 0.4544958770275116, "learning_rate": 1.873948023257767e-05, "loss": 0.0228, "step": 58340 }, { "grad_norm": 0.23738466203212738, "learning_rate": 1.8723352143108747e-05, "loss": 0.018, "step": 58350 }, { "grad_norm": 0.24745649099349976, "learning_rate": 1.870722939796352e-05, "loss": 0.0207, "step": 58360 }, { "grad_norm": 0.19004905223846436, "learning_rate": 1.869111199989693e-05, "loss": 0.0249, "step": 58370 }, { "grad_norm": 0.2469424307346344, "learning_rate": 1.8674999951663018e-05, "loss": 0.0263, "step": 58380 }, { "grad_norm": 0.19705557823181152, "learning_rate": 1.8658893256014857e-05, "loss": 0.0234, "step": 58390 }, { "grad_norm": 0.2883674204349518, "learning_rate": 1.8642791915704655e-05, "loss": 0.0214, "step": 58400 }, { "grad_norm": 0.2484285682439804, "learning_rate": 1.8626695933483697e-05, "loss": 0.0195, "step": 58410 }, { "grad_norm": 0.39583420753479004, "learning_rate": 1.8610605312102335e-05, "loss": 0.0311, "step": 58420 }, { "grad_norm": 0.35357728600502014, "learning_rate": 1.8594520054310032e-05, "loss": 0.0315, "step": 58430 }, { "grad_norm": 0.40005260705947876, "learning_rate": 1.8578440162855326e-05, "loss": 0.0172, "step": 58440 }, { "grad_norm": 0.26163315773010254, "learning_rate": 1.8562365640485774e-05, "loss": 0.0236, "step": 58450 }, { "grad_norm": 0.14167499542236328, "learning_rate": 1.854629648994815e-05, "loss": 0.0168, "step": 58460 }, { "grad_norm": 0.22720055282115936, "learning_rate": 1.8530232713988183e-05, "loss": 0.0216, "step": 58470 }, { "grad_norm": 0.3282960057258606, "learning_rate": 1.851417431535074e-05, "loss": 0.0219, "step": 58480 }, { "grad_norm": 0.2032632678747177, "learning_rate": 1.8498121296779785e-05, "loss": 0.0324, "step": 58490 }, { "grad_norm": 0.269883394241333, "learning_rate": 1.848207366101829e-05, "loss": 0.0274, "step": 58500 }, { "grad_norm": 0.2689962685108185, "learning_rate": 1.8466031410808422e-05, "loss": 0.026, "step": 58510 }, { "grad_norm": 0.45279932022094727, "learning_rate": 1.844999454889129e-05, "loss": 0.0265, "step": 58520 }, { "grad_norm": 0.3941851556301117, "learning_rate": 1.8433963078007237e-05, "loss": 0.0262, "step": 58530 }, { "grad_norm": 0.3006924092769623, "learning_rate": 1.841793700089554e-05, "loss": 0.0215, "step": 58540 }, { "grad_norm": 0.4264443814754486, "learning_rate": 1.8401916320294644e-05, "loss": 0.0218, "step": 58550 }, { "grad_norm": 0.39163273572921753, "learning_rate": 1.838590103894205e-05, "loss": 0.0228, "step": 58560 }, { "grad_norm": 0.28724274039268494, "learning_rate": 1.8369891159574292e-05, "loss": 0.0231, "step": 58570 }, { "grad_norm": 0.22648736834526062, "learning_rate": 1.835388668492708e-05, "loss": 0.0222, "step": 58580 }, { "grad_norm": 0.2608073055744171, "learning_rate": 1.8337887617735095e-05, "loss": 0.0207, "step": 58590 }, { "grad_norm": 0.38599497079849243, "learning_rate": 1.8321893960732157e-05, "loss": 0.0179, "step": 58600 }, { "grad_norm": 0.5275794863700867, "learning_rate": 1.8305905716651138e-05, "loss": 0.0217, "step": 58610 }, { "grad_norm": 0.8367440700531006, "learning_rate": 1.8289922888224e-05, "loss": 0.0234, "step": 58620 }, { "grad_norm": 0.4115997850894928, "learning_rate": 1.8273945478181765e-05, "loss": 0.027, "step": 58630 }, { "grad_norm": 0.28531280159950256, "learning_rate": 1.8257973489254558e-05, "loss": 0.0167, "step": 58640 }, { "grad_norm": 0.18864457309246063, "learning_rate": 1.824200692417152e-05, "loss": 0.0233, "step": 58650 }, { "grad_norm": 0.2089703232049942, "learning_rate": 1.8226045785660912e-05, "loss": 0.0225, "step": 58660 }, { "grad_norm": 0.39285480976104736, "learning_rate": 1.821009007645006e-05, "loss": 0.0258, "step": 58670 }, { "grad_norm": 0.2634750008583069, "learning_rate": 1.8194139799265357e-05, "loss": 0.0235, "step": 58680 }, { "grad_norm": 0.28986936807632446, "learning_rate": 1.8178194956832295e-05, "loss": 0.0221, "step": 58690 }, { "grad_norm": 0.23727154731750488, "learning_rate": 1.8162255551875346e-05, "loss": 0.0208, "step": 58700 }, { "grad_norm": 0.21130846440792084, "learning_rate": 1.8146321587118194e-05, "loss": 0.0205, "step": 58710 }, { "grad_norm": 0.20927461981773376, "learning_rate": 1.8130393065283448e-05, "loss": 0.0186, "step": 58720 }, { "grad_norm": 0.26283687353134155, "learning_rate": 1.8114469989092925e-05, "loss": 0.0227, "step": 58730 }, { "grad_norm": 0.3492216169834137, "learning_rate": 1.8098552361267397e-05, "loss": 0.0174, "step": 58740 }, { "grad_norm": 0.5085815191268921, "learning_rate": 1.8082640184526763e-05, "loss": 0.0202, "step": 58750 }, { "grad_norm": 0.24504348635673523, "learning_rate": 1.806673346158999e-05, "loss": 0.0247, "step": 58760 }, { "grad_norm": 0.27161625027656555, "learning_rate": 1.8050832195175067e-05, "loss": 0.0172, "step": 58770 }, { "grad_norm": 0.41222918033599854, "learning_rate": 1.8034936387999136e-05, "loss": 0.0267, "step": 58780 }, { "grad_norm": 0.3705587685108185, "learning_rate": 1.8019046042778315e-05, "loss": 0.0234, "step": 58790 }, { "grad_norm": 0.2510053813457489, "learning_rate": 1.800316116222785e-05, "loss": 0.021, "step": 58800 }, { "grad_norm": 0.20510363578796387, "learning_rate": 1.7987281749062018e-05, "loss": 0.0465, "step": 58810 }, { "grad_norm": 0.24045361578464508, "learning_rate": 1.7971407805994195e-05, "loss": 0.0286, "step": 58820 }, { "grad_norm": 0.204306498169899, "learning_rate": 1.7955539335736787e-05, "loss": 0.0211, "step": 58830 }, { "grad_norm": 0.2386288344860077, "learning_rate": 1.7939676341001304e-05, "loss": 0.0245, "step": 58840 }, { "grad_norm": 0.15291234850883484, "learning_rate": 1.7923818824498275e-05, "loss": 0.0175, "step": 58850 }, { "grad_norm": 0.33831003308296204, "learning_rate": 1.7907966788937315e-05, "loss": 0.0258, "step": 58860 }, { "grad_norm": 0.25642886757850647, "learning_rate": 1.7892120237027116e-05, "loss": 0.0266, "step": 58870 }, { "grad_norm": 0.23636388778686523, "learning_rate": 1.7876279171475413e-05, "loss": 0.025, "step": 58880 }, { "grad_norm": 0.5357676148414612, "learning_rate": 1.7860443594989028e-05, "loss": 0.0201, "step": 58890 }, { "grad_norm": 2.913088321685791, "learning_rate": 1.784461351027379e-05, "loss": 0.0213, "step": 58900 }, { "grad_norm": 0.3631083071231842, "learning_rate": 1.7828788920034677e-05, "loss": 0.0305, "step": 58910 }, { "grad_norm": 0.30259355902671814, "learning_rate": 1.7812969826975623e-05, "loss": 0.0239, "step": 58920 }, { "grad_norm": 0.3061559200286865, "learning_rate": 1.7797156233799738e-05, "loss": 0.0267, "step": 58930 }, { "grad_norm": 0.21797965466976166, "learning_rate": 1.7781348143209094e-05, "loss": 0.0224, "step": 58940 }, { "grad_norm": 0.2818240821361542, "learning_rate": 1.7765545557904862e-05, "loss": 0.0378, "step": 58950 }, { "grad_norm": 0.39042189717292786, "learning_rate": 1.7749748480587302e-05, "loss": 0.0277, "step": 58960 }, { "grad_norm": 0.26675844192504883, "learning_rate": 1.773395691395564e-05, "loss": 0.0214, "step": 58970 }, { "grad_norm": 0.2683558166027069, "learning_rate": 1.7718170860708305e-05, "loss": 0.0161, "step": 58980 }, { "grad_norm": 0.3403158485889435, "learning_rate": 1.770239032354264e-05, "loss": 0.0229, "step": 58990 }, { "grad_norm": 0.40156978368759155, "learning_rate": 1.7686615305155124e-05, "loss": 0.0279, "step": 59000 }, { "grad_norm": 0.25064337253570557, "learning_rate": 1.767084580824128e-05, "loss": 0.0187, "step": 59010 }, { "grad_norm": 0.5429121255874634, "learning_rate": 1.765508183549569e-05, "loss": 0.0305, "step": 59020 }, { "grad_norm": 0.3532862067222595, "learning_rate": 1.763932338961199e-05, "loss": 0.0245, "step": 59030 }, { "grad_norm": 0.2390071004629135, "learning_rate": 1.762357047328284e-05, "loss": 0.024, "step": 59040 }, { "grad_norm": 0.3142307698726654, "learning_rate": 1.7607823089199997e-05, "loss": 0.0267, "step": 59050 }, { "grad_norm": 0.1328781247138977, "learning_rate": 1.7592081240054265e-05, "loss": 0.0208, "step": 59060 }, { "grad_norm": 0.22713878750801086, "learning_rate": 1.7576344928535488e-05, "loss": 0.0238, "step": 59070 }, { "grad_norm": 0.5998172760009766, "learning_rate": 1.7560614157332573e-05, "loss": 0.0271, "step": 59080 }, { "grad_norm": 0.3117187023162842, "learning_rate": 1.7544888929133495e-05, "loss": 0.0236, "step": 59090 }, { "grad_norm": 0.30316439270973206, "learning_rate": 1.752916924662522e-05, "loss": 0.0246, "step": 59100 }, { "grad_norm": 0.3053973317146301, "learning_rate": 1.751345511249387e-05, "loss": 0.018, "step": 59110 }, { "grad_norm": 0.3572933077812195, "learning_rate": 1.749774652942452e-05, "loss": 0.0188, "step": 59120 }, { "grad_norm": 0.24142980575561523, "learning_rate": 1.748204350010135e-05, "loss": 0.0274, "step": 59130 }, { "grad_norm": 0.3039323687553406, "learning_rate": 1.7466346027207574e-05, "loss": 0.0206, "step": 59140 }, { "grad_norm": 0.27168044447898865, "learning_rate": 1.745065411342547e-05, "loss": 0.0228, "step": 59150 }, { "grad_norm": 0.4875268042087555, "learning_rate": 1.7434967761436366e-05, "loss": 0.0301, "step": 59160 }, { "grad_norm": 0.25945010781288147, "learning_rate": 1.741928697392058e-05, "loss": 0.025, "step": 59170 }, { "grad_norm": 0.43185165524482727, "learning_rate": 1.7403611753557597e-05, "loss": 0.0218, "step": 59180 }, { "grad_norm": 0.22320108115673065, "learning_rate": 1.738794210302584e-05, "loss": 0.0164, "step": 59190 }, { "grad_norm": 0.7145889401435852, "learning_rate": 1.7372278025002837e-05, "loss": 0.0197, "step": 59200 }, { "grad_norm": 0.22498804330825806, "learning_rate": 1.7356619522165164e-05, "loss": 0.0226, "step": 59210 }, { "grad_norm": 0.43182238936424255, "learning_rate": 1.7340966597188378e-05, "loss": 0.0189, "step": 59220 }, { "grad_norm": 0.3351042866706848, "learning_rate": 1.732531925274722e-05, "loss": 0.0189, "step": 59230 }, { "grad_norm": 0.2698439061641693, "learning_rate": 1.7309677491515318e-05, "loss": 0.0195, "step": 59240 }, { "grad_norm": 0.26990190148353577, "learning_rate": 1.7294041316165455e-05, "loss": 0.0177, "step": 59250 }, { "grad_norm": 0.30642855167388916, "learning_rate": 1.727841072936942e-05, "loss": 0.0232, "step": 59260 }, { "grad_norm": 0.29036158323287964, "learning_rate": 1.7262785733798058e-05, "loss": 0.0215, "step": 59270 }, { "grad_norm": 0.2910021245479584, "learning_rate": 1.7247166332121246e-05, "loss": 0.0255, "step": 59280 }, { "grad_norm": 0.4100711941719055, "learning_rate": 1.7231552527007933e-05, "loss": 0.02, "step": 59290 }, { "grad_norm": 0.6331471800804138, "learning_rate": 1.721594432112606e-05, "loss": 0.0219, "step": 59300 }, { "grad_norm": 0.2613619267940521, "learning_rate": 1.7200341717142655e-05, "loss": 0.0215, "step": 59310 }, { "grad_norm": 0.40190818905830383, "learning_rate": 1.718474471772377e-05, "loss": 0.0235, "step": 59320 }, { "grad_norm": 0.34432482719421387, "learning_rate": 1.7169153325534528e-05, "loss": 0.0192, "step": 59330 }, { "grad_norm": 0.1900249421596527, "learning_rate": 1.7153567543239045e-05, "loss": 0.0219, "step": 59340 }, { "grad_norm": 0.2401784509420395, "learning_rate": 1.7137987373500525e-05, "loss": 0.0171, "step": 59350 }, { "grad_norm": 0.2185879498720169, "learning_rate": 1.7122412818981198e-05, "loss": 0.0257, "step": 59360 }, { "grad_norm": 0.5357773303985596, "learning_rate": 1.7106843882342283e-05, "loss": 0.0181, "step": 59370 }, { "grad_norm": 0.2928759455680847, "learning_rate": 1.709128056624415e-05, "loss": 0.0152, "step": 59380 }, { "grad_norm": 0.3968493938446045, "learning_rate": 1.70757228733461e-05, "loss": 0.0168, "step": 59390 }, { "grad_norm": 0.1748322993516922, "learning_rate": 1.706017080630653e-05, "loss": 0.0239, "step": 59400 }, { "grad_norm": 0.6851019859313965, "learning_rate": 1.7044624367782873e-05, "loss": 0.0188, "step": 59410 }, { "grad_norm": 0.27231329679489136, "learning_rate": 1.7029083560431553e-05, "loss": 0.0196, "step": 59420 }, { "grad_norm": 0.2701322138309479, "learning_rate": 1.701354838690813e-05, "loss": 0.0193, "step": 59430 }, { "grad_norm": 0.28057652711868286, "learning_rate": 1.6998018849867087e-05, "loss": 0.0173, "step": 59440 }, { "grad_norm": 0.17143909633159637, "learning_rate": 1.698249495196202e-05, "loss": 0.0173, "step": 59450 }, { "grad_norm": 0.2338363230228424, "learning_rate": 1.6966976695845528e-05, "loss": 0.0235, "step": 59460 }, { "grad_norm": 0.2148517221212387, "learning_rate": 1.6951464084169268e-05, "loss": 0.0187, "step": 59470 }, { "grad_norm": 0.3834516704082489, "learning_rate": 1.6935957119583916e-05, "loss": 0.0258, "step": 59480 }, { "grad_norm": 0.24264416098594666, "learning_rate": 1.6920455804739205e-05, "loss": 0.0215, "step": 59490 }, { "grad_norm": 0.27898481488227844, "learning_rate": 1.6904960142283855e-05, "loss": 0.0262, "step": 59500 }, { "grad_norm": 0.14788547158241272, "learning_rate": 1.6889470134865666e-05, "loss": 0.0206, "step": 59510 }, { "grad_norm": 0.2546187937259674, "learning_rate": 1.687398578513145e-05, "loss": 0.024, "step": 59520 }, { "grad_norm": 0.30559036135673523, "learning_rate": 1.6858507095727066e-05, "loss": 0.0184, "step": 59530 }, { "grad_norm": 0.4031335413455963, "learning_rate": 1.6843034069297403e-05, "loss": 0.0199, "step": 59540 }, { "grad_norm": 0.38690388202667236, "learning_rate": 1.682756670848637e-05, "loss": 0.0289, "step": 59550 }, { "grad_norm": 0.22996841371059418, "learning_rate": 1.6812105015936937e-05, "loss": 0.0127, "step": 59560 }, { "grad_norm": 0.38250619173049927, "learning_rate": 1.6796648994291027e-05, "loss": 0.0269, "step": 59570 }, { "grad_norm": 0.23928916454315186, "learning_rate": 1.678119864618973e-05, "loss": 0.016, "step": 59580 }, { "grad_norm": 0.2293129414319992, "learning_rate": 1.6765753974273023e-05, "loss": 0.0166, "step": 59590 }, { "grad_norm": 0.3345593214035034, "learning_rate": 1.675031498118001e-05, "loss": 0.0195, "step": 59600 }, { "grad_norm": 0.2754210829734802, "learning_rate": 1.6734881669548804e-05, "loss": 0.0181, "step": 59610 }, { "grad_norm": 0.5663847923278809, "learning_rate": 1.671945404201648e-05, "loss": 0.023, "step": 59620 }, { "grad_norm": 0.2812214195728302, "learning_rate": 1.670403210121927e-05, "loss": 0.0206, "step": 59630 }, { "grad_norm": 0.3267306387424469, "learning_rate": 1.6688615849792312e-05, "loss": 0.0193, "step": 59640 }, { "grad_norm": 0.37409546971321106, "learning_rate": 1.6673205290369832e-05, "loss": 0.0275, "step": 59650 }, { "grad_norm": 0.3172617554664612, "learning_rate": 1.6657800425585086e-05, "loss": 0.0234, "step": 59660 }, { "grad_norm": 0.3319850564002991, "learning_rate": 1.6642401258070327e-05, "loss": 0.0351, "step": 59670 }, { "grad_norm": 0.3427027463912964, "learning_rate": 1.6627007790456884e-05, "loss": 0.0308, "step": 59680 }, { "grad_norm": 0.2628306448459625, "learning_rate": 1.6611620025375037e-05, "loss": 0.0217, "step": 59690 }, { "grad_norm": 0.22048088908195496, "learning_rate": 1.6596237965454154e-05, "loss": 0.0237, "step": 59700 }, { "grad_norm": 0.42561572790145874, "learning_rate": 1.6580861613322607e-05, "loss": 0.0227, "step": 59710 }, { "grad_norm": 0.3204241096973419, "learning_rate": 1.6565490971607796e-05, "loss": 0.0197, "step": 59720 }, { "grad_norm": 0.24785588681697845, "learning_rate": 1.655012604293614e-05, "loss": 0.0203, "step": 59730 }, { "grad_norm": 0.43279388546943665, "learning_rate": 1.653476682993309e-05, "loss": 0.0262, "step": 59740 }, { "grad_norm": 0.32378169894218445, "learning_rate": 1.651941333522311e-05, "loss": 0.0203, "step": 59750 }, { "grad_norm": 0.3165704905986786, "learning_rate": 1.6504065561429715e-05, "loss": 0.0208, "step": 59760 }, { "grad_norm": 0.31554317474365234, "learning_rate": 1.6488723511175385e-05, "loss": 0.0207, "step": 59770 }, { "grad_norm": 0.3670714497566223, "learning_rate": 1.6473387187081668e-05, "loss": 0.0221, "step": 59780 }, { "grad_norm": 0.22459010779857635, "learning_rate": 1.6458056591769123e-05, "loss": 0.0264, "step": 59790 }, { "grad_norm": 0.37137889862060547, "learning_rate": 1.6442731727857335e-05, "loss": 0.0266, "step": 59800 }, { "grad_norm": 0.19963474571704865, "learning_rate": 1.642741259796492e-05, "loss": 0.0171, "step": 59810 }, { "grad_norm": 0.22628025710582733, "learning_rate": 1.641209920470944e-05, "loss": 0.0242, "step": 59820 }, { "grad_norm": 0.2071022093296051, "learning_rate": 1.639679155070762e-05, "loss": 0.0165, "step": 59830 }, { "grad_norm": 0.22801129519939423, "learning_rate": 1.6381489638575048e-05, "loss": 0.0167, "step": 59840 }, { "grad_norm": 0.2478765994310379, "learning_rate": 1.636619347092643e-05, "loss": 0.0183, "step": 59850 }, { "grad_norm": 0.30967292189598083, "learning_rate": 1.6350903050375476e-05, "loss": 0.0193, "step": 59860 }, { "grad_norm": 0.30500292778015137, "learning_rate": 1.6335618379534856e-05, "loss": 0.0203, "step": 59870 }, { "grad_norm": 0.21465206146240234, "learning_rate": 1.6320339461016364e-05, "loss": 0.0199, "step": 59880 }, { "grad_norm": 0.1651068925857544, "learning_rate": 1.6305066297430687e-05, "loss": 0.0165, "step": 59890 }, { "grad_norm": 0.1446124017238617, "learning_rate": 1.6289798891387654e-05, "loss": 0.0229, "step": 59900 }, { "grad_norm": 0.15961812436580658, "learning_rate": 1.6274537245495995e-05, "loss": 0.018, "step": 59910 }, { "grad_norm": 0.3979131281375885, "learning_rate": 1.6259281362363527e-05, "loss": 0.0181, "step": 59920 }, { "grad_norm": 0.3401411473751068, "learning_rate": 1.6244031244597068e-05, "loss": 0.0291, "step": 59930 }, { "grad_norm": 0.19267107546329498, "learning_rate": 1.622878689480244e-05, "loss": 0.0173, "step": 59940 }, { "grad_norm": 0.35059064626693726, "learning_rate": 1.6213548315584498e-05, "loss": 0.0297, "step": 59950 }, { "grad_norm": 0.40016505122184753, "learning_rate": 1.6198315509547074e-05, "loss": 0.0229, "step": 59960 }, { "grad_norm": 0.18307319283485413, "learning_rate": 1.6183088479293056e-05, "loss": 0.021, "step": 59970 }, { "grad_norm": 0.3060983717441559, "learning_rate": 1.6167867227424316e-05, "loss": 0.0199, "step": 59980 }, { "grad_norm": 0.22857245802879333, "learning_rate": 1.615265175654176e-05, "loss": 0.022, "step": 59990 }, { "grad_norm": 1.1168103218078613, "learning_rate": 1.613744206924529e-05, "loss": 0.0231, "step": 60000 }, { "grad_norm": 0.19301903247833252, "learning_rate": 1.6122238168133845e-05, "loss": 0.0218, "step": 60010 }, { "grad_norm": 0.3951355516910553, "learning_rate": 1.6107040055805305e-05, "loss": 0.0205, "step": 60020 }, { "grad_norm": 0.3297105133533478, "learning_rate": 1.6091847734856673e-05, "loss": 0.0207, "step": 60030 }, { "grad_norm": 0.3110640048980713, "learning_rate": 1.6076661207883865e-05, "loss": 0.018, "step": 60040 }, { "grad_norm": 0.27184486389160156, "learning_rate": 1.6061480477481848e-05, "loss": 0.0207, "step": 60050 }, { "grad_norm": 0.28404754400253296, "learning_rate": 1.604630554624461e-05, "loss": 0.018, "step": 60060 }, { "grad_norm": 0.3808056712150574, "learning_rate": 1.603113641676509e-05, "loss": 0.0188, "step": 60070 }, { "grad_norm": 0.21259666979312897, "learning_rate": 1.6015973091635338e-05, "loss": 0.0193, "step": 60080 }, { "grad_norm": 0.23368337750434875, "learning_rate": 1.6000815573446288e-05, "loss": 0.0191, "step": 60090 }, { "grad_norm": 0.29139092564582825, "learning_rate": 1.5985663864788002e-05, "loss": 0.0313, "step": 60100 }, { "grad_norm": 0.5341275334358215, "learning_rate": 1.597051796824946e-05, "loss": 0.0218, "step": 60110 }, { "grad_norm": 0.21067936718463898, "learning_rate": 1.5955377886418682e-05, "loss": 0.023, "step": 60120 }, { "grad_norm": 0.41797488927841187, "learning_rate": 1.5940243621882704e-05, "loss": 0.0207, "step": 60130 }, { "grad_norm": 0.25004804134368896, "learning_rate": 1.5925115177227555e-05, "loss": 0.0281, "step": 60140 }, { "grad_norm": 0.33890387415885925, "learning_rate": 1.5909992555038288e-05, "loss": 0.0204, "step": 60150 }, { "grad_norm": 0.3726755678653717, "learning_rate": 1.5894875757898912e-05, "loss": 0.0297, "step": 60160 }, { "grad_norm": 0.2909717261791229, "learning_rate": 1.58797647883925e-05, "loss": 0.0199, "step": 60170 }, { "grad_norm": 0.41461947560310364, "learning_rate": 1.586465964910109e-05, "loss": 0.0216, "step": 60180 }, { "grad_norm": 0.32668957114219666, "learning_rate": 1.5849560342605734e-05, "loss": 0.0283, "step": 60190 }, { "grad_norm": 0.19835536181926727, "learning_rate": 1.583446687148651e-05, "loss": 0.0161, "step": 60200 }, { "grad_norm": 0.31807947158813477, "learning_rate": 1.581937923832248e-05, "loss": 0.0182, "step": 60210 }, { "grad_norm": 0.4308936297893524, "learning_rate": 1.5804297445691663e-05, "loss": 0.022, "step": 60220 }, { "grad_norm": 0.2857956290245056, "learning_rate": 1.578922149617119e-05, "loss": 0.0261, "step": 60230 }, { "grad_norm": 0.42315202951431274, "learning_rate": 1.5774151392337084e-05, "loss": 0.0204, "step": 60240 }, { "grad_norm": 0.48957639932632446, "learning_rate": 1.575908713676442e-05, "loss": 0.0265, "step": 60250 }, { "grad_norm": 0.3197627365589142, "learning_rate": 1.5744028732027293e-05, "loss": 0.027, "step": 60260 }, { "grad_norm": 0.3873068690299988, "learning_rate": 1.5728976180698723e-05, "loss": 0.0216, "step": 60270 }, { "grad_norm": 0.24389886856079102, "learning_rate": 1.5713929485350836e-05, "loss": 0.0238, "step": 60280 }, { "grad_norm": 0.24948002398014069, "learning_rate": 1.569888864855465e-05, "loss": 0.021, "step": 60290 }, { "grad_norm": 0.23680634796619415, "learning_rate": 1.568385367288028e-05, "loss": 0.0193, "step": 60300 }, { "grad_norm": 0.5020309686660767, "learning_rate": 1.5668824560896755e-05, "loss": 0.0202, "step": 60310 }, { "grad_norm": 0.2122618854045868, "learning_rate": 1.5653801315172156e-05, "loss": 0.0236, "step": 60320 }, { "grad_norm": 0.3121502995491028, "learning_rate": 1.5638783938273554e-05, "loss": 0.0163, "step": 60330 }, { "grad_norm": 0.31469354033470154, "learning_rate": 1.5623772432766966e-05, "loss": 0.0302, "step": 60340 }, { "grad_norm": 0.43744927644729614, "learning_rate": 1.5608766801217507e-05, "loss": 0.0225, "step": 60350 }, { "grad_norm": 0.5173324942588806, "learning_rate": 1.5593767046189184e-05, "loss": 0.0194, "step": 60360 }, { "grad_norm": 0.23730355501174927, "learning_rate": 1.5578773170245064e-05, "loss": 0.0205, "step": 60370 }, { "grad_norm": 0.2515738010406494, "learning_rate": 1.5563785175947182e-05, "loss": 0.0259, "step": 60380 }, { "grad_norm": 0.3309130072593689, "learning_rate": 1.5548803065856582e-05, "loss": 0.0206, "step": 60390 }, { "grad_norm": 0.2946045994758606, "learning_rate": 1.55338268425333e-05, "loss": 0.0209, "step": 60400 }, { "grad_norm": 0.15185818076133728, "learning_rate": 1.5518856508536373e-05, "loss": 0.0179, "step": 60410 }, { "grad_norm": 0.5878046154975891, "learning_rate": 1.5503892066423786e-05, "loss": 0.021, "step": 60420 }, { "grad_norm": 0.443411648273468, "learning_rate": 1.548893351875258e-05, "loss": 0.0351, "step": 60430 }, { "grad_norm": 0.3811165392398834, "learning_rate": 1.547398086807875e-05, "loss": 0.022, "step": 60440 }, { "grad_norm": 0.4785853624343872, "learning_rate": 1.5459034116957305e-05, "loss": 0.0318, "step": 60450 }, { "grad_norm": 0.7714568972587585, "learning_rate": 1.544409326794225e-05, "loss": 0.0297, "step": 60460 }, { "grad_norm": 0.2979590892791748, "learning_rate": 1.542915832358651e-05, "loss": 0.0211, "step": 60470 }, { "grad_norm": 0.25660327076911926, "learning_rate": 1.541422928644214e-05, "loss": 0.0216, "step": 60480 }, { "grad_norm": 0.46684595942497253, "learning_rate": 1.5399306159060024e-05, "loss": 0.0307, "step": 60490 }, { "grad_norm": 0.21356996893882751, "learning_rate": 1.5384388943990185e-05, "loss": 0.0194, "step": 60500 }, { "grad_norm": 0.24785643815994263, "learning_rate": 1.5369477643781526e-05, "loss": 0.0185, "step": 60510 }, { "grad_norm": 0.33605337142944336, "learning_rate": 1.5354572260981985e-05, "loss": 0.0241, "step": 60520 }, { "grad_norm": 0.3623518943786621, "learning_rate": 1.533967279813851e-05, "loss": 0.0201, "step": 60530 }, { "grad_norm": 0.428684800863266, "learning_rate": 1.5324779257796956e-05, "loss": 0.0209, "step": 60540 }, { "grad_norm": 0.14774203300476074, "learning_rate": 1.5309891642502293e-05, "loss": 0.0195, "step": 60550 }, { "grad_norm": 0.21446986496448517, "learning_rate": 1.5295009954798357e-05, "loss": 0.0181, "step": 60560 }, { "grad_norm": 0.3674318194389343, "learning_rate": 1.5280134197228036e-05, "loss": 0.0223, "step": 60570 }, { "grad_norm": 0.2894720137119293, "learning_rate": 1.526526437233319e-05, "loss": 0.0234, "step": 60580 }, { "grad_norm": 0.17637352645397186, "learning_rate": 1.525040048265466e-05, "loss": 0.0239, "step": 60590 }, { "grad_norm": 0.20280411839485168, "learning_rate": 1.5235542530732288e-05, "loss": 0.0169, "step": 60600 }, { "grad_norm": 0.3363274931907654, "learning_rate": 1.5220690519104901e-05, "loss": 0.0186, "step": 60610 }, { "grad_norm": 0.41563865542411804, "learning_rate": 1.5205844450310275e-05, "loss": 0.0228, "step": 60620 }, { "grad_norm": 0.2161657214164734, "learning_rate": 1.519100432688521e-05, "loss": 0.0213, "step": 60630 }, { "grad_norm": 0.22475609183311462, "learning_rate": 1.5176170151365476e-05, "loss": 0.0204, "step": 60640 }, { "grad_norm": 0.4262295663356781, "learning_rate": 1.5161341926285833e-05, "loss": 0.0198, "step": 60650 }, { "grad_norm": 0.4362201392650604, "learning_rate": 1.5146519654180025e-05, "loss": 0.0154, "step": 60660 }, { "grad_norm": 0.43313735723495483, "learning_rate": 1.5131703337580739e-05, "loss": 0.0214, "step": 60670 }, { "grad_norm": 0.27153101563453674, "learning_rate": 1.5116892979019731e-05, "loss": 0.0248, "step": 60680 }, { "grad_norm": 0.262256383895874, "learning_rate": 1.5102088581027623e-05, "loss": 0.0196, "step": 60690 }, { "grad_norm": 0.23672649264335632, "learning_rate": 1.5087290146134154e-05, "loss": 0.0175, "step": 60700 }, { "grad_norm": 0.32411906123161316, "learning_rate": 1.5072497676867915e-05, "loss": 0.0245, "step": 60710 }, { "grad_norm": 0.2517481744289398, "learning_rate": 1.505771117575655e-05, "loss": 0.0251, "step": 60720 }, { "grad_norm": 0.3208097219467163, "learning_rate": 1.5042930645326691e-05, "loss": 0.0218, "step": 60730 }, { "grad_norm": 0.27281254529953003, "learning_rate": 1.5028156088103879e-05, "loss": 0.0194, "step": 60740 }, { "grad_norm": 0.3782269060611725, "learning_rate": 1.5013387506612735e-05, "loss": 0.0204, "step": 60750 }, { "grad_norm": 0.3835781514644623, "learning_rate": 1.4998624903376767e-05, "loss": 0.0203, "step": 60760 }, { "grad_norm": 0.5478364825248718, "learning_rate": 1.4983868280918517e-05, "loss": 0.0197, "step": 60770 }, { "grad_norm": 0.44626033306121826, "learning_rate": 1.4969117641759478e-05, "loss": 0.0255, "step": 60780 }, { "grad_norm": 0.4920036494731903, "learning_rate": 1.495437298842014e-05, "loss": 0.0263, "step": 60790 }, { "grad_norm": 0.26280298829078674, "learning_rate": 1.4939634323419976e-05, "loss": 0.0233, "step": 60800 }, { "grad_norm": 0.35081931948661804, "learning_rate": 1.4924901649277384e-05, "loss": 0.0241, "step": 60810 }, { "grad_norm": 0.26443445682525635, "learning_rate": 1.4910174968509793e-05, "loss": 0.0181, "step": 60820 }, { "grad_norm": 0.7514113783836365, "learning_rate": 1.4895454283633598e-05, "loss": 0.0202, "step": 60830 }, { "grad_norm": 0.6075317859649658, "learning_rate": 1.488073959716415e-05, "loss": 0.0218, "step": 60840 }, { "grad_norm": 0.24004250764846802, "learning_rate": 1.4866030911615791e-05, "loss": 0.0214, "step": 60850 }, { "grad_norm": 0.7084845304489136, "learning_rate": 1.4851328229501849e-05, "loss": 0.0162, "step": 60860 }, { "grad_norm": 0.20839907228946686, "learning_rate": 1.4836631553334562e-05, "loss": 0.0211, "step": 60870 }, { "grad_norm": 0.33286622166633606, "learning_rate": 1.4821940885625251e-05, "loss": 0.0195, "step": 60880 }, { "grad_norm": 0.27747026085853577, "learning_rate": 1.4807256228884109e-05, "loss": 0.0194, "step": 60890 }, { "grad_norm": 0.44528430700302124, "learning_rate": 1.4792577585620353e-05, "loss": 0.0211, "step": 60900 }, { "grad_norm": 0.2286888211965561, "learning_rate": 1.4777904958342164e-05, "loss": 0.0273, "step": 60910 }, { "grad_norm": 0.30335763096809387, "learning_rate": 1.4763238349556691e-05, "loss": 0.0209, "step": 60920 }, { "grad_norm": 0.36597350239753723, "learning_rate": 1.4748577761770072e-05, "loss": 0.0248, "step": 60930 }, { "grad_norm": 0.5789167881011963, "learning_rate": 1.4733923197487354e-05, "loss": 0.0188, "step": 60940 }, { "grad_norm": 0.39572590589523315, "learning_rate": 1.4719274659212662e-05, "loss": 0.0183, "step": 60950 }, { "grad_norm": 0.2734399437904358, "learning_rate": 1.4704632149448983e-05, "loss": 0.0234, "step": 60960 }, { "grad_norm": 0.3159467279911041, "learning_rate": 1.4689995670698343e-05, "loss": 0.0244, "step": 60970 }, { "grad_norm": 0.24237032234668732, "learning_rate": 1.4675365225461728e-05, "loss": 0.0232, "step": 60980 }, { "grad_norm": 0.3387330174446106, "learning_rate": 1.4660740816239032e-05, "loss": 0.0208, "step": 60990 }, { "grad_norm": 0.2974424362182617, "learning_rate": 1.464612244552923e-05, "loss": 0.0208, "step": 61000 }, { "grad_norm": 0.4633997678756714, "learning_rate": 1.4631510115830161e-05, "loss": 0.0198, "step": 61010 }, { "grad_norm": 0.3608981668949127, "learning_rate": 1.4616903829638679e-05, "loss": 0.0174, "step": 61020 }, { "grad_norm": 0.3432910740375519, "learning_rate": 1.46023035894506e-05, "loss": 0.0218, "step": 61030 }, { "grad_norm": 0.5515931844711304, "learning_rate": 1.4587709397760713e-05, "loss": 0.0233, "step": 61040 }, { "grad_norm": 0.27817586064338684, "learning_rate": 1.4573121257062755e-05, "loss": 0.0213, "step": 61050 }, { "grad_norm": 0.4459230303764343, "learning_rate": 1.4558539169849472e-05, "loss": 0.0252, "step": 61060 }, { "grad_norm": 0.23279081284999847, "learning_rate": 1.4543963138612499e-05, "loss": 0.0172, "step": 61070 }, { "grad_norm": 0.27484700083732605, "learning_rate": 1.4529393165842498e-05, "loss": 0.0178, "step": 61080 }, { "grad_norm": 0.46381717920303345, "learning_rate": 1.4514829254029084e-05, "loss": 0.0208, "step": 61090 }, { "grad_norm": 0.5326412320137024, "learning_rate": 1.450027140566083e-05, "loss": 0.0226, "step": 61100 }, { "grad_norm": 0.18862612545490265, "learning_rate": 1.4485719623225268e-05, "loss": 0.0201, "step": 61110 }, { "grad_norm": 0.20897312462329865, "learning_rate": 1.4471173909208912e-05, "loss": 0.0197, "step": 61120 }, { "grad_norm": 0.24043036997318268, "learning_rate": 1.4456634266097236e-05, "loss": 0.0199, "step": 61130 }, { "grad_norm": 0.20990337431430817, "learning_rate": 1.444210069637461e-05, "loss": 0.0295, "step": 61140 }, { "grad_norm": 0.5095260143280029, "learning_rate": 1.4427573202524502e-05, "loss": 0.0238, "step": 61150 }, { "grad_norm": 0.2696464955806732, "learning_rate": 1.4413051787029208e-05, "loss": 0.0266, "step": 61160 }, { "grad_norm": 0.18112309277057648, "learning_rate": 1.439853645237006e-05, "loss": 0.0186, "step": 61170 }, { "grad_norm": 0.2657583951950073, "learning_rate": 1.4384027201027344e-05, "loss": 0.0202, "step": 61180 }, { "grad_norm": 0.2883061170578003, "learning_rate": 1.4369524035480253e-05, "loss": 0.0215, "step": 61190 }, { "grad_norm": 0.2535642087459564, "learning_rate": 1.4355026958207035e-05, "loss": 0.0176, "step": 61200 }, { "grad_norm": 0.6198858022689819, "learning_rate": 1.434053597168481e-05, "loss": 0.0199, "step": 61210 }, { "grad_norm": 0.4000510573387146, "learning_rate": 1.432605107838969e-05, "loss": 0.0213, "step": 61220 }, { "grad_norm": 0.3243001699447632, "learning_rate": 1.4311572280796759e-05, "loss": 0.0209, "step": 61230 }, { "grad_norm": 0.23659153282642365, "learning_rate": 1.4297099581380047e-05, "loss": 0.0262, "step": 61240 }, { "grad_norm": 0.3204019367694855, "learning_rate": 1.4282632982612538e-05, "loss": 0.0206, "step": 61250 }, { "grad_norm": 0.21598279476165771, "learning_rate": 1.4268172486966197e-05, "loss": 0.0297, "step": 61260 }, { "grad_norm": 0.19065946340560913, "learning_rate": 1.4253718096911894e-05, "loss": 0.0198, "step": 61270 }, { "grad_norm": 0.21512341499328613, "learning_rate": 1.4239269814919504e-05, "loss": 0.0303, "step": 61280 }, { "grad_norm": 0.3097594976425171, "learning_rate": 1.4224827643457843e-05, "loss": 0.0232, "step": 61290 }, { "grad_norm": 0.2136273980140686, "learning_rate": 1.4210391584994686e-05, "loss": 0.0208, "step": 61300 }, { "grad_norm": 0.4163212180137634, "learning_rate": 1.4195961641996763e-05, "loss": 0.0232, "step": 61310 }, { "grad_norm": 0.4752059280872345, "learning_rate": 1.4181537816929751e-05, "loss": 0.0202, "step": 61320 }, { "grad_norm": 0.20296253263950348, "learning_rate": 1.4167120112258303e-05, "loss": 0.0292, "step": 61330 }, { "grad_norm": 0.39341193437576294, "learning_rate": 1.4152708530445963e-05, "loss": 0.0195, "step": 61340 }, { "grad_norm": 0.5466586351394653, "learning_rate": 1.4138303073955344e-05, "loss": 0.0196, "step": 61350 }, { "grad_norm": 0.3776569664478302, "learning_rate": 1.4123903745247897e-05, "loss": 0.0247, "step": 61360 }, { "grad_norm": 0.26102545857429504, "learning_rate": 1.4109510546784082e-05, "loss": 0.0166, "step": 61370 }, { "grad_norm": 0.4513031840324402, "learning_rate": 1.4095123481023325e-05, "loss": 0.0301, "step": 61380 }, { "grad_norm": 0.27691182494163513, "learning_rate": 1.4080742550423936e-05, "loss": 0.0216, "step": 61390 }, { "grad_norm": 0.17895591259002686, "learning_rate": 1.4066367757443278e-05, "loss": 0.0215, "step": 61400 }, { "grad_norm": 0.27117329835891724, "learning_rate": 1.4051999104537572e-05, "loss": 0.0192, "step": 61410 }, { "grad_norm": 0.28775209188461304, "learning_rate": 1.4037636594162034e-05, "loss": 0.024, "step": 61420 }, { "grad_norm": 0.24158093333244324, "learning_rate": 1.402328022877083e-05, "loss": 0.0194, "step": 61430 }, { "grad_norm": 0.20417042076587677, "learning_rate": 1.4008930010817073e-05, "loss": 0.0194, "step": 61440 }, { "grad_norm": 0.32955843210220337, "learning_rate": 1.3994585942752831e-05, "loss": 0.0227, "step": 61450 }, { "grad_norm": 0.5456863641738892, "learning_rate": 1.398024802702909e-05, "loss": 0.0171, "step": 61460 }, { "grad_norm": 0.28228190541267395, "learning_rate": 1.3965916266095819e-05, "loss": 0.0199, "step": 61470 }, { "grad_norm": 0.24671277403831482, "learning_rate": 1.395159066240192e-05, "loss": 0.0195, "step": 61480 }, { "grad_norm": 0.3192525804042816, "learning_rate": 1.3937271218395248e-05, "loss": 0.0254, "step": 61490 }, { "grad_norm": 0.39068976044654846, "learning_rate": 1.3922957936522613e-05, "loss": 0.0258, "step": 61500 }, { "grad_norm": 0.4177992343902588, "learning_rate": 1.3908650819229758e-05, "loss": 0.023, "step": 61510 }, { "grad_norm": 0.2251422256231308, "learning_rate": 1.389434986896137e-05, "loss": 0.0212, "step": 61520 }, { "grad_norm": 0.26951655745506287, "learning_rate": 1.388005508816112e-05, "loss": 0.0195, "step": 61530 }, { "grad_norm": 0.22344709932804108, "learning_rate": 1.3865766479271557e-05, "loss": 0.024, "step": 61540 }, { "grad_norm": 0.14458350837230682, "learning_rate": 1.3851484044734225e-05, "loss": 0.0282, "step": 61550 }, { "grad_norm": 0.17967940866947174, "learning_rate": 1.3837207786989608e-05, "loss": 0.0231, "step": 61560 }, { "grad_norm": 0.5024517178535461, "learning_rate": 1.382293770847713e-05, "loss": 0.0267, "step": 61570 }, { "grad_norm": 1.2348549365997314, "learning_rate": 1.3808673811635159e-05, "loss": 0.0245, "step": 61580 }, { "grad_norm": 0.2412905991077423, "learning_rate": 1.3794416098900975e-05, "loss": 0.0311, "step": 61590 }, { "grad_norm": 0.38362917304039, "learning_rate": 1.3780164572710886e-05, "loss": 0.0306, "step": 61600 }, { "grad_norm": 0.2191459834575653, "learning_rate": 1.3765919235500035e-05, "loss": 0.0282, "step": 61610 }, { "grad_norm": 0.2773745656013489, "learning_rate": 1.375168008970259e-05, "loss": 0.0199, "step": 61620 }, { "grad_norm": 0.1806030422449112, "learning_rate": 1.3737447137751635e-05, "loss": 0.0172, "step": 61630 }, { "grad_norm": 0.27902913093566895, "learning_rate": 1.372322038207915e-05, "loss": 0.0201, "step": 61640 }, { "grad_norm": 0.3881990909576416, "learning_rate": 1.3708999825116159e-05, "loss": 0.0337, "step": 61650 }, { "grad_norm": 0.32941365242004395, "learning_rate": 1.3694785469292526e-05, "loss": 0.0228, "step": 61660 }, { "grad_norm": 0.5289288759231567, "learning_rate": 1.3680577317037101e-05, "loss": 0.0265, "step": 61670 }, { "grad_norm": 0.21097463369369507, "learning_rate": 1.3666375370777678e-05, "loss": 0.0162, "step": 61680 }, { "grad_norm": 0.25549763441085815, "learning_rate": 1.3652179632940981e-05, "loss": 0.0161, "step": 61690 }, { "grad_norm": 0.1738465428352356, "learning_rate": 1.3637990105952664e-05, "loss": 0.0231, "step": 61700 }, { "grad_norm": 0.30176252126693726, "learning_rate": 1.3623806792237337e-05, "loss": 0.021, "step": 61710 }, { "grad_norm": 0.2671346068382263, "learning_rate": 1.3609629694218551e-05, "loss": 0.0229, "step": 61720 }, { "grad_norm": 0.4851442277431488, "learning_rate": 1.3595458814318762e-05, "loss": 0.0219, "step": 61730 }, { "grad_norm": 0.2679431438446045, "learning_rate": 1.3581294154959389e-05, "loss": 0.022, "step": 61740 }, { "grad_norm": 0.30026933550834656, "learning_rate": 1.3567135718560792e-05, "loss": 0.0192, "step": 61750 }, { "grad_norm": 0.2623862326145172, "learning_rate": 1.3552983507542261e-05, "loss": 0.0201, "step": 61760 }, { "grad_norm": 0.15895219147205353, "learning_rate": 1.3538837524322023e-05, "loss": 0.0273, "step": 61770 }, { "grad_norm": 0.29712215065956116, "learning_rate": 1.3524697771317251e-05, "loss": 0.0233, "step": 61780 }, { "grad_norm": 0.224267840385437, "learning_rate": 1.3510564250943997e-05, "loss": 0.0215, "step": 61790 }, { "grad_norm": 0.24143898487091064, "learning_rate": 1.3496436965617353e-05, "loss": 0.0197, "step": 61800 }, { "grad_norm": 0.24554550647735596, "learning_rate": 1.3482315917751243e-05, "loss": 0.0239, "step": 61810 }, { "grad_norm": 0.4043256938457489, "learning_rate": 1.3468201109758582e-05, "loss": 0.0113, "step": 61820 }, { "grad_norm": 0.24027679860591888, "learning_rate": 1.3454092544051222e-05, "loss": 0.0214, "step": 61830 }, { "grad_norm": 0.29198211431503296, "learning_rate": 1.3439990223039878e-05, "loss": 0.0165, "step": 61840 }, { "grad_norm": 0.31545162200927734, "learning_rate": 1.3425894149134321e-05, "loss": 0.0173, "step": 61850 }, { "grad_norm": 0.32366734743118286, "learning_rate": 1.3411804324743132e-05, "loss": 0.0198, "step": 61860 }, { "grad_norm": 0.23589593172073364, "learning_rate": 1.3397720752273896e-05, "loss": 0.0214, "step": 61870 }, { "grad_norm": 0.23250992596149445, "learning_rate": 1.3383643434133108e-05, "loss": 0.0175, "step": 61880 }, { "grad_norm": 0.3037295937538147, "learning_rate": 1.3369572372726196e-05, "loss": 0.0237, "step": 61890 }, { "grad_norm": 0.23586305975914001, "learning_rate": 1.3355507570457525e-05, "loss": 0.021, "step": 61900 }, { "grad_norm": 0.294147253036499, "learning_rate": 1.334144902973038e-05, "loss": 0.0191, "step": 61910 }, { "grad_norm": 0.18539895117282867, "learning_rate": 1.3327396752946997e-05, "loss": 0.0211, "step": 61920 }, { "grad_norm": 0.22640644013881683, "learning_rate": 1.3313350742508496e-05, "loss": 0.0232, "step": 61930 }, { "grad_norm": 0.2851831912994385, "learning_rate": 1.3299311000814973e-05, "loss": 0.0255, "step": 61940 }, { "grad_norm": 0.3027383089065552, "learning_rate": 1.3285277530265432e-05, "loss": 0.0214, "step": 61950 }, { "grad_norm": 0.3086686432361603, "learning_rate": 1.3271250333257813e-05, "loss": 0.0204, "step": 61960 }, { "grad_norm": 0.1964484006166458, "learning_rate": 1.3257229412188977e-05, "loss": 0.0205, "step": 61970 }, { "grad_norm": 0.3651156425476074, "learning_rate": 1.3243214769454726e-05, "loss": 0.0183, "step": 61980 }, { "grad_norm": 0.29076915979385376, "learning_rate": 1.3229206407449751e-05, "loss": 0.0189, "step": 61990 }, { "grad_norm": 0.2706384062767029, "learning_rate": 1.3215204328567742e-05, "loss": 0.0204, "step": 62000 }, { "grad_norm": 0.2568756639957428, "learning_rate": 1.3201208535201232e-05, "loss": 0.0224, "step": 62010 }, { "grad_norm": 0.34785276651382446, "learning_rate": 1.3187219029741737e-05, "loss": 0.0245, "step": 62020 }, { "grad_norm": 0.43103259801864624, "learning_rate": 1.3173235814579693e-05, "loss": 0.025, "step": 62030 }, { "grad_norm": 0.3119252622127533, "learning_rate": 1.3159258892104398e-05, "loss": 0.022, "step": 62040 }, { "grad_norm": 0.27838534116744995, "learning_rate": 1.3145288264704198e-05, "loss": 0.0267, "step": 62050 }, { "grad_norm": 0.32099971175193787, "learning_rate": 1.3131323934766237e-05, "loss": 0.022, "step": 62060 }, { "grad_norm": 0.2865030765533447, "learning_rate": 1.311736590467666e-05, "loss": 0.0212, "step": 62070 }, { "grad_norm": 0.31836044788360596, "learning_rate": 1.3103414176820505e-05, "loss": 0.0223, "step": 62080 }, { "grad_norm": 0.20850609242916107, "learning_rate": 1.308946875358174e-05, "loss": 0.0257, "step": 62090 }, { "grad_norm": 0.45598870515823364, "learning_rate": 1.3075529637343276e-05, "loss": 0.0317, "step": 62100 }, { "grad_norm": 0.3520731031894684, "learning_rate": 1.3061596830486883e-05, "loss": 0.0219, "step": 62110 }, { "grad_norm": 1.4202529191970825, "learning_rate": 1.3047670335393353e-05, "loss": 0.02, "step": 62120 }, { "grad_norm": 0.26063603162765503, "learning_rate": 1.3033750154442298e-05, "loss": 0.0245, "step": 62130 }, { "grad_norm": 0.2432185560464859, "learning_rate": 1.3019836290012316e-05, "loss": 0.021, "step": 62140 }, { "grad_norm": 0.19333839416503906, "learning_rate": 1.30059287444809e-05, "loss": 0.0245, "step": 62150 }, { "grad_norm": 0.38979974389076233, "learning_rate": 1.299202752022447e-05, "loss": 0.0368, "step": 62160 }, { "grad_norm": 0.49194538593292236, "learning_rate": 1.2978132619618371e-05, "loss": 0.0195, "step": 62170 }, { "grad_norm": 0.2995436191558838, "learning_rate": 1.2964244045036866e-05, "loss": 0.0183, "step": 62180 }, { "grad_norm": 0.13091345131397247, "learning_rate": 1.295036179885311e-05, "loss": 0.0157, "step": 62190 }, { "grad_norm": 0.14591218531131744, "learning_rate": 1.293648588343922e-05, "loss": 0.0254, "step": 62200 }, { "grad_norm": 0.2534043788909912, "learning_rate": 1.2922616301166196e-05, "loss": 0.0184, "step": 62210 }, { "grad_norm": 0.23392407596111298, "learning_rate": 1.2908753054403976e-05, "loss": 0.0159, "step": 62220 }, { "grad_norm": 0.3063286244869232, "learning_rate": 1.2894896145521429e-05, "loss": 0.0191, "step": 62230 }, { "grad_norm": 0.5441034436225891, "learning_rate": 1.2881045576886275e-05, "loss": 0.0254, "step": 62240 }, { "grad_norm": 0.23707015812397003, "learning_rate": 1.2867201350865254e-05, "loss": 0.017, "step": 62250 }, { "grad_norm": 0.4634708762168884, "learning_rate": 1.2853363469823914e-05, "loss": 0.0156, "step": 62260 }, { "grad_norm": 0.3354969322681427, "learning_rate": 1.283953193612682e-05, "loss": 0.0184, "step": 62270 }, { "grad_norm": 0.24518531560897827, "learning_rate": 1.2825706752137372e-05, "loss": 0.0187, "step": 62280 }, { "grad_norm": 0.48121947050094604, "learning_rate": 1.2811887920217896e-05, "loss": 0.0186, "step": 62290 }, { "grad_norm": 0.2876710593700409, "learning_rate": 1.2798075442729707e-05, "loss": 0.0251, "step": 62300 }, { "grad_norm": 0.20712170004844666, "learning_rate": 1.2784269322032922e-05, "loss": 0.0224, "step": 62310 }, { "grad_norm": 0.26152852177619934, "learning_rate": 1.2770469560486686e-05, "loss": 0.0244, "step": 62320 }, { "grad_norm": 0.24702438712120056, "learning_rate": 1.2756676160448956e-05, "loss": 0.0214, "step": 62330 }, { "grad_norm": 0.45737946033477783, "learning_rate": 1.2742889124276663e-05, "loss": 0.0216, "step": 62340 }, { "grad_norm": 0.20353420078754425, "learning_rate": 1.2729108454325639e-05, "loss": 0.0189, "step": 62350 }, { "grad_norm": 0.320651650428772, "learning_rate": 1.2715334152950614e-05, "loss": 0.0209, "step": 62360 }, { "grad_norm": 0.36403557658195496, "learning_rate": 1.2701566222505246e-05, "loss": 0.0173, "step": 62370 }, { "grad_norm": 0.4580729603767395, "learning_rate": 1.2687804665342107e-05, "loss": 0.0214, "step": 62380 }, { "grad_norm": 0.49070173501968384, "learning_rate": 1.267404948381265e-05, "loss": 0.0217, "step": 62390 }, { "grad_norm": 0.23995892703533173, "learning_rate": 1.2660300680267267e-05, "loss": 0.0109, "step": 62400 }, { "grad_norm": 0.6818722486495972, "learning_rate": 1.2646558257055257e-05, "loss": 0.0187, "step": 62410 }, { "grad_norm": 0.19379077851772308, "learning_rate": 1.263282221652482e-05, "loss": 0.0157, "step": 62420 }, { "grad_norm": 0.29902902245521545, "learning_rate": 1.2619092561023088e-05, "loss": 0.0215, "step": 62430 }, { "grad_norm": 0.21447648108005524, "learning_rate": 1.2605369292896036e-05, "loss": 0.0194, "step": 62440 }, { "grad_norm": 0.2332761436700821, "learning_rate": 1.2591652414488658e-05, "loss": 0.0234, "step": 62450 }, { "grad_norm": 0.4714881479740143, "learning_rate": 1.2577941928144732e-05, "loss": 0.0228, "step": 62460 }, { "grad_norm": 0.24822555482387543, "learning_rate": 1.256423783620706e-05, "loss": 0.0241, "step": 62470 }, { "grad_norm": 0.3818178176879883, "learning_rate": 1.2550540141017264e-05, "loss": 0.0185, "step": 62480 }, { "grad_norm": 0.3286839723587036, "learning_rate": 1.253684884491591e-05, "loss": 0.0208, "step": 62490 }, { "grad_norm": 0.2709086835384369, "learning_rate": 1.2523163950242483e-05, "loss": 0.0188, "step": 62500 }, { "grad_norm": 0.2989709973335266, "learning_rate": 1.2509485459335313e-05, "loss": 0.0211, "step": 62510 }, { "grad_norm": 0.2710147500038147, "learning_rate": 1.2495813374531739e-05, "loss": 0.0248, "step": 62520 }, { "grad_norm": 0.5740447044372559, "learning_rate": 1.2482147698167907e-05, "loss": 0.0219, "step": 62530 }, { "grad_norm": 0.41162651777267456, "learning_rate": 1.2468488432578911e-05, "loss": 0.0175, "step": 62540 }, { "grad_norm": 0.6274034976959229, "learning_rate": 1.2454835580098761e-05, "loss": 0.0205, "step": 62550 }, { "grad_norm": 0.40138688683509827, "learning_rate": 1.2441189143060338e-05, "loss": 0.0221, "step": 62560 }, { "grad_norm": 0.5414290428161621, "learning_rate": 1.242754912379548e-05, "loss": 0.022, "step": 62570 }, { "grad_norm": 0.3700636923313141, "learning_rate": 1.2413915524634844e-05, "loss": 0.0327, "step": 62580 }, { "grad_norm": 0.305808961391449, "learning_rate": 1.2400288347908073e-05, "loss": 0.0228, "step": 62590 }, { "grad_norm": 0.5390132665634155, "learning_rate": 1.2386667595943663e-05, "loss": 0.0233, "step": 62600 }, { "grad_norm": 0.32656028866767883, "learning_rate": 1.2373053271069035e-05, "loss": 0.0219, "step": 62610 }, { "grad_norm": 0.21918407082557678, "learning_rate": 1.2359445375610501e-05, "loss": 0.0195, "step": 62620 }, { "grad_norm": 0.25273141264915466, "learning_rate": 1.2345843911893301e-05, "loss": 0.0182, "step": 62630 }, { "grad_norm": 0.3071579933166504, "learning_rate": 1.2332248882241498e-05, "loss": 0.0272, "step": 62640 }, { "grad_norm": 0.19812549650669098, "learning_rate": 1.2318660288978178e-05, "loss": 0.0173, "step": 62650 }, { "grad_norm": 0.2101358026266098, "learning_rate": 1.2305078134425213e-05, "loss": 0.0198, "step": 62660 }, { "grad_norm": 0.31846943497657776, "learning_rate": 1.2291502420903434e-05, "loss": 0.02, "step": 62670 }, { "grad_norm": 0.3140392601490021, "learning_rate": 1.2277933150732567e-05, "loss": 0.0187, "step": 62680 }, { "grad_norm": 0.39201974868774414, "learning_rate": 1.2264370326231216e-05, "loss": 0.0225, "step": 62690 }, { "grad_norm": 0.2669579088687897, "learning_rate": 1.2250813949716927e-05, "loss": 0.0168, "step": 62700 }, { "grad_norm": 0.17458446323871613, "learning_rate": 1.2237264023506063e-05, "loss": 0.0179, "step": 62710 }, { "grad_norm": 0.2656118869781494, "learning_rate": 1.2223720549913987e-05, "loss": 0.0208, "step": 62720 }, { "grad_norm": 0.3889247179031372, "learning_rate": 1.221018353125487e-05, "loss": 0.0156, "step": 62730 }, { "grad_norm": 0.29920512437820435, "learning_rate": 1.2196652969841837e-05, "loss": 0.0243, "step": 62740 }, { "grad_norm": 0.3434966504573822, "learning_rate": 1.2183128867986904e-05, "loss": 0.0226, "step": 62750 }, { "grad_norm": 0.5362032651901245, "learning_rate": 1.2169611228000927e-05, "loss": 0.0193, "step": 62760 }, { "grad_norm": 0.1937846839427948, "learning_rate": 1.2156100052193752e-05, "loss": 0.0192, "step": 62770 }, { "grad_norm": 0.2972629964351654, "learning_rate": 1.214259534287403e-05, "loss": 0.0188, "step": 62780 }, { "grad_norm": 0.6274147629737854, "learning_rate": 1.2129097102349363e-05, "loss": 0.0161, "step": 62790 }, { "grad_norm": 0.21964745223522186, "learning_rate": 1.2115605332926227e-05, "loss": 0.0229, "step": 62800 }, { "grad_norm": 0.26620569825172424, "learning_rate": 1.2102120036910003e-05, "loss": 0.0264, "step": 62810 }, { "grad_norm": 0.17867206037044525, "learning_rate": 1.2088641216604956e-05, "loss": 0.0209, "step": 62820 }, { "grad_norm": 0.1909906417131424, "learning_rate": 1.2075168874314264e-05, "loss": 0.0233, "step": 62830 }, { "grad_norm": 0.42008277773857117, "learning_rate": 1.2061703012339942e-05, "loss": 0.0154, "step": 62840 }, { "grad_norm": 0.24877014756202698, "learning_rate": 1.204824363298297e-05, "loss": 0.0233, "step": 62850 }, { "grad_norm": 0.324919730424881, "learning_rate": 1.2034790738543173e-05, "loss": 0.021, "step": 62860 }, { "grad_norm": 0.6727544069290161, "learning_rate": 1.2021344331319284e-05, "loss": 0.0183, "step": 62870 }, { "grad_norm": 0.3401646018028259, "learning_rate": 1.200790441360894e-05, "loss": 0.0278, "step": 62880 }, { "grad_norm": 0.3167474567890167, "learning_rate": 1.199447098770864e-05, "loss": 0.0172, "step": 62890 }, { "grad_norm": 0.2449588179588318, "learning_rate": 1.198104405591381e-05, "loss": 0.0213, "step": 62900 }, { "grad_norm": 0.4542151093482971, "learning_rate": 1.1967623620518697e-05, "loss": 0.0199, "step": 62910 }, { "grad_norm": 0.3385967016220093, "learning_rate": 1.1954209683816554e-05, "loss": 0.0192, "step": 62920 }, { "grad_norm": 0.2826205790042877, "learning_rate": 1.1940802248099402e-05, "loss": 0.0238, "step": 62930 }, { "grad_norm": 0.292510062456131, "learning_rate": 1.1927401315658232e-05, "loss": 0.0194, "step": 62940 }, { "grad_norm": 0.4073469936847687, "learning_rate": 1.1914006888782898e-05, "loss": 0.021, "step": 62950 }, { "grad_norm": 0.21436654031276703, "learning_rate": 1.1900618969762106e-05, "loss": 0.0171, "step": 62960 }, { "grad_norm": 0.24415330588817596, "learning_rate": 1.188723756088354e-05, "loss": 0.0149, "step": 62970 }, { "grad_norm": 0.25050491094589233, "learning_rate": 1.1873862664433672e-05, "loss": 0.0182, "step": 62980 }, { "grad_norm": 0.2470264583826065, "learning_rate": 1.1860494282697927e-05, "loss": 0.0226, "step": 62990 }, { "grad_norm": 0.5177316069602966, "learning_rate": 1.1847132417960588e-05, "loss": 0.0202, "step": 63000 }, { "grad_norm": 0.39974984526634216, "learning_rate": 1.1833777072504832e-05, "loss": 0.0244, "step": 63010 }, { "grad_norm": 0.2379353940486908, "learning_rate": 1.1820428248612731e-05, "loss": 0.0184, "step": 63020 }, { "grad_norm": 0.3944835364818573, "learning_rate": 1.1807085948565245e-05, "loss": 0.0189, "step": 63030 }, { "grad_norm": 0.2893998622894287, "learning_rate": 1.1793750174642172e-05, "loss": 0.023, "step": 63040 }, { "grad_norm": 0.15740559995174408, "learning_rate": 1.1780420929122254e-05, "loss": 0.0182, "step": 63050 }, { "grad_norm": 0.3557577133178711, "learning_rate": 1.176709821428309e-05, "loss": 0.0179, "step": 63060 }, { "grad_norm": 0.2352972775697708, "learning_rate": 1.1753782032401173e-05, "loss": 0.0194, "step": 63070 }, { "grad_norm": 0.26903873682022095, "learning_rate": 1.1740472385751866e-05, "loss": 0.0204, "step": 63080 }, { "grad_norm": 0.2863631546497345, "learning_rate": 1.172716927660943e-05, "loss": 0.0202, "step": 63090 }, { "grad_norm": 0.42587968707084656, "learning_rate": 1.1713872707247015e-05, "loss": 0.0241, "step": 63100 }, { "grad_norm": 0.2887174189090729, "learning_rate": 1.1700582679936595e-05, "loss": 0.0168, "step": 63110 }, { "grad_norm": 0.21459662914276123, "learning_rate": 1.1687299196949136e-05, "loss": 0.0284, "step": 63120 }, { "grad_norm": 0.2645370662212372, "learning_rate": 1.1674022260554374e-05, "loss": 0.0149, "step": 63130 }, { "grad_norm": 0.3988793194293976, "learning_rate": 1.1660751873020987e-05, "loss": 0.0146, "step": 63140 }, { "grad_norm": 0.23086915910243988, "learning_rate": 1.1647488036616538e-05, "loss": 0.0224, "step": 63150 }, { "grad_norm": 0.9805566668510437, "learning_rate": 1.1634230753607417e-05, "loss": 0.0262, "step": 63160 }, { "grad_norm": 0.47782284021377563, "learning_rate": 1.1620980026258982e-05, "loss": 0.0223, "step": 63170 }, { "grad_norm": 0.2623453140258789, "learning_rate": 1.1607735856835373e-05, "loss": 0.0197, "step": 63180 }, { "grad_norm": 0.2774195373058319, "learning_rate": 1.1594498247599677e-05, "loss": 0.02, "step": 63190 }, { "grad_norm": 0.3331698775291443, "learning_rate": 1.158126720081384e-05, "loss": 0.0287, "step": 63200 }, { "grad_norm": 0.36332932114601135, "learning_rate": 1.156804271873868e-05, "loss": 0.024, "step": 63210 }, { "grad_norm": 0.28942522406578064, "learning_rate": 1.1554824803633924e-05, "loss": 0.0205, "step": 63220 }, { "grad_norm": 0.25607359409332275, "learning_rate": 1.154161345775811e-05, "loss": 0.0174, "step": 63230 }, { "grad_norm": 0.2929711937904358, "learning_rate": 1.1528408683368724e-05, "loss": 0.0198, "step": 63240 }, { "grad_norm": 0.44568029046058655, "learning_rate": 1.1515210482722088e-05, "loss": 0.0274, "step": 63250 }, { "grad_norm": 0.25211647152900696, "learning_rate": 1.150201885807342e-05, "loss": 0.0238, "step": 63260 }, { "grad_norm": 0.19514282047748566, "learning_rate": 1.1488833811676807e-05, "loss": 0.0185, "step": 63270 }, { "grad_norm": 0.2095293551683426, "learning_rate": 1.1475655345785213e-05, "loss": 0.0251, "step": 63280 }, { "grad_norm": 0.5622780919075012, "learning_rate": 1.1462483462650481e-05, "loss": 0.0185, "step": 63290 }, { "grad_norm": 0.25433892011642456, "learning_rate": 1.1449318164523331e-05, "loss": 0.0188, "step": 63300 }, { "grad_norm": 0.16473743319511414, "learning_rate": 1.1436159453653334e-05, "loss": 0.0194, "step": 63310 }, { "grad_norm": 0.24456198513507843, "learning_rate": 1.1423007332288955e-05, "loss": 0.0223, "step": 63320 }, { "grad_norm": 0.37902146577835083, "learning_rate": 1.1409861802677546e-05, "loss": 0.018, "step": 63330 }, { "grad_norm": 0.2834765315055847, "learning_rate": 1.1396722867065313e-05, "loss": 0.0177, "step": 63340 }, { "grad_norm": 0.2076549232006073, "learning_rate": 1.1383590527697352e-05, "loss": 0.0197, "step": 63350 }, { "grad_norm": 0.6503766775131226, "learning_rate": 1.137046478681758e-05, "loss": 0.0178, "step": 63360 }, { "grad_norm": 0.4152877628803253, "learning_rate": 1.1357345646668888e-05, "loss": 0.0289, "step": 63370 }, { "grad_norm": 0.3091849684715271, "learning_rate": 1.1344233109492924e-05, "loss": 0.0198, "step": 63380 }, { "grad_norm": 0.24666553735733032, "learning_rate": 1.1331127177530292e-05, "loss": 0.0165, "step": 63390 }, { "grad_norm": 0.43432530760765076, "learning_rate": 1.1318027853020441e-05, "loss": 0.0181, "step": 63400 }, { "grad_norm": 0.3000679314136505, "learning_rate": 1.1304935138201645e-05, "loss": 0.023, "step": 63410 }, { "grad_norm": 0.25621888041496277, "learning_rate": 1.1291849035311153e-05, "loss": 0.0205, "step": 63420 }, { "grad_norm": 0.1991882026195526, "learning_rate": 1.1278769546584972e-05, "loss": 0.0126, "step": 63430 }, { "grad_norm": 0.17479148507118225, "learning_rate": 1.1265696674258052e-05, "loss": 0.0163, "step": 63440 }, { "grad_norm": 0.44641774892807007, "learning_rate": 1.1252630420564186e-05, "loss": 0.0203, "step": 63450 }, { "grad_norm": 0.6495863795280457, "learning_rate": 1.1239570787736036e-05, "loss": 0.0266, "step": 63460 }, { "grad_norm": 0.19808191061019897, "learning_rate": 1.1226517778005135e-05, "loss": 0.0173, "step": 63470 }, { "grad_norm": 0.25508949160575867, "learning_rate": 1.1213471393601893e-05, "loss": 0.0203, "step": 63480 }, { "grad_norm": 0.21396154165267944, "learning_rate": 1.1200431636755587e-05, "loss": 0.0231, "step": 63490 }, { "grad_norm": 0.3395642340183258, "learning_rate": 1.1187398509694336e-05, "loss": 0.0253, "step": 63500 }, { "grad_norm": 0.6846911311149597, "learning_rate": 1.1174372014645146e-05, "loss": 0.0176, "step": 63510 }, { "grad_norm": 0.3135453164577484, "learning_rate": 1.1161352153833899e-05, "loss": 0.0284, "step": 63520 }, { "grad_norm": 0.2616482377052307, "learning_rate": 1.1148338929485325e-05, "loss": 0.0242, "step": 63530 }, { "grad_norm": 0.18558041751384735, "learning_rate": 1.113533234382304e-05, "loss": 0.0152, "step": 63540 }, { "grad_norm": 0.1698831468820572, "learning_rate": 1.1122332399069513e-05, "loss": 0.0222, "step": 63550 }, { "grad_norm": 0.2676144540309906, "learning_rate": 1.1109339097446047e-05, "loss": 0.0239, "step": 63560 }, { "grad_norm": 0.41376742720603943, "learning_rate": 1.1096352441172897e-05, "loss": 0.0264, "step": 63570 }, { "grad_norm": 0.36664140224456787, "learning_rate": 1.1083372432469086e-05, "loss": 0.0202, "step": 63580 }, { "grad_norm": 0.27725112438201904, "learning_rate": 1.107039907355255e-05, "loss": 0.0257, "step": 63590 }, { "grad_norm": 0.34025049209594727, "learning_rate": 1.1057432366640103e-05, "loss": 0.0227, "step": 63600 }, { "grad_norm": 0.44233453273773193, "learning_rate": 1.1044472313947352e-05, "loss": 0.0275, "step": 63610 }, { "grad_norm": 0.32869213819503784, "learning_rate": 1.1031518917688877e-05, "loss": 0.0204, "step": 63620 }, { "grad_norm": 0.27657195925712585, "learning_rate": 1.1018572180078007e-05, "loss": 0.0272, "step": 63630 }, { "grad_norm": 0.46361061930656433, "learning_rate": 1.1005632103327018e-05, "loss": 0.0198, "step": 63640 }, { "grad_norm": 0.2672359347343445, "learning_rate": 1.0992698689646996e-05, "loss": 0.0222, "step": 63650 }, { "grad_norm": 0.31508901715278625, "learning_rate": 1.0979771941247919e-05, "loss": 0.0244, "step": 63660 }, { "grad_norm": 0.2800842821598053, "learning_rate": 1.0966851860338611e-05, "loss": 0.0185, "step": 63670 }, { "grad_norm": 0.19830667972564697, "learning_rate": 1.0953938449126766e-05, "loss": 0.0248, "step": 63680 }, { "grad_norm": 0.6592746376991272, "learning_rate": 1.0941031709818933e-05, "loss": 0.0366, "step": 63690 }, { "grad_norm": 0.1799587458372116, "learning_rate": 1.0928131644620509e-05, "loss": 0.0296, "step": 63700 }, { "grad_norm": 0.48571357131004333, "learning_rate": 1.0915238255735766e-05, "loss": 0.0222, "step": 63710 }, { "grad_norm": 0.13973112404346466, "learning_rate": 1.0902351545367833e-05, "loss": 0.0197, "step": 63720 }, { "grad_norm": 0.43512487411499023, "learning_rate": 1.0889471515718702e-05, "loss": 0.0233, "step": 63730 }, { "grad_norm": 0.45179852843284607, "learning_rate": 1.087659816898921e-05, "loss": 0.021, "step": 63740 }, { "grad_norm": 0.542558491230011, "learning_rate": 1.0863731507379082e-05, "loss": 0.0192, "step": 63750 }, { "grad_norm": 0.3567960858345032, "learning_rate": 1.0850871533086827e-05, "loss": 0.0248, "step": 63760 }, { "grad_norm": 0.31769561767578125, "learning_rate": 1.0838018248309927e-05, "loss": 0.0233, "step": 63770 }, { "grad_norm": 0.48467564582824707, "learning_rate": 1.0825171655244615e-05, "loss": 0.0266, "step": 63780 }, { "grad_norm": 0.5884653925895691, "learning_rate": 1.0812331756086025e-05, "loss": 0.0177, "step": 63790 }, { "grad_norm": 0.4667803943157196, "learning_rate": 1.079949855302817e-05, "loss": 0.0238, "step": 63800 }, { "grad_norm": 0.43748003244400024, "learning_rate": 1.0786672048263852e-05, "loss": 0.0283, "step": 63810 }, { "grad_norm": 0.27171674370765686, "learning_rate": 1.0773852243984817e-05, "loss": 0.0234, "step": 63820 }, { "grad_norm": 0.2238018810749054, "learning_rate": 1.0761039142381586e-05, "loss": 0.0249, "step": 63830 }, { "grad_norm": 0.32970163226127625, "learning_rate": 1.0748232745643577e-05, "loss": 0.0189, "step": 63840 }, { "grad_norm": 0.24725127220153809, "learning_rate": 1.0735433055959055e-05, "loss": 0.0187, "step": 63850 }, { "grad_norm": 0.1932400017976761, "learning_rate": 1.0722640075515133e-05, "loss": 0.0156, "step": 63860 }, { "grad_norm": 0.28991806507110596, "learning_rate": 1.0709853806497795e-05, "loss": 0.0186, "step": 63870 }, { "grad_norm": 0.39815255999565125, "learning_rate": 1.0697074251091831e-05, "loss": 0.0195, "step": 63880 }, { "grad_norm": 0.2769047021865845, "learning_rate": 1.0684301411480962e-05, "loss": 0.0194, "step": 63890 }, { "grad_norm": 0.36800387501716614, "learning_rate": 1.067153528984769e-05, "loss": 0.0173, "step": 63900 }, { "grad_norm": 0.5028465390205383, "learning_rate": 1.0658775888373395e-05, "loss": 0.0211, "step": 63910 }, { "grad_norm": 0.2321668416261673, "learning_rate": 1.0646023209238314e-05, "loss": 0.0245, "step": 63920 }, { "grad_norm": 0.1776280403137207, "learning_rate": 1.0633277254621537e-05, "loss": 0.0152, "step": 63930 }, { "grad_norm": 0.21549636125564575, "learning_rate": 1.0620538026700994e-05, "loss": 0.0194, "step": 63940 }, { "grad_norm": 0.4218829274177551, "learning_rate": 1.0607805527653486e-05, "loss": 0.0189, "step": 63950 }, { "grad_norm": 0.24892696738243103, "learning_rate": 1.059507975965462e-05, "loss": 0.0219, "step": 63960 }, { "grad_norm": 0.2305069863796234, "learning_rate": 1.0582360724878898e-05, "loss": 0.0185, "step": 63970 }, { "grad_norm": 0.22021594643592834, "learning_rate": 1.0569648425499651e-05, "loss": 0.0241, "step": 63980 }, { "grad_norm": 0.22133299708366394, "learning_rate": 1.0556942863689063e-05, "loss": 0.0202, "step": 63990 }, { "grad_norm": 0.36373668909072876, "learning_rate": 1.054424404161819e-05, "loss": 0.0202, "step": 64000 }, { "grad_norm": 0.2841576933860779, "learning_rate": 1.053155196145686e-05, "loss": 0.0228, "step": 64010 }, { "grad_norm": 0.14687591791152954, "learning_rate": 1.0518866625373863e-05, "loss": 0.0175, "step": 64020 }, { "grad_norm": 0.25059592723846436, "learning_rate": 1.0506188035536735e-05, "loss": 0.018, "step": 64030 }, { "grad_norm": 0.7621211409568787, "learning_rate": 1.0493516194111919e-05, "loss": 0.0219, "step": 64040 }, { "grad_norm": 0.29330089688301086, "learning_rate": 1.0480851103264688e-05, "loss": 0.0147, "step": 64050 }, { "grad_norm": 0.1529855728149414, "learning_rate": 1.046819276515913e-05, "loss": 0.0167, "step": 64060 }, { "grad_norm": 0.31221237778663635, "learning_rate": 1.0455541181958256e-05, "loss": 0.0208, "step": 64070 }, { "grad_norm": 0.2770227789878845, "learning_rate": 1.0442896355823822e-05, "loss": 0.0192, "step": 64080 }, { "grad_norm": 0.266509085893631, "learning_rate": 1.0430258288916539e-05, "loss": 0.0188, "step": 64090 }, { "grad_norm": 0.1374715119600296, "learning_rate": 1.0417626983395868e-05, "loss": 0.0192, "step": 64100 }, { "grad_norm": 0.2084847241640091, "learning_rate": 1.0405002441420165e-05, "loss": 0.016, "step": 64110 }, { "grad_norm": 1.0351223945617676, "learning_rate": 1.039238466514662e-05, "loss": 0.0193, "step": 64120 }, { "grad_norm": 0.20694105327129364, "learning_rate": 1.0379773656731262e-05, "loss": 0.0211, "step": 64130 }, { "grad_norm": 0.31032606959342957, "learning_rate": 1.0367169418328986e-05, "loss": 0.0204, "step": 64140 }, { "grad_norm": 0.37791550159454346, "learning_rate": 1.0354571952093484e-05, "loss": 0.0202, "step": 64150 }, { "grad_norm": 0.5720548629760742, "learning_rate": 1.034198126017733e-05, "loss": 0.0179, "step": 64160 }, { "grad_norm": 0.47388648986816406, "learning_rate": 1.0329397344731928e-05, "loss": 0.0236, "step": 64170 }, { "grad_norm": 0.31667909026145935, "learning_rate": 1.0316820207907524e-05, "loss": 0.0196, "step": 64180 }, { "grad_norm": 0.16389840841293335, "learning_rate": 1.0304249851853215e-05, "loss": 0.0185, "step": 64190 }, { "grad_norm": 0.45615822076797485, "learning_rate": 1.0291686278716933e-05, "loss": 0.02, "step": 64200 }, { "grad_norm": 0.3364444375038147, "learning_rate": 1.0279129490645418e-05, "loss": 0.0176, "step": 64210 }, { "grad_norm": 0.49734652042388916, "learning_rate": 1.0266579489784328e-05, "loss": 0.0194, "step": 64220 }, { "grad_norm": 0.29049161076545715, "learning_rate": 1.0254036278278084e-05, "loss": 0.0224, "step": 64230 }, { "grad_norm": 0.267453670501709, "learning_rate": 1.0241499858269982e-05, "loss": 0.0207, "step": 64240 }, { "grad_norm": 0.2489791214466095, "learning_rate": 1.0228970231902169e-05, "loss": 0.0159, "step": 64250 }, { "grad_norm": 0.21583905816078186, "learning_rate": 1.0216447401315582e-05, "loss": 0.0198, "step": 64260 }, { "grad_norm": 0.36833226680755615, "learning_rate": 1.020393136865007e-05, "loss": 0.0169, "step": 64270 }, { "grad_norm": 0.7178710103034973, "learning_rate": 1.0191422136044242e-05, "loss": 0.0297, "step": 64280 }, { "grad_norm": 0.2390059232711792, "learning_rate": 1.017891970563563e-05, "loss": 0.0188, "step": 64290 }, { "grad_norm": 0.40551498532295227, "learning_rate": 1.0166424079560516e-05, "loss": 0.0261, "step": 64300 }, { "grad_norm": 0.3537582457065582, "learning_rate": 1.0153935259954078e-05, "loss": 0.0172, "step": 64310 }, { "grad_norm": 0.3916313946247101, "learning_rate": 1.0141453248950311e-05, "loss": 0.0165, "step": 64320 }, { "grad_norm": 0.25964972376823425, "learning_rate": 1.0128978048682054e-05, "loss": 0.0236, "step": 64330 }, { "grad_norm": 0.43417564034461975, "learning_rate": 1.0116509661280982e-05, "loss": 0.0184, "step": 64340 }, { "grad_norm": 0.2813921272754669, "learning_rate": 1.0104048088877576e-05, "loss": 0.0204, "step": 64350 }, { "grad_norm": 0.3347698450088501, "learning_rate": 1.0091593333601201e-05, "loss": 0.0245, "step": 64360 }, { "grad_norm": 0.22356410324573517, "learning_rate": 1.0079145397580031e-05, "loss": 0.0169, "step": 64370 }, { "grad_norm": 0.4041755497455597, "learning_rate": 1.006670428294107e-05, "loss": 0.0212, "step": 64380 }, { "grad_norm": 0.3173668086528778, "learning_rate": 1.0054269991810166e-05, "loss": 0.0173, "step": 64390 }, { "grad_norm": 0.25521382689476013, "learning_rate": 1.0041842526312024e-05, "loss": 0.0179, "step": 64400 }, { "grad_norm": 0.2478872686624527, "learning_rate": 1.0029421888570101e-05, "loss": 0.0177, "step": 64410 }, { "grad_norm": 0.3346942067146301, "learning_rate": 1.0017008080706813e-05, "loss": 0.0295, "step": 64420 }, { "grad_norm": 0.1864338368177414, "learning_rate": 1.0004601104843287e-05, "loss": 0.0191, "step": 64430 }, { "grad_norm": 0.3175943195819855, "learning_rate": 9.992200963099562e-06, "loss": 0.0262, "step": 64440 }, { "grad_norm": 0.7690631747245789, "learning_rate": 9.979807657594486e-06, "loss": 0.0239, "step": 64450 }, { "grad_norm": 0.4238695502281189, "learning_rate": 9.967421190445703e-06, "loss": 0.0197, "step": 64460 }, { "grad_norm": 0.18014709651470184, "learning_rate": 9.955041563769769e-06, "loss": 0.017, "step": 64470 }, { "grad_norm": 0.5913445949554443, "learning_rate": 9.942668779681974e-06, "loss": 0.0215, "step": 64480 }, { "grad_norm": 0.3556358516216278, "learning_rate": 9.930302840296541e-06, "loss": 0.0194, "step": 64490 }, { "grad_norm": 0.4431513249874115, "learning_rate": 9.917943747726426e-06, "loss": 0.0209, "step": 64500 }, { "grad_norm": 0.663320779800415, "learning_rate": 9.905591504083484e-06, "loss": 0.0209, "step": 64510 }, { "grad_norm": 0.30134162306785583, "learning_rate": 9.893246111478382e-06, "loss": 0.0165, "step": 64520 }, { "grad_norm": 0.49412548542022705, "learning_rate": 9.88090757202057e-06, "loss": 0.0186, "step": 64530 }, { "grad_norm": 0.2064189463853836, "learning_rate": 9.868575887818421e-06, "loss": 0.0246, "step": 64540 }, { "grad_norm": 0.24380972981452942, "learning_rate": 9.856251060979044e-06, "loss": 0.0161, "step": 64550 }, { "grad_norm": 0.23624834418296814, "learning_rate": 9.843933093608426e-06, "loss": 0.0185, "step": 64560 }, { "grad_norm": 0.1759524792432785, "learning_rate": 9.831621987811368e-06, "loss": 0.0267, "step": 64570 }, { "grad_norm": 0.12987685203552246, "learning_rate": 9.819317745691509e-06, "loss": 0.0161, "step": 64580 }, { "grad_norm": 0.3540555238723755, "learning_rate": 9.8070203693513e-06, "loss": 0.0195, "step": 64590 }, { "grad_norm": 0.4483533501625061, "learning_rate": 9.794729860892048e-06, "loss": 0.0243, "step": 64600 }, { "grad_norm": 0.3251306712627411, "learning_rate": 9.782446222413827e-06, "loss": 0.0287, "step": 64610 }, { "grad_norm": 0.3995510935783386, "learning_rate": 9.770169456015598e-06, "loss": 0.0224, "step": 64620 }, { "grad_norm": 0.422063410282135, "learning_rate": 9.75789956379512e-06, "loss": 0.022, "step": 64630 }, { "grad_norm": 0.19361214339733124, "learning_rate": 9.74563654784898e-06, "loss": 0.0201, "step": 64640 }, { "grad_norm": 0.2494840770959854, "learning_rate": 9.733380410272596e-06, "loss": 0.0174, "step": 64650 }, { "grad_norm": 0.19196781516075134, "learning_rate": 9.721131153160207e-06, "loss": 0.0189, "step": 64660 }, { "grad_norm": 0.28184106945991516, "learning_rate": 9.708888778604886e-06, "loss": 0.0212, "step": 64670 }, { "grad_norm": 0.4118787944316864, "learning_rate": 9.696653288698488e-06, "loss": 0.0196, "step": 64680 }, { "grad_norm": 0.8507457971572876, "learning_rate": 9.684424685531763e-06, "loss": 0.027, "step": 64690 }, { "grad_norm": 0.22815987467765808, "learning_rate": 9.672202971194216e-06, "loss": 0.0206, "step": 64700 }, { "grad_norm": 0.1981123834848404, "learning_rate": 9.659988147774213e-06, "loss": 0.0233, "step": 64710 }, { "grad_norm": 0.12498676776885986, "learning_rate": 9.647780217358942e-06, "loss": 0.0173, "step": 64720 }, { "grad_norm": 0.23172380030155182, "learning_rate": 9.635579182034376e-06, "loss": 0.0201, "step": 64730 }, { "grad_norm": 0.3625258803367615, "learning_rate": 9.623385043885386e-06, "loss": 0.0275, "step": 64740 }, { "grad_norm": 0.27232688665390015, "learning_rate": 9.61119780499557e-06, "loss": 0.0178, "step": 64750 }, { "grad_norm": 0.3061824142932892, "learning_rate": 9.599017467447418e-06, "loss": 0.023, "step": 64760 }, { "grad_norm": 0.24447737634181976, "learning_rate": 9.586844033322206e-06, "loss": 0.0219, "step": 64770 }, { "grad_norm": 0.5327280163764954, "learning_rate": 9.574677504700052e-06, "loss": 0.0216, "step": 64780 }, { "grad_norm": 0.24511796236038208, "learning_rate": 9.562517883659877e-06, "loss": 0.0278, "step": 64790 }, { "grad_norm": 0.16770543158054352, "learning_rate": 9.55036517227944e-06, "loss": 0.0166, "step": 64800 }, { "grad_norm": 0.30171099305152893, "learning_rate": 9.538219372635282e-06, "loss": 0.0214, "step": 64810 }, { "grad_norm": 0.328764408826828, "learning_rate": 9.526080486802802e-06, "loss": 0.0194, "step": 64820 }, { "grad_norm": 0.17853336036205292, "learning_rate": 9.513948516856203e-06, "loss": 0.0206, "step": 64830 }, { "grad_norm": 0.3184701204299927, "learning_rate": 9.501823464868504e-06, "loss": 0.0177, "step": 64840 }, { "grad_norm": 0.2398167997598648, "learning_rate": 9.489705332911547e-06, "loss": 0.0163, "step": 64850 }, { "grad_norm": 0.2727256417274475, "learning_rate": 9.477594123055994e-06, "loss": 0.0162, "step": 64860 }, { "grad_norm": 0.2899407148361206, "learning_rate": 9.465489837371321e-06, "loss": 0.0242, "step": 64870 }, { "grad_norm": 0.2865319848060608, "learning_rate": 9.453392477925794e-06, "loss": 0.018, "step": 64880 }, { "grad_norm": 0.7349603772163391, "learning_rate": 9.441302046786566e-06, "loss": 0.0156, "step": 64890 }, { "grad_norm": 0.2777253985404968, "learning_rate": 9.429218546019519e-06, "loss": 0.0161, "step": 64900 }, { "grad_norm": 0.27134278416633606, "learning_rate": 9.41714197768941e-06, "loss": 0.0222, "step": 64910 }, { "grad_norm": 0.19236110150814056, "learning_rate": 9.405072343859805e-06, "loss": 0.017, "step": 64920 }, { "grad_norm": 0.3162670135498047, "learning_rate": 9.393009646593043e-06, "loss": 0.017, "step": 64930 }, { "grad_norm": 0.27542778849601746, "learning_rate": 9.38095388795035e-06, "loss": 0.0215, "step": 64940 }, { "grad_norm": 0.44841063022613525, "learning_rate": 9.36890506999169e-06, "loss": 0.02, "step": 64950 }, { "grad_norm": 0.2565057575702667, "learning_rate": 9.356863194775894e-06, "loss": 0.0223, "step": 64960 }, { "grad_norm": 1.2591311931610107, "learning_rate": 9.344828264360583e-06, "loss": 0.0219, "step": 64970 }, { "grad_norm": 0.3403938412666321, "learning_rate": 9.332800280802201e-06, "loss": 0.02, "step": 64980 }, { "grad_norm": 0.5321314334869385, "learning_rate": 9.32077924615602e-06, "loss": 0.0178, "step": 64990 }, { "grad_norm": 0.32176902890205383, "learning_rate": 9.308765162476063e-06, "loss": 0.0226, "step": 65000 }, { "grad_norm": 0.3335490822792053, "learning_rate": 9.296758031815239e-06, "loss": 0.0234, "step": 65010 }, { "grad_norm": 0.2503722012042999, "learning_rate": 9.284757856225229e-06, "loss": 0.0246, "step": 65020 }, { "grad_norm": 0.10459569096565247, "learning_rate": 9.272764637756538e-06, "loss": 0.0177, "step": 65030 }, { "grad_norm": 0.3354027569293976, "learning_rate": 9.260778378458479e-06, "loss": 0.0219, "step": 65040 }, { "grad_norm": 0.3576543629169464, "learning_rate": 9.248799080379172e-06, "loss": 0.0207, "step": 65050 }, { "grad_norm": 0.26920732855796814, "learning_rate": 9.236826745565558e-06, "loss": 0.022, "step": 65060 }, { "grad_norm": 0.19107849895954132, "learning_rate": 9.224861376063388e-06, "loss": 0.0223, "step": 65070 }, { "grad_norm": 0.25272274017333984, "learning_rate": 9.212902973917192e-06, "loss": 0.0178, "step": 65080 }, { "grad_norm": 0.8768083453178406, "learning_rate": 9.20095154117035e-06, "loss": 0.0137, "step": 65090 }, { "grad_norm": 0.15968537330627441, "learning_rate": 9.189007079865036e-06, "loss": 0.0176, "step": 65100 }, { "grad_norm": 0.40942397713661194, "learning_rate": 9.177069592042226e-06, "loss": 0.0182, "step": 65110 }, { "grad_norm": 0.5752711296081543, "learning_rate": 9.165139079741724e-06, "loss": 0.0261, "step": 65120 }, { "grad_norm": 0.27942314743995667, "learning_rate": 9.153215545002098e-06, "loss": 0.0259, "step": 65130 }, { "grad_norm": 0.28180065751075745, "learning_rate": 9.141298989860798e-06, "loss": 0.0258, "step": 65140 }, { "grad_norm": 0.1586105078458786, "learning_rate": 9.129389416353994e-06, "loss": 0.016, "step": 65150 }, { "grad_norm": 0.19990389049053192, "learning_rate": 9.11748682651673e-06, "loss": 0.0182, "step": 65160 }, { "grad_norm": 0.3215380907058716, "learning_rate": 9.105591222382837e-06, "loss": 0.0223, "step": 65170 }, { "grad_norm": 0.3559882640838623, "learning_rate": 9.093702605984915e-06, "loss": 0.0178, "step": 65180 }, { "grad_norm": 0.42788955569267273, "learning_rate": 9.081820979354455e-06, "loss": 0.0181, "step": 65190 }, { "grad_norm": 0.3929653763771057, "learning_rate": 9.069946344521663e-06, "loss": 0.0209, "step": 65200 }, { "grad_norm": 0.942274272441864, "learning_rate": 9.058078703515598e-06, "loss": 0.0195, "step": 65210 }, { "grad_norm": 0.22761142253875732, "learning_rate": 9.046218058364125e-06, "loss": 0.0188, "step": 65220 }, { "grad_norm": 0.33015793561935425, "learning_rate": 9.034364411093893e-06, "loss": 0.0157, "step": 65230 }, { "grad_norm": 0.16497567296028137, "learning_rate": 9.022517763730371e-06, "loss": 0.0196, "step": 65240 }, { "grad_norm": 0.6121340990066528, "learning_rate": 9.010678118297827e-06, "loss": 0.0228, "step": 65250 }, { "grad_norm": 0.3635931611061096, "learning_rate": 8.998845476819345e-06, "loss": 0.016, "step": 65260 }, { "grad_norm": 0.35453179478645325, "learning_rate": 8.987019841316773e-06, "loss": 0.022, "step": 65270 }, { "grad_norm": 0.44589293003082275, "learning_rate": 8.975201213810802e-06, "loss": 0.0173, "step": 65280 }, { "grad_norm": 0.31799668073654175, "learning_rate": 8.963389596320915e-06, "loss": 0.0217, "step": 65290 }, { "grad_norm": 0.34307852387428284, "learning_rate": 8.951584990865391e-06, "loss": 0.0214, "step": 65300 }, { "grad_norm": 0.4988681375980377, "learning_rate": 8.939787399461319e-06, "loss": 0.0201, "step": 65310 }, { "grad_norm": 0.48184290528297424, "learning_rate": 8.927996824124591e-06, "loss": 0.0227, "step": 65320 }, { "grad_norm": 0.5014515519142151, "learning_rate": 8.916213266869854e-06, "loss": 0.0235, "step": 65330 }, { "grad_norm": 0.2911550998687744, "learning_rate": 8.904436729710658e-06, "loss": 0.0155, "step": 65340 }, { "grad_norm": 0.556866466999054, "learning_rate": 8.892667214659245e-06, "loss": 0.0155, "step": 65350 }, { "grad_norm": 0.32620203495025635, "learning_rate": 8.880904723726713e-06, "loss": 0.0157, "step": 65360 }, { "grad_norm": 0.20362423360347748, "learning_rate": 8.869149258922971e-06, "loss": 0.0146, "step": 65370 }, { "grad_norm": 0.21232059597969055, "learning_rate": 8.857400822256662e-06, "loss": 0.0195, "step": 65380 }, { "grad_norm": 0.24709893763065338, "learning_rate": 8.845659415735324e-06, "loss": 0.0225, "step": 65390 }, { "grad_norm": 0.26336610317230225, "learning_rate": 8.83392504136521e-06, "loss": 0.0217, "step": 65400 }, { "grad_norm": 0.5963007807731628, "learning_rate": 8.822197701151407e-06, "loss": 0.0223, "step": 65410 }, { "grad_norm": 0.15748274326324463, "learning_rate": 8.810477397097804e-06, "loss": 0.0183, "step": 65420 }, { "grad_norm": 0.49871760606765747, "learning_rate": 8.798764131207077e-06, "loss": 0.018, "step": 65430 }, { "grad_norm": 0.1305701583623886, "learning_rate": 8.787057905480706e-06, "loss": 0.018, "step": 65440 }, { "grad_norm": 0.20202931761741638, "learning_rate": 8.775358721918958e-06, "loss": 0.0172, "step": 65450 }, { "grad_norm": 0.22626745700836182, "learning_rate": 8.763666582520923e-06, "loss": 0.0206, "step": 65460 }, { "grad_norm": 0.4127160310745239, "learning_rate": 8.751981489284445e-06, "loss": 0.019, "step": 65470 }, { "grad_norm": 0.23354752361774445, "learning_rate": 8.740303444206188e-06, "loss": 0.0205, "step": 65480 }, { "grad_norm": 0.2768862247467041, "learning_rate": 8.72863244928162e-06, "loss": 0.0195, "step": 65490 }, { "grad_norm": 0.2851773202419281, "learning_rate": 8.716968506504991e-06, "loss": 0.0282, "step": 65500 }, { "grad_norm": 0.3169936239719391, "learning_rate": 8.70531161786935e-06, "loss": 0.0171, "step": 65510 }, { "grad_norm": 0.35925570130348206, "learning_rate": 8.693661785366558e-06, "loss": 0.0268, "step": 65520 }, { "grad_norm": 0.3017876148223877, "learning_rate": 8.682019010987208e-06, "loss": 0.0256, "step": 65530 }, { "grad_norm": 0.2409389615058899, "learning_rate": 8.670383296720786e-06, "loss": 0.0185, "step": 65540 }, { "grad_norm": 0.1908060759305954, "learning_rate": 8.658754644555478e-06, "loss": 0.0213, "step": 65550 }, { "grad_norm": 0.22430279850959778, "learning_rate": 8.647133056478313e-06, "loss": 0.0234, "step": 65560 }, { "grad_norm": 0.36661916971206665, "learning_rate": 8.635518534475123e-06, "loss": 0.0173, "step": 65570 }, { "grad_norm": 0.24450743198394775, "learning_rate": 8.623911080530467e-06, "loss": 0.0207, "step": 65580 }, { "grad_norm": 0.26848217844963074, "learning_rate": 8.61231069662779e-06, "loss": 0.021, "step": 65590 }, { "grad_norm": 0.38659432530403137, "learning_rate": 8.600717384749252e-06, "loss": 0.0203, "step": 65600 }, { "grad_norm": 0.3457944989204407, "learning_rate": 8.58913114687584e-06, "loss": 0.022, "step": 65610 }, { "grad_norm": 0.42646434903144836, "learning_rate": 8.57755198498732e-06, "loss": 0.0235, "step": 65620 }, { "grad_norm": 0.2682492434978485, "learning_rate": 8.565979901062265e-06, "loss": 0.0258, "step": 65630 }, { "grad_norm": 0.2360558658838272, "learning_rate": 8.554414897078033e-06, "loss": 0.0228, "step": 65640 }, { "grad_norm": 0.43718937039375305, "learning_rate": 8.542856975010727e-06, "loss": 0.0173, "step": 65650 }, { "grad_norm": 0.4413819909095764, "learning_rate": 8.531306136835337e-06, "loss": 0.0212, "step": 65660 }, { "grad_norm": 0.3454114496707916, "learning_rate": 8.519762384525548e-06, "loss": 0.0232, "step": 65670 }, { "grad_norm": 0.4662169814109802, "learning_rate": 8.508225720053875e-06, "loss": 0.0245, "step": 65680 }, { "grad_norm": 0.26147255301475525, "learning_rate": 8.496696145391625e-06, "loss": 0.0182, "step": 65690 }, { "grad_norm": 0.31785643100738525, "learning_rate": 8.485173662508889e-06, "loss": 0.019, "step": 65700 }, { "grad_norm": 0.18250672519207, "learning_rate": 8.473658273374536e-06, "loss": 0.013, "step": 65710 }, { "grad_norm": 0.16494227945804596, "learning_rate": 8.462149979956253e-06, "loss": 0.0121, "step": 65720 }, { "grad_norm": 0.32442706823349, "learning_rate": 8.450648784220461e-06, "loss": 0.0201, "step": 65730 }, { "grad_norm": 0.2916001081466675, "learning_rate": 8.439154688132417e-06, "loss": 0.0179, "step": 65740 }, { "grad_norm": 0.35498398542404175, "learning_rate": 8.427667693656143e-06, "loss": 0.0194, "step": 65750 }, { "grad_norm": 0.26902174949645996, "learning_rate": 8.416187802754454e-06, "loss": 0.0178, "step": 65760 }, { "grad_norm": 0.1961744725704193, "learning_rate": 8.404715017388965e-06, "loss": 0.0168, "step": 65770 }, { "grad_norm": 0.2732551693916321, "learning_rate": 8.393249339520015e-06, "loss": 0.022, "step": 65780 }, { "grad_norm": 0.18437343835830688, "learning_rate": 8.381790771106834e-06, "loss": 0.0159, "step": 65790 }, { "grad_norm": 0.6163358688354492, "learning_rate": 8.370339314107339e-06, "loss": 0.0158, "step": 65800 }, { "grad_norm": 0.3586806058883667, "learning_rate": 8.358894970478281e-06, "loss": 0.0214, "step": 65810 }, { "grad_norm": 0.6316561102867126, "learning_rate": 8.347457742175196e-06, "loss": 0.0183, "step": 65820 }, { "grad_norm": 0.3040401339530945, "learning_rate": 8.33602763115236e-06, "loss": 0.0171, "step": 65830 }, { "grad_norm": 0.26446056365966797, "learning_rate": 8.324604639362916e-06, "loss": 0.0229, "step": 65840 }, { "grad_norm": 0.3241110146045685, "learning_rate": 8.31318876875869e-06, "loss": 0.021, "step": 65850 }, { "grad_norm": 0.3080770671367645, "learning_rate": 8.30178002129039e-06, "loss": 0.0197, "step": 65860 }, { "grad_norm": 0.24398232996463776, "learning_rate": 8.290378398907423e-06, "loss": 0.0185, "step": 65870 }, { "grad_norm": 0.18629196286201477, "learning_rate": 8.278983903558029e-06, "loss": 0.0187, "step": 65880 }, { "grad_norm": 0.26848891377449036, "learning_rate": 8.26759653718921e-06, "loss": 0.0214, "step": 65890 }, { "grad_norm": 0.48267579078674316, "learning_rate": 8.25621630174676e-06, "loss": 0.0202, "step": 65900 }, { "grad_norm": 0.42517003417015076, "learning_rate": 8.244843199175261e-06, "loss": 0.0281, "step": 65910 }, { "grad_norm": 0.6557698845863342, "learning_rate": 8.233477231418046e-06, "loss": 0.0204, "step": 65920 }, { "grad_norm": 0.5424200892448425, "learning_rate": 8.22211840041725e-06, "loss": 0.021, "step": 65930 }, { "grad_norm": 0.49607256054878235, "learning_rate": 8.210766708113792e-06, "loss": 0.0238, "step": 65940 }, { "grad_norm": 0.2642417848110199, "learning_rate": 8.199422156447367e-06, "loss": 0.0227, "step": 65950 }, { "grad_norm": 0.2847851514816284, "learning_rate": 8.188084747356451e-06, "loss": 0.0224, "step": 65960 }, { "grad_norm": 0.2728894054889679, "learning_rate": 8.176754482778299e-06, "loss": 0.0247, "step": 65970 }, { "grad_norm": 0.3219202756881714, "learning_rate": 8.165431364648918e-06, "loss": 0.0202, "step": 65980 }, { "grad_norm": 0.2287338376045227, "learning_rate": 8.154115394903162e-06, "loss": 0.0156, "step": 65990 }, { "grad_norm": 0.4437606930732727, "learning_rate": 8.142806575474582e-06, "loss": 0.0188, "step": 66000 }, { "grad_norm": 0.7239159941673279, "learning_rate": 8.131504908295562e-06, "loss": 0.0199, "step": 66010 }, { "grad_norm": 0.5354074239730835, "learning_rate": 8.120210395297262e-06, "loss": 0.0157, "step": 66020 }, { "grad_norm": 0.22233125567436218, "learning_rate": 8.108923038409565e-06, "loss": 0.0226, "step": 66030 }, { "grad_norm": 0.3037063479423523, "learning_rate": 8.097642839561226e-06, "loss": 0.0187, "step": 66040 }, { "grad_norm": 0.25305259227752686, "learning_rate": 8.08636980067966e-06, "loss": 0.0261, "step": 66050 }, { "grad_norm": 0.8086663484573364, "learning_rate": 8.075103923691186e-06, "loss": 0.0198, "step": 66060 }, { "grad_norm": 0.34006741642951965, "learning_rate": 8.063845210520793e-06, "loss": 0.024, "step": 66070 }, { "grad_norm": 0.2615557909011841, "learning_rate": 8.052593663092295e-06, "loss": 0.0214, "step": 66080 }, { "grad_norm": 0.2259102612733841, "learning_rate": 8.04134928332827e-06, "loss": 0.0161, "step": 66090 }, { "grad_norm": 0.8153185248374939, "learning_rate": 8.030112073150086e-06, "loss": 0.0183, "step": 66100 }, { "grad_norm": 0.309329628944397, "learning_rate": 8.018882034477881e-06, "loss": 0.0158, "step": 66110 }, { "grad_norm": 0.36571264266967773, "learning_rate": 8.007659169230541e-06, "loss": 0.0171, "step": 66120 }, { "grad_norm": 0.5661012530326843, "learning_rate": 7.996443479325755e-06, "loss": 0.0189, "step": 66130 }, { "grad_norm": 0.18889431655406952, "learning_rate": 7.985234966679977e-06, "loss": 0.0227, "step": 66140 }, { "grad_norm": 0.973439633846283, "learning_rate": 7.974033633208438e-06, "loss": 0.0178, "step": 66150 }, { "grad_norm": 1.103581428527832, "learning_rate": 7.962839480825135e-06, "loss": 0.0187, "step": 66160 }, { "grad_norm": 0.5742293000221252, "learning_rate": 7.951652511442858e-06, "loss": 0.024, "step": 66170 }, { "grad_norm": 0.3429897427558899, "learning_rate": 7.940472726973125e-06, "loss": 0.0248, "step": 66180 }, { "grad_norm": 0.26000040769577026, "learning_rate": 7.929300129326289e-06, "loss": 0.0155, "step": 66190 }, { "grad_norm": 0.23517993092536926, "learning_rate": 7.91813472041142e-06, "loss": 0.0224, "step": 66200 }, { "grad_norm": 0.3507995009422302, "learning_rate": 7.906976502136376e-06, "loss": 0.0191, "step": 66210 }, { "grad_norm": 0.213362917304039, "learning_rate": 7.89582547640782e-06, "loss": 0.0225, "step": 66220 }, { "grad_norm": 0.3563345670700073, "learning_rate": 7.884681645131115e-06, "loss": 0.0224, "step": 66230 }, { "grad_norm": 0.3532874882221222, "learning_rate": 7.87354501021048e-06, "loss": 0.0192, "step": 66240 }, { "grad_norm": 0.14215251803398132, "learning_rate": 7.86241557354882e-06, "loss": 0.0153, "step": 66250 }, { "grad_norm": 0.3095495402812958, "learning_rate": 7.8512933370479e-06, "loss": 0.0192, "step": 66260 }, { "grad_norm": 0.26558223366737366, "learning_rate": 7.840178302608158e-06, "loss": 0.0183, "step": 66270 }, { "grad_norm": 0.6889671683311462, "learning_rate": 7.82907047212888e-06, "loss": 0.0208, "step": 66280 }, { "grad_norm": 0.29971012473106384, "learning_rate": 7.81796984750809e-06, "loss": 0.0194, "step": 66290 }, { "grad_norm": 0.30777764320373535, "learning_rate": 7.806876430642546e-06, "loss": 0.0184, "step": 66300 }, { "grad_norm": 0.3080613911151886, "learning_rate": 7.795790223427862e-06, "loss": 0.0201, "step": 66310 }, { "grad_norm": 0.40224847197532654, "learning_rate": 7.784711227758324e-06, "loss": 0.0237, "step": 66320 }, { "grad_norm": 0.25480830669403076, "learning_rate": 7.773639445527053e-06, "loss": 0.0187, "step": 66330 }, { "grad_norm": 0.31044718623161316, "learning_rate": 7.762574878625905e-06, "loss": 0.0162, "step": 66340 }, { "grad_norm": 0.3531248867511749, "learning_rate": 7.751517528945513e-06, "loss": 0.0147, "step": 66350 }, { "grad_norm": 0.41942593455314636, "learning_rate": 7.740467398375278e-06, "loss": 0.0201, "step": 66360 }, { "grad_norm": 0.30732887983322144, "learning_rate": 7.729424488803378e-06, "loss": 0.0158, "step": 66370 }, { "grad_norm": 0.7920588850975037, "learning_rate": 7.71838880211671e-06, "loss": 0.026, "step": 66380 }, { "grad_norm": 0.5040701031684875, "learning_rate": 7.707360340200997e-06, "loss": 0.0147, "step": 66390 }, { "grad_norm": 0.28787893056869507, "learning_rate": 7.696339104940697e-06, "loss": 0.0173, "step": 66400 }, { "grad_norm": 0.24718759953975677, "learning_rate": 7.685325098219038e-06, "loss": 0.0194, "step": 66410 }, { "grad_norm": 0.3413603901863098, "learning_rate": 7.674318321918017e-06, "loss": 0.0254, "step": 66420 }, { "grad_norm": 0.3600670397281647, "learning_rate": 7.663318777918366e-06, "loss": 0.0149, "step": 66430 }, { "grad_norm": 0.271165668964386, "learning_rate": 7.652326468099647e-06, "loss": 0.0229, "step": 66440 }, { "grad_norm": 0.21373865008354187, "learning_rate": 7.641341394340096e-06, "loss": 0.0152, "step": 66450 }, { "grad_norm": 0.41259533166885376, "learning_rate": 7.630363558516818e-06, "loss": 0.0214, "step": 66460 }, { "grad_norm": 0.6196686029434204, "learning_rate": 7.619392962505578e-06, "loss": 0.0182, "step": 66470 }, { "grad_norm": 0.2069580852985382, "learning_rate": 7.6084296081809725e-06, "loss": 0.0215, "step": 66480 }, { "grad_norm": 0.36082160472869873, "learning_rate": 7.597473497416347e-06, "loss": 0.0199, "step": 66490 }, { "grad_norm": 0.12222053110599518, "learning_rate": 7.586524632083764e-06, "loss": 0.0142, "step": 66500 }, { "grad_norm": 0.2447621375322342, "learning_rate": 7.5755830140541326e-06, "loss": 0.0168, "step": 66510 }, { "grad_norm": 0.2089080512523651, "learning_rate": 7.5646486451970475e-06, "loss": 0.0189, "step": 66520 }, { "grad_norm": 0.26144567131996155, "learning_rate": 7.553721527380897e-06, "loss": 0.0184, "step": 66530 }, { "grad_norm": 0.22951170802116394, "learning_rate": 7.542801662472826e-06, "loss": 0.0165, "step": 66540 }, { "grad_norm": 0.7378123998641968, "learning_rate": 7.5318890523387475e-06, "loss": 0.0187, "step": 66550 }, { "grad_norm": 0.5648078322410583, "learning_rate": 7.52098369884332e-06, "loss": 0.0217, "step": 66560 }, { "grad_norm": 0.2601727843284607, "learning_rate": 7.510085603849992e-06, "loss": 0.0222, "step": 66570 }, { "grad_norm": 0.1804959625005722, "learning_rate": 7.499194769220918e-06, "loss": 0.0154, "step": 66580 }, { "grad_norm": 0.4413888156414032, "learning_rate": 7.48831119681706e-06, "loss": 0.0227, "step": 66590 }, { "grad_norm": 0.1771756261587143, "learning_rate": 7.477434888498119e-06, "loss": 0.0174, "step": 66600 }, { "grad_norm": 0.3011566698551178, "learning_rate": 7.466565846122564e-06, "loss": 0.0198, "step": 66610 }, { "grad_norm": 0.22826431691646576, "learning_rate": 7.455704071547626e-06, "loss": 0.0178, "step": 66620 }, { "grad_norm": 0.31957876682281494, "learning_rate": 7.444849566629247e-06, "loss": 0.0182, "step": 66630 }, { "grad_norm": 0.3062915503978729, "learning_rate": 7.434002333222212e-06, "loss": 0.0162, "step": 66640 }, { "grad_norm": 0.45450255274772644, "learning_rate": 7.423162373179976e-06, "loss": 0.0155, "step": 66650 }, { "grad_norm": 0.2706794738769531, "learning_rate": 7.412329688354835e-06, "loss": 0.0235, "step": 66660 }, { "grad_norm": 0.2542688250541687, "learning_rate": 7.40150428059776e-06, "loss": 0.0183, "step": 66670 }, { "grad_norm": 0.4534758925437927, "learning_rate": 7.3906861517585354e-06, "loss": 0.0237, "step": 66680 }, { "grad_norm": 0.7831277847290039, "learning_rate": 7.37987530368569e-06, "loss": 0.0173, "step": 66690 }, { "grad_norm": 0.25854039192199707, "learning_rate": 7.369071738226474e-06, "loss": 0.0162, "step": 66700 }, { "grad_norm": 0.2787981629371643, "learning_rate": 7.358275457226954e-06, "loss": 0.0152, "step": 66710 }, { "grad_norm": 0.3425036072731018, "learning_rate": 7.347486462531899e-06, "loss": 0.0203, "step": 66720 }, { "grad_norm": 0.18760354816913605, "learning_rate": 7.336704755984858e-06, "loss": 0.0139, "step": 66730 }, { "grad_norm": 0.10221223533153534, "learning_rate": 7.325930339428133e-06, "loss": 0.0199, "step": 66740 }, { "grad_norm": 0.2583830952644348, "learning_rate": 7.315163214702769e-06, "loss": 0.019, "step": 66750 }, { "grad_norm": 0.46458038687705994, "learning_rate": 7.304403383648595e-06, "loss": 0.0174, "step": 66760 }, { "grad_norm": 0.23793764412403107, "learning_rate": 7.293650848104139e-06, "loss": 0.0138, "step": 66770 }, { "grad_norm": 0.45336127281188965, "learning_rate": 7.2829056099067374e-06, "loss": 0.0147, "step": 66780 }, { "grad_norm": 0.20374129712581635, "learning_rate": 7.2721676708924494e-06, "loss": 0.0199, "step": 66790 }, { "grad_norm": 1.1214640140533447, "learning_rate": 7.261437032896096e-06, "loss": 0.0359, "step": 66800 }, { "grad_norm": 0.15082786977291107, "learning_rate": 7.250713697751255e-06, "loss": 0.0194, "step": 66810 }, { "grad_norm": 0.292020320892334, "learning_rate": 7.239997667290255e-06, "loss": 0.0139, "step": 66820 }, { "grad_norm": 0.274774968624115, "learning_rate": 7.2292889433441425e-06, "loss": 0.0166, "step": 66830 }, { "grad_norm": 0.6074391007423401, "learning_rate": 7.218587527742793e-06, "loss": 0.0138, "step": 66840 }, { "grad_norm": 0.2606668770313263, "learning_rate": 7.207893422314749e-06, "loss": 0.0213, "step": 66850 }, { "grad_norm": 0.367437481880188, "learning_rate": 7.1972066288873545e-06, "loss": 0.0201, "step": 66860 }, { "grad_norm": 0.23499810695648193, "learning_rate": 7.186527149286687e-06, "loss": 0.0339, "step": 66870 }, { "grad_norm": 0.2978322505950928, "learning_rate": 7.175854985337576e-06, "loss": 0.0191, "step": 66880 }, { "grad_norm": 0.226979598402977, "learning_rate": 7.16519013886362e-06, "loss": 0.0227, "step": 66890 }, { "grad_norm": 0.27340462803840637, "learning_rate": 7.154532611687109e-06, "loss": 0.0179, "step": 66900 }, { "grad_norm": 0.27155008912086487, "learning_rate": 7.143882405629177e-06, "loss": 0.0164, "step": 66910 }, { "grad_norm": 0.15707089006900787, "learning_rate": 7.133239522509605e-06, "loss": 0.0152, "step": 66920 }, { "grad_norm": 1.2413619756698608, "learning_rate": 7.1226039641469956e-06, "loss": 0.0187, "step": 66930 }, { "grad_norm": 0.25888267159461975, "learning_rate": 7.111975732358678e-06, "loss": 0.018, "step": 66940 }, { "grad_norm": 0.7137120962142944, "learning_rate": 7.101354828960693e-06, "loss": 0.0136, "step": 66950 }, { "grad_norm": 0.19247299432754517, "learning_rate": 7.090741255767918e-06, "loss": 0.0158, "step": 66960 }, { "grad_norm": 0.2708955407142639, "learning_rate": 7.080135014593875e-06, "loss": 0.0138, "step": 66970 }, { "grad_norm": 1.1016383171081543, "learning_rate": 7.069536107250896e-06, "loss": 0.0145, "step": 66980 }, { "grad_norm": 0.5876256823539734, "learning_rate": 7.058944535550049e-06, "loss": 0.0183, "step": 66990 }, { "grad_norm": 0.5074091553688049, "learning_rate": 7.048360301301138e-06, "loss": 0.0218, "step": 67000 }, { "grad_norm": 0.308940052986145, "learning_rate": 7.03778340631272e-06, "loss": 0.0182, "step": 67010 }, { "grad_norm": 0.34526363015174866, "learning_rate": 7.0272138523921e-06, "loss": 0.0169, "step": 67020 }, { "grad_norm": 0.23883292078971863, "learning_rate": 7.016651641345334e-06, "loss": 0.0228, "step": 67030 }, { "grad_norm": 0.24939802289009094, "learning_rate": 7.0060967749772e-06, "loss": 0.0165, "step": 67040 }, { "grad_norm": 0.5907725691795349, "learning_rate": 6.995549255091238e-06, "loss": 0.0243, "step": 67050 }, { "grad_norm": 0.4495316445827484, "learning_rate": 6.98500908348973e-06, "loss": 0.0232, "step": 67060 }, { "grad_norm": 0.22505007684230804, "learning_rate": 6.974476261973711e-06, "loss": 0.0223, "step": 67070 }, { "grad_norm": 0.5558612942695618, "learning_rate": 6.963950792342949e-06, "loss": 0.0135, "step": 67080 }, { "grad_norm": 0.2066812664270401, "learning_rate": 6.9534326763959715e-06, "loss": 0.0178, "step": 67090 }, { "grad_norm": 0.46606889367103577, "learning_rate": 6.942921915929995e-06, "loss": 0.0181, "step": 67100 }, { "grad_norm": 0.218128964304924, "learning_rate": 6.9324185127410735e-06, "loss": 0.0202, "step": 67110 }, { "grad_norm": 0.25056520104408264, "learning_rate": 6.921922468623921e-06, "loss": 0.0229, "step": 67120 }, { "grad_norm": 0.5467324256896973, "learning_rate": 6.911433785372023e-06, "loss": 0.0164, "step": 67130 }, { "grad_norm": 0.31557345390319824, "learning_rate": 6.900952464777632e-06, "loss": 0.0191, "step": 67140 }, { "grad_norm": 0.3451196253299713, "learning_rate": 6.8904785086316815e-06, "loss": 0.0231, "step": 67150 }, { "grad_norm": 0.3584006130695343, "learning_rate": 6.880011918723927e-06, "loss": 0.0235, "step": 67160 }, { "grad_norm": 0.2386394441127777, "learning_rate": 6.86955269684279e-06, "loss": 0.0173, "step": 67170 }, { "grad_norm": 0.30838316679000854, "learning_rate": 6.859100844775473e-06, "loss": 0.0279, "step": 67180 }, { "grad_norm": 0.30851560831069946, "learning_rate": 6.84865636430792e-06, "loss": 0.0184, "step": 67190 }, { "grad_norm": 0.33305779099464417, "learning_rate": 6.838219257224804e-06, "loss": 0.0204, "step": 67200 }, { "grad_norm": 0.14072132110595703, "learning_rate": 6.827789525309536e-06, "loss": 0.0177, "step": 67210 }, { "grad_norm": 0.2837110161781311, "learning_rate": 6.81736717034428e-06, "loss": 0.0289, "step": 67220 }, { "grad_norm": 0.16710184514522552, "learning_rate": 6.806952194109933e-06, "loss": 0.025, "step": 67230 }, { "grad_norm": 0.4128713309764862, "learning_rate": 6.796544598386112e-06, "loss": 0.0177, "step": 67240 }, { "grad_norm": 0.3834582567214966, "learning_rate": 6.786144384951204e-06, "loss": 0.0207, "step": 67250 }, { "grad_norm": 0.3584897816181183, "learning_rate": 6.775751555582322e-06, "loss": 0.0211, "step": 67260 }, { "grad_norm": 0.2620667517185211, "learning_rate": 6.76536611205531e-06, "loss": 0.0184, "step": 67270 }, { "grad_norm": 0.20715339481830597, "learning_rate": 6.754988056144762e-06, "loss": 0.0198, "step": 67280 }, { "grad_norm": 0.33808669447898865, "learning_rate": 6.744617389624014e-06, "loss": 0.0166, "step": 67290 }, { "grad_norm": 0.3373194932937622, "learning_rate": 6.734254114265087e-06, "loss": 0.0215, "step": 67300 }, { "grad_norm": 0.39307624101638794, "learning_rate": 6.723898231838843e-06, "loss": 0.0168, "step": 67310 }, { "grad_norm": 0.2365100383758545, "learning_rate": 6.713549744114766e-06, "loss": 0.0202, "step": 67320 }, { "grad_norm": 0.25810593366622925, "learning_rate": 6.703208652861159e-06, "loss": 0.0167, "step": 67330 }, { "grad_norm": 0.23151688277721405, "learning_rate": 6.69287495984503e-06, "loss": 0.0272, "step": 67340 }, { "grad_norm": 0.18358145654201508, "learning_rate": 6.6825486668321e-06, "loss": 0.0189, "step": 67350 }, { "grad_norm": 0.2129538655281067, "learning_rate": 6.672229775586886e-06, "loss": 0.0217, "step": 67360 }, { "grad_norm": 0.3078274428844452, "learning_rate": 6.661918287872576e-06, "loss": 0.0196, "step": 67370 }, { "grad_norm": 0.376253604888916, "learning_rate": 6.6516142054511346e-06, "loss": 0.0251, "step": 67380 }, { "grad_norm": 0.25211334228515625, "learning_rate": 6.641317530083241e-06, "loss": 0.0188, "step": 67390 }, { "grad_norm": 0.2656126916408539, "learning_rate": 6.631028263528322e-06, "loss": 0.0186, "step": 67400 }, { "grad_norm": 0.1737181544303894, "learning_rate": 6.620746407544537e-06, "loss": 0.018, "step": 67410 }, { "grad_norm": 0.2821296155452728, "learning_rate": 6.610471963888742e-06, "loss": 0.0224, "step": 67420 }, { "grad_norm": 0.4062161445617676, "learning_rate": 6.600204934316606e-06, "loss": 0.0179, "step": 67430 }, { "grad_norm": 0.1622193306684494, "learning_rate": 6.589945320582452e-06, "loss": 0.0199, "step": 67440 }, { "grad_norm": 0.3402160406112671, "learning_rate": 6.579693124439374e-06, "loss": 0.0202, "step": 67450 }, { "grad_norm": 0.20443883538246155, "learning_rate": 6.56944834763919e-06, "loss": 0.0147, "step": 67460 }, { "grad_norm": 0.344416081905365, "learning_rate": 6.5592109919324575e-06, "loss": 0.0147, "step": 67470 }, { "grad_norm": 0.2583988606929779, "learning_rate": 6.54898105906846e-06, "loss": 0.02, "step": 67480 }, { "grad_norm": 0.6186719536781311, "learning_rate": 6.5387585507952155e-06, "loss": 0.0244, "step": 67490 }, { "grad_norm": 0.3419128656387329, "learning_rate": 6.528543468859461e-06, "loss": 0.0208, "step": 67500 }, { "grad_norm": 0.2622070014476776, "learning_rate": 6.518335815006682e-06, "loss": 0.021, "step": 67510 }, { "grad_norm": 0.3357198238372803, "learning_rate": 6.5081355909810845e-06, "loss": 0.0232, "step": 67520 }, { "grad_norm": 0.21471987664699554, "learning_rate": 6.497942798525608e-06, "loss": 0.0225, "step": 67530 }, { "grad_norm": 0.5674304962158203, "learning_rate": 6.487757439381936e-06, "loss": 0.0209, "step": 67540 }, { "grad_norm": 0.34068337082862854, "learning_rate": 6.477579515290433e-06, "loss": 0.0192, "step": 67550 }, { "grad_norm": 0.23345616459846497, "learning_rate": 6.467409027990273e-06, "loss": 0.0247, "step": 67560 }, { "grad_norm": 0.48651713132858276, "learning_rate": 6.457245979219279e-06, "loss": 0.0271, "step": 67570 }, { "grad_norm": 0.3426220715045929, "learning_rate": 6.447090370714054e-06, "loss": 0.0257, "step": 67580 }, { "grad_norm": 0.29538920521736145, "learning_rate": 6.436942204209917e-06, "loss": 0.024, "step": 67590 }, { "grad_norm": 0.24971428513526917, "learning_rate": 6.4268014814408804e-06, "loss": 0.0231, "step": 67600 }, { "grad_norm": 0.2845902144908905, "learning_rate": 6.41666820413977e-06, "loss": 0.0207, "step": 67610 }, { "grad_norm": 0.4157751798629761, "learning_rate": 6.406542374038033e-06, "loss": 0.0235, "step": 67620 }, { "grad_norm": 1.208593487739563, "learning_rate": 6.396423992865935e-06, "loss": 0.0192, "step": 67630 }, { "grad_norm": 0.31311723589897156, "learning_rate": 6.386313062352412e-06, "loss": 0.0173, "step": 67640 }, { "grad_norm": 0.2512560784816742, "learning_rate": 6.376209584225152e-06, "loss": 0.0147, "step": 67650 }, { "grad_norm": 0.1832316815853119, "learning_rate": 6.366113560210557e-06, "loss": 0.0198, "step": 67660 }, { "grad_norm": 0.5881224274635315, "learning_rate": 6.356024992033766e-06, "loss": 0.0212, "step": 67670 }, { "grad_norm": 0.3167887032032013, "learning_rate": 6.345943881418648e-06, "loss": 0.018, "step": 67680 }, { "grad_norm": 0.26300057768821716, "learning_rate": 6.33587023008777e-06, "loss": 0.0179, "step": 67690 }, { "grad_norm": 0.28693529963493347, "learning_rate": 6.32580403976245e-06, "loss": 0.0211, "step": 67700 }, { "grad_norm": 0.2072140872478485, "learning_rate": 6.31574531216273e-06, "loss": 0.0221, "step": 67710 }, { "grad_norm": 0.20634402334690094, "learning_rate": 6.305694049007371e-06, "loss": 0.0188, "step": 67720 }, { "grad_norm": 0.9119769930839539, "learning_rate": 6.2956502520138575e-06, "loss": 0.02, "step": 67730 }, { "grad_norm": 0.43677017092704773, "learning_rate": 6.285613922898409e-06, "loss": 0.0195, "step": 67740 }, { "grad_norm": 0.5112770199775696, "learning_rate": 6.275585063375927e-06, "loss": 0.0205, "step": 67750 }, { "grad_norm": 0.2606634199619293, "learning_rate": 6.265563675160113e-06, "loss": 0.0184, "step": 67760 }, { "grad_norm": 0.45054587721824646, "learning_rate": 6.25554975996332e-06, "loss": 0.0172, "step": 67770 }, { "grad_norm": 0.2544254660606384, "learning_rate": 6.245543319496661e-06, "loss": 0.0182, "step": 67780 }, { "grad_norm": 0.2539328634738922, "learning_rate": 6.2355443554699685e-06, "loss": 0.0214, "step": 67790 }, { "grad_norm": 0.31470057368278503, "learning_rate": 6.225552869591766e-06, "loss": 0.0154, "step": 67800 }, { "grad_norm": 0.17660939693450928, "learning_rate": 6.215568863569365e-06, "loss": 0.0224, "step": 67810 }, { "grad_norm": 0.4639906883239746, "learning_rate": 6.205592339108712e-06, "loss": 0.0191, "step": 67820 }, { "grad_norm": 0.4660347104072571, "learning_rate": 6.195623297914577e-06, "loss": 0.0137, "step": 67830 }, { "grad_norm": 0.20077745616436005, "learning_rate": 6.185661741690357e-06, "loss": 0.0177, "step": 67840 }, { "grad_norm": 0.09229013323783875, "learning_rate": 6.1757076721382145e-06, "loss": 0.0242, "step": 67850 }, { "grad_norm": 0.42447033524513245, "learning_rate": 6.165761090959038e-06, "loss": 0.0267, "step": 67860 }, { "grad_norm": 0.12981240451335907, "learning_rate": 6.155821999852424e-06, "loss": 0.0178, "step": 67870 }, { "grad_norm": 0.1540789157152176, "learning_rate": 6.145890400516696e-06, "loss": 0.0164, "step": 67880 }, { "grad_norm": 0.20211298763751984, "learning_rate": 6.1359662946488816e-06, "loss": 0.0199, "step": 67890 }, { "grad_norm": 0.16454780101776123, "learning_rate": 6.12604968394474e-06, "loss": 0.0163, "step": 67900 }, { "grad_norm": 0.3162687122821808, "learning_rate": 6.11614057009875e-06, "loss": 0.016, "step": 67910 }, { "grad_norm": 0.41267985105514526, "learning_rate": 6.106238954804111e-06, "loss": 0.0198, "step": 67920 }, { "grad_norm": 0.4298267364501953, "learning_rate": 6.096344839752738e-06, "loss": 0.0199, "step": 67930 }, { "grad_norm": 0.2585894763469696, "learning_rate": 6.086458226635278e-06, "loss": 0.0177, "step": 67940 }, { "grad_norm": 0.27038246393203735, "learning_rate": 6.076579117141046e-06, "loss": 0.0272, "step": 67950 }, { "grad_norm": 0.5115615129470825, "learning_rate": 6.066707512958153e-06, "loss": 0.0213, "step": 67960 }, { "grad_norm": 0.4706535041332245, "learning_rate": 6.056843415773361e-06, "loss": 0.0219, "step": 67970 }, { "grad_norm": 0.2320690155029297, "learning_rate": 6.0469868272721776e-06, "loss": 0.0163, "step": 67980 }, { "grad_norm": 0.1879487782716751, "learning_rate": 6.037137749138844e-06, "loss": 0.0201, "step": 67990 }, { "grad_norm": 0.31955188512802124, "learning_rate": 6.027296183056252e-06, "loss": 0.0257, "step": 68000 }, { "grad_norm": 0.28853198885917664, "learning_rate": 6.017462130706114e-06, "loss": 0.0167, "step": 68010 }, { "grad_norm": 0.39960411190986633, "learning_rate": 6.0076355937687516e-06, "loss": 0.0219, "step": 68020 }, { "grad_norm": 0.1953015774488449, "learning_rate": 5.9978165739232925e-06, "loss": 0.0173, "step": 68030 }, { "grad_norm": 0.4621765911579132, "learning_rate": 5.988005072847508e-06, "loss": 0.0193, "step": 68040 }, { "grad_norm": 0.23885831236839294, "learning_rate": 5.978201092217928e-06, "loss": 0.0244, "step": 68050 }, { "grad_norm": 0.23405255377292633, "learning_rate": 5.9684046337097895e-06, "loss": 0.0136, "step": 68060 }, { "grad_norm": 0.24720726907253265, "learning_rate": 5.958615698997016e-06, "loss": 0.0227, "step": 68070 }, { "grad_norm": 0.33093446493148804, "learning_rate": 5.948834289752303e-06, "loss": 0.0168, "step": 68080 }, { "grad_norm": 0.3580750524997711, "learning_rate": 5.939060407646996e-06, "loss": 0.0221, "step": 68090 }, { "grad_norm": 0.22326253354549408, "learning_rate": 5.929294054351198e-06, "loss": 0.0199, "step": 68100 }, { "grad_norm": 0.28933221101760864, "learning_rate": 5.919535231533707e-06, "loss": 0.016, "step": 68110 }, { "grad_norm": 0.41419318318367004, "learning_rate": 5.909783940862046e-06, "loss": 0.0184, "step": 68120 }, { "grad_norm": 0.28712138533592224, "learning_rate": 5.900040184002436e-06, "loss": 0.0178, "step": 68130 }, { "grad_norm": 0.3171223998069763, "learning_rate": 5.890303962619831e-06, "loss": 0.0195, "step": 68140 }, { "grad_norm": 0.25694751739501953, "learning_rate": 5.88057527837787e-06, "loss": 0.0191, "step": 68150 }, { "grad_norm": 0.25929325819015503, "learning_rate": 5.870854132938924e-06, "loss": 0.0194, "step": 68160 }, { "grad_norm": 0.23776479065418243, "learning_rate": 5.861140527964071e-06, "loss": 0.0211, "step": 68170 }, { "grad_norm": 0.3221209943294525, "learning_rate": 5.851434465113098e-06, "loss": 0.0133, "step": 68180 }, { "grad_norm": 0.21623557806015015, "learning_rate": 5.841735946044524e-06, "loss": 0.0193, "step": 68190 }, { "grad_norm": 0.33866605162620544, "learning_rate": 5.832044972415523e-06, "loss": 0.018, "step": 68200 }, { "grad_norm": 0.2476108968257904, "learning_rate": 5.8223615458820655e-06, "loss": 0.0169, "step": 68210 }, { "grad_norm": 0.4341956377029419, "learning_rate": 5.812685668098733e-06, "loss": 0.0214, "step": 68220 }, { "grad_norm": 0.3266698718070984, "learning_rate": 5.803017340718913e-06, "loss": 0.0239, "step": 68230 }, { "grad_norm": 0.4489668905735016, "learning_rate": 5.793356565394636e-06, "loss": 0.0225, "step": 68240 }, { "grad_norm": 0.2707217037677765, "learning_rate": 5.783703343776658e-06, "loss": 0.0165, "step": 68250 }, { "grad_norm": 0.3668607771396637, "learning_rate": 5.774057677514477e-06, "loss": 0.0155, "step": 68260 }, { "grad_norm": 0.2826434373855591, "learning_rate": 5.764419568256235e-06, "loss": 0.0227, "step": 68270 }, { "grad_norm": 0.5511506199836731, "learning_rate": 5.754789017648865e-06, "loss": 0.0242, "step": 68280 }, { "grad_norm": 0.394911527633667, "learning_rate": 5.745166027337934e-06, "loss": 0.0239, "step": 68290 }, { "grad_norm": 0.808316171169281, "learning_rate": 5.735550598967754e-06, "loss": 0.0206, "step": 68300 }, { "grad_norm": 0.2782607972621918, "learning_rate": 5.725942734181339e-06, "loss": 0.0215, "step": 68310 }, { "grad_norm": 0.14648807048797607, "learning_rate": 5.7163424346204146e-06, "loss": 0.0177, "step": 68320 }, { "grad_norm": 0.3518840968608856, "learning_rate": 5.706749701925407e-06, "loss": 0.0189, "step": 68330 }, { "grad_norm": 0.13741245865821838, "learning_rate": 5.697164537735461e-06, "loss": 0.0194, "step": 68340 }, { "grad_norm": 0.2335474044084549, "learning_rate": 5.687586943688406e-06, "loss": 0.0195, "step": 68350 }, { "grad_norm": 0.27628493309020996, "learning_rate": 5.678016921420792e-06, "loss": 0.0173, "step": 68360 }, { "grad_norm": 0.41003847122192383, "learning_rate": 5.668454472567875e-06, "loss": 0.0207, "step": 68370 }, { "grad_norm": 0.42358556389808655, "learning_rate": 5.658899598763617e-06, "loss": 0.0189, "step": 68380 }, { "grad_norm": 0.30736809968948364, "learning_rate": 5.6493523016407035e-06, "loss": 0.015, "step": 68390 }, { "grad_norm": 0.16427096724510193, "learning_rate": 5.6398125828304615e-06, "loss": 0.0144, "step": 68400 }, { "grad_norm": 0.3907046318054199, "learning_rate": 5.630280443963015e-06, "loss": 0.0229, "step": 68410 }, { "grad_norm": 0.23960717022418976, "learning_rate": 5.620755886667112e-06, "loss": 0.0131, "step": 68420 }, { "grad_norm": 0.18240521848201752, "learning_rate": 5.611238912570271e-06, "loss": 0.0181, "step": 68430 }, { "grad_norm": 0.47068876028060913, "learning_rate": 5.601729523298649e-06, "loss": 0.0184, "step": 68440 }, { "grad_norm": 0.47709324955940247, "learning_rate": 5.592227720477161e-06, "loss": 0.0179, "step": 68450 }, { "grad_norm": 0.21864347159862518, "learning_rate": 5.582733505729415e-06, "loss": 0.0232, "step": 68460 }, { "grad_norm": 0.19688652455806732, "learning_rate": 5.573246880677668e-06, "loss": 0.0182, "step": 68470 }, { "grad_norm": 0.21980401873588562, "learning_rate": 5.56376784694298e-06, "loss": 0.0163, "step": 68480 }, { "grad_norm": 0.3494890630245209, "learning_rate": 5.554296406145027e-06, "loss": 0.025, "step": 68490 }, { "grad_norm": 1.1603184938430786, "learning_rate": 5.544832559902219e-06, "loss": 0.0247, "step": 68500 }, { "grad_norm": 0.5939328670501709, "learning_rate": 5.53537630983168e-06, "loss": 0.0269, "step": 68510 }, { "grad_norm": 0.12816749513149261, "learning_rate": 5.5259276575492125e-06, "loss": 0.0259, "step": 68520 }, { "grad_norm": 0.22191160917282104, "learning_rate": 5.516486604669357e-06, "loss": 0.0202, "step": 68530 }, { "grad_norm": 0.238425612449646, "learning_rate": 5.507053152805303e-06, "loss": 0.0205, "step": 68540 }, { "grad_norm": 0.40925851464271545, "learning_rate": 5.497627303568975e-06, "loss": 0.0234, "step": 68550 }, { "grad_norm": 0.14112187922000885, "learning_rate": 5.488209058571003e-06, "loss": 0.0141, "step": 68560 }, { "grad_norm": 0.2831352949142456, "learning_rate": 5.4787984194207054e-06, "loss": 0.0206, "step": 68570 }, { "grad_norm": 0.20873858034610748, "learning_rate": 5.469395387726095e-06, "loss": 0.0236, "step": 68580 }, { "grad_norm": 0.6043825149536133, "learning_rate": 5.45999996509391e-06, "loss": 0.0181, "step": 68590 }, { "grad_norm": 0.1596972644329071, "learning_rate": 5.450612153129536e-06, "loss": 0.0203, "step": 68600 }, { "grad_norm": 0.1758805513381958, "learning_rate": 5.4412319534371426e-06, "loss": 0.0158, "step": 68610 }, { "grad_norm": 0.23230725526809692, "learning_rate": 5.431859367619513e-06, "loss": 0.0184, "step": 68620 }, { "grad_norm": 0.24749639630317688, "learning_rate": 5.422494397278172e-06, "loss": 0.0185, "step": 68630 }, { "grad_norm": 0.42399054765701294, "learning_rate": 5.413137044013344e-06, "loss": 0.0193, "step": 68640 }, { "grad_norm": 0.2766035795211792, "learning_rate": 5.403787309423941e-06, "loss": 0.0222, "step": 68650 }, { "grad_norm": 0.32674044370651245, "learning_rate": 5.3944451951075835e-06, "loss": 0.0161, "step": 68660 }, { "grad_norm": 0.44764378666877747, "learning_rate": 5.385110702660562e-06, "loss": 0.0203, "step": 68670 }, { "grad_norm": 0.29436272382736206, "learning_rate": 5.375783833677922e-06, "loss": 0.0209, "step": 68680 }, { "grad_norm": 0.24390096962451935, "learning_rate": 5.36646458975334e-06, "loss": 0.0177, "step": 68690 }, { "grad_norm": 0.32742738723754883, "learning_rate": 5.3571529724792294e-06, "loss": 0.0156, "step": 68700 }, { "grad_norm": 0.3545425832271576, "learning_rate": 5.347848983446702e-06, "loss": 0.0208, "step": 68710 }, { "grad_norm": 0.22303539514541626, "learning_rate": 5.3385526242455185e-06, "loss": 0.0153, "step": 68720 }, { "grad_norm": 0.4799765348434448, "learning_rate": 5.329263896464226e-06, "loss": 0.0167, "step": 68730 }, { "grad_norm": 1.0918492078781128, "learning_rate": 5.3199828016899715e-06, "loss": 0.0235, "step": 68740 }, { "grad_norm": 0.23492686450481415, "learning_rate": 5.310709341508657e-06, "loss": 0.0192, "step": 68750 }, { "grad_norm": 0.16820098459720612, "learning_rate": 5.301443517504861e-06, "loss": 0.0175, "step": 68760 }, { "grad_norm": 0.31399428844451904, "learning_rate": 5.292185331261862e-06, "loss": 0.0136, "step": 68770 }, { "grad_norm": 0.24448321759700775, "learning_rate": 5.28293478436162e-06, "loss": 0.0186, "step": 68780 }, { "grad_norm": 0.22490514814853668, "learning_rate": 5.273691878384829e-06, "loss": 0.0198, "step": 68790 }, { "grad_norm": 0.3213980793952942, "learning_rate": 5.264456614910812e-06, "loss": 0.0159, "step": 68800 }, { "grad_norm": 0.2287832796573639, "learning_rate": 5.255228995517647e-06, "loss": 0.0185, "step": 68810 }, { "grad_norm": 0.2152387499809265, "learning_rate": 5.246009021782067e-06, "loss": 0.0161, "step": 68820 }, { "grad_norm": 0.20161683857440948, "learning_rate": 5.2367966952795225e-06, "loss": 0.0203, "step": 68830 }, { "grad_norm": 0.19481654465198517, "learning_rate": 5.2275920175841485e-06, "loss": 0.0167, "step": 68840 }, { "grad_norm": 0.12054982781410217, "learning_rate": 5.218394990268771e-06, "loss": 0.0167, "step": 68850 }, { "grad_norm": 0.5936324596405029, "learning_rate": 5.209205614904916e-06, "loss": 0.023, "step": 68860 }, { "grad_norm": 0.38775137066841125, "learning_rate": 5.200023893062777e-06, "loss": 0.0184, "step": 68870 }, { "grad_norm": 2.423717737197876, "learning_rate": 5.190849826311289e-06, "loss": 0.0176, "step": 68880 }, { "grad_norm": 2.204887866973877, "learning_rate": 5.181683416218025e-06, "loss": 0.0215, "step": 68890 }, { "grad_norm": 0.27516114711761475, "learning_rate": 5.172524664349276e-06, "loss": 0.0161, "step": 68900 }, { "grad_norm": 1.3237121105194092, "learning_rate": 5.1633735722700416e-06, "loss": 0.0177, "step": 68910 }, { "grad_norm": 0.1741948425769806, "learning_rate": 5.154230141543958e-06, "loss": 0.0185, "step": 68920 }, { "grad_norm": 0.24409843981266022, "learning_rate": 5.145094373733433e-06, "loss": 0.0198, "step": 68930 }, { "grad_norm": 0.11938358098268509, "learning_rate": 5.135966270399478e-06, "loss": 0.0197, "step": 68940 }, { "grad_norm": 0.3920260965824127, "learning_rate": 5.126845833101857e-06, "loss": 0.0174, "step": 68950 }, { "grad_norm": 0.30792132019996643, "learning_rate": 5.1177330633989994e-06, "loss": 0.0169, "step": 68960 }, { "grad_norm": 0.3957652151584625, "learning_rate": 5.108627962848033e-06, "loss": 0.0282, "step": 68970 }, { "grad_norm": 0.514990508556366, "learning_rate": 5.099530533004759e-06, "loss": 0.0183, "step": 68980 }, { "grad_norm": 0.30506056547164917, "learning_rate": 5.090440775423699e-06, "loss": 0.0155, "step": 68990 }, { "grad_norm": 0.2645440995693207, "learning_rate": 5.081358691658022e-06, "loss": 0.0212, "step": 69000 }, { "grad_norm": 0.4318664073944092, "learning_rate": 5.072284283259621e-06, "loss": 0.0256, "step": 69010 }, { "grad_norm": 0.17371146380901337, "learning_rate": 5.063217551779053e-06, "loss": 0.0126, "step": 69020 }, { "grad_norm": 0.4226192235946655, "learning_rate": 5.054158498765588e-06, "loss": 0.0179, "step": 69030 }, { "grad_norm": 0.36200472712516785, "learning_rate": 5.0451071257671625e-06, "loss": 0.0176, "step": 69040 }, { "grad_norm": 0.3932947516441345, "learning_rate": 5.036063434330407e-06, "loss": 0.016, "step": 69050 }, { "grad_norm": 0.1463812291622162, "learning_rate": 5.027027426000652e-06, "loss": 0.017, "step": 69060 }, { "grad_norm": 0.41196316480636597, "learning_rate": 5.017999102321886e-06, "loss": 0.016, "step": 69070 }, { "grad_norm": 0.28936344385147095, "learning_rate": 5.0089784648368224e-06, "loss": 0.0177, "step": 69080 }, { "grad_norm": 0.1518372893333435, "learning_rate": 4.999965515086829e-06, "loss": 0.0183, "step": 69090 }, { "grad_norm": 0.3763445317745209, "learning_rate": 4.990960254611976e-06, "loss": 0.0162, "step": 69100 }, { "grad_norm": 0.4230433404445648, "learning_rate": 4.9819626849510194e-06, "loss": 0.0229, "step": 69110 }, { "grad_norm": 0.5099437236785889, "learning_rate": 4.97297280764138e-06, "loss": 0.0281, "step": 69120 }, { "grad_norm": 0.21256260573863983, "learning_rate": 4.963990624219211e-06, "loss": 0.02, "step": 69130 }, { "grad_norm": 0.317282497882843, "learning_rate": 4.955016136219298e-06, "loss": 0.0168, "step": 69140 }, { "grad_norm": 0.23481477797031403, "learning_rate": 4.946049345175146e-06, "loss": 0.0163, "step": 69150 }, { "grad_norm": 0.35429447889328003, "learning_rate": 4.937090252618937e-06, "loss": 0.0175, "step": 69160 }, { "grad_norm": 0.2695292532444, "learning_rate": 4.928138860081521e-06, "loss": 0.0213, "step": 69170 }, { "grad_norm": 0.13311149179935455, "learning_rate": 4.919195169092472e-06, "loss": 0.0178, "step": 69180 }, { "grad_norm": 0.19646857678890228, "learning_rate": 4.910259181179994e-06, "loss": 0.0224, "step": 69190 }, { "grad_norm": 0.3844822347164154, "learning_rate": 4.901330897871015e-06, "loss": 0.0203, "step": 69200 }, { "grad_norm": 0.4002983272075653, "learning_rate": 4.8924103206911375e-06, "loss": 0.016, "step": 69210 }, { "grad_norm": 0.47427845001220703, "learning_rate": 4.883497451164637e-06, "loss": 0.0309, "step": 69220 }, { "grad_norm": 0.42259976267814636, "learning_rate": 4.874592290814484e-06, "loss": 0.0164, "step": 69230 }, { "grad_norm": 0.6220788359642029, "learning_rate": 4.865694841162327e-06, "loss": 0.0269, "step": 69240 }, { "grad_norm": 0.5685251355171204, "learning_rate": 4.856805103728496e-06, "loss": 0.0206, "step": 69250 }, { "grad_norm": 0.3316754698753357, "learning_rate": 4.847923080032007e-06, "loss": 0.0154, "step": 69260 }, { "grad_norm": 0.2280765175819397, "learning_rate": 4.8390487715905486e-06, "loss": 0.0181, "step": 69270 }, { "grad_norm": 0.2466830611228943, "learning_rate": 4.8301821799205e-06, "loss": 0.0148, "step": 69280 }, { "grad_norm": 0.17477896809577942, "learning_rate": 4.821323306536918e-06, "loss": 0.0189, "step": 69290 }, { "grad_norm": 0.2378297597169876, "learning_rate": 4.8124721529535455e-06, "loss": 0.0176, "step": 69300 }, { "grad_norm": 0.2897128462791443, "learning_rate": 4.803628720682807e-06, "loss": 0.023, "step": 69310 }, { "grad_norm": 0.29597339034080505, "learning_rate": 4.794793011235776e-06, "loss": 0.0232, "step": 69320 }, { "grad_norm": 0.3637852668762207, "learning_rate": 4.78596502612228e-06, "loss": 0.0218, "step": 69330 }, { "grad_norm": 1.2930428981781006, "learning_rate": 4.777144766850738e-06, "loss": 0.0223, "step": 69340 }, { "grad_norm": 0.19489364326000214, "learning_rate": 4.768332234928313e-06, "loss": 0.0161, "step": 69350 }, { "grad_norm": 0.2084302455186844, "learning_rate": 4.759527431860828e-06, "loss": 0.0161, "step": 69360 }, { "grad_norm": 0.4008682370185852, "learning_rate": 4.750730359152755e-06, "loss": 0.0192, "step": 69370 }, { "grad_norm": 0.20092718303203583, "learning_rate": 4.741941018307311e-06, "loss": 0.0191, "step": 69380 }, { "grad_norm": 0.5036243200302124, "learning_rate": 4.733159410826321e-06, "loss": 0.0207, "step": 69390 }, { "grad_norm": 0.20363733172416687, "learning_rate": 4.724385538210357e-06, "loss": 0.0168, "step": 69400 }, { "grad_norm": 0.27584895491600037, "learning_rate": 4.715619401958599e-06, "loss": 0.0188, "step": 69410 }, { "grad_norm": 0.3017244040966034, "learning_rate": 4.706861003568958e-06, "loss": 0.0202, "step": 69420 }, { "grad_norm": 0.26788023114204407, "learning_rate": 4.698110344538003e-06, "loss": 0.0196, "step": 69430 }, { "grad_norm": 0.22134685516357422, "learning_rate": 4.689367426360975e-06, "loss": 0.0185, "step": 69440 }, { "grad_norm": 0.3317965865135193, "learning_rate": 4.680632250531819e-06, "loss": 0.0184, "step": 69450 }, { "grad_norm": 0.24945209920406342, "learning_rate": 4.671904818543115e-06, "loss": 0.0176, "step": 69460 }, { "grad_norm": 0.16141517460346222, "learning_rate": 4.66318513188615e-06, "loss": 0.0161, "step": 69470 }, { "grad_norm": 1.4344983100891113, "learning_rate": 4.654473192050884e-06, "loss": 0.0164, "step": 69480 }, { "grad_norm": 0.24687187373638153, "learning_rate": 4.64576900052594e-06, "loss": 0.0162, "step": 69490 }, { "grad_norm": 0.1573270559310913, "learning_rate": 4.637072558798638e-06, "loss": 0.0153, "step": 69500 }, { "grad_norm": 0.20247583091259003, "learning_rate": 4.628383868354969e-06, "loss": 0.0182, "step": 69510 }, { "grad_norm": 0.42142391204833984, "learning_rate": 4.6197029306795595e-06, "loss": 0.014, "step": 69520 }, { "grad_norm": 0.3204450011253357, "learning_rate": 4.611029747255779e-06, "loss": 0.0167, "step": 69530 }, { "grad_norm": 0.3367277681827545, "learning_rate": 4.6023643195656164e-06, "loss": 0.0144, "step": 69540 }, { "grad_norm": 0.06996982544660568, "learning_rate": 4.593706649089768e-06, "loss": 0.0139, "step": 69550 }, { "grad_norm": 0.42225226759910583, "learning_rate": 4.585056737307597e-06, "loss": 0.0194, "step": 69560 }, { "grad_norm": 0.2223326712846756, "learning_rate": 4.576414585697103e-06, "loss": 0.0156, "step": 69570 }, { "grad_norm": 0.3130648136138916, "learning_rate": 4.567780195735044e-06, "loss": 0.0238, "step": 69580 }, { "grad_norm": 0.31334826350212097, "learning_rate": 4.559153568896757e-06, "loss": 0.0111, "step": 69590 }, { "grad_norm": 0.5114609599113464, "learning_rate": 4.550534706656329e-06, "loss": 0.017, "step": 69600 }, { "grad_norm": 0.2541210949420929, "learning_rate": 4.541923610486465e-06, "loss": 0.0147, "step": 69610 }, { "grad_norm": 0.2261892408132553, "learning_rate": 4.533320281858578e-06, "loss": 0.0192, "step": 69620 }, { "grad_norm": 0.4380232095718384, "learning_rate": 4.52472472224274e-06, "loss": 0.0226, "step": 69630 }, { "grad_norm": 0.2827264368534088, "learning_rate": 4.51613693310769e-06, "loss": 0.0211, "step": 69640 }, { "grad_norm": 0.3672930598258972, "learning_rate": 4.507556915920868e-06, "loss": 0.026, "step": 69650 }, { "grad_norm": 0.18776100873947144, "learning_rate": 4.498984672148332e-06, "loss": 0.0158, "step": 69660 }, { "grad_norm": 0.4868040382862091, "learning_rate": 4.490420203254864e-06, "loss": 0.0226, "step": 69670 }, { "grad_norm": 0.7418146729469299, "learning_rate": 4.481863510703893e-06, "loss": 0.0199, "step": 69680 }, { "grad_norm": 0.30355170369148254, "learning_rate": 4.473314595957523e-06, "loss": 0.0172, "step": 69690 }, { "grad_norm": 0.4704231023788452, "learning_rate": 4.464773460476535e-06, "loss": 0.016, "step": 69700 }, { "grad_norm": 0.2177627682685852, "learning_rate": 4.456240105720372e-06, "loss": 0.0172, "step": 69710 }, { "grad_norm": 0.19636477530002594, "learning_rate": 4.4477145331471405e-06, "loss": 0.024, "step": 69720 }, { "grad_norm": 0.4544883668422699, "learning_rate": 4.439196744213653e-06, "loss": 0.0191, "step": 69730 }, { "grad_norm": 0.2272215336561203, "learning_rate": 4.430686740375339e-06, "loss": 0.0191, "step": 69740 }, { "grad_norm": 0.27556276321411133, "learning_rate": 4.422184523086342e-06, "loss": 0.0164, "step": 69750 }, { "grad_norm": 0.26275503635406494, "learning_rate": 4.41369009379946e-06, "loss": 0.0165, "step": 69760 }, { "grad_norm": 0.3378657102584839, "learning_rate": 4.405203453966139e-06, "loss": 0.0183, "step": 69770 }, { "grad_norm": 0.2659624218940735, "learning_rate": 4.396724605036539e-06, "loss": 0.0173, "step": 69780 }, { "grad_norm": 0.1929461658000946, "learning_rate": 4.388253548459437e-06, "loss": 0.0213, "step": 69790 }, { "grad_norm": 0.34017881751060486, "learning_rate": 4.3797902856823395e-06, "loss": 0.0165, "step": 69800 }, { "grad_norm": 0.33570802211761475, "learning_rate": 4.371334818151357e-06, "loss": 0.0163, "step": 69810 }, { "grad_norm": 0.36710500717163086, "learning_rate": 4.362887147311306e-06, "loss": 0.0151, "step": 69820 }, { "grad_norm": 1.6455870866775513, "learning_rate": 4.354447274605672e-06, "loss": 0.0227, "step": 69830 }, { "grad_norm": 0.24946053326129913, "learning_rate": 4.346015201476572e-06, "loss": 0.021, "step": 69840 }, { "grad_norm": 0.25626811385154724, "learning_rate": 4.337590929364855e-06, "loss": 0.0171, "step": 69850 }, { "grad_norm": 0.7024551033973694, "learning_rate": 4.329174459709973e-06, "loss": 0.0166, "step": 69860 }, { "grad_norm": 0.3603462278842926, "learning_rate": 4.320765793950071e-06, "loss": 0.0184, "step": 69870 }, { "grad_norm": 0.26926758885383606, "learning_rate": 4.312364933521962e-06, "loss": 0.0203, "step": 69880 }, { "grad_norm": 0.23506461083889008, "learning_rate": 4.303971879861129e-06, "loss": 0.0134, "step": 69890 }, { "grad_norm": 0.21935077011585236, "learning_rate": 4.295586634401716e-06, "loss": 0.0207, "step": 69900 }, { "grad_norm": 0.16765588521957397, "learning_rate": 4.287209198576536e-06, "loss": 0.0152, "step": 69910 }, { "grad_norm": 1.0024064779281616, "learning_rate": 4.278839573817045e-06, "loss": 0.019, "step": 69920 }, { "grad_norm": 0.2715826630592346, "learning_rate": 4.270477761553399e-06, "loss": 0.023, "step": 69930 }, { "grad_norm": 0.15918660163879395, "learning_rate": 4.262123763214393e-06, "loss": 0.0139, "step": 69940 }, { "grad_norm": 0.3052125871181488, "learning_rate": 4.253777580227508e-06, "loss": 0.0178, "step": 69950 }, { "grad_norm": 0.192348450422287, "learning_rate": 4.2454392140188755e-06, "loss": 0.0164, "step": 69960 }, { "grad_norm": 0.16377972066402435, "learning_rate": 4.2371086660132785e-06, "loss": 0.02, "step": 69970 }, { "grad_norm": 0.6731185913085938, "learning_rate": 4.228785937634205e-06, "loss": 0.0148, "step": 69980 }, { "grad_norm": 0.10671055316925049, "learning_rate": 4.220471030303758e-06, "loss": 0.0148, "step": 69990 }, { "grad_norm": 0.22601290047168732, "learning_rate": 4.212163945442754e-06, "loss": 0.0166, "step": 70000 }, { "grad_norm": 0.3343638479709625, "learning_rate": 4.203864684470621e-06, "loss": 0.0165, "step": 70010 }, { "grad_norm": 0.32869189977645874, "learning_rate": 4.19557324880549e-06, "loss": 0.0172, "step": 70020 }, { "grad_norm": 0.35592758655548096, "learning_rate": 4.187289639864145e-06, "loss": 0.0183, "step": 70030 }, { "grad_norm": 0.2113570272922516, "learning_rate": 4.1790138590619974e-06, "loss": 0.0126, "step": 70040 }, { "grad_norm": 0.19283415377140045, "learning_rate": 4.170745907813195e-06, "loss": 0.0183, "step": 70050 }, { "grad_norm": 0.1559048593044281, "learning_rate": 4.162485787530479e-06, "loss": 0.0178, "step": 70060 }, { "grad_norm": 0.5732389092445374, "learning_rate": 4.154233499625282e-06, "loss": 0.0179, "step": 70070 }, { "grad_norm": 0.23533853888511658, "learning_rate": 4.145989045507692e-06, "loss": 0.0206, "step": 70080 }, { "grad_norm": 0.32242801785469055, "learning_rate": 4.1377524265864666e-06, "loss": 0.0216, "step": 70090 }, { "grad_norm": 0.18093349039554596, "learning_rate": 4.1295236442690175e-06, "loss": 0.0142, "step": 70100 }, { "grad_norm": 0.2119293063879013, "learning_rate": 4.121302699961421e-06, "loss": 0.0161, "step": 70110 }, { "grad_norm": 0.3097749948501587, "learning_rate": 4.113089595068403e-06, "loss": 0.0224, "step": 70120 }, { "grad_norm": 0.32974809408187866, "learning_rate": 4.104884330993364e-06, "loss": 0.0203, "step": 70130 }, { "grad_norm": 0.2846795618534088, "learning_rate": 4.0966869091383585e-06, "loss": 0.021, "step": 70140 }, { "grad_norm": 0.2428632378578186, "learning_rate": 4.088497330904101e-06, "loss": 0.0164, "step": 70150 }, { "grad_norm": 0.28059008717536926, "learning_rate": 4.080315597689976e-06, "loss": 0.0201, "step": 70160 }, { "grad_norm": 0.16271327435970306, "learning_rate": 4.072141710893995e-06, "loss": 0.0255, "step": 70170 }, { "grad_norm": 0.20492316782474518, "learning_rate": 4.063975671912879e-06, "loss": 0.0224, "step": 70180 }, { "grad_norm": 0.7973341941833496, "learning_rate": 4.055817482141949e-06, "loss": 0.0184, "step": 70190 }, { "grad_norm": 0.18066757917404175, "learning_rate": 4.047667142975259e-06, "loss": 0.0186, "step": 70200 }, { "grad_norm": 0.3764706254005432, "learning_rate": 4.039524655805443e-06, "loss": 0.0229, "step": 70210 }, { "grad_norm": 0.2704368829727173, "learning_rate": 4.03139002202384e-06, "loss": 0.0165, "step": 70220 }, { "grad_norm": 0.23126769065856934, "learning_rate": 4.023263243020447e-06, "loss": 0.0197, "step": 70230 }, { "grad_norm": 0.2429526299238205, "learning_rate": 4.015144320183884e-06, "loss": 0.0206, "step": 70240 }, { "grad_norm": 0.1929822862148285, "learning_rate": 4.007033254901482e-06, "loss": 0.0156, "step": 70250 }, { "grad_norm": 0.181778684258461, "learning_rate": 3.9989300485591795e-06, "loss": 0.0193, "step": 70260 }, { "grad_norm": 0.2747252583503723, "learning_rate": 3.990834702541601e-06, "loss": 0.0269, "step": 70270 }, { "grad_norm": 0.3895820677280426, "learning_rate": 3.982747218232019e-06, "loss": 0.0212, "step": 70280 }, { "grad_norm": 0.3531956076622009, "learning_rate": 3.97466759701236e-06, "loss": 0.0229, "step": 70290 }, { "grad_norm": 0.3252827227115631, "learning_rate": 3.966595840263226e-06, "loss": 0.0173, "step": 70300 }, { "grad_norm": 0.2714249789714813, "learning_rate": 3.958531949363836e-06, "loss": 0.0209, "step": 70310 }, { "grad_norm": 0.24668948352336884, "learning_rate": 3.950475925692098e-06, "loss": 0.0257, "step": 70320 }, { "grad_norm": 0.20577777922153473, "learning_rate": 3.9424277706245685e-06, "loss": 0.0158, "step": 70330 }, { "grad_norm": 0.3076515793800354, "learning_rate": 3.934387485536451e-06, "loss": 0.017, "step": 70340 }, { "grad_norm": 0.4146469235420227, "learning_rate": 3.926355071801619e-06, "loss": 0.016, "step": 70350 }, { "grad_norm": 0.23038485646247864, "learning_rate": 3.9183305307925965e-06, "loss": 0.0173, "step": 70360 }, { "grad_norm": 0.2503615617752075, "learning_rate": 3.910313863880533e-06, "loss": 0.0171, "step": 70370 }, { "grad_norm": 0.3042183816432953, "learning_rate": 3.902305072435292e-06, "loss": 0.0138, "step": 70380 }, { "grad_norm": 0.29993098974227905, "learning_rate": 3.894304157825329e-06, "loss": 0.0158, "step": 70390 }, { "grad_norm": 0.11189917474985123, "learning_rate": 3.886311121417791e-06, "loss": 0.017, "step": 70400 }, { "grad_norm": 0.3067598044872284, "learning_rate": 3.878325964578472e-06, "loss": 0.0153, "step": 70410 }, { "grad_norm": 0.2575879991054535, "learning_rate": 3.870348688671815e-06, "loss": 0.0147, "step": 70420 }, { "grad_norm": 0.22027894854545593, "learning_rate": 3.862379295060931e-06, "loss": 0.0156, "step": 70430 }, { "grad_norm": 0.3221525549888611, "learning_rate": 3.854417785107534e-06, "loss": 0.0272, "step": 70440 }, { "grad_norm": 0.27584001421928406, "learning_rate": 3.8464641601720755e-06, "loss": 0.0192, "step": 70450 }, { "grad_norm": 0.3711896538734436, "learning_rate": 3.838518421613579e-06, "loss": 0.0204, "step": 70460 }, { "grad_norm": 0.536460280418396, "learning_rate": 3.830580570789766e-06, "loss": 0.022, "step": 70470 }, { "grad_norm": 0.3854113221168518, "learning_rate": 3.822650609057005e-06, "loss": 0.0177, "step": 70480 }, { "grad_norm": 0.3619668483734131, "learning_rate": 3.814728537770285e-06, "loss": 0.0149, "step": 70490 }, { "grad_norm": 0.2650378346443176, "learning_rate": 3.806814358283306e-06, "loss": 0.0186, "step": 70500 }, { "grad_norm": 0.7930605411529541, "learning_rate": 3.7989080719483596e-06, "loss": 0.0203, "step": 70510 }, { "grad_norm": 0.11911597102880478, "learning_rate": 3.7910096801164143e-06, "loss": 0.0159, "step": 70520 }, { "grad_norm": 0.41492927074432373, "learning_rate": 3.7831191841371016e-06, "loss": 0.0185, "step": 70530 }, { "grad_norm": 0.36492642760276794, "learning_rate": 3.775236585358688e-06, "loss": 0.0176, "step": 70540 }, { "grad_norm": 0.6160640120506287, "learning_rate": 3.7673618851280843e-06, "loss": 0.0208, "step": 70550 }, { "grad_norm": 0.2934822142124176, "learning_rate": 3.759495084790887e-06, "loss": 0.0167, "step": 70560 }, { "grad_norm": 0.3339497447013855, "learning_rate": 3.751636185691282e-06, "loss": 0.0169, "step": 70570 }, { "grad_norm": 0.7933708429336548, "learning_rate": 3.7437851891721607e-06, "loss": 0.0197, "step": 70580 }, { "grad_norm": 0.20593348145484924, "learning_rate": 3.7359420965750404e-06, "loss": 0.0182, "step": 70590 }, { "grad_norm": 0.23549067974090576, "learning_rate": 3.7281069092400922e-06, "loss": 0.0179, "step": 70600 }, { "grad_norm": 0.3421452045440674, "learning_rate": 3.7202796285061348e-06, "loss": 0.012, "step": 70610 }, { "grad_norm": 0.2045801728963852, "learning_rate": 3.712460255710637e-06, "loss": 0.0212, "step": 70620 }, { "grad_norm": 0.29514017701148987, "learning_rate": 3.704648792189719e-06, "loss": 0.017, "step": 70630 }, { "grad_norm": 0.27606630325317383, "learning_rate": 3.696845239278124e-06, "loss": 0.0228, "step": 70640 }, { "grad_norm": 0.2087450921535492, "learning_rate": 3.689049598309302e-06, "loss": 0.0136, "step": 70650 }, { "grad_norm": 0.30638840794563293, "learning_rate": 3.681261870615288e-06, "loss": 0.0226, "step": 70660 }, { "grad_norm": 0.18920226395130157, "learning_rate": 3.6734820575268004e-06, "loss": 0.0318, "step": 70670 }, { "grad_norm": 0.2312917411327362, "learning_rate": 3.665710160373209e-06, "loss": 0.0193, "step": 70680 }, { "grad_norm": 0.41705095767974854, "learning_rate": 3.65794618048248e-06, "loss": 0.0179, "step": 70690 }, { "grad_norm": 0.26439759135246277, "learning_rate": 3.6501901191813125e-06, "loss": 0.0151, "step": 70700 }, { "grad_norm": 0.35927125811576843, "learning_rate": 3.642441977794975e-06, "loss": 0.0223, "step": 70710 }, { "grad_norm": 0.29882606863975525, "learning_rate": 3.634701757647424e-06, "loss": 0.0192, "step": 70720 }, { "grad_norm": 0.14515002071857452, "learning_rate": 3.6269694600612468e-06, "loss": 0.0124, "step": 70730 }, { "grad_norm": 0.4191781282424927, "learning_rate": 3.619245086357681e-06, "loss": 0.0206, "step": 70740 }, { "grad_norm": 0.41918379068374634, "learning_rate": 3.611528637856615e-06, "loss": 0.0196, "step": 70750 }, { "grad_norm": 0.514447033405304, "learning_rate": 3.6038201158765884e-06, "loss": 0.0127, "step": 70760 }, { "grad_norm": 0.6144407391548157, "learning_rate": 3.5961195217347534e-06, "loss": 0.0255, "step": 70770 }, { "grad_norm": 0.25670358538627625, "learning_rate": 3.588426856746946e-06, "loss": 0.0191, "step": 70780 }, { "grad_norm": 0.22239680588245392, "learning_rate": 3.5807421222276316e-06, "loss": 0.0268, "step": 70790 }, { "grad_norm": 0.18968656659126282, "learning_rate": 3.5730653194899156e-06, "loss": 0.017, "step": 70800 }, { "grad_norm": 0.2630009055137634, "learning_rate": 3.565396449845554e-06, "loss": 0.0202, "step": 70810 }, { "grad_norm": 0.3897550106048584, "learning_rate": 3.557735514604954e-06, "loss": 0.0193, "step": 70820 }, { "grad_norm": 0.5250493288040161, "learning_rate": 3.5500825150771633e-06, "loss": 0.0163, "step": 70830 }, { "grad_norm": 0.2595770061016083, "learning_rate": 3.5424374525698466e-06, "loss": 0.0183, "step": 70840 }, { "grad_norm": 0.2742473781108856, "learning_rate": 3.5348003283893704e-06, "loss": 0.0143, "step": 70850 }, { "grad_norm": 0.419699490070343, "learning_rate": 3.527171143840685e-06, "loss": 0.0176, "step": 70860 }, { "grad_norm": 0.37042543292045593, "learning_rate": 3.519549900227409e-06, "loss": 0.0182, "step": 70870 }, { "grad_norm": 0.3520730435848236, "learning_rate": 3.511936598851828e-06, "loss": 0.0211, "step": 70880 }, { "grad_norm": 0.5110542178153992, "learning_rate": 3.504331241014813e-06, "loss": 0.0173, "step": 70890 }, { "grad_norm": 0.2352212816476822, "learning_rate": 3.4967338280159414e-06, "loss": 0.0188, "step": 70900 }, { "grad_norm": 0.26918357610702515, "learning_rate": 3.4891443611533846e-06, "loss": 0.0223, "step": 70910 }, { "grad_norm": 0.2153431624174118, "learning_rate": 3.481562841723984e-06, "loss": 0.0242, "step": 70920 }, { "grad_norm": 0.31763961911201477, "learning_rate": 3.473989271023215e-06, "loss": 0.0207, "step": 70930 }, { "grad_norm": 0.19483396410942078, "learning_rate": 3.4664236503451854e-06, "loss": 0.0184, "step": 70940 }, { "grad_norm": 0.18662706017494202, "learning_rate": 3.4588659809826674e-06, "loss": 0.0157, "step": 70950 }, { "grad_norm": 0.28421154618263245, "learning_rate": 3.451316264227039e-06, "loss": 0.0268, "step": 70960 }, { "grad_norm": 0.1733541190624237, "learning_rate": 3.4437745013683574e-06, "loss": 0.0212, "step": 70970 }, { "grad_norm": 0.3121406137943268, "learning_rate": 3.4362406936952916e-06, "loss": 0.0166, "step": 70980 }, { "grad_norm": 0.20393966138362885, "learning_rate": 3.4287148424951786e-06, "loss": 0.018, "step": 70990 }, { "grad_norm": 0.2514900863170624, "learning_rate": 3.4211969490539617e-06, "loss": 0.0165, "step": 71000 }, { "grad_norm": 0.29410380125045776, "learning_rate": 3.4136870146562584e-06, "loss": 0.0192, "step": 71010 }, { "grad_norm": 0.5242627859115601, "learning_rate": 3.406185040585308e-06, "loss": 0.019, "step": 71020 }, { "grad_norm": 0.4975563585758209, "learning_rate": 3.3986910281229966e-06, "loss": 0.0232, "step": 71030 }, { "grad_norm": 0.7844718098640442, "learning_rate": 3.391204978549828e-06, "loss": 0.0253, "step": 71040 }, { "grad_norm": 0.30244502425193787, "learning_rate": 3.3837268931449785e-06, "loss": 0.0246, "step": 71050 }, { "grad_norm": 0.28004884719848633, "learning_rate": 3.376256773186248e-06, "loss": 0.021, "step": 71060 }, { "grad_norm": 0.22821955382823944, "learning_rate": 3.3687946199500664e-06, "loss": 0.0245, "step": 71070 }, { "grad_norm": 0.3896181285381317, "learning_rate": 3.3613404347115295e-06, "loss": 0.0137, "step": 71080 }, { "grad_norm": 0.18737275898456573, "learning_rate": 3.35389421874433e-06, "loss": 0.0153, "step": 71090 }, { "grad_norm": 0.38717710971832275, "learning_rate": 3.3464559733208446e-06, "loss": 0.0182, "step": 71100 }, { "grad_norm": 0.3924485445022583, "learning_rate": 3.3390256997120505e-06, "loss": 0.0193, "step": 71110 }, { "grad_norm": 0.18048466742038727, "learning_rate": 3.331603399187583e-06, "loss": 0.0213, "step": 71120 }, { "grad_norm": 0.26304513216018677, "learning_rate": 3.324189073015721e-06, "loss": 0.0252, "step": 71130 }, { "grad_norm": 0.3002978563308716, "learning_rate": 3.3167827224633465e-06, "loss": 0.0188, "step": 71140 }, { "grad_norm": 0.12348097562789917, "learning_rate": 3.309384348796024e-06, "loss": 0.0184, "step": 71150 }, { "grad_norm": 0.7463522553443909, "learning_rate": 3.301993953277921e-06, "loss": 0.019, "step": 71160 }, { "grad_norm": 0.44836148619651794, "learning_rate": 3.29461153717186e-06, "loss": 0.0167, "step": 71170 }, { "grad_norm": 0.44566449522972107, "learning_rate": 3.287237101739293e-06, "loss": 0.0171, "step": 71180 }, { "grad_norm": 0.3021615743637085, "learning_rate": 3.2798706482403075e-06, "loss": 0.021, "step": 71190 }, { "grad_norm": 0.3116420805454254, "learning_rate": 3.2725121779336285e-06, "loss": 0.0199, "step": 71200 }, { "grad_norm": 0.22173750400543213, "learning_rate": 3.265161692076618e-06, "loss": 0.019, "step": 71210 }, { "grad_norm": 0.4229615032672882, "learning_rate": 3.257819191925282e-06, "loss": 0.019, "step": 71220 }, { "grad_norm": 0.20993836224079132, "learning_rate": 3.2504846787342392e-06, "loss": 0.0172, "step": 71230 }, { "grad_norm": 0.6281777024269104, "learning_rate": 3.243158153756759e-06, "loss": 0.0181, "step": 71240 }, { "grad_norm": 0.49344202876091003, "learning_rate": 3.2358396182447514e-06, "loss": 0.0149, "step": 71250 }, { "grad_norm": 0.26867935061454773, "learning_rate": 3.2285290734487496e-06, "loss": 0.0178, "step": 71260 }, { "grad_norm": 2.061933994293213, "learning_rate": 3.221226520617926e-06, "loss": 0.0178, "step": 71270 }, { "grad_norm": 0.2925349175930023, "learning_rate": 3.213931961000094e-06, "loss": 0.0193, "step": 71280 }, { "grad_norm": 0.27906927466392517, "learning_rate": 3.2066453958416733e-06, "loss": 0.0181, "step": 71290 }, { "grad_norm": 0.20971369743347168, "learning_rate": 3.199366826387773e-06, "loss": 0.0226, "step": 71300 }, { "grad_norm": 0.20182545483112335, "learning_rate": 3.192096253882071e-06, "loss": 0.0179, "step": 71310 }, { "grad_norm": 0.28543373942375183, "learning_rate": 3.1848336795669177e-06, "loss": 0.0155, "step": 71320 }, { "grad_norm": 0.46102598309516907, "learning_rate": 3.1775791046833035e-06, "loss": 0.0254, "step": 71330 }, { "grad_norm": 0.22436794638633728, "learning_rate": 3.170332530470804e-06, "loss": 0.0189, "step": 71340 }, { "grad_norm": 0.3639447093009949, "learning_rate": 3.1630939581677012e-06, "loss": 0.023, "step": 71350 }, { "grad_norm": 0.4373251497745514, "learning_rate": 3.155863389010838e-06, "loss": 0.0256, "step": 71360 }, { "grad_norm": 0.2294677346944809, "learning_rate": 3.1486408242357323e-06, "loss": 0.0225, "step": 71370 }, { "grad_norm": 0.5863028168678284, "learning_rate": 3.1414262650765248e-06, "loss": 0.0338, "step": 71380 }, { "grad_norm": 0.32435333728790283, "learning_rate": 3.134219712765979e-06, "loss": 0.018, "step": 71390 }, { "grad_norm": 0.40921375155448914, "learning_rate": 3.1270211685355044e-06, "loss": 0.0197, "step": 71400 }, { "grad_norm": 0.273389607667923, "learning_rate": 3.1198306336151338e-06, "loss": 0.0181, "step": 71410 }, { "grad_norm": 0.2303021103143692, "learning_rate": 3.11264810923354e-06, "loss": 0.022, "step": 71420 }, { "grad_norm": 0.41982659697532654, "learning_rate": 3.105473596618008e-06, "loss": 0.0188, "step": 71430 }, { "grad_norm": 0.35037508606910706, "learning_rate": 3.0983070969944683e-06, "loss": 0.0216, "step": 71440 }, { "grad_norm": 0.21595922112464905, "learning_rate": 3.09114861158748e-06, "loss": 0.0168, "step": 71450 }, { "grad_norm": 1.1322582960128784, "learning_rate": 3.0839981416202314e-06, "loss": 0.0213, "step": 71460 }, { "grad_norm": 0.21618804335594177, "learning_rate": 3.0768556883145507e-06, "loss": 0.0105, "step": 71470 }, { "grad_norm": 0.6623212099075317, "learning_rate": 3.0697212528908834e-06, "loss": 0.016, "step": 71480 }, { "grad_norm": 0.20452085137367249, "learning_rate": 3.0625948365683e-06, "loss": 0.0173, "step": 71490 }, { "grad_norm": 0.816993772983551, "learning_rate": 3.055476440564531e-06, "loss": 0.0186, "step": 71500 }, { "grad_norm": 0.27286285161972046, "learning_rate": 3.0483660660958924e-06, "loss": 0.0151, "step": 71510 }, { "grad_norm": 0.3556099236011505, "learning_rate": 3.0412637143773624e-06, "loss": 0.0233, "step": 71520 }, { "grad_norm": 0.21321448683738708, "learning_rate": 3.034169386622554e-06, "loss": 0.0162, "step": 71530 }, { "grad_norm": 0.27096331119537354, "learning_rate": 3.027083084043658e-06, "loss": 0.0168, "step": 71540 }, { "grad_norm": 0.3345318138599396, "learning_rate": 3.0200048078515676e-06, "loss": 0.018, "step": 71550 }, { "grad_norm": 0.433918833732605, "learning_rate": 3.012934559255742e-06, "loss": 0.0175, "step": 71560 }, { "grad_norm": 0.2547891438007355, "learning_rate": 3.0058723394642994e-06, "loss": 0.0202, "step": 71570 }, { "grad_norm": 0.4466042220592499, "learning_rate": 2.998818149683985e-06, "loss": 0.0232, "step": 71580 }, { "grad_norm": 1.0822935104370117, "learning_rate": 2.9917719911201627e-06, "loss": 0.0202, "step": 71590 }, { "grad_norm": 0.31337442994117737, "learning_rate": 2.9847338649768352e-06, "loss": 0.0193, "step": 71600 }, { "grad_norm": 0.17844483256340027, "learning_rate": 2.977703772456608e-06, "loss": 0.019, "step": 71610 }, { "grad_norm": 0.3144424557685852, "learning_rate": 2.9706817147607535e-06, "loss": 0.0185, "step": 71620 }, { "grad_norm": 0.32003509998321533, "learning_rate": 2.9636676930891393e-06, "loss": 0.0193, "step": 71630 }, { "grad_norm": 0.19749431312084198, "learning_rate": 2.9566617086402625e-06, "loss": 0.0166, "step": 71640 }, { "grad_norm": 0.2139124572277069, "learning_rate": 2.9496637626112655e-06, "loss": 0.0195, "step": 71650 }, { "grad_norm": 0.32205480337142944, "learning_rate": 2.9426738561979027e-06, "loss": 0.0182, "step": 71660 }, { "grad_norm": 0.42092201113700867, "learning_rate": 2.9356919905945524e-06, "loss": 0.0378, "step": 71670 }, { "grad_norm": 0.22783301770687103, "learning_rate": 2.928718166994243e-06, "loss": 0.0146, "step": 71680 }, { "grad_norm": 0.5711779594421387, "learning_rate": 2.9217523865885833e-06, "loss": 0.0193, "step": 71690 }, { "grad_norm": 0.26339900493621826, "learning_rate": 2.9147946505678537e-06, "loss": 0.0178, "step": 71700 }, { "grad_norm": 0.667037308216095, "learning_rate": 2.9078449601209313e-06, "loss": 0.0153, "step": 71710 }, { "grad_norm": 0.4244081377983093, "learning_rate": 2.900903316435333e-06, "loss": 0.0212, "step": 71720 }, { "grad_norm": 0.20667116343975067, "learning_rate": 2.8939697206971983e-06, "loss": 0.0186, "step": 71730 }, { "grad_norm": 0.23430457711219788, "learning_rate": 2.8870441740912746e-06, "loss": 0.0145, "step": 71740 }, { "grad_norm": 0.2678099572658539, "learning_rate": 2.8801266778009762e-06, "loss": 0.0184, "step": 71750 }, { "grad_norm": 0.7841646075248718, "learning_rate": 2.8732172330082798e-06, "loss": 0.0182, "step": 71760 }, { "grad_norm": 0.37955889105796814, "learning_rate": 2.8663158408938517e-06, "loss": 0.0246, "step": 71770 }, { "grad_norm": 0.227370485663414, "learning_rate": 2.859422502636938e-06, "loss": 0.0176, "step": 71780 }, { "grad_norm": 0.20933254063129425, "learning_rate": 2.8525372194154076e-06, "loss": 0.0184, "step": 71790 }, { "grad_norm": 0.3407324552536011, "learning_rate": 2.8456599924057913e-06, "loss": 0.0148, "step": 71800 }, { "grad_norm": 0.21363860368728638, "learning_rate": 2.8387908227831995e-06, "loss": 0.0216, "step": 71810 }, { "grad_norm": 0.23000243306159973, "learning_rate": 2.831929711721404e-06, "loss": 0.0191, "step": 71820 }, { "grad_norm": 0.15550316870212555, "learning_rate": 2.825076660392767e-06, "loss": 0.0133, "step": 71830 }, { "grad_norm": 0.26565563678741455, "learning_rate": 2.8182316699682908e-06, "loss": 0.025, "step": 71840 }, { "grad_norm": 0.16694113612174988, "learning_rate": 2.811394741617601e-06, "loss": 0.0214, "step": 71850 }, { "grad_norm": 0.4683057367801666, "learning_rate": 2.8045658765089356e-06, "loss": 0.0206, "step": 71860 }, { "grad_norm": 0.21474537253379822, "learning_rate": 2.7977450758091605e-06, "loss": 0.0296, "step": 71870 }, { "grad_norm": 0.45472005009651184, "learning_rate": 2.790932340683783e-06, "loss": 0.0195, "step": 71880 }, { "grad_norm": 0.2672796845436096, "learning_rate": 2.7841276722968823e-06, "loss": 0.017, "step": 71890 }, { "grad_norm": 0.23502284288406372, "learning_rate": 2.7773310718112067e-06, "loss": 0.0161, "step": 71900 }, { "grad_norm": 0.25174859166145325, "learning_rate": 2.7705425403881102e-06, "loss": 0.0206, "step": 71910 }, { "grad_norm": 0.4734589159488678, "learning_rate": 2.763762079187565e-06, "loss": 0.0148, "step": 71920 }, { "grad_norm": 0.37638580799102783, "learning_rate": 2.756989689368178e-06, "loss": 0.0238, "step": 71930 }, { "grad_norm": 0.27763158082962036, "learning_rate": 2.750225372087134e-06, "loss": 0.0154, "step": 71940 }, { "grad_norm": 0.48660048842430115, "learning_rate": 2.7434691285003033e-06, "loss": 0.0155, "step": 71950 }, { "grad_norm": 0.6256928443908691, "learning_rate": 2.7367209597621178e-06, "loss": 0.0251, "step": 71960 }, { "grad_norm": 0.3325519263744354, "learning_rate": 2.729980867025683e-06, "loss": 0.0202, "step": 71970 }, { "grad_norm": 0.303799033164978, "learning_rate": 2.7232488514426724e-06, "loss": 0.0155, "step": 71980 }, { "grad_norm": 0.5018167495727539, "learning_rate": 2.7165249141634097e-06, "loss": 0.019, "step": 71990 }, { "grad_norm": 0.24302244186401367, "learning_rate": 2.709809056336837e-06, "loss": 0.0174, "step": 72000 }, { "grad_norm": 0.5586658120155334, "learning_rate": 2.703101279110498e-06, "loss": 0.0175, "step": 72010 }, { "grad_norm": 0.27323678135871887, "learning_rate": 2.696401583630587e-06, "loss": 0.0197, "step": 72020 }, { "grad_norm": 0.42724287509918213, "learning_rate": 2.689709971041887e-06, "loss": 0.0139, "step": 72030 }, { "grad_norm": 0.28011834621429443, "learning_rate": 2.6830264424878126e-06, "loss": 0.0215, "step": 72040 }, { "grad_norm": 0.288075715303421, "learning_rate": 2.676350999110394e-06, "loss": 0.0271, "step": 72050 }, { "grad_norm": 0.8373278975486755, "learning_rate": 2.6696836420502856e-06, "loss": 0.0181, "step": 72060 }, { "grad_norm": 0.18154139816761017, "learning_rate": 2.663024372446765e-06, "loss": 0.0172, "step": 72070 }, { "grad_norm": 0.2987005412578583, "learning_rate": 2.6563731914377055e-06, "loss": 0.0167, "step": 72080 }, { "grad_norm": 0.19097763299942017, "learning_rate": 2.6497301001596087e-06, "loss": 0.0223, "step": 72090 }, { "grad_norm": 0.461238294839859, "learning_rate": 2.643095099747611e-06, "loss": 0.0234, "step": 72100 }, { "grad_norm": 0.19174884259700775, "learning_rate": 2.636468191335445e-06, "loss": 0.0195, "step": 72110 }, { "grad_norm": 0.22652126848697662, "learning_rate": 2.629849376055471e-06, "loss": 0.0177, "step": 72120 }, { "grad_norm": 0.21620915830135345, "learning_rate": 2.6232386550386678e-06, "loss": 0.0187, "step": 72130 }, { "grad_norm": 0.4113794267177582, "learning_rate": 2.6166360294146097e-06, "loss": 0.0194, "step": 72140 }, { "grad_norm": 0.46220579743385315, "learning_rate": 2.6100415003115275e-06, "loss": 0.0205, "step": 72150 }, { "grad_norm": 0.2159947007894516, "learning_rate": 2.603455068856225e-06, "loss": 0.0191, "step": 72160 }, { "grad_norm": 0.16677305102348328, "learning_rate": 2.5968767361741584e-06, "loss": 0.018, "step": 72170 }, { "grad_norm": 0.29652342200279236, "learning_rate": 2.590306503389378e-06, "loss": 0.021, "step": 72180 }, { "grad_norm": 0.3489532470703125, "learning_rate": 2.583744371624558e-06, "loss": 0.0282, "step": 72190 }, { "grad_norm": 0.36030590534210205, "learning_rate": 2.57719034200099e-06, "loss": 0.0139, "step": 72200 }, { "grad_norm": 0.25222888588905334, "learning_rate": 2.570644415638568e-06, "loss": 0.0189, "step": 72210 }, { "grad_norm": 0.28254666924476624, "learning_rate": 2.5641065936558295e-06, "loss": 0.016, "step": 72220 }, { "grad_norm": 0.5416622757911682, "learning_rate": 2.557576877169898e-06, "loss": 0.0184, "step": 72230 }, { "grad_norm": 0.9625721573829651, "learning_rate": 2.5510552672965205e-06, "loss": 0.0183, "step": 72240 }, { "grad_norm": 0.2238149493932724, "learning_rate": 2.544541765150077e-06, "loss": 0.0161, "step": 72250 }, { "grad_norm": 0.4261472523212433, "learning_rate": 2.5380363718435165e-06, "loss": 0.0178, "step": 72260 }, { "grad_norm": 0.24522048234939575, "learning_rate": 2.5315390884884714e-06, "loss": 0.0155, "step": 72270 }, { "grad_norm": 0.2434535175561905, "learning_rate": 2.5250499161951214e-06, "loss": 0.0234, "step": 72280 }, { "grad_norm": 0.29668450355529785, "learning_rate": 2.518568856072301e-06, "loss": 0.0162, "step": 72290 }, { "grad_norm": 0.20532462000846863, "learning_rate": 2.512095909227441e-06, "loss": 0.0159, "step": 72300 }, { "grad_norm": 0.4672865569591522, "learning_rate": 2.50563107676659e-06, "loss": 0.0237, "step": 72310 }, { "grad_norm": 0.3643985390663147, "learning_rate": 2.4991743597944152e-06, "loss": 0.018, "step": 72320 }, { "grad_norm": 0.21491995453834534, "learning_rate": 2.492725759414205e-06, "loss": 0.0209, "step": 72330 }, { "grad_norm": 0.45015281438827515, "learning_rate": 2.4862852767278234e-06, "loss": 0.016, "step": 72340 }, { "grad_norm": 0.25046253204345703, "learning_rate": 2.479852912835784e-06, "loss": 0.0125, "step": 72350 }, { "grad_norm": 0.21983963251113892, "learning_rate": 2.4734286688372075e-06, "loss": 0.0198, "step": 72360 }, { "grad_norm": 0.346123069524765, "learning_rate": 2.467012545829811e-06, "loss": 0.0235, "step": 72370 }, { "grad_norm": 1.056784987449646, "learning_rate": 2.460604544909945e-06, "loss": 0.0188, "step": 72380 }, { "grad_norm": 0.8603350520133972, "learning_rate": 2.454204667172555e-06, "loss": 0.0131, "step": 72390 }, { "grad_norm": 0.20115748047828674, "learning_rate": 2.447812913711217e-06, "loss": 0.0151, "step": 72400 }, { "grad_norm": 0.44437289237976074, "learning_rate": 2.44142928561808e-06, "loss": 0.0207, "step": 72410 }, { "grad_norm": 0.24412913620471954, "learning_rate": 2.435053783983965e-06, "loss": 0.0219, "step": 72420 }, { "grad_norm": 0.5716400742530823, "learning_rate": 2.4286864098982453e-06, "loss": 0.0187, "step": 72430 }, { "grad_norm": 0.4133310616016388, "learning_rate": 2.4223271644489397e-06, "loss": 0.0176, "step": 72440 }, { "grad_norm": 0.2939050793647766, "learning_rate": 2.415976048722679e-06, "loss": 0.0169, "step": 72450 }, { "grad_norm": 0.16468596458435059, "learning_rate": 2.4096330638046673e-06, "loss": 0.0154, "step": 72460 }, { "grad_norm": 0.17770174145698547, "learning_rate": 2.4032982107787816e-06, "loss": 0.0161, "step": 72470 }, { "grad_norm": 0.41646847128868103, "learning_rate": 2.396971490727451e-06, "loss": 0.0222, "step": 72480 }, { "grad_norm": 0.28459328413009644, "learning_rate": 2.3906529047317493e-06, "loss": 0.0164, "step": 72490 }, { "grad_norm": 0.269862562417984, "learning_rate": 2.3843424538713465e-06, "loss": 0.0189, "step": 72500 }, { "grad_norm": 0.3281400501728058, "learning_rate": 2.3780401392245298e-06, "loss": 0.0178, "step": 72510 }, { "grad_norm": 0.3105693459510803, "learning_rate": 2.3717459618681883e-06, "loss": 0.0189, "step": 72520 }, { "grad_norm": 0.3829765319824219, "learning_rate": 2.3654599228778274e-06, "loss": 0.0176, "step": 72530 }, { "grad_norm": 0.3648001551628113, "learning_rate": 2.3591820233275607e-06, "loss": 0.0214, "step": 72540 }, { "grad_norm": 0.23093847930431366, "learning_rate": 2.3529122642901024e-06, "loss": 0.0156, "step": 72550 }, { "grad_norm": 0.31573736667633057, "learning_rate": 2.3466506468367845e-06, "loss": 0.0267, "step": 72560 }, { "grad_norm": 0.3413887023925781, "learning_rate": 2.3403971720375506e-06, "loss": 0.0224, "step": 72570 }, { "grad_norm": 0.3456851840019226, "learning_rate": 2.334151840960952e-06, "loss": 0.0154, "step": 72580 }, { "grad_norm": 5.560844421386719, "learning_rate": 2.3279146546741347e-06, "loss": 0.0212, "step": 72590 }, { "grad_norm": 0.1432197540998459, "learning_rate": 2.32168561424288e-06, "loss": 0.0144, "step": 72600 }, { "grad_norm": 0.21985195577144623, "learning_rate": 2.3154647207315307e-06, "loss": 0.0176, "step": 72610 }, { "grad_norm": 0.32864341139793396, "learning_rate": 2.309251975203103e-06, "loss": 0.0194, "step": 72620 }, { "grad_norm": 0.22339192032814026, "learning_rate": 2.303047378719159e-06, "loss": 0.0186, "step": 72630 }, { "grad_norm": 0.5200536251068115, "learning_rate": 2.2968509323399e-06, "loss": 0.0193, "step": 72640 }, { "grad_norm": 0.8170130848884583, "learning_rate": 2.290662637124147e-06, "loss": 0.0175, "step": 72650 }, { "grad_norm": 0.27221059799194336, "learning_rate": 2.2844824941292807e-06, "loss": 0.0144, "step": 72660 }, { "grad_norm": 0.41293665766716003, "learning_rate": 2.2783105044113406e-06, "loss": 0.0267, "step": 72670 }, { "grad_norm": 0.6152616143226624, "learning_rate": 2.272146669024944e-06, "loss": 0.0155, "step": 72680 }, { "grad_norm": 0.2643820345401764, "learning_rate": 2.26599098902332e-06, "loss": 0.0158, "step": 72690 }, { "grad_norm": 0.2285330891609192, "learning_rate": 2.2598434654583113e-06, "loss": 0.0154, "step": 72700 }, { "grad_norm": 0.2636253535747528, "learning_rate": 2.253704099380355e-06, "loss": 0.0256, "step": 72710 }, { "grad_norm": 0.3654307425022125, "learning_rate": 2.247572891838512e-06, "loss": 0.02, "step": 72720 }, { "grad_norm": 0.390485554933548, "learning_rate": 2.241449843880422e-06, "loss": 0.0203, "step": 72730 }, { "grad_norm": 0.24245411157608032, "learning_rate": 2.235334956552354e-06, "loss": 0.0205, "step": 72740 }, { "grad_norm": 0.23118378221988678, "learning_rate": 2.2292282308991775e-06, "loss": 0.0175, "step": 72750 }, { "grad_norm": 0.2719807028770447, "learning_rate": 2.223129667964363e-06, "loss": 0.0148, "step": 72760 }, { "grad_norm": 0.4748784899711609, "learning_rate": 2.2170392687899834e-06, "loss": 0.0146, "step": 72770 }, { "grad_norm": 0.275485634803772, "learning_rate": 2.210957034416733e-06, "loss": 0.013, "step": 72780 }, { "grad_norm": 0.1530395746231079, "learning_rate": 2.2048829658838867e-06, "loss": 0.0179, "step": 72790 }, { "grad_norm": 0.12799473106861115, "learning_rate": 2.1988170642293525e-06, "loss": 0.0131, "step": 72800 }, { "grad_norm": 0.3287983536720276, "learning_rate": 2.1927593304896075e-06, "loss": 0.0192, "step": 72810 }, { "grad_norm": 0.23086795210838318, "learning_rate": 2.1867097656997626e-06, "loss": 0.0187, "step": 72820 }, { "grad_norm": 0.7238967418670654, "learning_rate": 2.180668370893524e-06, "loss": 0.0179, "step": 72830 }, { "grad_norm": 0.366567462682724, "learning_rate": 2.174635147103199e-06, "loss": 0.0208, "step": 72840 }, { "grad_norm": 0.28262075781822205, "learning_rate": 2.1686100953597075e-06, "loss": 0.019, "step": 72850 }, { "grad_norm": 0.43691501021385193, "learning_rate": 2.1625932166925433e-06, "loss": 0.0157, "step": 72860 }, { "grad_norm": 0.28573811054229736, "learning_rate": 2.1565845121298556e-06, "loss": 0.0164, "step": 72870 }, { "grad_norm": 0.2284722477197647, "learning_rate": 2.150583982698351e-06, "loss": 0.0206, "step": 72880 }, { "grad_norm": 0.44649770855903625, "learning_rate": 2.144591629423359e-06, "loss": 0.0179, "step": 72890 }, { "grad_norm": 0.6910433769226074, "learning_rate": 2.138607453328817e-06, "loss": 0.0125, "step": 72900 }, { "grad_norm": 0.22311623394489288, "learning_rate": 2.132631455437234e-06, "loss": 0.0197, "step": 72910 }, { "grad_norm": 0.8585556149482727, "learning_rate": 2.1266636367697714e-06, "loss": 0.0281, "step": 72920 }, { "grad_norm": 1.4315330982208252, "learning_rate": 2.120703998346152e-06, "loss": 0.0163, "step": 72930 }, { "grad_norm": 0.29486793279647827, "learning_rate": 2.1147525411847114e-06, "loss": 0.0162, "step": 72940 }, { "grad_norm": 1.1485298871994019, "learning_rate": 2.108809266302403e-06, "loss": 0.0155, "step": 72950 }, { "grad_norm": 0.18837076425552368, "learning_rate": 2.102874174714764e-06, "loss": 0.0192, "step": 72960 }, { "grad_norm": 0.3917728662490845, "learning_rate": 2.0969472674359393e-06, "loss": 0.0158, "step": 72970 }, { "grad_norm": 0.2345460206270218, "learning_rate": 2.091028545478668e-06, "loss": 0.0226, "step": 72980 }, { "grad_norm": 0.31155112385749817, "learning_rate": 2.08511800985432e-06, "loss": 0.018, "step": 72990 }, { "grad_norm": 0.19545604288578033, "learning_rate": 2.07921566157282e-06, "loss": 0.0247, "step": 73000 }, { "grad_norm": 0.6743872761726379, "learning_rate": 2.073321501642728e-06, "loss": 0.0192, "step": 73010 }, { "grad_norm": 0.21830090880393982, "learning_rate": 2.0674355310711937e-06, "loss": 0.0235, "step": 73020 }, { "grad_norm": 0.4192482531070709, "learning_rate": 2.0615577508639682e-06, "loss": 0.0133, "step": 73030 }, { "grad_norm": 0.2791588008403778, "learning_rate": 2.055688162025399e-06, "loss": 0.0177, "step": 73040 }, { "grad_norm": 0.1384575366973877, "learning_rate": 2.0498267655584546e-06, "loss": 0.0216, "step": 73050 }, { "grad_norm": 0.2901966869831085, "learning_rate": 2.0439735624646626e-06, "loss": 0.0233, "step": 73060 }, { "grad_norm": 0.2583131492137909, "learning_rate": 2.0381285537442008e-06, "loss": 0.0148, "step": 73070 }, { "grad_norm": 0.26653072237968445, "learning_rate": 2.032291740395803e-06, "loss": 0.0183, "step": 73080 }, { "grad_norm": 0.25964078307151794, "learning_rate": 2.0264631234168276e-06, "loss": 0.022, "step": 73090 }, { "grad_norm": 0.40763065218925476, "learning_rate": 2.0206427038032326e-06, "loss": 0.0172, "step": 73100 }, { "grad_norm": 0.18464210629463196, "learning_rate": 2.0148304825495457e-06, "loss": 0.0148, "step": 73110 }, { "grad_norm": 1.051370620727539, "learning_rate": 2.0090264606489496e-06, "loss": 0.0163, "step": 73120 }, { "grad_norm": 0.2860417366027832, "learning_rate": 2.003230639093162e-06, "loss": 0.0185, "step": 73130 }, { "grad_norm": 0.3131665289402008, "learning_rate": 1.9974430188725525e-06, "loss": 0.0184, "step": 73140 }, { "grad_norm": 0.2722138464450836, "learning_rate": 1.9916636009760513e-06, "loss": 0.0191, "step": 73150 }, { "grad_norm": 0.20052604377269745, "learning_rate": 1.985892386391208e-06, "loss": 0.0247, "step": 73160 }, { "grad_norm": 0.9175277352333069, "learning_rate": 1.980129376104173e-06, "loss": 0.0181, "step": 73170 }, { "grad_norm": 0.22961486876010895, "learning_rate": 1.9743745710996796e-06, "loss": 0.015, "step": 73180 }, { "grad_norm": 0.14497791230678558, "learning_rate": 1.9686279723610757e-06, "loss": 0.0191, "step": 73190 }, { "grad_norm": 0.43408650159835815, "learning_rate": 1.962889580870281e-06, "loss": 0.0254, "step": 73200 }, { "grad_norm": 0.3129642605781555, "learning_rate": 1.9571593976078385e-06, "loss": 0.0191, "step": 73210 }, { "grad_norm": 0.24804289638996124, "learning_rate": 1.9514374235528876e-06, "loss": 0.0188, "step": 73220 }, { "grad_norm": 0.4169212877750397, "learning_rate": 1.9457236596831408e-06, "loss": 0.018, "step": 73230 }, { "grad_norm": 0.15935741364955902, "learning_rate": 1.940018106974939e-06, "loss": 0.0221, "step": 73240 }, { "grad_norm": 0.17891831696033478, "learning_rate": 1.9343207664032025e-06, "loss": 0.0165, "step": 73250 }, { "grad_norm": 0.7125946879386902, "learning_rate": 1.928631638941436e-06, "loss": 0.0156, "step": 73260 }, { "grad_norm": 0.22271320223808289, "learning_rate": 1.922950725561784e-06, "loss": 0.0172, "step": 73270 }, { "grad_norm": 0.1610146015882492, "learning_rate": 1.9172780272349312e-06, "loss": 0.0138, "step": 73280 }, { "grad_norm": 0.23666948080062866, "learning_rate": 1.911613544930202e-06, "loss": 0.0185, "step": 73290 }, { "grad_norm": 0.2177770882844925, "learning_rate": 1.9059572796155e-06, "loss": 0.0217, "step": 73300 }, { "grad_norm": 0.28030887246131897, "learning_rate": 1.9003092322573192e-06, "loss": 0.0154, "step": 73310 }, { "grad_norm": 0.15907637774944305, "learning_rate": 1.8946694038207647e-06, "loss": 0.0204, "step": 73320 }, { "grad_norm": 0.20337556302547455, "learning_rate": 1.8890377952695215e-06, "loss": 0.0146, "step": 73330 }, { "grad_norm": 0.28299441933631897, "learning_rate": 1.8834144075658865e-06, "loss": 0.0183, "step": 73340 }, { "grad_norm": 0.16353999078273773, "learning_rate": 1.8777992416707357e-06, "loss": 0.0228, "step": 73350 }, { "grad_norm": 0.23807065188884735, "learning_rate": 1.8721922985435458e-06, "loss": 0.0151, "step": 73360 }, { "grad_norm": 0.5812783241271973, "learning_rate": 1.8665935791424006e-06, "loss": 0.02, "step": 73370 }, { "grad_norm": 0.23805643618106842, "learning_rate": 1.8610030844239512e-06, "loss": 0.0148, "step": 73380 }, { "grad_norm": 0.22185565531253815, "learning_rate": 1.8554208153434838e-06, "loss": 0.0187, "step": 73390 }, { "grad_norm": 0.31459829211235046, "learning_rate": 1.8498467728548296e-06, "loss": 0.0133, "step": 73400 }, { "grad_norm": 0.34762364625930786, "learning_rate": 1.8442809579104547e-06, "loss": 0.0211, "step": 73410 }, { "grad_norm": 0.44731131196022034, "learning_rate": 1.8387233714614039e-06, "loss": 0.0177, "step": 73420 }, { "grad_norm": 0.563502848148346, "learning_rate": 1.8331740144573174e-06, "loss": 0.0287, "step": 73430 }, { "grad_norm": 0.1836545169353485, "learning_rate": 1.8276328878464255e-06, "loss": 0.0177, "step": 73440 }, { "grad_norm": 0.2731770873069763, "learning_rate": 1.82209999257556e-06, "loss": 0.0252, "step": 73450 }, { "grad_norm": 0.2392425835132599, "learning_rate": 1.8165753295901311e-06, "loss": 0.0149, "step": 73460 }, { "grad_norm": 0.29258131980895996, "learning_rate": 1.8110588998341616e-06, "loss": 0.02, "step": 73470 }, { "grad_norm": 0.25829678773880005, "learning_rate": 1.805550704250253e-06, "loss": 0.0229, "step": 73480 }, { "grad_norm": 0.4400016665458679, "learning_rate": 1.8000507437796077e-06, "loss": 0.016, "step": 73490 }, { "grad_norm": 0.2527415454387665, "learning_rate": 1.7945590193620242e-06, "loss": 0.0162, "step": 73500 }, { "grad_norm": 0.44520413875579834, "learning_rate": 1.7890755319358742e-06, "loss": 0.0194, "step": 73510 }, { "grad_norm": 0.26505136489868164, "learning_rate": 1.7836002824381525e-06, "loss": 0.0165, "step": 73520 }, { "grad_norm": 0.35845252871513367, "learning_rate": 1.7781332718044165e-06, "loss": 0.0155, "step": 73530 }, { "grad_norm": 0.35866081714630127, "learning_rate": 1.772674500968835e-06, "loss": 0.0181, "step": 73540 }, { "grad_norm": 0.3422084450721741, "learning_rate": 1.7672239708641624e-06, "loss": 0.02, "step": 73550 }, { "grad_norm": 0.263507604598999, "learning_rate": 1.7617816824217371e-06, "loss": 0.0153, "step": 73560 }, { "grad_norm": 0.4501745402812958, "learning_rate": 1.7563476365715148e-06, "loss": 0.0285, "step": 73570 }, { "grad_norm": 1.0665836334228516, "learning_rate": 1.750921834242003e-06, "loss": 0.0246, "step": 73580 }, { "grad_norm": 0.09515129029750824, "learning_rate": 1.7455042763603436e-06, "loss": 0.0182, "step": 73590 }, { "grad_norm": 0.18648932874202728, "learning_rate": 1.7400949638522345e-06, "loss": 0.0121, "step": 73600 }, { "grad_norm": 0.4421311616897583, "learning_rate": 1.734693897641987e-06, "loss": 0.0305, "step": 73610 }, { "grad_norm": 0.36367562413215637, "learning_rate": 1.7293010786524955e-06, "loss": 0.0151, "step": 73620 }, { "grad_norm": 0.39475321769714355, "learning_rate": 1.72391650780524e-06, "loss": 0.0153, "step": 73630 }, { "grad_norm": 0.2948645055294037, "learning_rate": 1.718540186020301e-06, "loss": 0.0165, "step": 73640 }, { "grad_norm": 0.46607398986816406, "learning_rate": 1.7131721142163437e-06, "loss": 0.0153, "step": 73650 }, { "grad_norm": 0.4669557213783264, "learning_rate": 1.7078122933106233e-06, "loss": 0.0235, "step": 73660 }, { "grad_norm": 0.4870930314064026, "learning_rate": 1.7024607242189905e-06, "loss": 0.0179, "step": 73670 }, { "grad_norm": 0.26728272438049316, "learning_rate": 1.6971174078558749e-06, "loss": 0.0148, "step": 73680 }, { "grad_norm": 0.22518181800842285, "learning_rate": 1.6917823451343073e-06, "loss": 0.0192, "step": 73690 }, { "grad_norm": 0.2357710748910904, "learning_rate": 1.6864555369659142e-06, "loss": 0.0171, "step": 73700 }, { "grad_norm": 0.3811670243740082, "learning_rate": 1.681136984260878e-06, "loss": 0.0211, "step": 73710 }, { "grad_norm": 0.30365490913391113, "learning_rate": 1.6758266879280172e-06, "loss": 0.0205, "step": 73720 }, { "grad_norm": 0.3895171582698822, "learning_rate": 1.6705246488746996e-06, "loss": 0.0171, "step": 73730 }, { "grad_norm": 0.2279505729675293, "learning_rate": 1.6652308680069062e-06, "loss": 0.0182, "step": 73740 }, { "grad_norm": 0.45280420780181885, "learning_rate": 1.6599453462292081e-06, "loss": 0.0203, "step": 73750 }, { "grad_norm": 0.561007022857666, "learning_rate": 1.6546680844447326e-06, "loss": 0.0216, "step": 73760 }, { "grad_norm": 0.22204042971134186, "learning_rate": 1.6493990835552475e-06, "loss": 0.019, "step": 73770 }, { "grad_norm": 0.22184410691261292, "learning_rate": 1.6441383444610547e-06, "loss": 0.0153, "step": 73780 }, { "grad_norm": 0.2867499589920044, "learning_rate": 1.6388858680610908e-06, "loss": 0.0162, "step": 73790 }, { "grad_norm": 0.36284762620925903, "learning_rate": 1.6336416552528543e-06, "loss": 0.0185, "step": 73800 }, { "grad_norm": 0.32000747323036194, "learning_rate": 1.6284057069324288e-06, "loss": 0.0184, "step": 73810 }, { "grad_norm": 0.19084271788597107, "learning_rate": 1.623178023994504e-06, "loss": 0.0273, "step": 73820 }, { "grad_norm": 0.238777294754982, "learning_rate": 1.6179586073323483e-06, "loss": 0.0181, "step": 73830 }, { "grad_norm": 0.5000972747802734, "learning_rate": 1.612747457837821e-06, "loss": 0.0203, "step": 73840 }, { "grad_norm": 0.41315650939941406, "learning_rate": 1.607544576401354e-06, "loss": 0.0173, "step": 73850 }, { "grad_norm": 0.17217113077640533, "learning_rate": 1.6023499639119754e-06, "loss": 0.0179, "step": 73860 }, { "grad_norm": 0.24932605028152466, "learning_rate": 1.5971636212573138e-06, "loss": 0.016, "step": 73870 }, { "grad_norm": 0.19493967294692993, "learning_rate": 1.5919855493235714e-06, "loss": 0.016, "step": 73880 }, { "grad_norm": 0.27864691615104675, "learning_rate": 1.586815748995535e-06, "loss": 0.017, "step": 73890 }, { "grad_norm": 0.3732370138168335, "learning_rate": 1.5816542211565866e-06, "loss": 0.0137, "step": 73900 }, { "grad_norm": 0.5691859126091003, "learning_rate": 1.5765009666886765e-06, "loss": 0.0168, "step": 73910 }, { "grad_norm": 0.32024693489074707, "learning_rate": 1.5713559864723781e-06, "loss": 0.0179, "step": 73920 }, { "grad_norm": 0.26653820276260376, "learning_rate": 1.5662192813868104e-06, "loss": 0.0162, "step": 73930 }, { "grad_norm": 0.32331210374832153, "learning_rate": 1.561090852309699e-06, "loss": 0.0196, "step": 73940 }, { "grad_norm": 0.21116861701011658, "learning_rate": 1.5559707001173651e-06, "loss": 0.0202, "step": 73950 }, { "grad_norm": 0.29567480087280273, "learning_rate": 1.5508588256846757e-06, "loss": 0.0164, "step": 73960 }, { "grad_norm": 0.22709737718105316, "learning_rate": 1.5457552298851319e-06, "loss": 0.0172, "step": 73970 }, { "grad_norm": 0.2738942503929138, "learning_rate": 1.5406599135907918e-06, "loss": 0.0219, "step": 73980 }, { "grad_norm": 0.9023508429527283, "learning_rate": 1.5355728776723088e-06, "loss": 0.0199, "step": 73990 }, { "grad_norm": 0.20490308105945587, "learning_rate": 1.5304941229989155e-06, "loss": 0.017, "step": 74000 }, { "grad_norm": 0.2704129219055176, "learning_rate": 1.5254236504384345e-06, "loss": 0.0183, "step": 74010 }, { "grad_norm": 0.18469572067260742, "learning_rate": 1.5203614608572726e-06, "loss": 0.0221, "step": 74020 }, { "grad_norm": 0.6743265986442566, "learning_rate": 1.5153075551204044e-06, "loss": 0.0176, "step": 74030 }, { "grad_norm": 0.24028116464614868, "learning_rate": 1.5102619340914225e-06, "loss": 0.0294, "step": 74040 }, { "grad_norm": 0.17281776666641235, "learning_rate": 1.505224598632482e-06, "loss": 0.0141, "step": 74050 }, { "grad_norm": 0.6886785626411438, "learning_rate": 1.5001955496043164e-06, "loss": 0.0163, "step": 74060 }, { "grad_norm": 0.3131624162197113, "learning_rate": 1.4951747878662604e-06, "loss": 0.0197, "step": 74070 }, { "grad_norm": 0.3385818898677826, "learning_rate": 1.4901623142762221e-06, "loss": 0.0172, "step": 74080 }, { "grad_norm": 0.13122819364070892, "learning_rate": 1.4851581296907001e-06, "loss": 0.0206, "step": 74090 }, { "grad_norm": 0.2150295376777649, "learning_rate": 1.4801622349647714e-06, "loss": 0.0153, "step": 74100 }, { "grad_norm": 0.8024460673332214, "learning_rate": 1.4751746309520976e-06, "loss": 0.0211, "step": 74110 }, { "grad_norm": 0.19973506033420563, "learning_rate": 1.4701953185049132e-06, "loss": 0.0173, "step": 74120 }, { "grad_norm": 0.25391146540641785, "learning_rate": 1.4652242984740661e-06, "loss": 0.0244, "step": 74130 }, { "grad_norm": 0.3143176734447479, "learning_rate": 1.4602615717089484e-06, "loss": 0.025, "step": 74140 }, { "grad_norm": 0.25986623764038086, "learning_rate": 1.4553071390575657e-06, "loss": 0.0199, "step": 74150 }, { "grad_norm": 0.8167429566383362, "learning_rate": 1.450361001366496e-06, "loss": 0.0224, "step": 74160 }, { "grad_norm": 0.7681439518928528, "learning_rate": 1.4454231594809021e-06, "loss": 0.023, "step": 74170 }, { "grad_norm": 0.36127737164497375, "learning_rate": 1.4404936142445036e-06, "loss": 0.0211, "step": 74180 }, { "grad_norm": 0.3189730644226074, "learning_rate": 1.4355723664996546e-06, "loss": 0.0179, "step": 74190 }, { "grad_norm": 0.34481650590896606, "learning_rate": 1.4306594170872433e-06, "loss": 0.0164, "step": 74200 }, { "grad_norm": 0.3603495955467224, "learning_rate": 1.4257547668467598e-06, "loss": 0.0254, "step": 74210 }, { "grad_norm": 0.2572503685951233, "learning_rate": 1.4208584166162886e-06, "loss": 0.0166, "step": 74220 }, { "grad_norm": 0.23119693994522095, "learning_rate": 1.4159703672324553e-06, "loss": 0.0216, "step": 74230 }, { "grad_norm": 0.3053383231163025, "learning_rate": 1.4110906195305252e-06, "loss": 0.0222, "step": 74240 }, { "grad_norm": 0.1412181407213211, "learning_rate": 1.406219174344292e-06, "loss": 0.0223, "step": 74250 }, { "grad_norm": 0.5810796618461609, "learning_rate": 1.401356032506157e-06, "loss": 0.0196, "step": 74260 }, { "grad_norm": 0.2057872712612152, "learning_rate": 1.3965011948471051e-06, "loss": 0.0226, "step": 74270 }, { "grad_norm": 0.442669540643692, "learning_rate": 1.3916546621966842e-06, "loss": 0.0175, "step": 74280 }, { "grad_norm": 0.31073102355003357, "learning_rate": 1.3868164353830482e-06, "loss": 0.0139, "step": 74290 }, { "grad_norm": 0.3838580846786499, "learning_rate": 1.3819865152329082e-06, "loss": 0.0133, "step": 74300 }, { "grad_norm": 0.25331389904022217, "learning_rate": 1.377164902571565e-06, "loss": 0.0242, "step": 74310 }, { "grad_norm": 0.3717452883720398, "learning_rate": 1.372351598222904e-06, "loss": 0.0197, "step": 74320 }, { "grad_norm": 0.2720194458961487, "learning_rate": 1.3675466030093843e-06, "loss": 0.0175, "step": 74330 }, { "grad_norm": 0.1541908234357834, "learning_rate": 1.3627499177520486e-06, "loss": 0.0195, "step": 74340 }, { "grad_norm": 0.35206443071365356, "learning_rate": 1.3579615432705196e-06, "loss": 0.0204, "step": 74350 }, { "grad_norm": 0.28245049715042114, "learning_rate": 1.3531814803829978e-06, "loss": 0.023, "step": 74360 }, { "grad_norm": 0.2978273034095764, "learning_rate": 1.3484097299062803e-06, "loss": 0.0164, "step": 74370 }, { "grad_norm": 0.2088695615530014, "learning_rate": 1.343646292655698e-06, "loss": 0.0137, "step": 74380 }, { "grad_norm": 1.0172544717788696, "learning_rate": 1.3388911694452277e-06, "loss": 0.0191, "step": 74390 }, { "grad_norm": 0.37486353516578674, "learning_rate": 1.3341443610873638e-06, "loss": 0.0198, "step": 74400 }, { "grad_norm": 0.21326366066932678, "learning_rate": 1.329405868393213e-06, "loss": 0.0196, "step": 74410 }, { "grad_norm": 0.45067286491394043, "learning_rate": 1.3246756921724613e-06, "loss": 0.0209, "step": 74420 }, { "grad_norm": 0.2282441407442093, "learning_rate": 1.3199538332333506e-06, "loss": 0.0172, "step": 74430 }, { "grad_norm": 0.2990746796131134, "learning_rate": 1.3152402923827411e-06, "loss": 0.0196, "step": 74440 }, { "grad_norm": 0.36562272906303406, "learning_rate": 1.3105350704260277e-06, "loss": 0.0164, "step": 74450 }, { "grad_norm": 0.2421802282333374, "learning_rate": 1.305838168167206e-06, "loss": 0.0177, "step": 74460 }, { "grad_norm": 0.6823996305465698, "learning_rate": 1.301149586408862e-06, "loss": 0.0136, "step": 74470 }, { "grad_norm": 0.3907480239868164, "learning_rate": 1.2964693259521322e-06, "loss": 0.0172, "step": 74480 }, { "grad_norm": 0.26900771260261536, "learning_rate": 1.2917973875967548e-06, "loss": 0.0195, "step": 74490 }, { "grad_norm": 0.5313870906829834, "learning_rate": 1.2871337721410249e-06, "loss": 0.0201, "step": 74500 }, { "grad_norm": 0.3681374192237854, "learning_rate": 1.2824784803818379e-06, "loss": 0.0196, "step": 74510 }, { "grad_norm": 0.4454384446144104, "learning_rate": 1.2778315131146467e-06, "loss": 0.0145, "step": 74520 }, { "grad_norm": 0.37710505723953247, "learning_rate": 1.2731928711334994e-06, "loss": 0.0166, "step": 74530 }, { "grad_norm": 0.6195645928382874, "learning_rate": 1.2685625552310122e-06, "loss": 0.0176, "step": 74540 }, { "grad_norm": 0.6435472965240479, "learning_rate": 1.263940566198374e-06, "loss": 0.0143, "step": 74550 }, { "grad_norm": 0.22218897938728333, "learning_rate": 1.259326904825353e-06, "loss": 0.0168, "step": 74560 }, { "grad_norm": 0.3384207785129547, "learning_rate": 1.2547215719003137e-06, "loss": 0.0215, "step": 74570 }, { "grad_norm": 0.329578697681427, "learning_rate": 1.2501245682101703e-06, "loss": 0.0188, "step": 74580 }, { "grad_norm": 0.27256008982658386, "learning_rate": 1.2455358945404171e-06, "loss": 0.0164, "step": 74590 }, { "grad_norm": 0.46280229091644287, "learning_rate": 1.2409555516751493e-06, "loss": 0.0156, "step": 74600 }, { "grad_norm": 0.27717921137809753, "learning_rate": 1.2363835403970125e-06, "loss": 0.0171, "step": 74610 }, { "grad_norm": 0.3160826563835144, "learning_rate": 1.2318198614872489e-06, "loss": 0.0186, "step": 74620 }, { "grad_norm": 0.3147968649864197, "learning_rate": 1.2272645157256457e-06, "loss": 0.0202, "step": 74630 }, { "grad_norm": 0.23465682566165924, "learning_rate": 1.222717503890608e-06, "loss": 0.0214, "step": 74640 }, { "grad_norm": 0.5224910378456116, "learning_rate": 1.218178826759081e-06, "loss": 0.0194, "step": 74650 }, { "grad_norm": 0.31355035305023193, "learning_rate": 1.213648485106611e-06, "loss": 0.0219, "step": 74660 }, { "grad_norm": 0.2267833799123764, "learning_rate": 1.2091264797073066e-06, "loss": 0.0183, "step": 74670 }, { "grad_norm": 0.2705574035644531, "learning_rate": 1.2046128113338494e-06, "loss": 0.0358, "step": 74680 }, { "grad_norm": 0.9504396319389343, "learning_rate": 1.200107480757512e-06, "loss": 0.0202, "step": 74690 }, { "grad_norm": 0.16473418474197388, "learning_rate": 1.1956104887481168e-06, "loss": 0.0256, "step": 74700 }, { "grad_norm": 0.29752641916275024, "learning_rate": 1.191121836074094e-06, "loss": 0.0168, "step": 74710 }, { "grad_norm": 0.6933927536010742, "learning_rate": 1.1866415235024186e-06, "loss": 0.0151, "step": 74720 }, { "grad_norm": 0.32939785718917847, "learning_rate": 1.182169551798662e-06, "loss": 0.0222, "step": 74730 }, { "grad_norm": 0.2707032561302185, "learning_rate": 1.177705921726957e-06, "loss": 0.0184, "step": 74740 }, { "grad_norm": 0.2702066898345947, "learning_rate": 1.173250634050016e-06, "loss": 0.0205, "step": 74750 }, { "grad_norm": 0.42161428928375244, "learning_rate": 1.1688036895291354e-06, "loss": 0.021, "step": 74760 }, { "grad_norm": 0.542750358581543, "learning_rate": 1.1643650889241574e-06, "loss": 0.0183, "step": 74770 }, { "grad_norm": 0.566524863243103, "learning_rate": 1.1599348329935311e-06, "loss": 0.0222, "step": 74780 }, { "grad_norm": 0.3024331033229828, "learning_rate": 1.1555129224942673e-06, "loss": 0.0264, "step": 74790 }, { "grad_norm": 0.4865538775920868, "learning_rate": 1.1510993581819396e-06, "loss": 0.0148, "step": 74800 }, { "grad_norm": 0.23869262635707855, "learning_rate": 1.1466941408107112e-06, "loss": 0.0261, "step": 74810 }, { "grad_norm": 0.30110111832618713, "learning_rate": 1.1422972711333247e-06, "loss": 0.0171, "step": 74820 }, { "grad_norm": 0.5103939175605774, "learning_rate": 1.1379087499010565e-06, "loss": 0.0224, "step": 74830 }, { "grad_norm": 0.25332319736480713, "learning_rate": 1.1335285778638127e-06, "loss": 0.0132, "step": 74840 }, { "grad_norm": 0.36194756627082825, "learning_rate": 1.1291567557700333e-06, "loss": 0.0179, "step": 74850 }, { "grad_norm": 0.9603939056396484, "learning_rate": 1.124793284366743e-06, "loss": 0.0171, "step": 74860 }, { "grad_norm": 0.3488251268863678, "learning_rate": 1.1204381643995455e-06, "loss": 0.0141, "step": 74870 }, { "grad_norm": 0.23926934599876404, "learning_rate": 1.116091396612595e-06, "loss": 0.0167, "step": 74880 }, { "grad_norm": 0.3272155225276947, "learning_rate": 1.1117529817486538e-06, "loss": 0.0207, "step": 74890 }, { "grad_norm": 0.261545330286026, "learning_rate": 1.107422920549034e-06, "loss": 0.0144, "step": 74900 }, { "grad_norm": 0.3443794846534729, "learning_rate": 1.1031012137536157e-06, "loss": 0.0172, "step": 74910 }, { "grad_norm": 0.339569628238678, "learning_rate": 1.0987878621008695e-06, "loss": 0.0193, "step": 74920 }, { "grad_norm": 0.2491903156042099, "learning_rate": 1.0944828663278284e-06, "loss": 0.0194, "step": 74930 }, { "grad_norm": 0.23153682053089142, "learning_rate": 1.0901862271700925e-06, "loss": 0.0223, "step": 74940 }, { "grad_norm": 0.4898313879966736, "learning_rate": 1.0858979453618467e-06, "loss": 0.0161, "step": 74950 }, { "grad_norm": 1.0844902992248535, "learning_rate": 1.0816180216358441e-06, "loss": 0.0191, "step": 74960 }, { "grad_norm": 0.39752188324928284, "learning_rate": 1.0773464567233937e-06, "loss": 0.0163, "step": 74970 }, { "grad_norm": 0.5934195518493652, "learning_rate": 1.0730832513543953e-06, "loss": 0.0205, "step": 74980 }, { "grad_norm": 0.7529779076576233, "learning_rate": 1.0688284062573218e-06, "loss": 0.0223, "step": 74990 }, { "grad_norm": 0.40267375111579895, "learning_rate": 1.0645819221591969e-06, "loss": 0.0152, "step": 75000 }, { "grad_norm": 0.3274322748184204, "learning_rate": 1.0603437997856347e-06, "loss": 0.0174, "step": 75010 }, { "grad_norm": 0.1750950813293457, "learning_rate": 1.0561140398608228e-06, "loss": 0.0147, "step": 75020 }, { "grad_norm": 0.30010440945625305, "learning_rate": 1.0518926431074937e-06, "loss": 0.0184, "step": 75030 }, { "grad_norm": 0.287809282541275, "learning_rate": 1.0476796102469877e-06, "loss": 0.0159, "step": 75040 }, { "grad_norm": 0.36479347944259644, "learning_rate": 1.0434749419991786e-06, "loss": 0.0188, "step": 75050 }, { "grad_norm": 0.30756524205207825, "learning_rate": 1.0392786390825415e-06, "loss": 0.0181, "step": 75060 }, { "grad_norm": 0.35394856333732605, "learning_rate": 1.0350907022141087e-06, "loss": 0.0173, "step": 75070 }, { "grad_norm": 0.23102298378944397, "learning_rate": 1.0309111321094744e-06, "loss": 0.0164, "step": 75080 }, { "grad_norm": 1.3936703205108643, "learning_rate": 1.026739929482834e-06, "loss": 0.0155, "step": 75090 }, { "grad_norm": 0.3404157757759094, "learning_rate": 1.022577095046906e-06, "loss": 0.0212, "step": 75100 }, { "grad_norm": 0.2026640772819519, "learning_rate": 1.0184226295130217e-06, "loss": 0.0172, "step": 75110 }, { "grad_norm": 0.3185576796531677, "learning_rate": 1.0142765335910575e-06, "loss": 0.0218, "step": 75120 }, { "grad_norm": 0.13136470317840576, "learning_rate": 1.0101388079894746e-06, "loss": 0.0155, "step": 75130 }, { "grad_norm": 0.5000818371772766, "learning_rate": 1.0060094534152908e-06, "loss": 0.0233, "step": 75140 }, { "grad_norm": 0.18925559520721436, "learning_rate": 1.0018884705741028e-06, "loss": 0.0147, "step": 75150 }, { "grad_norm": 0.2956635653972626, "learning_rate": 9.977758601700805e-07, "loss": 0.0248, "step": 75160 }, { "grad_norm": 0.798133134841919, "learning_rate": 9.93671622905945e-07, "loss": 0.0225, "step": 75170 }, { "grad_norm": 0.18890886008739471, "learning_rate": 9.895757594830024e-07, "loss": 0.0165, "step": 75180 }, { "grad_norm": 0.27420303225517273, "learning_rate": 9.854882706011204e-07, "loss": 0.0143, "step": 75190 }, { "grad_norm": 0.22322975099086761, "learning_rate": 9.814091569587513e-07, "loss": 0.0162, "step": 75200 }, { "grad_norm": 0.3346371352672577, "learning_rate": 9.77338419252888e-07, "loss": 0.0238, "step": 75210 }, { "grad_norm": 0.2945425808429718, "learning_rate": 9.732760581791289e-07, "loss": 0.0205, "step": 75220 }, { "grad_norm": 0.3175554871559143, "learning_rate": 9.692220744315972e-07, "loss": 0.0165, "step": 75230 }, { "grad_norm": 0.0799325481057167, "learning_rate": 9.651764687030162e-07, "loss": 0.0162, "step": 75240 }, { "grad_norm": 0.31119292974472046, "learning_rate": 9.611392416846776e-07, "loss": 0.0138, "step": 75250 }, { "grad_norm": 1.4300380945205688, "learning_rate": 9.571103940664238e-07, "loss": 0.0147, "step": 75260 }, { "grad_norm": 1.293534755706787, "learning_rate": 9.530899265366822e-07, "loss": 0.0193, "step": 75270 }, { "grad_norm": 0.12254633754491806, "learning_rate": 9.490778397824307e-07, "loss": 0.0187, "step": 75280 }, { "grad_norm": 1.4706668853759766, "learning_rate": 9.450741344892322e-07, "loss": 0.0185, "step": 75290 }, { "grad_norm": 0.30303388833999634, "learning_rate": 9.410788113412116e-07, "loss": 0.0165, "step": 75300 }, { "grad_norm": 0.4497573971748352, "learning_rate": 9.370918710210563e-07, "loss": 0.0143, "step": 75310 }, { "grad_norm": 0.22354382276535034, "learning_rate": 9.331133142100268e-07, "loss": 0.0158, "step": 75320 }, { "grad_norm": 0.18837134540081024, "learning_rate": 9.291431415879459e-07, "loss": 0.0188, "step": 75330 }, { "grad_norm": 0.360870897769928, "learning_rate": 9.251813538332155e-07, "loss": 0.0235, "step": 75340 }, { "grad_norm": 0.6996094584465027, "learning_rate": 9.212279516227884e-07, "loss": 0.0156, "step": 75350 }, { "grad_norm": 0.3825102746486664, "learning_rate": 9.172829356322021e-07, "loss": 0.0179, "step": 75360 }, { "grad_norm": 0.38292306661605835, "learning_rate": 9.133463065355452e-07, "loss": 0.0195, "step": 75370 }, { "grad_norm": 0.45634105801582336, "learning_rate": 9.094180650054796e-07, "loss": 0.0236, "step": 75380 }, { "grad_norm": 0.24942049384117126, "learning_rate": 9.054982117132404e-07, "loss": 0.0184, "step": 75390 }, { "grad_norm": 0.39016884565353394, "learning_rate": 9.015867473286143e-07, "loss": 0.0191, "step": 75400 }, { "grad_norm": 0.44318056106567383, "learning_rate": 8.976836725199778e-07, "loss": 0.0174, "step": 75410 }, { "grad_norm": 0.25970351696014404, "learning_rate": 8.937889879542416e-07, "loss": 0.0156, "step": 75420 }, { "grad_norm": 0.36381617188453674, "learning_rate": 8.899026942969068e-07, "loss": 0.0188, "step": 75430 }, { "grad_norm": 0.443684458732605, "learning_rate": 8.860247922120424e-07, "loss": 0.0154, "step": 75440 }, { "grad_norm": 0.32736262679100037, "learning_rate": 8.821552823622737e-07, "loss": 0.0193, "step": 75450 }, { "grad_norm": 0.47327882051467896, "learning_rate": 8.782941654087884e-07, "loss": 0.0175, "step": 75460 }, { "grad_norm": 0.3452114462852478, "learning_rate": 8.744414420113478e-07, "loss": 0.0219, "step": 75470 }, { "grad_norm": 0.38850826025009155, "learning_rate": 8.705971128282753e-07, "loss": 0.016, "step": 75480 }, { "grad_norm": 0.35429316759109497, "learning_rate": 8.66761178516473e-07, "loss": 0.0178, "step": 75490 }, { "grad_norm": 0.5721496343612671, "learning_rate": 8.629336397313781e-07, "loss": 0.0217, "step": 75500 }, { "grad_norm": 0.1537563055753708, "learning_rate": 8.591144971270282e-07, "loss": 0.021, "step": 75510 }, { "grad_norm": 0.26137790083885193, "learning_rate": 8.553037513560069e-07, "loss": 0.0255, "step": 75520 }, { "grad_norm": 0.27618542313575745, "learning_rate": 8.515014030694546e-07, "loss": 0.0148, "step": 75530 }, { "grad_norm": 0.2266199141740799, "learning_rate": 8.47707452917107e-07, "loss": 0.0175, "step": 75540 }, { "grad_norm": 0.7308844923973083, "learning_rate": 8.439219015472343e-07, "loss": 0.0174, "step": 75550 }, { "grad_norm": 0.22628501057624817, "learning_rate": 8.401447496066861e-07, "loss": 0.0135, "step": 75560 }, { "grad_norm": 0.2885206639766693, "learning_rate": 8.363759977408792e-07, "loss": 0.0142, "step": 75570 }, { "grad_norm": 0.2592752277851105, "learning_rate": 8.326156465937817e-07, "loss": 0.0189, "step": 75580 }, { "grad_norm": 0.6048799753189087, "learning_rate": 8.28863696807941e-07, "loss": 0.0269, "step": 75590 }, { "grad_norm": 0.2731459438800812, "learning_rate": 8.251201490244609e-07, "loss": 0.0203, "step": 75600 }, { "grad_norm": 1.0765115022659302, "learning_rate": 8.213850038830129e-07, "loss": 0.0154, "step": 75610 }, { "grad_norm": 0.1957743614912033, "learning_rate": 8.176582620218309e-07, "loss": 0.0192, "step": 75620 }, { "grad_norm": 0.28699252009391785, "learning_rate": 8.139399240777057e-07, "loss": 0.0114, "step": 75630 }, { "grad_norm": 0.6708692908287048, "learning_rate": 8.102299906860122e-07, "loss": 0.0227, "step": 75640 }, { "grad_norm": 0.341251015663147, "learning_rate": 8.065284624806657e-07, "loss": 0.0195, "step": 75650 }, { "grad_norm": 0.2981089651584625, "learning_rate": 8.028353400941601e-07, "loss": 0.0191, "step": 75660 }, { "grad_norm": 0.22466683387756348, "learning_rate": 7.991506241575575e-07, "loss": 0.0233, "step": 75670 }, { "grad_norm": 0.38105303049087524, "learning_rate": 7.954743153004541e-07, "loss": 0.0176, "step": 75680 }, { "grad_norm": 0.3919644355773926, "learning_rate": 7.918064141510528e-07, "loss": 0.0163, "step": 75690 }, { "grad_norm": 0.4269086420536041, "learning_rate": 7.881469213360859e-07, "loss": 0.0215, "step": 75700 }, { "grad_norm": 0.2208535373210907, "learning_rate": 7.844958374808642e-07, "loss": 0.0131, "step": 75710 }, { "grad_norm": 0.21370500326156616, "learning_rate": 7.808531632092608e-07, "loss": 0.0154, "step": 75720 }, { "grad_norm": 0.27402564883232117, "learning_rate": 7.772188991436946e-07, "loss": 0.0148, "step": 75730 }, { "grad_norm": 0.15285784006118774, "learning_rate": 7.735930459051799e-07, "loss": 0.0182, "step": 75740 }, { "grad_norm": 0.20273558795452118, "learning_rate": 7.699756041132655e-07, "loss": 0.0164, "step": 75750 }, { "grad_norm": 0.14932852983474731, "learning_rate": 7.663665743860793e-07, "loss": 0.0182, "step": 75760 }, { "grad_norm": 0.14593932032585144, "learning_rate": 7.627659573403001e-07, "loss": 0.0149, "step": 75770 }, { "grad_norm": 0.790799617767334, "learning_rate": 7.591737535911802e-07, "loss": 0.0175, "step": 75780 }, { "grad_norm": 0.2693912088871002, "learning_rate": 7.555899637525288e-07, "loss": 0.0205, "step": 75790 }, { "grad_norm": 0.3113752007484436, "learning_rate": 7.520145884367058e-07, "loss": 0.0172, "step": 75800 }, { "grad_norm": 0.5636363625526428, "learning_rate": 7.484476282546615e-07, "loss": 0.0155, "step": 75810 }, { "grad_norm": 0.2693956196308136, "learning_rate": 7.448890838158806e-07, "loss": 0.0204, "step": 75820 }, { "grad_norm": 0.3251126706600189, "learning_rate": 7.413389557284267e-07, "loss": 0.0222, "step": 75830 }, { "grad_norm": 0.2628958523273468, "learning_rate": 7.377972445989201e-07, "loss": 0.0186, "step": 75840 }, { "grad_norm": 0.5134045481681824, "learning_rate": 7.342639510325377e-07, "loss": 0.0197, "step": 75850 }, { "grad_norm": 1.1671644449234009, "learning_rate": 7.307390756330246e-07, "loss": 0.0246, "step": 75860 }, { "grad_norm": 0.2518017888069153, "learning_rate": 7.272226190026876e-07, "loss": 0.0148, "step": 75870 }, { "grad_norm": 0.2429250031709671, "learning_rate": 7.237145817423907e-07, "loss": 0.0124, "step": 75880 }, { "grad_norm": 0.24513261020183563, "learning_rate": 7.202149644515654e-07, "loss": 0.0156, "step": 75890 }, { "grad_norm": 0.4661323130130768, "learning_rate": 7.167237677281946e-07, "loss": 0.0206, "step": 75900 }, { "grad_norm": 0.19103224575519562, "learning_rate": 7.132409921688288e-07, "loss": 0.0202, "step": 75910 }, { "grad_norm": 0.3739233911037445, "learning_rate": 7.09766638368592e-07, "loss": 0.0204, "step": 75920 }, { "grad_norm": 0.2561556398868561, "learning_rate": 7.063007069211313e-07, "loss": 0.0179, "step": 75930 }, { "grad_norm": 0.3299544155597687, "learning_rate": 7.028431984187067e-07, "loss": 0.0157, "step": 75940 }, { "grad_norm": 0.19150002300739288, "learning_rate": 6.993941134520898e-07, "loss": 0.0151, "step": 75950 }, { "grad_norm": 0.17314793169498444, "learning_rate": 6.959534526106537e-07, "loss": 0.0167, "step": 75960 }, { "grad_norm": 0.20864835381507874, "learning_rate": 6.925212164822948e-07, "loss": 0.0145, "step": 75970 }, { "grad_norm": 0.32016876339912415, "learning_rate": 6.890974056535049e-07, "loss": 0.0254, "step": 75980 }, { "grad_norm": 0.48175618052482605, "learning_rate": 6.856820207093106e-07, "loss": 0.0184, "step": 75990 }, { "grad_norm": 0.4872777760028839, "learning_rate": 6.822750622333063e-07, "loss": 0.0143, "step": 76000 }, { "grad_norm": 0.4398421049118042, "learning_rate": 6.788765308076539e-07, "loss": 0.0183, "step": 76010 }, { "grad_norm": 0.3308489918708801, "learning_rate": 6.754864270130668e-07, "loss": 0.0128, "step": 76020 }, { "grad_norm": 1.2456448078155518, "learning_rate": 6.721047514288203e-07, "loss": 0.0163, "step": 76030 }, { "grad_norm": 0.3348568379878998, "learning_rate": 6.687315046327469e-07, "loss": 0.0146, "step": 76040 }, { "grad_norm": 0.7287457585334778, "learning_rate": 6.65366687201252e-07, "loss": 0.0178, "step": 76050 }, { "grad_norm": 0.2202511578798294, "learning_rate": 6.620102997092814e-07, "loss": 0.0208, "step": 76060 }, { "grad_norm": 0.5580980777740479, "learning_rate": 6.586623427303596e-07, "loss": 0.0146, "step": 76070 }, { "grad_norm": 0.47661092877388, "learning_rate": 6.553228168365455e-07, "loss": 0.0199, "step": 76080 }, { "grad_norm": 1.1801836490631104, "learning_rate": 6.519917225984884e-07, "loss": 0.0186, "step": 76090 }, { "grad_norm": 0.9220129251480103, "learning_rate": 6.486690605853718e-07, "loss": 0.018, "step": 76100 }, { "grad_norm": 0.3379957377910614, "learning_rate": 6.453548313649527e-07, "loss": 0.0152, "step": 76110 }, { "grad_norm": 0.5746924877166748, "learning_rate": 6.420490355035446e-07, "loss": 0.0195, "step": 76120 }, { "grad_norm": 0.18658503890037537, "learning_rate": 6.387516735660071e-07, "loss": 0.0172, "step": 76130 }, { "grad_norm": 0.3581477105617523, "learning_rate": 6.35462746115778e-07, "loss": 0.0154, "step": 76140 }, { "grad_norm": 0.2402796596288681, "learning_rate": 6.321822537148358e-07, "loss": 0.0194, "step": 76150 }, { "grad_norm": 0.37551403045654297, "learning_rate": 6.289101969237432e-07, "loss": 0.0157, "step": 76160 }, { "grad_norm": 0.29958054423332214, "learning_rate": 6.256465763015918e-07, "loss": 0.0194, "step": 76170 }, { "grad_norm": 0.3089480698108673, "learning_rate": 6.22391392406052e-07, "loss": 0.0181, "step": 76180 }, { "grad_norm": 0.2603748142719269, "learning_rate": 6.191446457933403e-07, "loss": 0.0181, "step": 76190 }, { "grad_norm": 0.322477251291275, "learning_rate": 6.159063370182406e-07, "loss": 0.0145, "step": 76200 }, { "grad_norm": 0.34914833307266235, "learning_rate": 6.126764666340879e-07, "loss": 0.0161, "step": 76210 }, { "grad_norm": 0.2172713428735733, "learning_rate": 6.094550351927852e-07, "loss": 0.0167, "step": 76220 }, { "grad_norm": 0.4734705090522766, "learning_rate": 6.062420432447757e-07, "loss": 0.0155, "step": 76230 }, { "grad_norm": 0.45151403546333313, "learning_rate": 6.030374913390813e-07, "loss": 0.0178, "step": 76240 }, { "grad_norm": 0.2757217288017273, "learning_rate": 5.998413800232694e-07, "loss": 0.0164, "step": 76250 }, { "grad_norm": 0.22471529245376587, "learning_rate": 5.966537098434755e-07, "loss": 0.0221, "step": 76260 }, { "grad_norm": 0.8352425694465637, "learning_rate": 5.934744813443694e-07, "loss": 0.0193, "step": 76270 }, { "grad_norm": 0.3485296666622162, "learning_rate": 5.903036950692054e-07, "loss": 0.0255, "step": 76280 }, { "grad_norm": 0.23750647902488708, "learning_rate": 5.871413515597835e-07, "loss": 0.017, "step": 76290 }, { "grad_norm": 0.6280633807182312, "learning_rate": 5.839874513564547e-07, "loss": 0.013, "step": 76300 }, { "grad_norm": 0.7196628451347351, "learning_rate": 5.808419949981436e-07, "loss": 0.0189, "step": 76310 }, { "grad_norm": 0.3442879617214203, "learning_rate": 5.777049830223257e-07, "loss": 0.0161, "step": 76320 }, { "grad_norm": 0.2581399977207184, "learning_rate": 5.745764159650114e-07, "loss": 0.0129, "step": 76330 }, { "grad_norm": 0.30965450406074524, "learning_rate": 5.71456294360806e-07, "loss": 0.0209, "step": 76340 }, { "grad_norm": 0.25064167380332947, "learning_rate": 5.683446187428443e-07, "loss": 0.0218, "step": 76350 }, { "grad_norm": 0.29717618227005005, "learning_rate": 5.652413896428288e-07, "loss": 0.0152, "step": 76360 }, { "grad_norm": 0.3477522134780884, "learning_rate": 5.621466075910131e-07, "loss": 0.0173, "step": 76370 }, { "grad_norm": 0.15045766532421112, "learning_rate": 5.590602731162187e-07, "loss": 0.0147, "step": 76380 }, { "grad_norm": 0.3391945958137512, "learning_rate": 5.559823867458125e-07, "loss": 0.0136, "step": 76390 }, { "grad_norm": 0.32770583033561707, "learning_rate": 5.529129490057128e-07, "loss": 0.026, "step": 76400 }, { "grad_norm": 0.1663677841424942, "learning_rate": 5.498519604204167e-07, "loss": 0.0181, "step": 76410 }, { "grad_norm": 0.2831283509731293, "learning_rate": 5.467994215129501e-07, "loss": 0.0152, "step": 76420 }, { "grad_norm": 0.2709849178791046, "learning_rate": 5.437553328049183e-07, "loss": 0.0205, "step": 76430 }, { "grad_norm": 0.16755031049251556, "learning_rate": 5.407196948164661e-07, "loss": 0.0143, "step": 76440 }, { "grad_norm": 0.5952063798904419, "learning_rate": 5.376925080663009e-07, "loss": 0.0152, "step": 76450 }, { "grad_norm": 0.5349721312522888, "learning_rate": 5.346737730716977e-07, "loss": 0.0246, "step": 76460 }, { "grad_norm": 0.2275862693786621, "learning_rate": 5.316634903484607e-07, "loss": 0.0151, "step": 76470 }, { "grad_norm": 0.18417677283287048, "learning_rate": 5.286616604109728e-07, "loss": 0.0165, "step": 76480 }, { "grad_norm": 0.2807301878929138, "learning_rate": 5.256682837721627e-07, "loss": 0.0214, "step": 76490 }, { "grad_norm": 0.2750479578971863, "learning_rate": 5.226833609435156e-07, "loss": 0.0158, "step": 76500 }, { "grad_norm": 0.15236249566078186, "learning_rate": 5.19706892435079e-07, "loss": 0.0175, "step": 76510 }, { "grad_norm": 0.25523367524147034, "learning_rate": 5.167388787554406e-07, "loss": 0.0199, "step": 76520 }, { "grad_norm": 0.20736296474933624, "learning_rate": 5.137793204117614e-07, "loss": 0.0152, "step": 76530 }, { "grad_norm": 0.3666602671146393, "learning_rate": 5.108282179097479e-07, "loss": 0.0197, "step": 76540 }, { "grad_norm": 0.348995566368103, "learning_rate": 5.078855717536523e-07, "loss": 0.0197, "step": 76550 }, { "grad_norm": 0.33663782477378845, "learning_rate": 5.049513824463059e-07, "loss": 0.0178, "step": 76560 }, { "grad_norm": 0.21252793073654175, "learning_rate": 5.020256504890742e-07, "loss": 0.0224, "step": 76570 }, { "grad_norm": 0.15266531705856323, "learning_rate": 4.991083763818849e-07, "loss": 0.0181, "step": 76580 }, { "grad_norm": 0.3058796525001526, "learning_rate": 4.961995606232228e-07, "loss": 0.0153, "step": 76590 }, { "grad_norm": 0.41973838210105896, "learning_rate": 4.932992037101236e-07, "loss": 0.017, "step": 76600 }, { "grad_norm": 0.259082168340683, "learning_rate": 4.904073061381798e-07, "loss": 0.0202, "step": 76610 }, { "grad_norm": 0.40588414669036865, "learning_rate": 4.875238684015349e-07, "loss": 0.0202, "step": 76620 }, { "grad_norm": 0.2645553648471832, "learning_rate": 4.846488909928948e-07, "loss": 0.0192, "step": 76630 }, { "grad_norm": 0.3017762303352356, "learning_rate": 4.81782374403511e-07, "loss": 0.0206, "step": 76640 }, { "grad_norm": 0.5037373304367065, "learning_rate": 4.789243191231918e-07, "loss": 0.0158, "step": 76650 }, { "grad_norm": 0.17794564366340637, "learning_rate": 4.76074725640302e-07, "loss": 0.0212, "step": 76660 }, { "grad_norm": 0.20964884757995605, "learning_rate": 4.732335944417632e-07, "loss": 0.0206, "step": 76670 }, { "grad_norm": 0.3667774200439453, "learning_rate": 4.704009260130371e-07, "loss": 0.0208, "step": 76680 }, { "grad_norm": 0.3691938519477844, "learning_rate": 4.675767208381587e-07, "loss": 0.0144, "step": 76690 }, { "grad_norm": 0.3050735294818878, "learning_rate": 4.647609793997032e-07, "loss": 0.0175, "step": 76700 }, { "grad_norm": 0.17538732290267944, "learning_rate": 4.619537021788023e-07, "loss": 0.0144, "step": 76710 }, { "grad_norm": 0.2157488763332367, "learning_rate": 4.5915488965515005e-07, "loss": 0.0149, "step": 76720 }, { "grad_norm": 0.22239138185977936, "learning_rate": 4.563645423069807e-07, "loss": 0.0189, "step": 76730 }, { "grad_norm": 0.28246596455574036, "learning_rate": 4.535826606110849e-07, "loss": 0.0164, "step": 76740 }, { "grad_norm": 0.25515732169151306, "learning_rate": 4.508092450428214e-07, "loss": 0.0199, "step": 76750 }, { "grad_norm": 0.345998615026474, "learning_rate": 4.480442960760778e-07, "loss": 0.0197, "step": 76760 }, { "grad_norm": 0.8615365028381348, "learning_rate": 4.4528781418332053e-07, "loss": 0.0185, "step": 76770 }, { "grad_norm": 0.37494438886642456, "learning_rate": 4.4253979983555073e-07, "loss": 0.0242, "step": 76780 }, { "grad_norm": 0.4057652950286865, "learning_rate": 4.3980025350233154e-07, "loss": 0.0157, "step": 76790 }, { "grad_norm": 0.47858425974845886, "learning_rate": 4.370691756517664e-07, "loss": 0.0152, "step": 76800 }, { "grad_norm": 0.5265386700630188, "learning_rate": 4.3434656675053753e-07, "loss": 0.0202, "step": 76810 }, { "grad_norm": 0.27726325392723083, "learning_rate": 4.3163242726385613e-07, "loss": 0.0182, "step": 76820 }, { "grad_norm": 0.39435115456581116, "learning_rate": 4.289267576554956e-07, "loss": 0.0202, "step": 76830 }, { "grad_norm": 0.3612188696861267, "learning_rate": 4.262295583877807e-07, "loss": 0.0182, "step": 76840 }, { "grad_norm": 0.17031966149806976, "learning_rate": 4.235408299215815e-07, "loss": 0.0192, "step": 76850 }, { "grad_norm": 0.7367569804191589, "learning_rate": 4.2086057271634173e-07, "loss": 0.0159, "step": 76860 }, { "grad_norm": 0.23496223986148834, "learning_rate": 4.181887872300394e-07, "loss": 0.0168, "step": 76870 }, { "grad_norm": 1.6727049350738525, "learning_rate": 4.155254739192038e-07, "loss": 0.0182, "step": 76880 }, { "grad_norm": 0.20087334513664246, "learning_rate": 4.128706332389265e-07, "loss": 0.0232, "step": 76890 }, { "grad_norm": 0.32812759280204773, "learning_rate": 4.1022426564284453e-07, "loss": 0.0184, "step": 76900 }, { "grad_norm": 0.37291890382766724, "learning_rate": 4.075863715831574e-07, "loss": 0.0173, "step": 76910 }, { "grad_norm": 0.5219247341156006, "learning_rate": 4.0495695151059887e-07, "loss": 0.0207, "step": 76920 }, { "grad_norm": 0.31344354152679443, "learning_rate": 4.023360058744763e-07, "loss": 0.0143, "step": 76930 }, { "grad_norm": 0.2590302526950836, "learning_rate": 3.997235351226314e-07, "loss": 0.012, "step": 76940 }, { "grad_norm": 0.3679355978965759, "learning_rate": 3.971195397014571e-07, "loss": 0.0209, "step": 76950 }, { "grad_norm": 0.26872897148132324, "learning_rate": 3.9452402005591417e-07, "loss": 0.026, "step": 76960 }, { "grad_norm": 0.7775712013244629, "learning_rate": 3.9193697662950334e-07, "loss": 0.0181, "step": 76970 }, { "grad_norm": 0.7383076548576355, "learning_rate": 3.893584098642822e-07, "loss": 0.0271, "step": 76980 }, { "grad_norm": 0.21667258441448212, "learning_rate": 3.867883202008538e-07, "loss": 0.0221, "step": 76990 }, { "grad_norm": 0.28633272647857666, "learning_rate": 3.842267080783779e-07, "loss": 0.0191, "step": 77000 }, { "grad_norm": 0.21185599267482758, "learning_rate": 3.816735739345656e-07, "loss": 0.0195, "step": 77010 }, { "grad_norm": 0.18136505782604218, "learning_rate": 3.791289182056679e-07, "loss": 0.0195, "step": 77020 }, { "grad_norm": 0.4312654137611389, "learning_rate": 3.7659274132650913e-07, "loss": 0.0219, "step": 77030 }, { "grad_norm": 0.23670288920402527, "learning_rate": 3.7406504373044826e-07, "loss": 0.0157, "step": 77040 }, { "grad_norm": 0.30655017495155334, "learning_rate": 3.7154582584939533e-07, "loss": 0.0165, "step": 77050 }, { "grad_norm": 0.43174993991851807, "learning_rate": 3.690350881138227e-07, "loss": 0.0162, "step": 77060 }, { "grad_norm": 0.17207126319408417, "learning_rate": 3.665328309527427e-07, "loss": 0.0132, "step": 77070 }, { "grad_norm": 0.3413521945476532, "learning_rate": 3.640390547937245e-07, "loss": 0.0203, "step": 77080 }, { "grad_norm": 1.0115617513656616, "learning_rate": 3.615537600628882e-07, "loss": 0.019, "step": 77090 }, { "grad_norm": 0.26279494166374207, "learning_rate": 3.5907694718489427e-07, "loss": 0.0187, "step": 77100 }, { "grad_norm": 0.27154695987701416, "learning_rate": 3.566086165829707e-07, "loss": 0.0176, "step": 77110 }, { "grad_norm": 0.9880568981170654, "learning_rate": 3.5414876867888024e-07, "loss": 0.0181, "step": 77120 }, { "grad_norm": 0.3484429717063904, "learning_rate": 3.5169740389295326e-07, "loss": 0.0255, "step": 77130 }, { "grad_norm": 0.26212063431739807, "learning_rate": 3.4925452264405466e-07, "loss": 0.0168, "step": 77140 }, { "grad_norm": 0.3645387291908264, "learning_rate": 3.468201253496062e-07, "loss": 0.0168, "step": 77150 }, { "grad_norm": 0.2639573812484741, "learning_rate": 3.443942124255861e-07, "loss": 0.0195, "step": 77160 }, { "grad_norm": 0.4468762278556824, "learning_rate": 3.4197678428650183e-07, "loss": 0.0206, "step": 77170 }, { "grad_norm": 0.253719687461853, "learning_rate": 3.3956784134544504e-07, "loss": 0.017, "step": 77180 }, { "grad_norm": 0.1695486605167389, "learning_rate": 3.3716738401402547e-07, "loss": 0.0142, "step": 77190 }, { "grad_norm": 0.2233036756515503, "learning_rate": 3.3477541270241495e-07, "loss": 0.0164, "step": 77200 }, { "grad_norm": 0.8699365854263306, "learning_rate": 3.3239192781934215e-07, "loss": 0.0201, "step": 77210 }, { "grad_norm": 0.32309964299201965, "learning_rate": 3.3001692977207563e-07, "loss": 0.022, "step": 77220 }, { "grad_norm": 0.35317525267601013, "learning_rate": 3.276504189664409e-07, "loss": 0.018, "step": 77230 }, { "grad_norm": 0.2315220981836319, "learning_rate": 3.252923958068088e-07, "loss": 0.0168, "step": 77240 }, { "grad_norm": 0.11661120504140854, "learning_rate": 3.2294286069609046e-07, "loss": 0.0156, "step": 77250 }, { "grad_norm": 0.3342747986316681, "learning_rate": 3.2060181403577583e-07, "loss": 0.0139, "step": 77260 }, { "grad_norm": 0.2439201921224594, "learning_rate": 3.1826925622587823e-07, "loss": 0.0178, "step": 77270 }, { "grad_norm": 0.24636171758174896, "learning_rate": 3.159451876649622e-07, "loss": 0.0156, "step": 77280 }, { "grad_norm": 0.4127085506916046, "learning_rate": 3.1362960875015444e-07, "loss": 0.0195, "step": 77290 }, { "grad_norm": 0.41844823956489563, "learning_rate": 3.1132251987711637e-07, "loss": 0.0245, "step": 77300 }, { "grad_norm": 0.33211109042167664, "learning_rate": 3.0902392144007695e-07, "loss": 0.0199, "step": 77310 }, { "grad_norm": 0.28383785486221313, "learning_rate": 3.0673381383179436e-07, "loss": 0.0149, "step": 77320 }, { "grad_norm": 0.45936572551727295, "learning_rate": 3.0445219744358875e-07, "loss": 0.0175, "step": 77330 }, { "grad_norm": 0.5987603664398193, "learning_rate": 3.021790726653262e-07, "loss": 0.0209, "step": 77340 }, { "grad_norm": 0.5052998661994934, "learning_rate": 2.9991443988542366e-07, "loss": 0.0149, "step": 77350 }, { "grad_norm": 0.4303431212902069, "learning_rate": 2.976582994908439e-07, "loss": 0.0162, "step": 77360 }, { "grad_norm": 0.18267802894115448, "learning_rate": 2.954106518671007e-07, "loss": 0.0182, "step": 77370 }, { "grad_norm": 0.27840855717658997, "learning_rate": 2.9317149739825356e-07, "loss": 0.0155, "step": 77380 }, { "grad_norm": 0.512131929397583, "learning_rate": 2.909408364669075e-07, "loss": 0.0245, "step": 77390 }, { "grad_norm": 0.5510477423667908, "learning_rate": 2.887186694542299e-07, "loss": 0.0189, "step": 77400 }, { "grad_norm": 0.33557525277137756, "learning_rate": 2.865049967399225e-07, "loss": 0.0223, "step": 77410 }, { "grad_norm": 0.5090655088424683, "learning_rate": 2.842998187022439e-07, "loss": 0.0178, "step": 77420 }, { "grad_norm": 0.18526437878608704, "learning_rate": 2.8210313571800374e-07, "loss": 0.0136, "step": 77430 }, { "grad_norm": 0.37848037481307983, "learning_rate": 2.7991494816255184e-07, "loss": 0.0154, "step": 77440 }, { "grad_norm": 0.20307371020317078, "learning_rate": 2.777352564097779e-07, "loss": 0.0167, "step": 77450 }, { "grad_norm": 0.2307002991437912, "learning_rate": 2.755640608321508e-07, "loss": 0.0147, "step": 77460 }, { "grad_norm": 0.3151826560497284, "learning_rate": 2.73401361800657e-07, "loss": 0.0161, "step": 77470 }, { "grad_norm": 0.35775837302207947, "learning_rate": 2.7124715968484537e-07, "loss": 0.0182, "step": 77480 }, { "grad_norm": 0.29830899834632874, "learning_rate": 2.691014548528104e-07, "loss": 0.0184, "step": 77490 }, { "grad_norm": 0.6929206252098083, "learning_rate": 2.669642476711864e-07, "loss": 0.0181, "step": 77500 }, { "grad_norm": 0.3152700662612915, "learning_rate": 2.648355385051815e-07, "loss": 0.0207, "step": 77510 }, { "grad_norm": 0.32635095715522766, "learning_rate": 2.627153277185157e-07, "loss": 0.0189, "step": 77520 }, { "grad_norm": 0.22120030224323273, "learning_rate": 2.606036156734881e-07, "loss": 0.0145, "step": 77530 }, { "grad_norm": 0.33623936772346497, "learning_rate": 2.5850040273092127e-07, "loss": 0.0206, "step": 77540 }, { "grad_norm": 0.198281928896904, "learning_rate": 2.5640568925020536e-07, "loss": 0.0261, "step": 77550 }, { "grad_norm": 0.6357131004333496, "learning_rate": 2.5431947558927083e-07, "loss": 0.0189, "step": 77560 }, { "grad_norm": 0.2529788017272949, "learning_rate": 2.522417621045825e-07, "loss": 0.015, "step": 77570 }, { "grad_norm": 0.33389243483543396, "learning_rate": 2.501725491511786e-07, "loss": 0.0215, "step": 77580 }, { "grad_norm": 0.6923825144767761, "learning_rate": 2.481118370826263e-07, "loss": 0.0186, "step": 77590 }, { "grad_norm": 0.2390104979276657, "learning_rate": 2.4605962625104393e-07, "loss": 0.0145, "step": 77600 }, { "grad_norm": 0.1894754320383072, "learning_rate": 2.440159170070955e-07, "loss": 0.02, "step": 77610 }, { "grad_norm": 0.37097200751304626, "learning_rate": 2.41980709699996e-07, "loss": 0.0167, "step": 77620 }, { "grad_norm": 0.27578580379486084, "learning_rate": 2.3995400467751174e-07, "loss": 0.0164, "step": 77630 }, { "grad_norm": 0.17581410706043243, "learning_rate": 2.3793580228594902e-07, "loss": 0.0273, "step": 77640 }, { "grad_norm": 0.33919769525527954, "learning_rate": 2.3592610287015982e-07, "loss": 0.0179, "step": 77650 }, { "grad_norm": 0.3937433958053589, "learning_rate": 2.3392490677354718e-07, "loss": 0.0142, "step": 77660 }, { "grad_norm": 0.41607651114463806, "learning_rate": 2.3193221433806533e-07, "loss": 0.0161, "step": 77670 }, { "grad_norm": 0.5154563784599304, "learning_rate": 2.299480259042086e-07, "loss": 0.0185, "step": 77680 }, { "grad_norm": 0.15537823736667633, "learning_rate": 2.2797234181102244e-07, "loss": 0.0189, "step": 77690 }, { "grad_norm": 0.1970391869544983, "learning_rate": 2.2600516239609238e-07, "loss": 0.0164, "step": 77700 }, { "grad_norm": 0.17855386435985565, "learning_rate": 2.240464879955606e-07, "loss": 0.0156, "step": 77710 }, { "grad_norm": 0.22377333045005798, "learning_rate": 2.2209631894410387e-07, "loss": 0.0169, "step": 77720 }, { "grad_norm": 0.3886922001838684, "learning_rate": 2.2015465557496117e-07, "loss": 0.0206, "step": 77730 }, { "grad_norm": 0.3474125266075134, "learning_rate": 2.1822149821990602e-07, "loss": 0.0219, "step": 77740 }, { "grad_norm": 0.43760472536087036, "learning_rate": 2.1629684720926303e-07, "loss": 0.0209, "step": 77750 }, { "grad_norm": 0.7942317128181458, "learning_rate": 2.1438070287189693e-07, "loss": 0.0186, "step": 77760 }, { "grad_norm": 1.3402705192565918, "learning_rate": 2.1247306553523472e-07, "loss": 0.0167, "step": 77770 }, { "grad_norm": 1.2279144525527954, "learning_rate": 2.105739355252323e-07, "loss": 0.0182, "step": 77780 }, { "grad_norm": 0.23545615375041962, "learning_rate": 2.0868331316639678e-07, "loss": 0.0138, "step": 77790 }, { "grad_norm": 0.2967003583908081, "learning_rate": 2.0680119878179193e-07, "loss": 0.0188, "step": 77800 }, { "grad_norm": 1.1178035736083984, "learning_rate": 2.0492759269301054e-07, "loss": 0.0194, "step": 77810 }, { "grad_norm": 0.28008827567100525, "learning_rate": 2.0306249522021315e-07, "loss": 0.0126, "step": 77820 }, { "grad_norm": 0.26445281505584717, "learning_rate": 2.0120590668207816e-07, "loss": 0.0195, "step": 77830 }, { "grad_norm": 0.4757964015007019, "learning_rate": 1.993578273958574e-07, "loss": 0.0162, "step": 77840 }, { "grad_norm": 0.12254171073436737, "learning_rate": 1.975182576773371e-07, "loss": 0.0158, "step": 77850 }, { "grad_norm": 0.42524802684783936, "learning_rate": 1.9568719784083812e-07, "loss": 0.0206, "step": 77860 }, { "grad_norm": 0.5664888024330139, "learning_rate": 1.9386464819924898e-07, "loss": 0.0226, "step": 77870 }, { "grad_norm": 0.19216367602348328, "learning_rate": 1.9205060906399285e-07, "loss": 0.0192, "step": 77880 }, { "grad_norm": 0.4345351755619049, "learning_rate": 1.9024508074503845e-07, "loss": 0.0146, "step": 77890 }, { "grad_norm": 0.20965617895126343, "learning_rate": 1.8844806355089452e-07, "loss": 0.0223, "step": 77900 }, { "grad_norm": 0.48819872736930847, "learning_rate": 1.8665955778863208e-07, "loss": 0.025, "step": 77910 }, { "grad_norm": 0.3943646550178528, "learning_rate": 1.848795637638512e-07, "loss": 0.017, "step": 77920 }, { "grad_norm": 0.29626455903053284, "learning_rate": 1.831080817807085e-07, "loss": 0.0188, "step": 77930 }, { "grad_norm": 0.22457590699195862, "learning_rate": 1.813451121418952e-07, "loss": 0.0257, "step": 77940 }, { "grad_norm": 0.6272268295288086, "learning_rate": 1.795906551486648e-07, "loss": 0.0158, "step": 77950 }, { "grad_norm": 0.8451939225196838, "learning_rate": 1.7784471110079414e-07, "loss": 0.0168, "step": 77960 }, { "grad_norm": 0.4384794235229492, "learning_rate": 1.7610728029662793e-07, "loss": 0.0174, "step": 77970 }, { "grad_norm": 0.18418431282043457, "learning_rate": 1.7437836303303979e-07, "loss": 0.016, "step": 77980 }, { "grad_norm": 0.40512070059776306, "learning_rate": 1.726579596054545e-07, "loss": 0.0139, "step": 77990 }, { "grad_norm": 0.5248584151268005, "learning_rate": 1.7094607030784803e-07, "loss": 0.0179, "step": 78000 }, { "grad_norm": 0.7073989510536194, "learning_rate": 1.6924269543272532e-07, "loss": 0.0158, "step": 78010 }, { "grad_norm": 0.5920606255531311, "learning_rate": 1.6754783527115348e-07, "loss": 0.0163, "step": 78020 }, { "grad_norm": 0.14018233120441437, "learning_rate": 1.6586149011273422e-07, "loss": 0.017, "step": 78030 }, { "grad_norm": 0.2771703004837036, "learning_rate": 1.6418366024562038e-07, "loss": 0.0258, "step": 78040 }, { "grad_norm": 0.5380282998085022, "learning_rate": 1.6251434595651037e-07, "loss": 0.0193, "step": 78050 }, { "grad_norm": 0.18051116168498993, "learning_rate": 1.608535475306372e-07, "loss": 0.0183, "step": 78060 }, { "grad_norm": 0.5934971570968628, "learning_rate": 1.5920126525179048e-07, "loss": 0.0147, "step": 78070 }, { "grad_norm": 0.18472813069820404, "learning_rate": 1.5755749940229435e-07, "loss": 0.0127, "step": 78080 }, { "grad_norm": 0.65616774559021, "learning_rate": 1.559222502630353e-07, "loss": 0.0185, "step": 78090 }, { "grad_norm": 0.26784050464630127, "learning_rate": 1.5429551811341757e-07, "loss": 0.016, "step": 78100 }, { "grad_norm": 0.2696925103664398, "learning_rate": 1.5267730323141882e-07, "loss": 0.0198, "step": 78110 }, { "grad_norm": 0.3080157935619354, "learning_rate": 1.5106760589353454e-07, "loss": 0.0184, "step": 78120 }, { "grad_norm": 0.20898397266864777, "learning_rate": 1.4946642637483355e-07, "loss": 0.0174, "step": 78130 }, { "grad_norm": 0.31120434403419495, "learning_rate": 1.4787376494889703e-07, "loss": 0.0161, "step": 78140 }, { "grad_norm": 0.36466389894485474, "learning_rate": 1.4628962188787955e-07, "loss": 0.0136, "step": 78150 }, { "grad_norm": 0.2557185888290405, "learning_rate": 1.4471399746247006e-07, "loss": 0.0199, "step": 78160 }, { "grad_norm": 0.5733533501625061, "learning_rate": 1.4314689194188103e-07, "loss": 0.0155, "step": 78170 }, { "grad_norm": 0.4258490800857544, "learning_rate": 1.4158830559390933e-07, "loss": 0.0203, "step": 78180 }, { "grad_norm": 0.3951478898525238, "learning_rate": 1.4003823868486422e-07, "loss": 0.0265, "step": 78190 }, { "grad_norm": 0.34258443117141724, "learning_rate": 1.3849669147960598e-07, "loss": 0.02, "step": 78200 }, { "grad_norm": 0.40672311186790466, "learning_rate": 1.3696366424155726e-07, "loss": 0.0166, "step": 78210 }, { "grad_norm": 0.680377721786499, "learning_rate": 1.35439157232653e-07, "loss": 0.0211, "step": 78220 }, { "grad_norm": 0.1560877114534378, "learning_rate": 1.3392317071340144e-07, "loss": 0.0148, "step": 78230 }, { "grad_norm": 0.3317056894302368, "learning_rate": 1.3241570494283984e-07, "loss": 0.0163, "step": 78240 }, { "grad_norm": 0.2085685133934021, "learning_rate": 1.3091676017855103e-07, "loss": 0.0215, "step": 78250 }, { "grad_norm": 0.24384069442749023, "learning_rate": 1.2942633667666904e-07, "loss": 0.017, "step": 78260 }, { "grad_norm": 2.390561819076538, "learning_rate": 1.2794443469185679e-07, "loss": 0.0228, "step": 78270 }, { "grad_norm": 0.3356262147426605, "learning_rate": 1.2647105447734508e-07, "loss": 0.0117, "step": 78280 }, { "grad_norm": 0.18009589612483978, "learning_rate": 1.2500619628488254e-07, "loss": 0.0193, "step": 78290 }, { "grad_norm": 0.7598472237586975, "learning_rate": 1.2354986036477446e-07, "loss": 0.0238, "step": 78300 }, { "grad_norm": 0.26632094383239746, "learning_rate": 1.221020469658718e-07, "loss": 0.0207, "step": 78310 }, { "grad_norm": 0.8529630303382874, "learning_rate": 1.2066275633556556e-07, "loss": 0.0165, "step": 78320 }, { "grad_norm": 0.23566588759422302, "learning_rate": 1.192319887197979e-07, "loss": 0.016, "step": 78330 }, { "grad_norm": 0.3600490987300873, "learning_rate": 1.1780974436303438e-07, "loss": 0.0141, "step": 78340 }, { "grad_norm": 0.35544025897979736, "learning_rate": 1.1639602350830836e-07, "loss": 0.0151, "step": 78350 }, { "grad_norm": 0.23271049559116364, "learning_rate": 1.1499082639718217e-07, "loss": 0.0185, "step": 78360 }, { "grad_norm": 0.08927909284830093, "learning_rate": 1.1359415326976374e-07, "loss": 0.015, "step": 78370 }, { "grad_norm": 0.25593453645706177, "learning_rate": 1.1220600436470663e-07, "loss": 0.0163, "step": 78380 }, { "grad_norm": 0.24572667479515076, "learning_rate": 1.1082637991920996e-07, "loss": 0.0149, "step": 78390 }, { "grad_norm": 0.24991969764232635, "learning_rate": 1.0945528016901851e-07, "loss": 0.0194, "step": 78400 }, { "grad_norm": 0.4987271726131439, "learning_rate": 1.0809270534840599e-07, "loss": 0.0176, "step": 78410 }, { "grad_norm": 0.3215004503726959, "learning_rate": 1.067386556902028e-07, "loss": 0.0151, "step": 78420 }, { "grad_norm": 0.1831200122833252, "learning_rate": 1.05393131425785e-07, "loss": 0.0107, "step": 78430 }, { "grad_norm": 0.38944968581199646, "learning_rate": 1.0405613278505199e-07, "loss": 0.0171, "step": 78440 }, { "grad_norm": 0.27115342020988464, "learning_rate": 1.027276599964766e-07, "loss": 0.0128, "step": 78450 }, { "grad_norm": 0.5256133079528809, "learning_rate": 1.0140771328704391e-07, "loss": 0.0183, "step": 78460 }, { "grad_norm": 0.3848491609096527, "learning_rate": 1.0009629288231237e-07, "loss": 0.0181, "step": 78470 }, { "grad_norm": 0.2577888071537018, "learning_rate": 9.879339900635276e-08, "loss": 0.0174, "step": 78480 }, { "grad_norm": 0.3256710469722748, "learning_rate": 9.74990318817981e-08, "loss": 0.023, "step": 78490 }, { "grad_norm": 0.2698575556278229, "learning_rate": 9.621319172982701e-08, "loss": 0.0185, "step": 78500 }, { "grad_norm": 0.16563230752944946, "learning_rate": 9.493587877015264e-08, "loss": 0.0179, "step": 78510 }, { "grad_norm": 0.28301477432250977, "learning_rate": 9.366709322102263e-08, "loss": 0.0176, "step": 78520 }, { "grad_norm": 0.8813892006874084, "learning_rate": 9.240683529924688e-08, "loss": 0.0165, "step": 78530 }, { "grad_norm": 0.4302455186843872, "learning_rate": 9.115510522016979e-08, "loss": 0.0147, "step": 78540 }, { "grad_norm": 0.19713501632213593, "learning_rate": 8.991190319767583e-08, "loss": 0.0163, "step": 78550 }, { "grad_norm": 0.22235029935836792, "learning_rate": 8.867722944419509e-08, "loss": 0.0133, "step": 78560 }, { "grad_norm": 0.3364603817462921, "learning_rate": 8.745108417069214e-08, "loss": 0.0165, "step": 78570 }, { "grad_norm": 0.16832280158996582, "learning_rate": 8.623346758669381e-08, "loss": 0.0122, "step": 78580 }, { "grad_norm": 0.3961941599845886, "learning_rate": 8.502437990025037e-08, "loss": 0.0147, "step": 78590 }, { "grad_norm": 0.22205811738967896, "learning_rate": 8.382382131795764e-08, "loss": 0.017, "step": 78600 }, { "grad_norm": 1.229482889175415, "learning_rate": 8.263179204496818e-08, "loss": 0.0225, "step": 78610 }, { "grad_norm": 0.6339410543441772, "learning_rate": 8.144829228496354e-08, "loss": 0.0189, "step": 78620 }, { "grad_norm": 0.26684755086898804, "learning_rate": 8.027332224016526e-08, "loss": 0.0199, "step": 78630 }, { "grad_norm": 0.33948466181755066, "learning_rate": 7.910688211135164e-08, "loss": 0.0236, "step": 78640 }, { "grad_norm": 0.4949132204055786, "learning_rate": 7.794897209783546e-08, "loss": 0.0191, "step": 78650 }, { "grad_norm": 0.9324135780334473, "learning_rate": 7.679959239746403e-08, "loss": 0.0163, "step": 78660 }, { "grad_norm": 0.6559485197067261, "learning_rate": 7.565874320664135e-08, "loss": 0.0172, "step": 78670 }, { "grad_norm": 0.412244588136673, "learning_rate": 7.45264247203059e-08, "loss": 0.0209, "step": 78680 }, { "grad_norm": 0.18914245069026947, "learning_rate": 7.340263713194184e-08, "loss": 0.0155, "step": 78690 }, { "grad_norm": 0.5779626369476318, "learning_rate": 7.228738063356777e-08, "loss": 0.018, "step": 78700 }, { "grad_norm": 0.39186015725135803, "learning_rate": 7.118065541575903e-08, "loss": 0.0222, "step": 78710 }, { "grad_norm": 0.09236432611942291, "learning_rate": 7.008246166761435e-08, "loss": 0.0141, "step": 78720 }, { "grad_norm": 0.18535801768302917, "learning_rate": 6.899279957679472e-08, "loss": 0.0171, "step": 78730 }, { "grad_norm": 0.5049319267272949, "learning_rate": 6.791166932949011e-08, "loss": 0.0177, "step": 78740 }, { "grad_norm": 0.3400198817253113, "learning_rate": 6.683907111043608e-08, "loss": 0.0174, "step": 78750 }, { "grad_norm": 0.21940290927886963, "learning_rate": 6.577500510290824e-08, "loss": 0.0178, "step": 78760 }, { "grad_norm": 0.3971763551235199, "learning_rate": 6.47194714887278e-08, "loss": 0.0357, "step": 78770 }, { "grad_norm": 0.2689003348350525, "learning_rate": 6.367247044825608e-08, "loss": 0.018, "step": 78780 }, { "grad_norm": 0.326164573431015, "learning_rate": 6.263400216039994e-08, "loss": 0.0143, "step": 78790 }, { "grad_norm": 0.37747398018836975, "learning_rate": 6.160406680260078e-08, "loss": 0.0299, "step": 78800 }, { "grad_norm": 0.18176713585853577, "learning_rate": 6.058266455084561e-08, "loss": 0.0184, "step": 78810 }, { "grad_norm": 0.9168412089347839, "learning_rate": 5.956979557967257e-08, "loss": 0.0147, "step": 78820 }, { "grad_norm": 0.739059567451477, "learning_rate": 5.856546006214325e-08, "loss": 0.0234, "step": 78830 }, { "grad_norm": 0.3335282504558563, "learning_rate": 5.7569658169881466e-08, "loss": 0.0134, "step": 78840 }, { "grad_norm": 0.12759532034397125, "learning_rate": 5.658239007303445e-08, "loss": 0.0152, "step": 78850 }, { "grad_norm": 0.44767841696739197, "learning_rate": 5.560365594030059e-08, "loss": 0.0237, "step": 78860 }, { "grad_norm": 0.43578773736953735, "learning_rate": 5.463345593891833e-08, "loss": 0.0197, "step": 78870 }, { "grad_norm": 0.3862093985080719, "learning_rate": 5.367179023467173e-08, "loss": 0.0223, "step": 78880 }, { "grad_norm": 0.6674002408981323, "learning_rate": 5.271865899187378e-08, "loss": 0.0167, "step": 78890 }, { "grad_norm": 0.4311682879924774, "learning_rate": 5.177406237340532e-08, "loss": 0.0293, "step": 78900 }, { "grad_norm": 0.44620752334594727, "learning_rate": 5.083800054065946e-08, "loss": 0.0151, "step": 78910 }, { "grad_norm": 0.262433797121048, "learning_rate": 4.991047365358603e-08, "loss": 0.015, "step": 78920 }, { "grad_norm": 0.46437370777130127, "learning_rate": 4.899148187067493e-08, "loss": 0.0257, "step": 78930 }, { "grad_norm": 0.3854385316371918, "learning_rate": 4.808102534895609e-08, "loss": 0.0137, "step": 78940 }, { "grad_norm": 0.2460242658853531, "learning_rate": 4.717910424400507e-08, "loss": 0.0221, "step": 78950 }, { "grad_norm": 0.40517088770866394, "learning_rate": 4.628571870993193e-08, "loss": 0.0216, "step": 78960 }, { "grad_norm": 0.4369730055332184, "learning_rate": 4.5400868899392324e-08, "loss": 0.024, "step": 78970 }, { "grad_norm": 0.43202102184295654, "learning_rate": 4.4524554963576436e-08, "loss": 0.0224, "step": 78980 }, { "grad_norm": 0.38633885979652405, "learning_rate": 4.36567770522367e-08, "loss": 0.0184, "step": 78990 }, { "grad_norm": 0.19999516010284424, "learning_rate": 4.279753531364339e-08, "loss": 0.014, "step": 79000 }, { "grad_norm": 0.46987900137901306, "learning_rate": 4.1946829894617955e-08, "loss": 0.0167, "step": 79010 }, { "grad_norm": 0.31596067547798157, "learning_rate": 4.11046609405219e-08, "loss": 0.0213, "step": 79020 }, { "grad_norm": 0.25018060207366943, "learning_rate": 4.027102859526233e-08, "loss": 0.0126, "step": 79030 }, { "grad_norm": 0.24070365726947784, "learning_rate": 3.9445933001280856e-08, "loss": 0.0173, "step": 79040 }, { "grad_norm": 0.2853321433067322, "learning_rate": 3.8629374299564704e-08, "loss": 0.0181, "step": 79050 }, { "grad_norm": 0.30778536200523376, "learning_rate": 3.78213526296467e-08, "loss": 0.0204, "step": 79060 }, { "grad_norm": 0.2180715799331665, "learning_rate": 3.702186812958308e-08, "loss": 0.0181, "step": 79070 }, { "grad_norm": 0.35355016589164734, "learning_rate": 3.623092093599789e-08, "loss": 0.0145, "step": 79080 }, { "grad_norm": 0.3003993332386017, "learning_rate": 3.544851118403303e-08, "loss": 0.0165, "step": 79090 }, { "grad_norm": 0.3890445828437805, "learning_rate": 3.46746390073871e-08, "loss": 0.0177, "step": 79100 }, { "grad_norm": 0.24784013628959656, "learning_rate": 3.390930453829322e-08, "loss": 0.0171, "step": 79110 }, { "grad_norm": 0.13609309494495392, "learning_rate": 3.3152507907519005e-08, "loss": 0.0146, "step": 79120 }, { "grad_norm": 0.27266353368759155, "learning_rate": 3.240424924438323e-08, "loss": 0.0167, "step": 79130 }, { "grad_norm": 0.2192535251379013, "learning_rate": 3.1664528676750295e-08, "loss": 0.0159, "step": 79140 }, { "grad_norm": 0.5112272500991821, "learning_rate": 3.093334633100797e-08, "loss": 0.0287, "step": 79150 }, { "grad_norm": 0.24074327945709229, "learning_rate": 3.021070233210077e-08, "loss": 0.0154, "step": 79160 }, { "grad_norm": 0.7096522450447083, "learning_rate": 2.9496596803507692e-08, "loss": 0.0179, "step": 79170 }, { "grad_norm": 0.26363468170166016, "learning_rate": 2.879102986725335e-08, "loss": 0.0183, "step": 79180 }, { "grad_norm": 0.2775930166244507, "learning_rate": 2.809400164389131e-08, "loss": 0.0219, "step": 79190 }, { "grad_norm": 0.22097450494766235, "learning_rate": 2.740551225253185e-08, "loss": 0.0164, "step": 79200 }, { "grad_norm": 0.27497509121894836, "learning_rate": 2.6725561810819754e-08, "loss": 0.0213, "step": 79210 }, { "grad_norm": 0.34207668900489807, "learning_rate": 2.605415043493431e-08, "loss": 0.0155, "step": 79220 }, { "grad_norm": 0.18378975987434387, "learning_rate": 2.5391278239605965e-08, "loss": 0.018, "step": 79230 }, { "grad_norm": 0.318583607673645, "learning_rate": 2.4736945338094118e-08, "loss": 0.0192, "step": 79240 }, { "grad_norm": 0.24295879900455475, "learning_rate": 2.4091151842214887e-08, "loss": 0.0167, "step": 79250 }, { "grad_norm": 0.19927892088890076, "learning_rate": 2.3453897862318884e-08, "loss": 0.0149, "step": 79260 }, { "grad_norm": 0.19103726744651794, "learning_rate": 2.2825183507285686e-08, "loss": 0.0166, "step": 79270 }, { "grad_norm": 0.4617324471473694, "learning_rate": 2.220500888455157e-08, "loss": 0.0209, "step": 79280 }, { "grad_norm": 0.2056758552789688, "learning_rate": 2.1593374100081777e-08, "loss": 0.0238, "step": 79290 }, { "grad_norm": 0.2654021680355072, "learning_rate": 2.09902792583927e-08, "loss": 0.0264, "step": 79300 }, { "grad_norm": 0.20329703390598297, "learning_rate": 2.0395724462540788e-08, "loss": 0.0156, "step": 79310 }, { "grad_norm": 0.31042012572288513, "learning_rate": 1.9809709814111453e-08, "loss": 0.0186, "step": 79320 }, { "grad_norm": 0.1413334608078003, "learning_rate": 1.923223541324126e-08, "loss": 0.0172, "step": 79330 }, { "grad_norm": 0.27205386757850647, "learning_rate": 1.8663301358606833e-08, "loss": 0.0231, "step": 79340 }, { "grad_norm": 0.194636270403862, "learning_rate": 1.8102907747419295e-08, "loss": 0.0233, "step": 79350 }, { "grad_norm": 0.7650447487831116, "learning_rate": 1.7551054675435385e-08, "loss": 0.0192, "step": 79360 }, { "grad_norm": 0.42401060461997986, "learning_rate": 1.7007742236957447e-08, "loss": 0.0188, "step": 79370 }, { "grad_norm": 0.5440788865089417, "learning_rate": 1.647297052481678e-08, "loss": 0.0202, "step": 79380 }, { "grad_norm": 0.24848328530788422, "learning_rate": 1.5946739630390285e-08, "loss": 0.0199, "step": 79390 }, { "grad_norm": 0.16737833619117737, "learning_rate": 1.5429049643606032e-08, "loss": 0.0198, "step": 79400 }, { "grad_norm": 0.13736140727996826, "learning_rate": 1.4919900652909935e-08, "loss": 0.0219, "step": 79410 }, { "grad_norm": 0.14441031217575073, "learning_rate": 1.4419292745310175e-08, "loss": 0.019, "step": 79420 }, { "grad_norm": 0.24591569602489471, "learning_rate": 1.3927226006343885e-08, "loss": 0.0184, "step": 79430 }, { "grad_norm": 0.29881641268730164, "learning_rate": 1.3443700520093805e-08, "loss": 0.0218, "step": 79440 }, { "grad_norm": 0.19435274600982666, "learning_rate": 1.296871636917718e-08, "loss": 0.0163, "step": 79450 }, { "grad_norm": 1.5520553588867188, "learning_rate": 1.2502273634762419e-08, "loss": 0.0139, "step": 79460 }, { "grad_norm": 0.33597898483276367, "learning_rate": 1.2044372396546876e-08, "loss": 0.0168, "step": 79470 }, { "grad_norm": 0.41187307238578796, "learning_rate": 1.1595012732773524e-08, "loss": 0.0198, "step": 79480 }, { "grad_norm": 0.2504931688308716, "learning_rate": 1.1154194720225386e-08, "loss": 0.0168, "step": 79490 }, { "grad_norm": 0.23583069443702698, "learning_rate": 1.0721918434231093e-08, "loss": 0.0154, "step": 79500 }, { "grad_norm": 1.625802755355835, "learning_rate": 1.0298183948648232e-08, "loss": 0.0204, "step": 79510 }, { "grad_norm": 0.2785186469554901, "learning_rate": 9.88299133588555e-09, "loss": 0.0184, "step": 79520 }, { "grad_norm": 0.5515143871307373, "learning_rate": 9.476340666891847e-09, "loss": 0.0179, "step": 79530 }, { "grad_norm": 0.20412376523017883, "learning_rate": 9.078232011139332e-09, "loss": 0.0131, "step": 79540 }, { "grad_norm": 0.20603206753730774, "learning_rate": 8.68866543666802e-09, "loss": 0.0148, "step": 79550 }, { "grad_norm": 0.357711523771286, "learning_rate": 8.307641010035782e-09, "loss": 0.0184, "step": 79560 }, { "grad_norm": 0.3922427296638489, "learning_rate": 7.93515879635165e-09, "loss": 0.0159, "step": 79570 }, { "grad_norm": 0.2741034924983978, "learning_rate": 7.571218859264706e-09, "loss": 0.0121, "step": 79580 }, { "grad_norm": 0.6659750938415527, "learning_rate": 7.215821260958544e-09, "loss": 0.0191, "step": 79590 }, { "grad_norm": 0.3550127446651459, "learning_rate": 6.868966062162363e-09, "loss": 0.0221, "step": 79600 }, { "grad_norm": 0.43969595432281494, "learning_rate": 6.530653322145419e-09, "loss": 0.0156, "step": 79610 }, { "grad_norm": 0.2244899570941925, "learning_rate": 6.200883098717025e-09, "loss": 0.0199, "step": 79620 }, { "grad_norm": 0.4928900897502899, "learning_rate": 5.879655448226551e-09, "loss": 0.023, "step": 79630 }, { "grad_norm": 0.5006168484687805, "learning_rate": 5.566970425557872e-09, "loss": 0.0167, "step": 79640 }, { "grad_norm": 0.3827381134033203, "learning_rate": 5.26282808414047e-09, "loss": 0.016, "step": 79650 }, { "grad_norm": 0.16313740611076355, "learning_rate": 4.967228475949437e-09, "loss": 0.0152, "step": 79660 }, { "grad_norm": 0.20693513751029968, "learning_rate": 4.680171651494369e-09, "loss": 0.0173, "step": 79670 }, { "grad_norm": 0.2928350865840912, "learning_rate": 4.40165765981937e-09, "loss": 0.0179, "step": 79680 }, { "grad_norm": 0.21889926493167877, "learning_rate": 4.131686548519698e-09, "loss": 0.0142, "step": 79690 }, { "grad_norm": 0.26086679100990295, "learning_rate": 3.8702583637251214e-09, "loss": 0.014, "step": 79700 }, { "grad_norm": 0.13897868990898132, "learning_rate": 3.617373150105463e-09, "loss": 0.0141, "step": 79710 }, { "grad_norm": 0.2817053496837616, "learning_rate": 3.3730309508706036e-09, "loss": 0.0139, "step": 79720 }, { "grad_norm": 0.4943157136440277, "learning_rate": 3.137231807781582e-09, "loss": 0.0157, "step": 79730 }, { "grad_norm": 0.1301524043083191, "learning_rate": 2.9099757611172894e-09, "loss": 0.0193, "step": 79740 }, { "grad_norm": 0.20998142659664154, "learning_rate": 2.6912628497133275e-09, "loss": 0.0174, "step": 79750 }, { "grad_norm": 0.669145941734314, "learning_rate": 2.481093110945354e-09, "loss": 0.0179, "step": 79760 }, { "grad_norm": 0.7613884806632996, "learning_rate": 2.2794665807235327e-09, "loss": 0.0216, "step": 79770 }, { "grad_norm": 0.16709773242473602, "learning_rate": 2.0863832934980843e-09, "loss": 0.0143, "step": 79780 }, { "grad_norm": 0.1914851814508438, "learning_rate": 1.901843282264837e-09, "loss": 0.0222, "step": 79790 }, { "grad_norm": 0.37685468792915344, "learning_rate": 1.7258465785541245e-09, "loss": 0.0198, "step": 79800 }, { "grad_norm": 0.2881806790828705, "learning_rate": 1.5583932124418887e-09, "loss": 0.0188, "step": 79810 }, { "grad_norm": 0.43427079916000366, "learning_rate": 1.3994832125441282e-09, "loss": 0.0173, "step": 79820 }, { "grad_norm": 0.2398228645324707, "learning_rate": 1.2491166060057958e-09, "loss": 0.0196, "step": 79830 }, { "grad_norm": 0.14783379435539246, "learning_rate": 1.1072934185230032e-09, "loss": 0.0175, "step": 79840 }, { "grad_norm": 0.2755509912967682, "learning_rate": 9.740136743319195e-10, "loss": 0.0166, "step": 79850 }, { "grad_norm": 1.0371137857437134, "learning_rate": 8.492773962087696e-10, "loss": 0.0177, "step": 79860 }, { "grad_norm": 0.4603922665119171, "learning_rate": 7.330846054587337e-10, "loss": 0.0256, "step": 79870 }, { "grad_norm": 0.3896399736404419, "learning_rate": 6.254353219492526e-10, "loss": 0.0149, "step": 79880 }, { "grad_norm": 0.2402418702840805, "learning_rate": 5.263295640600685e-10, "loss": 0.0158, "step": 79890 }, { "grad_norm": 0.6451910138130188, "learning_rate": 4.357673487387359e-10, "loss": 0.0165, "step": 79900 }, { "grad_norm": 0.3060459494590759, "learning_rate": 3.537486914506616e-10, "loss": 0.0148, "step": 79910 }, { "grad_norm": 0.3586720824241638, "learning_rate": 2.8027360621241117e-10, "loss": 0.0231, "step": 79920 }, { "grad_norm": 2.4549341201782227, "learning_rate": 2.153421055806071e-10, "loss": 0.0144, "step": 79930 }, { "grad_norm": 0.35950416326522827, "learning_rate": 1.5895420064637733e-10, "loss": 0.017, "step": 79940 }, { "grad_norm": 0.320956826210022, "learning_rate": 1.1110990105200891e-10, "loss": 0.0188, "step": 79950 }, { "grad_norm": 0.31847333908081055, "learning_rate": 7.180921496874326e-11, "loss": 0.0175, "step": 79960 }, { "grad_norm": 0.5140222907066345, "learning_rate": 4.10521491134297e-11, "loss": 0.0191, "step": 79970 }, { "grad_norm": 0.5678942799568176, "learning_rate": 1.883870874297422e-11, "loss": 0.0172, "step": 79980 }, { "grad_norm": 0.2877882122993469, "learning_rate": 5.1688976432373584e-12, "loss": 0.0207, "step": 79990 }, { "grad_norm": 0.42737773060798645, "learning_rate": 4.271816234080461e-14, "loss": 0.0154, "step": 80000 } ], "logging_steps": 10, "max_steps": 80000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 20000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 8, "trial_name": null, "trial_params": null }