{ "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": 1.200006127357483, "learning_rate": 2.25e-07, "loss": 1.4409, "step": 10 }, { "grad_norm": 1.2063723802566528, "learning_rate": 4.75e-07, "loss": 1.4262, "step": 20 }, { "grad_norm": 1.2707029581069946, "learning_rate": 7.25e-07, "loss": 1.4333, "step": 30 }, { "grad_norm": 1.0610071420669556, "learning_rate": 9.75e-07, "loss": 1.4414, "step": 40 }, { "grad_norm": 1.1525142192840576, "learning_rate": 1.2250000000000001e-06, "loss": 1.4226, "step": 50 }, { "grad_norm": 1.7864906787872314, "learning_rate": 1.475e-06, "loss": 1.3706, "step": 60 }, { "grad_norm": 1.7839109897613525, "learning_rate": 1.7250000000000002e-06, "loss": 1.2422, "step": 70 }, { "grad_norm": 2.2468364238739014, "learning_rate": 1.975e-06, "loss": 1.2061, "step": 80 }, { "grad_norm": 1.483929991722107, "learning_rate": 2.225e-06, "loss": 1.2341, "step": 90 }, { "grad_norm": 2.063420295715332, "learning_rate": 2.4750000000000004e-06, "loss": 1.1741, "step": 100 }, { "grad_norm": 1.5929186344146729, "learning_rate": 2.725e-06, "loss": 1.1469, "step": 110 }, { "grad_norm": 1.5417307615280151, "learning_rate": 2.975e-06, "loss": 1.1555, "step": 120 }, { "grad_norm": 1.531018853187561, "learning_rate": 3.225e-06, "loss": 1.1233, "step": 130 }, { "grad_norm": 1.448808193206787, "learning_rate": 3.4750000000000006e-06, "loss": 1.1236, "step": 140 }, { "grad_norm": 1.5196325778961182, "learning_rate": 3.725e-06, "loss": 1.0768, "step": 150 }, { "grad_norm": 1.8622503280639648, "learning_rate": 3.975e-06, "loss": 1.0736, "step": 160 }, { "grad_norm": 1.6263045072555542, "learning_rate": 4.225e-06, "loss": 1.0679, "step": 170 }, { "grad_norm": 1.4813590049743652, "learning_rate": 4.475e-06, "loss": 1.0633, "step": 180 }, { "grad_norm": 1.3066281080245972, "learning_rate": 4.7250000000000005e-06, "loss": 1.0799, "step": 190 }, { "grad_norm": 1.8845033645629883, "learning_rate": 4.975000000000001e-06, "loss": 1.0645, "step": 200 }, { "grad_norm": 1.3808010816574097, "learning_rate": 5.225e-06, "loss": 1.044, "step": 210 }, { "grad_norm": 1.2643654346466064, "learning_rate": 5.475e-06, "loss": 1.0683, "step": 220 }, { "grad_norm": 1.4082114696502686, "learning_rate": 5.725e-06, "loss": 1.0516, "step": 230 }, { "grad_norm": 1.7742204666137695, "learning_rate": 5.975e-06, "loss": 1.0544, "step": 240 }, { "grad_norm": 1.3008482456207275, "learning_rate": 6.2250000000000005e-06, "loss": 1.0465, "step": 250 }, { "grad_norm": 1.1290251016616821, "learning_rate": 6.475000000000001e-06, "loss": 1.0263, "step": 260 }, { "grad_norm": 1.563814640045166, "learning_rate": 6.725000000000001e-06, "loss": 1.0478, "step": 270 }, { "grad_norm": 1.2906379699707031, "learning_rate": 6.975000000000001e-06, "loss": 1.0407, "step": 280 }, { "grad_norm": 1.7122255563735962, "learning_rate": 7.2249999999999994e-06, "loss": 1.0404, "step": 290 }, { "grad_norm": 1.0537691116333008, "learning_rate": 7.4750000000000004e-06, "loss": 1.0285, "step": 300 }, { "grad_norm": 1.2008037567138672, "learning_rate": 7.725e-06, "loss": 1.0496, "step": 310 }, { "grad_norm": 2.077982187271118, "learning_rate": 7.975e-06, "loss": 1.0305, "step": 320 }, { "grad_norm": 1.4955880641937256, "learning_rate": 8.225e-06, "loss": 1.0132, "step": 330 }, { "grad_norm": 2.20841121673584, "learning_rate": 8.475000000000001e-06, "loss": 1.007, "step": 340 }, { "grad_norm": 1.8666841983795166, "learning_rate": 8.725e-06, "loss": 0.9892, "step": 350 }, { "grad_norm": 1.674803376197815, "learning_rate": 8.975e-06, "loss": 0.9502, "step": 360 }, { "grad_norm": 2.1492748260498047, "learning_rate": 9.225e-06, "loss": 0.9561, "step": 370 }, { "grad_norm": 2.4265661239624023, "learning_rate": 9.475e-06, "loss": 0.914, "step": 380 }, { "grad_norm": 2.005295753479004, "learning_rate": 9.725000000000001e-06, "loss": 0.8856, "step": 390 }, { "grad_norm": 2.7577269077301025, "learning_rate": 9.975e-06, "loss": 0.8667, "step": 400 }, { "grad_norm": 2.4309558868408203, "learning_rate": 1.0225e-05, "loss": 0.8369, "step": 410 }, { "grad_norm": 2.381694793701172, "learning_rate": 1.0475e-05, "loss": 0.8134, "step": 420 }, { "grad_norm": 2.495739459991455, "learning_rate": 1.0725e-05, "loss": 0.7806, "step": 430 }, { "grad_norm": 2.118256092071533, "learning_rate": 1.0975e-05, "loss": 0.7701, "step": 440 }, { "grad_norm": 2.7149157524108887, "learning_rate": 1.1225e-05, "loss": 0.714, "step": 450 }, { "grad_norm": 2.7888669967651367, "learning_rate": 1.1475000000000001e-05, "loss": 0.7065, "step": 460 }, { "grad_norm": 2.399935722351074, "learning_rate": 1.1725e-05, "loss": 0.6893, "step": 470 }, { "grad_norm": 2.913469076156616, "learning_rate": 1.1975e-05, "loss": 0.6774, "step": 480 }, { "grad_norm": 2.899045705795288, "learning_rate": 1.2225e-05, "loss": 0.6401, "step": 490 }, { "grad_norm": 2.808506488800049, "learning_rate": 1.2475e-05, "loss": 0.6038, "step": 500 }, { "grad_norm": 2.572528123855591, "learning_rate": 1.2725000000000001e-05, "loss": 0.5939, "step": 510 }, { "grad_norm": 2.908060073852539, "learning_rate": 1.2975e-05, "loss": 0.5833, "step": 520 }, { "grad_norm": 3.056756019592285, "learning_rate": 1.3225000000000001e-05, "loss": 0.5731, "step": 530 }, { "grad_norm": 2.3631210327148438, "learning_rate": 1.3475000000000002e-05, "loss": 0.4907, "step": 540 }, { "grad_norm": 2.9230501651763916, "learning_rate": 1.3725000000000002e-05, "loss": 0.5311, "step": 550 }, { "grad_norm": 3.602245807647705, "learning_rate": 1.3975000000000003e-05, "loss": 0.5182, "step": 560 }, { "grad_norm": 2.4906649589538574, "learning_rate": 1.4225e-05, "loss": 0.4543, "step": 570 }, { "grad_norm": 2.824880838394165, "learning_rate": 1.4475e-05, "loss": 0.4374, "step": 580 }, { "grad_norm": 2.9583728313446045, "learning_rate": 1.4725e-05, "loss": 0.4187, "step": 590 }, { "grad_norm": 2.464958906173706, "learning_rate": 1.4975e-05, "loss": 0.4122, "step": 600 }, { "grad_norm": 2.1925864219665527, "learning_rate": 1.5225e-05, "loss": 0.4016, "step": 610 }, { "grad_norm": 3.542553424835205, "learning_rate": 1.5475e-05, "loss": 0.3792, "step": 620 }, { "grad_norm": 2.3703863620758057, "learning_rate": 1.5725e-05, "loss": 0.3797, "step": 630 }, { "grad_norm": 3.155971050262451, "learning_rate": 1.5975000000000002e-05, "loss": 0.3418, "step": 640 }, { "grad_norm": 3.2014236450195312, "learning_rate": 1.6225e-05, "loss": 0.3214, "step": 650 }, { "grad_norm": 3.561274290084839, "learning_rate": 1.6475e-05, "loss": 0.3275, "step": 660 }, { "grad_norm": 2.9534718990325928, "learning_rate": 1.6725000000000003e-05, "loss": 0.289, "step": 670 }, { "grad_norm": 2.7297427654266357, "learning_rate": 1.6975000000000003e-05, "loss": 0.29, "step": 680 }, { "grad_norm": 2.741000175476074, "learning_rate": 1.7225e-05, "loss": 0.274, "step": 690 }, { "grad_norm": 3.4030635356903076, "learning_rate": 1.7475e-05, "loss": 0.2873, "step": 700 }, { "grad_norm": 3.4968550205230713, "learning_rate": 1.7725e-05, "loss": 0.2637, "step": 710 }, { "grad_norm": 3.0175373554229736, "learning_rate": 1.7975e-05, "loss": 0.262, "step": 720 }, { "grad_norm": 2.4470367431640625, "learning_rate": 1.8225e-05, "loss": 0.2466, "step": 730 }, { "grad_norm": 1.9853612184524536, "learning_rate": 1.8475000000000002e-05, "loss": 0.2145, "step": 740 }, { "grad_norm": 3.039370059967041, "learning_rate": 1.8725e-05, "loss": 0.2437, "step": 750 }, { "grad_norm": 2.3843820095062256, "learning_rate": 1.8975e-05, "loss": 0.2376, "step": 760 }, { "grad_norm": 2.590575933456421, "learning_rate": 1.9225e-05, "loss": 0.234, "step": 770 }, { "grad_norm": 2.3592917919158936, "learning_rate": 1.9475000000000002e-05, "loss": 0.2452, "step": 780 }, { "grad_norm": 2.2335643768310547, "learning_rate": 1.9725000000000002e-05, "loss": 0.1933, "step": 790 }, { "grad_norm": 2.778855800628662, "learning_rate": 1.9975e-05, "loss": 0.1779, "step": 800 }, { "grad_norm": 3.3723740577697754, "learning_rate": 2.0225000000000004e-05, "loss": 0.1788, "step": 810 }, { "grad_norm": 2.0036020278930664, "learning_rate": 2.0475e-05, "loss": 0.1926, "step": 820 }, { "grad_norm": 2.3441011905670166, "learning_rate": 2.0725e-05, "loss": 0.2139, "step": 830 }, { "grad_norm": 2.051159620285034, "learning_rate": 2.0975e-05, "loss": 0.1923, "step": 840 }, { "grad_norm": 1.9833259582519531, "learning_rate": 2.1225e-05, "loss": 0.1962, "step": 850 }, { "grad_norm": 2.471378803253174, "learning_rate": 2.1475e-05, "loss": 0.1663, "step": 860 }, { "grad_norm": 2.4104418754577637, "learning_rate": 2.1725e-05, "loss": 0.1949, "step": 870 }, { "grad_norm": 2.8768396377563477, "learning_rate": 2.1975000000000002e-05, "loss": 0.1706, "step": 880 }, { "grad_norm": 1.7689071893692017, "learning_rate": 2.2225e-05, "loss": 0.1731, "step": 890 }, { "grad_norm": 2.609632730484009, "learning_rate": 2.2475e-05, "loss": 0.1816, "step": 900 }, { "grad_norm": 1.7236528396606445, "learning_rate": 2.2725000000000003e-05, "loss": 0.1617, "step": 910 }, { "grad_norm": 1.9648983478546143, "learning_rate": 2.2975000000000003e-05, "loss": 0.1527, "step": 920 }, { "grad_norm": 2.71356463432312, "learning_rate": 2.3225000000000002e-05, "loss": 0.1573, "step": 930 }, { "grad_norm": 2.5599687099456787, "learning_rate": 2.3475e-05, "loss": 0.1693, "step": 940 }, { "grad_norm": 2.638166904449463, "learning_rate": 2.3725e-05, "loss": 0.1251, "step": 950 }, { "grad_norm": 2.2881219387054443, "learning_rate": 2.3975e-05, "loss": 0.1587, "step": 960 }, { "grad_norm": 2.157236099243164, "learning_rate": 2.4225e-05, "loss": 0.1777, "step": 970 }, { "grad_norm": 1.9583543539047241, "learning_rate": 2.4475000000000002e-05, "loss": 0.143, "step": 980 }, { "grad_norm": 2.5867509841918945, "learning_rate": 2.4725e-05, "loss": 0.1515, "step": 990 }, { "grad_norm": 1.9369537830352783, "learning_rate": 2.4975e-05, "loss": 0.1404, "step": 1000 }, { "grad_norm": 2.068150043487549, "learning_rate": 2.5225e-05, "loss": 0.1512, "step": 1010 }, { "grad_norm": 2.238969087600708, "learning_rate": 2.5475e-05, "loss": 0.1642, "step": 1020 }, { "grad_norm": 1.9801806211471558, "learning_rate": 2.5725e-05, "loss": 0.1293, "step": 1030 }, { "grad_norm": 2.0403168201446533, "learning_rate": 2.5974999999999998e-05, "loss": 0.1307, "step": 1040 }, { "grad_norm": 2.5083158016204834, "learning_rate": 2.6225e-05, "loss": 0.156, "step": 1050 }, { "grad_norm": 2.460367202758789, "learning_rate": 2.6475e-05, "loss": 0.1616, "step": 1060 }, { "grad_norm": 2.146677255630493, "learning_rate": 2.6725e-05, "loss": 0.1678, "step": 1070 }, { "grad_norm": 2.554915428161621, "learning_rate": 2.6975000000000002e-05, "loss": 0.1705, "step": 1080 }, { "grad_norm": 2.103191614151001, "learning_rate": 2.7225e-05, "loss": 0.1546, "step": 1090 }, { "grad_norm": 1.9311569929122925, "learning_rate": 2.7475e-05, "loss": 0.1308, "step": 1100 }, { "grad_norm": 2.4710428714752197, "learning_rate": 2.7725e-05, "loss": 0.1872, "step": 1110 }, { "grad_norm": 1.7848671674728394, "learning_rate": 2.7975000000000002e-05, "loss": 0.116, "step": 1120 }, { "grad_norm": 2.365664005279541, "learning_rate": 2.8225e-05, "loss": 0.1582, "step": 1130 }, { "grad_norm": 2.596031665802002, "learning_rate": 2.8475e-05, "loss": 0.1874, "step": 1140 }, { "grad_norm": 2.203458547592163, "learning_rate": 2.8725e-05, "loss": 0.1825, "step": 1150 }, { "grad_norm": 1.4601255655288696, "learning_rate": 2.8975000000000003e-05, "loss": 0.1531, "step": 1160 }, { "grad_norm": 2.2051124572753906, "learning_rate": 2.9225000000000002e-05, "loss": 0.1453, "step": 1170 }, { "grad_norm": 1.8866993188858032, "learning_rate": 2.9475e-05, "loss": 0.1259, "step": 1180 }, { "grad_norm": 1.7644076347351074, "learning_rate": 2.9725000000000004e-05, "loss": 0.145, "step": 1190 }, { "grad_norm": 3.198972702026367, "learning_rate": 2.9975000000000004e-05, "loss": 0.1467, "step": 1200 }, { "grad_norm": 3.100167751312256, "learning_rate": 3.0225000000000003e-05, "loss": 0.1727, "step": 1210 }, { "grad_norm": 1.9414368867874146, "learning_rate": 3.0475000000000002e-05, "loss": 0.148, "step": 1220 }, { "grad_norm": 1.6191805601119995, "learning_rate": 3.0725e-05, "loss": 0.1381, "step": 1230 }, { "grad_norm": 1.8288869857788086, "learning_rate": 3.0975e-05, "loss": 0.1496, "step": 1240 }, { "grad_norm": 2.0254032611846924, "learning_rate": 3.122500000000001e-05, "loss": 0.1289, "step": 1250 }, { "grad_norm": 2.0874152183532715, "learning_rate": 3.1475e-05, "loss": 0.1595, "step": 1260 }, { "grad_norm": 2.0687036514282227, "learning_rate": 3.1725e-05, "loss": 0.1252, "step": 1270 }, { "grad_norm": 1.422391653060913, "learning_rate": 3.1975e-05, "loss": 0.1501, "step": 1280 }, { "grad_norm": 1.1003530025482178, "learning_rate": 3.2225e-05, "loss": 0.131, "step": 1290 }, { "grad_norm": 1.8686761856079102, "learning_rate": 3.2474999999999997e-05, "loss": 0.1384, "step": 1300 }, { "grad_norm": 1.7751697301864624, "learning_rate": 3.2725e-05, "loss": 0.1293, "step": 1310 }, { "grad_norm": 2.0422396659851074, "learning_rate": 3.2975e-05, "loss": 0.1306, "step": 1320 }, { "grad_norm": 1.7604106664657593, "learning_rate": 3.3225e-05, "loss": 0.1218, "step": 1330 }, { "grad_norm": 1.4723535776138306, "learning_rate": 3.3475e-05, "loss": 0.1233, "step": 1340 }, { "grad_norm": 1.8281171321868896, "learning_rate": 3.3725e-05, "loss": 0.156, "step": 1350 }, { "grad_norm": 2.0118885040283203, "learning_rate": 3.3975e-05, "loss": 0.142, "step": 1360 }, { "grad_norm": 1.9674983024597168, "learning_rate": 3.4225e-05, "loss": 0.1265, "step": 1370 }, { "grad_norm": 2.0086653232574463, "learning_rate": 3.4475000000000005e-05, "loss": 0.1272, "step": 1380 }, { "grad_norm": 2.197181463241577, "learning_rate": 3.4725000000000004e-05, "loss": 0.1297, "step": 1390 }, { "grad_norm": 1.5416083335876465, "learning_rate": 3.4975e-05, "loss": 0.1254, "step": 1400 }, { "grad_norm": 1.4232113361358643, "learning_rate": 3.5225e-05, "loss": 0.1369, "step": 1410 }, { "grad_norm": 1.8009371757507324, "learning_rate": 3.5475e-05, "loss": 0.1302, "step": 1420 }, { "grad_norm": 1.257946491241455, "learning_rate": 3.5725e-05, "loss": 0.1033, "step": 1430 }, { "grad_norm": 1.4271154403686523, "learning_rate": 3.5975e-05, "loss": 0.1343, "step": 1440 }, { "grad_norm": 1.7395323514938354, "learning_rate": 3.6225000000000006e-05, "loss": 0.1416, "step": 1450 }, { "grad_norm": 1.6901028156280518, "learning_rate": 3.6475000000000006e-05, "loss": 0.1487, "step": 1460 }, { "grad_norm": 1.5812478065490723, "learning_rate": 3.6725000000000005e-05, "loss": 0.1454, "step": 1470 }, { "grad_norm": 1.5004240274429321, "learning_rate": 3.6975000000000004e-05, "loss": 0.1263, "step": 1480 }, { "grad_norm": 1.3130886554718018, "learning_rate": 3.7225000000000004e-05, "loss": 0.1282, "step": 1490 }, { "grad_norm": 1.1710255146026611, "learning_rate": 3.7475e-05, "loss": 0.1258, "step": 1500 }, { "grad_norm": 1.621403455734253, "learning_rate": 3.7725e-05, "loss": 0.1184, "step": 1510 }, { "grad_norm": 2.1786909103393555, "learning_rate": 3.7975e-05, "loss": 0.143, "step": 1520 }, { "grad_norm": 2.222222089767456, "learning_rate": 3.8225e-05, "loss": 0.1459, "step": 1530 }, { "grad_norm": 1.4123997688293457, "learning_rate": 3.8475e-05, "loss": 0.1382, "step": 1540 }, { "grad_norm": 1.1879957914352417, "learning_rate": 3.8725e-05, "loss": 0.1127, "step": 1550 }, { "grad_norm": 1.2847527265548706, "learning_rate": 3.8975e-05, "loss": 0.1344, "step": 1560 }, { "grad_norm": 1.6257960796356201, "learning_rate": 3.9225e-05, "loss": 0.1062, "step": 1570 }, { "grad_norm": 1.727474331855774, "learning_rate": 3.9475000000000004e-05, "loss": 0.1269, "step": 1580 }, { "grad_norm": 1.4727996587753296, "learning_rate": 3.9725e-05, "loss": 0.1271, "step": 1590 }, { "grad_norm": 1.396260380744934, "learning_rate": 3.9975e-05, "loss": 0.1165, "step": 1600 }, { "grad_norm": 1.36361825466156, "learning_rate": 4.0225e-05, "loss": 0.1132, "step": 1610 }, { "grad_norm": 1.3595613241195679, "learning_rate": 4.0475e-05, "loss": 0.1339, "step": 1620 }, { "grad_norm": 1.3133952617645264, "learning_rate": 4.0725e-05, "loss": 0.151, "step": 1630 }, { "grad_norm": 1.2549607753753662, "learning_rate": 4.0975e-05, "loss": 0.1313, "step": 1640 }, { "grad_norm": 1.5340842008590698, "learning_rate": 4.1225e-05, "loss": 0.1277, "step": 1650 }, { "grad_norm": 1.7942551374435425, "learning_rate": 4.1475000000000005e-05, "loss": 0.1343, "step": 1660 }, { "grad_norm": 1.198857307434082, "learning_rate": 4.1725000000000005e-05, "loss": 0.1139, "step": 1670 }, { "grad_norm": 1.5236679315567017, "learning_rate": 4.1975000000000004e-05, "loss": 0.1131, "step": 1680 }, { "grad_norm": 1.7947601079940796, "learning_rate": 4.2225e-05, "loss": 0.1212, "step": 1690 }, { "grad_norm": 1.2278493642807007, "learning_rate": 4.2475e-05, "loss": 0.124, "step": 1700 }, { "grad_norm": 1.84466552734375, "learning_rate": 4.2725e-05, "loss": 0.1568, "step": 1710 }, { "grad_norm": 1.135420799255371, "learning_rate": 4.2975e-05, "loss": 0.1241, "step": 1720 }, { "grad_norm": 1.3684732913970947, "learning_rate": 4.322500000000001e-05, "loss": 0.1183, "step": 1730 }, { "grad_norm": 1.390057921409607, "learning_rate": 4.3475000000000006e-05, "loss": 0.125, "step": 1740 }, { "grad_norm": 1.5323642492294312, "learning_rate": 4.3725000000000006e-05, "loss": 0.1376, "step": 1750 }, { "grad_norm": 1.8069888353347778, "learning_rate": 4.3975e-05, "loss": 0.1291, "step": 1760 }, { "grad_norm": 1.233368992805481, "learning_rate": 4.4225e-05, "loss": 0.1041, "step": 1770 }, { "grad_norm": 1.24197256565094, "learning_rate": 4.4475e-05, "loss": 0.1367, "step": 1780 }, { "grad_norm": 1.5370229482650757, "learning_rate": 4.4725e-05, "loss": 0.1055, "step": 1790 }, { "grad_norm": 1.046623945236206, "learning_rate": 4.4975e-05, "loss": 0.119, "step": 1800 }, { "grad_norm": 1.1560121774673462, "learning_rate": 4.5225e-05, "loss": 0.1104, "step": 1810 }, { "grad_norm": 1.6299195289611816, "learning_rate": 4.5475e-05, "loss": 0.1222, "step": 1820 }, { "grad_norm": 1.565991997718811, "learning_rate": 4.5725e-05, "loss": 0.1249, "step": 1830 }, { "grad_norm": 1.4722057580947876, "learning_rate": 4.5975e-05, "loss": 0.1371, "step": 1840 }, { "grad_norm": 1.048601508140564, "learning_rate": 4.6225e-05, "loss": 0.1487, "step": 1850 }, { "grad_norm": 1.2002617120742798, "learning_rate": 4.6475000000000005e-05, "loss": 0.145, "step": 1860 }, { "grad_norm": 1.0299510955810547, "learning_rate": 4.6725000000000004e-05, "loss": 0.128, "step": 1870 }, { "grad_norm": 1.1071223020553589, "learning_rate": 4.6975000000000003e-05, "loss": 0.098, "step": 1880 }, { "grad_norm": 1.2915793657302856, "learning_rate": 4.7225e-05, "loss": 0.1245, "step": 1890 }, { "grad_norm": 1.172791600227356, "learning_rate": 4.7475e-05, "loss": 0.1121, "step": 1900 }, { "grad_norm": 1.426775574684143, "learning_rate": 4.7725e-05, "loss": 0.1447, "step": 1910 }, { "grad_norm": 1.2364236116409302, "learning_rate": 4.7975e-05, "loss": 0.1194, "step": 1920 }, { "grad_norm": 1.0580295324325562, "learning_rate": 4.822500000000001e-05, "loss": 0.1033, "step": 1930 }, { "grad_norm": 1.4343231916427612, "learning_rate": 4.8475000000000006e-05, "loss": 0.1143, "step": 1940 }, { "grad_norm": 0.9685896635055542, "learning_rate": 4.8725000000000005e-05, "loss": 0.096, "step": 1950 }, { "grad_norm": 1.5057989358901978, "learning_rate": 4.8975000000000005e-05, "loss": 0.1398, "step": 1960 }, { "grad_norm": 0.8895019888877869, "learning_rate": 4.9225000000000004e-05, "loss": 0.0923, "step": 1970 }, { "grad_norm": 1.7174408435821533, "learning_rate": 4.9475e-05, "loss": 0.1255, "step": 1980 }, { "grad_norm": 1.3461326360702515, "learning_rate": 4.9725e-05, "loss": 0.1048, "step": 1990 }, { "grad_norm": 1.3915812969207764, "learning_rate": 4.9975e-05, "loss": 0.1246, "step": 2000 }, { "grad_norm": 1.15657639503479, "learning_rate": 5.0225e-05, "loss": 0.1359, "step": 2010 }, { "grad_norm": 1.4956872463226318, "learning_rate": 5.047500000000001e-05, "loss": 0.1228, "step": 2020 }, { "grad_norm": 1.2760807275772095, "learning_rate": 5.0725e-05, "loss": 0.1003, "step": 2030 }, { "grad_norm": 1.0020864009857178, "learning_rate": 5.0975000000000006e-05, "loss": 0.1195, "step": 2040 }, { "grad_norm": 1.041778326034546, "learning_rate": 5.1225e-05, "loss": 0.1169, "step": 2050 }, { "grad_norm": 1.1370898485183716, "learning_rate": 5.1475000000000004e-05, "loss": 0.1062, "step": 2060 }, { "grad_norm": 1.471795678138733, "learning_rate": 5.1725000000000004e-05, "loss": 0.1173, "step": 2070 }, { "grad_norm": 1.3704320192337036, "learning_rate": 5.197500000000001e-05, "loss": 0.1108, "step": 2080 }, { "grad_norm": 1.1045684814453125, "learning_rate": 5.2225e-05, "loss": 0.1225, "step": 2090 }, { "grad_norm": 0.7435891628265381, "learning_rate": 5.247500000000001e-05, "loss": 0.1073, "step": 2100 }, { "grad_norm": 1.082478642463684, "learning_rate": 5.2725e-05, "loss": 0.1409, "step": 2110 }, { "grad_norm": 1.001111388206482, "learning_rate": 5.297500000000001e-05, "loss": 0.1057, "step": 2120 }, { "grad_norm": 0.8942968249320984, "learning_rate": 5.3225e-05, "loss": 0.1094, "step": 2130 }, { "grad_norm": 1.2505449056625366, "learning_rate": 5.3475e-05, "loss": 0.0954, "step": 2140 }, { "grad_norm": 1.0819755792617798, "learning_rate": 5.3725000000000005e-05, "loss": 0.1299, "step": 2150 }, { "grad_norm": 0.9901742935180664, "learning_rate": 5.3975e-05, "loss": 0.1078, "step": 2160 }, { "grad_norm": 0.9293856024742126, "learning_rate": 5.4225000000000003e-05, "loss": 0.1116, "step": 2170 }, { "grad_norm": 0.9618505835533142, "learning_rate": 5.4474999999999996e-05, "loss": 0.1318, "step": 2180 }, { "grad_norm": 1.0832659006118774, "learning_rate": 5.4725e-05, "loss": 0.1023, "step": 2190 }, { "grad_norm": 0.7375123500823975, "learning_rate": 5.4975e-05, "loss": 0.0968, "step": 2200 }, { "grad_norm": 0.8558865785598755, "learning_rate": 5.522500000000001e-05, "loss": 0.0926, "step": 2210 }, { "grad_norm": 1.113889217376709, "learning_rate": 5.5475e-05, "loss": 0.0939, "step": 2220 }, { "grad_norm": 1.0418509244918823, "learning_rate": 5.5725000000000006e-05, "loss": 0.1124, "step": 2230 }, { "grad_norm": 1.107362985610962, "learning_rate": 5.5975e-05, "loss": 0.098, "step": 2240 }, { "grad_norm": 0.9225648045539856, "learning_rate": 5.6225000000000005e-05, "loss": 0.1234, "step": 2250 }, { "grad_norm": 0.9637777805328369, "learning_rate": 5.6475e-05, "loss": 0.095, "step": 2260 }, { "grad_norm": 0.9307835698127747, "learning_rate": 5.6725e-05, "loss": 0.1047, "step": 2270 }, { "grad_norm": 1.0596357583999634, "learning_rate": 5.6975e-05, "loss": 0.104, "step": 2280 }, { "grad_norm": 1.284986138343811, "learning_rate": 5.722500000000001e-05, "loss": 0.1447, "step": 2290 }, { "grad_norm": 1.1940913200378418, "learning_rate": 5.7475e-05, "loss": 0.11, "step": 2300 }, { "grad_norm": 0.7232175469398499, "learning_rate": 5.772500000000001e-05, "loss": 0.1078, "step": 2310 }, { "grad_norm": 1.0041872262954712, "learning_rate": 5.7975e-05, "loss": 0.096, "step": 2320 }, { "grad_norm": 0.9682076573371887, "learning_rate": 5.8225000000000006e-05, "loss": 0.1113, "step": 2330 }, { "grad_norm": 0.8083800077438354, "learning_rate": 5.8475000000000005e-05, "loss": 0.1047, "step": 2340 }, { "grad_norm": 0.8743035197257996, "learning_rate": 5.8725000000000004e-05, "loss": 0.1147, "step": 2350 }, { "grad_norm": 0.9988933801651001, "learning_rate": 5.8975000000000004e-05, "loss": 0.1119, "step": 2360 }, { "grad_norm": 1.1235376596450806, "learning_rate": 5.922500000000001e-05, "loss": 0.1253, "step": 2370 }, { "grad_norm": 1.1392112970352173, "learning_rate": 5.9475e-05, "loss": 0.1237, "step": 2380 }, { "grad_norm": 0.7725063562393188, "learning_rate": 5.9724999999999995e-05, "loss": 0.127, "step": 2390 }, { "grad_norm": 0.9502177834510803, "learning_rate": 5.9975e-05, "loss": 0.1038, "step": 2400 }, { "grad_norm": 0.8516688942909241, "learning_rate": 6.0225e-05, "loss": 0.0976, "step": 2410 }, { "grad_norm": 1.020812749862671, "learning_rate": 6.0475000000000006e-05, "loss": 0.1385, "step": 2420 }, { "grad_norm": 1.0618716478347778, "learning_rate": 6.0725e-05, "loss": 0.1148, "step": 2430 }, { "grad_norm": 1.14437997341156, "learning_rate": 6.0975000000000005e-05, "loss": 0.1157, "step": 2440 }, { "grad_norm": 1.03252112865448, "learning_rate": 6.1225e-05, "loss": 0.1115, "step": 2450 }, { "grad_norm": 0.8213791847229004, "learning_rate": 6.1475e-05, "loss": 0.0984, "step": 2460 }, { "grad_norm": 1.0797209739685059, "learning_rate": 6.1725e-05, "loss": 0.103, "step": 2470 }, { "grad_norm": 0.766682505607605, "learning_rate": 6.1975e-05, "loss": 0.1167, "step": 2480 }, { "grad_norm": 0.9603060483932495, "learning_rate": 6.2225e-05, "loss": 0.1028, "step": 2490 }, { "grad_norm": 0.7363983392715454, "learning_rate": 6.2475e-05, "loss": 0.0899, "step": 2500 }, { "grad_norm": 0.7997924089431763, "learning_rate": 6.2725e-05, "loss": 0.0928, "step": 2510 }, { "grad_norm": 0.772529125213623, "learning_rate": 6.297500000000001e-05, "loss": 0.1022, "step": 2520 }, { "grad_norm": 0.6824080944061279, "learning_rate": 6.3225e-05, "loss": 0.087, "step": 2530 }, { "grad_norm": 0.6993765830993652, "learning_rate": 6.347500000000001e-05, "loss": 0.0925, "step": 2540 }, { "grad_norm": 0.4831967353820801, "learning_rate": 6.3725e-05, "loss": 0.1037, "step": 2550 }, { "grad_norm": 0.7605200409889221, "learning_rate": 6.397500000000001e-05, "loss": 0.1062, "step": 2560 }, { "grad_norm": 0.8606461882591248, "learning_rate": 6.4225e-05, "loss": 0.0949, "step": 2570 }, { "grad_norm": 0.9823527932167053, "learning_rate": 6.447500000000001e-05, "loss": 0.0975, "step": 2580 }, { "grad_norm": 0.768942654132843, "learning_rate": 6.4725e-05, "loss": 0.0982, "step": 2590 }, { "grad_norm": 1.0189505815505981, "learning_rate": 6.497500000000001e-05, "loss": 0.1139, "step": 2600 }, { "grad_norm": 0.8324097394943237, "learning_rate": 6.5225e-05, "loss": 0.1042, "step": 2610 }, { "grad_norm": 0.8145201206207275, "learning_rate": 6.5475e-05, "loss": 0.093, "step": 2620 }, { "grad_norm": 0.7576887011528015, "learning_rate": 6.5725e-05, "loss": 0.11, "step": 2630 }, { "grad_norm": 0.8778746128082275, "learning_rate": 6.5975e-05, "loss": 0.1127, "step": 2640 }, { "grad_norm": 1.0284696817398071, "learning_rate": 6.6225e-05, "loss": 0.1166, "step": 2650 }, { "grad_norm": 0.9593722224235535, "learning_rate": 6.6475e-05, "loss": 0.1106, "step": 2660 }, { "grad_norm": 0.9299062490463257, "learning_rate": 6.672500000000001e-05, "loss": 0.1202, "step": 2670 }, { "grad_norm": 1.1876518726348877, "learning_rate": 6.6975e-05, "loss": 0.1234, "step": 2680 }, { "grad_norm": 0.6658750176429749, "learning_rate": 6.722500000000001e-05, "loss": 0.1104, "step": 2690 }, { "grad_norm": 1.01437246799469, "learning_rate": 6.7475e-05, "loss": 0.0922, "step": 2700 }, { "grad_norm": 0.6734445691108704, "learning_rate": 6.7725e-05, "loss": 0.1019, "step": 2710 }, { "grad_norm": 0.7645560503005981, "learning_rate": 6.7975e-05, "loss": 0.1216, "step": 2720 }, { "grad_norm": 0.8959755897521973, "learning_rate": 6.8225e-05, "loss": 0.1022, "step": 2730 }, { "grad_norm": 0.7587870359420776, "learning_rate": 6.8475e-05, "loss": 0.1071, "step": 2740 }, { "grad_norm": 0.9375230073928833, "learning_rate": 6.8725e-05, "loss": 0.1139, "step": 2750 }, { "grad_norm": 1.0871840715408325, "learning_rate": 6.8975e-05, "loss": 0.0978, "step": 2760 }, { "grad_norm": 0.8946226835250854, "learning_rate": 6.9225e-05, "loss": 0.1017, "step": 2770 }, { "grad_norm": 0.8101896047592163, "learning_rate": 6.9475e-05, "loss": 0.1131, "step": 2780 }, { "grad_norm": 0.954254686832428, "learning_rate": 6.9725e-05, "loss": 0.1105, "step": 2790 }, { "grad_norm": 0.8505905866622925, "learning_rate": 6.997500000000001e-05, "loss": 0.1117, "step": 2800 }, { "grad_norm": 0.7953459620475769, "learning_rate": 7.022500000000001e-05, "loss": 0.1356, "step": 2810 }, { "grad_norm": 1.0297133922576904, "learning_rate": 7.0475e-05, "loss": 0.1082, "step": 2820 }, { "grad_norm": 0.7251672148704529, "learning_rate": 7.072500000000001e-05, "loss": 0.0867, "step": 2830 }, { "grad_norm": 0.7616082429885864, "learning_rate": 7.0975e-05, "loss": 0.1208, "step": 2840 }, { "grad_norm": 0.6713007092475891, "learning_rate": 7.122500000000001e-05, "loss": 0.1251, "step": 2850 }, { "grad_norm": 0.8329033255577087, "learning_rate": 7.1475e-05, "loss": 0.1012, "step": 2860 }, { "grad_norm": 0.7222850918769836, "learning_rate": 7.172500000000001e-05, "loss": 0.0987, "step": 2870 }, { "grad_norm": 0.7164271473884583, "learning_rate": 7.1975e-05, "loss": 0.0843, "step": 2880 }, { "grad_norm": 0.8076629638671875, "learning_rate": 7.2225e-05, "loss": 0.1075, "step": 2890 }, { "grad_norm": 0.6591405272483826, "learning_rate": 7.2475e-05, "loss": 0.0961, "step": 2900 }, { "grad_norm": 0.9765083193778992, "learning_rate": 7.272499999999999e-05, "loss": 0.0927, "step": 2910 }, { "grad_norm": 0.6548402905464172, "learning_rate": 7.2975e-05, "loss": 0.1017, "step": 2920 }, { "grad_norm": 0.9387034773826599, "learning_rate": 7.3225e-05, "loss": 0.0839, "step": 2930 }, { "grad_norm": 0.7274496555328369, "learning_rate": 7.347500000000001e-05, "loss": 0.0822, "step": 2940 }, { "grad_norm": 0.7471102476119995, "learning_rate": 7.3725e-05, "loss": 0.0831, "step": 2950 }, { "grad_norm": 0.7942771315574646, "learning_rate": 7.397500000000001e-05, "loss": 0.0945, "step": 2960 }, { "grad_norm": 0.8196765184402466, "learning_rate": 7.4225e-05, "loss": 0.09, "step": 2970 }, { "grad_norm": 0.8125753998756409, "learning_rate": 7.447500000000001e-05, "loss": 0.097, "step": 2980 }, { "grad_norm": 0.9137815237045288, "learning_rate": 7.4725e-05, "loss": 0.1064, "step": 2990 }, { "grad_norm": 0.7821735739707947, "learning_rate": 7.4975e-05, "loss": 0.1072, "step": 3000 }, { "grad_norm": 0.7150062322616577, "learning_rate": 7.5225e-05, "loss": 0.0955, "step": 3010 }, { "grad_norm": 0.92818683385849, "learning_rate": 7.5475e-05, "loss": 0.0863, "step": 3020 }, { "grad_norm": 0.7029780149459839, "learning_rate": 7.5725e-05, "loss": 0.0915, "step": 3030 }, { "grad_norm": 0.7915005683898926, "learning_rate": 7.5975e-05, "loss": 0.0916, "step": 3040 }, { "grad_norm": 1.1385353803634644, "learning_rate": 7.6225e-05, "loss": 0.1071, "step": 3050 }, { "grad_norm": 0.7713677287101746, "learning_rate": 7.6475e-05, "loss": 0.0973, "step": 3060 }, { "grad_norm": 0.962304413318634, "learning_rate": 7.672500000000001e-05, "loss": 0.0876, "step": 3070 }, { "grad_norm": 0.764313817024231, "learning_rate": 7.697500000000001e-05, "loss": 0.1039, "step": 3080 }, { "grad_norm": 0.9552737474441528, "learning_rate": 7.722500000000001e-05, "loss": 0.1166, "step": 3090 }, { "grad_norm": 1.076857566833496, "learning_rate": 7.747500000000001e-05, "loss": 0.1015, "step": 3100 }, { "grad_norm": 0.7948228716850281, "learning_rate": 7.7725e-05, "loss": 0.0895, "step": 3110 }, { "grad_norm": 0.6591798663139343, "learning_rate": 7.797500000000001e-05, "loss": 0.0836, "step": 3120 }, { "grad_norm": 0.7237152457237244, "learning_rate": 7.8225e-05, "loss": 0.0855, "step": 3130 }, { "grad_norm": 0.7819040417671204, "learning_rate": 7.8475e-05, "loss": 0.101, "step": 3140 }, { "grad_norm": 0.6974128484725952, "learning_rate": 7.8725e-05, "loss": 0.089, "step": 3150 }, { "grad_norm": 0.7526759505271912, "learning_rate": 7.8975e-05, "loss": 0.1023, "step": 3160 }, { "grad_norm": 0.7242497205734253, "learning_rate": 7.9225e-05, "loss": 0.1019, "step": 3170 }, { "grad_norm": 0.855621874332428, "learning_rate": 7.9475e-05, "loss": 0.1032, "step": 3180 }, { "grad_norm": 0.7770329713821411, "learning_rate": 7.9725e-05, "loss": 0.1121, "step": 3190 }, { "grad_norm": 0.6193979978561401, "learning_rate": 7.9975e-05, "loss": 0.0893, "step": 3200 }, { "grad_norm": 0.6234460473060608, "learning_rate": 8.022500000000001e-05, "loss": 0.0934, "step": 3210 }, { "grad_norm": 1.0221447944641113, "learning_rate": 8.0475e-05, "loss": 0.1261, "step": 3220 }, { "grad_norm": 0.8795571327209473, "learning_rate": 8.072500000000001e-05, "loss": 0.1061, "step": 3230 }, { "grad_norm": 0.6572255492210388, "learning_rate": 8.0975e-05, "loss": 0.0924, "step": 3240 }, { "grad_norm": 0.6863035559654236, "learning_rate": 8.122500000000001e-05, "loss": 0.0982, "step": 3250 }, { "grad_norm": 1.1653310060501099, "learning_rate": 8.1475e-05, "loss": 0.1025, "step": 3260 }, { "grad_norm": 0.876767098903656, "learning_rate": 8.172500000000001e-05, "loss": 0.1025, "step": 3270 }, { "grad_norm": 0.76087886095047, "learning_rate": 8.1975e-05, "loss": 0.0887, "step": 3280 }, { "grad_norm": 0.7436174154281616, "learning_rate": 8.2225e-05, "loss": 0.0902, "step": 3290 }, { "grad_norm": 0.6285713315010071, "learning_rate": 8.2475e-05, "loss": 0.0915, "step": 3300 }, { "grad_norm": 0.8374627828598022, "learning_rate": 8.2725e-05, "loss": 0.1112, "step": 3310 }, { "grad_norm": 0.5527814030647278, "learning_rate": 8.2975e-05, "loss": 0.115, "step": 3320 }, { "grad_norm": 0.9074457287788391, "learning_rate": 8.3225e-05, "loss": 0.1092, "step": 3330 }, { "grad_norm": 0.9206497669219971, "learning_rate": 8.347500000000001e-05, "loss": 0.1015, "step": 3340 }, { "grad_norm": 0.8723562955856323, "learning_rate": 8.3725e-05, "loss": 0.1092, "step": 3350 }, { "grad_norm": 0.9136738181114197, "learning_rate": 8.397500000000001e-05, "loss": 0.0982, "step": 3360 }, { "grad_norm": 0.7477405071258545, "learning_rate": 8.422500000000001e-05, "loss": 0.108, "step": 3370 }, { "grad_norm": 0.746319055557251, "learning_rate": 8.447500000000001e-05, "loss": 0.1468, "step": 3380 }, { "grad_norm": 0.5257405638694763, "learning_rate": 8.4725e-05, "loss": 0.0962, "step": 3390 }, { "grad_norm": 0.9091600775718689, "learning_rate": 8.4975e-05, "loss": 0.1022, "step": 3400 }, { "grad_norm": 0.8260389566421509, "learning_rate": 8.5225e-05, "loss": 0.1037, "step": 3410 }, { "grad_norm": 0.8412608504295349, "learning_rate": 8.5475e-05, "loss": 0.0847, "step": 3420 }, { "grad_norm": 0.96625155210495, "learning_rate": 8.5725e-05, "loss": 0.0996, "step": 3430 }, { "grad_norm": 0.6844078302383423, "learning_rate": 8.5975e-05, "loss": 0.1156, "step": 3440 }, { "grad_norm": 0.7152140140533447, "learning_rate": 8.6225e-05, "loss": 0.0978, "step": 3450 }, { "grad_norm": 0.684718668460846, "learning_rate": 8.6475e-05, "loss": 0.0863, "step": 3460 }, { "grad_norm": 0.6613653302192688, "learning_rate": 8.672500000000001e-05, "loss": 0.1095, "step": 3470 }, { "grad_norm": 0.7316390872001648, "learning_rate": 8.6975e-05, "loss": 0.094, "step": 3480 }, { "grad_norm": 0.8396455645561218, "learning_rate": 8.7225e-05, "loss": 0.0955, "step": 3490 }, { "grad_norm": 0.8394449949264526, "learning_rate": 8.747500000000001e-05, "loss": 0.0971, "step": 3500 }, { "grad_norm": 0.8634128570556641, "learning_rate": 8.7725e-05, "loss": 0.1146, "step": 3510 }, { "grad_norm": 0.7947315573692322, "learning_rate": 8.797500000000001e-05, "loss": 0.0931, "step": 3520 }, { "grad_norm": 0.8263389468193054, "learning_rate": 8.8225e-05, "loss": 0.1063, "step": 3530 }, { "grad_norm": 0.6612719297409058, "learning_rate": 8.847500000000001e-05, "loss": 0.091, "step": 3540 }, { "grad_norm": 0.7165521383285522, "learning_rate": 8.8725e-05, "loss": 0.0933, "step": 3550 }, { "grad_norm": 0.530287504196167, "learning_rate": 8.897500000000001e-05, "loss": 0.0777, "step": 3560 }, { "grad_norm": 0.5415716767311096, "learning_rate": 8.9225e-05, "loss": 0.1014, "step": 3570 }, { "grad_norm": 0.999649703502655, "learning_rate": 8.9475e-05, "loss": 0.1094, "step": 3580 }, { "grad_norm": 0.7387140393257141, "learning_rate": 8.9725e-05, "loss": 0.1085, "step": 3590 }, { "grad_norm": 0.6996066570281982, "learning_rate": 8.9975e-05, "loss": 0.1023, "step": 3600 }, { "grad_norm": 0.6784666776657104, "learning_rate": 9.0225e-05, "loss": 0.0957, "step": 3610 }, { "grad_norm": 0.5687133073806763, "learning_rate": 9.0475e-05, "loss": 0.0946, "step": 3620 }, { "grad_norm": 0.6101356744766235, "learning_rate": 9.072500000000001e-05, "loss": 0.0815, "step": 3630 }, { "grad_norm": 0.7574860453605652, "learning_rate": 9.0975e-05, "loss": 0.0939, "step": 3640 }, { "grad_norm": 0.8618793487548828, "learning_rate": 9.122500000000001e-05, "loss": 0.0978, "step": 3650 }, { "grad_norm": 0.9166448712348938, "learning_rate": 9.1475e-05, "loss": 0.0922, "step": 3660 }, { "grad_norm": 0.6694897413253784, "learning_rate": 9.172500000000001e-05, "loss": 0.0944, "step": 3670 }, { "grad_norm": 0.7160492539405823, "learning_rate": 9.1975e-05, "loss": 0.0984, "step": 3680 }, { "grad_norm": 0.7927610278129578, "learning_rate": 9.2225e-05, "loss": 0.0775, "step": 3690 }, { "grad_norm": 0.681839108467102, "learning_rate": 9.2475e-05, "loss": 0.1002, "step": 3700 }, { "grad_norm": 0.8444750308990479, "learning_rate": 9.2725e-05, "loss": 0.0968, "step": 3710 }, { "grad_norm": 0.8702434301376343, "learning_rate": 9.2975e-05, "loss": 0.0985, "step": 3720 }, { "grad_norm": 0.6914610266685486, "learning_rate": 9.3225e-05, "loss": 0.0824, "step": 3730 }, { "grad_norm": 0.5511742830276489, "learning_rate": 9.3475e-05, "loss": 0.0805, "step": 3740 }, { "grad_norm": 0.7843304872512817, "learning_rate": 9.3725e-05, "loss": 0.0888, "step": 3750 }, { "grad_norm": 0.5712798833847046, "learning_rate": 9.397500000000001e-05, "loss": 0.07, "step": 3760 }, { "grad_norm": 0.744808554649353, "learning_rate": 9.422500000000001e-05, "loss": 0.0969, "step": 3770 }, { "grad_norm": 0.5994338989257812, "learning_rate": 9.4475e-05, "loss": 0.0974, "step": 3780 }, { "grad_norm": 0.697416365146637, "learning_rate": 9.472500000000001e-05, "loss": 0.0843, "step": 3790 }, { "grad_norm": 0.8094901442527771, "learning_rate": 9.4975e-05, "loss": 0.1138, "step": 3800 }, { "grad_norm": 0.9196756482124329, "learning_rate": 9.522500000000001e-05, "loss": 0.0763, "step": 3810 }, { "grad_norm": 0.9631965160369873, "learning_rate": 9.5475e-05, "loss": 0.1212, "step": 3820 }, { "grad_norm": 0.6915984153747559, "learning_rate": 9.572500000000001e-05, "loss": 0.0885, "step": 3830 }, { "grad_norm": 0.6547321081161499, "learning_rate": 9.5975e-05, "loss": 0.0783, "step": 3840 }, { "grad_norm": 0.5786947011947632, "learning_rate": 9.622500000000001e-05, "loss": 0.0949, "step": 3850 }, { "grad_norm": 0.9468570947647095, "learning_rate": 9.6475e-05, "loss": 0.0832, "step": 3860 }, { "grad_norm": 0.7664632201194763, "learning_rate": 9.6725e-05, "loss": 0.0869, "step": 3870 }, { "grad_norm": 0.5306453108787537, "learning_rate": 9.6975e-05, "loss": 0.0761, "step": 3880 }, { "grad_norm": 0.8288766741752625, "learning_rate": 9.7225e-05, "loss": 0.1043, "step": 3890 }, { "grad_norm": 0.6902810335159302, "learning_rate": 9.747500000000001e-05, "loss": 0.0967, "step": 3900 }, { "grad_norm": 0.7254154086112976, "learning_rate": 9.7725e-05, "loss": 0.0903, "step": 3910 }, { "grad_norm": 0.7697563171386719, "learning_rate": 9.797500000000001e-05, "loss": 0.108, "step": 3920 }, { "grad_norm": 0.6415933966636658, "learning_rate": 9.8225e-05, "loss": 0.0788, "step": 3930 }, { "grad_norm": 0.5344766974449158, "learning_rate": 9.847500000000001e-05, "loss": 0.1031, "step": 3940 }, { "grad_norm": 0.7106142640113831, "learning_rate": 9.8725e-05, "loss": 0.098, "step": 3950 }, { "grad_norm": 0.6740069389343262, "learning_rate": 9.897500000000001e-05, "loss": 0.1067, "step": 3960 }, { "grad_norm": 0.8095660209655762, "learning_rate": 9.9225e-05, "loss": 0.0865, "step": 3970 }, { "grad_norm": 0.6464433073997498, "learning_rate": 9.9475e-05, "loss": 0.108, "step": 3980 }, { "grad_norm": 0.8566124439239502, "learning_rate": 9.9725e-05, "loss": 0.1091, "step": 3990 }, { "grad_norm": 0.962486743927002, "learning_rate": 9.9975e-05, "loss": 0.0905, "step": 4000 }, { "grad_norm": 0.7311718463897705, "learning_rate": 9.999999653982884e-05, "loss": 0.0859, "step": 4010 }, { "grad_norm": 0.4900173246860504, "learning_rate": 9.999998457874392e-05, "loss": 0.0936, "step": 4020 }, { "grad_norm": 0.772788405418396, "learning_rate": 9.999996407402913e-05, "loss": 0.1, "step": 4030 }, { "grad_norm": 0.6067790985107422, "learning_rate": 9.999993502568801e-05, "loss": 0.0751, "step": 4040 }, { "grad_norm": 0.8188025951385498, "learning_rate": 9.999989743372548e-05, "loss": 0.0948, "step": 4050 }, { "grad_norm": 0.5144171714782715, "learning_rate": 9.999985129814798e-05, "loss": 0.0906, "step": 4060 }, { "grad_norm": 0.6118952631950378, "learning_rate": 9.99997966189634e-05, "loss": 0.0875, "step": 4070 }, { "grad_norm": 0.6522000432014465, "learning_rate": 9.999973339618107e-05, "loss": 0.0898, "step": 4080 }, { "grad_norm": 0.6205890774726868, "learning_rate": 9.999966162981179e-05, "loss": 0.0994, "step": 4090 }, { "grad_norm": 0.6432575583457947, "learning_rate": 9.999958131986784e-05, "loss": 0.0852, "step": 4100 }, { "grad_norm": 0.7834622859954834, "learning_rate": 9.999949246636293e-05, "loss": 0.0924, "step": 4110 }, { "grad_norm": 0.9317238926887512, "learning_rate": 9.999939506931224e-05, "loss": 0.108, "step": 4120 }, { "grad_norm": 0.8515656590461731, "learning_rate": 9.999928912873243e-05, "loss": 0.1054, "step": 4130 }, { "grad_norm": 0.6621122360229492, "learning_rate": 9.999917464464159e-05, "loss": 0.1037, "step": 4140 }, { "grad_norm": 0.6603320837020874, "learning_rate": 9.999905161705929e-05, "loss": 0.1115, "step": 4150 }, { "grad_norm": 0.8225124478340149, "learning_rate": 9.999892004600653e-05, "loss": 0.0831, "step": 4160 }, { "grad_norm": 0.6348512768745422, "learning_rate": 9.999877993150581e-05, "loss": 0.0783, "step": 4170 }, { "grad_norm": 0.6004526019096375, "learning_rate": 9.999863127358108e-05, "loss": 0.0801, "step": 4180 }, { "grad_norm": 0.5434044599533081, "learning_rate": 9.999847407225773e-05, "loss": 0.0825, "step": 4190 }, { "grad_norm": 0.6728059649467468, "learning_rate": 9.999830832756262e-05, "loss": 0.0826, "step": 4200 }, { "grad_norm": 0.5804798007011414, "learning_rate": 9.999813403952407e-05, "loss": 0.0937, "step": 4210 }, { "grad_norm": 0.7585635185241699, "learning_rate": 9.999795120817187e-05, "loss": 0.0981, "step": 4220 }, { "grad_norm": 0.8891108632087708, "learning_rate": 9.999775983353725e-05, "loss": 0.1205, "step": 4230 }, { "grad_norm": 0.6942175626754761, "learning_rate": 9.999755991565292e-05, "loss": 0.1079, "step": 4240 }, { "grad_norm": 0.8860461115837097, "learning_rate": 9.999735145455303e-05, "loss": 0.1111, "step": 4250 }, { "grad_norm": 0.5394997000694275, "learning_rate": 9.99971344502732e-05, "loss": 0.122, "step": 4260 }, { "grad_norm": 0.700219452381134, "learning_rate": 9.999690890285053e-05, "loss": 0.1029, "step": 4270 }, { "grad_norm": 0.7035778760910034, "learning_rate": 9.999667481232356e-05, "loss": 0.0909, "step": 4280 }, { "grad_norm": 0.8596712350845337, "learning_rate": 9.999643217873225e-05, "loss": 0.0938, "step": 4290 }, { "grad_norm": 0.8209409117698669, "learning_rate": 9.999618100211809e-05, "loss": 0.0832, "step": 4300 }, { "grad_norm": 0.6194326877593994, "learning_rate": 9.999592128252402e-05, "loss": 0.1044, "step": 4310 }, { "grad_norm": 0.677061915397644, "learning_rate": 9.999565301999437e-05, "loss": 0.1145, "step": 4320 }, { "grad_norm": 0.5560362339019775, "learning_rate": 9.999537621457502e-05, "loss": 0.0825, "step": 4330 }, { "grad_norm": 0.5672511458396912, "learning_rate": 9.999509086631323e-05, "loss": 0.0727, "step": 4340 }, { "grad_norm": 0.5995607972145081, "learning_rate": 9.99947969752578e-05, "loss": 0.0758, "step": 4350 }, { "grad_norm": 0.7886203527450562, "learning_rate": 9.999449454145891e-05, "loss": 0.0865, "step": 4360 }, { "grad_norm": 0.7477434873580933, "learning_rate": 9.999418356496827e-05, "loss": 0.082, "step": 4370 }, { "grad_norm": 0.47004109621047974, "learning_rate": 9.999386404583899e-05, "loss": 0.0895, "step": 4380 }, { "grad_norm": 0.7924354672431946, "learning_rate": 9.999353598412568e-05, "loss": 0.0829, "step": 4390 }, { "grad_norm": 0.592119574546814, "learning_rate": 9.999319937988442e-05, "loss": 0.0933, "step": 4400 }, { "grad_norm": 0.5133336186408997, "learning_rate": 9.999285423317268e-05, "loss": 0.0877, "step": 4410 }, { "grad_norm": 0.6858371496200562, "learning_rate": 9.999250054404947e-05, "loss": 0.1057, "step": 4420 }, { "grad_norm": 0.5327174663543701, "learning_rate": 9.99921383125752e-05, "loss": 0.076, "step": 4430 }, { "grad_norm": 0.5252895355224609, "learning_rate": 9.99917675388118e-05, "loss": 0.0971, "step": 4440 }, { "grad_norm": 0.7478704452514648, "learning_rate": 9.99913882228226e-05, "loss": 0.1086, "step": 4450 }, { "grad_norm": 0.6026208400726318, "learning_rate": 9.999100036467242e-05, "loss": 0.0887, "step": 4460 }, { "grad_norm": 0.8246649503707886, "learning_rate": 9.999060396442753e-05, "loss": 0.0809, "step": 4470 }, { "grad_norm": 0.6148375868797302, "learning_rate": 9.999019902215566e-05, "loss": 0.0988, "step": 4480 }, { "grad_norm": 0.6725165843963623, "learning_rate": 9.998978553792602e-05, "loss": 0.1036, "step": 4490 }, { "grad_norm": 0.7345813512802124, "learning_rate": 9.998936351180926e-05, "loss": 0.0791, "step": 4500 }, { "grad_norm": 0.7545507550239563, "learning_rate": 9.998893294387747e-05, "loss": 0.1086, "step": 4510 }, { "grad_norm": 1.026557207107544, "learning_rate": 9.998849383420426e-05, "loss": 0.0975, "step": 4520 }, { "grad_norm": 0.6478188633918762, "learning_rate": 9.998804618286465e-05, "loss": 0.0886, "step": 4530 }, { "grad_norm": 0.784540057182312, "learning_rate": 9.99875899899351e-05, "loss": 0.1106, "step": 4540 }, { "grad_norm": 0.7004691958427429, "learning_rate": 9.99871252554936e-05, "loss": 0.1053, "step": 4550 }, { "grad_norm": 0.5469407439231873, "learning_rate": 9.998665197961955e-05, "loss": 0.0905, "step": 4560 }, { "grad_norm": 0.7623053789138794, "learning_rate": 9.998617016239379e-05, "loss": 0.1078, "step": 4570 }, { "grad_norm": 1.2632991075515747, "learning_rate": 9.998567980389869e-05, "loss": 0.11, "step": 4580 }, { "grad_norm": 0.6546228528022766, "learning_rate": 9.998518090421802e-05, "loss": 0.094, "step": 4590 }, { "grad_norm": 1.0341155529022217, "learning_rate": 9.998467346343703e-05, "loss": 0.0947, "step": 4600 }, { "grad_norm": 0.6328631639480591, "learning_rate": 9.998415748164243e-05, "loss": 0.0904, "step": 4610 }, { "grad_norm": 0.6303930282592773, "learning_rate": 9.998363295892238e-05, "loss": 0.0889, "step": 4620 }, { "grad_norm": 0.8196330070495605, "learning_rate": 9.998309989536652e-05, "loss": 0.0994, "step": 4630 }, { "grad_norm": 0.5784780979156494, "learning_rate": 9.998255829106593e-05, "loss": 0.0914, "step": 4640 }, { "grad_norm": 0.9163643717765808, "learning_rate": 9.998200814611316e-05, "loss": 0.0845, "step": 4650 }, { "grad_norm": 0.6372926235198975, "learning_rate": 9.998144946060219e-05, "loss": 0.0917, "step": 4660 }, { "grad_norm": 0.8742778301239014, "learning_rate": 9.998088223462852e-05, "loss": 0.0803, "step": 4670 }, { "grad_norm": 0.6033772826194763, "learning_rate": 9.998030646828905e-05, "loss": 0.0752, "step": 4680 }, { "grad_norm": 0.5549464225769043, "learning_rate": 9.997972216168217e-05, "loss": 0.0879, "step": 4690 }, { "grad_norm": 0.8468894958496094, "learning_rate": 9.997912931490771e-05, "loss": 0.0874, "step": 4700 }, { "grad_norm": 0.6126832962036133, "learning_rate": 9.9978527928067e-05, "loss": 0.087, "step": 4710 }, { "grad_norm": 0.4593890607357025, "learning_rate": 9.997791800126277e-05, "loss": 0.0732, "step": 4720 }, { "grad_norm": 0.6225712895393372, "learning_rate": 9.997729953459927e-05, "loss": 0.102, "step": 4730 }, { "grad_norm": 0.5701650977134705, "learning_rate": 9.997667252818214e-05, "loss": 0.0866, "step": 4740 }, { "grad_norm": 0.4970948100090027, "learning_rate": 9.997603698211855e-05, "loss": 0.0754, "step": 4750 }, { "grad_norm": 0.711207389831543, "learning_rate": 9.99753928965171e-05, "loss": 0.0919, "step": 4760 }, { "grad_norm": 0.728497326374054, "learning_rate": 9.997474027148781e-05, "loss": 0.0906, "step": 4770 }, { "grad_norm": 0.7971452474594116, "learning_rate": 9.997407910714223e-05, "loss": 0.1058, "step": 4780 }, { "grad_norm": 0.5540167689323425, "learning_rate": 9.997340940359332e-05, "loss": 0.0794, "step": 4790 }, { "grad_norm": 0.6768692135810852, "learning_rate": 9.997273116095552e-05, "loss": 0.0776, "step": 4800 }, { "grad_norm": 0.6569985747337341, "learning_rate": 9.997204437934473e-05, "loss": 0.0892, "step": 4810 }, { "grad_norm": 0.7367137670516968, "learning_rate": 9.997134905887829e-05, "loss": 0.0881, "step": 4820 }, { "grad_norm": 0.5108292102813721, "learning_rate": 9.997064519967501e-05, "loss": 0.0859, "step": 4830 }, { "grad_norm": 0.6801072359085083, "learning_rate": 9.996993280185517e-05, "loss": 0.0974, "step": 4840 }, { "grad_norm": 0.7806439995765686, "learning_rate": 9.99692118655405e-05, "loss": 0.0825, "step": 4850 }, { "grad_norm": 0.566500723361969, "learning_rate": 9.996848239085417e-05, "loss": 0.0863, "step": 4860 }, { "grad_norm": 0.6260115504264832, "learning_rate": 9.996774437792085e-05, "loss": 0.0801, "step": 4870 }, { "grad_norm": 0.6480888724327087, "learning_rate": 9.996699782686664e-05, "loss": 0.0776, "step": 4880 }, { "grad_norm": 0.8614834547042847, "learning_rate": 9.996624273781909e-05, "loss": 0.0942, "step": 4890 }, { "grad_norm": 0.6140703558921814, "learning_rate": 9.996547911090725e-05, "loss": 0.0841, "step": 4900 }, { "grad_norm": 0.6786672472953796, "learning_rate": 9.996470694626157e-05, "loss": 0.0794, "step": 4910 }, { "grad_norm": 0.5019354224205017, "learning_rate": 9.996392624401403e-05, "loss": 0.0909, "step": 4920 }, { "grad_norm": 0.5277796983718872, "learning_rate": 9.996313700429801e-05, "loss": 0.0902, "step": 4930 }, { "grad_norm": 0.6521398425102234, "learning_rate": 9.996233922724836e-05, "loss": 0.0888, "step": 4940 }, { "grad_norm": 0.622067928314209, "learning_rate": 9.996153291300141e-05, "loss": 0.0901, "step": 4950 }, { "grad_norm": 0.5511813163757324, "learning_rate": 9.996071806169494e-05, "loss": 0.0762, "step": 4960 }, { "grad_norm": 0.534307062625885, "learning_rate": 9.995989467346817e-05, "loss": 0.0916, "step": 4970 }, { "grad_norm": 0.6178767085075378, "learning_rate": 9.995906274846183e-05, "loss": 0.094, "step": 4980 }, { "grad_norm": 0.616421639919281, "learning_rate": 9.995822228681803e-05, "loss": 0.0872, "step": 4990 }, { "grad_norm": 0.8179839253425598, "learning_rate": 9.99573732886804e-05, "loss": 0.0949, "step": 5000 }, { "grad_norm": 0.5490018725395203, "learning_rate": 9.995651575419402e-05, "loss": 0.0813, "step": 5010 }, { "grad_norm": 0.6885210275650024, "learning_rate": 9.995564968350541e-05, "loss": 0.0856, "step": 5020 }, { "grad_norm": 0.46702420711517334, "learning_rate": 9.995477507676256e-05, "loss": 0.0705, "step": 5030 }, { "grad_norm": 0.5650734901428223, "learning_rate": 9.995389193411493e-05, "loss": 0.0773, "step": 5040 }, { "grad_norm": 0.5763862133026123, "learning_rate": 9.995300025571339e-05, "loss": 0.084, "step": 5050 }, { "grad_norm": 0.5238661170005798, "learning_rate": 9.995210004171034e-05, "loss": 0.058, "step": 5060 }, { "grad_norm": 0.7001574635505676, "learning_rate": 9.995119129225956e-05, "loss": 0.0956, "step": 5070 }, { "grad_norm": 0.6833851337432861, "learning_rate": 9.995027400751637e-05, "loss": 0.0876, "step": 5080 }, { "grad_norm": 0.5825093388557434, "learning_rate": 9.994934818763751e-05, "loss": 0.072, "step": 5090 }, { "grad_norm": 0.5910405516624451, "learning_rate": 9.994841383278115e-05, "loss": 0.0716, "step": 5100 }, { "grad_norm": 0.5049072504043579, "learning_rate": 9.994747094310695e-05, "loss": 0.0817, "step": 5110 }, { "grad_norm": 0.6250694394111633, "learning_rate": 9.994651951877604e-05, "loss": 0.0925, "step": 5120 }, { "grad_norm": 0.6658788919448853, "learning_rate": 9.994555955995099e-05, "loss": 0.0917, "step": 5130 }, { "grad_norm": 0.6166968941688538, "learning_rate": 9.994459106679581e-05, "loss": 0.0769, "step": 5140 }, { "grad_norm": 0.5003327131271362, "learning_rate": 9.994361403947603e-05, "loss": 0.0702, "step": 5150 }, { "grad_norm": 0.6606822609901428, "learning_rate": 9.994262847815854e-05, "loss": 0.0952, "step": 5160 }, { "grad_norm": 0.6745010614395142, "learning_rate": 9.99416343830118e-05, "loss": 0.0731, "step": 5170 }, { "grad_norm": 0.5277637243270874, "learning_rate": 9.994063175420565e-05, "loss": 0.0653, "step": 5180 }, { "grad_norm": 0.5696021318435669, "learning_rate": 9.99396205919114e-05, "loss": 0.0919, "step": 5190 }, { "grad_norm": 0.6609699726104736, "learning_rate": 9.993860089630185e-05, "loss": 0.0672, "step": 5200 }, { "grad_norm": 0.7508015632629395, "learning_rate": 9.993757266755123e-05, "loss": 0.0747, "step": 5210 }, { "grad_norm": 0.6512196063995361, "learning_rate": 9.993653590583522e-05, "loss": 0.0886, "step": 5220 }, { "grad_norm": 0.6711141467094421, "learning_rate": 9.993549061133102e-05, "loss": 0.09, "step": 5230 }, { "grad_norm": 0.5571534633636475, "learning_rate": 9.993443678421719e-05, "loss": 0.0752, "step": 5240 }, { "grad_norm": 0.5702741146087646, "learning_rate": 9.993337442467384e-05, "loss": 0.096, "step": 5250 }, { "grad_norm": 0.6787834167480469, "learning_rate": 9.993230353288248e-05, "loss": 0.0708, "step": 5260 }, { "grad_norm": 0.6845564246177673, "learning_rate": 9.993122410902608e-05, "loss": 0.0811, "step": 5270 }, { "grad_norm": 0.6586082577705383, "learning_rate": 9.993013615328912e-05, "loss": 0.0933, "step": 5280 }, { "grad_norm": 0.5391817688941956, "learning_rate": 9.992903966585747e-05, "loss": 0.0948, "step": 5290 }, { "grad_norm": 0.555772066116333, "learning_rate": 9.992793464691852e-05, "loss": 0.0772, "step": 5300 }, { "grad_norm": 0.6765225529670715, "learning_rate": 9.992682109666105e-05, "loss": 0.0672, "step": 5310 }, { "grad_norm": 0.5258799195289612, "learning_rate": 9.992569901527538e-05, "loss": 0.0848, "step": 5320 }, { "grad_norm": 0.6532728672027588, "learning_rate": 9.99245684029532e-05, "loss": 0.0917, "step": 5330 }, { "grad_norm": 0.6207607388496399, "learning_rate": 9.992342925988774e-05, "loss": 0.0917, "step": 5340 }, { "grad_norm": 0.5355722308158875, "learning_rate": 9.992228158627361e-05, "loss": 0.0706, "step": 5350 }, { "grad_norm": 0.6426975131034851, "learning_rate": 9.992112538230693e-05, "loss": 0.0793, "step": 5360 }, { "grad_norm": 0.47208523750305176, "learning_rate": 9.991996064818527e-05, "loss": 0.0602, "step": 5370 }, { "grad_norm": 0.6063511371612549, "learning_rate": 9.991878738410768e-05, "loss": 0.0729, "step": 5380 }, { "grad_norm": 0.5740436911582947, "learning_rate": 9.991760559027457e-05, "loss": 0.0716, "step": 5390 }, { "grad_norm": 0.5121457576751709, "learning_rate": 9.991641526688793e-05, "loss": 0.0814, "step": 5400 }, { "grad_norm": 0.5245126485824585, "learning_rate": 9.991521641415113e-05, "loss": 0.0624, "step": 5410 }, { "grad_norm": 0.5176752209663391, "learning_rate": 9.991400903226904e-05, "loss": 0.0667, "step": 5420 }, { "grad_norm": 0.795061469078064, "learning_rate": 9.991279312144794e-05, "loss": 0.0741, "step": 5430 }, { "grad_norm": 0.543960452079773, "learning_rate": 9.991156868189564e-05, "loss": 0.0882, "step": 5440 }, { "grad_norm": 0.7123515605926514, "learning_rate": 9.991033571382131e-05, "loss": 0.064, "step": 5450 }, { "grad_norm": 0.7694076895713806, "learning_rate": 9.990909421743569e-05, "loss": 0.0735, "step": 5460 }, { "grad_norm": 0.5651885271072388, "learning_rate": 9.990784419295085e-05, "loss": 0.0699, "step": 5470 }, { "grad_norm": 0.5563700795173645, "learning_rate": 9.990658564058044e-05, "loss": 0.0831, "step": 5480 }, { "grad_norm": 0.5389290452003479, "learning_rate": 9.990531856053948e-05, "loss": 0.0798, "step": 5490 }, { "grad_norm": 0.6268002986907959, "learning_rate": 9.99040429530445e-05, "loss": 0.0792, "step": 5500 }, { "grad_norm": 0.55698162317276, "learning_rate": 9.990275881831346e-05, "loss": 0.0725, "step": 5510 }, { "grad_norm": 0.6406004428863525, "learning_rate": 9.990146615656577e-05, "loss": 0.0645, "step": 5520 }, { "grad_norm": 0.5405983924865723, "learning_rate": 9.990016496802233e-05, "loss": 0.067, "step": 5530 }, { "grad_norm": 0.6211642026901245, "learning_rate": 9.989885525290548e-05, "loss": 0.0647, "step": 5540 }, { "grad_norm": 0.5444684624671936, "learning_rate": 9.989753701143897e-05, "loss": 0.0662, "step": 5550 }, { "grad_norm": 0.5494222640991211, "learning_rate": 9.989621024384812e-05, "loss": 0.0701, "step": 5560 }, { "grad_norm": 0.5456683039665222, "learning_rate": 9.989487495035959e-05, "loss": 0.0676, "step": 5570 }, { "grad_norm": 0.7154433727264404, "learning_rate": 9.989353113120156e-05, "loss": 0.0959, "step": 5580 }, { "grad_norm": 0.7019950151443481, "learning_rate": 9.989217878660366e-05, "loss": 0.0684, "step": 5590 }, { "grad_norm": 0.5233292579650879, "learning_rate": 9.989081791679695e-05, "loss": 0.0845, "step": 5600 }, { "grad_norm": 0.38986119627952576, "learning_rate": 9.988944852201397e-05, "loss": 0.0734, "step": 5610 }, { "grad_norm": 0.49004414677619934, "learning_rate": 9.988807060248873e-05, "loss": 0.0719, "step": 5620 }, { "grad_norm": 0.49537384510040283, "learning_rate": 9.988668415845665e-05, "loss": 0.0603, "step": 5630 }, { "grad_norm": 0.528716504573822, "learning_rate": 9.988528919015466e-05, "loss": 0.0631, "step": 5640 }, { "grad_norm": 0.5960839986801147, "learning_rate": 9.988388569782112e-05, "loss": 0.079, "step": 5650 }, { "grad_norm": 0.5344069600105286, "learning_rate": 9.988247368169583e-05, "loss": 0.0703, "step": 5660 }, { "grad_norm": 0.7021846771240234, "learning_rate": 9.988105314202007e-05, "loss": 0.076, "step": 5670 }, { "grad_norm": 0.6220323443412781, "learning_rate": 9.987962407903659e-05, "loss": 0.0834, "step": 5680 }, { "grad_norm": 0.6788151264190674, "learning_rate": 9.987818649298957e-05, "loss": 0.0783, "step": 5690 }, { "grad_norm": 0.5171329379081726, "learning_rate": 9.987674038412465e-05, "loss": 0.0719, "step": 5700 }, { "grad_norm": 0.37608397006988525, "learning_rate": 9.987528575268891e-05, "loss": 0.0831, "step": 5710 }, { "grad_norm": 0.6554349660873413, "learning_rate": 9.987382259893095e-05, "loss": 0.0789, "step": 5720 }, { "grad_norm": 0.7734267115592957, "learning_rate": 9.987235092310074e-05, "loss": 0.0785, "step": 5730 }, { "grad_norm": 0.506623387336731, "learning_rate": 9.987087072544978e-05, "loss": 0.0943, "step": 5740 }, { "grad_norm": 0.4474213123321533, "learning_rate": 9.9869382006231e-05, "loss": 0.0764, "step": 5750 }, { "grad_norm": 0.544563889503479, "learning_rate": 9.986788476569875e-05, "loss": 0.0749, "step": 5760 }, { "grad_norm": 0.5963106751441956, "learning_rate": 9.986637900410887e-05, "loss": 0.0793, "step": 5770 }, { "grad_norm": 0.7394647598266602, "learning_rate": 9.986486472171869e-05, "loss": 0.0854, "step": 5780 }, { "grad_norm": 0.6714674234390259, "learning_rate": 9.986334191878692e-05, "loss": 0.0729, "step": 5790 }, { "grad_norm": 0.648161768913269, "learning_rate": 9.986181059557378e-05, "loss": 0.0856, "step": 5800 }, { "grad_norm": 0.6059330701828003, "learning_rate": 9.986027075234094e-05, "loss": 0.0754, "step": 5810 }, { "grad_norm": 0.7035854458808899, "learning_rate": 9.985872238935152e-05, "loss": 0.0843, "step": 5820 }, { "grad_norm": 0.7408549189567566, "learning_rate": 9.985716550687008e-05, "loss": 0.092, "step": 5830 }, { "grad_norm": 0.49011436104774475, "learning_rate": 9.985560010516264e-05, "loss": 0.0754, "step": 5840 }, { "grad_norm": 0.6202415227890015, "learning_rate": 9.985402618449668e-05, "loss": 0.0697, "step": 5850 }, { "grad_norm": 0.6768419146537781, "learning_rate": 9.985244374514118e-05, "loss": 0.0968, "step": 5860 }, { "grad_norm": 0.713105320930481, "learning_rate": 9.985085278736651e-05, "loss": 0.0896, "step": 5870 }, { "grad_norm": 0.5526664853096008, "learning_rate": 9.984925331144452e-05, "loss": 0.0892, "step": 5880 }, { "grad_norm": 0.4502007067203522, "learning_rate": 9.984764531764851e-05, "loss": 0.0721, "step": 5890 }, { "grad_norm": 0.5072143077850342, "learning_rate": 9.984602880625326e-05, "loss": 0.0928, "step": 5900 }, { "grad_norm": 0.5065177083015442, "learning_rate": 9.9844403777535e-05, "loss": 0.0856, "step": 5910 }, { "grad_norm": 0.7563795447349548, "learning_rate": 9.984277023177135e-05, "loss": 0.073, "step": 5920 }, { "grad_norm": 0.5390533804893494, "learning_rate": 9.984112816924148e-05, "loss": 0.0607, "step": 5930 }, { "grad_norm": 0.7828633785247803, "learning_rate": 9.983947759022596e-05, "loss": 0.0863, "step": 5940 }, { "grad_norm": 0.491927832365036, "learning_rate": 9.983781849500682e-05, "loss": 0.0733, "step": 5950 }, { "grad_norm": 0.5303486585617065, "learning_rate": 9.98361508838676e-05, "loss": 0.0676, "step": 5960 }, { "grad_norm": 0.8301680088043213, "learning_rate": 9.98344747570932e-05, "loss": 0.0735, "step": 5970 }, { "grad_norm": 0.6053450703620911, "learning_rate": 9.983279011497004e-05, "loss": 0.0735, "step": 5980 }, { "grad_norm": 0.5901185274124146, "learning_rate": 9.983109695778596e-05, "loss": 0.0882, "step": 5990 }, { "grad_norm": 0.5336601138114929, "learning_rate": 9.982939528583032e-05, "loss": 0.0866, "step": 6000 }, { "grad_norm": 0.718687117099762, "learning_rate": 9.982768509939385e-05, "loss": 0.0725, "step": 6010 }, { "grad_norm": 0.43591639399528503, "learning_rate": 9.982596639876879e-05, "loss": 0.0727, "step": 6020 }, { "grad_norm": 0.5014497637748718, "learning_rate": 9.982423918424881e-05, "loss": 0.0796, "step": 6030 }, { "grad_norm": 0.5640562176704407, "learning_rate": 9.982250345612908e-05, "loss": 0.0729, "step": 6040 }, { "grad_norm": 0.5787804126739502, "learning_rate": 9.982075921470611e-05, "loss": 0.0755, "step": 6050 }, { "grad_norm": 0.5774450302124023, "learning_rate": 9.981900646027802e-05, "loss": 0.0788, "step": 6060 }, { "grad_norm": 0.5977222323417664, "learning_rate": 9.981724519314425e-05, "loss": 0.0696, "step": 6070 }, { "grad_norm": 0.7016507983207703, "learning_rate": 9.981547541360581e-05, "loss": 0.0872, "step": 6080 }, { "grad_norm": 0.6319453120231628, "learning_rate": 9.981369712196508e-05, "loss": 0.0762, "step": 6090 }, { "grad_norm": 0.5863970518112183, "learning_rate": 9.981191031852592e-05, "loss": 0.103, "step": 6100 }, { "grad_norm": 0.5929358005523682, "learning_rate": 9.981011500359362e-05, "loss": 0.0756, "step": 6110 }, { "grad_norm": 0.567919909954071, "learning_rate": 9.9808311177475e-05, "loss": 0.079, "step": 6120 }, { "grad_norm": 0.5397199988365173, "learning_rate": 9.980649884047826e-05, "loss": 0.0707, "step": 6130 }, { "grad_norm": 0.5450133085250854, "learning_rate": 9.980467799291307e-05, "loss": 0.0799, "step": 6140 }, { "grad_norm": 0.6004332304000854, "learning_rate": 9.980284863509058e-05, "loss": 0.1039, "step": 6150 }, { "grad_norm": 0.5985152125358582, "learning_rate": 9.980101076732334e-05, "loss": 0.1008, "step": 6160 }, { "grad_norm": 0.7144010066986084, "learning_rate": 9.979916438992544e-05, "loss": 0.0837, "step": 6170 }, { "grad_norm": 0.6462729573249817, "learning_rate": 9.979730950321237e-05, "loss": 0.0739, "step": 6180 }, { "grad_norm": 0.4803813397884369, "learning_rate": 9.979544610750104e-05, "loss": 0.0872, "step": 6190 }, { "grad_norm": 0.6137504577636719, "learning_rate": 9.97935742031099e-05, "loss": 0.0781, "step": 6200 }, { "grad_norm": 0.8876283764839172, "learning_rate": 9.979169379035878e-05, "loss": 0.087, "step": 6210 }, { "grad_norm": 0.6271437406539917, "learning_rate": 9.978980486956899e-05, "loss": 0.0848, "step": 6220 }, { "grad_norm": 0.4619094431400299, "learning_rate": 9.978790744106332e-05, "loss": 0.0781, "step": 6230 }, { "grad_norm": 0.6608024835586548, "learning_rate": 9.978600150516594e-05, "loss": 0.0814, "step": 6240 }, { "grad_norm": 0.6373734474182129, "learning_rate": 9.978408706220259e-05, "loss": 0.0864, "step": 6250 }, { "grad_norm": 0.5617753863334656, "learning_rate": 9.978216411250032e-05, "loss": 0.0611, "step": 6260 }, { "grad_norm": 0.6068397164344788, "learning_rate": 9.978023265638778e-05, "loss": 0.0625, "step": 6270 }, { "grad_norm": 0.5880513191223145, "learning_rate": 9.977829269419495e-05, "loss": 0.0799, "step": 6280 }, { "grad_norm": 0.62727952003479, "learning_rate": 9.977634422625335e-05, "loss": 0.0743, "step": 6290 }, { "grad_norm": 0.514182448387146, "learning_rate": 9.97743872528959e-05, "loss": 0.0715, "step": 6300 }, { "grad_norm": 0.8010825514793396, "learning_rate": 9.9772421774457e-05, "loss": 0.0713, "step": 6310 }, { "grad_norm": 0.7266963124275208, "learning_rate": 9.977044779127252e-05, "loss": 0.0745, "step": 6320 }, { "grad_norm": 0.5579684376716614, "learning_rate": 9.976846530367971e-05, "loss": 0.0881, "step": 6330 }, { "grad_norm": 1.1744409799575806, "learning_rate": 9.976647431201735e-05, "loss": 0.0957, "step": 6340 }, { "grad_norm": 1.4612772464752197, "learning_rate": 9.976447481662568e-05, "loss": 0.0899, "step": 6350 }, { "grad_norm": 0.7346428632736206, "learning_rate": 9.976246681784629e-05, "loss": 0.0897, "step": 6360 }, { "grad_norm": 0.5745630860328674, "learning_rate": 9.976045031602234e-05, "loss": 0.0833, "step": 6370 }, { "grad_norm": 0.786571204662323, "learning_rate": 9.975842531149837e-05, "loss": 0.0772, "step": 6380 }, { "grad_norm": 0.6148983836174011, "learning_rate": 9.975639180462043e-05, "loss": 0.0895, "step": 6390 }, { "grad_norm": 0.5811973214149475, "learning_rate": 9.975434979573596e-05, "loss": 0.0845, "step": 6400 }, { "grad_norm": 0.639248251914978, "learning_rate": 9.97522992851939e-05, "loss": 0.0746, "step": 6410 }, { "grad_norm": 0.37706416845321655, "learning_rate": 9.975024027334461e-05, "loss": 0.0797, "step": 6420 }, { "grad_norm": 0.46079862117767334, "learning_rate": 9.974817276053993e-05, "loss": 0.0705, "step": 6430 }, { "grad_norm": 0.7819085717201233, "learning_rate": 9.974609674713315e-05, "loss": 0.0827, "step": 6440 }, { "grad_norm": 0.638825535774231, "learning_rate": 9.9744012233479e-05, "loss": 0.0736, "step": 6450 }, { "grad_norm": 0.681870698928833, "learning_rate": 9.974191921993366e-05, "loss": 0.0777, "step": 6460 }, { "grad_norm": 0.6300003528594971, "learning_rate": 9.973981770685474e-05, "loss": 0.0874, "step": 6470 }, { "grad_norm": 0.573358416557312, "learning_rate": 9.97377076946014e-05, "loss": 0.0708, "step": 6480 }, { "grad_norm": 0.7557254433631897, "learning_rate": 9.973558918353412e-05, "loss": 0.097, "step": 6490 }, { "grad_norm": 0.46418336033821106, "learning_rate": 9.973346217401494e-05, "loss": 0.0763, "step": 6500 }, { "grad_norm": 0.6024410128593445, "learning_rate": 9.973132666640726e-05, "loss": 0.0807, "step": 6510 }, { "grad_norm": 0.4550490081310272, "learning_rate": 9.972918266107602e-05, "loss": 0.0769, "step": 6520 }, { "grad_norm": 0.5039570331573486, "learning_rate": 9.972703015838756e-05, "loss": 0.0645, "step": 6530 }, { "grad_norm": 0.40966469049453735, "learning_rate": 9.97248691587097e-05, "loss": 0.0758, "step": 6540 }, { "grad_norm": 0.6703236103057861, "learning_rate": 9.972269966241166e-05, "loss": 0.0648, "step": 6550 }, { "grad_norm": 0.5171713829040527, "learning_rate": 9.972052166986417e-05, "loss": 0.0792, "step": 6560 }, { "grad_norm": 0.45623403787612915, "learning_rate": 9.971833518143938e-05, "loss": 0.087, "step": 6570 }, { "grad_norm": 0.6203022599220276, "learning_rate": 9.971614019751093e-05, "loss": 0.0632, "step": 6580 }, { "grad_norm": 0.40901532769203186, "learning_rate": 9.971393671845383e-05, "loss": 0.0783, "step": 6590 }, { "grad_norm": 0.6678228974342346, "learning_rate": 9.971172474464464e-05, "loss": 0.0917, "step": 6600 }, { "grad_norm": 0.5073179006576538, "learning_rate": 9.97095042764613e-05, "loss": 0.0652, "step": 6610 }, { "grad_norm": 0.57925945520401, "learning_rate": 9.970727531428324e-05, "loss": 0.0671, "step": 6620 }, { "grad_norm": 0.5386475324630737, "learning_rate": 9.970503785849132e-05, "loss": 0.0733, "step": 6630 }, { "grad_norm": 0.5201948285102844, "learning_rate": 9.970279190946788e-05, "loss": 0.072, "step": 6640 }, { "grad_norm": 0.6248980760574341, "learning_rate": 9.970053746759667e-05, "loss": 0.0886, "step": 6650 }, { "grad_norm": 0.4721408486366272, "learning_rate": 9.969827453326292e-05, "loss": 0.0683, "step": 6660 }, { "grad_norm": 0.4190390110015869, "learning_rate": 9.969600310685332e-05, "loss": 0.0617, "step": 6670 }, { "grad_norm": 0.5972900390625, "learning_rate": 9.969372318875596e-05, "loss": 0.0824, "step": 6680 }, { "grad_norm": 0.5665847659111023, "learning_rate": 9.969143477936043e-05, "loss": 0.0724, "step": 6690 }, { "grad_norm": 0.47464439272880554, "learning_rate": 9.968913787905775e-05, "loss": 0.0854, "step": 6700 }, { "grad_norm": 0.5737347602844238, "learning_rate": 9.968683248824045e-05, "loss": 0.0958, "step": 6710 }, { "grad_norm": 0.5507679581642151, "learning_rate": 9.968451860730238e-05, "loss": 0.0824, "step": 6720 }, { "grad_norm": 0.5559749603271484, "learning_rate": 9.968219623663896e-05, "loss": 0.0792, "step": 6730 }, { "grad_norm": 0.7557348012924194, "learning_rate": 9.967986537664702e-05, "loss": 0.0927, "step": 6740 }, { "grad_norm": 0.5735617280006409, "learning_rate": 9.967752602772483e-05, "loss": 0.0713, "step": 6750 }, { "grad_norm": 0.44739583134651184, "learning_rate": 9.967517819027212e-05, "loss": 0.0853, "step": 6760 }, { "grad_norm": 0.5550826787948608, "learning_rate": 9.967282186469009e-05, "loss": 0.0737, "step": 6770 }, { "grad_norm": 0.5789654850959778, "learning_rate": 9.967045705138135e-05, "loss": 0.0827, "step": 6780 }, { "grad_norm": 0.4421570599079132, "learning_rate": 9.966808375074998e-05, "loss": 0.0737, "step": 6790 }, { "grad_norm": 0.5904650688171387, "learning_rate": 9.966570196320154e-05, "loss": 0.0825, "step": 6800 }, { "grad_norm": 0.6288572549819946, "learning_rate": 9.966331168914299e-05, "loss": 0.0707, "step": 6810 }, { "grad_norm": 0.5048525929450989, "learning_rate": 9.966091292898277e-05, "loss": 0.0831, "step": 6820 }, { "grad_norm": 0.5379375219345093, "learning_rate": 9.965850568313076e-05, "loss": 0.0765, "step": 6830 }, { "grad_norm": 0.5070455074310303, "learning_rate": 9.965608995199827e-05, "loss": 0.07, "step": 6840 }, { "grad_norm": 0.47188904881477356, "learning_rate": 9.965366573599812e-05, "loss": 0.072, "step": 6850 }, { "grad_norm": 0.5382623672485352, "learning_rate": 9.965123303554453e-05, "loss": 0.0662, "step": 6860 }, { "grad_norm": 0.8112098574638367, "learning_rate": 9.964879185105317e-05, "loss": 0.0954, "step": 6870 }, { "grad_norm": 0.5188286900520325, "learning_rate": 9.964634218294119e-05, "loss": 0.0784, "step": 6880 }, { "grad_norm": 0.4800379276275635, "learning_rate": 9.964388403162714e-05, "loss": 0.0691, "step": 6890 }, { "grad_norm": 0.7261210680007935, "learning_rate": 9.96414173975311e-05, "loss": 0.0723, "step": 6900 }, { "grad_norm": 0.6198140382766724, "learning_rate": 9.963894228107451e-05, "loss": 0.0673, "step": 6910 }, { "grad_norm": 0.47229304909706116, "learning_rate": 9.963645868268032e-05, "loss": 0.0735, "step": 6920 }, { "grad_norm": 0.6618477702140808, "learning_rate": 9.963396660277289e-05, "loss": 0.0788, "step": 6930 }, { "grad_norm": 0.5825967788696289, "learning_rate": 9.963146604177807e-05, "loss": 0.0659, "step": 6940 }, { "grad_norm": 0.4330836832523346, "learning_rate": 9.962895700012311e-05, "loss": 0.0659, "step": 6950 }, { "grad_norm": 0.49334239959716797, "learning_rate": 9.962643947823677e-05, "loss": 0.0725, "step": 6960 }, { "grad_norm": 0.43862384557724, "learning_rate": 9.962391347654921e-05, "loss": 0.0718, "step": 6970 }, { "grad_norm": 0.45068755745887756, "learning_rate": 9.962137899549204e-05, "loss": 0.0762, "step": 6980 }, { "grad_norm": 0.5765212178230286, "learning_rate": 9.961883603549835e-05, "loss": 0.0563, "step": 6990 }, { "grad_norm": 0.5494239926338196, "learning_rate": 9.961628459700267e-05, "loss": 0.0673, "step": 7000 }, { "grad_norm": 0.443722128868103, "learning_rate": 9.961372468044095e-05, "loss": 0.0802, "step": 7010 }, { "grad_norm": 0.42710810899734497, "learning_rate": 9.961115628625062e-05, "loss": 0.065, "step": 7020 }, { "grad_norm": 0.565321147441864, "learning_rate": 9.960857941487056e-05, "loss": 0.0789, "step": 7030 }, { "grad_norm": 0.5818016529083252, "learning_rate": 9.960599406674106e-05, "loss": 0.0767, "step": 7040 }, { "grad_norm": 0.49598342180252075, "learning_rate": 9.960340024230393e-05, "loss": 0.0678, "step": 7050 }, { "grad_norm": 0.5576131343841553, "learning_rate": 9.960079794200232e-05, "loss": 0.0631, "step": 7060 }, { "grad_norm": 0.5745858550071716, "learning_rate": 9.959818716628096e-05, "loss": 0.0699, "step": 7070 }, { "grad_norm": 0.6382286548614502, "learning_rate": 9.95955679155859e-05, "loss": 0.0812, "step": 7080 }, { "grad_norm": 0.698692262172699, "learning_rate": 9.959294019036472e-05, "loss": 0.0815, "step": 7090 }, { "grad_norm": 0.5298591256141663, "learning_rate": 9.959030399106646e-05, "loss": 0.0614, "step": 7100 }, { "grad_norm": 0.45194125175476074, "learning_rate": 9.958765931814153e-05, "loss": 0.0811, "step": 7110 }, { "grad_norm": 0.44119659066200256, "learning_rate": 9.958500617204184e-05, "loss": 0.0618, "step": 7120 }, { "grad_norm": 0.34569719433784485, "learning_rate": 9.958234455322075e-05, "loss": 0.0619, "step": 7130 }, { "grad_norm": 0.5333142876625061, "learning_rate": 9.957967446213308e-05, "loss": 0.0791, "step": 7140 }, { "grad_norm": 0.39675959944725037, "learning_rate": 9.957699589923501e-05, "loss": 0.0755, "step": 7150 }, { "grad_norm": 0.7225584983825684, "learning_rate": 9.957430886498431e-05, "loss": 0.0741, "step": 7160 }, { "grad_norm": 0.47951996326446533, "learning_rate": 9.957161335984008e-05, "loss": 0.0721, "step": 7170 }, { "grad_norm": 0.5063344836235046, "learning_rate": 9.956890938426291e-05, "loss": 0.0667, "step": 7180 }, { "grad_norm": 0.5401932001113892, "learning_rate": 9.956619693871482e-05, "loss": 0.0775, "step": 7190 }, { "grad_norm": 0.6678482890129089, "learning_rate": 9.956347602365934e-05, "loss": 0.0788, "step": 7200 }, { "grad_norm": 0.41869333386421204, "learning_rate": 9.956074663956135e-05, "loss": 0.0721, "step": 7210 }, { "grad_norm": 0.54111647605896, "learning_rate": 9.955800878688726e-05, "loss": 0.0727, "step": 7220 }, { "grad_norm": 0.48890072107315063, "learning_rate": 9.955526246610489e-05, "loss": 0.0589, "step": 7230 }, { "grad_norm": 0.6309622526168823, "learning_rate": 9.955250767768349e-05, "loss": 0.0773, "step": 7240 }, { "grad_norm": 0.4149240553379059, "learning_rate": 9.95497444220938e-05, "loss": 0.0828, "step": 7250 }, { "grad_norm": 0.4505969285964966, "learning_rate": 9.954697269980797e-05, "loss": 0.0585, "step": 7260 }, { "grad_norm": 0.43965384364128113, "learning_rate": 9.954419251129962e-05, "loss": 0.0709, "step": 7270 }, { "grad_norm": 0.5516752004623413, "learning_rate": 9.95414038570438e-05, "loss": 0.0657, "step": 7280 }, { "grad_norm": 0.7275952100753784, "learning_rate": 9.953860673751703e-05, "loss": 0.0724, "step": 7290 }, { "grad_norm": 0.48266151547431946, "learning_rate": 9.953580115319725e-05, "loss": 0.0697, "step": 7300 }, { "grad_norm": 0.5103833675384521, "learning_rate": 9.953298710456387e-05, "loss": 0.0733, "step": 7310 }, { "grad_norm": 0.5066249370574951, "learning_rate": 9.953016459209771e-05, "loss": 0.0609, "step": 7320 }, { "grad_norm": 0.4472573697566986, "learning_rate": 9.952733361628108e-05, "loss": 0.0657, "step": 7330 }, { "grad_norm": 0.6715189814567566, "learning_rate": 9.952449417759772e-05, "loss": 0.065, "step": 7340 }, { "grad_norm": 0.5587707161903381, "learning_rate": 9.952164627653279e-05, "loss": 0.0613, "step": 7350 }, { "grad_norm": 0.6681423187255859, "learning_rate": 9.951878991357292e-05, "loss": 0.0678, "step": 7360 }, { "grad_norm": 0.5757527947425842, "learning_rate": 9.951592508920622e-05, "loss": 0.0761, "step": 7370 }, { "grad_norm": 0.3906566798686981, "learning_rate": 9.951305180392219e-05, "loss": 0.0606, "step": 7380 }, { "grad_norm": 0.6281169056892395, "learning_rate": 9.951017005821178e-05, "loss": 0.0794, "step": 7390 }, { "grad_norm": 0.5498671531677246, "learning_rate": 9.95072798525674e-05, "loss": 0.061, "step": 7400 }, { "grad_norm": 0.6515856385231018, "learning_rate": 9.950438118748293e-05, "loss": 0.0709, "step": 7410 }, { "grad_norm": 0.7332574725151062, "learning_rate": 9.950147406345366e-05, "loss": 0.0813, "step": 7420 }, { "grad_norm": 0.47901201248168945, "learning_rate": 9.949855848097635e-05, "loss": 0.0785, "step": 7430 }, { "grad_norm": 0.6077988743782043, "learning_rate": 9.949563444054916e-05, "loss": 0.0766, "step": 7440 }, { "grad_norm": 0.5048779845237732, "learning_rate": 9.949270194267178e-05, "loss": 0.0755, "step": 7450 }, { "grad_norm": 0.4355463683605194, "learning_rate": 9.948976098784526e-05, "loss": 0.0753, "step": 7460 }, { "grad_norm": 0.49437054991722107, "learning_rate": 9.948681157657213e-05, "loss": 0.075, "step": 7470 }, { "grad_norm": 0.6297447681427002, "learning_rate": 9.948385370935638e-05, "loss": 0.0713, "step": 7480 }, { "grad_norm": 0.8528733253479004, "learning_rate": 9.94808873867034e-05, "loss": 0.0757, "step": 7490 }, { "grad_norm": 0.5286040902137756, "learning_rate": 9.947791260912009e-05, "loss": 0.0756, "step": 7500 }, { "grad_norm": 0.639883816242218, "learning_rate": 9.947492937711474e-05, "loss": 0.0573, "step": 7510 }, { "grad_norm": 0.5973951816558838, "learning_rate": 9.947193769119707e-05, "loss": 0.0626, "step": 7520 }, { "grad_norm": 0.4800683259963989, "learning_rate": 9.946893755187834e-05, "loss": 0.0625, "step": 7530 }, { "grad_norm": 0.49747976660728455, "learning_rate": 9.946592895967115e-05, "loss": 0.0654, "step": 7540 }, { "grad_norm": 0.5777784585952759, "learning_rate": 9.94629119150896e-05, "loss": 0.0637, "step": 7550 }, { "grad_norm": 0.33928969502449036, "learning_rate": 9.94598864186492e-05, "loss": 0.0655, "step": 7560 }, { "grad_norm": 0.5158814787864685, "learning_rate": 9.945685247086696e-05, "loss": 0.0809, "step": 7570 }, { "grad_norm": 0.49390870332717896, "learning_rate": 9.945381007226129e-05, "loss": 0.0738, "step": 7580 }, { "grad_norm": 0.428954541683197, "learning_rate": 9.945075922335203e-05, "loss": 0.0778, "step": 7590 }, { "grad_norm": 0.6784534454345703, "learning_rate": 9.944769992466049e-05, "loss": 0.0902, "step": 7600 }, { "grad_norm": 1.1957206726074219, "learning_rate": 9.944463217670945e-05, "loss": 0.0897, "step": 7610 }, { "grad_norm": 0.5908552408218384, "learning_rate": 9.944155598002307e-05, "loss": 0.0904, "step": 7620 }, { "grad_norm": 0.6183780431747437, "learning_rate": 9.943847133512701e-05, "loss": 0.0951, "step": 7630 }, { "grad_norm": 0.6401304006576538, "learning_rate": 9.943537824254834e-05, "loss": 0.0737, "step": 7640 }, { "grad_norm": 0.5548728108406067, "learning_rate": 9.943227670281559e-05, "loss": 0.0703, "step": 7650 }, { "grad_norm": 0.46167534589767456, "learning_rate": 9.942916671645873e-05, "loss": 0.0675, "step": 7660 }, { "grad_norm": 0.5759041905403137, "learning_rate": 9.942604828400916e-05, "loss": 0.0741, "step": 7670 }, { "grad_norm": 0.49125584959983826, "learning_rate": 9.942292140599975e-05, "loss": 0.0763, "step": 7680 }, { "grad_norm": 0.5452987551689148, "learning_rate": 9.94197860829648e-05, "loss": 0.0675, "step": 7690 }, { "grad_norm": 0.6196016073226929, "learning_rate": 9.941664231544004e-05, "loss": 0.0562, "step": 7700 }, { "grad_norm": 0.572903573513031, "learning_rate": 9.941349010396264e-05, "loss": 0.0758, "step": 7710 }, { "grad_norm": 0.540608823299408, "learning_rate": 9.941032944907125e-05, "loss": 0.0682, "step": 7720 }, { "grad_norm": 0.6081307530403137, "learning_rate": 9.940716035130596e-05, "loss": 0.0698, "step": 7730 }, { "grad_norm": 0.6600297689437866, "learning_rate": 9.940398281120821e-05, "loss": 0.0728, "step": 7740 }, { "grad_norm": 0.5032612681388855, "learning_rate": 9.940079682932102e-05, "loss": 0.0638, "step": 7750 }, { "grad_norm": 0.4892394244670868, "learning_rate": 9.939760240618877e-05, "loss": 0.0714, "step": 7760 }, { "grad_norm": 0.46477004885673523, "learning_rate": 9.939439954235729e-05, "loss": 0.0678, "step": 7770 }, { "grad_norm": 0.6043347716331482, "learning_rate": 9.939118823837387e-05, "loss": 0.0762, "step": 7780 }, { "grad_norm": 0.5683153867721558, "learning_rate": 9.938796849478725e-05, "loss": 0.0726, "step": 7790 }, { "grad_norm": 0.5477091073989868, "learning_rate": 9.938474031214755e-05, "loss": 0.0636, "step": 7800 }, { "grad_norm": 0.5008305907249451, "learning_rate": 9.938150369100643e-05, "loss": 0.0653, "step": 7810 }, { "grad_norm": 0.3665500283241272, "learning_rate": 9.93782586319169e-05, "loss": 0.0532, "step": 7820 }, { "grad_norm": 0.7107347249984741, "learning_rate": 9.937500513543348e-05, "loss": 0.0688, "step": 7830 }, { "grad_norm": 0.5122933387756348, "learning_rate": 9.937174320211207e-05, "loss": 0.0558, "step": 7840 }, { "grad_norm": 0.4057683050632477, "learning_rate": 9.936847283251009e-05, "loss": 0.0601, "step": 7850 }, { "grad_norm": 0.5001450181007385, "learning_rate": 9.936519402718632e-05, "loss": 0.0694, "step": 7860 }, { "grad_norm": 0.4944629371166229, "learning_rate": 9.936190678670102e-05, "loss": 0.0808, "step": 7870 }, { "grad_norm": 0.4740886092185974, "learning_rate": 9.935861111161593e-05, "loss": 0.0768, "step": 7880 }, { "grad_norm": 0.4646363854408264, "learning_rate": 9.935530700249416e-05, "loss": 0.0698, "step": 7890 }, { "grad_norm": 0.4636951684951782, "learning_rate": 9.935199445990028e-05, "loss": 0.0784, "step": 7900 }, { "grad_norm": 0.37709519267082214, "learning_rate": 9.934867348440033e-05, "loss": 0.0704, "step": 7910 }, { "grad_norm": 0.6391862034797668, "learning_rate": 9.934534407656176e-05, "loss": 0.0933, "step": 7920 }, { "grad_norm": 0.6200513243675232, "learning_rate": 9.93420062369535e-05, "loss": 0.0729, "step": 7930 }, { "grad_norm": 0.6063004732131958, "learning_rate": 9.933865996614589e-05, "loss": 0.0653, "step": 7940 }, { "grad_norm": 0.4772430956363678, "learning_rate": 9.933530526471068e-05, "loss": 0.0641, "step": 7950 }, { "grad_norm": 0.5762190222740173, "learning_rate": 9.933194213322114e-05, "loss": 0.0631, "step": 7960 }, { "grad_norm": 0.46835631132125854, "learning_rate": 9.932857057225192e-05, "loss": 0.0686, "step": 7970 }, { "grad_norm": 0.46436503529548645, "learning_rate": 9.932519058237912e-05, "loss": 0.063, "step": 7980 }, { "grad_norm": 0.5160051584243774, "learning_rate": 9.932180216418032e-05, "loss": 0.0723, "step": 7990 }, { "grad_norm": 0.5355035066604614, "learning_rate": 9.931840531823446e-05, "loss": 0.0617, "step": 8000 }, { "grad_norm": 0.4664841592311859, "learning_rate": 9.9315000045122e-05, "loss": 0.0535, "step": 8010 }, { "grad_norm": 0.3615034222602844, "learning_rate": 9.931158634542481e-05, "loss": 0.0547, "step": 8020 }, { "grad_norm": 0.38160276412963867, "learning_rate": 9.930816421972617e-05, "loss": 0.0546, "step": 8030 }, { "grad_norm": 0.5298890471458435, "learning_rate": 9.930473366861086e-05, "loss": 0.0596, "step": 8040 }, { "grad_norm": 0.5021455883979797, "learning_rate": 9.930129469266505e-05, "loss": 0.0748, "step": 8050 }, { "grad_norm": 0.49332311749458313, "learning_rate": 9.929784729247638e-05, "loss": 0.0583, "step": 8060 }, { "grad_norm": 0.43523672223091125, "learning_rate": 9.929439146863389e-05, "loss": 0.0679, "step": 8070 }, { "grad_norm": 0.7976360321044922, "learning_rate": 9.92909272217281e-05, "loss": 0.0857, "step": 8080 }, { "grad_norm": 0.5476901531219482, "learning_rate": 9.928745455235097e-05, "loss": 0.0646, "step": 8090 }, { "grad_norm": 0.4418884813785553, "learning_rate": 9.928397346109588e-05, "loss": 0.0736, "step": 8100 }, { "grad_norm": 0.7524136900901794, "learning_rate": 9.928048394855762e-05, "loss": 0.0759, "step": 8110 }, { "grad_norm": 0.4879102110862732, "learning_rate": 9.92769860153325e-05, "loss": 0.0735, "step": 8120 }, { "grad_norm": 0.5620765686035156, "learning_rate": 9.927347966201819e-05, "loss": 0.0818, "step": 8130 }, { "grad_norm": 0.6255418658256531, "learning_rate": 9.926996488921383e-05, "loss": 0.0702, "step": 8140 }, { "grad_norm": 0.6566426753997803, "learning_rate": 9.926644169752001e-05, "loss": 0.0663, "step": 8150 }, { "grad_norm": 0.5802999138832092, "learning_rate": 9.926291008753875e-05, "loss": 0.0809, "step": 8160 }, { "grad_norm": 0.46422284841537476, "learning_rate": 9.92593700598735e-05, "loss": 0.0752, "step": 8170 }, { "grad_norm": 0.44805005192756653, "learning_rate": 9.925582161512915e-05, "loss": 0.0681, "step": 8180 }, { "grad_norm": 0.45058223605155945, "learning_rate": 9.925226475391205e-05, "loss": 0.0746, "step": 8190 }, { "grad_norm": 0.5108029246330261, "learning_rate": 9.924869947682993e-05, "loss": 0.0723, "step": 8200 }, { "grad_norm": 0.5867398381233215, "learning_rate": 9.924512578449204e-05, "loss": 0.0638, "step": 8210 }, { "grad_norm": 0.44280797243118286, "learning_rate": 9.924154367750901e-05, "loss": 0.0754, "step": 8220 }, { "grad_norm": 0.6868221163749695, "learning_rate": 9.923795315649293e-05, "loss": 0.0665, "step": 8230 }, { "grad_norm": 0.42269033193588257, "learning_rate": 9.92343542220573e-05, "loss": 0.071, "step": 8240 }, { "grad_norm": 0.5377534627914429, "learning_rate": 9.92307468748171e-05, "loss": 0.0639, "step": 8250 }, { "grad_norm": 0.42475631833076477, "learning_rate": 9.922713111538873e-05, "loss": 0.0578, "step": 8260 }, { "grad_norm": 0.6612101793289185, "learning_rate": 9.922350694439003e-05, "loss": 0.0653, "step": 8270 }, { "grad_norm": 0.6511193513870239, "learning_rate": 9.921987436244024e-05, "loss": 0.0763, "step": 8280 }, { "grad_norm": 0.546349287033081, "learning_rate": 9.921623337016008e-05, "loss": 0.058, "step": 8290 }, { "grad_norm": 0.4558015763759613, "learning_rate": 9.921258396817172e-05, "loss": 0.0713, "step": 8300 }, { "grad_norm": 0.42223605513572693, "learning_rate": 9.920892615709874e-05, "loss": 0.0626, "step": 8310 }, { "grad_norm": 0.4090694785118103, "learning_rate": 9.920525993756612e-05, "loss": 0.0532, "step": 8320 }, { "grad_norm": 0.5921701788902283, "learning_rate": 9.920158531020036e-05, "loss": 0.0685, "step": 8330 }, { "grad_norm": 0.5191918015480042, "learning_rate": 9.919790227562933e-05, "loss": 0.0537, "step": 8340 }, { "grad_norm": 0.5003188848495483, "learning_rate": 9.919421083448237e-05, "loss": 0.0764, "step": 8350 }, { "grad_norm": 0.5354146957397461, "learning_rate": 9.919051098739022e-05, "loss": 0.0602, "step": 8360 }, { "grad_norm": 0.5996882319450378, "learning_rate": 9.918680273498514e-05, "loss": 0.0599, "step": 8370 }, { "grad_norm": 0.5240553617477417, "learning_rate": 9.918308607790072e-05, "loss": 0.0717, "step": 8380 }, { "grad_norm": 0.38007447123527527, "learning_rate": 9.917936101677205e-05, "loss": 0.06, "step": 8390 }, { "grad_norm": 0.534784734249115, "learning_rate": 9.917562755223564e-05, "loss": 0.0637, "step": 8400 }, { "grad_norm": 0.7014754414558411, "learning_rate": 9.917188568492944e-05, "loss": 0.0685, "step": 8410 }, { "grad_norm": 0.7115383744239807, "learning_rate": 9.916813541549283e-05, "loss": 0.0586, "step": 8420 }, { "grad_norm": 0.45118144154548645, "learning_rate": 9.916437674456663e-05, "loss": 0.0572, "step": 8430 }, { "grad_norm": 0.6558455228805542, "learning_rate": 9.916060967279308e-05, "loss": 0.062, "step": 8440 }, { "grad_norm": 0.5045239329338074, "learning_rate": 9.91568342008159e-05, "loss": 0.0565, "step": 8450 }, { "grad_norm": 0.45776334404945374, "learning_rate": 9.915305032928019e-05, "loss": 0.0579, "step": 8460 }, { "grad_norm": 0.42451122403144836, "learning_rate": 9.914925805883253e-05, "loss": 0.0597, "step": 8470 }, { "grad_norm": 0.4580782651901245, "learning_rate": 9.914545739012088e-05, "loss": 0.0559, "step": 8480 }, { "grad_norm": 0.6724032163619995, "learning_rate": 9.91416483237947e-05, "loss": 0.0722, "step": 8490 }, { "grad_norm": 0.5578646659851074, "learning_rate": 9.913783086050485e-05, "loss": 0.0829, "step": 8500 }, { "grad_norm": 0.3799038827419281, "learning_rate": 9.913400500090364e-05, "loss": 0.0636, "step": 8510 }, { "grad_norm": 0.5316835641860962, "learning_rate": 9.913017074564479e-05, "loss": 0.0549, "step": 8520 }, { "grad_norm": 0.4375012516975403, "learning_rate": 9.912632809538348e-05, "loss": 0.064, "step": 8530 }, { "grad_norm": 0.500950813293457, "learning_rate": 9.912247705077629e-05, "loss": 0.0652, "step": 8540 }, { "grad_norm": 0.34826719760894775, "learning_rate": 9.911861761248127e-05, "loss": 0.0717, "step": 8550 }, { "grad_norm": 0.4815778136253357, "learning_rate": 9.91147497811579e-05, "loss": 0.0677, "step": 8560 }, { "grad_norm": 0.6736196875572205, "learning_rate": 9.911087355746709e-05, "loss": 0.0994, "step": 8570 }, { "grad_norm": 0.5797512531280518, "learning_rate": 9.910698894207117e-05, "loss": 0.0742, "step": 8580 }, { "grad_norm": 0.3693646490573883, "learning_rate": 9.910309593563392e-05, "loss": 0.0663, "step": 8590 }, { "grad_norm": 0.585914134979248, "learning_rate": 9.909919453882057e-05, "loss": 0.0551, "step": 8600 }, { "grad_norm": 0.6216839551925659, "learning_rate": 9.90952847522977e-05, "loss": 0.0568, "step": 8610 }, { "grad_norm": 0.5739368200302124, "learning_rate": 9.909136657673346e-05, "loss": 0.0704, "step": 8620 }, { "grad_norm": 0.43617165088653564, "learning_rate": 9.908744001279731e-05, "loss": 0.0569, "step": 8630 }, { "grad_norm": 0.6953545212745667, "learning_rate": 9.90835050611602e-05, "loss": 0.0701, "step": 8640 }, { "grad_norm": 0.5726929903030396, "learning_rate": 9.90795617224945e-05, "loss": 0.074, "step": 8650 }, { "grad_norm": 0.5626887679100037, "learning_rate": 9.907560999747405e-05, "loss": 0.0853, "step": 8660 }, { "grad_norm": 0.5825428366661072, "learning_rate": 9.907164988677408e-05, "loss": 0.0617, "step": 8670 }, { "grad_norm": 0.4063337445259094, "learning_rate": 9.906768139107124e-05, "loss": 0.0562, "step": 8680 }, { "grad_norm": 0.5876708030700684, "learning_rate": 9.906370451104367e-05, "loss": 0.0568, "step": 8690 }, { "grad_norm": 0.3373733162879944, "learning_rate": 9.905971924737088e-05, "loss": 0.0513, "step": 8700 }, { "grad_norm": 0.4748618006706238, "learning_rate": 9.905572560073387e-05, "loss": 0.067, "step": 8710 }, { "grad_norm": 0.7930983901023865, "learning_rate": 9.905172357181501e-05, "loss": 0.0795, "step": 8720 }, { "grad_norm": 0.6043055653572083, "learning_rate": 9.904771316129817e-05, "loss": 0.0631, "step": 8730 }, { "grad_norm": 0.442768394947052, "learning_rate": 9.904369436986862e-05, "loss": 0.0585, "step": 8740 }, { "grad_norm": 0.4923272430896759, "learning_rate": 9.903966719821303e-05, "loss": 0.0695, "step": 8750 }, { "grad_norm": 0.46637478470802307, "learning_rate": 9.903563164701956e-05, "loss": 0.062, "step": 8760 }, { "grad_norm": 0.6296637058258057, "learning_rate": 9.903158771697778e-05, "loss": 0.064, "step": 8770 }, { "grad_norm": 0.6915973424911499, "learning_rate": 9.902753540877867e-05, "loss": 0.088, "step": 8780 }, { "grad_norm": 0.4792475402355194, "learning_rate": 9.902347472311466e-05, "loss": 0.0643, "step": 8790 }, { "grad_norm": 0.4524202346801758, "learning_rate": 9.901940566067962e-05, "loss": 0.0499, "step": 8800 }, { "grad_norm": 0.595626950263977, "learning_rate": 9.901532822216883e-05, "loss": 0.0654, "step": 8810 }, { "grad_norm": 0.45025163888931274, "learning_rate": 9.901124240827904e-05, "loss": 0.0611, "step": 8820 }, { "grad_norm": 0.6021426320075989, "learning_rate": 9.900714821970835e-05, "loss": 0.0662, "step": 8830 }, { "grad_norm": 0.43518730998039246, "learning_rate": 9.900304565715641e-05, "loss": 0.0551, "step": 8840 }, { "grad_norm": 0.4227001368999481, "learning_rate": 9.899893472132419e-05, "loss": 0.0539, "step": 8850 }, { "grad_norm": 0.40061089396476746, "learning_rate": 9.899481541291415e-05, "loss": 0.0545, "step": 8860 }, { "grad_norm": 0.5859637260437012, "learning_rate": 9.899068773263016e-05, "loss": 0.0569, "step": 8870 }, { "grad_norm": 0.4237913489341736, "learning_rate": 9.898655168117754e-05, "loss": 0.0628, "step": 8880 }, { "grad_norm": 0.40029779076576233, "learning_rate": 9.898240725926302e-05, "loss": 0.0708, "step": 8890 }, { "grad_norm": 0.5398467183113098, "learning_rate": 9.897825446759478e-05, "loss": 0.0632, "step": 8900 }, { "grad_norm": 0.4764927327632904, "learning_rate": 9.897409330688241e-05, "loss": 0.0656, "step": 8910 }, { "grad_norm": 0.6567390561103821, "learning_rate": 9.896992377783692e-05, "loss": 0.0827, "step": 8920 }, { "grad_norm": 0.5169298052787781, "learning_rate": 9.89657458811708e-05, "loss": 0.0546, "step": 8930 }, { "grad_norm": 0.49452540278434753, "learning_rate": 9.896155961759792e-05, "loss": 0.0644, "step": 8940 }, { "grad_norm": 0.49178609251976013, "learning_rate": 9.895736498783361e-05, "loss": 0.0585, "step": 8950 }, { "grad_norm": 0.4403664469718933, "learning_rate": 9.895316199259462e-05, "loss": 0.0528, "step": 8960 }, { "grad_norm": 0.6041280627250671, "learning_rate": 9.894895063259909e-05, "loss": 0.0622, "step": 8970 }, { "grad_norm": 0.42254653573036194, "learning_rate": 9.894473090856667e-05, "loss": 0.0673, "step": 8980 }, { "grad_norm": 0.5655211806297302, "learning_rate": 9.894050282121839e-05, "loss": 0.082, "step": 8990 }, { "grad_norm": 0.5692877769470215, "learning_rate": 9.893626637127668e-05, "loss": 0.0635, "step": 9000 }, { "grad_norm": 0.4596206545829773, "learning_rate": 9.893202155946546e-05, "loss": 0.062, "step": 9010 }, { "grad_norm": 0.5390011668205261, "learning_rate": 9.892776838651006e-05, "loss": 0.0601, "step": 9020 }, { "grad_norm": 0.5449743270874023, "learning_rate": 9.892350685313722e-05, "loss": 0.0711, "step": 9030 }, { "grad_norm": 0.4052562713623047, "learning_rate": 9.891923696007513e-05, "loss": 0.0773, "step": 9040 }, { "grad_norm": 0.785422682762146, "learning_rate": 9.891495870805336e-05, "loss": 0.0645, "step": 9050 }, { "grad_norm": 0.5325893759727478, "learning_rate": 9.891067209780298e-05, "loss": 0.0777, "step": 9060 }, { "grad_norm": 0.38065826892852783, "learning_rate": 9.890637713005646e-05, "loss": 0.0541, "step": 9070 }, { "grad_norm": 0.44769465923309326, "learning_rate": 9.890207380554767e-05, "loss": 0.0577, "step": 9080 }, { "grad_norm": 0.4608071744441986, "learning_rate": 9.889776212501196e-05, "loss": 0.0756, "step": 9090 }, { "grad_norm": 0.720150351524353, "learning_rate": 9.889344208918605e-05, "loss": 0.0762, "step": 9100 }, { "grad_norm": 0.4952170252799988, "learning_rate": 9.888911369880812e-05, "loss": 0.0715, "step": 9110 }, { "grad_norm": 0.5531207919120789, "learning_rate": 9.888477695461777e-05, "loss": 0.0654, "step": 9120 }, { "grad_norm": 0.49839842319488525, "learning_rate": 9.888043185735607e-05, "loss": 0.0756, "step": 9130 }, { "grad_norm": 0.4412424862384796, "learning_rate": 9.887607840776542e-05, "loss": 0.0597, "step": 9140 }, { "grad_norm": 0.5811328291893005, "learning_rate": 9.887171660658975e-05, "loss": 0.0734, "step": 9150 }, { "grad_norm": 0.5245799422264099, "learning_rate": 9.886734645457435e-05, "loss": 0.0613, "step": 9160 }, { "grad_norm": 0.47897326946258545, "learning_rate": 9.886296795246597e-05, "loss": 0.056, "step": 9170 }, { "grad_norm": 0.4557494819164276, "learning_rate": 9.885858110101276e-05, "loss": 0.0562, "step": 9180 }, { "grad_norm": 0.41418886184692383, "learning_rate": 9.885418590096434e-05, "loss": 0.0533, "step": 9190 }, { "grad_norm": 0.48035290837287903, "learning_rate": 9.88497823530717e-05, "loss": 0.0605, "step": 9200 }, { "grad_norm": 0.40429583191871643, "learning_rate": 9.884537045808732e-05, "loss": 0.0742, "step": 9210 }, { "grad_norm": 0.5421463251113892, "learning_rate": 9.884095021676502e-05, "loss": 0.0711, "step": 9220 }, { "grad_norm": 0.4536348879337311, "learning_rate": 9.883652162986017e-05, "loss": 0.0647, "step": 9230 }, { "grad_norm": 0.3588198721408844, "learning_rate": 9.883208469812943e-05, "loss": 0.067, "step": 9240 }, { "grad_norm": 0.3618038594722748, "learning_rate": 9.882763942233098e-05, "loss": 0.051, "step": 9250 }, { "grad_norm": 0.42757168412208557, "learning_rate": 9.882318580322441e-05, "loss": 0.0491, "step": 9260 }, { "grad_norm": 0.4420641362667084, "learning_rate": 9.881872384157067e-05, "loss": 0.0631, "step": 9270 }, { "grad_norm": 0.38514840602874756, "learning_rate": 9.881425353813225e-05, "loss": 0.0623, "step": 9280 }, { "grad_norm": 0.7620610594749451, "learning_rate": 9.880977489367296e-05, "loss": 0.0822, "step": 9290 }, { "grad_norm": 0.47597646713256836, "learning_rate": 9.88052879089581e-05, "loss": 0.0658, "step": 9300 }, { "grad_norm": 0.5928020477294922, "learning_rate": 9.880079258475434e-05, "loss": 0.064, "step": 9310 }, { "grad_norm": 0.434081107378006, "learning_rate": 9.879628892182985e-05, "loss": 0.0855, "step": 9320 }, { "grad_norm": 0.39512428641319275, "learning_rate": 9.879177692095416e-05, "loss": 0.0559, "step": 9330 }, { "grad_norm": 0.39920079708099365, "learning_rate": 9.878725658289825e-05, "loss": 0.0676, "step": 9340 }, { "grad_norm": 0.4783356785774231, "learning_rate": 9.878272790843454e-05, "loss": 0.0627, "step": 9350 }, { "grad_norm": 0.4656180441379547, "learning_rate": 9.877819089833682e-05, "loss": 0.0628, "step": 9360 }, { "grad_norm": 0.5020067095756531, "learning_rate": 9.877364555338038e-05, "loss": 0.0742, "step": 9370 }, { "grad_norm": 0.4161931276321411, "learning_rate": 9.876909187434186e-05, "loss": 0.0678, "step": 9380 }, { "grad_norm": 0.497624933719635, "learning_rate": 9.876452986199939e-05, "loss": 0.0596, "step": 9390 }, { "grad_norm": 0.5174853205680847, "learning_rate": 9.875995951713248e-05, "loss": 0.0857, "step": 9400 }, { "grad_norm": 0.637118399143219, "learning_rate": 9.875538084052207e-05, "loss": 0.0856, "step": 9410 }, { "grad_norm": 0.7450485229492188, "learning_rate": 9.875079383295053e-05, "loss": 0.066, "step": 9420 }, { "grad_norm": 0.4007801413536072, "learning_rate": 9.874619849520167e-05, "loss": 0.0592, "step": 9430 }, { "grad_norm": 0.6661776900291443, "learning_rate": 9.874159482806069e-05, "loss": 0.0605, "step": 9440 }, { "grad_norm": 0.5268365144729614, "learning_rate": 9.873698283231426e-05, "loss": 0.0584, "step": 9450 }, { "grad_norm": 0.46512189507484436, "learning_rate": 9.87323625087504e-05, "loss": 0.0605, "step": 9460 }, { "grad_norm": 0.4536266624927521, "learning_rate": 9.872773385815863e-05, "loss": 0.0735, "step": 9470 }, { "grad_norm": 0.38692399859428406, "learning_rate": 9.872309688132986e-05, "loss": 0.0618, "step": 9480 }, { "grad_norm": 0.4433746635913849, "learning_rate": 9.871845157905639e-05, "loss": 0.0558, "step": 9490 }, { "grad_norm": 0.37469378113746643, "learning_rate": 9.871379795213201e-05, "loss": 0.0514, "step": 9500 }, { "grad_norm": 0.45917531847953796, "learning_rate": 9.87091360013519e-05, "loss": 0.0613, "step": 9510 }, { "grad_norm": 0.5467556715011597, "learning_rate": 9.870446572751262e-05, "loss": 0.0608, "step": 9520 }, { "grad_norm": 0.4500690698623657, "learning_rate": 9.869978713141224e-05, "loss": 0.0669, "step": 9530 }, { "grad_norm": 0.9199455976486206, "learning_rate": 9.869510021385016e-05, "loss": 0.069, "step": 9540 }, { "grad_norm": 0.3145845830440521, "learning_rate": 9.869040497562727e-05, "loss": 0.056, "step": 9550 }, { "grad_norm": 0.5159204602241516, "learning_rate": 9.868570141754587e-05, "loss": 0.0647, "step": 9560 }, { "grad_norm": 0.4528343081474304, "learning_rate": 9.868098954040965e-05, "loss": 0.0756, "step": 9570 }, { "grad_norm": 0.5186842679977417, "learning_rate": 9.867626934502374e-05, "loss": 0.0693, "step": 9580 }, { "grad_norm": 0.41486817598342896, "learning_rate": 9.86715408321947e-05, "loss": 0.0608, "step": 9590 }, { "grad_norm": 0.516328752040863, "learning_rate": 9.86668040027305e-05, "loss": 0.0603, "step": 9600 }, { "grad_norm": 0.6512332558631897, "learning_rate": 9.866205885744053e-05, "loss": 0.0568, "step": 9610 }, { "grad_norm": 0.4132401943206787, "learning_rate": 9.865730539713563e-05, "loss": 0.0585, "step": 9620 }, { "grad_norm": 0.3265419900417328, "learning_rate": 9.8652543622628e-05, "loss": 0.0538, "step": 9630 }, { "grad_norm": 0.5123006701469421, "learning_rate": 9.864777353473132e-05, "loss": 0.0627, "step": 9640 }, { "grad_norm": 0.47589895129203796, "learning_rate": 9.864299513426068e-05, "loss": 0.0655, "step": 9650 }, { "grad_norm": 0.67950439453125, "learning_rate": 9.863820842203254e-05, "loss": 0.0697, "step": 9660 }, { "grad_norm": 0.3933061957359314, "learning_rate": 9.863341339886483e-05, "loss": 0.0618, "step": 9670 }, { "grad_norm": 0.499762624502182, "learning_rate": 9.86286100655769e-05, "loss": 0.0501, "step": 9680 }, { "grad_norm": 0.6228466629981995, "learning_rate": 9.862379842298953e-05, "loss": 0.0652, "step": 9690 }, { "grad_norm": 0.4400955140590668, "learning_rate": 9.861897847192485e-05, "loss": 0.0554, "step": 9700 }, { "grad_norm": 0.5880815386772156, "learning_rate": 9.86141502132065e-05, "loss": 0.058, "step": 9710 }, { "grad_norm": 0.6256262063980103, "learning_rate": 9.860931364765946e-05, "loss": 0.0652, "step": 9720 }, { "grad_norm": 0.5001572370529175, "learning_rate": 9.860446877611021e-05, "loss": 0.0741, "step": 9730 }, { "grad_norm": 0.6684988737106323, "learning_rate": 9.859961559938655e-05, "loss": 0.0609, "step": 9740 }, { "grad_norm": 0.7348450422286987, "learning_rate": 9.85947541183178e-05, "loss": 0.0673, "step": 9750 }, { "grad_norm": 0.45130687952041626, "learning_rate": 9.858988433373463e-05, "loss": 0.0682, "step": 9760 }, { "grad_norm": 0.5196598768234253, "learning_rate": 9.858500624646918e-05, "loss": 0.0658, "step": 9770 }, { "grad_norm": 0.656175971031189, "learning_rate": 9.858011985735497e-05, "loss": 0.0656, "step": 9780 }, { "grad_norm": 0.5591351985931396, "learning_rate": 9.857522516722693e-05, "loss": 0.0629, "step": 9790 }, { "grad_norm": 0.43958625197410583, "learning_rate": 9.857032217692145e-05, "loss": 0.0593, "step": 9800 }, { "grad_norm": 0.6336467862129211, "learning_rate": 9.856541088727631e-05, "loss": 0.0649, "step": 9810 }, { "grad_norm": 0.4294448792934418, "learning_rate": 9.856049129913072e-05, "loss": 0.0651, "step": 9820 }, { "grad_norm": 0.5915256142616272, "learning_rate": 9.85555634133253e-05, "loss": 0.0767, "step": 9830 }, { "grad_norm": 0.4371525049209595, "learning_rate": 9.855062723070208e-05, "loss": 0.0851, "step": 9840 }, { "grad_norm": 0.5555716156959534, "learning_rate": 9.854568275210454e-05, "loss": 0.0829, "step": 9850 }, { "grad_norm": 0.5306034088134766, "learning_rate": 9.854072997837754e-05, "loss": 0.0551, "step": 9860 }, { "grad_norm": 0.32513147592544556, "learning_rate": 9.853576891036737e-05, "loss": 0.0556, "step": 9870 }, { "grad_norm": 0.5176831483840942, "learning_rate": 9.853079954892177e-05, "loss": 0.0714, "step": 9880 }, { "grad_norm": 0.7759354114532471, "learning_rate": 9.852582189488983e-05, "loss": 0.0637, "step": 9890 }, { "grad_norm": 0.49148139357566833, "learning_rate": 9.852083594912212e-05, "loss": 0.0738, "step": 9900 }, { "grad_norm": 0.47106078267097473, "learning_rate": 9.851584171247058e-05, "loss": 0.0637, "step": 9910 }, { "grad_norm": 0.6506679058074951, "learning_rate": 9.851083918578863e-05, "loss": 0.0733, "step": 9920 }, { "grad_norm": 0.5868524312973022, "learning_rate": 9.850582836993103e-05, "loss": 0.0622, "step": 9930 }, { "grad_norm": 0.47910556197166443, "learning_rate": 9.850080926575397e-05, "loss": 0.0618, "step": 9940 }, { "grad_norm": 0.6428040862083435, "learning_rate": 9.849578187411515e-05, "loss": 0.057, "step": 9950 }, { "grad_norm": 0.5703024864196777, "learning_rate": 9.849074619587354e-05, "loss": 0.0598, "step": 9960 }, { "grad_norm": 0.5455280542373657, "learning_rate": 9.848570223188964e-05, "loss": 0.0726, "step": 9970 }, { "grad_norm": 0.48333480954170227, "learning_rate": 9.848064998302531e-05, "loss": 0.0638, "step": 9980 }, { "grad_norm": 0.5567018985748291, "learning_rate": 9.847558945014386e-05, "loss": 0.0706, "step": 9990 }, { "grad_norm": 0.5965126752853394, "learning_rate": 9.847052063410996e-05, "loss": 0.077, "step": 10000 }, { "grad_norm": 0.5180984139442444, "learning_rate": 9.846544353578977e-05, "loss": 0.0771, "step": 10010 }, { "grad_norm": 0.5300484895706177, "learning_rate": 9.846035815605081e-05, "loss": 0.0775, "step": 10020 }, { "grad_norm": 0.6422309875488281, "learning_rate": 9.845526449576204e-05, "loss": 0.0626, "step": 10030 }, { "grad_norm": 0.6586577892303467, "learning_rate": 9.845016255579383e-05, "loss": 0.0789, "step": 10040 }, { "grad_norm": 0.4619106948375702, "learning_rate": 9.844505233701794e-05, "loss": 0.0578, "step": 10050 }, { "grad_norm": 0.5428905487060547, "learning_rate": 9.843993384030757e-05, "loss": 0.0625, "step": 10060 }, { "grad_norm": 0.3995421826839447, "learning_rate": 9.843480706653737e-05, "loss": 0.0503, "step": 10070 }, { "grad_norm": 0.3730994164943695, "learning_rate": 9.84296720165833e-05, "loss": 0.0583, "step": 10080 }, { "grad_norm": 0.4261143207550049, "learning_rate": 9.842452869132286e-05, "loss": 0.0588, "step": 10090 }, { "grad_norm": 0.5929514765739441, "learning_rate": 9.841937709163489e-05, "loss": 0.0613, "step": 10100 }, { "grad_norm": 0.4754594564437866, "learning_rate": 9.841421721839962e-05, "loss": 0.0483, "step": 10110 }, { "grad_norm": 1.2748106718063354, "learning_rate": 9.840904907249879e-05, "loss": 0.0623, "step": 10120 }, { "grad_norm": 0.694590151309967, "learning_rate": 9.840387265481545e-05, "loss": 0.0657, "step": 10130 }, { "grad_norm": 0.5569397211074829, "learning_rate": 9.839868796623411e-05, "loss": 0.0591, "step": 10140 }, { "grad_norm": 0.4927350580692291, "learning_rate": 9.839349500764072e-05, "loss": 0.075, "step": 10150 }, { "grad_norm": 0.5687325596809387, "learning_rate": 9.83882937799226e-05, "loss": 0.0681, "step": 10160 }, { "grad_norm": 0.42004942893981934, "learning_rate": 9.838308428396849e-05, "loss": 0.0646, "step": 10170 }, { "grad_norm": 0.5871571898460388, "learning_rate": 9.837786652066854e-05, "loss": 0.0761, "step": 10180 }, { "grad_norm": 0.6849088668823242, "learning_rate": 9.837264049091437e-05, "loss": 0.0868, "step": 10190 }, { "grad_norm": 1.0205074548721313, "learning_rate": 9.836740619559893e-05, "loss": 0.0713, "step": 10200 }, { "grad_norm": 0.5473470091819763, "learning_rate": 9.836216363561659e-05, "loss": 0.0685, "step": 10210 }, { "grad_norm": 0.5454677939414978, "learning_rate": 9.835691281186322e-05, "loss": 0.0635, "step": 10220 }, { "grad_norm": 0.5457859635353088, "learning_rate": 9.8351653725236e-05, "loss": 0.0623, "step": 10230 }, { "grad_norm": 0.4843190908432007, "learning_rate": 9.83463863766336e-05, "loss": 0.0718, "step": 10240 }, { "grad_norm": 0.43605875968933105, "learning_rate": 9.834111076695602e-05, "loss": 0.0685, "step": 10250 }, { "grad_norm": 0.45000606775283813, "learning_rate": 9.833582689710477e-05, "loss": 0.0517, "step": 10260 }, { "grad_norm": 0.588458776473999, "learning_rate": 9.833053476798268e-05, "loss": 0.0607, "step": 10270 }, { "grad_norm": 0.6555659770965576, "learning_rate": 9.832523438049404e-05, "loss": 0.065, "step": 10280 }, { "grad_norm": 0.4476284682750702, "learning_rate": 9.831992573554454e-05, "loss": 0.0678, "step": 10290 }, { "grad_norm": 0.5106391906738281, "learning_rate": 9.831460883404128e-05, "loss": 0.0528, "step": 10300 }, { "grad_norm": 0.5887622237205505, "learning_rate": 9.830928367689278e-05, "loss": 0.0552, "step": 10310 }, { "grad_norm": 0.5064799785614014, "learning_rate": 9.830395026500896e-05, "loss": 0.0619, "step": 10320 }, { "grad_norm": 0.6158971190452576, "learning_rate": 9.829860859930115e-05, "loss": 0.0596, "step": 10330 }, { "grad_norm": 0.5287237167358398, "learning_rate": 9.829325868068212e-05, "loss": 0.066, "step": 10340 }, { "grad_norm": 0.5247213840484619, "learning_rate": 9.8287900510066e-05, "loss": 0.0659, "step": 10350 }, { "grad_norm": 0.6267023682594299, "learning_rate": 9.828253408836834e-05, "loss": 0.0611, "step": 10360 }, { "grad_norm": 0.46759477257728577, "learning_rate": 9.827715941650615e-05, "loss": 0.0594, "step": 10370 }, { "grad_norm": 0.4293432831764221, "learning_rate": 9.82717764953978e-05, "loss": 0.0665, "step": 10380 }, { "grad_norm": 0.5310377478599548, "learning_rate": 9.826638532596308e-05, "loss": 0.0633, "step": 10390 }, { "grad_norm": 0.5136206746101379, "learning_rate": 9.82609859091232e-05, "loss": 0.0633, "step": 10400 }, { "grad_norm": 0.41305679082870483, "learning_rate": 9.825557824580076e-05, "loss": 0.0634, "step": 10410 }, { "grad_norm": 0.6039775013923645, "learning_rate": 9.82501623369198e-05, "loss": 0.0735, "step": 10420 }, { "grad_norm": 0.4808042645454407, "learning_rate": 9.824473818340574e-05, "loss": 0.0533, "step": 10430 }, { "grad_norm": 0.49461978673934937, "learning_rate": 9.823930578618541e-05, "loss": 0.0558, "step": 10440 }, { "grad_norm": 0.5831822752952576, "learning_rate": 9.823386514618709e-05, "loss": 0.0591, "step": 10450 }, { "grad_norm": 0.5753390192985535, "learning_rate": 9.82284162643404e-05, "loss": 0.059, "step": 10460 }, { "grad_norm": 0.44596534967422485, "learning_rate": 9.822295914157642e-05, "loss": 0.0604, "step": 10470 }, { "grad_norm": 0.41425812244415283, "learning_rate": 9.821749377882763e-05, "loss": 0.0587, "step": 10480 }, { "grad_norm": 0.44457441568374634, "learning_rate": 9.821202017702791e-05, "loss": 0.0512, "step": 10490 }, { "grad_norm": 0.47280678153038025, "learning_rate": 9.820653833711253e-05, "loss": 0.0516, "step": 10500 }, { "grad_norm": 0.4988566040992737, "learning_rate": 9.820104826001822e-05, "loss": 0.0618, "step": 10510 }, { "grad_norm": 0.47752541303634644, "learning_rate": 9.819554994668305e-05, "loss": 0.0531, "step": 10520 }, { "grad_norm": 0.41572484374046326, "learning_rate": 9.819004339804654e-05, "loss": 0.0527, "step": 10530 }, { "grad_norm": 0.4024781286716461, "learning_rate": 9.818452861504961e-05, "loss": 0.0522, "step": 10540 }, { "grad_norm": 0.448723703622818, "learning_rate": 9.81790055986346e-05, "loss": 0.0514, "step": 10550 }, { "grad_norm": 0.3943040668964386, "learning_rate": 9.817347434974523e-05, "loss": 0.0552, "step": 10560 }, { "grad_norm": 0.4291345477104187, "learning_rate": 9.816793486932664e-05, "loss": 0.0496, "step": 10570 }, { "grad_norm": 0.4299463927745819, "learning_rate": 9.816238715832538e-05, "loss": 0.061, "step": 10580 }, { "grad_norm": 0.4492337703704834, "learning_rate": 9.815683121768939e-05, "loss": 0.064, "step": 10590 }, { "grad_norm": 0.749173104763031, "learning_rate": 9.815126704836804e-05, "loss": 0.0714, "step": 10600 }, { "grad_norm": 0.4792734980583191, "learning_rate": 9.81456946513121e-05, "loss": 0.0871, "step": 10610 }, { "grad_norm": 0.6209853291511536, "learning_rate": 9.814011402747373e-05, "loss": 0.0852, "step": 10620 }, { "grad_norm": 0.5281957387924194, "learning_rate": 9.813452517780651e-05, "loss": 0.0631, "step": 10630 }, { "grad_norm": 0.31516972184181213, "learning_rate": 9.81289281032654e-05, "loss": 0.0623, "step": 10640 }, { "grad_norm": 0.6695922017097473, "learning_rate": 9.812332280480683e-05, "loss": 0.0748, "step": 10650 }, { "grad_norm": 0.5147625803947449, "learning_rate": 9.811770928338854e-05, "loss": 0.0725, "step": 10660 }, { "grad_norm": 0.5144379734992981, "learning_rate": 9.811208753996979e-05, "loss": 0.0704, "step": 10670 }, { "grad_norm": 0.3968760669231415, "learning_rate": 9.810645757551113e-05, "loss": 0.0525, "step": 10680 }, { "grad_norm": 0.4393942654132843, "learning_rate": 9.810081939097459e-05, "loss": 0.0518, "step": 10690 }, { "grad_norm": 0.5751966238021851, "learning_rate": 9.809517298732356e-05, "loss": 0.082, "step": 10700 }, { "grad_norm": 0.519834578037262, "learning_rate": 9.80895183655229e-05, "loss": 0.0566, "step": 10710 }, { "grad_norm": 0.5500138998031616, "learning_rate": 9.808385552653877e-05, "loss": 0.0557, "step": 10720 }, { "grad_norm": 0.4865476191043854, "learning_rate": 9.807818447133886e-05, "loss": 0.0533, "step": 10730 }, { "grad_norm": 0.509782075881958, "learning_rate": 9.807250520089215e-05, "loss": 0.0594, "step": 10740 }, { "grad_norm": 0.5968432426452637, "learning_rate": 9.806681771616908e-05, "loss": 0.0637, "step": 10750 }, { "grad_norm": 0.4821838438510895, "learning_rate": 9.80611220181415e-05, "loss": 0.067, "step": 10760 }, { "grad_norm": 0.5048344731330872, "learning_rate": 9.805541810778264e-05, "loss": 0.0626, "step": 10770 }, { "grad_norm": 0.5622633695602417, "learning_rate": 9.804970598606716e-05, "loss": 0.0699, "step": 10780 }, { "grad_norm": 0.690930187702179, "learning_rate": 9.804398565397106e-05, "loss": 0.0753, "step": 10790 }, { "grad_norm": 0.43870431184768677, "learning_rate": 9.803825711247183e-05, "loss": 0.0674, "step": 10800 }, { "grad_norm": 0.4661857783794403, "learning_rate": 9.803252036254831e-05, "loss": 0.0607, "step": 10810 }, { "grad_norm": 0.5929409265518188, "learning_rate": 9.802677540518076e-05, "loss": 0.0729, "step": 10820 }, { "grad_norm": 0.49231410026550293, "learning_rate": 9.802102224135081e-05, "loss": 0.0657, "step": 10830 }, { "grad_norm": 0.4993318021297455, "learning_rate": 9.801526087204155e-05, "loss": 0.0799, "step": 10840 }, { "grad_norm": 0.5323322415351868, "learning_rate": 9.800949129823743e-05, "loss": 0.0686, "step": 10850 }, { "grad_norm": 0.43572309613227844, "learning_rate": 9.80037135209243e-05, "loss": 0.0562, "step": 10860 }, { "grad_norm": 0.5533536672592163, "learning_rate": 9.799792754108946e-05, "loss": 0.0749, "step": 10870 }, { "grad_norm": 0.4710093140602112, "learning_rate": 9.799213335972152e-05, "loss": 0.0687, "step": 10880 }, { "grad_norm": 0.44611993432044983, "learning_rate": 9.798633097781058e-05, "loss": 0.071, "step": 10890 }, { "grad_norm": 0.43890678882598877, "learning_rate": 9.79805203963481e-05, "loss": 0.0665, "step": 10900 }, { "grad_norm": 0.6271626353263855, "learning_rate": 9.797470161632697e-05, "loss": 0.0698, "step": 10910 }, { "grad_norm": 0.5212019681930542, "learning_rate": 9.796887463874145e-05, "loss": 0.0747, "step": 10920 }, { "grad_norm": 0.44313105940818787, "learning_rate": 9.796303946458718e-05, "loss": 0.0664, "step": 10930 }, { "grad_norm": 0.5327914953231812, "learning_rate": 9.795719609486127e-05, "loss": 0.0726, "step": 10940 }, { "grad_norm": 0.594642698764801, "learning_rate": 9.795134453056219e-05, "loss": 0.0504, "step": 10950 }, { "grad_norm": 0.4590568542480469, "learning_rate": 9.794548477268979e-05, "loss": 0.0503, "step": 10960 }, { "grad_norm": 0.4892662763595581, "learning_rate": 9.793961682224537e-05, "loss": 0.0475, "step": 10970 }, { "grad_norm": 0.49132028222084045, "learning_rate": 9.793374068023156e-05, "loss": 0.06, "step": 10980 }, { "grad_norm": 0.5324029326438904, "learning_rate": 9.792785634765247e-05, "loss": 0.0636, "step": 10990 }, { "grad_norm": 0.42588353157043457, "learning_rate": 9.792196382551357e-05, "loss": 0.0556, "step": 11000 }, { "grad_norm": 0.4224613308906555, "learning_rate": 9.791606311482171e-05, "loss": 0.0653, "step": 11010 }, { "grad_norm": 0.595743715763092, "learning_rate": 9.791015421658518e-05, "loss": 0.0611, "step": 11020 }, { "grad_norm": 0.5869330167770386, "learning_rate": 9.790423713181362e-05, "loss": 0.06, "step": 11030 }, { "grad_norm": 0.49493807554244995, "learning_rate": 9.789831186151814e-05, "loss": 0.0643, "step": 11040 }, { "grad_norm": 0.5655615925788879, "learning_rate": 9.789237840671118e-05, "loss": 0.0561, "step": 11050 }, { "grad_norm": 0.4847574234008789, "learning_rate": 9.78864367684066e-05, "loss": 0.0632, "step": 11060 }, { "grad_norm": 0.5882359147071838, "learning_rate": 9.788048694761968e-05, "loss": 0.0579, "step": 11070 }, { "grad_norm": 0.6066553592681885, "learning_rate": 9.787452894536709e-05, "loss": 0.0601, "step": 11080 }, { "grad_norm": 0.42101427912712097, "learning_rate": 9.786856276266685e-05, "loss": 0.0586, "step": 11090 }, { "grad_norm": 0.6356210708618164, "learning_rate": 9.786258840053845e-05, "loss": 0.0679, "step": 11100 }, { "grad_norm": 0.5125594139099121, "learning_rate": 9.785660586000273e-05, "loss": 0.0684, "step": 11110 }, { "grad_norm": 0.37218865752220154, "learning_rate": 9.785061514208196e-05, "loss": 0.0661, "step": 11120 }, { "grad_norm": 0.55350661277771, "learning_rate": 9.784461624779977e-05, "loss": 0.0747, "step": 11130 }, { "grad_norm": 0.49050289392471313, "learning_rate": 9.783860917818123e-05, "loss": 0.0588, "step": 11140 }, { "grad_norm": 0.5303327441215515, "learning_rate": 9.783259393425277e-05, "loss": 0.047, "step": 11150 }, { "grad_norm": 0.5319327712059021, "learning_rate": 9.782657051704221e-05, "loss": 0.052, "step": 11160 }, { "grad_norm": 0.506359875202179, "learning_rate": 9.782053892757883e-05, "loss": 0.062, "step": 11170 }, { "grad_norm": 0.5431830883026123, "learning_rate": 9.781449916689324e-05, "loss": 0.0622, "step": 11180 }, { "grad_norm": 0.4419137239456177, "learning_rate": 9.780845123601746e-05, "loss": 0.0592, "step": 11190 }, { "grad_norm": 0.4393882751464844, "learning_rate": 9.780239513598492e-05, "loss": 0.0637, "step": 11200 }, { "grad_norm": 0.6273784041404724, "learning_rate": 9.779633086783047e-05, "loss": 0.0694, "step": 11210 }, { "grad_norm": 0.5023050308227539, "learning_rate": 9.779025843259031e-05, "loss": 0.0588, "step": 11220 }, { "grad_norm": 0.5246263742446899, "learning_rate": 9.778417783130204e-05, "loss": 0.0717, "step": 11230 }, { "grad_norm": 0.4633031189441681, "learning_rate": 9.777808906500468e-05, "loss": 0.0598, "step": 11240 }, { "grad_norm": 0.5911616086959839, "learning_rate": 9.777199213473862e-05, "loss": 0.0515, "step": 11250 }, { "grad_norm": 0.429211288690567, "learning_rate": 9.77658870415457e-05, "loss": 0.0472, "step": 11260 }, { "grad_norm": 0.45774686336517334, "learning_rate": 9.775977378646906e-05, "loss": 0.0672, "step": 11270 }, { "grad_norm": 0.858040452003479, "learning_rate": 9.775365237055331e-05, "loss": 0.0775, "step": 11280 }, { "grad_norm": 0.407552033662796, "learning_rate": 9.774752279484445e-05, "loss": 0.0613, "step": 11290 }, { "grad_norm": 0.5086733102798462, "learning_rate": 9.774138506038984e-05, "loss": 0.0592, "step": 11300 }, { "grad_norm": 0.5339058041572571, "learning_rate": 9.773523916823826e-05, "loss": 0.0615, "step": 11310 }, { "grad_norm": 0.497771292924881, "learning_rate": 9.772908511943986e-05, "loss": 0.0502, "step": 11320 }, { "grad_norm": 0.49788808822631836, "learning_rate": 9.77229229150462e-05, "loss": 0.0514, "step": 11330 }, { "grad_norm": 0.5386170744895935, "learning_rate": 9.771675255611024e-05, "loss": 0.0434, "step": 11340 }, { "grad_norm": 0.4192185401916504, "learning_rate": 9.771057404368632e-05, "loss": 0.0687, "step": 11350 }, { "grad_norm": 0.42757776379585266, "learning_rate": 9.770438737883018e-05, "loss": 0.0514, "step": 11360 }, { "grad_norm": 0.5025882720947266, "learning_rate": 9.769819256259898e-05, "loss": 0.0442, "step": 11370 }, { "grad_norm": 0.41335198283195496, "learning_rate": 9.769198959605119e-05, "loss": 0.0531, "step": 11380 }, { "grad_norm": 0.4695151448249817, "learning_rate": 9.768577848024678e-05, "loss": 0.0627, "step": 11390 }, { "grad_norm": 0.5191007256507874, "learning_rate": 9.767955921624702e-05, "loss": 0.0697, "step": 11400 }, { "grad_norm": 0.5901297330856323, "learning_rate": 9.767333180511465e-05, "loss": 0.0525, "step": 11410 }, { "grad_norm": 0.4690435230731964, "learning_rate": 9.766709624791373e-05, "loss": 0.0708, "step": 11420 }, { "grad_norm": 0.47630494832992554, "learning_rate": 9.766085254570975e-05, "loss": 0.0581, "step": 11430 }, { "grad_norm": 0.5304346680641174, "learning_rate": 9.76546006995696e-05, "loss": 0.0537, "step": 11440 }, { "grad_norm": 0.5889133810997009, "learning_rate": 9.764834071056155e-05, "loss": 0.0598, "step": 11450 }, { "grad_norm": 0.44881671667099, "learning_rate": 9.764207257975526e-05, "loss": 0.0586, "step": 11460 }, { "grad_norm": 0.6242299675941467, "learning_rate": 9.763579630822179e-05, "loss": 0.0495, "step": 11470 }, { "grad_norm": 0.5192617177963257, "learning_rate": 9.762951189703356e-05, "loss": 0.051, "step": 11480 }, { "grad_norm": 0.5755698680877686, "learning_rate": 9.762321934726442e-05, "loss": 0.0605, "step": 11490 }, { "grad_norm": 0.4618361294269562, "learning_rate": 9.761691865998959e-05, "loss": 0.0562, "step": 11500 }, { "grad_norm": 0.5291464924812317, "learning_rate": 9.76106098362857e-05, "loss": 0.0526, "step": 11510 }, { "grad_norm": 0.38015416264533997, "learning_rate": 9.760429287723072e-05, "loss": 0.0595, "step": 11520 }, { "grad_norm": 0.46259573101997375, "learning_rate": 9.759796778390406e-05, "loss": 0.056, "step": 11530 }, { "grad_norm": 0.4486137628555298, "learning_rate": 9.759163455738653e-05, "loss": 0.0636, "step": 11540 }, { "grad_norm": 0.5499234199523926, "learning_rate": 9.75852931987603e-05, "loss": 0.0637, "step": 11550 }, { "grad_norm": 0.3662637174129486, "learning_rate": 9.757894370910891e-05, "loss": 0.0569, "step": 11560 }, { "grad_norm": 0.5378787517547607, "learning_rate": 9.757258608951733e-05, "loss": 0.0552, "step": 11570 }, { "grad_norm": 0.5958003997802734, "learning_rate": 9.75662203410719e-05, "loss": 0.0629, "step": 11580 }, { "grad_norm": 0.5546146035194397, "learning_rate": 9.755984646486034e-05, "loss": 0.0724, "step": 11590 }, { "grad_norm": 0.4766956567764282, "learning_rate": 9.75534644619718e-05, "loss": 0.0585, "step": 11600 }, { "grad_norm": 0.43782296776771545, "learning_rate": 9.754707433349676e-05, "loss": 0.0616, "step": 11610 }, { "grad_norm": 0.6836637854576111, "learning_rate": 9.754067608052715e-05, "loss": 0.0712, "step": 11620 }, { "grad_norm": 0.5777985453605652, "learning_rate": 9.753426970415622e-05, "loss": 0.0845, "step": 11630 }, { "grad_norm": 0.4155181348323822, "learning_rate": 9.752785520547868e-05, "loss": 0.0577, "step": 11640 }, { "grad_norm": 0.44485220313072205, "learning_rate": 9.752143258559056e-05, "loss": 0.0579, "step": 11650 }, { "grad_norm": 0.5821259021759033, "learning_rate": 9.751500184558933e-05, "loss": 0.0596, "step": 11660 }, { "grad_norm": 0.3251696825027466, "learning_rate": 9.750856298657383e-05, "loss": 0.0521, "step": 11670 }, { "grad_norm": 0.47691407799720764, "learning_rate": 9.750211600964428e-05, "loss": 0.0632, "step": 11680 }, { "grad_norm": 0.5529038310050964, "learning_rate": 9.749566091590226e-05, "loss": 0.0519, "step": 11690 }, { "grad_norm": 0.37648913264274597, "learning_rate": 9.748919770645083e-05, "loss": 0.0581, "step": 11700 }, { "grad_norm": 0.5654134750366211, "learning_rate": 9.748272638239432e-05, "loss": 0.065, "step": 11710 }, { "grad_norm": 0.5885326862335205, "learning_rate": 9.747624694483855e-05, "loss": 0.0586, "step": 11720 }, { "grad_norm": 0.5498083233833313, "learning_rate": 9.746975939489065e-05, "loss": 0.0492, "step": 11730 }, { "grad_norm": 0.5417864918708801, "learning_rate": 9.746326373365918e-05, "loss": 0.0618, "step": 11740 }, { "grad_norm": 0.40801793336868286, "learning_rate": 9.745675996225403e-05, "loss": 0.0559, "step": 11750 }, { "grad_norm": 0.4708767533302307, "learning_rate": 9.745024808178657e-05, "loss": 0.0464, "step": 11760 }, { "grad_norm": 0.42166227102279663, "learning_rate": 9.744372809336947e-05, "loss": 0.0581, "step": 11770 }, { "grad_norm": 0.5163028836250305, "learning_rate": 9.743719999811682e-05, "loss": 0.0526, "step": 11780 }, { "grad_norm": 0.7244133949279785, "learning_rate": 9.743066379714412e-05, "loss": 0.0527, "step": 11790 }, { "grad_norm": 0.3545202910900116, "learning_rate": 9.74241194915682e-05, "loss": 0.0512, "step": 11800 }, { "grad_norm": 0.4921993017196655, "learning_rate": 9.741756708250731e-05, "loss": 0.0684, "step": 11810 }, { "grad_norm": 0.4162862300872803, "learning_rate": 9.741100657108109e-05, "loss": 0.058, "step": 11820 }, { "grad_norm": 0.5039846301078796, "learning_rate": 9.740443795841054e-05, "loss": 0.0628, "step": 11830 }, { "grad_norm": 0.4211704730987549, "learning_rate": 9.739786124561805e-05, "loss": 0.0491, "step": 11840 }, { "grad_norm": 0.42677196860313416, "learning_rate": 9.73912764338274e-05, "loss": 0.0494, "step": 11850 }, { "grad_norm": 0.41602060198783875, "learning_rate": 9.738468352416377e-05, "loss": 0.0627, "step": 11860 }, { "grad_norm": 0.5213252902030945, "learning_rate": 9.737808251775369e-05, "loss": 0.0508, "step": 11870 }, { "grad_norm": 0.34682348370552063, "learning_rate": 9.737147341572512e-05, "loss": 0.0714, "step": 11880 }, { "grad_norm": 0.3374637961387634, "learning_rate": 9.736485621920735e-05, "loss": 0.0522, "step": 11890 }, { "grad_norm": 0.4804217219352722, "learning_rate": 9.735823092933108e-05, "loss": 0.0569, "step": 11900 }, { "grad_norm": 0.38339370489120483, "learning_rate": 9.735159754722838e-05, "loss": 0.0572, "step": 11910 }, { "grad_norm": 0.48467549681663513, "learning_rate": 9.734495607403275e-05, "loss": 0.0584, "step": 11920 }, { "grad_norm": 0.44855996966362, "learning_rate": 9.733830651087901e-05, "loss": 0.0525, "step": 11930 }, { "grad_norm": 0.480838418006897, "learning_rate": 9.733164885890338e-05, "loss": 0.0566, "step": 11940 }, { "grad_norm": 0.35291028022766113, "learning_rate": 9.732498311924349e-05, "loss": 0.0472, "step": 11950 }, { "grad_norm": 0.5350884795188904, "learning_rate": 9.731830929303833e-05, "loss": 0.0563, "step": 11960 }, { "grad_norm": 0.6261168718338013, "learning_rate": 9.731162738142827e-05, "loss": 0.0501, "step": 11970 }, { "grad_norm": 0.44204530119895935, "learning_rate": 9.730493738555506e-05, "loss": 0.0536, "step": 11980 }, { "grad_norm": 0.3876287639141083, "learning_rate": 9.729823930656186e-05, "loss": 0.0525, "step": 11990 }, { "grad_norm": 0.44433850049972534, "learning_rate": 9.729153314559316e-05, "loss": 0.0663, "step": 12000 }, { "grad_norm": 0.5782232880592346, "learning_rate": 9.728481890379486e-05, "loss": 0.0582, "step": 12010 }, { "grad_norm": 0.3556321859359741, "learning_rate": 9.727809658231428e-05, "loss": 0.0512, "step": 12020 }, { "grad_norm": 0.4959840476512909, "learning_rate": 9.727136618230003e-05, "loss": 0.0578, "step": 12030 }, { "grad_norm": 0.48931220173835754, "learning_rate": 9.726462770490219e-05, "loss": 0.0534, "step": 12040 }, { "grad_norm": 0.5882696509361267, "learning_rate": 9.725788115127214e-05, "loss": 0.0606, "step": 12050 }, { "grad_norm": 0.35501959919929504, "learning_rate": 9.725112652256274e-05, "loss": 0.0574, "step": 12060 }, { "grad_norm": 0.5637917518615723, "learning_rate": 9.724436381992812e-05, "loss": 0.0782, "step": 12070 }, { "grad_norm": 0.5217328667640686, "learning_rate": 9.723759304452387e-05, "loss": 0.0564, "step": 12080 }, { "grad_norm": 0.4181738495826721, "learning_rate": 9.72308141975069e-05, "loss": 0.043, "step": 12090 }, { "grad_norm": 0.4651417136192322, "learning_rate": 9.722402728003557e-05, "loss": 0.0599, "step": 12100 }, { "grad_norm": 0.606221079826355, "learning_rate": 9.721723229326953e-05, "loss": 0.0524, "step": 12110 }, { "grad_norm": 0.48176851868629456, "learning_rate": 9.721042923836992e-05, "loss": 0.0493, "step": 12120 }, { "grad_norm": 0.47097304463386536, "learning_rate": 9.720361811649914e-05, "loss": 0.0537, "step": 12130 }, { "grad_norm": 0.43793562054634094, "learning_rate": 9.719679892882106e-05, "loss": 0.0483, "step": 12140 }, { "grad_norm": 0.5163471698760986, "learning_rate": 9.718997167650085e-05, "loss": 0.0497, "step": 12150 }, { "grad_norm": 0.49941956996917725, "learning_rate": 9.718313636070515e-05, "loss": 0.0485, "step": 12160 }, { "grad_norm": 0.5792545676231384, "learning_rate": 9.717629298260192e-05, "loss": 0.0456, "step": 12170 }, { "grad_norm": 0.5591253042221069, "learning_rate": 9.716944154336047e-05, "loss": 0.0617, "step": 12180 }, { "grad_norm": 0.6212420463562012, "learning_rate": 9.716258204415157e-05, "loss": 0.0619, "step": 12190 }, { "grad_norm": 0.5545949339866638, "learning_rate": 9.715571448614728e-05, "loss": 0.0686, "step": 12200 }, { "grad_norm": 0.4512195289134979, "learning_rate": 9.71488388705211e-05, "loss": 0.0782, "step": 12210 }, { "grad_norm": 0.527969479560852, "learning_rate": 9.714195519844788e-05, "loss": 0.0562, "step": 12220 }, { "grad_norm": 0.42978301644325256, "learning_rate": 9.713506347110386e-05, "loss": 0.0611, "step": 12230 }, { "grad_norm": 0.6457648873329163, "learning_rate": 9.712816368966662e-05, "loss": 0.0582, "step": 12240 }, { "grad_norm": 0.5113154053688049, "learning_rate": 9.712125585531517e-05, "loss": 0.0697, "step": 12250 }, { "grad_norm": 0.5061265230178833, "learning_rate": 9.711433996922988e-05, "loss": 0.0622, "step": 12260 }, { "grad_norm": 0.4852219521999359, "learning_rate": 9.710741603259245e-05, "loss": 0.0521, "step": 12270 }, { "grad_norm": 0.42159196734428406, "learning_rate": 9.710048404658603e-05, "loss": 0.0563, "step": 12280 }, { "grad_norm": 0.4361456036567688, "learning_rate": 9.709354401239508e-05, "loss": 0.0636, "step": 12290 }, { "grad_norm": 0.493865430355072, "learning_rate": 9.708659593120546e-05, "loss": 0.0489, "step": 12300 }, { "grad_norm": 0.47518229484558105, "learning_rate": 9.707963980420443e-05, "loss": 0.056, "step": 12310 }, { "grad_norm": 0.727881908416748, "learning_rate": 9.707267563258058e-05, "loss": 0.0643, "step": 12320 }, { "grad_norm": 0.5291418433189392, "learning_rate": 9.70657034175239e-05, "loss": 0.0564, "step": 12330 }, { "grad_norm": 0.46348243951797485, "learning_rate": 9.705872316022577e-05, "loss": 0.058, "step": 12340 }, { "grad_norm": 0.48819586634635925, "learning_rate": 9.705173486187891e-05, "loss": 0.0577, "step": 12350 }, { "grad_norm": 0.5116863250732422, "learning_rate": 9.704473852367741e-05, "loss": 0.0612, "step": 12360 }, { "grad_norm": 0.4874386787414551, "learning_rate": 9.70377341468168e-05, "loss": 0.054, "step": 12370 }, { "grad_norm": 0.46925613284111023, "learning_rate": 9.703072173249389e-05, "loss": 0.0552, "step": 12380 }, { "grad_norm": 0.5643797516822815, "learning_rate": 9.702370128190693e-05, "loss": 0.0724, "step": 12390 }, { "grad_norm": 0.5283221006393433, "learning_rate": 9.701667279625552e-05, "loss": 0.0624, "step": 12400 }, { "grad_norm": 0.5643315315246582, "learning_rate": 9.700963627674065e-05, "loss": 0.0624, "step": 12410 }, { "grad_norm": 0.46909812092781067, "learning_rate": 9.700259172456466e-05, "loss": 0.0548, "step": 12420 }, { "grad_norm": 0.5474047660827637, "learning_rate": 9.699553914093124e-05, "loss": 0.0681, "step": 12430 }, { "grad_norm": 0.43255484104156494, "learning_rate": 9.698847852704553e-05, "loss": 0.0643, "step": 12440 }, { "grad_norm": 0.5010390877723694, "learning_rate": 9.6981409884114e-05, "loss": 0.048, "step": 12450 }, { "grad_norm": 0.4800815284252167, "learning_rate": 9.697433321334443e-05, "loss": 0.0583, "step": 12460 }, { "grad_norm": 0.3693055808544159, "learning_rate": 9.696724851594607e-05, "loss": 0.0473, "step": 12470 }, { "grad_norm": 0.36722084879875183, "learning_rate": 9.696015579312952e-05, "loss": 0.0495, "step": 12480 }, { "grad_norm": 0.4310552775859833, "learning_rate": 9.695305504610668e-05, "loss": 0.0534, "step": 12490 }, { "grad_norm": 0.3054564297199249, "learning_rate": 9.694594627609092e-05, "loss": 0.0624, "step": 12500 }, { "grad_norm": 0.5270587801933289, "learning_rate": 9.693882948429691e-05, "loss": 0.0635, "step": 12510 }, { "grad_norm": 0.5116704106330872, "learning_rate": 9.693170467194071e-05, "loss": 0.0518, "step": 12520 }, { "grad_norm": 0.4632102847099304, "learning_rate": 9.692457184023977e-05, "loss": 0.0519, "step": 12530 }, { "grad_norm": 0.3928021192550659, "learning_rate": 9.691743099041291e-05, "loss": 0.0632, "step": 12540 }, { "grad_norm": 0.6382891535758972, "learning_rate": 9.691028212368027e-05, "loss": 0.0564, "step": 12550 }, { "grad_norm": 0.5372134447097778, "learning_rate": 9.690312524126342e-05, "loss": 0.0589, "step": 12560 }, { "grad_norm": 0.44086596369743347, "learning_rate": 9.689596034438527e-05, "loss": 0.0567, "step": 12570 }, { "grad_norm": 0.47057557106018066, "learning_rate": 9.688878743427012e-05, "loss": 0.0564, "step": 12580 }, { "grad_norm": 0.481797456741333, "learning_rate": 9.688160651214359e-05, "loss": 0.067, "step": 12590 }, { "grad_norm": 0.42104724049568176, "learning_rate": 9.687441757923273e-05, "loss": 0.0481, "step": 12600 }, { "grad_norm": 0.6273165941238403, "learning_rate": 9.68672206367659e-05, "loss": 0.0583, "step": 12610 }, { "grad_norm": 0.4635733664035797, "learning_rate": 9.686001568597291e-05, "loss": 0.0476, "step": 12620 }, { "grad_norm": 0.3553585112094879, "learning_rate": 9.685280272808486e-05, "loss": 0.0502, "step": 12630 }, { "grad_norm": 0.5057521462440491, "learning_rate": 9.684558176433424e-05, "loss": 0.0652, "step": 12640 }, { "grad_norm": 0.6750718951225281, "learning_rate": 9.683835279595495e-05, "loss": 0.0593, "step": 12650 }, { "grad_norm": 0.5274878740310669, "learning_rate": 9.683111582418216e-05, "loss": 0.0687, "step": 12660 }, { "grad_norm": 0.38051989674568176, "learning_rate": 9.682387085025254e-05, "loss": 0.0556, "step": 12670 }, { "grad_norm": 0.46182310581207275, "learning_rate": 9.681661787540401e-05, "loss": 0.0578, "step": 12680 }, { "grad_norm": 0.4430384933948517, "learning_rate": 9.680935690087593e-05, "loss": 0.0559, "step": 12690 }, { "grad_norm": 0.5042674541473389, "learning_rate": 9.680208792790901e-05, "loss": 0.0505, "step": 12700 }, { "grad_norm": 0.49085259437561035, "learning_rate": 9.679481095774529e-05, "loss": 0.0541, "step": 12710 }, { "grad_norm": 0.5307323336601257, "learning_rate": 9.678752599162822e-05, "loss": 0.0612, "step": 12720 }, { "grad_norm": 0.6438298225402832, "learning_rate": 9.678023303080259e-05, "loss": 0.0533, "step": 12730 }, { "grad_norm": 0.41883987188339233, "learning_rate": 9.677293207651459e-05, "loss": 0.0485, "step": 12740 }, { "grad_norm": 0.3215048313140869, "learning_rate": 9.676562313001173e-05, "loss": 0.0483, "step": 12750 }, { "grad_norm": 0.5932105779647827, "learning_rate": 9.675830619254293e-05, "loss": 0.06, "step": 12760 }, { "grad_norm": 0.44104304909706116, "learning_rate": 9.675098126535843e-05, "loss": 0.0632, "step": 12770 }, { "grad_norm": 0.5979925990104675, "learning_rate": 9.674364834970988e-05, "loss": 0.0632, "step": 12780 }, { "grad_norm": 0.4656454622745514, "learning_rate": 9.673630744685028e-05, "loss": 0.0538, "step": 12790 }, { "grad_norm": 0.3618309497833252, "learning_rate": 9.672895855803397e-05, "loss": 0.0547, "step": 12800 }, { "grad_norm": 0.3792458474636078, "learning_rate": 9.672160168451667e-05, "loss": 0.0474, "step": 12810 }, { "grad_norm": 0.4040696918964386, "learning_rate": 9.671423682755549e-05, "loss": 0.0595, "step": 12820 }, { "grad_norm": 0.6411328315734863, "learning_rate": 9.670686398840888e-05, "loss": 0.0615, "step": 12830 }, { "grad_norm": 0.41875162720680237, "learning_rate": 9.669948316833664e-05, "loss": 0.0527, "step": 12840 }, { "grad_norm": 0.5120159387588501, "learning_rate": 9.669209436859997e-05, "loss": 0.0562, "step": 12850 }, { "grad_norm": 0.4746646583080292, "learning_rate": 9.66846975904614e-05, "loss": 0.0692, "step": 12860 }, { "grad_norm": 0.5939496159553528, "learning_rate": 9.667729283518483e-05, "loss": 0.0559, "step": 12870 }, { "grad_norm": 0.5168810486793518, "learning_rate": 9.666988010403557e-05, "loss": 0.0425, "step": 12880 }, { "grad_norm": 0.5190283060073853, "learning_rate": 9.66624593982802e-05, "loss": 0.0464, "step": 12890 }, { "grad_norm": 0.4258296489715576, "learning_rate": 9.665503071918675e-05, "loss": 0.0589, "step": 12900 }, { "grad_norm": 0.36868390440940857, "learning_rate": 9.664759406802456e-05, "loss": 0.046, "step": 12910 }, { "grad_norm": 0.4316399097442627, "learning_rate": 9.664014944606437e-05, "loss": 0.041, "step": 12920 }, { "grad_norm": 0.4380510747432709, "learning_rate": 9.663269685457822e-05, "loss": 0.0471, "step": 12930 }, { "grad_norm": 0.450238436460495, "learning_rate": 9.662523629483962e-05, "loss": 0.0698, "step": 12940 }, { "grad_norm": 0.39022061228752136, "learning_rate": 9.661776776812333e-05, "loss": 0.0448, "step": 12950 }, { "grad_norm": 0.5142343044281006, "learning_rate": 9.661029127570553e-05, "loss": 0.0486, "step": 12960 }, { "grad_norm": 0.5417535305023193, "learning_rate": 9.660280681886373e-05, "loss": 0.0515, "step": 12970 }, { "grad_norm": 0.5962253212928772, "learning_rate": 9.659531439887685e-05, "loss": 0.0473, "step": 12980 }, { "grad_norm": 0.4230723977088928, "learning_rate": 9.658781401702511e-05, "loss": 0.046, "step": 12990 }, { "grad_norm": 0.4486565887928009, "learning_rate": 9.658030567459015e-05, "loss": 0.0538, "step": 13000 }, { "grad_norm": 0.5421573519706726, "learning_rate": 9.65727893728549e-05, "loss": 0.0629, "step": 13010 }, { "grad_norm": 0.591059148311615, "learning_rate": 9.656526511310375e-05, "loss": 0.0622, "step": 13020 }, { "grad_norm": 0.443368524312973, "learning_rate": 9.655773289662233e-05, "loss": 0.0675, "step": 13030 }, { "grad_norm": 0.47989094257354736, "learning_rate": 9.655019272469772e-05, "loss": 0.0582, "step": 13040 }, { "grad_norm": 0.4416965842247009, "learning_rate": 9.654264459861832e-05, "loss": 0.0452, "step": 13050 }, { "grad_norm": 0.41708824038505554, "learning_rate": 9.653508851967391e-05, "loss": 0.055, "step": 13060 }, { "grad_norm": 0.5029757618904114, "learning_rate": 9.65275244891556e-05, "loss": 0.0661, "step": 13070 }, { "grad_norm": 0.44126030802726746, "learning_rate": 9.651995250835591e-05, "loss": 0.0509, "step": 13080 }, { "grad_norm": 0.38329699635505676, "learning_rate": 9.651237257856862e-05, "loss": 0.0575, "step": 13090 }, { "grad_norm": 0.4689536988735199, "learning_rate": 9.6504784701089e-05, "loss": 0.0523, "step": 13100 }, { "grad_norm": 0.45359715819358826, "learning_rate": 9.649718887721357e-05, "loss": 0.0513, "step": 13110 }, { "grad_norm": 0.44948482513427734, "learning_rate": 9.648958510824028e-05, "loss": 0.0509, "step": 13120 }, { "grad_norm": 0.5195497870445251, "learning_rate": 9.648197339546837e-05, "loss": 0.0576, "step": 13130 }, { "grad_norm": 0.5110699534416199, "learning_rate": 9.647435374019851e-05, "loss": 0.0556, "step": 13140 }, { "grad_norm": 0.5076176524162292, "learning_rate": 9.646672614373266e-05, "loss": 0.0566, "step": 13150 }, { "grad_norm": 0.5448322892189026, "learning_rate": 9.645909060737418e-05, "loss": 0.044, "step": 13160 }, { "grad_norm": 0.4132154583930969, "learning_rate": 9.645144713242778e-05, "loss": 0.0512, "step": 13170 }, { "grad_norm": 0.45914122462272644, "learning_rate": 9.64437957201995e-05, "loss": 0.0505, "step": 13180 }, { "grad_norm": 0.4150124788284302, "learning_rate": 9.643613637199678e-05, "loss": 0.0499, "step": 13190 }, { "grad_norm": 0.47876253724098206, "learning_rate": 9.642846908912839e-05, "loss": 0.057, "step": 13200 }, { "grad_norm": 0.4581277072429657, "learning_rate": 9.642079387290444e-05, "loss": 0.0481, "step": 13210 }, { "grad_norm": 0.6810750365257263, "learning_rate": 9.641311072463644e-05, "loss": 0.0558, "step": 13220 }, { "grad_norm": 0.7641499638557434, "learning_rate": 9.640541964563722e-05, "loss": 0.059, "step": 13230 }, { "grad_norm": 0.5529658794403076, "learning_rate": 9.639772063722096e-05, "loss": 0.0614, "step": 13240 }, { "grad_norm": 0.42246732115745544, "learning_rate": 9.639001370070324e-05, "loss": 0.0541, "step": 13250 }, { "grad_norm": 0.5950027704238892, "learning_rate": 9.638229883740095e-05, "loss": 0.0605, "step": 13260 }, { "grad_norm": 0.40356799960136414, "learning_rate": 9.637457604863233e-05, "loss": 0.0552, "step": 13270 }, { "grad_norm": 0.4497414231300354, "learning_rate": 9.636684533571703e-05, "loss": 0.0512, "step": 13280 }, { "grad_norm": 0.42795658111572266, "learning_rate": 9.635910669997599e-05, "loss": 0.0443, "step": 13290 }, { "grad_norm": 0.4382409453392029, "learning_rate": 9.635136014273154e-05, "loss": 0.046, "step": 13300 }, { "grad_norm": 0.5565652847290039, "learning_rate": 9.634360566530735e-05, "loss": 0.0659, "step": 13310 }, { "grad_norm": 0.5482224822044373, "learning_rate": 9.633584326902845e-05, "loss": 0.0713, "step": 13320 }, { "grad_norm": 0.5469977855682373, "learning_rate": 9.632807295522124e-05, "loss": 0.0596, "step": 13330 }, { "grad_norm": 0.536787748336792, "learning_rate": 9.632029472521342e-05, "loss": 0.0501, "step": 13340 }, { "grad_norm": 0.5409473776817322, "learning_rate": 9.631250858033409e-05, "loss": 0.0533, "step": 13350 }, { "grad_norm": 0.4851268529891968, "learning_rate": 9.630471452191371e-05, "loss": 0.0479, "step": 13360 }, { "grad_norm": 0.44094735383987427, "learning_rate": 9.629691255128405e-05, "loss": 0.0628, "step": 13370 }, { "grad_norm": 0.593145489692688, "learning_rate": 9.628910266977825e-05, "loss": 0.0503, "step": 13380 }, { "grad_norm": 0.41949036717414856, "learning_rate": 9.628128487873083e-05, "loss": 0.0616, "step": 13390 }, { "grad_norm": 0.41881296038627625, "learning_rate": 9.627345917947761e-05, "loss": 0.0657, "step": 13400 }, { "grad_norm": 0.39108341932296753, "learning_rate": 9.626562557335579e-05, "loss": 0.0574, "step": 13410 }, { "grad_norm": 0.5633480548858643, "learning_rate": 9.625778406170393e-05, "loss": 0.0523, "step": 13420 }, { "grad_norm": 0.488107830286026, "learning_rate": 9.624993464586193e-05, "loss": 0.0546, "step": 13430 }, { "grad_norm": 0.5409313440322876, "learning_rate": 9.624207732717105e-05, "loss": 0.0512, "step": 13440 }, { "grad_norm": 0.5104399919509888, "learning_rate": 9.623421210697386e-05, "loss": 0.0518, "step": 13450 }, { "grad_norm": 0.5749576091766357, "learning_rate": 9.622633898661434e-05, "loss": 0.0589, "step": 13460 }, { "grad_norm": 0.4759705364704132, "learning_rate": 9.621845796743778e-05, "loss": 0.0693, "step": 13470 }, { "grad_norm": 0.4582413136959076, "learning_rate": 9.621056905079082e-05, "loss": 0.0636, "step": 13480 }, { "grad_norm": 0.48876601457595825, "learning_rate": 9.620267223802149e-05, "loss": 0.0649, "step": 13490 }, { "grad_norm": 0.4354824721813202, "learning_rate": 9.619476753047911e-05, "loss": 0.0439, "step": 13500 }, { "grad_norm": 0.5836232304573059, "learning_rate": 9.618685492951438e-05, "loss": 0.0666, "step": 13510 }, { "grad_norm": 0.3802037537097931, "learning_rate": 9.617893443647938e-05, "loss": 0.0625, "step": 13520 }, { "grad_norm": 0.6142165660858154, "learning_rate": 9.617100605272746e-05, "loss": 0.0632, "step": 13530 }, { "grad_norm": 0.37920257449150085, "learning_rate": 9.616306977961338e-05, "loss": 0.0568, "step": 13540 }, { "grad_norm": 0.4322676658630371, "learning_rate": 9.615512561849326e-05, "loss": 0.0451, "step": 13550 }, { "grad_norm": 0.37481117248535156, "learning_rate": 9.61471735707245e-05, "loss": 0.0543, "step": 13560 }, { "grad_norm": 0.5402946472167969, "learning_rate": 9.613921363766592e-05, "loss": 0.0523, "step": 13570 }, { "grad_norm": 0.5387834310531616, "learning_rate": 9.613124582067763e-05, "loss": 0.0502, "step": 13580 }, { "grad_norm": 0.389156311750412, "learning_rate": 9.612327012112112e-05, "loss": 0.0675, "step": 13590 }, { "grad_norm": 0.45347169041633606, "learning_rate": 9.611528654035921e-05, "loss": 0.0752, "step": 13600 }, { "grad_norm": 0.3512389361858368, "learning_rate": 9.610729507975611e-05, "loss": 0.0486, "step": 13610 }, { "grad_norm": 0.42518889904022217, "learning_rate": 9.609929574067731e-05, "loss": 0.0594, "step": 13620 }, { "grad_norm": 0.6305511593818665, "learning_rate": 9.609128852448967e-05, "loss": 0.0502, "step": 13630 }, { "grad_norm": 0.4417577385902405, "learning_rate": 9.608327343256143e-05, "loss": 0.056, "step": 13640 }, { "grad_norm": 0.46185511350631714, "learning_rate": 9.607525046626216e-05, "loss": 0.0583, "step": 13650 }, { "grad_norm": 0.5359417796134949, "learning_rate": 9.606721962696272e-05, "loss": 0.0622, "step": 13660 }, { "grad_norm": 0.5162296295166016, "learning_rate": 9.60591809160354e-05, "loss": 0.0463, "step": 13670 }, { "grad_norm": 0.410521537065506, "learning_rate": 9.605113433485378e-05, "loss": 0.0552, "step": 13680 }, { "grad_norm": 0.5824503302574158, "learning_rate": 9.604307988479279e-05, "loss": 0.0541, "step": 13690 }, { "grad_norm": 0.4183662533760071, "learning_rate": 9.603501756722876e-05, "loss": 0.0473, "step": 13700 }, { "grad_norm": 0.4406742751598358, "learning_rate": 9.602694738353927e-05, "loss": 0.0454, "step": 13710 }, { "grad_norm": 0.3903163969516754, "learning_rate": 9.601886933510331e-05, "loss": 0.0485, "step": 13720 }, { "grad_norm": 0.5012121200561523, "learning_rate": 9.60107834233012e-05, "loss": 0.0499, "step": 13730 }, { "grad_norm": 0.6095306277275085, "learning_rate": 9.60026896495146e-05, "loss": 0.0445, "step": 13740 }, { "grad_norm": 0.3318592607975006, "learning_rate": 9.599458801512652e-05, "loss": 0.0522, "step": 13750 }, { "grad_norm": 0.5793579816818237, "learning_rate": 9.598647852152129e-05, "loss": 0.0752, "step": 13760 }, { "grad_norm": 0.3882354199886322, "learning_rate": 9.597836117008462e-05, "loss": 0.045, "step": 13770 }, { "grad_norm": 0.4845195412635803, "learning_rate": 9.597023596220356e-05, "loss": 0.0528, "step": 13780 }, { "grad_norm": 0.5414882898330688, "learning_rate": 9.596210289926643e-05, "loss": 0.0537, "step": 13790 }, { "grad_norm": 0.7694136500358582, "learning_rate": 9.5953961982663e-05, "loss": 0.0611, "step": 13800 }, { "grad_norm": 0.541191041469574, "learning_rate": 9.594581321378431e-05, "loss": 0.068, "step": 13810 }, { "grad_norm": 0.5286421179771423, "learning_rate": 9.593765659402276e-05, "loss": 0.0616, "step": 13820 }, { "grad_norm": 0.5387779474258423, "learning_rate": 9.59294921247721e-05, "loss": 0.074, "step": 13830 }, { "grad_norm": 0.4939768612384796, "learning_rate": 9.59213198074274e-05, "loss": 0.0559, "step": 13840 }, { "grad_norm": 0.5473998785018921, "learning_rate": 9.59131396433851e-05, "loss": 0.0504, "step": 13850 }, { "grad_norm": 0.48255273699760437, "learning_rate": 9.590495163404297e-05, "loss": 0.0575, "step": 13860 }, { "grad_norm": 0.5365680456161499, "learning_rate": 9.589675578080009e-05, "loss": 0.0671, "step": 13870 }, { "grad_norm": 0.3851783871650696, "learning_rate": 9.588855208505694e-05, "loss": 0.0511, "step": 13880 }, { "grad_norm": 0.584943413734436, "learning_rate": 9.588034054821529e-05, "loss": 0.0695, "step": 13890 }, { "grad_norm": 0.3760607838630676, "learning_rate": 9.587212117167826e-05, "loss": 0.0521, "step": 13900 }, { "grad_norm": 0.5811701416969299, "learning_rate": 9.586389395685033e-05, "loss": 0.0533, "step": 13910 }, { "grad_norm": 0.45144206285476685, "learning_rate": 9.585565890513733e-05, "loss": 0.0511, "step": 13920 }, { "grad_norm": 0.4382421672344208, "learning_rate": 9.584741601794636e-05, "loss": 0.0485, "step": 13930 }, { "grad_norm": 0.45785075426101685, "learning_rate": 9.58391652966859e-05, "loss": 0.0505, "step": 13940 }, { "grad_norm": 0.41102102398872375, "learning_rate": 9.583090674276583e-05, "loss": 0.0573, "step": 13950 }, { "grad_norm": 0.534747838973999, "learning_rate": 9.582264035759726e-05, "loss": 0.0563, "step": 13960 }, { "grad_norm": 0.44986116886138916, "learning_rate": 9.58143661425927e-05, "loss": 0.071, "step": 13970 }, { "grad_norm": 0.532214343547821, "learning_rate": 9.580608409916601e-05, "loss": 0.0697, "step": 13980 }, { "grad_norm": 0.5341051816940308, "learning_rate": 9.579779422873233e-05, "loss": 0.0547, "step": 13990 }, { "grad_norm": 0.5589120984077454, "learning_rate": 9.578949653270819e-05, "loss": 0.055, "step": 14000 }, { "grad_norm": 0.36599189043045044, "learning_rate": 9.578119101251144e-05, "loss": 0.0505, "step": 14010 }, { "grad_norm": 0.5721931457519531, "learning_rate": 9.577287766956127e-05, "loss": 0.0631, "step": 14020 }, { "grad_norm": 0.5259125232696533, "learning_rate": 9.57645565052782e-05, "loss": 0.0618, "step": 14030 }, { "grad_norm": 0.5180391669273376, "learning_rate": 9.575622752108407e-05, "loss": 0.0506, "step": 14040 }, { "grad_norm": 0.5082559585571289, "learning_rate": 9.57478907184021e-05, "loss": 0.0504, "step": 14050 }, { "grad_norm": 0.36748653650283813, "learning_rate": 9.573954609865681e-05, "loss": 0.0545, "step": 14060 }, { "grad_norm": 0.4466114640235901, "learning_rate": 9.573119366327408e-05, "loss": 0.0548, "step": 14070 }, { "grad_norm": 0.6346703767776489, "learning_rate": 9.57228334136811e-05, "loss": 0.0493, "step": 14080 }, { "grad_norm": 0.38413673639297485, "learning_rate": 9.571446535130641e-05, "loss": 0.0451, "step": 14090 }, { "grad_norm": 0.3812517523765564, "learning_rate": 9.570608947757988e-05, "loss": 0.0521, "step": 14100 }, { "grad_norm": 0.4674017131328583, "learning_rate": 9.569770579393274e-05, "loss": 0.0553, "step": 14110 }, { "grad_norm": 0.34542495012283325, "learning_rate": 9.56893143017975e-05, "loss": 0.0429, "step": 14120 }, { "grad_norm": 0.5497802495956421, "learning_rate": 9.568091500260806e-05, "loss": 0.0616, "step": 14130 }, { "grad_norm": 0.43161678314208984, "learning_rate": 9.567250789779961e-05, "loss": 0.0502, "step": 14140 }, { "grad_norm": 0.3916386365890503, "learning_rate": 9.566409298880872e-05, "loss": 0.0498, "step": 14150 }, { "grad_norm": 0.7429450750350952, "learning_rate": 9.565567027707326e-05, "loss": 0.067, "step": 14160 }, { "grad_norm": 0.4327518343925476, "learning_rate": 9.56472397640324e-05, "loss": 0.0548, "step": 14170 }, { "grad_norm": 0.5025187730789185, "learning_rate": 9.563880145112675e-05, "loss": 0.0565, "step": 14180 }, { "grad_norm": 0.6980935335159302, "learning_rate": 9.563035533979814e-05, "loss": 0.0562, "step": 14190 }, { "grad_norm": 0.8492335677146912, "learning_rate": 9.562190143148981e-05, "loss": 0.0441, "step": 14200 }, { "grad_norm": 0.5144391655921936, "learning_rate": 9.561343972764627e-05, "loss": 0.0574, "step": 14210 }, { "grad_norm": 0.46691372990608215, "learning_rate": 9.560497022971343e-05, "loss": 0.0622, "step": 14220 }, { "grad_norm": 0.5182954668998718, "learning_rate": 9.559649293913847e-05, "loss": 0.0636, "step": 14230 }, { "grad_norm": 0.5944312810897827, "learning_rate": 9.558800785736993e-05, "loss": 0.0496, "step": 14240 }, { "grad_norm": 0.4675234854221344, "learning_rate": 9.557951498585767e-05, "loss": 0.0533, "step": 14250 }, { "grad_norm": 0.3846336305141449, "learning_rate": 9.557101432605293e-05, "loss": 0.0594, "step": 14260 }, { "grad_norm": 0.43605276942253113, "learning_rate": 9.556250587940818e-05, "loss": 0.0652, "step": 14270 }, { "grad_norm": 0.5651015043258667, "learning_rate": 9.555398964737734e-05, "loss": 0.0504, "step": 14280 }, { "grad_norm": 0.3844316005706787, "learning_rate": 9.554546563141555e-05, "loss": 0.0485, "step": 14290 }, { "grad_norm": 0.4707232713699341, "learning_rate": 9.553693383297937e-05, "loss": 0.0597, "step": 14300 }, { "grad_norm": 0.5647017359733582, "learning_rate": 9.552839425352663e-05, "loss": 0.0519, "step": 14310 }, { "grad_norm": 0.4671505093574524, "learning_rate": 9.551984689451652e-05, "loss": 0.0701, "step": 14320 }, { "grad_norm": 0.3902425169944763, "learning_rate": 9.551129175740953e-05, "loss": 0.052, "step": 14330 }, { "grad_norm": 0.5415008664131165, "learning_rate": 9.550272884366754e-05, "loss": 0.0576, "step": 14340 }, { "grad_norm": 0.5647459030151367, "learning_rate": 9.549415815475369e-05, "loss": 0.0562, "step": 14350 }, { "grad_norm": 0.5443434715270996, "learning_rate": 9.548557969213247e-05, "loss": 0.0622, "step": 14360 }, { "grad_norm": 0.5523458123207092, "learning_rate": 9.547699345726972e-05, "loss": 0.0575, "step": 14370 }, { "grad_norm": 0.3604947626590729, "learning_rate": 9.546839945163257e-05, "loss": 0.0547, "step": 14380 }, { "grad_norm": 0.452946275472641, "learning_rate": 9.545979767668953e-05, "loss": 0.0623, "step": 14390 }, { "grad_norm": 0.4772244095802307, "learning_rate": 9.54511881339104e-05, "loss": 0.0582, "step": 14400 }, { "grad_norm": 0.6691036820411682, "learning_rate": 9.54425708247663e-05, "loss": 0.0438, "step": 14410 }, { "grad_norm": 0.48447611927986145, "learning_rate": 9.543394575072972e-05, "loss": 0.0562, "step": 14420 }, { "grad_norm": 0.4608376920223236, "learning_rate": 9.542531291327441e-05, "loss": 0.0542, "step": 14430 }, { "grad_norm": 0.6108322143554688, "learning_rate": 9.541667231387552e-05, "loss": 0.0499, "step": 14440 }, { "grad_norm": 0.7162543535232544, "learning_rate": 9.540802395400949e-05, "loss": 0.0669, "step": 14450 }, { "grad_norm": 0.4797457754611969, "learning_rate": 9.539936783515406e-05, "loss": 0.0615, "step": 14460 }, { "grad_norm": 0.3593195080757141, "learning_rate": 9.539070395878835e-05, "loss": 0.0505, "step": 14470 }, { "grad_norm": 0.4530123174190521, "learning_rate": 9.538203232639277e-05, "loss": 0.0436, "step": 14480 }, { "grad_norm": 0.537391722202301, "learning_rate": 9.537335293944907e-05, "loss": 0.0485, "step": 14490 }, { "grad_norm": 0.6105149388313293, "learning_rate": 9.536466579944032e-05, "loss": 0.0695, "step": 14500 }, { "grad_norm": 0.4966566860675812, "learning_rate": 9.535597090785091e-05, "loss": 0.056, "step": 14510 }, { "grad_norm": 0.39378803968429565, "learning_rate": 9.534726826616656e-05, "loss": 0.0491, "step": 14520 }, { "grad_norm": 0.773104190826416, "learning_rate": 9.53385578758743e-05, "loss": 0.0863, "step": 14530 }, { "grad_norm": 0.42428919672966003, "learning_rate": 9.532983973846252e-05, "loss": 0.0523, "step": 14540 }, { "grad_norm": 0.4698517620563507, "learning_rate": 9.53211138554209e-05, "loss": 0.0471, "step": 14550 }, { "grad_norm": 0.4804444909095764, "learning_rate": 9.531238022824047e-05, "loss": 0.0605, "step": 14560 }, { "grad_norm": 0.37203651666641235, "learning_rate": 9.530363885841355e-05, "loss": 0.0396, "step": 14570 }, { "grad_norm": 0.6097116470336914, "learning_rate": 9.52948897474338e-05, "loss": 0.057, "step": 14580 }, { "grad_norm": 0.6232292056083679, "learning_rate": 9.528613289679622e-05, "loss": 0.0673, "step": 14590 }, { "grad_norm": 0.4834263026714325, "learning_rate": 9.52773683079971e-05, "loss": 0.0564, "step": 14600 }, { "grad_norm": 0.361028254032135, "learning_rate": 9.526859598253407e-05, "loss": 0.0492, "step": 14610 }, { "grad_norm": 0.44769158959388733, "learning_rate": 9.525981592190609e-05, "loss": 0.0483, "step": 14620 }, { "grad_norm": 0.44235995411872864, "learning_rate": 9.525102812761342e-05, "loss": 0.0521, "step": 14630 }, { "grad_norm": 0.5086811780929565, "learning_rate": 9.524223260115768e-05, "loss": 0.0406, "step": 14640 }, { "grad_norm": 0.44831934571266174, "learning_rate": 9.523342934404175e-05, "loss": 0.0516, "step": 14650 }, { "grad_norm": 0.41846248507499695, "learning_rate": 9.522461835776989e-05, "loss": 0.0441, "step": 14660 }, { "grad_norm": 0.47795847058296204, "learning_rate": 9.521579964384764e-05, "loss": 0.0512, "step": 14670 }, { "grad_norm": 0.45485618710517883, "learning_rate": 9.52069732037819e-05, "loss": 0.0456, "step": 14680 }, { "grad_norm": 0.3837031424045563, "learning_rate": 9.519813903908083e-05, "loss": 0.0509, "step": 14690 }, { "grad_norm": 0.45672741532325745, "learning_rate": 9.5189297151254e-05, "loss": 0.0533, "step": 14700 }, { "grad_norm": 0.6660682559013367, "learning_rate": 9.518044754181218e-05, "loss": 0.0564, "step": 14710 }, { "grad_norm": 0.4622816741466522, "learning_rate": 9.51715902122676e-05, "loss": 0.0553, "step": 14720 }, { "grad_norm": 0.5021299123764038, "learning_rate": 9.516272516413368e-05, "loss": 0.0562, "step": 14730 }, { "grad_norm": 0.36048561334609985, "learning_rate": 9.515385239892525e-05, "loss": 0.0571, "step": 14740 }, { "grad_norm": 0.561651885509491, "learning_rate": 9.514497191815839e-05, "loss": 0.0448, "step": 14750 }, { "grad_norm": 0.3847068250179291, "learning_rate": 9.513608372335055e-05, "loss": 0.0538, "step": 14760 }, { "grad_norm": 0.4278341829776764, "learning_rate": 9.512718781602045e-05, "loss": 0.053, "step": 14770 }, { "grad_norm": 0.42738083004951477, "learning_rate": 9.511828419768823e-05, "loss": 0.0546, "step": 14780 }, { "grad_norm": 0.5335391163825989, "learning_rate": 9.510937286987521e-05, "loss": 0.0516, "step": 14790 }, { "grad_norm": 0.6308862566947937, "learning_rate": 9.510045383410408e-05, "loss": 0.0715, "step": 14800 }, { "grad_norm": 0.48486408591270447, "learning_rate": 9.509152709189892e-05, "loss": 0.0567, "step": 14810 }, { "grad_norm": 0.3681463301181793, "learning_rate": 9.508259264478504e-05, "loss": 0.0616, "step": 14820 }, { "grad_norm": 0.48871660232543945, "learning_rate": 9.507365049428909e-05, "loss": 0.0769, "step": 14830 }, { "grad_norm": 0.7165206670761108, "learning_rate": 9.506470064193902e-05, "loss": 0.0719, "step": 14840 }, { "grad_norm": 0.45030128955841064, "learning_rate": 9.505574308926414e-05, "loss": 0.0699, "step": 14850 }, { "grad_norm": 0.48166927695274353, "learning_rate": 9.504677783779505e-05, "loss": 0.0527, "step": 14860 }, { "grad_norm": 0.32876360416412354, "learning_rate": 9.503780488906365e-05, "loss": 0.0479, "step": 14870 }, { "grad_norm": 0.48043587803840637, "learning_rate": 9.502882424460319e-05, "loss": 0.0597, "step": 14880 }, { "grad_norm": 0.6571159362792969, "learning_rate": 9.501983590594821e-05, "loss": 0.0623, "step": 14890 }, { "grad_norm": 0.42800089716911316, "learning_rate": 9.501083987463455e-05, "loss": 0.0526, "step": 14900 }, { "grad_norm": 0.46146950125694275, "learning_rate": 9.500183615219942e-05, "loss": 0.0534, "step": 14910 }, { "grad_norm": 0.4623904228210449, "learning_rate": 9.49928247401813e-05, "loss": 0.0454, "step": 14920 }, { "grad_norm": 0.48502469062805176, "learning_rate": 9.498380564011997e-05, "loss": 0.0443, "step": 14930 }, { "grad_norm": 0.3770906627178192, "learning_rate": 9.497477885355656e-05, "loss": 0.0586, "step": 14940 }, { "grad_norm": 0.5316478610038757, "learning_rate": 9.496574438203353e-05, "loss": 0.0597, "step": 14950 }, { "grad_norm": 0.3980375826358795, "learning_rate": 9.495670222709459e-05, "loss": 0.0538, "step": 14960 }, { "grad_norm": 0.39652401208877563, "learning_rate": 9.494765239028483e-05, "loss": 0.0509, "step": 14970 }, { "grad_norm": 0.3452291190624237, "learning_rate": 9.493859487315057e-05, "loss": 0.053, "step": 14980 }, { "grad_norm": 0.38844767212867737, "learning_rate": 9.492952967723953e-05, "loss": 0.0474, "step": 14990 }, { "grad_norm": 0.45099836587905884, "learning_rate": 9.492045680410068e-05, "loss": 0.0562, "step": 15000 }, { "grad_norm": 0.3363848030567169, "learning_rate": 9.491137625528436e-05, "loss": 0.0358, "step": 15010 }, { "grad_norm": 0.49462392926216125, "learning_rate": 9.490228803234215e-05, "loss": 0.052, "step": 15020 }, { "grad_norm": 0.5183454155921936, "learning_rate": 9.489319213682701e-05, "loss": 0.0482, "step": 15030 }, { "grad_norm": 0.4923206865787506, "learning_rate": 9.488408857029316e-05, "loss": 0.0611, "step": 15040 }, { "grad_norm": 0.4179174304008484, "learning_rate": 9.487497733429616e-05, "loss": 0.0617, "step": 15050 }, { "grad_norm": 0.5426543951034546, "learning_rate": 9.486585843039286e-05, "loss": 0.0537, "step": 15060 }, { "grad_norm": 0.8470214605331421, "learning_rate": 9.485673186014143e-05, "loss": 0.052, "step": 15070 }, { "grad_norm": 0.4290822148323059, "learning_rate": 9.484759762510137e-05, "loss": 0.058, "step": 15080 }, { "grad_norm": 0.5341482162475586, "learning_rate": 9.483845572683346e-05, "loss": 0.0603, "step": 15090 }, { "grad_norm": 0.4587027132511139, "learning_rate": 9.48293061668998e-05, "loss": 0.0623, "step": 15100 }, { "grad_norm": 0.5525330901145935, "learning_rate": 9.48201489468638e-05, "loss": 0.0515, "step": 15110 }, { "grad_norm": 0.4321880638599396, "learning_rate": 9.481098406829016e-05, "loss": 0.054, "step": 15120 }, { "grad_norm": 0.4439099431037903, "learning_rate": 9.480181153274495e-05, "loss": 0.0518, "step": 15130 }, { "grad_norm": 0.6491591930389404, "learning_rate": 9.479263134179548e-05, "loss": 0.0448, "step": 15140 }, { "grad_norm": 0.4275960624217987, "learning_rate": 9.478344349701039e-05, "loss": 0.0487, "step": 15150 }, { "grad_norm": 0.3028605878353119, "learning_rate": 9.477424799995964e-05, "loss": 0.0403, "step": 15160 }, { "grad_norm": 0.5741164684295654, "learning_rate": 9.476504485221448e-05, "loss": 0.0586, "step": 15170 }, { "grad_norm": 0.478837251663208, "learning_rate": 9.475583405534748e-05, "loss": 0.0534, "step": 15180 }, { "grad_norm": 0.32231229543685913, "learning_rate": 9.474661561093251e-05, "loss": 0.047, "step": 15190 }, { "grad_norm": 0.5598993301391602, "learning_rate": 9.473738952054478e-05, "loss": 0.0532, "step": 15200 }, { "grad_norm": 0.37955164909362793, "learning_rate": 9.472815578576073e-05, "loss": 0.0413, "step": 15210 }, { "grad_norm": 0.5860170722007751, "learning_rate": 9.471891440815817e-05, "loss": 0.0475, "step": 15220 }, { "grad_norm": 0.37027356028556824, "learning_rate": 9.470966538931621e-05, "loss": 0.0448, "step": 15230 }, { "grad_norm": 0.3738586902618408, "learning_rate": 9.470040873081525e-05, "loss": 0.0379, "step": 15240 }, { "grad_norm": 0.4573763608932495, "learning_rate": 9.469114443423698e-05, "loss": 0.0574, "step": 15250 }, { "grad_norm": 0.4318070709705353, "learning_rate": 9.468187250116445e-05, "loss": 0.051, "step": 15260 }, { "grad_norm": 0.26630499958992004, "learning_rate": 9.467259293318197e-05, "loss": 0.0436, "step": 15270 }, { "grad_norm": 0.40650802850723267, "learning_rate": 9.466330573187514e-05, "loss": 0.0469, "step": 15280 }, { "grad_norm": 0.4560418426990509, "learning_rate": 9.46540108988309e-05, "loss": 0.0567, "step": 15290 }, { "grad_norm": 0.3803328573703766, "learning_rate": 9.46447084356375e-05, "loss": 0.051, "step": 15300 }, { "grad_norm": 0.40638697147369385, "learning_rate": 9.463539834388447e-05, "loss": 0.0668, "step": 15310 }, { "grad_norm": 0.5581241250038147, "learning_rate": 9.462608062516263e-05, "loss": 0.0534, "step": 15320 }, { "grad_norm": 0.5204662680625916, "learning_rate": 9.461675528106413e-05, "loss": 0.042, "step": 15330 }, { "grad_norm": 0.42423683404922485, "learning_rate": 9.460742231318244e-05, "loss": 0.0524, "step": 15340 }, { "grad_norm": 0.5152906775474548, "learning_rate": 9.459808172311229e-05, "loss": 0.0572, "step": 15350 }, { "grad_norm": 0.5107575058937073, "learning_rate": 9.458873351244972e-05, "loss": 0.0463, "step": 15360 }, { "grad_norm": 0.40424293279647827, "learning_rate": 9.457937768279211e-05, "loss": 0.0459, "step": 15370 }, { "grad_norm": 0.35307520627975464, "learning_rate": 9.45700142357381e-05, "loss": 0.0481, "step": 15380 }, { "grad_norm": 0.5453061461448669, "learning_rate": 9.456064317288765e-05, "loss": 0.0427, "step": 15390 }, { "grad_norm": 0.41878196597099304, "learning_rate": 9.455126449584201e-05, "loss": 0.0442, "step": 15400 }, { "grad_norm": 0.4799332022666931, "learning_rate": 9.454187820620375e-05, "loss": 0.037, "step": 15410 }, { "grad_norm": 0.31417444348335266, "learning_rate": 9.453248430557673e-05, "loss": 0.0458, "step": 15420 }, { "grad_norm": 0.4289153516292572, "learning_rate": 9.452308279556611e-05, "loss": 0.0695, "step": 15430 }, { "grad_norm": 0.3662467300891876, "learning_rate": 9.451367367777835e-05, "loss": 0.0525, "step": 15440 }, { "grad_norm": 0.4923093318939209, "learning_rate": 9.450425695382122e-05, "loss": 0.061, "step": 15450 }, { "grad_norm": 0.4841388463973999, "learning_rate": 9.449483262530375e-05, "loss": 0.0465, "step": 15460 }, { "grad_norm": 0.701975405216217, "learning_rate": 9.448540069383633e-05, "loss": 0.0423, "step": 15470 }, { "grad_norm": 0.44991928339004517, "learning_rate": 9.447596116103061e-05, "loss": 0.048, "step": 15480 }, { "grad_norm": 0.48298361897468567, "learning_rate": 9.446651402849955e-05, "loss": 0.0559, "step": 15490 }, { "grad_norm": 0.6438928246498108, "learning_rate": 9.44570592978574e-05, "loss": 0.0496, "step": 15500 }, { "grad_norm": 0.3975922465324402, "learning_rate": 9.444759697071972e-05, "loss": 0.0539, "step": 15510 }, { "grad_norm": 0.38002318143844604, "learning_rate": 9.443812704870336e-05, "loss": 0.0419, "step": 15520 }, { "grad_norm": 0.43144041299819946, "learning_rate": 9.442864953342649e-05, "loss": 0.0438, "step": 15530 }, { "grad_norm": 0.3920728266239166, "learning_rate": 9.441916442650852e-05, "loss": 0.051, "step": 15540 }, { "grad_norm": 0.42895445227622986, "learning_rate": 9.440967172957023e-05, "loss": 0.0439, "step": 15550 }, { "grad_norm": 0.4995504915714264, "learning_rate": 9.440017144423364e-05, "loss": 0.0607, "step": 15560 }, { "grad_norm": 0.4236956536769867, "learning_rate": 9.439066357212209e-05, "loss": 0.0496, "step": 15570 }, { "grad_norm": 0.47913485765457153, "learning_rate": 9.438114811486022e-05, "loss": 0.0631, "step": 15580 }, { "grad_norm": 0.43328461050987244, "learning_rate": 9.4371625074074e-05, "loss": 0.043, "step": 15590 }, { "grad_norm": 0.46769559383392334, "learning_rate": 9.436209445139059e-05, "loss": 0.0445, "step": 15600 }, { "grad_norm": 0.35552978515625, "learning_rate": 9.435255624843855e-05, "loss": 0.0579, "step": 15610 }, { "grad_norm": 0.3748505115509033, "learning_rate": 9.43430104668477e-05, "loss": 0.0429, "step": 15620 }, { "grad_norm": 0.4881869852542877, "learning_rate": 9.433345710824914e-05, "loss": 0.0464, "step": 15630 }, { "grad_norm": 0.5382202863693237, "learning_rate": 9.432389617427529e-05, "loss": 0.048, "step": 15640 }, { "grad_norm": 0.3611733615398407, "learning_rate": 9.431432766655984e-05, "loss": 0.0508, "step": 15650 }, { "grad_norm": 0.4909653067588806, "learning_rate": 9.430475158673778e-05, "loss": 0.0507, "step": 15660 }, { "grad_norm": 0.48364681005477905, "learning_rate": 9.429516793644542e-05, "loss": 0.0432, "step": 15670 }, { "grad_norm": 0.43098023533821106, "learning_rate": 9.428557671732034e-05, "loss": 0.0411, "step": 15680 }, { "grad_norm": 0.37545904517173767, "learning_rate": 9.42759779310014e-05, "loss": 0.0433, "step": 15690 }, { "grad_norm": 0.4046085774898529, "learning_rate": 9.426637157912879e-05, "loss": 0.0449, "step": 15700 }, { "grad_norm": 0.6270937323570251, "learning_rate": 9.425675766334397e-05, "loss": 0.056, "step": 15710 }, { "grad_norm": 0.543235182762146, "learning_rate": 9.424713618528968e-05, "loss": 0.0419, "step": 15720 }, { "grad_norm": 0.46611160039901733, "learning_rate": 9.423750714661e-05, "loss": 0.0502, "step": 15730 }, { "grad_norm": 0.4446655511856079, "learning_rate": 9.422787054895022e-05, "loss": 0.044, "step": 15740 }, { "grad_norm": 0.4048289358615875, "learning_rate": 9.4218226393957e-05, "loss": 0.0469, "step": 15750 }, { "grad_norm": 0.4600493609905243, "learning_rate": 9.420857468327828e-05, "loss": 0.0485, "step": 15760 }, { "grad_norm": 0.42697426676750183, "learning_rate": 9.419891541856323e-05, "loss": 0.053, "step": 15770 }, { "grad_norm": 0.32044118642807007, "learning_rate": 9.41892486014624e-05, "loss": 0.0454, "step": 15780 }, { "grad_norm": 0.40885913372039795, "learning_rate": 9.417957423362756e-05, "loss": 0.0549, "step": 15790 }, { "grad_norm": 0.4074307978153229, "learning_rate": 9.416989231671178e-05, "loss": 0.0483, "step": 15800 }, { "grad_norm": 0.6638741493225098, "learning_rate": 9.416020285236946e-05, "loss": 0.0508, "step": 15810 }, { "grad_norm": 0.392665833234787, "learning_rate": 9.415050584225626e-05, "loss": 0.0433, "step": 15820 }, { "grad_norm": 0.45502445101737976, "learning_rate": 9.414080128802914e-05, "loss": 0.0523, "step": 15830 }, { "grad_norm": 0.5561049580574036, "learning_rate": 9.413108919134632e-05, "loss": 0.0486, "step": 15840 }, { "grad_norm": 0.4519883990287781, "learning_rate": 9.412136955386734e-05, "loss": 0.0546, "step": 15850 }, { "grad_norm": 0.372522234916687, "learning_rate": 9.411164237725303e-05, "loss": 0.0476, "step": 15860 }, { "grad_norm": 0.37244319915771484, "learning_rate": 9.41019076631655e-05, "loss": 0.049, "step": 15870 }, { "grad_norm": 0.38269761204719543, "learning_rate": 9.409216541326815e-05, "loss": 0.059, "step": 15880 }, { "grad_norm": 0.3614004850387573, "learning_rate": 9.408241562922564e-05, "loss": 0.0562, "step": 15890 }, { "grad_norm": 0.4081343412399292, "learning_rate": 9.407265831270395e-05, "loss": 0.0368, "step": 15900 }, { "grad_norm": 0.4942696988582611, "learning_rate": 9.406289346537035e-05, "loss": 0.0511, "step": 15910 }, { "grad_norm": 0.42987459897994995, "learning_rate": 9.405312108889339e-05, "loss": 0.0518, "step": 15920 }, { "grad_norm": 0.44385719299316406, "learning_rate": 9.404334118494288e-05, "loss": 0.0526, "step": 15930 }, { "grad_norm": 0.37752318382263184, "learning_rate": 9.403355375518995e-05, "loss": 0.0491, "step": 15940 }, { "grad_norm": 0.5114177465438843, "learning_rate": 9.4023758801307e-05, "loss": 0.0452, "step": 15950 }, { "grad_norm": 0.3368467390537262, "learning_rate": 9.401395632496774e-05, "loss": 0.0413, "step": 15960 }, { "grad_norm": 0.4067502021789551, "learning_rate": 9.400414632784711e-05, "loss": 0.0508, "step": 15970 }, { "grad_norm": 0.5151896476745605, "learning_rate": 9.39943288116214e-05, "loss": 0.0388, "step": 15980 }, { "grad_norm": 0.37930846214294434, "learning_rate": 9.398450377796815e-05, "loss": 0.0494, "step": 15990 }, { "grad_norm": 0.43741893768310547, "learning_rate": 9.397467122856616e-05, "loss": 0.0481, "step": 16000 }, { "grad_norm": 0.3491474986076355, "learning_rate": 9.396483116509558e-05, "loss": 0.0459, "step": 16010 }, { "grad_norm": 0.2968284785747528, "learning_rate": 9.39549835892378e-05, "loss": 0.0361, "step": 16020 }, { "grad_norm": 0.4472975730895996, "learning_rate": 9.39451285026755e-05, "loss": 0.0416, "step": 16030 }, { "grad_norm": 0.4750782549381256, "learning_rate": 9.393526590709262e-05, "loss": 0.0438, "step": 16040 }, { "grad_norm": 0.39635851979255676, "learning_rate": 9.392539580417444e-05, "loss": 0.0468, "step": 16050 }, { "grad_norm": 0.670904278755188, "learning_rate": 9.391551819560747e-05, "loss": 0.051, "step": 16060 }, { "grad_norm": 0.5024341940879822, "learning_rate": 9.390563308307955e-05, "loss": 0.0467, "step": 16070 }, { "grad_norm": 0.3572394549846649, "learning_rate": 9.389574046827974e-05, "loss": 0.0456, "step": 16080 }, { "grad_norm": 0.4971412718296051, "learning_rate": 9.388584035289845e-05, "loss": 0.0481, "step": 16090 }, { "grad_norm": 0.35381999611854553, "learning_rate": 9.387593273862732e-05, "loss": 0.05, "step": 16100 }, { "grad_norm": 0.38557174801826477, "learning_rate": 9.386601762715929e-05, "loss": 0.046, "step": 16110 }, { "grad_norm": 0.5669581294059753, "learning_rate": 9.38560950201886e-05, "loss": 0.0514, "step": 16120 }, { "grad_norm": 0.4291325509548187, "learning_rate": 9.384616491941071e-05, "loss": 0.045, "step": 16130 }, { "grad_norm": 0.43921545147895813, "learning_rate": 9.383622732652245e-05, "loss": 0.0395, "step": 16140 }, { "grad_norm": 0.3951163589954376, "learning_rate": 9.382628224322187e-05, "loss": 0.055, "step": 16150 }, { "grad_norm": 0.3365824520587921, "learning_rate": 9.381632967120829e-05, "loss": 0.0508, "step": 16160 }, { "grad_norm": 0.5898481607437134, "learning_rate": 9.380636961218235e-05, "loss": 0.0533, "step": 16170 }, { "grad_norm": 0.420830100774765, "learning_rate": 9.379640206784597e-05, "loss": 0.0399, "step": 16180 }, { "grad_norm": 0.5279235243797302, "learning_rate": 9.378642703990229e-05, "loss": 0.0442, "step": 16190 }, { "grad_norm": 0.3406590521335602, "learning_rate": 9.37764445300558e-05, "loss": 0.0379, "step": 16200 }, { "grad_norm": 0.34161585569381714, "learning_rate": 9.376645454001222e-05, "loss": 0.0548, "step": 16210 }, { "grad_norm": 0.38359469175338745, "learning_rate": 9.375645707147858e-05, "loss": 0.0464, "step": 16220 }, { "grad_norm": 0.41893166303634644, "learning_rate": 9.374645212616316e-05, "loss": 0.0635, "step": 16230 }, { "grad_norm": 0.7228212952613831, "learning_rate": 9.373643970577555e-05, "loss": 0.0619, "step": 16240 }, { "grad_norm": 0.5165945291519165, "learning_rate": 9.372641981202659e-05, "loss": 0.0532, "step": 16250 }, { "grad_norm": 0.5385264158248901, "learning_rate": 9.37163924466284e-05, "loss": 0.0564, "step": 16260 }, { "grad_norm": 0.37140777707099915, "learning_rate": 9.370635761129438e-05, "loss": 0.0451, "step": 16270 }, { "grad_norm": 0.4290462136268616, "learning_rate": 9.36963153077392e-05, "loss": 0.0534, "step": 16280 }, { "grad_norm": 0.4490756690502167, "learning_rate": 9.368626553767888e-05, "loss": 0.0477, "step": 16290 }, { "grad_norm": 0.43521520495414734, "learning_rate": 9.367620830283057e-05, "loss": 0.0461, "step": 16300 }, { "grad_norm": 0.6305320858955383, "learning_rate": 9.366614360491281e-05, "loss": 0.054, "step": 16310 }, { "grad_norm": 0.29867881536483765, "learning_rate": 9.365607144564539e-05, "loss": 0.0467, "step": 16320 }, { "grad_norm": 0.6049960255622864, "learning_rate": 9.364599182674934e-05, "loss": 0.0592, "step": 16330 }, { "grad_norm": 0.38929978013038635, "learning_rate": 9.3635904749947e-05, "loss": 0.0472, "step": 16340 }, { "grad_norm": 0.47107237577438354, "learning_rate": 9.362581021696202e-05, "loss": 0.0389, "step": 16350 }, { "grad_norm": 0.4429638087749481, "learning_rate": 9.361570822951921e-05, "loss": 0.0479, "step": 16360 }, { "grad_norm": 0.5020559430122375, "learning_rate": 9.360559878934476e-05, "loss": 0.0447, "step": 16370 }, { "grad_norm": 0.2850310206413269, "learning_rate": 9.359548189816611e-05, "loss": 0.0405, "step": 16380 }, { "grad_norm": 0.5543897151947021, "learning_rate": 9.358535755771193e-05, "loss": 0.0596, "step": 16390 }, { "grad_norm": 0.4533514380455017, "learning_rate": 9.357522576971221e-05, "loss": 0.0542, "step": 16400 }, { "grad_norm": 0.40216800570487976, "learning_rate": 9.356508653589819e-05, "loss": 0.051, "step": 16410 }, { "grad_norm": 0.4323791265487671, "learning_rate": 9.355493985800237e-05, "loss": 0.0428, "step": 16420 }, { "grad_norm": 0.5899932980537415, "learning_rate": 9.354478573775857e-05, "loss": 0.0533, "step": 16430 }, { "grad_norm": 0.4284421503543854, "learning_rate": 9.353462417690186e-05, "loss": 0.0468, "step": 16440 }, { "grad_norm": 0.42272067070007324, "learning_rate": 9.352445517716853e-05, "loss": 0.0405, "step": 16450 }, { "grad_norm": 0.39689451456069946, "learning_rate": 9.351427874029621e-05, "loss": 0.0561, "step": 16460 }, { "grad_norm": 0.5808475613594055, "learning_rate": 9.350409486802379e-05, "loss": 0.0468, "step": 16470 }, { "grad_norm": 0.4763403534889221, "learning_rate": 9.349390356209138e-05, "loss": 0.0532, "step": 16480 }, { "grad_norm": 0.3190506398677826, "learning_rate": 9.348370482424042e-05, "loss": 0.0533, "step": 16490 }, { "grad_norm": 0.5571724772453308, "learning_rate": 9.347349865621357e-05, "loss": 0.0466, "step": 16500 }, { "grad_norm": 0.38863298296928406, "learning_rate": 9.346328505975481e-05, "loss": 0.0501, "step": 16510 }, { "grad_norm": 0.45723381638526917, "learning_rate": 9.345306403660936e-05, "loss": 0.0592, "step": 16520 }, { "grad_norm": 0.49705153703689575, "learning_rate": 9.344283558852371e-05, "loss": 0.0471, "step": 16530 }, { "grad_norm": 0.46602216362953186, "learning_rate": 9.343259971724563e-05, "loss": 0.0516, "step": 16540 }, { "grad_norm": 0.48706257343292236, "learning_rate": 9.342235642452413e-05, "loss": 0.0554, "step": 16550 }, { "grad_norm": 0.5426137447357178, "learning_rate": 9.341210571210954e-05, "loss": 0.0611, "step": 16560 }, { "grad_norm": 0.5023689866065979, "learning_rate": 9.340184758175338e-05, "loss": 0.055, "step": 16570 }, { "grad_norm": 0.492335706949234, "learning_rate": 9.339158203520854e-05, "loss": 0.0597, "step": 16580 }, { "grad_norm": 0.3452984094619751, "learning_rate": 9.338130907422908e-05, "loss": 0.0438, "step": 16590 }, { "grad_norm": 0.2958562970161438, "learning_rate": 9.337102870057037e-05, "loss": 0.0397, "step": 16600 }, { "grad_norm": 0.476155549287796, "learning_rate": 9.336074091598907e-05, "loss": 0.0548, "step": 16610 }, { "grad_norm": 0.4023856520652771, "learning_rate": 9.335044572224306e-05, "loss": 0.0402, "step": 16620 }, { "grad_norm": 0.6112650632858276, "learning_rate": 9.334014312109151e-05, "loss": 0.0511, "step": 16630 }, { "grad_norm": 0.5041534304618835, "learning_rate": 9.332983311429486e-05, "loss": 0.0551, "step": 16640 }, { "grad_norm": 0.4511313736438751, "learning_rate": 9.33195157036148e-05, "loss": 0.0566, "step": 16650 }, { "grad_norm": 0.3704313635826111, "learning_rate": 9.330919089081432e-05, "loss": 0.0476, "step": 16660 }, { "grad_norm": 0.3136764466762543, "learning_rate": 9.32988586776576e-05, "loss": 0.0659, "step": 16670 }, { "grad_norm": 0.7494912147521973, "learning_rate": 9.328851906591016e-05, "loss": 0.045, "step": 16680 }, { "grad_norm": 0.496341347694397, "learning_rate": 9.327817205733875e-05, "loss": 0.0477, "step": 16690 }, { "grad_norm": 0.49554890394210815, "learning_rate": 9.326781765371142e-05, "loss": 0.0495, "step": 16700 }, { "grad_norm": 0.463453084230423, "learning_rate": 9.325745585679741e-05, "loss": 0.0556, "step": 16710 }, { "grad_norm": 0.44593068957328796, "learning_rate": 9.32470866683673e-05, "loss": 0.0612, "step": 16720 }, { "grad_norm": 0.5434566140174866, "learning_rate": 9.323671009019288e-05, "loss": 0.0684, "step": 16730 }, { "grad_norm": 0.314218670129776, "learning_rate": 9.322632612404725e-05, "loss": 0.0414, "step": 16740 }, { "grad_norm": 0.4485793113708496, "learning_rate": 9.321593477170471e-05, "loss": 0.0502, "step": 16750 }, { "grad_norm": 0.3321504592895508, "learning_rate": 9.320553603494088e-05, "loss": 0.0415, "step": 16760 }, { "grad_norm": 0.420813649892807, "learning_rate": 9.319512991553261e-05, "loss": 0.0438, "step": 16770 }, { "grad_norm": 0.4286409318447113, "learning_rate": 9.318471641525803e-05, "loss": 0.0436, "step": 16780 }, { "grad_norm": 0.4246872365474701, "learning_rate": 9.317429553589652e-05, "loss": 0.0413, "step": 16790 }, { "grad_norm": 0.4167291820049286, "learning_rate": 9.316386727922873e-05, "loss": 0.0439, "step": 16800 }, { "grad_norm": 0.39206722378730774, "learning_rate": 9.315343164703656e-05, "loss": 0.0377, "step": 16810 }, { "grad_norm": 0.523218035697937, "learning_rate": 9.314298864110316e-05, "loss": 0.0408, "step": 16820 }, { "grad_norm": 0.328228235244751, "learning_rate": 9.313253826321295e-05, "loss": 0.0376, "step": 16830 }, { "grad_norm": 0.4669145345687866, "learning_rate": 9.312208051515165e-05, "loss": 0.0384, "step": 16840 }, { "grad_norm": 0.39172372221946716, "learning_rate": 9.311161539870618e-05, "loss": 0.0413, "step": 16850 }, { "grad_norm": 0.46231284737586975, "learning_rate": 9.310114291566474e-05, "loss": 0.0523, "step": 16860 }, { "grad_norm": 0.4731047451496124, "learning_rate": 9.309066306781679e-05, "loss": 0.0396, "step": 16870 }, { "grad_norm": 0.41563889384269714, "learning_rate": 9.308017585695306e-05, "loss": 0.0415, "step": 16880 }, { "grad_norm": 0.5100786089897156, "learning_rate": 9.306968128486552e-05, "loss": 0.0438, "step": 16890 }, { "grad_norm": 0.522802472114563, "learning_rate": 9.30591793533474e-05, "loss": 0.0374, "step": 16900 }, { "grad_norm": 0.3422398567199707, "learning_rate": 9.304867006419321e-05, "loss": 0.0339, "step": 16910 }, { "grad_norm": 0.3281499147415161, "learning_rate": 9.303815341919868e-05, "loss": 0.0472, "step": 16920 }, { "grad_norm": 0.6428055763244629, "learning_rate": 9.302762942016084e-05, "loss": 0.0448, "step": 16930 }, { "grad_norm": 0.37502026557922363, "learning_rate": 9.301709806887792e-05, "loss": 0.0463, "step": 16940 }, { "grad_norm": 0.3185441195964813, "learning_rate": 9.300655936714948e-05, "loss": 0.0557, "step": 16950 }, { "grad_norm": 0.4252351224422455, "learning_rate": 9.299601331677627e-05, "loss": 0.0423, "step": 16960 }, { "grad_norm": 0.4419748783111572, "learning_rate": 9.298545991956033e-05, "loss": 0.046, "step": 16970 }, { "grad_norm": 0.437738299369812, "learning_rate": 9.297489917730493e-05, "loss": 0.0431, "step": 16980 }, { "grad_norm": 0.6360257267951965, "learning_rate": 9.296433109181464e-05, "loss": 0.0523, "step": 16990 }, { "grad_norm": 0.5724454522132874, "learning_rate": 9.295375566489523e-05, "loss": 0.0467, "step": 17000 }, { "grad_norm": 0.5719842314720154, "learning_rate": 9.294317289835379e-05, "loss": 0.0602, "step": 17010 }, { "grad_norm": 0.4534606337547302, "learning_rate": 9.293258279399859e-05, "loss": 0.0554, "step": 17020 }, { "grad_norm": 0.4268302321434021, "learning_rate": 9.292198535363919e-05, "loss": 0.0413, "step": 17030 }, { "grad_norm": 0.4107276499271393, "learning_rate": 9.291138057908641e-05, "loss": 0.0383, "step": 17040 }, { "grad_norm": 0.4164542257785797, "learning_rate": 9.290076847215234e-05, "loss": 0.0432, "step": 17050 }, { "grad_norm": 0.7705464363098145, "learning_rate": 9.289014903465025e-05, "loss": 0.049, "step": 17060 }, { "grad_norm": 0.43463262915611267, "learning_rate": 9.287952226839475e-05, "loss": 0.0427, "step": 17070 }, { "grad_norm": 0.48803576827049255, "learning_rate": 9.286888817520164e-05, "loss": 0.0465, "step": 17080 }, { "grad_norm": 0.4345944821834564, "learning_rate": 9.285824675688803e-05, "loss": 0.0464, "step": 17090 }, { "grad_norm": 0.39923059940338135, "learning_rate": 9.28475980152722e-05, "loss": 0.0444, "step": 17100 }, { "grad_norm": 0.49905747175216675, "learning_rate": 9.283694195217379e-05, "loss": 0.0504, "step": 17110 }, { "grad_norm": 0.392814964056015, "learning_rate": 9.282627856941356e-05, "loss": 0.0475, "step": 17120 }, { "grad_norm": 0.44895392656326294, "learning_rate": 9.281560786881363e-05, "loss": 0.0407, "step": 17130 }, { "grad_norm": 0.31787848472595215, "learning_rate": 9.280492985219733e-05, "loss": 0.0433, "step": 17140 }, { "grad_norm": 0.4187532961368561, "learning_rate": 9.279424452138924e-05, "loss": 0.0461, "step": 17150 }, { "grad_norm": 0.3758182227611542, "learning_rate": 9.278355187821517e-05, "loss": 0.0496, "step": 17160 }, { "grad_norm": 0.3448913097381592, "learning_rate": 9.277285192450224e-05, "loss": 0.045, "step": 17170 }, { "grad_norm": 0.406650185585022, "learning_rate": 9.276214466207875e-05, "loss": 0.0387, "step": 17180 }, { "grad_norm": 0.4754611551761627, "learning_rate": 9.275143009277427e-05, "loss": 0.0515, "step": 17190 }, { "grad_norm": 0.46639198064804077, "learning_rate": 9.274070821841964e-05, "loss": 0.0489, "step": 17200 }, { "grad_norm": 0.38182398676872253, "learning_rate": 9.272997904084696e-05, "loss": 0.0415, "step": 17210 }, { "grad_norm": 0.442173033952713, "learning_rate": 9.271924256188951e-05, "loss": 0.0497, "step": 17220 }, { "grad_norm": 0.4835744798183441, "learning_rate": 9.270849878338189e-05, "loss": 0.0454, "step": 17230 }, { "grad_norm": 0.36918243765830994, "learning_rate": 9.269774770715991e-05, "loss": 0.044, "step": 17240 }, { "grad_norm": 0.445242702960968, "learning_rate": 9.268698933506061e-05, "loss": 0.0636, "step": 17250 }, { "grad_norm": 0.2914799153804779, "learning_rate": 9.267622366892235e-05, "loss": 0.0455, "step": 17260 }, { "grad_norm": 0.4060397744178772, "learning_rate": 9.266545071058465e-05, "loss": 0.0429, "step": 17270 }, { "grad_norm": 0.45616984367370605, "learning_rate": 9.265467046188833e-05, "loss": 0.0453, "step": 17280 }, { "grad_norm": 0.4161578416824341, "learning_rate": 9.264388292467543e-05, "loss": 0.0476, "step": 17290 }, { "grad_norm": 0.42729946970939636, "learning_rate": 9.263308810078926e-05, "loss": 0.04, "step": 17300 }, { "grad_norm": 0.33883774280548096, "learning_rate": 9.262228599207434e-05, "loss": 0.0444, "step": 17310 }, { "grad_norm": 0.4254365861415863, "learning_rate": 9.261147660037647e-05, "loss": 0.0361, "step": 17320 }, { "grad_norm": 0.468013733625412, "learning_rate": 9.26006599275427e-05, "loss": 0.0396, "step": 17330 }, { "grad_norm": 0.4317743182182312, "learning_rate": 9.258983597542124e-05, "loss": 0.0506, "step": 17340 }, { "grad_norm": 0.3925175070762634, "learning_rate": 9.257900474586167e-05, "loss": 0.0501, "step": 17350 }, { "grad_norm": 0.420213907957077, "learning_rate": 9.256816624071471e-05, "loss": 0.0498, "step": 17360 }, { "grad_norm": 0.4493803083896637, "learning_rate": 9.25573204618324e-05, "loss": 0.0394, "step": 17370 }, { "grad_norm": 0.43801119923591614, "learning_rate": 9.254646741106796e-05, "loss": 0.043, "step": 17380 }, { "grad_norm": 0.46276751160621643, "learning_rate": 9.253560709027589e-05, "loss": 0.0414, "step": 17390 }, { "grad_norm": 0.3679051101207733, "learning_rate": 9.252473950131192e-05, "loss": 0.0466, "step": 17400 }, { "grad_norm": 0.5186166763305664, "learning_rate": 9.251386464603302e-05, "loss": 0.0362, "step": 17410 }, { "grad_norm": 0.4222957193851471, "learning_rate": 9.250298252629741e-05, "loss": 0.0416, "step": 17420 }, { "grad_norm": 0.48449087142944336, "learning_rate": 9.249209314396454e-05, "loss": 0.0393, "step": 17430 }, { "grad_norm": 0.5114085674285889, "learning_rate": 9.248119650089513e-05, "loss": 0.0524, "step": 17440 }, { "grad_norm": 0.4425680637359619, "learning_rate": 9.247029259895108e-05, "loss": 0.0479, "step": 17450 }, { "grad_norm": 0.3789154589176178, "learning_rate": 9.24593814399956e-05, "loss": 0.0416, "step": 17460 }, { "grad_norm": 0.305258572101593, "learning_rate": 9.244846302589309e-05, "loss": 0.0378, "step": 17470 }, { "grad_norm": 0.44897809624671936, "learning_rate": 9.243753735850923e-05, "loss": 0.0441, "step": 17480 }, { "grad_norm": 0.4127237796783447, "learning_rate": 9.24266044397109e-05, "loss": 0.0404, "step": 17490 }, { "grad_norm": 0.3701590299606323, "learning_rate": 9.241566427136624e-05, "loss": 0.0352, "step": 17500 }, { "grad_norm": 0.5097962617874146, "learning_rate": 9.240471685534463e-05, "loss": 0.0462, "step": 17510 }, { "grad_norm": 0.38371413946151733, "learning_rate": 9.239376219351667e-05, "loss": 0.0475, "step": 17520 }, { "grad_norm": 0.606538712978363, "learning_rate": 9.238280028775425e-05, "loss": 0.0438, "step": 17530 }, { "grad_norm": 0.6694178581237793, "learning_rate": 9.237183113993041e-05, "loss": 0.0529, "step": 17540 }, { "grad_norm": 0.5208279490470886, "learning_rate": 9.236085475191952e-05, "loss": 0.0446, "step": 17550 }, { "grad_norm": 0.5104385614395142, "learning_rate": 9.234987112559709e-05, "loss": 0.0605, "step": 17560 }, { "grad_norm": 0.6816447973251343, "learning_rate": 9.233888026283999e-05, "loss": 0.0565, "step": 17570 }, { "grad_norm": 0.5039337277412415, "learning_rate": 9.232788216552619e-05, "loss": 0.0458, "step": 17580 }, { "grad_norm": 0.41232943534851074, "learning_rate": 9.231687683553502e-05, "loss": 0.048, "step": 17590 }, { "grad_norm": 0.47105371952056885, "learning_rate": 9.230586427474698e-05, "loss": 0.0653, "step": 17600 }, { "grad_norm": 0.37166792154312134, "learning_rate": 9.229484448504379e-05, "loss": 0.0476, "step": 17610 }, { "grad_norm": 0.2997848093509674, "learning_rate": 9.228381746830843e-05, "loss": 0.0446, "step": 17620 }, { "grad_norm": 0.28804415464401245, "learning_rate": 9.227278322642514e-05, "loss": 0.0382, "step": 17630 }, { "grad_norm": 0.4794318377971649, "learning_rate": 9.226174176127937e-05, "loss": 0.0458, "step": 17640 }, { "grad_norm": 0.3847494423389435, "learning_rate": 9.22506930747578e-05, "loss": 0.0452, "step": 17650 }, { "grad_norm": 0.5178632736206055, "learning_rate": 9.223963716874831e-05, "loss": 0.0555, "step": 17660 }, { "grad_norm": 0.4829467236995697, "learning_rate": 9.222857404514012e-05, "loss": 0.053, "step": 17670 }, { "grad_norm": 0.4507746994495392, "learning_rate": 9.221750370582355e-05, "loss": 0.0421, "step": 17680 }, { "grad_norm": 0.46279841661453247, "learning_rate": 9.220642615269028e-05, "loss": 0.039, "step": 17690 }, { "grad_norm": 0.32735317945480347, "learning_rate": 9.219534138763311e-05, "loss": 0.0505, "step": 17700 }, { "grad_norm": 0.46799132227897644, "learning_rate": 9.218424941254613e-05, "loss": 0.0501, "step": 17710 }, { "grad_norm": 0.4591183066368103, "learning_rate": 9.217315022932468e-05, "loss": 0.0503, "step": 17720 }, { "grad_norm": 0.334475040435791, "learning_rate": 9.216204383986528e-05, "loss": 0.04, "step": 17730 }, { "grad_norm": 0.40922605991363525, "learning_rate": 9.215093024606574e-05, "loss": 0.0614, "step": 17740 }, { "grad_norm": 0.7197480201721191, "learning_rate": 9.213980944982506e-05, "loss": 0.0586, "step": 17750 }, { "grad_norm": 0.623845636844635, "learning_rate": 9.212868145304346e-05, "loss": 0.0602, "step": 17760 }, { "grad_norm": 0.3662109076976776, "learning_rate": 9.211754625762241e-05, "loss": 0.0405, "step": 17770 }, { "grad_norm": 0.4482369124889374, "learning_rate": 9.210640386546463e-05, "loss": 0.0547, "step": 17780 }, { "grad_norm": 0.494878351688385, "learning_rate": 9.209525427847405e-05, "loss": 0.0394, "step": 17790 }, { "grad_norm": 0.5748140215873718, "learning_rate": 9.208409749855583e-05, "loss": 0.0421, "step": 17800 }, { "grad_norm": 0.47451889514923096, "learning_rate": 9.207293352761633e-05, "loss": 0.0503, "step": 17810 }, { "grad_norm": 1.0738861560821533, "learning_rate": 9.206176236756319e-05, "loss": 0.0563, "step": 17820 }, { "grad_norm": 0.3079618811607361, "learning_rate": 9.205058402030525e-05, "loss": 0.0473, "step": 17830 }, { "grad_norm": 0.5396894216537476, "learning_rate": 9.203939848775259e-05, "loss": 0.047, "step": 17840 }, { "grad_norm": 0.45621511340141296, "learning_rate": 9.202820577181652e-05, "loss": 0.046, "step": 17850 }, { "grad_norm": 0.5379212498664856, "learning_rate": 9.201700587440953e-05, "loss": 0.0489, "step": 17860 }, { "grad_norm": 0.413372665643692, "learning_rate": 9.200579879744544e-05, "loss": 0.0449, "step": 17870 }, { "grad_norm": 0.8512937426567078, "learning_rate": 9.199458454283918e-05, "loss": 0.0489, "step": 17880 }, { "grad_norm": 0.2980969548225403, "learning_rate": 9.198336311250697e-05, "loss": 0.0423, "step": 17890 }, { "grad_norm": 0.6551498770713806, "learning_rate": 9.197213450836626e-05, "loss": 0.0497, "step": 17900 }, { "grad_norm": 0.6496584415435791, "learning_rate": 9.19608987323357e-05, "loss": 0.0516, "step": 17910 }, { "grad_norm": 0.5034224390983582, "learning_rate": 9.194965578633517e-05, "loss": 0.0554, "step": 17920 }, { "grad_norm": 0.376816987991333, "learning_rate": 9.193840567228582e-05, "loss": 0.0468, "step": 17930 }, { "grad_norm": 0.47689518332481384, "learning_rate": 9.192714839210994e-05, "loss": 0.0468, "step": 17940 }, { "grad_norm": 0.3967446982860565, "learning_rate": 9.19158839477311e-05, "loss": 0.0464, "step": 17950 }, { "grad_norm": 0.41894102096557617, "learning_rate": 9.190461234107411e-05, "loss": 0.0465, "step": 17960 }, { "grad_norm": 0.4696371853351593, "learning_rate": 9.189333357406496e-05, "loss": 0.0368, "step": 17970 }, { "grad_norm": 0.31553247570991516, "learning_rate": 9.188204764863089e-05, "loss": 0.052, "step": 17980 }, { "grad_norm": 0.5126311779022217, "learning_rate": 9.187075456670033e-05, "loss": 0.0374, "step": 17990 }, { "grad_norm": 0.27019166946411133, "learning_rate": 9.1859454330203e-05, "loss": 0.0516, "step": 18000 }, { "grad_norm": 0.40080270171165466, "learning_rate": 9.18481469410698e-05, "loss": 0.0551, "step": 18010 }, { "grad_norm": 0.4507271945476532, "learning_rate": 9.183683240123281e-05, "loss": 0.0492, "step": 18020 }, { "grad_norm": 0.5324230790138245, "learning_rate": 9.182551071262541e-05, "loss": 0.0457, "step": 18030 }, { "grad_norm": 0.468974232673645, "learning_rate": 9.181418187718218e-05, "loss": 0.0499, "step": 18040 }, { "grad_norm": 0.5135787129402161, "learning_rate": 9.180284589683888e-05, "loss": 0.0425, "step": 18050 }, { "grad_norm": 0.38507118821144104, "learning_rate": 9.17915027735325e-05, "loss": 0.0404, "step": 18060 }, { "grad_norm": 0.3894917964935303, "learning_rate": 9.178015250920133e-05, "loss": 0.0495, "step": 18070 }, { "grad_norm": 0.3814805746078491, "learning_rate": 9.176879510578477e-05, "loss": 0.0356, "step": 18080 }, { "grad_norm": 0.5248968601226807, "learning_rate": 9.17574305652235e-05, "loss": 0.0427, "step": 18090 }, { "grad_norm": 0.44174668192863464, "learning_rate": 9.174605888945942e-05, "loss": 0.0533, "step": 18100 }, { "grad_norm": 0.4936108887195587, "learning_rate": 9.173468008043564e-05, "loss": 0.0495, "step": 18110 }, { "grad_norm": 0.5072592496871948, "learning_rate": 9.172329414009648e-05, "loss": 0.0479, "step": 18120 }, { "grad_norm": 0.48226046562194824, "learning_rate": 9.171190107038747e-05, "loss": 0.0426, "step": 18130 }, { "grad_norm": 0.2945230305194855, "learning_rate": 9.170050087325541e-05, "loss": 0.0383, "step": 18140 }, { "grad_norm": 0.5210660696029663, "learning_rate": 9.168909355064824e-05, "loss": 0.0512, "step": 18150 }, { "grad_norm": 0.517433762550354, "learning_rate": 9.167767910451519e-05, "loss": 0.0424, "step": 18160 }, { "grad_norm": 0.3620709478855133, "learning_rate": 9.166625753680669e-05, "loss": 0.0436, "step": 18170 }, { "grad_norm": 0.5980707406997681, "learning_rate": 9.165482884947431e-05, "loss": 0.037, "step": 18180 }, { "grad_norm": 0.3366106152534485, "learning_rate": 9.164339304447098e-05, "loss": 0.045, "step": 18190 }, { "grad_norm": 0.373019278049469, "learning_rate": 9.163195012375072e-05, "loss": 0.0313, "step": 18200 }, { "grad_norm": 0.45461586117744446, "learning_rate": 9.16205000892688e-05, "loss": 0.0402, "step": 18210 }, { "grad_norm": 0.49301865696907043, "learning_rate": 9.160904294298175e-05, "loss": 0.0389, "step": 18220 }, { "grad_norm": 0.5099497437477112, "learning_rate": 9.159757868684727e-05, "loss": 0.0413, "step": 18230 }, { "grad_norm": 0.4835754930973053, "learning_rate": 9.15861073228243e-05, "loss": 0.042, "step": 18240 }, { "grad_norm": 0.45381057262420654, "learning_rate": 9.157462885287296e-05, "loss": 0.0495, "step": 18250 }, { "grad_norm": 0.40230023860931396, "learning_rate": 9.156314327895461e-05, "loss": 0.0435, "step": 18260 }, { "grad_norm": 0.4218267798423767, "learning_rate": 9.155165060303185e-05, "loss": 0.0415, "step": 18270 }, { "grad_norm": 0.4617537260055542, "learning_rate": 9.154015082706841e-05, "loss": 0.039, "step": 18280 }, { "grad_norm": 0.40193822979927063, "learning_rate": 9.152864395302936e-05, "loss": 0.0412, "step": 18290 }, { "grad_norm": 0.4235732853412628, "learning_rate": 9.151712998288085e-05, "loss": 0.0396, "step": 18300 }, { "grad_norm": 0.3374328911304474, "learning_rate": 9.150560891859031e-05, "loss": 0.0426, "step": 18310 }, { "grad_norm": 0.36794957518577576, "learning_rate": 9.14940807621264e-05, "loss": 0.0416, "step": 18320 }, { "grad_norm": 0.4037502408027649, "learning_rate": 9.148254551545894e-05, "loss": 0.0412, "step": 18330 }, { "grad_norm": 0.45682835578918457, "learning_rate": 9.147100318055901e-05, "loss": 0.0455, "step": 18340 }, { "grad_norm": 0.6011837720870972, "learning_rate": 9.145945375939888e-05, "loss": 0.0483, "step": 18350 }, { "grad_norm": 0.42995119094848633, "learning_rate": 9.144789725395203e-05, "loss": 0.0486, "step": 18360 }, { "grad_norm": 0.6007620096206665, "learning_rate": 9.14363336661931e-05, "loss": 0.0427, "step": 18370 }, { "grad_norm": 0.43467018008232117, "learning_rate": 9.142476299809806e-05, "loss": 0.0453, "step": 18380 }, { "grad_norm": 0.47521594166755676, "learning_rate": 9.1413185251644e-05, "loss": 0.0488, "step": 18390 }, { "grad_norm": 0.49052149057388306, "learning_rate": 9.140160042880923e-05, "loss": 0.0581, "step": 18400 }, { "grad_norm": 0.4717075228691101, "learning_rate": 9.139000853157327e-05, "loss": 0.055, "step": 18410 }, { "grad_norm": 0.4743734300136566, "learning_rate": 9.137840956191688e-05, "loss": 0.0478, "step": 18420 }, { "grad_norm": 0.5299521088600159, "learning_rate": 9.136680352182199e-05, "loss": 0.0457, "step": 18430 }, { "grad_norm": 0.5032039880752563, "learning_rate": 9.135519041327177e-05, "loss": 0.0465, "step": 18440 }, { "grad_norm": 0.4833422601222992, "learning_rate": 9.134357023825058e-05, "loss": 0.0534, "step": 18450 }, { "grad_norm": 0.5716691017150879, "learning_rate": 9.133194299874398e-05, "loss": 0.0494, "step": 18460 }, { "grad_norm": 0.6292024254798889, "learning_rate": 9.132030869673876e-05, "loss": 0.0507, "step": 18470 }, { "grad_norm": 0.47032785415649414, "learning_rate": 9.130866733422288e-05, "loss": 0.0518, "step": 18480 }, { "grad_norm": 0.4801740050315857, "learning_rate": 9.129701891318556e-05, "loss": 0.0435, "step": 18490 }, { "grad_norm": 0.27204784750938416, "learning_rate": 9.128536343561718e-05, "loss": 0.0582, "step": 18500 }, { "grad_norm": 0.40412741899490356, "learning_rate": 9.127370090350934e-05, "loss": 0.0417, "step": 18510 }, { "grad_norm": 0.5006762146949768, "learning_rate": 9.126203131885487e-05, "loss": 0.0452, "step": 18520 }, { "grad_norm": 0.4904908239841461, "learning_rate": 9.125035468364775e-05, "loss": 0.0397, "step": 18530 }, { "grad_norm": 0.3186231553554535, "learning_rate": 9.123867099988322e-05, "loss": 0.0331, "step": 18540 }, { "grad_norm": 0.3719456195831299, "learning_rate": 9.122698026955769e-05, "loss": 0.0422, "step": 18550 }, { "grad_norm": 0.3963716924190521, "learning_rate": 9.12152824946688e-05, "loss": 0.0713, "step": 18560 }, { "grad_norm": 0.5571818947792053, "learning_rate": 9.120357767721538e-05, "loss": 0.0511, "step": 18570 }, { "grad_norm": 0.5000022649765015, "learning_rate": 9.119186581919745e-05, "loss": 0.0453, "step": 18580 }, { "grad_norm": 0.5440114736557007, "learning_rate": 9.118014692261624e-05, "loss": 0.0442, "step": 18590 }, { "grad_norm": 0.4631412923336029, "learning_rate": 9.116842098947422e-05, "loss": 0.0592, "step": 18600 }, { "grad_norm": 0.3632313311100006, "learning_rate": 9.115668802177499e-05, "loss": 0.0385, "step": 18610 }, { "grad_norm": 0.38835638761520386, "learning_rate": 9.114494802152342e-05, "loss": 0.0544, "step": 18620 }, { "grad_norm": 0.6560245156288147, "learning_rate": 9.113320099072555e-05, "loss": 0.0697, "step": 18630 }, { "grad_norm": 0.49396422505378723, "learning_rate": 9.112144693138864e-05, "loss": 0.0454, "step": 18640 }, { "grad_norm": 0.46495917439460754, "learning_rate": 9.110968584552111e-05, "loss": 0.0406, "step": 18650 }, { "grad_norm": 0.3669126331806183, "learning_rate": 9.109791773513264e-05, "loss": 0.0473, "step": 18660 }, { "grad_norm": 0.45503926277160645, "learning_rate": 9.108614260223403e-05, "loss": 0.044, "step": 18670 }, { "grad_norm": 0.4424252212047577, "learning_rate": 9.107436044883738e-05, "loss": 0.0457, "step": 18680 }, { "grad_norm": 0.4587230086326599, "learning_rate": 9.10625712769559e-05, "loss": 0.0402, "step": 18690 }, { "grad_norm": 0.3677635192871094, "learning_rate": 9.105077508860406e-05, "loss": 0.0379, "step": 18700 }, { "grad_norm": 0.39835748076438904, "learning_rate": 9.103897188579751e-05, "loss": 0.051, "step": 18710 }, { "grad_norm": 0.45010998845100403, "learning_rate": 9.102716167055308e-05, "loss": 0.0375, "step": 18720 }, { "grad_norm": 0.5988395810127258, "learning_rate": 9.10153444448888e-05, "loss": 0.0701, "step": 18730 }, { "grad_norm": 0.5226265788078308, "learning_rate": 9.100352021082393e-05, "loss": 0.0435, "step": 18740 }, { "grad_norm": 0.47029873728752136, "learning_rate": 9.099168897037891e-05, "loss": 0.0434, "step": 18750 }, { "grad_norm": 0.5241556167602539, "learning_rate": 9.097985072557538e-05, "loss": 0.0405, "step": 18760 }, { "grad_norm": 0.34605494141578674, "learning_rate": 9.096800547843615e-05, "loss": 0.0631, "step": 18770 }, { "grad_norm": 0.4389539361000061, "learning_rate": 9.095615323098526e-05, "loss": 0.0514, "step": 18780 }, { "grad_norm": 0.5388430953025818, "learning_rate": 9.094429398524795e-05, "loss": 0.0375, "step": 18790 }, { "grad_norm": 0.4556245505809784, "learning_rate": 9.093242774325061e-05, "loss": 0.04, "step": 18800 }, { "grad_norm": 0.23013539612293243, "learning_rate": 9.092055450702088e-05, "loss": 0.0378, "step": 18810 }, { "grad_norm": 0.45522111654281616, "learning_rate": 9.090867427858756e-05, "loss": 0.0405, "step": 18820 }, { "grad_norm": 0.47412005066871643, "learning_rate": 9.089678705998066e-05, "loss": 0.0376, "step": 18830 }, { "grad_norm": 0.4032900929450989, "learning_rate": 9.088489285323139e-05, "loss": 0.0413, "step": 18840 }, { "grad_norm": 0.5091952681541443, "learning_rate": 9.087299166037212e-05, "loss": 0.0449, "step": 18850 }, { "grad_norm": 0.5385580062866211, "learning_rate": 9.086108348343647e-05, "loss": 0.0503, "step": 18860 }, { "grad_norm": 0.5221367478370667, "learning_rate": 9.08491683244592e-05, "loss": 0.0481, "step": 18870 }, { "grad_norm": 0.27915826439857483, "learning_rate": 9.08372461854763e-05, "loss": 0.0391, "step": 18880 }, { "grad_norm": 0.571182370185852, "learning_rate": 9.082531706852492e-05, "loss": 0.0418, "step": 18890 }, { "grad_norm": 0.4325554072856903, "learning_rate": 9.081338097564342e-05, "loss": 0.0506, "step": 18900 }, { "grad_norm": 0.5100898742675781, "learning_rate": 9.080143790887137e-05, "loss": 0.0441, "step": 18910 }, { "grad_norm": 0.4606996476650238, "learning_rate": 9.07894878702495e-05, "loss": 0.0373, "step": 18920 }, { "grad_norm": 0.345101535320282, "learning_rate": 9.077753086181974e-05, "loss": 0.0533, "step": 18930 }, { "grad_norm": 0.30766645073890686, "learning_rate": 9.076556688562524e-05, "loss": 0.0331, "step": 18940 }, { "grad_norm": 0.5616613030433655, "learning_rate": 9.075359594371029e-05, "loss": 0.0336, "step": 18950 }, { "grad_norm": 0.4223569631576538, "learning_rate": 9.07416180381204e-05, "loss": 0.0341, "step": 18960 }, { "grad_norm": 0.41901537775993347, "learning_rate": 9.072963317090228e-05, "loss": 0.0382, "step": 18970 }, { "grad_norm": 0.5116147994995117, "learning_rate": 9.071764134410382e-05, "loss": 0.0401, "step": 18980 }, { "grad_norm": 0.5832217931747437, "learning_rate": 9.070564255977407e-05, "loss": 0.0364, "step": 18990 }, { "grad_norm": 0.4735008180141449, "learning_rate": 9.06936368199633e-05, "loss": 0.0552, "step": 19000 }, { "grad_norm": 0.4882502555847168, "learning_rate": 9.0681624126723e-05, "loss": 0.0507, "step": 19010 }, { "grad_norm": 0.4299039840698242, "learning_rate": 9.066960448210576e-05, "loss": 0.0416, "step": 19020 }, { "grad_norm": 0.553083062171936, "learning_rate": 9.065757788816543e-05, "loss": 0.0427, "step": 19030 }, { "grad_norm": 0.6690390706062317, "learning_rate": 9.064554434695705e-05, "loss": 0.0551, "step": 19040 }, { "grad_norm": 0.4725050926208496, "learning_rate": 9.063350386053677e-05, "loss": 0.0484, "step": 19050 }, { "grad_norm": 0.40329909324645996, "learning_rate": 9.062145643096202e-05, "loss": 0.0418, "step": 19060 }, { "grad_norm": 0.35324111580848694, "learning_rate": 9.060940206029136e-05, "loss": 0.048, "step": 19070 }, { "grad_norm": 0.6121883392333984, "learning_rate": 9.059734075058457e-05, "loss": 0.0414, "step": 19080 }, { "grad_norm": 0.28556007146835327, "learning_rate": 9.058527250390257e-05, "loss": 0.0497, "step": 19090 }, { "grad_norm": 0.3507901430130005, "learning_rate": 9.057319732230752e-05, "loss": 0.0573, "step": 19100 }, { "grad_norm": 0.3061073422431946, "learning_rate": 9.056111520786273e-05, "loss": 0.0453, "step": 19110 }, { "grad_norm": 0.3310404121875763, "learning_rate": 9.054902616263268e-05, "loss": 0.0371, "step": 19120 }, { "grad_norm": 0.293076753616333, "learning_rate": 9.05369301886831e-05, "loss": 0.0537, "step": 19130 }, { "grad_norm": 0.4839780032634735, "learning_rate": 9.052482728808083e-05, "loss": 0.0423, "step": 19140 }, { "grad_norm": 0.37127164006233215, "learning_rate": 9.051271746289391e-05, "loss": 0.0257, "step": 19150 }, { "grad_norm": 0.4036156237125397, "learning_rate": 9.050060071519162e-05, "loss": 0.0396, "step": 19160 }, { "grad_norm": 0.4002767503261566, "learning_rate": 9.048847704704437e-05, "loss": 0.0339, "step": 19170 }, { "grad_norm": 0.3719679117202759, "learning_rate": 9.047634646052376e-05, "loss": 0.0419, "step": 19180 }, { "grad_norm": 0.4830949604511261, "learning_rate": 9.046420895770256e-05, "loss": 0.0503, "step": 19190 }, { "grad_norm": 0.24486733973026276, "learning_rate": 9.045206454065473e-05, "loss": 0.0329, "step": 19200 }, { "grad_norm": 0.4006473124027252, "learning_rate": 9.043991321145546e-05, "loss": 0.0349, "step": 19210 }, { "grad_norm": 0.4699331820011139, "learning_rate": 9.042775497218105e-05, "loss": 0.0347, "step": 19220 }, { "grad_norm": 0.4364800453186035, "learning_rate": 9.041558982490901e-05, "loss": 0.0428, "step": 19230 }, { "grad_norm": 0.3894549608230591, "learning_rate": 9.040341777171805e-05, "loss": 0.0425, "step": 19240 }, { "grad_norm": 0.3463621437549591, "learning_rate": 9.039123881468802e-05, "loss": 0.0348, "step": 19250 }, { "grad_norm": 0.39422300457954407, "learning_rate": 9.037905295589998e-05, "loss": 0.0412, "step": 19260 }, { "grad_norm": 0.4472205936908722, "learning_rate": 9.036686019743617e-05, "loss": 0.0349, "step": 19270 }, { "grad_norm": 0.3648942708969116, "learning_rate": 9.035466054137997e-05, "loss": 0.0377, "step": 19280 }, { "grad_norm": 0.3384242057800293, "learning_rate": 9.0342453989816e-05, "loss": 0.0456, "step": 19290 }, { "grad_norm": 0.3010440468788147, "learning_rate": 9.033024054483e-05, "loss": 0.0419, "step": 19300 }, { "grad_norm": 0.3881862163543701, "learning_rate": 9.031802020850894e-05, "loss": 0.0404, "step": 19310 }, { "grad_norm": 0.4212355613708496, "learning_rate": 9.030579298294092e-05, "loss": 0.052, "step": 19320 }, { "grad_norm": 0.4180741608142853, "learning_rate": 9.029355887021524e-05, "loss": 0.055, "step": 19330 }, { "grad_norm": 0.525132954120636, "learning_rate": 9.028131787242238e-05, "loss": 0.0635, "step": 19340 }, { "grad_norm": 0.4699709117412567, "learning_rate": 9.026906999165399e-05, "loss": 0.0508, "step": 19350 }, { "grad_norm": 0.2915371060371399, "learning_rate": 9.025681523000291e-05, "loss": 0.037, "step": 19360 }, { "grad_norm": 0.4748466908931732, "learning_rate": 9.024455358956315e-05, "loss": 0.0422, "step": 19370 }, { "grad_norm": 0.39150822162628174, "learning_rate": 9.023228507242984e-05, "loss": 0.0464, "step": 19380 }, { "grad_norm": 0.27862632274627686, "learning_rate": 9.022000968069937e-05, "loss": 0.0381, "step": 19390 }, { "grad_norm": 0.6994941830635071, "learning_rate": 9.020772741646928e-05, "loss": 0.0404, "step": 19400 }, { "grad_norm": 0.46917566657066345, "learning_rate": 9.019543828183826e-05, "loss": 0.041, "step": 19410 }, { "grad_norm": 0.4021453857421875, "learning_rate": 9.018314227890616e-05, "loss": 0.0394, "step": 19420 }, { "grad_norm": 0.4759809374809265, "learning_rate": 9.017083940977408e-05, "loss": 0.0551, "step": 19430 }, { "grad_norm": 0.3129808306694031, "learning_rate": 9.015852967654422e-05, "loss": 0.04, "step": 19440 }, { "grad_norm": 0.499656081199646, "learning_rate": 9.014621308131996e-05, "loss": 0.0328, "step": 19450 }, { "grad_norm": 0.44096916913986206, "learning_rate": 9.01338896262059e-05, "loss": 0.0456, "step": 19460 }, { "grad_norm": 0.4644589424133301, "learning_rate": 9.012155931330777e-05, "loss": 0.0517, "step": 19470 }, { "grad_norm": 0.414767324924469, "learning_rate": 9.010922214473246e-05, "loss": 0.0358, "step": 19480 }, { "grad_norm": 0.3981494903564453, "learning_rate": 9.009687812258808e-05, "loss": 0.0553, "step": 19490 }, { "grad_norm": 0.3665803074836731, "learning_rate": 9.00845272489839e-05, "loss": 0.043, "step": 19500 }, { "grad_norm": 0.4312187433242798, "learning_rate": 9.007216952603031e-05, "loss": 0.0563, "step": 19510 }, { "grad_norm": 0.3295075297355652, "learning_rate": 9.005980495583894e-05, "loss": 0.042, "step": 19520 }, { "grad_norm": 0.4126676023006439, "learning_rate": 9.004743354052252e-05, "loss": 0.0442, "step": 19530 }, { "grad_norm": 0.5335502028465271, "learning_rate": 9.003505528219503e-05, "loss": 0.0401, "step": 19540 }, { "grad_norm": 0.4486042559146881, "learning_rate": 9.002267018297154e-05, "loss": 0.0452, "step": 19550 }, { "grad_norm": 0.42409753799438477, "learning_rate": 9.001027824496834e-05, "loss": 0.0456, "step": 19560 }, { "grad_norm": 0.266405314207077, "learning_rate": 8.999787947030287e-05, "loss": 0.0439, "step": 19570 }, { "grad_norm": 0.46993815898895264, "learning_rate": 8.998547386109376e-05, "loss": 0.0381, "step": 19580 }, { "grad_norm": 0.3942272961139679, "learning_rate": 8.997306141946073e-05, "loss": 0.0418, "step": 19590 }, { "grad_norm": 0.42628517746925354, "learning_rate": 8.996064214752481e-05, "loss": 0.0452, "step": 19600 }, { "grad_norm": 0.3628118336200714, "learning_rate": 8.994821604740806e-05, "loss": 0.0416, "step": 19610 }, { "grad_norm": 0.3887074589729309, "learning_rate": 8.993578312123377e-05, "loss": 0.0426, "step": 19620 }, { "grad_norm": 0.32702821493148804, "learning_rate": 8.992334337112639e-05, "loss": 0.0444, "step": 19630 }, { "grad_norm": 0.3690127432346344, "learning_rate": 8.991089679921154e-05, "loss": 0.0307, "step": 19640 }, { "grad_norm": 0.4446401000022888, "learning_rate": 8.989844340761599e-05, "loss": 0.0382, "step": 19650 }, { "grad_norm": 0.4752258062362671, "learning_rate": 8.988598319846768e-05, "loss": 0.0375, "step": 19660 }, { "grad_norm": 0.36194509267807007, "learning_rate": 8.987351617389574e-05, "loss": 0.0342, "step": 19670 }, { "grad_norm": 0.4079916775226593, "learning_rate": 8.98610423360304e-05, "loss": 0.0362, "step": 19680 }, { "grad_norm": 0.429412841796875, "learning_rate": 8.984856168700317e-05, "loss": 0.0495, "step": 19690 }, { "grad_norm": 0.40929391980171204, "learning_rate": 8.983607422894658e-05, "loss": 0.038, "step": 19700 }, { "grad_norm": 0.41225019097328186, "learning_rate": 8.982357996399442e-05, "loss": 0.0452, "step": 19710 }, { "grad_norm": 0.41892048716545105, "learning_rate": 8.981107889428164e-05, "loss": 0.0513, "step": 19720 }, { "grad_norm": 0.3406420648097992, "learning_rate": 8.979857102194428e-05, "loss": 0.0392, "step": 19730 }, { "grad_norm": 0.386940062046051, "learning_rate": 8.978605634911968e-05, "loss": 0.0439, "step": 19740 }, { "grad_norm": 0.34032633900642395, "learning_rate": 8.977353487794616e-05, "loss": 0.0347, "step": 19750 }, { "grad_norm": 0.46985700726509094, "learning_rate": 8.976100661056334e-05, "loss": 0.0496, "step": 19760 }, { "grad_norm": 0.40302926301956177, "learning_rate": 8.974847154911197e-05, "loss": 0.034, "step": 19770 }, { "grad_norm": 0.40561985969543457, "learning_rate": 8.973592969573393e-05, "loss": 0.0478, "step": 19780 }, { "grad_norm": 0.47962310910224915, "learning_rate": 8.972338105257228e-05, "loss": 0.0498, "step": 19790 }, { "grad_norm": 0.3388932943344116, "learning_rate": 8.971082562177125e-05, "loss": 0.0384, "step": 19800 }, { "grad_norm": 0.4924905598163605, "learning_rate": 8.96982634054762e-05, "loss": 0.0403, "step": 19810 }, { "grad_norm": 0.41669368743896484, "learning_rate": 8.96856944058337e-05, "loss": 0.0359, "step": 19820 }, { "grad_norm": 0.3180446922779083, "learning_rate": 8.967311862499144e-05, "loss": 0.0269, "step": 19830 }, { "grad_norm": 0.30073854327201843, "learning_rate": 8.966053606509825e-05, "loss": 0.0354, "step": 19840 }, { "grad_norm": 0.5405938029289246, "learning_rate": 8.964794672830417e-05, "loss": 0.0495, "step": 19850 }, { "grad_norm": 0.4571440517902374, "learning_rate": 8.963535061676038e-05, "loss": 0.0539, "step": 19860 }, { "grad_norm": 0.5467488169670105, "learning_rate": 8.962274773261918e-05, "loss": 0.0446, "step": 19870 }, { "grad_norm": 0.48885366320610046, "learning_rate": 8.961013807803409e-05, "loss": 0.0518, "step": 19880 }, { "grad_norm": 0.6212040185928345, "learning_rate": 8.959752165515973e-05, "loss": 0.0465, "step": 19890 }, { "grad_norm": 0.6106945872306824, "learning_rate": 8.958489846615193e-05, "loss": 0.0619, "step": 19900 }, { "grad_norm": 0.533935010433197, "learning_rate": 8.957226851316762e-05, "loss": 0.0395, "step": 19910 }, { "grad_norm": 0.3684922754764557, "learning_rate": 8.955963179836493e-05, "loss": 0.044, "step": 19920 }, { "grad_norm": 0.36660894751548767, "learning_rate": 8.954698832390312e-05, "loss": 0.041, "step": 19930 }, { "grad_norm": 0.40803682804107666, "learning_rate": 8.953433809194263e-05, "loss": 0.0415, "step": 19940 }, { "grad_norm": 0.3980361819267273, "learning_rate": 8.9521681104645e-05, "loss": 0.0348, "step": 19950 }, { "grad_norm": 2.2580599784851074, "learning_rate": 8.9509017364173e-05, "loss": 0.0492, "step": 19960 }, { "grad_norm": 0.6053842902183533, "learning_rate": 8.949634687269052e-05, "loss": 0.0402, "step": 19970 }, { "grad_norm": 0.44506749510765076, "learning_rate": 8.948366963236259e-05, "loss": 0.053, "step": 19980 }, { "grad_norm": 0.39221036434173584, "learning_rate": 8.947098564535538e-05, "loss": 0.038, "step": 19990 }, { "grad_norm": 0.4727820158004761, "learning_rate": 8.945829491383627e-05, "loss": 0.0403, "step": 20000 }, { "grad_norm": 0.4128730595111847, "learning_rate": 8.944559743997374e-05, "loss": 0.041, "step": 20010 }, { "grad_norm": 0.25189995765686035, "learning_rate": 8.943289322593746e-05, "loss": 0.0401, "step": 20020 }, { "grad_norm": 0.46559396386146545, "learning_rate": 8.942018227389821e-05, "loss": 0.0367, "step": 20030 }, { "grad_norm": 0.5255997776985168, "learning_rate": 8.940746458602795e-05, "loss": 0.0629, "step": 20040 }, { "grad_norm": 0.45725640654563904, "learning_rate": 8.939474016449979e-05, "loss": 0.0489, "step": 20050 }, { "grad_norm": 0.34900256991386414, "learning_rate": 8.938200901148799e-05, "loss": 0.0388, "step": 20060 }, { "grad_norm": 0.5780149102210999, "learning_rate": 8.936927112916795e-05, "loss": 0.0471, "step": 20070 }, { "grad_norm": 0.40647387504577637, "learning_rate": 8.935652651971622e-05, "loss": 0.0377, "step": 20080 }, { "grad_norm": 0.5297650694847107, "learning_rate": 8.934377518531052e-05, "loss": 0.0383, "step": 20090 }, { "grad_norm": 0.47056761384010315, "learning_rate": 8.933101712812967e-05, "loss": 0.0424, "step": 20100 }, { "grad_norm": 0.330459326505661, "learning_rate": 8.931825235035374e-05, "loss": 0.0437, "step": 20110 }, { "grad_norm": 0.33655741810798645, "learning_rate": 8.930548085416382e-05, "loss": 0.038, "step": 20120 }, { "grad_norm": 0.32427290081977844, "learning_rate": 8.92927026417422e-05, "loss": 0.0385, "step": 20130 }, { "grad_norm": 0.4064313471317291, "learning_rate": 8.92799177152724e-05, "loss": 0.034, "step": 20140 }, { "grad_norm": 0.4248999357223511, "learning_rate": 8.926712607693895e-05, "loss": 0.0395, "step": 20150 }, { "grad_norm": 0.6447193622589111, "learning_rate": 8.925432772892762e-05, "loss": 0.048, "step": 20160 }, { "grad_norm": 0.5298957824707031, "learning_rate": 8.924152267342529e-05, "loss": 0.0548, "step": 20170 }, { "grad_norm": 0.43138259649276733, "learning_rate": 8.922871091261998e-05, "loss": 0.0499, "step": 20180 }, { "grad_norm": 0.3154129385948181, "learning_rate": 8.92158924487009e-05, "loss": 0.0397, "step": 20190 }, { "grad_norm": 0.40767672657966614, "learning_rate": 8.920306728385834e-05, "loss": 0.0358, "step": 20200 }, { "grad_norm": 0.5527401566505432, "learning_rate": 8.919023542028379e-05, "loss": 0.0342, "step": 20210 }, { "grad_norm": 0.30589717626571655, "learning_rate": 8.917739686016988e-05, "loss": 0.0393, "step": 20220 }, { "grad_norm": 0.4575427174568176, "learning_rate": 8.916455160571033e-05, "loss": 0.0579, "step": 20230 }, { "grad_norm": 0.41978001594543457, "learning_rate": 8.915169965910008e-05, "loss": 0.0331, "step": 20240 }, { "grad_norm": 0.3799489736557007, "learning_rate": 8.913884102253514e-05, "loss": 0.0385, "step": 20250 }, { "grad_norm": 0.3259912431240082, "learning_rate": 8.912597569821273e-05, "loss": 0.0334, "step": 20260 }, { "grad_norm": 0.32520681619644165, "learning_rate": 8.911310368833118e-05, "loss": 0.0381, "step": 20270 }, { "grad_norm": 0.399837464094162, "learning_rate": 8.910022499508994e-05, "loss": 0.0311, "step": 20280 }, { "grad_norm": 0.29777392745018005, "learning_rate": 8.908733962068965e-05, "loss": 0.0398, "step": 20290 }, { "grad_norm": 0.3244324326515198, "learning_rate": 8.907444756733207e-05, "loss": 0.0299, "step": 20300 }, { "grad_norm": 0.44158172607421875, "learning_rate": 8.906154883722006e-05, "loss": 0.0436, "step": 20310 }, { "grad_norm": 0.3440762460231781, "learning_rate": 8.904864343255773e-05, "loss": 0.0412, "step": 20320 }, { "grad_norm": 0.45317742228507996, "learning_rate": 8.90357313555502e-05, "loss": 0.0412, "step": 20330 }, { "grad_norm": 0.33718183636665344, "learning_rate": 8.90228126084038e-05, "loss": 0.0449, "step": 20340 }, { "grad_norm": 0.3618198335170746, "learning_rate": 8.900988719332601e-05, "loss": 0.0472, "step": 20350 }, { "grad_norm": 0.2814125716686249, "learning_rate": 8.899695511252542e-05, "loss": 0.0441, "step": 20360 }, { "grad_norm": 0.39855122566223145, "learning_rate": 8.898401636821176e-05, "loss": 0.0501, "step": 20370 }, { "grad_norm": 0.5621644854545593, "learning_rate": 8.897107096259593e-05, "loss": 0.0558, "step": 20380 }, { "grad_norm": 0.39733070135116577, "learning_rate": 8.895811889788994e-05, "loss": 0.0464, "step": 20390 }, { "grad_norm": 0.5178106427192688, "learning_rate": 8.894516017630692e-05, "loss": 0.0473, "step": 20400 }, { "grad_norm": 0.47382187843322754, "learning_rate": 8.893219480006118e-05, "loss": 0.0443, "step": 20410 }, { "grad_norm": 0.428452730178833, "learning_rate": 8.891922277136817e-05, "loss": 0.0411, "step": 20420 }, { "grad_norm": 0.4365878701210022, "learning_rate": 8.890624409244441e-05, "loss": 0.0359, "step": 20430 }, { "grad_norm": 0.4242461323738098, "learning_rate": 8.889325876550763e-05, "loss": 0.0441, "step": 20440 }, { "grad_norm": 0.27130329608917236, "learning_rate": 8.888026679277666e-05, "loss": 0.0352, "step": 20450 }, { "grad_norm": 0.4063988924026489, "learning_rate": 8.886726817647147e-05, "loss": 0.0624, "step": 20460 }, { "grad_norm": 0.42650702595710754, "learning_rate": 8.885426291881319e-05, "loss": 0.056, "step": 20470 }, { "grad_norm": 0.35970455408096313, "learning_rate": 8.884125102202401e-05, "loss": 0.0412, "step": 20480 }, { "grad_norm": 0.3473772704601288, "learning_rate": 8.882823248832736e-05, "loss": 0.0276, "step": 20490 }, { "grad_norm": 0.464207261800766, "learning_rate": 8.881520731994772e-05, "loss": 0.0376, "step": 20500 }, { "grad_norm": 0.3169001042842865, "learning_rate": 8.880217551911077e-05, "loss": 0.0348, "step": 20510 }, { "grad_norm": 0.4107460379600525, "learning_rate": 8.878913708804323e-05, "loss": 0.0315, "step": 20520 }, { "grad_norm": 0.33481279015541077, "learning_rate": 8.877609202897308e-05, "loss": 0.0447, "step": 20530 }, { "grad_norm": 0.3438739478588104, "learning_rate": 8.876304034412933e-05, "loss": 0.0295, "step": 20540 }, { "grad_norm": 0.4393289089202881, "learning_rate": 8.874998203574214e-05, "loss": 0.045, "step": 20550 }, { "grad_norm": 0.4310166835784912, "learning_rate": 8.873691710604284e-05, "loss": 0.0393, "step": 20560 }, { "grad_norm": 0.4094174802303314, "learning_rate": 8.872384555726387e-05, "loss": 0.0381, "step": 20570 }, { "grad_norm": 0.45462286472320557, "learning_rate": 8.871076739163878e-05, "loss": 0.0375, "step": 20580 }, { "grad_norm": 0.40198975801467896, "learning_rate": 8.86976826114023e-05, "loss": 0.0383, "step": 20590 }, { "grad_norm": 0.41892170906066895, "learning_rate": 8.868459121879023e-05, "loss": 0.0417, "step": 20600 }, { "grad_norm": 0.5405136942863464, "learning_rate": 8.867149321603956e-05, "loss": 0.0469, "step": 20610 }, { "grad_norm": 0.33323314785957336, "learning_rate": 8.865838860538835e-05, "loss": 0.0393, "step": 20620 }, { "grad_norm": 0.4961497485637665, "learning_rate": 8.864527738907585e-05, "loss": 0.0401, "step": 20630 }, { "grad_norm": 0.4773169457912445, "learning_rate": 8.863215956934239e-05, "loss": 0.0491, "step": 20640 }, { "grad_norm": 0.3450741767883301, "learning_rate": 8.861903514842947e-05, "loss": 0.0322, "step": 20650 }, { "grad_norm": 0.29975149035453796, "learning_rate": 8.860590412857966e-05, "loss": 0.0366, "step": 20660 }, { "grad_norm": 0.3629112243652344, "learning_rate": 8.85927665120367e-05, "loss": 0.0371, "step": 20670 }, { "grad_norm": 0.4126106798648834, "learning_rate": 8.857962230104546e-05, "loss": 0.0334, "step": 20680 }, { "grad_norm": 0.41341835260391235, "learning_rate": 8.856647149785193e-05, "loss": 0.0394, "step": 20690 }, { "grad_norm": 0.3170984089374542, "learning_rate": 8.855331410470322e-05, "loss": 0.0402, "step": 20700 }, { "grad_norm": 0.6713211536407471, "learning_rate": 8.854015012384756e-05, "loss": 0.0356, "step": 20710 }, { "grad_norm": 0.9906827211380005, "learning_rate": 8.852697955753433e-05, "loss": 0.0508, "step": 20720 }, { "grad_norm": 0.5635508298873901, "learning_rate": 8.851380240801399e-05, "loss": 0.0472, "step": 20730 }, { "grad_norm": 0.42696529626846313, "learning_rate": 8.850061867753818e-05, "loss": 0.0344, "step": 20740 }, { "grad_norm": 0.45098876953125, "learning_rate": 8.848742836835963e-05, "loss": 0.0557, "step": 20750 }, { "grad_norm": 0.40836548805236816, "learning_rate": 8.847423148273221e-05, "loss": 0.0383, "step": 20760 }, { "grad_norm": 0.30196669697761536, "learning_rate": 8.846102802291092e-05, "loss": 0.0473, "step": 20770 }, { "grad_norm": 0.4786294102668762, "learning_rate": 8.844781799115183e-05, "loss": 0.0435, "step": 20780 }, { "grad_norm": 0.411031574010849, "learning_rate": 8.84346013897122e-05, "loss": 0.0342, "step": 20790 }, { "grad_norm": 0.39238110184669495, "learning_rate": 8.842137822085038e-05, "loss": 0.0475, "step": 20800 }, { "grad_norm": 0.45646968483924866, "learning_rate": 8.840814848682585e-05, "loss": 0.0332, "step": 20810 }, { "grad_norm": 0.3189028203487396, "learning_rate": 8.83949121898992e-05, "loss": 0.0478, "step": 20820 }, { "grad_norm": 0.4595479667186737, "learning_rate": 8.838166933233217e-05, "loss": 0.0321, "step": 20830 }, { "grad_norm": 0.46122804284095764, "learning_rate": 8.83684199163876e-05, "loss": 0.0469, "step": 20840 }, { "grad_norm": 0.36059483885765076, "learning_rate": 8.835516394432943e-05, "loss": 0.0484, "step": 20850 }, { "grad_norm": 0.4387120008468628, "learning_rate": 8.834190141842276e-05, "loss": 0.0411, "step": 20860 }, { "grad_norm": 0.44443029165267944, "learning_rate": 8.83286323409338e-05, "loss": 0.0358, "step": 20870 }, { "grad_norm": 0.3778626620769501, "learning_rate": 8.831535671412986e-05, "loss": 0.0427, "step": 20880 }, { "grad_norm": 0.4406082034111023, "learning_rate": 8.830207454027938e-05, "loss": 0.0432, "step": 20890 }, { "grad_norm": 0.40795600414276123, "learning_rate": 8.828878582165192e-05, "loss": 0.0391, "step": 20900 }, { "grad_norm": 0.35057663917541504, "learning_rate": 8.827549056051818e-05, "loss": 0.0388, "step": 20910 }, { "grad_norm": 0.3684220016002655, "learning_rate": 8.826218875914993e-05, "loss": 0.0361, "step": 20920 }, { "grad_norm": 0.3657449781894684, "learning_rate": 8.82488804198201e-05, "loss": 0.0418, "step": 20930 }, { "grad_norm": 0.24803192913532257, "learning_rate": 8.82355655448027e-05, "loss": 0.0362, "step": 20940 }, { "grad_norm": 0.46448683738708496, "learning_rate": 8.822224413637293e-05, "loss": 0.0445, "step": 20950 }, { "grad_norm": 0.4145902693271637, "learning_rate": 8.820891619680697e-05, "loss": 0.0314, "step": 20960 }, { "grad_norm": 0.355011522769928, "learning_rate": 8.819558172838227e-05, "loss": 0.0367, "step": 20970 }, { "grad_norm": 0.3475629985332489, "learning_rate": 8.818224073337731e-05, "loss": 0.0301, "step": 20980 }, { "grad_norm": 0.3391627073287964, "learning_rate": 8.816889321407169e-05, "loss": 0.0295, "step": 20990 }, { "grad_norm": 0.4244330823421478, "learning_rate": 8.815553917274615e-05, "loss": 0.0418, "step": 21000 }, { "grad_norm": 0.42505714297294617, "learning_rate": 8.81421786116825e-05, "loss": 0.0468, "step": 21010 }, { "grad_norm": 0.43696141242980957, "learning_rate": 8.812881153316373e-05, "loss": 0.0373, "step": 21020 }, { "grad_norm": 0.2640332281589508, "learning_rate": 8.81154379394739e-05, "loss": 0.0416, "step": 21030 }, { "grad_norm": 0.40212497115135193, "learning_rate": 8.810205783289818e-05, "loss": 0.0412, "step": 21040 }, { "grad_norm": 0.4608577489852905, "learning_rate": 8.808867121572286e-05, "loss": 0.0419, "step": 21050 }, { "grad_norm": 0.3933468461036682, "learning_rate": 8.807527809023537e-05, "loss": 0.0428, "step": 21060 }, { "grad_norm": 0.37156549096107483, "learning_rate": 8.80618784587242e-05, "loss": 0.0393, "step": 21070 }, { "grad_norm": 0.33368954062461853, "learning_rate": 8.804847232347902e-05, "loss": 0.0384, "step": 21080 }, { "grad_norm": 0.41188520193099976, "learning_rate": 8.803505968679054e-05, "loss": 0.0486, "step": 21090 }, { "grad_norm": 0.4868846833705902, "learning_rate": 8.802164055095061e-05, "loss": 0.0467, "step": 21100 }, { "grad_norm": 0.7664022445678711, "learning_rate": 8.80082149182522e-05, "loss": 0.037, "step": 21110 }, { "grad_norm": 0.35886451601982117, "learning_rate": 8.79947827909894e-05, "loss": 0.0392, "step": 21120 }, { "grad_norm": 0.31698673963546753, "learning_rate": 8.798134417145738e-05, "loss": 0.047, "step": 21130 }, { "grad_norm": 0.3916842043399811, "learning_rate": 8.796789906195243e-05, "loss": 0.047, "step": 21140 }, { "grad_norm": 0.8077332973480225, "learning_rate": 8.795444746477195e-05, "loss": 0.0416, "step": 21150 }, { "grad_norm": 0.5352985262870789, "learning_rate": 8.794098938221446e-05, "loss": 0.0403, "step": 21160 }, { "grad_norm": 0.6619068384170532, "learning_rate": 8.792752481657957e-05, "loss": 0.0449, "step": 21170 }, { "grad_norm": 0.3570172190666199, "learning_rate": 8.791405377016802e-05, "loss": 0.0311, "step": 21180 }, { "grad_norm": 0.45361757278442383, "learning_rate": 8.790057624528163e-05, "loss": 0.0412, "step": 21190 }, { "grad_norm": 0.34685489535331726, "learning_rate": 8.788709224422333e-05, "loss": 0.0385, "step": 21200 }, { "grad_norm": 0.37359127402305603, "learning_rate": 8.787360176929717e-05, "loss": 0.0385, "step": 21210 }, { "grad_norm": 0.47243115305900574, "learning_rate": 8.786010482280834e-05, "loss": 0.0349, "step": 21220 }, { "grad_norm": 0.3260488212108612, "learning_rate": 8.784660140706306e-05, "loss": 0.0424, "step": 21230 }, { "grad_norm": 0.38993915915489197, "learning_rate": 8.783309152436872e-05, "loss": 0.0296, "step": 21240 }, { "grad_norm": 0.3425178527832031, "learning_rate": 8.781957517703375e-05, "loss": 0.0424, "step": 21250 }, { "grad_norm": 0.5905033349990845, "learning_rate": 8.780605236736776e-05, "loss": 0.0431, "step": 21260 }, { "grad_norm": 0.30596205592155457, "learning_rate": 8.779252309768142e-05, "loss": 0.042, "step": 21270 }, { "grad_norm": 0.38125866651535034, "learning_rate": 8.777898737028652e-05, "loss": 0.0439, "step": 21280 }, { "grad_norm": 0.3887932300567627, "learning_rate": 8.776544518749591e-05, "loss": 0.0356, "step": 21290 }, { "grad_norm": 0.42684152722358704, "learning_rate": 8.775189655162364e-05, "loss": 0.0334, "step": 21300 }, { "grad_norm": 0.5248486995697021, "learning_rate": 8.773834146498474e-05, "loss": 0.0356, "step": 21310 }, { "grad_norm": 0.45558616518974304, "learning_rate": 8.772477992989545e-05, "loss": 0.0366, "step": 21320 }, { "grad_norm": 1.0601632595062256, "learning_rate": 8.771121194867304e-05, "loss": 0.0489, "step": 21330 }, { "grad_norm": 0.31522583961486816, "learning_rate": 8.769763752363589e-05, "loss": 0.0436, "step": 21340 }, { "grad_norm": 0.48862308263778687, "learning_rate": 8.768405665710352e-05, "loss": 0.0428, "step": 21350 }, { "grad_norm": 0.502994954586029, "learning_rate": 8.767046935139655e-05, "loss": 0.0481, "step": 21360 }, { "grad_norm": 0.8951451778411865, "learning_rate": 8.765687560883666e-05, "loss": 0.0446, "step": 21370 }, { "grad_norm": 0.444148987531662, "learning_rate": 8.764327543174664e-05, "loss": 0.0393, "step": 21380 }, { "grad_norm": 0.5112702250480652, "learning_rate": 8.762966882245038e-05, "loss": 0.0382, "step": 21390 }, { "grad_norm": 0.4404652416706085, "learning_rate": 8.761605578327291e-05, "loss": 0.043, "step": 21400 }, { "grad_norm": 0.3072231113910675, "learning_rate": 8.76024363165403e-05, "loss": 0.0481, "step": 21410 }, { "grad_norm": 0.41475990414619446, "learning_rate": 8.758881042457976e-05, "loss": 0.0408, "step": 21420 }, { "grad_norm": 0.4199408292770386, "learning_rate": 8.757517810971957e-05, "loss": 0.0349, "step": 21430 }, { "grad_norm": 0.5052826404571533, "learning_rate": 8.756153937428913e-05, "loss": 0.0462, "step": 21440 }, { "grad_norm": 0.29420265555381775, "learning_rate": 8.754789422061889e-05, "loss": 0.0431, "step": 21450 }, { "grad_norm": 0.30460822582244873, "learning_rate": 8.753424265104052e-05, "loss": 0.0398, "step": 21460 }, { "grad_norm": 0.4737696647644043, "learning_rate": 8.752058466788659e-05, "loss": 0.0505, "step": 21470 }, { "grad_norm": 0.3179749846458435, "learning_rate": 8.750692027349097e-05, "loss": 0.0352, "step": 21480 }, { "grad_norm": 0.3475671112537384, "learning_rate": 8.749324947018847e-05, "loss": 0.0364, "step": 21490 }, { "grad_norm": 0.38562527298927307, "learning_rate": 8.747957226031507e-05, "loss": 0.0327, "step": 21500 }, { "grad_norm": 0.36202412843704224, "learning_rate": 8.746588864620787e-05, "loss": 0.04, "step": 21510 }, { "grad_norm": 0.35717225074768066, "learning_rate": 8.745219863020498e-05, "loss": 0.031, "step": 21520 }, { "grad_norm": 0.3649734854698181, "learning_rate": 8.743850221464564e-05, "loss": 0.0323, "step": 21530 }, { "grad_norm": 0.2725231647491455, "learning_rate": 8.742479940187026e-05, "loss": 0.0377, "step": 21540 }, { "grad_norm": 0.3502742350101471, "learning_rate": 8.74110901942202e-05, "loss": 0.0357, "step": 21550 }, { "grad_norm": 0.5249586701393127, "learning_rate": 8.739737459403803e-05, "loss": 0.042, "step": 21560 }, { "grad_norm": 0.4105794131755829, "learning_rate": 8.738365260366737e-05, "loss": 0.0409, "step": 21570 }, { "grad_norm": 0.390813410282135, "learning_rate": 8.736992422545292e-05, "loss": 0.0288, "step": 21580 }, { "grad_norm": 0.49595770239830017, "learning_rate": 8.73561894617405e-05, "loss": 0.046, "step": 21590 }, { "grad_norm": 0.30753135681152344, "learning_rate": 8.734244831487697e-05, "loss": 0.0375, "step": 21600 }, { "grad_norm": 0.5015090107917786, "learning_rate": 8.732870078721035e-05, "loss": 0.0458, "step": 21610 }, { "grad_norm": 0.30716007947921753, "learning_rate": 8.731494688108972e-05, "loss": 0.0365, "step": 21620 }, { "grad_norm": 0.555317759513855, "learning_rate": 8.730118659886523e-05, "loss": 0.0379, "step": 21630 }, { "grad_norm": 0.43101373314857483, "learning_rate": 8.728741994288814e-05, "loss": 0.036, "step": 21640 }, { "grad_norm": 0.4276820719242096, "learning_rate": 8.727364691551079e-05, "loss": 0.0345, "step": 21650 }, { "grad_norm": 0.36803966760635376, "learning_rate": 8.725986751908661e-05, "loss": 0.0461, "step": 21660 }, { "grad_norm": 0.34461185336112976, "learning_rate": 8.724608175597016e-05, "loss": 0.0367, "step": 21670 }, { "grad_norm": 0.3678639531135559, "learning_rate": 8.723228962851699e-05, "loss": 0.0352, "step": 21680 }, { "grad_norm": 0.39963582158088684, "learning_rate": 8.721849113908385e-05, "loss": 0.0369, "step": 21690 }, { "grad_norm": 0.4858255088329315, "learning_rate": 8.720468629002848e-05, "loss": 0.039, "step": 21700 }, { "grad_norm": 0.41231781244277954, "learning_rate": 8.719087508370978e-05, "loss": 0.0382, "step": 21710 }, { "grad_norm": 0.45068588852882385, "learning_rate": 8.717705752248772e-05, "loss": 0.0498, "step": 21720 }, { "grad_norm": 0.5598697662353516, "learning_rate": 8.71632336087233e-05, "loss": 0.0308, "step": 21730 }, { "grad_norm": 0.43516239523887634, "learning_rate": 8.71494033447787e-05, "loss": 0.0364, "step": 21740 }, { "grad_norm": 0.3351822793483734, "learning_rate": 8.713556673301708e-05, "loss": 0.0356, "step": 21750 }, { "grad_norm": 0.3520665466785431, "learning_rate": 8.712172377580278e-05, "loss": 0.0492, "step": 21760 }, { "grad_norm": 0.34074512124061584, "learning_rate": 8.710787447550114e-05, "loss": 0.0389, "step": 21770 }, { "grad_norm": 0.5308542847633362, "learning_rate": 8.70940188344787e-05, "loss": 0.0597, "step": 21780 }, { "grad_norm": 0.4221341907978058, "learning_rate": 8.708015685510293e-05, "loss": 0.0391, "step": 21790 }, { "grad_norm": 0.36209923028945923, "learning_rate": 8.706628853974252e-05, "loss": 0.0441, "step": 21800 }, { "grad_norm": 0.36836114525794983, "learning_rate": 8.705241389076715e-05, "loss": 0.0318, "step": 21810 }, { "grad_norm": 0.28803467750549316, "learning_rate": 8.703853291054764e-05, "loss": 0.0315, "step": 21820 }, { "grad_norm": 0.40787604451179504, "learning_rate": 8.702464560145587e-05, "loss": 0.0403, "step": 21830 }, { "grad_norm": 0.41958728432655334, "learning_rate": 8.701075196586476e-05, "loss": 0.0363, "step": 21840 }, { "grad_norm": 0.32146552205085754, "learning_rate": 8.699685200614842e-05, "loss": 0.034, "step": 21850 }, { "grad_norm": 0.30704838037490845, "learning_rate": 8.698294572468193e-05, "loss": 0.0486, "step": 21860 }, { "grad_norm": 0.38194721937179565, "learning_rate": 8.696903312384148e-05, "loss": 0.0436, "step": 21870 }, { "grad_norm": 0.4125576615333557, "learning_rate": 8.695511420600439e-05, "loss": 0.0315, "step": 21880 }, { "grad_norm": 0.3940443694591522, "learning_rate": 8.694118897354901e-05, "loss": 0.0352, "step": 21890 }, { "grad_norm": 0.5603779554367065, "learning_rate": 8.692725742885478e-05, "loss": 0.039, "step": 21900 }, { "grad_norm": 0.620466411113739, "learning_rate": 8.691331957430221e-05, "loss": 0.0394, "step": 21910 }, { "grad_norm": 0.30626970529556274, "learning_rate": 8.68993754122729e-05, "loss": 0.0375, "step": 21920 }, { "grad_norm": 0.43976229429244995, "learning_rate": 8.688542494514955e-05, "loss": 0.0345, "step": 21930 }, { "grad_norm": 0.4880855083465576, "learning_rate": 8.68714681753159e-05, "loss": 0.0432, "step": 21940 }, { "grad_norm": 0.344146192073822, "learning_rate": 8.685750510515676e-05, "loss": 0.0293, "step": 21950 }, { "grad_norm": 0.44101381301879883, "learning_rate": 8.684353573705805e-05, "loss": 0.0377, "step": 21960 }, { "grad_norm": 0.4071092903614044, "learning_rate": 8.682956007340677e-05, "loss": 0.029, "step": 21970 }, { "grad_norm": 0.3862667977809906, "learning_rate": 8.681557811659095e-05, "loss": 0.0342, "step": 21980 }, { "grad_norm": 0.2840996980667114, "learning_rate": 8.680158986899974e-05, "loss": 0.0289, "step": 21990 }, { "grad_norm": 0.40971285104751587, "learning_rate": 8.678759533302335e-05, "loss": 0.0347, "step": 22000 }, { "grad_norm": 0.3512570858001709, "learning_rate": 8.677359451105308e-05, "loss": 0.0311, "step": 22010 }, { "grad_norm": 0.4688318073749542, "learning_rate": 8.675958740548123e-05, "loss": 0.0325, "step": 22020 }, { "grad_norm": 0.4301697909832001, "learning_rate": 8.674557401870129e-05, "loss": 0.0368, "step": 22030 }, { "grad_norm": 0.5544517040252686, "learning_rate": 8.673155435310775e-05, "loss": 0.0407, "step": 22040 }, { "grad_norm": 0.35894858837127686, "learning_rate": 8.671752841109617e-05, "loss": 0.0285, "step": 22050 }, { "grad_norm": 0.36858460307121277, "learning_rate": 8.670349619506321e-05, "loss": 0.0352, "step": 22060 }, { "grad_norm": 0.5211271047592163, "learning_rate": 8.66894577074066e-05, "loss": 0.0386, "step": 22070 }, { "grad_norm": 0.2983255088329315, "learning_rate": 8.667541295052513e-05, "loss": 0.0323, "step": 22080 }, { "grad_norm": 0.36256682872772217, "learning_rate": 8.666136192681865e-05, "loss": 0.0384, "step": 22090 }, { "grad_norm": 0.3075794279575348, "learning_rate": 8.664730463868811e-05, "loss": 0.034, "step": 22100 }, { "grad_norm": 0.4553016424179077, "learning_rate": 8.663324108853552e-05, "loss": 0.0424, "step": 22110 }, { "grad_norm": 0.44222891330718994, "learning_rate": 8.661917127876395e-05, "loss": 0.0567, "step": 22120 }, { "grad_norm": 0.3910183310508728, "learning_rate": 8.660509521177754e-05, "loss": 0.0459, "step": 22130 }, { "grad_norm": 0.4786834120750427, "learning_rate": 8.65910128899815e-05, "loss": 0.036, "step": 22140 }, { "grad_norm": 0.43660956621170044, "learning_rate": 8.657692431578214e-05, "loss": 0.0383, "step": 22150 }, { "grad_norm": 0.504960834980011, "learning_rate": 8.656282949158679e-05, "loss": 0.0358, "step": 22160 }, { "grad_norm": 0.31664329767227173, "learning_rate": 8.654872841980388e-05, "loss": 0.0263, "step": 22170 }, { "grad_norm": 0.6598951816558838, "learning_rate": 8.653462110284289e-05, "loss": 0.0297, "step": 22180 }, { "grad_norm": 0.3395708501338959, "learning_rate": 8.652050754311437e-05, "loss": 0.0373, "step": 22190 }, { "grad_norm": 0.30816635489463806, "learning_rate": 8.650638774302995e-05, "loss": 0.0356, "step": 22200 }, { "grad_norm": 0.43280401825904846, "learning_rate": 8.649226170500233e-05, "loss": 0.04, "step": 22210 }, { "grad_norm": 0.34662866592407227, "learning_rate": 8.647812943144524e-05, "loss": 0.0376, "step": 22220 }, { "grad_norm": 0.3457833528518677, "learning_rate": 8.646399092477351e-05, "loss": 0.0326, "step": 22230 }, { "grad_norm": 0.422377347946167, "learning_rate": 8.644984618740301e-05, "loss": 0.0475, "step": 22240 }, { "grad_norm": 0.3644537925720215, "learning_rate": 8.643569522175073e-05, "loss": 0.0284, "step": 22250 }, { "grad_norm": 0.34595730900764465, "learning_rate": 8.642153803023463e-05, "loss": 0.0366, "step": 22260 }, { "grad_norm": 0.4220948815345764, "learning_rate": 8.640737461527383e-05, "loss": 0.0333, "step": 22270 }, { "grad_norm": 0.3312874436378479, "learning_rate": 8.639320497928845e-05, "loss": 0.0379, "step": 22280 }, { "grad_norm": 0.36400291323661804, "learning_rate": 8.637902912469969e-05, "loss": 0.03, "step": 22290 }, { "grad_norm": 0.3594789505004883, "learning_rate": 8.636484705392982e-05, "loss": 0.0298, "step": 22300 }, { "grad_norm": 0.4181278645992279, "learning_rate": 8.635065876940216e-05, "loss": 0.0305, "step": 22310 }, { "grad_norm": 0.3963790237903595, "learning_rate": 8.633646427354112e-05, "loss": 0.0372, "step": 22320 }, { "grad_norm": 0.2973575294017792, "learning_rate": 8.632226356877213e-05, "loss": 0.0302, "step": 22330 }, { "grad_norm": 0.25661611557006836, "learning_rate": 8.630805665752173e-05, "loss": 0.0404, "step": 22340 }, { "grad_norm": 0.3332061767578125, "learning_rate": 8.629384354221748e-05, "loss": 0.0406, "step": 22350 }, { "grad_norm": 0.3692713677883148, "learning_rate": 8.627962422528797e-05, "loss": 0.033, "step": 22360 }, { "grad_norm": 0.3429107666015625, "learning_rate": 8.626539870916296e-05, "loss": 0.0503, "step": 22370 }, { "grad_norm": 0.29212793707847595, "learning_rate": 8.625116699627317e-05, "loss": 0.0327, "step": 22380 }, { "grad_norm": 0.3581005930900574, "learning_rate": 8.623692908905041e-05, "loss": 0.0333, "step": 22390 }, { "grad_norm": 0.42424502968788147, "learning_rate": 8.622268498992755e-05, "loss": 0.0452, "step": 22400 }, { "grad_norm": 0.4798191785812378, "learning_rate": 8.620843470133851e-05, "loss": 0.0552, "step": 22410 }, { "grad_norm": 0.42990002036094666, "learning_rate": 8.619417822571829e-05, "loss": 0.0394, "step": 22420 }, { "grad_norm": 0.3932005763053894, "learning_rate": 8.617991556550292e-05, "loss": 0.0345, "step": 22430 }, { "grad_norm": 0.4993608593940735, "learning_rate": 8.616564672312952e-05, "loss": 0.038, "step": 22440 }, { "grad_norm": 0.405141681432724, "learning_rate": 8.61513717010362e-05, "loss": 0.0414, "step": 22450 }, { "grad_norm": 0.44851580262184143, "learning_rate": 8.613709050166221e-05, "loss": 0.0435, "step": 22460 }, { "grad_norm": 0.3206596374511719, "learning_rate": 8.61228031274478e-05, "loss": 0.0384, "step": 22470 }, { "grad_norm": 0.39574113488197327, "learning_rate": 8.610850958083431e-05, "loss": 0.0372, "step": 22480 }, { "grad_norm": 0.44375792145729065, "learning_rate": 8.609420986426409e-05, "loss": 0.0324, "step": 22490 }, { "grad_norm": 0.3994612693786621, "learning_rate": 8.60799039801806e-05, "loss": 0.0334, "step": 22500 }, { "grad_norm": 0.47957751154899597, "learning_rate": 8.606559193102828e-05, "loss": 0.0354, "step": 22510 }, { "grad_norm": 0.3897410035133362, "learning_rate": 8.605127371925273e-05, "loss": 0.0435, "step": 22520 }, { "grad_norm": 0.44741153717041016, "learning_rate": 8.603694934730047e-05, "loss": 0.0435, "step": 22530 }, { "grad_norm": 0.24103277921676636, "learning_rate": 8.602261881761919e-05, "loss": 0.026, "step": 22540 }, { "grad_norm": 0.3290890157222748, "learning_rate": 8.600828213265759e-05, "loss": 0.0496, "step": 22550 }, { "grad_norm": 0.38738828897476196, "learning_rate": 8.599393929486539e-05, "loss": 0.045, "step": 22560 }, { "grad_norm": 0.4002281427383423, "learning_rate": 8.59795903066934e-05, "loss": 0.0368, "step": 22570 }, { "grad_norm": 0.2925921678543091, "learning_rate": 8.596523517059347e-05, "loss": 0.0348, "step": 22580 }, { "grad_norm": 0.34746265411376953, "learning_rate": 8.59508738890185e-05, "loss": 0.0287, "step": 22590 }, { "grad_norm": 0.45788511633872986, "learning_rate": 8.593650646442246e-05, "loss": 0.0399, "step": 22600 }, { "grad_norm": 0.3489592671394348, "learning_rate": 8.59221328992603e-05, "loss": 0.0343, "step": 22610 }, { "grad_norm": 0.349319189786911, "learning_rate": 8.590775319598813e-05, "loss": 0.0423, "step": 22620 }, { "grad_norm": 0.4123348295688629, "learning_rate": 8.589336735706301e-05, "loss": 0.0444, "step": 22630 }, { "grad_norm": 0.3749636709690094, "learning_rate": 8.587897538494307e-05, "loss": 0.0535, "step": 22640 }, { "grad_norm": 0.4211873710155487, "learning_rate": 8.586457728208756e-05, "loss": 0.0425, "step": 22650 }, { "grad_norm": 0.481619268655777, "learning_rate": 8.585017305095667e-05, "loss": 0.0404, "step": 22660 }, { "grad_norm": 0.48666951060295105, "learning_rate": 8.583576269401173e-05, "loss": 0.0413, "step": 22670 }, { "grad_norm": 0.6656152009963989, "learning_rate": 8.582134621371504e-05, "loss": 0.0438, "step": 22680 }, { "grad_norm": 0.20995843410491943, "learning_rate": 8.580692361253e-05, "loss": 0.0304, "step": 22690 }, { "grad_norm": 0.4468720853328705, "learning_rate": 8.579249489292104e-05, "loss": 0.0455, "step": 22700 }, { "grad_norm": 0.36606869101524353, "learning_rate": 8.577806005735363e-05, "loss": 0.0448, "step": 22710 }, { "grad_norm": 0.3407146632671356, "learning_rate": 8.576361910829429e-05, "loss": 0.0416, "step": 22720 }, { "grad_norm": 0.31894034147262573, "learning_rate": 8.574917204821057e-05, "loss": 0.0391, "step": 22730 }, { "grad_norm": 0.3297191560268402, "learning_rate": 8.57347188795711e-05, "loss": 0.038, "step": 22740 }, { "grad_norm": 0.461936891078949, "learning_rate": 8.572025960484551e-05, "loss": 0.0424, "step": 22750 }, { "grad_norm": 0.5257735252380371, "learning_rate": 8.57057942265045e-05, "loss": 0.0477, "step": 22760 }, { "grad_norm": 0.4679870307445526, "learning_rate": 8.569132274701984e-05, "loss": 0.0404, "step": 22770 }, { "grad_norm": 0.32236602902412415, "learning_rate": 8.567684516886427e-05, "loss": 0.0341, "step": 22780 }, { "grad_norm": 0.28994113206863403, "learning_rate": 8.56623614945116e-05, "loss": 0.0292, "step": 22790 }, { "grad_norm": 0.38553813099861145, "learning_rate": 8.564787172643675e-05, "loss": 0.0366, "step": 22800 }, { "grad_norm": 0.4609874188899994, "learning_rate": 8.563337586711559e-05, "loss": 0.0315, "step": 22810 }, { "grad_norm": 0.33359313011169434, "learning_rate": 8.561887391902506e-05, "loss": 0.0458, "step": 22820 }, { "grad_norm": 0.27590590715408325, "learning_rate": 8.560436588464316e-05, "loss": 0.0307, "step": 22830 }, { "grad_norm": 0.32885685563087463, "learning_rate": 8.55898517664489e-05, "loss": 0.0453, "step": 22840 }, { "grad_norm": 0.4664807617664337, "learning_rate": 8.557533156692236e-05, "loss": 0.0547, "step": 22850 }, { "grad_norm": 0.35455992817878723, "learning_rate": 8.556080528854467e-05, "loss": 0.0344, "step": 22860 }, { "grad_norm": 0.4692040681838989, "learning_rate": 8.554627293379791e-05, "loss": 0.0341, "step": 22870 }, { "grad_norm": 0.463808536529541, "learning_rate": 8.553173450516531e-05, "loss": 0.0424, "step": 22880 }, { "grad_norm": 0.33669307827949524, "learning_rate": 8.551719000513108e-05, "loss": 0.0381, "step": 22890 }, { "grad_norm": 0.43793144822120667, "learning_rate": 8.550263943618049e-05, "loss": 0.0556, "step": 22900 }, { "grad_norm": 0.3263702392578125, "learning_rate": 8.54880828007998e-05, "loss": 0.037, "step": 22910 }, { "grad_norm": 0.38281822204589844, "learning_rate": 8.547352010147637e-05, "loss": 0.047, "step": 22920 }, { "grad_norm": 0.40830469131469727, "learning_rate": 8.545895134069855e-05, "loss": 0.0317, "step": 22930 }, { "grad_norm": 0.40228062868118286, "learning_rate": 8.544437652095576e-05, "loss": 0.0358, "step": 22940 }, { "grad_norm": 0.4756705164909363, "learning_rate": 8.542979564473843e-05, "loss": 0.0516, "step": 22950 }, { "grad_norm": 0.34940406680107117, "learning_rate": 8.541520871453802e-05, "loss": 0.0316, "step": 22960 }, { "grad_norm": 0.38229385018348694, "learning_rate": 8.540061573284705e-05, "loss": 0.0369, "step": 22970 }, { "grad_norm": 0.3445529639720917, "learning_rate": 8.538601670215906e-05, "loss": 0.0334, "step": 22980 }, { "grad_norm": 0.4146343767642975, "learning_rate": 8.537141162496864e-05, "loss": 0.0456, "step": 22990 }, { "grad_norm": 0.4261532425880432, "learning_rate": 8.535680050377137e-05, "loss": 0.033, "step": 23000 }, { "grad_norm": 0.25500497221946716, "learning_rate": 8.534218334106391e-05, "loss": 0.0324, "step": 23010 }, { "grad_norm": 0.36997899413108826, "learning_rate": 8.532756013934393e-05, "loss": 0.032, "step": 23020 }, { "grad_norm": 0.3238561749458313, "learning_rate": 8.531293090111012e-05, "loss": 0.0396, "step": 23030 }, { "grad_norm": 0.4020002484321594, "learning_rate": 8.529829562886225e-05, "loss": 0.0393, "step": 23040 }, { "grad_norm": 0.46812379360198975, "learning_rate": 8.528365432510105e-05, "loss": 0.0527, "step": 23050 }, { "grad_norm": 0.3279491364955902, "learning_rate": 8.526900699232833e-05, "loss": 0.0421, "step": 23060 }, { "grad_norm": 0.4937368631362915, "learning_rate": 8.525435363304695e-05, "loss": 0.0387, "step": 23070 }, { "grad_norm": 0.4146554172039032, "learning_rate": 8.523969424976072e-05, "loss": 0.0373, "step": 23080 }, { "grad_norm": 0.337894469499588, "learning_rate": 8.522502884497457e-05, "loss": 0.0365, "step": 23090 }, { "grad_norm": 0.45866674184799194, "learning_rate": 8.521035742119437e-05, "loss": 0.0425, "step": 23100 }, { "grad_norm": 0.3770507872104645, "learning_rate": 8.519567998092712e-05, "loss": 0.0421, "step": 23110 }, { "grad_norm": 0.439993679523468, "learning_rate": 8.518099652668075e-05, "loss": 0.039, "step": 23120 }, { "grad_norm": 0.32559511065483093, "learning_rate": 8.516630706096429e-05, "loss": 0.0412, "step": 23130 }, { "grad_norm": 0.45694106817245483, "learning_rate": 8.515161158628773e-05, "loss": 0.0324, "step": 23140 }, { "grad_norm": 0.406690388917923, "learning_rate": 8.513691010516216e-05, "loss": 0.0516, "step": 23150 }, { "grad_norm": 0.3240995407104492, "learning_rate": 8.512220262009966e-05, "loss": 0.0314, "step": 23160 }, { "grad_norm": 0.3950210511684418, "learning_rate": 8.510748913361332e-05, "loss": 0.0453, "step": 23170 }, { "grad_norm": 0.34007933735847473, "learning_rate": 8.509276964821726e-05, "loss": 0.0333, "step": 23180 }, { "grad_norm": 0.3341401219367981, "learning_rate": 8.507804416642669e-05, "loss": 0.0402, "step": 23190 }, { "grad_norm": 0.4742424190044403, "learning_rate": 8.506331269075774e-05, "loss": 0.0307, "step": 23200 }, { "grad_norm": 0.33046239614486694, "learning_rate": 8.504857522372765e-05, "loss": 0.0324, "step": 23210 }, { "grad_norm": 0.3544442653656006, "learning_rate": 8.503383176785461e-05, "loss": 0.0391, "step": 23220 }, { "grad_norm": 0.43877843022346497, "learning_rate": 8.501908232565792e-05, "loss": 0.0371, "step": 23230 }, { "grad_norm": 0.37781888246536255, "learning_rate": 8.50043268996578e-05, "loss": 0.0392, "step": 23240 }, { "grad_norm": 0.36421719193458557, "learning_rate": 8.498956549237562e-05, "loss": 0.0366, "step": 23250 }, { "grad_norm": 0.30582529306411743, "learning_rate": 8.497479810633366e-05, "loss": 0.0293, "step": 23260 }, { "grad_norm": 0.43990257382392883, "learning_rate": 8.496002474405525e-05, "loss": 0.0424, "step": 23270 }, { "grad_norm": 0.3013226389884949, "learning_rate": 8.494524540806478e-05, "loss": 0.0461, "step": 23280 }, { "grad_norm": 0.35618966817855835, "learning_rate": 8.493046010088761e-05, "loss": 0.0528, "step": 23290 }, { "grad_norm": 0.563441276550293, "learning_rate": 8.491566882505018e-05, "loss": 0.0388, "step": 23300 }, { "grad_norm": 0.38112327456474304, "learning_rate": 8.490087158307988e-05, "loss": 0.0498, "step": 23310 }, { "grad_norm": 0.5792088508605957, "learning_rate": 8.488606837750518e-05, "loss": 0.0435, "step": 23320 }, { "grad_norm": 0.3694356381893158, "learning_rate": 8.487125921085552e-05, "loss": 0.0332, "step": 23330 }, { "grad_norm": 0.3631080090999603, "learning_rate": 8.485644408566141e-05, "loss": 0.0388, "step": 23340 }, { "grad_norm": 0.3071943521499634, "learning_rate": 8.484162300445431e-05, "loss": 0.0362, "step": 23350 }, { "grad_norm": 0.3412810266017914, "learning_rate": 8.482679596976676e-05, "loss": 0.0307, "step": 23360 }, { "grad_norm": 0.5032842755317688, "learning_rate": 8.48119629841323e-05, "loss": 0.0361, "step": 23370 }, { "grad_norm": 0.4239846467971802, "learning_rate": 8.479712405008547e-05, "loss": 0.0399, "step": 23380 }, { "grad_norm": 0.28274697065353394, "learning_rate": 8.478227917016184e-05, "loss": 0.0447, "step": 23390 }, { "grad_norm": 0.2793786823749542, "learning_rate": 8.476742834689801e-05, "loss": 0.0407, "step": 23400 }, { "grad_norm": 0.4276030659675598, "learning_rate": 8.475257158283157e-05, "loss": 0.0371, "step": 23410 }, { "grad_norm": 0.5039665699005127, "learning_rate": 8.473770888050112e-05, "loss": 0.0322, "step": 23420 }, { "grad_norm": 0.32160642743110657, "learning_rate": 8.47228402424463e-05, "loss": 0.0319, "step": 23430 }, { "grad_norm": 1.0617812871932983, "learning_rate": 8.470796567120775e-05, "loss": 0.049, "step": 23440 }, { "grad_norm": 0.3752743899822235, "learning_rate": 8.469308516932714e-05, "loss": 0.0383, "step": 23450 }, { "grad_norm": 0.3855254054069519, "learning_rate": 8.467819873934714e-05, "loss": 0.0363, "step": 23460 }, { "grad_norm": 0.31121253967285156, "learning_rate": 8.466330638381143e-05, "loss": 0.0393, "step": 23470 }, { "grad_norm": 0.43557363748550415, "learning_rate": 8.464840810526469e-05, "loss": 0.0367, "step": 23480 }, { "grad_norm": 0.49267280101776123, "learning_rate": 8.463350390625264e-05, "loss": 0.0319, "step": 23490 }, { "grad_norm": 0.2835557758808136, "learning_rate": 8.4618593789322e-05, "loss": 0.0442, "step": 23500 }, { "grad_norm": 0.33449962735176086, "learning_rate": 8.46036777570205e-05, "loss": 0.0389, "step": 23510 }, { "grad_norm": 0.5306292176246643, "learning_rate": 8.458875581189688e-05, "loss": 0.0638, "step": 23520 }, { "grad_norm": 0.31140124797821045, "learning_rate": 8.457382795650092e-05, "loss": 0.0321, "step": 23530 }, { "grad_norm": 0.31555095314979553, "learning_rate": 8.455889419338335e-05, "loss": 0.0419, "step": 23540 }, { "grad_norm": 0.4468403160572052, "learning_rate": 8.454395452509593e-05, "loss": 0.0351, "step": 23550 }, { "grad_norm": 0.3546769320964813, "learning_rate": 8.452900895419146e-05, "loss": 0.0354, "step": 23560 }, { "grad_norm": 0.25888413190841675, "learning_rate": 8.451405748322376e-05, "loss": 0.0314, "step": 23570 }, { "grad_norm": 0.34036338329315186, "learning_rate": 8.449910011474759e-05, "loss": 0.0296, "step": 23580 }, { "grad_norm": 0.44486793875694275, "learning_rate": 8.448413685131876e-05, "loss": 0.0464, "step": 23590 }, { "grad_norm": 0.35394492745399475, "learning_rate": 8.446916769549407e-05, "loss": 0.0415, "step": 23600 }, { "grad_norm": 0.3511267602443695, "learning_rate": 8.445419264983136e-05, "loss": 0.043, "step": 23610 }, { "grad_norm": 0.37849465012550354, "learning_rate": 8.443921171688947e-05, "loss": 0.0378, "step": 23620 }, { "grad_norm": 0.31241485476493835, "learning_rate": 8.442422489922819e-05, "loss": 0.0352, "step": 23630 }, { "grad_norm": 0.34089386463165283, "learning_rate": 8.440923219940838e-05, "loss": 0.0411, "step": 23640 }, { "grad_norm": 0.4840799868106842, "learning_rate": 8.439423361999189e-05, "loss": 0.0435, "step": 23650 }, { "grad_norm": 0.350101500749588, "learning_rate": 8.437922916354155e-05, "loss": 0.0482, "step": 23660 }, { "grad_norm": 0.38815221190452576, "learning_rate": 8.436421883262123e-05, "loss": 0.0362, "step": 23670 }, { "grad_norm": 0.306321918964386, "learning_rate": 8.434920262979577e-05, "loss": 0.0538, "step": 23680 }, { "grad_norm": 0.5546888113021851, "learning_rate": 8.433418055763104e-05, "loss": 0.0302, "step": 23690 }, { "grad_norm": 0.3202863931655884, "learning_rate": 8.431915261869389e-05, "loss": 0.0409, "step": 23700 }, { "grad_norm": 0.544313371181488, "learning_rate": 8.43041188155522e-05, "loss": 0.0446, "step": 23710 }, { "grad_norm": 0.489568829536438, "learning_rate": 8.428907915077481e-05, "loss": 0.0382, "step": 23720 }, { "grad_norm": 0.4808547794818878, "learning_rate": 8.42740336269316e-05, "loss": 0.0452, "step": 23730 }, { "grad_norm": 0.26371240615844727, "learning_rate": 8.425898224659345e-05, "loss": 0.033, "step": 23740 }, { "grad_norm": 0.6125783920288086, "learning_rate": 8.42439250123322e-05, "loss": 0.0343, "step": 23750 }, { "grad_norm": 0.4273366630077362, "learning_rate": 8.422886192672076e-05, "loss": 0.0322, "step": 23760 }, { "grad_norm": 0.4233294129371643, "learning_rate": 8.421379299233297e-05, "loss": 0.0333, "step": 23770 }, { "grad_norm": 0.3543696403503418, "learning_rate": 8.419871821174371e-05, "loss": 0.0342, "step": 23780 }, { "grad_norm": 0.6241199970245361, "learning_rate": 8.418363758752884e-05, "loss": 0.0309, "step": 23790 }, { "grad_norm": 0.44880974292755127, "learning_rate": 8.416855112226523e-05, "loss": 0.0401, "step": 23800 }, { "grad_norm": 0.3038944900035858, "learning_rate": 8.415345881853075e-05, "loss": 0.0341, "step": 23810 }, { "grad_norm": 0.4178375005722046, "learning_rate": 8.413836067890426e-05, "loss": 0.0324, "step": 23820 }, { "grad_norm": 0.24533307552337646, "learning_rate": 8.41232567059656e-05, "loss": 0.0291, "step": 23830 }, { "grad_norm": 0.3886314928531647, "learning_rate": 8.410814690229565e-05, "loss": 0.0348, "step": 23840 }, { "grad_norm": 0.2360018789768219, "learning_rate": 8.409303127047626e-05, "loss": 0.0337, "step": 23850 }, { "grad_norm": 0.5401549339294434, "learning_rate": 8.407790981309028e-05, "loss": 0.0352, "step": 23860 }, { "grad_norm": 0.409821480512619, "learning_rate": 8.406278253272153e-05, "loss": 0.0375, "step": 23870 }, { "grad_norm": 0.46352437138557434, "learning_rate": 8.404764943195487e-05, "loss": 0.0311, "step": 23880 }, { "grad_norm": 0.3434855043888092, "learning_rate": 8.403251051337613e-05, "loss": 0.0437, "step": 23890 }, { "grad_norm": 0.30998164415359497, "learning_rate": 8.401736577957214e-05, "loss": 0.0331, "step": 23900 }, { "grad_norm": 0.21204999089241028, "learning_rate": 8.40022152331307e-05, "loss": 0.0371, "step": 23910 }, { "grad_norm": 0.38089150190353394, "learning_rate": 8.398705887664064e-05, "loss": 0.0455, "step": 23920 }, { "grad_norm": 0.28255555033683777, "learning_rate": 8.397189671269177e-05, "loss": 0.0347, "step": 23930 }, { "grad_norm": 0.2994585633277893, "learning_rate": 8.395672874387488e-05, "loss": 0.0252, "step": 23940 }, { "grad_norm": 0.2539703845977783, "learning_rate": 8.394155497278177e-05, "loss": 0.0344, "step": 23950 }, { "grad_norm": 0.4604555368423462, "learning_rate": 8.392637540200523e-05, "loss": 0.0481, "step": 23960 }, { "grad_norm": 0.3727436065673828, "learning_rate": 8.391119003413902e-05, "loss": 0.0371, "step": 23970 }, { "grad_norm": 0.42135098576545715, "learning_rate": 8.38959988717779e-05, "loss": 0.0367, "step": 23980 }, { "grad_norm": 0.4256518483161926, "learning_rate": 8.388080191751764e-05, "loss": 0.0388, "step": 23990 }, { "grad_norm": 0.36673250794410706, "learning_rate": 8.386559917395496e-05, "loss": 0.0377, "step": 24000 }, { "grad_norm": 0.4357382357120514, "learning_rate": 8.385039064368761e-05, "loss": 0.0274, "step": 24010 }, { "grad_norm": 0.4169408082962036, "learning_rate": 8.383517632931431e-05, "loss": 0.0262, "step": 24020 }, { "grad_norm": 0.3266875743865967, "learning_rate": 8.381995623343477e-05, "loss": 0.0358, "step": 24030 }, { "grad_norm": 0.49487367272377014, "learning_rate": 8.380473035864968e-05, "loss": 0.0379, "step": 24040 }, { "grad_norm": 0.29716360569000244, "learning_rate": 8.378949870756076e-05, "loss": 0.0324, "step": 24050 }, { "grad_norm": 0.5328670144081116, "learning_rate": 8.377426128277063e-05, "loss": 0.0373, "step": 24060 }, { "grad_norm": 0.4090198874473572, "learning_rate": 8.375901808688298e-05, "loss": 0.0281, "step": 24070 }, { "grad_norm": 0.38094618916511536, "learning_rate": 8.374376912250246e-05, "loss": 0.0401, "step": 24080 }, { "grad_norm": 0.41548794507980347, "learning_rate": 8.372851439223468e-05, "loss": 0.0346, "step": 24090 }, { "grad_norm": 0.31283462047576904, "learning_rate": 8.371325389868627e-05, "loss": 0.0383, "step": 24100 }, { "grad_norm": 0.43483030796051025, "learning_rate": 8.369798764446482e-05, "loss": 0.0316, "step": 24110 }, { "grad_norm": 0.3807775676250458, "learning_rate": 8.368271563217893e-05, "loss": 0.0334, "step": 24120 }, { "grad_norm": 0.4137595593929291, "learning_rate": 8.366743786443817e-05, "loss": 0.0392, "step": 24130 }, { "grad_norm": 0.3806188404560089, "learning_rate": 8.365215434385309e-05, "loss": 0.0444, "step": 24140 }, { "grad_norm": 0.4384664297103882, "learning_rate": 8.36368650730352e-05, "loss": 0.0445, "step": 24150 }, { "grad_norm": 0.320303738117218, "learning_rate": 8.362157005459705e-05, "loss": 0.04, "step": 24160 }, { "grad_norm": 0.4416672885417938, "learning_rate": 8.360626929115213e-05, "loss": 0.0309, "step": 24170 }, { "grad_norm": 0.45055949687957764, "learning_rate": 8.359096278531492e-05, "loss": 0.0325, "step": 24180 }, { "grad_norm": 0.3277461528778076, "learning_rate": 8.357565053970088e-05, "loss": 0.0324, "step": 24190 }, { "grad_norm": 0.3816486895084381, "learning_rate": 8.356033255692647e-05, "loss": 0.0295, "step": 24200 }, { "grad_norm": 0.4892994463443756, "learning_rate": 8.354500883960911e-05, "loss": 0.0333, "step": 24210 }, { "grad_norm": 0.2577022910118103, "learning_rate": 8.352967939036717e-05, "loss": 0.0334, "step": 24220 }, { "grad_norm": 0.3919335901737213, "learning_rate": 8.35143442118201e-05, "loss": 0.0405, "step": 24230 }, { "grad_norm": 0.25035133957862854, "learning_rate": 8.349900330658819e-05, "loss": 0.0292, "step": 24240 }, { "grad_norm": 0.41210120916366577, "learning_rate": 8.348365667729284e-05, "loss": 0.0394, "step": 24250 }, { "grad_norm": 0.41454291343688965, "learning_rate": 8.346830432655633e-05, "loss": 0.0308, "step": 24260 }, { "grad_norm": 0.41332170367240906, "learning_rate": 8.345294625700195e-05, "loss": 0.0294, "step": 24270 }, { "grad_norm": 0.4508712589740753, "learning_rate": 8.343758247125402e-05, "loss": 0.0369, "step": 24280 }, { "grad_norm": 0.36066779494285583, "learning_rate": 8.342221297193776e-05, "loss": 0.0379, "step": 24290 }, { "grad_norm": 0.6636274456977844, "learning_rate": 8.34068377616794e-05, "loss": 0.0447, "step": 24300 }, { "grad_norm": 0.5329453349113464, "learning_rate": 8.339145684310615e-05, "loss": 0.0425, "step": 24310 }, { "grad_norm": 0.44809436798095703, "learning_rate": 8.337607021884618e-05, "loss": 0.0351, "step": 24320 }, { "grad_norm": 0.27634814381599426, "learning_rate": 8.336067789152867e-05, "loss": 0.0274, "step": 24330 }, { "grad_norm": 0.3614826500415802, "learning_rate": 8.334527986378369e-05, "loss": 0.0365, "step": 24340 }, { "grad_norm": 0.32760873436927795, "learning_rate": 8.332987613824239e-05, "loss": 0.0411, "step": 24350 }, { "grad_norm": 0.3316931426525116, "learning_rate": 8.331446671753685e-05, "loss": 0.0365, "step": 24360 }, { "grad_norm": 0.3982713520526886, "learning_rate": 8.329905160430007e-05, "loss": 0.0355, "step": 24370 }, { "grad_norm": 0.3814690411090851, "learning_rate": 8.328363080116611e-05, "loss": 0.0293, "step": 24380 }, { "grad_norm": 0.25680485367774963, "learning_rate": 8.326820431076997e-05, "loss": 0.031, "step": 24390 }, { "grad_norm": 0.38747134804725647, "learning_rate": 8.325277213574759e-05, "loss": 0.0364, "step": 24400 }, { "grad_norm": 0.4468469023704529, "learning_rate": 8.32373342787359e-05, "loss": 0.0409, "step": 24410 }, { "grad_norm": 0.4267910122871399, "learning_rate": 8.322189074237285e-05, "loss": 0.0354, "step": 24420 }, { "grad_norm": 0.34379565715789795, "learning_rate": 8.32064415292973e-05, "loss": 0.0359, "step": 24430 }, { "grad_norm": 0.3605424761772156, "learning_rate": 8.319098664214907e-05, "loss": 0.0371, "step": 24440 }, { "grad_norm": 0.48109233379364014, "learning_rate": 8.3175526083569e-05, "loss": 0.0402, "step": 24450 }, { "grad_norm": 0.3349946141242981, "learning_rate": 8.316005985619889e-05, "loss": 0.0355, "step": 24460 }, { "grad_norm": 0.4109368920326233, "learning_rate": 8.314458796268147e-05, "loss": 0.0357, "step": 24470 }, { "grad_norm": 0.43083158135414124, "learning_rate": 8.312911040566047e-05, "loss": 0.0341, "step": 24480 }, { "grad_norm": 0.3905399739742279, "learning_rate": 8.31136271877806e-05, "loss": 0.0344, "step": 24490 }, { "grad_norm": 0.43285316228866577, "learning_rate": 8.309813831168748e-05, "loss": 0.039, "step": 24500 }, { "grad_norm": 0.4181622564792633, "learning_rate": 8.308264378002777e-05, "loss": 0.0319, "step": 24510 }, { "grad_norm": 0.33514270186424255, "learning_rate": 8.306714359544906e-05, "loss": 0.0395, "step": 24520 }, { "grad_norm": 0.35178789496421814, "learning_rate": 8.30516377605999e-05, "loss": 0.0331, "step": 24530 }, { "grad_norm": 0.4585232734680176, "learning_rate": 8.30361262781298e-05, "loss": 0.0381, "step": 24540 }, { "grad_norm": 0.4458655118942261, "learning_rate": 8.302060915068924e-05, "loss": 0.0413, "step": 24550 }, { "grad_norm": 0.40780892968177795, "learning_rate": 8.300508638092972e-05, "loss": 0.0336, "step": 24560 }, { "grad_norm": 0.5178930759429932, "learning_rate": 8.298955797150361e-05, "loss": 0.0353, "step": 24570 }, { "grad_norm": 0.48118868470191956, "learning_rate": 8.297402392506433e-05, "loss": 0.0354, "step": 24580 }, { "grad_norm": 0.3739701509475708, "learning_rate": 8.295848424426617e-05, "loss": 0.0358, "step": 24590 }, { "grad_norm": 0.24117770791053772, "learning_rate": 8.29429389317645e-05, "loss": 0.0307, "step": 24600 }, { "grad_norm": 0.535323441028595, "learning_rate": 8.292738799021556e-05, "loss": 0.0454, "step": 24610 }, { "grad_norm": 0.43704208731651306, "learning_rate": 8.291183142227656e-05, "loss": 0.0282, "step": 24620 }, { "grad_norm": 0.28921592235565186, "learning_rate": 8.289626923060572e-05, "loss": 0.0419, "step": 24630 }, { "grad_norm": 0.33810168504714966, "learning_rate": 8.288070141786218e-05, "loss": 0.0377, "step": 24640 }, { "grad_norm": 0.26014184951782227, "learning_rate": 8.286512798670605e-05, "loss": 0.0347, "step": 24650 }, { "grad_norm": 0.3618883788585663, "learning_rate": 8.284954893979842e-05, "loss": 0.0407, "step": 24660 }, { "grad_norm": 0.36896830797195435, "learning_rate": 8.283396427980131e-05, "loss": 0.047, "step": 24670 }, { "grad_norm": 0.3469155430793762, "learning_rate": 8.281837400937771e-05, "loss": 0.0392, "step": 24680 }, { "grad_norm": 0.48833197355270386, "learning_rate": 8.28027781311916e-05, "loss": 0.0358, "step": 24690 }, { "grad_norm": 0.6937330365180969, "learning_rate": 8.278717664790785e-05, "loss": 0.0456, "step": 24700 }, { "grad_norm": 0.3208114206790924, "learning_rate": 8.277156956219234e-05, "loss": 0.0352, "step": 24710 }, { "grad_norm": 0.44093453884124756, "learning_rate": 8.275595687671189e-05, "loss": 0.0347, "step": 24720 }, { "grad_norm": 0.48015469312667847, "learning_rate": 8.27403385941343e-05, "loss": 0.0374, "step": 24730 }, { "grad_norm": 0.3557146191596985, "learning_rate": 8.272471471712828e-05, "loss": 0.036, "step": 24740 }, { "grad_norm": 0.40116727352142334, "learning_rate": 8.270908524836355e-05, "loss": 0.0389, "step": 24750 }, { "grad_norm": 0.4295109212398529, "learning_rate": 8.269345019051074e-05, "loss": 0.0373, "step": 24760 }, { "grad_norm": 0.3257793188095093, "learning_rate": 8.267780954624147e-05, "loss": 0.0361, "step": 24770 }, { "grad_norm": 0.449340283870697, "learning_rate": 8.266216331822827e-05, "loss": 0.0401, "step": 24780 }, { "grad_norm": 0.40679383277893066, "learning_rate": 8.264651150914469e-05, "loss": 0.0371, "step": 24790 }, { "grad_norm": 0.3332923948764801, "learning_rate": 8.263085412166517e-05, "loss": 0.035, "step": 24800 }, { "grad_norm": 0.4257059693336487, "learning_rate": 8.261519115846514e-05, "loss": 0.0358, "step": 24810 }, { "grad_norm": 0.3980359733104706, "learning_rate": 8.259952262222096e-05, "loss": 0.0314, "step": 24820 }, { "grad_norm": 0.37245962023735046, "learning_rate": 8.258384851560997e-05, "loss": 0.0359, "step": 24830 }, { "grad_norm": 0.6462658047676086, "learning_rate": 8.256816884131044e-05, "loss": 0.0368, "step": 24840 }, { "grad_norm": 0.4438754916191101, "learning_rate": 8.255248360200159e-05, "loss": 0.0372, "step": 24850 }, { "grad_norm": 0.38161325454711914, "learning_rate": 8.253679280036359e-05, "loss": 0.0307, "step": 24860 }, { "grad_norm": 0.4277664124965668, "learning_rate": 8.252109643907762e-05, "loss": 0.0442, "step": 24870 }, { "grad_norm": 0.40324920415878296, "learning_rate": 8.250539452082569e-05, "loss": 0.0345, "step": 24880 }, { "grad_norm": 0.3277840316295624, "learning_rate": 8.248968704829087e-05, "loss": 0.0341, "step": 24890 }, { "grad_norm": 0.42347341775894165, "learning_rate": 8.247397402415714e-05, "loss": 0.0497, "step": 24900 }, { "grad_norm": 0.3706074655056, "learning_rate": 8.24582554511094e-05, "loss": 0.0372, "step": 24910 }, { "grad_norm": 0.26449358463287354, "learning_rate": 8.244253133183355e-05, "loss": 0.0304, "step": 24920 }, { "grad_norm": 0.5454270839691162, "learning_rate": 8.24268016690164e-05, "loss": 0.032, "step": 24930 }, { "grad_norm": 0.37366822361946106, "learning_rate": 8.241106646534571e-05, "loss": 0.0377, "step": 24940 }, { "grad_norm": 0.5561658143997192, "learning_rate": 8.23953257235102e-05, "loss": 0.0434, "step": 24950 }, { "grad_norm": 0.4975191354751587, "learning_rate": 8.237957944619956e-05, "loss": 0.0388, "step": 24960 }, { "grad_norm": 0.32981395721435547, "learning_rate": 8.236382763610437e-05, "loss": 0.0392, "step": 24970 }, { "grad_norm": 0.4316488802433014, "learning_rate": 8.234807029591619e-05, "loss": 0.0398, "step": 24980 }, { "grad_norm": 0.48743966221809387, "learning_rate": 8.233230742832752e-05, "loss": 0.0491, "step": 24990 }, { "grad_norm": 0.4988745152950287, "learning_rate": 8.231653903603178e-05, "loss": 0.041, "step": 25000 }, { "grad_norm": 0.2675533592700958, "learning_rate": 8.23007651217234e-05, "loss": 0.0375, "step": 25010 }, { "grad_norm": 0.3923912048339844, "learning_rate": 8.228498568809769e-05, "loss": 0.0344, "step": 25020 }, { "grad_norm": 0.4471108913421631, "learning_rate": 8.22692007378509e-05, "loss": 0.041, "step": 25030 }, { "grad_norm": 0.5013864040374756, "learning_rate": 8.225341027368028e-05, "loss": 0.0339, "step": 25040 }, { "grad_norm": 0.33160144090652466, "learning_rate": 8.223761429828399e-05, "loss": 0.0378, "step": 25050 }, { "grad_norm": 0.44101953506469727, "learning_rate": 8.22218128143611e-05, "loss": 0.0506, "step": 25060 }, { "grad_norm": 0.3373933434486389, "learning_rate": 8.220600582461166e-05, "loss": 0.0351, "step": 25070 }, { "grad_norm": 0.24755948781967163, "learning_rate": 8.219019333173668e-05, "loss": 0.0409, "step": 25080 }, { "grad_norm": 0.616894006729126, "learning_rate": 8.217437533843805e-05, "loss": 0.0463, "step": 25090 }, { "grad_norm": 0.25866398215293884, "learning_rate": 8.215855184741867e-05, "loss": 0.0372, "step": 25100 }, { "grad_norm": 0.4529836177825928, "learning_rate": 8.21427228613823e-05, "loss": 0.039, "step": 25110 }, { "grad_norm": 0.4141418933868408, "learning_rate": 8.21268883830337e-05, "loss": 0.0389, "step": 25120 }, { "grad_norm": 0.36457210779190063, "learning_rate": 8.211104841507855e-05, "loss": 0.0463, "step": 25130 }, { "grad_norm": 0.5865569710731506, "learning_rate": 8.209520296022346e-05, "loss": 0.0377, "step": 25140 }, { "grad_norm": 0.4053426682949066, "learning_rate": 8.207935202117599e-05, "loss": 0.0408, "step": 25150 }, { "grad_norm": 0.36327648162841797, "learning_rate": 8.206349560064463e-05, "loss": 0.0375, "step": 25160 }, { "grad_norm": 0.3842640221118927, "learning_rate": 8.204763370133881e-05, "loss": 0.0258, "step": 25170 }, { "grad_norm": 0.6000558137893677, "learning_rate": 8.203176632596892e-05, "loss": 0.0392, "step": 25180 }, { "grad_norm": 0.44097182154655457, "learning_rate": 8.20158934772462e-05, "loss": 0.0402, "step": 25190 }, { "grad_norm": 0.2949269413948059, "learning_rate": 8.200001515788294e-05, "loss": 0.0315, "step": 25200 }, { "grad_norm": 0.3823906183242798, "learning_rate": 8.198413137059228e-05, "loss": 0.0367, "step": 25210 }, { "grad_norm": 0.3267136514186859, "learning_rate": 8.196824211808835e-05, "loss": 0.0327, "step": 25220 }, { "grad_norm": 0.38008543848991394, "learning_rate": 8.195234740308617e-05, "loss": 0.0422, "step": 25230 }, { "grad_norm": 0.34149980545043945, "learning_rate": 8.193644722830171e-05, "loss": 0.0313, "step": 25240 }, { "grad_norm": 0.373505562543869, "learning_rate": 8.19205415964519e-05, "loss": 0.028, "step": 25250 }, { "grad_norm": 0.39326176047325134, "learning_rate": 8.190463051025456e-05, "loss": 0.0285, "step": 25260 }, { "grad_norm": 0.332173228263855, "learning_rate": 8.188871397242843e-05, "loss": 0.0358, "step": 25270 }, { "grad_norm": 0.31159326434135437, "learning_rate": 8.187279198569326e-05, "loss": 0.0329, "step": 25280 }, { "grad_norm": 0.3006640374660492, "learning_rate": 8.185686455276966e-05, "loss": 0.0291, "step": 25290 }, { "grad_norm": 0.5483241081237793, "learning_rate": 8.184093167637921e-05, "loss": 0.0374, "step": 25300 }, { "grad_norm": 0.5387920141220093, "learning_rate": 8.182499335924437e-05, "loss": 0.0446, "step": 25310 }, { "grad_norm": 0.41642287373542786, "learning_rate": 8.18090496040886e-05, "loss": 0.0352, "step": 25320 }, { "grad_norm": 0.34919899702072144, "learning_rate": 8.179310041363621e-05, "loss": 0.047, "step": 25330 }, { "grad_norm": 0.3954295814037323, "learning_rate": 8.17771457906125e-05, "loss": 0.0385, "step": 25340 }, { "grad_norm": 0.45070362091064453, "learning_rate": 8.176118573774371e-05, "loss": 0.0321, "step": 25350 }, { "grad_norm": 0.4298710823059082, "learning_rate": 8.174522025775692e-05, "loss": 0.0385, "step": 25360 }, { "grad_norm": 0.4968530833721161, "learning_rate": 8.172924935338022e-05, "loss": 0.0419, "step": 25370 }, { "grad_norm": 0.5984996557235718, "learning_rate": 8.171327302734262e-05, "loss": 0.0353, "step": 25380 }, { "grad_norm": 0.7332407236099243, "learning_rate": 8.169729128237401e-05, "loss": 0.0343, "step": 25390 }, { "grad_norm": 0.410427987575531, "learning_rate": 8.168130412120525e-05, "loss": 0.038, "step": 25400 }, { "grad_norm": 0.46724289655685425, "learning_rate": 8.16653115465681e-05, "loss": 0.0313, "step": 25410 }, { "grad_norm": 0.6809641122817993, "learning_rate": 8.164931356119526e-05, "loss": 0.0306, "step": 25420 }, { "grad_norm": 0.553731381893158, "learning_rate": 8.163331016782032e-05, "loss": 0.0414, "step": 25430 }, { "grad_norm": 0.33667272329330444, "learning_rate": 8.161730136917785e-05, "loss": 0.0329, "step": 25440 }, { "grad_norm": 0.5819829106330872, "learning_rate": 8.160128716800333e-05, "loss": 0.0412, "step": 25450 }, { "grad_norm": 0.3751387894153595, "learning_rate": 8.158526756703313e-05, "loss": 0.0287, "step": 25460 }, { "grad_norm": 0.4671899676322937, "learning_rate": 8.156924256900455e-05, "loss": 0.0378, "step": 25470 }, { "grad_norm": 0.3939632773399353, "learning_rate": 8.155321217665584e-05, "loss": 0.0331, "step": 25480 }, { "grad_norm": 0.4265965521335602, "learning_rate": 8.153717639272614e-05, "loss": 0.0332, "step": 25490 }, { "grad_norm": 0.33412057161331177, "learning_rate": 8.152113521995555e-05, "loss": 0.0329, "step": 25500 }, { "grad_norm": 0.3649296164512634, "learning_rate": 8.150508866108505e-05, "loss": 0.0315, "step": 25510 }, { "grad_norm": 0.43567293882369995, "learning_rate": 8.148903671885657e-05, "loss": 0.031, "step": 25520 }, { "grad_norm": 0.4026566445827484, "learning_rate": 8.147297939601292e-05, "loss": 0.0486, "step": 25530 }, { "grad_norm": 0.3825751543045044, "learning_rate": 8.145691669529792e-05, "loss": 0.03, "step": 25540 }, { "grad_norm": 0.4517919421195984, "learning_rate": 8.144084861945618e-05, "loss": 0.035, "step": 25550 }, { "grad_norm": 0.4383591115474701, "learning_rate": 8.142477517123333e-05, "loss": 0.0376, "step": 25560 }, { "grad_norm": 0.37217143177986145, "learning_rate": 8.140869635337586e-05, "loss": 0.0321, "step": 25570 }, { "grad_norm": 0.2099478393793106, "learning_rate": 8.139261216863123e-05, "loss": 0.0265, "step": 25580 }, { "grad_norm": 0.35654860734939575, "learning_rate": 8.137652261974776e-05, "loss": 0.0368, "step": 25590 }, { "grad_norm": 0.4054618179798126, "learning_rate": 8.136042770947472e-05, "loss": 0.0312, "step": 25600 }, { "grad_norm": 0.2760172486305237, "learning_rate": 8.134432744056228e-05, "loss": 0.0393, "step": 25610 }, { "grad_norm": 0.5210407376289368, "learning_rate": 8.132822181576158e-05, "loss": 0.04, "step": 25620 }, { "grad_norm": 0.3166964650154114, "learning_rate": 8.131211083782459e-05, "loss": 0.0319, "step": 25630 }, { "grad_norm": 0.2725782096385956, "learning_rate": 8.129599450950424e-05, "loss": 0.0297, "step": 25640 }, { "grad_norm": 0.5130347609519958, "learning_rate": 8.127987283355438e-05, "loss": 0.0395, "step": 25650 }, { "grad_norm": 0.40127211809158325, "learning_rate": 8.126374581272976e-05, "loss": 0.0289, "step": 25660 }, { "grad_norm": 0.43959107995033264, "learning_rate": 8.124761344978605e-05, "loss": 0.0476, "step": 25670 }, { "grad_norm": 0.2804381549358368, "learning_rate": 8.12314757474798e-05, "loss": 0.0283, "step": 25680 }, { "grad_norm": 0.4527730345726013, "learning_rate": 8.121533270856856e-05, "loss": 0.0302, "step": 25690 }, { "grad_norm": 0.31399378180503845, "learning_rate": 8.119918433581069e-05, "loss": 0.0341, "step": 25700 }, { "grad_norm": 0.45009252429008484, "learning_rate": 8.118303063196551e-05, "loss": 0.0368, "step": 25710 }, { "grad_norm": 0.28194862604141235, "learning_rate": 8.116687159979326e-05, "loss": 0.0334, "step": 25720 }, { "grad_norm": 0.44633781909942627, "learning_rate": 8.115070724205508e-05, "loss": 0.0344, "step": 25730 }, { "grad_norm": 0.4112047851085663, "learning_rate": 8.113453756151296e-05, "loss": 0.0334, "step": 25740 }, { "grad_norm": 0.5139113664627075, "learning_rate": 8.111836256092995e-05, "loss": 0.0305, "step": 25750 }, { "grad_norm": 0.24674628674983978, "learning_rate": 8.110218224306985e-05, "loss": 0.0336, "step": 25760 }, { "grad_norm": 0.3351249396800995, "learning_rate": 8.108599661069745e-05, "loss": 0.0318, "step": 25770 }, { "grad_norm": 0.3927977383136749, "learning_rate": 8.106980566657845e-05, "loss": 0.0273, "step": 25780 }, { "grad_norm": 0.30887672305107117, "learning_rate": 8.10536094134794e-05, "loss": 0.0335, "step": 25790 }, { "grad_norm": 0.36336401104927063, "learning_rate": 8.103740785416783e-05, "loss": 0.0292, "step": 25800 }, { "grad_norm": 0.4037589728832245, "learning_rate": 8.102120099141212e-05, "loss": 0.0374, "step": 25810 }, { "grad_norm": 0.3288711905479431, "learning_rate": 8.100498882798163e-05, "loss": 0.0259, "step": 25820 }, { "grad_norm": 0.3102432191371918, "learning_rate": 8.09887713666465e-05, "loss": 0.0262, "step": 25830 }, { "grad_norm": 0.36534780263900757, "learning_rate": 8.09725486101779e-05, "loss": 0.0306, "step": 25840 }, { "grad_norm": 0.3190428614616394, "learning_rate": 8.095632056134784e-05, "loss": 0.0353, "step": 25850 }, { "grad_norm": 0.2899703085422516, "learning_rate": 8.094008722292925e-05, "loss": 0.0322, "step": 25860 }, { "grad_norm": 0.29175320267677307, "learning_rate": 8.092384859769598e-05, "loss": 0.0314, "step": 25870 }, { "grad_norm": 0.32139137387275696, "learning_rate": 8.090760468842275e-05, "loss": 0.0398, "step": 25880 }, { "grad_norm": 0.26874488592147827, "learning_rate": 8.089135549788521e-05, "loss": 0.0349, "step": 25890 }, { "grad_norm": 0.37720149755477905, "learning_rate": 8.087510102885987e-05, "loss": 0.0248, "step": 25900 }, { "grad_norm": 0.3820059299468994, "learning_rate": 8.085884128412422e-05, "loss": 0.0361, "step": 25910 }, { "grad_norm": 0.43450090289115906, "learning_rate": 8.084257626645659e-05, "loss": 0.0328, "step": 25920 }, { "grad_norm": 0.4443514049053192, "learning_rate": 8.08263059786362e-05, "loss": 0.0346, "step": 25930 }, { "grad_norm": 0.48460954427719116, "learning_rate": 8.081003042344325e-05, "loss": 0.0353, "step": 25940 }, { "grad_norm": 0.2780267000198364, "learning_rate": 8.079374960365872e-05, "loss": 0.0345, "step": 25950 }, { "grad_norm": 0.38738423585891724, "learning_rate": 8.077746352206463e-05, "loss": 0.0363, "step": 25960 }, { "grad_norm": 0.4894638955593109, "learning_rate": 8.076117218144377e-05, "loss": 0.0384, "step": 25970 }, { "grad_norm": 0.40969741344451904, "learning_rate": 8.074487558457991e-05, "loss": 0.0339, "step": 25980 }, { "grad_norm": 0.27302876114845276, "learning_rate": 8.072857373425768e-05, "loss": 0.0336, "step": 25990 }, { "grad_norm": 1.5551221370697021, "learning_rate": 8.071226663326264e-05, "loss": 0.044, "step": 26000 }, { "grad_norm": 0.4271532893180847, "learning_rate": 8.069595428438121e-05, "loss": 0.0501, "step": 26010 }, { "grad_norm": 0.37175965309143066, "learning_rate": 8.067963669040072e-05, "loss": 0.0444, "step": 26020 }, { "grad_norm": 0.6474875807762146, "learning_rate": 8.066331385410942e-05, "loss": 0.0463, "step": 26030 }, { "grad_norm": 0.3940403461456299, "learning_rate": 8.064698577829641e-05, "loss": 0.04, "step": 26040 }, { "grad_norm": 0.32198768854141235, "learning_rate": 8.063065246575175e-05, "loss": 0.0371, "step": 26050 }, { "grad_norm": 0.3600611984729767, "learning_rate": 8.061431391926631e-05, "loss": 0.032, "step": 26060 }, { "grad_norm": 0.33096006512641907, "learning_rate": 8.059797014163195e-05, "loss": 0.0414, "step": 26070 }, { "grad_norm": 0.4904477298259735, "learning_rate": 8.058162113564133e-05, "loss": 0.0475, "step": 26080 }, { "grad_norm": 0.3331770896911621, "learning_rate": 8.056526690408806e-05, "loss": 0.0403, "step": 26090 }, { "grad_norm": 0.3746601939201355, "learning_rate": 8.054890744976666e-05, "loss": 0.0382, "step": 26100 }, { "grad_norm": 0.33299145102500916, "learning_rate": 8.053254277547248e-05, "loss": 0.0278, "step": 26110 }, { "grad_norm": 0.5025100111961365, "learning_rate": 8.051617288400182e-05, "loss": 0.0324, "step": 26120 }, { "grad_norm": 0.32709139585494995, "learning_rate": 8.049979777815182e-05, "loss": 0.0337, "step": 26130 }, { "grad_norm": 0.3738020360469818, "learning_rate": 8.048341746072054e-05, "loss": 0.0317, "step": 26140 }, { "grad_norm": 0.34721630811691284, "learning_rate": 8.046703193450696e-05, "loss": 0.0396, "step": 26150 }, { "grad_norm": 0.37539172172546387, "learning_rate": 8.04506412023109e-05, "loss": 0.0296, "step": 26160 }, { "grad_norm": 0.41528138518333435, "learning_rate": 8.043424526693306e-05, "loss": 0.0302, "step": 26170 }, { "grad_norm": 0.30622828006744385, "learning_rate": 8.04178441311751e-05, "loss": 0.0328, "step": 26180 }, { "grad_norm": 0.37961623072624207, "learning_rate": 8.04014377978395e-05, "loss": 0.0321, "step": 26190 }, { "grad_norm": 0.23820774257183075, "learning_rate": 8.038502626972967e-05, "loss": 0.0327, "step": 26200 }, { "grad_norm": 0.3758956789970398, "learning_rate": 8.036860954964989e-05, "loss": 0.0369, "step": 26210 }, { "grad_norm": 0.3189254701137543, "learning_rate": 8.035218764040531e-05, "loss": 0.0396, "step": 26220 }, { "grad_norm": 0.33710846304893494, "learning_rate": 8.033576054480199e-05, "loss": 0.0378, "step": 26230 }, { "grad_norm": 0.44882431626319885, "learning_rate": 8.031932826564688e-05, "loss": 0.0296, "step": 26240 }, { "grad_norm": 0.30127450823783875, "learning_rate": 8.030289080574782e-05, "loss": 0.0337, "step": 26250 }, { "grad_norm": 0.38022667169570923, "learning_rate": 8.028644816791349e-05, "loss": 0.0396, "step": 26260 }, { "grad_norm": 0.5010046362876892, "learning_rate": 8.027000035495351e-05, "loss": 0.0294, "step": 26270 }, { "grad_norm": 0.38325703144073486, "learning_rate": 8.025354736967836e-05, "loss": 0.0543, "step": 26280 }, { "grad_norm": 0.584912896156311, "learning_rate": 8.023708921489941e-05, "loss": 0.0522, "step": 26290 }, { "grad_norm": 0.46157681941986084, "learning_rate": 8.022062589342887e-05, "loss": 0.0442, "step": 26300 }, { "grad_norm": 0.3989030122756958, "learning_rate": 8.020415740807993e-05, "loss": 0.0448, "step": 26310 }, { "grad_norm": 1.2431731224060059, "learning_rate": 8.018768376166656e-05, "loss": 0.0359, "step": 26320 }, { "grad_norm": 0.37143367528915405, "learning_rate": 8.017120495700368e-05, "loss": 0.0344, "step": 26330 }, { "grad_norm": 0.40109431743621826, "learning_rate": 8.015472099690704e-05, "loss": 0.0344, "step": 26340 }, { "grad_norm": 0.3364817500114441, "learning_rate": 8.013823188419332e-05, "loss": 0.033, "step": 26350 }, { "grad_norm": 0.2422792911529541, "learning_rate": 8.012173762168006e-05, "loss": 0.032, "step": 26360 }, { "grad_norm": 0.39152565598487854, "learning_rate": 8.010523821218567e-05, "loss": 0.0288, "step": 26370 }, { "grad_norm": 0.37011849880218506, "learning_rate": 8.008873365852945e-05, "loss": 0.0303, "step": 26380 }, { "grad_norm": 0.6416373252868652, "learning_rate": 8.007222396353157e-05, "loss": 0.0406, "step": 26390 }, { "grad_norm": 0.43371689319610596, "learning_rate": 8.00557091300131e-05, "loss": 0.0299, "step": 26400 }, { "grad_norm": 0.29196083545684814, "learning_rate": 8.003918916079597e-05, "loss": 0.0346, "step": 26410 }, { "grad_norm": 0.28018465638160706, "learning_rate": 8.002266405870298e-05, "loss": 0.0293, "step": 26420 }, { "grad_norm": 0.28191813826560974, "learning_rate": 8.000613382655782e-05, "loss": 0.0422, "step": 26430 }, { "grad_norm": 0.3988160192966461, "learning_rate": 7.998959846718505e-05, "loss": 0.0397, "step": 26440 }, { "grad_norm": 0.33238956332206726, "learning_rate": 7.997305798341012e-05, "loss": 0.0403, "step": 26450 }, { "grad_norm": 0.3768060505390167, "learning_rate": 7.995651237805937e-05, "loss": 0.0344, "step": 26460 }, { "grad_norm": 0.3784100413322449, "learning_rate": 7.993996165395996e-05, "loss": 0.0338, "step": 26470 }, { "grad_norm": 0.3367038071155548, "learning_rate": 7.992340581393996e-05, "loss": 0.0291, "step": 26480 }, { "grad_norm": 0.3796812891960144, "learning_rate": 7.990684486082831e-05, "loss": 0.0412, "step": 26490 }, { "grad_norm": 0.4058787524700165, "learning_rate": 7.989027879745482e-05, "loss": 0.04, "step": 26500 }, { "grad_norm": 0.3677363991737366, "learning_rate": 7.98737076266502e-05, "loss": 0.046, "step": 26510 }, { "grad_norm": 0.4304004907608032, "learning_rate": 7.985713135124598e-05, "loss": 0.0373, "step": 26520 }, { "grad_norm": 0.39748042821884155, "learning_rate": 7.98405499740746e-05, "loss": 0.035, "step": 26530 }, { "grad_norm": 0.4751295745372772, "learning_rate": 7.98239634979694e-05, "loss": 0.033, "step": 26540 }, { "grad_norm": 0.41155409812927246, "learning_rate": 7.98073719257645e-05, "loss": 0.0297, "step": 26550 }, { "grad_norm": 0.3431420922279358, "learning_rate": 7.979077526029499e-05, "loss": 0.0349, "step": 26560 }, { "grad_norm": 0.4268607497215271, "learning_rate": 7.977417350439675e-05, "loss": 0.0287, "step": 26570 }, { "grad_norm": 0.3492754101753235, "learning_rate": 7.97575666609066e-05, "loss": 0.0332, "step": 26580 }, { "grad_norm": 0.398622065782547, "learning_rate": 7.974095473266216e-05, "loss": 0.0326, "step": 26590 }, { "grad_norm": 0.3232441544532776, "learning_rate": 7.972433772250198e-05, "loss": 0.035, "step": 26600 }, { "grad_norm": 0.36239123344421387, "learning_rate": 7.970771563326544e-05, "loss": 0.029, "step": 26610 }, { "grad_norm": 0.3604343831539154, "learning_rate": 7.96910884677928e-05, "loss": 0.0342, "step": 26620 }, { "grad_norm": 0.468720942735672, "learning_rate": 7.967445622892523e-05, "loss": 0.0367, "step": 26630 }, { "grad_norm": 0.3323296010494232, "learning_rate": 7.965781891950465e-05, "loss": 0.0325, "step": 26640 }, { "grad_norm": 0.568403422832489, "learning_rate": 7.964117654237397e-05, "loss": 0.03, "step": 26650 }, { "grad_norm": 0.3203636109828949, "learning_rate": 7.962452910037692e-05, "loss": 0.0361, "step": 26660 }, { "grad_norm": 0.2963915467262268, "learning_rate": 7.96078765963581e-05, "loss": 0.0322, "step": 26670 }, { "grad_norm": 0.4389905333518982, "learning_rate": 7.95912190331629e-05, "loss": 0.0415, "step": 26680 }, { "grad_norm": 0.43026894330978394, "learning_rate": 7.957455641363772e-05, "loss": 0.0374, "step": 26690 }, { "grad_norm": 0.4443652927875519, "learning_rate": 7.955788874062968e-05, "loss": 0.0462, "step": 26700 }, { "grad_norm": 0.3728642761707306, "learning_rate": 7.95412160169869e-05, "loss": 0.0327, "step": 26710 }, { "grad_norm": 0.5793817043304443, "learning_rate": 7.952453824555824e-05, "loss": 0.0409, "step": 26720 }, { "grad_norm": 0.48007404804229736, "learning_rate": 7.95078554291935e-05, "loss": 0.0338, "step": 26730 }, { "grad_norm": 0.3445594310760498, "learning_rate": 7.94911675707433e-05, "loss": 0.0331, "step": 26740 }, { "grad_norm": 0.3379964530467987, "learning_rate": 7.947447467305915e-05, "loss": 0.0373, "step": 26750 }, { "grad_norm": 0.3634724020957947, "learning_rate": 7.94577767389934e-05, "loss": 0.0314, "step": 26760 }, { "grad_norm": 0.31728795170783997, "learning_rate": 7.944107377139928e-05, "loss": 0.0254, "step": 26770 }, { "grad_norm": 0.28369563817977905, "learning_rate": 7.942436577313088e-05, "loss": 0.0321, "step": 26780 }, { "grad_norm": 0.35839882493019104, "learning_rate": 7.940765274704312e-05, "loss": 0.0305, "step": 26790 }, { "grad_norm": 0.5018502473831177, "learning_rate": 7.939093469599181e-05, "loss": 0.0322, "step": 26800 }, { "grad_norm": 0.3517337143421173, "learning_rate": 7.93742116228336e-05, "loss": 0.0333, "step": 26810 }, { "grad_norm": 0.29996922612190247, "learning_rate": 7.935748353042602e-05, "loss": 0.026, "step": 26820 }, { "grad_norm": 0.40694937109947205, "learning_rate": 7.934075042162744e-05, "loss": 0.0357, "step": 26830 }, { "grad_norm": 0.749197244644165, "learning_rate": 7.932401229929705e-05, "loss": 0.0399, "step": 26840 }, { "grad_norm": 0.4899929165840149, "learning_rate": 7.9307269166295e-05, "loss": 0.0353, "step": 26850 }, { "grad_norm": 0.42791783809661865, "learning_rate": 7.92905210254822e-05, "loss": 0.0315, "step": 26860 }, { "grad_norm": 0.5126593708992004, "learning_rate": 7.927376787972045e-05, "loss": 0.0327, "step": 26870 }, { "grad_norm": 0.3470296561717987, "learning_rate": 7.92570097318724e-05, "loss": 0.0405, "step": 26880 }, { "grad_norm": 0.3299737274646759, "learning_rate": 7.924024658480158e-05, "loss": 0.0375, "step": 26890 }, { "grad_norm": 0.35473155975341797, "learning_rate": 7.922347844137233e-05, "loss": 0.0335, "step": 26900 }, { "grad_norm": 0.49158555269241333, "learning_rate": 7.92067053044499e-05, "loss": 0.0366, "step": 26910 }, { "grad_norm": 0.49690619111061096, "learning_rate": 7.918992717690031e-05, "loss": 0.0409, "step": 26920 }, { "grad_norm": 0.4045504331588745, "learning_rate": 7.917314406159053e-05, "loss": 0.0358, "step": 26930 }, { "grad_norm": 0.5993085503578186, "learning_rate": 7.915635596138832e-05, "loss": 0.0521, "step": 26940 }, { "grad_norm": 0.3920290172100067, "learning_rate": 7.913956287916228e-05, "loss": 0.0375, "step": 26950 }, { "grad_norm": 0.2985778748989105, "learning_rate": 7.912276481778193e-05, "loss": 0.0265, "step": 26960 }, { "grad_norm": 0.39541321992874146, "learning_rate": 7.910596178011759e-05, "loss": 0.0436, "step": 26970 }, { "grad_norm": 0.27434152364730835, "learning_rate": 7.908915376904043e-05, "loss": 0.0232, "step": 26980 }, { "grad_norm": 0.34152254462242126, "learning_rate": 7.907234078742247e-05, "loss": 0.0356, "step": 26990 }, { "grad_norm": 0.29680222272872925, "learning_rate": 7.90555228381366e-05, "loss": 0.0287, "step": 27000 }, { "grad_norm": 0.33201634883880615, "learning_rate": 7.903869992405656e-05, "loss": 0.0234, "step": 27010 }, { "grad_norm": 0.30162909626960754, "learning_rate": 7.902187204805691e-05, "loss": 0.0305, "step": 27020 }, { "grad_norm": 0.3537808060646057, "learning_rate": 7.900503921301308e-05, "loss": 0.0245, "step": 27030 }, { "grad_norm": 0.3329593539237976, "learning_rate": 7.898820142180133e-05, "loss": 0.0431, "step": 27040 }, { "grad_norm": 0.35471129417419434, "learning_rate": 7.897135867729879e-05, "loss": 0.0338, "step": 27050 }, { "grad_norm": 0.47744572162628174, "learning_rate": 7.89545109823834e-05, "loss": 0.0311, "step": 27060 }, { "grad_norm": 0.3177913427352905, "learning_rate": 7.8937658339934e-05, "loss": 0.0308, "step": 27070 }, { "grad_norm": 0.360048770904541, "learning_rate": 7.892080075283026e-05, "loss": 0.034, "step": 27080 }, { "grad_norm": 0.3493550419807434, "learning_rate": 7.890393822395263e-05, "loss": 0.0352, "step": 27090 }, { "grad_norm": 0.28734612464904785, "learning_rate": 7.88870707561825e-05, "loss": 0.0289, "step": 27100 }, { "grad_norm": 0.629193902015686, "learning_rate": 7.887019835240203e-05, "loss": 0.0315, "step": 27110 }, { "grad_norm": 0.3501876890659332, "learning_rate": 7.885332101549427e-05, "loss": 0.034, "step": 27120 }, { "grad_norm": 0.28195467591285706, "learning_rate": 7.883643874834308e-05, "loss": 0.0284, "step": 27130 }, { "grad_norm": 0.37400466203689575, "learning_rate": 7.881955155383321e-05, "loss": 0.0436, "step": 27140 }, { "grad_norm": 0.3973286747932434, "learning_rate": 7.880265943485017e-05, "loss": 0.0319, "step": 27150 }, { "grad_norm": 0.4060206413269043, "learning_rate": 7.878576239428038e-05, "loss": 0.0359, "step": 27160 }, { "grad_norm": 0.4312794804573059, "learning_rate": 7.87688604350111e-05, "loss": 0.0409, "step": 27170 }, { "grad_norm": 0.6051466464996338, "learning_rate": 7.875195355993042e-05, "loss": 0.0398, "step": 27180 }, { "grad_norm": 0.3780898451805115, "learning_rate": 7.873504177192724e-05, "loss": 0.0381, "step": 27190 }, { "grad_norm": 0.40554776787757874, "learning_rate": 7.87181250738913e-05, "loss": 0.0434, "step": 27200 }, { "grad_norm": 0.3884384036064148, "learning_rate": 7.870120346871324e-05, "loss": 0.0317, "step": 27210 }, { "grad_norm": 0.3649645447731018, "learning_rate": 7.86842769592845e-05, "loss": 0.0305, "step": 27220 }, { "grad_norm": 0.39767956733703613, "learning_rate": 7.866734554849732e-05, "loss": 0.0303, "step": 27230 }, { "grad_norm": 0.3404722511768341, "learning_rate": 7.865040923924486e-05, "loss": 0.0345, "step": 27240 }, { "grad_norm": 0.44511619210243225, "learning_rate": 7.863346803442104e-05, "loss": 0.0435, "step": 27250 }, { "grad_norm": 0.45017027854919434, "learning_rate": 7.861652193692067e-05, "loss": 0.0292, "step": 27260 }, { "grad_norm": 0.34396663308143616, "learning_rate": 7.859957094963937e-05, "loss": 0.0273, "step": 27270 }, { "grad_norm": 0.3268871307373047, "learning_rate": 7.858261507547357e-05, "loss": 0.0269, "step": 27280 }, { "grad_norm": 0.39846092462539673, "learning_rate": 7.856565431732061e-05, "loss": 0.026, "step": 27290 }, { "grad_norm": 0.3388707935810089, "learning_rate": 7.854868867807859e-05, "loss": 0.0264, "step": 27300 }, { "grad_norm": 0.4404560625553131, "learning_rate": 7.85317181606465e-05, "loss": 0.0288, "step": 27310 }, { "grad_norm": 0.34482404589653015, "learning_rate": 7.85147427679241e-05, "loss": 0.0299, "step": 27320 }, { "grad_norm": 0.34981027245521545, "learning_rate": 7.849776250281205e-05, "loss": 0.0255, "step": 27330 }, { "grad_norm": 0.2518521249294281, "learning_rate": 7.84807773682118e-05, "loss": 0.0311, "step": 27340 }, { "grad_norm": 0.36319056153297424, "learning_rate": 7.846378736702565e-05, "loss": 0.0393, "step": 27350 }, { "grad_norm": 0.371356338262558, "learning_rate": 7.844679250215671e-05, "loss": 0.0372, "step": 27360 }, { "grad_norm": 0.3471572995185852, "learning_rate": 7.842979277650898e-05, "loss": 0.0292, "step": 27370 }, { "grad_norm": 0.33939334750175476, "learning_rate": 7.84127881929872e-05, "loss": 0.0339, "step": 27380 }, { "grad_norm": 0.3804114758968353, "learning_rate": 7.839577875449704e-05, "loss": 0.0285, "step": 27390 }, { "grad_norm": 0.30348342657089233, "learning_rate": 7.837876446394489e-05, "loss": 0.0246, "step": 27400 }, { "grad_norm": 0.28381556272506714, "learning_rate": 7.836174532423805e-05, "loss": 0.0342, "step": 27410 }, { "grad_norm": 0.4153187572956085, "learning_rate": 7.834472133828466e-05, "loss": 0.0297, "step": 27420 }, { "grad_norm": 0.4274923801422119, "learning_rate": 7.832769250899359e-05, "loss": 0.0377, "step": 27430 }, { "grad_norm": 0.5352441668510437, "learning_rate": 7.831065883927464e-05, "loss": 0.0379, "step": 27440 }, { "grad_norm": 0.38190245628356934, "learning_rate": 7.829362033203841e-05, "loss": 0.0322, "step": 27450 }, { "grad_norm": 0.3432736396789551, "learning_rate": 7.827657699019628e-05, "loss": 0.0274, "step": 27460 }, { "grad_norm": 0.3652041256427765, "learning_rate": 7.825952881666052e-05, "loss": 0.0435, "step": 27470 }, { "grad_norm": 0.4587835371494293, "learning_rate": 7.824247581434418e-05, "loss": 0.0348, "step": 27480 }, { "grad_norm": 0.4364752769470215, "learning_rate": 7.822541798616116e-05, "loss": 0.0351, "step": 27490 }, { "grad_norm": 0.3338066637516022, "learning_rate": 7.820835533502617e-05, "loss": 0.0444, "step": 27500 }, { "grad_norm": 0.4772060513496399, "learning_rate": 7.819128786385475e-05, "loss": 0.0365, "step": 27510 }, { "grad_norm": 0.28433990478515625, "learning_rate": 7.817421557556329e-05, "loss": 0.0381, "step": 27520 }, { "grad_norm": 0.344246506690979, "learning_rate": 7.815713847306893e-05, "loss": 0.0357, "step": 27530 }, { "grad_norm": 0.33713972568511963, "learning_rate": 7.81400565592897e-05, "loss": 0.0287, "step": 27540 }, { "grad_norm": 0.4315725564956665, "learning_rate": 7.812296983714444e-05, "loss": 0.0396, "step": 27550 }, { "grad_norm": 0.3968774378299713, "learning_rate": 7.810587830955281e-05, "loss": 0.0337, "step": 27560 }, { "grad_norm": 0.4961390793323517, "learning_rate": 7.808878197943528e-05, "loss": 0.0347, "step": 27570 }, { "grad_norm": 0.45568183064460754, "learning_rate": 7.807168084971312e-05, "loss": 0.0335, "step": 27580 }, { "grad_norm": 0.2507712244987488, "learning_rate": 7.805457492330849e-05, "loss": 0.0294, "step": 27590 }, { "grad_norm": 0.45791828632354736, "learning_rate": 7.803746420314428e-05, "loss": 0.0385, "step": 27600 }, { "grad_norm": 0.3444938659667969, "learning_rate": 7.802034869214428e-05, "loss": 0.0388, "step": 27610 }, { "grad_norm": 0.42233484983444214, "learning_rate": 7.800322839323303e-05, "loss": 0.0334, "step": 27620 }, { "grad_norm": 0.26727285981178284, "learning_rate": 7.798610330933593e-05, "loss": 0.0281, "step": 27630 }, { "grad_norm": 0.28353312611579895, "learning_rate": 7.796897344337922e-05, "loss": 0.0324, "step": 27640 }, { "grad_norm": 0.39616212248802185, "learning_rate": 7.795183879828989e-05, "loss": 0.0382, "step": 27650 }, { "grad_norm": 0.42685529589653015, "learning_rate": 7.793469937699579e-05, "loss": 0.0479, "step": 27660 }, { "grad_norm": 0.26343420147895813, "learning_rate": 7.791755518242558e-05, "loss": 0.0316, "step": 27670 }, { "grad_norm": 0.6618173122406006, "learning_rate": 7.790040621750876e-05, "loss": 0.0494, "step": 27680 }, { "grad_norm": 0.4615534543991089, "learning_rate": 7.788325248517558e-05, "loss": 0.0332, "step": 27690 }, { "grad_norm": 0.4543582499027252, "learning_rate": 7.786609398835715e-05, "loss": 0.0318, "step": 27700 }, { "grad_norm": 0.2980612814426422, "learning_rate": 7.784893072998541e-05, "loss": 0.0275, "step": 27710 }, { "grad_norm": 0.42928802967071533, "learning_rate": 7.783176271299306e-05, "loss": 0.0298, "step": 27720 }, { "grad_norm": 0.356452614068985, "learning_rate": 7.781458994031368e-05, "loss": 0.0241, "step": 27730 }, { "grad_norm": 0.3221862018108368, "learning_rate": 7.779741241488161e-05, "loss": 0.0274, "step": 27740 }, { "grad_norm": 0.3040800094604492, "learning_rate": 7.7780230139632e-05, "loss": 0.0232, "step": 27750 }, { "grad_norm": 0.38448256254196167, "learning_rate": 7.776304311750087e-05, "loss": 0.0298, "step": 27760 }, { "grad_norm": 0.3841462731361389, "learning_rate": 7.7745851351425e-05, "loss": 0.0412, "step": 27770 }, { "grad_norm": 0.34536904096603394, "learning_rate": 7.772865484434197e-05, "loss": 0.028, "step": 27780 }, { "grad_norm": 0.4823185205459595, "learning_rate": 7.77114535991902e-05, "loss": 0.0434, "step": 27790 }, { "grad_norm": 0.43752479553222656, "learning_rate": 7.769424761890893e-05, "loss": 0.0378, "step": 27800 }, { "grad_norm": 0.44709718227386475, "learning_rate": 7.767703690643817e-05, "loss": 0.0409, "step": 27810 }, { "grad_norm": 0.2819918692111969, "learning_rate": 7.76598214647188e-05, "loss": 0.032, "step": 27820 }, { "grad_norm": 0.8040570616722107, "learning_rate": 7.764260129669241e-05, "loss": 0.0391, "step": 27830 }, { "grad_norm": 0.2919732928276062, "learning_rate": 7.76253764053015e-05, "loss": 0.0395, "step": 27840 }, { "grad_norm": 0.5053101181983948, "learning_rate": 7.760814679348932e-05, "loss": 0.0302, "step": 27850 }, { "grad_norm": 0.3537231683731079, "learning_rate": 7.759091246419992e-05, "loss": 0.0309, "step": 27860 }, { "grad_norm": 0.23033268749713898, "learning_rate": 7.75736734203782e-05, "loss": 0.0246, "step": 27870 }, { "grad_norm": 0.4644320607185364, "learning_rate": 7.755642966496985e-05, "loss": 0.0489, "step": 27880 }, { "grad_norm": 0.361909955739975, "learning_rate": 7.753918120092132e-05, "loss": 0.0398, "step": 27890 }, { "grad_norm": 0.39941978454589844, "learning_rate": 7.752192803117993e-05, "loss": 0.0317, "step": 27900 }, { "grad_norm": 0.4699914753437042, "learning_rate": 7.750467015869377e-05, "loss": 0.0353, "step": 27910 }, { "grad_norm": 0.3129115700721741, "learning_rate": 7.748740758641174e-05, "loss": 0.0302, "step": 27920 }, { "grad_norm": 0.3406308591365814, "learning_rate": 7.74701403172835e-05, "loss": 0.0352, "step": 27930 }, { "grad_norm": 0.3880663514137268, "learning_rate": 7.745286835425962e-05, "loss": 0.0355, "step": 27940 }, { "grad_norm": 0.5981401205062866, "learning_rate": 7.743559170029138e-05, "loss": 0.0384, "step": 27950 }, { "grad_norm": 0.42440393567085266, "learning_rate": 7.741831035833087e-05, "loss": 0.0347, "step": 27960 }, { "grad_norm": 0.26552286744117737, "learning_rate": 7.740102433133102e-05, "loss": 0.0298, "step": 27970 }, { "grad_norm": 0.2521427571773529, "learning_rate": 7.738373362224553e-05, "loss": 0.0291, "step": 27980 }, { "grad_norm": 0.3150288760662079, "learning_rate": 7.73664382340289e-05, "loss": 0.0306, "step": 27990 }, { "grad_norm": 0.3676430881023407, "learning_rate": 7.734913816963647e-05, "loss": 0.0365, "step": 28000 }, { "grad_norm": 0.27925705909729004, "learning_rate": 7.73318334320243e-05, "loss": 0.028, "step": 28010 }, { "grad_norm": 0.405240535736084, "learning_rate": 7.731452402414934e-05, "loss": 0.0378, "step": 28020 }, { "grad_norm": 0.5754651427268982, "learning_rate": 7.729720994896928e-05, "loss": 0.0306, "step": 28030 }, { "grad_norm": 0.5467476844787598, "learning_rate": 7.727989120944262e-05, "loss": 0.0327, "step": 28040 }, { "grad_norm": 0.3322755992412567, "learning_rate": 7.726256780852865e-05, "loss": 0.0283, "step": 28050 }, { "grad_norm": 0.434957891702652, "learning_rate": 7.724523974918749e-05, "loss": 0.0279, "step": 28060 }, { "grad_norm": 0.5109728574752808, "learning_rate": 7.722790703438002e-05, "loss": 0.0296, "step": 28070 }, { "grad_norm": 0.5143767595291138, "learning_rate": 7.72105696670679e-05, "loss": 0.0264, "step": 28080 }, { "grad_norm": 0.5442364811897278, "learning_rate": 7.719322765021364e-05, "loss": 0.0348, "step": 28090 }, { "grad_norm": 0.30463260412216187, "learning_rate": 7.717588098678051e-05, "loss": 0.0315, "step": 28100 }, { "grad_norm": 0.38179221749305725, "learning_rate": 7.715852967973258e-05, "loss": 0.0287, "step": 28110 }, { "grad_norm": 0.35545992851257324, "learning_rate": 7.714117373203474e-05, "loss": 0.033, "step": 28120 }, { "grad_norm": 0.7798536419868469, "learning_rate": 7.712381314665259e-05, "loss": 0.0304, "step": 28130 }, { "grad_norm": 0.3365922272205353, "learning_rate": 7.710644792655261e-05, "loss": 0.0291, "step": 28140 }, { "grad_norm": 0.5619332790374756, "learning_rate": 7.708907807470207e-05, "loss": 0.0464, "step": 28150 }, { "grad_norm": 0.38042545318603516, "learning_rate": 7.707170359406896e-05, "loss": 0.0346, "step": 28160 }, { "grad_norm": 0.33315110206604004, "learning_rate": 7.705432448762213e-05, "loss": 0.0282, "step": 28170 }, { "grad_norm": 0.3422733247280121, "learning_rate": 7.703694075833117e-05, "loss": 0.0435, "step": 28180 }, { "grad_norm": 0.3328172564506531, "learning_rate": 7.70195524091665e-05, "loss": 0.0429, "step": 28190 }, { "grad_norm": 0.39534032344818115, "learning_rate": 7.70021594430993e-05, "loss": 0.0384, "step": 28200 }, { "grad_norm": 0.3188535273075104, "learning_rate": 7.698476186310157e-05, "loss": 0.0263, "step": 28210 }, { "grad_norm": 0.3265352249145508, "learning_rate": 7.696735967214608e-05, "loss": 0.0329, "step": 28220 }, { "grad_norm": 0.3470901548862457, "learning_rate": 7.694995287320636e-05, "loss": 0.0382, "step": 28230 }, { "grad_norm": 0.4638511836528778, "learning_rate": 7.693254146925679e-05, "loss": 0.0292, "step": 28240 }, { "grad_norm": 0.4070777893066406, "learning_rate": 7.691512546327251e-05, "loss": 0.0386, "step": 28250 }, { "grad_norm": 0.39532262086868286, "learning_rate": 7.689770485822939e-05, "loss": 0.0388, "step": 28260 }, { "grad_norm": 0.3124167323112488, "learning_rate": 7.688027965710416e-05, "loss": 0.0286, "step": 28270 }, { "grad_norm": 0.2634918987751007, "learning_rate": 7.686284986287433e-05, "loss": 0.0357, "step": 28280 }, { "grad_norm": 0.315676748752594, "learning_rate": 7.684541547851817e-05, "loss": 0.0322, "step": 28290 }, { "grad_norm": 0.2076706886291504, "learning_rate": 7.68279765070147e-05, "loss": 0.0326, "step": 28300 }, { "grad_norm": 0.4560741186141968, "learning_rate": 7.68105329513438e-05, "loss": 0.0309, "step": 28310 }, { "grad_norm": 0.3372134566307068, "learning_rate": 7.67930848144861e-05, "loss": 0.0297, "step": 28320 }, { "grad_norm": 0.3421724736690521, "learning_rate": 7.6775632099423e-05, "loss": 0.0362, "step": 28330 }, { "grad_norm": 0.2919286787509918, "learning_rate": 7.675817480913667e-05, "loss": 0.0343, "step": 28340 }, { "grad_norm": 0.39288488030433655, "learning_rate": 7.674071294661011e-05, "loss": 0.0252, "step": 28350 }, { "grad_norm": 0.3893623650074005, "learning_rate": 7.672324651482707e-05, "loss": 0.0281, "step": 28360 }, { "grad_norm": 0.3457542359828949, "learning_rate": 7.670577551677209e-05, "loss": 0.0342, "step": 28370 }, { "grad_norm": 0.3791887164115906, "learning_rate": 7.668829995543047e-05, "loss": 0.0288, "step": 28380 }, { "grad_norm": 0.27480486035346985, "learning_rate": 7.667081983378832e-05, "loss": 0.0344, "step": 28390 }, { "grad_norm": 0.4780431091785431, "learning_rate": 7.66533351548325e-05, "loss": 0.029, "step": 28400 }, { "grad_norm": 0.4719846844673157, "learning_rate": 7.663584592155069e-05, "loss": 0.0348, "step": 28410 }, { "grad_norm": 0.4159947633743286, "learning_rate": 7.661835213693129e-05, "loss": 0.0384, "step": 28420 }, { "grad_norm": 0.4235331118106842, "learning_rate": 7.660085380396353e-05, "loss": 0.0348, "step": 28430 }, { "grad_norm": 0.37992578744888306, "learning_rate": 7.658335092563738e-05, "loss": 0.0286, "step": 28440 }, { "grad_norm": 0.27054959535598755, "learning_rate": 7.656584350494362e-05, "loss": 0.0273, "step": 28450 }, { "grad_norm": 0.40380507707595825, "learning_rate": 7.654833154487378e-05, "loss": 0.0325, "step": 28460 }, { "grad_norm": 0.3106308579444885, "learning_rate": 7.653081504842017e-05, "loss": 0.0298, "step": 28470 }, { "grad_norm": 0.4055485129356384, "learning_rate": 7.65132940185759e-05, "loss": 0.0349, "step": 28480 }, { "grad_norm": 0.41141387820243835, "learning_rate": 7.649576845833481e-05, "loss": 0.0296, "step": 28490 }, { "grad_norm": 0.35169896483421326, "learning_rate": 7.647823837069156e-05, "loss": 0.0363, "step": 28500 }, { "grad_norm": 0.34128016233444214, "learning_rate": 7.646070375864156e-05, "loss": 0.0297, "step": 28510 }, { "grad_norm": 0.5309964418411255, "learning_rate": 7.644316462518099e-05, "loss": 0.0366, "step": 28520 }, { "grad_norm": 0.3352823257446289, "learning_rate": 7.642562097330679e-05, "loss": 0.0296, "step": 28530 }, { "grad_norm": 0.26604244112968445, "learning_rate": 7.640807280601671e-05, "loss": 0.0302, "step": 28540 }, { "grad_norm": 0.4108028709888458, "learning_rate": 7.639052012630927e-05, "loss": 0.0421, "step": 28550 }, { "grad_norm": 0.3153860569000244, "learning_rate": 7.63729629371837e-05, "loss": 0.0308, "step": 28560 }, { "grad_norm": 0.3496893346309662, "learning_rate": 7.635540124164009e-05, "loss": 0.0405, "step": 28570 }, { "grad_norm": 0.32845816016197205, "learning_rate": 7.633783504267922e-05, "loss": 0.0367, "step": 28580 }, { "grad_norm": 0.5120717287063599, "learning_rate": 7.632026434330269e-05, "loss": 0.0359, "step": 28590 }, { "grad_norm": 0.30877912044525146, "learning_rate": 7.630268914651282e-05, "loss": 0.0313, "step": 28600 }, { "grad_norm": 0.46231377124786377, "learning_rate": 7.628510945531278e-05, "loss": 0.0479, "step": 28610 }, { "grad_norm": 0.3496294617652893, "learning_rate": 7.626752527270641e-05, "loss": 0.0374, "step": 28620 }, { "grad_norm": 0.2903108298778534, "learning_rate": 7.62499366016984e-05, "loss": 0.0256, "step": 28630 }, { "grad_norm": 0.250936895608902, "learning_rate": 7.623234344529416e-05, "loss": 0.0349, "step": 28640 }, { "grad_norm": 0.24341997504234314, "learning_rate": 7.62147458064999e-05, "loss": 0.0269, "step": 28650 }, { "grad_norm": 0.4546489417552948, "learning_rate": 7.619714368832254e-05, "loss": 0.0364, "step": 28660 }, { "grad_norm": 0.5441449880599976, "learning_rate": 7.61795370937698e-05, "loss": 0.0338, "step": 28670 }, { "grad_norm": 0.24462804198265076, "learning_rate": 7.61619260258502e-05, "loss": 0.0315, "step": 28680 }, { "grad_norm": 0.3162288963794708, "learning_rate": 7.614431048757298e-05, "loss": 0.0368, "step": 28690 }, { "grad_norm": 0.21540813148021698, "learning_rate": 7.612669048194814e-05, "loss": 0.0333, "step": 28700 }, { "grad_norm": 0.3393607437610626, "learning_rate": 7.610906601198646e-05, "loss": 0.0285, "step": 28710 }, { "grad_norm": 0.3221277892589569, "learning_rate": 7.60914370806995e-05, "loss": 0.0373, "step": 28720 }, { "grad_norm": 0.3264603018760681, "learning_rate": 7.607380369109953e-05, "loss": 0.05, "step": 28730 }, { "grad_norm": 0.3384207487106323, "learning_rate": 7.605616584619961e-05, "loss": 0.0307, "step": 28740 }, { "grad_norm": 0.3876554071903229, "learning_rate": 7.603852354901362e-05, "loss": 0.0323, "step": 28750 }, { "grad_norm": 0.36261722445487976, "learning_rate": 7.602087680255609e-05, "loss": 0.0431, "step": 28760 }, { "grad_norm": 0.37923282384872437, "learning_rate": 7.600322560984238e-05, "loss": 0.0328, "step": 28770 }, { "grad_norm": 0.26985159516334534, "learning_rate": 7.598556997388863e-05, "loss": 0.0298, "step": 28780 }, { "grad_norm": 0.5513350367546082, "learning_rate": 7.596790989771166e-05, "loss": 0.042, "step": 28790 }, { "grad_norm": 0.32837235927581787, "learning_rate": 7.595024538432914e-05, "loss": 0.0342, "step": 28800 }, { "grad_norm": 0.4205987751483917, "learning_rate": 7.59325764367594e-05, "loss": 0.0282, "step": 28810 }, { "grad_norm": 0.34927791357040405, "learning_rate": 7.59149030580216e-05, "loss": 0.0292, "step": 28820 }, { "grad_norm": 0.29764196276664734, "learning_rate": 7.589722525113562e-05, "loss": 0.0262, "step": 28830 }, { "grad_norm": 0.3682582676410675, "learning_rate": 7.587954301912216e-05, "loss": 0.0265, "step": 28840 }, { "grad_norm": 0.5357165336608887, "learning_rate": 7.586185636500263e-05, "loss": 0.0337, "step": 28850 }, { "grad_norm": 0.3255724012851715, "learning_rate": 7.584416529179914e-05, "loss": 0.035, "step": 28860 }, { "grad_norm": 0.3141203820705414, "learning_rate": 7.582646980253465e-05, "loss": 0.0329, "step": 28870 }, { "grad_norm": 0.4284389019012451, "learning_rate": 7.580876990023282e-05, "loss": 0.0415, "step": 28880 }, { "grad_norm": 0.3585684597492218, "learning_rate": 7.579106558791809e-05, "loss": 0.0338, "step": 28890 }, { "grad_norm": 0.25834307074546814, "learning_rate": 7.577335686861565e-05, "loss": 0.0263, "step": 28900 }, { "grad_norm": 0.5777397155761719, "learning_rate": 7.575564374535141e-05, "loss": 0.0313, "step": 28910 }, { "grad_norm": 0.27768823504447937, "learning_rate": 7.573792622115207e-05, "loss": 0.0245, "step": 28920 }, { "grad_norm": 0.375179260969162, "learning_rate": 7.572020429904507e-05, "loss": 0.0283, "step": 28930 }, { "grad_norm": 0.44336333870887756, "learning_rate": 7.570247798205861e-05, "loss": 0.036, "step": 28940 }, { "grad_norm": 0.31640684604644775, "learning_rate": 7.568474727322164e-05, "loss": 0.0326, "step": 28950 }, { "grad_norm": 0.41115686297416687, "learning_rate": 7.566701217556384e-05, "loss": 0.0283, "step": 28960 }, { "grad_norm": 0.39235222339630127, "learning_rate": 7.564927269211564e-05, "loss": 0.0297, "step": 28970 }, { "grad_norm": 0.3683130443096161, "learning_rate": 7.563152882590824e-05, "loss": 0.0258, "step": 28980 }, { "grad_norm": 0.5890649557113647, "learning_rate": 7.56137805799736e-05, "loss": 0.0355, "step": 28990 }, { "grad_norm": 0.4639614224433899, "learning_rate": 7.559602795734439e-05, "loss": 0.0346, "step": 29000 }, { "grad_norm": 0.3716409206390381, "learning_rate": 7.557827096105408e-05, "loss": 0.0337, "step": 29010 }, { "grad_norm": 0.517916202545166, "learning_rate": 7.55605095941368e-05, "loss": 0.0373, "step": 29020 }, { "grad_norm": 0.29634952545166016, "learning_rate": 7.554274385962752e-05, "loss": 0.0329, "step": 29030 }, { "grad_norm": 0.4178181290626526, "learning_rate": 7.552497376056191e-05, "loss": 0.033, "step": 29040 }, { "grad_norm": 0.2991536557674408, "learning_rate": 7.550719929997639e-05, "loss": 0.0312, "step": 29050 }, { "grad_norm": 0.418792188167572, "learning_rate": 7.548942048090813e-05, "loss": 0.0302, "step": 29060 }, { "grad_norm": 0.4336537718772888, "learning_rate": 7.547163730639506e-05, "loss": 0.0292, "step": 29070 }, { "grad_norm": 0.36301955580711365, "learning_rate": 7.545384977947583e-05, "loss": 0.033, "step": 29080 }, { "grad_norm": 0.4598860740661621, "learning_rate": 7.543605790318981e-05, "loss": 0.0277, "step": 29090 }, { "grad_norm": 0.30236247181892395, "learning_rate": 7.54182616805772e-05, "loss": 0.0308, "step": 29100 }, { "grad_norm": 0.2434448003768921, "learning_rate": 7.540046111467885e-05, "loss": 0.0301, "step": 29110 }, { "grad_norm": 0.2963416278362274, "learning_rate": 7.53826562085364e-05, "loss": 0.0314, "step": 29120 }, { "grad_norm": 0.3959903419017792, "learning_rate": 7.536484696519221e-05, "loss": 0.0198, "step": 29130 }, { "grad_norm": 0.47796621918678284, "learning_rate": 7.534703338768942e-05, "loss": 0.0297, "step": 29140 }, { "grad_norm": 0.34591588377952576, "learning_rate": 7.532921547907185e-05, "loss": 0.0321, "step": 29150 }, { "grad_norm": 0.35124680399894714, "learning_rate": 7.531139324238412e-05, "loss": 0.0429, "step": 29160 }, { "grad_norm": 0.46733999252319336, "learning_rate": 7.529356668067157e-05, "loss": 0.0435, "step": 29170 }, { "grad_norm": 0.41995298862457275, "learning_rate": 7.527573579698023e-05, "loss": 0.0323, "step": 29180 }, { "grad_norm": 0.49021679162979126, "learning_rate": 7.525790059435693e-05, "loss": 0.0299, "step": 29190 }, { "grad_norm": 0.3552551567554474, "learning_rate": 7.524006107584926e-05, "loss": 0.0368, "step": 29200 }, { "grad_norm": 0.3772048056125641, "learning_rate": 7.522221724450544e-05, "loss": 0.0312, "step": 29210 }, { "grad_norm": 0.2689831256866455, "learning_rate": 7.520436910337451e-05, "loss": 0.037, "step": 29220 }, { "grad_norm": 0.3168630599975586, "learning_rate": 7.518651665550627e-05, "loss": 0.0302, "step": 29230 }, { "grad_norm": 0.24880734086036682, "learning_rate": 7.516865990395117e-05, "loss": 0.036, "step": 29240 }, { "grad_norm": 0.39949649572372437, "learning_rate": 7.515079885176047e-05, "loss": 0.0315, "step": 29250 }, { "grad_norm": 0.42931097745895386, "learning_rate": 7.513293350198612e-05, "loss": 0.0374, "step": 29260 }, { "grad_norm": 0.2710149884223938, "learning_rate": 7.511506385768081e-05, "loss": 0.0384, "step": 29270 }, { "grad_norm": 0.3502786457538605, "learning_rate": 7.509718992189801e-05, "loss": 0.0314, "step": 29280 }, { "grad_norm": 0.41509944200515747, "learning_rate": 7.507931169769182e-05, "loss": 0.0327, "step": 29290 }, { "grad_norm": 0.4245878756046295, "learning_rate": 7.506142918811722e-05, "loss": 0.0324, "step": 29300 }, { "grad_norm": 0.3007393479347229, "learning_rate": 7.504354239622978e-05, "loss": 0.0456, "step": 29310 }, { "grad_norm": 0.4716688096523285, "learning_rate": 7.50256513250859e-05, "loss": 0.0313, "step": 29320 }, { "grad_norm": 0.5452431440353394, "learning_rate": 7.500775597774265e-05, "loss": 0.0326, "step": 29330 }, { "grad_norm": 0.34930557012557983, "learning_rate": 7.498985635725788e-05, "loss": 0.0286, "step": 29340 }, { "grad_norm": 0.5097277164459229, "learning_rate": 7.497195246669012e-05, "loss": 0.0402, "step": 29350 }, { "grad_norm": 0.3619084060192108, "learning_rate": 7.495404430909868e-05, "loss": 0.0285, "step": 29360 }, { "grad_norm": 0.46659111976623535, "learning_rate": 7.493613188754356e-05, "loss": 0.0345, "step": 29370 }, { "grad_norm": 0.33062344789505005, "learning_rate": 7.49182152050855e-05, "loss": 0.0274, "step": 29380 }, { "grad_norm": 0.27679988741874695, "learning_rate": 7.490029426478598e-05, "loss": 0.0222, "step": 29390 }, { "grad_norm": 0.31337106227874756, "learning_rate": 7.488236906970719e-05, "loss": 0.0277, "step": 29400 }, { "grad_norm": 0.4838293194770813, "learning_rate": 7.486443962291207e-05, "loss": 0.0295, "step": 29410 }, { "grad_norm": 0.3605203926563263, "learning_rate": 7.484650592746424e-05, "loss": 0.0298, "step": 29420 }, { "grad_norm": 0.31057682633399963, "learning_rate": 7.482856798642811e-05, "loss": 0.0287, "step": 29430 }, { "grad_norm": 0.25397980213165283, "learning_rate": 7.481062580286878e-05, "loss": 0.0294, "step": 29440 }, { "grad_norm": 0.3272542655467987, "learning_rate": 7.479267937985208e-05, "loss": 0.0366, "step": 29450 }, { "grad_norm": 0.4465557336807251, "learning_rate": 7.477472872044456e-05, "loss": 0.0298, "step": 29460 }, { "grad_norm": 0.4090407192707062, "learning_rate": 7.475677382771347e-05, "loss": 0.0316, "step": 29470 }, { "grad_norm": 0.4551040530204773, "learning_rate": 7.473881470472683e-05, "loss": 0.0351, "step": 29480 }, { "grad_norm": 0.4337782561779022, "learning_rate": 7.472085135455337e-05, "loss": 0.0376, "step": 29490 }, { "grad_norm": 0.3465005159378052, "learning_rate": 7.470288378026256e-05, "loss": 0.0365, "step": 29500 }, { "grad_norm": 0.3799416124820709, "learning_rate": 7.468491198492451e-05, "loss": 0.0312, "step": 29510 }, { "grad_norm": 0.4376366138458252, "learning_rate": 7.466693597161013e-05, "loss": 0.033, "step": 29520 }, { "grad_norm": 0.3673129677772522, "learning_rate": 7.464895574339104e-05, "loss": 0.0322, "step": 29530 }, { "grad_norm": 0.24106518924236298, "learning_rate": 7.463097130333958e-05, "loss": 0.0404, "step": 29540 }, { "grad_norm": 0.2901785969734192, "learning_rate": 7.461298265452876e-05, "loss": 0.0325, "step": 29550 }, { "grad_norm": 0.2673550248146057, "learning_rate": 7.459498980003239e-05, "loss": 0.0293, "step": 29560 }, { "grad_norm": 0.4938179552555084, "learning_rate": 7.457699274292493e-05, "loss": 0.0323, "step": 29570 }, { "grad_norm": 0.3058892786502838, "learning_rate": 7.455899148628159e-05, "loss": 0.0346, "step": 29580 }, { "grad_norm": 0.4616132378578186, "learning_rate": 7.45409860331783e-05, "loss": 0.0316, "step": 29590 }, { "grad_norm": 0.4862196445465088, "learning_rate": 7.452297638669169e-05, "loss": 0.0271, "step": 29600 }, { "grad_norm": 0.33795633912086487, "learning_rate": 7.450496254989911e-05, "loss": 0.0206, "step": 29610 }, { "grad_norm": 0.390600323677063, "learning_rate": 7.448694452587866e-05, "loss": 0.0261, "step": 29620 }, { "grad_norm": 0.45327913761138916, "learning_rate": 7.44689223177091e-05, "loss": 0.0335, "step": 29630 }, { "grad_norm": 0.3400671184062958, "learning_rate": 7.445089592846994e-05, "loss": 0.0275, "step": 29640 }, { "grad_norm": 0.332492858171463, "learning_rate": 7.443286536124141e-05, "loss": 0.0365, "step": 29650 }, { "grad_norm": 0.36662301421165466, "learning_rate": 7.441483061910443e-05, "loss": 0.0334, "step": 29660 }, { "grad_norm": 0.38238808512687683, "learning_rate": 7.439679170514064e-05, "loss": 0.0376, "step": 29670 }, { "grad_norm": 0.3116876184940338, "learning_rate": 7.43787486224324e-05, "loss": 0.0299, "step": 29680 }, { "grad_norm": 0.33228400349617004, "learning_rate": 7.436070137406276e-05, "loss": 0.0293, "step": 29690 }, { "grad_norm": 0.4364342987537384, "learning_rate": 7.434264996311556e-05, "loss": 0.0288, "step": 29700 }, { "grad_norm": 0.5610917210578918, "learning_rate": 7.432459439267525e-05, "loss": 0.0431, "step": 29710 }, { "grad_norm": 0.3340577483177185, "learning_rate": 7.430653466582701e-05, "loss": 0.0348, "step": 29720 }, { "grad_norm": 0.3856643736362457, "learning_rate": 7.42884707856568e-05, "loss": 0.0411, "step": 29730 }, { "grad_norm": 0.3146122097969055, "learning_rate": 7.427040275525122e-05, "loss": 0.0209, "step": 29740 }, { "grad_norm": 0.31199899315834045, "learning_rate": 7.42523305776976e-05, "loss": 0.0395, "step": 29750 }, { "grad_norm": 0.2649189233779907, "learning_rate": 7.4234254256084e-05, "loss": 0.0288, "step": 29760 }, { "grad_norm": 0.4380274713039398, "learning_rate": 7.421617379349915e-05, "loss": 0.0367, "step": 29770 }, { "grad_norm": 0.3193580210208893, "learning_rate": 7.41980891930325e-05, "loss": 0.0256, "step": 29780 }, { "grad_norm": 0.31042155623435974, "learning_rate": 7.418000045777425e-05, "loss": 0.0308, "step": 29790 }, { "grad_norm": 0.3271487355232239, "learning_rate": 7.416190759081523e-05, "loss": 0.0294, "step": 29800 }, { "grad_norm": 0.37665823101997375, "learning_rate": 7.414381059524704e-05, "loss": 0.0252, "step": 29810 }, { "grad_norm": 0.346444308757782, "learning_rate": 7.412570947416195e-05, "loss": 0.0275, "step": 29820 }, { "grad_norm": 0.6385610699653625, "learning_rate": 7.410760423065295e-05, "loss": 0.0401, "step": 29830 }, { "grad_norm": 0.5240139961242676, "learning_rate": 7.408949486781372e-05, "loss": 0.0404, "step": 29840 }, { "grad_norm": 0.5243621468544006, "learning_rate": 7.407138138873868e-05, "loss": 0.0411, "step": 29850 }, { "grad_norm": 0.43079060316085815, "learning_rate": 7.405326379652292e-05, "loss": 0.0363, "step": 29860 }, { "grad_norm": 0.30532196164131165, "learning_rate": 7.403514209426222e-05, "loss": 0.0349, "step": 29870 }, { "grad_norm": 0.39702653884887695, "learning_rate": 7.40170162850531e-05, "loss": 0.0356, "step": 29880 }, { "grad_norm": 0.25956088304519653, "learning_rate": 7.399888637199278e-05, "loss": 0.0321, "step": 29890 }, { "grad_norm": 0.5128048658370972, "learning_rate": 7.398075235817914e-05, "loss": 0.0441, "step": 29900 }, { "grad_norm": 0.28215816617012024, "learning_rate": 7.39626142467108e-05, "loss": 0.0296, "step": 29910 }, { "grad_norm": 0.4279323220252991, "learning_rate": 7.394447204068706e-05, "loss": 0.0303, "step": 29920 }, { "grad_norm": 0.6073793172836304, "learning_rate": 7.392632574320793e-05, "loss": 0.0406, "step": 29930 }, { "grad_norm": 0.3421599566936493, "learning_rate": 7.390817535737411e-05, "loss": 0.0266, "step": 29940 }, { "grad_norm": 0.3390958309173584, "learning_rate": 7.389002088628703e-05, "loss": 0.0319, "step": 29950 }, { "grad_norm": 0.3087449371814728, "learning_rate": 7.387186233304877e-05, "loss": 0.0308, "step": 29960 }, { "grad_norm": 0.3024505078792572, "learning_rate": 7.385369970076212e-05, "loss": 0.0323, "step": 29970 }, { "grad_norm": 0.36992403864860535, "learning_rate": 7.38355329925306e-05, "loss": 0.0327, "step": 29980 }, { "grad_norm": 0.3304376006126404, "learning_rate": 7.381736221145838e-05, "loss": 0.0385, "step": 29990 }, { "grad_norm": 0.4708016812801361, "learning_rate": 7.37991873606504e-05, "loss": 0.0269, "step": 30000 }, { "grad_norm": 0.2871071696281433, "learning_rate": 7.378100844321218e-05, "loss": 0.0363, "step": 30010 }, { "grad_norm": 0.3319011926651001, "learning_rate": 7.376282546225004e-05, "loss": 0.0299, "step": 30020 }, { "grad_norm": 0.3659678101539612, "learning_rate": 7.374463842087094e-05, "loss": 0.0298, "step": 30030 }, { "grad_norm": 0.3683062195777893, "learning_rate": 7.372644732218254e-05, "loss": 0.0321, "step": 30040 }, { "grad_norm": 0.39503440260887146, "learning_rate": 7.370825216929322e-05, "loss": 0.0443, "step": 30050 }, { "grad_norm": 0.34453457593917847, "learning_rate": 7.369005296531205e-05, "loss": 0.0357, "step": 30060 }, { "grad_norm": 0.3266637623310089, "learning_rate": 7.367184971334873e-05, "loss": 0.033, "step": 30070 }, { "grad_norm": 0.26972293853759766, "learning_rate": 7.365364241651371e-05, "loss": 0.0337, "step": 30080 }, { "grad_norm": 0.31353646516799927, "learning_rate": 7.363543107791815e-05, "loss": 0.0296, "step": 30090 }, { "grad_norm": 0.34145647287368774, "learning_rate": 7.361721570067384e-05, "loss": 0.0267, "step": 30100 }, { "grad_norm": 0.3500685691833496, "learning_rate": 7.359899628789331e-05, "loss": 0.0285, "step": 30110 }, { "grad_norm": 0.376833975315094, "learning_rate": 7.358077284268974e-05, "loss": 0.0309, "step": 30120 }, { "grad_norm": 0.3604468107223511, "learning_rate": 7.356254536817702e-05, "loss": 0.0301, "step": 30130 }, { "grad_norm": 0.3906501531600952, "learning_rate": 7.354431386746973e-05, "loss": 0.0247, "step": 30140 }, { "grad_norm": 0.3623389005661011, "learning_rate": 7.352607834368316e-05, "loss": 0.0277, "step": 30150 }, { "grad_norm": 0.371385395526886, "learning_rate": 7.350783879993324e-05, "loss": 0.024, "step": 30160 }, { "grad_norm": 0.36640894412994385, "learning_rate": 7.348959523933658e-05, "loss": 0.0339, "step": 30170 }, { "grad_norm": 0.4262334406375885, "learning_rate": 7.347134766501057e-05, "loss": 0.0268, "step": 30180 }, { "grad_norm": 0.3084201216697693, "learning_rate": 7.345309608007315e-05, "loss": 0.0228, "step": 30190 }, { "grad_norm": 0.31937962770462036, "learning_rate": 7.343484048764308e-05, "loss": 0.0299, "step": 30200 }, { "grad_norm": 0.30383142828941345, "learning_rate": 7.341658089083972e-05, "loss": 0.0275, "step": 30210 }, { "grad_norm": 0.23928895592689514, "learning_rate": 7.339831729278313e-05, "loss": 0.0236, "step": 30220 }, { "grad_norm": 0.32635045051574707, "learning_rate": 7.338004969659404e-05, "loss": 0.03, "step": 30230 }, { "grad_norm": 0.40453675389289856, "learning_rate": 7.336177810539391e-05, "loss": 0.0339, "step": 30240 }, { "grad_norm": 0.3286522924900055, "learning_rate": 7.334350252230485e-05, "loss": 0.025, "step": 30250 }, { "grad_norm": 0.3452153205871582, "learning_rate": 7.332522295044965e-05, "loss": 0.0222, "step": 30260 }, { "grad_norm": 0.4354587197303772, "learning_rate": 7.33069393929518e-05, "loss": 0.0305, "step": 30270 }, { "grad_norm": 0.457611083984375, "learning_rate": 7.328865185293545e-05, "loss": 0.0423, "step": 30280 }, { "grad_norm": 0.41862767934799194, "learning_rate": 7.327036033352546e-05, "loss": 0.0319, "step": 30290 }, { "grad_norm": 0.38603636622428894, "learning_rate": 7.325206483784733e-05, "loss": 0.0309, "step": 30300 }, { "grad_norm": 0.40848538279533386, "learning_rate": 7.323376536902724e-05, "loss": 0.0384, "step": 30310 }, { "grad_norm": 0.6162024140357971, "learning_rate": 7.321546193019213e-05, "loss": 0.039, "step": 30320 }, { "grad_norm": 0.33832284808158875, "learning_rate": 7.31971545244695e-05, "loss": 0.035, "step": 30330 }, { "grad_norm": 0.24075929820537567, "learning_rate": 7.31788431549876e-05, "loss": 0.0306, "step": 30340 }, { "grad_norm": 0.5129883885383606, "learning_rate": 7.316052782487534e-05, "loss": 0.0275, "step": 30350 }, { "grad_norm": 0.32282912731170654, "learning_rate": 7.314220853726234e-05, "loss": 0.0341, "step": 30360 }, { "grad_norm": 0.4785570502281189, "learning_rate": 7.312388529527884e-05, "loss": 0.0366, "step": 30370 }, { "grad_norm": 0.35992592573165894, "learning_rate": 7.310555810205577e-05, "loss": 0.0317, "step": 30380 }, { "grad_norm": 0.23306535184383392, "learning_rate": 7.308722696072476e-05, "loss": 0.0303, "step": 30390 }, { "grad_norm": 0.38208818435668945, "learning_rate": 7.306889187441811e-05, "loss": 0.0252, "step": 30400 }, { "grad_norm": 0.2571341097354889, "learning_rate": 7.305055284626876e-05, "loss": 0.024, "step": 30410 }, { "grad_norm": 0.33706769347190857, "learning_rate": 7.303220987941037e-05, "loss": 0.0321, "step": 30420 }, { "grad_norm": 0.30287542939186096, "learning_rate": 7.301386297697726e-05, "loss": 0.0366, "step": 30430 }, { "grad_norm": 0.4137742519378662, "learning_rate": 7.299551214210438e-05, "loss": 0.0337, "step": 30440 }, { "grad_norm": 0.3311412036418915, "learning_rate": 7.297715737792738e-05, "loss": 0.0293, "step": 30450 }, { "grad_norm": 0.2754884362220764, "learning_rate": 7.295879868758265e-05, "loss": 0.0342, "step": 30460 }, { "grad_norm": 0.4834209084510803, "learning_rate": 7.294043607420713e-05, "loss": 0.0349, "step": 30470 }, { "grad_norm": 0.4037656784057617, "learning_rate": 7.292206954093852e-05, "loss": 0.0278, "step": 30480 }, { "grad_norm": 0.35674723982810974, "learning_rate": 7.290369909091515e-05, "loss": 0.0258, "step": 30490 }, { "grad_norm": 0.33887380361557007, "learning_rate": 7.2885324727276e-05, "loss": 0.0306, "step": 30500 }, { "grad_norm": 0.42233309149742126, "learning_rate": 7.286694645316076e-05, "loss": 0.0304, "step": 30510 }, { "grad_norm": 0.3875638544559479, "learning_rate": 7.284856427170982e-05, "loss": 0.0265, "step": 30520 }, { "grad_norm": 0.379787802696228, "learning_rate": 7.283017818606414e-05, "loss": 0.0351, "step": 30530 }, { "grad_norm": 0.4118642508983612, "learning_rate": 7.28117881993654e-05, "loss": 0.0301, "step": 30540 }, { "grad_norm": 0.24748414754867554, "learning_rate": 7.279339431475598e-05, "loss": 0.0233, "step": 30550 }, { "grad_norm": 0.23438851535320282, "learning_rate": 7.277499653537887e-05, "loss": 0.0211, "step": 30560 }, { "grad_norm": 0.28847789764404297, "learning_rate": 7.275659486437776e-05, "loss": 0.0265, "step": 30570 }, { "grad_norm": 0.2596013844013214, "learning_rate": 7.273818930489695e-05, "loss": 0.0339, "step": 30580 }, { "grad_norm": 0.4106093645095825, "learning_rate": 7.271977986008151e-05, "loss": 0.0256, "step": 30590 }, { "grad_norm": 0.3361077606678009, "learning_rate": 7.270136653307705e-05, "loss": 0.0249, "step": 30600 }, { "grad_norm": 0.2864668071269989, "learning_rate": 7.268294932702994e-05, "loss": 0.0258, "step": 30610 }, { "grad_norm": 0.3146439492702484, "learning_rate": 7.266452824508719e-05, "loss": 0.0315, "step": 30620 }, { "grad_norm": 0.30724456906318665, "learning_rate": 7.264610329039643e-05, "loss": 0.0294, "step": 30630 }, { "grad_norm": 0.3568369448184967, "learning_rate": 7.262767446610599e-05, "loss": 0.0257, "step": 30640 }, { "grad_norm": 0.39783090353012085, "learning_rate": 7.260924177536485e-05, "loss": 0.0318, "step": 30650 }, { "grad_norm": 0.25842615962028503, "learning_rate": 7.259080522132265e-05, "loss": 0.0285, "step": 30660 }, { "grad_norm": 0.3614245653152466, "learning_rate": 7.257236480712972e-05, "loss": 0.0317, "step": 30670 }, { "grad_norm": 0.39629998803138733, "learning_rate": 7.255392053593697e-05, "loss": 0.0237, "step": 30680 }, { "grad_norm": 0.3322148621082306, "learning_rate": 7.253547241089607e-05, "loss": 0.0277, "step": 30690 }, { "grad_norm": 0.6198861002922058, "learning_rate": 7.251702043515927e-05, "loss": 0.0261, "step": 30700 }, { "grad_norm": 0.3226798176765442, "learning_rate": 7.249856461187952e-05, "loss": 0.0272, "step": 30710 }, { "grad_norm": 0.32021021842956543, "learning_rate": 7.248010494421042e-05, "loss": 0.024, "step": 30720 }, { "grad_norm": 0.2891733646392822, "learning_rate": 7.246164143530622e-05, "loss": 0.0353, "step": 30730 }, { "grad_norm": 0.32598960399627686, "learning_rate": 7.244317408832181e-05, "loss": 0.028, "step": 30740 }, { "grad_norm": 0.32066404819488525, "learning_rate": 7.242470290641279e-05, "loss": 0.0271, "step": 30750 }, { "grad_norm": 0.2717796266078949, "learning_rate": 7.240622789273536e-05, "loss": 0.0256, "step": 30760 }, { "grad_norm": 0.3351847231388092, "learning_rate": 7.238774905044638e-05, "loss": 0.0277, "step": 30770 }, { "grad_norm": 0.2911614179611206, "learning_rate": 7.236926638270341e-05, "loss": 0.0288, "step": 30780 }, { "grad_norm": 0.3821761906147003, "learning_rate": 7.23507798926646e-05, "loss": 0.0321, "step": 30790 }, { "grad_norm": 0.26361826062202454, "learning_rate": 7.23322895834888e-05, "loss": 0.0314, "step": 30800 }, { "grad_norm": 0.46473702788352966, "learning_rate": 7.231379545833552e-05, "loss": 0.0328, "step": 30810 }, { "grad_norm": 0.26650187373161316, "learning_rate": 7.229529752036487e-05, "loss": 0.0304, "step": 30820 }, { "grad_norm": 0.27420395612716675, "learning_rate": 7.227679577273765e-05, "loss": 0.0232, "step": 30830 }, { "grad_norm": 0.4410402476787567, "learning_rate": 7.225829021861529e-05, "loss": 0.0333, "step": 30840 }, { "grad_norm": 0.23063136637210846, "learning_rate": 7.223978086115992e-05, "loss": 0.0229, "step": 30850 }, { "grad_norm": 0.2639407515525818, "learning_rate": 7.222126770353425e-05, "loss": 0.0368, "step": 30860 }, { "grad_norm": 0.24900631606578827, "learning_rate": 7.22027507489017e-05, "loss": 0.0247, "step": 30870 }, { "grad_norm": 0.286692351102829, "learning_rate": 7.218423000042627e-05, "loss": 0.029, "step": 30880 }, { "grad_norm": 0.3917371928691864, "learning_rate": 7.216570546127268e-05, "loss": 0.0267, "step": 30890 }, { "grad_norm": 0.3221350610256195, "learning_rate": 7.214717713460626e-05, "loss": 0.0295, "step": 30900 }, { "grad_norm": 0.3250807225704193, "learning_rate": 7.2128645023593e-05, "loss": 0.0277, "step": 30910 }, { "grad_norm": 0.2594098746776581, "learning_rate": 7.211010913139951e-05, "loss": 0.0319, "step": 30920 }, { "grad_norm": 0.3176972568035126, "learning_rate": 7.209156946119308e-05, "loss": 0.0251, "step": 30930 }, { "grad_norm": 0.2644885182380676, "learning_rate": 7.207302601614166e-05, "loss": 0.0228, "step": 30940 }, { "grad_norm": 0.2977452576160431, "learning_rate": 7.205447879941378e-05, "loss": 0.0292, "step": 30950 }, { "grad_norm": 0.3670976161956787, "learning_rate": 7.203592781417866e-05, "loss": 0.0276, "step": 30960 }, { "grad_norm": 0.4472581148147583, "learning_rate": 7.201737306360617e-05, "loss": 0.0274, "step": 30970 }, { "grad_norm": 0.47185343503952026, "learning_rate": 7.19988145508668e-05, "loss": 0.024, "step": 30980 }, { "grad_norm": 0.36909887194633484, "learning_rate": 7.198025227913168e-05, "loss": 0.0279, "step": 30990 }, { "grad_norm": 0.36288025975227356, "learning_rate": 7.196168625157261e-05, "loss": 0.0309, "step": 31000 }, { "grad_norm": 0.30083754658699036, "learning_rate": 7.194311647136201e-05, "loss": 0.0268, "step": 31010 }, { "grad_norm": 0.4153949022293091, "learning_rate": 7.192454294167297e-05, "loss": 0.0224, "step": 31020 }, { "grad_norm": 0.4247564375400543, "learning_rate": 7.190596566567917e-05, "loss": 0.0312, "step": 31030 }, { "grad_norm": 0.4720446467399597, "learning_rate": 7.188738464655496e-05, "loss": 0.0358, "step": 31040 }, { "grad_norm": 0.4404860734939575, "learning_rate": 7.186879988747533e-05, "loss": 0.0286, "step": 31050 }, { "grad_norm": 0.300152987241745, "learning_rate": 7.185021139161592e-05, "loss": 0.0265, "step": 31060 }, { "grad_norm": 0.26521027088165283, "learning_rate": 7.1831619162153e-05, "loss": 0.0278, "step": 31070 }, { "grad_norm": 0.44545871019363403, "learning_rate": 7.181302320226345e-05, "loss": 0.0289, "step": 31080 }, { "grad_norm": 0.31459325551986694, "learning_rate": 7.179442351512482e-05, "loss": 0.0364, "step": 31090 }, { "grad_norm": 0.5760052800178528, "learning_rate": 7.177582010391528e-05, "loss": 0.0426, "step": 31100 }, { "grad_norm": 0.393038272857666, "learning_rate": 7.175721297181366e-05, "loss": 0.0416, "step": 31110 }, { "grad_norm": 0.39407968521118164, "learning_rate": 7.173860212199942e-05, "loss": 0.0304, "step": 31120 }, { "grad_norm": 0.43531662225723267, "learning_rate": 7.171998755765263e-05, "loss": 0.0309, "step": 31130 }, { "grad_norm": 0.4204598069190979, "learning_rate": 7.170136928195398e-05, "loss": 0.0268, "step": 31140 }, { "grad_norm": 0.29433974623680115, "learning_rate": 7.168274729808489e-05, "loss": 0.0227, "step": 31150 }, { "grad_norm": 0.32340574264526367, "learning_rate": 7.166412160922728e-05, "loss": 0.0288, "step": 31160 }, { "grad_norm": 0.4207293689250946, "learning_rate": 7.164549221856382e-05, "loss": 0.0278, "step": 31170 }, { "grad_norm": 0.2708541750907898, "learning_rate": 7.162685912927775e-05, "loss": 0.0411, "step": 31180 }, { "grad_norm": 0.20716775953769684, "learning_rate": 7.160822234455294e-05, "loss": 0.0233, "step": 31190 }, { "grad_norm": 0.2506794333457947, "learning_rate": 7.158958186757391e-05, "loss": 0.026, "step": 31200 }, { "grad_norm": 0.4375135898590088, "learning_rate": 7.157093770152582e-05, "loss": 0.0325, "step": 31210 }, { "grad_norm": 0.3974294066429138, "learning_rate": 7.155228984959446e-05, "loss": 0.0294, "step": 31220 }, { "grad_norm": 0.32059141993522644, "learning_rate": 7.153363831496621e-05, "loss": 0.0327, "step": 31230 }, { "grad_norm": 0.42228075861930847, "learning_rate": 7.151498310082811e-05, "loss": 0.0357, "step": 31240 }, { "grad_norm": 0.39025193452835083, "learning_rate": 7.149632421036784e-05, "loss": 0.0292, "step": 31250 }, { "grad_norm": 0.45825305581092834, "learning_rate": 7.147766164677369e-05, "loss": 0.0472, "step": 31260 }, { "grad_norm": 0.4456443190574646, "learning_rate": 7.145899541323459e-05, "loss": 0.0296, "step": 31270 }, { "grad_norm": 0.39988282322883606, "learning_rate": 7.144032551294007e-05, "loss": 0.0376, "step": 31280 }, { "grad_norm": 0.42043060064315796, "learning_rate": 7.14216519490803e-05, "loss": 0.0362, "step": 31290 }, { "grad_norm": 0.40857580304145813, "learning_rate": 7.140297472484609e-05, "loss": 0.0409, "step": 31300 }, { "grad_norm": 0.23111924529075623, "learning_rate": 7.138429384342891e-05, "loss": 0.0418, "step": 31310 }, { "grad_norm": 0.218017578125, "learning_rate": 7.136560930802074e-05, "loss": 0.0249, "step": 31320 }, { "grad_norm": 0.23773741722106934, "learning_rate": 7.134692112181431e-05, "loss": 0.0262, "step": 31330 }, { "grad_norm": 0.318055659532547, "learning_rate": 7.13282292880029e-05, "loss": 0.0314, "step": 31340 }, { "grad_norm": 0.33812883496284485, "learning_rate": 7.130953380978043e-05, "loss": 0.0317, "step": 31350 }, { "grad_norm": 0.4366829991340637, "learning_rate": 7.129083469034144e-05, "loss": 0.0406, "step": 31360 }, { "grad_norm": 0.37739259004592896, "learning_rate": 7.127213193288112e-05, "loss": 0.0267, "step": 31370 }, { "grad_norm": 0.2807546555995941, "learning_rate": 7.125342554059522e-05, "loss": 0.0266, "step": 31380 }, { "grad_norm": 0.2968219518661499, "learning_rate": 7.12347155166802e-05, "loss": 0.0327, "step": 31390 }, { "grad_norm": 0.24586232006549835, "learning_rate": 7.121600186433306e-05, "loss": 0.0244, "step": 31400 }, { "grad_norm": 0.2624934911727905, "learning_rate": 7.119728458675148e-05, "loss": 0.0339, "step": 31410 }, { "grad_norm": 0.42316779494285583, "learning_rate": 7.117856368713369e-05, "loss": 0.032, "step": 31420 }, { "grad_norm": 0.4584256112575531, "learning_rate": 7.115983916867861e-05, "loss": 0.035, "step": 31430 }, { "grad_norm": 0.20785655081272125, "learning_rate": 7.114111103458574e-05, "loss": 0.0266, "step": 31440 }, { "grad_norm": 0.2943979799747467, "learning_rate": 7.11223792880552e-05, "loss": 0.0295, "step": 31450 }, { "grad_norm": 0.5587091445922852, "learning_rate": 7.110364393228773e-05, "loss": 0.0349, "step": 31460 }, { "grad_norm": 0.3698756992816925, "learning_rate": 7.108490497048471e-05, "loss": 0.0291, "step": 31470 }, { "grad_norm": 0.4063109755516052, "learning_rate": 7.10661624058481e-05, "loss": 0.0324, "step": 31480 }, { "grad_norm": 0.23724673688411713, "learning_rate": 7.10474162415805e-05, "loss": 0.0307, "step": 31490 }, { "grad_norm": 0.3191184401512146, "learning_rate": 7.102866648088511e-05, "loss": 0.0304, "step": 31500 }, { "grad_norm": 0.42428404092788696, "learning_rate": 7.100991312696576e-05, "loss": 0.0296, "step": 31510 }, { "grad_norm": 0.309794157743454, "learning_rate": 7.099115618302686e-05, "loss": 0.0349, "step": 31520 }, { "grad_norm": 0.3602893352508545, "learning_rate": 7.097239565227349e-05, "loss": 0.0372, "step": 31530 }, { "grad_norm": 0.3926042914390564, "learning_rate": 7.09536315379113e-05, "loss": 0.0322, "step": 31540 }, { "grad_norm": 0.4349432587623596, "learning_rate": 7.093486384314656e-05, "loss": 0.0299, "step": 31550 }, { "grad_norm": 0.4392605721950531, "learning_rate": 7.091609257118616e-05, "loss": 0.0297, "step": 31560 }, { "grad_norm": 0.4363221824169159, "learning_rate": 7.08973177252376e-05, "loss": 0.0285, "step": 31570 }, { "grad_norm": 0.34154343605041504, "learning_rate": 7.087853930850898e-05, "loss": 0.0268, "step": 31580 }, { "grad_norm": 0.30731403827667236, "learning_rate": 7.085975732420903e-05, "loss": 0.0268, "step": 31590 }, { "grad_norm": 0.2707277834415436, "learning_rate": 7.084097177554706e-05, "loss": 0.0212, "step": 31600 }, { "grad_norm": 0.30026739835739136, "learning_rate": 7.082218266573301e-05, "loss": 0.034, "step": 31610 }, { "grad_norm": 0.4066373407840729, "learning_rate": 7.080338999797743e-05, "loss": 0.0299, "step": 31620 }, { "grad_norm": 0.28668293356895447, "learning_rate": 7.07845937754915e-05, "loss": 0.034, "step": 31630 }, { "grad_norm": 0.5932155847549438, "learning_rate": 7.076579400148693e-05, "loss": 0.0399, "step": 31640 }, { "grad_norm": 0.33142101764678955, "learning_rate": 7.074699067917611e-05, "loss": 0.0348, "step": 31650 }, { "grad_norm": 0.3307495713233948, "learning_rate": 7.072818381177201e-05, "loss": 0.0297, "step": 31660 }, { "grad_norm": 0.3374773859977722, "learning_rate": 7.070937340248823e-05, "loss": 0.0274, "step": 31670 }, { "grad_norm": 0.34295475482940674, "learning_rate": 7.069055945453893e-05, "loss": 0.0259, "step": 31680 }, { "grad_norm": 0.46361038088798523, "learning_rate": 7.067174197113892e-05, "loss": 0.0273, "step": 31690 }, { "grad_norm": 0.348393976688385, "learning_rate": 7.065292095550355e-05, "loss": 0.0325, "step": 31700 }, { "grad_norm": 0.4247334599494934, "learning_rate": 7.063409641084887e-05, "loss": 0.0311, "step": 31710 }, { "grad_norm": 0.3235889673233032, "learning_rate": 7.061526834039145e-05, "loss": 0.0431, "step": 31720 }, { "grad_norm": 0.40433454513549805, "learning_rate": 7.05964367473485e-05, "loss": 0.0266, "step": 31730 }, { "grad_norm": 0.3010402023792267, "learning_rate": 7.057760163493783e-05, "loss": 0.0333, "step": 31740 }, { "grad_norm": 0.45132020115852356, "learning_rate": 7.055876300637783e-05, "loss": 0.0316, "step": 31750 }, { "grad_norm": 0.262014776468277, "learning_rate": 7.053992086488753e-05, "loss": 0.0301, "step": 31760 }, { "grad_norm": 0.3395308554172516, "learning_rate": 7.052107521368651e-05, "loss": 0.0374, "step": 31770 }, { "grad_norm": 0.36044880747795105, "learning_rate": 7.0502226055995e-05, "loss": 0.0354, "step": 31780 }, { "grad_norm": 0.31073328852653503, "learning_rate": 7.048337339503379e-05, "loss": 0.0402, "step": 31790 }, { "grad_norm": 0.42438429594039917, "learning_rate": 7.046451723402427e-05, "loss": 0.0271, "step": 31800 }, { "grad_norm": 0.3469749689102173, "learning_rate": 7.044565757618848e-05, "loss": 0.0401, "step": 31810 }, { "grad_norm": 0.37154167890548706, "learning_rate": 7.042679442474899e-05, "loss": 0.027, "step": 31820 }, { "grad_norm": 0.2681135833263397, "learning_rate": 7.040792778292902e-05, "loss": 0.037, "step": 31830 }, { "grad_norm": 0.3481564521789551, "learning_rate": 7.038905765395234e-05, "loss": 0.0346, "step": 31840 }, { "grad_norm": 0.4813290536403656, "learning_rate": 7.037018404104334e-05, "loss": 0.0364, "step": 31850 }, { "grad_norm": 0.2937580645084381, "learning_rate": 7.035130694742702e-05, "loss": 0.0272, "step": 31860 }, { "grad_norm": 0.3094002604484558, "learning_rate": 7.033242637632897e-05, "loss": 0.0319, "step": 31870 }, { "grad_norm": 0.36228716373443604, "learning_rate": 7.031354233097534e-05, "loss": 0.0274, "step": 31880 }, { "grad_norm": 0.28371238708496094, "learning_rate": 7.029465481459289e-05, "loss": 0.0385, "step": 31890 }, { "grad_norm": 0.26452478766441345, "learning_rate": 7.027576383040898e-05, "loss": 0.0247, "step": 31900 }, { "grad_norm": 0.3008354902267456, "learning_rate": 7.025686938165159e-05, "loss": 0.032, "step": 31910 }, { "grad_norm": 0.29340997338294983, "learning_rate": 7.023797147154924e-05, "loss": 0.0342, "step": 31920 }, { "grad_norm": 0.41567912697792053, "learning_rate": 7.021907010333111e-05, "loss": 0.0256, "step": 31930 }, { "grad_norm": 0.30256617069244385, "learning_rate": 7.020016528022685e-05, "loss": 0.0226, "step": 31940 }, { "grad_norm": 0.3616390526294708, "learning_rate": 7.018125700546683e-05, "loss": 0.0314, "step": 31950 }, { "grad_norm": 0.31257471442222595, "learning_rate": 7.016234528228196e-05, "loss": 0.0335, "step": 31960 }, { "grad_norm": 0.4628230631351471, "learning_rate": 7.014343011390372e-05, "loss": 0.0288, "step": 31970 }, { "grad_norm": 0.29614385962486267, "learning_rate": 7.01245115035642e-05, "loss": 0.0193, "step": 31980 }, { "grad_norm": 0.32385966181755066, "learning_rate": 7.010558945449606e-05, "loss": 0.0376, "step": 31990 }, { "grad_norm": 0.2967683970928192, "learning_rate": 7.008666396993258e-05, "loss": 0.0237, "step": 32000 }, { "grad_norm": 0.40924200415611267, "learning_rate": 7.006773505310759e-05, "loss": 0.0328, "step": 32010 }, { "grad_norm": 0.3366188406944275, "learning_rate": 7.004880270725553e-05, "loss": 0.0285, "step": 32020 }, { "grad_norm": 0.29426491260528564, "learning_rate": 7.002986693561144e-05, "loss": 0.0266, "step": 32030 }, { "grad_norm": 0.37212514877319336, "learning_rate": 7.001092774141089e-05, "loss": 0.0309, "step": 32040 }, { "grad_norm": 0.29768988490104675, "learning_rate": 6.999198512789009e-05, "loss": 0.0259, "step": 32050 }, { "grad_norm": 0.38512328267097473, "learning_rate": 6.997303909828584e-05, "loss": 0.0266, "step": 32060 }, { "grad_norm": 0.4648575186729431, "learning_rate": 6.995408965583544e-05, "loss": 0.027, "step": 32070 }, { "grad_norm": 0.3555191159248352, "learning_rate": 6.993513680377688e-05, "loss": 0.0214, "step": 32080 }, { "grad_norm": 0.2795117497444153, "learning_rate": 6.991618054534868e-05, "loss": 0.0352, "step": 32090 }, { "grad_norm": 0.5576463341712952, "learning_rate": 6.989722088378991e-05, "loss": 0.023, "step": 32100 }, { "grad_norm": 0.7857797145843506, "learning_rate": 6.987825782234027e-05, "loss": 0.034, "step": 32110 }, { "grad_norm": 0.25821253657341003, "learning_rate": 6.985929136424006e-05, "loss": 0.0232, "step": 32120 }, { "grad_norm": 0.32299792766571045, "learning_rate": 6.984032151273012e-05, "loss": 0.0257, "step": 32130 }, { "grad_norm": 0.524107813835144, "learning_rate": 6.982134827105186e-05, "loss": 0.0317, "step": 32140 }, { "grad_norm": 0.3545394837856293, "learning_rate": 6.980237164244729e-05, "loss": 0.0262, "step": 32150 }, { "grad_norm": 0.2973231375217438, "learning_rate": 6.9783391630159e-05, "loss": 0.0249, "step": 32160 }, { "grad_norm": 0.3423917591571808, "learning_rate": 6.976440823743015e-05, "loss": 0.0286, "step": 32170 }, { "grad_norm": 0.28604656457901, "learning_rate": 6.974542146750451e-05, "loss": 0.0232, "step": 32180 }, { "grad_norm": 0.2726486921310425, "learning_rate": 6.972643132362637e-05, "loss": 0.0211, "step": 32190 }, { "grad_norm": 0.277508407831192, "learning_rate": 6.970743780904064e-05, "loss": 0.024, "step": 32200 }, { "grad_norm": 0.34361732006073, "learning_rate": 6.968844092699277e-05, "loss": 0.025, "step": 32210 }, { "grad_norm": 0.33927613496780396, "learning_rate": 6.966944068072883e-05, "loss": 0.0226, "step": 32220 }, { "grad_norm": 0.3504614531993866, "learning_rate": 6.965043707349545e-05, "loss": 0.0219, "step": 32230 }, { "grad_norm": 0.5828375816345215, "learning_rate": 6.963143010853982e-05, "loss": 0.0324, "step": 32240 }, { "grad_norm": 0.519805908203125, "learning_rate": 6.961241978910971e-05, "loss": 0.0322, "step": 32250 }, { "grad_norm": 0.5540837645530701, "learning_rate": 6.959340611845344e-05, "loss": 0.0259, "step": 32260 }, { "grad_norm": 0.3799452483654022, "learning_rate": 6.957438909981995e-05, "loss": 0.0267, "step": 32270 }, { "grad_norm": 0.3163715600967407, "learning_rate": 6.955536873645872e-05, "loss": 0.0222, "step": 32280 }, { "grad_norm": 0.3734000027179718, "learning_rate": 6.953634503161982e-05, "loss": 0.0314, "step": 32290 }, { "grad_norm": 0.33663660287857056, "learning_rate": 6.951731798855387e-05, "loss": 0.0287, "step": 32300 }, { "grad_norm": 0.3577512800693512, "learning_rate": 6.949828761051208e-05, "loss": 0.0353, "step": 32310 }, { "grad_norm": 0.32623350620269775, "learning_rate": 6.947925390074622e-05, "loss": 0.0214, "step": 32320 }, { "grad_norm": 0.2585921287536621, "learning_rate": 6.946021686250863e-05, "loss": 0.0244, "step": 32330 }, { "grad_norm": 0.24636103212833405, "learning_rate": 6.94411764990522e-05, "loss": 0.0246, "step": 32340 }, { "grad_norm": 0.29042863845825195, "learning_rate": 6.942213281363044e-05, "loss": 0.0215, "step": 32350 }, { "grad_norm": 0.40098828077316284, "learning_rate": 6.940308580949737e-05, "loss": 0.0275, "step": 32360 }, { "grad_norm": 0.2352500855922699, "learning_rate": 6.93840354899076e-05, "loss": 0.0244, "step": 32370 }, { "grad_norm": 0.32061442732810974, "learning_rate": 6.936498185811633e-05, "loss": 0.0287, "step": 32380 }, { "grad_norm": 0.2618753910064697, "learning_rate": 6.93459249173793e-05, "loss": 0.0227, "step": 32390 }, { "grad_norm": 0.5095513463020325, "learning_rate": 6.932686467095279e-05, "loss": 0.0307, "step": 32400 }, { "grad_norm": 0.3635343611240387, "learning_rate": 6.930780112209373e-05, "loss": 0.0266, "step": 32410 }, { "grad_norm": 0.36793842911720276, "learning_rate": 6.92887342740595e-05, "loss": 0.0303, "step": 32420 }, { "grad_norm": 0.28242865204811096, "learning_rate": 6.926966413010816e-05, "loss": 0.0321, "step": 32430 }, { "grad_norm": 0.5336448550224304, "learning_rate": 6.925059069349824e-05, "loss": 0.0215, "step": 32440 }, { "grad_norm": 0.6997686624526978, "learning_rate": 6.923151396748886e-05, "loss": 0.0287, "step": 32450 }, { "grad_norm": 0.38534507155418396, "learning_rate": 6.921243395533974e-05, "loss": 0.038, "step": 32460 }, { "grad_norm": 0.5158618688583374, "learning_rate": 6.919335066031109e-05, "loss": 0.0445, "step": 32470 }, { "grad_norm": 0.4600653350353241, "learning_rate": 6.917426408566379e-05, "loss": 0.0422, "step": 32480 }, { "grad_norm": 0.5298210382461548, "learning_rate": 6.915517423465916e-05, "loss": 0.0295, "step": 32490 }, { "grad_norm": 0.3117005228996277, "learning_rate": 6.913608111055914e-05, "loss": 0.0302, "step": 32500 }, { "grad_norm": 0.22482267022132874, "learning_rate": 6.911698471662623e-05, "loss": 0.0242, "step": 32510 }, { "grad_norm": 0.3051961064338684, "learning_rate": 6.90978850561235e-05, "loss": 0.0245, "step": 32520 }, { "grad_norm": 0.2415563315153122, "learning_rate": 6.907878213231454e-05, "loss": 0.0253, "step": 32530 }, { "grad_norm": 0.39881643652915955, "learning_rate": 6.90596759484635e-05, "loss": 0.0289, "step": 32540 }, { "grad_norm": 0.36785247921943665, "learning_rate": 6.904056650783514e-05, "loss": 0.0291, "step": 32550 }, { "grad_norm": 0.40186893939971924, "learning_rate": 6.902145381369471e-05, "loss": 0.0317, "step": 32560 }, { "grad_norm": 0.4168166518211365, "learning_rate": 6.900233786930808e-05, "loss": 0.0301, "step": 32570 }, { "grad_norm": 0.3786267340183258, "learning_rate": 6.898321867794161e-05, "loss": 0.0315, "step": 32580 }, { "grad_norm": 0.18431231379508972, "learning_rate": 6.896409624286226e-05, "loss": 0.0233, "step": 32590 }, { "grad_norm": 0.3584957420825958, "learning_rate": 6.894497056733754e-05, "loss": 0.0323, "step": 32600 }, { "grad_norm": 0.3223895728588104, "learning_rate": 6.89258416546355e-05, "loss": 0.025, "step": 32610 }, { "grad_norm": 0.38158345222473145, "learning_rate": 6.890670950802474e-05, "loss": 0.0288, "step": 32620 }, { "grad_norm": 0.2945604920387268, "learning_rate": 6.88875741307744e-05, "loss": 0.0304, "step": 32630 }, { "grad_norm": 0.37092265486717224, "learning_rate": 6.886843552615425e-05, "loss": 0.0283, "step": 32640 }, { "grad_norm": 0.4076721966266632, "learning_rate": 6.884929369743451e-05, "loss": 0.0329, "step": 32650 }, { "grad_norm": 0.4311048090457916, "learning_rate": 6.8830148647886e-05, "loss": 0.0351, "step": 32660 }, { "grad_norm": 0.2928626239299774, "learning_rate": 6.88110003807801e-05, "loss": 0.0362, "step": 32670 }, { "grad_norm": 0.3563023507595062, "learning_rate": 6.87918488993887e-05, "loss": 0.0265, "step": 32680 }, { "grad_norm": 0.5087290406227112, "learning_rate": 6.877269420698431e-05, "loss": 0.0327, "step": 32690 }, { "grad_norm": 0.505003035068512, "learning_rate": 6.875353630683989e-05, "loss": 0.0295, "step": 32700 }, { "grad_norm": 0.5381101965904236, "learning_rate": 6.873437520222905e-05, "loss": 0.0342, "step": 32710 }, { "grad_norm": 0.33456453680992126, "learning_rate": 6.871521089642585e-05, "loss": 0.0308, "step": 32720 }, { "grad_norm": 0.523762583732605, "learning_rate": 6.869604339270498e-05, "loss": 0.0343, "step": 32730 }, { "grad_norm": 0.3382290005683899, "learning_rate": 6.867687269434164e-05, "loss": 0.0355, "step": 32740 }, { "grad_norm": 0.3317751884460449, "learning_rate": 6.865769880461156e-05, "loss": 0.0374, "step": 32750 }, { "grad_norm": 0.36020076274871826, "learning_rate": 6.863852172679104e-05, "loss": 0.03, "step": 32760 }, { "grad_norm": 0.4261024594306946, "learning_rate": 6.861934146415693e-05, "loss": 0.0308, "step": 32770 }, { "grad_norm": 0.330814927816391, "learning_rate": 6.86001580199866e-05, "loss": 0.0312, "step": 32780 }, { "grad_norm": 0.33678728342056274, "learning_rate": 6.858097139755798e-05, "loss": 0.0257, "step": 32790 }, { "grad_norm": 0.40743711590766907, "learning_rate": 6.856178160014955e-05, "loss": 0.03, "step": 32800 }, { "grad_norm": 0.3323182165622711, "learning_rate": 6.85425886310403e-05, "loss": 0.0281, "step": 32810 }, { "grad_norm": 0.4034193456172943, "learning_rate": 6.852339249350979e-05, "loss": 0.0359, "step": 32820 }, { "grad_norm": 0.47394266724586487, "learning_rate": 6.850419319083812e-05, "loss": 0.0369, "step": 32830 }, { "grad_norm": 0.2731002867221832, "learning_rate": 6.848499072630592e-05, "loss": 0.0197, "step": 32840 }, { "grad_norm": 0.3475596308708191, "learning_rate": 6.846578510319439e-05, "loss": 0.0358, "step": 32850 }, { "grad_norm": 0.3681623935699463, "learning_rate": 6.844657632478519e-05, "loss": 0.0346, "step": 32860 }, { "grad_norm": 0.3903600871562958, "learning_rate": 6.842736439436063e-05, "loss": 0.0244, "step": 32870 }, { "grad_norm": 0.32069358229637146, "learning_rate": 6.84081493152035e-05, "loss": 0.0256, "step": 32880 }, { "grad_norm": 0.35484394431114197, "learning_rate": 6.83889310905971e-05, "loss": 0.0255, "step": 32890 }, { "grad_norm": 0.3703971803188324, "learning_rate": 6.836970972382533e-05, "loss": 0.0365, "step": 32900 }, { "grad_norm": 0.3203161358833313, "learning_rate": 6.835048521817257e-05, "loss": 0.0363, "step": 32910 }, { "grad_norm": 0.52857905626297, "learning_rate": 6.833125757692379e-05, "loss": 0.0284, "step": 32920 }, { "grad_norm": 0.5092107057571411, "learning_rate": 6.831202680336441e-05, "loss": 0.0272, "step": 32930 }, { "grad_norm": 0.25892820954322815, "learning_rate": 6.829279290078052e-05, "loss": 0.0276, "step": 32940 }, { "grad_norm": 0.29782500863075256, "learning_rate": 6.827355587245863e-05, "loss": 0.0294, "step": 32950 }, { "grad_norm": 0.28832170367240906, "learning_rate": 6.82543157216858e-05, "loss": 0.022, "step": 32960 }, { "grad_norm": 0.33897095918655396, "learning_rate": 6.823507245174969e-05, "loss": 0.0218, "step": 32970 }, { "grad_norm": 0.2935929000377655, "learning_rate": 6.821582606593841e-05, "loss": 0.0316, "step": 32980 }, { "grad_norm": 0.33042824268341064, "learning_rate": 6.81965765675407e-05, "loss": 0.0337, "step": 32990 }, { "grad_norm": 0.3937261998653412, "learning_rate": 6.81773239598457e-05, "loss": 0.0348, "step": 33000 }, { "grad_norm": 0.41565266251564026, "learning_rate": 6.815806824614319e-05, "loss": 0.034, "step": 33010 }, { "grad_norm": 0.2557268440723419, "learning_rate": 6.813880942972343e-05, "loss": 0.0221, "step": 33020 }, { "grad_norm": 0.34162044525146484, "learning_rate": 6.811954751387726e-05, "loss": 0.0285, "step": 33030 }, { "grad_norm": 0.2116652876138687, "learning_rate": 6.810028250189598e-05, "loss": 0.0189, "step": 33040 }, { "grad_norm": 0.42692556977272034, "learning_rate": 6.808101439707147e-05, "loss": 0.043, "step": 33050 }, { "grad_norm": 0.35454729199409485, "learning_rate": 6.806174320269609e-05, "loss": 0.0344, "step": 33060 }, { "grad_norm": 0.4415973722934723, "learning_rate": 6.804246892206281e-05, "loss": 0.036, "step": 33070 }, { "grad_norm": 0.2787016034126282, "learning_rate": 6.802319155846506e-05, "loss": 0.0404, "step": 33080 }, { "grad_norm": 0.35142242908477783, "learning_rate": 6.800391111519679e-05, "loss": 0.0298, "step": 33090 }, { "grad_norm": 0.39997467398643494, "learning_rate": 6.798462759555253e-05, "loss": 0.0333, "step": 33100 }, { "grad_norm": 0.35668912529945374, "learning_rate": 6.79653410028273e-05, "loss": 0.0272, "step": 33110 }, { "grad_norm": 0.34604978561401367, "learning_rate": 6.794605134031663e-05, "loss": 0.0336, "step": 33120 }, { "grad_norm": 0.30439746379852295, "learning_rate": 6.792675861131661e-05, "loss": 0.03, "step": 33130 }, { "grad_norm": 0.30598214268684387, "learning_rate": 6.790746281912386e-05, "loss": 0.0314, "step": 33140 }, { "grad_norm": 0.3466852307319641, "learning_rate": 6.788816396703546e-05, "loss": 0.0245, "step": 33150 }, { "grad_norm": 0.35027194023132324, "learning_rate": 6.78688620583491e-05, "loss": 0.0298, "step": 33160 }, { "grad_norm": 0.37700653076171875, "learning_rate": 6.784955709636292e-05, "loss": 0.0354, "step": 33170 }, { "grad_norm": 0.4515920579433441, "learning_rate": 6.783024908437564e-05, "loss": 0.0338, "step": 33180 }, { "grad_norm": 0.2833057641983032, "learning_rate": 6.781093802568641e-05, "loss": 0.0221, "step": 33190 }, { "grad_norm": 0.3409282863140106, "learning_rate": 6.779162392359504e-05, "loss": 0.0294, "step": 33200 }, { "grad_norm": 0.43333950638771057, "learning_rate": 6.777230678140172e-05, "loss": 0.0313, "step": 33210 }, { "grad_norm": 0.3062725365161896, "learning_rate": 6.775298660240726e-05, "loss": 0.03, "step": 33220 }, { "grad_norm": 0.27587461471557617, "learning_rate": 6.773366338991292e-05, "loss": 0.0393, "step": 33230 }, { "grad_norm": 0.4784789979457855, "learning_rate": 6.771433714722052e-05, "loss": 0.0274, "step": 33240 }, { "grad_norm": 0.4121173918247223, "learning_rate": 6.769500787763239e-05, "loss": 0.0325, "step": 33250 }, { "grad_norm": 0.3632875978946686, "learning_rate": 6.76756755844514e-05, "loss": 0.0308, "step": 33260 }, { "grad_norm": 0.24823518097400665, "learning_rate": 6.765634027098087e-05, "loss": 0.0271, "step": 33270 }, { "grad_norm": 0.3428182899951935, "learning_rate": 6.763700194052468e-05, "loss": 0.0291, "step": 33280 }, { "grad_norm": 0.43881556391716003, "learning_rate": 6.761766059638723e-05, "loss": 0.0245, "step": 33290 }, { "grad_norm": 0.38298994302749634, "learning_rate": 6.759831624187345e-05, "loss": 0.0249, "step": 33300 }, { "grad_norm": 0.3419291079044342, "learning_rate": 6.757896888028871e-05, "loss": 0.024, "step": 33310 }, { "grad_norm": 0.3015671372413635, "learning_rate": 6.7559618514939e-05, "loss": 0.0268, "step": 33320 }, { "grad_norm": 0.3108513355255127, "learning_rate": 6.754026514913073e-05, "loss": 0.0293, "step": 33330 }, { "grad_norm": 0.3433510363101959, "learning_rate": 6.752090878617087e-05, "loss": 0.0383, "step": 33340 }, { "grad_norm": 0.3403770923614502, "learning_rate": 6.75015494293669e-05, "loss": 0.0272, "step": 33350 }, { "grad_norm": 0.33399301767349243, "learning_rate": 6.74821870820268e-05, "loss": 0.0226, "step": 33360 }, { "grad_norm": 0.2705196440219879, "learning_rate": 6.746282174745907e-05, "loss": 0.0357, "step": 33370 }, { "grad_norm": 0.3785693943500519, "learning_rate": 6.744345342897271e-05, "loss": 0.035, "step": 33380 }, { "grad_norm": 0.45232170820236206, "learning_rate": 6.742408212987724e-05, "loss": 0.0311, "step": 33390 }, { "grad_norm": 0.4162551760673523, "learning_rate": 6.740470785348269e-05, "loss": 0.0215, "step": 33400 }, { "grad_norm": 0.3358141779899597, "learning_rate": 6.738533060309958e-05, "loss": 0.0233, "step": 33410 }, { "grad_norm": 0.2581629455089569, "learning_rate": 6.736595038203894e-05, "loss": 0.0251, "step": 33420 }, { "grad_norm": 0.32815036177635193, "learning_rate": 6.734656719361236e-05, "loss": 0.0288, "step": 33430 }, { "grad_norm": 0.32561126351356506, "learning_rate": 6.732718104113189e-05, "loss": 0.0266, "step": 33440 }, { "grad_norm": 0.2906859517097473, "learning_rate": 6.730779192791006e-05, "loss": 0.0205, "step": 33450 }, { "grad_norm": 0.28738877177238464, "learning_rate": 6.728839985725997e-05, "loss": 0.0247, "step": 33460 }, { "grad_norm": 0.28872615098953247, "learning_rate": 6.726900483249517e-05, "loss": 0.026, "step": 33470 }, { "grad_norm": 0.5478929877281189, "learning_rate": 6.724960685692976e-05, "loss": 0.0299, "step": 33480 }, { "grad_norm": 0.2871231138706207, "learning_rate": 6.723020593387833e-05, "loss": 0.0253, "step": 33490 }, { "grad_norm": 0.493680477142334, "learning_rate": 6.721080206665593e-05, "loss": 0.0304, "step": 33500 }, { "grad_norm": 0.3181903064250946, "learning_rate": 6.719139525857819e-05, "loss": 0.0274, "step": 33510 }, { "grad_norm": 0.3962055742740631, "learning_rate": 6.717198551296117e-05, "loss": 0.0262, "step": 33520 }, { "grad_norm": 0.45453205704689026, "learning_rate": 6.715257283312148e-05, "loss": 0.0297, "step": 33530 }, { "grad_norm": 0.4210343360900879, "learning_rate": 6.713315722237623e-05, "loss": 0.0286, "step": 33540 }, { "grad_norm": 0.3353002071380615, "learning_rate": 6.7113738684043e-05, "loss": 0.0302, "step": 33550 }, { "grad_norm": 0.33109375834465027, "learning_rate": 6.709431722143989e-05, "loss": 0.0287, "step": 33560 }, { "grad_norm": 0.34952276945114136, "learning_rate": 6.70748928378855e-05, "loss": 0.0283, "step": 33570 }, { "grad_norm": 0.22823566198349, "learning_rate": 6.705546553669891e-05, "loss": 0.0195, "step": 33580 }, { "grad_norm": 0.2792244553565979, "learning_rate": 6.703603532119974e-05, "loss": 0.0208, "step": 33590 }, { "grad_norm": 0.3159826695919037, "learning_rate": 6.701660219470808e-05, "loss": 0.0228, "step": 33600 }, { "grad_norm": 0.4403702914714813, "learning_rate": 6.69971661605445e-05, "loss": 0.0247, "step": 33610 }, { "grad_norm": 0.3602367043495178, "learning_rate": 6.697772722203008e-05, "loss": 0.0302, "step": 33620 }, { "grad_norm": 0.3221903145313263, "learning_rate": 6.695828538248643e-05, "loss": 0.0276, "step": 33630 }, { "grad_norm": 0.41271206736564636, "learning_rate": 6.693884064523563e-05, "loss": 0.027, "step": 33640 }, { "grad_norm": 0.3018544316291809, "learning_rate": 6.691939301360023e-05, "loss": 0.0297, "step": 33650 }, { "grad_norm": 0.3033134341239929, "learning_rate": 6.689994249090333e-05, "loss": 0.0239, "step": 33660 }, { "grad_norm": 0.4697960913181305, "learning_rate": 6.688048908046845e-05, "loss": 0.0354, "step": 33670 }, { "grad_norm": 0.2513802945613861, "learning_rate": 6.686103278561969e-05, "loss": 0.0251, "step": 33680 }, { "grad_norm": 0.3116767704486847, "learning_rate": 6.684157360968156e-05, "loss": 0.027, "step": 33690 }, { "grad_norm": 0.3731110095977783, "learning_rate": 6.682211155597911e-05, "loss": 0.0221, "step": 33700 }, { "grad_norm": 0.32684215903282166, "learning_rate": 6.680264662783789e-05, "loss": 0.0256, "step": 33710 }, { "grad_norm": 0.38136550784111023, "learning_rate": 6.678317882858391e-05, "loss": 0.0293, "step": 33720 }, { "grad_norm": 0.320980042219162, "learning_rate": 6.67637081615437e-05, "loss": 0.0202, "step": 33730 }, { "grad_norm": 0.22032983601093292, "learning_rate": 6.674423463004427e-05, "loss": 0.0176, "step": 33740 }, { "grad_norm": 0.3647470474243164, "learning_rate": 6.672475823741308e-05, "loss": 0.0236, "step": 33750 }, { "grad_norm": 0.8042439818382263, "learning_rate": 6.670527898697811e-05, "loss": 0.0302, "step": 33760 }, { "grad_norm": 0.35347265005111694, "learning_rate": 6.668579688206788e-05, "loss": 0.0267, "step": 33770 }, { "grad_norm": 0.4918930232524872, "learning_rate": 6.666631192601131e-05, "loss": 0.0262, "step": 33780 }, { "grad_norm": 0.5074679851531982, "learning_rate": 6.664682412213785e-05, "loss": 0.0298, "step": 33790 }, { "grad_norm": 0.31842443346977234, "learning_rate": 6.662733347377745e-05, "loss": 0.0282, "step": 33800 }, { "grad_norm": 0.20004917681217194, "learning_rate": 6.660783998426051e-05, "loss": 0.0234, "step": 33810 }, { "grad_norm": 0.33545467257499695, "learning_rate": 6.658834365691794e-05, "loss": 0.0237, "step": 33820 }, { "grad_norm": 0.25250107049942017, "learning_rate": 6.656884449508115e-05, "loss": 0.0195, "step": 33830 }, { "grad_norm": 0.222221240401268, "learning_rate": 6.654934250208198e-05, "loss": 0.0244, "step": 33840 }, { "grad_norm": 0.2593976557254791, "learning_rate": 6.65298376812528e-05, "loss": 0.0256, "step": 33850 }, { "grad_norm": 0.32228508591651917, "learning_rate": 6.651033003592646e-05, "loss": 0.0277, "step": 33860 }, { "grad_norm": 0.21284104883670807, "learning_rate": 6.649081956943626e-05, "loss": 0.0214, "step": 33870 }, { "grad_norm": 0.37875837087631226, "learning_rate": 6.647130628511604e-05, "loss": 0.0297, "step": 33880 }, { "grad_norm": 0.2833975851535797, "learning_rate": 6.645179018630005e-05, "loss": 0.0284, "step": 33890 }, { "grad_norm": 0.2374112606048584, "learning_rate": 6.643227127632309e-05, "loss": 0.0257, "step": 33900 }, { "grad_norm": 0.33083832263946533, "learning_rate": 6.641274955852038e-05, "loss": 0.0247, "step": 33910 }, { "grad_norm": 0.4129982888698578, "learning_rate": 6.639322503622768e-05, "loss": 0.0441, "step": 33920 }, { "grad_norm": 0.4287928640842438, "learning_rate": 6.637369771278116e-05, "loss": 0.0284, "step": 33930 }, { "grad_norm": 0.60951828956604, "learning_rate": 6.635416759151751e-05, "loss": 0.0214, "step": 33940 }, { "grad_norm": 0.3683372437953949, "learning_rate": 6.633463467577394e-05, "loss": 0.045, "step": 33950 }, { "grad_norm": 0.3290882408618927, "learning_rate": 6.631509896888803e-05, "loss": 0.0302, "step": 33960 }, { "grad_norm": 0.24772155284881592, "learning_rate": 6.629556047419794e-05, "loss": 0.0247, "step": 33970 }, { "grad_norm": 0.23121055960655212, "learning_rate": 6.627601919504223e-05, "loss": 0.0236, "step": 33980 }, { "grad_norm": 0.32214048504829407, "learning_rate": 6.625647513476001e-05, "loss": 0.0199, "step": 33990 }, { "grad_norm": 0.4009026885032654, "learning_rate": 6.62369282966908e-05, "loss": 0.034, "step": 34000 }, { "grad_norm": 0.3699410855770111, "learning_rate": 6.621737868417464e-05, "loss": 0.0286, "step": 34010 }, { "grad_norm": 0.3283851742744446, "learning_rate": 6.619782630055198e-05, "loss": 0.0279, "step": 34020 }, { "grad_norm": 0.37806016206741333, "learning_rate": 6.617827114916382e-05, "loss": 0.026, "step": 34030 }, { "grad_norm": 0.3136194348335266, "learning_rate": 6.615871323335161e-05, "loss": 0.0283, "step": 34040 }, { "grad_norm": 0.3160253167152405, "learning_rate": 6.613915255645725e-05, "loss": 0.0308, "step": 34050 }, { "grad_norm": 0.40607598423957825, "learning_rate": 6.611958912182312e-05, "loss": 0.0324, "step": 34060 }, { "grad_norm": 0.28700172901153564, "learning_rate": 6.610002293279207e-05, "loss": 0.0251, "step": 34070 }, { "grad_norm": 0.26972705125808716, "learning_rate": 6.608045399270746e-05, "loss": 0.0273, "step": 34080 }, { "grad_norm": 0.31609129905700684, "learning_rate": 6.606088230491304e-05, "loss": 0.0242, "step": 34090 }, { "grad_norm": 0.3508830964565277, "learning_rate": 6.604130787275312e-05, "loss": 0.0234, "step": 34100 }, { "grad_norm": 0.28393155336380005, "learning_rate": 6.602173069957242e-05, "loss": 0.0263, "step": 34110 }, { "grad_norm": 0.49455341696739197, "learning_rate": 6.600215078871612e-05, "loss": 0.0252, "step": 34120 }, { "grad_norm": 0.3155761659145355, "learning_rate": 6.598256814352992e-05, "loss": 0.0248, "step": 34130 }, { "grad_norm": 0.33020979166030884, "learning_rate": 6.596298276735995e-05, "loss": 0.0205, "step": 34140 }, { "grad_norm": 0.31914934515953064, "learning_rate": 6.594339466355282e-05, "loss": 0.0278, "step": 34150 }, { "grad_norm": 0.4956086277961731, "learning_rate": 6.592380383545558e-05, "loss": 0.023, "step": 34160 }, { "grad_norm": 0.389556348323822, "learning_rate": 6.590421028641577e-05, "loss": 0.0362, "step": 34170 }, { "grad_norm": 0.3479129374027252, "learning_rate": 6.588461401978143e-05, "loss": 0.0298, "step": 34180 }, { "grad_norm": 0.3220599889755249, "learning_rate": 6.586501503890099e-05, "loss": 0.0347, "step": 34190 }, { "grad_norm": 0.37123918533325195, "learning_rate": 6.584541334712338e-05, "loss": 0.0336, "step": 34200 }, { "grad_norm": 0.5297973155975342, "learning_rate": 6.582580894779802e-05, "loss": 0.0309, "step": 34210 }, { "grad_norm": 0.255703330039978, "learning_rate": 6.580620184427473e-05, "loss": 0.0242, "step": 34220 }, { "grad_norm": 0.23578694462776184, "learning_rate": 6.578659203990385e-05, "loss": 0.026, "step": 34230 }, { "grad_norm": 0.30924221873283386, "learning_rate": 6.576697953803615e-05, "loss": 0.0208, "step": 34240 }, { "grad_norm": 0.2953323721885681, "learning_rate": 6.57473643420229e-05, "loss": 0.0224, "step": 34250 }, { "grad_norm": 0.31066736578941345, "learning_rate": 6.572774645521574e-05, "loss": 0.028, "step": 34260 }, { "grad_norm": 0.6029891967773438, "learning_rate": 6.570812588096688e-05, "loss": 0.0297, "step": 34270 }, { "grad_norm": 0.2045934945344925, "learning_rate": 6.568850262262893e-05, "loss": 0.0276, "step": 34280 }, { "grad_norm": 0.36300480365753174, "learning_rate": 6.566887668355497e-05, "loss": 0.024, "step": 34290 }, { "grad_norm": 0.3663865625858307, "learning_rate": 6.564924806709851e-05, "loss": 0.0248, "step": 34300 }, { "grad_norm": 0.3913494050502777, "learning_rate": 6.562961677661359e-05, "loss": 0.0327, "step": 34310 }, { "grad_norm": 0.42498326301574707, "learning_rate": 6.56099828154546e-05, "loss": 0.0276, "step": 34320 }, { "grad_norm": 0.37097233533859253, "learning_rate": 6.55903461869765e-05, "loss": 0.0269, "step": 34330 }, { "grad_norm": 0.33006635308265686, "learning_rate": 6.557070689453465e-05, "loss": 0.0225, "step": 34340 }, { "grad_norm": 0.21983970701694489, "learning_rate": 6.555106494148482e-05, "loss": 0.0303, "step": 34350 }, { "grad_norm": 0.38927724957466125, "learning_rate": 6.553142033118333e-05, "loss": 0.026, "step": 34360 }, { "grad_norm": 0.33707091212272644, "learning_rate": 6.551177306698688e-05, "loss": 0.0271, "step": 34370 }, { "grad_norm": 0.34458082914352417, "learning_rate": 6.549212315225267e-05, "loss": 0.0324, "step": 34380 }, { "grad_norm": 0.39501112699508667, "learning_rate": 6.547247059033833e-05, "loss": 0.0324, "step": 34390 }, { "grad_norm": 0.3578024208545685, "learning_rate": 6.545281538460193e-05, "loss": 0.0281, "step": 34400 }, { "grad_norm": 0.32430776953697205, "learning_rate": 6.543315753840202e-05, "loss": 0.0274, "step": 34410 }, { "grad_norm": 0.44063693284988403, "learning_rate": 6.541349705509758e-05, "loss": 0.0255, "step": 34420 }, { "grad_norm": 0.39670947194099426, "learning_rate": 6.539383393804805e-05, "loss": 0.0279, "step": 34430 }, { "grad_norm": 0.3601522743701935, "learning_rate": 6.537416819061333e-05, "loss": 0.0309, "step": 34440 }, { "grad_norm": 0.3533051311969757, "learning_rate": 6.535449981615375e-05, "loss": 0.0265, "step": 34450 }, { "grad_norm": 0.31081047654151917, "learning_rate": 6.53348288180301e-05, "loss": 0.0305, "step": 34460 }, { "grad_norm": 0.40017563104629517, "learning_rate": 6.531515519960361e-05, "loss": 0.0377, "step": 34470 }, { "grad_norm": 0.3736017942428589, "learning_rate": 6.529547896423597e-05, "loss": 0.0305, "step": 34480 }, { "grad_norm": 0.20441938936710358, "learning_rate": 6.52758001152893e-05, "loss": 0.0268, "step": 34490 }, { "grad_norm": 0.2938135266304016, "learning_rate": 6.525611865612618e-05, "loss": 0.0223, "step": 34500 }, { "grad_norm": 0.4369048774242401, "learning_rate": 6.523643459010966e-05, "loss": 0.0255, "step": 34510 }, { "grad_norm": 0.5107825398445129, "learning_rate": 6.521674792060317e-05, "loss": 0.0237, "step": 34520 }, { "grad_norm": 0.38623201847076416, "learning_rate": 6.519705865097063e-05, "loss": 0.0285, "step": 34530 }, { "grad_norm": 0.4038742780685425, "learning_rate": 6.517736678457641e-05, "loss": 0.0238, "step": 34540 }, { "grad_norm": 0.3280414938926697, "learning_rate": 6.515767232478534e-05, "loss": 0.0269, "step": 34550 }, { "grad_norm": 0.4126679301261902, "learning_rate": 6.51379752749626e-05, "loss": 0.0219, "step": 34560 }, { "grad_norm": 0.2641867697238922, "learning_rate": 6.511827563847393e-05, "loss": 0.0221, "step": 34570 }, { "grad_norm": 0.2902062237262726, "learning_rate": 6.509857341868542e-05, "loss": 0.0265, "step": 34580 }, { "grad_norm": 0.2859358787536621, "learning_rate": 6.507886861896367e-05, "loss": 0.0306, "step": 34590 }, { "grad_norm": 0.46614232659339905, "learning_rate": 6.505916124267567e-05, "loss": 0.0318, "step": 34600 }, { "grad_norm": 0.6446811556816101, "learning_rate": 6.503945129318891e-05, "loss": 0.0284, "step": 34610 }, { "grad_norm": 0.34342435002326965, "learning_rate": 6.501973877387122e-05, "loss": 0.0226, "step": 34620 }, { "grad_norm": 0.42564502358436584, "learning_rate": 6.500002368809098e-05, "loss": 0.0256, "step": 34630 }, { "grad_norm": 0.3624362647533417, "learning_rate": 6.498030603921694e-05, "loss": 0.0293, "step": 34640 }, { "grad_norm": 0.4055345952510834, "learning_rate": 6.496058583061832e-05, "loss": 0.0333, "step": 34650 }, { "grad_norm": 0.32309427857398987, "learning_rate": 6.494086306566475e-05, "loss": 0.0287, "step": 34660 }, { "grad_norm": 0.2848092317581177, "learning_rate": 6.492113774772632e-05, "loss": 0.0324, "step": 34670 }, { "grad_norm": 0.420859694480896, "learning_rate": 6.490140988017354e-05, "loss": 0.0387, "step": 34680 }, { "grad_norm": 0.39663583040237427, "learning_rate": 6.488167946637736e-05, "loss": 0.041, "step": 34690 }, { "grad_norm": 0.5039676427841187, "learning_rate": 6.486194650970915e-05, "loss": 0.0352, "step": 34700 }, { "grad_norm": 0.4263099133968353, "learning_rate": 6.48422110135408e-05, "loss": 0.0357, "step": 34710 }, { "grad_norm": 0.33712929487228394, "learning_rate": 6.482247298124451e-05, "loss": 0.045, "step": 34720 }, { "grad_norm": 0.39799362421035767, "learning_rate": 6.480273241619297e-05, "loss": 0.0267, "step": 34730 }, { "grad_norm": 0.38913166522979736, "learning_rate": 6.478298932175933e-05, "loss": 0.0266, "step": 34740 }, { "grad_norm": 0.25853779911994934, "learning_rate": 6.476324370131712e-05, "loss": 0.0261, "step": 34750 }, { "grad_norm": 0.46820124983787537, "learning_rate": 6.474349555824036e-05, "loss": 0.031, "step": 34760 }, { "grad_norm": 0.2706189751625061, "learning_rate": 6.472374489590342e-05, "loss": 0.0216, "step": 34770 }, { "grad_norm": 0.3956297039985657, "learning_rate": 6.470399171768118e-05, "loss": 0.0388, "step": 34780 }, { "grad_norm": 0.43301263451576233, "learning_rate": 6.468423602694891e-05, "loss": 0.0265, "step": 34790 }, { "grad_norm": 0.34632429480552673, "learning_rate": 6.466447782708232e-05, "loss": 0.032, "step": 34800 }, { "grad_norm": 0.3898659348487854, "learning_rate": 6.464471712145754e-05, "loss": 0.0283, "step": 34810 }, { "grad_norm": 0.3116031587123871, "learning_rate": 6.462495391345114e-05, "loss": 0.0253, "step": 34820 }, { "grad_norm": 0.23747770488262177, "learning_rate": 6.46051882064401e-05, "loss": 0.0248, "step": 34830 }, { "grad_norm": 0.4285765588283539, "learning_rate": 6.458542000380186e-05, "loss": 0.0266, "step": 34840 }, { "grad_norm": 0.3143494725227356, "learning_rate": 6.456564930891424e-05, "loss": 0.0213, "step": 34850 }, { "grad_norm": 0.2819913625717163, "learning_rate": 6.454587612515555e-05, "loss": 0.0258, "step": 34860 }, { "grad_norm": 0.410124808549881, "learning_rate": 6.452610045590444e-05, "loss": 0.0303, "step": 34870 }, { "grad_norm": 0.31080326437950134, "learning_rate": 6.450632230454005e-05, "loss": 0.025, "step": 34880 }, { "grad_norm": 0.3373957872390747, "learning_rate": 6.448654167444195e-05, "loss": 0.0374, "step": 34890 }, { "grad_norm": 0.3730693757534027, "learning_rate": 6.446675856899005e-05, "loss": 0.0258, "step": 34900 }, { "grad_norm": 0.40712878108024597, "learning_rate": 6.444697299156481e-05, "loss": 0.031, "step": 34910 }, { "grad_norm": 0.4844147861003876, "learning_rate": 6.442718494554701e-05, "loss": 0.0293, "step": 34920 }, { "grad_norm": 0.3411736786365509, "learning_rate": 6.440739443431787e-05, "loss": 0.0239, "step": 34930 }, { "grad_norm": 0.27882978320121765, "learning_rate": 6.438760146125906e-05, "loss": 0.0219, "step": 34940 }, { "grad_norm": 0.550361156463623, "learning_rate": 6.436780602975267e-05, "loss": 0.0333, "step": 34950 }, { "grad_norm": 0.36173659563064575, "learning_rate": 6.43480081431812e-05, "loss": 0.0304, "step": 34960 }, { "grad_norm": 0.5546711087226868, "learning_rate": 6.432820780492756e-05, "loss": 0.0398, "step": 34970 }, { "grad_norm": 0.327696830034256, "learning_rate": 6.430840501837506e-05, "loss": 0.0307, "step": 34980 }, { "grad_norm": 0.388715535402298, "learning_rate": 6.428859978690748e-05, "loss": 0.0267, "step": 34990 }, { "grad_norm": 0.34412717819213867, "learning_rate": 6.426879211390901e-05, "loss": 0.0262, "step": 35000 }, { "grad_norm": 0.4576735198497772, "learning_rate": 6.424898200276422e-05, "loss": 0.0284, "step": 35010 }, { "grad_norm": 0.33526867628097534, "learning_rate": 6.42291694568581e-05, "loss": 0.0216, "step": 35020 }, { "grad_norm": 0.261588990688324, "learning_rate": 6.42093544795761e-05, "loss": 0.0331, "step": 35030 }, { "grad_norm": 0.2582116723060608, "learning_rate": 6.418953707430403e-05, "loss": 0.0274, "step": 35040 }, { "grad_norm": 0.4441741704940796, "learning_rate": 6.416971724442819e-05, "loss": 0.029, "step": 35050 }, { "grad_norm": 0.44424161314964294, "learning_rate": 6.414989499333519e-05, "loss": 0.0318, "step": 35060 }, { "grad_norm": 0.33735528588294983, "learning_rate": 6.413007032441214e-05, "loss": 0.0335, "step": 35070 }, { "grad_norm": 0.3528691232204437, "learning_rate": 6.411024324104653e-05, "loss": 0.0331, "step": 35080 }, { "grad_norm": 0.3767465054988861, "learning_rate": 6.409041374662628e-05, "loss": 0.0259, "step": 35090 }, { "grad_norm": 0.33648499846458435, "learning_rate": 6.407058184453967e-05, "loss": 0.0406, "step": 35100 }, { "grad_norm": 0.29010745882987976, "learning_rate": 6.405074753817548e-05, "loss": 0.0261, "step": 35110 }, { "grad_norm": 0.5161163806915283, "learning_rate": 6.40309108309228e-05, "loss": 0.0278, "step": 35120 }, { "grad_norm": 0.33087432384490967, "learning_rate": 6.401107172617122e-05, "loss": 0.0265, "step": 35130 }, { "grad_norm": 0.3779878318309784, "learning_rate": 6.399123022731068e-05, "loss": 0.0351, "step": 35140 }, { "grad_norm": 0.3102956712245941, "learning_rate": 6.397138633773157e-05, "loss": 0.0249, "step": 35150 }, { "grad_norm": 0.5206102728843689, "learning_rate": 6.395154006082463e-05, "loss": 0.0329, "step": 35160 }, { "grad_norm": 0.26042062044143677, "learning_rate": 6.393169139998109e-05, "loss": 0.0257, "step": 35170 }, { "grad_norm": 0.31631895899772644, "learning_rate": 6.39118403585925e-05, "loss": 0.0278, "step": 35180 }, { "grad_norm": 0.36257290840148926, "learning_rate": 6.38919869400509e-05, "loss": 0.0247, "step": 35190 }, { "grad_norm": 0.1772301346063614, "learning_rate": 6.387213114774865e-05, "loss": 0.024, "step": 35200 }, { "grad_norm": 0.3098105192184448, "learning_rate": 6.385227298507863e-05, "loss": 0.0452, "step": 35210 }, { "grad_norm": 0.4077107906341553, "learning_rate": 6.3832412455434e-05, "loss": 0.0194, "step": 35220 }, { "grad_norm": 0.2210332602262497, "learning_rate": 6.381254956220841e-05, "loss": 0.0251, "step": 35230 }, { "grad_norm": 0.23701924085617065, "learning_rate": 6.379268430879586e-05, "loss": 0.0256, "step": 35240 }, { "grad_norm": 0.3046592175960541, "learning_rate": 6.37728166985908e-05, "loss": 0.0238, "step": 35250 }, { "grad_norm": 0.2623172700405121, "learning_rate": 6.375294673498804e-05, "loss": 0.0215, "step": 35260 }, { "grad_norm": 0.3752667009830475, "learning_rate": 6.373307442138284e-05, "loss": 0.0271, "step": 35270 }, { "grad_norm": 0.29532238841056824, "learning_rate": 6.371319976117081e-05, "loss": 0.0324, "step": 35280 }, { "grad_norm": 0.35120344161987305, "learning_rate": 6.3693322757748e-05, "loss": 0.0243, "step": 35290 }, { "grad_norm": 0.33842331171035767, "learning_rate": 6.367344341451086e-05, "loss": 0.0325, "step": 35300 }, { "grad_norm": 0.3653583824634552, "learning_rate": 6.36535617348562e-05, "loss": 0.0286, "step": 35310 }, { "grad_norm": 0.33795225620269775, "learning_rate": 6.363367772218128e-05, "loss": 0.025, "step": 35320 }, { "grad_norm": 0.5360882878303528, "learning_rate": 6.36137913798837e-05, "loss": 0.0286, "step": 35330 }, { "grad_norm": 0.3270680606365204, "learning_rate": 6.359390271136151e-05, "loss": 0.0218, "step": 35340 }, { "grad_norm": 0.29643145203590393, "learning_rate": 6.357401172001314e-05, "loss": 0.0264, "step": 35350 }, { "grad_norm": 0.24067223072052002, "learning_rate": 6.355411840923742e-05, "loss": 0.0233, "step": 35360 }, { "grad_norm": 0.46575814485549927, "learning_rate": 6.353422278243358e-05, "loss": 0.028, "step": 35370 }, { "grad_norm": 0.3346884846687317, "learning_rate": 6.351432484300121e-05, "loss": 0.0271, "step": 35380 }, { "grad_norm": 0.3491145968437195, "learning_rate": 6.349442459434036e-05, "loss": 0.0313, "step": 35390 }, { "grad_norm": 0.2820169925689697, "learning_rate": 6.34745220398514e-05, "loss": 0.0339, "step": 35400 }, { "grad_norm": 0.35578182339668274, "learning_rate": 6.345461718293518e-05, "loss": 0.0261, "step": 35410 }, { "grad_norm": 0.2934359312057495, "learning_rate": 6.343471002699286e-05, "loss": 0.0317, "step": 35420 }, { "grad_norm": 0.38613808155059814, "learning_rate": 6.341480057542602e-05, "loss": 0.0335, "step": 35430 }, { "grad_norm": 0.3660021126270294, "learning_rate": 6.339488883163667e-05, "loss": 0.0262, "step": 35440 }, { "grad_norm": 0.24424666166305542, "learning_rate": 6.337497479902716e-05, "loss": 0.0229, "step": 35450 }, { "grad_norm": 0.3132448196411133, "learning_rate": 6.335505848100027e-05, "loss": 0.023, "step": 35460 }, { "grad_norm": 0.3724532127380371, "learning_rate": 6.333513988095915e-05, "loss": 0.021, "step": 35470 }, { "grad_norm": 0.24682928621768951, "learning_rate": 6.331521900230735e-05, "loss": 0.0223, "step": 35480 }, { "grad_norm": 0.23878753185272217, "learning_rate": 6.329529584844878e-05, "loss": 0.0218, "step": 35490 }, { "grad_norm": 0.35610008239746094, "learning_rate": 6.327537042278777e-05, "loss": 0.0282, "step": 35500 }, { "grad_norm": 0.37253206968307495, "learning_rate": 6.325544272872905e-05, "loss": 0.0251, "step": 35510 }, { "grad_norm": 0.37161606550216675, "learning_rate": 6.323551276967771e-05, "loss": 0.0325, "step": 35520 }, { "grad_norm": 0.3216058611869812, "learning_rate": 6.321558054903922e-05, "loss": 0.0231, "step": 35530 }, { "grad_norm": 0.3182275593280792, "learning_rate": 6.319564607021947e-05, "loss": 0.029, "step": 35540 }, { "grad_norm": 0.24898004531860352, "learning_rate": 6.31757093366247e-05, "loss": 0.022, "step": 35550 }, { "grad_norm": 0.331838458776474, "learning_rate": 6.315577035166154e-05, "loss": 0.0251, "step": 35560 }, { "grad_norm": 0.46139466762542725, "learning_rate": 6.313582911873708e-05, "loss": 0.0241, "step": 35570 }, { "grad_norm": 0.2993537187576294, "learning_rate": 6.311588564125865e-05, "loss": 0.0261, "step": 35580 }, { "grad_norm": 0.5127886533737183, "learning_rate": 6.30959399226341e-05, "loss": 0.0306, "step": 35590 }, { "grad_norm": 0.307932585477829, "learning_rate": 6.30759919662716e-05, "loss": 0.0302, "step": 35600 }, { "grad_norm": 0.26685982942581177, "learning_rate": 6.30560417755797e-05, "loss": 0.0245, "step": 35610 }, { "grad_norm": 0.20920807123184204, "learning_rate": 6.303608935396735e-05, "loss": 0.0198, "step": 35620 }, { "grad_norm": 0.27881893515586853, "learning_rate": 6.301613470484386e-05, "loss": 0.0261, "step": 35630 }, { "grad_norm": 0.3363640010356903, "learning_rate": 6.299617783161893e-05, "loss": 0.0251, "step": 35640 }, { "grad_norm": 0.3209003806114197, "learning_rate": 6.297621873770266e-05, "loss": 0.0253, "step": 35650 }, { "grad_norm": 0.27805736660957336, "learning_rate": 6.29562574265055e-05, "loss": 0.0258, "step": 35660 }, { "grad_norm": 0.46072620153427124, "learning_rate": 6.293629390143834e-05, "loss": 0.0271, "step": 35670 }, { "grad_norm": 0.3286251425743103, "learning_rate": 6.291632816591232e-05, "loss": 0.0245, "step": 35680 }, { "grad_norm": 0.24018752574920654, "learning_rate": 6.28963602233391e-05, "loss": 0.0217, "step": 35690 }, { "grad_norm": 0.3727167546749115, "learning_rate": 6.287639007713062e-05, "loss": 0.0307, "step": 35700 }, { "grad_norm": 0.2737526595592499, "learning_rate": 6.285641773069926e-05, "loss": 0.0238, "step": 35710 }, { "grad_norm": 0.24851885437965393, "learning_rate": 6.283644318745773e-05, "loss": 0.0199, "step": 35720 }, { "grad_norm": 0.28573670983314514, "learning_rate": 6.281646645081912e-05, "loss": 0.0273, "step": 35730 }, { "grad_norm": 0.2942921221256256, "learning_rate": 6.279648752419693e-05, "loss": 0.0254, "step": 35740 }, { "grad_norm": 0.3743562400341034, "learning_rate": 6.2776506411005e-05, "loss": 0.0257, "step": 35750 }, { "grad_norm": 0.46878278255462646, "learning_rate": 6.275652311465758e-05, "loss": 0.0225, "step": 35760 }, { "grad_norm": 0.2814807593822479, "learning_rate": 6.273653763856926e-05, "loss": 0.0203, "step": 35770 }, { "grad_norm": 0.31582212448120117, "learning_rate": 6.271654998615501e-05, "loss": 0.0345, "step": 35780 }, { "grad_norm": 0.36535343527793884, "learning_rate": 6.269656016083013e-05, "loss": 0.0239, "step": 35790 }, { "grad_norm": 0.42740896344184875, "learning_rate": 6.267656816601038e-05, "loss": 0.0362, "step": 35800 }, { "grad_norm": 0.34653395414352417, "learning_rate": 6.265657400511185e-05, "loss": 0.0289, "step": 35810 }, { "grad_norm": 0.38382688164711, "learning_rate": 6.263657768155098e-05, "loss": 0.0278, "step": 35820 }, { "grad_norm": 0.3512437343597412, "learning_rate": 6.261657919874457e-05, "loss": 0.0251, "step": 35830 }, { "grad_norm": 0.40686264634132385, "learning_rate": 6.259657856010986e-05, "loss": 0.0319, "step": 35840 }, { "grad_norm": 0.3283431828022003, "learning_rate": 6.257657576906439e-05, "loss": 0.0266, "step": 35850 }, { "grad_norm": 0.4602685570716858, "learning_rate": 6.255657082902609e-05, "loss": 0.0258, "step": 35860 }, { "grad_norm": 0.28780168294906616, "learning_rate": 6.253656374341325e-05, "loss": 0.026, "step": 35870 }, { "grad_norm": 0.38708367943763733, "learning_rate": 6.251655451564457e-05, "loss": 0.0286, "step": 35880 }, { "grad_norm": 0.2788851857185364, "learning_rate": 6.249654314913902e-05, "loss": 0.0203, "step": 35890 }, { "grad_norm": 0.23026952147483826, "learning_rate": 6.247652964731604e-05, "loss": 0.0243, "step": 35900 }, { "grad_norm": 0.2641853988170624, "learning_rate": 6.245651401359537e-05, "loss": 0.0443, "step": 35910 }, { "grad_norm": 0.3320379853248596, "learning_rate": 6.243649625139715e-05, "loss": 0.0284, "step": 35920 }, { "grad_norm": 0.26835593581199646, "learning_rate": 6.241647636414185e-05, "loss": 0.0227, "step": 35930 }, { "grad_norm": 0.5485971570014954, "learning_rate": 6.239645435525034e-05, "loss": 0.0325, "step": 35940 }, { "grad_norm": 0.7935574054718018, "learning_rate": 6.237643022814381e-05, "loss": 0.0249, "step": 35950 }, { "grad_norm": 0.383373886346817, "learning_rate": 6.235640398624386e-05, "loss": 0.0238, "step": 35960 }, { "grad_norm": 0.31369099020957947, "learning_rate": 6.233637563297243e-05, "loss": 0.0306, "step": 35970 }, { "grad_norm": 0.4012422561645508, "learning_rate": 6.23163451717518e-05, "loss": 0.0321, "step": 35980 }, { "grad_norm": 0.35285040736198425, "learning_rate": 6.229631260600463e-05, "loss": 0.0269, "step": 35990 }, { "grad_norm": 0.40435805916786194, "learning_rate": 6.227627793915392e-05, "loss": 0.0353, "step": 36000 }, { "grad_norm": 0.3581738770008087, "learning_rate": 6.225624117462309e-05, "loss": 0.0267, "step": 36010 }, { "grad_norm": 0.3104953169822693, "learning_rate": 6.223620231583586e-05, "loss": 0.025, "step": 36020 }, { "grad_norm": 0.25324201583862305, "learning_rate": 6.221616136621629e-05, "loss": 0.0252, "step": 36030 }, { "grad_norm": 0.3458791673183441, "learning_rate": 6.219611832918887e-05, "loss": 0.0267, "step": 36040 }, { "grad_norm": 0.29009705781936646, "learning_rate": 6.217607320817838e-05, "loss": 0.0216, "step": 36050 }, { "grad_norm": 0.2704320549964905, "learning_rate": 6.215602600661001e-05, "loss": 0.0236, "step": 36060 }, { "grad_norm": 0.43715399503707886, "learning_rate": 6.213597672790925e-05, "loss": 0.03, "step": 36070 }, { "grad_norm": 0.3337031304836273, "learning_rate": 6.2115925375502e-05, "loss": 0.0264, "step": 36080 }, { "grad_norm": 0.572075605392456, "learning_rate": 6.209587195281447e-05, "loss": 0.0288, "step": 36090 }, { "grad_norm": 0.5467438697814941, "learning_rate": 6.207581646327324e-05, "loss": 0.0293, "step": 36100 }, { "grad_norm": 0.18307340145111084, "learning_rate": 6.205575891030526e-05, "loss": 0.0178, "step": 36110 }, { "grad_norm": 0.2446993738412857, "learning_rate": 6.203569929733781e-05, "loss": 0.0257, "step": 36120 }, { "grad_norm": 0.3253853917121887, "learning_rate": 6.201563762779852e-05, "loss": 0.0259, "step": 36130 }, { "grad_norm": 0.18920265138149261, "learning_rate": 6.199557390511538e-05, "loss": 0.0257, "step": 36140 }, { "grad_norm": 0.3416122794151306, "learning_rate": 6.197550813271675e-05, "loss": 0.027, "step": 36150 }, { "grad_norm": 0.30043333768844604, "learning_rate": 6.195544031403131e-05, "loss": 0.0326, "step": 36160 }, { "grad_norm": 0.2986289858818054, "learning_rate": 6.19353704524881e-05, "loss": 0.0218, "step": 36170 }, { "grad_norm": 0.24846267700195312, "learning_rate": 6.191529855151652e-05, "loss": 0.0199, "step": 36180 }, { "grad_norm": 0.3967738747596741, "learning_rate": 6.189522461454629e-05, "loss": 0.0247, "step": 36190 }, { "grad_norm": 0.2781318128108978, "learning_rate": 6.187514864500752e-05, "loss": 0.0263, "step": 36200 }, { "grad_norm": 0.442780077457428, "learning_rate": 6.185507064633062e-05, "loss": 0.0263, "step": 36210 }, { "grad_norm": 0.490308940410614, "learning_rate": 6.18349906219464e-05, "loss": 0.0268, "step": 36220 }, { "grad_norm": 0.4768640100955963, "learning_rate": 6.181490857528596e-05, "loss": 0.0257, "step": 36230 }, { "grad_norm": 0.27944767475128174, "learning_rate": 6.179482450978077e-05, "loss": 0.0264, "step": 36240 }, { "grad_norm": 0.39189857244491577, "learning_rate": 6.177473842886269e-05, "loss": 0.028, "step": 36250 }, { "grad_norm": 0.3097172677516937, "learning_rate": 6.175465033596382e-05, "loss": 0.0321, "step": 36260 }, { "grad_norm": 0.34056565165519714, "learning_rate": 6.173456023451671e-05, "loss": 0.0269, "step": 36270 }, { "grad_norm": 0.40845978260040283, "learning_rate": 6.171446812795422e-05, "loss": 0.0306, "step": 36280 }, { "grad_norm": 0.244633287191391, "learning_rate": 6.169437401970949e-05, "loss": 0.0355, "step": 36290 }, { "grad_norm": 0.17470622062683105, "learning_rate": 6.16742779132161e-05, "loss": 0.021, "step": 36300 }, { "grad_norm": 0.265927255153656, "learning_rate": 6.165417981190789e-05, "loss": 0.024, "step": 36310 }, { "grad_norm": 0.38698190450668335, "learning_rate": 6.16340797192191e-05, "loss": 0.0227, "step": 36320 }, { "grad_norm": 0.2959088981151581, "learning_rate": 6.161397763858427e-05, "loss": 0.0227, "step": 36330 }, { "grad_norm": 0.269369512796402, "learning_rate": 6.159387357343834e-05, "loss": 0.0202, "step": 36340 }, { "grad_norm": 0.331415593624115, "learning_rate": 6.157376752721648e-05, "loss": 0.0242, "step": 36350 }, { "grad_norm": 0.28486737608909607, "learning_rate": 6.155365950335428e-05, "loss": 0.0287, "step": 36360 }, { "grad_norm": 0.48234301805496216, "learning_rate": 6.153354950528768e-05, "loss": 0.0246, "step": 36370 }, { "grad_norm": 0.29302921891212463, "learning_rate": 6.151343753645293e-05, "loss": 0.026, "step": 36380 }, { "grad_norm": 0.31269633769989014, "learning_rate": 6.149332360028657e-05, "loss": 0.0309, "step": 36390 }, { "grad_norm": 0.2499934285879135, "learning_rate": 6.147320770022555e-05, "loss": 0.0206, "step": 36400 }, { "grad_norm": 0.26872801780700684, "learning_rate": 6.145308983970715e-05, "loss": 0.0297, "step": 36410 }, { "grad_norm": 0.1863371878862381, "learning_rate": 6.143297002216892e-05, "loss": 0.0231, "step": 36420 }, { "grad_norm": 0.24369506537914276, "learning_rate": 6.141284825104882e-05, "loss": 0.0233, "step": 36430 }, { "grad_norm": 0.3993298411369324, "learning_rate": 6.13927245297851e-05, "loss": 0.0305, "step": 36440 }, { "grad_norm": 0.3309917449951172, "learning_rate": 6.137259886181633e-05, "loss": 0.0323, "step": 36450 }, { "grad_norm": 0.33395183086395264, "learning_rate": 6.135247125058145e-05, "loss": 0.0316, "step": 36460 }, { "grad_norm": 0.2653490900993347, "learning_rate": 6.133234169951974e-05, "loss": 0.0207, "step": 36470 }, { "grad_norm": 0.23726914823055267, "learning_rate": 6.131221021207078e-05, "loss": 0.0305, "step": 36480 }, { "grad_norm": 0.2719564735889435, "learning_rate": 6.129207679167448e-05, "loss": 0.0187, "step": 36490 }, { "grad_norm": 0.2665988802909851, "learning_rate": 6.127194144177109e-05, "loss": 0.0185, "step": 36500 }, { "grad_norm": 0.3915807604789734, "learning_rate": 6.125180416580118e-05, "loss": 0.0208, "step": 36510 }, { "grad_norm": 0.308644562959671, "learning_rate": 6.123166496720571e-05, "loss": 0.0208, "step": 36520 }, { "grad_norm": 0.23165039718151093, "learning_rate": 6.121152384942588e-05, "loss": 0.024, "step": 36530 }, { "grad_norm": 0.2802601754665375, "learning_rate": 6.119138081590324e-05, "loss": 0.0249, "step": 36540 }, { "grad_norm": 0.3249892294406891, "learning_rate": 6.117123587007971e-05, "loss": 0.0213, "step": 36550 }, { "grad_norm": 0.2567669153213501, "learning_rate": 6.11510890153975e-05, "loss": 0.0291, "step": 36560 }, { "grad_norm": 0.24312350153923035, "learning_rate": 6.113094025529916e-05, "loss": 0.0226, "step": 36570 }, { "grad_norm": 0.3329765796661377, "learning_rate": 6.111078959322757e-05, "loss": 0.0282, "step": 36580 }, { "grad_norm": 0.48775357007980347, "learning_rate": 6.109063703262592e-05, "loss": 0.0314, "step": 36590 }, { "grad_norm": 0.5292181968688965, "learning_rate": 6.107048257693772e-05, "loss": 0.0286, "step": 36600 }, { "grad_norm": 0.2535587549209595, "learning_rate": 6.105032622960683e-05, "loss": 0.0245, "step": 36610 }, { "grad_norm": 0.3213077485561371, "learning_rate": 6.103016799407743e-05, "loss": 0.0212, "step": 36620 }, { "grad_norm": 0.30457696318626404, "learning_rate": 6.1010007873793984e-05, "loss": 0.029, "step": 36630 }, { "grad_norm": 0.3187440037727356, "learning_rate": 6.098984587220131e-05, "loss": 0.0236, "step": 36640 }, { "grad_norm": 0.2984723746776581, "learning_rate": 6.096968199274456e-05, "loss": 0.028, "step": 36650 }, { "grad_norm": 0.31568217277526855, "learning_rate": 6.0949516238869166e-05, "loss": 0.0267, "step": 36660 }, { "grad_norm": 0.3946670591831207, "learning_rate": 6.092934861402092e-05, "loss": 0.0306, "step": 36670 }, { "grad_norm": 0.375176340341568, "learning_rate": 6.0909179121645924e-05, "loss": 0.0251, "step": 36680 }, { "grad_norm": 0.384289413690567, "learning_rate": 6.0889007765190576e-05, "loss": 0.0277, "step": 36690 }, { "grad_norm": 0.22571048140525818, "learning_rate": 6.086883454810162e-05, "loss": 0.0201, "step": 36700 }, { "grad_norm": 0.26885268092155457, "learning_rate": 6.0848659473826084e-05, "loss": 0.0216, "step": 36710 }, { "grad_norm": 0.33899903297424316, "learning_rate": 6.082848254581138e-05, "loss": 0.0279, "step": 36720 }, { "grad_norm": 0.39822259545326233, "learning_rate": 6.080830376750517e-05, "loss": 0.0283, "step": 36730 }, { "grad_norm": 0.23543496429920197, "learning_rate": 6.0788123142355445e-05, "loss": 0.0217, "step": 36740 }, { "grad_norm": 0.3274632692337036, "learning_rate": 6.076794067381052e-05, "loss": 0.0259, "step": 36750 }, { "grad_norm": 0.20764654874801636, "learning_rate": 6.074775636531905e-05, "loss": 0.0253, "step": 36760 }, { "grad_norm": 0.39131858944892883, "learning_rate": 6.072757022032997e-05, "loss": 0.021, "step": 36770 }, { "grad_norm": 0.2633682191371918, "learning_rate": 6.070738224229253e-05, "loss": 0.0304, "step": 36780 }, { "grad_norm": 0.2809239625930786, "learning_rate": 6.0687192434656314e-05, "loss": 0.0272, "step": 36790 }, { "grad_norm": 0.43270817399024963, "learning_rate": 6.066700080087121e-05, "loss": 0.0251, "step": 36800 }, { "grad_norm": 0.23554883897304535, "learning_rate": 6.0646807344387424e-05, "loss": 0.0213, "step": 36810 }, { "grad_norm": 0.4003365933895111, "learning_rate": 6.062661206865543e-05, "loss": 0.0302, "step": 36820 }, { "grad_norm": 0.2386820763349533, "learning_rate": 6.06064149771261e-05, "loss": 0.0221, "step": 36830 }, { "grad_norm": 0.3161543905735016, "learning_rate": 6.058621607325051e-05, "loss": 0.0226, "step": 36840 }, { "grad_norm": 0.33431902527809143, "learning_rate": 6.056601536048014e-05, "loss": 0.0293, "step": 36850 }, { "grad_norm": 0.2554515302181244, "learning_rate": 6.0545812842266725e-05, "loss": 0.0283, "step": 36860 }, { "grad_norm": 0.2771773338317871, "learning_rate": 6.052560852206232e-05, "loss": 0.0285, "step": 36870 }, { "grad_norm": 0.2900908887386322, "learning_rate": 6.05054024033193e-05, "loss": 0.0337, "step": 36880 }, { "grad_norm": 0.29625529050827026, "learning_rate": 6.048519448949032e-05, "loss": 0.0237, "step": 36890 }, { "grad_norm": 0.309256374835968, "learning_rate": 6.046498478402839e-05, "loss": 0.0304, "step": 36900 }, { "grad_norm": 0.2987845242023468, "learning_rate": 6.044477329038677e-05, "loss": 0.0266, "step": 36910 }, { "grad_norm": 0.3389832377433777, "learning_rate": 6.042456001201906e-05, "loss": 0.0237, "step": 36920 }, { "grad_norm": 0.20289385318756104, "learning_rate": 6.040434495237917e-05, "loss": 0.0258, "step": 36930 }, { "grad_norm": 0.35594674944877625, "learning_rate": 6.0384128114921256e-05, "loss": 0.0307, "step": 36940 }, { "grad_norm": 0.31436923146247864, "learning_rate": 6.036390950309987e-05, "loss": 0.0179, "step": 36950 }, { "grad_norm": 0.324798583984375, "learning_rate": 6.0343689120369805e-05, "loss": 0.0259, "step": 36960 }, { "grad_norm": 0.2989186942577362, "learning_rate": 6.032346697018616e-05, "loss": 0.0223, "step": 36970 }, { "grad_norm": 0.3364205062389374, "learning_rate": 6.0303243056004375e-05, "loss": 0.034, "step": 36980 }, { "grad_norm": 0.2294752597808838, "learning_rate": 6.0283017381280136e-05, "loss": 0.0229, "step": 36990 }, { "grad_norm": 0.37022650241851807, "learning_rate": 6.026278994946948e-05, "loss": 0.0226, "step": 37000 }, { "grad_norm": 0.2826552093029022, "learning_rate": 6.02425607640287e-05, "loss": 0.0266, "step": 37010 }, { "grad_norm": 0.37426501512527466, "learning_rate": 6.022232982841441e-05, "loss": 0.0257, "step": 37020 }, { "grad_norm": 0.5872080326080322, "learning_rate": 6.020209714608355e-05, "loss": 0.0174, "step": 37030 }, { "grad_norm": 0.272018164396286, "learning_rate": 6.018186272049331e-05, "loss": 0.0288, "step": 37040 }, { "grad_norm": 0.3207513093948364, "learning_rate": 6.01616265551012e-05, "loss": 0.0337, "step": 37050 }, { "grad_norm": 0.34973251819610596, "learning_rate": 6.014138865336503e-05, "loss": 0.0254, "step": 37060 }, { "grad_norm": 0.2532137334346771, "learning_rate": 6.0121149018742905e-05, "loss": 0.0325, "step": 37070 }, { "grad_norm": 0.3014000654220581, "learning_rate": 6.010090765469325e-05, "loss": 0.0257, "step": 37080 }, { "grad_norm": 0.484541118144989, "learning_rate": 6.008066456467473e-05, "loss": 0.0241, "step": 37090 }, { "grad_norm": 0.300481379032135, "learning_rate": 6.0060419752146335e-05, "loss": 0.026, "step": 37100 }, { "grad_norm": 0.32869887351989746, "learning_rate": 6.0040173220567353e-05, "loss": 0.0248, "step": 37110 }, { "grad_norm": 0.2604449689388275, "learning_rate": 6.001992497339737e-05, "loss": 0.0259, "step": 37120 }, { "grad_norm": 0.29728928208351135, "learning_rate": 5.999967501409626e-05, "loss": 0.0237, "step": 37130 }, { "grad_norm": 0.20829246938228607, "learning_rate": 5.997942334612418e-05, "loss": 0.0211, "step": 37140 }, { "grad_norm": 0.2913860082626343, "learning_rate": 5.995916997294158e-05, "loss": 0.0243, "step": 37150 }, { "grad_norm": 0.30094560980796814, "learning_rate": 5.9938914898009214e-05, "loss": 0.0296, "step": 37160 }, { "grad_norm": 0.3135974109172821, "learning_rate": 5.991865812478813e-05, "loss": 0.019, "step": 37170 }, { "grad_norm": 0.507815420627594, "learning_rate": 5.989839965673964e-05, "loss": 0.0395, "step": 37180 }, { "grad_norm": 0.17588791251182556, "learning_rate": 5.987813949732539e-05, "loss": 0.0251, "step": 37190 }, { "grad_norm": 0.22777673602104187, "learning_rate": 5.9857877650007255e-05, "loss": 0.0196, "step": 37200 }, { "grad_norm": 0.36513838171958923, "learning_rate": 5.983761411824744e-05, "loss": 0.0235, "step": 37210 }, { "grad_norm": 0.38394084572792053, "learning_rate": 5.981734890550844e-05, "loss": 0.0297, "step": 37220 }, { "grad_norm": 0.4084254205226898, "learning_rate": 5.979708201525301e-05, "loss": 0.035, "step": 37230 }, { "grad_norm": 0.3919559717178345, "learning_rate": 5.977681345094422e-05, "loss": 0.028, "step": 37240 }, { "grad_norm": 0.2939518392086029, "learning_rate": 5.97565432160454e-05, "loss": 0.0361, "step": 37250 }, { "grad_norm": 0.45129185914993286, "learning_rate": 5.9736271314020186e-05, "loss": 0.0272, "step": 37260 }, { "grad_norm": 0.3174450993537903, "learning_rate": 5.971599774833251e-05, "loss": 0.0263, "step": 37270 }, { "grad_norm": 0.34599173069000244, "learning_rate": 5.9695722522446525e-05, "loss": 0.0323, "step": 37280 }, { "grad_norm": 1.3434603214263916, "learning_rate": 5.9675445639826765e-05, "loss": 0.0314, "step": 37290 }, { "grad_norm": 0.3020554184913635, "learning_rate": 5.965516710393796e-05, "loss": 0.0263, "step": 37300 }, { "grad_norm": 0.49536651372909546, "learning_rate": 5.963488691824516e-05, "loss": 0.0407, "step": 37310 }, { "grad_norm": 0.4534091353416443, "learning_rate": 5.96146050862137e-05, "loss": 0.0303, "step": 37320 }, { "grad_norm": 0.3653222918510437, "learning_rate": 5.959432161130919e-05, "loss": 0.0266, "step": 37330 }, { "grad_norm": 0.249502494931221, "learning_rate": 5.9574036496997545e-05, "loss": 0.0383, "step": 37340 }, { "grad_norm": 0.3176068663597107, "learning_rate": 5.955374974674488e-05, "loss": 0.0281, "step": 37350 }, { "grad_norm": 0.33789071440696716, "learning_rate": 5.9533461364017696e-05, "loss": 0.0236, "step": 37360 }, { "grad_norm": 0.28937265276908875, "learning_rate": 5.9513171352282716e-05, "loss": 0.0241, "step": 37370 }, { "grad_norm": 0.3627585768699646, "learning_rate": 5.949287971500692e-05, "loss": 0.0252, "step": 37380 }, { "grad_norm": 0.2526628077030182, "learning_rate": 5.947258645565762e-05, "loss": 0.021, "step": 37390 }, { "grad_norm": 0.24760660529136658, "learning_rate": 5.945229157770237e-05, "loss": 0.0181, "step": 37400 }, { "grad_norm": 0.5303140878677368, "learning_rate": 5.9431995084609006e-05, "loss": 0.0266, "step": 37410 }, { "grad_norm": 0.31369948387145996, "learning_rate": 5.941169697984564e-05, "loss": 0.0248, "step": 37420 }, { "grad_norm": 0.39691096544265747, "learning_rate": 5.9391397266880675e-05, "loss": 0.0257, "step": 37430 }, { "grad_norm": 0.38445785641670227, "learning_rate": 5.937109594918279e-05, "loss": 0.0343, "step": 37440 }, { "grad_norm": 0.3148820698261261, "learning_rate": 5.9350793030220884e-05, "loss": 0.0355, "step": 37450 }, { "grad_norm": 0.32942524552345276, "learning_rate": 5.933048851346421e-05, "loss": 0.0201, "step": 37460 }, { "grad_norm": 0.3281354010105133, "learning_rate": 5.931018240238224e-05, "loss": 0.0399, "step": 37470 }, { "grad_norm": 0.39795422554016113, "learning_rate": 5.928987470044471e-05, "loss": 0.0218, "step": 37480 }, { "grad_norm": 0.46631911396980286, "learning_rate": 5.9269565411121695e-05, "loss": 0.0252, "step": 37490 }, { "grad_norm": 0.31565770506858826, "learning_rate": 5.924925453788347e-05, "loss": 0.0255, "step": 37500 }, { "grad_norm": 0.30588194727897644, "learning_rate": 5.92289420842006e-05, "loss": 0.0329, "step": 37510 }, { "grad_norm": 0.47021761536598206, "learning_rate": 5.9208628053543945e-05, "loss": 0.0301, "step": 37520 }, { "grad_norm": 0.5061547756195068, "learning_rate": 5.918831244938462e-05, "loss": 0.0238, "step": 37530 }, { "grad_norm": 0.37935197353363037, "learning_rate": 5.9167995275194e-05, "loss": 0.0266, "step": 37540 }, { "grad_norm": 0.34239429235458374, "learning_rate": 5.914767653444373e-05, "loss": 0.0251, "step": 37550 }, { "grad_norm": 0.5297122001647949, "learning_rate": 5.912735623060572e-05, "loss": 0.0306, "step": 37560 }, { "grad_norm": 0.40930530428886414, "learning_rate": 5.910703436715217e-05, "loss": 0.0224, "step": 37570 }, { "grad_norm": 0.1803533136844635, "learning_rate": 5.908671094755552e-05, "loss": 0.023, "step": 37580 }, { "grad_norm": 0.3061126470565796, "learning_rate": 5.906638597528851e-05, "loss": 0.0267, "step": 37590 }, { "grad_norm": 0.2579355537891388, "learning_rate": 5.9046059453824076e-05, "loss": 0.0233, "step": 37600 }, { "grad_norm": 0.2816016674041748, "learning_rate": 5.9025731386635505e-05, "loss": 0.0204, "step": 37610 }, { "grad_norm": 0.3705773651599884, "learning_rate": 5.900540177719629e-05, "loss": 0.0227, "step": 37620 }, { "grad_norm": 0.4034065008163452, "learning_rate": 5.898507062898021e-05, "loss": 0.0264, "step": 37630 }, { "grad_norm": 0.25948938727378845, "learning_rate": 5.8964737945461316e-05, "loss": 0.0243, "step": 37640 }, { "grad_norm": 0.203206405043602, "learning_rate": 5.8944403730113885e-05, "loss": 0.025, "step": 37650 }, { "grad_norm": 0.45122864842414856, "learning_rate": 5.892406798641248e-05, "loss": 0.0252, "step": 37660 }, { "grad_norm": 0.3354063630104065, "learning_rate": 5.890373071783193e-05, "loss": 0.024, "step": 37670 }, { "grad_norm": 0.2616364657878876, "learning_rate": 5.888339192784732e-05, "loss": 0.0209, "step": 37680 }, { "grad_norm": 0.1765364706516266, "learning_rate": 5.8863051619934003e-05, "loss": 0.0249, "step": 37690 }, { "grad_norm": 0.3459242582321167, "learning_rate": 5.8842709797567554e-05, "loss": 0.0229, "step": 37700 }, { "grad_norm": 0.27356261014938354, "learning_rate": 5.8822366464223855e-05, "loss": 0.0188, "step": 37710 }, { "grad_norm": 0.2983633577823639, "learning_rate": 5.880202162337901e-05, "loss": 0.0285, "step": 37720 }, { "grad_norm": 0.3415180742740631, "learning_rate": 5.8781675278509405e-05, "loss": 0.0179, "step": 37730 }, { "grad_norm": 0.6918655633926392, "learning_rate": 5.8761327433091696e-05, "loss": 0.0239, "step": 37740 }, { "grad_norm": 0.4353291392326355, "learning_rate": 5.874097809060275e-05, "loss": 0.0321, "step": 37750 }, { "grad_norm": 0.42258405685424805, "learning_rate": 5.87206272545197e-05, "loss": 0.038, "step": 37760 }, { "grad_norm": 0.278231143951416, "learning_rate": 5.8700274928319955e-05, "loss": 0.0264, "step": 37770 }, { "grad_norm": 0.378447949886322, "learning_rate": 5.867992111548118e-05, "loss": 0.0249, "step": 37780 }, { "grad_norm": 0.3719976842403412, "learning_rate": 5.865956581948131e-05, "loss": 0.0223, "step": 37790 }, { "grad_norm": 0.36209115386009216, "learning_rate": 5.863920904379845e-05, "loss": 0.0223, "step": 37800 }, { "grad_norm": 0.33077937364578247, "learning_rate": 5.8618850791911064e-05, "loss": 0.0205, "step": 37810 }, { "grad_norm": 0.1900574266910553, "learning_rate": 5.859849106729779e-05, "loss": 0.0217, "step": 37820 }, { "grad_norm": 0.3074796199798584, "learning_rate": 5.857812987343758e-05, "loss": 0.0246, "step": 37830 }, { "grad_norm": 0.42290061712265015, "learning_rate": 5.855776721380957e-05, "loss": 0.0267, "step": 37840 }, { "grad_norm": 0.6907604336738586, "learning_rate": 5.8537403091893217e-05, "loss": 0.0253, "step": 37850 }, { "grad_norm": 0.32059481739997864, "learning_rate": 5.851703751116816e-05, "loss": 0.0242, "step": 37860 }, { "grad_norm": 0.3314518332481384, "learning_rate": 5.8496670475114336e-05, "loss": 0.0222, "step": 37870 }, { "grad_norm": 0.3358500003814697, "learning_rate": 5.84763019872119e-05, "loss": 0.0283, "step": 37880 }, { "grad_norm": 0.29197558760643005, "learning_rate": 5.845593205094131e-05, "loss": 0.0202, "step": 37890 }, { "grad_norm": 0.29118412733078003, "learning_rate": 5.843556066978318e-05, "loss": 0.0206, "step": 37900 }, { "grad_norm": 0.24511951208114624, "learning_rate": 5.8415187847218455e-05, "loss": 0.0253, "step": 37910 }, { "grad_norm": 0.3559531569480896, "learning_rate": 5.839481358672827e-05, "loss": 0.0254, "step": 37920 }, { "grad_norm": 0.4179244041442871, "learning_rate": 5.837443789179407e-05, "loss": 0.0247, "step": 37930 }, { "grad_norm": 0.4434855878353119, "learning_rate": 5.8354060765897445e-05, "loss": 0.0298, "step": 37940 }, { "grad_norm": 0.23004363477230072, "learning_rate": 5.8333682212520334e-05, "loss": 0.0255, "step": 37950 }, { "grad_norm": 0.46014049649238586, "learning_rate": 5.831330223514486e-05, "loss": 0.0277, "step": 37960 }, { "grad_norm": 0.24795155227184296, "learning_rate": 5.8292920837253396e-05, "loss": 0.0257, "step": 37970 }, { "grad_norm": 0.27421143651008606, "learning_rate": 5.827253802232857e-05, "loss": 0.0178, "step": 37980 }, { "grad_norm": 0.486259788274765, "learning_rate": 5.825215379385327e-05, "loss": 0.0216, "step": 37990 }, { "grad_norm": 0.3627333641052246, "learning_rate": 5.823176815531057e-05, "loss": 0.0433, "step": 38000 }, { "grad_norm": 0.23800335824489594, "learning_rate": 5.8211381110183826e-05, "loss": 0.0319, "step": 38010 }, { "grad_norm": 0.40418195724487305, "learning_rate": 5.8190992661956645e-05, "loss": 0.023, "step": 38020 }, { "grad_norm": 0.2978518605232239, "learning_rate": 5.817060281411284e-05, "loss": 0.0238, "step": 38030 }, { "grad_norm": 0.33097174763679504, "learning_rate": 5.815021157013647e-05, "loss": 0.0233, "step": 38040 }, { "grad_norm": 0.34801390767097473, "learning_rate": 5.8129818933511856e-05, "loss": 0.0324, "step": 38050 }, { "grad_norm": 0.18294952809810638, "learning_rate": 5.8109424907723544e-05, "loss": 0.027, "step": 38060 }, { "grad_norm": 0.3075748682022095, "learning_rate": 5.80890294962563e-05, "loss": 0.0231, "step": 38070 }, { "grad_norm": 0.3798944354057312, "learning_rate": 5.806863270259515e-05, "loss": 0.0284, "step": 38080 }, { "grad_norm": 0.304455041885376, "learning_rate": 5.804823453022536e-05, "loss": 0.0232, "step": 38090 }, { "grad_norm": 0.36053407192230225, "learning_rate": 5.80278349826324e-05, "loss": 0.0288, "step": 38100 }, { "grad_norm": 0.2918621599674225, "learning_rate": 5.8007434063302014e-05, "loss": 0.0244, "step": 38110 }, { "grad_norm": 0.3694441616535187, "learning_rate": 5.7987031775720136e-05, "loss": 0.0267, "step": 38120 }, { "grad_norm": 0.42821305990219116, "learning_rate": 5.7966628123372976e-05, "loss": 0.0295, "step": 38130 }, { "grad_norm": 0.5737369060516357, "learning_rate": 5.7946223109746956e-05, "loss": 0.0365, "step": 38140 }, { "grad_norm": 0.32410818338394165, "learning_rate": 5.7925816738328754e-05, "loss": 0.0229, "step": 38150 }, { "grad_norm": 0.4036501944065094, "learning_rate": 5.790540901260521e-05, "loss": 0.0304, "step": 38160 }, { "grad_norm": 0.3995535373687744, "learning_rate": 5.788499993606351e-05, "loss": 0.0261, "step": 38170 }, { "grad_norm": 0.32231447100639343, "learning_rate": 5.786458951219096e-05, "loss": 0.0286, "step": 38180 }, { "grad_norm": 0.30304256081581116, "learning_rate": 5.784417774447517e-05, "loss": 0.0335, "step": 38190 }, { "grad_norm": 0.28792449831962585, "learning_rate": 5.782376463640393e-05, "loss": 0.0274, "step": 38200 }, { "grad_norm": 0.3556525409221649, "learning_rate": 5.780335019146531e-05, "loss": 0.0276, "step": 38210 }, { "grad_norm": 0.32933056354522705, "learning_rate": 5.778293441314755e-05, "loss": 0.0247, "step": 38220 }, { "grad_norm": 0.3683755397796631, "learning_rate": 5.776251730493917e-05, "loss": 0.0276, "step": 38230 }, { "grad_norm": 0.2612221837043762, "learning_rate": 5.774209887032887e-05, "loss": 0.0274, "step": 38240 }, { "grad_norm": 0.4130256772041321, "learning_rate": 5.772167911280565e-05, "loss": 0.024, "step": 38250 }, { "grad_norm": 0.38410377502441406, "learning_rate": 5.770125803585864e-05, "loss": 0.0278, "step": 38260 }, { "grad_norm": 0.2956967353820801, "learning_rate": 5.768083564297726e-05, "loss": 0.0285, "step": 38270 }, { "grad_norm": 0.462889164686203, "learning_rate": 5.766041193765114e-05, "loss": 0.0301, "step": 38280 }, { "grad_norm": 0.31535860896110535, "learning_rate": 5.763998692337015e-05, "loss": 0.0203, "step": 38290 }, { "grad_norm": 0.3490214943885803, "learning_rate": 5.761956060362433e-05, "loss": 0.0261, "step": 38300 }, { "grad_norm": 0.3387981951236725, "learning_rate": 5.7599132981904e-05, "loss": 0.0283, "step": 38310 }, { "grad_norm": 0.32523593306541443, "learning_rate": 5.75787040616997e-05, "loss": 0.0298, "step": 38320 }, { "grad_norm": 0.3917085826396942, "learning_rate": 5.755827384650212e-05, "loss": 0.0287, "step": 38330 }, { "grad_norm": 0.36113476753234863, "learning_rate": 5.753784233980228e-05, "loss": 0.0294, "step": 38340 }, { "grad_norm": 0.23803703486919403, "learning_rate": 5.751740954509135e-05, "loss": 0.0199, "step": 38350 }, { "grad_norm": 0.16380895674228668, "learning_rate": 5.7496975465860715e-05, "loss": 0.0156, "step": 38360 }, { "grad_norm": 0.30657246708869934, "learning_rate": 5.747654010560202e-05, "loss": 0.021, "step": 38370 }, { "grad_norm": 0.2863733470439911, "learning_rate": 5.7456103467807097e-05, "loss": 0.0303, "step": 38380 }, { "grad_norm": 0.3608875274658203, "learning_rate": 5.7435665555968046e-05, "loss": 0.0195, "step": 38390 }, { "grad_norm": 0.29398465156555176, "learning_rate": 5.74152263735771e-05, "loss": 0.0294, "step": 38400 }, { "grad_norm": 0.3389441668987274, "learning_rate": 5.739478592412677e-05, "loss": 0.0198, "step": 38410 }, { "grad_norm": 0.5028718709945679, "learning_rate": 5.7374344211109766e-05, "loss": 0.0348, "step": 38420 }, { "grad_norm": 0.506659984588623, "learning_rate": 5.735390123801904e-05, "loss": 0.0283, "step": 38430 }, { "grad_norm": 0.5659242272377014, "learning_rate": 5.7333457008347704e-05, "loss": 0.0308, "step": 38440 }, { "grad_norm": 0.32000887393951416, "learning_rate": 5.7313011525589156e-05, "loss": 0.0292, "step": 38450 }, { "grad_norm": 0.2128814160823822, "learning_rate": 5.729256479323694e-05, "loss": 0.0252, "step": 38460 }, { "grad_norm": 0.4437682628631592, "learning_rate": 5.727211681478485e-05, "loss": 0.0251, "step": 38470 }, { "grad_norm": 0.2874296307563782, "learning_rate": 5.7251667593726886e-05, "loss": 0.0286, "step": 38480 }, { "grad_norm": 0.2531462609767914, "learning_rate": 5.723121713355728e-05, "loss": 0.0308, "step": 38490 }, { "grad_norm": 0.28601157665252686, "learning_rate": 5.721076543777044e-05, "loss": 0.0297, "step": 38500 }, { "grad_norm": 0.24895882606506348, "learning_rate": 5.7190312509860986e-05, "loss": 0.0226, "step": 38510 }, { "grad_norm": 0.21881063282489777, "learning_rate": 5.716985835332379e-05, "loss": 0.024, "step": 38520 }, { "grad_norm": 0.3164096176624298, "learning_rate": 5.714940297165389e-05, "loss": 0.024, "step": 38530 }, { "grad_norm": 0.26886963844299316, "learning_rate": 5.712894636834656e-05, "loss": 0.0275, "step": 38540 }, { "grad_norm": 0.33069875836372375, "learning_rate": 5.7108488546897276e-05, "loss": 0.0215, "step": 38550 }, { "grad_norm": 0.24367953836917877, "learning_rate": 5.708802951080172e-05, "loss": 0.0207, "step": 38560 }, { "grad_norm": 0.4378907382488251, "learning_rate": 5.706756926355576e-05, "loss": 0.0284, "step": 38570 }, { "grad_norm": 0.2879047393798828, "learning_rate": 5.704710780865554e-05, "loss": 0.0314, "step": 38580 }, { "grad_norm": 0.3641643524169922, "learning_rate": 5.7026645149597325e-05, "loss": 0.0254, "step": 38590 }, { "grad_norm": 0.47818833589553833, "learning_rate": 5.700618128987764e-05, "loss": 0.0343, "step": 38600 }, { "grad_norm": 0.23531489074230194, "learning_rate": 5.698571623299317e-05, "loss": 0.0185, "step": 38610 }, { "grad_norm": 0.2703884541988373, "learning_rate": 5.696524998244086e-05, "loss": 0.0214, "step": 38620 }, { "grad_norm": 0.2492457628250122, "learning_rate": 5.6944782541717836e-05, "loss": 0.0307, "step": 38630 }, { "grad_norm": 0.2704962193965912, "learning_rate": 5.69243139143214e-05, "loss": 0.0203, "step": 38640 }, { "grad_norm": 0.2914630174636841, "learning_rate": 5.6903844103749125e-05, "loss": 0.0216, "step": 38650 }, { "grad_norm": 0.31514719128608704, "learning_rate": 5.688337311349869e-05, "loss": 0.0241, "step": 38660 }, { "grad_norm": 0.37719330191612244, "learning_rate": 5.6862900947068074e-05, "loss": 0.0256, "step": 38670 }, { "grad_norm": 0.38582244515419006, "learning_rate": 5.6842427607955374e-05, "loss": 0.0289, "step": 38680 }, { "grad_norm": 0.2650983929634094, "learning_rate": 5.682195309965893e-05, "loss": 0.0203, "step": 38690 }, { "grad_norm": 0.3862419128417969, "learning_rate": 5.6801477425677294e-05, "loss": 0.0296, "step": 38700 }, { "grad_norm": 0.3787176311016083, "learning_rate": 5.678100058950917e-05, "loss": 0.0318, "step": 38710 }, { "grad_norm": 0.4673768877983093, "learning_rate": 5.676052259465352e-05, "loss": 0.0295, "step": 38720 }, { "grad_norm": 0.6174710392951965, "learning_rate": 5.674004344460945e-05, "loss": 0.0247, "step": 38730 }, { "grad_norm": 0.234082892537117, "learning_rate": 5.6719563142876295e-05, "loss": 0.0219, "step": 38740 }, { "grad_norm": 0.26777032017707825, "learning_rate": 5.669908169295359e-05, "loss": 0.023, "step": 38750 }, { "grad_norm": 0.4712390899658203, "learning_rate": 5.667859909834105e-05, "loss": 0.0267, "step": 38760 }, { "grad_norm": 0.30618271231651306, "learning_rate": 5.6658115362538565e-05, "loss": 0.0242, "step": 38770 }, { "grad_norm": 0.3843710422515869, "learning_rate": 5.663763048904628e-05, "loss": 0.0209, "step": 38780 }, { "grad_norm": 0.32798218727111816, "learning_rate": 5.661714448136447e-05, "loss": 0.0203, "step": 38790 }, { "grad_norm": 0.23153135180473328, "learning_rate": 5.659665734299366e-05, "loss": 0.021, "step": 38800 }, { "grad_norm": 0.3016614019870758, "learning_rate": 5.6576169077434516e-05, "loss": 0.0257, "step": 38810 }, { "grad_norm": 0.29608961939811707, "learning_rate": 5.6555679688187944e-05, "loss": 0.0282, "step": 38820 }, { "grad_norm": 0.5222342014312744, "learning_rate": 5.6535189178755e-05, "loss": 0.0299, "step": 38830 }, { "grad_norm": 0.2954566478729248, "learning_rate": 5.651469755263698e-05, "loss": 0.0299, "step": 38840 }, { "grad_norm": 0.1831355094909668, "learning_rate": 5.6494204813335316e-05, "loss": 0.032, "step": 38850 }, { "grad_norm": 0.28314682841300964, "learning_rate": 5.647371096435168e-05, "loss": 0.0196, "step": 38860 }, { "grad_norm": 0.30998918414115906, "learning_rate": 5.645321600918788e-05, "loss": 0.0163, "step": 38870 }, { "grad_norm": 0.35423558950424194, "learning_rate": 5.643271995134597e-05, "loss": 0.0341, "step": 38880 }, { "grad_norm": 0.2848629951477051, "learning_rate": 5.641222279432814e-05, "loss": 0.0239, "step": 38890 }, { "grad_norm": 0.388150691986084, "learning_rate": 5.6391724541636834e-05, "loss": 0.0203, "step": 38900 }, { "grad_norm": 0.36983561515808105, "learning_rate": 5.6371225196774605e-05, "loss": 0.029, "step": 38910 }, { "grad_norm": 0.3641084134578705, "learning_rate": 5.635072476324423e-05, "loss": 0.0307, "step": 38920 }, { "grad_norm": 0.3181314170360565, "learning_rate": 5.63302232445487e-05, "loss": 0.0205, "step": 38930 }, { "grad_norm": 0.2565745711326599, "learning_rate": 5.6309720644191144e-05, "loss": 0.0308, "step": 38940 }, { "grad_norm": 0.42504626512527466, "learning_rate": 5.628921696567491e-05, "loss": 0.0257, "step": 38950 }, { "grad_norm": 0.269197553396225, "learning_rate": 5.62687122125035e-05, "loss": 0.0268, "step": 38960 }, { "grad_norm": 0.3809939920902252, "learning_rate": 5.624820638818062e-05, "loss": 0.0263, "step": 38970 }, { "grad_norm": 0.3677290976047516, "learning_rate": 5.6227699496210164e-05, "loss": 0.0381, "step": 38980 }, { "grad_norm": 0.35748812556266785, "learning_rate": 5.6207191540096195e-05, "loss": 0.0287, "step": 38990 }, { "grad_norm": 0.356854647397995, "learning_rate": 5.618668252334296e-05, "loss": 0.0287, "step": 39000 }, { "grad_norm": 0.38487550616264343, "learning_rate": 5.616617244945488e-05, "loss": 0.0247, "step": 39010 }, { "grad_norm": 0.2577574849128723, "learning_rate": 5.614566132193656e-05, "loss": 0.022, "step": 39020 }, { "grad_norm": 0.3062780499458313, "learning_rate": 5.612514914429282e-05, "loss": 0.032, "step": 39030 }, { "grad_norm": 0.38672152161598206, "learning_rate": 5.610463592002863e-05, "loss": 0.0245, "step": 39040 }, { "grad_norm": 0.41724589467048645, "learning_rate": 5.608412165264909e-05, "loss": 0.0254, "step": 39050 }, { "grad_norm": 0.24614636600017548, "learning_rate": 5.606360634565959e-05, "loss": 0.0189, "step": 39060 }, { "grad_norm": 0.26072171330451965, "learning_rate": 5.604309000256559e-05, "loss": 0.0228, "step": 39070 }, { "grad_norm": 0.3357182741165161, "learning_rate": 5.6022572626872785e-05, "loss": 0.0211, "step": 39080 }, { "grad_norm": 0.24262544512748718, "learning_rate": 5.600205422208704e-05, "loss": 0.023, "step": 39090 }, { "grad_norm": 0.338894784450531, "learning_rate": 5.59815347917144e-05, "loss": 0.0283, "step": 39100 }, { "grad_norm": 0.4033489227294922, "learning_rate": 5.596101433926103e-05, "loss": 0.0224, "step": 39110 }, { "grad_norm": 0.3439798057079315, "learning_rate": 5.5940492868233364e-05, "loss": 0.0173, "step": 39120 }, { "grad_norm": 0.2891896963119507, "learning_rate": 5.591997038213793e-05, "loss": 0.0258, "step": 39130 }, { "grad_norm": 0.42899248003959656, "learning_rate": 5.5899446884481475e-05, "loss": 0.0233, "step": 39140 }, { "grad_norm": 0.3385428786277771, "learning_rate": 5.5878922378770906e-05, "loss": 0.0368, "step": 39150 }, { "grad_norm": 0.40439194440841675, "learning_rate": 5.5858396868513285e-05, "loss": 0.0282, "step": 39160 }, { "grad_norm": 0.29089635610580444, "learning_rate": 5.583787035721586e-05, "loss": 0.0208, "step": 39170 }, { "grad_norm": 0.27795958518981934, "learning_rate": 5.581734284838606e-05, "loss": 0.0215, "step": 39180 }, { "grad_norm": 0.4485384523868561, "learning_rate": 5.579681434553147e-05, "loss": 0.0255, "step": 39190 }, { "grad_norm": 0.21293824911117554, "learning_rate": 5.5776284852159854e-05, "loss": 0.0182, "step": 39200 }, { "grad_norm": 0.3552558422088623, "learning_rate": 5.575575437177913e-05, "loss": 0.0272, "step": 39210 }, { "grad_norm": 0.2961656153202057, "learning_rate": 5.573522290789742e-05, "loss": 0.0205, "step": 39220 }, { "grad_norm": 0.537380576133728, "learning_rate": 5.571469046402298e-05, "loss": 0.0275, "step": 39230 }, { "grad_norm": 0.22785471379756927, "learning_rate": 5.5694157043664205e-05, "loss": 0.0258, "step": 39240 }, { "grad_norm": 0.19074270129203796, "learning_rate": 5.567362265032975e-05, "loss": 0.0147, "step": 39250 }, { "grad_norm": 0.23537616431713104, "learning_rate": 5.565308728752836e-05, "loss": 0.0215, "step": 39260 }, { "grad_norm": 0.5645987391471863, "learning_rate": 5.5632550958768956e-05, "loss": 0.0221, "step": 39270 }, { "grad_norm": 0.47095540165901184, "learning_rate": 5.5612013667560636e-05, "loss": 0.0246, "step": 39280 }, { "grad_norm": 0.337556391954422, "learning_rate": 5.5591475417412676e-05, "loss": 0.0366, "step": 39290 }, { "grad_norm": 0.35933399200439453, "learning_rate": 5.557093621183451e-05, "loss": 0.0265, "step": 39300 }, { "grad_norm": 0.237836554646492, "learning_rate": 5.55503960543357e-05, "loss": 0.0239, "step": 39310 }, { "grad_norm": 0.3819912075996399, "learning_rate": 5.552985494842601e-05, "loss": 0.0281, "step": 39320 }, { "grad_norm": 0.2947006821632385, "learning_rate": 5.550931289761534e-05, "loss": 0.0233, "step": 39330 }, { "grad_norm": 0.4289660155773163, "learning_rate": 5.548876990541378e-05, "loss": 0.0223, "step": 39340 }, { "grad_norm": 0.46778857707977295, "learning_rate": 5.5468225975331564e-05, "loss": 0.0281, "step": 39350 }, { "grad_norm": 0.3557933270931244, "learning_rate": 5.544768111087909e-05, "loss": 0.0242, "step": 39360 }, { "grad_norm": 0.24844959378242493, "learning_rate": 5.54271353155669e-05, "loss": 0.0283, "step": 39370 }, { "grad_norm": 0.31068018078804016, "learning_rate": 5.5406588592905715e-05, "loss": 0.025, "step": 39380 }, { "grad_norm": 0.23664239048957825, "learning_rate": 5.5386040946406416e-05, "loss": 0.0273, "step": 39390 }, { "grad_norm": 0.37997305393218994, "learning_rate": 5.536549237958004e-05, "loss": 0.027, "step": 39400 }, { "grad_norm": 0.2908143103122711, "learning_rate": 5.5344942895937744e-05, "loss": 0.0363, "step": 39410 }, { "grad_norm": 0.2651037871837616, "learning_rate": 5.5324392498990904e-05, "loss": 0.0195, "step": 39420 }, { "grad_norm": 0.4625116288661957, "learning_rate": 5.5303841192251e-05, "loss": 0.0223, "step": 39430 }, { "grad_norm": 0.3619956374168396, "learning_rate": 5.52832889792297e-05, "loss": 0.0276, "step": 39440 }, { "grad_norm": 0.2762981355190277, "learning_rate": 5.526273586343881e-05, "loss": 0.0244, "step": 39450 }, { "grad_norm": 0.385276198387146, "learning_rate": 5.5242181848390306e-05, "loss": 0.0221, "step": 39460 }, { "grad_norm": 0.3482774496078491, "learning_rate": 5.5221626937596285e-05, "loss": 0.0175, "step": 39470 }, { "grad_norm": 0.49491095542907715, "learning_rate": 5.520107113456903e-05, "loss": 0.03, "step": 39480 }, { "grad_norm": 0.5097066760063171, "learning_rate": 5.5180514442820974e-05, "loss": 0.0376, "step": 39490 }, { "grad_norm": 0.4165729880332947, "learning_rate": 5.515995686586469e-05, "loss": 0.0207, "step": 39500 }, { "grad_norm": 0.30752843618392944, "learning_rate": 5.5139398407212916e-05, "loss": 0.0193, "step": 39510 }, { "grad_norm": 0.3198004961013794, "learning_rate": 5.511883907037849e-05, "loss": 0.0242, "step": 39520 }, { "grad_norm": 0.4059794843196869, "learning_rate": 5.509827885887449e-05, "loss": 0.0261, "step": 39530 }, { "grad_norm": 0.37995287775993347, "learning_rate": 5.507771777621406e-05, "loss": 0.0205, "step": 39540 }, { "grad_norm": 0.17342740297317505, "learning_rate": 5.505715582591052e-05, "loss": 0.0315, "step": 39550 }, { "grad_norm": 0.34048861265182495, "learning_rate": 5.50365930114774e-05, "loss": 0.0286, "step": 39560 }, { "grad_norm": 0.32797688245773315, "learning_rate": 5.5016029336428255e-05, "loss": 0.0207, "step": 39570 }, { "grad_norm": 0.3623182773590088, "learning_rate": 5.49954648042769e-05, "loss": 0.0259, "step": 39580 }, { "grad_norm": 0.40556201338768005, "learning_rate": 5.4974899418537226e-05, "loss": 0.0329, "step": 39590 }, { "grad_norm": 0.47957926988601685, "learning_rate": 5.4954333182723316e-05, "loss": 0.0245, "step": 39600 }, { "grad_norm": 0.47470057010650635, "learning_rate": 5.493376610034937e-05, "loss": 0.0267, "step": 39610 }, { "grad_norm": 0.30132588744163513, "learning_rate": 5.4913198174929735e-05, "loss": 0.0229, "step": 39620 }, { "grad_norm": 0.29413464665412903, "learning_rate": 5.489262940997889e-05, "loss": 0.0283, "step": 39630 }, { "grad_norm": 0.20945297181606293, "learning_rate": 5.487205980901151e-05, "loss": 0.0263, "step": 39640 }, { "grad_norm": 0.4958445727825165, "learning_rate": 5.485148937554234e-05, "loss": 0.0437, "step": 39650 }, { "grad_norm": 0.24833548069000244, "learning_rate": 5.483091811308635e-05, "loss": 0.0246, "step": 39660 }, { "grad_norm": 0.3342766761779785, "learning_rate": 5.4810346025158564e-05, "loss": 0.025, "step": 39670 }, { "grad_norm": 0.29681649804115295, "learning_rate": 5.478977311527421e-05, "loss": 0.0257, "step": 39680 }, { "grad_norm": 0.2751859426498413, "learning_rate": 5.476919938694863e-05, "loss": 0.025, "step": 39690 }, { "grad_norm": 0.25815626978874207, "learning_rate": 5.474862484369733e-05, "loss": 0.019, "step": 39700 }, { "grad_norm": 0.21320632100105286, "learning_rate": 5.472804948903589e-05, "loss": 0.0237, "step": 39710 }, { "grad_norm": 0.2890036106109619, "learning_rate": 5.470747332648013e-05, "loss": 0.0209, "step": 39720 }, { "grad_norm": 0.3292621672153473, "learning_rate": 5.468689635954591e-05, "loss": 0.0371, "step": 39730 }, { "grad_norm": 0.19289736449718475, "learning_rate": 5.46663185917493e-05, "loss": 0.0234, "step": 39740 }, { "grad_norm": 0.3019534945487976, "learning_rate": 5.464574002660645e-05, "loss": 0.0226, "step": 39750 }, { "grad_norm": 0.17094939947128296, "learning_rate": 5.4625160667633724e-05, "loss": 0.0261, "step": 39760 }, { "grad_norm": 0.23365069925785065, "learning_rate": 5.4604580518347505e-05, "loss": 0.0289, "step": 39770 }, { "grad_norm": 0.34682607650756836, "learning_rate": 5.458399958226442e-05, "loss": 0.0266, "step": 39780 }, { "grad_norm": 0.17223860323429108, "learning_rate": 5.456341786290119e-05, "loss": 0.0239, "step": 39790 }, { "grad_norm": 0.36308231949806213, "learning_rate": 5.454283536377465e-05, "loss": 0.0313, "step": 39800 }, { "grad_norm": 0.48171281814575195, "learning_rate": 5.452225208840179e-05, "loss": 0.0336, "step": 39810 }, { "grad_norm": 0.41206425428390503, "learning_rate": 5.450166804029975e-05, "loss": 0.035, "step": 39820 }, { "grad_norm": 0.4146319329738617, "learning_rate": 5.448108322298574e-05, "loss": 0.0228, "step": 39830 }, { "grad_norm": 0.20941835641860962, "learning_rate": 5.446049763997717e-05, "loss": 0.0213, "step": 39840 }, { "grad_norm": 0.39519360661506653, "learning_rate": 5.4439911294791546e-05, "loss": 0.0272, "step": 39850 }, { "grad_norm": 0.29191774129867554, "learning_rate": 5.441932419094652e-05, "loss": 0.0239, "step": 39860 }, { "grad_norm": 0.19471685588359833, "learning_rate": 5.439873633195985e-05, "loss": 0.0195, "step": 39870 }, { "grad_norm": 0.25543639063835144, "learning_rate": 5.437814772134947e-05, "loss": 0.0291, "step": 39880 }, { "grad_norm": 0.3247123062610626, "learning_rate": 5.4357558362633366e-05, "loss": 0.0231, "step": 39890 }, { "grad_norm": 0.18370765447616577, "learning_rate": 5.4336968259329726e-05, "loss": 0.0183, "step": 39900 }, { "grad_norm": 0.2317463606595993, "learning_rate": 5.431637741495681e-05, "loss": 0.0282, "step": 39910 }, { "grad_norm": 0.5704128742218018, "learning_rate": 5.429578583303307e-05, "loss": 0.0278, "step": 39920 }, { "grad_norm": 0.24798229336738586, "learning_rate": 5.427519351707701e-05, "loss": 0.0247, "step": 39930 }, { "grad_norm": 0.17102494835853577, "learning_rate": 5.42546004706073e-05, "loss": 0.0197, "step": 39940 }, { "grad_norm": 0.32016605138778687, "learning_rate": 5.4234006697142735e-05, "loss": 0.0227, "step": 39950 }, { "grad_norm": 0.19772252440452576, "learning_rate": 5.421341220020224e-05, "loss": 0.0229, "step": 39960 }, { "grad_norm": 0.30263227224349976, "learning_rate": 5.419281698330482e-05, "loss": 0.0246, "step": 39970 }, { "grad_norm": 0.17571009695529938, "learning_rate": 5.4172221049969665e-05, "loss": 0.0214, "step": 39980 }, { "grad_norm": 0.4221314787864685, "learning_rate": 5.415162440371604e-05, "loss": 0.0219, "step": 39990 }, { "grad_norm": 0.3331620395183563, "learning_rate": 5.413102704806334e-05, "loss": 0.027, "step": 40000 }, { "grad_norm": 0.27931535243988037, "learning_rate": 5.41104289865311e-05, "loss": 0.0255, "step": 40010 }, { "grad_norm": 0.3012451231479645, "learning_rate": 5.408983022263898e-05, "loss": 0.0202, "step": 40020 }, { "grad_norm": 0.31054526567459106, "learning_rate": 5.406923075990673e-05, "loss": 0.0288, "step": 40030 }, { "grad_norm": 0.21515440940856934, "learning_rate": 5.404863060185423e-05, "loss": 0.0186, "step": 40040 }, { "grad_norm": 0.2786232829093933, "learning_rate": 5.40280297520015e-05, "loss": 0.0306, "step": 40050 }, { "grad_norm": 0.298629492521286, "learning_rate": 5.400742821386865e-05, "loss": 0.0184, "step": 40060 }, { "grad_norm": 0.30651772022247314, "learning_rate": 5.398682599097592e-05, "loss": 0.0257, "step": 40070 }, { "grad_norm": 0.2962241768836975, "learning_rate": 5.396622308684367e-05, "loss": 0.0257, "step": 40080 }, { "grad_norm": 0.46362319588661194, "learning_rate": 5.394561950499236e-05, "loss": 0.0307, "step": 40090 }, { "grad_norm": 0.2970355153083801, "learning_rate": 5.39250152489426e-05, "loss": 0.0288, "step": 40100 }, { "grad_norm": 0.2800734341144562, "learning_rate": 5.390441032221507e-05, "loss": 0.0203, "step": 40110 }, { "grad_norm": 0.35439327359199524, "learning_rate": 5.388380472833062e-05, "loss": 0.0268, "step": 40120 }, { "grad_norm": 0.22317466139793396, "learning_rate": 5.386319847081016e-05, "loss": 0.0212, "step": 40130 }, { "grad_norm": 0.3842763602733612, "learning_rate": 5.384259155317473e-05, "loss": 0.0315, "step": 40140 }, { "grad_norm": 0.4157813489437103, "learning_rate": 5.38219839789455e-05, "loss": 0.0287, "step": 40150 }, { "grad_norm": 0.3385896682739258, "learning_rate": 5.380137575164376e-05, "loss": 0.0368, "step": 40160 }, { "grad_norm": 0.23189899325370789, "learning_rate": 5.378076687479085e-05, "loss": 0.0232, "step": 40170 }, { "grad_norm": 0.22562149167060852, "learning_rate": 5.3760157351908305e-05, "loss": 0.0205, "step": 40180 }, { "grad_norm": 0.35533028841018677, "learning_rate": 5.373954718651768e-05, "loss": 0.0307, "step": 40190 }, { "grad_norm": 0.23397085070610046, "learning_rate": 5.371893638214074e-05, "loss": 0.0195, "step": 40200 }, { "grad_norm": 0.3446224331855774, "learning_rate": 5.369832494229927e-05, "loss": 0.0207, "step": 40210 }, { "grad_norm": 0.386707067489624, "learning_rate": 5.367771287051524e-05, "loss": 0.0255, "step": 40220 }, { "grad_norm": 0.17806516587734222, "learning_rate": 5.365710017031065e-05, "loss": 0.0194, "step": 40230 }, { "grad_norm": 0.3840942978858948, "learning_rate": 5.3636486845207654e-05, "loss": 0.0329, "step": 40240 }, { "grad_norm": 0.3225451409816742, "learning_rate": 5.3615872898728526e-05, "loss": 0.0213, "step": 40250 }, { "grad_norm": 0.4270770251750946, "learning_rate": 5.3595258334395614e-05, "loss": 0.0265, "step": 40260 }, { "grad_norm": 0.3145243525505066, "learning_rate": 5.3574643155731396e-05, "loss": 0.0314, "step": 40270 }, { "grad_norm": 0.2831507921218872, "learning_rate": 5.3554027366258405e-05, "loss": 0.0214, "step": 40280 }, { "grad_norm": 0.3875751197338104, "learning_rate": 5.3533410969499355e-05, "loss": 0.0252, "step": 40290 }, { "grad_norm": 0.20617195963859558, "learning_rate": 5.3512793968977e-05, "loss": 0.0222, "step": 40300 }, { "grad_norm": 0.43991392850875854, "learning_rate": 5.349217636821423e-05, "loss": 0.0261, "step": 40310 }, { "grad_norm": 0.34951165318489075, "learning_rate": 5.347155817073405e-05, "loss": 0.0205, "step": 40320 }, { "grad_norm": 0.266879677772522, "learning_rate": 5.34509393800595e-05, "loss": 0.0216, "step": 40330 }, { "grad_norm": 0.2530824840068817, "learning_rate": 5.343031999971381e-05, "loss": 0.0206, "step": 40340 }, { "grad_norm": 0.24282710254192352, "learning_rate": 5.340970003322026e-05, "loss": 0.0281, "step": 40350 }, { "grad_norm": 0.3132883310317993, "learning_rate": 5.33890794841022e-05, "loss": 0.0203, "step": 40360 }, { "grad_norm": 0.25942692160606384, "learning_rate": 5.336845835588318e-05, "loss": 0.0217, "step": 40370 }, { "grad_norm": 0.4035344123840332, "learning_rate": 5.334783665208674e-05, "loss": 0.0258, "step": 40380 }, { "grad_norm": 0.3274015486240387, "learning_rate": 5.332721437623657e-05, "loss": 0.0209, "step": 40390 }, { "grad_norm": 0.22409026324748993, "learning_rate": 5.3306591531856464e-05, "loss": 0.0252, "step": 40400 }, { "grad_norm": 0.24042004346847534, "learning_rate": 5.3285968122470295e-05, "loss": 0.0182, "step": 40410 }, { "grad_norm": 0.2723466753959656, "learning_rate": 5.326534415160207e-05, "loss": 0.0283, "step": 40420 }, { "grad_norm": 0.1822127401828766, "learning_rate": 5.324471962277582e-05, "loss": 0.0315, "step": 40430 }, { "grad_norm": 0.2981732189655304, "learning_rate": 5.3224094539515746e-05, "loss": 0.0244, "step": 40440 }, { "grad_norm": 0.2655649483203888, "learning_rate": 5.3203468905346076e-05, "loss": 0.0196, "step": 40450 }, { "grad_norm": 0.39033910632133484, "learning_rate": 5.3182842723791195e-05, "loss": 0.0274, "step": 40460 }, { "grad_norm": 0.34030821919441223, "learning_rate": 5.316221599837554e-05, "loss": 0.0236, "step": 40470 }, { "grad_norm": 0.30815190076828003, "learning_rate": 5.314158873262366e-05, "loss": 0.0293, "step": 40480 }, { "grad_norm": 0.1963586062192917, "learning_rate": 5.312096093006018e-05, "loss": 0.0175, "step": 40490 }, { "grad_norm": 0.3537607491016388, "learning_rate": 5.3100332594209846e-05, "loss": 0.0224, "step": 40500 }, { "grad_norm": 0.32262569665908813, "learning_rate": 5.3079703728597454e-05, "loss": 0.0255, "step": 40510 }, { "grad_norm": 0.3074199855327606, "learning_rate": 5.305907433674794e-05, "loss": 0.023, "step": 40520 }, { "grad_norm": 0.2496064156293869, "learning_rate": 5.3038444422186284e-05, "loss": 0.0225, "step": 40530 }, { "grad_norm": 0.2847975194454193, "learning_rate": 5.301781398843757e-05, "loss": 0.0167, "step": 40540 }, { "grad_norm": 0.30381298065185547, "learning_rate": 5.299718303902699e-05, "loss": 0.0279, "step": 40550 }, { "grad_norm": 0.2067761868238449, "learning_rate": 5.297655157747979e-05, "loss": 0.0275, "step": 40560 }, { "grad_norm": 0.3003930151462555, "learning_rate": 5.295591960732136e-05, "loss": 0.0212, "step": 40570 }, { "grad_norm": 0.21861879527568817, "learning_rate": 5.293528713207708e-05, "loss": 0.0233, "step": 40580 }, { "grad_norm": 0.23899328708648682, "learning_rate": 5.291465415527253e-05, "loss": 0.0305, "step": 40590 }, { "grad_norm": 0.5876684784889221, "learning_rate": 5.2894020680433296e-05, "loss": 0.0213, "step": 40600 }, { "grad_norm": 0.25664880871772766, "learning_rate": 5.287338671108507e-05, "loss": 0.0204, "step": 40610 }, { "grad_norm": 0.37536072731018066, "learning_rate": 5.285275225075367e-05, "loss": 0.0239, "step": 40620 }, { "grad_norm": 0.2807018756866455, "learning_rate": 5.283211730296492e-05, "loss": 0.022, "step": 40630 }, { "grad_norm": 0.38240721821784973, "learning_rate": 5.281148187124477e-05, "loss": 0.0233, "step": 40640 }, { "grad_norm": 0.27730005979537964, "learning_rate": 5.279084595911927e-05, "loss": 0.0185, "step": 40650 }, { "grad_norm": 0.49518898129463196, "learning_rate": 5.277020957011453e-05, "loss": 0.0346, "step": 40660 }, { "grad_norm": 0.42408987879753113, "learning_rate": 5.274957270775673e-05, "loss": 0.0278, "step": 40670 }, { "grad_norm": 0.19702845811843872, "learning_rate": 5.2728935375572164e-05, "loss": 0.0177, "step": 40680 }, { "grad_norm": 0.3049654960632324, "learning_rate": 5.2708297577087176e-05, "loss": 0.0239, "step": 40690 }, { "grad_norm": 0.4665529429912567, "learning_rate": 5.2687659315828196e-05, "loss": 0.0217, "step": 40700 }, { "grad_norm": 0.30034777522087097, "learning_rate": 5.266702059532175e-05, "loss": 0.0232, "step": 40710 }, { "grad_norm": 0.3414801061153412, "learning_rate": 5.264638141909444e-05, "loss": 0.0268, "step": 40720 }, { "grad_norm": 0.1985570192337036, "learning_rate": 5.2625741790672925e-05, "loss": 0.0248, "step": 40730 }, { "grad_norm": 0.21042786538600922, "learning_rate": 5.2605101713583935e-05, "loss": 0.0215, "step": 40740 }, { "grad_norm": 0.45709073543548584, "learning_rate": 5.2584461191354315e-05, "loss": 0.0235, "step": 40750 }, { "grad_norm": 0.365816205739975, "learning_rate": 5.2563820227510964e-05, "loss": 0.0162, "step": 40760 }, { "grad_norm": 0.2405107319355011, "learning_rate": 5.2543178825580864e-05, "loss": 0.0258, "step": 40770 }, { "grad_norm": 0.4823632538318634, "learning_rate": 5.2522536989091054e-05, "loss": 0.0295, "step": 40780 }, { "grad_norm": 0.4051269590854645, "learning_rate": 5.2501894721568655e-05, "loss": 0.0212, "step": 40790 }, { "grad_norm": 0.28447088599205017, "learning_rate": 5.248125202654089e-05, "loss": 0.0199, "step": 40800 }, { "grad_norm": 0.45389923453330994, "learning_rate": 5.246060890753501e-05, "loss": 0.0349, "step": 40810 }, { "grad_norm": 0.28459176421165466, "learning_rate": 5.243996536807837e-05, "loss": 0.0205, "step": 40820 }, { "grad_norm": 0.30095306038856506, "learning_rate": 5.2419321411698384e-05, "loss": 0.0227, "step": 40830 }, { "grad_norm": 0.38341495394706726, "learning_rate": 5.239867704192253e-05, "loss": 0.0237, "step": 40840 }, { "grad_norm": 0.16800761222839355, "learning_rate": 5.237803226227838e-05, "loss": 0.0175, "step": 40850 }, { "grad_norm": 0.21761925518512726, "learning_rate": 5.235738707629354e-05, "loss": 0.0196, "step": 40860 }, { "grad_norm": 0.2890659272670746, "learning_rate": 5.233674148749575e-05, "loss": 0.0252, "step": 40870 }, { "grad_norm": 0.5177103877067566, "learning_rate": 5.231609549941272e-05, "loss": 0.037, "step": 40880 }, { "grad_norm": 0.3043244183063507, "learning_rate": 5.229544911557233e-05, "loss": 0.03, "step": 40890 }, { "grad_norm": 0.4382261335849762, "learning_rate": 5.227480233950246e-05, "loss": 0.0252, "step": 40900 }, { "grad_norm": 0.3023073673248291, "learning_rate": 5.22541551747311e-05, "loss": 0.0238, "step": 40910 }, { "grad_norm": 0.406731516122818, "learning_rate": 5.223350762478626e-05, "loss": 0.0193, "step": 40920 }, { "grad_norm": 0.19894076883792877, "learning_rate": 5.221285969319606e-05, "loss": 0.0212, "step": 40930 }, { "grad_norm": 0.14993077516555786, "learning_rate": 5.219221138348865e-05, "loss": 0.0212, "step": 40940 }, { "grad_norm": 0.30432575941085815, "learning_rate": 5.217156269919228e-05, "loss": 0.0233, "step": 40950 }, { "grad_norm": 0.1947956383228302, "learning_rate": 5.215091364383523e-05, "loss": 0.0211, "step": 40960 }, { "grad_norm": 0.45402178168296814, "learning_rate": 5.213026422094588e-05, "loss": 0.0307, "step": 40970 }, { "grad_norm": 0.3000980019569397, "learning_rate": 5.210961443405262e-05, "loss": 0.0254, "step": 40980 }, { "grad_norm": 0.3212009370326996, "learning_rate": 5.208896428668396e-05, "loss": 0.0184, "step": 40990 }, { "grad_norm": 0.4314519166946411, "learning_rate": 5.206831378236845e-05, "loss": 0.0295, "step": 41000 }, { "grad_norm": 0.23495657742023468, "learning_rate": 5.2047662924634666e-05, "loss": 0.0201, "step": 41010 }, { "grad_norm": 0.32595521211624146, "learning_rate": 5.20270117170113e-05, "loss": 0.0221, "step": 41020 }, { "grad_norm": 0.4358173906803131, "learning_rate": 5.200636016302707e-05, "loss": 0.0337, "step": 41030 }, { "grad_norm": 0.31554967164993286, "learning_rate": 5.1985708266210754e-05, "loss": 0.0278, "step": 41040 }, { "grad_norm": 0.38842275738716125, "learning_rate": 5.1965056030091206e-05, "loss": 0.0279, "step": 41050 }, { "grad_norm": 0.2615293860435486, "learning_rate": 5.194440345819732e-05, "loss": 0.0209, "step": 41060 }, { "grad_norm": 0.3677246868610382, "learning_rate": 5.1923750554058084e-05, "loss": 0.0242, "step": 41070 }, { "grad_norm": 0.37569382786750793, "learning_rate": 5.1903097321202476e-05, "loss": 0.0169, "step": 41080 }, { "grad_norm": 0.35145142674446106, "learning_rate": 5.18824437631596e-05, "loss": 0.0368, "step": 41090 }, { "grad_norm": 0.225617915391922, "learning_rate": 5.186178988345856e-05, "loss": 0.0243, "step": 41100 }, { "grad_norm": 0.3303466737270355, "learning_rate": 5.184113568562855e-05, "loss": 0.0306, "step": 41110 }, { "grad_norm": 0.32847756147384644, "learning_rate": 5.18204811731988e-05, "loss": 0.0318, "step": 41120 }, { "grad_norm": 0.33263272047042847, "learning_rate": 5.179982634969862e-05, "loss": 0.0231, "step": 41130 }, { "grad_norm": 0.41963326930999756, "learning_rate": 5.177917121865734e-05, "loss": 0.0232, "step": 41140 }, { "grad_norm": 0.4090001583099365, "learning_rate": 5.1758515783604346e-05, "loss": 0.0226, "step": 41150 }, { "grad_norm": 0.4545804262161255, "learning_rate": 5.173786004806911e-05, "loss": 0.029, "step": 41160 }, { "grad_norm": 0.3674626648426056, "learning_rate": 5.1717204015581135e-05, "loss": 0.0288, "step": 41170 }, { "grad_norm": 0.2895728051662445, "learning_rate": 5.169654768966994e-05, "loss": 0.0222, "step": 41180 }, { "grad_norm": 0.27003076672554016, "learning_rate": 5.1675891073865156e-05, "loss": 0.0199, "step": 41190 }, { "grad_norm": 0.305957555770874, "learning_rate": 5.1655234171696424e-05, "loss": 0.0193, "step": 41200 }, { "grad_norm": 0.3155774176120758, "learning_rate": 5.163457698669343e-05, "loss": 0.0174, "step": 41210 }, { "grad_norm": 0.2587355971336365, "learning_rate": 5.1613919522385936e-05, "loss": 0.0208, "step": 41220 }, { "grad_norm": 0.2843765616416931, "learning_rate": 5.1593261782303746e-05, "loss": 0.0296, "step": 41230 }, { "grad_norm": 0.23337014019489288, "learning_rate": 5.157260376997669e-05, "loss": 0.0196, "step": 41240 }, { "grad_norm": 0.42985326051712036, "learning_rate": 5.155194548893466e-05, "loss": 0.024, "step": 41250 }, { "grad_norm": 0.2656107544898987, "learning_rate": 5.153128694270759e-05, "loss": 0.0333, "step": 41260 }, { "grad_norm": 0.27734172344207764, "learning_rate": 5.151062813482548e-05, "loss": 0.0271, "step": 41270 }, { "grad_norm": 0.33883655071258545, "learning_rate": 5.148996906881832e-05, "loss": 0.0196, "step": 41280 }, { "grad_norm": 0.32960137724876404, "learning_rate": 5.1469309748216196e-05, "loss": 0.0251, "step": 41290 }, { "grad_norm": 0.4050391614437103, "learning_rate": 5.144865017654923e-05, "loss": 0.0219, "step": 41300 }, { "grad_norm": 0.32593876123428345, "learning_rate": 5.1427990357347556e-05, "loss": 0.0289, "step": 41310 }, { "grad_norm": 0.3250696063041687, "learning_rate": 5.140733029414139e-05, "loss": 0.0254, "step": 41320 }, { "grad_norm": 0.2394818663597107, "learning_rate": 5.138666999046098e-05, "loss": 0.0224, "step": 41330 }, { "grad_norm": 0.255596786737442, "learning_rate": 5.136600944983658e-05, "loss": 0.0213, "step": 41340 }, { "grad_norm": 0.2300318479537964, "learning_rate": 5.134534867579853e-05, "loss": 0.0306, "step": 41350 }, { "grad_norm": 0.2701605260372162, "learning_rate": 5.1324687671877184e-05, "loss": 0.0183, "step": 41360 }, { "grad_norm": 0.4110190272331238, "learning_rate": 5.130402644160296e-05, "loss": 0.0289, "step": 41370 }, { "grad_norm": 0.30813923478126526, "learning_rate": 5.128336498850628e-05, "loss": 0.0231, "step": 41380 }, { "grad_norm": 0.18923045694828033, "learning_rate": 5.126270331611761e-05, "loss": 0.0207, "step": 41390 }, { "grad_norm": 0.3655222952365875, "learning_rate": 5.124204142796748e-05, "loss": 0.0204, "step": 41400 }, { "grad_norm": 0.28991109132766724, "learning_rate": 5.122137932758644e-05, "loss": 0.0231, "step": 41410 }, { "grad_norm": 0.25565722584724426, "learning_rate": 5.1200717018505074e-05, "loss": 0.0225, "step": 41420 }, { "grad_norm": 0.24553897976875305, "learning_rate": 5.118005450425403e-05, "loss": 0.0201, "step": 41430 }, { "grad_norm": 0.3543528914451599, "learning_rate": 5.115939178836392e-05, "loss": 0.0176, "step": 41440 }, { "grad_norm": 0.2973175346851349, "learning_rate": 5.113872887436547e-05, "loss": 0.029, "step": 41450 }, { "grad_norm": 0.3553742468357086, "learning_rate": 5.11180657657894e-05, "loss": 0.0269, "step": 41460 }, { "grad_norm": 0.2748781740665436, "learning_rate": 5.109740246616649e-05, "loss": 0.0212, "step": 41470 }, { "grad_norm": 0.5107971429824829, "learning_rate": 5.10767389790275e-05, "loss": 0.0324, "step": 41480 }, { "grad_norm": 0.35095474123954773, "learning_rate": 5.1056075307903265e-05, "loss": 0.0279, "step": 41490 }, { "grad_norm": 0.4146108329296112, "learning_rate": 5.1035411456324634e-05, "loss": 0.021, "step": 41500 }, { "grad_norm": 0.18573959171772003, "learning_rate": 5.1014747427822504e-05, "loss": 0.0242, "step": 41510 }, { "grad_norm": 0.35019436478614807, "learning_rate": 5.0994083225927804e-05, "loss": 0.024, "step": 41520 }, { "grad_norm": 0.28921860456466675, "learning_rate": 5.0973418854171475e-05, "loss": 0.0246, "step": 41530 }, { "grad_norm": 0.32087796926498413, "learning_rate": 5.095275431608447e-05, "loss": 0.0201, "step": 41540 }, { "grad_norm": 0.4505974054336548, "learning_rate": 5.093208961519782e-05, "loss": 0.0238, "step": 41550 }, { "grad_norm": 0.29354146122932434, "learning_rate": 5.091142475504255e-05, "loss": 0.02, "step": 41560 }, { "grad_norm": 0.4216310977935791, "learning_rate": 5.089075973914971e-05, "loss": 0.0254, "step": 41570 }, { "grad_norm": 0.26275619864463806, "learning_rate": 5.08700945710504e-05, "loss": 0.0334, "step": 41580 }, { "grad_norm": 0.30101504921913147, "learning_rate": 5.0849429254275714e-05, "loss": 0.0207, "step": 41590 }, { "grad_norm": 0.21154841780662537, "learning_rate": 5.0828763792356804e-05, "loss": 0.0197, "step": 41600 }, { "grad_norm": 0.2712859809398651, "learning_rate": 5.080809818882483e-05, "loss": 0.0256, "step": 41610 }, { "grad_norm": 0.33230066299438477, "learning_rate": 5.078743244721097e-05, "loss": 0.025, "step": 41620 }, { "grad_norm": 0.24479997158050537, "learning_rate": 5.0766766571046455e-05, "loss": 0.0229, "step": 41630 }, { "grad_norm": 0.4689547121524811, "learning_rate": 5.07461005638625e-05, "loss": 0.02, "step": 41640 }, { "grad_norm": 0.24244324862957, "learning_rate": 5.072543442919037e-05, "loss": 0.0236, "step": 41650 }, { "grad_norm": 0.3223833441734314, "learning_rate": 5.070476817056132e-05, "loss": 0.0279, "step": 41660 }, { "grad_norm": 0.5174487233161926, "learning_rate": 5.068410179150668e-05, "loss": 0.0207, "step": 41670 }, { "grad_norm": 0.35148265957832336, "learning_rate": 5.066343529555775e-05, "loss": 0.0206, "step": 41680 }, { "grad_norm": 0.22667501866817474, "learning_rate": 5.064276868624587e-05, "loss": 0.0207, "step": 41690 }, { "grad_norm": 0.31766876578330994, "learning_rate": 5.0622101967102396e-05, "loss": 0.0219, "step": 41700 }, { "grad_norm": 0.2774638235569, "learning_rate": 5.0601435141658714e-05, "loss": 0.0188, "step": 41710 }, { "grad_norm": 0.4035339057445526, "learning_rate": 5.058076821344622e-05, "loss": 0.0238, "step": 41720 }, { "grad_norm": 0.5298086404800415, "learning_rate": 5.056010118599632e-05, "loss": 0.0347, "step": 41730 }, { "grad_norm": 0.2307213842868805, "learning_rate": 5.053943406284044e-05, "loss": 0.0217, "step": 41740 }, { "grad_norm": 0.288768470287323, "learning_rate": 5.051876684751006e-05, "loss": 0.0211, "step": 41750 }, { "grad_norm": 0.2777804136276245, "learning_rate": 5.0498099543536584e-05, "loss": 0.0195, "step": 41760 }, { "grad_norm": 0.282020628452301, "learning_rate": 5.047743215445152e-05, "loss": 0.0247, "step": 41770 }, { "grad_norm": 0.3648545742034912, "learning_rate": 5.045676468378637e-05, "loss": 0.0221, "step": 41780 }, { "grad_norm": 0.3240503966808319, "learning_rate": 5.0436097135072626e-05, "loss": 0.0285, "step": 41790 }, { "grad_norm": 0.2600509822368622, "learning_rate": 5.041542951184181e-05, "loss": 0.0258, "step": 41800 }, { "grad_norm": 0.20286162197589874, "learning_rate": 5.039476181762545e-05, "loss": 0.0211, "step": 41810 }, { "grad_norm": 0.2010015994310379, "learning_rate": 5.037409405595508e-05, "loss": 0.0211, "step": 41820 }, { "grad_norm": 0.3424580693244934, "learning_rate": 5.035342623036229e-05, "loss": 0.023, "step": 41830 }, { "grad_norm": 0.3390337824821472, "learning_rate": 5.033275834437862e-05, "loss": 0.0242, "step": 41840 }, { "grad_norm": 0.3399718701839447, "learning_rate": 5.031209040153564e-05, "loss": 0.022, "step": 41850 }, { "grad_norm": 0.26876768469810486, "learning_rate": 5.0291422405364955e-05, "loss": 0.0331, "step": 41860 }, { "grad_norm": 0.25318512320518494, "learning_rate": 5.0270754359398133e-05, "loss": 0.0184, "step": 41870 }, { "grad_norm": 0.1928115040063858, "learning_rate": 5.025008626716682e-05, "loss": 0.0175, "step": 41880 }, { "grad_norm": 0.29461580514907837, "learning_rate": 5.0229418132202585e-05, "loss": 0.0179, "step": 41890 }, { "grad_norm": 0.17613005638122559, "learning_rate": 5.020874995803707e-05, "loss": 0.0242, "step": 41900 }, { "grad_norm": 0.3728494644165039, "learning_rate": 5.01880817482019e-05, "loss": 0.025, "step": 41910 }, { "grad_norm": 0.39020413160324097, "learning_rate": 5.01674135062287e-05, "loss": 0.0181, "step": 41920 }, { "grad_norm": 0.3025552034378052, "learning_rate": 5.0146745235649115e-05, "loss": 0.0226, "step": 41930 }, { "grad_norm": 0.4620332717895508, "learning_rate": 5.0126076939994795e-05, "loss": 0.0221, "step": 41940 }, { "grad_norm": 0.31097689270973206, "learning_rate": 5.010540862279736e-05, "loss": 0.0266, "step": 41950 }, { "grad_norm": 0.2049938440322876, "learning_rate": 5.008474028758846e-05, "loss": 0.0205, "step": 41960 }, { "grad_norm": 0.32163524627685547, "learning_rate": 5.0064071937899784e-05, "loss": 0.0264, "step": 41970 }, { "grad_norm": 0.4648847281932831, "learning_rate": 5.004340357726296e-05, "loss": 0.0242, "step": 41980 }, { "grad_norm": 0.21850872039794922, "learning_rate": 5.002273520920965e-05, "loss": 0.0217, "step": 41990 }, { "grad_norm": 0.2130977064371109, "learning_rate": 5.000206683727151e-05, "loss": 0.0195, "step": 42000 }, { "grad_norm": 0.19717751443386078, "learning_rate": 4.998139846498021e-05, "loss": 0.0228, "step": 42010 }, { "grad_norm": 0.29022642970085144, "learning_rate": 4.9960730095867405e-05, "loss": 0.0234, "step": 42020 }, { "grad_norm": 0.21381568908691406, "learning_rate": 4.9940061733464755e-05, "loss": 0.0252, "step": 42030 }, { "grad_norm": 1.765181064605713, "learning_rate": 4.991939338130392e-05, "loss": 0.0286, "step": 42040 }, { "grad_norm": 0.42385804653167725, "learning_rate": 4.989872504291653e-05, "loss": 0.0259, "step": 42050 }, { "grad_norm": 0.2634970545768738, "learning_rate": 4.9878056721834273e-05, "loss": 0.0207, "step": 42060 }, { "grad_norm": 0.3499259054660797, "learning_rate": 4.98573884215888e-05, "loss": 0.0255, "step": 42070 }, { "grad_norm": 0.32579147815704346, "learning_rate": 4.9836720145711715e-05, "loss": 0.0267, "step": 42080 }, { "grad_norm": 0.3136959671974182, "learning_rate": 4.9816051897734725e-05, "loss": 0.0258, "step": 42090 }, { "grad_norm": 0.2881386876106262, "learning_rate": 4.979538368118941e-05, "loss": 0.0239, "step": 42100 }, { "grad_norm": 0.38804396986961365, "learning_rate": 4.9774715499607446e-05, "loss": 0.021, "step": 42110 }, { "grad_norm": 0.2493661344051361, "learning_rate": 4.9754047356520416e-05, "loss": 0.03, "step": 42120 }, { "grad_norm": 0.2537655532360077, "learning_rate": 4.973337925545999e-05, "loss": 0.0231, "step": 42130 }, { "grad_norm": 0.3892008662223816, "learning_rate": 4.9712711199957746e-05, "loss": 0.0247, "step": 42140 }, { "grad_norm": 0.4157654047012329, "learning_rate": 4.969204319354531e-05, "loss": 0.0277, "step": 42150 }, { "grad_norm": 0.31678661704063416, "learning_rate": 4.9671375239754267e-05, "loss": 0.0217, "step": 42160 }, { "grad_norm": 0.24385786056518555, "learning_rate": 4.9650707342116195e-05, "loss": 0.0199, "step": 42170 }, { "grad_norm": 0.21040569245815277, "learning_rate": 4.96300395041627e-05, "loss": 0.0213, "step": 42180 }, { "grad_norm": 0.42039111256599426, "learning_rate": 4.960937172942532e-05, "loss": 0.0221, "step": 42190 }, { "grad_norm": 0.22328387200832367, "learning_rate": 4.958870402143565e-05, "loss": 0.0202, "step": 42200 }, { "grad_norm": 0.2946512997150421, "learning_rate": 4.9568036383725186e-05, "loss": 0.0216, "step": 42210 }, { "grad_norm": 0.38476958870887756, "learning_rate": 4.9547368819825516e-05, "loss": 0.027, "step": 42220 }, { "grad_norm": 0.19513846933841705, "learning_rate": 4.952670133326812e-05, "loss": 0.0243, "step": 42230 }, { "grad_norm": 0.3306169807910919, "learning_rate": 4.950603392758453e-05, "loss": 0.0233, "step": 42240 }, { "grad_norm": 0.2207086831331253, "learning_rate": 4.948536660630621e-05, "loss": 0.0211, "step": 42250 }, { "grad_norm": 0.3059311509132385, "learning_rate": 4.9464699372964685e-05, "loss": 0.0253, "step": 42260 }, { "grad_norm": 0.22595158219337463, "learning_rate": 4.9444032231091395e-05, "loss": 0.0255, "step": 42270 }, { "grad_norm": 0.34563225507736206, "learning_rate": 4.9423365184217776e-05, "loss": 0.0186, "step": 42280 }, { "grad_norm": 0.35587453842163086, "learning_rate": 4.940269823587529e-05, "loss": 0.0201, "step": 42290 }, { "grad_norm": 0.4286033809185028, "learning_rate": 4.938203138959533e-05, "loss": 0.0189, "step": 42300 }, { "grad_norm": 0.23196813464164734, "learning_rate": 4.9361364648909325e-05, "loss": 0.0265, "step": 42310 }, { "grad_norm": 0.27503806352615356, "learning_rate": 4.9340698017348605e-05, "loss": 0.0196, "step": 42320 }, { "grad_norm": 0.31966933608055115, "learning_rate": 4.932003149844458e-05, "loss": 0.0216, "step": 42330 }, { "grad_norm": 0.44080567359924316, "learning_rate": 4.929936509572857e-05, "loss": 0.0311, "step": 42340 }, { "grad_norm": 0.4081614017486572, "learning_rate": 4.927869881273191e-05, "loss": 0.0253, "step": 42350 }, { "grad_norm": 0.3151772618293762, "learning_rate": 4.9258032652985894e-05, "loss": 0.0245, "step": 42360 }, { "grad_norm": 0.24279959499835968, "learning_rate": 4.9237366620021786e-05, "loss": 0.0242, "step": 42370 }, { "grad_norm": 0.32954469323158264, "learning_rate": 4.921670071737089e-05, "loss": 0.0213, "step": 42380 }, { "grad_norm": 0.4079456031322479, "learning_rate": 4.91960349485644e-05, "loss": 0.0299, "step": 42390 }, { "grad_norm": 0.30811741948127747, "learning_rate": 4.9175369317133566e-05, "loss": 0.0232, "step": 42400 }, { "grad_norm": 0.3217805027961731, "learning_rate": 4.915470382660953e-05, "loss": 0.0244, "step": 42410 }, { "grad_norm": 0.22129371762275696, "learning_rate": 4.9134038480523524e-05, "loss": 0.0207, "step": 42420 }, { "grad_norm": 0.2714349925518036, "learning_rate": 4.911337328240664e-05, "loss": 0.0282, "step": 42430 }, { "grad_norm": 0.2705337703227997, "learning_rate": 4.909270823579003e-05, "loss": 0.0267, "step": 42440 }, { "grad_norm": 0.375984787940979, "learning_rate": 4.907204334420476e-05, "loss": 0.0207, "step": 42450 }, { "grad_norm": 0.34521958231925964, "learning_rate": 4.9051378611181884e-05, "loss": 0.0275, "step": 42460 }, { "grad_norm": 0.250343382358551, "learning_rate": 4.9030714040252486e-05, "loss": 0.0156, "step": 42470 }, { "grad_norm": 0.20959219336509705, "learning_rate": 4.901004963494752e-05, "loss": 0.0171, "step": 42480 }, { "grad_norm": 0.4206004738807678, "learning_rate": 4.898938539879802e-05, "loss": 0.0267, "step": 42490 }, { "grad_norm": 0.25930866599082947, "learning_rate": 4.8968721335334904e-05, "loss": 0.0229, "step": 42500 }, { "grad_norm": 0.2310275435447693, "learning_rate": 4.894805744808912e-05, "loss": 0.0232, "step": 42510 }, { "grad_norm": 0.2205248326063156, "learning_rate": 4.8927393740591524e-05, "loss": 0.0203, "step": 42520 }, { "grad_norm": 0.3790871202945709, "learning_rate": 4.890673021637302e-05, "loss": 0.0207, "step": 42530 }, { "grad_norm": 0.27378755807876587, "learning_rate": 4.888606687896442e-05, "loss": 0.0184, "step": 42540 }, { "grad_norm": 0.2732101082801819, "learning_rate": 4.886540373189652e-05, "loss": 0.0216, "step": 42550 }, { "grad_norm": 0.2575031518936157, "learning_rate": 4.88447407787001e-05, "loss": 0.0145, "step": 42560 }, { "grad_norm": 0.2851046025753021, "learning_rate": 4.8824078022905864e-05, "loss": 0.0177, "step": 42570 }, { "grad_norm": 0.2567559778690338, "learning_rate": 4.880341546804455e-05, "loss": 0.015, "step": 42580 }, { "grad_norm": 0.3339442014694214, "learning_rate": 4.878275311764679e-05, "loss": 0.0188, "step": 42590 }, { "grad_norm": 0.4348843991756439, "learning_rate": 4.8762090975243254e-05, "loss": 0.022, "step": 42600 }, { "grad_norm": 0.2445719838142395, "learning_rate": 4.874142904436448e-05, "loss": 0.0206, "step": 42610 }, { "grad_norm": 0.41474297642707825, "learning_rate": 4.8720767328541094e-05, "loss": 0.0305, "step": 42620 }, { "grad_norm": 0.36460939049720764, "learning_rate": 4.870010583130358e-05, "loss": 0.0203, "step": 42630 }, { "grad_norm": 0.30183249711990356, "learning_rate": 4.867944455618241e-05, "loss": 0.0239, "step": 42640 }, { "grad_norm": 0.21805472671985626, "learning_rate": 4.865878350670807e-05, "loss": 0.0187, "step": 42650 }, { "grad_norm": 0.33302995562553406, "learning_rate": 4.8638122686410914e-05, "loss": 0.0196, "step": 42660 }, { "grad_norm": 0.2850074768066406, "learning_rate": 4.861746209882137e-05, "loss": 0.0168, "step": 42670 }, { "grad_norm": 0.2695937156677246, "learning_rate": 4.859680174746972e-05, "loss": 0.0228, "step": 42680 }, { "grad_norm": 0.2574308216571808, "learning_rate": 4.857614163588629e-05, "loss": 0.0223, "step": 42690 }, { "grad_norm": 0.2536945939064026, "learning_rate": 4.855548176760131e-05, "loss": 0.0191, "step": 42700 }, { "grad_norm": 0.2990882992744446, "learning_rate": 4.8534822146144986e-05, "loss": 0.0192, "step": 42710 }, { "grad_norm": 0.34305912256240845, "learning_rate": 4.851416277504749e-05, "loss": 0.0331, "step": 42720 }, { "grad_norm": 0.446056067943573, "learning_rate": 4.8493503657838923e-05, "loss": 0.0168, "step": 42730 }, { "grad_norm": 0.17934812605381012, "learning_rate": 4.84728447980494e-05, "loss": 0.0206, "step": 42740 }, { "grad_norm": 0.4585248827934265, "learning_rate": 4.8452186199208914e-05, "loss": 0.0219, "step": 42750 }, { "grad_norm": 0.33069100975990295, "learning_rate": 4.843152786484749e-05, "loss": 0.0227, "step": 42760 }, { "grad_norm": 0.2741270661354065, "learning_rate": 4.8410869798495035e-05, "loss": 0.0207, "step": 42770 }, { "grad_norm": 0.21931616961956024, "learning_rate": 4.8390212003681486e-05, "loss": 0.0225, "step": 42780 }, { "grad_norm": 0.3394681215286255, "learning_rate": 4.836955448393667e-05, "loss": 0.024, "step": 42790 }, { "grad_norm": 0.2675538957118988, "learning_rate": 4.8348897242790406e-05, "loss": 0.0251, "step": 42800 }, { "grad_norm": 0.24760954082012177, "learning_rate": 4.832824028377243e-05, "loss": 0.0162, "step": 42810 }, { "grad_norm": 0.2676766514778137, "learning_rate": 4.830758361041249e-05, "loss": 0.0195, "step": 42820 }, { "grad_norm": 0.44379281997680664, "learning_rate": 4.828692722624022e-05, "loss": 0.0278, "step": 42830 }, { "grad_norm": 0.22002458572387695, "learning_rate": 4.826627113478522e-05, "loss": 0.0325, "step": 42840 }, { "grad_norm": 0.3382912874221802, "learning_rate": 4.824561533957708e-05, "loss": 0.0199, "step": 42850 }, { "grad_norm": 0.29382431507110596, "learning_rate": 4.822495984414527e-05, "loss": 0.0205, "step": 42860 }, { "grad_norm": 0.27545249462127686, "learning_rate": 4.8204304652019304e-05, "loss": 0.0195, "step": 42870 }, { "grad_norm": 0.3068251609802246, "learning_rate": 4.8183649766728537e-05, "loss": 0.0239, "step": 42880 }, { "grad_norm": 0.5006483793258667, "learning_rate": 4.816299519180237e-05, "loss": 0.0216, "step": 42890 }, { "grad_norm": 0.3092568814754486, "learning_rate": 4.8142340930770066e-05, "loss": 0.0201, "step": 42900 }, { "grad_norm": 0.30034998059272766, "learning_rate": 4.81216869871609e-05, "loss": 0.0201, "step": 42910 }, { "grad_norm": 0.25329679250717163, "learning_rate": 4.8101033364504064e-05, "loss": 0.0195, "step": 42920 }, { "grad_norm": 0.2659328579902649, "learning_rate": 4.808038006632866e-05, "loss": 0.0241, "step": 42930 }, { "grad_norm": 0.4138922393321991, "learning_rate": 4.805972709616383e-05, "loss": 0.0207, "step": 42940 }, { "grad_norm": 0.2988135516643524, "learning_rate": 4.8039074457538556e-05, "loss": 0.0205, "step": 42950 }, { "grad_norm": 0.2651868164539337, "learning_rate": 4.801842215398184e-05, "loss": 0.0304, "step": 42960 }, { "grad_norm": 0.3761458694934845, "learning_rate": 4.799777018902256e-05, "loss": 0.0245, "step": 42970 }, { "grad_norm": 0.29733946919441223, "learning_rate": 4.797711856618961e-05, "loss": 0.0255, "step": 42980 }, { "grad_norm": 0.4756872355937958, "learning_rate": 4.7956467289011765e-05, "loss": 0.0212, "step": 42990 }, { "grad_norm": 0.29069778323173523, "learning_rate": 4.793581636101778e-05, "loss": 0.0274, "step": 43000 }, { "grad_norm": 0.43491312861442566, "learning_rate": 4.7915165785736326e-05, "loss": 0.025, "step": 43010 }, { "grad_norm": 0.3392865061759949, "learning_rate": 4.789451556669599e-05, "loss": 0.0168, "step": 43020 }, { "grad_norm": 0.27255067229270935, "learning_rate": 4.7873865707425384e-05, "loss": 0.024, "step": 43030 }, { "grad_norm": 0.3693639636039734, "learning_rate": 4.7853216211452975e-05, "loss": 0.0276, "step": 43040 }, { "grad_norm": 0.3836826682090759, "learning_rate": 4.78325670823072e-05, "loss": 0.0218, "step": 43050 }, { "grad_norm": 0.19368785619735718, "learning_rate": 4.781191832351641e-05, "loss": 0.0239, "step": 43060 }, { "grad_norm": 0.541862964630127, "learning_rate": 4.7791269938608965e-05, "loss": 0.0259, "step": 43070 }, { "grad_norm": 0.25241103768348694, "learning_rate": 4.777062193111305e-05, "loss": 0.0247, "step": 43080 }, { "grad_norm": 0.44367867708206177, "learning_rate": 4.77499743045569e-05, "loss": 0.0216, "step": 43090 }, { "grad_norm": 0.40194669365882874, "learning_rate": 4.7729327062468596e-05, "loss": 0.0221, "step": 43100 }, { "grad_norm": 0.34537985920906067, "learning_rate": 4.770868020837619e-05, "loss": 0.0208, "step": 43110 }, { "grad_norm": 0.22069554030895233, "learning_rate": 4.768803374580768e-05, "loss": 0.016, "step": 43120 }, { "grad_norm": 0.23327898979187012, "learning_rate": 4.766738767829094e-05, "loss": 0.0193, "step": 43130 }, { "grad_norm": 0.23305290937423706, "learning_rate": 4.764674200935388e-05, "loss": 0.0212, "step": 43140 }, { "grad_norm": 0.30432578921318054, "learning_rate": 4.762609674252424e-05, "loss": 0.0201, "step": 43150 }, { "grad_norm": 0.2072683870792389, "learning_rate": 4.760545188132974e-05, "loss": 0.0348, "step": 43160 }, { "grad_norm": 0.21785752475261688, "learning_rate": 4.758480742929801e-05, "loss": 0.0216, "step": 43170 }, { "grad_norm": 0.3568132221698761, "learning_rate": 4.756416338995664e-05, "loss": 0.0282, "step": 43180 }, { "grad_norm": 0.5212578773498535, "learning_rate": 4.7543519766833145e-05, "loss": 0.0192, "step": 43190 }, { "grad_norm": 0.19009923934936523, "learning_rate": 4.752287656345492e-05, "loss": 0.0253, "step": 43200 }, { "grad_norm": 0.22942174971103668, "learning_rate": 4.7502233783349355e-05, "loss": 0.0208, "step": 43210 }, { "grad_norm": 0.3041106164455414, "learning_rate": 4.7481591430043694e-05, "loss": 0.0211, "step": 43220 }, { "grad_norm": 0.2751219868659973, "learning_rate": 4.746094950706521e-05, "loss": 0.0262, "step": 43230 }, { "grad_norm": 0.17583905160427094, "learning_rate": 4.744030801794099e-05, "loss": 0.0215, "step": 43240 }, { "grad_norm": 0.46588462591171265, "learning_rate": 4.741966696619813e-05, "loss": 0.0296, "step": 43250 }, { "grad_norm": 0.26491668820381165, "learning_rate": 4.739902635536361e-05, "loss": 0.0208, "step": 43260 }, { "grad_norm": 0.3402687907218933, "learning_rate": 4.737838618896436e-05, "loss": 0.0237, "step": 43270 }, { "grad_norm": 0.49127647280693054, "learning_rate": 4.7357746470527203e-05, "loss": 0.0249, "step": 43280 }, { "grad_norm": 0.3188643157482147, "learning_rate": 4.73371072035789e-05, "loss": 0.0237, "step": 43290 }, { "grad_norm": 0.4672718346118927, "learning_rate": 4.731646839164616e-05, "loss": 0.025, "step": 43300 }, { "grad_norm": 0.32733777165412903, "learning_rate": 4.729583003825556e-05, "loss": 0.0244, "step": 43310 }, { "grad_norm": 0.23128826916217804, "learning_rate": 4.727519214693367e-05, "loss": 0.0238, "step": 43320 }, { "grad_norm": 0.2576501667499542, "learning_rate": 4.725455472120689e-05, "loss": 0.0182, "step": 43330 }, { "grad_norm": 0.22752536833286285, "learning_rate": 4.723391776460164e-05, "loss": 0.0226, "step": 43340 }, { "grad_norm": 0.2825258672237396, "learning_rate": 4.7213281280644186e-05, "loss": 0.0212, "step": 43350 }, { "grad_norm": 0.2461729198694229, "learning_rate": 4.719264527286075e-05, "loss": 0.0195, "step": 43360 }, { "grad_norm": 0.2085573822259903, "learning_rate": 4.717200974477744e-05, "loss": 0.0184, "step": 43370 }, { "grad_norm": 0.32849863171577454, "learning_rate": 4.715137469992034e-05, "loss": 0.0225, "step": 43380 }, { "grad_norm": 0.47783705592155457, "learning_rate": 4.713074014181539e-05, "loss": 0.0193, "step": 43390 }, { "grad_norm": 0.45432719588279724, "learning_rate": 4.711010607398846e-05, "loss": 0.0184, "step": 43400 }, { "grad_norm": 0.2436126470565796, "learning_rate": 4.708947249996538e-05, "loss": 0.0202, "step": 43410 }, { "grad_norm": 0.299703449010849, "learning_rate": 4.706883942327183e-05, "loss": 0.0211, "step": 43420 }, { "grad_norm": 0.20695947110652924, "learning_rate": 4.704820684743347e-05, "loss": 0.0216, "step": 43430 }, { "grad_norm": 0.3313417136669159, "learning_rate": 4.702757477597581e-05, "loss": 0.0261, "step": 43440 }, { "grad_norm": 0.2889827489852905, "learning_rate": 4.700694321242435e-05, "loss": 0.0197, "step": 43450 }, { "grad_norm": 0.43720147013664246, "learning_rate": 4.698631216030441e-05, "loss": 0.0221, "step": 43460 }, { "grad_norm": 0.28029951453208923, "learning_rate": 4.6965681623141314e-05, "loss": 0.0198, "step": 43470 }, { "grad_norm": 0.2551896572113037, "learning_rate": 4.694505160446024e-05, "loss": 0.0163, "step": 43480 }, { "grad_norm": 0.2236066311597824, "learning_rate": 4.692442210778628e-05, "loss": 0.0211, "step": 43490 }, { "grad_norm": 0.4175005555152893, "learning_rate": 4.690379313664448e-05, "loss": 0.0227, "step": 43500 }, { "grad_norm": 0.28267958760261536, "learning_rate": 4.688316469455973e-05, "loss": 0.0221, "step": 43510 }, { "grad_norm": 0.25688400864601135, "learning_rate": 4.6862536785056914e-05, "loss": 0.0231, "step": 43520 }, { "grad_norm": 0.370497465133667, "learning_rate": 4.684190941166073e-05, "loss": 0.0222, "step": 43530 }, { "grad_norm": 0.31165218353271484, "learning_rate": 4.682128257789587e-05, "loss": 0.0181, "step": 43540 }, { "grad_norm": 0.40930667519569397, "learning_rate": 4.680065628728687e-05, "loss": 0.0194, "step": 43550 }, { "grad_norm": 0.5583431124687195, "learning_rate": 4.678003054335822e-05, "loss": 0.0287, "step": 43560 }, { "grad_norm": 0.3398895859718323, "learning_rate": 4.675940534963428e-05, "loss": 0.0249, "step": 43570 }, { "grad_norm": 0.2225189357995987, "learning_rate": 4.673878070963931e-05, "loss": 0.0224, "step": 43580 }, { "grad_norm": 0.1632993221282959, "learning_rate": 4.671815662689756e-05, "loss": 0.0178, "step": 43590 }, { "grad_norm": 0.20641577243804932, "learning_rate": 4.669753310493306e-05, "loss": 0.0189, "step": 43600 }, { "grad_norm": 0.3915430009365082, "learning_rate": 4.6676910147269845e-05, "loss": 0.0187, "step": 43610 }, { "grad_norm": 0.23931884765625, "learning_rate": 4.665628775743177e-05, "loss": 0.0197, "step": 43620 }, { "grad_norm": 0.17004181444644928, "learning_rate": 4.66356659389427e-05, "loss": 0.0211, "step": 43630 }, { "grad_norm": 0.35750812292099, "learning_rate": 4.661504469532629e-05, "loss": 0.0212, "step": 43640 }, { "grad_norm": 0.2948574125766754, "learning_rate": 4.659442403010617e-05, "loss": 0.0175, "step": 43650 }, { "grad_norm": 0.25903528928756714, "learning_rate": 4.6573803946805845e-05, "loss": 0.0172, "step": 43660 }, { "grad_norm": 0.2618074119091034, "learning_rate": 4.6553184448948695e-05, "loss": 0.0181, "step": 43670 }, { "grad_norm": 0.377765417098999, "learning_rate": 4.653256554005807e-05, "loss": 0.0194, "step": 43680 }, { "grad_norm": 0.33621495962142944, "learning_rate": 4.6511947223657145e-05, "loss": 0.03, "step": 43690 }, { "grad_norm": 0.3369534909725189, "learning_rate": 4.649132950326906e-05, "loss": 0.0284, "step": 43700 }, { "grad_norm": 0.33705779910087585, "learning_rate": 4.647071238241679e-05, "loss": 0.0264, "step": 43710 }, { "grad_norm": 0.316935271024704, "learning_rate": 4.6450095864623264e-05, "loss": 0.0161, "step": 43720 }, { "grad_norm": 0.3669913113117218, "learning_rate": 4.642947995341123e-05, "loss": 0.0239, "step": 43730 }, { "grad_norm": 0.2241402566432953, "learning_rate": 4.640886465230345e-05, "loss": 0.0208, "step": 43740 }, { "grad_norm": 0.3052922487258911, "learning_rate": 4.6388249964822485e-05, "loss": 0.02, "step": 43750 }, { "grad_norm": 0.34334278106689453, "learning_rate": 4.63676358944908e-05, "loss": 0.0185, "step": 43760 }, { "grad_norm": 0.4304589331150055, "learning_rate": 4.634702244483081e-05, "loss": 0.0272, "step": 43770 }, { "grad_norm": 0.3244872987270355, "learning_rate": 4.632640961936475e-05, "loss": 0.0203, "step": 43780 }, { "grad_norm": 0.28122448921203613, "learning_rate": 4.6305797421614835e-05, "loss": 0.0184, "step": 43790 }, { "grad_norm": 0.3549800515174866, "learning_rate": 4.6285185855103105e-05, "loss": 0.0224, "step": 43800 }, { "grad_norm": 0.29882362484931946, "learning_rate": 4.626457492335151e-05, "loss": 0.0224, "step": 43810 }, { "grad_norm": 0.2075117975473404, "learning_rate": 4.624396462988188e-05, "loss": 0.0186, "step": 43820 }, { "grad_norm": 0.3057893216609955, "learning_rate": 4.6223354978215985e-05, "loss": 0.0176, "step": 43830 }, { "grad_norm": 0.3312540054321289, "learning_rate": 4.620274597187544e-05, "loss": 0.0227, "step": 43840 }, { "grad_norm": 0.23844455182552338, "learning_rate": 4.6182137614381736e-05, "loss": 0.0245, "step": 43850 }, { "grad_norm": 0.26144200563430786, "learning_rate": 4.616152990925631e-05, "loss": 0.0174, "step": 43860 }, { "grad_norm": 0.26639705896377563, "learning_rate": 4.614092286002041e-05, "loss": 0.018, "step": 43870 }, { "grad_norm": 0.2608698308467865, "learning_rate": 4.612031647019528e-05, "loss": 0.0168, "step": 43880 }, { "grad_norm": 0.24269023537635803, "learning_rate": 4.609971074330193e-05, "loss": 0.0213, "step": 43890 }, { "grad_norm": 0.40707212686538696, "learning_rate": 4.607910568286136e-05, "loss": 0.0255, "step": 43900 }, { "grad_norm": 0.3103732466697693, "learning_rate": 4.605850129239438e-05, "loss": 0.0209, "step": 43910 }, { "grad_norm": 0.30671390891075134, "learning_rate": 4.603789757542174e-05, "loss": 0.0199, "step": 43920 }, { "grad_norm": 0.4332347810268402, "learning_rate": 4.601729453546404e-05, "loss": 0.0235, "step": 43930 }, { "grad_norm": 0.3256300389766693, "learning_rate": 4.599669217604177e-05, "loss": 0.0248, "step": 43940 }, { "grad_norm": 0.295894056558609, "learning_rate": 4.597609050067532e-05, "loss": 0.0253, "step": 43950 }, { "grad_norm": 0.2513763904571533, "learning_rate": 4.595548951288495e-05, "loss": 0.0261, "step": 43960 }, { "grad_norm": 0.3852772116661072, "learning_rate": 4.593488921619081e-05, "loss": 0.0293, "step": 43970 }, { "grad_norm": 0.46282655000686646, "learning_rate": 4.591428961411289e-05, "loss": 0.0244, "step": 43980 }, { "grad_norm": 0.2770220935344696, "learning_rate": 4.589369071017117e-05, "loss": 0.0259, "step": 43990 }, { "grad_norm": 0.23778265714645386, "learning_rate": 4.587309250788538e-05, "loss": 0.0182, "step": 44000 }, { "grad_norm": 0.29817795753479004, "learning_rate": 4.585249501077522e-05, "loss": 0.0177, "step": 44010 }, { "grad_norm": 0.1667138934135437, "learning_rate": 4.583189822236021e-05, "loss": 0.025, "step": 44020 }, { "grad_norm": 0.30446964502334595, "learning_rate": 4.5811302146159816e-05, "loss": 0.0205, "step": 44030 }, { "grad_norm": 0.2271825075149536, "learning_rate": 4.579070678569332e-05, "loss": 0.0167, "step": 44040 }, { "grad_norm": 0.33517366647720337, "learning_rate": 4.5770112144479904e-05, "loss": 0.0166, "step": 44050 }, { "grad_norm": 0.26085591316223145, "learning_rate": 4.5749518226038645e-05, "loss": 0.0174, "step": 44060 }, { "grad_norm": 0.361806720495224, "learning_rate": 4.572892503388845e-05, "loss": 0.0221, "step": 44070 }, { "grad_norm": 0.18193207681179047, "learning_rate": 4.570833257154817e-05, "loss": 0.0231, "step": 44080 }, { "grad_norm": 0.2687632739543915, "learning_rate": 4.568774084253646e-05, "loss": 0.0273, "step": 44090 }, { "grad_norm": 0.3196795582771301, "learning_rate": 4.566714985037191e-05, "loss": 0.0217, "step": 44100 }, { "grad_norm": 0.37827712297439575, "learning_rate": 4.564655959857295e-05, "loss": 0.0259, "step": 44110 }, { "grad_norm": 0.32168686389923096, "learning_rate": 4.5625970090657893e-05, "loss": 0.0183, "step": 44120 }, { "grad_norm": 0.3163423240184784, "learning_rate": 4.560538133014491e-05, "loss": 0.0222, "step": 44130 }, { "grad_norm": 0.25059741735458374, "learning_rate": 4.5584793320552055e-05, "loss": 0.0154, "step": 44140 }, { "grad_norm": 0.19277353584766388, "learning_rate": 4.556420606539728e-05, "loss": 0.0198, "step": 44150 }, { "grad_norm": 0.18523567914962769, "learning_rate": 4.554361956819836e-05, "loss": 0.0169, "step": 44160 }, { "grad_norm": 0.24649828672409058, "learning_rate": 4.552303383247299e-05, "loss": 0.0191, "step": 44170 }, { "grad_norm": 0.3133244514465332, "learning_rate": 4.5502448861738664e-05, "loss": 0.0214, "step": 44180 }, { "grad_norm": 0.2446671426296234, "learning_rate": 4.548186465951284e-05, "loss": 0.0227, "step": 44190 }, { "grad_norm": 0.3600080907344818, "learning_rate": 4.546128122931277e-05, "loss": 0.0249, "step": 44200 }, { "grad_norm": 0.23338532447814941, "learning_rate": 4.544069857465561e-05, "loss": 0.0227, "step": 44210 }, { "grad_norm": 0.3735496699810028, "learning_rate": 4.542011669905837e-05, "loss": 0.0192, "step": 44220 }, { "grad_norm": 0.18174655735492706, "learning_rate": 4.53995356060379e-05, "loss": 0.0187, "step": 44230 }, { "grad_norm": 0.3031900227069855, "learning_rate": 4.5378955299110994e-05, "loss": 0.0149, "step": 44240 }, { "grad_norm": 0.5070690512657166, "learning_rate": 4.5358375781794233e-05, "loss": 0.0274, "step": 44250 }, { "grad_norm": 0.24532823264598846, "learning_rate": 4.533779705760411e-05, "loss": 0.0225, "step": 44260 }, { "grad_norm": 0.18530181050300598, "learning_rate": 4.5317219130056934e-05, "loss": 0.0217, "step": 44270 }, { "grad_norm": 0.3337535262107849, "learning_rate": 4.5296642002668964e-05, "loss": 0.025, "step": 44280 }, { "grad_norm": 0.45500630140304565, "learning_rate": 4.5276065678956216e-05, "loss": 0.0228, "step": 44290 }, { "grad_norm": 0.33007416129112244, "learning_rate": 4.525549016243466e-05, "loss": 0.0209, "step": 44300 }, { "grad_norm": 0.34316539764404297, "learning_rate": 4.523491545662008e-05, "loss": 0.0211, "step": 44310 }, { "grad_norm": 0.4086569547653198, "learning_rate": 4.5214341565028104e-05, "loss": 0.0236, "step": 44320 }, { "grad_norm": 0.3486386835575104, "learning_rate": 4.519376849117428e-05, "loss": 0.0152, "step": 44330 }, { "grad_norm": 0.2779925763607025, "learning_rate": 4.517319623857395e-05, "loss": 0.0239, "step": 44340 }, { "grad_norm": 0.29638776183128357, "learning_rate": 4.5152624810742394e-05, "loss": 0.0214, "step": 44350 }, { "grad_norm": 0.36014848947525024, "learning_rate": 4.5132054211194664e-05, "loss": 0.0222, "step": 44360 }, { "grad_norm": 0.3271130323410034, "learning_rate": 4.511148444344574e-05, "loss": 0.0171, "step": 44370 }, { "grad_norm": 0.5116726756095886, "learning_rate": 4.509091551101041e-05, "loss": 0.0302, "step": 44380 }, { "grad_norm": 0.26025041937828064, "learning_rate": 4.507034741740338e-05, "loss": 0.022, "step": 44390 }, { "grad_norm": 0.351716011762619, "learning_rate": 4.5049780166139145e-05, "loss": 0.023, "step": 44400 }, { "grad_norm": 0.35479483008384705, "learning_rate": 4.5029213760732075e-05, "loss": 0.0266, "step": 44410 }, { "grad_norm": 0.3628246486186981, "learning_rate": 4.5008648204696434e-05, "loss": 0.0237, "step": 44420 }, { "grad_norm": 0.37670648097991943, "learning_rate": 4.4988083501546284e-05, "loss": 0.0212, "step": 44430 }, { "grad_norm": 0.2057880014181137, "learning_rate": 4.4967519654795606e-05, "loss": 0.0146, "step": 44440 }, { "grad_norm": 0.27684810757637024, "learning_rate": 4.494695666795816e-05, "loss": 0.0203, "step": 44450 }, { "grad_norm": 0.290221631526947, "learning_rate": 4.4926394544547644e-05, "loss": 0.0279, "step": 44460 }, { "grad_norm": 0.40405935049057007, "learning_rate": 4.490583328807752e-05, "loss": 0.022, "step": 44470 }, { "grad_norm": 0.19870807230472565, "learning_rate": 4.488527290206117e-05, "loss": 0.0187, "step": 44480 }, { "grad_norm": 0.6695476770401001, "learning_rate": 4.48647133900118e-05, "loss": 0.0208, "step": 44490 }, { "grad_norm": 0.3702149987220764, "learning_rate": 4.484415475544244e-05, "loss": 0.0213, "step": 44500 }, { "grad_norm": 0.3169005513191223, "learning_rate": 4.4823597001866035e-05, "loss": 0.023, "step": 44510 }, { "grad_norm": 0.31773558259010315, "learning_rate": 4.480304013279532e-05, "loss": 0.0247, "step": 44520 }, { "grad_norm": 0.2301868051290512, "learning_rate": 4.478248415174292e-05, "loss": 0.0199, "step": 44530 }, { "grad_norm": 0.2787541151046753, "learning_rate": 4.476192906222126e-05, "loss": 0.0265, "step": 44540 }, { "grad_norm": 0.3454297184944153, "learning_rate": 4.474137486774268e-05, "loss": 0.026, "step": 44550 }, { "grad_norm": 0.2721575200557709, "learning_rate": 4.4720821571819296e-05, "loss": 0.0199, "step": 44560 }, { "grad_norm": 0.31298041343688965, "learning_rate": 4.470026917796314e-05, "loss": 0.0292, "step": 44570 }, { "grad_norm": 0.302592396736145, "learning_rate": 4.4679717689686005e-05, "loss": 0.0399, "step": 44580 }, { "grad_norm": 0.3500297963619232, "learning_rate": 4.465916711049963e-05, "loss": 0.019, "step": 44590 }, { "grad_norm": 0.40721073746681213, "learning_rate": 4.4638617443915524e-05, "loss": 0.0259, "step": 44600 }, { "grad_norm": 0.43622180819511414, "learning_rate": 4.4618068693445055e-05, "loss": 0.0217, "step": 44610 }, { "grad_norm": 0.5877621173858643, "learning_rate": 4.459752086259946e-05, "loss": 0.024, "step": 44620 }, { "grad_norm": 0.4536008834838867, "learning_rate": 4.457697395488977e-05, "loss": 0.027, "step": 44630 }, { "grad_norm": 0.21843016147613525, "learning_rate": 4.455642797382693e-05, "loss": 0.0297, "step": 44640 }, { "grad_norm": 0.2948354184627533, "learning_rate": 4.4535882922921643e-05, "loss": 0.0228, "step": 44650 }, { "grad_norm": 0.38736578822135925, "learning_rate": 4.451533880568455e-05, "loss": 0.0202, "step": 44660 }, { "grad_norm": 0.40764421224594116, "learning_rate": 4.449479562562603e-05, "loss": 0.0243, "step": 44670 }, { "grad_norm": 0.3615700900554657, "learning_rate": 4.447425338625639e-05, "loss": 0.0217, "step": 44680 }, { "grad_norm": 0.48305249214172363, "learning_rate": 4.4453712091085705e-05, "loss": 0.0231, "step": 44690 }, { "grad_norm": 0.3098808526992798, "learning_rate": 4.443317174362392e-05, "loss": 0.0247, "step": 44700 }, { "grad_norm": 0.4012118875980377, "learning_rate": 4.441263234738085e-05, "loss": 0.0182, "step": 44710 }, { "grad_norm": 0.2746395170688629, "learning_rate": 4.4392093905866086e-05, "loss": 0.0208, "step": 44720 }, { "grad_norm": 0.45677611231803894, "learning_rate": 4.437155642258911e-05, "loss": 0.0206, "step": 44730 }, { "grad_norm": 0.39884209632873535, "learning_rate": 4.4351019901059177e-05, "loss": 0.0309, "step": 44740 }, { "grad_norm": 0.30949220061302185, "learning_rate": 4.433048434478547e-05, "loss": 0.0322, "step": 44750 }, { "grad_norm": 0.3432745039463043, "learning_rate": 4.4309949757276906e-05, "loss": 0.0268, "step": 44760 }, { "grad_norm": 0.31589215993881226, "learning_rate": 4.428941614204233e-05, "loss": 0.0316, "step": 44770 }, { "grad_norm": 0.2455577254295349, "learning_rate": 4.426888350259034e-05, "loss": 0.0243, "step": 44780 }, { "grad_norm": 0.33988866209983826, "learning_rate": 4.424835184242939e-05, "loss": 0.0304, "step": 44790 }, { "grad_norm": 0.2548803687095642, "learning_rate": 4.422782116506784e-05, "loss": 0.0206, "step": 44800 }, { "grad_norm": 0.34065350890159607, "learning_rate": 4.4207291474013754e-05, "loss": 0.0243, "step": 44810 }, { "grad_norm": 0.29289886355400085, "learning_rate": 4.418676277277514e-05, "loss": 0.0262, "step": 44820 }, { "grad_norm": 0.34872967004776, "learning_rate": 4.416623506485976e-05, "loss": 0.0291, "step": 44830 }, { "grad_norm": 0.3745933771133423, "learning_rate": 4.414570835377527e-05, "loss": 0.0218, "step": 44840 }, { "grad_norm": 0.16548161208629608, "learning_rate": 4.412518264302909e-05, "loss": 0.0183, "step": 44850 }, { "grad_norm": 0.33904942870140076, "learning_rate": 4.4104657936128535e-05, "loss": 0.019, "step": 44860 }, { "grad_norm": 0.8618848323822021, "learning_rate": 4.408413423658071e-05, "loss": 0.027, "step": 44870 }, { "grad_norm": 0.3678087890148163, "learning_rate": 4.406361154789253e-05, "loss": 0.0234, "step": 44880 }, { "grad_norm": 0.4610062837600708, "learning_rate": 4.4043089873570776e-05, "loss": 0.0236, "step": 44890 }, { "grad_norm": 0.40128135681152344, "learning_rate": 4.4022569217122025e-05, "loss": 0.0183, "step": 44900 }, { "grad_norm": 0.476409375667572, "learning_rate": 4.400204958205274e-05, "loss": 0.0195, "step": 44910 }, { "grad_norm": 0.3841497600078583, "learning_rate": 4.3981530971869125e-05, "loss": 0.0289, "step": 44920 }, { "grad_norm": 0.2978695034980774, "learning_rate": 4.396101339007727e-05, "loss": 0.0158, "step": 44930 }, { "grad_norm": 0.18968194723129272, "learning_rate": 4.394049684018304e-05, "loss": 0.0189, "step": 44940 }, { "grad_norm": 0.3969269394874573, "learning_rate": 4.3919981325692186e-05, "loss": 0.0189, "step": 44950 }, { "grad_norm": 0.16946497559547424, "learning_rate": 4.389946685011024e-05, "loss": 0.0207, "step": 44960 }, { "grad_norm": 0.2905491888523102, "learning_rate": 4.387895341694255e-05, "loss": 0.0356, "step": 44970 }, { "grad_norm": 0.24410490691661835, "learning_rate": 4.3858441029694324e-05, "loss": 0.024, "step": 44980 }, { "grad_norm": 0.24977847933769226, "learning_rate": 4.3837929691870527e-05, "loss": 0.0203, "step": 44990 }, { "grad_norm": 0.30899471044540405, "learning_rate": 4.381741940697604e-05, "loss": 0.0225, "step": 45000 }, { "grad_norm": 0.19498763978481293, "learning_rate": 4.379691017851547e-05, "loss": 0.0191, "step": 45010 }, { "grad_norm": 0.34106072783470154, "learning_rate": 4.3776402009993304e-05, "loss": 0.0284, "step": 45020 }, { "grad_norm": 0.40842291712760925, "learning_rate": 4.3755894904913794e-05, "loss": 0.0229, "step": 45030 }, { "grad_norm": 0.3505430519580841, "learning_rate": 4.373538886678109e-05, "loss": 0.0167, "step": 45040 }, { "grad_norm": 0.26546213030815125, "learning_rate": 4.371488389909909e-05, "loss": 0.0198, "step": 45050 }, { "grad_norm": 0.4006539583206177, "learning_rate": 4.3694380005371515e-05, "loss": 0.0199, "step": 45060 }, { "grad_norm": 0.3870304524898529, "learning_rate": 4.367387718910196e-05, "loss": 0.0229, "step": 45070 }, { "grad_norm": 0.3469510078430176, "learning_rate": 4.3653375453793764e-05, "loss": 0.0164, "step": 45080 }, { "grad_norm": 0.2044171243906021, "learning_rate": 4.3632874802950136e-05, "loss": 0.0212, "step": 45090 }, { "grad_norm": 0.29644083976745605, "learning_rate": 4.3612375240074034e-05, "loss": 0.0206, "step": 45100 }, { "grad_norm": 0.2158868908882141, "learning_rate": 4.3591876768668325e-05, "loss": 0.0197, "step": 45110 }, { "grad_norm": 0.28129932284355164, "learning_rate": 4.3571379392235605e-05, "loss": 0.0163, "step": 45120 }, { "grad_norm": 0.48422953486442566, "learning_rate": 4.3550883114278335e-05, "loss": 0.0262, "step": 45130 }, { "grad_norm": 0.27176764607429504, "learning_rate": 4.353038793829876e-05, "loss": 0.019, "step": 45140 }, { "grad_norm": 0.39769673347473145, "learning_rate": 4.350989386779891e-05, "loss": 0.0194, "step": 45150 }, { "grad_norm": 0.29928550124168396, "learning_rate": 4.3489400906280724e-05, "loss": 0.0312, "step": 45160 }, { "grad_norm": 0.31167125701904297, "learning_rate": 4.3468909057245845e-05, "loss": 0.0223, "step": 45170 }, { "grad_norm": 0.2502678632736206, "learning_rate": 4.3448418324195794e-05, "loss": 0.0248, "step": 45180 }, { "grad_norm": 0.1979970782995224, "learning_rate": 4.342792871063184e-05, "loss": 0.0241, "step": 45190 }, { "grad_norm": 0.24255000054836273, "learning_rate": 4.3407440220055145e-05, "loss": 0.0204, "step": 45200 }, { "grad_norm": 0.22208696603775024, "learning_rate": 4.33869528559666e-05, "loss": 0.0202, "step": 45210 }, { "grad_norm": 0.288930207490921, "learning_rate": 4.336646662186696e-05, "loss": 0.0155, "step": 45220 }, { "grad_norm": 0.24763713777065277, "learning_rate": 4.334598152125672e-05, "loss": 0.0259, "step": 45230 }, { "grad_norm": 0.21549557149410248, "learning_rate": 4.3325497557636276e-05, "loss": 0.0269, "step": 45240 }, { "grad_norm": 0.19819985330104828, "learning_rate": 4.330501473450574e-05, "loss": 0.0221, "step": 45250 }, { "grad_norm": 0.4319770634174347, "learning_rate": 4.328453305536507e-05, "loss": 0.0199, "step": 45260 }, { "grad_norm": 0.2358665019273758, "learning_rate": 4.326405252371404e-05, "loss": 0.0183, "step": 45270 }, { "grad_norm": 0.3162055015563965, "learning_rate": 4.324357314305221e-05, "loss": 0.0219, "step": 45280 }, { "grad_norm": 0.25573208928108215, "learning_rate": 4.3223094916878945e-05, "loss": 0.0156, "step": 45290 }, { "grad_norm": 0.2591626048088074, "learning_rate": 4.320261784869338e-05, "loss": 0.0291, "step": 45300 }, { "grad_norm": 0.1662486493587494, "learning_rate": 4.318214194199455e-05, "loss": 0.025, "step": 45310 }, { "grad_norm": 0.19448143243789673, "learning_rate": 4.316166720028118e-05, "loss": 0.0172, "step": 45320 }, { "grad_norm": 0.32122305035591125, "learning_rate": 4.3141193627051864e-05, "loss": 0.0212, "step": 45330 }, { "grad_norm": 0.34533601999282837, "learning_rate": 4.312072122580496e-05, "loss": 0.0241, "step": 45340 }, { "grad_norm": 0.2413005381822586, "learning_rate": 4.3100250000038646e-05, "loss": 0.0261, "step": 45350 }, { "grad_norm": 0.383046418428421, "learning_rate": 4.307977995325091e-05, "loss": 0.0262, "step": 45360 }, { "grad_norm": 0.2826808989048004, "learning_rate": 4.30593110889395e-05, "loss": 0.0269, "step": 45370 }, { "grad_norm": 0.3102991282939911, "learning_rate": 4.3038843410602016e-05, "loss": 0.0205, "step": 45380 }, { "grad_norm": 0.1839456856250763, "learning_rate": 4.3018376921735774e-05, "loss": 0.0237, "step": 45390 }, { "grad_norm": 0.18204565346240997, "learning_rate": 4.299791162583799e-05, "loss": 0.0168, "step": 45400 }, { "grad_norm": 0.5363615155220032, "learning_rate": 4.29774475264056e-05, "loss": 0.0189, "step": 45410 }, { "grad_norm": 0.3612312078475952, "learning_rate": 4.2956984626935365e-05, "loss": 0.0167, "step": 45420 }, { "grad_norm": 0.16722214221954346, "learning_rate": 4.293652293092383e-05, "loss": 0.0192, "step": 45430 }, { "grad_norm": 0.4959866404533386, "learning_rate": 4.2916062441867324e-05, "loss": 0.0209, "step": 45440 }, { "grad_norm": 0.31360822916030884, "learning_rate": 4.289560316326201e-05, "loss": 0.0219, "step": 45450 }, { "grad_norm": 0.3966946303844452, "learning_rate": 4.28751450986038e-05, "loss": 0.0222, "step": 45460 }, { "grad_norm": 0.41728582978248596, "learning_rate": 4.2854688251388444e-05, "loss": 0.0265, "step": 45470 }, { "grad_norm": 0.31300199031829834, "learning_rate": 4.2834232625111425e-05, "loss": 0.0262, "step": 45480 }, { "grad_norm": 0.2968125641345978, "learning_rate": 4.2813778223268086e-05, "loss": 0.0169, "step": 45490 }, { "grad_norm": 0.2840751111507416, "learning_rate": 4.2793325049353477e-05, "loss": 0.0192, "step": 45500 }, { "grad_norm": 0.22532273828983307, "learning_rate": 4.2772873106862535e-05, "loss": 0.0189, "step": 45510 }, { "grad_norm": 0.305851012468338, "learning_rate": 4.275242239928993e-05, "loss": 0.0231, "step": 45520 }, { "grad_norm": 0.1988649070262909, "learning_rate": 4.273197293013009e-05, "loss": 0.0148, "step": 45530 }, { "grad_norm": 0.1594642698764801, "learning_rate": 4.271152470287731e-05, "loss": 0.0175, "step": 45540 }, { "grad_norm": 0.2895384728908539, "learning_rate": 4.2691077721025594e-05, "loss": 0.0175, "step": 45550 }, { "grad_norm": 0.28398436307907104, "learning_rate": 4.267063198806883e-05, "loss": 0.0201, "step": 45560 }, { "grad_norm": 0.3629932999610901, "learning_rate": 4.2650187507500574e-05, "loss": 0.0155, "step": 45570 }, { "grad_norm": 0.16643774509429932, "learning_rate": 4.2629744282814275e-05, "loss": 0.0184, "step": 45580 }, { "grad_norm": 0.24885320663452148, "learning_rate": 4.2609302317503074e-05, "loss": 0.0148, "step": 45590 }, { "grad_norm": 0.2813657522201538, "learning_rate": 4.258886161505999e-05, "loss": 0.0207, "step": 45600 }, { "grad_norm": 0.26169127225875854, "learning_rate": 4.2568422178977775e-05, "loss": 0.0193, "step": 45610 }, { "grad_norm": 0.2725646197795868, "learning_rate": 4.254798401274894e-05, "loss": 0.0275, "step": 45620 }, { "grad_norm": 0.3149981200695038, "learning_rate": 4.252754711986583e-05, "loss": 0.0192, "step": 45630 }, { "grad_norm": 0.33005306124687195, "learning_rate": 4.250711150382052e-05, "loss": 0.0206, "step": 45640 }, { "grad_norm": 0.23798999190330505, "learning_rate": 4.248667716810495e-05, "loss": 0.0196, "step": 45650 }, { "grad_norm": 0.2206575870513916, "learning_rate": 4.246624411621074e-05, "loss": 0.0206, "step": 45660 }, { "grad_norm": 0.3347111940383911, "learning_rate": 4.244581235162938e-05, "loss": 0.0239, "step": 45670 }, { "grad_norm": 0.1704559028148651, "learning_rate": 4.2425381877852075e-05, "loss": 0.024, "step": 45680 }, { "grad_norm": 0.4441549479961395, "learning_rate": 4.2404952698369856e-05, "loss": 0.0166, "step": 45690 }, { "grad_norm": 0.34527063369750977, "learning_rate": 4.238452481667349e-05, "loss": 0.0221, "step": 45700 }, { "grad_norm": 0.3025987148284912, "learning_rate": 4.2364098236253526e-05, "loss": 0.0275, "step": 45710 }, { "grad_norm": 0.23025576770305634, "learning_rate": 4.2343672960600356e-05, "loss": 0.0224, "step": 45720 }, { "grad_norm": 0.30701354146003723, "learning_rate": 4.232324899320406e-05, "loss": 0.0183, "step": 45730 }, { "grad_norm": 0.23880398273468018, "learning_rate": 4.230282633755457e-05, "loss": 0.0232, "step": 45740 }, { "grad_norm": 0.21966256201267242, "learning_rate": 4.2282404997141515e-05, "loss": 0.0218, "step": 45750 }, { "grad_norm": 0.2618862986564636, "learning_rate": 4.2261984975454397e-05, "loss": 0.02, "step": 45760 }, { "grad_norm": 0.4459080994129181, "learning_rate": 4.224156627598239e-05, "loss": 0.0195, "step": 45770 }, { "grad_norm": 0.2635466158390045, "learning_rate": 4.222114890221453e-05, "loss": 0.0182, "step": 45780 }, { "grad_norm": 0.34875375032424927, "learning_rate": 4.2200732857639546e-05, "loss": 0.023, "step": 45790 }, { "grad_norm": 0.20502381026744843, "learning_rate": 4.2180318145746035e-05, "loss": 0.0156, "step": 45800 }, { "grad_norm": 0.3380125164985657, "learning_rate": 4.215990477002227e-05, "loss": 0.0213, "step": 45810 }, { "grad_norm": 0.42106181383132935, "learning_rate": 4.2139492733956356e-05, "loss": 0.0184, "step": 45820 }, { "grad_norm": 0.3082994520664215, "learning_rate": 4.211908204103615e-05, "loss": 0.0246, "step": 45830 }, { "grad_norm": 0.387215793132782, "learning_rate": 4.2098672694749265e-05, "loss": 0.0255, "step": 45840 }, { "grad_norm": 0.3528743386268616, "learning_rate": 4.2078264698583133e-05, "loss": 0.0285, "step": 45850 }, { "grad_norm": 0.29604843258857727, "learning_rate": 4.205785805602488e-05, "loss": 0.0198, "step": 45860 }, { "grad_norm": 0.3056868016719818, "learning_rate": 4.203745277056149e-05, "loss": 0.0217, "step": 45870 }, { "grad_norm": 0.17955738306045532, "learning_rate": 4.201704884567964e-05, "loss": 0.0217, "step": 45880 }, { "grad_norm": 0.396930456161499, "learning_rate": 4.1996646284865816e-05, "loss": 0.0274, "step": 45890 }, { "grad_norm": 0.23566746711730957, "learning_rate": 4.197624509160625e-05, "loss": 0.0214, "step": 45900 }, { "grad_norm": 0.3825487494468689, "learning_rate": 4.195584526938692e-05, "loss": 0.0166, "step": 45910 }, { "grad_norm": 0.22019968926906586, "learning_rate": 4.193544682169365e-05, "loss": 0.0168, "step": 45920 }, { "grad_norm": 0.45169341564178467, "learning_rate": 4.1915049752011946e-05, "loss": 0.017, "step": 45930 }, { "grad_norm": 0.21143972873687744, "learning_rate": 4.189465406382712e-05, "loss": 0.0203, "step": 45940 }, { "grad_norm": 0.2121296525001526, "learning_rate": 4.187425976062422e-05, "loss": 0.0152, "step": 45950 }, { "grad_norm": 0.379334419965744, "learning_rate": 4.18538668458881e-05, "loss": 0.02, "step": 45960 }, { "grad_norm": 0.2661967873573303, "learning_rate": 4.183347532310333e-05, "loss": 0.0171, "step": 45970 }, { "grad_norm": 0.361305296421051, "learning_rate": 4.181308519575429e-05, "loss": 0.0196, "step": 45980 }, { "grad_norm": 0.3201330900192261, "learning_rate": 4.179269646732507e-05, "loss": 0.0328, "step": 45990 }, { "grad_norm": 0.16447605192661285, "learning_rate": 4.177230914129954e-05, "loss": 0.0207, "step": 46000 }, { "grad_norm": 0.2515504062175751, "learning_rate": 4.175192322116136e-05, "loss": 0.0213, "step": 46010 }, { "grad_norm": 1.6387587785720825, "learning_rate": 4.173153871039391e-05, "loss": 0.0201, "step": 46020 }, { "grad_norm": 0.4434303641319275, "learning_rate": 4.171115561248036e-05, "loss": 0.0231, "step": 46030 }, { "grad_norm": 0.24462367594242096, "learning_rate": 4.16907739309036e-05, "loss": 0.0185, "step": 46040 }, { "grad_norm": 0.2202760875225067, "learning_rate": 4.167039366914633e-05, "loss": 0.0169, "step": 46050 }, { "grad_norm": 0.39723071455955505, "learning_rate": 4.165001483069096e-05, "loss": 0.0224, "step": 46060 }, { "grad_norm": 0.2015564888715744, "learning_rate": 4.16296374190197e-05, "loss": 0.0264, "step": 46070 }, { "grad_norm": 0.2950245440006256, "learning_rate": 4.1609261437614464e-05, "loss": 0.0222, "step": 46080 }, { "grad_norm": 0.4327012300491333, "learning_rate": 4.158888688995696e-05, "loss": 0.0243, "step": 46090 }, { "grad_norm": 0.3183061182498932, "learning_rate": 4.1568513779528645e-05, "loss": 0.0165, "step": 46100 }, { "grad_norm": 0.301245778799057, "learning_rate": 4.1548142109810704e-05, "loss": 0.0259, "step": 46110 }, { "grad_norm": 0.3642260432243347, "learning_rate": 4.152777188428414e-05, "loss": 0.0214, "step": 46120 }, { "grad_norm": 0.19782671332359314, "learning_rate": 4.1507403106429646e-05, "loss": 0.0164, "step": 46130 }, { "grad_norm": 0.32959043979644775, "learning_rate": 4.148703577972768e-05, "loss": 0.0189, "step": 46140 }, { "grad_norm": 0.31414294242858887, "learning_rate": 4.146666990765846e-05, "loss": 0.021, "step": 46150 }, { "grad_norm": 0.2817588448524475, "learning_rate": 4.1446305493701986e-05, "loss": 0.0222, "step": 46160 }, { "grad_norm": 0.2732887268066406, "learning_rate": 4.142594254133796e-05, "loss": 0.0196, "step": 46170 }, { "grad_norm": 0.3815535306930542, "learning_rate": 4.1405581054045835e-05, "loss": 0.0218, "step": 46180 }, { "grad_norm": 0.3320947587490082, "learning_rate": 4.1385221035304864e-05, "loss": 0.0206, "step": 46190 }, { "grad_norm": 0.30567046999931335, "learning_rate": 4.136486248859398e-05, "loss": 0.0214, "step": 46200 }, { "grad_norm": 0.16937817633152008, "learning_rate": 4.1344505417391955e-05, "loss": 0.0226, "step": 46210 }, { "grad_norm": 0.2575649917125702, "learning_rate": 4.132414982517721e-05, "loss": 0.0193, "step": 46220 }, { "grad_norm": 0.21012289822101593, "learning_rate": 4.130379571542798e-05, "loss": 0.0222, "step": 46230 }, { "grad_norm": 0.25756382942199707, "learning_rate": 4.12834430916222e-05, "loss": 0.0188, "step": 46240 }, { "grad_norm": 0.23454239964485168, "learning_rate": 4.126309195723763e-05, "loss": 0.0212, "step": 46250 }, { "grad_norm": 0.28004905581474304, "learning_rate": 4.124274231575168e-05, "loss": 0.0244, "step": 46260 }, { "grad_norm": 0.19755998253822327, "learning_rate": 4.122239417064154e-05, "loss": 0.0183, "step": 46270 }, { "grad_norm": 0.12978389859199524, "learning_rate": 4.1202047525384184e-05, "loss": 0.0127, "step": 46280 }, { "grad_norm": 0.2576746642589569, "learning_rate": 4.118170238345627e-05, "loss": 0.0195, "step": 46290 }, { "grad_norm": 0.207465261220932, "learning_rate": 4.1161358748334256e-05, "loss": 0.0184, "step": 46300 }, { "grad_norm": 0.2562565207481384, "learning_rate": 4.1141016623494266e-05, "loss": 0.0171, "step": 46310 }, { "grad_norm": 0.3163454532623291, "learning_rate": 4.112067601241227e-05, "loss": 0.0212, "step": 46320 }, { "grad_norm": 0.23203612864017487, "learning_rate": 4.110033691856387e-05, "loss": 0.0185, "step": 46330 }, { "grad_norm": 0.45656439661979675, "learning_rate": 4.107999934542451e-05, "loss": 0.0259, "step": 46340 }, { "grad_norm": 0.19467070698738098, "learning_rate": 4.105966329646928e-05, "loss": 0.0166, "step": 46350 }, { "grad_norm": 0.2354860007762909, "learning_rate": 4.103932877517308e-05, "loss": 0.0178, "step": 46360 }, { "grad_norm": 0.24694374203681946, "learning_rate": 4.101899578501052e-05, "loss": 0.0169, "step": 46370 }, { "grad_norm": 0.267137736082077, "learning_rate": 4.099866432945595e-05, "loss": 0.0255, "step": 46380 }, { "grad_norm": 0.3683546781539917, "learning_rate": 4.097833441198346e-05, "loss": 0.0184, "step": 46390 }, { "grad_norm": 0.2445686161518097, "learning_rate": 4.095800603606685e-05, "loss": 0.0189, "step": 46400 }, { "grad_norm": 0.3175711929798126, "learning_rate": 4.093767920517975e-05, "loss": 0.0226, "step": 46410 }, { "grad_norm": 0.26589682698249817, "learning_rate": 4.091735392279539e-05, "loss": 0.0176, "step": 46420 }, { "grad_norm": 0.17815569043159485, "learning_rate": 4.089703019238685e-05, "loss": 0.0187, "step": 46430 }, { "grad_norm": 0.26920685172080994, "learning_rate": 4.0876708017426866e-05, "loss": 0.0197, "step": 46440 }, { "grad_norm": 0.3925413191318512, "learning_rate": 4.085638740138798e-05, "loss": 0.0189, "step": 46450 }, { "grad_norm": 0.3309965133666992, "learning_rate": 4.08360683477424e-05, "loss": 0.024, "step": 46460 }, { "grad_norm": 0.25648149847984314, "learning_rate": 4.0815750859962085e-05, "loss": 0.0208, "step": 46470 }, { "grad_norm": 0.30276182293891907, "learning_rate": 4.079543494151879e-05, "loss": 0.0239, "step": 46480 }, { "grad_norm": 0.2625916004180908, "learning_rate": 4.07751205958839e-05, "loss": 0.015, "step": 46490 }, { "grad_norm": 0.3186560273170471, "learning_rate": 4.0754807826528615e-05, "loss": 0.0179, "step": 46500 }, { "grad_norm": 0.23724643886089325, "learning_rate": 4.073449663692379e-05, "loss": 0.025, "step": 46510 }, { "grad_norm": 0.1651175618171692, "learning_rate": 4.07141870305401e-05, "loss": 0.0187, "step": 46520 }, { "grad_norm": 0.41452866792678833, "learning_rate": 4.0693879010847866e-05, "loss": 0.0384, "step": 46530 }, { "grad_norm": 0.32459744811058044, "learning_rate": 4.06735725813172e-05, "loss": 0.0196, "step": 46540 }, { "grad_norm": 0.31156429648399353, "learning_rate": 4.0653267745417903e-05, "loss": 0.0172, "step": 46550 }, { "grad_norm": 0.15243521332740784, "learning_rate": 4.063296450661949e-05, "loss": 0.0138, "step": 46560 }, { "grad_norm": 0.23562052845954895, "learning_rate": 4.061266286839128e-05, "loss": 0.0228, "step": 46570 }, { "grad_norm": 0.6111936569213867, "learning_rate": 4.0592362834202225e-05, "loss": 0.0205, "step": 46580 }, { "grad_norm": 0.23496460914611816, "learning_rate": 4.057206440752107e-05, "loss": 0.0139, "step": 46590 }, { "grad_norm": 0.21927808225154877, "learning_rate": 4.0551767591816245e-05, "loss": 0.0208, "step": 46600 }, { "grad_norm": 0.2755119800567627, "learning_rate": 4.0531472390555935e-05, "loss": 0.0222, "step": 46610 }, { "grad_norm": 0.41206103563308716, "learning_rate": 4.051117880720802e-05, "loss": 0.0213, "step": 46620 }, { "grad_norm": 0.2665349543094635, "learning_rate": 4.049088684524015e-05, "loss": 0.0222, "step": 46630 }, { "grad_norm": 0.29353511333465576, "learning_rate": 4.0470596508119636e-05, "loss": 0.0226, "step": 46640 }, { "grad_norm": 0.4278445839881897, "learning_rate": 4.0450307799313524e-05, "loss": 0.0234, "step": 46650 }, { "grad_norm": 0.25212839245796204, "learning_rate": 4.0430020722288656e-05, "loss": 0.0212, "step": 46660 }, { "grad_norm": 0.2227126657962799, "learning_rate": 4.040973528051148e-05, "loss": 0.0215, "step": 46670 }, { "grad_norm": 0.4009597897529602, "learning_rate": 4.038945147744827e-05, "loss": 0.0231, "step": 46680 }, { "grad_norm": 0.2957228720188141, "learning_rate": 4.0369169316564945e-05, "loss": 0.0223, "step": 46690 }, { "grad_norm": 0.2944653034210205, "learning_rate": 4.034888880132718e-05, "loss": 0.0234, "step": 46700 }, { "grad_norm": 0.36374664306640625, "learning_rate": 4.032860993520035e-05, "loss": 0.0315, "step": 46710 }, { "grad_norm": 0.2847362458705902, "learning_rate": 4.030833272164959e-05, "loss": 0.0231, "step": 46720 }, { "grad_norm": 0.2979359030723572, "learning_rate": 4.0288057164139705e-05, "loss": 0.0259, "step": 46730 }, { "grad_norm": 0.26684486865997314, "learning_rate": 4.0267783266135205e-05, "loss": 0.0203, "step": 46740 }, { "grad_norm": 0.2557869553565979, "learning_rate": 4.024751103110039e-05, "loss": 0.0216, "step": 46750 }, { "grad_norm": 0.30987846851348877, "learning_rate": 4.0227240462499176e-05, "loss": 0.0319, "step": 46760 }, { "grad_norm": 0.31558874249458313, "learning_rate": 4.020697156379531e-05, "loss": 0.021, "step": 46770 }, { "grad_norm": 0.3749668300151825, "learning_rate": 4.018670433845215e-05, "loss": 0.0245, "step": 46780 }, { "grad_norm": 0.32743173837661743, "learning_rate": 4.016643878993284e-05, "loss": 0.0218, "step": 46790 }, { "grad_norm": 0.2525940537452698, "learning_rate": 4.014617492170017e-05, "loss": 0.0241, "step": 46800 }, { "grad_norm": 0.27635204792022705, "learning_rate": 4.0125912737216726e-05, "loss": 0.0319, "step": 46810 }, { "grad_norm": 0.32717254757881165, "learning_rate": 4.0105652239944735e-05, "loss": 0.0269, "step": 46820 }, { "grad_norm": 0.23143164813518524, "learning_rate": 4.0085393433346144e-05, "loss": 0.0224, "step": 46830 }, { "grad_norm": 0.39848536252975464, "learning_rate": 4.006513632088268e-05, "loss": 0.0189, "step": 46840 }, { "grad_norm": 0.22740568220615387, "learning_rate": 4.004488090601567e-05, "loss": 0.0203, "step": 46850 }, { "grad_norm": 0.2509952783584595, "learning_rate": 4.002462719220626e-05, "loss": 0.0163, "step": 46860 }, { "grad_norm": 0.229559525847435, "learning_rate": 4.000437518291522e-05, "loss": 0.0223, "step": 46870 }, { "grad_norm": 0.2476794272661209, "learning_rate": 3.9984124881603094e-05, "loss": 0.0206, "step": 46880 }, { "grad_norm": 0.35449883341789246, "learning_rate": 3.9963876291730086e-05, "loss": 0.0198, "step": 46890 }, { "grad_norm": 0.3251779079437256, "learning_rate": 3.994362941675614e-05, "loss": 0.0239, "step": 46900 }, { "grad_norm": 0.3752645254135132, "learning_rate": 3.992338426014088e-05, "loss": 0.0199, "step": 46910 }, { "grad_norm": 0.33704352378845215, "learning_rate": 3.9903140825343636e-05, "loss": 0.0189, "step": 46920 }, { "grad_norm": 0.2040957808494568, "learning_rate": 3.98828991158235e-05, "loss": 0.0163, "step": 46930 }, { "grad_norm": 0.34707367420196533, "learning_rate": 3.9862659135039185e-05, "loss": 0.0265, "step": 46940 }, { "grad_norm": 0.2963702976703644, "learning_rate": 3.984242088644918e-05, "loss": 0.0207, "step": 46950 }, { "grad_norm": 0.19634568691253662, "learning_rate": 3.9822184373511615e-05, "loss": 0.0149, "step": 46960 }, { "grad_norm": 0.3586180806159973, "learning_rate": 3.98019495996844e-05, "loss": 0.0206, "step": 46970 }, { "grad_norm": 0.3066565692424774, "learning_rate": 3.978171656842507e-05, "loss": 0.0206, "step": 46980 }, { "grad_norm": 0.42606696486473083, "learning_rate": 3.976148528319091e-05, "loss": 0.0178, "step": 46990 }, { "grad_norm": 0.18723078072071075, "learning_rate": 3.974125574743888e-05, "loss": 0.0211, "step": 47000 }, { "grad_norm": 0.3315550982952118, "learning_rate": 3.9721027964625686e-05, "loss": 0.0212, "step": 47010 }, { "grad_norm": 0.4023805558681488, "learning_rate": 3.9700801938207676e-05, "loss": 0.0187, "step": 47020 }, { "grad_norm": 0.21030904352664948, "learning_rate": 3.9680577671640916e-05, "loss": 0.024, "step": 47030 }, { "grad_norm": 0.27017083764076233, "learning_rate": 3.966035516838121e-05, "loss": 0.0257, "step": 47040 }, { "grad_norm": 0.8135321140289307, "learning_rate": 3.9640134431884014e-05, "loss": 0.0191, "step": 47050 }, { "grad_norm": 0.15112613141536713, "learning_rate": 3.961991546560451e-05, "loss": 0.0154, "step": 47060 }, { "grad_norm": 0.18416965007781982, "learning_rate": 3.959969827299753e-05, "loss": 0.0192, "step": 47070 }, { "grad_norm": 0.35186830163002014, "learning_rate": 3.9579482857517684e-05, "loss": 0.0218, "step": 47080 }, { "grad_norm": 0.2601968050003052, "learning_rate": 3.955926922261921e-05, "loss": 0.0189, "step": 47090 }, { "grad_norm": 0.21156692504882812, "learning_rate": 3.9539057371756084e-05, "loss": 0.0204, "step": 47100 }, { "grad_norm": 0.2787921726703644, "learning_rate": 3.951884730838195e-05, "loss": 0.0174, "step": 47110 }, { "grad_norm": 0.3831470012664795, "learning_rate": 3.949863903595012e-05, "loss": 0.0178, "step": 47120 }, { "grad_norm": 0.28940460085868835, "learning_rate": 3.947843255791369e-05, "loss": 0.0228, "step": 47130 }, { "grad_norm": 0.5237001776695251, "learning_rate": 3.9458227877725364e-05, "loss": 0.0274, "step": 47140 }, { "grad_norm": 0.30064570903778076, "learning_rate": 3.943802499883758e-05, "loss": 0.0169, "step": 47150 }, { "grad_norm": 0.30335888266563416, "learning_rate": 3.9417823924702437e-05, "loss": 0.0231, "step": 47160 }, { "grad_norm": 0.28750959038734436, "learning_rate": 3.939762465877178e-05, "loss": 0.0223, "step": 47170 }, { "grad_norm": 0.2738797068595886, "learning_rate": 3.937742720449708e-05, "loss": 0.0202, "step": 47180 }, { "grad_norm": 0.20624610781669617, "learning_rate": 3.9357231565329563e-05, "loss": 0.0176, "step": 47190 }, { "grad_norm": 0.24818432331085205, "learning_rate": 3.933703774472008e-05, "loss": 0.0147, "step": 47200 }, { "grad_norm": 0.2724575996398926, "learning_rate": 3.93168457461192e-05, "loss": 0.0171, "step": 47210 }, { "grad_norm": 0.2609916925430298, "learning_rate": 3.9296655572977216e-05, "loss": 0.0193, "step": 47220 }, { "grad_norm": 0.289302259683609, "learning_rate": 3.927646722874404e-05, "loss": 0.0242, "step": 47230 }, { "grad_norm": 0.26624795794487, "learning_rate": 3.925628071686934e-05, "loss": 0.0199, "step": 47240 }, { "grad_norm": 0.3050977289676666, "learning_rate": 3.9236096040802415e-05, "loss": 0.0264, "step": 47250 }, { "grad_norm": 0.49523845314979553, "learning_rate": 3.9215913203992294e-05, "loss": 0.0276, "step": 47260 }, { "grad_norm": 0.3131139278411865, "learning_rate": 3.9195732209887645e-05, "loss": 0.0217, "step": 47270 }, { "grad_norm": 0.20901906490325928, "learning_rate": 3.9175553061936875e-05, "loss": 0.0195, "step": 47280 }, { "grad_norm": 0.21517352759838104, "learning_rate": 3.9155375763588045e-05, "loss": 0.0255, "step": 47290 }, { "grad_norm": 0.30155062675476074, "learning_rate": 3.913520031828889e-05, "loss": 0.0183, "step": 47300 }, { "grad_norm": 0.3031724691390991, "learning_rate": 3.911502672948685e-05, "loss": 0.0171, "step": 47310 }, { "grad_norm": 0.288991779088974, "learning_rate": 3.9094855000629014e-05, "loss": 0.0178, "step": 47320 }, { "grad_norm": 0.17412294447422028, "learning_rate": 3.907468513516223e-05, "loss": 0.0188, "step": 47330 }, { "grad_norm": 0.34790971875190735, "learning_rate": 3.905451713653294e-05, "loss": 0.0186, "step": 47340 }, { "grad_norm": 0.33780637383461, "learning_rate": 3.903435100818731e-05, "loss": 0.0207, "step": 47350 }, { "grad_norm": 0.4212513864040375, "learning_rate": 3.901418675357117e-05, "loss": 0.021, "step": 47360 }, { "grad_norm": 0.22314321994781494, "learning_rate": 3.8994024376130075e-05, "loss": 0.0267, "step": 47370 }, { "grad_norm": 0.32359007000923157, "learning_rate": 3.8973863879309194e-05, "loss": 0.0196, "step": 47380 }, { "grad_norm": 0.3118680715560913, "learning_rate": 3.8953705266553394e-05, "loss": 0.018, "step": 47390 }, { "grad_norm": 0.2417842000722885, "learning_rate": 3.893354854130727e-05, "loss": 0.0254, "step": 47400 }, { "grad_norm": 0.29590052366256714, "learning_rate": 3.8913393707015006e-05, "loss": 0.0156, "step": 47410 }, { "grad_norm": 0.39286351203918457, "learning_rate": 3.889324076712056e-05, "loss": 0.0252, "step": 47420 }, { "grad_norm": 0.28522607684135437, "learning_rate": 3.8873089725067476e-05, "loss": 0.0183, "step": 47430 }, { "grad_norm": 0.4487517178058624, "learning_rate": 3.885294058429905e-05, "loss": 0.0281, "step": 47440 }, { "grad_norm": 0.296211838722229, "learning_rate": 3.8832793348258206e-05, "loss": 0.02, "step": 47450 }, { "grad_norm": 0.2755186855792999, "learning_rate": 3.881264802038756e-05, "loss": 0.0302, "step": 47460 }, { "grad_norm": 0.501200795173645, "learning_rate": 3.879250460412939e-05, "loss": 0.0284, "step": 47470 }, { "grad_norm": 0.2551383674144745, "learning_rate": 3.8772363102925644e-05, "loss": 0.0167, "step": 47480 }, { "grad_norm": 0.20159004628658295, "learning_rate": 3.875222352021798e-05, "loss": 0.0184, "step": 47490 }, { "grad_norm": 0.2650522291660309, "learning_rate": 3.8732085859447686e-05, "loss": 0.0217, "step": 47500 }, { "grad_norm": 0.3789679706096649, "learning_rate": 3.871195012405575e-05, "loss": 0.0184, "step": 47510 }, { "grad_norm": 0.3661682605743408, "learning_rate": 3.869181631748278e-05, "loss": 0.021, "step": 47520 }, { "grad_norm": 0.3501816689968109, "learning_rate": 3.867168444316915e-05, "loss": 0.0186, "step": 47530 }, { "grad_norm": 0.3751649558544159, "learning_rate": 3.865155450455481e-05, "loss": 0.0212, "step": 47540 }, { "grad_norm": 0.34618401527404785, "learning_rate": 3.8631426505079426e-05, "loss": 0.0199, "step": 47550 }, { "grad_norm": 0.35378918051719666, "learning_rate": 3.8611300448182304e-05, "loss": 0.0315, "step": 47560 }, { "grad_norm": 0.2735805809497833, "learning_rate": 3.859117633730248e-05, "loss": 0.018, "step": 47570 }, { "grad_norm": 0.3636065721511841, "learning_rate": 3.857105417587858e-05, "loss": 0.0246, "step": 47580 }, { "grad_norm": 0.42868685722351074, "learning_rate": 3.855093396734894e-05, "loss": 0.018, "step": 47590 }, { "grad_norm": 0.25281667709350586, "learning_rate": 3.8530815715151545e-05, "loss": 0.031, "step": 47600 }, { "grad_norm": 0.3037244379520416, "learning_rate": 3.851069942272405e-05, "loss": 0.0198, "step": 47610 }, { "grad_norm": 0.23729805648326874, "learning_rate": 3.849058509350382e-05, "loss": 0.0213, "step": 47620 }, { "grad_norm": 0.39833295345306396, "learning_rate": 3.8470472730927783e-05, "loss": 0.0293, "step": 47630 }, { "grad_norm": 0.4321565330028534, "learning_rate": 3.845036233843264e-05, "loss": 0.0163, "step": 47640 }, { "grad_norm": 0.3091018497943878, "learning_rate": 3.843025391945469e-05, "loss": 0.0173, "step": 47650 }, { "grad_norm": 0.34197577834129333, "learning_rate": 3.841014747742992e-05, "loss": 0.0206, "step": 47660 }, { "grad_norm": 0.23543868958950043, "learning_rate": 3.839004301579397e-05, "loss": 0.016, "step": 47670 }, { "grad_norm": 0.39116808772087097, "learning_rate": 3.8369940537982097e-05, "loss": 0.0196, "step": 47680 }, { "grad_norm": 0.2700236439704895, "learning_rate": 3.834984004742933e-05, "loss": 0.0197, "step": 47690 }, { "grad_norm": 0.2008826732635498, "learning_rate": 3.832974154757026e-05, "loss": 0.0151, "step": 47700 }, { "grad_norm": 0.2929617762565613, "learning_rate": 3.830964504183919e-05, "loss": 0.0159, "step": 47710 }, { "grad_norm": 0.22544822096824646, "learning_rate": 3.828955053367003e-05, "loss": 0.0221, "step": 47720 }, { "grad_norm": 0.33313772082328796, "learning_rate": 3.826945802649642e-05, "loss": 0.0167, "step": 47730 }, { "grad_norm": 0.1895698755979538, "learning_rate": 3.824936752375159e-05, "loss": 0.0285, "step": 47740 }, { "grad_norm": 0.2509361803531647, "learning_rate": 3.822927902886848e-05, "loss": 0.019, "step": 47750 }, { "grad_norm": 0.2184525728225708, "learning_rate": 3.8209192545279653e-05, "loss": 0.0147, "step": 47760 }, { "grad_norm": 0.15495993196964264, "learning_rate": 3.8189108076417326e-05, "loss": 0.0176, "step": 47770 }, { "grad_norm": 0.34698012471199036, "learning_rate": 3.816902562571342e-05, "loss": 0.0263, "step": 47780 }, { "grad_norm": 0.25528547167778015, "learning_rate": 3.814894519659944e-05, "loss": 0.0184, "step": 47790 }, { "grad_norm": 0.31716758012771606, "learning_rate": 3.812886679250661e-05, "loss": 0.0169, "step": 47800 }, { "grad_norm": 0.22235020995140076, "learning_rate": 3.810879041686575e-05, "loss": 0.0134, "step": 47810 }, { "grad_norm": 0.5521735548973083, "learning_rate": 3.808871607310741e-05, "loss": 0.0237, "step": 47820 }, { "grad_norm": 0.41263002157211304, "learning_rate": 3.806864376466169e-05, "loss": 0.0208, "step": 47830 }, { "grad_norm": 0.2876332104206085, "learning_rate": 3.804857349495845e-05, "loss": 0.0206, "step": 47840 }, { "grad_norm": 0.40293097496032715, "learning_rate": 3.802850526742713e-05, "loss": 0.0319, "step": 47850 }, { "grad_norm": 0.22451777756214142, "learning_rate": 3.800843908549683e-05, "loss": 0.0146, "step": 47860 }, { "grad_norm": 0.2000725418329239, "learning_rate": 3.7988374952596325e-05, "loss": 0.0166, "step": 47870 }, { "grad_norm": 0.30962637066841125, "learning_rate": 3.7968312872154e-05, "loss": 0.0202, "step": 47880 }, { "grad_norm": 0.3121219873428345, "learning_rate": 3.7948252847597965e-05, "loss": 0.0218, "step": 47890 }, { "grad_norm": 0.3935304284095764, "learning_rate": 3.7928194882355885e-05, "loss": 0.0174, "step": 47900 }, { "grad_norm": 0.39179402589797974, "learning_rate": 3.790813897985515e-05, "loss": 0.021, "step": 47910 }, { "grad_norm": 0.32156312465667725, "learning_rate": 3.7888085143522726e-05, "loss": 0.0237, "step": 47920 }, { "grad_norm": 0.39609092473983765, "learning_rate": 3.7868033376785314e-05, "loss": 0.0219, "step": 47930 }, { "grad_norm": 0.23166526854038239, "learning_rate": 3.784798368306919e-05, "loss": 0.0167, "step": 47940 }, { "grad_norm": 0.24865561723709106, "learning_rate": 3.782793606580029e-05, "loss": 0.0155, "step": 47950 }, { "grad_norm": 0.17583952844142914, "learning_rate": 3.7807890528404205e-05, "loss": 0.0172, "step": 47960 }, { "grad_norm": 0.24235770106315613, "learning_rate": 3.778784707430616e-05, "loss": 0.0213, "step": 47970 }, { "grad_norm": 0.23098471760749817, "learning_rate": 3.776780570693107e-05, "loss": 0.0156, "step": 47980 }, { "grad_norm": 0.15991029143333435, "learning_rate": 3.774776642970342e-05, "loss": 0.0213, "step": 47990 }, { "grad_norm": 0.15817932784557343, "learning_rate": 3.77277292460474e-05, "loss": 0.0173, "step": 48000 }, { "grad_norm": 0.23106607794761658, "learning_rate": 3.770769415938678e-05, "loss": 0.0155, "step": 48010 }, { "grad_norm": 0.30033236742019653, "learning_rate": 3.768766117314506e-05, "loss": 0.0274, "step": 48020 }, { "grad_norm": 0.27283796668052673, "learning_rate": 3.76676302907453e-05, "loss": 0.0154, "step": 48030 }, { "grad_norm": 0.2449711710214615, "learning_rate": 3.764760151561021e-05, "loss": 0.021, "step": 48040 }, { "grad_norm": 0.20381829142570496, "learning_rate": 3.76275748511622e-05, "loss": 0.0228, "step": 48050 }, { "grad_norm": 0.19282500445842743, "learning_rate": 3.7607550300823255e-05, "loss": 0.0207, "step": 48060 }, { "grad_norm": 0.5069169402122498, "learning_rate": 3.7587527868015044e-05, "loss": 0.0359, "step": 48070 }, { "grad_norm": 0.32984310388565063, "learning_rate": 3.756750755615881e-05, "loss": 0.0242, "step": 48080 }, { "grad_norm": 0.3176979720592499, "learning_rate": 3.754748936867553e-05, "loss": 0.0229, "step": 48090 }, { "grad_norm": 0.33272841572761536, "learning_rate": 3.752747330898573e-05, "loss": 0.021, "step": 48100 }, { "grad_norm": 0.21959127485752106, "learning_rate": 3.750745938050962e-05, "loss": 0.0214, "step": 48110 }, { "grad_norm": 0.24070852994918823, "learning_rate": 3.7487447586667025e-05, "loss": 0.0199, "step": 48120 }, { "grad_norm": 0.35878846049308777, "learning_rate": 3.7467437930877426e-05, "loss": 0.0194, "step": 48130 }, { "grad_norm": 0.3075270652770996, "learning_rate": 3.744743041655992e-05, "loss": 0.0182, "step": 48140 }, { "grad_norm": 0.2829923629760742, "learning_rate": 3.742742504713324e-05, "loss": 0.0195, "step": 48150 }, { "grad_norm": 0.27080288529396057, "learning_rate": 3.740742182601576e-05, "loss": 0.0188, "step": 48160 }, { "grad_norm": 0.6607614159584045, "learning_rate": 3.7387420756625464e-05, "loss": 0.0238, "step": 48170 }, { "grad_norm": 0.4045337736606598, "learning_rate": 3.736742184238002e-05, "loss": 0.0198, "step": 48180 }, { "grad_norm": 0.19578443467617035, "learning_rate": 3.7347425086696684e-05, "loss": 0.0222, "step": 48190 }, { "grad_norm": 0.3825362026691437, "learning_rate": 3.732743049299235e-05, "loss": 0.0267, "step": 48200 }, { "grad_norm": 0.33603155612945557, "learning_rate": 3.730743806468354e-05, "loss": 0.0169, "step": 48210 }, { "grad_norm": 0.1892089992761612, "learning_rate": 3.7287447805186436e-05, "loss": 0.0148, "step": 48220 }, { "grad_norm": 0.4077284634113312, "learning_rate": 3.726745971791682e-05, "loss": 0.0203, "step": 48230 }, { "grad_norm": 0.518254816532135, "learning_rate": 3.724747380629008e-05, "loss": 0.032, "step": 48240 }, { "grad_norm": 0.31557783484458923, "learning_rate": 3.72274900737213e-05, "loss": 0.0223, "step": 48250 }, { "grad_norm": 0.3960276246070862, "learning_rate": 3.7207508523625123e-05, "loss": 0.0212, "step": 48260 }, { "grad_norm": 0.18987858295440674, "learning_rate": 3.718752915941588e-05, "loss": 0.0156, "step": 48270 }, { "grad_norm": 0.3525373935699463, "learning_rate": 3.7167551984507464e-05, "loss": 0.0195, "step": 48280 }, { "grad_norm": 0.31637701392173767, "learning_rate": 3.714757700231346e-05, "loss": 0.0153, "step": 48290 }, { "grad_norm": 0.2822297215461731, "learning_rate": 3.712760421624703e-05, "loss": 0.022, "step": 48300 }, { "grad_norm": 0.25838321447372437, "learning_rate": 3.710763362972099e-05, "loss": 0.0179, "step": 48310 }, { "grad_norm": 0.2626938223838806, "learning_rate": 3.708766524614774e-05, "loss": 0.0206, "step": 48320 }, { "grad_norm": 0.21596336364746094, "learning_rate": 3.7067699068939335e-05, "loss": 0.0149, "step": 48330 }, { "grad_norm": 0.19249629974365234, "learning_rate": 3.704773510150748e-05, "loss": 0.0187, "step": 48340 }, { "grad_norm": 0.3968699872493744, "learning_rate": 3.702777334726344e-05, "loss": 0.026, "step": 48350 }, { "grad_norm": 0.18993575870990753, "learning_rate": 3.7007813809618164e-05, "loss": 0.0128, "step": 48360 }, { "grad_norm": 0.3656080961227417, "learning_rate": 3.6987856491982145e-05, "loss": 0.0255, "step": 48370 }, { "grad_norm": 0.2535090446472168, "learning_rate": 3.696790139776558e-05, "loss": 0.0227, "step": 48380 }, { "grad_norm": 0.22810426354408264, "learning_rate": 3.6947948530378235e-05, "loss": 0.0191, "step": 48390 }, { "grad_norm": 0.24368302524089813, "learning_rate": 3.6927997893229516e-05, "loss": 0.0188, "step": 48400 }, { "grad_norm": 0.4004852771759033, "learning_rate": 3.690804948972845e-05, "loss": 0.0175, "step": 48410 }, { "grad_norm": 0.19363488256931305, "learning_rate": 3.688810332328363e-05, "loss": 0.0212, "step": 48420 }, { "grad_norm": 0.21872898936271667, "learning_rate": 3.686815939730336e-05, "loss": 0.0233, "step": 48430 }, { "grad_norm": 0.21981865167617798, "learning_rate": 3.684821771519548e-05, "loss": 0.017, "step": 48440 }, { "grad_norm": 0.2494828999042511, "learning_rate": 3.68282782803675e-05, "loss": 0.0174, "step": 48450 }, { "grad_norm": 0.17606042325496674, "learning_rate": 3.6808341096226504e-05, "loss": 0.0225, "step": 48460 }, { "grad_norm": 0.3939107358455658, "learning_rate": 3.678840616617924e-05, "loss": 0.0194, "step": 48470 }, { "grad_norm": 0.22021907567977905, "learning_rate": 3.6768473493632e-05, "loss": 0.0167, "step": 48480 }, { "grad_norm": 0.2491064965724945, "learning_rate": 3.6748543081990783e-05, "loss": 0.0177, "step": 48490 }, { "grad_norm": 0.19444945454597473, "learning_rate": 3.672861493466112e-05, "loss": 0.0165, "step": 48500 }, { "grad_norm": 0.30357518792152405, "learning_rate": 3.670868905504818e-05, "loss": 0.0205, "step": 48510 }, { "grad_norm": 0.317751407623291, "learning_rate": 3.6688765446556784e-05, "loss": 0.021, "step": 48520 }, { "grad_norm": 0.18440303206443787, "learning_rate": 3.6668844112591276e-05, "loss": 0.0217, "step": 48530 }, { "grad_norm": 0.30213531851768494, "learning_rate": 3.664892505655573e-05, "loss": 0.0165, "step": 48540 }, { "grad_norm": 0.26363831758499146, "learning_rate": 3.662900828185373e-05, "loss": 0.0218, "step": 48550 }, { "grad_norm": 0.19472788274288177, "learning_rate": 3.6609093791888516e-05, "loss": 0.0167, "step": 48560 }, { "grad_norm": 0.33977267146110535, "learning_rate": 3.658918159006292e-05, "loss": 0.0162, "step": 48570 }, { "grad_norm": 0.29281172156333923, "learning_rate": 3.656927167977942e-05, "loss": 0.0243, "step": 48580 }, { "grad_norm": 0.2494223713874817, "learning_rate": 3.654936406444006e-05, "loss": 0.0235, "step": 48590 }, { "grad_norm": 0.2511403262615204, "learning_rate": 3.65294587474465e-05, "loss": 0.0133, "step": 48600 }, { "grad_norm": 0.26908814907073975, "learning_rate": 3.650955573220002e-05, "loss": 0.0484, "step": 48610 }, { "grad_norm": 0.2807060182094574, "learning_rate": 3.648965502210149e-05, "loss": 0.0159, "step": 48620 }, { "grad_norm": 0.3251800239086151, "learning_rate": 3.646975662055142e-05, "loss": 0.0183, "step": 48630 }, { "grad_norm": 0.13060028851032257, "learning_rate": 3.644986053094987e-05, "loss": 0.0175, "step": 48640 }, { "grad_norm": 0.3114323318004608, "learning_rate": 3.642996675669659e-05, "loss": 0.0254, "step": 48650 }, { "grad_norm": 0.23278871178627014, "learning_rate": 3.641007530119083e-05, "loss": 0.0151, "step": 48660 }, { "grad_norm": 0.251547247171402, "learning_rate": 3.639018616783153e-05, "loss": 0.0196, "step": 48670 }, { "grad_norm": 0.21721096336841583, "learning_rate": 3.637029936001719e-05, "loss": 0.0185, "step": 48680 }, { "grad_norm": 0.32113775610923767, "learning_rate": 3.6350414881145886e-05, "loss": 0.0259, "step": 48690 }, { "grad_norm": 0.27557358145713806, "learning_rate": 3.6330532734615386e-05, "loss": 0.0167, "step": 48700 }, { "grad_norm": 0.15773464739322662, "learning_rate": 3.6310652923822975e-05, "loss": 0.0177, "step": 48710 }, { "grad_norm": 0.2111290842294693, "learning_rate": 3.629077545216558e-05, "loss": 0.0184, "step": 48720 }, { "grad_norm": 0.2693169116973877, "learning_rate": 3.627090032303969e-05, "loss": 0.0279, "step": 48730 }, { "grad_norm": 0.2556695342063904, "learning_rate": 3.625102753984146e-05, "loss": 0.029, "step": 48740 }, { "grad_norm": 0.2612229883670807, "learning_rate": 3.623115710596659e-05, "loss": 0.0179, "step": 48750 }, { "grad_norm": 0.4854457676410675, "learning_rate": 3.6211289024810395e-05, "loss": 0.0254, "step": 48760 }, { "grad_norm": 0.25705456733703613, "learning_rate": 3.619142329976777e-05, "loss": 0.0148, "step": 48770 }, { "grad_norm": 0.23158098757266998, "learning_rate": 3.6171559934233247e-05, "loss": 0.0145, "step": 48780 }, { "grad_norm": 0.25483593344688416, "learning_rate": 3.615169893160093e-05, "loss": 0.02, "step": 48790 }, { "grad_norm": 0.24035871028900146, "learning_rate": 3.61318402952645e-05, "loss": 0.0189, "step": 48800 }, { "grad_norm": 0.3229842185974121, "learning_rate": 3.6111984028617285e-05, "loss": 0.0218, "step": 48810 }, { "grad_norm": 0.17513853311538696, "learning_rate": 3.6092130135052134e-05, "loss": 0.0221, "step": 48820 }, { "grad_norm": 0.17577487230300903, "learning_rate": 3.6072278617961584e-05, "loss": 0.021, "step": 48830 }, { "grad_norm": 0.3350175619125366, "learning_rate": 3.605242948073767e-05, "loss": 0.0203, "step": 48840 }, { "grad_norm": 0.27050673961639404, "learning_rate": 3.603258272677212e-05, "loss": 0.0197, "step": 48850 }, { "grad_norm": 0.24933993816375732, "learning_rate": 3.601273835945616e-05, "loss": 0.0189, "step": 48860 }, { "grad_norm": 0.305351585149765, "learning_rate": 3.5992896382180664e-05, "loss": 0.0193, "step": 48870 }, { "grad_norm": 0.283146470785141, "learning_rate": 3.597305679833609e-05, "loss": 0.0194, "step": 48880 }, { "grad_norm": 0.21913059055805206, "learning_rate": 3.595321961131245e-05, "loss": 0.0221, "step": 48890 }, { "grad_norm": 0.21907538175582886, "learning_rate": 3.593338482449942e-05, "loss": 0.0252, "step": 48900 }, { "grad_norm": 0.23935312032699585, "learning_rate": 3.591355244128618e-05, "loss": 0.0312, "step": 48910 }, { "grad_norm": 0.3373996913433075, "learning_rate": 3.589372246506158e-05, "loss": 0.0228, "step": 48920 }, { "grad_norm": 0.24240250885486603, "learning_rate": 3.5873894899213984e-05, "loss": 0.0173, "step": 48930 }, { "grad_norm": 0.2755258083343506, "learning_rate": 3.5854069747131416e-05, "loss": 0.0157, "step": 48940 }, { "grad_norm": 0.2939777672290802, "learning_rate": 3.583424701220143e-05, "loss": 0.0184, "step": 48950 }, { "grad_norm": 0.2431904524564743, "learning_rate": 3.581442669781121e-05, "loss": 0.0207, "step": 48960 }, { "grad_norm": 0.4461933672428131, "learning_rate": 3.579460880734749e-05, "loss": 0.0201, "step": 48970 }, { "grad_norm": 0.3065183162689209, "learning_rate": 3.577479334419657e-05, "loss": 0.0221, "step": 48980 }, { "grad_norm": 0.21597212553024292, "learning_rate": 3.575498031174444e-05, "loss": 0.0199, "step": 48990 }, { "grad_norm": 0.2544010877609253, "learning_rate": 3.573516971337657e-05, "loss": 0.0209, "step": 49000 }, { "grad_norm": 0.4706483781337738, "learning_rate": 3.5715361552478046e-05, "loss": 0.0219, "step": 49010 }, { "grad_norm": 0.2875955402851105, "learning_rate": 3.5695555832433536e-05, "loss": 0.0174, "step": 49020 }, { "grad_norm": 0.4151100218296051, "learning_rate": 3.5675752556627325e-05, "loss": 0.0241, "step": 49030 }, { "grad_norm": 0.33396291732788086, "learning_rate": 3.565595172844322e-05, "loss": 0.0214, "step": 49040 }, { "grad_norm": 0.30971789360046387, "learning_rate": 3.5636153351264666e-05, "loss": 0.0197, "step": 49050 }, { "grad_norm": 0.29692983627319336, "learning_rate": 3.5616357428474655e-05, "loss": 0.0198, "step": 49060 }, { "grad_norm": 0.2546626329421997, "learning_rate": 3.559656396345575e-05, "loss": 0.0205, "step": 49070 }, { "grad_norm": 0.2268538773059845, "learning_rate": 3.5576772959590146e-05, "loss": 0.0197, "step": 49080 }, { "grad_norm": 0.29734566807746887, "learning_rate": 3.5556984420259545e-05, "loss": 0.0183, "step": 49090 }, { "grad_norm": 0.1727256327867508, "learning_rate": 3.5537198348845305e-05, "loss": 0.0159, "step": 49100 }, { "grad_norm": 0.2515961825847626, "learning_rate": 3.551741474872831e-05, "loss": 0.0198, "step": 49110 }, { "grad_norm": 0.22694554924964905, "learning_rate": 3.549763362328903e-05, "loss": 0.0214, "step": 49120 }, { "grad_norm": 0.2833094000816345, "learning_rate": 3.5477854975907515e-05, "loss": 0.0202, "step": 49130 }, { "grad_norm": 0.22448331117630005, "learning_rate": 3.5458078809963416e-05, "loss": 0.0202, "step": 49140 }, { "grad_norm": 0.398351788520813, "learning_rate": 3.543830512883594e-05, "loss": 0.0185, "step": 49150 }, { "grad_norm": 0.3179776072502136, "learning_rate": 3.5418533935903824e-05, "loss": 0.0173, "step": 49160 }, { "grad_norm": 0.36619833111763, "learning_rate": 3.539876523454547e-05, "loss": 0.0225, "step": 49170 }, { "grad_norm": 0.22787293791770935, "learning_rate": 3.537899902813878e-05, "loss": 0.0165, "step": 49180 }, { "grad_norm": 0.27733147144317627, "learning_rate": 3.5359235320061293e-05, "loss": 0.0222, "step": 49190 }, { "grad_norm": 0.18013115227222443, "learning_rate": 3.533947411369003e-05, "loss": 0.0145, "step": 49200 }, { "grad_norm": 0.19986172020435333, "learning_rate": 3.53197154124017e-05, "loss": 0.0183, "step": 49210 }, { "grad_norm": 0.20761437714099884, "learning_rate": 3.52999592195725e-05, "loss": 0.0186, "step": 49220 }, { "grad_norm": 0.24530445039272308, "learning_rate": 3.5280205538578224e-05, "loss": 0.014, "step": 49230 }, { "grad_norm": 0.290113240480423, "learning_rate": 3.5260454372794236e-05, "loss": 0.0209, "step": 49240 }, { "grad_norm": 0.32371455430984497, "learning_rate": 3.524070572559545e-05, "loss": 0.0178, "step": 49250 }, { "grad_norm": 0.23668691515922546, "learning_rate": 3.5220959600356395e-05, "loss": 0.0167, "step": 49260 }, { "grad_norm": 0.21152956783771515, "learning_rate": 3.5201216000451145e-05, "loss": 0.0142, "step": 49270 }, { "grad_norm": 0.22197279334068298, "learning_rate": 3.5181474929253335e-05, "loss": 0.0168, "step": 49280 }, { "grad_norm": 0.419080525636673, "learning_rate": 3.516173639013615e-05, "loss": 0.0188, "step": 49290 }, { "grad_norm": 0.3634704053401947, "learning_rate": 3.5142000386472406e-05, "loss": 0.0179, "step": 49300 }, { "grad_norm": 0.2485390454530716, "learning_rate": 3.5122266921634427e-05, "loss": 0.0152, "step": 49310 }, { "grad_norm": 0.3243288993835449, "learning_rate": 3.510253599899413e-05, "loss": 0.0215, "step": 49320 }, { "grad_norm": 0.20993396639823914, "learning_rate": 3.5082807621922965e-05, "loss": 0.019, "step": 49330 }, { "grad_norm": 0.3792305290699005, "learning_rate": 3.506308179379201e-05, "loss": 0.0222, "step": 49340 }, { "grad_norm": 0.2933465838432312, "learning_rate": 3.5043358517971844e-05, "loss": 0.0163, "step": 49350 }, { "grad_norm": 0.19134624302387238, "learning_rate": 3.502363779783264e-05, "loss": 0.0153, "step": 49360 }, { "grad_norm": 0.2878609299659729, "learning_rate": 3.500391963674415e-05, "loss": 0.0169, "step": 49370 }, { "grad_norm": 0.3952065408229828, "learning_rate": 3.4984204038075615e-05, "loss": 0.0239, "step": 49380 }, { "grad_norm": 0.2597118616104126, "learning_rate": 3.496449100519595e-05, "loss": 0.0174, "step": 49390 }, { "grad_norm": 0.35515299439430237, "learning_rate": 3.494478054147354e-05, "loss": 0.0194, "step": 49400 }, { "grad_norm": 0.27072036266326904, "learning_rate": 3.4925072650276395e-05, "loss": 0.0156, "step": 49410 }, { "grad_norm": 0.28182902932167053, "learning_rate": 3.4905367334972016e-05, "loss": 0.0152, "step": 49420 }, { "grad_norm": 0.4053989350795746, "learning_rate": 3.488566459892752e-05, "loss": 0.0239, "step": 49430 }, { "grad_norm": 0.28005340695381165, "learning_rate": 3.4865964445509585e-05, "loss": 0.0228, "step": 49440 }, { "grad_norm": 0.36585766077041626, "learning_rate": 3.484626687808438e-05, "loss": 0.02, "step": 49450 }, { "grad_norm": 0.23009303212165833, "learning_rate": 3.4826571900017735e-05, "loss": 0.0138, "step": 49460 }, { "grad_norm": 0.23246566951274872, "learning_rate": 3.480687951467495e-05, "loss": 0.0142, "step": 49470 }, { "grad_norm": 0.24314790964126587, "learning_rate": 3.4787189725420925e-05, "loss": 0.0175, "step": 49480 }, { "grad_norm": 0.14974209666252136, "learning_rate": 3.4767502535620086e-05, "loss": 0.0177, "step": 49490 }, { "grad_norm": 0.3492067754268646, "learning_rate": 3.474781794863648e-05, "loss": 0.0249, "step": 49500 }, { "grad_norm": 0.386655330657959, "learning_rate": 3.472813596783363e-05, "loss": 0.0233, "step": 49510 }, { "grad_norm": 0.2707095444202423, "learning_rate": 3.470845659657466e-05, "loss": 0.0212, "step": 49520 }, { "grad_norm": 0.28907129168510437, "learning_rate": 3.468877983822223e-05, "loss": 0.0174, "step": 49530 }, { "grad_norm": 0.34768936038017273, "learning_rate": 3.466910569613855e-05, "loss": 0.02, "step": 49540 }, { "grad_norm": 0.3076463043689728, "learning_rate": 3.464943417368542e-05, "loss": 0.0244, "step": 49550 }, { "grad_norm": 0.29984238743782043, "learning_rate": 3.462976527422415e-05, "loss": 0.0153, "step": 49560 }, { "grad_norm": 0.38983383774757385, "learning_rate": 3.461009900111562e-05, "loss": 0.0235, "step": 49570 }, { "grad_norm": 0.4476201832294464, "learning_rate": 3.459043535772023e-05, "loss": 0.0218, "step": 49580 }, { "grad_norm": 0.2854754626750946, "learning_rate": 3.4570774347398014e-05, "loss": 0.0236, "step": 49590 }, { "grad_norm": 0.32695910334587097, "learning_rate": 3.4551115973508454e-05, "loss": 0.0213, "step": 49600 }, { "grad_norm": 0.3632450997829437, "learning_rate": 3.453146023941066e-05, "loss": 0.0216, "step": 49610 }, { "grad_norm": 0.3504357635974884, "learning_rate": 3.451180714846325e-05, "loss": 0.0216, "step": 49620 }, { "grad_norm": 0.22153794765472412, "learning_rate": 3.449215670402438e-05, "loss": 0.0179, "step": 49630 }, { "grad_norm": 0.459898442029953, "learning_rate": 3.447250890945181e-05, "loss": 0.0194, "step": 49640 }, { "grad_norm": 0.27883705496788025, "learning_rate": 3.4452863768102754e-05, "loss": 0.0171, "step": 49650 }, { "grad_norm": 0.304015189409256, "learning_rate": 3.443322128333409e-05, "loss": 0.0195, "step": 49660 }, { "grad_norm": 0.38188809156417847, "learning_rate": 3.441358145850215e-05, "loss": 0.0211, "step": 49670 }, { "grad_norm": 0.2538140118122101, "learning_rate": 3.439394429696286e-05, "loss": 0.0245, "step": 49680 }, { "grad_norm": 0.38178062438964844, "learning_rate": 3.4374309802071644e-05, "loss": 0.019, "step": 49690 }, { "grad_norm": 0.18080544471740723, "learning_rate": 3.435467797718353e-05, "loss": 0.0182, "step": 49700 }, { "grad_norm": 0.20904922485351562, "learning_rate": 3.433504882565306e-05, "loss": 0.0212, "step": 49710 }, { "grad_norm": 0.18576578795909882, "learning_rate": 3.43154223508343e-05, "loss": 0.0224, "step": 49720 }, { "grad_norm": 0.2811628580093384, "learning_rate": 3.429579855608089e-05, "loss": 0.0175, "step": 49730 }, { "grad_norm": 0.2845042645931244, "learning_rate": 3.427617744474597e-05, "loss": 0.0192, "step": 49740 }, { "grad_norm": 0.32819271087646484, "learning_rate": 3.425655902018231e-05, "loss": 0.0229, "step": 49750 }, { "grad_norm": 0.2874004542827606, "learning_rate": 3.423694328574211e-05, "loss": 0.0165, "step": 49760 }, { "grad_norm": 0.32934874296188354, "learning_rate": 3.42173302447772e-05, "loss": 0.0209, "step": 49770 }, { "grad_norm": 0.23930111527442932, "learning_rate": 3.419771990063886e-05, "loss": 0.0167, "step": 49780 }, { "grad_norm": 0.17271849513053894, "learning_rate": 3.417811225667803e-05, "loss": 0.0152, "step": 49790 }, { "grad_norm": 0.3446783125400543, "learning_rate": 3.415850731624508e-05, "loss": 0.0218, "step": 49800 }, { "grad_norm": 0.4029639959335327, "learning_rate": 3.4138905082689945e-05, "loss": 0.0224, "step": 49810 }, { "grad_norm": 0.2664554715156555, "learning_rate": 3.4119305559362145e-05, "loss": 0.0232, "step": 49820 }, { "grad_norm": 0.3162483870983124, "learning_rate": 3.4099708749610684e-05, "loss": 0.0232, "step": 49830 }, { "grad_norm": 0.3161710500717163, "learning_rate": 3.408011465678413e-05, "loss": 0.0269, "step": 49840 }, { "grad_norm": 0.2902384400367737, "learning_rate": 3.406052328423055e-05, "loss": 0.0245, "step": 49850 }, { "grad_norm": 0.4210156500339508, "learning_rate": 3.4040934635297615e-05, "loss": 0.0195, "step": 49860 }, { "grad_norm": 0.30918505787849426, "learning_rate": 3.4021348713332466e-05, "loss": 0.0166, "step": 49870 }, { "grad_norm": 0.275447279214859, "learning_rate": 3.4001765521681807e-05, "loss": 0.0177, "step": 49880 }, { "grad_norm": 0.24883782863616943, "learning_rate": 3.398218506369188e-05, "loss": 0.0253, "step": 49890 }, { "grad_norm": 0.31904324889183044, "learning_rate": 3.3962607342708404e-05, "loss": 0.0205, "step": 49900 }, { "grad_norm": 0.27537593245506287, "learning_rate": 3.394303236207673e-05, "loss": 0.0176, "step": 49910 }, { "grad_norm": 0.23949624598026276, "learning_rate": 3.392346012514166e-05, "loss": 0.0202, "step": 49920 }, { "grad_norm": 0.22413764894008636, "learning_rate": 3.390389063524757e-05, "loss": 0.0152, "step": 49930 }, { "grad_norm": 0.17316804826259613, "learning_rate": 3.3884323895738324e-05, "loss": 0.016, "step": 49940 }, { "grad_norm": 0.36180010437965393, "learning_rate": 3.386475990995738e-05, "loss": 0.0224, "step": 49950 }, { "grad_norm": 0.509109616279602, "learning_rate": 3.384519868124765e-05, "loss": 0.0169, "step": 49960 }, { "grad_norm": 0.2630612552165985, "learning_rate": 3.3825640212951645e-05, "loss": 0.0197, "step": 49970 }, { "grad_norm": 0.2319304198026657, "learning_rate": 3.380608450841134e-05, "loss": 0.019, "step": 49980 }, { "grad_norm": 0.3333796560764313, "learning_rate": 3.3786531570968305e-05, "loss": 0.0183, "step": 49990 }, { "grad_norm": 0.2424515187740326, "learning_rate": 3.3766981403963584e-05, "loss": 0.021, "step": 50000 }, { "grad_norm": 0.423819899559021, "learning_rate": 3.374743401073775e-05, "loss": 0.0201, "step": 50010 }, { "grad_norm": 0.40202784538269043, "learning_rate": 3.372788939463095e-05, "loss": 0.0185, "step": 50020 }, { "grad_norm": 0.25690606236457825, "learning_rate": 3.370834755898281e-05, "loss": 0.0175, "step": 50030 }, { "grad_norm": 0.40111902356147766, "learning_rate": 3.3688808507132493e-05, "loss": 0.0211, "step": 50040 }, { "grad_norm": 0.18535448610782623, "learning_rate": 3.3669272242418685e-05, "loss": 0.0162, "step": 50050 }, { "grad_norm": 0.17685139179229736, "learning_rate": 3.364973876817961e-05, "loss": 0.0191, "step": 50060 }, { "grad_norm": 0.3563542366027832, "learning_rate": 3.363020808775299e-05, "loss": 0.0186, "step": 50070 }, { "grad_norm": 0.22387878596782684, "learning_rate": 3.361068020447611e-05, "loss": 0.0151, "step": 50080 }, { "grad_norm": 0.3118166923522949, "learning_rate": 3.3591155121685724e-05, "loss": 0.0166, "step": 50090 }, { "grad_norm": 0.2586359679698944, "learning_rate": 3.357163284271814e-05, "loss": 0.0181, "step": 50100 }, { "grad_norm": 0.3382960259914398, "learning_rate": 3.355211337090919e-05, "loss": 0.0184, "step": 50110 }, { "grad_norm": 0.2548075318336487, "learning_rate": 3.353259670959421e-05, "loss": 0.0205, "step": 50120 }, { "grad_norm": 0.2717721462249756, "learning_rate": 3.351308286210808e-05, "loss": 0.0192, "step": 50130 }, { "grad_norm": 0.22751273214817047, "learning_rate": 3.3493571831785156e-05, "loss": 0.0159, "step": 50140 }, { "grad_norm": 0.28651899099349976, "learning_rate": 3.347406362195936e-05, "loss": 0.0167, "step": 50150 }, { "grad_norm": 0.23090337216854095, "learning_rate": 3.345455823596411e-05, "loss": 0.0203, "step": 50160 }, { "grad_norm": 0.4320247173309326, "learning_rate": 3.3435055677132346e-05, "loss": 0.0179, "step": 50170 }, { "grad_norm": 0.34040430188179016, "learning_rate": 3.3415555948796505e-05, "loss": 0.0174, "step": 50180 }, { "grad_norm": 0.31947532296180725, "learning_rate": 3.3396059054288556e-05, "loss": 0.0172, "step": 50190 }, { "grad_norm": 0.24822160601615906, "learning_rate": 3.3376564996940015e-05, "loss": 0.0282, "step": 50200 }, { "grad_norm": 0.22322769463062286, "learning_rate": 3.3357073780081836e-05, "loss": 0.0155, "step": 50210 }, { "grad_norm": 0.24791760742664337, "learning_rate": 3.333758540704459e-05, "loss": 0.0156, "step": 50220 }, { "grad_norm": 0.26151880621910095, "learning_rate": 3.331809988115827e-05, "loss": 0.0189, "step": 50230 }, { "grad_norm": 0.2974138557910919, "learning_rate": 3.329861720575244e-05, "loss": 0.0187, "step": 50240 }, { "grad_norm": 0.23687703907489777, "learning_rate": 3.3279137384156126e-05, "loss": 0.0127, "step": 50250 }, { "grad_norm": 0.39909473061561584, "learning_rate": 3.3259660419697934e-05, "loss": 0.0223, "step": 50260 }, { "grad_norm": 0.2593061327934265, "learning_rate": 3.3240186315705926e-05, "loss": 0.0199, "step": 50270 }, { "grad_norm": 0.15771035850048065, "learning_rate": 3.322071507550769e-05, "loss": 0.0219, "step": 50280 }, { "grad_norm": 0.21298660337924957, "learning_rate": 3.320124670243033e-05, "loss": 0.0184, "step": 50290 }, { "grad_norm": 0.19819393754005432, "learning_rate": 3.318178119980045e-05, "loss": 0.0153, "step": 50300 }, { "grad_norm": 0.303732305765152, "learning_rate": 3.316231857094421e-05, "loss": 0.0205, "step": 50310 }, { "grad_norm": 0.31190210580825806, "learning_rate": 3.31428588191872e-05, "loss": 0.0233, "step": 50320 }, { "grad_norm": 0.452451229095459, "learning_rate": 3.312340194785458e-05, "loss": 0.02, "step": 50330 }, { "grad_norm": 0.31430819630622864, "learning_rate": 3.3103947960270975e-05, "loss": 0.0181, "step": 50340 }, { "grad_norm": 0.4035484790802002, "learning_rate": 3.308449685976058e-05, "loss": 0.0163, "step": 50350 }, { "grad_norm": 0.2726944088935852, "learning_rate": 3.3065048649647024e-05, "loss": 0.0185, "step": 50360 }, { "grad_norm": 0.26862627267837524, "learning_rate": 3.304560333325348e-05, "loss": 0.0206, "step": 50370 }, { "grad_norm": 0.3885861933231354, "learning_rate": 3.3026160913902624e-05, "loss": 0.0207, "step": 50380 }, { "grad_norm": 0.4433850049972534, "learning_rate": 3.300672139491662e-05, "loss": 0.024, "step": 50390 }, { "grad_norm": 0.2436968833208084, "learning_rate": 3.298728477961717e-05, "loss": 0.0162, "step": 50400 }, { "grad_norm": 0.2492787092924118, "learning_rate": 3.296785107132544e-05, "loss": 0.0255, "step": 50410 }, { "grad_norm": 0.37156593799591064, "learning_rate": 3.2948420273362156e-05, "loss": 0.0186, "step": 50420 }, { "grad_norm": 0.253379762172699, "learning_rate": 3.292899238904747e-05, "loss": 0.0182, "step": 50430 }, { "grad_norm": 0.23460917174816132, "learning_rate": 3.29095674217011e-05, "loss": 0.0209, "step": 50440 }, { "grad_norm": 0.273367315530777, "learning_rate": 3.289014537464224e-05, "loss": 0.0187, "step": 50450 }, { "grad_norm": 0.26268231868743896, "learning_rate": 3.287072625118955e-05, "loss": 0.0179, "step": 50460 }, { "grad_norm": 0.5495156049728394, "learning_rate": 3.285131005466129e-05, "loss": 0.0171, "step": 50470 }, { "grad_norm": 0.25998303294181824, "learning_rate": 3.2831896788375105e-05, "loss": 0.0139, "step": 50480 }, { "grad_norm": 0.28766369819641113, "learning_rate": 3.281248645564822e-05, "loss": 0.0193, "step": 50490 }, { "grad_norm": 0.11691438406705856, "learning_rate": 3.2793079059797306e-05, "loss": 0.0235, "step": 50500 }, { "grad_norm": 0.20703266561031342, "learning_rate": 3.277367460413859e-05, "loss": 0.0168, "step": 50510 }, { "grad_norm": 0.332658588886261, "learning_rate": 3.275427309198773e-05, "loss": 0.0232, "step": 50520 }, { "grad_norm": 0.21236379444599152, "learning_rate": 3.273487452665993e-05, "loss": 0.0176, "step": 50530 }, { "grad_norm": 0.2766038477420807, "learning_rate": 3.271547891146986e-05, "loss": 0.019, "step": 50540 }, { "grad_norm": 0.3108541667461395, "learning_rate": 3.269608624973173e-05, "loss": 0.0156, "step": 50550 }, { "grad_norm": 0.32232409715652466, "learning_rate": 3.26766965447592e-05, "loss": 0.0196, "step": 50560 }, { "grad_norm": 0.28133097290992737, "learning_rate": 3.2657309799865424e-05, "loss": 0.0185, "step": 50570 }, { "grad_norm": 1.1816084384918213, "learning_rate": 3.2637926018363084e-05, "loss": 0.0194, "step": 50580 }, { "grad_norm": 0.328844279050827, "learning_rate": 3.261854520356432e-05, "loss": 0.0172, "step": 50590 }, { "grad_norm": 0.14529329538345337, "learning_rate": 3.2599167358780816e-05, "loss": 0.0165, "step": 50600 }, { "grad_norm": 0.3899267911911011, "learning_rate": 3.2579792487323676e-05, "loss": 0.0128, "step": 50610 }, { "grad_norm": 0.3500383198261261, "learning_rate": 3.256042059250358e-05, "loss": 0.0197, "step": 50620 }, { "grad_norm": 0.4992017149925232, "learning_rate": 3.254105167763062e-05, "loss": 0.0263, "step": 50630 }, { "grad_norm": 0.25868016481399536, "learning_rate": 3.252168574601443e-05, "loss": 0.0253, "step": 50640 }, { "grad_norm": 0.3694976270198822, "learning_rate": 3.2502322800964115e-05, "loss": 0.0179, "step": 50650 }, { "grad_norm": 0.3445854187011719, "learning_rate": 3.248296284578825e-05, "loss": 0.016, "step": 50660 }, { "grad_norm": 0.19213995337486267, "learning_rate": 3.246360588379497e-05, "loss": 0.0173, "step": 50670 }, { "grad_norm": 0.30935755372047424, "learning_rate": 3.24442519182918e-05, "loss": 0.0329, "step": 50680 }, { "grad_norm": 0.2557614743709564, "learning_rate": 3.2424900952585845e-05, "loss": 0.0163, "step": 50690 }, { "grad_norm": 0.3377721905708313, "learning_rate": 3.240555298998362e-05, "loss": 0.0182, "step": 50700 }, { "grad_norm": 0.3326871991157532, "learning_rate": 3.2386208033791194e-05, "loss": 0.0181, "step": 50710 }, { "grad_norm": 0.2631453275680542, "learning_rate": 3.2366866087314064e-05, "loss": 0.0201, "step": 50720 }, { "grad_norm": 0.26042240858078003, "learning_rate": 3.234752715385727e-05, "loss": 0.0211, "step": 50730 }, { "grad_norm": 0.39029526710510254, "learning_rate": 3.232819123672529e-05, "loss": 0.0245, "step": 50740 }, { "grad_norm": 0.4942508935928345, "learning_rate": 3.2308858339222076e-05, "loss": 0.024, "step": 50750 }, { "grad_norm": 0.1967172622680664, "learning_rate": 3.2289528464651144e-05, "loss": 0.0169, "step": 50760 }, { "grad_norm": 0.21269680559635162, "learning_rate": 3.22702016163154e-05, "loss": 0.0217, "step": 50770 }, { "grad_norm": 0.2355351448059082, "learning_rate": 3.2250877797517295e-05, "loss": 0.0175, "step": 50780 }, { "grad_norm": 0.2853463590145111, "learning_rate": 3.223155701155872e-05, "loss": 0.0198, "step": 50790 }, { "grad_norm": 0.26044419407844543, "learning_rate": 3.2212239261741105e-05, "loss": 0.0279, "step": 50800 }, { "grad_norm": 0.33056455850601196, "learning_rate": 3.219292455136528e-05, "loss": 0.0166, "step": 50810 }, { "grad_norm": 0.20141054689884186, "learning_rate": 3.217361288373165e-05, "loss": 0.016, "step": 50820 }, { "grad_norm": 0.16496911644935608, "learning_rate": 3.215430426214002e-05, "loss": 0.0206, "step": 50830 }, { "grad_norm": 0.21982012689113617, "learning_rate": 3.2134998689889705e-05, "loss": 0.0141, "step": 50840 }, { "grad_norm": 0.2098032385110855, "learning_rate": 3.211569617027952e-05, "loss": 0.0196, "step": 50850 }, { "grad_norm": 0.20118635892868042, "learning_rate": 3.209639670660769e-05, "loss": 0.0182, "step": 50860 }, { "grad_norm": 0.5084896087646484, "learning_rate": 3.207710030217202e-05, "loss": 0.0179, "step": 50870 }, { "grad_norm": 0.3038143813610077, "learning_rate": 3.205780696026972e-05, "loss": 0.0168, "step": 50880 }, { "grad_norm": 0.3804613947868347, "learning_rate": 3.203851668419749e-05, "loss": 0.0156, "step": 50890 }, { "grad_norm": 0.2966812252998352, "learning_rate": 3.201922947725149e-05, "loss": 0.0159, "step": 50900 }, { "grad_norm": 0.2029699981212616, "learning_rate": 3.199994534272742e-05, "loss": 0.014, "step": 50910 }, { "grad_norm": 0.1984240561723709, "learning_rate": 3.1980664283920394e-05, "loss": 0.0223, "step": 50920 }, { "grad_norm": 0.33548760414123535, "learning_rate": 3.196138630412499e-05, "loss": 0.018, "step": 50930 }, { "grad_norm": 0.4940071403980255, "learning_rate": 3.194211140663532e-05, "loss": 0.0158, "step": 50940 }, { "grad_norm": 0.3190608024597168, "learning_rate": 3.1922839594744914e-05, "loss": 0.0217, "step": 50950 }, { "grad_norm": 0.2503771185874939, "learning_rate": 3.190357087174683e-05, "loss": 0.0209, "step": 50960 }, { "grad_norm": 0.19394728541374207, "learning_rate": 3.188430524093353e-05, "loss": 0.016, "step": 50970 }, { "grad_norm": 0.299681693315506, "learning_rate": 3.1865042705597014e-05, "loss": 0.0208, "step": 50980 }, { "grad_norm": 0.25565093755722046, "learning_rate": 3.184578326902868e-05, "loss": 0.0215, "step": 50990 }, { "grad_norm": 0.2796597182750702, "learning_rate": 3.182652693451949e-05, "loss": 0.021, "step": 51000 }, { "grad_norm": 0.3655809760093689, "learning_rate": 3.180727370535981e-05, "loss": 0.0171, "step": 51010 }, { "grad_norm": 0.2612963318824768, "learning_rate": 3.178802358483945e-05, "loss": 0.0168, "step": 51020 }, { "grad_norm": 0.3090062439441681, "learning_rate": 3.17687765762478e-05, "loss": 0.0139, "step": 51030 }, { "grad_norm": 0.2388303428888321, "learning_rate": 3.174953268287357e-05, "loss": 0.0224, "step": 51040 }, { "grad_norm": 0.3616490662097931, "learning_rate": 3.173029190800508e-05, "loss": 0.0147, "step": 51050 }, { "grad_norm": 0.3513353168964386, "learning_rate": 3.171105425493e-05, "loss": 0.0188, "step": 51060 }, { "grad_norm": 0.209461972117424, "learning_rate": 3.169181972693557e-05, "loss": 0.0134, "step": 51070 }, { "grad_norm": 0.2005862593650818, "learning_rate": 3.1672588327308404e-05, "loss": 0.0145, "step": 51080 }, { "grad_norm": 0.41792142391204834, "learning_rate": 3.1653360059334646e-05, "loss": 0.0197, "step": 51090 }, { "grad_norm": 0.4668949544429779, "learning_rate": 3.163413492629985e-05, "loss": 0.0292, "step": 51100 }, { "grad_norm": 0.3204275965690613, "learning_rate": 3.161491293148912e-05, "loss": 0.0204, "step": 51110 }, { "grad_norm": 0.18639054894447327, "learning_rate": 3.1595694078186925e-05, "loss": 0.0187, "step": 51120 }, { "grad_norm": 0.37990802526474, "learning_rate": 3.157647836967724e-05, "loss": 0.0166, "step": 51130 }, { "grad_norm": 0.34131312370300293, "learning_rate": 3.155726580924354e-05, "loss": 0.0168, "step": 51140 }, { "grad_norm": 0.25966134667396545, "learning_rate": 3.1538056400168676e-05, "loss": 0.0289, "step": 51150 }, { "grad_norm": 0.19298908114433289, "learning_rate": 3.151885014573506e-05, "loss": 0.0287, "step": 51160 }, { "grad_norm": 0.30408304929733276, "learning_rate": 3.149964704922448e-05, "loss": 0.0273, "step": 51170 }, { "grad_norm": 0.3361104428768158, "learning_rate": 3.148044711391825e-05, "loss": 0.0167, "step": 51180 }, { "grad_norm": 0.23617655038833618, "learning_rate": 3.146125034309707e-05, "loss": 0.0192, "step": 51190 }, { "grad_norm": 0.18663719296455383, "learning_rate": 3.1442056740041195e-05, "loss": 0.0248, "step": 51200 }, { "grad_norm": 0.2845157980918884, "learning_rate": 3.142286630803026e-05, "loss": 0.0313, "step": 51210 }, { "grad_norm": 0.20207944512367249, "learning_rate": 3.140367905034337e-05, "loss": 0.0179, "step": 51220 }, { "grad_norm": 0.23558366298675537, "learning_rate": 3.138449497025914e-05, "loss": 0.0177, "step": 51230 }, { "grad_norm": 0.2997815012931824, "learning_rate": 3.136531407105557e-05, "loss": 0.0173, "step": 51240 }, { "grad_norm": 0.20352302491664886, "learning_rate": 3.1346136356010184e-05, "loss": 0.0166, "step": 51250 }, { "grad_norm": 0.1510700285434723, "learning_rate": 3.132696182839988e-05, "loss": 0.0165, "step": 51260 }, { "grad_norm": 0.25168076157569885, "learning_rate": 3.130779049150111e-05, "loss": 0.0208, "step": 51270 }, { "grad_norm": 0.24469277262687683, "learning_rate": 3.128862234858971e-05, "loss": 0.0189, "step": 51280 }, { "grad_norm": 0.18192709982395172, "learning_rate": 3.1269457402941005e-05, "loss": 0.0172, "step": 51290 }, { "grad_norm": 0.20573118329048157, "learning_rate": 3.125029565782974e-05, "loss": 0.0181, "step": 51300 }, { "grad_norm": 0.26786985993385315, "learning_rate": 3.123113711653013e-05, "loss": 0.0161, "step": 51310 }, { "grad_norm": 0.19112946093082428, "learning_rate": 3.121198178231587e-05, "loss": 0.0188, "step": 51320 }, { "grad_norm": 0.38195371627807617, "learning_rate": 3.119282965846006e-05, "loss": 0.0184, "step": 51330 }, { "grad_norm": 0.3260185420513153, "learning_rate": 3.1173680748235303e-05, "loss": 0.0232, "step": 51340 }, { "grad_norm": 0.27555134892463684, "learning_rate": 3.115453505491358e-05, "loss": 0.0205, "step": 51350 }, { "grad_norm": 0.3170340657234192, "learning_rate": 3.1135392581766404e-05, "loss": 0.0212, "step": 51360 }, { "grad_norm": 0.339123010635376, "learning_rate": 3.111625333206467e-05, "loss": 0.016, "step": 51370 }, { "grad_norm": 0.15797089040279388, "learning_rate": 3.109711730907881e-05, "loss": 0.0135, "step": 51380 }, { "grad_norm": 0.30305975675582886, "learning_rate": 3.107798451607859e-05, "loss": 0.016, "step": 51390 }, { "grad_norm": 0.23283033072948456, "learning_rate": 3.105885495633328e-05, "loss": 0.0189, "step": 51400 }, { "grad_norm": 0.13071954250335693, "learning_rate": 3.103972863311163e-05, "loss": 0.0139, "step": 51410 }, { "grad_norm": 0.20810961723327637, "learning_rate": 3.102060554968178e-05, "loss": 0.0176, "step": 51420 }, { "grad_norm": 0.2914315164089203, "learning_rate": 3.100148570931137e-05, "loss": 0.0192, "step": 51430 }, { "grad_norm": 0.33991485834121704, "learning_rate": 3.098236911526744e-05, "loss": 0.0175, "step": 51440 }, { "grad_norm": 0.3107491433620453, "learning_rate": 3.0963255770816495e-05, "loss": 0.0163, "step": 51450 }, { "grad_norm": 0.3583585321903229, "learning_rate": 3.094414567922447e-05, "loss": 0.0198, "step": 51460 }, { "grad_norm": 0.17969278991222382, "learning_rate": 3.092503884375678e-05, "loss": 0.0176, "step": 51470 }, { "grad_norm": 0.528134286403656, "learning_rate": 3.0905935267678255e-05, "loss": 0.018, "step": 51480 }, { "grad_norm": 0.16637548804283142, "learning_rate": 3.0886834954253166e-05, "loss": 0.0212, "step": 51490 }, { "grad_norm": 0.3139779567718506, "learning_rate": 3.086773790674524e-05, "loss": 0.0195, "step": 51500 }, { "grad_norm": 0.2084810733795166, "learning_rate": 3.0848644128417617e-05, "loss": 0.0177, "step": 51510 }, { "grad_norm": 0.21508370339870453, "learning_rate": 3.082955362253294e-05, "loss": 0.0249, "step": 51520 }, { "grad_norm": 0.38548293709754944, "learning_rate": 3.081046639235323e-05, "loss": 0.0182, "step": 51530 }, { "grad_norm": 0.3034413754940033, "learning_rate": 3.0791382441139986e-05, "loss": 0.0145, "step": 51540 }, { "grad_norm": 0.256417840719223, "learning_rate": 3.077230177215411e-05, "loss": 0.023, "step": 51550 }, { "grad_norm": 0.17680513858795166, "learning_rate": 3.0753224388656e-05, "loss": 0.0149, "step": 51560 }, { "grad_norm": 0.12300707399845123, "learning_rate": 3.073415029390544e-05, "loss": 0.0164, "step": 51570 }, { "grad_norm": 0.30172476172447205, "learning_rate": 3.071507949116166e-05, "loss": 0.0153, "step": 51580 }, { "grad_norm": 0.2863595187664032, "learning_rate": 3.069601198368337e-05, "loss": 0.0139, "step": 51590 }, { "grad_norm": 0.2122754454612732, "learning_rate": 3.067694777472864e-05, "loss": 0.0125, "step": 51600 }, { "grad_norm": 0.37517720460891724, "learning_rate": 3.065788686755508e-05, "loss": 0.0207, "step": 51610 }, { "grad_norm": 0.25744307041168213, "learning_rate": 3.063882926541961e-05, "loss": 0.0152, "step": 51620 }, { "grad_norm": 0.30423927307128906, "learning_rate": 3.061977497157872e-05, "loss": 0.0166, "step": 51630 }, { "grad_norm": 0.3326818346977234, "learning_rate": 3.0600723989288224e-05, "loss": 0.0191, "step": 51640 }, { "grad_norm": 0.25568586587905884, "learning_rate": 3.058167632180343e-05, "loss": 0.0187, "step": 51650 }, { "grad_norm": 0.21728113293647766, "learning_rate": 3.0562631972379074e-05, "loss": 0.0168, "step": 51660 }, { "grad_norm": 0.2479478418827057, "learning_rate": 3.0543590944269276e-05, "loss": 0.0181, "step": 51670 }, { "grad_norm": 0.49573755264282227, "learning_rate": 3.052455324072766e-05, "loss": 0.0255, "step": 51680 }, { "grad_norm": 0.2724520266056061, "learning_rate": 3.0505518865007244e-05, "loss": 0.0158, "step": 51690 }, { "grad_norm": 0.21580012142658234, "learning_rate": 3.048648782036048e-05, "loss": 0.0154, "step": 51700 }, { "grad_norm": 0.15923769772052765, "learning_rate": 3.0467460110039236e-05, "loss": 0.0191, "step": 51710 }, { "grad_norm": 0.26952505111694336, "learning_rate": 3.0448435737294866e-05, "loss": 0.0182, "step": 51720 }, { "grad_norm": 0.16830125451087952, "learning_rate": 3.042941470537808e-05, "loss": 0.0164, "step": 51730 }, { "grad_norm": 0.2662731111049652, "learning_rate": 3.0410397017539072e-05, "loss": 0.0158, "step": 51740 }, { "grad_norm": 0.28602829575538635, "learning_rate": 3.039138267702742e-05, "loss": 0.0155, "step": 51750 }, { "grad_norm": 0.2626630663871765, "learning_rate": 3.0372371687092193e-05, "loss": 0.0187, "step": 51760 }, { "grad_norm": 0.31528154015541077, "learning_rate": 3.0353364050981823e-05, "loss": 0.0198, "step": 51770 }, { "grad_norm": 0.3753126561641693, "learning_rate": 3.033435977194418e-05, "loss": 0.0175, "step": 51780 }, { "grad_norm": 0.16528001427650452, "learning_rate": 3.0315358853226618e-05, "loss": 0.0148, "step": 51790 }, { "grad_norm": 0.13274730741977692, "learning_rate": 3.0296361298075844e-05, "loss": 0.0177, "step": 51800 }, { "grad_norm": 0.16710902750492096, "learning_rate": 3.027736710973803e-05, "loss": 0.0204, "step": 51810 }, { "grad_norm": 0.21733665466308594, "learning_rate": 3.025837629145875e-05, "loss": 0.0138, "step": 51820 }, { "grad_norm": 0.29033133387565613, "learning_rate": 3.0239388846483048e-05, "loss": 0.0186, "step": 51830 }, { "grad_norm": 0.24561169743537903, "learning_rate": 3.022040477805532e-05, "loss": 0.0216, "step": 51840 }, { "grad_norm": 0.3247775435447693, "learning_rate": 3.020142408941946e-05, "loss": 0.0159, "step": 51850 }, { "grad_norm": 0.237799733877182, "learning_rate": 3.018244678381873e-05, "loss": 0.0175, "step": 51860 }, { "grad_norm": 0.4208906888961792, "learning_rate": 3.0163472864495812e-05, "loss": 0.0229, "step": 51870 }, { "grad_norm": 0.42649850249290466, "learning_rate": 3.014450233469287e-05, "loss": 0.0179, "step": 51880 }, { "grad_norm": 0.25853702425956726, "learning_rate": 3.0125535197651422e-05, "loss": 0.0208, "step": 51890 }, { "grad_norm": 0.20014135539531708, "learning_rate": 3.0106571456612448e-05, "loss": 0.0139, "step": 51900 }, { "grad_norm": 0.31241312623023987, "learning_rate": 3.0087611114816305e-05, "loss": 0.0241, "step": 51910 }, { "grad_norm": 0.17861321568489075, "learning_rate": 3.006865417550284e-05, "loss": 0.014, "step": 51920 }, { "grad_norm": 0.2705116271972656, "learning_rate": 3.0049700641911242e-05, "loss": 0.0165, "step": 51930 }, { "grad_norm": 0.1581663340330124, "learning_rate": 3.0030750517280175e-05, "loss": 0.0198, "step": 51940 }, { "grad_norm": 0.20847982168197632, "learning_rate": 3.0011803804847682e-05, "loss": 0.0206, "step": 51950 }, { "grad_norm": 0.43446263670921326, "learning_rate": 2.9992860507851224e-05, "loss": 0.0202, "step": 51960 }, { "grad_norm": 0.26938697695732117, "learning_rate": 2.9973920629527723e-05, "loss": 0.0166, "step": 51970 }, { "grad_norm": 0.15755237638950348, "learning_rate": 2.9954984173113453e-05, "loss": 0.0189, "step": 51980 }, { "grad_norm": 0.6970972418785095, "learning_rate": 2.993605114184418e-05, "loss": 0.022, "step": 51990 }, { "grad_norm": 0.374846875667572, "learning_rate": 2.9917121538955005e-05, "loss": 0.0265, "step": 52000 }, { "grad_norm": 0.3261309564113617, "learning_rate": 2.98981953676805e-05, "loss": 0.0284, "step": 52010 }, { "grad_norm": 0.1695801317691803, "learning_rate": 2.9879272631254594e-05, "loss": 0.0168, "step": 52020 }, { "grad_norm": 0.21876674890518188, "learning_rate": 2.986035333291073e-05, "loss": 0.0215, "step": 52030 }, { "grad_norm": 0.15781652927398682, "learning_rate": 2.9841437475881646e-05, "loss": 0.023, "step": 52040 }, { "grad_norm": 0.19772745668888092, "learning_rate": 2.9822525063399555e-05, "loss": 0.0176, "step": 52050 }, { "grad_norm": 0.27418214082717896, "learning_rate": 2.9803616098696087e-05, "loss": 0.0182, "step": 52060 }, { "grad_norm": 0.2599453926086426, "learning_rate": 2.9784710585002228e-05, "loss": 0.0167, "step": 52070 }, { "grad_norm": 0.43916189670562744, "learning_rate": 2.9765808525548467e-05, "loss": 0.0149, "step": 52080 }, { "grad_norm": 0.2415178120136261, "learning_rate": 2.974690992356461e-05, "loss": 0.0156, "step": 52090 }, { "grad_norm": 0.2707664966583252, "learning_rate": 2.9728014782279934e-05, "loss": 0.0208, "step": 52100 }, { "grad_norm": 0.3877686858177185, "learning_rate": 2.970912310492307e-05, "loss": 0.0172, "step": 52110 }, { "grad_norm": 0.20554353296756744, "learning_rate": 2.9690234894722134e-05, "loss": 0.019, "step": 52120 }, { "grad_norm": 0.2526033818721771, "learning_rate": 2.9671350154904577e-05, "loss": 0.0195, "step": 52130 }, { "grad_norm": 0.2700880765914917, "learning_rate": 2.9652468888697282e-05, "loss": 0.0177, "step": 52140 }, { "grad_norm": 0.2577906548976898, "learning_rate": 2.9633591099326562e-05, "loss": 0.0158, "step": 52150 }, { "grad_norm": 0.2735428810119629, "learning_rate": 2.9614716790018072e-05, "loss": 0.0166, "step": 52160 }, { "grad_norm": 0.331615686416626, "learning_rate": 2.959584596399697e-05, "loss": 0.0166, "step": 52170 }, { "grad_norm": 0.34707099199295044, "learning_rate": 2.9576978624487717e-05, "loss": 0.0179, "step": 52180 }, { "grad_norm": 0.2761499881744385, "learning_rate": 2.9558114774714263e-05, "loss": 0.0198, "step": 52190 }, { "grad_norm": 0.25737741589546204, "learning_rate": 2.9539254417899897e-05, "loss": 0.019, "step": 52200 }, { "grad_norm": 0.24215784668922424, "learning_rate": 2.9520397557267365e-05, "loss": 0.0127, "step": 52210 }, { "grad_norm": 0.3025803565979004, "learning_rate": 2.9501544196038765e-05, "loss": 0.019, "step": 52220 }, { "grad_norm": 0.28816795349121094, "learning_rate": 2.9482694337435614e-05, "loss": 0.0237, "step": 52230 }, { "grad_norm": 0.2770148515701294, "learning_rate": 2.946384798467887e-05, "loss": 0.0172, "step": 52240 }, { "grad_norm": 0.13846711814403534, "learning_rate": 2.9445005140988824e-05, "loss": 0.0124, "step": 52250 }, { "grad_norm": 0.22714059054851532, "learning_rate": 2.942616580958524e-05, "loss": 0.0194, "step": 52260 }, { "grad_norm": 0.32288363575935364, "learning_rate": 2.9407329993687193e-05, "loss": 0.0183, "step": 52270 }, { "grad_norm": 0.21091370284557343, "learning_rate": 2.938849769651326e-05, "loss": 0.0161, "step": 52280 }, { "grad_norm": 0.22968798875808716, "learning_rate": 2.9369668921281345e-05, "loss": 0.0169, "step": 52290 }, { "grad_norm": 0.3842567503452301, "learning_rate": 2.9350843671208773e-05, "loss": 0.0143, "step": 52300 }, { "grad_norm": 0.2946847379207611, "learning_rate": 2.9332021949512245e-05, "loss": 0.0156, "step": 52310 }, { "grad_norm": 0.3861977756023407, "learning_rate": 2.9313203759407908e-05, "loss": 0.0182, "step": 52320 }, { "grad_norm": 0.21229401230812073, "learning_rate": 2.929438910411127e-05, "loss": 0.0293, "step": 52330 }, { "grad_norm": 0.27970394492149353, "learning_rate": 2.927557798683722e-05, "loss": 0.0164, "step": 52340 }, { "grad_norm": 0.21630153059959412, "learning_rate": 2.9256770410800095e-05, "loss": 0.013, "step": 52350 }, { "grad_norm": 0.20465828478336334, "learning_rate": 2.9237966379213554e-05, "loss": 0.0174, "step": 52360 }, { "grad_norm": 0.29058927297592163, "learning_rate": 2.9219165895290736e-05, "loss": 0.0202, "step": 52370 }, { "grad_norm": 0.2890315353870392, "learning_rate": 2.9200368962244083e-05, "loss": 0.0193, "step": 52380 }, { "grad_norm": 0.2441192865371704, "learning_rate": 2.9181575583285536e-05, "loss": 0.0146, "step": 52390 }, { "grad_norm": 0.25375446677207947, "learning_rate": 2.916278576162632e-05, "loss": 0.0162, "step": 52400 }, { "grad_norm": 0.30440548062324524, "learning_rate": 2.9143999500477116e-05, "loss": 0.0167, "step": 52410 }, { "grad_norm": 0.2527625262737274, "learning_rate": 2.9125216803048004e-05, "loss": 0.0199, "step": 52420 }, { "grad_norm": 0.23538050055503845, "learning_rate": 2.9106437672548403e-05, "loss": 0.015, "step": 52430 }, { "grad_norm": 0.19526177644729614, "learning_rate": 2.9087662112187154e-05, "loss": 0.0133, "step": 52440 }, { "grad_norm": 0.3000887930393219, "learning_rate": 2.9068890125172498e-05, "loss": 0.0147, "step": 52450 }, { "grad_norm": 0.23459266126155853, "learning_rate": 2.9050121714712058e-05, "loss": 0.0162, "step": 52460 }, { "grad_norm": 0.3082418441772461, "learning_rate": 2.9031356884012835e-05, "loss": 0.0212, "step": 52470 }, { "grad_norm": 0.3168932795524597, "learning_rate": 2.9012595636281247e-05, "loss": 0.0197, "step": 52480 }, { "grad_norm": 0.2900688052177429, "learning_rate": 2.8993837974723016e-05, "loss": 0.0294, "step": 52490 }, { "grad_norm": 0.20054222643375397, "learning_rate": 2.8975083902543394e-05, "loss": 0.0178, "step": 52500 }, { "grad_norm": 0.3130894601345062, "learning_rate": 2.895633342294688e-05, "loss": 0.0188, "step": 52510 }, { "grad_norm": 0.2560247480869293, "learning_rate": 2.893758653913744e-05, "loss": 0.0189, "step": 52520 }, { "grad_norm": 0.23333171010017395, "learning_rate": 2.8918843254318423e-05, "loss": 0.0193, "step": 52530 }, { "grad_norm": 0.273172527551651, "learning_rate": 2.8900103571692483e-05, "loss": 0.0225, "step": 52540 }, { "grad_norm": 0.23532423377037048, "learning_rate": 2.8881367494461796e-05, "loss": 0.017, "step": 52550 }, { "grad_norm": 0.173100546002388, "learning_rate": 2.886263502582779e-05, "loss": 0.015, "step": 52560 }, { "grad_norm": 0.2455870509147644, "learning_rate": 2.8843906168991353e-05, "loss": 0.0162, "step": 52570 }, { "grad_norm": 0.3442647159099579, "learning_rate": 2.8825180927152728e-05, "loss": 0.0145, "step": 52580 }, { "grad_norm": 0.3845903277397156, "learning_rate": 2.8806459303511547e-05, "loss": 0.0151, "step": 52590 }, { "grad_norm": 0.28957900404930115, "learning_rate": 2.8787741301266852e-05, "loss": 0.0172, "step": 52600 }, { "grad_norm": 0.33798453211784363, "learning_rate": 2.876902692361697e-05, "loss": 0.0147, "step": 52610 }, { "grad_norm": 0.2961741089820862, "learning_rate": 2.875031617375976e-05, "loss": 0.0207, "step": 52620 }, { "grad_norm": 0.27204567193984985, "learning_rate": 2.8731609054892318e-05, "loss": 0.0175, "step": 52630 }, { "grad_norm": 0.20178090035915375, "learning_rate": 2.871290557021119e-05, "loss": 0.0147, "step": 52640 }, { "grad_norm": 0.2122863382101059, "learning_rate": 2.8694205722912298e-05, "loss": 0.0208, "step": 52650 }, { "grad_norm": 0.2861441969871521, "learning_rate": 2.8675509516190936e-05, "loss": 0.0196, "step": 52660 }, { "grad_norm": 0.22154268622398376, "learning_rate": 2.8656816953241772e-05, "loss": 0.0202, "step": 52670 }, { "grad_norm": 0.9386155605316162, "learning_rate": 2.863812803725887e-05, "loss": 0.0163, "step": 52680 }, { "grad_norm": 0.2824147343635559, "learning_rate": 2.8619442771435623e-05, "loss": 0.0158, "step": 52690 }, { "grad_norm": 0.18756388127803802, "learning_rate": 2.860076115896484e-05, "loss": 0.0202, "step": 52700 }, { "grad_norm": 0.187560573220253, "learning_rate": 2.8582083203038712e-05, "loss": 0.0175, "step": 52710 }, { "grad_norm": 0.17425455152988434, "learning_rate": 2.8563408906848778e-05, "loss": 0.0157, "step": 52720 }, { "grad_norm": 0.392685204744339, "learning_rate": 2.8544738273585993e-05, "loss": 0.018, "step": 52730 }, { "grad_norm": 0.26620885729789734, "learning_rate": 2.8526071306440595e-05, "loss": 0.0156, "step": 52740 }, { "grad_norm": 0.3021385371685028, "learning_rate": 2.850740800860233e-05, "loss": 0.0139, "step": 52750 }, { "grad_norm": 0.1984863132238388, "learning_rate": 2.8488748383260178e-05, "loss": 0.0161, "step": 52760 }, { "grad_norm": 0.20518703758716583, "learning_rate": 2.8470092433602635e-05, "loss": 0.0225, "step": 52770 }, { "grad_norm": 0.2746381461620331, "learning_rate": 2.845144016281745e-05, "loss": 0.0148, "step": 52780 }, { "grad_norm": 0.3195258677005768, "learning_rate": 2.8432791574091744e-05, "loss": 0.0299, "step": 52790 }, { "grad_norm": 0.32491394877433777, "learning_rate": 2.8414146670612134e-05, "loss": 0.0233, "step": 52800 }, { "grad_norm": 0.20757368206977844, "learning_rate": 2.8395505455564446e-05, "loss": 0.016, "step": 52810 }, { "grad_norm": 0.24154536426067352, "learning_rate": 2.837686793213403e-05, "loss": 0.0149, "step": 52820 }, { "grad_norm": 0.3314729630947113, "learning_rate": 2.8358234103505477e-05, "loss": 0.0152, "step": 52830 }, { "grad_norm": 0.2805687189102173, "learning_rate": 2.8339603972862806e-05, "loss": 0.0159, "step": 52840 }, { "grad_norm": 0.24908234179019928, "learning_rate": 2.83209775433894e-05, "loss": 0.0191, "step": 52850 }, { "grad_norm": 0.4209405183792114, "learning_rate": 2.8302354818268013e-05, "loss": 0.0183, "step": 52860 }, { "grad_norm": 0.31903186440467834, "learning_rate": 2.8283735800680754e-05, "loss": 0.0182, "step": 52870 }, { "grad_norm": 0.2451992630958557, "learning_rate": 2.826512049380913e-05, "loss": 0.0205, "step": 52880 }, { "grad_norm": 0.20084527134895325, "learning_rate": 2.8246508900833935e-05, "loss": 0.0191, "step": 52890 }, { "grad_norm": 0.26340746879577637, "learning_rate": 2.822790102493541e-05, "loss": 0.0216, "step": 52900 }, { "grad_norm": 0.20781277120113373, "learning_rate": 2.820929686929314e-05, "loss": 0.0162, "step": 52910 }, { "grad_norm": 0.23107323050498962, "learning_rate": 2.819069643708605e-05, "loss": 0.0163, "step": 52920 }, { "grad_norm": 0.3545750081539154, "learning_rate": 2.8172099731492484e-05, "loss": 0.0168, "step": 52930 }, { "grad_norm": 0.322295606136322, "learning_rate": 2.815350675569003e-05, "loss": 0.0245, "step": 52940 }, { "grad_norm": 0.25748392939567566, "learning_rate": 2.8134917512855807e-05, "loss": 0.0208, "step": 52950 }, { "grad_norm": 0.19704768061637878, "learning_rate": 2.8116332006166134e-05, "loss": 0.0146, "step": 52960 }, { "grad_norm": 0.2777819335460663, "learning_rate": 2.809775023879685e-05, "loss": 0.0226, "step": 52970 }, { "grad_norm": 0.24727016687393188, "learning_rate": 2.807917221392299e-05, "loss": 0.0185, "step": 52980 }, { "grad_norm": 0.22592173516750336, "learning_rate": 2.8060597934719067e-05, "loss": 0.0187, "step": 52990 }, { "grad_norm": 0.5665838718414307, "learning_rate": 2.8042027404358927e-05, "loss": 0.0164, "step": 53000 }, { "grad_norm": 0.3150286078453064, "learning_rate": 2.802346062601572e-05, "loss": 0.0181, "step": 53010 }, { "grad_norm": 0.35300958156585693, "learning_rate": 2.800489760286207e-05, "loss": 0.0215, "step": 53020 }, { "grad_norm": 0.3159750998020172, "learning_rate": 2.7986338338069834e-05, "loss": 0.0176, "step": 53030 }, { "grad_norm": 0.3828384578227997, "learning_rate": 2.7967782834810297e-05, "loss": 0.0215, "step": 53040 }, { "grad_norm": 0.23333941400051117, "learning_rate": 2.794923109625409e-05, "loss": 0.0129, "step": 53050 }, { "grad_norm": 0.35804563760757446, "learning_rate": 2.7930683125571205e-05, "loss": 0.0165, "step": 53060 }, { "grad_norm": 0.13104617595672607, "learning_rate": 2.7912138925930997e-05, "loss": 0.0189, "step": 53070 }, { "grad_norm": 0.2250393033027649, "learning_rate": 2.7893598500502117e-05, "loss": 0.0264, "step": 53080 }, { "grad_norm": 0.37495845556259155, "learning_rate": 2.7875061852452644e-05, "loss": 0.0241, "step": 53090 }, { "grad_norm": 0.39958706498146057, "learning_rate": 2.7856528984949982e-05, "loss": 0.0171, "step": 53100 }, { "grad_norm": 0.2268192172050476, "learning_rate": 2.7837999901160888e-05, "loss": 0.0187, "step": 53110 }, { "grad_norm": 0.21787558495998383, "learning_rate": 2.7819474604251484e-05, "loss": 0.0152, "step": 53120 }, { "grad_norm": 0.2552178204059601, "learning_rate": 2.780095309738725e-05, "loss": 0.02, "step": 53130 }, { "grad_norm": 0.362101674079895, "learning_rate": 2.778243538373294e-05, "loss": 0.0159, "step": 53140 }, { "grad_norm": 0.2139495462179184, "learning_rate": 2.7763921466452826e-05, "loss": 0.0182, "step": 53150 }, { "grad_norm": 0.3151180148124695, "learning_rate": 2.7745411348710336e-05, "loss": 0.0241, "step": 53160 }, { "grad_norm": 0.47484761476516724, "learning_rate": 2.7726905033668395e-05, "loss": 0.0241, "step": 53170 }, { "grad_norm": 0.4432283043861389, "learning_rate": 2.7708402524489214e-05, "loss": 0.0193, "step": 53180 }, { "grad_norm": 0.2861799895763397, "learning_rate": 2.7689903824334364e-05, "loss": 0.0241, "step": 53190 }, { "grad_norm": 0.3062306344509125, "learning_rate": 2.7671408936364785e-05, "loss": 0.0154, "step": 53200 }, { "grad_norm": 0.40353819727897644, "learning_rate": 2.765291786374069e-05, "loss": 0.0147, "step": 53210 }, { "grad_norm": 0.16313782334327698, "learning_rate": 2.7634430609621775e-05, "loss": 0.0131, "step": 53220 }, { "grad_norm": 0.46054545044898987, "learning_rate": 2.7615947177166956e-05, "loss": 0.0197, "step": 53230 }, { "grad_norm": 0.20082324743270874, "learning_rate": 2.7597467569534553e-05, "loss": 0.0148, "step": 53240 }, { "grad_norm": 0.38855451345443726, "learning_rate": 2.757899178988226e-05, "loss": 0.0183, "step": 53250 }, { "grad_norm": 0.27678704261779785, "learning_rate": 2.7560519841367005e-05, "loss": 0.0154, "step": 53260 }, { "grad_norm": 0.21196293830871582, "learning_rate": 2.7542051727145237e-05, "loss": 0.0169, "step": 53270 }, { "grad_norm": 0.3091333210468292, "learning_rate": 2.7523587450372578e-05, "loss": 0.0225, "step": 53280 }, { "grad_norm": 0.31161361932754517, "learning_rate": 2.750512701420409e-05, "loss": 0.0165, "step": 53290 }, { "grad_norm": 0.3729662597179413, "learning_rate": 2.7486670421794158e-05, "loss": 0.0201, "step": 53300 }, { "grad_norm": 0.2272035926580429, "learning_rate": 2.7468217676296515e-05, "loss": 0.0251, "step": 53310 }, { "grad_norm": 0.29625171422958374, "learning_rate": 2.7449768780864226e-05, "loss": 0.0202, "step": 53320 }, { "grad_norm": 0.18946412205696106, "learning_rate": 2.7431323738649724e-05, "loss": 0.0164, "step": 53330 }, { "grad_norm": 0.30484768748283386, "learning_rate": 2.7412882552804713e-05, "loss": 0.018, "step": 53340 }, { "grad_norm": 0.33533981442451477, "learning_rate": 2.739444522648032e-05, "loss": 0.013, "step": 53350 }, { "grad_norm": 0.24424150586128235, "learning_rate": 2.737601176282697e-05, "loss": 0.0217, "step": 53360 }, { "grad_norm": 0.42612650990486145, "learning_rate": 2.735758216499445e-05, "loss": 0.0161, "step": 53370 }, { "grad_norm": 0.2351795881986618, "learning_rate": 2.7339156436131864e-05, "loss": 0.0155, "step": 53380 }, { "grad_norm": 0.2960587441921234, "learning_rate": 2.7320734579387663e-05, "loss": 0.0182, "step": 53390 }, { "grad_norm": 0.21329492330551147, "learning_rate": 2.730231659790966e-05, "loss": 0.0142, "step": 53400 }, { "grad_norm": 0.23776639997959137, "learning_rate": 2.728390249484494e-05, "loss": 0.0227, "step": 53410 }, { "grad_norm": 0.27536606788635254, "learning_rate": 2.7265492273340032e-05, "loss": 0.0187, "step": 53420 }, { "grad_norm": 0.24568119645118713, "learning_rate": 2.7247085936540678e-05, "loss": 0.0151, "step": 53430 }, { "grad_norm": 0.23114529252052307, "learning_rate": 2.722868348759205e-05, "loss": 0.0181, "step": 53440 }, { "grad_norm": 0.3035378158092499, "learning_rate": 2.7210284929638635e-05, "loss": 0.0212, "step": 53450 }, { "grad_norm": 0.3343859910964966, "learning_rate": 2.7191890265824183e-05, "loss": 0.0155, "step": 53460 }, { "grad_norm": 0.20403318107128143, "learning_rate": 2.7173499499291926e-05, "loss": 0.0135, "step": 53470 }, { "grad_norm": 0.29487451910972595, "learning_rate": 2.7155112633184277e-05, "loss": 0.0143, "step": 53480 }, { "grad_norm": 0.22710421681404114, "learning_rate": 2.713672967064307e-05, "loss": 0.0245, "step": 53490 }, { "grad_norm": 0.29128336906433105, "learning_rate": 2.711835061480945e-05, "loss": 0.013, "step": 53500 }, { "grad_norm": 0.2551833391189575, "learning_rate": 2.7099975468823896e-05, "loss": 0.0236, "step": 53510 }, { "grad_norm": 0.30079567432403564, "learning_rate": 2.708160423582622e-05, "loss": 0.0154, "step": 53520 }, { "grad_norm": 0.5588594079017639, "learning_rate": 2.706323691895557e-05, "loss": 0.0191, "step": 53530 }, { "grad_norm": 0.30494916439056396, "learning_rate": 2.70448735213504e-05, "loss": 0.019, "step": 53540 }, { "grad_norm": 0.22703246772289276, "learning_rate": 2.702651404614852e-05, "loss": 0.0165, "step": 53550 }, { "grad_norm": 0.24052929878234863, "learning_rate": 2.7008158496487056e-05, "loss": 0.0148, "step": 53560 }, { "grad_norm": 0.21815170347690582, "learning_rate": 2.6989806875502487e-05, "loss": 0.0151, "step": 53570 }, { "grad_norm": 0.2137247920036316, "learning_rate": 2.6971459186330584e-05, "loss": 0.0125, "step": 53580 }, { "grad_norm": 0.265857458114624, "learning_rate": 2.695311543210648e-05, "loss": 0.0162, "step": 53590 }, { "grad_norm": 0.25186121463775635, "learning_rate": 2.693477561596463e-05, "loss": 0.0162, "step": 53600 }, { "grad_norm": 0.2654254138469696, "learning_rate": 2.6916439741038756e-05, "loss": 0.0166, "step": 53610 }, { "grad_norm": 0.45676007866859436, "learning_rate": 2.689810781046203e-05, "loss": 0.0192, "step": 53620 }, { "grad_norm": 0.25526732206344604, "learning_rate": 2.6879779827366823e-05, "loss": 0.0169, "step": 53630 }, { "grad_norm": 0.24874910712242126, "learning_rate": 2.6861455794884904e-05, "loss": 0.0178, "step": 53640 }, { "grad_norm": 0.22260133922100067, "learning_rate": 2.6843135716147373e-05, "loss": 0.0164, "step": 53650 }, { "grad_norm": 0.26861342787742615, "learning_rate": 2.6824819594284556e-05, "loss": 0.0205, "step": 53660 }, { "grad_norm": 0.2752002775669098, "learning_rate": 2.6806507432426275e-05, "loss": 0.015, "step": 53670 }, { "grad_norm": 0.32883429527282715, "learning_rate": 2.6788199233701512e-05, "loss": 0.0178, "step": 53680 }, { "grad_norm": 0.30939483642578125, "learning_rate": 2.6769895001238652e-05, "loss": 0.0197, "step": 53690 }, { "grad_norm": 0.2608521282672882, "learning_rate": 2.67515947381654e-05, "loss": 0.0155, "step": 53700 }, { "grad_norm": 0.15326568484306335, "learning_rate": 2.6733298447608745e-05, "loss": 0.0147, "step": 53710 }, { "grad_norm": 0.19006270170211792, "learning_rate": 2.6715006132695074e-05, "loss": 0.0211, "step": 53720 }, { "grad_norm": 0.2514991760253906, "learning_rate": 2.6696717796549985e-05, "loss": 0.0245, "step": 53730 }, { "grad_norm": 0.3033209443092346, "learning_rate": 2.667843344229848e-05, "loss": 0.0172, "step": 53740 }, { "grad_norm": 0.4244177043437958, "learning_rate": 2.6660153073064846e-05, "loss": 0.0161, "step": 53750 }, { "grad_norm": 0.26009607315063477, "learning_rate": 2.664187669197271e-05, "loss": 0.0313, "step": 53760 }, { "grad_norm": 0.3270077109336853, "learning_rate": 2.6623604302145005e-05, "loss": 0.0175, "step": 53770 }, { "grad_norm": 0.3877028524875641, "learning_rate": 2.6605335906703975e-05, "loss": 0.021, "step": 53780 }, { "grad_norm": 0.15483561158180237, "learning_rate": 2.65870715087712e-05, "loss": 0.0182, "step": 53790 }, { "grad_norm": 0.3154638409614563, "learning_rate": 2.6568811111467573e-05, "loss": 0.0135, "step": 53800 }, { "grad_norm": 0.14571291208267212, "learning_rate": 2.6550554717913258e-05, "loss": 0.0177, "step": 53810 }, { "grad_norm": 0.25029703974723816, "learning_rate": 2.6532302331227805e-05, "loss": 0.0161, "step": 53820 }, { "grad_norm": 0.29971620440483093, "learning_rate": 2.651405395453004e-05, "loss": 0.0112, "step": 53830 }, { "grad_norm": 0.19263087213039398, "learning_rate": 2.6495809590938115e-05, "loss": 0.0248, "step": 53840 }, { "grad_norm": 0.6343530416488647, "learning_rate": 2.64775692435695e-05, "loss": 0.0153, "step": 53850 }, { "grad_norm": 0.23919261991977692, "learning_rate": 2.6459332915540928e-05, "loss": 0.0181, "step": 53860 }, { "grad_norm": 0.25368255376815796, "learning_rate": 2.644110060996856e-05, "loss": 0.0237, "step": 53870 }, { "grad_norm": 0.2951374053955078, "learning_rate": 2.642287232996774e-05, "loss": 0.0139, "step": 53880 }, { "grad_norm": 0.19448526203632355, "learning_rate": 2.6404648078653205e-05, "loss": 0.0107, "step": 53890 }, { "grad_norm": 0.22208121418952942, "learning_rate": 2.6386427859139002e-05, "loss": 0.0144, "step": 53900 }, { "grad_norm": 0.44277575612068176, "learning_rate": 2.63682116745384e-05, "loss": 0.0222, "step": 53910 }, { "grad_norm": 0.30603528022766113, "learning_rate": 2.6349999527964138e-05, "loss": 0.017, "step": 53920 }, { "grad_norm": 0.3014073371887207, "learning_rate": 2.63317914225281e-05, "loss": 0.0235, "step": 53930 }, { "grad_norm": 0.3006307780742645, "learning_rate": 2.6313587361341585e-05, "loss": 0.0184, "step": 53940 }, { "grad_norm": 0.4430227279663086, "learning_rate": 2.6295387347515165e-05, "loss": 0.025, "step": 53950 }, { "grad_norm": 0.2076224684715271, "learning_rate": 2.6277191384158727e-05, "loss": 0.0194, "step": 53960 }, { "grad_norm": 0.32286369800567627, "learning_rate": 2.6258999474381457e-05, "loss": 0.0208, "step": 53970 }, { "grad_norm": 0.3026307225227356, "learning_rate": 2.624081162129186e-05, "loss": 0.014, "step": 53980 }, { "grad_norm": 0.2642614543437958, "learning_rate": 2.6222627827997765e-05, "loss": 0.026, "step": 53990 }, { "grad_norm": 0.4549041986465454, "learning_rate": 2.6204448097606236e-05, "loss": 0.0168, "step": 54000 }, { "grad_norm": 0.22567477822303772, "learning_rate": 2.6186272433223726e-05, "loss": 0.0198, "step": 54010 }, { "grad_norm": 0.215250164270401, "learning_rate": 2.6168100837955943e-05, "loss": 0.0174, "step": 54020 }, { "grad_norm": 0.2601664066314697, "learning_rate": 2.6149933314907926e-05, "loss": 0.024, "step": 54030 }, { "grad_norm": 0.3555780053138733, "learning_rate": 2.6131769867184e-05, "loss": 0.0259, "step": 54040 }, { "grad_norm": 0.23790152370929718, "learning_rate": 2.611361049788783e-05, "loss": 0.0154, "step": 54050 }, { "grad_norm": 0.3779798746109009, "learning_rate": 2.6095455210122292e-05, "loss": 0.0155, "step": 54060 }, { "grad_norm": 0.1812872737646103, "learning_rate": 2.6077304006989712e-05, "loss": 0.0143, "step": 54070 }, { "grad_norm": 0.2668403387069702, "learning_rate": 2.6059156891591562e-05, "loss": 0.0187, "step": 54080 }, { "grad_norm": 0.24510034918785095, "learning_rate": 2.6041013867028718e-05, "loss": 0.0148, "step": 54090 }, { "grad_norm": 0.36256512999534607, "learning_rate": 2.6022874936401347e-05, "loss": 0.0156, "step": 54100 }, { "grad_norm": 0.33979499340057373, "learning_rate": 2.6004740102808832e-05, "loss": 0.0203, "step": 54110 }, { "grad_norm": 0.25904422998428345, "learning_rate": 2.598660936935e-05, "loss": 0.0175, "step": 54120 }, { "grad_norm": 0.3022860288619995, "learning_rate": 2.5968482739122845e-05, "loss": 0.0223, "step": 54130 }, { "grad_norm": 0.280984491109848, "learning_rate": 2.595036021522472e-05, "loss": 0.0187, "step": 54140 }, { "grad_norm": 0.29735878109931946, "learning_rate": 2.5932241800752278e-05, "loss": 0.0149, "step": 54150 }, { "grad_norm": 0.3547646403312683, "learning_rate": 2.5914127498801453e-05, "loss": 0.0274, "step": 54160 }, { "grad_norm": 0.19388100504875183, "learning_rate": 2.5896017312467497e-05, "loss": 0.0206, "step": 54170 }, { "grad_norm": 0.2582090497016907, "learning_rate": 2.587791124484493e-05, "loss": 0.0123, "step": 54180 }, { "grad_norm": 0.22676895558834076, "learning_rate": 2.5859809299027615e-05, "loss": 0.0187, "step": 54190 }, { "grad_norm": 0.3145180642604828, "learning_rate": 2.5841711478108632e-05, "loss": 0.018, "step": 54200 }, { "grad_norm": 0.34182730317115784, "learning_rate": 2.582361778518043e-05, "loss": 0.0197, "step": 54210 }, { "grad_norm": 0.5589137077331543, "learning_rate": 2.580552822333472e-05, "loss": 0.0213, "step": 54220 }, { "grad_norm": 0.4370710551738739, "learning_rate": 2.578744279566252e-05, "loss": 0.0144, "step": 54230 }, { "grad_norm": 0.280286967754364, "learning_rate": 2.576936150525413e-05, "loss": 0.0224, "step": 54240 }, { "grad_norm": 0.17446833848953247, "learning_rate": 2.5751284355199168e-05, "loss": 0.0137, "step": 54250 }, { "grad_norm": 0.41890013217926025, "learning_rate": 2.573321134858646e-05, "loss": 0.0175, "step": 54260 }, { "grad_norm": 0.23738396167755127, "learning_rate": 2.5715142488504286e-05, "loss": 0.0141, "step": 54270 }, { "grad_norm": 0.22873751819133759, "learning_rate": 2.5697077778040042e-05, "loss": 0.0179, "step": 54280 }, { "grad_norm": 0.22605206072330475, "learning_rate": 2.5679017220280522e-05, "loss": 0.0252, "step": 54290 }, { "grad_norm": 0.21846719086170197, "learning_rate": 2.5660960818311796e-05, "loss": 0.0185, "step": 54300 }, { "grad_norm": 0.3789588212966919, "learning_rate": 2.564290857521915e-05, "loss": 0.0194, "step": 54310 }, { "grad_norm": 0.2623119652271271, "learning_rate": 2.5624860494087298e-05, "loss": 0.0167, "step": 54320 }, { "grad_norm": 0.24994608759880066, "learning_rate": 2.5606816578000115e-05, "loss": 0.0201, "step": 54330 }, { "grad_norm": 0.3460100293159485, "learning_rate": 2.558877683004082e-05, "loss": 0.0153, "step": 54340 }, { "grad_norm": 0.15428876876831055, "learning_rate": 2.557074125329192e-05, "loss": 0.015, "step": 54350 }, { "grad_norm": 0.21410802006721497, "learning_rate": 2.5552709850835195e-05, "loss": 0.0165, "step": 54360 }, { "grad_norm": 0.1926053762435913, "learning_rate": 2.5534682625751738e-05, "loss": 0.0142, "step": 54370 }, { "grad_norm": 0.23549380898475647, "learning_rate": 2.551665958112186e-05, "loss": 0.0267, "step": 54380 }, { "grad_norm": 0.41962239146232605, "learning_rate": 2.549864072002527e-05, "loss": 0.0164, "step": 54390 }, { "grad_norm": 0.26301032304763794, "learning_rate": 2.5480626045540858e-05, "loss": 0.0178, "step": 54400 }, { "grad_norm": 0.29683971405029297, "learning_rate": 2.546261556074684e-05, "loss": 0.0161, "step": 54410 }, { "grad_norm": 0.25983351469039917, "learning_rate": 2.5444609268720726e-05, "loss": 0.015, "step": 54420 }, { "grad_norm": 0.28525710105895996, "learning_rate": 2.5426607172539297e-05, "loss": 0.0166, "step": 54430 }, { "grad_norm": 0.321927934885025, "learning_rate": 2.5408609275278617e-05, "loss": 0.0263, "step": 54440 }, { "grad_norm": 0.16853103041648865, "learning_rate": 2.5390615580014055e-05, "loss": 0.0144, "step": 54450 }, { "grad_norm": 0.1806253343820572, "learning_rate": 2.5372626089820207e-05, "loss": 0.0212, "step": 54460 }, { "grad_norm": 0.27650776505470276, "learning_rate": 2.5354640807770997e-05, "loss": 0.0128, "step": 54470 }, { "grad_norm": 0.3038273751735687, "learning_rate": 2.5336659736939622e-05, "loss": 0.0197, "step": 54480 }, { "grad_norm": 0.2749944031238556, "learning_rate": 2.5318682880398554e-05, "loss": 0.0181, "step": 54490 }, { "grad_norm": 0.22459809482097626, "learning_rate": 2.530071024121956e-05, "loss": 0.0141, "step": 54500 }, { "grad_norm": 0.24479612708091736, "learning_rate": 2.5282741822473627e-05, "loss": 0.0155, "step": 54510 }, { "grad_norm": 0.30798837542533875, "learning_rate": 2.526477762723114e-05, "loss": 0.017, "step": 54520 }, { "grad_norm": 0.20909324288368225, "learning_rate": 2.5246817658561618e-05, "loss": 0.0182, "step": 54530 }, { "grad_norm": 0.19489338994026184, "learning_rate": 2.5228861919533965e-05, "loss": 0.0146, "step": 54540 }, { "grad_norm": 0.19300420582294464, "learning_rate": 2.5210910413216326e-05, "loss": 0.0278, "step": 54550 }, { "grad_norm": 0.19705407321453094, "learning_rate": 2.5192963142676086e-05, "loss": 0.0148, "step": 54560 }, { "grad_norm": 0.25766658782958984, "learning_rate": 2.517502011098001e-05, "loss": 0.0158, "step": 54570 }, { "grad_norm": 0.4084837734699249, "learning_rate": 2.5157081321193987e-05, "loss": 0.0191, "step": 54580 }, { "grad_norm": 0.09495576471090317, "learning_rate": 2.5139146776383356e-05, "loss": 0.0154, "step": 54590 }, { "grad_norm": 0.20614871382713318, "learning_rate": 2.5121216479612575e-05, "loss": 0.0149, "step": 54600 }, { "grad_norm": 0.23791231215000153, "learning_rate": 2.510329043394546e-05, "loss": 0.0214, "step": 54610 }, { "grad_norm": 0.43695104122161865, "learning_rate": 2.508536864244508e-05, "loss": 0.0209, "step": 54620 }, { "grad_norm": 0.23995187878608704, "learning_rate": 2.5067451108173778e-05, "loss": 0.0146, "step": 54630 }, { "grad_norm": 0.23089484870433807, "learning_rate": 2.5049537834193204e-05, "loss": 0.0117, "step": 54640 }, { "grad_norm": 0.2076849490404129, "learning_rate": 2.5031628823564194e-05, "loss": 0.0177, "step": 54650 }, { "grad_norm": 0.3534650206565857, "learning_rate": 2.5013724079346933e-05, "loss": 0.0167, "step": 54660 }, { "grad_norm": 0.3378548324108124, "learning_rate": 2.4995823604600854e-05, "loss": 0.0135, "step": 54670 }, { "grad_norm": 0.32804813981056213, "learning_rate": 2.497792740238465e-05, "loss": 0.0169, "step": 54680 }, { "grad_norm": 0.26016101241111755, "learning_rate": 2.49600354757563e-05, "loss": 0.0216, "step": 54690 }, { "grad_norm": 0.19970352947711945, "learning_rate": 2.494214782777306e-05, "loss": 0.0199, "step": 54700 }, { "grad_norm": 0.36277005076408386, "learning_rate": 2.4924264461491386e-05, "loss": 0.0173, "step": 54710 }, { "grad_norm": 0.28471311926841736, "learning_rate": 2.4906385379967133e-05, "loss": 0.021, "step": 54720 }, { "grad_norm": 0.2294517606496811, "learning_rate": 2.4888510586255285e-05, "loss": 0.0188, "step": 54730 }, { "grad_norm": 0.19133077561855316, "learning_rate": 2.487064008341018e-05, "loss": 0.0169, "step": 54740 }, { "grad_norm": 0.2875811457633972, "learning_rate": 2.4852773874485407e-05, "loss": 0.0157, "step": 54750 }, { "grad_norm": 0.30769580602645874, "learning_rate": 2.483491196253377e-05, "loss": 0.0193, "step": 54760 }, { "grad_norm": 0.46048736572265625, "learning_rate": 2.4817054350607443e-05, "loss": 0.0226, "step": 54770 }, { "grad_norm": 0.22657980024814606, "learning_rate": 2.4799201041757742e-05, "loss": 0.0159, "step": 54780 }, { "grad_norm": 0.3335862159729004, "learning_rate": 2.4781352039035373e-05, "loss": 0.0175, "step": 54790 }, { "grad_norm": 0.30492398142814636, "learning_rate": 2.4763507345490194e-05, "loss": 0.0179, "step": 54800 }, { "grad_norm": 0.33845317363739014, "learning_rate": 2.4745666964171386e-05, "loss": 0.0219, "step": 54810 }, { "grad_norm": 0.3432948887348175, "learning_rate": 2.4727830898127387e-05, "loss": 0.02, "step": 54820 }, { "grad_norm": 0.24745868146419525, "learning_rate": 2.4709999150405895e-05, "loss": 0.0152, "step": 54830 }, { "grad_norm": 0.4101811945438385, "learning_rate": 2.469217172405388e-05, "loss": 0.0179, "step": 54840 }, { "grad_norm": 0.12878011167049408, "learning_rate": 2.4674348622117527e-05, "loss": 0.0117, "step": 54850 }, { "grad_norm": 0.2169317603111267, "learning_rate": 2.465652984764234e-05, "loss": 0.0144, "step": 54860 }, { "grad_norm": 0.24594895541667938, "learning_rate": 2.4638715403673056e-05, "loss": 0.0116, "step": 54870 }, { "grad_norm": 0.28491759300231934, "learning_rate": 2.4620905293253676e-05, "loss": 0.0203, "step": 54880 }, { "grad_norm": 0.21638575196266174, "learning_rate": 2.4603099519427463e-05, "loss": 0.0166, "step": 54890 }, { "grad_norm": 0.29137012362480164, "learning_rate": 2.458529808523695e-05, "loss": 0.0183, "step": 54900 }, { "grad_norm": 0.21828541159629822, "learning_rate": 2.456750099372387e-05, "loss": 0.0151, "step": 54910 }, { "grad_norm": 0.19316411018371582, "learning_rate": 2.4549708247929327e-05, "loss": 0.0188, "step": 54920 }, { "grad_norm": 0.18586014211177826, "learning_rate": 2.4531919850893554e-05, "loss": 0.0127, "step": 54930 }, { "grad_norm": 0.22087258100509644, "learning_rate": 2.4514135805656125e-05, "loss": 0.0152, "step": 54940 }, { "grad_norm": 0.42000502347946167, "learning_rate": 2.449635611525587e-05, "loss": 0.02, "step": 54950 }, { "grad_norm": 0.26805898547172546, "learning_rate": 2.447858078273079e-05, "loss": 0.0206, "step": 54960 }, { "grad_norm": 0.35169073939323425, "learning_rate": 2.4460809811118275e-05, "loss": 0.0323, "step": 54970 }, { "grad_norm": 0.24416323006153107, "learning_rate": 2.444304320345483e-05, "loss": 0.0182, "step": 54980 }, { "grad_norm": 0.31552615761756897, "learning_rate": 2.4425280962776354e-05, "loss": 0.0219, "step": 54990 }, { "grad_norm": 0.4019542634487152, "learning_rate": 2.4407523092117875e-05, "loss": 0.0162, "step": 55000 }, { "grad_norm": 0.33800768852233887, "learning_rate": 2.438976959451374e-05, "loss": 0.0178, "step": 55010 }, { "grad_norm": 0.2798241674900055, "learning_rate": 2.4372020472997565e-05, "loss": 0.015, "step": 55020 }, { "grad_norm": 0.2623806595802307, "learning_rate": 2.4354275730602122e-05, "loss": 0.0182, "step": 55030 }, { "grad_norm": 0.27198275923728943, "learning_rate": 2.433653537035958e-05, "loss": 0.0226, "step": 55040 }, { "grad_norm": 0.26816925406455994, "learning_rate": 2.431879939530123e-05, "loss": 0.0173, "step": 55050 }, { "grad_norm": 0.42596954107284546, "learning_rate": 2.4301067808457684e-05, "loss": 0.0172, "step": 55060 }, { "grad_norm": 0.27893388271331787, "learning_rate": 2.428334061285878e-05, "loss": 0.0155, "step": 55070 }, { "grad_norm": 0.31238797307014465, "learning_rate": 2.426561781153361e-05, "loss": 0.0165, "step": 55080 }, { "grad_norm": 0.23438797891139984, "learning_rate": 2.424789940751052e-05, "loss": 0.0181, "step": 55090 }, { "grad_norm": 0.23311051726341248, "learning_rate": 2.423018540381712e-05, "loss": 0.0162, "step": 55100 }, { "grad_norm": 0.17714251577854156, "learning_rate": 2.421247580348021e-05, "loss": 0.0184, "step": 55110 }, { "grad_norm": 0.26281672716140747, "learning_rate": 2.4194770609525897e-05, "loss": 0.0162, "step": 55120 }, { "grad_norm": 0.25407156348228455, "learning_rate": 2.417706982497951e-05, "loss": 0.0202, "step": 55130 }, { "grad_norm": 0.25415506958961487, "learning_rate": 2.415937345286563e-05, "loss": 0.0172, "step": 55140 }, { "grad_norm": 0.2673182785511017, "learning_rate": 2.4141681496208087e-05, "loss": 0.0135, "step": 55150 }, { "grad_norm": 0.21922893822193146, "learning_rate": 2.4123993958029946e-05, "loss": 0.0157, "step": 55160 }, { "grad_norm": 0.2883332669734955, "learning_rate": 2.4106310841353548e-05, "loss": 0.0166, "step": 55170 }, { "grad_norm": 0.19330181181430817, "learning_rate": 2.4088632149200398e-05, "loss": 0.0178, "step": 55180 }, { "grad_norm": 0.19358190894126892, "learning_rate": 2.4070957884591367e-05, "loss": 0.0133, "step": 55190 }, { "grad_norm": 0.4763127267360687, "learning_rate": 2.4053288050546464e-05, "loss": 0.0219, "step": 55200 }, { "grad_norm": 0.2933948338031769, "learning_rate": 2.403562265008498e-05, "loss": 0.0185, "step": 55210 }, { "grad_norm": 0.33245742321014404, "learning_rate": 2.4017961686225483e-05, "loss": 0.0188, "step": 55220 }, { "grad_norm": 0.22976848483085632, "learning_rate": 2.400030516198568e-05, "loss": 0.0226, "step": 55230 }, { "grad_norm": 0.21008887887001038, "learning_rate": 2.3982653080382673e-05, "loss": 0.0148, "step": 55240 }, { "grad_norm": 0.16623583436012268, "learning_rate": 2.396500544443266e-05, "loss": 0.017, "step": 55250 }, { "grad_norm": 0.32924938201904297, "learning_rate": 2.3947362257151156e-05, "loss": 0.016, "step": 55260 }, { "grad_norm": 0.3192521631717682, "learning_rate": 2.39297235215529e-05, "loss": 0.0219, "step": 55270 }, { "grad_norm": 0.21451188623905182, "learning_rate": 2.3912089240651873e-05, "loss": 0.0183, "step": 55280 }, { "grad_norm": 0.32500991225242615, "learning_rate": 2.389445941746129e-05, "loss": 0.0168, "step": 55290 }, { "grad_norm": 0.2533804774284363, "learning_rate": 2.3876834054993625e-05, "loss": 0.0147, "step": 55300 }, { "grad_norm": 0.16973216831684113, "learning_rate": 2.3859213156260522e-05, "loss": 0.0168, "step": 55310 }, { "grad_norm": 0.34509098529815674, "learning_rate": 2.3841596724272948e-05, "loss": 0.0181, "step": 55320 }, { "grad_norm": 0.26569488644599915, "learning_rate": 2.382398476204106e-05, "loss": 0.016, "step": 55330 }, { "grad_norm": 0.19625939428806305, "learning_rate": 2.3806377272574254e-05, "loss": 0.0141, "step": 55340 }, { "grad_norm": 0.5032418966293335, "learning_rate": 2.3788774258881174e-05, "loss": 0.0185, "step": 55350 }, { "grad_norm": 0.2613016664981842, "learning_rate": 2.37711757239697e-05, "loss": 0.0177, "step": 55360 }, { "grad_norm": 0.4429532587528229, "learning_rate": 2.3753581670846954e-05, "loss": 0.0169, "step": 55370 }, { "grad_norm": 0.11488247662782669, "learning_rate": 2.3735992102519216e-05, "loss": 0.017, "step": 55380 }, { "grad_norm": 0.255878746509552, "learning_rate": 2.371840702199215e-05, "loss": 0.0153, "step": 55390 }, { "grad_norm": 0.20747947692871094, "learning_rate": 2.3700826432270494e-05, "loss": 0.0143, "step": 55400 }, { "grad_norm": 0.3481852114200592, "learning_rate": 2.3683250336358326e-05, "loss": 0.0243, "step": 55410 }, { "grad_norm": 0.2293512374162674, "learning_rate": 2.3665678737258923e-05, "loss": 0.0171, "step": 55420 }, { "grad_norm": 0.1499265730381012, "learning_rate": 2.3648111637974745e-05, "loss": 0.0118, "step": 55430 }, { "grad_norm": 0.21459342539310455, "learning_rate": 2.36305490415076e-05, "loss": 0.0189, "step": 55440 }, { "grad_norm": 0.339639812707901, "learning_rate": 2.3612990950858392e-05, "loss": 0.0149, "step": 55450 }, { "grad_norm": 0.3054294288158417, "learning_rate": 2.359543736902735e-05, "loss": 0.0157, "step": 55460 }, { "grad_norm": 0.20844733715057373, "learning_rate": 2.3577888299013896e-05, "loss": 0.0181, "step": 55470 }, { "grad_norm": 0.3113037049770355, "learning_rate": 2.3560343743816683e-05, "loss": 0.0158, "step": 55480 }, { "grad_norm": 0.28241845965385437, "learning_rate": 2.354280370643362e-05, "loss": 0.0192, "step": 55490 }, { "grad_norm": 0.31119877099990845, "learning_rate": 2.3525268189861777e-05, "loss": 0.0172, "step": 55500 }, { "grad_norm": 0.30992361903190613, "learning_rate": 2.3507737197097513e-05, "loss": 0.0128, "step": 55510 }, { "grad_norm": 0.31770938634872437, "learning_rate": 2.34902107311364e-05, "loss": 0.0173, "step": 55520 }, { "grad_norm": 0.2851061224937439, "learning_rate": 2.3472688794973225e-05, "loss": 0.0163, "step": 55530 }, { "grad_norm": 0.3572022318840027, "learning_rate": 2.3455171391602016e-05, "loss": 0.0192, "step": 55540 }, { "grad_norm": 0.37190330028533936, "learning_rate": 2.343765852401601e-05, "loss": 0.015, "step": 55550 }, { "grad_norm": 0.16257107257843018, "learning_rate": 2.342015019520768e-05, "loss": 0.0184, "step": 55560 }, { "grad_norm": 0.16782481968402863, "learning_rate": 2.3402646408168742e-05, "loss": 0.0132, "step": 55570 }, { "grad_norm": 0.26974180340766907, "learning_rate": 2.3385147165890074e-05, "loss": 0.0179, "step": 55580 }, { "grad_norm": 0.2863371670246124, "learning_rate": 2.336765247136184e-05, "loss": 0.0185, "step": 55590 }, { "grad_norm": 0.40977850556373596, "learning_rate": 2.33501623275734e-05, "loss": 0.0219, "step": 55600 }, { "grad_norm": 0.27936652302742004, "learning_rate": 2.333267673751334e-05, "loss": 0.0132, "step": 55610 }, { "grad_norm": 0.2037186324596405, "learning_rate": 2.33151957041695e-05, "loss": 0.0128, "step": 55620 }, { "grad_norm": 0.29316484928131104, "learning_rate": 2.329771923052884e-05, "loss": 0.0165, "step": 55630 }, { "grad_norm": 0.27126583456993103, "learning_rate": 2.3280247319577697e-05, "loss": 0.0149, "step": 55640 }, { "grad_norm": 0.5624045729637146, "learning_rate": 2.3262779974301473e-05, "loss": 0.0203, "step": 55650 }, { "grad_norm": 0.2602764964103699, "learning_rate": 2.3245317197684895e-05, "loss": 0.0152, "step": 55660 }, { "grad_norm": 0.23068979382514954, "learning_rate": 2.322785899271188e-05, "loss": 0.013, "step": 55670 }, { "grad_norm": 0.22911575436592102, "learning_rate": 2.3210405362365507e-05, "loss": 0.0235, "step": 55680 }, { "grad_norm": 0.22334173321723938, "learning_rate": 2.31929563096282e-05, "loss": 0.0228, "step": 55690 }, { "grad_norm": 0.2109064757823944, "learning_rate": 2.317551183748146e-05, "loss": 0.0179, "step": 55700 }, { "grad_norm": 0.2317010462284088, "learning_rate": 2.3158071948906103e-05, "loss": 0.015, "step": 55710 }, { "grad_norm": 0.2016441971063614, "learning_rate": 2.314063664688212e-05, "loss": 0.0155, "step": 55720 }, { "grad_norm": 0.31039267778396606, "learning_rate": 2.3123205934388725e-05, "loss": 0.0148, "step": 55730 }, { "grad_norm": 0.15998801589012146, "learning_rate": 2.310577981440436e-05, "loss": 0.023, "step": 55740 }, { "grad_norm": 0.2155030369758606, "learning_rate": 2.3088358289906665e-05, "loss": 0.0146, "step": 55750 }, { "grad_norm": 0.3215336203575134, "learning_rate": 2.307094136387252e-05, "loss": 0.0167, "step": 55760 }, { "grad_norm": 0.25516510009765625, "learning_rate": 2.305352903927796e-05, "loss": 0.0195, "step": 55770 }, { "grad_norm": 0.16620610654354095, "learning_rate": 2.303612131909831e-05, "loss": 0.0139, "step": 55780 }, { "grad_norm": 0.2806977331638336, "learning_rate": 2.3018718206308054e-05, "loss": 0.0203, "step": 55790 }, { "grad_norm": 0.37066972255706787, "learning_rate": 2.3001319703880925e-05, "loss": 0.0212, "step": 55800 }, { "grad_norm": 0.23940107226371765, "learning_rate": 2.2983925814789835e-05, "loss": 0.016, "step": 55810 }, { "grad_norm": 0.30265575647354126, "learning_rate": 2.2966536542006957e-05, "loss": 0.0163, "step": 55820 }, { "grad_norm": 0.13613666594028473, "learning_rate": 2.294915188850358e-05, "loss": 0.0164, "step": 55830 }, { "grad_norm": 0.26878201961517334, "learning_rate": 2.2931771857250333e-05, "loss": 0.0187, "step": 55840 }, { "grad_norm": 0.3492538332939148, "learning_rate": 2.2914396451216946e-05, "loss": 0.0168, "step": 55850 }, { "grad_norm": 0.36362117528915405, "learning_rate": 2.2897025673372412e-05, "loss": 0.0157, "step": 55860 }, { "grad_norm": 0.2404070496559143, "learning_rate": 2.287965952668494e-05, "loss": 0.0196, "step": 55870 }, { "grad_norm": 0.1707657277584076, "learning_rate": 2.2862298014121873e-05, "loss": 0.0148, "step": 55880 }, { "grad_norm": 0.29270413517951965, "learning_rate": 2.28449411386499e-05, "loss": 0.0168, "step": 55890 }, { "grad_norm": 0.2667963206768036, "learning_rate": 2.282758890323477e-05, "loss": 0.0167, "step": 55900 }, { "grad_norm": 0.4579697847366333, "learning_rate": 2.2810241310841528e-05, "loss": 0.0152, "step": 55910 }, { "grad_norm": 0.18817923963069916, "learning_rate": 2.2792898364434412e-05, "loss": 0.0171, "step": 55920 }, { "grad_norm": 0.20686842501163483, "learning_rate": 2.2775560066976843e-05, "loss": 0.0155, "step": 55930 }, { "grad_norm": 0.1705368310213089, "learning_rate": 2.275822642143147e-05, "loss": 0.0162, "step": 55940 }, { "grad_norm": 0.20705589652061462, "learning_rate": 2.274089743076014e-05, "loss": 0.0159, "step": 55950 }, { "grad_norm": 0.3334802985191345, "learning_rate": 2.2723573097923917e-05, "loss": 0.0149, "step": 55960 }, { "grad_norm": 0.2954220771789551, "learning_rate": 2.2706253425883024e-05, "loss": 0.0151, "step": 55970 }, { "grad_norm": 0.37559181451797485, "learning_rate": 2.2688938417596933e-05, "loss": 0.0154, "step": 55980 }, { "grad_norm": 0.16208641231060028, "learning_rate": 2.2671628076024305e-05, "loss": 0.0165, "step": 55990 }, { "grad_norm": 0.1907423883676529, "learning_rate": 2.2654322404123008e-05, "loss": 0.0159, "step": 56000 }, { "grad_norm": 0.11901059746742249, "learning_rate": 2.2637021404850105e-05, "loss": 0.0119, "step": 56010 }, { "grad_norm": 0.36413702368736267, "learning_rate": 2.2619725081161885e-05, "loss": 0.0195, "step": 56020 }, { "grad_norm": 0.26270946860313416, "learning_rate": 2.260243343601376e-05, "loss": 0.016, "step": 56030 }, { "grad_norm": 0.212241530418396, "learning_rate": 2.2585146472360473e-05, "loss": 0.0125, "step": 56040 }, { "grad_norm": 0.27567487955093384, "learning_rate": 2.2567864193155834e-05, "loss": 0.0159, "step": 56050 }, { "grad_norm": 0.5328746438026428, "learning_rate": 2.255058660135294e-05, "loss": 0.0192, "step": 56060 }, { "grad_norm": 0.4429912865161896, "learning_rate": 2.253331369990407e-05, "loss": 0.0222, "step": 56070 }, { "grad_norm": 0.1739724576473236, "learning_rate": 2.251604549176063e-05, "loss": 0.0175, "step": 56080 }, { "grad_norm": 0.23785370588302612, "learning_rate": 2.249878197987337e-05, "loss": 0.0164, "step": 56090 }, { "grad_norm": 0.23698411881923676, "learning_rate": 2.2481523167192087e-05, "loss": 0.0143, "step": 56100 }, { "grad_norm": 0.2791900336742401, "learning_rate": 2.2464269056665867e-05, "loss": 0.0143, "step": 56110 }, { "grad_norm": 0.2722364366054535, "learning_rate": 2.2447019651242958e-05, "loss": 0.0248, "step": 56120 }, { "grad_norm": 0.30227744579315186, "learning_rate": 2.2429774953870818e-05, "loss": 0.0194, "step": 56130 }, { "grad_norm": 0.3217359185218811, "learning_rate": 2.241253496749611e-05, "loss": 0.0199, "step": 56140 }, { "grad_norm": 0.19150598347187042, "learning_rate": 2.2395299695064614e-05, "loss": 0.0136, "step": 56150 }, { "grad_norm": 0.27901309728622437, "learning_rate": 2.237806913952145e-05, "loss": 0.0164, "step": 56160 }, { "grad_norm": 0.26789215207099915, "learning_rate": 2.2360843303810798e-05, "loss": 0.0179, "step": 56170 }, { "grad_norm": 0.3683387041091919, "learning_rate": 2.2343622190876084e-05, "loss": 0.0152, "step": 56180 }, { "grad_norm": 0.278045654296875, "learning_rate": 2.2326405803659935e-05, "loss": 0.0187, "step": 56190 }, { "grad_norm": 0.2918742001056671, "learning_rate": 2.230919414510416e-05, "loss": 0.018, "step": 56200 }, { "grad_norm": 0.39974114298820496, "learning_rate": 2.229198721814976e-05, "loss": 0.0182, "step": 56210 }, { "grad_norm": 0.24905720353126526, "learning_rate": 2.2274785025736948e-05, "loss": 0.0133, "step": 56220 }, { "grad_norm": 0.13273289799690247, "learning_rate": 2.225758757080507e-05, "loss": 0.0171, "step": 56230 }, { "grad_norm": 0.398443341255188, "learning_rate": 2.2240394856292723e-05, "loss": 0.0199, "step": 56240 }, { "grad_norm": 0.3891907036304474, "learning_rate": 2.2223206885137664e-05, "loss": 0.0146, "step": 56250 }, { "grad_norm": 0.3902481198310852, "learning_rate": 2.2206023660276853e-05, "loss": 0.0178, "step": 56260 }, { "grad_norm": 0.41186264157295227, "learning_rate": 2.218884518464645e-05, "loss": 0.0171, "step": 56270 }, { "grad_norm": 0.3645915389060974, "learning_rate": 2.2171671461181732e-05, "loss": 0.0181, "step": 56280 }, { "grad_norm": 0.2077406644821167, "learning_rate": 2.2154502492817292e-05, "loss": 0.0193, "step": 56290 }, { "grad_norm": 0.11040529608726501, "learning_rate": 2.2137338282486782e-05, "loss": 0.0106, "step": 56300 }, { "grad_norm": 0.2887025773525238, "learning_rate": 2.2120178833123113e-05, "loss": 0.0194, "step": 56310 }, { "grad_norm": 0.37339088320732117, "learning_rate": 2.210302414765838e-05, "loss": 0.0153, "step": 56320 }, { "grad_norm": 0.15897785127162933, "learning_rate": 2.2085874229023808e-05, "loss": 0.0113, "step": 56330 }, { "grad_norm": 0.3032463788986206, "learning_rate": 2.2068729080149907e-05, "loss": 0.0139, "step": 56340 }, { "grad_norm": 0.46470215916633606, "learning_rate": 2.205158870396625e-05, "loss": 0.018, "step": 56350 }, { "grad_norm": 0.16535378992557526, "learning_rate": 2.2034453103401732e-05, "loss": 0.0167, "step": 56360 }, { "grad_norm": 0.18001866340637207, "learning_rate": 2.20173222813843e-05, "loss": 0.0123, "step": 56370 }, { "grad_norm": 0.22718262672424316, "learning_rate": 2.200019624084116e-05, "loss": 0.018, "step": 56380 }, { "grad_norm": 0.20631539821624756, "learning_rate": 2.1983074984698687e-05, "loss": 0.0193, "step": 56390 }, { "grad_norm": 0.3838234543800354, "learning_rate": 2.1965958515882433e-05, "loss": 0.0178, "step": 56400 }, { "grad_norm": 0.18917933106422424, "learning_rate": 2.1948846837317162e-05, "loss": 0.02, "step": 56410 }, { "grad_norm": 0.30297762155532837, "learning_rate": 2.1931739951926738e-05, "loss": 0.0133, "step": 56420 }, { "grad_norm": 0.1815367341041565, "learning_rate": 2.191463786263429e-05, "loss": 0.0086, "step": 56430 }, { "grad_norm": 0.18092359602451324, "learning_rate": 2.1897540572362095e-05, "loss": 0.016, "step": 56440 }, { "grad_norm": 0.3038730025291443, "learning_rate": 2.1880448084031614e-05, "loss": 0.0154, "step": 56450 }, { "grad_norm": 0.3028568625450134, "learning_rate": 2.1863360400563482e-05, "loss": 0.0132, "step": 56460 }, { "grad_norm": 0.25992530584335327, "learning_rate": 2.184627752487754e-05, "loss": 0.0167, "step": 56470 }, { "grad_norm": 0.1644384115934372, "learning_rate": 2.1829199459892725e-05, "loss": 0.0133, "step": 56480 }, { "grad_norm": 0.20965959131717682, "learning_rate": 2.1812126208527282e-05, "loss": 0.0157, "step": 56490 }, { "grad_norm": 0.31005290150642395, "learning_rate": 2.179505777369852e-05, "loss": 0.0201, "step": 56500 }, { "grad_norm": 0.19757604598999023, "learning_rate": 2.1777994158322974e-05, "loss": 0.0138, "step": 56510 }, { "grad_norm": 0.2881490886211395, "learning_rate": 2.176093536531637e-05, "loss": 0.0199, "step": 56520 }, { "grad_norm": 0.18252985179424286, "learning_rate": 2.1743881397593537e-05, "loss": 0.0127, "step": 56530 }, { "grad_norm": 0.31324127316474915, "learning_rate": 2.1726832258068595e-05, "loss": 0.0173, "step": 56540 }, { "grad_norm": 0.4303874373435974, "learning_rate": 2.170978794965472e-05, "loss": 0.0182, "step": 56550 }, { "grad_norm": 0.270812451839447, "learning_rate": 2.169274847526438e-05, "loss": 0.0152, "step": 56560 }, { "grad_norm": 0.20685164630413055, "learning_rate": 2.1675713837809103e-05, "loss": 0.0192, "step": 56570 }, { "grad_norm": 0.21326714754104614, "learning_rate": 2.1658684040199655e-05, "loss": 0.0155, "step": 56580 }, { "grad_norm": 0.26838037371635437, "learning_rate": 2.1641659085345974e-05, "loss": 0.0143, "step": 56590 }, { "grad_norm": 0.16084372997283936, "learning_rate": 2.1624638976157154e-05, "loss": 0.0196, "step": 56600 }, { "grad_norm": 0.3596755266189575, "learning_rate": 2.1607623715541476e-05, "loss": 0.0219, "step": 56610 }, { "grad_norm": 0.17421555519104004, "learning_rate": 2.159061330640636e-05, "loss": 0.0162, "step": 56620 }, { "grad_norm": 0.21350160241127014, "learning_rate": 2.1573607751658425e-05, "loss": 0.0108, "step": 56630 }, { "grad_norm": 0.28914305567741394, "learning_rate": 2.155660705420347e-05, "loss": 0.0151, "step": 56640 }, { "grad_norm": 0.2529950737953186, "learning_rate": 2.153961121694644e-05, "loss": 0.0166, "step": 56650 }, { "grad_norm": 0.24279586970806122, "learning_rate": 2.152262024279145e-05, "loss": 0.0178, "step": 56660 }, { "grad_norm": 0.4893406331539154, "learning_rate": 2.150563413464183e-05, "loss": 0.0185, "step": 56670 }, { "grad_norm": 0.2503156065940857, "learning_rate": 2.1488652895399974e-05, "loss": 0.013, "step": 56680 }, { "grad_norm": 0.4130651652812958, "learning_rate": 2.1471676527967587e-05, "loss": 0.0141, "step": 56690 }, { "grad_norm": 0.27840864658355713, "learning_rate": 2.1454705035245414e-05, "loss": 0.0198, "step": 56700 }, { "grad_norm": 0.23740869760513306, "learning_rate": 2.143773842013343e-05, "loss": 0.0139, "step": 56710 }, { "grad_norm": 0.3445427417755127, "learning_rate": 2.1420776685530796e-05, "loss": 0.0195, "step": 56720 }, { "grad_norm": 0.32814761996269226, "learning_rate": 2.1403819834335742e-05, "loss": 0.0147, "step": 56730 }, { "grad_norm": 0.23540855944156647, "learning_rate": 2.1386867869445808e-05, "loss": 0.0194, "step": 56740 }, { "grad_norm": 0.2452658712863922, "learning_rate": 2.1369920793757548e-05, "loss": 0.013, "step": 56750 }, { "grad_norm": 0.2617538869380951, "learning_rate": 2.1352978610166828e-05, "loss": 0.0141, "step": 56760 }, { "grad_norm": 0.25910741090774536, "learning_rate": 2.1336041321568546e-05, "loss": 0.0146, "step": 56770 }, { "grad_norm": 0.3029642105102539, "learning_rate": 2.131910893085684e-05, "loss": 0.0142, "step": 56780 }, { "grad_norm": 0.2455010563135147, "learning_rate": 2.1302181440925012e-05, "loss": 0.0195, "step": 56790 }, { "grad_norm": 0.20348045229911804, "learning_rate": 2.128525885466546e-05, "loss": 0.0159, "step": 56800 }, { "grad_norm": 0.16740168631076813, "learning_rate": 2.1268341174969847e-05, "loss": 0.0158, "step": 56810 }, { "grad_norm": 0.2724494934082031, "learning_rate": 2.1251428404728894e-05, "loss": 0.0139, "step": 56820 }, { "grad_norm": 0.20070838928222656, "learning_rate": 2.123452054683256e-05, "loss": 0.013, "step": 56830 }, { "grad_norm": 0.3683346211910248, "learning_rate": 2.1217617604169927e-05, "loss": 0.0181, "step": 56840 }, { "grad_norm": 0.4479416012763977, "learning_rate": 2.120071957962924e-05, "loss": 0.0172, "step": 56850 }, { "grad_norm": 0.31640955805778503, "learning_rate": 2.1183826476097917e-05, "loss": 0.0248, "step": 56860 }, { "grad_norm": 0.27738872170448303, "learning_rate": 2.116693829646254e-05, "loss": 0.0179, "step": 56870 }, { "grad_norm": 0.4288199841976166, "learning_rate": 2.1150055043608806e-05, "loss": 0.0148, "step": 56880 }, { "grad_norm": 0.47875723242759705, "learning_rate": 2.113317672042162e-05, "loss": 0.0178, "step": 56890 }, { "grad_norm": 0.25182750821113586, "learning_rate": 2.1116303329785025e-05, "loss": 0.0144, "step": 56900 }, { "grad_norm": 0.4019762873649597, "learning_rate": 2.109943487458222e-05, "loss": 0.0197, "step": 56910 }, { "grad_norm": 0.2816798985004425, "learning_rate": 2.1082571357695574e-05, "loss": 0.014, "step": 56920 }, { "grad_norm": 0.14529095590114594, "learning_rate": 2.1065712782006557e-05, "loss": 0.0136, "step": 56930 }, { "grad_norm": 0.183552548289299, "learning_rate": 2.104885915039591e-05, "loss": 0.0157, "step": 56940 }, { "grad_norm": 0.2325613796710968, "learning_rate": 2.103201046574338e-05, "loss": 0.0125, "step": 56950 }, { "grad_norm": 0.42459002137184143, "learning_rate": 2.1015166730928022e-05, "loss": 0.0183, "step": 56960 }, { "grad_norm": 0.377819687128067, "learning_rate": 2.0998327948827912e-05, "loss": 0.0209, "step": 56970 }, { "grad_norm": 0.4280973970890045, "learning_rate": 2.0981494122320363e-05, "loss": 0.0212, "step": 56980 }, { "grad_norm": 0.29620176553726196, "learning_rate": 2.0964665254281822e-05, "loss": 0.0193, "step": 56990 }, { "grad_norm": 0.3328436315059662, "learning_rate": 2.094784134758784e-05, "loss": 0.0176, "step": 57000 }, { "grad_norm": 0.22017395496368408, "learning_rate": 2.0931022405113226e-05, "loss": 0.0227, "step": 57010 }, { "grad_norm": 0.2497401088476181, "learning_rate": 2.091420842973183e-05, "loss": 0.0209, "step": 57020 }, { "grad_norm": 0.3923429548740387, "learning_rate": 2.0897399424316715e-05, "loss": 0.0128, "step": 57030 }, { "grad_norm": 0.36453762650489807, "learning_rate": 2.0880595391740078e-05, "loss": 0.0189, "step": 57040 }, { "grad_norm": 0.22607722878456116, "learning_rate": 2.0863796334873277e-05, "loss": 0.0172, "step": 57050 }, { "grad_norm": 0.1909366250038147, "learning_rate": 2.08470022565868e-05, "loss": 0.0107, "step": 57060 }, { "grad_norm": 0.20973041653633118, "learning_rate": 2.0830213159750317e-05, "loss": 0.0117, "step": 57070 }, { "grad_norm": 0.17868036031723022, "learning_rate": 2.0813429047232596e-05, "loss": 0.0169, "step": 57080 }, { "grad_norm": 0.25125035643577576, "learning_rate": 2.0796649921901594e-05, "loss": 0.0135, "step": 57090 }, { "grad_norm": 0.16745981574058533, "learning_rate": 2.077987578662441e-05, "loss": 0.0129, "step": 57100 }, { "grad_norm": 0.2671736776828766, "learning_rate": 2.0763106644267277e-05, "loss": 0.0175, "step": 57110 }, { "grad_norm": 0.25727611780166626, "learning_rate": 2.0746342497695607e-05, "loss": 0.0139, "step": 57120 }, { "grad_norm": 0.1878269612789154, "learning_rate": 2.0729583349773886e-05, "loss": 0.011, "step": 57130 }, { "grad_norm": 0.3215610980987549, "learning_rate": 2.0712829203365853e-05, "loss": 0.0182, "step": 57140 }, { "grad_norm": 0.20532537996768951, "learning_rate": 2.0696080061334267e-05, "loss": 0.0171, "step": 57150 }, { "grad_norm": 0.24548125267028809, "learning_rate": 2.067933592654117e-05, "loss": 0.0163, "step": 57160 }, { "grad_norm": 0.2891632318496704, "learning_rate": 2.066259680184763e-05, "loss": 0.0186, "step": 57170 }, { "grad_norm": 0.21302159130573273, "learning_rate": 2.0645862690113908e-05, "loss": 0.0121, "step": 57180 }, { "grad_norm": 0.26572513580322266, "learning_rate": 2.0629133594199436e-05, "loss": 0.0174, "step": 57190 }, { "grad_norm": 0.30126431584358215, "learning_rate": 2.0612409516962704e-05, "loss": 0.0129, "step": 57200 }, { "grad_norm": 0.3108040690422058, "learning_rate": 2.0595690461261467e-05, "loss": 0.015, "step": 57210 }, { "grad_norm": 0.2729089856147766, "learning_rate": 2.0578976429952503e-05, "loss": 0.0116, "step": 57220 }, { "grad_norm": 0.31667912006378174, "learning_rate": 2.0562267425891802e-05, "loss": 0.0157, "step": 57230 }, { "grad_norm": 0.3545595705509186, "learning_rate": 2.0545563451934467e-05, "loss": 0.0163, "step": 57240 }, { "grad_norm": 0.2886025905609131, "learning_rate": 2.0528864510934764e-05, "loss": 0.0193, "step": 57250 }, { "grad_norm": 0.30517834424972534, "learning_rate": 2.0512170605746096e-05, "loss": 0.0229, "step": 57260 }, { "grad_norm": 0.17647257447242737, "learning_rate": 2.0495481739220963e-05, "loss": 0.013, "step": 57270 }, { "grad_norm": 0.374681293964386, "learning_rate": 2.0478797914211045e-05, "loss": 0.0188, "step": 57280 }, { "grad_norm": 0.30573132634162903, "learning_rate": 2.046211913356716e-05, "loss": 0.0188, "step": 57290 }, { "grad_norm": 0.20601791143417358, "learning_rate": 2.0445445400139247e-05, "loss": 0.0188, "step": 57300 }, { "grad_norm": 0.35681775212287903, "learning_rate": 2.0428776716776405e-05, "loss": 0.0166, "step": 57310 }, { "grad_norm": 0.4816761314868927, "learning_rate": 2.0412113086326856e-05, "loss": 0.0161, "step": 57320 }, { "grad_norm": 0.3265724182128906, "learning_rate": 2.0395454511637918e-05, "loss": 0.0129, "step": 57330 }, { "grad_norm": 0.5364927053451538, "learning_rate": 2.037880099555616e-05, "loss": 0.016, "step": 57340 }, { "grad_norm": 0.2134971171617508, "learning_rate": 2.0362152540927144e-05, "loss": 0.0114, "step": 57350 }, { "grad_norm": 0.3250398635864258, "learning_rate": 2.0345509150595666e-05, "loss": 0.022, "step": 57360 }, { "grad_norm": 0.30302998423576355, "learning_rate": 2.0328870827405617e-05, "loss": 0.0194, "step": 57370 }, { "grad_norm": 0.18860532343387604, "learning_rate": 2.0312237574200043e-05, "loss": 0.011, "step": 57380 }, { "grad_norm": 0.3538421094417572, "learning_rate": 2.029560939382112e-05, "loss": 0.0185, "step": 57390 }, { "grad_norm": 0.27447187900543213, "learning_rate": 2.0278986289110097e-05, "loss": 0.0165, "step": 57400 }, { "grad_norm": 0.3366053104400635, "learning_rate": 2.0262368262907484e-05, "loss": 0.0163, "step": 57410 }, { "grad_norm": 0.32901522517204285, "learning_rate": 2.024575531805279e-05, "loss": 0.0218, "step": 57420 }, { "grad_norm": 0.19908536970615387, "learning_rate": 2.0229147457384735e-05, "loss": 0.0121, "step": 57430 }, { "grad_norm": 0.2093457281589508, "learning_rate": 2.0212544683741157e-05, "loss": 0.0235, "step": 57440 }, { "grad_norm": 0.21257305145263672, "learning_rate": 2.0195946999958976e-05, "loss": 0.0237, "step": 57450 }, { "grad_norm": 0.26703837513923645, "learning_rate": 2.017935440887434e-05, "loss": 0.0186, "step": 57460 }, { "grad_norm": 0.22263610363006592, "learning_rate": 2.0162766913322423e-05, "loss": 0.0152, "step": 57470 }, { "grad_norm": 0.2819395065307617, "learning_rate": 2.0146184516137588e-05, "loss": 0.0205, "step": 57480 }, { "grad_norm": 0.22671538591384888, "learning_rate": 2.012960722015332e-05, "loss": 0.0162, "step": 57490 }, { "grad_norm": 0.326707124710083, "learning_rate": 2.0113035028202214e-05, "loss": 0.018, "step": 57500 }, { "grad_norm": 0.2154536247253418, "learning_rate": 2.009646794311602e-05, "loss": 0.0204, "step": 57510 }, { "grad_norm": 0.16210158169269562, "learning_rate": 2.007990596772559e-05, "loss": 0.0149, "step": 57520 }, { "grad_norm": 0.2769346535205841, "learning_rate": 2.0063349104860923e-05, "loss": 0.0182, "step": 57530 }, { "grad_norm": 0.1894507110118866, "learning_rate": 2.0046797357351116e-05, "loss": 0.0188, "step": 57540 }, { "grad_norm": 0.30952179431915283, "learning_rate": 2.0030250728024412e-05, "loss": 0.0128, "step": 57550 }, { "grad_norm": 0.19401153922080994, "learning_rate": 2.001370921970819e-05, "loss": 0.0124, "step": 57560 }, { "grad_norm": 0.3175167143344879, "learning_rate": 1.9997172835228932e-05, "loss": 0.0143, "step": 57570 }, { "grad_norm": 0.2761119306087494, "learning_rate": 1.9980641577412262e-05, "loss": 0.0129, "step": 57580 }, { "grad_norm": 0.4129018485546112, "learning_rate": 1.9964115449082925e-05, "loss": 0.015, "step": 57590 }, { "grad_norm": 0.20774245262145996, "learning_rate": 1.9947594453064742e-05, "loss": 0.0177, "step": 57600 }, { "grad_norm": 0.29011860489845276, "learning_rate": 1.9931078592180774e-05, "loss": 0.0165, "step": 57610 }, { "grad_norm": 0.36305806040763855, "learning_rate": 1.9914567869253065e-05, "loss": 0.0158, "step": 57620 }, { "grad_norm": 0.3925957977771759, "learning_rate": 1.989806228710287e-05, "loss": 0.0216, "step": 57630 }, { "grad_norm": 0.21140369772911072, "learning_rate": 1.9881561848550555e-05, "loss": 0.0136, "step": 57640 }, { "grad_norm": 0.243360236287117, "learning_rate": 1.9865066556415544e-05, "loss": 0.0191, "step": 57650 }, { "grad_norm": 0.24321117997169495, "learning_rate": 1.98485764135165e-05, "loss": 0.0157, "step": 57660 }, { "grad_norm": 0.13864296674728394, "learning_rate": 1.983209142267109e-05, "loss": 0.0151, "step": 57670 }, { "grad_norm": 0.32306063175201416, "learning_rate": 1.9815611586696165e-05, "loss": 0.0132, "step": 57680 }, { "grad_norm": 0.310315877199173, "learning_rate": 1.9799136908407667e-05, "loss": 0.0136, "step": 57690 }, { "grad_norm": 0.14138992130756378, "learning_rate": 1.9782667390620678e-05, "loss": 0.0173, "step": 57700 }, { "grad_norm": 0.16026777029037476, "learning_rate": 1.976620303614939e-05, "loss": 0.01, "step": 57710 }, { "grad_norm": 0.22244274616241455, "learning_rate": 1.9749743847807108e-05, "loss": 0.0152, "step": 57720 }, { "grad_norm": 0.43490949273109436, "learning_rate": 1.9733289828406272e-05, "loss": 0.0152, "step": 57730 }, { "grad_norm": 0.33146804571151733, "learning_rate": 1.9716840980758382e-05, "loss": 0.0159, "step": 57740 }, { "grad_norm": 0.1475677490234375, "learning_rate": 1.9700397307674134e-05, "loss": 0.0122, "step": 57750 }, { "grad_norm": 0.14919723570346832, "learning_rate": 1.968395881196328e-05, "loss": 0.0188, "step": 57760 }, { "grad_norm": 0.4042636752128601, "learning_rate": 1.966752549643473e-05, "loss": 0.0146, "step": 57770 }, { "grad_norm": 0.2587693929672241, "learning_rate": 1.965109736389647e-05, "loss": 0.0177, "step": 57780 }, { "grad_norm": 0.2517254054546356, "learning_rate": 1.9634674417155645e-05, "loss": 0.0132, "step": 57790 }, { "grad_norm": 0.3051289916038513, "learning_rate": 1.9618256659018434e-05, "loss": 0.0313, "step": 57800 }, { "grad_norm": 0.3728892505168915, "learning_rate": 1.9601844092290257e-05, "loss": 0.017, "step": 57810 }, { "grad_norm": 0.32805293798446655, "learning_rate": 1.9585436719775512e-05, "loss": 0.0122, "step": 57820 }, { "grad_norm": 0.11395885795354843, "learning_rate": 1.9569034544277793e-05, "loss": 0.014, "step": 57830 }, { "grad_norm": 0.22445030510425568, "learning_rate": 1.9552637568599798e-05, "loss": 0.0152, "step": 57840 }, { "grad_norm": 0.22001345455646515, "learning_rate": 1.953624579554327e-05, "loss": 0.0164, "step": 57850 }, { "grad_norm": 0.34568196535110474, "learning_rate": 1.951985922790918e-05, "loss": 0.0241, "step": 57860 }, { "grad_norm": 0.24783048033714294, "learning_rate": 1.9503477868497505e-05, "loss": 0.0135, "step": 57870 }, { "grad_norm": 0.4417233169078827, "learning_rate": 1.9487101720107375e-05, "loss": 0.0174, "step": 57880 }, { "grad_norm": 0.22753918170928955, "learning_rate": 1.9470730785537032e-05, "loss": 0.0127, "step": 57890 }, { "grad_norm": 0.3731500506401062, "learning_rate": 1.9454365067583823e-05, "loss": 0.0144, "step": 57900 }, { "grad_norm": 0.28136569261550903, "learning_rate": 1.9438004569044215e-05, "loss": 0.0152, "step": 57910 }, { "grad_norm": 0.22977115213871002, "learning_rate": 1.9421649292713724e-05, "loss": 0.0318, "step": 57920 }, { "grad_norm": 0.36474767327308655, "learning_rate": 1.9405299241387076e-05, "loss": 0.0164, "step": 57930 }, { "grad_norm": 0.5822710394859314, "learning_rate": 1.9388954417858007e-05, "loss": 0.0136, "step": 57940 }, { "grad_norm": 0.4089372456073761, "learning_rate": 1.9372614824919417e-05, "loss": 0.0117, "step": 57950 }, { "grad_norm": 0.3031758964061737, "learning_rate": 1.9356280465363284e-05, "loss": 0.0239, "step": 57960 }, { "grad_norm": 0.32809266448020935, "learning_rate": 1.9339951341980723e-05, "loss": 0.0132, "step": 57970 }, { "grad_norm": 0.2509666681289673, "learning_rate": 1.9323627457561916e-05, "loss": 0.0163, "step": 57980 }, { "grad_norm": 0.16562509536743164, "learning_rate": 1.9307308814896198e-05, "loss": 0.0082, "step": 57990 }, { "grad_norm": 0.23061281442642212, "learning_rate": 1.9290995416771935e-05, "loss": 0.0135, "step": 58000 }, { "grad_norm": 0.18563726544380188, "learning_rate": 1.9274687265976665e-05, "loss": 0.0151, "step": 58010 }, { "grad_norm": 0.28051382303237915, "learning_rate": 1.9258384365297e-05, "loss": 0.0154, "step": 58020 }, { "grad_norm": 0.2614441215991974, "learning_rate": 1.924208671751866e-05, "loss": 0.016, "step": 58030 }, { "grad_norm": 0.23595920205116272, "learning_rate": 1.9225794325426492e-05, "loss": 0.0211, "step": 58040 }, { "grad_norm": 0.21494527161121368, "learning_rate": 1.920950719180436e-05, "loss": 0.0121, "step": 58050 }, { "grad_norm": 0.3562576472759247, "learning_rate": 1.919322531943536e-05, "loss": 0.0131, "step": 58060 }, { "grad_norm": 0.1982421725988388, "learning_rate": 1.917694871110157e-05, "loss": 0.0175, "step": 58070 }, { "grad_norm": 0.3197686970233917, "learning_rate": 1.9160677369584234e-05, "loss": 0.015, "step": 58080 }, { "grad_norm": 0.29437577724456787, "learning_rate": 1.9144411297663694e-05, "loss": 0.0184, "step": 58090 }, { "grad_norm": 0.2194858193397522, "learning_rate": 1.9128150498119328e-05, "loss": 0.0174, "step": 58100 }, { "grad_norm": 0.17162403464317322, "learning_rate": 1.9111894973729726e-05, "loss": 0.0107, "step": 58110 }, { "grad_norm": 0.3818584084510803, "learning_rate": 1.9095644727272454e-05, "loss": 0.0178, "step": 58120 }, { "grad_norm": 0.3509967029094696, "learning_rate": 1.907939976152429e-05, "loss": 0.0181, "step": 58130 }, { "grad_norm": 0.3258964717388153, "learning_rate": 1.906316007926101e-05, "loss": 0.0176, "step": 58140 }, { "grad_norm": 0.27688559889793396, "learning_rate": 1.904692568325755e-05, "loss": 0.0156, "step": 58150 }, { "grad_norm": 0.27707648277282715, "learning_rate": 1.9030696576287925e-05, "loss": 0.017, "step": 58160 }, { "grad_norm": 0.22332118451595306, "learning_rate": 1.9014472761125242e-05, "loss": 0.014, "step": 58170 }, { "grad_norm": 0.20506715774536133, "learning_rate": 1.899825424054172e-05, "loss": 0.0138, "step": 58180 }, { "grad_norm": 0.26556405425071716, "learning_rate": 1.898204101730863e-05, "loss": 0.0142, "step": 58190 }, { "grad_norm": 0.2559739351272583, "learning_rate": 1.8965833094196394e-05, "loss": 0.0257, "step": 58200 }, { "grad_norm": 0.24217107892036438, "learning_rate": 1.8949630473974495e-05, "loss": 0.0158, "step": 58210 }, { "grad_norm": 0.26781243085861206, "learning_rate": 1.8933433159411517e-05, "loss": 0.0182, "step": 58220 }, { "grad_norm": 0.21037843823432922, "learning_rate": 1.891724115327514e-05, "loss": 0.016, "step": 58230 }, { "grad_norm": 0.5214473009109497, "learning_rate": 1.8901054458332152e-05, "loss": 0.0191, "step": 58240 }, { "grad_norm": 0.13014701008796692, "learning_rate": 1.8884873077348364e-05, "loss": 0.0168, "step": 58250 }, { "grad_norm": 0.3114890456199646, "learning_rate": 1.88686970130888e-05, "loss": 0.0181, "step": 58260 }, { "grad_norm": 0.2704869508743286, "learning_rate": 1.8852526268317456e-05, "loss": 0.0181, "step": 58270 }, { "grad_norm": 0.17500393092632294, "learning_rate": 1.883636084579749e-05, "loss": 0.0168, "step": 58280 }, { "grad_norm": 0.3376254439353943, "learning_rate": 1.8820200748291145e-05, "loss": 0.0149, "step": 58290 }, { "grad_norm": 0.1728498488664627, "learning_rate": 1.8804045978559686e-05, "loss": 0.0122, "step": 58300 }, { "grad_norm": 0.19736631214618683, "learning_rate": 1.8787896539363594e-05, "loss": 0.0151, "step": 58310 }, { "grad_norm": 0.11102066189050674, "learning_rate": 1.8771752433462298e-05, "loss": 0.0112, "step": 58320 }, { "grad_norm": 0.31929436326026917, "learning_rate": 1.8755613663614447e-05, "loss": 0.0153, "step": 58330 }, { "grad_norm": 0.3349132537841797, "learning_rate": 1.873948023257767e-05, "loss": 0.0164, "step": 58340 }, { "grad_norm": 0.4243611693382263, "learning_rate": 1.8723352143108747e-05, "loss": 0.0159, "step": 58350 }, { "grad_norm": 0.21412764489650726, "learning_rate": 1.870722939796352e-05, "loss": 0.0166, "step": 58360 }, { "grad_norm": 0.4810888171195984, "learning_rate": 1.869111199989693e-05, "loss": 0.016, "step": 58370 }, { "grad_norm": 0.2923409938812256, "learning_rate": 1.8674999951663018e-05, "loss": 0.0161, "step": 58380 }, { "grad_norm": 0.10390080511569977, "learning_rate": 1.8658893256014857e-05, "loss": 0.0132, "step": 58390 }, { "grad_norm": 0.16907572746276855, "learning_rate": 1.8642791915704655e-05, "loss": 0.015, "step": 58400 }, { "grad_norm": 0.312060683965683, "learning_rate": 1.8626695933483697e-05, "loss": 0.0158, "step": 58410 }, { "grad_norm": 0.2793748080730438, "learning_rate": 1.8610605312102335e-05, "loss": 0.0153, "step": 58420 }, { "grad_norm": 0.26868587732315063, "learning_rate": 1.8594520054310032e-05, "loss": 0.0175, "step": 58430 }, { "grad_norm": 0.28281986713409424, "learning_rate": 1.8578440162855326e-05, "loss": 0.0123, "step": 58440 }, { "grad_norm": 0.21632111072540283, "learning_rate": 1.8562365640485774e-05, "loss": 0.0158, "step": 58450 }, { "grad_norm": 0.22741122543811798, "learning_rate": 1.854629648994815e-05, "loss": 0.0157, "step": 58460 }, { "grad_norm": 0.31435897946357727, "learning_rate": 1.8530232713988183e-05, "loss": 0.0141, "step": 58470 }, { "grad_norm": 0.2917799949645996, "learning_rate": 1.851417431535074e-05, "loss": 0.0146, "step": 58480 }, { "grad_norm": 0.2871326208114624, "learning_rate": 1.8498121296779785e-05, "loss": 0.0206, "step": 58490 }, { "grad_norm": 0.23955467343330383, "learning_rate": 1.848207366101829e-05, "loss": 0.0204, "step": 58500 }, { "grad_norm": 0.20461826026439667, "learning_rate": 1.8466031410808422e-05, "loss": 0.0126, "step": 58510 }, { "grad_norm": 0.4545135498046875, "learning_rate": 1.844999454889129e-05, "loss": 0.0166, "step": 58520 }, { "grad_norm": 0.29529014229774475, "learning_rate": 1.8433963078007237e-05, "loss": 0.0178, "step": 58530 }, { "grad_norm": 0.311286985874176, "learning_rate": 1.841793700089554e-05, "loss": 0.017, "step": 58540 }, { "grad_norm": 0.24616731703281403, "learning_rate": 1.8401916320294644e-05, "loss": 0.0139, "step": 58550 }, { "grad_norm": 0.3668922185897827, "learning_rate": 1.838590103894205e-05, "loss": 0.0148, "step": 58560 }, { "grad_norm": 0.2714920938014984, "learning_rate": 1.8369891159574292e-05, "loss": 0.0156, "step": 58570 }, { "grad_norm": 0.2814900875091553, "learning_rate": 1.835388668492708e-05, "loss": 0.0171, "step": 58580 }, { "grad_norm": 0.14784371852874756, "learning_rate": 1.8337887617735095e-05, "loss": 0.0152, "step": 58590 }, { "grad_norm": 0.2145923525094986, "learning_rate": 1.8321893960732157e-05, "loss": 0.0185, "step": 58600 }, { "grad_norm": 0.38212060928344727, "learning_rate": 1.8305905716651138e-05, "loss": 0.0144, "step": 58610 }, { "grad_norm": 0.18273285031318665, "learning_rate": 1.8289922888224e-05, "loss": 0.0133, "step": 58620 }, { "grad_norm": 0.2887842655181885, "learning_rate": 1.8273945478181765e-05, "loss": 0.0176, "step": 58630 }, { "grad_norm": 0.382199227809906, "learning_rate": 1.8257973489254558e-05, "loss": 0.0163, "step": 58640 }, { "grad_norm": 0.24687713384628296, "learning_rate": 1.824200692417152e-05, "loss": 0.016, "step": 58650 }, { "grad_norm": 0.2719595432281494, "learning_rate": 1.8226045785660912e-05, "loss": 0.0139, "step": 58660 }, { "grad_norm": 0.3614833950996399, "learning_rate": 1.821009007645006e-05, "loss": 0.0128, "step": 58670 }, { "grad_norm": 0.2559647858142853, "learning_rate": 1.8194139799265357e-05, "loss": 0.0108, "step": 58680 }, { "grad_norm": 0.40110570192337036, "learning_rate": 1.8178194956832295e-05, "loss": 0.0167, "step": 58690 }, { "grad_norm": 0.3257652223110199, "learning_rate": 1.8162255551875346e-05, "loss": 0.0164, "step": 58700 }, { "grad_norm": 0.21059240400791168, "learning_rate": 1.8146321587118194e-05, "loss": 0.019, "step": 58710 }, { "grad_norm": 0.2258375883102417, "learning_rate": 1.8130393065283448e-05, "loss": 0.0089, "step": 58720 }, { "grad_norm": 0.2625857889652252, "learning_rate": 1.8114469989092925e-05, "loss": 0.0146, "step": 58730 }, { "grad_norm": 0.3339325189590454, "learning_rate": 1.8098552361267397e-05, "loss": 0.0119, "step": 58740 }, { "grad_norm": 0.24156992137432098, "learning_rate": 1.8082640184526763e-05, "loss": 0.0116, "step": 58750 }, { "grad_norm": 0.20670367777347565, "learning_rate": 1.806673346158999e-05, "loss": 0.0136, "step": 58760 }, { "grad_norm": 0.14223721623420715, "learning_rate": 1.8050832195175067e-05, "loss": 0.0185, "step": 58770 }, { "grad_norm": 0.28770577907562256, "learning_rate": 1.8034936387999136e-05, "loss": 0.0181, "step": 58780 }, { "grad_norm": 0.3285593092441559, "learning_rate": 1.8019046042778315e-05, "loss": 0.0161, "step": 58790 }, { "grad_norm": 0.32065045833587646, "learning_rate": 1.800316116222785e-05, "loss": 0.0137, "step": 58800 }, { "grad_norm": 0.26383551955223083, "learning_rate": 1.7987281749062018e-05, "loss": 0.0362, "step": 58810 }, { "grad_norm": 0.2098148763179779, "learning_rate": 1.7971407805994195e-05, "loss": 0.0201, "step": 58820 }, { "grad_norm": 0.14467091858386993, "learning_rate": 1.7955539335736787e-05, "loss": 0.0172, "step": 58830 }, { "grad_norm": 0.19650426506996155, "learning_rate": 1.7939676341001304e-05, "loss": 0.014, "step": 58840 }, { "grad_norm": 0.31212687492370605, "learning_rate": 1.7923818824498275e-05, "loss": 0.0129, "step": 58850 }, { "grad_norm": 0.26823121309280396, "learning_rate": 1.7907966788937315e-05, "loss": 0.0234, "step": 58860 }, { "grad_norm": 0.19812631607055664, "learning_rate": 1.7892120237027116e-05, "loss": 0.0194, "step": 58870 }, { "grad_norm": 0.2804360091686249, "learning_rate": 1.7876279171475413e-05, "loss": 0.0184, "step": 58880 }, { "grad_norm": 0.29518577456474304, "learning_rate": 1.7860443594989028e-05, "loss": 0.0162, "step": 58890 }, { "grad_norm": 0.26371899247169495, "learning_rate": 1.784461351027379e-05, "loss": 0.0153, "step": 58900 }, { "grad_norm": 0.17379909753799438, "learning_rate": 1.7828788920034677e-05, "loss": 0.0219, "step": 58910 }, { "grad_norm": 0.32840216159820557, "learning_rate": 1.7812969826975623e-05, "loss": 0.0175, "step": 58920 }, { "grad_norm": 0.2141897827386856, "learning_rate": 1.7797156233799738e-05, "loss": 0.0197, "step": 58930 }, { "grad_norm": 0.22094179689884186, "learning_rate": 1.7781348143209094e-05, "loss": 0.0138, "step": 58940 }, { "grad_norm": 0.18235929310321808, "learning_rate": 1.7765545557904862e-05, "loss": 0.0251, "step": 58950 }, { "grad_norm": 0.26749053597450256, "learning_rate": 1.7749748480587302e-05, "loss": 0.0133, "step": 58960 }, { "grad_norm": 0.16109232604503632, "learning_rate": 1.773395691395564e-05, "loss": 0.012, "step": 58970 }, { "grad_norm": 0.22017090022563934, "learning_rate": 1.7718170860708305e-05, "loss": 0.015, "step": 58980 }, { "grad_norm": 0.20543722808361053, "learning_rate": 1.770239032354264e-05, "loss": 0.0148, "step": 58990 }, { "grad_norm": 0.30432334542274475, "learning_rate": 1.7686615305155124e-05, "loss": 0.0186, "step": 59000 }, { "grad_norm": 0.22910170257091522, "learning_rate": 1.767084580824128e-05, "loss": 0.0088, "step": 59010 }, { "grad_norm": 0.5465424060821533, "learning_rate": 1.765508183549569e-05, "loss": 0.0232, "step": 59020 }, { "grad_norm": 0.34574222564697266, "learning_rate": 1.763932338961199e-05, "loss": 0.0158, "step": 59030 }, { "grad_norm": 0.31468361616134644, "learning_rate": 1.762357047328284e-05, "loss": 0.0156, "step": 59040 }, { "grad_norm": 0.29704198241233826, "learning_rate": 1.7607823089199997e-05, "loss": 0.0213, "step": 59050 }, { "grad_norm": 0.23348015546798706, "learning_rate": 1.7592081240054265e-05, "loss": 0.0156, "step": 59060 }, { "grad_norm": 0.25593462586402893, "learning_rate": 1.7576344928535488e-05, "loss": 0.0153, "step": 59070 }, { "grad_norm": 0.2821762263774872, "learning_rate": 1.7560614157332573e-05, "loss": 0.0135, "step": 59080 }, { "grad_norm": 0.18461281061172485, "learning_rate": 1.7544888929133495e-05, "loss": 0.0129, "step": 59090 }, { "grad_norm": 0.25122395157814026, "learning_rate": 1.752916924662522e-05, "loss": 0.0195, "step": 59100 }, { "grad_norm": 0.3328208327293396, "learning_rate": 1.751345511249387e-05, "loss": 0.0114, "step": 59110 }, { "grad_norm": 0.24963819980621338, "learning_rate": 1.749774652942452e-05, "loss": 0.015, "step": 59120 }, { "grad_norm": 0.28214147686958313, "learning_rate": 1.748204350010135e-05, "loss": 0.0155, "step": 59130 }, { "grad_norm": 0.15437811613082886, "learning_rate": 1.7466346027207574e-05, "loss": 0.0141, "step": 59140 }, { "grad_norm": 0.1913299858570099, "learning_rate": 1.745065411342547e-05, "loss": 0.0131, "step": 59150 }, { "grad_norm": 0.2882164716720581, "learning_rate": 1.7434967761436366e-05, "loss": 0.0181, "step": 59160 }, { "grad_norm": 0.35236018896102905, "learning_rate": 1.741928697392058e-05, "loss": 0.013, "step": 59170 }, { "grad_norm": 0.36417701840400696, "learning_rate": 1.7403611753557597e-05, "loss": 0.0145, "step": 59180 }, { "grad_norm": 0.29678434133529663, "learning_rate": 1.738794210302584e-05, "loss": 0.0121, "step": 59190 }, { "grad_norm": 0.11786433309316635, "learning_rate": 1.7372278025002837e-05, "loss": 0.0122, "step": 59200 }, { "grad_norm": 0.2620387077331543, "learning_rate": 1.7356619522165164e-05, "loss": 0.0152, "step": 59210 }, { "grad_norm": 0.18403665721416473, "learning_rate": 1.7340966597188378e-05, "loss": 0.012, "step": 59220 }, { "grad_norm": 0.21300667524337769, "learning_rate": 1.732531925274722e-05, "loss": 0.0119, "step": 59230 }, { "grad_norm": 0.17322418093681335, "learning_rate": 1.7309677491515318e-05, "loss": 0.0155, "step": 59240 }, { "grad_norm": 0.29119303822517395, "learning_rate": 1.7294041316165455e-05, "loss": 0.0158, "step": 59250 }, { "grad_norm": 0.32551121711730957, "learning_rate": 1.727841072936942e-05, "loss": 0.019, "step": 59260 }, { "grad_norm": 0.2595538794994354, "learning_rate": 1.7262785733798058e-05, "loss": 0.0124, "step": 59270 }, { "grad_norm": 0.25916555523872375, "learning_rate": 1.7247166332121246e-05, "loss": 0.0137, "step": 59280 }, { "grad_norm": 0.22646941244602203, "learning_rate": 1.7231552527007933e-05, "loss": 0.0144, "step": 59290 }, { "grad_norm": 0.31105902791023254, "learning_rate": 1.721594432112606e-05, "loss": 0.0144, "step": 59300 }, { "grad_norm": 0.20631329715251923, "learning_rate": 1.7200341717142655e-05, "loss": 0.02, "step": 59310 }, { "grad_norm": 0.37790778279304504, "learning_rate": 1.718474471772377e-05, "loss": 0.0162, "step": 59320 }, { "grad_norm": 0.1992480456829071, "learning_rate": 1.7169153325534528e-05, "loss": 0.0122, "step": 59330 }, { "grad_norm": 0.22750750184059143, "learning_rate": 1.7153567543239045e-05, "loss": 0.013, "step": 59340 }, { "grad_norm": 0.19977913796901703, "learning_rate": 1.7137987373500525e-05, "loss": 0.0157, "step": 59350 }, { "grad_norm": 0.35262781381607056, "learning_rate": 1.7122412818981198e-05, "loss": 0.017, "step": 59360 }, { "grad_norm": 0.28313443064689636, "learning_rate": 1.7106843882342283e-05, "loss": 0.0149, "step": 59370 }, { "grad_norm": 0.15998320281505585, "learning_rate": 1.709128056624415e-05, "loss": 0.0144, "step": 59380 }, { "grad_norm": 0.33000677824020386, "learning_rate": 1.70757228733461e-05, "loss": 0.0125, "step": 59390 }, { "grad_norm": 0.14602671563625336, "learning_rate": 1.706017080630653e-05, "loss": 0.0105, "step": 59400 }, { "grad_norm": 0.3269295394420624, "learning_rate": 1.7044624367782873e-05, "loss": 0.0184, "step": 59410 }, { "grad_norm": 0.442179411649704, "learning_rate": 1.7029083560431553e-05, "loss": 0.0157, "step": 59420 }, { "grad_norm": 0.16956526041030884, "learning_rate": 1.701354838690813e-05, "loss": 0.0181, "step": 59430 }, { "grad_norm": 0.17325036227703094, "learning_rate": 1.6998018849867087e-05, "loss": 0.0123, "step": 59440 }, { "grad_norm": 0.17210251092910767, "learning_rate": 1.698249495196202e-05, "loss": 0.0143, "step": 59450 }, { "grad_norm": 0.2339795082807541, "learning_rate": 1.6966976695845528e-05, "loss": 0.0147, "step": 59460 }, { "grad_norm": 0.20815131068229675, "learning_rate": 1.6951464084169268e-05, "loss": 0.0135, "step": 59470 }, { "grad_norm": 0.21113884449005127, "learning_rate": 1.6935957119583916e-05, "loss": 0.0166, "step": 59480 }, { "grad_norm": 0.19049642980098724, "learning_rate": 1.6920455804739205e-05, "loss": 0.0137, "step": 59490 }, { "grad_norm": 0.29102540016174316, "learning_rate": 1.6904960142283855e-05, "loss": 0.0175, "step": 59500 }, { "grad_norm": 0.20130281150341034, "learning_rate": 1.6889470134865666e-05, "loss": 0.0139, "step": 59510 }, { "grad_norm": 0.2519213557243347, "learning_rate": 1.687398578513145e-05, "loss": 0.0168, "step": 59520 }, { "grad_norm": 0.22494861483573914, "learning_rate": 1.6858507095727066e-05, "loss": 0.0159, "step": 59530 }, { "grad_norm": 0.24533171951770782, "learning_rate": 1.6843034069297403e-05, "loss": 0.0132, "step": 59540 }, { "grad_norm": 0.34744009375572205, "learning_rate": 1.682756670848637e-05, "loss": 0.0269, "step": 59550 }, { "grad_norm": 0.19067604839801788, "learning_rate": 1.6812105015936937e-05, "loss": 0.01, "step": 59560 }, { "grad_norm": 0.2603560984134674, "learning_rate": 1.6796648994291027e-05, "loss": 0.018, "step": 59570 }, { "grad_norm": 0.18144913017749786, "learning_rate": 1.678119864618973e-05, "loss": 0.0099, "step": 59580 }, { "grad_norm": 0.32744869589805603, "learning_rate": 1.6765753974273023e-05, "loss": 0.0114, "step": 59590 }, { "grad_norm": 0.17606684565544128, "learning_rate": 1.675031498118001e-05, "loss": 0.0116, "step": 59600 }, { "grad_norm": 0.13997618854045868, "learning_rate": 1.6734881669548804e-05, "loss": 0.0122, "step": 59610 }, { "grad_norm": 0.2361157089471817, "learning_rate": 1.671945404201648e-05, "loss": 0.0226, "step": 59620 }, { "grad_norm": 0.29195117950439453, "learning_rate": 1.670403210121927e-05, "loss": 0.0102, "step": 59630 }, { "grad_norm": 0.16861657798290253, "learning_rate": 1.6688615849792312e-05, "loss": 0.0154, "step": 59640 }, { "grad_norm": 0.4833557903766632, "learning_rate": 1.6673205290369832e-05, "loss": 0.0204, "step": 59650 }, { "grad_norm": 0.3280336558818817, "learning_rate": 1.6657800425585086e-05, "loss": 0.0161, "step": 59660 }, { "grad_norm": 0.2871969938278198, "learning_rate": 1.6642401258070327e-05, "loss": 0.0289, "step": 59670 }, { "grad_norm": 0.3312750458717346, "learning_rate": 1.6627007790456884e-05, "loss": 0.0214, "step": 59680 }, { "grad_norm": 0.20739704370498657, "learning_rate": 1.6611620025375037e-05, "loss": 0.011, "step": 59690 }, { "grad_norm": 0.2325717657804489, "learning_rate": 1.6596237965454154e-05, "loss": 0.0164, "step": 59700 }, { "grad_norm": 0.36113980412483215, "learning_rate": 1.6580861613322607e-05, "loss": 0.0153, "step": 59710 }, { "grad_norm": 0.37946638464927673, "learning_rate": 1.6565490971607796e-05, "loss": 0.012, "step": 59720 }, { "grad_norm": 0.3316003680229187, "learning_rate": 1.655012604293614e-05, "loss": 0.0131, "step": 59730 }, { "grad_norm": 0.2515253722667694, "learning_rate": 1.653476682993309e-05, "loss": 0.0198, "step": 59740 }, { "grad_norm": 0.23481398820877075, "learning_rate": 1.651941333522311e-05, "loss": 0.0135, "step": 59750 }, { "grad_norm": 0.25429072976112366, "learning_rate": 1.6504065561429715e-05, "loss": 0.0164, "step": 59760 }, { "grad_norm": 0.21004478633403778, "learning_rate": 1.6488723511175385e-05, "loss": 0.0178, "step": 59770 }, { "grad_norm": 0.5378463268280029, "learning_rate": 1.6473387187081668e-05, "loss": 0.0145, "step": 59780 }, { "grad_norm": 0.138713538646698, "learning_rate": 1.6458056591769123e-05, "loss": 0.0101, "step": 59790 }, { "grad_norm": 0.2568652629852295, "learning_rate": 1.6442731727857335e-05, "loss": 0.0158, "step": 59800 }, { "grad_norm": 0.21776612102985382, "learning_rate": 1.642741259796492e-05, "loss": 0.012, "step": 59810 }, { "grad_norm": 0.23887315392494202, "learning_rate": 1.641209920470944e-05, "loss": 0.0122, "step": 59820 }, { "grad_norm": 0.386989563703537, "learning_rate": 1.639679155070762e-05, "loss": 0.0166, "step": 59830 }, { "grad_norm": 0.3387318551540375, "learning_rate": 1.6381489638575048e-05, "loss": 0.0157, "step": 59840 }, { "grad_norm": 0.19724181294441223, "learning_rate": 1.636619347092643e-05, "loss": 0.0132, "step": 59850 }, { "grad_norm": 0.24144408106803894, "learning_rate": 1.6350903050375476e-05, "loss": 0.0131, "step": 59860 }, { "grad_norm": 0.3206653594970703, "learning_rate": 1.6335618379534856e-05, "loss": 0.0152, "step": 59870 }, { "grad_norm": 0.2598188817501068, "learning_rate": 1.6320339461016364e-05, "loss": 0.013, "step": 59880 }, { "grad_norm": 0.19146257638931274, "learning_rate": 1.6305066297430687e-05, "loss": 0.0135, "step": 59890 }, { "grad_norm": 0.21092376112937927, "learning_rate": 1.6289798891387654e-05, "loss": 0.0187, "step": 59900 }, { "grad_norm": 0.2198619693517685, "learning_rate": 1.6274537245495995e-05, "loss": 0.0104, "step": 59910 }, { "grad_norm": 0.9852266907691956, "learning_rate": 1.6259281362363527e-05, "loss": 0.0143, "step": 59920 }, { "grad_norm": 0.23976634442806244, "learning_rate": 1.6244031244597068e-05, "loss": 0.0227, "step": 59930 }, { "grad_norm": 0.16562166810035706, "learning_rate": 1.622878689480244e-05, "loss": 0.0129, "step": 59940 }, { "grad_norm": 0.41320914030075073, "learning_rate": 1.6213548315584498e-05, "loss": 0.0186, "step": 59950 }, { "grad_norm": 0.24903346598148346, "learning_rate": 1.6198315509547074e-05, "loss": 0.016, "step": 59960 }, { "grad_norm": 0.12413537502288818, "learning_rate": 1.6183088479293056e-05, "loss": 0.0118, "step": 59970 }, { "grad_norm": 0.22668224573135376, "learning_rate": 1.6167867227424316e-05, "loss": 0.0145, "step": 59980 }, { "grad_norm": 0.38390958309173584, "learning_rate": 1.615265175654176e-05, "loss": 0.0191, "step": 59990 }, { "grad_norm": 0.3727737069129944, "learning_rate": 1.613744206924529e-05, "loss": 0.0174, "step": 60000 }, { "grad_norm": 0.17779850959777832, "learning_rate": 1.6122238168133845e-05, "loss": 0.0143, "step": 60010 }, { "grad_norm": 0.21929030120372772, "learning_rate": 1.6107040055805305e-05, "loss": 0.0117, "step": 60020 }, { "grad_norm": 0.20779351890087128, "learning_rate": 1.6091847734856673e-05, "loss": 0.0184, "step": 60030 }, { "grad_norm": 0.30118173360824585, "learning_rate": 1.6076661207883865e-05, "loss": 0.0144, "step": 60040 }, { "grad_norm": 0.39744463562965393, "learning_rate": 1.6061480477481848e-05, "loss": 0.0148, "step": 60050 }, { "grad_norm": 0.2815764546394348, "learning_rate": 1.604630554624461e-05, "loss": 0.0115, "step": 60060 }, { "grad_norm": 0.3999308943748474, "learning_rate": 1.603113641676509e-05, "loss": 0.0152, "step": 60070 }, { "grad_norm": 0.3211721181869507, "learning_rate": 1.6015973091635338e-05, "loss": 0.0147, "step": 60080 }, { "grad_norm": 0.29367098212242126, "learning_rate": 1.6000815573446288e-05, "loss": 0.013, "step": 60090 }, { "grad_norm": 0.18246285617351532, "learning_rate": 1.5985663864788002e-05, "loss": 0.0206, "step": 60100 }, { "grad_norm": 0.17030851542949677, "learning_rate": 1.597051796824946e-05, "loss": 0.014, "step": 60110 }, { "grad_norm": 0.2665877342224121, "learning_rate": 1.5955377886418682e-05, "loss": 0.018, "step": 60120 }, { "grad_norm": 0.40279337763786316, "learning_rate": 1.5940243621882704e-05, "loss": 0.0135, "step": 60130 }, { "grad_norm": 0.33561718463897705, "learning_rate": 1.5925115177227555e-05, "loss": 0.0197, "step": 60140 }, { "grad_norm": 0.4402807354927063, "learning_rate": 1.5909992555038288e-05, "loss": 0.0165, "step": 60150 }, { "grad_norm": 3.0268895626068115, "learning_rate": 1.5894875757898912e-05, "loss": 0.0238, "step": 60160 }, { "grad_norm": 0.251738041639328, "learning_rate": 1.58797647883925e-05, "loss": 0.0149, "step": 60170 }, { "grad_norm": 0.1880682110786438, "learning_rate": 1.586465964910109e-05, "loss": 0.0159, "step": 60180 }, { "grad_norm": 0.18383395671844482, "learning_rate": 1.5849560342605734e-05, "loss": 0.0144, "step": 60190 }, { "grad_norm": 0.17579086124897003, "learning_rate": 1.583446687148651e-05, "loss": 0.0112, "step": 60200 }, { "grad_norm": 0.16904260218143463, "learning_rate": 1.581937923832248e-05, "loss": 0.0156, "step": 60210 }, { "grad_norm": 0.24149858951568604, "learning_rate": 1.5804297445691663e-05, "loss": 0.0155, "step": 60220 }, { "grad_norm": 0.3023631274700165, "learning_rate": 1.578922149617119e-05, "loss": 0.0193, "step": 60230 }, { "grad_norm": 0.2724807858467102, "learning_rate": 1.5774151392337084e-05, "loss": 0.0151, "step": 60240 }, { "grad_norm": 0.29661044478416443, "learning_rate": 1.575908713676442e-05, "loss": 0.0142, "step": 60250 }, { "grad_norm": 0.43790528178215027, "learning_rate": 1.5744028732027293e-05, "loss": 0.0207, "step": 60260 }, { "grad_norm": 0.2876560389995575, "learning_rate": 1.5728976180698723e-05, "loss": 0.0162, "step": 60270 }, { "grad_norm": 0.22709304094314575, "learning_rate": 1.5713929485350836e-05, "loss": 0.0173, "step": 60280 }, { "grad_norm": 0.2153117060661316, "learning_rate": 1.569888864855465e-05, "loss": 0.0151, "step": 60290 }, { "grad_norm": 0.19177138805389404, "learning_rate": 1.568385367288028e-05, "loss": 0.0111, "step": 60300 }, { "grad_norm": 0.1767629235982895, "learning_rate": 1.5668824560896755e-05, "loss": 0.0163, "step": 60310 }, { "grad_norm": 0.1902145892381668, "learning_rate": 1.5653801315172156e-05, "loss": 0.014, "step": 60320 }, { "grad_norm": 0.2580455243587494, "learning_rate": 1.5638783938273554e-05, "loss": 0.0106, "step": 60330 }, { "grad_norm": 0.3554404675960541, "learning_rate": 1.5623772432766966e-05, "loss": 0.0143, "step": 60340 }, { "grad_norm": 0.5058126449584961, "learning_rate": 1.5608766801217507e-05, "loss": 0.0161, "step": 60350 }, { "grad_norm": 0.3203747272491455, "learning_rate": 1.5593767046189184e-05, "loss": 0.0144, "step": 60360 }, { "grad_norm": 0.28000935912132263, "learning_rate": 1.5578773170245064e-05, "loss": 0.0109, "step": 60370 }, { "grad_norm": 0.38423874974250793, "learning_rate": 1.5563785175947182e-05, "loss": 0.02, "step": 60380 }, { "grad_norm": 0.20073308050632477, "learning_rate": 1.5548803065856582e-05, "loss": 0.0119, "step": 60390 }, { "grad_norm": 0.3187839984893799, "learning_rate": 1.55338268425333e-05, "loss": 0.0152, "step": 60400 }, { "grad_norm": 0.2463352382183075, "learning_rate": 1.5518856508536373e-05, "loss": 0.0174, "step": 60410 }, { "grad_norm": 0.3539135158061981, "learning_rate": 1.5503892066423786e-05, "loss": 0.0176, "step": 60420 }, { "grad_norm": 0.489584356546402, "learning_rate": 1.548893351875258e-05, "loss": 0.0229, "step": 60430 }, { "grad_norm": 0.3352455198764801, "learning_rate": 1.547398086807875e-05, "loss": 0.0193, "step": 60440 }, { "grad_norm": 0.2503279745578766, "learning_rate": 1.5459034116957305e-05, "loss": 0.0238, "step": 60450 }, { "grad_norm": 0.1992761492729187, "learning_rate": 1.544409326794225e-05, "loss": 0.0182, "step": 60460 }, { "grad_norm": 0.1762438863515854, "learning_rate": 1.542915832358651e-05, "loss": 0.0163, "step": 60470 }, { "grad_norm": 0.2298070639371872, "learning_rate": 1.541422928644214e-05, "loss": 0.0131, "step": 60480 }, { "grad_norm": 0.3418499827384949, "learning_rate": 1.5399306159060024e-05, "loss": 0.0208, "step": 60490 }, { "grad_norm": 0.2985387444496155, "learning_rate": 1.5384388943990185e-05, "loss": 0.0195, "step": 60500 }, { "grad_norm": 0.29824593663215637, "learning_rate": 1.5369477643781526e-05, "loss": 0.0157, "step": 60510 }, { "grad_norm": 0.29765594005584717, "learning_rate": 1.5354572260981985e-05, "loss": 0.0166, "step": 60520 }, { "grad_norm": 0.2359454333782196, "learning_rate": 1.533967279813851e-05, "loss": 0.0098, "step": 60530 }, { "grad_norm": 0.2920534014701843, "learning_rate": 1.5324779257796956e-05, "loss": 0.0156, "step": 60540 }, { "grad_norm": 0.14824919402599335, "learning_rate": 1.5309891642502293e-05, "loss": 0.0163, "step": 60550 }, { "grad_norm": 0.23557667434215546, "learning_rate": 1.5295009954798357e-05, "loss": 0.01, "step": 60560 }, { "grad_norm": 0.23439525067806244, "learning_rate": 1.5280134197228036e-05, "loss": 0.0144, "step": 60570 }, { "grad_norm": 0.1736963391304016, "learning_rate": 1.526526437233319e-05, "loss": 0.0187, "step": 60580 }, { "grad_norm": 0.18508431315422058, "learning_rate": 1.525040048265466e-05, "loss": 0.0108, "step": 60590 }, { "grad_norm": 0.21268483996391296, "learning_rate": 1.5235542530732288e-05, "loss": 0.0122, "step": 60600 }, { "grad_norm": 0.1527978628873825, "learning_rate": 1.5220690519104901e-05, "loss": 0.014, "step": 60610 }, { "grad_norm": 0.24184896051883698, "learning_rate": 1.5205844450310275e-05, "loss": 0.0145, "step": 60620 }, { "grad_norm": 0.22368738055229187, "learning_rate": 1.519100432688521e-05, "loss": 0.015, "step": 60630 }, { "grad_norm": 0.433597594499588, "learning_rate": 1.5176170151365476e-05, "loss": 0.0186, "step": 60640 }, { "grad_norm": 0.2694525420665741, "learning_rate": 1.5161341926285833e-05, "loss": 0.0108, "step": 60650 }, { "grad_norm": 0.1378856599330902, "learning_rate": 1.5146519654180025e-05, "loss": 0.0145, "step": 60660 }, { "grad_norm": 0.25029245018959045, "learning_rate": 1.5131703337580739e-05, "loss": 0.0182, "step": 60670 }, { "grad_norm": 0.19994565844535828, "learning_rate": 1.5116892979019731e-05, "loss": 0.0156, "step": 60680 }, { "grad_norm": 0.24186605215072632, "learning_rate": 1.5102088581027623e-05, "loss": 0.0129, "step": 60690 }, { "grad_norm": 0.23338010907173157, "learning_rate": 1.5087290146134154e-05, "loss": 0.0105, "step": 60700 }, { "grad_norm": 0.1257435530424118, "learning_rate": 1.5072497676867915e-05, "loss": 0.0113, "step": 60710 }, { "grad_norm": 0.3903782367706299, "learning_rate": 1.505771117575655e-05, "loss": 0.0181, "step": 60720 }, { "grad_norm": 0.4128604531288147, "learning_rate": 1.5042930645326691e-05, "loss": 0.0133, "step": 60730 }, { "grad_norm": 0.33340805768966675, "learning_rate": 1.5028156088103879e-05, "loss": 0.0156, "step": 60740 }, { "grad_norm": 0.4478631913661957, "learning_rate": 1.5013387506612735e-05, "loss": 0.0142, "step": 60750 }, { "grad_norm": 0.11238206923007965, "learning_rate": 1.4998624903376767e-05, "loss": 0.015, "step": 60760 }, { "grad_norm": 0.6035367250442505, "learning_rate": 1.4983868280918517e-05, "loss": 0.0133, "step": 60770 }, { "grad_norm": 0.21307124197483063, "learning_rate": 1.4969117641759478e-05, "loss": 0.0181, "step": 60780 }, { "grad_norm": 0.18819686770439148, "learning_rate": 1.495437298842014e-05, "loss": 0.0172, "step": 60790 }, { "grad_norm": 0.3491462171077728, "learning_rate": 1.4939634323419976e-05, "loss": 0.0183, "step": 60800 }, { "grad_norm": 0.28747618198394775, "learning_rate": 1.4924901649277384e-05, "loss": 0.0168, "step": 60810 }, { "grad_norm": 0.2997187674045563, "learning_rate": 1.4910174968509793e-05, "loss": 0.0119, "step": 60820 }, { "grad_norm": 0.3022850453853607, "learning_rate": 1.4895454283633598e-05, "loss": 0.0168, "step": 60830 }, { "grad_norm": 0.32283198833465576, "learning_rate": 1.488073959716415e-05, "loss": 0.0154, "step": 60840 }, { "grad_norm": 0.2568976581096649, "learning_rate": 1.4866030911615791e-05, "loss": 0.0161, "step": 60850 }, { "grad_norm": 0.23033972084522247, "learning_rate": 1.4851328229501849e-05, "loss": 0.0181, "step": 60860 }, { "grad_norm": 0.203863263130188, "learning_rate": 1.4836631553334562e-05, "loss": 0.0126, "step": 60870 }, { "grad_norm": 0.34346452355384827, "learning_rate": 1.4821940885625251e-05, "loss": 0.0142, "step": 60880 }, { "grad_norm": 0.19063781201839447, "learning_rate": 1.4807256228884109e-05, "loss": 0.0128, "step": 60890 }, { "grad_norm": 0.15741115808486938, "learning_rate": 1.4792577585620353e-05, "loss": 0.0152, "step": 60900 }, { "grad_norm": 0.35475969314575195, "learning_rate": 1.4777904958342164e-05, "loss": 0.0226, "step": 60910 }, { "grad_norm": 0.24045996367931366, "learning_rate": 1.4763238349556691e-05, "loss": 0.0155, "step": 60920 }, { "grad_norm": 0.28391018509864807, "learning_rate": 1.4748577761770072e-05, "loss": 0.0183, "step": 60930 }, { "grad_norm": 0.15989969670772552, "learning_rate": 1.4733923197487354e-05, "loss": 0.0158, "step": 60940 }, { "grad_norm": 0.2665354311466217, "learning_rate": 1.4719274659212662e-05, "loss": 0.0183, "step": 60950 }, { "grad_norm": 0.3279563784599304, "learning_rate": 1.4704632149448983e-05, "loss": 0.0174, "step": 60960 }, { "grad_norm": 0.3469580113887787, "learning_rate": 1.4689995670698343e-05, "loss": 0.0218, "step": 60970 }, { "grad_norm": 0.2595861256122589, "learning_rate": 1.4675365225461728e-05, "loss": 0.0205, "step": 60980 }, { "grad_norm": 0.3180769383907318, "learning_rate": 1.4660740816239032e-05, "loss": 0.0121, "step": 60990 }, { "grad_norm": 0.2540958523750305, "learning_rate": 1.464612244552923e-05, "loss": 0.0183, "step": 61000 }, { "grad_norm": 0.17552396655082703, "learning_rate": 1.4631510115830161e-05, "loss": 0.0159, "step": 61010 }, { "grad_norm": 0.2272872030735016, "learning_rate": 1.4616903829638679e-05, "loss": 0.0111, "step": 61020 }, { "grad_norm": 0.13015934824943542, "learning_rate": 1.46023035894506e-05, "loss": 0.015, "step": 61030 }, { "grad_norm": 0.40238356590270996, "learning_rate": 1.4587709397760713e-05, "loss": 0.0152, "step": 61040 }, { "grad_norm": 0.1819818764925003, "learning_rate": 1.4573121257062755e-05, "loss": 0.016, "step": 61050 }, { "grad_norm": 0.3750613331794739, "learning_rate": 1.4558539169849472e-05, "loss": 0.0215, "step": 61060 }, { "grad_norm": 0.2791703939437866, "learning_rate": 1.4543963138612499e-05, "loss": 0.0127, "step": 61070 }, { "grad_norm": 0.27186647057533264, "learning_rate": 1.4529393165842498e-05, "loss": 0.0144, "step": 61080 }, { "grad_norm": 0.2803604006767273, "learning_rate": 1.4514829254029084e-05, "loss": 0.0174, "step": 61090 }, { "grad_norm": 0.1553741991519928, "learning_rate": 1.450027140566083e-05, "loss": 0.0113, "step": 61100 }, { "grad_norm": 0.3938332796096802, "learning_rate": 1.4485719623225268e-05, "loss": 0.0135, "step": 61110 }, { "grad_norm": 0.2416401356458664, "learning_rate": 1.4471173909208912e-05, "loss": 0.016, "step": 61120 }, { "grad_norm": 0.29977089166641235, "learning_rate": 1.4456634266097236e-05, "loss": 0.0152, "step": 61130 }, { "grad_norm": 0.31415802240371704, "learning_rate": 1.444210069637461e-05, "loss": 0.0225, "step": 61140 }, { "grad_norm": 0.22763167321681976, "learning_rate": 1.4427573202524502e-05, "loss": 0.0093, "step": 61150 }, { "grad_norm": 0.15663444995880127, "learning_rate": 1.4413051787029208e-05, "loss": 0.0154, "step": 61160 }, { "grad_norm": 0.4410029947757721, "learning_rate": 1.439853645237006e-05, "loss": 0.016, "step": 61170 }, { "grad_norm": 0.28348714113235474, "learning_rate": 1.4384027201027344e-05, "loss": 0.0142, "step": 61180 }, { "grad_norm": 0.15061624348163605, "learning_rate": 1.4369524035480253e-05, "loss": 0.0155, "step": 61190 }, { "grad_norm": 0.1713387817144394, "learning_rate": 1.4355026958207035e-05, "loss": 0.0129, "step": 61200 }, { "grad_norm": 0.28890472650527954, "learning_rate": 1.434053597168481e-05, "loss": 0.0135, "step": 61210 }, { "grad_norm": 0.25242897868156433, "learning_rate": 1.432605107838969e-05, "loss": 0.0156, "step": 61220 }, { "grad_norm": 0.3213488757610321, "learning_rate": 1.4311572280796759e-05, "loss": 0.0198, "step": 61230 }, { "grad_norm": 0.2339707762002945, "learning_rate": 1.4297099581380047e-05, "loss": 0.0183, "step": 61240 }, { "grad_norm": 0.26230406761169434, "learning_rate": 1.4282632982612538e-05, "loss": 0.0159, "step": 61250 }, { "grad_norm": 0.22986064851284027, "learning_rate": 1.4268172486966197e-05, "loss": 0.0176, "step": 61260 }, { "grad_norm": 0.1858847737312317, "learning_rate": 1.4253718096911894e-05, "loss": 0.015, "step": 61270 }, { "grad_norm": 0.302778959274292, "learning_rate": 1.4239269814919504e-05, "loss": 0.0196, "step": 61280 }, { "grad_norm": 0.339534193277359, "learning_rate": 1.4224827643457843e-05, "loss": 0.0258, "step": 61290 }, { "grad_norm": 0.3123270571231842, "learning_rate": 1.4210391584994686e-05, "loss": 0.0118, "step": 61300 }, { "grad_norm": 0.25217682123184204, "learning_rate": 1.4195961641996763e-05, "loss": 0.0171, "step": 61310 }, { "grad_norm": 0.27988994121551514, "learning_rate": 1.4181537816929751e-05, "loss": 0.0182, "step": 61320 }, { "grad_norm": 0.2616284489631653, "learning_rate": 1.4167120112258303e-05, "loss": 0.0208, "step": 61330 }, { "grad_norm": 0.19715091586112976, "learning_rate": 1.4152708530445963e-05, "loss": 0.0117, "step": 61340 }, { "grad_norm": 0.2525090277194977, "learning_rate": 1.4138303073955344e-05, "loss": 0.0148, "step": 61350 }, { "grad_norm": 0.23319409787654877, "learning_rate": 1.4123903745247897e-05, "loss": 0.0135, "step": 61360 }, { "grad_norm": 0.12377496063709259, "learning_rate": 1.4109510546784082e-05, "loss": 0.0117, "step": 61370 }, { "grad_norm": 0.1490655541419983, "learning_rate": 1.4095123481023325e-05, "loss": 0.0171, "step": 61380 }, { "grad_norm": 0.2192276120185852, "learning_rate": 1.4080742550423936e-05, "loss": 0.0153, "step": 61390 }, { "grad_norm": 0.21898888051509857, "learning_rate": 1.4066367757443278e-05, "loss": 0.0125, "step": 61400 }, { "grad_norm": 0.3539740741252899, "learning_rate": 1.4051999104537572e-05, "loss": 0.0167, "step": 61410 }, { "grad_norm": 0.35660871863365173, "learning_rate": 1.4037636594162034e-05, "loss": 0.0135, "step": 61420 }, { "grad_norm": 0.25858572125434875, "learning_rate": 1.402328022877083e-05, "loss": 0.014, "step": 61430 }, { "grad_norm": 0.16426682472229004, "learning_rate": 1.4008930010817073e-05, "loss": 0.0167, "step": 61440 }, { "grad_norm": 0.22487585246562958, "learning_rate": 1.3994585942752831e-05, "loss": 0.0119, "step": 61450 }, { "grad_norm": 0.10294366627931595, "learning_rate": 1.398024802702909e-05, "loss": 0.013, "step": 61460 }, { "grad_norm": 0.19667527079582214, "learning_rate": 1.3965916266095819e-05, "loss": 0.0165, "step": 61470 }, { "grad_norm": 0.35714584589004517, "learning_rate": 1.395159066240192e-05, "loss": 0.0159, "step": 61480 }, { "grad_norm": 0.23635387420654297, "learning_rate": 1.3937271218395248e-05, "loss": 0.015, "step": 61490 }, { "grad_norm": 0.18548740446567535, "learning_rate": 1.3922957936522613e-05, "loss": 0.0156, "step": 61500 }, { "grad_norm": 0.5379549860954285, "learning_rate": 1.3908650819229758e-05, "loss": 0.0146, "step": 61510 }, { "grad_norm": 0.40169456601142883, "learning_rate": 1.389434986896137e-05, "loss": 0.0154, "step": 61520 }, { "grad_norm": 0.25861284136772156, "learning_rate": 1.388005508816112e-05, "loss": 0.0103, "step": 61530 }, { "grad_norm": 0.426114946603775, "learning_rate": 1.3865766479271557e-05, "loss": 0.018, "step": 61540 }, { "grad_norm": 0.24147099256515503, "learning_rate": 1.3851484044734225e-05, "loss": 0.0196, "step": 61550 }, { "grad_norm": 0.23009197413921356, "learning_rate": 1.3837207786989608e-05, "loss": 0.0182, "step": 61560 }, { "grad_norm": 0.2849201261997223, "learning_rate": 1.382293770847713e-05, "loss": 0.0137, "step": 61570 }, { "grad_norm": 0.39804190397262573, "learning_rate": 1.3808673811635159e-05, "loss": 0.0139, "step": 61580 }, { "grad_norm": 0.27081748843193054, "learning_rate": 1.3794416098900975e-05, "loss": 0.015, "step": 61590 }, { "grad_norm": 0.3621445894241333, "learning_rate": 1.3780164572710886e-05, "loss": 0.0183, "step": 61600 }, { "grad_norm": 0.31640559434890747, "learning_rate": 1.3765919235500035e-05, "loss": 0.0196, "step": 61610 }, { "grad_norm": 0.2345142364501953, "learning_rate": 1.375168008970259e-05, "loss": 0.0166, "step": 61620 }, { "grad_norm": 0.10552321374416351, "learning_rate": 1.3737447137751635e-05, "loss": 0.016, "step": 61630 }, { "grad_norm": 0.3873700797557831, "learning_rate": 1.372322038207915e-05, "loss": 0.0149, "step": 61640 }, { "grad_norm": 0.27543848752975464, "learning_rate": 1.3708999825116159e-05, "loss": 0.0168, "step": 61650 }, { "grad_norm": 0.37638914585113525, "learning_rate": 1.3694785469292526e-05, "loss": 0.0201, "step": 61660 }, { "grad_norm": 0.24814778566360474, "learning_rate": 1.3680577317037101e-05, "loss": 0.0139, "step": 61670 }, { "grad_norm": 0.35229310393333435, "learning_rate": 1.3666375370777678e-05, "loss": 0.015, "step": 61680 }, { "grad_norm": 0.2626833915710449, "learning_rate": 1.3652179632940981e-05, "loss": 0.0137, "step": 61690 }, { "grad_norm": 0.15480612218379974, "learning_rate": 1.3637990105952664e-05, "loss": 0.0141, "step": 61700 }, { "grad_norm": 0.1656237244606018, "learning_rate": 1.3623806792237337e-05, "loss": 0.0148, "step": 61710 }, { "grad_norm": 0.2788442075252533, "learning_rate": 1.3609629694218551e-05, "loss": 0.0177, "step": 61720 }, { "grad_norm": 0.2069205343723297, "learning_rate": 1.3595458814318762e-05, "loss": 0.0126, "step": 61730 }, { "grad_norm": 0.18944722414016724, "learning_rate": 1.3581294154959389e-05, "loss": 0.016, "step": 61740 }, { "grad_norm": 0.30981600284576416, "learning_rate": 1.3567135718560792e-05, "loss": 0.0133, "step": 61750 }, { "grad_norm": 0.2418190836906433, "learning_rate": 1.3552983507542261e-05, "loss": 0.0135, "step": 61760 }, { "grad_norm": 0.17211222648620605, "learning_rate": 1.3538837524322023e-05, "loss": 0.0163, "step": 61770 }, { "grad_norm": 0.15634629130363464, "learning_rate": 1.3524697771317251e-05, "loss": 0.0201, "step": 61780 }, { "grad_norm": 0.19353654980659485, "learning_rate": 1.3510564250943997e-05, "loss": 0.0151, "step": 61790 }, { "grad_norm": 0.19756098091602325, "learning_rate": 1.3496436965617353e-05, "loss": 0.0128, "step": 61800 }, { "grad_norm": 0.29560819268226624, "learning_rate": 1.3482315917751243e-05, "loss": 0.017, "step": 61810 }, { "grad_norm": 0.2265034019947052, "learning_rate": 1.3468201109758582e-05, "loss": 0.0117, "step": 61820 }, { "grad_norm": 0.20987851917743683, "learning_rate": 1.3454092544051222e-05, "loss": 0.0174, "step": 61830 }, { "grad_norm": 0.2809975743293762, "learning_rate": 1.3439990223039878e-05, "loss": 0.0153, "step": 61840 }, { "grad_norm": 0.1562342792749405, "learning_rate": 1.3425894149134321e-05, "loss": 0.0134, "step": 61850 }, { "grad_norm": 0.2723090648651123, "learning_rate": 1.3411804324743132e-05, "loss": 0.014, "step": 61860 }, { "grad_norm": 0.31283754110336304, "learning_rate": 1.3397720752273896e-05, "loss": 0.0179, "step": 61870 }, { "grad_norm": 0.22220686078071594, "learning_rate": 1.3383643434133108e-05, "loss": 0.0123, "step": 61880 }, { "grad_norm": 0.2715665400028229, "learning_rate": 1.3369572372726196e-05, "loss": 0.0141, "step": 61890 }, { "grad_norm": 0.2028755247592926, "learning_rate": 1.3355507570457525e-05, "loss": 0.0131, "step": 61900 }, { "grad_norm": 0.16739287972450256, "learning_rate": 1.334144902973038e-05, "loss": 0.0135, "step": 61910 }, { "grad_norm": 0.1322897970676422, "learning_rate": 1.3327396752946997e-05, "loss": 0.0124, "step": 61920 }, { "grad_norm": 0.23201043903827667, "learning_rate": 1.3313350742508496e-05, "loss": 0.0129, "step": 61930 }, { "grad_norm": 0.3043639659881592, "learning_rate": 1.3299311000814973e-05, "loss": 0.0171, "step": 61940 }, { "grad_norm": 0.20334680378437042, "learning_rate": 1.3285277530265432e-05, "loss": 0.0164, "step": 61950 }, { "grad_norm": 0.3419192433357239, "learning_rate": 1.3271250333257813e-05, "loss": 0.0158, "step": 61960 }, { "grad_norm": 0.19476495683193207, "learning_rate": 1.3257229412188977e-05, "loss": 0.0104, "step": 61970 }, { "grad_norm": 0.3194551467895508, "learning_rate": 1.3243214769454726e-05, "loss": 0.0145, "step": 61980 }, { "grad_norm": 0.31290844082832336, "learning_rate": 1.3229206407449751e-05, "loss": 0.0118, "step": 61990 }, { "grad_norm": 0.19644193351268768, "learning_rate": 1.3215204328567742e-05, "loss": 0.0173, "step": 62000 }, { "grad_norm": 0.1952521950006485, "learning_rate": 1.3201208535201232e-05, "loss": 0.0124, "step": 62010 }, { "grad_norm": 0.18353073298931122, "learning_rate": 1.3187219029741737e-05, "loss": 0.0122, "step": 62020 }, { "grad_norm": 0.3954136371612549, "learning_rate": 1.3173235814579693e-05, "loss": 0.0124, "step": 62030 }, { "grad_norm": 0.2515047788619995, "learning_rate": 1.3159258892104398e-05, "loss": 0.0191, "step": 62040 }, { "grad_norm": 0.12270504236221313, "learning_rate": 1.3145288264704198e-05, "loss": 0.0144, "step": 62050 }, { "grad_norm": 0.0851830542087555, "learning_rate": 1.3131323934766237e-05, "loss": 0.0111, "step": 62060 }, { "grad_norm": 0.184568852186203, "learning_rate": 1.311736590467666e-05, "loss": 0.0148, "step": 62070 }, { "grad_norm": 0.4764483571052551, "learning_rate": 1.3103414176820505e-05, "loss": 0.0121, "step": 62080 }, { "grad_norm": 0.2853168845176697, "learning_rate": 1.308946875358174e-05, "loss": 0.024, "step": 62090 }, { "grad_norm": 0.2130938470363617, "learning_rate": 1.3075529637343276e-05, "loss": 0.0224, "step": 62100 }, { "grad_norm": 0.30810028314590454, "learning_rate": 1.3061596830486883e-05, "loss": 0.0153, "step": 62110 }, { "grad_norm": 0.34417295455932617, "learning_rate": 1.3047670335393353e-05, "loss": 0.0135, "step": 62120 }, { "grad_norm": 0.22708772122859955, "learning_rate": 1.3033750154442298e-05, "loss": 0.0214, "step": 62130 }, { "grad_norm": 0.27140259742736816, "learning_rate": 1.3019836290012316e-05, "loss": 0.0125, "step": 62140 }, { "grad_norm": 0.26068246364593506, "learning_rate": 1.30059287444809e-05, "loss": 0.016, "step": 62150 }, { "grad_norm": 0.2916980981826782, "learning_rate": 1.299202752022447e-05, "loss": 0.0226, "step": 62160 }, { "grad_norm": 0.1940343677997589, "learning_rate": 1.2978132619618371e-05, "loss": 0.0102, "step": 62170 }, { "grad_norm": 0.31287601590156555, "learning_rate": 1.2964244045036866e-05, "loss": 0.0199, "step": 62180 }, { "grad_norm": 0.29189398884773254, "learning_rate": 1.295036179885311e-05, "loss": 0.0146, "step": 62190 }, { "grad_norm": 0.18589241802692413, "learning_rate": 1.293648588343922e-05, "loss": 0.0155, "step": 62200 }, { "grad_norm": 0.24877311289310455, "learning_rate": 1.2922616301166196e-05, "loss": 0.0149, "step": 62210 }, { "grad_norm": 0.31607216596603394, "learning_rate": 1.2908753054403976e-05, "loss": 0.0165, "step": 62220 }, { "grad_norm": 0.25889113545417786, "learning_rate": 1.2894896145521429e-05, "loss": 0.0123, "step": 62230 }, { "grad_norm": 0.21419325470924377, "learning_rate": 1.2881045576886275e-05, "loss": 0.0135, "step": 62240 }, { "grad_norm": 0.270045667886734, "learning_rate": 1.2867201350865254e-05, "loss": 0.0181, "step": 62250 }, { "grad_norm": 0.23979534208774567, "learning_rate": 1.2853363469823914e-05, "loss": 0.0195, "step": 62260 }, { "grad_norm": 0.2815132439136505, "learning_rate": 1.283953193612682e-05, "loss": 0.0131, "step": 62270 }, { "grad_norm": 0.3670964241027832, "learning_rate": 1.2825706752137372e-05, "loss": 0.0134, "step": 62280 }, { "grad_norm": 0.42547887563705444, "learning_rate": 1.2811887920217896e-05, "loss": 0.0144, "step": 62290 }, { "grad_norm": 0.18712569773197174, "learning_rate": 1.2798075442729707e-05, "loss": 0.0148, "step": 62300 }, { "grad_norm": 0.22162072360515594, "learning_rate": 1.2784269322032922e-05, "loss": 0.0137, "step": 62310 }, { "grad_norm": 0.20256103575229645, "learning_rate": 1.2770469560486686e-05, "loss": 0.0157, "step": 62320 }, { "grad_norm": 0.248846635222435, "learning_rate": 1.2756676160448956e-05, "loss": 0.0149, "step": 62330 }, { "grad_norm": 0.19774624705314636, "learning_rate": 1.2742889124276663e-05, "loss": 0.0119, "step": 62340 }, { "grad_norm": 0.1470554769039154, "learning_rate": 1.2729108454325639e-05, "loss": 0.0149, "step": 62350 }, { "grad_norm": 0.3124602735042572, "learning_rate": 1.2715334152950614e-05, "loss": 0.0164, "step": 62360 }, { "grad_norm": 0.16945958137512207, "learning_rate": 1.2701566222505246e-05, "loss": 0.0137, "step": 62370 }, { "grad_norm": 0.3147033154964447, "learning_rate": 1.2687804665342107e-05, "loss": 0.0158, "step": 62380 }, { "grad_norm": 0.3061855137348175, "learning_rate": 1.267404948381265e-05, "loss": 0.0141, "step": 62390 }, { "grad_norm": 0.1846502125263214, "learning_rate": 1.2660300680267267e-05, "loss": 0.0105, "step": 62400 }, { "grad_norm": 0.27847784757614136, "learning_rate": 1.2646558257055257e-05, "loss": 0.0128, "step": 62410 }, { "grad_norm": 0.27899789810180664, "learning_rate": 1.263282221652482e-05, "loss": 0.0096, "step": 62420 }, { "grad_norm": 0.22655029594898224, "learning_rate": 1.2619092561023088e-05, "loss": 0.0159, "step": 62430 }, { "grad_norm": 0.45517632365226746, "learning_rate": 1.2605369292896036e-05, "loss": 0.0168, "step": 62440 }, { "grad_norm": 0.22328892350196838, "learning_rate": 1.2591652414488658e-05, "loss": 0.0128, "step": 62450 }, { "grad_norm": 0.27185872197151184, "learning_rate": 1.2577941928144732e-05, "loss": 0.0147, "step": 62460 }, { "grad_norm": 0.24606238305568695, "learning_rate": 1.256423783620706e-05, "loss": 0.0151, "step": 62470 }, { "grad_norm": 0.28315746784210205, "learning_rate": 1.2550540141017264e-05, "loss": 0.0143, "step": 62480 }, { "grad_norm": 0.25876981019973755, "learning_rate": 1.253684884491591e-05, "loss": 0.0113, "step": 62490 }, { "grad_norm": 0.3260371685028076, "learning_rate": 1.2523163950242483e-05, "loss": 0.0112, "step": 62500 }, { "grad_norm": 0.25968316197395325, "learning_rate": 1.2509485459335313e-05, "loss": 0.0132, "step": 62510 }, { "grad_norm": 0.1904831826686859, "learning_rate": 1.2495813374531739e-05, "loss": 0.0123, "step": 62520 }, { "grad_norm": 0.23737294971942902, "learning_rate": 1.2482147698167907e-05, "loss": 0.0111, "step": 62530 }, { "grad_norm": 0.14495672285556793, "learning_rate": 1.2468488432578911e-05, "loss": 0.0139, "step": 62540 }, { "grad_norm": 0.24745064973831177, "learning_rate": 1.2454835580098761e-05, "loss": 0.0151, "step": 62550 }, { "grad_norm": 0.319993257522583, "learning_rate": 1.2441189143060338e-05, "loss": 0.0131, "step": 62560 }, { "grad_norm": 0.25729456543922424, "learning_rate": 1.242754912379548e-05, "loss": 0.0135, "step": 62570 }, { "grad_norm": 0.5057744979858398, "learning_rate": 1.2413915524634844e-05, "loss": 0.0265, "step": 62580 }, { "grad_norm": 0.2099703699350357, "learning_rate": 1.2400288347908073e-05, "loss": 0.016, "step": 62590 }, { "grad_norm": 0.3010994493961334, "learning_rate": 1.2386667595943663e-05, "loss": 0.0127, "step": 62600 }, { "grad_norm": 0.34255465865135193, "learning_rate": 1.2373053271069035e-05, "loss": 0.0109, "step": 62610 }, { "grad_norm": 0.3305455148220062, "learning_rate": 1.2359445375610501e-05, "loss": 0.0139, "step": 62620 }, { "grad_norm": 0.1383880078792572, "learning_rate": 1.2345843911893301e-05, "loss": 0.011, "step": 62630 }, { "grad_norm": 0.22010304033756256, "learning_rate": 1.2332248882241498e-05, "loss": 0.0219, "step": 62640 }, { "grad_norm": 0.284098356962204, "learning_rate": 1.2318660288978178e-05, "loss": 0.0145, "step": 62650 }, { "grad_norm": 0.1615922898054123, "learning_rate": 1.2305078134425213e-05, "loss": 0.0131, "step": 62660 }, { "grad_norm": 0.19674888253211975, "learning_rate": 1.2291502420903434e-05, "loss": 0.0139, "step": 62670 }, { "grad_norm": 0.3218769431114197, "learning_rate": 1.2277933150732567e-05, "loss": 0.018, "step": 62680 }, { "grad_norm": 0.13175925612449646, "learning_rate": 1.2264370326231216e-05, "loss": 0.0117, "step": 62690 }, { "grad_norm": 0.3355604112148285, "learning_rate": 1.2250813949716927e-05, "loss": 0.0152, "step": 62700 }, { "grad_norm": 0.17754115164279938, "learning_rate": 1.2237264023506063e-05, "loss": 0.0163, "step": 62710 }, { "grad_norm": 0.5752263069152832, "learning_rate": 1.2223720549913987e-05, "loss": 0.0119, "step": 62720 }, { "grad_norm": 0.19041800498962402, "learning_rate": 1.221018353125487e-05, "loss": 0.013, "step": 62730 }, { "grad_norm": 0.18886688351631165, "learning_rate": 1.2196652969841837e-05, "loss": 0.0124, "step": 62740 }, { "grad_norm": 0.27087777853012085, "learning_rate": 1.2183128867986904e-05, "loss": 0.02, "step": 62750 }, { "grad_norm": 0.28432005643844604, "learning_rate": 1.2169611228000927e-05, "loss": 0.0118, "step": 62760 }, { "grad_norm": 0.19554395973682404, "learning_rate": 1.2156100052193752e-05, "loss": 0.0158, "step": 62770 }, { "grad_norm": 0.19581350684165955, "learning_rate": 1.214259534287403e-05, "loss": 0.012, "step": 62780 }, { "grad_norm": 0.12161851674318314, "learning_rate": 1.2129097102349363e-05, "loss": 0.0114, "step": 62790 }, { "grad_norm": 0.5779053568840027, "learning_rate": 1.2115605332926227e-05, "loss": 0.0149, "step": 62800 }, { "grad_norm": 0.3030970096588135, "learning_rate": 1.2102120036910003e-05, "loss": 0.0199, "step": 62810 }, { "grad_norm": 0.25455284118652344, "learning_rate": 1.2088641216604956e-05, "loss": 0.0115, "step": 62820 }, { "grad_norm": 0.29794996976852417, "learning_rate": 1.2075168874314264e-05, "loss": 0.0125, "step": 62830 }, { "grad_norm": 0.14134062826633453, "learning_rate": 1.2061703012339942e-05, "loss": 0.0107, "step": 62840 }, { "grad_norm": 0.28104862570762634, "learning_rate": 1.204824363298297e-05, "loss": 0.0163, "step": 62850 }, { "grad_norm": 0.4100854992866516, "learning_rate": 1.2034790738543173e-05, "loss": 0.0155, "step": 62860 }, { "grad_norm": 0.1860300749540329, "learning_rate": 1.2021344331319284e-05, "loss": 0.0137, "step": 62870 }, { "grad_norm": 0.2336660772562027, "learning_rate": 1.200790441360894e-05, "loss": 0.0194, "step": 62880 }, { "grad_norm": 0.15725238621234894, "learning_rate": 1.199447098770864e-05, "loss": 0.0116, "step": 62890 }, { "grad_norm": 0.24131730198860168, "learning_rate": 1.198104405591381e-05, "loss": 0.0138, "step": 62900 }, { "grad_norm": 0.21501311659812927, "learning_rate": 1.1967623620518697e-05, "loss": 0.0103, "step": 62910 }, { "grad_norm": 0.32718971371650696, "learning_rate": 1.1954209683816554e-05, "loss": 0.013, "step": 62920 }, { "grad_norm": 0.21326328814029694, "learning_rate": 1.1940802248099402e-05, "loss": 0.0111, "step": 62930 }, { "grad_norm": 0.32924461364746094, "learning_rate": 1.1927401315658232e-05, "loss": 0.0155, "step": 62940 }, { "grad_norm": 0.29571810364723206, "learning_rate": 1.1914006888782898e-05, "loss": 0.0158, "step": 62950 }, { "grad_norm": 0.2939685583114624, "learning_rate": 1.1900618969762106e-05, "loss": 0.0143, "step": 62960 }, { "grad_norm": 0.36787763237953186, "learning_rate": 1.188723756088354e-05, "loss": 0.0173, "step": 62970 }, { "grad_norm": 0.22760996222496033, "learning_rate": 1.1873862664433672e-05, "loss": 0.0156, "step": 62980 }, { "grad_norm": 0.2493571937084198, "learning_rate": 1.1860494282697927e-05, "loss": 0.0131, "step": 62990 }, { "grad_norm": 0.35289475321769714, "learning_rate": 1.1847132417960588e-05, "loss": 0.0147, "step": 63000 }, { "grad_norm": 0.1372191458940506, "learning_rate": 1.1833777072504832e-05, "loss": 0.0167, "step": 63010 }, { "grad_norm": 0.20245838165283203, "learning_rate": 1.1820428248612731e-05, "loss": 0.0142, "step": 63020 }, { "grad_norm": 0.4109179377555847, "learning_rate": 1.1807085948565245e-05, "loss": 0.0108, "step": 63030 }, { "grad_norm": 0.15462611615657806, "learning_rate": 1.1793750174642172e-05, "loss": 0.0129, "step": 63040 }, { "grad_norm": 0.1325763761997223, "learning_rate": 1.1780420929122254e-05, "loss": 0.0124, "step": 63050 }, { "grad_norm": 0.2146165370941162, "learning_rate": 1.176709821428309e-05, "loss": 0.0158, "step": 63060 }, { "grad_norm": 0.19396042823791504, "learning_rate": 1.1753782032401173e-05, "loss": 0.0176, "step": 63070 }, { "grad_norm": 0.27348679304122925, "learning_rate": 1.1740472385751866e-05, "loss": 0.0153, "step": 63080 }, { "grad_norm": 0.136832594871521, "learning_rate": 1.172716927660943e-05, "loss": 0.012, "step": 63090 }, { "grad_norm": 0.4090326130390167, "learning_rate": 1.1713872707247015e-05, "loss": 0.0129, "step": 63100 }, { "grad_norm": 0.1489722728729248, "learning_rate": 1.1700582679936595e-05, "loss": 0.0139, "step": 63110 }, { "grad_norm": 0.2636091709136963, "learning_rate": 1.1687299196949136e-05, "loss": 0.0198, "step": 63120 }, { "grad_norm": 0.22030337154865265, "learning_rate": 1.1674022260554374e-05, "loss": 0.0119, "step": 63130 }, { "grad_norm": 0.3159515857696533, "learning_rate": 1.1660751873020987e-05, "loss": 0.0115, "step": 63140 }, { "grad_norm": 0.3973003327846527, "learning_rate": 1.1647488036616538e-05, "loss": 0.015, "step": 63150 }, { "grad_norm": 0.25769171118736267, "learning_rate": 1.1634230753607417e-05, "loss": 0.0169, "step": 63160 }, { "grad_norm": 0.34522756934165955, "learning_rate": 1.1620980026258982e-05, "loss": 0.0135, "step": 63170 }, { "grad_norm": 0.43743717670440674, "learning_rate": 1.1607735856835373e-05, "loss": 0.014, "step": 63180 }, { "grad_norm": 0.18093612790107727, "learning_rate": 1.1594498247599677e-05, "loss": 0.0112, "step": 63190 }, { "grad_norm": 0.19774530827999115, "learning_rate": 1.158126720081384e-05, "loss": 0.0191, "step": 63200 }, { "grad_norm": 0.3951408565044403, "learning_rate": 1.156804271873868e-05, "loss": 0.0129, "step": 63210 }, { "grad_norm": 0.3659495413303375, "learning_rate": 1.1554824803633924e-05, "loss": 0.0148, "step": 63220 }, { "grad_norm": 0.26084569096565247, "learning_rate": 1.154161345775811e-05, "loss": 0.0118, "step": 63230 }, { "grad_norm": 0.17654606699943542, "learning_rate": 1.1528408683368724e-05, "loss": 0.0144, "step": 63240 }, { "grad_norm": 0.4179673492908478, "learning_rate": 1.1515210482722088e-05, "loss": 0.0137, "step": 63250 }, { "grad_norm": 0.2165822982788086, "learning_rate": 1.150201885807342e-05, "loss": 0.0185, "step": 63260 }, { "grad_norm": 0.22556209564208984, "learning_rate": 1.1488833811676807e-05, "loss": 0.0123, "step": 63270 }, { "grad_norm": 0.2621397078037262, "learning_rate": 1.1475655345785213e-05, "loss": 0.0133, "step": 63280 }, { "grad_norm": 0.2473410815000534, "learning_rate": 1.1462483462650481e-05, "loss": 0.0128, "step": 63290 }, { "grad_norm": 0.2628549039363861, "learning_rate": 1.1449318164523331e-05, "loss": 0.0187, "step": 63300 }, { "grad_norm": 0.20024846494197845, "learning_rate": 1.1436159453653334e-05, "loss": 0.009, "step": 63310 }, { "grad_norm": 0.25690197944641113, "learning_rate": 1.1423007332288955e-05, "loss": 0.0141, "step": 63320 }, { "grad_norm": 0.385133296251297, "learning_rate": 1.1409861802677546e-05, "loss": 0.0154, "step": 63330 }, { "grad_norm": 0.33662328124046326, "learning_rate": 1.1396722867065313e-05, "loss": 0.0124, "step": 63340 }, { "grad_norm": 0.198449045419693, "learning_rate": 1.1383590527697352e-05, "loss": 0.0132, "step": 63350 }, { "grad_norm": 0.14235654473304749, "learning_rate": 1.137046478681758e-05, "loss": 0.0098, "step": 63360 }, { "grad_norm": 0.33973193168640137, "learning_rate": 1.1357345646668888e-05, "loss": 0.026, "step": 63370 }, { "grad_norm": 0.34158819913864136, "learning_rate": 1.1344233109492924e-05, "loss": 0.015, "step": 63380 }, { "grad_norm": 0.30527734756469727, "learning_rate": 1.1331127177530292e-05, "loss": 0.0135, "step": 63390 }, { "grad_norm": 0.1825786530971527, "learning_rate": 1.1318027853020441e-05, "loss": 0.014, "step": 63400 }, { "grad_norm": 0.35774168372154236, "learning_rate": 1.1304935138201645e-05, "loss": 0.0135, "step": 63410 }, { "grad_norm": 0.28129833936691284, "learning_rate": 1.1291849035311153e-05, "loss": 0.0128, "step": 63420 }, { "grad_norm": 0.16210485994815826, "learning_rate": 1.1278769546584972e-05, "loss": 0.0098, "step": 63430 }, { "grad_norm": 0.23808553814888, "learning_rate": 1.1265696674258052e-05, "loss": 0.0117, "step": 63440 }, { "grad_norm": 0.21807818114757538, "learning_rate": 1.1252630420564186e-05, "loss": 0.0106, "step": 63450 }, { "grad_norm": 0.3409232199192047, "learning_rate": 1.1239570787736036e-05, "loss": 0.0159, "step": 63460 }, { "grad_norm": 0.28197765350341797, "learning_rate": 1.1226517778005135e-05, "loss": 0.0127, "step": 63470 }, { "grad_norm": 0.17772461473941803, "learning_rate": 1.1213471393601893e-05, "loss": 0.0141, "step": 63480 }, { "grad_norm": 0.486485093832016, "learning_rate": 1.1200431636755587e-05, "loss": 0.0144, "step": 63490 }, { "grad_norm": 0.35544833540916443, "learning_rate": 1.1187398509694336e-05, "loss": 0.0163, "step": 63500 }, { "grad_norm": 0.40531644225120544, "learning_rate": 1.1174372014645146e-05, "loss": 0.0127, "step": 63510 }, { "grad_norm": 0.17245373129844666, "learning_rate": 1.1161352153833899e-05, "loss": 0.0247, "step": 63520 }, { "grad_norm": 0.2526829242706299, "learning_rate": 1.1148338929485325e-05, "loss": 0.0107, "step": 63530 }, { "grad_norm": 0.08507826924324036, "learning_rate": 1.113533234382304e-05, "loss": 0.0117, "step": 63540 }, { "grad_norm": 0.23120462894439697, "learning_rate": 1.1122332399069513e-05, "loss": 0.0169, "step": 63550 }, { "grad_norm": 0.22751814126968384, "learning_rate": 1.1109339097446047e-05, "loss": 0.013, "step": 63560 }, { "grad_norm": 0.33633846044540405, "learning_rate": 1.1096352441172897e-05, "loss": 0.0239, "step": 63570 }, { "grad_norm": 0.23233439028263092, "learning_rate": 1.1083372432469086e-05, "loss": 0.0108, "step": 63580 }, { "grad_norm": 0.33426013588905334, "learning_rate": 1.107039907355255e-05, "loss": 0.0168, "step": 63590 }, { "grad_norm": 0.3404170870780945, "learning_rate": 1.1057432366640103e-05, "loss": 0.0125, "step": 63600 }, { "grad_norm": 0.366403192281723, "learning_rate": 1.1044472313947352e-05, "loss": 0.0174, "step": 63610 }, { "grad_norm": 0.21822386980056763, "learning_rate": 1.1031518917688877e-05, "loss": 0.0116, "step": 63620 }, { "grad_norm": 0.1860574632883072, "learning_rate": 1.1018572180078007e-05, "loss": 0.0228, "step": 63630 }, { "grad_norm": 0.34940651059150696, "learning_rate": 1.1005632103327018e-05, "loss": 0.0153, "step": 63640 }, { "grad_norm": 0.4578801095485687, "learning_rate": 1.0992698689646996e-05, "loss": 0.0149, "step": 63650 }, { "grad_norm": 0.19039735198020935, "learning_rate": 1.0979771941247919e-05, "loss": 0.0169, "step": 63660 }, { "grad_norm": 0.37195295095443726, "learning_rate": 1.0966851860338611e-05, "loss": 0.0123, "step": 63670 }, { "grad_norm": 0.37097829580307007, "learning_rate": 1.0953938449126766e-05, "loss": 0.0155, "step": 63680 }, { "grad_norm": 0.5071254968643188, "learning_rate": 1.0941031709818933e-05, "loss": 0.0236, "step": 63690 }, { "grad_norm": 0.4451442062854767, "learning_rate": 1.0928131644620509e-05, "loss": 0.0245, "step": 63700 }, { "grad_norm": 0.2858929932117462, "learning_rate": 1.0915238255735766e-05, "loss": 0.0127, "step": 63710 }, { "grad_norm": 0.28649091720581055, "learning_rate": 1.0902351545367833e-05, "loss": 0.0143, "step": 63720 }, { "grad_norm": 0.25013014674186707, "learning_rate": 1.0889471515718702e-05, "loss": 0.0114, "step": 63730 }, { "grad_norm": 0.19142045080661774, "learning_rate": 1.087659816898921e-05, "loss": 0.0115, "step": 63740 }, { "grad_norm": 0.2717091739177704, "learning_rate": 1.0863731507379082e-05, "loss": 0.015, "step": 63750 }, { "grad_norm": 0.3289482891559601, "learning_rate": 1.0850871533086827e-05, "loss": 0.0156, "step": 63760 }, { "grad_norm": 0.33493417501449585, "learning_rate": 1.0838018248309927e-05, "loss": 0.016, "step": 63770 }, { "grad_norm": 0.5593939423561096, "learning_rate": 1.0825171655244615e-05, "loss": 0.0211, "step": 63780 }, { "grad_norm": 0.2219119668006897, "learning_rate": 1.0812331756086025e-05, "loss": 0.0139, "step": 63790 }, { "grad_norm": 0.4176770746707916, "learning_rate": 1.079949855302817e-05, "loss": 0.0149, "step": 63800 }, { "grad_norm": 0.1981491595506668, "learning_rate": 1.0786672048263852e-05, "loss": 0.0144, "step": 63810 }, { "grad_norm": 0.1494363397359848, "learning_rate": 1.0773852243984817e-05, "loss": 0.0147, "step": 63820 }, { "grad_norm": 0.244155153632164, "learning_rate": 1.0761039142381586e-05, "loss": 0.0163, "step": 63830 }, { "grad_norm": 0.3021831214427948, "learning_rate": 1.0748232745643577e-05, "loss": 0.0157, "step": 63840 }, { "grad_norm": 0.1671970933675766, "learning_rate": 1.0735433055959055e-05, "loss": 0.0152, "step": 63850 }, { "grad_norm": 0.38274431228637695, "learning_rate": 1.0722640075515133e-05, "loss": 0.0139, "step": 63860 }, { "grad_norm": 0.22246135771274567, "learning_rate": 1.0709853806497795e-05, "loss": 0.0137, "step": 63870 }, { "grad_norm": 0.34251973032951355, "learning_rate": 1.0697074251091831e-05, "loss": 0.015, "step": 63880 }, { "grad_norm": 0.2397487610578537, "learning_rate": 1.0684301411480962e-05, "loss": 0.0117, "step": 63890 }, { "grad_norm": 0.2201499193906784, "learning_rate": 1.067153528984769e-05, "loss": 0.0111, "step": 63900 }, { "grad_norm": 0.20725373923778534, "learning_rate": 1.0658775888373395e-05, "loss": 0.0154, "step": 63910 }, { "grad_norm": 0.17418533563613892, "learning_rate": 1.0646023209238314e-05, "loss": 0.0149, "step": 63920 }, { "grad_norm": 0.38165438175201416, "learning_rate": 1.0633277254621537e-05, "loss": 0.0161, "step": 63930 }, { "grad_norm": 0.23676562309265137, "learning_rate": 1.0620538026700994e-05, "loss": 0.0123, "step": 63940 }, { "grad_norm": 0.20061063766479492, "learning_rate": 1.0607805527653486e-05, "loss": 0.0143, "step": 63950 }, { "grad_norm": 0.2814335823059082, "learning_rate": 1.059507975965462e-05, "loss": 0.0165, "step": 63960 }, { "grad_norm": 0.1917390376329422, "learning_rate": 1.0582360724878898e-05, "loss": 0.011, "step": 63970 }, { "grad_norm": 0.20367828011512756, "learning_rate": 1.0569648425499651e-05, "loss": 0.0153, "step": 63980 }, { "grad_norm": 0.18214309215545654, "learning_rate": 1.0556942863689063e-05, "loss": 0.0137, "step": 63990 }, { "grad_norm": 0.28191256523132324, "learning_rate": 1.054424404161819e-05, "loss": 0.0149, "step": 64000 }, { "grad_norm": 0.17406140267848969, "learning_rate": 1.053155196145686e-05, "loss": 0.0093, "step": 64010 }, { "grad_norm": 0.3046879172325134, "learning_rate": 1.0518866625373863e-05, "loss": 0.013, "step": 64020 }, { "grad_norm": 0.3098256587982178, "learning_rate": 1.0506188035536735e-05, "loss": 0.0105, "step": 64030 }, { "grad_norm": 0.21570219099521637, "learning_rate": 1.0493516194111919e-05, "loss": 0.0171, "step": 64040 }, { "grad_norm": 0.21595539152622223, "learning_rate": 1.0480851103264688e-05, "loss": 0.0117, "step": 64050 }, { "grad_norm": 0.2846289575099945, "learning_rate": 1.046819276515913e-05, "loss": 0.0122, "step": 64060 }, { "grad_norm": 0.2765819728374481, "learning_rate": 1.0455541181958256e-05, "loss": 0.014, "step": 64070 }, { "grad_norm": 0.3043547570705414, "learning_rate": 1.0442896355823822e-05, "loss": 0.012, "step": 64080 }, { "grad_norm": 0.2045176774263382, "learning_rate": 1.0430258288916539e-05, "loss": 0.0138, "step": 64090 }, { "grad_norm": 0.27002957463264465, "learning_rate": 1.0417626983395868e-05, "loss": 0.0133, "step": 64100 }, { "grad_norm": 0.1968483179807663, "learning_rate": 1.0405002441420165e-05, "loss": 0.0117, "step": 64110 }, { "grad_norm": 0.37188342213630676, "learning_rate": 1.039238466514662e-05, "loss": 0.0121, "step": 64120 }, { "grad_norm": 0.14158745110034943, "learning_rate": 1.0379773656731262e-05, "loss": 0.0146, "step": 64130 }, { "grad_norm": 0.3647041618824005, "learning_rate": 1.0367169418328986e-05, "loss": 0.0112, "step": 64140 }, { "grad_norm": 0.3517342507839203, "learning_rate": 1.0354571952093484e-05, "loss": 0.0144, "step": 64150 }, { "grad_norm": 0.22168202698230743, "learning_rate": 1.034198126017733e-05, "loss": 0.0136, "step": 64160 }, { "grad_norm": 0.17693807184696198, "learning_rate": 1.0329397344731928e-05, "loss": 0.0152, "step": 64170 }, { "grad_norm": 0.3602202832698822, "learning_rate": 1.0316820207907524e-05, "loss": 0.0146, "step": 64180 }, { "grad_norm": 0.16176967322826385, "learning_rate": 1.0304249851853215e-05, "loss": 0.0154, "step": 64190 }, { "grad_norm": 0.31993913650512695, "learning_rate": 1.0291686278716933e-05, "loss": 0.0117, "step": 64200 }, { "grad_norm": 0.23940178751945496, "learning_rate": 1.0279129490645418e-05, "loss": 0.0133, "step": 64210 }, { "grad_norm": 0.25862058997154236, "learning_rate": 1.0266579489784328e-05, "loss": 0.0129, "step": 64220 }, { "grad_norm": 0.2998587191104889, "learning_rate": 1.0254036278278084e-05, "loss": 0.0126, "step": 64230 }, { "grad_norm": 0.23589138686656952, "learning_rate": 1.0241499858269982e-05, "loss": 0.0136, "step": 64240 }, { "grad_norm": 0.22535327076911926, "learning_rate": 1.0228970231902169e-05, "loss": 0.013, "step": 64250 }, { "grad_norm": 0.28933632373809814, "learning_rate": 1.0216447401315582e-05, "loss": 0.0146, "step": 64260 }, { "grad_norm": 0.28321573138237, "learning_rate": 1.020393136865007e-05, "loss": 0.0101, "step": 64270 }, { "grad_norm": 0.3231853246688843, "learning_rate": 1.0191422136044242e-05, "loss": 0.0204, "step": 64280 }, { "grad_norm": 0.282867431640625, "learning_rate": 1.017891970563563e-05, "loss": 0.0168, "step": 64290 }, { "grad_norm": 0.23392517864704132, "learning_rate": 1.0166424079560516e-05, "loss": 0.0236, "step": 64300 }, { "grad_norm": 0.32225313782691956, "learning_rate": 1.0153935259954078e-05, "loss": 0.0135, "step": 64310 }, { "grad_norm": 0.4714916944503784, "learning_rate": 1.0141453248950311e-05, "loss": 0.0152, "step": 64320 }, { "grad_norm": 0.2802489995956421, "learning_rate": 1.0128978048682054e-05, "loss": 0.0157, "step": 64330 }, { "grad_norm": 0.323015421628952, "learning_rate": 1.0116509661280982e-05, "loss": 0.0204, "step": 64340 }, { "grad_norm": 0.13975101709365845, "learning_rate": 1.0104048088877576e-05, "loss": 0.0109, "step": 64350 }, { "grad_norm": 0.2225569635629654, "learning_rate": 1.0091593333601201e-05, "loss": 0.0108, "step": 64360 }, { "grad_norm": 0.23344478011131287, "learning_rate": 1.0079145397580031e-05, "loss": 0.0128, "step": 64370 }, { "grad_norm": 0.3660249412059784, "learning_rate": 1.006670428294107e-05, "loss": 0.0194, "step": 64380 }, { "grad_norm": 0.19057856500148773, "learning_rate": 1.0054269991810166e-05, "loss": 0.0121, "step": 64390 }, { "grad_norm": 0.32402315735816956, "learning_rate": 1.0041842526312024e-05, "loss": 0.0148, "step": 64400 }, { "grad_norm": 0.1329936385154724, "learning_rate": 1.0029421888570101e-05, "loss": 0.0097, "step": 64410 }, { "grad_norm": 0.1332169622182846, "learning_rate": 1.0017008080706813e-05, "loss": 0.0226, "step": 64420 }, { "grad_norm": 0.0943998470902443, "learning_rate": 1.0004601104843287e-05, "loss": 0.0124, "step": 64430 }, { "grad_norm": 0.3132537305355072, "learning_rate": 9.992200963099562e-06, "loss": 0.0121, "step": 64440 }, { "grad_norm": 0.19606122374534607, "learning_rate": 9.979807657594486e-06, "loss": 0.0123, "step": 64450 }, { "grad_norm": 0.400771826505661, "learning_rate": 9.967421190445703e-06, "loss": 0.0152, "step": 64460 }, { "grad_norm": 0.2728797495365143, "learning_rate": 9.955041563769769e-06, "loss": 0.0141, "step": 64470 }, { "grad_norm": 0.12747345864772797, "learning_rate": 9.942668779681974e-06, "loss": 0.0158, "step": 64480 }, { "grad_norm": 0.18343253433704376, "learning_rate": 9.930302840296541e-06, "loss": 0.0111, "step": 64490 }, { "grad_norm": 0.19521866738796234, "learning_rate": 9.917943747726426e-06, "loss": 0.0115, "step": 64500 }, { "grad_norm": 0.25793060660362244, "learning_rate": 9.905591504083484e-06, "loss": 0.0129, "step": 64510 }, { "grad_norm": 0.226531982421875, "learning_rate": 9.893246111478382e-06, "loss": 0.0144, "step": 64520 }, { "grad_norm": 0.34205004572868347, "learning_rate": 9.88090757202057e-06, "loss": 0.011, "step": 64530 }, { "grad_norm": 0.3095267117023468, "learning_rate": 9.868575887818421e-06, "loss": 0.0168, "step": 64540 }, { "grad_norm": 0.15056274831295013, "learning_rate": 9.856251060979044e-06, "loss": 0.0122, "step": 64550 }, { "grad_norm": 0.3652210831642151, "learning_rate": 9.843933093608426e-06, "loss": 0.0166, "step": 64560 }, { "grad_norm": 0.3663386106491089, "learning_rate": 9.831621987811368e-06, "loss": 0.0173, "step": 64570 }, { "grad_norm": 0.2140471488237381, "learning_rate": 9.819317745691509e-06, "loss": 0.0124, "step": 64580 }, { "grad_norm": 0.33350706100463867, "learning_rate": 9.8070203693513e-06, "loss": 0.0138, "step": 64590 }, { "grad_norm": 0.37982460856437683, "learning_rate": 9.794729860892048e-06, "loss": 0.0115, "step": 64600 }, { "grad_norm": 0.3190619647502899, "learning_rate": 9.782446222413827e-06, "loss": 0.0229, "step": 64610 }, { "grad_norm": 0.3091850280761719, "learning_rate": 9.770169456015598e-06, "loss": 0.0171, "step": 64620 }, { "grad_norm": 0.1513470858335495, "learning_rate": 9.75789956379512e-06, "loss": 0.0121, "step": 64630 }, { "grad_norm": 0.27993446588516235, "learning_rate": 9.74563654784898e-06, "loss": 0.0129, "step": 64640 }, { "grad_norm": 0.2707008123397827, "learning_rate": 9.733380410272596e-06, "loss": 0.0138, "step": 64650 }, { "grad_norm": 0.31085655093193054, "learning_rate": 9.721131153160207e-06, "loss": 0.0143, "step": 64660 }, { "grad_norm": 0.1592181771993637, "learning_rate": 9.708888778604886e-06, "loss": 0.0146, "step": 64670 }, { "grad_norm": 0.39836621284484863, "learning_rate": 9.696653288698488e-06, "loss": 0.0125, "step": 64680 }, { "grad_norm": 0.15153592824935913, "learning_rate": 9.684424685531763e-06, "loss": 0.0197, "step": 64690 }, { "grad_norm": 0.22373421490192413, "learning_rate": 9.672202971194216e-06, "loss": 0.0185, "step": 64700 }, { "grad_norm": 0.10562105476856232, "learning_rate": 9.659988147774213e-06, "loss": 0.0176, "step": 64710 }, { "grad_norm": 0.1502363383769989, "learning_rate": 9.647780217358942e-06, "loss": 0.0173, "step": 64720 }, { "grad_norm": 0.21795450150966644, "learning_rate": 9.635579182034376e-06, "loss": 0.015, "step": 64730 }, { "grad_norm": 0.18656836450099945, "learning_rate": 9.623385043885386e-06, "loss": 0.0164, "step": 64740 }, { "grad_norm": 0.23335956037044525, "learning_rate": 9.61119780499557e-06, "loss": 0.009, "step": 64750 }, { "grad_norm": 0.3058326542377472, "learning_rate": 9.599017467447418e-06, "loss": 0.0139, "step": 64760 }, { "grad_norm": 0.2650890052318573, "learning_rate": 9.586844033322206e-06, "loss": 0.0172, "step": 64770 }, { "grad_norm": 0.5176159143447876, "learning_rate": 9.574677504700052e-06, "loss": 0.0203, "step": 64780 }, { "grad_norm": 0.24126732349395752, "learning_rate": 9.562517883659877e-06, "loss": 0.0209, "step": 64790 }, { "grad_norm": 0.3139866590499878, "learning_rate": 9.55036517227944e-06, "loss": 0.0142, "step": 64800 }, { "grad_norm": 0.39705008268356323, "learning_rate": 9.538219372635282e-06, "loss": 0.0149, "step": 64810 }, { "grad_norm": 0.20912671089172363, "learning_rate": 9.526080486802802e-06, "loss": 0.0121, "step": 64820 }, { "grad_norm": 0.23539775609970093, "learning_rate": 9.513948516856203e-06, "loss": 0.0154, "step": 64830 }, { "grad_norm": 0.20927751064300537, "learning_rate": 9.501823464868504e-06, "loss": 0.0103, "step": 64840 }, { "grad_norm": 0.25435876846313477, "learning_rate": 9.489705332911547e-06, "loss": 0.0117, "step": 64850 }, { "grad_norm": 0.34154441952705383, "learning_rate": 9.477594123055994e-06, "loss": 0.0162, "step": 64860 }, { "grad_norm": 0.13343746960163116, "learning_rate": 9.465489837371321e-06, "loss": 0.0195, "step": 64870 }, { "grad_norm": 0.19628024101257324, "learning_rate": 9.453392477925794e-06, "loss": 0.0137, "step": 64880 }, { "grad_norm": 0.21548454463481903, "learning_rate": 9.441302046786566e-06, "loss": 0.0153, "step": 64890 }, { "grad_norm": 0.28476840257644653, "learning_rate": 9.429218546019519e-06, "loss": 0.0124, "step": 64900 }, { "grad_norm": 0.36655694246292114, "learning_rate": 9.41714197768941e-06, "loss": 0.0227, "step": 64910 }, { "grad_norm": 0.2888195216655731, "learning_rate": 9.405072343859805e-06, "loss": 0.0135, "step": 64920 }, { "grad_norm": 0.2632458806037903, "learning_rate": 9.393009646593043e-06, "loss": 0.0139, "step": 64930 }, { "grad_norm": 0.2364794909954071, "learning_rate": 9.38095388795035e-06, "loss": 0.0161, "step": 64940 }, { "grad_norm": 0.29625198245048523, "learning_rate": 9.36890506999169e-06, "loss": 0.0139, "step": 64950 }, { "grad_norm": 0.2019534856081009, "learning_rate": 9.356863194775894e-06, "loss": 0.0157, "step": 64960 }, { "grad_norm": 0.28409501910209656, "learning_rate": 9.344828264360583e-06, "loss": 0.0173, "step": 64970 }, { "grad_norm": 0.17645922303199768, "learning_rate": 9.332800280802201e-06, "loss": 0.0173, "step": 64980 }, { "grad_norm": 0.1024688184261322, "learning_rate": 9.32077924615602e-06, "loss": 0.0144, "step": 64990 }, { "grad_norm": 0.24803197383880615, "learning_rate": 9.308765162476063e-06, "loss": 0.0127, "step": 65000 }, { "grad_norm": 0.26733657717704773, "learning_rate": 9.296758031815239e-06, "loss": 0.015, "step": 65010 }, { "grad_norm": 0.33215707540512085, "learning_rate": 9.284757856225229e-06, "loss": 0.0113, "step": 65020 }, { "grad_norm": 0.16248224675655365, "learning_rate": 9.272764637756538e-06, "loss": 0.0102, "step": 65030 }, { "grad_norm": 0.4681997299194336, "learning_rate": 9.260778378458479e-06, "loss": 0.0176, "step": 65040 }, { "grad_norm": 0.1327275037765503, "learning_rate": 9.248799080379172e-06, "loss": 0.0136, "step": 65050 }, { "grad_norm": 0.2931485176086426, "learning_rate": 9.236826745565558e-06, "loss": 0.0148, "step": 65060 }, { "grad_norm": 0.2299109697341919, "learning_rate": 9.224861376063388e-06, "loss": 0.0153, "step": 65070 }, { "grad_norm": 0.2715487778186798, "learning_rate": 9.212902973917192e-06, "loss": 0.0106, "step": 65080 }, { "grad_norm": 0.3769717216491699, "learning_rate": 9.20095154117035e-06, "loss": 0.0101, "step": 65090 }, { "grad_norm": 0.2740950882434845, "learning_rate": 9.189007079865036e-06, "loss": 0.0101, "step": 65100 }, { "grad_norm": 0.316837877035141, "learning_rate": 9.177069592042226e-06, "loss": 0.014, "step": 65110 }, { "grad_norm": 0.20251910388469696, "learning_rate": 9.165139079741724e-06, "loss": 0.0169, "step": 65120 }, { "grad_norm": 0.657052218914032, "learning_rate": 9.153215545002098e-06, "loss": 0.017, "step": 65130 }, { "grad_norm": 0.3674430549144745, "learning_rate": 9.141298989860798e-06, "loss": 0.0192, "step": 65140 }, { "grad_norm": 0.2580872178077698, "learning_rate": 9.129389416353994e-06, "loss": 0.011, "step": 65150 }, { "grad_norm": 0.3458716571331024, "learning_rate": 9.11748682651673e-06, "loss": 0.011, "step": 65160 }, { "grad_norm": 0.28273335099220276, "learning_rate": 9.105591222382837e-06, "loss": 0.0127, "step": 65170 }, { "grad_norm": 0.1267794519662857, "learning_rate": 9.093702605984915e-06, "loss": 0.0146, "step": 65180 }, { "grad_norm": 0.40273329615592957, "learning_rate": 9.081820979354455e-06, "loss": 0.0144, "step": 65190 }, { "grad_norm": 0.21963246166706085, "learning_rate": 9.069946344521663e-06, "loss": 0.0175, "step": 65200 }, { "grad_norm": 0.1697779893875122, "learning_rate": 9.058078703515598e-06, "loss": 0.0141, "step": 65210 }, { "grad_norm": 0.32027485966682434, "learning_rate": 9.046218058364125e-06, "loss": 0.0214, "step": 65220 }, { "grad_norm": 0.2717057168483734, "learning_rate": 9.034364411093893e-06, "loss": 0.0142, "step": 65230 }, { "grad_norm": 0.9356179237365723, "learning_rate": 9.022517763730371e-06, "loss": 0.0152, "step": 65240 }, { "grad_norm": 0.433685839176178, "learning_rate": 9.010678118297827e-06, "loss": 0.0148, "step": 65250 }, { "grad_norm": 0.36409255862236023, "learning_rate": 8.998845476819345e-06, "loss": 0.0143, "step": 65260 }, { "grad_norm": 0.40722987055778503, "learning_rate": 8.987019841316773e-06, "loss": 0.0149, "step": 65270 }, { "grad_norm": 0.3078833818435669, "learning_rate": 8.975201213810802e-06, "loss": 0.0081, "step": 65280 }, { "grad_norm": 0.39294135570526123, "learning_rate": 8.963389596320915e-06, "loss": 0.014, "step": 65290 }, { "grad_norm": 0.2251298725605011, "learning_rate": 8.951584990865391e-06, "loss": 0.0137, "step": 65300 }, { "grad_norm": 0.4154379367828369, "learning_rate": 8.939787399461319e-06, "loss": 0.0171, "step": 65310 }, { "grad_norm": 0.3924334943294525, "learning_rate": 8.927996824124591e-06, "loss": 0.014, "step": 65320 }, { "grad_norm": 0.27299466729164124, "learning_rate": 8.916213266869854e-06, "loss": 0.0181, "step": 65330 }, { "grad_norm": 0.2713710069656372, "learning_rate": 8.904436729710658e-06, "loss": 0.0125, "step": 65340 }, { "grad_norm": 0.24091719090938568, "learning_rate": 8.892667214659245e-06, "loss": 0.0114, "step": 65350 }, { "grad_norm": 0.15992827713489532, "learning_rate": 8.880904723726713e-06, "loss": 0.0135, "step": 65360 }, { "grad_norm": 0.2762642800807953, "learning_rate": 8.869149258922971e-06, "loss": 0.0105, "step": 65370 }, { "grad_norm": 0.24791881442070007, "learning_rate": 8.857400822256662e-06, "loss": 0.0124, "step": 65380 }, { "grad_norm": 0.28649699687957764, "learning_rate": 8.845659415735324e-06, "loss": 0.0167, "step": 65390 }, { "grad_norm": 0.4311709403991699, "learning_rate": 8.83392504136521e-06, "loss": 0.0138, "step": 65400 }, { "grad_norm": 0.12825196981430054, "learning_rate": 8.822197701151407e-06, "loss": 0.0205, "step": 65410 }, { "grad_norm": 0.15353311598300934, "learning_rate": 8.810477397097804e-06, "loss": 0.0141, "step": 65420 }, { "grad_norm": 0.13855499029159546, "learning_rate": 8.798764131207077e-06, "loss": 0.0117, "step": 65430 }, { "grad_norm": 0.262715220451355, "learning_rate": 8.787057905480706e-06, "loss": 0.0159, "step": 65440 }, { "grad_norm": 0.22054921090602875, "learning_rate": 8.775358721918958e-06, "loss": 0.0132, "step": 65450 }, { "grad_norm": 0.3076813817024231, "learning_rate": 8.763666582520923e-06, "loss": 0.0137, "step": 65460 }, { "grad_norm": 0.17551656067371368, "learning_rate": 8.751981489284445e-06, "loss": 0.0119, "step": 65470 }, { "grad_norm": 0.10675327479839325, "learning_rate": 8.740303444206188e-06, "loss": 0.0102, "step": 65480 }, { "grad_norm": 0.19540423154830933, "learning_rate": 8.72863244928162e-06, "loss": 0.0121, "step": 65490 }, { "grad_norm": 0.22368444502353668, "learning_rate": 8.716968506504991e-06, "loss": 0.0189, "step": 65500 }, { "grad_norm": 0.28571203351020813, "learning_rate": 8.70531161786935e-06, "loss": 0.0094, "step": 65510 }, { "grad_norm": 0.15254731476306915, "learning_rate": 8.693661785366558e-06, "loss": 0.019, "step": 65520 }, { "grad_norm": 0.31780537962913513, "learning_rate": 8.682019010987208e-06, "loss": 0.0184, "step": 65530 }, { "grad_norm": 0.4863680303096771, "learning_rate": 8.670383296720786e-06, "loss": 0.0116, "step": 65540 }, { "grad_norm": 0.22294512391090393, "learning_rate": 8.658754644555478e-06, "loss": 0.0131, "step": 65550 }, { "grad_norm": 0.28470584750175476, "learning_rate": 8.647133056478313e-06, "loss": 0.0134, "step": 65560 }, { "grad_norm": 0.25970056653022766, "learning_rate": 8.635518534475123e-06, "loss": 0.0134, "step": 65570 }, { "grad_norm": 0.3653990924358368, "learning_rate": 8.623911080530467e-06, "loss": 0.0153, "step": 65580 }, { "grad_norm": 0.10810814797878265, "learning_rate": 8.61231069662779e-06, "loss": 0.014, "step": 65590 }, { "grad_norm": 0.336606502532959, "learning_rate": 8.600717384749252e-06, "loss": 0.011, "step": 65600 }, { "grad_norm": 0.3320348262786865, "learning_rate": 8.58913114687584e-06, "loss": 0.0112, "step": 65610 }, { "grad_norm": 0.3026059865951538, "learning_rate": 8.57755198498732e-06, "loss": 0.0157, "step": 65620 }, { "grad_norm": 0.29758328199386597, "learning_rate": 8.565979901062265e-06, "loss": 0.0185, "step": 65630 }, { "grad_norm": 0.1808280348777771, "learning_rate": 8.554414897078033e-06, "loss": 0.0145, "step": 65640 }, { "grad_norm": 0.17872266471385956, "learning_rate": 8.542856975010727e-06, "loss": 0.0113, "step": 65650 }, { "grad_norm": 0.19550679624080658, "learning_rate": 8.531306136835337e-06, "loss": 0.011, "step": 65660 }, { "grad_norm": 0.20942479372024536, "learning_rate": 8.519762384525548e-06, "loss": 0.012, "step": 65670 }, { "grad_norm": 0.24364826083183289, "learning_rate": 8.508225720053875e-06, "loss": 0.0111, "step": 65680 }, { "grad_norm": 0.20918786525726318, "learning_rate": 8.496696145391625e-06, "loss": 0.0099, "step": 65690 }, { "grad_norm": 0.233210489153862, "learning_rate": 8.485173662508889e-06, "loss": 0.0124, "step": 65700 }, { "grad_norm": 0.21320101618766785, "learning_rate": 8.473658273374536e-06, "loss": 0.0102, "step": 65710 }, { "grad_norm": 0.17728616297245026, "learning_rate": 8.462149979956253e-06, "loss": 0.009, "step": 65720 }, { "grad_norm": 0.2973248064517975, "learning_rate": 8.450648784220461e-06, "loss": 0.0118, "step": 65730 }, { "grad_norm": 0.43227556347846985, "learning_rate": 8.439154688132417e-06, "loss": 0.0137, "step": 65740 }, { "grad_norm": 0.33665916323661804, "learning_rate": 8.427667693656143e-06, "loss": 0.0156, "step": 65750 }, { "grad_norm": 0.2849646508693695, "learning_rate": 8.416187802754454e-06, "loss": 0.0128, "step": 65760 }, { "grad_norm": 0.1353636085987091, "learning_rate": 8.404715017388965e-06, "loss": 0.0099, "step": 65770 }, { "grad_norm": 0.23047171533107758, "learning_rate": 8.393249339520015e-06, "loss": 0.0123, "step": 65780 }, { "grad_norm": 0.27014732360839844, "learning_rate": 8.381790771106834e-06, "loss": 0.0106, "step": 65790 }, { "grad_norm": 0.35312142968177795, "learning_rate": 8.370339314107339e-06, "loss": 0.0123, "step": 65800 }, { "grad_norm": 0.4035831391811371, "learning_rate": 8.358894970478281e-06, "loss": 0.013, "step": 65810 }, { "grad_norm": 0.23079337179660797, "learning_rate": 8.347457742175196e-06, "loss": 0.0151, "step": 65820 }, { "grad_norm": 0.3217587172985077, "learning_rate": 8.33602763115236e-06, "loss": 0.0095, "step": 65830 }, { "grad_norm": 0.28221607208251953, "learning_rate": 8.324604639362916e-06, "loss": 0.0212, "step": 65840 }, { "grad_norm": 0.13751941919326782, "learning_rate": 8.31318876875869e-06, "loss": 0.01, "step": 65850 }, { "grad_norm": 0.2913569509983063, "learning_rate": 8.30178002129039e-06, "loss": 0.0089, "step": 65860 }, { "grad_norm": 0.268257737159729, "learning_rate": 8.290378398907423e-06, "loss": 0.0133, "step": 65870 }, { "grad_norm": 0.23078984022140503, "learning_rate": 8.278983903558029e-06, "loss": 0.0175, "step": 65880 }, { "grad_norm": 0.796183168888092, "learning_rate": 8.26759653718921e-06, "loss": 0.0134, "step": 65890 }, { "grad_norm": 0.19025655090808868, "learning_rate": 8.25621630174676e-06, "loss": 0.0159, "step": 65900 }, { "grad_norm": 0.47061678767204285, "learning_rate": 8.244843199175261e-06, "loss": 0.0199, "step": 65910 }, { "grad_norm": 0.19378311932086945, "learning_rate": 8.233477231418046e-06, "loss": 0.0107, "step": 65920 }, { "grad_norm": 0.17628636956214905, "learning_rate": 8.22211840041725e-06, "loss": 0.0135, "step": 65930 }, { "grad_norm": 0.3581065237522125, "learning_rate": 8.210766708113792e-06, "loss": 0.016, "step": 65940 }, { "grad_norm": 0.27662861347198486, "learning_rate": 8.199422156447367e-06, "loss": 0.0148, "step": 65950 }, { "grad_norm": 0.24271154403686523, "learning_rate": 8.188084747356451e-06, "loss": 0.0176, "step": 65960 }, { "grad_norm": 0.28994154930114746, "learning_rate": 8.176754482778299e-06, "loss": 0.0113, "step": 65970 }, { "grad_norm": 0.35915234684944153, "learning_rate": 8.165431364648918e-06, "loss": 0.0162, "step": 65980 }, { "grad_norm": 0.2973582446575165, "learning_rate": 8.154115394903162e-06, "loss": 0.0125, "step": 65990 }, { "grad_norm": 0.224399134516716, "learning_rate": 8.142806575474582e-06, "loss": 0.0119, "step": 66000 }, { "grad_norm": 0.17424066364765167, "learning_rate": 8.131504908295562e-06, "loss": 0.0142, "step": 66010 }, { "grad_norm": 0.24739603698253632, "learning_rate": 8.120210395297262e-06, "loss": 0.012, "step": 66020 }, { "grad_norm": 0.1758357137441635, "learning_rate": 8.108923038409565e-06, "loss": 0.0151, "step": 66030 }, { "grad_norm": 0.3229387104511261, "learning_rate": 8.097642839561226e-06, "loss": 0.0149, "step": 66040 }, { "grad_norm": 0.30236297845840454, "learning_rate": 8.08636980067966e-06, "loss": 0.0168, "step": 66050 }, { "grad_norm": 0.2334919273853302, "learning_rate": 8.075103923691186e-06, "loss": 0.0088, "step": 66060 }, { "grad_norm": 0.20339399576187134, "learning_rate": 8.063845210520793e-06, "loss": 0.0096, "step": 66070 }, { "grad_norm": 0.2942443788051605, "learning_rate": 8.052593663092295e-06, "loss": 0.014, "step": 66080 }, { "grad_norm": 0.1834854930639267, "learning_rate": 8.04134928332827e-06, "loss": 0.0119, "step": 66090 }, { "grad_norm": 0.23427152633666992, "learning_rate": 8.030112073150086e-06, "loss": 0.0201, "step": 66100 }, { "grad_norm": 0.16233055293560028, "learning_rate": 8.018882034477881e-06, "loss": 0.0076, "step": 66110 }, { "grad_norm": 0.22117899358272552, "learning_rate": 8.007659169230541e-06, "loss": 0.0094, "step": 66120 }, { "grad_norm": 0.20232322812080383, "learning_rate": 7.996443479325755e-06, "loss": 0.0113, "step": 66130 }, { "grad_norm": 0.21748995780944824, "learning_rate": 7.985234966679977e-06, "loss": 0.019, "step": 66140 }, { "grad_norm": 0.18486377596855164, "learning_rate": 7.974033633208438e-06, "loss": 0.0178, "step": 66150 }, { "grad_norm": 0.39896276593208313, "learning_rate": 7.962839480825135e-06, "loss": 0.0134, "step": 66160 }, { "grad_norm": 0.4388880431652069, "learning_rate": 7.951652511442858e-06, "loss": 0.0161, "step": 66170 }, { "grad_norm": 0.3303360342979431, "learning_rate": 7.940472726973125e-06, "loss": 0.0183, "step": 66180 }, { "grad_norm": 0.21735168993473053, "learning_rate": 7.929300129326289e-06, "loss": 0.0133, "step": 66190 }, { "grad_norm": 0.19692423939704895, "learning_rate": 7.91813472041142e-06, "loss": 0.0119, "step": 66200 }, { "grad_norm": 0.34564611315727234, "learning_rate": 7.906976502136376e-06, "loss": 0.0128, "step": 66210 }, { "grad_norm": 0.2944095730781555, "learning_rate": 7.89582547640782e-06, "loss": 0.0117, "step": 66220 }, { "grad_norm": 0.3981044590473175, "learning_rate": 7.884681645131115e-06, "loss": 0.0117, "step": 66230 }, { "grad_norm": 0.36609145998954773, "learning_rate": 7.87354501021048e-06, "loss": 0.0126, "step": 66240 }, { "grad_norm": 0.2314472794532776, "learning_rate": 7.86241557354882e-06, "loss": 0.0157, "step": 66250 }, { "grad_norm": 0.25930696725845337, "learning_rate": 7.8512933370479e-06, "loss": 0.0131, "step": 66260 }, { "grad_norm": 0.20366668701171875, "learning_rate": 7.840178302608158e-06, "loss": 0.0138, "step": 66270 }, { "grad_norm": 0.18715675175189972, "learning_rate": 7.82907047212888e-06, "loss": 0.0155, "step": 66280 }, { "grad_norm": 0.24822601675987244, "learning_rate": 7.81796984750809e-06, "loss": 0.0119, "step": 66290 }, { "grad_norm": 0.21140339970588684, "learning_rate": 7.806876430642546e-06, "loss": 0.0138, "step": 66300 }, { "grad_norm": 0.3435109257698059, "learning_rate": 7.795790223427862e-06, "loss": 0.0141, "step": 66310 }, { "grad_norm": 0.38205602765083313, "learning_rate": 7.784711227758324e-06, "loss": 0.014, "step": 66320 }, { "grad_norm": 0.16402851045131683, "learning_rate": 7.773639445527053e-06, "loss": 0.0135, "step": 66330 }, { "grad_norm": 0.12619735300540924, "learning_rate": 7.762574878625905e-06, "loss": 0.014, "step": 66340 }, { "grad_norm": 0.30363377928733826, "learning_rate": 7.751517528945513e-06, "loss": 0.0107, "step": 66350 }, { "grad_norm": 0.3772883415222168, "learning_rate": 7.740467398375278e-06, "loss": 0.0132, "step": 66360 }, { "grad_norm": 0.22363480925559998, "learning_rate": 7.729424488803378e-06, "loss": 0.0099, "step": 66370 }, { "grad_norm": 0.19470706582069397, "learning_rate": 7.71838880211671e-06, "loss": 0.0193, "step": 66380 }, { "grad_norm": 0.4010833501815796, "learning_rate": 7.707360340200997e-06, "loss": 0.012, "step": 66390 }, { "grad_norm": 0.20651985704898834, "learning_rate": 7.696339104940697e-06, "loss": 0.0118, "step": 66400 }, { "grad_norm": 0.3592011630535126, "learning_rate": 7.685325098219038e-06, "loss": 0.0152, "step": 66410 }, { "grad_norm": 0.13563048839569092, "learning_rate": 7.674318321918017e-06, "loss": 0.0145, "step": 66420 }, { "grad_norm": 0.4177559018135071, "learning_rate": 7.663318777918366e-06, "loss": 0.0185, "step": 66430 }, { "grad_norm": 0.39404815435409546, "learning_rate": 7.652326468099647e-06, "loss": 0.0167, "step": 66440 }, { "grad_norm": 0.3741471767425537, "learning_rate": 7.641341394340096e-06, "loss": 0.0132, "step": 66450 }, { "grad_norm": 0.3678719103336334, "learning_rate": 7.630363558516818e-06, "loss": 0.0115, "step": 66460 }, { "grad_norm": 0.44025224447250366, "learning_rate": 7.619392962505578e-06, "loss": 0.017, "step": 66470 }, { "grad_norm": 0.19001540541648865, "learning_rate": 7.6084296081809725e-06, "loss": 0.0183, "step": 66480 }, { "grad_norm": 0.2854800224304199, "learning_rate": 7.597473497416347e-06, "loss": 0.017, "step": 66490 }, { "grad_norm": 0.28473350405693054, "learning_rate": 7.586524632083764e-06, "loss": 0.0122, "step": 66500 }, { "grad_norm": 0.18600550293922424, "learning_rate": 7.5755830140541326e-06, "loss": 0.0144, "step": 66510 }, { "grad_norm": 0.3136345148086548, "learning_rate": 7.5646486451970475e-06, "loss": 0.013, "step": 66520 }, { "grad_norm": 0.4956844747066498, "learning_rate": 7.553721527380897e-06, "loss": 0.012, "step": 66530 }, { "grad_norm": 0.2105613499879837, "learning_rate": 7.542801662472826e-06, "loss": 0.0123, "step": 66540 }, { "grad_norm": 0.19836115837097168, "learning_rate": 7.5318890523387475e-06, "loss": 0.015, "step": 66550 }, { "grad_norm": 0.20301035046577454, "learning_rate": 7.52098369884332e-06, "loss": 0.0158, "step": 66560 }, { "grad_norm": 0.16133342683315277, "learning_rate": 7.510085603849992e-06, "loss": 0.0115, "step": 66570 }, { "grad_norm": 0.41970735788345337, "learning_rate": 7.499194769220918e-06, "loss": 0.0136, "step": 66580 }, { "grad_norm": 0.47191426157951355, "learning_rate": 7.48831119681706e-06, "loss": 0.0181, "step": 66590 }, { "grad_norm": 0.4243001937866211, "learning_rate": 7.477434888498119e-06, "loss": 0.0094, "step": 66600 }, { "grad_norm": 0.22877535223960876, "learning_rate": 7.466565846122564e-06, "loss": 0.0234, "step": 66610 }, { "grad_norm": 0.28830841183662415, "learning_rate": 7.455704071547626e-06, "loss": 0.0129, "step": 66620 }, { "grad_norm": 0.1496066451072693, "learning_rate": 7.444849566629247e-06, "loss": 0.0131, "step": 66630 }, { "grad_norm": 0.4334636926651001, "learning_rate": 7.434002333222212e-06, "loss": 0.0113, "step": 66640 }, { "grad_norm": 0.2638738751411438, "learning_rate": 7.423162373179976e-06, "loss": 0.0157, "step": 66650 }, { "grad_norm": 0.300481379032135, "learning_rate": 7.412329688354835e-06, "loss": 0.0133, "step": 66660 }, { "grad_norm": 0.221902996301651, "learning_rate": 7.40150428059776e-06, "loss": 0.0116, "step": 66670 }, { "grad_norm": 0.3733210861682892, "learning_rate": 7.3906861517585354e-06, "loss": 0.0176, "step": 66680 }, { "grad_norm": 0.3214443027973175, "learning_rate": 7.37987530368569e-06, "loss": 0.0152, "step": 66690 }, { "grad_norm": 0.3145394027233124, "learning_rate": 7.369071738226474e-06, "loss": 0.0128, "step": 66700 }, { "grad_norm": 0.3126503527164459, "learning_rate": 7.358275457226954e-06, "loss": 0.0149, "step": 66710 }, { "grad_norm": 0.27388373017311096, "learning_rate": 7.347486462531899e-06, "loss": 0.0115, "step": 66720 }, { "grad_norm": 0.12595954537391663, "learning_rate": 7.336704755984858e-06, "loss": 0.0082, "step": 66730 }, { "grad_norm": 0.20012183487415314, "learning_rate": 7.325930339428133e-06, "loss": 0.0148, "step": 66740 }, { "grad_norm": 0.8040987253189087, "learning_rate": 7.315163214702769e-06, "loss": 0.0103, "step": 66750 }, { "grad_norm": 0.21604737639427185, "learning_rate": 7.304403383648595e-06, "loss": 0.0107, "step": 66760 }, { "grad_norm": 0.1255723237991333, "learning_rate": 7.293650848104139e-06, "loss": 0.0148, "step": 66770 }, { "grad_norm": 0.21043047308921814, "learning_rate": 7.2829056099067374e-06, "loss": 0.0142, "step": 66780 }, { "grad_norm": 0.20279061794281006, "learning_rate": 7.2721676708924494e-06, "loss": 0.0135, "step": 66790 }, { "grad_norm": 0.37335580587387085, "learning_rate": 7.261437032896096e-06, "loss": 0.0211, "step": 66800 }, { "grad_norm": 0.2256942242383957, "learning_rate": 7.250713697751255e-06, "loss": 0.0133, "step": 66810 }, { "grad_norm": 0.18298718333244324, "learning_rate": 7.239997667290255e-06, "loss": 0.0087, "step": 66820 }, { "grad_norm": 0.2150638997554779, "learning_rate": 7.2292889433441425e-06, "loss": 0.0123, "step": 66830 }, { "grad_norm": 0.2272353172302246, "learning_rate": 7.218587527742793e-06, "loss": 0.0105, "step": 66840 }, { "grad_norm": 0.24768702685832977, "learning_rate": 7.207893422314749e-06, "loss": 0.0112, "step": 66850 }, { "grad_norm": 0.3590388298034668, "learning_rate": 7.1972066288873545e-06, "loss": 0.0107, "step": 66860 }, { "grad_norm": 0.3539550006389618, "learning_rate": 7.186527149286687e-06, "loss": 0.0225, "step": 66870 }, { "grad_norm": 0.3352464735507965, "learning_rate": 7.175854985337576e-06, "loss": 0.0131, "step": 66880 }, { "grad_norm": 0.1904878169298172, "learning_rate": 7.16519013886362e-06, "loss": 0.0156, "step": 66890 }, { "grad_norm": 0.2868805229663849, "learning_rate": 7.154532611687109e-06, "loss": 0.0102, "step": 66900 }, { "grad_norm": 0.16346393525600433, "learning_rate": 7.143882405629177e-06, "loss": 0.0125, "step": 66910 }, { "grad_norm": 0.2107207328081131, "learning_rate": 7.133239522509605e-06, "loss": 0.0108, "step": 66920 }, { "grad_norm": 0.1922309845685959, "learning_rate": 7.1226039641469956e-06, "loss": 0.0118, "step": 66930 }, { "grad_norm": 0.24645690619945526, "learning_rate": 7.111975732358678e-06, "loss": 0.0116, "step": 66940 }, { "grad_norm": 0.2357638031244278, "learning_rate": 7.101354828960693e-06, "loss": 0.0116, "step": 66950 }, { "grad_norm": 0.5068174600601196, "learning_rate": 7.090741255767918e-06, "loss": 0.0119, "step": 66960 }, { "grad_norm": 0.22469303011894226, "learning_rate": 7.080135014593875e-06, "loss": 0.0126, "step": 66970 }, { "grad_norm": 0.19879385828971863, "learning_rate": 7.069536107250896e-06, "loss": 0.0148, "step": 66980 }, { "grad_norm": 0.07633872330188751, "learning_rate": 7.058944535550049e-06, "loss": 0.0108, "step": 66990 }, { "grad_norm": 0.33974477648735046, "learning_rate": 7.048360301301138e-06, "loss": 0.0121, "step": 67000 }, { "grad_norm": 0.2588740885257721, "learning_rate": 7.03778340631272e-06, "loss": 0.0136, "step": 67010 }, { "grad_norm": 0.2289605587720871, "learning_rate": 7.0272138523921e-06, "loss": 0.0135, "step": 67020 }, { "grad_norm": 0.17358270287513733, "learning_rate": 7.016651641345334e-06, "loss": 0.0134, "step": 67030 }, { "grad_norm": 0.28098928928375244, "learning_rate": 7.0060967749772e-06, "loss": 0.01, "step": 67040 }, { "grad_norm": 0.5653812885284424, "learning_rate": 6.995549255091238e-06, "loss": 0.0212, "step": 67050 }, { "grad_norm": 0.2309621125459671, "learning_rate": 6.98500908348973e-06, "loss": 0.0169, "step": 67060 }, { "grad_norm": 0.6266295313835144, "learning_rate": 6.974476261973711e-06, "loss": 0.0176, "step": 67070 }, { "grad_norm": 0.1650133728981018, "learning_rate": 6.963950792342949e-06, "loss": 0.0091, "step": 67080 }, { "grad_norm": 0.1769980490207672, "learning_rate": 6.9534326763959715e-06, "loss": 0.011, "step": 67090 }, { "grad_norm": 0.29123854637145996, "learning_rate": 6.942921915929995e-06, "loss": 0.0157, "step": 67100 }, { "grad_norm": 0.2217540442943573, "learning_rate": 6.9324185127410735e-06, "loss": 0.0106, "step": 67110 }, { "grad_norm": 0.2383781224489212, "learning_rate": 6.921922468623921e-06, "loss": 0.0115, "step": 67120 }, { "grad_norm": 0.18019042909145355, "learning_rate": 6.911433785372023e-06, "loss": 0.0125, "step": 67130 }, { "grad_norm": 0.27281519770622253, "learning_rate": 6.900952464777632e-06, "loss": 0.0139, "step": 67140 }, { "grad_norm": 0.29175758361816406, "learning_rate": 6.8904785086316815e-06, "loss": 0.015, "step": 67150 }, { "grad_norm": 0.2735166847705841, "learning_rate": 6.880011918723927e-06, "loss": 0.0154, "step": 67160 }, { "grad_norm": 0.30642396211624146, "learning_rate": 6.86955269684279e-06, "loss": 0.01, "step": 67170 }, { "grad_norm": 0.26318737864494324, "learning_rate": 6.859100844775473e-06, "loss": 0.0188, "step": 67180 }, { "grad_norm": 0.20358966290950775, "learning_rate": 6.84865636430792e-06, "loss": 0.0137, "step": 67190 }, { "grad_norm": 0.1803887039422989, "learning_rate": 6.838219257224804e-06, "loss": 0.0096, "step": 67200 }, { "grad_norm": 0.11211074888706207, "learning_rate": 6.827789525309536e-06, "loss": 0.0154, "step": 67210 }, { "grad_norm": 0.2872600555419922, "learning_rate": 6.81736717034428e-06, "loss": 0.016, "step": 67220 }, { "grad_norm": 0.13691939413547516, "learning_rate": 6.806952194109933e-06, "loss": 0.0199, "step": 67230 }, { "grad_norm": 0.3712601661682129, "learning_rate": 6.796544598386112e-06, "loss": 0.008, "step": 67240 }, { "grad_norm": 0.2832328975200653, "learning_rate": 6.786144384951204e-06, "loss": 0.0132, "step": 67250 }, { "grad_norm": 0.2888994812965393, "learning_rate": 6.775751555582322e-06, "loss": 0.0111, "step": 67260 }, { "grad_norm": 0.25305554270744324, "learning_rate": 6.76536611205531e-06, "loss": 0.0122, "step": 67270 }, { "grad_norm": 0.16377785801887512, "learning_rate": 6.754988056144762e-06, "loss": 0.0154, "step": 67280 }, { "grad_norm": 0.34264951944351196, "learning_rate": 6.744617389624014e-06, "loss": 0.0153, "step": 67290 }, { "grad_norm": 0.3741759657859802, "learning_rate": 6.734254114265087e-06, "loss": 0.0135, "step": 67300 }, { "grad_norm": 0.34447094798088074, "learning_rate": 6.723898231838843e-06, "loss": 0.0103, "step": 67310 }, { "grad_norm": 0.3946772515773773, "learning_rate": 6.713549744114766e-06, "loss": 0.0162, "step": 67320 }, { "grad_norm": 0.27601444721221924, "learning_rate": 6.703208652861159e-06, "loss": 0.012, "step": 67330 }, { "grad_norm": 0.4348648488521576, "learning_rate": 6.69287495984503e-06, "loss": 0.015, "step": 67340 }, { "grad_norm": 0.3342808783054352, "learning_rate": 6.6825486668321e-06, "loss": 0.0157, "step": 67350 }, { "grad_norm": 0.17796237766742706, "learning_rate": 6.672229775586886e-06, "loss": 0.013, "step": 67360 }, { "grad_norm": 0.24744229018688202, "learning_rate": 6.661918287872576e-06, "loss": 0.0112, "step": 67370 }, { "grad_norm": 0.5493019819259644, "learning_rate": 6.6516142054511346e-06, "loss": 0.0144, "step": 67380 }, { "grad_norm": 0.2096758782863617, "learning_rate": 6.641317530083241e-06, "loss": 0.0105, "step": 67390 }, { "grad_norm": 0.18538010120391846, "learning_rate": 6.631028263528322e-06, "loss": 0.0153, "step": 67400 }, { "grad_norm": 0.25042957067489624, "learning_rate": 6.620746407544537e-06, "loss": 0.0127, "step": 67410 }, { "grad_norm": 0.2170678973197937, "learning_rate": 6.610471963888742e-06, "loss": 0.0111, "step": 67420 }, { "grad_norm": 0.18757185339927673, "learning_rate": 6.600204934316606e-06, "loss": 0.0113, "step": 67430 }, { "grad_norm": 0.343441903591156, "learning_rate": 6.589945320582452e-06, "loss": 0.012, "step": 67440 }, { "grad_norm": 0.24308176338672638, "learning_rate": 6.579693124439374e-06, "loss": 0.013, "step": 67450 }, { "grad_norm": 0.2225215882062912, "learning_rate": 6.56944834763919e-06, "loss": 0.0098, "step": 67460 }, { "grad_norm": 0.23041771352291107, "learning_rate": 6.5592109919324575e-06, "loss": 0.0115, "step": 67470 }, { "grad_norm": 0.4074367880821228, "learning_rate": 6.54898105906846e-06, "loss": 0.0128, "step": 67480 }, { "grad_norm": 0.15510740876197815, "learning_rate": 6.5387585507952155e-06, "loss": 0.0197, "step": 67490 }, { "grad_norm": 0.3557296395301819, "learning_rate": 6.528543468859461e-06, "loss": 0.0125, "step": 67500 }, { "grad_norm": 0.3323947489261627, "learning_rate": 6.518335815006682e-06, "loss": 0.0176, "step": 67510 }, { "grad_norm": 0.2723732888698578, "learning_rate": 6.5081355909810845e-06, "loss": 0.014, "step": 67520 }, { "grad_norm": 0.23040036857128143, "learning_rate": 6.497942798525608e-06, "loss": 0.018, "step": 67530 }, { "grad_norm": 0.5270251631736755, "learning_rate": 6.487757439381936e-06, "loss": 0.0184, "step": 67540 }, { "grad_norm": 0.17339947819709778, "learning_rate": 6.477579515290433e-06, "loss": 0.0169, "step": 67550 }, { "grad_norm": 0.20975114405155182, "learning_rate": 6.467409027990273e-06, "loss": 0.016, "step": 67560 }, { "grad_norm": 0.14131678640842438, "learning_rate": 6.457245979219279e-06, "loss": 0.0199, "step": 67570 }, { "grad_norm": 0.3005703389644623, "learning_rate": 6.447090370714054e-06, "loss": 0.0199, "step": 67580 }, { "grad_norm": 0.12671193480491638, "learning_rate": 6.436942204209917e-06, "loss": 0.0164, "step": 67590 }, { "grad_norm": 0.24395684897899628, "learning_rate": 6.4268014814408804e-06, "loss": 0.0143, "step": 67600 }, { "grad_norm": 0.20972727239131927, "learning_rate": 6.41666820413977e-06, "loss": 0.0119, "step": 67610 }, { "grad_norm": 0.3170303404331207, "learning_rate": 6.406542374038033e-06, "loss": 0.017, "step": 67620 }, { "grad_norm": 0.23870795965194702, "learning_rate": 6.396423992865935e-06, "loss": 0.0094, "step": 67630 }, { "grad_norm": 0.2048799991607666, "learning_rate": 6.386313062352412e-06, "loss": 0.0133, "step": 67640 }, { "grad_norm": 0.29966530203819275, "learning_rate": 6.376209584225152e-06, "loss": 0.0096, "step": 67650 }, { "grad_norm": 0.5828222632408142, "learning_rate": 6.366113560210557e-06, "loss": 0.0113, "step": 67660 }, { "grad_norm": 0.43119341135025024, "learning_rate": 6.356024992033766e-06, "loss": 0.0133, "step": 67670 }, { "grad_norm": 0.3339326083660126, "learning_rate": 6.345943881418648e-06, "loss": 0.0111, "step": 67680 }, { "grad_norm": 0.22070389986038208, "learning_rate": 6.33587023008777e-06, "loss": 0.0106, "step": 67690 }, { "grad_norm": 0.14854615926742554, "learning_rate": 6.32580403976245e-06, "loss": 0.0098, "step": 67700 }, { "grad_norm": 0.32067060470581055, "learning_rate": 6.31574531216273e-06, "loss": 0.015, "step": 67710 }, { "grad_norm": 0.2558537423610687, "learning_rate": 6.305694049007371e-06, "loss": 0.0153, "step": 67720 }, { "grad_norm": 0.31722429394721985, "learning_rate": 6.2956502520138575e-06, "loss": 0.0127, "step": 67730 }, { "grad_norm": 0.262207567691803, "learning_rate": 6.285613922898409e-06, "loss": 0.0103, "step": 67740 }, { "grad_norm": 0.20226530730724335, "learning_rate": 6.275585063375927e-06, "loss": 0.0173, "step": 67750 }, { "grad_norm": 0.08118193596601486, "learning_rate": 6.265563675160113e-06, "loss": 0.0134, "step": 67760 }, { "grad_norm": 0.1526537537574768, "learning_rate": 6.25554975996332e-06, "loss": 0.0116, "step": 67770 }, { "grad_norm": 0.2947080731391907, "learning_rate": 6.245543319496661e-06, "loss": 0.0105, "step": 67780 }, { "grad_norm": 0.22244849801063538, "learning_rate": 6.2355443554699685e-06, "loss": 0.0148, "step": 67790 }, { "grad_norm": 0.36756470799446106, "learning_rate": 6.225552869591766e-06, "loss": 0.0111, "step": 67800 }, { "grad_norm": 0.19974854588508606, "learning_rate": 6.215568863569365e-06, "loss": 0.0126, "step": 67810 }, { "grad_norm": 0.391589492559433, "learning_rate": 6.205592339108712e-06, "loss": 0.0188, "step": 67820 }, { "grad_norm": 0.36687982082366943, "learning_rate": 6.195623297914577e-06, "loss": 0.0105, "step": 67830 }, { "grad_norm": 0.18379102647304535, "learning_rate": 6.185661741690357e-06, "loss": 0.0136, "step": 67840 }, { "grad_norm": 0.31324297189712524, "learning_rate": 6.1757076721382145e-06, "loss": 0.0174, "step": 67850 }, { "grad_norm": 0.21470579504966736, "learning_rate": 6.165761090959038e-06, "loss": 0.0163, "step": 67860 }, { "grad_norm": 0.32500508427619934, "learning_rate": 6.155821999852424e-06, "loss": 0.0118, "step": 67870 }, { "grad_norm": 0.16467075049877167, "learning_rate": 6.145890400516696e-06, "loss": 0.0127, "step": 67880 }, { "grad_norm": 0.2545892298221588, "learning_rate": 6.1359662946488816e-06, "loss": 0.0137, "step": 67890 }, { "grad_norm": 0.411687433719635, "learning_rate": 6.12604968394474e-06, "loss": 0.0105, "step": 67900 }, { "grad_norm": 0.2803582549095154, "learning_rate": 6.11614057009875e-06, "loss": 0.014, "step": 67910 }, { "grad_norm": 0.26560747623443604, "learning_rate": 6.106238954804111e-06, "loss": 0.0111, "step": 67920 }, { "grad_norm": 0.2503238618373871, "learning_rate": 6.096344839752738e-06, "loss": 0.0156, "step": 67930 }, { "grad_norm": 0.18568788468837738, "learning_rate": 6.086458226635278e-06, "loss": 0.0119, "step": 67940 }, { "grad_norm": 0.171750009059906, "learning_rate": 6.076579117141046e-06, "loss": 0.015, "step": 67950 }, { "grad_norm": 0.25367528200149536, "learning_rate": 6.066707512958153e-06, "loss": 0.0095, "step": 67960 }, { "grad_norm": 0.24439099431037903, "learning_rate": 6.056843415773361e-06, "loss": 0.0175, "step": 67970 }, { "grad_norm": 0.34280556440353394, "learning_rate": 6.0469868272721776e-06, "loss": 0.0165, "step": 67980 }, { "grad_norm": 0.5742165446281433, "learning_rate": 6.037137749138844e-06, "loss": 0.0159, "step": 67990 }, { "grad_norm": 0.3357759714126587, "learning_rate": 6.027296183056252e-06, "loss": 0.0166, "step": 68000 }, { "grad_norm": 0.28034672141075134, "learning_rate": 6.017462130706114e-06, "loss": 0.0125, "step": 68010 }, { "grad_norm": 0.14257843792438507, "learning_rate": 6.0076355937687516e-06, "loss": 0.0157, "step": 68020 }, { "grad_norm": 0.2251233607530594, "learning_rate": 5.9978165739232925e-06, "loss": 0.0124, "step": 68030 }, { "grad_norm": 0.22612562775611877, "learning_rate": 5.988005072847508e-06, "loss": 0.0121, "step": 68040 }, { "grad_norm": 0.269619882106781, "learning_rate": 5.978201092217928e-06, "loss": 0.0219, "step": 68050 }, { "grad_norm": 0.18737946450710297, "learning_rate": 5.9684046337097895e-06, "loss": 0.0094, "step": 68060 }, { "grad_norm": 0.0932295024394989, "learning_rate": 5.958615698997016e-06, "loss": 0.0146, "step": 68070 }, { "grad_norm": 0.15949147939682007, "learning_rate": 5.948834289752303e-06, "loss": 0.0109, "step": 68080 }, { "grad_norm": 0.2013389766216278, "learning_rate": 5.939060407646996e-06, "loss": 0.0182, "step": 68090 }, { "grad_norm": 0.32902416586875916, "learning_rate": 5.929294054351198e-06, "loss": 0.0123, "step": 68100 }, { "grad_norm": 0.319216787815094, "learning_rate": 5.919535231533707e-06, "loss": 0.0126, "step": 68110 }, { "grad_norm": 0.27889347076416016, "learning_rate": 5.909783940862046e-06, "loss": 0.0148, "step": 68120 }, { "grad_norm": 0.2581186294555664, "learning_rate": 5.900040184002436e-06, "loss": 0.0103, "step": 68130 }, { "grad_norm": 0.6452799439430237, "learning_rate": 5.890303962619831e-06, "loss": 0.0107, "step": 68140 }, { "grad_norm": 0.3071441650390625, "learning_rate": 5.88057527837787e-06, "loss": 0.0147, "step": 68150 }, { "grad_norm": 0.1489536017179489, "learning_rate": 5.870854132938924e-06, "loss": 0.0161, "step": 68160 }, { "grad_norm": 0.2372041940689087, "learning_rate": 5.861140527964071e-06, "loss": 0.0191, "step": 68170 }, { "grad_norm": 0.17951227724552155, "learning_rate": 5.851434465113098e-06, "loss": 0.0103, "step": 68180 }, { "grad_norm": 0.27470889687538147, "learning_rate": 5.841735946044524e-06, "loss": 0.0149, "step": 68190 }, { "grad_norm": 0.13435156643390656, "learning_rate": 5.832044972415523e-06, "loss": 0.0113, "step": 68200 }, { "grad_norm": 0.16546691954135895, "learning_rate": 5.8223615458820655e-06, "loss": 0.0148, "step": 68210 }, { "grad_norm": 0.14631947875022888, "learning_rate": 5.812685668098733e-06, "loss": 0.0129, "step": 68220 }, { "grad_norm": 0.14953747391700745, "learning_rate": 5.803017340718913e-06, "loss": 0.0145, "step": 68230 }, { "grad_norm": 0.3530823290348053, "learning_rate": 5.793356565394636e-06, "loss": 0.0119, "step": 68240 }, { "grad_norm": 0.24487099051475525, "learning_rate": 5.783703343776658e-06, "loss": 0.0108, "step": 68250 }, { "grad_norm": 0.16889232397079468, "learning_rate": 5.774057677514477e-06, "loss": 0.0104, "step": 68260 }, { "grad_norm": 0.11101408302783966, "learning_rate": 5.764419568256235e-06, "loss": 0.0134, "step": 68270 }, { "grad_norm": 0.38947588205337524, "learning_rate": 5.754789017648865e-06, "loss": 0.0193, "step": 68280 }, { "grad_norm": 0.17109881341457367, "learning_rate": 5.745166027337934e-06, "loss": 0.019, "step": 68290 }, { "grad_norm": 0.10792802274227142, "learning_rate": 5.735550598967754e-06, "loss": 0.0165, "step": 68300 }, { "grad_norm": 0.23031464219093323, "learning_rate": 5.725942734181339e-06, "loss": 0.0135, "step": 68310 }, { "grad_norm": 0.3549956977367401, "learning_rate": 5.7163424346204146e-06, "loss": 0.0129, "step": 68320 }, { "grad_norm": 0.2314586043357849, "learning_rate": 5.706749701925407e-06, "loss": 0.0118, "step": 68330 }, { "grad_norm": 0.2427460253238678, "learning_rate": 5.697164537735461e-06, "loss": 0.0101, "step": 68340 }, { "grad_norm": 0.4203832745552063, "learning_rate": 5.687586943688406e-06, "loss": 0.0081, "step": 68350 }, { "grad_norm": 0.16982105374336243, "learning_rate": 5.678016921420792e-06, "loss": 0.0123, "step": 68360 }, { "grad_norm": 0.3121841847896576, "learning_rate": 5.668454472567875e-06, "loss": 0.0226, "step": 68370 }, { "grad_norm": 0.38233375549316406, "learning_rate": 5.658899598763617e-06, "loss": 0.0137, "step": 68380 }, { "grad_norm": 0.1938065141439438, "learning_rate": 5.6493523016407035e-06, "loss": 0.0092, "step": 68390 }, { "grad_norm": 0.19277185201644897, "learning_rate": 5.6398125828304615e-06, "loss": 0.0115, "step": 68400 }, { "grad_norm": 0.3368407189846039, "learning_rate": 5.630280443963015e-06, "loss": 0.0134, "step": 68410 }, { "grad_norm": 0.26178354024887085, "learning_rate": 5.620755886667112e-06, "loss": 0.0132, "step": 68420 }, { "grad_norm": 0.15946578979492188, "learning_rate": 5.611238912570271e-06, "loss": 0.0129, "step": 68430 }, { "grad_norm": 0.318811297416687, "learning_rate": 5.601729523298649e-06, "loss": 0.0176, "step": 68440 }, { "grad_norm": 0.23562833666801453, "learning_rate": 5.592227720477161e-06, "loss": 0.0119, "step": 68450 }, { "grad_norm": 0.18265241384506226, "learning_rate": 5.582733505729415e-06, "loss": 0.0158, "step": 68460 }, { "grad_norm": 0.28276628255844116, "learning_rate": 5.573246880677668e-06, "loss": 0.0126, "step": 68470 }, { "grad_norm": 0.9522616267204285, "learning_rate": 5.56376784694298e-06, "loss": 0.0134, "step": 68480 }, { "grad_norm": 0.37713420391082764, "learning_rate": 5.554296406145027e-06, "loss": 0.0155, "step": 68490 }, { "grad_norm": 0.23369580507278442, "learning_rate": 5.544832559902219e-06, "loss": 0.0121, "step": 68500 }, { "grad_norm": 0.2653406858444214, "learning_rate": 5.53537630983168e-06, "loss": 0.0139, "step": 68510 }, { "grad_norm": 0.2369907647371292, "learning_rate": 5.5259276575492125e-06, "loss": 0.0173, "step": 68520 }, { "grad_norm": 0.27107393741607666, "learning_rate": 5.516486604669357e-06, "loss": 0.0162, "step": 68530 }, { "grad_norm": 0.2731119990348816, "learning_rate": 5.507053152805303e-06, "loss": 0.0165, "step": 68540 }, { "grad_norm": 0.2674970328807831, "learning_rate": 5.497627303568975e-06, "loss": 0.0171, "step": 68550 }, { "grad_norm": 0.1972525715827942, "learning_rate": 5.488209058571003e-06, "loss": 0.0078, "step": 68560 }, { "grad_norm": 0.3127971291542053, "learning_rate": 5.4787984194207054e-06, "loss": 0.0162, "step": 68570 }, { "grad_norm": 0.2574707865715027, "learning_rate": 5.469395387726095e-06, "loss": 0.0111, "step": 68580 }, { "grad_norm": 0.2708113193511963, "learning_rate": 5.45999996509391e-06, "loss": 0.0118, "step": 68590 }, { "grad_norm": 0.2667766213417053, "learning_rate": 5.450612153129536e-06, "loss": 0.0135, "step": 68600 }, { "grad_norm": 0.1792241930961609, "learning_rate": 5.4412319534371426e-06, "loss": 0.0139, "step": 68610 }, { "grad_norm": 0.3000185787677765, "learning_rate": 5.431859367619513e-06, "loss": 0.0102, "step": 68620 }, { "grad_norm": 0.18640705943107605, "learning_rate": 5.422494397278172e-06, "loss": 0.0128, "step": 68630 }, { "grad_norm": 0.47629085183143616, "learning_rate": 5.413137044013344e-06, "loss": 0.0152, "step": 68640 }, { "grad_norm": 0.27641698718070984, "learning_rate": 5.403787309423941e-06, "loss": 0.0167, "step": 68650 }, { "grad_norm": 0.12668439745903015, "learning_rate": 5.3944451951075835e-06, "loss": 0.007, "step": 68660 }, { "grad_norm": 0.21849636733531952, "learning_rate": 5.385110702660562e-06, "loss": 0.0113, "step": 68670 }, { "grad_norm": 0.1757870465517044, "learning_rate": 5.375783833677922e-06, "loss": 0.0119, "step": 68680 }, { "grad_norm": 0.2586023807525635, "learning_rate": 5.36646458975334e-06, "loss": 0.0126, "step": 68690 }, { "grad_norm": 0.5434878468513489, "learning_rate": 5.3571529724792294e-06, "loss": 0.0115, "step": 68700 }, { "grad_norm": 0.3963862657546997, "learning_rate": 5.347848983446702e-06, "loss": 0.0148, "step": 68710 }, { "grad_norm": 0.16234071552753448, "learning_rate": 5.3385526242455185e-06, "loss": 0.0133, "step": 68720 }, { "grad_norm": 0.4479745030403137, "learning_rate": 5.329263896464226e-06, "loss": 0.0094, "step": 68730 }, { "grad_norm": 0.13787710666656494, "learning_rate": 5.3199828016899715e-06, "loss": 0.01, "step": 68740 }, { "grad_norm": 0.27529460191726685, "learning_rate": 5.310709341508657e-06, "loss": 0.0124, "step": 68750 }, { "grad_norm": 0.24202966690063477, "learning_rate": 5.301443517504861e-06, "loss": 0.0105, "step": 68760 }, { "grad_norm": 0.17372865974903107, "learning_rate": 5.292185331261862e-06, "loss": 0.0085, "step": 68770 }, { "grad_norm": 0.18479712307453156, "learning_rate": 5.28293478436162e-06, "loss": 0.0124, "step": 68780 }, { "grad_norm": 0.671730637550354, "learning_rate": 5.273691878384829e-06, "loss": 0.0152, "step": 68790 }, { "grad_norm": 0.2408944070339203, "learning_rate": 5.264456614910812e-06, "loss": 0.0132, "step": 68800 }, { "grad_norm": 0.25458547472953796, "learning_rate": 5.255228995517647e-06, "loss": 0.0222, "step": 68810 }, { "grad_norm": 0.14508414268493652, "learning_rate": 5.246009021782067e-06, "loss": 0.0108, "step": 68820 }, { "grad_norm": 0.25991761684417725, "learning_rate": 5.2367966952795225e-06, "loss": 0.0179, "step": 68830 }, { "grad_norm": 0.2220412641763687, "learning_rate": 5.2275920175841485e-06, "loss": 0.012, "step": 68840 }, { "grad_norm": 0.19080780446529388, "learning_rate": 5.218394990268771e-06, "loss": 0.0115, "step": 68850 }, { "grad_norm": 0.3945019543170929, "learning_rate": 5.209205614904916e-06, "loss": 0.0154, "step": 68860 }, { "grad_norm": 0.20986855030059814, "learning_rate": 5.200023893062777e-06, "loss": 0.0096, "step": 68870 }, { "grad_norm": 0.17603789269924164, "learning_rate": 5.190849826311289e-06, "loss": 0.0101, "step": 68880 }, { "grad_norm": 0.45807990431785583, "learning_rate": 5.181683416218025e-06, "loss": 0.0154, "step": 68890 }, { "grad_norm": 0.17793913185596466, "learning_rate": 5.172524664349276e-06, "loss": 0.0115, "step": 68900 }, { "grad_norm": 0.411228209733963, "learning_rate": 5.1633735722700416e-06, "loss": 0.017, "step": 68910 }, { "grad_norm": 0.29389500617980957, "learning_rate": 5.154230141543958e-06, "loss": 0.0114, "step": 68920 }, { "grad_norm": 0.2832350432872772, "learning_rate": 5.145094373733433e-06, "loss": 0.0124, "step": 68930 }, { "grad_norm": 0.13933300971984863, "learning_rate": 5.135966270399478e-06, "loss": 0.0103, "step": 68940 }, { "grad_norm": 0.49102145433425903, "learning_rate": 5.126845833101857e-06, "loss": 0.0148, "step": 68950 }, { "grad_norm": 0.17794270813465118, "learning_rate": 5.1177330633989994e-06, "loss": 0.0132, "step": 68960 }, { "grad_norm": 0.2796994745731354, "learning_rate": 5.108627962848033e-06, "loss": 0.0274, "step": 68970 }, { "grad_norm": 0.27248409390449524, "learning_rate": 5.099530533004759e-06, "loss": 0.0171, "step": 68980 }, { "grad_norm": 0.2554261088371277, "learning_rate": 5.090440775423699e-06, "loss": 0.0123, "step": 68990 }, { "grad_norm": 0.20306168496608734, "learning_rate": 5.081358691658022e-06, "loss": 0.0135, "step": 69000 }, { "grad_norm": 0.9016535878181458, "learning_rate": 5.072284283259621e-06, "loss": 0.0156, "step": 69010 }, { "grad_norm": 0.2245628833770752, "learning_rate": 5.063217551779053e-06, "loss": 0.0176, "step": 69020 }, { "grad_norm": 0.2839782238006592, "learning_rate": 5.054158498765588e-06, "loss": 0.0103, "step": 69030 }, { "grad_norm": 0.3372479975223541, "learning_rate": 5.0451071257671625e-06, "loss": 0.0123, "step": 69040 }, { "grad_norm": 0.4793764650821686, "learning_rate": 5.036063434330407e-06, "loss": 0.0201, "step": 69050 }, { "grad_norm": 0.17097043991088867, "learning_rate": 5.027027426000652e-06, "loss": 0.0127, "step": 69060 }, { "grad_norm": 0.2180471271276474, "learning_rate": 5.017999102321886e-06, "loss": 0.0143, "step": 69070 }, { "grad_norm": 0.323973149061203, "learning_rate": 5.0089784648368224e-06, "loss": 0.0145, "step": 69080 }, { "grad_norm": 0.1159965917468071, "learning_rate": 4.999965515086829e-06, "loss": 0.0092, "step": 69090 }, { "grad_norm": 0.20887491106987, "learning_rate": 4.990960254611976e-06, "loss": 0.0116, "step": 69100 }, { "grad_norm": 0.40104302763938904, "learning_rate": 4.9819626849510194e-06, "loss": 0.0168, "step": 69110 }, { "grad_norm": 0.23139281570911407, "learning_rate": 4.97297280764138e-06, "loss": 0.0171, "step": 69120 }, { "grad_norm": 0.21250639855861664, "learning_rate": 4.963990624219211e-06, "loss": 0.0134, "step": 69130 }, { "grad_norm": 0.3815348446369171, "learning_rate": 4.955016136219298e-06, "loss": 0.0131, "step": 69140 }, { "grad_norm": 0.20110994577407837, "learning_rate": 4.946049345175146e-06, "loss": 0.0106, "step": 69150 }, { "grad_norm": 0.35992154479026794, "learning_rate": 4.937090252618937e-06, "loss": 0.0141, "step": 69160 }, { "grad_norm": 0.32480210065841675, "learning_rate": 4.928138860081521e-06, "loss": 0.0156, "step": 69170 }, { "grad_norm": 0.19762298464775085, "learning_rate": 4.919195169092472e-06, "loss": 0.0173, "step": 69180 }, { "grad_norm": 0.1445176899433136, "learning_rate": 4.910259181179994e-06, "loss": 0.0133, "step": 69190 }, { "grad_norm": 0.22938847541809082, "learning_rate": 4.901330897871015e-06, "loss": 0.0131, "step": 69200 }, { "grad_norm": 0.24025721848011017, "learning_rate": 4.8924103206911375e-06, "loss": 0.0103, "step": 69210 }, { "grad_norm": 0.20434871315956116, "learning_rate": 4.883497451164637e-06, "loss": 0.0216, "step": 69220 }, { "grad_norm": 0.1372319459915161, "learning_rate": 4.874592290814484e-06, "loss": 0.0111, "step": 69230 }, { "grad_norm": 0.5845024585723877, "learning_rate": 4.865694841162327e-06, "loss": 0.0248, "step": 69240 }, { "grad_norm": 0.19571660459041595, "learning_rate": 4.856805103728496e-06, "loss": 0.014, "step": 69250 }, { "grad_norm": 0.1351248025894165, "learning_rate": 4.847923080032007e-06, "loss": 0.0103, "step": 69260 }, { "grad_norm": 0.2900935709476471, "learning_rate": 4.8390487715905486e-06, "loss": 0.0112, "step": 69270 }, { "grad_norm": 1.0097824335098267, "learning_rate": 4.8301821799205e-06, "loss": 0.0097, "step": 69280 }, { "grad_norm": 0.22985540330410004, "learning_rate": 4.821323306536918e-06, "loss": 0.0115, "step": 69290 }, { "grad_norm": 0.18328316509723663, "learning_rate": 4.8124721529535455e-06, "loss": 0.0107, "step": 69300 }, { "grad_norm": 0.24098631739616394, "learning_rate": 4.803628720682807e-06, "loss": 0.0158, "step": 69310 }, { "grad_norm": 0.22660958766937256, "learning_rate": 4.794793011235776e-06, "loss": 0.0137, "step": 69320 }, { "grad_norm": 0.19177113473415375, "learning_rate": 4.78596502612228e-06, "loss": 0.0154, "step": 69330 }, { "grad_norm": 0.3133293092250824, "learning_rate": 4.777144766850738e-06, "loss": 0.0153, "step": 69340 }, { "grad_norm": 0.17412933707237244, "learning_rate": 4.768332234928313e-06, "loss": 0.0138, "step": 69350 }, { "grad_norm": 0.0848008468747139, "learning_rate": 4.759527431860828e-06, "loss": 0.0124, "step": 69360 }, { "grad_norm": 0.23703095316886902, "learning_rate": 4.750730359152755e-06, "loss": 0.015, "step": 69370 }, { "grad_norm": 0.14749978482723236, "learning_rate": 4.741941018307311e-06, "loss": 0.0134, "step": 69380 }, { "grad_norm": 0.18595215678215027, "learning_rate": 4.733159410826321e-06, "loss": 0.0171, "step": 69390 }, { "grad_norm": 0.26941776275634766, "learning_rate": 4.724385538210357e-06, "loss": 0.012, "step": 69400 }, { "grad_norm": 0.2634221017360687, "learning_rate": 4.715619401958599e-06, "loss": 0.015, "step": 69410 }, { "grad_norm": 0.8368701934814453, "learning_rate": 4.706861003568958e-06, "loss": 0.0165, "step": 69420 }, { "grad_norm": 0.25999075174331665, "learning_rate": 4.698110344538003e-06, "loss": 0.0135, "step": 69430 }, { "grad_norm": 0.24445965886116028, "learning_rate": 4.689367426360975e-06, "loss": 0.0123, "step": 69440 }, { "grad_norm": 0.2870791256427765, "learning_rate": 4.680632250531819e-06, "loss": 0.0138, "step": 69450 }, { "grad_norm": 0.20172527432441711, "learning_rate": 4.671904818543115e-06, "loss": 0.0122, "step": 69460 }, { "grad_norm": 0.2852933704853058, "learning_rate": 4.66318513188615e-06, "loss": 0.0121, "step": 69470 }, { "grad_norm": 0.21913458406925201, "learning_rate": 4.654473192050884e-06, "loss": 0.0117, "step": 69480 }, { "grad_norm": 0.1902472823858261, "learning_rate": 4.64576900052594e-06, "loss": 0.0095, "step": 69490 }, { "grad_norm": 0.25760942697525024, "learning_rate": 4.637072558798638e-06, "loss": 0.0107, "step": 69500 }, { "grad_norm": 0.23033051192760468, "learning_rate": 4.628383868354969e-06, "loss": 0.0169, "step": 69510 }, { "grad_norm": 0.25750207901000977, "learning_rate": 4.6197029306795595e-06, "loss": 0.0087, "step": 69520 }, { "grad_norm": 0.35301709175109863, "learning_rate": 4.611029747255779e-06, "loss": 0.0094, "step": 69530 }, { "grad_norm": 0.22299623489379883, "learning_rate": 4.6023643195656164e-06, "loss": 0.0125, "step": 69540 }, { "grad_norm": 0.2068226933479309, "learning_rate": 4.593706649089768e-06, "loss": 0.0111, "step": 69550 }, { "grad_norm": 0.4853406548500061, "learning_rate": 4.585056737307597e-06, "loss": 0.0102, "step": 69560 }, { "grad_norm": 0.1534854769706726, "learning_rate": 4.576414585697103e-06, "loss": 0.0106, "step": 69570 }, { "grad_norm": 0.3997928500175476, "learning_rate": 4.567780195735044e-06, "loss": 0.0158, "step": 69580 }, { "grad_norm": 0.16572105884552002, "learning_rate": 4.559153568896757e-06, "loss": 0.0113, "step": 69590 }, { "grad_norm": 0.33061036467552185, "learning_rate": 4.550534706656329e-06, "loss": 0.0144, "step": 69600 }, { "grad_norm": 0.3373679518699646, "learning_rate": 4.541923610486465e-06, "loss": 0.0115, "step": 69610 }, { "grad_norm": 0.23147818446159363, "learning_rate": 4.533320281858578e-06, "loss": 0.0125, "step": 69620 }, { "grad_norm": 0.34944814443588257, "learning_rate": 4.52472472224274e-06, "loss": 0.0166, "step": 69630 }, { "grad_norm": 0.1842947006225586, "learning_rate": 4.51613693310769e-06, "loss": 0.012, "step": 69640 }, { "grad_norm": 0.25968557596206665, "learning_rate": 4.507556915920868e-06, "loss": 0.0128, "step": 69650 }, { "grad_norm": 0.22342094779014587, "learning_rate": 4.498984672148332e-06, "loss": 0.0105, "step": 69660 }, { "grad_norm": 0.2486022561788559, "learning_rate": 4.490420203254864e-06, "loss": 0.0154, "step": 69670 }, { "grad_norm": 0.6672602295875549, "learning_rate": 4.481863510703893e-06, "loss": 0.0184, "step": 69680 }, { "grad_norm": 0.15884877741336823, "learning_rate": 4.473314595957523e-06, "loss": 0.0117, "step": 69690 }, { "grad_norm": 0.313421368598938, "learning_rate": 4.464773460476535e-06, "loss": 0.0133, "step": 69700 }, { "grad_norm": 0.1661117970943451, "learning_rate": 4.456240105720372e-06, "loss": 0.011, "step": 69710 }, { "grad_norm": 0.22635619342327118, "learning_rate": 4.4477145331471405e-06, "loss": 0.0186, "step": 69720 }, { "grad_norm": 0.20929640531539917, "learning_rate": 4.439196744213653e-06, "loss": 0.0094, "step": 69730 }, { "grad_norm": 0.14800776541233063, "learning_rate": 4.430686740375339e-06, "loss": 0.0124, "step": 69740 }, { "grad_norm": 0.26312950253486633, "learning_rate": 4.422184523086342e-06, "loss": 0.0104, "step": 69750 }, { "grad_norm": 0.22854773700237274, "learning_rate": 4.41369009379946e-06, "loss": 0.0104, "step": 69760 }, { "grad_norm": 0.15619607269763947, "learning_rate": 4.405203453966139e-06, "loss": 0.0149, "step": 69770 }, { "grad_norm": 0.4257729947566986, "learning_rate": 4.396724605036539e-06, "loss": 0.0124, "step": 69780 }, { "grad_norm": 0.2299482524394989, "learning_rate": 4.388253548459437e-06, "loss": 0.0107, "step": 69790 }, { "grad_norm": 0.3089291751384735, "learning_rate": 4.3797902856823395e-06, "loss": 0.0125, "step": 69800 }, { "grad_norm": 0.13502733409404755, "learning_rate": 4.371334818151357e-06, "loss": 0.0099, "step": 69810 }, { "grad_norm": 0.28610101342201233, "learning_rate": 4.362887147311306e-06, "loss": 0.0131, "step": 69820 }, { "grad_norm": 0.2676157057285309, "learning_rate": 4.354447274605672e-06, "loss": 0.0123, "step": 69830 }, { "grad_norm": 0.17465710639953613, "learning_rate": 4.346015201476572e-06, "loss": 0.0198, "step": 69840 }, { "grad_norm": 0.19891390204429626, "learning_rate": 4.337590929364855e-06, "loss": 0.0104, "step": 69850 }, { "grad_norm": 0.33054623007774353, "learning_rate": 4.329174459709973e-06, "loss": 0.0113, "step": 69860 }, { "grad_norm": 0.159047931432724, "learning_rate": 4.320765793950071e-06, "loss": 0.013, "step": 69870 }, { "grad_norm": 0.2032114714384079, "learning_rate": 4.312364933521962e-06, "loss": 0.0147, "step": 69880 }, { "grad_norm": 0.19691058993339539, "learning_rate": 4.303971879861129e-06, "loss": 0.0122, "step": 69890 }, { "grad_norm": 0.3774654269218445, "learning_rate": 4.295586634401716e-06, "loss": 0.0155, "step": 69900 }, { "grad_norm": 0.1907101571559906, "learning_rate": 4.287209198576536e-06, "loss": 0.0103, "step": 69910 }, { "grad_norm": 0.22267453372478485, "learning_rate": 4.278839573817045e-06, "loss": 0.0149, "step": 69920 }, { "grad_norm": 0.6370172500610352, "learning_rate": 4.270477761553399e-06, "loss": 0.0134, "step": 69930 }, { "grad_norm": 0.24498522281646729, "learning_rate": 4.262123763214393e-06, "loss": 0.0094, "step": 69940 }, { "grad_norm": 0.23360194265842438, "learning_rate": 4.253777580227508e-06, "loss": 0.0104, "step": 69950 }, { "grad_norm": 0.46805065870285034, "learning_rate": 4.2454392140188755e-06, "loss": 0.0137, "step": 69960 }, { "grad_norm": 0.3078257441520691, "learning_rate": 4.2371086660132785e-06, "loss": 0.0129, "step": 69970 }, { "grad_norm": 0.318859726190567, "learning_rate": 4.228785937634205e-06, "loss": 0.0108, "step": 69980 }, { "grad_norm": 0.22463195025920868, "learning_rate": 4.220471030303758e-06, "loss": 0.0108, "step": 69990 }, { "grad_norm": 0.210575670003891, "learning_rate": 4.212163945442754e-06, "loss": 0.0098, "step": 70000 }, { "grad_norm": 0.38577648997306824, "learning_rate": 4.203864684470621e-06, "loss": 0.0092, "step": 70010 }, { "grad_norm": 0.32455939054489136, "learning_rate": 4.19557324880549e-06, "loss": 0.0151, "step": 70020 }, { "grad_norm": 0.17026007175445557, "learning_rate": 4.187289639864145e-06, "loss": 0.012, "step": 70030 }, { "grad_norm": 0.17341893911361694, "learning_rate": 4.1790138590619974e-06, "loss": 0.0113, "step": 70040 }, { "grad_norm": 0.27908846735954285, "learning_rate": 4.170745907813195e-06, "loss": 0.0155, "step": 70050 }, { "grad_norm": 0.15706127882003784, "learning_rate": 4.162485787530479e-06, "loss": 0.0127, "step": 70060 }, { "grad_norm": 0.3413764536380768, "learning_rate": 4.154233499625282e-06, "loss": 0.0139, "step": 70070 }, { "grad_norm": 0.17342956364154816, "learning_rate": 4.145989045507692e-06, "loss": 0.0159, "step": 70080 }, { "grad_norm": 0.5191631317138672, "learning_rate": 4.1377524265864666e-06, "loss": 0.0121, "step": 70090 }, { "grad_norm": 0.2399926334619522, "learning_rate": 4.1295236442690175e-06, "loss": 0.0105, "step": 70100 }, { "grad_norm": 0.3498480021953583, "learning_rate": 4.121302699961421e-06, "loss": 0.012, "step": 70110 }, { "grad_norm": 0.34853434562683105, "learning_rate": 4.113089595068403e-06, "loss": 0.0128, "step": 70120 }, { "grad_norm": 0.2001490443944931, "learning_rate": 4.104884330993364e-06, "loss": 0.0131, "step": 70130 }, { "grad_norm": 0.2101440131664276, "learning_rate": 4.0966869091383585e-06, "loss": 0.018, "step": 70140 }, { "grad_norm": 0.26662102341651917, "learning_rate": 4.088497330904101e-06, "loss": 0.0138, "step": 70150 }, { "grad_norm": 0.27703583240509033, "learning_rate": 4.080315597689976e-06, "loss": 0.0146, "step": 70160 }, { "grad_norm": 0.15494310855865479, "learning_rate": 4.072141710893995e-06, "loss": 0.0163, "step": 70170 }, { "grad_norm": 0.16989536583423615, "learning_rate": 4.063975671912879e-06, "loss": 0.0169, "step": 70180 }, { "grad_norm": 0.18631912767887115, "learning_rate": 4.055817482141949e-06, "loss": 0.0092, "step": 70190 }, { "grad_norm": 0.31151899695396423, "learning_rate": 4.047667142975259e-06, "loss": 0.0111, "step": 70200 }, { "grad_norm": 0.2670440077781677, "learning_rate": 4.039524655805443e-06, "loss": 0.0221, "step": 70210 }, { "grad_norm": 0.23807668685913086, "learning_rate": 4.03139002202384e-06, "loss": 0.01, "step": 70220 }, { "grad_norm": 0.42955002188682556, "learning_rate": 4.023263243020447e-06, "loss": 0.012, "step": 70230 }, { "grad_norm": 0.36216577887535095, "learning_rate": 4.015144320183884e-06, "loss": 0.0155, "step": 70240 }, { "grad_norm": 0.22955267131328583, "learning_rate": 4.007033254901482e-06, "loss": 0.0111, "step": 70250 }, { "grad_norm": 0.25335559248924255, "learning_rate": 3.9989300485591795e-06, "loss": 0.0129, "step": 70260 }, { "grad_norm": 0.16156038641929626, "learning_rate": 3.990834702541601e-06, "loss": 0.0158, "step": 70270 }, { "grad_norm": 0.25908058881759644, "learning_rate": 3.982747218232019e-06, "loss": 0.0124, "step": 70280 }, { "grad_norm": 0.18257389962673187, "learning_rate": 3.97466759701236e-06, "loss": 0.0111, "step": 70290 }, { "grad_norm": 0.4047677516937256, "learning_rate": 3.966595840263226e-06, "loss": 0.0132, "step": 70300 }, { "grad_norm": 0.27802637219429016, "learning_rate": 3.958531949363836e-06, "loss": 0.0113, "step": 70310 }, { "grad_norm": 0.20798878371715546, "learning_rate": 3.950475925692098e-06, "loss": 0.0126, "step": 70320 }, { "grad_norm": 0.20671412348747253, "learning_rate": 3.9424277706245685e-06, "loss": 0.0124, "step": 70330 }, { "grad_norm": 0.2013767510652542, "learning_rate": 3.934387485536451e-06, "loss": 0.0109, "step": 70340 }, { "grad_norm": 1.2554903030395508, "learning_rate": 3.926355071801619e-06, "loss": 0.0095, "step": 70350 }, { "grad_norm": 0.24109821021556854, "learning_rate": 3.9183305307925965e-06, "loss": 0.012, "step": 70360 }, { "grad_norm": 0.12149015814065933, "learning_rate": 3.910313863880533e-06, "loss": 0.007, "step": 70370 }, { "grad_norm": 0.2650837302207947, "learning_rate": 3.902305072435292e-06, "loss": 0.0114, "step": 70380 }, { "grad_norm": 0.24610435962677002, "learning_rate": 3.894304157825329e-06, "loss": 0.0108, "step": 70390 }, { "grad_norm": 0.2686416804790497, "learning_rate": 3.886311121417791e-06, "loss": 0.013, "step": 70400 }, { "grad_norm": 0.2082161009311676, "learning_rate": 3.878325964578472e-06, "loss": 0.0105, "step": 70410 }, { "grad_norm": 0.1794104129076004, "learning_rate": 3.870348688671815e-06, "loss": 0.0089, "step": 70420 }, { "grad_norm": 0.28493037819862366, "learning_rate": 3.862379295060931e-06, "loss": 0.0106, "step": 70430 }, { "grad_norm": 0.2207559049129486, "learning_rate": 3.854417785107534e-06, "loss": 0.0214, "step": 70440 }, { "grad_norm": 0.28223103284835815, "learning_rate": 3.8464641601720755e-06, "loss": 0.011, "step": 70450 }, { "grad_norm": 0.34369662404060364, "learning_rate": 3.838518421613579e-06, "loss": 0.0154, "step": 70460 }, { "grad_norm": 0.5303446054458618, "learning_rate": 3.830580570789766e-06, "loss": 0.0167, "step": 70470 }, { "grad_norm": 0.24953798949718475, "learning_rate": 3.822650609057005e-06, "loss": 0.0097, "step": 70480 }, { "grad_norm": 0.2710561752319336, "learning_rate": 3.814728537770285e-06, "loss": 0.0121, "step": 70490 }, { "grad_norm": 0.6424095630645752, "learning_rate": 3.806814358283306e-06, "loss": 0.0133, "step": 70500 }, { "grad_norm": 0.3296862542629242, "learning_rate": 3.7989080719483596e-06, "loss": 0.0125, "step": 70510 }, { "grad_norm": 0.1176658645272255, "learning_rate": 3.7910096801164143e-06, "loss": 0.0155, "step": 70520 }, { "grad_norm": 0.25500616431236267, "learning_rate": 3.7831191841371016e-06, "loss": 0.0107, "step": 70530 }, { "grad_norm": 0.30086150765419006, "learning_rate": 3.775236585358688e-06, "loss": 0.0135, "step": 70540 }, { "grad_norm": 0.09510167688131332, "learning_rate": 3.7673618851280843e-06, "loss": 0.0127, "step": 70550 }, { "grad_norm": 0.34766513109207153, "learning_rate": 3.759495084790887e-06, "loss": 0.0118, "step": 70560 }, { "grad_norm": 0.2367592304944992, "learning_rate": 3.751636185691282e-06, "loss": 0.0148, "step": 70570 }, { "grad_norm": 0.3298431932926178, "learning_rate": 3.7437851891721607e-06, "loss": 0.0135, "step": 70580 }, { "grad_norm": 0.3777453601360321, "learning_rate": 3.7359420965750404e-06, "loss": 0.0126, "step": 70590 }, { "grad_norm": 0.1743740439414978, "learning_rate": 3.7281069092400922e-06, "loss": 0.0105, "step": 70600 }, { "grad_norm": 0.34147781133651733, "learning_rate": 3.7202796285061348e-06, "loss": 0.0128, "step": 70610 }, { "grad_norm": 0.2655351459980011, "learning_rate": 3.712460255710637e-06, "loss": 0.0147, "step": 70620 }, { "grad_norm": 0.13530121743679047, "learning_rate": 3.704648792189719e-06, "loss": 0.0094, "step": 70630 }, { "grad_norm": 0.5423099994659424, "learning_rate": 3.696845239278124e-06, "loss": 0.0115, "step": 70640 }, { "grad_norm": 0.4269520938396454, "learning_rate": 3.689049598309302e-06, "loss": 0.0146, "step": 70650 }, { "grad_norm": 0.2600570619106293, "learning_rate": 3.681261870615288e-06, "loss": 0.0206, "step": 70660 }, { "grad_norm": 0.23364748060703278, "learning_rate": 3.6734820575268004e-06, "loss": 0.0211, "step": 70670 }, { "grad_norm": 0.13070663809776306, "learning_rate": 3.665710160373209e-06, "loss": 0.0112, "step": 70680 }, { "grad_norm": 0.28248122334480286, "learning_rate": 3.65794618048248e-06, "loss": 0.0104, "step": 70690 }, { "grad_norm": 0.2010720670223236, "learning_rate": 3.6501901191813125e-06, "loss": 0.0097, "step": 70700 }, { "grad_norm": 0.14878104627132416, "learning_rate": 3.642441977794975e-06, "loss": 0.0138, "step": 70710 }, { "grad_norm": 0.2760995030403137, "learning_rate": 3.634701757647424e-06, "loss": 0.0149, "step": 70720 }, { "grad_norm": 0.4873889982700348, "learning_rate": 3.6269694600612468e-06, "loss": 0.0109, "step": 70730 }, { "grad_norm": 0.3768450617790222, "learning_rate": 3.619245086357681e-06, "loss": 0.0124, "step": 70740 }, { "grad_norm": 0.2809118330478668, "learning_rate": 3.611528637856615e-06, "loss": 0.0136, "step": 70750 }, { "grad_norm": 0.45762550830841064, "learning_rate": 3.6038201158765884e-06, "loss": 0.012, "step": 70760 }, { "grad_norm": 0.5383567810058594, "learning_rate": 3.5961195217347534e-06, "loss": 0.0204, "step": 70770 }, { "grad_norm": 0.18659943342208862, "learning_rate": 3.588426856746946e-06, "loss": 0.0115, "step": 70780 }, { "grad_norm": 0.1590569019317627, "learning_rate": 3.5807421222276316e-06, "loss": 0.0173, "step": 70790 }, { "grad_norm": 0.22311489284038544, "learning_rate": 3.5730653194899156e-06, "loss": 0.0144, "step": 70800 }, { "grad_norm": 0.13062216341495514, "learning_rate": 3.565396449845554e-06, "loss": 0.008, "step": 70810 }, { "grad_norm": 0.2991241216659546, "learning_rate": 3.557735514604954e-06, "loss": 0.0129, "step": 70820 }, { "grad_norm": 0.9036766886711121, "learning_rate": 3.5500825150771633e-06, "loss": 0.0124, "step": 70830 }, { "grad_norm": 0.13437378406524658, "learning_rate": 3.5424374525698466e-06, "loss": 0.0126, "step": 70840 }, { "grad_norm": 0.2592385709285736, "learning_rate": 3.5348003283893704e-06, "loss": 0.013, "step": 70850 }, { "grad_norm": 0.16503159701824188, "learning_rate": 3.527171143840685e-06, "loss": 0.0122, "step": 70860 }, { "grad_norm": 0.27634555101394653, "learning_rate": 3.519549900227409e-06, "loss": 0.0139, "step": 70870 }, { "grad_norm": 0.24631735682487488, "learning_rate": 3.511936598851828e-06, "loss": 0.0103, "step": 70880 }, { "grad_norm": 0.17564481496810913, "learning_rate": 3.504331241014813e-06, "loss": 0.0143, "step": 70890 }, { "grad_norm": 0.29128697514533997, "learning_rate": 3.4967338280159414e-06, "loss": 0.0115, "step": 70900 }, { "grad_norm": 0.27486905455589294, "learning_rate": 3.4891443611533846e-06, "loss": 0.0156, "step": 70910 }, { "grad_norm": 0.1665736734867096, "learning_rate": 3.481562841723984e-06, "loss": 0.0144, "step": 70920 }, { "grad_norm": 0.1776556521654129, "learning_rate": 3.473989271023215e-06, "loss": 0.0113, "step": 70930 }, { "grad_norm": 0.600597083568573, "learning_rate": 3.4664236503451854e-06, "loss": 0.0158, "step": 70940 }, { "grad_norm": 0.12977257370948792, "learning_rate": 3.4588659809826674e-06, "loss": 0.008, "step": 70950 }, { "grad_norm": 0.27292031049728394, "learning_rate": 3.451316264227039e-06, "loss": 0.0177, "step": 70960 }, { "grad_norm": 0.2533012926578522, "learning_rate": 3.4437745013683574e-06, "loss": 0.0102, "step": 70970 }, { "grad_norm": 0.2762291431427002, "learning_rate": 3.4362406936952916e-06, "loss": 0.012, "step": 70980 }, { "grad_norm": 0.5300081372261047, "learning_rate": 3.4287148424951786e-06, "loss": 0.0116, "step": 70990 }, { "grad_norm": 0.16096700727939606, "learning_rate": 3.4211969490539617e-06, "loss": 0.0118, "step": 71000 }, { "grad_norm": 0.27008986473083496, "learning_rate": 3.4136870146562584e-06, "loss": 0.0129, "step": 71010 }, { "grad_norm": 0.5444952249526978, "learning_rate": 3.406185040585308e-06, "loss": 0.0126, "step": 71020 }, { "grad_norm": 0.23425212502479553, "learning_rate": 3.3986910281229966e-06, "loss": 0.0141, "step": 71030 }, { "grad_norm": 0.3788716495037079, "learning_rate": 3.391204978549828e-06, "loss": 0.0154, "step": 71040 }, { "grad_norm": 0.3471268117427826, "learning_rate": 3.3837268931449785e-06, "loss": 0.0176, "step": 71050 }, { "grad_norm": 0.2674367129802704, "learning_rate": 3.376256773186248e-06, "loss": 0.014, "step": 71060 }, { "grad_norm": 0.16309669613838196, "learning_rate": 3.3687946199500664e-06, "loss": 0.0134, "step": 71070 }, { "grad_norm": 0.2941732704639435, "learning_rate": 3.3613404347115295e-06, "loss": 0.0117, "step": 71080 }, { "grad_norm": 0.2800297439098358, "learning_rate": 3.35389421874433e-06, "loss": 0.0106, "step": 71090 }, { "grad_norm": 0.15378884971141815, "learning_rate": 3.3464559733208446e-06, "loss": 0.0094, "step": 71100 }, { "grad_norm": 0.32722172141075134, "learning_rate": 3.3390256997120505e-06, "loss": 0.0108, "step": 71110 }, { "grad_norm": 0.23518820106983185, "learning_rate": 3.331603399187583e-06, "loss": 0.0113, "step": 71120 }, { "grad_norm": 0.17203077673912048, "learning_rate": 3.324189073015721e-06, "loss": 0.0097, "step": 71130 }, { "grad_norm": 0.13959850370883942, "learning_rate": 3.3167827224633465e-06, "loss": 0.0114, "step": 71140 }, { "grad_norm": 0.18060222268104553, "learning_rate": 3.309384348796024e-06, "loss": 0.0114, "step": 71150 }, { "grad_norm": 0.25946956872940063, "learning_rate": 3.301993953277921e-06, "loss": 0.015, "step": 71160 }, { "grad_norm": 0.3324938416481018, "learning_rate": 3.29461153717186e-06, "loss": 0.0125, "step": 71170 }, { "grad_norm": 0.2647894322872162, "learning_rate": 3.287237101739293e-06, "loss": 0.019, "step": 71180 }, { "grad_norm": 0.21553343534469604, "learning_rate": 3.2798706482403075e-06, "loss": 0.0137, "step": 71190 }, { "grad_norm": 0.2613854706287384, "learning_rate": 3.2725121779336285e-06, "loss": 0.0149, "step": 71200 }, { "grad_norm": 0.2578642964363098, "learning_rate": 3.265161692076618e-06, "loss": 0.0131, "step": 71210 }, { "grad_norm": 0.11690250784158707, "learning_rate": 3.257819191925282e-06, "loss": 0.01, "step": 71220 }, { "grad_norm": 0.19368919730186462, "learning_rate": 3.2504846787342392e-06, "loss": 0.0116, "step": 71230 }, { "grad_norm": 0.09561622142791748, "learning_rate": 3.243158153756759e-06, "loss": 0.0125, "step": 71240 }, { "grad_norm": 0.3265935778617859, "learning_rate": 3.2358396182447514e-06, "loss": 0.0107, "step": 71250 }, { "grad_norm": 1.169721007347107, "learning_rate": 3.2285290734487496e-06, "loss": 0.0101, "step": 71260 }, { "grad_norm": 0.39081060886383057, "learning_rate": 3.221226520617926e-06, "loss": 0.0102, "step": 71270 }, { "grad_norm": 0.20379069447517395, "learning_rate": 3.213931961000094e-06, "loss": 0.0176, "step": 71280 }, { "grad_norm": 0.3812651038169861, "learning_rate": 3.2066453958416733e-06, "loss": 0.0119, "step": 71290 }, { "grad_norm": 0.25212159752845764, "learning_rate": 3.199366826387773e-06, "loss": 0.0175, "step": 71300 }, { "grad_norm": 0.21277955174446106, "learning_rate": 3.192096253882071e-06, "loss": 0.014, "step": 71310 }, { "grad_norm": 0.2785114049911499, "learning_rate": 3.1848336795669177e-06, "loss": 0.0133, "step": 71320 }, { "grad_norm": 0.15410082042217255, "learning_rate": 3.1775791046833035e-06, "loss": 0.0114, "step": 71330 }, { "grad_norm": 0.15073156356811523, "learning_rate": 3.170332530470804e-06, "loss": 0.0123, "step": 71340 }, { "grad_norm": 0.39051342010498047, "learning_rate": 3.1630939581677012e-06, "loss": 0.0153, "step": 71350 }, { "grad_norm": 0.27932432293891907, "learning_rate": 3.155863389010838e-06, "loss": 0.0085, "step": 71360 }, { "grad_norm": 0.5756843090057373, "learning_rate": 3.1486408242357323e-06, "loss": 0.0147, "step": 71370 }, { "grad_norm": 0.8412149548530579, "learning_rate": 3.1414262650765248e-06, "loss": 0.0258, "step": 71380 }, { "grad_norm": 0.27103784680366516, "learning_rate": 3.134219712765979e-06, "loss": 0.016, "step": 71390 }, { "grad_norm": 0.5742740035057068, "learning_rate": 3.1270211685355044e-06, "loss": 0.017, "step": 71400 }, { "grad_norm": 0.39324140548706055, "learning_rate": 3.1198306336151338e-06, "loss": 0.0131, "step": 71410 }, { "grad_norm": 0.34075161814689636, "learning_rate": 3.11264810923354e-06, "loss": 0.0129, "step": 71420 }, { "grad_norm": 0.14603285491466522, "learning_rate": 3.105473596618008e-06, "loss": 0.0103, "step": 71430 }, { "grad_norm": 0.27581384778022766, "learning_rate": 3.0983070969944683e-06, "loss": 0.0145, "step": 71440 }, { "grad_norm": 0.20795631408691406, "learning_rate": 3.09114861158748e-06, "loss": 0.0099, "step": 71450 }, { "grad_norm": 0.1666027158498764, "learning_rate": 3.0839981416202314e-06, "loss": 0.0161, "step": 71460 }, { "grad_norm": 0.1612568348646164, "learning_rate": 3.0768556883145507e-06, "loss": 0.0122, "step": 71470 }, { "grad_norm": 0.3446834683418274, "learning_rate": 3.0697212528908834e-06, "loss": 0.0148, "step": 71480 }, { "grad_norm": 0.23106423020362854, "learning_rate": 3.0625948365683e-06, "loss": 0.0105, "step": 71490 }, { "grad_norm": 0.1277570277452469, "learning_rate": 3.055476440564531e-06, "loss": 0.0175, "step": 71500 }, { "grad_norm": 0.13109925389289856, "learning_rate": 3.0483660660958924e-06, "loss": 0.0124, "step": 71510 }, { "grad_norm": 0.3063754737377167, "learning_rate": 3.0412637143773624e-06, "loss": 0.015, "step": 71520 }, { "grad_norm": 0.19307219982147217, "learning_rate": 3.034169386622554e-06, "loss": 0.0129, "step": 71530 }, { "grad_norm": 0.2177611142396927, "learning_rate": 3.027083084043658e-06, "loss": 0.0111, "step": 71540 }, { "grad_norm": 0.26943859457969666, "learning_rate": 3.0200048078515676e-06, "loss": 0.0131, "step": 71550 }, { "grad_norm": 0.3920181095600128, "learning_rate": 3.012934559255742e-06, "loss": 0.0114, "step": 71560 }, { "grad_norm": 0.2211814969778061, "learning_rate": 3.0058723394642994e-06, "loss": 0.0148, "step": 71570 }, { "grad_norm": 0.1472996473312378, "learning_rate": 2.998818149683985e-06, "loss": 0.0155, "step": 71580 }, { "grad_norm": 0.2516174614429474, "learning_rate": 2.9917719911201627e-06, "loss": 0.0118, "step": 71590 }, { "grad_norm": 0.4787015914916992, "learning_rate": 2.9847338649768352e-06, "loss": 0.0119, "step": 71600 }, { "grad_norm": 0.15652284026145935, "learning_rate": 2.977703772456608e-06, "loss": 0.0118, "step": 71610 }, { "grad_norm": 0.25011900067329407, "learning_rate": 2.9706817147607535e-06, "loss": 0.011, "step": 71620 }, { "grad_norm": 0.2693939805030823, "learning_rate": 2.9636676930891393e-06, "loss": 0.0118, "step": 71630 }, { "grad_norm": 0.3003142178058624, "learning_rate": 2.9566617086402625e-06, "loss": 0.0096, "step": 71640 }, { "grad_norm": 0.43331992626190186, "learning_rate": 2.9496637626112655e-06, "loss": 0.0119, "step": 71650 }, { "grad_norm": 0.277992308139801, "learning_rate": 2.9426738561979027e-06, "loss": 0.0105, "step": 71660 }, { "grad_norm": 0.3561018109321594, "learning_rate": 2.9356919905945524e-06, "loss": 0.0269, "step": 71670 }, { "grad_norm": 0.2162589132785797, "learning_rate": 2.928718166994243e-06, "loss": 0.011, "step": 71680 }, { "grad_norm": 0.2848966121673584, "learning_rate": 2.9217523865885833e-06, "loss": 0.0162, "step": 71690 }, { "grad_norm": 0.15787889063358307, "learning_rate": 2.9147946505678537e-06, "loss": 0.0124, "step": 71700 }, { "grad_norm": 0.2429332584142685, "learning_rate": 2.9078449601209313e-06, "loss": 0.0092, "step": 71710 }, { "grad_norm": 0.6064149737358093, "learning_rate": 2.900903316435333e-06, "loss": 0.0183, "step": 71720 }, { "grad_norm": 0.41548341512680054, "learning_rate": 2.8939697206971983e-06, "loss": 0.0153, "step": 71730 }, { "grad_norm": 0.2417738288640976, "learning_rate": 2.8870441740912746e-06, "loss": 0.012, "step": 71740 }, { "grad_norm": 0.3747917413711548, "learning_rate": 2.8801266778009762e-06, "loss": 0.0144, "step": 71750 }, { "grad_norm": 0.24069814383983612, "learning_rate": 2.8732172330082798e-06, "loss": 0.015, "step": 71760 }, { "grad_norm": 0.24297700822353363, "learning_rate": 2.8663158408938517e-06, "loss": 0.0143, "step": 71770 }, { "grad_norm": 0.5860773324966431, "learning_rate": 2.859422502636938e-06, "loss": 0.0129, "step": 71780 }, { "grad_norm": 0.24212248623371124, "learning_rate": 2.8525372194154076e-06, "loss": 0.012, "step": 71790 }, { "grad_norm": 0.1400633007287979, "learning_rate": 2.8456599924057913e-06, "loss": 0.01, "step": 71800 }, { "grad_norm": 0.15243852138519287, "learning_rate": 2.8387908227831995e-06, "loss": 0.012, "step": 71810 }, { "grad_norm": 0.17674146592617035, "learning_rate": 2.831929711721404e-06, "loss": 0.0128, "step": 71820 }, { "grad_norm": 0.2074471116065979, "learning_rate": 2.825076660392767e-06, "loss": 0.0114, "step": 71830 }, { "grad_norm": 0.5195015072822571, "learning_rate": 2.8182316699682908e-06, "loss": 0.0179, "step": 71840 }, { "grad_norm": 0.3495255410671234, "learning_rate": 2.811394741617601e-06, "loss": 0.0149, "step": 71850 }, { "grad_norm": 0.4063493013381958, "learning_rate": 2.8045658765089356e-06, "loss": 0.0143, "step": 71860 }, { "grad_norm": 0.22435303032398224, "learning_rate": 2.7977450758091605e-06, "loss": 0.018, "step": 71870 }, { "grad_norm": 0.2682998478412628, "learning_rate": 2.790932340683783e-06, "loss": 0.0158, "step": 71880 }, { "grad_norm": 0.2233860045671463, "learning_rate": 2.7841276722968823e-06, "loss": 0.0148, "step": 71890 }, { "grad_norm": 0.3408818542957306, "learning_rate": 2.7773310718112067e-06, "loss": 0.011, "step": 71900 }, { "grad_norm": 0.17027214169502258, "learning_rate": 2.7705425403881102e-06, "loss": 0.0127, "step": 71910 }, { "grad_norm": 0.17673350870609283, "learning_rate": 2.763762079187565e-06, "loss": 0.0105, "step": 71920 }, { "grad_norm": 0.3613740801811218, "learning_rate": 2.756989689368178e-06, "loss": 0.0137, "step": 71930 }, { "grad_norm": 0.2186768501996994, "learning_rate": 2.750225372087134e-06, "loss": 0.0133, "step": 71940 }, { "grad_norm": 0.32455021142959595, "learning_rate": 2.7434691285003033e-06, "loss": 0.0131, "step": 71950 }, { "grad_norm": 0.3048557639122009, "learning_rate": 2.7367209597621178e-06, "loss": 0.013, "step": 71960 }, { "grad_norm": 0.213412806391716, "learning_rate": 2.729980867025683e-06, "loss": 0.0166, "step": 71970 }, { "grad_norm": 0.2694293260574341, "learning_rate": 2.7232488514426724e-06, "loss": 0.0147, "step": 71980 }, { "grad_norm": 0.156792551279068, "learning_rate": 2.7165249141634097e-06, "loss": 0.0116, "step": 71990 }, { "grad_norm": 0.30148443579673767, "learning_rate": 2.709809056336837e-06, "loss": 0.0153, "step": 72000 }, { "grad_norm": 0.21271169185638428, "learning_rate": 2.703101279110498e-06, "loss": 0.0112, "step": 72010 }, { "grad_norm": 0.2280469685792923, "learning_rate": 2.696401583630587e-06, "loss": 0.0132, "step": 72020 }, { "grad_norm": 0.3478056788444519, "learning_rate": 2.689709971041887e-06, "loss": 0.0108, "step": 72030 }, { "grad_norm": 0.4274371266365051, "learning_rate": 2.6830264424878126e-06, "loss": 0.0136, "step": 72040 }, { "grad_norm": 0.3091392517089844, "learning_rate": 2.676350999110394e-06, "loss": 0.0131, "step": 72050 }, { "grad_norm": 0.23561158776283264, "learning_rate": 2.6696836420502856e-06, "loss": 0.0197, "step": 72060 }, { "grad_norm": 0.15663664042949677, "learning_rate": 2.663024372446765e-06, "loss": 0.0109, "step": 72070 }, { "grad_norm": 0.2504604756832123, "learning_rate": 2.6563731914377055e-06, "loss": 0.0117, "step": 72080 }, { "grad_norm": 0.22134636342525482, "learning_rate": 2.6497301001596087e-06, "loss": 0.0213, "step": 72090 }, { "grad_norm": 0.4632272720336914, "learning_rate": 2.643095099747611e-06, "loss": 0.0178, "step": 72100 }, { "grad_norm": 0.307511568069458, "learning_rate": 2.636468191335445e-06, "loss": 0.0136, "step": 72110 }, { "grad_norm": 0.212770015001297, "learning_rate": 2.629849376055471e-06, "loss": 0.0149, "step": 72120 }, { "grad_norm": 0.13353167474269867, "learning_rate": 2.6232386550386678e-06, "loss": 0.0119, "step": 72130 }, { "grad_norm": 0.33449751138687134, "learning_rate": 2.6166360294146097e-06, "loss": 0.011, "step": 72140 }, { "grad_norm": 0.23481276631355286, "learning_rate": 2.6100415003115275e-06, "loss": 0.0177, "step": 72150 }, { "grad_norm": 0.3671877980232239, "learning_rate": 2.603455068856225e-06, "loss": 0.0159, "step": 72160 }, { "grad_norm": 0.24628640711307526, "learning_rate": 2.5968767361741584e-06, "loss": 0.0143, "step": 72170 }, { "grad_norm": 0.23940375447273254, "learning_rate": 2.590306503389378e-06, "loss": 0.015, "step": 72180 }, { "grad_norm": 0.18223491311073303, "learning_rate": 2.583744371624558e-06, "loss": 0.0214, "step": 72190 }, { "grad_norm": 0.24894367158412933, "learning_rate": 2.57719034200099e-06, "loss": 0.0115, "step": 72200 }, { "grad_norm": 0.20591263473033905, "learning_rate": 2.570644415638568e-06, "loss": 0.0144, "step": 72210 }, { "grad_norm": 0.1846092790365219, "learning_rate": 2.5641065936558295e-06, "loss": 0.0164, "step": 72220 }, { "grad_norm": 0.364454448223114, "learning_rate": 2.557576877169898e-06, "loss": 0.0143, "step": 72230 }, { "grad_norm": 0.15851648151874542, "learning_rate": 2.5510552672965205e-06, "loss": 0.0078, "step": 72240 }, { "grad_norm": 0.5391554832458496, "learning_rate": 2.544541765150077e-06, "loss": 0.0125, "step": 72250 }, { "grad_norm": 0.24067361652851105, "learning_rate": 2.5380363718435165e-06, "loss": 0.0152, "step": 72260 }, { "grad_norm": 0.3658750355243683, "learning_rate": 2.5315390884884714e-06, "loss": 0.0135, "step": 72270 }, { "grad_norm": 0.1316455900669098, "learning_rate": 2.5250499161951214e-06, "loss": 0.0123, "step": 72280 }, { "grad_norm": 0.318727046251297, "learning_rate": 2.518568856072301e-06, "loss": 0.0117, "step": 72290 }, { "grad_norm": 0.1556268334388733, "learning_rate": 2.512095909227441e-06, "loss": 0.0119, "step": 72300 }, { "grad_norm": 0.18957243859767914, "learning_rate": 2.50563107676659e-06, "loss": 0.0144, "step": 72310 }, { "grad_norm": 0.1998429298400879, "learning_rate": 2.4991743597944152e-06, "loss": 0.0115, "step": 72320 }, { "grad_norm": 0.09936624765396118, "learning_rate": 2.492725759414205e-06, "loss": 0.0136, "step": 72330 }, { "grad_norm": 0.3882032334804535, "learning_rate": 2.4862852767278234e-06, "loss": 0.0109, "step": 72340 }, { "grad_norm": 0.10020897537469864, "learning_rate": 2.479852912835784e-06, "loss": 0.0084, "step": 72350 }, { "grad_norm": 0.2807200253009796, "learning_rate": 2.4734286688372075e-06, "loss": 0.0189, "step": 72360 }, { "grad_norm": 0.3450246751308441, "learning_rate": 2.467012545829811e-06, "loss": 0.016, "step": 72370 }, { "grad_norm": 0.20878565311431885, "learning_rate": 2.460604544909945e-06, "loss": 0.0106, "step": 72380 }, { "grad_norm": 0.3661392033100128, "learning_rate": 2.454204667172555e-06, "loss": 0.009, "step": 72390 }, { "grad_norm": 0.24571119248867035, "learning_rate": 2.447812913711217e-06, "loss": 0.0143, "step": 72400 }, { "grad_norm": 0.17954516410827637, "learning_rate": 2.44142928561808e-06, "loss": 0.0129, "step": 72410 }, { "grad_norm": 0.18904832005500793, "learning_rate": 2.435053783983965e-06, "loss": 0.01, "step": 72420 }, { "grad_norm": 0.38751205801963806, "learning_rate": 2.4286864098982453e-06, "loss": 0.017, "step": 72430 }, { "grad_norm": 0.276889443397522, "learning_rate": 2.4223271644489397e-06, "loss": 0.0122, "step": 72440 }, { "grad_norm": 0.09114424884319305, "learning_rate": 2.415976048722679e-06, "loss": 0.0074, "step": 72450 }, { "grad_norm": 0.4651961922645569, "learning_rate": 2.4096330638046673e-06, "loss": 0.0139, "step": 72460 }, { "grad_norm": 0.26111865043640137, "learning_rate": 2.4032982107787816e-06, "loss": 0.0133, "step": 72470 }, { "grad_norm": 0.17625582218170166, "learning_rate": 2.396971490727451e-06, "loss": 0.0123, "step": 72480 }, { "grad_norm": 0.18529754877090454, "learning_rate": 2.3906529047317493e-06, "loss": 0.009, "step": 72490 }, { "grad_norm": 0.32851648330688477, "learning_rate": 2.3843424538713465e-06, "loss": 0.0115, "step": 72500 }, { "grad_norm": 0.37246447801589966, "learning_rate": 2.3780401392245298e-06, "loss": 0.0141, "step": 72510 }, { "grad_norm": 0.3899514675140381, "learning_rate": 2.3717459618681883e-06, "loss": 0.0133, "step": 72520 }, { "grad_norm": 0.19299890100955963, "learning_rate": 2.3654599228778274e-06, "loss": 0.0147, "step": 72530 }, { "grad_norm": 0.2131156474351883, "learning_rate": 2.3591820233275607e-06, "loss": 0.0137, "step": 72540 }, { "grad_norm": 0.1842530518770218, "learning_rate": 2.3529122642901024e-06, "loss": 0.0159, "step": 72550 }, { "grad_norm": 0.32446372509002686, "learning_rate": 2.3466506468367845e-06, "loss": 0.0223, "step": 72560 }, { "grad_norm": 0.13926880061626434, "learning_rate": 2.3403971720375506e-06, "loss": 0.0119, "step": 72570 }, { "grad_norm": 0.17378437519073486, "learning_rate": 2.334151840960952e-06, "loss": 0.008, "step": 72580 }, { "grad_norm": 0.2728867530822754, "learning_rate": 2.3279146546741347e-06, "loss": 0.0109, "step": 72590 }, { "grad_norm": 0.37558072805404663, "learning_rate": 2.32168561424288e-06, "loss": 0.0118, "step": 72600 }, { "grad_norm": 0.2463637888431549, "learning_rate": 2.3154647207315307e-06, "loss": 0.0104, "step": 72610 }, { "grad_norm": 0.09786392748355865, "learning_rate": 2.309251975203103e-06, "loss": 0.0111, "step": 72620 }, { "grad_norm": 0.1929904669523239, "learning_rate": 2.303047378719159e-06, "loss": 0.0175, "step": 72630 }, { "grad_norm": 0.27386850118637085, "learning_rate": 2.2968509323399e-06, "loss": 0.0122, "step": 72640 }, { "grad_norm": 0.13680285215377808, "learning_rate": 2.290662637124147e-06, "loss": 0.0111, "step": 72650 }, { "grad_norm": 0.20767399668693542, "learning_rate": 2.2844824941292807e-06, "loss": 0.0085, "step": 72660 }, { "grad_norm": 0.2755945026874542, "learning_rate": 2.2783105044113406e-06, "loss": 0.0145, "step": 72670 }, { "grad_norm": 0.26275551319122314, "learning_rate": 2.272146669024944e-06, "loss": 0.0101, "step": 72680 }, { "grad_norm": 0.1982155740261078, "learning_rate": 2.26599098902332e-06, "loss": 0.0099, "step": 72690 }, { "grad_norm": 0.1332615464925766, "learning_rate": 2.2598434654583113e-06, "loss": 0.0093, "step": 72700 }, { "grad_norm": 0.36214515566825867, "learning_rate": 2.253704099380355e-06, "loss": 0.0146, "step": 72710 }, { "grad_norm": 0.24467827379703522, "learning_rate": 2.247572891838512e-06, "loss": 0.0147, "step": 72720 }, { "grad_norm": 0.27050063014030457, "learning_rate": 2.241449843880422e-06, "loss": 0.0129, "step": 72730 }, { "grad_norm": 0.13336396217346191, "learning_rate": 2.235334956552354e-06, "loss": 0.0176, "step": 72740 }, { "grad_norm": 0.1714780628681183, "learning_rate": 2.2292282308991775e-06, "loss": 0.014, "step": 72750 }, { "grad_norm": 0.3876652419567108, "learning_rate": 2.223129667964363e-06, "loss": 0.0111, "step": 72760 }, { "grad_norm": 0.16100434958934784, "learning_rate": 2.2170392687899834e-06, "loss": 0.0096, "step": 72770 }, { "grad_norm": 0.3118613660335541, "learning_rate": 2.210957034416733e-06, "loss": 0.0101, "step": 72780 }, { "grad_norm": 0.061667781323194504, "learning_rate": 2.2048829658838867e-06, "loss": 0.0113, "step": 72790 }, { "grad_norm": 0.22797146439552307, "learning_rate": 2.1988170642293525e-06, "loss": 0.0107, "step": 72800 }, { "grad_norm": 0.11839891970157623, "learning_rate": 2.1927593304896075e-06, "loss": 0.0108, "step": 72810 }, { "grad_norm": 0.2195233702659607, "learning_rate": 2.1867097656997626e-06, "loss": 0.0119, "step": 72820 }, { "grad_norm": 0.18617555499076843, "learning_rate": 2.180668370893524e-06, "loss": 0.0124, "step": 72830 }, { "grad_norm": 0.3704597055912018, "learning_rate": 2.174635147103199e-06, "loss": 0.0148, "step": 72840 }, { "grad_norm": 0.19856177270412445, "learning_rate": 2.1686100953597075e-06, "loss": 0.0088, "step": 72850 }, { "grad_norm": 0.3008286654949188, "learning_rate": 2.1625932166925433e-06, "loss": 0.0137, "step": 72860 }, { "grad_norm": 0.239911288022995, "learning_rate": 2.1565845121298556e-06, "loss": 0.0125, "step": 72870 }, { "grad_norm": 0.18427588045597076, "learning_rate": 2.150583982698351e-06, "loss": 0.0126, "step": 72880 }, { "grad_norm": 0.24587665498256683, "learning_rate": 2.144591629423359e-06, "loss": 0.0083, "step": 72890 }, { "grad_norm": 0.2470763921737671, "learning_rate": 2.138607453328817e-06, "loss": 0.0135, "step": 72900 }, { "grad_norm": 0.11121644824743271, "learning_rate": 2.132631455437234e-06, "loss": 0.0136, "step": 72910 }, { "grad_norm": 0.2537422776222229, "learning_rate": 2.1266636367697714e-06, "loss": 0.016, "step": 72920 }, { "grad_norm": 0.35955119132995605, "learning_rate": 2.120703998346152e-06, "loss": 0.0099, "step": 72930 }, { "grad_norm": 0.20902766287326813, "learning_rate": 2.1147525411847114e-06, "loss": 0.0115, "step": 72940 }, { "grad_norm": 0.2710510194301605, "learning_rate": 2.108809266302403e-06, "loss": 0.0104, "step": 72950 }, { "grad_norm": 0.242833212018013, "learning_rate": 2.102874174714764e-06, "loss": 0.0118, "step": 72960 }, { "grad_norm": 0.31862983107566833, "learning_rate": 2.0969472674359393e-06, "loss": 0.0146, "step": 72970 }, { "grad_norm": 0.24440763890743256, "learning_rate": 2.091028545478668e-06, "loss": 0.0161, "step": 72980 }, { "grad_norm": 0.15319442749023438, "learning_rate": 2.08511800985432e-06, "loss": 0.0079, "step": 72990 }, { "grad_norm": 0.2418946474790573, "learning_rate": 2.07921566157282e-06, "loss": 0.0178, "step": 73000 }, { "grad_norm": 0.23088762164115906, "learning_rate": 2.073321501642728e-06, "loss": 0.011, "step": 73010 }, { "grad_norm": 0.24230030179023743, "learning_rate": 2.0674355310711937e-06, "loss": 0.0101, "step": 73020 }, { "grad_norm": 0.22107350826263428, "learning_rate": 2.0615577508639682e-06, "loss": 0.009, "step": 73030 }, { "grad_norm": 0.29837819933891296, "learning_rate": 2.055688162025399e-06, "loss": 0.0114, "step": 73040 }, { "grad_norm": 1.2498515844345093, "learning_rate": 2.0498267655584546e-06, "loss": 0.0191, "step": 73050 }, { "grad_norm": 0.1948431432247162, "learning_rate": 2.0439735624646626e-06, "loss": 0.0091, "step": 73060 }, { "grad_norm": 0.2721415162086487, "learning_rate": 2.0381285537442008e-06, "loss": 0.0085, "step": 73070 }, { "grad_norm": 0.20255522429943085, "learning_rate": 2.032291740395803e-06, "loss": 0.0149, "step": 73080 }, { "grad_norm": 0.3843134343624115, "learning_rate": 2.0264631234168276e-06, "loss": 0.0165, "step": 73090 }, { "grad_norm": 0.23187269270420074, "learning_rate": 2.0206427038032326e-06, "loss": 0.0139, "step": 73100 }, { "grad_norm": 0.23567521572113037, "learning_rate": 2.0148304825495457e-06, "loss": 0.009, "step": 73110 }, { "grad_norm": 0.3376823365688324, "learning_rate": 2.0090264606489496e-06, "loss": 0.0082, "step": 73120 }, { "grad_norm": 0.4918532371520996, "learning_rate": 2.003230639093162e-06, "loss": 0.0141, "step": 73130 }, { "grad_norm": 0.15565653145313263, "learning_rate": 1.9974430188725525e-06, "loss": 0.0164, "step": 73140 }, { "grad_norm": 0.28568437695503235, "learning_rate": 1.9916636009760513e-06, "loss": 0.0134, "step": 73150 }, { "grad_norm": 0.22065411508083344, "learning_rate": 1.985892386391208e-06, "loss": 0.0144, "step": 73160 }, { "grad_norm": 0.2766842544078827, "learning_rate": 1.980129376104173e-06, "loss": 0.0127, "step": 73170 }, { "grad_norm": 0.1366690844297409, "learning_rate": 1.9743745710996796e-06, "loss": 0.0158, "step": 73180 }, { "grad_norm": 0.18558953702449799, "learning_rate": 1.9686279723610757e-06, "loss": 0.0115, "step": 73190 }, { "grad_norm": 0.3920474946498871, "learning_rate": 1.962889580870281e-06, "loss": 0.0121, "step": 73200 }, { "grad_norm": 0.1897437423467636, "learning_rate": 1.9571593976078385e-06, "loss": 0.0104, "step": 73210 }, { "grad_norm": 0.22385507822036743, "learning_rate": 1.9514374235528876e-06, "loss": 0.0121, "step": 73220 }, { "grad_norm": 0.4787673056125641, "learning_rate": 1.9457236596831408e-06, "loss": 0.0096, "step": 73230 }, { "grad_norm": 0.26373225450515747, "learning_rate": 1.940018106974939e-06, "loss": 0.0161, "step": 73240 }, { "grad_norm": 0.20236648619174957, "learning_rate": 1.9343207664032025e-06, "loss": 0.0095, "step": 73250 }, { "grad_norm": 0.1321093887090683, "learning_rate": 1.928631638941436e-06, "loss": 0.0072, "step": 73260 }, { "grad_norm": 0.20643696188926697, "learning_rate": 1.922950725561784e-06, "loss": 0.0092, "step": 73270 }, { "grad_norm": 0.21429851651191711, "learning_rate": 1.9172780272349312e-06, "loss": 0.0093, "step": 73280 }, { "grad_norm": 0.13063949346542358, "learning_rate": 1.911613544930202e-06, "loss": 0.0135, "step": 73290 }, { "grad_norm": 0.2969662547111511, "learning_rate": 1.9059572796155e-06, "loss": 0.0129, "step": 73300 }, { "grad_norm": 0.302350789308548, "learning_rate": 1.9003092322573192e-06, "loss": 0.0123, "step": 73310 }, { "grad_norm": 0.23878289759159088, "learning_rate": 1.8946694038207647e-06, "loss": 0.0133, "step": 73320 }, { "grad_norm": 0.5061202049255371, "learning_rate": 1.8890377952695215e-06, "loss": 0.011, "step": 73330 }, { "grad_norm": 0.21148595213890076, "learning_rate": 1.8834144075658865e-06, "loss": 0.0114, "step": 73340 }, { "grad_norm": 0.32989177107810974, "learning_rate": 1.8777992416707357e-06, "loss": 0.0124, "step": 73350 }, { "grad_norm": 0.2040010541677475, "learning_rate": 1.8721922985435458e-06, "loss": 0.0079, "step": 73360 }, { "grad_norm": 0.3328333795070648, "learning_rate": 1.8665935791424006e-06, "loss": 0.0132, "step": 73370 }, { "grad_norm": 0.20035777986049652, "learning_rate": 1.8610030844239512e-06, "loss": 0.0106, "step": 73380 }, { "grad_norm": 0.870049774646759, "learning_rate": 1.8554208153434838e-06, "loss": 0.0129, "step": 73390 }, { "grad_norm": 0.15101630985736847, "learning_rate": 1.8498467728548296e-06, "loss": 0.008, "step": 73400 }, { "grad_norm": 0.18429672718048096, "learning_rate": 1.8442809579104547e-06, "loss": 0.0092, "step": 73410 }, { "grad_norm": 0.23171471059322357, "learning_rate": 1.8387233714614039e-06, "loss": 0.0128, "step": 73420 }, { "grad_norm": 0.23081813752651215, "learning_rate": 1.8331740144573174e-06, "loss": 0.0138, "step": 73430 }, { "grad_norm": 0.2733449637889862, "learning_rate": 1.8276328878464255e-06, "loss": 0.0111, "step": 73440 }, { "grad_norm": 0.2839990556240082, "learning_rate": 1.82209999257556e-06, "loss": 0.0109, "step": 73450 }, { "grad_norm": 0.15738141536712646, "learning_rate": 1.8165753295901311e-06, "loss": 0.0133, "step": 73460 }, { "grad_norm": 0.14784079790115356, "learning_rate": 1.8110588998341616e-06, "loss": 0.014, "step": 73470 }, { "grad_norm": 0.19641269743442535, "learning_rate": 1.805550704250253e-06, "loss": 0.0168, "step": 73480 }, { "grad_norm": 0.2441188395023346, "learning_rate": 1.8000507437796077e-06, "loss": 0.0081, "step": 73490 }, { "grad_norm": 0.24611994624137878, "learning_rate": 1.7945590193620242e-06, "loss": 0.0127, "step": 73500 }, { "grad_norm": 0.19705593585968018, "learning_rate": 1.7890755319358742e-06, "loss": 0.0131, "step": 73510 }, { "grad_norm": 0.36925849318504333, "learning_rate": 1.7836002824381525e-06, "loss": 0.0153, "step": 73520 }, { "grad_norm": 0.24126432836055756, "learning_rate": 1.7781332718044165e-06, "loss": 0.0082, "step": 73530 }, { "grad_norm": 0.3022380471229553, "learning_rate": 1.772674500968835e-06, "loss": 0.0145, "step": 73540 }, { "grad_norm": 0.27306661009788513, "learning_rate": 1.7672239708641624e-06, "loss": 0.0148, "step": 73550 }, { "grad_norm": 0.09878332912921906, "learning_rate": 1.7617816824217371e-06, "loss": 0.0102, "step": 73560 }, { "grad_norm": 0.4064081609249115, "learning_rate": 1.7563476365715148e-06, "loss": 0.0151, "step": 73570 }, { "grad_norm": 0.10889681428670883, "learning_rate": 1.750921834242003e-06, "loss": 0.0147, "step": 73580 }, { "grad_norm": 0.3543998599052429, "learning_rate": 1.7455042763603436e-06, "loss": 0.0137, "step": 73590 }, { "grad_norm": 0.1543065756559372, "learning_rate": 1.7400949638522345e-06, "loss": 0.0115, "step": 73600 }, { "grad_norm": 0.3296581506729126, "learning_rate": 1.734693897641987e-06, "loss": 0.0233, "step": 73610 }, { "grad_norm": 0.30682840943336487, "learning_rate": 1.7293010786524955e-06, "loss": 0.011, "step": 73620 }, { "grad_norm": 0.271076500415802, "learning_rate": 1.72391650780524e-06, "loss": 0.0118, "step": 73630 }, { "grad_norm": 0.28803086280822754, "learning_rate": 1.718540186020301e-06, "loss": 0.0155, "step": 73640 }, { "grad_norm": 0.4180053770542145, "learning_rate": 1.7131721142163437e-06, "loss": 0.0121, "step": 73650 }, { "grad_norm": 0.29937106370925903, "learning_rate": 1.7078122933106233e-06, "loss": 0.0173, "step": 73660 }, { "grad_norm": 0.24933256208896637, "learning_rate": 1.7024607242189905e-06, "loss": 0.0136, "step": 73670 }, { "grad_norm": 0.23088112473487854, "learning_rate": 1.6971174078558749e-06, "loss": 0.01, "step": 73680 }, { "grad_norm": 0.2072000950574875, "learning_rate": 1.6917823451343073e-06, "loss": 0.0135, "step": 73690 }, { "grad_norm": 0.33676254749298096, "learning_rate": 1.6864555369659142e-06, "loss": 0.0106, "step": 73700 }, { "grad_norm": 0.32547616958618164, "learning_rate": 1.681136984260878e-06, "loss": 0.0112, "step": 73710 }, { "grad_norm": 0.1296796351671219, "learning_rate": 1.6758266879280172e-06, "loss": 0.0102, "step": 73720 }, { "grad_norm": 0.2840048372745514, "learning_rate": 1.6705246488746996e-06, "loss": 0.0103, "step": 73730 }, { "grad_norm": 0.08577614277601242, "learning_rate": 1.6652308680069062e-06, "loss": 0.0105, "step": 73740 }, { "grad_norm": 0.23841126263141632, "learning_rate": 1.6599453462292081e-06, "loss": 0.0123, "step": 73750 }, { "grad_norm": 0.4626079499721527, "learning_rate": 1.6546680844447326e-06, "loss": 0.0179, "step": 73760 }, { "grad_norm": 0.2957635819911957, "learning_rate": 1.6493990835552475e-06, "loss": 0.0139, "step": 73770 }, { "grad_norm": 0.15143705904483795, "learning_rate": 1.6441383444610547e-06, "loss": 0.0103, "step": 73780 }, { "grad_norm": 0.2005009800195694, "learning_rate": 1.6388858680610908e-06, "loss": 0.0162, "step": 73790 }, { "grad_norm": 0.2827439308166504, "learning_rate": 1.6336416552528543e-06, "loss": 0.0138, "step": 73800 }, { "grad_norm": 0.20221225917339325, "learning_rate": 1.6284057069324288e-06, "loss": 0.0144, "step": 73810 }, { "grad_norm": 0.18991927802562714, "learning_rate": 1.623178023994504e-06, "loss": 0.0189, "step": 73820 }, { "grad_norm": 0.3692372143268585, "learning_rate": 1.6179586073323483e-06, "loss": 0.0157, "step": 73830 }, { "grad_norm": 0.33548587560653687, "learning_rate": 1.612747457837821e-06, "loss": 0.0099, "step": 73840 }, { "grad_norm": 0.12379491329193115, "learning_rate": 1.607544576401354e-06, "loss": 0.0086, "step": 73850 }, { "grad_norm": 0.2653753459453583, "learning_rate": 1.6023499639119754e-06, "loss": 0.0138, "step": 73860 }, { "grad_norm": 0.24344265460968018, "learning_rate": 1.5971636212573138e-06, "loss": 0.0099, "step": 73870 }, { "grad_norm": 0.2862207293510437, "learning_rate": 1.5919855493235714e-06, "loss": 0.0106, "step": 73880 }, { "grad_norm": 0.16679391264915466, "learning_rate": 1.586815748995535e-06, "loss": 0.0089, "step": 73890 }, { "grad_norm": 0.3859945833683014, "learning_rate": 1.5816542211565866e-06, "loss": 0.0152, "step": 73900 }, { "grad_norm": 0.24556207656860352, "learning_rate": 1.5765009666886765e-06, "loss": 0.0121, "step": 73910 }, { "grad_norm": 0.39972493052482605, "learning_rate": 1.5713559864723781e-06, "loss": 0.0147, "step": 73920 }, { "grad_norm": 0.16034749150276184, "learning_rate": 1.5662192813868104e-06, "loss": 0.0117, "step": 73930 }, { "grad_norm": 0.21737511456012726, "learning_rate": 1.561090852309699e-06, "loss": 0.0176, "step": 73940 }, { "grad_norm": 0.19749437272548676, "learning_rate": 1.5559707001173651e-06, "loss": 0.0084, "step": 73950 }, { "grad_norm": 0.41181156039237976, "learning_rate": 1.5508588256846757e-06, "loss": 0.0143, "step": 73960 }, { "grad_norm": 0.19927474856376648, "learning_rate": 1.5457552298851319e-06, "loss": 0.0137, "step": 73970 }, { "grad_norm": 0.19328252971172333, "learning_rate": 1.5406599135907918e-06, "loss": 0.008, "step": 73980 }, { "grad_norm": 0.47287312150001526, "learning_rate": 1.5355728776723088e-06, "loss": 0.0101, "step": 73990 }, { "grad_norm": 0.19078348577022552, "learning_rate": 1.5304941229989155e-06, "loss": 0.012, "step": 74000 }, { "grad_norm": 0.25109758973121643, "learning_rate": 1.5254236504384345e-06, "loss": 0.0125, "step": 74010 }, { "grad_norm": 0.17849600315093994, "learning_rate": 1.5203614608572726e-06, "loss": 0.0137, "step": 74020 }, { "grad_norm": 0.22170238196849823, "learning_rate": 1.5153075551204044e-06, "loss": 0.0112, "step": 74030 }, { "grad_norm": 0.2323610633611679, "learning_rate": 1.5102619340914225e-06, "loss": 0.0224, "step": 74040 }, { "grad_norm": 0.47469958662986755, "learning_rate": 1.505224598632482e-06, "loss": 0.0107, "step": 74050 }, { "grad_norm": 0.19995816051959991, "learning_rate": 1.5001955496043164e-06, "loss": 0.0162, "step": 74060 }, { "grad_norm": 0.3183620870113373, "learning_rate": 1.4951747878662604e-06, "loss": 0.0136, "step": 74070 }, { "grad_norm": 0.14239172637462616, "learning_rate": 1.4901623142762221e-06, "loss": 0.0111, "step": 74080 }, { "grad_norm": 0.22613035142421722, "learning_rate": 1.4851581296907001e-06, "loss": 0.0158, "step": 74090 }, { "grad_norm": 0.19559264183044434, "learning_rate": 1.4801622349647714e-06, "loss": 0.0082, "step": 74100 }, { "grad_norm": 0.18389691412448883, "learning_rate": 1.4751746309520976e-06, "loss": 0.0145, "step": 74110 }, { "grad_norm": 0.14875230193138123, "learning_rate": 1.4701953185049132e-06, "loss": 0.0155, "step": 74120 }, { "grad_norm": 0.17407318949699402, "learning_rate": 1.4652242984740661e-06, "loss": 0.017, "step": 74130 }, { "grad_norm": 0.17800936102867126, "learning_rate": 1.4602615717089484e-06, "loss": 0.0159, "step": 74140 }, { "grad_norm": 0.2299906611442566, "learning_rate": 1.4553071390575657e-06, "loss": 0.0149, "step": 74150 }, { "grad_norm": 0.17905603349208832, "learning_rate": 1.450361001366496e-06, "loss": 0.0159, "step": 74160 }, { "grad_norm": 0.2791375517845154, "learning_rate": 1.4454231594809021e-06, "loss": 0.014, "step": 74170 }, { "grad_norm": 0.9051436185836792, "learning_rate": 1.4404936142445036e-06, "loss": 0.0123, "step": 74180 }, { "grad_norm": 0.22863414883613586, "learning_rate": 1.4355723664996546e-06, "loss": 0.01, "step": 74190 }, { "grad_norm": 0.28136566281318665, "learning_rate": 1.4306594170872433e-06, "loss": 0.0125, "step": 74200 }, { "grad_norm": 0.27569863200187683, "learning_rate": 1.4257547668467598e-06, "loss": 0.0175, "step": 74210 }, { "grad_norm": 0.487872451543808, "learning_rate": 1.4208584166162886e-06, "loss": 0.0113, "step": 74220 }, { "grad_norm": 0.2691513001918793, "learning_rate": 1.4159703672324553e-06, "loss": 0.0119, "step": 74230 }, { "grad_norm": 0.1508644074201584, "learning_rate": 1.4110906195305252e-06, "loss": 0.0129, "step": 74240 }, { "grad_norm": 0.24537047743797302, "learning_rate": 1.406219174344292e-06, "loss": 0.0131, "step": 74250 }, { "grad_norm": 0.1722390204668045, "learning_rate": 1.401356032506157e-06, "loss": 0.0126, "step": 74260 }, { "grad_norm": 0.1505410075187683, "learning_rate": 1.3965011948471051e-06, "loss": 0.0181, "step": 74270 }, { "grad_norm": 0.33986571431159973, "learning_rate": 1.3916546621966842e-06, "loss": 0.013, "step": 74280 }, { "grad_norm": 0.1549524962902069, "learning_rate": 1.3868164353830482e-06, "loss": 0.0104, "step": 74290 }, { "grad_norm": 0.4077055752277374, "learning_rate": 1.3819865152329082e-06, "loss": 0.0103, "step": 74300 }, { "grad_norm": 0.1420901119709015, "learning_rate": 1.377164902571565e-06, "loss": 0.0179, "step": 74310 }, { "grad_norm": 0.2534139156341553, "learning_rate": 1.372351598222904e-06, "loss": 0.0132, "step": 74320 }, { "grad_norm": 0.1979922652244568, "learning_rate": 1.3675466030093843e-06, "loss": 0.0103, "step": 74330 }, { "grad_norm": 0.18348875641822815, "learning_rate": 1.3627499177520486e-06, "loss": 0.0158, "step": 74340 }, { "grad_norm": 0.23457852005958557, "learning_rate": 1.3579615432705196e-06, "loss": 0.0103, "step": 74350 }, { "grad_norm": 0.4070093035697937, "learning_rate": 1.3531814803829978e-06, "loss": 0.0154, "step": 74360 }, { "grad_norm": 0.19129392504692078, "learning_rate": 1.3484097299062803e-06, "loss": 0.0092, "step": 74370 }, { "grad_norm": 0.08324958384037018, "learning_rate": 1.343646292655698e-06, "loss": 0.0104, "step": 74380 }, { "grad_norm": 0.24880178272724152, "learning_rate": 1.3388911694452277e-06, "loss": 0.0143, "step": 74390 }, { "grad_norm": 0.5420093536376953, "learning_rate": 1.3341443610873638e-06, "loss": 0.0175, "step": 74400 }, { "grad_norm": 0.2871691584587097, "learning_rate": 1.329405868393213e-06, "loss": 0.0105, "step": 74410 }, { "grad_norm": 0.23509728908538818, "learning_rate": 1.3246756921724613e-06, "loss": 0.0157, "step": 74420 }, { "grad_norm": 0.16524794697761536, "learning_rate": 1.3199538332333506e-06, "loss": 0.012, "step": 74430 }, { "grad_norm": 0.5912845134735107, "learning_rate": 1.3152402923827411e-06, "loss": 0.015, "step": 74440 }, { "grad_norm": 0.20883412659168243, "learning_rate": 1.3105350704260277e-06, "loss": 0.0117, "step": 74450 }, { "grad_norm": 0.2717360258102417, "learning_rate": 1.305838168167206e-06, "loss": 0.0099, "step": 74460 }, { "grad_norm": 0.24784930050373077, "learning_rate": 1.301149586408862e-06, "loss": 0.0087, "step": 74470 }, { "grad_norm": 0.160552516579628, "learning_rate": 1.2964693259521322e-06, "loss": 0.0149, "step": 74480 }, { "grad_norm": 0.39506471157073975, "learning_rate": 1.2917973875967548e-06, "loss": 0.0135, "step": 74490 }, { "grad_norm": 0.3083910048007965, "learning_rate": 1.2871337721410249e-06, "loss": 0.0123, "step": 74500 }, { "grad_norm": 0.20257486402988434, "learning_rate": 1.2824784803818379e-06, "loss": 0.0135, "step": 74510 }, { "grad_norm": 0.5428654551506042, "learning_rate": 1.2778315131146467e-06, "loss": 0.009, "step": 74520 }, { "grad_norm": 0.2895706593990326, "learning_rate": 1.2731928711334994e-06, "loss": 0.0107, "step": 74530 }, { "grad_norm": 0.2226068377494812, "learning_rate": 1.2685625552310122e-06, "loss": 0.0099, "step": 74540 }, { "grad_norm": 0.2075204849243164, "learning_rate": 1.263940566198374e-06, "loss": 0.0108, "step": 74550 }, { "grad_norm": 0.19901543855667114, "learning_rate": 1.259326904825353e-06, "loss": 0.0127, "step": 74560 }, { "grad_norm": 0.333135187625885, "learning_rate": 1.2547215719003137e-06, "loss": 0.0128, "step": 74570 }, { "grad_norm": 0.26256105303764343, "learning_rate": 1.2501245682101703e-06, "loss": 0.0115, "step": 74580 }, { "grad_norm": 0.31691864132881165, "learning_rate": 1.2455358945404171e-06, "loss": 0.0094, "step": 74590 }, { "grad_norm": 0.07472057640552521, "learning_rate": 1.2409555516751493e-06, "loss": 0.0087, "step": 74600 }, { "grad_norm": 0.23307295143604279, "learning_rate": 1.2363835403970125e-06, "loss": 0.0123, "step": 74610 }, { "grad_norm": 0.2905156910419464, "learning_rate": 1.2318198614872489e-06, "loss": 0.0148, "step": 74620 }, { "grad_norm": 0.31061312556266785, "learning_rate": 1.2272645157256457e-06, "loss": 0.011, "step": 74630 }, { "grad_norm": 0.2524862289428711, "learning_rate": 1.222717503890608e-06, "loss": 0.0184, "step": 74640 }, { "grad_norm": 0.49955621361732483, "learning_rate": 1.218178826759081e-06, "loss": 0.0148, "step": 74650 }, { "grad_norm": 0.39935845136642456, "learning_rate": 1.213648485106611e-06, "loss": 0.0146, "step": 74660 }, { "grad_norm": 0.17213919758796692, "learning_rate": 1.2091264797073066e-06, "loss": 0.0129, "step": 74670 }, { "grad_norm": 0.27885350584983826, "learning_rate": 1.2046128113338494e-06, "loss": 0.0256, "step": 74680 }, { "grad_norm": 0.6933188438415527, "learning_rate": 1.200107480757512e-06, "loss": 0.0083, "step": 74690 }, { "grad_norm": 0.11697094142436981, "learning_rate": 1.1956104887481168e-06, "loss": 0.0196, "step": 74700 }, { "grad_norm": 0.2851586937904358, "learning_rate": 1.191121836074094e-06, "loss": 0.0101, "step": 74710 }, { "grad_norm": 0.17284397780895233, "learning_rate": 1.1866415235024186e-06, "loss": 0.0099, "step": 74720 }, { "grad_norm": 0.23113806545734406, "learning_rate": 1.182169551798662e-06, "loss": 0.0148, "step": 74730 }, { "grad_norm": 0.2368156611919403, "learning_rate": 1.177705921726957e-06, "loss": 0.0124, "step": 74740 }, { "grad_norm": 0.09470243752002716, "learning_rate": 1.173250634050016e-06, "loss": 0.0102, "step": 74750 }, { "grad_norm": 0.2074500471353531, "learning_rate": 1.1688036895291354e-06, "loss": 0.0118, "step": 74760 }, { "grad_norm": 0.27795153856277466, "learning_rate": 1.1643650889241574e-06, "loss": 0.015, "step": 74770 }, { "grad_norm": 0.2261592149734497, "learning_rate": 1.1599348329935311e-06, "loss": 0.0108, "step": 74780 }, { "grad_norm": 0.17405755817890167, "learning_rate": 1.1555129224942673e-06, "loss": 0.019, "step": 74790 }, { "grad_norm": 0.17305231094360352, "learning_rate": 1.1510993581819396e-06, "loss": 0.0164, "step": 74800 }, { "grad_norm": 0.15764112770557404, "learning_rate": 1.1466941408107112e-06, "loss": 0.0216, "step": 74810 }, { "grad_norm": 0.2602249085903168, "learning_rate": 1.1422972711333247e-06, "loss": 0.0102, "step": 74820 }, { "grad_norm": 0.22267308831214905, "learning_rate": 1.1379087499010565e-06, "loss": 0.0099, "step": 74830 }, { "grad_norm": 0.22514104843139648, "learning_rate": 1.1335285778638127e-06, "loss": 0.0099, "step": 74840 }, { "grad_norm": 0.21576115489006042, "learning_rate": 1.1291567557700333e-06, "loss": 0.0164, "step": 74850 }, { "grad_norm": 0.24592743813991547, "learning_rate": 1.124793284366743e-06, "loss": 0.0138, "step": 74860 }, { "grad_norm": 0.7551577091217041, "learning_rate": 1.1204381643995455e-06, "loss": 0.0116, "step": 74870 }, { "grad_norm": 0.25285661220550537, "learning_rate": 1.116091396612595e-06, "loss": 0.0142, "step": 74880 }, { "grad_norm": 0.14162246882915497, "learning_rate": 1.1117529817486538e-06, "loss": 0.0132, "step": 74890 }, { "grad_norm": 0.1672276258468628, "learning_rate": 1.107422920549034e-06, "loss": 0.01, "step": 74900 }, { "grad_norm": 0.2990128695964813, "learning_rate": 1.1031012137536157e-06, "loss": 0.0136, "step": 74910 }, { "grad_norm": 0.3468335270881653, "learning_rate": 1.0987878621008695e-06, "loss": 0.0135, "step": 74920 }, { "grad_norm": 0.31545490026474, "learning_rate": 1.0944828663278284e-06, "loss": 0.0111, "step": 74930 }, { "grad_norm": 0.11012784391641617, "learning_rate": 1.0901862271700925e-06, "loss": 0.0132, "step": 74940 }, { "grad_norm": 0.20585238933563232, "learning_rate": 1.0858979453618467e-06, "loss": 0.0119, "step": 74950 }, { "grad_norm": 0.4460451304912567, "learning_rate": 1.0816180216358441e-06, "loss": 0.0136, "step": 74960 }, { "grad_norm": 0.3408544361591339, "learning_rate": 1.0773464567233937e-06, "loss": 0.0122, "step": 74970 }, { "grad_norm": 0.3066015839576721, "learning_rate": 1.0730832513543953e-06, "loss": 0.0124, "step": 74980 }, { "grad_norm": 0.3073086142539978, "learning_rate": 1.0688284062573218e-06, "loss": 0.0162, "step": 74990 }, { "grad_norm": 0.20716510713100433, "learning_rate": 1.0645819221591969e-06, "loss": 0.0086, "step": 75000 }, { "grad_norm": 0.3107036352157593, "learning_rate": 1.0603437997856347e-06, "loss": 0.0135, "step": 75010 }, { "grad_norm": 0.3022039830684662, "learning_rate": 1.0561140398608228e-06, "loss": 0.0129, "step": 75020 }, { "grad_norm": 0.4652179479598999, "learning_rate": 1.0518926431074937e-06, "loss": 0.0116, "step": 75030 }, { "grad_norm": 0.18307934701442719, "learning_rate": 1.0476796102469877e-06, "loss": 0.0132, "step": 75040 }, { "grad_norm": 0.17445045709609985, "learning_rate": 1.0434749419991786e-06, "loss": 0.0116, "step": 75050 }, { "grad_norm": 0.25498446822166443, "learning_rate": 1.0392786390825415e-06, "loss": 0.0113, "step": 75060 }, { "grad_norm": 0.30568283796310425, "learning_rate": 1.0350907022141087e-06, "loss": 0.0164, "step": 75070 }, { "grad_norm": 0.35859906673431396, "learning_rate": 1.0309111321094744e-06, "loss": 0.014, "step": 75080 }, { "grad_norm": 0.223048597574234, "learning_rate": 1.026739929482834e-06, "loss": 0.0096, "step": 75090 }, { "grad_norm": 0.4140911400318146, "learning_rate": 1.022577095046906e-06, "loss": 0.0137, "step": 75100 }, { "grad_norm": 0.2212071418762207, "learning_rate": 1.0184226295130217e-06, "loss": 0.0092, "step": 75110 }, { "grad_norm": 0.33385875821113586, "learning_rate": 1.0142765335910575e-06, "loss": 0.0127, "step": 75120 }, { "grad_norm": 0.24350211024284363, "learning_rate": 1.0101388079894746e-06, "loss": 0.012, "step": 75130 }, { "grad_norm": 0.18933327496051788, "learning_rate": 1.0060094534152908e-06, "loss": 0.0124, "step": 75140 }, { "grad_norm": 0.26997196674346924, "learning_rate": 1.0018884705741028e-06, "loss": 0.0111, "step": 75150 }, { "grad_norm": 0.19916300475597382, "learning_rate": 9.977758601700805e-07, "loss": 0.0175, "step": 75160 }, { "grad_norm": 0.2152506411075592, "learning_rate": 9.93671622905945e-07, "loss": 0.0208, "step": 75170 }, { "grad_norm": 0.24377627670764923, "learning_rate": 9.895757594830024e-07, "loss": 0.0113, "step": 75180 }, { "grad_norm": 0.2637578248977661, "learning_rate": 9.854882706011204e-07, "loss": 0.0099, "step": 75190 }, { "grad_norm": 0.40947791934013367, "learning_rate": 9.814091569587513e-07, "loss": 0.0104, "step": 75200 }, { "grad_norm": 0.27702972292900085, "learning_rate": 9.77338419252888e-07, "loss": 0.0195, "step": 75210 }, { "grad_norm": 0.342197060585022, "learning_rate": 9.732760581791289e-07, "loss": 0.0139, "step": 75220 }, { "grad_norm": 0.2083951234817505, "learning_rate": 9.692220744315972e-07, "loss": 0.0109, "step": 75230 }, { "grad_norm": 0.25153836607933044, "learning_rate": 9.651764687030162e-07, "loss": 0.0115, "step": 75240 }, { "grad_norm": 0.17285755276679993, "learning_rate": 9.611392416846776e-07, "loss": 0.0116, "step": 75250 }, { "grad_norm": 0.20103107392787933, "learning_rate": 9.571103940664238e-07, "loss": 0.0131, "step": 75260 }, { "grad_norm": 0.22617550194263458, "learning_rate": 9.530899265366822e-07, "loss": 0.0127, "step": 75270 }, { "grad_norm": 0.22116297483444214, "learning_rate": 9.490778397824307e-07, "loss": 0.0116, "step": 75280 }, { "grad_norm": 0.28986984491348267, "learning_rate": 9.450741344892322e-07, "loss": 0.0106, "step": 75290 }, { "grad_norm": 0.32535552978515625, "learning_rate": 9.410788113412116e-07, "loss": 0.0106, "step": 75300 }, { "grad_norm": 1.4755841493606567, "learning_rate": 9.370918710210563e-07, "loss": 0.0108, "step": 75310 }, { "grad_norm": 0.19214732944965363, "learning_rate": 9.331133142100268e-07, "loss": 0.0116, "step": 75320 }, { "grad_norm": 0.20082420110702515, "learning_rate": 9.291431415879459e-07, "loss": 0.0093, "step": 75330 }, { "grad_norm": 0.3450695276260376, "learning_rate": 9.251813538332155e-07, "loss": 0.0153, "step": 75340 }, { "grad_norm": 0.13027998805046082, "learning_rate": 9.212279516227884e-07, "loss": 0.0098, "step": 75350 }, { "grad_norm": 0.4078802466392517, "learning_rate": 9.172829356322021e-07, "loss": 0.0102, "step": 75360 }, { "grad_norm": 0.3780941367149353, "learning_rate": 9.133463065355452e-07, "loss": 0.0119, "step": 75370 }, { "grad_norm": 0.30473029613494873, "learning_rate": 9.094180650054796e-07, "loss": 0.013, "step": 75380 }, { "grad_norm": 0.34778547286987305, "learning_rate": 9.054982117132404e-07, "loss": 0.0098, "step": 75390 }, { "grad_norm": 0.2797555923461914, "learning_rate": 9.015867473286143e-07, "loss": 0.0126, "step": 75400 }, { "grad_norm": 0.27753186225891113, "learning_rate": 8.976836725199778e-07, "loss": 0.0116, "step": 75410 }, { "grad_norm": 0.22333981096744537, "learning_rate": 8.937889879542416e-07, "loss": 0.0114, "step": 75420 }, { "grad_norm": 0.49413296580314636, "learning_rate": 8.899026942969068e-07, "loss": 0.0118, "step": 75430 }, { "grad_norm": 0.2827043831348419, "learning_rate": 8.860247922120424e-07, "loss": 0.013, "step": 75440 }, { "grad_norm": 0.2655622065067291, "learning_rate": 8.821552823622737e-07, "loss": 0.0113, "step": 75450 }, { "grad_norm": 0.09551899135112762, "learning_rate": 8.782941654087884e-07, "loss": 0.0111, "step": 75460 }, { "grad_norm": 0.5229438543319702, "learning_rate": 8.744414420113478e-07, "loss": 0.0152, "step": 75470 }, { "grad_norm": 0.15395280718803406, "learning_rate": 8.705971128282753e-07, "loss": 0.0098, "step": 75480 }, { "grad_norm": 0.0802658274769783, "learning_rate": 8.66761178516473e-07, "loss": 0.0068, "step": 75490 }, { "grad_norm": 0.21065330505371094, "learning_rate": 8.629336397313781e-07, "loss": 0.0096, "step": 75500 }, { "grad_norm": 0.2467866688966751, "learning_rate": 8.591144971270282e-07, "loss": 0.0113, "step": 75510 }, { "grad_norm": 0.12196194380521774, "learning_rate": 8.553037513560069e-07, "loss": 0.0088, "step": 75520 }, { "grad_norm": 0.29683467745780945, "learning_rate": 8.515014030694546e-07, "loss": 0.0107, "step": 75530 }, { "grad_norm": 0.15685635805130005, "learning_rate": 8.47707452917107e-07, "loss": 0.0121, "step": 75540 }, { "grad_norm": 0.2623763680458069, "learning_rate": 8.439219015472343e-07, "loss": 0.0132, "step": 75550 }, { "grad_norm": 0.1765582114458084, "learning_rate": 8.401447496066861e-07, "loss": 0.0143, "step": 75560 }, { "grad_norm": 0.30239176750183105, "learning_rate": 8.363759977408792e-07, "loss": 0.0102, "step": 75570 }, { "grad_norm": 0.16804084181785583, "learning_rate": 8.326156465937817e-07, "loss": 0.0115, "step": 75580 }, { "grad_norm": 0.5296583771705627, "learning_rate": 8.28863696807941e-07, "loss": 0.0174, "step": 75590 }, { "grad_norm": 0.11333145201206207, "learning_rate": 8.251201490244609e-07, "loss": 0.0106, "step": 75600 }, { "grad_norm": 0.32752251625061035, "learning_rate": 8.213850038830129e-07, "loss": 0.0097, "step": 75610 }, { "grad_norm": 0.15320676565170288, "learning_rate": 8.176582620218309e-07, "loss": 0.0124, "step": 75620 }, { "grad_norm": 0.24259993433952332, "learning_rate": 8.139399240777057e-07, "loss": 0.0129, "step": 75630 }, { "grad_norm": 0.30965346097946167, "learning_rate": 8.102299906860122e-07, "loss": 0.0152, "step": 75640 }, { "grad_norm": 0.39760392904281616, "learning_rate": 8.065284624806657e-07, "loss": 0.012, "step": 75650 }, { "grad_norm": 0.24725914001464844, "learning_rate": 8.028353400941601e-07, "loss": 0.0124, "step": 75660 }, { "grad_norm": 0.21535958349704742, "learning_rate": 7.991506241575575e-07, "loss": 0.0124, "step": 75670 }, { "grad_norm": 0.3802390694618225, "learning_rate": 7.954743153004541e-07, "loss": 0.0151, "step": 75680 }, { "grad_norm": 0.25872114300727844, "learning_rate": 7.918064141510528e-07, "loss": 0.0114, "step": 75690 }, { "grad_norm": 0.22050869464874268, "learning_rate": 7.881469213360859e-07, "loss": 0.0127, "step": 75700 }, { "grad_norm": 0.3232828378677368, "learning_rate": 7.844958374808642e-07, "loss": 0.0134, "step": 75710 }, { "grad_norm": 0.5880516171455383, "learning_rate": 7.808531632092608e-07, "loss": 0.0092, "step": 75720 }, { "grad_norm": 0.251056045293808, "learning_rate": 7.772188991436946e-07, "loss": 0.0108, "step": 75730 }, { "grad_norm": 0.2377086728811264, "learning_rate": 7.735930459051799e-07, "loss": 0.0114, "step": 75740 }, { "grad_norm": 0.2831280529499054, "learning_rate": 7.699756041132655e-07, "loss": 0.0135, "step": 75750 }, { "grad_norm": 0.21738918125629425, "learning_rate": 7.663665743860793e-07, "loss": 0.011, "step": 75760 }, { "grad_norm": 0.2606866657733917, "learning_rate": 7.627659573403001e-07, "loss": 0.0122, "step": 75770 }, { "grad_norm": 0.18491940200328827, "learning_rate": 7.591737535911802e-07, "loss": 0.0097, "step": 75780 }, { "grad_norm": 0.31088942289352417, "learning_rate": 7.555899637525288e-07, "loss": 0.0123, "step": 75790 }, { "grad_norm": 0.3087916672229767, "learning_rate": 7.520145884367058e-07, "loss": 0.0113, "step": 75800 }, { "grad_norm": 0.07231105118989944, "learning_rate": 7.484476282546615e-07, "loss": 0.0118, "step": 75810 }, { "grad_norm": 0.32966476678848267, "learning_rate": 7.448890838158806e-07, "loss": 0.0159, "step": 75820 }, { "grad_norm": 0.34180283546447754, "learning_rate": 7.413389557284267e-07, "loss": 0.0116, "step": 75830 }, { "grad_norm": 0.29499202966690063, "learning_rate": 7.377972445989201e-07, "loss": 0.012, "step": 75840 }, { "grad_norm": 0.19406451284885406, "learning_rate": 7.342639510325377e-07, "loss": 0.0134, "step": 75850 }, { "grad_norm": 0.370279997587204, "learning_rate": 7.307390756330246e-07, "loss": 0.0148, "step": 75860 }, { "grad_norm": 0.3760235905647278, "learning_rate": 7.272226190026876e-07, "loss": 0.01, "step": 75870 }, { "grad_norm": 0.1776365041732788, "learning_rate": 7.237145817423907e-07, "loss": 0.0111, "step": 75880 }, { "grad_norm": 0.6843975186347961, "learning_rate": 7.202149644515654e-07, "loss": 0.0124, "step": 75890 }, { "grad_norm": 0.2988341450691223, "learning_rate": 7.167237677281946e-07, "loss": 0.0127, "step": 75900 }, { "grad_norm": 0.22736825048923492, "learning_rate": 7.132409921688288e-07, "loss": 0.0092, "step": 75910 }, { "grad_norm": 0.2997974455356598, "learning_rate": 7.09766638368592e-07, "loss": 0.0086, "step": 75920 }, { "grad_norm": 0.10264194756746292, "learning_rate": 7.063007069211313e-07, "loss": 0.0131, "step": 75930 }, { "grad_norm": 0.3288814425468445, "learning_rate": 7.028431984187067e-07, "loss": 0.0095, "step": 75940 }, { "grad_norm": 0.18744508922100067, "learning_rate": 6.993941134520898e-07, "loss": 0.0139, "step": 75950 }, { "grad_norm": 0.10734272748231888, "learning_rate": 6.959534526106537e-07, "loss": 0.0113, "step": 75960 }, { "grad_norm": 0.2504228353500366, "learning_rate": 6.925212164822948e-07, "loss": 0.011, "step": 75970 }, { "grad_norm": 0.2254820615053177, "learning_rate": 6.890974056535049e-07, "loss": 0.0148, "step": 75980 }, { "grad_norm": 0.21388570964336395, "learning_rate": 6.856820207093106e-07, "loss": 0.0112, "step": 75990 }, { "grad_norm": 0.2236836850643158, "learning_rate": 6.822750622333063e-07, "loss": 0.0098, "step": 76000 }, { "grad_norm": 0.3718152940273285, "learning_rate": 6.788765308076539e-07, "loss": 0.0164, "step": 76010 }, { "grad_norm": 0.21626518666744232, "learning_rate": 6.754864270130668e-07, "loss": 0.0115, "step": 76020 }, { "grad_norm": 0.373188853263855, "learning_rate": 6.721047514288203e-07, "loss": 0.011, "step": 76030 }, { "grad_norm": 0.25736117362976074, "learning_rate": 6.687315046327469e-07, "loss": 0.0128, "step": 76040 }, { "grad_norm": 0.3008553981781006, "learning_rate": 6.65366687201252e-07, "loss": 0.0135, "step": 76050 }, { "grad_norm": 0.35147297382354736, "learning_rate": 6.620102997092814e-07, "loss": 0.0126, "step": 76060 }, { "grad_norm": 0.22074121236801147, "learning_rate": 6.586623427303596e-07, "loss": 0.0096, "step": 76070 }, { "grad_norm": 0.07799699157476425, "learning_rate": 6.553228168365455e-07, "loss": 0.0134, "step": 76080 }, { "grad_norm": 0.39331895112991333, "learning_rate": 6.519917225984884e-07, "loss": 0.0184, "step": 76090 }, { "grad_norm": 0.3299313485622406, "learning_rate": 6.486690605853718e-07, "loss": 0.0127, "step": 76100 }, { "grad_norm": 0.30948925018310547, "learning_rate": 6.453548313649527e-07, "loss": 0.0119, "step": 76110 }, { "grad_norm": 0.3971824645996094, "learning_rate": 6.420490355035446e-07, "loss": 0.0116, "step": 76120 }, { "grad_norm": 0.21495427191257477, "learning_rate": 6.387516735660071e-07, "loss": 0.0125, "step": 76130 }, { "grad_norm": 0.46019676327705383, "learning_rate": 6.35462746115778e-07, "loss": 0.0107, "step": 76140 }, { "grad_norm": 0.12471312284469604, "learning_rate": 6.321822537148358e-07, "loss": 0.0124, "step": 76150 }, { "grad_norm": 0.2517455816268921, "learning_rate": 6.289101969237432e-07, "loss": 0.0136, "step": 76160 }, { "grad_norm": 0.42349836230278015, "learning_rate": 6.256465763015918e-07, "loss": 0.0126, "step": 76170 }, { "grad_norm": 0.2686823308467865, "learning_rate": 6.22391392406052e-07, "loss": 0.0113, "step": 76180 }, { "grad_norm": 0.1839926689863205, "learning_rate": 6.191446457933403e-07, "loss": 0.0082, "step": 76190 }, { "grad_norm": 0.8375608921051025, "learning_rate": 6.159063370182406e-07, "loss": 0.0121, "step": 76200 }, { "grad_norm": 0.25003254413604736, "learning_rate": 6.126764666340879e-07, "loss": 0.0103, "step": 76210 }, { "grad_norm": 0.33156514167785645, "learning_rate": 6.094550351927852e-07, "loss": 0.0121, "step": 76220 }, { "grad_norm": 0.33405202627182007, "learning_rate": 6.062420432447757e-07, "loss": 0.0131, "step": 76230 }, { "grad_norm": 0.17551955580711365, "learning_rate": 6.030374913390813e-07, "loss": 0.011, "step": 76240 }, { "grad_norm": 0.7246965169906616, "learning_rate": 5.998413800232694e-07, "loss": 0.0129, "step": 76250 }, { "grad_norm": 0.4521164000034332, "learning_rate": 5.966537098434755e-07, "loss": 0.0104, "step": 76260 }, { "grad_norm": 0.3147049844264984, "learning_rate": 5.934744813443694e-07, "loss": 0.0117, "step": 76270 }, { "grad_norm": 0.29024332761764526, "learning_rate": 5.903036950692054e-07, "loss": 0.0114, "step": 76280 }, { "grad_norm": 0.2604461908340454, "learning_rate": 5.871413515597835e-07, "loss": 0.0173, "step": 76290 }, { "grad_norm": 0.2341592013835907, "learning_rate": 5.839874513564547e-07, "loss": 0.008, "step": 76300 }, { "grad_norm": 0.17248405516147614, "learning_rate": 5.808419949981436e-07, "loss": 0.0133, "step": 76310 }, { "grad_norm": 0.2108277827501297, "learning_rate": 5.777049830223257e-07, "loss": 0.0111, "step": 76320 }, { "grad_norm": 0.35412001609802246, "learning_rate": 5.745764159650114e-07, "loss": 0.0104, "step": 76330 }, { "grad_norm": 0.27228817343711853, "learning_rate": 5.71456294360806e-07, "loss": 0.0119, "step": 76340 }, { "grad_norm": 0.3573414981365204, "learning_rate": 5.683446187428443e-07, "loss": 0.0148, "step": 76350 }, { "grad_norm": 0.26722514629364014, "learning_rate": 5.652413896428288e-07, "loss": 0.0103, "step": 76360 }, { "grad_norm": 0.28694719076156616, "learning_rate": 5.621466075910131e-07, "loss": 0.0114, "step": 76370 }, { "grad_norm": 0.14085376262664795, "learning_rate": 5.590602731162187e-07, "loss": 0.0123, "step": 76380 }, { "grad_norm": 0.29393306374549866, "learning_rate": 5.559823867458125e-07, "loss": 0.0098, "step": 76390 }, { "grad_norm": 0.17018190026283264, "learning_rate": 5.529129490057128e-07, "loss": 0.0162, "step": 76400 }, { "grad_norm": 0.21864740550518036, "learning_rate": 5.498519604204167e-07, "loss": 0.0128, "step": 76410 }, { "grad_norm": 0.36686161160469055, "learning_rate": 5.467994215129501e-07, "loss": 0.0092, "step": 76420 }, { "grad_norm": 0.36765849590301514, "learning_rate": 5.437553328049183e-07, "loss": 0.0123, "step": 76430 }, { "grad_norm": 0.2046944499015808, "learning_rate": 5.407196948164661e-07, "loss": 0.0111, "step": 76440 }, { "grad_norm": 0.186876580119133, "learning_rate": 5.376925080663009e-07, "loss": 0.0135, "step": 76450 }, { "grad_norm": 0.22444921731948853, "learning_rate": 5.346737730716977e-07, "loss": 0.0131, "step": 76460 }, { "grad_norm": 0.17694570124149323, "learning_rate": 5.316634903484607e-07, "loss": 0.0084, "step": 76470 }, { "grad_norm": 0.7243052124977112, "learning_rate": 5.286616604109728e-07, "loss": 0.0093, "step": 76480 }, { "grad_norm": 0.23330873250961304, "learning_rate": 5.256682837721627e-07, "loss": 0.016, "step": 76490 }, { "grad_norm": 0.07868429273366928, "learning_rate": 5.226833609435156e-07, "loss": 0.0099, "step": 76500 }, { "grad_norm": 0.17442545294761658, "learning_rate": 5.19706892435079e-07, "loss": 0.014, "step": 76510 }, { "grad_norm": 0.13743799924850464, "learning_rate": 5.167388787554406e-07, "loss": 0.0103, "step": 76520 }, { "grad_norm": 0.15544967353343964, "learning_rate": 5.137793204117614e-07, "loss": 0.0099, "step": 76530 }, { "grad_norm": 0.3573435842990875, "learning_rate": 5.108282179097479e-07, "loss": 0.0129, "step": 76540 }, { "grad_norm": 0.2496735006570816, "learning_rate": 5.078855717536523e-07, "loss": 0.0115, "step": 76550 }, { "grad_norm": 0.2719821035861969, "learning_rate": 5.049513824463059e-07, "loss": 0.0094, "step": 76560 }, { "grad_norm": 0.29158228635787964, "learning_rate": 5.020256504890742e-07, "loss": 0.0132, "step": 76570 }, { "grad_norm": 0.20356103777885437, "learning_rate": 4.991083763818849e-07, "loss": 0.0131, "step": 76580 }, { "grad_norm": 0.6699511408805847, "learning_rate": 4.961995606232228e-07, "loss": 0.0131, "step": 76590 }, { "grad_norm": 0.15085317194461823, "learning_rate": 4.932992037101236e-07, "loss": 0.0098, "step": 76600 }, { "grad_norm": 0.4795602858066559, "learning_rate": 4.904073061381798e-07, "loss": 0.0158, "step": 76610 }, { "grad_norm": 0.1580697000026703, "learning_rate": 4.875238684015349e-07, "loss": 0.0105, "step": 76620 }, { "grad_norm": 0.29067865014076233, "learning_rate": 4.846488909928948e-07, "loss": 0.0088, "step": 76630 }, { "grad_norm": 0.15885162353515625, "learning_rate": 4.81782374403511e-07, "loss": 0.0141, "step": 76640 }, { "grad_norm": 0.2232382446527481, "learning_rate": 4.789243191231918e-07, "loss": 0.0091, "step": 76650 }, { "grad_norm": 0.37723079323768616, "learning_rate": 4.76074725640302e-07, "loss": 0.0167, "step": 76660 }, { "grad_norm": 0.21946346759796143, "learning_rate": 4.732335944417632e-07, "loss": 0.0164, "step": 76670 }, { "grad_norm": 0.23597484827041626, "learning_rate": 4.704009260130371e-07, "loss": 0.0169, "step": 76680 }, { "grad_norm": 0.1746000498533249, "learning_rate": 4.675767208381587e-07, "loss": 0.0092, "step": 76690 }, { "grad_norm": 0.20811696350574493, "learning_rate": 4.647609793997032e-07, "loss": 0.0104, "step": 76700 }, { "grad_norm": 0.21127884089946747, "learning_rate": 4.619537021788023e-07, "loss": 0.0109, "step": 76710 }, { "grad_norm": 0.23694516718387604, "learning_rate": 4.5915488965515005e-07, "loss": 0.0119, "step": 76720 }, { "grad_norm": 0.22912459075450897, "learning_rate": 4.563645423069807e-07, "loss": 0.0121, "step": 76730 }, { "grad_norm": 0.3353300988674164, "learning_rate": 4.535826606110849e-07, "loss": 0.013, "step": 76740 }, { "grad_norm": 0.38536328077316284, "learning_rate": 4.508092450428214e-07, "loss": 0.0153, "step": 76750 }, { "grad_norm": 0.187288299202919, "learning_rate": 4.480442960760778e-07, "loss": 0.0112, "step": 76760 }, { "grad_norm": 0.17235355079174042, "learning_rate": 4.4528781418332053e-07, "loss": 0.0087, "step": 76770 }, { "grad_norm": 0.3423316776752472, "learning_rate": 4.4253979983555073e-07, "loss": 0.0146, "step": 76780 }, { "grad_norm": 0.15998223423957825, "learning_rate": 4.3980025350233154e-07, "loss": 0.0093, "step": 76790 }, { "grad_norm": 0.21271894872188568, "learning_rate": 4.370691756517664e-07, "loss": 0.0121, "step": 76800 }, { "grad_norm": 0.20349851250648499, "learning_rate": 4.3434656675053753e-07, "loss": 0.0143, "step": 76810 }, { "grad_norm": 0.27225005626678467, "learning_rate": 4.3163242726385613e-07, "loss": 0.0134, "step": 76820 }, { "grad_norm": 0.19706283509731293, "learning_rate": 4.289267576554956e-07, "loss": 0.0149, "step": 76830 }, { "grad_norm": 0.227415069937706, "learning_rate": 4.262295583877807e-07, "loss": 0.0114, "step": 76840 }, { "grad_norm": 0.08963766694068909, "learning_rate": 4.235408299215815e-07, "loss": 0.017, "step": 76850 }, { "grad_norm": 0.22255848348140717, "learning_rate": 4.2086057271634173e-07, "loss": 0.0132, "step": 76860 }, { "grad_norm": 0.1924043893814087, "learning_rate": 4.181887872300394e-07, "loss": 0.014, "step": 76870 }, { "grad_norm": 0.2766725420951843, "learning_rate": 4.155254739192038e-07, "loss": 0.0098, "step": 76880 }, { "grad_norm": 0.20631282031536102, "learning_rate": 4.128706332389265e-07, "loss": 0.0127, "step": 76890 }, { "grad_norm": 0.36979928612709045, "learning_rate": 4.1022426564284453e-07, "loss": 0.0155, "step": 76900 }, { "grad_norm": 0.198560893535614, "learning_rate": 4.075863715831574e-07, "loss": 0.0129, "step": 76910 }, { "grad_norm": 0.30324098467826843, "learning_rate": 4.0495695151059887e-07, "loss": 0.0147, "step": 76920 }, { "grad_norm": 0.5382579565048218, "learning_rate": 4.023360058744763e-07, "loss": 0.0103, "step": 76930 }, { "grad_norm": 0.1965634673833847, "learning_rate": 3.997235351226314e-07, "loss": 0.0095, "step": 76940 }, { "grad_norm": 0.2584383189678192, "learning_rate": 3.971195397014571e-07, "loss": 0.014, "step": 76950 }, { "grad_norm": 0.2587873339653015, "learning_rate": 3.9452402005591417e-07, "loss": 0.0148, "step": 76960 }, { "grad_norm": 0.19390344619750977, "learning_rate": 3.9193697662950334e-07, "loss": 0.0104, "step": 76970 }, { "grad_norm": 0.5295302271842957, "learning_rate": 3.893584098642822e-07, "loss": 0.0114, "step": 76980 }, { "grad_norm": 0.17218151688575745, "learning_rate": 3.867883202008538e-07, "loss": 0.0159, "step": 76990 }, { "grad_norm": 0.2075888216495514, "learning_rate": 3.842267080783779e-07, "loss": 0.0095, "step": 77000 }, { "grad_norm": 0.16453512012958527, "learning_rate": 3.816735739345656e-07, "loss": 0.0112, "step": 77010 }, { "grad_norm": 0.1006728932261467, "learning_rate": 3.791289182056679e-07, "loss": 0.01, "step": 77020 }, { "grad_norm": 0.22876323759555817, "learning_rate": 3.7659274132650913e-07, "loss": 0.0121, "step": 77030 }, { "grad_norm": 0.23990841209888458, "learning_rate": 3.7406504373044826e-07, "loss": 0.0097, "step": 77040 }, { "grad_norm": 0.14977344870567322, "learning_rate": 3.7154582584939533e-07, "loss": 0.0103, "step": 77050 }, { "grad_norm": 0.2355661243200302, "learning_rate": 3.690350881138227e-07, "loss": 0.0124, "step": 77060 }, { "grad_norm": 0.3820171654224396, "learning_rate": 3.665328309527427e-07, "loss": 0.0115, "step": 77070 }, { "grad_norm": 0.3106326758861542, "learning_rate": 3.640390547937245e-07, "loss": 0.0141, "step": 77080 }, { "grad_norm": 0.29538601636886597, "learning_rate": 3.615537600628882e-07, "loss": 0.0147, "step": 77090 }, { "grad_norm": 0.270663857460022, "learning_rate": 3.5907694718489427e-07, "loss": 0.0123, "step": 77100 }, { "grad_norm": 0.284674733877182, "learning_rate": 3.566086165829707e-07, "loss": 0.017, "step": 77110 }, { "grad_norm": 0.4789116382598877, "learning_rate": 3.5414876867888024e-07, "loss": 0.012, "step": 77120 }, { "grad_norm": 0.28775033354759216, "learning_rate": 3.5169740389295326e-07, "loss": 0.0158, "step": 77130 }, { "grad_norm": 0.19634605944156647, "learning_rate": 3.4925452264405466e-07, "loss": 0.0119, "step": 77140 }, { "grad_norm": 0.20606523752212524, "learning_rate": 3.468201253496062e-07, "loss": 0.011, "step": 77150 }, { "grad_norm": 0.2988036870956421, "learning_rate": 3.443942124255861e-07, "loss": 0.0141, "step": 77160 }, { "grad_norm": 0.31264033913612366, "learning_rate": 3.4197678428650183e-07, "loss": 0.0145, "step": 77170 }, { "grad_norm": 0.2103724479675293, "learning_rate": 3.3956784134544504e-07, "loss": 0.0129, "step": 77180 }, { "grad_norm": 0.23713742196559906, "learning_rate": 3.3716738401402547e-07, "loss": 0.0114, "step": 77190 }, { "grad_norm": 0.2554042637348175, "learning_rate": 3.3477541270241495e-07, "loss": 0.01, "step": 77200 }, { "grad_norm": 0.20873938500881195, "learning_rate": 3.3239192781934215e-07, "loss": 0.0092, "step": 77210 }, { "grad_norm": 0.1450725495815277, "learning_rate": 3.3001692977207563e-07, "loss": 0.0107, "step": 77220 }, { "grad_norm": 0.12229887396097183, "learning_rate": 3.276504189664409e-07, "loss": 0.01, "step": 77230 }, { "grad_norm": 0.21784645318984985, "learning_rate": 3.252923958068088e-07, "loss": 0.0131, "step": 77240 }, { "grad_norm": 0.21076390147209167, "learning_rate": 3.2294286069609046e-07, "loss": 0.012, "step": 77250 }, { "grad_norm": 0.2530565857887268, "learning_rate": 3.2060181403577583e-07, "loss": 0.0107, "step": 77260 }, { "grad_norm": 0.3102262616157532, "learning_rate": 3.1826925622587823e-07, "loss": 0.0102, "step": 77270 }, { "grad_norm": 0.19990263879299164, "learning_rate": 3.159451876649622e-07, "loss": 0.0137, "step": 77280 }, { "grad_norm": 0.17481869459152222, "learning_rate": 3.1362960875015444e-07, "loss": 0.0087, "step": 77290 }, { "grad_norm": 0.34294649958610535, "learning_rate": 3.1132251987711637e-07, "loss": 0.0126, "step": 77300 }, { "grad_norm": 0.2670419216156006, "learning_rate": 3.0902392144007695e-07, "loss": 0.0105, "step": 77310 }, { "grad_norm": 0.1629924327135086, "learning_rate": 3.0673381383179436e-07, "loss": 0.0118, "step": 77320 }, { "grad_norm": 0.2670247554779053, "learning_rate": 3.0445219744358875e-07, "loss": 0.0111, "step": 77330 }, { "grad_norm": 0.16288040578365326, "learning_rate": 3.021790726653262e-07, "loss": 0.0161, "step": 77340 }, { "grad_norm": 0.3542883098125458, "learning_rate": 2.9991443988542366e-07, "loss": 0.0139, "step": 77350 }, { "grad_norm": 0.4131176471710205, "learning_rate": 2.976582994908439e-07, "loss": 0.0104, "step": 77360 }, { "grad_norm": 0.17418727278709412, "learning_rate": 2.954106518671007e-07, "loss": 0.0096, "step": 77370 }, { "grad_norm": 0.2433803677558899, "learning_rate": 2.9317149739825356e-07, "loss": 0.0141, "step": 77380 }, { "grad_norm": 0.49296262860298157, "learning_rate": 2.909408364669075e-07, "loss": 0.0151, "step": 77390 }, { "grad_norm": 0.3390159010887146, "learning_rate": 2.887186694542299e-07, "loss": 0.0134, "step": 77400 }, { "grad_norm": 0.1543547660112381, "learning_rate": 2.865049967399225e-07, "loss": 0.011, "step": 77410 }, { "grad_norm": 0.17373822629451752, "learning_rate": 2.842998187022439e-07, "loss": 0.0115, "step": 77420 }, { "grad_norm": 0.20523656904697418, "learning_rate": 2.8210313571800374e-07, "loss": 0.0125, "step": 77430 }, { "grad_norm": 0.23814968764781952, "learning_rate": 2.7991494816255184e-07, "loss": 0.0096, "step": 77440 }, { "grad_norm": 0.24674898386001587, "learning_rate": 2.777352564097779e-07, "loss": 0.013, "step": 77450 }, { "grad_norm": 0.22305381298065186, "learning_rate": 2.755640608321508e-07, "loss": 0.0117, "step": 77460 }, { "grad_norm": 0.26146337389945984, "learning_rate": 2.73401361800657e-07, "loss": 0.0115, "step": 77470 }, { "grad_norm": 0.26351869106292725, "learning_rate": 2.7124715968484537e-07, "loss": 0.0122, "step": 77480 }, { "grad_norm": 0.2903648614883423, "learning_rate": 2.691014548528104e-07, "loss": 0.0151, "step": 77490 }, { "grad_norm": 0.27308985590934753, "learning_rate": 2.669642476711864e-07, "loss": 0.0157, "step": 77500 }, { "grad_norm": 0.2814517319202423, "learning_rate": 2.648355385051815e-07, "loss": 0.0122, "step": 77510 }, { "grad_norm": 0.5041233897209167, "learning_rate": 2.627153277185157e-07, "loss": 0.0128, "step": 77520 }, { "grad_norm": 0.13906626403331757, "learning_rate": 2.606036156734881e-07, "loss": 0.0126, "step": 77530 }, { "grad_norm": 0.09558157622814178, "learning_rate": 2.5850040273092127e-07, "loss": 0.0119, "step": 77540 }, { "grad_norm": 0.2232598215341568, "learning_rate": 2.5640568925020536e-07, "loss": 0.0208, "step": 77550 }, { "grad_norm": 0.19294360280036926, "learning_rate": 2.5431947558927083e-07, "loss": 0.012, "step": 77560 }, { "grad_norm": 0.20837239921092987, "learning_rate": 2.522417621045825e-07, "loss": 0.0101, "step": 77570 }, { "grad_norm": 0.20929524302482605, "learning_rate": 2.501725491511786e-07, "loss": 0.0158, "step": 77580 }, { "grad_norm": 0.214542418718338, "learning_rate": 2.481118370826263e-07, "loss": 0.0145, "step": 77590 }, { "grad_norm": 0.20117716491222382, "learning_rate": 2.4605962625104393e-07, "loss": 0.0113, "step": 77600 }, { "grad_norm": 0.37691548466682434, "learning_rate": 2.440159170070955e-07, "loss": 0.0154, "step": 77610 }, { "grad_norm": 0.33895477652549744, "learning_rate": 2.41980709699996e-07, "loss": 0.0125, "step": 77620 }, { "grad_norm": 0.17230980098247528, "learning_rate": 2.3995400467751174e-07, "loss": 0.0122, "step": 77630 }, { "grad_norm": 0.19291497766971588, "learning_rate": 2.3793580228594902e-07, "loss": 0.016, "step": 77640 }, { "grad_norm": 0.23358121514320374, "learning_rate": 2.3592610287015982e-07, "loss": 0.0094, "step": 77650 }, { "grad_norm": 0.163704514503479, "learning_rate": 2.3392490677354718e-07, "loss": 0.0104, "step": 77660 }, { "grad_norm": 0.2463969737291336, "learning_rate": 2.3193221433806533e-07, "loss": 0.009, "step": 77670 }, { "grad_norm": 0.2362125664949417, "learning_rate": 2.299480259042086e-07, "loss": 0.013, "step": 77680 }, { "grad_norm": 0.2700558006763458, "learning_rate": 2.2797234181102244e-07, "loss": 0.0152, "step": 77690 }, { "grad_norm": 0.16955842077732086, "learning_rate": 2.2600516239609238e-07, "loss": 0.01, "step": 77700 }, { "grad_norm": 0.22670283913612366, "learning_rate": 2.240464879955606e-07, "loss": 0.011, "step": 77710 }, { "grad_norm": 0.3039676249027252, "learning_rate": 2.2209631894410387e-07, "loss": 0.0131, "step": 77720 }, { "grad_norm": 0.25818249583244324, "learning_rate": 2.2015465557496117e-07, "loss": 0.0128, "step": 77730 }, { "grad_norm": 0.3136031925678253, "learning_rate": 2.1822149821990602e-07, "loss": 0.015, "step": 77740 }, { "grad_norm": 0.39737388491630554, "learning_rate": 2.1629684720926303e-07, "loss": 0.0166, "step": 77750 }, { "grad_norm": 0.20035618543624878, "learning_rate": 2.1438070287189693e-07, "loss": 0.0101, "step": 77760 }, { "grad_norm": 0.18436218798160553, "learning_rate": 2.1247306553523472e-07, "loss": 0.013, "step": 77770 }, { "grad_norm": 0.2333696186542511, "learning_rate": 2.105739355252323e-07, "loss": 0.01, "step": 77780 }, { "grad_norm": 0.13584725558757782, "learning_rate": 2.0868331316639678e-07, "loss": 0.0096, "step": 77790 }, { "grad_norm": 0.33545008301734924, "learning_rate": 2.0680119878179193e-07, "loss": 0.0109, "step": 77800 }, { "grad_norm": 0.38708850741386414, "learning_rate": 2.0492759269301054e-07, "loss": 0.0086, "step": 77810 }, { "grad_norm": 0.47287485003471375, "learning_rate": 2.0306249522021315e-07, "loss": 0.0115, "step": 77820 }, { "grad_norm": 0.13083548843860626, "learning_rate": 2.0120590668207816e-07, "loss": 0.0104, "step": 77830 }, { "grad_norm": 0.1416497677564621, "learning_rate": 1.993578273958574e-07, "loss": 0.0128, "step": 77840 }, { "grad_norm": 0.19237658381462097, "learning_rate": 1.975182576773371e-07, "loss": 0.0085, "step": 77850 }, { "grad_norm": 0.51931232213974, "learning_rate": 1.9568719784083812e-07, "loss": 0.0147, "step": 77860 }, { "grad_norm": 0.22801558673381805, "learning_rate": 1.9386464819924898e-07, "loss": 0.0125, "step": 77870 }, { "grad_norm": 0.10252997279167175, "learning_rate": 1.9205060906399285e-07, "loss": 0.0099, "step": 77880 }, { "grad_norm": 0.4372989535331726, "learning_rate": 1.9024508074503845e-07, "loss": 0.014, "step": 77890 }, { "grad_norm": 0.29914218187332153, "learning_rate": 1.8844806355089452e-07, "loss": 0.015, "step": 77900 }, { "grad_norm": 0.2728278338909149, "learning_rate": 1.8665955778863208e-07, "loss": 0.0146, "step": 77910 }, { "grad_norm": 0.23221349716186523, "learning_rate": 1.848795637638512e-07, "loss": 0.0116, "step": 77920 }, { "grad_norm": 0.2533069849014282, "learning_rate": 1.831080817807085e-07, "loss": 0.0138, "step": 77930 }, { "grad_norm": 0.4182930886745453, "learning_rate": 1.813451121418952e-07, "loss": 0.0206, "step": 77940 }, { "grad_norm": 0.5009698271751404, "learning_rate": 1.795906551486648e-07, "loss": 0.0112, "step": 77950 }, { "grad_norm": 0.23263132572174072, "learning_rate": 1.7784471110079414e-07, "loss": 0.0133, "step": 77960 }, { "grad_norm": 0.26524797081947327, "learning_rate": 1.7610728029662793e-07, "loss": 0.0116, "step": 77970 }, { "grad_norm": 0.1839156150817871, "learning_rate": 1.7437836303303979e-07, "loss": 0.011, "step": 77980 }, { "grad_norm": 0.13147781789302826, "learning_rate": 1.726579596054545e-07, "loss": 0.008, "step": 77990 }, { "grad_norm": 0.37589022517204285, "learning_rate": 1.7094607030784803e-07, "loss": 0.0117, "step": 78000 }, { "grad_norm": 0.18436045944690704, "learning_rate": 1.6924269543272532e-07, "loss": 0.0105, "step": 78010 }, { "grad_norm": 0.33557674288749695, "learning_rate": 1.6754783527115348e-07, "loss": 0.0141, "step": 78020 }, { "grad_norm": 0.1089787408709526, "learning_rate": 1.6586149011273422e-07, "loss": 0.0126, "step": 78030 }, { "grad_norm": 0.3381425738334656, "learning_rate": 1.6418366024562038e-07, "loss": 0.0194, "step": 78040 }, { "grad_norm": 0.24880081415176392, "learning_rate": 1.6251434595651037e-07, "loss": 0.0129, "step": 78050 }, { "grad_norm": 0.08333530277013779, "learning_rate": 1.608535475306372e-07, "loss": 0.0124, "step": 78060 }, { "grad_norm": 0.17853973805904388, "learning_rate": 1.5920126525179048e-07, "loss": 0.0083, "step": 78070 }, { "grad_norm": 0.13802453875541687, "learning_rate": 1.5755749940229435e-07, "loss": 0.0092, "step": 78080 }, { "grad_norm": 0.2545092701911926, "learning_rate": 1.559222502630353e-07, "loss": 0.011, "step": 78090 }, { "grad_norm": 0.2797316014766693, "learning_rate": 1.5429551811341757e-07, "loss": 0.0145, "step": 78100 }, { "grad_norm": 0.3311934471130371, "learning_rate": 1.5267730323141882e-07, "loss": 0.0123, "step": 78110 }, { "grad_norm": 0.24238142371177673, "learning_rate": 1.5106760589353454e-07, "loss": 0.0106, "step": 78120 }, { "grad_norm": 0.31080105900764465, "learning_rate": 1.4946642637483355e-07, "loss": 0.0102, "step": 78130 }, { "grad_norm": 0.2828293442726135, "learning_rate": 1.4787376494889703e-07, "loss": 0.015, "step": 78140 }, { "grad_norm": 0.18298225104808807, "learning_rate": 1.4628962188787955e-07, "loss": 0.0115, "step": 78150 }, { "grad_norm": 0.2816208600997925, "learning_rate": 1.4471399746247006e-07, "loss": 0.0131, "step": 78160 }, { "grad_norm": 0.15091170370578766, "learning_rate": 1.4314689194188103e-07, "loss": 0.0128, "step": 78170 }, { "grad_norm": 0.5718413591384888, "learning_rate": 1.4158830559390933e-07, "loss": 0.0111, "step": 78180 }, { "grad_norm": 0.21233806014060974, "learning_rate": 1.4003823868486422e-07, "loss": 0.0235, "step": 78190 }, { "grad_norm": 0.28241032361984253, "learning_rate": 1.3849669147960598e-07, "loss": 0.0161, "step": 78200 }, { "grad_norm": 0.2346065193414688, "learning_rate": 1.3696366424155726e-07, "loss": 0.012, "step": 78210 }, { "grad_norm": 0.507568359375, "learning_rate": 1.35439157232653e-07, "loss": 0.0204, "step": 78220 }, { "grad_norm": 0.3340911269187927, "learning_rate": 1.3392317071340144e-07, "loss": 0.012, "step": 78230 }, { "grad_norm": 0.16630207002162933, "learning_rate": 1.3241570494283984e-07, "loss": 0.0111, "step": 78240 }, { "grad_norm": 0.22288484871387482, "learning_rate": 1.3091676017855103e-07, "loss": 0.0175, "step": 78250 }, { "grad_norm": 0.26986023783683777, "learning_rate": 1.2942633667666904e-07, "loss": 0.009, "step": 78260 }, { "grad_norm": 0.2725581228733063, "learning_rate": 1.2794443469185679e-07, "loss": 0.0166, "step": 78270 }, { "grad_norm": 0.6407107710838318, "learning_rate": 1.2647105447734508e-07, "loss": 0.0102, "step": 78280 }, { "grad_norm": 0.2509458661079407, "learning_rate": 1.2500619628488254e-07, "loss": 0.0116, "step": 78290 }, { "grad_norm": 0.3528507649898529, "learning_rate": 1.2354986036477446e-07, "loss": 0.0157, "step": 78300 }, { "grad_norm": 0.6127824783325195, "learning_rate": 1.221020469658718e-07, "loss": 0.015, "step": 78310 }, { "grad_norm": 0.33068621158599854, "learning_rate": 1.2066275633556556e-07, "loss": 0.0136, "step": 78320 }, { "grad_norm": 0.36836016178131104, "learning_rate": 1.192319887197979e-07, "loss": 0.0122, "step": 78330 }, { "grad_norm": 0.21372054517269135, "learning_rate": 1.1780974436303438e-07, "loss": 0.0095, "step": 78340 }, { "grad_norm": 0.18069668114185333, "learning_rate": 1.1639602350830836e-07, "loss": 0.0116, "step": 78350 }, { "grad_norm": 0.26598337292671204, "learning_rate": 1.1499082639718217e-07, "loss": 0.0111, "step": 78360 }, { "grad_norm": 0.16822311282157898, "learning_rate": 1.1359415326976374e-07, "loss": 0.0111, "step": 78370 }, { "grad_norm": 0.15673723816871643, "learning_rate": 1.1220600436470663e-07, "loss": 0.0119, "step": 78380 }, { "grad_norm": 0.2626444697380066, "learning_rate": 1.1082637991920996e-07, "loss": 0.0142, "step": 78390 }, { "grad_norm": 0.16060492396354675, "learning_rate": 1.0945528016901851e-07, "loss": 0.0118, "step": 78400 }, { "grad_norm": 0.3804877698421478, "learning_rate": 1.0809270534840599e-07, "loss": 0.01, "step": 78410 }, { "grad_norm": 0.1457984298467636, "learning_rate": 1.067386556902028e-07, "loss": 0.0097, "step": 78420 }, { "grad_norm": 0.13197508454322815, "learning_rate": 1.05393131425785e-07, "loss": 0.0069, "step": 78430 }, { "grad_norm": 0.3641049563884735, "learning_rate": 1.0405613278505199e-07, "loss": 0.0116, "step": 78440 }, { "grad_norm": 0.3236944079399109, "learning_rate": 1.027276599964766e-07, "loss": 0.0125, "step": 78450 }, { "grad_norm": 0.2364392876625061, "learning_rate": 1.0140771328704391e-07, "loss": 0.0111, "step": 78460 }, { "grad_norm": 0.25030046701431274, "learning_rate": 1.0009629288231237e-07, "loss": 0.0075, "step": 78470 }, { "grad_norm": 0.33153969049453735, "learning_rate": 9.879339900635276e-08, "loss": 0.0135, "step": 78480 }, { "grad_norm": 0.33213549852371216, "learning_rate": 9.74990318817981e-08, "loss": 0.0125, "step": 78490 }, { "grad_norm": 0.3712794780731201, "learning_rate": 9.621319172982701e-08, "loss": 0.0123, "step": 78500 }, { "grad_norm": 0.2348291426897049, "learning_rate": 9.493587877015264e-08, "loss": 0.0098, "step": 78510 }, { "grad_norm": 0.2030535787343979, "learning_rate": 9.366709322102263e-08, "loss": 0.012, "step": 78520 }, { "grad_norm": 0.2920306324958801, "learning_rate": 9.240683529924688e-08, "loss": 0.0129, "step": 78530 }, { "grad_norm": 0.3318738043308258, "learning_rate": 9.115510522016979e-08, "loss": 0.0121, "step": 78540 }, { "grad_norm": 0.2550452947616577, "learning_rate": 8.991190319767583e-08, "loss": 0.0103, "step": 78550 }, { "grad_norm": 0.13841864466667175, "learning_rate": 8.867722944419509e-08, "loss": 0.0102, "step": 78560 }, { "grad_norm": 0.1758863627910614, "learning_rate": 8.745108417069214e-08, "loss": 0.0109, "step": 78570 }, { "grad_norm": 0.2508632242679596, "learning_rate": 8.623346758669381e-08, "loss": 0.0125, "step": 78580 }, { "grad_norm": 0.36047273874282837, "learning_rate": 8.502437990025037e-08, "loss": 0.0114, "step": 78590 }, { "grad_norm": 0.21620389819145203, "learning_rate": 8.382382131795764e-08, "loss": 0.0129, "step": 78600 }, { "grad_norm": 0.22547020018100739, "learning_rate": 8.263179204496818e-08, "loss": 0.0131, "step": 78610 }, { "grad_norm": 0.37166866660118103, "learning_rate": 8.144829228496354e-08, "loss": 0.0113, "step": 78620 }, { "grad_norm": 0.30753999948501587, "learning_rate": 8.027332224016526e-08, "loss": 0.0134, "step": 78630 }, { "grad_norm": 0.255830317735672, "learning_rate": 7.910688211135164e-08, "loss": 0.0126, "step": 78640 }, { "grad_norm": 0.2929418385028839, "learning_rate": 7.794897209783546e-08, "loss": 0.0126, "step": 78650 }, { "grad_norm": 0.22109834849834442, "learning_rate": 7.679959239746403e-08, "loss": 0.0149, "step": 78660 }, { "grad_norm": 0.20021624863147736, "learning_rate": 7.565874320664135e-08, "loss": 0.0125, "step": 78670 }, { "grad_norm": 0.48667824268341064, "learning_rate": 7.45264247203059e-08, "loss": 0.0142, "step": 78680 }, { "grad_norm": 0.1568262130022049, "learning_rate": 7.340263713194184e-08, "loss": 0.0113, "step": 78690 }, { "grad_norm": 0.2145504206418991, "learning_rate": 7.228738063356777e-08, "loss": 0.0088, "step": 78700 }, { "grad_norm": 0.31400275230407715, "learning_rate": 7.118065541575903e-08, "loss": 0.0105, "step": 78710 }, { "grad_norm": 0.30579251050949097, "learning_rate": 7.008246166761435e-08, "loss": 0.0114, "step": 78720 }, { "grad_norm": 0.2044864147901535, "learning_rate": 6.899279957679472e-08, "loss": 0.0092, "step": 78730 }, { "grad_norm": 0.1516205370426178, "learning_rate": 6.791166932949011e-08, "loss": 0.012, "step": 78740 }, { "grad_norm": 0.3379724621772766, "learning_rate": 6.683907111043608e-08, "loss": 0.0095, "step": 78750 }, { "grad_norm": 0.18094712495803833, "learning_rate": 6.577500510290824e-08, "loss": 0.012, "step": 78760 }, { "grad_norm": 0.28043538331985474, "learning_rate": 6.47194714887278e-08, "loss": 0.0201, "step": 78770 }, { "grad_norm": 0.26440417766571045, "learning_rate": 6.367247044825608e-08, "loss": 0.0141, "step": 78780 }, { "grad_norm": 0.5718141794204712, "learning_rate": 6.263400216039994e-08, "loss": 0.0091, "step": 78790 }, { "grad_norm": 0.7666188478469849, "learning_rate": 6.160406680260078e-08, "loss": 0.0208, "step": 78800 }, { "grad_norm": 0.2263244390487671, "learning_rate": 6.058266455084561e-08, "loss": 0.0117, "step": 78810 }, { "grad_norm": 0.2650584280490875, "learning_rate": 5.956979557967257e-08, "loss": 0.0124, "step": 78820 }, { "grad_norm": 0.09183287620544434, "learning_rate": 5.856546006214325e-08, "loss": 0.0116, "step": 78830 }, { "grad_norm": 0.22520025074481964, "learning_rate": 5.7569658169881466e-08, "loss": 0.0108, "step": 78840 }, { "grad_norm": 0.15156199038028717, "learning_rate": 5.658239007303445e-08, "loss": 0.0146, "step": 78850 }, { "grad_norm": 0.2711886763572693, "learning_rate": 5.560365594030059e-08, "loss": 0.0145, "step": 78860 }, { "grad_norm": 0.16724829375743866, "learning_rate": 5.463345593891833e-08, "loss": 0.0085, "step": 78870 }, { "grad_norm": 0.21830196678638458, "learning_rate": 5.367179023467173e-08, "loss": 0.0134, "step": 78880 }, { "grad_norm": 0.10491058975458145, "learning_rate": 5.271865899187378e-08, "loss": 0.0109, "step": 78890 }, { "grad_norm": 0.25377944111824036, "learning_rate": 5.177406237340532e-08, "loss": 0.0187, "step": 78900 }, { "grad_norm": 0.22543293237686157, "learning_rate": 5.083800054065946e-08, "loss": 0.0106, "step": 78910 }, { "grad_norm": 0.22492723166942596, "learning_rate": 4.991047365358603e-08, "loss": 0.0115, "step": 78920 }, { "grad_norm": 0.5829906463623047, "learning_rate": 4.899148187067493e-08, "loss": 0.0143, "step": 78930 }, { "grad_norm": 0.16292504966259003, "learning_rate": 4.808102534895609e-08, "loss": 0.009, "step": 78940 }, { "grad_norm": 0.3269560933113098, "learning_rate": 4.717910424400507e-08, "loss": 0.0117, "step": 78950 }, { "grad_norm": 0.31575483083724976, "learning_rate": 4.628571870993193e-08, "loss": 0.0136, "step": 78960 }, { "grad_norm": 0.24107442796230316, "learning_rate": 4.5400868899392324e-08, "loss": 0.0135, "step": 78970 }, { "grad_norm": 0.4351900517940521, "learning_rate": 4.4524554963576436e-08, "loss": 0.0162, "step": 78980 }, { "grad_norm": 0.34357941150665283, "learning_rate": 4.36567770522367e-08, "loss": 0.0141, "step": 78990 }, { "grad_norm": 0.17860379815101624, "learning_rate": 4.279753531364339e-08, "loss": 0.0094, "step": 79000 }, { "grad_norm": 0.5136991739273071, "learning_rate": 4.1946829894617955e-08, "loss": 0.0127, "step": 79010 }, { "grad_norm": 0.2936519980430603, "learning_rate": 4.11046609405219e-08, "loss": 0.0167, "step": 79020 }, { "grad_norm": 0.27744290232658386, "learning_rate": 4.027102859526233e-08, "loss": 0.0067, "step": 79030 }, { "grad_norm": 0.46544149518013, "learning_rate": 3.9445933001280856e-08, "loss": 0.0119, "step": 79040 }, { "grad_norm": 0.19590923190116882, "learning_rate": 3.8629374299564704e-08, "loss": 0.0121, "step": 79050 }, { "grad_norm": 0.22570361196994781, "learning_rate": 3.78213526296467e-08, "loss": 0.0151, "step": 79060 }, { "grad_norm": 0.19512146711349487, "learning_rate": 3.702186812958308e-08, "loss": 0.0144, "step": 79070 }, { "grad_norm": 0.2512279748916626, "learning_rate": 3.623092093599789e-08, "loss": 0.0111, "step": 79080 }, { "grad_norm": 0.19703620672225952, "learning_rate": 3.544851118403303e-08, "loss": 0.0104, "step": 79090 }, { "grad_norm": 1.407485008239746, "learning_rate": 3.46746390073871e-08, "loss": 0.0154, "step": 79100 }, { "grad_norm": 0.20896029472351074, "learning_rate": 3.390930453829322e-08, "loss": 0.0187, "step": 79110 }, { "grad_norm": 0.07854368537664413, "learning_rate": 3.3152507907519005e-08, "loss": 0.0104, "step": 79120 }, { "grad_norm": 0.44141316413879395, "learning_rate": 3.240424924438323e-08, "loss": 0.0132, "step": 79130 }, { "grad_norm": 0.4119659960269928, "learning_rate": 3.1664528676750295e-08, "loss": 0.0122, "step": 79140 }, { "grad_norm": 0.28365838527679443, "learning_rate": 3.093334633100797e-08, "loss": 0.0153, "step": 79150 }, { "grad_norm": 0.2604682445526123, "learning_rate": 3.021070233210077e-08, "loss": 0.0131, "step": 79160 }, { "grad_norm": 0.34239259362220764, "learning_rate": 2.9496596803507692e-08, "loss": 0.0097, "step": 79170 }, { "grad_norm": 0.16335764527320862, "learning_rate": 2.879102986725335e-08, "loss": 0.0101, "step": 79180 }, { "grad_norm": 0.2195822298526764, "learning_rate": 2.809400164389131e-08, "loss": 0.0116, "step": 79190 }, { "grad_norm": 0.21240589022636414, "learning_rate": 2.740551225253185e-08, "loss": 0.0146, "step": 79200 }, { "grad_norm": 0.29899609088897705, "learning_rate": 2.6725561810819754e-08, "loss": 0.011, "step": 79210 }, { "grad_norm": 0.2601870000362396, "learning_rate": 2.605415043493431e-08, "loss": 0.0123, "step": 79220 }, { "grad_norm": 0.4408564269542694, "learning_rate": 2.5391278239605965e-08, "loss": 0.0145, "step": 79230 }, { "grad_norm": 0.1483674943447113, "learning_rate": 2.4736945338094118e-08, "loss": 0.0108, "step": 79240 }, { "grad_norm": 0.15682293474674225, "learning_rate": 2.4091151842214887e-08, "loss": 0.0091, "step": 79250 }, { "grad_norm": 0.4466910660266876, "learning_rate": 2.3453897862318884e-08, "loss": 0.0168, "step": 79260 }, { "grad_norm": 0.269436776638031, "learning_rate": 2.2825183507285686e-08, "loss": 0.0134, "step": 79270 }, { "grad_norm": 0.34575602412223816, "learning_rate": 2.220500888455157e-08, "loss": 0.0186, "step": 79280 }, { "grad_norm": 0.2586832344532013, "learning_rate": 2.1593374100081777e-08, "loss": 0.0147, "step": 79290 }, { "grad_norm": 0.17256779968738556, "learning_rate": 2.09902792583927e-08, "loss": 0.0172, "step": 79300 }, { "grad_norm": 0.15852318704128265, "learning_rate": 2.0395724462540788e-08, "loss": 0.0124, "step": 79310 }, { "grad_norm": 0.351777583360672, "learning_rate": 1.9809709814111453e-08, "loss": 0.0113, "step": 79320 }, { "grad_norm": 0.13445784151554108, "learning_rate": 1.923223541324126e-08, "loss": 0.0185, "step": 79330 }, { "grad_norm": 0.17518995702266693, "learning_rate": 1.8663301358606833e-08, "loss": 0.0149, "step": 79340 }, { "grad_norm": 0.39355704188346863, "learning_rate": 1.8102907747419295e-08, "loss": 0.0151, "step": 79350 }, { "grad_norm": 0.22993113100528717, "learning_rate": 1.7551054675435385e-08, "loss": 0.01, "step": 79360 }, { "grad_norm": 0.22346779704093933, "learning_rate": 1.7007742236957447e-08, "loss": 0.0122, "step": 79370 }, { "grad_norm": 0.25044047832489014, "learning_rate": 1.647297052481678e-08, "loss": 0.0135, "step": 79380 }, { "grad_norm": 0.18052983283996582, "learning_rate": 1.5946739630390285e-08, "loss": 0.0109, "step": 79390 }, { "grad_norm": 0.12086163461208344, "learning_rate": 1.5429049643606032e-08, "loss": 0.0142, "step": 79400 }, { "grad_norm": 0.12373796105384827, "learning_rate": 1.4919900652909935e-08, "loss": 0.0126, "step": 79410 }, { "grad_norm": 0.18670155107975006, "learning_rate": 1.4419292745310175e-08, "loss": 0.0127, "step": 79420 }, { "grad_norm": 0.14871488511562347, "learning_rate": 1.3927226006343885e-08, "loss": 0.0111, "step": 79430 }, { "grad_norm": 0.30981001257896423, "learning_rate": 1.3443700520093805e-08, "loss": 0.0126, "step": 79440 }, { "grad_norm": 0.1598019301891327, "learning_rate": 1.296871636917718e-08, "loss": 0.0077, "step": 79450 }, { "grad_norm": 0.3323567807674408, "learning_rate": 1.2502273634762419e-08, "loss": 0.0117, "step": 79460 }, { "grad_norm": 0.2460843026638031, "learning_rate": 1.2044372396546876e-08, "loss": 0.0127, "step": 79470 }, { "grad_norm": 0.1715315729379654, "learning_rate": 1.1595012732773524e-08, "loss": 0.0138, "step": 79480 }, { "grad_norm": 0.24731336534023285, "learning_rate": 1.1154194720225386e-08, "loss": 0.0138, "step": 79490 }, { "grad_norm": 0.25399938225746155, "learning_rate": 1.0721918434231093e-08, "loss": 0.0112, "step": 79500 }, { "grad_norm": 0.3717067539691925, "learning_rate": 1.0298183948648232e-08, "loss": 0.012, "step": 79510 }, { "grad_norm": 0.18323248624801636, "learning_rate": 9.88299133588555e-09, "loss": 0.0098, "step": 79520 }, { "grad_norm": 0.23444822430610657, "learning_rate": 9.476340666891847e-09, "loss": 0.0154, "step": 79530 }, { "grad_norm": 0.15298520028591156, "learning_rate": 9.078232011139332e-09, "loss": 0.0131, "step": 79540 }, { "grad_norm": 0.2610815763473511, "learning_rate": 8.68866543666802e-09, "loss": 0.0116, "step": 79550 }, { "grad_norm": 0.27691367268562317, "learning_rate": 8.307641010035782e-09, "loss": 0.0129, "step": 79560 }, { "grad_norm": 0.28327617049217224, "learning_rate": 7.93515879635165e-09, "loss": 0.0106, "step": 79570 }, { "grad_norm": 0.6908968091011047, "learning_rate": 7.571218859264706e-09, "loss": 0.01, "step": 79580 }, { "grad_norm": 0.2967697083950043, "learning_rate": 7.215821260958544e-09, "loss": 0.0122, "step": 79590 }, { "grad_norm": 0.11057543754577637, "learning_rate": 6.868966062162363e-09, "loss": 0.01, "step": 79600 }, { "grad_norm": 0.13515238463878632, "learning_rate": 6.530653322145419e-09, "loss": 0.0112, "step": 79610 }, { "grad_norm": 0.19087667763233185, "learning_rate": 6.200883098717025e-09, "loss": 0.0092, "step": 79620 }, { "grad_norm": 0.4686487019062042, "learning_rate": 5.879655448226551e-09, "loss": 0.0165, "step": 79630 }, { "grad_norm": 0.303678423166275, "learning_rate": 5.566970425557872e-09, "loss": 0.0098, "step": 79640 }, { "grad_norm": 0.198717400431633, "learning_rate": 5.26282808414047e-09, "loss": 0.0079, "step": 79650 }, { "grad_norm": 0.2615604102611542, "learning_rate": 4.967228475949437e-09, "loss": 0.0131, "step": 79660 }, { "grad_norm": 0.20281830430030823, "learning_rate": 4.680171651494369e-09, "loss": 0.012, "step": 79670 }, { "grad_norm": 0.2838912606239319, "learning_rate": 4.40165765981937e-09, "loss": 0.0119, "step": 79680 }, { "grad_norm": 0.133696511387825, "learning_rate": 4.131686548519698e-09, "loss": 0.0069, "step": 79690 }, { "grad_norm": 0.302727073431015, "learning_rate": 3.8702583637251214e-09, "loss": 0.0099, "step": 79700 }, { "grad_norm": 0.23737213015556335, "learning_rate": 3.617373150105463e-09, "loss": 0.013, "step": 79710 }, { "grad_norm": 0.2118947058916092, "learning_rate": 3.3730309508706036e-09, "loss": 0.0116, "step": 79720 }, { "grad_norm": 0.27862346172332764, "learning_rate": 3.137231807781582e-09, "loss": 0.0092, "step": 79730 }, { "grad_norm": 0.24024330079555511, "learning_rate": 2.9099757611172894e-09, "loss": 0.0113, "step": 79740 }, { "grad_norm": 0.25360921025276184, "learning_rate": 2.6912628497133275e-09, "loss": 0.0125, "step": 79750 }, { "grad_norm": 0.2244144082069397, "learning_rate": 2.481093110945354e-09, "loss": 0.0144, "step": 79760 }, { "grad_norm": 0.1850534975528717, "learning_rate": 2.2794665807235327e-09, "loss": 0.0143, "step": 79770 }, { "grad_norm": 0.1795879751443863, "learning_rate": 2.0863832934980843e-09, "loss": 0.012, "step": 79780 }, { "grad_norm": 0.7997534871101379, "learning_rate": 1.901843282264837e-09, "loss": 0.0137, "step": 79790 }, { "grad_norm": 0.1392565220594406, "learning_rate": 1.7258465785541245e-09, "loss": 0.0197, "step": 79800 }, { "grad_norm": 0.1255989372730255, "learning_rate": 1.5583932124418887e-09, "loss": 0.0131, "step": 79810 }, { "grad_norm": 0.20831599831581116, "learning_rate": 1.3994832125441282e-09, "loss": 0.0095, "step": 79820 }, { "grad_norm": 0.20354169607162476, "learning_rate": 1.2491166060057958e-09, "loss": 0.011, "step": 79830 }, { "grad_norm": 0.19483450055122375, "learning_rate": 1.1072934185230032e-09, "loss": 0.0118, "step": 79840 }, { "grad_norm": 0.2381574809551239, "learning_rate": 9.740136743319195e-10, "loss": 0.0131, "step": 79850 }, { "grad_norm": 0.2715480923652649, "learning_rate": 8.492773962087696e-10, "loss": 0.0091, "step": 79860 }, { "grad_norm": 0.31454792618751526, "learning_rate": 7.330846054587337e-10, "loss": 0.0213, "step": 79870 }, { "grad_norm": 0.3061279356479645, "learning_rate": 6.254353219492526e-10, "loss": 0.0098, "step": 79880 }, { "grad_norm": 0.17269445955753326, "learning_rate": 5.263295640600685e-10, "loss": 0.0108, "step": 79890 }, { "grad_norm": 1.2002744674682617, "learning_rate": 4.357673487387359e-10, "loss": 0.0101, "step": 79900 }, { "grad_norm": 0.21687154471874237, "learning_rate": 3.537486914506616e-10, "loss": 0.0092, "step": 79910 }, { "grad_norm": 0.19282352924346924, "learning_rate": 2.8027360621241117e-10, "loss": 0.0126, "step": 79920 }, { "grad_norm": 0.16142484545707703, "learning_rate": 2.153421055806071e-10, "loss": 0.0112, "step": 79930 }, { "grad_norm": 0.18141846358776093, "learning_rate": 1.5895420064637733e-10, "loss": 0.0144, "step": 79940 }, { "grad_norm": 0.20182771980762482, "learning_rate": 1.1110990105200891e-10, "loss": 0.0116, "step": 79950 }, { "grad_norm": 0.22515660524368286, "learning_rate": 7.180921496874326e-11, "loss": 0.0149, "step": 79960 }, { "grad_norm": 0.2784842252731323, "learning_rate": 4.10521491134297e-11, "loss": 0.0102, "step": 79970 }, { "grad_norm": 0.30907967686653137, "learning_rate": 1.883870874297422e-11, "loss": 0.012, "step": 79980 }, { "grad_norm": 0.29883912205696106, "learning_rate": 5.1688976432373584e-12, "loss": 0.0111, "step": 79990 }, { "grad_norm": 0.2104547619819641, "learning_rate": 4.271816234080461e-14, "loss": 0.0083, "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 }