{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 80000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 0.7675600051879883, "learning_rate": 2.25e-07, "loss": 1.2776, "step": 10 }, { "grad_norm": 0.6772830486297607, "learning_rate": 4.75e-07, "loss": 1.3009, "step": 20 }, { "grad_norm": 0.31328195333480835, "learning_rate": 7.25e-07, "loss": 1.2688, "step": 30 }, { "grad_norm": 0.4928935170173645, "learning_rate": 9.75e-07, "loss": 1.2302, "step": 40 }, { "grad_norm": 0.6813535690307617, "learning_rate": 1.2250000000000001e-06, "loss": 1.2516, "step": 50 }, { "grad_norm": 0.5918872356414795, "learning_rate": 1.475e-06, "loss": 1.2524, "step": 60 }, { "grad_norm": 0.8409765958786011, "learning_rate": 1.7250000000000002e-06, "loss": 1.2121, "step": 70 }, { "grad_norm": 1.0194175243377686, "learning_rate": 1.975e-06, "loss": 1.243, "step": 80 }, { "grad_norm": 0.9446123242378235, "learning_rate": 2.225e-06, "loss": 1.2196, "step": 90 }, { "grad_norm": 0.705577552318573, "learning_rate": 2.4750000000000004e-06, "loss": 1.177, "step": 100 }, { "grad_norm": 0.6493461728096008, "learning_rate": 2.725e-06, "loss": 1.2376, "step": 110 }, { "grad_norm": 0.4987231492996216, "learning_rate": 2.975e-06, "loss": 1.2397, "step": 120 }, { "grad_norm": 0.8049596548080444, "learning_rate": 3.225e-06, "loss": 1.207, "step": 130 }, { "grad_norm": 1.1114877462387085, "learning_rate": 3.4750000000000006e-06, "loss": 1.2292, "step": 140 }, { "grad_norm": 2.002180337905884, "learning_rate": 3.725e-06, "loss": 1.219, "step": 150 }, { "grad_norm": 1.2606565952301025, "learning_rate": 3.975e-06, "loss": 1.1903, "step": 160 }, { "grad_norm": 2.4981422424316406, "learning_rate": 4.225e-06, "loss": 1.1784, "step": 170 }, { "grad_norm": 1.5010380744934082, "learning_rate": 4.475e-06, "loss": 1.1619, "step": 180 }, { "grad_norm": 1.5850430727005005, "learning_rate": 4.7250000000000005e-06, "loss": 1.1376, "step": 190 }, { "grad_norm": 6.888300895690918, "learning_rate": 4.975000000000001e-06, "loss": 1.1172, "step": 200 }, { "grad_norm": 1.4503662586212158, "learning_rate": 5.225e-06, "loss": 1.0896, "step": 210 }, { "grad_norm": 1.5438742637634277, "learning_rate": 5.475e-06, "loss": 1.0675, "step": 220 }, { "grad_norm": 1.2920721769332886, "learning_rate": 5.725e-06, "loss": 1.0763, "step": 230 }, { "grad_norm": 1.136579155921936, "learning_rate": 5.975e-06, "loss": 1.0748, "step": 240 }, { "grad_norm": 3.140918016433716, "learning_rate": 6.2250000000000005e-06, "loss": 1.0727, "step": 250 }, { "grad_norm": 1.177444577217102, "learning_rate": 6.475000000000001e-06, "loss": 1.0584, "step": 260 }, { "grad_norm": 2.413207769393921, "learning_rate": 6.725000000000001e-06, "loss": 1.0698, "step": 270 }, { "grad_norm": 1.3367022275924683, "learning_rate": 6.975000000000001e-06, "loss": 1.039, "step": 280 }, { "grad_norm": 1.4792331457138062, "learning_rate": 7.2249999999999994e-06, "loss": 1.0627, "step": 290 }, { "grad_norm": 0.9747526049613953, "learning_rate": 7.4750000000000004e-06, "loss": 1.0835, "step": 300 }, { "grad_norm": 0.9525136351585388, "learning_rate": 7.725e-06, "loss": 1.064, "step": 310 }, { "grad_norm": 0.9804520606994629, "learning_rate": 7.975e-06, "loss": 1.0605, "step": 320 }, { "grad_norm": 0.9824955463409424, "learning_rate": 8.225e-06, "loss": 1.063, "step": 330 }, { "grad_norm": 0.8015322089195251, "learning_rate": 8.475000000000001e-06, "loss": 1.0516, "step": 340 }, { "grad_norm": 1.2815232276916504, "learning_rate": 8.725e-06, "loss": 1.0564, "step": 350 }, { "grad_norm": 1.1160982847213745, "learning_rate": 8.975e-06, "loss": 1.0487, "step": 360 }, { "grad_norm": 1.4402161836624146, "learning_rate": 9.225e-06, "loss": 1.0651, "step": 370 }, { "grad_norm": 1.076098918914795, "learning_rate": 9.475e-06, "loss": 1.0501, "step": 380 }, { "grad_norm": 1.1875407695770264, "learning_rate": 9.725000000000001e-06, "loss": 1.0129, "step": 390 }, { "grad_norm": 1.1221766471862793, "learning_rate": 9.975e-06, "loss": 1.0466, "step": 400 }, { "grad_norm": 1.2103142738342285, "learning_rate": 1.0225e-05, "loss": 1.0255, "step": 410 }, { "grad_norm": 1.7919456958770752, "learning_rate": 1.0475e-05, "loss": 0.9905, "step": 420 }, { "grad_norm": 1.165640950202942, "learning_rate": 1.0725e-05, "loss": 0.9618, "step": 430 }, { "grad_norm": 1.5666236877441406, "learning_rate": 1.0975e-05, "loss": 0.9602, "step": 440 }, { "grad_norm": 1.649256706237793, "learning_rate": 1.1225e-05, "loss": 0.9186, "step": 450 }, { "grad_norm": 1.8142743110656738, "learning_rate": 1.1475000000000001e-05, "loss": 0.9047, "step": 460 }, { "grad_norm": 1.5309401750564575, "learning_rate": 1.1725e-05, "loss": 0.9076, "step": 470 }, { "grad_norm": 1.476155161857605, "learning_rate": 1.1975e-05, "loss": 0.9024, "step": 480 }, { "grad_norm": 1.722969889640808, "learning_rate": 1.2225e-05, "loss": 0.8648, "step": 490 }, { "grad_norm": 1.8954607248306274, "learning_rate": 1.2475e-05, "loss": 0.834, "step": 500 }, { "grad_norm": 2.394925832748413, "learning_rate": 1.2725000000000001e-05, "loss": 0.8295, "step": 510 }, { "grad_norm": 1.6791545152664185, "learning_rate": 1.2975e-05, "loss": 0.8035, "step": 520 }, { "grad_norm": 2.007439613342285, "learning_rate": 1.3225000000000001e-05, "loss": 0.7717, "step": 530 }, { "grad_norm": 2.3799474239349365, "learning_rate": 1.3475000000000002e-05, "loss": 0.7148, "step": 540 }, { "grad_norm": 2.7959558963775635, "learning_rate": 1.3725000000000002e-05, "loss": 0.7323, "step": 550 }, { "grad_norm": 1.9085721969604492, "learning_rate": 1.3975000000000003e-05, "loss": 0.6993, "step": 560 }, { "grad_norm": 2.2951271533966064, "learning_rate": 1.4225e-05, "loss": 0.6276, "step": 570 }, { "grad_norm": 1.9560697078704834, "learning_rate": 1.4475e-05, "loss": 0.5942, "step": 580 }, { "grad_norm": 2.237274646759033, "learning_rate": 1.4725e-05, "loss": 0.5944, "step": 590 }, { "grad_norm": 2.323291301727295, "learning_rate": 1.4975e-05, "loss": 0.5707, "step": 600 }, { "grad_norm": 2.60132098197937, "learning_rate": 1.5225e-05, "loss": 0.5303, "step": 610 }, { "grad_norm": 2.4012677669525146, "learning_rate": 1.5475e-05, "loss": 0.5163, "step": 620 }, { "grad_norm": 2.1559152603149414, "learning_rate": 1.5725e-05, "loss": 0.4977, "step": 630 }, { "grad_norm": 2.601882219314575, "learning_rate": 1.5975000000000002e-05, "loss": 0.4856, "step": 640 }, { "grad_norm": 2.6874277591705322, "learning_rate": 1.6225e-05, "loss": 0.446, "step": 650 }, { "grad_norm": 3.5242936611175537, "learning_rate": 1.6475e-05, "loss": 0.4471, "step": 660 }, { "grad_norm": 2.4286625385284424, "learning_rate": 1.6725000000000003e-05, "loss": 0.3873, "step": 670 }, { "grad_norm": 3.0158653259277344, "learning_rate": 1.6975000000000003e-05, "loss": 0.3763, "step": 680 }, { "grad_norm": 2.7735908031463623, "learning_rate": 1.7225e-05, "loss": 0.3691, "step": 690 }, { "grad_norm": 2.7270636558532715, "learning_rate": 1.7475e-05, "loss": 0.3742, "step": 700 }, { "grad_norm": 2.843867778778076, "learning_rate": 1.7725e-05, "loss": 0.3193, "step": 710 }, { "grad_norm": 2.9282655715942383, "learning_rate": 1.7975e-05, "loss": 0.3402, "step": 720 }, { "grad_norm": 2.7114763259887695, "learning_rate": 1.8225e-05, "loss": 0.3065, "step": 730 }, { "grad_norm": 2.0520215034484863, "learning_rate": 1.8475000000000002e-05, "loss": 0.2687, "step": 740 }, { "grad_norm": 3.1413333415985107, "learning_rate": 1.8725e-05, "loss": 0.2861, "step": 750 }, { "grad_norm": 2.016857147216797, "learning_rate": 1.8975e-05, "loss": 0.2636, "step": 760 }, { "grad_norm": 2.021374225616455, "learning_rate": 1.9225e-05, "loss": 0.2788, "step": 770 }, { "grad_norm": 2.6315972805023193, "learning_rate": 1.9475000000000002e-05, "loss": 0.3112, "step": 780 }, { "grad_norm": 2.563811779022217, "learning_rate": 1.9725000000000002e-05, "loss": 0.2421, "step": 790 }, { "grad_norm": 2.6643569469451904, "learning_rate": 1.9975e-05, "loss": 0.2397, "step": 800 }, { "grad_norm": 2.2867729663848877, "learning_rate": 2.0225000000000004e-05, "loss": 0.2344, "step": 810 }, { "grad_norm": 2.379420757293701, "learning_rate": 2.0475e-05, "loss": 0.2354, "step": 820 }, { "grad_norm": 2.8618698120117188, "learning_rate": 2.0725e-05, "loss": 0.236, "step": 830 }, { "grad_norm": 2.137392520904541, "learning_rate": 2.0975e-05, "loss": 0.2267, "step": 840 }, { "grad_norm": 2.108530282974243, "learning_rate": 2.1225e-05, "loss": 0.2259, "step": 850 }, { "grad_norm": 2.403055191040039, "learning_rate": 2.1475e-05, "loss": 0.1979, "step": 860 }, { "grad_norm": 2.519089698791504, "learning_rate": 2.1725e-05, "loss": 0.2243, "step": 870 }, { "grad_norm": 2.8199076652526855, "learning_rate": 2.1975000000000002e-05, "loss": 0.182, "step": 880 }, { "grad_norm": 2.444859266281128, "learning_rate": 2.2225e-05, "loss": 0.1946, "step": 890 }, { "grad_norm": 2.0967862606048584, "learning_rate": 2.2475e-05, "loss": 0.1936, "step": 900 }, { "grad_norm": 1.8562750816345215, "learning_rate": 2.2725000000000003e-05, "loss": 0.1938, "step": 910 }, { "grad_norm": 2.179626703262329, "learning_rate": 2.2975000000000003e-05, "loss": 0.1709, "step": 920 }, { "grad_norm": 2.1243491172790527, "learning_rate": 2.3225000000000002e-05, "loss": 0.1781, "step": 930 }, { "grad_norm": 2.1206250190734863, "learning_rate": 2.3475e-05, "loss": 0.1841, "step": 940 }, { "grad_norm": 2.3309969902038574, "learning_rate": 2.3725e-05, "loss": 0.17, "step": 950 }, { "grad_norm": 2.5369713306427, "learning_rate": 2.3975e-05, "loss": 0.1713, "step": 960 }, { "grad_norm": 2.4073727130889893, "learning_rate": 2.4225e-05, "loss": 0.2092, "step": 970 }, { "grad_norm": 2.1116840839385986, "learning_rate": 2.4475000000000002e-05, "loss": 0.1597, "step": 980 }, { "grad_norm": 2.2320761680603027, "learning_rate": 2.4725e-05, "loss": 0.1491, "step": 990 }, { "grad_norm": 1.9845266342163086, "learning_rate": 2.4975e-05, "loss": 0.1625, "step": 1000 }, { "grad_norm": 1.6547075510025024, "learning_rate": 2.5225e-05, "loss": 0.1662, "step": 1010 }, { "grad_norm": 2.2449605464935303, "learning_rate": 2.5475e-05, "loss": 0.182, "step": 1020 }, { "grad_norm": 1.6529868841171265, "learning_rate": 2.5725e-05, "loss": 0.1449, "step": 1030 }, { "grad_norm": 1.8803585767745972, "learning_rate": 2.5974999999999998e-05, "loss": 0.1562, "step": 1040 }, { "grad_norm": 1.9913036823272705, "learning_rate": 2.6225e-05, "loss": 0.1662, "step": 1050 }, { "grad_norm": 1.8656096458435059, "learning_rate": 2.6475e-05, "loss": 0.1674, "step": 1060 }, { "grad_norm": 2.9095890522003174, "learning_rate": 2.6725e-05, "loss": 0.1719, "step": 1070 }, { "grad_norm": 2.279845714569092, "learning_rate": 2.6975000000000002e-05, "loss": 0.177, "step": 1080 }, { "grad_norm": 1.5784858465194702, "learning_rate": 2.7225e-05, "loss": 0.1615, "step": 1090 }, { "grad_norm": 2.4028334617614746, "learning_rate": 2.7475e-05, "loss": 0.129, "step": 1100 }, { "grad_norm": 2.5671961307525635, "learning_rate": 2.7725e-05, "loss": 0.1911, "step": 1110 }, { "grad_norm": 1.4068734645843506, "learning_rate": 2.7975000000000002e-05, "loss": 0.1309, "step": 1120 }, { "grad_norm": 1.5161688327789307, "learning_rate": 2.8225e-05, "loss": 0.1469, "step": 1130 }, { "grad_norm": 2.281217336654663, "learning_rate": 2.8475e-05, "loss": 0.1906, "step": 1140 }, { "grad_norm": 2.0462393760681152, "learning_rate": 2.8725e-05, "loss": 0.1893, "step": 1150 }, { "grad_norm": 1.420461654663086, "learning_rate": 2.8975000000000003e-05, "loss": 0.1394, "step": 1160 }, { "grad_norm": 1.8167310953140259, "learning_rate": 2.9225000000000002e-05, "loss": 0.1547, "step": 1170 }, { "grad_norm": 1.77003812789917, "learning_rate": 2.9475e-05, "loss": 0.1341, "step": 1180 }, { "grad_norm": 1.6916000843048096, "learning_rate": 2.9725000000000004e-05, "loss": 0.1518, "step": 1190 }, { "grad_norm": 3.955242395401001, "learning_rate": 2.9975000000000004e-05, "loss": 0.1752, "step": 1200 }, { "grad_norm": 2.46894907951355, "learning_rate": 3.0225000000000003e-05, "loss": 0.181, "step": 1210 }, { "grad_norm": 1.690334439277649, "learning_rate": 3.0475000000000002e-05, "loss": 0.1433, "step": 1220 }, { "grad_norm": 2.1591455936431885, "learning_rate": 3.0725e-05, "loss": 0.1545, "step": 1230 }, { "grad_norm": 1.79021155834198, "learning_rate": 3.0975e-05, "loss": 0.1563, "step": 1240 }, { "grad_norm": 2.165506362915039, "learning_rate": 3.122500000000001e-05, "loss": 0.1446, "step": 1250 }, { "grad_norm": 1.8591197729110718, "learning_rate": 3.1475e-05, "loss": 0.1579, "step": 1260 }, { "grad_norm": 1.64596688747406, "learning_rate": 3.1725e-05, "loss": 0.1262, "step": 1270 }, { "grad_norm": 1.776603102684021, "learning_rate": 3.1975e-05, "loss": 0.1567, "step": 1280 }, { "grad_norm": 1.7678378820419312, "learning_rate": 3.2225e-05, "loss": 0.1538, "step": 1290 }, { "grad_norm": 1.2600648403167725, "learning_rate": 3.2474999999999997e-05, "loss": 0.1446, "step": 1300 }, { "grad_norm": 1.8170685768127441, "learning_rate": 3.2725e-05, "loss": 0.1511, "step": 1310 }, { "grad_norm": 1.887209415435791, "learning_rate": 3.2975e-05, "loss": 0.1586, "step": 1320 }, { "grad_norm": 1.570766806602478, "learning_rate": 3.3225e-05, "loss": 0.1233, "step": 1330 }, { "grad_norm": 1.4861334562301636, "learning_rate": 3.3475e-05, "loss": 0.1194, "step": 1340 }, { "grad_norm": 1.722685694694519, "learning_rate": 3.3725e-05, "loss": 0.1499, "step": 1350 }, { "grad_norm": 2.061434745788574, "learning_rate": 3.3975e-05, "loss": 0.1589, "step": 1360 }, { "grad_norm": 1.8958797454833984, "learning_rate": 3.4225e-05, "loss": 0.1209, "step": 1370 }, { "grad_norm": 1.3562705516815186, "learning_rate": 3.4475000000000005e-05, "loss": 0.1173, "step": 1380 }, { "grad_norm": 2.613753080368042, "learning_rate": 3.4725000000000004e-05, "loss": 0.1481, "step": 1390 }, { "grad_norm": 1.6385668516159058, "learning_rate": 3.4975e-05, "loss": 0.1245, "step": 1400 }, { "grad_norm": 1.2770930528640747, "learning_rate": 3.5225e-05, "loss": 0.139, "step": 1410 }, { "grad_norm": 1.4020421504974365, "learning_rate": 3.5475e-05, "loss": 0.1375, "step": 1420 }, { "grad_norm": 1.787766933441162, "learning_rate": 3.5725e-05, "loss": 0.1041, "step": 1430 }, { "grad_norm": 1.4302425384521484, "learning_rate": 3.5975e-05, "loss": 0.1467, "step": 1440 }, { "grad_norm": 1.3899329900741577, "learning_rate": 3.6225000000000006e-05, "loss": 0.1497, "step": 1450 }, { "grad_norm": 1.0915272235870361, "learning_rate": 3.6475000000000006e-05, "loss": 0.1368, "step": 1460 }, { "grad_norm": 1.412078619003296, "learning_rate": 3.6725000000000005e-05, "loss": 0.1425, "step": 1470 }, { "grad_norm": 1.4270477294921875, "learning_rate": 3.6975000000000004e-05, "loss": 0.1198, "step": 1480 }, { "grad_norm": 1.0230270624160767, "learning_rate": 3.7225000000000004e-05, "loss": 0.1281, "step": 1490 }, { "grad_norm": 1.506360411643982, "learning_rate": 3.7475e-05, "loss": 0.1367, "step": 1500 }, { "grad_norm": 1.7787809371948242, "learning_rate": 3.7725e-05, "loss": 0.1279, "step": 1510 }, { "grad_norm": 1.8123857975006104, "learning_rate": 3.7975e-05, "loss": 0.1348, "step": 1520 }, { "grad_norm": 1.4654587507247925, "learning_rate": 3.8225e-05, "loss": 0.1387, "step": 1530 }, { "grad_norm": 1.5048261880874634, "learning_rate": 3.8475e-05, "loss": 0.1297, "step": 1540 }, { "grad_norm": 1.3688552379608154, "learning_rate": 3.8725e-05, "loss": 0.1166, "step": 1550 }, { "grad_norm": 1.1924787759780884, "learning_rate": 3.8975e-05, "loss": 0.1424, "step": 1560 }, { "grad_norm": 1.4346565008163452, "learning_rate": 3.9225e-05, "loss": 0.1096, "step": 1570 }, { "grad_norm": 1.9663671255111694, "learning_rate": 3.9475000000000004e-05, "loss": 0.1339, "step": 1580 }, { "grad_norm": 1.7629313468933105, "learning_rate": 3.9725e-05, "loss": 0.1493, "step": 1590 }, { "grad_norm": 1.3072853088378906, "learning_rate": 3.9975e-05, "loss": 0.1309, "step": 1600 }, { "grad_norm": 1.6652079820632935, "learning_rate": 4.0225e-05, "loss": 0.1119, "step": 1610 }, { "grad_norm": 1.0771514177322388, "learning_rate": 4.0475e-05, "loss": 0.1404, "step": 1620 }, { "grad_norm": 1.2944055795669556, "learning_rate": 4.0725e-05, "loss": 0.1486, "step": 1630 }, { "grad_norm": 1.396806240081787, "learning_rate": 4.0975e-05, "loss": 0.1289, "step": 1640 }, { "grad_norm": 1.9361907243728638, "learning_rate": 4.1225e-05, "loss": 0.1387, "step": 1650 }, { "grad_norm": 1.2679293155670166, "learning_rate": 4.1475000000000005e-05, "loss": 0.14, "step": 1660 }, { "grad_norm": 1.4898725748062134, "learning_rate": 4.1725000000000005e-05, "loss": 0.11, "step": 1670 }, { "grad_norm": 1.4128738641738892, "learning_rate": 4.1975000000000004e-05, "loss": 0.1195, "step": 1680 }, { "grad_norm": 1.0075860023498535, "learning_rate": 4.2225e-05, "loss": 0.121, "step": 1690 }, { "grad_norm": 1.0865082740783691, "learning_rate": 4.2475e-05, "loss": 0.1228, "step": 1700 }, { "grad_norm": 1.755789875984192, "learning_rate": 4.2725e-05, "loss": 0.135, "step": 1710 }, { "grad_norm": 1.7018139362335205, "learning_rate": 4.2975e-05, "loss": 0.1411, "step": 1720 }, { "grad_norm": 1.2239410877227783, "learning_rate": 4.322500000000001e-05, "loss": 0.1386, "step": 1730 }, { "grad_norm": 1.5622532367706299, "learning_rate": 4.3475000000000006e-05, "loss": 0.1263, "step": 1740 }, { "grad_norm": 1.438529372215271, "learning_rate": 4.3725000000000006e-05, "loss": 0.1377, "step": 1750 }, { "grad_norm": 1.4736679792404175, "learning_rate": 4.3975e-05, "loss": 0.1305, "step": 1760 }, { "grad_norm": 1.7162383794784546, "learning_rate": 4.4225e-05, "loss": 0.1168, "step": 1770 }, { "grad_norm": 1.386346697807312, "learning_rate": 4.4475e-05, "loss": 0.1531, "step": 1780 }, { "grad_norm": 1.613215446472168, "learning_rate": 4.4725e-05, "loss": 0.1238, "step": 1790 }, { "grad_norm": 1.3321360349655151, "learning_rate": 4.4975e-05, "loss": 0.13, "step": 1800 }, { "grad_norm": 1.112078309059143, "learning_rate": 4.5225e-05, "loss": 0.1228, "step": 1810 }, { "grad_norm": 1.7382183074951172, "learning_rate": 4.5475e-05, "loss": 0.1303, "step": 1820 }, { "grad_norm": 1.8574788570404053, "learning_rate": 4.5725e-05, "loss": 0.1352, "step": 1830 }, { "grad_norm": 1.1157501935958862, "learning_rate": 4.5975e-05, "loss": 0.1302, "step": 1840 }, { "grad_norm": 1.017736792564392, "learning_rate": 4.6225e-05, "loss": 0.1604, "step": 1850 }, { "grad_norm": 1.074082851409912, "learning_rate": 4.6475000000000005e-05, "loss": 0.1494, "step": 1860 }, { "grad_norm": 1.306754469871521, "learning_rate": 4.6725000000000004e-05, "loss": 0.1128, "step": 1870 }, { "grad_norm": 1.478721261024475, "learning_rate": 4.6975000000000003e-05, "loss": 0.0981, "step": 1880 }, { "grad_norm": 1.1686164140701294, "learning_rate": 4.7225e-05, "loss": 0.1219, "step": 1890 }, { "grad_norm": 0.9225304126739502, "learning_rate": 4.7475e-05, "loss": 0.1156, "step": 1900 }, { "grad_norm": 1.5309666395187378, "learning_rate": 4.7725e-05, "loss": 0.1423, "step": 1910 }, { "grad_norm": 1.5407756567001343, "learning_rate": 4.7975e-05, "loss": 0.1338, "step": 1920 }, { "grad_norm": 0.9958724975585938, "learning_rate": 4.822500000000001e-05, "loss": 0.1163, "step": 1930 }, { "grad_norm": 1.2308284044265747, "learning_rate": 4.8475000000000006e-05, "loss": 0.1225, "step": 1940 }, { "grad_norm": 1.0644952058792114, "learning_rate": 4.8725000000000005e-05, "loss": 0.101, "step": 1950 }, { "grad_norm": 1.589260220527649, "learning_rate": 4.8975000000000005e-05, "loss": 0.1487, "step": 1960 }, { "grad_norm": 1.3113194704055786, "learning_rate": 4.9225000000000004e-05, "loss": 0.1043, "step": 1970 }, { "grad_norm": 1.4993351697921753, "learning_rate": 4.9475e-05, "loss": 0.1312, "step": 1980 }, { "grad_norm": 1.2103297710418701, "learning_rate": 4.9725e-05, "loss": 0.1099, "step": 1990 }, { "grad_norm": 1.0125404596328735, "learning_rate": 4.9975e-05, "loss": 0.1182, "step": 2000 }, { "grad_norm": 1.3699476718902588, "learning_rate": 5.0225e-05, "loss": 0.1431, "step": 2010 }, { "grad_norm": 1.417313575744629, "learning_rate": 5.047500000000001e-05, "loss": 0.115, "step": 2020 }, { "grad_norm": 1.2006542682647705, "learning_rate": 5.0725e-05, "loss": 0.1039, "step": 2030 }, { "grad_norm": 0.9757172465324402, "learning_rate": 5.0975000000000006e-05, "loss": 0.1295, "step": 2040 }, { "grad_norm": 1.394248366355896, "learning_rate": 5.1225e-05, "loss": 0.1261, "step": 2050 }, { "grad_norm": 1.2426031827926636, "learning_rate": 5.1475000000000004e-05, "loss": 0.1096, "step": 2060 }, { "grad_norm": 0.900567352771759, "learning_rate": 5.1725000000000004e-05, "loss": 0.119, "step": 2070 }, { "grad_norm": 1.1073161363601685, "learning_rate": 5.197500000000001e-05, "loss": 0.1132, "step": 2080 }, { "grad_norm": 1.1523520946502686, "learning_rate": 5.2225e-05, "loss": 0.1288, "step": 2090 }, { "grad_norm": 1.2825157642364502, "learning_rate": 5.247500000000001e-05, "loss": 0.1222, "step": 2100 }, { "grad_norm": 1.8189921379089355, "learning_rate": 5.2725e-05, "loss": 0.1565, "step": 2110 }, { "grad_norm": 1.0641710758209229, "learning_rate": 5.297500000000001e-05, "loss": 0.1358, "step": 2120 }, { "grad_norm": 1.2935794591903687, "learning_rate": 5.3225e-05, "loss": 0.1292, "step": 2130 }, { "grad_norm": 1.0691204071044922, "learning_rate": 5.3475e-05, "loss": 0.1046, "step": 2140 }, { "grad_norm": 1.0430645942687988, "learning_rate": 5.3725000000000005e-05, "loss": 0.144, "step": 2150 }, { "grad_norm": 1.114429235458374, "learning_rate": 5.3975e-05, "loss": 0.1234, "step": 2160 }, { "grad_norm": 0.8274062275886536, "learning_rate": 5.4225000000000003e-05, "loss": 0.1107, "step": 2170 }, { "grad_norm": 1.1735920906066895, "learning_rate": 5.4474999999999996e-05, "loss": 0.1297, "step": 2180 }, { "grad_norm": 0.9730892777442932, "learning_rate": 5.4725e-05, "loss": 0.1114, "step": 2190 }, { "grad_norm": 0.8602146506309509, "learning_rate": 5.4975e-05, "loss": 0.0993, "step": 2200 }, { "grad_norm": 1.0606237649917603, "learning_rate": 5.522500000000001e-05, "loss": 0.1186, "step": 2210 }, { "grad_norm": 1.0127733945846558, "learning_rate": 5.5475e-05, "loss": 0.1158, "step": 2220 }, { "grad_norm": 1.4431134462356567, "learning_rate": 5.5725000000000006e-05, "loss": 0.1109, "step": 2230 }, { "grad_norm": 0.8849617838859558, "learning_rate": 5.5975e-05, "loss": 0.113, "step": 2240 }, { "grad_norm": 0.8608994483947754, "learning_rate": 5.6225000000000005e-05, "loss": 0.1228, "step": 2250 }, { "grad_norm": 0.8387059569358826, "learning_rate": 5.6475e-05, "loss": 0.1025, "step": 2260 }, { "grad_norm": 0.7468574047088623, "learning_rate": 5.6725e-05, "loss": 0.0982, "step": 2270 }, { "grad_norm": 0.6850477457046509, "learning_rate": 5.6975e-05, "loss": 0.1016, "step": 2280 }, { "grad_norm": 1.1858141422271729, "learning_rate": 5.722500000000001e-05, "loss": 0.1399, "step": 2290 }, { "grad_norm": 0.8706744313240051, "learning_rate": 5.7475e-05, "loss": 0.0923, "step": 2300 }, { "grad_norm": 0.9488977789878845, "learning_rate": 5.772500000000001e-05, "loss": 0.1051, "step": 2310 }, { "grad_norm": 0.875896155834198, "learning_rate": 5.7975e-05, "loss": 0.1015, "step": 2320 }, { "grad_norm": 1.0068644285202026, "learning_rate": 5.8225000000000006e-05, "loss": 0.1316, "step": 2330 }, { "grad_norm": 1.16116201877594, "learning_rate": 5.8475000000000005e-05, "loss": 0.1183, "step": 2340 }, { "grad_norm": 1.022660255432129, "learning_rate": 5.8725000000000004e-05, "loss": 0.1156, "step": 2350 }, { "grad_norm": 0.928996741771698, "learning_rate": 5.8975000000000004e-05, "loss": 0.1198, "step": 2360 }, { "grad_norm": 0.8517217636108398, "learning_rate": 5.922500000000001e-05, "loss": 0.1146, "step": 2370 }, { "grad_norm": 1.0135302543640137, "learning_rate": 5.9475e-05, "loss": 0.1155, "step": 2380 }, { "grad_norm": 1.1076533794403076, "learning_rate": 5.9724999999999995e-05, "loss": 0.1237, "step": 2390 }, { "grad_norm": 0.8993136882781982, "learning_rate": 5.9975e-05, "loss": 0.1179, "step": 2400 }, { "grad_norm": 0.8759363293647766, "learning_rate": 6.0225e-05, "loss": 0.1138, "step": 2410 }, { "grad_norm": 0.9100046157836914, "learning_rate": 6.0475000000000006e-05, "loss": 0.1579, "step": 2420 }, { "grad_norm": 1.0316320657730103, "learning_rate": 6.0725e-05, "loss": 0.1253, "step": 2430 }, { "grad_norm": 1.0060168504714966, "learning_rate": 6.0975000000000005e-05, "loss": 0.1287, "step": 2440 }, { "grad_norm": 1.1485779285430908, "learning_rate": 6.1225e-05, "loss": 0.1109, "step": 2450 }, { "grad_norm": 0.6616907119750977, "learning_rate": 6.1475e-05, "loss": 0.0971, "step": 2460 }, { "grad_norm": 0.8932501673698425, "learning_rate": 6.1725e-05, "loss": 0.1046, "step": 2470 }, { "grad_norm": 0.972791314125061, "learning_rate": 6.1975e-05, "loss": 0.1283, "step": 2480 }, { "grad_norm": 0.768410861492157, "learning_rate": 6.2225e-05, "loss": 0.1048, "step": 2490 }, { "grad_norm": 0.7348693013191223, "learning_rate": 6.2475e-05, "loss": 0.0936, "step": 2500 }, { "grad_norm": 1.039549708366394, "learning_rate": 6.2725e-05, "loss": 0.0992, "step": 2510 }, { "grad_norm": 1.0321340560913086, "learning_rate": 6.297500000000001e-05, "loss": 0.1094, "step": 2520 }, { "grad_norm": 0.6722991466522217, "learning_rate": 6.3225e-05, "loss": 0.0944, "step": 2530 }, { "grad_norm": 0.8916482329368591, "learning_rate": 6.347500000000001e-05, "loss": 0.0955, "step": 2540 }, { "grad_norm": 0.8066254258155823, "learning_rate": 6.3725e-05, "loss": 0.112, "step": 2550 }, { "grad_norm": 0.8698080778121948, "learning_rate": 6.397500000000001e-05, "loss": 0.1142, "step": 2560 }, { "grad_norm": 0.6542173624038696, "learning_rate": 6.4225e-05, "loss": 0.1034, "step": 2570 }, { "grad_norm": 0.8447185158729553, "learning_rate": 6.447500000000001e-05, "loss": 0.1086, "step": 2580 }, { "grad_norm": 0.8812508583068848, "learning_rate": 6.4725e-05, "loss": 0.1032, "step": 2590 }, { "grad_norm": 0.9047253727912903, "learning_rate": 6.497500000000001e-05, "loss": 0.1153, "step": 2600 }, { "grad_norm": 0.8623126149177551, "learning_rate": 6.5225e-05, "loss": 0.1117, "step": 2610 }, { "grad_norm": 0.7288796901702881, "learning_rate": 6.5475e-05, "loss": 0.0898, "step": 2620 }, { "grad_norm": 0.9343096613883972, "learning_rate": 6.5725e-05, "loss": 0.1087, "step": 2630 }, { "grad_norm": 0.934818685054779, "learning_rate": 6.5975e-05, "loss": 0.1328, "step": 2640 }, { "grad_norm": 0.9401887655258179, "learning_rate": 6.6225e-05, "loss": 0.1228, "step": 2650 }, { "grad_norm": 0.8365387320518494, "learning_rate": 6.6475e-05, "loss": 0.1217, "step": 2660 }, { "grad_norm": 0.9499730467796326, "learning_rate": 6.672500000000001e-05, "loss": 0.1252, "step": 2670 }, { "grad_norm": 1.0397710800170898, "learning_rate": 6.6975e-05, "loss": 0.1355, "step": 2680 }, { "grad_norm": 0.8052604794502258, "learning_rate": 6.722500000000001e-05, "loss": 0.1122, "step": 2690 }, { "grad_norm": 0.8090829253196716, "learning_rate": 6.7475e-05, "loss": 0.1016, "step": 2700 }, { "grad_norm": 0.655761182308197, "learning_rate": 6.7725e-05, "loss": 0.106, "step": 2710 }, { "grad_norm": 1.5561333894729614, "learning_rate": 6.7975e-05, "loss": 0.1307, "step": 2720 }, { "grad_norm": 1.0275682210922241, "learning_rate": 6.8225e-05, "loss": 0.1223, "step": 2730 }, { "grad_norm": 1.0037025213241577, "learning_rate": 6.8475e-05, "loss": 0.1019, "step": 2740 }, { "grad_norm": 0.661521852016449, "learning_rate": 6.8725e-05, "loss": 0.1167, "step": 2750 }, { "grad_norm": 1.1848423480987549, "learning_rate": 6.8975e-05, "loss": 0.1056, "step": 2760 }, { "grad_norm": 0.8000900745391846, "learning_rate": 6.9225e-05, "loss": 0.1027, "step": 2770 }, { "grad_norm": 0.8303605914115906, "learning_rate": 6.9475e-05, "loss": 0.1198, "step": 2780 }, { "grad_norm": 0.8821954131126404, "learning_rate": 6.9725e-05, "loss": 0.1142, "step": 2790 }, { "grad_norm": 0.8104286193847656, "learning_rate": 6.997500000000001e-05, "loss": 0.1053, "step": 2800 }, { "grad_norm": 0.7675477862358093, "learning_rate": 7.022500000000001e-05, "loss": 0.1482, "step": 2810 }, { "grad_norm": 1.084057092666626, "learning_rate": 7.0475e-05, "loss": 0.1263, "step": 2820 }, { "grad_norm": 0.8276844024658203, "learning_rate": 7.072500000000001e-05, "loss": 0.1011, "step": 2830 }, { "grad_norm": 0.8172222375869751, "learning_rate": 7.0975e-05, "loss": 0.1149, "step": 2840 }, { "grad_norm": 0.7680665850639343, "learning_rate": 7.122500000000001e-05, "loss": 0.1203, "step": 2850 }, { "grad_norm": 0.7451033592224121, "learning_rate": 7.1475e-05, "loss": 0.1106, "step": 2860 }, { "grad_norm": 0.6704304218292236, "learning_rate": 7.172500000000001e-05, "loss": 0.101, "step": 2870 }, { "grad_norm": 0.7190694808959961, "learning_rate": 7.1975e-05, "loss": 0.0971, "step": 2880 }, { "grad_norm": 0.793613076210022, "learning_rate": 7.2225e-05, "loss": 0.1084, "step": 2890 }, { "grad_norm": 0.7907263040542603, "learning_rate": 7.2475e-05, "loss": 0.1093, "step": 2900 }, { "grad_norm": 0.9115076065063477, "learning_rate": 7.272499999999999e-05, "loss": 0.1048, "step": 2910 }, { "grad_norm": 0.6345772743225098, "learning_rate": 7.2975e-05, "loss": 0.1112, "step": 2920 }, { "grad_norm": 0.7538565993309021, "learning_rate": 7.3225e-05, "loss": 0.0868, "step": 2930 }, { "grad_norm": 0.7238234877586365, "learning_rate": 7.347500000000001e-05, "loss": 0.0867, "step": 2940 }, { "grad_norm": 0.6430583596229553, "learning_rate": 7.3725e-05, "loss": 0.0921, "step": 2950 }, { "grad_norm": 0.702982485294342, "learning_rate": 7.397500000000001e-05, "loss": 0.1039, "step": 2960 }, { "grad_norm": 0.7095206379890442, "learning_rate": 7.4225e-05, "loss": 0.0951, "step": 2970 }, { "grad_norm": 0.910490095615387, "learning_rate": 7.447500000000001e-05, "loss": 0.098, "step": 2980 }, { "grad_norm": 0.846595823764801, "learning_rate": 7.4725e-05, "loss": 0.1141, "step": 2990 }, { "grad_norm": 0.8354681134223938, "learning_rate": 7.4975e-05, "loss": 0.1285, "step": 3000 }, { "grad_norm": 1.4702095985412598, "learning_rate": 7.5225e-05, "loss": 0.0902, "step": 3010 }, { "grad_norm": 0.6256293058395386, "learning_rate": 7.5475e-05, "loss": 0.1023, "step": 3020 }, { "grad_norm": 0.8258270621299744, "learning_rate": 7.5725e-05, "loss": 0.1057, "step": 3030 }, { "grad_norm": 0.6442311406135559, "learning_rate": 7.5975e-05, "loss": 0.1139, "step": 3040 }, { "grad_norm": 0.601428747177124, "learning_rate": 7.6225e-05, "loss": 0.1068, "step": 3050 }, { "grad_norm": 0.7803087830543518, "learning_rate": 7.6475e-05, "loss": 0.1086, "step": 3060 }, { "grad_norm": 0.6929958462715149, "learning_rate": 7.672500000000001e-05, "loss": 0.0842, "step": 3070 }, { "grad_norm": 0.7823947668075562, "learning_rate": 7.697500000000001e-05, "loss": 0.1029, "step": 3080 }, { "grad_norm": 0.9950148463249207, "learning_rate": 7.722500000000001e-05, "loss": 0.1195, "step": 3090 }, { "grad_norm": 0.8142846822738647, "learning_rate": 7.747500000000001e-05, "loss": 0.1001, "step": 3100 }, { "grad_norm": 0.7627983689308167, "learning_rate": 7.7725e-05, "loss": 0.0859, "step": 3110 }, { "grad_norm": 0.6921958923339844, "learning_rate": 7.797500000000001e-05, "loss": 0.0836, "step": 3120 }, { "grad_norm": 0.8011678457260132, "learning_rate": 7.8225e-05, "loss": 0.0938, "step": 3130 }, { "grad_norm": 0.8639916181564331, "learning_rate": 7.8475e-05, "loss": 0.1012, "step": 3140 }, { "grad_norm": 0.7544994354248047, "learning_rate": 7.8725e-05, "loss": 0.1055, "step": 3150 }, { "grad_norm": 0.6329687237739563, "learning_rate": 7.8975e-05, "loss": 0.1044, "step": 3160 }, { "grad_norm": 0.7913590669631958, "learning_rate": 7.9225e-05, "loss": 0.0945, "step": 3170 }, { "grad_norm": 0.7114991545677185, "learning_rate": 7.9475e-05, "loss": 0.1115, "step": 3180 }, { "grad_norm": 0.7477418184280396, "learning_rate": 7.9725e-05, "loss": 0.1045, "step": 3190 }, { "grad_norm": 0.613694429397583, "learning_rate": 7.9975e-05, "loss": 0.1024, "step": 3200 }, { "grad_norm": 0.5847439169883728, "learning_rate": 8.022500000000001e-05, "loss": 0.0844, "step": 3210 }, { "grad_norm": 0.956060528755188, "learning_rate": 8.0475e-05, "loss": 0.1227, "step": 3220 }, { "grad_norm": 0.5704800486564636, "learning_rate": 8.072500000000001e-05, "loss": 0.1111, "step": 3230 }, { "grad_norm": 0.6472850441932678, "learning_rate": 8.0975e-05, "loss": 0.0933, "step": 3240 }, { "grad_norm": 0.610710620880127, "learning_rate": 8.122500000000001e-05, "loss": 0.1072, "step": 3250 }, { "grad_norm": 0.9765323996543884, "learning_rate": 8.1475e-05, "loss": 0.1024, "step": 3260 }, { "grad_norm": 0.8671872019767761, "learning_rate": 8.172500000000001e-05, "loss": 0.1058, "step": 3270 }, { "grad_norm": 0.8144862651824951, "learning_rate": 8.1975e-05, "loss": 0.0885, "step": 3280 }, { "grad_norm": 0.6860392093658447, "learning_rate": 8.2225e-05, "loss": 0.0954, "step": 3290 }, { "grad_norm": 0.8487151265144348, "learning_rate": 8.2475e-05, "loss": 0.1007, "step": 3300 }, { "grad_norm": 0.7611774802207947, "learning_rate": 8.2725e-05, "loss": 0.1102, "step": 3310 }, { "grad_norm": 0.7009250521659851, "learning_rate": 8.2975e-05, "loss": 0.1251, "step": 3320 }, { "grad_norm": 0.7464485764503479, "learning_rate": 8.3225e-05, "loss": 0.1013, "step": 3330 }, { "grad_norm": 0.8103556632995605, "learning_rate": 8.347500000000001e-05, "loss": 0.0996, "step": 3340 }, { "grad_norm": 0.7274894714355469, "learning_rate": 8.3725e-05, "loss": 0.0959, "step": 3350 }, { "grad_norm": 0.9970154166221619, "learning_rate": 8.397500000000001e-05, "loss": 0.0939, "step": 3360 }, { "grad_norm": 0.5641809701919556, "learning_rate": 8.422500000000001e-05, "loss": 0.0915, "step": 3370 }, { "grad_norm": 0.9448446035385132, "learning_rate": 8.447500000000001e-05, "loss": 0.145, "step": 3380 }, { "grad_norm": 0.712092399597168, "learning_rate": 8.4725e-05, "loss": 0.1048, "step": 3390 }, { "grad_norm": 0.7765471935272217, "learning_rate": 8.4975e-05, "loss": 0.1084, "step": 3400 }, { "grad_norm": 0.7755013108253479, "learning_rate": 8.5225e-05, "loss": 0.1072, "step": 3410 }, { "grad_norm": 0.7426000833511353, "learning_rate": 8.5475e-05, "loss": 0.0868, "step": 3420 }, { "grad_norm": 0.6131981015205383, "learning_rate": 8.5725e-05, "loss": 0.0968, "step": 3430 }, { "grad_norm": 0.6354411244392395, "learning_rate": 8.5975e-05, "loss": 0.1098, "step": 3440 }, { "grad_norm": 0.7165333032608032, "learning_rate": 8.6225e-05, "loss": 0.0984, "step": 3450 }, { "grad_norm": 0.6285755038261414, "learning_rate": 8.6475e-05, "loss": 0.0868, "step": 3460 }, { "grad_norm": 0.6855711340904236, "learning_rate": 8.672500000000001e-05, "loss": 0.1155, "step": 3470 }, { "grad_norm": 0.9406072497367859, "learning_rate": 8.6975e-05, "loss": 0.1138, "step": 3480 }, { "grad_norm": 0.7464175224304199, "learning_rate": 8.7225e-05, "loss": 0.1128, "step": 3490 }, { "grad_norm": 0.6611889600753784, "learning_rate": 8.747500000000001e-05, "loss": 0.086, "step": 3500 }, { "grad_norm": 0.6006784439086914, "learning_rate": 8.7725e-05, "loss": 0.0962, "step": 3510 }, { "grad_norm": 0.887994647026062, "learning_rate": 8.797500000000001e-05, "loss": 0.0948, "step": 3520 }, { "grad_norm": 0.5820920467376709, "learning_rate": 8.8225e-05, "loss": 0.0899, "step": 3530 }, { "grad_norm": 0.6522953510284424, "learning_rate": 8.847500000000001e-05, "loss": 0.0908, "step": 3540 }, { "grad_norm": 0.6856897473335266, "learning_rate": 8.8725e-05, "loss": 0.0971, "step": 3550 }, { "grad_norm": 0.7139050960540771, "learning_rate": 8.897500000000001e-05, "loss": 0.0758, "step": 3560 }, { "grad_norm": 0.7412662506103516, "learning_rate": 8.9225e-05, "loss": 0.1012, "step": 3570 }, { "grad_norm": 0.554200291633606, "learning_rate": 8.9475e-05, "loss": 0.1147, "step": 3580 }, { "grad_norm": 0.7256154417991638, "learning_rate": 8.9725e-05, "loss": 0.1019, "step": 3590 }, { "grad_norm": 0.807249128818512, "learning_rate": 8.9975e-05, "loss": 0.0991, "step": 3600 }, { "grad_norm": 0.532363772392273, "learning_rate": 9.0225e-05, "loss": 0.0965, "step": 3610 }, { "grad_norm": 0.7083023190498352, "learning_rate": 9.0475e-05, "loss": 0.0934, "step": 3620 }, { "grad_norm": 0.6563848853111267, "learning_rate": 9.072500000000001e-05, "loss": 0.0752, "step": 3630 }, { "grad_norm": 0.6792817115783691, "learning_rate": 9.0975e-05, "loss": 0.0868, "step": 3640 }, { "grad_norm": 0.5785808563232422, "learning_rate": 9.122500000000001e-05, "loss": 0.0963, "step": 3650 }, { "grad_norm": 0.6701889634132385, "learning_rate": 9.1475e-05, "loss": 0.0944, "step": 3660 }, { "grad_norm": 0.6874712109565735, "learning_rate": 9.172500000000001e-05, "loss": 0.0972, "step": 3670 }, { "grad_norm": 0.6522179245948792, "learning_rate": 9.1975e-05, "loss": 0.1039, "step": 3680 }, { "grad_norm": 0.7178080677986145, "learning_rate": 9.2225e-05, "loss": 0.0937, "step": 3690 }, { "grad_norm": 0.6291228532791138, "learning_rate": 9.2475e-05, "loss": 0.0993, "step": 3700 }, { "grad_norm": 0.6330315470695496, "learning_rate": 9.2725e-05, "loss": 0.0866, "step": 3710 }, { "grad_norm": 0.6059632301330566, "learning_rate": 9.2975e-05, "loss": 0.0926, "step": 3720 }, { "grad_norm": 0.8025416135787964, "learning_rate": 9.3225e-05, "loss": 0.0904, "step": 3730 }, { "grad_norm": 0.6233128905296326, "learning_rate": 9.3475e-05, "loss": 0.0929, "step": 3740 }, { "grad_norm": 0.6955448389053345, "learning_rate": 9.3725e-05, "loss": 0.1029, "step": 3750 }, { "grad_norm": 0.8129780292510986, "learning_rate": 9.397500000000001e-05, "loss": 0.0844, "step": 3760 }, { "grad_norm": 0.6741318702697754, "learning_rate": 9.422500000000001e-05, "loss": 0.1026, "step": 3770 }, { "grad_norm": 0.8067159056663513, "learning_rate": 9.4475e-05, "loss": 0.1009, "step": 3780 }, { "grad_norm": 0.6419830918312073, "learning_rate": 9.472500000000001e-05, "loss": 0.0886, "step": 3790 }, { "grad_norm": 0.5897074341773987, "learning_rate": 9.4975e-05, "loss": 0.1234, "step": 3800 }, { "grad_norm": 0.6673485636711121, "learning_rate": 9.522500000000001e-05, "loss": 0.0829, "step": 3810 }, { "grad_norm": 0.823235034942627, "learning_rate": 9.5475e-05, "loss": 0.1045, "step": 3820 }, { "grad_norm": 0.667453944683075, "learning_rate": 9.572500000000001e-05, "loss": 0.0876, "step": 3830 }, { "grad_norm": 0.6508010029792786, "learning_rate": 9.5975e-05, "loss": 0.0785, "step": 3840 }, { "grad_norm": 0.6539556980133057, "learning_rate": 9.622500000000001e-05, "loss": 0.0969, "step": 3850 }, { "grad_norm": 0.7380409240722656, "learning_rate": 9.6475e-05, "loss": 0.078, "step": 3860 }, { "grad_norm": 0.6408751606941223, "learning_rate": 9.6725e-05, "loss": 0.0826, "step": 3870 }, { "grad_norm": 0.7683277130126953, "learning_rate": 9.6975e-05, "loss": 0.0826, "step": 3880 }, { "grad_norm": 0.62336665391922, "learning_rate": 9.7225e-05, "loss": 0.0981, "step": 3890 }, { "grad_norm": 0.5808391571044922, "learning_rate": 9.747500000000001e-05, "loss": 0.0937, "step": 3900 }, { "grad_norm": 0.7012321352958679, "learning_rate": 9.7725e-05, "loss": 0.0973, "step": 3910 }, { "grad_norm": 0.649345874786377, "learning_rate": 9.797500000000001e-05, "loss": 0.102, "step": 3920 }, { "grad_norm": 0.6988878846168518, "learning_rate": 9.8225e-05, "loss": 0.0905, "step": 3930 }, { "grad_norm": 0.4973021149635315, "learning_rate": 9.847500000000001e-05, "loss": 0.1033, "step": 3940 }, { "grad_norm": 0.5272164344787598, "learning_rate": 9.8725e-05, "loss": 0.0905, "step": 3950 }, { "grad_norm": 0.6633888483047485, "learning_rate": 9.897500000000001e-05, "loss": 0.0995, "step": 3960 }, { "grad_norm": 0.7624949812889099, "learning_rate": 9.9225e-05, "loss": 0.0831, "step": 3970 }, { "grad_norm": 0.5508296489715576, "learning_rate": 9.9475e-05, "loss": 0.098, "step": 3980 }, { "grad_norm": 0.6629477739334106, "learning_rate": 9.9725e-05, "loss": 0.0983, "step": 3990 }, { "grad_norm": 0.8630056381225586, "learning_rate": 9.9975e-05, "loss": 0.1006, "step": 4000 }, { "grad_norm": 0.6720333695411682, "learning_rate": 9.999999653982884e-05, "loss": 0.0898, "step": 4010 }, { "grad_norm": 0.5836352705955505, "learning_rate": 9.999998457874392e-05, "loss": 0.0905, "step": 4020 }, { "grad_norm": 0.695839524269104, "learning_rate": 9.999996407402913e-05, "loss": 0.1003, "step": 4030 }, { "grad_norm": 0.5455872416496277, "learning_rate": 9.999993502568801e-05, "loss": 0.0787, "step": 4040 }, { "grad_norm": 0.6152843832969666, "learning_rate": 9.999989743372548e-05, "loss": 0.0938, "step": 4050 }, { "grad_norm": 0.5419327616691589, "learning_rate": 9.999985129814798e-05, "loss": 0.0892, "step": 4060 }, { "grad_norm": 0.605067789554596, "learning_rate": 9.99997966189634e-05, "loss": 0.0762, "step": 4070 }, { "grad_norm": 0.7876788377761841, "learning_rate": 9.999973339618107e-05, "loss": 0.1063, "step": 4080 }, { "grad_norm": 0.633581280708313, "learning_rate": 9.999966162981179e-05, "loss": 0.0945, "step": 4090 }, { "grad_norm": 0.5121256709098816, "learning_rate": 9.999958131986784e-05, "loss": 0.0778, "step": 4100 }, { "grad_norm": 0.556123673915863, "learning_rate": 9.999949246636293e-05, "loss": 0.0869, "step": 4110 }, { "grad_norm": 0.8190159797668457, "learning_rate": 9.999939506931224e-05, "loss": 0.0938, "step": 4120 }, { "grad_norm": 0.5894370079040527, "learning_rate": 9.999928912873243e-05, "loss": 0.0832, "step": 4130 }, { "grad_norm": 0.4486235976219177, "learning_rate": 9.999917464464159e-05, "loss": 0.0856, "step": 4140 }, { "grad_norm": 0.7011795043945312, "learning_rate": 9.999905161705929e-05, "loss": 0.0934, "step": 4150 }, { "grad_norm": 0.5967922806739807, "learning_rate": 9.999892004600653e-05, "loss": 0.0844, "step": 4160 }, { "grad_norm": 0.5422819256782532, "learning_rate": 9.999877993150581e-05, "loss": 0.0829, "step": 4170 }, { "grad_norm": 0.5692723393440247, "learning_rate": 9.999863127358108e-05, "loss": 0.0747, "step": 4180 }, { "grad_norm": 0.5298850536346436, "learning_rate": 9.999847407225773e-05, "loss": 0.0844, "step": 4190 }, { "grad_norm": 0.6258960366249084, "learning_rate": 9.999830832756262e-05, "loss": 0.0778, "step": 4200 }, { "grad_norm": 0.6790006160736084, "learning_rate": 9.999813403952407e-05, "loss": 0.0847, "step": 4210 }, { "grad_norm": 0.6099565029144287, "learning_rate": 9.999795120817187e-05, "loss": 0.0991, "step": 4220 }, { "grad_norm": 0.7179709076881409, "learning_rate": 9.999775983353725e-05, "loss": 0.107, "step": 4230 }, { "grad_norm": 0.5726455450057983, "learning_rate": 9.999755991565292e-05, "loss": 0.1029, "step": 4240 }, { "grad_norm": 0.4421703815460205, "learning_rate": 9.999735145455303e-05, "loss": 0.0977, "step": 4250 }, { "grad_norm": 0.630166232585907, "learning_rate": 9.99971344502732e-05, "loss": 0.0987, "step": 4260 }, { "grad_norm": 0.4558042287826538, "learning_rate": 9.999690890285053e-05, "loss": 0.0869, "step": 4270 }, { "grad_norm": 0.5093953609466553, "learning_rate": 9.999667481232356e-05, "loss": 0.0769, "step": 4280 }, { "grad_norm": 1.2538617849349976, "learning_rate": 9.999643217873225e-05, "loss": 0.0915, "step": 4290 }, { "grad_norm": 0.49808552861213684, "learning_rate": 9.999618100211809e-05, "loss": 0.0767, "step": 4300 }, { "grad_norm": 0.5110644698143005, "learning_rate": 9.999592128252402e-05, "loss": 0.0844, "step": 4310 }, { "grad_norm": 0.6965031027793884, "learning_rate": 9.999565301999437e-05, "loss": 0.1157, "step": 4320 }, { "grad_norm": 0.5878109335899353, "learning_rate": 9.999537621457502e-05, "loss": 0.0836, "step": 4330 }, { "grad_norm": 0.5796393752098083, "learning_rate": 9.999509086631323e-05, "loss": 0.0753, "step": 4340 }, { "grad_norm": 0.7333493828773499, "learning_rate": 9.99947969752578e-05, "loss": 0.0886, "step": 4350 }, { "grad_norm": 0.6708207726478577, "learning_rate": 9.999449454145891e-05, "loss": 0.0793, "step": 4360 }, { "grad_norm": 0.5040528774261475, "learning_rate": 9.999418356496827e-05, "loss": 0.0854, "step": 4370 }, { "grad_norm": 0.7723519206047058, "learning_rate": 9.999386404583899e-05, "loss": 0.0966, "step": 4380 }, { "grad_norm": 0.7283061742782593, "learning_rate": 9.999353598412568e-05, "loss": 0.0928, "step": 4390 }, { "grad_norm": 0.6884890794754028, "learning_rate": 9.999319937988442e-05, "loss": 0.0974, "step": 4400 }, { "grad_norm": 0.5689586997032166, "learning_rate": 9.999285423317268e-05, "loss": 0.0899, "step": 4410 }, { "grad_norm": 0.5875795483589172, "learning_rate": 9.999250054404947e-05, "loss": 0.1022, "step": 4420 }, { "grad_norm": 0.6655839085578918, "learning_rate": 9.99921383125752e-05, "loss": 0.0977, "step": 4430 }, { "grad_norm": 0.6370548605918884, "learning_rate": 9.99917675388118e-05, "loss": 0.0931, "step": 4440 }, { "grad_norm": 0.6216027140617371, "learning_rate": 9.99913882228226e-05, "loss": 0.1133, "step": 4450 }, { "grad_norm": 0.5697878003120422, "learning_rate": 9.999100036467242e-05, "loss": 0.0959, "step": 4460 }, { "grad_norm": 0.6847835779190063, "learning_rate": 9.999060396442753e-05, "loss": 0.0861, "step": 4470 }, { "grad_norm": 0.5870493650436401, "learning_rate": 9.999019902215566e-05, "loss": 0.0984, "step": 4480 }, { "grad_norm": 0.5995389819145203, "learning_rate": 9.998978553792602e-05, "loss": 0.0983, "step": 4490 }, { "grad_norm": 0.7595927715301514, "learning_rate": 9.998936351180926e-05, "loss": 0.0946, "step": 4500 }, { "grad_norm": 0.5194354057312012, "learning_rate": 9.998893294387747e-05, "loss": 0.1198, "step": 4510 }, { "grad_norm": 0.6867132186889648, "learning_rate": 9.998849383420426e-05, "loss": 0.0978, "step": 4520 }, { "grad_norm": 0.6438674330711365, "learning_rate": 9.998804618286465e-05, "loss": 0.0933, "step": 4530 }, { "grad_norm": 0.6782230734825134, "learning_rate": 9.99875899899351e-05, "loss": 0.1185, "step": 4540 }, { "grad_norm": 0.5298407673835754, "learning_rate": 9.99871252554936e-05, "loss": 0.1018, "step": 4550 }, { "grad_norm": 0.7806804776191711, "learning_rate": 9.998665197961955e-05, "loss": 0.0838, "step": 4560 }, { "grad_norm": 0.6327135562896729, "learning_rate": 9.998617016239379e-05, "loss": 0.1018, "step": 4570 }, { "grad_norm": 0.951499342918396, "learning_rate": 9.998567980389869e-05, "loss": 0.0959, "step": 4580 }, { "grad_norm": 0.5042281150817871, "learning_rate": 9.998518090421802e-05, "loss": 0.0918, "step": 4590 }, { "grad_norm": 0.6619171500205994, "learning_rate": 9.998467346343703e-05, "loss": 0.0903, "step": 4600 }, { "grad_norm": 0.7358041405677795, "learning_rate": 9.998415748164243e-05, "loss": 0.0832, "step": 4610 }, { "grad_norm": 0.6261961460113525, "learning_rate": 9.998363295892238e-05, "loss": 0.0736, "step": 4620 }, { "grad_norm": 0.681090235710144, "learning_rate": 9.998309989536652e-05, "loss": 0.0893, "step": 4630 }, { "grad_norm": 0.5726740956306458, "learning_rate": 9.998255829106593e-05, "loss": 0.0919, "step": 4640 }, { "grad_norm": 0.6608253717422485, "learning_rate": 9.998200814611316e-05, "loss": 0.0789, "step": 4650 }, { "grad_norm": 0.48664048314094543, "learning_rate": 9.998144946060219e-05, "loss": 0.0852, "step": 4660 }, { "grad_norm": 0.828132152557373, "learning_rate": 9.998088223462852e-05, "loss": 0.0801, "step": 4670 }, { "grad_norm": 0.6973253488540649, "learning_rate": 9.998030646828905e-05, "loss": 0.0842, "step": 4680 }, { "grad_norm": 0.5754653215408325, "learning_rate": 9.997972216168217e-05, "loss": 0.0903, "step": 4690 }, { "grad_norm": 0.5791105628013611, "learning_rate": 9.997912931490771e-05, "loss": 0.0873, "step": 4700 }, { "grad_norm": 0.7221914529800415, "learning_rate": 9.9978527928067e-05, "loss": 0.0864, "step": 4710 }, { "grad_norm": 0.6820574402809143, "learning_rate": 9.997791800126277e-05, "loss": 0.09, "step": 4720 }, { "grad_norm": 0.675128161907196, "learning_rate": 9.997729953459927e-05, "loss": 0.0916, "step": 4730 }, { "grad_norm": 2.6558215618133545, "learning_rate": 9.997667252818214e-05, "loss": 0.0908, "step": 4740 }, { "grad_norm": 0.5774450898170471, "learning_rate": 9.997603698211855e-05, "loss": 0.077, "step": 4750 }, { "grad_norm": 0.4572569727897644, "learning_rate": 9.99753928965171e-05, "loss": 0.0836, "step": 4760 }, { "grad_norm": 0.7488162517547607, "learning_rate": 9.997474027148781e-05, "loss": 0.1074, "step": 4770 }, { "grad_norm": 0.6300467252731323, "learning_rate": 9.997407910714223e-05, "loss": 0.1193, "step": 4780 }, { "grad_norm": 0.6150078773498535, "learning_rate": 9.997340940359332e-05, "loss": 0.0872, "step": 4790 }, { "grad_norm": 0.6268617510795593, "learning_rate": 9.997273116095552e-05, "loss": 0.0876, "step": 4800 }, { "grad_norm": 0.5483289957046509, "learning_rate": 9.997204437934473e-05, "loss": 0.0941, "step": 4810 }, { "grad_norm": 0.6484104990959167, "learning_rate": 9.997134905887829e-05, "loss": 0.1075, "step": 4820 }, { "grad_norm": 0.5786786675453186, "learning_rate": 9.997064519967501e-05, "loss": 0.0913, "step": 4830 }, { "grad_norm": 0.6752810478210449, "learning_rate": 9.996993280185517e-05, "loss": 0.0945, "step": 4840 }, { "grad_norm": 0.5846556425094604, "learning_rate": 9.99692118655405e-05, "loss": 0.0893, "step": 4850 }, { "grad_norm": 0.7257446646690369, "learning_rate": 9.996848239085417e-05, "loss": 0.0848, "step": 4860 }, { "grad_norm": 0.597976803779602, "learning_rate": 9.996774437792085e-05, "loss": 0.0856, "step": 4870 }, { "grad_norm": 0.6420459747314453, "learning_rate": 9.996699782686664e-05, "loss": 0.0812, "step": 4880 }, { "grad_norm": 0.6996833086013794, "learning_rate": 9.996624273781909e-05, "loss": 0.1104, "step": 4890 }, { "grad_norm": 0.5624186396598816, "learning_rate": 9.996547911090725e-05, "loss": 0.0826, "step": 4900 }, { "grad_norm": 0.56802898645401, "learning_rate": 9.996470694626157e-05, "loss": 0.0838, "step": 4910 }, { "grad_norm": 0.5719650983810425, "learning_rate": 9.996392624401403e-05, "loss": 0.0907, "step": 4920 }, { "grad_norm": 0.5820479393005371, "learning_rate": 9.996313700429801e-05, "loss": 0.0881, "step": 4930 }, { "grad_norm": 0.49220579862594604, "learning_rate": 9.996233922724836e-05, "loss": 0.0712, "step": 4940 }, { "grad_norm": 0.6555395722389221, "learning_rate": 9.996153291300141e-05, "loss": 0.0914, "step": 4950 }, { "grad_norm": 0.5090513229370117, "learning_rate": 9.996071806169494e-05, "loss": 0.0889, "step": 4960 }, { "grad_norm": 0.3945305049419403, "learning_rate": 9.995989467346817e-05, "loss": 0.0939, "step": 4970 }, { "grad_norm": 0.5608388185501099, "learning_rate": 9.995906274846183e-05, "loss": 0.0851, "step": 4980 }, { "grad_norm": 0.4637146592140198, "learning_rate": 9.995822228681803e-05, "loss": 0.0753, "step": 4990 }, { "grad_norm": 0.48224174976348877, "learning_rate": 9.99573732886804e-05, "loss": 0.0842, "step": 5000 }, { "grad_norm": 0.5125570893287659, "learning_rate": 9.995651575419402e-05, "loss": 0.0844, "step": 5010 }, { "grad_norm": 0.6883642673492432, "learning_rate": 9.995564968350541e-05, "loss": 0.0823, "step": 5020 }, { "grad_norm": 0.5817533731460571, "learning_rate": 9.995477507676256e-05, "loss": 0.0701, "step": 5030 }, { "grad_norm": 0.5820673108100891, "learning_rate": 9.995389193411493e-05, "loss": 0.071, "step": 5040 }, { "grad_norm": 0.6729530096054077, "learning_rate": 9.995300025571339e-05, "loss": 0.0863, "step": 5050 }, { "grad_norm": 0.6950698494911194, "learning_rate": 9.995210004171034e-05, "loss": 0.0577, "step": 5060 }, { "grad_norm": 0.5733428001403809, "learning_rate": 9.995119129225956e-05, "loss": 0.1009, "step": 5070 }, { "grad_norm": 0.5071480870246887, "learning_rate": 9.995027400751637e-05, "loss": 0.0891, "step": 5080 }, { "grad_norm": 0.6860296130180359, "learning_rate": 9.994934818763751e-05, "loss": 0.0769, "step": 5090 }, { "grad_norm": 0.7009074687957764, "learning_rate": 9.994841383278115e-05, "loss": 0.0775, "step": 5100 }, { "grad_norm": 0.5691312551498413, "learning_rate": 9.994747094310695e-05, "loss": 0.0861, "step": 5110 }, { "grad_norm": 0.603405237197876, "learning_rate": 9.994651951877604e-05, "loss": 0.0963, "step": 5120 }, { "grad_norm": 0.5807543992996216, "learning_rate": 9.994555955995099e-05, "loss": 0.0924, "step": 5130 }, { "grad_norm": 0.48155027627944946, "learning_rate": 9.994459106679581e-05, "loss": 0.0741, "step": 5140 }, { "grad_norm": 0.44997715950012207, "learning_rate": 9.994361403947603e-05, "loss": 0.0687, "step": 5150 }, { "grad_norm": 0.45776230096817017, "learning_rate": 9.994262847815854e-05, "loss": 0.0911, "step": 5160 }, { "grad_norm": 0.5363849997520447, "learning_rate": 9.99416343830118e-05, "loss": 0.0754, "step": 5170 }, { "grad_norm": 0.6038907170295715, "learning_rate": 9.994063175420565e-05, "loss": 0.066, "step": 5180 }, { "grad_norm": 0.6109164357185364, "learning_rate": 9.99396205919114e-05, "loss": 0.1081, "step": 5190 }, { "grad_norm": 0.5048061013221741, "learning_rate": 9.993860089630185e-05, "loss": 0.065, "step": 5200 }, { "grad_norm": 0.6306725144386292, "learning_rate": 9.993757266755123e-05, "loss": 0.0835, "step": 5210 }, { "grad_norm": 0.5881379246711731, "learning_rate": 9.993653590583522e-05, "loss": 0.078, "step": 5220 }, { "grad_norm": 0.38056179881095886, "learning_rate": 9.993549061133102e-05, "loss": 0.0748, "step": 5230 }, { "grad_norm": 0.4845488667488098, "learning_rate": 9.993443678421719e-05, "loss": 0.0784, "step": 5240 }, { "grad_norm": 0.556121289730072, "learning_rate": 9.993337442467384e-05, "loss": 0.0912, "step": 5250 }, { "grad_norm": 0.6582193970680237, "learning_rate": 9.993230353288248e-05, "loss": 0.0719, "step": 5260 }, { "grad_norm": 0.5986446142196655, "learning_rate": 9.993122410902608e-05, "loss": 0.0769, "step": 5270 }, { "grad_norm": 0.5785627365112305, "learning_rate": 9.993013615328912e-05, "loss": 0.0806, "step": 5280 }, { "grad_norm": 0.4724591076374054, "learning_rate": 9.992903966585747e-05, "loss": 0.0833, "step": 5290 }, { "grad_norm": 0.7008654475212097, "learning_rate": 9.992793464691852e-05, "loss": 0.0729, "step": 5300 }, { "grad_norm": 0.5226662158966064, "learning_rate": 9.992682109666105e-05, "loss": 0.0638, "step": 5310 }, { "grad_norm": 0.5917443633079529, "learning_rate": 9.992569901527538e-05, "loss": 0.0759, "step": 5320 }, { "grad_norm": 0.630059003829956, "learning_rate": 9.99245684029532e-05, "loss": 0.0918, "step": 5330 }, { "grad_norm": 0.4774397015571594, "learning_rate": 9.992342925988774e-05, "loss": 0.0748, "step": 5340 }, { "grad_norm": 0.3612774908542633, "learning_rate": 9.992228158627361e-05, "loss": 0.0628, "step": 5350 }, { "grad_norm": 0.43246176838874817, "learning_rate": 9.992112538230693e-05, "loss": 0.0688, "step": 5360 }, { "grad_norm": 0.5111973881721497, "learning_rate": 9.991996064818527e-05, "loss": 0.0617, "step": 5370 }, { "grad_norm": 0.531684935092926, "learning_rate": 9.991878738410768e-05, "loss": 0.072, "step": 5380 }, { "grad_norm": 0.48126915097236633, "learning_rate": 9.991760559027457e-05, "loss": 0.0717, "step": 5390 }, { "grad_norm": 0.5176061391830444, "learning_rate": 9.991641526688793e-05, "loss": 0.081, "step": 5400 }, { "grad_norm": 0.516690731048584, "learning_rate": 9.991521641415113e-05, "loss": 0.0688, "step": 5410 }, { "grad_norm": 0.37536758184432983, "learning_rate": 9.991400903226904e-05, "loss": 0.0713, "step": 5420 }, { "grad_norm": 0.6110721230506897, "learning_rate": 9.991279312144794e-05, "loss": 0.0669, "step": 5430 }, { "grad_norm": 0.580979585647583, "learning_rate": 9.991156868189564e-05, "loss": 0.0869, "step": 5440 }, { "grad_norm": 0.5040397644042969, "learning_rate": 9.991033571382131e-05, "loss": 0.0597, "step": 5450 }, { "grad_norm": 0.551862359046936, "learning_rate": 9.990909421743569e-05, "loss": 0.0695, "step": 5460 }, { "grad_norm": 0.4612298905849457, "learning_rate": 9.990784419295085e-05, "loss": 0.0652, "step": 5470 }, { "grad_norm": 0.49232012033462524, "learning_rate": 9.990658564058044e-05, "loss": 0.0766, "step": 5480 }, { "grad_norm": 0.39483219385147095, "learning_rate": 9.990531856053948e-05, "loss": 0.0643, "step": 5490 }, { "grad_norm": 0.5680502653121948, "learning_rate": 9.99040429530445e-05, "loss": 0.0749, "step": 5500 }, { "grad_norm": 0.5477913022041321, "learning_rate": 9.990275881831346e-05, "loss": 0.0811, "step": 5510 }, { "grad_norm": 0.5559313297271729, "learning_rate": 9.990146615656577e-05, "loss": 0.0671, "step": 5520 }, { "grad_norm": 0.4871159791946411, "learning_rate": 9.990016496802233e-05, "loss": 0.0721, "step": 5530 }, { "grad_norm": 0.47092872858047485, "learning_rate": 9.989885525290548e-05, "loss": 0.0589, "step": 5540 }, { "grad_norm": 0.452085018157959, "learning_rate": 9.989753701143897e-05, "loss": 0.0633, "step": 5550 }, { "grad_norm": 0.5739067792892456, "learning_rate": 9.989621024384812e-05, "loss": 0.0638, "step": 5560 }, { "grad_norm": 0.5817940831184387, "learning_rate": 9.989487495035959e-05, "loss": 0.069, "step": 5570 }, { "grad_norm": 0.5491105318069458, "learning_rate": 9.989353113120156e-05, "loss": 0.0845, "step": 5580 }, { "grad_norm": 0.5547860860824585, "learning_rate": 9.989217878660366e-05, "loss": 0.0757, "step": 5590 }, { "grad_norm": 0.4489729404449463, "learning_rate": 9.989081791679695e-05, "loss": 0.0792, "step": 5600 }, { "grad_norm": 0.5179898142814636, "learning_rate": 9.988944852201397e-05, "loss": 0.0703, "step": 5610 }, { "grad_norm": 0.5266959071159363, "learning_rate": 9.988807060248873e-05, "loss": 0.0653, "step": 5620 }, { "grad_norm": 0.3309868574142456, "learning_rate": 9.988668415845665e-05, "loss": 0.06, "step": 5630 }, { "grad_norm": 0.5027449727058411, "learning_rate": 9.988528919015466e-05, "loss": 0.0665, "step": 5640 }, { "grad_norm": 0.4987538754940033, "learning_rate": 9.988388569782112e-05, "loss": 0.0819, "step": 5650 }, { "grad_norm": 0.5911209583282471, "learning_rate": 9.988247368169583e-05, "loss": 0.0694, "step": 5660 }, { "grad_norm": 0.4854319393634796, "learning_rate": 9.988105314202007e-05, "loss": 0.0742, "step": 5670 }, { "grad_norm": 0.5581628680229187, "learning_rate": 9.987962407903659e-05, "loss": 0.0818, "step": 5680 }, { "grad_norm": 0.5571326613426208, "learning_rate": 9.987818649298957e-05, "loss": 0.0783, "step": 5690 }, { "grad_norm": 0.4368312656879425, "learning_rate": 9.987674038412465e-05, "loss": 0.0677, "step": 5700 }, { "grad_norm": 0.45265042781829834, "learning_rate": 9.987528575268891e-05, "loss": 0.0799, "step": 5710 }, { "grad_norm": 0.574731707572937, "learning_rate": 9.987382259893095e-05, "loss": 0.0668, "step": 5720 }, { "grad_norm": 0.4605560302734375, "learning_rate": 9.987235092310074e-05, "loss": 0.0685, "step": 5730 }, { "grad_norm": 0.34634703397750854, "learning_rate": 9.987087072544978e-05, "loss": 0.0861, "step": 5740 }, { "grad_norm": 0.48375722765922546, "learning_rate": 9.9869382006231e-05, "loss": 0.0663, "step": 5750 }, { "grad_norm": 0.43096715211868286, "learning_rate": 9.986788476569875e-05, "loss": 0.0679, "step": 5760 }, { "grad_norm": 0.4065420627593994, "learning_rate": 9.986637900410887e-05, "loss": 0.0687, "step": 5770 }, { "grad_norm": 0.44615012407302856, "learning_rate": 9.986486472171869e-05, "loss": 0.0856, "step": 5780 }, { "grad_norm": 0.47678548097610474, "learning_rate": 9.986334191878692e-05, "loss": 0.0618, "step": 5790 }, { "grad_norm": 0.6352498531341553, "learning_rate": 9.986181059557378e-05, "loss": 0.0818, "step": 5800 }, { "grad_norm": 0.5413223505020142, "learning_rate": 9.986027075234094e-05, "loss": 0.0655, "step": 5810 }, { "grad_norm": 0.4442448019981384, "learning_rate": 9.985872238935152e-05, "loss": 0.0672, "step": 5820 }, { "grad_norm": 0.4189223051071167, "learning_rate": 9.985716550687008e-05, "loss": 0.0724, "step": 5830 }, { "grad_norm": 0.7144829034805298, "learning_rate": 9.985560010516264e-05, "loss": 0.0621, "step": 5840 }, { "grad_norm": 0.38404226303100586, "learning_rate": 9.985402618449668e-05, "loss": 0.0568, "step": 5850 }, { "grad_norm": 0.47639036178588867, "learning_rate": 9.985244374514118e-05, "loss": 0.0843, "step": 5860 }, { "grad_norm": 0.4081534147262573, "learning_rate": 9.985085278736651e-05, "loss": 0.0784, "step": 5870 }, { "grad_norm": 0.5450401306152344, "learning_rate": 9.984925331144452e-05, "loss": 0.0864, "step": 5880 }, { "grad_norm": 0.5224577188491821, "learning_rate": 9.984764531764851e-05, "loss": 0.0751, "step": 5890 }, { "grad_norm": 0.4936981201171875, "learning_rate": 9.984602880625326e-05, "loss": 0.0948, "step": 5900 }, { "grad_norm": 0.4095041751861572, "learning_rate": 9.9844403777535e-05, "loss": 0.0855, "step": 5910 }, { "grad_norm": 0.6478757858276367, "learning_rate": 9.984277023177135e-05, "loss": 0.0751, "step": 5920 }, { "grad_norm": 0.5275310277938843, "learning_rate": 9.984112816924148e-05, "loss": 0.0726, "step": 5930 }, { "grad_norm": 0.4686587452888489, "learning_rate": 9.983947759022596e-05, "loss": 0.0738, "step": 5940 }, { "grad_norm": 0.41884034872055054, "learning_rate": 9.983781849500682e-05, "loss": 0.0682, "step": 5950 }, { "grad_norm": 0.6573699712753296, "learning_rate": 9.98361508838676e-05, "loss": 0.0662, "step": 5960 }, { "grad_norm": 0.7164794206619263, "learning_rate": 9.98344747570932e-05, "loss": 0.0741, "step": 5970 }, { "grad_norm": 0.4621574878692627, "learning_rate": 9.983279011497004e-05, "loss": 0.074, "step": 5980 }, { "grad_norm": 0.6066710948944092, "learning_rate": 9.983109695778596e-05, "loss": 0.0812, "step": 5990 }, { "grad_norm": 0.36414507031440735, "learning_rate": 9.982939528583032e-05, "loss": 0.0912, "step": 6000 }, { "grad_norm": 0.46865949034690857, "learning_rate": 9.982768509939385e-05, "loss": 0.0752, "step": 6010 }, { "grad_norm": 0.4263301491737366, "learning_rate": 9.982596639876879e-05, "loss": 0.0704, "step": 6020 }, { "grad_norm": 0.5232917666435242, "learning_rate": 9.982423918424881e-05, "loss": 0.0742, "step": 6030 }, { "grad_norm": 0.5610369443893433, "learning_rate": 9.982250345612908e-05, "loss": 0.0679, "step": 6040 }, { "grad_norm": 0.4726651906967163, "learning_rate": 9.982075921470611e-05, "loss": 0.0844, "step": 6050 }, { "grad_norm": 0.4575302004814148, "learning_rate": 9.981900646027802e-05, "loss": 0.0563, "step": 6060 }, { "grad_norm": 0.5181905627250671, "learning_rate": 9.981724519314425e-05, "loss": 0.071, "step": 6070 }, { "grad_norm": 0.5308888554573059, "learning_rate": 9.981547541360581e-05, "loss": 0.0752, "step": 6080 }, { "grad_norm": 0.5587815642356873, "learning_rate": 9.981369712196508e-05, "loss": 0.0772, "step": 6090 }, { "grad_norm": 0.4234645366668701, "learning_rate": 9.981191031852592e-05, "loss": 0.0869, "step": 6100 }, { "grad_norm": 0.37819865345954895, "learning_rate": 9.981011500359362e-05, "loss": 0.0552, "step": 6110 }, { "grad_norm": 0.3967142403125763, "learning_rate": 9.9808311177475e-05, "loss": 0.0648, "step": 6120 }, { "grad_norm": 0.4472222924232483, "learning_rate": 9.980649884047826e-05, "loss": 0.0564, "step": 6130 }, { "grad_norm": 0.44870254397392273, "learning_rate": 9.980467799291307e-05, "loss": 0.0685, "step": 6140 }, { "grad_norm": 0.5110296010971069, "learning_rate": 9.980284863509058e-05, "loss": 0.0866, "step": 6150 }, { "grad_norm": 0.5466628074645996, "learning_rate": 9.980101076732334e-05, "loss": 0.0862, "step": 6160 }, { "grad_norm": 0.6059700846672058, "learning_rate": 9.979916438992544e-05, "loss": 0.0821, "step": 6170 }, { "grad_norm": 0.47286778688430786, "learning_rate": 9.979730950321237e-05, "loss": 0.0657, "step": 6180 }, { "grad_norm": 0.36299678683280945, "learning_rate": 9.979544610750104e-05, "loss": 0.0831, "step": 6190 }, { "grad_norm": 0.37461531162261963, "learning_rate": 9.97935742031099e-05, "loss": 0.0701, "step": 6200 }, { "grad_norm": 0.5259912610054016, "learning_rate": 9.979169379035878e-05, "loss": 0.0813, "step": 6210 }, { "grad_norm": 0.5177221298217773, "learning_rate": 9.978980486956899e-05, "loss": 0.0608, "step": 6220 }, { "grad_norm": 0.3743232488632202, "learning_rate": 9.978790744106332e-05, "loss": 0.0676, "step": 6230 }, { "grad_norm": 0.5126904845237732, "learning_rate": 9.978600150516594e-05, "loss": 0.07, "step": 6240 }, { "grad_norm": 0.4179547131061554, "learning_rate": 9.978408706220259e-05, "loss": 0.0781, "step": 6250 }, { "grad_norm": 0.5054476261138916, "learning_rate": 9.978216411250032e-05, "loss": 0.0593, "step": 6260 }, { "grad_norm": 0.4866030216217041, "learning_rate": 9.978023265638778e-05, "loss": 0.059, "step": 6270 }, { "grad_norm": 0.50723797082901, "learning_rate": 9.977829269419495e-05, "loss": 0.0668, "step": 6280 }, { "grad_norm": 0.5008875727653503, "learning_rate": 9.977634422625335e-05, "loss": 0.0656, "step": 6290 }, { "grad_norm": 0.4033929705619812, "learning_rate": 9.97743872528959e-05, "loss": 0.0677, "step": 6300 }, { "grad_norm": 0.5443859696388245, "learning_rate": 9.9772421774457e-05, "loss": 0.0595, "step": 6310 }, { "grad_norm": 0.39563703536987305, "learning_rate": 9.977044779127252e-05, "loss": 0.0571, "step": 6320 }, { "grad_norm": 0.5752232670783997, "learning_rate": 9.976846530367971e-05, "loss": 0.0669, "step": 6330 }, { "grad_norm": 0.574953019618988, "learning_rate": 9.976647431201735e-05, "loss": 0.0776, "step": 6340 }, { "grad_norm": 0.5030823349952698, "learning_rate": 9.976447481662568e-05, "loss": 0.0625, "step": 6350 }, { "grad_norm": 0.4695771038532257, "learning_rate": 9.976246681784629e-05, "loss": 0.0669, "step": 6360 }, { "grad_norm": 0.4548564553260803, "learning_rate": 9.976045031602234e-05, "loss": 0.0651, "step": 6370 }, { "grad_norm": 0.5175975561141968, "learning_rate": 9.975842531149837e-05, "loss": 0.0609, "step": 6380 }, { "grad_norm": 0.5349797010421753, "learning_rate": 9.975639180462043e-05, "loss": 0.081, "step": 6390 }, { "grad_norm": 0.4613693952560425, "learning_rate": 9.975434979573596e-05, "loss": 0.0793, "step": 6400 }, { "grad_norm": 0.43298205733299255, "learning_rate": 9.97522992851939e-05, "loss": 0.0518, "step": 6410 }, { "grad_norm": 0.4289155900478363, "learning_rate": 9.975024027334461e-05, "loss": 0.0731, "step": 6420 }, { "grad_norm": 0.499947190284729, "learning_rate": 9.974817276053993e-05, "loss": 0.0667, "step": 6430 }, { "grad_norm": 0.4205218553543091, "learning_rate": 9.974609674713315e-05, "loss": 0.0647, "step": 6440 }, { "grad_norm": 0.42333394289016724, "learning_rate": 9.9744012233479e-05, "loss": 0.0615, "step": 6450 }, { "grad_norm": 0.40587830543518066, "learning_rate": 9.974191921993366e-05, "loss": 0.0624, "step": 6460 }, { "grad_norm": 0.3901880383491516, "learning_rate": 9.973981770685474e-05, "loss": 0.0623, "step": 6470 }, { "grad_norm": 0.5479671955108643, "learning_rate": 9.97377076946014e-05, "loss": 0.0617, "step": 6480 }, { "grad_norm": 0.49592456221580505, "learning_rate": 9.973558918353412e-05, "loss": 0.0747, "step": 6490 }, { "grad_norm": 0.40140169858932495, "learning_rate": 9.973346217401494e-05, "loss": 0.0581, "step": 6500 }, { "grad_norm": 0.44924476742744446, "learning_rate": 9.973132666640726e-05, "loss": 0.0562, "step": 6510 }, { "grad_norm": 0.47995725274086, "learning_rate": 9.972918266107602e-05, "loss": 0.0622, "step": 6520 }, { "grad_norm": 0.34570327401161194, "learning_rate": 9.972703015838756e-05, "loss": 0.0615, "step": 6530 }, { "grad_norm": 0.4686564803123474, "learning_rate": 9.97248691587097e-05, "loss": 0.0706, "step": 6540 }, { "grad_norm": 0.6192429661750793, "learning_rate": 9.972269966241166e-05, "loss": 0.0589, "step": 6550 }, { "grad_norm": 0.4608435332775116, "learning_rate": 9.972052166986417e-05, "loss": 0.0666, "step": 6560 }, { "grad_norm": 0.5157927870750427, "learning_rate": 9.971833518143938e-05, "loss": 0.0738, "step": 6570 }, { "grad_norm": 0.3666037917137146, "learning_rate": 9.971614019751093e-05, "loss": 0.0551, "step": 6580 }, { "grad_norm": 0.5203360915184021, "learning_rate": 9.971393671845383e-05, "loss": 0.073, "step": 6590 }, { "grad_norm": 0.47890588641166687, "learning_rate": 9.971172474464464e-05, "loss": 0.0787, "step": 6600 }, { "grad_norm": 0.5799616575241089, "learning_rate": 9.97095042764613e-05, "loss": 0.066, "step": 6610 }, { "grad_norm": 0.4487362504005432, "learning_rate": 9.970727531428324e-05, "loss": 0.0626, "step": 6620 }, { "grad_norm": 0.43695634603500366, "learning_rate": 9.970503785849132e-05, "loss": 0.0666, "step": 6630 }, { "grad_norm": 0.3855609595775604, "learning_rate": 9.970279190946788e-05, "loss": 0.0569, "step": 6640 }, { "grad_norm": 0.47553256154060364, "learning_rate": 9.970053746759667e-05, "loss": 0.0786, "step": 6650 }, { "grad_norm": 0.57091224193573, "learning_rate": 9.969827453326292e-05, "loss": 0.0581, "step": 6660 }, { "grad_norm": 0.9114806652069092, "learning_rate": 9.969600310685332e-05, "loss": 0.0553, "step": 6670 }, { "grad_norm": 0.3823832869529724, "learning_rate": 9.969372318875596e-05, "loss": 0.0768, "step": 6680 }, { "grad_norm": 0.5530168414115906, "learning_rate": 9.969143477936043e-05, "loss": 0.076, "step": 6690 }, { "grad_norm": 0.5195163488388062, "learning_rate": 9.968913787905775e-05, "loss": 0.0734, "step": 6700 }, { "grad_norm": 0.543339729309082, "learning_rate": 9.968683248824045e-05, "loss": 0.0773, "step": 6710 }, { "grad_norm": 0.44750717282295227, "learning_rate": 9.968451860730238e-05, "loss": 0.0695, "step": 6720 }, { "grad_norm": 0.4837262034416199, "learning_rate": 9.968219623663896e-05, "loss": 0.0755, "step": 6730 }, { "grad_norm": 0.44509080052375793, "learning_rate": 9.967986537664702e-05, "loss": 0.0949, "step": 6740 }, { "grad_norm": 0.38647404313087463, "learning_rate": 9.967752602772483e-05, "loss": 0.0622, "step": 6750 }, { "grad_norm": 0.5405877232551575, "learning_rate": 9.967517819027212e-05, "loss": 0.0684, "step": 6760 }, { "grad_norm": 0.4333084225654602, "learning_rate": 9.967282186469009e-05, "loss": 0.0637, "step": 6770 }, { "grad_norm": 0.4795316755771637, "learning_rate": 9.967045705138135e-05, "loss": 0.0783, "step": 6780 }, { "grad_norm": 0.537815272808075, "learning_rate": 9.966808375074998e-05, "loss": 0.0593, "step": 6790 }, { "grad_norm": 0.42700669169425964, "learning_rate": 9.966570196320154e-05, "loss": 0.0662, "step": 6800 }, { "grad_norm": 0.42746350169181824, "learning_rate": 9.966331168914299e-05, "loss": 0.0545, "step": 6810 }, { "grad_norm": 0.5287580490112305, "learning_rate": 9.966091292898277e-05, "loss": 0.0842, "step": 6820 }, { "grad_norm": 0.6390794515609741, "learning_rate": 9.965850568313076e-05, "loss": 0.0699, "step": 6830 }, { "grad_norm": 0.4526643753051758, "learning_rate": 9.965608995199827e-05, "loss": 0.0739, "step": 6840 }, { "grad_norm": 0.400973379611969, "learning_rate": 9.965366573599812e-05, "loss": 0.0697, "step": 6850 }, { "grad_norm": 0.46484753489494324, "learning_rate": 9.965123303554453e-05, "loss": 0.0622, "step": 6860 }, { "grad_norm": 0.5477197766304016, "learning_rate": 9.964879185105317e-05, "loss": 0.0884, "step": 6870 }, { "grad_norm": 0.28679853677749634, "learning_rate": 9.964634218294119e-05, "loss": 0.0763, "step": 6880 }, { "grad_norm": 0.43441271781921387, "learning_rate": 9.964388403162714e-05, "loss": 0.0754, "step": 6890 }, { "grad_norm": 0.5513740181922913, "learning_rate": 9.96414173975311e-05, "loss": 0.0767, "step": 6900 }, { "grad_norm": 0.49588069319725037, "learning_rate": 9.963894228107451e-05, "loss": 0.0651, "step": 6910 }, { "grad_norm": 0.4039798378944397, "learning_rate": 9.963645868268032e-05, "loss": 0.0794, "step": 6920 }, { "grad_norm": 0.4169510304927826, "learning_rate": 9.963396660277289e-05, "loss": 0.0789, "step": 6930 }, { "grad_norm": 0.5643049478530884, "learning_rate": 9.963146604177807e-05, "loss": 0.059, "step": 6940 }, { "grad_norm": 0.5043057799339294, "learning_rate": 9.962895700012311e-05, "loss": 0.0657, "step": 6950 }, { "grad_norm": 0.5203849673271179, "learning_rate": 9.962643947823677e-05, "loss": 0.0674, "step": 6960 }, { "grad_norm": 0.3823876976966858, "learning_rate": 9.962391347654921e-05, "loss": 0.0719, "step": 6970 }, { "grad_norm": 0.38331422209739685, "learning_rate": 9.962137899549204e-05, "loss": 0.073, "step": 6980 }, { "grad_norm": 0.4356664717197418, "learning_rate": 9.961883603549835e-05, "loss": 0.063, "step": 6990 }, { "grad_norm": 0.5082233548164368, "learning_rate": 9.961628459700267e-05, "loss": 0.0675, "step": 7000 }, { "grad_norm": 0.40483608841896057, "learning_rate": 9.961372468044095e-05, "loss": 0.0759, "step": 7010 }, { "grad_norm": 0.48367390036582947, "learning_rate": 9.961115628625062e-05, "loss": 0.0674, "step": 7020 }, { "grad_norm": 0.5567945837974548, "learning_rate": 9.960857941487056e-05, "loss": 0.0859, "step": 7030 }, { "grad_norm": 0.39065083861351013, "learning_rate": 9.960599406674106e-05, "loss": 0.0648, "step": 7040 }, { "grad_norm": 0.5229772329330444, "learning_rate": 9.960340024230393e-05, "loss": 0.0639, "step": 7050 }, { "grad_norm": 0.3361578583717346, "learning_rate": 9.960079794200232e-05, "loss": 0.0669, "step": 7060 }, { "grad_norm": 0.5305434465408325, "learning_rate": 9.959818716628096e-05, "loss": 0.0702, "step": 7070 }, { "grad_norm": 0.3906662166118622, "learning_rate": 9.95955679155859e-05, "loss": 0.0782, "step": 7080 }, { "grad_norm": 0.6564850807189941, "learning_rate": 9.959294019036472e-05, "loss": 0.0692, "step": 7090 }, { "grad_norm": 0.4978812038898468, "learning_rate": 9.959030399106646e-05, "loss": 0.0583, "step": 7100 }, { "grad_norm": 0.5834566950798035, "learning_rate": 9.958765931814153e-05, "loss": 0.0788, "step": 7110 }, { "grad_norm": 0.39128777384757996, "learning_rate": 9.958500617204184e-05, "loss": 0.0515, "step": 7120 }, { "grad_norm": 0.38477256894111633, "learning_rate": 9.958234455322075e-05, "loss": 0.0564, "step": 7130 }, { "grad_norm": 0.5107999444007874, "learning_rate": 9.957967446213308e-05, "loss": 0.0773, "step": 7140 }, { "grad_norm": 0.43322113156318665, "learning_rate": 9.957699589923501e-05, "loss": 0.0794, "step": 7150 }, { "grad_norm": 0.685595691204071, "learning_rate": 9.957430886498431e-05, "loss": 0.0721, "step": 7160 }, { "grad_norm": 0.403987854719162, "learning_rate": 9.957161335984008e-05, "loss": 0.0686, "step": 7170 }, { "grad_norm": 0.4681679308414459, "learning_rate": 9.956890938426291e-05, "loss": 0.0628, "step": 7180 }, { "grad_norm": 0.4064110517501831, "learning_rate": 9.956619693871482e-05, "loss": 0.0725, "step": 7190 }, { "grad_norm": 0.4935013949871063, "learning_rate": 9.956347602365934e-05, "loss": 0.0707, "step": 7200 }, { "grad_norm": 0.3521007001399994, "learning_rate": 9.956074663956135e-05, "loss": 0.0669, "step": 7210 }, { "grad_norm": 0.51341313123703, "learning_rate": 9.955800878688726e-05, "loss": 0.0751, "step": 7220 }, { "grad_norm": 0.431506484746933, "learning_rate": 9.955526246610489e-05, "loss": 0.0571, "step": 7230 }, { "grad_norm": 0.6196004748344421, "learning_rate": 9.955250767768349e-05, "loss": 0.0716, "step": 7240 }, { "grad_norm": 0.3396226167678833, "learning_rate": 9.95497444220938e-05, "loss": 0.0698, "step": 7250 }, { "grad_norm": 0.34108832478523254, "learning_rate": 9.954697269980797e-05, "loss": 0.0564, "step": 7260 }, { "grad_norm": 0.26748034358024597, "learning_rate": 9.954419251129962e-05, "loss": 0.0558, "step": 7270 }, { "grad_norm": 0.4468843638896942, "learning_rate": 9.95414038570438e-05, "loss": 0.0591, "step": 7280 }, { "grad_norm": 0.5164341330528259, "learning_rate": 9.953860673751703e-05, "loss": 0.0688, "step": 7290 }, { "grad_norm": 0.3642774522304535, "learning_rate": 9.953580115319725e-05, "loss": 0.0597, "step": 7300 }, { "grad_norm": 0.4808535873889923, "learning_rate": 9.953298710456387e-05, "loss": 0.0674, "step": 7310 }, { "grad_norm": 0.3938741087913513, "learning_rate": 9.953016459209771e-05, "loss": 0.0603, "step": 7320 }, { "grad_norm": 0.4229210913181305, "learning_rate": 9.952733361628108e-05, "loss": 0.0657, "step": 7330 }, { "grad_norm": 0.4423636794090271, "learning_rate": 9.952449417759772e-05, "loss": 0.0693, "step": 7340 }, { "grad_norm": 0.4088185429573059, "learning_rate": 9.952164627653279e-05, "loss": 0.0527, "step": 7350 }, { "grad_norm": 0.445626825094223, "learning_rate": 9.951878991357292e-05, "loss": 0.0638, "step": 7360 }, { "grad_norm": 0.4756673276424408, "learning_rate": 9.951592508920622e-05, "loss": 0.0664, "step": 7370 }, { "grad_norm": 0.3543103039264679, "learning_rate": 9.951305180392219e-05, "loss": 0.0544, "step": 7380 }, { "grad_norm": 0.4755264222621918, "learning_rate": 9.951017005821178e-05, "loss": 0.0613, "step": 7390 }, { "grad_norm": 0.3999435305595398, "learning_rate": 9.95072798525674e-05, "loss": 0.0575, "step": 7400 }, { "grad_norm": 0.38941165804862976, "learning_rate": 9.950438118748293e-05, "loss": 0.0654, "step": 7410 }, { "grad_norm": 0.4526243805885315, "learning_rate": 9.950147406345366e-05, "loss": 0.0776, "step": 7420 }, { "grad_norm": 0.43548768758773804, "learning_rate": 9.949855848097635e-05, "loss": 0.0702, "step": 7430 }, { "grad_norm": 0.6550454497337341, "learning_rate": 9.949563444054916e-05, "loss": 0.0624, "step": 7440 }, { "grad_norm": 0.5520855188369751, "learning_rate": 9.949270194267178e-05, "loss": 0.0652, "step": 7450 }, { "grad_norm": 0.4147232472896576, "learning_rate": 9.948976098784526e-05, "loss": 0.0616, "step": 7460 }, { "grad_norm": 0.5941852927207947, "learning_rate": 9.948681157657213e-05, "loss": 0.0732, "step": 7470 }, { "grad_norm": 0.4941081702709198, "learning_rate": 9.948385370935638e-05, "loss": 0.0756, "step": 7480 }, { "grad_norm": 0.606399655342102, "learning_rate": 9.94808873867034e-05, "loss": 0.0741, "step": 7490 }, { "grad_norm": 0.5683377385139465, "learning_rate": 9.947791260912009e-05, "loss": 0.0641, "step": 7500 }, { "grad_norm": 0.41064029932022095, "learning_rate": 9.947492937711474e-05, "loss": 0.0539, "step": 7510 }, { "grad_norm": 0.44236132502555847, "learning_rate": 9.947193769119707e-05, "loss": 0.0554, "step": 7520 }, { "grad_norm": 0.4140375554561615, "learning_rate": 9.946893755187834e-05, "loss": 0.0523, "step": 7530 }, { "grad_norm": 0.42532312870025635, "learning_rate": 9.946592895967115e-05, "loss": 0.0568, "step": 7540 }, { "grad_norm": 0.3915844261646271, "learning_rate": 9.94629119150896e-05, "loss": 0.0605, "step": 7550 }, { "grad_norm": 0.40977635979652405, "learning_rate": 9.94598864186492e-05, "loss": 0.0575, "step": 7560 }, { "grad_norm": 0.4184496998786926, "learning_rate": 9.945685247086696e-05, "loss": 0.0804, "step": 7570 }, { "grad_norm": 0.4504960775375366, "learning_rate": 9.945381007226129e-05, "loss": 0.068, "step": 7580 }, { "grad_norm": 0.4704066812992096, "learning_rate": 9.945075922335203e-05, "loss": 0.0752, "step": 7590 }, { "grad_norm": 0.5199998021125793, "learning_rate": 9.944769992466049e-05, "loss": 0.0829, "step": 7600 }, { "grad_norm": 0.45270320773124695, "learning_rate": 9.944463217670945e-05, "loss": 0.0791, "step": 7610 }, { "grad_norm": 0.37918007373809814, "learning_rate": 9.944155598002307e-05, "loss": 0.0632, "step": 7620 }, { "grad_norm": 0.4473378360271454, "learning_rate": 9.943847133512701e-05, "loss": 0.0693, "step": 7630 }, { "grad_norm": 0.45080286264419556, "learning_rate": 9.943537824254834e-05, "loss": 0.061, "step": 7640 }, { "grad_norm": 0.36417630314826965, "learning_rate": 9.943227670281559e-05, "loss": 0.0603, "step": 7650 }, { "grad_norm": 0.33939898014068604, "learning_rate": 9.942916671645873e-05, "loss": 0.0588, "step": 7660 }, { "grad_norm": 0.5177952647209167, "learning_rate": 9.942604828400916e-05, "loss": 0.0697, "step": 7670 }, { "grad_norm": 0.41406339406967163, "learning_rate": 9.942292140599975e-05, "loss": 0.0672, "step": 7680 }, { "grad_norm": 0.40734031796455383, "learning_rate": 9.94197860829648e-05, "loss": 0.0638, "step": 7690 }, { "grad_norm": 0.3041496276855469, "learning_rate": 9.941664231544004e-05, "loss": 0.056, "step": 7700 }, { "grad_norm": 0.6825931072235107, "learning_rate": 9.941349010396264e-05, "loss": 0.0786, "step": 7710 }, { "grad_norm": 0.40499117970466614, "learning_rate": 9.941032944907125e-05, "loss": 0.0631, "step": 7720 }, { "grad_norm": 0.5047183036804199, "learning_rate": 9.940716035130596e-05, "loss": 0.0718, "step": 7730 }, { "grad_norm": 0.6115058660507202, "learning_rate": 9.940398281120821e-05, "loss": 0.0733, "step": 7740 }, { "grad_norm": 0.4251208007335663, "learning_rate": 9.940079682932102e-05, "loss": 0.0599, "step": 7750 }, { "grad_norm": 0.45511093735694885, "learning_rate": 9.939760240618877e-05, "loss": 0.0641, "step": 7760 }, { "grad_norm": 0.4739494025707245, "learning_rate": 9.939439954235729e-05, "loss": 0.0729, "step": 7770 }, { "grad_norm": 0.4938061535358429, "learning_rate": 9.939118823837387e-05, "loss": 0.0674, "step": 7780 }, { "grad_norm": 0.5032980442047119, "learning_rate": 9.938796849478725e-05, "loss": 0.0649, "step": 7790 }, { "grad_norm": 0.43002378940582275, "learning_rate": 9.938474031214755e-05, "loss": 0.0632, "step": 7800 }, { "grad_norm": 0.4496627449989319, "learning_rate": 9.938150369100643e-05, "loss": 0.0673, "step": 7810 }, { "grad_norm": 0.37243351340293884, "learning_rate": 9.93782586319169e-05, "loss": 0.0571, "step": 7820 }, { "grad_norm": 0.43469399213790894, "learning_rate": 9.937500513543348e-05, "loss": 0.0618, "step": 7830 }, { "grad_norm": 0.42603668570518494, "learning_rate": 9.937174320211207e-05, "loss": 0.0538, "step": 7840 }, { "grad_norm": 0.3059886693954468, "learning_rate": 9.936847283251009e-05, "loss": 0.0599, "step": 7850 }, { "grad_norm": 0.5322601795196533, "learning_rate": 9.936519402718632e-05, "loss": 0.0655, "step": 7860 }, { "grad_norm": 0.5095965266227722, "learning_rate": 9.936190678670102e-05, "loss": 0.0779, "step": 7870 }, { "grad_norm": 0.5525032877922058, "learning_rate": 9.935861111161593e-05, "loss": 0.0726, "step": 7880 }, { "grad_norm": 0.3034631311893463, "learning_rate": 9.935530700249416e-05, "loss": 0.0596, "step": 7890 }, { "grad_norm": 0.47691717743873596, "learning_rate": 9.935199445990028e-05, "loss": 0.0768, "step": 7900 }, { "grad_norm": 0.45454394817352295, "learning_rate": 9.934867348440033e-05, "loss": 0.0725, "step": 7910 }, { "grad_norm": 0.38727518916130066, "learning_rate": 9.934534407656176e-05, "loss": 0.0948, "step": 7920 }, { "grad_norm": 0.39106839895248413, "learning_rate": 9.93420062369535e-05, "loss": 0.0584, "step": 7930 }, { "grad_norm": 0.46287715435028076, "learning_rate": 9.933865996614589e-05, "loss": 0.0633, "step": 7940 }, { "grad_norm": 0.49031317234039307, "learning_rate": 9.933530526471068e-05, "loss": 0.0488, "step": 7950 }, { "grad_norm": 0.44852790236473083, "learning_rate": 9.933194213322114e-05, "loss": 0.0538, "step": 7960 }, { "grad_norm": 0.45028722286224365, "learning_rate": 9.932857057225192e-05, "loss": 0.0534, "step": 7970 }, { "grad_norm": 0.5937226414680481, "learning_rate": 9.932519058237912e-05, "loss": 0.0594, "step": 7980 }, { "grad_norm": 0.4884411692619324, "learning_rate": 9.932180216418032e-05, "loss": 0.058, "step": 7990 }, { "grad_norm": 0.3860866129398346, "learning_rate": 9.931840531823446e-05, "loss": 0.0567, "step": 8000 }, { "grad_norm": 0.3834932744503021, "learning_rate": 9.9315000045122e-05, "loss": 0.0492, "step": 8010 }, { "grad_norm": 0.4616653025150299, "learning_rate": 9.931158634542481e-05, "loss": 0.0608, "step": 8020 }, { "grad_norm": 0.3850545883178711, "learning_rate": 9.930816421972617e-05, "loss": 0.0577, "step": 8030 }, { "grad_norm": 0.5191356539726257, "learning_rate": 9.930473366861086e-05, "loss": 0.0571, "step": 8040 }, { "grad_norm": 0.42569538950920105, "learning_rate": 9.930129469266505e-05, "loss": 0.0636, "step": 8050 }, { "grad_norm": 0.3428359925746918, "learning_rate": 9.929784729247638e-05, "loss": 0.0535, "step": 8060 }, { "grad_norm": 0.46362030506134033, "learning_rate": 9.929439146863389e-05, "loss": 0.0595, "step": 8070 }, { "grad_norm": 0.5814775228500366, "learning_rate": 9.92909272217281e-05, "loss": 0.0749, "step": 8080 }, { "grad_norm": 0.4199341833591461, "learning_rate": 9.928745455235097e-05, "loss": 0.0621, "step": 8090 }, { "grad_norm": 0.43558332324028015, "learning_rate": 9.928397346109588e-05, "loss": 0.0687, "step": 8100 }, { "grad_norm": 0.6191354393959045, "learning_rate": 9.928048394855762e-05, "loss": 0.0704, "step": 8110 }, { "grad_norm": 0.38305285573005676, "learning_rate": 9.92769860153325e-05, "loss": 0.0539, "step": 8120 }, { "grad_norm": 0.4137927293777466, "learning_rate": 9.927347966201819e-05, "loss": 0.0727, "step": 8130 }, { "grad_norm": 0.4164320230484009, "learning_rate": 9.926996488921383e-05, "loss": 0.0698, "step": 8140 }, { "grad_norm": 0.4649472236633301, "learning_rate": 9.926644169752001e-05, "loss": 0.0604, "step": 8150 }, { "grad_norm": 0.3971524238586426, "learning_rate": 9.926291008753875e-05, "loss": 0.073, "step": 8160 }, { "grad_norm": 0.3997914493083954, "learning_rate": 9.92593700598735e-05, "loss": 0.066, "step": 8170 }, { "grad_norm": 0.47882670164108276, "learning_rate": 9.925582161512915e-05, "loss": 0.0595, "step": 8180 }, { "grad_norm": 0.3836771249771118, "learning_rate": 9.925226475391205e-05, "loss": 0.0579, "step": 8190 }, { "grad_norm": 0.5204121470451355, "learning_rate": 9.924869947682993e-05, "loss": 0.0579, "step": 8200 }, { "grad_norm": 0.4835962653160095, "learning_rate": 9.924512578449204e-05, "loss": 0.0515, "step": 8210 }, { "grad_norm": 0.5008636116981506, "learning_rate": 9.924154367750901e-05, "loss": 0.0627, "step": 8220 }, { "grad_norm": 0.5457053184509277, "learning_rate": 9.923795315649293e-05, "loss": 0.0632, "step": 8230 }, { "grad_norm": 0.38256993889808655, "learning_rate": 9.92343542220573e-05, "loss": 0.0615, "step": 8240 }, { "grad_norm": 0.3292929530143738, "learning_rate": 9.92307468748171e-05, "loss": 0.0646, "step": 8250 }, { "grad_norm": 0.5101869106292725, "learning_rate": 9.922713111538873e-05, "loss": 0.0539, "step": 8260 }, { "grad_norm": 0.42142829298973083, "learning_rate": 9.922350694439003e-05, "loss": 0.0534, "step": 8270 }, { "grad_norm": 0.5008801221847534, "learning_rate": 9.921987436244024e-05, "loss": 0.0685, "step": 8280 }, { "grad_norm": 0.39616265892982483, "learning_rate": 9.921623337016008e-05, "loss": 0.055, "step": 8290 }, { "grad_norm": 1.8480390310287476, "learning_rate": 9.921258396817172e-05, "loss": 0.0749, "step": 8300 }, { "grad_norm": 0.626742959022522, "learning_rate": 9.920892615709874e-05, "loss": 0.0612, "step": 8310 }, { "grad_norm": 0.4291565716266632, "learning_rate": 9.920525993756612e-05, "loss": 0.0558, "step": 8320 }, { "grad_norm": 0.37244656682014465, "learning_rate": 9.920158531020036e-05, "loss": 0.0632, "step": 8330 }, { "grad_norm": 0.6643531918525696, "learning_rate": 9.919790227562933e-05, "loss": 0.0623, "step": 8340 }, { "grad_norm": 0.4209541976451874, "learning_rate": 9.919421083448237e-05, "loss": 0.0868, "step": 8350 }, { "grad_norm": 0.6939061284065247, "learning_rate": 9.919051098739022e-05, "loss": 0.0672, "step": 8360 }, { "grad_norm": 0.4619895815849304, "learning_rate": 9.918680273498514e-05, "loss": 0.0601, "step": 8370 }, { "grad_norm": 0.5200046300888062, "learning_rate": 9.918308607790072e-05, "loss": 0.0721, "step": 8380 }, { "grad_norm": 0.5108407139778137, "learning_rate": 9.917936101677205e-05, "loss": 0.0653, "step": 8390 }, { "grad_norm": 0.48874491453170776, "learning_rate": 9.917562755223564e-05, "loss": 0.0632, "step": 8400 }, { "grad_norm": 0.4207007586956024, "learning_rate": 9.917188568492944e-05, "loss": 0.0774, "step": 8410 }, { "grad_norm": 0.4089634418487549, "learning_rate": 9.916813541549283e-05, "loss": 0.0649, "step": 8420 }, { "grad_norm": 0.6236392259597778, "learning_rate": 9.916437674456663e-05, "loss": 0.0697, "step": 8430 }, { "grad_norm": 0.5259183645248413, "learning_rate": 9.916060967279308e-05, "loss": 0.0648, "step": 8440 }, { "grad_norm": 0.4677213728427887, "learning_rate": 9.91568342008159e-05, "loss": 0.0657, "step": 8450 }, { "grad_norm": 0.5903244018554688, "learning_rate": 9.915305032928019e-05, "loss": 0.0574, "step": 8460 }, { "grad_norm": 0.469819575548172, "learning_rate": 9.914925805883253e-05, "loss": 0.058, "step": 8470 }, { "grad_norm": 0.5146578550338745, "learning_rate": 9.914545739012088e-05, "loss": 0.0576, "step": 8480 }, { "grad_norm": 0.5542132258415222, "learning_rate": 9.91416483237947e-05, "loss": 0.0695, "step": 8490 }, { "grad_norm": 0.5050496459007263, "learning_rate": 9.913783086050485e-05, "loss": 0.0734, "step": 8500 }, { "grad_norm": 0.48250076174736023, "learning_rate": 9.913400500090364e-05, "loss": 0.0737, "step": 8510 }, { "grad_norm": 0.4737887382507324, "learning_rate": 9.913017074564479e-05, "loss": 0.0526, "step": 8520 }, { "grad_norm": 0.3287966847419739, "learning_rate": 9.912632809538348e-05, "loss": 0.0634, "step": 8530 }, { "grad_norm": 0.5023689866065979, "learning_rate": 9.912247705077629e-05, "loss": 0.0703, "step": 8540 }, { "grad_norm": 0.4029146134853363, "learning_rate": 9.911861761248127e-05, "loss": 0.0648, "step": 8550 }, { "grad_norm": 0.6474781036376953, "learning_rate": 9.91147497811579e-05, "loss": 0.0687, "step": 8560 }, { "grad_norm": 0.6075762510299683, "learning_rate": 9.911087355746709e-05, "loss": 0.1041, "step": 8570 }, { "grad_norm": 0.48518896102905273, "learning_rate": 9.910698894207117e-05, "loss": 0.08, "step": 8580 }, { "grad_norm": 0.607736349105835, "learning_rate": 9.910309593563392e-05, "loss": 0.0736, "step": 8590 }, { "grad_norm": 0.5066065788269043, "learning_rate": 9.909919453882057e-05, "loss": 0.0673, "step": 8600 }, { "grad_norm": 0.8320226669311523, "learning_rate": 9.90952847522977e-05, "loss": 0.0706, "step": 8610 }, { "grad_norm": 0.45556163787841797, "learning_rate": 9.909136657673346e-05, "loss": 0.0757, "step": 8620 }, { "grad_norm": 0.4976423680782318, "learning_rate": 9.908744001279731e-05, "loss": 0.0655, "step": 8630 }, { "grad_norm": 0.6781976819038391, "learning_rate": 9.90835050611602e-05, "loss": 0.0768, "step": 8640 }, { "grad_norm": 0.48412591218948364, "learning_rate": 9.90795617224945e-05, "loss": 0.0754, "step": 8650 }, { "grad_norm": 0.5757626891136169, "learning_rate": 9.907560999747405e-05, "loss": 0.0713, "step": 8660 }, { "grad_norm": 0.4444155991077423, "learning_rate": 9.907164988677408e-05, "loss": 0.0608, "step": 8670 }, { "grad_norm": 0.44934409856796265, "learning_rate": 9.906768139107124e-05, "loss": 0.0556, "step": 8680 }, { "grad_norm": 0.4141960144042969, "learning_rate": 9.906370451104367e-05, "loss": 0.0602, "step": 8690 }, { "grad_norm": 0.37551775574684143, "learning_rate": 9.905971924737088e-05, "loss": 0.0666, "step": 8700 }, { "grad_norm": 0.4737502336502075, "learning_rate": 9.905572560073387e-05, "loss": 0.0765, "step": 8710 }, { "grad_norm": 0.5856487154960632, "learning_rate": 9.905172357181501e-05, "loss": 0.0865, "step": 8720 }, { "grad_norm": 0.2980101406574249, "learning_rate": 9.904771316129817e-05, "loss": 0.0579, "step": 8730 }, { "grad_norm": 0.5436538457870483, "learning_rate": 9.904369436986862e-05, "loss": 0.0574, "step": 8740 }, { "grad_norm": 0.35644426941871643, "learning_rate": 9.903966719821303e-05, "loss": 0.068, "step": 8750 }, { "grad_norm": 0.4058797061443329, "learning_rate": 9.903563164701956e-05, "loss": 0.0542, "step": 8760 }, { "grad_norm": 0.5054779648780823, "learning_rate": 9.903158771697778e-05, "loss": 0.0644, "step": 8770 }, { "grad_norm": 0.7178212404251099, "learning_rate": 9.902753540877867e-05, "loss": 0.0734, "step": 8780 }, { "grad_norm": 0.3900347948074341, "learning_rate": 9.902347472311466e-05, "loss": 0.0599, "step": 8790 }, { "grad_norm": 0.47039103507995605, "learning_rate": 9.901940566067962e-05, "loss": 0.048, "step": 8800 }, { "grad_norm": 0.38962823152542114, "learning_rate": 9.901532822216883e-05, "loss": 0.0628, "step": 8810 }, { "grad_norm": 0.475521445274353, "learning_rate": 9.901124240827904e-05, "loss": 0.0593, "step": 8820 }, { "grad_norm": 0.40127477049827576, "learning_rate": 9.900714821970835e-05, "loss": 0.0556, "step": 8830 }, { "grad_norm": 0.45841333270072937, "learning_rate": 9.900304565715641e-05, "loss": 0.0588, "step": 8840 }, { "grad_norm": 0.37768831849098206, "learning_rate": 9.899893472132419e-05, "loss": 0.0532, "step": 8850 }, { "grad_norm": 0.5824100375175476, "learning_rate": 9.899481541291415e-05, "loss": 0.06, "step": 8860 }, { "grad_norm": 0.5343264937400818, "learning_rate": 9.899068773263016e-05, "loss": 0.0575, "step": 8870 }, { "grad_norm": 0.5022764801979065, "learning_rate": 9.898655168117754e-05, "loss": 0.0647, "step": 8880 }, { "grad_norm": 0.4645509421825409, "learning_rate": 9.898240725926302e-05, "loss": 0.0626, "step": 8890 }, { "grad_norm": 0.4692747890949249, "learning_rate": 9.897825446759478e-05, "loss": 0.0608, "step": 8900 }, { "grad_norm": 0.4010113775730133, "learning_rate": 9.897409330688241e-05, "loss": 0.0592, "step": 8910 }, { "grad_norm": 0.527428925037384, "learning_rate": 9.896992377783692e-05, "loss": 0.0732, "step": 8920 }, { "grad_norm": 0.3835698366165161, "learning_rate": 9.89657458811708e-05, "loss": 0.0582, "step": 8930 }, { "grad_norm": 0.368964284658432, "learning_rate": 9.896155961759792e-05, "loss": 0.0618, "step": 8940 }, { "grad_norm": 0.4996245205402374, "learning_rate": 9.895736498783361e-05, "loss": 0.0583, "step": 8950 }, { "grad_norm": 0.4140966832637787, "learning_rate": 9.895316199259462e-05, "loss": 0.0519, "step": 8960 }, { "grad_norm": 0.44503697752952576, "learning_rate": 9.894895063259909e-05, "loss": 0.0525, "step": 8970 }, { "grad_norm": 0.43931666016578674, "learning_rate": 9.894473090856667e-05, "loss": 0.0613, "step": 8980 }, { "grad_norm": 0.4806536138057709, "learning_rate": 9.894050282121839e-05, "loss": 0.0885, "step": 8990 }, { "grad_norm": 0.3592313528060913, "learning_rate": 9.893626637127668e-05, "loss": 0.0608, "step": 9000 }, { "grad_norm": 0.29472464323043823, "learning_rate": 9.893202155946546e-05, "loss": 0.0505, "step": 9010 }, { "grad_norm": 0.4832611680030823, "learning_rate": 9.892776838651006e-05, "loss": 0.0628, "step": 9020 }, { "grad_norm": 0.40021756291389465, "learning_rate": 9.892350685313722e-05, "loss": 0.076, "step": 9030 }, { "grad_norm": 0.4677134156227112, "learning_rate": 9.891923696007513e-05, "loss": 0.0679, "step": 9040 }, { "grad_norm": 0.41245388984680176, "learning_rate": 9.891495870805336e-05, "loss": 0.0607, "step": 9050 }, { "grad_norm": 0.49258971214294434, "learning_rate": 9.891067209780298e-05, "loss": 0.0683, "step": 9060 }, { "grad_norm": 0.42145854234695435, "learning_rate": 9.890637713005646e-05, "loss": 0.051, "step": 9070 }, { "grad_norm": 0.3889317512512207, "learning_rate": 9.890207380554767e-05, "loss": 0.055, "step": 9080 }, { "grad_norm": 0.4030986726284027, "learning_rate": 9.889776212501196e-05, "loss": 0.0648, "step": 9090 }, { "grad_norm": 0.4464252293109894, "learning_rate": 9.889344208918605e-05, "loss": 0.0709, "step": 9100 }, { "grad_norm": 0.37563812732696533, "learning_rate": 9.888911369880812e-05, "loss": 0.0586, "step": 9110 }, { "grad_norm": 0.3791442811489105, "learning_rate": 9.888477695461777e-05, "loss": 0.0557, "step": 9120 }, { "grad_norm": 0.48269936442375183, "learning_rate": 9.888043185735607e-05, "loss": 0.0618, "step": 9130 }, { "grad_norm": 0.4067701995372772, "learning_rate": 9.887607840776542e-05, "loss": 0.0481, "step": 9140 }, { "grad_norm": 0.44277694821357727, "learning_rate": 9.887171660658975e-05, "loss": 0.0734, "step": 9150 }, { "grad_norm": 0.4321296513080597, "learning_rate": 9.886734645457435e-05, "loss": 0.0527, "step": 9160 }, { "grad_norm": 0.6588608622550964, "learning_rate": 9.886296795246597e-05, "loss": 0.0556, "step": 9170 }, { "grad_norm": 0.4456315338611603, "learning_rate": 9.885858110101276e-05, "loss": 0.0598, "step": 9180 }, { "grad_norm": 0.2984200417995453, "learning_rate": 9.885418590096434e-05, "loss": 0.0421, "step": 9190 }, { "grad_norm": 0.28823786973953247, "learning_rate": 9.88497823530717e-05, "loss": 0.0561, "step": 9200 }, { "grad_norm": 0.4305708110332489, "learning_rate": 9.884537045808732e-05, "loss": 0.0706, "step": 9210 }, { "grad_norm": 0.4086962044239044, "learning_rate": 9.884095021676502e-05, "loss": 0.0675, "step": 9220 }, { "grad_norm": 0.39331817626953125, "learning_rate": 9.883652162986017e-05, "loss": 0.0572, "step": 9230 }, { "grad_norm": 0.40095391869544983, "learning_rate": 9.883208469812943e-05, "loss": 0.0655, "step": 9240 }, { "grad_norm": 0.35463371872901917, "learning_rate": 9.882763942233098e-05, "loss": 0.048, "step": 9250 }, { "grad_norm": 0.3443649709224701, "learning_rate": 9.882318580322441e-05, "loss": 0.0523, "step": 9260 }, { "grad_norm": 0.4392170310020447, "learning_rate": 9.881872384157067e-05, "loss": 0.0608, "step": 9270 }, { "grad_norm": 0.611570417881012, "learning_rate": 9.881425353813225e-05, "loss": 0.0627, "step": 9280 }, { "grad_norm": 0.6132785081863403, "learning_rate": 9.880977489367296e-05, "loss": 0.0715, "step": 9290 }, { "grad_norm": 0.39160194993019104, "learning_rate": 9.88052879089581e-05, "loss": 0.0725, "step": 9300 }, { "grad_norm": 0.9954825639724731, "learning_rate": 9.880079258475434e-05, "loss": 0.0684, "step": 9310 }, { "grad_norm": 0.5548824667930603, "learning_rate": 9.879628892182985e-05, "loss": 0.0849, "step": 9320 }, { "grad_norm": 0.7036041021347046, "learning_rate": 9.879177692095416e-05, "loss": 0.0736, "step": 9330 }, { "grad_norm": 0.4890304207801819, "learning_rate": 9.878725658289825e-05, "loss": 0.0644, "step": 9340 }, { "grad_norm": 0.3848324120044708, "learning_rate": 9.878272790843454e-05, "loss": 0.0614, "step": 9350 }, { "grad_norm": 0.7064926028251648, "learning_rate": 9.877819089833682e-05, "loss": 0.0566, "step": 9360 }, { "grad_norm": 0.2929384112358093, "learning_rate": 9.877364555338038e-05, "loss": 0.0813, "step": 9370 }, { "grad_norm": 0.4438900351524353, "learning_rate": 9.876909187434186e-05, "loss": 0.0616, "step": 9380 }, { "grad_norm": 0.41002926230430603, "learning_rate": 9.876452986199939e-05, "loss": 0.0617, "step": 9390 }, { "grad_norm": 0.42654868960380554, "learning_rate": 9.875995951713248e-05, "loss": 0.0681, "step": 9400 }, { "grad_norm": 0.3851604163646698, "learning_rate": 9.875538084052207e-05, "loss": 0.0679, "step": 9410 }, { "grad_norm": 0.5839819312095642, "learning_rate": 9.875079383295053e-05, "loss": 0.0615, "step": 9420 }, { "grad_norm": 0.37535834312438965, "learning_rate": 9.874619849520167e-05, "loss": 0.055, "step": 9430 }, { "grad_norm": 0.4313284158706665, "learning_rate": 9.874159482806069e-05, "loss": 0.061, "step": 9440 }, { "grad_norm": 0.4465057849884033, "learning_rate": 9.873698283231426e-05, "loss": 0.0566, "step": 9450 }, { "grad_norm": 0.5054265856742859, "learning_rate": 9.87323625087504e-05, "loss": 0.0564, "step": 9460 }, { "grad_norm": 0.48470455408096313, "learning_rate": 9.872773385815863e-05, "loss": 0.0796, "step": 9470 }, { "grad_norm": 0.314301460981369, "learning_rate": 9.872309688132986e-05, "loss": 0.0551, "step": 9480 }, { "grad_norm": 0.46387025713920593, "learning_rate": 9.871845157905639e-05, "loss": 0.0533, "step": 9490 }, { "grad_norm": 0.4440445005893707, "learning_rate": 9.871379795213201e-05, "loss": 0.0501, "step": 9500 }, { "grad_norm": 0.3328014314174652, "learning_rate": 9.87091360013519e-05, "loss": 0.0537, "step": 9510 }, { "grad_norm": 1.118445873260498, "learning_rate": 9.870446572751262e-05, "loss": 0.0632, "step": 9520 }, { "grad_norm": 0.3776147961616516, "learning_rate": 9.869978713141224e-05, "loss": 0.0629, "step": 9530 }, { "grad_norm": 0.5530865788459778, "learning_rate": 9.869510021385016e-05, "loss": 0.0783, "step": 9540 }, { "grad_norm": 0.44879865646362305, "learning_rate": 9.869040497562727e-05, "loss": 0.0704, "step": 9550 }, { "grad_norm": 0.4240950644016266, "learning_rate": 9.868570141754587e-05, "loss": 0.0805, "step": 9560 }, { "grad_norm": 0.38969799876213074, "learning_rate": 9.868098954040965e-05, "loss": 0.0795, "step": 9570 }, { "grad_norm": 0.5501499176025391, "learning_rate": 9.867626934502374e-05, "loss": 0.0678, "step": 9580 }, { "grad_norm": 0.5457320809364319, "learning_rate": 9.86715408321947e-05, "loss": 0.0641, "step": 9590 }, { "grad_norm": 0.3799542486667633, "learning_rate": 9.86668040027305e-05, "loss": 0.058, "step": 9600 }, { "grad_norm": 0.5004258155822754, "learning_rate": 9.866205885744053e-05, "loss": 0.0578, "step": 9610 }, { "grad_norm": 0.48555269837379456, "learning_rate": 9.865730539713563e-05, "loss": 0.0671, "step": 9620 }, { "grad_norm": 0.314377099275589, "learning_rate": 9.8652543622628e-05, "loss": 0.0511, "step": 9630 }, { "grad_norm": 0.4444960057735443, "learning_rate": 9.864777353473132e-05, "loss": 0.0497, "step": 9640 }, { "grad_norm": 0.4463218152523041, "learning_rate": 9.864299513426068e-05, "loss": 0.0604, "step": 9650 }, { "grad_norm": 0.39245522022247314, "learning_rate": 9.863820842203254e-05, "loss": 0.0543, "step": 9660 }, { "grad_norm": 0.3689267039299011, "learning_rate": 9.863341339886483e-05, "loss": 0.066, "step": 9670 }, { "grad_norm": 0.35837414860725403, "learning_rate": 9.86286100655769e-05, "loss": 0.0508, "step": 9680 }, { "grad_norm": 0.40552592277526855, "learning_rate": 9.862379842298953e-05, "loss": 0.0572, "step": 9690 }, { "grad_norm": 0.3705735504627228, "learning_rate": 9.861897847192485e-05, "loss": 0.0503, "step": 9700 }, { "grad_norm": 0.357021301984787, "learning_rate": 9.86141502132065e-05, "loss": 0.0515, "step": 9710 }, { "grad_norm": 0.5064004063606262, "learning_rate": 9.860931364765946e-05, "loss": 0.0592, "step": 9720 }, { "grad_norm": 0.5426502823829651, "learning_rate": 9.860446877611021e-05, "loss": 0.0691, "step": 9730 }, { "grad_norm": 0.44356194138526917, "learning_rate": 9.859961559938655e-05, "loss": 0.0582, "step": 9740 }, { "grad_norm": 0.32588568329811096, "learning_rate": 9.85947541183178e-05, "loss": 0.0531, "step": 9750 }, { "grad_norm": 0.42997390031814575, "learning_rate": 9.858988433373463e-05, "loss": 0.061, "step": 9760 }, { "grad_norm": 0.4561527967453003, "learning_rate": 9.858500624646918e-05, "loss": 0.0612, "step": 9770 }, { "grad_norm": 0.4193861186504364, "learning_rate": 9.858011985735497e-05, "loss": 0.0585, "step": 9780 }, { "grad_norm": 0.36507096886634827, "learning_rate": 9.857522516722693e-05, "loss": 0.0556, "step": 9790 }, { "grad_norm": 0.5014362931251526, "learning_rate": 9.857032217692145e-05, "loss": 0.0531, "step": 9800 }, { "grad_norm": 0.4673563539981842, "learning_rate": 9.856541088727631e-05, "loss": 0.0574, "step": 9810 }, { "grad_norm": 0.3289443552494049, "learning_rate": 9.856049129913072e-05, "loss": 0.0642, "step": 9820 }, { "grad_norm": 0.37803053855895996, "learning_rate": 9.85555634133253e-05, "loss": 0.0595, "step": 9830 }, { "grad_norm": 0.37217292189598083, "learning_rate": 9.855062723070208e-05, "loss": 0.064, "step": 9840 }, { "grad_norm": 0.30298665165901184, "learning_rate": 9.854568275210454e-05, "loss": 0.059, "step": 9850 }, { "grad_norm": 0.35944992303848267, "learning_rate": 9.854072997837754e-05, "loss": 0.0441, "step": 9860 }, { "grad_norm": 0.3305811285972595, "learning_rate": 9.853576891036737e-05, "loss": 0.0502, "step": 9870 }, { "grad_norm": 0.3859904706478119, "learning_rate": 9.853079954892177e-05, "loss": 0.0613, "step": 9880 }, { "grad_norm": 0.32416820526123047, "learning_rate": 9.852582189488983e-05, "loss": 0.0393, "step": 9890 }, { "grad_norm": 0.40665245056152344, "learning_rate": 9.852083594912212e-05, "loss": 0.0563, "step": 9900 }, { "grad_norm": 0.45879754424095154, "learning_rate": 9.851584171247058e-05, "loss": 0.0585, "step": 9910 }, { "grad_norm": 0.4367566406726837, "learning_rate": 9.851083918578863e-05, "loss": 0.0646, "step": 9920 }, { "grad_norm": 0.38245508074760437, "learning_rate": 9.850582836993103e-05, "loss": 0.0526, "step": 9930 }, { "grad_norm": 0.33901405334472656, "learning_rate": 9.850080926575397e-05, "loss": 0.0603, "step": 9940 }, { "grad_norm": 0.43150776624679565, "learning_rate": 9.849578187411515e-05, "loss": 0.0487, "step": 9950 }, { "grad_norm": 0.3199410140514374, "learning_rate": 9.849074619587354e-05, "loss": 0.0487, "step": 9960 }, { "grad_norm": 0.42479825019836426, "learning_rate": 9.848570223188964e-05, "loss": 0.0597, "step": 9970 }, { "grad_norm": 0.42261552810668945, "learning_rate": 9.848064998302531e-05, "loss": 0.0513, "step": 9980 }, { "grad_norm": 0.3781043291091919, "learning_rate": 9.847558945014386e-05, "loss": 0.0613, "step": 9990 }, { "grad_norm": 0.8051804900169373, "learning_rate": 9.847052063410996e-05, "loss": 0.0672, "step": 10000 }, { "grad_norm": 0.4796614348888397, "learning_rate": 9.846544353578977e-05, "loss": 0.0685, "step": 10010 }, { "grad_norm": 0.41267815232276917, "learning_rate": 9.846035815605081e-05, "loss": 0.0615, "step": 10020 }, { "grad_norm": 0.3564523160457611, "learning_rate": 9.845526449576204e-05, "loss": 0.0519, "step": 10030 }, { "grad_norm": 0.4223172068595886, "learning_rate": 9.845016255579383e-05, "loss": 0.0672, "step": 10040 }, { "grad_norm": 0.4274049401283264, "learning_rate": 9.844505233701794e-05, "loss": 0.0641, "step": 10050 }, { "grad_norm": 0.5422792434692383, "learning_rate": 9.843993384030757e-05, "loss": 0.0612, "step": 10060 }, { "grad_norm": 0.4139353036880493, "learning_rate": 9.843480706653737e-05, "loss": 0.0459, "step": 10070 }, { "grad_norm": 0.31284135580062866, "learning_rate": 9.84296720165833e-05, "loss": 0.0525, "step": 10080 }, { "grad_norm": 0.3706265091896057, "learning_rate": 9.842452869132286e-05, "loss": 0.0581, "step": 10090 }, { "grad_norm": 0.38582322001457214, "learning_rate": 9.841937709163489e-05, "loss": 0.0539, "step": 10100 }, { "grad_norm": 0.43325603008270264, "learning_rate": 9.841421721839962e-05, "loss": 0.0482, "step": 10110 }, { "grad_norm": 0.4851837754249573, "learning_rate": 9.840904907249879e-05, "loss": 0.0541, "step": 10120 }, { "grad_norm": 0.5633379220962524, "learning_rate": 9.840387265481545e-05, "loss": 0.063, "step": 10130 }, { "grad_norm": 0.37693798542022705, "learning_rate": 9.839868796623411e-05, "loss": 0.0572, "step": 10140 }, { "grad_norm": 0.3636018633842468, "learning_rate": 9.839349500764072e-05, "loss": 0.0632, "step": 10150 }, { "grad_norm": 0.47858402132987976, "learning_rate": 9.83882937799226e-05, "loss": 0.0545, "step": 10160 }, { "grad_norm": 0.32258859276771545, "learning_rate": 9.838308428396849e-05, "loss": 0.0525, "step": 10170 }, { "grad_norm": 0.34450310468673706, "learning_rate": 9.837786652066854e-05, "loss": 0.062, "step": 10180 }, { "grad_norm": 0.5141222476959229, "learning_rate": 9.837264049091437e-05, "loss": 0.0704, "step": 10190 }, { "grad_norm": 0.46510738134384155, "learning_rate": 9.836740619559893e-05, "loss": 0.0637, "step": 10200 }, { "grad_norm": 0.37966442108154297, "learning_rate": 9.836216363561659e-05, "loss": 0.0557, "step": 10210 }, { "grad_norm": 0.3814912736415863, "learning_rate": 9.835691281186322e-05, "loss": 0.0499, "step": 10220 }, { "grad_norm": 0.4218849241733551, "learning_rate": 9.8351653725236e-05, "loss": 0.0552, "step": 10230 }, { "grad_norm": 0.28761693835258484, "learning_rate": 9.83463863766336e-05, "loss": 0.0517, "step": 10240 }, { "grad_norm": 0.3422049283981323, "learning_rate": 9.834111076695602e-05, "loss": 0.0538, "step": 10250 }, { "grad_norm": 0.4954816401004791, "learning_rate": 9.833582689710477e-05, "loss": 0.0553, "step": 10260 }, { "grad_norm": 0.545324981212616, "learning_rate": 9.833053476798268e-05, "loss": 0.0915, "step": 10270 }, { "grad_norm": 0.5254679322242737, "learning_rate": 9.832523438049404e-05, "loss": 0.0779, "step": 10280 }, { "grad_norm": 0.4726778566837311, "learning_rate": 9.831992573554454e-05, "loss": 0.0743, "step": 10290 }, { "grad_norm": 0.44456255435943604, "learning_rate": 9.831460883404128e-05, "loss": 0.0557, "step": 10300 }, { "grad_norm": 0.5161610245704651, "learning_rate": 9.830928367689278e-05, "loss": 0.0505, "step": 10310 }, { "grad_norm": 0.6046238541603088, "learning_rate": 9.830395026500896e-05, "loss": 0.0655, "step": 10320 }, { "grad_norm": 0.5819413661956787, "learning_rate": 9.829860859930115e-05, "loss": 0.0621, "step": 10330 }, { "grad_norm": 0.6848849654197693, "learning_rate": 9.829325868068212e-05, "loss": 0.066, "step": 10340 }, { "grad_norm": 0.49049055576324463, "learning_rate": 9.8287900510066e-05, "loss": 0.0735, "step": 10350 }, { "grad_norm": 0.4393662214279175, "learning_rate": 9.828253408836834e-05, "loss": 0.0487, "step": 10360 }, { "grad_norm": 0.3037395775318146, "learning_rate": 9.827715941650615e-05, "loss": 0.0566, "step": 10370 }, { "grad_norm": 0.5625744462013245, "learning_rate": 9.82717764953978e-05, "loss": 0.0613, "step": 10380 }, { "grad_norm": 0.4202463626861572, "learning_rate": 9.826638532596308e-05, "loss": 0.0574, "step": 10390 }, { "grad_norm": 0.6618595123291016, "learning_rate": 9.82609859091232e-05, "loss": 0.0596, "step": 10400 }, { "grad_norm": 0.39835649728775024, "learning_rate": 9.825557824580076e-05, "loss": 0.0629, "step": 10410 }, { "grad_norm": 0.4026983380317688, "learning_rate": 9.82501623369198e-05, "loss": 0.0605, "step": 10420 }, { "grad_norm": 0.42086079716682434, "learning_rate": 9.824473818340574e-05, "loss": 0.0523, "step": 10430 }, { "grad_norm": 0.593402624130249, "learning_rate": 9.823930578618541e-05, "loss": 0.0509, "step": 10440 }, { "grad_norm": 0.41503968834877014, "learning_rate": 9.823386514618709e-05, "loss": 0.0633, "step": 10450 }, { "grad_norm": 0.3603450059890747, "learning_rate": 9.82284162643404e-05, "loss": 0.0551, "step": 10460 }, { "grad_norm": 0.4364190101623535, "learning_rate": 9.822295914157642e-05, "loss": 0.0543, "step": 10470 }, { "grad_norm": 0.3579353988170624, "learning_rate": 9.821749377882763e-05, "loss": 0.0518, "step": 10480 }, { "grad_norm": 0.4199584722518921, "learning_rate": 9.821202017702791e-05, "loss": 0.0497, "step": 10490 }, { "grad_norm": 0.29177650809288025, "learning_rate": 9.820653833711253e-05, "loss": 0.0468, "step": 10500 }, { "grad_norm": 0.410020649433136, "learning_rate": 9.820104826001822e-05, "loss": 0.0639, "step": 10510 }, { "grad_norm": 0.4553353190422058, "learning_rate": 9.819554994668305e-05, "loss": 0.0462, "step": 10520 }, { "grad_norm": 0.33853909373283386, "learning_rate": 9.819004339804654e-05, "loss": 0.0539, "step": 10530 }, { "grad_norm": 0.37819913029670715, "learning_rate": 9.818452861504961e-05, "loss": 0.0501, "step": 10540 }, { "grad_norm": 0.4541976749897003, "learning_rate": 9.81790055986346e-05, "loss": 0.0454, "step": 10550 }, { "grad_norm": 0.38521143794059753, "learning_rate": 9.817347434974523e-05, "loss": 0.0414, "step": 10560 }, { "grad_norm": 0.3512662947177887, "learning_rate": 9.816793486932664e-05, "loss": 0.0465, "step": 10570 }, { "grad_norm": 0.4431373178958893, "learning_rate": 9.816238715832538e-05, "loss": 0.0507, "step": 10580 }, { "grad_norm": 0.5677350759506226, "learning_rate": 9.815683121768939e-05, "loss": 0.0512, "step": 10590 }, { "grad_norm": 0.37406930327415466, "learning_rate": 9.815126704836804e-05, "loss": 0.07, "step": 10600 }, { "grad_norm": 0.3797992467880249, "learning_rate": 9.81456946513121e-05, "loss": 0.0793, "step": 10610 }, { "grad_norm": 0.4703540503978729, "learning_rate": 9.814011402747373e-05, "loss": 0.0734, "step": 10620 }, { "grad_norm": 0.45814311504364014, "learning_rate": 9.813452517780651e-05, "loss": 0.0534, "step": 10630 }, { "grad_norm": 0.37347668409347534, "learning_rate": 9.81289281032654e-05, "loss": 0.0723, "step": 10640 }, { "grad_norm": 0.2956002354621887, "learning_rate": 9.812332280480683e-05, "loss": 0.0663, "step": 10650 }, { "grad_norm": 0.4870022237300873, "learning_rate": 9.811770928338854e-05, "loss": 0.0609, "step": 10660 }, { "grad_norm": 0.5221078991889954, "learning_rate": 9.811208753996979e-05, "loss": 0.0649, "step": 10670 }, { "grad_norm": 0.3970504105091095, "learning_rate": 9.810645757551113e-05, "loss": 0.0444, "step": 10680 }, { "grad_norm": 0.3266879916191101, "learning_rate": 9.810081939097459e-05, "loss": 0.0473, "step": 10690 }, { "grad_norm": 0.4438785910606384, "learning_rate": 9.809517298732356e-05, "loss": 0.0703, "step": 10700 }, { "grad_norm": 0.3628922998905182, "learning_rate": 9.80895183655229e-05, "loss": 0.0596, "step": 10710 }, { "grad_norm": 0.32209640741348267, "learning_rate": 9.808385552653877e-05, "loss": 0.0526, "step": 10720 }, { "grad_norm": 0.47348612546920776, "learning_rate": 9.807818447133886e-05, "loss": 0.0506, "step": 10730 }, { "grad_norm": 0.38248512148857117, "learning_rate": 9.807250520089215e-05, "loss": 0.0467, "step": 10740 }, { "grad_norm": 0.43018755316734314, "learning_rate": 9.806681771616908e-05, "loss": 0.053, "step": 10750 }, { "grad_norm": 0.35842329263687134, "learning_rate": 9.80611220181415e-05, "loss": 0.0583, "step": 10760 }, { "grad_norm": 0.551926851272583, "learning_rate": 9.805541810778264e-05, "loss": 0.0587, "step": 10770 }, { "grad_norm": 0.4876195192337036, "learning_rate": 9.804970598606716e-05, "loss": 0.0587, "step": 10780 }, { "grad_norm": 0.3790760636329651, "learning_rate": 9.804398565397106e-05, "loss": 0.0514, "step": 10790 }, { "grad_norm": 0.40555423498153687, "learning_rate": 9.803825711247183e-05, "loss": 0.0552, "step": 10800 }, { "grad_norm": 0.4803519546985626, "learning_rate": 9.803252036254831e-05, "loss": 0.0477, "step": 10810 }, { "grad_norm": 0.34519699215888977, "learning_rate": 9.802677540518076e-05, "loss": 0.0528, "step": 10820 }, { "grad_norm": 0.25213822722435, "learning_rate": 9.802102224135081e-05, "loss": 0.0513, "step": 10830 }, { "grad_norm": 0.28844591975212097, "learning_rate": 9.801526087204155e-05, "loss": 0.0615, "step": 10840 }, { "grad_norm": 0.4602600038051605, "learning_rate": 9.800949129823743e-05, "loss": 0.0562, "step": 10850 }, { "grad_norm": 0.33096882700920105, "learning_rate": 9.80037135209243e-05, "loss": 0.0519, "step": 10860 }, { "grad_norm": 0.3316621482372284, "learning_rate": 9.799792754108946e-05, "loss": 0.0535, "step": 10870 }, { "grad_norm": 0.4420135021209717, "learning_rate": 9.799213335972152e-05, "loss": 0.0532, "step": 10880 }, { "grad_norm": 0.413678914308548, "learning_rate": 9.798633097781058e-05, "loss": 0.0491, "step": 10890 }, { "grad_norm": 0.4630952477455139, "learning_rate": 9.79805203963481e-05, "loss": 0.0615, "step": 10900 }, { "grad_norm": 0.3789059519767761, "learning_rate": 9.797470161632697e-05, "loss": 0.0527, "step": 10910 }, { "grad_norm": 0.5872753858566284, "learning_rate": 9.796887463874145e-05, "loss": 0.0535, "step": 10920 }, { "grad_norm": 0.3587530553340912, "learning_rate": 9.796303946458718e-05, "loss": 0.0626, "step": 10930 }, { "grad_norm": 0.4736200273036957, "learning_rate": 9.795719609486127e-05, "loss": 0.0682, "step": 10940 }, { "grad_norm": 0.42546650767326355, "learning_rate": 9.795134453056219e-05, "loss": 0.0533, "step": 10950 }, { "grad_norm": 0.31570160388946533, "learning_rate": 9.794548477268979e-05, "loss": 0.0493, "step": 10960 }, { "grad_norm": 0.3997395932674408, "learning_rate": 9.793961682224537e-05, "loss": 0.0455, "step": 10970 }, { "grad_norm": 0.3061942458152771, "learning_rate": 9.793374068023156e-05, "loss": 0.0476, "step": 10980 }, { "grad_norm": 0.30890145897865295, "learning_rate": 9.792785634765247e-05, "loss": 0.0479, "step": 10990 }, { "grad_norm": 0.35957232117652893, "learning_rate": 9.792196382551357e-05, "loss": 0.0533, "step": 11000 }, { "grad_norm": 0.2516695261001587, "learning_rate": 9.791606311482171e-05, "loss": 0.0495, "step": 11010 }, { "grad_norm": 0.3758942484855652, "learning_rate": 9.791015421658518e-05, "loss": 0.0539, "step": 11020 }, { "grad_norm": 0.3816754221916199, "learning_rate": 9.790423713181362e-05, "loss": 0.0512, "step": 11030 }, { "grad_norm": 0.3399656414985657, "learning_rate": 9.789831186151814e-05, "loss": 0.0465, "step": 11040 }, { "grad_norm": 0.4057309925556183, "learning_rate": 9.789237840671118e-05, "loss": 0.0442, "step": 11050 }, { "grad_norm": 0.37690800428390503, "learning_rate": 9.78864367684066e-05, "loss": 0.0525, "step": 11060 }, { "grad_norm": 0.44064760208129883, "learning_rate": 9.788048694761968e-05, "loss": 0.0473, "step": 11070 }, { "grad_norm": 0.48832985758781433, "learning_rate": 9.787452894536709e-05, "loss": 0.0515, "step": 11080 }, { "grad_norm": 0.3466123640537262, "learning_rate": 9.786856276266685e-05, "loss": 0.0498, "step": 11090 }, { "grad_norm": 0.3688209056854248, "learning_rate": 9.786258840053845e-05, "loss": 0.0551, "step": 11100 }, { "grad_norm": 0.38671109080314636, "learning_rate": 9.785660586000273e-05, "loss": 0.0566, "step": 11110 }, { "grad_norm": 0.4447542726993561, "learning_rate": 9.785061514208196e-05, "loss": 0.0486, "step": 11120 }, { "grad_norm": 0.3954307436943054, "learning_rate": 9.784461624779977e-05, "loss": 0.0601, "step": 11130 }, { "grad_norm": 0.40413615107536316, "learning_rate": 9.783860917818123e-05, "loss": 0.052, "step": 11140 }, { "grad_norm": 0.30928564071655273, "learning_rate": 9.783259393425277e-05, "loss": 0.0381, "step": 11150 }, { "grad_norm": 0.46652141213417053, "learning_rate": 9.782657051704221e-05, "loss": 0.0509, "step": 11160 }, { "grad_norm": 0.35088953375816345, "learning_rate": 9.782053892757883e-05, "loss": 0.0436, "step": 11170 }, { "grad_norm": 0.4107770621776581, "learning_rate": 9.781449916689324e-05, "loss": 0.0486, "step": 11180 }, { "grad_norm": 0.3104659616947174, "learning_rate": 9.780845123601746e-05, "loss": 0.0563, "step": 11190 }, { "grad_norm": 0.35207751393318176, "learning_rate": 9.780239513598492e-05, "loss": 0.0535, "step": 11200 }, { "grad_norm": 0.4393053352832794, "learning_rate": 9.779633086783047e-05, "loss": 0.0607, "step": 11210 }, { "grad_norm": 0.329891175031662, "learning_rate": 9.779025843259031e-05, "loss": 0.0432, "step": 11220 }, { "grad_norm": 0.3808453381061554, "learning_rate": 9.778417783130204e-05, "loss": 0.058, "step": 11230 }, { "grad_norm": 0.27698737382888794, "learning_rate": 9.777808906500468e-05, "loss": 0.0425, "step": 11240 }, { "grad_norm": 0.3432190418243408, "learning_rate": 9.777199213473862e-05, "loss": 0.0373, "step": 11250 }, { "grad_norm": 0.5300766229629517, "learning_rate": 9.77658870415457e-05, "loss": 0.0406, "step": 11260 }, { "grad_norm": 0.38180699944496155, "learning_rate": 9.775977378646906e-05, "loss": 0.0441, "step": 11270 }, { "grad_norm": 0.4445250332355499, "learning_rate": 9.775365237055331e-05, "loss": 0.0578, "step": 11280 }, { "grad_norm": 0.43648383021354675, "learning_rate": 9.774752279484445e-05, "loss": 0.0489, "step": 11290 }, { "grad_norm": 0.35968217253685, "learning_rate": 9.774138506038984e-05, "loss": 0.0492, "step": 11300 }, { "grad_norm": 0.4454306960105896, "learning_rate": 9.773523916823826e-05, "loss": 0.045, "step": 11310 }, { "grad_norm": 0.24243761599063873, "learning_rate": 9.772908511943986e-05, "loss": 0.0471, "step": 11320 }, { "grad_norm": 0.29683947563171387, "learning_rate": 9.77229229150462e-05, "loss": 0.0448, "step": 11330 }, { "grad_norm": 0.4098474085330963, "learning_rate": 9.771675255611024e-05, "loss": 0.0397, "step": 11340 }, { "grad_norm": 0.2652212083339691, "learning_rate": 9.771057404368632e-05, "loss": 0.0661, "step": 11350 }, { "grad_norm": 0.4087679386138916, "learning_rate": 9.770438737883018e-05, "loss": 0.0472, "step": 11360 }, { "grad_norm": 0.2738187909126282, "learning_rate": 9.769819256259898e-05, "loss": 0.0401, "step": 11370 }, { "grad_norm": 0.24931451678276062, "learning_rate": 9.769198959605119e-05, "loss": 0.0453, "step": 11380 }, { "grad_norm": 0.44463345408439636, "learning_rate": 9.768577848024678e-05, "loss": 0.0527, "step": 11390 }, { "grad_norm": 0.3247741162776947, "learning_rate": 9.767955921624702e-05, "loss": 0.0532, "step": 11400 }, { "grad_norm": 0.5179909467697144, "learning_rate": 9.767333180511465e-05, "loss": 0.0502, "step": 11410 }, { "grad_norm": 0.2926132082939148, "learning_rate": 9.766709624791373e-05, "loss": 0.0618, "step": 11420 }, { "grad_norm": 0.35456693172454834, "learning_rate": 9.766085254570975e-05, "loss": 0.0478, "step": 11430 }, { "grad_norm": 0.4057016968727112, "learning_rate": 9.76546006995696e-05, "loss": 0.0493, "step": 11440 }, { "grad_norm": 0.4728052318096161, "learning_rate": 9.764834071056155e-05, "loss": 0.0483, "step": 11450 }, { "grad_norm": 0.3765319287776947, "learning_rate": 9.764207257975526e-05, "loss": 0.0606, "step": 11460 }, { "grad_norm": 0.4745025336742401, "learning_rate": 9.763579630822179e-05, "loss": 0.0458, "step": 11470 }, { "grad_norm": 0.40866968035697937, "learning_rate": 9.762951189703356e-05, "loss": 0.0431, "step": 11480 }, { "grad_norm": 0.4842316210269928, "learning_rate": 9.762321934726442e-05, "loss": 0.0634, "step": 11490 }, { "grad_norm": 0.38931921124458313, "learning_rate": 9.761691865998959e-05, "loss": 0.0507, "step": 11500 }, { "grad_norm": 0.3504541516304016, "learning_rate": 9.76106098362857e-05, "loss": 0.041, "step": 11510 }, { "grad_norm": 0.3195827305316925, "learning_rate": 9.760429287723072e-05, "loss": 0.0472, "step": 11520 }, { "grad_norm": 0.43164175748825073, "learning_rate": 9.759796778390406e-05, "loss": 0.0429, "step": 11530 }, { "grad_norm": 0.35880109667778015, "learning_rate": 9.759163455738653e-05, "loss": 0.0535, "step": 11540 }, { "grad_norm": 0.45138102769851685, "learning_rate": 9.75852931987603e-05, "loss": 0.0503, "step": 11550 }, { "grad_norm": 0.3441612720489502, "learning_rate": 9.757894370910891e-05, "loss": 0.0508, "step": 11560 }, { "grad_norm": 0.40289759635925293, "learning_rate": 9.757258608951733e-05, "loss": 0.048, "step": 11570 }, { "grad_norm": 0.4482748210430145, "learning_rate": 9.75662203410719e-05, "loss": 0.0552, "step": 11580 }, { "grad_norm": 0.5680791735649109, "learning_rate": 9.755984646486034e-05, "loss": 0.0607, "step": 11590 }, { "grad_norm": 0.3614182770252228, "learning_rate": 9.75534644619718e-05, "loss": 0.049, "step": 11600 }, { "grad_norm": 0.35054656863212585, "learning_rate": 9.754707433349676e-05, "loss": 0.0696, "step": 11610 }, { "grad_norm": 0.46845561265945435, "learning_rate": 9.754067608052715e-05, "loss": 0.0696, "step": 11620 }, { "grad_norm": 0.5574074983596802, "learning_rate": 9.753426970415622e-05, "loss": 0.0752, "step": 11630 }, { "grad_norm": 0.3283177316188812, "learning_rate": 9.752785520547868e-05, "loss": 0.0515, "step": 11640 }, { "grad_norm": 0.5058126449584961, "learning_rate": 9.752143258559056e-05, "loss": 0.0607, "step": 11650 }, { "grad_norm": 0.32814261317253113, "learning_rate": 9.751500184558933e-05, "loss": 0.0522, "step": 11660 }, { "grad_norm": 0.2701498568058014, "learning_rate": 9.750856298657383e-05, "loss": 0.038, "step": 11670 }, { "grad_norm": 0.422206848859787, "learning_rate": 9.750211600964428e-05, "loss": 0.0556, "step": 11680 }, { "grad_norm": 0.4099365770816803, "learning_rate": 9.749566091590226e-05, "loss": 0.0469, "step": 11690 }, { "grad_norm": 0.38851532340049744, "learning_rate": 9.748919770645083e-05, "loss": 0.0475, "step": 11700 }, { "grad_norm": 0.3925086259841919, "learning_rate": 9.748272638239432e-05, "loss": 0.0417, "step": 11710 }, { "grad_norm": 0.40544864535331726, "learning_rate": 9.747624694483855e-05, "loss": 0.0487, "step": 11720 }, { "grad_norm": 0.27964988350868225, "learning_rate": 9.746975939489065e-05, "loss": 0.0406, "step": 11730 }, { "grad_norm": 0.4083704650402069, "learning_rate": 9.746326373365918e-05, "loss": 0.0524, "step": 11740 }, { "grad_norm": 0.2744259536266327, "learning_rate": 9.745675996225403e-05, "loss": 0.0535, "step": 11750 }, { "grad_norm": 0.3350454270839691, "learning_rate": 9.745024808178657e-05, "loss": 0.0441, "step": 11760 }, { "grad_norm": 0.3450547456741333, "learning_rate": 9.744372809336947e-05, "loss": 0.051, "step": 11770 }, { "grad_norm": 0.38093647360801697, "learning_rate": 9.743719999811682e-05, "loss": 0.0456, "step": 11780 }, { "grad_norm": 0.40405604243278503, "learning_rate": 9.743066379714412e-05, "loss": 0.0445, "step": 11790 }, { "grad_norm": 0.3141390383243561, "learning_rate": 9.74241194915682e-05, "loss": 0.0519, "step": 11800 }, { "grad_norm": 0.3620279133319855, "learning_rate": 9.741756708250731e-05, "loss": 0.0669, "step": 11810 }, { "grad_norm": 0.5243074297904968, "learning_rate": 9.741100657108109e-05, "loss": 0.0525, "step": 11820 }, { "grad_norm": 0.44831088185310364, "learning_rate": 9.740443795841054e-05, "loss": 0.0652, "step": 11830 }, { "grad_norm": 0.4359022378921509, "learning_rate": 9.739786124561805e-05, "loss": 0.0465, "step": 11840 }, { "grad_norm": 0.29912322759628296, "learning_rate": 9.73912764338274e-05, "loss": 0.0512, "step": 11850 }, { "grad_norm": 0.32577425241470337, "learning_rate": 9.738468352416377e-05, "loss": 0.055, "step": 11860 }, { "grad_norm": 0.37528717517852783, "learning_rate": 9.737808251775369e-05, "loss": 0.0454, "step": 11870 }, { "grad_norm": 0.3490842580795288, "learning_rate": 9.737147341572512e-05, "loss": 0.0628, "step": 11880 }, { "grad_norm": 0.2929342985153198, "learning_rate": 9.736485621920735e-05, "loss": 0.0449, "step": 11890 }, { "grad_norm": 0.4169508218765259, "learning_rate": 9.735823092933108e-05, "loss": 0.0494, "step": 11900 }, { "grad_norm": 0.3333892226219177, "learning_rate": 9.735159754722838e-05, "loss": 0.0483, "step": 11910 }, { "grad_norm": 0.4453542232513428, "learning_rate": 9.734495607403275e-05, "loss": 0.0434, "step": 11920 }, { "grad_norm": 0.4032983183860779, "learning_rate": 9.733830651087901e-05, "loss": 0.0524, "step": 11930 }, { "grad_norm": 0.39519891142845154, "learning_rate": 9.733164885890338e-05, "loss": 0.0558, "step": 11940 }, { "grad_norm": 0.28160184621810913, "learning_rate": 9.732498311924349e-05, "loss": 0.0366, "step": 11950 }, { "grad_norm": 0.46335598826408386, "learning_rate": 9.731830929303833e-05, "loss": 0.0407, "step": 11960 }, { "grad_norm": 0.27027198672294617, "learning_rate": 9.731162738142827e-05, "loss": 0.0433, "step": 11970 }, { "grad_norm": 0.5355331301689148, "learning_rate": 9.730493738555506e-05, "loss": 0.0541, "step": 11980 }, { "grad_norm": 0.420949250459671, "learning_rate": 9.729823930656186e-05, "loss": 0.0509, "step": 11990 }, { "grad_norm": 0.39752450585365295, "learning_rate": 9.729153314559316e-05, "loss": 0.0587, "step": 12000 }, { "grad_norm": 0.36561131477355957, "learning_rate": 9.728481890379486e-05, "loss": 0.0491, "step": 12010 }, { "grad_norm": 0.38618117570877075, "learning_rate": 9.727809658231428e-05, "loss": 0.0427, "step": 12020 }, { "grad_norm": 0.3714081943035126, "learning_rate": 9.727136618230003e-05, "loss": 0.0471, "step": 12030 }, { "grad_norm": 0.4857077896595001, "learning_rate": 9.726462770490219e-05, "loss": 0.0551, "step": 12040 }, { "grad_norm": 0.40279248356819153, "learning_rate": 9.725788115127214e-05, "loss": 0.0555, "step": 12050 }, { "grad_norm": 0.3974028527736664, "learning_rate": 9.725112652256274e-05, "loss": 0.0557, "step": 12060 }, { "grad_norm": 0.39018750190734863, "learning_rate": 9.724436381992812e-05, "loss": 0.0704, "step": 12070 }, { "grad_norm": 0.5172232389450073, "learning_rate": 9.723759304452387e-05, "loss": 0.0501, "step": 12080 }, { "grad_norm": 0.3403893709182739, "learning_rate": 9.72308141975069e-05, "loss": 0.0408, "step": 12090 }, { "grad_norm": 0.33700233697891235, "learning_rate": 9.722402728003557e-05, "loss": 0.0632, "step": 12100 }, { "grad_norm": 0.39765647053718567, "learning_rate": 9.721723229326953e-05, "loss": 0.0567, "step": 12110 }, { "grad_norm": 0.4320118725299835, "learning_rate": 9.721042923836992e-05, "loss": 0.0449, "step": 12120 }, { "grad_norm": 0.3517743945121765, "learning_rate": 9.720361811649914e-05, "loss": 0.0534, "step": 12130 }, { "grad_norm": 0.3343031704425812, "learning_rate": 9.719679892882106e-05, "loss": 0.0416, "step": 12140 }, { "grad_norm": 0.4044232666492462, "learning_rate": 9.718997167650085e-05, "loss": 0.0502, "step": 12150 }, { "grad_norm": 0.46273311972618103, "learning_rate": 9.718313636070515e-05, "loss": 0.0445, "step": 12160 }, { "grad_norm": 0.6641513705253601, "learning_rate": 9.717629298260192e-05, "loss": 0.0532, "step": 12170 }, { "grad_norm": 0.4261293411254883, "learning_rate": 9.716944154336047e-05, "loss": 0.0634, "step": 12180 }, { "grad_norm": 0.5187462568283081, "learning_rate": 9.716258204415157e-05, "loss": 0.0478, "step": 12190 }, { "grad_norm": 0.34695589542388916, "learning_rate": 9.715571448614728e-05, "loss": 0.0611, "step": 12200 }, { "grad_norm": 0.3999379873275757, "learning_rate": 9.71488388705211e-05, "loss": 0.069, "step": 12210 }, { "grad_norm": 0.564769446849823, "learning_rate": 9.714195519844788e-05, "loss": 0.0543, "step": 12220 }, { "grad_norm": 0.4212218225002289, "learning_rate": 9.713506347110386e-05, "loss": 0.0539, "step": 12230 }, { "grad_norm": 0.40515780448913574, "learning_rate": 9.712816368966662e-05, "loss": 0.0552, "step": 12240 }, { "grad_norm": 0.5163933634757996, "learning_rate": 9.712125585531517e-05, "loss": 0.0634, "step": 12250 }, { "grad_norm": 0.5299263596534729, "learning_rate": 9.711433996922988e-05, "loss": 0.0743, "step": 12260 }, { "grad_norm": 0.290642648935318, "learning_rate": 9.710741603259245e-05, "loss": 0.062, "step": 12270 }, { "grad_norm": 0.44428151845932007, "learning_rate": 9.710048404658603e-05, "loss": 0.0518, "step": 12280 }, { "grad_norm": 0.6352130174636841, "learning_rate": 9.709354401239508e-05, "loss": 0.0783, "step": 12290 }, { "grad_norm": 0.7808935642242432, "learning_rate": 9.708659593120546e-05, "loss": 0.0632, "step": 12300 }, { "grad_norm": 0.489632785320282, "learning_rate": 9.707963980420443e-05, "loss": 0.0605, "step": 12310 }, { "grad_norm": 0.44108152389526367, "learning_rate": 9.707267563258058e-05, "loss": 0.0661, "step": 12320 }, { "grad_norm": 0.37995225191116333, "learning_rate": 9.70657034175239e-05, "loss": 0.056, "step": 12330 }, { "grad_norm": 0.40711674094200134, "learning_rate": 9.705872316022577e-05, "loss": 0.0646, "step": 12340 }, { "grad_norm": 0.4258944094181061, "learning_rate": 9.705173486187891e-05, "loss": 0.0574, "step": 12350 }, { "grad_norm": 0.5900169610977173, "learning_rate": 9.704473852367741e-05, "loss": 0.0638, "step": 12360 }, { "grad_norm": 0.3915468156337738, "learning_rate": 9.70377341468168e-05, "loss": 0.0525, "step": 12370 }, { "grad_norm": 0.4068055748939514, "learning_rate": 9.703072173249389e-05, "loss": 0.0603, "step": 12380 }, { "grad_norm": 0.5325718522071838, "learning_rate": 9.702370128190693e-05, "loss": 0.0655, "step": 12390 }, { "grad_norm": 0.6486563682556152, "learning_rate": 9.701667279625552e-05, "loss": 0.0675, "step": 12400 }, { "grad_norm": 0.5146239399909973, "learning_rate": 9.700963627674065e-05, "loss": 0.0706, "step": 12410 }, { "grad_norm": 0.4796817898750305, "learning_rate": 9.700259172456466e-05, "loss": 0.0523, "step": 12420 }, { "grad_norm": 0.5186824202537537, "learning_rate": 9.699553914093124e-05, "loss": 0.0699, "step": 12430 }, { "grad_norm": 0.488886296749115, "learning_rate": 9.698847852704553e-05, "loss": 0.0641, "step": 12440 }, { "grad_norm": 0.49403050541877747, "learning_rate": 9.6981409884114e-05, "loss": 0.0519, "step": 12450 }, { "grad_norm": 0.5535376667976379, "learning_rate": 9.697433321334443e-05, "loss": 0.0541, "step": 12460 }, { "grad_norm": 0.26088446378707886, "learning_rate": 9.696724851594607e-05, "loss": 0.0524, "step": 12470 }, { "grad_norm": 0.386470764875412, "learning_rate": 9.696015579312952e-05, "loss": 0.042, "step": 12480 }, { "grad_norm": 0.3871738910675049, "learning_rate": 9.695305504610668e-05, "loss": 0.0423, "step": 12490 }, { "grad_norm": 0.3796779215335846, "learning_rate": 9.694594627609092e-05, "loss": 0.0497, "step": 12500 }, { "grad_norm": 0.47292065620422363, "learning_rate": 9.693882948429691e-05, "loss": 0.0661, "step": 12510 }, { "grad_norm": 0.4423923194408417, "learning_rate": 9.693170467194071e-05, "loss": 0.0545, "step": 12520 }, { "grad_norm": 0.44858020544052124, "learning_rate": 9.692457184023977e-05, "loss": 0.0538, "step": 12530 }, { "grad_norm": 0.4580138027667999, "learning_rate": 9.691743099041291e-05, "loss": 0.0621, "step": 12540 }, { "grad_norm": 0.5140674114227295, "learning_rate": 9.691028212368027e-05, "loss": 0.059, "step": 12550 }, { "grad_norm": 0.4929206669330597, "learning_rate": 9.690312524126342e-05, "loss": 0.0488, "step": 12560 }, { "grad_norm": 0.373826265335083, "learning_rate": 9.689596034438527e-05, "loss": 0.0494, "step": 12570 }, { "grad_norm": 0.3065679967403412, "learning_rate": 9.688878743427012e-05, "loss": 0.0526, "step": 12580 }, { "grad_norm": 0.35836905241012573, "learning_rate": 9.688160651214359e-05, "loss": 0.0604, "step": 12590 }, { "grad_norm": 0.48493146896362305, "learning_rate": 9.687441757923273e-05, "loss": 0.052, "step": 12600 }, { "grad_norm": 0.47386741638183594, "learning_rate": 9.68672206367659e-05, "loss": 0.0439, "step": 12610 }, { "grad_norm": 0.5049934387207031, "learning_rate": 9.686001568597291e-05, "loss": 0.0485, "step": 12620 }, { "grad_norm": 0.3353686034679413, "learning_rate": 9.685280272808486e-05, "loss": 0.0448, "step": 12630 }, { "grad_norm": 0.5205981135368347, "learning_rate": 9.684558176433424e-05, "loss": 0.0648, "step": 12640 }, { "grad_norm": 0.3817133605480194, "learning_rate": 9.683835279595495e-05, "loss": 0.0462, "step": 12650 }, { "grad_norm": 0.48926272988319397, "learning_rate": 9.683111582418216e-05, "loss": 0.0613, "step": 12660 }, { "grad_norm": 0.3774307072162628, "learning_rate": 9.682387085025254e-05, "loss": 0.0509, "step": 12670 }, { "grad_norm": 0.3498780131340027, "learning_rate": 9.681661787540401e-05, "loss": 0.0544, "step": 12680 }, { "grad_norm": 0.3870497941970825, "learning_rate": 9.680935690087593e-05, "loss": 0.0487, "step": 12690 }, { "grad_norm": 0.5100987553596497, "learning_rate": 9.680208792790901e-05, "loss": 0.0505, "step": 12700 }, { "grad_norm": 0.31505775451660156, "learning_rate": 9.679481095774529e-05, "loss": 0.0447, "step": 12710 }, { "grad_norm": 0.3572067618370056, "learning_rate": 9.678752599162822e-05, "loss": 0.0431, "step": 12720 }, { "grad_norm": 0.39237523078918457, "learning_rate": 9.678023303080259e-05, "loss": 0.0406, "step": 12730 }, { "grad_norm": 0.34737834334373474, "learning_rate": 9.677293207651459e-05, "loss": 0.0539, "step": 12740 }, { "grad_norm": 0.30137869715690613, "learning_rate": 9.676562313001173e-05, "loss": 0.0395, "step": 12750 }, { "grad_norm": 0.5215184688568115, "learning_rate": 9.675830619254293e-05, "loss": 0.057, "step": 12760 }, { "grad_norm": 0.41309282183647156, "learning_rate": 9.675098126535843e-05, "loss": 0.0607, "step": 12770 }, { "grad_norm": 0.5164836645126343, "learning_rate": 9.674364834970988e-05, "loss": 0.0546, "step": 12780 }, { "grad_norm": 0.36328643560409546, "learning_rate": 9.673630744685028e-05, "loss": 0.0588, "step": 12790 }, { "grad_norm": 0.3637029230594635, "learning_rate": 9.672895855803397e-05, "loss": 0.0576, "step": 12800 }, { "grad_norm": 0.372266948223114, "learning_rate": 9.672160168451667e-05, "loss": 0.0504, "step": 12810 }, { "grad_norm": 0.2872869670391083, "learning_rate": 9.671423682755549e-05, "loss": 0.0541, "step": 12820 }, { "grad_norm": 0.6240275502204895, "learning_rate": 9.670686398840888e-05, "loss": 0.0542, "step": 12830 }, { "grad_norm": 0.4901260733604431, "learning_rate": 9.669948316833664e-05, "loss": 0.048, "step": 12840 }, { "grad_norm": 0.33060526847839355, "learning_rate": 9.669209436859997e-05, "loss": 0.0396, "step": 12850 }, { "grad_norm": 0.3811427652835846, "learning_rate": 9.66846975904614e-05, "loss": 0.0511, "step": 12860 }, { "grad_norm": 0.4402015209197998, "learning_rate": 9.667729283518483e-05, "loss": 0.0457, "step": 12870 }, { "grad_norm": 0.3768898844718933, "learning_rate": 9.666988010403557e-05, "loss": 0.0436, "step": 12880 }, { "grad_norm": 0.293148010969162, "learning_rate": 9.66624593982802e-05, "loss": 0.0452, "step": 12890 }, { "grad_norm": 0.34407973289489746, "learning_rate": 9.665503071918675e-05, "loss": 0.056, "step": 12900 }, { "grad_norm": 0.2253444492816925, "learning_rate": 9.664759406802456e-05, "loss": 0.0426, "step": 12910 }, { "grad_norm": 0.34263432025909424, "learning_rate": 9.664014944606437e-05, "loss": 0.0353, "step": 12920 }, { "grad_norm": 0.30286335945129395, "learning_rate": 9.663269685457822e-05, "loss": 0.0431, "step": 12930 }, { "grad_norm": 0.42194223403930664, "learning_rate": 9.662523629483962e-05, "loss": 0.0452, "step": 12940 }, { "grad_norm": 0.3398895561695099, "learning_rate": 9.661776776812333e-05, "loss": 0.0375, "step": 12950 }, { "grad_norm": 0.44105958938598633, "learning_rate": 9.661029127570553e-05, "loss": 0.0405, "step": 12960 }, { "grad_norm": 0.3475692868232727, "learning_rate": 9.660280681886373e-05, "loss": 0.0429, "step": 12970 }, { "grad_norm": 0.3701843321323395, "learning_rate": 9.659531439887685e-05, "loss": 0.0374, "step": 12980 }, { "grad_norm": 0.3129587173461914, "learning_rate": 9.658781401702511e-05, "loss": 0.0403, "step": 12990 }, { "grad_norm": 0.38952013850212097, "learning_rate": 9.658030567459015e-05, "loss": 0.0401, "step": 13000 }, { "grad_norm": 0.3364516794681549, "learning_rate": 9.65727893728549e-05, "loss": 0.0618, "step": 13010 }, { "grad_norm": 0.5110674500465393, "learning_rate": 9.656526511310375e-05, "loss": 0.06, "step": 13020 }, { "grad_norm": 0.47779256105422974, "learning_rate": 9.655773289662233e-05, "loss": 0.0594, "step": 13030 }, { "grad_norm": 0.4008791148662567, "learning_rate": 9.655019272469772e-05, "loss": 0.0431, "step": 13040 }, { "grad_norm": 0.33628782629966736, "learning_rate": 9.654264459861832e-05, "loss": 0.0441, "step": 13050 }, { "grad_norm": 0.3100014925003052, "learning_rate": 9.653508851967391e-05, "loss": 0.0435, "step": 13060 }, { "grad_norm": 0.40606147050857544, "learning_rate": 9.65275244891556e-05, "loss": 0.0623, "step": 13070 }, { "grad_norm": 0.38136130571365356, "learning_rate": 9.651995250835591e-05, "loss": 0.0469, "step": 13080 }, { "grad_norm": 0.47566840052604675, "learning_rate": 9.651237257856862e-05, "loss": 0.05, "step": 13090 }, { "grad_norm": 0.3884623646736145, "learning_rate": 9.6504784701089e-05, "loss": 0.0454, "step": 13100 }, { "grad_norm": 0.4564158618450165, "learning_rate": 9.649718887721357e-05, "loss": 0.0382, "step": 13110 }, { "grad_norm": 0.4526863992214203, "learning_rate": 9.648958510824028e-05, "loss": 0.0446, "step": 13120 }, { "grad_norm": 0.45972174406051636, "learning_rate": 9.648197339546837e-05, "loss": 0.0557, "step": 13130 }, { "grad_norm": 0.4193723797798157, "learning_rate": 9.647435374019851e-05, "loss": 0.0567, "step": 13140 }, { "grad_norm": 0.4816511869430542, "learning_rate": 9.646672614373266e-05, "loss": 0.0532, "step": 13150 }, { "grad_norm": 0.4629300534725189, "learning_rate": 9.645909060737418e-05, "loss": 0.0384, "step": 13160 }, { "grad_norm": 0.3234824240207672, "learning_rate": 9.645144713242778e-05, "loss": 0.0466, "step": 13170 }, { "grad_norm": 0.43019166588783264, "learning_rate": 9.64437957201995e-05, "loss": 0.049, "step": 13180 }, { "grad_norm": 0.31219378113746643, "learning_rate": 9.643613637199678e-05, "loss": 0.0372, "step": 13190 }, { "grad_norm": 0.43445923924446106, "learning_rate": 9.642846908912839e-05, "loss": 0.0552, "step": 13200 }, { "grad_norm": 0.3308875262737274, "learning_rate": 9.642079387290444e-05, "loss": 0.034, "step": 13210 }, { "grad_norm": 0.3181224465370178, "learning_rate": 9.641311072463644e-05, "loss": 0.0381, "step": 13220 }, { "grad_norm": 0.49642375111579895, "learning_rate": 9.640541964563722e-05, "loss": 0.0484, "step": 13230 }, { "grad_norm": 0.5091738104820251, "learning_rate": 9.639772063722096e-05, "loss": 0.0593, "step": 13240 }, { "grad_norm": 0.3903161287307739, "learning_rate": 9.639001370070324e-05, "loss": 0.0468, "step": 13250 }, { "grad_norm": 0.4463886320590973, "learning_rate": 9.638229883740095e-05, "loss": 0.049, "step": 13260 }, { "grad_norm": 0.3589116334915161, "learning_rate": 9.637457604863233e-05, "loss": 0.0446, "step": 13270 }, { "grad_norm": 0.4503541886806488, "learning_rate": 9.636684533571703e-05, "loss": 0.0454, "step": 13280 }, { "grad_norm": 0.4225026071071625, "learning_rate": 9.635910669997599e-05, "loss": 0.0438, "step": 13290 }, { "grad_norm": 0.3122190833091736, "learning_rate": 9.635136014273154e-05, "loss": 0.0361, "step": 13300 }, { "grad_norm": 0.3683110177516937, "learning_rate": 9.634360566530735e-05, "loss": 0.0534, "step": 13310 }, { "grad_norm": 0.47143280506134033, "learning_rate": 9.633584326902845e-05, "loss": 0.0553, "step": 13320 }, { "grad_norm": 0.3980634808540344, "learning_rate": 9.632807295522124e-05, "loss": 0.0487, "step": 13330 }, { "grad_norm": 0.33348631858825684, "learning_rate": 9.632029472521342e-05, "loss": 0.0521, "step": 13340 }, { "grad_norm": 0.3734755516052246, "learning_rate": 9.631250858033409e-05, "loss": 0.0444, "step": 13350 }, { "grad_norm": 0.2740147113800049, "learning_rate": 9.630471452191371e-05, "loss": 0.0511, "step": 13360 }, { "grad_norm": 0.2815921902656555, "learning_rate": 9.629691255128405e-05, "loss": 0.0551, "step": 13370 }, { "grad_norm": 0.42637714743614197, "learning_rate": 9.628910266977825e-05, "loss": 0.045, "step": 13380 }, { "grad_norm": 0.4093478322029114, "learning_rate": 9.628128487873083e-05, "loss": 0.0526, "step": 13390 }, { "grad_norm": 0.36738529801368713, "learning_rate": 9.627345917947761e-05, "loss": 0.0554, "step": 13400 }, { "grad_norm": 0.45049595832824707, "learning_rate": 9.626562557335579e-05, "loss": 0.0521, "step": 13410 }, { "grad_norm": 0.459561288356781, "learning_rate": 9.625778406170393e-05, "loss": 0.0564, "step": 13420 }, { "grad_norm": 0.3616820275783539, "learning_rate": 9.624993464586193e-05, "loss": 0.0499, "step": 13430 }, { "grad_norm": 0.36687171459198, "learning_rate": 9.624207732717105e-05, "loss": 0.045, "step": 13440 }, { "grad_norm": 0.5235371589660645, "learning_rate": 9.623421210697386e-05, "loss": 0.0612, "step": 13450 }, { "grad_norm": 0.30301645398139954, "learning_rate": 9.622633898661434e-05, "loss": 0.0548, "step": 13460 }, { "grad_norm": 0.41591015458106995, "learning_rate": 9.621845796743778e-05, "loss": 0.061, "step": 13470 }, { "grad_norm": 0.37900954484939575, "learning_rate": 9.621056905079082e-05, "loss": 0.057, "step": 13480 }, { "grad_norm": 0.39173245429992676, "learning_rate": 9.620267223802149e-05, "loss": 0.0512, "step": 13490 }, { "grad_norm": 0.41298210620880127, "learning_rate": 9.619476753047911e-05, "loss": 0.0374, "step": 13500 }, { "grad_norm": 0.4371635317802429, "learning_rate": 9.618685492951438e-05, "loss": 0.0567, "step": 13510 }, { "grad_norm": 0.40607497096061707, "learning_rate": 9.617893443647938e-05, "loss": 0.0603, "step": 13520 }, { "grad_norm": 0.3869188725948334, "learning_rate": 9.617100605272746e-05, "loss": 0.0608, "step": 13530 }, { "grad_norm": 0.3430202603340149, "learning_rate": 9.616306977961338e-05, "loss": 0.0527, "step": 13540 }, { "grad_norm": 0.3108659088611603, "learning_rate": 9.615512561849326e-05, "loss": 0.0448, "step": 13550 }, { "grad_norm": 0.4292880594730377, "learning_rate": 9.61471735707245e-05, "loss": 0.0534, "step": 13560 }, { "grad_norm": 0.3148081302642822, "learning_rate": 9.613921363766592e-05, "loss": 0.0409, "step": 13570 }, { "grad_norm": 0.36326122283935547, "learning_rate": 9.613124582067763e-05, "loss": 0.0442, "step": 13580 }, { "grad_norm": 0.49709346890449524, "learning_rate": 9.612327012112112e-05, "loss": 0.0603, "step": 13590 }, { "grad_norm": 0.5343999862670898, "learning_rate": 9.611528654035921e-05, "loss": 0.0737, "step": 13600 }, { "grad_norm": 0.40890759229660034, "learning_rate": 9.610729507975611e-05, "loss": 0.0406, "step": 13610 }, { "grad_norm": 0.3695039451122284, "learning_rate": 9.609929574067731e-05, "loss": 0.0596, "step": 13620 }, { "grad_norm": 0.511406421661377, "learning_rate": 9.609128852448967e-05, "loss": 0.0482, "step": 13630 }, { "grad_norm": 0.45735883712768555, "learning_rate": 9.608327343256143e-05, "loss": 0.0474, "step": 13640 }, { "grad_norm": 0.3893585801124573, "learning_rate": 9.607525046626216e-05, "loss": 0.0532, "step": 13650 }, { "grad_norm": 0.4918266832828522, "learning_rate": 9.606721962696272e-05, "loss": 0.0547, "step": 13660 }, { "grad_norm": 0.3666170537471771, "learning_rate": 9.60591809160354e-05, "loss": 0.0465, "step": 13670 }, { "grad_norm": 0.32548969984054565, "learning_rate": 9.605113433485378e-05, "loss": 0.0514, "step": 13680 }, { "grad_norm": 0.49712643027305603, "learning_rate": 9.604307988479279e-05, "loss": 0.0463, "step": 13690 }, { "grad_norm": 0.3305007815361023, "learning_rate": 9.603501756722876e-05, "loss": 0.0466, "step": 13700 }, { "grad_norm": 0.3720790445804596, "learning_rate": 9.602694738353927e-05, "loss": 0.0421, "step": 13710 }, { "grad_norm": 0.2740316390991211, "learning_rate": 9.601886933510331e-05, "loss": 0.0379, "step": 13720 }, { "grad_norm": 0.41780659556388855, "learning_rate": 9.60107834233012e-05, "loss": 0.0412, "step": 13730 }, { "grad_norm": 0.37330833077430725, "learning_rate": 9.60026896495146e-05, "loss": 0.0439, "step": 13740 }, { "grad_norm": 0.30696073174476624, "learning_rate": 9.599458801512652e-05, "loss": 0.0453, "step": 13750 }, { "grad_norm": 0.4028429090976715, "learning_rate": 9.598647852152129e-05, "loss": 0.0679, "step": 13760 }, { "grad_norm": 0.28315484523773193, "learning_rate": 9.597836117008462e-05, "loss": 0.0423, "step": 13770 }, { "grad_norm": 0.34154751896858215, "learning_rate": 9.597023596220356e-05, "loss": 0.0432, "step": 13780 }, { "grad_norm": 0.5338660478591919, "learning_rate": 9.596210289926643e-05, "loss": 0.0479, "step": 13790 }, { "grad_norm": 0.45938217639923096, "learning_rate": 9.5953961982663e-05, "loss": 0.0475, "step": 13800 }, { "grad_norm": 0.46190470457077026, "learning_rate": 9.594581321378431e-05, "loss": 0.0527, "step": 13810 }, { "grad_norm": 0.39621594548225403, "learning_rate": 9.593765659402276e-05, "loss": 0.0508, "step": 13820 }, { "grad_norm": 0.39769282937049866, "learning_rate": 9.59294921247721e-05, "loss": 0.053, "step": 13830 }, { "grad_norm": 0.44154906272888184, "learning_rate": 9.59213198074274e-05, "loss": 0.0535, "step": 13840 }, { "grad_norm": 0.5081636309623718, "learning_rate": 9.59131396433851e-05, "loss": 0.0438, "step": 13850 }, { "grad_norm": 0.3886975944042206, "learning_rate": 9.590495163404297e-05, "loss": 0.0527, "step": 13860 }, { "grad_norm": 0.3507029414176941, "learning_rate": 9.589675578080009e-05, "loss": 0.0576, "step": 13870 }, { "grad_norm": 0.43333932757377625, "learning_rate": 9.588855208505694e-05, "loss": 0.0527, "step": 13880 }, { "grad_norm": 0.4674290120601654, "learning_rate": 9.588034054821529e-05, "loss": 0.0691, "step": 13890 }, { "grad_norm": 0.5111890435218811, "learning_rate": 9.587212117167826e-05, "loss": 0.0445, "step": 13900 }, { "grad_norm": 0.5105722546577454, "learning_rate": 9.586389395685033e-05, "loss": 0.0538, "step": 13910 }, { "grad_norm": 0.5364643335342407, "learning_rate": 9.585565890513733e-05, "loss": 0.0598, "step": 13920 }, { "grad_norm": 0.4849603772163391, "learning_rate": 9.584741601794636e-05, "loss": 0.0575, "step": 13930 }, { "grad_norm": 0.4065185487270355, "learning_rate": 9.58391652966859e-05, "loss": 0.0478, "step": 13940 }, { "grad_norm": 0.5483849048614502, "learning_rate": 9.583090674276583e-05, "loss": 0.0487, "step": 13950 }, { "grad_norm": 0.39791372418403625, "learning_rate": 9.582264035759726e-05, "loss": 0.0456, "step": 13960 }, { "grad_norm": 0.34475821256637573, "learning_rate": 9.58143661425927e-05, "loss": 0.0657, "step": 13970 }, { "grad_norm": 0.3978153467178345, "learning_rate": 9.580608409916601e-05, "loss": 0.0562, "step": 13980 }, { "grad_norm": 0.3753971755504608, "learning_rate": 9.579779422873233e-05, "loss": 0.0534, "step": 13990 }, { "grad_norm": 0.526714563369751, "learning_rate": 9.578949653270819e-05, "loss": 0.0547, "step": 14000 }, { "grad_norm": 0.4785560369491577, "learning_rate": 9.578119101251144e-05, "loss": 0.0486, "step": 14010 }, { "grad_norm": 0.37948745489120483, "learning_rate": 9.577287766956127e-05, "loss": 0.0601, "step": 14020 }, { "grad_norm": 0.3600105941295624, "learning_rate": 9.57645565052782e-05, "loss": 0.0532, "step": 14030 }, { "grad_norm": 0.38606205582618713, "learning_rate": 9.575622752108407e-05, "loss": 0.0407, "step": 14040 }, { "grad_norm": 0.35388118028640747, "learning_rate": 9.57478907184021e-05, "loss": 0.0398, "step": 14050 }, { "grad_norm": 0.3298298120498657, "learning_rate": 9.573954609865681e-05, "loss": 0.0486, "step": 14060 }, { "grad_norm": 0.44140544533729553, "learning_rate": 9.573119366327408e-05, "loss": 0.0546, "step": 14070 }, { "grad_norm": 0.35241997241973877, "learning_rate": 9.57228334136811e-05, "loss": 0.0449, "step": 14080 }, { "grad_norm": 0.3202846944332123, "learning_rate": 9.571446535130641e-05, "loss": 0.0463, "step": 14090 }, { "grad_norm": 0.31786319613456726, "learning_rate": 9.570608947757988e-05, "loss": 0.0533, "step": 14100 }, { "grad_norm": 0.3894791603088379, "learning_rate": 9.569770579393274e-05, "loss": 0.0485, "step": 14110 }, { "grad_norm": 0.37317124009132385, "learning_rate": 9.56893143017975e-05, "loss": 0.0415, "step": 14120 }, { "grad_norm": 0.4249415397644043, "learning_rate": 9.568091500260806e-05, "loss": 0.053, "step": 14130 }, { "grad_norm": 0.426482230424881, "learning_rate": 9.567250789779961e-05, "loss": 0.0476, "step": 14140 }, { "grad_norm": 0.3244151473045349, "learning_rate": 9.566409298880872e-05, "loss": 0.0375, "step": 14150 }, { "grad_norm": 0.5097789764404297, "learning_rate": 9.565567027707326e-05, "loss": 0.0659, "step": 14160 }, { "grad_norm": 0.6135331988334656, "learning_rate": 9.56472397640324e-05, "loss": 0.0647, "step": 14170 }, { "grad_norm": 0.4514925181865692, "learning_rate": 9.563880145112675e-05, "loss": 0.0579, "step": 14180 }, { "grad_norm": 0.4943186938762665, "learning_rate": 9.563035533979814e-05, "loss": 0.0582, "step": 14190 }, { "grad_norm": 0.4713784158229828, "learning_rate": 9.562190143148981e-05, "loss": 0.051, "step": 14200 }, { "grad_norm": 0.3299650251865387, "learning_rate": 9.561343972764627e-05, "loss": 0.0485, "step": 14210 }, { "grad_norm": 0.29559558629989624, "learning_rate": 9.560497022971343e-05, "loss": 0.0592, "step": 14220 }, { "grad_norm": 0.41032055020332336, "learning_rate": 9.559649293913847e-05, "loss": 0.0454, "step": 14230 }, { "grad_norm": 0.5264262557029724, "learning_rate": 9.558800785736993e-05, "loss": 0.0514, "step": 14240 }, { "grad_norm": 0.4462316334247589, "learning_rate": 9.557951498585767e-05, "loss": 0.0569, "step": 14250 }, { "grad_norm": 0.3138783276081085, "learning_rate": 9.557101432605293e-05, "loss": 0.0467, "step": 14260 }, { "grad_norm": 0.42064785957336426, "learning_rate": 9.556250587940818e-05, "loss": 0.0604, "step": 14270 }, { "grad_norm": 0.42240259051322937, "learning_rate": 9.555398964737734e-05, "loss": 0.0467, "step": 14280 }, { "grad_norm": 0.40919724106788635, "learning_rate": 9.554546563141555e-05, "loss": 0.0432, "step": 14290 }, { "grad_norm": 0.4402596056461334, "learning_rate": 9.553693383297937e-05, "loss": 0.0525, "step": 14300 }, { "grad_norm": 0.41357889771461487, "learning_rate": 9.552839425352663e-05, "loss": 0.0463, "step": 14310 }, { "grad_norm": 0.35777145624160767, "learning_rate": 9.551984689451652e-05, "loss": 0.0582, "step": 14320 }, { "grad_norm": 0.3175971508026123, "learning_rate": 9.551129175740953e-05, "loss": 0.0418, "step": 14330 }, { "grad_norm": 0.32146891951560974, "learning_rate": 9.550272884366754e-05, "loss": 0.0454, "step": 14340 }, { "grad_norm": 0.4966822862625122, "learning_rate": 9.549415815475369e-05, "loss": 0.0483, "step": 14350 }, { "grad_norm": 0.42361143231391907, "learning_rate": 9.548557969213247e-05, "loss": 0.0417, "step": 14360 }, { "grad_norm": 0.3845313787460327, "learning_rate": 9.547699345726972e-05, "loss": 0.0471, "step": 14370 }, { "grad_norm": 0.3232003152370453, "learning_rate": 9.546839945163257e-05, "loss": 0.0449, "step": 14380 }, { "grad_norm": 0.3553611636161804, "learning_rate": 9.545979767668953e-05, "loss": 0.0601, "step": 14390 }, { "grad_norm": 0.33711138367652893, "learning_rate": 9.54511881339104e-05, "loss": 0.052, "step": 14400 }, { "grad_norm": 0.3581922948360443, "learning_rate": 9.54425708247663e-05, "loss": 0.0429, "step": 14410 }, { "grad_norm": 0.3096526563167572, "learning_rate": 9.543394575072972e-05, "loss": 0.0426, "step": 14420 }, { "grad_norm": 0.37740129232406616, "learning_rate": 9.542531291327441e-05, "loss": 0.0422, "step": 14430 }, { "grad_norm": 0.3951590657234192, "learning_rate": 9.541667231387552e-05, "loss": 0.0411, "step": 14440 }, { "grad_norm": 0.4091281592845917, "learning_rate": 9.540802395400949e-05, "loss": 0.049, "step": 14450 }, { "grad_norm": 0.35154202580451965, "learning_rate": 9.539936783515406e-05, "loss": 0.0432, "step": 14460 }, { "grad_norm": 0.3111257255077362, "learning_rate": 9.539070395878835e-05, "loss": 0.0535, "step": 14470 }, { "grad_norm": 0.38898423314094543, "learning_rate": 9.538203232639277e-05, "loss": 0.0398, "step": 14480 }, { "grad_norm": 0.4382215440273285, "learning_rate": 9.537335293944907e-05, "loss": 0.0468, "step": 14490 }, { "grad_norm": 0.3081832528114319, "learning_rate": 9.536466579944032e-05, "loss": 0.0601, "step": 14500 }, { "grad_norm": 0.34070146083831787, "learning_rate": 9.535597090785091e-05, "loss": 0.0472, "step": 14510 }, { "grad_norm": 0.3206569254398346, "learning_rate": 9.534726826616656e-05, "loss": 0.0395, "step": 14520 }, { "grad_norm": 0.3547995388507843, "learning_rate": 9.53385578758743e-05, "loss": 0.0683, "step": 14530 }, { "grad_norm": 0.32659992575645447, "learning_rate": 9.532983973846252e-05, "loss": 0.0374, "step": 14540 }, { "grad_norm": 0.36209720373153687, "learning_rate": 9.53211138554209e-05, "loss": 0.0398, "step": 14550 }, { "grad_norm": 0.5601968169212341, "learning_rate": 9.531238022824047e-05, "loss": 0.0462, "step": 14560 }, { "grad_norm": 0.36380434036254883, "learning_rate": 9.530363885841355e-05, "loss": 0.0366, "step": 14570 }, { "grad_norm": 0.4064083397388458, "learning_rate": 9.52948897474338e-05, "loss": 0.0491, "step": 14580 }, { "grad_norm": 0.3481707274913788, "learning_rate": 9.528613289679622e-05, "loss": 0.0524, "step": 14590 }, { "grad_norm": 0.40738779306411743, "learning_rate": 9.52773683079971e-05, "loss": 0.0458, "step": 14600 }, { "grad_norm": 0.3623390793800354, "learning_rate": 9.526859598253407e-05, "loss": 0.0351, "step": 14610 }, { "grad_norm": 0.42082637548446655, "learning_rate": 9.525981592190609e-05, "loss": 0.0427, "step": 14620 }, { "grad_norm": 0.47301143407821655, "learning_rate": 9.525102812761342e-05, "loss": 0.0427, "step": 14630 }, { "grad_norm": 0.4720071256160736, "learning_rate": 9.524223260115768e-05, "loss": 0.042, "step": 14640 }, { "grad_norm": 0.3316323459148407, "learning_rate": 9.523342934404175e-05, "loss": 0.042, "step": 14650 }, { "grad_norm": 0.372660756111145, "learning_rate": 9.522461835776989e-05, "loss": 0.049, "step": 14660 }, { "grad_norm": 0.3054749071598053, "learning_rate": 9.521579964384764e-05, "loss": 0.0512, "step": 14670 }, { "grad_norm": 0.4081129729747772, "learning_rate": 9.52069732037819e-05, "loss": 0.0449, "step": 14680 }, { "grad_norm": 0.31546229124069214, "learning_rate": 9.519813903908083e-05, "loss": 0.0455, "step": 14690 }, { "grad_norm": 0.46001002192497253, "learning_rate": 9.5189297151254e-05, "loss": 0.0477, "step": 14700 }, { "grad_norm": 0.49588510394096375, "learning_rate": 9.518044754181218e-05, "loss": 0.0503, "step": 14710 }, { "grad_norm": 0.28023600578308105, "learning_rate": 9.51715902122676e-05, "loss": 0.0557, "step": 14720 }, { "grad_norm": 0.25385287404060364, "learning_rate": 9.516272516413368e-05, "loss": 0.0513, "step": 14730 }, { "grad_norm": 0.3247681260108948, "learning_rate": 9.515385239892525e-05, "loss": 0.0522, "step": 14740 }, { "grad_norm": 0.44828248023986816, "learning_rate": 9.514497191815839e-05, "loss": 0.0549, "step": 14750 }, { "grad_norm": 0.41961368918418884, "learning_rate": 9.513608372335055e-05, "loss": 0.0548, "step": 14760 }, { "grad_norm": 0.38487568497657776, "learning_rate": 9.512718781602045e-05, "loss": 0.048, "step": 14770 }, { "grad_norm": 0.39720237255096436, "learning_rate": 9.511828419768823e-05, "loss": 0.0437, "step": 14780 }, { "grad_norm": 0.3326103389263153, "learning_rate": 9.510937286987521e-05, "loss": 0.0395, "step": 14790 }, { "grad_norm": 0.4757642149925232, "learning_rate": 9.510045383410408e-05, "loss": 0.0453, "step": 14800 }, { "grad_norm": 0.3885006904602051, "learning_rate": 9.509152709189892e-05, "loss": 0.0393, "step": 14810 }, { "grad_norm": 0.3007967174053192, "learning_rate": 9.508259264478504e-05, "loss": 0.0455, "step": 14820 }, { "grad_norm": 0.4486839175224304, "learning_rate": 9.507365049428909e-05, "loss": 0.0532, "step": 14830 }, { "grad_norm": 0.2812560796737671, "learning_rate": 9.506470064193902e-05, "loss": 0.058, "step": 14840 }, { "grad_norm": 0.3655684292316437, "learning_rate": 9.505574308926414e-05, "loss": 0.0559, "step": 14850 }, { "grad_norm": 0.4279797077178955, "learning_rate": 9.504677783779505e-05, "loss": 0.0371, "step": 14860 }, { "grad_norm": 0.41810160875320435, "learning_rate": 9.503780488906365e-05, "loss": 0.0353, "step": 14870 }, { "grad_norm": 0.32822996377944946, "learning_rate": 9.502882424460319e-05, "loss": 0.0417, "step": 14880 }, { "grad_norm": 0.4689843952655792, "learning_rate": 9.501983590594821e-05, "loss": 0.0425, "step": 14890 }, { "grad_norm": 0.40643778443336487, "learning_rate": 9.501083987463455e-05, "loss": 0.0395, "step": 14900 }, { "grad_norm": 0.36024340987205505, "learning_rate": 9.500183615219942e-05, "loss": 0.0416, "step": 14910 }, { "grad_norm": 0.4835032820701599, "learning_rate": 9.49928247401813e-05, "loss": 0.0422, "step": 14920 }, { "grad_norm": 0.41923588514328003, "learning_rate": 9.498380564011997e-05, "loss": 0.0449, "step": 14930 }, { "grad_norm": 0.39967629313468933, "learning_rate": 9.497477885355656e-05, "loss": 0.0572, "step": 14940 }, { "grad_norm": 0.4112614691257477, "learning_rate": 9.496574438203353e-05, "loss": 0.0548, "step": 14950 }, { "grad_norm": 0.29770898818969727, "learning_rate": 9.495670222709459e-05, "loss": 0.0422, "step": 14960 }, { "grad_norm": 0.2787676155567169, "learning_rate": 9.494765239028483e-05, "loss": 0.0413, "step": 14970 }, { "grad_norm": 0.33386144042015076, "learning_rate": 9.493859487315057e-05, "loss": 0.0421, "step": 14980 }, { "grad_norm": 0.417217493057251, "learning_rate": 9.492952967723953e-05, "loss": 0.047, "step": 14990 }, { "grad_norm": 0.4077162444591522, "learning_rate": 9.492045680410068e-05, "loss": 0.0534, "step": 15000 }, { "grad_norm": 0.38041427731513977, "learning_rate": 9.491137625528436e-05, "loss": 0.0393, "step": 15010 }, { "grad_norm": 0.43279650807380676, "learning_rate": 9.490228803234215e-05, "loss": 0.043, "step": 15020 }, { "grad_norm": 0.41316941380500793, "learning_rate": 9.489319213682701e-05, "loss": 0.0409, "step": 15030 }, { "grad_norm": 0.28487005829811096, "learning_rate": 9.488408857029316e-05, "loss": 0.0532, "step": 15040 }, { "grad_norm": 0.35676881670951843, "learning_rate": 9.487497733429616e-05, "loss": 0.0594, "step": 15050 }, { "grad_norm": 0.45697811245918274, "learning_rate": 9.486585843039286e-05, "loss": 0.0616, "step": 15060 }, { "grad_norm": 0.3099939227104187, "learning_rate": 9.485673186014143e-05, "loss": 0.0474, "step": 15070 }, { "grad_norm": 0.35143324732780457, "learning_rate": 9.484759762510137e-05, "loss": 0.0489, "step": 15080 }, { "grad_norm": 0.4192099869251251, "learning_rate": 9.483845572683346e-05, "loss": 0.0494, "step": 15090 }, { "grad_norm": 0.38678884506225586, "learning_rate": 9.48293061668998e-05, "loss": 0.0443, "step": 15100 }, { "grad_norm": 0.3525322377681732, "learning_rate": 9.48201489468638e-05, "loss": 0.0406, "step": 15110 }, { "grad_norm": 0.2910543382167816, "learning_rate": 9.481098406829016e-05, "loss": 0.0347, "step": 15120 }, { "grad_norm": 0.32279863953590393, "learning_rate": 9.480181153274495e-05, "loss": 0.0393, "step": 15130 }, { "grad_norm": 0.33839061856269836, "learning_rate": 9.479263134179548e-05, "loss": 0.0311, "step": 15140 }, { "grad_norm": 0.4427463114261627, "learning_rate": 9.478344349701039e-05, "loss": 0.0444, "step": 15150 }, { "grad_norm": 0.306815505027771, "learning_rate": 9.477424799995964e-05, "loss": 0.0424, "step": 15160 }, { "grad_norm": 0.4292967915534973, "learning_rate": 9.476504485221448e-05, "loss": 0.0491, "step": 15170 }, { "grad_norm": 0.32616937160491943, "learning_rate": 9.475583405534748e-05, "loss": 0.0443, "step": 15180 }, { "grad_norm": 0.339767187833786, "learning_rate": 9.474661561093251e-05, "loss": 0.0371, "step": 15190 }, { "grad_norm": 0.4127395451068878, "learning_rate": 9.473738952054478e-05, "loss": 0.0522, "step": 15200 }, { "grad_norm": 0.3489748239517212, "learning_rate": 9.472815578576073e-05, "loss": 0.043, "step": 15210 }, { "grad_norm": 0.5345419049263, "learning_rate": 9.471891440815817e-05, "loss": 0.0407, "step": 15220 }, { "grad_norm": 0.2579619586467743, "learning_rate": 9.470966538931621e-05, "loss": 0.0351, "step": 15230 }, { "grad_norm": 0.38862600922584534, "learning_rate": 9.470040873081525e-05, "loss": 0.0361, "step": 15240 }, { "grad_norm": 0.45224887132644653, "learning_rate": 9.469114443423698e-05, "loss": 0.0583, "step": 15250 }, { "grad_norm": 0.3895515203475952, "learning_rate": 9.468187250116445e-05, "loss": 0.05, "step": 15260 }, { "grad_norm": 0.2755981683731079, "learning_rate": 9.467259293318197e-05, "loss": 0.0382, "step": 15270 }, { "grad_norm": 0.2231302261352539, "learning_rate": 9.466330573187514e-05, "loss": 0.0385, "step": 15280 }, { "grad_norm": 0.42665883898735046, "learning_rate": 9.46540108988309e-05, "loss": 0.0555, "step": 15290 }, { "grad_norm": 0.36704763770103455, "learning_rate": 9.46447084356375e-05, "loss": 0.0488, "step": 15300 }, { "grad_norm": 0.5911537408828735, "learning_rate": 9.463539834388447e-05, "loss": 0.0543, "step": 15310 }, { "grad_norm": 0.3706696331501007, "learning_rate": 9.462608062516263e-05, "loss": 0.0443, "step": 15320 }, { "grad_norm": 0.40308067202568054, "learning_rate": 9.461675528106413e-05, "loss": 0.0407, "step": 15330 }, { "grad_norm": 0.2977010905742645, "learning_rate": 9.460742231318244e-05, "loss": 0.0506, "step": 15340 }, { "grad_norm": 0.3920888304710388, "learning_rate": 9.459808172311229e-05, "loss": 0.0511, "step": 15350 }, { "grad_norm": 0.28932642936706543, "learning_rate": 9.458873351244972e-05, "loss": 0.0379, "step": 15360 }, { "grad_norm": 0.34041833877563477, "learning_rate": 9.457937768279211e-05, "loss": 0.0481, "step": 15370 }, { "grad_norm": 0.28668585419654846, "learning_rate": 9.45700142357381e-05, "loss": 0.0427, "step": 15380 }, { "grad_norm": 0.41653749346733093, "learning_rate": 9.456064317288765e-05, "loss": 0.0407, "step": 15390 }, { "grad_norm": 0.5015144944190979, "learning_rate": 9.455126449584201e-05, "loss": 0.0458, "step": 15400 }, { "grad_norm": 0.29849374294281006, "learning_rate": 9.454187820620375e-05, "loss": 0.0367, "step": 15410 }, { "grad_norm": 0.27823296189308167, "learning_rate": 9.453248430557673e-05, "loss": 0.0467, "step": 15420 }, { "grad_norm": 0.35529473423957825, "learning_rate": 9.452308279556611e-05, "loss": 0.0643, "step": 15430 }, { "grad_norm": 0.3208843469619751, "learning_rate": 9.451367367777835e-05, "loss": 0.0443, "step": 15440 }, { "grad_norm": 0.3375883996486664, "learning_rate": 9.450425695382122e-05, "loss": 0.0533, "step": 15450 }, { "grad_norm": 0.3992135226726532, "learning_rate": 9.449483262530375e-05, "loss": 0.0462, "step": 15460 }, { "grad_norm": 0.30047765374183655, "learning_rate": 9.448540069383633e-05, "loss": 0.0349, "step": 15470 }, { "grad_norm": 0.3909514546394348, "learning_rate": 9.447596116103061e-05, "loss": 0.0433, "step": 15480 }, { "grad_norm": 0.40634801983833313, "learning_rate": 9.446651402849955e-05, "loss": 0.0472, "step": 15490 }, { "grad_norm": 0.3540271818637848, "learning_rate": 9.44570592978574e-05, "loss": 0.041, "step": 15500 }, { "grad_norm": 0.444279283285141, "learning_rate": 9.444759697071972e-05, "loss": 0.0458, "step": 15510 }, { "grad_norm": 0.4082109332084656, "learning_rate": 9.443812704870336e-05, "loss": 0.0373, "step": 15520 }, { "grad_norm": 0.37326276302337646, "learning_rate": 9.442864953342649e-05, "loss": 0.04, "step": 15530 }, { "grad_norm": 0.39200884103775024, "learning_rate": 9.441916442650852e-05, "loss": 0.042, "step": 15540 }, { "grad_norm": 0.3505919575691223, "learning_rate": 9.440967172957023e-05, "loss": 0.0411, "step": 15550 }, { "grad_norm": 0.3573952615261078, "learning_rate": 9.440017144423364e-05, "loss": 0.0442, "step": 15560 }, { "grad_norm": 0.6191890239715576, "learning_rate": 9.439066357212209e-05, "loss": 0.0599, "step": 15570 }, { "grad_norm": 0.4265294671058655, "learning_rate": 9.438114811486022e-05, "loss": 0.0578, "step": 15580 }, { "grad_norm": 0.48367682099342346, "learning_rate": 9.4371625074074e-05, "loss": 0.0463, "step": 15590 }, { "grad_norm": 0.3436623215675354, "learning_rate": 9.436209445139059e-05, "loss": 0.0476, "step": 15600 }, { "grad_norm": 0.39914363622665405, "learning_rate": 9.435255624843855e-05, "loss": 0.0544, "step": 15610 }, { "grad_norm": 0.35047611594200134, "learning_rate": 9.43430104668477e-05, "loss": 0.0436, "step": 15620 }, { "grad_norm": 0.39106008410453796, "learning_rate": 9.433345710824914e-05, "loss": 0.0459, "step": 15630 }, { "grad_norm": 0.4034060835838318, "learning_rate": 9.432389617427529e-05, "loss": 0.0489, "step": 15640 }, { "grad_norm": 0.29975008964538574, "learning_rate": 9.431432766655984e-05, "loss": 0.0413, "step": 15650 }, { "grad_norm": 0.4035623371601105, "learning_rate": 9.430475158673778e-05, "loss": 0.0492, "step": 15660 }, { "grad_norm": 0.3777451813220978, "learning_rate": 9.429516793644542e-05, "loss": 0.0405, "step": 15670 }, { "grad_norm": 0.3237573504447937, "learning_rate": 9.428557671732034e-05, "loss": 0.0347, "step": 15680 }, { "grad_norm": 0.4788273870944977, "learning_rate": 9.42759779310014e-05, "loss": 0.0444, "step": 15690 }, { "grad_norm": 0.3991643190383911, "learning_rate": 9.426637157912879e-05, "loss": 0.0478, "step": 15700 }, { "grad_norm": 0.4855463206768036, "learning_rate": 9.425675766334397e-05, "loss": 0.0497, "step": 15710 }, { "grad_norm": 0.4395475387573242, "learning_rate": 9.424713618528968e-05, "loss": 0.0387, "step": 15720 }, { "grad_norm": 0.40075305104255676, "learning_rate": 9.423750714661e-05, "loss": 0.0554, "step": 15730 }, { "grad_norm": 0.37284815311431885, "learning_rate": 9.422787054895022e-05, "loss": 0.0403, "step": 15740 }, { "grad_norm": 0.37693119049072266, "learning_rate": 9.4218226393957e-05, "loss": 0.0502, "step": 15750 }, { "grad_norm": 0.38317441940307617, "learning_rate": 9.420857468327828e-05, "loss": 0.0463, "step": 15760 }, { "grad_norm": 0.513068437576294, "learning_rate": 9.419891541856323e-05, "loss": 0.0493, "step": 15770 }, { "grad_norm": 0.3054898977279663, "learning_rate": 9.41892486014624e-05, "loss": 0.0419, "step": 15780 }, { "grad_norm": 0.41827407479286194, "learning_rate": 9.417957423362756e-05, "loss": 0.052, "step": 15790 }, { "grad_norm": 0.31653034687042236, "learning_rate": 9.416989231671178e-05, "loss": 0.0465, "step": 15800 }, { "grad_norm": 0.3277879059314728, "learning_rate": 9.416020285236946e-05, "loss": 0.0431, "step": 15810 }, { "grad_norm": 0.3484608232975006, "learning_rate": 9.415050584225626e-05, "loss": 0.04, "step": 15820 }, { "grad_norm": 0.41131776571273804, "learning_rate": 9.414080128802914e-05, "loss": 0.04, "step": 15830 }, { "grad_norm": 0.38380420207977295, "learning_rate": 9.413108919134632e-05, "loss": 0.043, "step": 15840 }, { "grad_norm": 0.40975847840309143, "learning_rate": 9.412136955386734e-05, "loss": 0.0442, "step": 15850 }, { "grad_norm": 0.30216842889785767, "learning_rate": 9.411164237725303e-05, "loss": 0.0433, "step": 15860 }, { "grad_norm": 0.4440055787563324, "learning_rate": 9.41019076631655e-05, "loss": 0.0516, "step": 15870 }, { "grad_norm": 0.3496961295604706, "learning_rate": 9.409216541326815e-05, "loss": 0.0504, "step": 15880 }, { "grad_norm": 0.3632540702819824, "learning_rate": 9.408241562922564e-05, "loss": 0.0465, "step": 15890 }, { "grad_norm": 0.33620932698249817, "learning_rate": 9.407265831270395e-05, "loss": 0.0351, "step": 15900 }, { "grad_norm": 0.3632429540157318, "learning_rate": 9.406289346537035e-05, "loss": 0.0399, "step": 15910 }, { "grad_norm": 0.38607075810432434, "learning_rate": 9.405312108889339e-05, "loss": 0.0457, "step": 15920 }, { "grad_norm": 0.35351043939590454, "learning_rate": 9.404334118494288e-05, "loss": 0.0428, "step": 15930 }, { "grad_norm": 0.32130247354507446, "learning_rate": 9.403355375518995e-05, "loss": 0.0422, "step": 15940 }, { "grad_norm": 0.46919915080070496, "learning_rate": 9.4023758801307e-05, "loss": 0.0367, "step": 15950 }, { "grad_norm": 0.2862333655357361, "learning_rate": 9.401395632496774e-05, "loss": 0.0386, "step": 15960 }, { "grad_norm": 0.2812814712524414, "learning_rate": 9.400414632784711e-05, "loss": 0.0512, "step": 15970 }, { "grad_norm": 0.4360367953777313, "learning_rate": 9.39943288116214e-05, "loss": 0.0366, "step": 15980 }, { "grad_norm": 0.339820921421051, "learning_rate": 9.398450377796815e-05, "loss": 0.0468, "step": 15990 }, { "grad_norm": 0.3754011392593384, "learning_rate": 9.397467122856616e-05, "loss": 0.0446, "step": 16000 }, { "grad_norm": 0.3425680100917816, "learning_rate": 9.396483116509558e-05, "loss": 0.044, "step": 16010 }, { "grad_norm": 0.41865065693855286, "learning_rate": 9.39549835892378e-05, "loss": 0.0391, "step": 16020 }, { "grad_norm": 0.545530378818512, "learning_rate": 9.39451285026755e-05, "loss": 0.0469, "step": 16030 }, { "grad_norm": 0.40479519963264465, "learning_rate": 9.393526590709262e-05, "loss": 0.0405, "step": 16040 }, { "grad_norm": 0.39089056849479675, "learning_rate": 9.392539580417444e-05, "loss": 0.0408, "step": 16050 }, { "grad_norm": 0.36381271481513977, "learning_rate": 9.391551819560747e-05, "loss": 0.0418, "step": 16060 }, { "grad_norm": 0.3459925353527069, "learning_rate": 9.390563308307955e-05, "loss": 0.043, "step": 16070 }, { "grad_norm": 0.513481080532074, "learning_rate": 9.389574046827974e-05, "loss": 0.0454, "step": 16080 }, { "grad_norm": 0.3898126780986786, "learning_rate": 9.388584035289845e-05, "loss": 0.0438, "step": 16090 }, { "grad_norm": 0.4199931025505066, "learning_rate": 9.387593273862732e-05, "loss": 0.0492, "step": 16100 }, { "grad_norm": 0.4655994474887848, "learning_rate": 9.386601762715929e-05, "loss": 0.0497, "step": 16110 }, { "grad_norm": 0.4566764533519745, "learning_rate": 9.38560950201886e-05, "loss": 0.0537, "step": 16120 }, { "grad_norm": 0.4809117317199707, "learning_rate": 9.384616491941071e-05, "loss": 0.0436, "step": 16130 }, { "grad_norm": 0.442354679107666, "learning_rate": 9.383622732652245e-05, "loss": 0.0387, "step": 16140 }, { "grad_norm": 0.342912882566452, "learning_rate": 9.382628224322187e-05, "loss": 0.0477, "step": 16150 }, { "grad_norm": 0.3308088481426239, "learning_rate": 9.381632967120829e-05, "loss": 0.0501, "step": 16160 }, { "grad_norm": 0.3512941300868988, "learning_rate": 9.380636961218235e-05, "loss": 0.0411, "step": 16170 }, { "grad_norm": 0.3926427364349365, "learning_rate": 9.379640206784597e-05, "loss": 0.0393, "step": 16180 }, { "grad_norm": 0.4991394877433777, "learning_rate": 9.378642703990229e-05, "loss": 0.043, "step": 16190 }, { "grad_norm": 0.3013828992843628, "learning_rate": 9.37764445300558e-05, "loss": 0.043, "step": 16200 }, { "grad_norm": 0.28193163871765137, "learning_rate": 9.376645454001222e-05, "loss": 0.0508, "step": 16210 }, { "grad_norm": 0.24731427431106567, "learning_rate": 9.375645707147858e-05, "loss": 0.0377, "step": 16220 }, { "grad_norm": 0.28053054213523865, "learning_rate": 9.374645212616316e-05, "loss": 0.0503, "step": 16230 }, { "grad_norm": 0.4166642129421234, "learning_rate": 9.373643970577555e-05, "loss": 0.0466, "step": 16240 }, { "grad_norm": 0.5142087340354919, "learning_rate": 9.372641981202659e-05, "loss": 0.0461, "step": 16250 }, { "grad_norm": 0.4320780634880066, "learning_rate": 9.37163924466284e-05, "loss": 0.051, "step": 16260 }, { "grad_norm": 0.28017309308052063, "learning_rate": 9.370635761129438e-05, "loss": 0.0427, "step": 16270 }, { "grad_norm": 0.3497050106525421, "learning_rate": 9.36963153077392e-05, "loss": 0.0439, "step": 16280 }, { "grad_norm": 0.3140769600868225, "learning_rate": 9.368626553767888e-05, "loss": 0.0424, "step": 16290 }, { "grad_norm": 0.25205227732658386, "learning_rate": 9.367620830283057e-05, "loss": 0.0342, "step": 16300 }, { "grad_norm": 0.4808623790740967, "learning_rate": 9.366614360491281e-05, "loss": 0.0496, "step": 16310 }, { "grad_norm": 0.44814491271972656, "learning_rate": 9.365607144564539e-05, "loss": 0.0351, "step": 16320 }, { "grad_norm": 0.32051539421081543, "learning_rate": 9.364599182674934e-05, "loss": 0.0515, "step": 16330 }, { "grad_norm": 0.28748318552970886, "learning_rate": 9.3635904749947e-05, "loss": 0.0356, "step": 16340 }, { "grad_norm": 0.21321406960487366, "learning_rate": 9.362581021696202e-05, "loss": 0.0327, "step": 16350 }, { "grad_norm": 0.3684588372707367, "learning_rate": 9.361570822951921e-05, "loss": 0.0454, "step": 16360 }, { "grad_norm": 0.4741685092449188, "learning_rate": 9.360559878934476e-05, "loss": 0.0508, "step": 16370 }, { "grad_norm": 0.3579674959182739, "learning_rate": 9.359548189816611e-05, "loss": 0.0436, "step": 16380 }, { "grad_norm": 0.5117369890213013, "learning_rate": 9.358535755771193e-05, "loss": 0.0564, "step": 16390 }, { "grad_norm": 0.2971070110797882, "learning_rate": 9.357522576971221e-05, "loss": 0.0467, "step": 16400 }, { "grad_norm": 0.359355628490448, "learning_rate": 9.356508653589819e-05, "loss": 0.042, "step": 16410 }, { "grad_norm": 0.4080367088317871, "learning_rate": 9.355493985800237e-05, "loss": 0.0456, "step": 16420 }, { "grad_norm": 0.386947900056839, "learning_rate": 9.354478573775857e-05, "loss": 0.0516, "step": 16430 }, { "grad_norm": 0.3826042413711548, "learning_rate": 9.353462417690186e-05, "loss": 0.0459, "step": 16440 }, { "grad_norm": 0.38379743695259094, "learning_rate": 9.352445517716853e-05, "loss": 0.0445, "step": 16450 }, { "grad_norm": 0.41913458704948425, "learning_rate": 9.351427874029621e-05, "loss": 0.0493, "step": 16460 }, { "grad_norm": 0.24936281144618988, "learning_rate": 9.350409486802379e-05, "loss": 0.0371, "step": 16470 }, { "grad_norm": 0.33156079053878784, "learning_rate": 9.349390356209138e-05, "loss": 0.0501, "step": 16480 }, { "grad_norm": 0.38583216071128845, "learning_rate": 9.348370482424042e-05, "loss": 0.0466, "step": 16490 }, { "grad_norm": 0.5283547043800354, "learning_rate": 9.347349865621357e-05, "loss": 0.0361, "step": 16500 }, { "grad_norm": 0.422615647315979, "learning_rate": 9.346328505975481e-05, "loss": 0.0389, "step": 16510 }, { "grad_norm": 0.3047321140766144, "learning_rate": 9.345306403660936e-05, "loss": 0.0422, "step": 16520 }, { "grad_norm": 0.3831634521484375, "learning_rate": 9.344283558852371e-05, "loss": 0.0407, "step": 16530 }, { "grad_norm": 0.3596703112125397, "learning_rate": 9.343259971724563e-05, "loss": 0.0469, "step": 16540 }, { "grad_norm": 0.39384639263153076, "learning_rate": 9.342235642452413e-05, "loss": 0.0425, "step": 16550 }, { "grad_norm": 0.4375877380371094, "learning_rate": 9.341210571210954e-05, "loss": 0.0512, "step": 16560 }, { "grad_norm": 0.25907889008522034, "learning_rate": 9.340184758175338e-05, "loss": 0.0468, "step": 16570 }, { "grad_norm": 0.3409598469734192, "learning_rate": 9.339158203520854e-05, "loss": 0.049, "step": 16580 }, { "grad_norm": 0.29305005073547363, "learning_rate": 9.338130907422908e-05, "loss": 0.0333, "step": 16590 }, { "grad_norm": 0.2856234014034271, "learning_rate": 9.337102870057037e-05, "loss": 0.0358, "step": 16600 }, { "grad_norm": 0.3952939510345459, "learning_rate": 9.336074091598907e-05, "loss": 0.0318, "step": 16610 }, { "grad_norm": 0.3751756250858307, "learning_rate": 9.335044572224306e-05, "loss": 0.0312, "step": 16620 }, { "grad_norm": 0.30134207010269165, "learning_rate": 9.334014312109151e-05, "loss": 0.0312, "step": 16630 }, { "grad_norm": 0.5337060689926147, "learning_rate": 9.332983311429486e-05, "loss": 0.0501, "step": 16640 }, { "grad_norm": 0.3881077766418457, "learning_rate": 9.33195157036148e-05, "loss": 0.0448, "step": 16650 }, { "grad_norm": 0.4714665710926056, "learning_rate": 9.330919089081432e-05, "loss": 0.0406, "step": 16660 }, { "grad_norm": 0.32980233430862427, "learning_rate": 9.32988586776576e-05, "loss": 0.0527, "step": 16670 }, { "grad_norm": 0.36771756410598755, "learning_rate": 9.328851906591016e-05, "loss": 0.0409, "step": 16680 }, { "grad_norm": 0.43923184275627136, "learning_rate": 9.327817205733875e-05, "loss": 0.0456, "step": 16690 }, { "grad_norm": 0.38951316475868225, "learning_rate": 9.326781765371142e-05, "loss": 0.0397, "step": 16700 }, { "grad_norm": 0.3664199113845825, "learning_rate": 9.325745585679741e-05, "loss": 0.0412, "step": 16710 }, { "grad_norm": 0.43344050645828247, "learning_rate": 9.32470866683673e-05, "loss": 0.0598, "step": 16720 }, { "grad_norm": 0.4047490358352661, "learning_rate": 9.323671009019288e-05, "loss": 0.0485, "step": 16730 }, { "grad_norm": 0.33514508605003357, "learning_rate": 9.322632612404725e-05, "loss": 0.0404, "step": 16740 }, { "grad_norm": 0.3655657172203064, "learning_rate": 9.321593477170471e-05, "loss": 0.0523, "step": 16750 }, { "grad_norm": 0.4569515585899353, "learning_rate": 9.320553603494088e-05, "loss": 0.0407, "step": 16760 }, { "grad_norm": 0.34820717573165894, "learning_rate": 9.319512991553261e-05, "loss": 0.0391, "step": 16770 }, { "grad_norm": 0.443918913602829, "learning_rate": 9.318471641525803e-05, "loss": 0.0405, "step": 16780 }, { "grad_norm": 0.30242589116096497, "learning_rate": 9.317429553589652e-05, "loss": 0.035, "step": 16790 }, { "grad_norm": 0.3638206422328949, "learning_rate": 9.316386727922873e-05, "loss": 0.0398, "step": 16800 }, { "grad_norm": 0.5214499831199646, "learning_rate": 9.315343164703656e-05, "loss": 0.0342, "step": 16810 }, { "grad_norm": 0.3295597434043884, "learning_rate": 9.314298864110316e-05, "loss": 0.0356, "step": 16820 }, { "grad_norm": 0.2761304974555969, "learning_rate": 9.313253826321295e-05, "loss": 0.0382, "step": 16830 }, { "grad_norm": 0.2847634553909302, "learning_rate": 9.312208051515165e-05, "loss": 0.0319, "step": 16840 }, { "grad_norm": 0.3158356249332428, "learning_rate": 9.311161539870618e-05, "loss": 0.0382, "step": 16850 }, { "grad_norm": 0.2987338602542877, "learning_rate": 9.310114291566474e-05, "loss": 0.0484, "step": 16860 }, { "grad_norm": 0.2936008870601654, "learning_rate": 9.309066306781679e-05, "loss": 0.0354, "step": 16870 }, { "grad_norm": 0.4128132164478302, "learning_rate": 9.308017585695306e-05, "loss": 0.0378, "step": 16880 }, { "grad_norm": 0.26643142104148865, "learning_rate": 9.306968128486552e-05, "loss": 0.0389, "step": 16890 }, { "grad_norm": 0.3184453248977661, "learning_rate": 9.30591793533474e-05, "loss": 0.0326, "step": 16900 }, { "grad_norm": 0.35946527123451233, "learning_rate": 9.304867006419321e-05, "loss": 0.0342, "step": 16910 }, { "grad_norm": 0.503666877746582, "learning_rate": 9.303815341919868e-05, "loss": 0.0424, "step": 16920 }, { "grad_norm": 0.43975216150283813, "learning_rate": 9.302762942016084e-05, "loss": 0.0513, "step": 16930 }, { "grad_norm": 0.41356050968170166, "learning_rate": 9.301709806887792e-05, "loss": 0.0423, "step": 16940 }, { "grad_norm": 0.2950485944747925, "learning_rate": 9.300655936714948e-05, "loss": 0.0551, "step": 16950 }, { "grad_norm": 0.4600246846675873, "learning_rate": 9.299601331677627e-05, "loss": 0.0461, "step": 16960 }, { "grad_norm": 0.3669704496860504, "learning_rate": 9.298545991956033e-05, "loss": 0.0456, "step": 16970 }, { "grad_norm": 0.3220062255859375, "learning_rate": 9.297489917730493e-05, "loss": 0.0453, "step": 16980 }, { "grad_norm": 0.4787530303001404, "learning_rate": 9.296433109181464e-05, "loss": 0.0511, "step": 16990 }, { "grad_norm": 0.4869258999824524, "learning_rate": 9.295375566489523e-05, "loss": 0.0404, "step": 17000 }, { "grad_norm": 0.36399129033088684, "learning_rate": 9.294317289835379e-05, "loss": 0.051, "step": 17010 }, { "grad_norm": 0.36384275555610657, "learning_rate": 9.293258279399859e-05, "loss": 0.0559, "step": 17020 }, { "grad_norm": 0.2974782884120941, "learning_rate": 9.292198535363919e-05, "loss": 0.0353, "step": 17030 }, { "grad_norm": 0.40653106570243835, "learning_rate": 9.291138057908641e-05, "loss": 0.037, "step": 17040 }, { "grad_norm": 0.35072481632232666, "learning_rate": 9.290076847215234e-05, "loss": 0.0415, "step": 17050 }, { "grad_norm": 0.6029965281486511, "learning_rate": 9.289014903465025e-05, "loss": 0.0526, "step": 17060 }, { "grad_norm": 0.2957375645637512, "learning_rate": 9.287952226839475e-05, "loss": 0.0342, "step": 17070 }, { "grad_norm": 0.2837364971637726, "learning_rate": 9.286888817520164e-05, "loss": 0.0339, "step": 17080 }, { "grad_norm": 0.30728837847709656, "learning_rate": 9.285824675688803e-05, "loss": 0.0385, "step": 17090 }, { "grad_norm": 0.5079380869865417, "learning_rate": 9.28475980152722e-05, "loss": 0.0402, "step": 17100 }, { "grad_norm": 0.5383021831512451, "learning_rate": 9.283694195217379e-05, "loss": 0.0437, "step": 17110 }, { "grad_norm": 0.3914850354194641, "learning_rate": 9.282627856941356e-05, "loss": 0.0369, "step": 17120 }, { "grad_norm": 0.3452257513999939, "learning_rate": 9.281560786881363e-05, "loss": 0.039, "step": 17130 }, { "grad_norm": 0.28690966963768005, "learning_rate": 9.280492985219733e-05, "loss": 0.0364, "step": 17140 }, { "grad_norm": 0.31552958488464355, "learning_rate": 9.279424452138924e-05, "loss": 0.0412, "step": 17150 }, { "grad_norm": 0.38681766390800476, "learning_rate": 9.278355187821517e-05, "loss": 0.0392, "step": 17160 }, { "grad_norm": 0.3900647461414337, "learning_rate": 9.277285192450224e-05, "loss": 0.0461, "step": 17170 }, { "grad_norm": 0.40054193139076233, "learning_rate": 9.276214466207875e-05, "loss": 0.0425, "step": 17180 }, { "grad_norm": 0.3110947906970978, "learning_rate": 9.275143009277427e-05, "loss": 0.0347, "step": 17190 }, { "grad_norm": 0.41674530506134033, "learning_rate": 9.274070821841964e-05, "loss": 0.0451, "step": 17200 }, { "grad_norm": 0.47037363052368164, "learning_rate": 9.272997904084696e-05, "loss": 0.0399, "step": 17210 }, { "grad_norm": 0.31669923663139343, "learning_rate": 9.271924256188951e-05, "loss": 0.0335, "step": 17220 }, { "grad_norm": 0.3380902409553528, "learning_rate": 9.270849878338189e-05, "loss": 0.0341, "step": 17230 }, { "grad_norm": 0.41700005531311035, "learning_rate": 9.269774770715991e-05, "loss": 0.0386, "step": 17240 }, { "grad_norm": 0.3213154077529907, "learning_rate": 9.268698933506061e-05, "loss": 0.0523, "step": 17250 }, { "grad_norm": 0.3308027684688568, "learning_rate": 9.267622366892235e-05, "loss": 0.0391, "step": 17260 }, { "grad_norm": 0.48166635632514954, "learning_rate": 9.266545071058465e-05, "loss": 0.0378, "step": 17270 }, { "grad_norm": 0.3393622636795044, "learning_rate": 9.265467046188833e-05, "loss": 0.0323, "step": 17280 }, { "grad_norm": 0.40849629044532776, "learning_rate": 9.264388292467543e-05, "loss": 0.043, "step": 17290 }, { "grad_norm": 0.305668443441391, "learning_rate": 9.263308810078926e-05, "loss": 0.0363, "step": 17300 }, { "grad_norm": 0.32252660393714905, "learning_rate": 9.262228599207434e-05, "loss": 0.0395, "step": 17310 }, { "grad_norm": 0.2699960172176361, "learning_rate": 9.261147660037647e-05, "loss": 0.028, "step": 17320 }, { "grad_norm": 0.3274536728858948, "learning_rate": 9.26006599275427e-05, "loss": 0.0351, "step": 17330 }, { "grad_norm": 0.4036010503768921, "learning_rate": 9.258983597542124e-05, "loss": 0.0413, "step": 17340 }, { "grad_norm": 0.39173707365989685, "learning_rate": 9.257900474586167e-05, "loss": 0.0359, "step": 17350 }, { "grad_norm": 0.46178293228149414, "learning_rate": 9.256816624071471e-05, "loss": 0.0466, "step": 17360 }, { "grad_norm": 0.421760618686676, "learning_rate": 9.25573204618324e-05, "loss": 0.0454, "step": 17370 }, { "grad_norm": 0.3283107876777649, "learning_rate": 9.254646741106796e-05, "loss": 0.0373, "step": 17380 }, { "grad_norm": 0.3925209641456604, "learning_rate": 9.253560709027589e-05, "loss": 0.0345, "step": 17390 }, { "grad_norm": 0.37392640113830566, "learning_rate": 9.252473950131192e-05, "loss": 0.0438, "step": 17400 }, { "grad_norm": 0.28544676303863525, "learning_rate": 9.251386464603302e-05, "loss": 0.03, "step": 17410 }, { "grad_norm": 0.3339138627052307, "learning_rate": 9.250298252629741e-05, "loss": 0.0426, "step": 17420 }, { "grad_norm": 0.30599445104599, "learning_rate": 9.249209314396454e-05, "loss": 0.0357, "step": 17430 }, { "grad_norm": 0.3963446021080017, "learning_rate": 9.248119650089513e-05, "loss": 0.0478, "step": 17440 }, { "grad_norm": 0.3794762194156647, "learning_rate": 9.247029259895108e-05, "loss": 0.04, "step": 17450 }, { "grad_norm": 0.41989898681640625, "learning_rate": 9.24593814399956e-05, "loss": 0.038, "step": 17460 }, { "grad_norm": 0.34379005432128906, "learning_rate": 9.244846302589309e-05, "loss": 0.0339, "step": 17470 }, { "grad_norm": 0.37454530596733093, "learning_rate": 9.243753735850923e-05, "loss": 0.0395, "step": 17480 }, { "grad_norm": 0.3704153597354889, "learning_rate": 9.24266044397109e-05, "loss": 0.0336, "step": 17490 }, { "grad_norm": 0.2634260356426239, "learning_rate": 9.241566427136624e-05, "loss": 0.0312, "step": 17500 }, { "grad_norm": 0.4580734670162201, "learning_rate": 9.240471685534463e-05, "loss": 0.0407, "step": 17510 }, { "grad_norm": 0.3117523193359375, "learning_rate": 9.239376219351667e-05, "loss": 0.0398, "step": 17520 }, { "grad_norm": 0.47773492336273193, "learning_rate": 9.238280028775425e-05, "loss": 0.0354, "step": 17530 }, { "grad_norm": 0.3391498029232025, "learning_rate": 9.237183113993041e-05, "loss": 0.0501, "step": 17540 }, { "grad_norm": 0.3274223804473877, "learning_rate": 9.236085475191952e-05, "loss": 0.0372, "step": 17550 }, { "grad_norm": 0.3143867552280426, "learning_rate": 9.234987112559709e-05, "loss": 0.0539, "step": 17560 }, { "grad_norm": 0.36667364835739136, "learning_rate": 9.233888026283999e-05, "loss": 0.0506, "step": 17570 }, { "grad_norm": 0.4681747257709503, "learning_rate": 9.232788216552619e-05, "loss": 0.0475, "step": 17580 }, { "grad_norm": 0.41159266233444214, "learning_rate": 9.231687683553502e-05, "loss": 0.0527, "step": 17590 }, { "grad_norm": 0.3375900983810425, "learning_rate": 9.230586427474698e-05, "loss": 0.061, "step": 17600 }, { "grad_norm": 0.3157854378223419, "learning_rate": 9.229484448504379e-05, "loss": 0.0491, "step": 17610 }, { "grad_norm": 0.33440065383911133, "learning_rate": 9.228381746830843e-05, "loss": 0.0412, "step": 17620 }, { "grad_norm": 0.3238910436630249, "learning_rate": 9.227278322642514e-05, "loss": 0.0308, "step": 17630 }, { "grad_norm": 0.332862913608551, "learning_rate": 9.226174176127937e-05, "loss": 0.0417, "step": 17640 }, { "grad_norm": 0.23452480137348175, "learning_rate": 9.22506930747578e-05, "loss": 0.037, "step": 17650 }, { "grad_norm": 0.4203382134437561, "learning_rate": 9.223963716874831e-05, "loss": 0.05, "step": 17660 }, { "grad_norm": 0.3246818780899048, "learning_rate": 9.222857404514012e-05, "loss": 0.0423, "step": 17670 }, { "grad_norm": 0.5431563258171082, "learning_rate": 9.221750370582355e-05, "loss": 0.0404, "step": 17680 }, { "grad_norm": 0.3715410828590393, "learning_rate": 9.220642615269028e-05, "loss": 0.0444, "step": 17690 }, { "grad_norm": 0.43801116943359375, "learning_rate": 9.219534138763311e-05, "loss": 0.0437, "step": 17700 }, { "grad_norm": 0.3707423210144043, "learning_rate": 9.218424941254613e-05, "loss": 0.0505, "step": 17710 }, { "grad_norm": 0.385774165391922, "learning_rate": 9.217315022932468e-05, "loss": 0.053, "step": 17720 }, { "grad_norm": 0.44679421186447144, "learning_rate": 9.216204383986528e-05, "loss": 0.0461, "step": 17730 }, { "grad_norm": 0.36501097679138184, "learning_rate": 9.215093024606574e-05, "loss": 0.0518, "step": 17740 }, { "grad_norm": 0.330013245344162, "learning_rate": 9.213980944982506e-05, "loss": 0.0577, "step": 17750 }, { "grad_norm": 0.36591947078704834, "learning_rate": 9.212868145304346e-05, "loss": 0.0445, "step": 17760 }, { "grad_norm": 0.4838453531265259, "learning_rate": 9.211754625762241e-05, "loss": 0.0378, "step": 17770 }, { "grad_norm": 0.36755385994911194, "learning_rate": 9.210640386546463e-05, "loss": 0.0541, "step": 17780 }, { "grad_norm": 0.3060106337070465, "learning_rate": 9.209525427847405e-05, "loss": 0.028, "step": 17790 }, { "grad_norm": 0.33026009798049927, "learning_rate": 9.208409749855583e-05, "loss": 0.0351, "step": 17800 }, { "grad_norm": 0.40564483404159546, "learning_rate": 9.207293352761633e-05, "loss": 0.0509, "step": 17810 }, { "grad_norm": 0.44274792075157166, "learning_rate": 9.206176236756319e-05, "loss": 0.0384, "step": 17820 }, { "grad_norm": 0.30300381779670715, "learning_rate": 9.205058402030525e-05, "loss": 0.0423, "step": 17830 }, { "grad_norm": 0.3571844696998596, "learning_rate": 9.203939848775259e-05, "loss": 0.0376, "step": 17840 }, { "grad_norm": 0.3363604247570038, "learning_rate": 9.202820577181652e-05, "loss": 0.0373, "step": 17850 }, { "grad_norm": 1.1656358242034912, "learning_rate": 9.201700587440953e-05, "loss": 0.0413, "step": 17860 }, { "grad_norm": 0.42321160435676575, "learning_rate": 9.200579879744544e-05, "loss": 0.0468, "step": 17870 }, { "grad_norm": 0.3622592091560364, "learning_rate": 9.199458454283918e-05, "loss": 0.0406, "step": 17880 }, { "grad_norm": 0.5549548864364624, "learning_rate": 9.198336311250697e-05, "loss": 0.0463, "step": 17890 }, { "grad_norm": 0.49080923199653625, "learning_rate": 9.197213450836626e-05, "loss": 0.0462, "step": 17900 }, { "grad_norm": 0.5061219930648804, "learning_rate": 9.19608987323357e-05, "loss": 0.0457, "step": 17910 }, { "grad_norm": 0.41285282373428345, "learning_rate": 9.194965578633517e-05, "loss": 0.0452, "step": 17920 }, { "grad_norm": 0.3852699100971222, "learning_rate": 9.193840567228582e-05, "loss": 0.0448, "step": 17930 }, { "grad_norm": 0.34649547934532166, "learning_rate": 9.192714839210994e-05, "loss": 0.0384, "step": 17940 }, { "grad_norm": 0.41131699085235596, "learning_rate": 9.19158839477311e-05, "loss": 0.0491, "step": 17950 }, { "grad_norm": 0.41430914402008057, "learning_rate": 9.190461234107411e-05, "loss": 0.046, "step": 17960 }, { "grad_norm": 0.4114876091480255, "learning_rate": 9.189333357406496e-05, "loss": 0.0329, "step": 17970 }, { "grad_norm": 0.5066189169883728, "learning_rate": 9.188204764863089e-05, "loss": 0.0611, "step": 17980 }, { "grad_norm": 0.4640897810459137, "learning_rate": 9.187075456670033e-05, "loss": 0.0357, "step": 17990 }, { "grad_norm": 0.30395662784576416, "learning_rate": 9.1859454330203e-05, "loss": 0.0429, "step": 18000 }, { "grad_norm": 0.34776920080184937, "learning_rate": 9.18481469410698e-05, "loss": 0.0566, "step": 18010 }, { "grad_norm": 0.32769185304641724, "learning_rate": 9.183683240123281e-05, "loss": 0.0431, "step": 18020 }, { "grad_norm": 0.43581363558769226, "learning_rate": 9.182551071262541e-05, "loss": 0.0548, "step": 18030 }, { "grad_norm": 0.3742230534553528, "learning_rate": 9.181418187718218e-05, "loss": 0.0449, "step": 18040 }, { "grad_norm": 0.32859188318252563, "learning_rate": 9.180284589683888e-05, "loss": 0.0459, "step": 18050 }, { "grad_norm": 0.4665042459964752, "learning_rate": 9.17915027735325e-05, "loss": 0.0395, "step": 18060 }, { "grad_norm": 0.3165935277938843, "learning_rate": 9.178015250920133e-05, "loss": 0.0393, "step": 18070 }, { "grad_norm": 0.36813127994537354, "learning_rate": 9.176879510578477e-05, "loss": 0.0346, "step": 18080 }, { "grad_norm": 0.25714778900146484, "learning_rate": 9.17574305652235e-05, "loss": 0.0393, "step": 18090 }, { "grad_norm": 0.28992459177970886, "learning_rate": 9.174605888945942e-05, "loss": 0.0498, "step": 18100 }, { "grad_norm": 0.5551841259002686, "learning_rate": 9.173468008043564e-05, "loss": 0.0386, "step": 18110 }, { "grad_norm": 0.33235612511634827, "learning_rate": 9.172329414009648e-05, "loss": 0.0409, "step": 18120 }, { "grad_norm": 0.4444948136806488, "learning_rate": 9.171190107038747e-05, "loss": 0.0411, "step": 18130 }, { "grad_norm": 0.27230048179626465, "learning_rate": 9.170050087325541e-05, "loss": 0.0336, "step": 18140 }, { "grad_norm": 0.4165767729282379, "learning_rate": 9.168909355064824e-05, "loss": 0.0361, "step": 18150 }, { "grad_norm": 0.2612009048461914, "learning_rate": 9.167767910451519e-05, "loss": 0.0385, "step": 18160 }, { "grad_norm": 0.3533433973789215, "learning_rate": 9.166625753680669e-05, "loss": 0.0394, "step": 18170 }, { "grad_norm": 0.2969367206096649, "learning_rate": 9.165482884947431e-05, "loss": 0.0312, "step": 18180 }, { "grad_norm": 0.22521287202835083, "learning_rate": 9.164339304447098e-05, "loss": 0.0359, "step": 18190 }, { "grad_norm": 0.26101750135421753, "learning_rate": 9.163195012375072e-05, "loss": 0.0278, "step": 18200 }, { "grad_norm": 0.3069603145122528, "learning_rate": 9.16205000892688e-05, "loss": 0.0355, "step": 18210 }, { "grad_norm": 0.3418967127799988, "learning_rate": 9.160904294298175e-05, "loss": 0.0342, "step": 18220 }, { "grad_norm": 0.25015637278556824, "learning_rate": 9.159757868684727e-05, "loss": 0.0327, "step": 18230 }, { "grad_norm": 0.29523956775665283, "learning_rate": 9.15861073228243e-05, "loss": 0.035, "step": 18240 }, { "grad_norm": 0.3785862624645233, "learning_rate": 9.157462885287296e-05, "loss": 0.0477, "step": 18250 }, { "grad_norm": 0.41803503036499023, "learning_rate": 9.156314327895461e-05, "loss": 0.0357, "step": 18260 }, { "grad_norm": 0.4145306348800659, "learning_rate": 9.155165060303185e-05, "loss": 0.0441, "step": 18270 }, { "grad_norm": 0.35795167088508606, "learning_rate": 9.154015082706841e-05, "loss": 0.0354, "step": 18280 }, { "grad_norm": 0.3371224105358124, "learning_rate": 9.152864395302936e-05, "loss": 0.0377, "step": 18290 }, { "grad_norm": 0.39080530405044556, "learning_rate": 9.151712998288085e-05, "loss": 0.0328, "step": 18300 }, { "grad_norm": 0.2517135441303253, "learning_rate": 9.150560891859031e-05, "loss": 0.0342, "step": 18310 }, { "grad_norm": 0.31965553760528564, "learning_rate": 9.14940807621264e-05, "loss": 0.0388, "step": 18320 }, { "grad_norm": 0.2974845767021179, "learning_rate": 9.148254551545894e-05, "loss": 0.035, "step": 18330 }, { "grad_norm": 0.31113025546073914, "learning_rate": 9.147100318055901e-05, "loss": 0.0371, "step": 18340 }, { "grad_norm": 0.33805662393569946, "learning_rate": 9.145945375939888e-05, "loss": 0.0345, "step": 18350 }, { "grad_norm": 0.2320527583360672, "learning_rate": 9.144789725395203e-05, "loss": 0.0368, "step": 18360 }, { "grad_norm": 0.386930376291275, "learning_rate": 9.14363336661931e-05, "loss": 0.0397, "step": 18370 }, { "grad_norm": 0.43875712156295776, "learning_rate": 9.142476299809806e-05, "loss": 0.0391, "step": 18380 }, { "grad_norm": 0.3414101004600525, "learning_rate": 9.1413185251644e-05, "loss": 0.0375, "step": 18390 }, { "grad_norm": 0.2873205840587616, "learning_rate": 9.140160042880923e-05, "loss": 0.0456, "step": 18400 }, { "grad_norm": 0.27988168597221375, "learning_rate": 9.139000853157327e-05, "loss": 0.0447, "step": 18410 }, { "grad_norm": 0.33601558208465576, "learning_rate": 9.137840956191688e-05, "loss": 0.0508, "step": 18420 }, { "grad_norm": 0.2841019928455353, "learning_rate": 9.136680352182199e-05, "loss": 0.0391, "step": 18430 }, { "grad_norm": 0.4129248559474945, "learning_rate": 9.135519041327177e-05, "loss": 0.0354, "step": 18440 }, { "grad_norm": 0.3629591464996338, "learning_rate": 9.134357023825058e-05, "loss": 0.0483, "step": 18450 }, { "grad_norm": 0.44117048382759094, "learning_rate": 9.133194299874398e-05, "loss": 0.0396, "step": 18460 }, { "grad_norm": 0.49951881170272827, "learning_rate": 9.132030869673876e-05, "loss": 0.0457, "step": 18470 }, { "grad_norm": 0.3579710125923157, "learning_rate": 9.130866733422288e-05, "loss": 0.0378, "step": 18480 }, { "grad_norm": 0.313949316740036, "learning_rate": 9.129701891318556e-05, "loss": 0.0324, "step": 18490 }, { "grad_norm": 0.31006887555122375, "learning_rate": 9.128536343561718e-05, "loss": 0.0477, "step": 18500 }, { "grad_norm": 0.35651135444641113, "learning_rate": 9.127370090350934e-05, "loss": 0.0373, "step": 18510 }, { "grad_norm": 0.35906755924224854, "learning_rate": 9.126203131885487e-05, "loss": 0.0342, "step": 18520 }, { "grad_norm": 0.34554222226142883, "learning_rate": 9.125035468364775e-05, "loss": 0.0398, "step": 18530 }, { "grad_norm": 0.32800737023353577, "learning_rate": 9.123867099988322e-05, "loss": 0.035, "step": 18540 }, { "grad_norm": 0.325864315032959, "learning_rate": 9.122698026955769e-05, "loss": 0.0342, "step": 18550 }, { "grad_norm": 0.35391148924827576, "learning_rate": 9.12152824946688e-05, "loss": 0.0546, "step": 18560 }, { "grad_norm": 0.4458441436290741, "learning_rate": 9.120357767721538e-05, "loss": 0.0383, "step": 18570 }, { "grad_norm": 0.4388173520565033, "learning_rate": 9.119186581919745e-05, "loss": 0.0362, "step": 18580 }, { "grad_norm": 0.4632221758365631, "learning_rate": 9.118014692261624e-05, "loss": 0.0393, "step": 18590 }, { "grad_norm": 0.4153634011745453, "learning_rate": 9.116842098947422e-05, "loss": 0.0533, "step": 18600 }, { "grad_norm": 0.35375720262527466, "learning_rate": 9.115668802177499e-05, "loss": 0.0362, "step": 18610 }, { "grad_norm": 0.29179441928863525, "learning_rate": 9.114494802152342e-05, "loss": 0.0511, "step": 18620 }, { "grad_norm": 0.5499624013900757, "learning_rate": 9.113320099072555e-05, "loss": 0.0742, "step": 18630 }, { "grad_norm": 0.3818393051624298, "learning_rate": 9.112144693138864e-05, "loss": 0.0404, "step": 18640 }, { "grad_norm": 0.39382168650627136, "learning_rate": 9.110968584552111e-05, "loss": 0.0408, "step": 18650 }, { "grad_norm": 0.4133985936641693, "learning_rate": 9.109791773513264e-05, "loss": 0.0468, "step": 18660 }, { "grad_norm": 0.43211811780929565, "learning_rate": 9.108614260223403e-05, "loss": 0.0434, "step": 18670 }, { "grad_norm": 0.4077740013599396, "learning_rate": 9.107436044883738e-05, "loss": 0.042, "step": 18680 }, { "grad_norm": 0.3061167895793915, "learning_rate": 9.10625712769559e-05, "loss": 0.0367, "step": 18690 }, { "grad_norm": 0.3064802885055542, "learning_rate": 9.105077508860406e-05, "loss": 0.035, "step": 18700 }, { "grad_norm": 0.3394472002983093, "learning_rate": 9.103897188579751e-05, "loss": 0.0354, "step": 18710 }, { "grad_norm": 0.3709528148174286, "learning_rate": 9.102716167055308e-05, "loss": 0.0305, "step": 18720 }, { "grad_norm": 0.3240360617637634, "learning_rate": 9.10153444448888e-05, "loss": 0.0655, "step": 18730 }, { "grad_norm": 0.4014483392238617, "learning_rate": 9.100352021082393e-05, "loss": 0.0335, "step": 18740 }, { "grad_norm": 0.41800305247306824, "learning_rate": 9.099168897037891e-05, "loss": 0.0325, "step": 18750 }, { "grad_norm": 0.4387398660182953, "learning_rate": 9.097985072557538e-05, "loss": 0.039, "step": 18760 }, { "grad_norm": 0.30850842595100403, "learning_rate": 9.096800547843615e-05, "loss": 0.0566, "step": 18770 }, { "grad_norm": 0.5057215690612793, "learning_rate": 9.095615323098526e-05, "loss": 0.0494, "step": 18780 }, { "grad_norm": 0.46182507276535034, "learning_rate": 9.094429398524795e-05, "loss": 0.0403, "step": 18790 }, { "grad_norm": 0.3535148501396179, "learning_rate": 9.093242774325061e-05, "loss": 0.0356, "step": 18800 }, { "grad_norm": 0.39883625507354736, "learning_rate": 9.092055450702088e-05, "loss": 0.0364, "step": 18810 }, { "grad_norm": 0.24894915521144867, "learning_rate": 9.090867427858756e-05, "loss": 0.0398, "step": 18820 }, { "grad_norm": 0.23162320256233215, "learning_rate": 9.089678705998066e-05, "loss": 0.0348, "step": 18830 }, { "grad_norm": 0.3061756491661072, "learning_rate": 9.088489285323139e-05, "loss": 0.0385, "step": 18840 }, { "grad_norm": 0.36830437183380127, "learning_rate": 9.087299166037212e-05, "loss": 0.0398, "step": 18850 }, { "grad_norm": 0.3383004665374756, "learning_rate": 9.086108348343647e-05, "loss": 0.0453, "step": 18860 }, { "grad_norm": 0.36016932129859924, "learning_rate": 9.08491683244592e-05, "loss": 0.0413, "step": 18870 }, { "grad_norm": 0.25384750962257385, "learning_rate": 9.08372461854763e-05, "loss": 0.0371, "step": 18880 }, { "grad_norm": 0.27437683939933777, "learning_rate": 9.082531706852492e-05, "loss": 0.0372, "step": 18890 }, { "grad_norm": 0.22948989272117615, "learning_rate": 9.081338097564342e-05, "loss": 0.051, "step": 18900 }, { "grad_norm": 0.2811123728752136, "learning_rate": 9.080143790887137e-05, "loss": 0.033, "step": 18910 }, { "grad_norm": 0.3677198588848114, "learning_rate": 9.07894878702495e-05, "loss": 0.0336, "step": 18920 }, { "grad_norm": 0.3267376124858856, "learning_rate": 9.077753086181974e-05, "loss": 0.0404, "step": 18930 }, { "grad_norm": 0.3263775706291199, "learning_rate": 9.076556688562524e-05, "loss": 0.029, "step": 18940 }, { "grad_norm": 0.3433248698711395, "learning_rate": 9.075359594371029e-05, "loss": 0.0298, "step": 18950 }, { "grad_norm": 0.31791743636131287, "learning_rate": 9.07416180381204e-05, "loss": 0.0394, "step": 18960 }, { "grad_norm": 0.49628520011901855, "learning_rate": 9.072963317090228e-05, "loss": 0.04, "step": 18970 }, { "grad_norm": 0.338711142539978, "learning_rate": 9.071764134410382e-05, "loss": 0.0383, "step": 18980 }, { "grad_norm": 0.2806318402290344, "learning_rate": 9.070564255977407e-05, "loss": 0.0341, "step": 18990 }, { "grad_norm": 0.39687156677246094, "learning_rate": 9.06936368199633e-05, "loss": 0.0402, "step": 19000 }, { "grad_norm": 0.3175465762615204, "learning_rate": 9.0681624126723e-05, "loss": 0.0493, "step": 19010 }, { "grad_norm": 0.41311153769493103, "learning_rate": 9.066960448210576e-05, "loss": 0.038, "step": 19020 }, { "grad_norm": 0.3003394603729248, "learning_rate": 9.065757788816543e-05, "loss": 0.0314, "step": 19030 }, { "grad_norm": 0.2868156433105469, "learning_rate": 9.064554434695705e-05, "loss": 0.0453, "step": 19040 }, { "grad_norm": 0.3006803095340729, "learning_rate": 9.063350386053677e-05, "loss": 0.0453, "step": 19050 }, { "grad_norm": 0.42000797390937805, "learning_rate": 9.062145643096202e-05, "loss": 0.0387, "step": 19060 }, { "grad_norm": 0.30710798501968384, "learning_rate": 9.060940206029136e-05, "loss": 0.0485, "step": 19070 }, { "grad_norm": 0.3229750096797943, "learning_rate": 9.059734075058457e-05, "loss": 0.0378, "step": 19080 }, { "grad_norm": 0.3548387885093689, "learning_rate": 9.058527250390257e-05, "loss": 0.0486, "step": 19090 }, { "grad_norm": 0.24538660049438477, "learning_rate": 9.057319732230752e-05, "loss": 0.0496, "step": 19100 }, { "grad_norm": 0.28143173456192017, "learning_rate": 9.056111520786273e-05, "loss": 0.0371, "step": 19110 }, { "grad_norm": 0.3566204607486725, "learning_rate": 9.054902616263268e-05, "loss": 0.0313, "step": 19120 }, { "grad_norm": 0.3359836935997009, "learning_rate": 9.05369301886831e-05, "loss": 0.047, "step": 19130 }, { "grad_norm": 0.31467875838279724, "learning_rate": 9.052482728808083e-05, "loss": 0.0434, "step": 19140 }, { "grad_norm": 0.28953754901885986, "learning_rate": 9.051271746289391e-05, "loss": 0.0315, "step": 19150 }, { "grad_norm": 0.35622331500053406, "learning_rate": 9.050060071519162e-05, "loss": 0.0416, "step": 19160 }, { "grad_norm": 0.27053043246269226, "learning_rate": 9.048847704704437e-05, "loss": 0.0325, "step": 19170 }, { "grad_norm": 0.42397671937942505, "learning_rate": 9.047634646052376e-05, "loss": 0.0375, "step": 19180 }, { "grad_norm": 0.35236045718193054, "learning_rate": 9.046420895770256e-05, "loss": 0.0505, "step": 19190 }, { "grad_norm": 0.46262720227241516, "learning_rate": 9.045206454065473e-05, "loss": 0.0393, "step": 19200 }, { "grad_norm": 0.3883040249347687, "learning_rate": 9.043991321145546e-05, "loss": 0.0327, "step": 19210 }, { "grad_norm": 0.43271949887275696, "learning_rate": 9.042775497218105e-05, "loss": 0.0356, "step": 19220 }, { "grad_norm": 0.4744669795036316, "learning_rate": 9.041558982490901e-05, "loss": 0.0474, "step": 19230 }, { "grad_norm": 0.3109349310398102, "learning_rate": 9.040341777171805e-05, "loss": 0.0353, "step": 19240 }, { "grad_norm": 0.38420936465263367, "learning_rate": 9.039123881468802e-05, "loss": 0.0373, "step": 19250 }, { "grad_norm": 0.23624421656131744, "learning_rate": 9.037905295589998e-05, "loss": 0.0326, "step": 19260 }, { "grad_norm": 0.3286912143230438, "learning_rate": 9.036686019743617e-05, "loss": 0.0275, "step": 19270 }, { "grad_norm": 0.3392379581928253, "learning_rate": 9.035466054137997e-05, "loss": 0.0331, "step": 19280 }, { "grad_norm": 0.5714257955551147, "learning_rate": 9.0342453989816e-05, "loss": 0.0519, "step": 19290 }, { "grad_norm": 0.3868132531642914, "learning_rate": 9.033024054483e-05, "loss": 0.0388, "step": 19300 }, { "grad_norm": 0.8131640553474426, "learning_rate": 9.031802020850894e-05, "loss": 0.0438, "step": 19310 }, { "grad_norm": 0.5265710949897766, "learning_rate": 9.030579298294092e-05, "loss": 0.0494, "step": 19320 }, { "grad_norm": 0.48219484090805054, "learning_rate": 9.029355887021524e-05, "loss": 0.0585, "step": 19330 }, { "grad_norm": 0.5765953063964844, "learning_rate": 9.028131787242238e-05, "loss": 0.0624, "step": 19340 }, { "grad_norm": 0.3353717625141144, "learning_rate": 9.026906999165399e-05, "loss": 0.0508, "step": 19350 }, { "grad_norm": 0.35402560234069824, "learning_rate": 9.025681523000291e-05, "loss": 0.0339, "step": 19360 }, { "grad_norm": 0.40463125705718994, "learning_rate": 9.024455358956315e-05, "loss": 0.0339, "step": 19370 }, { "grad_norm": 0.30392447113990784, "learning_rate": 9.023228507242984e-05, "loss": 0.0433, "step": 19380 }, { "grad_norm": 0.3173176646232605, "learning_rate": 9.022000968069937e-05, "loss": 0.0413, "step": 19390 }, { "grad_norm": 0.31391647458076477, "learning_rate": 9.020772741646928e-05, "loss": 0.0334, "step": 19400 }, { "grad_norm": 0.3539845049381256, "learning_rate": 9.019543828183826e-05, "loss": 0.0376, "step": 19410 }, { "grad_norm": 0.3668762743473053, "learning_rate": 9.018314227890616e-05, "loss": 0.0351, "step": 19420 }, { "grad_norm": 0.33833053708076477, "learning_rate": 9.017083940977408e-05, "loss": 0.0521, "step": 19430 }, { "grad_norm": 0.2588155269622803, "learning_rate": 9.015852967654422e-05, "loss": 0.0318, "step": 19440 }, { "grad_norm": 0.28490105271339417, "learning_rate": 9.014621308131996e-05, "loss": 0.038, "step": 19450 }, { "grad_norm": 0.3048681914806366, "learning_rate": 9.01338896262059e-05, "loss": 0.034, "step": 19460 }, { "grad_norm": 0.34906089305877686, "learning_rate": 9.012155931330777e-05, "loss": 0.0475, "step": 19470 }, { "grad_norm": 0.4641868472099304, "learning_rate": 9.010922214473246e-05, "loss": 0.034, "step": 19480 }, { "grad_norm": 0.32357731461524963, "learning_rate": 9.009687812258808e-05, "loss": 0.0516, "step": 19490 }, { "grad_norm": 0.44601911306381226, "learning_rate": 9.00845272489839e-05, "loss": 0.0367, "step": 19500 }, { "grad_norm": 0.38136374950408936, "learning_rate": 9.007216952603031e-05, "loss": 0.0542, "step": 19510 }, { "grad_norm": 0.3312832713127136, "learning_rate": 9.005980495583894e-05, "loss": 0.0334, "step": 19520 }, { "grad_norm": 0.35979557037353516, "learning_rate": 9.004743354052252e-05, "loss": 0.037, "step": 19530 }, { "grad_norm": 0.37538278102874756, "learning_rate": 9.003505528219503e-05, "loss": 0.0336, "step": 19540 }, { "grad_norm": 0.3716221749782562, "learning_rate": 9.002267018297154e-05, "loss": 0.0369, "step": 19550 }, { "grad_norm": 0.23969417810440063, "learning_rate": 9.001027824496834e-05, "loss": 0.051, "step": 19560 }, { "grad_norm": 0.25878241658210754, "learning_rate": 8.999787947030287e-05, "loss": 0.0382, "step": 19570 }, { "grad_norm": 0.3666553795337677, "learning_rate": 8.998547386109376e-05, "loss": 0.0416, "step": 19580 }, { "grad_norm": 0.3745211362838745, "learning_rate": 8.997306141946073e-05, "loss": 0.0376, "step": 19590 }, { "grad_norm": 0.4059131145477295, "learning_rate": 8.996064214752481e-05, "loss": 0.0428, "step": 19600 }, { "grad_norm": 0.3428979218006134, "learning_rate": 8.994821604740806e-05, "loss": 0.0401, "step": 19610 }, { "grad_norm": 0.3468630909919739, "learning_rate": 8.993578312123377e-05, "loss": 0.0404, "step": 19620 }, { "grad_norm": 0.2287653684616089, "learning_rate": 8.992334337112639e-05, "loss": 0.0392, "step": 19630 }, { "grad_norm": 0.2941053509712219, "learning_rate": 8.991089679921154e-05, "loss": 0.0325, "step": 19640 }, { "grad_norm": 0.4007962942123413, "learning_rate": 8.989844340761599e-05, "loss": 0.0373, "step": 19650 }, { "grad_norm": 0.36282455921173096, "learning_rate": 8.988598319846768e-05, "loss": 0.0338, "step": 19660 }, { "grad_norm": 0.253379225730896, "learning_rate": 8.987351617389574e-05, "loss": 0.0358, "step": 19670 }, { "grad_norm": 0.35665789246559143, "learning_rate": 8.98610423360304e-05, "loss": 0.0412, "step": 19680 }, { "grad_norm": 0.28934574127197266, "learning_rate": 8.984856168700317e-05, "loss": 0.044, "step": 19690 }, { "grad_norm": 0.46341270208358765, "learning_rate": 8.983607422894658e-05, "loss": 0.0365, "step": 19700 }, { "grad_norm": 0.2352667599916458, "learning_rate": 8.982357996399442e-05, "loss": 0.0411, "step": 19710 }, { "grad_norm": 0.41151902079582214, "learning_rate": 8.981107889428164e-05, "loss": 0.0397, "step": 19720 }, { "grad_norm": 0.3594461977481842, "learning_rate": 8.979857102194428e-05, "loss": 0.0343, "step": 19730 }, { "grad_norm": 0.36716318130493164, "learning_rate": 8.978605634911968e-05, "loss": 0.0339, "step": 19740 }, { "grad_norm": 0.33647647500038147, "learning_rate": 8.977353487794616e-05, "loss": 0.027, "step": 19750 }, { "grad_norm": 0.348723441362381, "learning_rate": 8.976100661056334e-05, "loss": 0.0468, "step": 19760 }, { "grad_norm": 0.32155337929725647, "learning_rate": 8.974847154911197e-05, "loss": 0.0316, "step": 19770 }, { "grad_norm": 0.28777509927749634, "learning_rate": 8.973592969573393e-05, "loss": 0.0398, "step": 19780 }, { "grad_norm": 0.375971257686615, "learning_rate": 8.972338105257228e-05, "loss": 0.0498, "step": 19790 }, { "grad_norm": 0.2766897976398468, "learning_rate": 8.971082562177125e-05, "loss": 0.0365, "step": 19800 }, { "grad_norm": 0.35989829897880554, "learning_rate": 8.96982634054762e-05, "loss": 0.038, "step": 19810 }, { "grad_norm": 0.29879218339920044, "learning_rate": 8.96856944058337e-05, "loss": 0.0301, "step": 19820 }, { "grad_norm": 0.33463290333747864, "learning_rate": 8.967311862499144e-05, "loss": 0.0241, "step": 19830 }, { "grad_norm": 0.26750490069389343, "learning_rate": 8.966053606509825e-05, "loss": 0.0311, "step": 19840 }, { "grad_norm": 0.43592220544815063, "learning_rate": 8.964794672830417e-05, "loss": 0.0484, "step": 19850 }, { "grad_norm": 0.49434947967529297, "learning_rate": 8.963535061676038e-05, "loss": 0.0484, "step": 19860 }, { "grad_norm": 0.4159170687198639, "learning_rate": 8.962274773261918e-05, "loss": 0.0354, "step": 19870 }, { "grad_norm": 0.31503233313560486, "learning_rate": 8.961013807803409e-05, "loss": 0.046, "step": 19880 }, { "grad_norm": 0.3125114440917969, "learning_rate": 8.959752165515973e-05, "loss": 0.041, "step": 19890 }, { "grad_norm": 0.3484271764755249, "learning_rate": 8.958489846615193e-05, "loss": 0.0588, "step": 19900 }, { "grad_norm": 0.307608425617218, "learning_rate": 8.957226851316762e-05, "loss": 0.0377, "step": 19910 }, { "grad_norm": 0.3879927396774292, "learning_rate": 8.955963179836493e-05, "loss": 0.0374, "step": 19920 }, { "grad_norm": 0.35755351185798645, "learning_rate": 8.954698832390312e-05, "loss": 0.035, "step": 19930 }, { "grad_norm": 0.4283407926559448, "learning_rate": 8.953433809194263e-05, "loss": 0.0373, "step": 19940 }, { "grad_norm": 0.4270661175251007, "learning_rate": 8.9521681104645e-05, "loss": 0.0338, "step": 19950 }, { "grad_norm": 0.47168898582458496, "learning_rate": 8.9509017364173e-05, "loss": 0.0404, "step": 19960 }, { "grad_norm": 0.40518057346343994, "learning_rate": 8.949634687269052e-05, "loss": 0.0365, "step": 19970 }, { "grad_norm": 0.4282500445842743, "learning_rate": 8.948366963236259e-05, "loss": 0.0418, "step": 19980 }, { "grad_norm": 0.3100643754005432, "learning_rate": 8.947098564535538e-05, "loss": 0.0375, "step": 19990 }, { "grad_norm": 0.34277230501174927, "learning_rate": 8.945829491383627e-05, "loss": 0.0369, "step": 20000 }, { "grad_norm": 0.2983798086643219, "learning_rate": 8.944559743997374e-05, "loss": 0.031, "step": 20010 }, { "grad_norm": 0.3486606180667877, "learning_rate": 8.943289322593746e-05, "loss": 0.0334, "step": 20020 }, { "grad_norm": 0.4802112579345703, "learning_rate": 8.942018227389821e-05, "loss": 0.033, "step": 20030 }, { "grad_norm": 0.35629838705062866, "learning_rate": 8.940746458602795e-05, "loss": 0.0524, "step": 20040 }, { "grad_norm": 0.36252108216285706, "learning_rate": 8.939474016449979e-05, "loss": 0.0418, "step": 20050 }, { "grad_norm": 0.3049205541610718, "learning_rate": 8.938200901148799e-05, "loss": 0.0293, "step": 20060 }, { "grad_norm": 0.479642778635025, "learning_rate": 8.936927112916795e-05, "loss": 0.0471, "step": 20070 }, { "grad_norm": 0.2957889437675476, "learning_rate": 8.935652651971622e-05, "loss": 0.0348, "step": 20080 }, { "grad_norm": 0.34088313579559326, "learning_rate": 8.934377518531052e-05, "loss": 0.0385, "step": 20090 }, { "grad_norm": 0.36147740483283997, "learning_rate": 8.933101712812967e-05, "loss": 0.0363, "step": 20100 }, { "grad_norm": 0.336372435092926, "learning_rate": 8.931825235035374e-05, "loss": 0.0328, "step": 20110 }, { "grad_norm": 0.35762348771095276, "learning_rate": 8.930548085416382e-05, "loss": 0.0359, "step": 20120 }, { "grad_norm": 0.37347057461738586, "learning_rate": 8.92927026417422e-05, "loss": 0.0424, "step": 20130 }, { "grad_norm": 0.2708245813846588, "learning_rate": 8.92799177152724e-05, "loss": 0.03, "step": 20140 }, { "grad_norm": 0.2975698411464691, "learning_rate": 8.926712607693895e-05, "loss": 0.04, "step": 20150 }, { "grad_norm": 0.34782496094703674, "learning_rate": 8.925432772892762e-05, "loss": 0.0437, "step": 20160 }, { "grad_norm": 0.39588791131973267, "learning_rate": 8.924152267342529e-05, "loss": 0.0465, "step": 20170 }, { "grad_norm": 0.3074321448802948, "learning_rate": 8.922871091261998e-05, "loss": 0.0424, "step": 20180 }, { "grad_norm": 0.2601807713508606, "learning_rate": 8.92158924487009e-05, "loss": 0.0349, "step": 20190 }, { "grad_norm": 0.4086369574069977, "learning_rate": 8.920306728385834e-05, "loss": 0.0344, "step": 20200 }, { "grad_norm": 0.34179553389549255, "learning_rate": 8.919023542028379e-05, "loss": 0.0327, "step": 20210 }, { "grad_norm": 0.30686306953430176, "learning_rate": 8.917739686016988e-05, "loss": 0.037, "step": 20220 }, { "grad_norm": 0.4448140859603882, "learning_rate": 8.916455160571033e-05, "loss": 0.0472, "step": 20230 }, { "grad_norm": 0.42360424995422363, "learning_rate": 8.915169965910008e-05, "loss": 0.0402, "step": 20240 }, { "grad_norm": 0.441972553730011, "learning_rate": 8.913884102253514e-05, "loss": 0.0366, "step": 20250 }, { "grad_norm": 0.2962409555912018, "learning_rate": 8.912597569821273e-05, "loss": 0.0354, "step": 20260 }, { "grad_norm": 0.2628454566001892, "learning_rate": 8.911310368833118e-05, "loss": 0.03, "step": 20270 }, { "grad_norm": 0.26647379994392395, "learning_rate": 8.910022499508994e-05, "loss": 0.0318, "step": 20280 }, { "grad_norm": 0.23882363736629486, "learning_rate": 8.908733962068965e-05, "loss": 0.0467, "step": 20290 }, { "grad_norm": 0.23914547264575958, "learning_rate": 8.907444756733207e-05, "loss": 0.0318, "step": 20300 }, { "grad_norm": 0.41744646430015564, "learning_rate": 8.906154883722006e-05, "loss": 0.0393, "step": 20310 }, { "grad_norm": 0.35238832235336304, "learning_rate": 8.904864343255773e-05, "loss": 0.0337, "step": 20320 }, { "grad_norm": 0.3390534222126007, "learning_rate": 8.90357313555502e-05, "loss": 0.034, "step": 20330 }, { "grad_norm": 0.2622157335281372, "learning_rate": 8.90228126084038e-05, "loss": 0.0413, "step": 20340 }, { "grad_norm": 0.3791421055793762, "learning_rate": 8.900988719332601e-05, "loss": 0.039, "step": 20350 }, { "grad_norm": 0.3186052739620209, "learning_rate": 8.899695511252542e-05, "loss": 0.0415, "step": 20360 }, { "grad_norm": 0.2937624752521515, "learning_rate": 8.898401636821176e-05, "loss": 0.0431, "step": 20370 }, { "grad_norm": 0.4477388262748718, "learning_rate": 8.897107096259593e-05, "loss": 0.0526, "step": 20380 }, { "grad_norm": 0.3925635814666748, "learning_rate": 8.895811889788994e-05, "loss": 0.0369, "step": 20390 }, { "grad_norm": 0.36676302552223206, "learning_rate": 8.894516017630692e-05, "loss": 0.0364, "step": 20400 }, { "grad_norm": 0.45520031452178955, "learning_rate": 8.893219480006118e-05, "loss": 0.0385, "step": 20410 }, { "grad_norm": 0.3736468553543091, "learning_rate": 8.891922277136817e-05, "loss": 0.0338, "step": 20420 }, { "grad_norm": 0.28139951825141907, "learning_rate": 8.890624409244441e-05, "loss": 0.0287, "step": 20430 }, { "grad_norm": 0.36585158109664917, "learning_rate": 8.889325876550763e-05, "loss": 0.0347, "step": 20440 }, { "grad_norm": 0.25524693727493286, "learning_rate": 8.888026679277666e-05, "loss": 0.0331, "step": 20450 }, { "grad_norm": 0.2615645229816437, "learning_rate": 8.886726817647147e-05, "loss": 0.063, "step": 20460 }, { "grad_norm": 0.2690083086490631, "learning_rate": 8.885426291881319e-05, "loss": 0.0469, "step": 20470 }, { "grad_norm": 0.4590023159980774, "learning_rate": 8.884125102202401e-05, "loss": 0.0415, "step": 20480 }, { "grad_norm": 0.38347703218460083, "learning_rate": 8.882823248832736e-05, "loss": 0.0285, "step": 20490 }, { "grad_norm": 0.28243109583854675, "learning_rate": 8.881520731994772e-05, "loss": 0.0345, "step": 20500 }, { "grad_norm": 0.3324192762374878, "learning_rate": 8.880217551911077e-05, "loss": 0.0417, "step": 20510 }, { "grad_norm": 0.32491374015808105, "learning_rate": 8.878913708804323e-05, "loss": 0.0361, "step": 20520 }, { "grad_norm": 0.3447684943675995, "learning_rate": 8.877609202897308e-05, "loss": 0.0416, "step": 20530 }, { "grad_norm": 0.2993110418319702, "learning_rate": 8.876304034412933e-05, "loss": 0.0288, "step": 20540 }, { "grad_norm": 0.3253159523010254, "learning_rate": 8.874998203574214e-05, "loss": 0.0399, "step": 20550 }, { "grad_norm": 0.3839776813983917, "learning_rate": 8.873691710604284e-05, "loss": 0.0332, "step": 20560 }, { "grad_norm": 0.331225723028183, "learning_rate": 8.872384555726387e-05, "loss": 0.0325, "step": 20570 }, { "grad_norm": 0.23622344434261322, "learning_rate": 8.871076739163878e-05, "loss": 0.0319, "step": 20580 }, { "grad_norm": 0.31950250267982483, "learning_rate": 8.86976826114023e-05, "loss": 0.0389, "step": 20590 }, { "grad_norm": 0.3174765408039093, "learning_rate": 8.868459121879023e-05, "loss": 0.0396, "step": 20600 }, { "grad_norm": 0.3780258297920227, "learning_rate": 8.867149321603956e-05, "loss": 0.0513, "step": 20610 }, { "grad_norm": 0.41272032260894775, "learning_rate": 8.865838860538835e-05, "loss": 0.0395, "step": 20620 }, { "grad_norm": 0.34697219729423523, "learning_rate": 8.864527738907585e-05, "loss": 0.0417, "step": 20630 }, { "grad_norm": 0.38345688581466675, "learning_rate": 8.863215956934239e-05, "loss": 0.0391, "step": 20640 }, { "grad_norm": 0.3445656895637512, "learning_rate": 8.861903514842947e-05, "loss": 0.0294, "step": 20650 }, { "grad_norm": 0.3456600606441498, "learning_rate": 8.860590412857966e-05, "loss": 0.0346, "step": 20660 }, { "grad_norm": 0.2838993966579437, "learning_rate": 8.85927665120367e-05, "loss": 0.0289, "step": 20670 }, { "grad_norm": 0.384112149477005, "learning_rate": 8.857962230104546e-05, "loss": 0.0358, "step": 20680 }, { "grad_norm": 0.31582823395729065, "learning_rate": 8.856647149785193e-05, "loss": 0.0304, "step": 20690 }, { "grad_norm": 0.37990859150886536, "learning_rate": 8.855331410470322e-05, "loss": 0.0345, "step": 20700 }, { "grad_norm": 0.37223830819129944, "learning_rate": 8.854015012384756e-05, "loss": 0.0277, "step": 20710 }, { "grad_norm": 0.27597591280937195, "learning_rate": 8.852697955753433e-05, "loss": 0.0407, "step": 20720 }, { "grad_norm": 0.28691884875297546, "learning_rate": 8.851380240801399e-05, "loss": 0.0405, "step": 20730 }, { "grad_norm": 0.1947188526391983, "learning_rate": 8.850061867753818e-05, "loss": 0.0255, "step": 20740 }, { "grad_norm": 0.40139833092689514, "learning_rate": 8.848742836835963e-05, "loss": 0.0436, "step": 20750 }, { "grad_norm": 0.33634647727012634, "learning_rate": 8.847423148273221e-05, "loss": 0.0369, "step": 20760 }, { "grad_norm": 0.32697781920433044, "learning_rate": 8.846102802291092e-05, "loss": 0.045, "step": 20770 }, { "grad_norm": 0.3308013379573822, "learning_rate": 8.844781799115183e-05, "loss": 0.0455, "step": 20780 }, { "grad_norm": 0.2675623297691345, "learning_rate": 8.84346013897122e-05, "loss": 0.0321, "step": 20790 }, { "grad_norm": 0.3424820005893707, "learning_rate": 8.842137822085038e-05, "loss": 0.0462, "step": 20800 }, { "grad_norm": 0.27533310651779175, "learning_rate": 8.840814848682585e-05, "loss": 0.0342, "step": 20810 }, { "grad_norm": 0.23284660279750824, "learning_rate": 8.83949121898992e-05, "loss": 0.0484, "step": 20820 }, { "grad_norm": 0.3522247076034546, "learning_rate": 8.838166933233217e-05, "loss": 0.0332, "step": 20830 }, { "grad_norm": 0.3768649101257324, "learning_rate": 8.83684199163876e-05, "loss": 0.0434, "step": 20840 }, { "grad_norm": 0.3238702118396759, "learning_rate": 8.835516394432943e-05, "loss": 0.0447, "step": 20850 }, { "grad_norm": 0.33109238743782043, "learning_rate": 8.834190141842276e-05, "loss": 0.0476, "step": 20860 }, { "grad_norm": 0.41000592708587646, "learning_rate": 8.83286323409338e-05, "loss": 0.0385, "step": 20870 }, { "grad_norm": 0.3839711546897888, "learning_rate": 8.831535671412986e-05, "loss": 0.0422, "step": 20880 }, { "grad_norm": 0.32556992769241333, "learning_rate": 8.830207454027938e-05, "loss": 0.0453, "step": 20890 }, { "grad_norm": 0.3322961628437042, "learning_rate": 8.828878582165192e-05, "loss": 0.0367, "step": 20900 }, { "grad_norm": 0.42656514048576355, "learning_rate": 8.827549056051818e-05, "loss": 0.0424, "step": 20910 }, { "grad_norm": 0.37350696325302124, "learning_rate": 8.826218875914993e-05, "loss": 0.0359, "step": 20920 }, { "grad_norm": 0.2650122344493866, "learning_rate": 8.82488804198201e-05, "loss": 0.0288, "step": 20930 }, { "grad_norm": 0.23431304097175598, "learning_rate": 8.82355655448027e-05, "loss": 0.0331, "step": 20940 }, { "grad_norm": 0.31635957956314087, "learning_rate": 8.822224413637293e-05, "loss": 0.0368, "step": 20950 }, { "grad_norm": 0.30609482526779175, "learning_rate": 8.820891619680697e-05, "loss": 0.0389, "step": 20960 }, { "grad_norm": 0.2764101028442383, "learning_rate": 8.819558172838227e-05, "loss": 0.032, "step": 20970 }, { "grad_norm": 0.2987784147262573, "learning_rate": 8.818224073337731e-05, "loss": 0.0287, "step": 20980 }, { "grad_norm": 0.2496466189622879, "learning_rate": 8.816889321407169e-05, "loss": 0.0313, "step": 20990 }, { "grad_norm": 0.23985736072063446, "learning_rate": 8.815553917274615e-05, "loss": 0.0401, "step": 21000 }, { "grad_norm": 0.3127313256263733, "learning_rate": 8.81421786116825e-05, "loss": 0.0436, "step": 21010 }, { "grad_norm": 0.29648154973983765, "learning_rate": 8.812881153316373e-05, "loss": 0.0344, "step": 21020 }, { "grad_norm": 0.2715626060962677, "learning_rate": 8.81154379394739e-05, "loss": 0.0371, "step": 21030 }, { "grad_norm": 0.3848862051963806, "learning_rate": 8.810205783289818e-05, "loss": 0.0349, "step": 21040 }, { "grad_norm": 0.3457869291305542, "learning_rate": 8.808867121572286e-05, "loss": 0.038, "step": 21050 }, { "grad_norm": 0.3581317961215973, "learning_rate": 8.807527809023537e-05, "loss": 0.0429, "step": 21060 }, { "grad_norm": 0.23348680138587952, "learning_rate": 8.80618784587242e-05, "loss": 0.0325, "step": 21070 }, { "grad_norm": 0.25565412640571594, "learning_rate": 8.804847232347902e-05, "loss": 0.0335, "step": 21080 }, { "grad_norm": 0.30666622519493103, "learning_rate": 8.803505968679054e-05, "loss": 0.0406, "step": 21090 }, { "grad_norm": 0.3393286168575287, "learning_rate": 8.802164055095061e-05, "loss": 0.0442, "step": 21100 }, { "grad_norm": 0.3240087330341339, "learning_rate": 8.80082149182522e-05, "loss": 0.0346, "step": 21110 }, { "grad_norm": 0.30279606580734253, "learning_rate": 8.79947827909894e-05, "loss": 0.0312, "step": 21120 }, { "grad_norm": 0.31671199202537537, "learning_rate": 8.798134417145738e-05, "loss": 0.0439, "step": 21130 }, { "grad_norm": 0.3632678687572479, "learning_rate": 8.796789906195243e-05, "loss": 0.0355, "step": 21140 }, { "grad_norm": 0.40568843483924866, "learning_rate": 8.795444746477195e-05, "loss": 0.0374, "step": 21150 }, { "grad_norm": 0.32208487391471863, "learning_rate": 8.794098938221446e-05, "loss": 0.0341, "step": 21160 }, { "grad_norm": 0.48613253235816956, "learning_rate": 8.792752481657957e-05, "loss": 0.0409, "step": 21170 }, { "grad_norm": 0.3567437529563904, "learning_rate": 8.791405377016802e-05, "loss": 0.0337, "step": 21180 }, { "grad_norm": 0.365081250667572, "learning_rate": 8.790057624528163e-05, "loss": 0.0407, "step": 21190 }, { "grad_norm": 0.473076194524765, "learning_rate": 8.788709224422333e-05, "loss": 0.0396, "step": 21200 }, { "grad_norm": 0.33441826701164246, "learning_rate": 8.787360176929717e-05, "loss": 0.0349, "step": 21210 }, { "grad_norm": 0.38956740498542786, "learning_rate": 8.786010482280834e-05, "loss": 0.0407, "step": 21220 }, { "grad_norm": 0.24684493243694305, "learning_rate": 8.784660140706306e-05, "loss": 0.0391, "step": 21230 }, { "grad_norm": 0.29215607047080994, "learning_rate": 8.783309152436872e-05, "loss": 0.029, "step": 21240 }, { "grad_norm": 0.32093045115470886, "learning_rate": 8.781957517703375e-05, "loss": 0.0315, "step": 21250 }, { "grad_norm": 0.32422614097595215, "learning_rate": 8.780605236736776e-05, "loss": 0.0399, "step": 21260 }, { "grad_norm": 0.26285457611083984, "learning_rate": 8.779252309768142e-05, "loss": 0.0368, "step": 21270 }, { "grad_norm": 0.3434000313282013, "learning_rate": 8.777898737028652e-05, "loss": 0.0367, "step": 21280 }, { "grad_norm": 0.2989313006401062, "learning_rate": 8.776544518749591e-05, "loss": 0.0363, "step": 21290 }, { "grad_norm": 0.24190908670425415, "learning_rate": 8.775189655162364e-05, "loss": 0.0341, "step": 21300 }, { "grad_norm": 0.39665859937667847, "learning_rate": 8.773834146498474e-05, "loss": 0.0363, "step": 21310 }, { "grad_norm": 0.2792888879776001, "learning_rate": 8.772477992989545e-05, "loss": 0.0322, "step": 21320 }, { "grad_norm": 0.3054414391517639, "learning_rate": 8.771121194867304e-05, "loss": 0.0476, "step": 21330 }, { "grad_norm": 0.2740621864795685, "learning_rate": 8.769763752363589e-05, "loss": 0.0318, "step": 21340 }, { "grad_norm": 0.3345670998096466, "learning_rate": 8.768405665710352e-05, "loss": 0.0397, "step": 21350 }, { "grad_norm": 0.3448067605495453, "learning_rate": 8.767046935139655e-05, "loss": 0.0425, "step": 21360 }, { "grad_norm": 0.3159189224243164, "learning_rate": 8.765687560883666e-05, "loss": 0.037, "step": 21370 }, { "grad_norm": 0.4028487801551819, "learning_rate": 8.764327543174664e-05, "loss": 0.032, "step": 21380 }, { "grad_norm": 0.32093656063079834, "learning_rate": 8.762966882245038e-05, "loss": 0.0325, "step": 21390 }, { "grad_norm": 0.3781660795211792, "learning_rate": 8.761605578327291e-05, "loss": 0.042, "step": 21400 }, { "grad_norm": 0.32235923409461975, "learning_rate": 8.76024363165403e-05, "loss": 0.0465, "step": 21410 }, { "grad_norm": 0.27292194962501526, "learning_rate": 8.758881042457976e-05, "loss": 0.0395, "step": 21420 }, { "grad_norm": 0.26414787769317627, "learning_rate": 8.757517810971957e-05, "loss": 0.0321, "step": 21430 }, { "grad_norm": 0.30244940519332886, "learning_rate": 8.756153937428913e-05, "loss": 0.0379, "step": 21440 }, { "grad_norm": 0.33984896540641785, "learning_rate": 8.754789422061889e-05, "loss": 0.0546, "step": 21450 }, { "grad_norm": 0.5119495391845703, "learning_rate": 8.753424265104052e-05, "loss": 0.0463, "step": 21460 }, { "grad_norm": 0.4640659987926483, "learning_rate": 8.752058466788659e-05, "loss": 0.0566, "step": 21470 }, { "grad_norm": 0.381560742855072, "learning_rate": 8.750692027349097e-05, "loss": 0.0413, "step": 21480 }, { "grad_norm": 0.5558539032936096, "learning_rate": 8.749324947018847e-05, "loss": 0.0458, "step": 21490 }, { "grad_norm": 0.4595921039581299, "learning_rate": 8.747957226031507e-05, "loss": 0.0423, "step": 21500 }, { "grad_norm": 0.46118274331092834, "learning_rate": 8.746588864620787e-05, "loss": 0.046, "step": 21510 }, { "grad_norm": 0.264355331659317, "learning_rate": 8.745219863020498e-05, "loss": 0.0355, "step": 21520 }, { "grad_norm": 0.3195052444934845, "learning_rate": 8.743850221464564e-05, "loss": 0.0353, "step": 21530 }, { "grad_norm": 0.3011867105960846, "learning_rate": 8.742479940187026e-05, "loss": 0.039, "step": 21540 }, { "grad_norm": 0.5245694518089294, "learning_rate": 8.74110901942202e-05, "loss": 0.0423, "step": 21550 }, { "grad_norm": 0.31015992164611816, "learning_rate": 8.739737459403803e-05, "loss": 0.0385, "step": 21560 }, { "grad_norm": 0.28513675928115845, "learning_rate": 8.738365260366737e-05, "loss": 0.0352, "step": 21570 }, { "grad_norm": 0.31794580817222595, "learning_rate": 8.736992422545292e-05, "loss": 0.0304, "step": 21580 }, { "grad_norm": 0.28774914145469666, "learning_rate": 8.73561894617405e-05, "loss": 0.0454, "step": 21590 }, { "grad_norm": 0.296476811170578, "learning_rate": 8.734244831487697e-05, "loss": 0.034, "step": 21600 }, { "grad_norm": 0.37572017312049866, "learning_rate": 8.732870078721035e-05, "loss": 0.0457, "step": 21610 }, { "grad_norm": 0.31537696719169617, "learning_rate": 8.731494688108972e-05, "loss": 0.0337, "step": 21620 }, { "grad_norm": 0.4303317070007324, "learning_rate": 8.730118659886523e-05, "loss": 0.0403, "step": 21630 }, { "grad_norm": 0.42132675647735596, "learning_rate": 8.728741994288814e-05, "loss": 0.0372, "step": 21640 }, { "grad_norm": 0.3335736393928528, "learning_rate": 8.727364691551079e-05, "loss": 0.0355, "step": 21650 }, { "grad_norm": 0.3512105345726013, "learning_rate": 8.725986751908661e-05, "loss": 0.0408, "step": 21660 }, { "grad_norm": 0.3616984784603119, "learning_rate": 8.724608175597016e-05, "loss": 0.0348, "step": 21670 }, { "grad_norm": 0.3784736692905426, "learning_rate": 8.723228962851699e-05, "loss": 0.0324, "step": 21680 }, { "grad_norm": 0.4429040253162384, "learning_rate": 8.721849113908385e-05, "loss": 0.0372, "step": 21690 }, { "grad_norm": 0.4450344145298004, "learning_rate": 8.720468629002848e-05, "loss": 0.0349, "step": 21700 }, { "grad_norm": 0.41084370017051697, "learning_rate": 8.719087508370978e-05, "loss": 0.0328, "step": 21710 }, { "grad_norm": 0.4243687391281128, "learning_rate": 8.717705752248772e-05, "loss": 0.0501, "step": 21720 }, { "grad_norm": 0.40733280777931213, "learning_rate": 8.71632336087233e-05, "loss": 0.0329, "step": 21730 }, { "grad_norm": 0.23721235990524292, "learning_rate": 8.71494033447787e-05, "loss": 0.0375, "step": 21740 }, { "grad_norm": 0.2770586907863617, "learning_rate": 8.713556673301708e-05, "loss": 0.028, "step": 21750 }, { "grad_norm": 0.3207847774028778, "learning_rate": 8.712172377580278e-05, "loss": 0.0446, "step": 21760 }, { "grad_norm": 0.39839187264442444, "learning_rate": 8.710787447550114e-05, "loss": 0.0378, "step": 21770 }, { "grad_norm": 0.3537343144416809, "learning_rate": 8.70940188344787e-05, "loss": 0.0483, "step": 21780 }, { "grad_norm": 0.38797324895858765, "learning_rate": 8.708015685510293e-05, "loss": 0.0354, "step": 21790 }, { "grad_norm": 0.3791443109512329, "learning_rate": 8.706628853974252e-05, "loss": 0.0413, "step": 21800 }, { "grad_norm": 0.28083235025405884, "learning_rate": 8.705241389076715e-05, "loss": 0.0268, "step": 21810 }, { "grad_norm": 0.27148380875587463, "learning_rate": 8.703853291054764e-05, "loss": 0.0313, "step": 21820 }, { "grad_norm": 0.46675747632980347, "learning_rate": 8.702464560145587e-05, "loss": 0.0401, "step": 21830 }, { "grad_norm": 0.3937617540359497, "learning_rate": 8.701075196586476e-05, "loss": 0.0329, "step": 21840 }, { "grad_norm": 0.3177621364593506, "learning_rate": 8.699685200614842e-05, "loss": 0.0338, "step": 21850 }, { "grad_norm": 0.309712290763855, "learning_rate": 8.698294572468193e-05, "loss": 0.0419, "step": 21860 }, { "grad_norm": 0.29990237951278687, "learning_rate": 8.696903312384148e-05, "loss": 0.0352, "step": 21870 }, { "grad_norm": 0.29580581188201904, "learning_rate": 8.695511420600439e-05, "loss": 0.029, "step": 21880 }, { "grad_norm": 0.3096904158592224, "learning_rate": 8.694118897354901e-05, "loss": 0.0291, "step": 21890 }, { "grad_norm": 0.3373510241508484, "learning_rate": 8.692725742885478e-05, "loss": 0.033, "step": 21900 }, { "grad_norm": 0.47409799695014954, "learning_rate": 8.691331957430221e-05, "loss": 0.0386, "step": 21910 }, { "grad_norm": 0.2106814980506897, "learning_rate": 8.68993754122729e-05, "loss": 0.0335, "step": 21920 }, { "grad_norm": 0.2606680691242218, "learning_rate": 8.688542494514955e-05, "loss": 0.0262, "step": 21930 }, { "grad_norm": 0.318248450756073, "learning_rate": 8.68714681753159e-05, "loss": 0.0323, "step": 21940 }, { "grad_norm": 0.28720131516456604, "learning_rate": 8.685750510515676e-05, "loss": 0.0268, "step": 21950 }, { "grad_norm": 0.2954096496105194, "learning_rate": 8.684353573705805e-05, "loss": 0.0317, "step": 21960 }, { "grad_norm": 0.2546662390232086, "learning_rate": 8.682956007340677e-05, "loss": 0.0316, "step": 21970 }, { "grad_norm": 0.45572105050086975, "learning_rate": 8.681557811659095e-05, "loss": 0.0326, "step": 21980 }, { "grad_norm": 0.28820252418518066, "learning_rate": 8.680158986899974e-05, "loss": 0.0318, "step": 21990 }, { "grad_norm": 0.2800094485282898, "learning_rate": 8.678759533302335e-05, "loss": 0.0294, "step": 22000 }, { "grad_norm": 0.3402911424636841, "learning_rate": 8.677359451105308e-05, "loss": 0.0318, "step": 22010 }, { "grad_norm": 0.260919988155365, "learning_rate": 8.675958740548123e-05, "loss": 0.0428, "step": 22020 }, { "grad_norm": 0.3090541660785675, "learning_rate": 8.674557401870129e-05, "loss": 0.0378, "step": 22030 }, { "grad_norm": 0.27607882022857666, "learning_rate": 8.673155435310775e-05, "loss": 0.0336, "step": 22040 }, { "grad_norm": 0.3420400619506836, "learning_rate": 8.671752841109617e-05, "loss": 0.0301, "step": 22050 }, { "grad_norm": 0.39316800236701965, "learning_rate": 8.670349619506321e-05, "loss": 0.0266, "step": 22060 }, { "grad_norm": 0.2504676878452301, "learning_rate": 8.66894577074066e-05, "loss": 0.0305, "step": 22070 }, { "grad_norm": 0.26449403166770935, "learning_rate": 8.667541295052513e-05, "loss": 0.0256, "step": 22080 }, { "grad_norm": 0.32420769333839417, "learning_rate": 8.666136192681865e-05, "loss": 0.0452, "step": 22090 }, { "grad_norm": 0.39279183745384216, "learning_rate": 8.664730463868811e-05, "loss": 0.0344, "step": 22100 }, { "grad_norm": 0.3926219344139099, "learning_rate": 8.663324108853552e-05, "loss": 0.0388, "step": 22110 }, { "grad_norm": 0.30129218101501465, "learning_rate": 8.661917127876395e-05, "loss": 0.0456, "step": 22120 }, { "grad_norm": 0.3553774356842041, "learning_rate": 8.660509521177754e-05, "loss": 0.0351, "step": 22130 }, { "grad_norm": 0.274795800447464, "learning_rate": 8.65910128899815e-05, "loss": 0.0355, "step": 22140 }, { "grad_norm": 0.3756554424762726, "learning_rate": 8.657692431578214e-05, "loss": 0.0355, "step": 22150 }, { "grad_norm": 0.31302300095558167, "learning_rate": 8.656282949158679e-05, "loss": 0.0295, "step": 22160 }, { "grad_norm": 0.3316704332828522, "learning_rate": 8.654872841980388e-05, "loss": 0.0256, "step": 22170 }, { "grad_norm": 0.35963118076324463, "learning_rate": 8.653462110284289e-05, "loss": 0.0335, "step": 22180 }, { "grad_norm": 0.3289918005466461, "learning_rate": 8.652050754311437e-05, "loss": 0.0338, "step": 22190 }, { "grad_norm": 0.4547741115093231, "learning_rate": 8.650638774302995e-05, "loss": 0.0331, "step": 22200 }, { "grad_norm": 0.4293861389160156, "learning_rate": 8.649226170500233e-05, "loss": 0.0366, "step": 22210 }, { "grad_norm": 0.3643707036972046, "learning_rate": 8.647812943144524e-05, "loss": 0.0347, "step": 22220 }, { "grad_norm": 0.40528246760368347, "learning_rate": 8.646399092477351e-05, "loss": 0.0286, "step": 22230 }, { "grad_norm": 0.4175010919570923, "learning_rate": 8.644984618740301e-05, "loss": 0.0452, "step": 22240 }, { "grad_norm": 0.3813316822052002, "learning_rate": 8.643569522175073e-05, "loss": 0.0273, "step": 22250 }, { "grad_norm": 0.7034029364585876, "learning_rate": 8.642153803023463e-05, "loss": 0.0437, "step": 22260 }, { "grad_norm": 0.2480369359254837, "learning_rate": 8.640737461527383e-05, "loss": 0.0309, "step": 22270 }, { "grad_norm": 0.3477410674095154, "learning_rate": 8.639320497928845e-05, "loss": 0.0346, "step": 22280 }, { "grad_norm": 0.38461846113204956, "learning_rate": 8.637902912469969e-05, "loss": 0.0315, "step": 22290 }, { "grad_norm": 0.5115823745727539, "learning_rate": 8.636484705392982e-05, "loss": 0.0365, "step": 22300 }, { "grad_norm": 0.2780977487564087, "learning_rate": 8.635065876940216e-05, "loss": 0.0312, "step": 22310 }, { "grad_norm": 0.3317892551422119, "learning_rate": 8.633646427354112e-05, "loss": 0.0377, "step": 22320 }, { "grad_norm": 0.2932570278644562, "learning_rate": 8.632226356877213e-05, "loss": 0.027, "step": 22330 }, { "grad_norm": 0.23422496020793915, "learning_rate": 8.630805665752173e-05, "loss": 0.0402, "step": 22340 }, { "grad_norm": 0.3766602575778961, "learning_rate": 8.629384354221748e-05, "loss": 0.0403, "step": 22350 }, { "grad_norm": 0.2965952455997467, "learning_rate": 8.627962422528797e-05, "loss": 0.032, "step": 22360 }, { "grad_norm": 0.36538049578666687, "learning_rate": 8.626539870916296e-05, "loss": 0.0501, "step": 22370 }, { "grad_norm": 0.3251574635505676, "learning_rate": 8.625116699627317e-05, "loss": 0.036, "step": 22380 }, { "grad_norm": 0.3623878061771393, "learning_rate": 8.623692908905041e-05, "loss": 0.0248, "step": 22390 }, { "grad_norm": 0.3173255920410156, "learning_rate": 8.622268498992755e-05, "loss": 0.043, "step": 22400 }, { "grad_norm": 0.34947654604911804, "learning_rate": 8.620843470133851e-05, "loss": 0.0494, "step": 22410 }, { "grad_norm": 0.34756070375442505, "learning_rate": 8.619417822571829e-05, "loss": 0.0376, "step": 22420 }, { "grad_norm": 0.30625173449516296, "learning_rate": 8.617991556550292e-05, "loss": 0.0314, "step": 22430 }, { "grad_norm": 0.32209888100624084, "learning_rate": 8.616564672312952e-05, "loss": 0.0326, "step": 22440 }, { "grad_norm": 0.3266729414463043, "learning_rate": 8.61513717010362e-05, "loss": 0.0383, "step": 22450 }, { "grad_norm": 0.4439829885959625, "learning_rate": 8.613709050166221e-05, "loss": 0.0377, "step": 22460 }, { "grad_norm": 0.500389575958252, "learning_rate": 8.61228031274478e-05, "loss": 0.0388, "step": 22470 }, { "grad_norm": 0.5218641757965088, "learning_rate": 8.610850958083431e-05, "loss": 0.0371, "step": 22480 }, { "grad_norm": 0.37748247385025024, "learning_rate": 8.609420986426409e-05, "loss": 0.0335, "step": 22490 }, { "grad_norm": 0.2797091007232666, "learning_rate": 8.60799039801806e-05, "loss": 0.0305, "step": 22500 }, { "grad_norm": 0.30235058069229126, "learning_rate": 8.606559193102828e-05, "loss": 0.025, "step": 22510 }, { "grad_norm": 0.24604450166225433, "learning_rate": 8.605127371925273e-05, "loss": 0.0316, "step": 22520 }, { "grad_norm": 0.31622377038002014, "learning_rate": 8.603694934730047e-05, "loss": 0.0396, "step": 22530 }, { "grad_norm": 0.31658369302749634, "learning_rate": 8.602261881761919e-05, "loss": 0.0279, "step": 22540 }, { "grad_norm": 0.22855153679847717, "learning_rate": 8.600828213265759e-05, "loss": 0.0418, "step": 22550 }, { "grad_norm": 0.33750730752944946, "learning_rate": 8.599393929486539e-05, "loss": 0.047, "step": 22560 }, { "grad_norm": 0.2790418267250061, "learning_rate": 8.59795903066934e-05, "loss": 0.0279, "step": 22570 }, { "grad_norm": 0.25947684049606323, "learning_rate": 8.596523517059347e-05, "loss": 0.0261, "step": 22580 }, { "grad_norm": 0.2502368986606598, "learning_rate": 8.59508738890185e-05, "loss": 0.0255, "step": 22590 }, { "grad_norm": 0.27002808451652527, "learning_rate": 8.593650646442246e-05, "loss": 0.0323, "step": 22600 }, { "grad_norm": 0.383684366941452, "learning_rate": 8.59221328992603e-05, "loss": 0.0284, "step": 22610 }, { "grad_norm": 0.27092015743255615, "learning_rate": 8.590775319598813e-05, "loss": 0.034, "step": 22620 }, { "grad_norm": 0.3194811940193176, "learning_rate": 8.589336735706301e-05, "loss": 0.0285, "step": 22630 }, { "grad_norm": 0.44143790006637573, "learning_rate": 8.587897538494307e-05, "loss": 0.0443, "step": 22640 }, { "grad_norm": 0.35415518283843994, "learning_rate": 8.586457728208756e-05, "loss": 0.0335, "step": 22650 }, { "grad_norm": 0.42064836621284485, "learning_rate": 8.585017305095667e-05, "loss": 0.0362, "step": 22660 }, { "grad_norm": 0.22743190824985504, "learning_rate": 8.583576269401173e-05, "loss": 0.0259, "step": 22670 }, { "grad_norm": 0.3512086570262909, "learning_rate": 8.582134621371504e-05, "loss": 0.0342, "step": 22680 }, { "grad_norm": 0.23107556998729706, "learning_rate": 8.580692361253e-05, "loss": 0.0288, "step": 22690 }, { "grad_norm": 0.38007795810699463, "learning_rate": 8.579249489292104e-05, "loss": 0.0353, "step": 22700 }, { "grad_norm": 0.41691237688064575, "learning_rate": 8.577806005735363e-05, "loss": 0.0477, "step": 22710 }, { "grad_norm": 0.4126426875591278, "learning_rate": 8.576361910829429e-05, "loss": 0.0305, "step": 22720 }, { "grad_norm": 0.30132654309272766, "learning_rate": 8.574917204821057e-05, "loss": 0.0315, "step": 22730 }, { "grad_norm": 0.2156180441379547, "learning_rate": 8.57347188795711e-05, "loss": 0.0299, "step": 22740 }, { "grad_norm": 0.3770400881767273, "learning_rate": 8.572025960484551e-05, "loss": 0.0447, "step": 22750 }, { "grad_norm": 0.475717157125473, "learning_rate": 8.57057942265045e-05, "loss": 0.0427, "step": 22760 }, { "grad_norm": 0.28893139958381653, "learning_rate": 8.569132274701984e-05, "loss": 0.0355, "step": 22770 }, { "grad_norm": 0.2063145935535431, "learning_rate": 8.567684516886427e-05, "loss": 0.0315, "step": 22780 }, { "grad_norm": 0.28482022881507874, "learning_rate": 8.56623614945116e-05, "loss": 0.0298, "step": 22790 }, { "grad_norm": 0.426151841878891, "learning_rate": 8.564787172643675e-05, "loss": 0.0273, "step": 22800 }, { "grad_norm": 0.37784552574157715, "learning_rate": 8.563337586711559e-05, "loss": 0.0372, "step": 22810 }, { "grad_norm": 0.33126863837242126, "learning_rate": 8.561887391902506e-05, "loss": 0.0428, "step": 22820 }, { "grad_norm": 0.3026268780231476, "learning_rate": 8.560436588464316e-05, "loss": 0.0296, "step": 22830 }, { "grad_norm": 0.3166426718235016, "learning_rate": 8.55898517664489e-05, "loss": 0.0436, "step": 22840 }, { "grad_norm": 0.3602542281150818, "learning_rate": 8.557533156692236e-05, "loss": 0.0523, "step": 22850 }, { "grad_norm": 0.41339802742004395, "learning_rate": 8.556080528854467e-05, "loss": 0.0312, "step": 22860 }, { "grad_norm": 0.34961050748825073, "learning_rate": 8.554627293379791e-05, "loss": 0.0314, "step": 22870 }, { "grad_norm": 0.30888956785202026, "learning_rate": 8.553173450516531e-05, "loss": 0.0411, "step": 22880 }, { "grad_norm": 0.32151755690574646, "learning_rate": 8.551719000513108e-05, "loss": 0.0317, "step": 22890 }, { "grad_norm": 0.4449027180671692, "learning_rate": 8.550263943618049e-05, "loss": 0.0497, "step": 22900 }, { "grad_norm": 0.39288491010665894, "learning_rate": 8.54880828007998e-05, "loss": 0.0394, "step": 22910 }, { "grad_norm": 0.3578646183013916, "learning_rate": 8.547352010147637e-05, "loss": 0.0395, "step": 22920 }, { "grad_norm": 0.31036970019340515, "learning_rate": 8.545895134069855e-05, "loss": 0.0319, "step": 22930 }, { "grad_norm": 0.2805081307888031, "learning_rate": 8.544437652095576e-05, "loss": 0.0258, "step": 22940 }, { "grad_norm": 0.37608590722084045, "learning_rate": 8.542979564473843e-05, "loss": 0.0421, "step": 22950 }, { "grad_norm": 0.3202407658100128, "learning_rate": 8.541520871453802e-05, "loss": 0.0269, "step": 22960 }, { "grad_norm": 0.3312983214855194, "learning_rate": 8.540061573284705e-05, "loss": 0.0287, "step": 22970 }, { "grad_norm": 0.32559314370155334, "learning_rate": 8.538601670215906e-05, "loss": 0.0327, "step": 22980 }, { "grad_norm": 0.38230100274086, "learning_rate": 8.537141162496864e-05, "loss": 0.0403, "step": 22990 }, { "grad_norm": 0.322747141122818, "learning_rate": 8.535680050377137e-05, "loss": 0.0339, "step": 23000 }, { "grad_norm": 0.32048776745796204, "learning_rate": 8.534218334106391e-05, "loss": 0.033, "step": 23010 }, { "grad_norm": 0.31751886010169983, "learning_rate": 8.532756013934393e-05, "loss": 0.0323, "step": 23020 }, { "grad_norm": 0.3498108983039856, "learning_rate": 8.531293090111012e-05, "loss": 0.0405, "step": 23030 }, { "grad_norm": 0.3235180079936981, "learning_rate": 8.529829562886225e-05, "loss": 0.0321, "step": 23040 }, { "grad_norm": 0.3277817964553833, "learning_rate": 8.528365432510105e-05, "loss": 0.0441, "step": 23050 }, { "grad_norm": 0.2484062910079956, "learning_rate": 8.526900699232833e-05, "loss": 0.0294, "step": 23060 }, { "grad_norm": 0.264552503824234, "learning_rate": 8.525435363304695e-05, "loss": 0.0249, "step": 23070 }, { "grad_norm": 0.30297815799713135, "learning_rate": 8.523969424976072e-05, "loss": 0.0294, "step": 23080 }, { "grad_norm": 0.27051272988319397, "learning_rate": 8.522502884497457e-05, "loss": 0.0273, "step": 23090 }, { "grad_norm": 0.44019269943237305, "learning_rate": 8.521035742119437e-05, "loss": 0.0385, "step": 23100 }, { "grad_norm": 0.19067072868347168, "learning_rate": 8.519567998092712e-05, "loss": 0.0394, "step": 23110 }, { "grad_norm": 0.23189471662044525, "learning_rate": 8.518099652668075e-05, "loss": 0.0394, "step": 23120 }, { "grad_norm": 0.29955458641052246, "learning_rate": 8.516630706096429e-05, "loss": 0.031, "step": 23130 }, { "grad_norm": 0.31824615597724915, "learning_rate": 8.515161158628773e-05, "loss": 0.0269, "step": 23140 }, { "grad_norm": 0.4505695104598999, "learning_rate": 8.513691010516216e-05, "loss": 0.0491, "step": 23150 }, { "grad_norm": 0.2849963903427124, "learning_rate": 8.512220262009966e-05, "loss": 0.0276, "step": 23160 }, { "grad_norm": 0.31565967202186584, "learning_rate": 8.510748913361332e-05, "loss": 0.0321, "step": 23170 }, { "grad_norm": 0.32275572419166565, "learning_rate": 8.509276964821726e-05, "loss": 0.0271, "step": 23180 }, { "grad_norm": 0.3206689655780792, "learning_rate": 8.507804416642669e-05, "loss": 0.037, "step": 23190 }, { "grad_norm": 0.3510783910751343, "learning_rate": 8.506331269075774e-05, "loss": 0.0305, "step": 23200 }, { "grad_norm": 0.385390043258667, "learning_rate": 8.504857522372765e-05, "loss": 0.0321, "step": 23210 }, { "grad_norm": 0.26752379536628723, "learning_rate": 8.503383176785461e-05, "loss": 0.0296, "step": 23220 }, { "grad_norm": 0.40848785638809204, "learning_rate": 8.501908232565792e-05, "loss": 0.0299, "step": 23230 }, { "grad_norm": 0.3919762074947357, "learning_rate": 8.50043268996578e-05, "loss": 0.0293, "step": 23240 }, { "grad_norm": 0.2570716440677643, "learning_rate": 8.498956549237562e-05, "loss": 0.0287, "step": 23250 }, { "grad_norm": 0.21823027729988098, "learning_rate": 8.497479810633366e-05, "loss": 0.0283, "step": 23260 }, { "grad_norm": 0.41137272119522095, "learning_rate": 8.496002474405525e-05, "loss": 0.0317, "step": 23270 }, { "grad_norm": 0.2905580997467041, "learning_rate": 8.494524540806478e-05, "loss": 0.0351, "step": 23280 }, { "grad_norm": 0.3319880962371826, "learning_rate": 8.493046010088761e-05, "loss": 0.0386, "step": 23290 }, { "grad_norm": 0.44325652718544006, "learning_rate": 8.491566882505018e-05, "loss": 0.0336, "step": 23300 }, { "grad_norm": 0.39822304248809814, "learning_rate": 8.490087158307988e-05, "loss": 0.0408, "step": 23310 }, { "grad_norm": 0.5381467342376709, "learning_rate": 8.488606837750518e-05, "loss": 0.0368, "step": 23320 }, { "grad_norm": 0.38391122221946716, "learning_rate": 8.487125921085552e-05, "loss": 0.0357, "step": 23330 }, { "grad_norm": 0.43273961544036865, "learning_rate": 8.485644408566141e-05, "loss": 0.0385, "step": 23340 }, { "grad_norm": 0.41699934005737305, "learning_rate": 8.484162300445431e-05, "loss": 0.0368, "step": 23350 }, { "grad_norm": 0.47750207781791687, "learning_rate": 8.482679596976676e-05, "loss": 0.03, "step": 23360 }, { "grad_norm": 0.4093071222305298, "learning_rate": 8.48119629841323e-05, "loss": 0.0397, "step": 23370 }, { "grad_norm": 0.5062310695648193, "learning_rate": 8.479712405008547e-05, "loss": 0.0438, "step": 23380 }, { "grad_norm": 0.3194527328014374, "learning_rate": 8.478227917016184e-05, "loss": 0.039, "step": 23390 }, { "grad_norm": 0.3415508568286896, "learning_rate": 8.476742834689801e-05, "loss": 0.0448, "step": 23400 }, { "grad_norm": 0.3813536465167999, "learning_rate": 8.475257158283157e-05, "loss": 0.0459, "step": 23410 }, { "grad_norm": 0.2975156307220459, "learning_rate": 8.473770888050112e-05, "loss": 0.0285, "step": 23420 }, { "grad_norm": 0.47089698910713196, "learning_rate": 8.47228402424463e-05, "loss": 0.0331, "step": 23430 }, { "grad_norm": 0.3680345118045807, "learning_rate": 8.470796567120775e-05, "loss": 0.0301, "step": 23440 }, { "grad_norm": 0.29608914256095886, "learning_rate": 8.469308516932714e-05, "loss": 0.032, "step": 23450 }, { "grad_norm": 0.293899804353714, "learning_rate": 8.467819873934714e-05, "loss": 0.0305, "step": 23460 }, { "grad_norm": 0.29185786843299866, "learning_rate": 8.466330638381143e-05, "loss": 0.0318, "step": 23470 }, { "grad_norm": 0.36522841453552246, "learning_rate": 8.464840810526469e-05, "loss": 0.0276, "step": 23480 }, { "grad_norm": 0.3143647611141205, "learning_rate": 8.463350390625264e-05, "loss": 0.0355, "step": 23490 }, { "grad_norm": 0.24465663731098175, "learning_rate": 8.4618593789322e-05, "loss": 0.0468, "step": 23500 }, { "grad_norm": 0.23128701746463776, "learning_rate": 8.46036777570205e-05, "loss": 0.0324, "step": 23510 }, { "grad_norm": 0.3782735764980316, "learning_rate": 8.458875581189688e-05, "loss": 0.0551, "step": 23520 }, { "grad_norm": 0.3545584976673126, "learning_rate": 8.457382795650092e-05, "loss": 0.0286, "step": 23530 }, { "grad_norm": 0.3633802831172943, "learning_rate": 8.455889419338335e-05, "loss": 0.037, "step": 23540 }, { "grad_norm": 0.3786109685897827, "learning_rate": 8.454395452509593e-05, "loss": 0.0297, "step": 23550 }, { "grad_norm": 0.306057870388031, "learning_rate": 8.452900895419146e-05, "loss": 0.0325, "step": 23560 }, { "grad_norm": 0.3541865646839142, "learning_rate": 8.451405748322376e-05, "loss": 0.0294, "step": 23570 }, { "grad_norm": 0.304828405380249, "learning_rate": 8.449910011474759e-05, "loss": 0.0315, "step": 23580 }, { "grad_norm": 0.2945629954338074, "learning_rate": 8.448413685131876e-05, "loss": 0.0422, "step": 23590 }, { "grad_norm": 0.34258872270584106, "learning_rate": 8.446916769549407e-05, "loss": 0.0395, "step": 23600 }, { "grad_norm": 0.2782374620437622, "learning_rate": 8.445419264983136e-05, "loss": 0.0358, "step": 23610 }, { "grad_norm": 0.28819090127944946, "learning_rate": 8.443921171688947e-05, "loss": 0.0313, "step": 23620 }, { "grad_norm": 0.2693302631378174, "learning_rate": 8.442422489922819e-05, "loss": 0.0266, "step": 23630 }, { "grad_norm": 0.4947042763233185, "learning_rate": 8.440923219940838e-05, "loss": 0.0403, "step": 23640 }, { "grad_norm": 0.3681887984275818, "learning_rate": 8.439423361999189e-05, "loss": 0.0323, "step": 23650 }, { "grad_norm": 0.2933988869190216, "learning_rate": 8.437922916354155e-05, "loss": 0.0302, "step": 23660 }, { "grad_norm": 0.3018237054347992, "learning_rate": 8.436421883262123e-05, "loss": 0.0301, "step": 23670 }, { "grad_norm": 0.20801836252212524, "learning_rate": 8.434920262979577e-05, "loss": 0.0427, "step": 23680 }, { "grad_norm": 0.3249738812446594, "learning_rate": 8.433418055763104e-05, "loss": 0.0265, "step": 23690 }, { "grad_norm": 0.3229618966579437, "learning_rate": 8.431915261869389e-05, "loss": 0.0371, "step": 23700 }, { "grad_norm": 0.29605627059936523, "learning_rate": 8.43041188155522e-05, "loss": 0.0296, "step": 23710 }, { "grad_norm": 0.31491154432296753, "learning_rate": 8.428907915077481e-05, "loss": 0.0326, "step": 23720 }, { "grad_norm": 0.367727667093277, "learning_rate": 8.42740336269316e-05, "loss": 0.0409, "step": 23730 }, { "grad_norm": 0.3096657395362854, "learning_rate": 8.425898224659345e-05, "loss": 0.0272, "step": 23740 }, { "grad_norm": 0.2668198049068451, "learning_rate": 8.42439250123322e-05, "loss": 0.0302, "step": 23750 }, { "grad_norm": 0.2558976709842682, "learning_rate": 8.422886192672076e-05, "loss": 0.031, "step": 23760 }, { "grad_norm": 0.2862635850906372, "learning_rate": 8.421379299233297e-05, "loss": 0.0291, "step": 23770 }, { "grad_norm": 0.3066336512565613, "learning_rate": 8.419871821174371e-05, "loss": 0.03, "step": 23780 }, { "grad_norm": 0.31134533882141113, "learning_rate": 8.418363758752884e-05, "loss": 0.0296, "step": 23790 }, { "grad_norm": 0.2536793351173401, "learning_rate": 8.416855112226523e-05, "loss": 0.031, "step": 23800 }, { "grad_norm": 0.3037722706794739, "learning_rate": 8.415345881853075e-05, "loss": 0.031, "step": 23810 }, { "grad_norm": 0.35520175099372864, "learning_rate": 8.413836067890426e-05, "loss": 0.0287, "step": 23820 }, { "grad_norm": 0.3005349338054657, "learning_rate": 8.41232567059656e-05, "loss": 0.0278, "step": 23830 }, { "grad_norm": 0.3882456123828888, "learning_rate": 8.410814690229565e-05, "loss": 0.0368, "step": 23840 }, { "grad_norm": 0.2385256588459015, "learning_rate": 8.409303127047626e-05, "loss": 0.0266, "step": 23850 }, { "grad_norm": 0.25335389375686646, "learning_rate": 8.407790981309028e-05, "loss": 0.0252, "step": 23860 }, { "grad_norm": 0.3967244029045105, "learning_rate": 8.406278253272153e-05, "loss": 0.0344, "step": 23870 }, { "grad_norm": 0.253665953874588, "learning_rate": 8.404764943195487e-05, "loss": 0.0272, "step": 23880 }, { "grad_norm": 0.3276899456977844, "learning_rate": 8.403251051337613e-05, "loss": 0.0331, "step": 23890 }, { "grad_norm": 0.2931780517101288, "learning_rate": 8.401736577957214e-05, "loss": 0.0254, "step": 23900 }, { "grad_norm": 0.2897636294364929, "learning_rate": 8.40022152331307e-05, "loss": 0.0336, "step": 23910 }, { "grad_norm": 0.26385563611984253, "learning_rate": 8.398705887664064e-05, "loss": 0.0391, "step": 23920 }, { "grad_norm": 0.21301156282424927, "learning_rate": 8.397189671269177e-05, "loss": 0.0257, "step": 23930 }, { "grad_norm": 0.3191172480583191, "learning_rate": 8.395672874387488e-05, "loss": 0.023, "step": 23940 }, { "grad_norm": 0.1847859025001526, "learning_rate": 8.394155497278177e-05, "loss": 0.0258, "step": 23950 }, { "grad_norm": 0.41848546266555786, "learning_rate": 8.392637540200523e-05, "loss": 0.041, "step": 23960 }, { "grad_norm": 0.2761814594268799, "learning_rate": 8.391119003413902e-05, "loss": 0.0331, "step": 23970 }, { "grad_norm": 0.26018571853637695, "learning_rate": 8.38959988717779e-05, "loss": 0.0286, "step": 23980 }, { "grad_norm": 0.2529028058052063, "learning_rate": 8.388080191751764e-05, "loss": 0.0345, "step": 23990 }, { "grad_norm": 0.2921706438064575, "learning_rate": 8.386559917395496e-05, "loss": 0.0284, "step": 24000 }, { "grad_norm": 0.24024495482444763, "learning_rate": 8.385039064368761e-05, "loss": 0.0242, "step": 24010 }, { "grad_norm": 0.27422329783439636, "learning_rate": 8.383517632931431e-05, "loss": 0.0262, "step": 24020 }, { "grad_norm": 0.23514358699321747, "learning_rate": 8.381995623343477e-05, "loss": 0.0277, "step": 24030 }, { "grad_norm": 0.4727059602737427, "learning_rate": 8.380473035864968e-05, "loss": 0.0265, "step": 24040 }, { "grad_norm": 0.21993772685527802, "learning_rate": 8.378949870756076e-05, "loss": 0.0286, "step": 24050 }, { "grad_norm": 0.25632724165916443, "learning_rate": 8.377426128277063e-05, "loss": 0.0332, "step": 24060 }, { "grad_norm": 0.23608899116516113, "learning_rate": 8.375901808688298e-05, "loss": 0.0228, "step": 24070 }, { "grad_norm": 0.28231894969940186, "learning_rate": 8.374376912250246e-05, "loss": 0.0366, "step": 24080 }, { "grad_norm": 0.2389020472764969, "learning_rate": 8.372851439223468e-05, "loss": 0.0245, "step": 24090 }, { "grad_norm": 0.4079488515853882, "learning_rate": 8.371325389868627e-05, "loss": 0.0358, "step": 24100 }, { "grad_norm": 0.257267028093338, "learning_rate": 8.369798764446482e-05, "loss": 0.0309, "step": 24110 }, { "grad_norm": 0.4007214605808258, "learning_rate": 8.368271563217893e-05, "loss": 0.0239, "step": 24120 }, { "grad_norm": 0.3177538216114044, "learning_rate": 8.366743786443817e-05, "loss": 0.0404, "step": 24130 }, { "grad_norm": 0.3474205434322357, "learning_rate": 8.365215434385309e-05, "loss": 0.0403, "step": 24140 }, { "grad_norm": 0.33561083674430847, "learning_rate": 8.36368650730352e-05, "loss": 0.0294, "step": 24150 }, { "grad_norm": 0.20813968777656555, "learning_rate": 8.362157005459705e-05, "loss": 0.0308, "step": 24160 }, { "grad_norm": 0.34428828954696655, "learning_rate": 8.360626929115213e-05, "loss": 0.0268, "step": 24170 }, { "grad_norm": 0.41682887077331543, "learning_rate": 8.359096278531492e-05, "loss": 0.0265, "step": 24180 }, { "grad_norm": 0.2500552237033844, "learning_rate": 8.357565053970088e-05, "loss": 0.0274, "step": 24190 }, { "grad_norm": 0.2101975977420807, "learning_rate": 8.356033255692647e-05, "loss": 0.0256, "step": 24200 }, { "grad_norm": 0.6287606954574585, "learning_rate": 8.354500883960911e-05, "loss": 0.0282, "step": 24210 }, { "grad_norm": 0.3103533685207367, "learning_rate": 8.352967939036717e-05, "loss": 0.0348, "step": 24220 }, { "grad_norm": 0.35001158714294434, "learning_rate": 8.35143442118201e-05, "loss": 0.0336, "step": 24230 }, { "grad_norm": 0.22467242181301117, "learning_rate": 8.349900330658819e-05, "loss": 0.0267, "step": 24240 }, { "grad_norm": 0.2774863839149475, "learning_rate": 8.348365667729284e-05, "loss": 0.0326, "step": 24250 }, { "grad_norm": 0.45763060450553894, "learning_rate": 8.346830432655633e-05, "loss": 0.0275, "step": 24260 }, { "grad_norm": 0.3982834219932556, "learning_rate": 8.345294625700195e-05, "loss": 0.0302, "step": 24270 }, { "grad_norm": 0.3956606686115265, "learning_rate": 8.343758247125402e-05, "loss": 0.03, "step": 24280 }, { "grad_norm": 0.41598841547966003, "learning_rate": 8.342221297193776e-05, "loss": 0.0355, "step": 24290 }, { "grad_norm": 0.30289265513420105, "learning_rate": 8.34068377616794e-05, "loss": 0.0374, "step": 24300 }, { "grad_norm": 0.4080931544303894, "learning_rate": 8.339145684310615e-05, "loss": 0.0366, "step": 24310 }, { "grad_norm": 0.260989785194397, "learning_rate": 8.337607021884618e-05, "loss": 0.0272, "step": 24320 }, { "grad_norm": 0.7599778175354004, "learning_rate": 8.336067789152867e-05, "loss": 0.0247, "step": 24330 }, { "grad_norm": 0.34379199147224426, "learning_rate": 8.334527986378369e-05, "loss": 0.0355, "step": 24340 }, { "grad_norm": 0.4764408767223358, "learning_rate": 8.332987613824239e-05, "loss": 0.0455, "step": 24350 }, { "grad_norm": 0.2826247215270996, "learning_rate": 8.331446671753685e-05, "loss": 0.0309, "step": 24360 }, { "grad_norm": 0.31538063287734985, "learning_rate": 8.329905160430007e-05, "loss": 0.0347, "step": 24370 }, { "grad_norm": 0.31867414712905884, "learning_rate": 8.328363080116611e-05, "loss": 0.0244, "step": 24380 }, { "grad_norm": 0.2687215209007263, "learning_rate": 8.326820431076997e-05, "loss": 0.0254, "step": 24390 }, { "grad_norm": 0.35438552498817444, "learning_rate": 8.325277213574759e-05, "loss": 0.0298, "step": 24400 }, { "grad_norm": 0.25034257769584656, "learning_rate": 8.32373342787359e-05, "loss": 0.0247, "step": 24410 }, { "grad_norm": 0.37020421028137207, "learning_rate": 8.322189074237285e-05, "loss": 0.0314, "step": 24420 }, { "grad_norm": 0.3471143841743469, "learning_rate": 8.32064415292973e-05, "loss": 0.0332, "step": 24430 }, { "grad_norm": 0.281067818403244, "learning_rate": 8.319098664214907e-05, "loss": 0.0249, "step": 24440 }, { "grad_norm": 0.3795962631702423, "learning_rate": 8.3175526083569e-05, "loss": 0.0352, "step": 24450 }, { "grad_norm": 0.2481411099433899, "learning_rate": 8.316005985619889e-05, "loss": 0.0335, "step": 24460 }, { "grad_norm": 0.27919986844062805, "learning_rate": 8.314458796268147e-05, "loss": 0.0356, "step": 24470 }, { "grad_norm": 0.2857441008090973, "learning_rate": 8.312911040566047e-05, "loss": 0.0283, "step": 24480 }, { "grad_norm": 0.3361268937587738, "learning_rate": 8.31136271877806e-05, "loss": 0.0316, "step": 24490 }, { "grad_norm": 0.36973845958709717, "learning_rate": 8.309813831168748e-05, "loss": 0.0342, "step": 24500 }, { "grad_norm": 0.2978322505950928, "learning_rate": 8.308264378002777e-05, "loss": 0.0277, "step": 24510 }, { "grad_norm": 0.26517102122306824, "learning_rate": 8.306714359544906e-05, "loss": 0.0412, "step": 24520 }, { "grad_norm": 0.23595508933067322, "learning_rate": 8.30516377605999e-05, "loss": 0.0257, "step": 24530 }, { "grad_norm": 0.33862775564193726, "learning_rate": 8.30361262781298e-05, "loss": 0.0282, "step": 24540 }, { "grad_norm": 0.35878539085388184, "learning_rate": 8.302060915068924e-05, "loss": 0.0299, "step": 24550 }, { "grad_norm": 0.272518128156662, "learning_rate": 8.300508638092972e-05, "loss": 0.0224, "step": 24560 }, { "grad_norm": 0.23593589663505554, "learning_rate": 8.298955797150361e-05, "loss": 0.0253, "step": 24570 }, { "grad_norm": 0.43207111954689026, "learning_rate": 8.297402392506433e-05, "loss": 0.0311, "step": 24580 }, { "grad_norm": 0.28688451647758484, "learning_rate": 8.295848424426617e-05, "loss": 0.0273, "step": 24590 }, { "grad_norm": 0.2518344223499298, "learning_rate": 8.29429389317645e-05, "loss": 0.0277, "step": 24600 }, { "grad_norm": 0.306623637676239, "learning_rate": 8.292738799021556e-05, "loss": 0.0361, "step": 24610 }, { "grad_norm": 0.23918983340263367, "learning_rate": 8.291183142227656e-05, "loss": 0.025, "step": 24620 }, { "grad_norm": 0.2535260319709778, "learning_rate": 8.289626923060572e-05, "loss": 0.0392, "step": 24630 }, { "grad_norm": 0.28416070342063904, "learning_rate": 8.288070141786218e-05, "loss": 0.0323, "step": 24640 }, { "grad_norm": 0.2351817935705185, "learning_rate": 8.286512798670605e-05, "loss": 0.0265, "step": 24650 }, { "grad_norm": 0.26892584562301636, "learning_rate": 8.284954893979842e-05, "loss": 0.0326, "step": 24660 }, { "grad_norm": 0.22830696403980255, "learning_rate": 8.283396427980131e-05, "loss": 0.0367, "step": 24670 }, { "grad_norm": 0.3047146201133728, "learning_rate": 8.281837400937771e-05, "loss": 0.0304, "step": 24680 }, { "grad_norm": 0.30615851283073425, "learning_rate": 8.28027781311916e-05, "loss": 0.0332, "step": 24690 }, { "grad_norm": 0.365911602973938, "learning_rate": 8.278717664790785e-05, "loss": 0.0378, "step": 24700 }, { "grad_norm": 0.4043672978878021, "learning_rate": 8.277156956219234e-05, "loss": 0.0351, "step": 24710 }, { "grad_norm": 0.34539854526519775, "learning_rate": 8.275595687671189e-05, "loss": 0.031, "step": 24720 }, { "grad_norm": 0.35836297273635864, "learning_rate": 8.27403385941343e-05, "loss": 0.0309, "step": 24730 }, { "grad_norm": 0.44227325916290283, "learning_rate": 8.272471471712828e-05, "loss": 0.0329, "step": 24740 }, { "grad_norm": 0.46984222531318665, "learning_rate": 8.270908524836355e-05, "loss": 0.0349, "step": 24750 }, { "grad_norm": 0.29602426290512085, "learning_rate": 8.269345019051074e-05, "loss": 0.0404, "step": 24760 }, { "grad_norm": 0.2671529948711395, "learning_rate": 8.267780954624147e-05, "loss": 0.0304, "step": 24770 }, { "grad_norm": 0.3304803669452667, "learning_rate": 8.266216331822827e-05, "loss": 0.0338, "step": 24780 }, { "grad_norm": 0.28528499603271484, "learning_rate": 8.264651150914469e-05, "loss": 0.0315, "step": 24790 }, { "grad_norm": 0.2922361493110657, "learning_rate": 8.263085412166517e-05, "loss": 0.0245, "step": 24800 }, { "grad_norm": 0.30740103125572205, "learning_rate": 8.261519115846514e-05, "loss": 0.0319, "step": 24810 }, { "grad_norm": 0.28991174697875977, "learning_rate": 8.259952262222096e-05, "loss": 0.0293, "step": 24820 }, { "grad_norm": 0.3257848620414734, "learning_rate": 8.258384851560997e-05, "loss": 0.0323, "step": 24830 }, { "grad_norm": 0.4769248962402344, "learning_rate": 8.256816884131044e-05, "loss": 0.0324, "step": 24840 }, { "grad_norm": 0.2762592136859894, "learning_rate": 8.255248360200159e-05, "loss": 0.0314, "step": 24850 }, { "grad_norm": 0.27175450325012207, "learning_rate": 8.253679280036359e-05, "loss": 0.0274, "step": 24860 }, { "grad_norm": 0.35176271200180054, "learning_rate": 8.252109643907762e-05, "loss": 0.0433, "step": 24870 }, { "grad_norm": 0.253257155418396, "learning_rate": 8.250539452082569e-05, "loss": 0.0254, "step": 24880 }, { "grad_norm": 0.281649112701416, "learning_rate": 8.248968704829087e-05, "loss": 0.0303, "step": 24890 }, { "grad_norm": 0.3989909291267395, "learning_rate": 8.247397402415714e-05, "loss": 0.0375, "step": 24900 }, { "grad_norm": 0.2655233144760132, "learning_rate": 8.24582554511094e-05, "loss": 0.03, "step": 24910 }, { "grad_norm": 0.2926931381225586, "learning_rate": 8.244253133183355e-05, "loss": 0.0308, "step": 24920 }, { "grad_norm": 0.26964712142944336, "learning_rate": 8.24268016690164e-05, "loss": 0.0336, "step": 24930 }, { "grad_norm": 0.3090846538543701, "learning_rate": 8.241106646534571e-05, "loss": 0.0359, "step": 24940 }, { "grad_norm": 0.4060496389865875, "learning_rate": 8.23953257235102e-05, "loss": 0.0344, "step": 24950 }, { "grad_norm": 0.31644341349601746, "learning_rate": 8.237957944619956e-05, "loss": 0.0319, "step": 24960 }, { "grad_norm": 0.24707677960395813, "learning_rate": 8.236382763610437e-05, "loss": 0.0286, "step": 24970 }, { "grad_norm": 0.27302512526512146, "learning_rate": 8.234807029591619e-05, "loss": 0.0275, "step": 24980 }, { "grad_norm": 0.3140280544757843, "learning_rate": 8.233230742832752e-05, "loss": 0.0416, "step": 24990 }, { "grad_norm": 0.4149726331233978, "learning_rate": 8.231653903603178e-05, "loss": 0.0302, "step": 25000 }, { "grad_norm": 0.2601473927497864, "learning_rate": 8.23007651217234e-05, "loss": 0.0325, "step": 25010 }, { "grad_norm": 0.2650681436061859, "learning_rate": 8.228498568809769e-05, "loss": 0.0338, "step": 25020 }, { "grad_norm": 0.40464091300964355, "learning_rate": 8.22692007378509e-05, "loss": 0.0394, "step": 25030 }, { "grad_norm": 0.26669400930404663, "learning_rate": 8.225341027368028e-05, "loss": 0.0284, "step": 25040 }, { "grad_norm": 0.31558308005332947, "learning_rate": 8.223761429828399e-05, "loss": 0.0353, "step": 25050 }, { "grad_norm": 0.32786718010902405, "learning_rate": 8.22218128143611e-05, "loss": 0.0346, "step": 25060 }, { "grad_norm": 0.3956809639930725, "learning_rate": 8.220600582461166e-05, "loss": 0.0327, "step": 25070 }, { "grad_norm": 0.26564326882362366, "learning_rate": 8.219019333173668e-05, "loss": 0.0343, "step": 25080 }, { "grad_norm": 0.3947524130344391, "learning_rate": 8.217437533843805e-05, "loss": 0.037, "step": 25090 }, { "grad_norm": 0.292910099029541, "learning_rate": 8.215855184741867e-05, "loss": 0.0283, "step": 25100 }, { "grad_norm": 0.309454083442688, "learning_rate": 8.21427228613823e-05, "loss": 0.0303, "step": 25110 }, { "grad_norm": 0.26530763506889343, "learning_rate": 8.21268883830337e-05, "loss": 0.0302, "step": 25120 }, { "grad_norm": 0.32918038964271545, "learning_rate": 8.211104841507855e-05, "loss": 0.0437, "step": 25130 }, { "grad_norm": 0.23456014692783356, "learning_rate": 8.209520296022346e-05, "loss": 0.0335, "step": 25140 }, { "grad_norm": 0.3110148012638092, "learning_rate": 8.207935202117599e-05, "loss": 0.028, "step": 25150 }, { "grad_norm": 0.2626446783542633, "learning_rate": 8.206349560064463e-05, "loss": 0.0291, "step": 25160 }, { "grad_norm": 0.2822067439556122, "learning_rate": 8.204763370133881e-05, "loss": 0.0223, "step": 25170 }, { "grad_norm": 0.2809816896915436, "learning_rate": 8.203176632596892e-05, "loss": 0.0276, "step": 25180 }, { "grad_norm": 0.23694680631160736, "learning_rate": 8.20158934772462e-05, "loss": 0.036, "step": 25190 }, { "grad_norm": 0.2627525329589844, "learning_rate": 8.200001515788294e-05, "loss": 0.0223, "step": 25200 }, { "grad_norm": 0.29775965213775635, "learning_rate": 8.198413137059228e-05, "loss": 0.0311, "step": 25210 }, { "grad_norm": 0.3311699330806732, "learning_rate": 8.196824211808835e-05, "loss": 0.0282, "step": 25220 }, { "grad_norm": 0.3951110243797302, "learning_rate": 8.195234740308617e-05, "loss": 0.0348, "step": 25230 }, { "grad_norm": 0.2456294298171997, "learning_rate": 8.193644722830171e-05, "loss": 0.0254, "step": 25240 }, { "grad_norm": 0.3168751895427704, "learning_rate": 8.19205415964519e-05, "loss": 0.0238, "step": 25250 }, { "grad_norm": 0.2483016848564148, "learning_rate": 8.190463051025456e-05, "loss": 0.0233, "step": 25260 }, { "grad_norm": 0.29700466990470886, "learning_rate": 8.188871397242843e-05, "loss": 0.0275, "step": 25270 }, { "grad_norm": 0.25535598397254944, "learning_rate": 8.187279198569326e-05, "loss": 0.0283, "step": 25280 }, { "grad_norm": 0.31029587984085083, "learning_rate": 8.185686455276966e-05, "loss": 0.0265, "step": 25290 }, { "grad_norm": 0.32745498418807983, "learning_rate": 8.184093167637921e-05, "loss": 0.03, "step": 25300 }, { "grad_norm": 0.40598151087760925, "learning_rate": 8.182499335924437e-05, "loss": 0.0319, "step": 25310 }, { "grad_norm": 0.2549673020839691, "learning_rate": 8.18090496040886e-05, "loss": 0.0225, "step": 25320 }, { "grad_norm": 0.4255838096141815, "learning_rate": 8.179310041363621e-05, "loss": 0.0414, "step": 25330 }, { "grad_norm": 0.34208059310913086, "learning_rate": 8.17771457906125e-05, "loss": 0.027, "step": 25340 }, { "grad_norm": 0.31660550832748413, "learning_rate": 8.176118573774371e-05, "loss": 0.0307, "step": 25350 }, { "grad_norm": 0.3170666992664337, "learning_rate": 8.174522025775692e-05, "loss": 0.0273, "step": 25360 }, { "grad_norm": 0.31963440775871277, "learning_rate": 8.172924935338022e-05, "loss": 0.0329, "step": 25370 }, { "grad_norm": 0.28904151916503906, "learning_rate": 8.171327302734262e-05, "loss": 0.0242, "step": 25380 }, { "grad_norm": 0.2385827898979187, "learning_rate": 8.169729128237401e-05, "loss": 0.0259, "step": 25390 }, { "grad_norm": 0.22543205320835114, "learning_rate": 8.168130412120525e-05, "loss": 0.0244, "step": 25400 }, { "grad_norm": 0.354902058839798, "learning_rate": 8.16653115465681e-05, "loss": 0.0276, "step": 25410 }, { "grad_norm": 0.4567221403121948, "learning_rate": 8.164931356119526e-05, "loss": 0.0283, "step": 25420 }, { "grad_norm": 0.525468647480011, "learning_rate": 8.163331016782032e-05, "loss": 0.0407, "step": 25430 }, { "grad_norm": 0.28615984320640564, "learning_rate": 8.161730136917785e-05, "loss": 0.0287, "step": 25440 }, { "grad_norm": 0.3448057472705841, "learning_rate": 8.160128716800333e-05, "loss": 0.0368, "step": 25450 }, { "grad_norm": 0.29946619272232056, "learning_rate": 8.158526756703313e-05, "loss": 0.0281, "step": 25460 }, { "grad_norm": 0.3505348563194275, "learning_rate": 8.156924256900455e-05, "loss": 0.0343, "step": 25470 }, { "grad_norm": 0.28119611740112305, "learning_rate": 8.155321217665584e-05, "loss": 0.0251, "step": 25480 }, { "grad_norm": 0.2208322435617447, "learning_rate": 8.153717639272614e-05, "loss": 0.0266, "step": 25490 }, { "grad_norm": 0.6128384470939636, "learning_rate": 8.152113521995555e-05, "loss": 0.0337, "step": 25500 }, { "grad_norm": 0.33040380477905273, "learning_rate": 8.150508866108505e-05, "loss": 0.0303, "step": 25510 }, { "grad_norm": 0.34464240074157715, "learning_rate": 8.148903671885657e-05, "loss": 0.0275, "step": 25520 }, { "grad_norm": 0.4508190453052521, "learning_rate": 8.147297939601292e-05, "loss": 0.043, "step": 25530 }, { "grad_norm": 0.4453502297401428, "learning_rate": 8.145691669529792e-05, "loss": 0.0328, "step": 25540 }, { "grad_norm": 0.29864147305488586, "learning_rate": 8.144084861945618e-05, "loss": 0.0283, "step": 25550 }, { "grad_norm": 0.34688514471054077, "learning_rate": 8.142477517123333e-05, "loss": 0.0328, "step": 25560 }, { "grad_norm": 0.29271426796913147, "learning_rate": 8.140869635337586e-05, "loss": 0.0296, "step": 25570 }, { "grad_norm": 0.2557660639286041, "learning_rate": 8.139261216863123e-05, "loss": 0.0273, "step": 25580 }, { "grad_norm": 0.41783177852630615, "learning_rate": 8.137652261974776e-05, "loss": 0.0366, "step": 25590 }, { "grad_norm": 0.49546054005622864, "learning_rate": 8.136042770947472e-05, "loss": 0.0373, "step": 25600 }, { "grad_norm": 0.32520878314971924, "learning_rate": 8.134432744056228e-05, "loss": 0.0375, "step": 25610 }, { "grad_norm": 0.2938700318336487, "learning_rate": 8.132822181576158e-05, "loss": 0.036, "step": 25620 }, { "grad_norm": 0.4217250943183899, "learning_rate": 8.131211083782459e-05, "loss": 0.0338, "step": 25630 }, { "grad_norm": 0.25303202867507935, "learning_rate": 8.129599450950424e-05, "loss": 0.0259, "step": 25640 }, { "grad_norm": 0.3204381763935089, "learning_rate": 8.127987283355438e-05, "loss": 0.0332, "step": 25650 }, { "grad_norm": 0.2792738080024719, "learning_rate": 8.126374581272976e-05, "loss": 0.0222, "step": 25660 }, { "grad_norm": 0.4689328372478485, "learning_rate": 8.124761344978605e-05, "loss": 0.0441, "step": 25670 }, { "grad_norm": 0.3383304476737976, "learning_rate": 8.12314757474798e-05, "loss": 0.0262, "step": 25680 }, { "grad_norm": 0.20869357883930206, "learning_rate": 8.121533270856856e-05, "loss": 0.0291, "step": 25690 }, { "grad_norm": 0.45083630084991455, "learning_rate": 8.119918433581069e-05, "loss": 0.0275, "step": 25700 }, { "grad_norm": 0.3905434012413025, "learning_rate": 8.118303063196551e-05, "loss": 0.0285, "step": 25710 }, { "grad_norm": 0.3430732488632202, "learning_rate": 8.116687159979326e-05, "loss": 0.0227, "step": 25720 }, { "grad_norm": 0.3626424968242645, "learning_rate": 8.115070724205508e-05, "loss": 0.0359, "step": 25730 }, { "grad_norm": 0.34273120760917664, "learning_rate": 8.113453756151296e-05, "loss": 0.028, "step": 25740 }, { "grad_norm": 0.3228514790534973, "learning_rate": 8.111836256092995e-05, "loss": 0.0261, "step": 25750 }, { "grad_norm": 0.26422247290611267, "learning_rate": 8.110218224306985e-05, "loss": 0.0307, "step": 25760 }, { "grad_norm": 0.2690529227256775, "learning_rate": 8.108599661069745e-05, "loss": 0.0272, "step": 25770 }, { "grad_norm": 0.2695307433605194, "learning_rate": 8.106980566657845e-05, "loss": 0.0229, "step": 25780 }, { "grad_norm": 0.3657267391681671, "learning_rate": 8.10536094134794e-05, "loss": 0.0292, "step": 25790 }, { "grad_norm": 0.29315808415412903, "learning_rate": 8.103740785416783e-05, "loss": 0.0265, "step": 25800 }, { "grad_norm": 0.227586030960083, "learning_rate": 8.102120099141212e-05, "loss": 0.0253, "step": 25810 }, { "grad_norm": 0.17644034326076508, "learning_rate": 8.100498882798163e-05, "loss": 0.0259, "step": 25820 }, { "grad_norm": 0.27496564388275146, "learning_rate": 8.09887713666465e-05, "loss": 0.024, "step": 25830 }, { "grad_norm": 0.334898442029953, "learning_rate": 8.09725486101779e-05, "loss": 0.0335, "step": 25840 }, { "grad_norm": 0.27452802658081055, "learning_rate": 8.095632056134784e-05, "loss": 0.0306, "step": 25850 }, { "grad_norm": 0.2920527458190918, "learning_rate": 8.094008722292925e-05, "loss": 0.0269, "step": 25860 }, { "grad_norm": 0.26508212089538574, "learning_rate": 8.092384859769598e-05, "loss": 0.0288, "step": 25870 }, { "grad_norm": 0.3460388481616974, "learning_rate": 8.090760468842275e-05, "loss": 0.0511, "step": 25880 }, { "grad_norm": 0.36612412333488464, "learning_rate": 8.089135549788521e-05, "loss": 0.0279, "step": 25890 }, { "grad_norm": 0.20299582183361053, "learning_rate": 8.087510102885987e-05, "loss": 0.0244, "step": 25900 }, { "grad_norm": 0.33370739221572876, "learning_rate": 8.085884128412422e-05, "loss": 0.0343, "step": 25910 }, { "grad_norm": 0.3277893364429474, "learning_rate": 8.084257626645659e-05, "loss": 0.0288, "step": 25920 }, { "grad_norm": 0.3557719886302948, "learning_rate": 8.08263059786362e-05, "loss": 0.0268, "step": 25930 }, { "grad_norm": 0.25144460797309875, "learning_rate": 8.081003042344325e-05, "loss": 0.0213, "step": 25940 }, { "grad_norm": 0.32224103808403015, "learning_rate": 8.079374960365872e-05, "loss": 0.028, "step": 25950 }, { "grad_norm": 0.2195955067873001, "learning_rate": 8.077746352206463e-05, "loss": 0.0338, "step": 25960 }, { "grad_norm": 0.38477998971939087, "learning_rate": 8.076117218144377e-05, "loss": 0.0301, "step": 25970 }, { "grad_norm": 0.30632132291793823, "learning_rate": 8.074487558457991e-05, "loss": 0.027, "step": 25980 }, { "grad_norm": 0.34895461797714233, "learning_rate": 8.072857373425768e-05, "loss": 0.0279, "step": 25990 }, { "grad_norm": 0.40219372510910034, "learning_rate": 8.071226663326264e-05, "loss": 0.0293, "step": 26000 }, { "grad_norm": 0.2784111797809601, "learning_rate": 8.069595428438121e-05, "loss": 0.0343, "step": 26010 }, { "grad_norm": 0.4521828293800354, "learning_rate": 8.067963669040072e-05, "loss": 0.0269, "step": 26020 }, { "grad_norm": 0.4714627265930176, "learning_rate": 8.066331385410942e-05, "loss": 0.0305, "step": 26030 }, { "grad_norm": 0.3428077697753906, "learning_rate": 8.064698577829641e-05, "loss": 0.0333, "step": 26040 }, { "grad_norm": 0.35056254267692566, "learning_rate": 8.063065246575175e-05, "loss": 0.0348, "step": 26050 }, { "grad_norm": 0.27640387415885925, "learning_rate": 8.061431391926631e-05, "loss": 0.0256, "step": 26060 }, { "grad_norm": 0.2806272506713867, "learning_rate": 8.059797014163195e-05, "loss": 0.0374, "step": 26070 }, { "grad_norm": 0.5691486597061157, "learning_rate": 8.058162113564133e-05, "loss": 0.0421, "step": 26080 }, { "grad_norm": 0.35155561566352844, "learning_rate": 8.056526690408806e-05, "loss": 0.0389, "step": 26090 }, { "grad_norm": 0.3011760413646698, "learning_rate": 8.054890744976666e-05, "loss": 0.0355, "step": 26100 }, { "grad_norm": 0.25476402044296265, "learning_rate": 8.053254277547248e-05, "loss": 0.0279, "step": 26110 }, { "grad_norm": 0.34008297324180603, "learning_rate": 8.051617288400182e-05, "loss": 0.0311, "step": 26120 }, { "grad_norm": 0.33687883615493774, "learning_rate": 8.049979777815182e-05, "loss": 0.0286, "step": 26130 }, { "grad_norm": 0.21254561841487885, "learning_rate": 8.048341746072054e-05, "loss": 0.0299, "step": 26140 }, { "grad_norm": 0.3291355073451996, "learning_rate": 8.046703193450696e-05, "loss": 0.0368, "step": 26150 }, { "grad_norm": 0.26198968291282654, "learning_rate": 8.04506412023109e-05, "loss": 0.0267, "step": 26160 }, { "grad_norm": 0.27377477288246155, "learning_rate": 8.043424526693306e-05, "loss": 0.0278, "step": 26170 }, { "grad_norm": 0.2486613392829895, "learning_rate": 8.04178441311751e-05, "loss": 0.0287, "step": 26180 }, { "grad_norm": 0.42375731468200684, "learning_rate": 8.04014377978395e-05, "loss": 0.0306, "step": 26190 }, { "grad_norm": 0.6117240190505981, "learning_rate": 8.038502626972967e-05, "loss": 0.0311, "step": 26200 }, { "grad_norm": 0.2583341598510742, "learning_rate": 8.036860954964989e-05, "loss": 0.0313, "step": 26210 }, { "grad_norm": 0.29986926913261414, "learning_rate": 8.035218764040531e-05, "loss": 0.0333, "step": 26220 }, { "grad_norm": 0.3710647523403168, "learning_rate": 8.033576054480199e-05, "loss": 0.033, "step": 26230 }, { "grad_norm": 0.27624067664146423, "learning_rate": 8.031932826564688e-05, "loss": 0.0271, "step": 26240 }, { "grad_norm": 0.30044451355934143, "learning_rate": 8.030289080574782e-05, "loss": 0.0304, "step": 26250 }, { "grad_norm": 0.3310811221599579, "learning_rate": 8.028644816791349e-05, "loss": 0.0379, "step": 26260 }, { "grad_norm": 0.2978135049343109, "learning_rate": 8.027000035495351e-05, "loss": 0.0281, "step": 26270 }, { "grad_norm": 0.3032517731189728, "learning_rate": 8.025354736967836e-05, "loss": 0.0408, "step": 26280 }, { "grad_norm": 0.2321745753288269, "learning_rate": 8.023708921489941e-05, "loss": 0.0459, "step": 26290 }, { "grad_norm": 0.3728480041027069, "learning_rate": 8.022062589342887e-05, "loss": 0.0432, "step": 26300 }, { "grad_norm": 0.38252830505371094, "learning_rate": 8.020415740807993e-05, "loss": 0.0475, "step": 26310 }, { "grad_norm": 0.265424907207489, "learning_rate": 8.018768376166656e-05, "loss": 0.0299, "step": 26320 }, { "grad_norm": 0.3514787554740906, "learning_rate": 8.017120495700368e-05, "loss": 0.0306, "step": 26330 }, { "grad_norm": 0.3367747664451599, "learning_rate": 8.015472099690704e-05, "loss": 0.0304, "step": 26340 }, { "grad_norm": 0.33868905901908875, "learning_rate": 8.013823188419332e-05, "loss": 0.0304, "step": 26350 }, { "grad_norm": 0.31848835945129395, "learning_rate": 8.012173762168006e-05, "loss": 0.038, "step": 26360 }, { "grad_norm": 0.27359938621520996, "learning_rate": 8.010523821218567e-05, "loss": 0.0286, "step": 26370 }, { "grad_norm": 0.4159870445728302, "learning_rate": 8.008873365852945e-05, "loss": 0.0306, "step": 26380 }, { "grad_norm": 0.3237883746623993, "learning_rate": 8.007222396353157e-05, "loss": 0.0351, "step": 26390 }, { "grad_norm": 0.3278841972351074, "learning_rate": 8.00557091300131e-05, "loss": 0.0308, "step": 26400 }, { "grad_norm": 0.24398286640644073, "learning_rate": 8.003918916079597e-05, "loss": 0.0278, "step": 26410 }, { "grad_norm": 0.2830345034599304, "learning_rate": 8.002266405870298e-05, "loss": 0.0246, "step": 26420 }, { "grad_norm": 0.39161479473114014, "learning_rate": 8.000613382655782e-05, "loss": 0.0351, "step": 26430 }, { "grad_norm": 0.2690717279911041, "learning_rate": 7.998959846718505e-05, "loss": 0.0292, "step": 26440 }, { "grad_norm": 0.2551496624946594, "learning_rate": 7.997305798341012e-05, "loss": 0.0306, "step": 26450 }, { "grad_norm": 0.2711044251918793, "learning_rate": 7.995651237805937e-05, "loss": 0.0317, "step": 26460 }, { "grad_norm": 0.35740819573402405, "learning_rate": 7.993996165395996e-05, "loss": 0.0247, "step": 26470 }, { "grad_norm": 0.354999303817749, "learning_rate": 7.992340581393996e-05, "loss": 0.025, "step": 26480 }, { "grad_norm": 0.28595679998397827, "learning_rate": 7.990684486082831e-05, "loss": 0.031, "step": 26490 }, { "grad_norm": 0.3167290687561035, "learning_rate": 7.989027879745482e-05, "loss": 0.034, "step": 26500 }, { "grad_norm": 0.307781457901001, "learning_rate": 7.98737076266502e-05, "loss": 0.0431, "step": 26510 }, { "grad_norm": 0.43988698720932007, "learning_rate": 7.985713135124598e-05, "loss": 0.0334, "step": 26520 }, { "grad_norm": 0.5122446417808533, "learning_rate": 7.98405499740746e-05, "loss": 0.0355, "step": 26530 }, { "grad_norm": 0.33392074704170227, "learning_rate": 7.98239634979694e-05, "loss": 0.0344, "step": 26540 }, { "grad_norm": 0.3678765892982483, "learning_rate": 7.98073719257645e-05, "loss": 0.0321, "step": 26550 }, { "grad_norm": 0.3801378011703491, "learning_rate": 7.979077526029499e-05, "loss": 0.0356, "step": 26560 }, { "grad_norm": 0.22273042798042297, "learning_rate": 7.977417350439675e-05, "loss": 0.0258, "step": 26570 }, { "grad_norm": 0.3321053385734558, "learning_rate": 7.97575666609066e-05, "loss": 0.0252, "step": 26580 }, { "grad_norm": 0.403368204832077, "learning_rate": 7.974095473266216e-05, "loss": 0.0292, "step": 26590 }, { "grad_norm": 0.43095076084136963, "learning_rate": 7.972433772250198e-05, "loss": 0.0308, "step": 26600 }, { "grad_norm": 0.23526251316070557, "learning_rate": 7.970771563326544e-05, "loss": 0.0267, "step": 26610 }, { "grad_norm": 0.31215783953666687, "learning_rate": 7.96910884677928e-05, "loss": 0.0334, "step": 26620 }, { "grad_norm": 0.2979176938533783, "learning_rate": 7.967445622892523e-05, "loss": 0.0312, "step": 26630 }, { "grad_norm": 0.3898460566997528, "learning_rate": 7.965781891950465e-05, "loss": 0.0275, "step": 26640 }, { "grad_norm": 0.2192760556936264, "learning_rate": 7.964117654237397e-05, "loss": 0.0207, "step": 26650 }, { "grad_norm": 0.3497137129306793, "learning_rate": 7.962452910037692e-05, "loss": 0.0357, "step": 26660 }, { "grad_norm": 0.3203907608985901, "learning_rate": 7.96078765963581e-05, "loss": 0.0242, "step": 26670 }, { "grad_norm": 0.3461344540119171, "learning_rate": 7.95912190331629e-05, "loss": 0.0247, "step": 26680 }, { "grad_norm": 0.29029735922813416, "learning_rate": 7.957455641363772e-05, "loss": 0.0265, "step": 26690 }, { "grad_norm": 0.36052456498146057, "learning_rate": 7.955788874062968e-05, "loss": 0.0388, "step": 26700 }, { "grad_norm": 0.4958287179470062, "learning_rate": 7.95412160169869e-05, "loss": 0.0316, "step": 26710 }, { "grad_norm": 0.30912014842033386, "learning_rate": 7.952453824555824e-05, "loss": 0.0238, "step": 26720 }, { "grad_norm": 0.2732723355293274, "learning_rate": 7.95078554291935e-05, "loss": 0.023, "step": 26730 }, { "grad_norm": 0.2952338755130768, "learning_rate": 7.94911675707433e-05, "loss": 0.0262, "step": 26740 }, { "grad_norm": 0.30680814385414124, "learning_rate": 7.947447467305915e-05, "loss": 0.0341, "step": 26750 }, { "grad_norm": 0.28267374634742737, "learning_rate": 7.94577767389934e-05, "loss": 0.0277, "step": 26760 }, { "grad_norm": 0.2985621988773346, "learning_rate": 7.944107377139928e-05, "loss": 0.0216, "step": 26770 }, { "grad_norm": 0.26817843317985535, "learning_rate": 7.942436577313088e-05, "loss": 0.0295, "step": 26780 }, { "grad_norm": 0.24281474947929382, "learning_rate": 7.940765274704312e-05, "loss": 0.034, "step": 26790 }, { "grad_norm": 0.3596895635128021, "learning_rate": 7.939093469599181e-05, "loss": 0.027, "step": 26800 }, { "grad_norm": 0.33885055780410767, "learning_rate": 7.93742116228336e-05, "loss": 0.0241, "step": 26810 }, { "grad_norm": 0.24488602578639984, "learning_rate": 7.935748353042602e-05, "loss": 0.029, "step": 26820 }, { "grad_norm": 0.3323041796684265, "learning_rate": 7.934075042162744e-05, "loss": 0.0272, "step": 26830 }, { "grad_norm": 0.31195592880249023, "learning_rate": 7.932401229929705e-05, "loss": 0.0335, "step": 26840 }, { "grad_norm": 0.38170522451400757, "learning_rate": 7.9307269166295e-05, "loss": 0.0306, "step": 26850 }, { "grad_norm": 0.256427139043808, "learning_rate": 7.92905210254822e-05, "loss": 0.03, "step": 26860 }, { "grad_norm": 0.4059790372848511, "learning_rate": 7.927376787972045e-05, "loss": 0.0251, "step": 26870 }, { "grad_norm": 0.34516847133636475, "learning_rate": 7.92570097318724e-05, "loss": 0.0333, "step": 26880 }, { "grad_norm": 0.31435173749923706, "learning_rate": 7.924024658480158e-05, "loss": 0.031, "step": 26890 }, { "grad_norm": 0.2401813417673111, "learning_rate": 7.922347844137233e-05, "loss": 0.0287, "step": 26900 }, { "grad_norm": 0.25053903460502625, "learning_rate": 7.92067053044499e-05, "loss": 0.0359, "step": 26910 }, { "grad_norm": 0.49798405170440674, "learning_rate": 7.918992717690031e-05, "loss": 0.0387, "step": 26920 }, { "grad_norm": 0.36646246910095215, "learning_rate": 7.917314406159053e-05, "loss": 0.0295, "step": 26930 }, { "grad_norm": 0.48658764362335205, "learning_rate": 7.915635596138832e-05, "loss": 0.042, "step": 26940 }, { "grad_norm": 0.2872348427772522, "learning_rate": 7.913956287916228e-05, "loss": 0.0346, "step": 26950 }, { "grad_norm": 0.2166556715965271, "learning_rate": 7.912276481778193e-05, "loss": 0.0259, "step": 26960 }, { "grad_norm": 0.2591947317123413, "learning_rate": 7.910596178011759e-05, "loss": 0.037, "step": 26970 }, { "grad_norm": 0.2245386242866516, "learning_rate": 7.908915376904043e-05, "loss": 0.0214, "step": 26980 }, { "grad_norm": 0.2528972327709198, "learning_rate": 7.907234078742247e-05, "loss": 0.0336, "step": 26990 }, { "grad_norm": 0.40518251061439514, "learning_rate": 7.90555228381366e-05, "loss": 0.0254, "step": 27000 }, { "grad_norm": 0.32068678736686707, "learning_rate": 7.903869992405656e-05, "loss": 0.0239, "step": 27010 }, { "grad_norm": 0.290192574262619, "learning_rate": 7.902187204805691e-05, "loss": 0.0281, "step": 27020 }, { "grad_norm": 0.29801344871520996, "learning_rate": 7.900503921301308e-05, "loss": 0.023, "step": 27030 }, { "grad_norm": 0.2355516403913498, "learning_rate": 7.898820142180133e-05, "loss": 0.0383, "step": 27040 }, { "grad_norm": 0.319900244474411, "learning_rate": 7.897135867729879e-05, "loss": 0.0235, "step": 27050 }, { "grad_norm": 0.2885724902153015, "learning_rate": 7.89545109823834e-05, "loss": 0.0217, "step": 27060 }, { "grad_norm": 0.2451425939798355, "learning_rate": 7.8937658339934e-05, "loss": 0.024, "step": 27070 }, { "grad_norm": 0.23467035591602325, "learning_rate": 7.892080075283026e-05, "loss": 0.0281, "step": 27080 }, { "grad_norm": 0.27179133892059326, "learning_rate": 7.890393822395263e-05, "loss": 0.0234, "step": 27090 }, { "grad_norm": 0.4600229263305664, "learning_rate": 7.88870707561825e-05, "loss": 0.0218, "step": 27100 }, { "grad_norm": 0.39577528834342957, "learning_rate": 7.887019835240203e-05, "loss": 0.026, "step": 27110 }, { "grad_norm": 0.25275856256484985, "learning_rate": 7.885332101549427e-05, "loss": 0.0295, "step": 27120 }, { "grad_norm": 0.2607317268848419, "learning_rate": 7.883643874834308e-05, "loss": 0.0224, "step": 27130 }, { "grad_norm": 0.2583790719509125, "learning_rate": 7.881955155383321e-05, "loss": 0.0383, "step": 27140 }, { "grad_norm": 0.42935463786125183, "learning_rate": 7.880265943485017e-05, "loss": 0.029, "step": 27150 }, { "grad_norm": 0.2801440954208374, "learning_rate": 7.878576239428038e-05, "loss": 0.0285, "step": 27160 }, { "grad_norm": 0.37567490339279175, "learning_rate": 7.87688604350111e-05, "loss": 0.0369, "step": 27170 }, { "grad_norm": 0.5006158351898193, "learning_rate": 7.875195355993042e-05, "loss": 0.0367, "step": 27180 }, { "grad_norm": 0.2735961079597473, "learning_rate": 7.873504177192724e-05, "loss": 0.0338, "step": 27190 }, { "grad_norm": 0.4650057554244995, "learning_rate": 7.87181250738913e-05, "loss": 0.0439, "step": 27200 }, { "grad_norm": 0.1636897474527359, "learning_rate": 7.870120346871324e-05, "loss": 0.0267, "step": 27210 }, { "grad_norm": 0.2882481515407562, "learning_rate": 7.86842769592845e-05, "loss": 0.0238, "step": 27220 }, { "grad_norm": 0.32514333724975586, "learning_rate": 7.866734554849732e-05, "loss": 0.0255, "step": 27230 }, { "grad_norm": 0.31800469756126404, "learning_rate": 7.865040923924486e-05, "loss": 0.0229, "step": 27240 }, { "grad_norm": 0.4111006557941437, "learning_rate": 7.863346803442104e-05, "loss": 0.0365, "step": 27250 }, { "grad_norm": 0.3336755931377411, "learning_rate": 7.861652193692067e-05, "loss": 0.0245, "step": 27260 }, { "grad_norm": 0.28721553087234497, "learning_rate": 7.859957094963937e-05, "loss": 0.0182, "step": 27270 }, { "grad_norm": 0.353935182094574, "learning_rate": 7.858261507547357e-05, "loss": 0.0269, "step": 27280 }, { "grad_norm": 0.28817254304885864, "learning_rate": 7.856565431732061e-05, "loss": 0.0206, "step": 27290 }, { "grad_norm": 0.1901863068342209, "learning_rate": 7.854868867807859e-05, "loss": 0.0189, "step": 27300 }, { "grad_norm": 0.3489702641963959, "learning_rate": 7.85317181606465e-05, "loss": 0.0275, "step": 27310 }, { "grad_norm": 0.31449034810066223, "learning_rate": 7.85147427679241e-05, "loss": 0.0244, "step": 27320 }, { "grad_norm": 0.30818814039230347, "learning_rate": 7.849776250281205e-05, "loss": 0.0174, "step": 27330 }, { "grad_norm": 0.38917240500450134, "learning_rate": 7.84807773682118e-05, "loss": 0.0236, "step": 27340 }, { "grad_norm": 0.27426984906196594, "learning_rate": 7.846378736702565e-05, "loss": 0.0313, "step": 27350 }, { "grad_norm": 0.28185251355171204, "learning_rate": 7.844679250215671e-05, "loss": 0.0313, "step": 27360 }, { "grad_norm": 0.26396098732948303, "learning_rate": 7.842979277650898e-05, "loss": 0.0256, "step": 27370 }, { "grad_norm": 0.30680322647094727, "learning_rate": 7.84127881929872e-05, "loss": 0.0241, "step": 27380 }, { "grad_norm": 0.29665014147758484, "learning_rate": 7.839577875449704e-05, "loss": 0.0255, "step": 27390 }, { "grad_norm": 0.46697375178337097, "learning_rate": 7.837876446394489e-05, "loss": 0.0211, "step": 27400 }, { "grad_norm": 0.22070752084255219, "learning_rate": 7.836174532423805e-05, "loss": 0.0308, "step": 27410 }, { "grad_norm": 0.3121679425239563, "learning_rate": 7.834472133828466e-05, "loss": 0.0276, "step": 27420 }, { "grad_norm": 0.38837170600891113, "learning_rate": 7.832769250899359e-05, "loss": 0.0292, "step": 27430 }, { "grad_norm": 0.3284570276737213, "learning_rate": 7.831065883927464e-05, "loss": 0.0386, "step": 27440 }, { "grad_norm": 0.331633061170578, "learning_rate": 7.829362033203841e-05, "loss": 0.0312, "step": 27450 }, { "grad_norm": 0.2932037115097046, "learning_rate": 7.827657699019628e-05, "loss": 0.026, "step": 27460 }, { "grad_norm": 0.32532989978790283, "learning_rate": 7.825952881666052e-05, "loss": 0.0372, "step": 27470 }, { "grad_norm": 0.3599981665611267, "learning_rate": 7.824247581434418e-05, "loss": 0.0295, "step": 27480 }, { "grad_norm": 0.33944883942604065, "learning_rate": 7.822541798616116e-05, "loss": 0.0298, "step": 27490 }, { "grad_norm": 0.3423537313938141, "learning_rate": 7.820835533502617e-05, "loss": 0.0356, "step": 27500 }, { "grad_norm": 0.4138418436050415, "learning_rate": 7.819128786385475e-05, "loss": 0.0406, "step": 27510 }, { "grad_norm": 0.350673109292984, "learning_rate": 7.817421557556329e-05, "loss": 0.0383, "step": 27520 }, { "grad_norm": 0.3890722692012787, "learning_rate": 7.815713847306893e-05, "loss": 0.0327, "step": 27530 }, { "grad_norm": 0.24211376905441284, "learning_rate": 7.81400565592897e-05, "loss": 0.0281, "step": 27540 }, { "grad_norm": 0.369367390871048, "learning_rate": 7.812296983714444e-05, "loss": 0.0359, "step": 27550 }, { "grad_norm": 0.2013167142868042, "learning_rate": 7.810587830955281e-05, "loss": 0.0233, "step": 27560 }, { "grad_norm": 0.2940516471862793, "learning_rate": 7.808878197943528e-05, "loss": 0.0292, "step": 27570 }, { "grad_norm": 0.2761489748954773, "learning_rate": 7.807168084971312e-05, "loss": 0.0253, "step": 27580 }, { "grad_norm": 0.2766340970993042, "learning_rate": 7.805457492330849e-05, "loss": 0.0234, "step": 27590 }, { "grad_norm": 0.3208357095718384, "learning_rate": 7.803746420314428e-05, "loss": 0.0293, "step": 27600 }, { "grad_norm": 0.33168190717697144, "learning_rate": 7.802034869214428e-05, "loss": 0.0364, "step": 27610 }, { "grad_norm": 0.2759625017642975, "learning_rate": 7.800322839323303e-05, "loss": 0.0258, "step": 27620 }, { "grad_norm": 0.24267378449440002, "learning_rate": 7.798610330933593e-05, "loss": 0.0244, "step": 27630 }, { "grad_norm": 0.25233447551727295, "learning_rate": 7.796897344337922e-05, "loss": 0.0274, "step": 27640 }, { "grad_norm": 0.2713410258293152, "learning_rate": 7.795183879828989e-05, "loss": 0.0374, "step": 27650 }, { "grad_norm": 0.2547476291656494, "learning_rate": 7.793469937699579e-05, "loss": 0.0378, "step": 27660 }, { "grad_norm": 0.3007572591304779, "learning_rate": 7.791755518242558e-05, "loss": 0.0377, "step": 27670 }, { "grad_norm": 0.29760876297950745, "learning_rate": 7.790040621750876e-05, "loss": 0.0332, "step": 27680 }, { "grad_norm": 0.24404868483543396, "learning_rate": 7.788325248517558e-05, "loss": 0.0219, "step": 27690 }, { "grad_norm": 0.35906293988227844, "learning_rate": 7.786609398835715e-05, "loss": 0.0314, "step": 27700 }, { "grad_norm": 0.25251445174217224, "learning_rate": 7.784893072998541e-05, "loss": 0.0229, "step": 27710 }, { "grad_norm": 0.24331417679786682, "learning_rate": 7.783176271299306e-05, "loss": 0.0251, "step": 27720 }, { "grad_norm": 0.2485295981168747, "learning_rate": 7.781458994031368e-05, "loss": 0.0217, "step": 27730 }, { "grad_norm": 0.2521878182888031, "learning_rate": 7.779741241488161e-05, "loss": 0.026, "step": 27740 }, { "grad_norm": 0.3439041078090668, "learning_rate": 7.7780230139632e-05, "loss": 0.0276, "step": 27750 }, { "grad_norm": 0.3750625550746918, "learning_rate": 7.776304311750087e-05, "loss": 0.0256, "step": 27760 }, { "grad_norm": 0.24659080803394318, "learning_rate": 7.7745851351425e-05, "loss": 0.041, "step": 27770 }, { "grad_norm": 0.2850901782512665, "learning_rate": 7.772865484434197e-05, "loss": 0.0281, "step": 27780 }, { "grad_norm": 0.34831592440605164, "learning_rate": 7.77114535991902e-05, "loss": 0.0327, "step": 27790 }, { "grad_norm": 0.31378141045570374, "learning_rate": 7.769424761890893e-05, "loss": 0.0264, "step": 27800 }, { "grad_norm": 0.36389362812042236, "learning_rate": 7.767703690643817e-05, "loss": 0.0321, "step": 27810 }, { "grad_norm": 0.2192651331424713, "learning_rate": 7.76598214647188e-05, "loss": 0.021, "step": 27820 }, { "grad_norm": 0.27528682351112366, "learning_rate": 7.764260129669241e-05, "loss": 0.025, "step": 27830 }, { "grad_norm": 0.25497379899024963, "learning_rate": 7.76253764053015e-05, "loss": 0.0296, "step": 27840 }, { "grad_norm": 0.19373993575572968, "learning_rate": 7.760814679348932e-05, "loss": 0.0199, "step": 27850 }, { "grad_norm": 0.26570552587509155, "learning_rate": 7.759091246419992e-05, "loss": 0.0246, "step": 27860 }, { "grad_norm": 0.28045013546943665, "learning_rate": 7.75736734203782e-05, "loss": 0.0237, "step": 27870 }, { "grad_norm": 0.4410416781902313, "learning_rate": 7.755642966496985e-05, "loss": 0.0408, "step": 27880 }, { "grad_norm": 0.43326687812805176, "learning_rate": 7.753918120092132e-05, "loss": 0.0328, "step": 27890 }, { "grad_norm": 0.3445884585380554, "learning_rate": 7.752192803117993e-05, "loss": 0.0279, "step": 27900 }, { "grad_norm": 0.2702781856060028, "learning_rate": 7.750467015869377e-05, "loss": 0.0278, "step": 27910 }, { "grad_norm": 0.24008622765541077, "learning_rate": 7.748740758641174e-05, "loss": 0.0286, "step": 27920 }, { "grad_norm": 0.2834489047527313, "learning_rate": 7.74701403172835e-05, "loss": 0.0315, "step": 27930 }, { "grad_norm": 0.2587246596813202, "learning_rate": 7.745286835425962e-05, "loss": 0.0339, "step": 27940 }, { "grad_norm": 0.367698609828949, "learning_rate": 7.743559170029138e-05, "loss": 0.0297, "step": 27950 }, { "grad_norm": 0.36180388927459717, "learning_rate": 7.741831035833087e-05, "loss": 0.0382, "step": 27960 }, { "grad_norm": 0.31741568446159363, "learning_rate": 7.740102433133102e-05, "loss": 0.0293, "step": 27970 }, { "grad_norm": 0.41401243209838867, "learning_rate": 7.738373362224553e-05, "loss": 0.0228, "step": 27980 }, { "grad_norm": 0.35721784830093384, "learning_rate": 7.73664382340289e-05, "loss": 0.0291, "step": 27990 }, { "grad_norm": 0.3966934382915497, "learning_rate": 7.734913816963647e-05, "loss": 0.038, "step": 28000 }, { "grad_norm": 0.3026881515979767, "learning_rate": 7.73318334320243e-05, "loss": 0.0254, "step": 28010 }, { "grad_norm": 0.41360512375831604, "learning_rate": 7.731452402414934e-05, "loss": 0.0357, "step": 28020 }, { "grad_norm": 0.3674142062664032, "learning_rate": 7.729720994896928e-05, "loss": 0.0284, "step": 28030 }, { "grad_norm": 0.2560403048992157, "learning_rate": 7.727989120944262e-05, "loss": 0.0246, "step": 28040 }, { "grad_norm": 0.26860901713371277, "learning_rate": 7.726256780852865e-05, "loss": 0.0272, "step": 28050 }, { "grad_norm": 0.30178019404411316, "learning_rate": 7.724523974918749e-05, "loss": 0.0273, "step": 28060 }, { "grad_norm": 0.2976773977279663, "learning_rate": 7.722790703438002e-05, "loss": 0.0229, "step": 28070 }, { "grad_norm": 0.197932168841362, "learning_rate": 7.72105696670679e-05, "loss": 0.0245, "step": 28080 }, { "grad_norm": 0.3111152946949005, "learning_rate": 7.719322765021364e-05, "loss": 0.025, "step": 28090 }, { "grad_norm": 0.2330089956521988, "learning_rate": 7.717588098678051e-05, "loss": 0.0256, "step": 28100 }, { "grad_norm": 0.29770365357398987, "learning_rate": 7.715852967973258e-05, "loss": 0.0307, "step": 28110 }, { "grad_norm": 0.3177908658981323, "learning_rate": 7.714117373203474e-05, "loss": 0.0299, "step": 28120 }, { "grad_norm": 0.31523391604423523, "learning_rate": 7.712381314665259e-05, "loss": 0.0265, "step": 28130 }, { "grad_norm": 0.31264516711235046, "learning_rate": 7.710644792655261e-05, "loss": 0.0243, "step": 28140 }, { "grad_norm": 0.3363702893257141, "learning_rate": 7.708907807470207e-05, "loss": 0.0394, "step": 28150 }, { "grad_norm": 0.2554241418838501, "learning_rate": 7.707170359406896e-05, "loss": 0.03, "step": 28160 }, { "grad_norm": 0.30497801303863525, "learning_rate": 7.705432448762213e-05, "loss": 0.027, "step": 28170 }, { "grad_norm": 0.3418862223625183, "learning_rate": 7.703694075833117e-05, "loss": 0.0344, "step": 28180 }, { "grad_norm": 0.2656199038028717, "learning_rate": 7.70195524091665e-05, "loss": 0.0449, "step": 28190 }, { "grad_norm": 0.3429723381996155, "learning_rate": 7.70021594430993e-05, "loss": 0.0337, "step": 28200 }, { "grad_norm": 0.2943231463432312, "learning_rate": 7.698476186310157e-05, "loss": 0.0269, "step": 28210 }, { "grad_norm": 0.27150794863700867, "learning_rate": 7.696735967214608e-05, "loss": 0.0282, "step": 28220 }, { "grad_norm": 0.3532640337944031, "learning_rate": 7.694995287320636e-05, "loss": 0.0286, "step": 28230 }, { "grad_norm": 0.19599014520645142, "learning_rate": 7.693254146925679e-05, "loss": 0.0206, "step": 28240 }, { "grad_norm": 0.40415701270103455, "learning_rate": 7.691512546327251e-05, "loss": 0.0359, "step": 28250 }, { "grad_norm": 0.38111260533332825, "learning_rate": 7.689770485822939e-05, "loss": 0.0308, "step": 28260 }, { "grad_norm": 0.24219588935375214, "learning_rate": 7.688027965710416e-05, "loss": 0.0227, "step": 28270 }, { "grad_norm": 0.36180806159973145, "learning_rate": 7.686284986287433e-05, "loss": 0.0314, "step": 28280 }, { "grad_norm": 0.30816805362701416, "learning_rate": 7.684541547851817e-05, "loss": 0.0252, "step": 28290 }, { "grad_norm": 0.2652404010295868, "learning_rate": 7.68279765070147e-05, "loss": 0.0237, "step": 28300 }, { "grad_norm": 0.28383028507232666, "learning_rate": 7.68105329513438e-05, "loss": 0.0249, "step": 28310 }, { "grad_norm": 0.28505605459213257, "learning_rate": 7.67930848144861e-05, "loss": 0.0238, "step": 28320 }, { "grad_norm": 0.23152756690979004, "learning_rate": 7.6775632099423e-05, "loss": 0.0295, "step": 28330 }, { "grad_norm": 0.2998890280723572, "learning_rate": 7.675817480913667e-05, "loss": 0.0269, "step": 28340 }, { "grad_norm": 0.36564502120018005, "learning_rate": 7.674071294661011e-05, "loss": 0.0237, "step": 28350 }, { "grad_norm": 0.2780441343784332, "learning_rate": 7.672324651482707e-05, "loss": 0.0259, "step": 28360 }, { "grad_norm": 0.27908360958099365, "learning_rate": 7.670577551677209e-05, "loss": 0.033, "step": 28370 }, { "grad_norm": 0.2697327136993408, "learning_rate": 7.668829995543047e-05, "loss": 0.0241, "step": 28380 }, { "grad_norm": 0.23763933777809143, "learning_rate": 7.667081983378832e-05, "loss": 0.0283, "step": 28390 }, { "grad_norm": 0.3510109484195709, "learning_rate": 7.66533351548325e-05, "loss": 0.0246, "step": 28400 }, { "grad_norm": 0.46295374631881714, "learning_rate": 7.663584592155069e-05, "loss": 0.0292, "step": 28410 }, { "grad_norm": 0.26365941762924194, "learning_rate": 7.661835213693129e-05, "loss": 0.0309, "step": 28420 }, { "grad_norm": 0.4161548912525177, "learning_rate": 7.660085380396353e-05, "loss": 0.0263, "step": 28430 }, { "grad_norm": 0.31214264035224915, "learning_rate": 7.658335092563738e-05, "loss": 0.0258, "step": 28440 }, { "grad_norm": 0.240200012922287, "learning_rate": 7.656584350494362e-05, "loss": 0.0288, "step": 28450 }, { "grad_norm": 0.40285977721214294, "learning_rate": 7.654833154487378e-05, "loss": 0.0295, "step": 28460 }, { "grad_norm": 0.29010599851608276, "learning_rate": 7.653081504842017e-05, "loss": 0.0279, "step": 28470 }, { "grad_norm": 0.31353968381881714, "learning_rate": 7.65132940185759e-05, "loss": 0.0327, "step": 28480 }, { "grad_norm": 0.28360673785209656, "learning_rate": 7.649576845833481e-05, "loss": 0.0166, "step": 28490 }, { "grad_norm": 0.2605115473270416, "learning_rate": 7.647823837069156e-05, "loss": 0.0238, "step": 28500 }, { "grad_norm": 0.2605801820755005, "learning_rate": 7.646070375864156e-05, "loss": 0.0231, "step": 28510 }, { "grad_norm": 0.2861862778663635, "learning_rate": 7.644316462518099e-05, "loss": 0.0339, "step": 28520 }, { "grad_norm": 0.31406527757644653, "learning_rate": 7.642562097330679e-05, "loss": 0.0275, "step": 28530 }, { "grad_norm": 0.226642906665802, "learning_rate": 7.640807280601671e-05, "loss": 0.0256, "step": 28540 }, { "grad_norm": 0.2387235313653946, "learning_rate": 7.639052012630927e-05, "loss": 0.0307, "step": 28550 }, { "grad_norm": 0.2710375189781189, "learning_rate": 7.63729629371837e-05, "loss": 0.0278, "step": 28560 }, { "grad_norm": 0.36564719676971436, "learning_rate": 7.635540124164009e-05, "loss": 0.034, "step": 28570 }, { "grad_norm": 0.3705938756465912, "learning_rate": 7.633783504267922e-05, "loss": 0.0321, "step": 28580 }, { "grad_norm": 0.22751782834529877, "learning_rate": 7.632026434330269e-05, "loss": 0.0236, "step": 28590 }, { "grad_norm": 0.2240697294473648, "learning_rate": 7.630268914651282e-05, "loss": 0.0234, "step": 28600 }, { "grad_norm": 0.2264992743730545, "learning_rate": 7.628510945531278e-05, "loss": 0.0357, "step": 28610 }, { "grad_norm": 0.23900093138217926, "learning_rate": 7.626752527270641e-05, "loss": 0.0284, "step": 28620 }, { "grad_norm": 0.27996915578842163, "learning_rate": 7.62499366016984e-05, "loss": 0.0232, "step": 28630 }, { "grad_norm": 0.19464974105358124, "learning_rate": 7.623234344529416e-05, "loss": 0.0259, "step": 28640 }, { "grad_norm": 0.3381612002849579, "learning_rate": 7.62147458064999e-05, "loss": 0.0276, "step": 28650 }, { "grad_norm": 0.36372503638267517, "learning_rate": 7.619714368832254e-05, "loss": 0.0282, "step": 28660 }, { "grad_norm": 0.26115861535072327, "learning_rate": 7.61795370937698e-05, "loss": 0.0227, "step": 28670 }, { "grad_norm": 0.32357510924339294, "learning_rate": 7.61619260258502e-05, "loss": 0.0289, "step": 28680 }, { "grad_norm": 0.23010089993476868, "learning_rate": 7.614431048757298e-05, "loss": 0.0283, "step": 28690 }, { "grad_norm": 0.35285863280296326, "learning_rate": 7.612669048194814e-05, "loss": 0.031, "step": 28700 }, { "grad_norm": 0.2197616994380951, "learning_rate": 7.610906601198646e-05, "loss": 0.0218, "step": 28710 }, { "grad_norm": 0.22572945058345795, "learning_rate": 7.60914370806995e-05, "loss": 0.0313, "step": 28720 }, { "grad_norm": 0.32363834977149963, "learning_rate": 7.607380369109953e-05, "loss": 0.0452, "step": 28730 }, { "grad_norm": 0.22027075290679932, "learning_rate": 7.605616584619961e-05, "loss": 0.0221, "step": 28740 }, { "grad_norm": 0.3834301233291626, "learning_rate": 7.603852354901362e-05, "loss": 0.0287, "step": 28750 }, { "grad_norm": 0.23232874274253845, "learning_rate": 7.602087680255609e-05, "loss": 0.0395, "step": 28760 }, { "grad_norm": 0.2835109531879425, "learning_rate": 7.600322560984238e-05, "loss": 0.024, "step": 28770 }, { "grad_norm": 0.25997593998908997, "learning_rate": 7.598556997388863e-05, "loss": 0.0295, "step": 28780 }, { "grad_norm": 0.30946436524391174, "learning_rate": 7.596790989771166e-05, "loss": 0.0318, "step": 28790 }, { "grad_norm": 0.21479035913944244, "learning_rate": 7.595024538432914e-05, "loss": 0.0255, "step": 28800 }, { "grad_norm": 0.2680317163467407, "learning_rate": 7.59325764367594e-05, "loss": 0.0242, "step": 28810 }, { "grad_norm": 0.2593498229980469, "learning_rate": 7.59149030580216e-05, "loss": 0.0282, "step": 28820 }, { "grad_norm": 0.24953040480613708, "learning_rate": 7.589722525113562e-05, "loss": 0.0243, "step": 28830 }, { "grad_norm": 0.3691098690032959, "learning_rate": 7.587954301912216e-05, "loss": 0.0241, "step": 28840 }, { "grad_norm": 0.4048673212528229, "learning_rate": 7.586185636500263e-05, "loss": 0.0325, "step": 28850 }, { "grad_norm": 0.29089784622192383, "learning_rate": 7.584416529179914e-05, "loss": 0.0296, "step": 28860 }, { "grad_norm": 0.3138309419155121, "learning_rate": 7.582646980253465e-05, "loss": 0.0299, "step": 28870 }, { "grad_norm": 0.36362630128860474, "learning_rate": 7.580876990023282e-05, "loss": 0.0383, "step": 28880 }, { "grad_norm": 0.3292756676673889, "learning_rate": 7.579106558791809e-05, "loss": 0.0253, "step": 28890 }, { "grad_norm": 0.27217650413513184, "learning_rate": 7.577335686861565e-05, "loss": 0.0283, "step": 28900 }, { "grad_norm": 0.3263973295688629, "learning_rate": 7.575564374535141e-05, "loss": 0.0297, "step": 28910 }, { "grad_norm": 0.28073301911354065, "learning_rate": 7.573792622115207e-05, "loss": 0.023, "step": 28920 }, { "grad_norm": 0.25588899850845337, "learning_rate": 7.572020429904507e-05, "loss": 0.0249, "step": 28930 }, { "grad_norm": 0.3696938455104828, "learning_rate": 7.570247798205861e-05, "loss": 0.0258, "step": 28940 }, { "grad_norm": 0.2376185953617096, "learning_rate": 7.568474727322164e-05, "loss": 0.03, "step": 28950 }, { "grad_norm": 0.2289164960384369, "learning_rate": 7.566701217556384e-05, "loss": 0.0248, "step": 28960 }, { "grad_norm": 0.2433357536792755, "learning_rate": 7.564927269211564e-05, "loss": 0.0253, "step": 28970 }, { "grad_norm": 0.29453131556510925, "learning_rate": 7.563152882590824e-05, "loss": 0.0241, "step": 28980 }, { "grad_norm": 0.40923792123794556, "learning_rate": 7.56137805799736e-05, "loss": 0.0259, "step": 28990 }, { "grad_norm": 0.23029543459415436, "learning_rate": 7.559602795734439e-05, "loss": 0.0261, "step": 29000 }, { "grad_norm": 0.2274797260761261, "learning_rate": 7.557827096105408e-05, "loss": 0.0241, "step": 29010 }, { "grad_norm": 0.2259618192911148, "learning_rate": 7.55605095941368e-05, "loss": 0.0279, "step": 29020 }, { "grad_norm": 0.20690250396728516, "learning_rate": 7.554274385962752e-05, "loss": 0.0277, "step": 29030 }, { "grad_norm": 0.33868953585624695, "learning_rate": 7.552497376056191e-05, "loss": 0.0313, "step": 29040 }, { "grad_norm": 0.29383185505867004, "learning_rate": 7.550719929997639e-05, "loss": 0.0303, "step": 29050 }, { "grad_norm": 0.4200375974178314, "learning_rate": 7.548942048090813e-05, "loss": 0.0299, "step": 29060 }, { "grad_norm": 0.2671498656272888, "learning_rate": 7.547163730639506e-05, "loss": 0.027, "step": 29070 }, { "grad_norm": 0.2511545717716217, "learning_rate": 7.545384977947583e-05, "loss": 0.0244, "step": 29080 }, { "grad_norm": 0.2584078013896942, "learning_rate": 7.543605790318981e-05, "loss": 0.0224, "step": 29090 }, { "grad_norm": 0.24690698087215424, "learning_rate": 7.54182616805772e-05, "loss": 0.0279, "step": 29100 }, { "grad_norm": 0.31361910700798035, "learning_rate": 7.540046111467885e-05, "loss": 0.0228, "step": 29110 }, { "grad_norm": 0.2721880376338959, "learning_rate": 7.53826562085364e-05, "loss": 0.0253, "step": 29120 }, { "grad_norm": 0.275727778673172, "learning_rate": 7.536484696519221e-05, "loss": 0.0187, "step": 29130 }, { "grad_norm": 0.275113970041275, "learning_rate": 7.534703338768942e-05, "loss": 0.0244, "step": 29140 }, { "grad_norm": 0.2801564335823059, "learning_rate": 7.532921547907185e-05, "loss": 0.0247, "step": 29150 }, { "grad_norm": 0.2935205101966858, "learning_rate": 7.531139324238412e-05, "loss": 0.0331, "step": 29160 }, { "grad_norm": 0.3239131271839142, "learning_rate": 7.529356668067157e-05, "loss": 0.0387, "step": 29170 }, { "grad_norm": 0.30472293496131897, "learning_rate": 7.527573579698023e-05, "loss": 0.0285, "step": 29180 }, { "grad_norm": 0.31456515192985535, "learning_rate": 7.525790059435693e-05, "loss": 0.025, "step": 29190 }, { "grad_norm": 0.25380533933639526, "learning_rate": 7.524006107584926e-05, "loss": 0.0281, "step": 29200 }, { "grad_norm": 0.371824711561203, "learning_rate": 7.522221724450544e-05, "loss": 0.0269, "step": 29210 }, { "grad_norm": 0.32329875230789185, "learning_rate": 7.520436910337451e-05, "loss": 0.0378, "step": 29220 }, { "grad_norm": 0.24635447561740875, "learning_rate": 7.518651665550627e-05, "loss": 0.0235, "step": 29230 }, { "grad_norm": 0.24715808033943176, "learning_rate": 7.516865990395117e-05, "loss": 0.0363, "step": 29240 }, { "grad_norm": 0.2948109805583954, "learning_rate": 7.515079885176047e-05, "loss": 0.0251, "step": 29250 }, { "grad_norm": 0.2197135090827942, "learning_rate": 7.513293350198612e-05, "loss": 0.0295, "step": 29260 }, { "grad_norm": 0.18306607007980347, "learning_rate": 7.511506385768081e-05, "loss": 0.0257, "step": 29270 }, { "grad_norm": 0.3369191288948059, "learning_rate": 7.509718992189801e-05, "loss": 0.0275, "step": 29280 }, { "grad_norm": 0.35372158885002136, "learning_rate": 7.507931169769182e-05, "loss": 0.0263, "step": 29290 }, { "grad_norm": 0.29282453656196594, "learning_rate": 7.506142918811722e-05, "loss": 0.025, "step": 29300 }, { "grad_norm": 0.2860022783279419, "learning_rate": 7.504354239622978e-05, "loss": 0.0313, "step": 29310 }, { "grad_norm": 0.2970966100692749, "learning_rate": 7.50256513250859e-05, "loss": 0.0291, "step": 29320 }, { "grad_norm": 0.3652958273887634, "learning_rate": 7.500775597774265e-05, "loss": 0.024, "step": 29330 }, { "grad_norm": 0.23972848057746887, "learning_rate": 7.498985635725788e-05, "loss": 0.0217, "step": 29340 }, { "grad_norm": 0.37415438890457153, "learning_rate": 7.497195246669012e-05, "loss": 0.0313, "step": 29350 }, { "grad_norm": 0.26621273159980774, "learning_rate": 7.495404430909868e-05, "loss": 0.0225, "step": 29360 }, { "grad_norm": 0.33668431639671326, "learning_rate": 7.493613188754356e-05, "loss": 0.0323, "step": 29370 }, { "grad_norm": 0.2614683508872986, "learning_rate": 7.49182152050855e-05, "loss": 0.0238, "step": 29380 }, { "grad_norm": 0.22978991270065308, "learning_rate": 7.490029426478598e-05, "loss": 0.0188, "step": 29390 }, { "grad_norm": 0.27806219458580017, "learning_rate": 7.488236906970719e-05, "loss": 0.026, "step": 29400 }, { "grad_norm": 0.3489893078804016, "learning_rate": 7.486443962291207e-05, "loss": 0.0231, "step": 29410 }, { "grad_norm": 0.2293596714735031, "learning_rate": 7.484650592746424e-05, "loss": 0.0177, "step": 29420 }, { "grad_norm": 0.18281003832817078, "learning_rate": 7.482856798642811e-05, "loss": 0.0268, "step": 29430 }, { "grad_norm": 0.2394413948059082, "learning_rate": 7.481062580286878e-05, "loss": 0.0218, "step": 29440 }, { "grad_norm": 0.31601661443710327, "learning_rate": 7.479267937985208e-05, "loss": 0.025, "step": 29450 }, { "grad_norm": 0.25774243474006653, "learning_rate": 7.477472872044456e-05, "loss": 0.019, "step": 29460 }, { "grad_norm": 0.3548644483089447, "learning_rate": 7.475677382771347e-05, "loss": 0.023, "step": 29470 }, { "grad_norm": 0.3002500534057617, "learning_rate": 7.473881470472683e-05, "loss": 0.0333, "step": 29480 }, { "grad_norm": 0.3416748344898224, "learning_rate": 7.472085135455337e-05, "loss": 0.0284, "step": 29490 }, { "grad_norm": 0.38970687985420227, "learning_rate": 7.470288378026256e-05, "loss": 0.0341, "step": 29500 }, { "grad_norm": 0.26923710107803345, "learning_rate": 7.468491198492451e-05, "loss": 0.0256, "step": 29510 }, { "grad_norm": 0.29671579599380493, "learning_rate": 7.466693597161013e-05, "loss": 0.0242, "step": 29520 }, { "grad_norm": 0.34026432037353516, "learning_rate": 7.464895574339104e-05, "loss": 0.0275, "step": 29530 }, { "grad_norm": 0.243890643119812, "learning_rate": 7.463097130333958e-05, "loss": 0.036, "step": 29540 }, { "grad_norm": 0.27839797735214233, "learning_rate": 7.461298265452876e-05, "loss": 0.0261, "step": 29550 }, { "grad_norm": 0.20484547317028046, "learning_rate": 7.459498980003239e-05, "loss": 0.0215, "step": 29560 }, { "grad_norm": 0.3095286786556244, "learning_rate": 7.457699274292493e-05, "loss": 0.0256, "step": 29570 }, { "grad_norm": 0.3398088812828064, "learning_rate": 7.455899148628159e-05, "loss": 0.0301, "step": 29580 }, { "grad_norm": 0.35271286964416504, "learning_rate": 7.45409860331783e-05, "loss": 0.0276, "step": 29590 }, { "grad_norm": 0.35732904076576233, "learning_rate": 7.452297638669169e-05, "loss": 0.0245, "step": 29600 }, { "grad_norm": 0.2284100204706192, "learning_rate": 7.450496254989911e-05, "loss": 0.0208, "step": 29610 }, { "grad_norm": 0.2555176317691803, "learning_rate": 7.448694452587866e-05, "loss": 0.0209, "step": 29620 }, { "grad_norm": 0.24459561705589294, "learning_rate": 7.44689223177091e-05, "loss": 0.0317, "step": 29630 }, { "grad_norm": 0.31132861971855164, "learning_rate": 7.445089592846994e-05, "loss": 0.022, "step": 29640 }, { "grad_norm": 0.30433356761932373, "learning_rate": 7.443286536124141e-05, "loss": 0.0315, "step": 29650 }, { "grad_norm": 0.28013867139816284, "learning_rate": 7.441483061910443e-05, "loss": 0.0328, "step": 29660 }, { "grad_norm": 0.2589455246925354, "learning_rate": 7.439679170514064e-05, "loss": 0.0299, "step": 29670 }, { "grad_norm": 0.28032881021499634, "learning_rate": 7.43787486224324e-05, "loss": 0.0291, "step": 29680 }, { "grad_norm": 0.33873888850212097, "learning_rate": 7.436070137406276e-05, "loss": 0.0275, "step": 29690 }, { "grad_norm": 0.2407546490430832, "learning_rate": 7.434264996311556e-05, "loss": 0.021, "step": 29700 }, { "grad_norm": 0.33910268545150757, "learning_rate": 7.432459439267525e-05, "loss": 0.029, "step": 29710 }, { "grad_norm": 0.2606920003890991, "learning_rate": 7.430653466582701e-05, "loss": 0.0326, "step": 29720 }, { "grad_norm": 0.25019654631614685, "learning_rate": 7.42884707856568e-05, "loss": 0.0437, "step": 29730 }, { "grad_norm": 0.24776852130889893, "learning_rate": 7.427040275525122e-05, "loss": 0.024, "step": 29740 }, { "grad_norm": 0.22407422959804535, "learning_rate": 7.42523305776976e-05, "loss": 0.029, "step": 29750 }, { "grad_norm": 0.2689487040042877, "learning_rate": 7.4234254256084e-05, "loss": 0.0225, "step": 29760 }, { "grad_norm": 0.2318924516439438, "learning_rate": 7.421617379349915e-05, "loss": 0.0299, "step": 29770 }, { "grad_norm": 0.2472514808177948, "learning_rate": 7.41980891930325e-05, "loss": 0.0254, "step": 29780 }, { "grad_norm": 0.2509292662143707, "learning_rate": 7.418000045777425e-05, "loss": 0.0196, "step": 29790 }, { "grad_norm": 0.2564636170864105, "learning_rate": 7.416190759081523e-05, "loss": 0.0241, "step": 29800 }, { "grad_norm": 0.3976004123687744, "learning_rate": 7.414381059524704e-05, "loss": 0.0246, "step": 29810 }, { "grad_norm": 0.2435440570116043, "learning_rate": 7.412570947416195e-05, "loss": 0.0213, "step": 29820 }, { "grad_norm": 0.4285528063774109, "learning_rate": 7.410760423065295e-05, "loss": 0.0395, "step": 29830 }, { "grad_norm": 0.3184622824192047, "learning_rate": 7.408949486781372e-05, "loss": 0.034, "step": 29840 }, { "grad_norm": 0.7563765645027161, "learning_rate": 7.407138138873868e-05, "loss": 0.0348, "step": 29850 }, { "grad_norm": 0.367546021938324, "learning_rate": 7.405326379652292e-05, "loss": 0.0279, "step": 29860 }, { "grad_norm": 0.3281102478504181, "learning_rate": 7.403514209426222e-05, "loss": 0.0253, "step": 29870 }, { "grad_norm": 0.3355005085468292, "learning_rate": 7.40170162850531e-05, "loss": 0.0287, "step": 29880 }, { "grad_norm": 0.22288797795772552, "learning_rate": 7.399888637199278e-05, "loss": 0.024, "step": 29890 }, { "grad_norm": 0.6906057596206665, "learning_rate": 7.398075235817914e-05, "loss": 0.0425, "step": 29900 }, { "grad_norm": 0.2784418761730194, "learning_rate": 7.39626142467108e-05, "loss": 0.0265, "step": 29910 }, { "grad_norm": 0.2902994155883789, "learning_rate": 7.394447204068706e-05, "loss": 0.0305, "step": 29920 }, { "grad_norm": 0.3363906145095825, "learning_rate": 7.392632574320793e-05, "loss": 0.0398, "step": 29930 }, { "grad_norm": 0.375777930021286, "learning_rate": 7.390817535737411e-05, "loss": 0.0326, "step": 29940 }, { "grad_norm": 0.38557398319244385, "learning_rate": 7.389002088628703e-05, "loss": 0.0292, "step": 29950 }, { "grad_norm": 0.43932563066482544, "learning_rate": 7.387186233304877e-05, "loss": 0.0346, "step": 29960 }, { "grad_norm": 0.41764941811561584, "learning_rate": 7.385369970076212e-05, "loss": 0.0276, "step": 29970 }, { "grad_norm": 0.3632534444332123, "learning_rate": 7.38355329925306e-05, "loss": 0.031, "step": 29980 }, { "grad_norm": 0.22165116667747498, "learning_rate": 7.381736221145838e-05, "loss": 0.0257, "step": 29990 }, { "grad_norm": 0.2697186768054962, "learning_rate": 7.37991873606504e-05, "loss": 0.0231, "step": 30000 }, { "grad_norm": 0.35392895340919495, "learning_rate": 7.378100844321218e-05, "loss": 0.0279, "step": 30010 }, { "grad_norm": 0.2101486623287201, "learning_rate": 7.376282546225004e-05, "loss": 0.0236, "step": 30020 }, { "grad_norm": 0.2466091811656952, "learning_rate": 7.374463842087094e-05, "loss": 0.0237, "step": 30030 }, { "grad_norm": 0.28609585762023926, "learning_rate": 7.372644732218254e-05, "loss": 0.0242, "step": 30040 }, { "grad_norm": 0.4218393564224243, "learning_rate": 7.370825216929322e-05, "loss": 0.037, "step": 30050 }, { "grad_norm": 0.3391057550907135, "learning_rate": 7.369005296531205e-05, "loss": 0.0322, "step": 30060 }, { "grad_norm": 0.38847681879997253, "learning_rate": 7.367184971334873e-05, "loss": 0.0295, "step": 30070 }, { "grad_norm": 0.3323945701122284, "learning_rate": 7.365364241651371e-05, "loss": 0.0304, "step": 30080 }, { "grad_norm": 0.24120740592479706, "learning_rate": 7.363543107791815e-05, "loss": 0.0238, "step": 30090 }, { "grad_norm": 0.31473860144615173, "learning_rate": 7.361721570067384e-05, "loss": 0.026, "step": 30100 }, { "grad_norm": 0.3073801100254059, "learning_rate": 7.359899628789331e-05, "loss": 0.0345, "step": 30110 }, { "grad_norm": 0.3855524957180023, "learning_rate": 7.358077284268974e-05, "loss": 0.0339, "step": 30120 }, { "grad_norm": 0.30711957812309265, "learning_rate": 7.356254536817702e-05, "loss": 0.0246, "step": 30130 }, { "grad_norm": 0.3097672760486603, "learning_rate": 7.354431386746973e-05, "loss": 0.0308, "step": 30140 }, { "grad_norm": 0.39411240816116333, "learning_rate": 7.352607834368316e-05, "loss": 0.0268, "step": 30150 }, { "grad_norm": 0.23352296650409698, "learning_rate": 7.350783879993324e-05, "loss": 0.0212, "step": 30160 }, { "grad_norm": 0.34125658869743347, "learning_rate": 7.348959523933658e-05, "loss": 0.0282, "step": 30170 }, { "grad_norm": 0.2133331000804901, "learning_rate": 7.347134766501057e-05, "loss": 0.0242, "step": 30180 }, { "grad_norm": 0.22354422509670258, "learning_rate": 7.345309608007315e-05, "loss": 0.0214, "step": 30190 }, { "grad_norm": 0.27668771147727966, "learning_rate": 7.343484048764308e-05, "loss": 0.0258, "step": 30200 }, { "grad_norm": 0.29179009795188904, "learning_rate": 7.341658089083972e-05, "loss": 0.0243, "step": 30210 }, { "grad_norm": 0.18969063460826874, "learning_rate": 7.339831729278313e-05, "loss": 0.0203, "step": 30220 }, { "grad_norm": 0.26548948884010315, "learning_rate": 7.338004969659404e-05, "loss": 0.0218, "step": 30230 }, { "grad_norm": 0.3464377522468567, "learning_rate": 7.336177810539391e-05, "loss": 0.0292, "step": 30240 }, { "grad_norm": 0.2773746848106384, "learning_rate": 7.334350252230485e-05, "loss": 0.0203, "step": 30250 }, { "grad_norm": 0.3301772177219391, "learning_rate": 7.332522295044965e-05, "loss": 0.0223, "step": 30260 }, { "grad_norm": 0.33694738149642944, "learning_rate": 7.33069393929518e-05, "loss": 0.0286, "step": 30270 }, { "grad_norm": 0.2832348048686981, "learning_rate": 7.328865185293545e-05, "loss": 0.0395, "step": 30280 }, { "grad_norm": 0.21093319356441498, "learning_rate": 7.327036033352546e-05, "loss": 0.031, "step": 30290 }, { "grad_norm": 0.2512490749359131, "learning_rate": 7.325206483784733e-05, "loss": 0.0235, "step": 30300 }, { "grad_norm": 0.3073091208934784, "learning_rate": 7.323376536902724e-05, "loss": 0.0346, "step": 30310 }, { "grad_norm": 0.36859622597694397, "learning_rate": 7.321546193019213e-05, "loss": 0.0303, "step": 30320 }, { "grad_norm": 0.2777453660964966, "learning_rate": 7.31971545244695e-05, "loss": 0.0297, "step": 30330 }, { "grad_norm": 0.3283306062221527, "learning_rate": 7.31788431549876e-05, "loss": 0.0261, "step": 30340 }, { "grad_norm": 0.2259717881679535, "learning_rate": 7.316052782487534e-05, "loss": 0.023, "step": 30350 }, { "grad_norm": 0.37636327743530273, "learning_rate": 7.314220853726234e-05, "loss": 0.0302, "step": 30360 }, { "grad_norm": 0.2470361590385437, "learning_rate": 7.312388529527884e-05, "loss": 0.0323, "step": 30370 }, { "grad_norm": 0.35736405849456787, "learning_rate": 7.310555810205577e-05, "loss": 0.0307, "step": 30380 }, { "grad_norm": 0.32603734731674194, "learning_rate": 7.308722696072476e-05, "loss": 0.0335, "step": 30390 }, { "grad_norm": 0.30073416233062744, "learning_rate": 7.306889187441811e-05, "loss": 0.0239, "step": 30400 }, { "grad_norm": 0.1837541162967682, "learning_rate": 7.305055284626876e-05, "loss": 0.0183, "step": 30410 }, { "grad_norm": 0.3205423951148987, "learning_rate": 7.303220987941037e-05, "loss": 0.0288, "step": 30420 }, { "grad_norm": 0.24048550426959991, "learning_rate": 7.301386297697726e-05, "loss": 0.0267, "step": 30430 }, { "grad_norm": 0.3543432354927063, "learning_rate": 7.299551214210438e-05, "loss": 0.024, "step": 30440 }, { "grad_norm": 0.26760154962539673, "learning_rate": 7.297715737792738e-05, "loss": 0.0261, "step": 30450 }, { "grad_norm": 0.28331324458122253, "learning_rate": 7.295879868758265e-05, "loss": 0.0237, "step": 30460 }, { "grad_norm": 0.2626793086528778, "learning_rate": 7.294043607420713e-05, "loss": 0.025, "step": 30470 }, { "grad_norm": 0.2434108704328537, "learning_rate": 7.292206954093852e-05, "loss": 0.0227, "step": 30480 }, { "grad_norm": 0.3039710819721222, "learning_rate": 7.290369909091515e-05, "loss": 0.0241, "step": 30490 }, { "grad_norm": 0.24491049349308014, "learning_rate": 7.2885324727276e-05, "loss": 0.0283, "step": 30500 }, { "grad_norm": 0.25485891103744507, "learning_rate": 7.286694645316076e-05, "loss": 0.0233, "step": 30510 }, { "grad_norm": 0.28077057003974915, "learning_rate": 7.284856427170982e-05, "loss": 0.0221, "step": 30520 }, { "grad_norm": 0.2647637724876404, "learning_rate": 7.283017818606414e-05, "loss": 0.028, "step": 30530 }, { "grad_norm": 0.37159308791160583, "learning_rate": 7.28117881993654e-05, "loss": 0.0328, "step": 30540 }, { "grad_norm": 0.2734667956829071, "learning_rate": 7.279339431475598e-05, "loss": 0.0242, "step": 30550 }, { "grad_norm": 0.1865704506635666, "learning_rate": 7.277499653537887e-05, "loss": 0.0217, "step": 30560 }, { "grad_norm": 0.30771443247795105, "learning_rate": 7.275659486437776e-05, "loss": 0.0325, "step": 30570 }, { "grad_norm": 0.35740700364112854, "learning_rate": 7.273818930489695e-05, "loss": 0.0339, "step": 30580 }, { "grad_norm": 0.24022546410560608, "learning_rate": 7.271977986008151e-05, "loss": 0.029, "step": 30590 }, { "grad_norm": 0.28510913252830505, "learning_rate": 7.270136653307705e-05, "loss": 0.0239, "step": 30600 }, { "grad_norm": 0.2565404772758484, "learning_rate": 7.268294932702994e-05, "loss": 0.028, "step": 30610 }, { "grad_norm": 0.361847460269928, "learning_rate": 7.266452824508719e-05, "loss": 0.0298, "step": 30620 }, { "grad_norm": 0.25948894023895264, "learning_rate": 7.264610329039643e-05, "loss": 0.0246, "step": 30630 }, { "grad_norm": 0.22516809403896332, "learning_rate": 7.262767446610599e-05, "loss": 0.0201, "step": 30640 }, { "grad_norm": 0.3022496700286865, "learning_rate": 7.260924177536485e-05, "loss": 0.0295, "step": 30650 }, { "grad_norm": 0.23107169568538666, "learning_rate": 7.259080522132265e-05, "loss": 0.021, "step": 30660 }, { "grad_norm": 0.31894946098327637, "learning_rate": 7.257236480712972e-05, "loss": 0.0244, "step": 30670 }, { "grad_norm": 0.36321502923965454, "learning_rate": 7.255392053593697e-05, "loss": 0.0199, "step": 30680 }, { "grad_norm": 0.4143223762512207, "learning_rate": 7.253547241089607e-05, "loss": 0.0235, "step": 30690 }, { "grad_norm": 0.22525137662887573, "learning_rate": 7.251702043515927e-05, "loss": 0.0316, "step": 30700 }, { "grad_norm": 0.2402942180633545, "learning_rate": 7.249856461187952e-05, "loss": 0.0286, "step": 30710 }, { "grad_norm": 0.36903563141822815, "learning_rate": 7.248010494421042e-05, "loss": 0.0269, "step": 30720 }, { "grad_norm": 0.26880425214767456, "learning_rate": 7.246164143530622e-05, "loss": 0.0347, "step": 30730 }, { "grad_norm": 0.28107526898384094, "learning_rate": 7.244317408832181e-05, "loss": 0.0237, "step": 30740 }, { "grad_norm": 0.2285783737897873, "learning_rate": 7.242470290641279e-05, "loss": 0.023, "step": 30750 }, { "grad_norm": 0.3486366868019104, "learning_rate": 7.240622789273536e-05, "loss": 0.0209, "step": 30760 }, { "grad_norm": 0.21509228646755219, "learning_rate": 7.238774905044638e-05, "loss": 0.0211, "step": 30770 }, { "grad_norm": 0.2274996042251587, "learning_rate": 7.236926638270341e-05, "loss": 0.0221, "step": 30780 }, { "grad_norm": 0.2685796618461609, "learning_rate": 7.23507798926646e-05, "loss": 0.0278, "step": 30790 }, { "grad_norm": 0.32256147265434265, "learning_rate": 7.23322895834888e-05, "loss": 0.024, "step": 30800 }, { "grad_norm": 0.35807451605796814, "learning_rate": 7.231379545833552e-05, "loss": 0.0251, "step": 30810 }, { "grad_norm": 0.2226993441581726, "learning_rate": 7.229529752036487e-05, "loss": 0.0246, "step": 30820 }, { "grad_norm": 0.23331670463085175, "learning_rate": 7.227679577273765e-05, "loss": 0.0284, "step": 30830 }, { "grad_norm": 0.3573513627052307, "learning_rate": 7.225829021861529e-05, "loss": 0.0272, "step": 30840 }, { "grad_norm": 0.2917751967906952, "learning_rate": 7.223978086115992e-05, "loss": 0.019, "step": 30850 }, { "grad_norm": 0.2730124294757843, "learning_rate": 7.222126770353425e-05, "loss": 0.0317, "step": 30860 }, { "grad_norm": 0.2253316193819046, "learning_rate": 7.22027507489017e-05, "loss": 0.0206, "step": 30870 }, { "grad_norm": 0.3292413651943207, "learning_rate": 7.218423000042627e-05, "loss": 0.0195, "step": 30880 }, { "grad_norm": 0.2785123884677887, "learning_rate": 7.216570546127268e-05, "loss": 0.0267, "step": 30890 }, { "grad_norm": 0.2695539593696594, "learning_rate": 7.214717713460626e-05, "loss": 0.0324, "step": 30900 }, { "grad_norm": 0.283485472202301, "learning_rate": 7.2128645023593e-05, "loss": 0.0281, "step": 30910 }, { "grad_norm": 0.2752434313297272, "learning_rate": 7.211010913139951e-05, "loss": 0.0294, "step": 30920 }, { "grad_norm": 0.31864771246910095, "learning_rate": 7.209156946119308e-05, "loss": 0.0246, "step": 30930 }, { "grad_norm": 0.19498249888420105, "learning_rate": 7.207302601614166e-05, "loss": 0.0249, "step": 30940 }, { "grad_norm": 0.2533477246761322, "learning_rate": 7.205447879941378e-05, "loss": 0.0316, "step": 30950 }, { "grad_norm": 0.3586105704307556, "learning_rate": 7.203592781417866e-05, "loss": 0.0271, "step": 30960 }, { "grad_norm": 0.3353080153465271, "learning_rate": 7.201737306360617e-05, "loss": 0.0276, "step": 30970 }, { "grad_norm": 0.3122681677341461, "learning_rate": 7.19988145508668e-05, "loss": 0.0213, "step": 30980 }, { "grad_norm": 0.3195059299468994, "learning_rate": 7.198025227913168e-05, "loss": 0.0255, "step": 30990 }, { "grad_norm": 0.3236371874809265, "learning_rate": 7.196168625157261e-05, "loss": 0.0251, "step": 31000 }, { "grad_norm": 0.24543623626232147, "learning_rate": 7.194311647136201e-05, "loss": 0.022, "step": 31010 }, { "grad_norm": 0.24483297765254974, "learning_rate": 7.192454294167297e-05, "loss": 0.0194, "step": 31020 }, { "grad_norm": 0.28213948011398315, "learning_rate": 7.190596566567917e-05, "loss": 0.0246, "step": 31030 }, { "grad_norm": 0.29579633474349976, "learning_rate": 7.188738464655496e-05, "loss": 0.025, "step": 31040 }, { "grad_norm": 0.2121368944644928, "learning_rate": 7.186879988747533e-05, "loss": 0.0215, "step": 31050 }, { "grad_norm": 0.27121496200561523, "learning_rate": 7.185021139161592e-05, "loss": 0.0166, "step": 31060 }, { "grad_norm": 0.30623018741607666, "learning_rate": 7.1831619162153e-05, "loss": 0.0225, "step": 31070 }, { "grad_norm": 0.3100147247314453, "learning_rate": 7.181302320226345e-05, "loss": 0.0199, "step": 31080 }, { "grad_norm": 0.25555452704429626, "learning_rate": 7.179442351512482e-05, "loss": 0.032, "step": 31090 }, { "grad_norm": 0.28231608867645264, "learning_rate": 7.177582010391528e-05, "loss": 0.0235, "step": 31100 }, { "grad_norm": 0.3412419259548187, "learning_rate": 7.175721297181366e-05, "loss": 0.027, "step": 31110 }, { "grad_norm": 0.37590673565864563, "learning_rate": 7.173860212199942e-05, "loss": 0.0262, "step": 31120 }, { "grad_norm": 0.23901906609535217, "learning_rate": 7.171998755765263e-05, "loss": 0.0248, "step": 31130 }, { "grad_norm": 0.346799373626709, "learning_rate": 7.170136928195398e-05, "loss": 0.0276, "step": 31140 }, { "grad_norm": 0.28064754605293274, "learning_rate": 7.168274729808489e-05, "loss": 0.0222, "step": 31150 }, { "grad_norm": 0.2529304623603821, "learning_rate": 7.166412160922728e-05, "loss": 0.0221, "step": 31160 }, { "grad_norm": 0.24644088745117188, "learning_rate": 7.164549221856382e-05, "loss": 0.0211, "step": 31170 }, { "grad_norm": 0.2321680635213852, "learning_rate": 7.162685912927775e-05, "loss": 0.0369, "step": 31180 }, { "grad_norm": 0.22391383349895477, "learning_rate": 7.160822234455294e-05, "loss": 0.0249, "step": 31190 }, { "grad_norm": 0.27802932262420654, "learning_rate": 7.158958186757391e-05, "loss": 0.0213, "step": 31200 }, { "grad_norm": 0.3054168224334717, "learning_rate": 7.157093770152582e-05, "loss": 0.0238, "step": 31210 }, { "grad_norm": 1.327174186706543, "learning_rate": 7.155228984959446e-05, "loss": 0.0262, "step": 31220 }, { "grad_norm": 0.3066554367542267, "learning_rate": 7.153363831496621e-05, "loss": 0.0287, "step": 31230 }, { "grad_norm": 0.31074172258377075, "learning_rate": 7.151498310082811e-05, "loss": 0.0395, "step": 31240 }, { "grad_norm": 0.2439933717250824, "learning_rate": 7.149632421036784e-05, "loss": 0.0322, "step": 31250 }, { "grad_norm": 0.4107505977153778, "learning_rate": 7.147766164677369e-05, "loss": 0.0458, "step": 31260 }, { "grad_norm": 0.28251913189888, "learning_rate": 7.145899541323459e-05, "loss": 0.0202, "step": 31270 }, { "grad_norm": 0.23619656264781952, "learning_rate": 7.144032551294007e-05, "loss": 0.0355, "step": 31280 }, { "grad_norm": 0.3438284993171692, "learning_rate": 7.14216519490803e-05, "loss": 0.0291, "step": 31290 }, { "grad_norm": 0.4152251183986664, "learning_rate": 7.140297472484609e-05, "loss": 0.0366, "step": 31300 }, { "grad_norm": 0.16218169033527374, "learning_rate": 7.138429384342891e-05, "loss": 0.0471, "step": 31310 }, { "grad_norm": 0.30879077315330505, "learning_rate": 7.136560930802074e-05, "loss": 0.0229, "step": 31320 }, { "grad_norm": 0.34624114632606506, "learning_rate": 7.134692112181431e-05, "loss": 0.0242, "step": 31330 }, { "grad_norm": 0.2895081639289856, "learning_rate": 7.13282292880029e-05, "loss": 0.0231, "step": 31340 }, { "grad_norm": 0.18570640683174133, "learning_rate": 7.130953380978043e-05, "loss": 0.0234, "step": 31350 }, { "grad_norm": 0.3161744475364685, "learning_rate": 7.129083469034144e-05, "loss": 0.0338, "step": 31360 }, { "grad_norm": 0.343353271484375, "learning_rate": 7.127213193288112e-05, "loss": 0.0277, "step": 31370 }, { "grad_norm": 0.29632461071014404, "learning_rate": 7.125342554059522e-05, "loss": 0.0244, "step": 31380 }, { "grad_norm": 0.3451012670993805, "learning_rate": 7.12347155166802e-05, "loss": 0.0269, "step": 31390 }, { "grad_norm": 0.19784468412399292, "learning_rate": 7.121600186433306e-05, "loss": 0.0212, "step": 31400 }, { "grad_norm": 0.28341948986053467, "learning_rate": 7.119728458675148e-05, "loss": 0.0223, "step": 31410 }, { "grad_norm": 0.2663070261478424, "learning_rate": 7.117856368713369e-05, "loss": 0.0204, "step": 31420 }, { "grad_norm": 0.3099002540111542, "learning_rate": 7.115983916867861e-05, "loss": 0.0252, "step": 31430 }, { "grad_norm": 0.20919105410575867, "learning_rate": 7.114111103458574e-05, "loss": 0.0217, "step": 31440 }, { "grad_norm": 0.20479553937911987, "learning_rate": 7.11223792880552e-05, "loss": 0.0233, "step": 31450 }, { "grad_norm": 0.36616331338882446, "learning_rate": 7.110364393228773e-05, "loss": 0.0243, "step": 31460 }, { "grad_norm": 0.262099951505661, "learning_rate": 7.108490497048471e-05, "loss": 0.0191, "step": 31470 }, { "grad_norm": 0.21286045014858246, "learning_rate": 7.10661624058481e-05, "loss": 0.0256, "step": 31480 }, { "grad_norm": 0.1977788507938385, "learning_rate": 7.10474162415805e-05, "loss": 0.0239, "step": 31490 }, { "grad_norm": 0.2067641168832779, "learning_rate": 7.102866648088511e-05, "loss": 0.0267, "step": 31500 }, { "grad_norm": 0.25467395782470703, "learning_rate": 7.100991312696576e-05, "loss": 0.0223, "step": 31510 }, { "grad_norm": 0.2069297432899475, "learning_rate": 7.099115618302686e-05, "loss": 0.0214, "step": 31520 }, { "grad_norm": 0.247879296541214, "learning_rate": 7.097239565227349e-05, "loss": 0.0277, "step": 31530 }, { "grad_norm": 0.2807924747467041, "learning_rate": 7.09536315379113e-05, "loss": 0.0204, "step": 31540 }, { "grad_norm": 0.4474344849586487, "learning_rate": 7.093486384314656e-05, "loss": 0.0222, "step": 31550 }, { "grad_norm": 0.24494141340255737, "learning_rate": 7.091609257118616e-05, "loss": 0.0249, "step": 31560 }, { "grad_norm": 0.24939551949501038, "learning_rate": 7.08973177252376e-05, "loss": 0.0284, "step": 31570 }, { "grad_norm": 0.19954219460487366, "learning_rate": 7.087853930850898e-05, "loss": 0.0217, "step": 31580 }, { "grad_norm": 0.38479122519493103, "learning_rate": 7.085975732420903e-05, "loss": 0.023, "step": 31590 }, { "grad_norm": 0.336930513381958, "learning_rate": 7.084097177554706e-05, "loss": 0.0174, "step": 31600 }, { "grad_norm": 0.3024112582206726, "learning_rate": 7.082218266573301e-05, "loss": 0.0314, "step": 31610 }, { "grad_norm": 0.3677879571914673, "learning_rate": 7.080338999797743e-05, "loss": 0.025, "step": 31620 }, { "grad_norm": 0.2676643133163452, "learning_rate": 7.07845937754915e-05, "loss": 0.0271, "step": 31630 }, { "grad_norm": 0.23939140141010284, "learning_rate": 7.076579400148693e-05, "loss": 0.0299, "step": 31640 }, { "grad_norm": 0.28275999426841736, "learning_rate": 7.074699067917611e-05, "loss": 0.0257, "step": 31650 }, { "grad_norm": 0.2797185778617859, "learning_rate": 7.072818381177201e-05, "loss": 0.0237, "step": 31660 }, { "grad_norm": 0.24359586834907532, "learning_rate": 7.070937340248823e-05, "loss": 0.0237, "step": 31670 }, { "grad_norm": 0.2677321135997772, "learning_rate": 7.069055945453893e-05, "loss": 0.0258, "step": 31680 }, { "grad_norm": 0.26365119218826294, "learning_rate": 7.067174197113892e-05, "loss": 0.021, "step": 31690 }, { "grad_norm": 0.35403987765312195, "learning_rate": 7.065292095550355e-05, "loss": 0.0252, "step": 31700 }, { "grad_norm": 0.32618746161460876, "learning_rate": 7.063409641084887e-05, "loss": 0.0271, "step": 31710 }, { "grad_norm": 0.31286755204200745, "learning_rate": 7.061526834039145e-05, "loss": 0.0349, "step": 31720 }, { "grad_norm": 0.2629329562187195, "learning_rate": 7.05964367473485e-05, "loss": 0.0214, "step": 31730 }, { "grad_norm": 0.24439658224582672, "learning_rate": 7.057760163493783e-05, "loss": 0.0278, "step": 31740 }, { "grad_norm": 0.2610098421573639, "learning_rate": 7.055876300637783e-05, "loss": 0.023, "step": 31750 }, { "grad_norm": 0.311670184135437, "learning_rate": 7.053992086488753e-05, "loss": 0.0273, "step": 31760 }, { "grad_norm": 0.2961331605911255, "learning_rate": 7.052107521368651e-05, "loss": 0.0301, "step": 31770 }, { "grad_norm": 0.2911972403526306, "learning_rate": 7.0502226055995e-05, "loss": 0.0361, "step": 31780 }, { "grad_norm": 0.27374517917633057, "learning_rate": 7.048337339503379e-05, "loss": 0.0412, "step": 31790 }, { "grad_norm": 0.33982011675834656, "learning_rate": 7.046451723402427e-05, "loss": 0.0235, "step": 31800 }, { "grad_norm": 0.3197256028652191, "learning_rate": 7.044565757618848e-05, "loss": 0.0377, "step": 31810 }, { "grad_norm": 0.2937954366207123, "learning_rate": 7.042679442474899e-05, "loss": 0.0207, "step": 31820 }, { "grad_norm": 0.2373141497373581, "learning_rate": 7.040792778292902e-05, "loss": 0.0358, "step": 31830 }, { "grad_norm": 0.24554602801799774, "learning_rate": 7.038905765395234e-05, "loss": 0.0358, "step": 31840 }, { "grad_norm": 0.3747347295284271, "learning_rate": 7.037018404104334e-05, "loss": 0.0271, "step": 31850 }, { "grad_norm": 0.333111047744751, "learning_rate": 7.035130694742702e-05, "loss": 0.0217, "step": 31860 }, { "grad_norm": 0.31817251443862915, "learning_rate": 7.033242637632897e-05, "loss": 0.0265, "step": 31870 }, { "grad_norm": 0.35516536235809326, "learning_rate": 7.031354233097534e-05, "loss": 0.0276, "step": 31880 }, { "grad_norm": 0.3533261716365814, "learning_rate": 7.029465481459289e-05, "loss": 0.0383, "step": 31890 }, { "grad_norm": 0.24141165614128113, "learning_rate": 7.027576383040898e-05, "loss": 0.026, "step": 31900 }, { "grad_norm": 0.22446492314338684, "learning_rate": 7.025686938165159e-05, "loss": 0.0294, "step": 31910 }, { "grad_norm": 0.31080901622772217, "learning_rate": 7.023797147154924e-05, "loss": 0.0402, "step": 31920 }, { "grad_norm": 0.27180758118629456, "learning_rate": 7.021907010333111e-05, "loss": 0.0224, "step": 31930 }, { "grad_norm": 0.29350894689559937, "learning_rate": 7.020016528022685e-05, "loss": 0.0211, "step": 31940 }, { "grad_norm": 0.2559690773487091, "learning_rate": 7.018125700546683e-05, "loss": 0.0292, "step": 31950 }, { "grad_norm": 0.2839375436306, "learning_rate": 7.016234528228196e-05, "loss": 0.0278, "step": 31960 }, { "grad_norm": 0.361334890127182, "learning_rate": 7.014343011390372e-05, "loss": 0.0217, "step": 31970 }, { "grad_norm": 0.19432741403579712, "learning_rate": 7.01245115035642e-05, "loss": 0.0176, "step": 31980 }, { "grad_norm": 0.2806328535079956, "learning_rate": 7.010558945449606e-05, "loss": 0.0322, "step": 31990 }, { "grad_norm": 0.32361602783203125, "learning_rate": 7.008666396993258e-05, "loss": 0.0189, "step": 32000 }, { "grad_norm": 0.39132317900657654, "learning_rate": 7.006773505310759e-05, "loss": 0.0325, "step": 32010 }, { "grad_norm": 0.3115174472332001, "learning_rate": 7.004880270725553e-05, "loss": 0.0244, "step": 32020 }, { "grad_norm": 0.3513984680175781, "learning_rate": 7.002986693561144e-05, "loss": 0.021, "step": 32030 }, { "grad_norm": 0.2493661344051361, "learning_rate": 7.001092774141089e-05, "loss": 0.0275, "step": 32040 }, { "grad_norm": 0.20940391719341278, "learning_rate": 6.999198512789009e-05, "loss": 0.0243, "step": 32050 }, { "grad_norm": 0.3434496819972992, "learning_rate": 6.997303909828584e-05, "loss": 0.0299, "step": 32060 }, { "grad_norm": 0.5107635855674744, "learning_rate": 6.995408965583544e-05, "loss": 0.0285, "step": 32070 }, { "grad_norm": 0.29052895307540894, "learning_rate": 6.993513680377688e-05, "loss": 0.0172, "step": 32080 }, { "grad_norm": 0.21392658352851868, "learning_rate": 6.991618054534868e-05, "loss": 0.0354, "step": 32090 }, { "grad_norm": 0.40616554021835327, "learning_rate": 6.989722088378991e-05, "loss": 0.0248, "step": 32100 }, { "grad_norm": 0.33355990052223206, "learning_rate": 6.987825782234027e-05, "loss": 0.024, "step": 32110 }, { "grad_norm": 0.17597734928131104, "learning_rate": 6.985929136424006e-05, "loss": 0.0213, "step": 32120 }, { "grad_norm": 0.5232751369476318, "learning_rate": 6.984032151273012e-05, "loss": 0.029, "step": 32130 }, { "grad_norm": 0.25387027859687805, "learning_rate": 6.982134827105186e-05, "loss": 0.022, "step": 32140 }, { "grad_norm": 0.29364773631095886, "learning_rate": 6.980237164244729e-05, "loss": 0.0188, "step": 32150 }, { "grad_norm": 0.29368165135383606, "learning_rate": 6.9783391630159e-05, "loss": 0.0213, "step": 32160 }, { "grad_norm": 0.2312403917312622, "learning_rate": 6.976440823743015e-05, "loss": 0.0236, "step": 32170 }, { "grad_norm": 0.2706664204597473, "learning_rate": 6.974542146750451e-05, "loss": 0.0206, "step": 32180 }, { "grad_norm": 0.3552219867706299, "learning_rate": 6.972643132362637e-05, "loss": 0.0205, "step": 32190 }, { "grad_norm": 0.29072001576423645, "learning_rate": 6.970743780904064e-05, "loss": 0.0232, "step": 32200 }, { "grad_norm": 0.33126363158226013, "learning_rate": 6.968844092699277e-05, "loss": 0.0302, "step": 32210 }, { "grad_norm": 0.24509096145629883, "learning_rate": 6.966944068072883e-05, "loss": 0.025, "step": 32220 }, { "grad_norm": 0.24933038651943207, "learning_rate": 6.965043707349545e-05, "loss": 0.0213, "step": 32230 }, { "grad_norm": 0.47137635946273804, "learning_rate": 6.963143010853982e-05, "loss": 0.0283, "step": 32240 }, { "grad_norm": 0.3720119297504425, "learning_rate": 6.961241978910971e-05, "loss": 0.0283, "step": 32250 }, { "grad_norm": 0.32479962706565857, "learning_rate": 6.959340611845344e-05, "loss": 0.0262, "step": 32260 }, { "grad_norm": 0.27531036734580994, "learning_rate": 6.957438909981995e-05, "loss": 0.0203, "step": 32270 }, { "grad_norm": 0.2543330788612366, "learning_rate": 6.955536873645872e-05, "loss": 0.0161, "step": 32280 }, { "grad_norm": 0.3018813729286194, "learning_rate": 6.953634503161982e-05, "loss": 0.0228, "step": 32290 }, { "grad_norm": 0.22223125398159027, "learning_rate": 6.951731798855387e-05, "loss": 0.0238, "step": 32300 }, { "grad_norm": 0.24067486822605133, "learning_rate": 6.949828761051208e-05, "loss": 0.0304, "step": 32310 }, { "grad_norm": 0.5232728719711304, "learning_rate": 6.947925390074622e-05, "loss": 0.0185, "step": 32320 }, { "grad_norm": 0.22835834324359894, "learning_rate": 6.946021686250863e-05, "loss": 0.0185, "step": 32330 }, { "grad_norm": 0.21087457239627838, "learning_rate": 6.94411764990522e-05, "loss": 0.0246, "step": 32340 }, { "grad_norm": 0.22987402975559235, "learning_rate": 6.942213281363044e-05, "loss": 0.0171, "step": 32350 }, { "grad_norm": 0.30538684129714966, "learning_rate": 6.940308580949737e-05, "loss": 0.0222, "step": 32360 }, { "grad_norm": 0.23034600913524628, "learning_rate": 6.93840354899076e-05, "loss": 0.0212, "step": 32370 }, { "grad_norm": 0.2761320471763611, "learning_rate": 6.936498185811633e-05, "loss": 0.0233, "step": 32380 }, { "grad_norm": 0.18569789826869965, "learning_rate": 6.93459249173793e-05, "loss": 0.0145, "step": 32390 }, { "grad_norm": 0.3726232051849365, "learning_rate": 6.932686467095279e-05, "loss": 0.023, "step": 32400 }, { "grad_norm": 0.34136056900024414, "learning_rate": 6.930780112209373e-05, "loss": 0.0227, "step": 32410 }, { "grad_norm": 0.30155056715011597, "learning_rate": 6.92887342740595e-05, "loss": 0.025, "step": 32420 }, { "grad_norm": 0.2802553176879883, "learning_rate": 6.926966413010816e-05, "loss": 0.0324, "step": 32430 }, { "grad_norm": 0.2539100646972656, "learning_rate": 6.925059069349824e-05, "loss": 0.0193, "step": 32440 }, { "grad_norm": 0.24843581020832062, "learning_rate": 6.923151396748886e-05, "loss": 0.0261, "step": 32450 }, { "grad_norm": 0.24866029620170593, "learning_rate": 6.921243395533974e-05, "loss": 0.0206, "step": 32460 }, { "grad_norm": 0.3492155075073242, "learning_rate": 6.919335066031109e-05, "loss": 0.0341, "step": 32470 }, { "grad_norm": 0.34230005741119385, "learning_rate": 6.917426408566379e-05, "loss": 0.0239, "step": 32480 }, { "grad_norm": 0.37729939818382263, "learning_rate": 6.915517423465916e-05, "loss": 0.0224, "step": 32490 }, { "grad_norm": 0.33197346329689026, "learning_rate": 6.913608111055914e-05, "loss": 0.0237, "step": 32500 }, { "grad_norm": 0.4504473805427551, "learning_rate": 6.911698471662623e-05, "loss": 0.0254, "step": 32510 }, { "grad_norm": 0.2831518054008484, "learning_rate": 6.90978850561235e-05, "loss": 0.0211, "step": 32520 }, { "grad_norm": 0.3261747360229492, "learning_rate": 6.907878213231454e-05, "loss": 0.0223, "step": 32530 }, { "grad_norm": 0.5865008234977722, "learning_rate": 6.90596759484635e-05, "loss": 0.0244, "step": 32540 }, { "grad_norm": 0.3034939169883728, "learning_rate": 6.904056650783514e-05, "loss": 0.033, "step": 32550 }, { "grad_norm": 0.2641252875328064, "learning_rate": 6.902145381369471e-05, "loss": 0.0328, "step": 32560 }, { "grad_norm": 0.26080939173698425, "learning_rate": 6.900233786930808e-05, "loss": 0.0297, "step": 32570 }, { "grad_norm": 0.41596123576164246, "learning_rate": 6.898321867794161e-05, "loss": 0.0254, "step": 32580 }, { "grad_norm": 0.16418658196926117, "learning_rate": 6.896409624286226e-05, "loss": 0.024, "step": 32590 }, { "grad_norm": 0.24595682322978973, "learning_rate": 6.894497056733754e-05, "loss": 0.0267, "step": 32600 }, { "grad_norm": 0.3046455979347229, "learning_rate": 6.89258416546355e-05, "loss": 0.025, "step": 32610 }, { "grad_norm": 0.22108399868011475, "learning_rate": 6.890670950802474e-05, "loss": 0.0269, "step": 32620 }, { "grad_norm": 0.22135470807552338, "learning_rate": 6.88875741307744e-05, "loss": 0.0231, "step": 32630 }, { "grad_norm": 0.17994020879268646, "learning_rate": 6.886843552615425e-05, "loss": 0.0229, "step": 32640 }, { "grad_norm": 0.21639645099639893, "learning_rate": 6.884929369743451e-05, "loss": 0.0242, "step": 32650 }, { "grad_norm": 0.41211816668510437, "learning_rate": 6.8830148647886e-05, "loss": 0.0205, "step": 32660 }, { "grad_norm": 0.22468610107898712, "learning_rate": 6.88110003807801e-05, "loss": 0.0245, "step": 32670 }, { "grad_norm": 0.26244476437568665, "learning_rate": 6.87918488993887e-05, "loss": 0.0208, "step": 32680 }, { "grad_norm": 0.3616548180580139, "learning_rate": 6.877269420698431e-05, "loss": 0.025, "step": 32690 }, { "grad_norm": 0.30510398745536804, "learning_rate": 6.875353630683989e-05, "loss": 0.0259, "step": 32700 }, { "grad_norm": 0.41655439138412476, "learning_rate": 6.873437520222905e-05, "loss": 0.0285, "step": 32710 }, { "grad_norm": 0.1918834000825882, "learning_rate": 6.871521089642585e-05, "loss": 0.0225, "step": 32720 }, { "grad_norm": 0.21301323175430298, "learning_rate": 6.869604339270498e-05, "loss": 0.0213, "step": 32730 }, { "grad_norm": 0.24956656992435455, "learning_rate": 6.867687269434164e-05, "loss": 0.0285, "step": 32740 }, { "grad_norm": 0.23073232173919678, "learning_rate": 6.865769880461156e-05, "loss": 0.023, "step": 32750 }, { "grad_norm": 0.24945218861103058, "learning_rate": 6.863852172679104e-05, "loss": 0.0236, "step": 32760 }, { "grad_norm": 0.44800522923469543, "learning_rate": 6.861934146415693e-05, "loss": 0.0219, "step": 32770 }, { "grad_norm": 0.25199806690216064, "learning_rate": 6.86001580199866e-05, "loss": 0.0333, "step": 32780 }, { "grad_norm": 0.22231757640838623, "learning_rate": 6.858097139755798e-05, "loss": 0.0176, "step": 32790 }, { "grad_norm": 0.3375416100025177, "learning_rate": 6.856178160014955e-05, "loss": 0.0257, "step": 32800 }, { "grad_norm": 0.25825247168540955, "learning_rate": 6.85425886310403e-05, "loss": 0.0201, "step": 32810 }, { "grad_norm": 0.36807334423065186, "learning_rate": 6.852339249350979e-05, "loss": 0.0257, "step": 32820 }, { "grad_norm": 0.2631092369556427, "learning_rate": 6.850419319083812e-05, "loss": 0.0332, "step": 32830 }, { "grad_norm": 0.28751373291015625, "learning_rate": 6.848499072630592e-05, "loss": 0.0203, "step": 32840 }, { "grad_norm": 0.32096415758132935, "learning_rate": 6.846578510319439e-05, "loss": 0.0282, "step": 32850 }, { "grad_norm": 0.205630362033844, "learning_rate": 6.844657632478519e-05, "loss": 0.0275, "step": 32860 }, { "grad_norm": 0.23061266541481018, "learning_rate": 6.842736439436063e-05, "loss": 0.0215, "step": 32870 }, { "grad_norm": 0.31398922204971313, "learning_rate": 6.84081493152035e-05, "loss": 0.0259, "step": 32880 }, { "grad_norm": 0.25638943910598755, "learning_rate": 6.83889310905971e-05, "loss": 0.02, "step": 32890 }, { "grad_norm": 0.25320613384246826, "learning_rate": 6.836970972382533e-05, "loss": 0.0318, "step": 32900 }, { "grad_norm": 0.22657738626003265, "learning_rate": 6.835048521817257e-05, "loss": 0.0304, "step": 32910 }, { "grad_norm": 0.3178202509880066, "learning_rate": 6.833125757692379e-05, "loss": 0.0285, "step": 32920 }, { "grad_norm": 0.24917744100093842, "learning_rate": 6.831202680336441e-05, "loss": 0.0209, "step": 32930 }, { "grad_norm": 0.17769919335842133, "learning_rate": 6.829279290078052e-05, "loss": 0.0223, "step": 32940 }, { "grad_norm": 0.19948500394821167, "learning_rate": 6.827355587245863e-05, "loss": 0.023, "step": 32950 }, { "grad_norm": 0.26531288027763367, "learning_rate": 6.82543157216858e-05, "loss": 0.0207, "step": 32960 }, { "grad_norm": 0.23065784573554993, "learning_rate": 6.823507245174969e-05, "loss": 0.023, "step": 32970 }, { "grad_norm": 0.2950945794582367, "learning_rate": 6.821582606593841e-05, "loss": 0.0201, "step": 32980 }, { "grad_norm": 0.3786957859992981, "learning_rate": 6.81965765675407e-05, "loss": 0.0191, "step": 32990 }, { "grad_norm": 0.21423746645450592, "learning_rate": 6.81773239598457e-05, "loss": 0.0223, "step": 33000 }, { "grad_norm": 0.20322851836681366, "learning_rate": 6.815806824614319e-05, "loss": 0.0257, "step": 33010 }, { "grad_norm": 0.25859254598617554, "learning_rate": 6.813880942972343e-05, "loss": 0.0178, "step": 33020 }, { "grad_norm": 0.1973412185907364, "learning_rate": 6.811954751387726e-05, "loss": 0.0239, "step": 33030 }, { "grad_norm": 0.2916449308395386, "learning_rate": 6.810028250189598e-05, "loss": 0.0232, "step": 33040 }, { "grad_norm": 0.3841989040374756, "learning_rate": 6.808101439707147e-05, "loss": 0.0355, "step": 33050 }, { "grad_norm": 0.21557867527008057, "learning_rate": 6.806174320269609e-05, "loss": 0.03, "step": 33060 }, { "grad_norm": 0.3019848167896271, "learning_rate": 6.804246892206281e-05, "loss": 0.0262, "step": 33070 }, { "grad_norm": 0.2532775402069092, "learning_rate": 6.802319155846506e-05, "loss": 0.035, "step": 33080 }, { "grad_norm": 0.274679571390152, "learning_rate": 6.800391111519679e-05, "loss": 0.0216, "step": 33090 }, { "grad_norm": 0.2477036565542221, "learning_rate": 6.798462759555253e-05, "loss": 0.0312, "step": 33100 }, { "grad_norm": 0.214346781373024, "learning_rate": 6.79653410028273e-05, "loss": 0.0173, "step": 33110 }, { "grad_norm": 0.24875222146511078, "learning_rate": 6.794605134031663e-05, "loss": 0.0242, "step": 33120 }, { "grad_norm": 0.27596181631088257, "learning_rate": 6.792675861131661e-05, "loss": 0.0241, "step": 33130 }, { "grad_norm": 0.34544309973716736, "learning_rate": 6.790746281912386e-05, "loss": 0.0357, "step": 33140 }, { "grad_norm": 0.24232037365436554, "learning_rate": 6.788816396703546e-05, "loss": 0.0178, "step": 33150 }, { "grad_norm": 0.2726172208786011, "learning_rate": 6.78688620583491e-05, "loss": 0.0209, "step": 33160 }, { "grad_norm": 0.2729996144771576, "learning_rate": 6.784955709636292e-05, "loss": 0.0235, "step": 33170 }, { "grad_norm": 0.33749428391456604, "learning_rate": 6.783024908437564e-05, "loss": 0.026, "step": 33180 }, { "grad_norm": 0.24935734272003174, "learning_rate": 6.781093802568641e-05, "loss": 0.018, "step": 33190 }, { "grad_norm": 0.2672741711139679, "learning_rate": 6.779162392359504e-05, "loss": 0.02, "step": 33200 }, { "grad_norm": 0.45695897936820984, "learning_rate": 6.777230678140172e-05, "loss": 0.027, "step": 33210 }, { "grad_norm": 0.247549369931221, "learning_rate": 6.775298660240726e-05, "loss": 0.0247, "step": 33220 }, { "grad_norm": 0.21314950287342072, "learning_rate": 6.773366338991292e-05, "loss": 0.0355, "step": 33230 }, { "grad_norm": 0.26246294379234314, "learning_rate": 6.771433714722052e-05, "loss": 0.0252, "step": 33240 }, { "grad_norm": 0.2993510067462921, "learning_rate": 6.769500787763239e-05, "loss": 0.0308, "step": 33250 }, { "grad_norm": 0.21662692725658417, "learning_rate": 6.76756755844514e-05, "loss": 0.0232, "step": 33260 }, { "grad_norm": 0.24268431961536407, "learning_rate": 6.765634027098087e-05, "loss": 0.023, "step": 33270 }, { "grad_norm": 0.2301531434059143, "learning_rate": 6.763700194052468e-05, "loss": 0.0261, "step": 33280 }, { "grad_norm": 0.2541508078575134, "learning_rate": 6.761766059638723e-05, "loss": 0.0198, "step": 33290 }, { "grad_norm": 0.23340795934200287, "learning_rate": 6.759831624187345e-05, "loss": 0.0221, "step": 33300 }, { "grad_norm": 0.2671850025653839, "learning_rate": 6.757896888028871e-05, "loss": 0.0275, "step": 33310 }, { "grad_norm": 0.17439737915992737, "learning_rate": 6.7559618514939e-05, "loss": 0.0209, "step": 33320 }, { "grad_norm": 0.24851644039154053, "learning_rate": 6.754026514913073e-05, "loss": 0.0203, "step": 33330 }, { "grad_norm": 0.16004431247711182, "learning_rate": 6.752090878617087e-05, "loss": 0.0293, "step": 33340 }, { "grad_norm": 0.519420325756073, "learning_rate": 6.75015494293669e-05, "loss": 0.0238, "step": 33350 }, { "grad_norm": 0.285188764333725, "learning_rate": 6.74821870820268e-05, "loss": 0.0235, "step": 33360 }, { "grad_norm": 0.19611941277980804, "learning_rate": 6.746282174745907e-05, "loss": 0.0323, "step": 33370 }, { "grad_norm": 0.31616130471229553, "learning_rate": 6.744345342897271e-05, "loss": 0.0254, "step": 33380 }, { "grad_norm": 0.34559136629104614, "learning_rate": 6.742408212987724e-05, "loss": 0.0254, "step": 33390 }, { "grad_norm": 0.2767956256866455, "learning_rate": 6.740470785348269e-05, "loss": 0.0202, "step": 33400 }, { "grad_norm": 0.1795799285173416, "learning_rate": 6.738533060309958e-05, "loss": 0.0243, "step": 33410 }, { "grad_norm": 0.24488948285579681, "learning_rate": 6.736595038203894e-05, "loss": 0.021, "step": 33420 }, { "grad_norm": 0.26582685112953186, "learning_rate": 6.734656719361236e-05, "loss": 0.0253, "step": 33430 }, { "grad_norm": 0.2971818745136261, "learning_rate": 6.732718104113189e-05, "loss": 0.0261, "step": 33440 }, { "grad_norm": 0.4491939842700958, "learning_rate": 6.730779192791006e-05, "loss": 0.0187, "step": 33450 }, { "grad_norm": 0.2827315926551819, "learning_rate": 6.728839985725997e-05, "loss": 0.0219, "step": 33460 }, { "grad_norm": 0.2136816531419754, "learning_rate": 6.726900483249517e-05, "loss": 0.0206, "step": 33470 }, { "grad_norm": 0.3595332205295563, "learning_rate": 6.724960685692976e-05, "loss": 0.0239, "step": 33480 }, { "grad_norm": 0.22020164132118225, "learning_rate": 6.723020593387833e-05, "loss": 0.018, "step": 33490 }, { "grad_norm": 0.26250460743904114, "learning_rate": 6.721080206665593e-05, "loss": 0.0267, "step": 33500 }, { "grad_norm": 0.18309557437896729, "learning_rate": 6.719139525857819e-05, "loss": 0.0299, "step": 33510 }, { "grad_norm": 0.20218779146671295, "learning_rate": 6.717198551296117e-05, "loss": 0.0192, "step": 33520 }, { "grad_norm": 0.2923591732978821, "learning_rate": 6.715257283312148e-05, "loss": 0.0258, "step": 33530 }, { "grad_norm": 0.22694210708141327, "learning_rate": 6.713315722237623e-05, "loss": 0.0255, "step": 33540 }, { "grad_norm": 0.3061913847923279, "learning_rate": 6.7113738684043e-05, "loss": 0.03, "step": 33550 }, { "grad_norm": 0.24044877290725708, "learning_rate": 6.709431722143989e-05, "loss": 0.0266, "step": 33560 }, { "grad_norm": 0.25065362453460693, "learning_rate": 6.70748928378855e-05, "loss": 0.0286, "step": 33570 }, { "grad_norm": 0.19706781208515167, "learning_rate": 6.705546553669891e-05, "loss": 0.0173, "step": 33580 }, { "grad_norm": 0.17409130930900574, "learning_rate": 6.703603532119974e-05, "loss": 0.0168, "step": 33590 }, { "grad_norm": 0.22750146687030792, "learning_rate": 6.701660219470808e-05, "loss": 0.0239, "step": 33600 }, { "grad_norm": 0.2515023648738861, "learning_rate": 6.69971661605445e-05, "loss": 0.021, "step": 33610 }, { "grad_norm": 0.3506881594657898, "learning_rate": 6.697772722203008e-05, "loss": 0.0225, "step": 33620 }, { "grad_norm": 0.3154278099536896, "learning_rate": 6.695828538248643e-05, "loss": 0.0223, "step": 33630 }, { "grad_norm": 0.26635780930519104, "learning_rate": 6.693884064523563e-05, "loss": 0.0223, "step": 33640 }, { "grad_norm": 0.24209655821323395, "learning_rate": 6.691939301360023e-05, "loss": 0.0195, "step": 33650 }, { "grad_norm": 0.23056064546108246, "learning_rate": 6.689994249090333e-05, "loss": 0.0236, "step": 33660 }, { "grad_norm": 0.3661583662033081, "learning_rate": 6.688048908046845e-05, "loss": 0.0239, "step": 33670 }, { "grad_norm": 0.16008257865905762, "learning_rate": 6.686103278561969e-05, "loss": 0.0171, "step": 33680 }, { "grad_norm": 0.25537973642349243, "learning_rate": 6.684157360968156e-05, "loss": 0.021, "step": 33690 }, { "grad_norm": 0.3315347731113434, "learning_rate": 6.682211155597911e-05, "loss": 0.0179, "step": 33700 }, { "grad_norm": 0.21876423060894012, "learning_rate": 6.680264662783789e-05, "loss": 0.0215, "step": 33710 }, { "grad_norm": 0.3088197708129883, "learning_rate": 6.678317882858391e-05, "loss": 0.0223, "step": 33720 }, { "grad_norm": 0.26745566725730896, "learning_rate": 6.67637081615437e-05, "loss": 0.0199, "step": 33730 }, { "grad_norm": 0.22957147657871246, "learning_rate": 6.674423463004427e-05, "loss": 0.0159, "step": 33740 }, { "grad_norm": 0.214297816157341, "learning_rate": 6.672475823741308e-05, "loss": 0.0184, "step": 33750 }, { "grad_norm": 0.21683906018733978, "learning_rate": 6.670527898697811e-05, "loss": 0.0222, "step": 33760 }, { "grad_norm": 0.22826553881168365, "learning_rate": 6.668579688206788e-05, "loss": 0.024, "step": 33770 }, { "grad_norm": 0.2056548297405243, "learning_rate": 6.666631192601131e-05, "loss": 0.0185, "step": 33780 }, { "grad_norm": 0.19767306745052338, "learning_rate": 6.664682412213785e-05, "loss": 0.0235, "step": 33790 }, { "grad_norm": 0.32713860273361206, "learning_rate": 6.662733347377745e-05, "loss": 0.021, "step": 33800 }, { "grad_norm": 0.15183305740356445, "learning_rate": 6.660783998426051e-05, "loss": 0.0165, "step": 33810 }, { "grad_norm": 0.1932295560836792, "learning_rate": 6.658834365691794e-05, "loss": 0.018, "step": 33820 }, { "grad_norm": 0.1712164729833603, "learning_rate": 6.656884449508115e-05, "loss": 0.0171, "step": 33830 }, { "grad_norm": 0.2305416613817215, "learning_rate": 6.654934250208198e-05, "loss": 0.0176, "step": 33840 }, { "grad_norm": 0.27242311835289, "learning_rate": 6.65298376812528e-05, "loss": 0.02, "step": 33850 }, { "grad_norm": 0.29107218980789185, "learning_rate": 6.651033003592646e-05, "loss": 0.0207, "step": 33860 }, { "grad_norm": 0.3461274206638336, "learning_rate": 6.649081956943626e-05, "loss": 0.0184, "step": 33870 }, { "grad_norm": 0.2965593636035919, "learning_rate": 6.647130628511604e-05, "loss": 0.024, "step": 33880 }, { "grad_norm": 0.3060731887817383, "learning_rate": 6.645179018630005e-05, "loss": 0.021, "step": 33890 }, { "grad_norm": 0.238896444439888, "learning_rate": 6.643227127632309e-05, "loss": 0.0169, "step": 33900 }, { "grad_norm": 0.6503981947898865, "learning_rate": 6.641274955852038e-05, "loss": 0.0253, "step": 33910 }, { "grad_norm": 0.3628208637237549, "learning_rate": 6.639322503622768e-05, "loss": 0.0414, "step": 33920 }, { "grad_norm": 0.24925215542316437, "learning_rate": 6.637369771278116e-05, "loss": 0.021, "step": 33930 }, { "grad_norm": 0.27362167835235596, "learning_rate": 6.635416759151751e-05, "loss": 0.017, "step": 33940 }, { "grad_norm": 0.2820255756378174, "learning_rate": 6.633463467577394e-05, "loss": 0.0412, "step": 33950 }, { "grad_norm": 0.28432363271713257, "learning_rate": 6.631509896888803e-05, "loss": 0.0238, "step": 33960 }, { "grad_norm": 0.32823434472084045, "learning_rate": 6.629556047419794e-05, "loss": 0.0188, "step": 33970 }, { "grad_norm": 0.22790876030921936, "learning_rate": 6.627601919504223e-05, "loss": 0.0193, "step": 33980 }, { "grad_norm": 0.2212529182434082, "learning_rate": 6.625647513476001e-05, "loss": 0.0158, "step": 33990 }, { "grad_norm": 0.3001483976840973, "learning_rate": 6.62369282966908e-05, "loss": 0.0272, "step": 34000 }, { "grad_norm": 0.36297374963760376, "learning_rate": 6.621737868417464e-05, "loss": 0.0198, "step": 34010 }, { "grad_norm": 0.2505508065223694, "learning_rate": 6.619782630055198e-05, "loss": 0.0234, "step": 34020 }, { "grad_norm": 0.21586260199546814, "learning_rate": 6.617827114916382e-05, "loss": 0.0204, "step": 34030 }, { "grad_norm": 0.34126242995262146, "learning_rate": 6.615871323335161e-05, "loss": 0.0234, "step": 34040 }, { "grad_norm": 0.29915767908096313, "learning_rate": 6.613915255645725e-05, "loss": 0.0239, "step": 34050 }, { "grad_norm": 0.27937936782836914, "learning_rate": 6.611958912182312e-05, "loss": 0.0219, "step": 34060 }, { "grad_norm": 0.2312026172876358, "learning_rate": 6.610002293279207e-05, "loss": 0.0225, "step": 34070 }, { "grad_norm": 0.16458836197853088, "learning_rate": 6.608045399270746e-05, "loss": 0.0191, "step": 34080 }, { "grad_norm": 0.312565416097641, "learning_rate": 6.606088230491304e-05, "loss": 0.0156, "step": 34090 }, { "grad_norm": 0.3073199689388275, "learning_rate": 6.604130787275312e-05, "loss": 0.026, "step": 34100 }, { "grad_norm": 0.2720942795276642, "learning_rate": 6.602173069957242e-05, "loss": 0.0194, "step": 34110 }, { "grad_norm": 0.3165294826030731, "learning_rate": 6.600215078871612e-05, "loss": 0.0184, "step": 34120 }, { "grad_norm": 0.172150656580925, "learning_rate": 6.598256814352992e-05, "loss": 0.0218, "step": 34130 }, { "grad_norm": 0.2630215287208557, "learning_rate": 6.596298276735995e-05, "loss": 0.0152, "step": 34140 }, { "grad_norm": 0.22991929948329926, "learning_rate": 6.594339466355282e-05, "loss": 0.02, "step": 34150 }, { "grad_norm": 0.28246915340423584, "learning_rate": 6.592380383545558e-05, "loss": 0.021, "step": 34160 }, { "grad_norm": 0.25698190927505493, "learning_rate": 6.590421028641577e-05, "loss": 0.0301, "step": 34170 }, { "grad_norm": 0.28256940841674805, "learning_rate": 6.588461401978143e-05, "loss": 0.0263, "step": 34180 }, { "grad_norm": 0.3296908140182495, "learning_rate": 6.586501503890099e-05, "loss": 0.024, "step": 34190 }, { "grad_norm": 0.27541103959083557, "learning_rate": 6.584541334712338e-05, "loss": 0.0274, "step": 34200 }, { "grad_norm": 0.3347684442996979, "learning_rate": 6.582580894779802e-05, "loss": 0.0226, "step": 34210 }, { "grad_norm": 0.3840726912021637, "learning_rate": 6.580620184427473e-05, "loss": 0.0238, "step": 34220 }, { "grad_norm": 0.21305058896541595, "learning_rate": 6.578659203990385e-05, "loss": 0.0234, "step": 34230 }, { "grad_norm": 0.21440762281417847, "learning_rate": 6.576697953803615e-05, "loss": 0.0197, "step": 34240 }, { "grad_norm": 0.24368777871131897, "learning_rate": 6.57473643420229e-05, "loss": 0.0239, "step": 34250 }, { "grad_norm": 0.25387072563171387, "learning_rate": 6.572774645521574e-05, "loss": 0.0248, "step": 34260 }, { "grad_norm": 0.3674306869506836, "learning_rate": 6.570812588096688e-05, "loss": 0.0325, "step": 34270 }, { "grad_norm": 0.2545883357524872, "learning_rate": 6.568850262262893e-05, "loss": 0.0208, "step": 34280 }, { "grad_norm": 0.26425954699516296, "learning_rate": 6.566887668355497e-05, "loss": 0.0227, "step": 34290 }, { "grad_norm": 0.2289533168077469, "learning_rate": 6.564924806709851e-05, "loss": 0.0179, "step": 34300 }, { "grad_norm": 0.3228193521499634, "learning_rate": 6.562961677661359e-05, "loss": 0.0248, "step": 34310 }, { "grad_norm": 0.3660734295845032, "learning_rate": 6.56099828154546e-05, "loss": 0.0199, "step": 34320 }, { "grad_norm": 0.2763814330101013, "learning_rate": 6.55903461869765e-05, "loss": 0.0199, "step": 34330 }, { "grad_norm": 0.22162814438343048, "learning_rate": 6.557070689453465e-05, "loss": 0.0184, "step": 34340 }, { "grad_norm": 0.2764236032962799, "learning_rate": 6.555106494148482e-05, "loss": 0.0251, "step": 34350 }, { "grad_norm": 0.2868473529815674, "learning_rate": 6.553142033118333e-05, "loss": 0.0204, "step": 34360 }, { "grad_norm": 0.2484050989151001, "learning_rate": 6.551177306698688e-05, "loss": 0.0306, "step": 34370 }, { "grad_norm": 0.21848367154598236, "learning_rate": 6.549212315225267e-05, "loss": 0.0252, "step": 34380 }, { "grad_norm": 0.24074652791023254, "learning_rate": 6.547247059033833e-05, "loss": 0.0272, "step": 34390 }, { "grad_norm": 0.29173028469085693, "learning_rate": 6.545281538460193e-05, "loss": 0.0212, "step": 34400 }, { "grad_norm": 0.23952174186706543, "learning_rate": 6.543315753840202e-05, "loss": 0.0237, "step": 34410 }, { "grad_norm": 0.30672141909599304, "learning_rate": 6.541349705509758e-05, "loss": 0.0214, "step": 34420 }, { "grad_norm": 0.20512986183166504, "learning_rate": 6.539383393804805e-05, "loss": 0.0229, "step": 34430 }, { "grad_norm": 0.20076625049114227, "learning_rate": 6.537416819061333e-05, "loss": 0.0282, "step": 34440 }, { "grad_norm": 0.22221238911151886, "learning_rate": 6.535449981615375e-05, "loss": 0.0291, "step": 34450 }, { "grad_norm": 0.20719510316848755, "learning_rate": 6.53348288180301e-05, "loss": 0.021, "step": 34460 }, { "grad_norm": 0.2818510830402374, "learning_rate": 6.531515519960361e-05, "loss": 0.0259, "step": 34470 }, { "grad_norm": 0.28162866830825806, "learning_rate": 6.529547896423597e-05, "loss": 0.0187, "step": 34480 }, { "grad_norm": 0.24733713269233704, "learning_rate": 6.52758001152893e-05, "loss": 0.0244, "step": 34490 }, { "grad_norm": 0.21512863039970398, "learning_rate": 6.525611865612618e-05, "loss": 0.0197, "step": 34500 }, { "grad_norm": 0.2550695538520813, "learning_rate": 6.523643459010966e-05, "loss": 0.0202, "step": 34510 }, { "grad_norm": 0.3241499662399292, "learning_rate": 6.521674792060317e-05, "loss": 0.0221, "step": 34520 }, { "grad_norm": 0.2914218008518219, "learning_rate": 6.519705865097063e-05, "loss": 0.0255, "step": 34530 }, { "grad_norm": 0.28008952736854553, "learning_rate": 6.517736678457641e-05, "loss": 0.0168, "step": 34540 }, { "grad_norm": 0.2676158845424652, "learning_rate": 6.515767232478534e-05, "loss": 0.0215, "step": 34550 }, { "grad_norm": 0.26616889238357544, "learning_rate": 6.51379752749626e-05, "loss": 0.0165, "step": 34560 }, { "grad_norm": 0.2726342976093292, "learning_rate": 6.511827563847393e-05, "loss": 0.0206, "step": 34570 }, { "grad_norm": 0.22059500217437744, "learning_rate": 6.509857341868542e-05, "loss": 0.0246, "step": 34580 }, { "grad_norm": 0.15942531824111938, "learning_rate": 6.507886861896367e-05, "loss": 0.0206, "step": 34590 }, { "grad_norm": 0.32894405722618103, "learning_rate": 6.505916124267567e-05, "loss": 0.031, "step": 34600 }, { "grad_norm": 0.30041825771331787, "learning_rate": 6.503945129318891e-05, "loss": 0.0226, "step": 34610 }, { "grad_norm": 0.2421216070652008, "learning_rate": 6.501973877387122e-05, "loss": 0.0166, "step": 34620 }, { "grad_norm": 0.25703907012939453, "learning_rate": 6.500002368809098e-05, "loss": 0.0211, "step": 34630 }, { "grad_norm": 0.3804349899291992, "learning_rate": 6.498030603921694e-05, "loss": 0.0209, "step": 34640 }, { "grad_norm": 0.35113492608070374, "learning_rate": 6.496058583061832e-05, "loss": 0.0235, "step": 34650 }, { "grad_norm": 0.2396780401468277, "learning_rate": 6.494086306566475e-05, "loss": 0.0128, "step": 34660 }, { "grad_norm": 0.2701045572757721, "learning_rate": 6.492113774772632e-05, "loss": 0.0213, "step": 34670 }, { "grad_norm": 0.2893512547016144, "learning_rate": 6.490140988017354e-05, "loss": 0.0277, "step": 34680 }, { "grad_norm": 0.4806373715400696, "learning_rate": 6.488167946637736e-05, "loss": 0.0297, "step": 34690 }, { "grad_norm": 0.3227427899837494, "learning_rate": 6.486194650970915e-05, "loss": 0.0246, "step": 34700 }, { "grad_norm": 0.3062868118286133, "learning_rate": 6.48422110135408e-05, "loss": 0.0259, "step": 34710 }, { "grad_norm": 0.29035627841949463, "learning_rate": 6.482247298124451e-05, "loss": 0.0341, "step": 34720 }, { "grad_norm": 0.2093123495578766, "learning_rate": 6.480273241619297e-05, "loss": 0.0188, "step": 34730 }, { "grad_norm": 0.17823508381843567, "learning_rate": 6.478298932175933e-05, "loss": 0.0193, "step": 34740 }, { "grad_norm": 0.17017322778701782, "learning_rate": 6.476324370131712e-05, "loss": 0.0209, "step": 34750 }, { "grad_norm": 0.304412841796875, "learning_rate": 6.474349555824036e-05, "loss": 0.0211, "step": 34760 }, { "grad_norm": 0.18416571617126465, "learning_rate": 6.472374489590342e-05, "loss": 0.0158, "step": 34770 }, { "grad_norm": 0.2077171802520752, "learning_rate": 6.470399171768118e-05, "loss": 0.0345, "step": 34780 }, { "grad_norm": 0.3814149498939514, "learning_rate": 6.468423602694891e-05, "loss": 0.0251, "step": 34790 }, { "grad_norm": 0.5636371970176697, "learning_rate": 6.466447782708232e-05, "loss": 0.0167, "step": 34800 }, { "grad_norm": 0.283805251121521, "learning_rate": 6.464471712145754e-05, "loss": 0.0189, "step": 34810 }, { "grad_norm": 0.20815761387348175, "learning_rate": 6.462495391345114e-05, "loss": 0.0221, "step": 34820 }, { "grad_norm": 0.2039923369884491, "learning_rate": 6.46051882064401e-05, "loss": 0.0172, "step": 34830 }, { "grad_norm": 0.23937559127807617, "learning_rate": 6.458542000380186e-05, "loss": 0.0189, "step": 34840 }, { "grad_norm": 0.3626222014427185, "learning_rate": 6.456564930891424e-05, "loss": 0.0235, "step": 34850 }, { "grad_norm": 0.3392312526702881, "learning_rate": 6.454587612515555e-05, "loss": 0.0284, "step": 34860 }, { "grad_norm": 0.2197856605052948, "learning_rate": 6.452610045590444e-05, "loss": 0.0212, "step": 34870 }, { "grad_norm": 0.2542872428894043, "learning_rate": 6.450632230454005e-05, "loss": 0.0192, "step": 34880 }, { "grad_norm": 0.2905537784099579, "learning_rate": 6.448654167444195e-05, "loss": 0.0303, "step": 34890 }, { "grad_norm": 0.2763957381248474, "learning_rate": 6.446675856899005e-05, "loss": 0.0185, "step": 34900 }, { "grad_norm": 0.18977709114551544, "learning_rate": 6.444697299156481e-05, "loss": 0.0221, "step": 34910 }, { "grad_norm": 0.30118197202682495, "learning_rate": 6.442718494554701e-05, "loss": 0.0219, "step": 34920 }, { "grad_norm": 0.2653128504753113, "learning_rate": 6.440739443431787e-05, "loss": 0.0158, "step": 34930 }, { "grad_norm": 0.22594580054283142, "learning_rate": 6.438760146125906e-05, "loss": 0.0183, "step": 34940 }, { "grad_norm": 0.3289646506309509, "learning_rate": 6.436780602975267e-05, "loss": 0.0245, "step": 34950 }, { "grad_norm": 0.3143194615840912, "learning_rate": 6.43480081431812e-05, "loss": 0.0262, "step": 34960 }, { "grad_norm": 0.29770469665527344, "learning_rate": 6.432820780492756e-05, "loss": 0.0219, "step": 34970 }, { "grad_norm": 0.3187873959541321, "learning_rate": 6.430840501837506e-05, "loss": 0.0282, "step": 34980 }, { "grad_norm": 0.3857175409793854, "learning_rate": 6.428859978690748e-05, "loss": 0.0231, "step": 34990 }, { "grad_norm": 0.24064728617668152, "learning_rate": 6.426879211390901e-05, "loss": 0.0222, "step": 35000 }, { "grad_norm": 0.2115050107240677, "learning_rate": 6.424898200276422e-05, "loss": 0.0222, "step": 35010 }, { "grad_norm": 0.20583443343639374, "learning_rate": 6.42291694568581e-05, "loss": 0.0194, "step": 35020 }, { "grad_norm": 0.3847073018550873, "learning_rate": 6.42093544795761e-05, "loss": 0.0279, "step": 35030 }, { "grad_norm": 0.22916395962238312, "learning_rate": 6.418953707430403e-05, "loss": 0.0231, "step": 35040 }, { "grad_norm": 0.25757989287376404, "learning_rate": 6.416971724442819e-05, "loss": 0.0206, "step": 35050 }, { "grad_norm": 0.27354341745376587, "learning_rate": 6.414989499333519e-05, "loss": 0.0205, "step": 35060 }, { "grad_norm": 0.23288525640964508, "learning_rate": 6.413007032441214e-05, "loss": 0.0201, "step": 35070 }, { "grad_norm": 0.5188122987747192, "learning_rate": 6.411024324104653e-05, "loss": 0.0215, "step": 35080 }, { "grad_norm": 0.24443311989307404, "learning_rate": 6.409041374662628e-05, "loss": 0.0199, "step": 35090 }, { "grad_norm": 0.26058217883110046, "learning_rate": 6.407058184453967e-05, "loss": 0.0324, "step": 35100 }, { "grad_norm": 0.2784704864025116, "learning_rate": 6.405074753817548e-05, "loss": 0.0222, "step": 35110 }, { "grad_norm": 0.2462017983198166, "learning_rate": 6.40309108309228e-05, "loss": 0.0214, "step": 35120 }, { "grad_norm": 0.24910417199134827, "learning_rate": 6.401107172617122e-05, "loss": 0.0216, "step": 35130 }, { "grad_norm": 0.321469783782959, "learning_rate": 6.399123022731068e-05, "loss": 0.0455, "step": 35140 }, { "grad_norm": 0.28180310130119324, "learning_rate": 6.397138633773157e-05, "loss": 0.0195, "step": 35150 }, { "grad_norm": 0.33125558495521545, "learning_rate": 6.395154006082463e-05, "loss": 0.0195, "step": 35160 }, { "grad_norm": 0.20474287867546082, "learning_rate": 6.393169139998109e-05, "loss": 0.0187, "step": 35170 }, { "grad_norm": 0.24138516187667847, "learning_rate": 6.39118403585925e-05, "loss": 0.0197, "step": 35180 }, { "grad_norm": 0.25334009528160095, "learning_rate": 6.38919869400509e-05, "loss": 0.022, "step": 35190 }, { "grad_norm": 0.18865863978862762, "learning_rate": 6.387213114774865e-05, "loss": 0.0227, "step": 35200 }, { "grad_norm": 0.21338750422000885, "learning_rate": 6.385227298507863e-05, "loss": 0.0405, "step": 35210 }, { "grad_norm": 0.28932076692581177, "learning_rate": 6.3832412455434e-05, "loss": 0.0186, "step": 35220 }, { "grad_norm": 0.2648712992668152, "learning_rate": 6.381254956220841e-05, "loss": 0.02, "step": 35230 }, { "grad_norm": 0.18825885653495789, "learning_rate": 6.379268430879586e-05, "loss": 0.0199, "step": 35240 }, { "grad_norm": 0.21709007024765015, "learning_rate": 6.37728166985908e-05, "loss": 0.0188, "step": 35250 }, { "grad_norm": 0.25743404030799866, "learning_rate": 6.375294673498804e-05, "loss": 0.0167, "step": 35260 }, { "grad_norm": 0.2747577428817749, "learning_rate": 6.373307442138284e-05, "loss": 0.0204, "step": 35270 }, { "grad_norm": 0.2699142396450043, "learning_rate": 6.371319976117081e-05, "loss": 0.0329, "step": 35280 }, { "grad_norm": 0.2638385593891144, "learning_rate": 6.3693322757748e-05, "loss": 0.0188, "step": 35290 }, { "grad_norm": 0.29026997089385986, "learning_rate": 6.367344341451086e-05, "loss": 0.0213, "step": 35300 }, { "grad_norm": 0.29840657114982605, "learning_rate": 6.36535617348562e-05, "loss": 0.0235, "step": 35310 }, { "grad_norm": 0.3924548923969269, "learning_rate": 6.363367772218128e-05, "loss": 0.0224, "step": 35320 }, { "grad_norm": 0.3113507628440857, "learning_rate": 6.36137913798837e-05, "loss": 0.021, "step": 35330 }, { "grad_norm": 0.26722607016563416, "learning_rate": 6.359390271136151e-05, "loss": 0.0156, "step": 35340 }, { "grad_norm": 0.28901323676109314, "learning_rate": 6.357401172001314e-05, "loss": 0.0195, "step": 35350 }, { "grad_norm": 0.1851823776960373, "learning_rate": 6.355411840923742e-05, "loss": 0.0223, "step": 35360 }, { "grad_norm": 0.3646584153175354, "learning_rate": 6.353422278243358e-05, "loss": 0.0201, "step": 35370 }, { "grad_norm": 0.31353992223739624, "learning_rate": 6.351432484300121e-05, "loss": 0.0217, "step": 35380 }, { "grad_norm": 0.3194972276687622, "learning_rate": 6.349442459434036e-05, "loss": 0.0236, "step": 35390 }, { "grad_norm": 0.29161831736564636, "learning_rate": 6.34745220398514e-05, "loss": 0.0284, "step": 35400 }, { "grad_norm": 0.2668501138687134, "learning_rate": 6.345461718293518e-05, "loss": 0.0215, "step": 35410 }, { "grad_norm": 0.2896275520324707, "learning_rate": 6.343471002699286e-05, "loss": 0.0295, "step": 35420 }, { "grad_norm": 0.2733707129955292, "learning_rate": 6.341480057542602e-05, "loss": 0.0287, "step": 35430 }, { "grad_norm": 0.2487621307373047, "learning_rate": 6.339488883163667e-05, "loss": 0.0223, "step": 35440 }, { "grad_norm": 0.2390100061893463, "learning_rate": 6.337497479902716e-05, "loss": 0.0199, "step": 35450 }, { "grad_norm": 0.22827690839767456, "learning_rate": 6.335505848100027e-05, "loss": 0.0237, "step": 35460 }, { "grad_norm": 0.21971780061721802, "learning_rate": 6.333513988095915e-05, "loss": 0.0188, "step": 35470 }, { "grad_norm": 0.2974616587162018, "learning_rate": 6.331521900230735e-05, "loss": 0.0178, "step": 35480 }, { "grad_norm": 0.1849430501461029, "learning_rate": 6.329529584844878e-05, "loss": 0.0156, "step": 35490 }, { "grad_norm": 0.29374539852142334, "learning_rate": 6.327537042278777e-05, "loss": 0.0218, "step": 35500 }, { "grad_norm": 0.2715968191623688, "learning_rate": 6.325544272872905e-05, "loss": 0.0189, "step": 35510 }, { "grad_norm": 0.2568906545639038, "learning_rate": 6.323551276967771e-05, "loss": 0.0264, "step": 35520 }, { "grad_norm": 0.2475537657737732, "learning_rate": 6.321558054903922e-05, "loss": 0.0173, "step": 35530 }, { "grad_norm": 0.18264833092689514, "learning_rate": 6.319564607021947e-05, "loss": 0.0228, "step": 35540 }, { "grad_norm": 0.20996041595935822, "learning_rate": 6.31757093366247e-05, "loss": 0.016, "step": 35550 }, { "grad_norm": 0.38499224185943604, "learning_rate": 6.315577035166154e-05, "loss": 0.0209, "step": 35560 }, { "grad_norm": 0.2556651830673218, "learning_rate": 6.313582911873708e-05, "loss": 0.0176, "step": 35570 }, { "grad_norm": 0.30996304750442505, "learning_rate": 6.311588564125865e-05, "loss": 0.0188, "step": 35580 }, { "grad_norm": 0.3868354558944702, "learning_rate": 6.30959399226341e-05, "loss": 0.0284, "step": 35590 }, { "grad_norm": 0.2504328191280365, "learning_rate": 6.30759919662716e-05, "loss": 0.0362, "step": 35600 }, { "grad_norm": 0.24476437270641327, "learning_rate": 6.30560417755797e-05, "loss": 0.0189, "step": 35610 }, { "grad_norm": 0.2958660423755646, "learning_rate": 6.303608935396735e-05, "loss": 0.0176, "step": 35620 }, { "grad_norm": 0.24065659940242767, "learning_rate": 6.301613470484386e-05, "loss": 0.0178, "step": 35630 }, { "grad_norm": 0.22139036655426025, "learning_rate": 6.299617783161893e-05, "loss": 0.0212, "step": 35640 }, { "grad_norm": 0.29976600408554077, "learning_rate": 6.297621873770266e-05, "loss": 0.0216, "step": 35650 }, { "grad_norm": 0.19295604526996613, "learning_rate": 6.29562574265055e-05, "loss": 0.0198, "step": 35660 }, { "grad_norm": 0.24054773151874542, "learning_rate": 6.293629390143834e-05, "loss": 0.0202, "step": 35670 }, { "grad_norm": 0.1897367686033249, "learning_rate": 6.291632816591232e-05, "loss": 0.0157, "step": 35680 }, { "grad_norm": 0.1948748677968979, "learning_rate": 6.28963602233391e-05, "loss": 0.0179, "step": 35690 }, { "grad_norm": 0.27793189883232117, "learning_rate": 6.287639007713062e-05, "loss": 0.02, "step": 35700 }, { "grad_norm": 0.1844714730978012, "learning_rate": 6.285641773069926e-05, "loss": 0.0183, "step": 35710 }, { "grad_norm": 0.22812527418136597, "learning_rate": 6.283644318745773e-05, "loss": 0.0179, "step": 35720 }, { "grad_norm": 0.199338898062706, "learning_rate": 6.281646645081912e-05, "loss": 0.0183, "step": 35730 }, { "grad_norm": 0.19057868421077728, "learning_rate": 6.279648752419693e-05, "loss": 0.016, "step": 35740 }, { "grad_norm": 0.21941758692264557, "learning_rate": 6.2776506411005e-05, "loss": 0.0222, "step": 35750 }, { "grad_norm": 0.1989767849445343, "learning_rate": 6.275652311465758e-05, "loss": 0.0173, "step": 35760 }, { "grad_norm": 0.2885679602622986, "learning_rate": 6.273653763856926e-05, "loss": 0.0251, "step": 35770 }, { "grad_norm": 0.31147781014442444, "learning_rate": 6.271654998615501e-05, "loss": 0.0246, "step": 35780 }, { "grad_norm": 0.22000432014465332, "learning_rate": 6.269656016083013e-05, "loss": 0.0177, "step": 35790 }, { "grad_norm": 0.16674643754959106, "learning_rate": 6.267656816601038e-05, "loss": 0.0242, "step": 35800 }, { "grad_norm": 0.3520413339138031, "learning_rate": 6.265657400511185e-05, "loss": 0.024, "step": 35810 }, { "grad_norm": 0.19131411612033844, "learning_rate": 6.263657768155098e-05, "loss": 0.0177, "step": 35820 }, { "grad_norm": 0.2555549144744873, "learning_rate": 6.261657919874457e-05, "loss": 0.0183, "step": 35830 }, { "grad_norm": 0.23641441762447357, "learning_rate": 6.259657856010986e-05, "loss": 0.0244, "step": 35840 }, { "grad_norm": 0.362714946269989, "learning_rate": 6.257657576906439e-05, "loss": 0.0194, "step": 35850 }, { "grad_norm": 0.22971320152282715, "learning_rate": 6.255657082902609e-05, "loss": 0.0165, "step": 35860 }, { "grad_norm": 0.2599324584007263, "learning_rate": 6.253656374341325e-05, "loss": 0.0193, "step": 35870 }, { "grad_norm": 0.37982845306396484, "learning_rate": 6.251655451564457e-05, "loss": 0.0185, "step": 35880 }, { "grad_norm": 0.3310100734233856, "learning_rate": 6.249654314913902e-05, "loss": 0.0191, "step": 35890 }, { "grad_norm": 0.24706806242465973, "learning_rate": 6.247652964731604e-05, "loss": 0.0195, "step": 35900 }, { "grad_norm": 0.23979507386684418, "learning_rate": 6.245651401359537e-05, "loss": 0.0293, "step": 35910 }, { "grad_norm": 0.20646996796131134, "learning_rate": 6.243649625139715e-05, "loss": 0.0219, "step": 35920 }, { "grad_norm": 0.24978378415107727, "learning_rate": 6.241647636414185e-05, "loss": 0.0196, "step": 35930 }, { "grad_norm": 0.26291295886039734, "learning_rate": 6.239645435525034e-05, "loss": 0.0273, "step": 35940 }, { "grad_norm": 0.17989081144332886, "learning_rate": 6.237643022814381e-05, "loss": 0.0257, "step": 35950 }, { "grad_norm": 0.2286847084760666, "learning_rate": 6.235640398624386e-05, "loss": 0.0179, "step": 35960 }, { "grad_norm": 0.2581076920032501, "learning_rate": 6.233637563297243e-05, "loss": 0.0256, "step": 35970 }, { "grad_norm": 0.2449878752231598, "learning_rate": 6.23163451717518e-05, "loss": 0.0246, "step": 35980 }, { "grad_norm": 0.2240878790616989, "learning_rate": 6.229631260600463e-05, "loss": 0.0191, "step": 35990 }, { "grad_norm": 0.22469152510166168, "learning_rate": 6.227627793915392e-05, "loss": 0.0266, "step": 36000 }, { "grad_norm": 0.29770904779434204, "learning_rate": 6.225624117462309e-05, "loss": 0.0167, "step": 36010 }, { "grad_norm": 0.24077582359313965, "learning_rate": 6.223620231583586e-05, "loss": 0.0188, "step": 36020 }, { "grad_norm": 0.1921851634979248, "learning_rate": 6.221616136621629e-05, "loss": 0.0228, "step": 36030 }, { "grad_norm": 0.25967904925346375, "learning_rate": 6.219611832918887e-05, "loss": 0.0217, "step": 36040 }, { "grad_norm": 0.2071543037891388, "learning_rate": 6.217607320817838e-05, "loss": 0.018, "step": 36050 }, { "grad_norm": 0.216598778963089, "learning_rate": 6.215602600661001e-05, "loss": 0.0188, "step": 36060 }, { "grad_norm": 0.3521133065223694, "learning_rate": 6.213597672790925e-05, "loss": 0.0177, "step": 36070 }, { "grad_norm": 0.20949587225914001, "learning_rate": 6.2115925375502e-05, "loss": 0.0226, "step": 36080 }, { "grad_norm": 0.25498002767562866, "learning_rate": 6.209587195281447e-05, "loss": 0.0216, "step": 36090 }, { "grad_norm": 0.2494080811738968, "learning_rate": 6.207581646327324e-05, "loss": 0.0233, "step": 36100 }, { "grad_norm": 0.24858522415161133, "learning_rate": 6.205575891030526e-05, "loss": 0.0138, "step": 36110 }, { "grad_norm": 0.17314653098583221, "learning_rate": 6.203569929733781e-05, "loss": 0.0154, "step": 36120 }, { "grad_norm": 0.2744801342487335, "learning_rate": 6.201563762779852e-05, "loss": 0.0188, "step": 36130 }, { "grad_norm": 0.2789156138896942, "learning_rate": 6.199557390511538e-05, "loss": 0.0188, "step": 36140 }, { "grad_norm": 0.204999178647995, "learning_rate": 6.197550813271675e-05, "loss": 0.0156, "step": 36150 }, { "grad_norm": 0.242840975522995, "learning_rate": 6.195544031403131e-05, "loss": 0.0223, "step": 36160 }, { "grad_norm": 0.3982100784778595, "learning_rate": 6.19353704524881e-05, "loss": 0.0227, "step": 36170 }, { "grad_norm": 0.251899778842926, "learning_rate": 6.191529855151652e-05, "loss": 0.0156, "step": 36180 }, { "grad_norm": 0.25731903314590454, "learning_rate": 6.189522461454629e-05, "loss": 0.0242, "step": 36190 }, { "grad_norm": 0.25085631012916565, "learning_rate": 6.187514864500752e-05, "loss": 0.0185, "step": 36200 }, { "grad_norm": 0.4363897144794464, "learning_rate": 6.185507064633062e-05, "loss": 0.0266, "step": 36210 }, { "grad_norm": 0.31622588634490967, "learning_rate": 6.18349906219464e-05, "loss": 0.0243, "step": 36220 }, { "grad_norm": 0.2823767066001892, "learning_rate": 6.181490857528596e-05, "loss": 0.0194, "step": 36230 }, { "grad_norm": 0.22741387784481049, "learning_rate": 6.179482450978077e-05, "loss": 0.0128, "step": 36240 }, { "grad_norm": 0.23392392694950104, "learning_rate": 6.177473842886269e-05, "loss": 0.0188, "step": 36250 }, { "grad_norm": 0.1645856648683548, "learning_rate": 6.175465033596382e-05, "loss": 0.0206, "step": 36260 }, { "grad_norm": 0.24930140376091003, "learning_rate": 6.173456023451671e-05, "loss": 0.0201, "step": 36270 }, { "grad_norm": 0.3370155096054077, "learning_rate": 6.171446812795422e-05, "loss": 0.0223, "step": 36280 }, { "grad_norm": 0.20811717212200165, "learning_rate": 6.169437401970949e-05, "loss": 0.0314, "step": 36290 }, { "grad_norm": 0.1874491572380066, "learning_rate": 6.16742779132161e-05, "loss": 0.0134, "step": 36300 }, { "grad_norm": 0.2709580659866333, "learning_rate": 6.165417981190789e-05, "loss": 0.0171, "step": 36310 }, { "grad_norm": 0.20498201251029968, "learning_rate": 6.16340797192191e-05, "loss": 0.017, "step": 36320 }, { "grad_norm": 0.226202130317688, "learning_rate": 6.161397763858427e-05, "loss": 0.018, "step": 36330 }, { "grad_norm": 0.26703548431396484, "learning_rate": 6.159387357343834e-05, "loss": 0.0122, "step": 36340 }, { "grad_norm": 0.26240959763526917, "learning_rate": 6.157376752721648e-05, "loss": 0.0212, "step": 36350 }, { "grad_norm": 0.20208123326301575, "learning_rate": 6.155365950335428e-05, "loss": 0.0204, "step": 36360 }, { "grad_norm": 0.27392151951789856, "learning_rate": 6.153354950528768e-05, "loss": 0.0159, "step": 36370 }, { "grad_norm": 0.2897857427597046, "learning_rate": 6.151343753645293e-05, "loss": 0.0224, "step": 36380 }, { "grad_norm": 0.2264576256275177, "learning_rate": 6.149332360028657e-05, "loss": 0.0228, "step": 36390 }, { "grad_norm": 0.27509605884552, "learning_rate": 6.147320770022555e-05, "loss": 0.0217, "step": 36400 }, { "grad_norm": 0.2605011761188507, "learning_rate": 6.145308983970715e-05, "loss": 0.0264, "step": 36410 }, { "grad_norm": 0.2571014165878296, "learning_rate": 6.143297002216892e-05, "loss": 0.0192, "step": 36420 }, { "grad_norm": 0.2453702688217163, "learning_rate": 6.141284825104882e-05, "loss": 0.0221, "step": 36430 }, { "grad_norm": 0.2120192050933838, "learning_rate": 6.13927245297851e-05, "loss": 0.0274, "step": 36440 }, { "grad_norm": 0.19342930614948273, "learning_rate": 6.137259886181633e-05, "loss": 0.0244, "step": 36450 }, { "grad_norm": 0.265868604183197, "learning_rate": 6.135247125058145e-05, "loss": 0.0244, "step": 36460 }, { "grad_norm": 0.29177820682525635, "learning_rate": 6.133234169951974e-05, "loss": 0.0165, "step": 36470 }, { "grad_norm": 0.34895169734954834, "learning_rate": 6.131221021207078e-05, "loss": 0.0242, "step": 36480 }, { "grad_norm": 0.19929847121238708, "learning_rate": 6.129207679167448e-05, "loss": 0.0179, "step": 36490 }, { "grad_norm": 0.3293280303478241, "learning_rate": 6.127194144177109e-05, "loss": 0.016, "step": 36500 }, { "grad_norm": 0.26503267884254456, "learning_rate": 6.125180416580118e-05, "loss": 0.0169, "step": 36510 }, { "grad_norm": 0.22656914591789246, "learning_rate": 6.123166496720571e-05, "loss": 0.0164, "step": 36520 }, { "grad_norm": 0.32159265875816345, "learning_rate": 6.121152384942588e-05, "loss": 0.0194, "step": 36530 }, { "grad_norm": 0.21337296068668365, "learning_rate": 6.119138081590324e-05, "loss": 0.0215, "step": 36540 }, { "grad_norm": 0.18088282644748688, "learning_rate": 6.117123587007971e-05, "loss": 0.0164, "step": 36550 }, { "grad_norm": 0.25951245427131653, "learning_rate": 6.11510890153975e-05, "loss": 0.0253, "step": 36560 }, { "grad_norm": 0.2616865336894989, "learning_rate": 6.113094025529916e-05, "loss": 0.014, "step": 36570 }, { "grad_norm": 0.1957417130470276, "learning_rate": 6.111078959322757e-05, "loss": 0.0338, "step": 36580 }, { "grad_norm": 0.3780764043331146, "learning_rate": 6.109063703262592e-05, "loss": 0.0256, "step": 36590 }, { "grad_norm": 0.21874642372131348, "learning_rate": 6.107048257693772e-05, "loss": 0.0271, "step": 36600 }, { "grad_norm": 0.18032251298427582, "learning_rate": 6.105032622960683e-05, "loss": 0.0216, "step": 36610 }, { "grad_norm": 0.28114619851112366, "learning_rate": 6.103016799407743e-05, "loss": 0.0186, "step": 36620 }, { "grad_norm": 0.232025146484375, "learning_rate": 6.1010007873793984e-05, "loss": 0.0238, "step": 36630 }, { "grad_norm": 0.2118765413761139, "learning_rate": 6.098984587220131e-05, "loss": 0.0212, "step": 36640 }, { "grad_norm": 0.22145840525627136, "learning_rate": 6.096968199274456e-05, "loss": 0.0221, "step": 36650 }, { "grad_norm": 0.2516169250011444, "learning_rate": 6.0949516238869166e-05, "loss": 0.0177, "step": 36660 }, { "grad_norm": 0.2461695820093155, "learning_rate": 6.092934861402092e-05, "loss": 0.0278, "step": 36670 }, { "grad_norm": 0.3870866298675537, "learning_rate": 6.0909179121645924e-05, "loss": 0.0283, "step": 36680 }, { "grad_norm": 0.2527899146080017, "learning_rate": 6.0889007765190576e-05, "loss": 0.0263, "step": 36690 }, { "grad_norm": 0.22245171666145325, "learning_rate": 6.086883454810162e-05, "loss": 0.018, "step": 36700 }, { "grad_norm": 0.30118417739868164, "learning_rate": 6.0848659473826084e-05, "loss": 0.0258, "step": 36710 }, { "grad_norm": 0.3373793959617615, "learning_rate": 6.082848254581138e-05, "loss": 0.0241, "step": 36720 }, { "grad_norm": 0.3045825660228729, "learning_rate": 6.080830376750517e-05, "loss": 0.0199, "step": 36730 }, { "grad_norm": 0.22342315316200256, "learning_rate": 6.0788123142355445e-05, "loss": 0.015, "step": 36740 }, { "grad_norm": 0.25846627354621887, "learning_rate": 6.076794067381052e-05, "loss": 0.0199, "step": 36750 }, { "grad_norm": 0.2251823991537094, "learning_rate": 6.074775636531905e-05, "loss": 0.0238, "step": 36760 }, { "grad_norm": 0.3804880380630493, "learning_rate": 6.072757022032997e-05, "loss": 0.0232, "step": 36770 }, { "grad_norm": 0.16425642371177673, "learning_rate": 6.070738224229253e-05, "loss": 0.0244, "step": 36780 }, { "grad_norm": 0.28177252411842346, "learning_rate": 6.0687192434656314e-05, "loss": 0.0229, "step": 36790 }, { "grad_norm": 0.40316441655158997, "learning_rate": 6.066700080087121e-05, "loss": 0.018, "step": 36800 }, { "grad_norm": 0.26176342368125916, "learning_rate": 6.0646807344387424e-05, "loss": 0.0202, "step": 36810 }, { "grad_norm": 0.3478718101978302, "learning_rate": 6.062661206865543e-05, "loss": 0.0244, "step": 36820 }, { "grad_norm": 0.26201894879341125, "learning_rate": 6.06064149771261e-05, "loss": 0.022, "step": 36830 }, { "grad_norm": 0.24158166348934174, "learning_rate": 6.058621607325051e-05, "loss": 0.0219, "step": 36840 }, { "grad_norm": 0.24016758799552917, "learning_rate": 6.056601536048014e-05, "loss": 0.0222, "step": 36850 }, { "grad_norm": 0.32092955708503723, "learning_rate": 6.0545812842266725e-05, "loss": 0.0287, "step": 36860 }, { "grad_norm": 0.23185385763645172, "learning_rate": 6.052560852206232e-05, "loss": 0.0239, "step": 36870 }, { "grad_norm": 0.2690878212451935, "learning_rate": 6.05054024033193e-05, "loss": 0.0282, "step": 36880 }, { "grad_norm": 0.22753958404064178, "learning_rate": 6.048519448949032e-05, "loss": 0.0178, "step": 36890 }, { "grad_norm": 0.15745677053928375, "learning_rate": 6.046498478402839e-05, "loss": 0.0223, "step": 36900 }, { "grad_norm": 0.7788191437721252, "learning_rate": 6.044477329038677e-05, "loss": 0.022, "step": 36910 }, { "grad_norm": 0.26700708270072937, "learning_rate": 6.042456001201906e-05, "loss": 0.0237, "step": 36920 }, { "grad_norm": 0.21634025871753693, "learning_rate": 6.040434495237917e-05, "loss": 0.0228, "step": 36930 }, { "grad_norm": 0.3247472643852234, "learning_rate": 6.0384128114921256e-05, "loss": 0.0248, "step": 36940 }, { "grad_norm": 0.4715096354484558, "learning_rate": 6.036390950309987e-05, "loss": 0.0182, "step": 36950 }, { "grad_norm": 0.2129420042037964, "learning_rate": 6.0343689120369805e-05, "loss": 0.0255, "step": 36960 }, { "grad_norm": 0.2836648225784302, "learning_rate": 6.032346697018616e-05, "loss": 0.021, "step": 36970 }, { "grad_norm": 0.34617531299591064, "learning_rate": 6.0303243056004375e-05, "loss": 0.0379, "step": 36980 }, { "grad_norm": 0.17895375192165375, "learning_rate": 6.0283017381280136e-05, "loss": 0.0183, "step": 36990 }, { "grad_norm": 0.25002217292785645, "learning_rate": 6.026278994946948e-05, "loss": 0.0226, "step": 37000 }, { "grad_norm": 0.23617449402809143, "learning_rate": 6.02425607640287e-05, "loss": 0.0203, "step": 37010 }, { "grad_norm": 0.30759572982788086, "learning_rate": 6.022232982841441e-05, "loss": 0.0187, "step": 37020 }, { "grad_norm": 0.24066682159900665, "learning_rate": 6.020209714608355e-05, "loss": 0.0164, "step": 37030 }, { "grad_norm": 0.31087759137153625, "learning_rate": 6.018186272049331e-05, "loss": 0.0219, "step": 37040 }, { "grad_norm": 0.1968684196472168, "learning_rate": 6.01616265551012e-05, "loss": 0.0257, "step": 37050 }, { "grad_norm": 0.2171984612941742, "learning_rate": 6.014138865336503e-05, "loss": 0.0174, "step": 37060 }, { "grad_norm": 0.24273590743541718, "learning_rate": 6.0121149018742905e-05, "loss": 0.02, "step": 37070 }, { "grad_norm": 0.25918054580688477, "learning_rate": 6.010090765469325e-05, "loss": 0.0189, "step": 37080 }, { "grad_norm": 0.22797566652297974, "learning_rate": 6.008066456467473e-05, "loss": 0.0147, "step": 37090 }, { "grad_norm": 0.30052030086517334, "learning_rate": 6.0060419752146335e-05, "loss": 0.0205, "step": 37100 }, { "grad_norm": 0.21806582808494568, "learning_rate": 6.0040173220567353e-05, "loss": 0.0182, "step": 37110 }, { "grad_norm": 0.2948351204395294, "learning_rate": 6.001992497339737e-05, "loss": 0.0203, "step": 37120 }, { "grad_norm": 0.29230839014053345, "learning_rate": 5.999967501409626e-05, "loss": 0.0202, "step": 37130 }, { "grad_norm": 0.1952250450849533, "learning_rate": 5.997942334612418e-05, "loss": 0.0178, "step": 37140 }, { "grad_norm": 0.18025754392147064, "learning_rate": 5.995916997294158e-05, "loss": 0.0135, "step": 37150 }, { "grad_norm": 0.144478902220726, "learning_rate": 5.9938914898009214e-05, "loss": 0.0207, "step": 37160 }, { "grad_norm": 0.2259441465139389, "learning_rate": 5.991865812478813e-05, "loss": 0.0146, "step": 37170 }, { "grad_norm": 0.34154728055000305, "learning_rate": 5.989839965673964e-05, "loss": 0.0338, "step": 37180 }, { "grad_norm": 0.3017484247684479, "learning_rate": 5.987813949732539e-05, "loss": 0.0249, "step": 37190 }, { "grad_norm": 0.20017383992671967, "learning_rate": 5.9857877650007255e-05, "loss": 0.0197, "step": 37200 }, { "grad_norm": 0.2888721525669098, "learning_rate": 5.983761411824744e-05, "loss": 0.0179, "step": 37210 }, { "grad_norm": 0.24640683829784393, "learning_rate": 5.981734890550844e-05, "loss": 0.0225, "step": 37220 }, { "grad_norm": 0.23211748898029327, "learning_rate": 5.979708201525301e-05, "loss": 0.0239, "step": 37230 }, { "grad_norm": 0.2565743029117584, "learning_rate": 5.977681345094422e-05, "loss": 0.0199, "step": 37240 }, { "grad_norm": 0.20457513630390167, "learning_rate": 5.97565432160454e-05, "loss": 0.0197, "step": 37250 }, { "grad_norm": 0.20877501368522644, "learning_rate": 5.9736271314020186e-05, "loss": 0.0199, "step": 37260 }, { "grad_norm": 0.3794323801994324, "learning_rate": 5.971599774833251e-05, "loss": 0.0193, "step": 37270 }, { "grad_norm": 0.223756343126297, "learning_rate": 5.9695722522446525e-05, "loss": 0.0245, "step": 37280 }, { "grad_norm": 0.17067758738994598, "learning_rate": 5.9675445639826765e-05, "loss": 0.0189, "step": 37290 }, { "grad_norm": 0.5961388945579529, "learning_rate": 5.965516710393796e-05, "loss": 0.0192, "step": 37300 }, { "grad_norm": 0.27820438146591187, "learning_rate": 5.963488691824516e-05, "loss": 0.0262, "step": 37310 }, { "grad_norm": 0.30284491181373596, "learning_rate": 5.96146050862137e-05, "loss": 0.0277, "step": 37320 }, { "grad_norm": 0.27004876732826233, "learning_rate": 5.959432161130919e-05, "loss": 0.0185, "step": 37330 }, { "grad_norm": 0.24627381563186646, "learning_rate": 5.9574036496997545e-05, "loss": 0.0256, "step": 37340 }, { "grad_norm": 0.260061651468277, "learning_rate": 5.955374974674488e-05, "loss": 0.0224, "step": 37350 }, { "grad_norm": 0.2373119592666626, "learning_rate": 5.9533461364017696e-05, "loss": 0.0226, "step": 37360 }, { "grad_norm": 0.35422202944755554, "learning_rate": 5.9513171352282716e-05, "loss": 0.0196, "step": 37370 }, { "grad_norm": 0.31220510601997375, "learning_rate": 5.949287971500692e-05, "loss": 0.0163, "step": 37380 }, { "grad_norm": 0.17704223096370697, "learning_rate": 5.947258645565762e-05, "loss": 0.0176, "step": 37390 }, { "grad_norm": 0.2930271029472351, "learning_rate": 5.945229157770237e-05, "loss": 0.0176, "step": 37400 }, { "grad_norm": 0.24217239022254944, "learning_rate": 5.9431995084609006e-05, "loss": 0.0218, "step": 37410 }, { "grad_norm": 0.39501410722732544, "learning_rate": 5.941169697984564e-05, "loss": 0.0205, "step": 37420 }, { "grad_norm": 0.29337090253829956, "learning_rate": 5.9391397266880675e-05, "loss": 0.0161, "step": 37430 }, { "grad_norm": 0.2583591341972351, "learning_rate": 5.937109594918279e-05, "loss": 0.0313, "step": 37440 }, { "grad_norm": 0.18956196308135986, "learning_rate": 5.9350793030220884e-05, "loss": 0.0324, "step": 37450 }, { "grad_norm": 0.2437855303287506, "learning_rate": 5.933048851346421e-05, "loss": 0.0193, "step": 37460 }, { "grad_norm": 0.32579678297042847, "learning_rate": 5.931018240238224e-05, "loss": 0.0369, "step": 37470 }, { "grad_norm": 0.313193678855896, "learning_rate": 5.928987470044471e-05, "loss": 0.018, "step": 37480 }, { "grad_norm": 0.3104802072048187, "learning_rate": 5.9269565411121695e-05, "loss": 0.0204, "step": 37490 }, { "grad_norm": 0.3058614730834961, "learning_rate": 5.924925453788347e-05, "loss": 0.0259, "step": 37500 }, { "grad_norm": 0.20232534408569336, "learning_rate": 5.92289420842006e-05, "loss": 0.0248, "step": 37510 }, { "grad_norm": 0.3467719256877899, "learning_rate": 5.9208628053543945e-05, "loss": 0.0286, "step": 37520 }, { "grad_norm": 0.33912745118141174, "learning_rate": 5.918831244938462e-05, "loss": 0.0182, "step": 37530 }, { "grad_norm": 0.32641175389289856, "learning_rate": 5.9167995275194e-05, "loss": 0.0218, "step": 37540 }, { "grad_norm": 0.233133003115654, "learning_rate": 5.914767653444373e-05, "loss": 0.0218, "step": 37550 }, { "grad_norm": 0.3303965628147125, "learning_rate": 5.912735623060572e-05, "loss": 0.0263, "step": 37560 }, { "grad_norm": 0.1669832020998001, "learning_rate": 5.910703436715217e-05, "loss": 0.0165, "step": 37570 }, { "grad_norm": 0.16055849194526672, "learning_rate": 5.908671094755552e-05, "loss": 0.02, "step": 37580 }, { "grad_norm": 0.20804786682128906, "learning_rate": 5.906638597528851e-05, "loss": 0.0169, "step": 37590 }, { "grad_norm": 0.24606463313102722, "learning_rate": 5.9046059453824076e-05, "loss": 0.018, "step": 37600 }, { "grad_norm": 0.312480628490448, "learning_rate": 5.9025731386635505e-05, "loss": 0.0175, "step": 37610 }, { "grad_norm": 0.3621552288532257, "learning_rate": 5.900540177719629e-05, "loss": 0.0197, "step": 37620 }, { "grad_norm": 0.3254748582839966, "learning_rate": 5.898507062898021e-05, "loss": 0.022, "step": 37630 }, { "grad_norm": 0.3293412923812866, "learning_rate": 5.8964737945461316e-05, "loss": 0.0177, "step": 37640 }, { "grad_norm": 0.16695842146873474, "learning_rate": 5.8944403730113885e-05, "loss": 0.019, "step": 37650 }, { "grad_norm": 0.27738434076309204, "learning_rate": 5.892406798641248e-05, "loss": 0.027, "step": 37660 }, { "grad_norm": 0.2643584609031677, "learning_rate": 5.890373071783193e-05, "loss": 0.0174, "step": 37670 }, { "grad_norm": 0.24850568175315857, "learning_rate": 5.888339192784732e-05, "loss": 0.0203, "step": 37680 }, { "grad_norm": 0.18536017835140228, "learning_rate": 5.8863051619934003e-05, "loss": 0.0167, "step": 37690 }, { "grad_norm": 0.2258821576833725, "learning_rate": 5.8842709797567554e-05, "loss": 0.02, "step": 37700 }, { "grad_norm": 0.39606401324272156, "learning_rate": 5.8822366464223855e-05, "loss": 0.0127, "step": 37710 }, { "grad_norm": 0.18683776259422302, "learning_rate": 5.880202162337901e-05, "loss": 0.0243, "step": 37720 }, { "grad_norm": 0.20962531864643097, "learning_rate": 5.8781675278509405e-05, "loss": 0.0156, "step": 37730 }, { "grad_norm": 0.3468412458896637, "learning_rate": 5.8761327433091696e-05, "loss": 0.0189, "step": 37740 }, { "grad_norm": 0.276273638010025, "learning_rate": 5.874097809060275e-05, "loss": 0.0242, "step": 37750 }, { "grad_norm": 0.23937080800533295, "learning_rate": 5.87206272545197e-05, "loss": 0.0322, "step": 37760 }, { "grad_norm": 0.29004985094070435, "learning_rate": 5.8700274928319955e-05, "loss": 0.0232, "step": 37770 }, { "grad_norm": 0.3146826922893524, "learning_rate": 5.867992111548118e-05, "loss": 0.0198, "step": 37780 }, { "grad_norm": 0.22933921217918396, "learning_rate": 5.865956581948131e-05, "loss": 0.0167, "step": 37790 }, { "grad_norm": 0.25586768984794617, "learning_rate": 5.863920904379845e-05, "loss": 0.0167, "step": 37800 }, { "grad_norm": 0.21553121507167816, "learning_rate": 5.8618850791911064e-05, "loss": 0.0224, "step": 37810 }, { "grad_norm": 0.22491104900836945, "learning_rate": 5.859849106729779e-05, "loss": 0.0213, "step": 37820 }, { "grad_norm": 0.21074406802654266, "learning_rate": 5.857812987343758e-05, "loss": 0.0178, "step": 37830 }, { "grad_norm": 0.29734930396080017, "learning_rate": 5.855776721380957e-05, "loss": 0.0216, "step": 37840 }, { "grad_norm": 0.20238399505615234, "learning_rate": 5.8537403091893217e-05, "loss": 0.0178, "step": 37850 }, { "grad_norm": 0.26531997323036194, "learning_rate": 5.851703751116816e-05, "loss": 0.0196, "step": 37860 }, { "grad_norm": 0.24663928151130676, "learning_rate": 5.8496670475114336e-05, "loss": 0.0157, "step": 37870 }, { "grad_norm": 0.2731669843196869, "learning_rate": 5.84763019872119e-05, "loss": 0.0258, "step": 37880 }, { "grad_norm": 0.3248920440673828, "learning_rate": 5.845593205094131e-05, "loss": 0.0206, "step": 37890 }, { "grad_norm": 0.24700497090816498, "learning_rate": 5.843556066978318e-05, "loss": 0.0164, "step": 37900 }, { "grad_norm": 0.2146148681640625, "learning_rate": 5.8415187847218455e-05, "loss": 0.0228, "step": 37910 }, { "grad_norm": 0.20721624791622162, "learning_rate": 5.839481358672827e-05, "loss": 0.0162, "step": 37920 }, { "grad_norm": 0.19441768527030945, "learning_rate": 5.837443789179407e-05, "loss": 0.0188, "step": 37930 }, { "grad_norm": 0.3140900731086731, "learning_rate": 5.8354060765897445e-05, "loss": 0.0165, "step": 37940 }, { "grad_norm": 0.19396725296974182, "learning_rate": 5.8333682212520334e-05, "loss": 0.028, "step": 37950 }, { "grad_norm": 0.2993248999118805, "learning_rate": 5.831330223514486e-05, "loss": 0.0194, "step": 37960 }, { "grad_norm": 0.20845073461532593, "learning_rate": 5.8292920837253396e-05, "loss": 0.0215, "step": 37970 }, { "grad_norm": 0.31191903352737427, "learning_rate": 5.827253802232857e-05, "loss": 0.0178, "step": 37980 }, { "grad_norm": 0.22257503867149353, "learning_rate": 5.825215379385327e-05, "loss": 0.0173, "step": 37990 }, { "grad_norm": 0.1915523111820221, "learning_rate": 5.823176815531057e-05, "loss": 0.0384, "step": 38000 }, { "grad_norm": 0.2307330071926117, "learning_rate": 5.8211381110183826e-05, "loss": 0.0278, "step": 38010 }, { "grad_norm": 0.19734619557857513, "learning_rate": 5.8190992661956645e-05, "loss": 0.0175, "step": 38020 }, { "grad_norm": 0.1818557232618332, "learning_rate": 5.817060281411284e-05, "loss": 0.0198, "step": 38030 }, { "grad_norm": 0.20280954241752625, "learning_rate": 5.815021157013647e-05, "loss": 0.0175, "step": 38040 }, { "grad_norm": 0.27008116245269775, "learning_rate": 5.8129818933511856e-05, "loss": 0.031, "step": 38050 }, { "grad_norm": 0.2561416029930115, "learning_rate": 5.8109424907723544e-05, "loss": 0.0244, "step": 38060 }, { "grad_norm": 0.1765030026435852, "learning_rate": 5.80890294962563e-05, "loss": 0.0144, "step": 38070 }, { "grad_norm": 0.38172438740730286, "learning_rate": 5.806863270259515e-05, "loss": 0.0274, "step": 38080 }, { "grad_norm": 0.30231016874313354, "learning_rate": 5.804823453022536e-05, "loss": 0.0174, "step": 38090 }, { "grad_norm": 0.26184237003326416, "learning_rate": 5.80278349826324e-05, "loss": 0.0224, "step": 38100 }, { "grad_norm": 0.22098992764949799, "learning_rate": 5.8007434063302014e-05, "loss": 0.0198, "step": 38110 }, { "grad_norm": 0.33082500100135803, "learning_rate": 5.7987031775720136e-05, "loss": 0.0196, "step": 38120 }, { "grad_norm": 0.197537362575531, "learning_rate": 5.7966628123372976e-05, "loss": 0.0297, "step": 38130 }, { "grad_norm": 0.43431034684181213, "learning_rate": 5.7946223109746956e-05, "loss": 0.0325, "step": 38140 }, { "grad_norm": 0.2721039354801178, "learning_rate": 5.7925816738328754e-05, "loss": 0.0168, "step": 38150 }, { "grad_norm": 0.3201061487197876, "learning_rate": 5.790540901260521e-05, "loss": 0.022, "step": 38160 }, { "grad_norm": 0.29938578605651855, "learning_rate": 5.788499993606351e-05, "loss": 0.0226, "step": 38170 }, { "grad_norm": 0.2365482896566391, "learning_rate": 5.786458951219096e-05, "loss": 0.0243, "step": 38180 }, { "grad_norm": 0.24021603167057037, "learning_rate": 5.784417774447517e-05, "loss": 0.027, "step": 38190 }, { "grad_norm": 0.25166088342666626, "learning_rate": 5.782376463640393e-05, "loss": 0.0211, "step": 38200 }, { "grad_norm": 0.571654200553894, "learning_rate": 5.780335019146531e-05, "loss": 0.0219, "step": 38210 }, { "grad_norm": 0.2554699182510376, "learning_rate": 5.778293441314755e-05, "loss": 0.0195, "step": 38220 }, { "grad_norm": 0.34906283020973206, "learning_rate": 5.776251730493917e-05, "loss": 0.0323, "step": 38230 }, { "grad_norm": 0.2678050696849823, "learning_rate": 5.774209887032887e-05, "loss": 0.0231, "step": 38240 }, { "grad_norm": 0.1958552896976471, "learning_rate": 5.772167911280565e-05, "loss": 0.0187, "step": 38250 }, { "grad_norm": 0.28777578473091125, "learning_rate": 5.770125803585864e-05, "loss": 0.0184, "step": 38260 }, { "grad_norm": 0.34909555315971375, "learning_rate": 5.768083564297726e-05, "loss": 0.0302, "step": 38270 }, { "grad_norm": 0.22978085279464722, "learning_rate": 5.766041193765114e-05, "loss": 0.0215, "step": 38280 }, { "grad_norm": 0.2464255392551422, "learning_rate": 5.763998692337015e-05, "loss": 0.0193, "step": 38290 }, { "grad_norm": 0.3018675446510315, "learning_rate": 5.761956060362433e-05, "loss": 0.0226, "step": 38300 }, { "grad_norm": 0.2613265812397003, "learning_rate": 5.7599132981904e-05, "loss": 0.0202, "step": 38310 }, { "grad_norm": 0.23962630331516266, "learning_rate": 5.75787040616997e-05, "loss": 0.0252, "step": 38320 }, { "grad_norm": 0.3630169630050659, "learning_rate": 5.755827384650212e-05, "loss": 0.0189, "step": 38330 }, { "grad_norm": 0.27075570821762085, "learning_rate": 5.753784233980228e-05, "loss": 0.0202, "step": 38340 }, { "grad_norm": 0.2106432020664215, "learning_rate": 5.751740954509135e-05, "loss": 0.0145, "step": 38350 }, { "grad_norm": 0.15609031915664673, "learning_rate": 5.7496975465860715e-05, "loss": 0.0143, "step": 38360 }, { "grad_norm": 0.15497207641601562, "learning_rate": 5.747654010560202e-05, "loss": 0.0159, "step": 38370 }, { "grad_norm": 0.19676022231578827, "learning_rate": 5.7456103467807097e-05, "loss": 0.0219, "step": 38380 }, { "grad_norm": 0.2931155562400818, "learning_rate": 5.7435665555968046e-05, "loss": 0.0183, "step": 38390 }, { "grad_norm": 0.20722000300884247, "learning_rate": 5.74152263735771e-05, "loss": 0.0259, "step": 38400 }, { "grad_norm": 0.256562203168869, "learning_rate": 5.739478592412677e-05, "loss": 0.0201, "step": 38410 }, { "grad_norm": 0.29411181807518005, "learning_rate": 5.7374344211109766e-05, "loss": 0.0269, "step": 38420 }, { "grad_norm": 0.23177170753479004, "learning_rate": 5.735390123801904e-05, "loss": 0.0233, "step": 38430 }, { "grad_norm": 0.35888969898223877, "learning_rate": 5.7333457008347704e-05, "loss": 0.029, "step": 38440 }, { "grad_norm": 0.33015117049217224, "learning_rate": 5.7313011525589156e-05, "loss": 0.0236, "step": 38450 }, { "grad_norm": 0.24060456454753876, "learning_rate": 5.729256479323694e-05, "loss": 0.0188, "step": 38460 }, { "grad_norm": 0.24541470408439636, "learning_rate": 5.727211681478485e-05, "loss": 0.0224, "step": 38470 }, { "grad_norm": 0.23035983741283417, "learning_rate": 5.7251667593726886e-05, "loss": 0.0248, "step": 38480 }, { "grad_norm": 0.258213609457016, "learning_rate": 5.723121713355728e-05, "loss": 0.0333, "step": 38490 }, { "grad_norm": 0.1504814475774765, "learning_rate": 5.721076543777044e-05, "loss": 0.0229, "step": 38500 }, { "grad_norm": 0.1968914419412613, "learning_rate": 5.7190312509860986e-05, "loss": 0.018, "step": 38510 }, { "grad_norm": 0.20036780834197998, "learning_rate": 5.716985835332379e-05, "loss": 0.0205, "step": 38520 }, { "grad_norm": 0.3765339255332947, "learning_rate": 5.714940297165389e-05, "loss": 0.0191, "step": 38530 }, { "grad_norm": 0.2901117205619812, "learning_rate": 5.712894636834656e-05, "loss": 0.0192, "step": 38540 }, { "grad_norm": 0.1934562623500824, "learning_rate": 5.7108488546897276e-05, "loss": 0.0127, "step": 38550 }, { "grad_norm": 0.16173474490642548, "learning_rate": 5.708802951080172e-05, "loss": 0.0177, "step": 38560 }, { "grad_norm": 0.22822269797325134, "learning_rate": 5.706756926355576e-05, "loss": 0.0179, "step": 38570 }, { "grad_norm": 0.230693057179451, "learning_rate": 5.704710780865554e-05, "loss": 0.0177, "step": 38580 }, { "grad_norm": 0.20272685587406158, "learning_rate": 5.7026645149597325e-05, "loss": 0.0185, "step": 38590 }, { "grad_norm": 0.2670861482620239, "learning_rate": 5.700618128987764e-05, "loss": 0.0205, "step": 38600 }, { "grad_norm": 0.17990915477275848, "learning_rate": 5.698571623299317e-05, "loss": 0.02, "step": 38610 }, { "grad_norm": 0.21326209604740143, "learning_rate": 5.696524998244086e-05, "loss": 0.0158, "step": 38620 }, { "grad_norm": 0.17396984994411469, "learning_rate": 5.6944782541717836e-05, "loss": 0.0243, "step": 38630 }, { "grad_norm": 0.20757131278514862, "learning_rate": 5.69243139143214e-05, "loss": 0.0177, "step": 38640 }, { "grad_norm": 0.21231622993946075, "learning_rate": 5.6903844103749125e-05, "loss": 0.0221, "step": 38650 }, { "grad_norm": 0.22041364014148712, "learning_rate": 5.688337311349869e-05, "loss": 0.0154, "step": 38660 }, { "grad_norm": 0.17478644847869873, "learning_rate": 5.6862900947068074e-05, "loss": 0.0199, "step": 38670 }, { "grad_norm": 0.24743656814098358, "learning_rate": 5.6842427607955374e-05, "loss": 0.025, "step": 38680 }, { "grad_norm": 0.2176942378282547, "learning_rate": 5.682195309965893e-05, "loss": 0.0151, "step": 38690 }, { "grad_norm": 0.2722650468349457, "learning_rate": 5.6801477425677294e-05, "loss": 0.0198, "step": 38700 }, { "grad_norm": 0.2676020562648773, "learning_rate": 5.678100058950917e-05, "loss": 0.0234, "step": 38710 }, { "grad_norm": 0.31032103300094604, "learning_rate": 5.676052259465352e-05, "loss": 0.0212, "step": 38720 }, { "grad_norm": 0.1540268510580063, "learning_rate": 5.674004344460945e-05, "loss": 0.0162, "step": 38730 }, { "grad_norm": 0.255532830953598, "learning_rate": 5.6719563142876295e-05, "loss": 0.018, "step": 38740 }, { "grad_norm": 0.22131037712097168, "learning_rate": 5.669908169295359e-05, "loss": 0.0168, "step": 38750 }, { "grad_norm": 0.22612124681472778, "learning_rate": 5.667859909834105e-05, "loss": 0.0187, "step": 38760 }, { "grad_norm": 0.23141610622406006, "learning_rate": 5.6658115362538565e-05, "loss": 0.0181, "step": 38770 }, { "grad_norm": 0.3121827244758606, "learning_rate": 5.663763048904628e-05, "loss": 0.0181, "step": 38780 }, { "grad_norm": 0.3247644603252411, "learning_rate": 5.661714448136447e-05, "loss": 0.0156, "step": 38790 }, { "grad_norm": 0.15905123949050903, "learning_rate": 5.659665734299366e-05, "loss": 0.0143, "step": 38800 }, { "grad_norm": 0.2431437373161316, "learning_rate": 5.6576169077434516e-05, "loss": 0.0189, "step": 38810 }, { "grad_norm": 0.1729469895362854, "learning_rate": 5.6555679688187944e-05, "loss": 0.0164, "step": 38820 }, { "grad_norm": 0.4400544762611389, "learning_rate": 5.6535189178755e-05, "loss": 0.0209, "step": 38830 }, { "grad_norm": 0.25763869285583496, "learning_rate": 5.651469755263698e-05, "loss": 0.0222, "step": 38840 }, { "grad_norm": 0.17444247007369995, "learning_rate": 5.6494204813335316e-05, "loss": 0.0253, "step": 38850 }, { "grad_norm": 0.21168828010559082, "learning_rate": 5.647371096435168e-05, "loss": 0.016, "step": 38860 }, { "grad_norm": 0.21558824181556702, "learning_rate": 5.645321600918788e-05, "loss": 0.0152, "step": 38870 }, { "grad_norm": 0.3254982531070709, "learning_rate": 5.643271995134597e-05, "loss": 0.0257, "step": 38880 }, { "grad_norm": 0.19281399250030518, "learning_rate": 5.641222279432814e-05, "loss": 0.0193, "step": 38890 }, { "grad_norm": 0.23689217865467072, "learning_rate": 5.6391724541636834e-05, "loss": 0.017, "step": 38900 }, { "grad_norm": 0.316921591758728, "learning_rate": 5.6371225196774605e-05, "loss": 0.0246, "step": 38910 }, { "grad_norm": 0.28524962067604065, "learning_rate": 5.635072476324423e-05, "loss": 0.0238, "step": 38920 }, { "grad_norm": 0.2753354012966156, "learning_rate": 5.63302232445487e-05, "loss": 0.0181, "step": 38930 }, { "grad_norm": 0.17143982648849487, "learning_rate": 5.6309720644191144e-05, "loss": 0.0224, "step": 38940 }, { "grad_norm": 0.41397491097450256, "learning_rate": 5.628921696567491e-05, "loss": 0.0169, "step": 38950 }, { "grad_norm": 0.16987231373786926, "learning_rate": 5.62687122125035e-05, "loss": 0.0148, "step": 38960 }, { "grad_norm": 0.36628013849258423, "learning_rate": 5.624820638818062e-05, "loss": 0.0157, "step": 38970 }, { "grad_norm": 0.2942805886268616, "learning_rate": 5.6227699496210164e-05, "loss": 0.0221, "step": 38980 }, { "grad_norm": 0.23170998692512512, "learning_rate": 5.6207191540096195e-05, "loss": 0.021, "step": 38990 }, { "grad_norm": 0.15858037769794464, "learning_rate": 5.618668252334296e-05, "loss": 0.0145, "step": 39000 }, { "grad_norm": 0.2400592714548111, "learning_rate": 5.616617244945488e-05, "loss": 0.017, "step": 39010 }, { "grad_norm": 0.34083157777786255, "learning_rate": 5.614566132193656e-05, "loss": 0.0192, "step": 39020 }, { "grad_norm": 0.20864196121692657, "learning_rate": 5.612514914429282e-05, "loss": 0.0212, "step": 39030 }, { "grad_norm": 0.2822914123535156, "learning_rate": 5.610463592002863e-05, "loss": 0.0199, "step": 39040 }, { "grad_norm": 0.3738754689693451, "learning_rate": 5.608412165264909e-05, "loss": 0.0228, "step": 39050 }, { "grad_norm": 0.14278538525104523, "learning_rate": 5.606360634565959e-05, "loss": 0.0203, "step": 39060 }, { "grad_norm": 0.26132747530937195, "learning_rate": 5.604309000256559e-05, "loss": 0.0234, "step": 39070 }, { "grad_norm": 0.2952286899089813, "learning_rate": 5.6022572626872785e-05, "loss": 0.0197, "step": 39080 }, { "grad_norm": 0.2337682545185089, "learning_rate": 5.600205422208704e-05, "loss": 0.0221, "step": 39090 }, { "grad_norm": 0.24420860409736633, "learning_rate": 5.59815347917144e-05, "loss": 0.0236, "step": 39100 }, { "grad_norm": 0.27904725074768066, "learning_rate": 5.596101433926103e-05, "loss": 0.024, "step": 39110 }, { "grad_norm": 0.30876871943473816, "learning_rate": 5.5940492868233364e-05, "loss": 0.0174, "step": 39120 }, { "grad_norm": 0.24612170457839966, "learning_rate": 5.591997038213793e-05, "loss": 0.0182, "step": 39130 }, { "grad_norm": 0.33479437232017517, "learning_rate": 5.5899446884481475e-05, "loss": 0.0168, "step": 39140 }, { "grad_norm": 0.2245582491159439, "learning_rate": 5.5878922378770906e-05, "loss": 0.038, "step": 39150 }, { "grad_norm": 0.27465003728866577, "learning_rate": 5.5858396868513285e-05, "loss": 0.022, "step": 39160 }, { "grad_norm": 0.29657623171806335, "learning_rate": 5.583787035721586e-05, "loss": 0.0171, "step": 39170 }, { "grad_norm": 0.2402195930480957, "learning_rate": 5.581734284838606e-05, "loss": 0.0185, "step": 39180 }, { "grad_norm": 0.3711870014667511, "learning_rate": 5.579681434553147e-05, "loss": 0.0178, "step": 39190 }, { "grad_norm": 0.2504369616508484, "learning_rate": 5.5776284852159854e-05, "loss": 0.0186, "step": 39200 }, { "grad_norm": 0.2447347342967987, "learning_rate": 5.575575437177913e-05, "loss": 0.026, "step": 39210 }, { "grad_norm": 0.1944681853055954, "learning_rate": 5.573522290789742e-05, "loss": 0.0148, "step": 39220 }, { "grad_norm": 0.25986456871032715, "learning_rate": 5.571469046402298e-05, "loss": 0.021, "step": 39230 }, { "grad_norm": 0.29468557238578796, "learning_rate": 5.5694157043664205e-05, "loss": 0.0244, "step": 39240 }, { "grad_norm": 0.27478137612342834, "learning_rate": 5.567362265032975e-05, "loss": 0.0138, "step": 39250 }, { "grad_norm": 0.16351109743118286, "learning_rate": 5.565308728752836e-05, "loss": 0.0154, "step": 39260 }, { "grad_norm": 0.22490984201431274, "learning_rate": 5.5632550958768956e-05, "loss": 0.0166, "step": 39270 }, { "grad_norm": 0.3999177813529968, "learning_rate": 5.5612013667560636e-05, "loss": 0.0226, "step": 39280 }, { "grad_norm": 0.3727874159812927, "learning_rate": 5.5591475417412676e-05, "loss": 0.0355, "step": 39290 }, { "grad_norm": 0.380802720785141, "learning_rate": 5.557093621183451e-05, "loss": 0.0261, "step": 39300 }, { "grad_norm": 0.16120284795761108, "learning_rate": 5.55503960543357e-05, "loss": 0.0162, "step": 39310 }, { "grad_norm": 0.20715151727199554, "learning_rate": 5.552985494842601e-05, "loss": 0.0198, "step": 39320 }, { "grad_norm": 0.2309887707233429, "learning_rate": 5.550931289761534e-05, "loss": 0.0195, "step": 39330 }, { "grad_norm": 0.24677038192749023, "learning_rate": 5.548876990541378e-05, "loss": 0.0194, "step": 39340 }, { "grad_norm": 0.27539992332458496, "learning_rate": 5.5468225975331564e-05, "loss": 0.0252, "step": 39350 }, { "grad_norm": 0.6163018941879272, "learning_rate": 5.544768111087909e-05, "loss": 0.0213, "step": 39360 }, { "grad_norm": 0.22497712075710297, "learning_rate": 5.54271353155669e-05, "loss": 0.0222, "step": 39370 }, { "grad_norm": 0.1941821277141571, "learning_rate": 5.5406588592905715e-05, "loss": 0.0202, "step": 39380 }, { "grad_norm": 0.292273610830307, "learning_rate": 5.5386040946406416e-05, "loss": 0.0228, "step": 39390 }, { "grad_norm": 0.2448796033859253, "learning_rate": 5.536549237958004e-05, "loss": 0.0296, "step": 39400 }, { "grad_norm": 0.1865878403186798, "learning_rate": 5.5344942895937744e-05, "loss": 0.0306, "step": 39410 }, { "grad_norm": 0.24911938607692719, "learning_rate": 5.5324392498990904e-05, "loss": 0.0165, "step": 39420 }, { "grad_norm": 0.28576743602752686, "learning_rate": 5.5303841192251e-05, "loss": 0.0176, "step": 39430 }, { "grad_norm": 0.23130540549755096, "learning_rate": 5.52832889792297e-05, "loss": 0.02, "step": 39440 }, { "grad_norm": 0.1928742676973343, "learning_rate": 5.526273586343881e-05, "loss": 0.0187, "step": 39450 }, { "grad_norm": 0.18640662729740143, "learning_rate": 5.5242181848390306e-05, "loss": 0.0146, "step": 39460 }, { "grad_norm": 0.2524102032184601, "learning_rate": 5.5221626937596285e-05, "loss": 0.0123, "step": 39470 }, { "grad_norm": 0.19665074348449707, "learning_rate": 5.520107113456903e-05, "loss": 0.023, "step": 39480 }, { "grad_norm": 0.35704505443573, "learning_rate": 5.5180514442820974e-05, "loss": 0.0288, "step": 39490 }, { "grad_norm": 0.22804076969623566, "learning_rate": 5.515995686586469e-05, "loss": 0.0132, "step": 39500 }, { "grad_norm": 0.2399081289768219, "learning_rate": 5.5139398407212916e-05, "loss": 0.0155, "step": 39510 }, { "grad_norm": 0.2499360591173172, "learning_rate": 5.511883907037849e-05, "loss": 0.0163, "step": 39520 }, { "grad_norm": 0.2417306900024414, "learning_rate": 5.509827885887449e-05, "loss": 0.0182, "step": 39530 }, { "grad_norm": 0.3196011483669281, "learning_rate": 5.507771777621406e-05, "loss": 0.0184, "step": 39540 }, { "grad_norm": 0.24722164869308472, "learning_rate": 5.505715582591052e-05, "loss": 0.0313, "step": 39550 }, { "grad_norm": 0.2925489544868469, "learning_rate": 5.50365930114774e-05, "loss": 0.0253, "step": 39560 }, { "grad_norm": 0.1872486025094986, "learning_rate": 5.5016029336428255e-05, "loss": 0.0119, "step": 39570 }, { "grad_norm": 0.23842491209506989, "learning_rate": 5.49954648042769e-05, "loss": 0.0202, "step": 39580 }, { "grad_norm": 0.19902974367141724, "learning_rate": 5.4974899418537226e-05, "loss": 0.026, "step": 39590 }, { "grad_norm": 0.21572133898735046, "learning_rate": 5.4954333182723316e-05, "loss": 0.0213, "step": 39600 }, { "grad_norm": 0.2922287583351135, "learning_rate": 5.493376610034937e-05, "loss": 0.0178, "step": 39610 }, { "grad_norm": 0.20730267465114594, "learning_rate": 5.4913198174929735e-05, "loss": 0.0159, "step": 39620 }, { "grad_norm": 0.17366647720336914, "learning_rate": 5.489262940997889e-05, "loss": 0.0218, "step": 39630 }, { "grad_norm": 0.20300711691379547, "learning_rate": 5.487205980901151e-05, "loss": 0.0268, "step": 39640 }, { "grad_norm": 0.2486766278743744, "learning_rate": 5.485148937554234e-05, "loss": 0.0332, "step": 39650 }, { "grad_norm": 0.24031664431095123, "learning_rate": 5.483091811308635e-05, "loss": 0.0169, "step": 39660 }, { "grad_norm": 0.24429507553577423, "learning_rate": 5.4810346025158564e-05, "loss": 0.0186, "step": 39670 }, { "grad_norm": 0.3334853947162628, "learning_rate": 5.478977311527421e-05, "loss": 0.0164, "step": 39680 }, { "grad_norm": 0.307584285736084, "learning_rate": 5.476919938694863e-05, "loss": 0.0179, "step": 39690 }, { "grad_norm": 0.22575144469738007, "learning_rate": 5.474862484369733e-05, "loss": 0.0145, "step": 39700 }, { "grad_norm": 0.18582791090011597, "learning_rate": 5.472804948903589e-05, "loss": 0.0185, "step": 39710 }, { "grad_norm": 0.24665002524852753, "learning_rate": 5.470747332648013e-05, "loss": 0.0135, "step": 39720 }, { "grad_norm": 0.41835319995880127, "learning_rate": 5.468689635954591e-05, "loss": 0.0308, "step": 39730 }, { "grad_norm": 0.22819013893604279, "learning_rate": 5.46663185917493e-05, "loss": 0.0206, "step": 39740 }, { "grad_norm": 0.24856390058994293, "learning_rate": 5.464574002660645e-05, "loss": 0.0166, "step": 39750 }, { "grad_norm": 0.12596048414707184, "learning_rate": 5.4625160667633724e-05, "loss": 0.0204, "step": 39760 }, { "grad_norm": 0.13934163749217987, "learning_rate": 5.4604580518347505e-05, "loss": 0.0238, "step": 39770 }, { "grad_norm": 0.2149638682603836, "learning_rate": 5.458399958226442e-05, "loss": 0.0209, "step": 39780 }, { "grad_norm": 0.18265938758850098, "learning_rate": 5.456341786290119e-05, "loss": 0.0189, "step": 39790 }, { "grad_norm": 0.23999717831611633, "learning_rate": 5.454283536377465e-05, "loss": 0.0235, "step": 39800 }, { "grad_norm": 0.2388908714056015, "learning_rate": 5.452225208840179e-05, "loss": 0.0177, "step": 39810 }, { "grad_norm": 0.21643058955669403, "learning_rate": 5.450166804029975e-05, "loss": 0.0225, "step": 39820 }, { "grad_norm": 0.21353782713413239, "learning_rate": 5.448108322298574e-05, "loss": 0.015, "step": 39830 }, { "grad_norm": 0.12577186524868011, "learning_rate": 5.446049763997717e-05, "loss": 0.0161, "step": 39840 }, { "grad_norm": 0.2188078612089157, "learning_rate": 5.4439911294791546e-05, "loss": 0.0151, "step": 39850 }, { "grad_norm": 0.3089984357357025, "learning_rate": 5.441932419094652e-05, "loss": 0.0157, "step": 39860 }, { "grad_norm": 0.18978887796401978, "learning_rate": 5.439873633195985e-05, "loss": 0.0162, "step": 39870 }, { "grad_norm": 0.2574547529220581, "learning_rate": 5.437814772134947e-05, "loss": 0.0259, "step": 39880 }, { "grad_norm": 0.22680404782295227, "learning_rate": 5.4357558362633366e-05, "loss": 0.0196, "step": 39890 }, { "grad_norm": 0.2796447277069092, "learning_rate": 5.4336968259329726e-05, "loss": 0.0149, "step": 39900 }, { "grad_norm": 0.21051537990570068, "learning_rate": 5.431637741495681e-05, "loss": 0.0225, "step": 39910 }, { "grad_norm": 0.3117828369140625, "learning_rate": 5.429578583303307e-05, "loss": 0.0226, "step": 39920 }, { "grad_norm": 0.15452612936496735, "learning_rate": 5.427519351707701e-05, "loss": 0.0265, "step": 39930 }, { "grad_norm": 0.20059211552143097, "learning_rate": 5.42546004706073e-05, "loss": 0.0189, "step": 39940 }, { "grad_norm": 0.3078990578651428, "learning_rate": 5.4234006697142735e-05, "loss": 0.0197, "step": 39950 }, { "grad_norm": 0.35674288868904114, "learning_rate": 5.421341220020224e-05, "loss": 0.0192, "step": 39960 }, { "grad_norm": 0.1255258321762085, "learning_rate": 5.419281698330482e-05, "loss": 0.0229, "step": 39970 }, { "grad_norm": 0.19368191063404083, "learning_rate": 5.4172221049969665e-05, "loss": 0.0217, "step": 39980 }, { "grad_norm": 0.28799155354499817, "learning_rate": 5.415162440371604e-05, "loss": 0.0172, "step": 39990 }, { "grad_norm": 0.20824965834617615, "learning_rate": 5.413102704806334e-05, "loss": 0.0252, "step": 40000 }, { "grad_norm": 0.2475138157606125, "learning_rate": 5.41104289865311e-05, "loss": 0.0207, "step": 40010 }, { "grad_norm": 0.21622315049171448, "learning_rate": 5.408983022263898e-05, "loss": 0.0174, "step": 40020 }, { "grad_norm": 0.4542444348335266, "learning_rate": 5.406923075990673e-05, "loss": 0.0226, "step": 40030 }, { "grad_norm": 0.2287251204252243, "learning_rate": 5.404863060185423e-05, "loss": 0.0173, "step": 40040 }, { "grad_norm": 0.21418985724449158, "learning_rate": 5.40280297520015e-05, "loss": 0.0186, "step": 40050 }, { "grad_norm": 0.27610620856285095, "learning_rate": 5.400742821386865e-05, "loss": 0.0184, "step": 40060 }, { "grad_norm": 0.2894287705421448, "learning_rate": 5.398682599097592e-05, "loss": 0.0249, "step": 40070 }, { "grad_norm": 0.21779705584049225, "learning_rate": 5.396622308684367e-05, "loss": 0.0171, "step": 40080 }, { "grad_norm": 0.2556561231613159, "learning_rate": 5.394561950499236e-05, "loss": 0.0207, "step": 40090 }, { "grad_norm": 0.19653402268886566, "learning_rate": 5.39250152489426e-05, "loss": 0.0177, "step": 40100 }, { "grad_norm": 0.1527900993824005, "learning_rate": 5.390441032221507e-05, "loss": 0.014, "step": 40110 }, { "grad_norm": 0.16566617786884308, "learning_rate": 5.388380472833062e-05, "loss": 0.0161, "step": 40120 }, { "grad_norm": 0.19075898826122284, "learning_rate": 5.386319847081016e-05, "loss": 0.0177, "step": 40130 }, { "grad_norm": 0.37960946559906006, "learning_rate": 5.384259155317473e-05, "loss": 0.0215, "step": 40140 }, { "grad_norm": 0.32570767402648926, "learning_rate": 5.38219839789455e-05, "loss": 0.0249, "step": 40150 }, { "grad_norm": 0.2038409560918808, "learning_rate": 5.380137575164376e-05, "loss": 0.0213, "step": 40160 }, { "grad_norm": 0.2198556512594223, "learning_rate": 5.378076687479085e-05, "loss": 0.0158, "step": 40170 }, { "grad_norm": 0.1871410459280014, "learning_rate": 5.3760157351908305e-05, "loss": 0.0201, "step": 40180 }, { "grad_norm": 0.2736450731754303, "learning_rate": 5.373954718651768e-05, "loss": 0.0201, "step": 40190 }, { "grad_norm": 0.20969510078430176, "learning_rate": 5.371893638214074e-05, "loss": 0.0141, "step": 40200 }, { "grad_norm": 0.3263636827468872, "learning_rate": 5.369832494229927e-05, "loss": 0.022, "step": 40210 }, { "grad_norm": 0.29217877984046936, "learning_rate": 5.367771287051524e-05, "loss": 0.0255, "step": 40220 }, { "grad_norm": 0.23764380812644958, "learning_rate": 5.365710017031065e-05, "loss": 0.0177, "step": 40230 }, { "grad_norm": 0.22854620218276978, "learning_rate": 5.3636486845207654e-05, "loss": 0.0208, "step": 40240 }, { "grad_norm": 0.24559548497200012, "learning_rate": 5.3615872898728526e-05, "loss": 0.0144, "step": 40250 }, { "grad_norm": 0.37130147218704224, "learning_rate": 5.3595258334395614e-05, "loss": 0.0238, "step": 40260 }, { "grad_norm": 0.1473921239376068, "learning_rate": 5.3574643155731396e-05, "loss": 0.0276, "step": 40270 }, { "grad_norm": 0.18203668296337128, "learning_rate": 5.3554027366258405e-05, "loss": 0.0153, "step": 40280 }, { "grad_norm": 0.20178525149822235, "learning_rate": 5.3533410969499355e-05, "loss": 0.0222, "step": 40290 }, { "grad_norm": 0.22267493605613708, "learning_rate": 5.3512793968977e-05, "loss": 0.0208, "step": 40300 }, { "grad_norm": 0.3287551701068878, "learning_rate": 5.349217636821423e-05, "loss": 0.0214, "step": 40310 }, { "grad_norm": 0.14851616322994232, "learning_rate": 5.347155817073405e-05, "loss": 0.0153, "step": 40320 }, { "grad_norm": 0.2991069555282593, "learning_rate": 5.34509393800595e-05, "loss": 0.0161, "step": 40330 }, { "grad_norm": 0.27004310488700867, "learning_rate": 5.343031999971381e-05, "loss": 0.0173, "step": 40340 }, { "grad_norm": 0.22480395436286926, "learning_rate": 5.340970003322026e-05, "loss": 0.0188, "step": 40350 }, { "grad_norm": 0.22714872658252716, "learning_rate": 5.33890794841022e-05, "loss": 0.0131, "step": 40360 }, { "grad_norm": 0.19415724277496338, "learning_rate": 5.336845835588318e-05, "loss": 0.0187, "step": 40370 }, { "grad_norm": 0.23897133767604828, "learning_rate": 5.334783665208674e-05, "loss": 0.0208, "step": 40380 }, { "grad_norm": 0.26717472076416016, "learning_rate": 5.332721437623657e-05, "loss": 0.0169, "step": 40390 }, { "grad_norm": 0.22002193331718445, "learning_rate": 5.3306591531856464e-05, "loss": 0.0242, "step": 40400 }, { "grad_norm": 0.2456635683774948, "learning_rate": 5.3285968122470295e-05, "loss": 0.0177, "step": 40410 }, { "grad_norm": 0.209879070520401, "learning_rate": 5.326534415160207e-05, "loss": 0.0239, "step": 40420 }, { "grad_norm": 0.2408129870891571, "learning_rate": 5.324471962277582e-05, "loss": 0.0234, "step": 40430 }, { "grad_norm": 0.31775733828544617, "learning_rate": 5.3224094539515746e-05, "loss": 0.0185, "step": 40440 }, { "grad_norm": 0.2449280023574829, "learning_rate": 5.3203468905346076e-05, "loss": 0.0156, "step": 40450 }, { "grad_norm": 0.22359542548656464, "learning_rate": 5.3182842723791195e-05, "loss": 0.0187, "step": 40460 }, { "grad_norm": 0.19353042542934418, "learning_rate": 5.316221599837554e-05, "loss": 0.016, "step": 40470 }, { "grad_norm": 0.18737414479255676, "learning_rate": 5.314158873262366e-05, "loss": 0.023, "step": 40480 }, { "grad_norm": 0.20211364328861237, "learning_rate": 5.312096093006018e-05, "loss": 0.0141, "step": 40490 }, { "grad_norm": 0.20117352902889252, "learning_rate": 5.3100332594209846e-05, "loss": 0.0168, "step": 40500 }, { "grad_norm": 0.20252251625061035, "learning_rate": 5.3079703728597454e-05, "loss": 0.0155, "step": 40510 }, { "grad_norm": 0.25558745861053467, "learning_rate": 5.305907433674794e-05, "loss": 0.0162, "step": 40520 }, { "grad_norm": 0.25490450859069824, "learning_rate": 5.3038444422186284e-05, "loss": 0.0159, "step": 40530 }, { "grad_norm": 0.1649860441684723, "learning_rate": 5.301781398843757e-05, "loss": 0.0128, "step": 40540 }, { "grad_norm": 0.2415364384651184, "learning_rate": 5.299718303902699e-05, "loss": 0.0191, "step": 40550 }, { "grad_norm": 0.2708030641078949, "learning_rate": 5.297655157747979e-05, "loss": 0.0182, "step": 40560 }, { "grad_norm": 0.3272817134857178, "learning_rate": 5.295591960732136e-05, "loss": 0.0154, "step": 40570 }, { "grad_norm": 0.2361634373664856, "learning_rate": 5.293528713207708e-05, "loss": 0.0197, "step": 40580 }, { "grad_norm": 0.17085985839366913, "learning_rate": 5.291465415527253e-05, "loss": 0.0196, "step": 40590 }, { "grad_norm": 0.2141404151916504, "learning_rate": 5.2894020680433296e-05, "loss": 0.0187, "step": 40600 }, { "grad_norm": 0.2567320466041565, "learning_rate": 5.287338671108507e-05, "loss": 0.0188, "step": 40610 }, { "grad_norm": 0.486779123544693, "learning_rate": 5.285275225075367e-05, "loss": 0.0177, "step": 40620 }, { "grad_norm": 0.24211053550243378, "learning_rate": 5.283211730296492e-05, "loss": 0.0176, "step": 40630 }, { "grad_norm": 0.2071545124053955, "learning_rate": 5.281148187124477e-05, "loss": 0.0187, "step": 40640 }, { "grad_norm": 0.211097851395607, "learning_rate": 5.279084595911927e-05, "loss": 0.0183, "step": 40650 }, { "grad_norm": 0.30067723989486694, "learning_rate": 5.277020957011453e-05, "loss": 0.0255, "step": 40660 }, { "grad_norm": 0.2562424838542938, "learning_rate": 5.274957270775673e-05, "loss": 0.0255, "step": 40670 }, { "grad_norm": 0.28139960765838623, "learning_rate": 5.2728935375572164e-05, "loss": 0.0149, "step": 40680 }, { "grad_norm": 0.19054941833019257, "learning_rate": 5.2708297577087176e-05, "loss": 0.0206, "step": 40690 }, { "grad_norm": 0.2372024953365326, "learning_rate": 5.2687659315828196e-05, "loss": 0.0149, "step": 40700 }, { "grad_norm": 0.2012869268655777, "learning_rate": 5.266702059532175e-05, "loss": 0.0151, "step": 40710 }, { "grad_norm": 0.2691461145877838, "learning_rate": 5.264638141909444e-05, "loss": 0.0205, "step": 40720 }, { "grad_norm": 0.21319444477558136, "learning_rate": 5.2625741790672925e-05, "loss": 0.0194, "step": 40730 }, { "grad_norm": 0.29948386549949646, "learning_rate": 5.2605101713583935e-05, "loss": 0.0162, "step": 40740 }, { "grad_norm": 0.2705387473106384, "learning_rate": 5.2584461191354315e-05, "loss": 0.0242, "step": 40750 }, { "grad_norm": 0.2089153230190277, "learning_rate": 5.2563820227510964e-05, "loss": 0.0165, "step": 40760 }, { "grad_norm": 0.22381147742271423, "learning_rate": 5.2543178825580864e-05, "loss": 0.0162, "step": 40770 }, { "grad_norm": 0.30541595816612244, "learning_rate": 5.2522536989091054e-05, "loss": 0.0194, "step": 40780 }, { "grad_norm": 0.25570476055145264, "learning_rate": 5.2501894721568655e-05, "loss": 0.0162, "step": 40790 }, { "grad_norm": 0.18467971682548523, "learning_rate": 5.248125202654089e-05, "loss": 0.0135, "step": 40800 }, { "grad_norm": 0.2603122889995575, "learning_rate": 5.246060890753501e-05, "loss": 0.0283, "step": 40810 }, { "grad_norm": 0.24109897017478943, "learning_rate": 5.243996536807837e-05, "loss": 0.0156, "step": 40820 }, { "grad_norm": 0.366982638835907, "learning_rate": 5.2419321411698384e-05, "loss": 0.0169, "step": 40830 }, { "grad_norm": 0.27026084065437317, "learning_rate": 5.239867704192253e-05, "loss": 0.0221, "step": 40840 }, { "grad_norm": 0.2953990399837494, "learning_rate": 5.237803226227838e-05, "loss": 0.017, "step": 40850 }, { "grad_norm": 0.1813458651304245, "learning_rate": 5.235738707629354e-05, "loss": 0.0164, "step": 40860 }, { "grad_norm": 0.26983699202537537, "learning_rate": 5.233674148749575e-05, "loss": 0.0198, "step": 40870 }, { "grad_norm": 0.3413364589214325, "learning_rate": 5.231609549941272e-05, "loss": 0.0244, "step": 40880 }, { "grad_norm": 0.21574753522872925, "learning_rate": 5.229544911557233e-05, "loss": 0.023, "step": 40890 }, { "grad_norm": 0.1284911334514618, "learning_rate": 5.227480233950246e-05, "loss": 0.0181, "step": 40900 }, { "grad_norm": 0.20604750514030457, "learning_rate": 5.22541551747311e-05, "loss": 0.02, "step": 40910 }, { "grad_norm": 0.20768298208713531, "learning_rate": 5.223350762478626e-05, "loss": 0.0137, "step": 40920 }, { "grad_norm": 0.23465949296951294, "learning_rate": 5.221285969319606e-05, "loss": 0.0152, "step": 40930 }, { "grad_norm": 0.2233252227306366, "learning_rate": 5.219221138348865e-05, "loss": 0.0136, "step": 40940 }, { "grad_norm": 0.20916903018951416, "learning_rate": 5.217156269919228e-05, "loss": 0.0189, "step": 40950 }, { "grad_norm": 0.1400802880525589, "learning_rate": 5.215091364383523e-05, "loss": 0.0161, "step": 40960 }, { "grad_norm": 0.31496480107307434, "learning_rate": 5.213026422094588e-05, "loss": 0.0205, "step": 40970 }, { "grad_norm": 0.14950686693191528, "learning_rate": 5.210961443405262e-05, "loss": 0.0145, "step": 40980 }, { "grad_norm": 0.1473088264465332, "learning_rate": 5.208896428668396e-05, "loss": 0.0115, "step": 40990 }, { "grad_norm": 0.3209363520145416, "learning_rate": 5.206831378236845e-05, "loss": 0.02, "step": 41000 }, { "grad_norm": 0.2025758922100067, "learning_rate": 5.2047662924634666e-05, "loss": 0.0128, "step": 41010 }, { "grad_norm": 0.3008691072463989, "learning_rate": 5.20270117170113e-05, "loss": 0.0133, "step": 41020 }, { "grad_norm": 0.34676623344421387, "learning_rate": 5.200636016302707e-05, "loss": 0.0239, "step": 41030 }, { "grad_norm": 0.24340683221817017, "learning_rate": 5.1985708266210754e-05, "loss": 0.0222, "step": 41040 }, { "grad_norm": 0.20172272622585297, "learning_rate": 5.1965056030091206e-05, "loss": 0.0194, "step": 41050 }, { "grad_norm": 0.24649901688098907, "learning_rate": 5.194440345819732e-05, "loss": 0.0112, "step": 41060 }, { "grad_norm": 0.21927236020565033, "learning_rate": 5.1923750554058084e-05, "loss": 0.0202, "step": 41070 }, { "grad_norm": 0.1672237515449524, "learning_rate": 5.1903097321202476e-05, "loss": 0.0144, "step": 41080 }, { "grad_norm": 0.28953441977500916, "learning_rate": 5.18824437631596e-05, "loss": 0.0326, "step": 41090 }, { "grad_norm": 0.2839179039001465, "learning_rate": 5.186178988345856e-05, "loss": 0.0198, "step": 41100 }, { "grad_norm": 0.2739031910896301, "learning_rate": 5.184113568562855e-05, "loss": 0.0201, "step": 41110 }, { "grad_norm": 0.31970667839050293, "learning_rate": 5.18204811731988e-05, "loss": 0.0249, "step": 41120 }, { "grad_norm": 0.3119506537914276, "learning_rate": 5.179982634969862e-05, "loss": 0.0216, "step": 41130 }, { "grad_norm": 0.4337979257106781, "learning_rate": 5.177917121865734e-05, "loss": 0.0237, "step": 41140 }, { "grad_norm": 0.13516190648078918, "learning_rate": 5.1758515783604346e-05, "loss": 0.0144, "step": 41150 }, { "grad_norm": 0.2575002610683441, "learning_rate": 5.173786004806911e-05, "loss": 0.0224, "step": 41160 }, { "grad_norm": 0.358966588973999, "learning_rate": 5.1717204015581135e-05, "loss": 0.0209, "step": 41170 }, { "grad_norm": 0.18433311581611633, "learning_rate": 5.169654768966994e-05, "loss": 0.0151, "step": 41180 }, { "grad_norm": 0.3075278401374817, "learning_rate": 5.1675891073865156e-05, "loss": 0.0208, "step": 41190 }, { "grad_norm": 0.18611235916614532, "learning_rate": 5.1655234171696424e-05, "loss": 0.0154, "step": 41200 }, { "grad_norm": 0.3751266300678253, "learning_rate": 5.163457698669343e-05, "loss": 0.0177, "step": 41210 }, { "grad_norm": 0.1666465848684311, "learning_rate": 5.1613919522385936e-05, "loss": 0.0148, "step": 41220 }, { "grad_norm": 0.26474523544311523, "learning_rate": 5.1593261782303746e-05, "loss": 0.0281, "step": 41230 }, { "grad_norm": 0.20239032804965973, "learning_rate": 5.157260376997669e-05, "loss": 0.0143, "step": 41240 }, { "grad_norm": 0.20592918992042542, "learning_rate": 5.155194548893466e-05, "loss": 0.0196, "step": 41250 }, { "grad_norm": 0.18653123080730438, "learning_rate": 5.153128694270759e-05, "loss": 0.0283, "step": 41260 }, { "grad_norm": 0.2491372674703598, "learning_rate": 5.151062813482548e-05, "loss": 0.024, "step": 41270 }, { "grad_norm": 0.2648584842681885, "learning_rate": 5.148996906881832e-05, "loss": 0.0175, "step": 41280 }, { "grad_norm": 0.3225833475589752, "learning_rate": 5.1469309748216196e-05, "loss": 0.015, "step": 41290 }, { "grad_norm": 0.21700671315193176, "learning_rate": 5.144865017654923e-05, "loss": 0.0133, "step": 41300 }, { "grad_norm": 0.45267540216445923, "learning_rate": 5.1427990357347556e-05, "loss": 0.0227, "step": 41310 }, { "grad_norm": 0.23439501225948334, "learning_rate": 5.140733029414139e-05, "loss": 0.0192, "step": 41320 }, { "grad_norm": 0.3435969650745392, "learning_rate": 5.138666999046098e-05, "loss": 0.0197, "step": 41330 }, { "grad_norm": 0.2585790455341339, "learning_rate": 5.136600944983658e-05, "loss": 0.0134, "step": 41340 }, { "grad_norm": 0.22553180158138275, "learning_rate": 5.134534867579853e-05, "loss": 0.0171, "step": 41350 }, { "grad_norm": 0.26361867785453796, "learning_rate": 5.1324687671877184e-05, "loss": 0.0146, "step": 41360 }, { "grad_norm": 0.31260859966278076, "learning_rate": 5.130402644160296e-05, "loss": 0.0192, "step": 41370 }, { "grad_norm": 0.2272499054670334, "learning_rate": 5.128336498850628e-05, "loss": 0.0167, "step": 41380 }, { "grad_norm": 0.21173620223999023, "learning_rate": 5.126270331611761e-05, "loss": 0.0148, "step": 41390 }, { "grad_norm": 0.2489870935678482, "learning_rate": 5.124204142796748e-05, "loss": 0.0154, "step": 41400 }, { "grad_norm": 0.13505801558494568, "learning_rate": 5.122137932758644e-05, "loss": 0.0141, "step": 41410 }, { "grad_norm": 0.17786391079425812, "learning_rate": 5.1200717018505074e-05, "loss": 0.013, "step": 41420 }, { "grad_norm": 0.2799195945262909, "learning_rate": 5.118005450425403e-05, "loss": 0.0138, "step": 41430 }, { "grad_norm": 0.16238312423229218, "learning_rate": 5.115939178836392e-05, "loss": 0.0106, "step": 41440 }, { "grad_norm": 0.19798576831817627, "learning_rate": 5.113872887436547e-05, "loss": 0.0158, "step": 41450 }, { "grad_norm": 0.17717911303043365, "learning_rate": 5.11180657657894e-05, "loss": 0.0224, "step": 41460 }, { "grad_norm": 0.20685572922229767, "learning_rate": 5.109740246616649e-05, "loss": 0.014, "step": 41470 }, { "grad_norm": 0.3818662166595459, "learning_rate": 5.10767389790275e-05, "loss": 0.0234, "step": 41480 }, { "grad_norm": 0.19700577855110168, "learning_rate": 5.1056075307903265e-05, "loss": 0.0141, "step": 41490 }, { "grad_norm": 0.1232951357960701, "learning_rate": 5.1035411456324634e-05, "loss": 0.0138, "step": 41500 }, { "grad_norm": 0.26735952496528625, "learning_rate": 5.1014747427822504e-05, "loss": 0.0206, "step": 41510 }, { "grad_norm": 0.3339139223098755, "learning_rate": 5.0994083225927804e-05, "loss": 0.0206, "step": 41520 }, { "grad_norm": 0.2174021452665329, "learning_rate": 5.0973418854171475e-05, "loss": 0.0215, "step": 41530 }, { "grad_norm": 0.17105187475681305, "learning_rate": 5.095275431608447e-05, "loss": 0.0171, "step": 41540 }, { "grad_norm": 0.23542509973049164, "learning_rate": 5.093208961519782e-05, "loss": 0.0161, "step": 41550 }, { "grad_norm": 0.1871601790189743, "learning_rate": 5.091142475504255e-05, "loss": 0.0134, "step": 41560 }, { "grad_norm": 0.3699232339859009, "learning_rate": 5.089075973914971e-05, "loss": 0.0235, "step": 41570 }, { "grad_norm": 0.2720256447792053, "learning_rate": 5.08700945710504e-05, "loss": 0.0287, "step": 41580 }, { "grad_norm": 0.24581435322761536, "learning_rate": 5.0849429254275714e-05, "loss": 0.0154, "step": 41590 }, { "grad_norm": 0.14299730956554413, "learning_rate": 5.0828763792356804e-05, "loss": 0.0106, "step": 41600 }, { "grad_norm": 0.27603814005851746, "learning_rate": 5.080809818882483e-05, "loss": 0.0205, "step": 41610 }, { "grad_norm": 0.3209514915943146, "learning_rate": 5.078743244721097e-05, "loss": 0.017, "step": 41620 }, { "grad_norm": 0.1693614423274994, "learning_rate": 5.0766766571046455e-05, "loss": 0.0178, "step": 41630 }, { "grad_norm": 0.23231565952301025, "learning_rate": 5.07461005638625e-05, "loss": 0.013, "step": 41640 }, { "grad_norm": 0.2697560787200928, "learning_rate": 5.072543442919037e-05, "loss": 0.0157, "step": 41650 }, { "grad_norm": 0.2297571450471878, "learning_rate": 5.070476817056132e-05, "loss": 0.0232, "step": 41660 }, { "grad_norm": 0.2203335016965866, "learning_rate": 5.068410179150668e-05, "loss": 0.0138, "step": 41670 }, { "grad_norm": 0.236046701669693, "learning_rate": 5.066343529555775e-05, "loss": 0.0165, "step": 41680 }, { "grad_norm": 0.16143731772899628, "learning_rate": 5.064276868624587e-05, "loss": 0.0154, "step": 41690 }, { "grad_norm": 0.18602748215198517, "learning_rate": 5.0622101967102396e-05, "loss": 0.0158, "step": 41700 }, { "grad_norm": 0.23954267799854279, "learning_rate": 5.0601435141658714e-05, "loss": 0.0155, "step": 41710 }, { "grad_norm": 0.32783398032188416, "learning_rate": 5.058076821344622e-05, "loss": 0.0218, "step": 41720 }, { "grad_norm": 0.2827344238758087, "learning_rate": 5.056010118599632e-05, "loss": 0.0247, "step": 41730 }, { "grad_norm": 0.3519423305988312, "learning_rate": 5.053943406284044e-05, "loss": 0.0176, "step": 41740 }, { "grad_norm": 0.26899850368499756, "learning_rate": 5.051876684751006e-05, "loss": 0.0189, "step": 41750 }, { "grad_norm": 0.25769081711769104, "learning_rate": 5.0498099543536584e-05, "loss": 0.0182, "step": 41760 }, { "grad_norm": 0.22345438599586487, "learning_rate": 5.047743215445152e-05, "loss": 0.0219, "step": 41770 }, { "grad_norm": 0.24751633405685425, "learning_rate": 5.045676468378637e-05, "loss": 0.0145, "step": 41780 }, { "grad_norm": 0.32784605026245117, "learning_rate": 5.0436097135072626e-05, "loss": 0.0199, "step": 41790 }, { "grad_norm": 0.22088387608528137, "learning_rate": 5.041542951184181e-05, "loss": 0.0165, "step": 41800 }, { "grad_norm": 0.17890280485153198, "learning_rate": 5.039476181762545e-05, "loss": 0.0127, "step": 41810 }, { "grad_norm": 0.16341005265712738, "learning_rate": 5.037409405595508e-05, "loss": 0.0106, "step": 41820 }, { "grad_norm": 0.3063884675502777, "learning_rate": 5.035342623036229e-05, "loss": 0.0163, "step": 41830 }, { "grad_norm": 0.2117612212896347, "learning_rate": 5.033275834437862e-05, "loss": 0.0208, "step": 41840 }, { "grad_norm": 0.26317909359931946, "learning_rate": 5.031209040153564e-05, "loss": 0.0172, "step": 41850 }, { "grad_norm": 0.19633004069328308, "learning_rate": 5.0291422405364955e-05, "loss": 0.0293, "step": 41860 }, { "grad_norm": 0.21272923052310944, "learning_rate": 5.0270754359398133e-05, "loss": 0.014, "step": 41870 }, { "grad_norm": 0.19943933188915253, "learning_rate": 5.025008626716682e-05, "loss": 0.0144, "step": 41880 }, { "grad_norm": 0.14043395221233368, "learning_rate": 5.0229418132202585e-05, "loss": 0.0174, "step": 41890 }, { "grad_norm": 0.3504430949687958, "learning_rate": 5.020874995803707e-05, "loss": 0.021, "step": 41900 }, { "grad_norm": 0.2585729956626892, "learning_rate": 5.01880817482019e-05, "loss": 0.0182, "step": 41910 }, { "grad_norm": 0.11622601747512817, "learning_rate": 5.01674135062287e-05, "loss": 0.0139, "step": 41920 }, { "grad_norm": 0.22536446154117584, "learning_rate": 5.0146745235649115e-05, "loss": 0.0168, "step": 41930 }, { "grad_norm": 0.35102739930152893, "learning_rate": 5.0126076939994795e-05, "loss": 0.0133, "step": 41940 }, { "grad_norm": 0.1459423154592514, "learning_rate": 5.010540862279736e-05, "loss": 0.0197, "step": 41950 }, { "grad_norm": 0.2760160565376282, "learning_rate": 5.008474028758846e-05, "loss": 0.0171, "step": 41960 }, { "grad_norm": 0.3158223330974579, "learning_rate": 5.0064071937899784e-05, "loss": 0.0162, "step": 41970 }, { "grad_norm": 0.302282452583313, "learning_rate": 5.004340357726296e-05, "loss": 0.0155, "step": 41980 }, { "grad_norm": 0.23765674233436584, "learning_rate": 5.002273520920965e-05, "loss": 0.0217, "step": 41990 }, { "grad_norm": 0.16402573883533478, "learning_rate": 5.000206683727151e-05, "loss": 0.0144, "step": 42000 }, { "grad_norm": 0.11839977651834488, "learning_rate": 4.998139846498021e-05, "loss": 0.0133, "step": 42010 }, { "grad_norm": 0.2307921200990677, "learning_rate": 4.9960730095867405e-05, "loss": 0.0173, "step": 42020 }, { "grad_norm": 0.22779043018817902, "learning_rate": 4.9940061733464755e-05, "loss": 0.0228, "step": 42030 }, { "grad_norm": 0.18658903241157532, "learning_rate": 4.991939338130392e-05, "loss": 0.0274, "step": 42040 }, { "grad_norm": 0.44656550884246826, "learning_rate": 4.989872504291653e-05, "loss": 0.0291, "step": 42050 }, { "grad_norm": 0.4263274669647217, "learning_rate": 4.9878056721834273e-05, "loss": 0.0173, "step": 42060 }, { "grad_norm": 0.28850364685058594, "learning_rate": 4.98573884215888e-05, "loss": 0.0175, "step": 42070 }, { "grad_norm": 0.3342483639717102, "learning_rate": 4.9836720145711715e-05, "loss": 0.0227, "step": 42080 }, { "grad_norm": 0.24429942667484283, "learning_rate": 4.9816051897734725e-05, "loss": 0.0198, "step": 42090 }, { "grad_norm": 0.23375755548477173, "learning_rate": 4.979538368118941e-05, "loss": 0.0211, "step": 42100 }, { "grad_norm": 0.2536936402320862, "learning_rate": 4.9774715499607446e-05, "loss": 0.0173, "step": 42110 }, { "grad_norm": 0.18818683922290802, "learning_rate": 4.9754047356520416e-05, "loss": 0.0241, "step": 42120 }, { "grad_norm": 0.214823916554451, "learning_rate": 4.973337925545999e-05, "loss": 0.0177, "step": 42130 }, { "grad_norm": 0.24613118171691895, "learning_rate": 4.9712711199957746e-05, "loss": 0.0153, "step": 42140 }, { "grad_norm": 0.2515344023704529, "learning_rate": 4.969204319354531e-05, "loss": 0.0214, "step": 42150 }, { "grad_norm": 0.3365756571292877, "learning_rate": 4.9671375239754267e-05, "loss": 0.016, "step": 42160 }, { "grad_norm": 0.19124636054039001, "learning_rate": 4.9650707342116195e-05, "loss": 0.014, "step": 42170 }, { "grad_norm": 0.19125258922576904, "learning_rate": 4.96300395041627e-05, "loss": 0.0154, "step": 42180 }, { "grad_norm": 0.23396988213062286, "learning_rate": 4.960937172942532e-05, "loss": 0.0178, "step": 42190 }, { "grad_norm": 0.18674036860466003, "learning_rate": 4.958870402143565e-05, "loss": 0.0139, "step": 42200 }, { "grad_norm": 0.17722122371196747, "learning_rate": 4.9568036383725186e-05, "loss": 0.0188, "step": 42210 }, { "grad_norm": 0.27890896797180176, "learning_rate": 4.9547368819825516e-05, "loss": 0.0161, "step": 42220 }, { "grad_norm": 0.2214491367340088, "learning_rate": 4.952670133326812e-05, "loss": 0.0165, "step": 42230 }, { "grad_norm": 0.3173501193523407, "learning_rate": 4.950603392758453e-05, "loss": 0.022, "step": 42240 }, { "grad_norm": 0.21323980391025543, "learning_rate": 4.948536660630621e-05, "loss": 0.0149, "step": 42250 }, { "grad_norm": 0.2667178809642792, "learning_rate": 4.9464699372964685e-05, "loss": 0.0185, "step": 42260 }, { "grad_norm": 0.2579270601272583, "learning_rate": 4.9444032231091395e-05, "loss": 0.0189, "step": 42270 }, { "grad_norm": 0.2107168436050415, "learning_rate": 4.9423365184217776e-05, "loss": 0.0154, "step": 42280 }, { "grad_norm": 0.26531222462654114, "learning_rate": 4.940269823587529e-05, "loss": 0.0161, "step": 42290 }, { "grad_norm": 0.14778971672058105, "learning_rate": 4.938203138959533e-05, "loss": 0.0146, "step": 42300 }, { "grad_norm": 0.2983139157295227, "learning_rate": 4.9361364648909325e-05, "loss": 0.017, "step": 42310 }, { "grad_norm": 0.1887986660003662, "learning_rate": 4.9340698017348605e-05, "loss": 0.0135, "step": 42320 }, { "grad_norm": 0.22930489480495453, "learning_rate": 4.932003149844458e-05, "loss": 0.0151, "step": 42330 }, { "grad_norm": 0.183763787150383, "learning_rate": 4.929936509572857e-05, "loss": 0.0268, "step": 42340 }, { "grad_norm": 0.18313103914260864, "learning_rate": 4.927869881273191e-05, "loss": 0.0169, "step": 42350 }, { "grad_norm": 0.23121331632137299, "learning_rate": 4.9258032652985894e-05, "loss": 0.0163, "step": 42360 }, { "grad_norm": 0.20377226173877716, "learning_rate": 4.9237366620021786e-05, "loss": 0.0225, "step": 42370 }, { "grad_norm": 0.19778862595558167, "learning_rate": 4.921670071737089e-05, "loss": 0.0128, "step": 42380 }, { "grad_norm": 0.31219321489334106, "learning_rate": 4.91960349485644e-05, "loss": 0.022, "step": 42390 }, { "grad_norm": 0.34909987449645996, "learning_rate": 4.9175369317133566e-05, "loss": 0.0195, "step": 42400 }, { "grad_norm": 0.20883384346961975, "learning_rate": 4.915470382660953e-05, "loss": 0.0174, "step": 42410 }, { "grad_norm": 0.17373938858509064, "learning_rate": 4.9134038480523524e-05, "loss": 0.0152, "step": 42420 }, { "grad_norm": 0.1894906908273697, "learning_rate": 4.911337328240664e-05, "loss": 0.0211, "step": 42430 }, { "grad_norm": 0.19682136178016663, "learning_rate": 4.909270823579003e-05, "loss": 0.0171, "step": 42440 }, { "grad_norm": 0.23052026331424713, "learning_rate": 4.907204334420476e-05, "loss": 0.0172, "step": 42450 }, { "grad_norm": 0.20435594022274017, "learning_rate": 4.9051378611181884e-05, "loss": 0.0182, "step": 42460 }, { "grad_norm": 0.21204450726509094, "learning_rate": 4.9030714040252486e-05, "loss": 0.0158, "step": 42470 }, { "grad_norm": 0.22409135103225708, "learning_rate": 4.901004963494752e-05, "loss": 0.0114, "step": 42480 }, { "grad_norm": 0.19256901741027832, "learning_rate": 4.898938539879802e-05, "loss": 0.0184, "step": 42490 }, { "grad_norm": 0.23537816107273102, "learning_rate": 4.8968721335334904e-05, "loss": 0.0153, "step": 42500 }, { "grad_norm": 0.21089453995227814, "learning_rate": 4.894805744808912e-05, "loss": 0.0196, "step": 42510 }, { "grad_norm": 0.2213401347398758, "learning_rate": 4.8927393740591524e-05, "loss": 0.0177, "step": 42520 }, { "grad_norm": 0.2459680736064911, "learning_rate": 4.890673021637302e-05, "loss": 0.0157, "step": 42530 }, { "grad_norm": 0.13473226130008698, "learning_rate": 4.888606687896442e-05, "loss": 0.0122, "step": 42540 }, { "grad_norm": 0.21978339552879333, "learning_rate": 4.886540373189652e-05, "loss": 0.0148, "step": 42550 }, { "grad_norm": 0.17180891335010529, "learning_rate": 4.88447407787001e-05, "loss": 0.0135, "step": 42560 }, { "grad_norm": 0.20559915900230408, "learning_rate": 4.8824078022905864e-05, "loss": 0.0158, "step": 42570 }, { "grad_norm": 0.2759209871292114, "learning_rate": 4.880341546804455e-05, "loss": 0.012, "step": 42580 }, { "grad_norm": 0.2984883189201355, "learning_rate": 4.878275311764679e-05, "loss": 0.0193, "step": 42590 }, { "grad_norm": 0.2340039759874344, "learning_rate": 4.8762090975243254e-05, "loss": 0.0148, "step": 42600 }, { "grad_norm": 0.1733052134513855, "learning_rate": 4.874142904436448e-05, "loss": 0.016, "step": 42610 }, { "grad_norm": 0.17522113025188446, "learning_rate": 4.8720767328541094e-05, "loss": 0.0252, "step": 42620 }, { "grad_norm": 0.28122687339782715, "learning_rate": 4.870010583130358e-05, "loss": 0.0163, "step": 42630 }, { "grad_norm": 0.31389138102531433, "learning_rate": 4.867944455618241e-05, "loss": 0.0151, "step": 42640 }, { "grad_norm": 0.2119641900062561, "learning_rate": 4.865878350670807e-05, "loss": 0.0201, "step": 42650 }, { "grad_norm": 0.4071997404098511, "learning_rate": 4.8638122686410914e-05, "loss": 0.0187, "step": 42660 }, { "grad_norm": 0.12627512216567993, "learning_rate": 4.861746209882137e-05, "loss": 0.0108, "step": 42670 }, { "grad_norm": 0.23778140544891357, "learning_rate": 4.859680174746972e-05, "loss": 0.0179, "step": 42680 }, { "grad_norm": 0.23618963360786438, "learning_rate": 4.857614163588629e-05, "loss": 0.0239, "step": 42690 }, { "grad_norm": 0.2671301066875458, "learning_rate": 4.855548176760131e-05, "loss": 0.0151, "step": 42700 }, { "grad_norm": 0.13577406108379364, "learning_rate": 4.8534822146144986e-05, "loss": 0.0168, "step": 42710 }, { "grad_norm": 0.2600196897983551, "learning_rate": 4.851416277504749e-05, "loss": 0.0255, "step": 42720 }, { "grad_norm": 0.22244958579540253, "learning_rate": 4.8493503657838923e-05, "loss": 0.012, "step": 42730 }, { "grad_norm": 0.13977457582950592, "learning_rate": 4.84728447980494e-05, "loss": 0.0151, "step": 42740 }, { "grad_norm": 0.27851593494415283, "learning_rate": 4.8452186199208914e-05, "loss": 0.0194, "step": 42750 }, { "grad_norm": 0.2704361081123352, "learning_rate": 4.843152786484749e-05, "loss": 0.0177, "step": 42760 }, { "grad_norm": 0.15421512722969055, "learning_rate": 4.8410869798495035e-05, "loss": 0.013, "step": 42770 }, { "grad_norm": 0.15238675475120544, "learning_rate": 4.8390212003681486e-05, "loss": 0.018, "step": 42780 }, { "grad_norm": 0.2749163508415222, "learning_rate": 4.836955448393667e-05, "loss": 0.0194, "step": 42790 }, { "grad_norm": 0.39641621708869934, "learning_rate": 4.8348897242790406e-05, "loss": 0.0177, "step": 42800 }, { "grad_norm": 0.20243029296398163, "learning_rate": 4.832824028377243e-05, "loss": 0.0158, "step": 42810 }, { "grad_norm": 0.21176555752754211, "learning_rate": 4.830758361041249e-05, "loss": 0.0165, "step": 42820 }, { "grad_norm": 0.40443292260169983, "learning_rate": 4.828692722624022e-05, "loss": 0.0195, "step": 42830 }, { "grad_norm": 0.21133333444595337, "learning_rate": 4.826627113478522e-05, "loss": 0.0344, "step": 42840 }, { "grad_norm": 0.2415618598461151, "learning_rate": 4.824561533957708e-05, "loss": 0.015, "step": 42850 }, { "grad_norm": 0.2038147896528244, "learning_rate": 4.822495984414527e-05, "loss": 0.0134, "step": 42860 }, { "grad_norm": 0.2212570756673813, "learning_rate": 4.8204304652019304e-05, "loss": 0.0143, "step": 42870 }, { "grad_norm": 0.23333171010017395, "learning_rate": 4.8183649766728537e-05, "loss": 0.0151, "step": 42880 }, { "grad_norm": 0.3120250105857849, "learning_rate": 4.816299519180237e-05, "loss": 0.0198, "step": 42890 }, { "grad_norm": 0.20868559181690216, "learning_rate": 4.8142340930770066e-05, "loss": 0.014, "step": 42900 }, { "grad_norm": 0.16085129976272583, "learning_rate": 4.81216869871609e-05, "loss": 0.0114, "step": 42910 }, { "grad_norm": 0.20625457167625427, "learning_rate": 4.8101033364504064e-05, "loss": 0.0139, "step": 42920 }, { "grad_norm": 0.15516628324985504, "learning_rate": 4.808038006632866e-05, "loss": 0.0244, "step": 42930 }, { "grad_norm": 0.22849304974079132, "learning_rate": 4.805972709616383e-05, "loss": 0.0157, "step": 42940 }, { "grad_norm": 0.17652669548988342, "learning_rate": 4.8039074457538556e-05, "loss": 0.0111, "step": 42950 }, { "grad_norm": 0.2365439236164093, "learning_rate": 4.801842215398184e-05, "loss": 0.0207, "step": 42960 }, { "grad_norm": 0.25946709513664246, "learning_rate": 4.799777018902256e-05, "loss": 0.024, "step": 42970 }, { "grad_norm": 0.2884349524974823, "learning_rate": 4.797711856618961e-05, "loss": 0.016, "step": 42980 }, { "grad_norm": 0.20217865705490112, "learning_rate": 4.7956467289011765e-05, "loss": 0.0214, "step": 42990 }, { "grad_norm": 0.2041839361190796, "learning_rate": 4.793581636101778e-05, "loss": 0.0163, "step": 43000 }, { "grad_norm": 0.2095564901828766, "learning_rate": 4.7915165785736326e-05, "loss": 0.0158, "step": 43010 }, { "grad_norm": 0.2430124580860138, "learning_rate": 4.789451556669599e-05, "loss": 0.0135, "step": 43020 }, { "grad_norm": 0.17272143065929413, "learning_rate": 4.7873865707425384e-05, "loss": 0.0211, "step": 43030 }, { "grad_norm": 0.2634792923927307, "learning_rate": 4.7853216211452975e-05, "loss": 0.0169, "step": 43040 }, { "grad_norm": 0.2797538638114929, "learning_rate": 4.78325670823072e-05, "loss": 0.0169, "step": 43050 }, { "grad_norm": 0.16129323840141296, "learning_rate": 4.781191832351641e-05, "loss": 0.0166, "step": 43060 }, { "grad_norm": 0.34473761916160583, "learning_rate": 4.7791269938608965e-05, "loss": 0.0197, "step": 43070 }, { "grad_norm": 0.20106162130832672, "learning_rate": 4.777062193111305e-05, "loss": 0.0178, "step": 43080 }, { "grad_norm": 0.19067628681659698, "learning_rate": 4.77499743045569e-05, "loss": 0.0167, "step": 43090 }, { "grad_norm": 0.17363674938678741, "learning_rate": 4.7729327062468596e-05, "loss": 0.0159, "step": 43100 }, { "grad_norm": 0.3162558972835541, "learning_rate": 4.770868020837619e-05, "loss": 0.0206, "step": 43110 }, { "grad_norm": 0.24947191774845123, "learning_rate": 4.768803374580768e-05, "loss": 0.0141, "step": 43120 }, { "grad_norm": 0.23832112550735474, "learning_rate": 4.766738767829094e-05, "loss": 0.0152, "step": 43130 }, { "grad_norm": 0.19101610779762268, "learning_rate": 4.764674200935388e-05, "loss": 0.0151, "step": 43140 }, { "grad_norm": 0.20997542142868042, "learning_rate": 4.762609674252424e-05, "loss": 0.0159, "step": 43150 }, { "grad_norm": 0.19171077013015747, "learning_rate": 4.760545188132974e-05, "loss": 0.0298, "step": 43160 }, { "grad_norm": 0.1851639300584793, "learning_rate": 4.758480742929801e-05, "loss": 0.0158, "step": 43170 }, { "grad_norm": 0.2283579260110855, "learning_rate": 4.756416338995664e-05, "loss": 0.0224, "step": 43180 }, { "grad_norm": 0.20228344202041626, "learning_rate": 4.7543519766833145e-05, "loss": 0.0156, "step": 43190 }, { "grad_norm": 0.16769015789031982, "learning_rate": 4.752287656345492e-05, "loss": 0.0155, "step": 43200 }, { "grad_norm": 0.25382864475250244, "learning_rate": 4.7502233783349355e-05, "loss": 0.015, "step": 43210 }, { "grad_norm": 0.19954490661621094, "learning_rate": 4.7481591430043694e-05, "loss": 0.0107, "step": 43220 }, { "grad_norm": 0.2822725772857666, "learning_rate": 4.746094950706521e-05, "loss": 0.0201, "step": 43230 }, { "grad_norm": 0.1275016963481903, "learning_rate": 4.744030801794099e-05, "loss": 0.0172, "step": 43240 }, { "grad_norm": 0.18029414117336273, "learning_rate": 4.741966696619813e-05, "loss": 0.0224, "step": 43250 }, { "grad_norm": 0.22822852432727814, "learning_rate": 4.739902635536361e-05, "loss": 0.0237, "step": 43260 }, { "grad_norm": 0.23508237302303314, "learning_rate": 4.737838618896436e-05, "loss": 0.0146, "step": 43270 }, { "grad_norm": 0.12288524210453033, "learning_rate": 4.7357746470527203e-05, "loss": 0.0178, "step": 43280 }, { "grad_norm": 0.22753506898880005, "learning_rate": 4.73371072035789e-05, "loss": 0.0189, "step": 43290 }, { "grad_norm": 0.2005021870136261, "learning_rate": 4.731646839164616e-05, "loss": 0.0184, "step": 43300 }, { "grad_norm": 0.27216455340385437, "learning_rate": 4.729583003825556e-05, "loss": 0.0164, "step": 43310 }, { "grad_norm": 0.18934577703475952, "learning_rate": 4.727519214693367e-05, "loss": 0.0186, "step": 43320 }, { "grad_norm": 0.3029271960258484, "learning_rate": 4.725455472120689e-05, "loss": 0.0194, "step": 43330 }, { "grad_norm": 0.23473221063613892, "learning_rate": 4.723391776460164e-05, "loss": 0.0199, "step": 43340 }, { "grad_norm": 0.1716386079788208, "learning_rate": 4.7213281280644186e-05, "loss": 0.0202, "step": 43350 }, { "grad_norm": 0.2759532332420349, "learning_rate": 4.719264527286075e-05, "loss": 0.016, "step": 43360 }, { "grad_norm": 0.2239551544189453, "learning_rate": 4.717200974477744e-05, "loss": 0.0217, "step": 43370 }, { "grad_norm": 0.2701862156391144, "learning_rate": 4.715137469992034e-05, "loss": 0.0181, "step": 43380 }, { "grad_norm": 0.23139147460460663, "learning_rate": 4.713074014181539e-05, "loss": 0.0152, "step": 43390 }, { "grad_norm": 0.17878387868404388, "learning_rate": 4.711010607398846e-05, "loss": 0.0131, "step": 43400 }, { "grad_norm": 0.2142796367406845, "learning_rate": 4.708947249996538e-05, "loss": 0.0112, "step": 43410 }, { "grad_norm": 0.15972313284873962, "learning_rate": 4.706883942327183e-05, "loss": 0.0136, "step": 43420 }, { "grad_norm": 0.34710755944252014, "learning_rate": 4.704820684743347e-05, "loss": 0.0182, "step": 43430 }, { "grad_norm": 0.3066822290420532, "learning_rate": 4.702757477597581e-05, "loss": 0.0261, "step": 43440 }, { "grad_norm": 0.16819632053375244, "learning_rate": 4.700694321242435e-05, "loss": 0.0184, "step": 43450 }, { "grad_norm": 0.3187330961227417, "learning_rate": 4.698631216030441e-05, "loss": 0.0166, "step": 43460 }, { "grad_norm": 0.17867743968963623, "learning_rate": 4.6965681623141314e-05, "loss": 0.0146, "step": 43470 }, { "grad_norm": 0.16010083258152008, "learning_rate": 4.694505160446024e-05, "loss": 0.0125, "step": 43480 }, { "grad_norm": 0.2528391480445862, "learning_rate": 4.692442210778628e-05, "loss": 0.0154, "step": 43490 }, { "grad_norm": 0.37105321884155273, "learning_rate": 4.690379313664448e-05, "loss": 0.0163, "step": 43500 }, { "grad_norm": 0.2804955542087555, "learning_rate": 4.688316469455973e-05, "loss": 0.0171, "step": 43510 }, { "grad_norm": 0.27016058564186096, "learning_rate": 4.6862536785056914e-05, "loss": 0.0144, "step": 43520 }, { "grad_norm": 0.22004300355911255, "learning_rate": 4.684190941166073e-05, "loss": 0.0151, "step": 43530 }, { "grad_norm": 0.1563119888305664, "learning_rate": 4.682128257789587e-05, "loss": 0.0123, "step": 43540 }, { "grad_norm": 0.28337740898132324, "learning_rate": 4.680065628728687e-05, "loss": 0.0131, "step": 43550 }, { "grad_norm": 0.23094482719898224, "learning_rate": 4.678003054335822e-05, "loss": 0.0177, "step": 43560 }, { "grad_norm": 0.24452228844165802, "learning_rate": 4.675940534963428e-05, "loss": 0.0203, "step": 43570 }, { "grad_norm": 0.3724339008331299, "learning_rate": 4.673878070963931e-05, "loss": 0.0175, "step": 43580 }, { "grad_norm": 0.1904250532388687, "learning_rate": 4.671815662689756e-05, "loss": 0.0112, "step": 43590 }, { "grad_norm": 0.1906314492225647, "learning_rate": 4.669753310493306e-05, "loss": 0.0147, "step": 43600 }, { "grad_norm": 0.19756019115447998, "learning_rate": 4.6676910147269845e-05, "loss": 0.018, "step": 43610 }, { "grad_norm": 0.3034808337688446, "learning_rate": 4.665628775743177e-05, "loss": 0.0162, "step": 43620 }, { "grad_norm": 0.21266292035579681, "learning_rate": 4.66356659389427e-05, "loss": 0.0155, "step": 43630 }, { "grad_norm": 0.2898203134536743, "learning_rate": 4.661504469532629e-05, "loss": 0.0162, "step": 43640 }, { "grad_norm": 0.19143615663051605, "learning_rate": 4.659442403010617e-05, "loss": 0.0157, "step": 43650 }, { "grad_norm": 0.25725144147872925, "learning_rate": 4.6573803946805845e-05, "loss": 0.0118, "step": 43660 }, { "grad_norm": 0.20720136165618896, "learning_rate": 4.6553184448948695e-05, "loss": 0.018, "step": 43670 }, { "grad_norm": 0.19020766019821167, "learning_rate": 4.653256554005807e-05, "loss": 0.0122, "step": 43680 }, { "grad_norm": 0.21496213972568512, "learning_rate": 4.6511947223657145e-05, "loss": 0.0211, "step": 43690 }, { "grad_norm": 0.25323179364204407, "learning_rate": 4.649132950326906e-05, "loss": 0.0177, "step": 43700 }, { "grad_norm": 0.2956455647945404, "learning_rate": 4.647071238241679e-05, "loss": 0.018, "step": 43710 }, { "grad_norm": 0.23149904608726501, "learning_rate": 4.6450095864623264e-05, "loss": 0.0128, "step": 43720 }, { "grad_norm": 0.18485423922538757, "learning_rate": 4.642947995341123e-05, "loss": 0.0107, "step": 43730 }, { "grad_norm": 0.1998690515756607, "learning_rate": 4.640886465230345e-05, "loss": 0.0176, "step": 43740 }, { "grad_norm": 0.23426668345928192, "learning_rate": 4.6388249964822485e-05, "loss": 0.016, "step": 43750 }, { "grad_norm": 0.2179674506187439, "learning_rate": 4.63676358944908e-05, "loss": 0.0138, "step": 43760 }, { "grad_norm": 0.340362012386322, "learning_rate": 4.634702244483081e-05, "loss": 0.021, "step": 43770 }, { "grad_norm": 0.20459730923175812, "learning_rate": 4.632640961936475e-05, "loss": 0.017, "step": 43780 }, { "grad_norm": 0.4398461878299713, "learning_rate": 4.6305797421614835e-05, "loss": 0.015, "step": 43790 }, { "grad_norm": 0.2806663513183594, "learning_rate": 4.6285185855103105e-05, "loss": 0.0176, "step": 43800 }, { "grad_norm": 0.25285106897354126, "learning_rate": 4.626457492335151e-05, "loss": 0.0165, "step": 43810 }, { "grad_norm": 0.23468267917633057, "learning_rate": 4.624396462988188e-05, "loss": 0.0131, "step": 43820 }, { "grad_norm": 0.1662662774324417, "learning_rate": 4.6223354978215985e-05, "loss": 0.0174, "step": 43830 }, { "grad_norm": 0.1748388707637787, "learning_rate": 4.620274597187544e-05, "loss": 0.0151, "step": 43840 }, { "grad_norm": 0.2693612277507782, "learning_rate": 4.6182137614381736e-05, "loss": 0.0152, "step": 43850 }, { "grad_norm": 0.29927271604537964, "learning_rate": 4.616152990925631e-05, "loss": 0.014, "step": 43860 }, { "grad_norm": 0.16308924555778503, "learning_rate": 4.614092286002041e-05, "loss": 0.0147, "step": 43870 }, { "grad_norm": 0.18456107378005981, "learning_rate": 4.612031647019528e-05, "loss": 0.0155, "step": 43880 }, { "grad_norm": 0.2384355068206787, "learning_rate": 4.609971074330193e-05, "loss": 0.0144, "step": 43890 }, { "grad_norm": 0.26696690917015076, "learning_rate": 4.607910568286136e-05, "loss": 0.0197, "step": 43900 }, { "grad_norm": 0.19022363424301147, "learning_rate": 4.605850129239438e-05, "loss": 0.0161, "step": 43910 }, { "grad_norm": 0.2649172246456146, "learning_rate": 4.603789757542174e-05, "loss": 0.0126, "step": 43920 }, { "grad_norm": 0.33473077416419983, "learning_rate": 4.601729453546404e-05, "loss": 0.0192, "step": 43930 }, { "grad_norm": 0.2231435924768448, "learning_rate": 4.599669217604177e-05, "loss": 0.0157, "step": 43940 }, { "grad_norm": 0.2213243842124939, "learning_rate": 4.597609050067532e-05, "loss": 0.0237, "step": 43950 }, { "grad_norm": 0.23761679232120514, "learning_rate": 4.595548951288495e-05, "loss": 0.0226, "step": 43960 }, { "grad_norm": 0.2503557503223419, "learning_rate": 4.593488921619081e-05, "loss": 0.0222, "step": 43970 }, { "grad_norm": 0.31663960218429565, "learning_rate": 4.591428961411289e-05, "loss": 0.0181, "step": 43980 }, { "grad_norm": 0.26619961857795715, "learning_rate": 4.589369071017117e-05, "loss": 0.0265, "step": 43990 }, { "grad_norm": 0.23620064556598663, "learning_rate": 4.587309250788538e-05, "loss": 0.0172, "step": 44000 }, { "grad_norm": 0.17564158141613007, "learning_rate": 4.585249501077522e-05, "loss": 0.0144, "step": 44010 }, { "grad_norm": 0.2143876999616623, "learning_rate": 4.583189822236021e-05, "loss": 0.0178, "step": 44020 }, { "grad_norm": 0.5813246965408325, "learning_rate": 4.5811302146159816e-05, "loss": 0.0143, "step": 44030 }, { "grad_norm": 0.239773690700531, "learning_rate": 4.579070678569332e-05, "loss": 0.0157, "step": 44040 }, { "grad_norm": 0.2777985632419586, "learning_rate": 4.5770112144479904e-05, "loss": 0.0173, "step": 44050 }, { "grad_norm": 0.25704121589660645, "learning_rate": 4.5749518226038645e-05, "loss": 0.0126, "step": 44060 }, { "grad_norm": 0.3088667690753937, "learning_rate": 4.572892503388845e-05, "loss": 0.0152, "step": 44070 }, { "grad_norm": 0.22467875480651855, "learning_rate": 4.570833257154817e-05, "loss": 0.0146, "step": 44080 }, { "grad_norm": 0.24577130377292633, "learning_rate": 4.568774084253646e-05, "loss": 0.0207, "step": 44090 }, { "grad_norm": 0.30640339851379395, "learning_rate": 4.566714985037191e-05, "loss": 0.0265, "step": 44100 }, { "grad_norm": 0.2681851089000702, "learning_rate": 4.564655959857295e-05, "loss": 0.0181, "step": 44110 }, { "grad_norm": 0.35307905077934265, "learning_rate": 4.5625970090657893e-05, "loss": 0.0184, "step": 44120 }, { "grad_norm": 0.17582039535045624, "learning_rate": 4.560538133014491e-05, "loss": 0.013, "step": 44130 }, { "grad_norm": 0.26225942373275757, "learning_rate": 4.5584793320552055e-05, "loss": 0.0134, "step": 44140 }, { "grad_norm": 0.21152985095977783, "learning_rate": 4.556420606539728e-05, "loss": 0.0173, "step": 44150 }, { "grad_norm": 0.21503587067127228, "learning_rate": 4.554361956819836e-05, "loss": 0.0128, "step": 44160 }, { "grad_norm": 0.1585896760225296, "learning_rate": 4.552303383247299e-05, "loss": 0.0121, "step": 44170 }, { "grad_norm": 0.16876302659511566, "learning_rate": 4.5502448861738664e-05, "loss": 0.0143, "step": 44180 }, { "grad_norm": 0.1950097680091858, "learning_rate": 4.548186465951284e-05, "loss": 0.0219, "step": 44190 }, { "grad_norm": 0.15786223113536835, "learning_rate": 4.546128122931277e-05, "loss": 0.0168, "step": 44200 }, { "grad_norm": 0.1648511439561844, "learning_rate": 4.544069857465561e-05, "loss": 0.0143, "step": 44210 }, { "grad_norm": 0.1733083873987198, "learning_rate": 4.542011669905837e-05, "loss": 0.0114, "step": 44220 }, { "grad_norm": 0.2052045464515686, "learning_rate": 4.53995356060379e-05, "loss": 0.013, "step": 44230 }, { "grad_norm": 0.2088044136762619, "learning_rate": 4.5378955299110994e-05, "loss": 0.011, "step": 44240 }, { "grad_norm": 0.265733927488327, "learning_rate": 4.5358375781794233e-05, "loss": 0.0191, "step": 44250 }, { "grad_norm": 0.17700651288032532, "learning_rate": 4.533779705760411e-05, "loss": 0.0213, "step": 44260 }, { "grad_norm": 0.18655720353126526, "learning_rate": 4.5317219130056934e-05, "loss": 0.0159, "step": 44270 }, { "grad_norm": 0.20589575171470642, "learning_rate": 4.5296642002668964e-05, "loss": 0.018, "step": 44280 }, { "grad_norm": 0.28617390990257263, "learning_rate": 4.5276065678956216e-05, "loss": 0.0176, "step": 44290 }, { "grad_norm": 0.2420060783624649, "learning_rate": 4.525549016243466e-05, "loss": 0.0146, "step": 44300 }, { "grad_norm": 0.19895458221435547, "learning_rate": 4.523491545662008e-05, "loss": 0.0145, "step": 44310 }, { "grad_norm": 0.20488421618938446, "learning_rate": 4.5214341565028104e-05, "loss": 0.0187, "step": 44320 }, { "grad_norm": 0.3309212923049927, "learning_rate": 4.519376849117428e-05, "loss": 0.0138, "step": 44330 }, { "grad_norm": 0.14240294694900513, "learning_rate": 4.517319623857395e-05, "loss": 0.0206, "step": 44340 }, { "grad_norm": 0.30453401803970337, "learning_rate": 4.5152624810742394e-05, "loss": 0.0159, "step": 44350 }, { "grad_norm": 0.14200110733509064, "learning_rate": 4.5132054211194664e-05, "loss": 0.0152, "step": 44360 }, { "grad_norm": 0.178582563996315, "learning_rate": 4.511148444344574e-05, "loss": 0.0145, "step": 44370 }, { "grad_norm": 0.23495405912399292, "learning_rate": 4.509091551101041e-05, "loss": 0.0194, "step": 44380 }, { "grad_norm": 0.2350444495677948, "learning_rate": 4.507034741740338e-05, "loss": 0.0157, "step": 44390 }, { "grad_norm": 0.2634269595146179, "learning_rate": 4.5049780166139145e-05, "loss": 0.0221, "step": 44400 }, { "grad_norm": 0.36773011088371277, "learning_rate": 4.5029213760732075e-05, "loss": 0.0266, "step": 44410 }, { "grad_norm": 0.3150296211242676, "learning_rate": 4.5008648204696434e-05, "loss": 0.0219, "step": 44420 }, { "grad_norm": 0.23171387612819672, "learning_rate": 4.4988083501546284e-05, "loss": 0.0163, "step": 44430 }, { "grad_norm": 0.23190727829933167, "learning_rate": 4.4967519654795606e-05, "loss": 0.0161, "step": 44440 }, { "grad_norm": 0.21671628952026367, "learning_rate": 4.494695666795816e-05, "loss": 0.0126, "step": 44450 }, { "grad_norm": 0.26530924439430237, "learning_rate": 4.4926394544547644e-05, "loss": 0.0177, "step": 44460 }, { "grad_norm": 0.19157977402210236, "learning_rate": 4.490583328807752e-05, "loss": 0.0185, "step": 44470 }, { "grad_norm": 0.18733341991901398, "learning_rate": 4.488527290206117e-05, "loss": 0.0176, "step": 44480 }, { "grad_norm": 0.1724219024181366, "learning_rate": 4.48647133900118e-05, "loss": 0.014, "step": 44490 }, { "grad_norm": 0.22398044168949127, "learning_rate": 4.484415475544244e-05, "loss": 0.0141, "step": 44500 }, { "grad_norm": 0.2886697053909302, "learning_rate": 4.4823597001866035e-05, "loss": 0.0156, "step": 44510 }, { "grad_norm": 0.20038634538650513, "learning_rate": 4.480304013279532e-05, "loss": 0.015, "step": 44520 }, { "grad_norm": 0.19230490922927856, "learning_rate": 4.478248415174292e-05, "loss": 0.0131, "step": 44530 }, { "grad_norm": 0.2501888871192932, "learning_rate": 4.476192906222126e-05, "loss": 0.0214, "step": 44540 }, { "grad_norm": 0.288203626871109, "learning_rate": 4.474137486774268e-05, "loss": 0.0307, "step": 44550 }, { "grad_norm": 0.22750402987003326, "learning_rate": 4.4720821571819296e-05, "loss": 0.0162, "step": 44560 }, { "grad_norm": 0.25286760926246643, "learning_rate": 4.470026917796314e-05, "loss": 0.0269, "step": 44570 }, { "grad_norm": 0.1690596342086792, "learning_rate": 4.4679717689686005e-05, "loss": 0.0351, "step": 44580 }, { "grad_norm": 0.3265182673931122, "learning_rate": 4.465916711049963e-05, "loss": 0.0155, "step": 44590 }, { "grad_norm": 0.2062050700187683, "learning_rate": 4.4638617443915524e-05, "loss": 0.0183, "step": 44600 }, { "grad_norm": 0.30986306071281433, "learning_rate": 4.4618068693445055e-05, "loss": 0.0198, "step": 44610 }, { "grad_norm": 0.3070514500141144, "learning_rate": 4.459752086259946e-05, "loss": 0.0168, "step": 44620 }, { "grad_norm": 0.4282083809375763, "learning_rate": 4.457697395488977e-05, "loss": 0.0227, "step": 44630 }, { "grad_norm": 0.2046995908021927, "learning_rate": 4.455642797382693e-05, "loss": 0.0222, "step": 44640 }, { "grad_norm": 0.2604435384273529, "learning_rate": 4.4535882922921643e-05, "loss": 0.0187, "step": 44650 }, { "grad_norm": 0.21823495626449585, "learning_rate": 4.451533880568455e-05, "loss": 0.0144, "step": 44660 }, { "grad_norm": 0.3305860161781311, "learning_rate": 4.449479562562603e-05, "loss": 0.0178, "step": 44670 }, { "grad_norm": 0.25359785556793213, "learning_rate": 4.447425338625639e-05, "loss": 0.0207, "step": 44680 }, { "grad_norm": 0.2507544755935669, "learning_rate": 4.4453712091085705e-05, "loss": 0.0162, "step": 44690 }, { "grad_norm": 0.2338474988937378, "learning_rate": 4.443317174362392e-05, "loss": 0.0248, "step": 44700 }, { "grad_norm": 0.2334875464439392, "learning_rate": 4.441263234738085e-05, "loss": 0.0147, "step": 44710 }, { "grad_norm": 0.142378568649292, "learning_rate": 4.4392093905866086e-05, "loss": 0.0106, "step": 44720 }, { "grad_norm": 0.42555829882621765, "learning_rate": 4.437155642258911e-05, "loss": 0.0149, "step": 44730 }, { "grad_norm": 0.14210735261440277, "learning_rate": 4.4351019901059177e-05, "loss": 0.0198, "step": 44740 }, { "grad_norm": 0.2709680199623108, "learning_rate": 4.433048434478547e-05, "loss": 0.0264, "step": 44750 }, { "grad_norm": 0.23403342068195343, "learning_rate": 4.4309949757276906e-05, "loss": 0.018, "step": 44760 }, { "grad_norm": 0.24726368486881256, "learning_rate": 4.428941614204233e-05, "loss": 0.0236, "step": 44770 }, { "grad_norm": 0.18225225806236267, "learning_rate": 4.426888350259034e-05, "loss": 0.0179, "step": 44780 }, { "grad_norm": 0.23318569362163544, "learning_rate": 4.424835184242939e-05, "loss": 0.0264, "step": 44790 }, { "grad_norm": 0.14528872072696686, "learning_rate": 4.422782116506784e-05, "loss": 0.0136, "step": 44800 }, { "grad_norm": 0.3685203790664673, "learning_rate": 4.4207291474013754e-05, "loss": 0.0203, "step": 44810 }, { "grad_norm": 0.18356095254421234, "learning_rate": 4.418676277277514e-05, "loss": 0.0309, "step": 44820 }, { "grad_norm": 0.2682843804359436, "learning_rate": 4.416623506485976e-05, "loss": 0.0247, "step": 44830 }, { "grad_norm": 0.23669597506523132, "learning_rate": 4.414570835377527e-05, "loss": 0.0155, "step": 44840 }, { "grad_norm": 0.16683052480220795, "learning_rate": 4.412518264302909e-05, "loss": 0.0138, "step": 44850 }, { "grad_norm": 0.2721329629421234, "learning_rate": 4.4104657936128535e-05, "loss": 0.0131, "step": 44860 }, { "grad_norm": 0.2060941904783249, "learning_rate": 4.408413423658071e-05, "loss": 0.0193, "step": 44870 }, { "grad_norm": 0.2146286517381668, "learning_rate": 4.406361154789253e-05, "loss": 0.0133, "step": 44880 }, { "grad_norm": 0.41215986013412476, "learning_rate": 4.4043089873570776e-05, "loss": 0.0182, "step": 44890 }, { "grad_norm": 0.3414762318134308, "learning_rate": 4.4022569217122025e-05, "loss": 0.0183, "step": 44900 }, { "grad_norm": 0.241881862282753, "learning_rate": 4.400204958205274e-05, "loss": 0.0173, "step": 44910 }, { "grad_norm": 0.30857545137405396, "learning_rate": 4.3981530971869125e-05, "loss": 0.0258, "step": 44920 }, { "grad_norm": 0.1755431741476059, "learning_rate": 4.396101339007727e-05, "loss": 0.0128, "step": 44930 }, { "grad_norm": 0.22632594406604767, "learning_rate": 4.394049684018304e-05, "loss": 0.0171, "step": 44940 }, { "grad_norm": 0.2137109339237213, "learning_rate": 4.3919981325692186e-05, "loss": 0.0161, "step": 44950 }, { "grad_norm": 0.1599102020263672, "learning_rate": 4.389946685011024e-05, "loss": 0.0115, "step": 44960 }, { "grad_norm": 0.27797240018844604, "learning_rate": 4.387895341694255e-05, "loss": 0.0225, "step": 44970 }, { "grad_norm": 0.2153625786304474, "learning_rate": 4.3858441029694324e-05, "loss": 0.0204, "step": 44980 }, { "grad_norm": 0.15241225063800812, "learning_rate": 4.3837929691870527e-05, "loss": 0.0132, "step": 44990 }, { "grad_norm": 0.20894324779510498, "learning_rate": 4.381741940697604e-05, "loss": 0.02, "step": 45000 }, { "grad_norm": 0.19590187072753906, "learning_rate": 4.379691017851547e-05, "loss": 0.0129, "step": 45010 }, { "grad_norm": 0.24573421478271484, "learning_rate": 4.3776402009993304e-05, "loss": 0.02, "step": 45020 }, { "grad_norm": 0.2105136215686798, "learning_rate": 4.3755894904913794e-05, "loss": 0.0135, "step": 45030 }, { "grad_norm": 0.16960059106349945, "learning_rate": 4.373538886678109e-05, "loss": 0.0123, "step": 45040 }, { "grad_norm": 0.18686825037002563, "learning_rate": 4.371488389909909e-05, "loss": 0.0145, "step": 45050 }, { "grad_norm": 0.3070353865623474, "learning_rate": 4.3694380005371515e-05, "loss": 0.0166, "step": 45060 }, { "grad_norm": 0.19116507470607758, "learning_rate": 4.367387718910196e-05, "loss": 0.0135, "step": 45070 }, { "grad_norm": 0.15575455129146576, "learning_rate": 4.3653375453793764e-05, "loss": 0.012, "step": 45080 }, { "grad_norm": 0.32408612966537476, "learning_rate": 4.3632874802950136e-05, "loss": 0.0133, "step": 45090 }, { "grad_norm": 0.23517826199531555, "learning_rate": 4.3612375240074034e-05, "loss": 0.0128, "step": 45100 }, { "grad_norm": 0.15823829174041748, "learning_rate": 4.3591876768668325e-05, "loss": 0.0157, "step": 45110 }, { "grad_norm": 0.3534281849861145, "learning_rate": 4.3571379392235605e-05, "loss": 0.0147, "step": 45120 }, { "grad_norm": 0.2775437831878662, "learning_rate": 4.3550883114278335e-05, "loss": 0.0178, "step": 45130 }, { "grad_norm": 0.22239355742931366, "learning_rate": 4.353038793829876e-05, "loss": 0.0151, "step": 45140 }, { "grad_norm": 0.2342100590467453, "learning_rate": 4.350989386779891e-05, "loss": 0.012, "step": 45150 }, { "grad_norm": 0.269224613904953, "learning_rate": 4.3489400906280724e-05, "loss": 0.025, "step": 45160 }, { "grad_norm": 0.27421489357948303, "learning_rate": 4.3468909057245845e-05, "loss": 0.0151, "step": 45170 }, { "grad_norm": 0.13260185718536377, "learning_rate": 4.3448418324195794e-05, "loss": 0.0178, "step": 45180 }, { "grad_norm": 0.2037147730588913, "learning_rate": 4.342792871063184e-05, "loss": 0.0164, "step": 45190 }, { "grad_norm": 0.18182623386383057, "learning_rate": 4.3407440220055145e-05, "loss": 0.0165, "step": 45200 }, { "grad_norm": 0.16536590456962585, "learning_rate": 4.33869528559666e-05, "loss": 0.0125, "step": 45210 }, { "grad_norm": 0.20062129199504852, "learning_rate": 4.336646662186696e-05, "loss": 0.0115, "step": 45220 }, { "grad_norm": 0.15827485918998718, "learning_rate": 4.334598152125672e-05, "loss": 0.017, "step": 45230 }, { "grad_norm": 0.15195322036743164, "learning_rate": 4.3325497557636276e-05, "loss": 0.0236, "step": 45240 }, { "grad_norm": 0.15662339329719543, "learning_rate": 4.330501473450574e-05, "loss": 0.0149, "step": 45250 }, { "grad_norm": 0.16893237829208374, "learning_rate": 4.328453305536507e-05, "loss": 0.0133, "step": 45260 }, { "grad_norm": 0.23005591332912445, "learning_rate": 4.326405252371404e-05, "loss": 0.0154, "step": 45270 }, { "grad_norm": 0.3471510410308838, "learning_rate": 4.324357314305221e-05, "loss": 0.0142, "step": 45280 }, { "grad_norm": 0.14459221065044403, "learning_rate": 4.3223094916878945e-05, "loss": 0.0138, "step": 45290 }, { "grad_norm": 0.1683640033006668, "learning_rate": 4.320261784869338e-05, "loss": 0.0239, "step": 45300 }, { "grad_norm": 0.21104830503463745, "learning_rate": 4.318214194199455e-05, "loss": 0.021, "step": 45310 }, { "grad_norm": 0.11971070617437363, "learning_rate": 4.316166720028118e-05, "loss": 0.0128, "step": 45320 }, { "grad_norm": 0.26213693618774414, "learning_rate": 4.3141193627051864e-05, "loss": 0.0163, "step": 45330 }, { "grad_norm": 0.256609708070755, "learning_rate": 4.312072122580496e-05, "loss": 0.0183, "step": 45340 }, { "grad_norm": 0.1510343700647354, "learning_rate": 4.3100250000038646e-05, "loss": 0.0142, "step": 45350 }, { "grad_norm": 0.310005247592926, "learning_rate": 4.307977995325091e-05, "loss": 0.0267, "step": 45360 }, { "grad_norm": 0.3805534839630127, "learning_rate": 4.30593110889395e-05, "loss": 0.0158, "step": 45370 }, { "grad_norm": 0.20549003779888153, "learning_rate": 4.3038843410602016e-05, "loss": 0.0143, "step": 45380 }, { "grad_norm": 0.18518686294555664, "learning_rate": 4.3018376921735774e-05, "loss": 0.018, "step": 45390 }, { "grad_norm": 0.12459999322891235, "learning_rate": 4.299791162583799e-05, "loss": 0.014, "step": 45400 }, { "grad_norm": 0.23958632349967957, "learning_rate": 4.29774475264056e-05, "loss": 0.0153, "step": 45410 }, { "grad_norm": 0.21631792187690735, "learning_rate": 4.2956984626935365e-05, "loss": 0.0127, "step": 45420 }, { "grad_norm": 0.1972798854112625, "learning_rate": 4.293652293092383e-05, "loss": 0.0124, "step": 45430 }, { "grad_norm": 0.15579083561897278, "learning_rate": 4.2916062441867324e-05, "loss": 0.0137, "step": 45440 }, { "grad_norm": 0.17235687375068665, "learning_rate": 4.289560316326201e-05, "loss": 0.0205, "step": 45450 }, { "grad_norm": 0.338419109582901, "learning_rate": 4.28751450986038e-05, "loss": 0.024, "step": 45460 }, { "grad_norm": 0.3128323554992676, "learning_rate": 4.2854688251388444e-05, "loss": 0.0198, "step": 45470 }, { "grad_norm": 0.18986503779888153, "learning_rate": 4.2834232625111425e-05, "loss": 0.0152, "step": 45480 }, { "grad_norm": 0.2145538330078125, "learning_rate": 4.2813778223268086e-05, "loss": 0.0151, "step": 45490 }, { "grad_norm": 0.2570869028568268, "learning_rate": 4.2793325049353477e-05, "loss": 0.0142, "step": 45500 }, { "grad_norm": 0.2872139811515808, "learning_rate": 4.2772873106862535e-05, "loss": 0.0145, "step": 45510 }, { "grad_norm": 0.18649151921272278, "learning_rate": 4.275242239928993e-05, "loss": 0.0139, "step": 45520 }, { "grad_norm": 0.14877083897590637, "learning_rate": 4.273197293013009e-05, "loss": 0.0092, "step": 45530 }, { "grad_norm": 0.1641170233488083, "learning_rate": 4.271152470287731e-05, "loss": 0.0121, "step": 45540 }, { "grad_norm": 0.18209177255630493, "learning_rate": 4.2691077721025594e-05, "loss": 0.0138, "step": 45550 }, { "grad_norm": 0.1318688839673996, "learning_rate": 4.267063198806883e-05, "loss": 0.0161, "step": 45560 }, { "grad_norm": 0.19156403839588165, "learning_rate": 4.2650187507500574e-05, "loss": 0.0121, "step": 45570 }, { "grad_norm": 0.22383682429790497, "learning_rate": 4.2629744282814275e-05, "loss": 0.0154, "step": 45580 }, { "grad_norm": 0.1499357968568802, "learning_rate": 4.2609302317503074e-05, "loss": 0.0087, "step": 45590 }, { "grad_norm": 0.1885296106338501, "learning_rate": 4.258886161505999e-05, "loss": 0.0123, "step": 45600 }, { "grad_norm": 0.3023361563682556, "learning_rate": 4.2568422178977775e-05, "loss": 0.0135, "step": 45610 }, { "grad_norm": 0.3104969561100006, "learning_rate": 4.254798401274894e-05, "loss": 0.0217, "step": 45620 }, { "grad_norm": 0.3677830398082733, "learning_rate": 4.252754711986583e-05, "loss": 0.0158, "step": 45630 }, { "grad_norm": 0.14662861824035645, "learning_rate": 4.250711150382052e-05, "loss": 0.0133, "step": 45640 }, { "grad_norm": 0.24622221291065216, "learning_rate": 4.248667716810495e-05, "loss": 0.0168, "step": 45650 }, { "grad_norm": 0.22855931520462036, "learning_rate": 4.246624411621074e-05, "loss": 0.0135, "step": 45660 }, { "grad_norm": 0.3543383479118347, "learning_rate": 4.244581235162938e-05, "loss": 0.0183, "step": 45670 }, { "grad_norm": 0.13070861995220184, "learning_rate": 4.2425381877852075e-05, "loss": 0.0233, "step": 45680 }, { "grad_norm": 0.25192269682884216, "learning_rate": 4.2404952698369856e-05, "loss": 0.0157, "step": 45690 }, { "grad_norm": 0.13458742201328278, "learning_rate": 4.238452481667349e-05, "loss": 0.0143, "step": 45700 }, { "grad_norm": 0.15028126537799835, "learning_rate": 4.2364098236253526e-05, "loss": 0.018, "step": 45710 }, { "grad_norm": 0.22850459814071655, "learning_rate": 4.2343672960600356e-05, "loss": 0.0152, "step": 45720 }, { "grad_norm": 0.20777331292629242, "learning_rate": 4.232324899320406e-05, "loss": 0.0105, "step": 45730 }, { "grad_norm": 0.19850659370422363, "learning_rate": 4.230282633755457e-05, "loss": 0.0137, "step": 45740 }, { "grad_norm": 0.2848781645298004, "learning_rate": 4.2282404997141515e-05, "loss": 0.0165, "step": 45750 }, { "grad_norm": 0.17643699049949646, "learning_rate": 4.2261984975454397e-05, "loss": 0.0186, "step": 45760 }, { "grad_norm": 0.2263023406267166, "learning_rate": 4.224156627598239e-05, "loss": 0.0123, "step": 45770 }, { "grad_norm": 0.16462954878807068, "learning_rate": 4.222114890221453e-05, "loss": 0.0166, "step": 45780 }, { "grad_norm": 0.32600125670433044, "learning_rate": 4.2200732857639546e-05, "loss": 0.0236, "step": 45790 }, { "grad_norm": 0.19793525338172913, "learning_rate": 4.2180318145746035e-05, "loss": 0.0162, "step": 45800 }, { "grad_norm": 0.3126102685928345, "learning_rate": 4.215990477002227e-05, "loss": 0.0212, "step": 45810 }, { "grad_norm": 0.2099713832139969, "learning_rate": 4.2139492733956356e-05, "loss": 0.0135, "step": 45820 }, { "grad_norm": 0.16364698112010956, "learning_rate": 4.211908204103615e-05, "loss": 0.0263, "step": 45830 }, { "grad_norm": 0.2559160590171814, "learning_rate": 4.2098672694749265e-05, "loss": 0.0319, "step": 45840 }, { "grad_norm": 0.19196556508541107, "learning_rate": 4.2078264698583133e-05, "loss": 0.0199, "step": 45850 }, { "grad_norm": 0.15781159698963165, "learning_rate": 4.205785805602488e-05, "loss": 0.0145, "step": 45860 }, { "grad_norm": 0.31249433755874634, "learning_rate": 4.203745277056149e-05, "loss": 0.0172, "step": 45870 }, { "grad_norm": 0.1479322463274002, "learning_rate": 4.201704884567964e-05, "loss": 0.0166, "step": 45880 }, { "grad_norm": 0.17316865921020508, "learning_rate": 4.1996646284865816e-05, "loss": 0.0236, "step": 45890 }, { "grad_norm": 0.10520104318857193, "learning_rate": 4.197624509160625e-05, "loss": 0.0142, "step": 45900 }, { "grad_norm": 0.1833336055278778, "learning_rate": 4.195584526938692e-05, "loss": 0.0102, "step": 45910 }, { "grad_norm": 0.14756551384925842, "learning_rate": 4.193544682169365e-05, "loss": 0.0122, "step": 45920 }, { "grad_norm": 0.18811434507369995, "learning_rate": 4.1915049752011946e-05, "loss": 0.0114, "step": 45930 }, { "grad_norm": 0.18019789457321167, "learning_rate": 4.189465406382712e-05, "loss": 0.0082, "step": 45940 }, { "grad_norm": 0.19292359054088593, "learning_rate": 4.187425976062422e-05, "loss": 0.0153, "step": 45950 }, { "grad_norm": 0.33707571029663086, "learning_rate": 4.18538668458881e-05, "loss": 0.0128, "step": 45960 }, { "grad_norm": 0.2427617907524109, "learning_rate": 4.183347532310333e-05, "loss": 0.0133, "step": 45970 }, { "grad_norm": 0.2361297607421875, "learning_rate": 4.181308519575429e-05, "loss": 0.0144, "step": 45980 }, { "grad_norm": 0.28920111060142517, "learning_rate": 4.179269646732507e-05, "loss": 0.0269, "step": 45990 }, { "grad_norm": 0.22549667954444885, "learning_rate": 4.177230914129954e-05, "loss": 0.0168, "step": 46000 }, { "grad_norm": 0.2854464650154114, "learning_rate": 4.175192322116136e-05, "loss": 0.0134, "step": 46010 }, { "grad_norm": 0.1992892324924469, "learning_rate": 4.173153871039391e-05, "loss": 0.0163, "step": 46020 }, { "grad_norm": 0.2733120620250702, "learning_rate": 4.171115561248036e-05, "loss": 0.0159, "step": 46030 }, { "grad_norm": 0.19487659633159637, "learning_rate": 4.16907739309036e-05, "loss": 0.0128, "step": 46040 }, { "grad_norm": 0.1879325658082962, "learning_rate": 4.167039366914633e-05, "loss": 0.0108, "step": 46050 }, { "grad_norm": 0.18824642896652222, "learning_rate": 4.165001483069096e-05, "loss": 0.0119, "step": 46060 }, { "grad_norm": 0.13784180581569672, "learning_rate": 4.16296374190197e-05, "loss": 0.0222, "step": 46070 }, { "grad_norm": 0.25297811627388, "learning_rate": 4.1609261437614464e-05, "loss": 0.0173, "step": 46080 }, { "grad_norm": 0.49060553312301636, "learning_rate": 4.158888688995696e-05, "loss": 0.0228, "step": 46090 }, { "grad_norm": 0.2626057267189026, "learning_rate": 4.1568513779528645e-05, "loss": 0.0137, "step": 46100 }, { "grad_norm": 0.18510127067565918, "learning_rate": 4.1548142109810704e-05, "loss": 0.0196, "step": 46110 }, { "grad_norm": 0.16763855516910553, "learning_rate": 4.152777188428414e-05, "loss": 0.0125, "step": 46120 }, { "grad_norm": 0.236003115773201, "learning_rate": 4.1507403106429646e-05, "loss": 0.0152, "step": 46130 }, { "grad_norm": 0.18051926791667938, "learning_rate": 4.148703577972768e-05, "loss": 0.0108, "step": 46140 }, { "grad_norm": 0.22693580389022827, "learning_rate": 4.146666990765846e-05, "loss": 0.0125, "step": 46150 }, { "grad_norm": 0.30882397294044495, "learning_rate": 4.1446305493701986e-05, "loss": 0.0141, "step": 46160 }, { "grad_norm": 0.19123317301273346, "learning_rate": 4.142594254133796e-05, "loss": 0.0168, "step": 46170 }, { "grad_norm": 0.20235803723335266, "learning_rate": 4.1405581054045835e-05, "loss": 0.0157, "step": 46180 }, { "grad_norm": 0.19909772276878357, "learning_rate": 4.1385221035304864e-05, "loss": 0.0126, "step": 46190 }, { "grad_norm": 0.21613827347755432, "learning_rate": 4.136486248859398e-05, "loss": 0.0186, "step": 46200 }, { "grad_norm": 0.13932938873767853, "learning_rate": 4.1344505417391955e-05, "loss": 0.0161, "step": 46210 }, { "grad_norm": 0.173112154006958, "learning_rate": 4.132414982517721e-05, "loss": 0.0112, "step": 46220 }, { "grad_norm": 0.1929190307855606, "learning_rate": 4.130379571542798e-05, "loss": 0.0144, "step": 46230 }, { "grad_norm": 0.2855220437049866, "learning_rate": 4.12834430916222e-05, "loss": 0.0187, "step": 46240 }, { "grad_norm": 0.23166784644126892, "learning_rate": 4.126309195723763e-05, "loss": 0.0142, "step": 46250 }, { "grad_norm": 0.16381986439228058, "learning_rate": 4.124274231575168e-05, "loss": 0.018, "step": 46260 }, { "grad_norm": 0.20022760331630707, "learning_rate": 4.122239417064154e-05, "loss": 0.012, "step": 46270 }, { "grad_norm": 0.14176389575004578, "learning_rate": 4.1202047525384184e-05, "loss": 0.0122, "step": 46280 }, { "grad_norm": 0.1871444433927536, "learning_rate": 4.118170238345627e-05, "loss": 0.0103, "step": 46290 }, { "grad_norm": 0.16363753378391266, "learning_rate": 4.1161358748334256e-05, "loss": 0.0108, "step": 46300 }, { "grad_norm": 0.2670673429965973, "learning_rate": 4.1141016623494266e-05, "loss": 0.0125, "step": 46310 }, { "grad_norm": 0.25516781210899353, "learning_rate": 4.112067601241227e-05, "loss": 0.0119, "step": 46320 }, { "grad_norm": 0.24831581115722656, "learning_rate": 4.110033691856387e-05, "loss": 0.0138, "step": 46330 }, { "grad_norm": 0.3319453299045563, "learning_rate": 4.107999934542451e-05, "loss": 0.0208, "step": 46340 }, { "grad_norm": 0.2431667596101761, "learning_rate": 4.105966329646928e-05, "loss": 0.0172, "step": 46350 }, { "grad_norm": 0.11043603718280792, "learning_rate": 4.103932877517308e-05, "loss": 0.0141, "step": 46360 }, { "grad_norm": 0.16250838339328766, "learning_rate": 4.101899578501052e-05, "loss": 0.0142, "step": 46370 }, { "grad_norm": 0.1730709969997406, "learning_rate": 4.099866432945595e-05, "loss": 0.0139, "step": 46380 }, { "grad_norm": 0.3215324282646179, "learning_rate": 4.097833441198346e-05, "loss": 0.0167, "step": 46390 }, { "grad_norm": 0.13073667883872986, "learning_rate": 4.095800603606685e-05, "loss": 0.0117, "step": 46400 }, { "grad_norm": 0.26497510075569153, "learning_rate": 4.093767920517975e-05, "loss": 0.0179, "step": 46410 }, { "grad_norm": 0.238525390625, "learning_rate": 4.091735392279539e-05, "loss": 0.0163, "step": 46420 }, { "grad_norm": 0.17090657353401184, "learning_rate": 4.089703019238685e-05, "loss": 0.0137, "step": 46430 }, { "grad_norm": 0.16124244034290314, "learning_rate": 4.0876708017426866e-05, "loss": 0.0162, "step": 46440 }, { "grad_norm": 0.20877569913864136, "learning_rate": 4.085638740138798e-05, "loss": 0.0146, "step": 46450 }, { "grad_norm": 0.21508705615997314, "learning_rate": 4.08360683477424e-05, "loss": 0.014, "step": 46460 }, { "grad_norm": 0.2626778483390808, "learning_rate": 4.0815750859962085e-05, "loss": 0.0225, "step": 46470 }, { "grad_norm": 0.189381942152977, "learning_rate": 4.079543494151879e-05, "loss": 0.0183, "step": 46480 }, { "grad_norm": 0.26884472370147705, "learning_rate": 4.07751205958839e-05, "loss": 0.0135, "step": 46490 }, { "grad_norm": 0.13770368695259094, "learning_rate": 4.0754807826528615e-05, "loss": 0.0137, "step": 46500 }, { "grad_norm": 0.19894897937774658, "learning_rate": 4.073449663692379e-05, "loss": 0.0201, "step": 46510 }, { "grad_norm": 0.15824134647846222, "learning_rate": 4.07141870305401e-05, "loss": 0.0126, "step": 46520 }, { "grad_norm": 0.2916513979434967, "learning_rate": 4.0693879010847866e-05, "loss": 0.0335, "step": 46530 }, { "grad_norm": 0.2003180831670761, "learning_rate": 4.06735725813172e-05, "loss": 0.015, "step": 46540 }, { "grad_norm": 0.21787406504154205, "learning_rate": 4.0653267745417903e-05, "loss": 0.0121, "step": 46550 }, { "grad_norm": 0.1838533580303192, "learning_rate": 4.063296450661949e-05, "loss": 0.0094, "step": 46560 }, { "grad_norm": 0.1467036008834839, "learning_rate": 4.061266286839128e-05, "loss": 0.0147, "step": 46570 }, { "grad_norm": 0.16416095197200775, "learning_rate": 4.0592362834202225e-05, "loss": 0.0169, "step": 46580 }, { "grad_norm": 0.28726106882095337, "learning_rate": 4.057206440752107e-05, "loss": 0.0101, "step": 46590 }, { "grad_norm": 0.21503013372421265, "learning_rate": 4.0551767591816245e-05, "loss": 0.012, "step": 46600 }, { "grad_norm": 0.1619863659143448, "learning_rate": 4.0531472390555935e-05, "loss": 0.0189, "step": 46610 }, { "grad_norm": 0.20540623366832733, "learning_rate": 4.051117880720802e-05, "loss": 0.0126, "step": 46620 }, { "grad_norm": 0.15212790668010712, "learning_rate": 4.049088684524015e-05, "loss": 0.0111, "step": 46630 }, { "grad_norm": 0.19442421197891235, "learning_rate": 4.0470596508119636e-05, "loss": 0.0111, "step": 46640 }, { "grad_norm": 0.18076828122138977, "learning_rate": 4.0450307799313524e-05, "loss": 0.0155, "step": 46650 }, { "grad_norm": 0.18299834430217743, "learning_rate": 4.0430020722288656e-05, "loss": 0.0173, "step": 46660 }, { "grad_norm": 0.16318883001804352, "learning_rate": 4.040973528051148e-05, "loss": 0.016, "step": 46670 }, { "grad_norm": 0.17311662435531616, "learning_rate": 4.038945147744827e-05, "loss": 0.0184, "step": 46680 }, { "grad_norm": 0.2083224654197693, "learning_rate": 4.0369169316564945e-05, "loss": 0.0133, "step": 46690 }, { "grad_norm": 0.2630438804626465, "learning_rate": 4.034888880132718e-05, "loss": 0.015, "step": 46700 }, { "grad_norm": 0.2728462517261505, "learning_rate": 4.032860993520035e-05, "loss": 0.017, "step": 46710 }, { "grad_norm": 0.16236823797225952, "learning_rate": 4.030833272164959e-05, "loss": 0.0127, "step": 46720 }, { "grad_norm": 0.2230105996131897, "learning_rate": 4.0288057164139705e-05, "loss": 0.0203, "step": 46730 }, { "grad_norm": 0.19029201567173004, "learning_rate": 4.0267783266135205e-05, "loss": 0.0102, "step": 46740 }, { "grad_norm": 0.1614799052476883, "learning_rate": 4.024751103110039e-05, "loss": 0.016, "step": 46750 }, { "grad_norm": 0.29196926951408386, "learning_rate": 4.0227240462499176e-05, "loss": 0.0187, "step": 46760 }, { "grad_norm": 0.18739987909793854, "learning_rate": 4.020697156379531e-05, "loss": 0.0178, "step": 46770 }, { "grad_norm": 0.24865227937698364, "learning_rate": 4.018670433845215e-05, "loss": 0.0186, "step": 46780 }, { "grad_norm": 0.24489851295948029, "learning_rate": 4.016643878993284e-05, "loss": 0.0111, "step": 46790 }, { "grad_norm": 0.24922321736812592, "learning_rate": 4.014617492170017e-05, "loss": 0.0179, "step": 46800 }, { "grad_norm": 0.21355271339416504, "learning_rate": 4.0125912737216726e-05, "loss": 0.0348, "step": 46810 }, { "grad_norm": 0.19119921326637268, "learning_rate": 4.0105652239944735e-05, "loss": 0.0179, "step": 46820 }, { "grad_norm": 0.2188633531332016, "learning_rate": 4.0085393433346144e-05, "loss": 0.017, "step": 46830 }, { "grad_norm": 0.1805163472890854, "learning_rate": 4.006513632088268e-05, "loss": 0.0106, "step": 46840 }, { "grad_norm": 0.15840719640254974, "learning_rate": 4.004488090601567e-05, "loss": 0.0114, "step": 46850 }, { "grad_norm": 0.3977821469306946, "learning_rate": 4.002462719220626e-05, "loss": 0.0105, "step": 46860 }, { "grad_norm": 0.134905144572258, "learning_rate": 4.000437518291522e-05, "loss": 0.0139, "step": 46870 }, { "grad_norm": 0.23397578299045563, "learning_rate": 3.9984124881603094e-05, "loss": 0.0115, "step": 46880 }, { "grad_norm": 0.16384513676166534, "learning_rate": 3.9963876291730086e-05, "loss": 0.0117, "step": 46890 }, { "grad_norm": 0.2141624242067337, "learning_rate": 3.994362941675614e-05, "loss": 0.0112, "step": 46900 }, { "grad_norm": 0.1552043855190277, "learning_rate": 3.992338426014088e-05, "loss": 0.0145, "step": 46910 }, { "grad_norm": 0.27126699686050415, "learning_rate": 3.9903140825343636e-05, "loss": 0.0151, "step": 46920 }, { "grad_norm": 0.2100166529417038, "learning_rate": 3.98828991158235e-05, "loss": 0.0124, "step": 46930 }, { "grad_norm": 0.19013740122318268, "learning_rate": 3.9862659135039185e-05, "loss": 0.0151, "step": 46940 }, { "grad_norm": 0.18907274305820465, "learning_rate": 3.984242088644918e-05, "loss": 0.0117, "step": 46950 }, { "grad_norm": 0.17748992145061493, "learning_rate": 3.9822184373511615e-05, "loss": 0.0103, "step": 46960 }, { "grad_norm": 0.22490522265434265, "learning_rate": 3.98019495996844e-05, "loss": 0.0169, "step": 46970 }, { "grad_norm": 0.15489622950553894, "learning_rate": 3.978171656842507e-05, "loss": 0.0152, "step": 46980 }, { "grad_norm": 0.2665409743785858, "learning_rate": 3.976148528319091e-05, "loss": 0.014, "step": 46990 }, { "grad_norm": 0.22330358624458313, "learning_rate": 3.974125574743888e-05, "loss": 0.0188, "step": 47000 }, { "grad_norm": 0.18796010315418243, "learning_rate": 3.9721027964625686e-05, "loss": 0.0168, "step": 47010 }, { "grad_norm": 0.1618557572364807, "learning_rate": 3.9700801938207676e-05, "loss": 0.0171, "step": 47020 }, { "grad_norm": 0.16058623790740967, "learning_rate": 3.9680577671640916e-05, "loss": 0.0125, "step": 47030 }, { "grad_norm": 0.17553192377090454, "learning_rate": 3.966035516838121e-05, "loss": 0.0191, "step": 47040 }, { "grad_norm": 0.18019388616085052, "learning_rate": 3.9640134431884014e-05, "loss": 0.011, "step": 47050 }, { "grad_norm": 0.15886907279491425, "learning_rate": 3.961991546560451e-05, "loss": 0.0096, "step": 47060 }, { "grad_norm": 0.1853455752134323, "learning_rate": 3.959969827299753e-05, "loss": 0.013, "step": 47070 }, { "grad_norm": 0.222225159406662, "learning_rate": 3.9579482857517684e-05, "loss": 0.0108, "step": 47080 }, { "grad_norm": 0.22792211174964905, "learning_rate": 3.955926922261921e-05, "loss": 0.0128, "step": 47090 }, { "grad_norm": 0.14859376847743988, "learning_rate": 3.9539057371756084e-05, "loss": 0.0147, "step": 47100 }, { "grad_norm": 0.17769479751586914, "learning_rate": 3.951884730838195e-05, "loss": 0.0108, "step": 47110 }, { "grad_norm": 0.14122335612773895, "learning_rate": 3.949863903595012e-05, "loss": 0.0101, "step": 47120 }, { "grad_norm": 0.2475224882364273, "learning_rate": 3.947843255791369e-05, "loss": 0.0154, "step": 47130 }, { "grad_norm": 0.1447104662656784, "learning_rate": 3.9458227877725364e-05, "loss": 0.0179, "step": 47140 }, { "grad_norm": 0.20264671742916107, "learning_rate": 3.943802499883758e-05, "loss": 0.0118, "step": 47150 }, { "grad_norm": 0.2684149742126465, "learning_rate": 3.9417823924702437e-05, "loss": 0.0164, "step": 47160 }, { "grad_norm": 0.2618919014930725, "learning_rate": 3.939762465877178e-05, "loss": 0.0171, "step": 47170 }, { "grad_norm": 0.16979266703128815, "learning_rate": 3.937742720449708e-05, "loss": 0.016, "step": 47180 }, { "grad_norm": 0.16523540019989014, "learning_rate": 3.9357231565329563e-05, "loss": 0.0122, "step": 47190 }, { "grad_norm": 0.17328286170959473, "learning_rate": 3.933703774472008e-05, "loss": 0.0108, "step": 47200 }, { "grad_norm": 0.15734845399856567, "learning_rate": 3.93168457461192e-05, "loss": 0.0118, "step": 47210 }, { "grad_norm": 0.1596650630235672, "learning_rate": 3.9296655572977216e-05, "loss": 0.0155, "step": 47220 }, { "grad_norm": 0.2689453661441803, "learning_rate": 3.927646722874404e-05, "loss": 0.0135, "step": 47230 }, { "grad_norm": 0.11392003297805786, "learning_rate": 3.925628071686934e-05, "loss": 0.0146, "step": 47240 }, { "grad_norm": 0.1499653160572052, "learning_rate": 3.9236096040802415e-05, "loss": 0.0151, "step": 47250 }, { "grad_norm": 0.36266031861305237, "learning_rate": 3.9215913203992294e-05, "loss": 0.0244, "step": 47260 }, { "grad_norm": 0.1552048623561859, "learning_rate": 3.9195732209887645e-05, "loss": 0.016, "step": 47270 }, { "grad_norm": 0.13494214415550232, "learning_rate": 3.9175553061936875e-05, "loss": 0.0128, "step": 47280 }, { "grad_norm": 0.18947510421276093, "learning_rate": 3.9155375763588045e-05, "loss": 0.0189, "step": 47290 }, { "grad_norm": 0.25869160890579224, "learning_rate": 3.913520031828889e-05, "loss": 0.0148, "step": 47300 }, { "grad_norm": 0.273598313331604, "learning_rate": 3.911502672948685e-05, "loss": 0.0147, "step": 47310 }, { "grad_norm": 0.28672537207603455, "learning_rate": 3.9094855000629014e-05, "loss": 0.0113, "step": 47320 }, { "grad_norm": 0.23215070366859436, "learning_rate": 3.907468513516223e-05, "loss": 0.0096, "step": 47330 }, { "grad_norm": 0.1435314267873764, "learning_rate": 3.905451713653294e-05, "loss": 0.0109, "step": 47340 }, { "grad_norm": 0.16191299259662628, "learning_rate": 3.903435100818731e-05, "loss": 0.0137, "step": 47350 }, { "grad_norm": 0.22959448397159576, "learning_rate": 3.901418675357117e-05, "loss": 0.0146, "step": 47360 }, { "grad_norm": 0.17985433340072632, "learning_rate": 3.8994024376130075e-05, "loss": 0.0224, "step": 47370 }, { "grad_norm": 0.26344063878059387, "learning_rate": 3.8973863879309194e-05, "loss": 0.0179, "step": 47380 }, { "grad_norm": 0.15649671852588654, "learning_rate": 3.8953705266553394e-05, "loss": 0.0108, "step": 47390 }, { "grad_norm": 0.22565515339374542, "learning_rate": 3.893354854130727e-05, "loss": 0.0171, "step": 47400 }, { "grad_norm": 0.18138617277145386, "learning_rate": 3.8913393707015006e-05, "loss": 0.0098, "step": 47410 }, { "grad_norm": 0.2571519911289215, "learning_rate": 3.889324076712056e-05, "loss": 0.0209, "step": 47420 }, { "grad_norm": 0.14953674376010895, "learning_rate": 3.8873089725067476e-05, "loss": 0.012, "step": 47430 }, { "grad_norm": 0.25636446475982666, "learning_rate": 3.885294058429905e-05, "loss": 0.023, "step": 47440 }, { "grad_norm": 0.18886107206344604, "learning_rate": 3.8832793348258206e-05, "loss": 0.0137, "step": 47450 }, { "grad_norm": 0.18170683085918427, "learning_rate": 3.881264802038756e-05, "loss": 0.02, "step": 47460 }, { "grad_norm": 0.27047136425971985, "learning_rate": 3.879250460412939e-05, "loss": 0.0201, "step": 47470 }, { "grad_norm": 0.18493300676345825, "learning_rate": 3.8772363102925644e-05, "loss": 0.0141, "step": 47480 }, { "grad_norm": 0.13365110754966736, "learning_rate": 3.875222352021798e-05, "loss": 0.0113, "step": 47490 }, { "grad_norm": 0.203910693526268, "learning_rate": 3.8732085859447686e-05, "loss": 0.0124, "step": 47500 }, { "grad_norm": 0.25413909554481506, "learning_rate": 3.871195012405575e-05, "loss": 0.0112, "step": 47510 }, { "grad_norm": 0.14948588609695435, "learning_rate": 3.869181631748278e-05, "loss": 0.013, "step": 47520 }, { "grad_norm": 0.3860786259174347, "learning_rate": 3.867168444316915e-05, "loss": 0.012, "step": 47530 }, { "grad_norm": 0.42522957921028137, "learning_rate": 3.865155450455481e-05, "loss": 0.0116, "step": 47540 }, { "grad_norm": 0.22090518474578857, "learning_rate": 3.8631426505079426e-05, "loss": 0.0168, "step": 47550 }, { "grad_norm": 0.251110702753067, "learning_rate": 3.8611300448182304e-05, "loss": 0.0237, "step": 47560 }, { "grad_norm": 0.17435798048973083, "learning_rate": 3.859117633730248e-05, "loss": 0.012, "step": 47570 }, { "grad_norm": 0.3001750707626343, "learning_rate": 3.857105417587858e-05, "loss": 0.0194, "step": 47580 }, { "grad_norm": 0.3039669394493103, "learning_rate": 3.855093396734894e-05, "loss": 0.0123, "step": 47590 }, { "grad_norm": 0.20053693652153015, "learning_rate": 3.8530815715151545e-05, "loss": 0.0275, "step": 47600 }, { "grad_norm": 0.24122776091098785, "learning_rate": 3.851069942272405e-05, "loss": 0.0131, "step": 47610 }, { "grad_norm": 0.28859710693359375, "learning_rate": 3.849058509350382e-05, "loss": 0.0149, "step": 47620 }, { "grad_norm": 0.25864946842193604, "learning_rate": 3.8470472730927783e-05, "loss": 0.0227, "step": 47630 }, { "grad_norm": 0.1398390829563141, "learning_rate": 3.845036233843264e-05, "loss": 0.0104, "step": 47640 }, { "grad_norm": 0.1544111669063568, "learning_rate": 3.843025391945469e-05, "loss": 0.0131, "step": 47650 }, { "grad_norm": 0.18901941180229187, "learning_rate": 3.841014747742992e-05, "loss": 0.0122, "step": 47660 }, { "grad_norm": 0.17613191902637482, "learning_rate": 3.839004301579397e-05, "loss": 0.0116, "step": 47670 }, { "grad_norm": 0.21958114206790924, "learning_rate": 3.8369940537982097e-05, "loss": 0.0196, "step": 47680 }, { "grad_norm": 0.15732432901859283, "learning_rate": 3.834984004742933e-05, "loss": 0.0108, "step": 47690 }, { "grad_norm": 0.19008374214172363, "learning_rate": 3.832974154757026e-05, "loss": 0.0098, "step": 47700 }, { "grad_norm": 0.20792067050933838, "learning_rate": 3.830964504183919e-05, "loss": 0.013, "step": 47710 }, { "grad_norm": 0.17368774116039276, "learning_rate": 3.828955053367003e-05, "loss": 0.0164, "step": 47720 }, { "grad_norm": 0.23266670107841492, "learning_rate": 3.826945802649642e-05, "loss": 0.0134, "step": 47730 }, { "grad_norm": 0.15543273091316223, "learning_rate": 3.824936752375159e-05, "loss": 0.0214, "step": 47740 }, { "grad_norm": 0.10273519158363342, "learning_rate": 3.822927902886848e-05, "loss": 0.0116, "step": 47750 }, { "grad_norm": 0.1885129064321518, "learning_rate": 3.8209192545279653e-05, "loss": 0.0127, "step": 47760 }, { "grad_norm": 0.21155479550361633, "learning_rate": 3.8189108076417326e-05, "loss": 0.0121, "step": 47770 }, { "grad_norm": 0.28650858998298645, "learning_rate": 3.816902562571342e-05, "loss": 0.0229, "step": 47780 }, { "grad_norm": 0.17773319780826569, "learning_rate": 3.814894519659944e-05, "loss": 0.0121, "step": 47790 }, { "grad_norm": 0.17473970353603363, "learning_rate": 3.812886679250661e-05, "loss": 0.0107, "step": 47800 }, { "grad_norm": 0.13934379816055298, "learning_rate": 3.810879041686575e-05, "loss": 0.0113, "step": 47810 }, { "grad_norm": 0.3315136134624481, "learning_rate": 3.808871607310741e-05, "loss": 0.0179, "step": 47820 }, { "grad_norm": 0.2646694779396057, "learning_rate": 3.806864376466169e-05, "loss": 0.0171, "step": 47830 }, { "grad_norm": 0.13482187688350677, "learning_rate": 3.804857349495845e-05, "loss": 0.0115, "step": 47840 }, { "grad_norm": 0.29620975255966187, "learning_rate": 3.802850526742713e-05, "loss": 0.0377, "step": 47850 }, { "grad_norm": 0.22485944628715515, "learning_rate": 3.800843908549683e-05, "loss": 0.0115, "step": 47860 }, { "grad_norm": 0.21187449991703033, "learning_rate": 3.7988374952596325e-05, "loss": 0.0103, "step": 47870 }, { "grad_norm": 0.17178916931152344, "learning_rate": 3.7968312872154e-05, "loss": 0.0119, "step": 47880 }, { "grad_norm": 0.21230612695217133, "learning_rate": 3.7948252847597965e-05, "loss": 0.0164, "step": 47890 }, { "grad_norm": 0.19754427671432495, "learning_rate": 3.7928194882355885e-05, "loss": 0.015, "step": 47900 }, { "grad_norm": 0.1815929412841797, "learning_rate": 3.790813897985515e-05, "loss": 0.0137, "step": 47910 }, { "grad_norm": 0.3319525420665741, "learning_rate": 3.7888085143522726e-05, "loss": 0.0163, "step": 47920 }, { "grad_norm": 0.19748319685459137, "learning_rate": 3.7868033376785314e-05, "loss": 0.0137, "step": 47930 }, { "grad_norm": 0.2048332542181015, "learning_rate": 3.784798368306919e-05, "loss": 0.0127, "step": 47940 }, { "grad_norm": 0.10947956144809723, "learning_rate": 3.782793606580029e-05, "loss": 0.0121, "step": 47950 }, { "grad_norm": 0.20048199594020844, "learning_rate": 3.7807890528404205e-05, "loss": 0.0163, "step": 47960 }, { "grad_norm": 0.24909281730651855, "learning_rate": 3.778784707430616e-05, "loss": 0.0104, "step": 47970 }, { "grad_norm": 0.1880127191543579, "learning_rate": 3.776780570693107e-05, "loss": 0.0118, "step": 47980 }, { "grad_norm": 0.1710738092660904, "learning_rate": 3.774776642970342e-05, "loss": 0.0149, "step": 47990 }, { "grad_norm": 0.22392593324184418, "learning_rate": 3.77277292460474e-05, "loss": 0.0103, "step": 48000 }, { "grad_norm": 0.2619422376155853, "learning_rate": 3.770769415938678e-05, "loss": 0.0112, "step": 48010 }, { "grad_norm": 0.20419842004776, "learning_rate": 3.768766117314506e-05, "loss": 0.0194, "step": 48020 }, { "grad_norm": 0.29533815383911133, "learning_rate": 3.76676302907453e-05, "loss": 0.014, "step": 48030 }, { "grad_norm": 0.21797145903110504, "learning_rate": 3.764760151561021e-05, "loss": 0.0135, "step": 48040 }, { "grad_norm": 0.25102120637893677, "learning_rate": 3.76275748511622e-05, "loss": 0.0115, "step": 48050 }, { "grad_norm": 0.16498254239559174, "learning_rate": 3.7607550300823255e-05, "loss": 0.0148, "step": 48060 }, { "grad_norm": 0.2805381417274475, "learning_rate": 3.7587527868015044e-05, "loss": 0.0294, "step": 48070 }, { "grad_norm": 0.44162312150001526, "learning_rate": 3.756750755615881e-05, "loss": 0.0172, "step": 48080 }, { "grad_norm": 0.17015016078948975, "learning_rate": 3.754748936867553e-05, "loss": 0.0154, "step": 48090 }, { "grad_norm": 0.27969837188720703, "learning_rate": 3.752747330898573e-05, "loss": 0.0138, "step": 48100 }, { "grad_norm": 0.23120547831058502, "learning_rate": 3.750745938050962e-05, "loss": 0.016, "step": 48110 }, { "grad_norm": 0.13110287487506866, "learning_rate": 3.7487447586667025e-05, "loss": 0.0139, "step": 48120 }, { "grad_norm": 0.24612070620059967, "learning_rate": 3.7467437930877426e-05, "loss": 0.0129, "step": 48130 }, { "grad_norm": 0.22561170160770416, "learning_rate": 3.744743041655992e-05, "loss": 0.012, "step": 48140 }, { "grad_norm": 0.22254151105880737, "learning_rate": 3.742742504713324e-05, "loss": 0.0142, "step": 48150 }, { "grad_norm": 0.15068772435188293, "learning_rate": 3.740742182601576e-05, "loss": 0.0133, "step": 48160 }, { "grad_norm": 0.15115444362163544, "learning_rate": 3.7387420756625464e-05, "loss": 0.0197, "step": 48170 }, { "grad_norm": 0.2143661230802536, "learning_rate": 3.736742184238002e-05, "loss": 0.0149, "step": 48180 }, { "grad_norm": 0.18859736621379852, "learning_rate": 3.7347425086696684e-05, "loss": 0.014, "step": 48190 }, { "grad_norm": 0.3033998906612396, "learning_rate": 3.732743049299235e-05, "loss": 0.0159, "step": 48200 }, { "grad_norm": 0.13865703344345093, "learning_rate": 3.730743806468354e-05, "loss": 0.0111, "step": 48210 }, { "grad_norm": 0.20337271690368652, "learning_rate": 3.7287447805186436e-05, "loss": 0.0097, "step": 48220 }, { "grad_norm": 0.22744204103946686, "learning_rate": 3.726745971791682e-05, "loss": 0.0199, "step": 48230 }, { "grad_norm": 0.3225991129875183, "learning_rate": 3.724747380629008e-05, "loss": 0.0284, "step": 48240 }, { "grad_norm": 0.3264772593975067, "learning_rate": 3.72274900737213e-05, "loss": 0.0187, "step": 48250 }, { "grad_norm": 0.2471785694360733, "learning_rate": 3.7207508523625123e-05, "loss": 0.0149, "step": 48260 }, { "grad_norm": 0.3026006519794464, "learning_rate": 3.718752915941588e-05, "loss": 0.0111, "step": 48270 }, { "grad_norm": 0.2378583550453186, "learning_rate": 3.7167551984507464e-05, "loss": 0.0134, "step": 48280 }, { "grad_norm": 0.2024964690208435, "learning_rate": 3.714757700231346e-05, "loss": 0.0134, "step": 48290 }, { "grad_norm": 0.2112136036157608, "learning_rate": 3.712760421624703e-05, "loss": 0.0118, "step": 48300 }, { "grad_norm": 0.207623690366745, "learning_rate": 3.710763362972099e-05, "loss": 0.014, "step": 48310 }, { "grad_norm": 0.26437878608703613, "learning_rate": 3.708766524614774e-05, "loss": 0.016, "step": 48320 }, { "grad_norm": 0.16126364469528198, "learning_rate": 3.7067699068939335e-05, "loss": 0.01, "step": 48330 }, { "grad_norm": 0.15360544621944427, "learning_rate": 3.704773510150748e-05, "loss": 0.0126, "step": 48340 }, { "grad_norm": 0.25417137145996094, "learning_rate": 3.702777334726344e-05, "loss": 0.0168, "step": 48350 }, { "grad_norm": 0.1685297191143036, "learning_rate": 3.7007813809618164e-05, "loss": 0.0113, "step": 48360 }, { "grad_norm": 0.1960141658782959, "learning_rate": 3.6987856491982145e-05, "loss": 0.0196, "step": 48370 }, { "grad_norm": 0.2060214877128601, "learning_rate": 3.696790139776558e-05, "loss": 0.0177, "step": 48380 }, { "grad_norm": 0.15110114216804504, "learning_rate": 3.6947948530378235e-05, "loss": 0.0178, "step": 48390 }, { "grad_norm": 0.17509996891021729, "learning_rate": 3.6927997893229516e-05, "loss": 0.0125, "step": 48400 }, { "grad_norm": 0.2616898715496063, "learning_rate": 3.690804948972845e-05, "loss": 0.012, "step": 48410 }, { "grad_norm": 0.16143909096717834, "learning_rate": 3.688810332328363e-05, "loss": 0.0124, "step": 48420 }, { "grad_norm": 0.28092461824417114, "learning_rate": 3.686815939730336e-05, "loss": 0.0169, "step": 48430 }, { "grad_norm": 0.2665722966194153, "learning_rate": 3.684821771519548e-05, "loss": 0.0121, "step": 48440 }, { "grad_norm": 0.1732402890920639, "learning_rate": 3.68282782803675e-05, "loss": 0.0114, "step": 48450 }, { "grad_norm": 0.17999492585659027, "learning_rate": 3.6808341096226504e-05, "loss": 0.0182, "step": 48460 }, { "grad_norm": 0.4246848225593567, "learning_rate": 3.678840616617924e-05, "loss": 0.0159, "step": 48470 }, { "grad_norm": 0.23762130737304688, "learning_rate": 3.6768473493632e-05, "loss": 0.0124, "step": 48480 }, { "grad_norm": 0.22864104807376862, "learning_rate": 3.6748543081990783e-05, "loss": 0.0123, "step": 48490 }, { "grad_norm": 0.19564130902290344, "learning_rate": 3.672861493466112e-05, "loss": 0.0136, "step": 48500 }, { "grad_norm": 0.19011786580085754, "learning_rate": 3.670868905504818e-05, "loss": 0.0142, "step": 48510 }, { "grad_norm": 0.21049928665161133, "learning_rate": 3.6688765446556784e-05, "loss": 0.0123, "step": 48520 }, { "grad_norm": 0.12732639908790588, "learning_rate": 3.6668844112591276e-05, "loss": 0.0155, "step": 48530 }, { "grad_norm": 0.16772043704986572, "learning_rate": 3.664892505655573e-05, "loss": 0.0133, "step": 48540 }, { "grad_norm": 0.1692620813846588, "learning_rate": 3.662900828185373e-05, "loss": 0.0167, "step": 48550 }, { "grad_norm": 0.17006175220012665, "learning_rate": 3.6609093791888516e-05, "loss": 0.0112, "step": 48560 }, { "grad_norm": 0.31868627667427063, "learning_rate": 3.658918159006292e-05, "loss": 0.0169, "step": 48570 }, { "grad_norm": 0.16977529227733612, "learning_rate": 3.656927167977942e-05, "loss": 0.0148, "step": 48580 }, { "grad_norm": 0.17783524096012115, "learning_rate": 3.654936406444006e-05, "loss": 0.0132, "step": 48590 }, { "grad_norm": 0.18332841992378235, "learning_rate": 3.65294587474465e-05, "loss": 0.01, "step": 48600 }, { "grad_norm": 0.2652488648891449, "learning_rate": 3.650955573220002e-05, "loss": 0.0476, "step": 48610 }, { "grad_norm": 0.30728858709335327, "learning_rate": 3.648965502210149e-05, "loss": 0.0141, "step": 48620 }, { "grad_norm": 0.21022938191890717, "learning_rate": 3.646975662055142e-05, "loss": 0.0146, "step": 48630 }, { "grad_norm": 0.13432176411151886, "learning_rate": 3.644986053094987e-05, "loss": 0.0122, "step": 48640 }, { "grad_norm": 0.2701756954193115, "learning_rate": 3.642996675669659e-05, "loss": 0.0163, "step": 48650 }, { "grad_norm": 0.22926023602485657, "learning_rate": 3.641007530119083e-05, "loss": 0.0115, "step": 48660 }, { "grad_norm": 0.23405639827251434, "learning_rate": 3.639018616783153e-05, "loss": 0.016, "step": 48670 }, { "grad_norm": 0.5303194522857666, "learning_rate": 3.637029936001719e-05, "loss": 0.012, "step": 48680 }, { "grad_norm": 0.17468218505382538, "learning_rate": 3.6350414881145886e-05, "loss": 0.0186, "step": 48690 }, { "grad_norm": 0.13362912833690643, "learning_rate": 3.6330532734615386e-05, "loss": 0.0176, "step": 48700 }, { "grad_norm": 0.12986642122268677, "learning_rate": 3.6310652923822975e-05, "loss": 0.0202, "step": 48710 }, { "grad_norm": 0.1792319416999817, "learning_rate": 3.629077545216558e-05, "loss": 0.0131, "step": 48720 }, { "grad_norm": 0.27223533391952515, "learning_rate": 3.627090032303969e-05, "loss": 0.0213, "step": 48730 }, { "grad_norm": 0.12549084424972534, "learning_rate": 3.625102753984146e-05, "loss": 0.0225, "step": 48740 }, { "grad_norm": 0.17807358503341675, "learning_rate": 3.623115710596659e-05, "loss": 0.0119, "step": 48750 }, { "grad_norm": 0.35709771513938904, "learning_rate": 3.6211289024810395e-05, "loss": 0.0187, "step": 48760 }, { "grad_norm": 0.1537739485502243, "learning_rate": 3.619142329976777e-05, "loss": 0.0113, "step": 48770 }, { "grad_norm": 0.25614669919013977, "learning_rate": 3.6171559934233247e-05, "loss": 0.0106, "step": 48780 }, { "grad_norm": 0.16526196897029877, "learning_rate": 3.615169893160093e-05, "loss": 0.0105, "step": 48790 }, { "grad_norm": 0.15239791572093964, "learning_rate": 3.61318402952645e-05, "loss": 0.0131, "step": 48800 }, { "grad_norm": 0.9369248151779175, "learning_rate": 3.6111984028617285e-05, "loss": 0.0169, "step": 48810 }, { "grad_norm": 0.20503266155719757, "learning_rate": 3.6092130135052134e-05, "loss": 0.0164, "step": 48820 }, { "grad_norm": 0.18579044938087463, "learning_rate": 3.6072278617961584e-05, "loss": 0.0193, "step": 48830 }, { "grad_norm": 0.27367401123046875, "learning_rate": 3.605242948073767e-05, "loss": 0.0211, "step": 48840 }, { "grad_norm": 0.294874906539917, "learning_rate": 3.603258272677212e-05, "loss": 0.0158, "step": 48850 }, { "grad_norm": 0.14555342495441437, "learning_rate": 3.601273835945616e-05, "loss": 0.012, "step": 48860 }, { "grad_norm": 0.37131020426750183, "learning_rate": 3.5992896382180664e-05, "loss": 0.0145, "step": 48870 }, { "grad_norm": 0.18793804943561554, "learning_rate": 3.597305679833609e-05, "loss": 0.0196, "step": 48880 }, { "grad_norm": 0.16444756090641022, "learning_rate": 3.595321961131245e-05, "loss": 0.0127, "step": 48890 }, { "grad_norm": 0.12153727561235428, "learning_rate": 3.593338482449942e-05, "loss": 0.0172, "step": 48900 }, { "grad_norm": 0.19436199963092804, "learning_rate": 3.591355244128618e-05, "loss": 0.0227, "step": 48910 }, { "grad_norm": 0.2599393427371979, "learning_rate": 3.589372246506158e-05, "loss": 0.0175, "step": 48920 }, { "grad_norm": 0.1852748692035675, "learning_rate": 3.5873894899213984e-05, "loss": 0.0115, "step": 48930 }, { "grad_norm": 0.20125332474708557, "learning_rate": 3.5854069747131416e-05, "loss": 0.0132, "step": 48940 }, { "grad_norm": 0.2903587818145752, "learning_rate": 3.583424701220143e-05, "loss": 0.0158, "step": 48950 }, { "grad_norm": 0.23631957173347473, "learning_rate": 3.581442669781121e-05, "loss": 0.0122, "step": 48960 }, { "grad_norm": 0.17457662522792816, "learning_rate": 3.579460880734749e-05, "loss": 0.0115, "step": 48970 }, { "grad_norm": 0.18352824449539185, "learning_rate": 3.577479334419657e-05, "loss": 0.0131, "step": 48980 }, { "grad_norm": 0.21169698238372803, "learning_rate": 3.575498031174444e-05, "loss": 0.0112, "step": 48990 }, { "grad_norm": 0.23909683525562286, "learning_rate": 3.573516971337657e-05, "loss": 0.0136, "step": 49000 }, { "grad_norm": 0.28753745555877686, "learning_rate": 3.5715361552478046e-05, "loss": 0.0165, "step": 49010 }, { "grad_norm": 0.1422409862279892, "learning_rate": 3.5695555832433536e-05, "loss": 0.0116, "step": 49020 }, { "grad_norm": 0.23601987957954407, "learning_rate": 3.5675752556627325e-05, "loss": 0.0144, "step": 49030 }, { "grad_norm": 0.2844562828540802, "learning_rate": 3.565595172844322e-05, "loss": 0.0162, "step": 49040 }, { "grad_norm": 0.162942573428154, "learning_rate": 3.5636153351264666e-05, "loss": 0.0121, "step": 49050 }, { "grad_norm": 0.17673373222351074, "learning_rate": 3.5616357428474655e-05, "loss": 0.0118, "step": 49060 }, { "grad_norm": 0.17157140374183655, "learning_rate": 3.559656396345575e-05, "loss": 0.0166, "step": 49070 }, { "grad_norm": 0.201247900724411, "learning_rate": 3.5576772959590146e-05, "loss": 0.014, "step": 49080 }, { "grad_norm": 0.15206734836101532, "learning_rate": 3.5556984420259545e-05, "loss": 0.0188, "step": 49090 }, { "grad_norm": 0.1802704781293869, "learning_rate": 3.5537198348845305e-05, "loss": 0.0144, "step": 49100 }, { "grad_norm": 0.1689099222421646, "learning_rate": 3.551741474872831e-05, "loss": 0.0141, "step": 49110 }, { "grad_norm": 0.24957527220249176, "learning_rate": 3.549763362328903e-05, "loss": 0.0173, "step": 49120 }, { "grad_norm": 0.2278563380241394, "learning_rate": 3.5477854975907515e-05, "loss": 0.0118, "step": 49130 }, { "grad_norm": 0.15098385512828827, "learning_rate": 3.5458078809963416e-05, "loss": 0.0152, "step": 49140 }, { "grad_norm": 0.2350846827030182, "learning_rate": 3.543830512883594e-05, "loss": 0.0124, "step": 49150 }, { "grad_norm": 0.29904037714004517, "learning_rate": 3.5418533935903824e-05, "loss": 0.0128, "step": 49160 }, { "grad_norm": 0.2540086507797241, "learning_rate": 3.539876523454547e-05, "loss": 0.0146, "step": 49170 }, { "grad_norm": 0.3767602741718292, "learning_rate": 3.537899902813878e-05, "loss": 0.0102, "step": 49180 }, { "grad_norm": 0.19280487298965454, "learning_rate": 3.5359235320061293e-05, "loss": 0.0147, "step": 49190 }, { "grad_norm": 0.16972695291042328, "learning_rate": 3.533947411369003e-05, "loss": 0.0132, "step": 49200 }, { "grad_norm": 0.1977277547121048, "learning_rate": 3.53197154124017e-05, "loss": 0.0133, "step": 49210 }, { "grad_norm": 0.35324984788894653, "learning_rate": 3.52999592195725e-05, "loss": 0.0176, "step": 49220 }, { "grad_norm": 0.22648833692073822, "learning_rate": 3.5280205538578224e-05, "loss": 0.012, "step": 49230 }, { "grad_norm": 0.8222048878669739, "learning_rate": 3.5260454372794236e-05, "loss": 0.0222, "step": 49240 }, { "grad_norm": 0.2273424118757248, "learning_rate": 3.524070572559545e-05, "loss": 0.0127, "step": 49250 }, { "grad_norm": 0.29166126251220703, "learning_rate": 3.5220959600356395e-05, "loss": 0.0171, "step": 49260 }, { "grad_norm": 0.27449271082878113, "learning_rate": 3.5201216000451145e-05, "loss": 0.0136, "step": 49270 }, { "grad_norm": 0.14864085614681244, "learning_rate": 3.5181474929253335e-05, "loss": 0.012, "step": 49280 }, { "grad_norm": 0.16226251423358917, "learning_rate": 3.516173639013615e-05, "loss": 0.0122, "step": 49290 }, { "grad_norm": 0.11327192932367325, "learning_rate": 3.5142000386472406e-05, "loss": 0.0123, "step": 49300 }, { "grad_norm": 0.18909130990505219, "learning_rate": 3.5122266921634427e-05, "loss": 0.0117, "step": 49310 }, { "grad_norm": 0.21126997470855713, "learning_rate": 3.510253599899413e-05, "loss": 0.0147, "step": 49320 }, { "grad_norm": 0.1527312695980072, "learning_rate": 3.5082807621922965e-05, "loss": 0.011, "step": 49330 }, { "grad_norm": 0.26415202021598816, "learning_rate": 3.506308179379201e-05, "loss": 0.0172, "step": 49340 }, { "grad_norm": 0.13587182760238647, "learning_rate": 3.5043358517971844e-05, "loss": 0.0113, "step": 49350 }, { "grad_norm": 0.15258029103279114, "learning_rate": 3.502363779783264e-05, "loss": 0.0113, "step": 49360 }, { "grad_norm": 0.18818381428718567, "learning_rate": 3.500391963674415e-05, "loss": 0.0092, "step": 49370 }, { "grad_norm": 0.2217591553926468, "learning_rate": 3.4984204038075615e-05, "loss": 0.0167, "step": 49380 }, { "grad_norm": 0.16193139553070068, "learning_rate": 3.496449100519595e-05, "loss": 0.0168, "step": 49390 }, { "grad_norm": 0.18851318955421448, "learning_rate": 3.494478054147354e-05, "loss": 0.0126, "step": 49400 }, { "grad_norm": 0.20632575452327728, "learning_rate": 3.4925072650276395e-05, "loss": 0.0103, "step": 49410 }, { "grad_norm": 0.14611691236495972, "learning_rate": 3.4905367334972016e-05, "loss": 0.0104, "step": 49420 }, { "grad_norm": 0.21062111854553223, "learning_rate": 3.488566459892752e-05, "loss": 0.0115, "step": 49430 }, { "grad_norm": 0.3674667179584503, "learning_rate": 3.4865964445509585e-05, "loss": 0.0123, "step": 49440 }, { "grad_norm": 0.19099946320056915, "learning_rate": 3.484626687808438e-05, "loss": 0.0123, "step": 49450 }, { "grad_norm": 0.23879189789295197, "learning_rate": 3.4826571900017735e-05, "loss": 0.0099, "step": 49460 }, { "grad_norm": 0.20467275381088257, "learning_rate": 3.480687951467495e-05, "loss": 0.0112, "step": 49470 }, { "grad_norm": 0.11356550455093384, "learning_rate": 3.4787189725420925e-05, "loss": 0.0114, "step": 49480 }, { "grad_norm": 0.2884548604488373, "learning_rate": 3.4767502535620086e-05, "loss": 0.0105, "step": 49490 }, { "grad_norm": 0.2160801887512207, "learning_rate": 3.474781794863648e-05, "loss": 0.0218, "step": 49500 }, { "grad_norm": 0.19824817776679993, "learning_rate": 3.472813596783363e-05, "loss": 0.0214, "step": 49510 }, { "grad_norm": 0.2641873359680176, "learning_rate": 3.470845659657466e-05, "loss": 0.0151, "step": 49520 }, { "grad_norm": 0.15466292202472687, "learning_rate": 3.468877983822223e-05, "loss": 0.0134, "step": 49530 }, { "grad_norm": 0.2095445692539215, "learning_rate": 3.466910569613855e-05, "loss": 0.0098, "step": 49540 }, { "grad_norm": 0.23195499181747437, "learning_rate": 3.464943417368542e-05, "loss": 0.0185, "step": 49550 }, { "grad_norm": 0.5046021342277527, "learning_rate": 3.462976527422415e-05, "loss": 0.0109, "step": 49560 }, { "grad_norm": 0.2347395420074463, "learning_rate": 3.461009900111562e-05, "loss": 0.0134, "step": 49570 }, { "grad_norm": 0.30630746483802795, "learning_rate": 3.459043535772023e-05, "loss": 0.0156, "step": 49580 }, { "grad_norm": 0.1868525892496109, "learning_rate": 3.4570774347398014e-05, "loss": 0.0159, "step": 49590 }, { "grad_norm": 0.13208557665348053, "learning_rate": 3.4551115973508454e-05, "loss": 0.0152, "step": 49600 }, { "grad_norm": 0.35705235600471497, "learning_rate": 3.453146023941066e-05, "loss": 0.0175, "step": 49610 }, { "grad_norm": 0.16052596271038055, "learning_rate": 3.451180714846325e-05, "loss": 0.0132, "step": 49620 }, { "grad_norm": 0.18864180147647858, "learning_rate": 3.449215670402438e-05, "loss": 0.015, "step": 49630 }, { "grad_norm": 0.239970862865448, "learning_rate": 3.447250890945181e-05, "loss": 0.0152, "step": 49640 }, { "grad_norm": 0.12033449113368988, "learning_rate": 3.4452863768102754e-05, "loss": 0.0129, "step": 49650 }, { "grad_norm": 0.19695526361465454, "learning_rate": 3.443322128333409e-05, "loss": 0.0162, "step": 49660 }, { "grad_norm": 0.3194115459918976, "learning_rate": 3.441358145850215e-05, "loss": 0.0183, "step": 49670 }, { "grad_norm": 0.15640951693058014, "learning_rate": 3.439394429696286e-05, "loss": 0.0139, "step": 49680 }, { "grad_norm": 0.20751510560512543, "learning_rate": 3.4374309802071644e-05, "loss": 0.0147, "step": 49690 }, { "grad_norm": 0.27116823196411133, "learning_rate": 3.435467797718353e-05, "loss": 0.0126, "step": 49700 }, { "grad_norm": 0.22883808612823486, "learning_rate": 3.433504882565306e-05, "loss": 0.0112, "step": 49710 }, { "grad_norm": 0.19390946626663208, "learning_rate": 3.43154223508343e-05, "loss": 0.019, "step": 49720 }, { "grad_norm": 0.10988721251487732, "learning_rate": 3.429579855608089e-05, "loss": 0.0131, "step": 49730 }, { "grad_norm": 0.12975575029850006, "learning_rate": 3.427617744474597e-05, "loss": 0.0116, "step": 49740 }, { "grad_norm": 0.24357382953166962, "learning_rate": 3.425655902018231e-05, "loss": 0.0151, "step": 49750 }, { "grad_norm": 0.2214909791946411, "learning_rate": 3.423694328574211e-05, "loss": 0.0132, "step": 49760 }, { "grad_norm": 0.2524205446243286, "learning_rate": 3.42173302447772e-05, "loss": 0.0134, "step": 49770 }, { "grad_norm": 0.28118571639060974, "learning_rate": 3.419771990063886e-05, "loss": 0.0093, "step": 49780 }, { "grad_norm": 0.16519257426261902, "learning_rate": 3.417811225667803e-05, "loss": 0.0111, "step": 49790 }, { "grad_norm": 0.2292332649230957, "learning_rate": 3.415850731624508e-05, "loss": 0.0178, "step": 49800 }, { "grad_norm": 0.19230976700782776, "learning_rate": 3.4138905082689945e-05, "loss": 0.0158, "step": 49810 }, { "grad_norm": 0.23470960557460785, "learning_rate": 3.4119305559362145e-05, "loss": 0.0179, "step": 49820 }, { "grad_norm": 0.17183712124824524, "learning_rate": 3.4099708749610684e-05, "loss": 0.011, "step": 49830 }, { "grad_norm": 0.1957906037569046, "learning_rate": 3.408011465678413e-05, "loss": 0.0177, "step": 49840 }, { "grad_norm": 0.20795755088329315, "learning_rate": 3.406052328423055e-05, "loss": 0.0174, "step": 49850 }, { "grad_norm": 0.1447196900844574, "learning_rate": 3.4040934635297615e-05, "loss": 0.0115, "step": 49860 }, { "grad_norm": 0.480523020029068, "learning_rate": 3.4021348713332466e-05, "loss": 0.0108, "step": 49870 }, { "grad_norm": 0.16279292106628418, "learning_rate": 3.4001765521681807e-05, "loss": 0.0117, "step": 49880 }, { "grad_norm": 0.20612174272537231, "learning_rate": 3.398218506369188e-05, "loss": 0.0208, "step": 49890 }, { "grad_norm": 0.3663932979106903, "learning_rate": 3.3962607342708404e-05, "loss": 0.0136, "step": 49900 }, { "grad_norm": 0.2874864935874939, "learning_rate": 3.394303236207673e-05, "loss": 0.012, "step": 49910 }, { "grad_norm": 0.16517458856105804, "learning_rate": 3.392346012514166e-05, "loss": 0.0134, "step": 49920 }, { "grad_norm": 0.1841207593679428, "learning_rate": 3.390389063524757e-05, "loss": 0.0111, "step": 49930 }, { "grad_norm": 0.12266390770673752, "learning_rate": 3.3884323895738324e-05, "loss": 0.0116, "step": 49940 }, { "grad_norm": 0.39142993092536926, "learning_rate": 3.386475990995738e-05, "loss": 0.018, "step": 49950 }, { "grad_norm": 0.1831558346748352, "learning_rate": 3.384519868124765e-05, "loss": 0.0138, "step": 49960 }, { "grad_norm": 0.220703586935997, "learning_rate": 3.3825640212951645e-05, "loss": 0.0117, "step": 49970 }, { "grad_norm": 0.18145038187503815, "learning_rate": 3.380608450841134e-05, "loss": 0.0128, "step": 49980 }, { "grad_norm": 0.2665539085865021, "learning_rate": 3.3786531570968305e-05, "loss": 0.0138, "step": 49990 }, { "grad_norm": 0.1773114800453186, "learning_rate": 3.3766981403963584e-05, "loss": 0.0143, "step": 50000 }, { "grad_norm": 0.14617682993412018, "learning_rate": 3.374743401073775e-05, "loss": 0.0117, "step": 50010 }, { "grad_norm": 0.22837823629379272, "learning_rate": 3.372788939463095e-05, "loss": 0.0093, "step": 50020 }, { "grad_norm": 0.1613357663154602, "learning_rate": 3.370834755898281e-05, "loss": 0.0107, "step": 50030 }, { "grad_norm": 0.2682904005050659, "learning_rate": 3.3688808507132493e-05, "loss": 0.015, "step": 50040 }, { "grad_norm": 0.14525531232357025, "learning_rate": 3.3669272242418685e-05, "loss": 0.0147, "step": 50050 }, { "grad_norm": 0.18837156891822815, "learning_rate": 3.364973876817961e-05, "loss": 0.0129, "step": 50060 }, { "grad_norm": 0.24848991632461548, "learning_rate": 3.363020808775299e-05, "loss": 0.0088, "step": 50070 }, { "grad_norm": 0.237962543964386, "learning_rate": 3.361068020447611e-05, "loss": 0.0126, "step": 50080 }, { "grad_norm": 0.15534032881259918, "learning_rate": 3.3591155121685724e-05, "loss": 0.0138, "step": 50090 }, { "grad_norm": 0.2233525663614273, "learning_rate": 3.357163284271814e-05, "loss": 0.0139, "step": 50100 }, { "grad_norm": 0.1299462765455246, "learning_rate": 3.355211337090919e-05, "loss": 0.0099, "step": 50110 }, { "grad_norm": 0.1788444072008133, "learning_rate": 3.353259670959421e-05, "loss": 0.0148, "step": 50120 }, { "grad_norm": 0.1631031036376953, "learning_rate": 3.351308286210808e-05, "loss": 0.0129, "step": 50130 }, { "grad_norm": 0.1854046732187271, "learning_rate": 3.3493571831785156e-05, "loss": 0.0121, "step": 50140 }, { "grad_norm": 0.2043033391237259, "learning_rate": 3.347406362195936e-05, "loss": 0.0102, "step": 50150 }, { "grad_norm": 0.1912824511528015, "learning_rate": 3.345455823596411e-05, "loss": 0.0134, "step": 50160 }, { "grad_norm": 0.30122125148773193, "learning_rate": 3.3435055677132346e-05, "loss": 0.0107, "step": 50170 }, { "grad_norm": 0.2229912281036377, "learning_rate": 3.3415555948796505e-05, "loss": 0.0116, "step": 50180 }, { "grad_norm": 0.3291977047920227, "learning_rate": 3.3396059054288556e-05, "loss": 0.0131, "step": 50190 }, { "grad_norm": 0.1392662674188614, "learning_rate": 3.3376564996940015e-05, "loss": 0.0239, "step": 50200 }, { "grad_norm": 0.16061407327651978, "learning_rate": 3.3357073780081836e-05, "loss": 0.0134, "step": 50210 }, { "grad_norm": 0.15035539865493774, "learning_rate": 3.333758540704459e-05, "loss": 0.0087, "step": 50220 }, { "grad_norm": 0.15151354670524597, "learning_rate": 3.331809988115827e-05, "loss": 0.01, "step": 50230 }, { "grad_norm": 0.23733621835708618, "learning_rate": 3.329861720575244e-05, "loss": 0.0094, "step": 50240 }, { "grad_norm": 0.15294094383716583, "learning_rate": 3.3279137384156126e-05, "loss": 0.009, "step": 50250 }, { "grad_norm": 0.24871595203876495, "learning_rate": 3.3259660419697934e-05, "loss": 0.0126, "step": 50260 }, { "grad_norm": 0.23497577011585236, "learning_rate": 3.3240186315705926e-05, "loss": 0.016, "step": 50270 }, { "grad_norm": 0.1364145129919052, "learning_rate": 3.322071507550769e-05, "loss": 0.0225, "step": 50280 }, { "grad_norm": 0.2182762622833252, "learning_rate": 3.320124670243033e-05, "loss": 0.011, "step": 50290 }, { "grad_norm": 0.12317114323377609, "learning_rate": 3.318178119980045e-05, "loss": 0.0112, "step": 50300 }, { "grad_norm": 0.19014571607112885, "learning_rate": 3.316231857094421e-05, "loss": 0.0112, "step": 50310 }, { "grad_norm": 0.21221911907196045, "learning_rate": 3.31428588191872e-05, "loss": 0.0125, "step": 50320 }, { "grad_norm": 0.28540849685668945, "learning_rate": 3.312340194785458e-05, "loss": 0.0151, "step": 50330 }, { "grad_norm": 0.2542571723461151, "learning_rate": 3.3103947960270975e-05, "loss": 0.0125, "step": 50340 }, { "grad_norm": 0.18720965087413788, "learning_rate": 3.308449685976058e-05, "loss": 0.0096, "step": 50350 }, { "grad_norm": 0.18068671226501465, "learning_rate": 3.3065048649647024e-05, "loss": 0.014, "step": 50360 }, { "grad_norm": 0.29286089539527893, "learning_rate": 3.304560333325348e-05, "loss": 0.0146, "step": 50370 }, { "grad_norm": 0.26652076840400696, "learning_rate": 3.3026160913902624e-05, "loss": 0.0146, "step": 50380 }, { "grad_norm": 0.3573322296142578, "learning_rate": 3.300672139491662e-05, "loss": 0.0192, "step": 50390 }, { "grad_norm": 0.22359538078308105, "learning_rate": 3.298728477961717e-05, "loss": 0.0154, "step": 50400 }, { "grad_norm": 0.1874690055847168, "learning_rate": 3.296785107132544e-05, "loss": 0.0171, "step": 50410 }, { "grad_norm": 0.1478564441204071, "learning_rate": 3.2948420273362156e-05, "loss": 0.0087, "step": 50420 }, { "grad_norm": 0.19898952543735504, "learning_rate": 3.292899238904747e-05, "loss": 0.012, "step": 50430 }, { "grad_norm": 0.18972983956336975, "learning_rate": 3.29095674217011e-05, "loss": 0.0124, "step": 50440 }, { "grad_norm": 0.18469104170799255, "learning_rate": 3.289014537464224e-05, "loss": 0.0156, "step": 50450 }, { "grad_norm": 0.15623518824577332, "learning_rate": 3.287072625118955e-05, "loss": 0.0162, "step": 50460 }, { "grad_norm": 0.2355046570301056, "learning_rate": 3.285131005466129e-05, "loss": 0.0107, "step": 50470 }, { "grad_norm": 0.2426232099533081, "learning_rate": 3.2831896788375105e-05, "loss": 0.0103, "step": 50480 }, { "grad_norm": 0.24218395352363586, "learning_rate": 3.281248645564822e-05, "loss": 0.0093, "step": 50490 }, { "grad_norm": 0.11592187732458115, "learning_rate": 3.2793079059797306e-05, "loss": 0.016, "step": 50500 }, { "grad_norm": 0.13889078795909882, "learning_rate": 3.277367460413859e-05, "loss": 0.0114, "step": 50510 }, { "grad_norm": 0.1430206596851349, "learning_rate": 3.275427309198773e-05, "loss": 0.0178, "step": 50520 }, { "grad_norm": 0.2267184853553772, "learning_rate": 3.273487452665993e-05, "loss": 0.0117, "step": 50530 }, { "grad_norm": 0.19205321371555328, "learning_rate": 3.271547891146986e-05, "loss": 0.0119, "step": 50540 }, { "grad_norm": 0.2664033770561218, "learning_rate": 3.269608624973173e-05, "loss": 0.0078, "step": 50550 }, { "grad_norm": 0.20401979982852936, "learning_rate": 3.26766965447592e-05, "loss": 0.0119, "step": 50560 }, { "grad_norm": 0.3240152895450592, "learning_rate": 3.2657309799865424e-05, "loss": 0.0161, "step": 50570 }, { "grad_norm": 0.15279662609100342, "learning_rate": 3.2637926018363084e-05, "loss": 0.0112, "step": 50580 }, { "grad_norm": 0.21797364950180054, "learning_rate": 3.261854520356432e-05, "loss": 0.0133, "step": 50590 }, { "grad_norm": 0.20458950102329254, "learning_rate": 3.2599167358780816e-05, "loss": 0.0113, "step": 50600 }, { "grad_norm": 0.2110082507133484, "learning_rate": 3.2579792487323676e-05, "loss": 0.0081, "step": 50610 }, { "grad_norm": 0.21815341711044312, "learning_rate": 3.256042059250358e-05, "loss": 0.0162, "step": 50620 }, { "grad_norm": 0.3221967816352844, "learning_rate": 3.254105167763062e-05, "loss": 0.0179, "step": 50630 }, { "grad_norm": 0.20332103967666626, "learning_rate": 3.252168574601443e-05, "loss": 0.0146, "step": 50640 }, { "grad_norm": 0.19301490485668182, "learning_rate": 3.2502322800964115e-05, "loss": 0.0103, "step": 50650 }, { "grad_norm": 0.28713035583496094, "learning_rate": 3.248296284578825e-05, "loss": 0.0115, "step": 50660 }, { "grad_norm": 0.17331813275814056, "learning_rate": 3.246360588379497e-05, "loss": 0.0155, "step": 50670 }, { "grad_norm": 0.1322159469127655, "learning_rate": 3.24442519182918e-05, "loss": 0.0233, "step": 50680 }, { "grad_norm": 0.17163696885108948, "learning_rate": 3.2424900952585845e-05, "loss": 0.0097, "step": 50690 }, { "grad_norm": 0.17055939137935638, "learning_rate": 3.240555298998362e-05, "loss": 0.0132, "step": 50700 }, { "grad_norm": 0.1945556253194809, "learning_rate": 3.2386208033791194e-05, "loss": 0.0112, "step": 50710 }, { "grad_norm": 0.20441460609436035, "learning_rate": 3.2366866087314064e-05, "loss": 0.0106, "step": 50720 }, { "grad_norm": 0.1619947850704193, "learning_rate": 3.234752715385727e-05, "loss": 0.0131, "step": 50730 }, { "grad_norm": 0.17438682913780212, "learning_rate": 3.232819123672529e-05, "loss": 0.0191, "step": 50740 }, { "grad_norm": 0.34523576498031616, "learning_rate": 3.2308858339222076e-05, "loss": 0.0195, "step": 50750 }, { "grad_norm": 0.26598241925239563, "learning_rate": 3.2289528464651144e-05, "loss": 0.0156, "step": 50760 }, { "grad_norm": 0.22342689335346222, "learning_rate": 3.22702016163154e-05, "loss": 0.0147, "step": 50770 }, { "grad_norm": 0.21633948385715485, "learning_rate": 3.2250877797517295e-05, "loss": 0.0164, "step": 50780 }, { "grad_norm": 0.13003042340278625, "learning_rate": 3.223155701155872e-05, "loss": 0.0173, "step": 50790 }, { "grad_norm": 0.19306504726409912, "learning_rate": 3.2212239261741105e-05, "loss": 0.0229, "step": 50800 }, { "grad_norm": 0.13098612427711487, "learning_rate": 3.219292455136528e-05, "loss": 0.0118, "step": 50810 }, { "grad_norm": 0.1469954550266266, "learning_rate": 3.217361288373165e-05, "loss": 0.0099, "step": 50820 }, { "grad_norm": 0.10574837028980255, "learning_rate": 3.215430426214002e-05, "loss": 0.0111, "step": 50830 }, { "grad_norm": 0.25918087363243103, "learning_rate": 3.2134998689889705e-05, "loss": 0.0116, "step": 50840 }, { "grad_norm": 0.19170384109020233, "learning_rate": 3.211569617027952e-05, "loss": 0.0137, "step": 50850 }, { "grad_norm": 0.09003143757581711, "learning_rate": 3.209639670660769e-05, "loss": 0.0134, "step": 50860 }, { "grad_norm": 0.4081428050994873, "learning_rate": 3.207710030217202e-05, "loss": 0.0158, "step": 50870 }, { "grad_norm": 0.2908669710159302, "learning_rate": 3.205780696026972e-05, "loss": 0.0105, "step": 50880 }, { "grad_norm": 0.13164883852005005, "learning_rate": 3.203851668419749e-05, "loss": 0.0116, "step": 50890 }, { "grad_norm": 0.15372024476528168, "learning_rate": 3.201922947725149e-05, "loss": 0.0137, "step": 50900 }, { "grad_norm": 0.18900513648986816, "learning_rate": 3.199994534272742e-05, "loss": 0.0119, "step": 50910 }, { "grad_norm": 0.18809078633785248, "learning_rate": 3.1980664283920394e-05, "loss": 0.0162, "step": 50920 }, { "grad_norm": 0.21162013709545135, "learning_rate": 3.196138630412499e-05, "loss": 0.0099, "step": 50930 }, { "grad_norm": 0.2580263316631317, "learning_rate": 3.194211140663532e-05, "loss": 0.0108, "step": 50940 }, { "grad_norm": 0.1921321153640747, "learning_rate": 3.1922839594744914e-05, "loss": 0.0184, "step": 50950 }, { "grad_norm": 0.2023678421974182, "learning_rate": 3.190357087174683e-05, "loss": 0.014, "step": 50960 }, { "grad_norm": 0.24044303596019745, "learning_rate": 3.188430524093353e-05, "loss": 0.0087, "step": 50970 }, { "grad_norm": 0.16887858510017395, "learning_rate": 3.1865042705597014e-05, "loss": 0.015, "step": 50980 }, { "grad_norm": 0.2775132656097412, "learning_rate": 3.184578326902868e-05, "loss": 0.0154, "step": 50990 }, { "grad_norm": 0.16926582157611847, "learning_rate": 3.182652693451949e-05, "loss": 0.014, "step": 51000 }, { "grad_norm": 0.1675052046775818, "learning_rate": 3.180727370535981e-05, "loss": 0.0122, "step": 51010 }, { "grad_norm": 0.3910232484340668, "learning_rate": 3.178802358483945e-05, "loss": 0.0124, "step": 51020 }, { "grad_norm": 0.16122998297214508, "learning_rate": 3.17687765762478e-05, "loss": 0.0125, "step": 51030 }, { "grad_norm": 0.11349353939294815, "learning_rate": 3.174953268287357e-05, "loss": 0.0162, "step": 51040 }, { "grad_norm": 0.26045823097229004, "learning_rate": 3.173029190800508e-05, "loss": 0.0138, "step": 51050 }, { "grad_norm": 0.20976372063159943, "learning_rate": 3.171105425493e-05, "loss": 0.0177, "step": 51060 }, { "grad_norm": 0.15740539133548737, "learning_rate": 3.169181972693557e-05, "loss": 0.0086, "step": 51070 }, { "grad_norm": 0.12577739357948303, "learning_rate": 3.1672588327308404e-05, "loss": 0.0106, "step": 51080 }, { "grad_norm": 0.2998232841491699, "learning_rate": 3.1653360059334646e-05, "loss": 0.0157, "step": 51090 }, { "grad_norm": 0.36768341064453125, "learning_rate": 3.163413492629985e-05, "loss": 0.028, "step": 51100 }, { "grad_norm": 0.22639305889606476, "learning_rate": 3.161491293148912e-05, "loss": 0.0126, "step": 51110 }, { "grad_norm": 0.18597736954689026, "learning_rate": 3.1595694078186925e-05, "loss": 0.0124, "step": 51120 }, { "grad_norm": 0.17866051197052002, "learning_rate": 3.157647836967724e-05, "loss": 0.0104, "step": 51130 }, { "grad_norm": 0.26949790120124817, "learning_rate": 3.155726580924354e-05, "loss": 0.0152, "step": 51140 }, { "grad_norm": 0.19015084207057953, "learning_rate": 3.1538056400168676e-05, "loss": 0.0222, "step": 51150 }, { "grad_norm": 0.18983954191207886, "learning_rate": 3.151885014573506e-05, "loss": 0.0202, "step": 51160 }, { "grad_norm": 0.1766977310180664, "learning_rate": 3.149964704922448e-05, "loss": 0.0213, "step": 51170 }, { "grad_norm": 0.15508989989757538, "learning_rate": 3.148044711391825e-05, "loss": 0.0111, "step": 51180 }, { "grad_norm": 0.35757768154144287, "learning_rate": 3.146125034309707e-05, "loss": 0.0109, "step": 51190 }, { "grad_norm": 0.18882592022418976, "learning_rate": 3.1442056740041195e-05, "loss": 0.0162, "step": 51200 }, { "grad_norm": 0.263801634311676, "learning_rate": 3.142286630803026e-05, "loss": 0.0246, "step": 51210 }, { "grad_norm": 0.20680221915245056, "learning_rate": 3.140367905034337e-05, "loss": 0.0122, "step": 51220 }, { "grad_norm": 0.20191676914691925, "learning_rate": 3.138449497025914e-05, "loss": 0.0134, "step": 51230 }, { "grad_norm": 0.1730862855911255, "learning_rate": 3.136531407105557e-05, "loss": 0.0126, "step": 51240 }, { "grad_norm": 0.2180105447769165, "learning_rate": 3.1346136356010184e-05, "loss": 0.0169, "step": 51250 }, { "grad_norm": 0.1677677035331726, "learning_rate": 3.132696182839988e-05, "loss": 0.0094, "step": 51260 }, { "grad_norm": 0.1576022207736969, "learning_rate": 3.130779049150111e-05, "loss": 0.0152, "step": 51270 }, { "grad_norm": 0.26806744933128357, "learning_rate": 3.128862234858971e-05, "loss": 0.0106, "step": 51280 }, { "grad_norm": 0.13882987201213837, "learning_rate": 3.1269457402941005e-05, "loss": 0.0104, "step": 51290 }, { "grad_norm": 0.12739738821983337, "learning_rate": 3.125029565782974e-05, "loss": 0.013, "step": 51300 }, { "grad_norm": 0.24853771924972534, "learning_rate": 3.123113711653013e-05, "loss": 0.0123, "step": 51310 }, { "grad_norm": 0.11581743508577347, "learning_rate": 3.121198178231587e-05, "loss": 0.0151, "step": 51320 }, { "grad_norm": 0.23011641204357147, "learning_rate": 3.119282965846006e-05, "loss": 0.0137, "step": 51330 }, { "grad_norm": 0.23430617153644562, "learning_rate": 3.1173680748235303e-05, "loss": 0.0153, "step": 51340 }, { "grad_norm": 0.15009856224060059, "learning_rate": 3.115453505491358e-05, "loss": 0.0152, "step": 51350 }, { "grad_norm": 0.1858477145433426, "learning_rate": 3.1135392581766404e-05, "loss": 0.012, "step": 51360 }, { "grad_norm": 0.16637717187404633, "learning_rate": 3.111625333206467e-05, "loss": 0.0117, "step": 51370 }, { "grad_norm": 0.13679437339305878, "learning_rate": 3.109711730907881e-05, "loss": 0.0088, "step": 51380 }, { "grad_norm": 0.21020056307315826, "learning_rate": 3.107798451607859e-05, "loss": 0.0108, "step": 51390 }, { "grad_norm": 0.12889665365219116, "learning_rate": 3.105885495633328e-05, "loss": 0.0119, "step": 51400 }, { "grad_norm": 0.11436288803815842, "learning_rate": 3.103972863311163e-05, "loss": 0.008, "step": 51410 }, { "grad_norm": 0.16319310665130615, "learning_rate": 3.102060554968178e-05, "loss": 0.0115, "step": 51420 }, { "grad_norm": 0.1756318360567093, "learning_rate": 3.100148570931137e-05, "loss": 0.0115, "step": 51430 }, { "grad_norm": 0.26464730501174927, "learning_rate": 3.098236911526744e-05, "loss": 0.017, "step": 51440 }, { "grad_norm": 0.13913968205451965, "learning_rate": 3.0963255770816495e-05, "loss": 0.0101, "step": 51450 }, { "grad_norm": 0.15767982602119446, "learning_rate": 3.094414567922447e-05, "loss": 0.0144, "step": 51460 }, { "grad_norm": 0.1396196037530899, "learning_rate": 3.092503884375678e-05, "loss": 0.0145, "step": 51470 }, { "grad_norm": 0.31226739287376404, "learning_rate": 3.0905935267678255e-05, "loss": 0.0141, "step": 51480 }, { "grad_norm": 0.2892463505268097, "learning_rate": 3.0886834954253166e-05, "loss": 0.0121, "step": 51490 }, { "grad_norm": 0.24720381200313568, "learning_rate": 3.086773790674524e-05, "loss": 0.0164, "step": 51500 }, { "grad_norm": 0.2511190176010132, "learning_rate": 3.0848644128417617e-05, "loss": 0.0135, "step": 51510 }, { "grad_norm": 0.4230724573135376, "learning_rate": 3.082955362253294e-05, "loss": 0.0187, "step": 51520 }, { "grad_norm": 0.22033686935901642, "learning_rate": 3.081046639235323e-05, "loss": 0.0141, "step": 51530 }, { "grad_norm": 0.2858951985836029, "learning_rate": 3.0791382441139986e-05, "loss": 0.0106, "step": 51540 }, { "grad_norm": 0.14400410652160645, "learning_rate": 3.077230177215411e-05, "loss": 0.0147, "step": 51550 }, { "grad_norm": 0.19533412158489227, "learning_rate": 3.0753224388656e-05, "loss": 0.0093, "step": 51560 }, { "grad_norm": 0.20355573296546936, "learning_rate": 3.073415029390544e-05, "loss": 0.0164, "step": 51570 }, { "grad_norm": 0.1652533859014511, "learning_rate": 3.071507949116166e-05, "loss": 0.0095, "step": 51580 }, { "grad_norm": 0.14808563888072968, "learning_rate": 3.069601198368337e-05, "loss": 0.0084, "step": 51590 }, { "grad_norm": 0.19209085404872894, "learning_rate": 3.067694777472864e-05, "loss": 0.0118, "step": 51600 }, { "grad_norm": 0.15860602259635925, "learning_rate": 3.065788686755508e-05, "loss": 0.0139, "step": 51610 }, { "grad_norm": 0.19616729021072388, "learning_rate": 3.063882926541961e-05, "loss": 0.0134, "step": 51620 }, { "grad_norm": 0.2200794219970703, "learning_rate": 3.061977497157872e-05, "loss": 0.01, "step": 51630 }, { "grad_norm": 0.2191067785024643, "learning_rate": 3.0600723989288224e-05, "loss": 0.0121, "step": 51640 }, { "grad_norm": 0.22228334844112396, "learning_rate": 3.058167632180343e-05, "loss": 0.0135, "step": 51650 }, { "grad_norm": 0.23553280532360077, "learning_rate": 3.0562631972379074e-05, "loss": 0.0144, "step": 51660 }, { "grad_norm": 0.18197104334831238, "learning_rate": 3.0543590944269276e-05, "loss": 0.0105, "step": 51670 }, { "grad_norm": 0.3655495047569275, "learning_rate": 3.052455324072766e-05, "loss": 0.0186, "step": 51680 }, { "grad_norm": 0.21620585024356842, "learning_rate": 3.0505518865007244e-05, "loss": 0.0151, "step": 51690 }, { "grad_norm": 0.15772514045238495, "learning_rate": 3.048648782036048e-05, "loss": 0.0118, "step": 51700 }, { "grad_norm": 0.17011629045009613, "learning_rate": 3.0467460110039236e-05, "loss": 0.0129, "step": 51710 }, { "grad_norm": 0.2895577847957611, "learning_rate": 3.0448435737294866e-05, "loss": 0.0131, "step": 51720 }, { "grad_norm": 0.23201623558998108, "learning_rate": 3.042941470537808e-05, "loss": 0.0078, "step": 51730 }, { "grad_norm": 0.12793326377868652, "learning_rate": 3.0410397017539072e-05, "loss": 0.0085, "step": 51740 }, { "grad_norm": 0.25106534361839294, "learning_rate": 3.039138267702742e-05, "loss": 0.0088, "step": 51750 }, { "grad_norm": 0.0947934091091156, "learning_rate": 3.0372371687092193e-05, "loss": 0.0087, "step": 51760 }, { "grad_norm": 0.14363525807857513, "learning_rate": 3.0353364050981823e-05, "loss": 0.0109, "step": 51770 }, { "grad_norm": 0.15522925555706024, "learning_rate": 3.033435977194418e-05, "loss": 0.013, "step": 51780 }, { "grad_norm": 0.15727701783180237, "learning_rate": 3.0315358853226618e-05, "loss": 0.0117, "step": 51790 }, { "grad_norm": 0.20155854523181915, "learning_rate": 3.0296361298075844e-05, "loss": 0.0108, "step": 51800 }, { "grad_norm": 0.20439040660858154, "learning_rate": 3.027736710973803e-05, "loss": 0.0153, "step": 51810 }, { "grad_norm": 0.13079427182674408, "learning_rate": 3.025837629145875e-05, "loss": 0.0087, "step": 51820 }, { "grad_norm": 0.17281606793403625, "learning_rate": 3.0239388846483048e-05, "loss": 0.0117, "step": 51830 }, { "grad_norm": 0.17808179557323456, "learning_rate": 3.022040477805532e-05, "loss": 0.0138, "step": 51840 }, { "grad_norm": 0.26169222593307495, "learning_rate": 3.020142408941946e-05, "loss": 0.0116, "step": 51850 }, { "grad_norm": 0.23878449201583862, "learning_rate": 3.018244678381873e-05, "loss": 0.0148, "step": 51860 }, { "grad_norm": 0.32981228828430176, "learning_rate": 3.0163472864495812e-05, "loss": 0.0188, "step": 51870 }, { "grad_norm": 0.18761712312698364, "learning_rate": 3.014450233469287e-05, "loss": 0.0138, "step": 51880 }, { "grad_norm": 0.23738111555576324, "learning_rate": 3.0125535197651422e-05, "loss": 0.0213, "step": 51890 }, { "grad_norm": 0.13980796933174133, "learning_rate": 3.0106571456612448e-05, "loss": 0.0091, "step": 51900 }, { "grad_norm": 0.31083470582962036, "learning_rate": 3.0087611114816305e-05, "loss": 0.0245, "step": 51910 }, { "grad_norm": 0.1729450225830078, "learning_rate": 3.006865417550284e-05, "loss": 0.0106, "step": 51920 }, { "grad_norm": 0.22171439230442047, "learning_rate": 3.0049700641911242e-05, "loss": 0.0147, "step": 51930 }, { "grad_norm": 1.8349368572235107, "learning_rate": 3.0030750517280175e-05, "loss": 0.014, "step": 51940 }, { "grad_norm": 0.2020130753517151, "learning_rate": 3.0011803804847682e-05, "loss": 0.0144, "step": 51950 }, { "grad_norm": 0.2983362078666687, "learning_rate": 2.9992860507851224e-05, "loss": 0.0142, "step": 51960 }, { "grad_norm": 0.1987074911594391, "learning_rate": 2.9973920629527723e-05, "loss": 0.0097, "step": 51970 }, { "grad_norm": 0.15564243495464325, "learning_rate": 2.9954984173113453e-05, "loss": 0.0122, "step": 51980 }, { "grad_norm": 0.3227483332157135, "learning_rate": 2.993605114184418e-05, "loss": 0.016, "step": 51990 }, { "grad_norm": 0.2379918247461319, "learning_rate": 2.9917121538955005e-05, "loss": 0.0207, "step": 52000 }, { "grad_norm": 0.187263622879982, "learning_rate": 2.98981953676805e-05, "loss": 0.0208, "step": 52010 }, { "grad_norm": 0.15726077556610107, "learning_rate": 2.9879272631254594e-05, "loss": 0.0119, "step": 52020 }, { "grad_norm": 0.14410895109176636, "learning_rate": 2.986035333291073e-05, "loss": 0.0108, "step": 52030 }, { "grad_norm": 0.14702892303466797, "learning_rate": 2.9841437475881646e-05, "loss": 0.0212, "step": 52040 }, { "grad_norm": 0.27448463439941406, "learning_rate": 2.9822525063399555e-05, "loss": 0.0117, "step": 52050 }, { "grad_norm": 0.21303339302539825, "learning_rate": 2.9803616098696087e-05, "loss": 0.0164, "step": 52060 }, { "grad_norm": 0.2180686593055725, "learning_rate": 2.9784710585002228e-05, "loss": 0.0142, "step": 52070 }, { "grad_norm": 0.2821659743785858, "learning_rate": 2.9765808525548467e-05, "loss": 0.0137, "step": 52080 }, { "grad_norm": 0.13828495144844055, "learning_rate": 2.974690992356461e-05, "loss": 0.0136, "step": 52090 }, { "grad_norm": 0.1258784383535385, "learning_rate": 2.9728014782279934e-05, "loss": 0.0172, "step": 52100 }, { "grad_norm": 0.3636297285556793, "learning_rate": 2.970912310492307e-05, "loss": 0.0115, "step": 52110 }, { "grad_norm": 0.12842957675457, "learning_rate": 2.9690234894722134e-05, "loss": 0.0117, "step": 52120 }, { "grad_norm": 0.16225582361221313, "learning_rate": 2.9671350154904577e-05, "loss": 0.0157, "step": 52130 }, { "grad_norm": 0.3931463956832886, "learning_rate": 2.9652468888697282e-05, "loss": 0.0129, "step": 52140 }, { "grad_norm": 0.16479340195655823, "learning_rate": 2.9633591099326562e-05, "loss": 0.0107, "step": 52150 }, { "grad_norm": 0.2740628123283386, "learning_rate": 2.9614716790018072e-05, "loss": 0.0102, "step": 52160 }, { "grad_norm": 0.194222092628479, "learning_rate": 2.959584596399697e-05, "loss": 0.0098, "step": 52170 }, { "grad_norm": 0.21445181965827942, "learning_rate": 2.9576978624487717e-05, "loss": 0.0161, "step": 52180 }, { "grad_norm": 0.20414409041404724, "learning_rate": 2.9558114774714263e-05, "loss": 0.0111, "step": 52190 }, { "grad_norm": 0.1636761873960495, "learning_rate": 2.9539254417899897e-05, "loss": 0.0127, "step": 52200 }, { "grad_norm": 0.14457182586193085, "learning_rate": 2.9520397557267365e-05, "loss": 0.0093, "step": 52210 }, { "grad_norm": 0.16799113154411316, "learning_rate": 2.9501544196038765e-05, "loss": 0.0119, "step": 52220 }, { "grad_norm": 0.18412749469280243, "learning_rate": 2.9482694337435614e-05, "loss": 0.0213, "step": 52230 }, { "grad_norm": 0.14125138521194458, "learning_rate": 2.946384798467887e-05, "loss": 0.0108, "step": 52240 }, { "grad_norm": 0.13201531767845154, "learning_rate": 2.9445005140988824e-05, "loss": 0.0081, "step": 52250 }, { "grad_norm": 0.11121626943349838, "learning_rate": 2.942616580958524e-05, "loss": 0.0102, "step": 52260 }, { "grad_norm": 0.1894652247428894, "learning_rate": 2.9407329993687193e-05, "loss": 0.0094, "step": 52270 }, { "grad_norm": 0.237808495759964, "learning_rate": 2.938849769651326e-05, "loss": 0.009, "step": 52280 }, { "grad_norm": 0.143040269613266, "learning_rate": 2.9369668921281345e-05, "loss": 0.0091, "step": 52290 }, { "grad_norm": 0.19455043971538544, "learning_rate": 2.9350843671208773e-05, "loss": 0.0093, "step": 52300 }, { "grad_norm": 0.16934478282928467, "learning_rate": 2.9332021949512245e-05, "loss": 0.0102, "step": 52310 }, { "grad_norm": 0.24752293527126312, "learning_rate": 2.9313203759407908e-05, "loss": 0.0113, "step": 52320 }, { "grad_norm": 0.20849430561065674, "learning_rate": 2.929438910411127e-05, "loss": 0.0228, "step": 52330 }, { "grad_norm": 0.3191393315792084, "learning_rate": 2.927557798683722e-05, "loss": 0.0138, "step": 52340 }, { "grad_norm": 0.1435224413871765, "learning_rate": 2.9256770410800095e-05, "loss": 0.0082, "step": 52350 }, { "grad_norm": 0.12630315124988556, "learning_rate": 2.9237966379213554e-05, "loss": 0.0091, "step": 52360 }, { "grad_norm": 0.25708574056625366, "learning_rate": 2.9219165895290736e-05, "loss": 0.0139, "step": 52370 }, { "grad_norm": 0.2593685984611511, "learning_rate": 2.9200368962244083e-05, "loss": 0.0153, "step": 52380 }, { "grad_norm": 0.18155807256698608, "learning_rate": 2.9181575583285536e-05, "loss": 0.0115, "step": 52390 }, { "grad_norm": 0.2499665915966034, "learning_rate": 2.916278576162632e-05, "loss": 0.0147, "step": 52400 }, { "grad_norm": 0.23413383960723877, "learning_rate": 2.9143999500477116e-05, "loss": 0.01, "step": 52410 }, { "grad_norm": 0.22165676951408386, "learning_rate": 2.9125216803048004e-05, "loss": 0.0162, "step": 52420 }, { "grad_norm": 0.20592404901981354, "learning_rate": 2.9106437672548403e-05, "loss": 0.0092, "step": 52430 }, { "grad_norm": 0.11124145984649658, "learning_rate": 2.9087662112187154e-05, "loss": 0.0101, "step": 52440 }, { "grad_norm": 0.14510643482208252, "learning_rate": 2.9068890125172498e-05, "loss": 0.009, "step": 52450 }, { "grad_norm": 0.1632167547941208, "learning_rate": 2.9050121714712058e-05, "loss": 0.0183, "step": 52460 }, { "grad_norm": 0.24495677649974823, "learning_rate": 2.9031356884012835e-05, "loss": 0.0109, "step": 52470 }, { "grad_norm": 0.1778874546289444, "learning_rate": 2.9012595636281247e-05, "loss": 0.0128, "step": 52480 }, { "grad_norm": 0.18465037643909454, "learning_rate": 2.8993837974723016e-05, "loss": 0.0234, "step": 52490 }, { "grad_norm": 0.2574763894081116, "learning_rate": 2.8975083902543394e-05, "loss": 0.0103, "step": 52500 }, { "grad_norm": 0.23566335439682007, "learning_rate": 2.895633342294688e-05, "loss": 0.0125, "step": 52510 }, { "grad_norm": 0.15765157341957092, "learning_rate": 2.893758653913744e-05, "loss": 0.0104, "step": 52520 }, { "grad_norm": 0.1504724770784378, "learning_rate": 2.8918843254318423e-05, "loss": 0.0097, "step": 52530 }, { "grad_norm": 0.31735777854919434, "learning_rate": 2.8900103571692483e-05, "loss": 0.012, "step": 52540 }, { "grad_norm": 0.18488650023937225, "learning_rate": 2.8881367494461796e-05, "loss": 0.0106, "step": 52550 }, { "grad_norm": 0.13939225673675537, "learning_rate": 2.886263502582779e-05, "loss": 0.0138, "step": 52560 }, { "grad_norm": 0.22332973778247833, "learning_rate": 2.8843906168991353e-05, "loss": 0.0102, "step": 52570 }, { "grad_norm": 0.16667070984840393, "learning_rate": 2.8825180927152728e-05, "loss": 0.0113, "step": 52580 }, { "grad_norm": 0.14578887820243835, "learning_rate": 2.8806459303511547e-05, "loss": 0.0096, "step": 52590 }, { "grad_norm": 0.21884344518184662, "learning_rate": 2.8787741301266852e-05, "loss": 0.0107, "step": 52600 }, { "grad_norm": 0.2507430613040924, "learning_rate": 2.876902692361697e-05, "loss": 0.0092, "step": 52610 }, { "grad_norm": 0.24138669669628143, "learning_rate": 2.875031617375976e-05, "loss": 0.0095, "step": 52620 }, { "grad_norm": 0.11891970038414001, "learning_rate": 2.8731609054892318e-05, "loss": 0.0085, "step": 52630 }, { "grad_norm": 0.20559751987457275, "learning_rate": 2.871290557021119e-05, "loss": 0.0116, "step": 52640 }, { "grad_norm": 0.1676540970802307, "learning_rate": 2.8694205722912298e-05, "loss": 0.0129, "step": 52650 }, { "grad_norm": 0.25859352946281433, "learning_rate": 2.8675509516190936e-05, "loss": 0.0149, "step": 52660 }, { "grad_norm": 0.3530738353729248, "learning_rate": 2.8656816953241772e-05, "loss": 0.0176, "step": 52670 }, { "grad_norm": 0.1670401394367218, "learning_rate": 2.863812803725887e-05, "loss": 0.0101, "step": 52680 }, { "grad_norm": 0.13008156418800354, "learning_rate": 2.8619442771435623e-05, "loss": 0.0086, "step": 52690 }, { "grad_norm": 0.24361814558506012, "learning_rate": 2.860076115896484e-05, "loss": 0.0161, "step": 52700 }, { "grad_norm": 0.1985911726951599, "learning_rate": 2.8582083203038712e-05, "loss": 0.0091, "step": 52710 }, { "grad_norm": 0.20030897855758667, "learning_rate": 2.8563408906848778e-05, "loss": 0.0092, "step": 52720 }, { "grad_norm": 0.20296819508075714, "learning_rate": 2.8544738273585993e-05, "loss": 0.0108, "step": 52730 }, { "grad_norm": 0.15530411899089813, "learning_rate": 2.8526071306440595e-05, "loss": 0.0079, "step": 52740 }, { "grad_norm": 0.24964028596878052, "learning_rate": 2.850740800860233e-05, "loss": 0.0102, "step": 52750 }, { "grad_norm": 0.1229800134897232, "learning_rate": 2.8488748383260178e-05, "loss": 0.0099, "step": 52760 }, { "grad_norm": 0.15942725539207458, "learning_rate": 2.8470092433602635e-05, "loss": 0.0159, "step": 52770 }, { "grad_norm": 0.1604471653699875, "learning_rate": 2.845144016281745e-05, "loss": 0.0114, "step": 52780 }, { "grad_norm": 0.19756445288658142, "learning_rate": 2.8432791574091744e-05, "loss": 0.0257, "step": 52790 }, { "grad_norm": 0.16900956630706787, "learning_rate": 2.8414146670612134e-05, "loss": 0.0153, "step": 52800 }, { "grad_norm": 0.14415661990642548, "learning_rate": 2.8395505455564446e-05, "loss": 0.015, "step": 52810 }, { "grad_norm": 0.15343698859214783, "learning_rate": 2.837686793213403e-05, "loss": 0.0107, "step": 52820 }, { "grad_norm": 0.20877808332443237, "learning_rate": 2.8358234103505477e-05, "loss": 0.0133, "step": 52830 }, { "grad_norm": 0.18826550245285034, "learning_rate": 2.8339603972862806e-05, "loss": 0.0107, "step": 52840 }, { "grad_norm": 0.1496615707874298, "learning_rate": 2.83209775433894e-05, "loss": 0.0108, "step": 52850 }, { "grad_norm": 0.20339104533195496, "learning_rate": 2.8302354818268013e-05, "loss": 0.0134, "step": 52860 }, { "grad_norm": 0.2774742841720581, "learning_rate": 2.8283735800680754e-05, "loss": 0.0148, "step": 52870 }, { "grad_norm": 0.15248043835163116, "learning_rate": 2.826512049380913e-05, "loss": 0.0144, "step": 52880 }, { "grad_norm": 0.17132745683193207, "learning_rate": 2.8246508900833935e-05, "loss": 0.019, "step": 52890 }, { "grad_norm": 0.16614104807376862, "learning_rate": 2.822790102493541e-05, "loss": 0.0133, "step": 52900 }, { "grad_norm": 0.14594466984272003, "learning_rate": 2.820929686929314e-05, "loss": 0.0099, "step": 52910 }, { "grad_norm": 0.2467995285987854, "learning_rate": 2.819069643708605e-05, "loss": 0.0116, "step": 52920 }, { "grad_norm": 0.23426537215709686, "learning_rate": 2.8172099731492484e-05, "loss": 0.0146, "step": 52930 }, { "grad_norm": 0.2511497437953949, "learning_rate": 2.815350675569003e-05, "loss": 0.0189, "step": 52940 }, { "grad_norm": 0.09916720539331436, "learning_rate": 2.8134917512855807e-05, "loss": 0.0203, "step": 52950 }, { "grad_norm": 0.15091700851917267, "learning_rate": 2.8116332006166134e-05, "loss": 0.0096, "step": 52960 }, { "grad_norm": 0.28174591064453125, "learning_rate": 2.809775023879685e-05, "loss": 0.0224, "step": 52970 }, { "grad_norm": 0.1230233684182167, "learning_rate": 2.807917221392299e-05, "loss": 0.0132, "step": 52980 }, { "grad_norm": 0.15355010330677032, "learning_rate": 2.8060597934719067e-05, "loss": 0.01, "step": 52990 }, { "grad_norm": 0.21262143552303314, "learning_rate": 2.8042027404358927e-05, "loss": 0.0128, "step": 53000 }, { "grad_norm": 0.2362869530916214, "learning_rate": 2.802346062601572e-05, "loss": 0.0108, "step": 53010 }, { "grad_norm": 0.30423593521118164, "learning_rate": 2.800489760286207e-05, "loss": 0.0183, "step": 53020 }, { "grad_norm": 0.20481203496456146, "learning_rate": 2.7986338338069834e-05, "loss": 0.0109, "step": 53030 }, { "grad_norm": 0.28837651014328003, "learning_rate": 2.7967782834810297e-05, "loss": 0.0187, "step": 53040 }, { "grad_norm": 0.1541733741760254, "learning_rate": 2.794923109625409e-05, "loss": 0.0127, "step": 53050 }, { "grad_norm": 0.24934570491313934, "learning_rate": 2.7930683125571205e-05, "loss": 0.0098, "step": 53060 }, { "grad_norm": 0.18605943024158478, "learning_rate": 2.7912138925930997e-05, "loss": 0.0144, "step": 53070 }, { "grad_norm": 0.13494302332401276, "learning_rate": 2.7893598500502117e-05, "loss": 0.027, "step": 53080 }, { "grad_norm": 0.35452187061309814, "learning_rate": 2.7875061852452644e-05, "loss": 0.0163, "step": 53090 }, { "grad_norm": 0.2889411747455597, "learning_rate": 2.7856528984949982e-05, "loss": 0.0122, "step": 53100 }, { "grad_norm": 0.16283194720745087, "learning_rate": 2.7837999901160888e-05, "loss": 0.0112, "step": 53110 }, { "grad_norm": 0.12997150421142578, "learning_rate": 2.7819474604251484e-05, "loss": 0.0094, "step": 53120 }, { "grad_norm": 0.15622283518314362, "learning_rate": 2.780095309738725e-05, "loss": 0.0107, "step": 53130 }, { "grad_norm": 0.1969609409570694, "learning_rate": 2.778243538373294e-05, "loss": 0.0116, "step": 53140 }, { "grad_norm": 0.134737029671669, "learning_rate": 2.7763921466452826e-05, "loss": 0.0128, "step": 53150 }, { "grad_norm": 0.18833589553833008, "learning_rate": 2.7745411348710336e-05, "loss": 0.0234, "step": 53160 }, { "grad_norm": 0.37151581048965454, "learning_rate": 2.7726905033668395e-05, "loss": 0.0209, "step": 53170 }, { "grad_norm": 0.24355529248714447, "learning_rate": 2.7708402524489214e-05, "loss": 0.0173, "step": 53180 }, { "grad_norm": 0.2660931944847107, "learning_rate": 2.7689903824334364e-05, "loss": 0.0151, "step": 53190 }, { "grad_norm": 0.2419872134923935, "learning_rate": 2.7671408936364785e-05, "loss": 0.0131, "step": 53200 }, { "grad_norm": 0.1672639548778534, "learning_rate": 2.765291786374069e-05, "loss": 0.0122, "step": 53210 }, { "grad_norm": 0.2665204405784607, "learning_rate": 2.7634430609621775e-05, "loss": 0.0097, "step": 53220 }, { "grad_norm": 0.2425096184015274, "learning_rate": 2.7615947177166956e-05, "loss": 0.0145, "step": 53230 }, { "grad_norm": 0.15152566134929657, "learning_rate": 2.7597467569534553e-05, "loss": 0.0079, "step": 53240 }, { "grad_norm": 0.15873606503009796, "learning_rate": 2.757899178988226e-05, "loss": 0.0109, "step": 53250 }, { "grad_norm": 0.18185245990753174, "learning_rate": 2.7560519841367005e-05, "loss": 0.0095, "step": 53260 }, { "grad_norm": 0.1566939651966095, "learning_rate": 2.7542051727145237e-05, "loss": 0.0114, "step": 53270 }, { "grad_norm": 0.23098352551460266, "learning_rate": 2.7523587450372578e-05, "loss": 0.0165, "step": 53280 }, { "grad_norm": 0.21278207004070282, "learning_rate": 2.750512701420409e-05, "loss": 0.0105, "step": 53290 }, { "grad_norm": 0.1283673793077469, "learning_rate": 2.7486670421794158e-05, "loss": 0.0105, "step": 53300 }, { "grad_norm": 0.1305108517408371, "learning_rate": 2.7468217676296515e-05, "loss": 0.0211, "step": 53310 }, { "grad_norm": 0.22888623178005219, "learning_rate": 2.7449768780864226e-05, "loss": 0.0147, "step": 53320 }, { "grad_norm": 0.18182650208473206, "learning_rate": 2.7431323738649724e-05, "loss": 0.01, "step": 53330 }, { "grad_norm": 0.344119131565094, "learning_rate": 2.7412882552804713e-05, "loss": 0.0155, "step": 53340 }, { "grad_norm": 0.14406231045722961, "learning_rate": 2.739444522648032e-05, "loss": 0.0089, "step": 53350 }, { "grad_norm": 0.14051081240177155, "learning_rate": 2.737601176282697e-05, "loss": 0.0115, "step": 53360 }, { "grad_norm": 0.24947963654994965, "learning_rate": 2.735758216499445e-05, "loss": 0.0104, "step": 53370 }, { "grad_norm": 0.1327831745147705, "learning_rate": 2.7339156436131864e-05, "loss": 0.0101, "step": 53380 }, { "grad_norm": 0.2188619077205658, "learning_rate": 2.7320734579387663e-05, "loss": 0.012, "step": 53390 }, { "grad_norm": 0.13422435522079468, "learning_rate": 2.730231659790966e-05, "loss": 0.008, "step": 53400 }, { "grad_norm": 0.14560635387897491, "learning_rate": 2.728390249484494e-05, "loss": 0.0177, "step": 53410 }, { "grad_norm": 0.17560100555419922, "learning_rate": 2.7265492273340032e-05, "loss": 0.0097, "step": 53420 }, { "grad_norm": 0.18596665561199188, "learning_rate": 2.7247085936540678e-05, "loss": 0.0097, "step": 53430 }, { "grad_norm": 0.12854912877082825, "learning_rate": 2.722868348759205e-05, "loss": 0.0112, "step": 53440 }, { "grad_norm": 0.1708814799785614, "learning_rate": 2.7210284929638635e-05, "loss": 0.0142, "step": 53450 }, { "grad_norm": 0.2665359079837799, "learning_rate": 2.7191890265824183e-05, "loss": 0.0115, "step": 53460 }, { "grad_norm": 0.10968462377786636, "learning_rate": 2.7173499499291926e-05, "loss": 0.0095, "step": 53470 }, { "grad_norm": 0.33045023679733276, "learning_rate": 2.7155112633184277e-05, "loss": 0.0126, "step": 53480 }, { "grad_norm": 0.2674945592880249, "learning_rate": 2.713672967064307e-05, "loss": 0.0151, "step": 53490 }, { "grad_norm": 0.15414740145206451, "learning_rate": 2.711835061480945e-05, "loss": 0.0088, "step": 53500 }, { "grad_norm": 0.21951280534267426, "learning_rate": 2.7099975468823896e-05, "loss": 0.0137, "step": 53510 }, { "grad_norm": 0.15447625517845154, "learning_rate": 2.708160423582622e-05, "loss": 0.0118, "step": 53520 }, { "grad_norm": 0.4175197184085846, "learning_rate": 2.706323691895557e-05, "loss": 0.0199, "step": 53530 }, { "grad_norm": 0.1635657250881195, "learning_rate": 2.70448735213504e-05, "loss": 0.0083, "step": 53540 }, { "grad_norm": 0.11191409081220627, "learning_rate": 2.702651404614852e-05, "loss": 0.0142, "step": 53550 }, { "grad_norm": 0.28394949436187744, "learning_rate": 2.7008158496487056e-05, "loss": 0.01, "step": 53560 }, { "grad_norm": 0.11003033071756363, "learning_rate": 2.6989806875502487e-05, "loss": 0.0098, "step": 53570 }, { "grad_norm": 0.19337889552116394, "learning_rate": 2.6971459186330584e-05, "loss": 0.0113, "step": 53580 }, { "grad_norm": 0.19248686730861664, "learning_rate": 2.695311543210648e-05, "loss": 0.011, "step": 53590 }, { "grad_norm": 0.22911246120929718, "learning_rate": 2.693477561596463e-05, "loss": 0.0131, "step": 53600 }, { "grad_norm": 0.24099883437156677, "learning_rate": 2.6916439741038756e-05, "loss": 0.0109, "step": 53610 }, { "grad_norm": 0.22942551970481873, "learning_rate": 2.689810781046203e-05, "loss": 0.0107, "step": 53620 }, { "grad_norm": 0.21892575919628143, "learning_rate": 2.6879779827366823e-05, "loss": 0.0108, "step": 53630 }, { "grad_norm": 0.19340291619300842, "learning_rate": 2.6861455794884904e-05, "loss": 0.0097, "step": 53640 }, { "grad_norm": 0.14817966520786285, "learning_rate": 2.6843135716147373e-05, "loss": 0.0113, "step": 53650 }, { "grad_norm": 0.23197811841964722, "learning_rate": 2.6824819594284556e-05, "loss": 0.0141, "step": 53660 }, { "grad_norm": 0.22190295159816742, "learning_rate": 2.6806507432426275e-05, "loss": 0.0104, "step": 53670 }, { "grad_norm": 0.14850591123104095, "learning_rate": 2.6788199233701512e-05, "loss": 0.0116, "step": 53680 }, { "grad_norm": 0.19243672490119934, "learning_rate": 2.6769895001238652e-05, "loss": 0.0169, "step": 53690 }, { "grad_norm": 0.2867184281349182, "learning_rate": 2.67515947381654e-05, "loss": 0.0117, "step": 53700 }, { "grad_norm": 0.15030305087566376, "learning_rate": 2.6733298447608745e-05, "loss": 0.01, "step": 53710 }, { "grad_norm": 0.10294989496469498, "learning_rate": 2.6715006132695074e-05, "loss": 0.0141, "step": 53720 }, { "grad_norm": 0.17289331555366516, "learning_rate": 2.6696717796549985e-05, "loss": 0.019, "step": 53730 }, { "grad_norm": 0.25208231806755066, "learning_rate": 2.667843344229848e-05, "loss": 0.0095, "step": 53740 }, { "grad_norm": 0.27571967244148254, "learning_rate": 2.6660153073064846e-05, "loss": 0.0137, "step": 53750 }, { "grad_norm": 0.10269142687320709, "learning_rate": 2.664187669197271e-05, "loss": 0.0188, "step": 53760 }, { "grad_norm": 0.2421049028635025, "learning_rate": 2.6623604302145005e-05, "loss": 0.01, "step": 53770 }, { "grad_norm": 0.28422021865844727, "learning_rate": 2.6605335906703975e-05, "loss": 0.0119, "step": 53780 }, { "grad_norm": 0.1355682909488678, "learning_rate": 2.65870715087712e-05, "loss": 0.0134, "step": 53790 }, { "grad_norm": 0.2604723870754242, "learning_rate": 2.6568811111467573e-05, "loss": 0.0105, "step": 53800 }, { "grad_norm": 0.14502008259296417, "learning_rate": 2.6550554717913258e-05, "loss": 0.0153, "step": 53810 }, { "grad_norm": 0.22646231949329376, "learning_rate": 2.6532302331227805e-05, "loss": 0.0134, "step": 53820 }, { "grad_norm": 0.15067531168460846, "learning_rate": 2.651405395453004e-05, "loss": 0.0097, "step": 53830 }, { "grad_norm": 0.10463234782218933, "learning_rate": 2.6495809590938115e-05, "loss": 0.0149, "step": 53840 }, { "grad_norm": 0.3384922444820404, "learning_rate": 2.64775692435695e-05, "loss": 0.0179, "step": 53850 }, { "grad_norm": 0.19411930441856384, "learning_rate": 2.6459332915540928e-05, "loss": 0.0083, "step": 53860 }, { "grad_norm": 0.302284836769104, "learning_rate": 2.644110060996856e-05, "loss": 0.0168, "step": 53870 }, { "grad_norm": 0.16970327496528625, "learning_rate": 2.642287232996774e-05, "loss": 0.0101, "step": 53880 }, { "grad_norm": 0.17271360754966736, "learning_rate": 2.6404648078653205e-05, "loss": 0.0085, "step": 53890 }, { "grad_norm": 0.14861001074314117, "learning_rate": 2.6386427859139002e-05, "loss": 0.0075, "step": 53900 }, { "grad_norm": 0.28030332922935486, "learning_rate": 2.63682116745384e-05, "loss": 0.0161, "step": 53910 }, { "grad_norm": 0.13799726963043213, "learning_rate": 2.6349999527964138e-05, "loss": 0.0084, "step": 53920 }, { "grad_norm": 0.15833508968353271, "learning_rate": 2.63317914225281e-05, "loss": 0.0192, "step": 53930 }, { "grad_norm": 0.19541411101818085, "learning_rate": 2.6313587361341585e-05, "loss": 0.0205, "step": 53940 }, { "grad_norm": 0.4012134373188019, "learning_rate": 2.6295387347515165e-05, "loss": 0.0197, "step": 53950 }, { "grad_norm": 0.207090824842453, "learning_rate": 2.6277191384158727e-05, "loss": 0.0193, "step": 53960 }, { "grad_norm": 0.13213428854942322, "learning_rate": 2.6258999474381457e-05, "loss": 0.0192, "step": 53970 }, { "grad_norm": 0.20170512795448303, "learning_rate": 2.624081162129186e-05, "loss": 0.0099, "step": 53980 }, { "grad_norm": 0.1833004504442215, "learning_rate": 2.6222627827997765e-05, "loss": 0.0233, "step": 53990 }, { "grad_norm": 0.187967911362648, "learning_rate": 2.6204448097606236e-05, "loss": 0.0092, "step": 54000 }, { "grad_norm": 0.16287070512771606, "learning_rate": 2.6186272433223726e-05, "loss": 0.013, "step": 54010 }, { "grad_norm": 0.13581284880638123, "learning_rate": 2.6168100837955943e-05, "loss": 0.012, "step": 54020 }, { "grad_norm": 0.16096490621566772, "learning_rate": 2.6149933314907926e-05, "loss": 0.0226, "step": 54030 }, { "grad_norm": 0.28838911652565, "learning_rate": 2.6131769867184e-05, "loss": 0.0163, "step": 54040 }, { "grad_norm": 0.11568349599838257, "learning_rate": 2.611361049788783e-05, "loss": 0.0084, "step": 54050 }, { "grad_norm": 0.22705860435962677, "learning_rate": 2.6095455210122292e-05, "loss": 0.0096, "step": 54060 }, { "grad_norm": 0.12687167525291443, "learning_rate": 2.6077304006989712e-05, "loss": 0.0075, "step": 54070 }, { "grad_norm": 0.24453912675380707, "learning_rate": 2.6059156891591562e-05, "loss": 0.0112, "step": 54080 }, { "grad_norm": 0.1730869710445404, "learning_rate": 2.6041013867028718e-05, "loss": 0.0099, "step": 54090 }, { "grad_norm": 0.2228143960237503, "learning_rate": 2.6022874936401347e-05, "loss": 0.0093, "step": 54100 }, { "grad_norm": 0.1898318976163864, "learning_rate": 2.6004740102808832e-05, "loss": 0.015, "step": 54110 }, { "grad_norm": 0.09741530567407608, "learning_rate": 2.598660936935e-05, "loss": 0.0085, "step": 54120 }, { "grad_norm": 0.26516205072402954, "learning_rate": 2.5968482739122845e-05, "loss": 0.0152, "step": 54130 }, { "grad_norm": 0.24821822345256805, "learning_rate": 2.595036021522472e-05, "loss": 0.0112, "step": 54140 }, { "grad_norm": 0.1880558580160141, "learning_rate": 2.5932241800752278e-05, "loss": 0.0126, "step": 54150 }, { "grad_norm": 0.2713954746723175, "learning_rate": 2.5914127498801453e-05, "loss": 0.0263, "step": 54160 }, { "grad_norm": 0.14193952083587646, "learning_rate": 2.5896017312467497e-05, "loss": 0.01, "step": 54170 }, { "grad_norm": 0.282900869846344, "learning_rate": 2.587791124484493e-05, "loss": 0.0116, "step": 54180 }, { "grad_norm": 0.09929417818784714, "learning_rate": 2.5859809299027615e-05, "loss": 0.0118, "step": 54190 }, { "grad_norm": 0.19417575001716614, "learning_rate": 2.5841711478108632e-05, "loss": 0.0186, "step": 54200 }, { "grad_norm": 0.19570867717266083, "learning_rate": 2.582361778518043e-05, "loss": 0.0116, "step": 54210 }, { "grad_norm": 0.11059019714593887, "learning_rate": 2.580552822333472e-05, "loss": 0.0124, "step": 54220 }, { "grad_norm": 0.18057848513126373, "learning_rate": 2.578744279566252e-05, "loss": 0.0101, "step": 54230 }, { "grad_norm": 0.14647522568702698, "learning_rate": 2.576936150525413e-05, "loss": 0.0205, "step": 54240 }, { "grad_norm": 0.12849746644496918, "learning_rate": 2.5751284355199168e-05, "loss": 0.0088, "step": 54250 }, { "grad_norm": 0.12880849838256836, "learning_rate": 2.573321134858646e-05, "loss": 0.0076, "step": 54260 }, { "grad_norm": 0.12502941489219666, "learning_rate": 2.5715142488504286e-05, "loss": 0.0084, "step": 54270 }, { "grad_norm": 0.09495427459478378, "learning_rate": 2.5697077778040042e-05, "loss": 0.0096, "step": 54280 }, { "grad_norm": 0.2094879448413849, "learning_rate": 2.5679017220280522e-05, "loss": 0.0246, "step": 54290 }, { "grad_norm": 0.10638003796339035, "learning_rate": 2.5660960818311796e-05, "loss": 0.0096, "step": 54300 }, { "grad_norm": 0.28128308057785034, "learning_rate": 2.564290857521915e-05, "loss": 0.0133, "step": 54310 }, { "grad_norm": 0.2267896682024002, "learning_rate": 2.5624860494087298e-05, "loss": 0.0114, "step": 54320 }, { "grad_norm": 0.13261710107326508, "learning_rate": 2.5606816578000115e-05, "loss": 0.018, "step": 54330 }, { "grad_norm": 0.19828057289123535, "learning_rate": 2.558877683004082e-05, "loss": 0.0113, "step": 54340 }, { "grad_norm": 0.1206427589058876, "learning_rate": 2.557074125329192e-05, "loss": 0.0082, "step": 54350 }, { "grad_norm": 0.19994504749774933, "learning_rate": 2.5552709850835195e-05, "loss": 0.0121, "step": 54360 }, { "grad_norm": 0.12338163703680038, "learning_rate": 2.5534682625751738e-05, "loss": 0.0084, "step": 54370 }, { "grad_norm": 0.19422981142997742, "learning_rate": 2.551665958112186e-05, "loss": 0.0211, "step": 54380 }, { "grad_norm": 0.1633588969707489, "learning_rate": 2.549864072002527e-05, "loss": 0.008, "step": 54390 }, { "grad_norm": 0.22645258903503418, "learning_rate": 2.5480626045540858e-05, "loss": 0.0128, "step": 54400 }, { "grad_norm": 0.2259775996208191, "learning_rate": 2.546261556074684e-05, "loss": 0.0099, "step": 54410 }, { "grad_norm": 0.12920568883419037, "learning_rate": 2.5444609268720726e-05, "loss": 0.0091, "step": 54420 }, { "grad_norm": 0.10308614373207092, "learning_rate": 2.5426607172539297e-05, "loss": 0.0092, "step": 54430 }, { "grad_norm": 0.11512648314237595, "learning_rate": 2.5408609275278617e-05, "loss": 0.019, "step": 54440 }, { "grad_norm": 0.13589617609977722, "learning_rate": 2.5390615580014055e-05, "loss": 0.0086, "step": 54450 }, { "grad_norm": 0.14352717995643616, "learning_rate": 2.5372626089820207e-05, "loss": 0.0155, "step": 54460 }, { "grad_norm": 0.17742659151554108, "learning_rate": 2.5354640807770997e-05, "loss": 0.0101, "step": 54470 }, { "grad_norm": 0.2618962526321411, "learning_rate": 2.5336659736939622e-05, "loss": 0.0125, "step": 54480 }, { "grad_norm": 0.12221042811870575, "learning_rate": 2.5318682880398554e-05, "loss": 0.0147, "step": 54490 }, { "grad_norm": 0.17295043170452118, "learning_rate": 2.530071024121956e-05, "loss": 0.0141, "step": 54500 }, { "grad_norm": 0.2676989734172821, "learning_rate": 2.5282741822473627e-05, "loss": 0.0099, "step": 54510 }, { "grad_norm": 0.2510339021682739, "learning_rate": 2.526477762723114e-05, "loss": 0.0139, "step": 54520 }, { "grad_norm": 0.16628773510456085, "learning_rate": 2.5246817658561618e-05, "loss": 0.0118, "step": 54530 }, { "grad_norm": 0.1529345065355301, "learning_rate": 2.5228861919533965e-05, "loss": 0.0101, "step": 54540 }, { "grad_norm": 0.18431422114372253, "learning_rate": 2.5210910413216326e-05, "loss": 0.0223, "step": 54550 }, { "grad_norm": 0.17894472181797028, "learning_rate": 2.5192963142676086e-05, "loss": 0.0062, "step": 54560 }, { "grad_norm": 0.22897133231163025, "learning_rate": 2.517502011098001e-05, "loss": 0.0085, "step": 54570 }, { "grad_norm": 0.13152317702770233, "learning_rate": 2.5157081321193987e-05, "loss": 0.0141, "step": 54580 }, { "grad_norm": 0.18449856340885162, "learning_rate": 2.5139146776383356e-05, "loss": 0.0094, "step": 54590 }, { "grad_norm": 0.14976464211940765, "learning_rate": 2.5121216479612575e-05, "loss": 0.0103, "step": 54600 }, { "grad_norm": 0.15944060683250427, "learning_rate": 2.510329043394546e-05, "loss": 0.0192, "step": 54610 }, { "grad_norm": 0.2011605203151703, "learning_rate": 2.508536864244508e-05, "loss": 0.0136, "step": 54620 }, { "grad_norm": 0.26173779368400574, "learning_rate": 2.5067451108173778e-05, "loss": 0.0115, "step": 54630 }, { "grad_norm": 0.12859004735946655, "learning_rate": 2.5049537834193204e-05, "loss": 0.0095, "step": 54640 }, { "grad_norm": 0.11223483085632324, "learning_rate": 2.5031628823564194e-05, "loss": 0.0103, "step": 54650 }, { "grad_norm": 0.22052496671676636, "learning_rate": 2.5013724079346933e-05, "loss": 0.011, "step": 54660 }, { "grad_norm": 0.18442431092262268, "learning_rate": 2.4995823604600854e-05, "loss": 0.0086, "step": 54670 }, { "grad_norm": 0.1428445726633072, "learning_rate": 2.497792740238465e-05, "loss": 0.0092, "step": 54680 }, { "grad_norm": 0.22017787396907806, "learning_rate": 2.49600354757563e-05, "loss": 0.0172, "step": 54690 }, { "grad_norm": 0.18731507658958435, "learning_rate": 2.494214782777306e-05, "loss": 0.0137, "step": 54700 }, { "grad_norm": 0.25776591897010803, "learning_rate": 2.4924264461491386e-05, "loss": 0.0128, "step": 54710 }, { "grad_norm": 0.251329630613327, "learning_rate": 2.4906385379967133e-05, "loss": 0.0191, "step": 54720 }, { "grad_norm": 0.15368014574050903, "learning_rate": 2.4888510586255285e-05, "loss": 0.0118, "step": 54730 }, { "grad_norm": 0.18275503814220428, "learning_rate": 2.487064008341018e-05, "loss": 0.0135, "step": 54740 }, { "grad_norm": 0.13908983767032623, "learning_rate": 2.4852773874485407e-05, "loss": 0.0083, "step": 54750 }, { "grad_norm": 0.17744697630405426, "learning_rate": 2.483491196253377e-05, "loss": 0.011, "step": 54760 }, { "grad_norm": 0.19938859343528748, "learning_rate": 2.4817054350607443e-05, "loss": 0.0126, "step": 54770 }, { "grad_norm": 0.1504499316215515, "learning_rate": 2.4799201041757742e-05, "loss": 0.0073, "step": 54780 }, { "grad_norm": 0.24279451370239258, "learning_rate": 2.4781352039035373e-05, "loss": 0.0105, "step": 54790 }, { "grad_norm": 0.20740915834903717, "learning_rate": 2.4763507345490194e-05, "loss": 0.0093, "step": 54800 }, { "grad_norm": 0.17132240533828735, "learning_rate": 2.4745666964171386e-05, "loss": 0.0143, "step": 54810 }, { "grad_norm": 0.3174591064453125, "learning_rate": 2.4727830898127387e-05, "loss": 0.0119, "step": 54820 }, { "grad_norm": 0.11797848343849182, "learning_rate": 2.4709999150405895e-05, "loss": 0.0195, "step": 54830 }, { "grad_norm": 0.3093706965446472, "learning_rate": 2.469217172405388e-05, "loss": 0.012, "step": 54840 }, { "grad_norm": 0.13071279227733612, "learning_rate": 2.4674348622117527e-05, "loss": 0.007, "step": 54850 }, { "grad_norm": 0.19997228682041168, "learning_rate": 2.465652984764234e-05, "loss": 0.0083, "step": 54860 }, { "grad_norm": 0.1664491444826126, "learning_rate": 2.4638715403673056e-05, "loss": 0.0075, "step": 54870 }, { "grad_norm": 0.25694629549980164, "learning_rate": 2.4620905293253676e-05, "loss": 0.012, "step": 54880 }, { "grad_norm": 0.2898372709751129, "learning_rate": 2.4603099519427463e-05, "loss": 0.0116, "step": 54890 }, { "grad_norm": 0.3468640148639679, "learning_rate": 2.458529808523695e-05, "loss": 0.0096, "step": 54900 }, { "grad_norm": 0.1717825084924698, "learning_rate": 2.456750099372387e-05, "loss": 0.0084, "step": 54910 }, { "grad_norm": 0.1719883233308792, "learning_rate": 2.4549708247929327e-05, "loss": 0.0167, "step": 54920 }, { "grad_norm": 0.17014120519161224, "learning_rate": 2.4531919850893554e-05, "loss": 0.0091, "step": 54930 }, { "grad_norm": 0.11201226711273193, "learning_rate": 2.4514135805656125e-05, "loss": 0.0082, "step": 54940 }, { "grad_norm": 0.24772332608699799, "learning_rate": 2.449635611525587e-05, "loss": 0.0153, "step": 54950 }, { "grad_norm": 0.15609192848205566, "learning_rate": 2.447858078273079e-05, "loss": 0.0132, "step": 54960 }, { "grad_norm": 0.2067539244890213, "learning_rate": 2.4460809811118275e-05, "loss": 0.0224, "step": 54970 }, { "grad_norm": 0.11619102209806442, "learning_rate": 2.444304320345483e-05, "loss": 0.0084, "step": 54980 }, { "grad_norm": 0.15937180817127228, "learning_rate": 2.4425280962776354e-05, "loss": 0.0133, "step": 54990 }, { "grad_norm": 0.2179512232542038, "learning_rate": 2.4407523092117875e-05, "loss": 0.014, "step": 55000 }, { "grad_norm": 0.27322089672088623, "learning_rate": 2.438976959451374e-05, "loss": 0.009, "step": 55010 }, { "grad_norm": 0.22648932039737701, "learning_rate": 2.4372020472997565e-05, "loss": 0.0109, "step": 55020 }, { "grad_norm": 0.3855288326740265, "learning_rate": 2.4354275730602122e-05, "loss": 0.0119, "step": 55030 }, { "grad_norm": 0.17403753101825714, "learning_rate": 2.433653537035958e-05, "loss": 0.0136, "step": 55040 }, { "grad_norm": 0.20971348881721497, "learning_rate": 2.431879939530123e-05, "loss": 0.0116, "step": 55050 }, { "grad_norm": 0.1980225145816803, "learning_rate": 2.4301067808457684e-05, "loss": 0.0105, "step": 55060 }, { "grad_norm": 0.16916882991790771, "learning_rate": 2.428334061285878e-05, "loss": 0.0125, "step": 55070 }, { "grad_norm": 0.29070088267326355, "learning_rate": 2.426561781153361e-05, "loss": 0.0107, "step": 55080 }, { "grad_norm": 0.18092463910579681, "learning_rate": 2.424789940751052e-05, "loss": 0.0087, "step": 55090 }, { "grad_norm": 0.30781373381614685, "learning_rate": 2.423018540381712e-05, "loss": 0.012, "step": 55100 }, { "grad_norm": 0.1591530442237854, "learning_rate": 2.421247580348021e-05, "loss": 0.015, "step": 55110 }, { "grad_norm": 0.20464932918548584, "learning_rate": 2.4194770609525897e-05, "loss": 0.008, "step": 55120 }, { "grad_norm": 0.1257612705230713, "learning_rate": 2.417706982497951e-05, "loss": 0.0131, "step": 55130 }, { "grad_norm": 0.19546538591384888, "learning_rate": 2.415937345286563e-05, "loss": 0.017, "step": 55140 }, { "grad_norm": 0.23470807075500488, "learning_rate": 2.4141681496208087e-05, "loss": 0.0113, "step": 55150 }, { "grad_norm": 0.20749229192733765, "learning_rate": 2.4123993958029946e-05, "loss": 0.0093, "step": 55160 }, { "grad_norm": 0.2186969816684723, "learning_rate": 2.4106310841353548e-05, "loss": 0.0087, "step": 55170 }, { "grad_norm": 0.17450357973575592, "learning_rate": 2.4088632149200398e-05, "loss": 0.017, "step": 55180 }, { "grad_norm": 0.13215523958206177, "learning_rate": 2.4070957884591367e-05, "loss": 0.0084, "step": 55190 }, { "grad_norm": 0.347584068775177, "learning_rate": 2.4053288050546464e-05, "loss": 0.0166, "step": 55200 }, { "grad_norm": 0.2441045194864273, "learning_rate": 2.403562265008498e-05, "loss": 0.0111, "step": 55210 }, { "grad_norm": 0.21990054845809937, "learning_rate": 2.4017961686225483e-05, "loss": 0.014, "step": 55220 }, { "grad_norm": 0.19484928250312805, "learning_rate": 2.400030516198568e-05, "loss": 0.0234, "step": 55230 }, { "grad_norm": 0.0959366038441658, "learning_rate": 2.3982653080382673e-05, "loss": 0.0094, "step": 55240 }, { "grad_norm": 0.18253841996192932, "learning_rate": 2.396500544443266e-05, "loss": 0.0097, "step": 55250 }, { "grad_norm": 0.16001127660274506, "learning_rate": 2.3947362257151156e-05, "loss": 0.0092, "step": 55260 }, { "grad_norm": 0.2516876757144928, "learning_rate": 2.39297235215529e-05, "loss": 0.0114, "step": 55270 }, { "grad_norm": 0.20285442471504211, "learning_rate": 2.3912089240651873e-05, "loss": 0.0149, "step": 55280 }, { "grad_norm": 0.13482150435447693, "learning_rate": 2.389445941746129e-05, "loss": 0.0187, "step": 55290 }, { "grad_norm": 0.28222012519836426, "learning_rate": 2.3876834054993625e-05, "loss": 0.0089, "step": 55300 }, { "grad_norm": 0.158816859126091, "learning_rate": 2.3859213156260522e-05, "loss": 0.0094, "step": 55310 }, { "grad_norm": 0.18633955717086792, "learning_rate": 2.3841596724272948e-05, "loss": 0.0126, "step": 55320 }, { "grad_norm": 0.16321654617786407, "learning_rate": 2.382398476204106e-05, "loss": 0.0121, "step": 55330 }, { "grad_norm": 0.13282817602157593, "learning_rate": 2.3806377272574254e-05, "loss": 0.0099, "step": 55340 }, { "grad_norm": 0.19431772828102112, "learning_rate": 2.3788774258881174e-05, "loss": 0.0111, "step": 55350 }, { "grad_norm": 0.3000277578830719, "learning_rate": 2.37711757239697e-05, "loss": 0.0102, "step": 55360 }, { "grad_norm": 0.28814905881881714, "learning_rate": 2.3753581670846954e-05, "loss": 0.0093, "step": 55370 }, { "grad_norm": 0.12443835288286209, "learning_rate": 2.3735992102519216e-05, "loss": 0.0091, "step": 55380 }, { "grad_norm": 0.12041394412517548, "learning_rate": 2.371840702199215e-05, "loss": 0.0101, "step": 55390 }, { "grad_norm": 0.14405658841133118, "learning_rate": 2.3700826432270494e-05, "loss": 0.0069, "step": 55400 }, { "grad_norm": 0.18789416551589966, "learning_rate": 2.3683250336358326e-05, "loss": 0.0186, "step": 55410 }, { "grad_norm": 0.16386869549751282, "learning_rate": 2.3665678737258923e-05, "loss": 0.018, "step": 55420 }, { "grad_norm": 0.08413026481866837, "learning_rate": 2.3648111637974745e-05, "loss": 0.0076, "step": 55430 }, { "grad_norm": 0.1698838621377945, "learning_rate": 2.36305490415076e-05, "loss": 0.0157, "step": 55440 }, { "grad_norm": 0.210637167096138, "learning_rate": 2.3612990950858392e-05, "loss": 0.0078, "step": 55450 }, { "grad_norm": 0.0959385558962822, "learning_rate": 2.359543736902735e-05, "loss": 0.0096, "step": 55460 }, { "grad_norm": 0.16257750988006592, "learning_rate": 2.3577888299013896e-05, "loss": 0.0107, "step": 55470 }, { "grad_norm": 0.29385632276535034, "learning_rate": 2.3560343743816683e-05, "loss": 0.0076, "step": 55480 }, { "grad_norm": 0.15748490393161774, "learning_rate": 2.354280370643362e-05, "loss": 0.016, "step": 55490 }, { "grad_norm": 0.20822253823280334, "learning_rate": 2.3525268189861777e-05, "loss": 0.0096, "step": 55500 }, { "grad_norm": 0.2759939134120941, "learning_rate": 2.3507737197097513e-05, "loss": 0.0096, "step": 55510 }, { "grad_norm": 0.19509069621562958, "learning_rate": 2.34902107311364e-05, "loss": 0.0104, "step": 55520 }, { "grad_norm": 0.14142900705337524, "learning_rate": 2.3472688794973225e-05, "loss": 0.0133, "step": 55530 }, { "grad_norm": 0.1320847123861313, "learning_rate": 2.3455171391602016e-05, "loss": 0.0099, "step": 55540 }, { "grad_norm": 0.2701331079006195, "learning_rate": 2.343765852401601e-05, "loss": 0.0087, "step": 55550 }, { "grad_norm": 0.10881424695253372, "learning_rate": 2.342015019520768e-05, "loss": 0.0132, "step": 55560 }, { "grad_norm": 0.12303429841995239, "learning_rate": 2.3402646408168742e-05, "loss": 0.0086, "step": 55570 }, { "grad_norm": 0.2904508113861084, "learning_rate": 2.3385147165890074e-05, "loss": 0.0168, "step": 55580 }, { "grad_norm": 0.18552061915397644, "learning_rate": 2.336765247136184e-05, "loss": 0.0111, "step": 55590 }, { "grad_norm": 0.28878897428512573, "learning_rate": 2.33501623275734e-05, "loss": 0.0172, "step": 55600 }, { "grad_norm": 0.11332985758781433, "learning_rate": 2.333267673751334e-05, "loss": 0.0077, "step": 55610 }, { "grad_norm": 0.13450634479522705, "learning_rate": 2.33151957041695e-05, "loss": 0.0084, "step": 55620 }, { "grad_norm": 0.202538400888443, "learning_rate": 2.329771923052884e-05, "loss": 0.011, "step": 55630 }, { "grad_norm": 0.140974223613739, "learning_rate": 2.3280247319577697e-05, "loss": 0.0103, "step": 55640 }, { "grad_norm": 0.2098013162612915, "learning_rate": 2.3262779974301473e-05, "loss": 0.012, "step": 55650 }, { "grad_norm": 0.16310280561447144, "learning_rate": 2.3245317197684895e-05, "loss": 0.0103, "step": 55660 }, { "grad_norm": 0.2019478976726532, "learning_rate": 2.322785899271188e-05, "loss": 0.0104, "step": 55670 }, { "grad_norm": 0.1601162999868393, "learning_rate": 2.3210405362365507e-05, "loss": 0.0161, "step": 55680 }, { "grad_norm": 0.18631288409233093, "learning_rate": 2.31929563096282e-05, "loss": 0.0123, "step": 55690 }, { "grad_norm": 0.21403108537197113, "learning_rate": 2.317551183748146e-05, "loss": 0.0092, "step": 55700 }, { "grad_norm": 0.12110218405723572, "learning_rate": 2.3158071948906103e-05, "loss": 0.0106, "step": 55710 }, { "grad_norm": 0.1167701855301857, "learning_rate": 2.314063664688212e-05, "loss": 0.0077, "step": 55720 }, { "grad_norm": 0.15729080140590668, "learning_rate": 2.3123205934388725e-05, "loss": 0.0099, "step": 55730 }, { "grad_norm": 0.20637129247188568, "learning_rate": 2.310577981440436e-05, "loss": 0.0191, "step": 55740 }, { "grad_norm": 0.10838453471660614, "learning_rate": 2.3088358289906665e-05, "loss": 0.0076, "step": 55750 }, { "grad_norm": 0.21550166606903076, "learning_rate": 2.307094136387252e-05, "loss": 0.0088, "step": 55760 }, { "grad_norm": 0.2131248265504837, "learning_rate": 2.305352903927796e-05, "loss": 0.0151, "step": 55770 }, { "grad_norm": 0.1594085842370987, "learning_rate": 2.303612131909831e-05, "loss": 0.0078, "step": 55780 }, { "grad_norm": 0.20993965864181519, "learning_rate": 2.3018718206308054e-05, "loss": 0.0155, "step": 55790 }, { "grad_norm": 0.22003141045570374, "learning_rate": 2.3001319703880925e-05, "loss": 0.0126, "step": 55800 }, { "grad_norm": 0.18325215578079224, "learning_rate": 2.2983925814789835e-05, "loss": 0.0107, "step": 55810 }, { "grad_norm": 0.17416946589946747, "learning_rate": 2.2966536542006957e-05, "loss": 0.0103, "step": 55820 }, { "grad_norm": 0.12965960800647736, "learning_rate": 2.294915188850358e-05, "loss": 0.0104, "step": 55830 }, { "grad_norm": 0.18213696777820587, "learning_rate": 2.2931771857250333e-05, "loss": 0.0136, "step": 55840 }, { "grad_norm": 0.10932203382253647, "learning_rate": 2.2914396451216946e-05, "loss": 0.01, "step": 55850 }, { "grad_norm": 0.1875612884759903, "learning_rate": 2.2897025673372412e-05, "loss": 0.0077, "step": 55860 }, { "grad_norm": 0.158952534198761, "learning_rate": 2.287965952668494e-05, "loss": 0.0088, "step": 55870 }, { "grad_norm": 0.20746268332004547, "learning_rate": 2.2862298014121873e-05, "loss": 0.0104, "step": 55880 }, { "grad_norm": 0.20168159902095795, "learning_rate": 2.28449411386499e-05, "loss": 0.0091, "step": 55890 }, { "grad_norm": 0.27400314807891846, "learning_rate": 2.282758890323477e-05, "loss": 0.0095, "step": 55900 }, { "grad_norm": 0.10780943930149078, "learning_rate": 2.2810241310841528e-05, "loss": 0.0111, "step": 55910 }, { "grad_norm": 0.12112802267074585, "learning_rate": 2.2792898364434412e-05, "loss": 0.0083, "step": 55920 }, { "grad_norm": 0.14119383692741394, "learning_rate": 2.2775560066976843e-05, "loss": 0.0095, "step": 55930 }, { "grad_norm": 0.11201678216457367, "learning_rate": 2.275822642143147e-05, "loss": 0.0109, "step": 55940 }, { "grad_norm": 0.1413993239402771, "learning_rate": 2.274089743076014e-05, "loss": 0.0094, "step": 55950 }, { "grad_norm": 0.18133501708507538, "learning_rate": 2.2723573097923917e-05, "loss": 0.0098, "step": 55960 }, { "grad_norm": 0.14566411077976227, "learning_rate": 2.2706253425883024e-05, "loss": 0.0065, "step": 55970 }, { "grad_norm": 0.20171484351158142, "learning_rate": 2.2688938417596933e-05, "loss": 0.0134, "step": 55980 }, { "grad_norm": 0.10364632308483124, "learning_rate": 2.2671628076024305e-05, "loss": 0.0108, "step": 55990 }, { "grad_norm": 0.26510700583457947, "learning_rate": 2.2654322404123008e-05, "loss": 0.0073, "step": 56000 }, { "grad_norm": 0.16937442123889923, "learning_rate": 2.2637021404850105e-05, "loss": 0.0093, "step": 56010 }, { "grad_norm": 0.2177325040102005, "learning_rate": 2.2619725081161885e-05, "loss": 0.0145, "step": 56020 }, { "grad_norm": 0.1544448733329773, "learning_rate": 2.260243343601376e-05, "loss": 0.0128, "step": 56030 }, { "grad_norm": 0.2341112196445465, "learning_rate": 2.2585146472360473e-05, "loss": 0.0121, "step": 56040 }, { "grad_norm": 0.2697688043117523, "learning_rate": 2.2567864193155834e-05, "loss": 0.0101, "step": 56050 }, { "grad_norm": 0.19420252740383148, "learning_rate": 2.255058660135294e-05, "loss": 0.0125, "step": 56060 }, { "grad_norm": 0.44656795263290405, "learning_rate": 2.253331369990407e-05, "loss": 0.0142, "step": 56070 }, { "grad_norm": 0.20879815518856049, "learning_rate": 2.251604549176063e-05, "loss": 0.0136, "step": 56080 }, { "grad_norm": 0.15669132769107819, "learning_rate": 2.249878197987337e-05, "loss": 0.01, "step": 56090 }, { "grad_norm": 0.20674146711826324, "learning_rate": 2.2481523167192087e-05, "loss": 0.0094, "step": 56100 }, { "grad_norm": 0.1327495276927948, "learning_rate": 2.2464269056665867e-05, "loss": 0.0084, "step": 56110 }, { "grad_norm": 0.18349173665046692, "learning_rate": 2.2447019651242958e-05, "loss": 0.0131, "step": 56120 }, { "grad_norm": 0.15936705470085144, "learning_rate": 2.2429774953870818e-05, "loss": 0.0073, "step": 56130 }, { "grad_norm": 0.37210801243782043, "learning_rate": 2.241253496749611e-05, "loss": 0.0122, "step": 56140 }, { "grad_norm": 0.1315714567899704, "learning_rate": 2.2395299695064614e-05, "loss": 0.0081, "step": 56150 }, { "grad_norm": 0.2064093053340912, "learning_rate": 2.237806913952145e-05, "loss": 0.0115, "step": 56160 }, { "grad_norm": 0.1533295214176178, "learning_rate": 2.2360843303810798e-05, "loss": 0.0092, "step": 56170 }, { "grad_norm": 0.2535569369792938, "learning_rate": 2.2343622190876084e-05, "loss": 0.0077, "step": 56180 }, { "grad_norm": 0.14924049377441406, "learning_rate": 2.2326405803659935e-05, "loss": 0.0089, "step": 56190 }, { "grad_norm": 0.20877420902252197, "learning_rate": 2.230919414510416e-05, "loss": 0.0094, "step": 56200 }, { "grad_norm": 0.3077507019042969, "learning_rate": 2.229198721814976e-05, "loss": 0.0111, "step": 56210 }, { "grad_norm": 0.1585344672203064, "learning_rate": 2.2274785025736948e-05, "loss": 0.0085, "step": 56220 }, { "grad_norm": 0.11473671346902847, "learning_rate": 2.225758757080507e-05, "loss": 0.0112, "step": 56230 }, { "grad_norm": 0.2649933993816376, "learning_rate": 2.2240394856292723e-05, "loss": 0.011, "step": 56240 }, { "grad_norm": 0.151372492313385, "learning_rate": 2.2223206885137664e-05, "loss": 0.0086, "step": 56250 }, { "grad_norm": 0.257610946893692, "learning_rate": 2.2206023660276853e-05, "loss": 0.0112, "step": 56260 }, { "grad_norm": 0.2599722743034363, "learning_rate": 2.218884518464645e-05, "loss": 0.014, "step": 56270 }, { "grad_norm": 0.28507497906684875, "learning_rate": 2.2171671461181732e-05, "loss": 0.0117, "step": 56280 }, { "grad_norm": 0.18439547717571259, "learning_rate": 2.2154502492817292e-05, "loss": 0.0134, "step": 56290 }, { "grad_norm": 0.1121886819601059, "learning_rate": 2.2137338282486782e-05, "loss": 0.0082, "step": 56300 }, { "grad_norm": 0.2943460941314697, "learning_rate": 2.2120178833123113e-05, "loss": 0.0105, "step": 56310 }, { "grad_norm": 0.15483854711055756, "learning_rate": 2.210302414765838e-05, "loss": 0.0096, "step": 56320 }, { "grad_norm": 0.212563619017601, "learning_rate": 2.2085874229023808e-05, "loss": 0.0094, "step": 56330 }, { "grad_norm": 0.24553132057189941, "learning_rate": 2.2068729080149907e-05, "loss": 0.0085, "step": 56340 }, { "grad_norm": 0.2508915066719055, "learning_rate": 2.205158870396625e-05, "loss": 0.0131, "step": 56350 }, { "grad_norm": 0.3141486644744873, "learning_rate": 2.2034453103401732e-05, "loss": 0.0091, "step": 56360 }, { "grad_norm": 0.23531818389892578, "learning_rate": 2.20173222813843e-05, "loss": 0.0064, "step": 56370 }, { "grad_norm": 0.21309733390808105, "learning_rate": 2.200019624084116e-05, "loss": 0.0122, "step": 56380 }, { "grad_norm": 0.16382984817028046, "learning_rate": 2.1983074984698687e-05, "loss": 0.0117, "step": 56390 }, { "grad_norm": 0.3002219498157501, "learning_rate": 2.1965958515882433e-05, "loss": 0.0164, "step": 56400 }, { "grad_norm": 0.11059944331645966, "learning_rate": 2.1948846837317162e-05, "loss": 0.0083, "step": 56410 }, { "grad_norm": 0.2958764135837555, "learning_rate": 2.1931739951926738e-05, "loss": 0.0065, "step": 56420 }, { "grad_norm": 0.15177032351493835, "learning_rate": 2.191463786263429e-05, "loss": 0.007, "step": 56430 }, { "grad_norm": 0.17344917356967926, "learning_rate": 2.1897540572362095e-05, "loss": 0.0116, "step": 56440 }, { "grad_norm": 0.12317054718732834, "learning_rate": 2.1880448084031614e-05, "loss": 0.0084, "step": 56450 }, { "grad_norm": 0.15001024305820465, "learning_rate": 2.1863360400563482e-05, "loss": 0.0116, "step": 56460 }, { "grad_norm": 0.18410691618919373, "learning_rate": 2.184627752487754e-05, "loss": 0.0099, "step": 56470 }, { "grad_norm": 0.10291719436645508, "learning_rate": 2.1829199459892725e-05, "loss": 0.0089, "step": 56480 }, { "grad_norm": 0.17621076107025146, "learning_rate": 2.1812126208527282e-05, "loss": 0.0115, "step": 56490 }, { "grad_norm": 0.1646353155374527, "learning_rate": 2.179505777369852e-05, "loss": 0.0155, "step": 56500 }, { "grad_norm": 0.17140376567840576, "learning_rate": 2.1777994158322974e-05, "loss": 0.0085, "step": 56510 }, { "grad_norm": 0.11149661242961884, "learning_rate": 2.176093536531637e-05, "loss": 0.0143, "step": 56520 }, { "grad_norm": 0.09258218109607697, "learning_rate": 2.1743881397593537e-05, "loss": 0.009, "step": 56530 }, { "grad_norm": 0.3667442202568054, "learning_rate": 2.1726832258068595e-05, "loss": 0.0248, "step": 56540 }, { "grad_norm": 0.31542298197746277, "learning_rate": 2.170978794965472e-05, "loss": 0.0118, "step": 56550 }, { "grad_norm": 0.11712423712015152, "learning_rate": 2.169274847526438e-05, "loss": 0.0102, "step": 56560 }, { "grad_norm": 0.1788201928138733, "learning_rate": 2.1675713837809103e-05, "loss": 0.0104, "step": 56570 }, { "grad_norm": 0.17612062394618988, "learning_rate": 2.1658684040199655e-05, "loss": 0.0102, "step": 56580 }, { "grad_norm": 0.24617637693881989, "learning_rate": 2.1641659085345974e-05, "loss": 0.0074, "step": 56590 }, { "grad_norm": 0.18315130472183228, "learning_rate": 2.1624638976157154e-05, "loss": 0.014, "step": 56600 }, { "grad_norm": 0.24300257861614227, "learning_rate": 2.1607623715541476e-05, "loss": 0.0162, "step": 56610 }, { "grad_norm": 0.13200615346431732, "learning_rate": 2.159061330640636e-05, "loss": 0.0081, "step": 56620 }, { "grad_norm": 0.14483749866485596, "learning_rate": 2.1573607751658425e-05, "loss": 0.0065, "step": 56630 }, { "grad_norm": 0.14233922958374023, "learning_rate": 2.155660705420347e-05, "loss": 0.0149, "step": 56640 }, { "grad_norm": 0.29310309886932373, "learning_rate": 2.153961121694644e-05, "loss": 0.0087, "step": 56650 }, { "grad_norm": 0.15767447650432587, "learning_rate": 2.152262024279145e-05, "loss": 0.0219, "step": 56660 }, { "grad_norm": 0.4187067449092865, "learning_rate": 2.150563413464183e-05, "loss": 0.0146, "step": 56670 }, { "grad_norm": 0.21002179384231567, "learning_rate": 2.1488652895399974e-05, "loss": 0.0088, "step": 56680 }, { "grad_norm": 0.26257869601249695, "learning_rate": 2.1471676527967587e-05, "loss": 0.0106, "step": 56690 }, { "grad_norm": 0.1495046615600586, "learning_rate": 2.1454705035245414e-05, "loss": 0.0137, "step": 56700 }, { "grad_norm": 0.16018353402614594, "learning_rate": 2.143773842013343e-05, "loss": 0.0069, "step": 56710 }, { "grad_norm": 0.27316299080848694, "learning_rate": 2.1420776685530796e-05, "loss": 0.0115, "step": 56720 }, { "grad_norm": 0.35522204637527466, "learning_rate": 2.1403819834335742e-05, "loss": 0.0097, "step": 56730 }, { "grad_norm": 0.13379190862178802, "learning_rate": 2.1386867869445808e-05, "loss": 0.0112, "step": 56740 }, { "grad_norm": 0.12058977037668228, "learning_rate": 2.1369920793757548e-05, "loss": 0.0129, "step": 56750 }, { "grad_norm": 0.26327359676361084, "learning_rate": 2.1352978610166828e-05, "loss": 0.008, "step": 56760 }, { "grad_norm": 0.17687271535396576, "learning_rate": 2.1336041321568546e-05, "loss": 0.01, "step": 56770 }, { "grad_norm": 0.19265325367450714, "learning_rate": 2.131910893085684e-05, "loss": 0.0083, "step": 56780 }, { "grad_norm": 0.17564868927001953, "learning_rate": 2.1302181440925012e-05, "loss": 0.0154, "step": 56790 }, { "grad_norm": 0.11114185303449631, "learning_rate": 2.128525885466546e-05, "loss": 0.0088, "step": 56800 }, { "grad_norm": 0.17929475009441376, "learning_rate": 2.1268341174969847e-05, "loss": 0.0139, "step": 56810 }, { "grad_norm": 0.14256735146045685, "learning_rate": 2.1251428404728894e-05, "loss": 0.0081, "step": 56820 }, { "grad_norm": 0.18214170634746552, "learning_rate": 2.123452054683256e-05, "loss": 0.0085, "step": 56830 }, { "grad_norm": 0.2495773434638977, "learning_rate": 2.1217617604169927e-05, "loss": 0.013, "step": 56840 }, { "grad_norm": 0.23694941401481628, "learning_rate": 2.120071957962924e-05, "loss": 0.011, "step": 56850 }, { "grad_norm": 0.19997721910476685, "learning_rate": 2.1183826476097917e-05, "loss": 0.0179, "step": 56860 }, { "grad_norm": 0.13754285871982574, "learning_rate": 2.116693829646254e-05, "loss": 0.0098, "step": 56870 }, { "grad_norm": 0.28083324432373047, "learning_rate": 2.1150055043608806e-05, "loss": 0.0092, "step": 56880 }, { "grad_norm": 0.31760674715042114, "learning_rate": 2.113317672042162e-05, "loss": 0.0141, "step": 56890 }, { "grad_norm": 0.11928020417690277, "learning_rate": 2.1116303329785025e-05, "loss": 0.0072, "step": 56900 }, { "grad_norm": 0.15209104120731354, "learning_rate": 2.109943487458222e-05, "loss": 0.0136, "step": 56910 }, { "grad_norm": 0.1919427365064621, "learning_rate": 2.1082571357695574e-05, "loss": 0.0082, "step": 56920 }, { "grad_norm": 0.20404598116874695, "learning_rate": 2.1065712782006557e-05, "loss": 0.012, "step": 56930 }, { "grad_norm": 0.13132251799106598, "learning_rate": 2.104885915039591e-05, "loss": 0.0092, "step": 56940 }, { "grad_norm": 0.2066221833229065, "learning_rate": 2.103201046574338e-05, "loss": 0.0077, "step": 56950 }, { "grad_norm": 0.23401817679405212, "learning_rate": 2.1015166730928022e-05, "loss": 0.011, "step": 56960 }, { "grad_norm": 0.20562253892421722, "learning_rate": 2.0998327948827912e-05, "loss": 0.0117, "step": 56970 }, { "grad_norm": 0.16460192203521729, "learning_rate": 2.0981494122320363e-05, "loss": 0.0095, "step": 56980 }, { "grad_norm": 0.21823428571224213, "learning_rate": 2.0964665254281822e-05, "loss": 0.0101, "step": 56990 }, { "grad_norm": 0.19088375568389893, "learning_rate": 2.094784134758784e-05, "loss": 0.009, "step": 57000 }, { "grad_norm": 0.1484912633895874, "learning_rate": 2.0931022405113226e-05, "loss": 0.0181, "step": 57010 }, { "grad_norm": 0.28212371468544006, "learning_rate": 2.091420842973183e-05, "loss": 0.0146, "step": 57020 }, { "grad_norm": 0.2155705839395523, "learning_rate": 2.0897399424316715e-05, "loss": 0.0071, "step": 57030 }, { "grad_norm": 0.2753356993198395, "learning_rate": 2.0880595391740078e-05, "loss": 0.0115, "step": 57040 }, { "grad_norm": 0.09183046966791153, "learning_rate": 2.0863796334873277e-05, "loss": 0.0147, "step": 57050 }, { "grad_norm": 0.15459372103214264, "learning_rate": 2.08470022565868e-05, "loss": 0.0078, "step": 57060 }, { "grad_norm": 0.12231855094432831, "learning_rate": 2.0830213159750317e-05, "loss": 0.0059, "step": 57070 }, { "grad_norm": 0.17237317562103271, "learning_rate": 2.0813429047232596e-05, "loss": 0.008, "step": 57080 }, { "grad_norm": 0.12605923414230347, "learning_rate": 2.0796649921901594e-05, "loss": 0.0101, "step": 57090 }, { "grad_norm": 0.1304553747177124, "learning_rate": 2.077987578662441e-05, "loss": 0.0088, "step": 57100 }, { "grad_norm": 0.3671083450317383, "learning_rate": 2.0763106644267277e-05, "loss": 0.0149, "step": 57110 }, { "grad_norm": 0.16431289911270142, "learning_rate": 2.0746342497695607e-05, "loss": 0.007, "step": 57120 }, { "grad_norm": 0.1188405454158783, "learning_rate": 2.0729583349773886e-05, "loss": 0.0054, "step": 57130 }, { "grad_norm": 0.35370397567749023, "learning_rate": 2.0712829203365853e-05, "loss": 0.0087, "step": 57140 }, { "grad_norm": 0.11689762026071548, "learning_rate": 2.0696080061334267e-05, "loss": 0.0097, "step": 57150 }, { "grad_norm": 0.10795674473047256, "learning_rate": 2.067933592654117e-05, "loss": 0.0071, "step": 57160 }, { "grad_norm": 0.16443489491939545, "learning_rate": 2.066259680184763e-05, "loss": 0.0078, "step": 57170 }, { "grad_norm": 0.16495352983474731, "learning_rate": 2.0645862690113908e-05, "loss": 0.0122, "step": 57180 }, { "grad_norm": 0.19416558742523193, "learning_rate": 2.0629133594199436e-05, "loss": 0.0158, "step": 57190 }, { "grad_norm": 0.1675838828086853, "learning_rate": 2.0612409516962704e-05, "loss": 0.0115, "step": 57200 }, { "grad_norm": 0.15909233689308167, "learning_rate": 2.0595690461261467e-05, "loss": 0.0091, "step": 57210 }, { "grad_norm": 0.1846071481704712, "learning_rate": 2.0578976429952503e-05, "loss": 0.0073, "step": 57220 }, { "grad_norm": 0.3083379566669464, "learning_rate": 2.0562267425891802e-05, "loss": 0.012, "step": 57230 }, { "grad_norm": 0.19262436032295227, "learning_rate": 2.0545563451934467e-05, "loss": 0.0093, "step": 57240 }, { "grad_norm": 0.1300806701183319, "learning_rate": 2.0528864510934764e-05, "loss": 0.0155, "step": 57250 }, { "grad_norm": 0.1524561643600464, "learning_rate": 2.0512170605746096e-05, "loss": 0.0177, "step": 57260 }, { "grad_norm": 0.13223162293434143, "learning_rate": 2.0495481739220963e-05, "loss": 0.0112, "step": 57270 }, { "grad_norm": 0.168637216091156, "learning_rate": 2.0478797914211045e-05, "loss": 0.012, "step": 57280 }, { "grad_norm": 0.40906915068626404, "learning_rate": 2.046211913356716e-05, "loss": 0.0113, "step": 57290 }, { "grad_norm": 0.129729226231575, "learning_rate": 2.0445445400139247e-05, "loss": 0.0152, "step": 57300 }, { "grad_norm": 0.3019374907016754, "learning_rate": 2.0428776716776405e-05, "loss": 0.0103, "step": 57310 }, { "grad_norm": 0.10509396344423294, "learning_rate": 2.0412113086326856e-05, "loss": 0.01, "step": 57320 }, { "grad_norm": 0.11886226385831833, "learning_rate": 2.0395454511637918e-05, "loss": 0.0079, "step": 57330 }, { "grad_norm": 0.17863228917121887, "learning_rate": 2.037880099555616e-05, "loss": 0.0095, "step": 57340 }, { "grad_norm": 0.12240158766508102, "learning_rate": 2.0362152540927144e-05, "loss": 0.0083, "step": 57350 }, { "grad_norm": 0.2132425606250763, "learning_rate": 2.0345509150595666e-05, "loss": 0.0219, "step": 57360 }, { "grad_norm": 0.16770808398723602, "learning_rate": 2.0328870827405617e-05, "loss": 0.0113, "step": 57370 }, { "grad_norm": 0.35030364990234375, "learning_rate": 2.0312237574200043e-05, "loss": 0.0086, "step": 57380 }, { "grad_norm": 0.20285740494728088, "learning_rate": 2.029560939382112e-05, "loss": 0.0089, "step": 57390 }, { "grad_norm": 0.11266366392374039, "learning_rate": 2.0278986289110097e-05, "loss": 0.0101, "step": 57400 }, { "grad_norm": 0.13214686512947083, "learning_rate": 2.0262368262907484e-05, "loss": 0.0092, "step": 57410 }, { "grad_norm": 0.2942379415035248, "learning_rate": 2.024575531805279e-05, "loss": 0.014, "step": 57420 }, { "grad_norm": 0.15991804003715515, "learning_rate": 2.0229147457384735e-05, "loss": 0.0063, "step": 57430 }, { "grad_norm": 0.15262968838214874, "learning_rate": 2.0212544683741157e-05, "loss": 0.0217, "step": 57440 }, { "grad_norm": 0.15139277279376984, "learning_rate": 2.0195946999958976e-05, "loss": 0.0149, "step": 57450 }, { "grad_norm": 0.13249897956848145, "learning_rate": 2.017935440887434e-05, "loss": 0.0126, "step": 57460 }, { "grad_norm": 0.09777332842350006, "learning_rate": 2.0162766913322423e-05, "loss": 0.0091, "step": 57470 }, { "grad_norm": 0.1753745973110199, "learning_rate": 2.0146184516137588e-05, "loss": 0.012, "step": 57480 }, { "grad_norm": 0.1314801275730133, "learning_rate": 2.012960722015332e-05, "loss": 0.012, "step": 57490 }, { "grad_norm": 0.17689360678195953, "learning_rate": 2.0113035028202214e-05, "loss": 0.0133, "step": 57500 }, { "grad_norm": 0.17512206733226776, "learning_rate": 2.009646794311602e-05, "loss": 0.0162, "step": 57510 }, { "grad_norm": 0.13515983521938324, "learning_rate": 2.007990596772559e-05, "loss": 0.0126, "step": 57520 }, { "grad_norm": 0.10562848299741745, "learning_rate": 2.0063349104860923e-05, "loss": 0.0112, "step": 57530 }, { "grad_norm": 0.14922139048576355, "learning_rate": 2.0046797357351116e-05, "loss": 0.019, "step": 57540 }, { "grad_norm": 0.19253458082675934, "learning_rate": 2.0030250728024412e-05, "loss": 0.0073, "step": 57550 }, { "grad_norm": 0.14922727644443512, "learning_rate": 2.001370921970819e-05, "loss": 0.0071, "step": 57560 }, { "grad_norm": 0.20963013172149658, "learning_rate": 1.9997172835228932e-05, "loss": 0.0077, "step": 57570 }, { "grad_norm": 0.23758134245872498, "learning_rate": 1.9980641577412262e-05, "loss": 0.008, "step": 57580 }, { "grad_norm": 0.18672223389148712, "learning_rate": 1.9964115449082925e-05, "loss": 0.0112, "step": 57590 }, { "grad_norm": 0.16210587322711945, "learning_rate": 1.9947594453064742e-05, "loss": 0.0141, "step": 57600 }, { "grad_norm": 0.14526721835136414, "learning_rate": 1.9931078592180774e-05, "loss": 0.0124, "step": 57610 }, { "grad_norm": 0.2524830996990204, "learning_rate": 1.9914567869253065e-05, "loss": 0.0107, "step": 57620 }, { "grad_norm": 0.13484010100364685, "learning_rate": 1.989806228710287e-05, "loss": 0.0183, "step": 57630 }, { "grad_norm": 0.1521233171224594, "learning_rate": 1.9881561848550555e-05, "loss": 0.0094, "step": 57640 }, { "grad_norm": 0.22590771317481995, "learning_rate": 1.9865066556415544e-05, "loss": 0.0136, "step": 57650 }, { "grad_norm": 0.1469297558069229, "learning_rate": 1.98485764135165e-05, "loss": 0.0079, "step": 57660 }, { "grad_norm": 0.1027943417429924, "learning_rate": 1.983209142267109e-05, "loss": 0.0093, "step": 57670 }, { "grad_norm": 0.15604117512702942, "learning_rate": 1.9815611586696165e-05, "loss": 0.009, "step": 57680 }, { "grad_norm": 0.10327167063951492, "learning_rate": 1.9799136908407667e-05, "loss": 0.0081, "step": 57690 }, { "grad_norm": 0.15636377036571503, "learning_rate": 1.9782667390620678e-05, "loss": 0.0086, "step": 57700 }, { "grad_norm": 0.20310109853744507, "learning_rate": 1.976620303614939e-05, "loss": 0.0067, "step": 57710 }, { "grad_norm": 0.112788125872612, "learning_rate": 1.9749743847807108e-05, "loss": 0.0081, "step": 57720 }, { "grad_norm": 0.19461452960968018, "learning_rate": 1.9733289828406272e-05, "loss": 0.0088, "step": 57730 }, { "grad_norm": 0.2589839994907379, "learning_rate": 1.9716840980758382e-05, "loss": 0.0113, "step": 57740 }, { "grad_norm": 0.16076897084712982, "learning_rate": 1.9700397307674134e-05, "loss": 0.0077, "step": 57750 }, { "grad_norm": 0.2243102341890335, "learning_rate": 1.968395881196328e-05, "loss": 0.0093, "step": 57760 }, { "grad_norm": 0.33900320529937744, "learning_rate": 1.966752549643473e-05, "loss": 0.0134, "step": 57770 }, { "grad_norm": 0.1427975445985794, "learning_rate": 1.965109736389647e-05, "loss": 0.0084, "step": 57780 }, { "grad_norm": 0.21060344576835632, "learning_rate": 1.9634674417155645e-05, "loss": 0.0077, "step": 57790 }, { "grad_norm": 0.14364400506019592, "learning_rate": 1.9618256659018434e-05, "loss": 0.032, "step": 57800 }, { "grad_norm": 0.24576538801193237, "learning_rate": 1.9601844092290257e-05, "loss": 0.0108, "step": 57810 }, { "grad_norm": 0.22873303294181824, "learning_rate": 1.9585436719775512e-05, "loss": 0.0089, "step": 57820 }, { "grad_norm": 0.20873399078845978, "learning_rate": 1.9569034544277793e-05, "loss": 0.0119, "step": 57830 }, { "grad_norm": 0.14133620262145996, "learning_rate": 1.9552637568599798e-05, "loss": 0.0103, "step": 57840 }, { "grad_norm": 0.2527254819869995, "learning_rate": 1.953624579554327e-05, "loss": 0.0162, "step": 57850 }, { "grad_norm": 0.13156819343566895, "learning_rate": 1.951985922790918e-05, "loss": 0.0169, "step": 57860 }, { "grad_norm": 0.13922731578350067, "learning_rate": 1.9503477868497505e-05, "loss": 0.008, "step": 57870 }, { "grad_norm": 0.2150232046842575, "learning_rate": 1.9487101720107375e-05, "loss": 0.0131, "step": 57880 }, { "grad_norm": 0.13014212250709534, "learning_rate": 1.9470730785537032e-05, "loss": 0.0118, "step": 57890 }, { "grad_norm": 0.13628675043582916, "learning_rate": 1.9454365067583823e-05, "loss": 0.0085, "step": 57900 }, { "grad_norm": 0.161021426320076, "learning_rate": 1.9438004569044215e-05, "loss": 0.0082, "step": 57910 }, { "grad_norm": 0.17374053597450256, "learning_rate": 1.9421649292713724e-05, "loss": 0.0253, "step": 57920 }, { "grad_norm": 0.24141515791416168, "learning_rate": 1.9405299241387076e-05, "loss": 0.0139, "step": 57930 }, { "grad_norm": 0.19226041436195374, "learning_rate": 1.9388954417858007e-05, "loss": 0.0099, "step": 57940 }, { "grad_norm": 0.17439930140972137, "learning_rate": 1.9372614824919417e-05, "loss": 0.0092, "step": 57950 }, { "grad_norm": 0.2739788293838501, "learning_rate": 1.9356280465363284e-05, "loss": 0.0131, "step": 57960 }, { "grad_norm": 0.2348632663488388, "learning_rate": 1.9339951341980723e-05, "loss": 0.008, "step": 57970 }, { "grad_norm": 0.18152041733264923, "learning_rate": 1.9323627457561916e-05, "loss": 0.01, "step": 57980 }, { "grad_norm": 0.17193599045276642, "learning_rate": 1.9307308814896198e-05, "loss": 0.008, "step": 57990 }, { "grad_norm": 0.09002909064292908, "learning_rate": 1.9290995416771935e-05, "loss": 0.0094, "step": 58000 }, { "grad_norm": 0.17638009786605835, "learning_rate": 1.9274687265976665e-05, "loss": 0.0099, "step": 58010 }, { "grad_norm": 0.18965595960617065, "learning_rate": 1.9258384365297e-05, "loss": 0.0113, "step": 58020 }, { "grad_norm": 0.16854730248451233, "learning_rate": 1.924208671751866e-05, "loss": 0.0079, "step": 58030 }, { "grad_norm": 0.1260114312171936, "learning_rate": 1.9225794325426492e-05, "loss": 0.0087, "step": 58040 }, { "grad_norm": 0.17038802802562714, "learning_rate": 1.920950719180436e-05, "loss": 0.0078, "step": 58050 }, { "grad_norm": 0.12818460166454315, "learning_rate": 1.919322531943536e-05, "loss": 0.0075, "step": 58060 }, { "grad_norm": 0.12784472107887268, "learning_rate": 1.917694871110157e-05, "loss": 0.0097, "step": 58070 }, { "grad_norm": 0.3096688389778137, "learning_rate": 1.9160677369584234e-05, "loss": 0.0094, "step": 58080 }, { "grad_norm": 0.13857638835906982, "learning_rate": 1.9144411297663694e-05, "loss": 0.0114, "step": 58090 }, { "grad_norm": 0.10134487599134445, "learning_rate": 1.9128150498119328e-05, "loss": 0.0148, "step": 58100 }, { "grad_norm": 0.152361199259758, "learning_rate": 1.9111894973729726e-05, "loss": 0.0071, "step": 58110 }, { "grad_norm": 0.3055894076824188, "learning_rate": 1.9095644727272454e-05, "loss": 0.0133, "step": 58120 }, { "grad_norm": 0.2513846755027771, "learning_rate": 1.907939976152429e-05, "loss": 0.0104, "step": 58130 }, { "grad_norm": 0.22970744967460632, "learning_rate": 1.906316007926101e-05, "loss": 0.0162, "step": 58140 }, { "grad_norm": 0.2676549255847931, "learning_rate": 1.904692568325755e-05, "loss": 0.0094, "step": 58150 }, { "grad_norm": 0.2708713412284851, "learning_rate": 1.9030696576287925e-05, "loss": 0.0119, "step": 58160 }, { "grad_norm": 0.11127372086048126, "learning_rate": 1.9014472761125242e-05, "loss": 0.0067, "step": 58170 }, { "grad_norm": 0.15768522024154663, "learning_rate": 1.899825424054172e-05, "loss": 0.0095, "step": 58180 }, { "grad_norm": 0.17197461426258087, "learning_rate": 1.898204101730863e-05, "loss": 0.0093, "step": 58190 }, { "grad_norm": 0.14708735048770905, "learning_rate": 1.8965833094196394e-05, "loss": 0.022, "step": 58200 }, { "grad_norm": 0.1514582633972168, "learning_rate": 1.8949630473974495e-05, "loss": 0.0089, "step": 58210 }, { "grad_norm": 0.09523973613977432, "learning_rate": 1.8933433159411517e-05, "loss": 0.0094, "step": 58220 }, { "grad_norm": 0.09745155274868011, "learning_rate": 1.891724115327514e-05, "loss": 0.0081, "step": 58230 }, { "grad_norm": 0.11999903619289398, "learning_rate": 1.8901054458332152e-05, "loss": 0.0111, "step": 58240 }, { "grad_norm": 0.11807417869567871, "learning_rate": 1.8884873077348364e-05, "loss": 0.0116, "step": 58250 }, { "grad_norm": 0.11996873468160629, "learning_rate": 1.88686970130888e-05, "loss": 0.0117, "step": 58260 }, { "grad_norm": 0.13161718845367432, "learning_rate": 1.8852526268317456e-05, "loss": 0.0113, "step": 58270 }, { "grad_norm": 0.15765240788459778, "learning_rate": 1.883636084579749e-05, "loss": 0.0075, "step": 58280 }, { "grad_norm": 0.29329606890678406, "learning_rate": 1.8820200748291145e-05, "loss": 0.0121, "step": 58290 }, { "grad_norm": 0.16327090561389923, "learning_rate": 1.8804045978559686e-05, "loss": 0.0093, "step": 58300 }, { "grad_norm": 0.25247955322265625, "learning_rate": 1.8787896539363594e-05, "loss": 0.0108, "step": 58310 }, { "grad_norm": 0.22511929273605347, "learning_rate": 1.8771752433462298e-05, "loss": 0.0082, "step": 58320 }, { "grad_norm": 0.17038898169994354, "learning_rate": 1.8755613663614447e-05, "loss": 0.011, "step": 58330 }, { "grad_norm": 0.3080219030380249, "learning_rate": 1.873948023257767e-05, "loss": 0.0087, "step": 58340 }, { "grad_norm": 0.10146388411521912, "learning_rate": 1.8723352143108747e-05, "loss": 0.0082, "step": 58350 }, { "grad_norm": 0.1649845689535141, "learning_rate": 1.870722939796352e-05, "loss": 0.0102, "step": 58360 }, { "grad_norm": 0.19119714200496674, "learning_rate": 1.869111199989693e-05, "loss": 0.0089, "step": 58370 }, { "grad_norm": 0.12919296324253082, "learning_rate": 1.8674999951663018e-05, "loss": 0.009, "step": 58380 }, { "grad_norm": 0.15746645629405975, "learning_rate": 1.8658893256014857e-05, "loss": 0.0126, "step": 58390 }, { "grad_norm": 0.18902148306369781, "learning_rate": 1.8642791915704655e-05, "loss": 0.0076, "step": 58400 }, { "grad_norm": 0.12819239497184753, "learning_rate": 1.8626695933483697e-05, "loss": 0.0069, "step": 58410 }, { "grad_norm": 0.1506287306547165, "learning_rate": 1.8610605312102335e-05, "loss": 0.0091, "step": 58420 }, { "grad_norm": 0.31070590019226074, "learning_rate": 1.8594520054310032e-05, "loss": 0.0161, "step": 58430 }, { "grad_norm": 0.20261219143867493, "learning_rate": 1.8578440162855326e-05, "loss": 0.0082, "step": 58440 }, { "grad_norm": 0.1357872635126114, "learning_rate": 1.8562365640485774e-05, "loss": 0.0102, "step": 58450 }, { "grad_norm": 0.09782223403453827, "learning_rate": 1.854629648994815e-05, "loss": 0.0058, "step": 58460 }, { "grad_norm": 0.13437007367610931, "learning_rate": 1.8530232713988183e-05, "loss": 0.0107, "step": 58470 }, { "grad_norm": 0.27211451530456543, "learning_rate": 1.851417431535074e-05, "loss": 0.0111, "step": 58480 }, { "grad_norm": 0.3352937698364258, "learning_rate": 1.8498121296779785e-05, "loss": 0.0204, "step": 58490 }, { "grad_norm": 0.14519409835338593, "learning_rate": 1.848207366101829e-05, "loss": 0.0133, "step": 58500 }, { "grad_norm": 0.11903634667396545, "learning_rate": 1.8466031410808422e-05, "loss": 0.0085, "step": 58510 }, { "grad_norm": 0.2780333161354065, "learning_rate": 1.844999454889129e-05, "loss": 0.0121, "step": 58520 }, { "grad_norm": 0.2438940703868866, "learning_rate": 1.8433963078007237e-05, "loss": 0.0115, "step": 58530 }, { "grad_norm": 0.11862664669752121, "learning_rate": 1.841793700089554e-05, "loss": 0.0092, "step": 58540 }, { "grad_norm": 0.17162661254405975, "learning_rate": 1.8401916320294644e-05, "loss": 0.0104, "step": 58550 }, { "grad_norm": 0.21243764460086823, "learning_rate": 1.838590103894205e-05, "loss": 0.0083, "step": 58560 }, { "grad_norm": 0.23874704539775848, "learning_rate": 1.8369891159574292e-05, "loss": 0.0084, "step": 58570 }, { "grad_norm": 0.2114422619342804, "learning_rate": 1.835388668492708e-05, "loss": 0.012, "step": 58580 }, { "grad_norm": 0.13024456799030304, "learning_rate": 1.8337887617735095e-05, "loss": 0.0082, "step": 58590 }, { "grad_norm": 0.2023780643939972, "learning_rate": 1.8321893960732157e-05, "loss": 0.0093, "step": 58600 }, { "grad_norm": 0.17110660672187805, "learning_rate": 1.8305905716651138e-05, "loss": 0.0078, "step": 58610 }, { "grad_norm": 0.15757596492767334, "learning_rate": 1.8289922888224e-05, "loss": 0.0093, "step": 58620 }, { "grad_norm": 0.21530325710773468, "learning_rate": 1.8273945478181765e-05, "loss": 0.0107, "step": 58630 }, { "grad_norm": 0.1645166426897049, "learning_rate": 1.8257973489254558e-05, "loss": 0.0079, "step": 58640 }, { "grad_norm": 0.09106626361608505, "learning_rate": 1.824200692417152e-05, "loss": 0.0099, "step": 58650 }, { "grad_norm": 0.11226974427700043, "learning_rate": 1.8226045785660912e-05, "loss": 0.0069, "step": 58660 }, { "grad_norm": 0.18294134736061096, "learning_rate": 1.821009007645006e-05, "loss": 0.0081, "step": 58670 }, { "grad_norm": 0.25447526574134827, "learning_rate": 1.8194139799265357e-05, "loss": 0.0081, "step": 58680 }, { "grad_norm": 0.14454099535942078, "learning_rate": 1.8178194956832295e-05, "loss": 0.0102, "step": 58690 }, { "grad_norm": 0.13246764242649078, "learning_rate": 1.8162255551875346e-05, "loss": 0.0077, "step": 58700 }, { "grad_norm": 0.16171526908874512, "learning_rate": 1.8146321587118194e-05, "loss": 0.0159, "step": 58710 }, { "grad_norm": 0.1884184181690216, "learning_rate": 1.8130393065283448e-05, "loss": 0.0076, "step": 58720 }, { "grad_norm": 0.19862006604671478, "learning_rate": 1.8114469989092925e-05, "loss": 0.007, "step": 58730 }, { "grad_norm": 0.148777037858963, "learning_rate": 1.8098552361267397e-05, "loss": 0.0074, "step": 58740 }, { "grad_norm": 0.2521968185901642, "learning_rate": 1.8082640184526763e-05, "loss": 0.0108, "step": 58750 }, { "grad_norm": 0.16286565363407135, "learning_rate": 1.806673346158999e-05, "loss": 0.0083, "step": 58760 }, { "grad_norm": 0.12367700785398483, "learning_rate": 1.8050832195175067e-05, "loss": 0.0124, "step": 58770 }, { "grad_norm": 0.2156285047531128, "learning_rate": 1.8034936387999136e-05, "loss": 0.0126, "step": 58780 }, { "grad_norm": 0.14800690114498138, "learning_rate": 1.8019046042778315e-05, "loss": 0.0082, "step": 58790 }, { "grad_norm": 0.15475402772426605, "learning_rate": 1.800316116222785e-05, "loss": 0.0085, "step": 58800 }, { "grad_norm": 0.16672328114509583, "learning_rate": 1.7987281749062018e-05, "loss": 0.0345, "step": 58810 }, { "grad_norm": 0.1379934698343277, "learning_rate": 1.7971407805994195e-05, "loss": 0.0185, "step": 58820 }, { "grad_norm": 0.12134523689746857, "learning_rate": 1.7955539335736787e-05, "loss": 0.0121, "step": 58830 }, { "grad_norm": 0.14714597165584564, "learning_rate": 1.7939676341001304e-05, "loss": 0.0173, "step": 58840 }, { "grad_norm": 0.16458624601364136, "learning_rate": 1.7923818824498275e-05, "loss": 0.0086, "step": 58850 }, { "grad_norm": 0.21894115209579468, "learning_rate": 1.7907966788937315e-05, "loss": 0.0098, "step": 58860 }, { "grad_norm": 0.2960582375526428, "learning_rate": 1.7892120237027116e-05, "loss": 0.0157, "step": 58870 }, { "grad_norm": 0.13208025693893433, "learning_rate": 1.7876279171475413e-05, "loss": 0.011, "step": 58880 }, { "grad_norm": 0.17232365906238556, "learning_rate": 1.7860443594989028e-05, "loss": 0.0078, "step": 58890 }, { "grad_norm": 0.11245673894882202, "learning_rate": 1.784461351027379e-05, "loss": 0.0076, "step": 58900 }, { "grad_norm": 0.14590363204479218, "learning_rate": 1.7828788920034677e-05, "loss": 0.0159, "step": 58910 }, { "grad_norm": 0.24071168899536133, "learning_rate": 1.7812969826975623e-05, "loss": 0.0132, "step": 58920 }, { "grad_norm": 0.24897591769695282, "learning_rate": 1.7797156233799738e-05, "loss": 0.0137, "step": 58930 }, { "grad_norm": 0.15717872977256775, "learning_rate": 1.7781348143209094e-05, "loss": 0.0135, "step": 58940 }, { "grad_norm": 0.18060146272182465, "learning_rate": 1.7765545557904862e-05, "loss": 0.0228, "step": 58950 }, { "grad_norm": 0.23696358501911163, "learning_rate": 1.7749748480587302e-05, "loss": 0.0109, "step": 58960 }, { "grad_norm": 0.15505856275558472, "learning_rate": 1.773395691395564e-05, "loss": 0.0088, "step": 58970 }, { "grad_norm": 0.08686832338571548, "learning_rate": 1.7718170860708305e-05, "loss": 0.0066, "step": 58980 }, { "grad_norm": 0.21226035058498383, "learning_rate": 1.770239032354264e-05, "loss": 0.0088, "step": 58990 }, { "grad_norm": 0.206813246011734, "learning_rate": 1.7686615305155124e-05, "loss": 0.0159, "step": 59000 }, { "grad_norm": 0.20371092855930328, "learning_rate": 1.767084580824128e-05, "loss": 0.0075, "step": 59010 }, { "grad_norm": 0.3941778540611267, "learning_rate": 1.765508183549569e-05, "loss": 0.0189, "step": 59020 }, { "grad_norm": 0.14010970294475555, "learning_rate": 1.763932338961199e-05, "loss": 0.01, "step": 59030 }, { "grad_norm": 0.10984008759260178, "learning_rate": 1.762357047328284e-05, "loss": 0.0093, "step": 59040 }, { "grad_norm": 0.25140053033828735, "learning_rate": 1.7607823089199997e-05, "loss": 0.0152, "step": 59050 }, { "grad_norm": 0.16387128829956055, "learning_rate": 1.7592081240054265e-05, "loss": 0.0117, "step": 59060 }, { "grad_norm": 0.16759653389453888, "learning_rate": 1.7576344928535488e-05, "loss": 0.012, "step": 59070 }, { "grad_norm": 0.208556666970253, "learning_rate": 1.7560614157332573e-05, "loss": 0.0122, "step": 59080 }, { "grad_norm": 0.17481611669063568, "learning_rate": 1.7544888929133495e-05, "loss": 0.0112, "step": 59090 }, { "grad_norm": 0.1414664089679718, "learning_rate": 1.752916924662522e-05, "loss": 0.012, "step": 59100 }, { "grad_norm": 0.17281703650951385, "learning_rate": 1.751345511249387e-05, "loss": 0.0078, "step": 59110 }, { "grad_norm": 0.2007315307855606, "learning_rate": 1.749774652942452e-05, "loss": 0.0099, "step": 59120 }, { "grad_norm": 0.17157070338726044, "learning_rate": 1.748204350010135e-05, "loss": 0.0122, "step": 59130 }, { "grad_norm": 0.15628232061862946, "learning_rate": 1.7466346027207574e-05, "loss": 0.0092, "step": 59140 }, { "grad_norm": 0.23685555160045624, "learning_rate": 1.745065411342547e-05, "loss": 0.0096, "step": 59150 }, { "grad_norm": 0.17363889515399933, "learning_rate": 1.7434967761436366e-05, "loss": 0.015, "step": 59160 }, { "grad_norm": 0.24485230445861816, "learning_rate": 1.741928697392058e-05, "loss": 0.0089, "step": 59170 }, { "grad_norm": 0.3036203384399414, "learning_rate": 1.7403611753557597e-05, "loss": 0.0069, "step": 59180 }, { "grad_norm": 0.1184002086520195, "learning_rate": 1.738794210302584e-05, "loss": 0.0055, "step": 59190 }, { "grad_norm": 0.10612405091524124, "learning_rate": 1.7372278025002837e-05, "loss": 0.0062, "step": 59200 }, { "grad_norm": 0.14885160326957703, "learning_rate": 1.7356619522165164e-05, "loss": 0.0085, "step": 59210 }, { "grad_norm": 0.20923329889774323, "learning_rate": 1.7340966597188378e-05, "loss": 0.0097, "step": 59220 }, { "grad_norm": 0.14827792346477509, "learning_rate": 1.732531925274722e-05, "loss": 0.0094, "step": 59230 }, { "grad_norm": 0.18201805651187897, "learning_rate": 1.7309677491515318e-05, "loss": 0.0079, "step": 59240 }, { "grad_norm": 0.23549096286296844, "learning_rate": 1.7294041316165455e-05, "loss": 0.0086, "step": 59250 }, { "grad_norm": 0.1310437023639679, "learning_rate": 1.727841072936942e-05, "loss": 0.0083, "step": 59260 }, { "grad_norm": 0.25781258940696716, "learning_rate": 1.7262785733798058e-05, "loss": 0.0074, "step": 59270 }, { "grad_norm": 0.2541799545288086, "learning_rate": 1.7247166332121246e-05, "loss": 0.0101, "step": 59280 }, { "grad_norm": 0.25979238748550415, "learning_rate": 1.7231552527007933e-05, "loss": 0.0079, "step": 59290 }, { "grad_norm": 0.4326173961162567, "learning_rate": 1.721594432112606e-05, "loss": 0.0075, "step": 59300 }, { "grad_norm": 0.1320560872554779, "learning_rate": 1.7200341717142655e-05, "loss": 0.013, "step": 59310 }, { "grad_norm": 0.265225887298584, "learning_rate": 1.718474471772377e-05, "loss": 0.0092, "step": 59320 }, { "grad_norm": 0.1825990378856659, "learning_rate": 1.7169153325534528e-05, "loss": 0.0076, "step": 59330 }, { "grad_norm": 0.1582985371351242, "learning_rate": 1.7153567543239045e-05, "loss": 0.009, "step": 59340 }, { "grad_norm": 0.14211823046207428, "learning_rate": 1.7137987373500525e-05, "loss": 0.0079, "step": 59350 }, { "grad_norm": 0.09003446251153946, "learning_rate": 1.7122412818981198e-05, "loss": 0.011, "step": 59360 }, { "grad_norm": 0.20437650382518768, "learning_rate": 1.7106843882342283e-05, "loss": 0.0099, "step": 59370 }, { "grad_norm": 0.3066485524177551, "learning_rate": 1.709128056624415e-05, "loss": 0.0093, "step": 59380 }, { "grad_norm": 0.24616120755672455, "learning_rate": 1.70757228733461e-05, "loss": 0.0076, "step": 59390 }, { "grad_norm": 0.1804358810186386, "learning_rate": 1.706017080630653e-05, "loss": 0.0089, "step": 59400 }, { "grad_norm": 0.11490166187286377, "learning_rate": 1.7044624367782873e-05, "loss": 0.0085, "step": 59410 }, { "grad_norm": 0.11595356464385986, "learning_rate": 1.7029083560431553e-05, "loss": 0.0071, "step": 59420 }, { "grad_norm": 0.16465814411640167, "learning_rate": 1.701354838690813e-05, "loss": 0.015, "step": 59430 }, { "grad_norm": 0.19133585691452026, "learning_rate": 1.6998018849867087e-05, "loss": 0.0073, "step": 59440 }, { "grad_norm": 0.1863844394683838, "learning_rate": 1.698249495196202e-05, "loss": 0.0097, "step": 59450 }, { "grad_norm": 0.14600864052772522, "learning_rate": 1.6966976695845528e-05, "loss": 0.0099, "step": 59460 }, { "grad_norm": 0.09395867586135864, "learning_rate": 1.6951464084169268e-05, "loss": 0.0076, "step": 59470 }, { "grad_norm": 0.20735333859920502, "learning_rate": 1.6935957119583916e-05, "loss": 0.0104, "step": 59480 }, { "grad_norm": 0.13552948832511902, "learning_rate": 1.6920455804739205e-05, "loss": 0.0088, "step": 59490 }, { "grad_norm": 0.20824189484119415, "learning_rate": 1.6904960142283855e-05, "loss": 0.0128, "step": 59500 }, { "grad_norm": 0.11369499564170837, "learning_rate": 1.6889470134865666e-05, "loss": 0.0063, "step": 59510 }, { "grad_norm": 0.1557944267988205, "learning_rate": 1.687398578513145e-05, "loss": 0.0086, "step": 59520 }, { "grad_norm": 0.11580207943916321, "learning_rate": 1.6858507095727066e-05, "loss": 0.0081, "step": 59530 }, { "grad_norm": 0.18247650563716888, "learning_rate": 1.6843034069297403e-05, "loss": 0.0071, "step": 59540 }, { "grad_norm": 0.18705137073993683, "learning_rate": 1.682756670848637e-05, "loss": 0.0186, "step": 59550 }, { "grad_norm": 0.1017235741019249, "learning_rate": 1.6812105015936937e-05, "loss": 0.006, "step": 59560 }, { "grad_norm": 0.18289366364479065, "learning_rate": 1.6796648994291027e-05, "loss": 0.0152, "step": 59570 }, { "grad_norm": 0.1602652221918106, "learning_rate": 1.678119864618973e-05, "loss": 0.0056, "step": 59580 }, { "grad_norm": 0.09394737333059311, "learning_rate": 1.6765753974273023e-05, "loss": 0.0054, "step": 59590 }, { "grad_norm": 0.10426755994558334, "learning_rate": 1.675031498118001e-05, "loss": 0.0068, "step": 59600 }, { "grad_norm": 0.1313449740409851, "learning_rate": 1.6734881669548804e-05, "loss": 0.004, "step": 59610 }, { "grad_norm": 0.15557727217674255, "learning_rate": 1.671945404201648e-05, "loss": 0.0125, "step": 59620 }, { "grad_norm": 0.20606936514377594, "learning_rate": 1.670403210121927e-05, "loss": 0.006, "step": 59630 }, { "grad_norm": 0.22352463006973267, "learning_rate": 1.6688615849792312e-05, "loss": 0.0077, "step": 59640 }, { "grad_norm": 0.3053678274154663, "learning_rate": 1.6673205290369832e-05, "loss": 0.0097, "step": 59650 }, { "grad_norm": 0.21033014357089996, "learning_rate": 1.6657800425585086e-05, "loss": 0.0113, "step": 59660 }, { "grad_norm": 0.17273087799549103, "learning_rate": 1.6642401258070327e-05, "loss": 0.0152, "step": 59670 }, { "grad_norm": 0.26517197489738464, "learning_rate": 1.6627007790456884e-05, "loss": 0.0134, "step": 59680 }, { "grad_norm": 0.13479796051979065, "learning_rate": 1.6611620025375037e-05, "loss": 0.0081, "step": 59690 }, { "grad_norm": 0.15756970643997192, "learning_rate": 1.6596237965454154e-05, "loss": 0.0146, "step": 59700 }, { "grad_norm": 0.2827577292919159, "learning_rate": 1.6580861613322607e-05, "loss": 0.0117, "step": 59710 }, { "grad_norm": 0.14140859246253967, "learning_rate": 1.6565490971607796e-05, "loss": 0.0055, "step": 59720 }, { "grad_norm": 0.27935662865638733, "learning_rate": 1.655012604293614e-05, "loss": 0.0095, "step": 59730 }, { "grad_norm": 0.1594454050064087, "learning_rate": 1.653476682993309e-05, "loss": 0.011, "step": 59740 }, { "grad_norm": 0.1633843332529068, "learning_rate": 1.651941333522311e-05, "loss": 0.0076, "step": 59750 }, { "grad_norm": 0.2249433547258377, "learning_rate": 1.6504065561429715e-05, "loss": 0.0097, "step": 59760 }, { "grad_norm": 0.15712715685367584, "learning_rate": 1.6488723511175385e-05, "loss": 0.0141, "step": 59770 }, { "grad_norm": 0.23788240551948547, "learning_rate": 1.6473387187081668e-05, "loss": 0.0104, "step": 59780 }, { "grad_norm": 0.16906115412712097, "learning_rate": 1.6458056591769123e-05, "loss": 0.0069, "step": 59790 }, { "grad_norm": 0.12110833823680878, "learning_rate": 1.6442731727857335e-05, "loss": 0.0073, "step": 59800 }, { "grad_norm": 0.2675979435443878, "learning_rate": 1.642741259796492e-05, "loss": 0.0079, "step": 59810 }, { "grad_norm": 0.14154131710529327, "learning_rate": 1.641209920470944e-05, "loss": 0.0089, "step": 59820 }, { "grad_norm": 0.1397903561592102, "learning_rate": 1.639679155070762e-05, "loss": 0.0069, "step": 59830 }, { "grad_norm": 0.24590255320072174, "learning_rate": 1.6381489638575048e-05, "loss": 0.0073, "step": 59840 }, { "grad_norm": 0.11954742670059204, "learning_rate": 1.636619347092643e-05, "loss": 0.0063, "step": 59850 }, { "grad_norm": 0.279387503862381, "learning_rate": 1.6350903050375476e-05, "loss": 0.008, "step": 59860 }, { "grad_norm": 0.1945524513721466, "learning_rate": 1.6335618379534856e-05, "loss": 0.0079, "step": 59870 }, { "grad_norm": 0.14563122391700745, "learning_rate": 1.6320339461016364e-05, "loss": 0.0063, "step": 59880 }, { "grad_norm": 0.10014659911394119, "learning_rate": 1.6305066297430687e-05, "loss": 0.0072, "step": 59890 }, { "grad_norm": 0.11588648706674576, "learning_rate": 1.6289798891387654e-05, "loss": 0.0111, "step": 59900 }, { "grad_norm": 0.14688338339328766, "learning_rate": 1.6274537245495995e-05, "loss": 0.0062, "step": 59910 }, { "grad_norm": 0.25117209553718567, "learning_rate": 1.6259281362363527e-05, "loss": 0.0083, "step": 59920 }, { "grad_norm": 0.08818996697664261, "learning_rate": 1.6244031244597068e-05, "loss": 0.0168, "step": 59930 }, { "grad_norm": 0.14954686164855957, "learning_rate": 1.622878689480244e-05, "loss": 0.0075, "step": 59940 }, { "grad_norm": 0.20831389725208282, "learning_rate": 1.6213548315584498e-05, "loss": 0.0123, "step": 59950 }, { "grad_norm": 0.17637301981449127, "learning_rate": 1.6198315509547074e-05, "loss": 0.0087, "step": 59960 }, { "grad_norm": 0.31354373693466187, "learning_rate": 1.6183088479293056e-05, "loss": 0.0064, "step": 59970 }, { "grad_norm": 0.1473359912633896, "learning_rate": 1.6167867227424316e-05, "loss": 0.0074, "step": 59980 }, { "grad_norm": 0.21396347880363464, "learning_rate": 1.615265175654176e-05, "loss": 0.0121, "step": 59990 }, { "grad_norm": 0.14485205709934235, "learning_rate": 1.613744206924529e-05, "loss": 0.0069, "step": 60000 }, { "grad_norm": 0.1430296152830124, "learning_rate": 1.6122238168133845e-05, "loss": 0.0084, "step": 60010 }, { "grad_norm": 0.11171312630176544, "learning_rate": 1.6107040055805305e-05, "loss": 0.0097, "step": 60020 }, { "grad_norm": 0.2377203106880188, "learning_rate": 1.6091847734856673e-05, "loss": 0.0081, "step": 60030 }, { "grad_norm": 0.28683772683143616, "learning_rate": 1.6076661207883865e-05, "loss": 0.0101, "step": 60040 }, { "grad_norm": 0.08558584004640579, "learning_rate": 1.6061480477481848e-05, "loss": 0.0074, "step": 60050 }, { "grad_norm": 0.1742011457681656, "learning_rate": 1.604630554624461e-05, "loss": 0.0063, "step": 60060 }, { "grad_norm": 0.1889689564704895, "learning_rate": 1.603113641676509e-05, "loss": 0.0069, "step": 60070 }, { "grad_norm": 0.09099573642015457, "learning_rate": 1.6015973091635338e-05, "loss": 0.0069, "step": 60080 }, { "grad_norm": 0.11995182931423187, "learning_rate": 1.6000815573446288e-05, "loss": 0.0079, "step": 60090 }, { "grad_norm": 0.18617841601371765, "learning_rate": 1.5985663864788002e-05, "loss": 0.0174, "step": 60100 }, { "grad_norm": 0.16752035915851593, "learning_rate": 1.597051796824946e-05, "loss": 0.0124, "step": 60110 }, { "grad_norm": 0.18349158763885498, "learning_rate": 1.5955377886418682e-05, "loss": 0.0082, "step": 60120 }, { "grad_norm": 0.19670312106609344, "learning_rate": 1.5940243621882704e-05, "loss": 0.0073, "step": 60130 }, { "grad_norm": 0.2626708745956421, "learning_rate": 1.5925115177227555e-05, "loss": 0.0123, "step": 60140 }, { "grad_norm": 0.20318621397018433, "learning_rate": 1.5909992555038288e-05, "loss": 0.0094, "step": 60150 }, { "grad_norm": 0.20561689138412476, "learning_rate": 1.5894875757898912e-05, "loss": 0.0147, "step": 60160 }, { "grad_norm": 0.16998693346977234, "learning_rate": 1.58797647883925e-05, "loss": 0.0087, "step": 60170 }, { "grad_norm": 0.15450908243656158, "learning_rate": 1.586465964910109e-05, "loss": 0.0126, "step": 60180 }, { "grad_norm": 0.11661290377378464, "learning_rate": 1.5849560342605734e-05, "loss": 0.0095, "step": 60190 }, { "grad_norm": 0.10404377430677414, "learning_rate": 1.583446687148651e-05, "loss": 0.0087, "step": 60200 }, { "grad_norm": 0.12625043094158173, "learning_rate": 1.581937923832248e-05, "loss": 0.0063, "step": 60210 }, { "grad_norm": 0.10585851222276688, "learning_rate": 1.5804297445691663e-05, "loss": 0.0089, "step": 60220 }, { "grad_norm": 0.2742137908935547, "learning_rate": 1.578922149617119e-05, "loss": 0.0129, "step": 60230 }, { "grad_norm": 0.15369297564029694, "learning_rate": 1.5774151392337084e-05, "loss": 0.0075, "step": 60240 }, { "grad_norm": 0.2504644989967346, "learning_rate": 1.575908713676442e-05, "loss": 0.0104, "step": 60250 }, { "grad_norm": 0.29708197712898254, "learning_rate": 1.5744028732027293e-05, "loss": 0.0107, "step": 60260 }, { "grad_norm": 0.302521675825119, "learning_rate": 1.5728976180698723e-05, "loss": 0.0092, "step": 60270 }, { "grad_norm": 0.13914546370506287, "learning_rate": 1.5713929485350836e-05, "loss": 0.0124, "step": 60280 }, { "grad_norm": 0.17779962718486786, "learning_rate": 1.569888864855465e-05, "loss": 0.0089, "step": 60290 }, { "grad_norm": 0.20782308280467987, "learning_rate": 1.568385367288028e-05, "loss": 0.0075, "step": 60300 }, { "grad_norm": 0.15391846001148224, "learning_rate": 1.5668824560896755e-05, "loss": 0.0087, "step": 60310 }, { "grad_norm": 0.19523033499717712, "learning_rate": 1.5653801315172156e-05, "loss": 0.0078, "step": 60320 }, { "grad_norm": 0.1908976137638092, "learning_rate": 1.5638783938273554e-05, "loss": 0.0074, "step": 60330 }, { "grad_norm": 0.2613268792629242, "learning_rate": 1.5623772432766966e-05, "loss": 0.0126, "step": 60340 }, { "grad_norm": 0.23475363850593567, "learning_rate": 1.5608766801217507e-05, "loss": 0.0093, "step": 60350 }, { "grad_norm": 0.1813865453004837, "learning_rate": 1.5593767046189184e-05, "loss": 0.0081, "step": 60360 }, { "grad_norm": 0.15552160143852234, "learning_rate": 1.5578773170245064e-05, "loss": 0.0071, "step": 60370 }, { "grad_norm": 0.2712749242782593, "learning_rate": 1.5563785175947182e-05, "loss": 0.01, "step": 60380 }, { "grad_norm": 0.13474129140377045, "learning_rate": 1.5548803065856582e-05, "loss": 0.0066, "step": 60390 }, { "grad_norm": 0.16202522814273834, "learning_rate": 1.55338268425333e-05, "loss": 0.0086, "step": 60400 }, { "grad_norm": 0.1250137984752655, "learning_rate": 1.5518856508536373e-05, "loss": 0.0069, "step": 60410 }, { "grad_norm": 0.24620790779590607, "learning_rate": 1.5503892066423786e-05, "loss": 0.01, "step": 60420 }, { "grad_norm": 0.2680128514766693, "learning_rate": 1.548893351875258e-05, "loss": 0.0185, "step": 60430 }, { "grad_norm": 0.2681436240673065, "learning_rate": 1.547398086807875e-05, "loss": 0.0125, "step": 60440 }, { "grad_norm": 0.2541140913963318, "learning_rate": 1.5459034116957305e-05, "loss": 0.0132, "step": 60450 }, { "grad_norm": 0.1316530704498291, "learning_rate": 1.544409326794225e-05, "loss": 0.0175, "step": 60460 }, { "grad_norm": 0.2691238224506378, "learning_rate": 1.542915832358651e-05, "loss": 0.012, "step": 60470 }, { "grad_norm": 0.15144531428813934, "learning_rate": 1.541422928644214e-05, "loss": 0.0075, "step": 60480 }, { "grad_norm": 0.21395641565322876, "learning_rate": 1.5399306159060024e-05, "loss": 0.0147, "step": 60490 }, { "grad_norm": 0.1345624029636383, "learning_rate": 1.5384388943990185e-05, "loss": 0.0074, "step": 60500 }, { "grad_norm": 0.15528631210327148, "learning_rate": 1.5369477643781526e-05, "loss": 0.0074, "step": 60510 }, { "grad_norm": 0.15073955059051514, "learning_rate": 1.5354572260981985e-05, "loss": 0.0084, "step": 60520 }, { "grad_norm": 0.2609146237373352, "learning_rate": 1.533967279813851e-05, "loss": 0.0061, "step": 60530 }, { "grad_norm": 0.2655382752418518, "learning_rate": 1.5324779257796956e-05, "loss": 0.0102, "step": 60540 }, { "grad_norm": 0.10337822139263153, "learning_rate": 1.5309891642502293e-05, "loss": 0.0097, "step": 60550 }, { "grad_norm": 0.12575462460517883, "learning_rate": 1.5295009954798357e-05, "loss": 0.0063, "step": 60560 }, { "grad_norm": 0.21135376393795013, "learning_rate": 1.5280134197228036e-05, "loss": 0.0116, "step": 60570 }, { "grad_norm": 0.12413696944713593, "learning_rate": 1.526526437233319e-05, "loss": 0.0092, "step": 60580 }, { "grad_norm": 0.15267400443553925, "learning_rate": 1.525040048265466e-05, "loss": 0.0083, "step": 60590 }, { "grad_norm": 0.09552446752786636, "learning_rate": 1.5235542530732288e-05, "loss": 0.0069, "step": 60600 }, { "grad_norm": 0.14574241638183594, "learning_rate": 1.5220690519104901e-05, "loss": 0.0088, "step": 60610 }, { "grad_norm": 0.15218476951122284, "learning_rate": 1.5205844450310275e-05, "loss": 0.0061, "step": 60620 }, { "grad_norm": 0.27801334857940674, "learning_rate": 1.519100432688521e-05, "loss": 0.0077, "step": 60630 }, { "grad_norm": 0.21304158866405487, "learning_rate": 1.5176170151365476e-05, "loss": 0.0074, "step": 60640 }, { "grad_norm": 0.11579263210296631, "learning_rate": 1.5161341926285833e-05, "loss": 0.0063, "step": 60650 }, { "grad_norm": 0.24181883037090302, "learning_rate": 1.5146519654180025e-05, "loss": 0.0064, "step": 60660 }, { "grad_norm": 0.18813703954219818, "learning_rate": 1.5131703337580739e-05, "loss": 0.0115, "step": 60670 }, { "grad_norm": 0.15478605031967163, "learning_rate": 1.5116892979019731e-05, "loss": 0.0135, "step": 60680 }, { "grad_norm": 0.11707736551761627, "learning_rate": 1.5102088581027623e-05, "loss": 0.0062, "step": 60690 }, { "grad_norm": 0.12089216709136963, "learning_rate": 1.5087290146134154e-05, "loss": 0.0073, "step": 60700 }, { "grad_norm": 0.19691956043243408, "learning_rate": 1.5072497676867915e-05, "loss": 0.0082, "step": 60710 }, { "grad_norm": 0.20364253222942352, "learning_rate": 1.505771117575655e-05, "loss": 0.0094, "step": 60720 }, { "grad_norm": 0.19256678223609924, "learning_rate": 1.5042930645326691e-05, "loss": 0.007, "step": 60730 }, { "grad_norm": 0.21778714656829834, "learning_rate": 1.5028156088103879e-05, "loss": 0.0105, "step": 60740 }, { "grad_norm": 0.4024431109428406, "learning_rate": 1.5013387506612735e-05, "loss": 0.0143, "step": 60750 }, { "grad_norm": 0.07858987152576447, "learning_rate": 1.4998624903376767e-05, "loss": 0.0109, "step": 60760 }, { "grad_norm": 0.2846398651599884, "learning_rate": 1.4983868280918517e-05, "loss": 0.0085, "step": 60770 }, { "grad_norm": 0.26060909032821655, "learning_rate": 1.4969117641759478e-05, "loss": 0.0173, "step": 60780 }, { "grad_norm": 0.09839010238647461, "learning_rate": 1.495437298842014e-05, "loss": 0.0095, "step": 60790 }, { "grad_norm": 0.19775277376174927, "learning_rate": 1.4939634323419976e-05, "loss": 0.0087, "step": 60800 }, { "grad_norm": 0.2942611575126648, "learning_rate": 1.4924901649277384e-05, "loss": 0.0113, "step": 60810 }, { "grad_norm": 0.10189679265022278, "learning_rate": 1.4910174968509793e-05, "loss": 0.0051, "step": 60820 }, { "grad_norm": 0.2569611966609955, "learning_rate": 1.4895454283633598e-05, "loss": 0.0074, "step": 60830 }, { "grad_norm": 0.44945842027664185, "learning_rate": 1.488073959716415e-05, "loss": 0.009, "step": 60840 }, { "grad_norm": 0.252946674823761, "learning_rate": 1.4866030911615791e-05, "loss": 0.0107, "step": 60850 }, { "grad_norm": 0.15430644154548645, "learning_rate": 1.4851328229501849e-05, "loss": 0.0101, "step": 60860 }, { "grad_norm": 0.2099466770887375, "learning_rate": 1.4836631553334562e-05, "loss": 0.0071, "step": 60870 }, { "grad_norm": 0.21358662843704224, "learning_rate": 1.4821940885625251e-05, "loss": 0.0084, "step": 60880 }, { "grad_norm": 0.2072809487581253, "learning_rate": 1.4807256228884109e-05, "loss": 0.0076, "step": 60890 }, { "grad_norm": 0.17620711028575897, "learning_rate": 1.4792577585620353e-05, "loss": 0.0064, "step": 60900 }, { "grad_norm": 0.16167649626731873, "learning_rate": 1.4777904958342164e-05, "loss": 0.0119, "step": 60910 }, { "grad_norm": 0.15575304627418518, "learning_rate": 1.4763238349556691e-05, "loss": 0.0072, "step": 60920 }, { "grad_norm": 0.1655287742614746, "learning_rate": 1.4748577761770072e-05, "loss": 0.0131, "step": 60930 }, { "grad_norm": 0.28727859258651733, "learning_rate": 1.4733923197487354e-05, "loss": 0.0086, "step": 60940 }, { "grad_norm": 0.21931086480617523, "learning_rate": 1.4719274659212662e-05, "loss": 0.0104, "step": 60950 }, { "grad_norm": 0.128085196018219, "learning_rate": 1.4704632149448983e-05, "loss": 0.0078, "step": 60960 }, { "grad_norm": 0.11825179308652878, "learning_rate": 1.4689995670698343e-05, "loss": 0.0181, "step": 60970 }, { "grad_norm": 0.16057837009429932, "learning_rate": 1.4675365225461728e-05, "loss": 0.0138, "step": 60980 }, { "grad_norm": 0.25364044308662415, "learning_rate": 1.4660740816239032e-05, "loss": 0.0083, "step": 60990 }, { "grad_norm": 0.14940369129180908, "learning_rate": 1.464612244552923e-05, "loss": 0.0098, "step": 61000 }, { "grad_norm": 0.08604249358177185, "learning_rate": 1.4631510115830161e-05, "loss": 0.0065, "step": 61010 }, { "grad_norm": 0.1451258808374405, "learning_rate": 1.4616903829638679e-05, "loss": 0.008, "step": 61020 }, { "grad_norm": 0.14190930128097534, "learning_rate": 1.46023035894506e-05, "loss": 0.0086, "step": 61030 }, { "grad_norm": 0.2344444990158081, "learning_rate": 1.4587709397760713e-05, "loss": 0.0104, "step": 61040 }, { "grad_norm": 0.11359509080648422, "learning_rate": 1.4573121257062755e-05, "loss": 0.0085, "step": 61050 }, { "grad_norm": 0.46536490321159363, "learning_rate": 1.4558539169849472e-05, "loss": 0.0138, "step": 61060 }, { "grad_norm": 0.26979735493659973, "learning_rate": 1.4543963138612499e-05, "loss": 0.008, "step": 61070 }, { "grad_norm": 0.1573380082845688, "learning_rate": 1.4529393165842498e-05, "loss": 0.0078, "step": 61080 }, { "grad_norm": 0.30696776509284973, "learning_rate": 1.4514829254029084e-05, "loss": 0.008, "step": 61090 }, { "grad_norm": 0.10090843588113785, "learning_rate": 1.450027140566083e-05, "loss": 0.008, "step": 61100 }, { "grad_norm": 0.17482887208461761, "learning_rate": 1.4485719623225268e-05, "loss": 0.0078, "step": 61110 }, { "grad_norm": 0.11261728405952454, "learning_rate": 1.4471173909208912e-05, "loss": 0.0069, "step": 61120 }, { "grad_norm": 0.1609513908624649, "learning_rate": 1.4456634266097236e-05, "loss": 0.0083, "step": 61130 }, { "grad_norm": 0.13052009046077728, "learning_rate": 1.444210069637461e-05, "loss": 0.0144, "step": 61140 }, { "grad_norm": 0.1280176043510437, "learning_rate": 1.4427573202524502e-05, "loss": 0.0058, "step": 61150 }, { "grad_norm": 0.16116547584533691, "learning_rate": 1.4413051787029208e-05, "loss": 0.0143, "step": 61160 }, { "grad_norm": 0.1476346105337143, "learning_rate": 1.439853645237006e-05, "loss": 0.0102, "step": 61170 }, { "grad_norm": 0.07910706847906113, "learning_rate": 1.4384027201027344e-05, "loss": 0.0072, "step": 61180 }, { "grad_norm": 0.11683440953493118, "learning_rate": 1.4369524035480253e-05, "loss": 0.0094, "step": 61190 }, { "grad_norm": 0.09965354949235916, "learning_rate": 1.4355026958207035e-05, "loss": 0.0086, "step": 61200 }, { "grad_norm": 0.13430939614772797, "learning_rate": 1.434053597168481e-05, "loss": 0.0097, "step": 61210 }, { "grad_norm": 0.12887734174728394, "learning_rate": 1.432605107838969e-05, "loss": 0.0093, "step": 61220 }, { "grad_norm": 0.23809631168842316, "learning_rate": 1.4311572280796759e-05, "loss": 0.0109, "step": 61230 }, { "grad_norm": 0.21703052520751953, "learning_rate": 1.4297099581380047e-05, "loss": 0.0129, "step": 61240 }, { "grad_norm": 0.18171291053295135, "learning_rate": 1.4282632982612538e-05, "loss": 0.012, "step": 61250 }, { "grad_norm": 0.08940233290195465, "learning_rate": 1.4268172486966197e-05, "loss": 0.0113, "step": 61260 }, { "grad_norm": 0.10156593471765518, "learning_rate": 1.4253718096911894e-05, "loss": 0.012, "step": 61270 }, { "grad_norm": 0.22478631138801575, "learning_rate": 1.4239269814919504e-05, "loss": 0.0153, "step": 61280 }, { "grad_norm": 0.18396347761154175, "learning_rate": 1.4224827643457843e-05, "loss": 0.014, "step": 61290 }, { "grad_norm": 0.1434643417596817, "learning_rate": 1.4210391584994686e-05, "loss": 0.0071, "step": 61300 }, { "grad_norm": 0.181910902261734, "learning_rate": 1.4195961641996763e-05, "loss": 0.0095, "step": 61310 }, { "grad_norm": 0.13597826659679413, "learning_rate": 1.4181537816929751e-05, "loss": 0.0068, "step": 61320 }, { "grad_norm": 0.24208945035934448, "learning_rate": 1.4167120112258303e-05, "loss": 0.0137, "step": 61330 }, { "grad_norm": 0.14582057297229767, "learning_rate": 1.4152708530445963e-05, "loss": 0.0052, "step": 61340 }, { "grad_norm": 0.09241156280040741, "learning_rate": 1.4138303073955344e-05, "loss": 0.0072, "step": 61350 }, { "grad_norm": 0.23287640511989594, "learning_rate": 1.4123903745247897e-05, "loss": 0.0098, "step": 61360 }, { "grad_norm": 0.14151079952716827, "learning_rate": 1.4109510546784082e-05, "loss": 0.0068, "step": 61370 }, { "grad_norm": 0.09764975309371948, "learning_rate": 1.4095123481023325e-05, "loss": 0.0141, "step": 61380 }, { "grad_norm": 0.18601101636886597, "learning_rate": 1.4080742550423936e-05, "loss": 0.0079, "step": 61390 }, { "grad_norm": 0.08181561529636383, "learning_rate": 1.4066367757443278e-05, "loss": 0.0081, "step": 61400 }, { "grad_norm": 0.18522286415100098, "learning_rate": 1.4051999104537572e-05, "loss": 0.0086, "step": 61410 }, { "grad_norm": 0.20318540930747986, "learning_rate": 1.4037636594162034e-05, "loss": 0.0097, "step": 61420 }, { "grad_norm": 0.1278901845216751, "learning_rate": 1.402328022877083e-05, "loss": 0.0053, "step": 61430 }, { "grad_norm": 0.22598311305046082, "learning_rate": 1.4008930010817073e-05, "loss": 0.0103, "step": 61440 }, { "grad_norm": 0.36201387643814087, "learning_rate": 1.3994585942752831e-05, "loss": 0.0075, "step": 61450 }, { "grad_norm": 0.11155307292938232, "learning_rate": 1.398024802702909e-05, "loss": 0.0061, "step": 61460 }, { "grad_norm": 0.1823987364768982, "learning_rate": 1.3965916266095819e-05, "loss": 0.0092, "step": 61470 }, { "grad_norm": 0.22267892956733704, "learning_rate": 1.395159066240192e-05, "loss": 0.0074, "step": 61480 }, { "grad_norm": 0.11359415203332901, "learning_rate": 1.3937271218395248e-05, "loss": 0.0076, "step": 61490 }, { "grad_norm": 0.2955538332462311, "learning_rate": 1.3922957936522613e-05, "loss": 0.0085, "step": 61500 }, { "grad_norm": 0.24961772561073303, "learning_rate": 1.3908650819229758e-05, "loss": 0.0114, "step": 61510 }, { "grad_norm": 0.1162494421005249, "learning_rate": 1.389434986896137e-05, "loss": 0.0075, "step": 61520 }, { "grad_norm": 0.1551981121301651, "learning_rate": 1.388005508816112e-05, "loss": 0.0068, "step": 61530 }, { "grad_norm": 0.14105458557605743, "learning_rate": 1.3865766479271557e-05, "loss": 0.0104, "step": 61540 }, { "grad_norm": 0.12366987764835358, "learning_rate": 1.3851484044734225e-05, "loss": 0.0183, "step": 61550 }, { "grad_norm": 0.21838364005088806, "learning_rate": 1.3837207786989608e-05, "loss": 0.0196, "step": 61560 }, { "grad_norm": 0.11349747329950333, "learning_rate": 1.382293770847713e-05, "loss": 0.0069, "step": 61570 }, { "grad_norm": 0.26679033041000366, "learning_rate": 1.3808673811635159e-05, "loss": 0.0092, "step": 61580 }, { "grad_norm": 0.28468942642211914, "learning_rate": 1.3794416098900975e-05, "loss": 0.0125, "step": 61590 }, { "grad_norm": 0.17131949961185455, "learning_rate": 1.3780164572710886e-05, "loss": 0.0154, "step": 61600 }, { "grad_norm": 0.09989948570728302, "learning_rate": 1.3765919235500035e-05, "loss": 0.0117, "step": 61610 }, { "grad_norm": 0.28114086389541626, "learning_rate": 1.375168008970259e-05, "loss": 0.0089, "step": 61620 }, { "grad_norm": 0.11414875835180283, "learning_rate": 1.3737447137751635e-05, "loss": 0.0065, "step": 61630 }, { "grad_norm": 0.17598965764045715, "learning_rate": 1.372322038207915e-05, "loss": 0.0086, "step": 61640 }, { "grad_norm": 0.23659484088420868, "learning_rate": 1.3708999825116159e-05, "loss": 0.0162, "step": 61650 }, { "grad_norm": 0.2782646119594574, "learning_rate": 1.3694785469292526e-05, "loss": 0.0148, "step": 61660 }, { "grad_norm": 0.13742904365062714, "learning_rate": 1.3680577317037101e-05, "loss": 0.0109, "step": 61670 }, { "grad_norm": 0.25111520290374756, "learning_rate": 1.3666375370777678e-05, "loss": 0.0074, "step": 61680 }, { "grad_norm": 0.10873088985681534, "learning_rate": 1.3652179632940981e-05, "loss": 0.0063, "step": 61690 }, { "grad_norm": 0.4353066682815552, "learning_rate": 1.3637990105952664e-05, "loss": 0.0086, "step": 61700 }, { "grad_norm": 0.12238821387290955, "learning_rate": 1.3623806792237337e-05, "loss": 0.0066, "step": 61710 }, { "grad_norm": 0.2578361928462982, "learning_rate": 1.3609629694218551e-05, "loss": 0.0086, "step": 61720 }, { "grad_norm": 0.1329210251569748, "learning_rate": 1.3595458814318762e-05, "loss": 0.007, "step": 61730 }, { "grad_norm": 0.2738262712955475, "learning_rate": 1.3581294154959389e-05, "loss": 0.0105, "step": 61740 }, { "grad_norm": 0.21107983589172363, "learning_rate": 1.3567135718560792e-05, "loss": 0.0057, "step": 61750 }, { "grad_norm": 0.14962905645370483, "learning_rate": 1.3552983507542261e-05, "loss": 0.0076, "step": 61760 }, { "grad_norm": 0.09118052572011948, "learning_rate": 1.3538837524322023e-05, "loss": 0.013, "step": 61770 }, { "grad_norm": 0.09051378816366196, "learning_rate": 1.3524697771317251e-05, "loss": 0.0148, "step": 61780 }, { "grad_norm": 0.12062003463506699, "learning_rate": 1.3510564250943997e-05, "loss": 0.0081, "step": 61790 }, { "grad_norm": 0.15334855020046234, "learning_rate": 1.3496436965617353e-05, "loss": 0.0063, "step": 61800 }, { "grad_norm": 0.17156969010829926, "learning_rate": 1.3482315917751243e-05, "loss": 0.0115, "step": 61810 }, { "grad_norm": 0.08954504877328873, "learning_rate": 1.3468201109758582e-05, "loss": 0.0069, "step": 61820 }, { "grad_norm": 0.11422216892242432, "learning_rate": 1.3454092544051222e-05, "loss": 0.011, "step": 61830 }, { "grad_norm": 0.2673143148422241, "learning_rate": 1.3439990223039878e-05, "loss": 0.0072, "step": 61840 }, { "grad_norm": 0.145237997174263, "learning_rate": 1.3425894149134321e-05, "loss": 0.0058, "step": 61850 }, { "grad_norm": 0.16867658495903015, "learning_rate": 1.3411804324743132e-05, "loss": 0.0097, "step": 61860 }, { "grad_norm": 0.1548888087272644, "learning_rate": 1.3397720752273896e-05, "loss": 0.0106, "step": 61870 }, { "grad_norm": 0.20380109548568726, "learning_rate": 1.3383643434133108e-05, "loss": 0.0079, "step": 61880 }, { "grad_norm": 0.10732901841402054, "learning_rate": 1.3369572372726196e-05, "loss": 0.0073, "step": 61890 }, { "grad_norm": 0.43523749709129333, "learning_rate": 1.3355507570457525e-05, "loss": 0.0068, "step": 61900 }, { "grad_norm": 0.28974640369415283, "learning_rate": 1.334144902973038e-05, "loss": 0.0066, "step": 61910 }, { "grad_norm": 0.15131910145282745, "learning_rate": 1.3327396752946997e-05, "loss": 0.0099, "step": 61920 }, { "grad_norm": 0.09741547703742981, "learning_rate": 1.3313350742508496e-05, "loss": 0.0093, "step": 61930 }, { "grad_norm": 0.10169877111911774, "learning_rate": 1.3299311000814973e-05, "loss": 0.0131, "step": 61940 }, { "grad_norm": 0.07307811081409454, "learning_rate": 1.3285277530265432e-05, "loss": 0.0132, "step": 61950 }, { "grad_norm": 0.19650420546531677, "learning_rate": 1.3271250333257813e-05, "loss": 0.0094, "step": 61960 }, { "grad_norm": 0.0726892277598381, "learning_rate": 1.3257229412188977e-05, "loss": 0.0061, "step": 61970 }, { "grad_norm": 0.25798872113227844, "learning_rate": 1.3243214769454726e-05, "loss": 0.0081, "step": 61980 }, { "grad_norm": 0.13148002326488495, "learning_rate": 1.3229206407449751e-05, "loss": 0.0073, "step": 61990 }, { "grad_norm": 0.22085309028625488, "learning_rate": 1.3215204328567742e-05, "loss": 0.0102, "step": 62000 }, { "grad_norm": 0.1659134477376938, "learning_rate": 1.3201208535201232e-05, "loss": 0.0071, "step": 62010 }, { "grad_norm": 0.07069988548755646, "learning_rate": 1.3187219029741737e-05, "loss": 0.0081, "step": 62020 }, { "grad_norm": 0.19719776511192322, "learning_rate": 1.3173235814579693e-05, "loss": 0.0073, "step": 62030 }, { "grad_norm": 0.22250428795814514, "learning_rate": 1.3159258892104398e-05, "loss": 0.0096, "step": 62040 }, { "grad_norm": 0.07211955636739731, "learning_rate": 1.3145288264704198e-05, "loss": 0.0106, "step": 62050 }, { "grad_norm": 0.13067860901355743, "learning_rate": 1.3131323934766237e-05, "loss": 0.0071, "step": 62060 }, { "grad_norm": 0.11220275610685349, "learning_rate": 1.311736590467666e-05, "loss": 0.0082, "step": 62070 }, { "grad_norm": 0.12012217938899994, "learning_rate": 1.3103414176820505e-05, "loss": 0.0101, "step": 62080 }, { "grad_norm": 0.12619003653526306, "learning_rate": 1.308946875358174e-05, "loss": 0.021, "step": 62090 }, { "grad_norm": 0.21931341290473938, "learning_rate": 1.3075529637343276e-05, "loss": 0.022, "step": 62100 }, { "grad_norm": 0.18340475857257843, "learning_rate": 1.3061596830486883e-05, "loss": 0.0116, "step": 62110 }, { "grad_norm": 0.23327656090259552, "learning_rate": 1.3047670335393353e-05, "loss": 0.0067, "step": 62120 }, { "grad_norm": 0.10267740488052368, "learning_rate": 1.3033750154442298e-05, "loss": 0.0092, "step": 62130 }, { "grad_norm": 0.08250294625759125, "learning_rate": 1.3019836290012316e-05, "loss": 0.0074, "step": 62140 }, { "grad_norm": 0.15937672555446625, "learning_rate": 1.30059287444809e-05, "loss": 0.0101, "step": 62150 }, { "grad_norm": 0.1472865492105484, "learning_rate": 1.299202752022447e-05, "loss": 0.0191, "step": 62160 }, { "grad_norm": 0.09890062361955643, "learning_rate": 1.2978132619618371e-05, "loss": 0.0056, "step": 62170 }, { "grad_norm": 0.1788104623556137, "learning_rate": 1.2964244045036866e-05, "loss": 0.0124, "step": 62180 }, { "grad_norm": 0.12866896390914917, "learning_rate": 1.295036179885311e-05, "loss": 0.0052, "step": 62190 }, { "grad_norm": 0.0906778946518898, "learning_rate": 1.293648588343922e-05, "loss": 0.0149, "step": 62200 }, { "grad_norm": 0.20620419085025787, "learning_rate": 1.2922616301166196e-05, "loss": 0.0093, "step": 62210 }, { "grad_norm": 0.1865997314453125, "learning_rate": 1.2908753054403976e-05, "loss": 0.0069, "step": 62220 }, { "grad_norm": 0.17512895166873932, "learning_rate": 1.2894896145521429e-05, "loss": 0.0065, "step": 62230 }, { "grad_norm": 0.20134630799293518, "learning_rate": 1.2881045576886275e-05, "loss": 0.0143, "step": 62240 }, { "grad_norm": 0.1285129189491272, "learning_rate": 1.2867201350865254e-05, "loss": 0.007, "step": 62250 }, { "grad_norm": 0.20429642498493195, "learning_rate": 1.2853363469823914e-05, "loss": 0.0077, "step": 62260 }, { "grad_norm": 0.17139935493469238, "learning_rate": 1.283953193612682e-05, "loss": 0.0058, "step": 62270 }, { "grad_norm": 0.19937266409397125, "learning_rate": 1.2825706752137372e-05, "loss": 0.0089, "step": 62280 }, { "grad_norm": 0.22256746888160706, "learning_rate": 1.2811887920217896e-05, "loss": 0.0106, "step": 62290 }, { "grad_norm": 0.15324868261814117, "learning_rate": 1.2798075442729707e-05, "loss": 0.007, "step": 62300 }, { "grad_norm": 0.13604190945625305, "learning_rate": 1.2784269322032922e-05, "loss": 0.0085, "step": 62310 }, { "grad_norm": 0.11015571653842926, "learning_rate": 1.2770469560486686e-05, "loss": 0.0082, "step": 62320 }, { "grad_norm": 0.09315891563892365, "learning_rate": 1.2756676160448956e-05, "loss": 0.0088, "step": 62330 }, { "grad_norm": 0.17491880059242249, "learning_rate": 1.2742889124276663e-05, "loss": 0.0092, "step": 62340 }, { "grad_norm": 0.15023764967918396, "learning_rate": 1.2729108454325639e-05, "loss": 0.0102, "step": 62350 }, { "grad_norm": 0.19113953411579132, "learning_rate": 1.2715334152950614e-05, "loss": 0.0105, "step": 62360 }, { "grad_norm": 0.07982853055000305, "learning_rate": 1.2701566222505246e-05, "loss": 0.0093, "step": 62370 }, { "grad_norm": 0.31488096714019775, "learning_rate": 1.2687804665342107e-05, "loss": 0.0099, "step": 62380 }, { "grad_norm": 0.12862837314605713, "learning_rate": 1.267404948381265e-05, "loss": 0.0078, "step": 62390 }, { "grad_norm": 0.19655698537826538, "learning_rate": 1.2660300680267267e-05, "loss": 0.0051, "step": 62400 }, { "grad_norm": 0.22686120867729187, "learning_rate": 1.2646558257055257e-05, "loss": 0.0075, "step": 62410 }, { "grad_norm": 0.26777413487434387, "learning_rate": 1.263282221652482e-05, "loss": 0.0076, "step": 62420 }, { "grad_norm": 0.15676796436309814, "learning_rate": 1.2619092561023088e-05, "loss": 0.0112, "step": 62430 }, { "grad_norm": 0.20526361465454102, "learning_rate": 1.2605369292896036e-05, "loss": 0.0104, "step": 62440 }, { "grad_norm": 0.16031552851200104, "learning_rate": 1.2591652414488658e-05, "loss": 0.0081, "step": 62450 }, { "grad_norm": 0.16672781109809875, "learning_rate": 1.2577941928144732e-05, "loss": 0.0077, "step": 62460 }, { "grad_norm": 0.14441272616386414, "learning_rate": 1.256423783620706e-05, "loss": 0.0106, "step": 62470 }, { "grad_norm": 0.3290478587150574, "learning_rate": 1.2550540141017264e-05, "loss": 0.0068, "step": 62480 }, { "grad_norm": 0.12018175423145294, "learning_rate": 1.253684884491591e-05, "loss": 0.0065, "step": 62490 }, { "grad_norm": 0.31775471568107605, "learning_rate": 1.2523163950242483e-05, "loss": 0.0056, "step": 62500 }, { "grad_norm": 0.16027162969112396, "learning_rate": 1.2509485459335313e-05, "loss": 0.0077, "step": 62510 }, { "grad_norm": 0.1357097327709198, "learning_rate": 1.2495813374531739e-05, "loss": 0.0068, "step": 62520 }, { "grad_norm": 0.22990106046199799, "learning_rate": 1.2482147698167907e-05, "loss": 0.0068, "step": 62530 }, { "grad_norm": 0.15674430131912231, "learning_rate": 1.2468488432578911e-05, "loss": 0.0086, "step": 62540 }, { "grad_norm": 0.21190203726291656, "learning_rate": 1.2454835580098761e-05, "loss": 0.0082, "step": 62550 }, { "grad_norm": 0.0924692451953888, "learning_rate": 1.2441189143060338e-05, "loss": 0.008, "step": 62560 }, { "grad_norm": 0.16499193012714386, "learning_rate": 1.242754912379548e-05, "loss": 0.0086, "step": 62570 }, { "grad_norm": 0.11263678222894669, "learning_rate": 1.2413915524634844e-05, "loss": 0.0194, "step": 62580 }, { "grad_norm": 0.289419949054718, "learning_rate": 1.2400288347908073e-05, "loss": 0.0095, "step": 62590 }, { "grad_norm": 0.13397610187530518, "learning_rate": 1.2386667595943663e-05, "loss": 0.007, "step": 62600 }, { "grad_norm": 0.16797703504562378, "learning_rate": 1.2373053271069035e-05, "loss": 0.0062, "step": 62610 }, { "grad_norm": 0.20308487117290497, "learning_rate": 1.2359445375610501e-05, "loss": 0.0114, "step": 62620 }, { "grad_norm": 0.08432861417531967, "learning_rate": 1.2345843911893301e-05, "loss": 0.0061, "step": 62630 }, { "grad_norm": 0.2860017418861389, "learning_rate": 1.2332248882241498e-05, "loss": 0.0174, "step": 62640 }, { "grad_norm": 0.2092975229024887, "learning_rate": 1.2318660288978178e-05, "loss": 0.0067, "step": 62650 }, { "grad_norm": 0.2098037153482437, "learning_rate": 1.2305078134425213e-05, "loss": 0.0064, "step": 62660 }, { "grad_norm": 0.09056627005338669, "learning_rate": 1.2291502420903434e-05, "loss": 0.0098, "step": 62670 }, { "grad_norm": 0.1327306479215622, "learning_rate": 1.2277933150732567e-05, "loss": 0.01, "step": 62680 }, { "grad_norm": 0.0735832005739212, "learning_rate": 1.2264370326231216e-05, "loss": 0.0077, "step": 62690 }, { "grad_norm": 0.11298537999391556, "learning_rate": 1.2250813949716927e-05, "loss": 0.0074, "step": 62700 }, { "grad_norm": 0.08131399750709534, "learning_rate": 1.2237264023506063e-05, "loss": 0.0086, "step": 62710 }, { "grad_norm": 0.15224780142307281, "learning_rate": 1.2223720549913987e-05, "loss": 0.0073, "step": 62720 }, { "grad_norm": 0.11897214502096176, "learning_rate": 1.221018353125487e-05, "loss": 0.0068, "step": 62730 }, { "grad_norm": 0.15735207498073578, "learning_rate": 1.2196652969841837e-05, "loss": 0.0087, "step": 62740 }, { "grad_norm": 0.22516542673110962, "learning_rate": 1.2183128867986904e-05, "loss": 0.0137, "step": 62750 }, { "grad_norm": 0.09716806560754776, "learning_rate": 1.2169611228000927e-05, "loss": 0.0084, "step": 62760 }, { "grad_norm": 0.13071638345718384, "learning_rate": 1.2156100052193752e-05, "loss": 0.0081, "step": 62770 }, { "grad_norm": 0.1233709529042244, "learning_rate": 1.214259534287403e-05, "loss": 0.0064, "step": 62780 }, { "grad_norm": 0.08942647278308868, "learning_rate": 1.2129097102349363e-05, "loss": 0.0068, "step": 62790 }, { "grad_norm": 0.16685043275356293, "learning_rate": 1.2115605332926227e-05, "loss": 0.0076, "step": 62800 }, { "grad_norm": 0.1531713753938675, "learning_rate": 1.2102120036910003e-05, "loss": 0.018, "step": 62810 }, { "grad_norm": 0.10132788866758347, "learning_rate": 1.2088641216604956e-05, "loss": 0.0095, "step": 62820 }, { "grad_norm": 0.11435941606760025, "learning_rate": 1.2075168874314264e-05, "loss": 0.0074, "step": 62830 }, { "grad_norm": 0.1075182557106018, "learning_rate": 1.2061703012339942e-05, "loss": 0.0065, "step": 62840 }, { "grad_norm": 0.2807382047176361, "learning_rate": 1.204824363298297e-05, "loss": 0.0107, "step": 62850 }, { "grad_norm": 0.20327942073345184, "learning_rate": 1.2034790738543173e-05, "loss": 0.0094, "step": 62860 }, { "grad_norm": 0.13238073885440826, "learning_rate": 1.2021344331319284e-05, "loss": 0.0071, "step": 62870 }, { "grad_norm": 0.2924475371837616, "learning_rate": 1.200790441360894e-05, "loss": 0.011, "step": 62880 }, { "grad_norm": 0.14701217412948608, "learning_rate": 1.199447098770864e-05, "loss": 0.0067, "step": 62890 }, { "grad_norm": 0.11939375847578049, "learning_rate": 1.198104405591381e-05, "loss": 0.0064, "step": 62900 }, { "grad_norm": 0.09156185388565063, "learning_rate": 1.1967623620518697e-05, "loss": 0.0066, "step": 62910 }, { "grad_norm": 0.42460423707962036, "learning_rate": 1.1954209683816554e-05, "loss": 0.0106, "step": 62920 }, { "grad_norm": 0.1415310651063919, "learning_rate": 1.1940802248099402e-05, "loss": 0.0109, "step": 62930 }, { "grad_norm": 0.1042453944683075, "learning_rate": 1.1927401315658232e-05, "loss": 0.0082, "step": 62940 }, { "grad_norm": 0.26134294271469116, "learning_rate": 1.1914006888782898e-05, "loss": 0.0086, "step": 62950 }, { "grad_norm": 0.22477447986602783, "learning_rate": 1.1900618969762106e-05, "loss": 0.0083, "step": 62960 }, { "grad_norm": 0.15228472650051117, "learning_rate": 1.188723756088354e-05, "loss": 0.007, "step": 62970 }, { "grad_norm": 0.11606626957654953, "learning_rate": 1.1873862664433672e-05, "loss": 0.0068, "step": 62980 }, { "grad_norm": 0.2014511376619339, "learning_rate": 1.1860494282697927e-05, "loss": 0.0076, "step": 62990 }, { "grad_norm": 0.24056021869182587, "learning_rate": 1.1847132417960588e-05, "loss": 0.0063, "step": 63000 }, { "grad_norm": 0.11915261298418045, "learning_rate": 1.1833777072504832e-05, "loss": 0.0099, "step": 63010 }, { "grad_norm": 0.14013069868087769, "learning_rate": 1.1820428248612731e-05, "loss": 0.0074, "step": 63020 }, { "grad_norm": 0.30640819668769836, "learning_rate": 1.1807085948565245e-05, "loss": 0.011, "step": 63030 }, { "grad_norm": 0.06184164807200432, "learning_rate": 1.1793750174642172e-05, "loss": 0.0079, "step": 63040 }, { "grad_norm": 0.1764453500509262, "learning_rate": 1.1780420929122254e-05, "loss": 0.011, "step": 63050 }, { "grad_norm": 0.1796666383743286, "learning_rate": 1.176709821428309e-05, "loss": 0.0074, "step": 63060 }, { "grad_norm": 0.1243012398481369, "learning_rate": 1.1753782032401173e-05, "loss": 0.0109, "step": 63070 }, { "grad_norm": 0.28641512989997864, "learning_rate": 1.1740472385751866e-05, "loss": 0.01, "step": 63080 }, { "grad_norm": 0.1368279904127121, "learning_rate": 1.172716927660943e-05, "loss": 0.0067, "step": 63090 }, { "grad_norm": 0.17299138009548187, "learning_rate": 1.1713872707247015e-05, "loss": 0.0097, "step": 63100 }, { "grad_norm": 0.17105485498905182, "learning_rate": 1.1700582679936595e-05, "loss": 0.0062, "step": 63110 }, { "grad_norm": 0.1789064258337021, "learning_rate": 1.1687299196949136e-05, "loss": 0.0155, "step": 63120 }, { "grad_norm": 0.19677039980888367, "learning_rate": 1.1674022260554374e-05, "loss": 0.0065, "step": 63130 }, { "grad_norm": 0.1354704201221466, "learning_rate": 1.1660751873020987e-05, "loss": 0.0071, "step": 63140 }, { "grad_norm": 0.21653325855731964, "learning_rate": 1.1647488036616538e-05, "loss": 0.0073, "step": 63150 }, { "grad_norm": 0.13601751625537872, "learning_rate": 1.1634230753607417e-05, "loss": 0.0123, "step": 63160 }, { "grad_norm": 0.15678836405277252, "learning_rate": 1.1620980026258982e-05, "loss": 0.0091, "step": 63170 }, { "grad_norm": 0.13746559619903564, "learning_rate": 1.1607735856835373e-05, "loss": 0.0063, "step": 63180 }, { "grad_norm": 0.10291440039873123, "learning_rate": 1.1594498247599677e-05, "loss": 0.007, "step": 63190 }, { "grad_norm": 0.2120392918586731, "learning_rate": 1.158126720081384e-05, "loss": 0.0119, "step": 63200 }, { "grad_norm": 0.23482121527194977, "learning_rate": 1.156804271873868e-05, "loss": 0.0059, "step": 63210 }, { "grad_norm": 0.16179275512695312, "learning_rate": 1.1554824803633924e-05, "loss": 0.0095, "step": 63220 }, { "grad_norm": 0.24344561994075775, "learning_rate": 1.154161345775811e-05, "loss": 0.0062, "step": 63230 }, { "grad_norm": 0.1627369374036789, "learning_rate": 1.1528408683368724e-05, "loss": 0.0086, "step": 63240 }, { "grad_norm": 0.3242795765399933, "learning_rate": 1.1515210482722088e-05, "loss": 0.0138, "step": 63250 }, { "grad_norm": 0.22449837625026703, "learning_rate": 1.150201885807342e-05, "loss": 0.0132, "step": 63260 }, { "grad_norm": 0.15899279713630676, "learning_rate": 1.1488833811676807e-05, "loss": 0.0071, "step": 63270 }, { "grad_norm": 0.1533692628145218, "learning_rate": 1.1475655345785213e-05, "loss": 0.0085, "step": 63280 }, { "grad_norm": 0.1448161005973816, "learning_rate": 1.1462483462650481e-05, "loss": 0.0062, "step": 63290 }, { "grad_norm": 0.1470293402671814, "learning_rate": 1.1449318164523331e-05, "loss": 0.0084, "step": 63300 }, { "grad_norm": 0.17978143692016602, "learning_rate": 1.1436159453653334e-05, "loss": 0.0058, "step": 63310 }, { "grad_norm": 0.14172200858592987, "learning_rate": 1.1423007332288955e-05, "loss": 0.0129, "step": 63320 }, { "grad_norm": 0.249791219830513, "learning_rate": 1.1409861802677546e-05, "loss": 0.0071, "step": 63330 }, { "grad_norm": 0.20099078118801117, "learning_rate": 1.1396722867065313e-05, "loss": 0.0078, "step": 63340 }, { "grad_norm": 0.20086757838726044, "learning_rate": 1.1383590527697352e-05, "loss": 0.0126, "step": 63350 }, { "grad_norm": 0.10243923217058182, "learning_rate": 1.137046478681758e-05, "loss": 0.0064, "step": 63360 }, { "grad_norm": 0.42611244320869446, "learning_rate": 1.1357345646668888e-05, "loss": 0.0226, "step": 63370 }, { "grad_norm": 0.22188317775726318, "learning_rate": 1.1344233109492924e-05, "loss": 0.0071, "step": 63380 }, { "grad_norm": 0.11401139944791794, "learning_rate": 1.1331127177530292e-05, "loss": 0.006, "step": 63390 }, { "grad_norm": 0.08165103942155838, "learning_rate": 1.1318027853020441e-05, "loss": 0.0045, "step": 63400 }, { "grad_norm": 0.12337684631347656, "learning_rate": 1.1304935138201645e-05, "loss": 0.0067, "step": 63410 }, { "grad_norm": 0.17308831214904785, "learning_rate": 1.1291849035311153e-05, "loss": 0.0087, "step": 63420 }, { "grad_norm": 0.1504172831773758, "learning_rate": 1.1278769546584972e-05, "loss": 0.0095, "step": 63430 }, { "grad_norm": 0.1318894773721695, "learning_rate": 1.1265696674258052e-05, "loss": 0.0068, "step": 63440 }, { "grad_norm": 0.21582183241844177, "learning_rate": 1.1252630420564186e-05, "loss": 0.0069, "step": 63450 }, { "grad_norm": 0.20924973487854004, "learning_rate": 1.1239570787736036e-05, "loss": 0.0109, "step": 63460 }, { "grad_norm": 0.12263736873865128, "learning_rate": 1.1226517778005135e-05, "loss": 0.0081, "step": 63470 }, { "grad_norm": 0.09543255716562271, "learning_rate": 1.1213471393601893e-05, "loss": 0.0054, "step": 63480 }, { "grad_norm": 0.13076528906822205, "learning_rate": 1.1200431636755587e-05, "loss": 0.0128, "step": 63490 }, { "grad_norm": 0.25913023948669434, "learning_rate": 1.1187398509694336e-05, "loss": 0.0128, "step": 63500 }, { "grad_norm": 0.17785444855690002, "learning_rate": 1.1174372014645146e-05, "loss": 0.0058, "step": 63510 }, { "grad_norm": 0.15591372549533844, "learning_rate": 1.1161352153833899e-05, "loss": 0.0168, "step": 63520 }, { "grad_norm": 0.1461789906024933, "learning_rate": 1.1148338929485325e-05, "loss": 0.0095, "step": 63530 }, { "grad_norm": 0.07651577144861221, "learning_rate": 1.113533234382304e-05, "loss": 0.0048, "step": 63540 }, { "grad_norm": 0.09466195851564407, "learning_rate": 1.1122332399069513e-05, "loss": 0.007, "step": 63550 }, { "grad_norm": 0.2433350533246994, "learning_rate": 1.1109339097446047e-05, "loss": 0.0094, "step": 63560 }, { "grad_norm": 0.19912776350975037, "learning_rate": 1.1096352441172897e-05, "loss": 0.0167, "step": 63570 }, { "grad_norm": 0.21227481961250305, "learning_rate": 1.1083372432469086e-05, "loss": 0.0056, "step": 63580 }, { "grad_norm": 0.11014621704816818, "learning_rate": 1.107039907355255e-05, "loss": 0.0118, "step": 63590 }, { "grad_norm": 0.1593308448791504, "learning_rate": 1.1057432366640103e-05, "loss": 0.0075, "step": 63600 }, { "grad_norm": 0.5338256359100342, "learning_rate": 1.1044472313947352e-05, "loss": 0.0123, "step": 63610 }, { "grad_norm": 0.11939886957406998, "learning_rate": 1.1031518917688877e-05, "loss": 0.009, "step": 63620 }, { "grad_norm": 0.10048642754554749, "learning_rate": 1.1018572180078007e-05, "loss": 0.0164, "step": 63630 }, { "grad_norm": 0.1432473510503769, "learning_rate": 1.1005632103327018e-05, "loss": 0.0069, "step": 63640 }, { "grad_norm": 0.16836415231227875, "learning_rate": 1.0992698689646996e-05, "loss": 0.0088, "step": 63650 }, { "grad_norm": 0.33235785365104675, "learning_rate": 1.0979771941247919e-05, "loss": 0.0139, "step": 63660 }, { "grad_norm": 0.1720137596130371, "learning_rate": 1.0966851860338611e-05, "loss": 0.0077, "step": 63670 }, { "grad_norm": 0.15265870094299316, "learning_rate": 1.0953938449126766e-05, "loss": 0.0069, "step": 63680 }, { "grad_norm": 0.3906601369380951, "learning_rate": 1.0941031709818933e-05, "loss": 0.0192, "step": 63690 }, { "grad_norm": 0.12495177984237671, "learning_rate": 1.0928131644620509e-05, "loss": 0.0178, "step": 63700 }, { "grad_norm": 0.2059883326292038, "learning_rate": 1.0915238255735766e-05, "loss": 0.0099, "step": 63710 }, { "grad_norm": 0.1370978057384491, "learning_rate": 1.0902351545367833e-05, "loss": 0.007, "step": 63720 }, { "grad_norm": 0.22688771784305573, "learning_rate": 1.0889471515718702e-05, "loss": 0.0068, "step": 63730 }, { "grad_norm": 0.10049678385257721, "learning_rate": 1.087659816898921e-05, "loss": 0.0047, "step": 63740 }, { "grad_norm": 0.15738768875598907, "learning_rate": 1.0863731507379082e-05, "loss": 0.0064, "step": 63750 }, { "grad_norm": 0.243474081158638, "learning_rate": 1.0850871533086827e-05, "loss": 0.0097, "step": 63760 }, { "grad_norm": 0.22348706424236298, "learning_rate": 1.0838018248309927e-05, "loss": 0.008, "step": 63770 }, { "grad_norm": 0.24305415153503418, "learning_rate": 1.0825171655244615e-05, "loss": 0.0111, "step": 63780 }, { "grad_norm": 0.13268402218818665, "learning_rate": 1.0812331756086025e-05, "loss": 0.0057, "step": 63790 }, { "grad_norm": 0.2703007757663727, "learning_rate": 1.079949855302817e-05, "loss": 0.0107, "step": 63800 }, { "grad_norm": 0.2297995686531067, "learning_rate": 1.0786672048263852e-05, "loss": 0.0112, "step": 63810 }, { "grad_norm": 0.0999022126197815, "learning_rate": 1.0773852243984817e-05, "loss": 0.0105, "step": 63820 }, { "grad_norm": 0.24330264329910278, "learning_rate": 1.0761039142381586e-05, "loss": 0.0163, "step": 63830 }, { "grad_norm": 0.18707874417304993, "learning_rate": 1.0748232745643577e-05, "loss": 0.0108, "step": 63840 }, { "grad_norm": 0.15148209035396576, "learning_rate": 1.0735433055959055e-05, "loss": 0.0114, "step": 63850 }, { "grad_norm": 0.3601014316082001, "learning_rate": 1.0722640075515133e-05, "loss": 0.0072, "step": 63860 }, { "grad_norm": 0.11437700688838959, "learning_rate": 1.0709853806497795e-05, "loss": 0.0055, "step": 63870 }, { "grad_norm": 0.10065323859453201, "learning_rate": 1.0697074251091831e-05, "loss": 0.0073, "step": 63880 }, { "grad_norm": 0.3100475072860718, "learning_rate": 1.0684301411480962e-05, "loss": 0.0066, "step": 63890 }, { "grad_norm": 0.13555319607257843, "learning_rate": 1.067153528984769e-05, "loss": 0.0053, "step": 63900 }, { "grad_norm": 0.09766841679811478, "learning_rate": 1.0658775888373395e-05, "loss": 0.0078, "step": 63910 }, { "grad_norm": 0.14843332767486572, "learning_rate": 1.0646023209238314e-05, "loss": 0.0117, "step": 63920 }, { "grad_norm": 0.11176733672618866, "learning_rate": 1.0633277254621537e-05, "loss": 0.007, "step": 63930 }, { "grad_norm": 0.11391901969909668, "learning_rate": 1.0620538026700994e-05, "loss": 0.007, "step": 63940 }, { "grad_norm": 0.1709270030260086, "learning_rate": 1.0607805527653486e-05, "loss": 0.0077, "step": 63950 }, { "grad_norm": 0.13700400292873383, "learning_rate": 1.059507975965462e-05, "loss": 0.007, "step": 63960 }, { "grad_norm": 0.09071321040391922, "learning_rate": 1.0582360724878898e-05, "loss": 0.0051, "step": 63970 }, { "grad_norm": 0.2025415003299713, "learning_rate": 1.0569648425499651e-05, "loss": 0.0087, "step": 63980 }, { "grad_norm": 0.09609650075435638, "learning_rate": 1.0556942863689063e-05, "loss": 0.0076, "step": 63990 }, { "grad_norm": 0.1447419673204422, "learning_rate": 1.054424404161819e-05, "loss": 0.0059, "step": 64000 }, { "grad_norm": 0.15368789434432983, "learning_rate": 1.053155196145686e-05, "loss": 0.0084, "step": 64010 }, { "grad_norm": 0.2070779800415039, "learning_rate": 1.0518866625373863e-05, "loss": 0.0065, "step": 64020 }, { "grad_norm": 0.0658968910574913, "learning_rate": 1.0506188035536735e-05, "loss": 0.0061, "step": 64030 }, { "grad_norm": 0.21535134315490723, "learning_rate": 1.0493516194111919e-05, "loss": 0.0078, "step": 64040 }, { "grad_norm": 0.12405633181333542, "learning_rate": 1.0480851103264688e-05, "loss": 0.0049, "step": 64050 }, { "grad_norm": 0.08859313279390335, "learning_rate": 1.046819276515913e-05, "loss": 0.0053, "step": 64060 }, { "grad_norm": 0.1126248762011528, "learning_rate": 1.0455541181958256e-05, "loss": 0.0066, "step": 64070 }, { "grad_norm": 0.13050197064876556, "learning_rate": 1.0442896355823822e-05, "loss": 0.0071, "step": 64080 }, { "grad_norm": 0.15244248509407043, "learning_rate": 1.0430258288916539e-05, "loss": 0.0065, "step": 64090 }, { "grad_norm": 0.12444748729467392, "learning_rate": 1.0417626983395868e-05, "loss": 0.0075, "step": 64100 }, { "grad_norm": 0.1995815485715866, "learning_rate": 1.0405002441420165e-05, "loss": 0.0063, "step": 64110 }, { "grad_norm": 0.14640583097934723, "learning_rate": 1.039238466514662e-05, "loss": 0.0074, "step": 64120 }, { "grad_norm": 0.17270581424236298, "learning_rate": 1.0379773656731262e-05, "loss": 0.0117, "step": 64130 }, { "grad_norm": 0.18940210342407227, "learning_rate": 1.0367169418328986e-05, "loss": 0.0071, "step": 64140 }, { "grad_norm": 0.3357461392879486, "learning_rate": 1.0354571952093484e-05, "loss": 0.0116, "step": 64150 }, { "grad_norm": 0.1751074492931366, "learning_rate": 1.034198126017733e-05, "loss": 0.0069, "step": 64160 }, { "grad_norm": 0.20507024228572845, "learning_rate": 1.0329397344731928e-05, "loss": 0.0089, "step": 64170 }, { "grad_norm": 0.14693452417850494, "learning_rate": 1.0316820207907524e-05, "loss": 0.0097, "step": 64180 }, { "grad_norm": 0.17269253730773926, "learning_rate": 1.0304249851853215e-05, "loss": 0.008, "step": 64190 }, { "grad_norm": 0.1944451481103897, "learning_rate": 1.0291686278716933e-05, "loss": 0.0065, "step": 64200 }, { "grad_norm": 0.14557693898677826, "learning_rate": 1.0279129490645418e-05, "loss": 0.0084, "step": 64210 }, { "grad_norm": 0.135201558470726, "learning_rate": 1.0266579489784328e-05, "loss": 0.0069, "step": 64220 }, { "grad_norm": 0.13577939569950104, "learning_rate": 1.0254036278278084e-05, "loss": 0.0081, "step": 64230 }, { "grad_norm": 0.10856785625219345, "learning_rate": 1.0241499858269982e-05, "loss": 0.0052, "step": 64240 }, { "grad_norm": 0.47951793670654297, "learning_rate": 1.0228970231902169e-05, "loss": 0.0062, "step": 64250 }, { "grad_norm": 0.182006374001503, "learning_rate": 1.0216447401315582e-05, "loss": 0.0073, "step": 64260 }, { "grad_norm": 0.17705830931663513, "learning_rate": 1.020393136865007e-05, "loss": 0.0053, "step": 64270 }, { "grad_norm": 0.25114089250564575, "learning_rate": 1.0191422136044242e-05, "loss": 0.0173, "step": 64280 }, { "grad_norm": 0.3317408561706543, "learning_rate": 1.017891970563563e-05, "loss": 0.0073, "step": 64290 }, { "grad_norm": 0.14142324030399323, "learning_rate": 1.0166424079560516e-05, "loss": 0.0121, "step": 64300 }, { "grad_norm": 0.1822972148656845, "learning_rate": 1.0153935259954078e-05, "loss": 0.0066, "step": 64310 }, { "grad_norm": 0.38274070620536804, "learning_rate": 1.0141453248950311e-05, "loss": 0.0099, "step": 64320 }, { "grad_norm": 0.1386435329914093, "learning_rate": 1.0128978048682054e-05, "loss": 0.0096, "step": 64330 }, { "grad_norm": 0.28311580419540405, "learning_rate": 1.0116509661280982e-05, "loss": 0.0096, "step": 64340 }, { "grad_norm": 0.1409751921892166, "learning_rate": 1.0104048088877576e-05, "loss": 0.0059, "step": 64350 }, { "grad_norm": 0.14618609845638275, "learning_rate": 1.0091593333601201e-05, "loss": 0.0082, "step": 64360 }, { "grad_norm": 0.1429850310087204, "learning_rate": 1.0079145397580031e-05, "loss": 0.0074, "step": 64370 }, { "grad_norm": 0.3113709092140198, "learning_rate": 1.006670428294107e-05, "loss": 0.0136, "step": 64380 }, { "grad_norm": 0.1414756029844284, "learning_rate": 1.0054269991810166e-05, "loss": 0.0061, "step": 64390 }, { "grad_norm": 0.14963221549987793, "learning_rate": 1.0041842526312024e-05, "loss": 0.006, "step": 64400 }, { "grad_norm": 0.18986564874649048, "learning_rate": 1.0029421888570101e-05, "loss": 0.0058, "step": 64410 }, { "grad_norm": 0.10886362940073013, "learning_rate": 1.0017008080706813e-05, "loss": 0.0177, "step": 64420 }, { "grad_norm": 0.08184482157230377, "learning_rate": 1.0004601104843287e-05, "loss": 0.0054, "step": 64430 }, { "grad_norm": 0.1535494476556778, "learning_rate": 9.992200963099562e-06, "loss": 0.0071, "step": 64440 }, { "grad_norm": 0.14621151983737946, "learning_rate": 9.979807657594486e-06, "loss": 0.006, "step": 64450 }, { "grad_norm": 0.19188325107097626, "learning_rate": 9.967421190445703e-06, "loss": 0.0071, "step": 64460 }, { "grad_norm": 0.18299199640750885, "learning_rate": 9.955041563769769e-06, "loss": 0.009, "step": 64470 }, { "grad_norm": 0.06975886970758438, "learning_rate": 9.942668779681974e-06, "loss": 0.0121, "step": 64480 }, { "grad_norm": 0.536794900894165, "learning_rate": 9.930302840296541e-06, "loss": 0.0083, "step": 64490 }, { "grad_norm": 0.12537124752998352, "learning_rate": 9.917943747726426e-06, "loss": 0.0055, "step": 64500 }, { "grad_norm": 0.22080522775650024, "learning_rate": 9.905591504083484e-06, "loss": 0.0071, "step": 64510 }, { "grad_norm": 0.14018265902996063, "learning_rate": 9.893246111478382e-06, "loss": 0.0063, "step": 64520 }, { "grad_norm": 0.2174081951379776, "learning_rate": 9.88090757202057e-06, "loss": 0.0071, "step": 64530 }, { "grad_norm": 0.12262458354234695, "learning_rate": 9.868575887818421e-06, "loss": 0.0114, "step": 64540 }, { "grad_norm": 0.14832833409309387, "learning_rate": 9.856251060979044e-06, "loss": 0.0084, "step": 64550 }, { "grad_norm": 0.15642324090003967, "learning_rate": 9.843933093608426e-06, "loss": 0.011, "step": 64560 }, { "grad_norm": 0.1719776690006256, "learning_rate": 9.831621987811368e-06, "loss": 0.0122, "step": 64570 }, { "grad_norm": 0.1015063226222992, "learning_rate": 9.819317745691509e-06, "loss": 0.0051, "step": 64580 }, { "grad_norm": 0.19151604175567627, "learning_rate": 9.8070203693513e-06, "loss": 0.0076, "step": 64590 }, { "grad_norm": 0.12393911182880402, "learning_rate": 9.794729860892048e-06, "loss": 0.0067, "step": 64600 }, { "grad_norm": 0.1949271559715271, "learning_rate": 9.782446222413827e-06, "loss": 0.0173, "step": 64610 }, { "grad_norm": 0.0748576745390892, "learning_rate": 9.770169456015598e-06, "loss": 0.0136, "step": 64620 }, { "grad_norm": 0.07760438323020935, "learning_rate": 9.75789956379512e-06, "loss": 0.0096, "step": 64630 }, { "grad_norm": 0.13363583385944366, "learning_rate": 9.74563654784898e-06, "loss": 0.0066, "step": 64640 }, { "grad_norm": 0.11282563209533691, "learning_rate": 9.733380410272596e-06, "loss": 0.0077, "step": 64650 }, { "grad_norm": 0.12967634201049805, "learning_rate": 9.721131153160207e-06, "loss": 0.0073, "step": 64660 }, { "grad_norm": 0.16489940881729126, "learning_rate": 9.708888778604886e-06, "loss": 0.0066, "step": 64670 }, { "grad_norm": 0.09972712397575378, "learning_rate": 9.696653288698488e-06, "loss": 0.0066, "step": 64680 }, { "grad_norm": 0.2025763988494873, "learning_rate": 9.684424685531763e-06, "loss": 0.0152, "step": 64690 }, { "grad_norm": 0.19875915348529816, "learning_rate": 9.672202971194216e-06, "loss": 0.0108, "step": 64700 }, { "grad_norm": 0.08757370710372925, "learning_rate": 9.659988147774213e-06, "loss": 0.0091, "step": 64710 }, { "grad_norm": 0.10416403412818909, "learning_rate": 9.647780217358942e-06, "loss": 0.0085, "step": 64720 }, { "grad_norm": 0.1048148050904274, "learning_rate": 9.635579182034376e-06, "loss": 0.0071, "step": 64730 }, { "grad_norm": 0.10827379673719406, "learning_rate": 9.623385043885386e-06, "loss": 0.0142, "step": 64740 }, { "grad_norm": 0.18430602550506592, "learning_rate": 9.61119780499557e-06, "loss": 0.0052, "step": 64750 }, { "grad_norm": 0.13728725910186768, "learning_rate": 9.599017467447418e-06, "loss": 0.007, "step": 64760 }, { "grad_norm": 0.19875389337539673, "learning_rate": 9.586844033322206e-06, "loss": 0.0073, "step": 64770 }, { "grad_norm": 0.36730465292930603, "learning_rate": 9.574677504700052e-06, "loss": 0.012, "step": 64780 }, { "grad_norm": 0.14297005534172058, "learning_rate": 9.562517883659877e-06, "loss": 0.0179, "step": 64790 }, { "grad_norm": 0.101052425801754, "learning_rate": 9.55036517227944e-06, "loss": 0.0061, "step": 64800 }, { "grad_norm": 0.21120700240135193, "learning_rate": 9.538219372635282e-06, "loss": 0.0075, "step": 64810 }, { "grad_norm": 0.1106119230389595, "learning_rate": 9.526080486802802e-06, "loss": 0.0082, "step": 64820 }, { "grad_norm": 0.11015824228525162, "learning_rate": 9.513948516856203e-06, "loss": 0.0093, "step": 64830 }, { "grad_norm": 0.1911003589630127, "learning_rate": 9.501823464868504e-06, "loss": 0.0062, "step": 64840 }, { "grad_norm": 0.1866876482963562, "learning_rate": 9.489705332911547e-06, "loss": 0.0054, "step": 64850 }, { "grad_norm": 0.1390742063522339, "learning_rate": 9.477594123055994e-06, "loss": 0.0084, "step": 64860 }, { "grad_norm": 0.1349690556526184, "learning_rate": 9.465489837371321e-06, "loss": 0.013, "step": 64870 }, { "grad_norm": 0.20166480541229248, "learning_rate": 9.453392477925794e-06, "loss": 0.006, "step": 64880 }, { "grad_norm": 0.11376860737800598, "learning_rate": 9.441302046786566e-06, "loss": 0.0078, "step": 64890 }, { "grad_norm": 0.09128717333078384, "learning_rate": 9.429218546019519e-06, "loss": 0.0055, "step": 64900 }, { "grad_norm": 0.25747889280319214, "learning_rate": 9.41714197768941e-06, "loss": 0.0141, "step": 64910 }, { "grad_norm": 0.23316353559494019, "learning_rate": 9.405072343859805e-06, "loss": 0.0083, "step": 64920 }, { "grad_norm": 0.18858617544174194, "learning_rate": 9.393009646593043e-06, "loss": 0.0066, "step": 64930 }, { "grad_norm": 0.16936594247817993, "learning_rate": 9.38095388795035e-06, "loss": 0.0101, "step": 64940 }, { "grad_norm": 0.27780452370643616, "learning_rate": 9.36890506999169e-06, "loss": 0.008, "step": 64950 }, { "grad_norm": 0.1790178120136261, "learning_rate": 9.356863194775894e-06, "loss": 0.0086, "step": 64960 }, { "grad_norm": 0.27104905247688293, "learning_rate": 9.344828264360583e-06, "loss": 0.0106, "step": 64970 }, { "grad_norm": 0.17435239255428314, "learning_rate": 9.332800280802201e-06, "loss": 0.0096, "step": 64980 }, { "grad_norm": 0.13387860357761383, "learning_rate": 9.32077924615602e-06, "loss": 0.0055, "step": 64990 }, { "grad_norm": 0.12251120805740356, "learning_rate": 9.308765162476063e-06, "loss": 0.0064, "step": 65000 }, { "grad_norm": 0.14045651257038116, "learning_rate": 9.296758031815239e-06, "loss": 0.0088, "step": 65010 }, { "grad_norm": 0.1781095415353775, "learning_rate": 9.284757856225229e-06, "loss": 0.0066, "step": 65020 }, { "grad_norm": 0.30808740854263306, "learning_rate": 9.272764637756538e-06, "loss": 0.0069, "step": 65030 }, { "grad_norm": 0.17758317291736603, "learning_rate": 9.260778378458479e-06, "loss": 0.01, "step": 65040 }, { "grad_norm": 0.10537897050380707, "learning_rate": 9.248799080379172e-06, "loss": 0.0078, "step": 65050 }, { "grad_norm": 0.1629696935415268, "learning_rate": 9.236826745565558e-06, "loss": 0.0078, "step": 65060 }, { "grad_norm": 0.20081622898578644, "learning_rate": 9.224861376063388e-06, "loss": 0.0072, "step": 65070 }, { "grad_norm": 0.0977916568517685, "learning_rate": 9.212902973917192e-06, "loss": 0.0041, "step": 65080 }, { "grad_norm": 0.08857109397649765, "learning_rate": 9.20095154117035e-06, "loss": 0.0056, "step": 65090 }, { "grad_norm": 0.17728960514068604, "learning_rate": 9.189007079865036e-06, "loss": 0.0054, "step": 65100 }, { "grad_norm": 0.33211976289749146, "learning_rate": 9.177069592042226e-06, "loss": 0.0073, "step": 65110 }, { "grad_norm": 0.2030588686466217, "learning_rate": 9.165139079741724e-06, "loss": 0.0093, "step": 65120 }, { "grad_norm": 0.13985362648963928, "learning_rate": 9.153215545002098e-06, "loss": 0.0138, "step": 65130 }, { "grad_norm": 0.18312110006809235, "learning_rate": 9.141298989860798e-06, "loss": 0.0169, "step": 65140 }, { "grad_norm": 0.09367120265960693, "learning_rate": 9.129389416353994e-06, "loss": 0.0065, "step": 65150 }, { "grad_norm": 0.19112038612365723, "learning_rate": 9.11748682651673e-06, "loss": 0.0054, "step": 65160 }, { "grad_norm": 0.17243161797523499, "learning_rate": 9.105591222382837e-06, "loss": 0.0077, "step": 65170 }, { "grad_norm": 0.10173477977514267, "learning_rate": 9.093702605984915e-06, "loss": 0.0056, "step": 65180 }, { "grad_norm": 0.19259558618068695, "learning_rate": 9.081820979354455e-06, "loss": 0.0099, "step": 65190 }, { "grad_norm": 0.24181267619132996, "learning_rate": 9.069946344521663e-06, "loss": 0.0084, "step": 65200 }, { "grad_norm": 0.1294994056224823, "learning_rate": 9.058078703515598e-06, "loss": 0.0077, "step": 65210 }, { "grad_norm": 0.21009039878845215, "learning_rate": 9.046218058364125e-06, "loss": 0.0134, "step": 65220 }, { "grad_norm": 0.1577986776828766, "learning_rate": 9.034364411093893e-06, "loss": 0.0082, "step": 65230 }, { "grad_norm": 0.2934344708919525, "learning_rate": 9.022517763730371e-06, "loss": 0.0093, "step": 65240 }, { "grad_norm": 0.11367896944284439, "learning_rate": 9.010678118297827e-06, "loss": 0.0054, "step": 65250 }, { "grad_norm": 0.2529325485229492, "learning_rate": 8.998845476819345e-06, "loss": 0.0101, "step": 65260 }, { "grad_norm": 0.17732778191566467, "learning_rate": 8.987019841316773e-06, "loss": 0.0068, "step": 65270 }, { "grad_norm": 0.18375635147094727, "learning_rate": 8.975201213810802e-06, "loss": 0.005, "step": 65280 }, { "grad_norm": 0.1518455147743225, "learning_rate": 8.963389596320915e-06, "loss": 0.0078, "step": 65290 }, { "grad_norm": 0.12325501441955566, "learning_rate": 8.951584990865391e-06, "loss": 0.0069, "step": 65300 }, { "grad_norm": 0.26300519704818726, "learning_rate": 8.939787399461319e-06, "loss": 0.0135, "step": 65310 }, { "grad_norm": 0.2745019793510437, "learning_rate": 8.927996824124591e-06, "loss": 0.0127, "step": 65320 }, { "grad_norm": 0.3048096299171448, "learning_rate": 8.916213266869854e-06, "loss": 0.0111, "step": 65330 }, { "grad_norm": 0.0885196402668953, "learning_rate": 8.904436729710658e-06, "loss": 0.0056, "step": 65340 }, { "grad_norm": 0.06728744506835938, "learning_rate": 8.892667214659245e-06, "loss": 0.005, "step": 65350 }, { "grad_norm": 0.14699533581733704, "learning_rate": 8.880904723726713e-06, "loss": 0.0064, "step": 65360 }, { "grad_norm": 0.21085599064826965, "learning_rate": 8.869149258922971e-06, "loss": 0.0051, "step": 65370 }, { "grad_norm": 0.05569514259696007, "learning_rate": 8.857400822256662e-06, "loss": 0.0045, "step": 65380 }, { "grad_norm": 0.09826978296041489, "learning_rate": 8.845659415735324e-06, "loss": 0.0151, "step": 65390 }, { "grad_norm": 0.12065640091896057, "learning_rate": 8.83392504136521e-06, "loss": 0.0069, "step": 65400 }, { "grad_norm": 0.11566030234098434, "learning_rate": 8.822197701151407e-06, "loss": 0.0094, "step": 65410 }, { "grad_norm": 0.20278163254261017, "learning_rate": 8.810477397097804e-06, "loss": 0.0083, "step": 65420 }, { "grad_norm": 0.1994723379611969, "learning_rate": 8.798764131207077e-06, "loss": 0.0073, "step": 65430 }, { "grad_norm": 0.1402653157711029, "learning_rate": 8.787057905480706e-06, "loss": 0.0086, "step": 65440 }, { "grad_norm": 0.12984265387058258, "learning_rate": 8.775358721918958e-06, "loss": 0.0064, "step": 65450 }, { "grad_norm": 0.10135593265295029, "learning_rate": 8.763666582520923e-06, "loss": 0.0084, "step": 65460 }, { "grad_norm": 0.08086635917425156, "learning_rate": 8.751981489284445e-06, "loss": 0.0057, "step": 65470 }, { "grad_norm": 0.20189033448696136, "learning_rate": 8.740303444206188e-06, "loss": 0.0084, "step": 65480 }, { "grad_norm": 0.1712365448474884, "learning_rate": 8.72863244928162e-06, "loss": 0.0059, "step": 65490 }, { "grad_norm": 0.1339159458875656, "learning_rate": 8.716968506504991e-06, "loss": 0.014, "step": 65500 }, { "grad_norm": 0.16655553877353668, "learning_rate": 8.70531161786935e-06, "loss": 0.0061, "step": 65510 }, { "grad_norm": 0.08164538443088531, "learning_rate": 8.693661785366558e-06, "loss": 0.0146, "step": 65520 }, { "grad_norm": 0.09777513891458511, "learning_rate": 8.682019010987208e-06, "loss": 0.0142, "step": 65530 }, { "grad_norm": 0.17277222871780396, "learning_rate": 8.670383296720786e-06, "loss": 0.006, "step": 65540 }, { "grad_norm": 0.2621573805809021, "learning_rate": 8.658754644555478e-06, "loss": 0.0076, "step": 65550 }, { "grad_norm": 0.0728544145822525, "learning_rate": 8.647133056478313e-06, "loss": 0.0084, "step": 65560 }, { "grad_norm": 0.18041540682315826, "learning_rate": 8.635518534475123e-06, "loss": 0.0072, "step": 65570 }, { "grad_norm": 0.12274531275033951, "learning_rate": 8.623911080530467e-06, "loss": 0.0092, "step": 65580 }, { "grad_norm": 0.12649057805538177, "learning_rate": 8.61231069662779e-06, "loss": 0.0064, "step": 65590 }, { "grad_norm": 0.0935673639178276, "learning_rate": 8.600717384749252e-06, "loss": 0.0065, "step": 65600 }, { "grad_norm": 0.14305561780929565, "learning_rate": 8.58913114687584e-06, "loss": 0.0067, "step": 65610 }, { "grad_norm": 0.3327738344669342, "learning_rate": 8.57755198498732e-06, "loss": 0.0118, "step": 65620 }, { "grad_norm": 0.1284782737493515, "learning_rate": 8.565979901062265e-06, "loss": 0.0155, "step": 65630 }, { "grad_norm": 0.1372298151254654, "learning_rate": 8.554414897078033e-06, "loss": 0.0061, "step": 65640 }, { "grad_norm": 0.10696634650230408, "learning_rate": 8.542856975010727e-06, "loss": 0.0063, "step": 65650 }, { "grad_norm": 0.13429376482963562, "learning_rate": 8.531306136835337e-06, "loss": 0.0046, "step": 65660 }, { "grad_norm": 0.18748319149017334, "learning_rate": 8.519762384525548e-06, "loss": 0.0082, "step": 65670 }, { "grad_norm": 0.20621874928474426, "learning_rate": 8.508225720053875e-06, "loss": 0.0078, "step": 65680 }, { "grad_norm": 0.22499828040599823, "learning_rate": 8.496696145391625e-06, "loss": 0.0069, "step": 65690 }, { "grad_norm": 0.12256471812725067, "learning_rate": 8.485173662508889e-06, "loss": 0.0047, "step": 65700 }, { "grad_norm": 0.2322717308998108, "learning_rate": 8.473658273374536e-06, "loss": 0.0075, "step": 65710 }, { "grad_norm": 0.09165988117456436, "learning_rate": 8.462149979956253e-06, "loss": 0.0044, "step": 65720 }, { "grad_norm": 0.2148849070072174, "learning_rate": 8.450648784220461e-06, "loss": 0.0071, "step": 65730 }, { "grad_norm": 0.19005760550498962, "learning_rate": 8.439154688132417e-06, "loss": 0.006, "step": 65740 }, { "grad_norm": 0.2262587994337082, "learning_rate": 8.427667693656143e-06, "loss": 0.007, "step": 65750 }, { "grad_norm": 0.33260828256607056, "learning_rate": 8.416187802754454e-06, "loss": 0.0075, "step": 65760 }, { "grad_norm": 0.1322745829820633, "learning_rate": 8.404715017388965e-06, "loss": 0.0053, "step": 65770 }, { "grad_norm": 0.20457591116428375, "learning_rate": 8.393249339520015e-06, "loss": 0.0096, "step": 65780 }, { "grad_norm": 0.11634767800569534, "learning_rate": 8.381790771106834e-06, "loss": 0.0055, "step": 65790 }, { "grad_norm": 0.2248682677745819, "learning_rate": 8.370339314107339e-06, "loss": 0.0052, "step": 65800 }, { "grad_norm": 0.1703198254108429, "learning_rate": 8.358894970478281e-06, "loss": 0.006, "step": 65810 }, { "grad_norm": 0.17884919047355652, "learning_rate": 8.347457742175196e-06, "loss": 0.0054, "step": 65820 }, { "grad_norm": 0.24656185507774353, "learning_rate": 8.33602763115236e-06, "loss": 0.007, "step": 65830 }, { "grad_norm": 0.24120692908763885, "learning_rate": 8.324604639362916e-06, "loss": 0.0132, "step": 65840 }, { "grad_norm": 0.3006303012371063, "learning_rate": 8.31318876875869e-06, "loss": 0.0077, "step": 65850 }, { "grad_norm": 0.12338969111442566, "learning_rate": 8.30178002129039e-06, "loss": 0.0057, "step": 65860 }, { "grad_norm": 0.09671696275472641, "learning_rate": 8.290378398907423e-06, "loss": 0.0071, "step": 65870 }, { "grad_norm": 0.0781046599149704, "learning_rate": 8.278983903558029e-06, "loss": 0.0113, "step": 65880 }, { "grad_norm": 0.21111628413200378, "learning_rate": 8.26759653718921e-06, "loss": 0.0069, "step": 65890 }, { "grad_norm": 0.11012173444032669, "learning_rate": 8.25621630174676e-06, "loss": 0.0072, "step": 65900 }, { "grad_norm": 0.10736466944217682, "learning_rate": 8.244843199175261e-06, "loss": 0.0137, "step": 65910 }, { "grad_norm": 0.0724552571773529, "learning_rate": 8.233477231418046e-06, "loss": 0.0057, "step": 65920 }, { "grad_norm": 0.20716002583503723, "learning_rate": 8.22211840041725e-06, "loss": 0.0098, "step": 65930 }, { "grad_norm": 0.2960454225540161, "learning_rate": 8.210766708113792e-06, "loss": 0.0128, "step": 65940 }, { "grad_norm": 0.14460302889347076, "learning_rate": 8.199422156447367e-06, "loss": 0.0096, "step": 65950 }, { "grad_norm": 0.13639315962791443, "learning_rate": 8.188084747356451e-06, "loss": 0.0094, "step": 65960 }, { "grad_norm": 0.2340257167816162, "learning_rate": 8.176754482778299e-06, "loss": 0.0068, "step": 65970 }, { "grad_norm": 0.14384272694587708, "learning_rate": 8.165431364648918e-06, "loss": 0.0107, "step": 65980 }, { "grad_norm": 0.12770837545394897, "learning_rate": 8.154115394903162e-06, "loss": 0.0077, "step": 65990 }, { "grad_norm": 0.1233021467924118, "learning_rate": 8.142806575474582e-06, "loss": 0.0083, "step": 66000 }, { "grad_norm": 0.20861196517944336, "learning_rate": 8.131504908295562e-06, "loss": 0.0065, "step": 66010 }, { "grad_norm": 0.14430852234363556, "learning_rate": 8.120210395297262e-06, "loss": 0.0058, "step": 66020 }, { "grad_norm": 0.1321989744901657, "learning_rate": 8.108923038409565e-06, "loss": 0.0097, "step": 66030 }, { "grad_norm": 0.15259599685668945, "learning_rate": 8.097642839561226e-06, "loss": 0.0098, "step": 66040 }, { "grad_norm": 0.1778450906276703, "learning_rate": 8.08636980067966e-06, "loss": 0.0075, "step": 66050 }, { "grad_norm": 0.10937119275331497, "learning_rate": 8.075103923691186e-06, "loss": 0.0047, "step": 66060 }, { "grad_norm": 0.13902482390403748, "learning_rate": 8.063845210520793e-06, "loss": 0.0068, "step": 66070 }, { "grad_norm": 0.19143164157867432, "learning_rate": 8.052593663092295e-06, "loss": 0.0089, "step": 66080 }, { "grad_norm": 0.09906777739524841, "learning_rate": 8.04134928332827e-06, "loss": 0.0077, "step": 66090 }, { "grad_norm": 0.11670403927564621, "learning_rate": 8.030112073150086e-06, "loss": 0.0119, "step": 66100 }, { "grad_norm": 0.5111179351806641, "learning_rate": 8.018882034477881e-06, "loss": 0.0054, "step": 66110 }, { "grad_norm": 0.12283544987440109, "learning_rate": 8.007659169230541e-06, "loss": 0.0045, "step": 66120 }, { "grad_norm": 0.15426519513130188, "learning_rate": 7.996443479325755e-06, "loss": 0.0061, "step": 66130 }, { "grad_norm": 0.08237076550722122, "learning_rate": 7.985234966679977e-06, "loss": 0.0105, "step": 66140 }, { "grad_norm": 0.12440727651119232, "learning_rate": 7.974033633208438e-06, "loss": 0.0069, "step": 66150 }, { "grad_norm": 0.15560854971408844, "learning_rate": 7.962839480825135e-06, "loss": 0.0063, "step": 66160 }, { "grad_norm": 0.4001365005970001, "learning_rate": 7.951652511442858e-06, "loss": 0.0119, "step": 66170 }, { "grad_norm": 0.1354498565196991, "learning_rate": 7.940472726973125e-06, "loss": 0.0085, "step": 66180 }, { "grad_norm": 0.11767410486936569, "learning_rate": 7.929300129326289e-06, "loss": 0.0052, "step": 66190 }, { "grad_norm": 0.1504564881324768, "learning_rate": 7.91813472041142e-06, "loss": 0.0062, "step": 66200 }, { "grad_norm": 0.2046940177679062, "learning_rate": 7.906976502136376e-06, "loss": 0.0086, "step": 66210 }, { "grad_norm": 0.1731315404176712, "learning_rate": 7.89582547640782e-06, "loss": 0.0077, "step": 66220 }, { "grad_norm": 0.0983923077583313, "learning_rate": 7.884681645131115e-06, "loss": 0.0067, "step": 66230 }, { "grad_norm": 0.2924439013004303, "learning_rate": 7.87354501021048e-06, "loss": 0.0074, "step": 66240 }, { "grad_norm": 0.08919552713632584, "learning_rate": 7.86241557354882e-06, "loss": 0.007, "step": 66250 }, { "grad_norm": 0.3832104206085205, "learning_rate": 7.8512933370479e-06, "loss": 0.0081, "step": 66260 }, { "grad_norm": 0.17217178642749786, "learning_rate": 7.840178302608158e-06, "loss": 0.0079, "step": 66270 }, { "grad_norm": 0.17733192443847656, "learning_rate": 7.82907047212888e-06, "loss": 0.0084, "step": 66280 }, { "grad_norm": 0.16830551624298096, "learning_rate": 7.81796984750809e-06, "loss": 0.0093, "step": 66290 }, { "grad_norm": 0.23378896713256836, "learning_rate": 7.806876430642546e-06, "loss": 0.009, "step": 66300 }, { "grad_norm": 0.18776728212833405, "learning_rate": 7.795790223427862e-06, "loss": 0.0079, "step": 66310 }, { "grad_norm": 0.25647979974746704, "learning_rate": 7.784711227758324e-06, "loss": 0.0094, "step": 66320 }, { "grad_norm": 0.11678169667720795, "learning_rate": 7.773639445527053e-06, "loss": 0.0086, "step": 66330 }, { "grad_norm": 0.15494108200073242, "learning_rate": 7.762574878625905e-06, "loss": 0.0071, "step": 66340 }, { "grad_norm": 0.08659423142671585, "learning_rate": 7.751517528945513e-06, "loss": 0.0047, "step": 66350 }, { "grad_norm": 0.09261471778154373, "learning_rate": 7.740467398375278e-06, "loss": 0.0088, "step": 66360 }, { "grad_norm": 0.15776832401752472, "learning_rate": 7.729424488803378e-06, "loss": 0.0061, "step": 66370 }, { "grad_norm": 0.16696639358997345, "learning_rate": 7.71838880211671e-06, "loss": 0.0141, "step": 66380 }, { "grad_norm": 0.07146161794662476, "learning_rate": 7.707360340200997e-06, "loss": 0.0049, "step": 66390 }, { "grad_norm": 0.08339033275842667, "learning_rate": 7.696339104940697e-06, "loss": 0.0066, "step": 66400 }, { "grad_norm": 0.11081632971763611, "learning_rate": 7.685325098219038e-06, "loss": 0.0096, "step": 66410 }, { "grad_norm": 0.10040804743766785, "learning_rate": 7.674318321918017e-06, "loss": 0.0087, "step": 66420 }, { "grad_norm": 0.07697252184152603, "learning_rate": 7.663318777918366e-06, "loss": 0.0068, "step": 66430 }, { "grad_norm": 0.061252694576978683, "learning_rate": 7.652326468099647e-06, "loss": 0.0097, "step": 66440 }, { "grad_norm": 0.0909859761595726, "learning_rate": 7.641341394340096e-06, "loss": 0.0052, "step": 66450 }, { "grad_norm": 0.37955936789512634, "learning_rate": 7.630363558516818e-06, "loss": 0.0114, "step": 66460 }, { "grad_norm": 0.15177232027053833, "learning_rate": 7.619392962505578e-06, "loss": 0.0102, "step": 66470 }, { "grad_norm": 0.19561202824115753, "learning_rate": 7.6084296081809725e-06, "loss": 0.0084, "step": 66480 }, { "grad_norm": 0.24229973554611206, "learning_rate": 7.597473497416347e-06, "loss": 0.0091, "step": 66490 }, { "grad_norm": 0.11632447689771652, "learning_rate": 7.586524632083764e-06, "loss": 0.0053, "step": 66500 }, { "grad_norm": 0.1687142699956894, "learning_rate": 7.5755830140541326e-06, "loss": 0.0067, "step": 66510 }, { "grad_norm": 0.057570476084947586, "learning_rate": 7.5646486451970475e-06, "loss": 0.0046, "step": 66520 }, { "grad_norm": 0.18916524946689606, "learning_rate": 7.553721527380897e-06, "loss": 0.0121, "step": 66530 }, { "grad_norm": 0.293041855096817, "learning_rate": 7.542801662472826e-06, "loss": 0.0076, "step": 66540 }, { "grad_norm": 0.09935431182384491, "learning_rate": 7.5318890523387475e-06, "loss": 0.0052, "step": 66550 }, { "grad_norm": 0.10927480459213257, "learning_rate": 7.52098369884332e-06, "loss": 0.0073, "step": 66560 }, { "grad_norm": 0.18591095507144928, "learning_rate": 7.510085603849992e-06, "loss": 0.0088, "step": 66570 }, { "grad_norm": 0.1243080124258995, "learning_rate": 7.499194769220918e-06, "loss": 0.0059, "step": 66580 }, { "grad_norm": 0.13808102905750275, "learning_rate": 7.48831119681706e-06, "loss": 0.0115, "step": 66590 }, { "grad_norm": 0.22212131321430206, "learning_rate": 7.477434888498119e-06, "loss": 0.0056, "step": 66600 }, { "grad_norm": 0.08903306722640991, "learning_rate": 7.466565846122564e-06, "loss": 0.0103, "step": 66610 }, { "grad_norm": 0.2216206192970276, "learning_rate": 7.455704071547626e-06, "loss": 0.0066, "step": 66620 }, { "grad_norm": 0.0963490754365921, "learning_rate": 7.444849566629247e-06, "loss": 0.0061, "step": 66630 }, { "grad_norm": 0.2130221724510193, "learning_rate": 7.434002333222212e-06, "loss": 0.0064, "step": 66640 }, { "grad_norm": 0.25578543543815613, "learning_rate": 7.423162373179976e-06, "loss": 0.0073, "step": 66650 }, { "grad_norm": 0.14861643314361572, "learning_rate": 7.412329688354835e-06, "loss": 0.0089, "step": 66660 }, { "grad_norm": 0.13454362750053406, "learning_rate": 7.40150428059776e-06, "loss": 0.0058, "step": 66670 }, { "grad_norm": 0.24678532779216766, "learning_rate": 7.3906861517585354e-06, "loss": 0.0097, "step": 66680 }, { "grad_norm": 0.2056690752506256, "learning_rate": 7.37987530368569e-06, "loss": 0.0074, "step": 66690 }, { "grad_norm": 0.063234843313694, "learning_rate": 7.369071738226474e-06, "loss": 0.0089, "step": 66700 }, { "grad_norm": 0.21430687606334686, "learning_rate": 7.358275457226954e-06, "loss": 0.0066, "step": 66710 }, { "grad_norm": 0.20076704025268555, "learning_rate": 7.347486462531899e-06, "loss": 0.0053, "step": 66720 }, { "grad_norm": 0.2951568067073822, "learning_rate": 7.336704755984858e-06, "loss": 0.0053, "step": 66730 }, { "grad_norm": 0.08277900516986847, "learning_rate": 7.325930339428133e-06, "loss": 0.0095, "step": 66740 }, { "grad_norm": 0.1723664402961731, "learning_rate": 7.315163214702769e-06, "loss": 0.0094, "step": 66750 }, { "grad_norm": 0.229012593626976, "learning_rate": 7.304403383648595e-06, "loss": 0.0072, "step": 66760 }, { "grad_norm": 0.10761835426092148, "learning_rate": 7.293650848104139e-06, "loss": 0.006, "step": 66770 }, { "grad_norm": 0.10323526710271835, "learning_rate": 7.2829056099067374e-06, "loss": 0.008, "step": 66780 }, { "grad_norm": 0.22089284658432007, "learning_rate": 7.2721676708924494e-06, "loss": 0.008, "step": 66790 }, { "grad_norm": 0.3898143470287323, "learning_rate": 7.261437032896096e-06, "loss": 0.022, "step": 66800 }, { "grad_norm": 0.15236926078796387, "learning_rate": 7.250713697751255e-06, "loss": 0.007, "step": 66810 }, { "grad_norm": 0.1300530582666397, "learning_rate": 7.239997667290255e-06, "loss": 0.0048, "step": 66820 }, { "grad_norm": 0.12175104767084122, "learning_rate": 7.2292889433441425e-06, "loss": 0.007, "step": 66830 }, { "grad_norm": 0.09608844667673111, "learning_rate": 7.218587527742793e-06, "loss": 0.0046, "step": 66840 }, { "grad_norm": 0.2039564549922943, "learning_rate": 7.207893422314749e-06, "loss": 0.0062, "step": 66850 }, { "grad_norm": 0.18233084678649902, "learning_rate": 7.1972066288873545e-06, "loss": 0.0062, "step": 66860 }, { "grad_norm": 0.7854871153831482, "learning_rate": 7.186527149286687e-06, "loss": 0.0201, "step": 66870 }, { "grad_norm": 0.1426917165517807, "learning_rate": 7.175854985337576e-06, "loss": 0.0071, "step": 66880 }, { "grad_norm": 0.11724075675010681, "learning_rate": 7.16519013886362e-06, "loss": 0.0141, "step": 66890 }, { "grad_norm": 0.2020956128835678, "learning_rate": 7.154532611687109e-06, "loss": 0.0056, "step": 66900 }, { "grad_norm": 0.108696848154068, "learning_rate": 7.143882405629177e-06, "loss": 0.0057, "step": 66910 }, { "grad_norm": 0.23369090259075165, "learning_rate": 7.133239522509605e-06, "loss": 0.0059, "step": 66920 }, { "grad_norm": 0.10090728104114532, "learning_rate": 7.1226039641469956e-06, "loss": 0.0059, "step": 66930 }, { "grad_norm": 0.1532106250524521, "learning_rate": 7.111975732358678e-06, "loss": 0.0072, "step": 66940 }, { "grad_norm": 0.08765382319688797, "learning_rate": 7.101354828960693e-06, "loss": 0.0056, "step": 66950 }, { "grad_norm": 0.0890984758734703, "learning_rate": 7.090741255767918e-06, "loss": 0.0064, "step": 66960 }, { "grad_norm": 0.22824552655220032, "learning_rate": 7.080135014593875e-06, "loss": 0.0069, "step": 66970 }, { "grad_norm": 0.17090362310409546, "learning_rate": 7.069536107250896e-06, "loss": 0.0066, "step": 66980 }, { "grad_norm": 0.07639095932245255, "learning_rate": 7.058944535550049e-06, "loss": 0.0037, "step": 66990 }, { "grad_norm": 0.12701234221458435, "learning_rate": 7.048360301301138e-06, "loss": 0.0073, "step": 67000 }, { "grad_norm": 0.09302342683076859, "learning_rate": 7.03778340631272e-06, "loss": 0.0065, "step": 67010 }, { "grad_norm": 0.17992016673088074, "learning_rate": 7.0272138523921e-06, "loss": 0.0063, "step": 67020 }, { "grad_norm": 0.20775215327739716, "learning_rate": 7.016651641345334e-06, "loss": 0.0104, "step": 67030 }, { "grad_norm": 0.1282671093940735, "learning_rate": 7.0060967749772e-06, "loss": 0.0066, "step": 67040 }, { "grad_norm": 0.3395046293735504, "learning_rate": 6.995549255091238e-06, "loss": 0.017, "step": 67050 }, { "grad_norm": 0.13418693840503693, "learning_rate": 6.98500908348973e-06, "loss": 0.0084, "step": 67060 }, { "grad_norm": 0.20246730744838715, "learning_rate": 6.974476261973711e-06, "loss": 0.0078, "step": 67070 }, { "grad_norm": 0.07290561497211456, "learning_rate": 6.963950792342949e-06, "loss": 0.0071, "step": 67080 }, { "grad_norm": 0.20298466086387634, "learning_rate": 6.9534326763959715e-06, "loss": 0.005, "step": 67090 }, { "grad_norm": 0.1772344708442688, "learning_rate": 6.942921915929995e-06, "loss": 0.0067, "step": 67100 }, { "grad_norm": 0.10561544448137283, "learning_rate": 6.9324185127410735e-06, "loss": 0.0072, "step": 67110 }, { "grad_norm": 0.2227640151977539, "learning_rate": 6.921922468623921e-06, "loss": 0.0101, "step": 67120 }, { "grad_norm": 0.11206340044736862, "learning_rate": 6.911433785372023e-06, "loss": 0.0063, "step": 67130 }, { "grad_norm": 0.10455548763275146, "learning_rate": 6.900952464777632e-06, "loss": 0.0086, "step": 67140 }, { "grad_norm": 0.24683639407157898, "learning_rate": 6.8904785086316815e-06, "loss": 0.0068, "step": 67150 }, { "grad_norm": 0.26956960558891296, "learning_rate": 6.880011918723927e-06, "loss": 0.0056, "step": 67160 }, { "grad_norm": 0.16647563874721527, "learning_rate": 6.86955269684279e-06, "loss": 0.0052, "step": 67170 }, { "grad_norm": 0.2032594084739685, "learning_rate": 6.859100844775473e-06, "loss": 0.0163, "step": 67180 }, { "grad_norm": 0.20052991807460785, "learning_rate": 6.84865636430792e-06, "loss": 0.0077, "step": 67190 }, { "grad_norm": 0.1215025782585144, "learning_rate": 6.838219257224804e-06, "loss": 0.0055, "step": 67200 }, { "grad_norm": 0.15909595787525177, "learning_rate": 6.827789525309536e-06, "loss": 0.0098, "step": 67210 }, { "grad_norm": 0.19461169838905334, "learning_rate": 6.81736717034428e-06, "loss": 0.0098, "step": 67220 }, { "grad_norm": 0.11144841462373734, "learning_rate": 6.806952194109933e-06, "loss": 0.016, "step": 67230 }, { "grad_norm": 0.1929745227098465, "learning_rate": 6.796544598386112e-06, "loss": 0.0045, "step": 67240 }, { "grad_norm": 0.26034650206565857, "learning_rate": 6.786144384951204e-06, "loss": 0.008, "step": 67250 }, { "grad_norm": 0.08053071796894073, "learning_rate": 6.775751555582322e-06, "loss": 0.0057, "step": 67260 }, { "grad_norm": 0.1938427984714508, "learning_rate": 6.76536611205531e-06, "loss": 0.0067, "step": 67270 }, { "grad_norm": 0.21171770989894867, "learning_rate": 6.754988056144762e-06, "loss": 0.0072, "step": 67280 }, { "grad_norm": 0.22217416763305664, "learning_rate": 6.744617389624014e-06, "loss": 0.0046, "step": 67290 }, { "grad_norm": 0.2040989100933075, "learning_rate": 6.734254114265087e-06, "loss": 0.0076, "step": 67300 }, { "grad_norm": 0.18070468306541443, "learning_rate": 6.723898231838843e-06, "loss": 0.0063, "step": 67310 }, { "grad_norm": 0.1255722939968109, "learning_rate": 6.713549744114766e-06, "loss": 0.0078, "step": 67320 }, { "grad_norm": 0.08169808238744736, "learning_rate": 6.703208652861159e-06, "loss": 0.0063, "step": 67330 }, { "grad_norm": 0.21221435070037842, "learning_rate": 6.69287495984503e-06, "loss": 0.0099, "step": 67340 }, { "grad_norm": 0.12781348824501038, "learning_rate": 6.6825486668321e-06, "loss": 0.0056, "step": 67350 }, { "grad_norm": 0.12173964828252792, "learning_rate": 6.672229775586886e-06, "loss": 0.0065, "step": 67360 }, { "grad_norm": 0.10452879220247269, "learning_rate": 6.661918287872576e-06, "loss": 0.0056, "step": 67370 }, { "grad_norm": 0.1231052577495575, "learning_rate": 6.6516142054511346e-06, "loss": 0.0118, "step": 67380 }, { "grad_norm": 0.13714838027954102, "learning_rate": 6.641317530083241e-06, "loss": 0.0058, "step": 67390 }, { "grad_norm": 0.22226884961128235, "learning_rate": 6.631028263528322e-06, "loss": 0.0089, "step": 67400 }, { "grad_norm": 0.18614351749420166, "learning_rate": 6.620746407544537e-06, "loss": 0.0074, "step": 67410 }, { "grad_norm": 0.16682758927345276, "learning_rate": 6.610471963888742e-06, "loss": 0.0091, "step": 67420 }, { "grad_norm": 0.21784526109695435, "learning_rate": 6.600204934316606e-06, "loss": 0.0053, "step": 67430 }, { "grad_norm": 0.10694603621959686, "learning_rate": 6.589945320582452e-06, "loss": 0.0061, "step": 67440 }, { "grad_norm": 0.21398165822029114, "learning_rate": 6.579693124439374e-06, "loss": 0.0061, "step": 67450 }, { "grad_norm": 0.2409149706363678, "learning_rate": 6.56944834763919e-06, "loss": 0.0061, "step": 67460 }, { "grad_norm": 0.1434798538684845, "learning_rate": 6.5592109919324575e-06, "loss": 0.0061, "step": 67470 }, { "grad_norm": 0.19435501098632812, "learning_rate": 6.54898105906846e-06, "loss": 0.0083, "step": 67480 }, { "grad_norm": 0.17015135288238525, "learning_rate": 6.5387585507952155e-06, "loss": 0.0117, "step": 67490 }, { "grad_norm": 0.10183979570865631, "learning_rate": 6.528543468859461e-06, "loss": 0.0081, "step": 67500 }, { "grad_norm": 0.12963539361953735, "learning_rate": 6.518335815006682e-06, "loss": 0.0122, "step": 67510 }, { "grad_norm": 0.2367352843284607, "learning_rate": 6.5081355909810845e-06, "loss": 0.0069, "step": 67520 }, { "grad_norm": 0.24180594086647034, "learning_rate": 6.497942798525608e-06, "loss": 0.0142, "step": 67530 }, { "grad_norm": 0.33107611536979675, "learning_rate": 6.487757439381936e-06, "loss": 0.0118, "step": 67540 }, { "grad_norm": 0.13251172006130219, "learning_rate": 6.477579515290433e-06, "loss": 0.0087, "step": 67550 }, { "grad_norm": 0.1408286839723587, "learning_rate": 6.467409027990273e-06, "loss": 0.0134, "step": 67560 }, { "grad_norm": 0.12184437364339828, "learning_rate": 6.457245979219279e-06, "loss": 0.012, "step": 67570 }, { "grad_norm": 0.25307872891426086, "learning_rate": 6.447090370714054e-06, "loss": 0.0139, "step": 67580 }, { "grad_norm": 0.13578923046588898, "learning_rate": 6.436942204209917e-06, "loss": 0.0117, "step": 67590 }, { "grad_norm": 0.08280348777770996, "learning_rate": 6.4268014814408804e-06, "loss": 0.011, "step": 67600 }, { "grad_norm": 0.18930844962596893, "learning_rate": 6.41666820413977e-06, "loss": 0.0051, "step": 67610 }, { "grad_norm": 0.20469023287296295, "learning_rate": 6.406542374038033e-06, "loss": 0.0096, "step": 67620 }, { "grad_norm": 0.1475832611322403, "learning_rate": 6.396423992865935e-06, "loss": 0.0064, "step": 67630 }, { "grad_norm": 0.2818075716495514, "learning_rate": 6.386313062352412e-06, "loss": 0.0067, "step": 67640 }, { "grad_norm": 0.1218504086136818, "learning_rate": 6.376209584225152e-06, "loss": 0.0052, "step": 67650 }, { "grad_norm": 0.2601471543312073, "learning_rate": 6.366113560210557e-06, "loss": 0.008, "step": 67660 }, { "grad_norm": 0.4319038391113281, "learning_rate": 6.356024992033766e-06, "loss": 0.0141, "step": 67670 }, { "grad_norm": 0.1037764623761177, "learning_rate": 6.345943881418648e-06, "loss": 0.0055, "step": 67680 }, { "grad_norm": 0.13523465394973755, "learning_rate": 6.33587023008777e-06, "loss": 0.0068, "step": 67690 }, { "grad_norm": 0.146647647023201, "learning_rate": 6.32580403976245e-06, "loss": 0.0043, "step": 67700 }, { "grad_norm": 0.09586120396852493, "learning_rate": 6.31574531216273e-06, "loss": 0.0052, "step": 67710 }, { "grad_norm": 0.1666111946105957, "learning_rate": 6.305694049007371e-06, "loss": 0.0087, "step": 67720 }, { "grad_norm": 0.24590162932872772, "learning_rate": 6.2956502520138575e-06, "loss": 0.0078, "step": 67730 }, { "grad_norm": 0.1553708016872406, "learning_rate": 6.285613922898409e-06, "loss": 0.0061, "step": 67740 }, { "grad_norm": 0.13825029134750366, "learning_rate": 6.275585063375927e-06, "loss": 0.0084, "step": 67750 }, { "grad_norm": 0.19742000102996826, "learning_rate": 6.265563675160113e-06, "loss": 0.009, "step": 67760 }, { "grad_norm": 0.08356115221977234, "learning_rate": 6.25554975996332e-06, "loss": 0.007, "step": 67770 }, { "grad_norm": 0.15284250676631927, "learning_rate": 6.245543319496661e-06, "loss": 0.0049, "step": 67780 }, { "grad_norm": 0.16711954772472382, "learning_rate": 6.2355443554699685e-06, "loss": 0.0114, "step": 67790 }, { "grad_norm": 0.11409159749746323, "learning_rate": 6.225552869591766e-06, "loss": 0.0054, "step": 67800 }, { "grad_norm": 0.13208553194999695, "learning_rate": 6.215568863569365e-06, "loss": 0.0084, "step": 67810 }, { "grad_norm": 0.3165893852710724, "learning_rate": 6.205592339108712e-06, "loss": 0.0147, "step": 67820 }, { "grad_norm": 0.1752793788909912, "learning_rate": 6.195623297914577e-06, "loss": 0.0082, "step": 67830 }, { "grad_norm": 0.20102308690547943, "learning_rate": 6.185661741690357e-06, "loss": 0.0073, "step": 67840 }, { "grad_norm": 0.10717589408159256, "learning_rate": 6.1757076721382145e-06, "loss": 0.0103, "step": 67850 }, { "grad_norm": 0.21125231683254242, "learning_rate": 6.165761090959038e-06, "loss": 0.0159, "step": 67860 }, { "grad_norm": 0.3000659942626953, "learning_rate": 6.155821999852424e-06, "loss": 0.008, "step": 67870 }, { "grad_norm": 0.12657737731933594, "learning_rate": 6.145890400516696e-06, "loss": 0.0069, "step": 67880 }, { "grad_norm": 0.1264527291059494, "learning_rate": 6.1359662946488816e-06, "loss": 0.0066, "step": 67890 }, { "grad_norm": 0.25617435574531555, "learning_rate": 6.12604968394474e-06, "loss": 0.0051, "step": 67900 }, { "grad_norm": 0.16461846232414246, "learning_rate": 6.11614057009875e-06, "loss": 0.0082, "step": 67910 }, { "grad_norm": 0.3036797046661377, "learning_rate": 6.106238954804111e-06, "loss": 0.0064, "step": 67920 }, { "grad_norm": 0.15403102338314056, "learning_rate": 6.096344839752738e-06, "loss": 0.0073, "step": 67930 }, { "grad_norm": 0.08659049868583679, "learning_rate": 6.086458226635278e-06, "loss": 0.0084, "step": 67940 }, { "grad_norm": 0.19719694554805756, "learning_rate": 6.076579117141046e-06, "loss": 0.0145, "step": 67950 }, { "grad_norm": 0.2166900932788849, "learning_rate": 6.066707512958153e-06, "loss": 0.0066, "step": 67960 }, { "grad_norm": 0.14270402491092682, "learning_rate": 6.056843415773361e-06, "loss": 0.0071, "step": 67970 }, { "grad_norm": 0.1488952487707138, "learning_rate": 6.0469868272721776e-06, "loss": 0.0058, "step": 67980 }, { "grad_norm": 0.09033553302288055, "learning_rate": 6.037137749138844e-06, "loss": 0.0128, "step": 67990 }, { "grad_norm": 0.20432963967323303, "learning_rate": 6.027296183056252e-06, "loss": 0.0136, "step": 68000 }, { "grad_norm": 0.23979419469833374, "learning_rate": 6.017462130706114e-06, "loss": 0.0085, "step": 68010 }, { "grad_norm": 0.12346333265304565, "learning_rate": 6.0076355937687516e-06, "loss": 0.0078, "step": 68020 }, { "grad_norm": 0.074558787047863, "learning_rate": 5.9978165739232925e-06, "loss": 0.0064, "step": 68030 }, { "grad_norm": 0.10870394855737686, "learning_rate": 5.988005072847508e-06, "loss": 0.007, "step": 68040 }, { "grad_norm": 0.16569188237190247, "learning_rate": 5.978201092217928e-06, "loss": 0.0148, "step": 68050 }, { "grad_norm": 0.1972598135471344, "learning_rate": 5.9684046337097895e-06, "loss": 0.0057, "step": 68060 }, { "grad_norm": 0.08021898567676544, "learning_rate": 5.958615698997016e-06, "loss": 0.0076, "step": 68070 }, { "grad_norm": 0.14079560339450836, "learning_rate": 5.948834289752303e-06, "loss": 0.0069, "step": 68080 }, { "grad_norm": 0.17211836576461792, "learning_rate": 5.939060407646996e-06, "loss": 0.0099, "step": 68090 }, { "grad_norm": 0.05825282260775566, "learning_rate": 5.929294054351198e-06, "loss": 0.0086, "step": 68100 }, { "grad_norm": 0.06732462346553802, "learning_rate": 5.919535231533707e-06, "loss": 0.0054, "step": 68110 }, { "grad_norm": 0.2201937884092331, "learning_rate": 5.909783940862046e-06, "loss": 0.0073, "step": 68120 }, { "grad_norm": 0.0986323133111, "learning_rate": 5.900040184002436e-06, "loss": 0.0054, "step": 68130 }, { "grad_norm": 0.2719455063343048, "learning_rate": 5.890303962619831e-06, "loss": 0.0079, "step": 68140 }, { "grad_norm": 0.18733690679073334, "learning_rate": 5.88057527837787e-06, "loss": 0.0066, "step": 68150 }, { "grad_norm": 0.11971567571163177, "learning_rate": 5.870854132938924e-06, "loss": 0.0072, "step": 68160 }, { "grad_norm": 0.21034547686576843, "learning_rate": 5.861140527964071e-06, "loss": 0.0104, "step": 68170 }, { "grad_norm": 0.09985937923192978, "learning_rate": 5.851434465113098e-06, "loss": 0.0048, "step": 68180 }, { "grad_norm": 0.09447168558835983, "learning_rate": 5.841735946044524e-06, "loss": 0.0051, "step": 68190 }, { "grad_norm": 0.25456640124320984, "learning_rate": 5.832044972415523e-06, "loss": 0.008, "step": 68200 }, { "grad_norm": 0.15792672336101532, "learning_rate": 5.8223615458820655e-06, "loss": 0.0052, "step": 68210 }, { "grad_norm": 0.08525262027978897, "learning_rate": 5.812685668098733e-06, "loss": 0.0073, "step": 68220 }, { "grad_norm": 0.05899905040860176, "learning_rate": 5.803017340718913e-06, "loss": 0.006, "step": 68230 }, { "grad_norm": 0.0515148788690567, "learning_rate": 5.793356565394636e-06, "loss": 0.0061, "step": 68240 }, { "grad_norm": 0.28323858976364136, "learning_rate": 5.783703343776658e-06, "loss": 0.0066, "step": 68250 }, { "grad_norm": 0.08207996189594269, "learning_rate": 5.774057677514477e-06, "loss": 0.0048, "step": 68260 }, { "grad_norm": 0.1444263607263565, "learning_rate": 5.764419568256235e-06, "loss": 0.0057, "step": 68270 }, { "grad_norm": 0.25929686427116394, "learning_rate": 5.754789017648865e-06, "loss": 0.0209, "step": 68280 }, { "grad_norm": 0.17032209038734436, "learning_rate": 5.745166027337934e-06, "loss": 0.0104, "step": 68290 }, { "grad_norm": 0.09921868145465851, "learning_rate": 5.735550598967754e-06, "loss": 0.008, "step": 68300 }, { "grad_norm": 0.23295515775680542, "learning_rate": 5.725942734181339e-06, "loss": 0.0093, "step": 68310 }, { "grad_norm": 0.13971582055091858, "learning_rate": 5.7163424346204146e-06, "loss": 0.0052, "step": 68320 }, { "grad_norm": 0.10547348111867905, "learning_rate": 5.706749701925407e-06, "loss": 0.0071, "step": 68330 }, { "grad_norm": 0.1089089959859848, "learning_rate": 5.697164537735461e-06, "loss": 0.0056, "step": 68340 }, { "grad_norm": 0.0826699435710907, "learning_rate": 5.687586943688406e-06, "loss": 0.0044, "step": 68350 }, { "grad_norm": 0.060873012989759445, "learning_rate": 5.678016921420792e-06, "loss": 0.0056, "step": 68360 }, { "grad_norm": 0.17336267232894897, "learning_rate": 5.668454472567875e-06, "loss": 0.0153, "step": 68370 }, { "grad_norm": 0.2443062961101532, "learning_rate": 5.658899598763617e-06, "loss": 0.0092, "step": 68380 }, { "grad_norm": 0.09228084236383438, "learning_rate": 5.6493523016407035e-06, "loss": 0.0041, "step": 68390 }, { "grad_norm": 0.11959195882081985, "learning_rate": 5.6398125828304615e-06, "loss": 0.0079, "step": 68400 }, { "grad_norm": 0.1845458745956421, "learning_rate": 5.630280443963015e-06, "loss": 0.0072, "step": 68410 }, { "grad_norm": 0.14758417010307312, "learning_rate": 5.620755886667112e-06, "loss": 0.0075, "step": 68420 }, { "grad_norm": 0.27113601565361023, "learning_rate": 5.611238912570271e-06, "loss": 0.0059, "step": 68430 }, { "grad_norm": 0.25782954692840576, "learning_rate": 5.601729523298649e-06, "loss": 0.013, "step": 68440 }, { "grad_norm": 0.11649148911237717, "learning_rate": 5.592227720477161e-06, "loss": 0.0063, "step": 68450 }, { "grad_norm": 0.2523511052131653, "learning_rate": 5.582733505729415e-06, "loss": 0.0092, "step": 68460 }, { "grad_norm": 0.14557228982448578, "learning_rate": 5.573246880677668e-06, "loss": 0.0074, "step": 68470 }, { "grad_norm": 0.16852639615535736, "learning_rate": 5.56376784694298e-06, "loss": 0.008, "step": 68480 }, { "grad_norm": 0.2484351396560669, "learning_rate": 5.554296406145027e-06, "loss": 0.0123, "step": 68490 }, { "grad_norm": 0.17665889859199524, "learning_rate": 5.544832559902219e-06, "loss": 0.0097, "step": 68500 }, { "grad_norm": 0.14092354476451874, "learning_rate": 5.53537630983168e-06, "loss": 0.0114, "step": 68510 }, { "grad_norm": 0.1403682827949524, "learning_rate": 5.5259276575492125e-06, "loss": 0.012, "step": 68520 }, { "grad_norm": 0.14519238471984863, "learning_rate": 5.516486604669357e-06, "loss": 0.0115, "step": 68530 }, { "grad_norm": 0.12978236377239227, "learning_rate": 5.507053152805303e-06, "loss": 0.0177, "step": 68540 }, { "grad_norm": 0.15902724862098694, "learning_rate": 5.497627303568975e-06, "loss": 0.0134, "step": 68550 }, { "grad_norm": 0.09776465594768524, "learning_rate": 5.488209058571003e-06, "loss": 0.0046, "step": 68560 }, { "grad_norm": 0.29572850465774536, "learning_rate": 5.4787984194207054e-06, "loss": 0.0062, "step": 68570 }, { "grad_norm": 0.09864270687103271, "learning_rate": 5.469395387726095e-06, "loss": 0.006, "step": 68580 }, { "grad_norm": 0.12550979852676392, "learning_rate": 5.45999996509391e-06, "loss": 0.0071, "step": 68590 }, { "grad_norm": 0.2236141860485077, "learning_rate": 5.450612153129536e-06, "loss": 0.0072, "step": 68600 }, { "grad_norm": 0.0451183021068573, "learning_rate": 5.4412319534371426e-06, "loss": 0.0072, "step": 68610 }, { "grad_norm": 0.16697299480438232, "learning_rate": 5.431859367619513e-06, "loss": 0.0063, "step": 68620 }, { "grad_norm": 0.11760128289461136, "learning_rate": 5.422494397278172e-06, "loss": 0.0066, "step": 68630 }, { "grad_norm": 0.43476322293281555, "learning_rate": 5.413137044013344e-06, "loss": 0.0109, "step": 68640 }, { "grad_norm": 0.14963531494140625, "learning_rate": 5.403787309423941e-06, "loss": 0.0091, "step": 68650 }, { "grad_norm": 0.10987047106027603, "learning_rate": 5.3944451951075835e-06, "loss": 0.0051, "step": 68660 }, { "grad_norm": 0.19610121846199036, "learning_rate": 5.385110702660562e-06, "loss": 0.0058, "step": 68670 }, { "grad_norm": 0.12718768417835236, "learning_rate": 5.375783833677922e-06, "loss": 0.0097, "step": 68680 }, { "grad_norm": 0.1096949651837349, "learning_rate": 5.36646458975334e-06, "loss": 0.0058, "step": 68690 }, { "grad_norm": 0.13207635283470154, "learning_rate": 5.3571529724792294e-06, "loss": 0.0054, "step": 68700 }, { "grad_norm": 0.3633706271648407, "learning_rate": 5.347848983446702e-06, "loss": 0.0082, "step": 68710 }, { "grad_norm": 0.08727380633354187, "learning_rate": 5.3385526242455185e-06, "loss": 0.009, "step": 68720 }, { "grad_norm": 0.16276690363883972, "learning_rate": 5.329263896464226e-06, "loss": 0.006, "step": 68730 }, { "grad_norm": 0.12274350970983505, "learning_rate": 5.3199828016899715e-06, "loss": 0.0077, "step": 68740 }, { "grad_norm": 0.1586175262928009, "learning_rate": 5.310709341508657e-06, "loss": 0.0058, "step": 68750 }, { "grad_norm": 0.13309815526008606, "learning_rate": 5.301443517504861e-06, "loss": 0.0048, "step": 68760 }, { "grad_norm": 0.2000378519296646, "learning_rate": 5.292185331261862e-06, "loss": 0.0054, "step": 68770 }, { "grad_norm": 0.16588394343852997, "learning_rate": 5.28293478436162e-06, "loss": 0.0064, "step": 68780 }, { "grad_norm": 0.12512348592281342, "learning_rate": 5.273691878384829e-06, "loss": 0.0091, "step": 68790 }, { "grad_norm": 0.0801742747426033, "learning_rate": 5.264456614910812e-06, "loss": 0.0064, "step": 68800 }, { "grad_norm": 0.1753787100315094, "learning_rate": 5.255228995517647e-06, "loss": 0.0146, "step": 68810 }, { "grad_norm": 0.1730462908744812, "learning_rate": 5.246009021782067e-06, "loss": 0.0074, "step": 68820 }, { "grad_norm": 0.20947352051734924, "learning_rate": 5.2367966952795225e-06, "loss": 0.0112, "step": 68830 }, { "grad_norm": 0.10316796600818634, "learning_rate": 5.2275920175841485e-06, "loss": 0.0049, "step": 68840 }, { "grad_norm": 0.1057596430182457, "learning_rate": 5.218394990268771e-06, "loss": 0.0066, "step": 68850 }, { "grad_norm": 0.33615928888320923, "learning_rate": 5.209205614904916e-06, "loss": 0.0076, "step": 68860 }, { "grad_norm": 0.19287459552288055, "learning_rate": 5.200023893062777e-06, "loss": 0.0072, "step": 68870 }, { "grad_norm": 0.13672307133674622, "learning_rate": 5.190849826311289e-06, "loss": 0.0046, "step": 68880 }, { "grad_norm": 0.2881356179714203, "learning_rate": 5.181683416218025e-06, "loss": 0.0084, "step": 68890 }, { "grad_norm": 0.13069486618041992, "learning_rate": 5.172524664349276e-06, "loss": 0.0046, "step": 68900 }, { "grad_norm": 0.12250684201717377, "learning_rate": 5.1633735722700416e-06, "loss": 0.0052, "step": 68910 }, { "grad_norm": 0.12788289785385132, "learning_rate": 5.154230141543958e-06, "loss": 0.0058, "step": 68920 }, { "grad_norm": 0.193992018699646, "learning_rate": 5.145094373733433e-06, "loss": 0.0069, "step": 68930 }, { "grad_norm": 0.07482478022575378, "learning_rate": 5.135966270399478e-06, "loss": 0.0039, "step": 68940 }, { "grad_norm": 0.33367830514907837, "learning_rate": 5.126845833101857e-06, "loss": 0.0101, "step": 68950 }, { "grad_norm": 0.0780666172504425, "learning_rate": 5.1177330633989994e-06, "loss": 0.0054, "step": 68960 }, { "grad_norm": 0.34673020243644714, "learning_rate": 5.108627962848033e-06, "loss": 0.0241, "step": 68970 }, { "grad_norm": 0.13529250025749207, "learning_rate": 5.099530533004759e-06, "loss": 0.0066, "step": 68980 }, { "grad_norm": 0.13627184927463531, "learning_rate": 5.090440775423699e-06, "loss": 0.0049, "step": 68990 }, { "grad_norm": 0.0846584364771843, "learning_rate": 5.081358691658022e-06, "loss": 0.0067, "step": 69000 }, { "grad_norm": 0.24354755878448486, "learning_rate": 5.072284283259621e-06, "loss": 0.0094, "step": 69010 }, { "grad_norm": 0.061918411403894424, "learning_rate": 5.063217551779053e-06, "loss": 0.0068, "step": 69020 }, { "grad_norm": 0.13290999829769135, "learning_rate": 5.054158498765588e-06, "loss": 0.0049, "step": 69030 }, { "grad_norm": 0.19014514982700348, "learning_rate": 5.0451071257671625e-06, "loss": 0.0088, "step": 69040 }, { "grad_norm": 0.18065416812896729, "learning_rate": 5.036063434330407e-06, "loss": 0.0085, "step": 69050 }, { "grad_norm": 0.3395198881626129, "learning_rate": 5.027027426000652e-06, "loss": 0.006, "step": 69060 }, { "grad_norm": 0.10618146508932114, "learning_rate": 5.017999102321886e-06, "loss": 0.0052, "step": 69070 }, { "grad_norm": 0.2048761248588562, "learning_rate": 5.0089784648368224e-06, "loss": 0.0077, "step": 69080 }, { "grad_norm": 0.2232203185558319, "learning_rate": 4.999965515086829e-06, "loss": 0.0072, "step": 69090 }, { "grad_norm": 0.08246451616287231, "learning_rate": 4.990960254611976e-06, "loss": 0.0045, "step": 69100 }, { "grad_norm": 0.30055999755859375, "learning_rate": 4.9819626849510194e-06, "loss": 0.0088, "step": 69110 }, { "grad_norm": 0.19186203181743622, "learning_rate": 4.97297280764138e-06, "loss": 0.0172, "step": 69120 }, { "grad_norm": 0.11276551336050034, "learning_rate": 4.963990624219211e-06, "loss": 0.007, "step": 69130 }, { "grad_norm": 0.20988622307777405, "learning_rate": 4.955016136219298e-06, "loss": 0.0062, "step": 69140 }, { "grad_norm": 0.18587012588977814, "learning_rate": 4.946049345175146e-06, "loss": 0.0063, "step": 69150 }, { "grad_norm": 0.2513361871242523, "learning_rate": 4.937090252618937e-06, "loss": 0.0068, "step": 69160 }, { "grad_norm": 0.0674784928560257, "learning_rate": 4.928138860081521e-06, "loss": 0.0135, "step": 69170 }, { "grad_norm": 0.06446032971143723, "learning_rate": 4.919195169092472e-06, "loss": 0.0079, "step": 69180 }, { "grad_norm": 0.3706144690513611, "learning_rate": 4.910259181179994e-06, "loss": 0.0095, "step": 69190 }, { "grad_norm": 0.16709457337856293, "learning_rate": 4.901330897871015e-06, "loss": 0.0063, "step": 69200 }, { "grad_norm": 0.0665082111954689, "learning_rate": 4.8924103206911375e-06, "loss": 0.0043, "step": 69210 }, { "grad_norm": 0.15538465976715088, "learning_rate": 4.883497451164637e-06, "loss": 0.0138, "step": 69220 }, { "grad_norm": 0.19626662135124207, "learning_rate": 4.874592290814484e-06, "loss": 0.0065, "step": 69230 }, { "grad_norm": 0.4047514796257019, "learning_rate": 4.865694841162327e-06, "loss": 0.0182, "step": 69240 }, { "grad_norm": 0.08163750916719437, "learning_rate": 4.856805103728496e-06, "loss": 0.0059, "step": 69250 }, { "grad_norm": 0.12598064541816711, "learning_rate": 4.847923080032007e-06, "loss": 0.0038, "step": 69260 }, { "grad_norm": 0.18326738476753235, "learning_rate": 4.8390487715905486e-06, "loss": 0.0055, "step": 69270 }, { "grad_norm": 0.40969163179397583, "learning_rate": 4.8301821799205e-06, "loss": 0.0054, "step": 69280 }, { "grad_norm": 0.16917763650417328, "learning_rate": 4.821323306536918e-06, "loss": 0.0046, "step": 69290 }, { "grad_norm": 0.0814218744635582, "learning_rate": 4.8124721529535455e-06, "loss": 0.0061, "step": 69300 }, { "grad_norm": 0.15990035235881805, "learning_rate": 4.803628720682807e-06, "loss": 0.0133, "step": 69310 }, { "grad_norm": 0.12319394946098328, "learning_rate": 4.794793011235776e-06, "loss": 0.0087, "step": 69320 }, { "grad_norm": 0.11292003840208054, "learning_rate": 4.78596502612228e-06, "loss": 0.0101, "step": 69330 }, { "grad_norm": 0.31703218817710876, "learning_rate": 4.777144766850738e-06, "loss": 0.0079, "step": 69340 }, { "grad_norm": 0.11232564598321915, "learning_rate": 4.768332234928313e-06, "loss": 0.0074, "step": 69350 }, { "grad_norm": 0.1182437464594841, "learning_rate": 4.759527431860828e-06, "loss": 0.0061, "step": 69360 }, { "grad_norm": 0.15591612458229065, "learning_rate": 4.750730359152755e-06, "loss": 0.0113, "step": 69370 }, { "grad_norm": 0.09201224893331528, "learning_rate": 4.741941018307311e-06, "loss": 0.0076, "step": 69380 }, { "grad_norm": 0.12675610184669495, "learning_rate": 4.733159410826321e-06, "loss": 0.0075, "step": 69390 }, { "grad_norm": 0.18607152998447418, "learning_rate": 4.724385538210357e-06, "loss": 0.0053, "step": 69400 }, { "grad_norm": 0.12409821152687073, "learning_rate": 4.715619401958599e-06, "loss": 0.0062, "step": 69410 }, { "grad_norm": 0.14935868978500366, "learning_rate": 4.706861003568958e-06, "loss": 0.0096, "step": 69420 }, { "grad_norm": 0.16321346163749695, "learning_rate": 4.698110344538003e-06, "loss": 0.0084, "step": 69430 }, { "grad_norm": 0.2750518321990967, "learning_rate": 4.689367426360975e-06, "loss": 0.0061, "step": 69440 }, { "grad_norm": 0.29102179408073425, "learning_rate": 4.680632250531819e-06, "loss": 0.0077, "step": 69450 }, { "grad_norm": 0.08937472105026245, "learning_rate": 4.671904818543115e-06, "loss": 0.0064, "step": 69460 }, { "grad_norm": 0.20093496143817902, "learning_rate": 4.66318513188615e-06, "loss": 0.0045, "step": 69470 }, { "grad_norm": 0.059603251516819, "learning_rate": 4.654473192050884e-06, "loss": 0.0072, "step": 69480 }, { "grad_norm": 0.050962310284376144, "learning_rate": 4.64576900052594e-06, "loss": 0.0067, "step": 69490 }, { "grad_norm": 0.14088809490203857, "learning_rate": 4.637072558798638e-06, "loss": 0.0059, "step": 69500 }, { "grad_norm": 0.21171098947525024, "learning_rate": 4.628383868354969e-06, "loss": 0.0117, "step": 69510 }, { "grad_norm": 0.10668369382619858, "learning_rate": 4.6197029306795595e-06, "loss": 0.0051, "step": 69520 }, { "grad_norm": 0.1553293764591217, "learning_rate": 4.611029747255779e-06, "loss": 0.0054, "step": 69530 }, { "grad_norm": 0.15650618076324463, "learning_rate": 4.6023643195656164e-06, "loss": 0.007, "step": 69540 }, { "grad_norm": 0.08218182623386383, "learning_rate": 4.593706649089768e-06, "loss": 0.0036, "step": 69550 }, { "grad_norm": 0.21658219397068024, "learning_rate": 4.585056737307597e-06, "loss": 0.0067, "step": 69560 }, { "grad_norm": 0.13421238958835602, "learning_rate": 4.576414585697103e-06, "loss": 0.0043, "step": 69570 }, { "grad_norm": 0.2574496269226074, "learning_rate": 4.567780195735044e-06, "loss": 0.0074, "step": 69580 }, { "grad_norm": 0.08617080003023148, "learning_rate": 4.559153568896757e-06, "loss": 0.0062, "step": 69590 }, { "grad_norm": 0.2559351921081543, "learning_rate": 4.550534706656329e-06, "loss": 0.0074, "step": 69600 }, { "grad_norm": 0.06282155215740204, "learning_rate": 4.541923610486465e-06, "loss": 0.0061, "step": 69610 }, { "grad_norm": 0.19240963459014893, "learning_rate": 4.533320281858578e-06, "loss": 0.0064, "step": 69620 }, { "grad_norm": 0.10913047939538956, "learning_rate": 4.52472472224274e-06, "loss": 0.0086, "step": 69630 }, { "grad_norm": 0.09668506681919098, "learning_rate": 4.51613693310769e-06, "loss": 0.0059, "step": 69640 }, { "grad_norm": 0.20110933482646942, "learning_rate": 4.507556915920868e-06, "loss": 0.0097, "step": 69650 }, { "grad_norm": 0.08268660306930542, "learning_rate": 4.498984672148332e-06, "loss": 0.0054, "step": 69660 }, { "grad_norm": 0.13857153058052063, "learning_rate": 4.490420203254864e-06, "loss": 0.0078, "step": 69670 }, { "grad_norm": 0.29660674929618835, "learning_rate": 4.481863510703893e-06, "loss": 0.0077, "step": 69680 }, { "grad_norm": 0.1681414693593979, "learning_rate": 4.473314595957523e-06, "loss": 0.0082, "step": 69690 }, { "grad_norm": 0.14098772406578064, "learning_rate": 4.464773460476535e-06, "loss": 0.0076, "step": 69700 }, { "grad_norm": 0.06979241222143173, "learning_rate": 4.456240105720372e-06, "loss": 0.0062, "step": 69710 }, { "grad_norm": 0.11902976781129837, "learning_rate": 4.4477145331471405e-06, "loss": 0.0161, "step": 69720 }, { "grad_norm": 0.07032778114080429, "learning_rate": 4.439196744213653e-06, "loss": 0.0059, "step": 69730 }, { "grad_norm": 0.1538575440645218, "learning_rate": 4.430686740375339e-06, "loss": 0.006, "step": 69740 }, { "grad_norm": 0.08822862803936005, "learning_rate": 4.422184523086342e-06, "loss": 0.0053, "step": 69750 }, { "grad_norm": 0.13818548619747162, "learning_rate": 4.41369009379946e-06, "loss": 0.0053, "step": 69760 }, { "grad_norm": 0.11807925254106522, "learning_rate": 4.405203453966139e-06, "loss": 0.0058, "step": 69770 }, { "grad_norm": 0.2129671573638916, "learning_rate": 4.396724605036539e-06, "loss": 0.0085, "step": 69780 }, { "grad_norm": 0.19340161979198456, "learning_rate": 4.388253548459437e-06, "loss": 0.0072, "step": 69790 }, { "grad_norm": 0.22941020131111145, "learning_rate": 4.3797902856823395e-06, "loss": 0.0063, "step": 69800 }, { "grad_norm": 0.11784593015909195, "learning_rate": 4.371334818151357e-06, "loss": 0.0058, "step": 69810 }, { "grad_norm": 0.1839981973171234, "learning_rate": 4.362887147311306e-06, "loss": 0.0053, "step": 69820 }, { "grad_norm": 0.11727640777826309, "learning_rate": 4.354447274605672e-06, "loss": 0.0049, "step": 69830 }, { "grad_norm": 0.14950522780418396, "learning_rate": 4.346015201476572e-06, "loss": 0.0141, "step": 69840 }, { "grad_norm": 0.25626683235168457, "learning_rate": 4.337590929364855e-06, "loss": 0.0064, "step": 69850 }, { "grad_norm": 0.1633554846048355, "learning_rate": 4.329174459709973e-06, "loss": 0.0057, "step": 69860 }, { "grad_norm": 0.2209860384464264, "learning_rate": 4.320765793950071e-06, "loss": 0.0087, "step": 69870 }, { "grad_norm": 0.23975983262062073, "learning_rate": 4.312364933521962e-06, "loss": 0.0093, "step": 69880 }, { "grad_norm": 0.240676149725914, "learning_rate": 4.303971879861129e-06, "loss": 0.0062, "step": 69890 }, { "grad_norm": 0.17958518862724304, "learning_rate": 4.295586634401716e-06, "loss": 0.0095, "step": 69900 }, { "grad_norm": 0.15961429476737976, "learning_rate": 4.287209198576536e-06, "loss": 0.0062, "step": 69910 }, { "grad_norm": 0.13397838175296783, "learning_rate": 4.278839573817045e-06, "loss": 0.011, "step": 69920 }, { "grad_norm": 0.2539516091346741, "learning_rate": 4.270477761553399e-06, "loss": 0.009, "step": 69930 }, { "grad_norm": 0.27613842487335205, "learning_rate": 4.262123763214393e-06, "loss": 0.0054, "step": 69940 }, { "grad_norm": 0.29550471901893616, "learning_rate": 4.253777580227508e-06, "loss": 0.0068, "step": 69950 }, { "grad_norm": 0.16876067221164703, "learning_rate": 4.2454392140188755e-06, "loss": 0.0072, "step": 69960 }, { "grad_norm": 0.10358277708292007, "learning_rate": 4.2371086660132785e-06, "loss": 0.0067, "step": 69970 }, { "grad_norm": 0.053267646580934525, "learning_rate": 4.228785937634205e-06, "loss": 0.0049, "step": 69980 }, { "grad_norm": 0.14352039992809296, "learning_rate": 4.220471030303758e-06, "loss": 0.0045, "step": 69990 }, { "grad_norm": 0.09478156268596649, "learning_rate": 4.212163945442754e-06, "loss": 0.0058, "step": 70000 }, { "grad_norm": 0.26249071955680847, "learning_rate": 4.203864684470621e-06, "loss": 0.0067, "step": 70010 }, { "grad_norm": 0.18738619983196259, "learning_rate": 4.19557324880549e-06, "loss": 0.0067, "step": 70020 }, { "grad_norm": 0.1391049325466156, "learning_rate": 4.187289639864145e-06, "loss": 0.0041, "step": 70030 }, { "grad_norm": 0.0399077944457531, "learning_rate": 4.1790138590619974e-06, "loss": 0.0038, "step": 70040 }, { "grad_norm": 0.15460337698459625, "learning_rate": 4.170745907813195e-06, "loss": 0.0069, "step": 70050 }, { "grad_norm": 0.15446655452251434, "learning_rate": 4.162485787530479e-06, "loss": 0.0061, "step": 70060 }, { "grad_norm": 0.13057510554790497, "learning_rate": 4.154233499625282e-06, "loss": 0.0049, "step": 70070 }, { "grad_norm": 0.12633226811885834, "learning_rate": 4.145989045507692e-06, "loss": 0.0058, "step": 70080 }, { "grad_norm": 0.16482919454574585, "learning_rate": 4.1377524265864666e-06, "loss": 0.0089, "step": 70090 }, { "grad_norm": 0.08057647943496704, "learning_rate": 4.1295236442690175e-06, "loss": 0.0059, "step": 70100 }, { "grad_norm": 0.1451271027326584, "learning_rate": 4.121302699961421e-06, "loss": 0.0051, "step": 70110 }, { "grad_norm": 0.19613833725452423, "learning_rate": 4.113089595068403e-06, "loss": 0.0106, "step": 70120 }, { "grad_norm": 0.1251760572195053, "learning_rate": 4.104884330993364e-06, "loss": 0.0095, "step": 70130 }, { "grad_norm": 0.15440934896469116, "learning_rate": 4.0966869091383585e-06, "loss": 0.0161, "step": 70140 }, { "grad_norm": 0.11643940955400467, "learning_rate": 4.088497330904101e-06, "loss": 0.0071, "step": 70150 }, { "grad_norm": 0.09977256506681442, "learning_rate": 4.080315597689976e-06, "loss": 0.0057, "step": 70160 }, { "grad_norm": 0.1138058602809906, "learning_rate": 4.072141710893995e-06, "loss": 0.0129, "step": 70170 }, { "grad_norm": 0.08528874814510345, "learning_rate": 4.063975671912879e-06, "loss": 0.0154, "step": 70180 }, { "grad_norm": 0.10565335303544998, "learning_rate": 4.055817482141949e-06, "loss": 0.0036, "step": 70190 }, { "grad_norm": 0.10639894753694534, "learning_rate": 4.047667142975259e-06, "loss": 0.0066, "step": 70200 }, { "grad_norm": 0.12747223675251007, "learning_rate": 4.039524655805443e-06, "loss": 0.0113, "step": 70210 }, { "grad_norm": 0.11144856363534927, "learning_rate": 4.03139002202384e-06, "loss": 0.0044, "step": 70220 }, { "grad_norm": 0.17910003662109375, "learning_rate": 4.023263243020447e-06, "loss": 0.0074, "step": 70230 }, { "grad_norm": 0.2204219400882721, "learning_rate": 4.015144320183884e-06, "loss": 0.0081, "step": 70240 }, { "grad_norm": 0.12095536291599274, "learning_rate": 4.007033254901482e-06, "loss": 0.0051, "step": 70250 }, { "grad_norm": 0.20107075572013855, "learning_rate": 3.9989300485591795e-06, "loss": 0.0074, "step": 70260 }, { "grad_norm": 0.09502371400594711, "learning_rate": 3.990834702541601e-06, "loss": 0.011, "step": 70270 }, { "grad_norm": 0.21651673316955566, "learning_rate": 3.982747218232019e-06, "loss": 0.005, "step": 70280 }, { "grad_norm": 0.1653796285390854, "learning_rate": 3.97466759701236e-06, "loss": 0.0107, "step": 70290 }, { "grad_norm": 0.17219749093055725, "learning_rate": 3.966595840263226e-06, "loss": 0.0059, "step": 70300 }, { "grad_norm": 0.2645966112613678, "learning_rate": 3.958531949363836e-06, "loss": 0.0073, "step": 70310 }, { "grad_norm": 0.12450313568115234, "learning_rate": 3.950475925692098e-06, "loss": 0.0099, "step": 70320 }, { "grad_norm": 0.14863049983978271, "learning_rate": 3.9424277706245685e-06, "loss": 0.0056, "step": 70330 }, { "grad_norm": 0.2613060176372528, "learning_rate": 3.934387485536451e-06, "loss": 0.0064, "step": 70340 }, { "grad_norm": 0.08033469319343567, "learning_rate": 3.926355071801619e-06, "loss": 0.0049, "step": 70350 }, { "grad_norm": 0.06000697985291481, "learning_rate": 3.9183305307925965e-06, "loss": 0.0046, "step": 70360 }, { "grad_norm": 0.08251404762268066, "learning_rate": 3.910313863880533e-06, "loss": 0.0042, "step": 70370 }, { "grad_norm": 0.09073170274496078, "learning_rate": 3.902305072435292e-06, "loss": 0.0051, "step": 70380 }, { "grad_norm": 0.12285982072353363, "learning_rate": 3.894304157825329e-06, "loss": 0.007, "step": 70390 }, { "grad_norm": 0.07899504899978638, "learning_rate": 3.886311121417791e-06, "loss": 0.0044, "step": 70400 }, { "grad_norm": 0.11189430952072144, "learning_rate": 3.878325964578472e-06, "loss": 0.0053, "step": 70410 }, { "grad_norm": 0.09240486472845078, "learning_rate": 3.870348688671815e-06, "loss": 0.0044, "step": 70420 }, { "grad_norm": 0.1378154456615448, "learning_rate": 3.862379295060931e-06, "loss": 0.0046, "step": 70430 }, { "grad_norm": 0.27261972427368164, "learning_rate": 3.854417785107534e-06, "loss": 0.0143, "step": 70440 }, { "grad_norm": 0.1842544674873352, "learning_rate": 3.8464641601720755e-06, "loss": 0.0063, "step": 70450 }, { "grad_norm": 0.19228456914424896, "learning_rate": 3.838518421613579e-06, "loss": 0.0065, "step": 70460 }, { "grad_norm": 0.1847243756055832, "learning_rate": 3.830580570789766e-06, "loss": 0.0084, "step": 70470 }, { "grad_norm": 0.13615673780441284, "learning_rate": 3.822650609057005e-06, "loss": 0.0063, "step": 70480 }, { "grad_norm": 0.08730112016201019, "learning_rate": 3.814728537770285e-06, "loss": 0.0059, "step": 70490 }, { "grad_norm": 0.13537178933620453, "learning_rate": 3.806814358283306e-06, "loss": 0.0068, "step": 70500 }, { "grad_norm": 0.13720907270908356, "learning_rate": 3.7989080719483596e-06, "loss": 0.0047, "step": 70510 }, { "grad_norm": 0.10195054858922958, "learning_rate": 3.7910096801164143e-06, "loss": 0.0084, "step": 70520 }, { "grad_norm": 0.15338066220283508, "learning_rate": 3.7831191841371016e-06, "loss": 0.0072, "step": 70530 }, { "grad_norm": 0.16680178046226501, "learning_rate": 3.775236585358688e-06, "loss": 0.008, "step": 70540 }, { "grad_norm": 0.20808959007263184, "learning_rate": 3.7673618851280843e-06, "loss": 0.0075, "step": 70550 }, { "grad_norm": 0.17440181970596313, "learning_rate": 3.759495084790887e-06, "loss": 0.0064, "step": 70560 }, { "grad_norm": 0.1717197746038437, "learning_rate": 3.751636185691282e-06, "loss": 0.0051, "step": 70570 }, { "grad_norm": 0.06124028190970421, "learning_rate": 3.7437851891721607e-06, "loss": 0.0055, "step": 70580 }, { "grad_norm": 0.2403302639722824, "learning_rate": 3.7359420965750404e-06, "loss": 0.0076, "step": 70590 }, { "grad_norm": 0.07372495532035828, "learning_rate": 3.7281069092400922e-06, "loss": 0.0049, "step": 70600 }, { "grad_norm": 0.1131361871957779, "learning_rate": 3.7202796285061348e-06, "loss": 0.0056, "step": 70610 }, { "grad_norm": 0.1543905884027481, "learning_rate": 3.712460255710637e-06, "loss": 0.0099, "step": 70620 }, { "grad_norm": 0.20388445258140564, "learning_rate": 3.704648792189719e-06, "loss": 0.0047, "step": 70630 }, { "grad_norm": 0.14119289815425873, "learning_rate": 3.696845239278124e-06, "loss": 0.0061, "step": 70640 }, { "grad_norm": 0.15098445117473602, "learning_rate": 3.689049598309302e-06, "loss": 0.0096, "step": 70650 }, { "grad_norm": 0.2453235238790512, "learning_rate": 3.681261870615288e-06, "loss": 0.0133, "step": 70660 }, { "grad_norm": 0.08751659095287323, "learning_rate": 3.6734820575268004e-06, "loss": 0.0151, "step": 70670 }, { "grad_norm": 0.1263769268989563, "learning_rate": 3.665710160373209e-06, "loss": 0.0047, "step": 70680 }, { "grad_norm": 0.1315121352672577, "learning_rate": 3.65794618048248e-06, "loss": 0.0055, "step": 70690 }, { "grad_norm": 0.06357283890247345, "learning_rate": 3.6501901191813125e-06, "loss": 0.0046, "step": 70700 }, { "grad_norm": 0.10827376693487167, "learning_rate": 3.642441977794975e-06, "loss": 0.0087, "step": 70710 }, { "grad_norm": 0.16904906928539276, "learning_rate": 3.634701757647424e-06, "loss": 0.0069, "step": 70720 }, { "grad_norm": 0.19151617586612701, "learning_rate": 3.6269694600612468e-06, "loss": 0.0037, "step": 70730 }, { "grad_norm": 0.23826970160007477, "learning_rate": 3.619245086357681e-06, "loss": 0.0058, "step": 70740 }, { "grad_norm": 0.21463410556316376, "learning_rate": 3.611528637856615e-06, "loss": 0.0073, "step": 70750 }, { "grad_norm": 0.15396453440189362, "learning_rate": 3.6038201158765884e-06, "loss": 0.0051, "step": 70760 }, { "grad_norm": 0.44220390915870667, "learning_rate": 3.5961195217347534e-06, "loss": 0.0158, "step": 70770 }, { "grad_norm": 0.1524381935596466, "learning_rate": 3.588426856746946e-06, "loss": 0.0074, "step": 70780 }, { "grad_norm": 0.1535586714744568, "learning_rate": 3.5807421222276316e-06, "loss": 0.0152, "step": 70790 }, { "grad_norm": 0.11928867548704147, "learning_rate": 3.5730653194899156e-06, "loss": 0.0071, "step": 70800 }, { "grad_norm": 0.18517616391181946, "learning_rate": 3.565396449845554e-06, "loss": 0.0061, "step": 70810 }, { "grad_norm": 0.3515370488166809, "learning_rate": 3.557735514604954e-06, "loss": 0.0105, "step": 70820 }, { "grad_norm": 0.11979477107524872, "learning_rate": 3.5500825150771633e-06, "loss": 0.0051, "step": 70830 }, { "grad_norm": 0.09337838739156723, "learning_rate": 3.5424374525698466e-06, "loss": 0.0065, "step": 70840 }, { "grad_norm": 0.14390315115451813, "learning_rate": 3.5348003283893704e-06, "loss": 0.0048, "step": 70850 }, { "grad_norm": 0.10240621864795685, "learning_rate": 3.527171143840685e-06, "loss": 0.0073, "step": 70860 }, { "grad_norm": 0.13682661950588226, "learning_rate": 3.519549900227409e-06, "loss": 0.0094, "step": 70870 }, { "grad_norm": 0.15062472224235535, "learning_rate": 3.511936598851828e-06, "loss": 0.0087, "step": 70880 }, { "grad_norm": 0.16868235170841217, "learning_rate": 3.504331241014813e-06, "loss": 0.0055, "step": 70890 }, { "grad_norm": 0.25276345014572144, "learning_rate": 3.4967338280159414e-06, "loss": 0.0077, "step": 70900 }, { "grad_norm": 0.09097284078598022, "learning_rate": 3.4891443611533846e-06, "loss": 0.0109, "step": 70910 }, { "grad_norm": 0.12434884905815125, "learning_rate": 3.481562841723984e-06, "loss": 0.0118, "step": 70920 }, { "grad_norm": 0.21371515095233917, "learning_rate": 3.473989271023215e-06, "loss": 0.0056, "step": 70930 }, { "grad_norm": 0.16206811368465424, "learning_rate": 3.4664236503451854e-06, "loss": 0.006, "step": 70940 }, { "grad_norm": 0.11535242944955826, "learning_rate": 3.4588659809826674e-06, "loss": 0.007, "step": 70950 }, { "grad_norm": 0.20977066457271576, "learning_rate": 3.451316264227039e-06, "loss": 0.0152, "step": 70960 }, { "grad_norm": 0.10440278798341751, "learning_rate": 3.4437745013683574e-06, "loss": 0.0059, "step": 70970 }, { "grad_norm": 0.2207888662815094, "learning_rate": 3.4362406936952916e-06, "loss": 0.0058, "step": 70980 }, { "grad_norm": 0.1528061479330063, "learning_rate": 3.4287148424951786e-06, "loss": 0.0064, "step": 70990 }, { "grad_norm": 0.12609714269638062, "learning_rate": 3.4211969490539617e-06, "loss": 0.0053, "step": 71000 }, { "grad_norm": 0.25997409224510193, "learning_rate": 3.4136870146562584e-06, "loss": 0.0051, "step": 71010 }, { "grad_norm": 0.20478862524032593, "learning_rate": 3.406185040585308e-06, "loss": 0.0051, "step": 71020 }, { "grad_norm": 0.1267831176519394, "learning_rate": 3.3986910281229966e-06, "loss": 0.0078, "step": 71030 }, { "grad_norm": 0.13207261264324188, "learning_rate": 3.391204978549828e-06, "loss": 0.0083, "step": 71040 }, { "grad_norm": 0.18512126803398132, "learning_rate": 3.3837268931449785e-06, "loss": 0.0124, "step": 71050 }, { "grad_norm": 0.09365303069353104, "learning_rate": 3.376256773186248e-06, "loss": 0.0125, "step": 71060 }, { "grad_norm": 0.10493876785039902, "learning_rate": 3.3687946199500664e-06, "loss": 0.0089, "step": 71070 }, { "grad_norm": 0.1692817360162735, "learning_rate": 3.3613404347115295e-06, "loss": 0.0041, "step": 71080 }, { "grad_norm": 0.16188466548919678, "learning_rate": 3.35389421874433e-06, "loss": 0.0051, "step": 71090 }, { "grad_norm": 0.12269073724746704, "learning_rate": 3.3464559733208446e-06, "loss": 0.0058, "step": 71100 }, { "grad_norm": 0.29999786615371704, "learning_rate": 3.3390256997120505e-06, "loss": 0.0116, "step": 71110 }, { "grad_norm": 0.10911145806312561, "learning_rate": 3.331603399187583e-06, "loss": 0.0115, "step": 71120 }, { "grad_norm": 0.18456870317459106, "learning_rate": 3.324189073015721e-06, "loss": 0.0084, "step": 71130 }, { "grad_norm": 0.20920182764530182, "learning_rate": 3.3167827224633465e-06, "loss": 0.0067, "step": 71140 }, { "grad_norm": 0.11566741019487381, "learning_rate": 3.309384348796024e-06, "loss": 0.0075, "step": 71150 }, { "grad_norm": 0.17665335536003113, "learning_rate": 3.301993953277921e-06, "loss": 0.0055, "step": 71160 }, { "grad_norm": 0.15695244073867798, "learning_rate": 3.29461153717186e-06, "loss": 0.004, "step": 71170 }, { "grad_norm": 0.1354185938835144, "learning_rate": 3.287237101739293e-06, "loss": 0.0135, "step": 71180 }, { "grad_norm": 0.3916688561439514, "learning_rate": 3.2798706482403075e-06, "loss": 0.0135, "step": 71190 }, { "grad_norm": 0.11367017775774002, "learning_rate": 3.2725121779336285e-06, "loss": 0.0069, "step": 71200 }, { "grad_norm": 0.2064627707004547, "learning_rate": 3.265161692076618e-06, "loss": 0.0099, "step": 71210 }, { "grad_norm": 0.08383526653051376, "learning_rate": 3.257819191925282e-06, "loss": 0.0052, "step": 71220 }, { "grad_norm": 0.1589747816324234, "learning_rate": 3.2504846787342392e-06, "loss": 0.0073, "step": 71230 }, { "grad_norm": 0.10024180263280869, "learning_rate": 3.243158153756759e-06, "loss": 0.0073, "step": 71240 }, { "grad_norm": 0.0989885926246643, "learning_rate": 3.2358396182447514e-06, "loss": 0.0058, "step": 71250 }, { "grad_norm": 0.054860666394233704, "learning_rate": 3.2285290734487496e-06, "loss": 0.0057, "step": 71260 }, { "grad_norm": 0.1289122998714447, "learning_rate": 3.221226520617926e-06, "loss": 0.0065, "step": 71270 }, { "grad_norm": 0.185776486992836, "learning_rate": 3.213931961000094e-06, "loss": 0.0088, "step": 71280 }, { "grad_norm": 0.13426503539085388, "learning_rate": 3.2066453958416733e-06, "loss": 0.0053, "step": 71290 }, { "grad_norm": 0.16197311878204346, "learning_rate": 3.199366826387773e-06, "loss": 0.0103, "step": 71300 }, { "grad_norm": 0.4781731367111206, "learning_rate": 3.192096253882071e-06, "loss": 0.0053, "step": 71310 }, { "grad_norm": 0.14281417429447174, "learning_rate": 3.1848336795669177e-06, "loss": 0.0064, "step": 71320 }, { "grad_norm": 0.11015976965427399, "learning_rate": 3.1775791046833035e-06, "loss": 0.0093, "step": 71330 }, { "grad_norm": 0.1427365094423294, "learning_rate": 3.170332530470804e-06, "loss": 0.0084, "step": 71340 }, { "grad_norm": 0.2290431708097458, "learning_rate": 3.1630939581677012e-06, "loss": 0.0077, "step": 71350 }, { "grad_norm": 0.09913822263479233, "learning_rate": 3.155863389010838e-06, "loss": 0.0096, "step": 71360 }, { "grad_norm": 0.08158905804157257, "learning_rate": 3.1486408242357323e-06, "loss": 0.0109, "step": 71370 }, { "grad_norm": 0.33227479457855225, "learning_rate": 3.1414262650765248e-06, "loss": 0.0235, "step": 71380 }, { "grad_norm": 0.21008436381816864, "learning_rate": 3.134219712765979e-06, "loss": 0.0074, "step": 71390 }, { "grad_norm": 0.11378796398639679, "learning_rate": 3.1270211685355044e-06, "loss": 0.0126, "step": 71400 }, { "grad_norm": 0.1200263574719429, "learning_rate": 3.1198306336151338e-06, "loss": 0.0061, "step": 71410 }, { "grad_norm": 0.06707712262868881, "learning_rate": 3.11264810923354e-06, "loss": 0.0083, "step": 71420 }, { "grad_norm": 0.09760845452547073, "learning_rate": 3.105473596618008e-06, "loss": 0.0041, "step": 71430 }, { "grad_norm": 0.26765432953834534, "learning_rate": 3.0983070969944683e-06, "loss": 0.0056, "step": 71440 }, { "grad_norm": 0.0841655284166336, "learning_rate": 3.09114861158748e-06, "loss": 0.0048, "step": 71450 }, { "grad_norm": 0.19949181377887726, "learning_rate": 3.0839981416202314e-06, "loss": 0.0093, "step": 71460 }, { "grad_norm": 0.09548134356737137, "learning_rate": 3.0768556883145507e-06, "loss": 0.0064, "step": 71470 }, { "grad_norm": 0.13428670167922974, "learning_rate": 3.0697212528908834e-06, "loss": 0.0058, "step": 71480 }, { "grad_norm": 0.19114241003990173, "learning_rate": 3.0625948365683e-06, "loss": 0.0062, "step": 71490 }, { "grad_norm": 0.0795864462852478, "learning_rate": 3.055476440564531e-06, "loss": 0.0072, "step": 71500 }, { "grad_norm": 0.07415567338466644, "learning_rate": 3.0483660660958924e-06, "loss": 0.005, "step": 71510 }, { "grad_norm": 0.3838105797767639, "learning_rate": 3.0412637143773624e-06, "loss": 0.0127, "step": 71520 }, { "grad_norm": 0.09973352402448654, "learning_rate": 3.034169386622554e-06, "loss": 0.0053, "step": 71530 }, { "grad_norm": 0.07812508940696716, "learning_rate": 3.027083084043658e-06, "loss": 0.0063, "step": 71540 }, { "grad_norm": 0.0938640758395195, "learning_rate": 3.0200048078515676e-06, "loss": 0.0054, "step": 71550 }, { "grad_norm": 0.20482167601585388, "learning_rate": 3.012934559255742e-06, "loss": 0.0042, "step": 71560 }, { "grad_norm": 0.17076475918293, "learning_rate": 3.0058723394642994e-06, "loss": 0.0073, "step": 71570 }, { "grad_norm": 0.29945388436317444, "learning_rate": 2.998818149683985e-06, "loss": 0.0146, "step": 71580 }, { "grad_norm": 0.15422679483890533, "learning_rate": 2.9917719911201627e-06, "loss": 0.0051, "step": 71590 }, { "grad_norm": 0.06884840130805969, "learning_rate": 2.9847338649768352e-06, "loss": 0.0051, "step": 71600 }, { "grad_norm": 0.09664638340473175, "learning_rate": 2.977703772456608e-06, "loss": 0.0053, "step": 71610 }, { "grad_norm": 0.11556775867938995, "learning_rate": 2.9706817147607535e-06, "loss": 0.0056, "step": 71620 }, { "grad_norm": 0.06832779198884964, "learning_rate": 2.9636676930891393e-06, "loss": 0.0061, "step": 71630 }, { "grad_norm": 0.10190997272729874, "learning_rate": 2.9566617086402625e-06, "loss": 0.0051, "step": 71640 }, { "grad_norm": 0.0995296835899353, "learning_rate": 2.9496637626112655e-06, "loss": 0.0056, "step": 71650 }, { "grad_norm": 0.11909503489732742, "learning_rate": 2.9426738561979027e-06, "loss": 0.007, "step": 71660 }, { "grad_norm": 0.261265367269516, "learning_rate": 2.9356919905945524e-06, "loss": 0.0209, "step": 71670 }, { "grad_norm": 0.30156266689300537, "learning_rate": 2.928718166994243e-06, "loss": 0.0082, "step": 71680 }, { "grad_norm": 0.20435714721679688, "learning_rate": 2.9217523865885833e-06, "loss": 0.0105, "step": 71690 }, { "grad_norm": 0.10564948618412018, "learning_rate": 2.9147946505678537e-06, "loss": 0.0058, "step": 71700 }, { "grad_norm": 0.10060321539640427, "learning_rate": 2.9078449601209313e-06, "loss": 0.0045, "step": 71710 }, { "grad_norm": 0.34365347027778625, "learning_rate": 2.900903316435333e-06, "loss": 0.0085, "step": 71720 }, { "grad_norm": 0.25315403938293457, "learning_rate": 2.8939697206971983e-06, "loss": 0.0091, "step": 71730 }, { "grad_norm": 0.08450125902891159, "learning_rate": 2.8870441740912746e-06, "loss": 0.0059, "step": 71740 }, { "grad_norm": 0.1465640813112259, "learning_rate": 2.8801266778009762e-06, "loss": 0.0084, "step": 71750 }, { "grad_norm": 0.08444442600011826, "learning_rate": 2.8732172330082798e-06, "loss": 0.0111, "step": 71760 }, { "grad_norm": 0.08517562597990036, "learning_rate": 2.8663158408938517e-06, "loss": 0.0066, "step": 71770 }, { "grad_norm": 0.08926097303628922, "learning_rate": 2.859422502636938e-06, "loss": 0.0071, "step": 71780 }, { "grad_norm": 0.25563254952430725, "learning_rate": 2.8525372194154076e-06, "loss": 0.0066, "step": 71790 }, { "grad_norm": 0.0702219158411026, "learning_rate": 2.8456599924057913e-06, "loss": 0.0048, "step": 71800 }, { "grad_norm": 0.17550165951251984, "learning_rate": 2.8387908227831995e-06, "loss": 0.0087, "step": 71810 }, { "grad_norm": 0.0863342136144638, "learning_rate": 2.831929711721404e-06, "loss": 0.0093, "step": 71820 }, { "grad_norm": 0.18367129564285278, "learning_rate": 2.825076660392767e-06, "loss": 0.0048, "step": 71830 }, { "grad_norm": 0.05856896936893463, "learning_rate": 2.8182316699682908e-06, "loss": 0.0094, "step": 71840 }, { "grad_norm": 0.11431185901165009, "learning_rate": 2.811394741617601e-06, "loss": 0.0086, "step": 71850 }, { "grad_norm": 0.3909386992454529, "learning_rate": 2.8045658765089356e-06, "loss": 0.0111, "step": 71860 }, { "grad_norm": 0.1549731343984604, "learning_rate": 2.7977450758091605e-06, "loss": 0.0151, "step": 71870 }, { "grad_norm": 0.2289043813943863, "learning_rate": 2.790932340683783e-06, "loss": 0.0118, "step": 71880 }, { "grad_norm": 0.20438852906227112, "learning_rate": 2.7841276722968823e-06, "loss": 0.0103, "step": 71890 }, { "grad_norm": 0.10953350365161896, "learning_rate": 2.7773310718112067e-06, "loss": 0.0053, "step": 71900 }, { "grad_norm": 0.11316566169261932, "learning_rate": 2.7705425403881102e-06, "loss": 0.0062, "step": 71910 }, { "grad_norm": 0.1929018199443817, "learning_rate": 2.763762079187565e-06, "loss": 0.0045, "step": 71920 }, { "grad_norm": 0.09239687770605087, "learning_rate": 2.756989689368178e-06, "loss": 0.0101, "step": 71930 }, { "grad_norm": 0.11188600212335587, "learning_rate": 2.750225372087134e-06, "loss": 0.0079, "step": 71940 }, { "grad_norm": 0.3048926293849945, "learning_rate": 2.7434691285003033e-06, "loss": 0.006, "step": 71950 }, { "grad_norm": 0.14469461143016815, "learning_rate": 2.7367209597621178e-06, "loss": 0.0092, "step": 71960 }, { "grad_norm": 0.1762285679578781, "learning_rate": 2.729980867025683e-06, "loss": 0.008, "step": 71970 }, { "grad_norm": 0.14110329747200012, "learning_rate": 2.7232488514426724e-06, "loss": 0.0056, "step": 71980 }, { "grad_norm": 0.11936303228139877, "learning_rate": 2.7165249141634097e-06, "loss": 0.0045, "step": 71990 }, { "grad_norm": 0.2542703151702881, "learning_rate": 2.709809056336837e-06, "loss": 0.0161, "step": 72000 }, { "grad_norm": 0.23212222754955292, "learning_rate": 2.703101279110498e-06, "loss": 0.0068, "step": 72010 }, { "grad_norm": 0.1667550951242447, "learning_rate": 2.696401583630587e-06, "loss": 0.013, "step": 72020 }, { "grad_norm": 0.19356589019298553, "learning_rate": 2.689709971041887e-06, "loss": 0.0051, "step": 72030 }, { "grad_norm": 0.1466618925333023, "learning_rate": 2.6830264424878126e-06, "loss": 0.0058, "step": 72040 }, { "grad_norm": 0.13589666783809662, "learning_rate": 2.676350999110394e-06, "loss": 0.0169, "step": 72050 }, { "grad_norm": 0.06343752890825272, "learning_rate": 2.6696836420502856e-06, "loss": 0.0112, "step": 72060 }, { "grad_norm": 0.0800781100988388, "learning_rate": 2.663024372446765e-06, "loss": 0.0053, "step": 72070 }, { "grad_norm": 0.14989855885505676, "learning_rate": 2.6563731914377055e-06, "loss": 0.0059, "step": 72080 }, { "grad_norm": 0.10587993264198303, "learning_rate": 2.6497301001596087e-06, "loss": 0.0109, "step": 72090 }, { "grad_norm": 0.335364431142807, "learning_rate": 2.643095099747611e-06, "loss": 0.009, "step": 72100 }, { "grad_norm": 0.14797624945640564, "learning_rate": 2.636468191335445e-06, "loss": 0.0098, "step": 72110 }, { "grad_norm": 0.19017057120800018, "learning_rate": 2.629849376055471e-06, "loss": 0.0058, "step": 72120 }, { "grad_norm": 0.09484095126390457, "learning_rate": 2.6232386550386678e-06, "loss": 0.0066, "step": 72130 }, { "grad_norm": 0.09769479930400848, "learning_rate": 2.6166360294146097e-06, "loss": 0.0068, "step": 72140 }, { "grad_norm": 0.11359597742557526, "learning_rate": 2.6100415003115275e-06, "loss": 0.0123, "step": 72150 }, { "grad_norm": 0.07305225729942322, "learning_rate": 2.603455068856225e-06, "loss": 0.0079, "step": 72160 }, { "grad_norm": 0.1532241851091385, "learning_rate": 2.5968767361741584e-06, "loss": 0.0075, "step": 72170 }, { "grad_norm": 0.13099519908428192, "learning_rate": 2.590306503389378e-06, "loss": 0.01, "step": 72180 }, { "grad_norm": 0.17541328072547913, "learning_rate": 2.583744371624558e-06, "loss": 0.0174, "step": 72190 }, { "grad_norm": 0.15067161619663239, "learning_rate": 2.57719034200099e-06, "loss": 0.0059, "step": 72200 }, { "grad_norm": 0.4528706669807434, "learning_rate": 2.570644415638568e-06, "loss": 0.0056, "step": 72210 }, { "grad_norm": 0.049760010093450546, "learning_rate": 2.5641065936558295e-06, "loss": 0.0151, "step": 72220 }, { "grad_norm": 0.10994402319192886, "learning_rate": 2.557576877169898e-06, "loss": 0.0112, "step": 72230 }, { "grad_norm": 0.08647482842206955, "learning_rate": 2.5510552672965205e-06, "loss": 0.0041, "step": 72240 }, { "grad_norm": 0.09396696090698242, "learning_rate": 2.544541765150077e-06, "loss": 0.0046, "step": 72250 }, { "grad_norm": 0.11995294690132141, "learning_rate": 2.5380363718435165e-06, "loss": 0.0104, "step": 72260 }, { "grad_norm": 0.18155641853809357, "learning_rate": 2.5315390884884714e-06, "loss": 0.0053, "step": 72270 }, { "grad_norm": 0.19863715767860413, "learning_rate": 2.5250499161951214e-06, "loss": 0.0082, "step": 72280 }, { "grad_norm": 0.14572761952877045, "learning_rate": 2.518568856072301e-06, "loss": 0.0046, "step": 72290 }, { "grad_norm": 0.26972073316574097, "learning_rate": 2.512095909227441e-06, "loss": 0.0074, "step": 72300 }, { "grad_norm": 0.14306777715682983, "learning_rate": 2.50563107676659e-06, "loss": 0.0065, "step": 72310 }, { "grad_norm": 0.23521339893341064, "learning_rate": 2.4991743597944152e-06, "loss": 0.0082, "step": 72320 }, { "grad_norm": 0.21791864931583405, "learning_rate": 2.492725759414205e-06, "loss": 0.0095, "step": 72330 }, { "grad_norm": 0.17187099158763885, "learning_rate": 2.4862852767278234e-06, "loss": 0.0054, "step": 72340 }, { "grad_norm": 0.1893341988325119, "learning_rate": 2.479852912835784e-06, "loss": 0.0036, "step": 72350 }, { "grad_norm": 0.15900036692619324, "learning_rate": 2.4734286688372075e-06, "loss": 0.0083, "step": 72360 }, { "grad_norm": 0.10629186779260635, "learning_rate": 2.467012545829811e-06, "loss": 0.0142, "step": 72370 }, { "grad_norm": 0.167661651968956, "learning_rate": 2.460604544909945e-06, "loss": 0.0047, "step": 72380 }, { "grad_norm": 0.14635242521762848, "learning_rate": 2.454204667172555e-06, "loss": 0.0046, "step": 72390 }, { "grad_norm": 0.12504923343658447, "learning_rate": 2.447812913711217e-06, "loss": 0.0061, "step": 72400 }, { "grad_norm": 0.07158686965703964, "learning_rate": 2.44142928561808e-06, "loss": 0.0094, "step": 72410 }, { "grad_norm": 0.11944399029016495, "learning_rate": 2.435053783983965e-06, "loss": 0.0064, "step": 72420 }, { "grad_norm": 0.17236527800559998, "learning_rate": 2.4286864098982453e-06, "loss": 0.0059, "step": 72430 }, { "grad_norm": 0.04846172779798508, "learning_rate": 2.4223271644489397e-06, "loss": 0.0057, "step": 72440 }, { "grad_norm": 0.10019838809967041, "learning_rate": 2.415976048722679e-06, "loss": 0.0051, "step": 72450 }, { "grad_norm": 0.11929348856210709, "learning_rate": 2.4096330638046673e-06, "loss": 0.0051, "step": 72460 }, { "grad_norm": 0.13314570486545563, "learning_rate": 2.4032982107787816e-06, "loss": 0.0076, "step": 72470 }, { "grad_norm": 0.09074375778436661, "learning_rate": 2.396971490727451e-06, "loss": 0.0064, "step": 72480 }, { "grad_norm": 0.3456208109855652, "learning_rate": 2.3906529047317493e-06, "loss": 0.0048, "step": 72490 }, { "grad_norm": 0.12606264650821686, "learning_rate": 2.3843424538713465e-06, "loss": 0.0045, "step": 72500 }, { "grad_norm": 0.1061730831861496, "learning_rate": 2.3780401392245298e-06, "loss": 0.006, "step": 72510 }, { "grad_norm": 0.09077824652194977, "learning_rate": 2.3717459618681883e-06, "loss": 0.006, "step": 72520 }, { "grad_norm": 0.19983229041099548, "learning_rate": 2.3654599228778274e-06, "loss": 0.0072, "step": 72530 }, { "grad_norm": 0.16539885103702545, "learning_rate": 2.3591820233275607e-06, "loss": 0.01, "step": 72540 }, { "grad_norm": 0.09319140762090683, "learning_rate": 2.3529122642901024e-06, "loss": 0.008, "step": 72550 }, { "grad_norm": 0.07794355601072311, "learning_rate": 2.3466506468367845e-06, "loss": 0.0126, "step": 72560 }, { "grad_norm": 0.10975934565067291, "learning_rate": 2.3403971720375506e-06, "loss": 0.005, "step": 72570 }, { "grad_norm": 0.11751443892717361, "learning_rate": 2.334151840960952e-06, "loss": 0.0054, "step": 72580 }, { "grad_norm": 0.10025513172149658, "learning_rate": 2.3279146546741347e-06, "loss": 0.0052, "step": 72590 }, { "grad_norm": 0.29829317331314087, "learning_rate": 2.32168561424288e-06, "loss": 0.0053, "step": 72600 }, { "grad_norm": 0.06755772978067398, "learning_rate": 2.3154647207315307e-06, "loss": 0.0083, "step": 72610 }, { "grad_norm": 0.116756372153759, "learning_rate": 2.309251975203103e-06, "loss": 0.0049, "step": 72620 }, { "grad_norm": 0.09318749606609344, "learning_rate": 2.303047378719159e-06, "loss": 0.0128, "step": 72630 }, { "grad_norm": 0.14540258049964905, "learning_rate": 2.2968509323399e-06, "loss": 0.0112, "step": 72640 }, { "grad_norm": 0.20218124985694885, "learning_rate": 2.290662637124147e-06, "loss": 0.0041, "step": 72650 }, { "grad_norm": 0.06547429412603378, "learning_rate": 2.2844824941292807e-06, "loss": 0.0039, "step": 72660 }, { "grad_norm": 0.2766132950782776, "learning_rate": 2.2783105044113406e-06, "loss": 0.0097, "step": 72670 }, { "grad_norm": 0.1431293487548828, "learning_rate": 2.272146669024944e-06, "loss": 0.0047, "step": 72680 }, { "grad_norm": 0.14430004358291626, "learning_rate": 2.26599098902332e-06, "loss": 0.0065, "step": 72690 }, { "grad_norm": 0.0994187518954277, "learning_rate": 2.2598434654583113e-06, "loss": 0.0049, "step": 72700 }, { "grad_norm": 0.22945430874824524, "learning_rate": 2.253704099380355e-06, "loss": 0.0131, "step": 72710 }, { "grad_norm": 0.21752631664276123, "learning_rate": 2.247572891838512e-06, "loss": 0.0079, "step": 72720 }, { "grad_norm": 0.20465178787708282, "learning_rate": 2.241449843880422e-06, "loss": 0.0092, "step": 72730 }, { "grad_norm": 0.1266821026802063, "learning_rate": 2.235334956552354e-06, "loss": 0.0103, "step": 72740 }, { "grad_norm": 0.22542229294776917, "learning_rate": 2.2292282308991775e-06, "loss": 0.0084, "step": 72750 }, { "grad_norm": 0.2510128319263458, "learning_rate": 2.223129667964363e-06, "loss": 0.0065, "step": 72760 }, { "grad_norm": 0.16604754328727722, "learning_rate": 2.2170392687899834e-06, "loss": 0.0048, "step": 72770 }, { "grad_norm": 0.10770231485366821, "learning_rate": 2.210957034416733e-06, "loss": 0.0048, "step": 72780 }, { "grad_norm": 0.1537703573703766, "learning_rate": 2.2048829658838867e-06, "loss": 0.0061, "step": 72790 }, { "grad_norm": 0.10190179944038391, "learning_rate": 2.1988170642293525e-06, "loss": 0.0042, "step": 72800 }, { "grad_norm": 0.05561227351427078, "learning_rate": 2.1927593304896075e-06, "loss": 0.0057, "step": 72810 }, { "grad_norm": 0.19032298028469086, "learning_rate": 2.1867097656997626e-06, "loss": 0.0055, "step": 72820 }, { "grad_norm": 0.12354236096143723, "learning_rate": 2.180668370893524e-06, "loss": 0.0066, "step": 72830 }, { "grad_norm": 0.12612128257751465, "learning_rate": 2.174635147103199e-06, "loss": 0.0065, "step": 72840 }, { "grad_norm": 0.13425017893314362, "learning_rate": 2.1686100953597075e-06, "loss": 0.0066, "step": 72850 }, { "grad_norm": 0.20436355471611023, "learning_rate": 2.1625932166925433e-06, "loss": 0.0061, "step": 72860 }, { "grad_norm": 0.14414989948272705, "learning_rate": 2.1565845121298556e-06, "loss": 0.0069, "step": 72870 }, { "grad_norm": 0.2197299599647522, "learning_rate": 2.150583982698351e-06, "loss": 0.0093, "step": 72880 }, { "grad_norm": 0.08152107149362564, "learning_rate": 2.144591629423359e-06, "loss": 0.004, "step": 72890 }, { "grad_norm": 0.16303984820842743, "learning_rate": 2.138607453328817e-06, "loss": 0.0073, "step": 72900 }, { "grad_norm": 0.10433997213840485, "learning_rate": 2.132631455437234e-06, "loss": 0.0074, "step": 72910 }, { "grad_norm": 0.16001851856708527, "learning_rate": 2.1266636367697714e-06, "loss": 0.0142, "step": 72920 }, { "grad_norm": 0.34096381068229675, "learning_rate": 2.120703998346152e-06, "loss": 0.0053, "step": 72930 }, { "grad_norm": 0.1364302635192871, "learning_rate": 2.1147525411847114e-06, "loss": 0.0086, "step": 72940 }, { "grad_norm": 0.20892849564552307, "learning_rate": 2.108809266302403e-06, "loss": 0.0051, "step": 72950 }, { "grad_norm": 0.20712938904762268, "learning_rate": 2.102874174714764e-06, "loss": 0.0099, "step": 72960 }, { "grad_norm": 0.18639160692691803, "learning_rate": 2.0969472674359393e-06, "loss": 0.0068, "step": 72970 }, { "grad_norm": 0.05865652114152908, "learning_rate": 2.091028545478668e-06, "loss": 0.0076, "step": 72980 }, { "grad_norm": 0.0895497053861618, "learning_rate": 2.08511800985432e-06, "loss": 0.0055, "step": 72990 }, { "grad_norm": 0.16068589687347412, "learning_rate": 2.07921566157282e-06, "loss": 0.0124, "step": 73000 }, { "grad_norm": 0.11670789122581482, "learning_rate": 2.073321501642728e-06, "loss": 0.0045, "step": 73010 }, { "grad_norm": 0.0963829830288887, "learning_rate": 2.0674355310711937e-06, "loss": 0.0092, "step": 73020 }, { "grad_norm": 0.11957364529371262, "learning_rate": 2.0615577508639682e-06, "loss": 0.0042, "step": 73030 }, { "grad_norm": 0.1060912013053894, "learning_rate": 2.055688162025399e-06, "loss": 0.0044, "step": 73040 }, { "grad_norm": 0.1112382784485817, "learning_rate": 2.0498267655584546e-06, "loss": 0.0081, "step": 73050 }, { "grad_norm": 0.12089399248361588, "learning_rate": 2.0439735624646626e-06, "loss": 0.0073, "step": 73060 }, { "grad_norm": 0.08505817502737045, "learning_rate": 2.0381285537442008e-06, "loss": 0.0047, "step": 73070 }, { "grad_norm": 0.24997462332248688, "learning_rate": 2.032291740395803e-06, "loss": 0.0051, "step": 73080 }, { "grad_norm": 0.14914782345294952, "learning_rate": 2.0264631234168276e-06, "loss": 0.0078, "step": 73090 }, { "grad_norm": 0.2529365122318268, "learning_rate": 2.0206427038032326e-06, "loss": 0.0062, "step": 73100 }, { "grad_norm": 0.14103484153747559, "learning_rate": 2.0148304825495457e-06, "loss": 0.004, "step": 73110 }, { "grad_norm": 0.132130965590477, "learning_rate": 2.0090264606489496e-06, "loss": 0.0043, "step": 73120 }, { "grad_norm": 0.14550039172172546, "learning_rate": 2.003230639093162e-06, "loss": 0.0054, "step": 73130 }, { "grad_norm": 0.15329957008361816, "learning_rate": 1.9974430188725525e-06, "loss": 0.0095, "step": 73140 }, { "grad_norm": 0.1356395035982132, "learning_rate": 1.9916636009760513e-06, "loss": 0.0085, "step": 73150 }, { "grad_norm": 0.20752814412117004, "learning_rate": 1.985892386391208e-06, "loss": 0.0117, "step": 73160 }, { "grad_norm": 0.07122992724180222, "learning_rate": 1.980129376104173e-06, "loss": 0.0082, "step": 73170 }, { "grad_norm": 0.14676514267921448, "learning_rate": 1.9743745710996796e-06, "loss": 0.0073, "step": 73180 }, { "grad_norm": 0.07329809665679932, "learning_rate": 1.9686279723610757e-06, "loss": 0.0054, "step": 73190 }, { "grad_norm": 0.12544278800487518, "learning_rate": 1.962889580870281e-06, "loss": 0.0103, "step": 73200 }, { "grad_norm": 0.1610136181116104, "learning_rate": 1.9571593976078385e-06, "loss": 0.0042, "step": 73210 }, { "grad_norm": 0.1112193763256073, "learning_rate": 1.9514374235528876e-06, "loss": 0.0051, "step": 73220 }, { "grad_norm": 0.15053929388523102, "learning_rate": 1.9457236596831408e-06, "loss": 0.0047, "step": 73230 }, { "grad_norm": 0.1027446836233139, "learning_rate": 1.940018106974939e-06, "loss": 0.0113, "step": 73240 }, { "grad_norm": 0.10819617658853531, "learning_rate": 1.9343207664032025e-06, "loss": 0.0049, "step": 73250 }, { "grad_norm": 0.08718249201774597, "learning_rate": 1.928631638941436e-06, "loss": 0.0039, "step": 73260 }, { "grad_norm": 0.0761813148856163, "learning_rate": 1.922950725561784e-06, "loss": 0.0045, "step": 73270 }, { "grad_norm": 0.05715432018041611, "learning_rate": 1.9172780272349312e-06, "loss": 0.0044, "step": 73280 }, { "grad_norm": 0.09075949341058731, "learning_rate": 1.911613544930202e-06, "loss": 0.0054, "step": 73290 }, { "grad_norm": 0.09314845502376556, "learning_rate": 1.9059572796155e-06, "loss": 0.0065, "step": 73300 }, { "grad_norm": 0.14916332066059113, "learning_rate": 1.9003092322573192e-06, "loss": 0.005, "step": 73310 }, { "grad_norm": 0.1185229942202568, "learning_rate": 1.8946694038207647e-06, "loss": 0.0147, "step": 73320 }, { "grad_norm": 0.20060895383358002, "learning_rate": 1.8890377952695215e-06, "loss": 0.0047, "step": 73330 }, { "grad_norm": 0.17961855232715607, "learning_rate": 1.8834144075658865e-06, "loss": 0.0046, "step": 73340 }, { "grad_norm": 0.2248038351535797, "learning_rate": 1.8777992416707357e-06, "loss": 0.0098, "step": 73350 }, { "grad_norm": 0.1368977278470993, "learning_rate": 1.8721922985435458e-06, "loss": 0.0048, "step": 73360 }, { "grad_norm": 0.1098538488149643, "learning_rate": 1.8665935791424006e-06, "loss": 0.0064, "step": 73370 }, { "grad_norm": 0.13430823385715485, "learning_rate": 1.8610030844239512e-06, "loss": 0.0052, "step": 73380 }, { "grad_norm": 0.060752470046281815, "learning_rate": 1.8554208153434838e-06, "loss": 0.0041, "step": 73390 }, { "grad_norm": 0.08274278789758682, "learning_rate": 1.8498467728548296e-06, "loss": 0.0033, "step": 73400 }, { "grad_norm": 0.14723080396652222, "learning_rate": 1.8442809579104547e-06, "loss": 0.0055, "step": 73410 }, { "grad_norm": 0.23637552559375763, "learning_rate": 1.8387233714614039e-06, "loss": 0.0063, "step": 73420 }, { "grad_norm": 0.09347698092460632, "learning_rate": 1.8331740144573174e-06, "loss": 0.0136, "step": 73430 }, { "grad_norm": 0.15559659898281097, "learning_rate": 1.8276328878464255e-06, "loss": 0.005, "step": 73440 }, { "grad_norm": 0.1901736706495285, "learning_rate": 1.82209999257556e-06, "loss": 0.0081, "step": 73450 }, { "grad_norm": 0.11742010712623596, "learning_rate": 1.8165753295901311e-06, "loss": 0.006, "step": 73460 }, { "grad_norm": 0.0916886106133461, "learning_rate": 1.8110588998341616e-06, "loss": 0.0068, "step": 73470 }, { "grad_norm": 0.18374000489711761, "learning_rate": 1.805550704250253e-06, "loss": 0.0183, "step": 73480 }, { "grad_norm": 0.19198480248451233, "learning_rate": 1.8000507437796077e-06, "loss": 0.0055, "step": 73490 }, { "grad_norm": 0.2002488672733307, "learning_rate": 1.7945590193620242e-06, "loss": 0.0058, "step": 73500 }, { "grad_norm": 0.15649402141571045, "learning_rate": 1.7890755319358742e-06, "loss": 0.0071, "step": 73510 }, { "grad_norm": 0.3181346356868744, "learning_rate": 1.7836002824381525e-06, "loss": 0.0071, "step": 73520 }, { "grad_norm": 0.040852636098861694, "learning_rate": 1.7781332718044165e-06, "loss": 0.0038, "step": 73530 }, { "grad_norm": 0.11087425798177719, "learning_rate": 1.772674500968835e-06, "loss": 0.0064, "step": 73540 }, { "grad_norm": 0.2767990231513977, "learning_rate": 1.7672239708641624e-06, "loss": 0.0081, "step": 73550 }, { "grad_norm": 0.18896666169166565, "learning_rate": 1.7617816824217371e-06, "loss": 0.0045, "step": 73560 }, { "grad_norm": 0.11365139484405518, "learning_rate": 1.7563476365715148e-06, "loss": 0.0119, "step": 73570 }, { "grad_norm": 0.14788544178009033, "learning_rate": 1.750921834242003e-06, "loss": 0.01, "step": 73580 }, { "grad_norm": 0.14740200340747833, "learning_rate": 1.7455042763603436e-06, "loss": 0.005, "step": 73590 }, { "grad_norm": 0.20755238831043243, "learning_rate": 1.7400949638522345e-06, "loss": 0.0044, "step": 73600 }, { "grad_norm": 0.16200748085975647, "learning_rate": 1.734693897641987e-06, "loss": 0.0197, "step": 73610 }, { "grad_norm": 0.15834815800189972, "learning_rate": 1.7293010786524955e-06, "loss": 0.0053, "step": 73620 }, { "grad_norm": 0.4203920066356659, "learning_rate": 1.72391650780524e-06, "loss": 0.0088, "step": 73630 }, { "grad_norm": 0.22316348552703857, "learning_rate": 1.718540186020301e-06, "loss": 0.0065, "step": 73640 }, { "grad_norm": 0.1473740190267563, "learning_rate": 1.7131721142163437e-06, "loss": 0.0068, "step": 73650 }, { "grad_norm": 0.18606530129909515, "learning_rate": 1.7078122933106233e-06, "loss": 0.01, "step": 73660 }, { "grad_norm": 0.4115734100341797, "learning_rate": 1.7024607242189905e-06, "loss": 0.0073, "step": 73670 }, { "grad_norm": 0.13881656527519226, "learning_rate": 1.6971174078558749e-06, "loss": 0.0041, "step": 73680 }, { "grad_norm": 0.07159169763326645, "learning_rate": 1.6917823451343073e-06, "loss": 0.0072, "step": 73690 }, { "grad_norm": 0.1225053146481514, "learning_rate": 1.6864555369659142e-06, "loss": 0.0072, "step": 73700 }, { "grad_norm": 0.11753232777118683, "learning_rate": 1.681136984260878e-06, "loss": 0.0084, "step": 73710 }, { "grad_norm": 0.09384652227163315, "learning_rate": 1.6758266879280172e-06, "loss": 0.005, "step": 73720 }, { "grad_norm": 0.1237700879573822, "learning_rate": 1.6705246488746996e-06, "loss": 0.0052, "step": 73730 }, { "grad_norm": 0.06968297064304352, "learning_rate": 1.6652308680069062e-06, "loss": 0.0072, "step": 73740 }, { "grad_norm": 0.1461871862411499, "learning_rate": 1.6599453462292081e-06, "loss": 0.0078, "step": 73750 }, { "grad_norm": 0.3251088559627533, "learning_rate": 1.6546680844447326e-06, "loss": 0.0085, "step": 73760 }, { "grad_norm": 0.15720222890377045, "learning_rate": 1.6493990835552475e-06, "loss": 0.0072, "step": 73770 }, { "grad_norm": 0.06516557931900024, "learning_rate": 1.6441383444610547e-06, "loss": 0.0055, "step": 73780 }, { "grad_norm": 0.04716743156313896, "learning_rate": 1.6388858680610908e-06, "loss": 0.0081, "step": 73790 }, { "grad_norm": 0.25581470131874084, "learning_rate": 1.6336416552528543e-06, "loss": 0.0053, "step": 73800 }, { "grad_norm": 0.11279289424419403, "learning_rate": 1.6284057069324288e-06, "loss": 0.0124, "step": 73810 }, { "grad_norm": 0.14749613404273987, "learning_rate": 1.623178023994504e-06, "loss": 0.0089, "step": 73820 }, { "grad_norm": 0.10636503994464874, "learning_rate": 1.6179586073323483e-06, "loss": 0.0063, "step": 73830 }, { "grad_norm": 0.260578453540802, "learning_rate": 1.612747457837821e-06, "loss": 0.0064, "step": 73840 }, { "grad_norm": 0.1561764031648636, "learning_rate": 1.607544576401354e-06, "loss": 0.0073, "step": 73850 }, { "grad_norm": 0.16746266186237335, "learning_rate": 1.6023499639119754e-06, "loss": 0.0053, "step": 73860 }, { "grad_norm": 0.1571003496646881, "learning_rate": 1.5971636212573138e-06, "loss": 0.0056, "step": 73870 }, { "grad_norm": 0.12381834536790848, "learning_rate": 1.5919855493235714e-06, "loss": 0.0058, "step": 73880 }, { "grad_norm": 0.23498496413230896, "learning_rate": 1.586815748995535e-06, "loss": 0.004, "step": 73890 }, { "grad_norm": 0.10597701370716095, "learning_rate": 1.5816542211565866e-06, "loss": 0.0046, "step": 73900 }, { "grad_norm": 0.08768032491207123, "learning_rate": 1.5765009666886765e-06, "loss": 0.0056, "step": 73910 }, { "grad_norm": 0.24014300107955933, "learning_rate": 1.5713559864723781e-06, "loss": 0.0078, "step": 73920 }, { "grad_norm": 0.23054100573062897, "learning_rate": 1.5662192813868104e-06, "loss": 0.0063, "step": 73930 }, { "grad_norm": 0.10414855927228928, "learning_rate": 1.561090852309699e-06, "loss": 0.0102, "step": 73940 }, { "grad_norm": 0.08724439144134521, "learning_rate": 1.5559707001173651e-06, "loss": 0.0056, "step": 73950 }, { "grad_norm": 0.1657896190881729, "learning_rate": 1.5508588256846757e-06, "loss": 0.0061, "step": 73960 }, { "grad_norm": 0.1490328311920166, "learning_rate": 1.5457552298851319e-06, "loss": 0.0059, "step": 73970 }, { "grad_norm": 0.19011855125427246, "learning_rate": 1.5406599135907918e-06, "loss": 0.0075, "step": 73980 }, { "grad_norm": 0.2784547805786133, "learning_rate": 1.5355728776723088e-06, "loss": 0.0084, "step": 73990 }, { "grad_norm": 0.06524841487407684, "learning_rate": 1.5304941229989155e-06, "loss": 0.0056, "step": 74000 }, { "grad_norm": 0.1585819572210312, "learning_rate": 1.5254236504384345e-06, "loss": 0.0078, "step": 74010 }, { "grad_norm": 0.17501018941402435, "learning_rate": 1.5203614608572726e-06, "loss": 0.0101, "step": 74020 }, { "grad_norm": 0.11761406064033508, "learning_rate": 1.5153075551204044e-06, "loss": 0.0046, "step": 74030 }, { "grad_norm": 0.10188047587871552, "learning_rate": 1.5102619340914225e-06, "loss": 0.0172, "step": 74040 }, { "grad_norm": 0.12708698213100433, "learning_rate": 1.505224598632482e-06, "loss": 0.0054, "step": 74050 }, { "grad_norm": 0.06859676539897919, "learning_rate": 1.5001955496043164e-06, "loss": 0.0118, "step": 74060 }, { "grad_norm": 0.15151207149028778, "learning_rate": 1.4951747878662604e-06, "loss": 0.0057, "step": 74070 }, { "grad_norm": 0.09779619425535202, "learning_rate": 1.4901623142762221e-06, "loss": 0.0055, "step": 74080 }, { "grad_norm": 0.08605344593524933, "learning_rate": 1.4851581296907001e-06, "loss": 0.0107, "step": 74090 }, { "grad_norm": 0.1274985671043396, "learning_rate": 1.4801622349647714e-06, "loss": 0.006, "step": 74100 }, { "grad_norm": 0.1519533395767212, "learning_rate": 1.4751746309520976e-06, "loss": 0.0059, "step": 74110 }, { "grad_norm": 0.13818682730197906, "learning_rate": 1.4701953185049132e-06, "loss": 0.0098, "step": 74120 }, { "grad_norm": 0.11901061236858368, "learning_rate": 1.4652242984740661e-06, "loss": 0.0093, "step": 74130 }, { "grad_norm": 0.24939130246639252, "learning_rate": 1.4602615717089484e-06, "loss": 0.0152, "step": 74140 }, { "grad_norm": 0.1032877191901207, "learning_rate": 1.4553071390575657e-06, "loss": 0.0057, "step": 74150 }, { "grad_norm": 0.13945485651493073, "learning_rate": 1.450361001366496e-06, "loss": 0.0109, "step": 74160 }, { "grad_norm": 0.1647447794675827, "learning_rate": 1.4454231594809021e-06, "loss": 0.0076, "step": 74170 }, { "grad_norm": 0.07898823916912079, "learning_rate": 1.4404936142445036e-06, "loss": 0.0076, "step": 74180 }, { "grad_norm": 0.12406095117330551, "learning_rate": 1.4355723664996546e-06, "loss": 0.0056, "step": 74190 }, { "grad_norm": 0.11905654519796371, "learning_rate": 1.4306594170872433e-06, "loss": 0.0049, "step": 74200 }, { "grad_norm": 0.17244648933410645, "learning_rate": 1.4257547668467598e-06, "loss": 0.0182, "step": 74210 }, { "grad_norm": 0.11583595722913742, "learning_rate": 1.4208584166162886e-06, "loss": 0.0092, "step": 74220 }, { "grad_norm": 0.13372433185577393, "learning_rate": 1.4159703672324553e-06, "loss": 0.0084, "step": 74230 }, { "grad_norm": 0.10657425969839096, "learning_rate": 1.4110906195305252e-06, "loss": 0.0081, "step": 74240 }, { "grad_norm": 0.24167458713054657, "learning_rate": 1.406219174344292e-06, "loss": 0.0054, "step": 74250 }, { "grad_norm": 0.10329338908195496, "learning_rate": 1.401356032506157e-06, "loss": 0.0048, "step": 74260 }, { "grad_norm": 0.07593322545289993, "learning_rate": 1.3965011948471051e-06, "loss": 0.01, "step": 74270 }, { "grad_norm": 0.12075767666101456, "learning_rate": 1.3916546621966842e-06, "loss": 0.0066, "step": 74280 }, { "grad_norm": 0.2648557722568512, "learning_rate": 1.3868164353830482e-06, "loss": 0.0042, "step": 74290 }, { "grad_norm": 0.12270697206258774, "learning_rate": 1.3819865152329082e-06, "loss": 0.0039, "step": 74300 }, { "grad_norm": 0.06844160705804825, "learning_rate": 1.377164902571565e-06, "loss": 0.0105, "step": 74310 }, { "grad_norm": 0.12084870785474777, "learning_rate": 1.372351598222904e-06, "loss": 0.0073, "step": 74320 }, { "grad_norm": 0.09682036191225052, "learning_rate": 1.3675466030093843e-06, "loss": 0.0055, "step": 74330 }, { "grad_norm": 0.28619393706321716, "learning_rate": 1.3627499177520486e-06, "loss": 0.0087, "step": 74340 }, { "grad_norm": 0.1453189104795456, "learning_rate": 1.3579615432705196e-06, "loss": 0.0086, "step": 74350 }, { "grad_norm": 0.11321140825748444, "learning_rate": 1.3531814803829978e-06, "loss": 0.0091, "step": 74360 }, { "grad_norm": 0.09051501005887985, "learning_rate": 1.3484097299062803e-06, "loss": 0.0075, "step": 74370 }, { "grad_norm": 0.056738514453172684, "learning_rate": 1.343646292655698e-06, "loss": 0.0074, "step": 74380 }, { "grad_norm": 0.13014903664588928, "learning_rate": 1.3388911694452277e-06, "loss": 0.0067, "step": 74390 }, { "grad_norm": 0.3006870448589325, "learning_rate": 1.3341443610873638e-06, "loss": 0.0085, "step": 74400 }, { "grad_norm": 0.08009789139032364, "learning_rate": 1.329405868393213e-06, "loss": 0.0049, "step": 74410 }, { "grad_norm": 0.10990506410598755, "learning_rate": 1.3246756921724613e-06, "loss": 0.0092, "step": 74420 }, { "grad_norm": 0.5959733724594116, "learning_rate": 1.3199538332333506e-06, "loss": 0.0058, "step": 74430 }, { "grad_norm": 0.2653404772281647, "learning_rate": 1.3152402923827411e-06, "loss": 0.0086, "step": 74440 }, { "grad_norm": 0.14389795064926147, "learning_rate": 1.3105350704260277e-06, "loss": 0.0056, "step": 74450 }, { "grad_norm": 0.10449519008398056, "learning_rate": 1.305838168167206e-06, "loss": 0.0042, "step": 74460 }, { "grad_norm": 0.2904629409313202, "learning_rate": 1.301149586408862e-06, "loss": 0.0038, "step": 74470 }, { "grad_norm": 0.09343932569026947, "learning_rate": 1.2964693259521322e-06, "loss": 0.0076, "step": 74480 }, { "grad_norm": 0.20337627828121185, "learning_rate": 1.2917973875967548e-06, "loss": 0.0044, "step": 74490 }, { "grad_norm": 0.3025585114955902, "learning_rate": 1.2871337721410249e-06, "loss": 0.01, "step": 74500 }, { "grad_norm": 0.1670227199792862, "learning_rate": 1.2824784803818379e-06, "loss": 0.0087, "step": 74510 }, { "grad_norm": 0.30663540959358215, "learning_rate": 1.2778315131146467e-06, "loss": 0.0049, "step": 74520 }, { "grad_norm": 0.20631606876850128, "learning_rate": 1.2731928711334994e-06, "loss": 0.0053, "step": 74530 }, { "grad_norm": 0.11132983863353729, "learning_rate": 1.2685625552310122e-06, "loss": 0.0044, "step": 74540 }, { "grad_norm": 0.09210878610610962, "learning_rate": 1.263940566198374e-06, "loss": 0.0055, "step": 74550 }, { "grad_norm": 0.07348942011594772, "learning_rate": 1.259326904825353e-06, "loss": 0.0049, "step": 74560 }, { "grad_norm": 0.1236652359366417, "learning_rate": 1.2547215719003137e-06, "loss": 0.0066, "step": 74570 }, { "grad_norm": 0.18831200897693634, "learning_rate": 1.2501245682101703e-06, "loss": 0.006, "step": 74580 }, { "grad_norm": 0.49983569979667664, "learning_rate": 1.2455358945404171e-06, "loss": 0.0067, "step": 74590 }, { "grad_norm": 0.11804670095443726, "learning_rate": 1.2409555516751493e-06, "loss": 0.0048, "step": 74600 }, { "grad_norm": 0.1973053216934204, "learning_rate": 1.2363835403970125e-06, "loss": 0.0064, "step": 74610 }, { "grad_norm": 0.08281877636909485, "learning_rate": 1.2318198614872489e-06, "loss": 0.0062, "step": 74620 }, { "grad_norm": 0.15242703258991241, "learning_rate": 1.2272645157256457e-06, "loss": 0.0048, "step": 74630 }, { "grad_norm": 0.12553781270980835, "learning_rate": 1.222717503890608e-06, "loss": 0.0077, "step": 74640 }, { "grad_norm": 0.3490310311317444, "learning_rate": 1.218178826759081e-06, "loss": 0.0134, "step": 74650 }, { "grad_norm": 0.34833312034606934, "learning_rate": 1.213648485106611e-06, "loss": 0.0072, "step": 74660 }, { "grad_norm": 0.10687897354364395, "learning_rate": 1.2091264797073066e-06, "loss": 0.0164, "step": 74670 }, { "grad_norm": 0.09693226218223572, "learning_rate": 1.2046128113338494e-06, "loss": 0.0178, "step": 74680 }, { "grad_norm": 0.1283983439207077, "learning_rate": 1.200107480757512e-06, "loss": 0.0092, "step": 74690 }, { "grad_norm": 0.09598871320486069, "learning_rate": 1.1956104887481168e-06, "loss": 0.0118, "step": 74700 }, { "grad_norm": 0.2083151936531067, "learning_rate": 1.191121836074094e-06, "loss": 0.0048, "step": 74710 }, { "grad_norm": 0.08132870495319366, "learning_rate": 1.1866415235024186e-06, "loss": 0.0042, "step": 74720 }, { "grad_norm": 0.18754161894321442, "learning_rate": 1.182169551798662e-06, "loss": 0.0103, "step": 74730 }, { "grad_norm": 0.1360727846622467, "learning_rate": 1.177705921726957e-06, "loss": 0.0049, "step": 74740 }, { "grad_norm": 0.18400070071220398, "learning_rate": 1.173250634050016e-06, "loss": 0.0071, "step": 74750 }, { "grad_norm": 0.20814481377601624, "learning_rate": 1.1688036895291354e-06, "loss": 0.007, "step": 74760 }, { "grad_norm": 0.07963085174560547, "learning_rate": 1.1643650889241574e-06, "loss": 0.0043, "step": 74770 }, { "grad_norm": 0.24418297410011292, "learning_rate": 1.1599348329935311e-06, "loss": 0.008, "step": 74780 }, { "grad_norm": 0.11519856005907059, "learning_rate": 1.1555129224942673e-06, "loss": 0.014, "step": 74790 }, { "grad_norm": 0.11441359668970108, "learning_rate": 1.1510993581819396e-06, "loss": 0.0086, "step": 74800 }, { "grad_norm": 1.068347454071045, "learning_rate": 1.1466941408107112e-06, "loss": 0.0149, "step": 74810 }, { "grad_norm": 0.1415995955467224, "learning_rate": 1.1422972711333247e-06, "loss": 0.0053, "step": 74820 }, { "grad_norm": 0.21267268061637878, "learning_rate": 1.1379087499010565e-06, "loss": 0.006, "step": 74830 }, { "grad_norm": 0.14316634833812714, "learning_rate": 1.1335285778638127e-06, "loss": 0.0045, "step": 74840 }, { "grad_norm": 0.14722301065921783, "learning_rate": 1.1291567557700333e-06, "loss": 0.0058, "step": 74850 }, { "grad_norm": 0.21770577132701874, "learning_rate": 1.124793284366743e-06, "loss": 0.007, "step": 74860 }, { "grad_norm": 0.11366604268550873, "learning_rate": 1.1204381643995455e-06, "loss": 0.0054, "step": 74870 }, { "grad_norm": 0.09068675339221954, "learning_rate": 1.116091396612595e-06, "loss": 0.0058, "step": 74880 }, { "grad_norm": 0.07448213547468185, "learning_rate": 1.1117529817486538e-06, "loss": 0.006, "step": 74890 }, { "grad_norm": 0.2742801606655121, "learning_rate": 1.107422920549034e-06, "loss": 0.0052, "step": 74900 }, { "grad_norm": 0.246832937002182, "learning_rate": 1.1031012137536157e-06, "loss": 0.0057, "step": 74910 }, { "grad_norm": 0.10368268191814423, "learning_rate": 1.0987878621008695e-06, "loss": 0.0056, "step": 74920 }, { "grad_norm": 0.16459859907627106, "learning_rate": 1.0944828663278284e-06, "loss": 0.007, "step": 74930 }, { "grad_norm": 0.11877844482660294, "learning_rate": 1.0901862271700925e-06, "loss": 0.0068, "step": 74940 }, { "grad_norm": 0.09010549634695053, "learning_rate": 1.0858979453618467e-06, "loss": 0.008, "step": 74950 }, { "grad_norm": 0.23400266468524933, "learning_rate": 1.0816180216358441e-06, "loss": 0.006, "step": 74960 }, { "grad_norm": 0.11287138611078262, "learning_rate": 1.0773464567233937e-06, "loss": 0.0049, "step": 74970 }, { "grad_norm": 0.23295387625694275, "learning_rate": 1.0730832513543953e-06, "loss": 0.0054, "step": 74980 }, { "grad_norm": 0.21932242810726166, "learning_rate": 1.0688284062573218e-06, "loss": 0.0093, "step": 74990 }, { "grad_norm": 0.13079765439033508, "learning_rate": 1.0645819221591969e-06, "loss": 0.0041, "step": 75000 }, { "grad_norm": 0.1493338942527771, "learning_rate": 1.0603437997856347e-06, "loss": 0.0058, "step": 75010 }, { "grad_norm": 0.08125023543834686, "learning_rate": 1.0561140398608228e-06, "loss": 0.0048, "step": 75020 }, { "grad_norm": 0.09724394232034683, "learning_rate": 1.0518926431074937e-06, "loss": 0.0059, "step": 75030 }, { "grad_norm": 0.09262627363204956, "learning_rate": 1.0476796102469877e-06, "loss": 0.0059, "step": 75040 }, { "grad_norm": 0.17752477526664734, "learning_rate": 1.0434749419991786e-06, "loss": 0.0064, "step": 75050 }, { "grad_norm": 0.10927051305770874, "learning_rate": 1.0392786390825415e-06, "loss": 0.0077, "step": 75060 }, { "grad_norm": 0.25836578011512756, "learning_rate": 1.0350907022141087e-06, "loss": 0.0068, "step": 75070 }, { "grad_norm": 0.19198469817638397, "learning_rate": 1.0309111321094744e-06, "loss": 0.0061, "step": 75080 }, { "grad_norm": 0.05731375887989998, "learning_rate": 1.026739929482834e-06, "loss": 0.004, "step": 75090 }, { "grad_norm": 0.17930512130260468, "learning_rate": 1.022577095046906e-06, "loss": 0.0075, "step": 75100 }, { "grad_norm": 0.21992462873458862, "learning_rate": 1.0184226295130217e-06, "loss": 0.0084, "step": 75110 }, { "grad_norm": 0.23578540980815887, "learning_rate": 1.0142765335910575e-06, "loss": 0.0096, "step": 75120 }, { "grad_norm": 0.2068357616662979, "learning_rate": 1.0101388079894746e-06, "loss": 0.005, "step": 75130 }, { "grad_norm": 0.05458296462893486, "learning_rate": 1.0060094534152908e-06, "loss": 0.0103, "step": 75140 }, { "grad_norm": 0.2392917424440384, "learning_rate": 1.0018884705741028e-06, "loss": 0.0046, "step": 75150 }, { "grad_norm": 0.186258003115654, "learning_rate": 9.977758601700805e-07, "loss": 0.01, "step": 75160 }, { "grad_norm": 0.11757984757423401, "learning_rate": 9.93671622905945e-07, "loss": 0.0143, "step": 75170 }, { "grad_norm": 0.10810183733701706, "learning_rate": 9.895757594830024e-07, "loss": 0.0042, "step": 75180 }, { "grad_norm": 0.15439163148403168, "learning_rate": 9.854882706011204e-07, "loss": 0.0047, "step": 75190 }, { "grad_norm": 0.10969268530607224, "learning_rate": 9.814091569587513e-07, "loss": 0.0046, "step": 75200 }, { "grad_norm": 0.23558467626571655, "learning_rate": 9.77338419252888e-07, "loss": 0.0101, "step": 75210 }, { "grad_norm": 0.08694876730442047, "learning_rate": 9.732760581791289e-07, "loss": 0.0065, "step": 75220 }, { "grad_norm": 0.21182358264923096, "learning_rate": 9.692220744315972e-07, "loss": 0.0054, "step": 75230 }, { "grad_norm": 0.08506401628255844, "learning_rate": 9.651764687030162e-07, "loss": 0.0046, "step": 75240 }, { "grad_norm": 0.13441774249076843, "learning_rate": 9.611392416846776e-07, "loss": 0.0046, "step": 75250 }, { "grad_norm": 0.18384680151939392, "learning_rate": 9.571103940664238e-07, "loss": 0.0067, "step": 75260 }, { "grad_norm": 0.10694798082113266, "learning_rate": 9.530899265366822e-07, "loss": 0.0086, "step": 75270 }, { "grad_norm": 0.10620048642158508, "learning_rate": 9.490778397824307e-07, "loss": 0.0073, "step": 75280 }, { "grad_norm": 0.11033564060926437, "learning_rate": 9.450741344892322e-07, "loss": 0.0055, "step": 75290 }, { "grad_norm": 0.19305622577667236, "learning_rate": 9.410788113412116e-07, "loss": 0.0056, "step": 75300 }, { "grad_norm": 0.056553784757852554, "learning_rate": 9.370918710210563e-07, "loss": 0.006, "step": 75310 }, { "grad_norm": 0.06897422671318054, "learning_rate": 9.331133142100268e-07, "loss": 0.006, "step": 75320 }, { "grad_norm": 0.12993788719177246, "learning_rate": 9.291431415879459e-07, "loss": 0.0039, "step": 75330 }, { "grad_norm": 0.1000954806804657, "learning_rate": 9.251813538332155e-07, "loss": 0.007, "step": 75340 }, { "grad_norm": 0.35297977924346924, "learning_rate": 9.212279516227884e-07, "loss": 0.0066, "step": 75350 }, { "grad_norm": 0.1363029032945633, "learning_rate": 9.172829356322021e-07, "loss": 0.0046, "step": 75360 }, { "grad_norm": 0.056498315185308456, "learning_rate": 9.133463065355452e-07, "loss": 0.0069, "step": 75370 }, { "grad_norm": 0.1160896047949791, "learning_rate": 9.094180650054796e-07, "loss": 0.0072, "step": 75380 }, { "grad_norm": 0.27496829628944397, "learning_rate": 9.054982117132404e-07, "loss": 0.0066, "step": 75390 }, { "grad_norm": 0.15358039736747742, "learning_rate": 9.015867473286143e-07, "loss": 0.0075, "step": 75400 }, { "grad_norm": 0.17686574161052704, "learning_rate": 8.976836725199778e-07, "loss": 0.0051, "step": 75410 }, { "grad_norm": 0.14324957132339478, "learning_rate": 8.937889879542416e-07, "loss": 0.0048, "step": 75420 }, { "grad_norm": 0.1240425631403923, "learning_rate": 8.899026942969068e-07, "loss": 0.0052, "step": 75430 }, { "grad_norm": 0.1902248114347458, "learning_rate": 8.860247922120424e-07, "loss": 0.0062, "step": 75440 }, { "grad_norm": 0.2737041115760803, "learning_rate": 8.821552823622737e-07, "loss": 0.0066, "step": 75450 }, { "grad_norm": 0.09981447458267212, "learning_rate": 8.782941654087884e-07, "loss": 0.005, "step": 75460 }, { "grad_norm": 0.3873743712902069, "learning_rate": 8.744414420113478e-07, "loss": 0.0081, "step": 75470 }, { "grad_norm": 0.10052957385778427, "learning_rate": 8.705971128282753e-07, "loss": 0.0045, "step": 75480 }, { "grad_norm": 0.09932845830917358, "learning_rate": 8.66761178516473e-07, "loss": 0.0061, "step": 75490 }, { "grad_norm": 0.1666150540113449, "learning_rate": 8.629336397313781e-07, "loss": 0.0049, "step": 75500 }, { "grad_norm": 0.24790769815444946, "learning_rate": 8.591144971270282e-07, "loss": 0.0072, "step": 75510 }, { "grad_norm": 0.07823685556650162, "learning_rate": 8.553037513560069e-07, "loss": 0.0126, "step": 75520 }, { "grad_norm": 0.14548300206661224, "learning_rate": 8.515014030694546e-07, "loss": 0.0046, "step": 75530 }, { "grad_norm": 0.2707090377807617, "learning_rate": 8.47707452917107e-07, "loss": 0.0049, "step": 75540 }, { "grad_norm": 0.15277691185474396, "learning_rate": 8.439219015472343e-07, "loss": 0.0091, "step": 75550 }, { "grad_norm": 0.21721941232681274, "learning_rate": 8.401447496066861e-07, "loss": 0.0048, "step": 75560 }, { "grad_norm": 0.10643897205591202, "learning_rate": 8.363759977408792e-07, "loss": 0.0049, "step": 75570 }, { "grad_norm": 0.2475498765707016, "learning_rate": 8.326156465937817e-07, "loss": 0.0046, "step": 75580 }, { "grad_norm": 0.3894655704498291, "learning_rate": 8.28863696807941e-07, "loss": 0.0155, "step": 75590 }, { "grad_norm": 0.0458364300429821, "learning_rate": 8.251201490244609e-07, "loss": 0.0076, "step": 75600 }, { "grad_norm": 0.11220157146453857, "learning_rate": 8.213850038830129e-07, "loss": 0.0043, "step": 75610 }, { "grad_norm": 0.20310820639133453, "learning_rate": 8.176582620218309e-07, "loss": 0.0053, "step": 75620 }, { "grad_norm": 0.15019617974758148, "learning_rate": 8.139399240777057e-07, "loss": 0.0048, "step": 75630 }, { "grad_norm": 0.0959615558385849, "learning_rate": 8.102299906860122e-07, "loss": 0.0087, "step": 75640 }, { "grad_norm": 0.31970521807670593, "learning_rate": 8.065284624806657e-07, "loss": 0.0125, "step": 75650 }, { "grad_norm": 0.0926356166601181, "learning_rate": 8.028353400941601e-07, "loss": 0.0048, "step": 75660 }, { "grad_norm": 0.08028669655323029, "learning_rate": 7.991506241575575e-07, "loss": 0.0104, "step": 75670 }, { "grad_norm": 0.30189529061317444, "learning_rate": 7.954743153004541e-07, "loss": 0.0081, "step": 75680 }, { "grad_norm": 0.07335628569126129, "learning_rate": 7.918064141510528e-07, "loss": 0.0068, "step": 75690 }, { "grad_norm": 0.08335200697183609, "learning_rate": 7.881469213360859e-07, "loss": 0.0099, "step": 75700 }, { "grad_norm": 0.20839840173721313, "learning_rate": 7.844958374808642e-07, "loss": 0.0044, "step": 75710 }, { "grad_norm": 0.04937035217881203, "learning_rate": 7.808531632092608e-07, "loss": 0.0057, "step": 75720 }, { "grad_norm": 0.08098378032445908, "learning_rate": 7.772188991436946e-07, "loss": 0.0071, "step": 75730 }, { "grad_norm": 0.1422397494316101, "learning_rate": 7.735930459051799e-07, "loss": 0.006, "step": 75740 }, { "grad_norm": 0.1478758007287979, "learning_rate": 7.699756041132655e-07, "loss": 0.0073, "step": 75750 }, { "grad_norm": 0.10041499882936478, "learning_rate": 7.663665743860793e-07, "loss": 0.006, "step": 75760 }, { "grad_norm": 0.28259190917015076, "learning_rate": 7.627659573403001e-07, "loss": 0.0052, "step": 75770 }, { "grad_norm": 0.05908133089542389, "learning_rate": 7.591737535911802e-07, "loss": 0.0049, "step": 75780 }, { "grad_norm": 0.044790711253881454, "learning_rate": 7.555899637525288e-07, "loss": 0.0061, "step": 75790 }, { "grad_norm": 0.07482421398162842, "learning_rate": 7.520145884367058e-07, "loss": 0.0037, "step": 75800 }, { "grad_norm": 0.1462910771369934, "learning_rate": 7.484476282546615e-07, "loss": 0.0051, "step": 75810 }, { "grad_norm": 0.18740633130073547, "learning_rate": 7.448890838158806e-07, "loss": 0.0103, "step": 75820 }, { "grad_norm": 0.16731226444244385, "learning_rate": 7.413389557284267e-07, "loss": 0.0067, "step": 75830 }, { "grad_norm": 0.20755480229854584, "learning_rate": 7.377972445989201e-07, "loss": 0.0064, "step": 75840 }, { "grad_norm": 0.14456380903720856, "learning_rate": 7.342639510325377e-07, "loss": 0.0068, "step": 75850 }, { "grad_norm": 0.2861665189266205, "learning_rate": 7.307390756330246e-07, "loss": 0.0098, "step": 75860 }, { "grad_norm": 0.2092103809118271, "learning_rate": 7.272226190026876e-07, "loss": 0.0038, "step": 75870 }, { "grad_norm": 0.15226991474628448, "learning_rate": 7.237145817423907e-07, "loss": 0.0039, "step": 75880 }, { "grad_norm": 0.24583761394023895, "learning_rate": 7.202149644515654e-07, "loss": 0.0058, "step": 75890 }, { "grad_norm": 0.10373135656118393, "learning_rate": 7.167237677281946e-07, "loss": 0.0046, "step": 75900 }, { "grad_norm": 0.21227237582206726, "learning_rate": 7.132409921688288e-07, "loss": 0.007, "step": 75910 }, { "grad_norm": 0.06824734061956406, "learning_rate": 7.09766638368592e-07, "loss": 0.0045, "step": 75920 }, { "grad_norm": 0.1366325467824936, "learning_rate": 7.063007069211313e-07, "loss": 0.0044, "step": 75930 }, { "grad_norm": 0.13517242670059204, "learning_rate": 7.028431984187067e-07, "loss": 0.0048, "step": 75940 }, { "grad_norm": 0.06273990869522095, "learning_rate": 6.993941134520898e-07, "loss": 0.0047, "step": 75950 }, { "grad_norm": 0.11551689356565475, "learning_rate": 6.959534526106537e-07, "loss": 0.0062, "step": 75960 }, { "grad_norm": 0.1392170935869217, "learning_rate": 6.925212164822948e-07, "loss": 0.0043, "step": 75970 }, { "grad_norm": 0.1342310607433319, "learning_rate": 6.890974056535049e-07, "loss": 0.009, "step": 75980 }, { "grad_norm": 0.40276411175727844, "learning_rate": 6.856820207093106e-07, "loss": 0.005, "step": 75990 }, { "grad_norm": 0.12386675179004669, "learning_rate": 6.822750622333063e-07, "loss": 0.0041, "step": 76000 }, { "grad_norm": 0.20783424377441406, "learning_rate": 6.788765308076539e-07, "loss": 0.0062, "step": 76010 }, { "grad_norm": 0.17318810522556305, "learning_rate": 6.754864270130668e-07, "loss": 0.0047, "step": 76020 }, { "grad_norm": 0.13114716112613678, "learning_rate": 6.721047514288203e-07, "loss": 0.0038, "step": 76030 }, { "grad_norm": 0.1922919899225235, "learning_rate": 6.687315046327469e-07, "loss": 0.0054, "step": 76040 }, { "grad_norm": 0.30763015151023865, "learning_rate": 6.65366687201252e-07, "loss": 0.0078, "step": 76050 }, { "grad_norm": 0.1358414590358734, "learning_rate": 6.620102997092814e-07, "loss": 0.0097, "step": 76060 }, { "grad_norm": 0.2898128628730774, "learning_rate": 6.586623427303596e-07, "loss": 0.0067, "step": 76070 }, { "grad_norm": 0.07762596756219864, "learning_rate": 6.553228168365455e-07, "loss": 0.008, "step": 76080 }, { "grad_norm": 0.2685689330101013, "learning_rate": 6.519917225984884e-07, "loss": 0.011, "step": 76090 }, { "grad_norm": 0.1617220640182495, "learning_rate": 6.486690605853718e-07, "loss": 0.0061, "step": 76100 }, { "grad_norm": 0.15487487614154816, "learning_rate": 6.453548313649527e-07, "loss": 0.0065, "step": 76110 }, { "grad_norm": 0.34963858127593994, "learning_rate": 6.420490355035446e-07, "loss": 0.0075, "step": 76120 }, { "grad_norm": 0.06964993476867676, "learning_rate": 6.387516735660071e-07, "loss": 0.0048, "step": 76130 }, { "grad_norm": 0.13597899675369263, "learning_rate": 6.35462746115778e-07, "loss": 0.0047, "step": 76140 }, { "grad_norm": 0.1689879149198532, "learning_rate": 6.321822537148358e-07, "loss": 0.0061, "step": 76150 }, { "grad_norm": 0.23550361394882202, "learning_rate": 6.289101969237432e-07, "loss": 0.0061, "step": 76160 }, { "grad_norm": 0.2139180451631546, "learning_rate": 6.256465763015918e-07, "loss": 0.0076, "step": 76170 }, { "grad_norm": 0.10944674164056778, "learning_rate": 6.22391392406052e-07, "loss": 0.0048, "step": 76180 }, { "grad_norm": 0.14815981686115265, "learning_rate": 6.191446457933403e-07, "loss": 0.0043, "step": 76190 }, { "grad_norm": 0.06958211213350296, "learning_rate": 6.159063370182406e-07, "loss": 0.0055, "step": 76200 }, { "grad_norm": 0.09175439924001694, "learning_rate": 6.126764666340879e-07, "loss": 0.0039, "step": 76210 }, { "grad_norm": 0.0626639574766159, "learning_rate": 6.094550351927852e-07, "loss": 0.0053, "step": 76220 }, { "grad_norm": 0.11395915597677231, "learning_rate": 6.062420432447757e-07, "loss": 0.0069, "step": 76230 }, { "grad_norm": 0.1294795572757721, "learning_rate": 6.030374913390813e-07, "loss": 0.0042, "step": 76240 }, { "grad_norm": 0.2572122812271118, "learning_rate": 5.998413800232694e-07, "loss": 0.0073, "step": 76250 }, { "grad_norm": 0.09779465198516846, "learning_rate": 5.966537098434755e-07, "loss": 0.0057, "step": 76260 }, { "grad_norm": 0.13247202336788177, "learning_rate": 5.934744813443694e-07, "loss": 0.0053, "step": 76270 }, { "grad_norm": 0.05701529234647751, "learning_rate": 5.903036950692054e-07, "loss": 0.0088, "step": 76280 }, { "grad_norm": 0.08702075481414795, "learning_rate": 5.871413515597835e-07, "loss": 0.0075, "step": 76290 }, { "grad_norm": 0.11226821690797806, "learning_rate": 5.839874513564547e-07, "loss": 0.005, "step": 76300 }, { "grad_norm": 0.1442432850599289, "learning_rate": 5.808419949981436e-07, "loss": 0.0071, "step": 76310 }, { "grad_norm": 0.10552451014518738, "learning_rate": 5.777049830223257e-07, "loss": 0.0079, "step": 76320 }, { "grad_norm": 0.08277374505996704, "learning_rate": 5.745764159650114e-07, "loss": 0.0035, "step": 76330 }, { "grad_norm": 0.11567513644695282, "learning_rate": 5.71456294360806e-07, "loss": 0.006, "step": 76340 }, { "grad_norm": 0.41710367798805237, "learning_rate": 5.683446187428443e-07, "loss": 0.0084, "step": 76350 }, { "grad_norm": 0.1417423039674759, "learning_rate": 5.652413896428288e-07, "loss": 0.0066, "step": 76360 }, { "grad_norm": 0.18534263968467712, "learning_rate": 5.621466075910131e-07, "loss": 0.0068, "step": 76370 }, { "grad_norm": 0.04607655480504036, "learning_rate": 5.590602731162187e-07, "loss": 0.0051, "step": 76380 }, { "grad_norm": 0.25136250257492065, "learning_rate": 5.559823867458125e-07, "loss": 0.0051, "step": 76390 }, { "grad_norm": 0.08411040902137756, "learning_rate": 5.529129490057128e-07, "loss": 0.0082, "step": 76400 }, { "grad_norm": 0.17666010558605194, "learning_rate": 5.498519604204167e-07, "loss": 0.0059, "step": 76410 }, { "grad_norm": 0.5098959803581238, "learning_rate": 5.467994215129501e-07, "loss": 0.0066, "step": 76420 }, { "grad_norm": 0.2669893801212311, "learning_rate": 5.437553328049183e-07, "loss": 0.0072, "step": 76430 }, { "grad_norm": 0.07161524146795273, "learning_rate": 5.407196948164661e-07, "loss": 0.0038, "step": 76440 }, { "grad_norm": 0.10626658797264099, "learning_rate": 5.376925080663009e-07, "loss": 0.0068, "step": 76450 }, { "grad_norm": 0.08366137742996216, "learning_rate": 5.346737730716977e-07, "loss": 0.0101, "step": 76460 }, { "grad_norm": 0.09181694686412811, "learning_rate": 5.316634903484607e-07, "loss": 0.0049, "step": 76470 }, { "grad_norm": 0.14351509511470795, "learning_rate": 5.286616604109728e-07, "loss": 0.0039, "step": 76480 }, { "grad_norm": 0.14164292812347412, "learning_rate": 5.256682837721627e-07, "loss": 0.0143, "step": 76490 }, { "grad_norm": 0.17770318686962128, "learning_rate": 5.226833609435156e-07, "loss": 0.0046, "step": 76500 }, { "grad_norm": 0.10277988016605377, "learning_rate": 5.19706892435079e-07, "loss": 0.0106, "step": 76510 }, { "grad_norm": 0.2368331253528595, "learning_rate": 5.167388787554406e-07, "loss": 0.0072, "step": 76520 }, { "grad_norm": 0.13147513568401337, "learning_rate": 5.137793204117614e-07, "loss": 0.0088, "step": 76530 }, { "grad_norm": 0.25278759002685547, "learning_rate": 5.108282179097479e-07, "loss": 0.0076, "step": 76540 }, { "grad_norm": 0.0807400792837143, "learning_rate": 5.078855717536523e-07, "loss": 0.0051, "step": 76550 }, { "grad_norm": 0.1795891970396042, "learning_rate": 5.049513824463059e-07, "loss": 0.005, "step": 76560 }, { "grad_norm": 0.2588193118572235, "learning_rate": 5.020256504890742e-07, "loss": 0.006, "step": 76570 }, { "grad_norm": 0.5412664413452148, "learning_rate": 4.991083763818849e-07, "loss": 0.0082, "step": 76580 }, { "grad_norm": 0.2547723352909088, "learning_rate": 4.961995606232228e-07, "loss": 0.0103, "step": 76590 }, { "grad_norm": 0.1485670804977417, "learning_rate": 4.932992037101236e-07, "loss": 0.005, "step": 76600 }, { "grad_norm": 0.16126468777656555, "learning_rate": 4.904073061381798e-07, "loss": 0.0108, "step": 76610 }, { "grad_norm": 0.13753755390644073, "learning_rate": 4.875238684015349e-07, "loss": 0.0057, "step": 76620 }, { "grad_norm": 0.12315277755260468, "learning_rate": 4.846488909928948e-07, "loss": 0.0049, "step": 76630 }, { "grad_norm": 0.12036009132862091, "learning_rate": 4.81782374403511e-07, "loss": 0.0131, "step": 76640 }, { "grad_norm": 0.06870155781507492, "learning_rate": 4.789243191231918e-07, "loss": 0.0052, "step": 76650 }, { "grad_norm": 0.13440369069576263, "learning_rate": 4.76074725640302e-07, "loss": 0.0128, "step": 76660 }, { "grad_norm": 0.2625438868999481, "learning_rate": 4.732335944417632e-07, "loss": 0.0131, "step": 76670 }, { "grad_norm": 0.21252429485321045, "learning_rate": 4.704009260130371e-07, "loss": 0.0068, "step": 76680 }, { "grad_norm": 0.15404725074768066, "learning_rate": 4.675767208381587e-07, "loss": 0.0054, "step": 76690 }, { "grad_norm": 0.1595609188079834, "learning_rate": 4.647609793997032e-07, "loss": 0.0057, "step": 76700 }, { "grad_norm": 0.10415266454219818, "learning_rate": 4.619537021788023e-07, "loss": 0.0042, "step": 76710 }, { "grad_norm": 0.12379235029220581, "learning_rate": 4.5915488965515005e-07, "loss": 0.0065, "step": 76720 }, { "grad_norm": 0.15041594207286835, "learning_rate": 4.563645423069807e-07, "loss": 0.0051, "step": 76730 }, { "grad_norm": 0.21785087883472443, "learning_rate": 4.535826606110849e-07, "loss": 0.0049, "step": 76740 }, { "grad_norm": 0.13901491463184357, "learning_rate": 4.508092450428214e-07, "loss": 0.0063, "step": 76750 }, { "grad_norm": 0.09261485934257507, "learning_rate": 4.480442960760778e-07, "loss": 0.0097, "step": 76760 }, { "grad_norm": 0.23560504615306854, "learning_rate": 4.4528781418332053e-07, "loss": 0.0043, "step": 76770 }, { "grad_norm": 0.23554666340351105, "learning_rate": 4.4253979983555073e-07, "loss": 0.011, "step": 76780 }, { "grad_norm": 0.08832169324159622, "learning_rate": 4.3980025350233154e-07, "loss": 0.0051, "step": 76790 }, { "grad_norm": 0.11589669436216354, "learning_rate": 4.370691756517664e-07, "loss": 0.0046, "step": 76800 }, { "grad_norm": 0.1452094316482544, "learning_rate": 4.3434656675053753e-07, "loss": 0.0045, "step": 76810 }, { "grad_norm": 0.24084950983524323, "learning_rate": 4.3163242726385613e-07, "loss": 0.0048, "step": 76820 }, { "grad_norm": 0.08225696533918381, "learning_rate": 4.289267576554956e-07, "loss": 0.0068, "step": 76830 }, { "grad_norm": 0.1981358528137207, "learning_rate": 4.262295583877807e-07, "loss": 0.0037, "step": 76840 }, { "grad_norm": 0.1269686222076416, "learning_rate": 4.235408299215815e-07, "loss": 0.0074, "step": 76850 }, { "grad_norm": 0.06507909297943115, "learning_rate": 4.2086057271634173e-07, "loss": 0.0038, "step": 76860 }, { "grad_norm": 0.11349983513355255, "learning_rate": 4.181887872300394e-07, "loss": 0.0049, "step": 76870 }, { "grad_norm": 0.0721784308552742, "learning_rate": 4.155254739192038e-07, "loss": 0.0039, "step": 76880 }, { "grad_norm": 0.12477954477071762, "learning_rate": 4.128706332389265e-07, "loss": 0.0135, "step": 76890 }, { "grad_norm": 0.30938172340393066, "learning_rate": 4.1022426564284453e-07, "loss": 0.0048, "step": 76900 }, { "grad_norm": 0.2882131338119507, "learning_rate": 4.075863715831574e-07, "loss": 0.0052, "step": 76910 }, { "grad_norm": 0.12197790294885635, "learning_rate": 4.0495695151059887e-07, "loss": 0.0074, "step": 76920 }, { "grad_norm": 0.2374269664287567, "learning_rate": 4.023360058744763e-07, "loss": 0.0057, "step": 76930 }, { "grad_norm": 0.18877500295639038, "learning_rate": 3.997235351226314e-07, "loss": 0.0033, "step": 76940 }, { "grad_norm": 0.10217658430337906, "learning_rate": 3.971195397014571e-07, "loss": 0.0064, "step": 76950 }, { "grad_norm": 0.11971528828144073, "learning_rate": 3.9452402005591417e-07, "loss": 0.0092, "step": 76960 }, { "grad_norm": 0.09960142523050308, "learning_rate": 3.9193697662950334e-07, "loss": 0.0044, "step": 76970 }, { "grad_norm": 0.45326411724090576, "learning_rate": 3.893584098642822e-07, "loss": 0.0121, "step": 76980 }, { "grad_norm": 0.09424950927495956, "learning_rate": 3.867883202008538e-07, "loss": 0.0066, "step": 76990 }, { "grad_norm": 0.1313074678182602, "learning_rate": 3.842267080783779e-07, "loss": 0.0057, "step": 77000 }, { "grad_norm": 0.146855890750885, "learning_rate": 3.816735739345656e-07, "loss": 0.0059, "step": 77010 }, { "grad_norm": 0.12259719520807266, "learning_rate": 3.791289182056679e-07, "loss": 0.0057, "step": 77020 }, { "grad_norm": 0.0775463879108429, "learning_rate": 3.7659274132650913e-07, "loss": 0.0062, "step": 77030 }, { "grad_norm": 0.09883712977170944, "learning_rate": 3.7406504373044826e-07, "loss": 0.007, "step": 77040 }, { "grad_norm": 0.16758348047733307, "learning_rate": 3.7154582584939533e-07, "loss": 0.004, "step": 77050 }, { "grad_norm": 0.1516275703907013, "learning_rate": 3.690350881138227e-07, "loss": 0.0056, "step": 77060 }, { "grad_norm": 0.12892182171344757, "learning_rate": 3.665328309527427e-07, "loss": 0.0049, "step": 77070 }, { "grad_norm": 0.08778209984302521, "learning_rate": 3.640390547937245e-07, "loss": 0.0065, "step": 77080 }, { "grad_norm": 0.13974237442016602, "learning_rate": 3.615537600628882e-07, "loss": 0.0083, "step": 77090 }, { "grad_norm": 0.18684454262256622, "learning_rate": 3.5907694718489427e-07, "loss": 0.0065, "step": 77100 }, { "grad_norm": 0.12394613027572632, "learning_rate": 3.566086165829707e-07, "loss": 0.0097, "step": 77110 }, { "grad_norm": 0.3822467029094696, "learning_rate": 3.5414876867888024e-07, "loss": 0.0079, "step": 77120 }, { "grad_norm": 0.20911160111427307, "learning_rate": 3.5169740389295326e-07, "loss": 0.0073, "step": 77130 }, { "grad_norm": 0.15281637012958527, "learning_rate": 3.4925452264405466e-07, "loss": 0.0053, "step": 77140 }, { "grad_norm": 0.15812326967716217, "learning_rate": 3.468201253496062e-07, "loss": 0.0049, "step": 77150 }, { "grad_norm": 0.22803358733654022, "learning_rate": 3.443942124255861e-07, "loss": 0.0083, "step": 77160 }, { "grad_norm": 0.18184825778007507, "learning_rate": 3.4197678428650183e-07, "loss": 0.0066, "step": 77170 }, { "grad_norm": 0.08325747400522232, "learning_rate": 3.3956784134544504e-07, "loss": 0.0056, "step": 77180 }, { "grad_norm": 0.05938224121928215, "learning_rate": 3.3716738401402547e-07, "loss": 0.0044, "step": 77190 }, { "grad_norm": 0.18606412410736084, "learning_rate": 3.3477541270241495e-07, "loss": 0.0057, "step": 77200 }, { "grad_norm": 0.0961214080452919, "learning_rate": 3.3239192781934215e-07, "loss": 0.005, "step": 77210 }, { "grad_norm": 0.2559587359428406, "learning_rate": 3.3001692977207563e-07, "loss": 0.0048, "step": 77220 }, { "grad_norm": 0.047294843941926956, "learning_rate": 3.276504189664409e-07, "loss": 0.0058, "step": 77230 }, { "grad_norm": 0.120116226375103, "learning_rate": 3.252923958068088e-07, "loss": 0.0059, "step": 77240 }, { "grad_norm": 0.08288370817899704, "learning_rate": 3.2294286069609046e-07, "loss": 0.0057, "step": 77250 }, { "grad_norm": 0.09800540655851364, "learning_rate": 3.2060181403577583e-07, "loss": 0.0059, "step": 77260 }, { "grad_norm": 0.21933086216449738, "learning_rate": 3.1826925622587823e-07, "loss": 0.0054, "step": 77270 }, { "grad_norm": 0.07673896849155426, "learning_rate": 3.159451876649622e-07, "loss": 0.0066, "step": 77280 }, { "grad_norm": 0.17046764492988586, "learning_rate": 3.1362960875015444e-07, "loss": 0.0046, "step": 77290 }, { "grad_norm": 0.27380552887916565, "learning_rate": 3.1132251987711637e-07, "loss": 0.0093, "step": 77300 }, { "grad_norm": 0.1059834361076355, "learning_rate": 3.0902392144007695e-07, "loss": 0.0063, "step": 77310 }, { "grad_norm": 0.158324733376503, "learning_rate": 3.0673381383179436e-07, "loss": 0.0053, "step": 77320 }, { "grad_norm": 0.42781272530555725, "learning_rate": 3.0445219744358875e-07, "loss": 0.006, "step": 77330 }, { "grad_norm": 0.37434646487236023, "learning_rate": 3.021790726653262e-07, "loss": 0.0128, "step": 77340 }, { "grad_norm": 0.30183085799217224, "learning_rate": 2.9991443988542366e-07, "loss": 0.0073, "step": 77350 }, { "grad_norm": 0.20113882422447205, "learning_rate": 2.976582994908439e-07, "loss": 0.0045, "step": 77360 }, { "grad_norm": 0.11337338387966156, "learning_rate": 2.954106518671007e-07, "loss": 0.0042, "step": 77370 }, { "grad_norm": 0.14193585515022278, "learning_rate": 2.9317149739825356e-07, "loss": 0.0041, "step": 77380 }, { "grad_norm": 0.35670965909957886, "learning_rate": 2.909408364669075e-07, "loss": 0.0083, "step": 77390 }, { "grad_norm": 0.21466435492038727, "learning_rate": 2.887186694542299e-07, "loss": 0.0103, "step": 77400 }, { "grad_norm": 0.0934065580368042, "learning_rate": 2.865049967399225e-07, "loss": 0.0067, "step": 77410 }, { "grad_norm": 0.081200510263443, "learning_rate": 2.842998187022439e-07, "loss": 0.0043, "step": 77420 }, { "grad_norm": 0.2298596352338791, "learning_rate": 2.8210313571800374e-07, "loss": 0.0047, "step": 77430 }, { "grad_norm": 0.13690316677093506, "learning_rate": 2.7991494816255184e-07, "loss": 0.0056, "step": 77440 }, { "grad_norm": 0.1512662023305893, "learning_rate": 2.777352564097779e-07, "loss": 0.0058, "step": 77450 }, { "grad_norm": 0.20407696068286896, "learning_rate": 2.755640608321508e-07, "loss": 0.0048, "step": 77460 }, { "grad_norm": 0.33011946082115173, "learning_rate": 2.73401361800657e-07, "loss": 0.0065, "step": 77470 }, { "grad_norm": 0.14182934165000916, "learning_rate": 2.7124715968484537e-07, "loss": 0.0062, "step": 77480 }, { "grad_norm": 0.10054469108581543, "learning_rate": 2.691014548528104e-07, "loss": 0.0056, "step": 77490 }, { "grad_norm": 0.07226429879665375, "learning_rate": 2.669642476711864e-07, "loss": 0.0082, "step": 77500 }, { "grad_norm": 0.11112065613269806, "learning_rate": 2.648355385051815e-07, "loss": 0.0068, "step": 77510 }, { "grad_norm": 0.14745578169822693, "learning_rate": 2.627153277185157e-07, "loss": 0.0076, "step": 77520 }, { "grad_norm": 0.05711745098233223, "learning_rate": 2.606036156734881e-07, "loss": 0.0042, "step": 77530 }, { "grad_norm": 0.08585722744464874, "learning_rate": 2.5850040273092127e-07, "loss": 0.004, "step": 77540 }, { "grad_norm": 0.41613703966140747, "learning_rate": 2.5640568925020536e-07, "loss": 0.01, "step": 77550 }, { "grad_norm": 0.18046967685222626, "learning_rate": 2.5431947558927083e-07, "loss": 0.0078, "step": 77560 }, { "grad_norm": 0.09400296211242676, "learning_rate": 2.522417621045825e-07, "loss": 0.0038, "step": 77570 }, { "grad_norm": 0.09052921086549759, "learning_rate": 2.501725491511786e-07, "loss": 0.007, "step": 77580 }, { "grad_norm": 0.05341582000255585, "learning_rate": 2.481118370826263e-07, "loss": 0.0043, "step": 77590 }, { "grad_norm": 0.1978582739830017, "learning_rate": 2.4605962625104393e-07, "loss": 0.0049, "step": 77600 }, { "grad_norm": 0.11936602741479874, "learning_rate": 2.440159170070955e-07, "loss": 0.0107, "step": 77610 }, { "grad_norm": 0.11736780405044556, "learning_rate": 2.41980709699996e-07, "loss": 0.0053, "step": 77620 }, { "grad_norm": 0.09579905867576599, "learning_rate": 2.3995400467751174e-07, "loss": 0.0056, "step": 77630 }, { "grad_norm": 0.14841942489147186, "learning_rate": 2.3793580228594902e-07, "loss": 0.0118, "step": 77640 }, { "grad_norm": 0.1673140972852707, "learning_rate": 2.3592610287015982e-07, "loss": 0.0047, "step": 77650 }, { "grad_norm": 0.09707457572221756, "learning_rate": 2.3392490677354718e-07, "loss": 0.0062, "step": 77660 }, { "grad_norm": 0.15880899131298065, "learning_rate": 2.3193221433806533e-07, "loss": 0.004, "step": 77670 }, { "grad_norm": 0.20482580363750458, "learning_rate": 2.299480259042086e-07, "loss": 0.0059, "step": 77680 }, { "grad_norm": 0.15211518108844757, "learning_rate": 2.2797234181102244e-07, "loss": 0.0052, "step": 77690 }, { "grad_norm": 0.15116462111473083, "learning_rate": 2.2600516239609238e-07, "loss": 0.0065, "step": 77700 }, { "grad_norm": 0.10888710618019104, "learning_rate": 2.240464879955606e-07, "loss": 0.0071, "step": 77710 }, { "grad_norm": 0.14470764994621277, "learning_rate": 2.2209631894410387e-07, "loss": 0.0047, "step": 77720 }, { "grad_norm": 0.12267784029245377, "learning_rate": 2.2015465557496117e-07, "loss": 0.0057, "step": 77730 }, { "grad_norm": 0.09234770387411118, "learning_rate": 2.1822149821990602e-07, "loss": 0.006, "step": 77740 }, { "grad_norm": 0.2854050099849701, "learning_rate": 2.1629684720926303e-07, "loss": 0.0099, "step": 77750 }, { "grad_norm": 0.219798743724823, "learning_rate": 2.1438070287189693e-07, "loss": 0.0061, "step": 77760 }, { "grad_norm": 0.046581435948610306, "learning_rate": 2.1247306553523472e-07, "loss": 0.0083, "step": 77770 }, { "grad_norm": 0.08653029054403305, "learning_rate": 2.105739355252323e-07, "loss": 0.0053, "step": 77780 }, { "grad_norm": 0.21058188378810883, "learning_rate": 2.0868331316639678e-07, "loss": 0.0048, "step": 77790 }, { "grad_norm": 0.08442594856023788, "learning_rate": 2.0680119878179193e-07, "loss": 0.0039, "step": 77800 }, { "grad_norm": 0.2147202491760254, "learning_rate": 2.0492759269301054e-07, "loss": 0.0075, "step": 77810 }, { "grad_norm": 0.20598441362380981, "learning_rate": 2.0306249522021315e-07, "loss": 0.0049, "step": 77820 }, { "grad_norm": 0.1239398792386055, "learning_rate": 2.0120590668207816e-07, "loss": 0.005, "step": 77830 }, { "grad_norm": 0.09642212837934494, "learning_rate": 1.993578273958574e-07, "loss": 0.0052, "step": 77840 }, { "grad_norm": 0.056993693113327026, "learning_rate": 1.975182576773371e-07, "loss": 0.0055, "step": 77850 }, { "grad_norm": 0.40568721294403076, "learning_rate": 1.9568719784083812e-07, "loss": 0.0124, "step": 77860 }, { "grad_norm": 0.08804386109113693, "learning_rate": 1.9386464819924898e-07, "loss": 0.0092, "step": 77870 }, { "grad_norm": 0.05326545983552933, "learning_rate": 1.9205060906399285e-07, "loss": 0.0058, "step": 77880 }, { "grad_norm": 0.08858409523963928, "learning_rate": 1.9024508074503845e-07, "loss": 0.006, "step": 77890 }, { "grad_norm": 0.3612346947193146, "learning_rate": 1.8844806355089452e-07, "loss": 0.0127, "step": 77900 }, { "grad_norm": 0.13677392899990082, "learning_rate": 1.8665955778863208e-07, "loss": 0.0145, "step": 77910 }, { "grad_norm": 0.0933137908577919, "learning_rate": 1.848795637638512e-07, "loss": 0.0046, "step": 77920 }, { "grad_norm": 0.14711163938045502, "learning_rate": 1.831080817807085e-07, "loss": 0.0089, "step": 77930 }, { "grad_norm": 0.05405840650200844, "learning_rate": 1.813451121418952e-07, "loss": 0.0137, "step": 77940 }, { "grad_norm": 0.13406531512737274, "learning_rate": 1.795906551486648e-07, "loss": 0.0038, "step": 77950 }, { "grad_norm": 0.1964215487241745, "learning_rate": 1.7784471110079414e-07, "loss": 0.0093, "step": 77960 }, { "grad_norm": 0.27483049035072327, "learning_rate": 1.7610728029662793e-07, "loss": 0.0065, "step": 77970 }, { "grad_norm": 0.05146399885416031, "learning_rate": 1.7437836303303979e-07, "loss": 0.0054, "step": 77980 }, { "grad_norm": 0.1114112064242363, "learning_rate": 1.726579596054545e-07, "loss": 0.0034, "step": 77990 }, { "grad_norm": 0.24178341031074524, "learning_rate": 1.7094607030784803e-07, "loss": 0.0087, "step": 78000 }, { "grad_norm": 0.09975431114435196, "learning_rate": 1.6924269543272532e-07, "loss": 0.0061, "step": 78010 }, { "grad_norm": 0.047928258776664734, "learning_rate": 1.6754783527115348e-07, "loss": 0.0052, "step": 78020 }, { "grad_norm": 0.05607568472623825, "learning_rate": 1.6586149011273422e-07, "loss": 0.0054, "step": 78030 }, { "grad_norm": 0.15057207643985748, "learning_rate": 1.6418366024562038e-07, "loss": 0.016, "step": 78040 }, { "grad_norm": 0.11305883526802063, "learning_rate": 1.6251434595651037e-07, "loss": 0.0067, "step": 78050 }, { "grad_norm": 0.07160530239343643, "learning_rate": 1.608535475306372e-07, "loss": 0.0066, "step": 78060 }, { "grad_norm": 0.07900314778089523, "learning_rate": 1.5920126525179048e-07, "loss": 0.0062, "step": 78070 }, { "grad_norm": 0.143800288438797, "learning_rate": 1.5755749940229435e-07, "loss": 0.0042, "step": 78080 }, { "grad_norm": 0.12418360263109207, "learning_rate": 1.559222502630353e-07, "loss": 0.0054, "step": 78090 }, { "grad_norm": 0.17441490292549133, "learning_rate": 1.5429551811341757e-07, "loss": 0.0071, "step": 78100 }, { "grad_norm": 0.14886359870433807, "learning_rate": 1.5267730323141882e-07, "loss": 0.0116, "step": 78110 }, { "grad_norm": 0.08694032579660416, "learning_rate": 1.5106760589353454e-07, "loss": 0.0048, "step": 78120 }, { "grad_norm": 0.13026249408721924, "learning_rate": 1.4946642637483355e-07, "loss": 0.0063, "step": 78130 }, { "grad_norm": 0.3102235198020935, "learning_rate": 1.4787376494889703e-07, "loss": 0.0056, "step": 78140 }, { "grad_norm": 0.11130964010953903, "learning_rate": 1.4628962188787955e-07, "loss": 0.0045, "step": 78150 }, { "grad_norm": 0.1308162361383438, "learning_rate": 1.4471399746247006e-07, "loss": 0.0071, "step": 78160 }, { "grad_norm": 0.10788922011852264, "learning_rate": 1.4314689194188103e-07, "loss": 0.0047, "step": 78170 }, { "grad_norm": 0.36061975359916687, "learning_rate": 1.4158830559390933e-07, "loss": 0.0099, "step": 78180 }, { "grad_norm": 0.06852435320615768, "learning_rate": 1.4003823868486422e-07, "loss": 0.0178, "step": 78190 }, { "grad_norm": 0.18452835083007812, "learning_rate": 1.3849669147960598e-07, "loss": 0.0072, "step": 78200 }, { "grad_norm": 0.12045714259147644, "learning_rate": 1.3696366424155726e-07, "loss": 0.006, "step": 78210 }, { "grad_norm": 0.35911765694618225, "learning_rate": 1.35439157232653e-07, "loss": 0.0158, "step": 78220 }, { "grad_norm": 0.09361103922128677, "learning_rate": 1.3392317071340144e-07, "loss": 0.0066, "step": 78230 }, { "grad_norm": 0.2871406078338623, "learning_rate": 1.3241570494283984e-07, "loss": 0.0068, "step": 78240 }, { "grad_norm": 0.1545448750257492, "learning_rate": 1.3091676017855103e-07, "loss": 0.0085, "step": 78250 }, { "grad_norm": 0.33499810099601746, "learning_rate": 1.2942633667666904e-07, "loss": 0.0072, "step": 78260 }, { "grad_norm": 0.0948454886674881, "learning_rate": 1.2794443469185679e-07, "loss": 0.0129, "step": 78270 }, { "grad_norm": 0.22502049803733826, "learning_rate": 1.2647105447734508e-07, "loss": 0.006, "step": 78280 }, { "grad_norm": 0.04948614537715912, "learning_rate": 1.2500619628488254e-07, "loss": 0.0039, "step": 78290 }, { "grad_norm": 0.11830177158117294, "learning_rate": 1.2354986036477446e-07, "loss": 0.0114, "step": 78300 }, { "grad_norm": 0.13996632397174835, "learning_rate": 1.221020469658718e-07, "loss": 0.0078, "step": 78310 }, { "grad_norm": 0.08285195380449295, "learning_rate": 1.2066275633556556e-07, "loss": 0.0055, "step": 78320 }, { "grad_norm": 0.15089143812656403, "learning_rate": 1.192319887197979e-07, "loss": 0.0051, "step": 78330 }, { "grad_norm": 0.10504128038883209, "learning_rate": 1.1780974436303438e-07, "loss": 0.0048, "step": 78340 }, { "grad_norm": 0.09903137385845184, "learning_rate": 1.1639602350830836e-07, "loss": 0.0061, "step": 78350 }, { "grad_norm": 0.08040538430213928, "learning_rate": 1.1499082639718217e-07, "loss": 0.0083, "step": 78360 }, { "grad_norm": 0.16268795728683472, "learning_rate": 1.1359415326976374e-07, "loss": 0.0048, "step": 78370 }, { "grad_norm": 0.25074172019958496, "learning_rate": 1.1220600436470663e-07, "loss": 0.0052, "step": 78380 }, { "grad_norm": 0.06704428046941757, "learning_rate": 1.1082637991920996e-07, "loss": 0.0054, "step": 78390 }, { "grad_norm": 0.17621292173862457, "learning_rate": 1.0945528016901851e-07, "loss": 0.0078, "step": 78400 }, { "grad_norm": 0.13468657433986664, "learning_rate": 1.0809270534840599e-07, "loss": 0.0075, "step": 78410 }, { "grad_norm": 0.09809894114732742, "learning_rate": 1.067386556902028e-07, "loss": 0.0041, "step": 78420 }, { "grad_norm": 0.1999521255493164, "learning_rate": 1.05393131425785e-07, "loss": 0.0034, "step": 78430 }, { "grad_norm": 0.14641287922859192, "learning_rate": 1.0405613278505199e-07, "loss": 0.005, "step": 78440 }, { "grad_norm": 0.17688608169555664, "learning_rate": 1.027276599964766e-07, "loss": 0.0053, "step": 78450 }, { "grad_norm": 0.0769200548529625, "learning_rate": 1.0140771328704391e-07, "loss": 0.0039, "step": 78460 }, { "grad_norm": 0.1100059226155281, "learning_rate": 1.0009629288231237e-07, "loss": 0.0049, "step": 78470 }, { "grad_norm": 0.2026599943637848, "learning_rate": 9.879339900635276e-08, "loss": 0.0082, "step": 78480 }, { "grad_norm": 0.3159602880477905, "learning_rate": 9.74990318817981e-08, "loss": 0.0094, "step": 78490 }, { "grad_norm": 0.20149368047714233, "learning_rate": 9.621319172982701e-08, "loss": 0.0058, "step": 78500 }, { "grad_norm": 0.09402872622013092, "learning_rate": 9.493587877015264e-08, "loss": 0.0066, "step": 78510 }, { "grad_norm": 0.12620295584201813, "learning_rate": 9.366709322102263e-08, "loss": 0.0054, "step": 78520 }, { "grad_norm": 0.2663940489292145, "learning_rate": 9.240683529924688e-08, "loss": 0.0077, "step": 78530 }, { "grad_norm": 0.18969985842704773, "learning_rate": 9.115510522016979e-08, "loss": 0.0049, "step": 78540 }, { "grad_norm": 0.33199524879455566, "learning_rate": 8.991190319767583e-08, "loss": 0.0055, "step": 78550 }, { "grad_norm": 0.07673544436693192, "learning_rate": 8.867722944419509e-08, "loss": 0.0041, "step": 78560 }, { "grad_norm": 0.11137028783559799, "learning_rate": 8.745108417069214e-08, "loss": 0.0058, "step": 78570 }, { "grad_norm": 0.07067939639091492, "learning_rate": 8.623346758669381e-08, "loss": 0.0049, "step": 78580 }, { "grad_norm": 0.16773952543735504, "learning_rate": 8.502437990025037e-08, "loss": 0.0058, "step": 78590 }, { "grad_norm": 0.18235263228416443, "learning_rate": 8.382382131795764e-08, "loss": 0.0048, "step": 78600 }, { "grad_norm": 0.10087459534406662, "learning_rate": 8.263179204496818e-08, "loss": 0.0092, "step": 78610 }, { "grad_norm": 0.1378817856311798, "learning_rate": 8.144829228496354e-08, "loss": 0.0047, "step": 78620 }, { "grad_norm": 0.17977049946784973, "learning_rate": 8.027332224016526e-08, "loss": 0.0052, "step": 78630 }, { "grad_norm": 0.16066661477088928, "learning_rate": 7.910688211135164e-08, "loss": 0.0069, "step": 78640 }, { "grad_norm": 0.26691314578056335, "learning_rate": 7.794897209783546e-08, "loss": 0.0074, "step": 78650 }, { "grad_norm": 0.16272127628326416, "learning_rate": 7.679959239746403e-08, "loss": 0.0074, "step": 78660 }, { "grad_norm": 0.2165769785642624, "learning_rate": 7.565874320664135e-08, "loss": 0.0085, "step": 78670 }, { "grad_norm": 0.3379037082195282, "learning_rate": 7.45264247203059e-08, "loss": 0.0071, "step": 78680 }, { "grad_norm": 0.0880211666226387, "learning_rate": 7.340263713194184e-08, "loss": 0.0038, "step": 78690 }, { "grad_norm": 0.14436118304729462, "learning_rate": 7.228738063356777e-08, "loss": 0.0041, "step": 78700 }, { "grad_norm": 0.4411477744579315, "learning_rate": 7.118065541575903e-08, "loss": 0.0096, "step": 78710 }, { "grad_norm": 0.14954280853271484, "learning_rate": 7.008246166761435e-08, "loss": 0.0044, "step": 78720 }, { "grad_norm": 0.12521186470985413, "learning_rate": 6.899279957679472e-08, "loss": 0.005, "step": 78730 }, { "grad_norm": 0.12412416189908981, "learning_rate": 6.791166932949011e-08, "loss": 0.0049, "step": 78740 }, { "grad_norm": 0.12265995889902115, "learning_rate": 6.683907111043608e-08, "loss": 0.0071, "step": 78750 }, { "grad_norm": 0.08643154799938202, "learning_rate": 6.577500510290824e-08, "loss": 0.0067, "step": 78760 }, { "grad_norm": 0.20711937546730042, "learning_rate": 6.47194714887278e-08, "loss": 0.0185, "step": 78770 }, { "grad_norm": 0.14140500128269196, "learning_rate": 6.367247044825608e-08, "loss": 0.0077, "step": 78780 }, { "grad_norm": 0.19051431119441986, "learning_rate": 6.263400216039994e-08, "loss": 0.006, "step": 78790 }, { "grad_norm": 0.20175215601921082, "learning_rate": 6.160406680260078e-08, "loss": 0.0194, "step": 78800 }, { "grad_norm": 0.11572075635194778, "learning_rate": 6.058266455084561e-08, "loss": 0.0048, "step": 78810 }, { "grad_norm": 0.0834839940071106, "learning_rate": 5.956979557967257e-08, "loss": 0.006, "step": 78820 }, { "grad_norm": 0.1481017917394638, "learning_rate": 5.856546006214325e-08, "loss": 0.008, "step": 78830 }, { "grad_norm": 0.1057661920785904, "learning_rate": 5.7569658169881466e-08, "loss": 0.0048, "step": 78840 }, { "grad_norm": 0.12535162270069122, "learning_rate": 5.658239007303445e-08, "loss": 0.007, "step": 78850 }, { "grad_norm": 0.36331936717033386, "learning_rate": 5.560365594030059e-08, "loss": 0.0097, "step": 78860 }, { "grad_norm": 0.3071640729904175, "learning_rate": 5.463345593891833e-08, "loss": 0.0052, "step": 78870 }, { "grad_norm": 0.08785640448331833, "learning_rate": 5.367179023467173e-08, "loss": 0.0052, "step": 78880 }, { "grad_norm": 0.1368289738893509, "learning_rate": 5.271865899187378e-08, "loss": 0.0053, "step": 78890 }, { "grad_norm": 0.2698056697845459, "learning_rate": 5.177406237340532e-08, "loss": 0.0195, "step": 78900 }, { "grad_norm": 0.11709540337324142, "learning_rate": 5.083800054065946e-08, "loss": 0.0034, "step": 78910 }, { "grad_norm": 0.10535359382629395, "learning_rate": 4.991047365358603e-08, "loss": 0.004, "step": 78920 }, { "grad_norm": 0.36592504382133484, "learning_rate": 4.899148187067493e-08, "loss": 0.017, "step": 78930 }, { "grad_norm": 0.1391177773475647, "learning_rate": 4.808102534895609e-08, "loss": 0.0057, "step": 78940 }, { "grad_norm": 0.16756002604961395, "learning_rate": 4.717910424400507e-08, "loss": 0.0051, "step": 78950 }, { "grad_norm": 0.2656387984752655, "learning_rate": 4.628571870993193e-08, "loss": 0.0068, "step": 78960 }, { "grad_norm": 0.2838825583457947, "learning_rate": 4.5400868899392324e-08, "loss": 0.0072, "step": 78970 }, { "grad_norm": 0.3149338960647583, "learning_rate": 4.4524554963576436e-08, "loss": 0.01, "step": 78980 }, { "grad_norm": 0.10917030274868011, "learning_rate": 4.36567770522367e-08, "loss": 0.0051, "step": 78990 }, { "grad_norm": 0.14181609451770782, "learning_rate": 4.279753531364339e-08, "loss": 0.0045, "step": 79000 }, { "grad_norm": 0.2478824257850647, "learning_rate": 4.1946829894617955e-08, "loss": 0.0077, "step": 79010 }, { "grad_norm": 0.11646518111228943, "learning_rate": 4.11046609405219e-08, "loss": 0.011, "step": 79020 }, { "grad_norm": 0.16079415380954742, "learning_rate": 4.027102859526233e-08, "loss": 0.0038, "step": 79030 }, { "grad_norm": 0.09884407371282578, "learning_rate": 3.9445933001280856e-08, "loss": 0.0067, "step": 79040 }, { "grad_norm": 0.09303553402423859, "learning_rate": 3.8629374299564704e-08, "loss": 0.0079, "step": 79050 }, { "grad_norm": 0.08782871812582016, "learning_rate": 3.78213526296467e-08, "loss": 0.0066, "step": 79060 }, { "grad_norm": 0.04887544363737106, "learning_rate": 3.702186812958308e-08, "loss": 0.0059, "step": 79070 }, { "grad_norm": 0.2163083702325821, "learning_rate": 3.623092093599789e-08, "loss": 0.0062, "step": 79080 }, { "grad_norm": 0.2208586037158966, "learning_rate": 3.544851118403303e-08, "loss": 0.0056, "step": 79090 }, { "grad_norm": 0.15780147910118103, "learning_rate": 3.46746390073871e-08, "loss": 0.007, "step": 79100 }, { "grad_norm": 0.08428903669118881, "learning_rate": 3.390930453829322e-08, "loss": 0.015, "step": 79110 }, { "grad_norm": 0.17989499866962433, "learning_rate": 3.3152507907519005e-08, "loss": 0.0037, "step": 79120 }, { "grad_norm": 0.13963644206523895, "learning_rate": 3.240424924438323e-08, "loss": 0.0052, "step": 79130 }, { "grad_norm": 0.1683541089296341, "learning_rate": 3.1664528676750295e-08, "loss": 0.0076, "step": 79140 }, { "grad_norm": 0.17646922171115875, "learning_rate": 3.093334633100797e-08, "loss": 0.0174, "step": 79150 }, { "grad_norm": 0.12369577586650848, "learning_rate": 3.021070233210077e-08, "loss": 0.0059, "step": 79160 }, { "grad_norm": 0.12098788470029831, "learning_rate": 2.9496596803507692e-08, "loss": 0.0062, "step": 79170 }, { "grad_norm": 0.21733060479164124, "learning_rate": 2.879102986725335e-08, "loss": 0.0063, "step": 79180 }, { "grad_norm": 0.07345584034919739, "learning_rate": 2.809400164389131e-08, "loss": 0.0056, "step": 79190 }, { "grad_norm": 0.1280825287103653, "learning_rate": 2.740551225253185e-08, "loss": 0.0059, "step": 79200 }, { "grad_norm": 0.0991913378238678, "learning_rate": 2.6725561810819754e-08, "loss": 0.0083, "step": 79210 }, { "grad_norm": 0.16955798864364624, "learning_rate": 2.605415043493431e-08, "loss": 0.0051, "step": 79220 }, { "grad_norm": 0.07625117152929306, "learning_rate": 2.5391278239605965e-08, "loss": 0.0047, "step": 79230 }, { "grad_norm": 0.08951558917760849, "learning_rate": 2.4736945338094118e-08, "loss": 0.0054, "step": 79240 }, { "grad_norm": 0.07128096371889114, "learning_rate": 2.4091151842214887e-08, "loss": 0.0046, "step": 79250 }, { "grad_norm": 0.10597986727952957, "learning_rate": 2.3453897862318884e-08, "loss": 0.0084, "step": 79260 }, { "grad_norm": 0.09298627078533173, "learning_rate": 2.2825183507285686e-08, "loss": 0.0054, "step": 79270 }, { "grad_norm": 0.07504747807979584, "learning_rate": 2.220500888455157e-08, "loss": 0.0097, "step": 79280 }, { "grad_norm": 0.30294302105903625, "learning_rate": 2.1593374100081777e-08, "loss": 0.0118, "step": 79290 }, { "grad_norm": 0.10510019958019257, "learning_rate": 2.09902792583927e-08, "loss": 0.0121, "step": 79300 }, { "grad_norm": 0.08836384862661362, "learning_rate": 2.0395724462540788e-08, "loss": 0.005, "step": 79310 }, { "grad_norm": 0.07828467339277267, "learning_rate": 1.9809709814111453e-08, "loss": 0.0037, "step": 79320 }, { "grad_norm": 0.05899541452527046, "learning_rate": 1.923223541324126e-08, "loss": 0.0127, "step": 79330 }, { "grad_norm": 0.15897482633590698, "learning_rate": 1.8663301358606833e-08, "loss": 0.0061, "step": 79340 }, { "grad_norm": 0.05381548032164574, "learning_rate": 1.8102907747419295e-08, "loss": 0.0137, "step": 79350 }, { "grad_norm": 0.07418574392795563, "learning_rate": 1.7551054675435385e-08, "loss": 0.006, "step": 79360 }, { "grad_norm": 0.05998067930340767, "learning_rate": 1.7007742236957447e-08, "loss": 0.0068, "step": 79370 }, { "grad_norm": 0.21437089145183563, "learning_rate": 1.647297052481678e-08, "loss": 0.0076, "step": 79380 }, { "grad_norm": 0.17232294380664825, "learning_rate": 1.5946739630390285e-08, "loss": 0.0086, "step": 79390 }, { "grad_norm": 0.36498451232910156, "learning_rate": 1.5429049643606032e-08, "loss": 0.0105, "step": 79400 }, { "grad_norm": 0.08733094483613968, "learning_rate": 1.4919900652909935e-08, "loss": 0.0093, "step": 79410 }, { "grad_norm": 0.07867586612701416, "learning_rate": 1.4419292745310175e-08, "loss": 0.0088, "step": 79420 }, { "grad_norm": 0.07475145161151886, "learning_rate": 1.3927226006343885e-08, "loss": 0.006, "step": 79430 }, { "grad_norm": 0.21146903932094574, "learning_rate": 1.3443700520093805e-08, "loss": 0.0069, "step": 79440 }, { "grad_norm": 0.08336326479911804, "learning_rate": 1.296871636917718e-08, "loss": 0.0056, "step": 79450 }, { "grad_norm": 0.2263249307870865, "learning_rate": 1.2502273634762419e-08, "loss": 0.0067, "step": 79460 }, { "grad_norm": 0.3599039614200592, "learning_rate": 1.2044372396546876e-08, "loss": 0.0079, "step": 79470 }, { "grad_norm": 0.16871845722198486, "learning_rate": 1.1595012732773524e-08, "loss": 0.0095, "step": 79480 }, { "grad_norm": 0.05319077894091606, "learning_rate": 1.1154194720225386e-08, "loss": 0.0066, "step": 79490 }, { "grad_norm": 0.22543033957481384, "learning_rate": 1.0721918434231093e-08, "loss": 0.0053, "step": 79500 }, { "grad_norm": 0.08335569500923157, "learning_rate": 1.0298183948648232e-08, "loss": 0.0048, "step": 79510 }, { "grad_norm": 0.2961924374103546, "learning_rate": 9.88299133588555e-09, "loss": 0.0054, "step": 79520 }, { "grad_norm": 0.29711517691612244, "learning_rate": 9.476340666891847e-09, "loss": 0.0114, "step": 79530 }, { "grad_norm": 0.07334393262863159, "learning_rate": 9.078232011139332e-09, "loss": 0.0042, "step": 79540 }, { "grad_norm": 0.14108097553253174, "learning_rate": 8.68866543666802e-09, "loss": 0.0063, "step": 79550 }, { "grad_norm": 0.07522191107273102, "learning_rate": 8.307641010035782e-09, "loss": 0.0072, "step": 79560 }, { "grad_norm": 0.07563011348247528, "learning_rate": 7.93515879635165e-09, "loss": 0.0064, "step": 79570 }, { "grad_norm": 0.08039194345474243, "learning_rate": 7.571218859264706e-09, "loss": 0.0049, "step": 79580 }, { "grad_norm": 0.11946171522140503, "learning_rate": 7.215821260958544e-09, "loss": 0.0094, "step": 79590 }, { "grad_norm": 0.16027551889419556, "learning_rate": 6.868966062162363e-09, "loss": 0.0059, "step": 79600 }, { "grad_norm": 0.1824624389410019, "learning_rate": 6.530653322145419e-09, "loss": 0.0041, "step": 79610 }, { "grad_norm": 0.11646176874637604, "learning_rate": 6.200883098717025e-09, "loss": 0.0031, "step": 79620 }, { "grad_norm": 0.2857043147087097, "learning_rate": 5.879655448226551e-09, "loss": 0.0127, "step": 79630 }, { "grad_norm": 0.15554383397102356, "learning_rate": 5.566970425557872e-09, "loss": 0.0034, "step": 79640 }, { "grad_norm": 0.28862258791923523, "learning_rate": 5.26282808414047e-09, "loss": 0.0042, "step": 79650 }, { "grad_norm": 0.12071027606725693, "learning_rate": 4.967228475949437e-09, "loss": 0.0069, "step": 79660 }, { "grad_norm": 0.10010311007499695, "learning_rate": 4.680171651494369e-09, "loss": 0.0056, "step": 79670 }, { "grad_norm": 0.15139439702033997, "learning_rate": 4.40165765981937e-09, "loss": 0.006, "step": 79680 }, { "grad_norm": 0.14338183403015137, "learning_rate": 4.131686548519698e-09, "loss": 0.0034, "step": 79690 }, { "grad_norm": 0.10720481723546982, "learning_rate": 3.8702583637251214e-09, "loss": 0.0048, "step": 79700 }, { "grad_norm": 0.10791261494159698, "learning_rate": 3.617373150105463e-09, "loss": 0.005, "step": 79710 }, { "grad_norm": 0.15009567141532898, "learning_rate": 3.3730309508706036e-09, "loss": 0.0055, "step": 79720 }, { "grad_norm": 0.18918348848819733, "learning_rate": 3.137231807781582e-09, "loss": 0.0045, "step": 79730 }, { "grad_norm": 0.1389179676771164, "learning_rate": 2.9099757611172894e-09, "loss": 0.0049, "step": 79740 }, { "grad_norm": 0.2751205861568451, "learning_rate": 2.6912628497133275e-09, "loss": 0.0065, "step": 79750 }, { "grad_norm": 0.2917356789112091, "learning_rate": 2.481093110945354e-09, "loss": 0.0096, "step": 79760 }, { "grad_norm": 0.05827521160244942, "learning_rate": 2.2794665807235327e-09, "loss": 0.0071, "step": 79770 }, { "grad_norm": 0.1315326988697052, "learning_rate": 2.0863832934980843e-09, "loss": 0.0056, "step": 79780 }, { "grad_norm": 0.1713462471961975, "learning_rate": 1.901843282264837e-09, "loss": 0.0066, "step": 79790 }, { "grad_norm": 0.15261641144752502, "learning_rate": 1.7258465785541245e-09, "loss": 0.0104, "step": 79800 }, { "grad_norm": 0.04770815372467041, "learning_rate": 1.5583932124418887e-09, "loss": 0.0075, "step": 79810 }, { "grad_norm": 0.08438723534345627, "learning_rate": 1.3994832125441282e-09, "loss": 0.0055, "step": 79820 }, { "grad_norm": 0.17303743958473206, "learning_rate": 1.2491166060057958e-09, "loss": 0.0063, "step": 79830 }, { "grad_norm": 0.10827421396970749, "learning_rate": 1.1072934185230032e-09, "loss": 0.0072, "step": 79840 }, { "grad_norm": 0.14224520325660706, "learning_rate": 9.740136743319195e-10, "loss": 0.0064, "step": 79850 }, { "grad_norm": 0.08736162632703781, "learning_rate": 8.492773962087696e-10, "loss": 0.0039, "step": 79860 }, { "grad_norm": 0.20674234628677368, "learning_rate": 7.330846054587337e-10, "loss": 0.0121, "step": 79870 }, { "grad_norm": 0.1850060522556305, "learning_rate": 6.254353219492526e-10, "loss": 0.0055, "step": 79880 }, { "grad_norm": 0.15224991738796234, "learning_rate": 5.263295640600685e-10, "loss": 0.0044, "step": 79890 }, { "grad_norm": 0.1087690219283104, "learning_rate": 4.357673487387359e-10, "loss": 0.0043, "step": 79900 }, { "grad_norm": 0.22641721367835999, "learning_rate": 3.537486914506616e-10, "loss": 0.0061, "step": 79910 }, { "grad_norm": 0.08275235444307327, "learning_rate": 2.8027360621241117e-10, "loss": 0.0057, "step": 79920 }, { "grad_norm": 0.22888264060020447, "learning_rate": 2.153421055806071e-10, "loss": 0.0057, "step": 79930 }, { "grad_norm": 0.07097417116165161, "learning_rate": 1.5895420064637733e-10, "loss": 0.0081, "step": 79940 }, { "grad_norm": 0.1437423825263977, "learning_rate": 1.1110990105200891e-10, "loss": 0.006, "step": 79950 }, { "grad_norm": 0.19020915031433105, "learning_rate": 7.180921496874326e-11, "loss": 0.0089, "step": 79960 }, { "grad_norm": 0.22150608897209167, "learning_rate": 4.10521491134297e-11, "loss": 0.0054, "step": 79970 }, { "grad_norm": 0.060337796807289124, "learning_rate": 1.883870874297422e-11, "loss": 0.0042, "step": 79980 }, { "grad_norm": 0.1912253201007843, "learning_rate": 5.1688976432373584e-12, "loss": 0.0075, "step": 79990 }, { "grad_norm": 0.1742105931043625, "learning_rate": 4.271816234080461e-14, "loss": 0.0045, "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 }