{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.6, "eval_steps": 500, "global_step": 60000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 1.2290912866592407, "learning_rate": 1.8e-07, "loss": 1.6012, "step": 10 }, { "grad_norm": 1.2460496425628662, "learning_rate": 3.8e-07, "loss": 1.6164, "step": 20 }, { "grad_norm": 1.0227303504943848, "learning_rate": 5.8e-07, "loss": 1.6227, "step": 30 }, { "grad_norm": 1.2758257389068604, "learning_rate": 7.8e-07, "loss": 1.6333, "step": 40 }, { "grad_norm": 1.1886036396026611, "learning_rate": 9.8e-07, "loss": 1.5865, "step": 50 }, { "grad_norm": 1.3831751346588135, "learning_rate": 1.18e-06, "loss": 1.5342, "step": 60 }, { "grad_norm": 1.3029847145080566, "learning_rate": 1.3800000000000001e-06, "loss": 1.426, "step": 70 }, { "grad_norm": 1.2573559284210205, "learning_rate": 1.5800000000000003e-06, "loss": 1.3724, "step": 80 }, { "grad_norm": 0.7010372877120972, "learning_rate": 1.7800000000000001e-06, "loss": 1.3593, "step": 90 }, { "grad_norm": 0.9436380863189697, "learning_rate": 1.98e-06, "loss": 1.3213, "step": 100 }, { "grad_norm": 1.1609714031219482, "learning_rate": 2.1800000000000003e-06, "loss": 1.3273, "step": 110 }, { "grad_norm": 1.1970487833023071, "learning_rate": 2.38e-06, "loss": 1.2789, "step": 120 }, { "grad_norm": 1.250502586364746, "learning_rate": 2.5800000000000003e-06, "loss": 1.2456, "step": 130 }, { "grad_norm": 2.416372537612915, "learning_rate": 2.78e-06, "loss": 1.2138, "step": 140 }, { "grad_norm": 1.3832491636276245, "learning_rate": 2.9800000000000003e-06, "loss": 1.1763, "step": 150 }, { "grad_norm": 0.9700374007225037, "learning_rate": 3.1800000000000005e-06, "loss": 1.1786, "step": 160 }, { "grad_norm": 1.4295426607131958, "learning_rate": 3.38e-06, "loss": 1.164, "step": 170 }, { "grad_norm": 1.4310730695724487, "learning_rate": 3.58e-06, "loss": 1.1589, "step": 180 }, { "grad_norm": 1.0520210266113281, "learning_rate": 3.7800000000000002e-06, "loss": 1.1439, "step": 190 }, { "grad_norm": 1.0056607723236084, "learning_rate": 3.98e-06, "loss": 1.1354, "step": 200 }, { "grad_norm": 1.0266286134719849, "learning_rate": 4.18e-06, "loss": 1.1391, "step": 210 }, { "grad_norm": 1.2878936529159546, "learning_rate": 4.38e-06, "loss": 1.1231, "step": 220 }, { "grad_norm": 1.1836620569229126, "learning_rate": 4.58e-06, "loss": 1.1106, "step": 230 }, { "grad_norm": 1.4053236246109009, "learning_rate": 4.780000000000001e-06, "loss": 1.1103, "step": 240 }, { "grad_norm": 0.9213263988494873, "learning_rate": 4.98e-06, "loss": 1.1031, "step": 250 }, { "grad_norm": 1.014990210533142, "learning_rate": 5.18e-06, "loss": 1.1006, "step": 260 }, { "grad_norm": 1.2848129272460938, "learning_rate": 5.38e-06, "loss": 1.0839, "step": 270 }, { "grad_norm": 1.542252779006958, "learning_rate": 5.580000000000001e-06, "loss": 1.0892, "step": 280 }, { "grad_norm": 1.2400487661361694, "learning_rate": 5.78e-06, "loss": 1.0813, "step": 290 }, { "grad_norm": 1.5386462211608887, "learning_rate": 5.98e-06, "loss": 1.0689, "step": 300 }, { "grad_norm": 1.2535550594329834, "learning_rate": 6.18e-06, "loss": 1.0761, "step": 310 }, { "grad_norm": 1.9339817762374878, "learning_rate": 6.38e-06, "loss": 1.0883, "step": 320 }, { "grad_norm": 1.8144413232803345, "learning_rate": 6.58e-06, "loss": 1.0685, "step": 330 }, { "grad_norm": 2.278562307357788, "learning_rate": 6.78e-06, "loss": 1.0669, "step": 340 }, { "grad_norm": 1.182571530342102, "learning_rate": 6.98e-06, "loss": 1.0632, "step": 350 }, { "grad_norm": 1.4468576908111572, "learning_rate": 7.180000000000001e-06, "loss": 1.0295, "step": 360 }, { "grad_norm": 1.1499576568603516, "learning_rate": 7.3800000000000005e-06, "loss": 1.0135, "step": 370 }, { "grad_norm": 1.3779709339141846, "learning_rate": 7.580000000000001e-06, "loss": 0.9846, "step": 380 }, { "grad_norm": 1.5467307567596436, "learning_rate": 7.78e-06, "loss": 0.9648, "step": 390 }, { "grad_norm": 1.8705132007598877, "learning_rate": 7.98e-06, "loss": 0.9101, "step": 400 }, { "grad_norm": 1.576177954673767, "learning_rate": 8.18e-06, "loss": 0.8784, "step": 410 }, { "grad_norm": 1.7473618984222412, "learning_rate": 8.380000000000001e-06, "loss": 0.8442, "step": 420 }, { "grad_norm": 1.8937230110168457, "learning_rate": 8.580000000000001e-06, "loss": 0.7999, "step": 430 }, { "grad_norm": 1.7929352521896362, "learning_rate": 8.78e-06, "loss": 0.7639, "step": 440 }, { "grad_norm": 1.9616237878799438, "learning_rate": 8.98e-06, "loss": 0.7272, "step": 450 }, { "grad_norm": 1.925363302230835, "learning_rate": 9.180000000000002e-06, "loss": 0.6912, "step": 460 }, { "grad_norm": 2.4074981212615967, "learning_rate": 9.38e-06, "loss": 0.67, "step": 470 }, { "grad_norm": 1.8589738607406616, "learning_rate": 9.58e-06, "loss": 0.6264, "step": 480 }, { "grad_norm": 2.520130157470703, "learning_rate": 9.78e-06, "loss": 0.5956, "step": 490 }, { "grad_norm": 1.8395726680755615, "learning_rate": 9.980000000000001e-06, "loss": 0.5588, "step": 500 }, { "grad_norm": 1.9660086631774902, "learning_rate": 1.018e-05, "loss": 0.5262, "step": 510 }, { "grad_norm": 2.0401809215545654, "learning_rate": 1.038e-05, "loss": 0.5027, "step": 520 }, { "grad_norm": 2.0323073863983154, "learning_rate": 1.058e-05, "loss": 0.4878, "step": 530 }, { "grad_norm": 2.0087082386016846, "learning_rate": 1.0780000000000002e-05, "loss": 0.459, "step": 540 }, { "grad_norm": 2.18056321144104, "learning_rate": 1.098e-05, "loss": 0.4473, "step": 550 }, { "grad_norm": 2.159303665161133, "learning_rate": 1.118e-05, "loss": 0.411, "step": 560 }, { "grad_norm": 2.5626964569091797, "learning_rate": 1.1380000000000001e-05, "loss": 0.3801, "step": 570 }, { "grad_norm": 2.761140823364258, "learning_rate": 1.1580000000000001e-05, "loss": 0.3546, "step": 580 }, { "grad_norm": 2.620248556137085, "learning_rate": 1.178e-05, "loss": 0.3177, "step": 590 }, { "grad_norm": 2.751621723175049, "learning_rate": 1.198e-05, "loss": 0.2962, "step": 600 }, { "grad_norm": 2.7359917163848877, "learning_rate": 1.2180000000000002e-05, "loss": 0.2811, "step": 610 }, { "grad_norm": 2.1962080001831055, "learning_rate": 1.238e-05, "loss": 0.2656, "step": 620 }, { "grad_norm": 2.904689311981201, "learning_rate": 1.258e-05, "loss": 0.2583, "step": 630 }, { "grad_norm": 2.5334489345550537, "learning_rate": 1.278e-05, "loss": 0.2333, "step": 640 }, { "grad_norm": 2.514230251312256, "learning_rate": 1.2980000000000001e-05, "loss": 0.2238, "step": 650 }, { "grad_norm": 2.091510534286499, "learning_rate": 1.3180000000000001e-05, "loss": 0.2268, "step": 660 }, { "grad_norm": 2.6753268241882324, "learning_rate": 1.338e-05, "loss": 0.2158, "step": 670 }, { "grad_norm": 1.9262925386428833, "learning_rate": 1.358e-05, "loss": 0.2283, "step": 680 }, { "grad_norm": 2.1982569694519043, "learning_rate": 1.3780000000000002e-05, "loss": 0.2197, "step": 690 }, { "grad_norm": 2.021962881088257, "learning_rate": 1.3980000000000002e-05, "loss": 0.2294, "step": 700 }, { "grad_norm": 2.1376867294311523, "learning_rate": 1.4180000000000001e-05, "loss": 0.1937, "step": 710 }, { "grad_norm": 1.9813554286956787, "learning_rate": 1.4380000000000001e-05, "loss": 0.1837, "step": 720 }, { "grad_norm": 2.408888816833496, "learning_rate": 1.4580000000000003e-05, "loss": 0.1982, "step": 730 }, { "grad_norm": 2.9145734310150146, "learning_rate": 1.4779999999999999e-05, "loss": 0.1909, "step": 740 }, { "grad_norm": 2.1620123386383057, "learning_rate": 1.4979999999999999e-05, "loss": 0.1809, "step": 750 }, { "grad_norm": 2.135772228240967, "learning_rate": 1.518e-05, "loss": 0.1942, "step": 760 }, { "grad_norm": 1.903408408164978, "learning_rate": 1.538e-05, "loss": 0.1948, "step": 770 }, { "grad_norm": 2.177030086517334, "learning_rate": 1.558e-05, "loss": 0.1787, "step": 780 }, { "grad_norm": 2.2047250270843506, "learning_rate": 1.578e-05, "loss": 0.1658, "step": 790 }, { "grad_norm": 1.9262462854385376, "learning_rate": 1.598e-05, "loss": 0.1808, "step": 800 }, { "grad_norm": 1.7759777307510376, "learning_rate": 1.618e-05, "loss": 0.1818, "step": 810 }, { "grad_norm": 1.547258734703064, "learning_rate": 1.6380000000000002e-05, "loss": 0.1731, "step": 820 }, { "grad_norm": 2.2078609466552734, "learning_rate": 1.658e-05, "loss": 0.161, "step": 830 }, { "grad_norm": 1.9968024492263794, "learning_rate": 1.6780000000000002e-05, "loss": 0.1584, "step": 840 }, { "grad_norm": 1.754744529724121, "learning_rate": 1.698e-05, "loss": 0.1509, "step": 850 }, { "grad_norm": 1.7173848152160645, "learning_rate": 1.718e-05, "loss": 0.1615, "step": 860 }, { "grad_norm": 2.1212944984436035, "learning_rate": 1.7380000000000003e-05, "loss": 0.1472, "step": 870 }, { "grad_norm": 1.991513967514038, "learning_rate": 1.758e-05, "loss": 0.1469, "step": 880 }, { "grad_norm": 2.043696403503418, "learning_rate": 1.7780000000000003e-05, "loss": 0.1599, "step": 890 }, { "grad_norm": 2.49912428855896, "learning_rate": 1.798e-05, "loss": 0.1383, "step": 900 }, { "grad_norm": 2.189675807952881, "learning_rate": 1.818e-05, "loss": 0.1462, "step": 910 }, { "grad_norm": 2.1053051948547363, "learning_rate": 1.838e-05, "loss": 0.1405, "step": 920 }, { "grad_norm": 1.9213342666625977, "learning_rate": 1.858e-05, "loss": 0.1436, "step": 930 }, { "grad_norm": 1.8916701078414917, "learning_rate": 1.878e-05, "loss": 0.1319, "step": 940 }, { "grad_norm": 1.7576944828033447, "learning_rate": 1.898e-05, "loss": 0.142, "step": 950 }, { "grad_norm": 1.6606614589691162, "learning_rate": 1.918e-05, "loss": 0.1313, "step": 960 }, { "grad_norm": 1.6617685556411743, "learning_rate": 1.938e-05, "loss": 0.1572, "step": 970 }, { "grad_norm": 1.8980330228805542, "learning_rate": 1.9580000000000002e-05, "loss": 0.1434, "step": 980 }, { "grad_norm": 2.186476469039917, "learning_rate": 1.978e-05, "loss": 0.1473, "step": 990 }, { "grad_norm": 1.9831750392913818, "learning_rate": 1.9980000000000002e-05, "loss": 0.1362, "step": 1000 }, { "grad_norm": 2.1972403526306152, "learning_rate": 2.0180000000000003e-05, "loss": 0.132, "step": 1010 }, { "grad_norm": 2.0710513591766357, "learning_rate": 2.038e-05, "loss": 0.1344, "step": 1020 }, { "grad_norm": 1.7182835340499878, "learning_rate": 2.0580000000000003e-05, "loss": 0.1327, "step": 1030 }, { "grad_norm": 1.4701626300811768, "learning_rate": 2.078e-05, "loss": 0.1126, "step": 1040 }, { "grad_norm": 1.8498966693878174, "learning_rate": 2.098e-05, "loss": 0.1424, "step": 1050 }, { "grad_norm": 1.600350022315979, "learning_rate": 2.118e-05, "loss": 0.125, "step": 1060 }, { "grad_norm": 1.6018701791763306, "learning_rate": 2.138e-05, "loss": 0.1183, "step": 1070 }, { "grad_norm": 1.6179834604263306, "learning_rate": 2.158e-05, "loss": 0.1285, "step": 1080 }, { "grad_norm": 1.5172916650772095, "learning_rate": 2.178e-05, "loss": 0.124, "step": 1090 }, { "grad_norm": 1.9335929155349731, "learning_rate": 2.198e-05, "loss": 0.1265, "step": 1100 }, { "grad_norm": 1.4130911827087402, "learning_rate": 2.218e-05, "loss": 0.1146, "step": 1110 }, { "grad_norm": 1.850577712059021, "learning_rate": 2.2380000000000003e-05, "loss": 0.1256, "step": 1120 }, { "grad_norm": 1.4143354892730713, "learning_rate": 2.258e-05, "loss": 0.1233, "step": 1130 }, { "grad_norm": 1.8339554071426392, "learning_rate": 2.2780000000000002e-05, "loss": 0.1224, "step": 1140 }, { "grad_norm": 1.4653862714767456, "learning_rate": 2.298e-05, "loss": 0.1204, "step": 1150 }, { "grad_norm": 1.778446912765503, "learning_rate": 2.318e-05, "loss": 0.1183, "step": 1160 }, { "grad_norm": 1.626905083656311, "learning_rate": 2.3380000000000003e-05, "loss": 0.1113, "step": 1170 }, { "grad_norm": 1.463860273361206, "learning_rate": 2.358e-05, "loss": 0.126, "step": 1180 }, { "grad_norm": 1.6939525604248047, "learning_rate": 2.3780000000000003e-05, "loss": 0.1253, "step": 1190 }, { "grad_norm": 1.430732250213623, "learning_rate": 2.398e-05, "loss": 0.1177, "step": 1200 }, { "grad_norm": 1.5416743755340576, "learning_rate": 2.418e-05, "loss": 0.1353, "step": 1210 }, { "grad_norm": 1.7357805967330933, "learning_rate": 2.438e-05, "loss": 0.1323, "step": 1220 }, { "grad_norm": 1.8270666599273682, "learning_rate": 2.4580000000000002e-05, "loss": 0.1098, "step": 1230 }, { "grad_norm": 1.630895733833313, "learning_rate": 2.478e-05, "loss": 0.1211, "step": 1240 }, { "grad_norm": 1.6119129657745361, "learning_rate": 2.498e-05, "loss": 0.1052, "step": 1250 }, { "grad_norm": 1.4018057584762573, "learning_rate": 2.5180000000000003e-05, "loss": 0.1171, "step": 1260 }, { "grad_norm": 1.8197362422943115, "learning_rate": 2.5380000000000004e-05, "loss": 0.1334, "step": 1270 }, { "grad_norm": 1.3923875093460083, "learning_rate": 2.5580000000000002e-05, "loss": 0.1182, "step": 1280 }, { "grad_norm": 1.041835904121399, "learning_rate": 2.5779999999999997e-05, "loss": 0.122, "step": 1290 }, { "grad_norm": 1.4869507551193237, "learning_rate": 2.598e-05, "loss": 0.1244, "step": 1300 }, { "grad_norm": 1.8758677244186401, "learning_rate": 2.618e-05, "loss": 0.1175, "step": 1310 }, { "grad_norm": 1.9378823041915894, "learning_rate": 2.6379999999999998e-05, "loss": 0.1096, "step": 1320 }, { "grad_norm": 1.723809003829956, "learning_rate": 2.658e-05, "loss": 0.1228, "step": 1330 }, { "grad_norm": 1.5816935300827026, "learning_rate": 2.678e-05, "loss": 0.119, "step": 1340 }, { "grad_norm": 1.7127572298049927, "learning_rate": 2.698e-05, "loss": 0.1124, "step": 1350 }, { "grad_norm": 1.5133707523345947, "learning_rate": 2.718e-05, "loss": 0.1141, "step": 1360 }, { "grad_norm": 1.4465445280075073, "learning_rate": 2.738e-05, "loss": 0.1149, "step": 1370 }, { "grad_norm": 1.5675820112228394, "learning_rate": 2.758e-05, "loss": 0.1158, "step": 1380 }, { "grad_norm": 1.5134754180908203, "learning_rate": 2.778e-05, "loss": 0.1124, "step": 1390 }, { "grad_norm": 1.872249722480774, "learning_rate": 2.798e-05, "loss": 0.1099, "step": 1400 }, { "grad_norm": 1.6044838428497314, "learning_rate": 2.818e-05, "loss": 0.115, "step": 1410 }, { "grad_norm": 1.6048882007598877, "learning_rate": 2.8380000000000003e-05, "loss": 0.1169, "step": 1420 }, { "grad_norm": 1.558488130569458, "learning_rate": 2.858e-05, "loss": 0.1205, "step": 1430 }, { "grad_norm": 1.6813035011291504, "learning_rate": 2.8780000000000002e-05, "loss": 0.1044, "step": 1440 }, { "grad_norm": 1.9473639726638794, "learning_rate": 2.898e-05, "loss": 0.1222, "step": 1450 }, { "grad_norm": 1.53206467628479, "learning_rate": 2.9180000000000002e-05, "loss": 0.1038, "step": 1460 }, { "grad_norm": 1.5294194221496582, "learning_rate": 2.9380000000000003e-05, "loss": 0.1188, "step": 1470 }, { "grad_norm": 1.4874261617660522, "learning_rate": 2.958e-05, "loss": 0.1113, "step": 1480 }, { "grad_norm": 1.4147335290908813, "learning_rate": 2.9780000000000003e-05, "loss": 0.1181, "step": 1490 }, { "grad_norm": 2.092658519744873, "learning_rate": 2.998e-05, "loss": 0.1149, "step": 1500 }, { "grad_norm": 1.7414278984069824, "learning_rate": 3.0180000000000002e-05, "loss": 0.1157, "step": 1510 }, { "grad_norm": 1.5015239715576172, "learning_rate": 3.0380000000000004e-05, "loss": 0.1043, "step": 1520 }, { "grad_norm": 1.6321384906768799, "learning_rate": 3.058e-05, "loss": 0.1067, "step": 1530 }, { "grad_norm": 1.3756071329116821, "learning_rate": 3.078e-05, "loss": 0.0968, "step": 1540 }, { "grad_norm": 1.4284217357635498, "learning_rate": 3.0980000000000005e-05, "loss": 0.0997, "step": 1550 }, { "grad_norm": 1.3079365491867065, "learning_rate": 3.118e-05, "loss": 0.105, "step": 1560 }, { "grad_norm": 1.4701611995697021, "learning_rate": 3.138e-05, "loss": 0.0998, "step": 1570 }, { "grad_norm": 1.1680314540863037, "learning_rate": 3.1580000000000006e-05, "loss": 0.1049, "step": 1580 }, { "grad_norm": 1.406514048576355, "learning_rate": 3.1780000000000004e-05, "loss": 0.1, "step": 1590 }, { "grad_norm": 1.0148186683654785, "learning_rate": 3.198e-05, "loss": 0.117, "step": 1600 }, { "grad_norm": 1.5542938709259033, "learning_rate": 3.218e-05, "loss": 0.0945, "step": 1610 }, { "grad_norm": 1.515797734260559, "learning_rate": 3.238e-05, "loss": 0.1141, "step": 1620 }, { "grad_norm": 1.2807338237762451, "learning_rate": 3.2579999999999996e-05, "loss": 0.1045, "step": 1630 }, { "grad_norm": 1.183184266090393, "learning_rate": 3.278e-05, "loss": 0.1094, "step": 1640 }, { "grad_norm": 1.2934596538543701, "learning_rate": 3.298e-05, "loss": 0.1112, "step": 1650 }, { "grad_norm": 1.2150341272354126, "learning_rate": 3.318e-05, "loss": 0.1071, "step": 1660 }, { "grad_norm": 1.137741208076477, "learning_rate": 3.338e-05, "loss": 0.1007, "step": 1670 }, { "grad_norm": 1.2264149188995361, "learning_rate": 3.358e-05, "loss": 0.1094, "step": 1680 }, { "grad_norm": 1.4653520584106445, "learning_rate": 3.378e-05, "loss": 0.1158, "step": 1690 }, { "grad_norm": 1.2986127138137817, "learning_rate": 3.398e-05, "loss": 0.1047, "step": 1700 }, { "grad_norm": 1.0463472604751587, "learning_rate": 3.418e-05, "loss": 0.1057, "step": 1710 }, { "grad_norm": 1.1150590181350708, "learning_rate": 3.438e-05, "loss": 0.1153, "step": 1720 }, { "grad_norm": 1.2571240663528442, "learning_rate": 3.4580000000000004e-05, "loss": 0.1024, "step": 1730 }, { "grad_norm": 1.0144950151443481, "learning_rate": 3.478e-05, "loss": 0.0981, "step": 1740 }, { "grad_norm": 1.1527057886123657, "learning_rate": 3.498e-05, "loss": 0.1005, "step": 1750 }, { "grad_norm": 1.4035841226577759, "learning_rate": 3.518e-05, "loss": 0.1129, "step": 1760 }, { "grad_norm": 1.1984410285949707, "learning_rate": 3.5380000000000003e-05, "loss": 0.1077, "step": 1770 }, { "grad_norm": 0.9347260594367981, "learning_rate": 3.558e-05, "loss": 0.0973, "step": 1780 }, { "grad_norm": 1.0611302852630615, "learning_rate": 3.578e-05, "loss": 0.0985, "step": 1790 }, { "grad_norm": 1.3737915754318237, "learning_rate": 3.5980000000000004e-05, "loss": 0.1007, "step": 1800 }, { "grad_norm": 1.3762823343276978, "learning_rate": 3.618e-05, "loss": 0.1025, "step": 1810 }, { "grad_norm": 1.2935340404510498, "learning_rate": 3.638e-05, "loss": 0.093, "step": 1820 }, { "grad_norm": 1.177607774734497, "learning_rate": 3.6580000000000006e-05, "loss": 0.1001, "step": 1830 }, { "grad_norm": 1.112083077430725, "learning_rate": 3.6780000000000004e-05, "loss": 0.0973, "step": 1840 }, { "grad_norm": 1.0434048175811768, "learning_rate": 3.698e-05, "loss": 0.0893, "step": 1850 }, { "grad_norm": 1.3256551027297974, "learning_rate": 3.7180000000000007e-05, "loss": 0.0875, "step": 1860 }, { "grad_norm": 0.8122597336769104, "learning_rate": 3.7380000000000005e-05, "loss": 0.0963, "step": 1870 }, { "grad_norm": 1.0861409902572632, "learning_rate": 3.758e-05, "loss": 0.0931, "step": 1880 }, { "grad_norm": 1.2610929012298584, "learning_rate": 3.778000000000001e-05, "loss": 0.0994, "step": 1890 }, { "grad_norm": 1.193092703819275, "learning_rate": 3.7980000000000006e-05, "loss": 0.096, "step": 1900 }, { "grad_norm": 1.3506171703338623, "learning_rate": 3.818e-05, "loss": 0.0926, "step": 1910 }, { "grad_norm": 1.4965968132019043, "learning_rate": 3.838e-05, "loss": 0.1079, "step": 1920 }, { "grad_norm": 1.053296685218811, "learning_rate": 3.858e-05, "loss": 0.106, "step": 1930 }, { "grad_norm": 1.1691193580627441, "learning_rate": 3.878e-05, "loss": 0.1009, "step": 1940 }, { "grad_norm": 1.0233510732650757, "learning_rate": 3.898e-05, "loss": 0.1015, "step": 1950 }, { "grad_norm": 0.8906300067901611, "learning_rate": 3.918e-05, "loss": 0.1002, "step": 1960 }, { "grad_norm": 0.8604732751846313, "learning_rate": 3.938e-05, "loss": 0.1016, "step": 1970 }, { "grad_norm": 1.2033056020736694, "learning_rate": 3.958e-05, "loss": 0.1032, "step": 1980 }, { "grad_norm": 1.0708104372024536, "learning_rate": 3.978e-05, "loss": 0.0881, "step": 1990 }, { "grad_norm": 1.2015386819839478, "learning_rate": 3.998e-05, "loss": 0.0949, "step": 2000 }, { "grad_norm": 0.937743067741394, "learning_rate": 4.018e-05, "loss": 0.1021, "step": 2010 }, { "grad_norm": 1.0698339939117432, "learning_rate": 4.038e-05, "loss": 0.0831, "step": 2020 }, { "grad_norm": 1.2554341554641724, "learning_rate": 4.058e-05, "loss": 0.0993, "step": 2030 }, { "grad_norm": 0.8914675116539001, "learning_rate": 4.078e-05, "loss": 0.092, "step": 2040 }, { "grad_norm": 1.3133528232574463, "learning_rate": 4.0980000000000004e-05, "loss": 0.1052, "step": 2050 }, { "grad_norm": 1.2768296003341675, "learning_rate": 4.118e-05, "loss": 0.0931, "step": 2060 }, { "grad_norm": 1.4005250930786133, "learning_rate": 4.138e-05, "loss": 0.0969, "step": 2070 }, { "grad_norm": 1.2988919019699097, "learning_rate": 4.1580000000000005e-05, "loss": 0.0951, "step": 2080 }, { "grad_norm": 0.8640112280845642, "learning_rate": 4.178e-05, "loss": 0.092, "step": 2090 }, { "grad_norm": 1.0795797109603882, "learning_rate": 4.198e-05, "loss": 0.0904, "step": 2100 }, { "grad_norm": 0.9410653710365295, "learning_rate": 4.2180000000000006e-05, "loss": 0.0872, "step": 2110 }, { "grad_norm": 1.1896841526031494, "learning_rate": 4.2380000000000004e-05, "loss": 0.097, "step": 2120 }, { "grad_norm": 1.2544245719909668, "learning_rate": 4.258e-05, "loss": 0.0918, "step": 2130 }, { "grad_norm": 1.1860486268997192, "learning_rate": 4.278e-05, "loss": 0.1043, "step": 2140 }, { "grad_norm": 1.045448660850525, "learning_rate": 4.2980000000000005e-05, "loss": 0.0819, "step": 2150 }, { "grad_norm": 1.3089593648910522, "learning_rate": 4.318e-05, "loss": 0.0876, "step": 2160 }, { "grad_norm": 1.4617528915405273, "learning_rate": 4.338e-05, "loss": 0.1123, "step": 2170 }, { "grad_norm": 0.9591245055198669, "learning_rate": 4.3580000000000006e-05, "loss": 0.1001, "step": 2180 }, { "grad_norm": 0.908771812915802, "learning_rate": 4.3780000000000004e-05, "loss": 0.0897, "step": 2190 }, { "grad_norm": 0.6492735147476196, "learning_rate": 4.398e-05, "loss": 0.0892, "step": 2200 }, { "grad_norm": 0.9633204340934753, "learning_rate": 4.418000000000001e-05, "loss": 0.1068, "step": 2210 }, { "grad_norm": 1.3645237684249878, "learning_rate": 4.438e-05, "loss": 0.0966, "step": 2220 }, { "grad_norm": 0.9686059951782227, "learning_rate": 4.458e-05, "loss": 0.0896, "step": 2230 }, { "grad_norm": 1.1634559631347656, "learning_rate": 4.478e-05, "loss": 0.0782, "step": 2240 }, { "grad_norm": 1.2163933515548706, "learning_rate": 4.498e-05, "loss": 0.0991, "step": 2250 }, { "grad_norm": 0.8722169995307922, "learning_rate": 4.518e-05, "loss": 0.0893, "step": 2260 }, { "grad_norm": 1.1959494352340698, "learning_rate": 4.538e-05, "loss": 0.1013, "step": 2270 }, { "grad_norm": 0.9531932473182678, "learning_rate": 4.558e-05, "loss": 0.0904, "step": 2280 }, { "grad_norm": 1.1789480447769165, "learning_rate": 4.578e-05, "loss": 0.0911, "step": 2290 }, { "grad_norm": 1.3253966569900513, "learning_rate": 4.5980000000000004e-05, "loss": 0.0886, "step": 2300 }, { "grad_norm": 0.9672907590866089, "learning_rate": 4.618e-05, "loss": 0.1002, "step": 2310 }, { "grad_norm": 0.9954025745391846, "learning_rate": 4.638e-05, "loss": 0.0944, "step": 2320 }, { "grad_norm": 1.0856378078460693, "learning_rate": 4.6580000000000005e-05, "loss": 0.1056, "step": 2330 }, { "grad_norm": 1.2152234315872192, "learning_rate": 4.678e-05, "loss": 0.0882, "step": 2340 }, { "grad_norm": 0.9829302430152893, "learning_rate": 4.698e-05, "loss": 0.0868, "step": 2350 }, { "grad_norm": 1.0614885091781616, "learning_rate": 4.718e-05, "loss": 0.0895, "step": 2360 }, { "grad_norm": 1.1140849590301514, "learning_rate": 4.7380000000000004e-05, "loss": 0.0905, "step": 2370 }, { "grad_norm": 0.8411664366722107, "learning_rate": 4.758e-05, "loss": 0.085, "step": 2380 }, { "grad_norm": 0.8826441168785095, "learning_rate": 4.778e-05, "loss": 0.0801, "step": 2390 }, { "grad_norm": 0.9238112568855286, "learning_rate": 4.7980000000000005e-05, "loss": 0.0828, "step": 2400 }, { "grad_norm": 0.9299654960632324, "learning_rate": 4.818e-05, "loss": 0.0855, "step": 2410 }, { "grad_norm": 1.0628398656845093, "learning_rate": 4.838e-05, "loss": 0.0822, "step": 2420 }, { "grad_norm": 1.1392028331756592, "learning_rate": 4.8580000000000006e-05, "loss": 0.09, "step": 2430 }, { "grad_norm": 0.9941686391830444, "learning_rate": 4.8780000000000004e-05, "loss": 0.08, "step": 2440 }, { "grad_norm": 1.6255854368209839, "learning_rate": 4.898e-05, "loss": 0.0967, "step": 2450 }, { "grad_norm": 0.8056438565254211, "learning_rate": 4.918000000000001e-05, "loss": 0.0871, "step": 2460 }, { "grad_norm": 0.8102939128875732, "learning_rate": 4.9380000000000005e-05, "loss": 0.0901, "step": 2470 }, { "grad_norm": 0.8373276591300964, "learning_rate": 4.958e-05, "loss": 0.0827, "step": 2480 }, { "grad_norm": 0.7185338139533997, "learning_rate": 4.978e-05, "loss": 0.0822, "step": 2490 }, { "grad_norm": 0.8493714332580566, "learning_rate": 4.9980000000000006e-05, "loss": 0.0883, "step": 2500 }, { "grad_norm": 0.8506907820701599, "learning_rate": 5.0180000000000004e-05, "loss": 0.0867, "step": 2510 }, { "grad_norm": 0.9758744239807129, "learning_rate": 5.038e-05, "loss": 0.0778, "step": 2520 }, { "grad_norm": 0.8658690452575684, "learning_rate": 5.058000000000001e-05, "loss": 0.0769, "step": 2530 }, { "grad_norm": 1.0286805629730225, "learning_rate": 5.0780000000000005e-05, "loss": 0.0783, "step": 2540 }, { "grad_norm": 0.9064648151397705, "learning_rate": 5.098e-05, "loss": 0.0883, "step": 2550 }, { "grad_norm": 0.9923864603042603, "learning_rate": 5.118000000000001e-05, "loss": 0.0765, "step": 2560 }, { "grad_norm": 0.8054091334342957, "learning_rate": 5.1380000000000006e-05, "loss": 0.0915, "step": 2570 }, { "grad_norm": 1.1013610363006592, "learning_rate": 5.1580000000000004e-05, "loss": 0.0835, "step": 2580 }, { "grad_norm": 1.0966627597808838, "learning_rate": 5.178000000000001e-05, "loss": 0.0954, "step": 2590 }, { "grad_norm": 0.9028438925743103, "learning_rate": 5.198000000000001e-05, "loss": 0.0878, "step": 2600 }, { "grad_norm": 0.7007715106010437, "learning_rate": 5.2180000000000005e-05, "loss": 0.0866, "step": 2610 }, { "grad_norm": 0.9387818574905396, "learning_rate": 5.238000000000001e-05, "loss": 0.0785, "step": 2620 }, { "grad_norm": 0.9752476215362549, "learning_rate": 5.258000000000001e-05, "loss": 0.0788, "step": 2630 }, { "grad_norm": 0.7812796235084534, "learning_rate": 5.2780000000000006e-05, "loss": 0.0916, "step": 2640 }, { "grad_norm": 0.7433364987373352, "learning_rate": 5.2980000000000004e-05, "loss": 0.075, "step": 2650 }, { "grad_norm": 0.7556120753288269, "learning_rate": 5.318000000000001e-05, "loss": 0.0924, "step": 2660 }, { "grad_norm": 0.7730178236961365, "learning_rate": 5.338000000000001e-05, "loss": 0.0839, "step": 2670 }, { "grad_norm": 0.6746904850006104, "learning_rate": 5.3580000000000005e-05, "loss": 0.0902, "step": 2680 }, { "grad_norm": 0.878186821937561, "learning_rate": 5.378e-05, "loss": 0.0877, "step": 2690 }, { "grad_norm": 0.7982309460639954, "learning_rate": 5.3979999999999995e-05, "loss": 0.0854, "step": 2700 }, { "grad_norm": 0.8585929870605469, "learning_rate": 5.418e-05, "loss": 0.0783, "step": 2710 }, { "grad_norm": 0.6710546612739563, "learning_rate": 5.438e-05, "loss": 0.0756, "step": 2720 }, { "grad_norm": 0.8046708106994629, "learning_rate": 5.4579999999999996e-05, "loss": 0.0888, "step": 2730 }, { "grad_norm": 0.798771858215332, "learning_rate": 5.478e-05, "loss": 0.088, "step": 2740 }, { "grad_norm": 1.0829910039901733, "learning_rate": 5.498e-05, "loss": 0.0756, "step": 2750 }, { "grad_norm": 0.9812852740287781, "learning_rate": 5.518e-05, "loss": 0.078, "step": 2760 }, { "grad_norm": 0.7689820528030396, "learning_rate": 5.538e-05, "loss": 0.0881, "step": 2770 }, { "grad_norm": 0.8602181077003479, "learning_rate": 5.558e-05, "loss": 0.0831, "step": 2780 }, { "grad_norm": 0.7148047089576721, "learning_rate": 5.578e-05, "loss": 0.0768, "step": 2790 }, { "grad_norm": 0.8527437448501587, "learning_rate": 5.5979999999999996e-05, "loss": 0.0745, "step": 2800 }, { "grad_norm": 0.8923451900482178, "learning_rate": 5.618e-05, "loss": 0.0859, "step": 2810 }, { "grad_norm": 0.8212958574295044, "learning_rate": 5.638e-05, "loss": 0.0746, "step": 2820 }, { "grad_norm": 0.7652589082717896, "learning_rate": 5.658e-05, "loss": 0.072, "step": 2830 }, { "grad_norm": 0.8929500579833984, "learning_rate": 5.678e-05, "loss": 0.0766, "step": 2840 }, { "grad_norm": 0.7901267409324646, "learning_rate": 5.698e-05, "loss": 0.085, "step": 2850 }, { "grad_norm": 0.8382807970046997, "learning_rate": 5.718e-05, "loss": 0.0884, "step": 2860 }, { "grad_norm": 0.6783472895622253, "learning_rate": 5.738e-05, "loss": 0.0818, "step": 2870 }, { "grad_norm": 0.7239498496055603, "learning_rate": 5.758e-05, "loss": 0.0737, "step": 2880 }, { "grad_norm": 0.7628580331802368, "learning_rate": 5.778e-05, "loss": 0.0716, "step": 2890 }, { "grad_norm": 1.0676965713500977, "learning_rate": 5.7980000000000004e-05, "loss": 0.0832, "step": 2900 }, { "grad_norm": 0.8924280405044556, "learning_rate": 5.818e-05, "loss": 0.0854, "step": 2910 }, { "grad_norm": 0.7455031871795654, "learning_rate": 5.838e-05, "loss": 0.083, "step": 2920 }, { "grad_norm": 0.9217103123664856, "learning_rate": 5.858e-05, "loss": 0.0841, "step": 2930 }, { "grad_norm": 0.985454261302948, "learning_rate": 5.878e-05, "loss": 0.0982, "step": 2940 }, { "grad_norm": 0.6373068690299988, "learning_rate": 5.898e-05, "loss": 0.085, "step": 2950 }, { "grad_norm": 0.7201591730117798, "learning_rate": 5.918e-05, "loss": 0.0811, "step": 2960 }, { "grad_norm": 0.7736907601356506, "learning_rate": 5.9380000000000004e-05, "loss": 0.0783, "step": 2970 }, { "grad_norm": 0.6466020941734314, "learning_rate": 5.958e-05, "loss": 0.0792, "step": 2980 }, { "grad_norm": 0.7938363552093506, "learning_rate": 5.978e-05, "loss": 0.0836, "step": 2990 }, { "grad_norm": 0.8457830548286438, "learning_rate": 5.9980000000000005e-05, "loss": 0.0666, "step": 3000 }, { "grad_norm": 0.7363418340682983, "learning_rate": 6.018e-05, "loss": 0.0781, "step": 3010 }, { "grad_norm": 0.686091959476471, "learning_rate": 6.038e-05, "loss": 0.0688, "step": 3020 }, { "grad_norm": 0.6579721570014954, "learning_rate": 6.0580000000000006e-05, "loss": 0.0943, "step": 3030 }, { "grad_norm": 0.7850069999694824, "learning_rate": 6.0780000000000004e-05, "loss": 0.0833, "step": 3040 }, { "grad_norm": 0.5588982701301575, "learning_rate": 6.098e-05, "loss": 0.0837, "step": 3050 }, { "grad_norm": 0.7701879739761353, "learning_rate": 6.118000000000001e-05, "loss": 0.088, "step": 3060 }, { "grad_norm": 0.7791573405265808, "learning_rate": 6.138e-05, "loss": 0.0804, "step": 3070 }, { "grad_norm": 0.7099859118461609, "learning_rate": 6.158e-05, "loss": 0.087, "step": 3080 }, { "grad_norm": 0.7828622460365295, "learning_rate": 6.178000000000001e-05, "loss": 0.0818, "step": 3090 }, { "grad_norm": 0.701026976108551, "learning_rate": 6.198e-05, "loss": 0.0806, "step": 3100 }, { "grad_norm": 0.8385618329048157, "learning_rate": 6.218e-05, "loss": 0.073, "step": 3110 }, { "grad_norm": 0.8139421939849854, "learning_rate": 6.238000000000001e-05, "loss": 0.0783, "step": 3120 }, { "grad_norm": 1.0593305826187134, "learning_rate": 6.258e-05, "loss": 0.0831, "step": 3130 }, { "grad_norm": 0.6710803508758545, "learning_rate": 6.278e-05, "loss": 0.0724, "step": 3140 }, { "grad_norm": 0.7578899264335632, "learning_rate": 6.298000000000001e-05, "loss": 0.0672, "step": 3150 }, { "grad_norm": 0.6445244550704956, "learning_rate": 6.318e-05, "loss": 0.077, "step": 3160 }, { "grad_norm": 0.6721904277801514, "learning_rate": 6.338e-05, "loss": 0.0719, "step": 3170 }, { "grad_norm": 0.658791184425354, "learning_rate": 6.358000000000001e-05, "loss": 0.0798, "step": 3180 }, { "grad_norm": 0.6775658130645752, "learning_rate": 6.378e-05, "loss": 0.0651, "step": 3190 }, { "grad_norm": 0.8936405181884766, "learning_rate": 6.398000000000001e-05, "loss": 0.0784, "step": 3200 }, { "grad_norm": 0.7268358469009399, "learning_rate": 6.418000000000001e-05, "loss": 0.0817, "step": 3210 }, { "grad_norm": 0.5887531042098999, "learning_rate": 6.438e-05, "loss": 0.0643, "step": 3220 }, { "grad_norm": 0.695975661277771, "learning_rate": 6.458000000000001e-05, "loss": 0.0779, "step": 3230 }, { "grad_norm": 0.709709882736206, "learning_rate": 6.478000000000001e-05, "loss": 0.0729, "step": 3240 }, { "grad_norm": 1.1079152822494507, "learning_rate": 6.498e-05, "loss": 0.0846, "step": 3250 }, { "grad_norm": 0.8895103931427002, "learning_rate": 6.518000000000001e-05, "loss": 0.0816, "step": 3260 }, { "grad_norm": 0.8708611130714417, "learning_rate": 6.538000000000001e-05, "loss": 0.0848, "step": 3270 }, { "grad_norm": 0.573621928691864, "learning_rate": 6.558e-05, "loss": 0.0803, "step": 3280 }, { "grad_norm": 0.5341253876686096, "learning_rate": 6.578000000000001e-05, "loss": 0.0736, "step": 3290 }, { "grad_norm": 0.7932272553443909, "learning_rate": 6.598e-05, "loss": 0.074, "step": 3300 }, { "grad_norm": 0.637997031211853, "learning_rate": 6.618e-05, "loss": 0.0764, "step": 3310 }, { "grad_norm": 0.7436404228210449, "learning_rate": 6.638e-05, "loss": 0.0825, "step": 3320 }, { "grad_norm": 0.7574580311775208, "learning_rate": 6.658e-05, "loss": 0.0878, "step": 3330 }, { "grad_norm": 0.6752119660377502, "learning_rate": 6.678e-05, "loss": 0.0736, "step": 3340 }, { "grad_norm": 0.6176100373268127, "learning_rate": 6.698e-05, "loss": 0.0753, "step": 3350 }, { "grad_norm": 0.6247387528419495, "learning_rate": 6.718e-05, "loss": 0.0706, "step": 3360 }, { "grad_norm": 0.6345912218093872, "learning_rate": 6.738e-05, "loss": 0.071, "step": 3370 }, { "grad_norm": 0.8483233451843262, "learning_rate": 6.758e-05, "loss": 0.0811, "step": 3380 }, { "grad_norm": 0.6762822270393372, "learning_rate": 6.778e-05, "loss": 0.0779, "step": 3390 }, { "grad_norm": 0.7730652689933777, "learning_rate": 6.798e-05, "loss": 0.0795, "step": 3400 }, { "grad_norm": 0.849593997001648, "learning_rate": 6.818e-05, "loss": 0.0754, "step": 3410 }, { "grad_norm": 0.5692761540412903, "learning_rate": 6.838e-05, "loss": 0.0684, "step": 3420 }, { "grad_norm": 0.7750732898712158, "learning_rate": 6.858e-05, "loss": 0.0788, "step": 3430 }, { "grad_norm": 0.8249272108078003, "learning_rate": 6.878e-05, "loss": 0.0727, "step": 3440 }, { "grad_norm": 0.8045344352722168, "learning_rate": 6.898e-05, "loss": 0.0784, "step": 3450 }, { "grad_norm": 0.7028372883796692, "learning_rate": 6.918e-05, "loss": 0.0775, "step": 3460 }, { "grad_norm": 0.6903756260871887, "learning_rate": 6.938e-05, "loss": 0.071, "step": 3470 }, { "grad_norm": 0.7190449833869934, "learning_rate": 6.958e-05, "loss": 0.0731, "step": 3480 }, { "grad_norm": 0.6560386419296265, "learning_rate": 6.978e-05, "loss": 0.0774, "step": 3490 }, { "grad_norm": 0.7669695615768433, "learning_rate": 6.998e-05, "loss": 0.0791, "step": 3500 }, { "grad_norm": 0.6766068339347839, "learning_rate": 7.018e-05, "loss": 0.0763, "step": 3510 }, { "grad_norm": 0.5767879486083984, "learning_rate": 7.038e-05, "loss": 0.075, "step": 3520 }, { "grad_norm": 0.9619874358177185, "learning_rate": 7.058e-05, "loss": 0.0851, "step": 3530 }, { "grad_norm": 0.7226930260658264, "learning_rate": 7.078e-05, "loss": 0.0767, "step": 3540 }, { "grad_norm": 0.6095721125602722, "learning_rate": 7.098e-05, "loss": 0.0782, "step": 3550 }, { "grad_norm": 0.6947830319404602, "learning_rate": 7.118e-05, "loss": 0.0774, "step": 3560 }, { "grad_norm": 0.7306665778160095, "learning_rate": 7.138e-05, "loss": 0.0726, "step": 3570 }, { "grad_norm": 0.7859469652175903, "learning_rate": 7.158e-05, "loss": 0.0773, "step": 3580 }, { "grad_norm": 0.939265787601471, "learning_rate": 7.178000000000001e-05, "loss": 0.0777, "step": 3590 }, { "grad_norm": 0.7816467881202698, "learning_rate": 7.198e-05, "loss": 0.0703, "step": 3600 }, { "grad_norm": 0.6888828873634338, "learning_rate": 7.218e-05, "loss": 0.0682, "step": 3610 }, { "grad_norm": 0.9029473662376404, "learning_rate": 7.238000000000001e-05, "loss": 0.0843, "step": 3620 }, { "grad_norm": 0.5385906100273132, "learning_rate": 7.258e-05, "loss": 0.0754, "step": 3630 }, { "grad_norm": 0.6447309255599976, "learning_rate": 7.278e-05, "loss": 0.0736, "step": 3640 }, { "grad_norm": 0.6336073875427246, "learning_rate": 7.298000000000001e-05, "loss": 0.0855, "step": 3650 }, { "grad_norm": 0.6040142774581909, "learning_rate": 7.318e-05, "loss": 0.0744, "step": 3660 }, { "grad_norm": 0.7830803394317627, "learning_rate": 7.338e-05, "loss": 0.0749, "step": 3670 }, { "grad_norm": 0.7362026572227478, "learning_rate": 7.358000000000001e-05, "loss": 0.0682, "step": 3680 }, { "grad_norm": 0.6518739461898804, "learning_rate": 7.378e-05, "loss": 0.0762, "step": 3690 }, { "grad_norm": 0.607718288898468, "learning_rate": 7.398e-05, "loss": 0.071, "step": 3700 }, { "grad_norm": 0.6504892110824585, "learning_rate": 7.418000000000001e-05, "loss": 0.0625, "step": 3710 }, { "grad_norm": 0.562951922416687, "learning_rate": 7.438e-05, "loss": 0.0763, "step": 3720 }, { "grad_norm": 0.5356259942054749, "learning_rate": 7.458000000000001e-05, "loss": 0.0685, "step": 3730 }, { "grad_norm": 0.6308535933494568, "learning_rate": 7.478e-05, "loss": 0.0722, "step": 3740 }, { "grad_norm": 0.7760159969329834, "learning_rate": 7.498e-05, "loss": 0.0681, "step": 3750 }, { "grad_norm": 0.6545782685279846, "learning_rate": 7.518000000000001e-05, "loss": 0.0678, "step": 3760 }, { "grad_norm": 0.6055973172187805, "learning_rate": 7.538e-05, "loss": 0.0645, "step": 3770 }, { "grad_norm": 0.5603523254394531, "learning_rate": 7.558e-05, "loss": 0.0627, "step": 3780 }, { "grad_norm": 0.6331689357757568, "learning_rate": 7.578000000000001e-05, "loss": 0.0707, "step": 3790 }, { "grad_norm": 0.5651906728744507, "learning_rate": 7.598e-05, "loss": 0.0594, "step": 3800 }, { "grad_norm": 0.5818098783493042, "learning_rate": 7.618e-05, "loss": 0.0759, "step": 3810 }, { "grad_norm": 0.7271901965141296, "learning_rate": 7.638000000000001e-05, "loss": 0.0745, "step": 3820 }, { "grad_norm": 0.7587158679962158, "learning_rate": 7.658e-05, "loss": 0.0758, "step": 3830 }, { "grad_norm": 0.5288383960723877, "learning_rate": 7.678000000000001e-05, "loss": 0.0648, "step": 3840 }, { "grad_norm": 0.5369072556495667, "learning_rate": 7.698000000000001e-05, "loss": 0.0795, "step": 3850 }, { "grad_norm": 0.7513662576675415, "learning_rate": 7.718e-05, "loss": 0.0741, "step": 3860 }, { "grad_norm": 0.5982961058616638, "learning_rate": 7.738000000000001e-05, "loss": 0.0786, "step": 3870 }, { "grad_norm": 0.5358338356018066, "learning_rate": 7.758000000000001e-05, "loss": 0.0695, "step": 3880 }, { "grad_norm": 0.5780497193336487, "learning_rate": 7.778e-05, "loss": 0.0795, "step": 3890 }, { "grad_norm": 0.4997905194759369, "learning_rate": 7.798000000000001e-05, "loss": 0.077, "step": 3900 }, { "grad_norm": 0.5546618700027466, "learning_rate": 7.818000000000001e-05, "loss": 0.0694, "step": 3910 }, { "grad_norm": 0.7473470568656921, "learning_rate": 7.838e-05, "loss": 0.0729, "step": 3920 }, { "grad_norm": 0.6219871044158936, "learning_rate": 7.858000000000001e-05, "loss": 0.066, "step": 3930 }, { "grad_norm": 0.7440235018730164, "learning_rate": 7.878e-05, "loss": 0.0776, "step": 3940 }, { "grad_norm": 0.6325212717056274, "learning_rate": 7.897999999999999e-05, "loss": 0.0761, "step": 3950 }, { "grad_norm": 0.4927770793437958, "learning_rate": 7.918e-05, "loss": 0.0816, "step": 3960 }, { "grad_norm": 0.682582437992096, "learning_rate": 7.938e-05, "loss": 0.0817, "step": 3970 }, { "grad_norm": 0.8074224591255188, "learning_rate": 7.958e-05, "loss": 0.0733, "step": 3980 }, { "grad_norm": 0.5221471786499023, "learning_rate": 7.978e-05, "loss": 0.0741, "step": 3990 }, { "grad_norm": 0.7943696975708008, "learning_rate": 7.998e-05, "loss": 0.0838, "step": 4000 }, { "grad_norm": 0.6027992963790894, "learning_rate": 8.018e-05, "loss": 0.0781, "step": 4010 }, { "grad_norm": 0.7089065909385681, "learning_rate": 8.038e-05, "loss": 0.0785, "step": 4020 }, { "grad_norm": 0.6590677499771118, "learning_rate": 8.058e-05, "loss": 0.0778, "step": 4030 }, { "grad_norm": 0.5493741035461426, "learning_rate": 8.078e-05, "loss": 0.0826, "step": 4040 }, { "grad_norm": 0.6736083030700684, "learning_rate": 8.098e-05, "loss": 0.0794, "step": 4050 }, { "grad_norm": 0.6986470818519592, "learning_rate": 8.118e-05, "loss": 0.0816, "step": 4060 }, { "grad_norm": 0.6084708571434021, "learning_rate": 8.138e-05, "loss": 0.0806, "step": 4070 }, { "grad_norm": 0.4387200176715851, "learning_rate": 8.158e-05, "loss": 0.0762, "step": 4080 }, { "grad_norm": 0.708766520023346, "learning_rate": 8.178e-05, "loss": 0.0741, "step": 4090 }, { "grad_norm": 0.6516726613044739, "learning_rate": 8.198e-05, "loss": 0.0725, "step": 4100 }, { "grad_norm": 0.5477268695831299, "learning_rate": 8.218e-05, "loss": 0.0734, "step": 4110 }, { "grad_norm": 0.49008721113204956, "learning_rate": 8.238000000000001e-05, "loss": 0.0736, "step": 4120 }, { "grad_norm": 0.592712938785553, "learning_rate": 8.258e-05, "loss": 0.073, "step": 4130 }, { "grad_norm": 0.5666195750236511, "learning_rate": 8.278e-05, "loss": 0.0729, "step": 4140 }, { "grad_norm": 0.7040342688560486, "learning_rate": 8.298000000000001e-05, "loss": 0.084, "step": 4150 }, { "grad_norm": 0.5405345559120178, "learning_rate": 8.318e-05, "loss": 0.0673, "step": 4160 }, { "grad_norm": 0.8642361760139465, "learning_rate": 8.338e-05, "loss": 0.0721, "step": 4170 }, { "grad_norm": 0.6923254728317261, "learning_rate": 8.358e-05, "loss": 0.0681, "step": 4180 }, { "grad_norm": 0.5574576258659363, "learning_rate": 8.378e-05, "loss": 0.0677, "step": 4190 }, { "grad_norm": 0.6036507487297058, "learning_rate": 8.398e-05, "loss": 0.0718, "step": 4200 }, { "grad_norm": 0.7173923254013062, "learning_rate": 8.418e-05, "loss": 0.078, "step": 4210 }, { "grad_norm": 1.3902225494384766, "learning_rate": 8.438e-05, "loss": 0.0796, "step": 4220 }, { "grad_norm": 0.6476540565490723, "learning_rate": 8.458e-05, "loss": 0.0756, "step": 4230 }, { "grad_norm": 0.5445621013641357, "learning_rate": 8.478e-05, "loss": 0.0789, "step": 4240 }, { "grad_norm": 0.5061116218566895, "learning_rate": 8.498e-05, "loss": 0.0774, "step": 4250 }, { "grad_norm": 0.6528274416923523, "learning_rate": 8.518000000000001e-05, "loss": 0.0781, "step": 4260 }, { "grad_norm": 0.552505373954773, "learning_rate": 8.538e-05, "loss": 0.0742, "step": 4270 }, { "grad_norm": 0.7733786106109619, "learning_rate": 8.558e-05, "loss": 0.0816, "step": 4280 }, { "grad_norm": 0.5476343035697937, "learning_rate": 8.578000000000001e-05, "loss": 0.0697, "step": 4290 }, { "grad_norm": 0.597711980342865, "learning_rate": 8.598e-05, "loss": 0.0768, "step": 4300 }, { "grad_norm": 0.671457052230835, "learning_rate": 8.618e-05, "loss": 0.0767, "step": 4310 }, { "grad_norm": 0.4863736927509308, "learning_rate": 8.638000000000001e-05, "loss": 0.0706, "step": 4320 }, { "grad_norm": 0.7833236455917358, "learning_rate": 8.658e-05, "loss": 0.0706, "step": 4330 }, { "grad_norm": 0.574954092502594, "learning_rate": 8.678e-05, "loss": 0.0744, "step": 4340 }, { "grad_norm": 0.848329484462738, "learning_rate": 8.698000000000001e-05, "loss": 0.069, "step": 4350 }, { "grad_norm": 0.5513610243797302, "learning_rate": 8.718e-05, "loss": 0.0752, "step": 4360 }, { "grad_norm": 0.6922655701637268, "learning_rate": 8.738000000000001e-05, "loss": 0.0801, "step": 4370 }, { "grad_norm": 0.6096819639205933, "learning_rate": 8.758000000000001e-05, "loss": 0.0747, "step": 4380 }, { "grad_norm": 0.5967550277709961, "learning_rate": 8.778e-05, "loss": 0.0686, "step": 4390 }, { "grad_norm": 0.5499820113182068, "learning_rate": 8.798000000000001e-05, "loss": 0.0755, "step": 4400 }, { "grad_norm": 0.6125385165214539, "learning_rate": 8.818000000000001e-05, "loss": 0.0806, "step": 4410 }, { "grad_norm": 0.6512491106987, "learning_rate": 8.838e-05, "loss": 0.0754, "step": 4420 }, { "grad_norm": 0.5914857983589172, "learning_rate": 8.858000000000001e-05, "loss": 0.0836, "step": 4430 }, { "grad_norm": 0.4492367208003998, "learning_rate": 8.878000000000001e-05, "loss": 0.0658, "step": 4440 }, { "grad_norm": 0.5012035965919495, "learning_rate": 8.898e-05, "loss": 0.0711, "step": 4450 }, { "grad_norm": 0.7608851790428162, "learning_rate": 8.918000000000001e-05, "loss": 0.0775, "step": 4460 }, { "grad_norm": 0.5314978957176208, "learning_rate": 8.938e-05, "loss": 0.0716, "step": 4470 }, { "grad_norm": 0.48411011695861816, "learning_rate": 8.958e-05, "loss": 0.0759, "step": 4480 }, { "grad_norm": 0.6099975109100342, "learning_rate": 8.978000000000001e-05, "loss": 0.0722, "step": 4490 }, { "grad_norm": 0.5218112468719482, "learning_rate": 8.998e-05, "loss": 0.0733, "step": 4500 }, { "grad_norm": 0.5964562296867371, "learning_rate": 9.018000000000001e-05, "loss": 0.0747, "step": 4510 }, { "grad_norm": 0.5776119828224182, "learning_rate": 9.038000000000001e-05, "loss": 0.0736, "step": 4520 }, { "grad_norm": 0.47986727952957153, "learning_rate": 9.058e-05, "loss": 0.0688, "step": 4530 }, { "grad_norm": 0.528157651424408, "learning_rate": 9.078000000000001e-05, "loss": 0.0748, "step": 4540 }, { "grad_norm": 0.58983314037323, "learning_rate": 9.098000000000001e-05, "loss": 0.0737, "step": 4550 }, { "grad_norm": 0.6674344539642334, "learning_rate": 9.118e-05, "loss": 0.075, "step": 4560 }, { "grad_norm": 0.5572109818458557, "learning_rate": 9.138e-05, "loss": 0.0784, "step": 4570 }, { "grad_norm": 0.6629700660705566, "learning_rate": 9.158e-05, "loss": 0.0825, "step": 4580 }, { "grad_norm": 0.5988973379135132, "learning_rate": 9.178e-05, "loss": 0.0735, "step": 4590 }, { "grad_norm": 0.5844960808753967, "learning_rate": 9.198e-05, "loss": 0.0825, "step": 4600 }, { "grad_norm": 0.6806857585906982, "learning_rate": 9.218e-05, "loss": 0.0871, "step": 4610 }, { "grad_norm": 0.6931775212287903, "learning_rate": 9.238e-05, "loss": 0.0754, "step": 4620 }, { "grad_norm": 0.611411988735199, "learning_rate": 9.258e-05, "loss": 0.0809, "step": 4630 }, { "grad_norm": 0.4795951545238495, "learning_rate": 9.278e-05, "loss": 0.0749, "step": 4640 }, { "grad_norm": 0.5294396877288818, "learning_rate": 9.298e-05, "loss": 0.0755, "step": 4650 }, { "grad_norm": 0.5398479104042053, "learning_rate": 9.318e-05, "loss": 0.0701, "step": 4660 }, { "grad_norm": 0.49637946486473083, "learning_rate": 9.338e-05, "loss": 0.0743, "step": 4670 }, { "grad_norm": 0.5349768996238708, "learning_rate": 9.358e-05, "loss": 0.0734, "step": 4680 }, { "grad_norm": 0.6235644221305847, "learning_rate": 9.378e-05, "loss": 0.0776, "step": 4690 }, { "grad_norm": 0.7026760578155518, "learning_rate": 9.398e-05, "loss": 0.0734, "step": 4700 }, { "grad_norm": 0.474298357963562, "learning_rate": 9.418e-05, "loss": 0.0756, "step": 4710 }, { "grad_norm": 0.5725098848342896, "learning_rate": 9.438e-05, "loss": 0.0819, "step": 4720 }, { "grad_norm": 0.6220020651817322, "learning_rate": 9.458e-05, "loss": 0.0755, "step": 4730 }, { "grad_norm": 0.6721327900886536, "learning_rate": 9.478e-05, "loss": 0.073, "step": 4740 }, { "grad_norm": 0.694243311882019, "learning_rate": 9.498e-05, "loss": 0.063, "step": 4750 }, { "grad_norm": 0.6536597013473511, "learning_rate": 9.518000000000001e-05, "loss": 0.0748, "step": 4760 }, { "grad_norm": 0.6888153553009033, "learning_rate": 9.538e-05, "loss": 0.0734, "step": 4770 }, { "grad_norm": 0.6048505306243896, "learning_rate": 9.558e-05, "loss": 0.0785, "step": 4780 }, { "grad_norm": 0.6198250651359558, "learning_rate": 9.578000000000001e-05, "loss": 0.0749, "step": 4790 }, { "grad_norm": 0.6689453721046448, "learning_rate": 9.598e-05, "loss": 0.0711, "step": 4800 }, { "grad_norm": 0.5142550468444824, "learning_rate": 9.618e-05, "loss": 0.0673, "step": 4810 }, { "grad_norm": 0.5017518401145935, "learning_rate": 9.638000000000001e-05, "loss": 0.0654, "step": 4820 }, { "grad_norm": 0.5227983593940735, "learning_rate": 9.658e-05, "loss": 0.0629, "step": 4830 }, { "grad_norm": 0.603459894657135, "learning_rate": 9.678e-05, "loss": 0.0664, "step": 4840 }, { "grad_norm": 0.5729115009307861, "learning_rate": 9.698000000000001e-05, "loss": 0.0707, "step": 4850 }, { "grad_norm": 0.5818468332290649, "learning_rate": 9.718e-05, "loss": 0.0715, "step": 4860 }, { "grad_norm": 0.48878568410873413, "learning_rate": 9.738e-05, "loss": 0.0662, "step": 4870 }, { "grad_norm": 0.4382123351097107, "learning_rate": 9.758000000000001e-05, "loss": 0.0705, "step": 4880 }, { "grad_norm": 0.4278465807437897, "learning_rate": 9.778e-05, "loss": 0.0679, "step": 4890 }, { "grad_norm": 0.4833517372608185, "learning_rate": 9.798000000000001e-05, "loss": 0.0666, "step": 4900 }, { "grad_norm": 0.4499589204788208, "learning_rate": 9.818000000000001e-05, "loss": 0.0722, "step": 4910 }, { "grad_norm": 0.4592512249946594, "learning_rate": 9.838e-05, "loss": 0.0695, "step": 4920 }, { "grad_norm": 0.5048083662986755, "learning_rate": 9.858000000000001e-05, "loss": 0.0735, "step": 4930 }, { "grad_norm": 0.3962958753108978, "learning_rate": 9.878e-05, "loss": 0.0659, "step": 4940 }, { "grad_norm": 0.5285344123840332, "learning_rate": 9.898e-05, "loss": 0.062, "step": 4950 }, { "grad_norm": 0.5859363079071045, "learning_rate": 9.918000000000001e-05, "loss": 0.0696, "step": 4960 }, { "grad_norm": 0.599980354309082, "learning_rate": 9.938e-05, "loss": 0.0637, "step": 4970 }, { "grad_norm": 0.5149223208427429, "learning_rate": 9.958e-05, "loss": 0.0618, "step": 4980 }, { "grad_norm": 0.602069079875946, "learning_rate": 9.978000000000001e-05, "loss": 0.0718, "step": 4990 }, { "grad_norm": 0.510648250579834, "learning_rate": 9.998e-05, "loss": 0.0678, "step": 5000 }, { "grad_norm": 0.5860572457313538, "learning_rate": 9.999999778549045e-05, "loss": 0.0718, "step": 5010 }, { "grad_norm": 0.524765133857727, "learning_rate": 9.999999013039593e-05, "loss": 0.0716, "step": 5020 }, { "grad_norm": 0.5203791856765747, "learning_rate": 9.999997700737766e-05, "loss": 0.083, "step": 5030 }, { "grad_norm": 0.5978304147720337, "learning_rate": 9.999995841643709e-05, "loss": 0.0694, "step": 5040 }, { "grad_norm": 0.6102453470230103, "learning_rate": 9.999993435757623e-05, "loss": 0.0741, "step": 5050 }, { "grad_norm": 0.8299005031585693, "learning_rate": 9.999990483079773e-05, "loss": 0.07, "step": 5060 }, { "grad_norm": 0.6680502891540527, "learning_rate": 9.999986983610481e-05, "loss": 0.0746, "step": 5070 }, { "grad_norm": 0.6042954325675964, "learning_rate": 9.99998293735013e-05, "loss": 0.0838, "step": 5080 }, { "grad_norm": 0.5273632407188416, "learning_rate": 9.999978344299161e-05, "loss": 0.0722, "step": 5090 }, { "grad_norm": 0.6281518340110779, "learning_rate": 9.99997320445808e-05, "loss": 0.0749, "step": 5100 }, { "grad_norm": 0.5066910982131958, "learning_rate": 9.999967517827444e-05, "loss": 0.0709, "step": 5110 }, { "grad_norm": 0.5761662125587463, "learning_rate": 9.999961284407879e-05, "loss": 0.0705, "step": 5120 }, { "grad_norm": 0.5905638933181763, "learning_rate": 9.999954504200067e-05, "loss": 0.0737, "step": 5130 }, { "grad_norm": 0.47844356298446655, "learning_rate": 9.999947177204744e-05, "loss": 0.0659, "step": 5140 }, { "grad_norm": 0.5486882925033569, "learning_rate": 9.999939303422718e-05, "loss": 0.0779, "step": 5150 }, { "grad_norm": 0.4975206255912781, "learning_rate": 9.999930882854847e-05, "loss": 0.0705, "step": 5160 }, { "grad_norm": 0.5569563508033752, "learning_rate": 9.999921915502051e-05, "loss": 0.0692, "step": 5170 }, { "grad_norm": 0.4385300278663635, "learning_rate": 9.99991240136531e-05, "loss": 0.0685, "step": 5180 }, { "grad_norm": 0.5247234106063843, "learning_rate": 9.999902340445668e-05, "loss": 0.069, "step": 5190 }, { "grad_norm": 0.6019571423530579, "learning_rate": 9.999891732744224e-05, "loss": 0.065, "step": 5200 }, { "grad_norm": 0.460910439491272, "learning_rate": 9.999880578262135e-05, "loss": 0.0687, "step": 5210 }, { "grad_norm": 0.5081547498703003, "learning_rate": 9.999868877000624e-05, "loss": 0.0715, "step": 5220 }, { "grad_norm": 0.46797043085098267, "learning_rate": 9.99985662896097e-05, "loss": 0.0638, "step": 5230 }, { "grad_norm": 0.5384241342544556, "learning_rate": 9.999843834144513e-05, "loss": 0.0733, "step": 5240 }, { "grad_norm": 0.6345313787460327, "learning_rate": 9.99983049255265e-05, "loss": 0.065, "step": 5250 }, { "grad_norm": 0.6413396596908569, "learning_rate": 9.999816604186843e-05, "loss": 0.0712, "step": 5260 }, { "grad_norm": 0.5505473017692566, "learning_rate": 9.999802169048609e-05, "loss": 0.0634, "step": 5270 }, { "grad_norm": 0.46517297625541687, "learning_rate": 9.999787187139527e-05, "loss": 0.0718, "step": 5280 }, { "grad_norm": 0.4482157528400421, "learning_rate": 9.999771658461234e-05, "loss": 0.0628, "step": 5290 }, { "grad_norm": 0.47481346130371094, "learning_rate": 9.999755583015431e-05, "loss": 0.0611, "step": 5300 }, { "grad_norm": 0.514595627784729, "learning_rate": 9.999738960803874e-05, "loss": 0.066, "step": 5310 }, { "grad_norm": 0.6671262383460999, "learning_rate": 9.99972179182838e-05, "loss": 0.0787, "step": 5320 }, { "grad_norm": 0.6182829141616821, "learning_rate": 9.99970407609083e-05, "loss": 0.0714, "step": 5330 }, { "grad_norm": 0.4871818423271179, "learning_rate": 9.999685813593159e-05, "loss": 0.0718, "step": 5340 }, { "grad_norm": 0.5860035419464111, "learning_rate": 9.999667004337362e-05, "loss": 0.0684, "step": 5350 }, { "grad_norm": 0.628444254398346, "learning_rate": 9.9996476483255e-05, "loss": 0.0734, "step": 5360 }, { "grad_norm": 0.6326653361320496, "learning_rate": 9.999627745559688e-05, "loss": 0.0851, "step": 5370 }, { "grad_norm": 0.5769596695899963, "learning_rate": 9.999607296042101e-05, "loss": 0.0746, "step": 5380 }, { "grad_norm": 0.6262946128845215, "learning_rate": 9.99958629977498e-05, "loss": 0.075, "step": 5390 }, { "grad_norm": 0.5318493843078613, "learning_rate": 9.999564756760615e-05, "loss": 0.0776, "step": 5400 }, { "grad_norm": 0.4935837984085083, "learning_rate": 9.999542667001366e-05, "loss": 0.0698, "step": 5410 }, { "grad_norm": 0.5301372408866882, "learning_rate": 9.999520030499647e-05, "loss": 0.0652, "step": 5420 }, { "grad_norm": 0.53269362449646, "learning_rate": 9.999496847257936e-05, "loss": 0.0677, "step": 5430 }, { "grad_norm": 0.5064731240272522, "learning_rate": 9.999473117278764e-05, "loss": 0.065, "step": 5440 }, { "grad_norm": 0.4838753044605255, "learning_rate": 9.999448840564731e-05, "loss": 0.0679, "step": 5450 }, { "grad_norm": 0.4321793019771576, "learning_rate": 9.999424017118488e-05, "loss": 0.0705, "step": 5460 }, { "grad_norm": 0.49262651801109314, "learning_rate": 9.999398646942751e-05, "loss": 0.068, "step": 5470 }, { "grad_norm": 0.48772886395454407, "learning_rate": 9.999372730040296e-05, "loss": 0.065, "step": 5480 }, { "grad_norm": 0.5892905592918396, "learning_rate": 9.999346266413953e-05, "loss": 0.0652, "step": 5490 }, { "grad_norm": 0.45037826895713806, "learning_rate": 9.99931925606662e-05, "loss": 0.0652, "step": 5500 }, { "grad_norm": 0.508467435836792, "learning_rate": 9.99929169900125e-05, "loss": 0.0632, "step": 5510 }, { "grad_norm": 0.48006731271743774, "learning_rate": 9.999263595220855e-05, "loss": 0.068, "step": 5520 }, { "grad_norm": 0.465902179479599, "learning_rate": 9.99923494472851e-05, "loss": 0.0768, "step": 5530 }, { "grad_norm": 0.35091322660446167, "learning_rate": 9.999205747527348e-05, "loss": 0.0697, "step": 5540 }, { "grad_norm": 0.49681204557418823, "learning_rate": 9.999176003620561e-05, "loss": 0.0612, "step": 5550 }, { "grad_norm": 0.5050622820854187, "learning_rate": 9.999145713011405e-05, "loss": 0.0743, "step": 5560 }, { "grad_norm": 0.5260747671127319, "learning_rate": 9.999114875703186e-05, "loss": 0.0652, "step": 5570 }, { "grad_norm": 0.6310049891471863, "learning_rate": 9.999083491699281e-05, "loss": 0.068, "step": 5580 }, { "grad_norm": 0.6106846332550049, "learning_rate": 9.999051561003123e-05, "loss": 0.0757, "step": 5590 }, { "grad_norm": 0.5187007188796997, "learning_rate": 9.999019083618202e-05, "loss": 0.0704, "step": 5600 }, { "grad_norm": 0.4837956130504608, "learning_rate": 9.99898605954807e-05, "loss": 0.0704, "step": 5610 }, { "grad_norm": 0.5037572383880615, "learning_rate": 9.998952488796338e-05, "loss": 0.0662, "step": 5620 }, { "grad_norm": 0.49410200119018555, "learning_rate": 9.998918371366676e-05, "loss": 0.0625, "step": 5630 }, { "grad_norm": 0.4746967554092407, "learning_rate": 9.99888370726282e-05, "loss": 0.0632, "step": 5640 }, { "grad_norm": 0.6312411427497864, "learning_rate": 9.998848496488556e-05, "loss": 0.0635, "step": 5650 }, { "grad_norm": 0.5988750457763672, "learning_rate": 9.998812739047736e-05, "loss": 0.069, "step": 5660 }, { "grad_norm": 0.5640274286270142, "learning_rate": 9.99877643494427e-05, "loss": 0.0638, "step": 5670 }, { "grad_norm": 0.5431151390075684, "learning_rate": 9.998739584182128e-05, "loss": 0.0704, "step": 5680 }, { "grad_norm": 0.4126555919647217, "learning_rate": 9.998702186765342e-05, "loss": 0.071, "step": 5690 }, { "grad_norm": 0.4403904378414154, "learning_rate": 9.998664242698e-05, "loss": 0.0623, "step": 5700 }, { "grad_norm": 0.6179052591323853, "learning_rate": 9.998625751984251e-05, "loss": 0.0615, "step": 5710 }, { "grad_norm": 0.4275966286659241, "learning_rate": 9.998586714628307e-05, "loss": 0.0624, "step": 5720 }, { "grad_norm": 0.6683180332183838, "learning_rate": 9.998547130634432e-05, "loss": 0.0682, "step": 5730 }, { "grad_norm": 0.508109986782074, "learning_rate": 9.99850700000696e-05, "loss": 0.0728, "step": 5740 }, { "grad_norm": 0.44037923216819763, "learning_rate": 9.998466322750278e-05, "loss": 0.066, "step": 5750 }, { "grad_norm": 0.49787580966949463, "learning_rate": 9.998425098868834e-05, "loss": 0.0716, "step": 5760 }, { "grad_norm": 0.441249281167984, "learning_rate": 9.998383328367136e-05, "loss": 0.0775, "step": 5770 }, { "grad_norm": 0.6406336426734924, "learning_rate": 9.99834101124975e-05, "loss": 0.073, "step": 5780 }, { "grad_norm": 0.5812379121780396, "learning_rate": 9.998298147521309e-05, "loss": 0.065, "step": 5790 }, { "grad_norm": 0.3768431842327118, "learning_rate": 9.998254737186496e-05, "loss": 0.063, "step": 5800 }, { "grad_norm": 0.5023033618927002, "learning_rate": 9.99821078025006e-05, "loss": 0.0702, "step": 5810 }, { "grad_norm": 0.6396487355232239, "learning_rate": 9.998166276716807e-05, "loss": 0.0765, "step": 5820 }, { "grad_norm": 0.500536322593689, "learning_rate": 9.998121226591606e-05, "loss": 0.0675, "step": 5830 }, { "grad_norm": 0.49315062165260315, "learning_rate": 9.998075629879382e-05, "loss": 0.0667, "step": 5840 }, { "grad_norm": 0.5549322962760925, "learning_rate": 9.99802948658512e-05, "loss": 0.0791, "step": 5850 }, { "grad_norm": 0.48272964358329773, "learning_rate": 9.99798279671387e-05, "loss": 0.0595, "step": 5860 }, { "grad_norm": 0.48752081394195557, "learning_rate": 9.997935560270734e-05, "loss": 0.0628, "step": 5870 }, { "grad_norm": 0.6614753603935242, "learning_rate": 9.997887777260879e-05, "loss": 0.064, "step": 5880 }, { "grad_norm": 0.4510793685913086, "learning_rate": 9.997839447689532e-05, "loss": 0.0691, "step": 5890 }, { "grad_norm": 0.4670790731906891, "learning_rate": 9.997790571561978e-05, "loss": 0.0698, "step": 5900 }, { "grad_norm": 0.4942845404148102, "learning_rate": 9.99774114888356e-05, "loss": 0.0597, "step": 5910 }, { "grad_norm": 0.46544042229652405, "learning_rate": 9.997691179659684e-05, "loss": 0.0657, "step": 5920 }, { "grad_norm": 0.5793338418006897, "learning_rate": 9.997640663895815e-05, "loss": 0.0612, "step": 5930 }, { "grad_norm": 0.39671584963798523, "learning_rate": 9.997589601597477e-05, "loss": 0.0704, "step": 5940 }, { "grad_norm": 0.5684154033660889, "learning_rate": 9.997537992770252e-05, "loss": 0.0732, "step": 5950 }, { "grad_norm": 0.5000068545341492, "learning_rate": 9.997485837419788e-05, "loss": 0.0673, "step": 5960 }, { "grad_norm": 0.40565431118011475, "learning_rate": 9.997433135551786e-05, "loss": 0.0648, "step": 5970 }, { "grad_norm": 0.4198535084724426, "learning_rate": 9.997379887172009e-05, "loss": 0.0671, "step": 5980 }, { "grad_norm": 0.445829838514328, "learning_rate": 9.997326092286281e-05, "loss": 0.0609, "step": 5990 }, { "grad_norm": 0.4819698631763458, "learning_rate": 9.997271750900486e-05, "loss": 0.06, "step": 6000 }, { "grad_norm": 0.5047538876533508, "learning_rate": 9.997216863020565e-05, "loss": 0.0693, "step": 6010 }, { "grad_norm": 0.5197441577911377, "learning_rate": 9.99716142865252e-05, "loss": 0.0644, "step": 6020 }, { "grad_norm": 0.46894675493240356, "learning_rate": 9.997105447802415e-05, "loss": 0.0661, "step": 6030 }, { "grad_norm": 0.5731987357139587, "learning_rate": 9.997048920476373e-05, "loss": 0.0634, "step": 6040 }, { "grad_norm": 0.4247092604637146, "learning_rate": 9.996991846680572e-05, "loss": 0.0589, "step": 6050 }, { "grad_norm": 0.4612569510936737, "learning_rate": 9.996934226421257e-05, "loss": 0.0579, "step": 6060 }, { "grad_norm": 0.5551966428756714, "learning_rate": 9.996876059704726e-05, "loss": 0.0604, "step": 6070 }, { "grad_norm": 0.41662323474884033, "learning_rate": 9.996817346537343e-05, "loss": 0.0625, "step": 6080 }, { "grad_norm": 0.5128140449523926, "learning_rate": 9.996758086925526e-05, "loss": 0.0659, "step": 6090 }, { "grad_norm": 0.458993524312973, "learning_rate": 9.996698280875759e-05, "loss": 0.0656, "step": 6100 }, { "grad_norm": 0.42190057039260864, "learning_rate": 9.99663792839458e-05, "loss": 0.0613, "step": 6110 }, { "grad_norm": 0.5187911987304688, "learning_rate": 9.99657702948859e-05, "loss": 0.0678, "step": 6120 }, { "grad_norm": 0.2993547022342682, "learning_rate": 9.996515584164448e-05, "loss": 0.0638, "step": 6130 }, { "grad_norm": 0.3701499402523041, "learning_rate": 9.996453592428873e-05, "loss": 0.0577, "step": 6140 }, { "grad_norm": 0.526172935962677, "learning_rate": 9.996391054288646e-05, "loss": 0.0586, "step": 6150 }, { "grad_norm": 0.5314487814903259, "learning_rate": 9.996327969750605e-05, "loss": 0.0619, "step": 6160 }, { "grad_norm": 0.4963778853416443, "learning_rate": 9.996264338821649e-05, "loss": 0.0573, "step": 6170 }, { "grad_norm": 0.45911821722984314, "learning_rate": 9.996200161508735e-05, "loss": 0.0627, "step": 6180 }, { "grad_norm": 0.5022353529930115, "learning_rate": 9.996135437818885e-05, "loss": 0.0579, "step": 6190 }, { "grad_norm": 0.5261334776878357, "learning_rate": 9.996070167759175e-05, "loss": 0.0583, "step": 6200 }, { "grad_norm": 0.40000396966934204, "learning_rate": 9.996004351336743e-05, "loss": 0.0585, "step": 6210 }, { "grad_norm": 0.46920621395111084, "learning_rate": 9.995937988558785e-05, "loss": 0.0607, "step": 6220 }, { "grad_norm": 0.5009639263153076, "learning_rate": 9.995871079432561e-05, "loss": 0.0628, "step": 6230 }, { "grad_norm": 0.5822951197624207, "learning_rate": 9.995803623965389e-05, "loss": 0.0604, "step": 6240 }, { "grad_norm": 0.438414990901947, "learning_rate": 9.995735622164641e-05, "loss": 0.057, "step": 6250 }, { "grad_norm": 0.4004955589771271, "learning_rate": 9.995667074037758e-05, "loss": 0.0562, "step": 6260 }, { "grad_norm": 0.5307245850563049, "learning_rate": 9.995597979592232e-05, "loss": 0.0608, "step": 6270 }, { "grad_norm": 0.40881961584091187, "learning_rate": 9.995528338835625e-05, "loss": 0.058, "step": 6280 }, { "grad_norm": 0.4004250168800354, "learning_rate": 9.995458151775547e-05, "loss": 0.0593, "step": 6290 }, { "grad_norm": 0.414103627204895, "learning_rate": 9.995387418419677e-05, "loss": 0.0641, "step": 6300 }, { "grad_norm": 0.4431852102279663, "learning_rate": 9.99531613877575e-05, "loss": 0.0653, "step": 6310 }, { "grad_norm": 0.5756431818008423, "learning_rate": 9.995244312851559e-05, "loss": 0.0708, "step": 6320 }, { "grad_norm": 0.5382875204086304, "learning_rate": 9.995171940654961e-05, "loss": 0.07, "step": 6330 }, { "grad_norm": 0.3961900770664215, "learning_rate": 9.995099022193871e-05, "loss": 0.0584, "step": 6340 }, { "grad_norm": 0.46450814604759216, "learning_rate": 9.995025557476261e-05, "loss": 0.0628, "step": 6350 }, { "grad_norm": 0.4483158588409424, "learning_rate": 9.994951546510165e-05, "loss": 0.0617, "step": 6360 }, { "grad_norm": 0.3938438594341278, "learning_rate": 9.994876989303679e-05, "loss": 0.0601, "step": 6370 }, { "grad_norm": 0.45788347721099854, "learning_rate": 9.994801885864955e-05, "loss": 0.0592, "step": 6380 }, { "grad_norm": 0.491054505109787, "learning_rate": 9.994726236202205e-05, "loss": 0.0653, "step": 6390 }, { "grad_norm": 0.5241134166717529, "learning_rate": 9.994650040323704e-05, "loss": 0.0612, "step": 6400 }, { "grad_norm": 0.5273036360740662, "learning_rate": 9.994573298237784e-05, "loss": 0.0689, "step": 6410 }, { "grad_norm": 0.6145632863044739, "learning_rate": 9.994496009952837e-05, "loss": 0.0658, "step": 6420 }, { "grad_norm": 0.514326810836792, "learning_rate": 9.994418175477316e-05, "loss": 0.069, "step": 6430 }, { "grad_norm": 0.47742846608161926, "learning_rate": 9.994339794819733e-05, "loss": 0.0615, "step": 6440 }, { "grad_norm": 0.5343334078788757, "learning_rate": 9.994260867988658e-05, "loss": 0.0625, "step": 6450 }, { "grad_norm": 0.529356837272644, "learning_rate": 9.994181394992723e-05, "loss": 0.0586, "step": 6460 }, { "grad_norm": 0.5890268087387085, "learning_rate": 9.994101375840618e-05, "loss": 0.0614, "step": 6470 }, { "grad_norm": 0.4472443163394928, "learning_rate": 9.994020810541098e-05, "loss": 0.068, "step": 6480 }, { "grad_norm": 0.43322840332984924, "learning_rate": 9.99393969910297e-05, "loss": 0.0703, "step": 6490 }, { "grad_norm": 0.4363170266151428, "learning_rate": 9.993858041535104e-05, "loss": 0.0674, "step": 6500 }, { "grad_norm": 0.4056730270385742, "learning_rate": 9.99377583784643e-05, "loss": 0.0638, "step": 6510 }, { "grad_norm": 0.46146607398986816, "learning_rate": 9.993693088045939e-05, "loss": 0.0614, "step": 6520 }, { "grad_norm": 0.45036542415618896, "learning_rate": 9.99360979214268e-05, "loss": 0.0639, "step": 6530 }, { "grad_norm": 0.4038468897342682, "learning_rate": 9.99352595014576e-05, "loss": 0.0611, "step": 6540 }, { "grad_norm": 0.4974776804447174, "learning_rate": 9.993441562064354e-05, "loss": 0.0749, "step": 6550 }, { "grad_norm": 0.5477267503738403, "learning_rate": 9.993356627907685e-05, "loss": 0.0584, "step": 6560 }, { "grad_norm": 0.5117640495300293, "learning_rate": 9.99327114768504e-05, "loss": 0.0597, "step": 6570 }, { "grad_norm": 0.38423141837120056, "learning_rate": 9.99318512140577e-05, "loss": 0.0574, "step": 6580 }, { "grad_norm": 0.4918487071990967, "learning_rate": 9.993098549079284e-05, "loss": 0.0621, "step": 6590 }, { "grad_norm": 0.3873863220214844, "learning_rate": 9.993011430715047e-05, "loss": 0.0583, "step": 6600 }, { "grad_norm": 0.5004062652587891, "learning_rate": 9.992923766322586e-05, "loss": 0.0596, "step": 6610 }, { "grad_norm": 0.5009154081344604, "learning_rate": 9.99283555591149e-05, "loss": 0.0684, "step": 6620 }, { "grad_norm": 0.35232681035995483, "learning_rate": 9.992746799491404e-05, "loss": 0.0583, "step": 6630 }, { "grad_norm": 0.40899497270584106, "learning_rate": 9.992657497072033e-05, "loss": 0.0551, "step": 6640 }, { "grad_norm": 0.4208568036556244, "learning_rate": 9.992567648663147e-05, "loss": 0.0549, "step": 6650 }, { "grad_norm": 0.6590387225151062, "learning_rate": 9.992477254274568e-05, "loss": 0.0667, "step": 6660 }, { "grad_norm": 0.42645779252052307, "learning_rate": 9.992386313916183e-05, "loss": 0.0618, "step": 6670 }, { "grad_norm": 0.4932340681552887, "learning_rate": 9.992294827597934e-05, "loss": 0.0656, "step": 6680 }, { "grad_norm": 0.44504430890083313, "learning_rate": 9.992202795329831e-05, "loss": 0.0662, "step": 6690 }, { "grad_norm": 0.3466145694255829, "learning_rate": 9.992110217121936e-05, "loss": 0.0545, "step": 6700 }, { "grad_norm": 0.4011939465999603, "learning_rate": 9.992017092984372e-05, "loss": 0.0539, "step": 6710 }, { "grad_norm": 0.45300012826919556, "learning_rate": 9.991923422927326e-05, "loss": 0.0661, "step": 6720 }, { "grad_norm": 0.43515732884407043, "learning_rate": 9.991829206961037e-05, "loss": 0.0558, "step": 6730 }, { "grad_norm": 0.5151969194412231, "learning_rate": 9.991734445095813e-05, "loss": 0.0575, "step": 6740 }, { "grad_norm": 0.4544309377670288, "learning_rate": 9.991639137342015e-05, "loss": 0.0595, "step": 6750 }, { "grad_norm": 0.4487329125404358, "learning_rate": 9.991543283710064e-05, "loss": 0.0653, "step": 6760 }, { "grad_norm": 0.43640148639678955, "learning_rate": 9.991446884210445e-05, "loss": 0.0589, "step": 6770 }, { "grad_norm": 0.4272100329399109, "learning_rate": 9.9913499388537e-05, "loss": 0.0567, "step": 6780 }, { "grad_norm": 0.4285908341407776, "learning_rate": 9.99125244765043e-05, "loss": 0.0552, "step": 6790 }, { "grad_norm": 0.5024617314338684, "learning_rate": 9.991154410611296e-05, "loss": 0.0549, "step": 6800 }, { "grad_norm": 0.6223841309547424, "learning_rate": 9.99105582774702e-05, "loss": 0.0578, "step": 6810 }, { "grad_norm": 0.44652965664863586, "learning_rate": 9.990956699068384e-05, "loss": 0.0547, "step": 6820 }, { "grad_norm": 0.42879077792167664, "learning_rate": 9.990857024586224e-05, "loss": 0.0552, "step": 6830 }, { "grad_norm": 0.40456047654151917, "learning_rate": 9.990756804311446e-05, "loss": 0.0524, "step": 6840 }, { "grad_norm": 0.45124897360801697, "learning_rate": 9.990656038255006e-05, "loss": 0.059, "step": 6850 }, { "grad_norm": 0.46750152111053467, "learning_rate": 9.990554726427926e-05, "loss": 0.0529, "step": 6860 }, { "grad_norm": 0.41565394401550293, "learning_rate": 9.990452868841284e-05, "loss": 0.0555, "step": 6870 }, { "grad_norm": 0.44582152366638184, "learning_rate": 9.99035046550622e-05, "loss": 0.0585, "step": 6880 }, { "grad_norm": 0.44850805401802063, "learning_rate": 9.99024751643393e-05, "loss": 0.0541, "step": 6890 }, { "grad_norm": 0.3938608467578888, "learning_rate": 9.990144021635677e-05, "loss": 0.0587, "step": 6900 }, { "grad_norm": 0.4664738178253174, "learning_rate": 9.990039981122775e-05, "loss": 0.0567, "step": 6910 }, { "grad_norm": 0.5158368945121765, "learning_rate": 9.989935394906602e-05, "loss": 0.0576, "step": 6920 }, { "grad_norm": 0.34577476978302, "learning_rate": 9.989830262998598e-05, "loss": 0.0576, "step": 6930 }, { "grad_norm": 0.48265963792800903, "learning_rate": 9.989724585410259e-05, "loss": 0.0618, "step": 6940 }, { "grad_norm": 0.45193910598754883, "learning_rate": 9.989618362153139e-05, "loss": 0.0633, "step": 6950 }, { "grad_norm": 0.36106088757514954, "learning_rate": 9.989511593238859e-05, "loss": 0.0603, "step": 6960 }, { "grad_norm": 0.3496214747428894, "learning_rate": 9.98940427867909e-05, "loss": 0.0481, "step": 6970 }, { "grad_norm": 0.3700890839099884, "learning_rate": 9.989296418485573e-05, "loss": 0.0561, "step": 6980 }, { "grad_norm": 0.3349691927433014, "learning_rate": 9.989188012670101e-05, "loss": 0.0552, "step": 6990 }, { "grad_norm": 0.395805686712265, "learning_rate": 9.989079061244528e-05, "loss": 0.0542, "step": 7000 }, { "grad_norm": 0.44368976354599, "learning_rate": 9.988969564220769e-05, "loss": 0.0525, "step": 7010 }, { "grad_norm": 0.40282928943634033, "learning_rate": 9.988859521610801e-05, "loss": 0.0491, "step": 7020 }, { "grad_norm": 0.47525152564048767, "learning_rate": 9.988748933426656e-05, "loss": 0.0522, "step": 7030 }, { "grad_norm": 0.36940616369247437, "learning_rate": 9.988637799680428e-05, "loss": 0.0585, "step": 7040 }, { "grad_norm": 0.5223163366317749, "learning_rate": 9.98852612038427e-05, "loss": 0.0551, "step": 7050 }, { "grad_norm": 0.39818453788757324, "learning_rate": 9.988413895550397e-05, "loss": 0.0592, "step": 7060 }, { "grad_norm": 0.36971399188041687, "learning_rate": 9.98830112519108e-05, "loss": 0.0611, "step": 7070 }, { "grad_norm": 0.47246482968330383, "learning_rate": 9.98818780931865e-05, "loss": 0.0611, "step": 7080 }, { "grad_norm": 0.4157969653606415, "learning_rate": 9.988073947945502e-05, "loss": 0.0525, "step": 7090 }, { "grad_norm": 0.4261585474014282, "learning_rate": 9.987959541084087e-05, "loss": 0.0564, "step": 7100 }, { "grad_norm": 0.4482017457485199, "learning_rate": 9.987844588746915e-05, "loss": 0.0597, "step": 7110 }, { "grad_norm": 0.4520516097545624, "learning_rate": 9.987729090946558e-05, "loss": 0.0591, "step": 7120 }, { "grad_norm": 0.4297524690628052, "learning_rate": 9.987613047695647e-05, "loss": 0.0579, "step": 7130 }, { "grad_norm": 0.39670997858047485, "learning_rate": 9.987496459006871e-05, "loss": 0.057, "step": 7140 }, { "grad_norm": 0.4551587700843811, "learning_rate": 9.987379324892982e-05, "loss": 0.0543, "step": 7150 }, { "grad_norm": 0.4484739601612091, "learning_rate": 9.987261645366788e-05, "loss": 0.0564, "step": 7160 }, { "grad_norm": 0.5580835342407227, "learning_rate": 9.987143420441158e-05, "loss": 0.0604, "step": 7170 }, { "grad_norm": 0.6088237166404724, "learning_rate": 9.987024650129022e-05, "loss": 0.0636, "step": 7180 }, { "grad_norm": 0.32037779688835144, "learning_rate": 9.986905334443368e-05, "loss": 0.0526, "step": 7190 }, { "grad_norm": 0.4058995842933655, "learning_rate": 9.986785473397245e-05, "loss": 0.0542, "step": 7200 }, { "grad_norm": 0.36504533886909485, "learning_rate": 9.98666506700376e-05, "loss": 0.0539, "step": 7210 }, { "grad_norm": 0.3850560486316681, "learning_rate": 9.986544115276081e-05, "loss": 0.0515, "step": 7220 }, { "grad_norm": 0.34163907170295715, "learning_rate": 9.986422618227433e-05, "loss": 0.0553, "step": 7230 }, { "grad_norm": 0.5090659260749817, "learning_rate": 9.986300575871106e-05, "loss": 0.0575, "step": 7240 }, { "grad_norm": 0.4446185827255249, "learning_rate": 9.986177988220444e-05, "loss": 0.0616, "step": 7250 }, { "grad_norm": 0.4146069288253784, "learning_rate": 9.986054855288856e-05, "loss": 0.0582, "step": 7260 }, { "grad_norm": 0.3478843569755554, "learning_rate": 9.985931177089802e-05, "loss": 0.0629, "step": 7270 }, { "grad_norm": 0.5521957874298096, "learning_rate": 9.985806953636814e-05, "loss": 0.0592, "step": 7280 }, { "grad_norm": 0.5520628690719604, "learning_rate": 9.985682184943471e-05, "loss": 0.0643, "step": 7290 }, { "grad_norm": 0.4737977683544159, "learning_rate": 9.98555687102342e-05, "loss": 0.0659, "step": 7300 }, { "grad_norm": 0.501708447933197, "learning_rate": 9.985431011890367e-05, "loss": 0.0682, "step": 7310 }, { "grad_norm": 0.38110214471817017, "learning_rate": 9.985304607558075e-05, "loss": 0.057, "step": 7320 }, { "grad_norm": 0.4827958345413208, "learning_rate": 9.985177658040364e-05, "loss": 0.0637, "step": 7330 }, { "grad_norm": 0.5733841061592102, "learning_rate": 9.985050163351119e-05, "loss": 0.0581, "step": 7340 }, { "grad_norm": 0.3847861886024475, "learning_rate": 9.984922123504286e-05, "loss": 0.0566, "step": 7350 }, { "grad_norm": 0.32923540472984314, "learning_rate": 9.984793538513862e-05, "loss": 0.0598, "step": 7360 }, { "grad_norm": 0.4439884424209595, "learning_rate": 9.984664408393912e-05, "loss": 0.0534, "step": 7370 }, { "grad_norm": 0.3985814154148102, "learning_rate": 9.984534733158556e-05, "loss": 0.0525, "step": 7380 }, { "grad_norm": 0.540309727191925, "learning_rate": 9.984404512821977e-05, "loss": 0.0537, "step": 7390 }, { "grad_norm": 0.4473859667778015, "learning_rate": 9.984273747398411e-05, "loss": 0.0576, "step": 7400 }, { "grad_norm": 0.48094648122787476, "learning_rate": 9.984142436902165e-05, "loss": 0.0579, "step": 7410 }, { "grad_norm": 0.40711894631385803, "learning_rate": 9.984010581347596e-05, "loss": 0.0566, "step": 7420 }, { "grad_norm": 0.4863482117652893, "learning_rate": 9.983878180749121e-05, "loss": 0.0532, "step": 7430 }, { "grad_norm": 0.4571806490421295, "learning_rate": 9.983745235121222e-05, "loss": 0.0577, "step": 7440 }, { "grad_norm": 0.4465266764163971, "learning_rate": 9.983611744478438e-05, "loss": 0.0556, "step": 7450 }, { "grad_norm": 0.4169601798057556, "learning_rate": 9.983477708835365e-05, "loss": 0.0608, "step": 7460 }, { "grad_norm": 0.36112672090530396, "learning_rate": 9.983343128206664e-05, "loss": 0.0532, "step": 7470 }, { "grad_norm": 0.3469308614730835, "learning_rate": 9.983208002607049e-05, "loss": 0.0497, "step": 7480 }, { "grad_norm": 0.4387678802013397, "learning_rate": 9.9830723320513e-05, "loss": 0.0543, "step": 7490 }, { "grad_norm": 0.4034811854362488, "learning_rate": 9.982936116554254e-05, "loss": 0.056, "step": 7500 }, { "grad_norm": 0.3597908914089203, "learning_rate": 9.982799356130803e-05, "loss": 0.0536, "step": 7510 }, { "grad_norm": 0.5198501348495483, "learning_rate": 9.982662050795908e-05, "loss": 0.0639, "step": 7520 }, { "grad_norm": 0.32107365131378174, "learning_rate": 9.982524200564583e-05, "loss": 0.0558, "step": 7530 }, { "grad_norm": 0.35286417603492737, "learning_rate": 9.982385805451901e-05, "loss": 0.052, "step": 7540 }, { "grad_norm": 0.35251814126968384, "learning_rate": 9.982246865472998e-05, "loss": 0.0473, "step": 7550 }, { "grad_norm": 0.37237149477005005, "learning_rate": 9.982107380643069e-05, "loss": 0.0544, "step": 7560 }, { "grad_norm": 0.3994311988353729, "learning_rate": 9.981967350977368e-05, "loss": 0.0467, "step": 7570 }, { "grad_norm": 0.4019874036312103, "learning_rate": 9.981826776491208e-05, "loss": 0.0565, "step": 7580 }, { "grad_norm": 0.3738962709903717, "learning_rate": 9.98168565719996e-05, "loss": 0.0532, "step": 7590 }, { "grad_norm": 0.4644791781902313, "learning_rate": 9.98154399311906e-05, "loss": 0.0491, "step": 7600 }, { "grad_norm": 0.3969637155532837, "learning_rate": 9.981401784263997e-05, "loss": 0.0565, "step": 7610 }, { "grad_norm": 0.3948717713356018, "learning_rate": 9.981259030650326e-05, "loss": 0.0553, "step": 7620 }, { "grad_norm": 0.4266799986362457, "learning_rate": 9.981115732293655e-05, "loss": 0.0631, "step": 7630 }, { "grad_norm": 0.45850253105163574, "learning_rate": 9.980971889209659e-05, "loss": 0.0545, "step": 7640 }, { "grad_norm": 0.40550366044044495, "learning_rate": 9.980827501414064e-05, "loss": 0.0545, "step": 7650 }, { "grad_norm": 0.4949432611465454, "learning_rate": 9.980682568922663e-05, "loss": 0.0582, "step": 7660 }, { "grad_norm": 0.3795875906944275, "learning_rate": 9.980537091751304e-05, "loss": 0.0545, "step": 7670 }, { "grad_norm": 0.27439114451408386, "learning_rate": 9.980391069915897e-05, "loss": 0.053, "step": 7680 }, { "grad_norm": 0.3975661098957062, "learning_rate": 9.98024450343241e-05, "loss": 0.0541, "step": 7690 }, { "grad_norm": 0.3941151797771454, "learning_rate": 9.980097392316872e-05, "loss": 0.0601, "step": 7700 }, { "grad_norm": 0.36469271779060364, "learning_rate": 9.97994973658537e-05, "loss": 0.0576, "step": 7710 }, { "grad_norm": 0.5172598361968994, "learning_rate": 9.979801536254054e-05, "loss": 0.0579, "step": 7720 }, { "grad_norm": 0.5323194861412048, "learning_rate": 9.979652791339127e-05, "loss": 0.0555, "step": 7730 }, { "grad_norm": 0.3963802754878998, "learning_rate": 9.97950350185686e-05, "loss": 0.0584, "step": 7740 }, { "grad_norm": 0.36184102296829224, "learning_rate": 9.979353667823574e-05, "loss": 0.0569, "step": 7750 }, { "grad_norm": 0.3869134187698364, "learning_rate": 9.979203289255658e-05, "loss": 0.0563, "step": 7760 }, { "grad_norm": 0.41875478625297546, "learning_rate": 9.979052366169557e-05, "loss": 0.062, "step": 7770 }, { "grad_norm": 0.5658137202262878, "learning_rate": 9.978900898581775e-05, "loss": 0.0524, "step": 7780 }, { "grad_norm": 0.44946590065956116, "learning_rate": 9.978748886508875e-05, "loss": 0.051, "step": 7790 }, { "grad_norm": 0.7403585910797119, "learning_rate": 9.978596329967484e-05, "loss": 0.0445, "step": 7800 }, { "grad_norm": 0.40170812606811523, "learning_rate": 9.978443228974284e-05, "loss": 0.0496, "step": 7810 }, { "grad_norm": 0.3842078447341919, "learning_rate": 9.978289583546015e-05, "loss": 0.0609, "step": 7820 }, { "grad_norm": 0.43350911140441895, "learning_rate": 9.978135393699484e-05, "loss": 0.0517, "step": 7830 }, { "grad_norm": 0.5071576237678528, "learning_rate": 9.977980659451548e-05, "loss": 0.0616, "step": 7840 }, { "grad_norm": 0.3819287419319153, "learning_rate": 9.977825380819135e-05, "loss": 0.065, "step": 7850 }, { "grad_norm": 0.4184328019618988, "learning_rate": 9.97766955781922e-05, "loss": 0.0692, "step": 7860 }, { "grad_norm": 0.46140867471694946, "learning_rate": 9.977513190468848e-05, "loss": 0.0642, "step": 7870 }, { "grad_norm": 0.49363961815834045, "learning_rate": 9.977356278785116e-05, "loss": 0.0568, "step": 7880 }, { "grad_norm": 0.44143471121788025, "learning_rate": 9.977198822785184e-05, "loss": 0.0589, "step": 7890 }, { "grad_norm": 0.5159215331077576, "learning_rate": 9.977040822486273e-05, "loss": 0.0549, "step": 7900 }, { "grad_norm": 0.47467613220214844, "learning_rate": 9.97688227790566e-05, "loss": 0.0594, "step": 7910 }, { "grad_norm": 0.3514975905418396, "learning_rate": 9.976723189060684e-05, "loss": 0.0519, "step": 7920 }, { "grad_norm": 0.35424599051475525, "learning_rate": 9.976563555968742e-05, "loss": 0.0481, "step": 7930 }, { "grad_norm": 0.4576451778411865, "learning_rate": 9.976403378647292e-05, "loss": 0.0591, "step": 7940 }, { "grad_norm": 0.36849290132522583, "learning_rate": 9.97624265711385e-05, "loss": 0.0529, "step": 7950 }, { "grad_norm": 0.37096911668777466, "learning_rate": 9.976081391385993e-05, "loss": 0.0535, "step": 7960 }, { "grad_norm": 0.4477829039096832, "learning_rate": 9.975919581481356e-05, "loss": 0.0574, "step": 7970 }, { "grad_norm": 0.38107162714004517, "learning_rate": 9.975757227417634e-05, "loss": 0.0532, "step": 7980 }, { "grad_norm": 0.4125936031341553, "learning_rate": 9.975594329212586e-05, "loss": 0.0581, "step": 7990 }, { "grad_norm": 0.42385461926460266, "learning_rate": 9.97543088688402e-05, "loss": 0.051, "step": 8000 }, { "grad_norm": 0.5159865021705627, "learning_rate": 9.975266900449814e-05, "loss": 0.0568, "step": 8010 }, { "grad_norm": 0.4957515597343445, "learning_rate": 9.975102369927898e-05, "loss": 0.0554, "step": 8020 }, { "grad_norm": 0.36537694931030273, "learning_rate": 9.974937295336269e-05, "loss": 0.0527, "step": 8030 }, { "grad_norm": 0.35498061776161194, "learning_rate": 9.974771676692975e-05, "loss": 0.0518, "step": 8040 }, { "grad_norm": 0.4854637086391449, "learning_rate": 9.974605514016131e-05, "loss": 0.0596, "step": 8050 }, { "grad_norm": 0.38414350152015686, "learning_rate": 9.974438807323907e-05, "loss": 0.0625, "step": 8060 }, { "grad_norm": 0.3391292989253998, "learning_rate": 9.974271556634535e-05, "loss": 0.0541, "step": 8070 }, { "grad_norm": 0.3872169554233551, "learning_rate": 9.974103761966302e-05, "loss": 0.0567, "step": 8080 }, { "grad_norm": 0.35389721393585205, "learning_rate": 9.973935423337563e-05, "loss": 0.055, "step": 8090 }, { "grad_norm": 0.4373152554035187, "learning_rate": 9.973766540766722e-05, "loss": 0.0565, "step": 8100 }, { "grad_norm": 0.6633663177490234, "learning_rate": 9.97359711427225e-05, "loss": 0.0522, "step": 8110 }, { "grad_norm": 0.2810637652873993, "learning_rate": 9.973427143872677e-05, "loss": 0.0594, "step": 8120 }, { "grad_norm": 0.3544493615627289, "learning_rate": 9.973256629586589e-05, "loss": 0.0499, "step": 8130 }, { "grad_norm": 0.3720862865447998, "learning_rate": 9.973085571432632e-05, "loss": 0.0628, "step": 8140 }, { "grad_norm": 0.40373528003692627, "learning_rate": 9.972913969429513e-05, "loss": 0.0561, "step": 8150 }, { "grad_norm": 0.4565989673137665, "learning_rate": 9.972741823596e-05, "loss": 0.0598, "step": 8160 }, { "grad_norm": 0.3813292682170868, "learning_rate": 9.972569133950917e-05, "loss": 0.0575, "step": 8170 }, { "grad_norm": 0.34899815917015076, "learning_rate": 9.972395900513151e-05, "loss": 0.0593, "step": 8180 }, { "grad_norm": 0.35875067114830017, "learning_rate": 9.972222123301645e-05, "loss": 0.0551, "step": 8190 }, { "grad_norm": 0.3970724642276764, "learning_rate": 9.972047802335403e-05, "loss": 0.0567, "step": 8200 }, { "grad_norm": 0.39821866154670715, "learning_rate": 9.971872937633488e-05, "loss": 0.0634, "step": 8210 }, { "grad_norm": 0.42805203795433044, "learning_rate": 9.971697529215024e-05, "loss": 0.0576, "step": 8220 }, { "grad_norm": 0.31377974152565, "learning_rate": 9.971521577099192e-05, "loss": 0.0562, "step": 8230 }, { "grad_norm": 0.4684707224369049, "learning_rate": 9.971345081305236e-05, "loss": 0.0453, "step": 8240 }, { "grad_norm": 0.38618186116218567, "learning_rate": 9.971168041852456e-05, "loss": 0.0596, "step": 8250 }, { "grad_norm": 0.5137717723846436, "learning_rate": 9.970990458760215e-05, "loss": 0.052, "step": 8260 }, { "grad_norm": 0.4718211591243744, "learning_rate": 9.970812332047929e-05, "loss": 0.0521, "step": 8270 }, { "grad_norm": 0.5512176156044006, "learning_rate": 9.97063366173508e-05, "loss": 0.0474, "step": 8280 }, { "grad_norm": 0.3812483549118042, "learning_rate": 9.970454447841207e-05, "loss": 0.0539, "step": 8290 }, { "grad_norm": 0.36078932881355286, "learning_rate": 9.970274690385909e-05, "loss": 0.0521, "step": 8300 }, { "grad_norm": 0.46582457423210144, "learning_rate": 9.970094389388844e-05, "loss": 0.048, "step": 8310 }, { "grad_norm": 0.447731077671051, "learning_rate": 9.969913544869728e-05, "loss": 0.0503, "step": 8320 }, { "grad_norm": 0.49788206815719604, "learning_rate": 9.96973215684834e-05, "loss": 0.0518, "step": 8330 }, { "grad_norm": 0.3911728858947754, "learning_rate": 9.969550225344513e-05, "loss": 0.0544, "step": 8340 }, { "grad_norm": 0.3926340937614441, "learning_rate": 9.969367750378147e-05, "loss": 0.0526, "step": 8350 }, { "grad_norm": 0.3701018691062927, "learning_rate": 9.969184731969194e-05, "loss": 0.061, "step": 8360 }, { "grad_norm": 0.3619856834411621, "learning_rate": 9.96900117013767e-05, "loss": 0.0557, "step": 8370 }, { "grad_norm": 0.38091203570365906, "learning_rate": 9.96881706490365e-05, "loss": 0.0544, "step": 8380 }, { "grad_norm": 0.5074671506881714, "learning_rate": 9.968632416287265e-05, "loss": 0.0564, "step": 8390 }, { "grad_norm": 0.7983483672142029, "learning_rate": 9.96844722430871e-05, "loss": 0.0497, "step": 8400 }, { "grad_norm": 0.29753077030181885, "learning_rate": 9.968261488988235e-05, "loss": 0.0526, "step": 8410 }, { "grad_norm": 0.4145583510398865, "learning_rate": 9.968075210346155e-05, "loss": 0.0504, "step": 8420 }, { "grad_norm": 0.4708218574523926, "learning_rate": 9.967888388402839e-05, "loss": 0.0583, "step": 8430 }, { "grad_norm": 0.44550666213035583, "learning_rate": 9.967701023178717e-05, "loss": 0.0573, "step": 8440 }, { "grad_norm": 0.4906218647956848, "learning_rate": 9.967513114694282e-05, "loss": 0.0577, "step": 8450 }, { "grad_norm": 0.36140337586402893, "learning_rate": 9.967324662970079e-05, "loss": 0.0501, "step": 8460 }, { "grad_norm": 0.4041441082954407, "learning_rate": 9.96713566802672e-05, "loss": 0.0552, "step": 8470 }, { "grad_norm": 0.30276480317115784, "learning_rate": 9.966946129884873e-05, "loss": 0.0534, "step": 8480 }, { "grad_norm": 0.3654511868953705, "learning_rate": 9.966756048565265e-05, "loss": 0.0514, "step": 8490 }, { "grad_norm": 0.39199844002723694, "learning_rate": 9.966565424088681e-05, "loss": 0.0617, "step": 8500 }, { "grad_norm": 0.34191328287124634, "learning_rate": 9.96637425647597e-05, "loss": 0.0482, "step": 8510 }, { "grad_norm": 0.3017869293689728, "learning_rate": 9.966182545748038e-05, "loss": 0.0428, "step": 8520 }, { "grad_norm": 0.49354448914527893, "learning_rate": 9.96599029192585e-05, "loss": 0.0656, "step": 8530 }, { "grad_norm": 0.3608805537223816, "learning_rate": 9.965797495030428e-05, "loss": 0.054, "step": 8540 }, { "grad_norm": 0.4072861969470978, "learning_rate": 9.96560415508286e-05, "loss": 0.0555, "step": 8550 }, { "grad_norm": 0.3463886082172394, "learning_rate": 9.965410272104286e-05, "loss": 0.0538, "step": 8560 }, { "grad_norm": 0.335723340511322, "learning_rate": 9.96521584611591e-05, "loss": 0.0533, "step": 8570 }, { "grad_norm": 0.36884310841560364, "learning_rate": 9.965020877138994e-05, "loss": 0.0474, "step": 8580 }, { "grad_norm": 0.31018826365470886, "learning_rate": 9.964825365194861e-05, "loss": 0.0518, "step": 8590 }, { "grad_norm": 0.4327804446220398, "learning_rate": 9.96462931030489e-05, "loss": 0.0472, "step": 8600 }, { "grad_norm": 0.3718584179878235, "learning_rate": 9.96443271249052e-05, "loss": 0.0512, "step": 8610 }, { "grad_norm": 0.39918258786201477, "learning_rate": 9.964235571773255e-05, "loss": 0.047, "step": 8620 }, { "grad_norm": 0.365741103887558, "learning_rate": 9.96403788817465e-05, "loss": 0.0454, "step": 8630 }, { "grad_norm": 0.33893725275993347, "learning_rate": 9.963839661716325e-05, "loss": 0.0447, "step": 8640 }, { "grad_norm": 0.36271458864212036, "learning_rate": 9.963640892419958e-05, "loss": 0.0421, "step": 8650 }, { "grad_norm": 0.35841840505599976, "learning_rate": 9.963441580307286e-05, "loss": 0.0553, "step": 8660 }, { "grad_norm": 0.4077622592449188, "learning_rate": 9.963241725400104e-05, "loss": 0.0514, "step": 8670 }, { "grad_norm": 0.444305419921875, "learning_rate": 9.963041327720271e-05, "loss": 0.0526, "step": 8680 }, { "grad_norm": 0.36249780654907227, "learning_rate": 9.962840387289697e-05, "loss": 0.0507, "step": 8690 }, { "grad_norm": 0.2950401306152344, "learning_rate": 9.962638904130363e-05, "loss": 0.0471, "step": 8700 }, { "grad_norm": 0.2998977601528168, "learning_rate": 9.962436878264298e-05, "loss": 0.0469, "step": 8710 }, { "grad_norm": 0.39026862382888794, "learning_rate": 9.962234309713598e-05, "loss": 0.0474, "step": 8720 }, { "grad_norm": 0.35245391726493835, "learning_rate": 9.962031198500414e-05, "loss": 0.0494, "step": 8730 }, { "grad_norm": 0.46778711676597595, "learning_rate": 9.961827544646958e-05, "loss": 0.0623, "step": 8740 }, { "grad_norm": 0.4116298258304596, "learning_rate": 9.961623348175501e-05, "loss": 0.0529, "step": 8750 }, { "grad_norm": 0.4940626621246338, "learning_rate": 9.961418609108377e-05, "loss": 0.0521, "step": 8760 }, { "grad_norm": 0.4402436912059784, "learning_rate": 9.961213327467971e-05, "loss": 0.0576, "step": 8770 }, { "grad_norm": 0.3534444570541382, "learning_rate": 9.961007503276736e-05, "loss": 0.0511, "step": 8780 }, { "grad_norm": 0.40426304936408997, "learning_rate": 9.960801136557179e-05, "loss": 0.0468, "step": 8790 }, { "grad_norm": 0.5764668583869934, "learning_rate": 9.960594227331866e-05, "loss": 0.0612, "step": 8800 }, { "grad_norm": 0.5342347025871277, "learning_rate": 9.960386775623429e-05, "loss": 0.0504, "step": 8810 }, { "grad_norm": 0.39874300360679626, "learning_rate": 9.96017878145455e-05, "loss": 0.0525, "step": 8820 }, { "grad_norm": 0.4372211694717407, "learning_rate": 9.959970244847977e-05, "loss": 0.0558, "step": 8830 }, { "grad_norm": 0.2859443724155426, "learning_rate": 9.959761165826518e-05, "loss": 0.0489, "step": 8840 }, { "grad_norm": 0.3135307729244232, "learning_rate": 9.959551544413033e-05, "loss": 0.0524, "step": 8850 }, { "grad_norm": 0.31215259432792664, "learning_rate": 9.959341380630448e-05, "loss": 0.048, "step": 8860 }, { "grad_norm": 0.3498096764087677, "learning_rate": 9.959130674501746e-05, "loss": 0.0538, "step": 8870 }, { "grad_norm": 0.44053518772125244, "learning_rate": 9.958919426049968e-05, "loss": 0.0436, "step": 8880 }, { "grad_norm": 0.33738499879837036, "learning_rate": 9.958707635298219e-05, "loss": 0.049, "step": 8890 }, { "grad_norm": 0.409353643655777, "learning_rate": 9.958495302269657e-05, "loss": 0.0501, "step": 8900 }, { "grad_norm": 0.3462561368942261, "learning_rate": 9.958282426987503e-05, "loss": 0.0585, "step": 8910 }, { "grad_norm": 0.39549657702445984, "learning_rate": 9.95806900947504e-05, "loss": 0.0464, "step": 8920 }, { "grad_norm": 0.3356640040874481, "learning_rate": 9.957855049755604e-05, "loss": 0.0595, "step": 8930 }, { "grad_norm": 0.38587918877601624, "learning_rate": 9.957640547852593e-05, "loss": 0.0572, "step": 8940 }, { "grad_norm": 0.462403267621994, "learning_rate": 9.957425503789466e-05, "loss": 0.0619, "step": 8950 }, { "grad_norm": 0.5055117607116699, "learning_rate": 9.957209917589738e-05, "loss": 0.0562, "step": 8960 }, { "grad_norm": 0.3754042088985443, "learning_rate": 9.956993789276987e-05, "loss": 0.0525, "step": 8970 }, { "grad_norm": 0.5313535928726196, "learning_rate": 9.956777118874847e-05, "loss": 0.0555, "step": 8980 }, { "grad_norm": 0.3723146319389343, "learning_rate": 9.956559906407016e-05, "loss": 0.0541, "step": 8990 }, { "grad_norm": 0.3742574155330658, "learning_rate": 9.956342151897245e-05, "loss": 0.0572, "step": 9000 }, { "grad_norm": 0.49175113439559937, "learning_rate": 9.956123855369346e-05, "loss": 0.0646, "step": 9010 }, { "grad_norm": 0.36307471990585327, "learning_rate": 9.955905016847196e-05, "loss": 0.0608, "step": 9020 }, { "grad_norm": 0.5760164260864258, "learning_rate": 9.955685636354723e-05, "loss": 0.0555, "step": 9030 }, { "grad_norm": 0.3909743130207062, "learning_rate": 9.95546571391592e-05, "loss": 0.0544, "step": 9040 }, { "grad_norm": 0.3858664929866791, "learning_rate": 9.955245249554837e-05, "loss": 0.0577, "step": 9050 }, { "grad_norm": 0.3347601890563965, "learning_rate": 9.955024243295582e-05, "loss": 0.0564, "step": 9060 }, { "grad_norm": 0.25364089012145996, "learning_rate": 9.954802695162328e-05, "loss": 0.0566, "step": 9070 }, { "grad_norm": 0.3258350193500519, "learning_rate": 9.954580605179302e-05, "loss": 0.0575, "step": 9080 }, { "grad_norm": 0.35279929637908936, "learning_rate": 9.954357973370788e-05, "loss": 0.0552, "step": 9090 }, { "grad_norm": 0.4379960894584656, "learning_rate": 9.954134799761135e-05, "loss": 0.058, "step": 9100 }, { "grad_norm": 0.38533303141593933, "learning_rate": 9.953911084374748e-05, "loss": 0.049, "step": 9110 }, { "grad_norm": 0.3569990396499634, "learning_rate": 9.953686827236093e-05, "loss": 0.0567, "step": 9120 }, { "grad_norm": 0.38090431690216064, "learning_rate": 9.953462028369695e-05, "loss": 0.0435, "step": 9130 }, { "grad_norm": 0.4360225200653076, "learning_rate": 9.953236687800136e-05, "loss": 0.0565, "step": 9140 }, { "grad_norm": 0.43725332617759705, "learning_rate": 9.95301080555206e-05, "loss": 0.0501, "step": 9150 }, { "grad_norm": 0.3611280620098114, "learning_rate": 9.952784381650171e-05, "loss": 0.0556, "step": 9160 }, { "grad_norm": 0.3485574424266815, "learning_rate": 9.952557416119226e-05, "loss": 0.0558, "step": 9170 }, { "grad_norm": 0.3526560068130493, "learning_rate": 9.95232990898405e-05, "loss": 0.0533, "step": 9180 }, { "grad_norm": 0.27648621797561646, "learning_rate": 9.95210186026952e-05, "loss": 0.0436, "step": 9190 }, { "grad_norm": 0.43216705322265625, "learning_rate": 9.951873270000576e-05, "loss": 0.0469, "step": 9200 }, { "grad_norm": 0.323789119720459, "learning_rate": 9.951644138202216e-05, "loss": 0.053, "step": 9210 }, { "grad_norm": 0.4499853551387787, "learning_rate": 9.951414464899498e-05, "loss": 0.0456, "step": 9220 }, { "grad_norm": 0.3795529901981354, "learning_rate": 9.951184250117538e-05, "loss": 0.0444, "step": 9230 }, { "grad_norm": 0.3160145878791809, "learning_rate": 9.950953493881513e-05, "loss": 0.0434, "step": 9240 }, { "grad_norm": 0.38552868366241455, "learning_rate": 9.950722196216658e-05, "loss": 0.0428, "step": 9250 }, { "grad_norm": 0.29236510396003723, "learning_rate": 9.950490357148265e-05, "loss": 0.0448, "step": 9260 }, { "grad_norm": 0.3811708390712738, "learning_rate": 9.950257976701692e-05, "loss": 0.0487, "step": 9270 }, { "grad_norm": 0.43121927976608276, "learning_rate": 9.950025054902348e-05, "loss": 0.0516, "step": 9280 }, { "grad_norm": 0.34685197472572327, "learning_rate": 9.949791591775706e-05, "loss": 0.0442, "step": 9290 }, { "grad_norm": 0.34184810519218445, "learning_rate": 9.949557587347298e-05, "loss": 0.0454, "step": 9300 }, { "grad_norm": 0.4253843426704407, "learning_rate": 9.949323041642713e-05, "loss": 0.0529, "step": 9310 }, { "grad_norm": 0.33675646781921387, "learning_rate": 9.949087954687602e-05, "loss": 0.0493, "step": 9320 }, { "grad_norm": 0.5187653303146362, "learning_rate": 9.948852326507672e-05, "loss": 0.0472, "step": 9330 }, { "grad_norm": 0.3426743149757385, "learning_rate": 9.948616157128694e-05, "loss": 0.0523, "step": 9340 }, { "grad_norm": 0.2807077467441559, "learning_rate": 9.948379446576493e-05, "loss": 0.0534, "step": 9350 }, { "grad_norm": 0.5233156681060791, "learning_rate": 9.948142194876952e-05, "loss": 0.0519, "step": 9360 }, { "grad_norm": 0.4832969605922699, "learning_rate": 9.947904402056024e-05, "loss": 0.0528, "step": 9370 }, { "grad_norm": 0.3738801181316376, "learning_rate": 9.947666068139708e-05, "loss": 0.0522, "step": 9380 }, { "grad_norm": 0.3331742286682129, "learning_rate": 9.947427193154071e-05, "loss": 0.0496, "step": 9390 }, { "grad_norm": 0.3461473882198334, "learning_rate": 9.947187777125233e-05, "loss": 0.0506, "step": 9400 }, { "grad_norm": 0.318248987197876, "learning_rate": 9.946947820079377e-05, "loss": 0.0502, "step": 9410 }, { "grad_norm": 0.3653043508529663, "learning_rate": 9.946707322042747e-05, "loss": 0.0526, "step": 9420 }, { "grad_norm": 0.3212428689002991, "learning_rate": 9.94646628304164e-05, "loss": 0.0489, "step": 9430 }, { "grad_norm": 0.44478943943977356, "learning_rate": 9.946224703102418e-05, "loss": 0.0524, "step": 9440 }, { "grad_norm": 0.32948610186576843, "learning_rate": 9.945982582251498e-05, "loss": 0.0551, "step": 9450 }, { "grad_norm": 0.3418050706386566, "learning_rate": 9.94573992051536e-05, "loss": 0.0523, "step": 9460 }, { "grad_norm": 0.26698586344718933, "learning_rate": 9.94549671792054e-05, "loss": 0.0503, "step": 9470 }, { "grad_norm": 0.45970889925956726, "learning_rate": 9.945252974493635e-05, "loss": 0.048, "step": 9480 }, { "grad_norm": 0.4575933516025543, "learning_rate": 9.9450086902613e-05, "loss": 0.0493, "step": 9490 }, { "grad_norm": 0.40629974007606506, "learning_rate": 9.944763865250248e-05, "loss": 0.0533, "step": 9500 }, { "grad_norm": 0.4606097638607025, "learning_rate": 9.944518499487254e-05, "loss": 0.0495, "step": 9510 }, { "grad_norm": 0.5255265235900879, "learning_rate": 9.944272592999151e-05, "loss": 0.0502, "step": 9520 }, { "grad_norm": 0.5499163269996643, "learning_rate": 9.94402614581283e-05, "loss": 0.0496, "step": 9530 }, { "grad_norm": 0.30558061599731445, "learning_rate": 9.943779157955244e-05, "loss": 0.0528, "step": 9540 }, { "grad_norm": 0.4467090666294098, "learning_rate": 9.943531629453403e-05, "loss": 0.0512, "step": 9550 }, { "grad_norm": 0.2561991512775421, "learning_rate": 9.943283560334375e-05, "loss": 0.0473, "step": 9560 }, { "grad_norm": 0.3109055161476135, "learning_rate": 9.943034950625288e-05, "loss": 0.0464, "step": 9570 }, { "grad_norm": 0.5127304792404175, "learning_rate": 9.942785800353332e-05, "loss": 0.0516, "step": 9580 }, { "grad_norm": 0.35189858078956604, "learning_rate": 9.942536109545751e-05, "loss": 0.0534, "step": 9590 }, { "grad_norm": 0.3851185142993927, "learning_rate": 9.942285878229853e-05, "loss": 0.0441, "step": 9600 }, { "grad_norm": 0.3093411326408386, "learning_rate": 9.942035106433001e-05, "loss": 0.0502, "step": 9610 }, { "grad_norm": 0.632173478603363, "learning_rate": 9.94178379418262e-05, "loss": 0.0573, "step": 9620 }, { "grad_norm": 0.45892879366874695, "learning_rate": 9.941531941506194e-05, "loss": 0.0531, "step": 9630 }, { "grad_norm": 0.39248988032341003, "learning_rate": 9.941279548431263e-05, "loss": 0.0532, "step": 9640 }, { "grad_norm": 0.30331966280937195, "learning_rate": 9.941026614985431e-05, "loss": 0.0544, "step": 9650 }, { "grad_norm": 0.3944055140018463, "learning_rate": 9.940773141196357e-05, "loss": 0.0571, "step": 9660 }, { "grad_norm": 0.3564179837703705, "learning_rate": 9.94051912709176e-05, "loss": 0.051, "step": 9670 }, { "grad_norm": 0.35905587673187256, "learning_rate": 9.940264572699421e-05, "loss": 0.0516, "step": 9680 }, { "grad_norm": 0.39132159948349, "learning_rate": 9.940009478047174e-05, "loss": 0.054, "step": 9690 }, { "grad_norm": 0.3519262969493866, "learning_rate": 9.939753843162918e-05, "loss": 0.0557, "step": 9700 }, { "grad_norm": 0.3575972318649292, "learning_rate": 9.939497668074609e-05, "loss": 0.0504, "step": 9710 }, { "grad_norm": 0.42956122756004333, "learning_rate": 9.93924095281026e-05, "loss": 0.0513, "step": 9720 }, { "grad_norm": 0.2798291742801666, "learning_rate": 9.938983697397948e-05, "loss": 0.0508, "step": 9730 }, { "grad_norm": 0.27055490016937256, "learning_rate": 9.938725901865805e-05, "loss": 0.0471, "step": 9740 }, { "grad_norm": 0.3330162465572357, "learning_rate": 9.93846756624202e-05, "loss": 0.0468, "step": 9750 }, { "grad_norm": 0.4588657021522522, "learning_rate": 9.938208690554849e-05, "loss": 0.0443, "step": 9760 }, { "grad_norm": 0.40373942255973816, "learning_rate": 9.9379492748326e-05, "loss": 0.0511, "step": 9770 }, { "grad_norm": 0.3162996470928192, "learning_rate": 9.937689319103641e-05, "loss": 0.0467, "step": 9780 }, { "grad_norm": 0.35621747374534607, "learning_rate": 9.937428823396404e-05, "loss": 0.0437, "step": 9790 }, { "grad_norm": 0.3031654357910156, "learning_rate": 9.937167787739372e-05, "loss": 0.0513, "step": 9800 }, { "grad_norm": 0.3405665457248688, "learning_rate": 9.936906212161095e-05, "loss": 0.0491, "step": 9810 }, { "grad_norm": 0.3401898741722107, "learning_rate": 9.936644096690176e-05, "loss": 0.0501, "step": 9820 }, { "grad_norm": 0.3171832263469696, "learning_rate": 9.936381441355282e-05, "loss": 0.0491, "step": 9830 }, { "grad_norm": 0.39691391587257385, "learning_rate": 9.936118246185136e-05, "loss": 0.0523, "step": 9840 }, { "grad_norm": 0.3883192241191864, "learning_rate": 9.935854511208518e-05, "loss": 0.0523, "step": 9850 }, { "grad_norm": 0.36951595544815063, "learning_rate": 9.935590236454272e-05, "loss": 0.0458, "step": 9860 }, { "grad_norm": 0.4334831237792969, "learning_rate": 9.935325421951298e-05, "loss": 0.0463, "step": 9870 }, { "grad_norm": 0.4386645555496216, "learning_rate": 9.935060067728557e-05, "loss": 0.0475, "step": 9880 }, { "grad_norm": 0.4013185501098633, "learning_rate": 9.934794173815067e-05, "loss": 0.0508, "step": 9890 }, { "grad_norm": 0.3811300992965698, "learning_rate": 9.934527740239906e-05, "loss": 0.0487, "step": 9900 }, { "grad_norm": 0.4229215681552887, "learning_rate": 9.934260767032209e-05, "loss": 0.044, "step": 9910 }, { "grad_norm": 0.4194489121437073, "learning_rate": 9.933993254221172e-05, "loss": 0.0544, "step": 9920 }, { "grad_norm": 0.34805744886398315, "learning_rate": 9.933725201836053e-05, "loss": 0.0485, "step": 9930 }, { "grad_norm": 0.3444404900074005, "learning_rate": 9.933456609906162e-05, "loss": 0.0499, "step": 9940 }, { "grad_norm": 0.3663182556629181, "learning_rate": 9.933187478460875e-05, "loss": 0.0431, "step": 9950 }, { "grad_norm": 0.29427191615104675, "learning_rate": 9.93291780752962e-05, "loss": 0.0415, "step": 9960 }, { "grad_norm": 0.3507573902606964, "learning_rate": 9.932647597141893e-05, "loss": 0.0445, "step": 9970 }, { "grad_norm": 0.30810484290122986, "learning_rate": 9.932376847327239e-05, "loss": 0.0443, "step": 9980 }, { "grad_norm": 0.5094755291938782, "learning_rate": 9.932105558115268e-05, "loss": 0.0501, "step": 9990 }, { "grad_norm": 0.36594709753990173, "learning_rate": 9.931833729535651e-05, "loss": 0.0509, "step": 10000 }, { "grad_norm": 0.27444741129875183, "learning_rate": 9.931561361618111e-05, "loss": 0.0476, "step": 10010 }, { "grad_norm": 0.3287566602230072, "learning_rate": 9.931288454392435e-05, "loss": 0.0436, "step": 10020 }, { "grad_norm": 0.34086042642593384, "learning_rate": 9.931015007888467e-05, "loss": 0.0483, "step": 10030 }, { "grad_norm": 0.47466954588890076, "learning_rate": 9.930741022136112e-05, "loss": 0.0578, "step": 10040 }, { "grad_norm": 0.49805107712745667, "learning_rate": 9.930466497165333e-05, "loss": 0.0431, "step": 10050 }, { "grad_norm": 0.3757033050060272, "learning_rate": 9.93019143300615e-05, "loss": 0.0559, "step": 10060 }, { "grad_norm": 0.3281531035900116, "learning_rate": 9.929915829688644e-05, "loss": 0.0407, "step": 10070 }, { "grad_norm": 0.3654947280883789, "learning_rate": 9.929639687242955e-05, "loss": 0.0507, "step": 10080 }, { "grad_norm": 0.3063330352306366, "learning_rate": 9.929363005699281e-05, "loss": 0.0476, "step": 10090 }, { "grad_norm": 0.4041636884212494, "learning_rate": 9.92908578508788e-05, "loss": 0.0481, "step": 10100 }, { "grad_norm": 0.34936538338661194, "learning_rate": 9.928808025439069e-05, "loss": 0.0515, "step": 10110 }, { "grad_norm": 0.5106599926948547, "learning_rate": 9.928529726783223e-05, "loss": 0.0508, "step": 10120 }, { "grad_norm": 0.34660831093788147, "learning_rate": 9.928250889150774e-05, "loss": 0.0482, "step": 10130 }, { "grad_norm": 0.32736554741859436, "learning_rate": 9.92797151257222e-05, "loss": 0.0468, "step": 10140 }, { "grad_norm": 0.44060128927230835, "learning_rate": 9.927691597078108e-05, "loss": 0.0523, "step": 10150 }, { "grad_norm": 0.3084362745285034, "learning_rate": 9.927411142699053e-05, "loss": 0.0444, "step": 10160 }, { "grad_norm": 0.3747149109840393, "learning_rate": 9.927130149465725e-05, "loss": 0.0495, "step": 10170 }, { "grad_norm": 0.5128858685493469, "learning_rate": 9.92684861740885e-05, "loss": 0.0528, "step": 10180 }, { "grad_norm": 0.46515920758247375, "learning_rate": 9.926566546559217e-05, "loss": 0.056, "step": 10190 }, { "grad_norm": 0.3386708199977875, "learning_rate": 9.926283936947673e-05, "loss": 0.055, "step": 10200 }, { "grad_norm": 0.4104672372341156, "learning_rate": 9.926000788605126e-05, "loss": 0.0484, "step": 10210 }, { "grad_norm": 0.4786820709705353, "learning_rate": 9.92571710156254e-05, "loss": 0.0493, "step": 10220 }, { "grad_norm": 0.46075278520584106, "learning_rate": 9.925432875850936e-05, "loss": 0.0544, "step": 10230 }, { "grad_norm": 0.4168274700641632, "learning_rate": 9.925148111501396e-05, "loss": 0.0499, "step": 10240 }, { "grad_norm": 0.42817187309265137, "learning_rate": 9.924862808545066e-05, "loss": 0.0439, "step": 10250 }, { "grad_norm": 0.3789260983467102, "learning_rate": 9.924576967013141e-05, "loss": 0.0501, "step": 10260 }, { "grad_norm": 0.29488030076026917, "learning_rate": 9.924290586936887e-05, "loss": 0.0473, "step": 10270 }, { "grad_norm": 0.4586491286754608, "learning_rate": 9.924003668347614e-05, "loss": 0.0525, "step": 10280 }, { "grad_norm": 0.35877519845962524, "learning_rate": 9.923716211276704e-05, "loss": 0.047, "step": 10290 }, { "grad_norm": 0.32751724123954773, "learning_rate": 9.923428215755594e-05, "loss": 0.0554, "step": 10300 }, { "grad_norm": 0.4647864103317261, "learning_rate": 9.923139681815775e-05, "loss": 0.0525, "step": 10310 }, { "grad_norm": 0.40411633253097534, "learning_rate": 9.922850609488801e-05, "loss": 0.047, "step": 10320 }, { "grad_norm": 0.3450279235839844, "learning_rate": 9.922560998806287e-05, "loss": 0.0458, "step": 10330 }, { "grad_norm": 0.37609100341796875, "learning_rate": 9.922270849799905e-05, "loss": 0.0443, "step": 10340 }, { "grad_norm": 0.40295690298080444, "learning_rate": 9.92198016250138e-05, "loss": 0.0524, "step": 10350 }, { "grad_norm": 0.3401024043560028, "learning_rate": 9.921688936942506e-05, "loss": 0.0407, "step": 10360 }, { "grad_norm": 0.372504323720932, "learning_rate": 9.921397173155129e-05, "loss": 0.0468, "step": 10370 }, { "grad_norm": 0.28201669454574585, "learning_rate": 9.921104871171157e-05, "loss": 0.0459, "step": 10380 }, { "grad_norm": 0.3293721079826355, "learning_rate": 9.920812031022554e-05, "loss": 0.0482, "step": 10390 }, { "grad_norm": 0.303854376077652, "learning_rate": 9.920518652741348e-05, "loss": 0.0466, "step": 10400 }, { "grad_norm": 0.3340703248977661, "learning_rate": 9.920224736359618e-05, "loss": 0.0487, "step": 10410 }, { "grad_norm": 0.2778138220310211, "learning_rate": 9.91993028190951e-05, "loss": 0.0414, "step": 10420 }, { "grad_norm": 0.41058480739593506, "learning_rate": 9.919635289423222e-05, "loss": 0.0425, "step": 10430 }, { "grad_norm": 0.26096048951148987, "learning_rate": 9.919339758933015e-05, "loss": 0.0387, "step": 10440 }, { "grad_norm": 0.33036768436431885, "learning_rate": 9.919043690471209e-05, "loss": 0.0389, "step": 10450 }, { "grad_norm": 0.35123056173324585, "learning_rate": 9.91874708407018e-05, "loss": 0.0558, "step": 10460 }, { "grad_norm": 0.3451860249042511, "learning_rate": 9.918449939762367e-05, "loss": 0.0472, "step": 10470 }, { "grad_norm": 0.308724045753479, "learning_rate": 9.91815225758026e-05, "loss": 0.0453, "step": 10480 }, { "grad_norm": 0.44020190834999084, "learning_rate": 9.917854037556419e-05, "loss": 0.0454, "step": 10490 }, { "grad_norm": 0.3556680977344513, "learning_rate": 9.917555279723454e-05, "loss": 0.0404, "step": 10500 }, { "grad_norm": 0.343187540769577, "learning_rate": 9.917255984114036e-05, "loss": 0.045, "step": 10510 }, { "grad_norm": 0.5314446687698364, "learning_rate": 9.916956150760896e-05, "loss": 0.0403, "step": 10520 }, { "grad_norm": 0.33040499687194824, "learning_rate": 9.916655779696826e-05, "loss": 0.0443, "step": 10530 }, { "grad_norm": 0.40431472659111023, "learning_rate": 9.916354870954671e-05, "loss": 0.0467, "step": 10540 }, { "grad_norm": 0.24941405653953552, "learning_rate": 9.91605342456734e-05, "loss": 0.0435, "step": 10550 }, { "grad_norm": 2.1443474292755127, "learning_rate": 9.915751440567795e-05, "loss": 0.0455, "step": 10560 }, { "grad_norm": 0.4861806333065033, "learning_rate": 9.915448918989066e-05, "loss": 0.0585, "step": 10570 }, { "grad_norm": 0.4029988944530487, "learning_rate": 9.915145859864232e-05, "loss": 0.0517, "step": 10580 }, { "grad_norm": 0.35801488161087036, "learning_rate": 9.914842263226437e-05, "loss": 0.0582, "step": 10590 }, { "grad_norm": 0.34087300300598145, "learning_rate": 9.914538129108882e-05, "loss": 0.0562, "step": 10600 }, { "grad_norm": 0.4736572802066803, "learning_rate": 9.914233457544825e-05, "loss": 0.0551, "step": 10610 }, { "grad_norm": 0.3913670480251312, "learning_rate": 9.913928248567586e-05, "loss": 0.0474, "step": 10620 }, { "grad_norm": 0.4505714178085327, "learning_rate": 9.913622502210542e-05, "loss": 0.0548, "step": 10630 }, { "grad_norm": 0.39323270320892334, "learning_rate": 9.913316218507128e-05, "loss": 0.0524, "step": 10640 }, { "grad_norm": 0.4497497081756592, "learning_rate": 9.91300939749084e-05, "loss": 0.051, "step": 10650 }, { "grad_norm": 0.35823729634284973, "learning_rate": 9.91270203919523e-05, "loss": 0.0527, "step": 10660 }, { "grad_norm": 0.37832629680633545, "learning_rate": 9.912394143653912e-05, "loss": 0.052, "step": 10670 }, { "grad_norm": 0.34262487292289734, "learning_rate": 9.912085710900555e-05, "loss": 0.0526, "step": 10680 }, { "grad_norm": 0.37038442492485046, "learning_rate": 9.911776740968892e-05, "loss": 0.0468, "step": 10690 }, { "grad_norm": 0.36616766452789307, "learning_rate": 9.911467233892709e-05, "loss": 0.0552, "step": 10700 }, { "grad_norm": 0.3142869174480438, "learning_rate": 9.911157189705853e-05, "loss": 0.0489, "step": 10710 }, { "grad_norm": 0.2971804141998291, "learning_rate": 9.910846608442229e-05, "loss": 0.0523, "step": 10720 }, { "grad_norm": 0.30865761637687683, "learning_rate": 9.910535490135805e-05, "loss": 0.0454, "step": 10730 }, { "grad_norm": 0.43348070979118347, "learning_rate": 9.910223834820603e-05, "loss": 0.0499, "step": 10740 }, { "grad_norm": 0.33535537123680115, "learning_rate": 9.909911642530703e-05, "loss": 0.0475, "step": 10750 }, { "grad_norm": 0.34207814931869507, "learning_rate": 9.909598913300249e-05, "loss": 0.0524, "step": 10760 }, { "grad_norm": 0.5386326313018799, "learning_rate": 9.909285647163438e-05, "loss": 0.0481, "step": 10770 }, { "grad_norm": 0.3358062505722046, "learning_rate": 9.908971844154531e-05, "loss": 0.0455, "step": 10780 }, { "grad_norm": 0.4265573024749756, "learning_rate": 9.908657504307843e-05, "loss": 0.0614, "step": 10790 }, { "grad_norm": 0.6316945552825928, "learning_rate": 9.908342627657751e-05, "loss": 0.0558, "step": 10800 }, { "grad_norm": 0.33776986598968506, "learning_rate": 9.908027214238689e-05, "loss": 0.0552, "step": 10810 }, { "grad_norm": 0.3784693777561188, "learning_rate": 9.90771126408515e-05, "loss": 0.0508, "step": 10820 }, { "grad_norm": 0.3195761740207672, "learning_rate": 9.907394777231685e-05, "loss": 0.0528, "step": 10830 }, { "grad_norm": 0.5827722549438477, "learning_rate": 9.907077753712905e-05, "loss": 0.0501, "step": 10840 }, { "grad_norm": 0.33947569131851196, "learning_rate": 9.906760193563482e-05, "loss": 0.0573, "step": 10850 }, { "grad_norm": 0.4012192487716675, "learning_rate": 9.906442096818139e-05, "loss": 0.0592, "step": 10860 }, { "grad_norm": 0.3558918237686157, "learning_rate": 9.906123463511665e-05, "loss": 0.0463, "step": 10870 }, { "grad_norm": 0.39759501814842224, "learning_rate": 9.905804293678907e-05, "loss": 0.0486, "step": 10880 }, { "grad_norm": 0.3254307508468628, "learning_rate": 9.905484587354766e-05, "loss": 0.0464, "step": 10890 }, { "grad_norm": 0.3079908490180969, "learning_rate": 9.905164344574205e-05, "loss": 0.0473, "step": 10900 }, { "grad_norm": 0.4814301133155823, "learning_rate": 9.904843565372248e-05, "loss": 0.0471, "step": 10910 }, { "grad_norm": 0.43842679262161255, "learning_rate": 9.904522249783972e-05, "loss": 0.045, "step": 10920 }, { "grad_norm": 0.33773332834243774, "learning_rate": 9.904200397844517e-05, "loss": 0.0459, "step": 10930 }, { "grad_norm": 0.3359193801879883, "learning_rate": 9.903878009589078e-05, "loss": 0.0474, "step": 10940 }, { "grad_norm": 0.38826507329940796, "learning_rate": 9.903555085052915e-05, "loss": 0.0503, "step": 10950 }, { "grad_norm": 0.40814369916915894, "learning_rate": 9.903231624271338e-05, "loss": 0.0452, "step": 10960 }, { "grad_norm": 0.354454904794693, "learning_rate": 9.902907627279724e-05, "loss": 0.052, "step": 10970 }, { "grad_norm": 0.34398216009140015, "learning_rate": 9.902583094113504e-05, "loss": 0.0504, "step": 10980 }, { "grad_norm": 0.3375708758831024, "learning_rate": 9.902258024808168e-05, "loss": 0.0482, "step": 10990 }, { "grad_norm": 0.3587714731693268, "learning_rate": 9.901932419399264e-05, "loss": 0.0533, "step": 11000 }, { "grad_norm": 0.3471125662326813, "learning_rate": 9.9016062779224e-05, "loss": 0.0456, "step": 11010 }, { "grad_norm": 0.3917613625526428, "learning_rate": 9.901279600413242e-05, "loss": 0.0423, "step": 11020 }, { "grad_norm": 0.30904266238212585, "learning_rate": 9.900952386907518e-05, "loss": 0.0476, "step": 11030 }, { "grad_norm": 0.3147369623184204, "learning_rate": 9.90062463744101e-05, "loss": 0.0479, "step": 11040 }, { "grad_norm": 0.3380044400691986, "learning_rate": 9.900296352049558e-05, "loss": 0.0467, "step": 11050 }, { "grad_norm": 0.38121479749679565, "learning_rate": 9.899967530769065e-05, "loss": 0.0523, "step": 11060 }, { "grad_norm": 0.3043902814388275, "learning_rate": 9.899638173635489e-05, "loss": 0.0538, "step": 11070 }, { "grad_norm": 0.35394036769866943, "learning_rate": 9.899308280684849e-05, "loss": 0.0473, "step": 11080 }, { "grad_norm": 0.34820839762687683, "learning_rate": 9.898977851953222e-05, "loss": 0.0452, "step": 11090 }, { "grad_norm": 0.32984811067581177, "learning_rate": 9.898646887476741e-05, "loss": 0.0506, "step": 11100 }, { "grad_norm": 0.6641027331352234, "learning_rate": 9.898315387291603e-05, "loss": 0.0453, "step": 11110 }, { "grad_norm": 0.4233356714248657, "learning_rate": 9.89798335143406e-05, "loss": 0.0517, "step": 11120 }, { "grad_norm": 0.3560253381729126, "learning_rate": 9.897650779940419e-05, "loss": 0.0497, "step": 11130 }, { "grad_norm": 0.36218398809432983, "learning_rate": 9.897317672847054e-05, "loss": 0.0473, "step": 11140 }, { "grad_norm": 0.34496909379959106, "learning_rate": 9.89698403019039e-05, "loss": 0.0437, "step": 11150 }, { "grad_norm": 0.330182284116745, "learning_rate": 9.896649852006917e-05, "loss": 0.0475, "step": 11160 }, { "grad_norm": 0.30452215671539307, "learning_rate": 9.896315138333177e-05, "loss": 0.0414, "step": 11170 }, { "grad_norm": 0.3584408164024353, "learning_rate": 9.895979889205774e-05, "loss": 0.0387, "step": 11180 }, { "grad_norm": 0.3498450219631195, "learning_rate": 9.895644104661372e-05, "loss": 0.048, "step": 11190 }, { "grad_norm": 0.34243959188461304, "learning_rate": 9.895307784736691e-05, "loss": 0.0401, "step": 11200 }, { "grad_norm": 0.3158833682537079, "learning_rate": 9.894970929468512e-05, "loss": 0.0431, "step": 11210 }, { "grad_norm": 0.5704686045646667, "learning_rate": 9.89463353889367e-05, "loss": 0.0388, "step": 11220 }, { "grad_norm": 0.3859318792819977, "learning_rate": 9.894295613049065e-05, "loss": 0.0532, "step": 11230 }, { "grad_norm": 0.38961169123649597, "learning_rate": 9.893957151971649e-05, "loss": 0.0446, "step": 11240 }, { "grad_norm": 0.3272552192211151, "learning_rate": 9.893618155698436e-05, "loss": 0.0447, "step": 11250 }, { "grad_norm": 0.39865919947624207, "learning_rate": 9.8932786242665e-05, "loss": 0.0507, "step": 11260 }, { "grad_norm": 0.3135634660720825, "learning_rate": 9.89293855771297e-05, "loss": 0.0412, "step": 11270 }, { "grad_norm": 0.36153554916381836, "learning_rate": 9.892597956075036e-05, "loss": 0.0408, "step": 11280 }, { "grad_norm": 0.26808756589889526, "learning_rate": 9.892256819389947e-05, "loss": 0.0483, "step": 11290 }, { "grad_norm": 0.32921186089515686, "learning_rate": 9.891915147695006e-05, "loss": 0.046, "step": 11300 }, { "grad_norm": 0.31383270025253296, "learning_rate": 9.891572941027577e-05, "loss": 0.0416, "step": 11310 }, { "grad_norm": 0.31917431950569153, "learning_rate": 9.89123019942509e-05, "loss": 0.0393, "step": 11320 }, { "grad_norm": 0.8540762662887573, "learning_rate": 9.89088692292502e-05, "loss": 0.0557, "step": 11330 }, { "grad_norm": 0.3539579510688782, "learning_rate": 9.89054311156491e-05, "loss": 0.0461, "step": 11340 }, { "grad_norm": 0.39999768137931824, "learning_rate": 9.890198765382357e-05, "loss": 0.0427, "step": 11350 }, { "grad_norm": 0.4229260981082916, "learning_rate": 9.889853884415021e-05, "loss": 0.0454, "step": 11360 }, { "grad_norm": 0.36393967270851135, "learning_rate": 9.889508468700614e-05, "loss": 0.0468, "step": 11370 }, { "grad_norm": 0.26623299717903137, "learning_rate": 9.889162518276915e-05, "loss": 0.0455, "step": 11380 }, { "grad_norm": 0.2856539189815521, "learning_rate": 9.888816033181752e-05, "loss": 0.0426, "step": 11390 }, { "grad_norm": 0.321359783411026, "learning_rate": 9.888469013453018e-05, "loss": 0.0403, "step": 11400 }, { "grad_norm": 0.32147008180618286, "learning_rate": 9.888121459128663e-05, "loss": 0.0482, "step": 11410 }, { "grad_norm": 0.37245193123817444, "learning_rate": 9.887773370246693e-05, "loss": 0.0493, "step": 11420 }, { "grad_norm": 0.33903563022613525, "learning_rate": 9.887424746845177e-05, "loss": 0.0528, "step": 11430 }, { "grad_norm": 0.32196348905563354, "learning_rate": 9.887075588962239e-05, "loss": 0.0482, "step": 11440 }, { "grad_norm": 0.24704810976982117, "learning_rate": 9.88672589663606e-05, "loss": 0.0424, "step": 11450 }, { "grad_norm": 0.3408089578151703, "learning_rate": 9.886375669904886e-05, "loss": 0.0431, "step": 11460 }, { "grad_norm": 0.3532370328903198, "learning_rate": 9.886024908807014e-05, "loss": 0.0458, "step": 11470 }, { "grad_norm": 0.5123305916786194, "learning_rate": 9.885673613380806e-05, "loss": 0.056, "step": 11480 }, { "grad_norm": 0.3266412317752838, "learning_rate": 9.885321783664676e-05, "loss": 0.0494, "step": 11490 }, { "grad_norm": 0.4434637129306793, "learning_rate": 9.884969419697101e-05, "loss": 0.0449, "step": 11500 }, { "grad_norm": 0.2779988646507263, "learning_rate": 9.884616521516614e-05, "loss": 0.0476, "step": 11510 }, { "grad_norm": 0.3475705683231354, "learning_rate": 9.88426308916181e-05, "loss": 0.0445, "step": 11520 }, { "grad_norm": 0.3704659342765808, "learning_rate": 9.883909122671335e-05, "loss": 0.042, "step": 11530 }, { "grad_norm": 0.3739197850227356, "learning_rate": 9.883554622083904e-05, "loss": 0.0444, "step": 11540 }, { "grad_norm": 0.6037025451660156, "learning_rate": 9.88319958743828e-05, "loss": 0.0477, "step": 11550 }, { "grad_norm": 0.4130690097808838, "learning_rate": 9.882844018773291e-05, "loss": 0.0451, "step": 11560 }, { "grad_norm": 0.38339298963546753, "learning_rate": 9.882487916127823e-05, "loss": 0.0446, "step": 11570 }, { "grad_norm": 0.3615727126598358, "learning_rate": 9.882131279540815e-05, "loss": 0.0453, "step": 11580 }, { "grad_norm": 0.5324529409408569, "learning_rate": 9.881774109051271e-05, "loss": 0.047, "step": 11590 }, { "grad_norm": 0.29738953709602356, "learning_rate": 9.881416404698252e-05, "loss": 0.0415, "step": 11600 }, { "grad_norm": 0.3471683859825134, "learning_rate": 9.881058166520873e-05, "loss": 0.0475, "step": 11610 }, { "grad_norm": 0.38899970054626465, "learning_rate": 9.880699394558311e-05, "loss": 0.0523, "step": 11620 }, { "grad_norm": 0.3326849341392517, "learning_rate": 9.880340088849801e-05, "loss": 0.042, "step": 11630 }, { "grad_norm": 0.305390864610672, "learning_rate": 9.879980249434637e-05, "loss": 0.0482, "step": 11640 }, { "grad_norm": 0.33421528339385986, "learning_rate": 9.879619876352168e-05, "loss": 0.0444, "step": 11650 }, { "grad_norm": 0.4538671374320984, "learning_rate": 9.879258969641809e-05, "loss": 0.0533, "step": 11660 }, { "grad_norm": 0.4136192798614502, "learning_rate": 9.878897529343023e-05, "loss": 0.0479, "step": 11670 }, { "grad_norm": 0.34111639857292175, "learning_rate": 9.878535555495338e-05, "loss": 0.0508, "step": 11680 }, { "grad_norm": 0.3153097927570343, "learning_rate": 9.87817304813834e-05, "loss": 0.0556, "step": 11690 }, { "grad_norm": 0.2800796627998352, "learning_rate": 9.877810007311671e-05, "loss": 0.0434, "step": 11700 }, { "grad_norm": 0.3412986099720001, "learning_rate": 9.877446433055035e-05, "loss": 0.0449, "step": 11710 }, { "grad_norm": 0.2910092771053314, "learning_rate": 9.877082325408191e-05, "loss": 0.0455, "step": 11720 }, { "grad_norm": 0.3607761859893799, "learning_rate": 9.876717684410954e-05, "loss": 0.0515, "step": 11730 }, { "grad_norm": 0.3508486747741699, "learning_rate": 9.876352510103204e-05, "loss": 0.047, "step": 11740 }, { "grad_norm": 0.3604107201099396, "learning_rate": 9.875986802524875e-05, "loss": 0.0438, "step": 11750 }, { "grad_norm": 0.3675523102283478, "learning_rate": 9.87562056171596e-05, "loss": 0.0485, "step": 11760 }, { "grad_norm": 0.3380461633205414, "learning_rate": 9.875253787716511e-05, "loss": 0.0432, "step": 11770 }, { "grad_norm": 0.3450464904308319, "learning_rate": 9.874886480566637e-05, "loss": 0.0473, "step": 11780 }, { "grad_norm": 0.38397321105003357, "learning_rate": 9.874518640306507e-05, "loss": 0.0454, "step": 11790 }, { "grad_norm": 0.26028886437416077, "learning_rate": 9.874150266976347e-05, "loss": 0.041, "step": 11800 }, { "grad_norm": 0.5001546740531921, "learning_rate": 9.873781360616443e-05, "loss": 0.0439, "step": 11810 }, { "grad_norm": 0.43883275985717773, "learning_rate": 9.873411921267137e-05, "loss": 0.0498, "step": 11820 }, { "grad_norm": 0.3473135828971863, "learning_rate": 9.873041948968829e-05, "loss": 0.0465, "step": 11830 }, { "grad_norm": 0.23877672851085663, "learning_rate": 9.872671443761981e-05, "loss": 0.0387, "step": 11840 }, { "grad_norm": 0.2604241669178009, "learning_rate": 9.872300405687109e-05, "loss": 0.0413, "step": 11850 }, { "grad_norm": 0.20465758442878723, "learning_rate": 9.871928834784792e-05, "loss": 0.0416, "step": 11860 }, { "grad_norm": 0.7388666868209839, "learning_rate": 9.871556731095661e-05, "loss": 0.0479, "step": 11870 }, { "grad_norm": 0.32062920928001404, "learning_rate": 9.871184094660411e-05, "loss": 0.0513, "step": 11880 }, { "grad_norm": 0.4539012312889099, "learning_rate": 9.870810925519791e-05, "loss": 0.0466, "step": 11890 }, { "grad_norm": 0.3044845163822174, "learning_rate": 9.870437223714612e-05, "loss": 0.0452, "step": 11900 }, { "grad_norm": 0.3181373178958893, "learning_rate": 9.87006298928574e-05, "loss": 0.0432, "step": 11910 }, { "grad_norm": 0.3492434322834015, "learning_rate": 9.869688222274103e-05, "loss": 0.0375, "step": 11920 }, { "grad_norm": 0.37575432658195496, "learning_rate": 9.869312922720681e-05, "loss": 0.0472, "step": 11930 }, { "grad_norm": 0.3235952854156494, "learning_rate": 9.868937090666521e-05, "loss": 0.0419, "step": 11940 }, { "grad_norm": 0.4188508093357086, "learning_rate": 9.86856072615272e-05, "loss": 0.0402, "step": 11950 }, { "grad_norm": 0.36848947405815125, "learning_rate": 9.868183829220438e-05, "loss": 0.0491, "step": 11960 }, { "grad_norm": 0.2938978374004364, "learning_rate": 9.867806399910893e-05, "loss": 0.0366, "step": 11970 }, { "grad_norm": 0.35207056999206543, "learning_rate": 9.867428438265356e-05, "loss": 0.0462, "step": 11980 }, { "grad_norm": 0.48251935839653015, "learning_rate": 9.867049944325165e-05, "loss": 0.0462, "step": 11990 }, { "grad_norm": 0.3347910940647125, "learning_rate": 9.86667091813171e-05, "loss": 0.041, "step": 12000 }, { "grad_norm": 0.35393357276916504, "learning_rate": 9.866291359726438e-05, "loss": 0.0439, "step": 12010 }, { "grad_norm": 1.0279409885406494, "learning_rate": 9.865911269150861e-05, "loss": 0.0465, "step": 12020 }, { "grad_norm": 0.3609522879123688, "learning_rate": 9.865530646446544e-05, "loss": 0.0491, "step": 12030 }, { "grad_norm": 0.28416869044303894, "learning_rate": 9.86514949165511e-05, "loss": 0.0504, "step": 12040 }, { "grad_norm": 0.27980682253837585, "learning_rate": 9.864767804818243e-05, "loss": 0.0438, "step": 12050 }, { "grad_norm": 0.3251280188560486, "learning_rate": 9.86438558597768e-05, "loss": 0.0375, "step": 12060 }, { "grad_norm": 0.361467182636261, "learning_rate": 9.864002835175225e-05, "loss": 0.044, "step": 12070 }, { "grad_norm": 0.3757465183734894, "learning_rate": 9.863619552452734e-05, "loss": 0.0459, "step": 12080 }, { "grad_norm": 0.28921976685523987, "learning_rate": 9.863235737852119e-05, "loss": 0.0385, "step": 12090 }, { "grad_norm": 0.37452617287635803, "learning_rate": 9.862851391415356e-05, "loss": 0.0437, "step": 12100 }, { "grad_norm": 0.3370845317840576, "learning_rate": 9.862466513184477e-05, "loss": 0.0461, "step": 12110 }, { "grad_norm": 0.41700634360313416, "learning_rate": 9.86208110320157e-05, "loss": 0.0431, "step": 12120 }, { "grad_norm": 0.31810441613197327, "learning_rate": 9.861695161508784e-05, "loss": 0.0471, "step": 12130 }, { "grad_norm": 0.29962286353111267, "learning_rate": 9.861308688148324e-05, "loss": 0.0367, "step": 12140 }, { "grad_norm": 0.3431469202041626, "learning_rate": 9.860921683162455e-05, "loss": 0.0517, "step": 12150 }, { "grad_norm": 0.3188369572162628, "learning_rate": 9.860534146593499e-05, "loss": 0.0547, "step": 12160 }, { "grad_norm": 0.2818196415901184, "learning_rate": 9.860146078483836e-05, "loss": 0.0467, "step": 12170 }, { "grad_norm": 0.4690272808074951, "learning_rate": 9.859757478875905e-05, "loss": 0.0568, "step": 12180 }, { "grad_norm": 0.3852277100086212, "learning_rate": 9.859368347812204e-05, "loss": 0.0504, "step": 12190 }, { "grad_norm": 0.31784787774086, "learning_rate": 9.858978685335285e-05, "loss": 0.0506, "step": 12200 }, { "grad_norm": 0.3731967806816101, "learning_rate": 9.858588491487763e-05, "loss": 0.0448, "step": 12210 }, { "grad_norm": 0.43887829780578613, "learning_rate": 9.858197766312308e-05, "loss": 0.0446, "step": 12220 }, { "grad_norm": 0.35786646604537964, "learning_rate": 9.857806509851649e-05, "loss": 0.0421, "step": 12230 }, { "grad_norm": 0.370095431804657, "learning_rate": 9.857414722148574e-05, "loss": 0.0482, "step": 12240 }, { "grad_norm": 0.37102454900741577, "learning_rate": 9.857022403245928e-05, "loss": 0.045, "step": 12250 }, { "grad_norm": 0.3240910768508911, "learning_rate": 9.856629553186615e-05, "loss": 0.0433, "step": 12260 }, { "grad_norm": 0.3926573097705841, "learning_rate": 9.856236172013595e-05, "loss": 0.0454, "step": 12270 }, { "grad_norm": 0.3444381356239319, "learning_rate": 9.85584225976989e-05, "loss": 0.0449, "step": 12280 }, { "grad_norm": 0.24569769203662872, "learning_rate": 9.855447816498575e-05, "loss": 0.0408, "step": 12290 }, { "grad_norm": 0.37716925144195557, "learning_rate": 9.855052842242787e-05, "loss": 0.0466, "step": 12300 }, { "grad_norm": 0.3676066994667053, "learning_rate": 9.85465733704572e-05, "loss": 0.0417, "step": 12310 }, { "grad_norm": 0.2893206775188446, "learning_rate": 9.854261300950624e-05, "loss": 0.042, "step": 12320 }, { "grad_norm": 0.37651360034942627, "learning_rate": 9.853864734000813e-05, "loss": 0.0484, "step": 12330 }, { "grad_norm": 0.2638469338417053, "learning_rate": 9.85346763623965e-05, "loss": 0.0395, "step": 12340 }, { "grad_norm": 0.371609091758728, "learning_rate": 9.853070007710564e-05, "loss": 0.0391, "step": 12350 }, { "grad_norm": 0.5240578651428223, "learning_rate": 9.85267184845704e-05, "loss": 0.0425, "step": 12360 }, { "grad_norm": 0.36323630809783936, "learning_rate": 9.852273158522616e-05, "loss": 0.0403, "step": 12370 }, { "grad_norm": 0.3025631606578827, "learning_rate": 9.851873937950896e-05, "loss": 0.0486, "step": 12380 }, { "grad_norm": 0.3687131702899933, "learning_rate": 9.851474186785537e-05, "loss": 0.0476, "step": 12390 }, { "grad_norm": 0.40406474471092224, "learning_rate": 9.851073905070254e-05, "loss": 0.0428, "step": 12400 }, { "grad_norm": 0.3803628981113434, "learning_rate": 9.850673092848824e-05, "loss": 0.0491, "step": 12410 }, { "grad_norm": 0.37411049008369446, "learning_rate": 9.850271750165077e-05, "loss": 0.0446, "step": 12420 }, { "grad_norm": 0.2603725790977478, "learning_rate": 9.849869877062902e-05, "loss": 0.0441, "step": 12430 }, { "grad_norm": 0.2538454234600067, "learning_rate": 9.849467473586252e-05, "loss": 0.0424, "step": 12440 }, { "grad_norm": 0.2584332227706909, "learning_rate": 9.849064539779127e-05, "loss": 0.0445, "step": 12450 }, { "grad_norm": 0.29250553250312805, "learning_rate": 9.848661075685594e-05, "loss": 0.0413, "step": 12460 }, { "grad_norm": 0.33436426520347595, "learning_rate": 9.848257081349778e-05, "loss": 0.0407, "step": 12470 }, { "grad_norm": 0.22946473956108093, "learning_rate": 9.847852556815856e-05, "loss": 0.0393, "step": 12480 }, { "grad_norm": 0.29625970125198364, "learning_rate": 9.847447502128067e-05, "loss": 0.0415, "step": 12490 }, { "grad_norm": 0.28697478771209717, "learning_rate": 9.847041917330708e-05, "loss": 0.0354, "step": 12500 }, { "grad_norm": 0.26451101899147034, "learning_rate": 9.846635802468132e-05, "loss": 0.0399, "step": 12510 }, { "grad_norm": 0.27140721678733826, "learning_rate": 9.84622915758475e-05, "loss": 0.0355, "step": 12520 }, { "grad_norm": 0.3051184415817261, "learning_rate": 9.845821982725034e-05, "loss": 0.0426, "step": 12530 }, { "grad_norm": 0.3038599193096161, "learning_rate": 9.845414277933514e-05, "loss": 0.0425, "step": 12540 }, { "grad_norm": 0.3647308349609375, "learning_rate": 9.845006043254771e-05, "loss": 0.0403, "step": 12550 }, { "grad_norm": 0.4311656355857849, "learning_rate": 9.844597278733451e-05, "loss": 0.0472, "step": 12560 }, { "grad_norm": 0.3657476603984833, "learning_rate": 9.844187984414259e-05, "loss": 0.0494, "step": 12570 }, { "grad_norm": 0.39163219928741455, "learning_rate": 9.84377816034195e-05, "loss": 0.046, "step": 12580 }, { "grad_norm": 0.3937895894050598, "learning_rate": 9.843367806561345e-05, "loss": 0.0442, "step": 12590 }, { "grad_norm": 0.3524342179298401, "learning_rate": 9.842956923117317e-05, "loss": 0.0524, "step": 12600 }, { "grad_norm": 0.3308761417865753, "learning_rate": 9.842545510054802e-05, "loss": 0.047, "step": 12610 }, { "grad_norm": 0.44375571608543396, "learning_rate": 9.842133567418792e-05, "loss": 0.0437, "step": 12620 }, { "grad_norm": 0.40669602155685425, "learning_rate": 9.841721095254333e-05, "loss": 0.0499, "step": 12630 }, { "grad_norm": 0.41039973497390747, "learning_rate": 9.841308093606537e-05, "loss": 0.0441, "step": 12640 }, { "grad_norm": 0.3247970938682556, "learning_rate": 9.840894562520565e-05, "loss": 0.0485, "step": 12650 }, { "grad_norm": 0.3355473279953003, "learning_rate": 9.840480502041642e-05, "loss": 0.0481, "step": 12660 }, { "grad_norm": 0.3597879111766815, "learning_rate": 9.840065912215049e-05, "loss": 0.0455, "step": 12670 }, { "grad_norm": 0.37918657064437866, "learning_rate": 9.839650793086124e-05, "loss": 0.0465, "step": 12680 }, { "grad_norm": 0.3330017030239105, "learning_rate": 9.839235144700265e-05, "loss": 0.0492, "step": 12690 }, { "grad_norm": 0.3832281827926636, "learning_rate": 9.838818967102926e-05, "loss": 0.0477, "step": 12700 }, { "grad_norm": 0.3111664950847626, "learning_rate": 9.83840226033962e-05, "loss": 0.048, "step": 12710 }, { "grad_norm": 0.3590354919433594, "learning_rate": 9.837985024455918e-05, "loss": 0.0405, "step": 12720 }, { "grad_norm": 0.30994194746017456, "learning_rate": 9.837567259497447e-05, "loss": 0.0442, "step": 12730 }, { "grad_norm": 0.3006303310394287, "learning_rate": 9.837148965509894e-05, "loss": 0.049, "step": 12740 }, { "grad_norm": 0.4070906341075897, "learning_rate": 9.836730142539001e-05, "loss": 0.0581, "step": 12750 }, { "grad_norm": 0.5396161079406738, "learning_rate": 9.836310790630574e-05, "loss": 0.0537, "step": 12760 }, { "grad_norm": 0.24693572521209717, "learning_rate": 9.83589090983047e-05, "loss": 0.0515, "step": 12770 }, { "grad_norm": 0.3619064688682556, "learning_rate": 9.835470500184605e-05, "loss": 0.0487, "step": 12780 }, { "grad_norm": 0.40405720472335815, "learning_rate": 9.835049561738957e-05, "loss": 0.0601, "step": 12790 }, { "grad_norm": 0.33164408802986145, "learning_rate": 9.834628094539558e-05, "loss": 0.0504, "step": 12800 }, { "grad_norm": 0.3678898513317108, "learning_rate": 9.834206098632499e-05, "loss": 0.0549, "step": 12810 }, { "grad_norm": 0.3773169219493866, "learning_rate": 9.833783574063931e-05, "loss": 0.0454, "step": 12820 }, { "grad_norm": 0.3440993130207062, "learning_rate": 9.833360520880058e-05, "loss": 0.0474, "step": 12830 }, { "grad_norm": 0.3673279583454132, "learning_rate": 9.832936939127144e-05, "loss": 0.0457, "step": 12840 }, { "grad_norm": 0.31010702252388, "learning_rate": 9.832512828851515e-05, "loss": 0.0441, "step": 12850 }, { "grad_norm": 0.33105334639549255, "learning_rate": 9.832088190099546e-05, "loss": 0.0467, "step": 12860 }, { "grad_norm": 0.32479968667030334, "learning_rate": 9.831663022917679e-05, "loss": 0.043, "step": 12870 }, { "grad_norm": 0.3674417734146118, "learning_rate": 9.831237327352407e-05, "loss": 0.0478, "step": 12880 }, { "grad_norm": 0.27304181456565857, "learning_rate": 9.830811103450286e-05, "loss": 0.0431, "step": 12890 }, { "grad_norm": 0.3449183702468872, "learning_rate": 9.830384351257924e-05, "loss": 0.0449, "step": 12900 }, { "grad_norm": 0.35488030314445496, "learning_rate": 9.829957070821993e-05, "loss": 0.045, "step": 12910 }, { "grad_norm": 0.43262988328933716, "learning_rate": 9.829529262189218e-05, "loss": 0.0426, "step": 12920 }, { "grad_norm": 0.3382638990879059, "learning_rate": 9.829100925406385e-05, "loss": 0.0435, "step": 12930 }, { "grad_norm": 0.3721224069595337, "learning_rate": 9.828672060520333e-05, "loss": 0.0418, "step": 12940 }, { "grad_norm": 0.3006106913089752, "learning_rate": 9.828242667577966e-05, "loss": 0.046, "step": 12950 }, { "grad_norm": 0.33593037724494934, "learning_rate": 9.82781274662624e-05, "loss": 0.048, "step": 12960 }, { "grad_norm": 0.30217182636260986, "learning_rate": 9.82738229771217e-05, "loss": 0.0427, "step": 12970 }, { "grad_norm": 0.23426881432533264, "learning_rate": 9.826951320882829e-05, "loss": 0.0384, "step": 12980 }, { "grad_norm": 0.28405696153640747, "learning_rate": 9.826519816185351e-05, "loss": 0.0466, "step": 12990 }, { "grad_norm": 0.36403951048851013, "learning_rate": 9.826087783666921e-05, "loss": 0.0457, "step": 13000 }, { "grad_norm": 0.2684200704097748, "learning_rate": 9.825655223374787e-05, "loss": 0.0371, "step": 13010 }, { "grad_norm": 0.3209739029407501, "learning_rate": 9.825222135356253e-05, "loss": 0.0384, "step": 13020 }, { "grad_norm": 0.2540209889411926, "learning_rate": 9.82478851965868e-05, "loss": 0.037, "step": 13030 }, { "grad_norm": 0.3644986152648926, "learning_rate": 9.82435437632949e-05, "loss": 0.0424, "step": 13040 }, { "grad_norm": 0.3495578467845917, "learning_rate": 9.823919705416158e-05, "loss": 0.0436, "step": 13050 }, { "grad_norm": 0.6231338977813721, "learning_rate": 9.82348450696622e-05, "loss": 0.0477, "step": 13060 }, { "grad_norm": 0.2772665023803711, "learning_rate": 9.823048781027268e-05, "loss": 0.0441, "step": 13070 }, { "grad_norm": 0.4475894868373871, "learning_rate": 9.822612527646953e-05, "loss": 0.0443, "step": 13080 }, { "grad_norm": 0.4276920557022095, "learning_rate": 9.822175746872984e-05, "loss": 0.0488, "step": 13090 }, { "grad_norm": 0.31993791460990906, "learning_rate": 9.821738438753123e-05, "loss": 0.0563, "step": 13100 }, { "grad_norm": 0.31398528814315796, "learning_rate": 9.821300603335196e-05, "loss": 0.0489, "step": 13110 }, { "grad_norm": 0.37714216113090515, "learning_rate": 9.820862240667085e-05, "loss": 0.0444, "step": 13120 }, { "grad_norm": 0.32742947340011597, "learning_rate": 9.820423350796726e-05, "loss": 0.0439, "step": 13130 }, { "grad_norm": 0.3170934319496155, "learning_rate": 9.819983933772118e-05, "loss": 0.0427, "step": 13140 }, { "grad_norm": 0.31373468041419983, "learning_rate": 9.819543989641314e-05, "loss": 0.0429, "step": 13150 }, { "grad_norm": 0.3484460115432739, "learning_rate": 9.819103518452423e-05, "loss": 0.0412, "step": 13160 }, { "grad_norm": 0.23133669793605804, "learning_rate": 9.818662520253618e-05, "loss": 0.0448, "step": 13170 }, { "grad_norm": 0.3155452311038971, "learning_rate": 9.818220995093126e-05, "loss": 0.0425, "step": 13180 }, { "grad_norm": 0.3352014422416687, "learning_rate": 9.817778943019228e-05, "loss": 0.0412, "step": 13190 }, { "grad_norm": 0.41701439023017883, "learning_rate": 9.81733636408027e-05, "loss": 0.0431, "step": 13200 }, { "grad_norm": 0.2569778561592102, "learning_rate": 9.816893258324649e-05, "loss": 0.0385, "step": 13210 }, { "grad_norm": 0.4296848773956299, "learning_rate": 9.816449625800823e-05, "loss": 0.0448, "step": 13220 }, { "grad_norm": 0.5203623175621033, "learning_rate": 9.816005466557308e-05, "loss": 0.0404, "step": 13230 }, { "grad_norm": 0.27819952368736267, "learning_rate": 9.815560780642674e-05, "loss": 0.0418, "step": 13240 }, { "grad_norm": 0.3332388699054718, "learning_rate": 9.815115568105555e-05, "loss": 0.0452, "step": 13250 }, { "grad_norm": 0.3823072910308838, "learning_rate": 9.814669828994638e-05, "loss": 0.0447, "step": 13260 }, { "grad_norm": 0.37948885560035706, "learning_rate": 9.814223563358665e-05, "loss": 0.0428, "step": 13270 }, { "grad_norm": 0.3601579964160919, "learning_rate": 9.813776771246443e-05, "loss": 0.0465, "step": 13280 }, { "grad_norm": 0.3459244966506958, "learning_rate": 9.813329452706829e-05, "loss": 0.0416, "step": 13290 }, { "grad_norm": 0.4017314314842224, "learning_rate": 9.812881607788744e-05, "loss": 0.0462, "step": 13300 }, { "grad_norm": 0.2650337815284729, "learning_rate": 9.812433236541163e-05, "loss": 0.04, "step": 13310 }, { "grad_norm": 0.26830723881721497, "learning_rate": 9.811984339013116e-05, "loss": 0.0369, "step": 13320 }, { "grad_norm": 0.6776134967803955, "learning_rate": 9.811534915253698e-05, "loss": 0.0422, "step": 13330 }, { "grad_norm": 0.19620148837566376, "learning_rate": 9.811084965312056e-05, "loss": 0.0437, "step": 13340 }, { "grad_norm": 0.30835872888565063, "learning_rate": 9.810634489237396e-05, "loss": 0.0412, "step": 13350 }, { "grad_norm": 0.33312585949897766, "learning_rate": 9.81018348707898e-05, "loss": 0.0428, "step": 13360 }, { "grad_norm": 0.2803276479244232, "learning_rate": 9.809731958886131e-05, "loss": 0.047, "step": 13370 }, { "grad_norm": 0.40544116497039795, "learning_rate": 9.809279904708224e-05, "loss": 0.0606, "step": 13380 }, { "grad_norm": 0.4326684772968292, "learning_rate": 9.808827324594699e-05, "loss": 0.0429, "step": 13390 }, { "grad_norm": 0.3641364276409149, "learning_rate": 9.808374218595046e-05, "loss": 0.041, "step": 13400 }, { "grad_norm": 0.30097314715385437, "learning_rate": 9.80792058675882e-05, "loss": 0.0463, "step": 13410 }, { "grad_norm": 0.3927001357078552, "learning_rate": 9.807466429135627e-05, "loss": 0.043, "step": 13420 }, { "grad_norm": 0.3377230763435364, "learning_rate": 9.807011745775132e-05, "loss": 0.0391, "step": 13430 }, { "grad_norm": 0.29042190313339233, "learning_rate": 9.806556536727061e-05, "loss": 0.0397, "step": 13440 }, { "grad_norm": 0.2990545630455017, "learning_rate": 9.806100802041193e-05, "loss": 0.0395, "step": 13450 }, { "grad_norm": 0.32782548666000366, "learning_rate": 9.805644541767368e-05, "loss": 0.0397, "step": 13460 }, { "grad_norm": 0.29837584495544434, "learning_rate": 9.805187755955478e-05, "loss": 0.0407, "step": 13470 }, { "grad_norm": 0.36222872138023376, "learning_rate": 9.804730444655483e-05, "loss": 0.0434, "step": 13480 }, { "grad_norm": 0.4232427477836609, "learning_rate": 9.804272607917388e-05, "loss": 0.039, "step": 13490 }, { "grad_norm": 0.36549925804138184, "learning_rate": 9.803814245791265e-05, "loss": 0.047, "step": 13500 }, { "grad_norm": 0.30652666091918945, "learning_rate": 9.803355358327239e-05, "loss": 0.0471, "step": 13510 }, { "grad_norm": 0.28278619050979614, "learning_rate": 9.802895945575492e-05, "loss": 0.0427, "step": 13520 }, { "grad_norm": 0.2941766381263733, "learning_rate": 9.802436007586266e-05, "loss": 0.042, "step": 13530 }, { "grad_norm": 0.37510406970977783, "learning_rate": 9.801975544409858e-05, "loss": 0.0446, "step": 13540 }, { "grad_norm": 0.36560603976249695, "learning_rate": 9.801514556096625e-05, "loss": 0.0448, "step": 13550 }, { "grad_norm": 0.3470667898654938, "learning_rate": 9.801053042696977e-05, "loss": 0.0489, "step": 13560 }, { "grad_norm": 0.3087363839149475, "learning_rate": 9.800591004261388e-05, "loss": 0.041, "step": 13570 }, { "grad_norm": 0.5995907783508301, "learning_rate": 9.800128440840385e-05, "loss": 0.0411, "step": 13580 }, { "grad_norm": 0.3070255517959595, "learning_rate": 9.799665352484552e-05, "loss": 0.0448, "step": 13590 }, { "grad_norm": 0.39407697319984436, "learning_rate": 9.799201739244532e-05, "loss": 0.0465, "step": 13600 }, { "grad_norm": 0.3065382242202759, "learning_rate": 9.798737601171025e-05, "loss": 0.0472, "step": 13610 }, { "grad_norm": 0.31503936648368835, "learning_rate": 9.79827293831479e-05, "loss": 0.0345, "step": 13620 }, { "grad_norm": 0.3045990467071533, "learning_rate": 9.797807750726638e-05, "loss": 0.0367, "step": 13630 }, { "grad_norm": 0.29756850004196167, "learning_rate": 9.797342038457446e-05, "loss": 0.0427, "step": 13640 }, { "grad_norm": 0.32596078515052795, "learning_rate": 9.796875801558141e-05, "loss": 0.042, "step": 13650 }, { "grad_norm": 0.26758578419685364, "learning_rate": 9.79640904007971e-05, "loss": 0.0403, "step": 13660 }, { "grad_norm": 0.30753469467163086, "learning_rate": 9.795941754073199e-05, "loss": 0.0513, "step": 13670 }, { "grad_norm": 0.5452101230621338, "learning_rate": 9.795473943589705e-05, "loss": 0.0465, "step": 13680 }, { "grad_norm": 0.3589220643043518, "learning_rate": 9.795005608680394e-05, "loss": 0.0423, "step": 13690 }, { "grad_norm": 0.3058546185493469, "learning_rate": 9.794536749396477e-05, "loss": 0.0419, "step": 13700 }, { "grad_norm": 0.32240965962409973, "learning_rate": 9.79406736578923e-05, "loss": 0.0415, "step": 13710 }, { "grad_norm": 0.3770883083343506, "learning_rate": 9.793597457909984e-05, "loss": 0.0431, "step": 13720 }, { "grad_norm": 0.3484858274459839, "learning_rate": 9.793127025810127e-05, "loss": 0.0418, "step": 13730 }, { "grad_norm": 0.3306919038295746, "learning_rate": 9.792656069541104e-05, "loss": 0.0462, "step": 13740 }, { "grad_norm": 0.25344008207321167, "learning_rate": 9.79218458915442e-05, "loss": 0.0365, "step": 13750 }, { "grad_norm": 0.31213629245758057, "learning_rate": 9.791712584701634e-05, "loss": 0.0369, "step": 13760 }, { "grad_norm": 0.3063492476940155, "learning_rate": 9.791240056234364e-05, "loss": 0.0416, "step": 13770 }, { "grad_norm": 0.29720476269721985, "learning_rate": 9.790767003804283e-05, "loss": 0.0442, "step": 13780 }, { "grad_norm": 0.3435783386230469, "learning_rate": 9.790293427463126e-05, "loss": 0.0468, "step": 13790 }, { "grad_norm": 0.3826882243156433, "learning_rate": 9.789819327262684e-05, "loss": 0.0459, "step": 13800 }, { "grad_norm": 0.4172306954860687, "learning_rate": 9.7893447032548e-05, "loss": 0.0435, "step": 13810 }, { "grad_norm": 0.29107484221458435, "learning_rate": 9.78886955549138e-05, "loss": 0.0433, "step": 13820 }, { "grad_norm": 0.4787232577800751, "learning_rate": 9.788393884024387e-05, "loss": 0.0432, "step": 13830 }, { "grad_norm": 0.2341480851173401, "learning_rate": 9.787917688905836e-05, "loss": 0.0466, "step": 13840 }, { "grad_norm": 0.3649059534072876, "learning_rate": 9.787440970187807e-05, "loss": 0.043, "step": 13850 }, { "grad_norm": 0.36825600266456604, "learning_rate": 9.786963727922429e-05, "loss": 0.039, "step": 13860 }, { "grad_norm": 0.3522152602672577, "learning_rate": 9.786485962161897e-05, "loss": 0.0421, "step": 13870 }, { "grad_norm": 0.2987143099308014, "learning_rate": 9.786007672958455e-05, "loss": 0.045, "step": 13880 }, { "grad_norm": 0.24955661594867706, "learning_rate": 9.78552886036441e-05, "loss": 0.0453, "step": 13890 }, { "grad_norm": 0.3717139661312103, "learning_rate": 9.785049524432124e-05, "loss": 0.0427, "step": 13900 }, { "grad_norm": 0.24686293303966522, "learning_rate": 9.784569665214016e-05, "loss": 0.0362, "step": 13910 }, { "grad_norm": 0.27299344539642334, "learning_rate": 9.784089282762563e-05, "loss": 0.045, "step": 13920 }, { "grad_norm": 0.32929345965385437, "learning_rate": 9.7836083771303e-05, "loss": 0.0415, "step": 13930 }, { "grad_norm": 0.30565375089645386, "learning_rate": 9.783126948369817e-05, "loss": 0.0371, "step": 13940 }, { "grad_norm": 0.5078455209732056, "learning_rate": 9.78264499653376e-05, "loss": 0.0399, "step": 13950 }, { "grad_norm": 0.37606942653656006, "learning_rate": 9.782162521674838e-05, "loss": 0.0411, "step": 13960 }, { "grad_norm": 0.3345758318901062, "learning_rate": 9.781679523845812e-05, "loss": 0.0438, "step": 13970 }, { "grad_norm": 0.28351733088493347, "learning_rate": 9.781196003099502e-05, "loss": 0.0487, "step": 13980 }, { "grad_norm": 0.2779580354690552, "learning_rate": 9.780711959488786e-05, "loss": 0.0404, "step": 13990 }, { "grad_norm": 0.2989889085292816, "learning_rate": 9.780227393066599e-05, "loss": 0.0388, "step": 14000 }, { "grad_norm": 0.34719017148017883, "learning_rate": 9.77974230388593e-05, "loss": 0.0491, "step": 14010 }, { "grad_norm": 0.34509024024009705, "learning_rate": 9.779256691999829e-05, "loss": 0.0449, "step": 14020 }, { "grad_norm": 0.3214462101459503, "learning_rate": 9.778770557461403e-05, "loss": 0.0414, "step": 14030 }, { "grad_norm": 0.31411662697792053, "learning_rate": 9.778283900323812e-05, "loss": 0.0394, "step": 14040 }, { "grad_norm": 0.26384952664375305, "learning_rate": 9.777796720640277e-05, "loss": 0.0342, "step": 14050 }, { "grad_norm": 0.36584505438804626, "learning_rate": 9.777309018464078e-05, "loss": 0.0348, "step": 14060 }, { "grad_norm": 0.34444546699523926, "learning_rate": 9.776820793848547e-05, "loss": 0.0383, "step": 14070 }, { "grad_norm": 0.32026076316833496, "learning_rate": 9.776332046847075e-05, "loss": 0.0368, "step": 14080 }, { "grad_norm": 0.3494466245174408, "learning_rate": 9.775842777513111e-05, "loss": 0.0356, "step": 14090 }, { "grad_norm": 0.30349305272102356, "learning_rate": 9.775352985900163e-05, "loss": 0.0369, "step": 14100 }, { "grad_norm": 0.27523866295814514, "learning_rate": 9.774862672061791e-05, "loss": 0.0358, "step": 14110 }, { "grad_norm": 0.26790687441825867, "learning_rate": 9.774371836051616e-05, "loss": 0.0362, "step": 14120 }, { "grad_norm": 0.3408817648887634, "learning_rate": 9.773880477923315e-05, "loss": 0.0345, "step": 14130 }, { "grad_norm": 0.3623236417770386, "learning_rate": 9.773388597730623e-05, "loss": 0.0367, "step": 14140 }, { "grad_norm": 0.23445779085159302, "learning_rate": 9.77289619552733e-05, "loss": 0.0434, "step": 14150 }, { "grad_norm": 0.25980454683303833, "learning_rate": 9.772403271367285e-05, "loss": 0.0398, "step": 14160 }, { "grad_norm": 0.2974863648414612, "learning_rate": 9.771909825304396e-05, "loss": 0.0432, "step": 14170 }, { "grad_norm": 0.3983615040779114, "learning_rate": 9.771415857392619e-05, "loss": 0.0413, "step": 14180 }, { "grad_norm": 0.3307344615459442, "learning_rate": 9.770921367685978e-05, "loss": 0.0389, "step": 14190 }, { "grad_norm": 0.2983647882938385, "learning_rate": 9.770426356238551e-05, "loss": 0.0409, "step": 14200 }, { "grad_norm": 0.3096766471862793, "learning_rate": 9.769930823104469e-05, "loss": 0.0406, "step": 14210 }, { "grad_norm": 0.30117496848106384, "learning_rate": 9.769434768337926e-05, "loss": 0.0383, "step": 14220 }, { "grad_norm": 0.3371264338493347, "learning_rate": 9.768938191993164e-05, "loss": 0.0351, "step": 14230 }, { "grad_norm": 0.31215542554855347, "learning_rate": 9.768441094124494e-05, "loss": 0.0381, "step": 14240 }, { "grad_norm": 0.3587605059146881, "learning_rate": 9.767943474786275e-05, "loss": 0.0387, "step": 14250 }, { "grad_norm": 0.2660154104232788, "learning_rate": 9.767445334032923e-05, "loss": 0.0461, "step": 14260 }, { "grad_norm": 0.33313578367233276, "learning_rate": 9.766946671918919e-05, "loss": 0.039, "step": 14270 }, { "grad_norm": 0.29012107849121094, "learning_rate": 9.766447488498796e-05, "loss": 0.0345, "step": 14280 }, { "grad_norm": 0.27200207114219666, "learning_rate": 9.765947783827139e-05, "loss": 0.0349, "step": 14290 }, { "grad_norm": 0.3683050870895386, "learning_rate": 9.765447557958599e-05, "loss": 0.0349, "step": 14300 }, { "grad_norm": 0.2375708669424057, "learning_rate": 9.764946810947879e-05, "loss": 0.0371, "step": 14310 }, { "grad_norm": 0.38202324509620667, "learning_rate": 9.764445542849738e-05, "loss": 0.0381, "step": 14320 }, { "grad_norm": 0.27934497594833374, "learning_rate": 9.763943753718998e-05, "loss": 0.0353, "step": 14330 }, { "grad_norm": 0.3221566379070282, "learning_rate": 9.76344144361053e-05, "loss": 0.0364, "step": 14340 }, { "grad_norm": 0.31162595748901367, "learning_rate": 9.762938612579269e-05, "loss": 0.0407, "step": 14350 }, { "grad_norm": 0.28857335448265076, "learning_rate": 9.762435260680202e-05, "loss": 0.0333, "step": 14360 }, { "grad_norm": 0.3165050745010376, "learning_rate": 9.761931387968373e-05, "loss": 0.0447, "step": 14370 }, { "grad_norm": 0.3213842213153839, "learning_rate": 9.76142699449889e-05, "loss": 0.0369, "step": 14380 }, { "grad_norm": 0.23493972420692444, "learning_rate": 9.760922080326908e-05, "loss": 0.0342, "step": 14390 }, { "grad_norm": 0.3201969563961029, "learning_rate": 9.760416645507644e-05, "loss": 0.0364, "step": 14400 }, { "grad_norm": 0.3258163332939148, "learning_rate": 9.759910690096375e-05, "loss": 0.0339, "step": 14410 }, { "grad_norm": 0.26989802718162537, "learning_rate": 9.759404214148429e-05, "loss": 0.0377, "step": 14420 }, { "grad_norm": 0.3168519139289856, "learning_rate": 9.758897217719191e-05, "loss": 0.0376, "step": 14430 }, { "grad_norm": 0.32964038848876953, "learning_rate": 9.758389700864113e-05, "loss": 0.0488, "step": 14440 }, { "grad_norm": 0.3428014814853668, "learning_rate": 9.757881663638688e-05, "loss": 0.0451, "step": 14450 }, { "grad_norm": 0.35265693068504333, "learning_rate": 9.757373106098478e-05, "loss": 0.0428, "step": 14460 }, { "grad_norm": 0.35460400581359863, "learning_rate": 9.756864028299097e-05, "loss": 0.049, "step": 14470 }, { "grad_norm": 0.34426581859588623, "learning_rate": 9.75635443029622e-05, "loss": 0.0468, "step": 14480 }, { "grad_norm": 0.28388798236846924, "learning_rate": 9.755844312145572e-05, "loss": 0.0381, "step": 14490 }, { "grad_norm": 0.4677591919898987, "learning_rate": 9.755333673902941e-05, "loss": 0.0467, "step": 14500 }, { "grad_norm": 0.35614895820617676, "learning_rate": 9.75482251562417e-05, "loss": 0.0477, "step": 14510 }, { "grad_norm": 0.3327988386154175, "learning_rate": 9.754310837365155e-05, "loss": 0.0388, "step": 14520 }, { "grad_norm": 0.2636396586894989, "learning_rate": 9.753798639181856e-05, "loss": 0.0393, "step": 14530 }, { "grad_norm": 0.28590333461761475, "learning_rate": 9.753285921130286e-05, "loss": 0.0362, "step": 14540 }, { "grad_norm": 0.24828217923641205, "learning_rate": 9.752772683266512e-05, "loss": 0.0391, "step": 14550 }, { "grad_norm": 0.30948662757873535, "learning_rate": 9.752258925646665e-05, "loss": 0.0422, "step": 14560 }, { "grad_norm": 0.22081328928470612, "learning_rate": 9.751744648326926e-05, "loss": 0.0341, "step": 14570 }, { "grad_norm": 0.27760156989097595, "learning_rate": 9.751229851363536e-05, "loss": 0.0352, "step": 14580 }, { "grad_norm": 0.3646862506866455, "learning_rate": 9.750714534812793e-05, "loss": 0.0419, "step": 14590 }, { "grad_norm": 0.3551965355873108, "learning_rate": 9.750198698731053e-05, "loss": 0.0414, "step": 14600 }, { "grad_norm": 0.31272247433662415, "learning_rate": 9.749682343174722e-05, "loss": 0.0369, "step": 14610 }, { "grad_norm": 0.3472641408443451, "learning_rate": 9.749165468200272e-05, "loss": 0.0462, "step": 14620 }, { "grad_norm": 0.3234409987926483, "learning_rate": 9.748648073864229e-05, "loss": 0.0419, "step": 14630 }, { "grad_norm": 0.5036682486534119, "learning_rate": 9.748130160223168e-05, "loss": 0.0425, "step": 14640 }, { "grad_norm": 0.30399882793426514, "learning_rate": 9.747611727333734e-05, "loss": 0.0417, "step": 14650 }, { "grad_norm": 0.29654037952423096, "learning_rate": 9.74709277525262e-05, "loss": 0.0418, "step": 14660 }, { "grad_norm": 0.36896735429763794, "learning_rate": 9.746573304036576e-05, "loss": 0.0457, "step": 14670 }, { "grad_norm": 0.2737372815608978, "learning_rate": 9.746053313742412e-05, "loss": 0.0398, "step": 14680 }, { "grad_norm": 0.3128187656402588, "learning_rate": 9.745532804426994e-05, "loss": 0.0438, "step": 14690 }, { "grad_norm": 0.31675103306770325, "learning_rate": 9.745011776147242e-05, "loss": 0.0421, "step": 14700 }, { "grad_norm": 0.29308465123176575, "learning_rate": 9.744490228960138e-05, "loss": 0.0436, "step": 14710 }, { "grad_norm": 0.31906744837760925, "learning_rate": 9.743968162922713e-05, "loss": 0.043, "step": 14720 }, { "grad_norm": 0.2359563559293747, "learning_rate": 9.743445578092064e-05, "loss": 0.0444, "step": 14730 }, { "grad_norm": 0.2993040382862091, "learning_rate": 9.742922474525338e-05, "loss": 0.043, "step": 14740 }, { "grad_norm": 0.3539595901966095, "learning_rate": 9.742398852279741e-05, "loss": 0.0395, "step": 14750 }, { "grad_norm": 0.4124974310398102, "learning_rate": 9.741874711412535e-05, "loss": 0.0432, "step": 14760 }, { "grad_norm": 0.264922171831131, "learning_rate": 9.741350051981042e-05, "loss": 0.0391, "step": 14770 }, { "grad_norm": 0.2911969721317291, "learning_rate": 9.740824874042633e-05, "loss": 0.0335, "step": 14780 }, { "grad_norm": 0.33882424235343933, "learning_rate": 9.740299177654746e-05, "loss": 0.0394, "step": 14790 }, { "grad_norm": 0.27414512634277344, "learning_rate": 9.739772962874867e-05, "loss": 0.0395, "step": 14800 }, { "grad_norm": 0.41163355112075806, "learning_rate": 9.739246229760541e-05, "loss": 0.0453, "step": 14810 }, { "grad_norm": 0.33393022418022156, "learning_rate": 9.738718978369376e-05, "loss": 0.0437, "step": 14820 }, { "grad_norm": 0.2469150573015213, "learning_rate": 9.738191208759025e-05, "loss": 0.0445, "step": 14830 }, { "grad_norm": 0.25026094913482666, "learning_rate": 9.73766292098721e-05, "loss": 0.0367, "step": 14840 }, { "grad_norm": 0.3689064085483551, "learning_rate": 9.737134115111699e-05, "loss": 0.0425, "step": 14850 }, { "grad_norm": 0.36653080582618713, "learning_rate": 9.736604791190323e-05, "loss": 0.0426, "step": 14860 }, { "grad_norm": 0.32297205924987793, "learning_rate": 9.73607494928097e-05, "loss": 0.0383, "step": 14870 }, { "grad_norm": 0.32639279961586, "learning_rate": 9.735544589441581e-05, "loss": 0.0378, "step": 14880 }, { "grad_norm": 0.2713908851146698, "learning_rate": 9.735013711730154e-05, "loss": 0.0399, "step": 14890 }, { "grad_norm": 0.3423907458782196, "learning_rate": 9.734482316204747e-05, "loss": 0.0373, "step": 14900 }, { "grad_norm": 0.2627764642238617, "learning_rate": 9.733950402923473e-05, "loss": 0.0332, "step": 14910 }, { "grad_norm": 0.33867692947387695, "learning_rate": 9.7334179719445e-05, "loss": 0.0365, "step": 14920 }, { "grad_norm": 0.2989160418510437, "learning_rate": 9.732885023326053e-05, "loss": 0.0344, "step": 14930 }, { "grad_norm": 0.29444175958633423, "learning_rate": 9.732351557126418e-05, "loss": 0.032, "step": 14940 }, { "grad_norm": 0.3155304193496704, "learning_rate": 9.731817573403929e-05, "loss": 0.0389, "step": 14950 }, { "grad_norm": 0.29310914874076843, "learning_rate": 9.731283072216985e-05, "loss": 0.0346, "step": 14960 }, { "grad_norm": 0.38362860679626465, "learning_rate": 9.730748053624039e-05, "loss": 0.038, "step": 14970 }, { "grad_norm": 0.34273117780685425, "learning_rate": 9.730212517683598e-05, "loss": 0.038, "step": 14980 }, { "grad_norm": 0.3255550265312195, "learning_rate": 9.729676464454228e-05, "loss": 0.0365, "step": 14990 }, { "grad_norm": 0.3577669560909271, "learning_rate": 9.72913989399455e-05, "loss": 0.0411, "step": 15000 }, { "grad_norm": 0.27556443214416504, "learning_rate": 9.728602806363242e-05, "loss": 0.037, "step": 15010 }, { "grad_norm": 0.33191055059432983, "learning_rate": 9.728065201619043e-05, "loss": 0.0381, "step": 15020 }, { "grad_norm": 0.6770281195640564, "learning_rate": 9.727527079820742e-05, "loss": 0.0433, "step": 15030 }, { "grad_norm": 0.3558996319770813, "learning_rate": 9.726988441027186e-05, "loss": 0.0433, "step": 15040 }, { "grad_norm": 0.30724427103996277, "learning_rate": 9.726449285297281e-05, "loss": 0.051, "step": 15050 }, { "grad_norm": 0.48862046003341675, "learning_rate": 9.72590961268999e-05, "loss": 0.0472, "step": 15060 }, { "grad_norm": 0.2842896580696106, "learning_rate": 9.725369423264328e-05, "loss": 0.0381, "step": 15070 }, { "grad_norm": 0.3528073728084564, "learning_rate": 9.72482871707937e-05, "loss": 0.0494, "step": 15080 }, { "grad_norm": 0.4563453495502472, "learning_rate": 9.724287494194247e-05, "loss": 0.0525, "step": 15090 }, { "grad_norm": 0.35613754391670227, "learning_rate": 9.723745754668147e-05, "loss": 0.0449, "step": 15100 }, { "grad_norm": 0.42017558217048645, "learning_rate": 9.723203498560313e-05, "loss": 0.0471, "step": 15110 }, { "grad_norm": 0.3069661855697632, "learning_rate": 9.722660725930046e-05, "loss": 0.0452, "step": 15120 }, { "grad_norm": 0.3419007360935211, "learning_rate": 9.722117436836702e-05, "loss": 0.054, "step": 15130 }, { "grad_norm": 0.43757498264312744, "learning_rate": 9.721573631339696e-05, "loss": 0.0503, "step": 15140 }, { "grad_norm": 0.30898433923721313, "learning_rate": 9.721029309498494e-05, "loss": 0.044, "step": 15150 }, { "grad_norm": 0.29291731119155884, "learning_rate": 9.720484471372627e-05, "loss": 0.0425, "step": 15160 }, { "grad_norm": 0.32245397567749023, "learning_rate": 9.719939117021673e-05, "loss": 0.048, "step": 15170 }, { "grad_norm": 0.5239185690879822, "learning_rate": 9.719393246505275e-05, "loss": 0.0434, "step": 15180 }, { "grad_norm": 0.40993258357048035, "learning_rate": 9.718846859883128e-05, "loss": 0.048, "step": 15190 }, { "grad_norm": 0.3054550886154175, "learning_rate": 9.718299957214982e-05, "loss": 0.0468, "step": 15200 }, { "grad_norm": 0.2557492256164551, "learning_rate": 9.717752538560646e-05, "loss": 0.0386, "step": 15210 }, { "grad_norm": 0.3751221299171448, "learning_rate": 9.717204603979986e-05, "loss": 0.0432, "step": 15220 }, { "grad_norm": 0.4310876131057739, "learning_rate": 9.716656153532922e-05, "loss": 0.0436, "step": 15230 }, { "grad_norm": 0.35638296604156494, "learning_rate": 9.716107187279434e-05, "loss": 0.0408, "step": 15240 }, { "grad_norm": 0.3280402421951294, "learning_rate": 9.715557705279555e-05, "loss": 0.038, "step": 15250 }, { "grad_norm": 0.3469272553920746, "learning_rate": 9.715007707593372e-05, "loss": 0.0389, "step": 15260 }, { "grad_norm": 0.39220699667930603, "learning_rate": 9.714457194281036e-05, "loss": 0.0383, "step": 15270 }, { "grad_norm": 0.30255329608917236, "learning_rate": 9.713906165402751e-05, "loss": 0.0407, "step": 15280 }, { "grad_norm": 0.28956538438796997, "learning_rate": 9.713354621018774e-05, "loss": 0.0418, "step": 15290 }, { "grad_norm": 0.2887735664844513, "learning_rate": 9.712802561189422e-05, "loss": 0.0425, "step": 15300 }, { "grad_norm": 0.31137773394584656, "learning_rate": 9.712249985975069e-05, "loss": 0.0374, "step": 15310 }, { "grad_norm": 0.359786719083786, "learning_rate": 9.71169689543614e-05, "loss": 0.0417, "step": 15320 }, { "grad_norm": 0.29800689220428467, "learning_rate": 9.711143289633123e-05, "loss": 0.0394, "step": 15330 }, { "grad_norm": 0.3243056833744049, "learning_rate": 9.710589168626561e-05, "loss": 0.0382, "step": 15340 }, { "grad_norm": 0.2861693501472473, "learning_rate": 9.710034532477048e-05, "loss": 0.0451, "step": 15350 }, { "grad_norm": 0.30342501401901245, "learning_rate": 9.709479381245239e-05, "loss": 0.0406, "step": 15360 }, { "grad_norm": 0.2910974621772766, "learning_rate": 9.708923714991847e-05, "loss": 0.0367, "step": 15370 }, { "grad_norm": 0.19377994537353516, "learning_rate": 9.708367533777638e-05, "loss": 0.0407, "step": 15380 }, { "grad_norm": 0.2752709686756134, "learning_rate": 9.707810837663431e-05, "loss": 0.0349, "step": 15390 }, { "grad_norm": 0.282927542924881, "learning_rate": 9.707253626710113e-05, "loss": 0.0344, "step": 15400 }, { "grad_norm": 0.364013135433197, "learning_rate": 9.706695900978613e-05, "loss": 0.0461, "step": 15410 }, { "grad_norm": 0.286006897687912, "learning_rate": 9.706137660529926e-05, "loss": 0.0422, "step": 15420 }, { "grad_norm": 0.3020152449607849, "learning_rate": 9.705578905425101e-05, "loss": 0.0378, "step": 15430 }, { "grad_norm": 0.3428325653076172, "learning_rate": 9.705019635725241e-05, "loss": 0.0437, "step": 15440 }, { "grad_norm": 0.3040279150009155, "learning_rate": 9.704459851491508e-05, "loss": 0.0332, "step": 15450 }, { "grad_norm": 0.2948179841041565, "learning_rate": 9.703899552785118e-05, "loss": 0.0374, "step": 15460 }, { "grad_norm": 0.2740309238433838, "learning_rate": 9.703338739667346e-05, "loss": 0.0361, "step": 15470 }, { "grad_norm": 0.3313779830932617, "learning_rate": 9.70277741219952e-05, "loss": 0.0437, "step": 15480 }, { "grad_norm": 0.2783018946647644, "learning_rate": 9.702215570443027e-05, "loss": 0.0395, "step": 15490 }, { "grad_norm": 0.27798205614089966, "learning_rate": 9.701653214459309e-05, "loss": 0.0389, "step": 15500 }, { "grad_norm": 0.29930153489112854, "learning_rate": 9.701090344309865e-05, "loss": 0.0391, "step": 15510 }, { "grad_norm": 0.36233535408973694, "learning_rate": 9.700526960056247e-05, "loss": 0.0369, "step": 15520 }, { "grad_norm": 0.2475321888923645, "learning_rate": 9.699963061760068e-05, "loss": 0.0423, "step": 15530 }, { "grad_norm": 0.25724267959594727, "learning_rate": 9.699398649482997e-05, "loss": 0.0397, "step": 15540 }, { "grad_norm": 0.3061915636062622, "learning_rate": 9.698833723286753e-05, "loss": 0.037, "step": 15550 }, { "grad_norm": 0.35624629259109497, "learning_rate": 9.698268283233118e-05, "loss": 0.0454, "step": 15560 }, { "grad_norm": 0.5295356512069702, "learning_rate": 9.697702329383929e-05, "loss": 0.0487, "step": 15570 }, { "grad_norm": 0.4273187816143036, "learning_rate": 9.697135861801074e-05, "loss": 0.0439, "step": 15580 }, { "grad_norm": 0.41410693526268005, "learning_rate": 9.696568880546505e-05, "loss": 0.046, "step": 15590 }, { "grad_norm": 0.3242546319961548, "learning_rate": 9.696001385682223e-05, "loss": 0.0398, "step": 15600 }, { "grad_norm": 0.2738421559333801, "learning_rate": 9.695433377270291e-05, "loss": 0.0443, "step": 15610 }, { "grad_norm": 0.3312127888202667, "learning_rate": 9.694864855372824e-05, "loss": 0.0363, "step": 15620 }, { "grad_norm": 0.31363412737846375, "learning_rate": 9.694295820051995e-05, "loss": 0.0439, "step": 15630 }, { "grad_norm": 0.40837109088897705, "learning_rate": 9.693726271370032e-05, "loss": 0.0466, "step": 15640 }, { "grad_norm": 0.33354783058166504, "learning_rate": 9.693156209389221e-05, "loss": 0.0403, "step": 15650 }, { "grad_norm": 0.30564066767692566, "learning_rate": 9.692585634171905e-05, "loss": 0.0394, "step": 15660 }, { "grad_norm": 0.2769453525543213, "learning_rate": 9.692014545780476e-05, "loss": 0.0371, "step": 15670 }, { "grad_norm": 0.27610012888908386, "learning_rate": 9.691442944277393e-05, "loss": 0.0395, "step": 15680 }, { "grad_norm": 0.23722948133945465, "learning_rate": 9.690870829725162e-05, "loss": 0.0366, "step": 15690 }, { "grad_norm": 0.30984944105148315, "learning_rate": 9.69029820218635e-05, "loss": 0.036, "step": 15700 }, { "grad_norm": 0.35883980989456177, "learning_rate": 9.689725061723579e-05, "loss": 0.0441, "step": 15710 }, { "grad_norm": 0.3098337948322296, "learning_rate": 9.689151408399527e-05, "loss": 0.0438, "step": 15720 }, { "grad_norm": 0.31170326471328735, "learning_rate": 9.688577242276924e-05, "loss": 0.0402, "step": 15730 }, { "grad_norm": 0.3501620590686798, "learning_rate": 9.688002563418566e-05, "loss": 0.0382, "step": 15740 }, { "grad_norm": 0.3079439699649811, "learning_rate": 9.687427371887293e-05, "loss": 0.0397, "step": 15750 }, { "grad_norm": 0.3149685263633728, "learning_rate": 9.686851667746012e-05, "loss": 0.0427, "step": 15760 }, { "grad_norm": 0.2577701807022095, "learning_rate": 9.686275451057677e-05, "loss": 0.041, "step": 15770 }, { "grad_norm": 0.4278131127357483, "learning_rate": 9.685698721885308e-05, "loss": 0.0464, "step": 15780 }, { "grad_norm": 0.2979361116886139, "learning_rate": 9.68512148029197e-05, "loss": 0.0373, "step": 15790 }, { "grad_norm": 0.2716276943683624, "learning_rate": 9.684543726340791e-05, "loss": 0.0414, "step": 15800 }, { "grad_norm": 0.32571297883987427, "learning_rate": 9.683965460094952e-05, "loss": 0.0397, "step": 15810 }, { "grad_norm": 0.24582278728485107, "learning_rate": 9.683386681617694e-05, "loss": 0.0381, "step": 15820 }, { "grad_norm": 0.25893932580947876, "learning_rate": 9.68280739097231e-05, "loss": 0.0367, "step": 15830 }, { "grad_norm": 0.26026397943496704, "learning_rate": 9.682227588222148e-05, "loss": 0.0386, "step": 15840 }, { "grad_norm": 0.2967575490474701, "learning_rate": 9.681647273430618e-05, "loss": 0.0341, "step": 15850 }, { "grad_norm": 0.37126436829566956, "learning_rate": 9.681066446661182e-05, "loss": 0.039, "step": 15860 }, { "grad_norm": 0.38119733333587646, "learning_rate": 9.680485107977357e-05, "loss": 0.0396, "step": 15870 }, { "grad_norm": 0.3044896125793457, "learning_rate": 9.679903257442716e-05, "loss": 0.0381, "step": 15880 }, { "grad_norm": 0.27082008123397827, "learning_rate": 9.679320895120891e-05, "loss": 0.04, "step": 15890 }, { "grad_norm": 0.36583465337753296, "learning_rate": 9.67873802107557e-05, "loss": 0.0418, "step": 15900 }, { "grad_norm": 0.5865141153335571, "learning_rate": 9.67815463537049e-05, "loss": 0.0589, "step": 15910 }, { "grad_norm": 0.24052655696868896, "learning_rate": 9.677570738069457e-05, "loss": 0.0422, "step": 15920 }, { "grad_norm": 0.265798419713974, "learning_rate": 9.676986329236318e-05, "loss": 0.0435, "step": 15930 }, { "grad_norm": 0.3752918541431427, "learning_rate": 9.676401408934987e-05, "loss": 0.041, "step": 15940 }, { "grad_norm": 0.2592118978500366, "learning_rate": 9.675815977229428e-05, "loss": 0.0368, "step": 15950 }, { "grad_norm": 0.3272685110569, "learning_rate": 9.675230034183664e-05, "loss": 0.0425, "step": 15960 }, { "grad_norm": 0.28791409730911255, "learning_rate": 9.674643579861773e-05, "loss": 0.0354, "step": 15970 }, { "grad_norm": 0.30072981119155884, "learning_rate": 9.674056614327886e-05, "loss": 0.0386, "step": 15980 }, { "grad_norm": 0.3238239586353302, "learning_rate": 9.673469137646198e-05, "loss": 0.0438, "step": 15990 }, { "grad_norm": 0.30201655626296997, "learning_rate": 9.67288114988095e-05, "loss": 0.0402, "step": 16000 }, { "grad_norm": 0.28084227442741394, "learning_rate": 9.672292651096447e-05, "loss": 0.0413, "step": 16010 }, { "grad_norm": 0.29302576184272766, "learning_rate": 9.671703641357042e-05, "loss": 0.0438, "step": 16020 }, { "grad_norm": 0.2873350977897644, "learning_rate": 9.67111412072715e-05, "loss": 0.0376, "step": 16030 }, { "grad_norm": 0.30220654606819153, "learning_rate": 9.670524089271242e-05, "loss": 0.0371, "step": 16040 }, { "grad_norm": 0.2823343276977539, "learning_rate": 9.669933547053842e-05, "loss": 0.0352, "step": 16050 }, { "grad_norm": 0.4347900450229645, "learning_rate": 9.669342494139531e-05, "loss": 0.0404, "step": 16060 }, { "grad_norm": 0.3904942572116852, "learning_rate": 9.668750930592943e-05, "loss": 0.0393, "step": 16070 }, { "grad_norm": 0.32735002040863037, "learning_rate": 9.668158856478775e-05, "loss": 0.0389, "step": 16080 }, { "grad_norm": 0.3856528401374817, "learning_rate": 9.66756627186177e-05, "loss": 0.0391, "step": 16090 }, { "grad_norm": 0.3942490816116333, "learning_rate": 9.666973176806737e-05, "loss": 0.0388, "step": 16100 }, { "grad_norm": 0.39242660999298096, "learning_rate": 9.666379571378534e-05, "loss": 0.0366, "step": 16110 }, { "grad_norm": 0.3041468858718872, "learning_rate": 9.665785455642076e-05, "loss": 0.0391, "step": 16120 }, { "grad_norm": 0.3011395335197449, "learning_rate": 9.665190829662337e-05, "loss": 0.0365, "step": 16130 }, { "grad_norm": 0.3890451192855835, "learning_rate": 9.664595693504342e-05, "loss": 0.039, "step": 16140 }, { "grad_norm": 0.2864048480987549, "learning_rate": 9.664000047233175e-05, "loss": 0.0429, "step": 16150 }, { "grad_norm": 0.3107631206512451, "learning_rate": 9.663403890913976e-05, "loss": 0.0472, "step": 16160 }, { "grad_norm": 0.3112051784992218, "learning_rate": 9.662807224611938e-05, "loss": 0.0378, "step": 16170 }, { "grad_norm": 0.32703399658203125, "learning_rate": 9.662210048392311e-05, "loss": 0.0385, "step": 16180 }, { "grad_norm": 0.2823988199234009, "learning_rate": 9.661612362320405e-05, "loss": 0.0436, "step": 16190 }, { "grad_norm": 0.3056599497795105, "learning_rate": 9.661014166461579e-05, "loss": 0.0413, "step": 16200 }, { "grad_norm": 0.28197023272514343, "learning_rate": 9.66041546088125e-05, "loss": 0.0352, "step": 16210 }, { "grad_norm": 0.28476789593696594, "learning_rate": 9.659816245644895e-05, "loss": 0.0436, "step": 16220 }, { "grad_norm": 0.2987506687641144, "learning_rate": 9.65921652081804e-05, "loss": 0.0362, "step": 16230 }, { "grad_norm": 0.26202473044395447, "learning_rate": 9.658616286466271e-05, "loss": 0.0381, "step": 16240 }, { "grad_norm": 0.3056700825691223, "learning_rate": 9.65801554265523e-05, "loss": 0.0423, "step": 16250 }, { "grad_norm": 0.38138943910598755, "learning_rate": 9.657414289450612e-05, "loss": 0.0417, "step": 16260 }, { "grad_norm": 0.3123602569103241, "learning_rate": 9.656812526918171e-05, "loss": 0.0355, "step": 16270 }, { "grad_norm": 0.31641432642936707, "learning_rate": 9.656210255123712e-05, "loss": 0.0397, "step": 16280 }, { "grad_norm": 0.37250563502311707, "learning_rate": 9.6556074741331e-05, "loss": 0.0399, "step": 16290 }, { "grad_norm": 0.34540697932243347, "learning_rate": 9.655004184012256e-05, "loss": 0.0366, "step": 16300 }, { "grad_norm": 0.27798768877983093, "learning_rate": 9.654400384827152e-05, "loss": 0.0316, "step": 16310 }, { "grad_norm": 0.3460087180137634, "learning_rate": 9.653796076643818e-05, "loss": 0.0394, "step": 16320 }, { "grad_norm": 0.31902560591697693, "learning_rate": 9.653191259528344e-05, "loss": 0.035, "step": 16330 }, { "grad_norm": 0.39732620120048523, "learning_rate": 9.65258593354687e-05, "loss": 0.0361, "step": 16340 }, { "grad_norm": 0.2532636225223541, "learning_rate": 9.651980098765591e-05, "loss": 0.0349, "step": 16350 }, { "grad_norm": 0.30868861079216003, "learning_rate": 9.651373755250765e-05, "loss": 0.0407, "step": 16360 }, { "grad_norm": 0.3459707498550415, "learning_rate": 9.650766903068697e-05, "loss": 0.0384, "step": 16370 }, { "grad_norm": 0.4229661524295807, "learning_rate": 9.650159542285753e-05, "loss": 0.0371, "step": 16380 }, { "grad_norm": 0.2762415111064911, "learning_rate": 9.649551672968353e-05, "loss": 0.037, "step": 16390 }, { "grad_norm": 0.3963911533355713, "learning_rate": 9.648943295182973e-05, "loss": 0.0373, "step": 16400 }, { "grad_norm": 0.30376720428466797, "learning_rate": 9.648334408996144e-05, "loss": 0.0389, "step": 16410 }, { "grad_norm": 0.27936527132987976, "learning_rate": 9.647725014474452e-05, "loss": 0.0435, "step": 16420 }, { "grad_norm": 0.3841704726219177, "learning_rate": 9.64711511168454e-05, "loss": 0.0419, "step": 16430 }, { "grad_norm": 0.3290550112724304, "learning_rate": 9.646504700693108e-05, "loss": 0.0437, "step": 16440 }, { "grad_norm": 0.32618406414985657, "learning_rate": 9.645893781566907e-05, "loss": 0.0367, "step": 16450 }, { "grad_norm": 0.36765414476394653, "learning_rate": 9.645282354372744e-05, "loss": 0.0403, "step": 16460 }, { "grad_norm": 0.4178431034088135, "learning_rate": 9.644670419177491e-05, "loss": 0.0423, "step": 16470 }, { "grad_norm": 0.26255813241004944, "learning_rate": 9.644057976048062e-05, "loss": 0.0341, "step": 16480 }, { "grad_norm": 0.3336696922779083, "learning_rate": 9.643445025051435e-05, "loss": 0.0371, "step": 16490 }, { "grad_norm": 0.24567387998104095, "learning_rate": 9.642831566254641e-05, "loss": 0.0326, "step": 16500 }, { "grad_norm": 0.3161414861679077, "learning_rate": 9.642217599724769e-05, "loss": 0.0403, "step": 16510 }, { "grad_norm": 0.3374008536338806, "learning_rate": 9.64160312552896e-05, "loss": 0.0433, "step": 16520 }, { "grad_norm": 0.2692716419696808, "learning_rate": 9.64098814373441e-05, "loss": 0.0391, "step": 16530 }, { "grad_norm": 0.2866562604904175, "learning_rate": 9.640372654408374e-05, "loss": 0.0414, "step": 16540 }, { "grad_norm": 0.2715506851673126, "learning_rate": 9.639756657618162e-05, "loss": 0.0418, "step": 16550 }, { "grad_norm": 0.2629857659339905, "learning_rate": 9.639140153431138e-05, "loss": 0.0374, "step": 16560 }, { "grad_norm": 0.3613930642604828, "learning_rate": 9.638523141914721e-05, "loss": 0.0348, "step": 16570 }, { "grad_norm": 0.2802883982658386, "learning_rate": 9.637905623136388e-05, "loss": 0.0356, "step": 16580 }, { "grad_norm": 0.37127259373664856, "learning_rate": 9.637287597163669e-05, "loss": 0.0371, "step": 16590 }, { "grad_norm": 0.26706168055534363, "learning_rate": 9.63666906406415e-05, "loss": 0.0341, "step": 16600 }, { "grad_norm": 0.32318931818008423, "learning_rate": 9.636050023905473e-05, "loss": 0.0356, "step": 16610 }, { "grad_norm": 0.3055196702480316, "learning_rate": 9.635430476755336e-05, "loss": 0.036, "step": 16620 }, { "grad_norm": 0.2631864547729492, "learning_rate": 9.63481042268149e-05, "loss": 0.0413, "step": 16630 }, { "grad_norm": 0.6769106388092041, "learning_rate": 9.634189861751745e-05, "loss": 0.0367, "step": 16640 }, { "grad_norm": 0.30335310101509094, "learning_rate": 9.633568794033967e-05, "loss": 0.0409, "step": 16650 }, { "grad_norm": 0.2588582932949066, "learning_rate": 9.63294721959607e-05, "loss": 0.0409, "step": 16660 }, { "grad_norm": 0.33905285596847534, "learning_rate": 9.63232513850603e-05, "loss": 0.04, "step": 16670 }, { "grad_norm": 0.3282262682914734, "learning_rate": 9.631702550831878e-05, "loss": 0.0391, "step": 16680 }, { "grad_norm": 0.3573174774646759, "learning_rate": 9.631079456641698e-05, "loss": 0.0379, "step": 16690 }, { "grad_norm": 0.27754610776901245, "learning_rate": 9.630455856003632e-05, "loss": 0.0412, "step": 16700 }, { "grad_norm": 0.33984750509262085, "learning_rate": 9.629831748985876e-05, "loss": 0.0384, "step": 16710 }, { "grad_norm": 0.32840728759765625, "learning_rate": 9.629207135656679e-05, "loss": 0.0444, "step": 16720 }, { "grad_norm": 0.2980872392654419, "learning_rate": 9.628582016084353e-05, "loss": 0.0417, "step": 16730 }, { "grad_norm": 0.25286006927490234, "learning_rate": 9.627956390337254e-05, "loss": 0.0362, "step": 16740 }, { "grad_norm": 0.3451392650604248, "learning_rate": 9.627330258483802e-05, "loss": 0.036, "step": 16750 }, { "grad_norm": 0.32793861627578735, "learning_rate": 9.62670362059247e-05, "loss": 0.0321, "step": 16760 }, { "grad_norm": 0.29118236899375916, "learning_rate": 9.626076476731786e-05, "loss": 0.0398, "step": 16770 }, { "grad_norm": 0.2859082520008087, "learning_rate": 9.625448826970336e-05, "loss": 0.031, "step": 16780 }, { "grad_norm": 0.33027711510658264, "learning_rate": 9.624820671376755e-05, "loss": 0.0376, "step": 16790 }, { "grad_norm": 0.2667204439640045, "learning_rate": 9.62419201001974e-05, "loss": 0.0357, "step": 16800 }, { "grad_norm": 0.3567417860031128, "learning_rate": 9.623562842968037e-05, "loss": 0.0491, "step": 16810 }, { "grad_norm": 0.2642751634120941, "learning_rate": 9.622933170290454e-05, "loss": 0.0358, "step": 16820 }, { "grad_norm": 0.34983518719673157, "learning_rate": 9.622302992055849e-05, "loss": 0.0367, "step": 16830 }, { "grad_norm": 0.2715960443019867, "learning_rate": 9.62167230833314e-05, "loss": 0.0427, "step": 16840 }, { "grad_norm": 0.3179127871990204, "learning_rate": 9.621041119191295e-05, "loss": 0.0459, "step": 16850 }, { "grad_norm": 0.310914009809494, "learning_rate": 9.620409424699342e-05, "loss": 0.0398, "step": 16860 }, { "grad_norm": 0.33890244364738464, "learning_rate": 9.619777224926359e-05, "loss": 0.0368, "step": 16870 }, { "grad_norm": 0.3212319314479828, "learning_rate": 9.619144519941485e-05, "loss": 0.0419, "step": 16880 }, { "grad_norm": 0.27023258805274963, "learning_rate": 9.618511309813912e-05, "loss": 0.0346, "step": 16890 }, { "grad_norm": 0.31821829080581665, "learning_rate": 9.617877594612886e-05, "loss": 0.0364, "step": 16900 }, { "grad_norm": 0.35137420892715454, "learning_rate": 9.617243374407707e-05, "loss": 0.0396, "step": 16910 }, { "grad_norm": 0.5347915887832642, "learning_rate": 9.616608649267736e-05, "loss": 0.0358, "step": 16920 }, { "grad_norm": 0.3147035241127014, "learning_rate": 9.615973419262385e-05, "loss": 0.0384, "step": 16930 }, { "grad_norm": 0.24787050485610962, "learning_rate": 9.615337684461119e-05, "loss": 0.0349, "step": 16940 }, { "grad_norm": 0.22012092173099518, "learning_rate": 9.614701444933465e-05, "loss": 0.0346, "step": 16950 }, { "grad_norm": 0.25866374373435974, "learning_rate": 9.614064700748997e-05, "loss": 0.0352, "step": 16960 }, { "grad_norm": 0.31257364153862, "learning_rate": 9.613427451977352e-05, "loss": 0.0377, "step": 16970 }, { "grad_norm": 0.25820785760879517, "learning_rate": 9.612789698688216e-05, "loss": 0.0403, "step": 16980 }, { "grad_norm": 0.3095071613788605, "learning_rate": 9.612151440951334e-05, "loss": 0.0379, "step": 16990 }, { "grad_norm": 0.22482089698314667, "learning_rate": 9.611512678836506e-05, "loss": 0.041, "step": 17000 }, { "grad_norm": 0.3632158637046814, "learning_rate": 9.610873412413584e-05, "loss": 0.0369, "step": 17010 }, { "grad_norm": 0.2851641774177551, "learning_rate": 9.610233641752476e-05, "loss": 0.0455, "step": 17020 }, { "grad_norm": 0.2826571762561798, "learning_rate": 9.609593366923151e-05, "loss": 0.0364, "step": 17030 }, { "grad_norm": 0.30497848987579346, "learning_rate": 9.608952587995625e-05, "loss": 0.0378, "step": 17040 }, { "grad_norm": 0.24523262679576874, "learning_rate": 9.608311305039972e-05, "loss": 0.0424, "step": 17050 }, { "grad_norm": 0.3125912547111511, "learning_rate": 9.607669518126326e-05, "loss": 0.0455, "step": 17060 }, { "grad_norm": 0.3897329568862915, "learning_rate": 9.607027227324866e-05, "loss": 0.0382, "step": 17070 }, { "grad_norm": 0.29137834906578064, "learning_rate": 9.606384432705837e-05, "loss": 0.0412, "step": 17080 }, { "grad_norm": 0.3425300419330597, "learning_rate": 9.60574113433953e-05, "loss": 0.0368, "step": 17090 }, { "grad_norm": 0.21462294459342957, "learning_rate": 9.6050973322963e-05, "loss": 0.0318, "step": 17100 }, { "grad_norm": 0.2786695063114166, "learning_rate": 9.604453026646547e-05, "loss": 0.034, "step": 17110 }, { "grad_norm": 0.32097041606903076, "learning_rate": 9.603808217460735e-05, "loss": 0.0354, "step": 17120 }, { "grad_norm": 0.2809235155582428, "learning_rate": 9.603162904809377e-05, "loss": 0.0387, "step": 17130 }, { "grad_norm": 0.3362914025783539, "learning_rate": 9.602517088763045e-05, "loss": 0.0386, "step": 17140 }, { "grad_norm": 0.2695412039756775, "learning_rate": 9.601870769392365e-05, "loss": 0.0326, "step": 17150 }, { "grad_norm": 0.2625236213207245, "learning_rate": 9.601223946768017e-05, "loss": 0.0347, "step": 17160 }, { "grad_norm": 0.3730710744857788, "learning_rate": 9.600576620960734e-05, "loss": 0.0361, "step": 17170 }, { "grad_norm": 0.3226581811904907, "learning_rate": 9.599928792041308e-05, "loss": 0.0392, "step": 17180 }, { "grad_norm": 0.2731984257698059, "learning_rate": 9.599280460080587e-05, "loss": 0.0363, "step": 17190 }, { "grad_norm": 0.2902558445930481, "learning_rate": 9.59863162514947e-05, "loss": 0.0356, "step": 17200 }, { "grad_norm": 0.3084840476512909, "learning_rate": 9.597982287318911e-05, "loss": 0.0356, "step": 17210 }, { "grad_norm": 0.28694501519203186, "learning_rate": 9.597332446659923e-05, "loss": 0.0379, "step": 17220 }, { "grad_norm": 0.46241477131843567, "learning_rate": 9.59668210324357e-05, "loss": 0.0351, "step": 17230 }, { "grad_norm": 0.35319003462791443, "learning_rate": 9.596031257140974e-05, "loss": 0.0326, "step": 17240 }, { "grad_norm": 0.3135063052177429, "learning_rate": 9.59537990842331e-05, "loss": 0.033, "step": 17250 }, { "grad_norm": 0.29288938641548157, "learning_rate": 9.594728057161806e-05, "loss": 0.0342, "step": 17260 }, { "grad_norm": 0.23839586973190308, "learning_rate": 9.594075703427752e-05, "loss": 0.0428, "step": 17270 }, { "grad_norm": 0.22460274398326874, "learning_rate": 9.593422847292486e-05, "loss": 0.0398, "step": 17280 }, { "grad_norm": 0.29157164692878723, "learning_rate": 9.592769488827402e-05, "loss": 0.0372, "step": 17290 }, { "grad_norm": 0.2772408425807953, "learning_rate": 9.592115628103952e-05, "loss": 0.0315, "step": 17300 }, { "grad_norm": 0.2725975215435028, "learning_rate": 9.591461265193643e-05, "loss": 0.036, "step": 17310 }, { "grad_norm": 0.3174715042114258, "learning_rate": 9.590806400168032e-05, "loss": 0.0302, "step": 17320 }, { "grad_norm": 0.26562410593032837, "learning_rate": 9.590151033098735e-05, "loss": 0.0362, "step": 17330 }, { "grad_norm": 0.30964893102645874, "learning_rate": 9.589495164057423e-05, "loss": 0.0374, "step": 17340 }, { "grad_norm": 0.24875487387180328, "learning_rate": 9.58883879311582e-05, "loss": 0.0335, "step": 17350 }, { "grad_norm": 0.25152140855789185, "learning_rate": 9.588181920345705e-05, "loss": 0.0357, "step": 17360 }, { "grad_norm": 0.3111923933029175, "learning_rate": 9.587524545818913e-05, "loss": 0.0382, "step": 17370 }, { "grad_norm": 0.31354573369026184, "learning_rate": 9.586866669607335e-05, "loss": 0.033, "step": 17380 }, { "grad_norm": 0.359882652759552, "learning_rate": 9.586208291782915e-05, "loss": 0.0395, "step": 17390 }, { "grad_norm": 0.2905906140804291, "learning_rate": 9.58554941241765e-05, "loss": 0.0321, "step": 17400 }, { "grad_norm": 0.2600795328617096, "learning_rate": 9.584890031583596e-05, "loss": 0.0345, "step": 17410 }, { "grad_norm": 0.3004423677921295, "learning_rate": 9.584230149352861e-05, "loss": 0.039, "step": 17420 }, { "grad_norm": 0.2251468300819397, "learning_rate": 9.58356976579761e-05, "loss": 0.0379, "step": 17430 }, { "grad_norm": 0.2542807459831238, "learning_rate": 9.58290888099006e-05, "loss": 0.0345, "step": 17440 }, { "grad_norm": 0.3080635368824005, "learning_rate": 9.582247495002486e-05, "loss": 0.0379, "step": 17450 }, { "grad_norm": 0.25770148634910583, "learning_rate": 9.581585607907214e-05, "loss": 0.0336, "step": 17460 }, { "grad_norm": 0.3328300416469574, "learning_rate": 9.580923219776628e-05, "loss": 0.0389, "step": 17470 }, { "grad_norm": 1.203275203704834, "learning_rate": 9.580260330683167e-05, "loss": 0.0371, "step": 17480 }, { "grad_norm": 0.33109530806541443, "learning_rate": 9.579596940699322e-05, "loss": 0.0346, "step": 17490 }, { "grad_norm": 0.31651926040649414, "learning_rate": 9.578933049897643e-05, "loss": 0.0417, "step": 17500 }, { "grad_norm": 0.3273641765117645, "learning_rate": 9.578268658350728e-05, "loss": 0.0454, "step": 17510 }, { "grad_norm": 0.2536168694496155, "learning_rate": 9.577603766131235e-05, "loss": 0.0357, "step": 17520 }, { "grad_norm": 0.29349789023399353, "learning_rate": 9.576938373311878e-05, "loss": 0.0393, "step": 17530 }, { "grad_norm": 0.3122764825820923, "learning_rate": 9.576272479965421e-05, "loss": 0.0404, "step": 17540 }, { "grad_norm": 0.41974005103111267, "learning_rate": 9.575606086164687e-05, "loss": 0.0411, "step": 17550 }, { "grad_norm": 0.39105308055877686, "learning_rate": 9.57493919198255e-05, "loss": 0.0392, "step": 17560 }, { "grad_norm": 0.2958887219429016, "learning_rate": 9.57427179749194e-05, "loss": 0.0402, "step": 17570 }, { "grad_norm": 0.3136330544948578, "learning_rate": 9.573603902765846e-05, "loss": 0.0414, "step": 17580 }, { "grad_norm": 0.26090845465660095, "learning_rate": 9.572935507877304e-05, "loss": 0.0464, "step": 17590 }, { "grad_norm": 0.3432660698890686, "learning_rate": 9.57226661289941e-05, "loss": 0.0443, "step": 17600 }, { "grad_norm": 0.33883777260780334, "learning_rate": 9.571597217905315e-05, "loss": 0.042, "step": 17610 }, { "grad_norm": 0.3268703818321228, "learning_rate": 9.57092732296822e-05, "loss": 0.0469, "step": 17620 }, { "grad_norm": 0.3191290497779846, "learning_rate": 9.570256928161385e-05, "loss": 0.0355, "step": 17630 }, { "grad_norm": 0.2908702492713928, "learning_rate": 9.569586033558126e-05, "loss": 0.0397, "step": 17640 }, { "grad_norm": 0.31248608231544495, "learning_rate": 9.568914639231807e-05, "loss": 0.0403, "step": 17650 }, { "grad_norm": 0.32630226016044617, "learning_rate": 9.568242745255852e-05, "loss": 0.0405, "step": 17660 }, { "grad_norm": 0.2935034930706024, "learning_rate": 9.567570351703739e-05, "loss": 0.0339, "step": 17670 }, { "grad_norm": 0.2986380159854889, "learning_rate": 9.566897458649001e-05, "loss": 0.0364, "step": 17680 }, { "grad_norm": 0.20895886421203613, "learning_rate": 9.566224066165221e-05, "loss": 0.0354, "step": 17690 }, { "grad_norm": 0.2743991017341614, "learning_rate": 9.565550174326043e-05, "loss": 0.0424, "step": 17700 }, { "grad_norm": 0.2716120183467865, "learning_rate": 9.564875783205162e-05, "loss": 0.0338, "step": 17710 }, { "grad_norm": 0.7761270403862, "learning_rate": 9.564200892876328e-05, "loss": 0.0369, "step": 17720 }, { "grad_norm": 0.34927836060523987, "learning_rate": 9.563525503413348e-05, "loss": 0.0399, "step": 17730 }, { "grad_norm": 0.3337514400482178, "learning_rate": 9.562849614890079e-05, "loss": 0.0453, "step": 17740 }, { "grad_norm": 0.43548545241355896, "learning_rate": 9.562173227380436e-05, "loss": 0.0451, "step": 17750 }, { "grad_norm": 0.37498724460601807, "learning_rate": 9.561496340958389e-05, "loss": 0.0394, "step": 17760 }, { "grad_norm": 0.3512674570083618, "learning_rate": 9.560818955697959e-05, "loss": 0.0355, "step": 17770 }, { "grad_norm": 0.24486853182315826, "learning_rate": 9.560141071673228e-05, "loss": 0.0362, "step": 17780 }, { "grad_norm": 0.23244410753250122, "learning_rate": 9.559462688958323e-05, "loss": 0.0338, "step": 17790 }, { "grad_norm": 0.27905094623565674, "learning_rate": 9.558783807627434e-05, "loss": 0.0344, "step": 17800 }, { "grad_norm": 0.33067774772644043, "learning_rate": 9.558104427754801e-05, "loss": 0.0322, "step": 17810 }, { "grad_norm": 0.2923247814178467, "learning_rate": 9.557424549414722e-05, "loss": 0.0378, "step": 17820 }, { "grad_norm": 0.270332932472229, "learning_rate": 9.556744172681546e-05, "loss": 0.0339, "step": 17830 }, { "grad_norm": 0.32437339425086975, "learning_rate": 9.556063297629677e-05, "loss": 0.0349, "step": 17840 }, { "grad_norm": 0.26931843161582947, "learning_rate": 9.555381924333578e-05, "loss": 0.0357, "step": 17850 }, { "grad_norm": 0.28443896770477295, "learning_rate": 9.554700052867758e-05, "loss": 0.0342, "step": 17860 }, { "grad_norm": 0.30152666568756104, "learning_rate": 9.554017683306789e-05, "loss": 0.0325, "step": 17870 }, { "grad_norm": 0.3377286493778229, "learning_rate": 9.553334815725294e-05, "loss": 0.0404, "step": 17880 }, { "grad_norm": 0.3807281255722046, "learning_rate": 9.552651450197949e-05, "loss": 0.0351, "step": 17890 }, { "grad_norm": 0.3465990126132965, "learning_rate": 9.551967586799486e-05, "loss": 0.0369, "step": 17900 }, { "grad_norm": 0.38770824670791626, "learning_rate": 9.551283225604692e-05, "loss": 0.0395, "step": 17910 }, { "grad_norm": 0.33356693387031555, "learning_rate": 9.550598366688406e-05, "loss": 0.0379, "step": 17920 }, { "grad_norm": 0.4224439859390259, "learning_rate": 9.549913010125526e-05, "loss": 0.0409, "step": 17930 }, { "grad_norm": 0.2901231050491333, "learning_rate": 9.549227155990999e-05, "loss": 0.0344, "step": 17940 }, { "grad_norm": 0.300426721572876, "learning_rate": 9.548540804359828e-05, "loss": 0.0366, "step": 17950 }, { "grad_norm": 0.27020084857940674, "learning_rate": 9.547853955307077e-05, "loss": 0.033, "step": 17960 }, { "grad_norm": 0.3573792278766632, "learning_rate": 9.547166608907853e-05, "loss": 0.0344, "step": 17970 }, { "grad_norm": 0.33713147044181824, "learning_rate": 9.546478765237326e-05, "loss": 0.0374, "step": 17980 }, { "grad_norm": 0.38726285099983215, "learning_rate": 9.545790424370715e-05, "loss": 0.0393, "step": 17990 }, { "grad_norm": 0.2972657084465027, "learning_rate": 9.5451015863833e-05, "loss": 0.0368, "step": 18000 }, { "grad_norm": 0.29002678394317627, "learning_rate": 9.544412251350408e-05, "loss": 0.0386, "step": 18010 }, { "grad_norm": 0.24214592576026917, "learning_rate": 9.543722419347422e-05, "loss": 0.0393, "step": 18020 }, { "grad_norm": 0.30298906564712524, "learning_rate": 9.543032090449788e-05, "loss": 0.0384, "step": 18030 }, { "grad_norm": 0.3433704078197479, "learning_rate": 9.542341264732992e-05, "loss": 0.0383, "step": 18040 }, { "grad_norm": 0.2466932237148285, "learning_rate": 9.541649942272585e-05, "loss": 0.0375, "step": 18050 }, { "grad_norm": 0.2708093225955963, "learning_rate": 9.54095812314417e-05, "loss": 0.0363, "step": 18060 }, { "grad_norm": 0.2605464458465576, "learning_rate": 9.540265807423401e-05, "loss": 0.0343, "step": 18070 }, { "grad_norm": 0.2187110036611557, "learning_rate": 9.53957299518599e-05, "loss": 0.0373, "step": 18080 }, { "grad_norm": 0.3129284679889679, "learning_rate": 9.5388796865077e-05, "loss": 0.0346, "step": 18090 }, { "grad_norm": 0.2336917221546173, "learning_rate": 9.538185881464353e-05, "loss": 0.0353, "step": 18100 }, { "grad_norm": 0.2693476676940918, "learning_rate": 9.537491580131821e-05, "loss": 0.0378, "step": 18110 }, { "grad_norm": 0.32758137583732605, "learning_rate": 9.53679678258603e-05, "loss": 0.0313, "step": 18120 }, { "grad_norm": 0.30243009328842163, "learning_rate": 9.536101488902966e-05, "loss": 0.0345, "step": 18130 }, { "grad_norm": 0.288645476102829, "learning_rate": 9.535405699158663e-05, "loss": 0.0377, "step": 18140 }, { "grad_norm": 0.28619906306266785, "learning_rate": 9.53470941342921e-05, "loss": 0.0376, "step": 18150 }, { "grad_norm": 0.2590637505054474, "learning_rate": 9.534012631790756e-05, "loss": 0.0355, "step": 18160 }, { "grad_norm": 0.36061230301856995, "learning_rate": 9.533315354319494e-05, "loss": 0.0368, "step": 18170 }, { "grad_norm": 0.27225422859191895, "learning_rate": 9.532617581091682e-05, "loss": 0.0424, "step": 18180 }, { "grad_norm": 0.23305687308311462, "learning_rate": 9.531919312183629e-05, "loss": 0.0372, "step": 18190 }, { "grad_norm": 0.37958723306655884, "learning_rate": 9.531220547671688e-05, "loss": 0.0464, "step": 18200 }, { "grad_norm": 0.3049033582210541, "learning_rate": 9.530521287632285e-05, "loss": 0.0358, "step": 18210 }, { "grad_norm": 0.29977327585220337, "learning_rate": 9.529821532141884e-05, "loss": 0.0368, "step": 18220 }, { "grad_norm": 0.5595778226852417, "learning_rate": 9.52912128127701e-05, "loss": 0.0437, "step": 18230 }, { "grad_norm": 0.3409932851791382, "learning_rate": 9.528420535114244e-05, "loss": 0.0429, "step": 18240 }, { "grad_norm": 0.2678205370903015, "learning_rate": 9.527719293730215e-05, "loss": 0.0395, "step": 18250 }, { "grad_norm": 0.31317138671875, "learning_rate": 9.527017557201611e-05, "loss": 0.0352, "step": 18260 }, { "grad_norm": 0.23997248709201813, "learning_rate": 9.526315325605176e-05, "loss": 0.0367, "step": 18270 }, { "grad_norm": 0.28682172298431396, "learning_rate": 9.525612599017699e-05, "loss": 0.0329, "step": 18280 }, { "grad_norm": 0.35202962160110474, "learning_rate": 9.524909377516033e-05, "loss": 0.0364, "step": 18290 }, { "grad_norm": 0.26805537939071655, "learning_rate": 9.524205661177081e-05, "loss": 0.0372, "step": 18300 }, { "grad_norm": 0.29652801156044006, "learning_rate": 9.523501450077801e-05, "loss": 0.0342, "step": 18310 }, { "grad_norm": 0.2527906596660614, "learning_rate": 9.522796744295202e-05, "loss": 0.0352, "step": 18320 }, { "grad_norm": 0.3309440016746521, "learning_rate": 9.522091543906352e-05, "loss": 0.0347, "step": 18330 }, { "grad_norm": 0.31337934732437134, "learning_rate": 9.521385848988369e-05, "loss": 0.0384, "step": 18340 }, { "grad_norm": 0.315447598695755, "learning_rate": 9.520679659618428e-05, "loss": 0.0329, "step": 18350 }, { "grad_norm": 0.3210081160068512, "learning_rate": 9.519972975873754e-05, "loss": 0.0328, "step": 18360 }, { "grad_norm": 0.3072379529476166, "learning_rate": 9.519265797831633e-05, "loss": 0.0391, "step": 18370 }, { "grad_norm": 0.37524548172950745, "learning_rate": 9.518558125569399e-05, "loss": 0.0417, "step": 18380 }, { "grad_norm": 0.2700424790382385, "learning_rate": 9.517849959164442e-05, "loss": 0.038, "step": 18390 }, { "grad_norm": 0.27296364307403564, "learning_rate": 9.517141298694205e-05, "loss": 0.0383, "step": 18400 }, { "grad_norm": 0.26264774799346924, "learning_rate": 9.516432144236188e-05, "loss": 0.0351, "step": 18410 }, { "grad_norm": 0.42257893085479736, "learning_rate": 9.515722495867941e-05, "loss": 0.0347, "step": 18420 }, { "grad_norm": 0.2771919369697571, "learning_rate": 9.515012353667072e-05, "loss": 0.0336, "step": 18430 }, { "grad_norm": 0.28078025579452515, "learning_rate": 9.51430171771124e-05, "loss": 0.0333, "step": 18440 }, { "grad_norm": 0.3616812825202942, "learning_rate": 9.513590588078159e-05, "loss": 0.0383, "step": 18450 }, { "grad_norm": 0.24792048335075378, "learning_rate": 9.512878964845597e-05, "loss": 0.0361, "step": 18460 }, { "grad_norm": 0.250845730304718, "learning_rate": 9.512166848091377e-05, "loss": 0.035, "step": 18470 }, { "grad_norm": 0.25421980023384094, "learning_rate": 9.511454237893376e-05, "loss": 0.036, "step": 18480 }, { "grad_norm": 0.2806874215602875, "learning_rate": 9.51074113432952e-05, "loss": 0.0356, "step": 18490 }, { "grad_norm": 0.22894933819770813, "learning_rate": 9.510027537477797e-05, "loss": 0.0344, "step": 18500 }, { "grad_norm": 0.2923920750617981, "learning_rate": 9.509313447416242e-05, "loss": 0.0386, "step": 18510 }, { "grad_norm": 0.30689263343811035, "learning_rate": 9.508598864222949e-05, "loss": 0.0347, "step": 18520 }, { "grad_norm": 0.32649046182632446, "learning_rate": 9.507883787976062e-05, "loss": 0.0331, "step": 18530 }, { "grad_norm": 0.2627509832382202, "learning_rate": 9.507168218753781e-05, "loss": 0.0372, "step": 18540 }, { "grad_norm": 0.23703478276729584, "learning_rate": 9.506452156634362e-05, "loss": 0.0327, "step": 18550 }, { "grad_norm": 0.28633081912994385, "learning_rate": 9.505735601696109e-05, "loss": 0.0335, "step": 18560 }, { "grad_norm": 0.27317872643470764, "learning_rate": 9.505018554017385e-05, "loss": 0.0332, "step": 18570 }, { "grad_norm": 0.2841385304927826, "learning_rate": 9.504301013676604e-05, "loss": 0.0337, "step": 18580 }, { "grad_norm": 0.3357928395271301, "learning_rate": 9.503582980752238e-05, "loss": 0.0374, "step": 18590 }, { "grad_norm": 0.31912118196487427, "learning_rate": 9.502864455322809e-05, "loss": 0.0289, "step": 18600 }, { "grad_norm": 0.2584540843963623, "learning_rate": 9.502145437466891e-05, "loss": 0.0408, "step": 18610 }, { "grad_norm": 0.25140154361724854, "learning_rate": 9.501425927263116e-05, "loss": 0.0317, "step": 18620 }, { "grad_norm": 0.24372711777687073, "learning_rate": 9.500705924790172e-05, "loss": 0.0325, "step": 18630 }, { "grad_norm": 0.29973331093788147, "learning_rate": 9.499985430126794e-05, "loss": 0.0407, "step": 18640 }, { "grad_norm": 0.2997944951057434, "learning_rate": 9.499264443351775e-05, "loss": 0.0343, "step": 18650 }, { "grad_norm": 0.30515968799591064, "learning_rate": 9.498542964543961e-05, "loss": 0.0356, "step": 18660 }, { "grad_norm": 0.3630238175392151, "learning_rate": 9.497820993782252e-05, "loss": 0.0384, "step": 18670 }, { "grad_norm": 0.25365564227104187, "learning_rate": 9.497098531145601e-05, "loss": 0.0344, "step": 18680 }, { "grad_norm": 0.3339015543460846, "learning_rate": 9.496375576713017e-05, "loss": 0.0441, "step": 18690 }, { "grad_norm": 0.33822938799858093, "learning_rate": 9.49565213056356e-05, "loss": 0.0385, "step": 18700 }, { "grad_norm": 0.28146040439605713, "learning_rate": 9.494928192776342e-05, "loss": 0.0406, "step": 18710 }, { "grad_norm": 0.36684155464172363, "learning_rate": 9.494203763430538e-05, "loss": 0.0432, "step": 18720 }, { "grad_norm": 0.4130955636501312, "learning_rate": 9.493478842605366e-05, "loss": 0.0482, "step": 18730 }, { "grad_norm": 0.8010186553001404, "learning_rate": 9.492753430380105e-05, "loss": 0.0498, "step": 18740 }, { "grad_norm": 0.344209223985672, "learning_rate": 9.492027526834083e-05, "loss": 0.0583, "step": 18750 }, { "grad_norm": 0.29350852966308594, "learning_rate": 9.491301132046684e-05, "loss": 0.0482, "step": 18760 }, { "grad_norm": 0.36145004630088806, "learning_rate": 9.490574246097345e-05, "loss": 0.0505, "step": 18770 }, { "grad_norm": 0.3269612491130829, "learning_rate": 9.48984686906556e-05, "loss": 0.0446, "step": 18780 }, { "grad_norm": 0.315528005361557, "learning_rate": 9.489119001030871e-05, "loss": 0.0447, "step": 18790 }, { "grad_norm": 0.28926777839660645, "learning_rate": 9.488390642072878e-05, "loss": 0.0377, "step": 18800 }, { "grad_norm": 0.33444055914878845, "learning_rate": 9.48766179227123e-05, "loss": 0.0404, "step": 18810 }, { "grad_norm": 0.30939510464668274, "learning_rate": 9.486932451705636e-05, "loss": 0.0365, "step": 18820 }, { "grad_norm": 0.29159653186798096, "learning_rate": 9.486202620455857e-05, "loss": 0.0434, "step": 18830 }, { "grad_norm": 0.34027260541915894, "learning_rate": 9.485472298601704e-05, "loss": 0.0444, "step": 18840 }, { "grad_norm": 0.3325185179710388, "learning_rate": 9.484741486223043e-05, "loss": 0.0404, "step": 18850 }, { "grad_norm": 0.32428938150405884, "learning_rate": 9.484010183399797e-05, "loss": 0.0427, "step": 18860 }, { "grad_norm": 0.3781747817993164, "learning_rate": 9.483278390211938e-05, "loss": 0.0396, "step": 18870 }, { "grad_norm": 0.37120088934898376, "learning_rate": 9.482546106739496e-05, "loss": 0.0356, "step": 18880 }, { "grad_norm": 0.2562500536441803, "learning_rate": 9.48181333306255e-05, "loss": 0.0336, "step": 18890 }, { "grad_norm": 0.3435106575489044, "learning_rate": 9.481080069261237e-05, "loss": 0.0371, "step": 18900 }, { "grad_norm": 0.24666783213615417, "learning_rate": 9.480346315415745e-05, "loss": 0.0365, "step": 18910 }, { "grad_norm": 0.2778903543949127, "learning_rate": 9.479612071606314e-05, "loss": 0.0426, "step": 18920 }, { "grad_norm": 0.33430778980255127, "learning_rate": 9.478877337913244e-05, "loss": 0.0381, "step": 18930 }, { "grad_norm": 0.2701009213924408, "learning_rate": 9.478142114416881e-05, "loss": 0.0369, "step": 18940 }, { "grad_norm": 0.3106098175048828, "learning_rate": 9.47740640119763e-05, "loss": 0.0408, "step": 18950 }, { "grad_norm": 0.2845015823841095, "learning_rate": 9.476670198335947e-05, "loss": 0.0327, "step": 18960 }, { "grad_norm": 0.31328320503234863, "learning_rate": 9.47593350591234e-05, "loss": 0.037, "step": 18970 }, { "grad_norm": 0.26648658514022827, "learning_rate": 9.475196324007376e-05, "loss": 0.0367, "step": 18980 }, { "grad_norm": 0.2660379707813263, "learning_rate": 9.474458652701669e-05, "loss": 0.0373, "step": 18990 }, { "grad_norm": 0.24749259650707245, "learning_rate": 9.473720492075892e-05, "loss": 0.0336, "step": 19000 }, { "grad_norm": 0.2885451316833496, "learning_rate": 9.472981842210768e-05, "loss": 0.0346, "step": 19010 }, { "grad_norm": 0.3173202872276306, "learning_rate": 9.472242703187074e-05, "loss": 0.0324, "step": 19020 }, { "grad_norm": 0.5223367810249329, "learning_rate": 9.471503075085643e-05, "loss": 0.0382, "step": 19030 }, { "grad_norm": 0.3115447163581848, "learning_rate": 9.470762957987359e-05, "loss": 0.0314, "step": 19040 }, { "grad_norm": 0.2744147777557373, "learning_rate": 9.470022351973158e-05, "loss": 0.0368, "step": 19050 }, { "grad_norm": 0.32628393173217773, "learning_rate": 9.469281257124034e-05, "loss": 0.0396, "step": 19060 }, { "grad_norm": 0.3244577646255493, "learning_rate": 9.46853967352103e-05, "loss": 0.0355, "step": 19070 }, { "grad_norm": 0.28875377774238586, "learning_rate": 9.467797601245246e-05, "loss": 0.0352, "step": 19080 }, { "grad_norm": 0.33956828713417053, "learning_rate": 9.467055040377834e-05, "loss": 0.0426, "step": 19090 }, { "grad_norm": 0.2911307215690613, "learning_rate": 9.466311990999999e-05, "loss": 0.0353, "step": 19100 }, { "grad_norm": 0.21841949224472046, "learning_rate": 9.465568453193e-05, "loss": 0.0389, "step": 19110 }, { "grad_norm": 0.3304734230041504, "learning_rate": 9.464824427038148e-05, "loss": 0.036, "step": 19120 }, { "grad_norm": 0.2681001126766205, "learning_rate": 9.46407991261681e-05, "loss": 0.0314, "step": 19130 }, { "grad_norm": 0.27186623215675354, "learning_rate": 9.463334910010404e-05, "loss": 0.0325, "step": 19140 }, { "grad_norm": 0.3030039072036743, "learning_rate": 9.462589419300403e-05, "loss": 0.0371, "step": 19150 }, { "grad_norm": 0.26235365867614746, "learning_rate": 9.461843440568333e-05, "loss": 0.0389, "step": 19160 }, { "grad_norm": 0.3515237867832184, "learning_rate": 9.461096973895773e-05, "loss": 0.0291, "step": 19170 }, { "grad_norm": 0.3994947075843811, "learning_rate": 9.460350019364355e-05, "loss": 0.0376, "step": 19180 }, { "grad_norm": 0.29883497953414917, "learning_rate": 9.459602577055764e-05, "loss": 0.0342, "step": 19190 }, { "grad_norm": 0.23580463230609894, "learning_rate": 9.45885464705174e-05, "loss": 0.0347, "step": 19200 }, { "grad_norm": 0.3004681468009949, "learning_rate": 9.458106229434076e-05, "loss": 0.0392, "step": 19210 }, { "grad_norm": 0.27012333273887634, "learning_rate": 9.457357324284617e-05, "loss": 0.0334, "step": 19220 }, { "grad_norm": 0.3455204963684082, "learning_rate": 9.456607931685262e-05, "loss": 0.034, "step": 19230 }, { "grad_norm": 0.25939199328422546, "learning_rate": 9.455858051717965e-05, "loss": 0.0348, "step": 19240 }, { "grad_norm": 0.3122013807296753, "learning_rate": 9.45510768446473e-05, "loss": 0.0337, "step": 19250 }, { "grad_norm": 0.2683207392692566, "learning_rate": 9.454356830007618e-05, "loss": 0.0366, "step": 19260 }, { "grad_norm": 0.28551024198532104, "learning_rate": 9.45360548842874e-05, "loss": 0.0329, "step": 19270 }, { "grad_norm": 0.22793005406856537, "learning_rate": 9.452853659810261e-05, "loss": 0.0308, "step": 19280 }, { "grad_norm": 0.26842260360717773, "learning_rate": 9.452101344234401e-05, "loss": 0.036, "step": 19290 }, { "grad_norm": 0.2782140374183655, "learning_rate": 9.451348541783431e-05, "loss": 0.0361, "step": 19300 }, { "grad_norm": 0.3655969798564911, "learning_rate": 9.450595252539678e-05, "loss": 0.0455, "step": 19310 }, { "grad_norm": 0.42203035950660706, "learning_rate": 9.449841476585518e-05, "loss": 0.0388, "step": 19320 }, { "grad_norm": 0.42982974648475647, "learning_rate": 9.449087214003384e-05, "loss": 0.0325, "step": 19330 }, { "grad_norm": 0.28357070684432983, "learning_rate": 9.448332464875765e-05, "loss": 0.0361, "step": 19340 }, { "grad_norm": 0.2897912263870239, "learning_rate": 9.447577229285192e-05, "loss": 0.0362, "step": 19350 }, { "grad_norm": 0.5728402733802795, "learning_rate": 9.446821507314261e-05, "loss": 0.0373, "step": 19360 }, { "grad_norm": 0.36491891741752625, "learning_rate": 9.446065299045617e-05, "loss": 0.0318, "step": 19370 }, { "grad_norm": 0.24255864322185516, "learning_rate": 9.445308604561955e-05, "loss": 0.0378, "step": 19380 }, { "grad_norm": 0.24131974577903748, "learning_rate": 9.444551423946028e-05, "loss": 0.0359, "step": 19390 }, { "grad_norm": 0.30867713689804077, "learning_rate": 9.443793757280638e-05, "loss": 0.0333, "step": 19400 }, { "grad_norm": 0.24941158294677734, "learning_rate": 9.443035604648646e-05, "loss": 0.0318, "step": 19410 }, { "grad_norm": 0.25873592495918274, "learning_rate": 9.44227696613296e-05, "loss": 0.0357, "step": 19420 }, { "grad_norm": 0.23689787089824677, "learning_rate": 9.441517841816542e-05, "loss": 0.0395, "step": 19430 }, { "grad_norm": 0.26019349694252014, "learning_rate": 9.440758231782413e-05, "loss": 0.0333, "step": 19440 }, { "grad_norm": 0.25287139415740967, "learning_rate": 9.439998136113639e-05, "loss": 0.0358, "step": 19450 }, { "grad_norm": 0.2740198075771332, "learning_rate": 9.439237554893344e-05, "loss": 0.0347, "step": 19460 }, { "grad_norm": 0.23501570522785187, "learning_rate": 9.438476488204705e-05, "loss": 0.0338, "step": 19470 }, { "grad_norm": 0.293277770280838, "learning_rate": 9.43771493613095e-05, "loss": 0.0326, "step": 19480 }, { "grad_norm": 0.27377018332481384, "learning_rate": 9.436952898755362e-05, "loss": 0.0334, "step": 19490 }, { "grad_norm": 0.21228042244911194, "learning_rate": 9.436190376161276e-05, "loss": 0.0337, "step": 19500 }, { "grad_norm": 0.2836637794971466, "learning_rate": 9.43542736843208e-05, "loss": 0.039, "step": 19510 }, { "grad_norm": 0.3088230788707733, "learning_rate": 9.434663875651216e-05, "loss": 0.034, "step": 19520 }, { "grad_norm": 0.34025299549102783, "learning_rate": 9.433899897902177e-05, "loss": 0.0287, "step": 19530 }, { "grad_norm": 0.2630127966403961, "learning_rate": 9.433135435268511e-05, "loss": 0.0363, "step": 19540 }, { "grad_norm": 0.3184562921524048, "learning_rate": 9.432370487833819e-05, "loss": 0.0404, "step": 19550 }, { "grad_norm": 0.31731534004211426, "learning_rate": 9.431605055681756e-05, "loss": 0.0382, "step": 19560 }, { "grad_norm": 0.29993152618408203, "learning_rate": 9.430839138896026e-05, "loss": 0.04, "step": 19570 }, { "grad_norm": 0.3337842524051666, "learning_rate": 9.43007273756039e-05, "loss": 0.0437, "step": 19580 }, { "grad_norm": 0.24634379148483276, "learning_rate": 9.429305851758658e-05, "loss": 0.0349, "step": 19590 }, { "grad_norm": 0.30793043971061707, "learning_rate": 9.428538481574699e-05, "loss": 0.0352, "step": 19600 }, { "grad_norm": 0.3239770829677582, "learning_rate": 9.42777062709243e-05, "loss": 0.0365, "step": 19610 }, { "grad_norm": 0.3152424395084381, "learning_rate": 9.427002288395821e-05, "loss": 0.0371, "step": 19620 }, { "grad_norm": 0.33306485414505005, "learning_rate": 9.426233465568898e-05, "loss": 0.0307, "step": 19630 }, { "grad_norm": 0.40245816111564636, "learning_rate": 9.42546415869574e-05, "loss": 0.039, "step": 19640 }, { "grad_norm": 0.3299529254436493, "learning_rate": 9.424694367860473e-05, "loss": 0.0306, "step": 19650 }, { "grad_norm": 0.270679235458374, "learning_rate": 9.423924093147284e-05, "loss": 0.0386, "step": 19660 }, { "grad_norm": 0.2668755352497101, "learning_rate": 9.423153334640407e-05, "loss": 0.0356, "step": 19670 }, { "grad_norm": 0.2651456594467163, "learning_rate": 9.42238209242413e-05, "loss": 0.0315, "step": 19680 }, { "grad_norm": 0.288823664188385, "learning_rate": 9.421610366582798e-05, "loss": 0.0341, "step": 19690 }, { "grad_norm": 0.30134376883506775, "learning_rate": 9.420838157200803e-05, "loss": 0.0323, "step": 19700 }, { "grad_norm": 0.2908545136451721, "learning_rate": 9.420065464362594e-05, "loss": 0.0344, "step": 19710 }, { "grad_norm": 0.38675862550735474, "learning_rate": 9.419292288152673e-05, "loss": 0.0308, "step": 19720 }, { "grad_norm": 0.3135015070438385, "learning_rate": 9.418518628655588e-05, "loss": 0.0347, "step": 19730 }, { "grad_norm": 0.2551846206188202, "learning_rate": 9.417744485955951e-05, "loss": 0.0337, "step": 19740 }, { "grad_norm": 0.2761509418487549, "learning_rate": 9.41696986013842e-05, "loss": 0.036, "step": 19750 }, { "grad_norm": 0.36914366483688354, "learning_rate": 9.416194751287705e-05, "loss": 0.0342, "step": 19760 }, { "grad_norm": 0.31763729453086853, "learning_rate": 9.415419159488572e-05, "loss": 0.0396, "step": 19770 }, { "grad_norm": 0.24151836335659027, "learning_rate": 9.414643084825837e-05, "loss": 0.0331, "step": 19780 }, { "grad_norm": 0.22252242267131805, "learning_rate": 9.413866527384372e-05, "loss": 0.0347, "step": 19790 }, { "grad_norm": 0.2933497130870819, "learning_rate": 9.4130894872491e-05, "loss": 0.0298, "step": 19800 }, { "grad_norm": 0.19840432703495026, "learning_rate": 9.412311964504998e-05, "loss": 0.0382, "step": 19810 }, { "grad_norm": 0.2842673063278198, "learning_rate": 9.411533959237091e-05, "loss": 0.0339, "step": 19820 }, { "grad_norm": 0.19863393902778625, "learning_rate": 9.410755471530464e-05, "loss": 0.0337, "step": 19830 }, { "grad_norm": 0.21903863549232483, "learning_rate": 9.40997650147025e-05, "loss": 0.0342, "step": 19840 }, { "grad_norm": 0.27701476216316223, "learning_rate": 9.409197049141637e-05, "loss": 0.03, "step": 19850 }, { "grad_norm": 0.2338973730802536, "learning_rate": 9.408417114629863e-05, "loss": 0.0319, "step": 19860 }, { "grad_norm": 0.39430978894233704, "learning_rate": 9.40763669802022e-05, "loss": 0.0328, "step": 19870 }, { "grad_norm": 0.28690221905708313, "learning_rate": 9.406855799398056e-05, "loss": 0.0342, "step": 19880 }, { "grad_norm": 0.21071264147758484, "learning_rate": 9.406074418848767e-05, "loss": 0.0334, "step": 19890 }, { "grad_norm": 0.2738555073738098, "learning_rate": 9.405292556457805e-05, "loss": 0.0345, "step": 19900 }, { "grad_norm": 0.2656964659690857, "learning_rate": 9.404510212310671e-05, "loss": 0.0355, "step": 19910 }, { "grad_norm": 0.2129662036895752, "learning_rate": 9.403727386492924e-05, "loss": 0.0395, "step": 19920 }, { "grad_norm": 0.3004094362258911, "learning_rate": 9.40294407909017e-05, "loss": 0.0358, "step": 19930 }, { "grad_norm": 0.25614020228385925, "learning_rate": 9.40216029018807e-05, "loss": 0.032, "step": 19940 }, { "grad_norm": 0.25893962383270264, "learning_rate": 9.401376019872338e-05, "loss": 0.0283, "step": 19950 }, { "grad_norm": 0.26711326837539673, "learning_rate": 9.400591268228746e-05, "loss": 0.0357, "step": 19960 }, { "grad_norm": 0.2411770224571228, "learning_rate": 9.399806035343106e-05, "loss": 0.0346, "step": 19970 }, { "grad_norm": 0.24754014611244202, "learning_rate": 9.399020321301294e-05, "loss": 0.0363, "step": 19980 }, { "grad_norm": 0.2549114525318146, "learning_rate": 9.398234126189234e-05, "loss": 0.0307, "step": 19990 }, { "grad_norm": 0.24494512379169464, "learning_rate": 9.397447450092902e-05, "loss": 0.0294, "step": 20000 }, { "grad_norm": 0.5573055744171143, "learning_rate": 9.39666029309833e-05, "loss": 0.0339, "step": 20010 }, { "grad_norm": 0.4221695065498352, "learning_rate": 9.395872655291596e-05, "loss": 0.0398, "step": 20020 }, { "grad_norm": 0.38071051239967346, "learning_rate": 9.395084536758838e-05, "loss": 0.0352, "step": 20030 }, { "grad_norm": 0.3797624409198761, "learning_rate": 9.394295937586243e-05, "loss": 0.041, "step": 20040 }, { "grad_norm": 0.3136899471282959, "learning_rate": 9.393506857860052e-05, "loss": 0.037, "step": 20050 }, { "grad_norm": 0.30193954706192017, "learning_rate": 9.392717297666555e-05, "loss": 0.0365, "step": 20060 }, { "grad_norm": 0.29312053322792053, "learning_rate": 9.391927257092101e-05, "loss": 0.0329, "step": 20070 }, { "grad_norm": 0.2645489573478699, "learning_rate": 9.391136736223085e-05, "loss": 0.036, "step": 20080 }, { "grad_norm": 0.35860180854797363, "learning_rate": 9.390345735145956e-05, "loss": 0.0371, "step": 20090 }, { "grad_norm": 0.24101127684116364, "learning_rate": 9.389554253947219e-05, "loss": 0.0344, "step": 20100 }, { "grad_norm": 0.29320815205574036, "learning_rate": 9.388762292713428e-05, "loss": 0.0297, "step": 20110 }, { "grad_norm": 0.24464024603366852, "learning_rate": 9.38796985153119e-05, "loss": 0.0303, "step": 20120 }, { "grad_norm": 0.6899362206459045, "learning_rate": 9.387176930487169e-05, "loss": 0.0343, "step": 20130 }, { "grad_norm": 0.308505117893219, "learning_rate": 9.386383529668072e-05, "loss": 0.0342, "step": 20140 }, { "grad_norm": 0.32426121830940247, "learning_rate": 9.385589649160669e-05, "loss": 0.0359, "step": 20150 }, { "grad_norm": 0.27759289741516113, "learning_rate": 9.384795289051775e-05, "loss": 0.037, "step": 20160 }, { "grad_norm": 0.28118768334388733, "learning_rate": 9.384000449428261e-05, "loss": 0.0321, "step": 20170 }, { "grad_norm": 0.3428255617618561, "learning_rate": 9.383205130377048e-05, "loss": 0.0337, "step": 20180 }, { "grad_norm": 0.3113052546977997, "learning_rate": 9.382409331985114e-05, "loss": 0.0437, "step": 20190 }, { "grad_norm": 0.27986350655555725, "learning_rate": 9.381613054339482e-05, "loss": 0.0352, "step": 20200 }, { "grad_norm": 0.25685107707977295, "learning_rate": 9.380816297527235e-05, "loss": 0.0352, "step": 20210 }, { "grad_norm": 0.2741990089416504, "learning_rate": 9.380019061635506e-05, "loss": 0.0372, "step": 20220 }, { "grad_norm": 0.3193974196910858, "learning_rate": 9.379221346751474e-05, "loss": 0.042, "step": 20230 }, { "grad_norm": 0.5965449213981628, "learning_rate": 9.378423152962382e-05, "loss": 0.0369, "step": 20240 }, { "grad_norm": 0.37415581941604614, "learning_rate": 9.377624480355517e-05, "loss": 0.0354, "step": 20250 }, { "grad_norm": 0.6163648366928101, "learning_rate": 9.376825329018219e-05, "loss": 0.0417, "step": 20260 }, { "grad_norm": 0.44717535376548767, "learning_rate": 9.376025699037884e-05, "loss": 0.044, "step": 20270 }, { "grad_norm": 0.29323887825012207, "learning_rate": 9.37522559050196e-05, "loss": 0.0351, "step": 20280 }, { "grad_norm": 0.29691100120544434, "learning_rate": 9.37442500349794e-05, "loss": 0.0343, "step": 20290 }, { "grad_norm": 0.36731937527656555, "learning_rate": 9.373623938113381e-05, "loss": 0.0411, "step": 20300 }, { "grad_norm": 0.2513483166694641, "learning_rate": 9.372822394435883e-05, "loss": 0.0356, "step": 20310 }, { "grad_norm": 0.25311899185180664, "learning_rate": 9.372020372553102e-05, "loss": 0.0375, "step": 20320 }, { "grad_norm": 0.2657555937767029, "learning_rate": 9.371217872552746e-05, "loss": 0.0332, "step": 20330 }, { "grad_norm": 0.30172261595726013, "learning_rate": 9.370414894522576e-05, "loss": 0.036, "step": 20340 }, { "grad_norm": 0.29589399695396423, "learning_rate": 9.369611438550406e-05, "loss": 0.0397, "step": 20350 }, { "grad_norm": 0.2836178243160248, "learning_rate": 9.368807504724095e-05, "loss": 0.031, "step": 20360 }, { "grad_norm": 0.2922353446483612, "learning_rate": 9.368003093131565e-05, "loss": 0.0342, "step": 20370 }, { "grad_norm": 0.23246946930885315, "learning_rate": 9.367198203860785e-05, "loss": 0.0343, "step": 20380 }, { "grad_norm": 0.3421041667461395, "learning_rate": 9.366392836999774e-05, "loss": 0.0395, "step": 20390 }, { "grad_norm": 0.24741941690444946, "learning_rate": 9.365586992636607e-05, "loss": 0.0392, "step": 20400 }, { "grad_norm": 0.22766661643981934, "learning_rate": 9.364780670859412e-05, "loss": 0.0312, "step": 20410 }, { "grad_norm": 0.36796289682388306, "learning_rate": 9.363973871756364e-05, "loss": 0.0355, "step": 20420 }, { "grad_norm": 0.3381396234035492, "learning_rate": 9.363166595415696e-05, "loss": 0.0377, "step": 20430 }, { "grad_norm": 0.3212019205093384, "learning_rate": 9.362358841925686e-05, "loss": 0.0335, "step": 20440 }, { "grad_norm": 0.33825352787971497, "learning_rate": 9.361550611374674e-05, "loss": 0.0342, "step": 20450 }, { "grad_norm": 0.2154557853937149, "learning_rate": 9.360741903851043e-05, "loss": 0.0285, "step": 20460 }, { "grad_norm": 0.24252523481845856, "learning_rate": 9.359932719443236e-05, "loss": 0.0325, "step": 20470 }, { "grad_norm": 0.28237637877464294, "learning_rate": 9.35912305823974e-05, "loss": 0.0366, "step": 20480 }, { "grad_norm": 0.23483118414878845, "learning_rate": 9.358312920329101e-05, "loss": 0.0325, "step": 20490 }, { "grad_norm": 0.26978838443756104, "learning_rate": 9.357502305799914e-05, "loss": 0.0345, "step": 20500 }, { "grad_norm": 0.2909323275089264, "learning_rate": 9.356691214740824e-05, "loss": 0.0344, "step": 20510 }, { "grad_norm": 0.2676534652709961, "learning_rate": 9.355879647240535e-05, "loss": 0.0325, "step": 20520 }, { "grad_norm": 0.3258497416973114, "learning_rate": 9.355067603387798e-05, "loss": 0.0313, "step": 20530 }, { "grad_norm": 0.24347622692584991, "learning_rate": 9.354255083271412e-05, "loss": 0.0419, "step": 20540 }, { "grad_norm": 0.2626056671142578, "learning_rate": 9.353442086980239e-05, "loss": 0.0285, "step": 20550 }, { "grad_norm": 0.2791796326637268, "learning_rate": 9.352628614603185e-05, "loss": 0.0347, "step": 20560 }, { "grad_norm": 0.2456272542476654, "learning_rate": 9.351814666229209e-05, "loss": 0.0339, "step": 20570 }, { "grad_norm": 0.26543954014778137, "learning_rate": 9.351000241947324e-05, "loss": 0.0369, "step": 20580 }, { "grad_norm": 0.2643379271030426, "learning_rate": 9.350185341846594e-05, "loss": 0.0357, "step": 20590 }, { "grad_norm": 0.3004196286201477, "learning_rate": 9.349369966016134e-05, "loss": 0.0336, "step": 20600 }, { "grad_norm": 0.21804001927375793, "learning_rate": 9.348554114545117e-05, "loss": 0.0323, "step": 20610 }, { "grad_norm": 0.2750760018825531, "learning_rate": 9.347737787522758e-05, "loss": 0.0342, "step": 20620 }, { "grad_norm": 0.32393375039100647, "learning_rate": 9.346920985038332e-05, "loss": 0.034, "step": 20630 }, { "grad_norm": 0.2476273626089096, "learning_rate": 9.346103707181162e-05, "loss": 0.0311, "step": 20640 }, { "grad_norm": 0.29078924655914307, "learning_rate": 9.345285954040626e-05, "loss": 0.0301, "step": 20650 }, { "grad_norm": 0.35968318581581116, "learning_rate": 9.34446772570615e-05, "loss": 0.0351, "step": 20660 }, { "grad_norm": 0.2853911221027374, "learning_rate": 9.343649022267214e-05, "loss": 0.0299, "step": 20670 }, { "grad_norm": 0.37530654668807983, "learning_rate": 9.342829843813353e-05, "loss": 0.0396, "step": 20680 }, { "grad_norm": 0.33197176456451416, "learning_rate": 9.342010190434149e-05, "loss": 0.0383, "step": 20690 }, { "grad_norm": 0.33180204033851624, "learning_rate": 9.34119006221924e-05, "loss": 0.0377, "step": 20700 }, { "grad_norm": 0.23908501863479614, "learning_rate": 9.340369459258313e-05, "loss": 0.0346, "step": 20710 }, { "grad_norm": 0.265953004360199, "learning_rate": 9.339548381641106e-05, "loss": 0.0337, "step": 20720 }, { "grad_norm": 0.27194085717201233, "learning_rate": 9.338726829457413e-05, "loss": 0.0352, "step": 20730 }, { "grad_norm": 0.2129693329334259, "learning_rate": 9.337904802797078e-05, "loss": 0.0297, "step": 20740 }, { "grad_norm": 0.3019654154777527, "learning_rate": 9.337082301749993e-05, "loss": 0.0321, "step": 20750 }, { "grad_norm": 0.27056995034217834, "learning_rate": 9.336259326406109e-05, "loss": 0.0313, "step": 20760 }, { "grad_norm": 0.2631112039089203, "learning_rate": 9.335435876855427e-05, "loss": 0.0341, "step": 20770 }, { "grad_norm": 0.3108314573764801, "learning_rate": 9.334611953187994e-05, "loss": 0.0359, "step": 20780 }, { "grad_norm": 0.3425011932849884, "learning_rate": 9.333787555493914e-05, "loss": 0.0371, "step": 20790 }, { "grad_norm": 0.31774669885635376, "learning_rate": 9.332962683863345e-05, "loss": 0.0364, "step": 20800 }, { "grad_norm": 0.24626924097537994, "learning_rate": 9.332137338386489e-05, "loss": 0.0392, "step": 20810 }, { "grad_norm": 0.5762718319892883, "learning_rate": 9.33131151915361e-05, "loss": 0.044, "step": 20820 }, { "grad_norm": 0.32839542627334595, "learning_rate": 9.330485226255012e-05, "loss": 0.0423, "step": 20830 }, { "grad_norm": 0.26598066091537476, "learning_rate": 9.329658459781061e-05, "loss": 0.0447, "step": 20840 }, { "grad_norm": 0.30551639199256897, "learning_rate": 9.328831219822172e-05, "loss": 0.0411, "step": 20850 }, { "grad_norm": 0.27943533658981323, "learning_rate": 9.328003506468808e-05, "loss": 0.0347, "step": 20860 }, { "grad_norm": 0.32123178243637085, "learning_rate": 9.327175319811488e-05, "loss": 0.0391, "step": 20870 }, { "grad_norm": 0.35101187229156494, "learning_rate": 9.326346659940781e-05, "loss": 0.0379, "step": 20880 }, { "grad_norm": 0.25687092542648315, "learning_rate": 9.325517526947308e-05, "loss": 0.0374, "step": 20890 }, { "grad_norm": 0.26068219542503357, "learning_rate": 9.32468792092174e-05, "loss": 0.0399, "step": 20900 }, { "grad_norm": 0.3846026360988617, "learning_rate": 9.323857841954803e-05, "loss": 0.0454, "step": 20910 }, { "grad_norm": 0.4344618320465088, "learning_rate": 9.323027290137276e-05, "loss": 0.0508, "step": 20920 }, { "grad_norm": 0.26141831278800964, "learning_rate": 9.322196265559981e-05, "loss": 0.0434, "step": 20930 }, { "grad_norm": 0.30034396052360535, "learning_rate": 9.321364768313803e-05, "loss": 0.0425, "step": 20940 }, { "grad_norm": 0.3282811641693115, "learning_rate": 9.32053279848967e-05, "loss": 0.044, "step": 20950 }, { "grad_norm": 0.3204282820224762, "learning_rate": 9.319700356178567e-05, "loss": 0.0347, "step": 20960 }, { "grad_norm": 0.47990599274635315, "learning_rate": 9.318867441471527e-05, "loss": 0.0355, "step": 20970 }, { "grad_norm": 0.34505003690719604, "learning_rate": 9.318034054459637e-05, "loss": 0.0373, "step": 20980 }, { "grad_norm": 0.40426549315452576, "learning_rate": 9.317200195234034e-05, "loss": 0.0363, "step": 20990 }, { "grad_norm": 0.2871544063091278, "learning_rate": 9.316365863885909e-05, "loss": 0.0382, "step": 21000 }, { "grad_norm": 0.32224324345588684, "learning_rate": 9.315531060506502e-05, "loss": 0.0367, "step": 21010 }, { "grad_norm": 0.24619172513484955, "learning_rate": 9.314695785187108e-05, "loss": 0.036, "step": 21020 }, { "grad_norm": 0.28639188408851624, "learning_rate": 9.313860038019069e-05, "loss": 0.0417, "step": 21030 }, { "grad_norm": 0.2943178415298462, "learning_rate": 9.313023819093782e-05, "loss": 0.0339, "step": 21040 }, { "grad_norm": 0.33986547589302063, "learning_rate": 9.312187128502695e-05, "loss": 0.0384, "step": 21050 }, { "grad_norm": 0.2998370826244354, "learning_rate": 9.311349966337307e-05, "loss": 0.0351, "step": 21060 }, { "grad_norm": 0.32225868105888367, "learning_rate": 9.310512332689169e-05, "loss": 0.0323, "step": 21070 }, { "grad_norm": 0.6140560507774353, "learning_rate": 9.309674227649883e-05, "loss": 0.0414, "step": 21080 }, { "grad_norm": 0.36629530787467957, "learning_rate": 9.308835651311103e-05, "loss": 0.0475, "step": 21090 }, { "grad_norm": 0.3031627833843231, "learning_rate": 9.307996603764533e-05, "loss": 0.046, "step": 21100 }, { "grad_norm": 0.33481869101524353, "learning_rate": 9.307157085101932e-05, "loss": 0.0373, "step": 21110 }, { "grad_norm": 0.27633920311927795, "learning_rate": 9.306317095415109e-05, "loss": 0.0381, "step": 21120 }, { "grad_norm": 0.33129221200942993, "learning_rate": 9.305476634795922e-05, "loss": 0.0386, "step": 21130 }, { "grad_norm": 0.3842407464981079, "learning_rate": 9.304635703336284e-05, "loss": 0.0371, "step": 21140 }, { "grad_norm": 0.30220669507980347, "learning_rate": 9.303794301128157e-05, "loss": 0.0374, "step": 21150 }, { "grad_norm": 0.2267349660396576, "learning_rate": 9.302952428263555e-05, "loss": 0.0418, "step": 21160 }, { "grad_norm": 0.2461182326078415, "learning_rate": 9.302110084834545e-05, "loss": 0.0356, "step": 21170 }, { "grad_norm": 0.2419830560684204, "learning_rate": 9.301267270933245e-05, "loss": 0.0373, "step": 21180 }, { "grad_norm": 0.3379487693309784, "learning_rate": 9.300423986651823e-05, "loss": 0.0409, "step": 21190 }, { "grad_norm": 0.2694736123085022, "learning_rate": 9.299580232082501e-05, "loss": 0.0306, "step": 21200 }, { "grad_norm": 0.28396305441856384, "learning_rate": 9.298736007317547e-05, "loss": 0.0402, "step": 21210 }, { "grad_norm": 0.27072182297706604, "learning_rate": 9.297891312449288e-05, "loss": 0.0318, "step": 21220 }, { "grad_norm": 0.27563610672950745, "learning_rate": 9.297046147570094e-05, "loss": 0.0284, "step": 21230 }, { "grad_norm": 0.30967840552330017, "learning_rate": 9.296200512772396e-05, "loss": 0.036, "step": 21240 }, { "grad_norm": 0.31991541385650635, "learning_rate": 9.295354408148668e-05, "loss": 0.0351, "step": 21250 }, { "grad_norm": 0.3408469557762146, "learning_rate": 9.294507833791441e-05, "loss": 0.0372, "step": 21260 }, { "grad_norm": 0.2692928612232208, "learning_rate": 9.293660789793295e-05, "loss": 0.0352, "step": 21270 }, { "grad_norm": 0.3905610740184784, "learning_rate": 9.292813276246858e-05, "loss": 0.0322, "step": 21280 }, { "grad_norm": 0.2614212930202484, "learning_rate": 9.291965293244816e-05, "loss": 0.0351, "step": 21290 }, { "grad_norm": 0.2954288125038147, "learning_rate": 9.291116840879904e-05, "loss": 0.0304, "step": 21300 }, { "grad_norm": 0.23844014108181, "learning_rate": 9.290267919244904e-05, "loss": 0.0315, "step": 21310 }, { "grad_norm": 0.5553337335586548, "learning_rate": 9.289418528432655e-05, "loss": 0.0343, "step": 21320 }, { "grad_norm": 0.2943166196346283, "learning_rate": 9.288568668536045e-05, "loss": 0.0362, "step": 21330 }, { "grad_norm": 0.23342251777648926, "learning_rate": 9.287718339648013e-05, "loss": 0.0426, "step": 21340 }, { "grad_norm": 0.312296986579895, "learning_rate": 9.28686754186155e-05, "loss": 0.0376, "step": 21350 }, { "grad_norm": 0.321835994720459, "learning_rate": 9.286016275269698e-05, "loss": 0.0354, "step": 21360 }, { "grad_norm": 0.29589349031448364, "learning_rate": 9.285164539965551e-05, "loss": 0.0327, "step": 21370 }, { "grad_norm": 0.23153500258922577, "learning_rate": 9.284312336042251e-05, "loss": 0.0339, "step": 21380 }, { "grad_norm": 0.2500738203525543, "learning_rate": 9.283459663592996e-05, "loss": 0.0327, "step": 21390 }, { "grad_norm": 0.3657819926738739, "learning_rate": 9.282606522711033e-05, "loss": 0.0341, "step": 21400 }, { "grad_norm": 0.2597118616104126, "learning_rate": 9.281752913489657e-05, "loss": 0.0313, "step": 21410 }, { "grad_norm": 0.32067006826400757, "learning_rate": 9.280898836022222e-05, "loss": 0.0313, "step": 21420 }, { "grad_norm": 0.23315751552581787, "learning_rate": 9.280044290402126e-05, "loss": 0.0321, "step": 21430 }, { "grad_norm": 0.2641756534576416, "learning_rate": 9.279189276722821e-05, "loss": 0.0312, "step": 21440 }, { "grad_norm": 0.2517768144607544, "learning_rate": 9.278333795077812e-05, "loss": 0.0356, "step": 21450 }, { "grad_norm": 0.37637943029403687, "learning_rate": 9.27747784556065e-05, "loss": 0.0307, "step": 21460 }, { "grad_norm": 0.2983406186103821, "learning_rate": 9.276621428264942e-05, "loss": 0.031, "step": 21470 }, { "grad_norm": 0.2969433069229126, "learning_rate": 9.275764543284345e-05, "loss": 0.0396, "step": 21480 }, { "grad_norm": 0.23869094252586365, "learning_rate": 9.274907190712566e-05, "loss": 0.0348, "step": 21490 }, { "grad_norm": 0.22348454594612122, "learning_rate": 9.274049370643363e-05, "loss": 0.0306, "step": 21500 }, { "grad_norm": 0.25016793608665466, "learning_rate": 9.273191083170547e-05, "loss": 0.0359, "step": 21510 }, { "grad_norm": 0.2504647970199585, "learning_rate": 9.27233232838798e-05, "loss": 0.0323, "step": 21520 }, { "grad_norm": 0.22566980123519897, "learning_rate": 9.27147310638957e-05, "loss": 0.0302, "step": 21530 }, { "grad_norm": 0.28266671299934387, "learning_rate": 9.270613417269286e-05, "loss": 0.0314, "step": 21540 }, { "grad_norm": 0.2590150535106659, "learning_rate": 9.269753261121138e-05, "loss": 0.0359, "step": 21550 }, { "grad_norm": 0.1891588270664215, "learning_rate": 9.268892638039194e-05, "loss": 0.0288, "step": 21560 }, { "grad_norm": 0.2976199984550476, "learning_rate": 9.268031548117569e-05, "loss": 0.0315, "step": 21570 }, { "grad_norm": 0.230403333902359, "learning_rate": 9.26716999145043e-05, "loss": 0.0312, "step": 21580 }, { "grad_norm": 0.23332931101322174, "learning_rate": 9.266307968131998e-05, "loss": 0.0299, "step": 21590 }, { "grad_norm": 0.3413526117801666, "learning_rate": 9.26544547825654e-05, "loss": 0.0372, "step": 21600 }, { "grad_norm": 0.23218685388565063, "learning_rate": 9.264582521918376e-05, "loss": 0.0273, "step": 21610 }, { "grad_norm": 0.3327529728412628, "learning_rate": 9.263719099211881e-05, "loss": 0.0421, "step": 21620 }, { "grad_norm": 0.3095029890537262, "learning_rate": 9.262855210231476e-05, "loss": 0.0373, "step": 21630 }, { "grad_norm": 0.26142361760139465, "learning_rate": 9.261990855071633e-05, "loss": 0.0316, "step": 21640 }, { "grad_norm": 0.24136710166931152, "learning_rate": 9.261126033826878e-05, "loss": 0.0304, "step": 21650 }, { "grad_norm": 0.20871853828430176, "learning_rate": 9.260260746591786e-05, "loss": 0.0339, "step": 21660 }, { "grad_norm": 0.29656392335891724, "learning_rate": 9.259394993460985e-05, "loss": 0.0353, "step": 21670 }, { "grad_norm": 0.3538704514503479, "learning_rate": 9.258528774529151e-05, "loss": 0.0331, "step": 21680 }, { "grad_norm": 0.27682217955589294, "learning_rate": 9.257662089891013e-05, "loss": 0.0338, "step": 21690 }, { "grad_norm": 0.2524005174636841, "learning_rate": 9.25679493964135e-05, "loss": 0.0318, "step": 21700 }, { "grad_norm": 0.29531407356262207, "learning_rate": 9.255927323874994e-05, "loss": 0.0315, "step": 21710 }, { "grad_norm": 0.30856916308403015, "learning_rate": 9.255059242686822e-05, "loss": 0.0345, "step": 21720 }, { "grad_norm": 0.23810972273349762, "learning_rate": 9.254190696171769e-05, "loss": 0.0298, "step": 21730 }, { "grad_norm": 0.33618709444999695, "learning_rate": 9.25332168442482e-05, "loss": 0.0377, "step": 21740 }, { "grad_norm": 0.29083311557769775, "learning_rate": 9.252452207541004e-05, "loss": 0.0359, "step": 21750 }, { "grad_norm": 0.2582060694694519, "learning_rate": 9.251582265615409e-05, "loss": 0.0293, "step": 21760 }, { "grad_norm": 0.23096963763237, "learning_rate": 9.250711858743169e-05, "loss": 0.0299, "step": 21770 }, { "grad_norm": 0.2948416769504547, "learning_rate": 9.24984098701947e-05, "loss": 0.0322, "step": 21780 }, { "grad_norm": 0.21357107162475586, "learning_rate": 9.248969650539552e-05, "loss": 0.0279, "step": 21790 }, { "grad_norm": 0.251380056142807, "learning_rate": 9.2480978493987e-05, "loss": 0.0341, "step": 21800 }, { "grad_norm": 0.2777077853679657, "learning_rate": 9.247225583692256e-05, "loss": 0.0324, "step": 21810 }, { "grad_norm": 0.2941746413707733, "learning_rate": 9.246352853515607e-05, "loss": 0.0311, "step": 21820 }, { "grad_norm": 0.2857985496520996, "learning_rate": 9.245479658964194e-05, "loss": 0.033, "step": 21830 }, { "grad_norm": 0.25114500522613525, "learning_rate": 9.244606000133507e-05, "loss": 0.0301, "step": 21840 }, { "grad_norm": 0.24836958944797516, "learning_rate": 9.24373187711909e-05, "loss": 0.0279, "step": 21850 }, { "grad_norm": 0.20312157273292542, "learning_rate": 9.242857290016537e-05, "loss": 0.033, "step": 21860 }, { "grad_norm": 0.2793785035610199, "learning_rate": 9.241982238921488e-05, "loss": 0.0324, "step": 21870 }, { "grad_norm": 0.28149330615997314, "learning_rate": 9.24110672392964e-05, "loss": 0.0288, "step": 21880 }, { "grad_norm": 0.32050177454948425, "learning_rate": 9.240230745136737e-05, "loss": 0.0378, "step": 21890 }, { "grad_norm": 0.30798691511154175, "learning_rate": 9.239354302638575e-05, "loss": 0.0318, "step": 21900 }, { "grad_norm": 0.2751286029815674, "learning_rate": 9.238477396531e-05, "loss": 0.0357, "step": 21910 }, { "grad_norm": 0.24168969690799713, "learning_rate": 9.23760002690991e-05, "loss": 0.0296, "step": 21920 }, { "grad_norm": 0.24193961918354034, "learning_rate": 9.236722193871252e-05, "loss": 0.0321, "step": 21930 }, { "grad_norm": 0.2563038170337677, "learning_rate": 9.235843897511023e-05, "loss": 0.0331, "step": 21940 }, { "grad_norm": 0.25749436020851135, "learning_rate": 9.234965137925276e-05, "loss": 0.0275, "step": 21950 }, { "grad_norm": 0.2515009045600891, "learning_rate": 9.234085915210108e-05, "loss": 0.0314, "step": 21960 }, { "grad_norm": 0.21073175966739655, "learning_rate": 9.23320622946167e-05, "loss": 0.0318, "step": 21970 }, { "grad_norm": 0.27322545647621155, "learning_rate": 9.232326080776163e-05, "loss": 0.0328, "step": 21980 }, { "grad_norm": 0.259453147649765, "learning_rate": 9.23144546924984e-05, "loss": 0.0313, "step": 21990 }, { "grad_norm": 0.5705735683441162, "learning_rate": 9.230564394979e-05, "loss": 0.038, "step": 22000 }, { "grad_norm": 0.22748860716819763, "learning_rate": 9.22968285806e-05, "loss": 0.0373, "step": 22010 }, { "grad_norm": 0.24004380404949188, "learning_rate": 9.228800858589242e-05, "loss": 0.0319, "step": 22020 }, { "grad_norm": 0.307222455739975, "learning_rate": 9.227918396663179e-05, "loss": 0.0379, "step": 22030 }, { "grad_norm": 0.4289611876010895, "learning_rate": 9.227035472378319e-05, "loss": 0.0389, "step": 22040 }, { "grad_norm": 0.30038538575172424, "learning_rate": 9.226152085831213e-05, "loss": 0.0411, "step": 22050 }, { "grad_norm": 0.3355734944343567, "learning_rate": 9.22526823711847e-05, "loss": 0.034, "step": 22060 }, { "grad_norm": 0.3331252932548523, "learning_rate": 9.224383926336745e-05, "loss": 0.0345, "step": 22070 }, { "grad_norm": 0.5955085158348083, "learning_rate": 9.223499153582744e-05, "loss": 0.0327, "step": 22080 }, { "grad_norm": 0.261843740940094, "learning_rate": 9.222613918953226e-05, "loss": 0.0326, "step": 22090 }, { "grad_norm": 0.33431899547576904, "learning_rate": 9.221728222544999e-05, "loss": 0.0356, "step": 22100 }, { "grad_norm": 0.2809661328792572, "learning_rate": 9.22084206445492e-05, "loss": 0.0262, "step": 22110 }, { "grad_norm": 0.23642560839653015, "learning_rate": 9.2199554447799e-05, "loss": 0.0358, "step": 22120 }, { "grad_norm": 0.22282536327838898, "learning_rate": 9.219068363616897e-05, "loss": 0.0297, "step": 22130 }, { "grad_norm": 0.2609015107154846, "learning_rate": 9.218180821062919e-05, "loss": 0.032, "step": 22140 }, { "grad_norm": 0.33234331011772156, "learning_rate": 9.21729281721503e-05, "loss": 0.0339, "step": 22150 }, { "grad_norm": 0.23161201179027557, "learning_rate": 9.216404352170339e-05, "loss": 0.0383, "step": 22160 }, { "grad_norm": 0.2536731958389282, "learning_rate": 9.215515426026007e-05, "loss": 0.0366, "step": 22170 }, { "grad_norm": 0.2481936812400818, "learning_rate": 9.214626038879246e-05, "loss": 0.0343, "step": 22180 }, { "grad_norm": 0.2907717525959015, "learning_rate": 9.21373619082732e-05, "loss": 0.0351, "step": 22190 }, { "grad_norm": 0.29590174555778503, "learning_rate": 9.212845881967535e-05, "loss": 0.0302, "step": 22200 }, { "grad_norm": 0.36297011375427246, "learning_rate": 9.211955112397262e-05, "loss": 0.0347, "step": 22210 }, { "grad_norm": 0.22655446827411652, "learning_rate": 9.211063882213909e-05, "loss": 0.0381, "step": 22220 }, { "grad_norm": 0.23528333008289337, "learning_rate": 9.210172191514942e-05, "loss": 0.0289, "step": 22230 }, { "grad_norm": 0.20847369730472565, "learning_rate": 9.209280040397874e-05, "loss": 0.03, "step": 22240 }, { "grad_norm": 0.22458770871162415, "learning_rate": 9.208387428960268e-05, "loss": 0.0341, "step": 22250 }, { "grad_norm": 0.3059251606464386, "learning_rate": 9.20749435729974e-05, "loss": 0.0373, "step": 22260 }, { "grad_norm": 0.2932429313659668, "learning_rate": 9.206600825513957e-05, "loss": 0.0337, "step": 22270 }, { "grad_norm": 0.2264278531074524, "learning_rate": 9.20570683370063e-05, "loss": 0.0305, "step": 22280 }, { "grad_norm": 0.21960493922233582, "learning_rate": 9.204812381957528e-05, "loss": 0.0291, "step": 22290 }, { "grad_norm": 0.34569424390792847, "learning_rate": 9.203917470382465e-05, "loss": 0.0346, "step": 22300 }, { "grad_norm": 0.23594744503498077, "learning_rate": 9.203022099073309e-05, "loss": 0.032, "step": 22310 }, { "grad_norm": 0.25766295194625854, "learning_rate": 9.202126268127976e-05, "loss": 0.0354, "step": 22320 }, { "grad_norm": 0.30674856901168823, "learning_rate": 9.20122997764443e-05, "loss": 0.0338, "step": 22330 }, { "grad_norm": 0.22023309767246246, "learning_rate": 9.200333227720692e-05, "loss": 0.0343, "step": 22340 }, { "grad_norm": 0.21349848806858063, "learning_rate": 9.199436018454826e-05, "loss": 0.0351, "step": 22350 }, { "grad_norm": 0.2641364634037018, "learning_rate": 9.198538349944952e-05, "loss": 0.0348, "step": 22360 }, { "grad_norm": 0.23434743285179138, "learning_rate": 9.197640222289234e-05, "loss": 0.031, "step": 22370 }, { "grad_norm": 0.26625844836235046, "learning_rate": 9.196741635585895e-05, "loss": 0.0315, "step": 22380 }, { "grad_norm": 0.2611999213695526, "learning_rate": 9.195842589933199e-05, "loss": 0.0268, "step": 22390 }, { "grad_norm": 0.25883233547210693, "learning_rate": 9.194943085429466e-05, "loss": 0.0295, "step": 22400 }, { "grad_norm": 0.27234065532684326, "learning_rate": 9.194043122173065e-05, "loss": 0.0288, "step": 22410 }, { "grad_norm": 0.26379433274269104, "learning_rate": 9.193142700262413e-05, "loss": 0.0325, "step": 22420 }, { "grad_norm": 0.32260510325431824, "learning_rate": 9.192241819795979e-05, "loss": 0.034, "step": 22430 }, { "grad_norm": 0.3049217760562897, "learning_rate": 9.191340480872284e-05, "loss": 0.028, "step": 22440 }, { "grad_norm": 0.3208746612071991, "learning_rate": 9.190438683589895e-05, "loss": 0.0281, "step": 22450 }, { "grad_norm": 0.25034719705581665, "learning_rate": 9.189536428047432e-05, "loss": 0.0311, "step": 22460 }, { "grad_norm": 0.21332737803459167, "learning_rate": 9.188633714343564e-05, "loss": 0.0324, "step": 22470 }, { "grad_norm": 0.3555116653442383, "learning_rate": 9.18773054257701e-05, "loss": 0.0336, "step": 22480 }, { "grad_norm": 0.3025357723236084, "learning_rate": 9.18682691284654e-05, "loss": 0.03, "step": 22490 }, { "grad_norm": 0.2790542542934418, "learning_rate": 9.185922825250974e-05, "loss": 0.0291, "step": 22500 }, { "grad_norm": 0.2161760777235031, "learning_rate": 9.185018279889181e-05, "loss": 0.0336, "step": 22510 }, { "grad_norm": 0.26434165239334106, "learning_rate": 9.184113276860082e-05, "loss": 0.0321, "step": 22520 }, { "grad_norm": 0.28641510009765625, "learning_rate": 9.183207816262645e-05, "loss": 0.0336, "step": 22530 }, { "grad_norm": 0.262395441532135, "learning_rate": 9.182301898195891e-05, "loss": 0.0292, "step": 22540 }, { "grad_norm": 0.3152386248111725, "learning_rate": 9.181395522758889e-05, "loss": 0.0374, "step": 22550 }, { "grad_norm": 0.25039219856262207, "learning_rate": 9.180488690050759e-05, "loss": 0.0316, "step": 22560 }, { "grad_norm": 0.2610468864440918, "learning_rate": 9.179581400170671e-05, "loss": 0.0339, "step": 22570 }, { "grad_norm": 0.20024578273296356, "learning_rate": 9.178673653217845e-05, "loss": 0.0304, "step": 22580 }, { "grad_norm": 0.2514799237251282, "learning_rate": 9.177765449291551e-05, "loss": 0.0329, "step": 22590 }, { "grad_norm": 0.3230438232421875, "learning_rate": 9.176856788491109e-05, "loss": 0.0311, "step": 22600 }, { "grad_norm": 0.3333125114440918, "learning_rate": 9.175947670915887e-05, "loss": 0.033, "step": 22610 }, { "grad_norm": 0.2163715809583664, "learning_rate": 9.175038096665309e-05, "loss": 0.0323, "step": 22620 }, { "grad_norm": 0.4105238616466522, "learning_rate": 9.17412806583884e-05, "loss": 0.0359, "step": 22630 }, { "grad_norm": 0.25331228971481323, "learning_rate": 9.173217578536002e-05, "loss": 0.0336, "step": 22640 }, { "grad_norm": 0.31540820002555847, "learning_rate": 9.172306634856362e-05, "loss": 0.0324, "step": 22650 }, { "grad_norm": 0.29619672894477844, "learning_rate": 9.171395234899545e-05, "loss": 0.0355, "step": 22660 }, { "grad_norm": 0.2771053910255432, "learning_rate": 9.170483378765214e-05, "loss": 0.0336, "step": 22670 }, { "grad_norm": 0.33219850063323975, "learning_rate": 9.169571066553091e-05, "loss": 0.036, "step": 22680 }, { "grad_norm": 0.30967646837234497, "learning_rate": 9.168658298362946e-05, "loss": 0.0378, "step": 22690 }, { "grad_norm": 0.24166859686374664, "learning_rate": 9.167745074294598e-05, "loss": 0.0332, "step": 22700 }, { "grad_norm": 0.345885694026947, "learning_rate": 9.166831394447913e-05, "loss": 0.0353, "step": 22710 }, { "grad_norm": 0.2680346667766571, "learning_rate": 9.165917258922812e-05, "loss": 0.0321, "step": 22720 }, { "grad_norm": 0.2646850049495697, "learning_rate": 9.165002667819262e-05, "loss": 0.0345, "step": 22730 }, { "grad_norm": 0.2599988579750061, "learning_rate": 9.164087621237282e-05, "loss": 0.0347, "step": 22740 }, { "grad_norm": 0.2460075467824936, "learning_rate": 9.163172119276942e-05, "loss": 0.0299, "step": 22750 }, { "grad_norm": 0.5592795014381409, "learning_rate": 9.162256162038358e-05, "loss": 0.0305, "step": 22760 }, { "grad_norm": 0.25662869215011597, "learning_rate": 9.161339749621698e-05, "loss": 0.029, "step": 22770 }, { "grad_norm": 0.23512418568134308, "learning_rate": 9.160422882127177e-05, "loss": 0.0306, "step": 22780 }, { "grad_norm": 0.24611727893352509, "learning_rate": 9.159505559655069e-05, "loss": 0.0344, "step": 22790 }, { "grad_norm": 0.3186180889606476, "learning_rate": 9.158587782305684e-05, "loss": 0.037, "step": 22800 }, { "grad_norm": 0.23750349879264832, "learning_rate": 9.157669550179391e-05, "loss": 0.0311, "step": 22810 }, { "grad_norm": 0.2657097280025482, "learning_rate": 9.156750863376609e-05, "loss": 0.0366, "step": 22820 }, { "grad_norm": 0.3094010651111603, "learning_rate": 9.155831721997801e-05, "loss": 0.0294, "step": 22830 }, { "grad_norm": 0.28740036487579346, "learning_rate": 9.154912126143484e-05, "loss": 0.0353, "step": 22840 }, { "grad_norm": 0.2735205888748169, "learning_rate": 9.153992075914224e-05, "loss": 0.0338, "step": 22850 }, { "grad_norm": 0.2637150287628174, "learning_rate": 9.153071571410635e-05, "loss": 0.0322, "step": 22860 }, { "grad_norm": 0.3710232973098755, "learning_rate": 9.152150612733384e-05, "loss": 0.0349, "step": 22870 }, { "grad_norm": 0.3183723986148834, "learning_rate": 9.151229199983184e-05, "loss": 0.0333, "step": 22880 }, { "grad_norm": 0.2596432864665985, "learning_rate": 9.150307333260802e-05, "loss": 0.0315, "step": 22890 }, { "grad_norm": 0.20147894322872162, "learning_rate": 9.149385012667048e-05, "loss": 0.0308, "step": 22900 }, { "grad_norm": 0.302336722612381, "learning_rate": 9.148462238302788e-05, "loss": 0.0333, "step": 22910 }, { "grad_norm": 0.32606568932533264, "learning_rate": 9.147539010268936e-05, "loss": 0.0364, "step": 22920 }, { "grad_norm": 0.29516884684562683, "learning_rate": 9.14661532866645e-05, "loss": 0.0321, "step": 22930 }, { "grad_norm": 0.31539902091026306, "learning_rate": 9.145691193596348e-05, "loss": 0.0338, "step": 22940 }, { "grad_norm": 0.2520931363105774, "learning_rate": 9.144766605159691e-05, "loss": 0.0284, "step": 22950 }, { "grad_norm": 0.284973680973053, "learning_rate": 9.14384156345759e-05, "loss": 0.0322, "step": 22960 }, { "grad_norm": 0.29331865906715393, "learning_rate": 9.142916068591204e-05, "loss": 0.0331, "step": 22970 }, { "grad_norm": 0.2929303050041199, "learning_rate": 9.141990120661746e-05, "loss": 0.0333, "step": 22980 }, { "grad_norm": 0.2976960837841034, "learning_rate": 9.141063719770475e-05, "loss": 0.0259, "step": 22990 }, { "grad_norm": 0.3896193206310272, "learning_rate": 9.140136866018704e-05, "loss": 0.0347, "step": 23000 }, { "grad_norm": 0.21854673326015472, "learning_rate": 9.139209559507788e-05, "loss": 0.0361, "step": 23010 }, { "grad_norm": 0.27674829959869385, "learning_rate": 9.13828180033914e-05, "loss": 0.0317, "step": 23020 }, { "grad_norm": 0.18349915742874146, "learning_rate": 9.137353588614212e-05, "loss": 0.0363, "step": 23030 }, { "grad_norm": 0.27902504801750183, "learning_rate": 9.136424924434519e-05, "loss": 0.0416, "step": 23040 }, { "grad_norm": 0.26963669061660767, "learning_rate": 9.135495807901615e-05, "loss": 0.0326, "step": 23050 }, { "grad_norm": 0.2110481560230255, "learning_rate": 9.134566239117108e-05, "loss": 0.0359, "step": 23060 }, { "grad_norm": 0.28238409757614136, "learning_rate": 9.13363621818265e-05, "loss": 0.0359, "step": 23070 }, { "grad_norm": 0.2941235601902008, "learning_rate": 9.132705745199953e-05, "loss": 0.0309, "step": 23080 }, { "grad_norm": 0.35827741026878357, "learning_rate": 9.131774820270768e-05, "loss": 0.0431, "step": 23090 }, { "grad_norm": 0.2624850869178772, "learning_rate": 9.130843443496901e-05, "loss": 0.0317, "step": 23100 }, { "grad_norm": 0.2438155561685562, "learning_rate": 9.129911614980206e-05, "loss": 0.0361, "step": 23110 }, { "grad_norm": 0.3181522786617279, "learning_rate": 9.128979334822584e-05, "loss": 0.029, "step": 23120 }, { "grad_norm": 0.30706384778022766, "learning_rate": 9.128046603125992e-05, "loss": 0.0292, "step": 23130 }, { "grad_norm": 0.28918468952178955, "learning_rate": 9.12711341999243e-05, "loss": 0.0286, "step": 23140 }, { "grad_norm": 0.3100024163722992, "learning_rate": 9.12617978552395e-05, "loss": 0.0364, "step": 23150 }, { "grad_norm": 0.3222707211971283, "learning_rate": 9.12524569982265e-05, "loss": 0.0347, "step": 23160 }, { "grad_norm": 0.2066129595041275, "learning_rate": 9.124311162990684e-05, "loss": 0.0292, "step": 23170 }, { "grad_norm": 0.28686466813087463, "learning_rate": 9.12337617513025e-05, "loss": 0.03, "step": 23180 }, { "grad_norm": 0.2839651107788086, "learning_rate": 9.122440736343596e-05, "loss": 0.032, "step": 23190 }, { "grad_norm": 0.2663485109806061, "learning_rate": 9.12150484673302e-05, "loss": 0.0336, "step": 23200 }, { "grad_norm": 0.28671595454216003, "learning_rate": 9.120568506400873e-05, "loss": 0.0325, "step": 23210 }, { "grad_norm": 0.3685995638370514, "learning_rate": 9.119631715449548e-05, "loss": 0.0343, "step": 23220 }, { "grad_norm": 0.27818673849105835, "learning_rate": 9.118694473981493e-05, "loss": 0.0355, "step": 23230 }, { "grad_norm": 0.21703708171844482, "learning_rate": 9.117756782099203e-05, "loss": 0.0293, "step": 23240 }, { "grad_norm": 0.26320213079452515, "learning_rate": 9.11681863990522e-05, "loss": 0.0343, "step": 23250 }, { "grad_norm": 0.2915642559528351, "learning_rate": 9.115880047502142e-05, "loss": 0.0337, "step": 23260 }, { "grad_norm": 0.2732219099998474, "learning_rate": 9.114941004992609e-05, "loss": 0.0311, "step": 23270 }, { "grad_norm": 0.2957462668418884, "learning_rate": 9.114001512479317e-05, "loss": 0.0299, "step": 23280 }, { "grad_norm": 0.31195226311683655, "learning_rate": 9.113061570065003e-05, "loss": 0.0341, "step": 23290 }, { "grad_norm": 0.3252130448818207, "learning_rate": 9.112121177852459e-05, "loss": 0.0379, "step": 23300 }, { "grad_norm": 0.32155174016952515, "learning_rate": 9.111180335944527e-05, "loss": 0.0312, "step": 23310 }, { "grad_norm": 0.3006870448589325, "learning_rate": 9.110239044444093e-05, "loss": 0.0327, "step": 23320 }, { "grad_norm": 0.3101575970649719, "learning_rate": 9.109297303454099e-05, "loss": 0.0377, "step": 23330 }, { "grad_norm": 0.23698736727237701, "learning_rate": 9.108355113077526e-05, "loss": 0.0353, "step": 23340 }, { "grad_norm": 0.3533897399902344, "learning_rate": 9.107412473417419e-05, "loss": 0.0336, "step": 23350 }, { "grad_norm": 0.32754695415496826, "learning_rate": 9.106469384576858e-05, "loss": 0.0288, "step": 23360 }, { "grad_norm": 0.2629201114177704, "learning_rate": 9.105525846658978e-05, "loss": 0.0262, "step": 23370 }, { "grad_norm": 0.35005900263786316, "learning_rate": 9.104581859766965e-05, "loss": 0.0336, "step": 23380 }, { "grad_norm": 0.26237404346466064, "learning_rate": 9.10363742400405e-05, "loss": 0.0291, "step": 23390 }, { "grad_norm": 0.24552644789218903, "learning_rate": 9.102692539473518e-05, "loss": 0.0307, "step": 23400 }, { "grad_norm": 0.2859744131565094, "learning_rate": 9.101747206278697e-05, "loss": 0.0326, "step": 23410 }, { "grad_norm": 0.28530746698379517, "learning_rate": 9.100801424522968e-05, "loss": 0.0301, "step": 23420 }, { "grad_norm": 0.28547555208206177, "learning_rate": 9.099855194309762e-05, "loss": 0.031, "step": 23430 }, { "grad_norm": 0.2506678104400635, "learning_rate": 9.098908515742554e-05, "loss": 0.0331, "step": 23440 }, { "grad_norm": 0.28299781680107117, "learning_rate": 9.097961388924873e-05, "loss": 0.0306, "step": 23450 }, { "grad_norm": 0.2684740722179413, "learning_rate": 9.097013813960298e-05, "loss": 0.0327, "step": 23460 }, { "grad_norm": 0.26576945185661316, "learning_rate": 9.09606579095245e-05, "loss": 0.033, "step": 23470 }, { "grad_norm": 0.21528972685337067, "learning_rate": 9.095117320005008e-05, "loss": 0.0288, "step": 23480 }, { "grad_norm": 0.28824201226234436, "learning_rate": 9.094168401221691e-05, "loss": 0.0288, "step": 23490 }, { "grad_norm": 0.25340914726257324, "learning_rate": 9.093219034706273e-05, "loss": 0.0342, "step": 23500 }, { "grad_norm": 0.33185964822769165, "learning_rate": 9.092269220562577e-05, "loss": 0.033, "step": 23510 }, { "grad_norm": 0.2873094975948334, "learning_rate": 9.09131895889447e-05, "loss": 0.0327, "step": 23520 }, { "grad_norm": 0.24931995570659637, "learning_rate": 9.090368249805873e-05, "loss": 0.0327, "step": 23530 }, { "grad_norm": 0.24115003645420074, "learning_rate": 9.089417093400754e-05, "loss": 0.0333, "step": 23540 }, { "grad_norm": 0.24252040684223175, "learning_rate": 9.088465489783131e-05, "loss": 0.0324, "step": 23550 }, { "grad_norm": 0.2123619019985199, "learning_rate": 9.087513439057068e-05, "loss": 0.0306, "step": 23560 }, { "grad_norm": 0.24543674290180206, "learning_rate": 9.08656094132668e-05, "loss": 0.0279, "step": 23570 }, { "grad_norm": 0.2832995653152466, "learning_rate": 9.085607996696134e-05, "loss": 0.0302, "step": 23580 }, { "grad_norm": 0.32709765434265137, "learning_rate": 9.084654605269639e-05, "loss": 0.0303, "step": 23590 }, { "grad_norm": 0.3358708322048187, "learning_rate": 9.083700767151457e-05, "loss": 0.0298, "step": 23600 }, { "grad_norm": 0.3255902826786041, "learning_rate": 9.082746482445898e-05, "loss": 0.0359, "step": 23610 }, { "grad_norm": 0.26755091547966003, "learning_rate": 9.081791751257325e-05, "loss": 0.0291, "step": 23620 }, { "grad_norm": 0.2702423632144928, "learning_rate": 9.080836573690142e-05, "loss": 0.0266, "step": 23630 }, { "grad_norm": 0.2543042302131653, "learning_rate": 9.079880949848805e-05, "loss": 0.0308, "step": 23640 }, { "grad_norm": 0.22319136559963226, "learning_rate": 9.078924879837822e-05, "loss": 0.0298, "step": 23650 }, { "grad_norm": 0.26765239238739014, "learning_rate": 9.077968363761747e-05, "loss": 0.0304, "step": 23660 }, { "grad_norm": 0.2917574942111969, "learning_rate": 9.077011401725182e-05, "loss": 0.0305, "step": 23670 }, { "grad_norm": 0.3141767382621765, "learning_rate": 9.07605399383278e-05, "loss": 0.0295, "step": 23680 }, { "grad_norm": 0.32890015840530396, "learning_rate": 9.075096140189243e-05, "loss": 0.0317, "step": 23690 }, { "grad_norm": 0.22289834916591644, "learning_rate": 9.074137840899318e-05, "loss": 0.0327, "step": 23700 }, { "grad_norm": 0.244684636592865, "learning_rate": 9.073179096067804e-05, "loss": 0.03, "step": 23710 }, { "grad_norm": 0.2432366907596588, "learning_rate": 9.072219905799549e-05, "loss": 0.0283, "step": 23720 }, { "grad_norm": 0.2924101948738098, "learning_rate": 9.071260270199447e-05, "loss": 0.0361, "step": 23730 }, { "grad_norm": 0.2353994995355606, "learning_rate": 9.070300189372441e-05, "loss": 0.0321, "step": 23740 }, { "grad_norm": 0.22242097556591034, "learning_rate": 9.069339663423528e-05, "loss": 0.0341, "step": 23750 }, { "grad_norm": 0.24729008972644806, "learning_rate": 9.068378692457747e-05, "loss": 0.0321, "step": 23760 }, { "grad_norm": 0.3082471787929535, "learning_rate": 9.067417276580189e-05, "loss": 0.0354, "step": 23770 }, { "grad_norm": 0.25974634289741516, "learning_rate": 9.066455415895993e-05, "loss": 0.0372, "step": 23780 }, { "grad_norm": 0.23926036059856415, "learning_rate": 9.065493110510346e-05, "loss": 0.0355, "step": 23790 }, { "grad_norm": 0.3379344344139099, "learning_rate": 9.064530360528484e-05, "loss": 0.0342, "step": 23800 }, { "grad_norm": 0.29010066390037537, "learning_rate": 9.063567166055695e-05, "loss": 0.0332, "step": 23810 }, { "grad_norm": 0.27317824959754944, "learning_rate": 9.062603527197308e-05, "loss": 0.0345, "step": 23820 }, { "grad_norm": 0.2406795173883438, "learning_rate": 9.06163944405871e-05, "loss": 0.0315, "step": 23830 }, { "grad_norm": 0.29529768228530884, "learning_rate": 9.060674916745327e-05, "loss": 0.0283, "step": 23840 }, { "grad_norm": 0.30513378977775574, "learning_rate": 9.05970994536264e-05, "loss": 0.0286, "step": 23850 }, { "grad_norm": 0.3080386221408844, "learning_rate": 9.05874453001618e-05, "loss": 0.0303, "step": 23860 }, { "grad_norm": 0.26914912462234497, "learning_rate": 9.057778670811517e-05, "loss": 0.0277, "step": 23870 }, { "grad_norm": 0.20994789898395538, "learning_rate": 9.056812367854281e-05, "loss": 0.0308, "step": 23880 }, { "grad_norm": 0.21059778332710266, "learning_rate": 9.055845621250143e-05, "loss": 0.0278, "step": 23890 }, { "grad_norm": 0.395325243473053, "learning_rate": 9.054878431104825e-05, "loss": 0.0298, "step": 23900 }, { "grad_norm": 0.24626916646957397, "learning_rate": 9.0539107975241e-05, "loss": 0.0309, "step": 23910 }, { "grad_norm": 0.2598765194416046, "learning_rate": 9.052942720613784e-05, "loss": 0.0292, "step": 23920 }, { "grad_norm": 0.20923520624637604, "learning_rate": 9.051974200479745e-05, "loss": 0.0286, "step": 23930 }, { "grad_norm": 0.23621512949466705, "learning_rate": 9.051005237227901e-05, "loss": 0.0281, "step": 23940 }, { "grad_norm": 0.24297845363616943, "learning_rate": 9.050035830964215e-05, "loss": 0.0315, "step": 23950 }, { "grad_norm": 0.3165528476238251, "learning_rate": 9.049065981794698e-05, "loss": 0.0301, "step": 23960 }, { "grad_norm": 0.3168933093547821, "learning_rate": 9.048095689825414e-05, "loss": 0.0302, "step": 23970 }, { "grad_norm": 0.26319247484207153, "learning_rate": 9.047124955162472e-05, "loss": 0.0304, "step": 23980 }, { "grad_norm": 0.256902277469635, "learning_rate": 9.046153777912028e-05, "loss": 0.031, "step": 23990 }, { "grad_norm": 0.2857876121997833, "learning_rate": 9.045182158180292e-05, "loss": 0.0381, "step": 24000 }, { "grad_norm": 0.2713342607021332, "learning_rate": 9.044210096073516e-05, "loss": 0.0275, "step": 24010 }, { "grad_norm": 0.23331187665462494, "learning_rate": 9.043237591698004e-05, "loss": 0.027, "step": 24020 }, { "grad_norm": 0.30637794733047485, "learning_rate": 9.04226464516011e-05, "loss": 0.0306, "step": 24030 }, { "grad_norm": 0.23940148949623108, "learning_rate": 9.041291256566229e-05, "loss": 0.0269, "step": 24040 }, { "grad_norm": 0.3006315529346466, "learning_rate": 9.040317426022814e-05, "loss": 0.0279, "step": 24050 }, { "grad_norm": 0.37845221161842346, "learning_rate": 9.03934315363636e-05, "loss": 0.0286, "step": 24060 }, { "grad_norm": 0.27635321021080017, "learning_rate": 9.038368439513409e-05, "loss": 0.0299, "step": 24070 }, { "grad_norm": 0.2868705987930298, "learning_rate": 9.03739328376056e-05, "loss": 0.0315, "step": 24080 }, { "grad_norm": 0.24029897153377533, "learning_rate": 9.036417686484451e-05, "loss": 0.0368, "step": 24090 }, { "grad_norm": 0.3582642078399658, "learning_rate": 9.035441647791773e-05, "loss": 0.0316, "step": 24100 }, { "grad_norm": 0.3779372274875641, "learning_rate": 9.034465167789263e-05, "loss": 0.0335, "step": 24110 }, { "grad_norm": 0.2649083435535431, "learning_rate": 9.033488246583706e-05, "loss": 0.0306, "step": 24120 }, { "grad_norm": 0.23120276629924774, "learning_rate": 9.032510884281941e-05, "loss": 0.0318, "step": 24130 }, { "grad_norm": 0.26619434356689453, "learning_rate": 9.031533080990848e-05, "loss": 0.0297, "step": 24140 }, { "grad_norm": 0.2602007985115051, "learning_rate": 9.030554836817358e-05, "loss": 0.0328, "step": 24150 }, { "grad_norm": 0.19406452775001526, "learning_rate": 9.029576151868451e-05, "loss": 0.0273, "step": 24160 }, { "grad_norm": 0.24482180178165436, "learning_rate": 9.028597026251155e-05, "loss": 0.0291, "step": 24170 }, { "grad_norm": 0.2727980315685272, "learning_rate": 9.027617460072547e-05, "loss": 0.0275, "step": 24180 }, { "grad_norm": 0.25771450996398926, "learning_rate": 9.026637453439745e-05, "loss": 0.027, "step": 24190 }, { "grad_norm": 0.1971546709537506, "learning_rate": 9.025657006459927e-05, "loss": 0.0302, "step": 24200 }, { "grad_norm": 0.26784858107566833, "learning_rate": 9.024676119240311e-05, "loss": 0.0238, "step": 24210 }, { "grad_norm": 0.24050457775592804, "learning_rate": 9.023694791888166e-05, "loss": 0.0302, "step": 24220 }, { "grad_norm": 0.3330393135547638, "learning_rate": 9.022713024510808e-05, "loss": 0.0378, "step": 24230 }, { "grad_norm": 0.4009268283843994, "learning_rate": 9.021730817215601e-05, "loss": 0.0353, "step": 24240 }, { "grad_norm": 0.33541813492774963, "learning_rate": 9.02074817010996e-05, "loss": 0.0341, "step": 24250 }, { "grad_norm": 0.3008006811141968, "learning_rate": 9.019765083301342e-05, "loss": 0.0319, "step": 24260 }, { "grad_norm": 0.2442713975906372, "learning_rate": 9.01878155689726e-05, "loss": 0.0313, "step": 24270 }, { "grad_norm": 0.24278214573860168, "learning_rate": 9.017797591005268e-05, "loss": 0.0332, "step": 24280 }, { "grad_norm": 0.2694127857685089, "learning_rate": 9.016813185732972e-05, "loss": 0.0302, "step": 24290 }, { "grad_norm": 0.28702312707901, "learning_rate": 9.015828341188027e-05, "loss": 0.0351, "step": 24300 }, { "grad_norm": 0.309924453496933, "learning_rate": 9.01484305747813e-05, "loss": 0.0386, "step": 24310 }, { "grad_norm": 0.31165358424186707, "learning_rate": 9.013857334711033e-05, "loss": 0.0329, "step": 24320 }, { "grad_norm": 0.3174089193344116, "learning_rate": 9.012871172994534e-05, "loss": 0.0297, "step": 24330 }, { "grad_norm": 0.28190791606903076, "learning_rate": 9.011884572436476e-05, "loss": 0.0373, "step": 24340 }, { "grad_norm": 0.2940385043621063, "learning_rate": 9.010897533144754e-05, "loss": 0.0241, "step": 24350 }, { "grad_norm": 0.28557050228118896, "learning_rate": 9.009910055227306e-05, "loss": 0.0287, "step": 24360 }, { "grad_norm": 0.25874385237693787, "learning_rate": 9.008922138792124e-05, "loss": 0.0308, "step": 24370 }, { "grad_norm": 0.3125855326652527, "learning_rate": 9.007933783947244e-05, "loss": 0.0282, "step": 24380 }, { "grad_norm": 0.24212850630283356, "learning_rate": 9.006944990800752e-05, "loss": 0.0272, "step": 24390 }, { "grad_norm": 0.22712677717208862, "learning_rate": 9.005955759460779e-05, "loss": 0.0266, "step": 24400 }, { "grad_norm": 0.29242441058158875, "learning_rate": 9.004966090035508e-05, "loss": 0.0406, "step": 24410 }, { "grad_norm": 0.26839324831962585, "learning_rate": 9.003975982633166e-05, "loss": 0.0327, "step": 24420 }, { "grad_norm": 0.287761390209198, "learning_rate": 9.00298543736203e-05, "loss": 0.0323, "step": 24430 }, { "grad_norm": 0.22247812151908875, "learning_rate": 9.001994454330427e-05, "loss": 0.0306, "step": 24440 }, { "grad_norm": 0.22773684561252594, "learning_rate": 9.001003033646727e-05, "loss": 0.0369, "step": 24450 }, { "grad_norm": 0.30217602849006653, "learning_rate": 9.00001117541935e-05, "loss": 0.0299, "step": 24460 }, { "grad_norm": 0.23553037643432617, "learning_rate": 8.999018879756764e-05, "loss": 0.0301, "step": 24470 }, { "grad_norm": 0.2953532636165619, "learning_rate": 8.998026146767487e-05, "loss": 0.0316, "step": 24480 }, { "grad_norm": 0.2423483282327652, "learning_rate": 8.99703297656008e-05, "loss": 0.0296, "step": 24490 }, { "grad_norm": 0.2650439143180847, "learning_rate": 8.996039369243156e-05, "loss": 0.0345, "step": 24500 }, { "grad_norm": 0.29350391030311584, "learning_rate": 8.995045324925378e-05, "loss": 0.0323, "step": 24510 }, { "grad_norm": 0.3005538284778595, "learning_rate": 8.994050843715448e-05, "loss": 0.0284, "step": 24520 }, { "grad_norm": 0.2725777328014374, "learning_rate": 8.993055925722121e-05, "loss": 0.0308, "step": 24530 }, { "grad_norm": 0.22446490824222565, "learning_rate": 8.992060571054202e-05, "loss": 0.0322, "step": 24540 }, { "grad_norm": 0.3931368291378021, "learning_rate": 8.991064779820542e-05, "loss": 0.0274, "step": 24550 }, { "grad_norm": 0.23482821881771088, "learning_rate": 8.990068552130036e-05, "loss": 0.0317, "step": 24560 }, { "grad_norm": 0.27113449573516846, "learning_rate": 8.989071888091634e-05, "loss": 0.0293, "step": 24570 }, { "grad_norm": 0.28357332944869995, "learning_rate": 8.988074787814329e-05, "loss": 0.0304, "step": 24580 }, { "grad_norm": 0.3188905715942383, "learning_rate": 8.987077251407158e-05, "loss": 0.0343, "step": 24590 }, { "grad_norm": 0.27490517497062683, "learning_rate": 8.986079278979216e-05, "loss": 0.0269, "step": 24600 }, { "grad_norm": 0.2574551999568939, "learning_rate": 8.985080870639635e-05, "loss": 0.0274, "step": 24610 }, { "grad_norm": 0.28671273589134216, "learning_rate": 8.984082026497603e-05, "loss": 0.0318, "step": 24620 }, { "grad_norm": 0.3030722737312317, "learning_rate": 8.98308274666235e-05, "loss": 0.0303, "step": 24630 }, { "grad_norm": 0.2950882911682129, "learning_rate": 8.982083031243155e-05, "loss": 0.0298, "step": 24640 }, { "grad_norm": 0.22330863773822784, "learning_rate": 8.98108288034935e-05, "loss": 0.028, "step": 24650 }, { "grad_norm": 0.212008997797966, "learning_rate": 8.980082294090305e-05, "loss": 0.0299, "step": 24660 }, { "grad_norm": 0.26501113176345825, "learning_rate": 8.979081272575443e-05, "loss": 0.031, "step": 24670 }, { "grad_norm": 0.3196823298931122, "learning_rate": 8.978079815914236e-05, "loss": 0.0325, "step": 24680 }, { "grad_norm": 0.2474205195903778, "learning_rate": 8.977077924216202e-05, "loss": 0.0292, "step": 24690 }, { "grad_norm": 0.26258713006973267, "learning_rate": 8.976075597590905e-05, "loss": 0.0306, "step": 24700 }, { "grad_norm": 0.27586475014686584, "learning_rate": 8.975072836147958e-05, "loss": 0.0328, "step": 24710 }, { "grad_norm": 0.25784924626350403, "learning_rate": 8.974069639997025e-05, "loss": 0.0292, "step": 24720 }, { "grad_norm": 0.2878739833831787, "learning_rate": 8.973066009247808e-05, "loss": 0.0348, "step": 24730 }, { "grad_norm": 0.25415945053100586, "learning_rate": 8.972061944010066e-05, "loss": 0.0309, "step": 24740 }, { "grad_norm": 0.25111737847328186, "learning_rate": 8.971057444393603e-05, "loss": 0.0297, "step": 24750 }, { "grad_norm": 0.3462611436843872, "learning_rate": 8.970052510508268e-05, "loss": 0.028, "step": 24760 }, { "grad_norm": 0.3669081926345825, "learning_rate": 8.969047142463959e-05, "loss": 0.0356, "step": 24770 }, { "grad_norm": 0.33997419476509094, "learning_rate": 8.968041340370621e-05, "loss": 0.0365, "step": 24780 }, { "grad_norm": 0.23579488694667816, "learning_rate": 8.96703510433825e-05, "loss": 0.035, "step": 24790 }, { "grad_norm": 0.28108686208724976, "learning_rate": 8.966028434476883e-05, "loss": 0.0353, "step": 24800 }, { "grad_norm": 0.2461758852005005, "learning_rate": 8.96502133089661e-05, "loss": 0.0293, "step": 24810 }, { "grad_norm": 0.23696257174015045, "learning_rate": 8.964013793707564e-05, "loss": 0.0298, "step": 24820 }, { "grad_norm": 0.2896569073200226, "learning_rate": 8.963005823019932e-05, "loss": 0.0294, "step": 24830 }, { "grad_norm": 0.4691542088985443, "learning_rate": 8.961997418943939e-05, "loss": 0.0321, "step": 24840 }, { "grad_norm": 0.27909913659095764, "learning_rate": 8.960988581589865e-05, "loss": 0.029, "step": 24850 }, { "grad_norm": 0.30534231662750244, "learning_rate": 8.959979311068037e-05, "loss": 0.0328, "step": 24860 }, { "grad_norm": 0.29984182119369507, "learning_rate": 8.958969607488823e-05, "loss": 0.0365, "step": 24870 }, { "grad_norm": 0.2450684756040573, "learning_rate": 8.957959470962647e-05, "loss": 0.0351, "step": 24880 }, { "grad_norm": 0.24917003512382507, "learning_rate": 8.956948901599971e-05, "loss": 0.0313, "step": 24890 }, { "grad_norm": 0.2801682651042938, "learning_rate": 8.955937899511315e-05, "loss": 0.0338, "step": 24900 }, { "grad_norm": 0.2647407054901123, "learning_rate": 8.954926464807238e-05, "loss": 0.0308, "step": 24910 }, { "grad_norm": 0.2774140536785126, "learning_rate": 8.953914597598347e-05, "loss": 0.031, "step": 24920 }, { "grad_norm": 0.24590572714805603, "learning_rate": 8.952902297995303e-05, "loss": 0.0367, "step": 24930 }, { "grad_norm": 0.26165908575057983, "learning_rate": 8.951889566108804e-05, "loss": 0.0272, "step": 24940 }, { "grad_norm": 0.23862454295158386, "learning_rate": 8.950876402049606e-05, "loss": 0.0308, "step": 24950 }, { "grad_norm": 0.2730882465839386, "learning_rate": 8.949862805928504e-05, "loss": 0.029, "step": 24960 }, { "grad_norm": 0.2825525403022766, "learning_rate": 8.948848777856343e-05, "loss": 0.037, "step": 24970 }, { "grad_norm": 0.3002350330352783, "learning_rate": 8.947834317944017e-05, "loss": 0.0336, "step": 24980 }, { "grad_norm": 0.2297813594341278, "learning_rate": 8.946819426302466e-05, "loss": 0.0317, "step": 24990 }, { "grad_norm": 0.325875461101532, "learning_rate": 8.945804103042676e-05, "loss": 0.0316, "step": 25000 }, { "grad_norm": 0.30237385630607605, "learning_rate": 8.944788348275681e-05, "loss": 0.0275, "step": 25010 }, { "grad_norm": 0.26680895686149597, "learning_rate": 8.943772162112565e-05, "loss": 0.03, "step": 25020 }, { "grad_norm": 0.32211756706237793, "learning_rate": 8.942755544664454e-05, "loss": 0.03, "step": 25030 }, { "grad_norm": 0.3046349585056305, "learning_rate": 8.941738496042525e-05, "loss": 0.0299, "step": 25040 }, { "grad_norm": 0.2519265413284302, "learning_rate": 8.940721016357999e-05, "loss": 0.0284, "step": 25050 }, { "grad_norm": 0.26642537117004395, "learning_rate": 8.939703105722148e-05, "loss": 0.0349, "step": 25060 }, { "grad_norm": 0.2589081823825836, "learning_rate": 8.93868476424629e-05, "loss": 0.0348, "step": 25070 }, { "grad_norm": 0.28290072083473206, "learning_rate": 8.937665992041786e-05, "loss": 0.0329, "step": 25080 }, { "grad_norm": 0.3262339234352112, "learning_rate": 8.93664678922005e-05, "loss": 0.0329, "step": 25090 }, { "grad_norm": 0.2534581124782562, "learning_rate": 8.93562715589254e-05, "loss": 0.0311, "step": 25100 }, { "grad_norm": 0.20673954486846924, "learning_rate": 8.934607092170762e-05, "loss": 0.029, "step": 25110 }, { "grad_norm": 0.2348257452249527, "learning_rate": 8.933586598166266e-05, "loss": 0.0412, "step": 25120 }, { "grad_norm": 0.22465947270393372, "learning_rate": 8.932565673990655e-05, "loss": 0.0322, "step": 25130 }, { "grad_norm": 0.2563099265098572, "learning_rate": 8.931544319755574e-05, "loss": 0.0312, "step": 25140 }, { "grad_norm": 0.21277262270450592, "learning_rate": 8.930522535572718e-05, "loss": 0.0295, "step": 25150 }, { "grad_norm": 0.23043681681156158, "learning_rate": 8.929500321553826e-05, "loss": 0.0322, "step": 25160 }, { "grad_norm": 0.27147096395492554, "learning_rate": 8.928477677810686e-05, "loss": 0.0312, "step": 25170 }, { "grad_norm": 0.2889135777950287, "learning_rate": 8.927454604455137e-05, "loss": 0.0335, "step": 25180 }, { "grad_norm": 0.23583486676216125, "learning_rate": 8.926431101599053e-05, "loss": 0.0313, "step": 25190 }, { "grad_norm": 0.3398309648036957, "learning_rate": 8.925407169354369e-05, "loss": 0.0309, "step": 25200 }, { "grad_norm": 0.2864283323287964, "learning_rate": 8.92438280783306e-05, "loss": 0.03, "step": 25210 }, { "grad_norm": 0.30148938298225403, "learning_rate": 8.923358017147146e-05, "loss": 0.0306, "step": 25220 }, { "grad_norm": 0.35097959637641907, "learning_rate": 8.922332797408697e-05, "loss": 0.0296, "step": 25230 }, { "grad_norm": 0.29249322414398193, "learning_rate": 8.921307148729831e-05, "loss": 0.0312, "step": 25240 }, { "grad_norm": 0.22055502235889435, "learning_rate": 8.920281071222712e-05, "loss": 0.0277, "step": 25250 }, { "grad_norm": 0.2844606637954712, "learning_rate": 8.919254564999548e-05, "loss": 0.0295, "step": 25260 }, { "grad_norm": 0.193317711353302, "learning_rate": 8.918227630172598e-05, "loss": 0.0311, "step": 25270 }, { "grad_norm": 0.22869732975959778, "learning_rate": 8.917200266854165e-05, "loss": 0.0292, "step": 25280 }, { "grad_norm": 0.2525438070297241, "learning_rate": 8.9161724751566e-05, "loss": 0.0294, "step": 25290 }, { "grad_norm": 0.21981975436210632, "learning_rate": 8.915144255192302e-05, "loss": 0.029, "step": 25300 }, { "grad_norm": 0.36073869466781616, "learning_rate": 8.914115607073714e-05, "loss": 0.0336, "step": 25310 }, { "grad_norm": 0.45496639609336853, "learning_rate": 8.913086530913327e-05, "loss": 0.0399, "step": 25320 }, { "grad_norm": 0.32555046677589417, "learning_rate": 8.912057026823681e-05, "loss": 0.0355, "step": 25330 }, { "grad_norm": 0.31546667218208313, "learning_rate": 8.91102709491736e-05, "loss": 0.0319, "step": 25340 }, { "grad_norm": 0.27412816882133484, "learning_rate": 8.909996735306996e-05, "loss": 0.0352, "step": 25350 }, { "grad_norm": 0.2499939203262329, "learning_rate": 8.908965948105268e-05, "loss": 0.0269, "step": 25360 }, { "grad_norm": 0.2555117905139923, "learning_rate": 8.907934733424901e-05, "loss": 0.031, "step": 25370 }, { "grad_norm": 0.2106734812259674, "learning_rate": 8.906903091378666e-05, "loss": 0.0285, "step": 25380 }, { "grad_norm": 0.24569249153137207, "learning_rate": 8.905871022079384e-05, "loss": 0.0281, "step": 25390 }, { "grad_norm": 0.3363354206085205, "learning_rate": 8.90483852563992e-05, "loss": 0.0345, "step": 25400 }, { "grad_norm": 0.26837852597236633, "learning_rate": 8.903805602173185e-05, "loss": 0.0302, "step": 25410 }, { "grad_norm": 0.29707399010658264, "learning_rate": 8.902772251792137e-05, "loss": 0.0372, "step": 25420 }, { "grad_norm": 0.2970879077911377, "learning_rate": 8.901738474609786e-05, "loss": 0.0312, "step": 25430 }, { "grad_norm": 0.26963120698928833, "learning_rate": 8.900704270739179e-05, "loss": 0.0319, "step": 25440 }, { "grad_norm": 0.2497844696044922, "learning_rate": 8.89966964029342e-05, "loss": 0.0333, "step": 25450 }, { "grad_norm": 0.2321600317955017, "learning_rate": 8.898634583385652e-05, "loss": 0.0312, "step": 25460 }, { "grad_norm": 0.2175353765487671, "learning_rate": 8.897599100129065e-05, "loss": 0.0252, "step": 25470 }, { "grad_norm": 0.19714458286762238, "learning_rate": 8.896563190636903e-05, "loss": 0.0278, "step": 25480 }, { "grad_norm": 0.2689904570579529, "learning_rate": 8.895526855022448e-05, "loss": 0.0271, "step": 25490 }, { "grad_norm": 0.23632211983203888, "learning_rate": 8.894490093399033e-05, "loss": 0.029, "step": 25500 }, { "grad_norm": 0.2644004821777344, "learning_rate": 8.893452905880035e-05, "loss": 0.0329, "step": 25510 }, { "grad_norm": 0.25316768884658813, "learning_rate": 8.892415292578883e-05, "loss": 0.0288, "step": 25520 }, { "grad_norm": 0.2595835328102112, "learning_rate": 8.891377253609046e-05, "loss": 0.0291, "step": 25530 }, { "grad_norm": 0.2246798574924469, "learning_rate": 8.890338789084043e-05, "loss": 0.0302, "step": 25540 }, { "grad_norm": 0.21187050640583038, "learning_rate": 8.88929989911744e-05, "loss": 0.0248, "step": 25550 }, { "grad_norm": 0.2816685438156128, "learning_rate": 8.888260583822847e-05, "loss": 0.0285, "step": 25560 }, { "grad_norm": 0.2627944350242615, "learning_rate": 8.887220843313921e-05, "loss": 0.0303, "step": 25570 }, { "grad_norm": 0.23619690537452698, "learning_rate": 8.88618067770437e-05, "loss": 0.0316, "step": 25580 }, { "grad_norm": 0.22691987454891205, "learning_rate": 8.885140087107942e-05, "loss": 0.0309, "step": 25590 }, { "grad_norm": 0.27967754006385803, "learning_rate": 8.884099071638436e-05, "loss": 0.0322, "step": 25600 }, { "grad_norm": 0.2969549596309662, "learning_rate": 8.883057631409695e-05, "loss": 0.0354, "step": 25610 }, { "grad_norm": 0.32082247734069824, "learning_rate": 8.882015766535608e-05, "loss": 0.0297, "step": 25620 }, { "grad_norm": 0.24516086280345917, "learning_rate": 8.880973477130115e-05, "loss": 0.0267, "step": 25630 }, { "grad_norm": 0.22009113430976868, "learning_rate": 8.879930763307197e-05, "loss": 0.0271, "step": 25640 }, { "grad_norm": 0.5068790316581726, "learning_rate": 8.878887625180884e-05, "loss": 0.0366, "step": 25650 }, { "grad_norm": 0.19887706637382507, "learning_rate": 8.877844062865253e-05, "loss": 0.0329, "step": 25660 }, { "grad_norm": 0.29716458916664124, "learning_rate": 8.876800076474424e-05, "loss": 0.0338, "step": 25670 }, { "grad_norm": 0.23181982338428497, "learning_rate": 8.875755666122568e-05, "loss": 0.028, "step": 25680 }, { "grad_norm": 0.23971828818321228, "learning_rate": 8.8747108319239e-05, "loss": 0.0288, "step": 25690 }, { "grad_norm": 0.2299007624387741, "learning_rate": 8.87366557399268e-05, "loss": 0.0293, "step": 25700 }, { "grad_norm": 0.26775187253952026, "learning_rate": 8.872619892443217e-05, "loss": 0.0325, "step": 25710 }, { "grad_norm": 0.924479603767395, "learning_rate": 8.871573787389865e-05, "loss": 0.0389, "step": 25720 }, { "grad_norm": 0.30040305852890015, "learning_rate": 8.870527258947024e-05, "loss": 0.0409, "step": 25730 }, { "grad_norm": 0.35683926939964294, "learning_rate": 8.869480307229143e-05, "loss": 0.0407, "step": 25740 }, { "grad_norm": 0.3454369902610779, "learning_rate": 8.868432932350712e-05, "loss": 0.0329, "step": 25750 }, { "grad_norm": 0.41862884163856506, "learning_rate": 8.867385134426272e-05, "loss": 0.0355, "step": 25760 }, { "grad_norm": 0.304065465927124, "learning_rate": 8.866336913570407e-05, "loss": 0.0293, "step": 25770 }, { "grad_norm": 0.273001104593277, "learning_rate": 8.865288269897751e-05, "loss": 0.0326, "step": 25780 }, { "grad_norm": 0.31616997718811035, "learning_rate": 8.864239203522981e-05, "loss": 0.03, "step": 25790 }, { "grad_norm": 0.29316356778144836, "learning_rate": 8.863189714560822e-05, "loss": 0.0336, "step": 25800 }, { "grad_norm": 0.2692314684391022, "learning_rate": 8.862139803126043e-05, "loss": 0.0301, "step": 25810 }, { "grad_norm": 0.25977349281311035, "learning_rate": 8.861089469333463e-05, "loss": 0.0291, "step": 25820 }, { "grad_norm": 0.34894686937332153, "learning_rate": 8.860038713297944e-05, "loss": 0.032, "step": 25830 }, { "grad_norm": 0.6267641186714172, "learning_rate": 8.858987535134394e-05, "loss": 0.0414, "step": 25840 }, { "grad_norm": 0.2938433289527893, "learning_rate": 8.857935934957769e-05, "loss": 0.0383, "step": 25850 }, { "grad_norm": 0.3315674662590027, "learning_rate": 8.856883912883071e-05, "loss": 0.032, "step": 25860 }, { "grad_norm": 0.18920357525348663, "learning_rate": 8.855831469025346e-05, "loss": 0.0323, "step": 25870 }, { "grad_norm": 0.3331491947174072, "learning_rate": 8.854778603499689e-05, "loss": 0.0316, "step": 25880 }, { "grad_norm": 0.254584938287735, "learning_rate": 8.85372531642124e-05, "loss": 0.0321, "step": 25890 }, { "grad_norm": 0.23399345576763153, "learning_rate": 8.852671607905185e-05, "loss": 0.0317, "step": 25900 }, { "grad_norm": 0.2748660147190094, "learning_rate": 8.851617478066754e-05, "loss": 0.0331, "step": 25910 }, { "grad_norm": 0.23483812808990479, "learning_rate": 8.850562927021227e-05, "loss": 0.0309, "step": 25920 }, { "grad_norm": 0.278764933347702, "learning_rate": 8.849507954883928e-05, "loss": 0.0304, "step": 25930 }, { "grad_norm": 0.27548274397850037, "learning_rate": 8.848452561770226e-05, "loss": 0.0316, "step": 25940 }, { "grad_norm": 0.20949006080627441, "learning_rate": 8.847396747795538e-05, "loss": 0.0278, "step": 25950 }, { "grad_norm": 0.18531914055347443, "learning_rate": 8.846340513075327e-05, "loss": 0.0273, "step": 25960 }, { "grad_norm": 0.2740154266357422, "learning_rate": 8.845283857725099e-05, "loss": 0.0293, "step": 25970 }, { "grad_norm": 0.2638716697692871, "learning_rate": 8.844226781860409e-05, "loss": 0.0289, "step": 25980 }, { "grad_norm": 0.24419544637203217, "learning_rate": 8.84316928559686e-05, "loss": 0.0302, "step": 25990 }, { "grad_norm": 0.23551268875598907, "learning_rate": 8.842111369050094e-05, "loss": 0.03, "step": 26000 }, { "grad_norm": 0.2667246460914612, "learning_rate": 8.841053032335808e-05, "loss": 0.0288, "step": 26010 }, { "grad_norm": 0.20150546729564667, "learning_rate": 8.839994275569735e-05, "loss": 0.0307, "step": 26020 }, { "grad_norm": 0.20502635836601257, "learning_rate": 8.838935098867662e-05, "loss": 0.0284, "step": 26030 }, { "grad_norm": 0.35996538400650024, "learning_rate": 8.837875502345418e-05, "loss": 0.0346, "step": 26040 }, { "grad_norm": 0.3367576003074646, "learning_rate": 8.83681548611888e-05, "loss": 0.0306, "step": 26050 }, { "grad_norm": 0.23595702648162842, "learning_rate": 8.835755050303969e-05, "loss": 0.0342, "step": 26060 }, { "grad_norm": 0.34052085876464844, "learning_rate": 8.834694195016653e-05, "loss": 0.0355, "step": 26070 }, { "grad_norm": 0.2961571216583252, "learning_rate": 8.833632920372942e-05, "loss": 0.0312, "step": 26080 }, { "grad_norm": 0.27105623483657837, "learning_rate": 8.832571226488903e-05, "loss": 0.029, "step": 26090 }, { "grad_norm": 0.26992908120155334, "learning_rate": 8.831509113480634e-05, "loss": 0.0293, "step": 26100 }, { "grad_norm": 0.24000683426856995, "learning_rate": 8.83044658146429e-05, "loss": 0.0253, "step": 26110 }, { "grad_norm": 0.24850694835186005, "learning_rate": 8.829383630556067e-05, "loss": 0.0284, "step": 26120 }, { "grad_norm": 0.22960488498210907, "learning_rate": 8.828320260872207e-05, "loss": 0.0284, "step": 26130 }, { "grad_norm": 0.2603646218776703, "learning_rate": 8.827256472529e-05, "loss": 0.0306, "step": 26140 }, { "grad_norm": 0.22416941821575165, "learning_rate": 8.826192265642778e-05, "loss": 0.0276, "step": 26150 }, { "grad_norm": 0.2607567310333252, "learning_rate": 8.825127640329923e-05, "loss": 0.0333, "step": 26160 }, { "grad_norm": 0.26053130626678467, "learning_rate": 8.824062596706861e-05, "loss": 0.0319, "step": 26170 }, { "grad_norm": 0.21859070658683777, "learning_rate": 8.822997134890062e-05, "loss": 0.0269, "step": 26180 }, { "grad_norm": 0.2671319246292114, "learning_rate": 8.821931254996044e-05, "loss": 0.0346, "step": 26190 }, { "grad_norm": 0.28247395157814026, "learning_rate": 8.82086495714137e-05, "loss": 0.0323, "step": 26200 }, { "grad_norm": 0.2105749398469925, "learning_rate": 8.81979824144265e-05, "loss": 0.0273, "step": 26210 }, { "grad_norm": 0.30643150210380554, "learning_rate": 8.818731108016536e-05, "loss": 0.0281, "step": 26220 }, { "grad_norm": 0.2491588145494461, "learning_rate": 8.81766355697973e-05, "loss": 0.0281, "step": 26230 }, { "grad_norm": 0.27688995003700256, "learning_rate": 8.816595588448977e-05, "loss": 0.0281, "step": 26240 }, { "grad_norm": 0.21278508007526398, "learning_rate": 8.81552720254107e-05, "loss": 0.027, "step": 26250 }, { "grad_norm": 0.2716599106788635, "learning_rate": 8.814458399372842e-05, "loss": 0.0332, "step": 26260 }, { "grad_norm": 0.20669744908809662, "learning_rate": 8.813389179061181e-05, "loss": 0.0309, "step": 26270 }, { "grad_norm": 0.20569555461406708, "learning_rate": 8.812319541723012e-05, "loss": 0.0264, "step": 26280 }, { "grad_norm": 0.24050554633140564, "learning_rate": 8.811249487475309e-05, "loss": 0.0342, "step": 26290 }, { "grad_norm": 0.23549260199069977, "learning_rate": 8.810179016435092e-05, "loss": 0.0304, "step": 26300 }, { "grad_norm": 0.22124072909355164, "learning_rate": 8.809108128719428e-05, "loss": 0.0301, "step": 26310 }, { "grad_norm": 0.23127077519893646, "learning_rate": 8.808036824445424e-05, "loss": 0.0318, "step": 26320 }, { "grad_norm": 0.2362799197435379, "learning_rate": 8.806965103730238e-05, "loss": 0.0327, "step": 26330 }, { "grad_norm": 0.33387789130210876, "learning_rate": 8.805892966691074e-05, "loss": 0.0344, "step": 26340 }, { "grad_norm": 0.30935409665107727, "learning_rate": 8.804820413445175e-05, "loss": 0.0297, "step": 26350 }, { "grad_norm": 0.2797938585281372, "learning_rate": 8.803747444109837e-05, "loss": 0.0315, "step": 26360 }, { "grad_norm": 0.25785019993782043, "learning_rate": 8.802674058802399e-05, "loss": 0.0334, "step": 26370 }, { "grad_norm": 0.33879512548446655, "learning_rate": 8.801600257640241e-05, "loss": 0.0344, "step": 26380 }, { "grad_norm": 0.2409130334854126, "learning_rate": 8.800526040740795e-05, "loss": 0.0297, "step": 26390 }, { "grad_norm": 0.2574874460697174, "learning_rate": 8.799451408221535e-05, "loss": 0.0309, "step": 26400 }, { "grad_norm": 0.42753180861473083, "learning_rate": 8.798376360199982e-05, "loss": 0.0349, "step": 26410 }, { "grad_norm": 0.2024017721414566, "learning_rate": 8.797300896793701e-05, "loss": 0.0339, "step": 26420 }, { "grad_norm": 0.3466433584690094, "learning_rate": 8.796225018120302e-05, "loss": 0.0307, "step": 26430 }, { "grad_norm": 0.19871780276298523, "learning_rate": 8.795148724297444e-05, "loss": 0.0305, "step": 26440 }, { "grad_norm": 0.21659894287586212, "learning_rate": 8.794072015442825e-05, "loss": 0.0385, "step": 26450 }, { "grad_norm": 0.37491703033447266, "learning_rate": 8.792994891674198e-05, "loss": 0.0337, "step": 26460 }, { "grad_norm": 0.2702803313732147, "learning_rate": 8.79191735310935e-05, "loss": 0.0355, "step": 26470 }, { "grad_norm": 0.29009079933166504, "learning_rate": 8.790839399866122e-05, "loss": 0.03, "step": 26480 }, { "grad_norm": 0.3062555491924286, "learning_rate": 8.789761032062397e-05, "loss": 0.0356, "step": 26490 }, { "grad_norm": 0.22886864840984344, "learning_rate": 8.788682249816103e-05, "loss": 0.0306, "step": 26500 }, { "grad_norm": 0.34130874276161194, "learning_rate": 8.787603053245215e-05, "loss": 0.0318, "step": 26510 }, { "grad_norm": 0.22116880118846893, "learning_rate": 8.78652344246775e-05, "loss": 0.0317, "step": 26520 }, { "grad_norm": 0.2833399772644043, "learning_rate": 8.785443417601776e-05, "loss": 0.0371, "step": 26530 }, { "grad_norm": 0.24860455095767975, "learning_rate": 8.784362978765401e-05, "loss": 0.0359, "step": 26540 }, { "grad_norm": 0.305483877658844, "learning_rate": 8.783282126076779e-05, "loss": 0.0357, "step": 26550 }, { "grad_norm": 0.3041270077228546, "learning_rate": 8.782200859654112e-05, "loss": 0.0301, "step": 26560 }, { "grad_norm": 0.3107556998729706, "learning_rate": 8.781119179615646e-05, "loss": 0.0357, "step": 26570 }, { "grad_norm": 0.25071531534194946, "learning_rate": 8.780037086079674e-05, "loss": 0.0293, "step": 26580 }, { "grad_norm": 0.26949501037597656, "learning_rate": 8.778954579164527e-05, "loss": 0.0292, "step": 26590 }, { "grad_norm": 0.2791302800178528, "learning_rate": 8.777871658988588e-05, "loss": 0.0301, "step": 26600 }, { "grad_norm": 0.19938655197620392, "learning_rate": 8.776788325670285e-05, "loss": 0.029, "step": 26610 }, { "grad_norm": 0.2673748731613159, "learning_rate": 8.775704579328089e-05, "loss": 0.028, "step": 26620 }, { "grad_norm": 0.3098308742046356, "learning_rate": 8.774620420080517e-05, "loss": 0.0282, "step": 26630 }, { "grad_norm": 0.2564246654510498, "learning_rate": 8.773535848046131e-05, "loss": 0.0281, "step": 26640 }, { "grad_norm": 0.24016891419887543, "learning_rate": 8.772450863343538e-05, "loss": 0.0308, "step": 26650 }, { "grad_norm": 0.22851011157035828, "learning_rate": 8.77136546609139e-05, "loss": 0.0302, "step": 26660 }, { "grad_norm": 0.2967146933078766, "learning_rate": 8.770279656408385e-05, "loss": 0.0292, "step": 26670 }, { "grad_norm": 0.243900865316391, "learning_rate": 8.769193434413265e-05, "loss": 0.0272, "step": 26680 }, { "grad_norm": 0.25408387184143066, "learning_rate": 8.76810680022482e-05, "loss": 0.0333, "step": 26690 }, { "grad_norm": 0.2813531458377838, "learning_rate": 8.767019753961878e-05, "loss": 0.0277, "step": 26700 }, { "grad_norm": 0.25429585576057434, "learning_rate": 8.765932295743321e-05, "loss": 0.0292, "step": 26710 }, { "grad_norm": 0.2538623809814453, "learning_rate": 8.764844425688068e-05, "loss": 0.0323, "step": 26720 }, { "grad_norm": 0.25609782338142395, "learning_rate": 8.763756143915092e-05, "loss": 0.0346, "step": 26730 }, { "grad_norm": 0.3628929853439331, "learning_rate": 8.7626674505434e-05, "loss": 0.0327, "step": 26740 }, { "grad_norm": 0.29585719108581543, "learning_rate": 8.761578345692053e-05, "loss": 0.0318, "step": 26750 }, { "grad_norm": 0.2708943784236908, "learning_rate": 8.760488829480156e-05, "loss": 0.0276, "step": 26760 }, { "grad_norm": 0.24330374598503113, "learning_rate": 8.759398902026854e-05, "loss": 0.0306, "step": 26770 }, { "grad_norm": 0.21205097436904907, "learning_rate": 8.758308563451339e-05, "loss": 0.0233, "step": 26780 }, { "grad_norm": 0.3111143112182617, "learning_rate": 8.75721781387285e-05, "loss": 0.0317, "step": 26790 }, { "grad_norm": 0.23092196881771088, "learning_rate": 8.75612665341067e-05, "loss": 0.027, "step": 26800 }, { "grad_norm": 0.25821536779403687, "learning_rate": 8.755035082184126e-05, "loss": 0.0269, "step": 26810 }, { "grad_norm": 0.28790122270584106, "learning_rate": 8.753943100312592e-05, "loss": 0.0284, "step": 26820 }, { "grad_norm": 0.2867448627948761, "learning_rate": 8.752850707915484e-05, "loss": 0.0288, "step": 26830 }, { "grad_norm": 0.2326114922761917, "learning_rate": 8.751757905112264e-05, "loss": 0.0296, "step": 26840 }, { "grad_norm": 0.2147192806005478, "learning_rate": 8.75066469202244e-05, "loss": 0.0301, "step": 26850 }, { "grad_norm": 0.2816747725009918, "learning_rate": 8.749571068765567e-05, "loss": 0.0274, "step": 26860 }, { "grad_norm": 0.20712816715240479, "learning_rate": 8.748477035461238e-05, "loss": 0.0287, "step": 26870 }, { "grad_norm": 0.2099238485097885, "learning_rate": 8.747382592229095e-05, "loss": 0.0303, "step": 26880 }, { "grad_norm": 0.31689417362213135, "learning_rate": 8.746287739188828e-05, "loss": 0.0375, "step": 26890 }, { "grad_norm": 0.25637686252593994, "learning_rate": 8.745192476460165e-05, "loss": 0.0342, "step": 26900 }, { "grad_norm": 0.29668426513671875, "learning_rate": 8.744096804162882e-05, "loss": 0.0302, "step": 26910 }, { "grad_norm": 0.25478875637054443, "learning_rate": 8.743000722416804e-05, "loss": 0.0327, "step": 26920 }, { "grad_norm": 0.3503742218017578, "learning_rate": 8.741904231341793e-05, "loss": 0.0348, "step": 26930 }, { "grad_norm": 0.2698379158973694, "learning_rate": 8.740807331057762e-05, "loss": 0.0345, "step": 26940 }, { "grad_norm": 0.34319034218788147, "learning_rate": 8.739710021684667e-05, "loss": 0.0298, "step": 26950 }, { "grad_norm": 0.31818488240242004, "learning_rate": 8.738612303342503e-05, "loss": 0.0296, "step": 26960 }, { "grad_norm": 0.27539074420928955, "learning_rate": 8.73751417615132e-05, "loss": 0.0325, "step": 26970 }, { "grad_norm": 0.2958248257637024, "learning_rate": 8.736415640231208e-05, "loss": 0.0288, "step": 26980 }, { "grad_norm": 0.4703603982925415, "learning_rate": 8.735316695702297e-05, "loss": 0.0312, "step": 26990 }, { "grad_norm": 0.4183773398399353, "learning_rate": 8.734217342684769e-05, "loss": 0.0334, "step": 27000 }, { "grad_norm": 0.3468129336833954, "learning_rate": 8.733117581298847e-05, "loss": 0.0277, "step": 27010 }, { "grad_norm": 0.28600555658340454, "learning_rate": 8.732017411664796e-05, "loss": 0.0387, "step": 27020 }, { "grad_norm": 0.2238410860300064, "learning_rate": 8.730916833902936e-05, "loss": 0.0292, "step": 27030 }, { "grad_norm": 0.22472108900547028, "learning_rate": 8.729815848133618e-05, "loss": 0.0328, "step": 27040 }, { "grad_norm": 0.2432767152786255, "learning_rate": 8.728714454477247e-05, "loss": 0.0345, "step": 27050 }, { "grad_norm": 0.2664840519428253, "learning_rate": 8.727612653054269e-05, "loss": 0.0284, "step": 27060 }, { "grad_norm": 0.2550053298473358, "learning_rate": 8.726510443985176e-05, "loss": 0.0306, "step": 27070 }, { "grad_norm": 0.28644615411758423, "learning_rate": 8.725407827390503e-05, "loss": 0.0285, "step": 27080 }, { "grad_norm": 0.25557610392570496, "learning_rate": 8.724304803390833e-05, "loss": 0.0269, "step": 27090 }, { "grad_norm": 0.3097231984138489, "learning_rate": 8.723201372106788e-05, "loss": 0.0279, "step": 27100 }, { "grad_norm": 0.28294914960861206, "learning_rate": 8.722097533659038e-05, "loss": 0.0362, "step": 27110 }, { "grad_norm": 0.23992720246315002, "learning_rate": 8.720993288168299e-05, "loss": 0.0322, "step": 27120 }, { "grad_norm": 0.2368941456079483, "learning_rate": 8.719888635755327e-05, "loss": 0.0276, "step": 27130 }, { "grad_norm": 0.3248007297515869, "learning_rate": 8.718783576540928e-05, "loss": 0.0268, "step": 27140 }, { "grad_norm": 0.21494020521640778, "learning_rate": 8.717678110645948e-05, "loss": 0.0302, "step": 27150 }, { "grad_norm": 0.29250186681747437, "learning_rate": 8.716572238191279e-05, "loss": 0.0355, "step": 27160 }, { "grad_norm": 0.2582803964614868, "learning_rate": 8.715465959297857e-05, "loss": 0.0301, "step": 27170 }, { "grad_norm": 0.3175666630268097, "learning_rate": 8.714359274086665e-05, "loss": 0.0302, "step": 27180 }, { "grad_norm": 0.2161819189786911, "learning_rate": 8.713252182678726e-05, "loss": 0.0318, "step": 27190 }, { "grad_norm": 0.17938563227653503, "learning_rate": 8.712144685195112e-05, "loss": 0.0291, "step": 27200 }, { "grad_norm": 0.34400492906570435, "learning_rate": 8.711036781756936e-05, "loss": 0.0318, "step": 27210 }, { "grad_norm": 0.20862418413162231, "learning_rate": 8.709928472485357e-05, "loss": 0.0321, "step": 27220 }, { "grad_norm": 0.2565416991710663, "learning_rate": 8.708819757501579e-05, "loss": 0.0321, "step": 27230 }, { "grad_norm": 0.2799316346645355, "learning_rate": 8.707710636926846e-05, "loss": 0.0298, "step": 27240 }, { "grad_norm": 0.18682517111301422, "learning_rate": 8.706601110882455e-05, "loss": 0.0323, "step": 27250 }, { "grad_norm": 0.2627207040786743, "learning_rate": 8.705491179489738e-05, "loss": 0.0336, "step": 27260 }, { "grad_norm": 0.22035278379917145, "learning_rate": 8.704380842870077e-05, "loss": 0.0319, "step": 27270 }, { "grad_norm": 0.24014544486999512, "learning_rate": 8.703270101144895e-05, "loss": 0.0298, "step": 27280 }, { "grad_norm": 0.3181326985359192, "learning_rate": 8.702158954435664e-05, "loss": 0.0367, "step": 27290 }, { "grad_norm": 0.22887775301933289, "learning_rate": 8.701047402863896e-05, "loss": 0.0336, "step": 27300 }, { "grad_norm": 0.31146013736724854, "learning_rate": 8.699935446551148e-05, "loss": 0.0315, "step": 27310 }, { "grad_norm": 0.2546696662902832, "learning_rate": 8.698823085619022e-05, "loss": 0.0282, "step": 27320 }, { "grad_norm": 0.7105696201324463, "learning_rate": 8.697710320189166e-05, "loss": 0.0345, "step": 27330 }, { "grad_norm": 0.32276609539985657, "learning_rate": 8.696597150383268e-05, "loss": 0.0371, "step": 27340 }, { "grad_norm": 0.2833329737186432, "learning_rate": 8.695483576323063e-05, "loss": 0.0346, "step": 27350 }, { "grad_norm": 0.2874600291252136, "learning_rate": 8.69436959813033e-05, "loss": 0.0299, "step": 27360 }, { "grad_norm": 0.21380867063999176, "learning_rate": 8.693255215926892e-05, "loss": 0.0297, "step": 27370 }, { "grad_norm": 0.23628173768520355, "learning_rate": 8.692140429834617e-05, "loss": 0.0291, "step": 27380 }, { "grad_norm": 0.22722278535366058, "learning_rate": 8.691025239975415e-05, "loss": 0.0305, "step": 27390 }, { "grad_norm": 0.20183134078979492, "learning_rate": 8.689909646471243e-05, "loss": 0.0262, "step": 27400 }, { "grad_norm": 0.2568448781967163, "learning_rate": 8.688793649444099e-05, "loss": 0.0298, "step": 27410 }, { "grad_norm": 0.25825434923171997, "learning_rate": 8.687677249016029e-05, "loss": 0.033, "step": 27420 }, { "grad_norm": 0.2719058692455292, "learning_rate": 8.686560445309118e-05, "loss": 0.0298, "step": 27430 }, { "grad_norm": 0.2599443793296814, "learning_rate": 8.685443238445499e-05, "loss": 0.0301, "step": 27440 }, { "grad_norm": 0.2768963575363159, "learning_rate": 8.68432562854735e-05, "loss": 0.0293, "step": 27450 }, { "grad_norm": 0.23048686981201172, "learning_rate": 8.683207615736887e-05, "loss": 0.0324, "step": 27460 }, { "grad_norm": 0.2345290333032608, "learning_rate": 8.682089200136379e-05, "loss": 0.0262, "step": 27470 }, { "grad_norm": 0.2123831957578659, "learning_rate": 8.680970381868132e-05, "loss": 0.0293, "step": 27480 }, { "grad_norm": 0.20408624410629272, "learning_rate": 8.679851161054498e-05, "loss": 0.026, "step": 27490 }, { "grad_norm": 0.294243723154068, "learning_rate": 8.678731537817873e-05, "loss": 0.0275, "step": 27500 }, { "grad_norm": 0.204739511013031, "learning_rate": 8.677611512280697e-05, "loss": 0.0323, "step": 27510 }, { "grad_norm": 0.22028344869613647, "learning_rate": 8.676491084565457e-05, "loss": 0.0215, "step": 27520 }, { "grad_norm": 0.2576832175254822, "learning_rate": 8.675370254794678e-05, "loss": 0.0294, "step": 27530 }, { "grad_norm": 0.23103253543376923, "learning_rate": 8.674249023090935e-05, "loss": 0.0247, "step": 27540 }, { "grad_norm": 0.19690339267253876, "learning_rate": 8.673127389576843e-05, "loss": 0.026, "step": 27550 }, { "grad_norm": 0.3087855577468872, "learning_rate": 8.67200535437506e-05, "loss": 0.0307, "step": 27560 }, { "grad_norm": 0.2155025154352188, "learning_rate": 8.670882917608296e-05, "loss": 0.0267, "step": 27570 }, { "grad_norm": 0.2844778001308441, "learning_rate": 8.669760079399292e-05, "loss": 0.0279, "step": 27580 }, { "grad_norm": 0.2708774209022522, "learning_rate": 8.668636839870845e-05, "loss": 0.0308, "step": 27590 }, { "grad_norm": 0.2686190605163574, "learning_rate": 8.667513199145789e-05, "loss": 0.0271, "step": 27600 }, { "grad_norm": 0.27506378293037415, "learning_rate": 8.666389157347002e-05, "loss": 0.0308, "step": 27610 }, { "grad_norm": 0.24749991297721863, "learning_rate": 8.66526471459741e-05, "loss": 0.0248, "step": 27620 }, { "grad_norm": 0.26372408866882324, "learning_rate": 8.66413987101998e-05, "loss": 0.0282, "step": 27630 }, { "grad_norm": 0.3005412518978119, "learning_rate": 8.663014626737723e-05, "loss": 0.0289, "step": 27640 }, { "grad_norm": 0.32063621282577515, "learning_rate": 8.661888981873691e-05, "loss": 0.0301, "step": 27650 }, { "grad_norm": 0.23865798115730286, "learning_rate": 8.660762936550988e-05, "loss": 0.027, "step": 27660 }, { "grad_norm": 0.3090013861656189, "learning_rate": 8.659636490892753e-05, "loss": 0.0275, "step": 27670 }, { "grad_norm": 0.27160319685935974, "learning_rate": 8.658509645022174e-05, "loss": 0.0249, "step": 27680 }, { "grad_norm": 0.24477560818195343, "learning_rate": 8.657382399062481e-05, "loss": 0.0287, "step": 27690 }, { "grad_norm": 0.3547475337982178, "learning_rate": 8.656254753136946e-05, "loss": 0.0273, "step": 27700 }, { "grad_norm": 0.298736572265625, "learning_rate": 8.655126707368891e-05, "loss": 0.0273, "step": 27710 }, { "grad_norm": 0.3223673701286316, "learning_rate": 8.653998261881672e-05, "loss": 0.0294, "step": 27720 }, { "grad_norm": 0.18249602615833282, "learning_rate": 8.652869416798699e-05, "loss": 0.0285, "step": 27730 }, { "grad_norm": 0.24309127032756805, "learning_rate": 8.651740172243417e-05, "loss": 0.0266, "step": 27740 }, { "grad_norm": 0.21791523694992065, "learning_rate": 8.65061052833932e-05, "loss": 0.0281, "step": 27750 }, { "grad_norm": 0.28080257773399353, "learning_rate": 8.649480485209945e-05, "loss": 0.0262, "step": 27760 }, { "grad_norm": 1.1484829187393188, "learning_rate": 8.64835004297887e-05, "loss": 0.0287, "step": 27770 }, { "grad_norm": 0.2899460792541504, "learning_rate": 8.64721920176972e-05, "loss": 0.0263, "step": 27780 }, { "grad_norm": 0.3391817808151245, "learning_rate": 8.646087961706164e-05, "loss": 0.0355, "step": 27790 }, { "grad_norm": 0.2471434772014618, "learning_rate": 8.644956322911908e-05, "loss": 0.031, "step": 27800 }, { "grad_norm": 0.25490936636924744, "learning_rate": 8.643824285510709e-05, "loss": 0.0288, "step": 27810 }, { "grad_norm": 0.2406267374753952, "learning_rate": 8.642691849626364e-05, "loss": 0.0276, "step": 27820 }, { "grad_norm": 0.2986033856868744, "learning_rate": 8.641559015382717e-05, "loss": 0.0291, "step": 27830 }, { "grad_norm": 0.212666317820549, "learning_rate": 8.640425782903649e-05, "loss": 0.0331, "step": 27840 }, { "grad_norm": 0.22746026515960693, "learning_rate": 8.639292152313091e-05, "loss": 0.0305, "step": 27850 }, { "grad_norm": 0.24438850581645966, "learning_rate": 8.638158123735015e-05, "loss": 0.0274, "step": 27860 }, { "grad_norm": 0.32578688859939575, "learning_rate": 8.637023697293436e-05, "loss": 0.0335, "step": 27870 }, { "grad_norm": 0.29744401574134827, "learning_rate": 8.635888873112414e-05, "loss": 0.0331, "step": 27880 }, { "grad_norm": 0.24325628578662872, "learning_rate": 8.634753651316052e-05, "loss": 0.0276, "step": 27890 }, { "grad_norm": 0.2706238925457001, "learning_rate": 8.633618032028496e-05, "loss": 0.0276, "step": 27900 }, { "grad_norm": 0.2123597264289856, "learning_rate": 8.632482015373934e-05, "loss": 0.0296, "step": 27910 }, { "grad_norm": 0.5857713222503662, "learning_rate": 8.6313456014766e-05, "loss": 0.0245, "step": 27920 }, { "grad_norm": 0.42630183696746826, "learning_rate": 8.630208790460771e-05, "loss": 0.0277, "step": 27930 }, { "grad_norm": 0.19464628398418427, "learning_rate": 8.629071582450768e-05, "loss": 0.0301, "step": 27940 }, { "grad_norm": 0.28361082077026367, "learning_rate": 8.62793397757095e-05, "loss": 0.0376, "step": 27950 }, { "grad_norm": 0.26176565885543823, "learning_rate": 8.626795975945729e-05, "loss": 0.0345, "step": 27960 }, { "grad_norm": 0.2847420275211334, "learning_rate": 8.625657577699551e-05, "loss": 0.0326, "step": 27970 }, { "grad_norm": 0.3042140305042267, "learning_rate": 8.624518782956914e-05, "loss": 0.0321, "step": 27980 }, { "grad_norm": 0.26893535256385803, "learning_rate": 8.62337959184235e-05, "loss": 0.0304, "step": 27990 }, { "grad_norm": 0.20253147184848785, "learning_rate": 8.622240004480441e-05, "loss": 0.0267, "step": 28000 }, { "grad_norm": 0.234086811542511, "learning_rate": 8.621100020995814e-05, "loss": 0.0258, "step": 28010 }, { "grad_norm": 0.26519402861595154, "learning_rate": 8.619959641513132e-05, "loss": 0.0263, "step": 28020 }, { "grad_norm": 0.33314749598503113, "learning_rate": 8.618818866157105e-05, "loss": 0.0297, "step": 28030 }, { "grad_norm": 0.28256747126579285, "learning_rate": 8.617677695052487e-05, "loss": 0.0313, "step": 28040 }, { "grad_norm": 0.281789094209671, "learning_rate": 8.616536128324078e-05, "loss": 0.0312, "step": 28050 }, { "grad_norm": 0.1916702687740326, "learning_rate": 8.615394166096712e-05, "loss": 0.0285, "step": 28060 }, { "grad_norm": 0.32113975286483765, "learning_rate": 8.614251808495279e-05, "loss": 0.0264, "step": 28070 }, { "grad_norm": 0.27132293581962585, "learning_rate": 8.6131090556447e-05, "loss": 0.0288, "step": 28080 }, { "grad_norm": 0.28284546732902527, "learning_rate": 8.611965907669947e-05, "loss": 0.031, "step": 28090 }, { "grad_norm": 0.30463653802871704, "learning_rate": 8.610822364696034e-05, "loss": 0.0324, "step": 28100 }, { "grad_norm": 0.5686002373695374, "learning_rate": 8.609678426848015e-05, "loss": 0.035, "step": 28110 }, { "grad_norm": 0.24210283160209656, "learning_rate": 8.60853409425099e-05, "loss": 0.0305, "step": 28120 }, { "grad_norm": 0.302970826625824, "learning_rate": 8.607389367030104e-05, "loss": 0.0311, "step": 28130 }, { "grad_norm": 0.2880082130432129, "learning_rate": 8.606244245310538e-05, "loss": 0.0302, "step": 28140 }, { "grad_norm": 0.2339506298303604, "learning_rate": 8.605098729217525e-05, "loss": 0.0271, "step": 28150 }, { "grad_norm": 0.2328348159790039, "learning_rate": 8.603952818876335e-05, "loss": 0.0269, "step": 28160 }, { "grad_norm": 0.26604902744293213, "learning_rate": 8.602806514412281e-05, "loss": 0.0277, "step": 28170 }, { "grad_norm": 0.33240142464637756, "learning_rate": 8.601659815950726e-05, "loss": 0.0282, "step": 28180 }, { "grad_norm": 0.2576879858970642, "learning_rate": 8.600512723617067e-05, "loss": 0.0277, "step": 28190 }, { "grad_norm": 0.3199092745780945, "learning_rate": 8.59936523753675e-05, "loss": 0.0279, "step": 28200 }, { "grad_norm": 0.24199914932250977, "learning_rate": 8.598217357835264e-05, "loss": 0.0264, "step": 28210 }, { "grad_norm": 0.3359038531780243, "learning_rate": 8.597069084638135e-05, "loss": 0.0338, "step": 28220 }, { "grad_norm": 0.26221713423728943, "learning_rate": 8.595920418070939e-05, "loss": 0.0242, "step": 28230 }, { "grad_norm": 0.24137923121452332, "learning_rate": 8.594771358259295e-05, "loss": 0.0227, "step": 28240 }, { "grad_norm": 0.1937115341424942, "learning_rate": 8.593621905328858e-05, "loss": 0.0282, "step": 28250 }, { "grad_norm": 0.2318277657032013, "learning_rate": 8.592472059405333e-05, "loss": 0.0257, "step": 28260 }, { "grad_norm": 0.2814060151576996, "learning_rate": 8.591321820614464e-05, "loss": 0.025, "step": 28270 }, { "grad_norm": 0.3032830059528351, "learning_rate": 8.590171189082041e-05, "loss": 0.0294, "step": 28280 }, { "grad_norm": 0.24969597160816193, "learning_rate": 8.589020164933894e-05, "loss": 0.0249, "step": 28290 }, { "grad_norm": 0.21740034222602844, "learning_rate": 8.587868748295898e-05, "loss": 0.0221, "step": 28300 }, { "grad_norm": 0.275697261095047, "learning_rate": 8.586716939293971e-05, "loss": 0.0254, "step": 28310 }, { "grad_norm": 0.27083516120910645, "learning_rate": 8.58556473805407e-05, "loss": 0.0272, "step": 28320 }, { "grad_norm": 0.2699764668941498, "learning_rate": 8.584412144702202e-05, "loss": 0.0274, "step": 28330 }, { "grad_norm": 0.249410480260849, "learning_rate": 8.58325915936441e-05, "loss": 0.0274, "step": 28340 }, { "grad_norm": 0.28666630387306213, "learning_rate": 8.582105782166783e-05, "loss": 0.0264, "step": 28350 }, { "grad_norm": 0.18321166932582855, "learning_rate": 8.580952013235455e-05, "loss": 0.0259, "step": 28360 }, { "grad_norm": 0.25856658816337585, "learning_rate": 8.579797852696596e-05, "loss": 0.026, "step": 28370 }, { "grad_norm": 0.2660110890865326, "learning_rate": 8.578643300676428e-05, "loss": 0.0288, "step": 28380 }, { "grad_norm": 0.3246751129627228, "learning_rate": 8.577488357301209e-05, "loss": 0.0268, "step": 28390 }, { "grad_norm": 0.2387290596961975, "learning_rate": 8.576333022697242e-05, "loss": 0.0336, "step": 28400 }, { "grad_norm": 0.2687186300754547, "learning_rate": 8.575177296990873e-05, "loss": 0.0288, "step": 28410 }, { "grad_norm": 0.25009819865226746, "learning_rate": 8.574021180308489e-05, "loss": 0.0258, "step": 28420 }, { "grad_norm": 0.23284941911697388, "learning_rate": 8.572864672776523e-05, "loss": 0.0306, "step": 28430 }, { "grad_norm": 0.29756203293800354, "learning_rate": 8.571707774521447e-05, "loss": 0.0293, "step": 28440 }, { "grad_norm": 0.31797850131988525, "learning_rate": 8.57055048566978e-05, "loss": 0.0344, "step": 28450 }, { "grad_norm": 0.2155226618051529, "learning_rate": 8.569392806348078e-05, "loss": 0.0286, "step": 28460 }, { "grad_norm": 0.29076728224754333, "learning_rate": 8.568234736682947e-05, "loss": 0.0287, "step": 28470 }, { "grad_norm": 0.31017014384269714, "learning_rate": 8.567076276801029e-05, "loss": 0.031, "step": 28480 }, { "grad_norm": 0.23827335238456726, "learning_rate": 8.565917426829013e-05, "loss": 0.0304, "step": 28490 }, { "grad_norm": 0.26678934693336487, "learning_rate": 8.564758186893628e-05, "loss": 0.0351, "step": 28500 }, { "grad_norm": 0.3191007375717163, "learning_rate": 8.563598557121649e-05, "loss": 0.0291, "step": 28510 }, { "grad_norm": 0.25499263405799866, "learning_rate": 8.562438537639888e-05, "loss": 0.0299, "step": 28520 }, { "grad_norm": 0.2746596038341522, "learning_rate": 8.561278128575206e-05, "loss": 0.0287, "step": 28530 }, { "grad_norm": 0.26485303044319153, "learning_rate": 8.5601173300545e-05, "loss": 0.0352, "step": 28540 }, { "grad_norm": 0.24239325523376465, "learning_rate": 8.558956142204717e-05, "loss": 0.0285, "step": 28550 }, { "grad_norm": 0.2534829080104828, "learning_rate": 8.55779456515284e-05, "loss": 0.0264, "step": 28560 }, { "grad_norm": 0.2705429494380951, "learning_rate": 8.556632599025898e-05, "loss": 0.0272, "step": 28570 }, { "grad_norm": 0.43279269337654114, "learning_rate": 8.555470243950964e-05, "loss": 0.0314, "step": 28580 }, { "grad_norm": 0.2913261353969574, "learning_rate": 8.554307500055148e-05, "loss": 0.0263, "step": 28590 }, { "grad_norm": 0.20830374956130981, "learning_rate": 8.553144367465609e-05, "loss": 0.0271, "step": 28600 }, { "grad_norm": 0.3036060333251953, "learning_rate": 8.551980846309544e-05, "loss": 0.0308, "step": 28610 }, { "grad_norm": 0.27052533626556396, "learning_rate": 8.550816936714193e-05, "loss": 0.0266, "step": 28620 }, { "grad_norm": 0.20204707980155945, "learning_rate": 8.549652638806841e-05, "loss": 0.0268, "step": 28630 }, { "grad_norm": 0.267611563205719, "learning_rate": 8.548487952714812e-05, "loss": 0.0318, "step": 28640 }, { "grad_norm": 0.30950435996055603, "learning_rate": 8.547322878565478e-05, "loss": 0.0278, "step": 28650 }, { "grad_norm": 0.24474826455116272, "learning_rate": 8.546157416486245e-05, "loss": 0.03, "step": 28660 }, { "grad_norm": 0.2764523923397064, "learning_rate": 8.54499156660457e-05, "loss": 0.0264, "step": 28670 }, { "grad_norm": 0.2930888831615448, "learning_rate": 8.543825329047947e-05, "loss": 0.0317, "step": 28680 }, { "grad_norm": 0.19455872476100922, "learning_rate": 8.542658703943913e-05, "loss": 0.0294, "step": 28690 }, { "grad_norm": 0.22505800426006317, "learning_rate": 8.541491691420051e-05, "loss": 0.0365, "step": 28700 }, { "grad_norm": 0.22877749800682068, "learning_rate": 8.54032429160398e-05, "loss": 0.0248, "step": 28710 }, { "grad_norm": 0.2468203604221344, "learning_rate": 8.539156504623369e-05, "loss": 0.0292, "step": 28720 }, { "grad_norm": 0.28499749302864075, "learning_rate": 8.537988330605923e-05, "loss": 0.0333, "step": 28730 }, { "grad_norm": 0.25020870566368103, "learning_rate": 8.536819769679393e-05, "loss": 0.035, "step": 28740 }, { "grad_norm": 0.2655138671398163, "learning_rate": 8.53565082197157e-05, "loss": 0.0333, "step": 28750 }, { "grad_norm": 0.25367647409439087, "learning_rate": 8.534481487610289e-05, "loss": 0.0252, "step": 28760 }, { "grad_norm": 0.25046050548553467, "learning_rate": 8.533311766723428e-05, "loss": 0.0252, "step": 28770 }, { "grad_norm": 0.27940842509269714, "learning_rate": 8.532141659438901e-05, "loss": 0.0282, "step": 28780 }, { "grad_norm": 0.22806516289710999, "learning_rate": 8.530971165884675e-05, "loss": 0.0279, "step": 28790 }, { "grad_norm": 0.24766045808792114, "learning_rate": 8.529800286188752e-05, "loss": 0.0266, "step": 28800 }, { "grad_norm": 0.24655045568943024, "learning_rate": 8.528629020479175e-05, "loss": 0.0245, "step": 28810 }, { "grad_norm": 0.20379893481731415, "learning_rate": 8.527457368884033e-05, "loss": 0.0259, "step": 28820 }, { "grad_norm": 0.23875518143177032, "learning_rate": 8.526285331531458e-05, "loss": 0.0265, "step": 28830 }, { "grad_norm": 0.24211980402469635, "learning_rate": 8.525112908549621e-05, "loss": 0.0247, "step": 28840 }, { "grad_norm": 0.2635290026664734, "learning_rate": 8.523940100066735e-05, "loss": 0.0275, "step": 28850 }, { "grad_norm": 0.2362983077764511, "learning_rate": 8.52276690621106e-05, "loss": 0.0277, "step": 28860 }, { "grad_norm": 0.27770164608955383, "learning_rate": 8.521593327110889e-05, "loss": 0.0245, "step": 28870 }, { "grad_norm": 0.2632129192352295, "learning_rate": 8.520419362894569e-05, "loss": 0.0316, "step": 28880 }, { "grad_norm": 0.2919059693813324, "learning_rate": 8.51924501369048e-05, "loss": 0.0272, "step": 28890 }, { "grad_norm": 0.25273966789245605, "learning_rate": 8.518070279627047e-05, "loss": 0.0248, "step": 28900 }, { "grad_norm": 0.22595685720443726, "learning_rate": 8.516895160832737e-05, "loss": 0.0245, "step": 28910 }, { "grad_norm": 0.216860830783844, "learning_rate": 8.515719657436061e-05, "loss": 0.0296, "step": 28920 }, { "grad_norm": 0.21179743111133575, "learning_rate": 8.514543769565568e-05, "loss": 0.0285, "step": 28930 }, { "grad_norm": 0.19401046633720398, "learning_rate": 8.513367497349853e-05, "loss": 0.0262, "step": 28940 }, { "grad_norm": 0.2651415467262268, "learning_rate": 8.51219084091755e-05, "loss": 0.0292, "step": 28950 }, { "grad_norm": 0.24536053836345673, "learning_rate": 8.511013800397338e-05, "loss": 0.0258, "step": 28960 }, { "grad_norm": 0.23702630400657654, "learning_rate": 8.509836375917937e-05, "loss": 0.0314, "step": 28970 }, { "grad_norm": 0.23794086277484894, "learning_rate": 8.508658567608104e-05, "loss": 0.0267, "step": 28980 }, { "grad_norm": 0.33378666639328003, "learning_rate": 8.507480375596647e-05, "loss": 0.0277, "step": 28990 }, { "grad_norm": 0.21657906472682953, "learning_rate": 8.506301800012408e-05, "loss": 0.0289, "step": 29000 }, { "grad_norm": 0.5268332958221436, "learning_rate": 8.505122840984278e-05, "loss": 0.0297, "step": 29010 }, { "grad_norm": 0.2453569769859314, "learning_rate": 8.503943498641182e-05, "loss": 0.0293, "step": 29020 }, { "grad_norm": 0.29734736680984497, "learning_rate": 8.502763773112095e-05, "loss": 0.0283, "step": 29030 }, { "grad_norm": 0.24560412764549255, "learning_rate": 8.501583664526026e-05, "loss": 0.0266, "step": 29040 }, { "grad_norm": 0.23585449159145355, "learning_rate": 8.500403173012032e-05, "loss": 0.0368, "step": 29050 }, { "grad_norm": 0.34691911935806274, "learning_rate": 8.499222298699211e-05, "loss": 0.0347, "step": 29060 }, { "grad_norm": 0.26792603731155396, "learning_rate": 8.498041041716701e-05, "loss": 0.0389, "step": 29070 }, { "grad_norm": 0.21263498067855835, "learning_rate": 8.496859402193681e-05, "loss": 0.0293, "step": 29080 }, { "grad_norm": 0.2885090708732605, "learning_rate": 8.495677380259374e-05, "loss": 0.0316, "step": 29090 }, { "grad_norm": 0.23876352608203888, "learning_rate": 8.494494976043045e-05, "loss": 0.0252, "step": 29100 }, { "grad_norm": 0.22228319942951202, "learning_rate": 8.493312189673998e-05, "loss": 0.026, "step": 29110 }, { "grad_norm": 0.24212655425071716, "learning_rate": 8.492129021281584e-05, "loss": 0.0248, "step": 29120 }, { "grad_norm": 0.28401118516921997, "learning_rate": 8.490945470995188e-05, "loss": 0.025, "step": 29130 }, { "grad_norm": 0.34648433327674866, "learning_rate": 8.489761538944247e-05, "loss": 0.0295, "step": 29140 }, { "grad_norm": 0.270978718996048, "learning_rate": 8.48857722525823e-05, "loss": 0.0315, "step": 29150 }, { "grad_norm": 0.27156350016593933, "learning_rate": 8.487392530066652e-05, "loss": 0.0323, "step": 29160 }, { "grad_norm": 0.32633310556411743, "learning_rate": 8.486207453499069e-05, "loss": 0.0298, "step": 29170 }, { "grad_norm": 0.3073897957801819, "learning_rate": 8.485021995685082e-05, "loss": 0.0296, "step": 29180 }, { "grad_norm": 0.25662556290626526, "learning_rate": 8.483836156754328e-05, "loss": 0.0304, "step": 29190 }, { "grad_norm": 0.1997276395559311, "learning_rate": 8.482649936836491e-05, "loss": 0.0268, "step": 29200 }, { "grad_norm": 0.2986949682235718, "learning_rate": 8.481463336061293e-05, "loss": 0.026, "step": 29210 }, { "grad_norm": 0.26414552330970764, "learning_rate": 8.480276354558496e-05, "loss": 0.032, "step": 29220 }, { "grad_norm": 0.2096778303384781, "learning_rate": 8.479088992457913e-05, "loss": 0.0257, "step": 29230 }, { "grad_norm": 0.19288334250450134, "learning_rate": 8.477901249889387e-05, "loss": 0.0256, "step": 29240 }, { "grad_norm": 0.17360328137874603, "learning_rate": 8.47671312698281e-05, "loss": 0.0294, "step": 29250 }, { "grad_norm": 0.27060240507125854, "learning_rate": 8.475524623868112e-05, "loss": 0.0252, "step": 29260 }, { "grad_norm": 0.21127411723136902, "learning_rate": 8.474335740675266e-05, "loss": 0.0247, "step": 29270 }, { "grad_norm": 0.3074692189693451, "learning_rate": 8.473146477534289e-05, "loss": 0.028, "step": 29280 }, { "grad_norm": 0.2348417192697525, "learning_rate": 8.471956834575232e-05, "loss": 0.0274, "step": 29290 }, { "grad_norm": 0.2558923661708832, "learning_rate": 8.470766811928197e-05, "loss": 0.0285, "step": 29300 }, { "grad_norm": 0.23553308844566345, "learning_rate": 8.469576409723323e-05, "loss": 0.0289, "step": 29310 }, { "grad_norm": 0.22350376844406128, "learning_rate": 8.468385628090788e-05, "loss": 0.0254, "step": 29320 }, { "grad_norm": 0.2832091748714447, "learning_rate": 8.467194467160815e-05, "loss": 0.0279, "step": 29330 }, { "grad_norm": 0.23289190232753754, "learning_rate": 8.466002927063667e-05, "loss": 0.0277, "step": 29340 }, { "grad_norm": 0.21312469244003296, "learning_rate": 8.464811007929651e-05, "loss": 0.0252, "step": 29350 }, { "grad_norm": 0.2535112500190735, "learning_rate": 8.463618709889114e-05, "loss": 0.0272, "step": 29360 }, { "grad_norm": 0.25301000475883484, "learning_rate": 8.462426033072442e-05, "loss": 0.0314, "step": 29370 }, { "grad_norm": 0.2855204641819, "learning_rate": 8.461232977610061e-05, "loss": 0.0332, "step": 29380 }, { "grad_norm": 0.3202858567237854, "learning_rate": 8.46003954363245e-05, "loss": 0.0362, "step": 29390 }, { "grad_norm": 0.3324684500694275, "learning_rate": 8.458845731270115e-05, "loss": 0.0301, "step": 29400 }, { "grad_norm": 0.23788870871067047, "learning_rate": 8.45765154065361e-05, "loss": 0.0306, "step": 29410 }, { "grad_norm": 0.22722628712654114, "learning_rate": 8.456456971913532e-05, "loss": 0.028, "step": 29420 }, { "grad_norm": 0.27513790130615234, "learning_rate": 8.455262025180517e-05, "loss": 0.0277, "step": 29430 }, { "grad_norm": 0.27278652787208557, "learning_rate": 8.454066700585242e-05, "loss": 0.0274, "step": 29440 }, { "grad_norm": 0.23645946383476257, "learning_rate": 8.452870998258423e-05, "loss": 0.0272, "step": 29450 }, { "grad_norm": 0.24881958961486816, "learning_rate": 8.451674918330825e-05, "loss": 0.0313, "step": 29460 }, { "grad_norm": 0.42532435059547424, "learning_rate": 8.450478460933246e-05, "loss": 0.0316, "step": 29470 }, { "grad_norm": 0.21859531104564667, "learning_rate": 8.449281626196532e-05, "loss": 0.0244, "step": 29480 }, { "grad_norm": 0.22620010375976562, "learning_rate": 8.448084414251564e-05, "loss": 0.0274, "step": 29490 }, { "grad_norm": 0.367156058549881, "learning_rate": 8.446886825229271e-05, "loss": 0.0294, "step": 29500 }, { "grad_norm": 0.268239289522171, "learning_rate": 8.445688859260615e-05, "loss": 0.0333, "step": 29510 }, { "grad_norm": 0.43973425030708313, "learning_rate": 8.444490516476606e-05, "loss": 0.0332, "step": 29520 }, { "grad_norm": 0.2520362436771393, "learning_rate": 8.443291797008293e-05, "loss": 0.0315, "step": 29530 }, { "grad_norm": 0.3071427345275879, "learning_rate": 8.442092700986765e-05, "loss": 0.0353, "step": 29540 }, { "grad_norm": 0.2211705595254898, "learning_rate": 8.440893228543156e-05, "loss": 0.034, "step": 29550 }, { "grad_norm": 0.28372058272361755, "learning_rate": 8.439693379808638e-05, "loss": 0.0276, "step": 29560 }, { "grad_norm": 0.21592211723327637, "learning_rate": 8.43849315491442e-05, "loss": 0.0288, "step": 29570 }, { "grad_norm": 0.2353755682706833, "learning_rate": 8.437292553991763e-05, "loss": 0.0278, "step": 29580 }, { "grad_norm": 0.22837740182876587, "learning_rate": 8.436091577171959e-05, "loss": 0.031, "step": 29590 }, { "grad_norm": 0.2510949671268463, "learning_rate": 8.434890224586347e-05, "loss": 0.0269, "step": 29600 }, { "grad_norm": 0.23065905272960663, "learning_rate": 8.433688496366303e-05, "loss": 0.0319, "step": 29610 }, { "grad_norm": 0.24298079311847687, "learning_rate": 8.432486392643248e-05, "loss": 0.0233, "step": 29620 }, { "grad_norm": 0.22177676856517792, "learning_rate": 8.431283913548643e-05, "loss": 0.027, "step": 29630 }, { "grad_norm": 0.2065637856721878, "learning_rate": 8.430081059213985e-05, "loss": 0.0269, "step": 29640 }, { "grad_norm": 0.21302691102027893, "learning_rate": 8.428877829770823e-05, "loss": 0.0272, "step": 29650 }, { "grad_norm": 0.2800058424472809, "learning_rate": 8.427674225350735e-05, "loss": 0.033, "step": 29660 }, { "grad_norm": 0.22510434687137604, "learning_rate": 8.426470246085347e-05, "loss": 0.0259, "step": 29670 }, { "grad_norm": 0.25162041187286377, "learning_rate": 8.425265892106324e-05, "loss": 0.029, "step": 29680 }, { "grad_norm": 0.19319167733192444, "learning_rate": 8.424061163545374e-05, "loss": 0.0242, "step": 29690 }, { "grad_norm": 0.3120652139186859, "learning_rate": 8.422856060534243e-05, "loss": 0.0306, "step": 29700 }, { "grad_norm": 0.3756107985973358, "learning_rate": 8.421650583204718e-05, "loss": 0.0299, "step": 29710 }, { "grad_norm": 0.23163758218288422, "learning_rate": 8.420444731688633e-05, "loss": 0.0291, "step": 29720 }, { "grad_norm": 0.33308279514312744, "learning_rate": 8.419238506117852e-05, "loss": 0.0276, "step": 29730 }, { "grad_norm": 0.20760475099086761, "learning_rate": 8.418031906624289e-05, "loss": 0.0253, "step": 29740 }, { "grad_norm": 0.27165016531944275, "learning_rate": 8.416824933339898e-05, "loss": 0.0274, "step": 29750 }, { "grad_norm": 0.2193240374326706, "learning_rate": 8.415617586396667e-05, "loss": 0.0254, "step": 29760 }, { "grad_norm": 0.2117098569869995, "learning_rate": 8.414409865926632e-05, "loss": 0.022, "step": 29770 }, { "grad_norm": 0.2554475963115692, "learning_rate": 8.413201772061867e-05, "loss": 0.0242, "step": 29780 }, { "grad_norm": 0.2420123666524887, "learning_rate": 8.411993304934488e-05, "loss": 0.0298, "step": 29790 }, { "grad_norm": 0.2726929783821106, "learning_rate": 8.410784464676654e-05, "loss": 0.0244, "step": 29800 }, { "grad_norm": 0.25824543833732605, "learning_rate": 8.409575251420556e-05, "loss": 0.0289, "step": 29810 }, { "grad_norm": 0.23323704302310944, "learning_rate": 8.408365665298435e-05, "loss": 0.0265, "step": 29820 }, { "grad_norm": 0.16622582077980042, "learning_rate": 8.40715570644257e-05, "loss": 0.0317, "step": 29830 }, { "grad_norm": 0.2708929479122162, "learning_rate": 8.40594537498528e-05, "loss": 0.0276, "step": 29840 }, { "grad_norm": 0.33572259545326233, "learning_rate": 8.404734671058924e-05, "loss": 0.0291, "step": 29850 }, { "grad_norm": 0.23304909467697144, "learning_rate": 8.403523594795902e-05, "loss": 0.0262, "step": 29860 }, { "grad_norm": 0.20408962666988373, "learning_rate": 8.402312146328659e-05, "loss": 0.0274, "step": 29870 }, { "grad_norm": 0.3316859006881714, "learning_rate": 8.401100325789675e-05, "loss": 0.0276, "step": 29880 }, { "grad_norm": 0.35880687832832336, "learning_rate": 8.399888133311472e-05, "loss": 0.0317, "step": 29890 }, { "grad_norm": 0.2600122094154358, "learning_rate": 8.398675569026613e-05, "loss": 0.0289, "step": 29900 }, { "grad_norm": 0.27163931727409363, "learning_rate": 8.397462633067705e-05, "loss": 0.0298, "step": 29910 }, { "grad_norm": 0.22560034692287445, "learning_rate": 8.396249325567392e-05, "loss": 0.0243, "step": 29920 }, { "grad_norm": 0.2247617393732071, "learning_rate": 8.395035646658357e-05, "loss": 0.0259, "step": 29930 }, { "grad_norm": 0.24794933199882507, "learning_rate": 8.39382159647333e-05, "loss": 0.0253, "step": 29940 }, { "grad_norm": 0.2597416937351227, "learning_rate": 8.392607175145075e-05, "loss": 0.0287, "step": 29950 }, { "grad_norm": 0.23187477886676788, "learning_rate": 8.3913923828064e-05, "loss": 0.0251, "step": 29960 }, { "grad_norm": 0.24364595115184784, "learning_rate": 8.390177219590152e-05, "loss": 0.0283, "step": 29970 }, { "grad_norm": 0.29040390253067017, "learning_rate": 8.388961685629222e-05, "loss": 0.0281, "step": 29980 }, { "grad_norm": 0.23082076013088226, "learning_rate": 8.387745781056536e-05, "loss": 0.0317, "step": 29990 }, { "grad_norm": 0.31489309668540955, "learning_rate": 8.386529506005065e-05, "loss": 0.0299, "step": 30000 }, { "grad_norm": 0.2840903699398041, "learning_rate": 8.38531286060782e-05, "loss": 0.0318, "step": 30010 }, { "grad_norm": 0.3241869807243347, "learning_rate": 8.384095844997849e-05, "loss": 0.0261, "step": 30020 }, { "grad_norm": 0.2371254414319992, "learning_rate": 8.382878459308245e-05, "loss": 0.0269, "step": 30030 }, { "grad_norm": 0.274812251329422, "learning_rate": 8.381660703672138e-05, "loss": 0.0263, "step": 30040 }, { "grad_norm": 0.1874818056821823, "learning_rate": 8.380442578222702e-05, "loss": 0.0243, "step": 30050 }, { "grad_norm": 0.25563883781433105, "learning_rate": 8.379224083093146e-05, "loss": 0.0262, "step": 30060 }, { "grad_norm": 0.2684374749660492, "learning_rate": 8.378005218416727e-05, "loss": 0.0305, "step": 30070 }, { "grad_norm": 0.4609244763851166, "learning_rate": 8.376785984326735e-05, "loss": 0.0242, "step": 30080 }, { "grad_norm": 0.22436459362506866, "learning_rate": 8.375566380956506e-05, "loss": 0.0299, "step": 30090 }, { "grad_norm": 0.20501458644866943, "learning_rate": 8.374346408439411e-05, "loss": 0.026, "step": 30100 }, { "grad_norm": 0.2605191469192505, "learning_rate": 8.373126066908868e-05, "loss": 0.0259, "step": 30110 }, { "grad_norm": 0.2892582416534424, "learning_rate": 8.371905356498326e-05, "loss": 0.0302, "step": 30120 }, { "grad_norm": 0.19990099966526031, "learning_rate": 8.370684277341288e-05, "loss": 0.0254, "step": 30130 }, { "grad_norm": 0.21943208575248718, "learning_rate": 8.369462829571282e-05, "loss": 0.0308, "step": 30140 }, { "grad_norm": 0.23432578146457672, "learning_rate": 8.36824101332189e-05, "loss": 0.0288, "step": 30150 }, { "grad_norm": 0.24402621388435364, "learning_rate": 8.367018828726721e-05, "loss": 0.0241, "step": 30160 }, { "grad_norm": 0.2386789470911026, "learning_rate": 8.365796275919438e-05, "loss": 0.0239, "step": 30170 }, { "grad_norm": 0.22630198299884796, "learning_rate": 8.364573355033734e-05, "loss": 0.0276, "step": 30180 }, { "grad_norm": 0.24771162867546082, "learning_rate": 8.363350066203346e-05, "loss": 0.028, "step": 30190 }, { "grad_norm": 0.23346886038780212, "learning_rate": 8.362126409562053e-05, "loss": 0.0264, "step": 30200 }, { "grad_norm": 0.20029208064079285, "learning_rate": 8.360902385243667e-05, "loss": 0.0216, "step": 30210 }, { "grad_norm": 0.3393164277076721, "learning_rate": 8.359677993382052e-05, "loss": 0.0305, "step": 30220 }, { "grad_norm": 0.24600672721862793, "learning_rate": 8.358453234111103e-05, "loss": 0.0281, "step": 30230 }, { "grad_norm": 0.26037105917930603, "learning_rate": 8.357228107564756e-05, "loss": 0.0299, "step": 30240 }, { "grad_norm": 0.24355025589466095, "learning_rate": 8.356002613876993e-05, "loss": 0.0252, "step": 30250 }, { "grad_norm": 0.2261502891778946, "learning_rate": 8.35477675318183e-05, "loss": 0.0248, "step": 30260 }, { "grad_norm": 0.19858026504516602, "learning_rate": 8.353550525613323e-05, "loss": 0.0269, "step": 30270 }, { "grad_norm": 0.26302245259284973, "learning_rate": 8.352323931305572e-05, "loss": 0.0261, "step": 30280 }, { "grad_norm": 0.27227720618247986, "learning_rate": 8.351096970392717e-05, "loss": 0.0245, "step": 30290 }, { "grad_norm": 0.24579168856143951, "learning_rate": 8.349869643008937e-05, "loss": 0.0245, "step": 30300 }, { "grad_norm": 0.3086435794830322, "learning_rate": 8.348641949288449e-05, "loss": 0.0299, "step": 30310 }, { "grad_norm": 0.20502080023288727, "learning_rate": 8.34741388936551e-05, "loss": 0.0213, "step": 30320 }, { "grad_norm": 0.2497856616973877, "learning_rate": 8.346185463374423e-05, "loss": 0.0281, "step": 30330 }, { "grad_norm": 0.2944256067276001, "learning_rate": 8.344956671449524e-05, "loss": 0.025, "step": 30340 }, { "grad_norm": 0.19927282631397247, "learning_rate": 8.343727513725192e-05, "loss": 0.0248, "step": 30350 }, { "grad_norm": 0.321687787771225, "learning_rate": 8.342497990335847e-05, "loss": 0.0268, "step": 30360 }, { "grad_norm": 0.31057441234588623, "learning_rate": 8.341268101415946e-05, "loss": 0.0278, "step": 30370 }, { "grad_norm": 0.31174492835998535, "learning_rate": 8.34003784709999e-05, "loss": 0.0293, "step": 30380 }, { "grad_norm": 0.2609703838825226, "learning_rate": 8.338807227522516e-05, "loss": 0.0289, "step": 30390 }, { "grad_norm": 0.297268271446228, "learning_rate": 8.337576242818103e-05, "loss": 0.0284, "step": 30400 }, { "grad_norm": 0.2611660659313202, "learning_rate": 8.336344893121372e-05, "loss": 0.0304, "step": 30410 }, { "grad_norm": 0.22758610546588898, "learning_rate": 8.335113178566977e-05, "loss": 0.0268, "step": 30420 }, { "grad_norm": 0.2449675053358078, "learning_rate": 8.33388109928962e-05, "loss": 0.0269, "step": 30430 }, { "grad_norm": 0.21783949434757233, "learning_rate": 8.33264865542404e-05, "loss": 0.0275, "step": 30440 }, { "grad_norm": 0.2270858883857727, "learning_rate": 8.331415847105013e-05, "loss": 0.0353, "step": 30450 }, { "grad_norm": 0.2070438712835312, "learning_rate": 8.330182674467357e-05, "loss": 0.0269, "step": 30460 }, { "grad_norm": 0.2484843134880066, "learning_rate": 8.32894913764593e-05, "loss": 0.0285, "step": 30470 }, { "grad_norm": 0.2866087257862091, "learning_rate": 8.327715236775633e-05, "loss": 0.0252, "step": 30480 }, { "grad_norm": 0.25174611806869507, "learning_rate": 8.326480971991398e-05, "loss": 0.0264, "step": 30490 }, { "grad_norm": 0.318915456533432, "learning_rate": 8.325246343428206e-05, "loss": 0.0281, "step": 30500 }, { "grad_norm": 0.32530927658081055, "learning_rate": 8.324011351221072e-05, "loss": 0.0261, "step": 30510 }, { "grad_norm": 0.2510374188423157, "learning_rate": 8.322775995505057e-05, "loss": 0.0312, "step": 30520 }, { "grad_norm": 0.2565203607082367, "learning_rate": 8.321540276415254e-05, "loss": 0.0318, "step": 30530 }, { "grad_norm": 0.21589352190494537, "learning_rate": 8.320304194086798e-05, "loss": 0.0314, "step": 30540 }, { "grad_norm": 0.30855879187583923, "learning_rate": 8.31906774865487e-05, "loss": 0.0317, "step": 30550 }, { "grad_norm": 0.26244547963142395, "learning_rate": 8.317830940254682e-05, "loss": 0.0256, "step": 30560 }, { "grad_norm": 0.27461883425712585, "learning_rate": 8.316593769021491e-05, "loss": 0.0254, "step": 30570 }, { "grad_norm": 0.3135557472705841, "learning_rate": 8.315356235090592e-05, "loss": 0.0296, "step": 30580 }, { "grad_norm": 0.26990294456481934, "learning_rate": 8.314118338597319e-05, "loss": 0.0262, "step": 30590 }, { "grad_norm": 0.31203120946884155, "learning_rate": 8.312880079677048e-05, "loss": 0.0292, "step": 30600 }, { "grad_norm": 0.24034589529037476, "learning_rate": 8.311641458465191e-05, "loss": 0.0293, "step": 30610 }, { "grad_norm": 0.2609521746635437, "learning_rate": 8.310402475097205e-05, "loss": 0.0258, "step": 30620 }, { "grad_norm": 0.24606819450855255, "learning_rate": 8.309163129708581e-05, "loss": 0.033, "step": 30630 }, { "grad_norm": 0.20903447270393372, "learning_rate": 8.307923422434852e-05, "loss": 0.0235, "step": 30640 }, { "grad_norm": 0.19279047846794128, "learning_rate": 8.30668335341159e-05, "loss": 0.0237, "step": 30650 }, { "grad_norm": 0.2345513552427292, "learning_rate": 8.305442922774408e-05, "loss": 0.026, "step": 30660 }, { "grad_norm": 0.29866042733192444, "learning_rate": 8.304202130658959e-05, "loss": 0.0312, "step": 30670 }, { "grad_norm": 0.20515592396259308, "learning_rate": 8.302960977200931e-05, "loss": 0.0215, "step": 30680 }, { "grad_norm": 0.22611527144908905, "learning_rate": 8.301719462536058e-05, "loss": 0.0278, "step": 30690 }, { "grad_norm": 0.21707968413829803, "learning_rate": 8.300477586800108e-05, "loss": 0.0291, "step": 30700 }, { "grad_norm": 0.2634601593017578, "learning_rate": 8.299235350128892e-05, "loss": 0.0263, "step": 30710 }, { "grad_norm": 0.23357829451560974, "learning_rate": 8.297992752658258e-05, "loss": 0.0302, "step": 30720 }, { "grad_norm": 0.24272768199443817, "learning_rate": 8.296749794524094e-05, "loss": 0.0266, "step": 30730 }, { "grad_norm": 0.21283721923828125, "learning_rate": 8.295506475862332e-05, "loss": 0.0245, "step": 30740 }, { "grad_norm": 0.20551849901676178, "learning_rate": 8.294262796808933e-05, "loss": 0.0278, "step": 30750 }, { "grad_norm": 0.173317089676857, "learning_rate": 8.293018757499909e-05, "loss": 0.0248, "step": 30760 }, { "grad_norm": 0.2774581015110016, "learning_rate": 8.291774358071305e-05, "loss": 0.025, "step": 30770 }, { "grad_norm": 0.20644840598106384, "learning_rate": 8.290529598659205e-05, "loss": 0.0234, "step": 30780 }, { "grad_norm": 0.2699173092842102, "learning_rate": 8.289284479399738e-05, "loss": 0.0316, "step": 30790 }, { "grad_norm": 0.2581027150154114, "learning_rate": 8.288039000429064e-05, "loss": 0.0299, "step": 30800 }, { "grad_norm": 0.2881316542625427, "learning_rate": 8.286793161883388e-05, "loss": 0.0289, "step": 30810 }, { "grad_norm": 0.20690517127513885, "learning_rate": 8.285546963898954e-05, "loss": 0.0248, "step": 30820 }, { "grad_norm": 0.22425776720046997, "learning_rate": 8.284300406612044e-05, "loss": 0.027, "step": 30830 }, { "grad_norm": 0.2884933650493622, "learning_rate": 8.283053490158978e-05, "loss": 0.029, "step": 30840 }, { "grad_norm": 0.19765520095825195, "learning_rate": 8.28180621467612e-05, "loss": 0.0266, "step": 30850 }, { "grad_norm": 0.2393619418144226, "learning_rate": 8.280558580299867e-05, "loss": 0.0287, "step": 30860 }, { "grad_norm": 0.316313773393631, "learning_rate": 8.279310587166661e-05, "loss": 0.0334, "step": 30870 }, { "grad_norm": 0.2874181866645813, "learning_rate": 8.278062235412978e-05, "loss": 0.0304, "step": 30880 }, { "grad_norm": 0.27513810992240906, "learning_rate": 8.276813525175339e-05, "loss": 0.031, "step": 30890 }, { "grad_norm": 0.19116835296154022, "learning_rate": 8.2755644565903e-05, "loss": 0.0261, "step": 30900 }, { "grad_norm": 0.2464233934879303, "learning_rate": 8.274315029794454e-05, "loss": 0.026, "step": 30910 }, { "grad_norm": 0.3065889775753021, "learning_rate": 8.273065244924443e-05, "loss": 0.029, "step": 30920 }, { "grad_norm": 0.257420152425766, "learning_rate": 8.271815102116936e-05, "loss": 0.0311, "step": 30930 }, { "grad_norm": 0.2166869044303894, "learning_rate": 8.270564601508648e-05, "loss": 0.0327, "step": 30940 }, { "grad_norm": 0.27205848693847656, "learning_rate": 8.269313743236333e-05, "loss": 0.0275, "step": 30950 }, { "grad_norm": 0.6440478563308716, "learning_rate": 8.268062527436783e-05, "loss": 0.0321, "step": 30960 }, { "grad_norm": 0.2453162521123886, "learning_rate": 8.266810954246828e-05, "loss": 0.0326, "step": 30970 }, { "grad_norm": 0.23970694839954376, "learning_rate": 8.265559023803338e-05, "loss": 0.0279, "step": 30980 }, { "grad_norm": 0.25907033681869507, "learning_rate": 8.264306736243223e-05, "loss": 0.0267, "step": 30990 }, { "grad_norm": 0.20750977098941803, "learning_rate": 8.263054091703432e-05, "loss": 0.0281, "step": 31000 }, { "grad_norm": 0.23794546723365784, "learning_rate": 8.26180109032095e-05, "loss": 0.0317, "step": 31010 }, { "grad_norm": 0.3549157679080963, "learning_rate": 8.260547732232807e-05, "loss": 0.034, "step": 31020 }, { "grad_norm": 0.2581976652145386, "learning_rate": 8.259294017576063e-05, "loss": 0.0319, "step": 31030 }, { "grad_norm": 0.2868545651435852, "learning_rate": 8.258039946487828e-05, "loss": 0.0276, "step": 31040 }, { "grad_norm": 0.22518086433410645, "learning_rate": 8.256785519105241e-05, "loss": 0.0257, "step": 31050 }, { "grad_norm": 0.26306384801864624, "learning_rate": 8.255530735565488e-05, "loss": 0.0305, "step": 31060 }, { "grad_norm": 0.2692229449748993, "learning_rate": 8.254275596005787e-05, "loss": 0.0321, "step": 31070 }, { "grad_norm": 0.29958242177963257, "learning_rate": 8.2530201005634e-05, "loss": 0.0286, "step": 31080 }, { "grad_norm": 0.16354826092720032, "learning_rate": 8.251764249375626e-05, "loss": 0.0248, "step": 31090 }, { "grad_norm": 0.23098979890346527, "learning_rate": 8.250508042579803e-05, "loss": 0.0252, "step": 31100 }, { "grad_norm": 0.25206127762794495, "learning_rate": 8.249251480313307e-05, "loss": 0.0259, "step": 31110 }, { "grad_norm": 0.1870948225259781, "learning_rate": 8.247994562713555e-05, "loss": 0.0266, "step": 31120 }, { "grad_norm": 0.23327912390232086, "learning_rate": 8.246737289917998e-05, "loss": 0.0236, "step": 31130 }, { "grad_norm": 0.25134655833244324, "learning_rate": 8.245479662064135e-05, "loss": 0.0261, "step": 31140 }, { "grad_norm": 0.25303539633750916, "learning_rate": 8.244221679289496e-05, "loss": 0.0259, "step": 31150 }, { "grad_norm": 0.20015652477741241, "learning_rate": 8.242963341731652e-05, "loss": 0.0227, "step": 31160 }, { "grad_norm": 0.1942436695098877, "learning_rate": 8.24170464952821e-05, "loss": 0.0279, "step": 31170 }, { "grad_norm": 0.21117837727069855, "learning_rate": 8.240445602816824e-05, "loss": 0.0283, "step": 31180 }, { "grad_norm": 0.21457897126674652, "learning_rate": 8.239186201735179e-05, "loss": 0.0245, "step": 31190 }, { "grad_norm": 0.23400288820266724, "learning_rate": 8.237926446420998e-05, "loss": 0.0274, "step": 31200 }, { "grad_norm": 0.24146798253059387, "learning_rate": 8.236666337012052e-05, "loss": 0.0228, "step": 31210 }, { "grad_norm": 0.2532918155193329, "learning_rate": 8.23540587364614e-05, "loss": 0.0248, "step": 31220 }, { "grad_norm": 0.24724183976650238, "learning_rate": 8.234145056461107e-05, "loss": 0.0257, "step": 31230 }, { "grad_norm": 0.2300424575805664, "learning_rate": 8.232883885594832e-05, "loss": 0.023, "step": 31240 }, { "grad_norm": 0.28682491183280945, "learning_rate": 8.231622361185236e-05, "loss": 0.0293, "step": 31250 }, { "grad_norm": 0.24216055870056152, "learning_rate": 8.230360483370278e-05, "loss": 0.0271, "step": 31260 }, { "grad_norm": 0.41253170371055603, "learning_rate": 8.229098252287953e-05, "loss": 0.0316, "step": 31270 }, { "grad_norm": 0.2443891167640686, "learning_rate": 8.2278356680763e-05, "loss": 0.0308, "step": 31280 }, { "grad_norm": 0.20859293639659882, "learning_rate": 8.22657273087339e-05, "loss": 0.0263, "step": 31290 }, { "grad_norm": 0.2303994745016098, "learning_rate": 8.225309440817336e-05, "loss": 0.0316, "step": 31300 }, { "grad_norm": 0.3172914683818817, "learning_rate": 8.224045798046293e-05, "loss": 0.0326, "step": 31310 }, { "grad_norm": 0.3391183018684387, "learning_rate": 8.22278180269845e-05, "loss": 0.0282, "step": 31320 }, { "grad_norm": 0.2756049335002899, "learning_rate": 8.221517454912031e-05, "loss": 0.0254, "step": 31330 }, { "grad_norm": 0.2231389582157135, "learning_rate": 8.220252754825308e-05, "loss": 0.0266, "step": 31340 }, { "grad_norm": 0.238222137093544, "learning_rate": 8.218987702576586e-05, "loss": 0.0275, "step": 31350 }, { "grad_norm": 0.2964929938316345, "learning_rate": 8.217722298304207e-05, "loss": 0.0284, "step": 31360 }, { "grad_norm": 0.2973387539386749, "learning_rate": 8.216456542146557e-05, "loss": 0.0345, "step": 31370 }, { "grad_norm": 0.3322126865386963, "learning_rate": 8.215190434242055e-05, "loss": 0.0395, "step": 31380 }, { "grad_norm": 0.1811201423406601, "learning_rate": 8.213923974729161e-05, "loss": 0.0332, "step": 31390 }, { "grad_norm": 0.23079752922058105, "learning_rate": 8.212657163746373e-05, "loss": 0.0271, "step": 31400 }, { "grad_norm": 0.2673124372959137, "learning_rate": 8.211390001432227e-05, "loss": 0.0305, "step": 31410 }, { "grad_norm": 0.2834646999835968, "learning_rate": 8.210122487925297e-05, "loss": 0.036, "step": 31420 }, { "grad_norm": 0.29141080379486084, "learning_rate": 8.208854623364202e-05, "loss": 0.0285, "step": 31430 }, { "grad_norm": 0.2678581774234772, "learning_rate": 8.207586407887589e-05, "loss": 0.0303, "step": 31440 }, { "grad_norm": 0.19590137898921967, "learning_rate": 8.206317841634148e-05, "loss": 0.0272, "step": 31450 }, { "grad_norm": 0.22883331775665283, "learning_rate": 8.205048924742609e-05, "loss": 0.0242, "step": 31460 }, { "grad_norm": 0.19439198076725006, "learning_rate": 8.203779657351738e-05, "loss": 0.0264, "step": 31470 }, { "grad_norm": 0.21179132163524628, "learning_rate": 8.20251003960034e-05, "loss": 0.0251, "step": 31480 }, { "grad_norm": 0.24337118864059448, "learning_rate": 8.201240071627258e-05, "loss": 0.0264, "step": 31490 }, { "grad_norm": 0.2437192052602768, "learning_rate": 8.199969753571377e-05, "loss": 0.0253, "step": 31500 }, { "grad_norm": 0.2752099633216858, "learning_rate": 8.198699085571615e-05, "loss": 0.0278, "step": 31510 }, { "grad_norm": 0.2491586059331894, "learning_rate": 8.197428067766928e-05, "loss": 0.0234, "step": 31520 }, { "grad_norm": 0.24962390959262848, "learning_rate": 8.196156700296316e-05, "loss": 0.0223, "step": 31530 }, { "grad_norm": 0.2792713940143585, "learning_rate": 8.194884983298814e-05, "loss": 0.0262, "step": 31540 }, { "grad_norm": 0.2538940906524658, "learning_rate": 8.193612916913491e-05, "loss": 0.0272, "step": 31550 }, { "grad_norm": 0.2544557750225067, "learning_rate": 8.192340501279463e-05, "loss": 0.0257, "step": 31560 }, { "grad_norm": 0.20147019624710083, "learning_rate": 8.191067736535876e-05, "loss": 0.0244, "step": 31570 }, { "grad_norm": 0.2482064664363861, "learning_rate": 8.18979462282192e-05, "loss": 0.0337, "step": 31580 }, { "grad_norm": 0.2638210356235504, "learning_rate": 8.188521160276819e-05, "loss": 0.0323, "step": 31590 }, { "grad_norm": 0.2628701329231262, "learning_rate": 8.187247349039837e-05, "loss": 0.0293, "step": 31600 }, { "grad_norm": 0.3025832772254944, "learning_rate": 8.185973189250278e-05, "loss": 0.0296, "step": 31610 }, { "grad_norm": 0.16626958549022675, "learning_rate": 8.184698681047482e-05, "loss": 0.0285, "step": 31620 }, { "grad_norm": 0.28888413310050964, "learning_rate": 8.183423824570827e-05, "loss": 0.0321, "step": 31630 }, { "grad_norm": 0.27504095435142517, "learning_rate": 8.182148619959725e-05, "loss": 0.026, "step": 31640 }, { "grad_norm": 0.29044950008392334, "learning_rate": 8.180873067353636e-05, "loss": 0.0332, "step": 31650 }, { "grad_norm": 0.22912119328975677, "learning_rate": 8.179597166892052e-05, "loss": 0.0259, "step": 31660 }, { "grad_norm": 0.23869244754314423, "learning_rate": 8.178320918714501e-05, "loss": 0.023, "step": 31670 }, { "grad_norm": 0.22442610561847687, "learning_rate": 8.177044322960554e-05, "loss": 0.0277, "step": 31680 }, { "grad_norm": 0.2112307846546173, "learning_rate": 8.175767379769816e-05, "loss": 0.0279, "step": 31690 }, { "grad_norm": 0.28059500455856323, "learning_rate": 8.174490089281932e-05, "loss": 0.03, "step": 31700 }, { "grad_norm": 0.3285745680332184, "learning_rate": 8.173212451636584e-05, "loss": 0.0314, "step": 31710 }, { "grad_norm": 0.2984437048435211, "learning_rate": 8.171934466973493e-05, "loss": 0.0272, "step": 31720 }, { "grad_norm": 0.24498242139816284, "learning_rate": 8.170656135432418e-05, "loss": 0.0276, "step": 31730 }, { "grad_norm": 0.23220430314540863, "learning_rate": 8.169377457153155e-05, "loss": 0.0258, "step": 31740 }, { "grad_norm": 0.25007128715515137, "learning_rate": 8.168098432275539e-05, "loss": 0.0285, "step": 31750 }, { "grad_norm": 0.23174819350242615, "learning_rate": 8.166819060939442e-05, "loss": 0.0281, "step": 31760 }, { "grad_norm": 0.20686496794223785, "learning_rate": 8.165539343284772e-05, "loss": 0.0259, "step": 31770 }, { "grad_norm": 0.27292680740356445, "learning_rate": 8.16425927945148e-05, "loss": 0.0282, "step": 31780 }, { "grad_norm": 0.3099048435688019, "learning_rate": 8.162978869579551e-05, "loss": 0.0294, "step": 31790 }, { "grad_norm": 0.27178385853767395, "learning_rate": 8.161698113809007e-05, "loss": 0.027, "step": 31800 }, { "grad_norm": 0.2703040838241577, "learning_rate": 8.160417012279911e-05, "loss": 0.0248, "step": 31810 }, { "grad_norm": 0.2308625727891922, "learning_rate": 8.159135565132363e-05, "loss": 0.026, "step": 31820 }, { "grad_norm": 0.22984807193279266, "learning_rate": 8.157853772506498e-05, "loss": 0.0259, "step": 31830 }, { "grad_norm": 0.19249080121517181, "learning_rate": 8.156571634542494e-05, "loss": 0.0236, "step": 31840 }, { "grad_norm": 0.2802889347076416, "learning_rate": 8.15528915138056e-05, "loss": 0.0279, "step": 31850 }, { "grad_norm": 0.25481322407722473, "learning_rate": 8.154006323160949e-05, "loss": 0.026, "step": 31860 }, { "grad_norm": 0.2549186944961548, "learning_rate": 8.152723150023949e-05, "loss": 0.0278, "step": 31870 }, { "grad_norm": 0.2105163037776947, "learning_rate": 8.151439632109886e-05, "loss": 0.0275, "step": 31880 }, { "grad_norm": 0.23417215049266815, "learning_rate": 8.150155769559122e-05, "loss": 0.0247, "step": 31890 }, { "grad_norm": 0.2855096161365509, "learning_rate": 8.148871562512058e-05, "loss": 0.0288, "step": 31900 }, { "grad_norm": 0.28077393770217896, "learning_rate": 8.147587011109136e-05, "loss": 0.0295, "step": 31910 }, { "grad_norm": 0.28179460763931274, "learning_rate": 8.14630211549083e-05, "loss": 0.0323, "step": 31920 }, { "grad_norm": 0.33337709307670593, "learning_rate": 8.145016875797655e-05, "loss": 0.0283, "step": 31930 }, { "grad_norm": 0.3053241968154907, "learning_rate": 8.143731292170164e-05, "loss": 0.0286, "step": 31940 }, { "grad_norm": 0.25979098677635193, "learning_rate": 8.142445364748944e-05, "loss": 0.0242, "step": 31950 }, { "grad_norm": 0.25040504336357117, "learning_rate": 8.141159093674624e-05, "loss": 0.0286, "step": 31960 }, { "grad_norm": 0.23065213859081268, "learning_rate": 8.139872479087869e-05, "loss": 0.0237, "step": 31970 }, { "grad_norm": 0.25816524028778076, "learning_rate": 8.13858552112938e-05, "loss": 0.0253, "step": 31980 }, { "grad_norm": 0.2724524140357971, "learning_rate": 8.137298219939895e-05, "loss": 0.0224, "step": 31990 }, { "grad_norm": 0.2354525923728943, "learning_rate": 8.136010575660196e-05, "loss": 0.0316, "step": 32000 }, { "grad_norm": 0.19178101420402527, "learning_rate": 8.134722588431095e-05, "loss": 0.0231, "step": 32010 }, { "grad_norm": 0.33985310792922974, "learning_rate": 8.133434258393444e-05, "loss": 0.0282, "step": 32020 }, { "grad_norm": 0.17302539944648743, "learning_rate": 8.132145585688134e-05, "loss": 0.0301, "step": 32030 }, { "grad_norm": 0.20751698315143585, "learning_rate": 8.130856570456093e-05, "loss": 0.028, "step": 32040 }, { "grad_norm": 0.22134344279766083, "learning_rate": 8.129567212838283e-05, "loss": 0.0284, "step": 32050 }, { "grad_norm": 0.1927187293767929, "learning_rate": 8.128277512975708e-05, "loss": 0.0263, "step": 32060 }, { "grad_norm": 0.1800108104944229, "learning_rate": 8.126987471009408e-05, "loss": 0.0243, "step": 32070 }, { "grad_norm": 0.2631456255912781, "learning_rate": 8.125697087080459e-05, "loss": 0.0239, "step": 32080 }, { "grad_norm": 0.1945788562297821, "learning_rate": 8.124406361329976e-05, "loss": 0.0258, "step": 32090 }, { "grad_norm": 0.24890092015266418, "learning_rate": 8.123115293899111e-05, "loss": 0.0237, "step": 32100 }, { "grad_norm": 0.3262605667114258, "learning_rate": 8.121823884929055e-05, "loss": 0.0258, "step": 32110 }, { "grad_norm": 0.25094208121299744, "learning_rate": 8.12053213456103e-05, "loss": 0.0241, "step": 32120 }, { "grad_norm": 0.3138270080089569, "learning_rate": 8.119240042936303e-05, "loss": 0.0266, "step": 32130 }, { "grad_norm": 0.2719496190547943, "learning_rate": 8.117947610196175e-05, "loss": 0.0272, "step": 32140 }, { "grad_norm": 0.17116306722164154, "learning_rate": 8.116654836481982e-05, "loss": 0.0233, "step": 32150 }, { "grad_norm": 0.2558582127094269, "learning_rate": 8.115361721935106e-05, "loss": 0.0243, "step": 32160 }, { "grad_norm": 0.26899516582489014, "learning_rate": 8.114068266696953e-05, "loss": 0.0257, "step": 32170 }, { "grad_norm": 0.23673389852046967, "learning_rate": 8.112774470908978e-05, "loss": 0.0252, "step": 32180 }, { "grad_norm": 0.24596458673477173, "learning_rate": 8.111480334712665e-05, "loss": 0.029, "step": 32190 }, { "grad_norm": 0.18348591029644012, "learning_rate": 8.110185858249542e-05, "loss": 0.0216, "step": 32200 }, { "grad_norm": 0.19773916900157928, "learning_rate": 8.108891041661168e-05, "loss": 0.028, "step": 32210 }, { "grad_norm": 0.3967902362346649, "learning_rate": 8.107595885089146e-05, "loss": 0.0254, "step": 32220 }, { "grad_norm": 0.22404000163078308, "learning_rate": 8.106300388675108e-05, "loss": 0.0322, "step": 32230 }, { "grad_norm": 0.2564097046852112, "learning_rate": 8.105004552560731e-05, "loss": 0.028, "step": 32240 }, { "grad_norm": 0.24550825357437134, "learning_rate": 8.103708376887724e-05, "loss": 0.0262, "step": 32250 }, { "grad_norm": 0.3280199468135834, "learning_rate": 8.102411861797836e-05, "loss": 0.0293, "step": 32260 }, { "grad_norm": 0.2821236252784729, "learning_rate": 8.101115007432851e-05, "loss": 0.0257, "step": 32270 }, { "grad_norm": 0.3205975592136383, "learning_rate": 8.09981781393459e-05, "loss": 0.0308, "step": 32280 }, { "grad_norm": 0.2654191851615906, "learning_rate": 8.098520281444915e-05, "loss": 0.0284, "step": 32290 }, { "grad_norm": 0.22912654280662537, "learning_rate": 8.09722241010572e-05, "loss": 0.0273, "step": 32300 }, { "grad_norm": 0.2749459743499756, "learning_rate": 8.095924200058939e-05, "loss": 0.0301, "step": 32310 }, { "grad_norm": 0.25590023398399353, "learning_rate": 8.094625651446541e-05, "loss": 0.0256, "step": 32320 }, { "grad_norm": 0.2345859408378601, "learning_rate": 8.093326764410536e-05, "loss": 0.0243, "step": 32330 }, { "grad_norm": 0.2518680989742279, "learning_rate": 8.092027539092966e-05, "loss": 0.0235, "step": 32340 }, { "grad_norm": 0.28951260447502136, "learning_rate": 8.090727975635913e-05, "loss": 0.0247, "step": 32350 }, { "grad_norm": 0.28059300780296326, "learning_rate": 8.089428074181497e-05, "loss": 0.0275, "step": 32360 }, { "grad_norm": 0.21498988568782806, "learning_rate": 8.088127834871871e-05, "loss": 0.0266, "step": 32370 }, { "grad_norm": 0.2081502079963684, "learning_rate": 8.086827257849226e-05, "loss": 0.0263, "step": 32380 }, { "grad_norm": 0.2927209436893463, "learning_rate": 8.085526343255795e-05, "loss": 0.0267, "step": 32390 }, { "grad_norm": 0.19227148592472076, "learning_rate": 8.084225091233842e-05, "loss": 0.0255, "step": 32400 }, { "grad_norm": 0.17940089106559753, "learning_rate": 8.082923501925668e-05, "loss": 0.0234, "step": 32410 }, { "grad_norm": 0.23665182292461395, "learning_rate": 8.081621575473617e-05, "loss": 0.0285, "step": 32420 }, { "grad_norm": 0.2532392144203186, "learning_rate": 8.080319312020064e-05, "loss": 0.0281, "step": 32430 }, { "grad_norm": 0.23660908639431, "learning_rate": 8.079016711707421e-05, "loss": 0.0257, "step": 32440 }, { "grad_norm": 0.27895835041999817, "learning_rate": 8.077713774678139e-05, "loss": 0.026, "step": 32450 }, { "grad_norm": 0.1941630244255066, "learning_rate": 8.076410501074707e-05, "loss": 0.0269, "step": 32460 }, { "grad_norm": 0.27276691794395447, "learning_rate": 8.075106891039647e-05, "loss": 0.0278, "step": 32470 }, { "grad_norm": 0.21507395803928375, "learning_rate": 8.073802944715523e-05, "loss": 0.0267, "step": 32480 }, { "grad_norm": 0.22728152573108673, "learning_rate": 8.072498662244929e-05, "loss": 0.0244, "step": 32490 }, { "grad_norm": 0.2900875508785248, "learning_rate": 8.0711940437705e-05, "loss": 0.0316, "step": 32500 }, { "grad_norm": 0.24556681513786316, "learning_rate": 8.06988908943491e-05, "loss": 0.0286, "step": 32510 }, { "grad_norm": 0.2843132019042969, "learning_rate": 8.068583799380863e-05, "loss": 0.0269, "step": 32520 }, { "grad_norm": 0.26029741764068604, "learning_rate": 8.067278173751104e-05, "loss": 0.0279, "step": 32530 }, { "grad_norm": 0.2797057032585144, "learning_rate": 8.065972212688417e-05, "loss": 0.0286, "step": 32540 }, { "grad_norm": 0.24070321023464203, "learning_rate": 8.064665916335618e-05, "loss": 0.0257, "step": 32550 }, { "grad_norm": 0.20625101029872894, "learning_rate": 8.063359284835564e-05, "loss": 0.0276, "step": 32560 }, { "grad_norm": 0.20951347053050995, "learning_rate": 8.062052318331142e-05, "loss": 0.0249, "step": 32570 }, { "grad_norm": 0.2670455873012543, "learning_rate": 8.060745016965283e-05, "loss": 0.0238, "step": 32580 }, { "grad_norm": 0.2331392616033554, "learning_rate": 8.059437380880952e-05, "loss": 0.0262, "step": 32590 }, { "grad_norm": 0.216817706823349, "learning_rate": 8.058129410221146e-05, "loss": 0.0262, "step": 32600 }, { "grad_norm": 0.23931290209293365, "learning_rate": 8.056821105128908e-05, "loss": 0.0242, "step": 32610 }, { "grad_norm": 0.24029722809791565, "learning_rate": 8.05551246574731e-05, "loss": 0.0286, "step": 32620 }, { "grad_norm": 0.22944621741771698, "learning_rate": 8.05420349221946e-05, "loss": 0.0248, "step": 32630 }, { "grad_norm": 0.2115623652935028, "learning_rate": 8.05289418468851e-05, "loss": 0.0218, "step": 32640 }, { "grad_norm": 0.20545628666877747, "learning_rate": 8.051584543297642e-05, "loss": 0.0225, "step": 32650 }, { "grad_norm": 0.1779240518808365, "learning_rate": 8.050274568190074e-05, "loss": 0.0258, "step": 32660 }, { "grad_norm": 0.23915694653987885, "learning_rate": 8.048964259509067e-05, "loss": 0.0258, "step": 32670 }, { "grad_norm": 0.19988276064395905, "learning_rate": 8.047653617397914e-05, "loss": 0.0255, "step": 32680 }, { "grad_norm": 0.27024683356285095, "learning_rate": 8.046342641999941e-05, "loss": 0.0259, "step": 32690 }, { "grad_norm": 0.25474220514297485, "learning_rate": 8.045031333458517e-05, "loss": 0.0287, "step": 32700 }, { "grad_norm": 0.20438945293426514, "learning_rate": 8.043719691917047e-05, "loss": 0.0272, "step": 32710 }, { "grad_norm": 0.2501339912414551, "learning_rate": 8.042407717518966e-05, "loss": 0.0282, "step": 32720 }, { "grad_norm": 0.3968904912471771, "learning_rate": 8.041095410407751e-05, "loss": 0.0295, "step": 32730 }, { "grad_norm": 0.2337350845336914, "learning_rate": 8.039782770726913e-05, "loss": 0.0262, "step": 32740 }, { "grad_norm": 0.20984996855258942, "learning_rate": 8.038469798620004e-05, "loss": 0.0271, "step": 32750 }, { "grad_norm": 0.25977659225463867, "learning_rate": 8.037156494230604e-05, "loss": 0.0288, "step": 32760 }, { "grad_norm": 0.24909143149852753, "learning_rate": 8.035842857702338e-05, "loss": 0.0274, "step": 32770 }, { "grad_norm": 0.2747247517108917, "learning_rate": 8.03452888917886e-05, "loss": 0.0278, "step": 32780 }, { "grad_norm": 0.3191737234592438, "learning_rate": 8.033214588803866e-05, "loss": 0.0263, "step": 32790 }, { "grad_norm": 0.2800706624984741, "learning_rate": 8.031899956721083e-05, "loss": 0.0217, "step": 32800 }, { "grad_norm": 0.2449851632118225, "learning_rate": 8.030584993074282e-05, "loss": 0.0245, "step": 32810 }, { "grad_norm": 0.2252771109342575, "learning_rate": 8.02926969800726e-05, "loss": 0.0236, "step": 32820 }, { "grad_norm": 0.24452215433120728, "learning_rate": 8.027954071663859e-05, "loss": 0.0257, "step": 32830 }, { "grad_norm": 0.2617591321468353, "learning_rate": 8.026638114187954e-05, "loss": 0.0237, "step": 32840 }, { "grad_norm": 0.30794477462768555, "learning_rate": 8.025321825723456e-05, "loss": 0.0224, "step": 32850 }, { "grad_norm": 0.2526240050792694, "learning_rate": 8.02400520641431e-05, "loss": 0.0237, "step": 32860 }, { "grad_norm": 0.3383212983608246, "learning_rate": 8.022688256404501e-05, "loss": 0.0318, "step": 32870 }, { "grad_norm": 0.287508487701416, "learning_rate": 8.02137097583805e-05, "loss": 0.0304, "step": 32880 }, { "grad_norm": 0.26247984170913696, "learning_rate": 8.02005336485901e-05, "loss": 0.0347, "step": 32890 }, { "grad_norm": 0.31644144654273987, "learning_rate": 8.018735423611476e-05, "loss": 0.0291, "step": 32900 }, { "grad_norm": 0.3075595498085022, "learning_rate": 8.017417152239574e-05, "loss": 0.0286, "step": 32910 }, { "grad_norm": 0.22405996918678284, "learning_rate": 8.01609855088747e-05, "loss": 0.0278, "step": 32920 }, { "grad_norm": 0.235768124461174, "learning_rate": 8.014779619699362e-05, "loss": 0.0311, "step": 32930 }, { "grad_norm": 0.2346818894147873, "learning_rate": 8.013460358819489e-05, "loss": 0.0239, "step": 32940 }, { "grad_norm": 0.19205906987190247, "learning_rate": 8.01214076839212e-05, "loss": 0.0237, "step": 32950 }, { "grad_norm": 0.19152463972568512, "learning_rate": 8.010820848561565e-05, "loss": 0.0229, "step": 32960 }, { "grad_norm": 0.19708240032196045, "learning_rate": 8.009500599472171e-05, "loss": 0.0243, "step": 32970 }, { "grad_norm": 0.25782060623168945, "learning_rate": 8.008180021268314e-05, "loss": 0.0265, "step": 32980 }, { "grad_norm": 0.26142796874046326, "learning_rate": 8.006859114094414e-05, "loss": 0.0249, "step": 32990 }, { "grad_norm": 0.2525329291820526, "learning_rate": 8.005537878094921e-05, "loss": 0.0265, "step": 33000 }, { "grad_norm": 0.26297688484191895, "learning_rate": 8.004216313414323e-05, "loss": 0.0218, "step": 33010 }, { "grad_norm": 0.2505999505519867, "learning_rate": 8.002894420197149e-05, "loss": 0.023, "step": 33020 }, { "grad_norm": 0.2187090665102005, "learning_rate": 8.001572198587954e-05, "loss": 0.0226, "step": 33030 }, { "grad_norm": 0.22744834423065186, "learning_rate": 8.000249648731338e-05, "loss": 0.0254, "step": 33040 }, { "grad_norm": 0.24195435643196106, "learning_rate": 7.998926770771928e-05, "loss": 0.0247, "step": 33050 }, { "grad_norm": 0.26075077056884766, "learning_rate": 7.997603564854397e-05, "loss": 0.0284, "step": 33060 }, { "grad_norm": 0.27593502402305603, "learning_rate": 7.996280031123448e-05, "loss": 0.0235, "step": 33070 }, { "grad_norm": 0.21335355937480927, "learning_rate": 7.994956169723818e-05, "loss": 0.0231, "step": 33080 }, { "grad_norm": 0.24439628422260284, "learning_rate": 7.993631980800285e-05, "loss": 0.0227, "step": 33090 }, { "grad_norm": 0.2303427904844284, "learning_rate": 7.992307464497659e-05, "loss": 0.0237, "step": 33100 }, { "grad_norm": 0.20959335565567017, "learning_rate": 7.990982620960787e-05, "loss": 0.0252, "step": 33110 }, { "grad_norm": 0.23418062925338745, "learning_rate": 7.989657450334554e-05, "loss": 0.0237, "step": 33120 }, { "grad_norm": 0.2516968548297882, "learning_rate": 7.988331952763877e-05, "loss": 0.0293, "step": 33130 }, { "grad_norm": 0.27089419960975647, "learning_rate": 7.987006128393709e-05, "loss": 0.0236, "step": 33140 }, { "grad_norm": 0.21286743879318237, "learning_rate": 7.985679977369043e-05, "loss": 0.0276, "step": 33150 }, { "grad_norm": 0.23796872794628143, "learning_rate": 7.984353499834902e-05, "loss": 0.0248, "step": 33160 }, { "grad_norm": 0.23757177591323853, "learning_rate": 7.983026695936351e-05, "loss": 0.0239, "step": 33170 }, { "grad_norm": 0.31961894035339355, "learning_rate": 7.981699565818486e-05, "loss": 0.0212, "step": 33180 }, { "grad_norm": 0.2704734206199646, "learning_rate": 7.980372109626437e-05, "loss": 0.0266, "step": 33190 }, { "grad_norm": 0.2615438997745514, "learning_rate": 7.979044327505375e-05, "loss": 0.0216, "step": 33200 }, { "grad_norm": 0.21211862564086914, "learning_rate": 7.977716219600506e-05, "loss": 0.0231, "step": 33210 }, { "grad_norm": 0.2621438801288605, "learning_rate": 7.97638778605707e-05, "loss": 0.0243, "step": 33220 }, { "grad_norm": 0.3346730172634125, "learning_rate": 7.975059027020338e-05, "loss": 0.0212, "step": 33230 }, { "grad_norm": 0.2552284896373749, "learning_rate": 7.973729942635623e-05, "loss": 0.0284, "step": 33240 }, { "grad_norm": 0.25368037819862366, "learning_rate": 7.972400533048273e-05, "loss": 0.0229, "step": 33250 }, { "grad_norm": 0.3741641342639923, "learning_rate": 7.97107079840367e-05, "loss": 0.0277, "step": 33260 }, { "grad_norm": 0.21857064962387085, "learning_rate": 7.969740738847231e-05, "loss": 0.0296, "step": 33270 }, { "grad_norm": 0.17018823325634003, "learning_rate": 7.968410354524411e-05, "loss": 0.0235, "step": 33280 }, { "grad_norm": 0.2130993902683258, "learning_rate": 7.967079645580697e-05, "loss": 0.022, "step": 33290 }, { "grad_norm": 0.20166835188865662, "learning_rate": 7.965748612161612e-05, "loss": 0.0253, "step": 33300 }, { "grad_norm": 0.2526130676269531, "learning_rate": 7.96441725441272e-05, "loss": 0.0254, "step": 33310 }, { "grad_norm": 0.2339572012424469, "learning_rate": 7.963085572479614e-05, "loss": 0.0234, "step": 33320 }, { "grad_norm": 0.24924921989440918, "learning_rate": 7.961753566507924e-05, "loss": 0.0223, "step": 33330 }, { "grad_norm": 0.6323918104171753, "learning_rate": 7.960421236643316e-05, "loss": 0.0272, "step": 33340 }, { "grad_norm": 0.2293563336133957, "learning_rate": 7.959088583031496e-05, "loss": 0.0257, "step": 33350 }, { "grad_norm": 0.23017968237400055, "learning_rate": 7.957755605818194e-05, "loss": 0.0247, "step": 33360 }, { "grad_norm": 0.2493630051612854, "learning_rate": 7.956422305149185e-05, "loss": 0.0278, "step": 33370 }, { "grad_norm": 0.19274266064167023, "learning_rate": 7.95508868117028e-05, "loss": 0.0226, "step": 33380 }, { "grad_norm": 0.21329046785831451, "learning_rate": 7.953754734027318e-05, "loss": 0.0285, "step": 33390 }, { "grad_norm": 0.1918436884880066, "learning_rate": 7.952420463866182e-05, "loss": 0.0246, "step": 33400 }, { "grad_norm": 0.22004148364067078, "learning_rate": 7.951085870832782e-05, "loss": 0.0224, "step": 33410 }, { "grad_norm": 0.28657981753349304, "learning_rate": 7.949750955073067e-05, "loss": 0.0227, "step": 33420 }, { "grad_norm": 0.19299794733524323, "learning_rate": 7.948415716733022e-05, "loss": 0.0208, "step": 33430 }, { "grad_norm": 0.28699547052383423, "learning_rate": 7.947080155958669e-05, "loss": 0.0277, "step": 33440 }, { "grad_norm": 0.30778226256370544, "learning_rate": 7.94574427289606e-05, "loss": 0.0277, "step": 33450 }, { "grad_norm": 0.22907014191150665, "learning_rate": 7.944408067691284e-05, "loss": 0.025, "step": 33460 }, { "grad_norm": 0.26867952942848206, "learning_rate": 7.943071540490473e-05, "loss": 0.0221, "step": 33470 }, { "grad_norm": 0.23765969276428223, "learning_rate": 7.94173469143978e-05, "loss": 0.0235, "step": 33480 }, { "grad_norm": 0.2237054854631424, "learning_rate": 7.940397520685406e-05, "loss": 0.022, "step": 33490 }, { "grad_norm": 0.24380911886692047, "learning_rate": 7.939060028373577e-05, "loss": 0.0242, "step": 33500 }, { "grad_norm": 0.28323429822921753, "learning_rate": 7.937722214650565e-05, "loss": 0.0253, "step": 33510 }, { "grad_norm": 0.20068907737731934, "learning_rate": 7.936384079662666e-05, "loss": 0.0262, "step": 33520 }, { "grad_norm": 0.2579534649848938, "learning_rate": 7.93504562355622e-05, "loss": 0.0233, "step": 33530 }, { "grad_norm": 0.18905596435070038, "learning_rate": 7.933706846477599e-05, "loss": 0.0259, "step": 33540 }, { "grad_norm": 0.24324168264865875, "learning_rate": 7.932367748573206e-05, "loss": 0.0257, "step": 33550 }, { "grad_norm": 0.2216711938381195, "learning_rate": 7.931028329989485e-05, "loss": 0.0216, "step": 33560 }, { "grad_norm": 0.32287412881851196, "learning_rate": 7.929688590872913e-05, "loss": 0.0266, "step": 33570 }, { "grad_norm": 0.31291264295578003, "learning_rate": 7.928348531370003e-05, "loss": 0.0327, "step": 33580 }, { "grad_norm": 0.38395261764526367, "learning_rate": 7.927008151627297e-05, "loss": 0.028, "step": 33590 }, { "grad_norm": 0.28325074911117554, "learning_rate": 7.925667451791383e-05, "loss": 0.0285, "step": 33600 }, { "grad_norm": 0.2978566884994507, "learning_rate": 7.924326432008874e-05, "loss": 0.0248, "step": 33610 }, { "grad_norm": 0.24284857511520386, "learning_rate": 7.922985092426422e-05, "loss": 0.0231, "step": 33620 }, { "grad_norm": 0.25371241569519043, "learning_rate": 7.921643433190717e-05, "loss": 0.0242, "step": 33630 }, { "grad_norm": 0.23717106878757477, "learning_rate": 7.920301454448478e-05, "loss": 0.0268, "step": 33640 }, { "grad_norm": 0.23796598613262177, "learning_rate": 7.918959156346461e-05, "loss": 0.0242, "step": 33650 }, { "grad_norm": 0.27830106019973755, "learning_rate": 7.91761653903146e-05, "loss": 0.0288, "step": 33660 }, { "grad_norm": 0.22601184248924255, "learning_rate": 7.916273602650302e-05, "loss": 0.0253, "step": 33670 }, { "grad_norm": 0.1820819228887558, "learning_rate": 7.914930347349847e-05, "loss": 0.023, "step": 33680 }, { "grad_norm": 0.19774870574474335, "learning_rate": 7.913586773276992e-05, "loss": 0.0245, "step": 33690 }, { "grad_norm": 0.25305160880088806, "learning_rate": 7.912242880578667e-05, "loss": 0.0249, "step": 33700 }, { "grad_norm": 0.27178576588630676, "learning_rate": 7.910898669401839e-05, "loss": 0.023, "step": 33710 }, { "grad_norm": 0.2488345354795456, "learning_rate": 7.909554139893511e-05, "loss": 0.0259, "step": 33720 }, { "grad_norm": 0.253324955701828, "learning_rate": 7.908209292200715e-05, "loss": 0.0275, "step": 33730 }, { "grad_norm": 0.2193860411643982, "learning_rate": 7.906864126470523e-05, "loss": 0.0242, "step": 33740 }, { "grad_norm": 0.23151357471942902, "learning_rate": 7.905518642850041e-05, "loss": 0.0233, "step": 33750 }, { "grad_norm": 0.24076494574546814, "learning_rate": 7.904172841486409e-05, "loss": 0.0242, "step": 33760 }, { "grad_norm": 0.21385321021080017, "learning_rate": 7.902826722526801e-05, "loss": 0.0239, "step": 33770 }, { "grad_norm": 0.2470235675573349, "learning_rate": 7.901480286118427e-05, "loss": 0.0272, "step": 33780 }, { "grad_norm": 0.18998995423316956, "learning_rate": 7.900133532408531e-05, "loss": 0.0237, "step": 33790 }, { "grad_norm": 0.21856696903705597, "learning_rate": 7.898786461544395e-05, "loss": 0.0245, "step": 33800 }, { "grad_norm": 0.3192938268184662, "learning_rate": 7.897439073673325e-05, "loss": 0.0297, "step": 33810 }, { "grad_norm": 0.24707989394664764, "learning_rate": 7.896091368942677e-05, "loss": 0.0302, "step": 33820 }, { "grad_norm": 0.24366804957389832, "learning_rate": 7.894743347499832e-05, "loss": 0.0244, "step": 33830 }, { "grad_norm": 0.23873840272426605, "learning_rate": 7.893395009492203e-05, "loss": 0.0264, "step": 33840 }, { "grad_norm": 0.20660628378391266, "learning_rate": 7.892046355067248e-05, "loss": 0.0226, "step": 33850 }, { "grad_norm": 0.23222562670707703, "learning_rate": 7.890697384372451e-05, "loss": 0.0232, "step": 33860 }, { "grad_norm": 0.21853229403495789, "learning_rate": 7.889348097555336e-05, "loss": 0.0217, "step": 33870 }, { "grad_norm": 0.24957258999347687, "learning_rate": 7.887998494763455e-05, "loss": 0.0296, "step": 33880 }, { "grad_norm": 0.28232550621032715, "learning_rate": 7.886648576144404e-05, "loss": 0.0273, "step": 33890 }, { "grad_norm": 0.23150140047073364, "learning_rate": 7.885298341845802e-05, "loss": 0.0251, "step": 33900 }, { "grad_norm": 0.2042376548051834, "learning_rate": 7.883947792015311e-05, "loss": 0.0282, "step": 33910 }, { "grad_norm": 0.21456791460514069, "learning_rate": 7.882596926800628e-05, "loss": 0.029, "step": 33920 }, { "grad_norm": 0.19774119555950165, "learning_rate": 7.881245746349477e-05, "loss": 0.0234, "step": 33930 }, { "grad_norm": 0.2321728765964508, "learning_rate": 7.879894250809623e-05, "loss": 0.0234, "step": 33940 }, { "grad_norm": 0.17333929240703583, "learning_rate": 7.878542440328865e-05, "loss": 0.0254, "step": 33950 }, { "grad_norm": 0.2156459540128708, "learning_rate": 7.877190315055031e-05, "loss": 0.0232, "step": 33960 }, { "grad_norm": 0.23719945549964905, "learning_rate": 7.875837875135991e-05, "loss": 0.0235, "step": 33970 }, { "grad_norm": 0.2321656495332718, "learning_rate": 7.874485120719646e-05, "loss": 0.025, "step": 33980 }, { "grad_norm": 0.19667665660381317, "learning_rate": 7.873132051953928e-05, "loss": 0.0272, "step": 33990 }, { "grad_norm": 0.22177328169345856, "learning_rate": 7.87177866898681e-05, "loss": 0.0265, "step": 34000 }, { "grad_norm": 0.2627960443496704, "learning_rate": 7.870424971966294e-05, "loss": 0.0294, "step": 34010 }, { "grad_norm": 0.24464315176010132, "learning_rate": 7.869070961040419e-05, "loss": 0.0243, "step": 34020 }, { "grad_norm": 0.26962703466415405, "learning_rate": 7.867716636357257e-05, "loss": 0.0281, "step": 34030 }, { "grad_norm": 0.32449981570243835, "learning_rate": 7.866361998064915e-05, "loss": 0.0301, "step": 34040 }, { "grad_norm": 0.27380678057670593, "learning_rate": 7.865007046311534e-05, "loss": 0.0312, "step": 34050 }, { "grad_norm": 0.27034950256347656, "learning_rate": 7.86365178124529e-05, "loss": 0.0287, "step": 34060 }, { "grad_norm": 0.26727423071861267, "learning_rate": 7.862296203014394e-05, "loss": 0.0238, "step": 34070 }, { "grad_norm": 0.22033119201660156, "learning_rate": 7.860940311767088e-05, "loss": 0.0265, "step": 34080 }, { "grad_norm": 0.17495566606521606, "learning_rate": 7.85958410765165e-05, "loss": 0.0228, "step": 34090 }, { "grad_norm": 0.20461346209049225, "learning_rate": 7.858227590816394e-05, "loss": 0.0324, "step": 34100 }, { "grad_norm": 0.1789553016424179, "learning_rate": 7.856870761409664e-05, "loss": 0.0266, "step": 34110 }, { "grad_norm": 0.1845931112766266, "learning_rate": 7.855513619579846e-05, "loss": 0.0255, "step": 34120 }, { "grad_norm": 0.1885620504617691, "learning_rate": 7.85415616547535e-05, "loss": 0.0301, "step": 34130 }, { "grad_norm": 0.29653704166412354, "learning_rate": 7.852798399244627e-05, "loss": 0.0287, "step": 34140 }, { "grad_norm": 0.20861349999904633, "learning_rate": 7.851440321036161e-05, "loss": 0.024, "step": 34150 }, { "grad_norm": 0.16384907066822052, "learning_rate": 7.850081930998468e-05, "loss": 0.0271, "step": 34160 }, { "grad_norm": 0.23338289558887482, "learning_rate": 7.8487232292801e-05, "loss": 0.0296, "step": 34170 }, { "grad_norm": 0.2027679681777954, "learning_rate": 7.847364216029642e-05, "loss": 0.023, "step": 34180 }, { "grad_norm": 0.1719490885734558, "learning_rate": 7.846004891395716e-05, "loss": 0.0232, "step": 34190 }, { "grad_norm": 0.20947252213954926, "learning_rate": 7.844645255526972e-05, "loss": 0.0238, "step": 34200 }, { "grad_norm": 0.26109954714775085, "learning_rate": 7.843285308572101e-05, "loss": 0.026, "step": 34210 }, { "grad_norm": 0.23185206949710846, "learning_rate": 7.841925050679823e-05, "loss": 0.0249, "step": 34220 }, { "grad_norm": 0.2691453695297241, "learning_rate": 7.840564481998895e-05, "loss": 0.0247, "step": 34230 }, { "grad_norm": 0.21994568407535553, "learning_rate": 7.839203602678105e-05, "loss": 0.0276, "step": 34240 }, { "grad_norm": 0.2493879795074463, "learning_rate": 7.837842412866279e-05, "loss": 0.0279, "step": 34250 }, { "grad_norm": 0.22915329039096832, "learning_rate": 7.836480912712272e-05, "loss": 0.0258, "step": 34260 }, { "grad_norm": 0.22139406204223633, "learning_rate": 7.835119102364976e-05, "loss": 0.0285, "step": 34270 }, { "grad_norm": 0.2568122446537018, "learning_rate": 7.83375698197332e-05, "loss": 0.0242, "step": 34280 }, { "grad_norm": 0.30818983912467957, "learning_rate": 7.83239455168626e-05, "loss": 0.0311, "step": 34290 }, { "grad_norm": 0.17566730082035065, "learning_rate": 7.83103181165279e-05, "loss": 0.0251, "step": 34300 }, { "grad_norm": 0.22310976684093475, "learning_rate": 7.829668762021937e-05, "loss": 0.0241, "step": 34310 }, { "grad_norm": 0.2586321532726288, "learning_rate": 7.828305402942764e-05, "loss": 0.0287, "step": 34320 }, { "grad_norm": 0.20407038927078247, "learning_rate": 7.826941734564363e-05, "loss": 0.025, "step": 34330 }, { "grad_norm": 0.21533244848251343, "learning_rate": 7.825577757035865e-05, "loss": 0.0251, "step": 34340 }, { "grad_norm": 0.2084359973669052, "learning_rate": 7.824213470506431e-05, "loss": 0.0233, "step": 34350 }, { "grad_norm": 0.2808699309825897, "learning_rate": 7.822848875125257e-05, "loss": 0.0275, "step": 34360 }, { "grad_norm": 0.2384951114654541, "learning_rate": 7.821483971041576e-05, "loss": 0.023, "step": 34370 }, { "grad_norm": 0.33546897768974304, "learning_rate": 7.820118758404649e-05, "loss": 0.023, "step": 34380 }, { "grad_norm": 0.25564590096473694, "learning_rate": 7.818753237363776e-05, "loss": 0.0264, "step": 34390 }, { "grad_norm": 0.19305376708507538, "learning_rate": 7.817387408068286e-05, "loss": 0.0232, "step": 34400 }, { "grad_norm": 0.20903119444847107, "learning_rate": 7.816021270667544e-05, "loss": 0.0246, "step": 34410 }, { "grad_norm": 0.1826261281967163, "learning_rate": 7.81465482531095e-05, "loss": 0.0246, "step": 34420 }, { "grad_norm": 0.27059832215309143, "learning_rate": 7.813288072147938e-05, "loss": 0.0261, "step": 34430 }, { "grad_norm": 0.2290775179862976, "learning_rate": 7.811921011327972e-05, "loss": 0.027, "step": 34440 }, { "grad_norm": 0.2760351598262787, "learning_rate": 7.810553643000549e-05, "loss": 0.0319, "step": 34450 }, { "grad_norm": 0.3002055287361145, "learning_rate": 7.809185967315206e-05, "loss": 0.0274, "step": 34460 }, { "grad_norm": 0.22653257846832275, "learning_rate": 7.80781798442151e-05, "loss": 0.0251, "step": 34470 }, { "grad_norm": 0.17329096794128418, "learning_rate": 7.806449694469061e-05, "loss": 0.0251, "step": 34480 }, { "grad_norm": 0.2706916332244873, "learning_rate": 7.805081097607492e-05, "loss": 0.0289, "step": 34490 }, { "grad_norm": 0.22588865458965302, "learning_rate": 7.803712193986474e-05, "loss": 0.025, "step": 34500 }, { "grad_norm": 0.25682884454727173, "learning_rate": 7.802342983755702e-05, "loss": 0.024, "step": 34510 }, { "grad_norm": 0.18842706084251404, "learning_rate": 7.800973467064918e-05, "loss": 0.0218, "step": 34520 }, { "grad_norm": 0.22802132368087769, "learning_rate": 7.799603644063884e-05, "loss": 0.0252, "step": 34530 }, { "grad_norm": 0.23660273849964142, "learning_rate": 7.798233514902405e-05, "loss": 0.029, "step": 34540 }, { "grad_norm": 0.20804718136787415, "learning_rate": 7.796863079730318e-05, "loss": 0.0224, "step": 34550 }, { "grad_norm": 0.23334801197052002, "learning_rate": 7.795492338697488e-05, "loss": 0.0245, "step": 34560 }, { "grad_norm": 0.2733229994773865, "learning_rate": 7.794121291953819e-05, "loss": 0.0219, "step": 34570 }, { "grad_norm": 0.2512863874435425, "learning_rate": 7.792749939649246e-05, "loss": 0.0228, "step": 34580 }, { "grad_norm": 0.1987116038799286, "learning_rate": 7.79137828193374e-05, "loss": 0.0264, "step": 34590 }, { "grad_norm": 0.2762336730957031, "learning_rate": 7.790006318957301e-05, "loss": 0.0237, "step": 34600 }, { "grad_norm": 0.21511779725551605, "learning_rate": 7.788634050869965e-05, "loss": 0.0238, "step": 34610 }, { "grad_norm": 0.2498491257429123, "learning_rate": 7.787261477821803e-05, "loss": 0.024, "step": 34620 }, { "grad_norm": 0.19012130796909332, "learning_rate": 7.785888599962916e-05, "loss": 0.0235, "step": 34630 }, { "grad_norm": 0.19310395419597626, "learning_rate": 7.784515417443439e-05, "loss": 0.0218, "step": 34640 }, { "grad_norm": 0.24628736078739166, "learning_rate": 7.783141930413545e-05, "loss": 0.0216, "step": 34650 }, { "grad_norm": 0.2118874192237854, "learning_rate": 7.78176813902343e-05, "loss": 0.0221, "step": 34660 }, { "grad_norm": 0.24016450345516205, "learning_rate": 7.780394043423336e-05, "loss": 0.0263, "step": 34670 }, { "grad_norm": 0.2257128804922104, "learning_rate": 7.77901964376353e-05, "loss": 0.0222, "step": 34680 }, { "grad_norm": 0.2640903890132904, "learning_rate": 7.777644940194316e-05, "loss": 0.0205, "step": 34690 }, { "grad_norm": 0.20823387801647186, "learning_rate": 7.776269932866023e-05, "loss": 0.0241, "step": 34700 }, { "grad_norm": 0.2262853980064392, "learning_rate": 7.774894621929026e-05, "loss": 0.0258, "step": 34710 }, { "grad_norm": 0.21648137271404266, "learning_rate": 7.773519007533725e-05, "loss": 0.0307, "step": 34720 }, { "grad_norm": 0.2057930827140808, "learning_rate": 7.772143089830556e-05, "loss": 0.0269, "step": 34730 }, { "grad_norm": 0.2770114839076996, "learning_rate": 7.770766868969985e-05, "loss": 0.027, "step": 34740 }, { "grad_norm": 0.18436706066131592, "learning_rate": 7.769390345102518e-05, "loss": 0.0197, "step": 34750 }, { "grad_norm": 0.20537258684635162, "learning_rate": 7.768013518378683e-05, "loss": 0.0239, "step": 34760 }, { "grad_norm": 0.25156185030937195, "learning_rate": 7.766636388949053e-05, "loss": 0.0259, "step": 34770 }, { "grad_norm": 0.22267653048038483, "learning_rate": 7.765258956964229e-05, "loss": 0.0244, "step": 34780 }, { "grad_norm": 0.26809829473495483, "learning_rate": 7.76388122257484e-05, "loss": 0.0264, "step": 34790 }, { "grad_norm": 0.18756003677845, "learning_rate": 7.762503185931558e-05, "loss": 0.0207, "step": 34800 }, { "grad_norm": 0.23048552870750427, "learning_rate": 7.76112484718508e-05, "loss": 0.0237, "step": 34810 }, { "grad_norm": 0.32593873143196106, "learning_rate": 7.75974620648614e-05, "loss": 0.0269, "step": 34820 }, { "grad_norm": 0.229127436876297, "learning_rate": 7.758367263985503e-05, "loss": 0.0267, "step": 34830 }, { "grad_norm": 0.19112256169319153, "learning_rate": 7.75698801983397e-05, "loss": 0.0236, "step": 34840 }, { "grad_norm": 0.19438838958740234, "learning_rate": 7.755608474182372e-05, "loss": 0.0251, "step": 34850 }, { "grad_norm": 0.25568047165870667, "learning_rate": 7.754228627181574e-05, "loss": 0.0262, "step": 34860 }, { "grad_norm": 0.1907448172569275, "learning_rate": 7.752848478982476e-05, "loss": 0.0212, "step": 34870 }, { "grad_norm": 0.18730755150318146, "learning_rate": 7.751468029736006e-05, "loss": 0.0247, "step": 34880 }, { "grad_norm": 0.204233318567276, "learning_rate": 7.750087279593129e-05, "loss": 0.0282, "step": 34890 }, { "grad_norm": 0.24089723825454712, "learning_rate": 7.748706228704843e-05, "loss": 0.0262, "step": 34900 }, { "grad_norm": 0.22558489441871643, "learning_rate": 7.747324877222176e-05, "loss": 0.0262, "step": 34910 }, { "grad_norm": 0.2590050995349884, "learning_rate": 7.745943225296188e-05, "loss": 0.0258, "step": 34920 }, { "grad_norm": 0.22859148681163788, "learning_rate": 7.744561273077981e-05, "loss": 0.0224, "step": 34930 }, { "grad_norm": 0.27094659209251404, "learning_rate": 7.743179020718678e-05, "loss": 0.0318, "step": 34940 }, { "grad_norm": 0.4479856491088867, "learning_rate": 7.741796468369443e-05, "loss": 0.0311, "step": 34950 }, { "grad_norm": 0.20426274836063385, "learning_rate": 7.740413616181466e-05, "loss": 0.028, "step": 34960 }, { "grad_norm": 0.28414154052734375, "learning_rate": 7.739030464305978e-05, "loss": 0.0306, "step": 34970 }, { "grad_norm": 0.22373400628566742, "learning_rate": 7.737647012894235e-05, "loss": 0.0256, "step": 34980 }, { "grad_norm": 0.22771160304546356, "learning_rate": 7.736263262097532e-05, "loss": 0.0231, "step": 34990 }, { "grad_norm": 0.2187865674495697, "learning_rate": 7.734879212067192e-05, "loss": 0.0249, "step": 35000 }, { "grad_norm": 0.2289605587720871, "learning_rate": 7.733494862954573e-05, "loss": 0.0237, "step": 35010 }, { "grad_norm": 0.23232004046440125, "learning_rate": 7.732110214911066e-05, "loss": 0.0226, "step": 35020 }, { "grad_norm": 0.2803749740123749, "learning_rate": 7.730725268088094e-05, "loss": 0.023, "step": 35030 }, { "grad_norm": 0.1995522379875183, "learning_rate": 7.729340022637112e-05, "loss": 0.025, "step": 35040 }, { "grad_norm": 0.279740571975708, "learning_rate": 7.727954478709607e-05, "loss": 0.0261, "step": 35050 }, { "grad_norm": 0.21438682079315186, "learning_rate": 7.726568636457103e-05, "loss": 0.0238, "step": 35060 }, { "grad_norm": 0.24400712549686432, "learning_rate": 7.725182496031153e-05, "loss": 0.0251, "step": 35070 }, { "grad_norm": 0.2327231615781784, "learning_rate": 7.723796057583341e-05, "loss": 0.0278, "step": 35080 }, { "grad_norm": 0.2843463122844696, "learning_rate": 7.722409321265287e-05, "loss": 0.0233, "step": 35090 }, { "grad_norm": 0.2565215826034546, "learning_rate": 7.721022287228645e-05, "loss": 0.0291, "step": 35100 }, { "grad_norm": 0.21396592259407043, "learning_rate": 7.719634955625094e-05, "loss": 0.0246, "step": 35110 }, { "grad_norm": 0.6929619312286377, "learning_rate": 7.718247326606355e-05, "loss": 0.0277, "step": 35120 }, { "grad_norm": 0.4240470230579376, "learning_rate": 7.716859400324175e-05, "loss": 0.03, "step": 35130 }, { "grad_norm": 0.25809401273727417, "learning_rate": 7.715471176930335e-05, "loss": 0.0299, "step": 35140 }, { "grad_norm": 0.31512340903282166, "learning_rate": 7.714082656576651e-05, "loss": 0.0285, "step": 35150 }, { "grad_norm": 0.25255727767944336, "learning_rate": 7.712693839414968e-05, "loss": 0.025, "step": 35160 }, { "grad_norm": 0.21138527989387512, "learning_rate": 7.711304725597164e-05, "loss": 0.0354, "step": 35170 }, { "grad_norm": 0.32132241129875183, "learning_rate": 7.709915315275151e-05, "loss": 0.0266, "step": 35180 }, { "grad_norm": 0.20903606712818146, "learning_rate": 7.708525608600876e-05, "loss": 0.0306, "step": 35190 }, { "grad_norm": 0.1964380294084549, "learning_rate": 7.707135605726311e-05, "loss": 0.0234, "step": 35200 }, { "grad_norm": 0.20063915848731995, "learning_rate": 7.705745306803466e-05, "loss": 0.0253, "step": 35210 }, { "grad_norm": 0.37811222672462463, "learning_rate": 7.704354711984383e-05, "loss": 0.0284, "step": 35220 }, { "grad_norm": 0.3607385456562042, "learning_rate": 7.702963821421133e-05, "loss": 0.0304, "step": 35230 }, { "grad_norm": 0.2959858179092407, "learning_rate": 7.701572635265826e-05, "loss": 0.0285, "step": 35240 }, { "grad_norm": 0.23044776916503906, "learning_rate": 7.700181153670596e-05, "loss": 0.0248, "step": 35250 }, { "grad_norm": 0.2626868784427643, "learning_rate": 7.698789376787616e-05, "loss": 0.0259, "step": 35260 }, { "grad_norm": 0.24143844842910767, "learning_rate": 7.697397304769085e-05, "loss": 0.025, "step": 35270 }, { "grad_norm": 0.329928457736969, "learning_rate": 7.696004937767241e-05, "loss": 0.0303, "step": 35280 }, { "grad_norm": 0.27602672576904297, "learning_rate": 7.694612275934352e-05, "loss": 0.0241, "step": 35290 }, { "grad_norm": 0.2616387903690338, "learning_rate": 7.693219319422714e-05, "loss": 0.0245, "step": 35300 }, { "grad_norm": 0.22985966503620148, "learning_rate": 7.69182606838466e-05, "loss": 0.0231, "step": 35310 }, { "grad_norm": 0.24028082191944122, "learning_rate": 7.690432522972558e-05, "loss": 0.0235, "step": 35320 }, { "grad_norm": 0.26973995566368103, "learning_rate": 7.689038683338796e-05, "loss": 0.0259, "step": 35330 }, { "grad_norm": 0.24256756901741028, "learning_rate": 7.687644549635808e-05, "loss": 0.022, "step": 35340 }, { "grad_norm": 0.26267585158348083, "learning_rate": 7.686250122016053e-05, "loss": 0.0217, "step": 35350 }, { "grad_norm": 0.2125687152147293, "learning_rate": 7.684855400632022e-05, "loss": 0.0213, "step": 35360 }, { "grad_norm": 0.23263229429721832, "learning_rate": 7.683460385636243e-05, "loss": 0.0269, "step": 35370 }, { "grad_norm": 0.21787187457084656, "learning_rate": 7.68206507718127e-05, "loss": 0.026, "step": 35380 }, { "grad_norm": 0.3205558657646179, "learning_rate": 7.680669475419693e-05, "loss": 0.0286, "step": 35390 }, { "grad_norm": 0.2197900265455246, "learning_rate": 7.679273580504132e-05, "loss": 0.0276, "step": 35400 }, { "grad_norm": 0.2747122347354889, "learning_rate": 7.67787739258724e-05, "loss": 0.0282, "step": 35410 }, { "grad_norm": 0.27227792143821716, "learning_rate": 7.676480911821705e-05, "loss": 0.0297, "step": 35420 }, { "grad_norm": 0.3536750078201294, "learning_rate": 7.675084138360238e-05, "loss": 0.026, "step": 35430 }, { "grad_norm": 0.21670520305633545, "learning_rate": 7.673687072355592e-05, "loss": 0.0249, "step": 35440 }, { "grad_norm": 0.2650170624256134, "learning_rate": 7.67228971396055e-05, "loss": 0.0282, "step": 35450 }, { "grad_norm": 0.2389940619468689, "learning_rate": 7.670892063327922e-05, "loss": 0.0282, "step": 35460 }, { "grad_norm": 0.22524209320545197, "learning_rate": 7.669494120610552e-05, "loss": 0.0264, "step": 35470 }, { "grad_norm": 0.21689948439598083, "learning_rate": 7.66809588596132e-05, "loss": 0.0306, "step": 35480 }, { "grad_norm": 0.22887270152568817, "learning_rate": 7.666697359533132e-05, "loss": 0.0245, "step": 35490 }, { "grad_norm": 0.24834251403808594, "learning_rate": 7.665298541478932e-05, "loss": 0.026, "step": 35500 }, { "grad_norm": 0.27783578634262085, "learning_rate": 7.66389943195169e-05, "loss": 0.0275, "step": 35510 }, { "grad_norm": 0.19986437261104584, "learning_rate": 7.66250003110441e-05, "loss": 0.0208, "step": 35520 }, { "grad_norm": 0.21595454216003418, "learning_rate": 7.66110033909013e-05, "loss": 0.0237, "step": 35530 }, { "grad_norm": 0.2149573415517807, "learning_rate": 7.659700356061918e-05, "loss": 0.0241, "step": 35540 }, { "grad_norm": 0.28516885638237, "learning_rate": 7.658300082172875e-05, "loss": 0.0268, "step": 35550 }, { "grad_norm": 0.2057037502527237, "learning_rate": 7.65689951757613e-05, "loss": 0.0293, "step": 35560 }, { "grad_norm": 0.21335631608963013, "learning_rate": 7.655498662424849e-05, "loss": 0.0227, "step": 35570 }, { "grad_norm": 0.19832895696163177, "learning_rate": 7.654097516872227e-05, "loss": 0.0265, "step": 35580 }, { "grad_norm": 0.26321321725845337, "learning_rate": 7.65269608107149e-05, "loss": 0.0278, "step": 35590 }, { "grad_norm": 0.2731079161167145, "learning_rate": 7.6512943551759e-05, "loss": 0.0293, "step": 35600 }, { "grad_norm": 0.2332087606191635, "learning_rate": 7.649892339338743e-05, "loss": 0.0248, "step": 35610 }, { "grad_norm": 0.2269972562789917, "learning_rate": 7.648490033713344e-05, "loss": 0.0211, "step": 35620 }, { "grad_norm": 0.26330137252807617, "learning_rate": 7.647087438453058e-05, "loss": 0.0231, "step": 35630 }, { "grad_norm": 0.25713929533958435, "learning_rate": 7.64568455371127e-05, "loss": 0.022, "step": 35640 }, { "grad_norm": 0.21258972585201263, "learning_rate": 7.644281379641396e-05, "loss": 0.0241, "step": 35650 }, { "grad_norm": 0.18975044786930084, "learning_rate": 7.642877916396887e-05, "loss": 0.0208, "step": 35660 }, { "grad_norm": 0.2830851972103119, "learning_rate": 7.641474164131221e-05, "loss": 0.0226, "step": 35670 }, { "grad_norm": 0.2610153555870056, "learning_rate": 7.640070122997913e-05, "loss": 0.0267, "step": 35680 }, { "grad_norm": 0.2085733413696289, "learning_rate": 7.638665793150505e-05, "loss": 0.0229, "step": 35690 }, { "grad_norm": 0.22975724935531616, "learning_rate": 7.637261174742574e-05, "loss": 0.0246, "step": 35700 }, { "grad_norm": 0.20685651898384094, "learning_rate": 7.635856267927725e-05, "loss": 0.026, "step": 35710 }, { "grad_norm": 0.2690337896347046, "learning_rate": 7.634451072859597e-05, "loss": 0.0243, "step": 35720 }, { "grad_norm": 0.2516947090625763, "learning_rate": 7.633045589691863e-05, "loss": 0.0258, "step": 35730 }, { "grad_norm": 0.17676278948783875, "learning_rate": 7.63163981857822e-05, "loss": 0.0217, "step": 35740 }, { "grad_norm": 0.20976798236370087, "learning_rate": 7.630233759672403e-05, "loss": 0.0256, "step": 35750 }, { "grad_norm": 0.30440348386764526, "learning_rate": 7.628827413128177e-05, "loss": 0.0252, "step": 35760 }, { "grad_norm": 0.2497008591890335, "learning_rate": 7.627420779099338e-05, "loss": 0.0237, "step": 35770 }, { "grad_norm": 0.2251412719488144, "learning_rate": 7.626013857739711e-05, "loss": 0.0226, "step": 35780 }, { "grad_norm": 0.26481756567955017, "learning_rate": 7.624606649203158e-05, "loss": 0.0216, "step": 35790 }, { "grad_norm": 0.3591984510421753, "learning_rate": 7.623199153643569e-05, "loss": 0.0263, "step": 35800 }, { "grad_norm": 0.20370616018772125, "learning_rate": 7.621791371214863e-05, "loss": 0.0239, "step": 35810 }, { "grad_norm": 0.21990785002708435, "learning_rate": 7.620383302070995e-05, "loss": 0.0211, "step": 35820 }, { "grad_norm": 0.19315309822559357, "learning_rate": 7.61897494636595e-05, "loss": 0.0214, "step": 35830 }, { "grad_norm": 0.24545808136463165, "learning_rate": 7.617566304253739e-05, "loss": 0.0226, "step": 35840 }, { "grad_norm": 0.1783958524465561, "learning_rate": 7.616157375888416e-05, "loss": 0.0221, "step": 35850 }, { "grad_norm": 0.18890924751758575, "learning_rate": 7.614748161424053e-05, "loss": 0.0233, "step": 35860 }, { "grad_norm": 0.19937944412231445, "learning_rate": 7.613338661014763e-05, "loss": 0.0239, "step": 35870 }, { "grad_norm": 0.24962149560451508, "learning_rate": 7.611928874814686e-05, "loss": 0.0219, "step": 35880 }, { "grad_norm": 0.20012325048446655, "learning_rate": 7.610518802977996e-05, "loss": 0.0305, "step": 35890 }, { "grad_norm": 0.18691162765026093, "learning_rate": 7.609108445658893e-05, "loss": 0.022, "step": 35900 }, { "grad_norm": 0.1772705316543579, "learning_rate": 7.607697803011612e-05, "loss": 0.0241, "step": 35910 }, { "grad_norm": 0.3044378459453583, "learning_rate": 7.606286875190421e-05, "loss": 0.0221, "step": 35920 }, { "grad_norm": 0.24075107276439667, "learning_rate": 7.604875662349617e-05, "loss": 0.0227, "step": 35930 }, { "grad_norm": 0.20574216544628143, "learning_rate": 7.603464164643525e-05, "loss": 0.0251, "step": 35940 }, { "grad_norm": 0.2115067094564438, "learning_rate": 7.602052382226507e-05, "loss": 0.0242, "step": 35950 }, { "grad_norm": 0.20775863528251648, "learning_rate": 7.600640315252954e-05, "loss": 0.024, "step": 35960 }, { "grad_norm": 0.19863294064998627, "learning_rate": 7.599227963877284e-05, "loss": 0.026, "step": 35970 }, { "grad_norm": 0.22909832000732422, "learning_rate": 7.597815328253954e-05, "loss": 0.0241, "step": 35980 }, { "grad_norm": 0.2807990610599518, "learning_rate": 7.596402408537443e-05, "loss": 0.0259, "step": 35990 }, { "grad_norm": 0.243717759847641, "learning_rate": 7.59498920488227e-05, "loss": 0.0262, "step": 36000 }, { "grad_norm": 0.22248418629169464, "learning_rate": 7.593575717442979e-05, "loss": 0.0271, "step": 36010 }, { "grad_norm": 0.21150006353855133, "learning_rate": 7.592161946374147e-05, "loss": 0.0207, "step": 36020 }, { "grad_norm": 0.22822579741477966, "learning_rate": 7.590747891830381e-05, "loss": 0.0242, "step": 36030 }, { "grad_norm": 0.25190654397010803, "learning_rate": 7.58933355396632e-05, "loss": 0.0284, "step": 36040 }, { "grad_norm": 0.2226772904396057, "learning_rate": 7.587918932936636e-05, "loss": 0.0226, "step": 36050 }, { "grad_norm": 0.2723824381828308, "learning_rate": 7.586504028896028e-05, "loss": 0.0261, "step": 36060 }, { "grad_norm": 0.22653080523014069, "learning_rate": 7.585088841999228e-05, "loss": 0.0235, "step": 36070 }, { "grad_norm": 0.2253563553094864, "learning_rate": 7.583673372400999e-05, "loss": 0.027, "step": 36080 }, { "grad_norm": 0.256389319896698, "learning_rate": 7.582257620256134e-05, "loss": 0.027, "step": 36090 }, { "grad_norm": 0.20896132290363312, "learning_rate": 7.580841585719458e-05, "loss": 0.0222, "step": 36100 }, { "grad_norm": 0.31856560707092285, "learning_rate": 7.579425268945825e-05, "loss": 0.0298, "step": 36110 }, { "grad_norm": 0.2288067787885666, "learning_rate": 7.578008670090127e-05, "loss": 0.0221, "step": 36120 }, { "grad_norm": 0.2893092930316925, "learning_rate": 7.576591789307272e-05, "loss": 0.0231, "step": 36130 }, { "grad_norm": 0.18820340931415558, "learning_rate": 7.575174626752216e-05, "loss": 0.024, "step": 36140 }, { "grad_norm": 0.24150903522968292, "learning_rate": 7.573757182579934e-05, "loss": 0.0217, "step": 36150 }, { "grad_norm": 0.3293170630931854, "learning_rate": 7.572339456945435e-05, "loss": 0.0252, "step": 36160 }, { "grad_norm": 0.2757301330566406, "learning_rate": 7.570921450003762e-05, "loss": 0.022, "step": 36170 }, { "grad_norm": 0.2423810511827469, "learning_rate": 7.569503161909983e-05, "loss": 0.0204, "step": 36180 }, { "grad_norm": 0.2294469028711319, "learning_rate": 7.568084592819202e-05, "loss": 0.0279, "step": 36190 }, { "grad_norm": 0.27090445160865784, "learning_rate": 7.566665742886551e-05, "loss": 0.0245, "step": 36200 }, { "grad_norm": 0.24865613877773285, "learning_rate": 7.565246612267194e-05, "loss": 0.0257, "step": 36210 }, { "grad_norm": 0.21410565078258514, "learning_rate": 7.563827201116325e-05, "loss": 0.0263, "step": 36220 }, { "grad_norm": 0.2376389056444168, "learning_rate": 7.562407509589164e-05, "loss": 0.0255, "step": 36230 }, { "grad_norm": 0.2352200150489807, "learning_rate": 7.560987537840973e-05, "loss": 0.0266, "step": 36240 }, { "grad_norm": 0.2611333429813385, "learning_rate": 7.559567286027036e-05, "loss": 0.0271, "step": 36250 }, { "grad_norm": 0.3511757552623749, "learning_rate": 7.558146754302668e-05, "loss": 0.0252, "step": 36260 }, { "grad_norm": 0.2535261809825897, "learning_rate": 7.556725942823217e-05, "loss": 0.0275, "step": 36270 }, { "grad_norm": 0.19282075762748718, "learning_rate": 7.555304851744061e-05, "loss": 0.0256, "step": 36280 }, { "grad_norm": 0.2341534048318863, "learning_rate": 7.553883481220608e-05, "loss": 0.028, "step": 36290 }, { "grad_norm": 0.27241647243499756, "learning_rate": 7.552461831408298e-05, "loss": 0.0264, "step": 36300 }, { "grad_norm": 0.14773030579090118, "learning_rate": 7.551039902462599e-05, "loss": 0.0229, "step": 36310 }, { "grad_norm": 0.25962069630622864, "learning_rate": 7.549617694539011e-05, "loss": 0.0278, "step": 36320 }, { "grad_norm": 0.1862228661775589, "learning_rate": 7.548195207793066e-05, "loss": 0.0233, "step": 36330 }, { "grad_norm": 0.23265184462070465, "learning_rate": 7.546772442380323e-05, "loss": 0.0254, "step": 36340 }, { "grad_norm": 0.24639451503753662, "learning_rate": 7.545349398456376e-05, "loss": 0.0249, "step": 36350 }, { "grad_norm": 0.2536628246307373, "learning_rate": 7.543926076176845e-05, "loss": 0.0228, "step": 36360 }, { "grad_norm": 0.23157835006713867, "learning_rate": 7.542502475697385e-05, "loss": 0.0263, "step": 36370 }, { "grad_norm": 0.21400363743305206, "learning_rate": 7.541078597173675e-05, "loss": 0.0222, "step": 36380 }, { "grad_norm": 0.2612849771976471, "learning_rate": 7.539654440761431e-05, "loss": 0.0219, "step": 36390 }, { "grad_norm": 0.20150171220302582, "learning_rate": 7.538230006616395e-05, "loss": 0.0252, "step": 36400 }, { "grad_norm": 0.22781439125537872, "learning_rate": 7.536805294894342e-05, "loss": 0.0219, "step": 36410 }, { "grad_norm": 0.20992761850357056, "learning_rate": 7.535380305751076e-05, "loss": 0.0234, "step": 36420 }, { "grad_norm": 0.19786983728408813, "learning_rate": 7.533955039342431e-05, "loss": 0.0209, "step": 36430 }, { "grad_norm": 0.24536819756031036, "learning_rate": 7.532529495824274e-05, "loss": 0.0246, "step": 36440 }, { "grad_norm": 0.17546901106834412, "learning_rate": 7.531103675352497e-05, "loss": 0.0266, "step": 36450 }, { "grad_norm": 0.2812611758708954, "learning_rate": 7.529677578083029e-05, "loss": 0.0258, "step": 36460 }, { "grad_norm": 0.22043874859809875, "learning_rate": 7.528251204171823e-05, "loss": 0.0224, "step": 36470 }, { "grad_norm": 0.2195114642381668, "learning_rate": 7.526824553774867e-05, "loss": 0.0227, "step": 36480 }, { "grad_norm": 0.20025596022605896, "learning_rate": 7.525397627048178e-05, "loss": 0.0202, "step": 36490 }, { "grad_norm": 0.1749761700630188, "learning_rate": 7.523970424147802e-05, "loss": 0.023, "step": 36500 }, { "grad_norm": 0.23850145936012268, "learning_rate": 7.522542945229813e-05, "loss": 0.0228, "step": 36510 }, { "grad_norm": 0.22266411781311035, "learning_rate": 7.521115190450321e-05, "loss": 0.0244, "step": 36520 }, { "grad_norm": 0.24661539494991302, "learning_rate": 7.519687159965462e-05, "loss": 0.0275, "step": 36530 }, { "grad_norm": 0.3006591796875, "learning_rate": 7.518258853931403e-05, "loss": 0.025, "step": 36540 }, { "grad_norm": 0.21154052019119263, "learning_rate": 7.516830272504342e-05, "loss": 0.0277, "step": 36550 }, { "grad_norm": 0.21821273863315582, "learning_rate": 7.515401415840509e-05, "loss": 0.0262, "step": 36560 }, { "grad_norm": 0.2810736298561096, "learning_rate": 7.513972284096155e-05, "loss": 0.0268, "step": 36570 }, { "grad_norm": 0.46718743443489075, "learning_rate": 7.512542877427574e-05, "loss": 0.0215, "step": 36580 }, { "grad_norm": 0.23185975849628448, "learning_rate": 7.51111319599108e-05, "loss": 0.025, "step": 36590 }, { "grad_norm": 0.39050114154815674, "learning_rate": 7.509683239943024e-05, "loss": 0.0292, "step": 36600 }, { "grad_norm": 0.2328527420759201, "learning_rate": 7.50825300943978e-05, "loss": 0.0232, "step": 36610 }, { "grad_norm": 0.1742914319038391, "learning_rate": 7.50682250463776e-05, "loss": 0.0252, "step": 36620 }, { "grad_norm": 0.25878196954727173, "learning_rate": 7.5053917256934e-05, "loss": 0.0238, "step": 36630 }, { "grad_norm": 0.44512173533439636, "learning_rate": 7.503960672763166e-05, "loss": 0.0263, "step": 36640 }, { "grad_norm": 0.22793178260326385, "learning_rate": 7.502529346003559e-05, "loss": 0.0193, "step": 36650 }, { "grad_norm": 0.19024759531021118, "learning_rate": 7.501097745571106e-05, "loss": 0.0234, "step": 36660 }, { "grad_norm": 0.22718888521194458, "learning_rate": 7.499665871622361e-05, "loss": 0.024, "step": 36670 }, { "grad_norm": 0.20858047902584076, "learning_rate": 7.498233724313917e-05, "loss": 0.024, "step": 36680 }, { "grad_norm": 0.22521519660949707, "learning_rate": 7.496801303802387e-05, "loss": 0.0232, "step": 36690 }, { "grad_norm": 0.2282027155160904, "learning_rate": 7.495368610244423e-05, "loss": 0.0189, "step": 36700 }, { "grad_norm": 0.21152007579803467, "learning_rate": 7.493935643796697e-05, "loss": 0.0285, "step": 36710 }, { "grad_norm": 0.24103038012981415, "learning_rate": 7.49250240461592e-05, "loss": 0.0203, "step": 36720 }, { "grad_norm": 0.21813057363033295, "learning_rate": 7.491068892858826e-05, "loss": 0.022, "step": 36730 }, { "grad_norm": 0.18629661202430725, "learning_rate": 7.489635108682184e-05, "loss": 0.0203, "step": 36740 }, { "grad_norm": 0.26166781783103943, "learning_rate": 7.48820105224279e-05, "loss": 0.0261, "step": 36750 }, { "grad_norm": 0.2241339236497879, "learning_rate": 7.48676672369747e-05, "loss": 0.0264, "step": 36760 }, { "grad_norm": 0.24894367158412933, "learning_rate": 7.485332123203079e-05, "loss": 0.0274, "step": 36770 }, { "grad_norm": 0.24947749078273773, "learning_rate": 7.483897250916504e-05, "loss": 0.0238, "step": 36780 }, { "grad_norm": 0.24106194078922272, "learning_rate": 7.48246210699466e-05, "loss": 0.019, "step": 36790 }, { "grad_norm": 0.24412313103675842, "learning_rate": 7.481026691594492e-05, "loss": 0.0214, "step": 36800 }, { "grad_norm": 0.2444833368062973, "learning_rate": 7.479591004872975e-05, "loss": 0.023, "step": 36810 }, { "grad_norm": 0.16711796820163727, "learning_rate": 7.478155046987114e-05, "loss": 0.0201, "step": 36820 }, { "grad_norm": 0.21525588631629944, "learning_rate": 7.476718818093944e-05, "loss": 0.0226, "step": 36830 }, { "grad_norm": 0.23083722591400146, "learning_rate": 7.475282318350527e-05, "loss": 0.0224, "step": 36840 }, { "grad_norm": 0.22657634317874908, "learning_rate": 7.473845547913957e-05, "loss": 0.0279, "step": 36850 }, { "grad_norm": 0.1757124662399292, "learning_rate": 7.472408506941357e-05, "loss": 0.0212, "step": 36860 }, { "grad_norm": 0.23511961102485657, "learning_rate": 7.470971195589881e-05, "loss": 0.022, "step": 36870 }, { "grad_norm": 0.2160911113023758, "learning_rate": 7.46953361401671e-05, "loss": 0.0257, "step": 36880 }, { "grad_norm": 0.23299458622932434, "learning_rate": 7.468095762379055e-05, "loss": 0.0266, "step": 36890 }, { "grad_norm": 0.16950292885303497, "learning_rate": 7.466657640834158e-05, "loss": 0.0244, "step": 36900 }, { "grad_norm": 0.2152286320924759, "learning_rate": 7.46521924953929e-05, "loss": 0.0234, "step": 36910 }, { "grad_norm": 0.2378849983215332, "learning_rate": 7.463780588651752e-05, "loss": 0.0279, "step": 36920 }, { "grad_norm": 0.2755930423736572, "learning_rate": 7.462341658328873e-05, "loss": 0.0244, "step": 36930 }, { "grad_norm": 0.22335605323314667, "learning_rate": 7.460902458728012e-05, "loss": 0.0235, "step": 36940 }, { "grad_norm": 0.2580615282058716, "learning_rate": 7.459462990006558e-05, "loss": 0.0242, "step": 36950 }, { "grad_norm": 0.2530224323272705, "learning_rate": 7.458023252321929e-05, "loss": 0.0265, "step": 36960 }, { "grad_norm": 0.20984628796577454, "learning_rate": 7.456583245831573e-05, "loss": 0.0255, "step": 36970 }, { "grad_norm": 0.2498079240322113, "learning_rate": 7.455142970692968e-05, "loss": 0.0276, "step": 36980 }, { "grad_norm": 0.21650445461273193, "learning_rate": 7.453702427063615e-05, "loss": 0.0274, "step": 36990 }, { "grad_norm": 0.25496813654899597, "learning_rate": 7.452261615101057e-05, "loss": 0.0283, "step": 37000 }, { "grad_norm": 0.25061383843421936, "learning_rate": 7.450820534962856e-05, "loss": 0.0261, "step": 37010 }, { "grad_norm": 0.25624027848243713, "learning_rate": 7.449379186806604e-05, "loss": 0.0289, "step": 37020 }, { "grad_norm": 0.18642188608646393, "learning_rate": 7.44793757078993e-05, "loss": 0.0221, "step": 37030 }, { "grad_norm": 0.25915318727493286, "learning_rate": 7.44649568707048e-05, "loss": 0.0241, "step": 37040 }, { "grad_norm": 0.25773489475250244, "learning_rate": 7.445053535805942e-05, "loss": 0.0213, "step": 37050 }, { "grad_norm": 0.22832971811294556, "learning_rate": 7.443611117154024e-05, "loss": 0.0235, "step": 37060 }, { "grad_norm": 0.24639996886253357, "learning_rate": 7.442168431272469e-05, "loss": 0.0276, "step": 37070 }, { "grad_norm": 0.2486051470041275, "learning_rate": 7.440725478319044e-05, "loss": 0.0237, "step": 37080 }, { "grad_norm": 0.21204283833503723, "learning_rate": 7.439282258451553e-05, "loss": 0.0235, "step": 37090 }, { "grad_norm": 0.1693676859140396, "learning_rate": 7.43783877182782e-05, "loss": 0.0204, "step": 37100 }, { "grad_norm": 0.2429494559764862, "learning_rate": 7.436395018605707e-05, "loss": 0.0238, "step": 37110 }, { "grad_norm": 0.188181072473526, "learning_rate": 7.434950998943094e-05, "loss": 0.0194, "step": 37120 }, { "grad_norm": 0.24376901984214783, "learning_rate": 7.433506712997902e-05, "loss": 0.0215, "step": 37130 }, { "grad_norm": 0.22866854071617126, "learning_rate": 7.432062160928077e-05, "loss": 0.0258, "step": 37140 }, { "grad_norm": 0.22972632944583893, "learning_rate": 7.430617342891588e-05, "loss": 0.0239, "step": 37150 }, { "grad_norm": 0.2625246047973633, "learning_rate": 7.429172259046442e-05, "loss": 0.0255, "step": 37160 }, { "grad_norm": 0.2455238699913025, "learning_rate": 7.42772690955067e-05, "loss": 0.0213, "step": 37170 }, { "grad_norm": 0.23450468480587006, "learning_rate": 7.426281294562332e-05, "loss": 0.023, "step": 37180 }, { "grad_norm": 0.19888834655284882, "learning_rate": 7.424835414239519e-05, "loss": 0.0238, "step": 37190 }, { "grad_norm": 0.2761878967285156, "learning_rate": 7.423389268740352e-05, "loss": 0.023, "step": 37200 }, { "grad_norm": 0.21446478366851807, "learning_rate": 7.42194285822298e-05, "loss": 0.0241, "step": 37210 }, { "grad_norm": 0.3221561312675476, "learning_rate": 7.420496182845576e-05, "loss": 0.0249, "step": 37220 }, { "grad_norm": 0.24368621408939362, "learning_rate": 7.419049242766348e-05, "loss": 0.0315, "step": 37230 }, { "grad_norm": 0.2143125981092453, "learning_rate": 7.417602038143533e-05, "loss": 0.0283, "step": 37240 }, { "grad_norm": 0.30796343088150024, "learning_rate": 7.416154569135393e-05, "loss": 0.0278, "step": 37250 }, { "grad_norm": 0.22285233438014984, "learning_rate": 7.414706835900222e-05, "loss": 0.0256, "step": 37260 }, { "grad_norm": 0.2273033708333969, "learning_rate": 7.413258838596345e-05, "loss": 0.0235, "step": 37270 }, { "grad_norm": 0.2116408348083496, "learning_rate": 7.411810577382106e-05, "loss": 0.025, "step": 37280 }, { "grad_norm": 0.22733311355113983, "learning_rate": 7.410362052415889e-05, "loss": 0.024, "step": 37290 }, { "grad_norm": 0.25943976640701294, "learning_rate": 7.408913263856102e-05, "loss": 0.0261, "step": 37300 }, { "grad_norm": 0.25081369280815125, "learning_rate": 7.407464211861183e-05, "loss": 0.0282, "step": 37310 }, { "grad_norm": 0.20027318596839905, "learning_rate": 7.406014896589597e-05, "loss": 0.0268, "step": 37320 }, { "grad_norm": 0.255777508020401, "learning_rate": 7.404565318199838e-05, "loss": 0.0226, "step": 37330 }, { "grad_norm": 0.252140074968338, "learning_rate": 7.403115476850432e-05, "loss": 0.0252, "step": 37340 }, { "grad_norm": 0.19227509200572968, "learning_rate": 7.401665372699932e-05, "loss": 0.0248, "step": 37350 }, { "grad_norm": 0.22050164639949799, "learning_rate": 7.400215005906916e-05, "loss": 0.0197, "step": 37360 }, { "grad_norm": 0.21665829420089722, "learning_rate": 7.398764376629996e-05, "loss": 0.0226, "step": 37370 }, { "grad_norm": 0.31456458568573, "learning_rate": 7.39731348502781e-05, "loss": 0.0273, "step": 37380 }, { "grad_norm": 0.2463359385728836, "learning_rate": 7.395862331259028e-05, "loss": 0.0236, "step": 37390 }, { "grad_norm": 0.21306751668453217, "learning_rate": 7.394410915482341e-05, "loss": 0.0243, "step": 37400 }, { "grad_norm": 0.18903934955596924, "learning_rate": 7.392959237856476e-05, "loss": 0.0226, "step": 37410 }, { "grad_norm": 0.23906449973583221, "learning_rate": 7.391507298540187e-05, "loss": 0.0235, "step": 37420 }, { "grad_norm": 0.2776004374027252, "learning_rate": 7.390055097692257e-05, "loss": 0.0258, "step": 37430 }, { "grad_norm": 0.19469736516475677, "learning_rate": 7.388602635471494e-05, "loss": 0.0226, "step": 37440 }, { "grad_norm": 0.220298171043396, "learning_rate": 7.387149912036738e-05, "loss": 0.026, "step": 37450 }, { "grad_norm": 0.2245488166809082, "learning_rate": 7.385696927546854e-05, "loss": 0.028, "step": 37460 }, { "grad_norm": 0.20255741477012634, "learning_rate": 7.384243682160743e-05, "loss": 0.0232, "step": 37470 }, { "grad_norm": 0.20744799077510834, "learning_rate": 7.382790176037329e-05, "loss": 0.0237, "step": 37480 }, { "grad_norm": 0.17758919298648834, "learning_rate": 7.381336409335562e-05, "loss": 0.0269, "step": 37490 }, { "grad_norm": 0.23010605573654175, "learning_rate": 7.379882382214426e-05, "loss": 0.0254, "step": 37500 }, { "grad_norm": 0.2611079514026642, "learning_rate": 7.378428094832931e-05, "loss": 0.0237, "step": 37510 }, { "grad_norm": 0.22502422332763672, "learning_rate": 7.376973547350114e-05, "loss": 0.0237, "step": 37520 }, { "grad_norm": 0.2687758207321167, "learning_rate": 7.375518739925046e-05, "loss": 0.0256, "step": 37530 }, { "grad_norm": 0.20325982570648193, "learning_rate": 7.374063672716818e-05, "loss": 0.0296, "step": 37540 }, { "grad_norm": 0.22377441823482513, "learning_rate": 7.372608345884558e-05, "loss": 0.0244, "step": 37550 }, { "grad_norm": 0.21647292375564575, "learning_rate": 7.371152759587415e-05, "loss": 0.0235, "step": 37560 }, { "grad_norm": 0.23960833251476288, "learning_rate": 7.369696913984572e-05, "loss": 0.0228, "step": 37570 }, { "grad_norm": 0.2472827434539795, "learning_rate": 7.368240809235237e-05, "loss": 0.0234, "step": 37580 }, { "grad_norm": 0.22442784905433655, "learning_rate": 7.366784445498649e-05, "loss": 0.0238, "step": 37590 }, { "grad_norm": 0.21737739443778992, "learning_rate": 7.36532782293407e-05, "loss": 0.0278, "step": 37600 }, { "grad_norm": 0.18513742089271545, "learning_rate": 7.363870941700797e-05, "loss": 0.0221, "step": 37610 }, { "grad_norm": 0.15291766822338104, "learning_rate": 7.362413801958152e-05, "loss": 0.0199, "step": 37620 }, { "grad_norm": 0.21229472756385803, "learning_rate": 7.360956403865486e-05, "loss": 0.0197, "step": 37630 }, { "grad_norm": 0.21743163466453552, "learning_rate": 7.359498747582177e-05, "loss": 0.0247, "step": 37640 }, { "grad_norm": 0.3590874671936035, "learning_rate": 7.358040833267634e-05, "loss": 0.0252, "step": 37650 }, { "grad_norm": 0.2375161200761795, "learning_rate": 7.356582661081288e-05, "loss": 0.0258, "step": 37660 }, { "grad_norm": 0.2500118017196655, "learning_rate": 7.355124231182607e-05, "loss": 0.0247, "step": 37670 }, { "grad_norm": 0.2669360041618347, "learning_rate": 7.353665543731083e-05, "loss": 0.0238, "step": 37680 }, { "grad_norm": 0.24746333062648773, "learning_rate": 7.352206598886229e-05, "loss": 0.0244, "step": 37690 }, { "grad_norm": 0.3007536828517914, "learning_rate": 7.350747396807601e-05, "loss": 0.0231, "step": 37700 }, { "grad_norm": 0.23289768397808075, "learning_rate": 7.349287937654772e-05, "loss": 0.0209, "step": 37710 }, { "grad_norm": 0.29877492785453796, "learning_rate": 7.347828221587345e-05, "loss": 0.0243, "step": 37720 }, { "grad_norm": 0.26083236932754517, "learning_rate": 7.346368248764955e-05, "loss": 0.0216, "step": 37730 }, { "grad_norm": 0.26743805408477783, "learning_rate": 7.34490801934726e-05, "loss": 0.0237, "step": 37740 }, { "grad_norm": 0.15560835599899292, "learning_rate": 7.343447533493947e-05, "loss": 0.0227, "step": 37750 }, { "grad_norm": 0.17401723563671112, "learning_rate": 7.341986791364739e-05, "loss": 0.02, "step": 37760 }, { "grad_norm": 0.2764962911605835, "learning_rate": 7.340525793119374e-05, "loss": 0.0258, "step": 37770 }, { "grad_norm": 0.21659763157367706, "learning_rate": 7.33906453891763e-05, "loss": 0.0218, "step": 37780 }, { "grad_norm": 0.2688233256340027, "learning_rate": 7.337603028919303e-05, "loss": 0.0231, "step": 37790 }, { "grad_norm": 0.20517684519290924, "learning_rate": 7.336141263284225e-05, "loss": 0.0236, "step": 37800 }, { "grad_norm": 0.20624111592769623, "learning_rate": 7.334679242172249e-05, "loss": 0.0209, "step": 37810 }, { "grad_norm": 0.21534709632396698, "learning_rate": 7.33321696574326e-05, "loss": 0.0233, "step": 37820 }, { "grad_norm": 0.31217530369758606, "learning_rate": 7.331754434157174e-05, "loss": 0.0238, "step": 37830 }, { "grad_norm": 0.19661042094230652, "learning_rate": 7.330291647573928e-05, "loss": 0.0225, "step": 37840 }, { "grad_norm": 0.21446838974952698, "learning_rate": 7.32882860615349e-05, "loss": 0.0231, "step": 37850 }, { "grad_norm": 0.20223236083984375, "learning_rate": 7.327365310055858e-05, "loss": 0.0224, "step": 37860 }, { "grad_norm": 0.19140899181365967, "learning_rate": 7.325901759441055e-05, "loss": 0.0217, "step": 37870 }, { "grad_norm": 0.22424574196338654, "learning_rate": 7.324437954469132e-05, "loss": 0.0307, "step": 37880 }, { "grad_norm": 0.2278982400894165, "learning_rate": 7.32297389530017e-05, "loss": 0.0259, "step": 37890 }, { "grad_norm": 0.42946702241897583, "learning_rate": 7.321509582094274e-05, "loss": 0.0214, "step": 37900 }, { "grad_norm": 0.24652019143104553, "learning_rate": 7.320045015011582e-05, "loss": 0.022, "step": 37910 }, { "grad_norm": 0.22915507853031158, "learning_rate": 7.318580194212253e-05, "loss": 0.0214, "step": 37920 }, { "grad_norm": 0.23010298609733582, "learning_rate": 7.317115119856483e-05, "loss": 0.0224, "step": 37930 }, { "grad_norm": 0.23145048320293427, "learning_rate": 7.315649792104485e-05, "loss": 0.0295, "step": 37940 }, { "grad_norm": 0.22788579761981964, "learning_rate": 7.314184211116507e-05, "loss": 0.0237, "step": 37950 }, { "grad_norm": 0.20787401497364044, "learning_rate": 7.312718377052823e-05, "loss": 0.0237, "step": 37960 }, { "grad_norm": 0.19828760623931885, "learning_rate": 7.311252290073737e-05, "loss": 0.0214, "step": 37970 }, { "grad_norm": 0.22084420919418335, "learning_rate": 7.309785950339572e-05, "loss": 0.0246, "step": 37980 }, { "grad_norm": 0.20864349603652954, "learning_rate": 7.308319358010692e-05, "loss": 0.0211, "step": 37990 }, { "grad_norm": 0.21703389286994934, "learning_rate": 7.306852513247474e-05, "loss": 0.02, "step": 38000 }, { "grad_norm": 0.2499249428510666, "learning_rate": 7.305385416210336e-05, "loss": 0.0245, "step": 38010 }, { "grad_norm": 0.18034932017326355, "learning_rate": 7.303918067059711e-05, "loss": 0.0207, "step": 38020 }, { "grad_norm": 0.22579607367515564, "learning_rate": 7.302450465956075e-05, "loss": 0.0294, "step": 38030 }, { "grad_norm": 0.21723899245262146, "learning_rate": 7.300982613059914e-05, "loss": 0.0212, "step": 38040 }, { "grad_norm": 0.24222852289676666, "learning_rate": 7.299514508531757e-05, "loss": 0.023, "step": 38050 }, { "grad_norm": 0.2364516705274582, "learning_rate": 7.298046152532148e-05, "loss": 0.0216, "step": 38060 }, { "grad_norm": 0.22244420647621155, "learning_rate": 7.296577545221668e-05, "loss": 0.0207, "step": 38070 }, { "grad_norm": 0.4064829647541046, "learning_rate": 7.29510868676092e-05, "loss": 0.0232, "step": 38080 }, { "grad_norm": 0.1888543963432312, "learning_rate": 7.293639577310538e-05, "loss": 0.0258, "step": 38090 }, { "grad_norm": 0.18283085525035858, "learning_rate": 7.292170217031179e-05, "loss": 0.025, "step": 38100 }, { "grad_norm": 0.1691489964723587, "learning_rate": 7.290700606083532e-05, "loss": 0.0233, "step": 38110 }, { "grad_norm": 0.2481217086315155, "learning_rate": 7.28923074462831e-05, "loss": 0.0203, "step": 38120 }, { "grad_norm": 0.2128852754831314, "learning_rate": 7.287760632826255e-05, "loss": 0.0247, "step": 38130 }, { "grad_norm": 0.3129708766937256, "learning_rate": 7.286290270838139e-05, "loss": 0.0308, "step": 38140 }, { "grad_norm": 0.22751682996749878, "learning_rate": 7.284819658824756e-05, "loss": 0.0262, "step": 38150 }, { "grad_norm": 0.2689087688922882, "learning_rate": 7.283348796946929e-05, "loss": 0.0233, "step": 38160 }, { "grad_norm": 0.24104875326156616, "learning_rate": 7.28187768536551e-05, "loss": 0.0237, "step": 38170 }, { "grad_norm": 0.2094767689704895, "learning_rate": 7.280406324241379e-05, "loss": 0.0226, "step": 38180 }, { "grad_norm": 0.2567066550254822, "learning_rate": 7.278934713735441e-05, "loss": 0.0239, "step": 38190 }, { "grad_norm": 0.20096275210380554, "learning_rate": 7.277462854008629e-05, "loss": 0.0248, "step": 38200 }, { "grad_norm": 0.263485312461853, "learning_rate": 7.275990745221904e-05, "loss": 0.024, "step": 38210 }, { "grad_norm": 0.20207618176937103, "learning_rate": 7.274518387536252e-05, "loss": 0.0231, "step": 38220 }, { "grad_norm": 0.2788422405719757, "learning_rate": 7.273045781112689e-05, "loss": 0.0219, "step": 38230 }, { "grad_norm": 0.2540900409221649, "learning_rate": 7.271572926112257e-05, "loss": 0.0242, "step": 38240 }, { "grad_norm": 0.21388153731822968, "learning_rate": 7.270099822696024e-05, "loss": 0.0198, "step": 38250 }, { "grad_norm": 0.21827684342861176, "learning_rate": 7.26862647102509e-05, "loss": 0.0222, "step": 38260 }, { "grad_norm": 0.33232635259628296, "learning_rate": 7.267152871260573e-05, "loss": 0.0234, "step": 38270 }, { "grad_norm": 0.20412984490394592, "learning_rate": 7.265679023563629e-05, "loss": 0.0254, "step": 38280 }, { "grad_norm": 0.18699346482753754, "learning_rate": 7.264204928095433e-05, "loss": 0.024, "step": 38290 }, { "grad_norm": 0.242629274725914, "learning_rate": 7.262730585017188e-05, "loss": 0.0236, "step": 38300 }, { "grad_norm": 0.20667780935764313, "learning_rate": 7.261255994490131e-05, "loss": 0.0215, "step": 38310 }, { "grad_norm": 0.2990456223487854, "learning_rate": 7.259781156675517e-05, "loss": 0.0237, "step": 38320 }, { "grad_norm": 0.23224113881587982, "learning_rate": 7.258306071734633e-05, "loss": 0.025, "step": 38330 }, { "grad_norm": 0.23455533385276794, "learning_rate": 7.256830739828792e-05, "loss": 0.0212, "step": 38340 }, { "grad_norm": 0.2914440631866455, "learning_rate": 7.255355161119336e-05, "loss": 0.0231, "step": 38350 }, { "grad_norm": 0.20829443633556366, "learning_rate": 7.253879335767628e-05, "loss": 0.0227, "step": 38360 }, { "grad_norm": 0.23794810473918915, "learning_rate": 7.252403263935067e-05, "loss": 0.0237, "step": 38370 }, { "grad_norm": 0.17913031578063965, "learning_rate": 7.25092694578307e-05, "loss": 0.0207, "step": 38380 }, { "grad_norm": 0.2300601452589035, "learning_rate": 7.249450381473087e-05, "loss": 0.0248, "step": 38390 }, { "grad_norm": 0.2077813446521759, "learning_rate": 7.247973571166593e-05, "loss": 0.0255, "step": 38400 }, { "grad_norm": 0.16152675449848175, "learning_rate": 7.246496515025089e-05, "loss": 0.0219, "step": 38410 }, { "grad_norm": 0.22223512828350067, "learning_rate": 7.245019213210101e-05, "loss": 0.0209, "step": 38420 }, { "grad_norm": 0.19694703817367554, "learning_rate": 7.24354166588319e-05, "loss": 0.024, "step": 38430 }, { "grad_norm": 0.20411841571331024, "learning_rate": 7.242063873205935e-05, "loss": 0.0234, "step": 38440 }, { "grad_norm": 0.2994633913040161, "learning_rate": 7.240585835339946e-05, "loss": 0.0263, "step": 38450 }, { "grad_norm": 0.23826153576374054, "learning_rate": 7.239107552446857e-05, "loss": 0.025, "step": 38460 }, { "grad_norm": 0.23936684429645538, "learning_rate": 7.237629024688334e-05, "loss": 0.0252, "step": 38470 }, { "grad_norm": 0.29823485016822815, "learning_rate": 7.236150252226064e-05, "loss": 0.025, "step": 38480 }, { "grad_norm": 0.1687944084405899, "learning_rate": 7.234671235221765e-05, "loss": 0.0204, "step": 38490 }, { "grad_norm": 0.1731472760438919, "learning_rate": 7.233191973837179e-05, "loss": 0.0213, "step": 38500 }, { "grad_norm": 0.22234977781772614, "learning_rate": 7.231712468234078e-05, "loss": 0.0249, "step": 38510 }, { "grad_norm": 0.3281284272670746, "learning_rate": 7.230232718574254e-05, "loss": 0.0222, "step": 38520 }, { "grad_norm": 0.1910300850868225, "learning_rate": 7.228752725019535e-05, "loss": 0.0198, "step": 38530 }, { "grad_norm": 0.21919956803321838, "learning_rate": 7.227272487731769e-05, "loss": 0.0258, "step": 38540 }, { "grad_norm": 0.20522357523441315, "learning_rate": 7.225792006872831e-05, "loss": 0.0185, "step": 38550 }, { "grad_norm": 0.20818999409675598, "learning_rate": 7.224311282604628e-05, "loss": 0.0268, "step": 38560 }, { "grad_norm": 0.25051942467689514, "learning_rate": 7.222830315089085e-05, "loss": 0.0205, "step": 38570 }, { "grad_norm": 0.23372073471546173, "learning_rate": 7.22134910448816e-05, "loss": 0.0228, "step": 38580 }, { "grad_norm": 0.248141348361969, "learning_rate": 7.219867650963839e-05, "loss": 0.0208, "step": 38590 }, { "grad_norm": 0.3381713926792145, "learning_rate": 7.21838595467813e-05, "loss": 0.0246, "step": 38600 }, { "grad_norm": 0.24297407269477844, "learning_rate": 7.216904015793067e-05, "loss": 0.0209, "step": 38610 }, { "grad_norm": 0.23237252235412598, "learning_rate": 7.215421834470713e-05, "loss": 0.0206, "step": 38620 }, { "grad_norm": 0.21200375258922577, "learning_rate": 7.213939410873158e-05, "loss": 0.026, "step": 38630 }, { "grad_norm": 0.2063789814710617, "learning_rate": 7.212456745162519e-05, "loss": 0.0221, "step": 38640 }, { "grad_norm": 0.19652701914310455, "learning_rate": 7.210973837500937e-05, "loss": 0.0217, "step": 38650 }, { "grad_norm": 0.25362691283226013, "learning_rate": 7.209490688050578e-05, "loss": 0.0208, "step": 38660 }, { "grad_norm": 0.25415894389152527, "learning_rate": 7.208007296973641e-05, "loss": 0.0267, "step": 38670 }, { "grad_norm": 0.219833105802536, "learning_rate": 7.206523664432345e-05, "loss": 0.0235, "step": 38680 }, { "grad_norm": 0.20715592801570892, "learning_rate": 7.205039790588939e-05, "loss": 0.0233, "step": 38690 }, { "grad_norm": 0.2085563987493515, "learning_rate": 7.203555675605697e-05, "loss": 0.022, "step": 38700 }, { "grad_norm": 0.245001420378685, "learning_rate": 7.202071319644917e-05, "loss": 0.0194, "step": 38710 }, { "grad_norm": 0.2699524164199829, "learning_rate": 7.200586722868932e-05, "loss": 0.0213, "step": 38720 }, { "grad_norm": 0.22072148323059082, "learning_rate": 7.199101885440088e-05, "loss": 0.0226, "step": 38730 }, { "grad_norm": 0.25191113352775574, "learning_rate": 7.197616807520771e-05, "loss": 0.0181, "step": 38740 }, { "grad_norm": 0.18590010702610016, "learning_rate": 7.196131489273381e-05, "loss": 0.0212, "step": 38750 }, { "grad_norm": 0.19994628429412842, "learning_rate": 7.194645930860355e-05, "loss": 0.0214, "step": 38760 }, { "grad_norm": 0.20479092001914978, "learning_rate": 7.19316013244415e-05, "loss": 0.019, "step": 38770 }, { "grad_norm": 0.25044187903404236, "learning_rate": 7.191674094187248e-05, "loss": 0.0208, "step": 38780 }, { "grad_norm": 0.27190709114074707, "learning_rate": 7.190187816252165e-05, "loss": 0.0221, "step": 38790 }, { "grad_norm": 0.2265734225511551, "learning_rate": 7.188701298801435e-05, "loss": 0.0238, "step": 38800 }, { "grad_norm": 0.25850361585617065, "learning_rate": 7.18721454199762e-05, "loss": 0.0212, "step": 38810 }, { "grad_norm": 0.17560447752475739, "learning_rate": 7.185727546003312e-05, "loss": 0.0237, "step": 38820 }, { "grad_norm": 0.277595192193985, "learning_rate": 7.184240310981126e-05, "loss": 0.0232, "step": 38830 }, { "grad_norm": 0.23122994601726532, "learning_rate": 7.182752837093704e-05, "loss": 0.0213, "step": 38840 }, { "grad_norm": 0.2401050180196762, "learning_rate": 7.181265124503711e-05, "loss": 0.0264, "step": 38850 }, { "grad_norm": 0.15202850103378296, "learning_rate": 7.179777173373847e-05, "loss": 0.0249, "step": 38860 }, { "grad_norm": 0.21025966107845306, "learning_rate": 7.178288983866826e-05, "loss": 0.0228, "step": 38870 }, { "grad_norm": 0.2327083945274353, "learning_rate": 7.176800556145397e-05, "loss": 0.0232, "step": 38880 }, { "grad_norm": 0.23030351102352142, "learning_rate": 7.175311890372334e-05, "loss": 0.0237, "step": 38890 }, { "grad_norm": 0.20068439841270447, "learning_rate": 7.17382298671043e-05, "loss": 0.0196, "step": 38900 }, { "grad_norm": 0.20588664710521698, "learning_rate": 7.172333845322515e-05, "loss": 0.0231, "step": 38910 }, { "grad_norm": 0.1940145492553711, "learning_rate": 7.170844466371436e-05, "loss": 0.0205, "step": 38920 }, { "grad_norm": 0.18308676779270172, "learning_rate": 7.16935485002007e-05, "loss": 0.0263, "step": 38930 }, { "grad_norm": 0.21510711312294006, "learning_rate": 7.167864996431319e-05, "loss": 0.0211, "step": 38940 }, { "grad_norm": 0.28321903944015503, "learning_rate": 7.166374905768111e-05, "loss": 0.023, "step": 38950 }, { "grad_norm": 0.21203704178333282, "learning_rate": 7.164884578193401e-05, "loss": 0.0218, "step": 38960 }, { "grad_norm": 0.262617290019989, "learning_rate": 7.163394013870168e-05, "loss": 0.025, "step": 38970 }, { "grad_norm": 0.2103716880083084, "learning_rate": 7.161903212961419e-05, "loss": 0.0283, "step": 38980 }, { "grad_norm": 0.26738548278808594, "learning_rate": 7.160412175630183e-05, "loss": 0.0301, "step": 38990 }, { "grad_norm": 0.2189614325761795, "learning_rate": 7.158920902039521e-05, "loss": 0.0223, "step": 39000 }, { "grad_norm": 0.2624810039997101, "learning_rate": 7.157429392352514e-05, "loss": 0.0246, "step": 39010 }, { "grad_norm": 0.28908830881118774, "learning_rate": 7.155937646732275e-05, "loss": 0.0281, "step": 39020 }, { "grad_norm": 0.28146812319755554, "learning_rate": 7.154445665341933e-05, "loss": 0.0284, "step": 39030 }, { "grad_norm": 0.2778870761394501, "learning_rate": 7.152953448344654e-05, "loss": 0.0255, "step": 39040 }, { "grad_norm": 0.22189569473266602, "learning_rate": 7.151460995903624e-05, "loss": 0.0246, "step": 39050 }, { "grad_norm": 0.21592655777931213, "learning_rate": 7.149968308182052e-05, "loss": 0.0234, "step": 39060 }, { "grad_norm": 0.21752159297466278, "learning_rate": 7.14847538534318e-05, "loss": 0.0224, "step": 39070 }, { "grad_norm": 0.2888203263282776, "learning_rate": 7.14698222755027e-05, "loss": 0.0242, "step": 39080 }, { "grad_norm": 0.26918676495552063, "learning_rate": 7.14548883496661e-05, "loss": 0.0269, "step": 39090 }, { "grad_norm": 0.24603666365146637, "learning_rate": 7.143995207755517e-05, "loss": 0.0278, "step": 39100 }, { "grad_norm": 0.24142087996006012, "learning_rate": 7.142501346080333e-05, "loss": 0.0216, "step": 39110 }, { "grad_norm": 0.19662028551101685, "learning_rate": 7.141007250104421e-05, "loss": 0.0192, "step": 39120 }, { "grad_norm": 0.20917260646820068, "learning_rate": 7.139512919991176e-05, "loss": 0.0201, "step": 39130 }, { "grad_norm": 0.2252759486436844, "learning_rate": 7.138018355904014e-05, "loss": 0.0279, "step": 39140 }, { "grad_norm": 0.24430640041828156, "learning_rate": 7.13652355800638e-05, "loss": 0.0205, "step": 39150 }, { "grad_norm": 0.19887733459472656, "learning_rate": 7.135028526461743e-05, "loss": 0.022, "step": 39160 }, { "grad_norm": 0.18860028684139252, "learning_rate": 7.133533261433594e-05, "loss": 0.0243, "step": 39170 }, { "grad_norm": 0.13965578377246857, "learning_rate": 7.132037763085457e-05, "loss": 0.0221, "step": 39180 }, { "grad_norm": 0.20291607081890106, "learning_rate": 7.130542031580875e-05, "loss": 0.0243, "step": 39190 }, { "grad_norm": 0.2325291782617569, "learning_rate": 7.12904606708342e-05, "loss": 0.0224, "step": 39200 }, { "grad_norm": 0.23835740983486176, "learning_rate": 7.127549869756687e-05, "loss": 0.0265, "step": 39210 }, { "grad_norm": 0.160496324300766, "learning_rate": 7.126053439764299e-05, "loss": 0.0256, "step": 39220 }, { "grad_norm": 0.2487425059080124, "learning_rate": 7.124556777269904e-05, "loss": 0.0237, "step": 39230 }, { "grad_norm": 0.2699713110923767, "learning_rate": 7.123059882437174e-05, "loss": 0.0214, "step": 39240 }, { "grad_norm": 0.25497889518737793, "learning_rate": 7.121562755429807e-05, "loss": 0.0233, "step": 39250 }, { "grad_norm": 0.1623646765947342, "learning_rate": 7.120065396411528e-05, "loss": 0.026, "step": 39260 }, { "grad_norm": 0.18577159941196442, "learning_rate": 7.118567805546084e-05, "loss": 0.0227, "step": 39270 }, { "grad_norm": 0.1655995398759842, "learning_rate": 7.117069982997248e-05, "loss": 0.0205, "step": 39280 }, { "grad_norm": 0.2951297163963318, "learning_rate": 7.115571928928825e-05, "loss": 0.0232, "step": 39290 }, { "grad_norm": 0.19812025129795074, "learning_rate": 7.114073643504635e-05, "loss": 0.0222, "step": 39300 }, { "grad_norm": 0.24507838487625122, "learning_rate": 7.11257512688853e-05, "loss": 0.0246, "step": 39310 }, { "grad_norm": 0.2274661660194397, "learning_rate": 7.111076379244384e-05, "loss": 0.0219, "step": 39320 }, { "grad_norm": 0.3631342351436615, "learning_rate": 7.109577400736101e-05, "loss": 0.0239, "step": 39330 }, { "grad_norm": 0.2199224978685379, "learning_rate": 7.108078191527605e-05, "loss": 0.0264, "step": 39340 }, { "grad_norm": 0.2965487241744995, "learning_rate": 7.106578751782847e-05, "loss": 0.0275, "step": 39350 }, { "grad_norm": 0.24816329777240753, "learning_rate": 7.105079081665803e-05, "loss": 0.027, "step": 39360 }, { "grad_norm": 0.1976923644542694, "learning_rate": 7.103579181340476e-05, "loss": 0.0241, "step": 39370 }, { "grad_norm": 0.17495477199554443, "learning_rate": 7.102079050970893e-05, "loss": 0.0253, "step": 39380 }, { "grad_norm": 0.19080138206481934, "learning_rate": 7.100578690721105e-05, "loss": 0.0222, "step": 39390 }, { "grad_norm": 0.21689186990261078, "learning_rate": 7.09907810075519e-05, "loss": 0.0229, "step": 39400 }, { "grad_norm": 0.1996515840291977, "learning_rate": 7.097577281237249e-05, "loss": 0.0225, "step": 39410 }, { "grad_norm": 0.23745927214622498, "learning_rate": 7.09607623233141e-05, "loss": 0.0269, "step": 39420 }, { "grad_norm": 0.2728751301765442, "learning_rate": 7.094574954201827e-05, "loss": 0.0234, "step": 39430 }, { "grad_norm": 0.19723477959632874, "learning_rate": 7.093073447012675e-05, "loss": 0.0211, "step": 39440 }, { "grad_norm": 0.2372138351202011, "learning_rate": 7.09157171092816e-05, "loss": 0.0221, "step": 39450 }, { "grad_norm": 0.27316758036613464, "learning_rate": 7.090069746112504e-05, "loss": 0.0217, "step": 39460 }, { "grad_norm": 0.21935483813285828, "learning_rate": 7.088567552729965e-05, "loss": 0.0213, "step": 39470 }, { "grad_norm": 0.2540722191333771, "learning_rate": 7.087065130944818e-05, "loss": 0.0198, "step": 39480 }, { "grad_norm": 0.22775909304618835, "learning_rate": 7.085562480921366e-05, "loss": 0.0227, "step": 39490 }, { "grad_norm": 0.18530060350894928, "learning_rate": 7.084059602823937e-05, "loss": 0.0233, "step": 39500 }, { "grad_norm": 0.19243265688419342, "learning_rate": 7.082556496816882e-05, "loss": 0.021, "step": 39510 }, { "grad_norm": 0.18461401760578156, "learning_rate": 7.081053163064582e-05, "loss": 0.0197, "step": 39520 }, { "grad_norm": 0.193012073636055, "learning_rate": 7.079549601731434e-05, "loss": 0.0225, "step": 39530 }, { "grad_norm": 0.3005278408527374, "learning_rate": 7.07804581298187e-05, "loss": 0.0281, "step": 39540 }, { "grad_norm": 0.2584727108478546, "learning_rate": 7.07654179698034e-05, "loss": 0.0221, "step": 39550 }, { "grad_norm": 0.27498164772987366, "learning_rate": 7.075037553891321e-05, "loss": 0.0264, "step": 39560 }, { "grad_norm": 0.2421235740184784, "learning_rate": 7.073533083879315e-05, "loss": 0.0246, "step": 39570 }, { "grad_norm": 0.20654024183750153, "learning_rate": 7.072028387108849e-05, "loss": 0.022, "step": 39580 }, { "grad_norm": 0.1467386782169342, "learning_rate": 7.070523463744472e-05, "loss": 0.0216, "step": 39590 }, { "grad_norm": 0.28480032086372375, "learning_rate": 7.069018313950763e-05, "loss": 0.0252, "step": 39600 }, { "grad_norm": 0.27305784821510315, "learning_rate": 7.067512937892322e-05, "loss": 0.0232, "step": 39610 }, { "grad_norm": 0.2245340496301651, "learning_rate": 7.066007335733775e-05, "loss": 0.0189, "step": 39620 }, { "grad_norm": 0.2319847196340561, "learning_rate": 7.064501507639772e-05, "loss": 0.0203, "step": 39630 }, { "grad_norm": 0.2655825912952423, "learning_rate": 7.062995453774987e-05, "loss": 0.0213, "step": 39640 }, { "grad_norm": 0.2176571637392044, "learning_rate": 7.061489174304121e-05, "loss": 0.0203, "step": 39650 }, { "grad_norm": 0.2032836377620697, "learning_rate": 7.0599826693919e-05, "loss": 0.02, "step": 39660 }, { "grad_norm": 0.26587337255477905, "learning_rate": 7.05847593920307e-05, "loss": 0.0202, "step": 39670 }, { "grad_norm": 0.18653692305088043, "learning_rate": 7.056968983902406e-05, "loss": 0.0221, "step": 39680 }, { "grad_norm": 0.17798788845539093, "learning_rate": 7.055461803654706e-05, "loss": 0.0227, "step": 39690 }, { "grad_norm": 0.18058712780475616, "learning_rate": 7.053954398624794e-05, "loss": 0.0207, "step": 39700 }, { "grad_norm": 0.28135761618614197, "learning_rate": 7.052446768977518e-05, "loss": 0.0273, "step": 39710 }, { "grad_norm": 0.1819266378879547, "learning_rate": 7.050938914877748e-05, "loss": 0.0191, "step": 39720 }, { "grad_norm": 0.2659221291542053, "learning_rate": 7.049430836490382e-05, "loss": 0.0242, "step": 39730 }, { "grad_norm": 0.28556084632873535, "learning_rate": 7.047922533980342e-05, "loss": 0.0259, "step": 39740 }, { "grad_norm": 0.20488448441028595, "learning_rate": 7.046414007512571e-05, "loss": 0.0202, "step": 39750 }, { "grad_norm": 0.20186462998390198, "learning_rate": 7.044905257252042e-05, "loss": 0.0233, "step": 39760 }, { "grad_norm": 0.20323966443538666, "learning_rate": 7.04339628336375e-05, "loss": 0.0265, "step": 39770 }, { "grad_norm": 0.2682804763317108, "learning_rate": 7.041887086012711e-05, "loss": 0.022, "step": 39780 }, { "grad_norm": 0.17626066505908966, "learning_rate": 7.040377665363969e-05, "loss": 0.0184, "step": 39790 }, { "grad_norm": 0.25423693656921387, "learning_rate": 7.038868021582594e-05, "loss": 0.0206, "step": 39800 }, { "grad_norm": 0.1952810287475586, "learning_rate": 7.037358154833679e-05, "loss": 0.0219, "step": 39810 }, { "grad_norm": 0.21982580423355103, "learning_rate": 7.035848065282339e-05, "loss": 0.0224, "step": 39820 }, { "grad_norm": 0.1851784884929657, "learning_rate": 7.034337753093714e-05, "loss": 0.0214, "step": 39830 }, { "grad_norm": 0.19189386069774628, "learning_rate": 7.032827218432972e-05, "loss": 0.0192, "step": 39840 }, { "grad_norm": 0.21595044434070587, "learning_rate": 7.031316461465302e-05, "loss": 0.0191, "step": 39850 }, { "grad_norm": 0.1905488520860672, "learning_rate": 7.029805482355915e-05, "loss": 0.0185, "step": 39860 }, { "grad_norm": 0.16600969433784485, "learning_rate": 7.028294281270055e-05, "loss": 0.0245, "step": 39870 }, { "grad_norm": 0.22089959681034088, "learning_rate": 7.026782858372982e-05, "loss": 0.0184, "step": 39880 }, { "grad_norm": 0.23269023001194, "learning_rate": 7.025271213829982e-05, "loss": 0.0229, "step": 39890 }, { "grad_norm": 0.18168973922729492, "learning_rate": 7.023759347806366e-05, "loss": 0.0207, "step": 39900 }, { "grad_norm": 0.19474613666534424, "learning_rate": 7.022247260467472e-05, "loss": 0.0281, "step": 39910 }, { "grad_norm": 0.25602126121520996, "learning_rate": 7.020734951978658e-05, "loss": 0.0241, "step": 39920 }, { "grad_norm": 0.20638105273246765, "learning_rate": 7.019222422505307e-05, "loss": 0.0223, "step": 39930 }, { "grad_norm": 0.2182944118976593, "learning_rate": 7.017709672212827e-05, "loss": 0.0209, "step": 39940 }, { "grad_norm": 0.203403040766716, "learning_rate": 7.016196701266652e-05, "loss": 0.0208, "step": 39950 }, { "grad_norm": 0.23524977266788483, "learning_rate": 7.014683509832238e-05, "loss": 0.023, "step": 39960 }, { "grad_norm": 0.19488458335399628, "learning_rate": 7.013170098075063e-05, "loss": 0.0232, "step": 39970 }, { "grad_norm": 0.17351040244102478, "learning_rate": 7.011656466160632e-05, "loss": 0.0249, "step": 39980 }, { "grad_norm": 0.23655572533607483, "learning_rate": 7.010142614254475e-05, "loss": 0.0214, "step": 39990 }, { "grad_norm": 0.2565668225288391, "learning_rate": 7.008628542522147e-05, "loss": 0.0273, "step": 40000 }, { "grad_norm": 0.182309091091156, "learning_rate": 7.007114251129218e-05, "loss": 0.0231, "step": 40010 }, { "grad_norm": 0.2415405511856079, "learning_rate": 7.005599740241293e-05, "loss": 0.025, "step": 40020 }, { "grad_norm": 0.26331380009651184, "learning_rate": 7.004085010023996e-05, "loss": 0.0234, "step": 40030 }, { "grad_norm": 0.2317269742488861, "learning_rate": 7.002570060642976e-05, "loss": 0.0263, "step": 40040 }, { "grad_norm": 0.2215440720319748, "learning_rate": 7.001054892263903e-05, "loss": 0.025, "step": 40050 }, { "grad_norm": 0.22340746223926544, "learning_rate": 6.999539505052477e-05, "loss": 0.0289, "step": 40060 }, { "grad_norm": 0.2002769261598587, "learning_rate": 6.998023899174418e-05, "loss": 0.0266, "step": 40070 }, { "grad_norm": 0.1944398432970047, "learning_rate": 6.996508074795467e-05, "loss": 0.0254, "step": 40080 }, { "grad_norm": 0.187722310423851, "learning_rate": 6.994992032081396e-05, "loss": 0.0216, "step": 40090 }, { "grad_norm": 0.2131974995136261, "learning_rate": 6.993475771197995e-05, "loss": 0.0243, "step": 40100 }, { "grad_norm": 0.21034054458141327, "learning_rate": 6.991959292311082e-05, "loss": 0.02, "step": 40110 }, { "grad_norm": 0.26897087693214417, "learning_rate": 6.990442595586495e-05, "loss": 0.0223, "step": 40120 }, { "grad_norm": 0.17730306088924408, "learning_rate": 6.988925681190098e-05, "loss": 0.0216, "step": 40130 }, { "grad_norm": 0.20849882066249847, "learning_rate": 6.987408549287778e-05, "loss": 0.023, "step": 40140 }, { "grad_norm": 0.21204528212547302, "learning_rate": 6.985891200045449e-05, "loss": 0.0223, "step": 40150 }, { "grad_norm": 0.20228585600852966, "learning_rate": 6.984373633629045e-05, "loss": 0.0235, "step": 40160 }, { "grad_norm": 0.28390073776245117, "learning_rate": 6.98285585020452e-05, "loss": 0.0253, "step": 40170 }, { "grad_norm": 0.23905833065509796, "learning_rate": 6.981337849937864e-05, "loss": 0.0224, "step": 40180 }, { "grad_norm": 0.2135322540998459, "learning_rate": 6.979819632995078e-05, "loss": 0.0261, "step": 40190 }, { "grad_norm": 0.23995378613471985, "learning_rate": 6.978301199542193e-05, "loss": 0.0193, "step": 40200 }, { "grad_norm": 0.28423500061035156, "learning_rate": 6.976782549745263e-05, "loss": 0.0205, "step": 40210 }, { "grad_norm": 0.23467311263084412, "learning_rate": 6.975263683770367e-05, "loss": 0.0248, "step": 40220 }, { "grad_norm": 0.2689182758331299, "learning_rate": 6.973744601783604e-05, "loss": 0.0226, "step": 40230 }, { "grad_norm": 0.2693234384059906, "learning_rate": 6.972225303951097e-05, "loss": 0.0226, "step": 40240 }, { "grad_norm": 0.2034071832895279, "learning_rate": 6.970705790438998e-05, "loss": 0.0218, "step": 40250 }, { "grad_norm": 0.2690271735191345, "learning_rate": 6.969186061413477e-05, "loss": 0.0195, "step": 40260 }, { "grad_norm": 0.23680514097213745, "learning_rate": 6.967666117040727e-05, "loss": 0.0202, "step": 40270 }, { "grad_norm": 0.24186712503433228, "learning_rate": 6.966145957486972e-05, "loss": 0.0214, "step": 40280 }, { "grad_norm": 0.2257242053747177, "learning_rate": 6.964625582918449e-05, "loss": 0.0211, "step": 40290 }, { "grad_norm": 0.30497387051582336, "learning_rate": 6.963104993501425e-05, "loss": 0.0261, "step": 40300 }, { "grad_norm": 0.2109086960554123, "learning_rate": 6.961584189402192e-05, "loss": 0.0202, "step": 40310 }, { "grad_norm": 0.19745904207229614, "learning_rate": 6.96006317078706e-05, "loss": 0.0206, "step": 40320 }, { "grad_norm": 0.22912977635860443, "learning_rate": 6.958541937822367e-05, "loss": 0.0241, "step": 40330 }, { "grad_norm": 0.1982191652059555, "learning_rate": 6.957020490674473e-05, "loss": 0.0189, "step": 40340 }, { "grad_norm": 0.2190854400396347, "learning_rate": 6.95549882950976e-05, "loss": 0.0242, "step": 40350 }, { "grad_norm": 0.22068792581558228, "learning_rate": 6.953976954494635e-05, "loss": 0.0249, "step": 40360 }, { "grad_norm": 0.2002863883972168, "learning_rate": 6.952454865795528e-05, "loss": 0.0213, "step": 40370 }, { "grad_norm": 0.21240058541297913, "learning_rate": 6.950932563578892e-05, "loss": 0.021, "step": 40380 }, { "grad_norm": 0.2953889071941376, "learning_rate": 6.949410048011206e-05, "loss": 0.0228, "step": 40390 }, { "grad_norm": 0.23277221620082855, "learning_rate": 6.947887319258966e-05, "loss": 0.0225, "step": 40400 }, { "grad_norm": 0.21849064528942108, "learning_rate": 6.946364377488696e-05, "loss": 0.0208, "step": 40410 }, { "grad_norm": 0.2009449154138565, "learning_rate": 6.944841222866947e-05, "loss": 0.0231, "step": 40420 }, { "grad_norm": 0.16854053735733032, "learning_rate": 6.943317855560284e-05, "loss": 0.0241, "step": 40430 }, { "grad_norm": 0.21604637801647186, "learning_rate": 6.941794275735302e-05, "loss": 0.0209, "step": 40440 }, { "grad_norm": 0.24603885412216187, "learning_rate": 6.94027048355862e-05, "loss": 0.023, "step": 40450 }, { "grad_norm": 0.2266319990158081, "learning_rate": 6.938746479196871e-05, "loss": 0.0204, "step": 40460 }, { "grad_norm": 0.23046515882015228, "learning_rate": 6.937222262816724e-05, "loss": 0.019, "step": 40470 }, { "grad_norm": 0.28533193469047546, "learning_rate": 6.935697834584865e-05, "loss": 0.0203, "step": 40480 }, { "grad_norm": 0.23833858966827393, "learning_rate": 6.934173194667998e-05, "loss": 0.0202, "step": 40490 }, { "grad_norm": 0.22994814813137054, "learning_rate": 6.93264834323286e-05, "loss": 0.0221, "step": 40500 }, { "grad_norm": 0.2192985564470291, "learning_rate": 6.931123280446206e-05, "loss": 0.0182, "step": 40510 }, { "grad_norm": 0.28044557571411133, "learning_rate": 6.929598006474811e-05, "loss": 0.021, "step": 40520 }, { "grad_norm": 0.18271899223327637, "learning_rate": 6.92807252148548e-05, "loss": 0.0228, "step": 40530 }, { "grad_norm": 0.2737583518028259, "learning_rate": 6.926546825645037e-05, "loss": 0.0204, "step": 40540 }, { "grad_norm": 0.18604758381843567, "learning_rate": 6.92502091912033e-05, "loss": 0.0201, "step": 40550 }, { "grad_norm": 0.21835795044898987, "learning_rate": 6.923494802078229e-05, "loss": 0.0211, "step": 40560 }, { "grad_norm": 0.21819967031478882, "learning_rate": 6.921968474685628e-05, "loss": 0.0234, "step": 40570 }, { "grad_norm": 0.16563957929611206, "learning_rate": 6.920441937109445e-05, "loss": 0.0211, "step": 40580 }, { "grad_norm": 0.1950923204421997, "learning_rate": 6.91891518951662e-05, "loss": 0.0209, "step": 40590 }, { "grad_norm": 0.2255650907754898, "learning_rate": 6.917388232074114e-05, "loss": 0.0213, "step": 40600 }, { "grad_norm": 0.21473640203475952, "learning_rate": 6.915861064948914e-05, "loss": 0.0204, "step": 40610 }, { "grad_norm": 0.22296378016471863, "learning_rate": 6.914333688308025e-05, "loss": 0.0232, "step": 40620 }, { "grad_norm": 0.2492218315601349, "learning_rate": 6.912806102318487e-05, "loss": 0.0233, "step": 40630 }, { "grad_norm": 0.2692922055721283, "learning_rate": 6.911278307147347e-05, "loss": 0.0249, "step": 40640 }, { "grad_norm": 0.295511931180954, "learning_rate": 6.909750302961684e-05, "loss": 0.0214, "step": 40650 }, { "grad_norm": 0.2041023075580597, "learning_rate": 6.9082220899286e-05, "loss": 0.0257, "step": 40660 }, { "grad_norm": 0.21111120283603668, "learning_rate": 6.906693668215219e-05, "loss": 0.0236, "step": 40670 }, { "grad_norm": 0.1907312572002411, "learning_rate": 6.905165037988683e-05, "loss": 0.0224, "step": 40680 }, { "grad_norm": 0.2463475465774536, "learning_rate": 6.903636199416164e-05, "loss": 0.0222, "step": 40690 }, { "grad_norm": 0.30946797132492065, "learning_rate": 6.902107152664851e-05, "loss": 0.0288, "step": 40700 }, { "grad_norm": 0.1882985681295395, "learning_rate": 6.90057789790196e-05, "loss": 0.021, "step": 40710 }, { "grad_norm": 0.2690953016281128, "learning_rate": 6.899048435294728e-05, "loss": 0.0202, "step": 40720 }, { "grad_norm": 0.3185952603816986, "learning_rate": 6.897518765010415e-05, "loss": 0.0255, "step": 40730 }, { "grad_norm": 0.2033916711807251, "learning_rate": 6.895988887216302e-05, "loss": 0.022, "step": 40740 }, { "grad_norm": 0.15589061379432678, "learning_rate": 6.894458802079694e-05, "loss": 0.0218, "step": 40750 }, { "grad_norm": 0.1974983662366867, "learning_rate": 6.892928509767922e-05, "loss": 0.0198, "step": 40760 }, { "grad_norm": 0.21057984232902527, "learning_rate": 6.891398010448333e-05, "loss": 0.0273, "step": 40770 }, { "grad_norm": 0.2864721715450287, "learning_rate": 6.8898673042883e-05, "loss": 0.0226, "step": 40780 }, { "grad_norm": 0.27631500363349915, "learning_rate": 6.888336391455222e-05, "loss": 0.0282, "step": 40790 }, { "grad_norm": 0.22823728621006012, "learning_rate": 6.886805272116513e-05, "loss": 0.0286, "step": 40800 }, { "grad_norm": 0.24597962200641632, "learning_rate": 6.885273946439617e-05, "loss": 0.0267, "step": 40810 }, { "grad_norm": 0.304504930973053, "learning_rate": 6.883742414591998e-05, "loss": 0.0262, "step": 40820 }, { "grad_norm": 0.25169187784194946, "learning_rate": 6.88221067674114e-05, "loss": 0.0243, "step": 40830 }, { "grad_norm": 0.28150439262390137, "learning_rate": 6.88067873305455e-05, "loss": 0.0255, "step": 40840 }, { "grad_norm": 0.25660884380340576, "learning_rate": 6.879146583699765e-05, "loss": 0.0197, "step": 40850 }, { "grad_norm": 0.210776686668396, "learning_rate": 6.877614228844334e-05, "loss": 0.0197, "step": 40860 }, { "grad_norm": 0.4161990284919739, "learning_rate": 6.876081668655832e-05, "loss": 0.0249, "step": 40870 }, { "grad_norm": 0.1864745318889618, "learning_rate": 6.874548903301861e-05, "loss": 0.0238, "step": 40880 }, { "grad_norm": 0.25671398639678955, "learning_rate": 6.873015932950039e-05, "loss": 0.0231, "step": 40890 }, { "grad_norm": 0.28017309308052063, "learning_rate": 6.871482757768012e-05, "loss": 0.0257, "step": 40900 }, { "grad_norm": 0.3079476058483124, "learning_rate": 6.869949377923442e-05, "loss": 0.0213, "step": 40910 }, { "grad_norm": 0.19785195589065552, "learning_rate": 6.868415793584022e-05, "loss": 0.0246, "step": 40920 }, { "grad_norm": 0.19918504357337952, "learning_rate": 6.86688200491746e-05, "loss": 0.0234, "step": 40930 }, { "grad_norm": 0.23733572661876678, "learning_rate": 6.865348012091488e-05, "loss": 0.0224, "step": 40940 }, { "grad_norm": 0.267322301864624, "learning_rate": 6.86381381527386e-05, "loss": 0.0247, "step": 40950 }, { "grad_norm": 0.25997042655944824, "learning_rate": 6.862279414632359e-05, "loss": 0.0223, "step": 40960 }, { "grad_norm": 0.2800977826118469, "learning_rate": 6.860744810334778e-05, "loss": 0.0235, "step": 40970 }, { "grad_norm": 0.16613642871379852, "learning_rate": 6.859210002548943e-05, "loss": 0.0222, "step": 40980 }, { "grad_norm": 0.26085567474365234, "learning_rate": 6.8576749914427e-05, "loss": 0.0224, "step": 40990 }, { "grad_norm": 0.2408471703529358, "learning_rate": 6.85613977718391e-05, "loss": 0.022, "step": 41000 }, { "grad_norm": 0.2179042249917984, "learning_rate": 6.854604359940464e-05, "loss": 0.0227, "step": 41010 }, { "grad_norm": 0.23292872309684753, "learning_rate": 6.853068739880276e-05, "loss": 0.0198, "step": 41020 }, { "grad_norm": 0.1821887195110321, "learning_rate": 6.851532917171276e-05, "loss": 0.0213, "step": 41030 }, { "grad_norm": 0.21379829943180084, "learning_rate": 6.849996891981418e-05, "loss": 0.0227, "step": 41040 }, { "grad_norm": 0.17313936352729797, "learning_rate": 6.848460664478684e-05, "loss": 0.0199, "step": 41050 }, { "grad_norm": 0.26691994071006775, "learning_rate": 6.846924234831069e-05, "loss": 0.0232, "step": 41060 }, { "grad_norm": 0.20284366607666016, "learning_rate": 6.845387603206597e-05, "loss": 0.0233, "step": 41070 }, { "grad_norm": 0.22367200255393982, "learning_rate": 6.843850769773312e-05, "loss": 0.0235, "step": 41080 }, { "grad_norm": 0.2773991823196411, "learning_rate": 6.842313734699278e-05, "loss": 0.0193, "step": 41090 }, { "grad_norm": 0.19936569035053253, "learning_rate": 6.840776498152584e-05, "loss": 0.0194, "step": 41100 }, { "grad_norm": 0.2637753188610077, "learning_rate": 6.83923906030134e-05, "loss": 0.0239, "step": 41110 }, { "grad_norm": 0.24480774998664856, "learning_rate": 6.837701421313677e-05, "loss": 0.0217, "step": 41120 }, { "grad_norm": 0.2391466349363327, "learning_rate": 6.83616358135775e-05, "loss": 0.0202, "step": 41130 }, { "grad_norm": 0.19743859767913818, "learning_rate": 6.834625540601734e-05, "loss": 0.0213, "step": 41140 }, { "grad_norm": 0.26577067375183105, "learning_rate": 6.833087299213829e-05, "loss": 0.0223, "step": 41150 }, { "grad_norm": 0.7787466049194336, "learning_rate": 6.83154885736225e-05, "loss": 0.025, "step": 41160 }, { "grad_norm": 0.21567051112651825, "learning_rate": 6.830010215215244e-05, "loss": 0.0242, "step": 41170 }, { "grad_norm": 0.25032883882522583, "learning_rate": 6.828471372941072e-05, "loss": 0.026, "step": 41180 }, { "grad_norm": 0.2601791024208069, "learning_rate": 6.826932330708018e-05, "loss": 0.0233, "step": 41190 }, { "grad_norm": 0.2477971315383911, "learning_rate": 6.825393088684393e-05, "loss": 0.0254, "step": 41200 }, { "grad_norm": 0.24559278786182404, "learning_rate": 6.823853647038524e-05, "loss": 0.0216, "step": 41210 }, { "grad_norm": 0.23376156389713287, "learning_rate": 6.822314005938762e-05, "loss": 0.0231, "step": 41220 }, { "grad_norm": 0.22377552092075348, "learning_rate": 6.820774165553479e-05, "loss": 0.0246, "step": 41230 }, { "grad_norm": 0.1883409470319748, "learning_rate": 6.819234126051072e-05, "loss": 0.0231, "step": 41240 }, { "grad_norm": 0.21471410989761353, "learning_rate": 6.817693887599956e-05, "loss": 0.0269, "step": 41250 }, { "grad_norm": 0.24273031949996948, "learning_rate": 6.816153450368568e-05, "loss": 0.0216, "step": 41260 }, { "grad_norm": 0.18960939347743988, "learning_rate": 6.814612814525372e-05, "loss": 0.0237, "step": 41270 }, { "grad_norm": 0.2005922794342041, "learning_rate": 6.813071980238846e-05, "loss": 0.0188, "step": 41280 }, { "grad_norm": 0.18558591604232788, "learning_rate": 6.811530947677492e-05, "loss": 0.0248, "step": 41290 }, { "grad_norm": 0.25795623660087585, "learning_rate": 6.809989717009839e-05, "loss": 0.021, "step": 41300 }, { "grad_norm": 0.21940010786056519, "learning_rate": 6.808448288404431e-05, "loss": 0.0213, "step": 41310 }, { "grad_norm": 0.2132710963487625, "learning_rate": 6.806906662029838e-05, "loss": 0.0208, "step": 41320 }, { "grad_norm": 0.23736461997032166, "learning_rate": 6.805364838054647e-05, "loss": 0.0209, "step": 41330 }, { "grad_norm": 0.21847184002399445, "learning_rate": 6.803822816647474e-05, "loss": 0.0188, "step": 41340 }, { "grad_norm": 0.1579228788614273, "learning_rate": 6.802280597976949e-05, "loss": 0.0212, "step": 41350 }, { "grad_norm": 0.2153356969356537, "learning_rate": 6.800738182211727e-05, "loss": 0.0212, "step": 41360 }, { "grad_norm": 0.17382846772670746, "learning_rate": 6.799195569520485e-05, "loss": 0.0188, "step": 41370 }, { "grad_norm": 0.19214004278182983, "learning_rate": 6.79765276007192e-05, "loss": 0.0208, "step": 41380 }, { "grad_norm": 0.2517099678516388, "learning_rate": 6.796109754034753e-05, "loss": 0.0187, "step": 41390 }, { "grad_norm": 0.1575021743774414, "learning_rate": 6.794566551577724e-05, "loss": 0.0223, "step": 41400 }, { "grad_norm": 0.27758949995040894, "learning_rate": 6.793023152869594e-05, "loss": 0.0197, "step": 41410 }, { "grad_norm": 0.21111853420734406, "learning_rate": 6.791479558079147e-05, "loss": 0.02, "step": 41420 }, { "grad_norm": 0.2191196233034134, "learning_rate": 6.789935767375191e-05, "loss": 0.0196, "step": 41430 }, { "grad_norm": 0.19303251802921295, "learning_rate": 6.78839178092655e-05, "loss": 0.0226, "step": 41440 }, { "grad_norm": 0.31148970127105713, "learning_rate": 6.786847598902072e-05, "loss": 0.0236, "step": 41450 }, { "grad_norm": 0.21620792150497437, "learning_rate": 6.785303221470627e-05, "loss": 0.0279, "step": 41460 }, { "grad_norm": 0.3233644366264343, "learning_rate": 6.783758648801108e-05, "loss": 0.0201, "step": 41470 }, { "grad_norm": 0.3325812518596649, "learning_rate": 6.782213881062422e-05, "loss": 0.0233, "step": 41480 }, { "grad_norm": 0.23472505807876587, "learning_rate": 6.780668918423508e-05, "loss": 0.0217, "step": 41490 }, { "grad_norm": 0.19219253957271576, "learning_rate": 6.779123761053317e-05, "loss": 0.0188, "step": 41500 }, { "grad_norm": 0.25340735912323, "learning_rate": 6.777578409120826e-05, "loss": 0.0195, "step": 41510 }, { "grad_norm": 0.22547829151153564, "learning_rate": 6.776032862795033e-05, "loss": 0.025, "step": 41520 }, { "grad_norm": 0.17731359601020813, "learning_rate": 6.774487122244956e-05, "loss": 0.0224, "step": 41530 }, { "grad_norm": 0.22515267133712769, "learning_rate": 6.772941187639637e-05, "loss": 0.0183, "step": 41540 }, { "grad_norm": 0.20543669164180756, "learning_rate": 6.771395059148134e-05, "loss": 0.0223, "step": 41550 }, { "grad_norm": 0.19870731234550476, "learning_rate": 6.76984873693953e-05, "loss": 0.0188, "step": 41560 }, { "grad_norm": 0.38098445534706116, "learning_rate": 6.768302221182931e-05, "loss": 0.0235, "step": 41570 }, { "grad_norm": 0.20194680988788605, "learning_rate": 6.766755512047457e-05, "loss": 0.0202, "step": 41580 }, { "grad_norm": 0.22757172584533691, "learning_rate": 6.765208609702259e-05, "loss": 0.0236, "step": 41590 }, { "grad_norm": 0.18981434404850006, "learning_rate": 6.763661514316499e-05, "loss": 0.0185, "step": 41600 }, { "grad_norm": 0.2156091183423996, "learning_rate": 6.76211422605937e-05, "loss": 0.0222, "step": 41610 }, { "grad_norm": 0.17041070759296417, "learning_rate": 6.760566745100076e-05, "loss": 0.019, "step": 41620 }, { "grad_norm": 0.26091763377189636, "learning_rate": 6.75901907160785e-05, "loss": 0.0242, "step": 41630 }, { "grad_norm": 0.22330951690673828, "learning_rate": 6.757471205751943e-05, "loss": 0.0255, "step": 41640 }, { "grad_norm": 0.2183196097612381, "learning_rate": 6.75592314770163e-05, "loss": 0.0244, "step": 41650 }, { "grad_norm": 0.18911010026931763, "learning_rate": 6.7543748976262e-05, "loss": 0.018, "step": 41660 }, { "grad_norm": 0.2669483423233032, "learning_rate": 6.752826455694968e-05, "loss": 0.0252, "step": 41670 }, { "grad_norm": 0.19783642888069153, "learning_rate": 6.751277822077271e-05, "loss": 0.0208, "step": 41680 }, { "grad_norm": 0.20560401678085327, "learning_rate": 6.749728996942465e-05, "loss": 0.0247, "step": 41690 }, { "grad_norm": 0.21743451058864594, "learning_rate": 6.748179980459924e-05, "loss": 0.0212, "step": 41700 }, { "grad_norm": 0.25707608461380005, "learning_rate": 6.746630772799052e-05, "loss": 0.0231, "step": 41710 }, { "grad_norm": 0.23092204332351685, "learning_rate": 6.745081374129262e-05, "loss": 0.0206, "step": 41720 }, { "grad_norm": 0.19040974974632263, "learning_rate": 6.743531784619998e-05, "loss": 0.0223, "step": 41730 }, { "grad_norm": 0.24671931564807892, "learning_rate": 6.741982004440719e-05, "loss": 0.0244, "step": 41740 }, { "grad_norm": 0.2582656741142273, "learning_rate": 6.740432033760907e-05, "loss": 0.0243, "step": 41750 }, { "grad_norm": 0.18713629245758057, "learning_rate": 6.738881872750066e-05, "loss": 0.0236, "step": 41760 }, { "grad_norm": 0.2562991678714752, "learning_rate": 6.737331521577715e-05, "loss": 0.0214, "step": 41770 }, { "grad_norm": 0.18627581000328064, "learning_rate": 6.735780980413403e-05, "loss": 0.0259, "step": 41780 }, { "grad_norm": 0.19581733644008636, "learning_rate": 6.734230249426692e-05, "loss": 0.0206, "step": 41790 }, { "grad_norm": 0.18451298773288727, "learning_rate": 6.732679328787168e-05, "loss": 0.0215, "step": 41800 }, { "grad_norm": 0.274425745010376, "learning_rate": 6.731128218664438e-05, "loss": 0.0249, "step": 41810 }, { "grad_norm": 0.2549055814743042, "learning_rate": 6.729576919228129e-05, "loss": 0.0211, "step": 41820 }, { "grad_norm": 0.2617138624191284, "learning_rate": 6.728025430647888e-05, "loss": 0.0218, "step": 41830 }, { "grad_norm": 0.2534711956977844, "learning_rate": 6.726473753093383e-05, "loss": 0.0217, "step": 41840 }, { "grad_norm": 0.23474903404712677, "learning_rate": 6.724921886734305e-05, "loss": 0.0249, "step": 41850 }, { "grad_norm": 0.23191943764686584, "learning_rate": 6.723369831740363e-05, "loss": 0.0188, "step": 41860 }, { "grad_norm": 0.2669115364551544, "learning_rate": 6.721817588281288e-05, "loss": 0.0218, "step": 41870 }, { "grad_norm": 0.28211525082588196, "learning_rate": 6.720265156526828e-05, "loss": 0.0213, "step": 41880 }, { "grad_norm": 0.26030105352401733, "learning_rate": 6.718712536646758e-05, "loss": 0.0232, "step": 41890 }, { "grad_norm": 0.1874251365661621, "learning_rate": 6.71715972881087e-05, "loss": 0.0224, "step": 41900 }, { "grad_norm": 0.26109403371810913, "learning_rate": 6.715606733188974e-05, "loss": 0.0249, "step": 41910 }, { "grad_norm": 0.1988108605146408, "learning_rate": 6.714053549950909e-05, "loss": 0.0226, "step": 41920 }, { "grad_norm": 0.18794548511505127, "learning_rate": 6.71250017926652e-05, "loss": 0.0198, "step": 41930 }, { "grad_norm": 0.16471333801746368, "learning_rate": 6.71094662130569e-05, "loss": 0.0208, "step": 41940 }, { "grad_norm": 0.24523136019706726, "learning_rate": 6.709392876238307e-05, "loss": 0.02, "step": 41950 }, { "grad_norm": 0.23762904107570648, "learning_rate": 6.707838944234289e-05, "loss": 0.0213, "step": 41960 }, { "grad_norm": 0.3240586519241333, "learning_rate": 6.706284825463574e-05, "loss": 0.0228, "step": 41970 }, { "grad_norm": 0.17693649232387543, "learning_rate": 6.704730520096114e-05, "loss": 0.0222, "step": 41980 }, { "grad_norm": 0.2937830984592438, "learning_rate": 6.703176028301888e-05, "loss": 0.0229, "step": 41990 }, { "grad_norm": 0.2552689015865326, "learning_rate": 6.701621350250892e-05, "loss": 0.0272, "step": 42000 }, { "grad_norm": 0.2688751220703125, "learning_rate": 6.700066486113144e-05, "loss": 0.024, "step": 42010 }, { "grad_norm": 0.19122987985610962, "learning_rate": 6.69851143605868e-05, "loss": 0.0212, "step": 42020 }, { "grad_norm": 0.1834849715232849, "learning_rate": 6.696956200257561e-05, "loss": 0.022, "step": 42030 }, { "grad_norm": 0.17380274832248688, "learning_rate": 6.695400778879863e-05, "loss": 0.0203, "step": 42040 }, { "grad_norm": 0.2810570001602173, "learning_rate": 6.693845172095683e-05, "loss": 0.0206, "step": 42050 }, { "grad_norm": 0.20368005335330963, "learning_rate": 6.692289380075142e-05, "loss": 0.0211, "step": 42060 }, { "grad_norm": 0.17736756801605225, "learning_rate": 6.690733402988379e-05, "loss": 0.0242, "step": 42070 }, { "grad_norm": 0.19968771934509277, "learning_rate": 6.689177241005553e-05, "loss": 0.0218, "step": 42080 }, { "grad_norm": 0.23448091745376587, "learning_rate": 6.687620894296844e-05, "loss": 0.0202, "step": 42090 }, { "grad_norm": 0.20893952250480652, "learning_rate": 6.686064363032451e-05, "loss": 0.0233, "step": 42100 }, { "grad_norm": 0.205468088388443, "learning_rate": 6.684507647382596e-05, "loss": 0.0256, "step": 42110 }, { "grad_norm": 0.24030697345733643, "learning_rate": 6.682950747517513e-05, "loss": 0.0211, "step": 42120 }, { "grad_norm": 0.1957574188709259, "learning_rate": 6.68139366360747e-05, "loss": 0.0209, "step": 42130 }, { "grad_norm": 0.17899401485919952, "learning_rate": 6.679836395822744e-05, "loss": 0.0225, "step": 42140 }, { "grad_norm": 0.17366105318069458, "learning_rate": 6.678278944333633e-05, "loss": 0.019, "step": 42150 }, { "grad_norm": 0.22845374047756195, "learning_rate": 6.676721309310462e-05, "loss": 0.0253, "step": 42160 }, { "grad_norm": 0.284775048494339, "learning_rate": 6.67516349092357e-05, "loss": 0.0193, "step": 42170 }, { "grad_norm": 0.2054770290851593, "learning_rate": 6.673605489343315e-05, "loss": 0.0207, "step": 42180 }, { "grad_norm": 0.1996784210205078, "learning_rate": 6.672047304740081e-05, "loss": 0.0255, "step": 42190 }, { "grad_norm": 0.2123512327671051, "learning_rate": 6.670488937284268e-05, "loss": 0.0241, "step": 42200 }, { "grad_norm": 0.268831342458725, "learning_rate": 6.668930387146294e-05, "loss": 0.0208, "step": 42210 }, { "grad_norm": 0.1990453451871872, "learning_rate": 6.667371654496605e-05, "loss": 0.0209, "step": 42220 }, { "grad_norm": 0.20512327551841736, "learning_rate": 6.665812739505659e-05, "loss": 0.0197, "step": 42230 }, { "grad_norm": 0.2749043405056, "learning_rate": 6.664253642343935e-05, "loss": 0.0198, "step": 42240 }, { "grad_norm": 0.23050138354301453, "learning_rate": 6.662694363181935e-05, "loss": 0.0201, "step": 42250 }, { "grad_norm": 0.19389203190803528, "learning_rate": 6.66113490219018e-05, "loss": 0.0228, "step": 42260 }, { "grad_norm": 0.22994640469551086, "learning_rate": 6.659575259539207e-05, "loss": 0.0228, "step": 42270 }, { "grad_norm": 0.1964968889951706, "learning_rate": 6.65801543539958e-05, "loss": 0.0197, "step": 42280 }, { "grad_norm": 0.2442346066236496, "learning_rate": 6.65645542994188e-05, "loss": 0.0206, "step": 42290 }, { "grad_norm": 0.23140190541744232, "learning_rate": 6.654895243336702e-05, "loss": 0.0216, "step": 42300 }, { "grad_norm": 0.21428711712360382, "learning_rate": 6.65333487575467e-05, "loss": 0.0231, "step": 42310 }, { "grad_norm": 0.18383295834064484, "learning_rate": 6.651774327366422e-05, "loss": 0.0214, "step": 42320 }, { "grad_norm": 0.23944269120693207, "learning_rate": 6.650213598342613e-05, "loss": 0.0234, "step": 42330 }, { "grad_norm": 0.1477167010307312, "learning_rate": 6.648652688853931e-05, "loss": 0.0196, "step": 42340 }, { "grad_norm": 0.1930953562259674, "learning_rate": 6.647091599071066e-05, "loss": 0.0189, "step": 42350 }, { "grad_norm": 0.21153433620929718, "learning_rate": 6.645530329164743e-05, "loss": 0.0204, "step": 42360 }, { "grad_norm": 0.2742076516151428, "learning_rate": 6.643968879305694e-05, "loss": 0.0233, "step": 42370 }, { "grad_norm": 0.2345179170370102, "learning_rate": 6.642407249664683e-05, "loss": 0.0236, "step": 42380 }, { "grad_norm": 0.19246245920658112, "learning_rate": 6.640845440412483e-05, "loss": 0.0218, "step": 42390 }, { "grad_norm": 0.215981587767601, "learning_rate": 6.639283451719893e-05, "loss": 0.0235, "step": 42400 }, { "grad_norm": 0.25105080008506775, "learning_rate": 6.637721283757729e-05, "loss": 0.0243, "step": 42410 }, { "grad_norm": 0.16209086775779724, "learning_rate": 6.636158936696828e-05, "loss": 0.0211, "step": 42420 }, { "grad_norm": 0.24503253400325775, "learning_rate": 6.634596410708047e-05, "loss": 0.0215, "step": 42430 }, { "grad_norm": 0.2506985366344452, "learning_rate": 6.633033705962257e-05, "loss": 0.0233, "step": 42440 }, { "grad_norm": 0.24117501080036163, "learning_rate": 6.631470822630359e-05, "loss": 0.0249, "step": 42450 }, { "grad_norm": 0.22244544327259064, "learning_rate": 6.629907760883263e-05, "loss": 0.0231, "step": 42460 }, { "grad_norm": 0.20854172110557556, "learning_rate": 6.628344520891905e-05, "loss": 0.0188, "step": 42470 }, { "grad_norm": 0.18341389298439026, "learning_rate": 6.62678110282724e-05, "loss": 0.0226, "step": 42480 }, { "grad_norm": 0.2860443592071533, "learning_rate": 6.625217506860237e-05, "loss": 0.023, "step": 42490 }, { "grad_norm": 0.2440931797027588, "learning_rate": 6.623653733161892e-05, "loss": 0.0236, "step": 42500 }, { "grad_norm": 0.1736789345741272, "learning_rate": 6.622089781903217e-05, "loss": 0.0206, "step": 42510 }, { "grad_norm": 0.19482038915157318, "learning_rate": 6.620525653255241e-05, "loss": 0.0281, "step": 42520 }, { "grad_norm": 0.2643069624900818, "learning_rate": 6.618961347389015e-05, "loss": 0.0202, "step": 42530 }, { "grad_norm": 0.18397405743598938, "learning_rate": 6.617396864475613e-05, "loss": 0.0237, "step": 42540 }, { "grad_norm": 0.20702655613422394, "learning_rate": 6.61583220468612e-05, "loss": 0.0171, "step": 42550 }, { "grad_norm": 0.19902676343917847, "learning_rate": 6.614267368191645e-05, "loss": 0.0203, "step": 42560 }, { "grad_norm": 0.4390590190887451, "learning_rate": 6.61270235516332e-05, "loss": 0.022, "step": 42570 }, { "grad_norm": 0.26121994853019714, "learning_rate": 6.611137165772287e-05, "loss": 0.0224, "step": 42580 }, { "grad_norm": 0.2704324722290039, "learning_rate": 6.60957180018972e-05, "loss": 0.0183, "step": 42590 }, { "grad_norm": 0.2016293853521347, "learning_rate": 6.608006258586797e-05, "loss": 0.0216, "step": 42600 }, { "grad_norm": 0.18635205924510956, "learning_rate": 6.606440541134731e-05, "loss": 0.0187, "step": 42610 }, { "grad_norm": 0.25781968235969543, "learning_rate": 6.60487464800474e-05, "loss": 0.0198, "step": 42620 }, { "grad_norm": 0.22485053539276123, "learning_rate": 6.603308579368071e-05, "loss": 0.0198, "step": 42630 }, { "grad_norm": 0.2230103760957718, "learning_rate": 6.601742335395987e-05, "loss": 0.0214, "step": 42640 }, { "grad_norm": 0.20228423178195953, "learning_rate": 6.600175916259769e-05, "loss": 0.019, "step": 42650 }, { "grad_norm": 0.2140170782804489, "learning_rate": 6.598609322130718e-05, "loss": 0.0179, "step": 42660 }, { "grad_norm": 0.1783076822757721, "learning_rate": 6.597042553180155e-05, "loss": 0.0187, "step": 42670 }, { "grad_norm": 0.15674085915088654, "learning_rate": 6.59547560957942e-05, "loss": 0.02, "step": 42680 }, { "grad_norm": 0.18799617886543274, "learning_rate": 6.59390849149987e-05, "loss": 0.0239, "step": 42690 }, { "grad_norm": 0.20359858870506287, "learning_rate": 6.592341199112886e-05, "loss": 0.0225, "step": 42700 }, { "grad_norm": 0.22353236377239227, "learning_rate": 6.590773732589861e-05, "loss": 0.0203, "step": 42710 }, { "grad_norm": 0.23544728755950928, "learning_rate": 6.589206092102213e-05, "loss": 0.0248, "step": 42720 }, { "grad_norm": 0.21198885142803192, "learning_rate": 6.587638277821375e-05, "loss": 0.0198, "step": 42730 }, { "grad_norm": 0.18979991972446442, "learning_rate": 6.586070289918803e-05, "loss": 0.0219, "step": 42740 }, { "grad_norm": 0.2591067850589752, "learning_rate": 6.584502128565968e-05, "loss": 0.0193, "step": 42750 }, { "grad_norm": 0.24361106753349304, "learning_rate": 6.582933793934363e-05, "loss": 0.0227, "step": 42760 }, { "grad_norm": 0.1939699500799179, "learning_rate": 6.581365286195498e-05, "loss": 0.0202, "step": 42770 }, { "grad_norm": 0.1969667226076126, "learning_rate": 6.579796605520903e-05, "loss": 0.0206, "step": 42780 }, { "grad_norm": 0.20734430849552155, "learning_rate": 6.578227752082127e-05, "loss": 0.0228, "step": 42790 }, { "grad_norm": 0.2095235288143158, "learning_rate": 6.576658726050735e-05, "loss": 0.0198, "step": 42800 }, { "grad_norm": 0.20255321264266968, "learning_rate": 6.575089527598316e-05, "loss": 0.0207, "step": 42810 }, { "grad_norm": 0.2258266806602478, "learning_rate": 6.573520156896476e-05, "loss": 0.022, "step": 42820 }, { "grad_norm": 0.21242013573646545, "learning_rate": 6.571950614116835e-05, "loss": 0.0215, "step": 42830 }, { "grad_norm": 0.22312019765377045, "learning_rate": 6.570380899431039e-05, "loss": 0.0246, "step": 42840 }, { "grad_norm": 0.22034920752048492, "learning_rate": 6.568811013010749e-05, "loss": 0.0201, "step": 42850 }, { "grad_norm": 0.20168578624725342, "learning_rate": 6.567240955027645e-05, "loss": 0.0228, "step": 42860 }, { "grad_norm": 0.2049509733915329, "learning_rate": 6.565670725653427e-05, "loss": 0.0194, "step": 42870 }, { "grad_norm": 0.18469952046871185, "learning_rate": 6.564100325059812e-05, "loss": 0.0203, "step": 42880 }, { "grad_norm": 0.24907855689525604, "learning_rate": 6.562529753418538e-05, "loss": 0.0217, "step": 42890 }, { "grad_norm": 0.19949990510940552, "learning_rate": 6.56095901090136e-05, "loss": 0.0216, "step": 42900 }, { "grad_norm": 0.2619963586330414, "learning_rate": 6.55938809768005e-05, "loss": 0.0213, "step": 42910 }, { "grad_norm": 0.22254112362861633, "learning_rate": 6.557817013926402e-05, "loss": 0.0187, "step": 42920 }, { "grad_norm": 0.2025921642780304, "learning_rate": 6.556245759812227e-05, "loss": 0.0188, "step": 42930 }, { "grad_norm": 0.2248309701681137, "learning_rate": 6.554674335509357e-05, "loss": 0.0236, "step": 42940 }, { "grad_norm": 0.2590397596359253, "learning_rate": 6.553102741189638e-05, "loss": 0.0205, "step": 42950 }, { "grad_norm": 0.2165997475385666, "learning_rate": 6.55153097702494e-05, "loss": 0.021, "step": 42960 }, { "grad_norm": 0.22551265358924866, "learning_rate": 6.549959043187146e-05, "loss": 0.0218, "step": 42970 }, { "grad_norm": 0.22422291338443756, "learning_rate": 6.54838693984816e-05, "loss": 0.0233, "step": 42980 }, { "grad_norm": 0.21656963229179382, "learning_rate": 6.546814667179909e-05, "loss": 0.0218, "step": 42990 }, { "grad_norm": 0.24275831878185272, "learning_rate": 6.545242225354328e-05, "loss": 0.0193, "step": 43000 }, { "grad_norm": 0.3064572811126709, "learning_rate": 6.543669614543383e-05, "loss": 0.0217, "step": 43010 }, { "grad_norm": 0.22097374498844147, "learning_rate": 6.542096834919049e-05, "loss": 0.0239, "step": 43020 }, { "grad_norm": 0.22479523718357086, "learning_rate": 6.540523886653324e-05, "loss": 0.0232, "step": 43030 }, { "grad_norm": 0.23698599636554718, "learning_rate": 6.538950769918222e-05, "loss": 0.0231, "step": 43040 }, { "grad_norm": 0.2358255684375763, "learning_rate": 6.537377484885779e-05, "loss": 0.0227, "step": 43050 }, { "grad_norm": 0.23992758989334106, "learning_rate": 6.535804031728045e-05, "loss": 0.0234, "step": 43060 }, { "grad_norm": 0.2174280732870102, "learning_rate": 6.534230410617092e-05, "loss": 0.0214, "step": 43070 }, { "grad_norm": 0.2021753191947937, "learning_rate": 6.532656621725007e-05, "loss": 0.0227, "step": 43080 }, { "grad_norm": 0.17224131524562836, "learning_rate": 6.531082665223898e-05, "loss": 0.0217, "step": 43090 }, { "grad_norm": 0.23287515342235565, "learning_rate": 6.529508541285889e-05, "loss": 0.0256, "step": 43100 }, { "grad_norm": 0.23635083436965942, "learning_rate": 6.527934250083128e-05, "loss": 0.0217, "step": 43110 }, { "grad_norm": 0.2669701874256134, "learning_rate": 6.526359791787775e-05, "loss": 0.0208, "step": 43120 }, { "grad_norm": 0.2107647806406021, "learning_rate": 6.524785166572009e-05, "loss": 0.0211, "step": 43130 }, { "grad_norm": 0.2280542552471161, "learning_rate": 6.523210374608029e-05, "loss": 0.0192, "step": 43140 }, { "grad_norm": 0.2288053035736084, "learning_rate": 6.521635416068054e-05, "loss": 0.0193, "step": 43150 }, { "grad_norm": 0.16912303864955902, "learning_rate": 6.520060291124317e-05, "loss": 0.0227, "step": 43160 }, { "grad_norm": 0.21437153220176697, "learning_rate": 6.518484999949072e-05, "loss": 0.017, "step": 43170 }, { "grad_norm": 0.20744822919368744, "learning_rate": 6.516909542714591e-05, "loss": 0.0179, "step": 43180 }, { "grad_norm": 0.1799403429031372, "learning_rate": 6.515333919593161e-05, "loss": 0.023, "step": 43190 }, { "grad_norm": 0.196995809674263, "learning_rate": 6.513758130757094e-05, "loss": 0.0215, "step": 43200 }, { "grad_norm": 0.3061525821685791, "learning_rate": 6.512182176378713e-05, "loss": 0.0208, "step": 43210 }, { "grad_norm": 0.23707158863544464, "learning_rate": 6.510606056630362e-05, "loss": 0.0222, "step": 43220 }, { "grad_norm": 0.27347853779792786, "learning_rate": 6.509029771684405e-05, "loss": 0.0214, "step": 43230 }, { "grad_norm": 0.2903575003147125, "learning_rate": 6.50745332171322e-05, "loss": 0.0249, "step": 43240 }, { "grad_norm": 0.1944480538368225, "learning_rate": 6.505876706889207e-05, "loss": 0.0226, "step": 43250 }, { "grad_norm": 0.23050831258296967, "learning_rate": 6.504299927384781e-05, "loss": 0.0223, "step": 43260 }, { "grad_norm": 0.27174869179725647, "learning_rate": 6.502722983372378e-05, "loss": 0.0209, "step": 43270 }, { "grad_norm": 0.20452994108200073, "learning_rate": 6.501145875024446e-05, "loss": 0.0226, "step": 43280 }, { "grad_norm": 0.23828180134296417, "learning_rate": 6.499568602513459e-05, "loss": 0.0232, "step": 43290 }, { "grad_norm": 0.20837217569351196, "learning_rate": 6.497991166011903e-05, "loss": 0.0216, "step": 43300 }, { "grad_norm": 0.21749569475650787, "learning_rate": 6.49641356569229e-05, "loss": 0.0224, "step": 43310 }, { "grad_norm": 0.2038729041814804, "learning_rate": 6.494835801727135e-05, "loss": 0.0229, "step": 43320 }, { "grad_norm": 0.21375106275081635, "learning_rate": 6.493257874288986e-05, "loss": 0.0219, "step": 43330 }, { "grad_norm": 0.18804346024990082, "learning_rate": 6.491679783550399e-05, "loss": 0.0202, "step": 43340 }, { "grad_norm": 0.25033438205718994, "learning_rate": 6.490101529683957e-05, "loss": 0.0204, "step": 43350 }, { "grad_norm": 0.2385934740304947, "learning_rate": 6.488523112862248e-05, "loss": 0.0222, "step": 43360 }, { "grad_norm": 0.20725484192371368, "learning_rate": 6.486944533257891e-05, "loss": 0.0211, "step": 43370 }, { "grad_norm": 0.18817417323589325, "learning_rate": 6.485365791043517e-05, "loss": 0.0249, "step": 43380 }, { "grad_norm": 0.16869881749153137, "learning_rate": 6.483786886391772e-05, "loss": 0.0199, "step": 43390 }, { "grad_norm": 0.21986903250217438, "learning_rate": 6.482207819475323e-05, "loss": 0.0184, "step": 43400 }, { "grad_norm": 0.16938996315002441, "learning_rate": 6.480628590466857e-05, "loss": 0.0219, "step": 43410 }, { "grad_norm": 0.18008852005004883, "learning_rate": 6.479049199539074e-05, "loss": 0.0214, "step": 43420 }, { "grad_norm": 0.25471338629722595, "learning_rate": 6.477469646864693e-05, "loss": 0.0199, "step": 43430 }, { "grad_norm": 0.17351533472537994, "learning_rate": 6.475889932616454e-05, "loss": 0.0209, "step": 43440 }, { "grad_norm": 0.21647174656391144, "learning_rate": 6.474310056967111e-05, "loss": 0.0204, "step": 43450 }, { "grad_norm": 0.21573379635810852, "learning_rate": 6.472730020089437e-05, "loss": 0.0212, "step": 43460 }, { "grad_norm": 0.26470521092414856, "learning_rate": 6.471149822156222e-05, "loss": 0.0215, "step": 43470 }, { "grad_norm": 0.22581566870212555, "learning_rate": 6.469569463340275e-05, "loss": 0.0234, "step": 43480 }, { "grad_norm": 0.25856176018714905, "learning_rate": 6.467988943814418e-05, "loss": 0.0223, "step": 43490 }, { "grad_norm": 0.2544693350791931, "learning_rate": 6.4664082637515e-05, "loss": 0.021, "step": 43500 }, { "grad_norm": 0.16132065653800964, "learning_rate": 6.464827423324377e-05, "loss": 0.0256, "step": 43510 }, { "grad_norm": 0.24862557649612427, "learning_rate": 6.463246422705931e-05, "loss": 0.0244, "step": 43520 }, { "grad_norm": 0.2500322163105011, "learning_rate": 6.461665262069055e-05, "loss": 0.0217, "step": 43530 }, { "grad_norm": 0.19567592442035675, "learning_rate": 6.460083941586665e-05, "loss": 0.023, "step": 43540 }, { "grad_norm": 0.22579461336135864, "learning_rate": 6.45850246143169e-05, "loss": 0.0239, "step": 43550 }, { "grad_norm": 0.17156505584716797, "learning_rate": 6.456920821777077e-05, "loss": 0.0218, "step": 43560 }, { "grad_norm": 0.18603840470314026, "learning_rate": 6.455339022795796e-05, "loss": 0.0233, "step": 43570 }, { "grad_norm": 0.19079473614692688, "learning_rate": 6.453757064660825e-05, "loss": 0.0236, "step": 43580 }, { "grad_norm": 0.20844954252243042, "learning_rate": 6.452174947545169e-05, "loss": 0.0209, "step": 43590 }, { "grad_norm": 0.27598118782043457, "learning_rate": 6.450592671621842e-05, "loss": 0.0256, "step": 43600 }, { "grad_norm": 0.23983581364154816, "learning_rate": 6.449010237063882e-05, "loss": 0.0204, "step": 43610 }, { "grad_norm": 0.2523171007633209, "learning_rate": 6.447427644044343e-05, "loss": 0.0205, "step": 43620 }, { "grad_norm": 0.205433651804924, "learning_rate": 6.44584489273629e-05, "loss": 0.0194, "step": 43630 }, { "grad_norm": 0.28314024209976196, "learning_rate": 6.444261983312817e-05, "loss": 0.0287, "step": 43640 }, { "grad_norm": 0.2114475518465042, "learning_rate": 6.442678915947023e-05, "loss": 0.0231, "step": 43650 }, { "grad_norm": 0.27858155965805054, "learning_rate": 6.441095690812032e-05, "loss": 0.0266, "step": 43660 }, { "grad_norm": 0.20074120163917542, "learning_rate": 6.439512308080982e-05, "loss": 0.0235, "step": 43670 }, { "grad_norm": 0.2380058616399765, "learning_rate": 6.43792876792703e-05, "loss": 0.0209, "step": 43680 }, { "grad_norm": 0.3003307580947876, "learning_rate": 6.43634507052335e-05, "loss": 0.0238, "step": 43690 }, { "grad_norm": 0.19092877209186554, "learning_rate": 6.434761216043135e-05, "loss": 0.0208, "step": 43700 }, { "grad_norm": 0.19307203590869904, "learning_rate": 6.433177204659589e-05, "loss": 0.0218, "step": 43710 }, { "grad_norm": 0.23810191452503204, "learning_rate": 6.431593036545938e-05, "loss": 0.0232, "step": 43720 }, { "grad_norm": 0.24370373785495758, "learning_rate": 6.430008711875426e-05, "loss": 0.0215, "step": 43730 }, { "grad_norm": 0.19134710729122162, "learning_rate": 6.42842423082131e-05, "loss": 0.0218, "step": 43740 }, { "grad_norm": 0.18830838799476624, "learning_rate": 6.426839593556868e-05, "loss": 0.0236, "step": 43750 }, { "grad_norm": 0.22962641716003418, "learning_rate": 6.425254800255392e-05, "loss": 0.0219, "step": 43760 }, { "grad_norm": 0.23085252940654755, "learning_rate": 6.423669851090196e-05, "loss": 0.0227, "step": 43770 }, { "grad_norm": 0.2689226567745209, "learning_rate": 6.422084746234604e-05, "loss": 0.0241, "step": 43780 }, { "grad_norm": 0.2157234251499176, "learning_rate": 6.420499485861965e-05, "loss": 0.0199, "step": 43790 }, { "grad_norm": 0.17014823853969574, "learning_rate": 6.418914070145636e-05, "loss": 0.0199, "step": 43800 }, { "grad_norm": 0.21101504564285278, "learning_rate": 6.417328499258996e-05, "loss": 0.0181, "step": 43810 }, { "grad_norm": 0.2788817584514618, "learning_rate": 6.415742773375446e-05, "loss": 0.0203, "step": 43820 }, { "grad_norm": 0.21371325850486755, "learning_rate": 6.414156892668393e-05, "loss": 0.0236, "step": 43830 }, { "grad_norm": 0.20947502553462982, "learning_rate": 6.412570857311267e-05, "loss": 0.0189, "step": 43840 }, { "grad_norm": 0.2928188145160675, "learning_rate": 6.410984667477518e-05, "loss": 0.0217, "step": 43850 }, { "grad_norm": 0.22487643361091614, "learning_rate": 6.409398323340607e-05, "loss": 0.022, "step": 43860 }, { "grad_norm": 0.2220068722963333, "learning_rate": 6.407811825074012e-05, "loss": 0.0265, "step": 43870 }, { "grad_norm": 0.263297975063324, "learning_rate": 6.406225172851234e-05, "loss": 0.0284, "step": 43880 }, { "grad_norm": 0.3330889344215393, "learning_rate": 6.404638366845786e-05, "loss": 0.0203, "step": 43890 }, { "grad_norm": 0.23566602170467377, "learning_rate": 6.403051407231196e-05, "loss": 0.022, "step": 43900 }, { "grad_norm": 0.2712399959564209, "learning_rate": 6.401464294181016e-05, "loss": 0.0204, "step": 43910 }, { "grad_norm": 0.3083360493183136, "learning_rate": 6.399877027868808e-05, "loss": 0.0245, "step": 43920 }, { "grad_norm": 0.17600755393505096, "learning_rate": 6.39828960846815e-05, "loss": 0.0197, "step": 43930 }, { "grad_norm": 0.23456566035747528, "learning_rate": 6.396702036152644e-05, "loss": 0.0233, "step": 43940 }, { "grad_norm": 0.19249080121517181, "learning_rate": 6.395114311095905e-05, "loss": 0.0223, "step": 43950 }, { "grad_norm": 0.19931332767009735, "learning_rate": 6.393526433471559e-05, "loss": 0.0218, "step": 43960 }, { "grad_norm": 0.17576876282691956, "learning_rate": 6.391938403453262e-05, "loss": 0.0207, "step": 43970 }, { "grad_norm": 0.25237777829170227, "learning_rate": 6.390350221214671e-05, "loss": 0.0261, "step": 43980 }, { "grad_norm": 0.20171776413917542, "learning_rate": 6.38876188692947e-05, "loss": 0.0208, "step": 43990 }, { "grad_norm": 0.2978047728538513, "learning_rate": 6.387173400771358e-05, "loss": 0.0278, "step": 44000 }, { "grad_norm": 0.2132893055677414, "learning_rate": 6.385584762914047e-05, "loss": 0.023, "step": 44010 }, { "grad_norm": 0.15543071925640106, "learning_rate": 6.38399597353127e-05, "loss": 0.0239, "step": 44020 }, { "grad_norm": 0.221104696393013, "learning_rate": 6.382407032796775e-05, "loss": 0.021, "step": 44030 }, { "grad_norm": 0.18532390892505646, "learning_rate": 6.380817940884325e-05, "loss": 0.0194, "step": 44040 }, { "grad_norm": 0.2789376378059387, "learning_rate": 6.379228697967702e-05, "loss": 0.0221, "step": 44050 }, { "grad_norm": 0.26945623755455017, "learning_rate": 6.3776393042207e-05, "loss": 0.0229, "step": 44060 }, { "grad_norm": 0.23251201212406158, "learning_rate": 6.376049759817137e-05, "loss": 0.0224, "step": 44070 }, { "grad_norm": 0.21007892489433289, "learning_rate": 6.37446006493084e-05, "loss": 0.0209, "step": 44080 }, { "grad_norm": 0.2294251024723053, "learning_rate": 6.372870219735656e-05, "loss": 0.0192, "step": 44090 }, { "grad_norm": 0.24467015266418457, "learning_rate": 6.371280224405452e-05, "loss": 0.0228, "step": 44100 }, { "grad_norm": 0.2075963020324707, "learning_rate": 6.369690079114101e-05, "loss": 0.0208, "step": 44110 }, { "grad_norm": 0.22654521465301514, "learning_rate": 6.368099784035504e-05, "loss": 0.0203, "step": 44120 }, { "grad_norm": 0.1664091944694519, "learning_rate": 6.366509339343572e-05, "loss": 0.0182, "step": 44130 }, { "grad_norm": 0.1800333708524704, "learning_rate": 6.364918745212232e-05, "loss": 0.0193, "step": 44140 }, { "grad_norm": 0.23712535202503204, "learning_rate": 6.363328001815431e-05, "loss": 0.0219, "step": 44150 }, { "grad_norm": 0.22688822448253632, "learning_rate": 6.361737109327128e-05, "loss": 0.0234, "step": 44160 }, { "grad_norm": 0.19456396996974945, "learning_rate": 6.360146067921301e-05, "loss": 0.0217, "step": 44170 }, { "grad_norm": 0.206015482544899, "learning_rate": 6.35855487777195e-05, "loss": 0.0175, "step": 44180 }, { "grad_norm": 0.2246965616941452, "learning_rate": 6.356963539053076e-05, "loss": 0.0216, "step": 44190 }, { "grad_norm": 0.20181827247142792, "learning_rate": 6.355372051938711e-05, "loss": 0.0201, "step": 44200 }, { "grad_norm": 0.15202926099300385, "learning_rate": 6.353780416602894e-05, "loss": 0.0199, "step": 44210 }, { "grad_norm": 0.2419348657131195, "learning_rate": 6.352188633219689e-05, "loss": 0.0195, "step": 44220 }, { "grad_norm": 0.2391827553510666, "learning_rate": 6.350596701963166e-05, "loss": 0.0194, "step": 44230 }, { "grad_norm": 0.19935116171836853, "learning_rate": 6.349004623007419e-05, "loss": 0.0207, "step": 44240 }, { "grad_norm": 0.24715596437454224, "learning_rate": 6.347412396526555e-05, "loss": 0.0209, "step": 44250 }, { "grad_norm": 0.2989742159843445, "learning_rate": 6.345820022694696e-05, "loss": 0.0224, "step": 44260 }, { "grad_norm": 0.2491610199213028, "learning_rate": 6.344227501685984e-05, "loss": 0.0259, "step": 44270 }, { "grad_norm": 0.1883103847503662, "learning_rate": 6.342634833674572e-05, "loss": 0.0206, "step": 44280 }, { "grad_norm": 0.21680913865566254, "learning_rate": 6.341042018834635e-05, "loss": 0.0199, "step": 44290 }, { "grad_norm": 0.23230503499507904, "learning_rate": 6.339449057340359e-05, "loss": 0.0204, "step": 44300 }, { "grad_norm": 0.22721169888973236, "learning_rate": 6.337855949365945e-05, "loss": 0.0232, "step": 44310 }, { "grad_norm": 0.2443893551826477, "learning_rate": 6.336262695085619e-05, "loss": 0.0227, "step": 44320 }, { "grad_norm": 0.2908831536769867, "learning_rate": 6.334669294673612e-05, "loss": 0.0224, "step": 44330 }, { "grad_norm": 0.17795535922050476, "learning_rate": 6.333075748304179e-05, "loss": 0.0197, "step": 44340 }, { "grad_norm": 0.19837774336338043, "learning_rate": 6.331482056151585e-05, "loss": 0.0188, "step": 44350 }, { "grad_norm": 0.2331663817167282, "learning_rate": 6.329888218390117e-05, "loss": 0.0258, "step": 44360 }, { "grad_norm": 0.2558695673942566, "learning_rate": 6.328294235194072e-05, "loss": 0.0246, "step": 44370 }, { "grad_norm": 0.24527224898338318, "learning_rate": 6.326700106737765e-05, "loss": 0.0209, "step": 44380 }, { "grad_norm": 0.20901456475257874, "learning_rate": 6.325105833195531e-05, "loss": 0.0215, "step": 44390 }, { "grad_norm": 0.2720675766468048, "learning_rate": 6.323511414741715e-05, "loss": 0.0204, "step": 44400 }, { "grad_norm": 0.2187551110982895, "learning_rate": 6.321916851550678e-05, "loss": 0.0233, "step": 44410 }, { "grad_norm": 0.20598092675209045, "learning_rate": 6.320322143796806e-05, "loss": 0.0252, "step": 44420 }, { "grad_norm": 0.28277790546417236, "learning_rate": 6.318727291654488e-05, "loss": 0.0229, "step": 44430 }, { "grad_norm": 0.2908828556537628, "learning_rate": 6.317132295298134e-05, "loss": 0.023, "step": 44440 }, { "grad_norm": 0.14921028912067413, "learning_rate": 6.315537154902173e-05, "loss": 0.0194, "step": 44450 }, { "grad_norm": 0.2007327526807785, "learning_rate": 6.313941870641048e-05, "loss": 0.0177, "step": 44460 }, { "grad_norm": 0.24572621285915375, "learning_rate": 6.312346442689214e-05, "loss": 0.0214, "step": 44470 }, { "grad_norm": 0.23817647993564606, "learning_rate": 6.310750871221147e-05, "loss": 0.0226, "step": 44480 }, { "grad_norm": 0.18423640727996826, "learning_rate": 6.309155156411335e-05, "loss": 0.0222, "step": 44490 }, { "grad_norm": 0.19970959424972534, "learning_rate": 6.307559298434284e-05, "loss": 0.0184, "step": 44500 }, { "grad_norm": 0.19563733041286469, "learning_rate": 6.305963297464512e-05, "loss": 0.0197, "step": 44510 }, { "grad_norm": 0.2123512625694275, "learning_rate": 6.304367153676561e-05, "loss": 0.0194, "step": 44520 }, { "grad_norm": 0.23608998954296112, "learning_rate": 6.302770867244978e-05, "loss": 0.0208, "step": 44530 }, { "grad_norm": 0.18964333832263947, "learning_rate": 6.301174438344328e-05, "loss": 0.0197, "step": 44540 }, { "grad_norm": 0.2463638335466385, "learning_rate": 6.299577867149202e-05, "loss": 0.0173, "step": 44550 }, { "grad_norm": 0.17246794700622559, "learning_rate": 6.297981153834192e-05, "loss": 0.0182, "step": 44560 }, { "grad_norm": 0.22855186462402344, "learning_rate": 6.296384298573916e-05, "loss": 0.0165, "step": 44570 }, { "grad_norm": 0.20054419338703156, "learning_rate": 6.294787301543001e-05, "loss": 0.0202, "step": 44580 }, { "grad_norm": 0.2743960916996002, "learning_rate": 6.293190162916095e-05, "loss": 0.028, "step": 44590 }, { "grad_norm": 0.2986934185028076, "learning_rate": 6.291592882867855e-05, "loss": 0.0209, "step": 44600 }, { "grad_norm": 0.2540855407714844, "learning_rate": 6.28999546157296e-05, "loss": 0.019, "step": 44610 }, { "grad_norm": 0.23877061903476715, "learning_rate": 6.288397899206102e-05, "loss": 0.0239, "step": 44620 }, { "grad_norm": 0.15301759541034698, "learning_rate": 6.286800195941984e-05, "loss": 0.0204, "step": 44630 }, { "grad_norm": 0.24043352901935577, "learning_rate": 6.285202351955334e-05, "loss": 0.0221, "step": 44640 }, { "grad_norm": 0.39155885577201843, "learning_rate": 6.283604367420887e-05, "loss": 0.0214, "step": 44650 }, { "grad_norm": 0.1808369755744934, "learning_rate": 6.282006242513394e-05, "loss": 0.0202, "step": 44660 }, { "grad_norm": 0.210830956697464, "learning_rate": 6.280407977407628e-05, "loss": 0.0176, "step": 44670 }, { "grad_norm": 0.28298065066337585, "learning_rate": 6.27880957227837e-05, "loss": 0.0217, "step": 44680 }, { "grad_norm": 0.1824493110179901, "learning_rate": 6.27721102730042e-05, "loss": 0.0228, "step": 44690 }, { "grad_norm": 0.20620980858802795, "learning_rate": 6.275612342648592e-05, "loss": 0.0221, "step": 44700 }, { "grad_norm": 0.16636322438716888, "learning_rate": 6.274013518497716e-05, "loss": 0.0202, "step": 44710 }, { "grad_norm": 0.28497645258903503, "learning_rate": 6.272414555022636e-05, "loss": 0.0199, "step": 44720 }, { "grad_norm": 0.21712546050548553, "learning_rate": 6.270815452398215e-05, "loss": 0.0252, "step": 44730 }, { "grad_norm": 0.16868340969085693, "learning_rate": 6.269216210799326e-05, "loss": 0.0209, "step": 44740 }, { "grad_norm": 0.2636770009994507, "learning_rate": 6.26761683040086e-05, "loss": 0.0199, "step": 44750 }, { "grad_norm": 0.18980221450328827, "learning_rate": 6.266017311377723e-05, "loss": 0.0194, "step": 44760 }, { "grad_norm": 0.19428227841854095, "learning_rate": 6.264417653904839e-05, "loss": 0.0212, "step": 44770 }, { "grad_norm": 0.1900111436843872, "learning_rate": 6.262817858157139e-05, "loss": 0.0202, "step": 44780 }, { "grad_norm": 0.23654143512248993, "learning_rate": 6.261217924309576e-05, "loss": 0.0215, "step": 44790 }, { "grad_norm": 0.22247256338596344, "learning_rate": 6.259617852537118e-05, "loss": 0.0182, "step": 44800 }, { "grad_norm": 0.2169320434331894, "learning_rate": 6.258017643014747e-05, "loss": 0.0206, "step": 44810 }, { "grad_norm": 0.2778957188129425, "learning_rate": 6.256417295917456e-05, "loss": 0.0231, "step": 44820 }, { "grad_norm": 0.22203879058361053, "learning_rate": 6.254816811420258e-05, "loss": 0.0206, "step": 44830 }, { "grad_norm": 0.2211761176586151, "learning_rate": 6.253216189698183e-05, "loss": 0.0189, "step": 44840 }, { "grad_norm": 0.21073687076568604, "learning_rate": 6.251615430926267e-05, "loss": 0.0223, "step": 44850 }, { "grad_norm": 0.18070726096630096, "learning_rate": 6.25001453527957e-05, "loss": 0.0194, "step": 44860 }, { "grad_norm": 0.19734600186347961, "learning_rate": 6.248413502933164e-05, "loss": 0.018, "step": 44870 }, { "grad_norm": 0.21530859172344208, "learning_rate": 6.246812334062133e-05, "loss": 0.0182, "step": 44880 }, { "grad_norm": 0.2817663252353668, "learning_rate": 6.245211028841579e-05, "loss": 0.0182, "step": 44890 }, { "grad_norm": 0.2177758812904358, "learning_rate": 6.24360958744662e-05, "loss": 0.0185, "step": 44900 }, { "grad_norm": 0.18099729716777802, "learning_rate": 6.242008010052387e-05, "loss": 0.0249, "step": 44910 }, { "grad_norm": 0.19993700087070465, "learning_rate": 6.240406296834024e-05, "loss": 0.0251, "step": 44920 }, { "grad_norm": 0.43739402294158936, "learning_rate": 6.238804447966694e-05, "loss": 0.0259, "step": 44930 }, { "grad_norm": 0.19545528292655945, "learning_rate": 6.237202463625573e-05, "loss": 0.0235, "step": 44940 }, { "grad_norm": 0.2873224914073944, "learning_rate": 6.235600343985848e-05, "loss": 0.0254, "step": 44950 }, { "grad_norm": 0.2291148453950882, "learning_rate": 6.233998089222729e-05, "loss": 0.0322, "step": 44960 }, { "grad_norm": 0.22191359102725983, "learning_rate": 6.232395699511433e-05, "loss": 0.0202, "step": 44970 }, { "grad_norm": 0.20700033009052277, "learning_rate": 6.230793175027199e-05, "loss": 0.0216, "step": 44980 }, { "grad_norm": 0.2721405029296875, "learning_rate": 6.22919051594527e-05, "loss": 0.021, "step": 44990 }, { "grad_norm": 0.25346556305885315, "learning_rate": 6.227587722440917e-05, "loss": 0.0183, "step": 45000 }, { "grad_norm": 0.22354647517204285, "learning_rate": 6.225984794689414e-05, "loss": 0.0206, "step": 45010 }, { "grad_norm": 0.17019671201705933, "learning_rate": 6.22438173286606e-05, "loss": 0.0176, "step": 45020 }, { "grad_norm": 0.1417592465877533, "learning_rate": 6.222778537146159e-05, "loss": 0.0185, "step": 45030 }, { "grad_norm": 0.16551151871681213, "learning_rate": 6.221175207705037e-05, "loss": 0.019, "step": 45040 }, { "grad_norm": 0.31370115280151367, "learning_rate": 6.219571744718027e-05, "loss": 0.0171, "step": 45050 }, { "grad_norm": 0.23853197693824768, "learning_rate": 6.217968148360487e-05, "loss": 0.0197, "step": 45060 }, { "grad_norm": 0.2802991569042206, "learning_rate": 6.216364418807782e-05, "loss": 0.0216, "step": 45070 }, { "grad_norm": 0.2715865969657898, "learning_rate": 6.214760556235292e-05, "loss": 0.0208, "step": 45080 }, { "grad_norm": 0.2619873881340027, "learning_rate": 6.213156560818414e-05, "loss": 0.0242, "step": 45090 }, { "grad_norm": 0.2316725105047226, "learning_rate": 6.211552432732559e-05, "loss": 0.0184, "step": 45100 }, { "grad_norm": 0.2338683307170868, "learning_rate": 6.20994817215315e-05, "loss": 0.0199, "step": 45110 }, { "grad_norm": 0.19872045516967773, "learning_rate": 6.208343779255628e-05, "loss": 0.0203, "step": 45120 }, { "grad_norm": 0.20332026481628418, "learning_rate": 6.206739254215449e-05, "loss": 0.0197, "step": 45130 }, { "grad_norm": 0.19161082804203033, "learning_rate": 6.205134597208077e-05, "loss": 0.0196, "step": 45140 }, { "grad_norm": 0.19175110757350922, "learning_rate": 6.203529808409e-05, "loss": 0.0221, "step": 45150 }, { "grad_norm": 0.22194375097751617, "learning_rate": 6.201924887993712e-05, "loss": 0.0238, "step": 45160 }, { "grad_norm": 0.21101412177085876, "learning_rate": 6.200319836137725e-05, "loss": 0.022, "step": 45170 }, { "grad_norm": 0.20137478411197662, "learning_rate": 6.198714653016565e-05, "loss": 0.0189, "step": 45180 }, { "grad_norm": 0.1650865375995636, "learning_rate": 6.197109338805774e-05, "loss": 0.0213, "step": 45190 }, { "grad_norm": 0.16274774074554443, "learning_rate": 6.195503893680903e-05, "loss": 0.02, "step": 45200 }, { "grad_norm": 0.2330704629421234, "learning_rate": 6.193898317817524e-05, "loss": 0.0193, "step": 45210 }, { "grad_norm": 0.29613763093948364, "learning_rate": 6.192292611391221e-05, "loss": 0.0235, "step": 45220 }, { "grad_norm": 0.20015251636505127, "learning_rate": 6.190686774577591e-05, "loss": 0.0228, "step": 45230 }, { "grad_norm": 0.2182372361421585, "learning_rate": 6.189080807552245e-05, "loss": 0.0224, "step": 45240 }, { "grad_norm": 0.21770140528678894, "learning_rate": 6.187474710490809e-05, "loss": 0.0226, "step": 45250 }, { "grad_norm": 0.30197614431381226, "learning_rate": 6.185868483568926e-05, "loss": 0.0205, "step": 45260 }, { "grad_norm": 0.22693143784999847, "learning_rate": 6.184262126962245e-05, "loss": 0.0224, "step": 45270 }, { "grad_norm": 0.262636661529541, "learning_rate": 6.182655640846442e-05, "loss": 0.0237, "step": 45280 }, { "grad_norm": 0.22141514718532562, "learning_rate": 6.181049025397196e-05, "loss": 0.0213, "step": 45290 }, { "grad_norm": 0.2131628841161728, "learning_rate": 6.179442280790202e-05, "loss": 0.0184, "step": 45300 }, { "grad_norm": 0.24478179216384888, "learning_rate": 6.177835407201174e-05, "loss": 0.023, "step": 45310 }, { "grad_norm": 0.2489771544933319, "learning_rate": 6.176228404805839e-05, "loss": 0.0239, "step": 45320 }, { "grad_norm": 0.20796862244606018, "learning_rate": 6.174621273779932e-05, "loss": 0.0186, "step": 45330 }, { "grad_norm": 0.18358558416366577, "learning_rate": 6.17301401429921e-05, "loss": 0.0199, "step": 45340 }, { "grad_norm": 0.24873477220535278, "learning_rate": 6.171406626539436e-05, "loss": 0.0207, "step": 45350 }, { "grad_norm": 0.1562817394733429, "learning_rate": 6.169799110676398e-05, "loss": 0.0219, "step": 45360 }, { "grad_norm": 0.19780495762825012, "learning_rate": 6.168191466885885e-05, "loss": 0.0187, "step": 45370 }, { "grad_norm": 0.23541918396949768, "learning_rate": 6.166583695343711e-05, "loss": 0.02, "step": 45380 }, { "grad_norm": 0.2174205631017685, "learning_rate": 6.164975796225698e-05, "loss": 0.0194, "step": 45390 }, { "grad_norm": 0.20206467807292938, "learning_rate": 6.163367769707683e-05, "loss": 0.0219, "step": 45400 }, { "grad_norm": 0.19400902092456818, "learning_rate": 6.161759615965519e-05, "loss": 0.0231, "step": 45410 }, { "grad_norm": 0.2213224470615387, "learning_rate": 6.16015133517507e-05, "loss": 0.0187, "step": 45420 }, { "grad_norm": 0.35358870029449463, "learning_rate": 6.158542927512214e-05, "loss": 0.0194, "step": 45430 }, { "grad_norm": 0.21949796378612518, "learning_rate": 6.156934393152846e-05, "loss": 0.0229, "step": 45440 }, { "grad_norm": 0.23513907194137573, "learning_rate": 6.15532573227287e-05, "loss": 0.0199, "step": 45450 }, { "grad_norm": 0.24183964729309082, "learning_rate": 6.153716945048212e-05, "loss": 0.0187, "step": 45460 }, { "grad_norm": 0.23898498713970184, "learning_rate": 6.152108031654802e-05, "loss": 0.022, "step": 45470 }, { "grad_norm": 0.16555237770080566, "learning_rate": 6.15049899226859e-05, "loss": 0.019, "step": 45480 }, { "grad_norm": 0.18536022305488586, "learning_rate": 6.148889827065537e-05, "loss": 0.0201, "step": 45490 }, { "grad_norm": 0.21466730535030365, "learning_rate": 6.147280536221622e-05, "loss": 0.0233, "step": 45500 }, { "grad_norm": 0.240334615111351, "learning_rate": 6.145671119912832e-05, "loss": 0.0217, "step": 45510 }, { "grad_norm": 0.15371912717819214, "learning_rate": 6.144061578315169e-05, "loss": 0.0193, "step": 45520 }, { "grad_norm": 0.1824382096529007, "learning_rate": 6.142451911604654e-05, "loss": 0.0199, "step": 45530 }, { "grad_norm": 0.1648166924715042, "learning_rate": 6.140842119957315e-05, "loss": 0.0179, "step": 45540 }, { "grad_norm": 0.2632385492324829, "learning_rate": 6.139232203549197e-05, "loss": 0.0233, "step": 45550 }, { "grad_norm": 0.23995408415794373, "learning_rate": 6.137622162556357e-05, "loss": 0.0228, "step": 45560 }, { "grad_norm": 0.23452050983905792, "learning_rate": 6.13601199715487e-05, "loss": 0.0222, "step": 45570 }, { "grad_norm": 0.18106946349143982, "learning_rate": 6.134401707520816e-05, "loss": 0.0187, "step": 45580 }, { "grad_norm": 0.2295931577682495, "learning_rate": 6.132791293830298e-05, "loss": 0.0171, "step": 45590 }, { "grad_norm": 0.22224073112010956, "learning_rate": 6.131180756259428e-05, "loss": 0.0197, "step": 45600 }, { "grad_norm": 0.21789175271987915, "learning_rate": 6.129570094984331e-05, "loss": 0.0234, "step": 45610 }, { "grad_norm": 0.22771579027175903, "learning_rate": 6.127959310181145e-05, "loss": 0.0233, "step": 45620 }, { "grad_norm": 0.17193709313869476, "learning_rate": 6.126348402026026e-05, "loss": 0.0184, "step": 45630 }, { "grad_norm": 0.17941924929618835, "learning_rate": 6.12473737069514e-05, "loss": 0.0168, "step": 45640 }, { "grad_norm": 0.18207331001758575, "learning_rate": 6.123126216364665e-05, "loss": 0.0197, "step": 45650 }, { "grad_norm": 0.22333072125911713, "learning_rate": 6.121514939210797e-05, "loss": 0.0199, "step": 45660 }, { "grad_norm": 0.19952714443206787, "learning_rate": 6.119903539409741e-05, "loss": 0.0168, "step": 45670 }, { "grad_norm": 0.21225230395793915, "learning_rate": 6.118292017137716e-05, "loss": 0.0175, "step": 45680 }, { "grad_norm": 0.20712296664714813, "learning_rate": 6.116680372570959e-05, "loss": 0.019, "step": 45690 }, { "grad_norm": 0.1894322782754898, "learning_rate": 6.115068605885713e-05, "loss": 0.0183, "step": 45700 }, { "grad_norm": 0.1930975317955017, "learning_rate": 6.113456717258243e-05, "loss": 0.0176, "step": 45710 }, { "grad_norm": 0.202150359749794, "learning_rate": 6.11184470686482e-05, "loss": 0.0177, "step": 45720 }, { "grad_norm": 0.2354465276002884, "learning_rate": 6.110232574881731e-05, "loss": 0.0191, "step": 45730 }, { "grad_norm": 0.18345770239830017, "learning_rate": 6.108620321485277e-05, "loss": 0.016, "step": 45740 }, { "grad_norm": 0.19521164894104004, "learning_rate": 6.107007946851773e-05, "loss": 0.0185, "step": 45750 }, { "grad_norm": 0.20294497907161713, "learning_rate": 6.105395451157542e-05, "loss": 0.0166, "step": 45760 }, { "grad_norm": 0.1812572032213211, "learning_rate": 6.103782834578928e-05, "loss": 0.0182, "step": 45770 }, { "grad_norm": 0.19544468820095062, "learning_rate": 6.102170097292281e-05, "loss": 0.0196, "step": 45780 }, { "grad_norm": 0.19909395277500153, "learning_rate": 6.1005572394739716e-05, "loss": 0.0178, "step": 45790 }, { "grad_norm": 0.1583031713962555, "learning_rate": 6.0989442613003765e-05, "loss": 0.0188, "step": 45800 }, { "grad_norm": 0.21821975708007812, "learning_rate": 6.097331162947888e-05, "loss": 0.0191, "step": 45810 }, { "grad_norm": 0.1948288232088089, "learning_rate": 6.095717944592914e-05, "loss": 0.0173, "step": 45820 }, { "grad_norm": 0.19240938127040863, "learning_rate": 6.094104606411873e-05, "loss": 0.0152, "step": 45830 }, { "grad_norm": 0.16171905398368835, "learning_rate": 6.0924911485811966e-05, "loss": 0.0152, "step": 45840 }, { "grad_norm": 0.2762337625026703, "learning_rate": 6.090877571277331e-05, "loss": 0.018, "step": 45850 }, { "grad_norm": 0.1633891761302948, "learning_rate": 6.089263874676736e-05, "loss": 0.0167, "step": 45860 }, { "grad_norm": 0.22829584777355194, "learning_rate": 6.0876500589558796e-05, "loss": 0.0202, "step": 45870 }, { "grad_norm": 0.19754670560359955, "learning_rate": 6.086036124291248e-05, "loss": 0.0193, "step": 45880 }, { "grad_norm": 0.18049752712249756, "learning_rate": 6.084422070859339e-05, "loss": 0.017, "step": 45890 }, { "grad_norm": 0.22613972425460815, "learning_rate": 6.082807898836663e-05, "loss": 0.0196, "step": 45900 }, { "grad_norm": 0.22779934108257294, "learning_rate": 6.081193608399742e-05, "loss": 0.019, "step": 45910 }, { "grad_norm": 0.20175394415855408, "learning_rate": 6.0795791997251164e-05, "loss": 0.0147, "step": 45920 }, { "grad_norm": 0.19982600212097168, "learning_rate": 6.0779646729893294e-05, "loss": 0.0185, "step": 45930 }, { "grad_norm": 0.18384486436843872, "learning_rate": 6.0763500283689476e-05, "loss": 0.0204, "step": 45940 }, { "grad_norm": 0.20444579422473907, "learning_rate": 6.0747352660405455e-05, "loss": 0.0187, "step": 45950 }, { "grad_norm": 0.22733356058597565, "learning_rate": 6.073120386180709e-05, "loss": 0.0169, "step": 45960 }, { "grad_norm": 0.1965760886669159, "learning_rate": 6.0715053889660425e-05, "loss": 0.0219, "step": 45970 }, { "grad_norm": 0.19381915032863617, "learning_rate": 6.069890274573157e-05, "loss": 0.0186, "step": 45980 }, { "grad_norm": 0.19910822808742523, "learning_rate": 6.068275043178679e-05, "loss": 0.0192, "step": 45990 }, { "grad_norm": 0.18741081655025482, "learning_rate": 6.066659694959248e-05, "loss": 0.0203, "step": 46000 }, { "grad_norm": 0.18606582283973694, "learning_rate": 6.0650442300915176e-05, "loss": 0.02, "step": 46010 }, { "grad_norm": 0.2027377486228943, "learning_rate": 6.063428648752152e-05, "loss": 0.0193, "step": 46020 }, { "grad_norm": 0.4088203012943268, "learning_rate": 6.0618129511178266e-05, "loss": 0.0177, "step": 46030 }, { "grad_norm": 0.18654991686344147, "learning_rate": 6.060197137365234e-05, "loss": 0.0172, "step": 46040 }, { "grad_norm": 0.22910016775131226, "learning_rate": 6.058581207671077e-05, "loss": 0.017, "step": 46050 }, { "grad_norm": 0.21167463064193726, "learning_rate": 6.056965162212072e-05, "loss": 0.0212, "step": 46060 }, { "grad_norm": 0.24308419227600098, "learning_rate": 6.055349001164943e-05, "loss": 0.0208, "step": 46070 }, { "grad_norm": 0.1957593709230423, "learning_rate": 6.0537327247064347e-05, "loss": 0.0218, "step": 46080 }, { "grad_norm": 0.221614807844162, "learning_rate": 6.0521163330133014e-05, "loss": 0.0163, "step": 46090 }, { "grad_norm": 0.20570774376392365, "learning_rate": 6.050499826262306e-05, "loss": 0.0225, "step": 46100 }, { "grad_norm": 0.182676762342453, "learning_rate": 6.0488832046302294e-05, "loss": 0.02, "step": 46110 }, { "grad_norm": 0.21521124243736267, "learning_rate": 6.0472664682938626e-05, "loss": 0.0192, "step": 46120 }, { "grad_norm": 0.2542572617530823, "learning_rate": 6.045649617430009e-05, "loss": 0.0201, "step": 46130 }, { "grad_norm": 0.17651578783988953, "learning_rate": 6.0440326522154866e-05, "loss": 0.0212, "step": 46140 }, { "grad_norm": 0.20760168135166168, "learning_rate": 6.0424155728271224e-05, "loss": 0.0205, "step": 46150 }, { "grad_norm": 0.21526598930358887, "learning_rate": 6.040798379441758e-05, "loss": 0.0195, "step": 46160 }, { "grad_norm": 0.17243073880672455, "learning_rate": 6.0391810722362485e-05, "loss": 0.0173, "step": 46170 }, { "grad_norm": 0.24691206216812134, "learning_rate": 6.037563651387458e-05, "loss": 0.0207, "step": 46180 }, { "grad_norm": 0.2870063781738281, "learning_rate": 6.0359461170722666e-05, "loss": 0.0226, "step": 46190 }, { "grad_norm": 0.2847156226634979, "learning_rate": 6.034328469467566e-05, "loss": 0.0262, "step": 46200 }, { "grad_norm": 0.19907139241695404, "learning_rate": 6.0327107087502596e-05, "loss": 0.0187, "step": 46210 }, { "grad_norm": 0.16317827999591827, "learning_rate": 6.031092835097262e-05, "loss": 0.021, "step": 46220 }, { "grad_norm": 0.20388263463974, "learning_rate": 6.0294748486855024e-05, "loss": 0.0212, "step": 46230 }, { "grad_norm": 0.21262627840042114, "learning_rate": 6.0278567496919216e-05, "loss": 0.021, "step": 46240 }, { "grad_norm": 0.2505829930305481, "learning_rate": 6.026238538293472e-05, "loss": 0.0192, "step": 46250 }, { "grad_norm": 0.2201787680387497, "learning_rate": 6.024620214667118e-05, "loss": 0.0217, "step": 46260 }, { "grad_norm": 0.24605737626552582, "learning_rate": 6.0230017789898384e-05, "loss": 0.0198, "step": 46270 }, { "grad_norm": 0.22104722261428833, "learning_rate": 6.021383231438622e-05, "loss": 0.0192, "step": 46280 }, { "grad_norm": 0.24869157373905182, "learning_rate": 6.0197645721904704e-05, "loss": 0.0203, "step": 46290 }, { "grad_norm": 0.27069172263145447, "learning_rate": 6.0181458014224e-05, "loss": 0.0215, "step": 46300 }, { "grad_norm": 0.21876974403858185, "learning_rate": 6.016526919311435e-05, "loss": 0.0187, "step": 46310 }, { "grad_norm": 0.2331724464893341, "learning_rate": 6.014907926034613e-05, "loss": 0.0207, "step": 46320 }, { "grad_norm": 0.1822429597377777, "learning_rate": 6.0132888217689875e-05, "loss": 0.0176, "step": 46330 }, { "grad_norm": 0.19612979888916016, "learning_rate": 6.01166960669162e-05, "loss": 0.0177, "step": 46340 }, { "grad_norm": 0.27160194516181946, "learning_rate": 6.0100502809795845e-05, "loss": 0.0179, "step": 46350 }, { "grad_norm": 0.18962295353412628, "learning_rate": 6.00843084480997e-05, "loss": 0.0192, "step": 46360 }, { "grad_norm": 0.2204764485359192, "learning_rate": 6.0068112983598736e-05, "loss": 0.0193, "step": 46370 }, { "grad_norm": 0.2046932727098465, "learning_rate": 6.005191641806407e-05, "loss": 0.0217, "step": 46380 }, { "grad_norm": 0.2220669835805893, "learning_rate": 6.003571875326694e-05, "loss": 0.0232, "step": 46390 }, { "grad_norm": 0.24658158421516418, "learning_rate": 6.00195199909787e-05, "loss": 0.0213, "step": 46400 }, { "grad_norm": 0.21703501045703888, "learning_rate": 6.000332013297082e-05, "loss": 0.0184, "step": 46410 }, { "grad_norm": 0.19704005122184753, "learning_rate": 5.998711918101487e-05, "loss": 0.0223, "step": 46420 }, { "grad_norm": 0.2578410506248474, "learning_rate": 5.997091713688261e-05, "loss": 0.0224, "step": 46430 }, { "grad_norm": 0.20969273149967194, "learning_rate": 5.9954714002345836e-05, "loss": 0.0198, "step": 46440 }, { "grad_norm": 0.20172294974327087, "learning_rate": 5.993850977917649e-05, "loss": 0.0173, "step": 46450 }, { "grad_norm": 0.22009404003620148, "learning_rate": 5.992230446914667e-05, "loss": 0.02, "step": 46460 }, { "grad_norm": 0.21101786196231842, "learning_rate": 5.990609807402855e-05, "loss": 0.0183, "step": 46470 }, { "grad_norm": 0.2365584522485733, "learning_rate": 5.988989059559443e-05, "loss": 0.0194, "step": 46480 }, { "grad_norm": 0.200452983379364, "learning_rate": 5.987368203561675e-05, "loss": 0.019, "step": 46490 }, { "grad_norm": 0.19846101105213165, "learning_rate": 5.9857472395868055e-05, "loss": 0.0231, "step": 46500 }, { "grad_norm": 0.23049408197402954, "learning_rate": 5.9841261678120983e-05, "loss": 0.0225, "step": 46510 }, { "grad_norm": 0.16021008789539337, "learning_rate": 5.982504988414834e-05, "loss": 0.0189, "step": 46520 }, { "grad_norm": 0.19434744119644165, "learning_rate": 5.9808837015723015e-05, "loss": 0.024, "step": 46530 }, { "grad_norm": 0.22518706321716309, "learning_rate": 5.9792623074618016e-05, "loss": 0.0219, "step": 46540 }, { "grad_norm": 0.19213300943374634, "learning_rate": 5.977640806260648e-05, "loss": 0.0228, "step": 46550 }, { "grad_norm": 0.1972597986459732, "learning_rate": 5.976019198146166e-05, "loss": 0.0234, "step": 46560 }, { "grad_norm": 0.213586688041687, "learning_rate": 5.974397483295692e-05, "loss": 0.0257, "step": 46570 }, { "grad_norm": 0.26151278614997864, "learning_rate": 5.972775661886575e-05, "loss": 0.0261, "step": 46580 }, { "grad_norm": 0.21693214774131775, "learning_rate": 5.971153734096173e-05, "loss": 0.0192, "step": 46590 }, { "grad_norm": 0.17536325752735138, "learning_rate": 5.969531700101859e-05, "loss": 0.0183, "step": 46600 }, { "grad_norm": 0.2436857521533966, "learning_rate": 5.9679095600810155e-05, "loss": 0.0239, "step": 46610 }, { "grad_norm": 0.2205153852701187, "learning_rate": 5.9662873142110384e-05, "loss": 0.0212, "step": 46620 }, { "grad_norm": 0.26183629035949707, "learning_rate": 5.964664962669333e-05, "loss": 0.021, "step": 46630 }, { "grad_norm": 0.20056982338428497, "learning_rate": 5.9630425056333186e-05, "loss": 0.0191, "step": 46640 }, { "grad_norm": 0.18631914258003235, "learning_rate": 5.961419943280422e-05, "loss": 0.0179, "step": 46650 }, { "grad_norm": 0.15189552307128906, "learning_rate": 5.959797275788087e-05, "loss": 0.0181, "step": 46660 }, { "grad_norm": 0.20358921587467194, "learning_rate": 5.958174503333765e-05, "loss": 0.0193, "step": 46670 }, { "grad_norm": 0.2106599509716034, "learning_rate": 5.9565516260949195e-05, "loss": 0.0179, "step": 46680 }, { "grad_norm": 0.19553470611572266, "learning_rate": 5.9549286442490273e-05, "loss": 0.0178, "step": 46690 }, { "grad_norm": 0.1651957929134369, "learning_rate": 5.953305557973572e-05, "loss": 0.0178, "step": 46700 }, { "grad_norm": 0.255116730928421, "learning_rate": 5.9516823674460564e-05, "loss": 0.0231, "step": 46710 }, { "grad_norm": 0.24690647423267365, "learning_rate": 5.9500590728439875e-05, "loss": 0.0202, "step": 46720 }, { "grad_norm": 0.3490903973579407, "learning_rate": 5.948435674344886e-05, "loss": 0.0195, "step": 46730 }, { "grad_norm": 0.18705375492572784, "learning_rate": 5.946812172126285e-05, "loss": 0.0195, "step": 46740 }, { "grad_norm": 0.20932988822460175, "learning_rate": 5.945188566365729e-05, "loss": 0.0185, "step": 46750 }, { "grad_norm": 0.19091494381427765, "learning_rate": 5.943564857240773e-05, "loss": 0.0208, "step": 46760 }, { "grad_norm": 0.24683675169944763, "learning_rate": 5.941941044928981e-05, "loss": 0.0174, "step": 46770 }, { "grad_norm": 0.18354028463363647, "learning_rate": 5.940317129607935e-05, "loss": 0.0208, "step": 46780 }, { "grad_norm": 0.21906882524490356, "learning_rate": 5.9386931114552204e-05, "loss": 0.0249, "step": 46790 }, { "grad_norm": 0.24113406240940094, "learning_rate": 5.937068990648438e-05, "loss": 0.0229, "step": 46800 }, { "grad_norm": 0.27035030722618103, "learning_rate": 5.935444767365199e-05, "loss": 0.0245, "step": 46810 }, { "grad_norm": 0.21001431345939636, "learning_rate": 5.933820441783129e-05, "loss": 0.0199, "step": 46820 }, { "grad_norm": 0.22147759795188904, "learning_rate": 5.932196014079857e-05, "loss": 0.0179, "step": 46830 }, { "grad_norm": 0.3203933835029602, "learning_rate": 5.930571484433032e-05, "loss": 0.0233, "step": 46840 }, { "grad_norm": 0.21324197947978973, "learning_rate": 5.928946853020309e-05, "loss": 0.02, "step": 46850 }, { "grad_norm": 0.3464517295360565, "learning_rate": 5.927322120019352e-05, "loss": 0.0183, "step": 46860 }, { "grad_norm": 0.280652791261673, "learning_rate": 5.925697285607844e-05, "loss": 0.0192, "step": 46870 }, { "grad_norm": 0.1706182062625885, "learning_rate": 5.924072349963472e-05, "loss": 0.0205, "step": 46880 }, { "grad_norm": 0.21723726391792297, "learning_rate": 5.922447313263938e-05, "loss": 0.019, "step": 46890 }, { "grad_norm": 0.17444033920764923, "learning_rate": 5.920822175686952e-05, "loss": 0.0163, "step": 46900 }, { "grad_norm": 0.24266134202480316, "learning_rate": 5.919196937410239e-05, "loss": 0.0203, "step": 46910 }, { "grad_norm": 0.18207624554634094, "learning_rate": 5.917571598611529e-05, "loss": 0.0216, "step": 46920 }, { "grad_norm": 0.15922574698925018, "learning_rate": 5.9159461594685686e-05, "loss": 0.0178, "step": 46930 }, { "grad_norm": 0.18687260150909424, "learning_rate": 5.914320620159114e-05, "loss": 0.0211, "step": 46940 }, { "grad_norm": 0.20962558686733246, "learning_rate": 5.912694980860931e-05, "loss": 0.0204, "step": 46950 }, { "grad_norm": 0.23262964189052582, "learning_rate": 5.9110692417517964e-05, "loss": 0.0193, "step": 46960 }, { "grad_norm": 0.18701040744781494, "learning_rate": 5.9094434030095e-05, "loss": 0.0179, "step": 46970 }, { "grad_norm": 0.21029086410999298, "learning_rate": 5.907817464811841e-05, "loss": 0.0172, "step": 46980 }, { "grad_norm": 0.20090708136558533, "learning_rate": 5.906191427336627e-05, "loss": 0.0202, "step": 46990 }, { "grad_norm": 0.21965955197811127, "learning_rate": 5.9045652907616835e-05, "loss": 0.0187, "step": 47000 }, { "grad_norm": 0.2024742066860199, "learning_rate": 5.902939055264838e-05, "loss": 0.0185, "step": 47010 }, { "grad_norm": 0.19136029481887817, "learning_rate": 5.901312721023935e-05, "loss": 0.0199, "step": 47020 }, { "grad_norm": 0.2589499056339264, "learning_rate": 5.8996862882168294e-05, "loss": 0.0258, "step": 47030 }, { "grad_norm": 0.2676827609539032, "learning_rate": 5.8980597570213826e-05, "loss": 0.0196, "step": 47040 }, { "grad_norm": 0.2883603572845459, "learning_rate": 5.896433127615471e-05, "loss": 0.0199, "step": 47050 }, { "grad_norm": 0.2586670219898224, "learning_rate": 5.894806400176981e-05, "loss": 0.0192, "step": 47060 }, { "grad_norm": 0.24331575632095337, "learning_rate": 5.893179574883808e-05, "loss": 0.0167, "step": 47070 }, { "grad_norm": 0.41573190689086914, "learning_rate": 5.8915526519138585e-05, "loss": 0.0188, "step": 47080 }, { "grad_norm": 0.2083490788936615, "learning_rate": 5.889925631445053e-05, "loss": 0.0216, "step": 47090 }, { "grad_norm": 0.13938280940055847, "learning_rate": 5.888298513655318e-05, "loss": 0.0194, "step": 47100 }, { "grad_norm": 0.23071803152561188, "learning_rate": 5.886671298722592e-05, "loss": 0.0184, "step": 47110 }, { "grad_norm": 0.161577045917511, "learning_rate": 5.885043986824828e-05, "loss": 0.0174, "step": 47120 }, { "grad_norm": 0.19726553559303284, "learning_rate": 5.883416578139982e-05, "loss": 0.0175, "step": 47130 }, { "grad_norm": 0.17710478603839874, "learning_rate": 5.881789072846029e-05, "loss": 0.0208, "step": 47140 }, { "grad_norm": 0.15238021314144135, "learning_rate": 5.8801614711209474e-05, "loss": 0.0175, "step": 47150 }, { "grad_norm": 0.20176678895950317, "learning_rate": 5.878533773142732e-05, "loss": 0.0175, "step": 47160 }, { "grad_norm": 0.20761731266975403, "learning_rate": 5.8769059790893834e-05, "loss": 0.0211, "step": 47170 }, { "grad_norm": 0.31147995591163635, "learning_rate": 5.8752780891389145e-05, "loss": 0.0186, "step": 47180 }, { "grad_norm": 0.16893555223941803, "learning_rate": 5.873650103469351e-05, "loss": 0.0193, "step": 47190 }, { "grad_norm": 0.20393115282058716, "learning_rate": 5.872022022258725e-05, "loss": 0.0186, "step": 47200 }, { "grad_norm": 0.21331366896629333, "learning_rate": 5.8703938456850814e-05, "loss": 0.0175, "step": 47210 }, { "grad_norm": 0.24955807626247406, "learning_rate": 5.868765573926477e-05, "loss": 0.0209, "step": 47220 }, { "grad_norm": 0.2068895697593689, "learning_rate": 5.867137207160974e-05, "loss": 0.0238, "step": 47230 }, { "grad_norm": 0.18306158483028412, "learning_rate": 5.86550874556665e-05, "loss": 0.02, "step": 47240 }, { "grad_norm": 0.2832311689853668, "learning_rate": 5.863880189321592e-05, "loss": 0.0188, "step": 47250 }, { "grad_norm": 0.21948423981666565, "learning_rate": 5.862251538603894e-05, "loss": 0.0184, "step": 47260 }, { "grad_norm": 0.17989087104797363, "learning_rate": 5.8606227935916656e-05, "loss": 0.0196, "step": 47270 }, { "grad_norm": 0.18717817962169647, "learning_rate": 5.858993954463021e-05, "loss": 0.0189, "step": 47280 }, { "grad_norm": 0.16164548695087433, "learning_rate": 5.8573650213960896e-05, "loss": 0.0206, "step": 47290 }, { "grad_norm": 0.2490403652191162, "learning_rate": 5.8557359945690084e-05, "loss": 0.0221, "step": 47300 }, { "grad_norm": 0.19308729469776154, "learning_rate": 5.8541068741599246e-05, "loss": 0.0191, "step": 47310 }, { "grad_norm": 0.20431840419769287, "learning_rate": 5.8524776603469985e-05, "loss": 0.0221, "step": 47320 }, { "grad_norm": 0.1723218709230423, "learning_rate": 5.850848353308397e-05, "loss": 0.0189, "step": 47330 }, { "grad_norm": 0.2076643407344818, "learning_rate": 5.849218953222297e-05, "loss": 0.0207, "step": 47340 }, { "grad_norm": 0.17571982741355896, "learning_rate": 5.847589460266891e-05, "loss": 0.0201, "step": 47350 }, { "grad_norm": 0.17473722994327545, "learning_rate": 5.845959874620375e-05, "loss": 0.0211, "step": 47360 }, { "grad_norm": 0.18826037645339966, "learning_rate": 5.84433019646096e-05, "loss": 0.0217, "step": 47370 }, { "grad_norm": 0.227006733417511, "learning_rate": 5.842700425966863e-05, "loss": 0.0189, "step": 47380 }, { "grad_norm": 0.19165465235710144, "learning_rate": 5.841070563316315e-05, "loss": 0.0191, "step": 47390 }, { "grad_norm": 0.17258958518505096, "learning_rate": 5.839440608687554e-05, "loss": 0.0193, "step": 47400 }, { "grad_norm": 0.31046512722969055, "learning_rate": 5.837810562258831e-05, "loss": 0.0197, "step": 47410 }, { "grad_norm": 0.25406381487846375, "learning_rate": 5.836180424208405e-05, "loss": 0.0177, "step": 47420 }, { "grad_norm": 0.20159952342510223, "learning_rate": 5.834550194714543e-05, "loss": 0.0187, "step": 47430 }, { "grad_norm": 0.22133618593215942, "learning_rate": 5.832919873955528e-05, "loss": 0.0175, "step": 47440 }, { "grad_norm": 0.22691717743873596, "learning_rate": 5.831289462109648e-05, "loss": 0.0189, "step": 47450 }, { "grad_norm": 0.2244863659143448, "learning_rate": 5.829658959355202e-05, "loss": 0.021, "step": 47460 }, { "grad_norm": 0.17636710405349731, "learning_rate": 5.8280283658704995e-05, "loss": 0.0168, "step": 47470 }, { "grad_norm": 0.16966292262077332, "learning_rate": 5.826397681833859e-05, "loss": 0.0227, "step": 47480 }, { "grad_norm": 0.18771392107009888, "learning_rate": 5.824766907423612e-05, "loss": 0.022, "step": 47490 }, { "grad_norm": 0.2861937880516052, "learning_rate": 5.8231360428180945e-05, "loss": 0.0183, "step": 47500 }, { "grad_norm": 0.5022716522216797, "learning_rate": 5.821505088195658e-05, "loss": 0.0225, "step": 47510 }, { "grad_norm": 0.21949148178100586, "learning_rate": 5.819874043734661e-05, "loss": 0.0186, "step": 47520 }, { "grad_norm": 0.18389494717121124, "learning_rate": 5.8182429096134695e-05, "loss": 0.019, "step": 47530 }, { "grad_norm": 0.21193717420101166, "learning_rate": 5.816611686010465e-05, "loss": 0.0192, "step": 47540 }, { "grad_norm": 0.21354304254055023, "learning_rate": 5.814980373104033e-05, "loss": 0.0165, "step": 47550 }, { "grad_norm": 0.16325822472572327, "learning_rate": 5.813348971072572e-05, "loss": 0.0194, "step": 47560 }, { "grad_norm": 0.17918136715888977, "learning_rate": 5.811717480094492e-05, "loss": 0.0205, "step": 47570 }, { "grad_norm": 0.16122330725193024, "learning_rate": 5.810085900348209e-05, "loss": 0.0158, "step": 47580 }, { "grad_norm": 0.24059529602527618, "learning_rate": 5.8084542320121483e-05, "loss": 0.0182, "step": 47590 }, { "grad_norm": 0.2346045970916748, "learning_rate": 5.8068224752647497e-05, "loss": 0.0179, "step": 47600 }, { "grad_norm": 0.18590644001960754, "learning_rate": 5.805190630284457e-05, "loss": 0.0189, "step": 47610 }, { "grad_norm": 0.20370054244995117, "learning_rate": 5.803558697249729e-05, "loss": 0.0182, "step": 47620 }, { "grad_norm": 0.21076519787311554, "learning_rate": 5.8019266763390276e-05, "loss": 0.0161, "step": 47630 }, { "grad_norm": 0.27712374925613403, "learning_rate": 5.800294567730833e-05, "loss": 0.0206, "step": 47640 }, { "grad_norm": 0.15738551318645477, "learning_rate": 5.798662371603628e-05, "loss": 0.0196, "step": 47650 }, { "grad_norm": 0.21745331585407257, "learning_rate": 5.797030088135904e-05, "loss": 0.0169, "step": 47660 }, { "grad_norm": 0.20811189711093903, "learning_rate": 5.7953977175061705e-05, "loss": 0.0183, "step": 47670 }, { "grad_norm": 0.1671813428401947, "learning_rate": 5.793765259892938e-05, "loss": 0.0175, "step": 47680 }, { "grad_norm": 0.21285322308540344, "learning_rate": 5.792132715474729e-05, "loss": 0.0231, "step": 47690 }, { "grad_norm": 0.17468108236789703, "learning_rate": 5.790500084430078e-05, "loss": 0.0199, "step": 47700 }, { "grad_norm": 0.1947506219148636, "learning_rate": 5.7888673669375274e-05, "loss": 0.0165, "step": 47710 }, { "grad_norm": 0.192851260304451, "learning_rate": 5.787234563175625e-05, "loss": 0.0187, "step": 47720 }, { "grad_norm": 0.1826542317867279, "learning_rate": 5.7856016733229355e-05, "loss": 0.0201, "step": 47730 }, { "grad_norm": 0.22817499935626984, "learning_rate": 5.7839686975580297e-05, "loss": 0.0194, "step": 47740 }, { "grad_norm": 0.20592531561851501, "learning_rate": 5.782335636059484e-05, "loss": 0.0191, "step": 47750 }, { "grad_norm": 0.19800081849098206, "learning_rate": 5.780702489005889e-05, "loss": 0.0179, "step": 47760 }, { "grad_norm": 0.1772828847169876, "learning_rate": 5.779069256575845e-05, "loss": 0.0206, "step": 47770 }, { "grad_norm": 0.19672949612140656, "learning_rate": 5.7774359389479574e-05, "loss": 0.0205, "step": 47780 }, { "grad_norm": 0.21276521682739258, "learning_rate": 5.775802536300845e-05, "loss": 0.0172, "step": 47790 }, { "grad_norm": 0.18015840649604797, "learning_rate": 5.774169048813134e-05, "loss": 0.0205, "step": 47800 }, { "grad_norm": 0.25200605392456055, "learning_rate": 5.77253547666346e-05, "loss": 0.0207, "step": 47810 }, { "grad_norm": 0.20096361637115479, "learning_rate": 5.770901820030465e-05, "loss": 0.0173, "step": 47820 }, { "grad_norm": 0.1884939968585968, "learning_rate": 5.769268079092809e-05, "loss": 0.018, "step": 47830 }, { "grad_norm": 0.20164179801940918, "learning_rate": 5.767634254029151e-05, "loss": 0.0177, "step": 47840 }, { "grad_norm": 0.2168823778629303, "learning_rate": 5.7660003450181655e-05, "loss": 0.0223, "step": 47850 }, { "grad_norm": 0.14473120868206024, "learning_rate": 5.764366352238534e-05, "loss": 0.0185, "step": 47860 }, { "grad_norm": 0.2825842797756195, "learning_rate": 5.7627322758689474e-05, "loss": 0.0184, "step": 47870 }, { "grad_norm": 0.23623046278953552, "learning_rate": 5.761098116088105e-05, "loss": 0.0209, "step": 47880 }, { "grad_norm": 0.19544953107833862, "learning_rate": 5.759463873074717e-05, "loss": 0.0161, "step": 47890 }, { "grad_norm": 0.19797013700008392, "learning_rate": 5.757829547007504e-05, "loss": 0.017, "step": 47900 }, { "grad_norm": 0.22430169582366943, "learning_rate": 5.756195138065189e-05, "loss": 0.021, "step": 47910 }, { "grad_norm": 0.16803064942359924, "learning_rate": 5.754560646426511e-05, "loss": 0.0196, "step": 47920 }, { "grad_norm": 0.26190492510795593, "learning_rate": 5.752926072270216e-05, "loss": 0.0188, "step": 47930 }, { "grad_norm": 0.24711647629737854, "learning_rate": 5.7512914157750563e-05, "loss": 0.0231, "step": 47940 }, { "grad_norm": 0.2437174916267395, "learning_rate": 5.749656677119798e-05, "loss": 0.0232, "step": 47950 }, { "grad_norm": 0.25416937470436096, "learning_rate": 5.7480218564832125e-05, "loss": 0.0214, "step": 47960 }, { "grad_norm": 0.21725520491600037, "learning_rate": 5.746386954044082e-05, "loss": 0.0186, "step": 47970 }, { "grad_norm": 0.20882967114448547, "learning_rate": 5.744751969981195e-05, "loss": 0.0203, "step": 47980 }, { "grad_norm": 0.17057284712791443, "learning_rate": 5.7431169044733526e-05, "loss": 0.0178, "step": 47990 }, { "grad_norm": 0.16895171999931335, "learning_rate": 5.741481757699364e-05, "loss": 0.0184, "step": 48000 }, { "grad_norm": 0.18645267188549042, "learning_rate": 5.7398465298380434e-05, "loss": 0.0193, "step": 48010 }, { "grad_norm": 0.13987573981285095, "learning_rate": 5.7382112210682193e-05, "loss": 0.0184, "step": 48020 }, { "grad_norm": 0.1868358850479126, "learning_rate": 5.7365758315687266e-05, "loss": 0.0175, "step": 48030 }, { "grad_norm": 0.19513152539730072, "learning_rate": 5.734940361518407e-05, "loss": 0.0198, "step": 48040 }, { "grad_norm": 0.2804327607154846, "learning_rate": 5.733304811096116e-05, "loss": 0.0208, "step": 48050 }, { "grad_norm": 0.24082684516906738, "learning_rate": 5.731669180480713e-05, "loss": 0.02, "step": 48060 }, { "grad_norm": 0.3560004234313965, "learning_rate": 5.730033469851067e-05, "loss": 0.0208, "step": 48070 }, { "grad_norm": 0.2086528241634369, "learning_rate": 5.7283976793860607e-05, "loss": 0.0228, "step": 48080 }, { "grad_norm": 0.23637759685516357, "learning_rate": 5.726761809264577e-05, "loss": 0.0188, "step": 48090 }, { "grad_norm": 0.23365555703639984, "learning_rate": 5.7251258596655155e-05, "loss": 0.0205, "step": 48100 }, { "grad_norm": 0.21825304627418518, "learning_rate": 5.72348983076778e-05, "loss": 0.0197, "step": 48110 }, { "grad_norm": 0.19242163002490997, "learning_rate": 5.7218537227502854e-05, "loss": 0.0185, "step": 48120 }, { "grad_norm": 0.24715200066566467, "learning_rate": 5.720217535791951e-05, "loss": 0.0188, "step": 48130 }, { "grad_norm": 0.19721771776676178, "learning_rate": 5.718581270071711e-05, "loss": 0.0208, "step": 48140 }, { "grad_norm": 0.16793204843997955, "learning_rate": 5.716944925768505e-05, "loss": 0.0199, "step": 48150 }, { "grad_norm": 0.2066030651330948, "learning_rate": 5.7153085030612786e-05, "loss": 0.0184, "step": 48160 }, { "grad_norm": 0.23478110134601593, "learning_rate": 5.713672002128989e-05, "loss": 0.0206, "step": 48170 }, { "grad_norm": 0.15946681797504425, "learning_rate": 5.712035423150602e-05, "loss": 0.0195, "step": 48180 }, { "grad_norm": 0.155979186296463, "learning_rate": 5.710398766305094e-05, "loss": 0.0181, "step": 48190 }, { "grad_norm": 0.16955460608005524, "learning_rate": 5.708762031771442e-05, "loss": 0.0159, "step": 48200 }, { "grad_norm": 0.1511451005935669, "learning_rate": 5.707125219728642e-05, "loss": 0.0178, "step": 48210 }, { "grad_norm": 0.18752050399780273, "learning_rate": 5.7054883303556905e-05, "loss": 0.0172, "step": 48220 }, { "grad_norm": 0.3738778829574585, "learning_rate": 5.703851363831595e-05, "loss": 0.0224, "step": 48230 }, { "grad_norm": 0.2338312715291977, "learning_rate": 5.702214320335374e-05, "loss": 0.0179, "step": 48240 }, { "grad_norm": 0.18515470623970032, "learning_rate": 5.7005772000460514e-05, "loss": 0.018, "step": 48250 }, { "grad_norm": 0.20402884483337402, "learning_rate": 5.698940003142656e-05, "loss": 0.0154, "step": 48260 }, { "grad_norm": 0.23992572724819183, "learning_rate": 5.697302729804236e-05, "loss": 0.0208, "step": 48270 }, { "grad_norm": 0.20379303395748138, "learning_rate": 5.695665380209837e-05, "loss": 0.018, "step": 48280 }, { "grad_norm": 0.2277950793504715, "learning_rate": 5.6940279545385165e-05, "loss": 0.0204, "step": 48290 }, { "grad_norm": 0.31527018547058105, "learning_rate": 5.692390452969344e-05, "loss": 0.0207, "step": 48300 }, { "grad_norm": 0.24971185624599457, "learning_rate": 5.690752875681392e-05, "loss": 0.0177, "step": 48310 }, { "grad_norm": 0.1616293042898178, "learning_rate": 5.6891152228537435e-05, "loss": 0.0195, "step": 48320 }, { "grad_norm": 0.21973057091236115, "learning_rate": 5.687477494665492e-05, "loss": 0.022, "step": 48330 }, { "grad_norm": 0.22255510091781616, "learning_rate": 5.685839691295733e-05, "loss": 0.0178, "step": 48340 }, { "grad_norm": 0.18898233771324158, "learning_rate": 5.6842018129235786e-05, "loss": 0.0172, "step": 48350 }, { "grad_norm": 0.14768365025520325, "learning_rate": 5.6825638597281404e-05, "loss": 0.0195, "step": 48360 }, { "grad_norm": 0.17086131870746613, "learning_rate": 5.680925831888546e-05, "loss": 0.0192, "step": 48370 }, { "grad_norm": 0.18753495812416077, "learning_rate": 5.6792877295839274e-05, "loss": 0.017, "step": 48380 }, { "grad_norm": 0.18838874995708466, "learning_rate": 5.6776495529934224e-05, "loss": 0.0189, "step": 48390 }, { "grad_norm": 0.20280617475509644, "learning_rate": 5.6760113022961824e-05, "loss": 0.0163, "step": 48400 }, { "grad_norm": 0.22506211698055267, "learning_rate": 5.6743729776713617e-05, "loss": 0.0202, "step": 48410 }, { "grad_norm": 0.23167011141777039, "learning_rate": 5.672734579298126e-05, "loss": 0.0203, "step": 48420 }, { "grad_norm": 0.2562592029571533, "learning_rate": 5.671096107355649e-05, "loss": 0.0196, "step": 48430 }, { "grad_norm": 0.23231446743011475, "learning_rate": 5.66945756202311e-05, "loss": 0.0215, "step": 48440 }, { "grad_norm": 0.20114409923553467, "learning_rate": 5.667818943479699e-05, "loss": 0.0153, "step": 48450 }, { "grad_norm": 0.20515723526477814, "learning_rate": 5.666180251904612e-05, "loss": 0.0164, "step": 48460 }, { "grad_norm": 0.19823288917541504, "learning_rate": 5.6645414874770555e-05, "loss": 0.0181, "step": 48470 }, { "grad_norm": 0.1838126927614212, "learning_rate": 5.66290265037624e-05, "loss": 0.019, "step": 48480 }, { "grad_norm": 0.19773735105991364, "learning_rate": 5.661263740781386e-05, "loss": 0.0162, "step": 48490 }, { "grad_norm": 0.1818491369485855, "learning_rate": 5.6596247588717254e-05, "loss": 0.019, "step": 48500 }, { "grad_norm": 0.18589025735855103, "learning_rate": 5.6579857048264926e-05, "loss": 0.0195, "step": 48510 }, { "grad_norm": 0.227531298995018, "learning_rate": 5.6563465788249314e-05, "loss": 0.0217, "step": 48520 }, { "grad_norm": 0.1912023425102234, "learning_rate": 5.6547073810462956e-05, "loss": 0.0188, "step": 48530 }, { "grad_norm": 0.18288971483707428, "learning_rate": 5.653068111669846e-05, "loss": 0.0158, "step": 48540 }, { "grad_norm": 0.25757676362991333, "learning_rate": 5.651428770874848e-05, "loss": 0.021, "step": 48550 }, { "grad_norm": 0.35900887846946716, "learning_rate": 5.64978935884058e-05, "loss": 0.0183, "step": 48560 }, { "grad_norm": 0.23226672410964966, "learning_rate": 5.6481498757463244e-05, "loss": 0.0183, "step": 48570 }, { "grad_norm": 0.1650865375995636, "learning_rate": 5.646510321771373e-05, "loss": 0.0178, "step": 48580 }, { "grad_norm": 0.15900999307632446, "learning_rate": 5.644870697095024e-05, "loss": 0.0179, "step": 48590 }, { "grad_norm": 0.17529703676700592, "learning_rate": 5.643231001896586e-05, "loss": 0.0186, "step": 48600 }, { "grad_norm": 0.15188345313072205, "learning_rate": 5.6415912363553726e-05, "loss": 0.018, "step": 48610 }, { "grad_norm": 0.3012668490409851, "learning_rate": 5.639951400650706e-05, "loss": 0.0191, "step": 48620 }, { "grad_norm": 0.21635670959949493, "learning_rate": 5.6383114949619165e-05, "loss": 0.0177, "step": 48630 }, { "grad_norm": 0.25187626481056213, "learning_rate": 5.636671519468342e-05, "loss": 0.0196, "step": 48640 }, { "grad_norm": 0.23625677824020386, "learning_rate": 5.635031474349327e-05, "loss": 0.0186, "step": 48650 }, { "grad_norm": 0.22567251324653625, "learning_rate": 5.6333913597842246e-05, "loss": 0.0174, "step": 48660 }, { "grad_norm": 0.21013198792934418, "learning_rate": 5.6317511759523955e-05, "loss": 0.0205, "step": 48670 }, { "grad_norm": 0.15876458585262299, "learning_rate": 5.630110923033207e-05, "loss": 0.0193, "step": 48680 }, { "grad_norm": 0.20065143704414368, "learning_rate": 5.628470601206036e-05, "loss": 0.0232, "step": 48690 }, { "grad_norm": 0.24188870191574097, "learning_rate": 5.626830210650266e-05, "loss": 0.0208, "step": 48700 }, { "grad_norm": 0.22099144756793976, "learning_rate": 5.625189751545285e-05, "loss": 0.0194, "step": 48710 }, { "grad_norm": 0.2049303650856018, "learning_rate": 5.6235492240704936e-05, "loss": 0.0177, "step": 48720 }, { "grad_norm": 0.17944316565990448, "learning_rate": 5.621908628405296e-05, "loss": 0.0208, "step": 48730 }, { "grad_norm": 0.25333020091056824, "learning_rate": 5.620267964729106e-05, "loss": 0.0186, "step": 48740 }, { "grad_norm": 0.1678200662136078, "learning_rate": 5.618627233221344e-05, "loss": 0.0185, "step": 48750 }, { "grad_norm": 0.17634378373622894, "learning_rate": 5.616986434061438e-05, "loss": 0.0195, "step": 48760 }, { "grad_norm": 0.22203929722309113, "learning_rate": 5.615345567428822e-05, "loss": 0.0193, "step": 48770 }, { "grad_norm": 0.21059885621070862, "learning_rate": 5.613704633502941e-05, "loss": 0.0182, "step": 48780 }, { "grad_norm": 0.2064809650182724, "learning_rate": 5.612063632463245e-05, "loss": 0.0217, "step": 48790 }, { "grad_norm": 0.1979590654373169, "learning_rate": 5.610422564489188e-05, "loss": 0.0198, "step": 48800 }, { "grad_norm": 0.2552591562271118, "learning_rate": 5.608781429760239e-05, "loss": 0.0193, "step": 48810 }, { "grad_norm": 0.2582206726074219, "learning_rate": 5.607140228455866e-05, "loss": 0.0216, "step": 48820 }, { "grad_norm": 0.198747456073761, "learning_rate": 5.605498960755553e-05, "loss": 0.019, "step": 48830 }, { "grad_norm": 0.22026054561138153, "learning_rate": 5.603857626838782e-05, "loss": 0.0197, "step": 48840 }, { "grad_norm": 0.22998669743537903, "learning_rate": 5.60221622688505e-05, "loss": 0.0179, "step": 48850 }, { "grad_norm": 0.15466895699501038, "learning_rate": 5.6005747610738565e-05, "loss": 0.0164, "step": 48860 }, { "grad_norm": 0.13529545068740845, "learning_rate": 5.59893322958471e-05, "loss": 0.0164, "step": 48870 }, { "grad_norm": 0.22951455414295197, "learning_rate": 5.5972916325971256e-05, "loss": 0.0186, "step": 48880 }, { "grad_norm": 0.18462321162223816, "learning_rate": 5.595649970290628e-05, "loss": 0.0166, "step": 48890 }, { "grad_norm": 0.22896088659763336, "learning_rate": 5.5940082428447426e-05, "loss": 0.0178, "step": 48900 }, { "grad_norm": 0.23864132165908813, "learning_rate": 5.592366450439012e-05, "loss": 0.0218, "step": 48910 }, { "grad_norm": 0.33398154377937317, "learning_rate": 5.590724593252975e-05, "loss": 0.0178, "step": 48920 }, { "grad_norm": 0.22011315822601318, "learning_rate": 5.589082671466184e-05, "loss": 0.0184, "step": 48930 }, { "grad_norm": 0.2620839476585388, "learning_rate": 5.587440685258199e-05, "loss": 0.0164, "step": 48940 }, { "grad_norm": 0.16640283167362213, "learning_rate": 5.585798634808583e-05, "loss": 0.0187, "step": 48950 }, { "grad_norm": 0.2050129622220993, "learning_rate": 5.584156520296909e-05, "loss": 0.0204, "step": 48960 }, { "grad_norm": 0.24309219419956207, "learning_rate": 5.582514341902757e-05, "loss": 0.0187, "step": 48970 }, { "grad_norm": 0.1915641874074936, "learning_rate": 5.580872099805713e-05, "loss": 0.016, "step": 48980 }, { "grad_norm": 0.15162600576877594, "learning_rate": 5.5792297941853674e-05, "loss": 0.0165, "step": 48990 }, { "grad_norm": 0.2361467480659485, "learning_rate": 5.5775874252213247e-05, "loss": 0.0185, "step": 49000 }, { "grad_norm": 0.20222985744476318, "learning_rate": 5.575944993093189e-05, "loss": 0.0184, "step": 49010 }, { "grad_norm": 0.20339183509349823, "learning_rate": 5.574302497980574e-05, "loss": 0.0161, "step": 49020 }, { "grad_norm": 0.2583071291446686, "learning_rate": 5.5726599400631005e-05, "loss": 0.0169, "step": 49030 }, { "grad_norm": 0.18649578094482422, "learning_rate": 5.571017319520399e-05, "loss": 0.016, "step": 49040 }, { "grad_norm": 0.2309008240699768, "learning_rate": 5.5693746365320984e-05, "loss": 0.0201, "step": 49050 }, { "grad_norm": 0.261712908744812, "learning_rate": 5.567731891277848e-05, "loss": 0.0194, "step": 49060 }, { "grad_norm": 0.22199788689613342, "learning_rate": 5.566089083937288e-05, "loss": 0.0221, "step": 49070 }, { "grad_norm": 0.16869710385799408, "learning_rate": 5.564446214690079e-05, "loss": 0.0184, "step": 49080 }, { "grad_norm": 0.2243313044309616, "learning_rate": 5.56280328371588e-05, "loss": 0.0166, "step": 49090 }, { "grad_norm": 0.26692843437194824, "learning_rate": 5.5611602911943596e-05, "loss": 0.017, "step": 49100 }, { "grad_norm": 0.21344493329524994, "learning_rate": 5.559517237305194e-05, "loss": 0.0178, "step": 49110 }, { "grad_norm": 0.19057516753673553, "learning_rate": 5.557874122228064e-05, "loss": 0.0169, "step": 49120 }, { "grad_norm": 0.22414329648017883, "learning_rate": 5.55623094614266e-05, "loss": 0.0181, "step": 49130 }, { "grad_norm": 0.19135582447052002, "learning_rate": 5.5545877092286743e-05, "loss": 0.0183, "step": 49140 }, { "grad_norm": 0.20626670122146606, "learning_rate": 5.552944411665811e-05, "loss": 0.0189, "step": 49150 }, { "grad_norm": 0.2538893222808838, "learning_rate": 5.5513010536337795e-05, "loss": 0.0202, "step": 49160 }, { "grad_norm": 0.23277659714221954, "learning_rate": 5.5496576353122934e-05, "loss": 0.0214, "step": 49170 }, { "grad_norm": 0.1932583600282669, "learning_rate": 5.548014156881074e-05, "loss": 0.0193, "step": 49180 }, { "grad_norm": 0.17783014476299286, "learning_rate": 5.5463706185198494e-05, "loss": 0.0202, "step": 49190 }, { "grad_norm": 0.15318740904331207, "learning_rate": 5.544727020408358e-05, "loss": 0.0176, "step": 49200 }, { "grad_norm": 0.25127241015434265, "learning_rate": 5.5430833627263366e-05, "loss": 0.0194, "step": 49210 }, { "grad_norm": 0.18314974009990692, "learning_rate": 5.5414396456535354e-05, "loss": 0.018, "step": 49220 }, { "grad_norm": 0.19328804314136505, "learning_rate": 5.539795869369709e-05, "loss": 0.0178, "step": 49230 }, { "grad_norm": 0.25751012563705444, "learning_rate": 5.538152034054618e-05, "loss": 0.0185, "step": 49240 }, { "grad_norm": 0.19116300344467163, "learning_rate": 5.536508139888028e-05, "loss": 0.0212, "step": 49250 }, { "grad_norm": 0.18329168856143951, "learning_rate": 5.534864187049716e-05, "loss": 0.0171, "step": 49260 }, { "grad_norm": 0.19986961781978607, "learning_rate": 5.5332201757194615e-05, "loss": 0.0195, "step": 49270 }, { "grad_norm": 0.2202739715576172, "learning_rate": 5.531576106077049e-05, "loss": 0.0202, "step": 49280 }, { "grad_norm": 0.25800973176956177, "learning_rate": 5.529931978302272e-05, "loss": 0.0234, "step": 49290 }, { "grad_norm": 0.22942118346691132, "learning_rate": 5.528287792574932e-05, "loss": 0.015, "step": 49300 }, { "grad_norm": 0.19912192225456238, "learning_rate": 5.5266435490748324e-05, "loss": 0.0208, "step": 49310 }, { "grad_norm": 0.2022780030965805, "learning_rate": 5.524999247981787e-05, "loss": 0.0167, "step": 49320 }, { "grad_norm": 0.24010266363620758, "learning_rate": 5.523354889475613e-05, "loss": 0.0164, "step": 49330 }, { "grad_norm": 0.21205931901931763, "learning_rate": 5.521710473736134e-05, "loss": 0.02, "step": 49340 }, { "grad_norm": 0.1872328370809555, "learning_rate": 5.5200660009431826e-05, "loss": 0.0165, "step": 49350 }, { "grad_norm": 0.18043716251850128, "learning_rate": 5.518421471276596e-05, "loss": 0.0163, "step": 49360 }, { "grad_norm": 0.2334698587656021, "learning_rate": 5.5167768849162174e-05, "loss": 0.0189, "step": 49370 }, { "grad_norm": 0.22535696625709534, "learning_rate": 5.515132242041893e-05, "loss": 0.0165, "step": 49380 }, { "grad_norm": 0.22403551638126373, "learning_rate": 5.513487542833483e-05, "loss": 0.0191, "step": 49390 }, { "grad_norm": 0.22754406929016113, "learning_rate": 5.5118427874708466e-05, "loss": 0.0213, "step": 49400 }, { "grad_norm": 0.24893011152744293, "learning_rate": 5.510197976133853e-05, "loss": 0.0174, "step": 49410 }, { "grad_norm": 0.2609057128429413, "learning_rate": 5.508553109002376e-05, "loss": 0.0182, "step": 49420 }, { "grad_norm": 0.1713026612997055, "learning_rate": 5.5069081862562957e-05, "loss": 0.0198, "step": 49430 }, { "grad_norm": 0.1415536254644394, "learning_rate": 5.5052632080754965e-05, "loss": 0.0163, "step": 49440 }, { "grad_norm": 0.1590377390384674, "learning_rate": 5.503618174639874e-05, "loss": 0.019, "step": 49450 }, { "grad_norm": 0.1932472139596939, "learning_rate": 5.501973086129325e-05, "loss": 0.0182, "step": 49460 }, { "grad_norm": 0.20526641607284546, "learning_rate": 5.500327942723753e-05, "loss": 0.0193, "step": 49470 }, { "grad_norm": 0.16717451810836792, "learning_rate": 5.498682744603071e-05, "loss": 0.0175, "step": 49480 }, { "grad_norm": 0.2017250508069992, "learning_rate": 5.497037491947194e-05, "loss": 0.0178, "step": 49490 }, { "grad_norm": 0.19605644047260284, "learning_rate": 5.4953921849360424e-05, "loss": 0.0173, "step": 49500 }, { "grad_norm": 0.15188942849636078, "learning_rate": 5.493746823749547e-05, "loss": 0.0157, "step": 49510 }, { "grad_norm": 0.23483943939208984, "learning_rate": 5.49210140856764e-05, "loss": 0.0175, "step": 49520 }, { "grad_norm": 0.1867867410182953, "learning_rate": 5.4904559395702647e-05, "loss": 0.0176, "step": 49530 }, { "grad_norm": 0.17753051221370697, "learning_rate": 5.488810416937364e-05, "loss": 0.019, "step": 49540 }, { "grad_norm": 0.20393066108226776, "learning_rate": 5.4871648408488926e-05, "loss": 0.0182, "step": 49550 }, { "grad_norm": 0.25068968534469604, "learning_rate": 5.485519211484807e-05, "loss": 0.0189, "step": 49560 }, { "grad_norm": 0.3278426229953766, "learning_rate": 5.4838735290250676e-05, "loss": 0.0181, "step": 49570 }, { "grad_norm": 0.25496333837509155, "learning_rate": 5.482227793649649e-05, "loss": 0.0164, "step": 49580 }, { "grad_norm": 0.19974416494369507, "learning_rate": 5.480582005538524e-05, "loss": 0.024, "step": 49590 }, { "grad_norm": 0.25206270813941956, "learning_rate": 5.478936164871671e-05, "loss": 0.0211, "step": 49600 }, { "grad_norm": 0.16075630486011505, "learning_rate": 5.477290271829081e-05, "loss": 0.0164, "step": 49610 }, { "grad_norm": 0.17256630957126617, "learning_rate": 5.475644326590744e-05, "loss": 0.0186, "step": 49620 }, { "grad_norm": 0.26079821586608887, "learning_rate": 5.473998329336658e-05, "loss": 0.0176, "step": 49630 }, { "grad_norm": 0.1981649547815323, "learning_rate": 5.4723522802468286e-05, "loss": 0.0227, "step": 49640 }, { "grad_norm": 0.251802921295166, "learning_rate": 5.470706179501264e-05, "loss": 0.0187, "step": 49650 }, { "grad_norm": 0.21577078104019165, "learning_rate": 5.4690600272799776e-05, "loss": 0.0175, "step": 49660 }, { "grad_norm": 0.18268385529518127, "learning_rate": 5.467413823762993e-05, "loss": 0.0184, "step": 49670 }, { "grad_norm": 0.23202309012413025, "learning_rate": 5.465767569130335e-05, "loss": 0.0201, "step": 49680 }, { "grad_norm": 0.18312503397464752, "learning_rate": 5.464121263562036e-05, "loss": 0.0175, "step": 49690 }, { "grad_norm": 0.2165450155735016, "learning_rate": 5.4624749072381343e-05, "loss": 0.0227, "step": 49700 }, { "grad_norm": 0.19888442754745483, "learning_rate": 5.460828500338672e-05, "loss": 0.0188, "step": 49710 }, { "grad_norm": 0.24060417711734772, "learning_rate": 5.459182043043698e-05, "loss": 0.02, "step": 49720 }, { "grad_norm": 0.2862420976161957, "learning_rate": 5.457535535533265e-05, "loss": 0.019, "step": 49730 }, { "grad_norm": 0.19459357857704163, "learning_rate": 5.4558889779874334e-05, "loss": 0.0215, "step": 49740 }, { "grad_norm": 0.32380354404449463, "learning_rate": 5.454242370586269e-05, "loss": 0.0187, "step": 49750 }, { "grad_norm": 0.25091880559921265, "learning_rate": 5.452595713509843e-05, "loss": 0.0188, "step": 49760 }, { "grad_norm": 0.18630951642990112, "learning_rate": 5.4509490069382274e-05, "loss": 0.0217, "step": 49770 }, { "grad_norm": 0.19779939949512482, "learning_rate": 5.449302251051509e-05, "loss": 0.0205, "step": 49780 }, { "grad_norm": 0.20256780087947845, "learning_rate": 5.44765544602977e-05, "loss": 0.0214, "step": 49790 }, { "grad_norm": 0.1803211122751236, "learning_rate": 5.446008592053107e-05, "loss": 0.0174, "step": 49800 }, { "grad_norm": 0.19362516701221466, "learning_rate": 5.4443616893016135e-05, "loss": 0.0189, "step": 49810 }, { "grad_norm": 0.20558221638202667, "learning_rate": 5.4427147379553934e-05, "loss": 0.017, "step": 49820 }, { "grad_norm": 0.22104261815547943, "learning_rate": 5.441067738194556e-05, "loss": 0.0193, "step": 49830 }, { "grad_norm": 0.19584260880947113, "learning_rate": 5.439420690199214e-05, "loss": 0.0192, "step": 49840 }, { "grad_norm": 0.2709973454475403, "learning_rate": 5.437773594149484e-05, "loss": 0.021, "step": 49850 }, { "grad_norm": 0.25330445170402527, "learning_rate": 5.436126450225495e-05, "loss": 0.0205, "step": 49860 }, { "grad_norm": 0.21872040629386902, "learning_rate": 5.434479258607371e-05, "loss": 0.0203, "step": 49870 }, { "grad_norm": 0.19255532324314117, "learning_rate": 5.43283201947525e-05, "loss": 0.0188, "step": 49880 }, { "grad_norm": 0.20008736848831177, "learning_rate": 5.431184733009268e-05, "loss": 0.0203, "step": 49890 }, { "grad_norm": 0.18908345699310303, "learning_rate": 5.4295373993895736e-05, "loss": 0.0163, "step": 49900 }, { "grad_norm": 0.22482502460479736, "learning_rate": 5.4278900187963157e-05, "loss": 0.016, "step": 49910 }, { "grad_norm": 0.2301001399755478, "learning_rate": 5.426242591409646e-05, "loss": 0.0217, "step": 49920 }, { "grad_norm": 0.13786610960960388, "learning_rate": 5.42459511740973e-05, "loss": 0.018, "step": 49930 }, { "grad_norm": 0.2393510788679123, "learning_rate": 5.422947596976729e-05, "loss": 0.0188, "step": 49940 }, { "grad_norm": 0.26177978515625, "learning_rate": 5.4213000302908134e-05, "loss": 0.0216, "step": 49950 }, { "grad_norm": 0.16599836945533752, "learning_rate": 5.419652417532162e-05, "loss": 0.0179, "step": 49960 }, { "grad_norm": 0.1645561307668686, "learning_rate": 5.4180047588809534e-05, "loss": 0.0189, "step": 49970 }, { "grad_norm": 0.21016040444374084, "learning_rate": 5.4163570545173704e-05, "loss": 0.0192, "step": 49980 }, { "grad_norm": 0.19223980605602264, "learning_rate": 5.414709304621608e-05, "loss": 0.0182, "step": 49990 }, { "grad_norm": 0.1850457638502121, "learning_rate": 5.413061509373858e-05, "loss": 0.0193, "step": 50000 }, { "grad_norm": 0.21721015870571136, "learning_rate": 5.411413668954324e-05, "loss": 0.0173, "step": 50010 }, { "grad_norm": 0.1924409568309784, "learning_rate": 5.4097657835432083e-05, "loss": 0.0227, "step": 50020 }, { "grad_norm": 0.22514113783836365, "learning_rate": 5.408117853320723e-05, "loss": 0.0177, "step": 50030 }, { "grad_norm": 0.1817917376756668, "learning_rate": 5.406469878467084e-05, "loss": 0.0209, "step": 50040 }, { "grad_norm": 0.19895131886005402, "learning_rate": 5.404821859162509e-05, "loss": 0.0189, "step": 50050 }, { "grad_norm": 0.21625842154026031, "learning_rate": 5.403173795587225e-05, "loss": 0.0179, "step": 50060 }, { "grad_norm": 0.20249097049236298, "learning_rate": 5.4015256879214606e-05, "loss": 0.018, "step": 50070 }, { "grad_norm": 0.22322674095630646, "learning_rate": 5.39987753634545e-05, "loss": 0.0175, "step": 50080 }, { "grad_norm": 0.27972668409347534, "learning_rate": 5.3982293410394335e-05, "loss": 0.0168, "step": 50090 }, { "grad_norm": 0.22556044161319733, "learning_rate": 5.396581102183655e-05, "loss": 0.0185, "step": 50100 }, { "grad_norm": 0.19902002811431885, "learning_rate": 5.3949328199583634e-05, "loss": 0.0209, "step": 50110 }, { "grad_norm": 0.22886145114898682, "learning_rate": 5.393284494543813e-05, "loss": 0.0224, "step": 50120 }, { "grad_norm": 0.15541964769363403, "learning_rate": 5.391636126120262e-05, "loss": 0.0216, "step": 50130 }, { "grad_norm": 0.15764571726322174, "learning_rate": 5.389987714867971e-05, "loss": 0.0199, "step": 50140 }, { "grad_norm": 0.20865917205810547, "learning_rate": 5.3883392609672114e-05, "loss": 0.0167, "step": 50150 }, { "grad_norm": 0.1769716888666153, "learning_rate": 5.3866907645982545e-05, "loss": 0.0157, "step": 50160 }, { "grad_norm": 0.1980895698070526, "learning_rate": 5.385042225941375e-05, "loss": 0.0189, "step": 50170 }, { "grad_norm": 0.21114306151866913, "learning_rate": 5.3833936451768585e-05, "loss": 0.0149, "step": 50180 }, { "grad_norm": 0.17684468626976013, "learning_rate": 5.3817450224849884e-05, "loss": 0.0158, "step": 50190 }, { "grad_norm": 0.19517147541046143, "learning_rate": 5.380096358046056e-05, "loss": 0.0187, "step": 50200 }, { "grad_norm": 0.21122190356254578, "learning_rate": 5.378447652040359e-05, "loss": 0.0187, "step": 50210 }, { "grad_norm": 0.25108686089515686, "learning_rate": 5.376798904648195e-05, "loss": 0.0188, "step": 50220 }, { "grad_norm": 0.1574217528104782, "learning_rate": 5.375150116049869e-05, "loss": 0.0217, "step": 50230 }, { "grad_norm": 0.2393856793642044, "learning_rate": 5.373501286425691e-05, "loss": 0.0199, "step": 50240 }, { "grad_norm": 0.25860315561294556, "learning_rate": 5.3718524159559726e-05, "loss": 0.0214, "step": 50250 }, { "grad_norm": 0.24197465181350708, "learning_rate": 5.370203504821034e-05, "loss": 0.0178, "step": 50260 }, { "grad_norm": 0.1991364061832428, "learning_rate": 5.368554553201196e-05, "loss": 0.0203, "step": 50270 }, { "grad_norm": 0.21197009086608887, "learning_rate": 5.3669055612767874e-05, "loss": 0.0184, "step": 50280 }, { "grad_norm": 0.20946788787841797, "learning_rate": 5.3652565292281376e-05, "loss": 0.0193, "step": 50290 }, { "grad_norm": 0.16949938237667084, "learning_rate": 5.3636074572355824e-05, "loss": 0.0152, "step": 50300 }, { "grad_norm": 0.14759747684001923, "learning_rate": 5.361958345479463e-05, "loss": 0.0186, "step": 50310 }, { "grad_norm": 0.20351409912109375, "learning_rate": 5.360309194140123e-05, "loss": 0.0178, "step": 50320 }, { "grad_norm": 0.20070086419582367, "learning_rate": 5.3586600033979106e-05, "loss": 0.0179, "step": 50330 }, { "grad_norm": 0.16835203766822815, "learning_rate": 5.357010773433181e-05, "loss": 0.0185, "step": 50340 }, { "grad_norm": 0.17899833619594574, "learning_rate": 5.3553615044262895e-05, "loss": 0.0191, "step": 50350 }, { "grad_norm": 0.2891761064529419, "learning_rate": 5.3537121965575964e-05, "loss": 0.0196, "step": 50360 }, { "grad_norm": 0.19366773962974548, "learning_rate": 5.352062850007471e-05, "loss": 0.0203, "step": 50370 }, { "grad_norm": 0.1712343543767929, "learning_rate": 5.350413464956282e-05, "loss": 0.0187, "step": 50380 }, { "grad_norm": 0.17174449563026428, "learning_rate": 5.348764041584403e-05, "loss": 0.0193, "step": 50390 }, { "grad_norm": 0.1961102932691574, "learning_rate": 5.347114580072212e-05, "loss": 0.0187, "step": 50400 }, { "grad_norm": 0.3207356035709381, "learning_rate": 5.345465080600093e-05, "loss": 0.017, "step": 50410 }, { "grad_norm": 0.16148535907268524, "learning_rate": 5.3438155433484326e-05, "loss": 0.0187, "step": 50420 }, { "grad_norm": 0.17574535310268402, "learning_rate": 5.3421659684976197e-05, "loss": 0.015, "step": 50430 }, { "grad_norm": 0.14868353307247162, "learning_rate": 5.340516356228052e-05, "loss": 0.0171, "step": 50440 }, { "grad_norm": 0.22616371512413025, "learning_rate": 5.338866706720128e-05, "loss": 0.0186, "step": 50450 }, { "grad_norm": 0.27014708518981934, "learning_rate": 5.337217020154249e-05, "loss": 0.0181, "step": 50460 }, { "grad_norm": 0.1976892203092575, "learning_rate": 5.335567296710825e-05, "loss": 0.0181, "step": 50470 }, { "grad_norm": 0.16579729318618774, "learning_rate": 5.333917536570265e-05, "loss": 0.019, "step": 50480 }, { "grad_norm": 0.22925423085689545, "learning_rate": 5.332267739912986e-05, "loss": 0.0173, "step": 50490 }, { "grad_norm": 0.1736288070678711, "learning_rate": 5.330617906919405e-05, "loss": 0.0179, "step": 50500 }, { "grad_norm": 0.4094230532646179, "learning_rate": 5.328968037769949e-05, "loss": 0.0168, "step": 50510 }, { "grad_norm": 0.16022081673145294, "learning_rate": 5.327318132645042e-05, "loss": 0.0206, "step": 50520 }, { "grad_norm": 0.2654649615287781, "learning_rate": 5.325668191725116e-05, "loss": 0.0205, "step": 50530 }, { "grad_norm": 0.1951848268508911, "learning_rate": 5.324018215190606e-05, "loss": 0.017, "step": 50540 }, { "grad_norm": 0.21425846219062805, "learning_rate": 5.3223682032219515e-05, "loss": 0.0182, "step": 50550 }, { "grad_norm": 0.4108874499797821, "learning_rate": 5.320718155999595e-05, "loss": 0.016, "step": 50560 }, { "grad_norm": 0.21614070236682892, "learning_rate": 5.3190680737039835e-05, "loss": 0.0174, "step": 50570 }, { "grad_norm": 0.1831386387348175, "learning_rate": 5.317417956515567e-05, "loss": 0.0263, "step": 50580 }, { "grad_norm": 0.22762185335159302, "learning_rate": 5.315767804614798e-05, "loss": 0.0161, "step": 50590 }, { "grad_norm": 0.21149423718452454, "learning_rate": 5.3141176181821395e-05, "loss": 0.0213, "step": 50600 }, { "grad_norm": 0.27193382382392883, "learning_rate": 5.31246739739805e-05, "loss": 0.0226, "step": 50610 }, { "grad_norm": 0.2076548933982849, "learning_rate": 5.310817142442994e-05, "loss": 0.0163, "step": 50620 }, { "grad_norm": 0.20308038592338562, "learning_rate": 5.309166853497445e-05, "loss": 0.0177, "step": 50630 }, { "grad_norm": 0.17328962683677673, "learning_rate": 5.307516530741873e-05, "loss": 0.0225, "step": 50640 }, { "grad_norm": 0.1656346321105957, "learning_rate": 5.305866174356754e-05, "loss": 0.0179, "step": 50650 }, { "grad_norm": 0.1741143763065338, "learning_rate": 5.304215784522571e-05, "loss": 0.0201, "step": 50660 }, { "grad_norm": 0.21538911759853363, "learning_rate": 5.302565361419808e-05, "loss": 0.0179, "step": 50670 }, { "grad_norm": 0.2406798005104065, "learning_rate": 5.3009149052289507e-05, "loss": 0.0188, "step": 50680 }, { "grad_norm": 0.23906712234020233, "learning_rate": 5.299264416130493e-05, "loss": 0.0175, "step": 50690 }, { "grad_norm": 0.24437563121318817, "learning_rate": 5.297613894304928e-05, "loss": 0.0172, "step": 50700 }, { "grad_norm": 0.14486056566238403, "learning_rate": 5.2959633399327534e-05, "loss": 0.0148, "step": 50710 }, { "grad_norm": 0.17313212156295776, "learning_rate": 5.294312753194476e-05, "loss": 0.0172, "step": 50720 }, { "grad_norm": 0.18899065256118774, "learning_rate": 5.292662134270596e-05, "loss": 0.0172, "step": 50730 }, { "grad_norm": 0.2039368748664856, "learning_rate": 5.291011483341626e-05, "loss": 0.0173, "step": 50740 }, { "grad_norm": 0.19586323201656342, "learning_rate": 5.2893608005880767e-05, "loss": 0.0171, "step": 50750 }, { "grad_norm": 0.2543921172618866, "learning_rate": 5.287710086190467e-05, "loss": 0.0214, "step": 50760 }, { "grad_norm": 0.25672608613967896, "learning_rate": 5.286059340329314e-05, "loss": 0.0171, "step": 50770 }, { "grad_norm": 0.15745024383068085, "learning_rate": 5.284408563185141e-05, "loss": 0.0165, "step": 50780 }, { "grad_norm": 0.20051193237304688, "learning_rate": 5.282757754938476e-05, "loss": 0.0187, "step": 50790 }, { "grad_norm": 0.1734285056591034, "learning_rate": 5.281106915769849e-05, "loss": 0.0164, "step": 50800 }, { "grad_norm": 0.16430191695690155, "learning_rate": 5.2794560458597897e-05, "loss": 0.0172, "step": 50810 }, { "grad_norm": 0.21307829022407532, "learning_rate": 5.27780514538884e-05, "loss": 0.019, "step": 50820 }, { "grad_norm": 0.22130616009235382, "learning_rate": 5.2761542145375365e-05, "loss": 0.016, "step": 50830 }, { "grad_norm": 0.19587017595767975, "learning_rate": 5.274503253486421e-05, "loss": 0.0157, "step": 50840 }, { "grad_norm": 0.15499727427959442, "learning_rate": 5.272852262416046e-05, "loss": 0.0164, "step": 50850 }, { "grad_norm": 0.15573318302631378, "learning_rate": 5.2712012415069555e-05, "loss": 0.017, "step": 50860 }, { "grad_norm": 0.20996379852294922, "learning_rate": 5.269550190939705e-05, "loss": 0.0156, "step": 50870 }, { "grad_norm": 0.18046274781227112, "learning_rate": 5.267899110894852e-05, "loss": 0.0168, "step": 50880 }, { "grad_norm": 0.19065631926059723, "learning_rate": 5.266248001552955e-05, "loss": 0.0158, "step": 50890 }, { "grad_norm": 0.19578708708286285, "learning_rate": 5.264596863094575e-05, "loss": 0.0193, "step": 50900 }, { "grad_norm": 0.17152957618236542, "learning_rate": 5.262945695700282e-05, "loss": 0.0168, "step": 50910 }, { "grad_norm": 0.28193098306655884, "learning_rate": 5.261294499550643e-05, "loss": 0.0233, "step": 50920 }, { "grad_norm": 0.17326979339122772, "learning_rate": 5.25964327482623e-05, "loss": 0.0167, "step": 50930 }, { "grad_norm": 0.19495438039302826, "learning_rate": 5.257992021707617e-05, "loss": 0.0179, "step": 50940 }, { "grad_norm": 0.2301308661699295, "learning_rate": 5.256340740375387e-05, "loss": 0.019, "step": 50950 }, { "grad_norm": 0.15596498548984528, "learning_rate": 5.254689431010117e-05, "loss": 0.0263, "step": 50960 }, { "grad_norm": 0.16627725958824158, "learning_rate": 5.253038093792395e-05, "loss": 0.0151, "step": 50970 }, { "grad_norm": 0.21117767691612244, "learning_rate": 5.251386728902806e-05, "loss": 0.016, "step": 50980 }, { "grad_norm": 0.2475709170103073, "learning_rate": 5.2497353365219446e-05, "loss": 0.0188, "step": 50990 }, { "grad_norm": 0.27679240703582764, "learning_rate": 5.2480839168304e-05, "loss": 0.0191, "step": 51000 }, { "grad_norm": 0.18365170061588287, "learning_rate": 5.246432470008773e-05, "loss": 0.0174, "step": 51010 }, { "grad_norm": 0.24489451944828033, "learning_rate": 5.2447809962376616e-05, "loss": 0.0191, "step": 51020 }, { "grad_norm": 0.22347459197044373, "learning_rate": 5.243129495697666e-05, "loss": 0.0208, "step": 51030 }, { "grad_norm": 0.20147255063056946, "learning_rate": 5.241477968569397e-05, "loss": 0.0173, "step": 51040 }, { "grad_norm": 0.23320533335208893, "learning_rate": 5.23982641503346e-05, "loss": 0.0184, "step": 51050 }, { "grad_norm": 0.19924882054328918, "learning_rate": 5.238174835270465e-05, "loss": 0.0157, "step": 51060 }, { "grad_norm": 0.18794690072536469, "learning_rate": 5.23652322946103e-05, "loss": 0.0165, "step": 51070 }, { "grad_norm": 0.24879801273345947, "learning_rate": 5.234871597785771e-05, "loss": 0.02, "step": 51080 }, { "grad_norm": 0.19679678976535797, "learning_rate": 5.233219940425306e-05, "loss": 0.0187, "step": 51090 }, { "grad_norm": 0.21632014214992523, "learning_rate": 5.2315682575602576e-05, "loss": 0.0154, "step": 51100 }, { "grad_norm": 0.20838195085525513, "learning_rate": 5.229916549371254e-05, "loss": 0.0174, "step": 51110 }, { "grad_norm": 0.2147676795721054, "learning_rate": 5.228264816038924e-05, "loss": 0.0161, "step": 51120 }, { "grad_norm": 0.20836235582828522, "learning_rate": 5.2266130577438935e-05, "loss": 0.0197, "step": 51130 }, { "grad_norm": 0.22340938448905945, "learning_rate": 5.224961274666802e-05, "loss": 0.0183, "step": 51140 }, { "grad_norm": 0.18225930631160736, "learning_rate": 5.223309466988283e-05, "loss": 0.0178, "step": 51150 }, { "grad_norm": 0.20760509371757507, "learning_rate": 5.221657634888976e-05, "loss": 0.017, "step": 51160 }, { "grad_norm": 0.186088427901268, "learning_rate": 5.2200057785495235e-05, "loss": 0.019, "step": 51170 }, { "grad_norm": 0.1901930421590805, "learning_rate": 5.2183538981505706e-05, "loss": 0.0216, "step": 51180 }, { "grad_norm": 0.17087171971797943, "learning_rate": 5.216701993872762e-05, "loss": 0.0191, "step": 51190 }, { "grad_norm": 0.21394452452659607, "learning_rate": 5.215050065896752e-05, "loss": 0.0188, "step": 51200 }, { "grad_norm": 0.23217672109603882, "learning_rate": 5.2133981144031875e-05, "loss": 0.0187, "step": 51210 }, { "grad_norm": 0.2254931628704071, "learning_rate": 5.211746139572727e-05, "loss": 0.0206, "step": 51220 }, { "grad_norm": 0.17657367885112762, "learning_rate": 5.210094141586026e-05, "loss": 0.0217, "step": 51230 }, { "grad_norm": 0.191436767578125, "learning_rate": 5.2084421206237476e-05, "loss": 0.0187, "step": 51240 }, { "grad_norm": 0.19795992970466614, "learning_rate": 5.20679007686655e-05, "loss": 0.0183, "step": 51250 }, { "grad_norm": 0.21880947053432465, "learning_rate": 5.205138010495102e-05, "loss": 0.0202, "step": 51260 }, { "grad_norm": 0.2128855139017105, "learning_rate": 5.20348592169007e-05, "loss": 0.017, "step": 51270 }, { "grad_norm": 0.18464885652065277, "learning_rate": 5.201833810632123e-05, "loss": 0.0233, "step": 51280 }, { "grad_norm": 0.15885092318058014, "learning_rate": 5.200181677501933e-05, "loss": 0.0167, "step": 51290 }, { "grad_norm": 0.23374474048614502, "learning_rate": 5.1985295224801777e-05, "loss": 0.0185, "step": 51300 }, { "grad_norm": 0.15012226998806, "learning_rate": 5.196877345747531e-05, "loss": 0.0167, "step": 51310 }, { "grad_norm": 0.20185774564743042, "learning_rate": 5.195225147484674e-05, "loss": 0.0189, "step": 51320 }, { "grad_norm": 0.24344095587730408, "learning_rate": 5.193572927872288e-05, "loss": 0.0199, "step": 51330 }, { "grad_norm": 0.2013285756111145, "learning_rate": 5.191920687091059e-05, "loss": 0.0201, "step": 51340 }, { "grad_norm": 0.32019057869911194, "learning_rate": 5.1902684253216695e-05, "loss": 0.0175, "step": 51350 }, { "grad_norm": 0.21146440505981445, "learning_rate": 5.1886161427448134e-05, "loss": 0.0181, "step": 51360 }, { "grad_norm": 0.23517973721027374, "learning_rate": 5.186963839541179e-05, "loss": 0.0163, "step": 51370 }, { "grad_norm": 0.22426177561283112, "learning_rate": 5.185311515891459e-05, "loss": 0.0198, "step": 51380 }, { "grad_norm": 0.22020067274570465, "learning_rate": 5.1836591719763516e-05, "loss": 0.0175, "step": 51390 }, { "grad_norm": 0.1976906955242157, "learning_rate": 5.182006807976554e-05, "loss": 0.013, "step": 51400 }, { "grad_norm": 0.2237738072872162, "learning_rate": 5.180354424072763e-05, "loss": 0.0172, "step": 51410 }, { "grad_norm": 0.17541766166687012, "learning_rate": 5.178702020445686e-05, "loss": 0.016, "step": 51420 }, { "grad_norm": 0.4694939851760864, "learning_rate": 5.177049597276024e-05, "loss": 0.0198, "step": 51430 }, { "grad_norm": 0.1796540915966034, "learning_rate": 5.175397154744482e-05, "loss": 0.0149, "step": 51440 }, { "grad_norm": 0.21056562662124634, "learning_rate": 5.1737446930317736e-05, "loss": 0.0175, "step": 51450 }, { "grad_norm": 0.23588113486766815, "learning_rate": 5.172092212318604e-05, "loss": 0.0236, "step": 51460 }, { "grad_norm": 0.15809127688407898, "learning_rate": 5.170439712785692e-05, "loss": 0.017, "step": 51470 }, { "grad_norm": 0.1647893339395523, "learning_rate": 5.1687871946137467e-05, "loss": 0.0167, "step": 51480 }, { "grad_norm": 0.18919208645820618, "learning_rate": 5.16713465798349e-05, "loss": 0.0166, "step": 51490 }, { "grad_norm": 0.2480635941028595, "learning_rate": 5.165482103075637e-05, "loss": 0.0184, "step": 51500 }, { "grad_norm": 0.15707631409168243, "learning_rate": 5.163829530070909e-05, "loss": 0.0199, "step": 51510 }, { "grad_norm": 0.1844170093536377, "learning_rate": 5.162176939150032e-05, "loss": 0.0153, "step": 51520 }, { "grad_norm": 0.21761728823184967, "learning_rate": 5.160524330493728e-05, "loss": 0.0204, "step": 51530 }, { "grad_norm": 0.18354777991771698, "learning_rate": 5.158871704282724e-05, "loss": 0.0188, "step": 51540 }, { "grad_norm": 0.1907922625541687, "learning_rate": 5.15721906069775e-05, "loss": 0.016, "step": 51550 }, { "grad_norm": 0.13600672781467438, "learning_rate": 5.155566399919536e-05, "loss": 0.016, "step": 51560 }, { "grad_norm": 0.13330325484275818, "learning_rate": 5.153913722128813e-05, "loss": 0.0141, "step": 51570 }, { "grad_norm": 0.1742459088563919, "learning_rate": 5.1522610275063176e-05, "loss": 0.0164, "step": 51580 }, { "grad_norm": 0.24676641821861267, "learning_rate": 5.150608316232786e-05, "loss": 0.0175, "step": 51590 }, { "grad_norm": 0.1967104822397232, "learning_rate": 5.148955588488955e-05, "loss": 0.0179, "step": 51600 }, { "grad_norm": 0.18749968707561493, "learning_rate": 5.147302844455564e-05, "loss": 0.0173, "step": 51610 }, { "grad_norm": 0.18553929030895233, "learning_rate": 5.145650084313356e-05, "loss": 0.0163, "step": 51620 }, { "grad_norm": 0.1745738834142685, "learning_rate": 5.1439973082430734e-05, "loss": 0.0165, "step": 51630 }, { "grad_norm": 0.19962134957313538, "learning_rate": 5.142344516425461e-05, "loss": 0.0157, "step": 51640 }, { "grad_norm": 0.16807027161121368, "learning_rate": 5.1406917090412673e-05, "loss": 0.0153, "step": 51650 }, { "grad_norm": 0.19571252167224884, "learning_rate": 5.1390388862712404e-05, "loss": 0.0151, "step": 51660 }, { "grad_norm": 0.33122897148132324, "learning_rate": 5.137386048296129e-05, "loss": 0.0239, "step": 51670 }, { "grad_norm": 0.19078713655471802, "learning_rate": 5.135733195296686e-05, "loss": 0.0172, "step": 51680 }, { "grad_norm": 0.18403345346450806, "learning_rate": 5.134080327453664e-05, "loss": 0.0186, "step": 51690 }, { "grad_norm": 0.22060030698776245, "learning_rate": 5.1324274449478204e-05, "loss": 0.0194, "step": 51700 }, { "grad_norm": 0.2512471079826355, "learning_rate": 5.130774547959909e-05, "loss": 0.0159, "step": 51710 }, { "grad_norm": 0.15302890539169312, "learning_rate": 5.129121636670692e-05, "loss": 0.022, "step": 51720 }, { "grad_norm": 0.19097915291786194, "learning_rate": 5.1274687112609246e-05, "loss": 0.0166, "step": 51730 }, { "grad_norm": 0.18188942968845367, "learning_rate": 5.125815771911372e-05, "loss": 0.0187, "step": 51740 }, { "grad_norm": 0.21246899664402008, "learning_rate": 5.1241628188027956e-05, "loss": 0.0149, "step": 51750 }, { "grad_norm": 0.261409729719162, "learning_rate": 5.1225098521159585e-05, "loss": 0.019, "step": 51760 }, { "grad_norm": 0.168845072388649, "learning_rate": 5.12085687203163e-05, "loss": 0.0202, "step": 51770 }, { "grad_norm": 0.25955209136009216, "learning_rate": 5.119203878730575e-05, "loss": 0.0189, "step": 51780 }, { "grad_norm": 0.15728849172592163, "learning_rate": 5.117550872393564e-05, "loss": 0.016, "step": 51790 }, { "grad_norm": 0.18896035850048065, "learning_rate": 5.1158978532013645e-05, "loss": 0.0217, "step": 51800 }, { "grad_norm": 0.30490171909332275, "learning_rate": 5.114244821334751e-05, "loss": 0.0209, "step": 51810 }, { "grad_norm": 0.18008707463741302, "learning_rate": 5.112591776974497e-05, "loss": 0.0156, "step": 51820 }, { "grad_norm": 0.18876560032367706, "learning_rate": 5.110938720301374e-05, "loss": 0.0186, "step": 51830 }, { "grad_norm": 0.14849388599395752, "learning_rate": 5.1092856514961595e-05, "loss": 0.0201, "step": 51840 }, { "grad_norm": 0.14686281979084015, "learning_rate": 5.107632570739631e-05, "loss": 0.0153, "step": 51850 }, { "grad_norm": 0.18347562849521637, "learning_rate": 5.1059794782125645e-05, "loss": 0.0233, "step": 51860 }, { "grad_norm": 0.18111540377140045, "learning_rate": 5.104326374095744e-05, "loss": 0.0195, "step": 51870 }, { "grad_norm": 0.19286352396011353, "learning_rate": 5.102673258569947e-05, "loss": 0.0219, "step": 51880 }, { "grad_norm": 0.21485401690006256, "learning_rate": 5.101020131815956e-05, "loss": 0.0184, "step": 51890 }, { "grad_norm": 0.22014136612415314, "learning_rate": 5.099366994014556e-05, "loss": 0.0207, "step": 51900 }, { "grad_norm": 0.2314479649066925, "learning_rate": 5.097713845346532e-05, "loss": 0.0206, "step": 51910 }, { "grad_norm": 0.15726566314697266, "learning_rate": 5.096060685992668e-05, "loss": 0.0198, "step": 51920 }, { "grad_norm": 0.1665259599685669, "learning_rate": 5.0944075161337514e-05, "loss": 0.0182, "step": 51930 }, { "grad_norm": 0.1404602974653244, "learning_rate": 5.09275433595057e-05, "loss": 0.0171, "step": 51940 }, { "grad_norm": 0.160282701253891, "learning_rate": 5.0911011456239157e-05, "loss": 0.0161, "step": 51950 }, { "grad_norm": 0.20603428781032562, "learning_rate": 5.089447945334574e-05, "loss": 0.0163, "step": 51960 }, { "grad_norm": 0.20040322840213776, "learning_rate": 5.087794735263343e-05, "loss": 0.0166, "step": 51970 }, { "grad_norm": 0.21799908578395844, "learning_rate": 5.086141515591011e-05, "loss": 0.0173, "step": 51980 }, { "grad_norm": 0.17237406969070435, "learning_rate": 5.084488286498371e-05, "loss": 0.0188, "step": 51990 }, { "grad_norm": 0.2129688709974289, "learning_rate": 5.08283504816622e-05, "loss": 0.0191, "step": 52000 }, { "grad_norm": 0.15262740850448608, "learning_rate": 5.081181800775353e-05, "loss": 0.0196, "step": 52010 }, { "grad_norm": 0.24833306670188904, "learning_rate": 5.079528544506565e-05, "loss": 0.0202, "step": 52020 }, { "grad_norm": 0.17676876485347748, "learning_rate": 5.077875279540657e-05, "loss": 0.0184, "step": 52030 }, { "grad_norm": 0.21867682039737701, "learning_rate": 5.076222006058424e-05, "loss": 0.018, "step": 52040 }, { "grad_norm": 0.2150750458240509, "learning_rate": 5.074568724240667e-05, "loss": 0.0213, "step": 52050 }, { "grad_norm": 0.2011292725801468, "learning_rate": 5.072915434268187e-05, "loss": 0.0155, "step": 52060 }, { "grad_norm": 0.15168039500713348, "learning_rate": 5.071262136321785e-05, "loss": 0.0154, "step": 52070 }, { "grad_norm": 0.2494329810142517, "learning_rate": 5.069608830582262e-05, "loss": 0.0162, "step": 52080 }, { "grad_norm": 0.18560291826725006, "learning_rate": 5.067955517230423e-05, "loss": 0.0194, "step": 52090 }, { "grad_norm": 0.19889725744724274, "learning_rate": 5.06630219644707e-05, "loss": 0.0175, "step": 52100 }, { "grad_norm": 0.17385847866535187, "learning_rate": 5.064648868413008e-05, "loss": 0.0153, "step": 52110 }, { "grad_norm": 0.21543675661087036, "learning_rate": 5.062995533309044e-05, "loss": 0.017, "step": 52120 }, { "grad_norm": 0.23380009829998016, "learning_rate": 5.061342191315983e-05, "loss": 0.0178, "step": 52130 }, { "grad_norm": 0.2390987128019333, "learning_rate": 5.0596888426146325e-05, "loss": 0.0189, "step": 52140 }, { "grad_norm": 0.24841924011707306, "learning_rate": 5.058035487385798e-05, "loss": 0.019, "step": 52150 }, { "grad_norm": 0.16082863509655, "learning_rate": 5.0563821258102915e-05, "loss": 0.0152, "step": 52160 }, { "grad_norm": 0.17632505297660828, "learning_rate": 5.05472875806892e-05, "loss": 0.0156, "step": 52170 }, { "grad_norm": 0.1773897260427475, "learning_rate": 5.053075384342494e-05, "loss": 0.0158, "step": 52180 }, { "grad_norm": 0.15431064367294312, "learning_rate": 5.0514220048118234e-05, "loss": 0.016, "step": 52190 }, { "grad_norm": 0.23743917047977448, "learning_rate": 5.049768619657721e-05, "loss": 0.0183, "step": 52200 }, { "grad_norm": 0.17032083868980408, "learning_rate": 5.048115229060996e-05, "loss": 0.0171, "step": 52210 }, { "grad_norm": 0.48611271381378174, "learning_rate": 5.046461833202462e-05, "loss": 0.0187, "step": 52220 }, { "grad_norm": 0.2083245813846588, "learning_rate": 5.044808432262932e-05, "loss": 0.0175, "step": 52230 }, { "grad_norm": 0.21656499803066254, "learning_rate": 5.043155026423219e-05, "loss": 0.0195, "step": 52240 }, { "grad_norm": 0.18410265445709229, "learning_rate": 5.041501615864137e-05, "loss": 0.0188, "step": 52250 }, { "grad_norm": 0.1588584929704666, "learning_rate": 5.039848200766502e-05, "loss": 0.0163, "step": 52260 }, { "grad_norm": 0.1661570519208908, "learning_rate": 5.038194781311125e-05, "loss": 0.0184, "step": 52270 }, { "grad_norm": 0.20581792294979095, "learning_rate": 5.0365413576788264e-05, "loss": 0.0191, "step": 52280 }, { "grad_norm": 0.21927601099014282, "learning_rate": 5.0348879300504194e-05, "loss": 0.0183, "step": 52290 }, { "grad_norm": 0.3004867434501648, "learning_rate": 5.033234498606721e-05, "loss": 0.0185, "step": 52300 }, { "grad_norm": 0.20913171768188477, "learning_rate": 5.0315810635285476e-05, "loss": 0.0199, "step": 52310 }, { "grad_norm": 0.16099666059017181, "learning_rate": 5.0299276249967164e-05, "loss": 0.0176, "step": 52320 }, { "grad_norm": 0.15112610161304474, "learning_rate": 5.0282741831920454e-05, "loss": 0.0165, "step": 52330 }, { "grad_norm": 0.17202988266944885, "learning_rate": 5.026620738295351e-05, "loss": 0.0191, "step": 52340 }, { "grad_norm": 0.20188786089420319, "learning_rate": 5.024967290487455e-05, "loss": 0.0154, "step": 52350 }, { "grad_norm": 0.16795027256011963, "learning_rate": 5.023313839949172e-05, "loss": 0.0172, "step": 52360 }, { "grad_norm": 0.19275139272212982, "learning_rate": 5.0216603868613224e-05, "loss": 0.0151, "step": 52370 }, { "grad_norm": 0.17492522299289703, "learning_rate": 5.020006931404726e-05, "loss": 0.0186, "step": 52380 }, { "grad_norm": 0.16577306389808655, "learning_rate": 5.0183534737602014e-05, "loss": 0.0184, "step": 52390 }, { "grad_norm": 0.17426902055740356, "learning_rate": 5.0167000141085686e-05, "loss": 0.0174, "step": 52400 }, { "grad_norm": 0.20389841496944427, "learning_rate": 5.015046552630647e-05, "loss": 0.0208, "step": 52410 }, { "grad_norm": 0.23162297904491425, "learning_rate": 5.013393089507258e-05, "loss": 0.0206, "step": 52420 }, { "grad_norm": 0.21447061002254486, "learning_rate": 5.011739624919219e-05, "loss": 0.0175, "step": 52430 }, { "grad_norm": 0.1960039883852005, "learning_rate": 5.0100861590473545e-05, "loss": 0.0179, "step": 52440 }, { "grad_norm": 0.1881413459777832, "learning_rate": 5.008432692072482e-05, "loss": 0.0164, "step": 52450 }, { "grad_norm": 0.2452923208475113, "learning_rate": 5.006779224175423e-05, "loss": 0.0187, "step": 52460 }, { "grad_norm": 0.19670653343200684, "learning_rate": 5.005125755536998e-05, "loss": 0.0141, "step": 52470 }, { "grad_norm": 0.22060318291187286, "learning_rate": 5.003472286338029e-05, "loss": 0.0148, "step": 52480 }, { "grad_norm": 0.2234545797109604, "learning_rate": 5.001818816759335e-05, "loss": 0.0145, "step": 52490 }, { "grad_norm": 0.15916670858860016, "learning_rate": 5.000165346981738e-05, "loss": 0.0164, "step": 52500 }, { "grad_norm": 0.21530784666538239, "learning_rate": 4.9985118771860595e-05, "loss": 0.0182, "step": 52510 }, { "grad_norm": 0.19938108325004578, "learning_rate": 4.996858407553119e-05, "loss": 0.0183, "step": 52520 }, { "grad_norm": 0.17451241612434387, "learning_rate": 4.99520493826374e-05, "loss": 0.0177, "step": 52530 }, { "grad_norm": 0.1806677281856537, "learning_rate": 4.9935514694987384e-05, "loss": 0.0143, "step": 52540 }, { "grad_norm": 0.19227851927280426, "learning_rate": 4.991898001438941e-05, "loss": 0.017, "step": 52550 }, { "grad_norm": 0.16608671844005585, "learning_rate": 4.990244534265166e-05, "loss": 0.0174, "step": 52560 }, { "grad_norm": 0.2243773341178894, "learning_rate": 4.988591068158234e-05, "loss": 0.0157, "step": 52570 }, { "grad_norm": 0.19688351452350616, "learning_rate": 4.986937603298965e-05, "loss": 0.0167, "step": 52580 }, { "grad_norm": 0.1971692442893982, "learning_rate": 4.985284139868181e-05, "loss": 0.0165, "step": 52590 }, { "grad_norm": 0.21093827486038208, "learning_rate": 4.9836306780467e-05, "loss": 0.017, "step": 52600 }, { "grad_norm": 0.26144665479660034, "learning_rate": 4.981977218015343e-05, "loss": 0.0159, "step": 52610 }, { "grad_norm": 0.18112146854400635, "learning_rate": 4.980323759954932e-05, "loss": 0.0153, "step": 52620 }, { "grad_norm": 0.16054502129554749, "learning_rate": 4.978670304046284e-05, "loss": 0.0175, "step": 52630 }, { "grad_norm": 0.17102953791618347, "learning_rate": 4.9770168504702195e-05, "loss": 0.0172, "step": 52640 }, { "grad_norm": 0.21547459065914154, "learning_rate": 4.975363399407559e-05, "loss": 0.0179, "step": 52650 }, { "grad_norm": 0.1874857395887375, "learning_rate": 4.9737099510391194e-05, "loss": 0.0193, "step": 52660 }, { "grad_norm": 0.2636081576347351, "learning_rate": 4.972056505545719e-05, "loss": 0.0208, "step": 52670 }, { "grad_norm": 0.21822133660316467, "learning_rate": 4.9704030631081775e-05, "loss": 0.0195, "step": 52680 }, { "grad_norm": 0.19868417084217072, "learning_rate": 4.968749623907313e-05, "loss": 0.02, "step": 52690 }, { "grad_norm": 0.18385614454746246, "learning_rate": 4.967096188123943e-05, "loss": 0.0179, "step": 52700 }, { "grad_norm": 0.13032494485378265, "learning_rate": 4.965442755938884e-05, "loss": 0.0157, "step": 52710 }, { "grad_norm": 0.16794368624687195, "learning_rate": 4.963789327532954e-05, "loss": 0.0137, "step": 52720 }, { "grad_norm": 0.19252009689807892, "learning_rate": 4.962135903086968e-05, "loss": 0.0149, "step": 52730 }, { "grad_norm": 0.18464632332324982, "learning_rate": 4.960482482781741e-05, "loss": 0.0127, "step": 52740 }, { "grad_norm": 0.2337939590215683, "learning_rate": 4.958829066798092e-05, "loss": 0.0194, "step": 52750 }, { "grad_norm": 0.2113730013370514, "learning_rate": 4.957175655316835e-05, "loss": 0.0217, "step": 52760 }, { "grad_norm": 0.22719153761863708, "learning_rate": 4.9555222485187834e-05, "loss": 0.0147, "step": 52770 }, { "grad_norm": 0.19475843012332916, "learning_rate": 4.953868846584752e-05, "loss": 0.016, "step": 52780 }, { "grad_norm": 0.19716760516166687, "learning_rate": 4.9522154496955544e-05, "loss": 0.0175, "step": 52790 }, { "grad_norm": 0.16862478852272034, "learning_rate": 4.950562058032001e-05, "loss": 0.0194, "step": 52800 }, { "grad_norm": 0.20858074724674225, "learning_rate": 4.948908671774909e-05, "loss": 0.0183, "step": 52810 }, { "grad_norm": 0.27122724056243896, "learning_rate": 4.947255291105087e-05, "loss": 0.0179, "step": 52820 }, { "grad_norm": 0.2046983391046524, "learning_rate": 4.945601916203348e-05, "loss": 0.0172, "step": 52830 }, { "grad_norm": 0.18702402710914612, "learning_rate": 4.9439485472505016e-05, "loss": 0.0198, "step": 52840 }, { "grad_norm": 0.16786783933639526, "learning_rate": 4.942295184427356e-05, "loss": 0.0203, "step": 52850 }, { "grad_norm": 0.17725259065628052, "learning_rate": 4.9406418279147234e-05, "loss": 0.017, "step": 52860 }, { "grad_norm": 0.15425027906894684, "learning_rate": 4.9389884778934116e-05, "loss": 0.017, "step": 52870 }, { "grad_norm": 0.15867933630943298, "learning_rate": 4.937335134544228e-05, "loss": 0.0168, "step": 52880 }, { "grad_norm": 0.12763996422290802, "learning_rate": 4.935681798047979e-05, "loss": 0.0152, "step": 52890 }, { "grad_norm": 0.22602635622024536, "learning_rate": 4.934028468585473e-05, "loss": 0.0191, "step": 52900 }, { "grad_norm": 0.14458027482032776, "learning_rate": 4.932375146337514e-05, "loss": 0.0167, "step": 52910 }, { "grad_norm": 0.19981668889522552, "learning_rate": 4.930721831484908e-05, "loss": 0.0191, "step": 52920 }, { "grad_norm": 0.18876489996910095, "learning_rate": 4.929068524208456e-05, "loss": 0.0161, "step": 52930 }, { "grad_norm": 0.26572179794311523, "learning_rate": 4.927415224688965e-05, "loss": 0.0191, "step": 52940 }, { "grad_norm": 0.17965909838676453, "learning_rate": 4.9257619331072375e-05, "loss": 0.0155, "step": 52950 }, { "grad_norm": 0.17096149921417236, "learning_rate": 4.924108649644072e-05, "loss": 0.0156, "step": 52960 }, { "grad_norm": 0.20866046845912933, "learning_rate": 4.922455374480271e-05, "loss": 0.0174, "step": 52970 }, { "grad_norm": 0.17123174667358398, "learning_rate": 4.920802107796634e-05, "loss": 0.0165, "step": 52980 }, { "grad_norm": 0.19973425567150116, "learning_rate": 4.9191488497739564e-05, "loss": 0.0163, "step": 52990 }, { "grad_norm": 0.2500126361846924, "learning_rate": 4.917495600593042e-05, "loss": 0.0172, "step": 53000 }, { "grad_norm": 0.2038763016462326, "learning_rate": 4.915842360434684e-05, "loss": 0.0171, "step": 53010 }, { "grad_norm": 0.1922164112329483, "learning_rate": 4.914189129479679e-05, "loss": 0.0175, "step": 53020 }, { "grad_norm": 0.18531319499015808, "learning_rate": 4.912535907908822e-05, "loss": 0.0173, "step": 53030 }, { "grad_norm": 0.1952197253704071, "learning_rate": 4.910882695902907e-05, "loss": 0.0159, "step": 53040 }, { "grad_norm": 0.20037581026554108, "learning_rate": 4.9092294936427234e-05, "loss": 0.0164, "step": 53050 }, { "grad_norm": 0.21386541426181793, "learning_rate": 4.9075763013090685e-05, "loss": 0.0173, "step": 53060 }, { "grad_norm": 0.18949145078659058, "learning_rate": 4.9059231190827305e-05, "loss": 0.0179, "step": 53070 }, { "grad_norm": 0.1615937054157257, "learning_rate": 4.904269947144499e-05, "loss": 0.0155, "step": 53080 }, { "grad_norm": 0.1664070338010788, "learning_rate": 4.902616785675161e-05, "loss": 0.0168, "step": 53090 }, { "grad_norm": 0.1925656646490097, "learning_rate": 4.900963634855505e-05, "loss": 0.0157, "step": 53100 }, { "grad_norm": 0.25456613302230835, "learning_rate": 4.8993104948663173e-05, "loss": 0.0195, "step": 53110 }, { "grad_norm": 0.1795198619365692, "learning_rate": 4.8976573658883823e-05, "loss": 0.0177, "step": 53120 }, { "grad_norm": 0.2592797577381134, "learning_rate": 4.896004248102484e-05, "loss": 0.0192, "step": 53130 }, { "grad_norm": 0.22179827094078064, "learning_rate": 4.8943511416894045e-05, "loss": 0.0164, "step": 53140 }, { "grad_norm": 0.16485176980495453, "learning_rate": 4.892698046829926e-05, "loss": 0.0176, "step": 53150 }, { "grad_norm": 0.1475653052330017, "learning_rate": 4.8910449637048275e-05, "loss": 0.0176, "step": 53160 }, { "grad_norm": 0.19371397793293, "learning_rate": 4.8893918924948884e-05, "loss": 0.0175, "step": 53170 }, { "grad_norm": 0.18103209137916565, "learning_rate": 4.8877388333808835e-05, "loss": 0.0182, "step": 53180 }, { "grad_norm": 0.2269100397825241, "learning_rate": 4.8860857865435925e-05, "loss": 0.0147, "step": 53190 }, { "grad_norm": 0.14176520705223083, "learning_rate": 4.884432752163789e-05, "loss": 0.0187, "step": 53200 }, { "grad_norm": 0.17427021265029907, "learning_rate": 4.8827797304222466e-05, "loss": 0.0178, "step": 53210 }, { "grad_norm": 0.19395464658737183, "learning_rate": 4.8811267214997366e-05, "loss": 0.0172, "step": 53220 }, { "grad_norm": 0.16620171070098877, "learning_rate": 4.879473725577029e-05, "loss": 0.0168, "step": 53230 }, { "grad_norm": 0.22057995200157166, "learning_rate": 4.877820742834891e-05, "loss": 0.0185, "step": 53240 }, { "grad_norm": 0.18090099096298218, "learning_rate": 4.8761677734540956e-05, "loss": 0.0162, "step": 53250 }, { "grad_norm": 0.22786743938922882, "learning_rate": 4.874514817615407e-05, "loss": 0.0157, "step": 53260 }, { "grad_norm": 0.20304858684539795, "learning_rate": 4.872861875499588e-05, "loss": 0.0166, "step": 53270 }, { "grad_norm": 0.25488823652267456, "learning_rate": 4.871208947287404e-05, "loss": 0.0174, "step": 53280 }, { "grad_norm": 0.21975794434547424, "learning_rate": 4.869556033159615e-05, "loss": 0.0171, "step": 53290 }, { "grad_norm": 0.14873667061328888, "learning_rate": 4.8679031332969835e-05, "loss": 0.0171, "step": 53300 }, { "grad_norm": 0.1845386028289795, "learning_rate": 4.8662502478802635e-05, "loss": 0.0197, "step": 53310 }, { "grad_norm": 0.2142050564289093, "learning_rate": 4.8645973770902176e-05, "loss": 0.0165, "step": 53320 }, { "grad_norm": 0.23614202439785004, "learning_rate": 4.862944521107599e-05, "loss": 0.0188, "step": 53330 }, { "grad_norm": 0.14981161057949066, "learning_rate": 4.861291680113162e-05, "loss": 0.0192, "step": 53340 }, { "grad_norm": 0.19281260669231415, "learning_rate": 4.8596388542876563e-05, "loss": 0.017, "step": 53350 }, { "grad_norm": 0.19791707396507263, "learning_rate": 4.857986043811836e-05, "loss": 0.0172, "step": 53360 }, { "grad_norm": 0.19456085562705994, "learning_rate": 4.856333248866448e-05, "loss": 0.0189, "step": 53370 }, { "grad_norm": 0.13814321160316467, "learning_rate": 4.85468046963224e-05, "loss": 0.016, "step": 53380 }, { "grad_norm": 0.2001998871564865, "learning_rate": 4.853027706289957e-05, "loss": 0.0168, "step": 53390 }, { "grad_norm": 0.18721874058246613, "learning_rate": 4.851374959020344e-05, "loss": 0.018, "step": 53400 }, { "grad_norm": 0.19240802526474, "learning_rate": 4.849722228004141e-05, "loss": 0.0158, "step": 53410 }, { "grad_norm": 0.1956101953983307, "learning_rate": 4.848069513422091e-05, "loss": 0.0162, "step": 53420 }, { "grad_norm": 0.22427642345428467, "learning_rate": 4.8464168154549294e-05, "loss": 0.0172, "step": 53430 }, { "grad_norm": 0.19860407710075378, "learning_rate": 4.844764134283392e-05, "loss": 0.018, "step": 53440 }, { "grad_norm": 0.2549300789833069, "learning_rate": 4.843111470088218e-05, "loss": 0.0176, "step": 53450 }, { "grad_norm": 0.1604147106409073, "learning_rate": 4.841458823050137e-05, "loss": 0.0179, "step": 53460 }, { "grad_norm": 0.17625391483306885, "learning_rate": 4.8398061933498816e-05, "loss": 0.0151, "step": 53470 }, { "grad_norm": 0.14690636098384857, "learning_rate": 4.83815358116818e-05, "loss": 0.0169, "step": 53480 }, { "grad_norm": 0.26647377014160156, "learning_rate": 4.8365009866857584e-05, "loss": 0.0189, "step": 53490 }, { "grad_norm": 0.1579408347606659, "learning_rate": 4.834848410083342e-05, "loss": 0.0185, "step": 53500 }, { "grad_norm": 0.7331882119178772, "learning_rate": 4.833195851541657e-05, "loss": 0.0154, "step": 53510 }, { "grad_norm": 0.2318931370973587, "learning_rate": 4.831543311241423e-05, "loss": 0.0244, "step": 53520 }, { "grad_norm": 0.241911843419075, "learning_rate": 4.829890789363359e-05, "loss": 0.0211, "step": 53530 }, { "grad_norm": 0.18426936864852905, "learning_rate": 4.828238286088182e-05, "loss": 0.0198, "step": 53540 }, { "grad_norm": 0.1796611100435257, "learning_rate": 4.826585801596609e-05, "loss": 0.0193, "step": 53550 }, { "grad_norm": 0.20242343842983246, "learning_rate": 4.8249333360693486e-05, "loss": 0.0177, "step": 53560 }, { "grad_norm": 0.2091234177350998, "learning_rate": 4.8232808896871185e-05, "loss": 0.0185, "step": 53570 }, { "grad_norm": 0.2639659643173218, "learning_rate": 4.821628462630624e-05, "loss": 0.016, "step": 53580 }, { "grad_norm": 0.1367916464805603, "learning_rate": 4.8199760550805735e-05, "loss": 0.0157, "step": 53590 }, { "grad_norm": 0.22611774504184723, "learning_rate": 4.818323667217669e-05, "loss": 0.019, "step": 53600 }, { "grad_norm": 0.1774034947156906, "learning_rate": 4.8166712992226165e-05, "loss": 0.0174, "step": 53610 }, { "grad_norm": 0.19495603442192078, "learning_rate": 4.8150189512761154e-05, "loss": 0.0222, "step": 53620 }, { "grad_norm": 0.22074101865291595, "learning_rate": 4.813366623558862e-05, "loss": 0.0188, "step": 53630 }, { "grad_norm": 0.18821465969085693, "learning_rate": 4.811714316251554e-05, "loss": 0.0185, "step": 53640 }, { "grad_norm": 0.22122228145599365, "learning_rate": 4.810062029534886e-05, "loss": 0.0206, "step": 53650 }, { "grad_norm": 0.21973764896392822, "learning_rate": 4.8084097635895505e-05, "loss": 0.0205, "step": 53660 }, { "grad_norm": 0.1788010150194168, "learning_rate": 4.8067575185962335e-05, "loss": 0.0164, "step": 53670 }, { "grad_norm": 0.1485755741596222, "learning_rate": 4.805105294735625e-05, "loss": 0.0172, "step": 53680 }, { "grad_norm": 0.15977449715137482, "learning_rate": 4.803453092188406e-05, "loss": 0.0163, "step": 53690 }, { "grad_norm": 0.1894797831773758, "learning_rate": 4.801800911135263e-05, "loss": 0.0188, "step": 53700 }, { "grad_norm": 0.16999293863773346, "learning_rate": 4.8001487517568744e-05, "loss": 0.0177, "step": 53710 }, { "grad_norm": 0.19007252156734467, "learning_rate": 4.798496614233918e-05, "loss": 0.0174, "step": 53720 }, { "grad_norm": 0.22358381748199463, "learning_rate": 4.796844498747069e-05, "loss": 0.0199, "step": 53730 }, { "grad_norm": 0.23136542737483978, "learning_rate": 4.795192405477e-05, "loss": 0.0154, "step": 53740 }, { "grad_norm": 0.19713228940963745, "learning_rate": 4.79354033460438e-05, "loss": 0.0163, "step": 53750 }, { "grad_norm": 0.22933802008628845, "learning_rate": 4.79188828630988e-05, "loss": 0.0196, "step": 53760 }, { "grad_norm": 0.17196112871170044, "learning_rate": 4.790236260774165e-05, "loss": 0.0198, "step": 53770 }, { "grad_norm": 0.14118634164333344, "learning_rate": 4.788584258177896e-05, "loss": 0.0174, "step": 53780 }, { "grad_norm": 0.14467160403728485, "learning_rate": 4.7869322787017356e-05, "loss": 0.0168, "step": 53790 }, { "grad_norm": 0.21085263788700104, "learning_rate": 4.78528032252634e-05, "loss": 0.0167, "step": 53800 }, { "grad_norm": 0.2362741082906723, "learning_rate": 4.783628389832366e-05, "loss": 0.0221, "step": 53810 }, { "grad_norm": 0.2324490249156952, "learning_rate": 4.7819764808004644e-05, "loss": 0.0203, "step": 53820 }, { "grad_norm": 0.16984812915325165, "learning_rate": 4.780324595611289e-05, "loss": 0.0148, "step": 53830 }, { "grad_norm": 0.21996469795703888, "learning_rate": 4.7786727344454854e-05, "loss": 0.0192, "step": 53840 }, { "grad_norm": 0.18541723489761353, "learning_rate": 4.777020897483697e-05, "loss": 0.0173, "step": 53850 }, { "grad_norm": 0.19385012984275818, "learning_rate": 4.775369084906569e-05, "loss": 0.0189, "step": 53860 }, { "grad_norm": 0.15157820284366608, "learning_rate": 4.773717296894742e-05, "loss": 0.0161, "step": 53870 }, { "grad_norm": 0.25092846155166626, "learning_rate": 4.772065533628848e-05, "loss": 0.0177, "step": 53880 }, { "grad_norm": 0.3335728943347931, "learning_rate": 4.7704137952895265e-05, "loss": 0.0205, "step": 53890 }, { "grad_norm": 0.7505324482917786, "learning_rate": 4.768762082057407e-05, "loss": 0.0211, "step": 53900 }, { "grad_norm": 0.17569227516651154, "learning_rate": 4.7671103941131186e-05, "loss": 0.0174, "step": 53910 }, { "grad_norm": 0.21591278910636902, "learning_rate": 4.765458731637287e-05, "loss": 0.0206, "step": 53920 }, { "grad_norm": 0.24694830179214478, "learning_rate": 4.763807094810537e-05, "loss": 0.021, "step": 53930 }, { "grad_norm": 0.21131181716918945, "learning_rate": 4.762155483813487e-05, "loss": 0.0194, "step": 53940 }, { "grad_norm": 0.18319612741470337, "learning_rate": 4.760503898826754e-05, "loss": 0.0209, "step": 53950 }, { "grad_norm": 0.1719888299703598, "learning_rate": 4.758852340030957e-05, "loss": 0.0179, "step": 53960 }, { "grad_norm": 0.23825877904891968, "learning_rate": 4.757200807606705e-05, "loss": 0.0183, "step": 53970 }, { "grad_norm": 0.1826046258211136, "learning_rate": 4.755549301734607e-05, "loss": 0.0166, "step": 53980 }, { "grad_norm": 0.19779247045516968, "learning_rate": 4.753897822595271e-05, "loss": 0.0163, "step": 53990 }, { "grad_norm": 0.17481626570224762, "learning_rate": 4.7522463703692994e-05, "loss": 0.0186, "step": 54000 }, { "grad_norm": 0.21034841239452362, "learning_rate": 4.75059494523729e-05, "loss": 0.0157, "step": 54010 }, { "grad_norm": 0.1734495908021927, "learning_rate": 4.748943547379843e-05, "loss": 0.0166, "step": 54020 }, { "grad_norm": 0.1887338012456894, "learning_rate": 4.7472921769775535e-05, "loss": 0.017, "step": 54030 }, { "grad_norm": 0.15205451846122742, "learning_rate": 4.745640834211011e-05, "loss": 0.0149, "step": 54040 }, { "grad_norm": 0.18474775552749634, "learning_rate": 4.7439895192608054e-05, "loss": 0.0163, "step": 54050 }, { "grad_norm": 0.18876497447490692, "learning_rate": 4.7423382323075186e-05, "loss": 0.0206, "step": 54060 }, { "grad_norm": 0.1984294205904007, "learning_rate": 4.740686973531737e-05, "loss": 0.0178, "step": 54070 }, { "grad_norm": 0.18508392572402954, "learning_rate": 4.7390357431140375e-05, "loss": 0.0234, "step": 54080 }, { "grad_norm": 0.1717800796031952, "learning_rate": 4.737384541234998e-05, "loss": 0.019, "step": 54090 }, { "grad_norm": 0.18903054296970367, "learning_rate": 4.735733368075189e-05, "loss": 0.0154, "step": 54100 }, { "grad_norm": 0.21525441110134125, "learning_rate": 4.734082223815182e-05, "loss": 0.0164, "step": 54110 }, { "grad_norm": 0.16642457246780396, "learning_rate": 4.7324311086355436e-05, "loss": 0.0173, "step": 54120 }, { "grad_norm": 0.16800543665885925, "learning_rate": 4.730780022716836e-05, "loss": 0.0159, "step": 54130 }, { "grad_norm": 0.19917382299900055, "learning_rate": 4.729128966239619e-05, "loss": 0.0162, "step": 54140 }, { "grad_norm": 0.16500355303287506, "learning_rate": 4.7274779393844536e-05, "loss": 0.0156, "step": 54150 }, { "grad_norm": 0.19638456404209137, "learning_rate": 4.725826942331891e-05, "loss": 0.0171, "step": 54160 }, { "grad_norm": 0.19084744155406952, "learning_rate": 4.724175975262481e-05, "loss": 0.0183, "step": 54170 }, { "grad_norm": 0.19114351272583008, "learning_rate": 4.722525038356772e-05, "loss": 0.0159, "step": 54180 }, { "grad_norm": 0.1777985841035843, "learning_rate": 4.7208741317953075e-05, "loss": 0.0169, "step": 54190 }, { "grad_norm": 0.2240452915430069, "learning_rate": 4.7192232557586264e-05, "loss": 0.0202, "step": 54200 }, { "grad_norm": 0.2656185030937195, "learning_rate": 4.717572410427269e-05, "loss": 0.0185, "step": 54210 }, { "grad_norm": 0.19412823021411896, "learning_rate": 4.71592159598177e-05, "loss": 0.0146, "step": 54220 }, { "grad_norm": 0.21830104291439056, "learning_rate": 4.714270812602657e-05, "loss": 0.0177, "step": 54230 }, { "grad_norm": 0.390792578458786, "learning_rate": 4.712620060470458e-05, "loss": 0.0178, "step": 54240 }, { "grad_norm": 0.18374556303024292, "learning_rate": 4.7109693397656986e-05, "loss": 0.0185, "step": 54250 }, { "grad_norm": 0.171469584107399, "learning_rate": 4.709318650668894e-05, "loss": 0.0191, "step": 54260 }, { "grad_norm": 0.22488848865032196, "learning_rate": 4.7076679933605676e-05, "loss": 0.0197, "step": 54270 }, { "grad_norm": 0.18879352509975433, "learning_rate": 4.70601736802123e-05, "loss": 0.0152, "step": 54280 }, { "grad_norm": 0.1864113211631775, "learning_rate": 4.704366774831391e-05, "loss": 0.0204, "step": 54290 }, { "grad_norm": 0.23545539379119873, "learning_rate": 4.702716213971557e-05, "loss": 0.0174, "step": 54300 }, { "grad_norm": 0.17894528806209564, "learning_rate": 4.70106568562223e-05, "loss": 0.0167, "step": 54310 }, { "grad_norm": 0.1563006490468979, "learning_rate": 4.6994151899639105e-05, "loss": 0.0155, "step": 54320 }, { "grad_norm": 0.19676867127418518, "learning_rate": 4.697764727177093e-05, "loss": 0.0177, "step": 54330 }, { "grad_norm": 0.22526952624320984, "learning_rate": 4.696114297442272e-05, "loss": 0.018, "step": 54340 }, { "grad_norm": 0.22748644649982452, "learning_rate": 4.694463900939933e-05, "loss": 0.0148, "step": 54350 }, { "grad_norm": 0.2113131582736969, "learning_rate": 4.692813537850564e-05, "loss": 0.0175, "step": 54360 }, { "grad_norm": 0.2060922235250473, "learning_rate": 4.6911632083546445e-05, "loss": 0.0189, "step": 54370 }, { "grad_norm": 0.20285840332508087, "learning_rate": 4.6895129126326525e-05, "loss": 0.0153, "step": 54380 }, { "grad_norm": 0.2164228856563568, "learning_rate": 4.687862650865059e-05, "loss": 0.0172, "step": 54390 }, { "grad_norm": 0.2488373965024948, "learning_rate": 4.686212423232339e-05, "loss": 0.0163, "step": 54400 }, { "grad_norm": 0.18211355805397034, "learning_rate": 4.6845622299149565e-05, "loss": 0.0181, "step": 54410 }, { "grad_norm": 0.17868433892726898, "learning_rate": 4.682912071093374e-05, "loss": 0.014, "step": 54420 }, { "grad_norm": 0.19152849912643433, "learning_rate": 4.681261946948052e-05, "loss": 0.016, "step": 54430 }, { "grad_norm": 0.2003607451915741, "learning_rate": 4.679611857659443e-05, "loss": 0.0156, "step": 54440 }, { "grad_norm": 0.1863427311182022, "learning_rate": 4.6779618034079975e-05, "loss": 0.0156, "step": 54450 }, { "grad_norm": 0.19828537106513977, "learning_rate": 4.6763117843741675e-05, "loss": 0.0165, "step": 54460 }, { "grad_norm": 0.14001025259494781, "learning_rate": 4.674661800738394e-05, "loss": 0.0167, "step": 54470 }, { "grad_norm": 0.1589716672897339, "learning_rate": 4.673011852681117e-05, "loss": 0.0156, "step": 54480 }, { "grad_norm": 0.18561778962612152, "learning_rate": 4.671361940382772e-05, "loss": 0.0164, "step": 54490 }, { "grad_norm": 0.16262993216514587, "learning_rate": 4.669712064023791e-05, "loss": 0.0171, "step": 54500 }, { "grad_norm": 0.2106868326663971, "learning_rate": 4.668062223784602e-05, "loss": 0.0165, "step": 54510 }, { "grad_norm": 0.14608129858970642, "learning_rate": 4.6664124198456274e-05, "loss": 0.0157, "step": 54520 }, { "grad_norm": 0.23877482116222382, "learning_rate": 4.6647626523872905e-05, "loss": 0.0171, "step": 54530 }, { "grad_norm": 0.1785871535539627, "learning_rate": 4.6631129215900066e-05, "loss": 0.0159, "step": 54540 }, { "grad_norm": 0.17027457058429718, "learning_rate": 4.661463227634187e-05, "loss": 0.013, "step": 54550 }, { "grad_norm": 0.1633918732404709, "learning_rate": 4.659813570700238e-05, "loss": 0.0182, "step": 54560 }, { "grad_norm": 0.20948933064937592, "learning_rate": 4.658163950968568e-05, "loss": 0.0144, "step": 54570 }, { "grad_norm": 0.16376475989818573, "learning_rate": 4.6565143686195726e-05, "loss": 0.0155, "step": 54580 }, { "grad_norm": 0.21700245141983032, "learning_rate": 4.654864823833649e-05, "loss": 0.0149, "step": 54590 }, { "grad_norm": 0.2648756802082062, "learning_rate": 4.65321531679119e-05, "loss": 0.016, "step": 54600 }, { "grad_norm": 0.14976617693901062, "learning_rate": 4.6515658476725834e-05, "loss": 0.0157, "step": 54610 }, { "grad_norm": 0.1961222141981125, "learning_rate": 4.649916416658212e-05, "loss": 0.0182, "step": 54620 }, { "grad_norm": 0.19898360967636108, "learning_rate": 4.648267023928455e-05, "loss": 0.0189, "step": 54630 }, { "grad_norm": 0.16800329089164734, "learning_rate": 4.646617669663687e-05, "loss": 0.0166, "step": 54640 }, { "grad_norm": 0.26305949687957764, "learning_rate": 4.644968354044277e-05, "loss": 0.0165, "step": 54650 }, { "grad_norm": 0.16808435320854187, "learning_rate": 4.643319077250596e-05, "loss": 0.015, "step": 54660 }, { "grad_norm": 0.14658187329769135, "learning_rate": 4.6416698394630055e-05, "loss": 0.0136, "step": 54670 }, { "grad_norm": 0.20560169219970703, "learning_rate": 4.640020640861862e-05, "loss": 0.0146, "step": 54680 }, { "grad_norm": 0.2103135585784912, "learning_rate": 4.6383714816275195e-05, "loss": 0.0167, "step": 54690 }, { "grad_norm": 0.24741913378238678, "learning_rate": 4.636722361940329e-05, "loss": 0.0166, "step": 54700 }, { "grad_norm": 0.15965482592582703, "learning_rate": 4.635073281980632e-05, "loss": 0.014, "step": 54710 }, { "grad_norm": 0.19542260468006134, "learning_rate": 4.633424241928775e-05, "loss": 0.0177, "step": 54720 }, { "grad_norm": 0.17452654242515564, "learning_rate": 4.6317752419650914e-05, "loss": 0.0164, "step": 54730 }, { "grad_norm": 0.1859019249677658, "learning_rate": 4.630126282269914e-05, "loss": 0.0206, "step": 54740 }, { "grad_norm": 0.178471639752388, "learning_rate": 4.628477363023569e-05, "loss": 0.0162, "step": 54750 }, { "grad_norm": 0.19956208765506744, "learning_rate": 4.626828484406383e-05, "loss": 0.0203, "step": 54760 }, { "grad_norm": 0.17324621975421906, "learning_rate": 4.625179646598669e-05, "loss": 0.0168, "step": 54770 }, { "grad_norm": 0.263126403093338, "learning_rate": 4.6235308497807487e-05, "loss": 0.0221, "step": 54780 }, { "grad_norm": 0.217945858836174, "learning_rate": 4.6218820941329276e-05, "loss": 0.0177, "step": 54790 }, { "grad_norm": 0.2215087115764618, "learning_rate": 4.620233379835513e-05, "loss": 0.0174, "step": 54800 }, { "grad_norm": 0.1744711846113205, "learning_rate": 4.618584707068804e-05, "loss": 0.0177, "step": 54810 }, { "grad_norm": 0.18095608055591583, "learning_rate": 4.616936076013099e-05, "loss": 0.0186, "step": 54820 }, { "grad_norm": 0.18204133212566376, "learning_rate": 4.615287486848688e-05, "loss": 0.0162, "step": 54830 }, { "grad_norm": 0.18079626560211182, "learning_rate": 4.6136389397558584e-05, "loss": 0.0135, "step": 54840 }, { "grad_norm": 0.14365604519844055, "learning_rate": 4.6119904349148935e-05, "loss": 0.0138, "step": 54850 }, { "grad_norm": 0.2433844953775406, "learning_rate": 4.6103419725060726e-05, "loss": 0.0188, "step": 54860 }, { "grad_norm": 0.21544356644153595, "learning_rate": 4.608693552709667e-05, "loss": 0.0169, "step": 54870 }, { "grad_norm": 0.16844667494297028, "learning_rate": 4.607045175705947e-05, "loss": 0.0166, "step": 54880 }, { "grad_norm": 0.19685682654380798, "learning_rate": 4.605396841675176e-05, "loss": 0.0167, "step": 54890 }, { "grad_norm": 0.24480408430099487, "learning_rate": 4.6037485507976094e-05, "loss": 0.0152, "step": 54900 }, { "grad_norm": 0.17459289729595184, "learning_rate": 4.602100303253509e-05, "loss": 0.0154, "step": 54910 }, { "grad_norm": 0.2094302475452423, "learning_rate": 4.600452099223122e-05, "loss": 0.0141, "step": 54920 }, { "grad_norm": 0.17147621512413025, "learning_rate": 4.598803938886691e-05, "loss": 0.0166, "step": 54930 }, { "grad_norm": 0.16393151879310608, "learning_rate": 4.597155822424459e-05, "loss": 0.016, "step": 54940 }, { "grad_norm": 0.17116300761699677, "learning_rate": 4.59550775001666e-05, "loss": 0.0158, "step": 54950 }, { "grad_norm": 0.17330302298069, "learning_rate": 4.5938597218435234e-05, "loss": 0.0179, "step": 54960 }, { "grad_norm": 0.22719456255435944, "learning_rate": 4.592211738085279e-05, "loss": 0.0164, "step": 54970 }, { "grad_norm": 0.15800820291042328, "learning_rate": 4.590563798922146e-05, "loss": 0.012, "step": 54980 }, { "grad_norm": 0.20085473358631134, "learning_rate": 4.5889159045343404e-05, "loss": 0.0153, "step": 54990 }, { "grad_norm": 0.16407988965511322, "learning_rate": 4.5872680551020734e-05, "loss": 0.0185, "step": 55000 }, { "grad_norm": 0.23417550325393677, "learning_rate": 4.585620250805552e-05, "loss": 0.0155, "step": 55010 }, { "grad_norm": 0.21940352022647858, "learning_rate": 4.583972491824975e-05, "loss": 0.0162, "step": 55020 }, { "grad_norm": 0.17909404635429382, "learning_rate": 4.582324778340541e-05, "loss": 0.0162, "step": 55030 }, { "grad_norm": 0.20757460594177246, "learning_rate": 4.5806771105324424e-05, "loss": 0.0176, "step": 55040 }, { "grad_norm": 0.16943901777267456, "learning_rate": 4.579029488580865e-05, "loss": 0.0159, "step": 55050 }, { "grad_norm": 0.15559925138950348, "learning_rate": 4.5773819126659876e-05, "loss": 0.0171, "step": 55060 }, { "grad_norm": 0.17203466594219208, "learning_rate": 4.575734382967991e-05, "loss": 0.02, "step": 55070 }, { "grad_norm": 0.27301254868507385, "learning_rate": 4.574086899667044e-05, "loss": 0.0164, "step": 55080 }, { "grad_norm": 0.19290190935134888, "learning_rate": 4.5724394629433106e-05, "loss": 0.0184, "step": 55090 }, { "grad_norm": 0.2562234699726105, "learning_rate": 4.570792072976958e-05, "loss": 0.0173, "step": 55100 }, { "grad_norm": 0.14980903267860413, "learning_rate": 4.569144729948138e-05, "loss": 0.0198, "step": 55110 }, { "grad_norm": 0.17479680478572845, "learning_rate": 4.5674974340370033e-05, "loss": 0.0156, "step": 55120 }, { "grad_norm": 0.23002170026302338, "learning_rate": 4.565850185423699e-05, "loss": 0.0159, "step": 55130 }, { "grad_norm": 0.1557602733373642, "learning_rate": 4.564202984288365e-05, "loss": 0.0141, "step": 55140 }, { "grad_norm": 0.38761040568351746, "learning_rate": 4.5625558308111356e-05, "loss": 0.0157, "step": 55150 }, { "grad_norm": 0.18040840327739716, "learning_rate": 4.560908725172145e-05, "loss": 0.0169, "step": 55160 }, { "grad_norm": 0.19043946266174316, "learning_rate": 4.559261667551516e-05, "loss": 0.0186, "step": 55170 }, { "grad_norm": 0.15682217478752136, "learning_rate": 4.5576146581293685e-05, "loss": 0.0162, "step": 55180 }, { "grad_norm": 0.18611866235733032, "learning_rate": 4.5559676970858164e-05, "loss": 0.0155, "step": 55190 }, { "grad_norm": 0.16201366484165192, "learning_rate": 4.5543207846009705e-05, "loss": 0.0143, "step": 55200 }, { "grad_norm": 0.27235785126686096, "learning_rate": 4.552673920854933e-05, "loss": 0.0236, "step": 55210 }, { "grad_norm": 0.11764062941074371, "learning_rate": 4.5510271060278006e-05, "loss": 0.0151, "step": 55220 }, { "grad_norm": 0.18296679854393005, "learning_rate": 4.5493803402996704e-05, "loss": 0.0133, "step": 55230 }, { "grad_norm": 0.21420106291770935, "learning_rate": 4.5477336238506295e-05, "loss": 0.0191, "step": 55240 }, { "grad_norm": 0.13700559735298157, "learning_rate": 4.546086956860759e-05, "loss": 0.0168, "step": 55250 }, { "grad_norm": 0.14867180585861206, "learning_rate": 4.5444403395101374e-05, "loss": 0.0163, "step": 55260 }, { "grad_norm": 0.19582369923591614, "learning_rate": 4.542793771978834e-05, "loss": 0.0174, "step": 55270 }, { "grad_norm": 0.24492168426513672, "learning_rate": 4.541147254446916e-05, "loss": 0.0143, "step": 55280 }, { "grad_norm": 0.20637363195419312, "learning_rate": 4.539500787094445e-05, "loss": 0.0149, "step": 55290 }, { "grad_norm": 0.1726594865322113, "learning_rate": 4.537854370101475e-05, "loss": 0.02, "step": 55300 }, { "grad_norm": 0.15802980959415436, "learning_rate": 4.5362080036480566e-05, "loss": 0.02, "step": 55310 }, { "grad_norm": 0.14418324828147888, "learning_rate": 4.534561687914235e-05, "loss": 0.016, "step": 55320 }, { "grad_norm": 0.17343266308307648, "learning_rate": 4.5329154230800466e-05, "loss": 0.0166, "step": 55330 }, { "grad_norm": 0.17485859990119934, "learning_rate": 4.531269209325525e-05, "loss": 0.0171, "step": 55340 }, { "grad_norm": 0.1661103069782257, "learning_rate": 4.529623046830696e-05, "loss": 0.0151, "step": 55350 }, { "grad_norm": 0.18986743688583374, "learning_rate": 4.5279769357755866e-05, "loss": 0.0195, "step": 55360 }, { "grad_norm": 0.1466992348432541, "learning_rate": 4.5263308763402084e-05, "loss": 0.0174, "step": 55370 }, { "grad_norm": 0.12013811618089676, "learning_rate": 4.524684868704574e-05, "loss": 0.0149, "step": 55380 }, { "grad_norm": 0.21365056931972504, "learning_rate": 4.523038913048688e-05, "loss": 0.0163, "step": 55390 }, { "grad_norm": 0.1538025140762329, "learning_rate": 4.521393009552549e-05, "loss": 0.0139, "step": 55400 }, { "grad_norm": 0.21848690509796143, "learning_rate": 4.5197471583961484e-05, "loss": 0.0152, "step": 55410 }, { "grad_norm": 0.19282345473766327, "learning_rate": 4.51810135975948e-05, "loss": 0.0145, "step": 55420 }, { "grad_norm": 0.16655981540679932, "learning_rate": 4.516455613822521e-05, "loss": 0.0161, "step": 55430 }, { "grad_norm": 0.14355385303497314, "learning_rate": 4.514809920765249e-05, "loss": 0.0173, "step": 55440 }, { "grad_norm": 0.1646609753370285, "learning_rate": 4.5131642807676356e-05, "loss": 0.0168, "step": 55450 }, { "grad_norm": 0.16385625302791595, "learning_rate": 4.511518694009644e-05, "loss": 0.0169, "step": 55460 }, { "grad_norm": 0.17505989968776703, "learning_rate": 4.509873160671231e-05, "loss": 0.0176, "step": 55470 }, { "grad_norm": 0.1648680865764618, "learning_rate": 4.508227680932356e-05, "loss": 0.0176, "step": 55480 }, { "grad_norm": 0.15015144646167755, "learning_rate": 4.5065822549729605e-05, "loss": 0.017, "step": 55490 }, { "grad_norm": 0.21232379972934723, "learning_rate": 4.504936882972989e-05, "loss": 0.0148, "step": 55500 }, { "grad_norm": 0.1790974736213684, "learning_rate": 4.503291565112374e-05, "loss": 0.0178, "step": 55510 }, { "grad_norm": 0.2130909413099289, "learning_rate": 4.501646301571048e-05, "loss": 0.0205, "step": 55520 }, { "grad_norm": 0.17320480942726135, "learning_rate": 4.500001092528933e-05, "loss": 0.0207, "step": 55530 }, { "grad_norm": 0.17433936893939972, "learning_rate": 4.4983559381659455e-05, "loss": 0.0192, "step": 55540 }, { "grad_norm": 0.20262356102466583, "learning_rate": 4.496710838661999e-05, "loss": 0.0245, "step": 55550 }, { "grad_norm": 0.20183105766773224, "learning_rate": 4.495065794196999e-05, "loss": 0.0169, "step": 55560 }, { "grad_norm": 0.1888856440782547, "learning_rate": 4.4934208049508455e-05, "loss": 0.0167, "step": 55570 }, { "grad_norm": 0.18051043152809143, "learning_rate": 4.4917758711034303e-05, "loss": 0.0178, "step": 55580 }, { "grad_norm": 0.14063744246959686, "learning_rate": 4.490130992834642e-05, "loss": 0.0158, "step": 55590 }, { "grad_norm": 0.163282573223114, "learning_rate": 4.488486170324359e-05, "loss": 0.0147, "step": 55600 }, { "grad_norm": 0.17111581563949585, "learning_rate": 4.486841403752463e-05, "loss": 0.0172, "step": 55610 }, { "grad_norm": 0.22532562911510468, "learning_rate": 4.4851966932988174e-05, "loss": 0.0186, "step": 55620 }, { "grad_norm": 0.1910749077796936, "learning_rate": 4.4835520391432894e-05, "loss": 0.0161, "step": 55630 }, { "grad_norm": 0.14608260989189148, "learning_rate": 4.481907441465733e-05, "loss": 0.0185, "step": 55640 }, { "grad_norm": 0.2285613715648651, "learning_rate": 4.480262900446e-05, "loss": 0.0182, "step": 55650 }, { "grad_norm": 0.19336958229541779, "learning_rate": 4.478618416263933e-05, "loss": 0.0188, "step": 55660 }, { "grad_norm": 0.1779138147830963, "learning_rate": 4.476973989099373e-05, "loss": 0.0176, "step": 55670 }, { "grad_norm": 0.1937202513217926, "learning_rate": 4.475329619132153e-05, "loss": 0.0168, "step": 55680 }, { "grad_norm": 0.21476958692073822, "learning_rate": 4.4736853065420956e-05, "loss": 0.0175, "step": 55690 }, { "grad_norm": 0.19003282487392426, "learning_rate": 4.472041051509023e-05, "loss": 0.0167, "step": 55700 }, { "grad_norm": 0.2285638451576233, "learning_rate": 4.470396854212746e-05, "loss": 0.0154, "step": 55710 }, { "grad_norm": 0.18173381686210632, "learning_rate": 4.468752714833072e-05, "loss": 0.0148, "step": 55720 }, { "grad_norm": 0.20938017964363098, "learning_rate": 4.467108633549803e-05, "loss": 0.0184, "step": 55730 }, { "grad_norm": 0.15111950039863586, "learning_rate": 4.4654646105427335e-05, "loss": 0.0152, "step": 55740 }, { "grad_norm": 0.2500550150871277, "learning_rate": 4.463820645991651e-05, "loss": 0.0198, "step": 55750 }, { "grad_norm": 0.18138258159160614, "learning_rate": 4.462176740076334e-05, "loss": 0.0159, "step": 55760 }, { "grad_norm": 0.15785132348537445, "learning_rate": 4.4605328929765624e-05, "loss": 0.016, "step": 55770 }, { "grad_norm": 0.19788606464862823, "learning_rate": 4.458889104872102e-05, "loss": 0.024, "step": 55780 }, { "grad_norm": 0.15143471956253052, "learning_rate": 4.457245375942715e-05, "loss": 0.0165, "step": 55790 }, { "grad_norm": 0.20116327702999115, "learning_rate": 4.455601706368157e-05, "loss": 0.0182, "step": 55800 }, { "grad_norm": 0.1609765589237213, "learning_rate": 4.453958096328179e-05, "loss": 0.0159, "step": 55810 }, { "grad_norm": 0.2138499617576599, "learning_rate": 4.452314546002523e-05, "loss": 0.0161, "step": 55820 }, { "grad_norm": 0.247023805975914, "learning_rate": 4.4506710555709246e-05, "loss": 0.0174, "step": 55830 }, { "grad_norm": 0.2467336505651474, "learning_rate": 4.449027625213114e-05, "loss": 0.0192, "step": 55840 }, { "grad_norm": 0.1826799362897873, "learning_rate": 4.4473842551088135e-05, "loss": 0.0183, "step": 55850 }, { "grad_norm": 0.1690484881401062, "learning_rate": 4.445740945437738e-05, "loss": 0.0159, "step": 55860 }, { "grad_norm": 0.2069840431213379, "learning_rate": 4.444097696379602e-05, "loss": 0.0207, "step": 55870 }, { "grad_norm": 0.26357218623161316, "learning_rate": 4.442454508114106e-05, "loss": 0.0187, "step": 55880 }, { "grad_norm": 0.1821317821741104, "learning_rate": 4.4408113808209465e-05, "loss": 0.0166, "step": 55890 }, { "grad_norm": 0.147181436419487, "learning_rate": 4.439168314679813e-05, "loss": 0.014, "step": 55900 }, { "grad_norm": 0.2288040816783905, "learning_rate": 4.43752530987039e-05, "loss": 0.0132, "step": 55910 }, { "grad_norm": 0.12918716669082642, "learning_rate": 4.4358823665723515e-05, "loss": 0.0143, "step": 55920 }, { "grad_norm": 0.1703684777021408, "learning_rate": 4.434239484965371e-05, "loss": 0.0167, "step": 55930 }, { "grad_norm": 0.25189271569252014, "learning_rate": 4.4325966652291103e-05, "loss": 0.0182, "step": 55940 }, { "grad_norm": 0.5116862058639526, "learning_rate": 4.430953907543225e-05, "loss": 0.0174, "step": 55950 }, { "grad_norm": 0.16587811708450317, "learning_rate": 4.4293112120873645e-05, "loss": 0.0152, "step": 55960 }, { "grad_norm": 0.12927602231502533, "learning_rate": 4.427668579041172e-05, "loss": 0.0129, "step": 55970 }, { "grad_norm": 0.18033744394779205, "learning_rate": 4.4260260085842816e-05, "loss": 0.0155, "step": 55980 }, { "grad_norm": 0.21116602420806885, "learning_rate": 4.4243835008963256e-05, "loss": 0.0211, "step": 55990 }, { "grad_norm": 0.2562488913536072, "learning_rate": 4.422741056156925e-05, "loss": 0.0145, "step": 56000 }, { "grad_norm": 0.29592165350914, "learning_rate": 4.421098674545693e-05, "loss": 0.0159, "step": 56010 }, { "grad_norm": 0.2032608687877655, "learning_rate": 4.419456356242241e-05, "loss": 0.0188, "step": 56020 }, { "grad_norm": 0.17050588130950928, "learning_rate": 4.4178141014261685e-05, "loss": 0.0177, "step": 56030 }, { "grad_norm": 0.22985588014125824, "learning_rate": 4.416171910277071e-05, "loss": 0.0161, "step": 56040 }, { "grad_norm": 0.18488436937332153, "learning_rate": 4.414529782974533e-05, "loss": 0.0165, "step": 56050 }, { "grad_norm": 0.20000869035720825, "learning_rate": 4.41288771969814e-05, "loss": 0.0194, "step": 56060 }, { "grad_norm": 0.22547723352909088, "learning_rate": 4.411245720627464e-05, "loss": 0.0152, "step": 56070 }, { "grad_norm": 0.1810460388660431, "learning_rate": 4.40960378594207e-05, "loss": 0.013, "step": 56080 }, { "grad_norm": 0.12490404397249222, "learning_rate": 4.407961915821519e-05, "loss": 0.014, "step": 56090 }, { "grad_norm": 0.20378383994102478, "learning_rate": 4.406320110445362e-05, "loss": 0.0184, "step": 56100 }, { "grad_norm": 0.1538500338792801, "learning_rate": 4.404678369993144e-05, "loss": 0.0188, "step": 56110 }, { "grad_norm": 0.22039446234703064, "learning_rate": 4.403036694644406e-05, "loss": 0.0147, "step": 56120 }, { "grad_norm": 0.16186343133449554, "learning_rate": 4.4013950845786764e-05, "loss": 0.0158, "step": 56130 }, { "grad_norm": 0.2096690684556961, "learning_rate": 4.399753539975482e-05, "loss": 0.0183, "step": 56140 }, { "grad_norm": 0.18689119815826416, "learning_rate": 4.398112061014337e-05, "loss": 0.0158, "step": 56150 }, { "grad_norm": 0.18131409585475922, "learning_rate": 4.396470647874753e-05, "loss": 0.018, "step": 56160 }, { "grad_norm": 0.23265260457992554, "learning_rate": 4.394829300736229e-05, "loss": 0.02, "step": 56170 }, { "grad_norm": 0.1937732994556427, "learning_rate": 4.393188019778265e-05, "loss": 0.0161, "step": 56180 }, { "grad_norm": 0.2153562307357788, "learning_rate": 4.391546805180347e-05, "loss": 0.0184, "step": 56190 }, { "grad_norm": 0.20418979227542877, "learning_rate": 4.389905657121955e-05, "loss": 0.0154, "step": 56200 }, { "grad_norm": 0.19227711856365204, "learning_rate": 4.3882645757825635e-05, "loss": 0.0144, "step": 56210 }, { "grad_norm": 0.2100050002336502, "learning_rate": 4.386623561341637e-05, "loss": 0.0185, "step": 56220 }, { "grad_norm": 0.19162452220916748, "learning_rate": 4.384982613978637e-05, "loss": 0.0153, "step": 56230 }, { "grad_norm": 0.1892939954996109, "learning_rate": 4.383341733873012e-05, "loss": 0.0147, "step": 56240 }, { "grad_norm": 0.21767111122608185, "learning_rate": 4.381700921204209e-05, "loss": 0.0182, "step": 56250 }, { "grad_norm": 0.16648732125759125, "learning_rate": 4.380060176151663e-05, "loss": 0.0128, "step": 56260 }, { "grad_norm": 0.15671314299106598, "learning_rate": 4.378419498894805e-05, "loss": 0.0163, "step": 56270 }, { "grad_norm": 0.24303171038627625, "learning_rate": 4.376778889613056e-05, "loss": 0.0175, "step": 56280 }, { "grad_norm": 0.2018212080001831, "learning_rate": 4.375138348485831e-05, "loss": 0.0148, "step": 56290 }, { "grad_norm": 0.19796763360500336, "learning_rate": 4.3734978756925346e-05, "loss": 0.0151, "step": 56300 }, { "grad_norm": 0.1636270433664322, "learning_rate": 4.3718574714125706e-05, "loss": 0.0143, "step": 56310 }, { "grad_norm": 0.19144392013549805, "learning_rate": 4.370217135825329e-05, "loss": 0.0155, "step": 56320 }, { "grad_norm": 0.178591787815094, "learning_rate": 4.368576869110194e-05, "loss": 0.0144, "step": 56330 }, { "grad_norm": 0.1873578280210495, "learning_rate": 4.366936671446544e-05, "loss": 0.0158, "step": 56340 }, { "grad_norm": 0.20527365803718567, "learning_rate": 4.3652965430137474e-05, "loss": 0.0154, "step": 56350 }, { "grad_norm": 0.23350025713443756, "learning_rate": 4.3636564839911646e-05, "loss": 0.0198, "step": 56360 }, { "grad_norm": 0.20317845046520233, "learning_rate": 4.3620164945581545e-05, "loss": 0.0159, "step": 56370 }, { "grad_norm": 0.1697932630777359, "learning_rate": 4.360376574894061e-05, "loss": 0.0137, "step": 56380 }, { "grad_norm": 0.18813535571098328, "learning_rate": 4.358736725178224e-05, "loss": 0.0124, "step": 56390 }, { "grad_norm": 0.15307697653770447, "learning_rate": 4.357096945589974e-05, "loss": 0.015, "step": 56400 }, { "grad_norm": 0.14065097272396088, "learning_rate": 4.355457236308636e-05, "loss": 0.0133, "step": 56410 }, { "grad_norm": 0.15641914308071136, "learning_rate": 4.353817597513526e-05, "loss": 0.0165, "step": 56420 }, { "grad_norm": 0.23910897970199585, "learning_rate": 4.352178029383948e-05, "loss": 0.0164, "step": 56430 }, { "grad_norm": 0.18094559013843536, "learning_rate": 4.3505385320992105e-05, "loss": 0.0157, "step": 56440 }, { "grad_norm": 0.25835904479026794, "learning_rate": 4.348899105838602e-05, "loss": 0.0174, "step": 56450 }, { "grad_norm": 0.18201123178005219, "learning_rate": 4.3472597507814087e-05, "loss": 0.0174, "step": 56460 }, { "grad_norm": 0.2091294378042221, "learning_rate": 4.3456204671069066e-05, "loss": 0.0185, "step": 56470 }, { "grad_norm": 0.23210905492305756, "learning_rate": 4.343981254994367e-05, "loss": 0.017, "step": 56480 }, { "grad_norm": 0.2238859236240387, "learning_rate": 4.34234211462305e-05, "loss": 0.0152, "step": 56490 }, { "grad_norm": 0.16099010407924652, "learning_rate": 4.340703046172213e-05, "loss": 0.0144, "step": 56500 }, { "grad_norm": 0.30909478664398193, "learning_rate": 4.339064049821097e-05, "loss": 0.0117, "step": 56510 }, { "grad_norm": 0.15934142470359802, "learning_rate": 4.3374251257489446e-05, "loss": 0.0138, "step": 56520 }, { "grad_norm": 0.12705476582050323, "learning_rate": 4.335786274134984e-05, "loss": 0.0136, "step": 56530 }, { "grad_norm": 0.19706378877162933, "learning_rate": 4.334147495158438e-05, "loss": 0.0169, "step": 56540 }, { "grad_norm": 0.17132748663425446, "learning_rate": 4.33250878899852e-05, "loss": 0.0172, "step": 56550 }, { "grad_norm": 0.22682252526283264, "learning_rate": 4.330870155834437e-05, "loss": 0.0145, "step": 56560 }, { "grad_norm": 0.2425321340560913, "learning_rate": 4.3292315958453884e-05, "loss": 0.017, "step": 56570 }, { "grad_norm": 0.17603662610054016, "learning_rate": 4.3275931092105646e-05, "loss": 0.0134, "step": 56580 }, { "grad_norm": 0.16061504185199738, "learning_rate": 4.325954696109147e-05, "loss": 0.0151, "step": 56590 }, { "grad_norm": 0.19623729586601257, "learning_rate": 4.32431635672031e-05, "loss": 0.0178, "step": 56600 }, { "grad_norm": 0.17855894565582275, "learning_rate": 4.3226780912232214e-05, "loss": 0.0151, "step": 56610 }, { "grad_norm": 0.15507441759109497, "learning_rate": 4.321039899797036e-05, "loss": 0.0196, "step": 56620 }, { "grad_norm": 0.22323891520500183, "learning_rate": 4.319401782620908e-05, "loss": 0.0148, "step": 56630 }, { "grad_norm": 0.2340627759695053, "learning_rate": 4.317763739873978e-05, "loss": 0.0178, "step": 56640 }, { "grad_norm": 0.1868567317724228, "learning_rate": 4.31612577173538e-05, "loss": 0.0169, "step": 56650 }, { "grad_norm": 0.21123163402080536, "learning_rate": 4.314487878384239e-05, "loss": 0.0163, "step": 56660 }, { "grad_norm": 0.21716007590293884, "learning_rate": 4.3128500599996736e-05, "loss": 0.0166, "step": 56670 }, { "grad_norm": 0.2839692234992981, "learning_rate": 4.31121231676079e-05, "loss": 0.0166, "step": 56680 }, { "grad_norm": 0.19194217026233673, "learning_rate": 4.309574648846694e-05, "loss": 0.0162, "step": 56690 }, { "grad_norm": 0.13896524906158447, "learning_rate": 4.3079370564364755e-05, "loss": 0.0153, "step": 56700 }, { "grad_norm": 0.21594253182411194, "learning_rate": 4.306299539709221e-05, "loss": 0.018, "step": 56710 }, { "grad_norm": 0.2042657434940338, "learning_rate": 4.3046620988440046e-05, "loss": 0.0175, "step": 56720 }, { "grad_norm": 0.266496866941452, "learning_rate": 4.303024734019897e-05, "loss": 0.0183, "step": 56730 }, { "grad_norm": 0.2193050980567932, "learning_rate": 4.3013874454159565e-05, "loss": 0.0168, "step": 56740 }, { "grad_norm": 0.2276228815317154, "learning_rate": 4.299750233211233e-05, "loss": 0.0194, "step": 56750 }, { "grad_norm": 0.12719108164310455, "learning_rate": 4.2981130975847715e-05, "loss": 0.0175, "step": 56760 }, { "grad_norm": 0.1901039332151413, "learning_rate": 4.2964760387156076e-05, "loss": 0.016, "step": 56770 }, { "grad_norm": 0.14624975621700287, "learning_rate": 4.294839056782767e-05, "loss": 0.0183, "step": 56780 }, { "grad_norm": 0.205959752202034, "learning_rate": 4.293202151965266e-05, "loss": 0.0162, "step": 56790 }, { "grad_norm": 0.16847892105579376, "learning_rate": 4.291565324442115e-05, "loss": 0.0174, "step": 56800 }, { "grad_norm": 0.24952447414398193, "learning_rate": 4.2899285743923135e-05, "loss": 0.0195, "step": 56810 }, { "grad_norm": 0.21304535865783691, "learning_rate": 4.288291901994857e-05, "loss": 0.0187, "step": 56820 }, { "grad_norm": 0.2193499356508255, "learning_rate": 4.286655307428729e-05, "loss": 0.0157, "step": 56830 }, { "grad_norm": 0.19686822593212128, "learning_rate": 4.285018790872903e-05, "loss": 0.0144, "step": 56840 }, { "grad_norm": 0.15558187663555145, "learning_rate": 4.283382352506347e-05, "loss": 0.0193, "step": 56850 }, { "grad_norm": 0.1967412531375885, "learning_rate": 4.2817459925080205e-05, "loss": 0.0162, "step": 56860 }, { "grad_norm": 0.19499991834163666, "learning_rate": 4.28010971105687e-05, "loss": 0.0157, "step": 56870 }, { "grad_norm": 0.16658790409564972, "learning_rate": 4.278473508331841e-05, "loss": 0.0181, "step": 56880 }, { "grad_norm": 0.19598622620105743, "learning_rate": 4.276837384511864e-05, "loss": 0.016, "step": 56890 }, { "grad_norm": 0.21009325981140137, "learning_rate": 4.275201339775864e-05, "loss": 0.0158, "step": 56900 }, { "grad_norm": 0.1564355045557022, "learning_rate": 4.273565374302756e-05, "loss": 0.0168, "step": 56910 }, { "grad_norm": 0.24144130945205688, "learning_rate": 4.2719294882714454e-05, "loss": 0.0158, "step": 56920 }, { "grad_norm": 0.22134807705879211, "learning_rate": 4.270293681860831e-05, "loss": 0.0184, "step": 56930 }, { "grad_norm": 0.19854170083999634, "learning_rate": 4.268657955249802e-05, "loss": 0.0169, "step": 56940 }, { "grad_norm": 0.3116559684276581, "learning_rate": 4.267022308617241e-05, "loss": 0.0169, "step": 56950 }, { "grad_norm": 0.14966797828674316, "learning_rate": 4.265386742142018e-05, "loss": 0.0175, "step": 56960 }, { "grad_norm": 0.18992994725704193, "learning_rate": 4.263751256002996e-05, "loss": 0.0177, "step": 56970 }, { "grad_norm": 0.23065495491027832, "learning_rate": 4.26211585037903e-05, "loss": 0.0156, "step": 56980 }, { "grad_norm": 0.218790665268898, "learning_rate": 4.2604805254489656e-05, "loss": 0.0173, "step": 56990 }, { "grad_norm": 0.1635393500328064, "learning_rate": 4.258845281391638e-05, "loss": 0.0148, "step": 57000 }, { "grad_norm": 0.160941019654274, "learning_rate": 4.257210118385876e-05, "loss": 0.0225, "step": 57010 }, { "grad_norm": 0.19115190207958221, "learning_rate": 4.2555750366104994e-05, "loss": 0.0142, "step": 57020 }, { "grad_norm": 0.1728702187538147, "learning_rate": 4.2539400362443184e-05, "loss": 0.0134, "step": 57030 }, { "grad_norm": 0.18958020210266113, "learning_rate": 4.252305117466133e-05, "loss": 0.0176, "step": 57040 }, { "grad_norm": 0.20617897808551788, "learning_rate": 4.2506702804547356e-05, "loss": 0.0177, "step": 57050 }, { "grad_norm": 0.24535949528217316, "learning_rate": 4.249035525388907e-05, "loss": 0.0154, "step": 57060 }, { "grad_norm": 0.14197397232055664, "learning_rate": 4.247400852447427e-05, "loss": 0.0173, "step": 57070 }, { "grad_norm": 0.15685583651065826, "learning_rate": 4.245766261809059e-05, "loss": 0.0165, "step": 57080 }, { "grad_norm": 0.21614664793014526, "learning_rate": 4.244131753652558e-05, "loss": 0.0191, "step": 57090 }, { "grad_norm": 0.24323725700378418, "learning_rate": 4.242497328156672e-05, "loss": 0.0157, "step": 57100 }, { "grad_norm": 0.21358448266983032, "learning_rate": 4.240862985500139e-05, "loss": 0.0148, "step": 57110 }, { "grad_norm": 0.16072198748588562, "learning_rate": 4.2392287258616894e-05, "loss": 0.014, "step": 57120 }, { "grad_norm": 0.14165745675563812, "learning_rate": 4.2375945494200396e-05, "loss": 0.0194, "step": 57130 }, { "grad_norm": 0.21684643626213074, "learning_rate": 4.2359604563539054e-05, "loss": 0.0196, "step": 57140 }, { "grad_norm": 0.16275158524513245, "learning_rate": 4.2343264468419885e-05, "loss": 0.0136, "step": 57150 }, { "grad_norm": 0.1581605076789856, "learning_rate": 4.232692521062979e-05, "loss": 0.0147, "step": 57160 }, { "grad_norm": 0.2627318799495697, "learning_rate": 4.231058679195562e-05, "loss": 0.0132, "step": 57170 }, { "grad_norm": 0.2337988168001175, "learning_rate": 4.229424921418411e-05, "loss": 0.0159, "step": 57180 }, { "grad_norm": 0.2614242434501648, "learning_rate": 4.2277912479101916e-05, "loss": 0.0168, "step": 57190 }, { "grad_norm": 0.1638471931219101, "learning_rate": 4.226157658849562e-05, "loss": 0.0175, "step": 57200 }, { "grad_norm": 0.2220000922679901, "learning_rate": 4.2245241544151664e-05, "loss": 0.0194, "step": 57210 }, { "grad_norm": 0.17201106250286102, "learning_rate": 4.2228907347856425e-05, "loss": 0.015, "step": 57220 }, { "grad_norm": 0.21755613386631012, "learning_rate": 4.2212574001396214e-05, "loss": 0.0222, "step": 57230 }, { "grad_norm": 0.17331141233444214, "learning_rate": 4.21962415065572e-05, "loss": 0.0166, "step": 57240 }, { "grad_norm": 0.16837233304977417, "learning_rate": 4.217990986512548e-05, "loss": 0.0161, "step": 57250 }, { "grad_norm": 0.23780381679534912, "learning_rate": 4.216357907888704e-05, "loss": 0.0163, "step": 57260 }, { "grad_norm": 0.16526083648204803, "learning_rate": 4.2147249149627824e-05, "loss": 0.0177, "step": 57270 }, { "grad_norm": 0.3001859486103058, "learning_rate": 4.213092007913364e-05, "loss": 0.0161, "step": 57280 }, { "grad_norm": 0.17022188007831573, "learning_rate": 4.2114591869190205e-05, "loss": 0.0154, "step": 57290 }, { "grad_norm": 0.14495688676834106, "learning_rate": 4.209826452158315e-05, "loss": 0.015, "step": 57300 }, { "grad_norm": 0.21669811010360718, "learning_rate": 4.2081938038098e-05, "loss": 0.0163, "step": 57310 }, { "grad_norm": 0.3438612222671509, "learning_rate": 4.2065612420520185e-05, "loss": 0.021, "step": 57320 }, { "grad_norm": 0.2142101675271988, "learning_rate": 4.204928767063509e-05, "loss": 0.0169, "step": 57330 }, { "grad_norm": 0.173627108335495, "learning_rate": 4.2032963790227944e-05, "loss": 0.0134, "step": 57340 }, { "grad_norm": 0.16003474593162537, "learning_rate": 4.2016640781083887e-05, "loss": 0.0132, "step": 57350 }, { "grad_norm": 0.17684441804885864, "learning_rate": 4.2000318644988004e-05, "loss": 0.0142, "step": 57360 }, { "grad_norm": 0.19110579788684845, "learning_rate": 4.198399738372524e-05, "loss": 0.0161, "step": 57370 }, { "grad_norm": 0.2341768741607666, "learning_rate": 4.196767699908045e-05, "loss": 0.0224, "step": 57380 }, { "grad_norm": 0.1654990166425705, "learning_rate": 4.195135749283845e-05, "loss": 0.0163, "step": 57390 }, { "grad_norm": 0.19387026131153107, "learning_rate": 4.1935038866783894e-05, "loss": 0.0148, "step": 57400 }, { "grad_norm": 0.21198363602161407, "learning_rate": 4.1918721122701355e-05, "loss": 0.0147, "step": 57410 }, { "grad_norm": 0.1886320859193802, "learning_rate": 4.190240426237533e-05, "loss": 0.0139, "step": 57420 }, { "grad_norm": 0.2659107744693756, "learning_rate": 4.188608828759019e-05, "loss": 0.0154, "step": 57430 }, { "grad_norm": 0.12339252978563309, "learning_rate": 4.1869773200130246e-05, "loss": 0.0177, "step": 57440 }, { "grad_norm": 0.23409567773342133, "learning_rate": 4.185345900177967e-05, "loss": 0.0181, "step": 57450 }, { "grad_norm": 0.15730828046798706, "learning_rate": 4.183714569432258e-05, "loss": 0.0138, "step": 57460 }, { "grad_norm": 0.23338276147842407, "learning_rate": 4.182083327954295e-05, "loss": 0.0139, "step": 57470 }, { "grad_norm": 0.312163770198822, "learning_rate": 4.1804521759224714e-05, "loss": 0.0164, "step": 57480 }, { "grad_norm": 0.14426782727241516, "learning_rate": 4.178821113515165e-05, "loss": 0.0147, "step": 57490 }, { "grad_norm": 0.14984869956970215, "learning_rate": 4.1771901409107476e-05, "loss": 0.0147, "step": 57500 }, { "grad_norm": 0.22871315479278564, "learning_rate": 4.175559258287577e-05, "loss": 0.0144, "step": 57510 }, { "grad_norm": 0.14698055386543274, "learning_rate": 4.173928465824008e-05, "loss": 0.017, "step": 57520 }, { "grad_norm": 0.18523655831813812, "learning_rate": 4.17229776369838e-05, "loss": 0.0156, "step": 57530 }, { "grad_norm": 0.24889156222343445, "learning_rate": 4.170667152089025e-05, "loss": 0.0158, "step": 57540 }, { "grad_norm": 0.15669357776641846, "learning_rate": 4.1690366311742626e-05, "loss": 0.0159, "step": 57550 }, { "grad_norm": 0.19822078943252563, "learning_rate": 4.1674062011324055e-05, "loss": 0.0165, "step": 57560 }, { "grad_norm": 0.3123476505279541, "learning_rate": 4.165775862141752e-05, "loss": 0.0168, "step": 57570 }, { "grad_norm": 0.20977546274662018, "learning_rate": 4.164145614380598e-05, "loss": 0.0136, "step": 57580 }, { "grad_norm": 0.15884459018707275, "learning_rate": 4.162515458027223e-05, "loss": 0.0141, "step": 57590 }, { "grad_norm": 0.22468458116054535, "learning_rate": 4.160885393259899e-05, "loss": 0.0137, "step": 57600 }, { "grad_norm": 0.21711993217468262, "learning_rate": 4.159255420256886e-05, "loss": 0.0171, "step": 57610 }, { "grad_norm": 0.2200091928243637, "learning_rate": 4.157625539196436e-05, "loss": 0.0164, "step": 57620 }, { "grad_norm": 0.18872012197971344, "learning_rate": 4.155995750256791e-05, "loss": 0.0144, "step": 57630 }, { "grad_norm": 0.15849463641643524, "learning_rate": 4.15436605361618e-05, "loss": 0.0173, "step": 57640 }, { "grad_norm": 0.3242653012275696, "learning_rate": 4.152736449452827e-05, "loss": 0.0163, "step": 57650 }, { "grad_norm": 0.15165166556835175, "learning_rate": 4.151106937944943e-05, "loss": 0.0167, "step": 57660 }, { "grad_norm": 0.1552266627550125, "learning_rate": 4.149477519270727e-05, "loss": 0.0135, "step": 57670 }, { "grad_norm": 0.17609186470508575, "learning_rate": 4.147848193608371e-05, "loss": 0.019, "step": 57680 }, { "grad_norm": 0.17370513081550598, "learning_rate": 4.146218961136056e-05, "loss": 0.0162, "step": 57690 }, { "grad_norm": 0.22326195240020752, "learning_rate": 4.14458982203195e-05, "loss": 0.0178, "step": 57700 }, { "grad_norm": 0.1784849315881729, "learning_rate": 4.1429607764742165e-05, "loss": 0.0164, "step": 57710 }, { "grad_norm": 0.1651400476694107, "learning_rate": 4.141331824641003e-05, "loss": 0.0158, "step": 57720 }, { "grad_norm": 0.23886919021606445, "learning_rate": 4.139702966710452e-05, "loss": 0.0177, "step": 57730 }, { "grad_norm": 0.15989123284816742, "learning_rate": 4.13807420286069e-05, "loss": 0.0136, "step": 57740 }, { "grad_norm": 0.21694685518741608, "learning_rate": 4.136445533269837e-05, "loss": 0.0186, "step": 57750 }, { "grad_norm": 0.15692485868930817, "learning_rate": 4.134816958116003e-05, "loss": 0.0147, "step": 57760 }, { "grad_norm": 0.17455296218395233, "learning_rate": 4.133188477577283e-05, "loss": 0.0127, "step": 57770 }, { "grad_norm": 0.17870883643627167, "learning_rate": 4.131560091831771e-05, "loss": 0.015, "step": 57780 }, { "grad_norm": 0.1393163502216339, "learning_rate": 4.1299318010575405e-05, "loss": 0.0158, "step": 57790 }, { "grad_norm": 0.1890982687473297, "learning_rate": 4.128303605432661e-05, "loss": 0.0165, "step": 57800 }, { "grad_norm": 0.1820962131023407, "learning_rate": 4.126675505135188e-05, "loss": 0.0157, "step": 57810 }, { "grad_norm": 0.19118764996528625, "learning_rate": 4.125047500343169e-05, "loss": 0.0155, "step": 57820 }, { "grad_norm": 0.11447852104902267, "learning_rate": 4.1234195912346376e-05, "loss": 0.0131, "step": 57830 }, { "grad_norm": 0.17510543763637543, "learning_rate": 4.1217917779876235e-05, "loss": 0.0147, "step": 57840 }, { "grad_norm": 0.2288040667772293, "learning_rate": 4.120164060780141e-05, "loss": 0.0176, "step": 57850 }, { "grad_norm": 0.20227216184139252, "learning_rate": 4.1185364397901935e-05, "loss": 0.0154, "step": 57860 }, { "grad_norm": 0.16727645695209503, "learning_rate": 4.116908915195776e-05, "loss": 0.0146, "step": 57870 }, { "grad_norm": 0.582584798336029, "learning_rate": 4.115281487174872e-05, "loss": 0.0157, "step": 57880 }, { "grad_norm": 0.1576758176088333, "learning_rate": 4.113654155905452e-05, "loss": 0.0157, "step": 57890 }, { "grad_norm": 0.25725018978118896, "learning_rate": 4.112026921565483e-05, "loss": 0.018, "step": 57900 }, { "grad_norm": 0.15751591324806213, "learning_rate": 4.1103997843329147e-05, "loss": 0.0145, "step": 57910 }, { "grad_norm": 0.2404097020626068, "learning_rate": 4.108772744385689e-05, "loss": 0.015, "step": 57920 }, { "grad_norm": 0.2419528216123581, "learning_rate": 4.1071458019017346e-05, "loss": 0.017, "step": 57930 }, { "grad_norm": 0.1715066134929657, "learning_rate": 4.1055189570589734e-05, "loss": 0.0173, "step": 57940 }, { "grad_norm": 0.18597739934921265, "learning_rate": 4.103892210035315e-05, "loss": 0.018, "step": 57950 }, { "grad_norm": 0.4356642961502075, "learning_rate": 4.102265561008656e-05, "loss": 0.0174, "step": 57960 }, { "grad_norm": 0.16120098531246185, "learning_rate": 4.100639010156885e-05, "loss": 0.0139, "step": 57970 }, { "grad_norm": 0.20617206394672394, "learning_rate": 4.099012557657882e-05, "loss": 0.0138, "step": 57980 }, { "grad_norm": 0.17201846837997437, "learning_rate": 4.09738620368951e-05, "loss": 0.0174, "step": 57990 }, { "grad_norm": 0.2614220976829529, "learning_rate": 4.095759948429626e-05, "loss": 0.0159, "step": 58000 }, { "grad_norm": 0.18173927068710327, "learning_rate": 4.094133792056075e-05, "loss": 0.016, "step": 58010 }, { "grad_norm": 0.19064170122146606, "learning_rate": 4.092507734746688e-05, "loss": 0.0145, "step": 58020 }, { "grad_norm": 0.20663118362426758, "learning_rate": 4.090881776679293e-05, "loss": 0.0168, "step": 58030 }, { "grad_norm": 0.2322959303855896, "learning_rate": 4.0892559180316994e-05, "loss": 0.015, "step": 58040 }, { "grad_norm": 0.20786792039871216, "learning_rate": 4.08763015898171e-05, "loss": 0.0137, "step": 58050 }, { "grad_norm": 0.17512686550617218, "learning_rate": 4.0860044997071145e-05, "loss": 0.0139, "step": 58060 }, { "grad_norm": 0.16633117198944092, "learning_rate": 4.084378940385693e-05, "loss": 0.0168, "step": 58070 }, { "grad_norm": 0.21948176622390747, "learning_rate": 4.082753481195211e-05, "loss": 0.0152, "step": 58080 }, { "grad_norm": 0.2470438927412033, "learning_rate": 4.0811281223134324e-05, "loss": 0.0165, "step": 58090 }, { "grad_norm": 0.23302224278450012, "learning_rate": 4.0795028639181e-05, "loss": 0.0188, "step": 58100 }, { "grad_norm": 0.21857817471027374, "learning_rate": 4.077877706186951e-05, "loss": 0.0177, "step": 58110 }, { "grad_norm": 0.1656983643770218, "learning_rate": 4.076252649297709e-05, "loss": 0.015, "step": 58120 }, { "grad_norm": 0.25747889280319214, "learning_rate": 4.07462769342809e-05, "loss": 0.0137, "step": 58130 }, { "grad_norm": 0.16951416432857513, "learning_rate": 4.073002838755793e-05, "loss": 0.0151, "step": 58140 }, { "grad_norm": 0.23070372641086578, "learning_rate": 4.0713780854585115e-05, "loss": 0.0162, "step": 58150 }, { "grad_norm": 0.15109148621559143, "learning_rate": 4.069753433713927e-05, "loss": 0.0155, "step": 58160 }, { "grad_norm": 0.2724620997905731, "learning_rate": 4.068128883699709e-05, "loss": 0.0156, "step": 58170 }, { "grad_norm": 0.2315356582403183, "learning_rate": 4.0665044355935155e-05, "loss": 0.0153, "step": 58180 }, { "grad_norm": 0.16306447982788086, "learning_rate": 4.064880089572993e-05, "loss": 0.0145, "step": 58190 }, { "grad_norm": 0.2525222599506378, "learning_rate": 4.063255845815779e-05, "loss": 0.016, "step": 58200 }, { "grad_norm": 0.1468953639268875, "learning_rate": 4.0616317044994945e-05, "loss": 0.0175, "step": 58210 }, { "grad_norm": 0.2045658975839615, "learning_rate": 4.0600076658017585e-05, "loss": 0.0147, "step": 58220 }, { "grad_norm": 0.18628983199596405, "learning_rate": 4.0583837299001706e-05, "loss": 0.0146, "step": 58230 }, { "grad_norm": 0.24866491556167603, "learning_rate": 4.056759896972322e-05, "loss": 0.0177, "step": 58240 }, { "grad_norm": 0.18214304745197296, "learning_rate": 4.055136167195793e-05, "loss": 0.0138, "step": 58250 }, { "grad_norm": 0.19107085466384888, "learning_rate": 4.053512540748152e-05, "loss": 0.0144, "step": 58260 }, { "grad_norm": 0.20302817225456238, "learning_rate": 4.051889017806955e-05, "loss": 0.0175, "step": 58270 }, { "grad_norm": 0.203121155500412, "learning_rate": 4.0502655985497516e-05, "loss": 0.0172, "step": 58280 }, { "grad_norm": 0.2207781970500946, "learning_rate": 4.048642283154074e-05, "loss": 0.0164, "step": 58290 }, { "grad_norm": 0.2331756353378296, "learning_rate": 4.0470190717974464e-05, "loss": 0.015, "step": 58300 }, { "grad_norm": 0.21619190275669098, "learning_rate": 4.045395964657381e-05, "loss": 0.0157, "step": 58310 }, { "grad_norm": 0.25348663330078125, "learning_rate": 4.043772961911376e-05, "loss": 0.0172, "step": 58320 }, { "grad_norm": 0.20231367647647858, "learning_rate": 4.042150063736924e-05, "loss": 0.0165, "step": 58330 }, { "grad_norm": 0.19282226264476776, "learning_rate": 4.0405272703114976e-05, "loss": 0.0194, "step": 58340 }, { "grad_norm": 0.24044521152973175, "learning_rate": 4.038904581812569e-05, "loss": 0.0148, "step": 58350 }, { "grad_norm": 0.1651594340801239, "learning_rate": 4.03728199841759e-05, "loss": 0.015, "step": 58360 }, { "grad_norm": 0.20159970223903656, "learning_rate": 4.0356595203040044e-05, "loss": 0.0145, "step": 58370 }, { "grad_norm": 0.24526239931583405, "learning_rate": 4.0340371476492426e-05, "loss": 0.0183, "step": 58380 }, { "grad_norm": 0.15314093232154846, "learning_rate": 4.032414880630727e-05, "loss": 0.0153, "step": 58390 }, { "grad_norm": 0.1986606866121292, "learning_rate": 4.0307927194258646e-05, "loss": 0.0149, "step": 58400 }, { "grad_norm": 0.22536346316337585, "learning_rate": 4.029170664212054e-05, "loss": 0.0163, "step": 58410 }, { "grad_norm": 0.13865633308887482, "learning_rate": 4.0275487151666794e-05, "loss": 0.0152, "step": 58420 }, { "grad_norm": 0.19765064120292664, "learning_rate": 4.025926872467116e-05, "loss": 0.0157, "step": 58430 }, { "grad_norm": 0.1744636744260788, "learning_rate": 4.024305136290725e-05, "loss": 0.0171, "step": 58440 }, { "grad_norm": 0.15716595947742462, "learning_rate": 4.022683506814858e-05, "loss": 0.0135, "step": 58450 }, { "grad_norm": 0.13828934729099274, "learning_rate": 4.021061984216853e-05, "loss": 0.0143, "step": 58460 }, { "grad_norm": 0.16570492088794708, "learning_rate": 4.019440568674035e-05, "loss": 0.0171, "step": 58470 }, { "grad_norm": 0.21651318669319153, "learning_rate": 4.017819260363724e-05, "loss": 0.0175, "step": 58480 }, { "grad_norm": 0.1650172770023346, "learning_rate": 4.016198059463223e-05, "loss": 0.0164, "step": 58490 }, { "grad_norm": 0.15896515548229218, "learning_rate": 4.0145769661498233e-05, "loss": 0.0155, "step": 58500 }, { "grad_norm": 0.20945370197296143, "learning_rate": 4.012955980600804e-05, "loss": 0.0155, "step": 58510 }, { "grad_norm": 0.14335153996944427, "learning_rate": 4.0113351029934355e-05, "loss": 0.0145, "step": 58520 }, { "grad_norm": 0.22463826835155487, "learning_rate": 4.0097143335049714e-05, "loss": 0.0154, "step": 58530 }, { "grad_norm": 0.17063313722610474, "learning_rate": 4.0080936723126606e-05, "loss": 0.0164, "step": 58540 }, { "grad_norm": 0.22056150436401367, "learning_rate": 4.006473119593735e-05, "loss": 0.0142, "step": 58550 }, { "grad_norm": 0.22259846329689026, "learning_rate": 4.004852675525415e-05, "loss": 0.0171, "step": 58560 }, { "grad_norm": 0.16551528871059418, "learning_rate": 4.00323234028491e-05, "loss": 0.0141, "step": 58570 }, { "grad_norm": 0.13644298911094666, "learning_rate": 4.0016121140494186e-05, "loss": 0.0139, "step": 58580 }, { "grad_norm": 0.1468796581029892, "learning_rate": 3.999991996996122e-05, "loss": 0.0162, "step": 58590 }, { "grad_norm": 0.18437941372394562, "learning_rate": 3.9983719893022e-05, "loss": 0.0152, "step": 58600 }, { "grad_norm": 0.1828266829252243, "learning_rate": 3.996752091144812e-05, "loss": 0.0165, "step": 58610 }, { "grad_norm": 0.2420729398727417, "learning_rate": 3.9951323027011055e-05, "loss": 0.0157, "step": 58620 }, { "grad_norm": 0.20844820141792297, "learning_rate": 3.993512624148219e-05, "loss": 0.0139, "step": 58630 }, { "grad_norm": 0.1838974803686142, "learning_rate": 3.99189305566328e-05, "loss": 0.0122, "step": 58640 }, { "grad_norm": 0.17095035314559937, "learning_rate": 3.990273597423401e-05, "loss": 0.015, "step": 58650 }, { "grad_norm": 0.29471662640571594, "learning_rate": 3.988654249605681e-05, "loss": 0.0136, "step": 58660 }, { "grad_norm": 0.1880331188440323, "learning_rate": 3.987035012387213e-05, "loss": 0.0149, "step": 58670 }, { "grad_norm": 0.18176418542861938, "learning_rate": 3.9854158859450735e-05, "loss": 0.0151, "step": 58680 }, { "grad_norm": 0.23147624731063843, "learning_rate": 3.983796870456328e-05, "loss": 0.0136, "step": 58690 }, { "grad_norm": 0.21856406331062317, "learning_rate": 3.982177966098028e-05, "loss": 0.0163, "step": 58700 }, { "grad_norm": 0.18418829143047333, "learning_rate": 3.980559173047216e-05, "loss": 0.0158, "step": 58710 }, { "grad_norm": 0.14908045530319214, "learning_rate": 3.9789404914809176e-05, "loss": 0.0139, "step": 58720 }, { "grad_norm": 0.17992599308490753, "learning_rate": 3.9773219215761544e-05, "loss": 0.0151, "step": 58730 }, { "grad_norm": 0.15434138476848602, "learning_rate": 3.975703463509928e-05, "loss": 0.0145, "step": 58740 }, { "grad_norm": 0.19987764954566956, "learning_rate": 3.974085117459231e-05, "loss": 0.0148, "step": 58750 }, { "grad_norm": 0.157212495803833, "learning_rate": 3.972466883601043e-05, "loss": 0.0146, "step": 58760 }, { "grad_norm": 0.11132406443357468, "learning_rate": 3.9708487621123316e-05, "loss": 0.0148, "step": 58770 }, { "grad_norm": 0.18526481091976166, "learning_rate": 3.969230753170051e-05, "loss": 0.0152, "step": 58780 }, { "grad_norm": 0.1625327467918396, "learning_rate": 3.967612856951146e-05, "loss": 0.016, "step": 58790 }, { "grad_norm": 0.18739521503448486, "learning_rate": 3.9659950736325476e-05, "loss": 0.014, "step": 58800 }, { "grad_norm": 0.1322970688343048, "learning_rate": 3.964377403391174e-05, "loss": 0.0132, "step": 58810 }, { "grad_norm": 0.1487988531589508, "learning_rate": 3.9627598464039303e-05, "loss": 0.015, "step": 58820 }, { "grad_norm": 0.1685970276594162, "learning_rate": 3.9611424028477096e-05, "loss": 0.0159, "step": 58830 }, { "grad_norm": 0.18465417623519897, "learning_rate": 3.959525072899394e-05, "loss": 0.0156, "step": 58840 }, { "grad_norm": 0.2274140864610672, "learning_rate": 3.9579078567358506e-05, "loss": 0.0162, "step": 58850 }, { "grad_norm": 0.15266390144824982, "learning_rate": 3.956290754533939e-05, "loss": 0.0124, "step": 58860 }, { "grad_norm": 0.14971670508384705, "learning_rate": 3.954673766470501e-05, "loss": 0.0147, "step": 58870 }, { "grad_norm": 0.17060257494449615, "learning_rate": 3.9530568927223674e-05, "loss": 0.0172, "step": 58880 }, { "grad_norm": 0.13687476515769958, "learning_rate": 3.95144013346636e-05, "loss": 0.0169, "step": 58890 }, { "grad_norm": 0.15356920659542084, "learning_rate": 3.949823488879282e-05, "loss": 0.0151, "step": 58900 }, { "grad_norm": 0.1635468751192093, "learning_rate": 3.9482069591379276e-05, "loss": 0.0137, "step": 58910 }, { "grad_norm": 0.1588008552789688, "learning_rate": 3.9465905444190784e-05, "loss": 0.0145, "step": 58920 }, { "grad_norm": 0.21016404032707214, "learning_rate": 3.944974244899505e-05, "loss": 0.0151, "step": 58930 }, { "grad_norm": 0.19696921110153198, "learning_rate": 3.9433580607559616e-05, "loss": 0.0166, "step": 58940 }, { "grad_norm": 0.1499110907316208, "learning_rate": 3.941741992165193e-05, "loss": 0.0195, "step": 58950 }, { "grad_norm": 0.10717349499464035, "learning_rate": 3.940126039303928e-05, "loss": 0.0154, "step": 58960 }, { "grad_norm": 0.1488315612077713, "learning_rate": 3.938510202348886e-05, "loss": 0.0141, "step": 58970 }, { "grad_norm": 0.15144889056682587, "learning_rate": 3.93689448147677e-05, "loss": 0.0158, "step": 58980 }, { "grad_norm": 0.1336035281419754, "learning_rate": 3.935278876864278e-05, "loss": 0.0125, "step": 58990 }, { "grad_norm": 0.18777857720851898, "learning_rate": 3.933663388688087e-05, "loss": 0.0138, "step": 59000 }, { "grad_norm": 0.14532774686813354, "learning_rate": 3.9320480171248645e-05, "loss": 0.0133, "step": 59010 }, { "grad_norm": 0.13002249598503113, "learning_rate": 3.930432762351265e-05, "loss": 0.0147, "step": 59020 }, { "grad_norm": 0.17903223633766174, "learning_rate": 3.928817624543931e-05, "loss": 0.0172, "step": 59030 }, { "grad_norm": 0.17736779153347015, "learning_rate": 3.9272026038794885e-05, "loss": 0.0128, "step": 59040 }, { "grad_norm": 0.1839931756258011, "learning_rate": 3.9255877005345586e-05, "loss": 0.0168, "step": 59050 }, { "grad_norm": 0.22155769169330597, "learning_rate": 3.923972914685743e-05, "loss": 0.0191, "step": 59060 }, { "grad_norm": 0.2498810887336731, "learning_rate": 3.922358246509631e-05, "loss": 0.0155, "step": 59070 }, { "grad_norm": 0.29074224829673767, "learning_rate": 3.920743696182801e-05, "loss": 0.0181, "step": 59080 }, { "grad_norm": 0.18747439980506897, "learning_rate": 3.9191292638818164e-05, "loss": 0.0146, "step": 59090 }, { "grad_norm": 0.22570791840553284, "learning_rate": 3.9175149497832316e-05, "loss": 0.0169, "step": 59100 }, { "grad_norm": 0.1907850056886673, "learning_rate": 3.915900754063584e-05, "loss": 0.0162, "step": 59110 }, { "grad_norm": 0.19282498955726624, "learning_rate": 3.9142866768994014e-05, "loss": 0.0132, "step": 59120 }, { "grad_norm": 0.17907464504241943, "learning_rate": 3.912672718467193e-05, "loss": 0.0182, "step": 59130 }, { "grad_norm": 0.2088755965232849, "learning_rate": 3.911058878943463e-05, "loss": 0.0149, "step": 59140 }, { "grad_norm": 0.17401662468910217, "learning_rate": 3.9094451585046974e-05, "loss": 0.015, "step": 59150 }, { "grad_norm": 0.2702566087245941, "learning_rate": 3.90783155732737e-05, "loss": 0.0175, "step": 59160 }, { "grad_norm": 0.14548884332180023, "learning_rate": 3.9062180755879373e-05, "loss": 0.0175, "step": 59170 }, { "grad_norm": 0.1988179236650467, "learning_rate": 3.904604713462855e-05, "loss": 0.0137, "step": 59180 }, { "grad_norm": 0.19034962356090546, "learning_rate": 3.902991471128554e-05, "loss": 0.0158, "step": 59190 }, { "grad_norm": 0.17738507688045502, "learning_rate": 3.901378348761456e-05, "loss": 0.0156, "step": 59200 }, { "grad_norm": 0.19091053307056427, "learning_rate": 3.899765346537971e-05, "loss": 0.0136, "step": 59210 }, { "grad_norm": 0.1925533562898636, "learning_rate": 3.898152464634493e-05, "loss": 0.0161, "step": 59220 }, { "grad_norm": 0.19339898228645325, "learning_rate": 3.896539703227402e-05, "loss": 0.0138, "step": 59230 }, { "grad_norm": 0.15659427642822266, "learning_rate": 3.894927062493073e-05, "loss": 0.0137, "step": 59240 }, { "grad_norm": 0.20793889462947845, "learning_rate": 3.893314542607859e-05, "loss": 0.0166, "step": 59250 }, { "grad_norm": 0.1772814244031906, "learning_rate": 3.891702143748103e-05, "loss": 0.0145, "step": 59260 }, { "grad_norm": 0.15572227537631989, "learning_rate": 3.890089866090133e-05, "loss": 0.0148, "step": 59270 }, { "grad_norm": 0.12219593673944473, "learning_rate": 3.888477709810267e-05, "loss": 0.0192, "step": 59280 }, { "grad_norm": 0.17120297253131866, "learning_rate": 3.886865675084806e-05, "loss": 0.0149, "step": 59290 }, { "grad_norm": 0.15601973235607147, "learning_rate": 3.885253762090043e-05, "loss": 0.0158, "step": 59300 }, { "grad_norm": 0.1796521693468094, "learning_rate": 3.8836419710022534e-05, "loss": 0.0157, "step": 59310 }, { "grad_norm": 0.1851840764284134, "learning_rate": 3.882030301997698e-05, "loss": 0.0177, "step": 59320 }, { "grad_norm": 0.14846271276474, "learning_rate": 3.880418755252629e-05, "loss": 0.0145, "step": 59330 }, { "grad_norm": 0.17620977759361267, "learning_rate": 3.87880733094328e-05, "loss": 0.0163, "step": 59340 }, { "grad_norm": 0.2273288518190384, "learning_rate": 3.877196029245877e-05, "loss": 0.0204, "step": 59350 }, { "grad_norm": 0.14051613211631775, "learning_rate": 3.875584850336627e-05, "loss": 0.0158, "step": 59360 }, { "grad_norm": 0.14647270739078522, "learning_rate": 3.873973794391728e-05, "loss": 0.0182, "step": 59370 }, { "grad_norm": 0.1304096281528473, "learning_rate": 3.872362861587361e-05, "loss": 0.0182, "step": 59380 }, { "grad_norm": 0.17949582636356354, "learning_rate": 3.870752052099698e-05, "loss": 0.0185, "step": 59390 }, { "grad_norm": 0.20321260392665863, "learning_rate": 3.8691413661048916e-05, "loss": 0.0169, "step": 59400 }, { "grad_norm": 0.2565016448497772, "learning_rate": 3.867530803779085e-05, "loss": 0.0148, "step": 59410 }, { "grad_norm": 0.17693577706813812, "learning_rate": 3.8659203652984054e-05, "loss": 0.0131, "step": 59420 }, { "grad_norm": 0.17561084032058716, "learning_rate": 3.864310050838972e-05, "loss": 0.0157, "step": 59430 }, { "grad_norm": 0.22768980264663696, "learning_rate": 3.8626998605768835e-05, "loss": 0.0147, "step": 59440 }, { "grad_norm": 0.19843454658985138, "learning_rate": 3.8610897946882294e-05, "loss": 0.0174, "step": 59450 }, { "grad_norm": 0.19027048349380493, "learning_rate": 3.859479853349082e-05, "loss": 0.0155, "step": 59460 }, { "grad_norm": 0.15079954266548157, "learning_rate": 3.857870036735503e-05, "loss": 0.0139, "step": 59470 }, { "grad_norm": 0.19039182364940643, "learning_rate": 3.856260345023538e-05, "loss": 0.0158, "step": 59480 }, { "grad_norm": 0.16973567008972168, "learning_rate": 3.854650778389224e-05, "loss": 0.0176, "step": 59490 }, { "grad_norm": 0.13933053612709045, "learning_rate": 3.8530413370085786e-05, "loss": 0.0142, "step": 59500 }, { "grad_norm": 0.20828551054000854, "learning_rate": 3.851432021057608e-05, "loss": 0.017, "step": 59510 }, { "grad_norm": 0.2829453945159912, "learning_rate": 3.849822830712306e-05, "loss": 0.0149, "step": 59520 }, { "grad_norm": 0.17399749159812927, "learning_rate": 3.848213766148649e-05, "loss": 0.014, "step": 59530 }, { "grad_norm": 0.1465304046869278, "learning_rate": 3.846604827542602e-05, "loss": 0.0143, "step": 59540 }, { "grad_norm": 0.33232948184013367, "learning_rate": 3.844996015070117e-05, "loss": 0.0173, "step": 59550 }, { "grad_norm": 0.17988735437393188, "learning_rate": 3.843387328907132e-05, "loss": 0.0192, "step": 59560 }, { "grad_norm": 0.17340300977230072, "learning_rate": 3.8417787692295696e-05, "loss": 0.0147, "step": 59570 }, { "grad_norm": 0.16645073890686035, "learning_rate": 3.84017033621334e-05, "loss": 0.0163, "step": 59580 }, { "grad_norm": 0.18006300926208496, "learning_rate": 3.838562030034338e-05, "loss": 0.0158, "step": 59590 }, { "grad_norm": 0.20068304240703583, "learning_rate": 3.836953850868447e-05, "loss": 0.0161, "step": 59600 }, { "grad_norm": 0.19001959264278412, "learning_rate": 3.835345798891533e-05, "loss": 0.0166, "step": 59610 }, { "grad_norm": 0.11750786006450653, "learning_rate": 3.833737874279452e-05, "loss": 0.0137, "step": 59620 }, { "grad_norm": 0.13095976412296295, "learning_rate": 3.8321300772080433e-05, "loss": 0.0148, "step": 59630 }, { "grad_norm": 0.19350506365299225, "learning_rate": 3.8305224078531345e-05, "loss": 0.0151, "step": 59640 }, { "grad_norm": 0.19948945939540863, "learning_rate": 3.828914866390537e-05, "loss": 0.0155, "step": 59650 }, { "grad_norm": 0.24345532059669495, "learning_rate": 3.827307452996048e-05, "loss": 0.0166, "step": 59660 }, { "grad_norm": 0.19183126091957092, "learning_rate": 3.825700167845454e-05, "loss": 0.0177, "step": 59670 }, { "grad_norm": 0.1429336965084076, "learning_rate": 3.8240930111145214e-05, "loss": 0.0138, "step": 59680 }, { "grad_norm": 0.1631501019001007, "learning_rate": 3.8224859829790116e-05, "loss": 0.0161, "step": 59690 }, { "grad_norm": 0.14793327450752258, "learning_rate": 3.820879083614664e-05, "loss": 0.0161, "step": 59700 }, { "grad_norm": 0.1734660118818283, "learning_rate": 3.819272313197208e-05, "loss": 0.0178, "step": 59710 }, { "grad_norm": 0.1645030975341797, "learning_rate": 3.817665671902355e-05, "loss": 0.0137, "step": 59720 }, { "grad_norm": 0.16577136516571045, "learning_rate": 3.816059159905808e-05, "loss": 0.0124, "step": 59730 }, { "grad_norm": 0.16524755954742432, "learning_rate": 3.8144527773832476e-05, "loss": 0.0145, "step": 59740 }, { "grad_norm": 0.17303724586963654, "learning_rate": 3.812846524510352e-05, "loss": 0.0128, "step": 59750 }, { "grad_norm": 0.1364641934633255, "learning_rate": 3.811240401462775e-05, "loss": 0.0149, "step": 59760 }, { "grad_norm": 0.16594786942005157, "learning_rate": 3.809634408416162e-05, "loss": 0.0174, "step": 59770 }, { "grad_norm": 0.19224384427070618, "learning_rate": 3.8080285455461394e-05, "loss": 0.0162, "step": 59780 }, { "grad_norm": 0.1833655834197998, "learning_rate": 3.806422813028323e-05, "loss": 0.0163, "step": 59790 }, { "grad_norm": 0.16770686209201813, "learning_rate": 3.8048172110383114e-05, "loss": 0.0139, "step": 59800 }, { "grad_norm": 0.17483149468898773, "learning_rate": 3.8032117397516944e-05, "loss": 0.0179, "step": 59810 }, { "grad_norm": 0.19188639521598816, "learning_rate": 3.801606399344042e-05, "loss": 0.0145, "step": 59820 }, { "grad_norm": 0.18685732781887054, "learning_rate": 3.800001189990913e-05, "loss": 0.0137, "step": 59830 }, { "grad_norm": 0.14410142600536346, "learning_rate": 3.798396111867847e-05, "loss": 0.0129, "step": 59840 }, { "grad_norm": 0.10751029849052429, "learning_rate": 3.796791165150378e-05, "loss": 0.0119, "step": 59850 }, { "grad_norm": 0.16450180113315582, "learning_rate": 3.795186350014018e-05, "loss": 0.0145, "step": 59860 }, { "grad_norm": 0.14821434020996094, "learning_rate": 3.793581666634266e-05, "loss": 0.0162, "step": 59870 }, { "grad_norm": 0.12584048509597778, "learning_rate": 3.791977115186609e-05, "loss": 0.0134, "step": 59880 }, { "grad_norm": 0.14160002768039703, "learning_rate": 3.79037269584652e-05, "loss": 0.0183, "step": 59890 }, { "grad_norm": 0.15204237401485443, "learning_rate": 3.7887684087894545e-05, "loss": 0.0142, "step": 59900 }, { "grad_norm": 0.20413275063037872, "learning_rate": 3.787164254190854e-05, "loss": 0.0156, "step": 59910 }, { "grad_norm": 0.14248248934745789, "learning_rate": 3.7855602322261473e-05, "loss": 0.0144, "step": 59920 }, { "grad_norm": 0.19471345841884613, "learning_rate": 3.783956343070746e-05, "loss": 0.0158, "step": 59930 }, { "grad_norm": 0.16779176890850067, "learning_rate": 3.7823525869000526e-05, "loss": 0.0136, "step": 59940 }, { "grad_norm": 0.18322671949863434, "learning_rate": 3.78074896388945e-05, "loss": 0.0143, "step": 59950 }, { "grad_norm": 0.26641929149627686, "learning_rate": 3.7791454742143075e-05, "loss": 0.0152, "step": 59960 }, { "grad_norm": 0.13605213165283203, "learning_rate": 3.7775421180499795e-05, "loss": 0.0168, "step": 59970 }, { "grad_norm": 0.18220342695713043, "learning_rate": 3.7759388955718086e-05, "loss": 0.0153, "step": 59980 }, { "grad_norm": 0.20189088582992554, "learning_rate": 3.774335806955117e-05, "loss": 0.014, "step": 59990 }, { "grad_norm": 0.2100561112165451, "learning_rate": 3.7727328523752215e-05, "loss": 0.0143, "step": 60000 } ], "logging_steps": 10, "max_steps": 100000, "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": false }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 28, "trial_name": null, "trial_params": null }