{ "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": 0.926266074180603, "learning_rate": 1.8e-07, "loss": 1.4438, "step": 10 }, { "grad_norm": 1.0640310049057007, "learning_rate": 3.8e-07, "loss": 1.4395, "step": 20 }, { "grad_norm": 0.9207378625869751, "learning_rate": 5.8e-07, "loss": 1.4342, "step": 30 }, { "grad_norm": 0.8789049386978149, "learning_rate": 7.8e-07, "loss": 1.4412, "step": 40 }, { "grad_norm": 0.7615863680839539, "learning_rate": 9.8e-07, "loss": 1.4241, "step": 50 }, { "grad_norm": 1.0124503374099731, "learning_rate": 1.18e-06, "loss": 1.376, "step": 60 }, { "grad_norm": 0.8852088451385498, "learning_rate": 1.3800000000000001e-06, "loss": 1.3341, "step": 70 }, { "grad_norm": 1.1166917085647583, "learning_rate": 1.5800000000000003e-06, "loss": 1.287, "step": 80 }, { "grad_norm": 0.8359726071357727, "learning_rate": 1.7800000000000001e-06, "loss": 1.2729, "step": 90 }, { "grad_norm": 0.7087129354476929, "learning_rate": 1.98e-06, "loss": 1.2444, "step": 100 }, { "grad_norm": 0.8514461517333984, "learning_rate": 2.1800000000000003e-06, "loss": 1.2798, "step": 110 }, { "grad_norm": 1.2569960355758667, "learning_rate": 2.38e-06, "loss": 1.2389, "step": 120 }, { "grad_norm": 1.465353012084961, "learning_rate": 2.5800000000000003e-06, "loss": 1.2177, "step": 130 }, { "grad_norm": 2.254542589187622, "learning_rate": 2.78e-06, "loss": 1.2076, "step": 140 }, { "grad_norm": 1.0675272941589355, "learning_rate": 2.9800000000000003e-06, "loss": 1.1924, "step": 150 }, { "grad_norm": 1.3136998414993286, "learning_rate": 3.1800000000000005e-06, "loss": 1.1717, "step": 160 }, { "grad_norm": 1.7947893142700195, "learning_rate": 3.38e-06, "loss": 1.1604, "step": 170 }, { "grad_norm": 1.1186102628707886, "learning_rate": 3.58e-06, "loss": 1.1338, "step": 180 }, { "grad_norm": 1.1510117053985596, "learning_rate": 3.7800000000000002e-06, "loss": 1.1316, "step": 190 }, { "grad_norm": 1.0984177589416504, "learning_rate": 3.98e-06, "loss": 1.1084, "step": 200 }, { "grad_norm": 1.1552013158798218, "learning_rate": 4.18e-06, "loss": 1.1082, "step": 210 }, { "grad_norm": 1.0869241952896118, "learning_rate": 4.38e-06, "loss": 1.1082, "step": 220 }, { "grad_norm": 1.1602262258529663, "learning_rate": 4.58e-06, "loss": 1.111, "step": 230 }, { "grad_norm": 1.7610721588134766, "learning_rate": 4.780000000000001e-06, "loss": 1.0903, "step": 240 }, { "grad_norm": 1.2524158954620361, "learning_rate": 4.98e-06, "loss": 1.0972, "step": 250 }, { "grad_norm": 1.145118236541748, "learning_rate": 5.18e-06, "loss": 1.0818, "step": 260 }, { "grad_norm": 1.0128108263015747, "learning_rate": 5.38e-06, "loss": 1.0724, "step": 270 }, { "grad_norm": 1.351258635520935, "learning_rate": 5.580000000000001e-06, "loss": 1.0873, "step": 280 }, { "grad_norm": 1.1126656532287598, "learning_rate": 5.78e-06, "loss": 1.0808, "step": 290 }, { "grad_norm": 1.5978598594665527, "learning_rate": 5.98e-06, "loss": 1.0694, "step": 300 }, { "grad_norm": 1.2440999746322632, "learning_rate": 6.18e-06, "loss": 1.0426, "step": 310 }, { "grad_norm": 1.4261150360107422, "learning_rate": 6.38e-06, "loss": 1.0258, "step": 320 }, { "grad_norm": 1.6032270193099976, "learning_rate": 6.58e-06, "loss": 1.0008, "step": 330 }, { "grad_norm": 1.5931432247161865, "learning_rate": 6.78e-06, "loss": 0.9674, "step": 340 }, { "grad_norm": 1.5402252674102783, "learning_rate": 6.98e-06, "loss": 0.962, "step": 350 }, { "grad_norm": 1.5855474472045898, "learning_rate": 7.180000000000001e-06, "loss": 0.9422, "step": 360 }, { "grad_norm": 1.9631766080856323, "learning_rate": 7.3800000000000005e-06, "loss": 0.9142, "step": 370 }, { "grad_norm": 1.8945468664169312, "learning_rate": 7.580000000000001e-06, "loss": 0.8874, "step": 380 }, { "grad_norm": 1.8411067724227905, "learning_rate": 7.78e-06, "loss": 0.8613, "step": 390 }, { "grad_norm": 1.7587705850601196, "learning_rate": 7.98e-06, "loss": 0.8239, "step": 400 }, { "grad_norm": 1.9953256845474243, "learning_rate": 8.18e-06, "loss": 0.7818, "step": 410 }, { "grad_norm": 1.5480085611343384, "learning_rate": 8.380000000000001e-06, "loss": 0.744, "step": 420 }, { "grad_norm": 2.051807165145874, "learning_rate": 8.580000000000001e-06, "loss": 0.7039, "step": 430 }, { "grad_norm": 1.8827357292175293, "learning_rate": 8.78e-06, "loss": 0.6799, "step": 440 }, { "grad_norm": 2.2053680419921875, "learning_rate": 8.98e-06, "loss": 0.6308, "step": 450 }, { "grad_norm": 2.365339994430542, "learning_rate": 9.180000000000002e-06, "loss": 0.601, "step": 460 }, { "grad_norm": 2.4449708461761475, "learning_rate": 9.38e-06, "loss": 0.5866, "step": 470 }, { "grad_norm": 2.2257940769195557, "learning_rate": 9.58e-06, "loss": 0.544, "step": 480 }, { "grad_norm": 2.1845319271087646, "learning_rate": 9.78e-06, "loss": 0.5304, "step": 490 }, { "grad_norm": 1.8719619512557983, "learning_rate": 9.980000000000001e-06, "loss": 0.4801, "step": 500 }, { "grad_norm": 2.23685884475708, "learning_rate": 1.018e-05, "loss": 0.4509, "step": 510 }, { "grad_norm": 2.2305705547332764, "learning_rate": 1.038e-05, "loss": 0.4353, "step": 520 }, { "grad_norm": 2.8304765224456787, "learning_rate": 1.058e-05, "loss": 0.4199, "step": 530 }, { "grad_norm": 2.6049983501434326, "learning_rate": 1.0780000000000002e-05, "loss": 0.3829, "step": 540 }, { "grad_norm": 2.4628560543060303, "learning_rate": 1.098e-05, "loss": 0.3924, "step": 550 }, { "grad_norm": 3.0925612449645996, "learning_rate": 1.118e-05, "loss": 0.3626, "step": 560 }, { "grad_norm": 2.828927516937256, "learning_rate": 1.1380000000000001e-05, "loss": 0.3571, "step": 570 }, { "grad_norm": 2.2817559242248535, "learning_rate": 1.1580000000000001e-05, "loss": 0.3344, "step": 580 }, { "grad_norm": 2.1635260581970215, "learning_rate": 1.178e-05, "loss": 0.3044, "step": 590 }, { "grad_norm": 2.560818672180176, "learning_rate": 1.198e-05, "loss": 0.3004, "step": 600 }, { "grad_norm": 3.1146340370178223, "learning_rate": 1.2180000000000002e-05, "loss": 0.2827, "step": 610 }, { "grad_norm": 2.7388076782226562, "learning_rate": 1.238e-05, "loss": 0.302, "step": 620 }, { "grad_norm": 2.4787237644195557, "learning_rate": 1.258e-05, "loss": 0.2454, "step": 630 }, { "grad_norm": 2.952630043029785, "learning_rate": 1.278e-05, "loss": 0.2523, "step": 640 }, { "grad_norm": 2.6432156562805176, "learning_rate": 1.2980000000000001e-05, "loss": 0.2293, "step": 650 }, { "grad_norm": 2.473900318145752, "learning_rate": 1.3180000000000001e-05, "loss": 0.2358, "step": 660 }, { "grad_norm": 2.6395559310913086, "learning_rate": 1.338e-05, "loss": 0.2321, "step": 670 }, { "grad_norm": 2.060318946838379, "learning_rate": 1.358e-05, "loss": 0.2122, "step": 680 }, { "grad_norm": 2.3797101974487305, "learning_rate": 1.3780000000000002e-05, "loss": 0.2131, "step": 690 }, { "grad_norm": 2.041769027709961, "learning_rate": 1.3980000000000002e-05, "loss": 0.2049, "step": 700 }, { "grad_norm": 2.726841688156128, "learning_rate": 1.4180000000000001e-05, "loss": 0.2173, "step": 710 }, { "grad_norm": 1.9584563970565796, "learning_rate": 1.4380000000000001e-05, "loss": 0.197, "step": 720 }, { "grad_norm": 2.1229758262634277, "learning_rate": 1.4580000000000003e-05, "loss": 0.2028, "step": 730 }, { "grad_norm": 2.2316715717315674, "learning_rate": 1.4779999999999999e-05, "loss": 0.1857, "step": 740 }, { "grad_norm": 3.030278205871582, "learning_rate": 1.4979999999999999e-05, "loss": 0.2036, "step": 750 }, { "grad_norm": 2.147660732269287, "learning_rate": 1.518e-05, "loss": 0.1753, "step": 760 }, { "grad_norm": 2.1291427612304688, "learning_rate": 1.538e-05, "loss": 0.1953, "step": 770 }, { "grad_norm": 2.3734099864959717, "learning_rate": 1.558e-05, "loss": 0.1811, "step": 780 }, { "grad_norm": 2.0445516109466553, "learning_rate": 1.578e-05, "loss": 0.1846, "step": 790 }, { "grad_norm": 2.276503801345825, "learning_rate": 1.598e-05, "loss": 0.1781, "step": 800 }, { "grad_norm": 2.500742197036743, "learning_rate": 1.618e-05, "loss": 0.1903, "step": 810 }, { "grad_norm": 1.7059069871902466, "learning_rate": 1.6380000000000002e-05, "loss": 0.1798, "step": 820 }, { "grad_norm": 2.3454294204711914, "learning_rate": 1.658e-05, "loss": 0.1775, "step": 830 }, { "grad_norm": 2.3162667751312256, "learning_rate": 1.6780000000000002e-05, "loss": 0.1608, "step": 840 }, { "grad_norm": 1.9290380477905273, "learning_rate": 1.698e-05, "loss": 0.1591, "step": 850 }, { "grad_norm": 1.9581745862960815, "learning_rate": 1.718e-05, "loss": 0.1661, "step": 860 }, { "grad_norm": 2.471449375152588, "learning_rate": 1.7380000000000003e-05, "loss": 0.1685, "step": 870 }, { "grad_norm": 2.3924989700317383, "learning_rate": 1.758e-05, "loss": 0.1626, "step": 880 }, { "grad_norm": 2.1243438720703125, "learning_rate": 1.7780000000000003e-05, "loss": 0.1702, "step": 890 }, { "grad_norm": 2.313042402267456, "learning_rate": 1.798e-05, "loss": 0.1661, "step": 900 }, { "grad_norm": 2.1570398807525635, "learning_rate": 1.818e-05, "loss": 0.1663, "step": 910 }, { "grad_norm": 2.036301612854004, "learning_rate": 1.838e-05, "loss": 0.1446, "step": 920 }, { "grad_norm": 1.9931927919387817, "learning_rate": 1.858e-05, "loss": 0.1615, "step": 930 }, { "grad_norm": 1.81403648853302, "learning_rate": 1.878e-05, "loss": 0.1689, "step": 940 }, { "grad_norm": 2.0347235202789307, "learning_rate": 1.898e-05, "loss": 0.156, "step": 950 }, { "grad_norm": 2.040017604827881, "learning_rate": 1.918e-05, "loss": 0.1376, "step": 960 }, { "grad_norm": 1.9952945709228516, "learning_rate": 1.938e-05, "loss": 0.153, "step": 970 }, { "grad_norm": 2.2728240489959717, "learning_rate": 1.9580000000000002e-05, "loss": 0.1355, "step": 980 }, { "grad_norm": 2.187716245651245, "learning_rate": 1.978e-05, "loss": 0.1315, "step": 990 }, { "grad_norm": 1.5979294776916504, "learning_rate": 1.9980000000000002e-05, "loss": 0.1487, "step": 1000 }, { "grad_norm": 2.006293773651123, "learning_rate": 2.0180000000000003e-05, "loss": 0.127, "step": 1010 }, { "grad_norm": 1.860627293586731, "learning_rate": 2.038e-05, "loss": 0.1404, "step": 1020 }, { "grad_norm": 2.1587531566619873, "learning_rate": 2.0580000000000003e-05, "loss": 0.1466, "step": 1030 }, { "grad_norm": 1.531379222869873, "learning_rate": 2.078e-05, "loss": 0.1331, "step": 1040 }, { "grad_norm": 2.1895930767059326, "learning_rate": 2.098e-05, "loss": 0.1232, "step": 1050 }, { "grad_norm": 1.7790223360061646, "learning_rate": 2.118e-05, "loss": 0.1397, "step": 1060 }, { "grad_norm": 1.8040825128555298, "learning_rate": 2.138e-05, "loss": 0.1427, "step": 1070 }, { "grad_norm": 1.5641379356384277, "learning_rate": 2.158e-05, "loss": 0.1426, "step": 1080 }, { "grad_norm": 2.0407676696777344, "learning_rate": 2.178e-05, "loss": 0.1416, "step": 1090 }, { "grad_norm": 1.5827224254608154, "learning_rate": 2.198e-05, "loss": 0.1253, "step": 1100 }, { "grad_norm": 2.0948946475982666, "learning_rate": 2.218e-05, "loss": 0.1442, "step": 1110 }, { "grad_norm": 1.7712286710739136, "learning_rate": 2.2380000000000003e-05, "loss": 0.1337, "step": 1120 }, { "grad_norm": 1.352770447731018, "learning_rate": 2.258e-05, "loss": 0.1467, "step": 1130 }, { "grad_norm": 1.789804220199585, "learning_rate": 2.2780000000000002e-05, "loss": 0.1326, "step": 1140 }, { "grad_norm": 1.6404792070388794, "learning_rate": 2.298e-05, "loss": 0.1334, "step": 1150 }, { "grad_norm": 1.9014173746109009, "learning_rate": 2.318e-05, "loss": 0.1269, "step": 1160 }, { "grad_norm": 1.6180914640426636, "learning_rate": 2.3380000000000003e-05, "loss": 0.1231, "step": 1170 }, { "grad_norm": 1.6588298082351685, "learning_rate": 2.358e-05, "loss": 0.1258, "step": 1180 }, { "grad_norm": 1.6564593315124512, "learning_rate": 2.3780000000000003e-05, "loss": 0.122, "step": 1190 }, { "grad_norm": 1.6317307949066162, "learning_rate": 2.398e-05, "loss": 0.1341, "step": 1200 }, { "grad_norm": 1.6443976163864136, "learning_rate": 2.418e-05, "loss": 0.1138, "step": 1210 }, { "grad_norm": 1.5773783922195435, "learning_rate": 2.438e-05, "loss": 0.1214, "step": 1220 }, { "grad_norm": 1.4465614557266235, "learning_rate": 2.4580000000000002e-05, "loss": 0.1214, "step": 1230 }, { "grad_norm": 1.6430245637893677, "learning_rate": 2.478e-05, "loss": 0.115, "step": 1240 }, { "grad_norm": 1.7517200708389282, "learning_rate": 2.498e-05, "loss": 0.1172, "step": 1250 }, { "grad_norm": 1.3756705522537231, "learning_rate": 2.5180000000000003e-05, "loss": 0.123, "step": 1260 }, { "grad_norm": 1.6590244770050049, "learning_rate": 2.5380000000000004e-05, "loss": 0.1198, "step": 1270 }, { "grad_norm": 1.6160808801651, "learning_rate": 2.5580000000000002e-05, "loss": 0.1115, "step": 1280 }, { "grad_norm": 1.2673935890197754, "learning_rate": 2.5779999999999997e-05, "loss": 0.1154, "step": 1290 }, { "grad_norm": 1.4574971199035645, "learning_rate": 2.598e-05, "loss": 0.1136, "step": 1300 }, { "grad_norm": 1.2327100038528442, "learning_rate": 2.618e-05, "loss": 0.1172, "step": 1310 }, { "grad_norm": 1.3506114482879639, "learning_rate": 2.6379999999999998e-05, "loss": 0.1287, "step": 1320 }, { "grad_norm": 1.7393161058425903, "learning_rate": 2.658e-05, "loss": 0.1195, "step": 1330 }, { "grad_norm": 1.0298277139663696, "learning_rate": 2.678e-05, "loss": 0.1124, "step": 1340 }, { "grad_norm": 1.3873794078826904, "learning_rate": 2.698e-05, "loss": 0.1088, "step": 1350 }, { "grad_norm": 1.1892354488372803, "learning_rate": 2.718e-05, "loss": 0.1098, "step": 1360 }, { "grad_norm": 1.558323621749878, "learning_rate": 2.738e-05, "loss": 0.1094, "step": 1370 }, { "grad_norm": 1.1874359846115112, "learning_rate": 2.758e-05, "loss": 0.1178, "step": 1380 }, { "grad_norm": 1.5623217821121216, "learning_rate": 2.778e-05, "loss": 0.1179, "step": 1390 }, { "grad_norm": 1.7260891199111938, "learning_rate": 2.798e-05, "loss": 0.1052, "step": 1400 }, { "grad_norm": 1.6796718835830688, "learning_rate": 2.818e-05, "loss": 0.1219, "step": 1410 }, { "grad_norm": 1.7223145961761475, "learning_rate": 2.8380000000000003e-05, "loss": 0.1226, "step": 1420 }, { "grad_norm": 1.462886929512024, "learning_rate": 2.858e-05, "loss": 0.1107, "step": 1430 }, { "grad_norm": 1.3069902658462524, "learning_rate": 2.8780000000000002e-05, "loss": 0.1177, "step": 1440 }, { "grad_norm": 1.3319061994552612, "learning_rate": 2.898e-05, "loss": 0.099, "step": 1450 }, { "grad_norm": 1.3519988059997559, "learning_rate": 2.9180000000000002e-05, "loss": 0.1116, "step": 1460 }, { "grad_norm": 1.008862853050232, "learning_rate": 2.9380000000000003e-05, "loss": 0.1095, "step": 1470 }, { "grad_norm": 1.6593354940414429, "learning_rate": 2.958e-05, "loss": 0.1175, "step": 1480 }, { "grad_norm": 1.3916335105895996, "learning_rate": 2.9780000000000003e-05, "loss": 0.1181, "step": 1490 }, { "grad_norm": 1.650828242301941, "learning_rate": 2.998e-05, "loss": 0.1146, "step": 1500 }, { "grad_norm": 1.2685716152191162, "learning_rate": 3.0180000000000002e-05, "loss": 0.1176, "step": 1510 }, { "grad_norm": 1.510684609413147, "learning_rate": 3.0380000000000004e-05, "loss": 0.112, "step": 1520 }, { "grad_norm": 1.432042121887207, "learning_rate": 3.058e-05, "loss": 0.1113, "step": 1530 }, { "grad_norm": 1.7365586757659912, "learning_rate": 3.078e-05, "loss": 0.1009, "step": 1540 }, { "grad_norm": 1.3169975280761719, "learning_rate": 3.0980000000000005e-05, "loss": 0.1088, "step": 1550 }, { "grad_norm": 1.4847275018692017, "learning_rate": 3.118e-05, "loss": 0.11, "step": 1560 }, { "grad_norm": 1.5091370344161987, "learning_rate": 3.138e-05, "loss": 0.1107, "step": 1570 }, { "grad_norm": 1.3874123096466064, "learning_rate": 3.1580000000000006e-05, "loss": 0.1096, "step": 1580 }, { "grad_norm": 1.6959023475646973, "learning_rate": 3.1780000000000004e-05, "loss": 0.1058, "step": 1590 }, { "grad_norm": 1.6006110906600952, "learning_rate": 3.198e-05, "loss": 0.1139, "step": 1600 }, { "grad_norm": 1.3088146448135376, "learning_rate": 3.218e-05, "loss": 0.1072, "step": 1610 }, { "grad_norm": 1.1733874082565308, "learning_rate": 3.238e-05, "loss": 0.1047, "step": 1620 }, { "grad_norm": 1.708408236503601, "learning_rate": 3.2579999999999996e-05, "loss": 0.1037, "step": 1630 }, { "grad_norm": 1.2776528596878052, "learning_rate": 3.278e-05, "loss": 0.1053, "step": 1640 }, { "grad_norm": 1.4028071165084839, "learning_rate": 3.298e-05, "loss": 0.1046, "step": 1650 }, { "grad_norm": 1.1378097534179688, "learning_rate": 3.318e-05, "loss": 0.1055, "step": 1660 }, { "grad_norm": 1.6558804512023926, "learning_rate": 3.338e-05, "loss": 0.1093, "step": 1670 }, { "grad_norm": 1.1552854776382446, "learning_rate": 3.358e-05, "loss": 0.0966, "step": 1680 }, { "grad_norm": 1.4640536308288574, "learning_rate": 3.378e-05, "loss": 0.1098, "step": 1690 }, { "grad_norm": 1.322249174118042, "learning_rate": 3.398e-05, "loss": 0.103, "step": 1700 }, { "grad_norm": 1.3498573303222656, "learning_rate": 3.418e-05, "loss": 0.1097, "step": 1710 }, { "grad_norm": 1.262957215309143, "learning_rate": 3.438e-05, "loss": 0.1034, "step": 1720 }, { "grad_norm": 1.4252980947494507, "learning_rate": 3.4580000000000004e-05, "loss": 0.1028, "step": 1730 }, { "grad_norm": 1.2226507663726807, "learning_rate": 3.478e-05, "loss": 0.0976, "step": 1740 }, { "grad_norm": 1.3647156953811646, "learning_rate": 3.498e-05, "loss": 0.1039, "step": 1750 }, { "grad_norm": 1.2517162561416626, "learning_rate": 3.518e-05, "loss": 0.1066, "step": 1760 }, { "grad_norm": 0.861154317855835, "learning_rate": 3.5380000000000003e-05, "loss": 0.0944, "step": 1770 }, { "grad_norm": 1.3860739469528198, "learning_rate": 3.558e-05, "loss": 0.1087, "step": 1780 }, { "grad_norm": 1.2621777057647705, "learning_rate": 3.578e-05, "loss": 0.0947, "step": 1790 }, { "grad_norm": 1.1672967672348022, "learning_rate": 3.5980000000000004e-05, "loss": 0.0979, "step": 1800 }, { "grad_norm": 1.1447454690933228, "learning_rate": 3.618e-05, "loss": 0.0967, "step": 1810 }, { "grad_norm": 1.183205485343933, "learning_rate": 3.638e-05, "loss": 0.0962, "step": 1820 }, { "grad_norm": 1.0532835721969604, "learning_rate": 3.6580000000000006e-05, "loss": 0.0954, "step": 1830 }, { "grad_norm": 1.1037495136260986, "learning_rate": 3.6780000000000004e-05, "loss": 0.0971, "step": 1840 }, { "grad_norm": 1.0365946292877197, "learning_rate": 3.698e-05, "loss": 0.0842, "step": 1850 }, { "grad_norm": 1.3786957263946533, "learning_rate": 3.7180000000000007e-05, "loss": 0.0956, "step": 1860 }, { "grad_norm": 1.1241066455841064, "learning_rate": 3.7380000000000005e-05, "loss": 0.0938, "step": 1870 }, { "grad_norm": 1.217797040939331, "learning_rate": 3.758e-05, "loss": 0.0885, "step": 1880 }, { "grad_norm": 1.19753098487854, "learning_rate": 3.778000000000001e-05, "loss": 0.0976, "step": 1890 }, { "grad_norm": 1.3078787326812744, "learning_rate": 3.7980000000000006e-05, "loss": 0.101, "step": 1900 }, { "grad_norm": 1.1819777488708496, "learning_rate": 3.818e-05, "loss": 0.0995, "step": 1910 }, { "grad_norm": 1.1380770206451416, "learning_rate": 3.838e-05, "loss": 0.1035, "step": 1920 }, { "grad_norm": 1.1521198749542236, "learning_rate": 3.858e-05, "loss": 0.1001, "step": 1930 }, { "grad_norm": 1.1119909286499023, "learning_rate": 3.878e-05, "loss": 0.1035, "step": 1940 }, { "grad_norm": 1.0273287296295166, "learning_rate": 3.898e-05, "loss": 0.1018, "step": 1950 }, { "grad_norm": 1.1000126600265503, "learning_rate": 3.918e-05, "loss": 0.1093, "step": 1960 }, { "grad_norm": 0.8933774828910828, "learning_rate": 3.938e-05, "loss": 0.1103, "step": 1970 }, { "grad_norm": 1.1168996095657349, "learning_rate": 3.958e-05, "loss": 0.0951, "step": 1980 }, { "grad_norm": 1.3976420164108276, "learning_rate": 3.978e-05, "loss": 0.0962, "step": 1990 }, { "grad_norm": 1.1261975765228271, "learning_rate": 3.998e-05, "loss": 0.1124, "step": 2000 }, { "grad_norm": 1.0097042322158813, "learning_rate": 4.018e-05, "loss": 0.0971, "step": 2010 }, { "grad_norm": 0.856428325176239, "learning_rate": 4.038e-05, "loss": 0.0952, "step": 2020 }, { "grad_norm": 0.8297505974769592, "learning_rate": 4.058e-05, "loss": 0.0967, "step": 2030 }, { "grad_norm": 0.9606473445892334, "learning_rate": 4.078e-05, "loss": 0.0957, "step": 2040 }, { "grad_norm": 1.0479447841644287, "learning_rate": 4.0980000000000004e-05, "loss": 0.0974, "step": 2050 }, { "grad_norm": 1.1613229513168335, "learning_rate": 4.118e-05, "loss": 0.1059, "step": 2060 }, { "grad_norm": 0.9652323126792908, "learning_rate": 4.138e-05, "loss": 0.1007, "step": 2070 }, { "grad_norm": 1.3068994283676147, "learning_rate": 4.1580000000000005e-05, "loss": 0.0918, "step": 2080 }, { "grad_norm": 0.9245490431785583, "learning_rate": 4.178e-05, "loss": 0.0946, "step": 2090 }, { "grad_norm": 1.0429331064224243, "learning_rate": 4.198e-05, "loss": 0.0931, "step": 2100 }, { "grad_norm": 1.0073847770690918, "learning_rate": 4.2180000000000006e-05, "loss": 0.0813, "step": 2110 }, { "grad_norm": 1.215219259262085, "learning_rate": 4.2380000000000004e-05, "loss": 0.0979, "step": 2120 }, { "grad_norm": 1.1025471687316895, "learning_rate": 4.258e-05, "loss": 0.0943, "step": 2130 }, { "grad_norm": 1.096553087234497, "learning_rate": 4.278e-05, "loss": 0.0907, "step": 2140 }, { "grad_norm": 0.8908363580703735, "learning_rate": 4.2980000000000005e-05, "loss": 0.09, "step": 2150 }, { "grad_norm": 1.1317164897918701, "learning_rate": 4.318e-05, "loss": 0.0877, "step": 2160 }, { "grad_norm": 0.8051722645759583, "learning_rate": 4.338e-05, "loss": 0.0889, "step": 2170 }, { "grad_norm": 1.2600350379943848, "learning_rate": 4.3580000000000006e-05, "loss": 0.0888, "step": 2180 }, { "grad_norm": 1.0949082374572754, "learning_rate": 4.3780000000000004e-05, "loss": 0.0962, "step": 2190 }, { "grad_norm": 0.9003851413726807, "learning_rate": 4.398e-05, "loss": 0.0942, "step": 2200 }, { "grad_norm": 0.9931448101997375, "learning_rate": 4.418000000000001e-05, "loss": 0.0817, "step": 2210 }, { "grad_norm": 1.5315364599227905, "learning_rate": 4.438e-05, "loss": 0.0909, "step": 2220 }, { "grad_norm": 0.9725093841552734, "learning_rate": 4.458e-05, "loss": 0.0934, "step": 2230 }, { "grad_norm": 0.7677587270736694, "learning_rate": 4.478e-05, "loss": 0.0974, "step": 2240 }, { "grad_norm": 0.784747302532196, "learning_rate": 4.498e-05, "loss": 0.1018, "step": 2250 }, { "grad_norm": 0.9413838386535645, "learning_rate": 4.518e-05, "loss": 0.0826, "step": 2260 }, { "grad_norm": 1.1606714725494385, "learning_rate": 4.538e-05, "loss": 0.079, "step": 2270 }, { "grad_norm": 1.0361566543579102, "learning_rate": 4.558e-05, "loss": 0.0975, "step": 2280 }, { "grad_norm": 1.0695616006851196, "learning_rate": 4.578e-05, "loss": 0.0869, "step": 2290 }, { "grad_norm": 0.8362181186676025, "learning_rate": 4.5980000000000004e-05, "loss": 0.0923, "step": 2300 }, { "grad_norm": 1.049444317817688, "learning_rate": 4.618e-05, "loss": 0.0934, "step": 2310 }, { "grad_norm": 0.773941695690155, "learning_rate": 4.638e-05, "loss": 0.0822, "step": 2320 }, { "grad_norm": 1.139686107635498, "learning_rate": 4.6580000000000005e-05, "loss": 0.091, "step": 2330 }, { "grad_norm": 0.7111517786979675, "learning_rate": 4.678e-05, "loss": 0.0876, "step": 2340 }, { "grad_norm": 1.08384108543396, "learning_rate": 4.698e-05, "loss": 0.0907, "step": 2350 }, { "grad_norm": 1.0586307048797607, "learning_rate": 4.718e-05, "loss": 0.0874, "step": 2360 }, { "grad_norm": 0.9089347124099731, "learning_rate": 4.7380000000000004e-05, "loss": 0.0907, "step": 2370 }, { "grad_norm": 0.8797197341918945, "learning_rate": 4.758e-05, "loss": 0.0879, "step": 2380 }, { "grad_norm": 0.9668692946434021, "learning_rate": 4.778e-05, "loss": 0.0996, "step": 2390 }, { "grad_norm": 0.8799086809158325, "learning_rate": 4.7980000000000005e-05, "loss": 0.0888, "step": 2400 }, { "grad_norm": 1.0102177858352661, "learning_rate": 4.818e-05, "loss": 0.0859, "step": 2410 }, { "grad_norm": 0.8223481178283691, "learning_rate": 4.838e-05, "loss": 0.0854, "step": 2420 }, { "grad_norm": 1.0871100425720215, "learning_rate": 4.8580000000000006e-05, "loss": 0.0848, "step": 2430 }, { "grad_norm": 0.8422275185585022, "learning_rate": 4.8780000000000004e-05, "loss": 0.0856, "step": 2440 }, { "grad_norm": 0.8731896877288818, "learning_rate": 4.898e-05, "loss": 0.0865, "step": 2450 }, { "grad_norm": 0.9828234910964966, "learning_rate": 4.918000000000001e-05, "loss": 0.0815, "step": 2460 }, { "grad_norm": 0.8781130313873291, "learning_rate": 4.9380000000000005e-05, "loss": 0.0853, "step": 2470 }, { "grad_norm": 0.8877680897712708, "learning_rate": 4.958e-05, "loss": 0.0932, "step": 2480 }, { "grad_norm": 0.8104763031005859, "learning_rate": 4.978e-05, "loss": 0.0824, "step": 2490 }, { "grad_norm": 0.7932062149047852, "learning_rate": 4.9980000000000006e-05, "loss": 0.0844, "step": 2500 }, { "grad_norm": 0.9729063510894775, "learning_rate": 5.0180000000000004e-05, "loss": 0.0724, "step": 2510 }, { "grad_norm": 0.8251408934593201, "learning_rate": 5.038e-05, "loss": 0.0763, "step": 2520 }, { "grad_norm": 0.812582790851593, "learning_rate": 5.058000000000001e-05, "loss": 0.091, "step": 2530 }, { "grad_norm": 0.7264870405197144, "learning_rate": 5.0780000000000005e-05, "loss": 0.0942, "step": 2540 }, { "grad_norm": 1.0052839517593384, "learning_rate": 5.098e-05, "loss": 0.0839, "step": 2550 }, { "grad_norm": 0.8519619703292847, "learning_rate": 5.118000000000001e-05, "loss": 0.0889, "step": 2560 }, { "grad_norm": 0.8198904395103455, "learning_rate": 5.1380000000000006e-05, "loss": 0.0849, "step": 2570 }, { "grad_norm": 0.9072335362434387, "learning_rate": 5.1580000000000004e-05, "loss": 0.0937, "step": 2580 }, { "grad_norm": 0.8841065764427185, "learning_rate": 5.178000000000001e-05, "loss": 0.0902, "step": 2590 }, { "grad_norm": 0.6875423789024353, "learning_rate": 5.198000000000001e-05, "loss": 0.088, "step": 2600 }, { "grad_norm": 0.857256293296814, "learning_rate": 5.2180000000000005e-05, "loss": 0.0856, "step": 2610 }, { "grad_norm": 0.9570732116699219, "learning_rate": 5.238000000000001e-05, "loss": 0.0816, "step": 2620 }, { "grad_norm": 0.6422113180160522, "learning_rate": 5.258000000000001e-05, "loss": 0.0868, "step": 2630 }, { "grad_norm": 0.9579894542694092, "learning_rate": 5.2780000000000006e-05, "loss": 0.1013, "step": 2640 }, { "grad_norm": 0.8130826950073242, "learning_rate": 5.2980000000000004e-05, "loss": 0.0845, "step": 2650 }, { "grad_norm": 0.8297514319419861, "learning_rate": 5.318000000000001e-05, "loss": 0.0771, "step": 2660 }, { "grad_norm": 0.9439654350280762, "learning_rate": 5.338000000000001e-05, "loss": 0.0807, "step": 2670 }, { "grad_norm": 0.6653961539268494, "learning_rate": 5.3580000000000005e-05, "loss": 0.0778, "step": 2680 }, { "grad_norm": 0.7107208967208862, "learning_rate": 5.378e-05, "loss": 0.0796, "step": 2690 }, { "grad_norm": 0.7031088471412659, "learning_rate": 5.3979999999999995e-05, "loss": 0.09, "step": 2700 }, { "grad_norm": 0.7514981031417847, "learning_rate": 5.418e-05, "loss": 0.0849, "step": 2710 }, { "grad_norm": 0.9804263114929199, "learning_rate": 5.438e-05, "loss": 0.0878, "step": 2720 }, { "grad_norm": 0.984539270401001, "learning_rate": 5.4579999999999996e-05, "loss": 0.0854, "step": 2730 }, { "grad_norm": 0.9428192377090454, "learning_rate": 5.478e-05, "loss": 0.0804, "step": 2740 }, { "grad_norm": 0.7810980677604675, "learning_rate": 5.498e-05, "loss": 0.076, "step": 2750 }, { "grad_norm": 0.584133505821228, "learning_rate": 5.518e-05, "loss": 0.0692, "step": 2760 }, { "grad_norm": 0.6875228881835938, "learning_rate": 5.538e-05, "loss": 0.0799, "step": 2770 }, { "grad_norm": 0.7095732688903809, "learning_rate": 5.558e-05, "loss": 0.0675, "step": 2780 }, { "grad_norm": 0.6867661476135254, "learning_rate": 5.578e-05, "loss": 0.0782, "step": 2790 }, { "grad_norm": 0.9318858981132507, "learning_rate": 5.5979999999999996e-05, "loss": 0.0775, "step": 2800 }, { "grad_norm": 0.6417056918144226, "learning_rate": 5.618e-05, "loss": 0.0722, "step": 2810 }, { "grad_norm": 0.8512111306190491, "learning_rate": 5.638e-05, "loss": 0.0714, "step": 2820 }, { "grad_norm": 0.8993717432022095, "learning_rate": 5.658e-05, "loss": 0.0692, "step": 2830 }, { "grad_norm": 0.6779329776763916, "learning_rate": 5.678e-05, "loss": 0.0702, "step": 2840 }, { "grad_norm": 0.5727977752685547, "learning_rate": 5.698e-05, "loss": 0.0744, "step": 2850 }, { "grad_norm": 0.8506995439529419, "learning_rate": 5.718e-05, "loss": 0.0917, "step": 2860 }, { "grad_norm": 0.6828613877296448, "learning_rate": 5.738e-05, "loss": 0.0726, "step": 2870 }, { "grad_norm": 0.7491737008094788, "learning_rate": 5.758e-05, "loss": 0.0774, "step": 2880 }, { "grad_norm": 0.9719417691230774, "learning_rate": 5.778e-05, "loss": 0.0788, "step": 2890 }, { "grad_norm": 0.6660922169685364, "learning_rate": 5.7980000000000004e-05, "loss": 0.0713, "step": 2900 }, { "grad_norm": 0.6533383727073669, "learning_rate": 5.818e-05, "loss": 0.0849, "step": 2910 }, { "grad_norm": 0.663981020450592, "learning_rate": 5.838e-05, "loss": 0.0844, "step": 2920 }, { "grad_norm": 0.719357430934906, "learning_rate": 5.858e-05, "loss": 0.0886, "step": 2930 }, { "grad_norm": 0.7175427079200745, "learning_rate": 5.878e-05, "loss": 0.0752, "step": 2940 }, { "grad_norm": 0.6750261187553406, "learning_rate": 5.898e-05, "loss": 0.0757, "step": 2950 }, { "grad_norm": 0.727814257144928, "learning_rate": 5.918e-05, "loss": 0.0766, "step": 2960 }, { "grad_norm": 0.7766228318214417, "learning_rate": 5.9380000000000004e-05, "loss": 0.0755, "step": 2970 }, { "grad_norm": 0.7372186779975891, "learning_rate": 5.958e-05, "loss": 0.0824, "step": 2980 }, { "grad_norm": 0.600877583026886, "learning_rate": 5.978e-05, "loss": 0.0654, "step": 2990 }, { "grad_norm": 0.7467097043991089, "learning_rate": 5.9980000000000005e-05, "loss": 0.0794, "step": 3000 }, { "grad_norm": 0.6808333992958069, "learning_rate": 6.018e-05, "loss": 0.078, "step": 3010 }, { "grad_norm": 0.495544970035553, "learning_rate": 6.038e-05, "loss": 0.0845, "step": 3020 }, { "grad_norm": 0.5764254927635193, "learning_rate": 6.0580000000000006e-05, "loss": 0.0753, "step": 3030 }, { "grad_norm": 0.846410870552063, "learning_rate": 6.0780000000000004e-05, "loss": 0.0801, "step": 3040 }, { "grad_norm": 0.6397631168365479, "learning_rate": 6.098e-05, "loss": 0.0723, "step": 3050 }, { "grad_norm": 0.5315824747085571, "learning_rate": 6.118000000000001e-05, "loss": 0.0673, "step": 3060 }, { "grad_norm": 0.8218473792076111, "learning_rate": 6.138e-05, "loss": 0.0783, "step": 3070 }, { "grad_norm": 0.6852237582206726, "learning_rate": 6.158e-05, "loss": 0.0803, "step": 3080 }, { "grad_norm": 0.7143826484680176, "learning_rate": 6.178000000000001e-05, "loss": 0.074, "step": 3090 }, { "grad_norm": 0.7483090758323669, "learning_rate": 6.198e-05, "loss": 0.0804, "step": 3100 }, { "grad_norm": 0.8317491412162781, "learning_rate": 6.218e-05, "loss": 0.0838, "step": 3110 }, { "grad_norm": 0.7646452784538269, "learning_rate": 6.238000000000001e-05, "loss": 0.0732, "step": 3120 }, { "grad_norm": 0.783631443977356, "learning_rate": 6.258e-05, "loss": 0.0843, "step": 3130 }, { "grad_norm": 0.7437312006950378, "learning_rate": 6.278e-05, "loss": 0.0842, "step": 3140 }, { "grad_norm": 0.6731628179550171, "learning_rate": 6.298000000000001e-05, "loss": 0.0745, "step": 3150 }, { "grad_norm": 0.7607911825180054, "learning_rate": 6.318e-05, "loss": 0.084, "step": 3160 }, { "grad_norm": 0.4912475645542145, "learning_rate": 6.338e-05, "loss": 0.068, "step": 3170 }, { "grad_norm": 0.7493554949760437, "learning_rate": 6.358000000000001e-05, "loss": 0.0776, "step": 3180 }, { "grad_norm": 0.9089020490646362, "learning_rate": 6.378e-05, "loss": 0.0894, "step": 3190 }, { "grad_norm": 0.6497768759727478, "learning_rate": 6.398000000000001e-05, "loss": 0.0781, "step": 3200 }, { "grad_norm": 0.9334494471549988, "learning_rate": 6.418000000000001e-05, "loss": 0.077, "step": 3210 }, { "grad_norm": 1.1049327850341797, "learning_rate": 6.438e-05, "loss": 0.0797, "step": 3220 }, { "grad_norm": 0.7053990960121155, "learning_rate": 6.458000000000001e-05, "loss": 0.0813, "step": 3230 }, { "grad_norm": 0.8259148597717285, "learning_rate": 6.478000000000001e-05, "loss": 0.0773, "step": 3240 }, { "grad_norm": 0.6796258091926575, "learning_rate": 6.498e-05, "loss": 0.082, "step": 3250 }, { "grad_norm": 0.5291226506233215, "learning_rate": 6.518000000000001e-05, "loss": 0.0777, "step": 3260 }, { "grad_norm": 0.6771771311759949, "learning_rate": 6.538000000000001e-05, "loss": 0.0702, "step": 3270 }, { "grad_norm": 0.5519546270370483, "learning_rate": 6.558e-05, "loss": 0.0795, "step": 3280 }, { "grad_norm": 0.7435193061828613, "learning_rate": 6.578000000000001e-05, "loss": 0.0745, "step": 3290 }, { "grad_norm": 0.6808909773826599, "learning_rate": 6.598e-05, "loss": 0.0761, "step": 3300 }, { "grad_norm": 0.6827627420425415, "learning_rate": 6.618e-05, "loss": 0.0788, "step": 3310 }, { "grad_norm": 0.5869776606559753, "learning_rate": 6.638e-05, "loss": 0.0696, "step": 3320 }, { "grad_norm": 0.9396904110908508, "learning_rate": 6.658e-05, "loss": 0.0718, "step": 3330 }, { "grad_norm": 0.727669894695282, "learning_rate": 6.678e-05, "loss": 0.0762, "step": 3340 }, { "grad_norm": 0.5646992921829224, "learning_rate": 6.698e-05, "loss": 0.078, "step": 3350 }, { "grad_norm": 0.5698143243789673, "learning_rate": 6.718e-05, "loss": 0.0709, "step": 3360 }, { "grad_norm": 0.7778666019439697, "learning_rate": 6.738e-05, "loss": 0.0712, "step": 3370 }, { "grad_norm": 1.0334298610687256, "learning_rate": 6.758e-05, "loss": 0.0769, "step": 3380 }, { "grad_norm": 0.7395018339157104, "learning_rate": 6.778e-05, "loss": 0.0737, "step": 3390 }, { "grad_norm": 0.8913915753364563, "learning_rate": 6.798e-05, "loss": 0.0843, "step": 3400 }, { "grad_norm": 0.5268679261207581, "learning_rate": 6.818e-05, "loss": 0.0723, "step": 3410 }, { "grad_norm": 0.6583300828933716, "learning_rate": 6.838e-05, "loss": 0.0661, "step": 3420 }, { "grad_norm": 0.7444570064544678, "learning_rate": 6.858e-05, "loss": 0.0767, "step": 3430 }, { "grad_norm": 0.6741772890090942, "learning_rate": 6.878e-05, "loss": 0.0717, "step": 3440 }, { "grad_norm": 0.6784478425979614, "learning_rate": 6.898e-05, "loss": 0.0751, "step": 3450 }, { "grad_norm": 0.7168152332305908, "learning_rate": 6.918e-05, "loss": 0.0673, "step": 3460 }, { "grad_norm": 0.5885259509086609, "learning_rate": 6.938e-05, "loss": 0.0773, "step": 3470 }, { "grad_norm": 0.625619113445282, "learning_rate": 6.958e-05, "loss": 0.0651, "step": 3480 }, { "grad_norm": 0.7113298177719116, "learning_rate": 6.978e-05, "loss": 0.0761, "step": 3490 }, { "grad_norm": 0.6929978728294373, "learning_rate": 6.998e-05, "loss": 0.0673, "step": 3500 }, { "grad_norm": 0.728690505027771, "learning_rate": 7.018e-05, "loss": 0.0736, "step": 3510 }, { "grad_norm": 0.6411104798316956, "learning_rate": 7.038e-05, "loss": 0.0621, "step": 3520 }, { "grad_norm": 0.6817249655723572, "learning_rate": 7.058e-05, "loss": 0.0689, "step": 3530 }, { "grad_norm": 0.5806167125701904, "learning_rate": 7.078e-05, "loss": 0.0834, "step": 3540 }, { "grad_norm": 0.5402979254722595, "learning_rate": 7.098e-05, "loss": 0.0711, "step": 3550 }, { "grad_norm": 0.5669298768043518, "learning_rate": 7.118e-05, "loss": 0.0718, "step": 3560 }, { "grad_norm": 0.6522741913795471, "learning_rate": 7.138e-05, "loss": 0.0714, "step": 3570 }, { "grad_norm": 1.0146198272705078, "learning_rate": 7.158e-05, "loss": 0.064, "step": 3580 }, { "grad_norm": 0.6896697282791138, "learning_rate": 7.178000000000001e-05, "loss": 0.0798, "step": 3590 }, { "grad_norm": 0.6256049275398254, "learning_rate": 7.198e-05, "loss": 0.0754, "step": 3600 }, { "grad_norm": 0.5702405571937561, "learning_rate": 7.218e-05, "loss": 0.0742, "step": 3610 }, { "grad_norm": 0.6473242044448853, "learning_rate": 7.238000000000001e-05, "loss": 0.0757, "step": 3620 }, { "grad_norm": 0.6914260387420654, "learning_rate": 7.258e-05, "loss": 0.0783, "step": 3630 }, { "grad_norm": 0.6713892817497253, "learning_rate": 7.278e-05, "loss": 0.0769, "step": 3640 }, { "grad_norm": 0.5596297383308411, "learning_rate": 7.298000000000001e-05, "loss": 0.0803, "step": 3650 }, { "grad_norm": 0.49075937271118164, "learning_rate": 7.318e-05, "loss": 0.0669, "step": 3660 }, { "grad_norm": 0.5055199265480042, "learning_rate": 7.338e-05, "loss": 0.0697, "step": 3670 }, { "grad_norm": 0.5792491436004639, "learning_rate": 7.358000000000001e-05, "loss": 0.0669, "step": 3680 }, { "grad_norm": 0.544502317905426, "learning_rate": 7.378e-05, "loss": 0.0693, "step": 3690 }, { "grad_norm": 0.6983371376991272, "learning_rate": 7.398e-05, "loss": 0.0692, "step": 3700 }, { "grad_norm": 0.5878961086273193, "learning_rate": 7.418000000000001e-05, "loss": 0.0687, "step": 3710 }, { "grad_norm": 0.7127764821052551, "learning_rate": 7.438e-05, "loss": 0.0656, "step": 3720 }, { "grad_norm": 0.6415865421295166, "learning_rate": 7.458000000000001e-05, "loss": 0.0686, "step": 3730 }, { "grad_norm": 0.5155027508735657, "learning_rate": 7.478e-05, "loss": 0.0678, "step": 3740 }, { "grad_norm": 0.583268404006958, "learning_rate": 7.498e-05, "loss": 0.0698, "step": 3750 }, { "grad_norm": 0.7324720621109009, "learning_rate": 7.518000000000001e-05, "loss": 0.0758, "step": 3760 }, { "grad_norm": 0.780573844909668, "learning_rate": 7.538e-05, "loss": 0.07, "step": 3770 }, { "grad_norm": 0.5533287525177002, "learning_rate": 7.558e-05, "loss": 0.0658, "step": 3780 }, { "grad_norm": 0.5688812136650085, "learning_rate": 7.578000000000001e-05, "loss": 0.0762, "step": 3790 }, { "grad_norm": 0.7023118138313293, "learning_rate": 7.598e-05, "loss": 0.0753, "step": 3800 }, { "grad_norm": 0.6739596128463745, "learning_rate": 7.618e-05, "loss": 0.0726, "step": 3810 }, { "grad_norm": 0.488687127828598, "learning_rate": 7.638000000000001e-05, "loss": 0.0709, "step": 3820 }, { "grad_norm": 0.565336287021637, "learning_rate": 7.658e-05, "loss": 0.0617, "step": 3830 }, { "grad_norm": 0.6788327693939209, "learning_rate": 7.678000000000001e-05, "loss": 0.0636, "step": 3840 }, { "grad_norm": 0.5672218203544617, "learning_rate": 7.698000000000001e-05, "loss": 0.0688, "step": 3850 }, { "grad_norm": 0.725165843963623, "learning_rate": 7.718e-05, "loss": 0.0667, "step": 3860 }, { "grad_norm": 0.5552905201911926, "learning_rate": 7.738000000000001e-05, "loss": 0.0732, "step": 3870 }, { "grad_norm": 0.5871935486793518, "learning_rate": 7.758000000000001e-05, "loss": 0.0804, "step": 3880 }, { "grad_norm": 0.812216579914093, "learning_rate": 7.778e-05, "loss": 0.0721, "step": 3890 }, { "grad_norm": 0.743348240852356, "learning_rate": 7.798000000000001e-05, "loss": 0.0711, "step": 3900 }, { "grad_norm": 0.4964960217475891, "learning_rate": 7.818000000000001e-05, "loss": 0.0639, "step": 3910 }, { "grad_norm": 0.608456552028656, "learning_rate": 7.838e-05, "loss": 0.0686, "step": 3920 }, { "grad_norm": 0.5461157560348511, "learning_rate": 7.858000000000001e-05, "loss": 0.0649, "step": 3930 }, { "grad_norm": 0.5798609256744385, "learning_rate": 7.878e-05, "loss": 0.0756, "step": 3940 }, { "grad_norm": 0.5873997807502747, "learning_rate": 7.897999999999999e-05, "loss": 0.076, "step": 3950 }, { "grad_norm": 0.49911317229270935, "learning_rate": 7.918e-05, "loss": 0.0767, "step": 3960 }, { "grad_norm": 0.5114102959632874, "learning_rate": 7.938e-05, "loss": 0.0697, "step": 3970 }, { "grad_norm": 0.5821915864944458, "learning_rate": 7.958e-05, "loss": 0.0655, "step": 3980 }, { "grad_norm": 0.8597725629806519, "learning_rate": 7.978e-05, "loss": 0.0671, "step": 3990 }, { "grad_norm": 0.5585454106330872, "learning_rate": 7.998e-05, "loss": 0.0823, "step": 4000 }, { "grad_norm": 0.43395811319351196, "learning_rate": 8.018e-05, "loss": 0.0743, "step": 4010 }, { "grad_norm": 0.5978971123695374, "learning_rate": 8.038e-05, "loss": 0.0798, "step": 4020 }, { "grad_norm": 0.5861995220184326, "learning_rate": 8.058e-05, "loss": 0.0799, "step": 4030 }, { "grad_norm": 0.5677093267440796, "learning_rate": 8.078e-05, "loss": 0.0709, "step": 4040 }, { "grad_norm": 0.6224440932273865, "learning_rate": 8.098e-05, "loss": 0.072, "step": 4050 }, { "grad_norm": 0.6574553847312927, "learning_rate": 8.118e-05, "loss": 0.0679, "step": 4060 }, { "grad_norm": 0.4835106134414673, "learning_rate": 8.138e-05, "loss": 0.0724, "step": 4070 }, { "grad_norm": 0.5309403538703918, "learning_rate": 8.158e-05, "loss": 0.0632, "step": 4080 }, { "grad_norm": 0.6054978370666504, "learning_rate": 8.178e-05, "loss": 0.0782, "step": 4090 }, { "grad_norm": 0.48820793628692627, "learning_rate": 8.198e-05, "loss": 0.065, "step": 4100 }, { "grad_norm": 0.6458432078361511, "learning_rate": 8.218e-05, "loss": 0.0653, "step": 4110 }, { "grad_norm": 0.492869108915329, "learning_rate": 8.238000000000001e-05, "loss": 0.07, "step": 4120 }, { "grad_norm": 0.5692965388298035, "learning_rate": 8.258e-05, "loss": 0.0682, "step": 4130 }, { "grad_norm": 0.6462418437004089, "learning_rate": 8.278e-05, "loss": 0.0749, "step": 4140 }, { "grad_norm": 0.5332359671592712, "learning_rate": 8.298000000000001e-05, "loss": 0.0763, "step": 4150 }, { "grad_norm": 0.48575395345687866, "learning_rate": 8.318e-05, "loss": 0.0684, "step": 4160 }, { "grad_norm": 0.6302199959754944, "learning_rate": 8.338e-05, "loss": 0.0756, "step": 4170 }, { "grad_norm": 0.5844523310661316, "learning_rate": 8.358e-05, "loss": 0.0718, "step": 4180 }, { "grad_norm": 0.5593042373657227, "learning_rate": 8.378e-05, "loss": 0.0648, "step": 4190 }, { "grad_norm": 0.472177654504776, "learning_rate": 8.398e-05, "loss": 0.0709, "step": 4200 }, { "grad_norm": 0.4204423427581787, "learning_rate": 8.418e-05, "loss": 0.0602, "step": 4210 }, { "grad_norm": 0.5694792866706848, "learning_rate": 8.438e-05, "loss": 0.0775, "step": 4220 }, { "grad_norm": 0.5901989936828613, "learning_rate": 8.458e-05, "loss": 0.0664, "step": 4230 }, { "grad_norm": 0.5694212317466736, "learning_rate": 8.478e-05, "loss": 0.0605, "step": 4240 }, { "grad_norm": 0.44711053371429443, "learning_rate": 8.498e-05, "loss": 0.058, "step": 4250 }, { "grad_norm": 0.6746779680252075, "learning_rate": 8.518000000000001e-05, "loss": 0.0833, "step": 4260 }, { "grad_norm": 0.5489562749862671, "learning_rate": 8.538e-05, "loss": 0.0693, "step": 4270 }, { "grad_norm": 0.5111076831817627, "learning_rate": 8.558e-05, "loss": 0.0696, "step": 4280 }, { "grad_norm": 0.7805970907211304, "learning_rate": 8.578000000000001e-05, "loss": 0.0673, "step": 4290 }, { "grad_norm": 0.5361025929450989, "learning_rate": 8.598e-05, "loss": 0.0664, "step": 4300 }, { "grad_norm": 0.5352867841720581, "learning_rate": 8.618e-05, "loss": 0.0595, "step": 4310 }, { "grad_norm": 0.43269598484039307, "learning_rate": 8.638000000000001e-05, "loss": 0.0588, "step": 4320 }, { "grad_norm": 0.5919831991195679, "learning_rate": 8.658e-05, "loss": 0.0564, "step": 4330 }, { "grad_norm": 0.5983070731163025, "learning_rate": 8.678e-05, "loss": 0.0741, "step": 4340 }, { "grad_norm": 0.45673853158950806, "learning_rate": 8.698000000000001e-05, "loss": 0.0628, "step": 4350 }, { "grad_norm": 0.5287368893623352, "learning_rate": 8.718e-05, "loss": 0.0728, "step": 4360 }, { "grad_norm": 0.5538188815116882, "learning_rate": 8.738000000000001e-05, "loss": 0.0641, "step": 4370 }, { "grad_norm": 0.4827328324317932, "learning_rate": 8.758000000000001e-05, "loss": 0.0653, "step": 4380 }, { "grad_norm": 0.3589581549167633, "learning_rate": 8.778e-05, "loss": 0.0581, "step": 4390 }, { "grad_norm": 0.5307368040084839, "learning_rate": 8.798000000000001e-05, "loss": 0.0653, "step": 4400 }, { "grad_norm": 0.5026375651359558, "learning_rate": 8.818000000000001e-05, "loss": 0.0527, "step": 4410 }, { "grad_norm": 0.47930335998535156, "learning_rate": 8.838e-05, "loss": 0.0621, "step": 4420 }, { "grad_norm": 0.45860275626182556, "learning_rate": 8.858000000000001e-05, "loss": 0.0591, "step": 4430 }, { "grad_norm": 0.5524290800094604, "learning_rate": 8.878000000000001e-05, "loss": 0.062, "step": 4440 }, { "grad_norm": 0.6018303632736206, "learning_rate": 8.898e-05, "loss": 0.062, "step": 4450 }, { "grad_norm": 0.5606292486190796, "learning_rate": 8.918000000000001e-05, "loss": 0.0727, "step": 4460 }, { "grad_norm": 0.5733949542045593, "learning_rate": 8.938e-05, "loss": 0.0799, "step": 4470 }, { "grad_norm": 0.5251554846763611, "learning_rate": 8.958e-05, "loss": 0.0659, "step": 4480 }, { "grad_norm": 0.6789089441299438, "learning_rate": 8.978000000000001e-05, "loss": 0.0781, "step": 4490 }, { "grad_norm": 0.6006895899772644, "learning_rate": 8.998e-05, "loss": 0.0668, "step": 4500 }, { "grad_norm": 0.5243695378303528, "learning_rate": 9.018000000000001e-05, "loss": 0.0786, "step": 4510 }, { "grad_norm": 0.5276608467102051, "learning_rate": 9.038000000000001e-05, "loss": 0.0793, "step": 4520 }, { "grad_norm": 0.4818456470966339, "learning_rate": 9.058e-05, "loss": 0.0748, "step": 4530 }, { "grad_norm": 0.46986573934555054, "learning_rate": 9.078000000000001e-05, "loss": 0.077, "step": 4540 }, { "grad_norm": 0.5033984184265137, "learning_rate": 9.098000000000001e-05, "loss": 0.0637, "step": 4550 }, { "grad_norm": 0.47905537486076355, "learning_rate": 9.118e-05, "loss": 0.0637, "step": 4560 }, { "grad_norm": 0.5633179545402527, "learning_rate": 9.138e-05, "loss": 0.0688, "step": 4570 }, { "grad_norm": 0.6283842325210571, "learning_rate": 9.158e-05, "loss": 0.0608, "step": 4580 }, { "grad_norm": 0.5044362545013428, "learning_rate": 9.178e-05, "loss": 0.0706, "step": 4590 }, { "grad_norm": 0.44225332140922546, "learning_rate": 9.198e-05, "loss": 0.0639, "step": 4600 }, { "grad_norm": 0.5093481540679932, "learning_rate": 9.218e-05, "loss": 0.0646, "step": 4610 }, { "grad_norm": 0.4953937828540802, "learning_rate": 9.238e-05, "loss": 0.0626, "step": 4620 }, { "grad_norm": 0.5530047416687012, "learning_rate": 9.258e-05, "loss": 0.0702, "step": 4630 }, { "grad_norm": 0.5669158697128296, "learning_rate": 9.278e-05, "loss": 0.0638, "step": 4640 }, { "grad_norm": 0.5714249014854431, "learning_rate": 9.298e-05, "loss": 0.067, "step": 4650 }, { "grad_norm": 0.5842466950416565, "learning_rate": 9.318e-05, "loss": 0.0742, "step": 4660 }, { "grad_norm": 0.5311502814292908, "learning_rate": 9.338e-05, "loss": 0.0636, "step": 4670 }, { "grad_norm": 0.49794629216194153, "learning_rate": 9.358e-05, "loss": 0.074, "step": 4680 }, { "grad_norm": 0.5987551212310791, "learning_rate": 9.378e-05, "loss": 0.0645, "step": 4690 }, { "grad_norm": 0.5080201029777527, "learning_rate": 9.398e-05, "loss": 0.0636, "step": 4700 }, { "grad_norm": 0.46786603331565857, "learning_rate": 9.418e-05, "loss": 0.0656, "step": 4710 }, { "grad_norm": 0.5118427276611328, "learning_rate": 9.438e-05, "loss": 0.0777, "step": 4720 }, { "grad_norm": 0.6808896660804749, "learning_rate": 9.458e-05, "loss": 0.0736, "step": 4730 }, { "grad_norm": 0.4309031367301941, "learning_rate": 9.478e-05, "loss": 0.065, "step": 4740 }, { "grad_norm": 0.47506821155548096, "learning_rate": 9.498e-05, "loss": 0.0731, "step": 4750 }, { "grad_norm": 0.5642704963684082, "learning_rate": 9.518000000000001e-05, "loss": 0.0744, "step": 4760 }, { "grad_norm": 0.4994054436683655, "learning_rate": 9.538e-05, "loss": 0.0664, "step": 4770 }, { "grad_norm": 0.5278519988059998, "learning_rate": 9.558e-05, "loss": 0.0671, "step": 4780 }, { "grad_norm": 0.5900089740753174, "learning_rate": 9.578000000000001e-05, "loss": 0.0698, "step": 4790 }, { "grad_norm": 0.4894128143787384, "learning_rate": 9.598e-05, "loss": 0.0728, "step": 4800 }, { "grad_norm": 0.5192451477050781, "learning_rate": 9.618e-05, "loss": 0.0647, "step": 4810 }, { "grad_norm": 0.38500213623046875, "learning_rate": 9.638000000000001e-05, "loss": 0.0651, "step": 4820 }, { "grad_norm": 0.4942716360092163, "learning_rate": 9.658e-05, "loss": 0.0671, "step": 4830 }, { "grad_norm": 0.4908624589443207, "learning_rate": 9.678e-05, "loss": 0.071, "step": 4840 }, { "grad_norm": 0.5899695754051208, "learning_rate": 9.698000000000001e-05, "loss": 0.0675, "step": 4850 }, { "grad_norm": 0.6551910042762756, "learning_rate": 9.718e-05, "loss": 0.0716, "step": 4860 }, { "grad_norm": 0.5707669258117676, "learning_rate": 9.738e-05, "loss": 0.069, "step": 4870 }, { "grad_norm": 0.4489106833934784, "learning_rate": 9.758000000000001e-05, "loss": 0.0592, "step": 4880 }, { "grad_norm": 0.42826634645462036, "learning_rate": 9.778e-05, "loss": 0.0653, "step": 4890 }, { "grad_norm": 0.4936700463294983, "learning_rate": 9.798000000000001e-05, "loss": 0.0648, "step": 4900 }, { "grad_norm": 0.5479432344436646, "learning_rate": 9.818000000000001e-05, "loss": 0.0731, "step": 4910 }, { "grad_norm": 0.6700789928436279, "learning_rate": 9.838e-05, "loss": 0.0697, "step": 4920 }, { "grad_norm": 0.48181942105293274, "learning_rate": 9.858000000000001e-05, "loss": 0.0643, "step": 4930 }, { "grad_norm": 0.49263232946395874, "learning_rate": 9.878e-05, "loss": 0.0659, "step": 4940 }, { "grad_norm": 0.638244092464447, "learning_rate": 9.898e-05, "loss": 0.0937, "step": 4950 }, { "grad_norm": 0.5617243051528931, "learning_rate": 9.918000000000001e-05, "loss": 0.0807, "step": 4960 }, { "grad_norm": 0.5582041144371033, "learning_rate": 9.938e-05, "loss": 0.0808, "step": 4970 }, { "grad_norm": 0.40571045875549316, "learning_rate": 9.958e-05, "loss": 0.0728, "step": 4980 }, { "grad_norm": 0.4709371328353882, "learning_rate": 9.978000000000001e-05, "loss": 0.0765, "step": 4990 }, { "grad_norm": 0.5428269505500793, "learning_rate": 9.998e-05, "loss": 0.0753, "step": 5000 }, { "grad_norm": 0.4644511938095093, "learning_rate": 9.999999778549045e-05, "loss": 0.0603, "step": 5010 }, { "grad_norm": 0.5012858510017395, "learning_rate": 9.999999013039593e-05, "loss": 0.0753, "step": 5020 }, { "grad_norm": 0.5124042630195618, "learning_rate": 9.999997700737766e-05, "loss": 0.0722, "step": 5030 }, { "grad_norm": 0.4216383099555969, "learning_rate": 9.999995841643709e-05, "loss": 0.0629, "step": 5040 }, { "grad_norm": 0.5065250992774963, "learning_rate": 9.999993435757623e-05, "loss": 0.0615, "step": 5050 }, { "grad_norm": 0.5301805138587952, "learning_rate": 9.999990483079773e-05, "loss": 0.0661, "step": 5060 }, { "grad_norm": 0.4691994786262512, "learning_rate": 9.999986983610481e-05, "loss": 0.0661, "step": 5070 }, { "grad_norm": 0.5693060755729675, "learning_rate": 9.99998293735013e-05, "loss": 0.0636, "step": 5080 }, { "grad_norm": 0.47503462433815, "learning_rate": 9.999978344299161e-05, "loss": 0.0716, "step": 5090 }, { "grad_norm": 0.5082978010177612, "learning_rate": 9.99997320445808e-05, "loss": 0.0679, "step": 5100 }, { "grad_norm": 0.4646107852458954, "learning_rate": 9.999967517827444e-05, "loss": 0.0575, "step": 5110 }, { "grad_norm": 0.4864778518676758, "learning_rate": 9.999961284407879e-05, "loss": 0.0648, "step": 5120 }, { "grad_norm": 0.6000898480415344, "learning_rate": 9.999954504200067e-05, "loss": 0.0639, "step": 5130 }, { "grad_norm": 0.5130972266197205, "learning_rate": 9.999947177204744e-05, "loss": 0.0608, "step": 5140 }, { "grad_norm": 0.4634381830692291, "learning_rate": 9.999939303422718e-05, "loss": 0.0727, "step": 5150 }, { "grad_norm": 0.44148072600364685, "learning_rate": 9.999930882854847e-05, "loss": 0.063, "step": 5160 }, { "grad_norm": 0.4357107877731323, "learning_rate": 9.999921915502051e-05, "loss": 0.0674, "step": 5170 }, { "grad_norm": 0.5137631297111511, "learning_rate": 9.99991240136531e-05, "loss": 0.0566, "step": 5180 }, { "grad_norm": 0.539608359336853, "learning_rate": 9.999902340445668e-05, "loss": 0.0665, "step": 5190 }, { "grad_norm": 0.45470643043518066, "learning_rate": 9.999891732744224e-05, "loss": 0.0652, "step": 5200 }, { "grad_norm": 0.4831715226173401, "learning_rate": 9.999880578262135e-05, "loss": 0.0665, "step": 5210 }, { "grad_norm": 0.5551472306251526, "learning_rate": 9.999868877000624e-05, "loss": 0.0676, "step": 5220 }, { "grad_norm": 0.41186386346817017, "learning_rate": 9.99985662896097e-05, "loss": 0.0635, "step": 5230 }, { "grad_norm": 0.4290256202220917, "learning_rate": 9.999843834144513e-05, "loss": 0.0689, "step": 5240 }, { "grad_norm": 0.4794294238090515, "learning_rate": 9.99983049255265e-05, "loss": 0.0686, "step": 5250 }, { "grad_norm": 0.34987038373947144, "learning_rate": 9.999816604186843e-05, "loss": 0.0629, "step": 5260 }, { "grad_norm": 0.426460325717926, "learning_rate": 9.999802169048609e-05, "loss": 0.0705, "step": 5270 }, { "grad_norm": 0.508003830909729, "learning_rate": 9.999787187139527e-05, "loss": 0.0681, "step": 5280 }, { "grad_norm": 0.41614627838134766, "learning_rate": 9.999771658461234e-05, "loss": 0.0649, "step": 5290 }, { "grad_norm": 0.514342725276947, "learning_rate": 9.999755583015431e-05, "loss": 0.0614, "step": 5300 }, { "grad_norm": 0.5054588317871094, "learning_rate": 9.999738960803874e-05, "loss": 0.0611, "step": 5310 }, { "grad_norm": 0.3895703852176666, "learning_rate": 9.99972179182838e-05, "loss": 0.0618, "step": 5320 }, { "grad_norm": 0.40275099873542786, "learning_rate": 9.99970407609083e-05, "loss": 0.065, "step": 5330 }, { "grad_norm": 0.33797889947891235, "learning_rate": 9.999685813593159e-05, "loss": 0.0556, "step": 5340 }, { "grad_norm": 0.4465845227241516, "learning_rate": 9.999667004337362e-05, "loss": 0.0602, "step": 5350 }, { "grad_norm": 0.44945380091667175, "learning_rate": 9.9996476483255e-05, "loss": 0.0587, "step": 5360 }, { "grad_norm": 0.5049943923950195, "learning_rate": 9.999627745559688e-05, "loss": 0.0629, "step": 5370 }, { "grad_norm": 0.46413522958755493, "learning_rate": 9.999607296042101e-05, "loss": 0.0634, "step": 5380 }, { "grad_norm": 0.48722729086875916, "learning_rate": 9.99958629977498e-05, "loss": 0.0658, "step": 5390 }, { "grad_norm": 0.37375590205192566, "learning_rate": 9.999564756760615e-05, "loss": 0.0553, "step": 5400 }, { "grad_norm": 0.43490612506866455, "learning_rate": 9.999542667001366e-05, "loss": 0.0651, "step": 5410 }, { "grad_norm": 0.5588338375091553, "learning_rate": 9.999520030499647e-05, "loss": 0.062, "step": 5420 }, { "grad_norm": 0.46720147132873535, "learning_rate": 9.999496847257936e-05, "loss": 0.0617, "step": 5430 }, { "grad_norm": 0.4278126657009125, "learning_rate": 9.999473117278764e-05, "loss": 0.0598, "step": 5440 }, { "grad_norm": 0.6067655682563782, "learning_rate": 9.999448840564731e-05, "loss": 0.0606, "step": 5450 }, { "grad_norm": 0.47906333208084106, "learning_rate": 9.999424017118488e-05, "loss": 0.0644, "step": 5460 }, { "grad_norm": 0.5337560772895813, "learning_rate": 9.999398646942751e-05, "loss": 0.0673, "step": 5470 }, { "grad_norm": 0.4326343238353729, "learning_rate": 9.999372730040296e-05, "loss": 0.0657, "step": 5480 }, { "grad_norm": 0.4527367949485779, "learning_rate": 9.999346266413953e-05, "loss": 0.0614, "step": 5490 }, { "grad_norm": 0.496304988861084, "learning_rate": 9.99931925606662e-05, "loss": 0.0662, "step": 5500 }, { "grad_norm": 0.4333779811859131, "learning_rate": 9.99929169900125e-05, "loss": 0.063, "step": 5510 }, { "grad_norm": 0.48398733139038086, "learning_rate": 9.999263595220855e-05, "loss": 0.0628, "step": 5520 }, { "grad_norm": 0.5379939079284668, "learning_rate": 9.99923494472851e-05, "loss": 0.0657, "step": 5530 }, { "grad_norm": 0.3578220009803772, "learning_rate": 9.999205747527348e-05, "loss": 0.0529, "step": 5540 }, { "grad_norm": 0.5203117728233337, "learning_rate": 9.999176003620561e-05, "loss": 0.0611, "step": 5550 }, { "grad_norm": 0.49612146615982056, "learning_rate": 9.999145713011405e-05, "loss": 0.0675, "step": 5560 }, { "grad_norm": 0.5242570638656616, "learning_rate": 9.999114875703186e-05, "loss": 0.0633, "step": 5570 }, { "grad_norm": 0.5583491921424866, "learning_rate": 9.999083491699281e-05, "loss": 0.0741, "step": 5580 }, { "grad_norm": 0.4255281388759613, "learning_rate": 9.999051561003123e-05, "loss": 0.0591, "step": 5590 }, { "grad_norm": 0.4800891876220703, "learning_rate": 9.999019083618202e-05, "loss": 0.0633, "step": 5600 }, { "grad_norm": 0.5458996891975403, "learning_rate": 9.99898605954807e-05, "loss": 0.0556, "step": 5610 }, { "grad_norm": 0.6131067872047424, "learning_rate": 9.998952488796338e-05, "loss": 0.0545, "step": 5620 }, { "grad_norm": 0.527489960193634, "learning_rate": 9.998918371366676e-05, "loss": 0.0514, "step": 5630 }, { "grad_norm": 0.4142662584781647, "learning_rate": 9.99888370726282e-05, "loss": 0.0544, "step": 5640 }, { "grad_norm": 0.4268535375595093, "learning_rate": 9.998848496488556e-05, "loss": 0.0592, "step": 5650 }, { "grad_norm": 0.5080171227455139, "learning_rate": 9.998812739047736e-05, "loss": 0.0643, "step": 5660 }, { "grad_norm": 0.4442363977432251, "learning_rate": 9.99877643494427e-05, "loss": 0.0709, "step": 5670 }, { "grad_norm": 0.42720091342926025, "learning_rate": 9.998739584182128e-05, "loss": 0.056, "step": 5680 }, { "grad_norm": 0.35932767391204834, "learning_rate": 9.998702186765342e-05, "loss": 0.0632, "step": 5690 }, { "grad_norm": 0.5374000072479248, "learning_rate": 9.998664242698e-05, "loss": 0.0571, "step": 5700 }, { "grad_norm": 0.8226503729820251, "learning_rate": 9.998625751984251e-05, "loss": 0.061, "step": 5710 }, { "grad_norm": 0.506425678730011, "learning_rate": 9.998586714628307e-05, "loss": 0.0603, "step": 5720 }, { "grad_norm": 0.5329839587211609, "learning_rate": 9.998547130634432e-05, "loss": 0.0628, "step": 5730 }, { "grad_norm": 0.5513833165168762, "learning_rate": 9.99850700000696e-05, "loss": 0.0586, "step": 5740 }, { "grad_norm": 0.4479277729988098, "learning_rate": 9.998466322750278e-05, "loss": 0.0624, "step": 5750 }, { "grad_norm": 0.5143094658851624, "learning_rate": 9.998425098868834e-05, "loss": 0.0545, "step": 5760 }, { "grad_norm": 0.40038713812828064, "learning_rate": 9.998383328367136e-05, "loss": 0.0614, "step": 5770 }, { "grad_norm": 0.3659718930721283, "learning_rate": 9.99834101124975e-05, "loss": 0.0527, "step": 5780 }, { "grad_norm": 0.4497443735599518, "learning_rate": 9.998298147521309e-05, "loss": 0.0559, "step": 5790 }, { "grad_norm": 0.5939666032791138, "learning_rate": 9.998254737186496e-05, "loss": 0.0596, "step": 5800 }, { "grad_norm": 0.41053974628448486, "learning_rate": 9.99821078025006e-05, "loss": 0.0545, "step": 5810 }, { "grad_norm": 0.49026474356651306, "learning_rate": 9.998166276716807e-05, "loss": 0.0618, "step": 5820 }, { "grad_norm": 0.3338240683078766, "learning_rate": 9.998121226591606e-05, "loss": 0.0531, "step": 5830 }, { "grad_norm": 0.42953768372535706, "learning_rate": 9.998075629879382e-05, "loss": 0.0542, "step": 5840 }, { "grad_norm": 0.42552536725997925, "learning_rate": 9.99802948658512e-05, "loss": 0.0604, "step": 5850 }, { "grad_norm": 0.41968509554862976, "learning_rate": 9.99798279671387e-05, "loss": 0.0701, "step": 5860 }, { "grad_norm": 0.3823744058609009, "learning_rate": 9.997935560270734e-05, "loss": 0.071, "step": 5870 }, { "grad_norm": 0.4100053608417511, "learning_rate": 9.997887777260879e-05, "loss": 0.0558, "step": 5880 }, { "grad_norm": 0.4328686594963074, "learning_rate": 9.997839447689532e-05, "loss": 0.0628, "step": 5890 }, { "grad_norm": 0.4530455470085144, "learning_rate": 9.997790571561978e-05, "loss": 0.0659, "step": 5900 }, { "grad_norm": 0.4285752773284912, "learning_rate": 9.99774114888356e-05, "loss": 0.0616, "step": 5910 }, { "grad_norm": 0.4455105662345886, "learning_rate": 9.997691179659684e-05, "loss": 0.0676, "step": 5920 }, { "grad_norm": 0.38748612999916077, "learning_rate": 9.997640663895815e-05, "loss": 0.0544, "step": 5930 }, { "grad_norm": 0.4677203893661499, "learning_rate": 9.997589601597477e-05, "loss": 0.0585, "step": 5940 }, { "grad_norm": 0.4211668372154236, "learning_rate": 9.997537992770252e-05, "loss": 0.0545, "step": 5950 }, { "grad_norm": 0.4213336408138275, "learning_rate": 9.997485837419788e-05, "loss": 0.058, "step": 5960 }, { "grad_norm": 0.4482356011867523, "learning_rate": 9.997433135551786e-05, "loss": 0.0613, "step": 5970 }, { "grad_norm": 0.5056943893432617, "learning_rate": 9.997379887172009e-05, "loss": 0.0564, "step": 5980 }, { "grad_norm": 0.4607807397842407, "learning_rate": 9.997326092286281e-05, "loss": 0.0638, "step": 5990 }, { "grad_norm": 0.614204466342926, "learning_rate": 9.997271750900486e-05, "loss": 0.0656, "step": 6000 }, { "grad_norm": 0.5554153919219971, "learning_rate": 9.997216863020565e-05, "loss": 0.0693, "step": 6010 }, { "grad_norm": 0.49864572286605835, "learning_rate": 9.99716142865252e-05, "loss": 0.0641, "step": 6020 }, { "grad_norm": 0.46066299080848694, "learning_rate": 9.997105447802415e-05, "loss": 0.0593, "step": 6030 }, { "grad_norm": 0.5524208545684814, "learning_rate": 9.997048920476373e-05, "loss": 0.0603, "step": 6040 }, { "grad_norm": 0.43039458990097046, "learning_rate": 9.996991846680572e-05, "loss": 0.0634, "step": 6050 }, { "grad_norm": 0.3720848560333252, "learning_rate": 9.996934226421257e-05, "loss": 0.0559, "step": 6060 }, { "grad_norm": 0.39239534735679626, "learning_rate": 9.996876059704726e-05, "loss": 0.0536, "step": 6070 }, { "grad_norm": 0.5595059394836426, "learning_rate": 9.996817346537343e-05, "loss": 0.0563, "step": 6080 }, { "grad_norm": 0.42064663767814636, "learning_rate": 9.996758086925526e-05, "loss": 0.0662, "step": 6090 }, { "grad_norm": 0.345564603805542, "learning_rate": 9.996698280875759e-05, "loss": 0.0562, "step": 6100 }, { "grad_norm": 0.3413321077823639, "learning_rate": 9.99663792839458e-05, "loss": 0.0595, "step": 6110 }, { "grad_norm": 0.3425450921058655, "learning_rate": 9.99657702948859e-05, "loss": 0.0558, "step": 6120 }, { "grad_norm": 0.3816792964935303, "learning_rate": 9.996515584164448e-05, "loss": 0.0524, "step": 6130 }, { "grad_norm": 0.8617020845413208, "learning_rate": 9.996453592428873e-05, "loss": 0.0665, "step": 6140 }, { "grad_norm": 0.3964192569255829, "learning_rate": 9.996391054288646e-05, "loss": 0.0585, "step": 6150 }, { "grad_norm": 0.4135698974132538, "learning_rate": 9.996327969750605e-05, "loss": 0.0636, "step": 6160 }, { "grad_norm": 0.4787285029888153, "learning_rate": 9.996264338821649e-05, "loss": 0.066, "step": 6170 }, { "grad_norm": 0.31758010387420654, "learning_rate": 9.996200161508735e-05, "loss": 0.0545, "step": 6180 }, { "grad_norm": 0.39767178893089294, "learning_rate": 9.996135437818885e-05, "loss": 0.0631, "step": 6190 }, { "grad_norm": 0.44828787446022034, "learning_rate": 9.996070167759175e-05, "loss": 0.0598, "step": 6200 }, { "grad_norm": 0.4983219504356384, "learning_rate": 9.996004351336743e-05, "loss": 0.0723, "step": 6210 }, { "grad_norm": 0.42601433396339417, "learning_rate": 9.995937988558785e-05, "loss": 0.058, "step": 6220 }, { "grad_norm": 0.39424562454223633, "learning_rate": 9.995871079432561e-05, "loss": 0.0576, "step": 6230 }, { "grad_norm": 0.34219321608543396, "learning_rate": 9.995803623965389e-05, "loss": 0.0504, "step": 6240 }, { "grad_norm": 0.42347344756126404, "learning_rate": 9.995735622164641e-05, "loss": 0.0582, "step": 6250 }, { "grad_norm": 0.4162420630455017, "learning_rate": 9.995667074037758e-05, "loss": 0.0638, "step": 6260 }, { "grad_norm": 0.3777949810028076, "learning_rate": 9.995597979592232e-05, "loss": 0.0605, "step": 6270 }, { "grad_norm": 0.4919721186161041, "learning_rate": 9.995528338835625e-05, "loss": 0.0583, "step": 6280 }, { "grad_norm": 0.4118482172489166, "learning_rate": 9.995458151775547e-05, "loss": 0.0645, "step": 6290 }, { "grad_norm": 0.3866031765937805, "learning_rate": 9.995387418419677e-05, "loss": 0.0599, "step": 6300 }, { "grad_norm": 0.37938040494918823, "learning_rate": 9.99531613877575e-05, "loss": 0.0534, "step": 6310 }, { "grad_norm": 0.4754638373851776, "learning_rate": 9.995244312851559e-05, "loss": 0.0551, "step": 6320 }, { "grad_norm": 0.3623580038547516, "learning_rate": 9.995171940654961e-05, "loss": 0.0498, "step": 6330 }, { "grad_norm": 0.40886566042900085, "learning_rate": 9.995099022193871e-05, "loss": 0.0556, "step": 6340 }, { "grad_norm": 0.449526309967041, "learning_rate": 9.995025557476261e-05, "loss": 0.0554, "step": 6350 }, { "grad_norm": 0.4195100963115692, "learning_rate": 9.994951546510165e-05, "loss": 0.0568, "step": 6360 }, { "grad_norm": 0.3547925651073456, "learning_rate": 9.994876989303679e-05, "loss": 0.0594, "step": 6370 }, { "grad_norm": 0.386791467666626, "learning_rate": 9.994801885864955e-05, "loss": 0.0599, "step": 6380 }, { "grad_norm": 0.35817351937294006, "learning_rate": 9.994726236202205e-05, "loss": 0.0539, "step": 6390 }, { "grad_norm": 0.512991189956665, "learning_rate": 9.994650040323704e-05, "loss": 0.0588, "step": 6400 }, { "grad_norm": 0.4047374427318573, "learning_rate": 9.994573298237784e-05, "loss": 0.0556, "step": 6410 }, { "grad_norm": 0.48971351981163025, "learning_rate": 9.994496009952837e-05, "loss": 0.0577, "step": 6420 }, { "grad_norm": 0.44855645298957825, "learning_rate": 9.994418175477316e-05, "loss": 0.0574, "step": 6430 }, { "grad_norm": 0.5468647480010986, "learning_rate": 9.994339794819733e-05, "loss": 0.0611, "step": 6440 }, { "grad_norm": 0.3880525231361389, "learning_rate": 9.994260867988658e-05, "loss": 0.0733, "step": 6450 }, { "grad_norm": 0.3562110960483551, "learning_rate": 9.994181394992723e-05, "loss": 0.0528, "step": 6460 }, { "grad_norm": 0.3369860053062439, "learning_rate": 9.994101375840618e-05, "loss": 0.0546, "step": 6470 }, { "grad_norm": 0.2956914007663727, "learning_rate": 9.994020810541098e-05, "loss": 0.0558, "step": 6480 }, { "grad_norm": 0.41209861636161804, "learning_rate": 9.99393969910297e-05, "loss": 0.0528, "step": 6490 }, { "grad_norm": 0.34301450848579407, "learning_rate": 9.993858041535104e-05, "loss": 0.0534, "step": 6500 }, { "grad_norm": 0.40438055992126465, "learning_rate": 9.99377583784643e-05, "loss": 0.052, "step": 6510 }, { "grad_norm": 0.4600892961025238, "learning_rate": 9.993693088045939e-05, "loss": 0.0529, "step": 6520 }, { "grad_norm": 0.3293594419956207, "learning_rate": 9.99360979214268e-05, "loss": 0.0517, "step": 6530 }, { "grad_norm": 0.33498111367225647, "learning_rate": 9.99352595014576e-05, "loss": 0.0533, "step": 6540 }, { "grad_norm": 0.38799095153808594, "learning_rate": 9.993441562064354e-05, "loss": 0.0502, "step": 6550 }, { "grad_norm": 0.379114031791687, "learning_rate": 9.993356627907685e-05, "loss": 0.0475, "step": 6560 }, { "grad_norm": 0.40418267250061035, "learning_rate": 9.99327114768504e-05, "loss": 0.0507, "step": 6570 }, { "grad_norm": 0.3614661693572998, "learning_rate": 9.99318512140577e-05, "loss": 0.05, "step": 6580 }, { "grad_norm": 0.3470165431499481, "learning_rate": 9.993098549079284e-05, "loss": 0.0516, "step": 6590 }, { "grad_norm": 0.31470492482185364, "learning_rate": 9.993011430715047e-05, "loss": 0.055, "step": 6600 }, { "grad_norm": 0.3346826732158661, "learning_rate": 9.992923766322586e-05, "loss": 0.0475, "step": 6610 }, { "grad_norm": 0.3468683660030365, "learning_rate": 9.99283555591149e-05, "loss": 0.0552, "step": 6620 }, { "grad_norm": 0.4406120479106903, "learning_rate": 9.992746799491404e-05, "loss": 0.0538, "step": 6630 }, { "grad_norm": 0.34432560205459595, "learning_rate": 9.992657497072033e-05, "loss": 0.0581, "step": 6640 }, { "grad_norm": 0.4283198416233063, "learning_rate": 9.992567648663147e-05, "loss": 0.0491, "step": 6650 }, { "grad_norm": 0.44196295738220215, "learning_rate": 9.992477254274568e-05, "loss": 0.0503, "step": 6660 }, { "grad_norm": 0.3720895051956177, "learning_rate": 9.992386313916183e-05, "loss": 0.0531, "step": 6670 }, { "grad_norm": 0.4168696999549866, "learning_rate": 9.992294827597934e-05, "loss": 0.059, "step": 6680 }, { "grad_norm": 0.5441315770149231, "learning_rate": 9.992202795329831e-05, "loss": 0.056, "step": 6690 }, { "grad_norm": 0.40624600648880005, "learning_rate": 9.992110217121936e-05, "loss": 0.0556, "step": 6700 }, { "grad_norm": 0.35226741433143616, "learning_rate": 9.992017092984372e-05, "loss": 0.0548, "step": 6710 }, { "grad_norm": 0.4956085979938507, "learning_rate": 9.991923422927326e-05, "loss": 0.0526, "step": 6720 }, { "grad_norm": 0.4150584638118744, "learning_rate": 9.991829206961037e-05, "loss": 0.0566, "step": 6730 }, { "grad_norm": 0.3187045454978943, "learning_rate": 9.991734445095813e-05, "loss": 0.0531, "step": 6740 }, { "grad_norm": 0.4119006395339966, "learning_rate": 9.991639137342015e-05, "loss": 0.0537, "step": 6750 }, { "grad_norm": 0.37421515583992004, "learning_rate": 9.991543283710064e-05, "loss": 0.0483, "step": 6760 }, { "grad_norm": 0.44766661524772644, "learning_rate": 9.991446884210445e-05, "loss": 0.0537, "step": 6770 }, { "grad_norm": 0.35091015696525574, "learning_rate": 9.9913499388537e-05, "loss": 0.0483, "step": 6780 }, { "grad_norm": 0.41174420714378357, "learning_rate": 9.99125244765043e-05, "loss": 0.0542, "step": 6790 }, { "grad_norm": 0.3496655821800232, "learning_rate": 9.991154410611296e-05, "loss": 0.0543, "step": 6800 }, { "grad_norm": 0.49570897221565247, "learning_rate": 9.99105582774702e-05, "loss": 0.0465, "step": 6810 }, { "grad_norm": 0.402892529964447, "learning_rate": 9.990956699068384e-05, "loss": 0.0604, "step": 6820 }, { "grad_norm": 0.39344704151153564, "learning_rate": 9.990857024586224e-05, "loss": 0.0538, "step": 6830 }, { "grad_norm": 0.36278533935546875, "learning_rate": 9.990756804311446e-05, "loss": 0.05, "step": 6840 }, { "grad_norm": 0.4516240358352661, "learning_rate": 9.990656038255006e-05, "loss": 0.0551, "step": 6850 }, { "grad_norm": 0.3395448625087738, "learning_rate": 9.990554726427926e-05, "loss": 0.0527, "step": 6860 }, { "grad_norm": 0.4092943072319031, "learning_rate": 9.990452868841284e-05, "loss": 0.0578, "step": 6870 }, { "grad_norm": 0.33714306354522705, "learning_rate": 9.99035046550622e-05, "loss": 0.0531, "step": 6880 }, { "grad_norm": 0.565619707107544, "learning_rate": 9.99024751643393e-05, "loss": 0.0478, "step": 6890 }, { "grad_norm": 0.40878769755363464, "learning_rate": 9.990144021635677e-05, "loss": 0.0588, "step": 6900 }, { "grad_norm": 0.4625092148780823, "learning_rate": 9.990039981122775e-05, "loss": 0.0551, "step": 6910 }, { "grad_norm": 0.39816930890083313, "learning_rate": 9.989935394906602e-05, "loss": 0.0551, "step": 6920 }, { "grad_norm": 0.4588523209095001, "learning_rate": 9.989830262998598e-05, "loss": 0.0555, "step": 6930 }, { "grad_norm": 0.49483925104141235, "learning_rate": 9.989724585410259e-05, "loss": 0.0584, "step": 6940 }, { "grad_norm": 0.39877450466156006, "learning_rate": 9.989618362153139e-05, "loss": 0.0517, "step": 6950 }, { "grad_norm": 0.4055769443511963, "learning_rate": 9.989511593238859e-05, "loss": 0.0541, "step": 6960 }, { "grad_norm": 0.4010087549686432, "learning_rate": 9.98940427867909e-05, "loss": 0.056, "step": 6970 }, { "grad_norm": 0.3573002517223358, "learning_rate": 9.989296418485573e-05, "loss": 0.0462, "step": 6980 }, { "grad_norm": 0.4189901351928711, "learning_rate": 9.989188012670101e-05, "loss": 0.0527, "step": 6990 }, { "grad_norm": 0.35880306363105774, "learning_rate": 9.989079061244528e-05, "loss": 0.0484, "step": 7000 }, { "grad_norm": 0.4238847494125366, "learning_rate": 9.988969564220769e-05, "loss": 0.0541, "step": 7010 }, { "grad_norm": 0.4250948429107666, "learning_rate": 9.988859521610801e-05, "loss": 0.0578, "step": 7020 }, { "grad_norm": 0.6419286727905273, "learning_rate": 9.988748933426656e-05, "loss": 0.0506, "step": 7030 }, { "grad_norm": 0.41450366377830505, "learning_rate": 9.988637799680428e-05, "loss": 0.0546, "step": 7040 }, { "grad_norm": 0.4254898726940155, "learning_rate": 9.98852612038427e-05, "loss": 0.0541, "step": 7050 }, { "grad_norm": 0.37049826979637146, "learning_rate": 9.988413895550397e-05, "loss": 0.052, "step": 7060 }, { "grad_norm": 0.4662364721298218, "learning_rate": 9.98830112519108e-05, "loss": 0.0548, "step": 7070 }, { "grad_norm": 0.3503289818763733, "learning_rate": 9.98818780931865e-05, "loss": 0.0527, "step": 7080 }, { "grad_norm": 0.2995888888835907, "learning_rate": 9.988073947945502e-05, "loss": 0.0514, "step": 7090 }, { "grad_norm": 0.38517123460769653, "learning_rate": 9.987959541084087e-05, "loss": 0.0487, "step": 7100 }, { "grad_norm": 0.3700242042541504, "learning_rate": 9.987844588746915e-05, "loss": 0.0557, "step": 7110 }, { "grad_norm": 0.40670472383499146, "learning_rate": 9.987729090946558e-05, "loss": 0.0528, "step": 7120 }, { "grad_norm": 0.3395320773124695, "learning_rate": 9.987613047695647e-05, "loss": 0.048, "step": 7130 }, { "grad_norm": 0.4132165312767029, "learning_rate": 9.987496459006871e-05, "loss": 0.0582, "step": 7140 }, { "grad_norm": 0.34410539269447327, "learning_rate": 9.987379324892982e-05, "loss": 0.0533, "step": 7150 }, { "grad_norm": 0.40609416365623474, "learning_rate": 9.987261645366788e-05, "loss": 0.0532, "step": 7160 }, { "grad_norm": 0.3591887354850769, "learning_rate": 9.987143420441158e-05, "loss": 0.0491, "step": 7170 }, { "grad_norm": 0.30287426710128784, "learning_rate": 9.987024650129022e-05, "loss": 0.0481, "step": 7180 }, { "grad_norm": 0.32588809728622437, "learning_rate": 9.986905334443368e-05, "loss": 0.0466, "step": 7190 }, { "grad_norm": 0.46134960651397705, "learning_rate": 9.986785473397245e-05, "loss": 0.0468, "step": 7200 }, { "grad_norm": 0.34240707755088806, "learning_rate": 9.98666506700376e-05, "loss": 0.0516, "step": 7210 }, { "grad_norm": 0.4140361249446869, "learning_rate": 9.986544115276081e-05, "loss": 0.0442, "step": 7220 }, { "grad_norm": 0.51502925157547, "learning_rate": 9.986422618227433e-05, "loss": 0.047, "step": 7230 }, { "grad_norm": 0.38299986720085144, "learning_rate": 9.986300575871106e-05, "loss": 0.0527, "step": 7240 }, { "grad_norm": 0.3472742736339569, "learning_rate": 9.986177988220444e-05, "loss": 0.0454, "step": 7250 }, { "grad_norm": 0.4065685570240021, "learning_rate": 9.986054855288856e-05, "loss": 0.0625, "step": 7260 }, { "grad_norm": 0.35621631145477295, "learning_rate": 9.985931177089802e-05, "loss": 0.0507, "step": 7270 }, { "grad_norm": 0.3625246286392212, "learning_rate": 9.985806953636814e-05, "loss": 0.0472, "step": 7280 }, { "grad_norm": 0.40893372893333435, "learning_rate": 9.985682184943471e-05, "loss": 0.0503, "step": 7290 }, { "grad_norm": 0.3218996822834015, "learning_rate": 9.98555687102342e-05, "loss": 0.043, "step": 7300 }, { "grad_norm": 0.3358239233493805, "learning_rate": 9.985431011890367e-05, "loss": 0.0517, "step": 7310 }, { "grad_norm": 0.3915499746799469, "learning_rate": 9.985304607558075e-05, "loss": 0.0466, "step": 7320 }, { "grad_norm": 0.5120510458946228, "learning_rate": 9.985177658040364e-05, "loss": 0.0467, "step": 7330 }, { "grad_norm": 0.43046095967292786, "learning_rate": 9.985050163351119e-05, "loss": 0.0506, "step": 7340 }, { "grad_norm": 0.384126752614975, "learning_rate": 9.984922123504286e-05, "loss": 0.056, "step": 7350 }, { "grad_norm": 0.31558293104171753, "learning_rate": 9.984793538513862e-05, "loss": 0.0468, "step": 7360 }, { "grad_norm": 0.43304744362831116, "learning_rate": 9.984664408393912e-05, "loss": 0.056, "step": 7370 }, { "grad_norm": 0.5177833437919617, "learning_rate": 9.984534733158556e-05, "loss": 0.0588, "step": 7380 }, { "grad_norm": 0.357433557510376, "learning_rate": 9.984404512821977e-05, "loss": 0.0488, "step": 7390 }, { "grad_norm": 0.31407463550567627, "learning_rate": 9.984273747398411e-05, "loss": 0.0479, "step": 7400 }, { "grad_norm": 0.3638628423213959, "learning_rate": 9.984142436902165e-05, "loss": 0.0519, "step": 7410 }, { "grad_norm": 0.3936910033226013, "learning_rate": 9.984010581347596e-05, "loss": 0.0516, "step": 7420 }, { "grad_norm": 0.32428547739982605, "learning_rate": 9.983878180749121e-05, "loss": 0.0472, "step": 7430 }, { "grad_norm": 0.4512236714363098, "learning_rate": 9.983745235121222e-05, "loss": 0.055, "step": 7440 }, { "grad_norm": 0.3775761127471924, "learning_rate": 9.983611744478438e-05, "loss": 0.0529, "step": 7450 }, { "grad_norm": 0.37046629190444946, "learning_rate": 9.983477708835365e-05, "loss": 0.051, "step": 7460 }, { "grad_norm": 0.3876641094684601, "learning_rate": 9.983343128206664e-05, "loss": 0.0555, "step": 7470 }, { "grad_norm": 0.673738956451416, "learning_rate": 9.983208002607049e-05, "loss": 0.0506, "step": 7480 }, { "grad_norm": 0.41357511281967163, "learning_rate": 9.9830723320513e-05, "loss": 0.0559, "step": 7490 }, { "grad_norm": 0.3796588182449341, "learning_rate": 9.982936116554254e-05, "loss": 0.0507, "step": 7500 }, { "grad_norm": 0.37393718957901, "learning_rate": 9.982799356130803e-05, "loss": 0.0535, "step": 7510 }, { "grad_norm": 0.33597666025161743, "learning_rate": 9.982662050795908e-05, "loss": 0.054, "step": 7520 }, { "grad_norm": 0.28676578402519226, "learning_rate": 9.982524200564583e-05, "loss": 0.0454, "step": 7530 }, { "grad_norm": 0.3263082802295685, "learning_rate": 9.982385805451901e-05, "loss": 0.0467, "step": 7540 }, { "grad_norm": 0.4436427056789398, "learning_rate": 9.982246865472998e-05, "loss": 0.0521, "step": 7550 }, { "grad_norm": 0.3014652729034424, "learning_rate": 9.982107380643069e-05, "loss": 0.0494, "step": 7560 }, { "grad_norm": 0.32462865114212036, "learning_rate": 9.981967350977368e-05, "loss": 0.0503, "step": 7570 }, { "grad_norm": 0.43280965089797974, "learning_rate": 9.981826776491208e-05, "loss": 0.0469, "step": 7580 }, { "grad_norm": 0.3608369827270508, "learning_rate": 9.98168565719996e-05, "loss": 0.0578, "step": 7590 }, { "grad_norm": 0.2721635699272156, "learning_rate": 9.98154399311906e-05, "loss": 0.0394, "step": 7600 }, { "grad_norm": 1.1957061290740967, "learning_rate": 9.981401784263997e-05, "loss": 0.0552, "step": 7610 }, { "grad_norm": 0.3119502365589142, "learning_rate": 9.981259030650326e-05, "loss": 0.058, "step": 7620 }, { "grad_norm": 0.33646491169929504, "learning_rate": 9.981115732293655e-05, "loss": 0.0536, "step": 7630 }, { "grad_norm": 0.40127241611480713, "learning_rate": 9.980971889209659e-05, "loss": 0.0589, "step": 7640 }, { "grad_norm": 0.4122830927371979, "learning_rate": 9.980827501414064e-05, "loss": 0.0537, "step": 7650 }, { "grad_norm": 0.4366772174835205, "learning_rate": 9.980682568922663e-05, "loss": 0.0506, "step": 7660 }, { "grad_norm": 0.33631303906440735, "learning_rate": 9.980537091751304e-05, "loss": 0.0545, "step": 7670 }, { "grad_norm": 0.26522013545036316, "learning_rate": 9.980391069915897e-05, "loss": 0.0507, "step": 7680 }, { "grad_norm": 0.3481731414794922, "learning_rate": 9.98024450343241e-05, "loss": 0.0533, "step": 7690 }, { "grad_norm": 0.36488795280456543, "learning_rate": 9.980097392316872e-05, "loss": 0.0556, "step": 7700 }, { "grad_norm": 0.3268296718597412, "learning_rate": 9.97994973658537e-05, "loss": 0.0561, "step": 7710 }, { "grad_norm": 0.5487844944000244, "learning_rate": 9.979801536254054e-05, "loss": 0.0564, "step": 7720 }, { "grad_norm": 0.419490784406662, "learning_rate": 9.979652791339127e-05, "loss": 0.0507, "step": 7730 }, { "grad_norm": 0.5390470027923584, "learning_rate": 9.97950350185686e-05, "loss": 0.0482, "step": 7740 }, { "grad_norm": 0.35406678915023804, "learning_rate": 9.979353667823574e-05, "loss": 0.053, "step": 7750 }, { "grad_norm": 0.39786049723625183, "learning_rate": 9.979203289255658e-05, "loss": 0.0598, "step": 7760 }, { "grad_norm": 0.3770468533039093, "learning_rate": 9.979052366169557e-05, "loss": 0.0615, "step": 7770 }, { "grad_norm": 0.3629057705402374, "learning_rate": 9.978900898581775e-05, "loss": 0.0568, "step": 7780 }, { "grad_norm": 0.38191038370132446, "learning_rate": 9.978748886508875e-05, "loss": 0.0583, "step": 7790 }, { "grad_norm": 0.36008182168006897, "learning_rate": 9.978596329967484e-05, "loss": 0.0624, "step": 7800 }, { "grad_norm": 0.41115814447402954, "learning_rate": 9.978443228974284e-05, "loss": 0.051, "step": 7810 }, { "grad_norm": 0.44092193245887756, "learning_rate": 9.978289583546015e-05, "loss": 0.0478, "step": 7820 }, { "grad_norm": 0.4015404284000397, "learning_rate": 9.978135393699484e-05, "loss": 0.05, "step": 7830 }, { "grad_norm": 0.4476138651371002, "learning_rate": 9.977980659451548e-05, "loss": 0.062, "step": 7840 }, { "grad_norm": 0.37302571535110474, "learning_rate": 9.977825380819135e-05, "loss": 0.049, "step": 7850 }, { "grad_norm": 0.41760873794555664, "learning_rate": 9.97766955781922e-05, "loss": 0.0496, "step": 7860 }, { "grad_norm": 0.3196607828140259, "learning_rate": 9.977513190468848e-05, "loss": 0.053, "step": 7870 }, { "grad_norm": 0.3525106906890869, "learning_rate": 9.977356278785116e-05, "loss": 0.049, "step": 7880 }, { "grad_norm": 0.2966672480106354, "learning_rate": 9.977198822785184e-05, "loss": 0.0452, "step": 7890 }, { "grad_norm": 0.37587741017341614, "learning_rate": 9.977040822486273e-05, "loss": 0.0503, "step": 7900 }, { "grad_norm": 0.3497025966644287, "learning_rate": 9.97688227790566e-05, "loss": 0.0524, "step": 7910 }, { "grad_norm": 0.3379359245300293, "learning_rate": 9.976723189060684e-05, "loss": 0.047, "step": 7920 }, { "grad_norm": 0.35920828580856323, "learning_rate": 9.976563555968742e-05, "loss": 0.0429, "step": 7930 }, { "grad_norm": 0.3310626447200775, "learning_rate": 9.976403378647292e-05, "loss": 0.0448, "step": 7940 }, { "grad_norm": 0.3087473213672638, "learning_rate": 9.97624265711385e-05, "loss": 0.0463, "step": 7950 }, { "grad_norm": 0.3178175091743469, "learning_rate": 9.976081391385993e-05, "loss": 0.0452, "step": 7960 }, { "grad_norm": 0.3578147888183594, "learning_rate": 9.975919581481356e-05, "loss": 0.0462, "step": 7970 }, { "grad_norm": 0.25471925735473633, "learning_rate": 9.975757227417634e-05, "loss": 0.0454, "step": 7980 }, { "grad_norm": 0.3742423951625824, "learning_rate": 9.975594329212586e-05, "loss": 0.0504, "step": 7990 }, { "grad_norm": 0.35239067673683167, "learning_rate": 9.97543088688402e-05, "loss": 0.0455, "step": 8000 }, { "grad_norm": 0.3326095640659332, "learning_rate": 9.975266900449814e-05, "loss": 0.0397, "step": 8010 }, { "grad_norm": 0.3884967863559723, "learning_rate": 9.975102369927898e-05, "loss": 0.0436, "step": 8020 }, { "grad_norm": 0.4008559286594391, "learning_rate": 9.974937295336269e-05, "loss": 0.0434, "step": 8030 }, { "grad_norm": 0.37743955850601196, "learning_rate": 9.974771676692975e-05, "loss": 0.0499, "step": 8040 }, { "grad_norm": 0.444466769695282, "learning_rate": 9.974605514016131e-05, "loss": 0.0515, "step": 8050 }, { "grad_norm": 0.47083550691604614, "learning_rate": 9.974438807323907e-05, "loss": 0.0513, "step": 8060 }, { "grad_norm": 0.325857937335968, "learning_rate": 9.974271556634535e-05, "loss": 0.0451, "step": 8070 }, { "grad_norm": 0.27605491876602173, "learning_rate": 9.974103761966302e-05, "loss": 0.0469, "step": 8080 }, { "grad_norm": 0.34423407912254333, "learning_rate": 9.973935423337563e-05, "loss": 0.0481, "step": 8090 }, { "grad_norm": 0.5677672028541565, "learning_rate": 9.973766540766722e-05, "loss": 0.0539, "step": 8100 }, { "grad_norm": 0.3998109996318817, "learning_rate": 9.97359711427225e-05, "loss": 0.0514, "step": 8110 }, { "grad_norm": 0.3092383146286011, "learning_rate": 9.973427143872677e-05, "loss": 0.0476, "step": 8120 }, { "grad_norm": 0.32017940282821655, "learning_rate": 9.973256629586589e-05, "loss": 0.0453, "step": 8130 }, { "grad_norm": 0.3527444899082184, "learning_rate": 9.973085571432632e-05, "loss": 0.0486, "step": 8140 }, { "grad_norm": 0.3089236319065094, "learning_rate": 9.972913969429513e-05, "loss": 0.0468, "step": 8150 }, { "grad_norm": 0.3043989837169647, "learning_rate": 9.972741823596e-05, "loss": 0.0487, "step": 8160 }, { "grad_norm": 0.30768758058547974, "learning_rate": 9.972569133950917e-05, "loss": 0.0441, "step": 8170 }, { "grad_norm": 0.257933109998703, "learning_rate": 9.972395900513151e-05, "loss": 0.0457, "step": 8180 }, { "grad_norm": 0.3420272171497345, "learning_rate": 9.972222123301645e-05, "loss": 0.0416, "step": 8190 }, { "grad_norm": 0.2994364798069, "learning_rate": 9.972047802335403e-05, "loss": 0.044, "step": 8200 }, { "grad_norm": 0.4104158282279968, "learning_rate": 9.971872937633488e-05, "loss": 0.0514, "step": 8210 }, { "grad_norm": 0.2749440371990204, "learning_rate": 9.971697529215024e-05, "loss": 0.0508, "step": 8220 }, { "grad_norm": 0.33501332998275757, "learning_rate": 9.971521577099192e-05, "loss": 0.0484, "step": 8230 }, { "grad_norm": 0.37572136521339417, "learning_rate": 9.971345081305236e-05, "loss": 0.0428, "step": 8240 }, { "grad_norm": 0.35907071828842163, "learning_rate": 9.971168041852456e-05, "loss": 0.0505, "step": 8250 }, { "grad_norm": 0.38814738392829895, "learning_rate": 9.970990458760215e-05, "loss": 0.054, "step": 8260 }, { "grad_norm": 0.3169155716896057, "learning_rate": 9.970812332047929e-05, "loss": 0.0456, "step": 8270 }, { "grad_norm": 0.37685197591781616, "learning_rate": 9.97063366173508e-05, "loss": 0.0487, "step": 8280 }, { "grad_norm": 0.41870319843292236, "learning_rate": 9.970454447841207e-05, "loss": 0.0474, "step": 8290 }, { "grad_norm": 0.3548044264316559, "learning_rate": 9.970274690385909e-05, "loss": 0.0513, "step": 8300 }, { "grad_norm": 0.37152817845344543, "learning_rate": 9.970094389388844e-05, "loss": 0.0461, "step": 8310 }, { "grad_norm": 0.4095340669155121, "learning_rate": 9.969913544869728e-05, "loss": 0.0485, "step": 8320 }, { "grad_norm": 0.27079278230667114, "learning_rate": 9.96973215684834e-05, "loss": 0.0492, "step": 8330 }, { "grad_norm": 0.34027135372161865, "learning_rate": 9.969550225344513e-05, "loss": 0.0516, "step": 8340 }, { "grad_norm": 0.43455618619918823, "learning_rate": 9.969367750378147e-05, "loss": 0.0487, "step": 8350 }, { "grad_norm": 0.3960484564304352, "learning_rate": 9.969184731969194e-05, "loss": 0.0483, "step": 8360 }, { "grad_norm": 0.4458215832710266, "learning_rate": 9.96900117013767e-05, "loss": 0.0492, "step": 8370 }, { "grad_norm": 0.3856180012226105, "learning_rate": 9.96881706490365e-05, "loss": 0.0515, "step": 8380 }, { "grad_norm": 0.37299445271492004, "learning_rate": 9.968632416287265e-05, "loss": 0.044, "step": 8390 }, { "grad_norm": 0.30591240525245667, "learning_rate": 9.96844722430871e-05, "loss": 0.0464, "step": 8400 }, { "grad_norm": 0.3312205672264099, "learning_rate": 9.968261488988235e-05, "loss": 0.0502, "step": 8410 }, { "grad_norm": 0.3004993498325348, "learning_rate": 9.968075210346155e-05, "loss": 0.0401, "step": 8420 }, { "grad_norm": 0.3403964340686798, "learning_rate": 9.967888388402839e-05, "loss": 0.0423, "step": 8430 }, { "grad_norm": 0.3361660838127136, "learning_rate": 9.967701023178717e-05, "loss": 0.0434, "step": 8440 }, { "grad_norm": 0.47279706597328186, "learning_rate": 9.967513114694282e-05, "loss": 0.049, "step": 8450 }, { "grad_norm": 0.3269754648208618, "learning_rate": 9.967324662970079e-05, "loss": 0.0394, "step": 8460 }, { "grad_norm": 0.2907603979110718, "learning_rate": 9.96713566802672e-05, "loss": 0.0381, "step": 8470 }, { "grad_norm": 0.31393423676490784, "learning_rate": 9.966946129884873e-05, "loss": 0.0489, "step": 8480 }, { "grad_norm": 0.3474614918231964, "learning_rate": 9.966756048565265e-05, "loss": 0.0476, "step": 8490 }, { "grad_norm": 0.30257952213287354, "learning_rate": 9.966565424088681e-05, "loss": 0.0427, "step": 8500 }, { "grad_norm": 0.36315110325813293, "learning_rate": 9.96637425647597e-05, "loss": 0.0466, "step": 8510 }, { "grad_norm": 0.3187565505504608, "learning_rate": 9.966182545748038e-05, "loss": 0.044, "step": 8520 }, { "grad_norm": 0.33957791328430176, "learning_rate": 9.96599029192585e-05, "loss": 0.0446, "step": 8530 }, { "grad_norm": 0.3106779158115387, "learning_rate": 9.965797495030428e-05, "loss": 0.0419, "step": 8540 }, { "grad_norm": 0.359354168176651, "learning_rate": 9.96560415508286e-05, "loss": 0.0476, "step": 8550 }, { "grad_norm": 0.3367865979671478, "learning_rate": 9.965410272104286e-05, "loss": 0.0421, "step": 8560 }, { "grad_norm": 0.4610930383205414, "learning_rate": 9.96521584611591e-05, "loss": 0.0454, "step": 8570 }, { "grad_norm": 0.34398430585861206, "learning_rate": 9.965020877138994e-05, "loss": 0.0442, "step": 8580 }, { "grad_norm": 0.3460339903831482, "learning_rate": 9.964825365194861e-05, "loss": 0.04, "step": 8590 }, { "grad_norm": 0.40391597151756287, "learning_rate": 9.96462931030489e-05, "loss": 0.0391, "step": 8600 }, { "grad_norm": 0.29682865738868713, "learning_rate": 9.96443271249052e-05, "loss": 0.0382, "step": 8610 }, { "grad_norm": 0.28675809502601624, "learning_rate": 9.964235571773255e-05, "loss": 0.0454, "step": 8620 }, { "grad_norm": 0.38405323028564453, "learning_rate": 9.96403788817465e-05, "loss": 0.0493, "step": 8630 }, { "grad_norm": 0.40547943115234375, "learning_rate": 9.963839661716325e-05, "loss": 0.0441, "step": 8640 }, { "grad_norm": 0.3506540060043335, "learning_rate": 9.963640892419958e-05, "loss": 0.0418, "step": 8650 }, { "grad_norm": 0.40888914465904236, "learning_rate": 9.963441580307286e-05, "loss": 0.043, "step": 8660 }, { "grad_norm": 0.39581069350242615, "learning_rate": 9.963241725400104e-05, "loss": 0.0441, "step": 8670 }, { "grad_norm": 0.4175892770290375, "learning_rate": 9.963041327720271e-05, "loss": 0.0456, "step": 8680 }, { "grad_norm": 0.29358506202697754, "learning_rate": 9.962840387289697e-05, "loss": 0.0437, "step": 8690 }, { "grad_norm": 0.45270875096321106, "learning_rate": 9.962638904130363e-05, "loss": 0.0515, "step": 8700 }, { "grad_norm": 0.324324369430542, "learning_rate": 9.962436878264298e-05, "loss": 0.0413, "step": 8710 }, { "grad_norm": 0.3553517460823059, "learning_rate": 9.962234309713598e-05, "loss": 0.0417, "step": 8720 }, { "grad_norm": 0.40642720460891724, "learning_rate": 9.962031198500414e-05, "loss": 0.0435, "step": 8730 }, { "grad_norm": 0.4186655580997467, "learning_rate": 9.961827544646958e-05, "loss": 0.0542, "step": 8740 }, { "grad_norm": 0.3363558351993561, "learning_rate": 9.961623348175501e-05, "loss": 0.0474, "step": 8750 }, { "grad_norm": 0.38522812724113464, "learning_rate": 9.961418609108377e-05, "loss": 0.0523, "step": 8760 }, { "grad_norm": 0.3852006196975708, "learning_rate": 9.961213327467971e-05, "loss": 0.0589, "step": 8770 }, { "grad_norm": 0.38542383909225464, "learning_rate": 9.961007503276736e-05, "loss": 0.0432, "step": 8780 }, { "grad_norm": 0.3933038115501404, "learning_rate": 9.960801136557179e-05, "loss": 0.0447, "step": 8790 }, { "grad_norm": 0.4073471426963806, "learning_rate": 9.960594227331866e-05, "loss": 0.0486, "step": 8800 }, { "grad_norm": 0.33812302350997925, "learning_rate": 9.960386775623429e-05, "loss": 0.0446, "step": 8810 }, { "grad_norm": 0.35340335965156555, "learning_rate": 9.96017878145455e-05, "loss": 0.0464, "step": 8820 }, { "grad_norm": 0.3580501675605774, "learning_rate": 9.959970244847977e-05, "loss": 0.0422, "step": 8830 }, { "grad_norm": 0.3120120167732239, "learning_rate": 9.959761165826518e-05, "loss": 0.0508, "step": 8840 }, { "grad_norm": 0.32864290475845337, "learning_rate": 9.959551544413033e-05, "loss": 0.0453, "step": 8850 }, { "grad_norm": 0.36085179448127747, "learning_rate": 9.959341380630448e-05, "loss": 0.039, "step": 8860 }, { "grad_norm": 0.3134787976741791, "learning_rate": 9.959130674501746e-05, "loss": 0.0426, "step": 8870 }, { "grad_norm": 0.39763814210891724, "learning_rate": 9.958919426049968e-05, "loss": 0.0468, "step": 8880 }, { "grad_norm": 0.33929961919784546, "learning_rate": 9.958707635298219e-05, "loss": 0.0435, "step": 8890 }, { "grad_norm": 0.43532365560531616, "learning_rate": 9.958495302269657e-05, "loss": 0.0485, "step": 8900 }, { "grad_norm": 0.36310142278671265, "learning_rate": 9.958282426987503e-05, "loss": 0.0438, "step": 8910 }, { "grad_norm": 0.33328741788864136, "learning_rate": 9.95806900947504e-05, "loss": 0.0474, "step": 8920 }, { "grad_norm": 0.25346270203590393, "learning_rate": 9.957855049755604e-05, "loss": 0.0445, "step": 8930 }, { "grad_norm": 0.3863539695739746, "learning_rate": 9.957640547852593e-05, "loss": 0.0391, "step": 8940 }, { "grad_norm": 0.4084285795688629, "learning_rate": 9.957425503789466e-05, "loss": 0.0461, "step": 8950 }, { "grad_norm": 0.3463532626628876, "learning_rate": 9.957209917589738e-05, "loss": 0.0393, "step": 8960 }, { "grad_norm": 0.3396458029747009, "learning_rate": 9.956993789276987e-05, "loss": 0.0439, "step": 8970 }, { "grad_norm": 0.4344041347503662, "learning_rate": 9.956777118874847e-05, "loss": 0.0532, "step": 8980 }, { "grad_norm": 0.3363199830055237, "learning_rate": 9.956559906407016e-05, "loss": 0.0483, "step": 8990 }, { "grad_norm": 0.2663528621196747, "learning_rate": 9.956342151897245e-05, "loss": 0.0391, "step": 9000 }, { "grad_norm": 0.24903780221939087, "learning_rate": 9.956123855369346e-05, "loss": 0.0353, "step": 9010 }, { "grad_norm": 0.3210415244102478, "learning_rate": 9.955905016847196e-05, "loss": 0.0461, "step": 9020 }, { "grad_norm": 0.29778915643692017, "learning_rate": 9.955685636354723e-05, "loss": 0.042, "step": 9030 }, { "grad_norm": 0.2771719992160797, "learning_rate": 9.95546571391592e-05, "loss": 0.0375, "step": 9040 }, { "grad_norm": 0.2951692044734955, "learning_rate": 9.955245249554837e-05, "loss": 0.0422, "step": 9050 }, { "grad_norm": 0.460600346326828, "learning_rate": 9.955024243295582e-05, "loss": 0.044, "step": 9060 }, { "grad_norm": 0.32214394211769104, "learning_rate": 9.954802695162328e-05, "loss": 0.0447, "step": 9070 }, { "grad_norm": 0.2572252154350281, "learning_rate": 9.954580605179302e-05, "loss": 0.0392, "step": 9080 }, { "grad_norm": 0.35640859603881836, "learning_rate": 9.954357973370788e-05, "loss": 0.0404, "step": 9090 }, { "grad_norm": 0.3587583899497986, "learning_rate": 9.954134799761135e-05, "loss": 0.0387, "step": 9100 }, { "grad_norm": 0.23519843816757202, "learning_rate": 9.953911084374748e-05, "loss": 0.0403, "step": 9110 }, { "grad_norm": 0.42086195945739746, "learning_rate": 9.953686827236093e-05, "loss": 0.0447, "step": 9120 }, { "grad_norm": 0.29912787675857544, "learning_rate": 9.953462028369695e-05, "loss": 0.0448, "step": 9130 }, { "grad_norm": 0.31225186586380005, "learning_rate": 9.953236687800136e-05, "loss": 0.0459, "step": 9140 }, { "grad_norm": 0.3248921036720276, "learning_rate": 9.95301080555206e-05, "loss": 0.0503, "step": 9150 }, { "grad_norm": 0.35685625672340393, "learning_rate": 9.952784381650171e-05, "loss": 0.0513, "step": 9160 }, { "grad_norm": 0.4581593871116638, "learning_rate": 9.952557416119226e-05, "loss": 0.0543, "step": 9170 }, { "grad_norm": 0.3208722472190857, "learning_rate": 9.95232990898405e-05, "loss": 0.0528, "step": 9180 }, { "grad_norm": 0.341362327337265, "learning_rate": 9.95210186026952e-05, "loss": 0.0492, "step": 9190 }, { "grad_norm": 0.3634656071662903, "learning_rate": 9.951873270000576e-05, "loss": 0.0456, "step": 9200 }, { "grad_norm": 0.3081986904144287, "learning_rate": 9.951644138202216e-05, "loss": 0.0502, "step": 9210 }, { "grad_norm": 0.44449466466903687, "learning_rate": 9.951414464899498e-05, "loss": 0.0541, "step": 9220 }, { "grad_norm": 0.3958427608013153, "learning_rate": 9.951184250117538e-05, "loss": 0.0499, "step": 9230 }, { "grad_norm": 0.32618752121925354, "learning_rate": 9.950953493881513e-05, "loss": 0.0547, "step": 9240 }, { "grad_norm": 0.33067166805267334, "learning_rate": 9.950722196216658e-05, "loss": 0.0446, "step": 9250 }, { "grad_norm": 0.2583971917629242, "learning_rate": 9.950490357148265e-05, "loss": 0.0459, "step": 9260 }, { "grad_norm": 0.4045358896255493, "learning_rate": 9.950257976701692e-05, "loss": 0.0507, "step": 9270 }, { "grad_norm": 0.451690137386322, "learning_rate": 9.950025054902348e-05, "loss": 0.0469, "step": 9280 }, { "grad_norm": 0.4254789352416992, "learning_rate": 9.949791591775706e-05, "loss": 0.0423, "step": 9290 }, { "grad_norm": 0.3314303159713745, "learning_rate": 9.949557587347298e-05, "loss": 0.0493, "step": 9300 }, { "grad_norm": 0.35585731267929077, "learning_rate": 9.949323041642713e-05, "loss": 0.0512, "step": 9310 }, { "grad_norm": 0.33633917570114136, "learning_rate": 9.949087954687602e-05, "loss": 0.0526, "step": 9320 }, { "grad_norm": 0.3284129202365875, "learning_rate": 9.948852326507672e-05, "loss": 0.0428, "step": 9330 }, { "grad_norm": 0.34552857279777527, "learning_rate": 9.948616157128694e-05, "loss": 0.0522, "step": 9340 }, { "grad_norm": 0.33601102232933044, "learning_rate": 9.948379446576493e-05, "loss": 0.0444, "step": 9350 }, { "grad_norm": 0.2768407166004181, "learning_rate": 9.948142194876952e-05, "loss": 0.0409, "step": 9360 }, { "grad_norm": 0.3854173421859741, "learning_rate": 9.947904402056024e-05, "loss": 0.0435, "step": 9370 }, { "grad_norm": 0.3321753740310669, "learning_rate": 9.947666068139708e-05, "loss": 0.04, "step": 9380 }, { "grad_norm": 0.301902174949646, "learning_rate": 9.947427193154071e-05, "loss": 0.0497, "step": 9390 }, { "grad_norm": 0.28820735216140747, "learning_rate": 9.947187777125233e-05, "loss": 0.0541, "step": 9400 }, { "grad_norm": 0.3476165533065796, "learning_rate": 9.946947820079377e-05, "loss": 0.0433, "step": 9410 }, { "grad_norm": 0.394025981426239, "learning_rate": 9.946707322042747e-05, "loss": 0.0476, "step": 9420 }, { "grad_norm": 0.2817972004413605, "learning_rate": 9.94646628304164e-05, "loss": 0.0457, "step": 9430 }, { "grad_norm": 0.30414488911628723, "learning_rate": 9.946224703102418e-05, "loss": 0.0426, "step": 9440 }, { "grad_norm": 0.3295677602291107, "learning_rate": 9.945982582251498e-05, "loss": 0.0437, "step": 9450 }, { "grad_norm": 0.3774215579032898, "learning_rate": 9.94573992051536e-05, "loss": 0.0425, "step": 9460 }, { "grad_norm": 0.3436443507671356, "learning_rate": 9.94549671792054e-05, "loss": 0.0433, "step": 9470 }, { "grad_norm": 0.4266938269138336, "learning_rate": 9.945252974493635e-05, "loss": 0.0477, "step": 9480 }, { "grad_norm": 0.33795368671417236, "learning_rate": 9.9450086902613e-05, "loss": 0.0454, "step": 9490 }, { "grad_norm": 0.35957568883895874, "learning_rate": 9.944763865250248e-05, "loss": 0.0463, "step": 9500 }, { "grad_norm": 0.3388185501098633, "learning_rate": 9.944518499487254e-05, "loss": 0.0431, "step": 9510 }, { "grad_norm": 0.3638243079185486, "learning_rate": 9.944272592999151e-05, "loss": 0.0506, "step": 9520 }, { "grad_norm": 0.3359801471233368, "learning_rate": 9.94402614581283e-05, "loss": 0.0472, "step": 9530 }, { "grad_norm": 0.31439515948295593, "learning_rate": 9.943779157955244e-05, "loss": 0.0461, "step": 9540 }, { "grad_norm": 0.31682834029197693, "learning_rate": 9.943531629453403e-05, "loss": 0.0453, "step": 9550 }, { "grad_norm": 0.28352123498916626, "learning_rate": 9.943283560334375e-05, "loss": 0.0407, "step": 9560 }, { "grad_norm": 0.2810952365398407, "learning_rate": 9.943034950625288e-05, "loss": 0.0419, "step": 9570 }, { "grad_norm": 0.3190231919288635, "learning_rate": 9.942785800353332e-05, "loss": 0.0455, "step": 9580 }, { "grad_norm": 0.24626639485359192, "learning_rate": 9.942536109545751e-05, "loss": 0.0463, "step": 9590 }, { "grad_norm": 0.3959740996360779, "learning_rate": 9.942285878229853e-05, "loss": 0.0457, "step": 9600 }, { "grad_norm": 0.2800191342830658, "learning_rate": 9.942035106433001e-05, "loss": 0.0371, "step": 9610 }, { "grad_norm": 0.3045077919960022, "learning_rate": 9.94178379418262e-05, "loss": 0.0426, "step": 9620 }, { "grad_norm": 0.35885918140411377, "learning_rate": 9.941531941506194e-05, "loss": 0.0445, "step": 9630 }, { "grad_norm": 0.27992063760757446, "learning_rate": 9.941279548431263e-05, "loss": 0.0398, "step": 9640 }, { "grad_norm": 0.2927576005458832, "learning_rate": 9.941026614985431e-05, "loss": 0.0466, "step": 9650 }, { "grad_norm": 0.26415857672691345, "learning_rate": 9.940773141196357e-05, "loss": 0.0411, "step": 9660 }, { "grad_norm": 0.4065438210964203, "learning_rate": 9.94051912709176e-05, "loss": 0.0413, "step": 9670 }, { "grad_norm": 0.3406978249549866, "learning_rate": 9.940264572699421e-05, "loss": 0.0456, "step": 9680 }, { "grad_norm": 0.3433470129966736, "learning_rate": 9.940009478047174e-05, "loss": 0.0456, "step": 9690 }, { "grad_norm": 0.40591999888420105, "learning_rate": 9.939753843162918e-05, "loss": 0.0427, "step": 9700 }, { "grad_norm": 0.28991150856018066, "learning_rate": 9.939497668074609e-05, "loss": 0.0398, "step": 9710 }, { "grad_norm": 0.26016002893447876, "learning_rate": 9.93924095281026e-05, "loss": 0.0469, "step": 9720 }, { "grad_norm": 0.31723204255104065, "learning_rate": 9.938983697397948e-05, "loss": 0.0476, "step": 9730 }, { "grad_norm": 0.3654789626598358, "learning_rate": 9.938725901865805e-05, "loss": 0.0452, "step": 9740 }, { "grad_norm": 0.345502644777298, "learning_rate": 9.93846756624202e-05, "loss": 0.0414, "step": 9750 }, { "grad_norm": 0.33298152685165405, "learning_rate": 9.938208690554849e-05, "loss": 0.0448, "step": 9760 }, { "grad_norm": 0.3901350200176239, "learning_rate": 9.9379492748326e-05, "loss": 0.0461, "step": 9770 }, { "grad_norm": 0.3039166033267975, "learning_rate": 9.937689319103641e-05, "loss": 0.0422, "step": 9780 }, { "grad_norm": 0.3351316750049591, "learning_rate": 9.937428823396404e-05, "loss": 0.0451, "step": 9790 }, { "grad_norm": 0.30264967679977417, "learning_rate": 9.937167787739372e-05, "loss": 0.0432, "step": 9800 }, { "grad_norm": 0.28139162063598633, "learning_rate": 9.936906212161095e-05, "loss": 0.0389, "step": 9810 }, { "grad_norm": 0.4950602948665619, "learning_rate": 9.936644096690176e-05, "loss": 0.0459, "step": 9820 }, { "grad_norm": 0.37116026878356934, "learning_rate": 9.936381441355282e-05, "loss": 0.0465, "step": 9830 }, { "grad_norm": 0.35893169045448303, "learning_rate": 9.936118246185136e-05, "loss": 0.0463, "step": 9840 }, { "grad_norm": 0.4369598925113678, "learning_rate": 9.935854511208518e-05, "loss": 0.0538, "step": 9850 }, { "grad_norm": 0.34265321493148804, "learning_rate": 9.935590236454272e-05, "loss": 0.0545, "step": 9860 }, { "grad_norm": 0.2758217751979828, "learning_rate": 9.935325421951298e-05, "loss": 0.0435, "step": 9870 }, { "grad_norm": 0.27397292852401733, "learning_rate": 9.935060067728557e-05, "loss": 0.0485, "step": 9880 }, { "grad_norm": 0.3757546842098236, "learning_rate": 9.934794173815067e-05, "loss": 0.0529, "step": 9890 }, { "grad_norm": 0.3037133812904358, "learning_rate": 9.934527740239906e-05, "loss": 0.0454, "step": 9900 }, { "grad_norm": 0.4031955301761627, "learning_rate": 9.934260767032209e-05, "loss": 0.0531, "step": 9910 }, { "grad_norm": 0.33055397868156433, "learning_rate": 9.933993254221172e-05, "loss": 0.0451, "step": 9920 }, { "grad_norm": 0.39158737659454346, "learning_rate": 9.933725201836053e-05, "loss": 0.0442, "step": 9930 }, { "grad_norm": 0.292324036359787, "learning_rate": 9.933456609906162e-05, "loss": 0.0396, "step": 9940 }, { "grad_norm": 0.37294435501098633, "learning_rate": 9.933187478460875e-05, "loss": 0.0548, "step": 9950 }, { "grad_norm": 0.4996308386325836, "learning_rate": 9.93291780752962e-05, "loss": 0.0473, "step": 9960 }, { "grad_norm": 0.3327767252922058, "learning_rate": 9.932647597141893e-05, "loss": 0.0506, "step": 9970 }, { "grad_norm": 0.46021005511283875, "learning_rate": 9.932376847327239e-05, "loss": 0.0554, "step": 9980 }, { "grad_norm": 0.35016465187072754, "learning_rate": 9.932105558115268e-05, "loss": 0.0467, "step": 9990 }, { "grad_norm": 0.36344021558761597, "learning_rate": 9.931833729535651e-05, "loss": 0.0478, "step": 10000 }, { "grad_norm": 0.24859300255775452, "learning_rate": 9.931561361618111e-05, "loss": 0.0416, "step": 10010 }, { "grad_norm": 0.3184868395328522, "learning_rate": 9.931288454392435e-05, "loss": 0.0494, "step": 10020 }, { "grad_norm": 0.3448464274406433, "learning_rate": 9.931015007888467e-05, "loss": 0.0458, "step": 10030 }, { "grad_norm": 0.37086889147758484, "learning_rate": 9.930741022136112e-05, "loss": 0.053, "step": 10040 }, { "grad_norm": 0.3342663049697876, "learning_rate": 9.930466497165333e-05, "loss": 0.0428, "step": 10050 }, { "grad_norm": 0.2962871789932251, "learning_rate": 9.93019143300615e-05, "loss": 0.0413, "step": 10060 }, { "grad_norm": 0.336031049489975, "learning_rate": 9.929915829688644e-05, "loss": 0.0404, "step": 10070 }, { "grad_norm": 0.3762799799442291, "learning_rate": 9.929639687242955e-05, "loss": 0.0464, "step": 10080 }, { "grad_norm": 0.3699324131011963, "learning_rate": 9.929363005699281e-05, "loss": 0.0462, "step": 10090 }, { "grad_norm": 0.4274086058139801, "learning_rate": 9.92908578508788e-05, "loss": 0.057, "step": 10100 }, { "grad_norm": 0.36405542492866516, "learning_rate": 9.928808025439069e-05, "loss": 0.0555, "step": 10110 }, { "grad_norm": 0.3813841640949249, "learning_rate": 9.928529726783223e-05, "loss": 0.0478, "step": 10120 }, { "grad_norm": 0.3978312611579895, "learning_rate": 9.928250889150774e-05, "loss": 0.0507, "step": 10130 }, { "grad_norm": 0.40949782729148865, "learning_rate": 9.92797151257222e-05, "loss": 0.0445, "step": 10140 }, { "grad_norm": 0.30184242129325867, "learning_rate": 9.927691597078108e-05, "loss": 0.0464, "step": 10150 }, { "grad_norm": 0.3587197959423065, "learning_rate": 9.927411142699053e-05, "loss": 0.043, "step": 10160 }, { "grad_norm": 0.32560721039772034, "learning_rate": 9.927130149465725e-05, "loss": 0.0421, "step": 10170 }, { "grad_norm": 0.38016006350517273, "learning_rate": 9.92684861740885e-05, "loss": 0.0437, "step": 10180 }, { "grad_norm": 0.35360363125801086, "learning_rate": 9.926566546559217e-05, "loss": 0.0435, "step": 10190 }, { "grad_norm": 0.41700801253318787, "learning_rate": 9.926283936947673e-05, "loss": 0.0473, "step": 10200 }, { "grad_norm": 0.3321857452392578, "learning_rate": 9.926000788605126e-05, "loss": 0.0475, "step": 10210 }, { "grad_norm": 0.5331854224205017, "learning_rate": 9.92571710156254e-05, "loss": 0.047, "step": 10220 }, { "grad_norm": 0.31534528732299805, "learning_rate": 9.925432875850936e-05, "loss": 0.0395, "step": 10230 }, { "grad_norm": 0.3046083152294159, "learning_rate": 9.925148111501396e-05, "loss": 0.0403, "step": 10240 }, { "grad_norm": 0.3628411293029785, "learning_rate": 9.924862808545066e-05, "loss": 0.0411, "step": 10250 }, { "grad_norm": 0.34374502301216125, "learning_rate": 9.924576967013141e-05, "loss": 0.0441, "step": 10260 }, { "grad_norm": 0.29366335272789, "learning_rate": 9.924290586936887e-05, "loss": 0.0433, "step": 10270 }, { "grad_norm": 0.28533831238746643, "learning_rate": 9.924003668347614e-05, "loss": 0.0426, "step": 10280 }, { "grad_norm": 0.4267674386501312, "learning_rate": 9.923716211276704e-05, "loss": 0.0453, "step": 10290 }, { "grad_norm": 0.3385354280471802, "learning_rate": 9.923428215755594e-05, "loss": 0.0503, "step": 10300 }, { "grad_norm": 0.3021430969238281, "learning_rate": 9.923139681815775e-05, "loss": 0.0469, "step": 10310 }, { "grad_norm": 0.36965978145599365, "learning_rate": 9.922850609488801e-05, "loss": 0.0446, "step": 10320 }, { "grad_norm": 0.42819932103157043, "learning_rate": 9.922560998806287e-05, "loss": 0.0448, "step": 10330 }, { "grad_norm": 0.3312870264053345, "learning_rate": 9.922270849799905e-05, "loss": 0.041, "step": 10340 }, { "grad_norm": 0.33989670872688293, "learning_rate": 9.92198016250138e-05, "loss": 0.0408, "step": 10350 }, { "grad_norm": 0.31195172667503357, "learning_rate": 9.921688936942506e-05, "loss": 0.0344, "step": 10360 }, { "grad_norm": 0.3219633102416992, "learning_rate": 9.921397173155129e-05, "loss": 0.048, "step": 10370 }, { "grad_norm": 0.42650261521339417, "learning_rate": 9.921104871171157e-05, "loss": 0.0471, "step": 10380 }, { "grad_norm": 0.32018062472343445, "learning_rate": 9.920812031022554e-05, "loss": 0.042, "step": 10390 }, { "grad_norm": 0.4623417854309082, "learning_rate": 9.920518652741348e-05, "loss": 0.042, "step": 10400 }, { "grad_norm": 0.33381709456443787, "learning_rate": 9.920224736359618e-05, "loss": 0.0411, "step": 10410 }, { "grad_norm": 0.38686758279800415, "learning_rate": 9.91993028190951e-05, "loss": 0.0509, "step": 10420 }, { "grad_norm": 0.4788787066936493, "learning_rate": 9.919635289423222e-05, "loss": 0.0446, "step": 10430 }, { "grad_norm": 0.2812008261680603, "learning_rate": 9.919339758933015e-05, "loss": 0.0499, "step": 10440 }, { "grad_norm": 0.30717021226882935, "learning_rate": 9.919043690471209e-05, "loss": 0.0411, "step": 10450 }, { "grad_norm": 0.3375486433506012, "learning_rate": 9.91874708407018e-05, "loss": 0.0469, "step": 10460 }, { "grad_norm": 0.39275622367858887, "learning_rate": 9.918449939762367e-05, "loss": 0.0374, "step": 10470 }, { "grad_norm": 0.2831420302391052, "learning_rate": 9.91815225758026e-05, "loss": 0.0368, "step": 10480 }, { "grad_norm": 0.33020302653312683, "learning_rate": 9.917854037556419e-05, "loss": 0.0434, "step": 10490 }, { "grad_norm": 0.2800186276435852, "learning_rate": 9.917555279723454e-05, "loss": 0.0482, "step": 10500 }, { "grad_norm": 0.3985837996006012, "learning_rate": 9.917255984114036e-05, "loss": 0.0471, "step": 10510 }, { "grad_norm": 0.5274776220321655, "learning_rate": 9.916956150760896e-05, "loss": 0.0404, "step": 10520 }, { "grad_norm": 0.3011280596256256, "learning_rate": 9.916655779696826e-05, "loss": 0.0453, "step": 10530 }, { "grad_norm": 0.3225436210632324, "learning_rate": 9.916354870954671e-05, "loss": 0.0448, "step": 10540 }, { "grad_norm": 0.33337417244911194, "learning_rate": 9.91605342456734e-05, "loss": 0.0482, "step": 10550 }, { "grad_norm": 0.34147772192955017, "learning_rate": 9.915751440567795e-05, "loss": 0.0397, "step": 10560 }, { "grad_norm": 0.30635735392570496, "learning_rate": 9.915448918989066e-05, "loss": 0.0449, "step": 10570 }, { "grad_norm": 0.38050973415374756, "learning_rate": 9.915145859864232e-05, "loss": 0.0507, "step": 10580 }, { "grad_norm": 0.3902590274810791, "learning_rate": 9.914842263226437e-05, "loss": 0.0492, "step": 10590 }, { "grad_norm": 0.3009953498840332, "learning_rate": 9.914538129108882e-05, "loss": 0.0481, "step": 10600 }, { "grad_norm": 0.33610832691192627, "learning_rate": 9.914233457544825e-05, "loss": 0.0496, "step": 10610 }, { "grad_norm": 0.32614395022392273, "learning_rate": 9.913928248567586e-05, "loss": 0.0486, "step": 10620 }, { "grad_norm": 0.3680899739265442, "learning_rate": 9.913622502210542e-05, "loss": 0.0483, "step": 10630 }, { "grad_norm": 0.4659446179866791, "learning_rate": 9.913316218507128e-05, "loss": 0.0479, "step": 10640 }, { "grad_norm": 0.3610993027687073, "learning_rate": 9.91300939749084e-05, "loss": 0.039, "step": 10650 }, { "grad_norm": 0.5849641561508179, "learning_rate": 9.91270203919523e-05, "loss": 0.0451, "step": 10660 }, { "grad_norm": 0.3997565805912018, "learning_rate": 9.912394143653912e-05, "loss": 0.0507, "step": 10670 }, { "grad_norm": 0.3227463960647583, "learning_rate": 9.912085710900555e-05, "loss": 0.0469, "step": 10680 }, { "grad_norm": 0.3463442921638489, "learning_rate": 9.911776740968892e-05, "loss": 0.0494, "step": 10690 }, { "grad_norm": 0.2627561092376709, "learning_rate": 9.911467233892709e-05, "loss": 0.0455, "step": 10700 }, { "grad_norm": 0.3110652565956116, "learning_rate": 9.911157189705853e-05, "loss": 0.0454, "step": 10710 }, { "grad_norm": 0.2524372637271881, "learning_rate": 9.910846608442229e-05, "loss": 0.0422, "step": 10720 }, { "grad_norm": 0.41449329257011414, "learning_rate": 9.910535490135805e-05, "loss": 0.0419, "step": 10730 }, { "grad_norm": 0.36881765723228455, "learning_rate": 9.910223834820603e-05, "loss": 0.0469, "step": 10740 }, { "grad_norm": 0.30542513728141785, "learning_rate": 9.909911642530703e-05, "loss": 0.0472, "step": 10750 }, { "grad_norm": 0.3256625831127167, "learning_rate": 9.909598913300249e-05, "loss": 0.0415, "step": 10760 }, { "grad_norm": 0.36690688133239746, "learning_rate": 9.909285647163438e-05, "loss": 0.0436, "step": 10770 }, { "grad_norm": 0.25294095277786255, "learning_rate": 9.908971844154531e-05, "loss": 0.0403, "step": 10780 }, { "grad_norm": 0.381345272064209, "learning_rate": 9.908657504307843e-05, "loss": 0.046, "step": 10790 }, { "grad_norm": 0.39558130502700806, "learning_rate": 9.908342627657751e-05, "loss": 0.0444, "step": 10800 }, { "grad_norm": 0.36712828278541565, "learning_rate": 9.908027214238689e-05, "loss": 0.0449, "step": 10810 }, { "grad_norm": 0.22823351621627808, "learning_rate": 9.90771126408515e-05, "loss": 0.0443, "step": 10820 }, { "grad_norm": 0.30898377299308777, "learning_rate": 9.907394777231685e-05, "loss": 0.0454, "step": 10830 }, { "grad_norm": 0.2721451222896576, "learning_rate": 9.907077753712905e-05, "loss": 0.0388, "step": 10840 }, { "grad_norm": 0.33023539185523987, "learning_rate": 9.906760193563482e-05, "loss": 0.0451, "step": 10850 }, { "grad_norm": 0.37579891085624695, "learning_rate": 9.906442096818139e-05, "loss": 0.0443, "step": 10860 }, { "grad_norm": 0.2726559638977051, "learning_rate": 9.906123463511665e-05, "loss": 0.0433, "step": 10870 }, { "grad_norm": 0.2549954354763031, "learning_rate": 9.905804293678907e-05, "loss": 0.039, "step": 10880 }, { "grad_norm": 0.325654000043869, "learning_rate": 9.905484587354766e-05, "loss": 0.0429, "step": 10890 }, { "grad_norm": 0.35401153564453125, "learning_rate": 9.905164344574205e-05, "loss": 0.0427, "step": 10900 }, { "grad_norm": 0.2860376834869385, "learning_rate": 9.904843565372248e-05, "loss": 0.0405, "step": 10910 }, { "grad_norm": 0.3491675853729248, "learning_rate": 9.904522249783972e-05, "loss": 0.0419, "step": 10920 }, { "grad_norm": 0.27252066135406494, "learning_rate": 9.904200397844517e-05, "loss": 0.0425, "step": 10930 }, { "grad_norm": 0.29211533069610596, "learning_rate": 9.903878009589078e-05, "loss": 0.0409, "step": 10940 }, { "grad_norm": 0.3451719284057617, "learning_rate": 9.903555085052915e-05, "loss": 0.0415, "step": 10950 }, { "grad_norm": 0.3939625322818756, "learning_rate": 9.903231624271338e-05, "loss": 0.0446, "step": 10960 }, { "grad_norm": 0.26346537470817566, "learning_rate": 9.902907627279724e-05, "loss": 0.0398, "step": 10970 }, { "grad_norm": 0.2721627354621887, "learning_rate": 9.902583094113504e-05, "loss": 0.0439, "step": 10980 }, { "grad_norm": 0.3069576919078827, "learning_rate": 9.902258024808168e-05, "loss": 0.039, "step": 10990 }, { "grad_norm": 0.29497674107551575, "learning_rate": 9.901932419399264e-05, "loss": 0.0376, "step": 11000 }, { "grad_norm": 0.3185654282569885, "learning_rate": 9.9016062779224e-05, "loss": 0.0403, "step": 11010 }, { "grad_norm": 0.3774182200431824, "learning_rate": 9.901279600413242e-05, "loss": 0.0443, "step": 11020 }, { "grad_norm": 0.3034294545650482, "learning_rate": 9.900952386907518e-05, "loss": 0.0483, "step": 11030 }, { "grad_norm": 0.32387179136276245, "learning_rate": 9.90062463744101e-05, "loss": 0.0436, "step": 11040 }, { "grad_norm": 0.24872350692749023, "learning_rate": 9.900296352049558e-05, "loss": 0.0433, "step": 11050 }, { "grad_norm": 0.33692285418510437, "learning_rate": 9.899967530769065e-05, "loss": 0.0435, "step": 11060 }, { "grad_norm": 0.30952510237693787, "learning_rate": 9.899638173635489e-05, "loss": 0.0393, "step": 11070 }, { "grad_norm": 0.30484604835510254, "learning_rate": 9.899308280684849e-05, "loss": 0.0467, "step": 11080 }, { "grad_norm": 0.33107760548591614, "learning_rate": 9.898977851953222e-05, "loss": 0.0442, "step": 11090 }, { "grad_norm": 0.2653965353965759, "learning_rate": 9.898646887476741e-05, "loss": 0.0386, "step": 11100 }, { "grad_norm": 0.23868173360824585, "learning_rate": 9.898315387291603e-05, "loss": 0.0405, "step": 11110 }, { "grad_norm": 0.2952445447444916, "learning_rate": 9.89798335143406e-05, "loss": 0.038, "step": 11120 }, { "grad_norm": 0.4213348925113678, "learning_rate": 9.897650779940419e-05, "loss": 0.0475, "step": 11130 }, { "grad_norm": 0.24832887947559357, "learning_rate": 9.897317672847054e-05, "loss": 0.0359, "step": 11140 }, { "grad_norm": 0.2631649374961853, "learning_rate": 9.89698403019039e-05, "loss": 0.0376, "step": 11150 }, { "grad_norm": 0.2896246016025543, "learning_rate": 9.896649852006917e-05, "loss": 0.0332, "step": 11160 }, { "grad_norm": 0.299116313457489, "learning_rate": 9.896315138333177e-05, "loss": 0.041, "step": 11170 }, { "grad_norm": 0.21951782703399658, "learning_rate": 9.895979889205774e-05, "loss": 0.0379, "step": 11180 }, { "grad_norm": 0.30716171860694885, "learning_rate": 9.895644104661372e-05, "loss": 0.0406, "step": 11190 }, { "grad_norm": 0.36420097947120667, "learning_rate": 9.895307784736691e-05, "loss": 0.034, "step": 11200 }, { "grad_norm": 0.33883583545684814, "learning_rate": 9.894970929468512e-05, "loss": 0.0401, "step": 11210 }, { "grad_norm": 0.2832435369491577, "learning_rate": 9.89463353889367e-05, "loss": 0.0363, "step": 11220 }, { "grad_norm": 0.5097227096557617, "learning_rate": 9.894295613049065e-05, "loss": 0.0393, "step": 11230 }, { "grad_norm": 0.38258272409439087, "learning_rate": 9.893957151971649e-05, "loss": 0.0382, "step": 11240 }, { "grad_norm": 0.30235159397125244, "learning_rate": 9.893618155698436e-05, "loss": 0.0366, "step": 11250 }, { "grad_norm": 0.291530579328537, "learning_rate": 9.8932786242665e-05, "loss": 0.0359, "step": 11260 }, { "grad_norm": 0.28082334995269775, "learning_rate": 9.89293855771297e-05, "loss": 0.0348, "step": 11270 }, { "grad_norm": 0.23567615449428558, "learning_rate": 9.892597956075036e-05, "loss": 0.037, "step": 11280 }, { "grad_norm": 0.3701847493648529, "learning_rate": 9.892256819389947e-05, "loss": 0.0387, "step": 11290 }, { "grad_norm": 0.30570510029792786, "learning_rate": 9.891915147695006e-05, "loss": 0.0393, "step": 11300 }, { "grad_norm": 0.40625447034835815, "learning_rate": 9.891572941027577e-05, "loss": 0.049, "step": 11310 }, { "grad_norm": 0.3070928156375885, "learning_rate": 9.89123019942509e-05, "loss": 0.0482, "step": 11320 }, { "grad_norm": 0.37601351737976074, "learning_rate": 9.89088692292502e-05, "loss": 0.0353, "step": 11330 }, { "grad_norm": 0.34135496616363525, "learning_rate": 9.89054311156491e-05, "loss": 0.0418, "step": 11340 }, { "grad_norm": 0.3029141426086426, "learning_rate": 9.890198765382357e-05, "loss": 0.0389, "step": 11350 }, { "grad_norm": 0.2839607894420624, "learning_rate": 9.889853884415021e-05, "loss": 0.0373, "step": 11360 }, { "grad_norm": 0.3418501019477844, "learning_rate": 9.889508468700614e-05, "loss": 0.0346, "step": 11370 }, { "grad_norm": 0.29151299595832825, "learning_rate": 9.889162518276915e-05, "loss": 0.0336, "step": 11380 }, { "grad_norm": 0.33813101053237915, "learning_rate": 9.888816033181752e-05, "loss": 0.0411, "step": 11390 }, { "grad_norm": 0.273104190826416, "learning_rate": 9.888469013453018e-05, "loss": 0.0395, "step": 11400 }, { "grad_norm": 0.2930062413215637, "learning_rate": 9.888121459128663e-05, "loss": 0.0359, "step": 11410 }, { "grad_norm": 0.33453255891799927, "learning_rate": 9.887773370246693e-05, "loss": 0.0389, "step": 11420 }, { "grad_norm": 0.3531169295310974, "learning_rate": 9.887424746845177e-05, "loss": 0.0447, "step": 11430 }, { "grad_norm": 0.28645089268684387, "learning_rate": 9.887075588962239e-05, "loss": 0.0406, "step": 11440 }, { "grad_norm": 0.39555346965789795, "learning_rate": 9.88672589663606e-05, "loss": 0.0391, "step": 11450 }, { "grad_norm": 0.3605438768863678, "learning_rate": 9.886375669904886e-05, "loss": 0.0476, "step": 11460 }, { "grad_norm": 0.2860826253890991, "learning_rate": 9.886024908807014e-05, "loss": 0.0484, "step": 11470 }, { "grad_norm": 0.3357045650482178, "learning_rate": 9.885673613380806e-05, "loss": 0.0446, "step": 11480 }, { "grad_norm": 0.39311495423316956, "learning_rate": 9.885321783664676e-05, "loss": 0.0443, "step": 11490 }, { "grad_norm": 0.30032217502593994, "learning_rate": 9.884969419697101e-05, "loss": 0.0407, "step": 11500 }, { "grad_norm": 0.27444255352020264, "learning_rate": 9.884616521516614e-05, "loss": 0.0424, "step": 11510 }, { "grad_norm": 0.3406539857387543, "learning_rate": 9.88426308916181e-05, "loss": 0.0346, "step": 11520 }, { "grad_norm": 0.3601529598236084, "learning_rate": 9.883909122671335e-05, "loss": 0.041, "step": 11530 }, { "grad_norm": 0.2949189841747284, "learning_rate": 9.883554622083904e-05, "loss": 0.0414, "step": 11540 }, { "grad_norm": 0.2768166661262512, "learning_rate": 9.88319958743828e-05, "loss": 0.0433, "step": 11550 }, { "grad_norm": 0.3266761004924774, "learning_rate": 9.882844018773291e-05, "loss": 0.0397, "step": 11560 }, { "grad_norm": 0.3443334102630615, "learning_rate": 9.882487916127823e-05, "loss": 0.0424, "step": 11570 }, { "grad_norm": 0.37295621633529663, "learning_rate": 9.882131279540815e-05, "loss": 0.0418, "step": 11580 }, { "grad_norm": 0.275917649269104, "learning_rate": 9.881774109051271e-05, "loss": 0.0373, "step": 11590 }, { "grad_norm": 0.24154943227767944, "learning_rate": 9.881416404698252e-05, "loss": 0.0426, "step": 11600 }, { "grad_norm": 0.36848869919776917, "learning_rate": 9.881058166520873e-05, "loss": 0.0355, "step": 11610 }, { "grad_norm": 0.39843177795410156, "learning_rate": 9.880699394558311e-05, "loss": 0.0425, "step": 11620 }, { "grad_norm": 0.3318515717983246, "learning_rate": 9.880340088849801e-05, "loss": 0.0445, "step": 11630 }, { "grad_norm": 0.3241727948188782, "learning_rate": 9.879980249434637e-05, "loss": 0.041, "step": 11640 }, { "grad_norm": 0.2857781648635864, "learning_rate": 9.879619876352168e-05, "loss": 0.0363, "step": 11650 }, { "grad_norm": 0.33428260684013367, "learning_rate": 9.879258969641809e-05, "loss": 0.0501, "step": 11660 }, { "grad_norm": 0.31921884417533875, "learning_rate": 9.878897529343023e-05, "loss": 0.0449, "step": 11670 }, { "grad_norm": 0.28233617544174194, "learning_rate": 9.878535555495338e-05, "loss": 0.0393, "step": 11680 }, { "grad_norm": 0.3514484763145447, "learning_rate": 9.87817304813834e-05, "loss": 0.0354, "step": 11690 }, { "grad_norm": 0.27810177206993103, "learning_rate": 9.877810007311671e-05, "loss": 0.0432, "step": 11700 }, { "grad_norm": 0.23549793660640717, "learning_rate": 9.877446433055035e-05, "loss": 0.0355, "step": 11710 }, { "grad_norm": 0.282041996717453, "learning_rate": 9.877082325408191e-05, "loss": 0.0446, "step": 11720 }, { "grad_norm": 0.3605155348777771, "learning_rate": 9.876717684410954e-05, "loss": 0.0377, "step": 11730 }, { "grad_norm": 0.3250211477279663, "learning_rate": 9.876352510103204e-05, "loss": 0.0453, "step": 11740 }, { "grad_norm": 0.2802806794643402, "learning_rate": 9.875986802524875e-05, "loss": 0.0377, "step": 11750 }, { "grad_norm": 0.4038347601890564, "learning_rate": 9.87562056171596e-05, "loss": 0.0445, "step": 11760 }, { "grad_norm": 0.3523809313774109, "learning_rate": 9.875253787716511e-05, "loss": 0.0386, "step": 11770 }, { "grad_norm": 0.27496159076690674, "learning_rate": 9.874886480566637e-05, "loss": 0.0371, "step": 11780 }, { "grad_norm": 0.3130829632282257, "learning_rate": 9.874518640306507e-05, "loss": 0.036, "step": 11790 }, { "grad_norm": 0.33539652824401855, "learning_rate": 9.874150266976347e-05, "loss": 0.049, "step": 11800 }, { "grad_norm": 0.3259206712245941, "learning_rate": 9.873781360616443e-05, "loss": 0.0465, "step": 11810 }, { "grad_norm": 0.30058059096336365, "learning_rate": 9.873411921267137e-05, "loss": 0.0435, "step": 11820 }, { "grad_norm": 0.270409494638443, "learning_rate": 9.873041948968829e-05, "loss": 0.0374, "step": 11830 }, { "grad_norm": 0.24771590530872345, "learning_rate": 9.872671443761981e-05, "loss": 0.0428, "step": 11840 }, { "grad_norm": 0.3428417146205902, "learning_rate": 9.872300405687109e-05, "loss": 0.0398, "step": 11850 }, { "grad_norm": 0.39156627655029297, "learning_rate": 9.871928834784792e-05, "loss": 0.0359, "step": 11860 }, { "grad_norm": 0.35724160075187683, "learning_rate": 9.871556731095661e-05, "loss": 0.0469, "step": 11870 }, { "grad_norm": 0.24775661528110504, "learning_rate": 9.871184094660411e-05, "loss": 0.0344, "step": 11880 }, { "grad_norm": 0.3193596601486206, "learning_rate": 9.870810925519791e-05, "loss": 0.0349, "step": 11890 }, { "grad_norm": 0.3267079293727875, "learning_rate": 9.870437223714612e-05, "loss": 0.0378, "step": 11900 }, { "grad_norm": 0.31687691807746887, "learning_rate": 9.87006298928574e-05, "loss": 0.0358, "step": 11910 }, { "grad_norm": 0.3065298795700073, "learning_rate": 9.869688222274103e-05, "loss": 0.0376, "step": 11920 }, { "grad_norm": 0.3170585334300995, "learning_rate": 9.869312922720681e-05, "loss": 0.039, "step": 11930 }, { "grad_norm": 0.4476315677165985, "learning_rate": 9.868937090666521e-05, "loss": 0.0432, "step": 11940 }, { "grad_norm": 0.3508242666721344, "learning_rate": 9.86856072615272e-05, "loss": 0.0419, "step": 11950 }, { "grad_norm": 0.2929801046848297, "learning_rate": 9.868183829220438e-05, "loss": 0.0354, "step": 11960 }, { "grad_norm": 0.2968622148036957, "learning_rate": 9.867806399910893e-05, "loss": 0.0369, "step": 11970 }, { "grad_norm": 0.2870660424232483, "learning_rate": 9.867428438265356e-05, "loss": 0.0422, "step": 11980 }, { "grad_norm": 0.3981765806674957, "learning_rate": 9.867049944325165e-05, "loss": 0.0397, "step": 11990 }, { "grad_norm": 0.3765154778957367, "learning_rate": 9.86667091813171e-05, "loss": 0.0421, "step": 12000 }, { "grad_norm": 0.26176607608795166, "learning_rate": 9.866291359726438e-05, "loss": 0.0362, "step": 12010 }, { "grad_norm": 0.31467172503471375, "learning_rate": 9.865911269150861e-05, "loss": 0.0386, "step": 12020 }, { "grad_norm": 0.2753302752971649, "learning_rate": 9.865530646446544e-05, "loss": 0.0391, "step": 12030 }, { "grad_norm": 0.37434467673301697, "learning_rate": 9.86514949165511e-05, "loss": 0.0348, "step": 12040 }, { "grad_norm": 0.2943544387817383, "learning_rate": 9.864767804818243e-05, "loss": 0.0321, "step": 12050 }, { "grad_norm": 0.2370724081993103, "learning_rate": 9.86438558597768e-05, "loss": 0.0368, "step": 12060 }, { "grad_norm": 0.250652551651001, "learning_rate": 9.864002835175225e-05, "loss": 0.0402, "step": 12070 }, { "grad_norm": 0.19835588335990906, "learning_rate": 9.863619552452734e-05, "loss": 0.0352, "step": 12080 }, { "grad_norm": 0.30014199018478394, "learning_rate": 9.863235737852119e-05, "loss": 0.0358, "step": 12090 }, { "grad_norm": 0.29690179228782654, "learning_rate": 9.862851391415356e-05, "loss": 0.0394, "step": 12100 }, { "grad_norm": 0.3365504741668701, "learning_rate": 9.862466513184477e-05, "loss": 0.0369, "step": 12110 }, { "grad_norm": 0.2485009729862213, "learning_rate": 9.86208110320157e-05, "loss": 0.0353, "step": 12120 }, { "grad_norm": 0.312532901763916, "learning_rate": 9.861695161508784e-05, "loss": 0.0385, "step": 12130 }, { "grad_norm": 0.37399107217788696, "learning_rate": 9.861308688148324e-05, "loss": 0.0413, "step": 12140 }, { "grad_norm": 0.3235595226287842, "learning_rate": 9.860921683162455e-05, "loss": 0.0477, "step": 12150 }, { "grad_norm": 0.40079742670059204, "learning_rate": 9.860534146593499e-05, "loss": 0.0407, "step": 12160 }, { "grad_norm": 0.26755762100219727, "learning_rate": 9.860146078483836e-05, "loss": 0.0323, "step": 12170 }, { "grad_norm": 0.23143692314624786, "learning_rate": 9.859757478875905e-05, "loss": 0.0357, "step": 12180 }, { "grad_norm": 0.3130905032157898, "learning_rate": 9.859368347812204e-05, "loss": 0.0381, "step": 12190 }, { "grad_norm": 0.3001139163970947, "learning_rate": 9.858978685335285e-05, "loss": 0.0445, "step": 12200 }, { "grad_norm": 0.40312671661376953, "learning_rate": 9.858588491487763e-05, "loss": 0.0418, "step": 12210 }, { "grad_norm": 0.25162145495414734, "learning_rate": 9.858197766312308e-05, "loss": 0.0362, "step": 12220 }, { "grad_norm": 0.2678479850292206, "learning_rate": 9.857806509851649e-05, "loss": 0.0448, "step": 12230 }, { "grad_norm": 0.2723527252674103, "learning_rate": 9.857414722148574e-05, "loss": 0.0352, "step": 12240 }, { "grad_norm": 0.2564423084259033, "learning_rate": 9.857022403245928e-05, "loss": 0.0353, "step": 12250 }, { "grad_norm": 0.44489234685897827, "learning_rate": 9.856629553186615e-05, "loss": 0.045, "step": 12260 }, { "grad_norm": 0.37573739886283875, "learning_rate": 9.856236172013595e-05, "loss": 0.0421, "step": 12270 }, { "grad_norm": 0.305762380361557, "learning_rate": 9.85584225976989e-05, "loss": 0.0414, "step": 12280 }, { "grad_norm": 0.32502710819244385, "learning_rate": 9.855447816498575e-05, "loss": 0.0447, "step": 12290 }, { "grad_norm": 0.39162978529930115, "learning_rate": 9.855052842242787e-05, "loss": 0.042, "step": 12300 }, { "grad_norm": 0.2845302224159241, "learning_rate": 9.85465733704572e-05, "loss": 0.0428, "step": 12310 }, { "grad_norm": 0.33502233028411865, "learning_rate": 9.854261300950624e-05, "loss": 0.0363, "step": 12320 }, { "grad_norm": 0.335854709148407, "learning_rate": 9.853864734000813e-05, "loss": 0.0436, "step": 12330 }, { "grad_norm": 0.3242794871330261, "learning_rate": 9.85346763623965e-05, "loss": 0.0355, "step": 12340 }, { "grad_norm": 0.26107627153396606, "learning_rate": 9.853070007710564e-05, "loss": 0.0398, "step": 12350 }, { "grad_norm": 0.30790066719055176, "learning_rate": 9.85267184845704e-05, "loss": 0.035, "step": 12360 }, { "grad_norm": 0.37130817770957947, "learning_rate": 9.852273158522616e-05, "loss": 0.0326, "step": 12370 }, { "grad_norm": 0.257451593875885, "learning_rate": 9.851873937950896e-05, "loss": 0.0337, "step": 12380 }, { "grad_norm": 0.35149985551834106, "learning_rate": 9.851474186785537e-05, "loss": 0.034, "step": 12390 }, { "grad_norm": 0.30886873602867126, "learning_rate": 9.851073905070254e-05, "loss": 0.0443, "step": 12400 }, { "grad_norm": 0.3352571129798889, "learning_rate": 9.850673092848824e-05, "loss": 0.04, "step": 12410 }, { "grad_norm": 0.22157521545886993, "learning_rate": 9.850271750165077e-05, "loss": 0.0379, "step": 12420 }, { "grad_norm": 0.2520051896572113, "learning_rate": 9.849869877062902e-05, "loss": 0.0393, "step": 12430 }, { "grad_norm": 0.2881471812725067, "learning_rate": 9.849467473586252e-05, "loss": 0.0418, "step": 12440 }, { "grad_norm": 0.28176820278167725, "learning_rate": 9.849064539779127e-05, "loss": 0.0363, "step": 12450 }, { "grad_norm": 0.36208418011665344, "learning_rate": 9.848661075685594e-05, "loss": 0.034, "step": 12460 }, { "grad_norm": 0.2994452118873596, "learning_rate": 9.848257081349778e-05, "loss": 0.0362, "step": 12470 }, { "grad_norm": 0.29282134771347046, "learning_rate": 9.847852556815856e-05, "loss": 0.0429, "step": 12480 }, { "grad_norm": 0.28885114192962646, "learning_rate": 9.847447502128067e-05, "loss": 0.0458, "step": 12490 }, { "grad_norm": 0.2539275586605072, "learning_rate": 9.847041917330708e-05, "loss": 0.0378, "step": 12500 }, { "grad_norm": 0.3396759331226349, "learning_rate": 9.846635802468132e-05, "loss": 0.0438, "step": 12510 }, { "grad_norm": 0.33753642439842224, "learning_rate": 9.84622915758475e-05, "loss": 0.0399, "step": 12520 }, { "grad_norm": 0.31997013092041016, "learning_rate": 9.845821982725034e-05, "loss": 0.0345, "step": 12530 }, { "grad_norm": 0.31236258149147034, "learning_rate": 9.845414277933514e-05, "loss": 0.0363, "step": 12540 }, { "grad_norm": 0.30606719851493835, "learning_rate": 9.845006043254771e-05, "loss": 0.0386, "step": 12550 }, { "grad_norm": 0.30326059460639954, "learning_rate": 9.844597278733451e-05, "loss": 0.0399, "step": 12560 }, { "grad_norm": 0.3414601981639862, "learning_rate": 9.844187984414259e-05, "loss": 0.0357, "step": 12570 }, { "grad_norm": 0.24585184454917908, "learning_rate": 9.84377816034195e-05, "loss": 0.0339, "step": 12580 }, { "grad_norm": 0.34284475445747375, "learning_rate": 9.843367806561345e-05, "loss": 0.0459, "step": 12590 }, { "grad_norm": 0.22346605360507965, "learning_rate": 9.842956923117317e-05, "loss": 0.0349, "step": 12600 }, { "grad_norm": 0.31821563839912415, "learning_rate": 9.842545510054802e-05, "loss": 0.0415, "step": 12610 }, { "grad_norm": 0.27117919921875, "learning_rate": 9.842133567418792e-05, "loss": 0.0397, "step": 12620 }, { "grad_norm": 0.3355554938316345, "learning_rate": 9.841721095254333e-05, "loss": 0.0409, "step": 12630 }, { "grad_norm": 0.29278460144996643, "learning_rate": 9.841308093606537e-05, "loss": 0.0362, "step": 12640 }, { "grad_norm": 0.4303464889526367, "learning_rate": 9.840894562520565e-05, "loss": 0.0421, "step": 12650 }, { "grad_norm": 0.28064239025115967, "learning_rate": 9.840480502041642e-05, "loss": 0.0366, "step": 12660 }, { "grad_norm": 0.25883379578590393, "learning_rate": 9.840065912215049e-05, "loss": 0.0435, "step": 12670 }, { "grad_norm": 0.2354414016008377, "learning_rate": 9.839650793086124e-05, "loss": 0.031, "step": 12680 }, { "grad_norm": 0.31784215569496155, "learning_rate": 9.839235144700265e-05, "loss": 0.0428, "step": 12690 }, { "grad_norm": 0.2722555100917816, "learning_rate": 9.838818967102926e-05, "loss": 0.0371, "step": 12700 }, { "grad_norm": 0.2251145839691162, "learning_rate": 9.83840226033962e-05, "loss": 0.0401, "step": 12710 }, { "grad_norm": 0.23562447726726532, "learning_rate": 9.837985024455918e-05, "loss": 0.041, "step": 12720 }, { "grad_norm": 0.2960880696773529, "learning_rate": 9.837567259497447e-05, "loss": 0.0406, "step": 12730 }, { "grad_norm": 0.36613893508911133, "learning_rate": 9.837148965509894e-05, "loss": 0.0388, "step": 12740 }, { "grad_norm": 0.25186261534690857, "learning_rate": 9.836730142539001e-05, "loss": 0.047, "step": 12750 }, { "grad_norm": 0.2797747850418091, "learning_rate": 9.836310790630574e-05, "loss": 0.0431, "step": 12760 }, { "grad_norm": 0.2778472900390625, "learning_rate": 9.83589090983047e-05, "loss": 0.0414, "step": 12770 }, { "grad_norm": 0.32753339409828186, "learning_rate": 9.835470500184605e-05, "loss": 0.0426, "step": 12780 }, { "grad_norm": 0.30349084734916687, "learning_rate": 9.835049561738957e-05, "loss": 0.0339, "step": 12790 }, { "grad_norm": 0.25011178851127625, "learning_rate": 9.834628094539558e-05, "loss": 0.04, "step": 12800 }, { "grad_norm": 0.3377917408943176, "learning_rate": 9.834206098632499e-05, "loss": 0.0459, "step": 12810 }, { "grad_norm": 0.3600689470767975, "learning_rate": 9.833783574063931e-05, "loss": 0.0443, "step": 12820 }, { "grad_norm": 0.3728122115135193, "learning_rate": 9.833360520880058e-05, "loss": 0.0431, "step": 12830 }, { "grad_norm": 0.3049481213092804, "learning_rate": 9.832936939127144e-05, "loss": 0.0386, "step": 12840 }, { "grad_norm": 0.3989103138446808, "learning_rate": 9.832512828851515e-05, "loss": 0.0395, "step": 12850 }, { "grad_norm": 0.5182158350944519, "learning_rate": 9.832088190099546e-05, "loss": 0.04, "step": 12860 }, { "grad_norm": 0.2828269898891449, "learning_rate": 9.831663022917679e-05, "loss": 0.039, "step": 12870 }, { "grad_norm": 0.3883172571659088, "learning_rate": 9.831237327352407e-05, "loss": 0.0439, "step": 12880 }, { "grad_norm": 0.287761926651001, "learning_rate": 9.830811103450286e-05, "loss": 0.0374, "step": 12890 }, { "grad_norm": 0.22414863109588623, "learning_rate": 9.830384351257924e-05, "loss": 0.0402, "step": 12900 }, { "grad_norm": 0.2984229624271393, "learning_rate": 9.829957070821993e-05, "loss": 0.0391, "step": 12910 }, { "grad_norm": 0.30395087599754333, "learning_rate": 9.829529262189218e-05, "loss": 0.0327, "step": 12920 }, { "grad_norm": 0.2308344542980194, "learning_rate": 9.829100925406385e-05, "loss": 0.0346, "step": 12930 }, { "grad_norm": 0.2660689353942871, "learning_rate": 9.828672060520333e-05, "loss": 0.0359, "step": 12940 }, { "grad_norm": 0.31662923097610474, "learning_rate": 9.828242667577966e-05, "loss": 0.0392, "step": 12950 }, { "grad_norm": 0.2898550033569336, "learning_rate": 9.82781274662624e-05, "loss": 0.0348, "step": 12960 }, { "grad_norm": 0.2589579224586487, "learning_rate": 9.82738229771217e-05, "loss": 0.0354, "step": 12970 }, { "grad_norm": 0.2938840985298157, "learning_rate": 9.826951320882829e-05, "loss": 0.0442, "step": 12980 }, { "grad_norm": 0.2623448669910431, "learning_rate": 9.826519816185351e-05, "loss": 0.0354, "step": 12990 }, { "grad_norm": 0.2845878303050995, "learning_rate": 9.826087783666921e-05, "loss": 0.0401, "step": 13000 }, { "grad_norm": 0.2940865755081177, "learning_rate": 9.825655223374787e-05, "loss": 0.0373, "step": 13010 }, { "grad_norm": 0.35209161043167114, "learning_rate": 9.825222135356253e-05, "loss": 0.0384, "step": 13020 }, { "grad_norm": 0.3647196888923645, "learning_rate": 9.82478851965868e-05, "loss": 0.0345, "step": 13030 }, { "grad_norm": 0.3474145233631134, "learning_rate": 9.82435437632949e-05, "loss": 0.0396, "step": 13040 }, { "grad_norm": 0.3238076865673065, "learning_rate": 9.823919705416158e-05, "loss": 0.0435, "step": 13050 }, { "grad_norm": 0.38507992029190063, "learning_rate": 9.82348450696622e-05, "loss": 0.0349, "step": 13060 }, { "grad_norm": 0.26351943612098694, "learning_rate": 9.823048781027268e-05, "loss": 0.0394, "step": 13070 }, { "grad_norm": 0.3206985890865326, "learning_rate": 9.822612527646953e-05, "loss": 0.0383, "step": 13080 }, { "grad_norm": 0.4123363792896271, "learning_rate": 9.822175746872984e-05, "loss": 0.0397, "step": 13090 }, { "grad_norm": 0.348063588142395, "learning_rate": 9.821738438753123e-05, "loss": 0.0503, "step": 13100 }, { "grad_norm": 0.3196176588535309, "learning_rate": 9.821300603335196e-05, "loss": 0.0392, "step": 13110 }, { "grad_norm": 0.2974485158920288, "learning_rate": 9.820862240667085e-05, "loss": 0.0485, "step": 13120 }, { "grad_norm": 0.325341135263443, "learning_rate": 9.820423350796726e-05, "loss": 0.0462, "step": 13130 }, { "grad_norm": 0.2600257694721222, "learning_rate": 9.819983933772118e-05, "loss": 0.049, "step": 13140 }, { "grad_norm": 0.27277135848999023, "learning_rate": 9.819543989641314e-05, "loss": 0.0379, "step": 13150 }, { "grad_norm": 0.2862090468406677, "learning_rate": 9.819103518452423e-05, "loss": 0.0369, "step": 13160 }, { "grad_norm": 0.2174135446548462, "learning_rate": 9.818662520253618e-05, "loss": 0.0394, "step": 13170 }, { "grad_norm": 0.2510294020175934, "learning_rate": 9.818220995093126e-05, "loss": 0.0376, "step": 13180 }, { "grad_norm": 0.2285630702972412, "learning_rate": 9.817778943019228e-05, "loss": 0.04, "step": 13190 }, { "grad_norm": 0.307352751493454, "learning_rate": 9.81733636408027e-05, "loss": 0.0423, "step": 13200 }, { "grad_norm": 0.34281855821609497, "learning_rate": 9.816893258324649e-05, "loss": 0.0365, "step": 13210 }, { "grad_norm": 0.31106260418891907, "learning_rate": 9.816449625800823e-05, "loss": 0.0477, "step": 13220 }, { "grad_norm": 0.31316009163856506, "learning_rate": 9.816005466557308e-05, "loss": 0.0356, "step": 13230 }, { "grad_norm": 0.281779021024704, "learning_rate": 9.815560780642674e-05, "loss": 0.0328, "step": 13240 }, { "grad_norm": 0.2996721565723419, "learning_rate": 9.815115568105555e-05, "loss": 0.0471, "step": 13250 }, { "grad_norm": 0.2887083888053894, "learning_rate": 9.814669828994638e-05, "loss": 0.0425, "step": 13260 }, { "grad_norm": 0.25593653321266174, "learning_rate": 9.814223563358665e-05, "loss": 0.0421, "step": 13270 }, { "grad_norm": 0.49793729186058044, "learning_rate": 9.813776771246443e-05, "loss": 0.0413, "step": 13280 }, { "grad_norm": 0.4494016766548157, "learning_rate": 9.813329452706829e-05, "loss": 0.0402, "step": 13290 }, { "grad_norm": 0.3432735502719879, "learning_rate": 9.812881607788744e-05, "loss": 0.0414, "step": 13300 }, { "grad_norm": 0.25481662154197693, "learning_rate": 9.812433236541163e-05, "loss": 0.0422, "step": 13310 }, { "grad_norm": 0.3458648920059204, "learning_rate": 9.811984339013116e-05, "loss": 0.0389, "step": 13320 }, { "grad_norm": 0.2631092667579651, "learning_rate": 9.811534915253698e-05, "loss": 0.0351, "step": 13330 }, { "grad_norm": 0.2747398316860199, "learning_rate": 9.811084965312056e-05, "loss": 0.0398, "step": 13340 }, { "grad_norm": 0.33792978525161743, "learning_rate": 9.810634489237396e-05, "loss": 0.0382, "step": 13350 }, { "grad_norm": 0.36521878838539124, "learning_rate": 9.81018348707898e-05, "loss": 0.0393, "step": 13360 }, { "grad_norm": 0.32720836997032166, "learning_rate": 9.809731958886131e-05, "loss": 0.0371, "step": 13370 }, { "grad_norm": 0.33459293842315674, "learning_rate": 9.809279904708224e-05, "loss": 0.0366, "step": 13380 }, { "grad_norm": 0.24548222124576569, "learning_rate": 9.808827324594699e-05, "loss": 0.0377, "step": 13390 }, { "grad_norm": 0.40546315908432007, "learning_rate": 9.808374218595046e-05, "loss": 0.039, "step": 13400 }, { "grad_norm": 0.31479284167289734, "learning_rate": 9.80792058675882e-05, "loss": 0.0372, "step": 13410 }, { "grad_norm": 0.32750841975212097, "learning_rate": 9.807466429135627e-05, "loss": 0.0404, "step": 13420 }, { "grad_norm": 0.334699809551239, "learning_rate": 9.807011745775132e-05, "loss": 0.0398, "step": 13430 }, { "grad_norm": 0.2570916712284088, "learning_rate": 9.806556536727061e-05, "loss": 0.0343, "step": 13440 }, { "grad_norm": 0.3018801212310791, "learning_rate": 9.806100802041193e-05, "loss": 0.0346, "step": 13450 }, { "grad_norm": 0.2877376079559326, "learning_rate": 9.805644541767368e-05, "loss": 0.0367, "step": 13460 }, { "grad_norm": 0.22874176502227783, "learning_rate": 9.805187755955478e-05, "loss": 0.0357, "step": 13470 }, { "grad_norm": 0.30522385239601135, "learning_rate": 9.804730444655483e-05, "loss": 0.0407, "step": 13480 }, { "grad_norm": 0.28973859548568726, "learning_rate": 9.804272607917388e-05, "loss": 0.0329, "step": 13490 }, { "grad_norm": 0.38678210973739624, "learning_rate": 9.803814245791265e-05, "loss": 0.0386, "step": 13500 }, { "grad_norm": 0.2578030824661255, "learning_rate": 9.803355358327239e-05, "loss": 0.0354, "step": 13510 }, { "grad_norm": 0.3041863739490509, "learning_rate": 9.802895945575492e-05, "loss": 0.0368, "step": 13520 }, { "grad_norm": 0.6265033483505249, "learning_rate": 9.802436007586266e-05, "loss": 0.0367, "step": 13530 }, { "grad_norm": 0.34223178029060364, "learning_rate": 9.801975544409858e-05, "loss": 0.0364, "step": 13540 }, { "grad_norm": 0.31234657764434814, "learning_rate": 9.801514556096625e-05, "loss": 0.0399, "step": 13550 }, { "grad_norm": 0.2395200878381729, "learning_rate": 9.801053042696977e-05, "loss": 0.0402, "step": 13560 }, { "grad_norm": 0.2798481583595276, "learning_rate": 9.800591004261388e-05, "loss": 0.0369, "step": 13570 }, { "grad_norm": 0.3158271014690399, "learning_rate": 9.800128440840385e-05, "loss": 0.0403, "step": 13580 }, { "grad_norm": 0.33565017580986023, "learning_rate": 9.799665352484552e-05, "loss": 0.0379, "step": 13590 }, { "grad_norm": 0.30948537588119507, "learning_rate": 9.799201739244532e-05, "loss": 0.0396, "step": 13600 }, { "grad_norm": 0.42942681908607483, "learning_rate": 9.798737601171025e-05, "loss": 0.0408, "step": 13610 }, { "grad_norm": 0.3347659111022949, "learning_rate": 9.79827293831479e-05, "loss": 0.0364, "step": 13620 }, { "grad_norm": 0.3458830714225769, "learning_rate": 9.797807750726638e-05, "loss": 0.0409, "step": 13630 }, { "grad_norm": 0.323188841342926, "learning_rate": 9.797342038457446e-05, "loss": 0.037, "step": 13640 }, { "grad_norm": 0.29470884799957275, "learning_rate": 9.796875801558141e-05, "loss": 0.0349, "step": 13650 }, { "grad_norm": 0.2546238899230957, "learning_rate": 9.79640904007971e-05, "loss": 0.0348, "step": 13660 }, { "grad_norm": 0.25182127952575684, "learning_rate": 9.795941754073199e-05, "loss": 0.0302, "step": 13670 }, { "grad_norm": 0.3128451704978943, "learning_rate": 9.795473943589705e-05, "loss": 0.0394, "step": 13680 }, { "grad_norm": 0.22666093707084656, "learning_rate": 9.795005608680394e-05, "loss": 0.0358, "step": 13690 }, { "grad_norm": 0.32842203974723816, "learning_rate": 9.794536749396477e-05, "loss": 0.0392, "step": 13700 }, { "grad_norm": 0.3588382601737976, "learning_rate": 9.79406736578923e-05, "loss": 0.0422, "step": 13710 }, { "grad_norm": 0.33090341091156006, "learning_rate": 9.793597457909984e-05, "loss": 0.0428, "step": 13720 }, { "grad_norm": 0.32738181948661804, "learning_rate": 9.793127025810127e-05, "loss": 0.0345, "step": 13730 }, { "grad_norm": 0.2513921856880188, "learning_rate": 9.792656069541104e-05, "loss": 0.0298, "step": 13740 }, { "grad_norm": 0.39710119366645813, "learning_rate": 9.79218458915442e-05, "loss": 0.0399, "step": 13750 }, { "grad_norm": 0.3247661888599396, "learning_rate": 9.791712584701634e-05, "loss": 0.0362, "step": 13760 }, { "grad_norm": 0.2883260250091553, "learning_rate": 9.791240056234364e-05, "loss": 0.033, "step": 13770 }, { "grad_norm": 0.2353980392217636, "learning_rate": 9.790767003804283e-05, "loss": 0.0315, "step": 13780 }, { "grad_norm": 0.30197983980178833, "learning_rate": 9.790293427463126e-05, "loss": 0.0342, "step": 13790 }, { "grad_norm": 0.3640075623989105, "learning_rate": 9.789819327262684e-05, "loss": 0.0323, "step": 13800 }, { "grad_norm": 0.26150476932525635, "learning_rate": 9.7893447032548e-05, "loss": 0.032, "step": 13810 }, { "grad_norm": 0.25291574001312256, "learning_rate": 9.78886955549138e-05, "loss": 0.0338, "step": 13820 }, { "grad_norm": 0.28551825881004333, "learning_rate": 9.788393884024387e-05, "loss": 0.0373, "step": 13830 }, { "grad_norm": 0.37465816736221313, "learning_rate": 9.787917688905836e-05, "loss": 0.0347, "step": 13840 }, { "grad_norm": 0.24215511977672577, "learning_rate": 9.787440970187807e-05, "loss": 0.0322, "step": 13850 }, { "grad_norm": 0.25642791390419006, "learning_rate": 9.786963727922429e-05, "loss": 0.0395, "step": 13860 }, { "grad_norm": 0.34010884165763855, "learning_rate": 9.786485962161897e-05, "loss": 0.0338, "step": 13870 }, { "grad_norm": 0.2224397510290146, "learning_rate": 9.786007672958455e-05, "loss": 0.0309, "step": 13880 }, { "grad_norm": 0.25739020109176636, "learning_rate": 9.78552886036441e-05, "loss": 0.0365, "step": 13890 }, { "grad_norm": 0.23785021901130676, "learning_rate": 9.785049524432124e-05, "loss": 0.032, "step": 13900 }, { "grad_norm": 0.27734771370887756, "learning_rate": 9.784569665214016e-05, "loss": 0.0379, "step": 13910 }, { "grad_norm": 0.2667106091976166, "learning_rate": 9.784089282762563e-05, "loss": 0.0343, "step": 13920 }, { "grad_norm": 0.2881315052509308, "learning_rate": 9.7836083771303e-05, "loss": 0.0345, "step": 13930 }, { "grad_norm": 0.21391765773296356, "learning_rate": 9.783126948369817e-05, "loss": 0.0323, "step": 13940 }, { "grad_norm": 0.23508913815021515, "learning_rate": 9.78264499653376e-05, "loss": 0.0353, "step": 13950 }, { "grad_norm": 0.23773498833179474, "learning_rate": 9.782162521674838e-05, "loss": 0.0331, "step": 13960 }, { "grad_norm": 0.27781692147254944, "learning_rate": 9.781679523845812e-05, "loss": 0.0356, "step": 13970 }, { "grad_norm": 0.26316747069358826, "learning_rate": 9.781196003099502e-05, "loss": 0.0313, "step": 13980 }, { "grad_norm": 0.2340203821659088, "learning_rate": 9.780711959488786e-05, "loss": 0.0359, "step": 13990 }, { "grad_norm": 0.24709133803844452, "learning_rate": 9.780227393066599e-05, "loss": 0.0326, "step": 14000 }, { "grad_norm": 0.3055279850959778, "learning_rate": 9.77974230388593e-05, "loss": 0.0336, "step": 14010 }, { "grad_norm": 0.21505403518676758, "learning_rate": 9.779256691999829e-05, "loss": 0.0428, "step": 14020 }, { "grad_norm": 0.24175554513931274, "learning_rate": 9.778770557461403e-05, "loss": 0.0364, "step": 14030 }, { "grad_norm": 0.2733537554740906, "learning_rate": 9.778283900323812e-05, "loss": 0.0373, "step": 14040 }, { "grad_norm": 0.30366867780685425, "learning_rate": 9.777796720640277e-05, "loss": 0.0351, "step": 14050 }, { "grad_norm": 0.3134421110153198, "learning_rate": 9.777309018464078e-05, "loss": 0.0318, "step": 14060 }, { "grad_norm": 0.29294469952583313, "learning_rate": 9.776820793848547e-05, "loss": 0.0336, "step": 14070 }, { "grad_norm": 0.31278249621391296, "learning_rate": 9.776332046847075e-05, "loss": 0.0359, "step": 14080 }, { "grad_norm": 0.32349318265914917, "learning_rate": 9.775842777513111e-05, "loss": 0.0366, "step": 14090 }, { "grad_norm": 0.2767930328845978, "learning_rate": 9.775352985900163e-05, "loss": 0.0336, "step": 14100 }, { "grad_norm": 0.32760149240493774, "learning_rate": 9.774862672061791e-05, "loss": 0.0322, "step": 14110 }, { "grad_norm": 0.32983383536338806, "learning_rate": 9.774371836051616e-05, "loss": 0.038, "step": 14120 }, { "grad_norm": 0.31600624322891235, "learning_rate": 9.773880477923315e-05, "loss": 0.0337, "step": 14130 }, { "grad_norm": 0.2180221527814865, "learning_rate": 9.773388597730623e-05, "loss": 0.0332, "step": 14140 }, { "grad_norm": 0.29690471291542053, "learning_rate": 9.77289619552733e-05, "loss": 0.0342, "step": 14150 }, { "grad_norm": 0.30264535546302795, "learning_rate": 9.772403271367285e-05, "loss": 0.0324, "step": 14160 }, { "grad_norm": 0.23706813156604767, "learning_rate": 9.771909825304396e-05, "loss": 0.0347, "step": 14170 }, { "grad_norm": 0.26583945751190186, "learning_rate": 9.771415857392619e-05, "loss": 0.0326, "step": 14180 }, { "grad_norm": 0.26536357402801514, "learning_rate": 9.770921367685978e-05, "loss": 0.0299, "step": 14190 }, { "grad_norm": 0.29222726821899414, "learning_rate": 9.770426356238551e-05, "loss": 0.0349, "step": 14200 }, { "grad_norm": 0.261313259601593, "learning_rate": 9.769930823104469e-05, "loss": 0.0359, "step": 14210 }, { "grad_norm": 0.31219425797462463, "learning_rate": 9.769434768337926e-05, "loss": 0.0343, "step": 14220 }, { "grad_norm": 0.37188994884490967, "learning_rate": 9.768938191993164e-05, "loss": 0.0338, "step": 14230 }, { "grad_norm": 0.43989139795303345, "learning_rate": 9.768441094124494e-05, "loss": 0.0366, "step": 14240 }, { "grad_norm": 0.2639882266521454, "learning_rate": 9.767943474786275e-05, "loss": 0.031, "step": 14250 }, { "grad_norm": 0.2928404211997986, "learning_rate": 9.767445334032923e-05, "loss": 0.0329, "step": 14260 }, { "grad_norm": 0.2521282732486725, "learning_rate": 9.766946671918919e-05, "loss": 0.0366, "step": 14270 }, { "grad_norm": 0.34267041087150574, "learning_rate": 9.766447488498796e-05, "loss": 0.0428, "step": 14280 }, { "grad_norm": 0.4284228980541229, "learning_rate": 9.765947783827139e-05, "loss": 0.033, "step": 14290 }, { "grad_norm": 0.31129714846611023, "learning_rate": 9.765447557958599e-05, "loss": 0.0369, "step": 14300 }, { "grad_norm": 0.24469445645809174, "learning_rate": 9.764946810947879e-05, "loss": 0.0347, "step": 14310 }, { "grad_norm": 0.2773723900318146, "learning_rate": 9.764445542849738e-05, "loss": 0.0331, "step": 14320 }, { "grad_norm": 0.3452497124671936, "learning_rate": 9.763943753718998e-05, "loss": 0.0337, "step": 14330 }, { "grad_norm": 0.2443310171365738, "learning_rate": 9.76344144361053e-05, "loss": 0.0317, "step": 14340 }, { "grad_norm": 0.2589702010154724, "learning_rate": 9.762938612579269e-05, "loss": 0.0391, "step": 14350 }, { "grad_norm": 0.21584828197956085, "learning_rate": 9.762435260680202e-05, "loss": 0.0358, "step": 14360 }, { "grad_norm": 0.3022271692752838, "learning_rate": 9.761931387968373e-05, "loss": 0.0352, "step": 14370 }, { "grad_norm": 0.32511141896247864, "learning_rate": 9.76142699449889e-05, "loss": 0.0327, "step": 14380 }, { "grad_norm": 0.30711454153060913, "learning_rate": 9.760922080326908e-05, "loss": 0.0355, "step": 14390 }, { "grad_norm": 0.2945803105831146, "learning_rate": 9.760416645507644e-05, "loss": 0.0362, "step": 14400 }, { "grad_norm": 0.27105847001075745, "learning_rate": 9.759910690096375e-05, "loss": 0.0351, "step": 14410 }, { "grad_norm": 0.29040056467056274, "learning_rate": 9.759404214148429e-05, "loss": 0.0436, "step": 14420 }, { "grad_norm": 0.30709895491600037, "learning_rate": 9.758897217719191e-05, "loss": 0.041, "step": 14430 }, { "grad_norm": 0.3356644809246063, "learning_rate": 9.758389700864113e-05, "loss": 0.0417, "step": 14440 }, { "grad_norm": 0.3098624050617218, "learning_rate": 9.757881663638688e-05, "loss": 0.0391, "step": 14450 }, { "grad_norm": 0.24552392959594727, "learning_rate": 9.757373106098478e-05, "loss": 0.0393, "step": 14460 }, { "grad_norm": 0.2766810655593872, "learning_rate": 9.756864028299097e-05, "loss": 0.0272, "step": 14470 }, { "grad_norm": 0.2884629964828491, "learning_rate": 9.75635443029622e-05, "loss": 0.0341, "step": 14480 }, { "grad_norm": 0.2554100751876831, "learning_rate": 9.755844312145572e-05, "loss": 0.0349, "step": 14490 }, { "grad_norm": 0.2588379383087158, "learning_rate": 9.755333673902941e-05, "loss": 0.0317, "step": 14500 }, { "grad_norm": 0.29842960834503174, "learning_rate": 9.75482251562417e-05, "loss": 0.0321, "step": 14510 }, { "grad_norm": 0.35492852330207825, "learning_rate": 9.754310837365155e-05, "loss": 0.0336, "step": 14520 }, { "grad_norm": 0.274381548166275, "learning_rate": 9.753798639181856e-05, "loss": 0.0325, "step": 14530 }, { "grad_norm": 0.25919368863105774, "learning_rate": 9.753285921130286e-05, "loss": 0.032, "step": 14540 }, { "grad_norm": 0.286914199590683, "learning_rate": 9.752772683266512e-05, "loss": 0.0322, "step": 14550 }, { "grad_norm": 0.20764924585819244, "learning_rate": 9.752258925646665e-05, "loss": 0.0317, "step": 14560 }, { "grad_norm": 0.2787712514400482, "learning_rate": 9.751744648326926e-05, "loss": 0.037, "step": 14570 }, { "grad_norm": 0.3015483021736145, "learning_rate": 9.751229851363536e-05, "loss": 0.0329, "step": 14580 }, { "grad_norm": 0.28849613666534424, "learning_rate": 9.750714534812793e-05, "loss": 0.0367, "step": 14590 }, { "grad_norm": 0.3413008451461792, "learning_rate": 9.750198698731053e-05, "loss": 0.0379, "step": 14600 }, { "grad_norm": 0.28975120186805725, "learning_rate": 9.749682343174722e-05, "loss": 0.0365, "step": 14610 }, { "grad_norm": 0.2986695468425751, "learning_rate": 9.749165468200272e-05, "loss": 0.038, "step": 14620 }, { "grad_norm": 0.2809639573097229, "learning_rate": 9.748648073864229e-05, "loss": 0.0317, "step": 14630 }, { "grad_norm": 0.30001765489578247, "learning_rate": 9.748130160223168e-05, "loss": 0.0351, "step": 14640 }, { "grad_norm": 0.303689181804657, "learning_rate": 9.747611727333734e-05, "loss": 0.0352, "step": 14650 }, { "grad_norm": 0.2826749384403229, "learning_rate": 9.74709277525262e-05, "loss": 0.0407, "step": 14660 }, { "grad_norm": 0.277803510427475, "learning_rate": 9.746573304036576e-05, "loss": 0.0327, "step": 14670 }, { "grad_norm": 0.2342057079076767, "learning_rate": 9.746053313742412e-05, "loss": 0.0324, "step": 14680 }, { "grad_norm": 0.3780374228954315, "learning_rate": 9.745532804426994e-05, "loss": 0.0315, "step": 14690 }, { "grad_norm": 0.2134404480457306, "learning_rate": 9.745011776147242e-05, "loss": 0.0435, "step": 14700 }, { "grad_norm": 0.2726178467273712, "learning_rate": 9.744490228960138e-05, "loss": 0.0341, "step": 14710 }, { "grad_norm": 0.26284128427505493, "learning_rate": 9.743968162922713e-05, "loss": 0.0328, "step": 14720 }, { "grad_norm": 0.2908937931060791, "learning_rate": 9.743445578092064e-05, "loss": 0.0342, "step": 14730 }, { "grad_norm": 0.24996766448020935, "learning_rate": 9.742922474525338e-05, "loss": 0.0326, "step": 14740 }, { "grad_norm": 0.2562509775161743, "learning_rate": 9.742398852279741e-05, "loss": 0.0303, "step": 14750 }, { "grad_norm": 0.3089108467102051, "learning_rate": 9.741874711412535e-05, "loss": 0.0349, "step": 14760 }, { "grad_norm": 0.2639855146408081, "learning_rate": 9.741350051981042e-05, "loss": 0.032, "step": 14770 }, { "grad_norm": 0.29578036069869995, "learning_rate": 9.740824874042633e-05, "loss": 0.0333, "step": 14780 }, { "grad_norm": 0.31100109219551086, "learning_rate": 9.740299177654746e-05, "loss": 0.0296, "step": 14790 }, { "grad_norm": 0.32179808616638184, "learning_rate": 9.739772962874867e-05, "loss": 0.0413, "step": 14800 }, { "grad_norm": 0.24398058652877808, "learning_rate": 9.739246229760541e-05, "loss": 0.0383, "step": 14810 }, { "grad_norm": 0.343957781791687, "learning_rate": 9.738718978369376e-05, "loss": 0.0357, "step": 14820 }, { "grad_norm": 0.2765892446041107, "learning_rate": 9.738191208759025e-05, "loss": 0.0336, "step": 14830 }, { "grad_norm": 0.21158742904663086, "learning_rate": 9.73766292098721e-05, "loss": 0.0325, "step": 14840 }, { "grad_norm": 0.25538092851638794, "learning_rate": 9.737134115111699e-05, "loss": 0.0373, "step": 14850 }, { "grad_norm": 0.23532238602638245, "learning_rate": 9.736604791190323e-05, "loss": 0.04, "step": 14860 }, { "grad_norm": 0.3320574164390564, "learning_rate": 9.73607494928097e-05, "loss": 0.0372, "step": 14870 }, { "grad_norm": 0.25583919882774353, "learning_rate": 9.735544589441581e-05, "loss": 0.0373, "step": 14880 }, { "grad_norm": 0.3537604808807373, "learning_rate": 9.735013711730154e-05, "loss": 0.0339, "step": 14890 }, { "grad_norm": 0.29651299118995667, "learning_rate": 9.734482316204747e-05, "loss": 0.0371, "step": 14900 }, { "grad_norm": 0.34486448764801025, "learning_rate": 9.733950402923473e-05, "loss": 0.0308, "step": 14910 }, { "grad_norm": 0.2635531425476074, "learning_rate": 9.7334179719445e-05, "loss": 0.0403, "step": 14920 }, { "grad_norm": 0.33360373973846436, "learning_rate": 9.732885023326053e-05, "loss": 0.042, "step": 14930 }, { "grad_norm": 0.3043363094329834, "learning_rate": 9.732351557126418e-05, "loss": 0.0395, "step": 14940 }, { "grad_norm": 0.3556044101715088, "learning_rate": 9.731817573403929e-05, "loss": 0.036, "step": 14950 }, { "grad_norm": 0.38771331310272217, "learning_rate": 9.731283072216985e-05, "loss": 0.0333, "step": 14960 }, { "grad_norm": 0.2898343801498413, "learning_rate": 9.730748053624039e-05, "loss": 0.0365, "step": 14970 }, { "grad_norm": 0.28056392073631287, "learning_rate": 9.730212517683598e-05, "loss": 0.0386, "step": 14980 }, { "grad_norm": 0.3652091920375824, "learning_rate": 9.729676464454228e-05, "loss": 0.0361, "step": 14990 }, { "grad_norm": 0.4435567855834961, "learning_rate": 9.72913989399455e-05, "loss": 0.0403, "step": 15000 }, { "grad_norm": 0.2660548686981201, "learning_rate": 9.728602806363242e-05, "loss": 0.037, "step": 15010 }, { "grad_norm": 0.3088749647140503, "learning_rate": 9.728065201619043e-05, "loss": 0.0371, "step": 15020 }, { "grad_norm": 0.31870508193969727, "learning_rate": 9.727527079820742e-05, "loss": 0.0412, "step": 15030 }, { "grad_norm": 0.2312532216310501, "learning_rate": 9.726988441027186e-05, "loss": 0.0415, "step": 15040 }, { "grad_norm": 0.25565671920776367, "learning_rate": 9.726449285297281e-05, "loss": 0.0347, "step": 15050 }, { "grad_norm": 0.367854505777359, "learning_rate": 9.72590961268999e-05, "loss": 0.0433, "step": 15060 }, { "grad_norm": 0.3445506989955902, "learning_rate": 9.725369423264328e-05, "loss": 0.0389, "step": 15070 }, { "grad_norm": 0.22267302870750427, "learning_rate": 9.72482871707937e-05, "loss": 0.0355, "step": 15080 }, { "grad_norm": 0.3181397020816803, "learning_rate": 9.724287494194247e-05, "loss": 0.0349, "step": 15090 }, { "grad_norm": 0.3054862916469574, "learning_rate": 9.723745754668147e-05, "loss": 0.0386, "step": 15100 }, { "grad_norm": 0.3583819568157196, "learning_rate": 9.723203498560313e-05, "loss": 0.0367, "step": 15110 }, { "grad_norm": 0.3903188109397888, "learning_rate": 9.722660725930046e-05, "loss": 0.0388, "step": 15120 }, { "grad_norm": 0.34163740277290344, "learning_rate": 9.722117436836702e-05, "loss": 0.0351, "step": 15130 }, { "grad_norm": 0.25849515199661255, "learning_rate": 9.721573631339696e-05, "loss": 0.0323, "step": 15140 }, { "grad_norm": 0.3084145486354828, "learning_rate": 9.721029309498494e-05, "loss": 0.036, "step": 15150 }, { "grad_norm": 0.27523893117904663, "learning_rate": 9.720484471372627e-05, "loss": 0.0384, "step": 15160 }, { "grad_norm": 0.27196428179740906, "learning_rate": 9.719939117021673e-05, "loss": 0.034, "step": 15170 }, { "grad_norm": 0.302689790725708, "learning_rate": 9.719393246505275e-05, "loss": 0.0285, "step": 15180 }, { "grad_norm": 0.25891387462615967, "learning_rate": 9.718846859883128e-05, "loss": 0.0366, "step": 15190 }, { "grad_norm": 0.2147124707698822, "learning_rate": 9.718299957214982e-05, "loss": 0.0301, "step": 15200 }, { "grad_norm": 0.3935260474681854, "learning_rate": 9.717752538560646e-05, "loss": 0.0368, "step": 15210 }, { "grad_norm": 0.29108133912086487, "learning_rate": 9.717204603979986e-05, "loss": 0.0337, "step": 15220 }, { "grad_norm": 0.26061591506004333, "learning_rate": 9.716656153532922e-05, "loss": 0.037, "step": 15230 }, { "grad_norm": 0.3091268837451935, "learning_rate": 9.716107187279434e-05, "loss": 0.0339, "step": 15240 }, { "grad_norm": 0.3037129044532776, "learning_rate": 9.715557705279555e-05, "loss": 0.0305, "step": 15250 }, { "grad_norm": 0.31030693650245667, "learning_rate": 9.715007707593372e-05, "loss": 0.0318, "step": 15260 }, { "grad_norm": 0.26006564497947693, "learning_rate": 9.714457194281036e-05, "loss": 0.0353, "step": 15270 }, { "grad_norm": 0.3438926935195923, "learning_rate": 9.713906165402751e-05, "loss": 0.0416, "step": 15280 }, { "grad_norm": 0.32606229186058044, "learning_rate": 9.713354621018774e-05, "loss": 0.0357, "step": 15290 }, { "grad_norm": 0.29445815086364746, "learning_rate": 9.712802561189422e-05, "loss": 0.037, "step": 15300 }, { "grad_norm": 0.2772170603275299, "learning_rate": 9.712249985975069e-05, "loss": 0.0309, "step": 15310 }, { "grad_norm": 0.3162381052970886, "learning_rate": 9.71169689543614e-05, "loss": 0.0285, "step": 15320 }, { "grad_norm": 0.3232446014881134, "learning_rate": 9.711143289633123e-05, "loss": 0.036, "step": 15330 }, { "grad_norm": 0.28883302211761475, "learning_rate": 9.710589168626561e-05, "loss": 0.0299, "step": 15340 }, { "grad_norm": 0.3230142295360565, "learning_rate": 9.710034532477048e-05, "loss": 0.0352, "step": 15350 }, { "grad_norm": 0.25552862882614136, "learning_rate": 9.709479381245239e-05, "loss": 0.0397, "step": 15360 }, { "grad_norm": 0.2649746835231781, "learning_rate": 9.708923714991847e-05, "loss": 0.0404, "step": 15370 }, { "grad_norm": 0.25485631823539734, "learning_rate": 9.708367533777638e-05, "loss": 0.0324, "step": 15380 }, { "grad_norm": 0.28962740302085876, "learning_rate": 9.707810837663431e-05, "loss": 0.0322, "step": 15390 }, { "grad_norm": 0.2901704013347626, "learning_rate": 9.707253626710113e-05, "loss": 0.0341, "step": 15400 }, { "grad_norm": 0.26659831404685974, "learning_rate": 9.706695900978613e-05, "loss": 0.0334, "step": 15410 }, { "grad_norm": 0.2651692032814026, "learning_rate": 9.706137660529926e-05, "loss": 0.0338, "step": 15420 }, { "grad_norm": 0.2606465220451355, "learning_rate": 9.705578905425101e-05, "loss": 0.033, "step": 15430 }, { "grad_norm": 0.24997463822364807, "learning_rate": 9.705019635725241e-05, "loss": 0.0325, "step": 15440 }, { "grad_norm": 0.3107466697692871, "learning_rate": 9.704459851491508e-05, "loss": 0.0362, "step": 15450 }, { "grad_norm": 0.2932913303375244, "learning_rate": 9.703899552785118e-05, "loss": 0.03, "step": 15460 }, { "grad_norm": 0.2737484276294708, "learning_rate": 9.703338739667346e-05, "loss": 0.0315, "step": 15470 }, { "grad_norm": 0.23612038791179657, "learning_rate": 9.70277741219952e-05, "loss": 0.0295, "step": 15480 }, { "grad_norm": 0.3314518630504608, "learning_rate": 9.702215570443027e-05, "loss": 0.0317, "step": 15490 }, { "grad_norm": 0.4634520709514618, "learning_rate": 9.701653214459309e-05, "loss": 0.0349, "step": 15500 }, { "grad_norm": 0.2508816719055176, "learning_rate": 9.701090344309865e-05, "loss": 0.0378, "step": 15510 }, { "grad_norm": 0.25462010502815247, "learning_rate": 9.700526960056247e-05, "loss": 0.0331, "step": 15520 }, { "grad_norm": 0.27112656831741333, "learning_rate": 9.699963061760068e-05, "loss": 0.0338, "step": 15530 }, { "grad_norm": 0.3347164988517761, "learning_rate": 9.699398649482997e-05, "loss": 0.0347, "step": 15540 }, { "grad_norm": 0.330297589302063, "learning_rate": 9.698833723286753e-05, "loss": 0.0326, "step": 15550 }, { "grad_norm": 0.35875001549720764, "learning_rate": 9.698268283233118e-05, "loss": 0.0355, "step": 15560 }, { "grad_norm": 0.35209181904792786, "learning_rate": 9.697702329383929e-05, "loss": 0.032, "step": 15570 }, { "grad_norm": 0.30227574706077576, "learning_rate": 9.697135861801074e-05, "loss": 0.0318, "step": 15580 }, { "grad_norm": 0.3369952142238617, "learning_rate": 9.696568880546505e-05, "loss": 0.0398, "step": 15590 }, { "grad_norm": 0.3250758647918701, "learning_rate": 9.696001385682223e-05, "loss": 0.0381, "step": 15600 }, { "grad_norm": 0.24348567426204681, "learning_rate": 9.695433377270291e-05, "loss": 0.0435, "step": 15610 }, { "grad_norm": 0.23934230208396912, "learning_rate": 9.694864855372824e-05, "loss": 0.0258, "step": 15620 }, { "grad_norm": 0.22117340564727783, "learning_rate": 9.694295820051995e-05, "loss": 0.0291, "step": 15630 }, { "grad_norm": 0.28450506925582886, "learning_rate": 9.693726271370032e-05, "loss": 0.0292, "step": 15640 }, { "grad_norm": 0.30020272731781006, "learning_rate": 9.693156209389221e-05, "loss": 0.0326, "step": 15650 }, { "grad_norm": 0.26302579045295715, "learning_rate": 9.692585634171905e-05, "loss": 0.0298, "step": 15660 }, { "grad_norm": 0.3392316699028015, "learning_rate": 9.692014545780476e-05, "loss": 0.0326, "step": 15670 }, { "grad_norm": 0.2756618559360504, "learning_rate": 9.691442944277393e-05, "loss": 0.0337, "step": 15680 }, { "grad_norm": 0.2535751461982727, "learning_rate": 9.690870829725162e-05, "loss": 0.0271, "step": 15690 }, { "grad_norm": 0.29212576150894165, "learning_rate": 9.69029820218635e-05, "loss": 0.0291, "step": 15700 }, { "grad_norm": 0.2659573554992676, "learning_rate": 9.689725061723579e-05, "loss": 0.027, "step": 15710 }, { "grad_norm": 0.3363509476184845, "learning_rate": 9.689151408399527e-05, "loss": 0.0348, "step": 15720 }, { "grad_norm": 0.3052576184272766, "learning_rate": 9.688577242276924e-05, "loss": 0.0341, "step": 15730 }, { "grad_norm": 0.3161926865577698, "learning_rate": 9.688002563418566e-05, "loss": 0.0291, "step": 15740 }, { "grad_norm": 0.28802409768104553, "learning_rate": 9.687427371887293e-05, "loss": 0.0325, "step": 15750 }, { "grad_norm": 0.28860583901405334, "learning_rate": 9.686851667746012e-05, "loss": 0.0374, "step": 15760 }, { "grad_norm": 0.3190818428993225, "learning_rate": 9.686275451057677e-05, "loss": 0.0389, "step": 15770 }, { "grad_norm": 0.295091450214386, "learning_rate": 9.685698721885308e-05, "loss": 0.0277, "step": 15780 }, { "grad_norm": 0.25909075140953064, "learning_rate": 9.68512148029197e-05, "loss": 0.0352, "step": 15790 }, { "grad_norm": 0.3073492646217346, "learning_rate": 9.684543726340791e-05, "loss": 0.0315, "step": 15800 }, { "grad_norm": 0.32102805376052856, "learning_rate": 9.683965460094952e-05, "loss": 0.0341, "step": 15810 }, { "grad_norm": 0.3586331903934479, "learning_rate": 9.683386681617694e-05, "loss": 0.0401, "step": 15820 }, { "grad_norm": 0.27487713098526, "learning_rate": 9.68280739097231e-05, "loss": 0.0347, "step": 15830 }, { "grad_norm": 0.22964301705360413, "learning_rate": 9.682227588222148e-05, "loss": 0.0319, "step": 15840 }, { "grad_norm": 0.25841134786605835, "learning_rate": 9.681647273430618e-05, "loss": 0.0312, "step": 15850 }, { "grad_norm": 0.3634115755558014, "learning_rate": 9.681066446661182e-05, "loss": 0.0396, "step": 15860 }, { "grad_norm": 0.20897699892520905, "learning_rate": 9.680485107977357e-05, "loss": 0.0341, "step": 15870 }, { "grad_norm": 0.34135884046554565, "learning_rate": 9.679903257442716e-05, "loss": 0.041, "step": 15880 }, { "grad_norm": 0.254241019487381, "learning_rate": 9.679320895120891e-05, "loss": 0.0314, "step": 15890 }, { "grad_norm": 0.24481873214244843, "learning_rate": 9.67873802107557e-05, "loss": 0.0299, "step": 15900 }, { "grad_norm": 0.362565815448761, "learning_rate": 9.67815463537049e-05, "loss": 0.0322, "step": 15910 }, { "grad_norm": 0.26674216985702515, "learning_rate": 9.677570738069457e-05, "loss": 0.0341, "step": 15920 }, { "grad_norm": 0.2935384213924408, "learning_rate": 9.676986329236318e-05, "loss": 0.0369, "step": 15930 }, { "grad_norm": 0.3533395230770111, "learning_rate": 9.676401408934987e-05, "loss": 0.0307, "step": 15940 }, { "grad_norm": 0.24060538411140442, "learning_rate": 9.675815977229428e-05, "loss": 0.0398, "step": 15950 }, { "grad_norm": 0.30329665541648865, "learning_rate": 9.675230034183664e-05, "loss": 0.0305, "step": 15960 }, { "grad_norm": 0.22869747877120972, "learning_rate": 9.674643579861773e-05, "loss": 0.0313, "step": 15970 }, { "grad_norm": 0.2744176685810089, "learning_rate": 9.674056614327886e-05, "loss": 0.0346, "step": 15980 }, { "grad_norm": 0.3115042448043823, "learning_rate": 9.673469137646198e-05, "loss": 0.0332, "step": 15990 }, { "grad_norm": 0.25747445225715637, "learning_rate": 9.67288114988095e-05, "loss": 0.0349, "step": 16000 }, { "grad_norm": 0.6891515254974365, "learning_rate": 9.672292651096447e-05, "loss": 0.0438, "step": 16010 }, { "grad_norm": 0.24196016788482666, "learning_rate": 9.671703641357042e-05, "loss": 0.0391, "step": 16020 }, { "grad_norm": 0.3456735610961914, "learning_rate": 9.67111412072715e-05, "loss": 0.0424, "step": 16030 }, { "grad_norm": 0.2535969316959381, "learning_rate": 9.670524089271242e-05, "loss": 0.0353, "step": 16040 }, { "grad_norm": 0.34694480895996094, "learning_rate": 9.669933547053842e-05, "loss": 0.032, "step": 16050 }, { "grad_norm": 0.24827633798122406, "learning_rate": 9.669342494139531e-05, "loss": 0.0362, "step": 16060 }, { "grad_norm": 0.3255065679550171, "learning_rate": 9.668750930592943e-05, "loss": 0.0362, "step": 16070 }, { "grad_norm": 0.36284568905830383, "learning_rate": 9.668158856478775e-05, "loss": 0.0441, "step": 16080 }, { "grad_norm": 0.28888988494873047, "learning_rate": 9.66756627186177e-05, "loss": 0.0354, "step": 16090 }, { "grad_norm": 0.2570813298225403, "learning_rate": 9.666973176806737e-05, "loss": 0.0344, "step": 16100 }, { "grad_norm": 0.26255255937576294, "learning_rate": 9.666379571378534e-05, "loss": 0.0339, "step": 16110 }, { "grad_norm": 0.9382214546203613, "learning_rate": 9.665785455642076e-05, "loss": 0.0337, "step": 16120 }, { "grad_norm": 0.4109303057193756, "learning_rate": 9.665190829662337e-05, "loss": 0.0411, "step": 16130 }, { "grad_norm": 0.28751224279403687, "learning_rate": 9.664595693504342e-05, "loss": 0.0353, "step": 16140 }, { "grad_norm": 0.2585670053958893, "learning_rate": 9.664000047233175e-05, "loss": 0.0341, "step": 16150 }, { "grad_norm": 0.3148718774318695, "learning_rate": 9.663403890913976e-05, "loss": 0.0372, "step": 16160 }, { "grad_norm": 0.20417141914367676, "learning_rate": 9.662807224611938e-05, "loss": 0.0331, "step": 16170 }, { "grad_norm": 0.28565117716789246, "learning_rate": 9.662210048392311e-05, "loss": 0.0339, "step": 16180 }, { "grad_norm": 0.29219794273376465, "learning_rate": 9.661612362320405e-05, "loss": 0.0343, "step": 16190 }, { "grad_norm": 0.4606810212135315, "learning_rate": 9.661014166461579e-05, "loss": 0.0302, "step": 16200 }, { "grad_norm": 0.334590882062912, "learning_rate": 9.66041546088125e-05, "loss": 0.0454, "step": 16210 }, { "grad_norm": 0.30920901894569397, "learning_rate": 9.659816245644895e-05, "loss": 0.0326, "step": 16220 }, { "grad_norm": 0.30135470628738403, "learning_rate": 9.65921652081804e-05, "loss": 0.0367, "step": 16230 }, { "grad_norm": 0.34073641896247864, "learning_rate": 9.658616286466271e-05, "loss": 0.0377, "step": 16240 }, { "grad_norm": 0.28693467378616333, "learning_rate": 9.65801554265523e-05, "loss": 0.0393, "step": 16250 }, { "grad_norm": 0.23382243514060974, "learning_rate": 9.657414289450612e-05, "loss": 0.0315, "step": 16260 }, { "grad_norm": 0.27280932664871216, "learning_rate": 9.656812526918171e-05, "loss": 0.0351, "step": 16270 }, { "grad_norm": 0.26217734813690186, "learning_rate": 9.656210255123712e-05, "loss": 0.037, "step": 16280 }, { "grad_norm": 0.29114386439323425, "learning_rate": 9.6556074741331e-05, "loss": 0.0299, "step": 16290 }, { "grad_norm": 0.2959083616733551, "learning_rate": 9.655004184012256e-05, "loss": 0.0358, "step": 16300 }, { "grad_norm": 0.24725013971328735, "learning_rate": 9.654400384827152e-05, "loss": 0.0385, "step": 16310 }, { "grad_norm": 0.3027280569076538, "learning_rate": 9.653796076643818e-05, "loss": 0.0358, "step": 16320 }, { "grad_norm": 0.3101494014263153, "learning_rate": 9.653191259528344e-05, "loss": 0.0352, "step": 16330 }, { "grad_norm": 0.27461034059524536, "learning_rate": 9.65258593354687e-05, "loss": 0.0346, "step": 16340 }, { "grad_norm": 0.2702934443950653, "learning_rate": 9.651980098765591e-05, "loss": 0.0298, "step": 16350 }, { "grad_norm": 0.2467641681432724, "learning_rate": 9.651373755250765e-05, "loss": 0.0348, "step": 16360 }, { "grad_norm": 0.24837104976177216, "learning_rate": 9.650766903068697e-05, "loss": 0.034, "step": 16370 }, { "grad_norm": 0.332442969083786, "learning_rate": 9.650159542285753e-05, "loss": 0.036, "step": 16380 }, { "grad_norm": 0.2666580379009247, "learning_rate": 9.649551672968353e-05, "loss": 0.0333, "step": 16390 }, { "grad_norm": 0.3104778826236725, "learning_rate": 9.648943295182973e-05, "loss": 0.0332, "step": 16400 }, { "grad_norm": 0.28920674324035645, "learning_rate": 9.648334408996144e-05, "loss": 0.0358, "step": 16410 }, { "grad_norm": 0.3061271011829376, "learning_rate": 9.647725014474452e-05, "loss": 0.0346, "step": 16420 }, { "grad_norm": 0.2480398714542389, "learning_rate": 9.64711511168454e-05, "loss": 0.0338, "step": 16430 }, { "grad_norm": 0.330721378326416, "learning_rate": 9.646504700693108e-05, "loss": 0.0319, "step": 16440 }, { "grad_norm": 0.27821531891822815, "learning_rate": 9.645893781566907e-05, "loss": 0.0336, "step": 16450 }, { "grad_norm": 0.26592886447906494, "learning_rate": 9.645282354372744e-05, "loss": 0.031, "step": 16460 }, { "grad_norm": 0.22388872504234314, "learning_rate": 9.644670419177491e-05, "loss": 0.0306, "step": 16470 }, { "grad_norm": 0.2518552839756012, "learning_rate": 9.644057976048062e-05, "loss": 0.0313, "step": 16480 }, { "grad_norm": 0.2702423334121704, "learning_rate": 9.643445025051435e-05, "loss": 0.0283, "step": 16490 }, { "grad_norm": 0.31991684436798096, "learning_rate": 9.642831566254641e-05, "loss": 0.0378, "step": 16500 }, { "grad_norm": 0.2608754634857178, "learning_rate": 9.642217599724769e-05, "loss": 0.0353, "step": 16510 }, { "grad_norm": 0.23827916383743286, "learning_rate": 9.64160312552896e-05, "loss": 0.0289, "step": 16520 }, { "grad_norm": 0.3084511458873749, "learning_rate": 9.64098814373441e-05, "loss": 0.0337, "step": 16530 }, { "grad_norm": 0.263764351606369, "learning_rate": 9.640372654408374e-05, "loss": 0.0347, "step": 16540 }, { "grad_norm": 0.38032904267311096, "learning_rate": 9.639756657618162e-05, "loss": 0.0338, "step": 16550 }, { "grad_norm": 0.2749463617801666, "learning_rate": 9.639140153431138e-05, "loss": 0.0364, "step": 16560 }, { "grad_norm": 0.2598888874053955, "learning_rate": 9.638523141914721e-05, "loss": 0.0339, "step": 16570 }, { "grad_norm": 0.25017642974853516, "learning_rate": 9.637905623136388e-05, "loss": 0.0313, "step": 16580 }, { "grad_norm": 0.20783784985542297, "learning_rate": 9.637287597163669e-05, "loss": 0.0326, "step": 16590 }, { "grad_norm": 0.23641709983348846, "learning_rate": 9.63666906406415e-05, "loss": 0.0269, "step": 16600 }, { "grad_norm": 0.28982046246528625, "learning_rate": 9.636050023905473e-05, "loss": 0.0325, "step": 16610 }, { "grad_norm": 0.2697049379348755, "learning_rate": 9.635430476755336e-05, "loss": 0.0339, "step": 16620 }, { "grad_norm": 0.3731878399848938, "learning_rate": 9.63481042268149e-05, "loss": 0.0361, "step": 16630 }, { "grad_norm": 0.27754175662994385, "learning_rate": 9.634189861751745e-05, "loss": 0.033, "step": 16640 }, { "grad_norm": 0.2108234316110611, "learning_rate": 9.633568794033967e-05, "loss": 0.0334, "step": 16650 }, { "grad_norm": 0.28251370787620544, "learning_rate": 9.63294721959607e-05, "loss": 0.0346, "step": 16660 }, { "grad_norm": 0.30468398332595825, "learning_rate": 9.63232513850603e-05, "loss": 0.0384, "step": 16670 }, { "grad_norm": 0.27548307180404663, "learning_rate": 9.631702550831878e-05, "loss": 0.0349, "step": 16680 }, { "grad_norm": 0.2918286621570587, "learning_rate": 9.631079456641698e-05, "loss": 0.0345, "step": 16690 }, { "grad_norm": 0.3228692412376404, "learning_rate": 9.630455856003632e-05, "loss": 0.0304, "step": 16700 }, { "grad_norm": 0.30974099040031433, "learning_rate": 9.629831748985876e-05, "loss": 0.0373, "step": 16710 }, { "grad_norm": 0.22570382058620453, "learning_rate": 9.629207135656679e-05, "loss": 0.0321, "step": 16720 }, { "grad_norm": 0.5321520566940308, "learning_rate": 9.628582016084353e-05, "loss": 0.0301, "step": 16730 }, { "grad_norm": 0.23256908357143402, "learning_rate": 9.627956390337254e-05, "loss": 0.0349, "step": 16740 }, { "grad_norm": 0.36735349893569946, "learning_rate": 9.627330258483802e-05, "loss": 0.034, "step": 16750 }, { "grad_norm": 0.264870285987854, "learning_rate": 9.62670362059247e-05, "loss": 0.0396, "step": 16760 }, { "grad_norm": 0.38924935460090637, "learning_rate": 9.626076476731786e-05, "loss": 0.0338, "step": 16770 }, { "grad_norm": 0.38380518555641174, "learning_rate": 9.625448826970336e-05, "loss": 0.0349, "step": 16780 }, { "grad_norm": 0.3029647767543793, "learning_rate": 9.624820671376755e-05, "loss": 0.0343, "step": 16790 }, { "grad_norm": 0.4331294596195221, "learning_rate": 9.62419201001974e-05, "loss": 0.0388, "step": 16800 }, { "grad_norm": 0.3989524245262146, "learning_rate": 9.623562842968037e-05, "loss": 0.0356, "step": 16810 }, { "grad_norm": 0.4200061559677124, "learning_rate": 9.622933170290454e-05, "loss": 0.0372, "step": 16820 }, { "grad_norm": 0.25751811265945435, "learning_rate": 9.622302992055849e-05, "loss": 0.0338, "step": 16830 }, { "grad_norm": 0.2654813826084137, "learning_rate": 9.62167230833314e-05, "loss": 0.0372, "step": 16840 }, { "grad_norm": 0.25905126333236694, "learning_rate": 9.621041119191295e-05, "loss": 0.029, "step": 16850 }, { "grad_norm": 0.23499339818954468, "learning_rate": 9.620409424699342e-05, "loss": 0.0353, "step": 16860 }, { "grad_norm": 0.2555527985095978, "learning_rate": 9.619777224926359e-05, "loss": 0.0309, "step": 16870 }, { "grad_norm": 0.22028748691082, "learning_rate": 9.619144519941485e-05, "loss": 0.0368, "step": 16880 }, { "grad_norm": 0.33574768900871277, "learning_rate": 9.618511309813912e-05, "loss": 0.031, "step": 16890 }, { "grad_norm": 0.30050769448280334, "learning_rate": 9.617877594612886e-05, "loss": 0.0337, "step": 16900 }, { "grad_norm": 0.36175310611724854, "learning_rate": 9.617243374407707e-05, "loss": 0.0273, "step": 16910 }, { "grad_norm": 0.277650386095047, "learning_rate": 9.616608649267736e-05, "loss": 0.0381, "step": 16920 }, { "grad_norm": 0.2631755471229553, "learning_rate": 9.615973419262385e-05, "loss": 0.0346, "step": 16930 }, { "grad_norm": 0.23145653307437897, "learning_rate": 9.615337684461119e-05, "loss": 0.0345, "step": 16940 }, { "grad_norm": 0.23230110108852386, "learning_rate": 9.614701444933465e-05, "loss": 0.0331, "step": 16950 }, { "grad_norm": 0.2687970995903015, "learning_rate": 9.614064700748997e-05, "loss": 0.0338, "step": 16960 }, { "grad_norm": 0.20456883311271667, "learning_rate": 9.613427451977352e-05, "loss": 0.0286, "step": 16970 }, { "grad_norm": 0.2771492898464203, "learning_rate": 9.612789698688216e-05, "loss": 0.0308, "step": 16980 }, { "grad_norm": 0.2699139416217804, "learning_rate": 9.612151440951334e-05, "loss": 0.0332, "step": 16990 }, { "grad_norm": 0.23374012112617493, "learning_rate": 9.611512678836506e-05, "loss": 0.0302, "step": 17000 }, { "grad_norm": 0.2455156445503235, "learning_rate": 9.610873412413584e-05, "loss": 0.0293, "step": 17010 }, { "grad_norm": 0.30970022082328796, "learning_rate": 9.610233641752476e-05, "loss": 0.035, "step": 17020 }, { "grad_norm": 0.23062995076179504, "learning_rate": 9.609593366923151e-05, "loss": 0.0283, "step": 17030 }, { "grad_norm": 0.3460191488265991, "learning_rate": 9.608952587995625e-05, "loss": 0.0355, "step": 17040 }, { "grad_norm": 0.25813567638397217, "learning_rate": 9.608311305039972e-05, "loss": 0.0307, "step": 17050 }, { "grad_norm": 0.2141980081796646, "learning_rate": 9.607669518126326e-05, "loss": 0.0327, "step": 17060 }, { "grad_norm": 0.28621137142181396, "learning_rate": 9.607027227324866e-05, "loss": 0.0329, "step": 17070 }, { "grad_norm": 0.23501300811767578, "learning_rate": 9.606384432705837e-05, "loss": 0.0295, "step": 17080 }, { "grad_norm": 0.2829117178916931, "learning_rate": 9.60574113433953e-05, "loss": 0.0324, "step": 17090 }, { "grad_norm": 0.3068890869617462, "learning_rate": 9.6050973322963e-05, "loss": 0.0297, "step": 17100 }, { "grad_norm": 0.2756653428077698, "learning_rate": 9.604453026646547e-05, "loss": 0.0328, "step": 17110 }, { "grad_norm": 0.22802750766277313, "learning_rate": 9.603808217460735e-05, "loss": 0.0349, "step": 17120 }, { "grad_norm": 0.30523088574409485, "learning_rate": 9.603162904809377e-05, "loss": 0.0306, "step": 17130 }, { "grad_norm": 0.320470929145813, "learning_rate": 9.602517088763045e-05, "loss": 0.0294, "step": 17140 }, { "grad_norm": 0.32993289828300476, "learning_rate": 9.601870769392365e-05, "loss": 0.0329, "step": 17150 }, { "grad_norm": 0.4023357927799225, "learning_rate": 9.601223946768017e-05, "loss": 0.0336, "step": 17160 }, { "grad_norm": 0.3142189085483551, "learning_rate": 9.600576620960734e-05, "loss": 0.0298, "step": 17170 }, { "grad_norm": 0.26000407338142395, "learning_rate": 9.599928792041308e-05, "loss": 0.0283, "step": 17180 }, { "grad_norm": 0.372917503118515, "learning_rate": 9.599280460080587e-05, "loss": 0.036, "step": 17190 }, { "grad_norm": 0.217434823513031, "learning_rate": 9.59863162514947e-05, "loss": 0.0244, "step": 17200 }, { "grad_norm": 0.2907239496707916, "learning_rate": 9.597982287318911e-05, "loss": 0.0287, "step": 17210 }, { "grad_norm": 0.2751898169517517, "learning_rate": 9.597332446659923e-05, "loss": 0.0291, "step": 17220 }, { "grad_norm": 0.2932552993297577, "learning_rate": 9.59668210324357e-05, "loss": 0.0319, "step": 17230 }, { "grad_norm": 0.2208014279603958, "learning_rate": 9.596031257140974e-05, "loss": 0.0283, "step": 17240 }, { "grad_norm": 0.22066451609134674, "learning_rate": 9.59537990842331e-05, "loss": 0.0229, "step": 17250 }, { "grad_norm": 0.31250083446502686, "learning_rate": 9.594728057161806e-05, "loss": 0.0298, "step": 17260 }, { "grad_norm": 0.24821113049983978, "learning_rate": 9.594075703427752e-05, "loss": 0.0318, "step": 17270 }, { "grad_norm": 0.2776879072189331, "learning_rate": 9.593422847292486e-05, "loss": 0.0307, "step": 17280 }, { "grad_norm": 0.24001707136631012, "learning_rate": 9.592769488827402e-05, "loss": 0.0321, "step": 17290 }, { "grad_norm": 0.3006582260131836, "learning_rate": 9.592115628103952e-05, "loss": 0.0301, "step": 17300 }, { "grad_norm": 0.33859965205192566, "learning_rate": 9.591461265193643e-05, "loss": 0.0367, "step": 17310 }, { "grad_norm": 0.23776037991046906, "learning_rate": 9.590806400168032e-05, "loss": 0.0319, "step": 17320 }, { "grad_norm": 0.2107274979352951, "learning_rate": 9.590151033098735e-05, "loss": 0.0305, "step": 17330 }, { "grad_norm": 0.3256738483905792, "learning_rate": 9.589495164057423e-05, "loss": 0.0357, "step": 17340 }, { "grad_norm": 0.32717663049697876, "learning_rate": 9.58883879311582e-05, "loss": 0.0349, "step": 17350 }, { "grad_norm": 0.40824130177497864, "learning_rate": 9.588181920345705e-05, "loss": 0.0401, "step": 17360 }, { "grad_norm": 0.29986080527305603, "learning_rate": 9.587524545818913e-05, "loss": 0.0365, "step": 17370 }, { "grad_norm": 0.33309993147850037, "learning_rate": 9.586866669607335e-05, "loss": 0.0339, "step": 17380 }, { "grad_norm": 0.30223459005355835, "learning_rate": 9.586208291782915e-05, "loss": 0.0258, "step": 17390 }, { "grad_norm": 0.344524621963501, "learning_rate": 9.58554941241765e-05, "loss": 0.0336, "step": 17400 }, { "grad_norm": 0.235386461019516, "learning_rate": 9.584890031583596e-05, "loss": 0.0308, "step": 17410 }, { "grad_norm": 0.18260271847248077, "learning_rate": 9.584230149352861e-05, "loss": 0.0241, "step": 17420 }, { "grad_norm": 0.2373114824295044, "learning_rate": 9.58356976579761e-05, "loss": 0.0254, "step": 17430 }, { "grad_norm": 0.35674774646759033, "learning_rate": 9.58290888099006e-05, "loss": 0.0295, "step": 17440 }, { "grad_norm": 0.19590981304645538, "learning_rate": 9.582247495002486e-05, "loss": 0.0305, "step": 17450 }, { "grad_norm": 0.28100329637527466, "learning_rate": 9.581585607907214e-05, "loss": 0.0314, "step": 17460 }, { "grad_norm": 0.23834510147571564, "learning_rate": 9.580923219776628e-05, "loss": 0.0299, "step": 17470 }, { "grad_norm": 0.24319647252559662, "learning_rate": 9.580260330683167e-05, "loss": 0.0301, "step": 17480 }, { "grad_norm": 0.21488618850708008, "learning_rate": 9.579596940699322e-05, "loss": 0.0295, "step": 17490 }, { "grad_norm": 0.2606527507305145, "learning_rate": 9.578933049897643e-05, "loss": 0.0348, "step": 17500 }, { "grad_norm": 0.2718023955821991, "learning_rate": 9.578268658350728e-05, "loss": 0.0283, "step": 17510 }, { "grad_norm": 0.2089015245437622, "learning_rate": 9.577603766131235e-05, "loss": 0.0303, "step": 17520 }, { "grad_norm": 0.31069910526275635, "learning_rate": 9.576938373311878e-05, "loss": 0.027, "step": 17530 }, { "grad_norm": 0.21880125999450684, "learning_rate": 9.576272479965421e-05, "loss": 0.0278, "step": 17540 }, { "grad_norm": 0.24819132685661316, "learning_rate": 9.575606086164687e-05, "loss": 0.0274, "step": 17550 }, { "grad_norm": 0.26477426290512085, "learning_rate": 9.57493919198255e-05, "loss": 0.0301, "step": 17560 }, { "grad_norm": 0.22116973996162415, "learning_rate": 9.57427179749194e-05, "loss": 0.0299, "step": 17570 }, { "grad_norm": 0.18797188997268677, "learning_rate": 9.573603902765846e-05, "loss": 0.0266, "step": 17580 }, { "grad_norm": 0.22689682245254517, "learning_rate": 9.572935507877304e-05, "loss": 0.0297, "step": 17590 }, { "grad_norm": 0.25124919414520264, "learning_rate": 9.57226661289941e-05, "loss": 0.037, "step": 17600 }, { "grad_norm": 0.305941641330719, "learning_rate": 9.571597217905315e-05, "loss": 0.0327, "step": 17610 }, { "grad_norm": 0.23489877581596375, "learning_rate": 9.57092732296822e-05, "loss": 0.0294, "step": 17620 }, { "grad_norm": 0.2810666263103485, "learning_rate": 9.570256928161385e-05, "loss": 0.0339, "step": 17630 }, { "grad_norm": 0.2148725986480713, "learning_rate": 9.569586033558126e-05, "loss": 0.032, "step": 17640 }, { "grad_norm": 0.2972099483013153, "learning_rate": 9.568914639231807e-05, "loss": 0.041, "step": 17650 }, { "grad_norm": 0.2592768669128418, "learning_rate": 9.568242745255852e-05, "loss": 0.0384, "step": 17660 }, { "grad_norm": 0.24582435190677643, "learning_rate": 9.567570351703739e-05, "loss": 0.0284, "step": 17670 }, { "grad_norm": 0.3358895778656006, "learning_rate": 9.566897458649001e-05, "loss": 0.0388, "step": 17680 }, { "grad_norm": 0.26496925950050354, "learning_rate": 9.566224066165221e-05, "loss": 0.0327, "step": 17690 }, { "grad_norm": 0.2963041365146637, "learning_rate": 9.565550174326043e-05, "loss": 0.0367, "step": 17700 }, { "grad_norm": 0.3055281937122345, "learning_rate": 9.564875783205162e-05, "loss": 0.0298, "step": 17710 }, { "grad_norm": 0.26434606313705444, "learning_rate": 9.564200892876328e-05, "loss": 0.0364, "step": 17720 }, { "grad_norm": 0.37890076637268066, "learning_rate": 9.563525503413348e-05, "loss": 0.0508, "step": 17730 }, { "grad_norm": 0.2694531977176666, "learning_rate": 9.562849614890079e-05, "loss": 0.0398, "step": 17740 }, { "grad_norm": 0.306747704744339, "learning_rate": 9.562173227380436e-05, "loss": 0.0395, "step": 17750 }, { "grad_norm": 0.27998533844947815, "learning_rate": 9.561496340958389e-05, "loss": 0.0351, "step": 17760 }, { "grad_norm": 0.3450969457626343, "learning_rate": 9.560818955697959e-05, "loss": 0.0395, "step": 17770 }, { "grad_norm": 0.25169578194618225, "learning_rate": 9.560141071673228e-05, "loss": 0.0325, "step": 17780 }, { "grad_norm": 0.23806972801685333, "learning_rate": 9.559462688958323e-05, "loss": 0.0289, "step": 17790 }, { "grad_norm": 0.26323002576828003, "learning_rate": 9.558783807627434e-05, "loss": 0.0267, "step": 17800 }, { "grad_norm": 0.2901581823825836, "learning_rate": 9.558104427754801e-05, "loss": 0.0309, "step": 17810 }, { "grad_norm": 0.3274593949317932, "learning_rate": 9.557424549414722e-05, "loss": 0.0312, "step": 17820 }, { "grad_norm": 0.3283700942993164, "learning_rate": 9.556744172681546e-05, "loss": 0.0292, "step": 17830 }, { "grad_norm": 0.31230688095092773, "learning_rate": 9.556063297629677e-05, "loss": 0.0315, "step": 17840 }, { "grad_norm": 0.3569613993167877, "learning_rate": 9.555381924333578e-05, "loss": 0.0353, "step": 17850 }, { "grad_norm": 0.21911215782165527, "learning_rate": 9.554700052867758e-05, "loss": 0.0277, "step": 17860 }, { "grad_norm": 0.30338165163993835, "learning_rate": 9.554017683306789e-05, "loss": 0.0296, "step": 17870 }, { "grad_norm": 0.36433956027030945, "learning_rate": 9.553334815725294e-05, "loss": 0.0351, "step": 17880 }, { "grad_norm": 0.2907641530036926, "learning_rate": 9.552651450197949e-05, "loss": 0.0335, "step": 17890 }, { "grad_norm": 0.2503392994403839, "learning_rate": 9.551967586799486e-05, "loss": 0.0344, "step": 17900 }, { "grad_norm": 0.2652073800563812, "learning_rate": 9.551283225604692e-05, "loss": 0.0382, "step": 17910 }, { "grad_norm": 0.20029449462890625, "learning_rate": 9.550598366688406e-05, "loss": 0.031, "step": 17920 }, { "grad_norm": 0.2996261417865753, "learning_rate": 9.549913010125526e-05, "loss": 0.03, "step": 17930 }, { "grad_norm": 0.2320801019668579, "learning_rate": 9.549227155990999e-05, "loss": 0.0302, "step": 17940 }, { "grad_norm": 0.5247793793678284, "learning_rate": 9.548540804359828e-05, "loss": 0.032, "step": 17950 }, { "grad_norm": 0.2849590480327606, "learning_rate": 9.547853955307077e-05, "loss": 0.0327, "step": 17960 }, { "grad_norm": 0.23742590844631195, "learning_rate": 9.547166608907853e-05, "loss": 0.0266, "step": 17970 }, { "grad_norm": 0.25098541378974915, "learning_rate": 9.546478765237326e-05, "loss": 0.0327, "step": 17980 }, { "grad_norm": 0.2819177210330963, "learning_rate": 9.545790424370715e-05, "loss": 0.0378, "step": 17990 }, { "grad_norm": 0.2920217216014862, "learning_rate": 9.5451015863833e-05, "loss": 0.035, "step": 18000 }, { "grad_norm": 0.25658854842185974, "learning_rate": 9.544412251350408e-05, "loss": 0.0334, "step": 18010 }, { "grad_norm": 0.3701631724834442, "learning_rate": 9.543722419347422e-05, "loss": 0.0344, "step": 18020 }, { "grad_norm": 0.25389769673347473, "learning_rate": 9.543032090449788e-05, "loss": 0.0308, "step": 18030 }, { "grad_norm": 0.29755479097366333, "learning_rate": 9.542341264732992e-05, "loss": 0.0313, "step": 18040 }, { "grad_norm": 0.20703023672103882, "learning_rate": 9.541649942272585e-05, "loss": 0.0313, "step": 18050 }, { "grad_norm": 0.20146366953849792, "learning_rate": 9.54095812314417e-05, "loss": 0.0341, "step": 18060 }, { "grad_norm": 0.25220394134521484, "learning_rate": 9.540265807423401e-05, "loss": 0.029, "step": 18070 }, { "grad_norm": 0.25114527344703674, "learning_rate": 9.53957299518599e-05, "loss": 0.031, "step": 18080 }, { "grad_norm": 0.24977223575115204, "learning_rate": 9.5388796865077e-05, "loss": 0.0224, "step": 18090 }, { "grad_norm": 0.2806992828845978, "learning_rate": 9.538185881464353e-05, "loss": 0.0332, "step": 18100 }, { "grad_norm": 0.286499947309494, "learning_rate": 9.537491580131821e-05, "loss": 0.0342, "step": 18110 }, { "grad_norm": 0.2915756404399872, "learning_rate": 9.53679678258603e-05, "loss": 0.0262, "step": 18120 }, { "grad_norm": 0.2727106809616089, "learning_rate": 9.536101488902966e-05, "loss": 0.0341, "step": 18130 }, { "grad_norm": 0.28088295459747314, "learning_rate": 9.535405699158663e-05, "loss": 0.0295, "step": 18140 }, { "grad_norm": 0.19800356030464172, "learning_rate": 9.53470941342921e-05, "loss": 0.0357, "step": 18150 }, { "grad_norm": 0.2673523426055908, "learning_rate": 9.534012631790756e-05, "loss": 0.0315, "step": 18160 }, { "grad_norm": 0.22908607125282288, "learning_rate": 9.533315354319494e-05, "loss": 0.0336, "step": 18170 }, { "grad_norm": 0.3071703016757965, "learning_rate": 9.532617581091682e-05, "loss": 0.0334, "step": 18180 }, { "grad_norm": 0.1775054782629013, "learning_rate": 9.531919312183629e-05, "loss": 0.0319, "step": 18190 }, { "grad_norm": 0.22459466755390167, "learning_rate": 9.531220547671688e-05, "loss": 0.0289, "step": 18200 }, { "grad_norm": 0.24070914089679718, "learning_rate": 9.530521287632285e-05, "loss": 0.0317, "step": 18210 }, { "grad_norm": 0.24624019861221313, "learning_rate": 9.529821532141884e-05, "loss": 0.0262, "step": 18220 }, { "grad_norm": 0.2721171975135803, "learning_rate": 9.52912128127701e-05, "loss": 0.0258, "step": 18230 }, { "grad_norm": 0.3256719708442688, "learning_rate": 9.528420535114244e-05, "loss": 0.0289, "step": 18240 }, { "grad_norm": 0.28470343351364136, "learning_rate": 9.527719293730215e-05, "loss": 0.0281, "step": 18250 }, { "grad_norm": 0.2804812788963318, "learning_rate": 9.527017557201611e-05, "loss": 0.0288, "step": 18260 }, { "grad_norm": 0.3503231406211853, "learning_rate": 9.526315325605176e-05, "loss": 0.0285, "step": 18270 }, { "grad_norm": 0.24878191947937012, "learning_rate": 9.525612599017699e-05, "loss": 0.0322, "step": 18280 }, { "grad_norm": 0.2537640631198883, "learning_rate": 9.524909377516033e-05, "loss": 0.0239, "step": 18290 }, { "grad_norm": 0.29796522855758667, "learning_rate": 9.524205661177081e-05, "loss": 0.0302, "step": 18300 }, { "grad_norm": 0.31740450859069824, "learning_rate": 9.523501450077801e-05, "loss": 0.0292, "step": 18310 }, { "grad_norm": 0.239588662981987, "learning_rate": 9.522796744295202e-05, "loss": 0.0284, "step": 18320 }, { "grad_norm": 0.348416268825531, "learning_rate": 9.522091543906352e-05, "loss": 0.0351, "step": 18330 }, { "grad_norm": 0.27575281262397766, "learning_rate": 9.521385848988369e-05, "loss": 0.029, "step": 18340 }, { "grad_norm": 0.250097393989563, "learning_rate": 9.520679659618428e-05, "loss": 0.0281, "step": 18350 }, { "grad_norm": 0.29368337988853455, "learning_rate": 9.519972975873754e-05, "loss": 0.0294, "step": 18360 }, { "grad_norm": 0.3314424753189087, "learning_rate": 9.519265797831633e-05, "loss": 0.033, "step": 18370 }, { "grad_norm": 0.2684437334537506, "learning_rate": 9.518558125569399e-05, "loss": 0.0328, "step": 18380 }, { "grad_norm": 0.23363441228866577, "learning_rate": 9.517849959164442e-05, "loss": 0.0331, "step": 18390 }, { "grad_norm": 0.33178427815437317, "learning_rate": 9.517141298694205e-05, "loss": 0.0378, "step": 18400 }, { "grad_norm": 0.32734501361846924, "learning_rate": 9.516432144236188e-05, "loss": 0.0386, "step": 18410 }, { "grad_norm": 0.3517460525035858, "learning_rate": 9.515722495867941e-05, "loss": 0.0285, "step": 18420 }, { "grad_norm": 0.2600436210632324, "learning_rate": 9.515012353667072e-05, "loss": 0.0353, "step": 18430 }, { "grad_norm": 0.2714396119117737, "learning_rate": 9.51430171771124e-05, "loss": 0.029, "step": 18440 }, { "grad_norm": 0.25096362829208374, "learning_rate": 9.513590588078159e-05, "loss": 0.0273, "step": 18450 }, { "grad_norm": 0.31805768609046936, "learning_rate": 9.512878964845597e-05, "loss": 0.0306, "step": 18460 }, { "grad_norm": 0.28352710604667664, "learning_rate": 9.512166848091377e-05, "loss": 0.0361, "step": 18470 }, { "grad_norm": 0.2904402017593384, "learning_rate": 9.511454237893376e-05, "loss": 0.0329, "step": 18480 }, { "grad_norm": 0.20694968104362488, "learning_rate": 9.51074113432952e-05, "loss": 0.0291, "step": 18490 }, { "grad_norm": 0.2240927368402481, "learning_rate": 9.510027537477797e-05, "loss": 0.031, "step": 18500 }, { "grad_norm": 0.336261510848999, "learning_rate": 9.509313447416242e-05, "loss": 0.0277, "step": 18510 }, { "grad_norm": 0.3450959622859955, "learning_rate": 9.508598864222949e-05, "loss": 0.0308, "step": 18520 }, { "grad_norm": 0.33992627263069153, "learning_rate": 9.507883787976062e-05, "loss": 0.0297, "step": 18530 }, { "grad_norm": 0.25519639253616333, "learning_rate": 9.507168218753781e-05, "loss": 0.0297, "step": 18540 }, { "grad_norm": 0.27072885632514954, "learning_rate": 9.506452156634362e-05, "loss": 0.0319, "step": 18550 }, { "grad_norm": 0.27800843119621277, "learning_rate": 9.505735601696109e-05, "loss": 0.0328, "step": 18560 }, { "grad_norm": 0.2309028059244156, "learning_rate": 9.505018554017385e-05, "loss": 0.0322, "step": 18570 }, { "grad_norm": 0.34458309412002563, "learning_rate": 9.504301013676604e-05, "loss": 0.0322, "step": 18580 }, { "grad_norm": 0.2498074620962143, "learning_rate": 9.503582980752238e-05, "loss": 0.0315, "step": 18590 }, { "grad_norm": 0.26286599040031433, "learning_rate": 9.502864455322809e-05, "loss": 0.0304, "step": 18600 }, { "grad_norm": 0.28092360496520996, "learning_rate": 9.502145437466891e-05, "loss": 0.0277, "step": 18610 }, { "grad_norm": 0.25479215383529663, "learning_rate": 9.501425927263116e-05, "loss": 0.0267, "step": 18620 }, { "grad_norm": 0.2578144073486328, "learning_rate": 9.500705924790172e-05, "loss": 0.0308, "step": 18630 }, { "grad_norm": 0.27617090940475464, "learning_rate": 9.499985430126794e-05, "loss": 0.0282, "step": 18640 }, { "grad_norm": 0.24216200411319733, "learning_rate": 9.499264443351775e-05, "loss": 0.0291, "step": 18650 }, { "grad_norm": 0.26896750926971436, "learning_rate": 9.498542964543961e-05, "loss": 0.0292, "step": 18660 }, { "grad_norm": 0.2640894949436188, "learning_rate": 9.497820993782252e-05, "loss": 0.0287, "step": 18670 }, { "grad_norm": 0.31302252411842346, "learning_rate": 9.497098531145601e-05, "loss": 0.0352, "step": 18680 }, { "grad_norm": 0.27899524569511414, "learning_rate": 9.496375576713017e-05, "loss": 0.0338, "step": 18690 }, { "grad_norm": 0.2556074857711792, "learning_rate": 9.49565213056356e-05, "loss": 0.0315, "step": 18700 }, { "grad_norm": 0.25562822818756104, "learning_rate": 9.494928192776342e-05, "loss": 0.0285, "step": 18710 }, { "grad_norm": 0.42743322253227234, "learning_rate": 9.494203763430538e-05, "loss": 0.0301, "step": 18720 }, { "grad_norm": 0.3691502511501312, "learning_rate": 9.493478842605366e-05, "loss": 0.0278, "step": 18730 }, { "grad_norm": 0.2789233326911926, "learning_rate": 9.492753430380105e-05, "loss": 0.0315, "step": 18740 }, { "grad_norm": 0.23351019620895386, "learning_rate": 9.492027526834083e-05, "loss": 0.0322, "step": 18750 }, { "grad_norm": 0.269034206867218, "learning_rate": 9.491301132046684e-05, "loss": 0.0317, "step": 18760 }, { "grad_norm": 0.24622084200382233, "learning_rate": 9.490574246097345e-05, "loss": 0.0253, "step": 18770 }, { "grad_norm": 0.3010234236717224, "learning_rate": 9.48984686906556e-05, "loss": 0.0233, "step": 18780 }, { "grad_norm": 0.2735303044319153, "learning_rate": 9.489119001030871e-05, "loss": 0.0249, "step": 18790 }, { "grad_norm": 0.2781805396080017, "learning_rate": 9.488390642072878e-05, "loss": 0.0279, "step": 18800 }, { "grad_norm": 0.21163690090179443, "learning_rate": 9.48766179227123e-05, "loss": 0.0259, "step": 18810 }, { "grad_norm": 0.2548230290412903, "learning_rate": 9.486932451705636e-05, "loss": 0.0262, "step": 18820 }, { "grad_norm": 0.21168826520442963, "learning_rate": 9.486202620455857e-05, "loss": 0.0258, "step": 18830 }, { "grad_norm": 0.2875572443008423, "learning_rate": 9.485472298601704e-05, "loss": 0.0282, "step": 18840 }, { "grad_norm": 0.28826308250427246, "learning_rate": 9.484741486223043e-05, "loss": 0.0367, "step": 18850 }, { "grad_norm": 0.3282787799835205, "learning_rate": 9.484010183399797e-05, "loss": 0.0321, "step": 18860 }, { "grad_norm": 0.28568559885025024, "learning_rate": 9.483278390211938e-05, "loss": 0.0345, "step": 18870 }, { "grad_norm": 0.3655472695827484, "learning_rate": 9.482546106739496e-05, "loss": 0.0267, "step": 18880 }, { "grad_norm": 0.2147955447435379, "learning_rate": 9.48181333306255e-05, "loss": 0.0339, "step": 18890 }, { "grad_norm": 0.2246023416519165, "learning_rate": 9.481080069261237e-05, "loss": 0.0273, "step": 18900 }, { "grad_norm": 0.31924566626548767, "learning_rate": 9.480346315415745e-05, "loss": 0.0257, "step": 18910 }, { "grad_norm": 0.22365063428878784, "learning_rate": 9.479612071606314e-05, "loss": 0.0293, "step": 18920 }, { "grad_norm": 0.25823238492012024, "learning_rate": 9.478877337913244e-05, "loss": 0.0302, "step": 18930 }, { "grad_norm": 0.284587562084198, "learning_rate": 9.478142114416881e-05, "loss": 0.0259, "step": 18940 }, { "grad_norm": 0.24407728016376495, "learning_rate": 9.47740640119763e-05, "loss": 0.029, "step": 18950 }, { "grad_norm": 0.29876917600631714, "learning_rate": 9.476670198335947e-05, "loss": 0.0312, "step": 18960 }, { "grad_norm": 0.23957933485507965, "learning_rate": 9.47593350591234e-05, "loss": 0.0317, "step": 18970 }, { "grad_norm": 0.25017470121383667, "learning_rate": 9.475196324007376e-05, "loss": 0.0296, "step": 18980 }, { "grad_norm": 0.23283202946186066, "learning_rate": 9.474458652701669e-05, "loss": 0.0311, "step": 18990 }, { "grad_norm": 0.24346435070037842, "learning_rate": 9.473720492075892e-05, "loss": 0.03, "step": 19000 }, { "grad_norm": 0.3140462636947632, "learning_rate": 9.472981842210768e-05, "loss": 0.0344, "step": 19010 }, { "grad_norm": 0.2245747298002243, "learning_rate": 9.472242703187074e-05, "loss": 0.0314, "step": 19020 }, { "grad_norm": 0.24815484881401062, "learning_rate": 9.471503075085643e-05, "loss": 0.028, "step": 19030 }, { "grad_norm": 0.2808056175708771, "learning_rate": 9.470762957987359e-05, "loss": 0.029, "step": 19040 }, { "grad_norm": 0.2569553256034851, "learning_rate": 9.470022351973158e-05, "loss": 0.0252, "step": 19050 }, { "grad_norm": 0.25365135073661804, "learning_rate": 9.469281257124034e-05, "loss": 0.0259, "step": 19060 }, { "grad_norm": 0.24626430869102478, "learning_rate": 9.46853967352103e-05, "loss": 0.0338, "step": 19070 }, { "grad_norm": 0.293247252702713, "learning_rate": 9.467797601245246e-05, "loss": 0.0282, "step": 19080 }, { "grad_norm": 0.3401992619037628, "learning_rate": 9.467055040377834e-05, "loss": 0.0332, "step": 19090 }, { "grad_norm": 0.27973341941833496, "learning_rate": 9.466311990999999e-05, "loss": 0.0346, "step": 19100 }, { "grad_norm": 0.3545305132865906, "learning_rate": 9.465568453193e-05, "loss": 0.031, "step": 19110 }, { "grad_norm": 0.30755430459976196, "learning_rate": 9.464824427038148e-05, "loss": 0.0335, "step": 19120 }, { "grad_norm": 0.3438699245452881, "learning_rate": 9.46407991261681e-05, "loss": 0.0294, "step": 19130 }, { "grad_norm": 0.2112903743982315, "learning_rate": 9.463334910010404e-05, "loss": 0.0321, "step": 19140 }, { "grad_norm": 0.2088727205991745, "learning_rate": 9.462589419300403e-05, "loss": 0.036, "step": 19150 }, { "grad_norm": 0.2672663927078247, "learning_rate": 9.461843440568333e-05, "loss": 0.0286, "step": 19160 }, { "grad_norm": 0.20804134011268616, "learning_rate": 9.461096973895773e-05, "loss": 0.0313, "step": 19170 }, { "grad_norm": 0.34312137961387634, "learning_rate": 9.460350019364355e-05, "loss": 0.0339, "step": 19180 }, { "grad_norm": 0.18554966151714325, "learning_rate": 9.459602577055764e-05, "loss": 0.0255, "step": 19190 }, { "grad_norm": 0.20622044801712036, "learning_rate": 9.45885464705174e-05, "loss": 0.0317, "step": 19200 }, { "grad_norm": 0.19893844425678253, "learning_rate": 9.458106229434076e-05, "loss": 0.0329, "step": 19210 }, { "grad_norm": 0.21821196377277374, "learning_rate": 9.457357324284617e-05, "loss": 0.0303, "step": 19220 }, { "grad_norm": 0.2035987377166748, "learning_rate": 9.456607931685262e-05, "loss": 0.0288, "step": 19230 }, { "grad_norm": 0.2414398342370987, "learning_rate": 9.455858051717965e-05, "loss": 0.026, "step": 19240 }, { "grad_norm": 0.230718731880188, "learning_rate": 9.45510768446473e-05, "loss": 0.0245, "step": 19250 }, { "grad_norm": 0.2461741864681244, "learning_rate": 9.454356830007618e-05, "loss": 0.0325, "step": 19260 }, { "grad_norm": 0.2656139135360718, "learning_rate": 9.45360548842874e-05, "loss": 0.039, "step": 19270 }, { "grad_norm": 0.2163236290216446, "learning_rate": 9.452853659810261e-05, "loss": 0.0299, "step": 19280 }, { "grad_norm": 0.20139314234256744, "learning_rate": 9.452101344234401e-05, "loss": 0.0318, "step": 19290 }, { "grad_norm": 0.2320900708436966, "learning_rate": 9.451348541783431e-05, "loss": 0.0293, "step": 19300 }, { "grad_norm": 0.30578792095184326, "learning_rate": 9.450595252539678e-05, "loss": 0.0334, "step": 19310 }, { "grad_norm": 0.2539439797401428, "learning_rate": 9.449841476585518e-05, "loss": 0.0301, "step": 19320 }, { "grad_norm": 0.3315797448158264, "learning_rate": 9.449087214003384e-05, "loss": 0.0287, "step": 19330 }, { "grad_norm": 0.28129273653030396, "learning_rate": 9.448332464875765e-05, "loss": 0.0317, "step": 19340 }, { "grad_norm": 0.24503248929977417, "learning_rate": 9.447577229285192e-05, "loss": 0.0291, "step": 19350 }, { "grad_norm": 0.2660091519355774, "learning_rate": 9.446821507314261e-05, "loss": 0.0277, "step": 19360 }, { "grad_norm": 0.270400732755661, "learning_rate": 9.446065299045617e-05, "loss": 0.0319, "step": 19370 }, { "grad_norm": 0.24186111986637115, "learning_rate": 9.445308604561955e-05, "loss": 0.032, "step": 19380 }, { "grad_norm": 0.18659576773643494, "learning_rate": 9.444551423946028e-05, "loss": 0.026, "step": 19390 }, { "grad_norm": 0.20280012488365173, "learning_rate": 9.443793757280638e-05, "loss": 0.0305, "step": 19400 }, { "grad_norm": 0.2681781053543091, "learning_rate": 9.443035604648646e-05, "loss": 0.0276, "step": 19410 }, { "grad_norm": 0.3600035309791565, "learning_rate": 9.44227696613296e-05, "loss": 0.0369, "step": 19420 }, { "grad_norm": 0.38596081733703613, "learning_rate": 9.441517841816542e-05, "loss": 0.037, "step": 19430 }, { "grad_norm": 0.31702664494514465, "learning_rate": 9.440758231782413e-05, "loss": 0.0323, "step": 19440 }, { "grad_norm": 0.26054880023002625, "learning_rate": 9.439998136113639e-05, "loss": 0.0325, "step": 19450 }, { "grad_norm": 0.45853081345558167, "learning_rate": 9.439237554893344e-05, "loss": 0.0297, "step": 19460 }, { "grad_norm": 0.23793408274650574, "learning_rate": 9.438476488204705e-05, "loss": 0.0331, "step": 19470 }, { "grad_norm": 0.26976001262664795, "learning_rate": 9.43771493613095e-05, "loss": 0.0323, "step": 19480 }, { "grad_norm": 0.2618996202945709, "learning_rate": 9.436952898755362e-05, "loss": 0.0279, "step": 19490 }, { "grad_norm": 0.28414103388786316, "learning_rate": 9.436190376161276e-05, "loss": 0.0356, "step": 19500 }, { "grad_norm": 0.31580308079719543, "learning_rate": 9.43542736843208e-05, "loss": 0.0316, "step": 19510 }, { "grad_norm": 0.2657964527606964, "learning_rate": 9.434663875651216e-05, "loss": 0.0304, "step": 19520 }, { "grad_norm": 0.2932145297527313, "learning_rate": 9.433899897902177e-05, "loss": 0.0315, "step": 19530 }, { "grad_norm": 0.2769638001918793, "learning_rate": 9.433135435268511e-05, "loss": 0.0277, "step": 19540 }, { "grad_norm": 0.21201978623867035, "learning_rate": 9.432370487833819e-05, "loss": 0.0319, "step": 19550 }, { "grad_norm": 0.2748103141784668, "learning_rate": 9.431605055681756e-05, "loss": 0.0285, "step": 19560 }, { "grad_norm": 0.23744745552539825, "learning_rate": 9.430839138896026e-05, "loss": 0.0321, "step": 19570 }, { "grad_norm": 0.23902875185012817, "learning_rate": 9.43007273756039e-05, "loss": 0.0238, "step": 19580 }, { "grad_norm": 0.23124168813228607, "learning_rate": 9.429305851758658e-05, "loss": 0.0285, "step": 19590 }, { "grad_norm": 0.2670486569404602, "learning_rate": 9.428538481574699e-05, "loss": 0.0282, "step": 19600 }, { "grad_norm": 0.18977013230323792, "learning_rate": 9.42777062709243e-05, "loss": 0.0409, "step": 19610 }, { "grad_norm": 0.30368998646736145, "learning_rate": 9.427002288395821e-05, "loss": 0.031, "step": 19620 }, { "grad_norm": 0.23841708898544312, "learning_rate": 9.426233465568898e-05, "loss": 0.0345, "step": 19630 }, { "grad_norm": 0.28237321972846985, "learning_rate": 9.42546415869574e-05, "loss": 0.0325, "step": 19640 }, { "grad_norm": 0.41705405712127686, "learning_rate": 9.424694367860473e-05, "loss": 0.0387, "step": 19650 }, { "grad_norm": 0.23061278462409973, "learning_rate": 9.423924093147284e-05, "loss": 0.0286, "step": 19660 }, { "grad_norm": 0.23239095509052277, "learning_rate": 9.423153334640407e-05, "loss": 0.0337, "step": 19670 }, { "grad_norm": 0.25082525610923767, "learning_rate": 9.42238209242413e-05, "loss": 0.0331, "step": 19680 }, { "grad_norm": 0.2984689176082611, "learning_rate": 9.421610366582798e-05, "loss": 0.035, "step": 19690 }, { "grad_norm": 0.26770803332328796, "learning_rate": 9.420838157200803e-05, "loss": 0.0312, "step": 19700 }, { "grad_norm": 0.27045807242393494, "learning_rate": 9.420065464362594e-05, "loss": 0.031, "step": 19710 }, { "grad_norm": 0.30282458662986755, "learning_rate": 9.419292288152673e-05, "loss": 0.0298, "step": 19720 }, { "grad_norm": 0.28029483556747437, "learning_rate": 9.418518628655588e-05, "loss": 0.0273, "step": 19730 }, { "grad_norm": 0.31875625252723694, "learning_rate": 9.417744485955951e-05, "loss": 0.027, "step": 19740 }, { "grad_norm": 0.2723501920700073, "learning_rate": 9.41696986013842e-05, "loss": 0.0257, "step": 19750 }, { "grad_norm": 0.31051573157310486, "learning_rate": 9.416194751287705e-05, "loss": 0.0258, "step": 19760 }, { "grad_norm": 0.31315240263938904, "learning_rate": 9.415419159488572e-05, "loss": 0.0285, "step": 19770 }, { "grad_norm": 0.25964003801345825, "learning_rate": 9.414643084825837e-05, "loss": 0.027, "step": 19780 }, { "grad_norm": 0.2741195559501648, "learning_rate": 9.413866527384372e-05, "loss": 0.0289, "step": 19790 }, { "grad_norm": 0.26668107509613037, "learning_rate": 9.4130894872491e-05, "loss": 0.0308, "step": 19800 }, { "grad_norm": 0.3786512017250061, "learning_rate": 9.412311964504998e-05, "loss": 0.0342, "step": 19810 }, { "grad_norm": 0.21981805562973022, "learning_rate": 9.411533959237091e-05, "loss": 0.0282, "step": 19820 }, { "grad_norm": 0.3404329717159271, "learning_rate": 9.410755471530464e-05, "loss": 0.0304, "step": 19830 }, { "grad_norm": 0.24586212635040283, "learning_rate": 9.40997650147025e-05, "loss": 0.0299, "step": 19840 }, { "grad_norm": 0.2995378375053406, "learning_rate": 9.409197049141637e-05, "loss": 0.0313, "step": 19850 }, { "grad_norm": 0.31417980790138245, "learning_rate": 9.408417114629863e-05, "loss": 0.03, "step": 19860 }, { "grad_norm": 0.2427179366350174, "learning_rate": 9.40763669802022e-05, "loss": 0.0269, "step": 19870 }, { "grad_norm": 0.3961351811885834, "learning_rate": 9.406855799398056e-05, "loss": 0.0322, "step": 19880 }, { "grad_norm": 0.32439151406288147, "learning_rate": 9.406074418848767e-05, "loss": 0.0376, "step": 19890 }, { "grad_norm": 0.20726360380649567, "learning_rate": 9.405292556457805e-05, "loss": 0.0355, "step": 19900 }, { "grad_norm": 0.29650136828422546, "learning_rate": 9.404510212310671e-05, "loss": 0.0292, "step": 19910 }, { "grad_norm": 0.3234997093677521, "learning_rate": 9.403727386492924e-05, "loss": 0.0312, "step": 19920 }, { "grad_norm": 0.2636062204837799, "learning_rate": 9.40294407909017e-05, "loss": 0.0295, "step": 19930 }, { "grad_norm": 0.17975759506225586, "learning_rate": 9.40216029018807e-05, "loss": 0.0261, "step": 19940 }, { "grad_norm": 0.20582866668701172, "learning_rate": 9.401376019872338e-05, "loss": 0.0266, "step": 19950 }, { "grad_norm": 0.24689774215221405, "learning_rate": 9.400591268228746e-05, "loss": 0.0238, "step": 19960 }, { "grad_norm": 0.29829198122024536, "learning_rate": 9.399806035343106e-05, "loss": 0.0298, "step": 19970 }, { "grad_norm": 0.27410823106765747, "learning_rate": 9.399020321301294e-05, "loss": 0.0375, "step": 19980 }, { "grad_norm": 0.21115006506443024, "learning_rate": 9.398234126189234e-05, "loss": 0.0334, "step": 19990 }, { "grad_norm": 0.2706219255924225, "learning_rate": 9.397447450092902e-05, "loss": 0.0314, "step": 20000 }, { "grad_norm": 0.25777989625930786, "learning_rate": 9.39666029309833e-05, "loss": 0.0278, "step": 20010 }, { "grad_norm": 0.3400562107563019, "learning_rate": 9.395872655291596e-05, "loss": 0.03, "step": 20020 }, { "grad_norm": 0.208025723695755, "learning_rate": 9.395084536758838e-05, "loss": 0.0274, "step": 20030 }, { "grad_norm": 0.2682800590991974, "learning_rate": 9.394295937586243e-05, "loss": 0.0344, "step": 20040 }, { "grad_norm": 0.2815673351287842, "learning_rate": 9.393506857860052e-05, "loss": 0.0297, "step": 20050 }, { "grad_norm": 0.29493448138237, "learning_rate": 9.392717297666555e-05, "loss": 0.0341, "step": 20060 }, { "grad_norm": 0.26549336314201355, "learning_rate": 9.391927257092101e-05, "loss": 0.0287, "step": 20070 }, { "grad_norm": 0.2439284324645996, "learning_rate": 9.391136736223085e-05, "loss": 0.0315, "step": 20080 }, { "grad_norm": 0.25063517689704895, "learning_rate": 9.390345735145956e-05, "loss": 0.0327, "step": 20090 }, { "grad_norm": 0.23203478753566742, "learning_rate": 9.389554253947219e-05, "loss": 0.0295, "step": 20100 }, { "grad_norm": 0.2801978290081024, "learning_rate": 9.388762292713428e-05, "loss": 0.0271, "step": 20110 }, { "grad_norm": 0.2624700963497162, "learning_rate": 9.38796985153119e-05, "loss": 0.0316, "step": 20120 }, { "grad_norm": 0.2075023502111435, "learning_rate": 9.387176930487169e-05, "loss": 0.0261, "step": 20130 }, { "grad_norm": 0.24277465045452118, "learning_rate": 9.386383529668072e-05, "loss": 0.028, "step": 20140 }, { "grad_norm": 0.3222155272960663, "learning_rate": 9.385589649160669e-05, "loss": 0.0367, "step": 20150 }, { "grad_norm": 0.245045006275177, "learning_rate": 9.384795289051775e-05, "loss": 0.0343, "step": 20160 }, { "grad_norm": 0.2740139961242676, "learning_rate": 9.384000449428261e-05, "loss": 0.0352, "step": 20170 }, { "grad_norm": 0.28787246346473694, "learning_rate": 9.383205130377048e-05, "loss": 0.0313, "step": 20180 }, { "grad_norm": 0.3848835229873657, "learning_rate": 9.382409331985114e-05, "loss": 0.0315, "step": 20190 }, { "grad_norm": 0.1743897646665573, "learning_rate": 9.381613054339482e-05, "loss": 0.0272, "step": 20200 }, { "grad_norm": 0.2891896963119507, "learning_rate": 9.380816297527235e-05, "loss": 0.0284, "step": 20210 }, { "grad_norm": 0.16661189496517181, "learning_rate": 9.380019061635506e-05, "loss": 0.0245, "step": 20220 }, { "grad_norm": 0.1884041428565979, "learning_rate": 9.379221346751474e-05, "loss": 0.0276, "step": 20230 }, { "grad_norm": 0.2756652534008026, "learning_rate": 9.378423152962382e-05, "loss": 0.0274, "step": 20240 }, { "grad_norm": 0.2601723372936249, "learning_rate": 9.377624480355517e-05, "loss": 0.0302, "step": 20250 }, { "grad_norm": 0.23613330721855164, "learning_rate": 9.376825329018219e-05, "loss": 0.0292, "step": 20260 }, { "grad_norm": 0.22646789252758026, "learning_rate": 9.376025699037884e-05, "loss": 0.0315, "step": 20270 }, { "grad_norm": 0.25366610288619995, "learning_rate": 9.37522559050196e-05, "loss": 0.0282, "step": 20280 }, { "grad_norm": 0.2458350509405136, "learning_rate": 9.37442500349794e-05, "loss": 0.0282, "step": 20290 }, { "grad_norm": 0.45263397693634033, "learning_rate": 9.373623938113381e-05, "loss": 0.0257, "step": 20300 }, { "grad_norm": 0.23932114243507385, "learning_rate": 9.372822394435883e-05, "loss": 0.0301, "step": 20310 }, { "grad_norm": 0.24077391624450684, "learning_rate": 9.372020372553102e-05, "loss": 0.0285, "step": 20320 }, { "grad_norm": 0.2997088134288788, "learning_rate": 9.371217872552746e-05, "loss": 0.0323, "step": 20330 }, { "grad_norm": 0.24006538093090057, "learning_rate": 9.370414894522576e-05, "loss": 0.0269, "step": 20340 }, { "grad_norm": 0.2977958619594574, "learning_rate": 9.369611438550406e-05, "loss": 0.0323, "step": 20350 }, { "grad_norm": 0.2651129364967346, "learning_rate": 9.368807504724095e-05, "loss": 0.0332, "step": 20360 }, { "grad_norm": 0.2531901001930237, "learning_rate": 9.368003093131565e-05, "loss": 0.0273, "step": 20370 }, { "grad_norm": 0.3362269401550293, "learning_rate": 9.367198203860785e-05, "loss": 0.0325, "step": 20380 }, { "grad_norm": 0.22045928239822388, "learning_rate": 9.366392836999774e-05, "loss": 0.0306, "step": 20390 }, { "grad_norm": 0.23647597432136536, "learning_rate": 9.365586992636607e-05, "loss": 0.0267, "step": 20400 }, { "grad_norm": 0.30338242650032043, "learning_rate": 9.364780670859412e-05, "loss": 0.0293, "step": 20410 }, { "grad_norm": 0.22910559177398682, "learning_rate": 9.363973871756364e-05, "loss": 0.0353, "step": 20420 }, { "grad_norm": 0.3489348590373993, "learning_rate": 9.363166595415696e-05, "loss": 0.0383, "step": 20430 }, { "grad_norm": 0.22966039180755615, "learning_rate": 9.362358841925686e-05, "loss": 0.0281, "step": 20440 }, { "grad_norm": 0.24378380179405212, "learning_rate": 9.361550611374674e-05, "loss": 0.0281, "step": 20450 }, { "grad_norm": 0.21705546975135803, "learning_rate": 9.360741903851043e-05, "loss": 0.0281, "step": 20460 }, { "grad_norm": 0.2279907464981079, "learning_rate": 9.359932719443236e-05, "loss": 0.0253, "step": 20470 }, { "grad_norm": 0.2410992830991745, "learning_rate": 9.35912305823974e-05, "loss": 0.0282, "step": 20480 }, { "grad_norm": 0.23715552687644958, "learning_rate": 9.358312920329101e-05, "loss": 0.0268, "step": 20490 }, { "grad_norm": 0.2885739505290985, "learning_rate": 9.357502305799914e-05, "loss": 0.042, "step": 20500 }, { "grad_norm": 0.2070334255695343, "learning_rate": 9.356691214740824e-05, "loss": 0.033, "step": 20510 }, { "grad_norm": 0.2853637933731079, "learning_rate": 9.355879647240535e-05, "loss": 0.0407, "step": 20520 }, { "grad_norm": 0.23484690487384796, "learning_rate": 9.355067603387798e-05, "loss": 0.0332, "step": 20530 }, { "grad_norm": 0.3275039792060852, "learning_rate": 9.354255083271412e-05, "loss": 0.0327, "step": 20540 }, { "grad_norm": 0.3554699718952179, "learning_rate": 9.353442086980239e-05, "loss": 0.0278, "step": 20550 }, { "grad_norm": 0.25123050808906555, "learning_rate": 9.352628614603185e-05, "loss": 0.0295, "step": 20560 }, { "grad_norm": 0.21755309402942657, "learning_rate": 9.351814666229209e-05, "loss": 0.0248, "step": 20570 }, { "grad_norm": 0.2611139714717865, "learning_rate": 9.351000241947324e-05, "loss": 0.0284, "step": 20580 }, { "grad_norm": 0.2522139549255371, "learning_rate": 9.350185341846594e-05, "loss": 0.0303, "step": 20590 }, { "grad_norm": 0.2451811283826828, "learning_rate": 9.349369966016134e-05, "loss": 0.0314, "step": 20600 }, { "grad_norm": 0.22928419709205627, "learning_rate": 9.348554114545117e-05, "loss": 0.0288, "step": 20610 }, { "grad_norm": 0.2548484206199646, "learning_rate": 9.347737787522758e-05, "loss": 0.0279, "step": 20620 }, { "grad_norm": 0.285766065120697, "learning_rate": 9.346920985038332e-05, "loss": 0.0308, "step": 20630 }, { "grad_norm": 0.30679851770401, "learning_rate": 9.346103707181162e-05, "loss": 0.0255, "step": 20640 }, { "grad_norm": 0.25324103236198425, "learning_rate": 9.345285954040626e-05, "loss": 0.0356, "step": 20650 }, { "grad_norm": 0.28301191329956055, "learning_rate": 9.34446772570615e-05, "loss": 0.029, "step": 20660 }, { "grad_norm": 0.24778367578983307, "learning_rate": 9.343649022267214e-05, "loss": 0.0275, "step": 20670 }, { "grad_norm": 0.28601086139678955, "learning_rate": 9.342829843813353e-05, "loss": 0.0255, "step": 20680 }, { "grad_norm": 0.22985364496707916, "learning_rate": 9.342010190434149e-05, "loss": 0.0277, "step": 20690 }, { "grad_norm": 0.287691205739975, "learning_rate": 9.34119006221924e-05, "loss": 0.0356, "step": 20700 }, { "grad_norm": 0.1963147222995758, "learning_rate": 9.340369459258313e-05, "loss": 0.0282, "step": 20710 }, { "grad_norm": 0.2833441197872162, "learning_rate": 9.339548381641106e-05, "loss": 0.0312, "step": 20720 }, { "grad_norm": 0.20875386893749237, "learning_rate": 9.338726829457413e-05, "loss": 0.025, "step": 20730 }, { "grad_norm": 0.227056622505188, "learning_rate": 9.337904802797078e-05, "loss": 0.028, "step": 20740 }, { "grad_norm": 0.2745210826396942, "learning_rate": 9.337082301749993e-05, "loss": 0.0273, "step": 20750 }, { "grad_norm": 0.30886536836624146, "learning_rate": 9.336259326406109e-05, "loss": 0.0295, "step": 20760 }, { "grad_norm": 0.3441995680332184, "learning_rate": 9.335435876855427e-05, "loss": 0.0293, "step": 20770 }, { "grad_norm": 0.24040578305721283, "learning_rate": 9.334611953187994e-05, "loss": 0.0321, "step": 20780 }, { "grad_norm": 0.24748852849006653, "learning_rate": 9.333787555493914e-05, "loss": 0.0281, "step": 20790 }, { "grad_norm": 0.3288484811782837, "learning_rate": 9.332962683863345e-05, "loss": 0.0284, "step": 20800 }, { "grad_norm": 0.23681162297725677, "learning_rate": 9.332137338386489e-05, "loss": 0.0268, "step": 20810 }, { "grad_norm": 0.2739810347557068, "learning_rate": 9.33131151915361e-05, "loss": 0.0319, "step": 20820 }, { "grad_norm": 0.23164398968219757, "learning_rate": 9.330485226255012e-05, "loss": 0.0245, "step": 20830 }, { "grad_norm": 0.17980913817882538, "learning_rate": 9.329658459781061e-05, "loss": 0.0255, "step": 20840 }, { "grad_norm": 0.24895994365215302, "learning_rate": 9.328831219822172e-05, "loss": 0.027, "step": 20850 }, { "grad_norm": 0.3084070384502411, "learning_rate": 9.328003506468808e-05, "loss": 0.0353, "step": 20860 }, { "grad_norm": 0.2515881061553955, "learning_rate": 9.327175319811488e-05, "loss": 0.0245, "step": 20870 }, { "grad_norm": 0.30325475335121155, "learning_rate": 9.326346659940781e-05, "loss": 0.0308, "step": 20880 }, { "grad_norm": 0.2653599977493286, "learning_rate": 9.325517526947308e-05, "loss": 0.0268, "step": 20890 }, { "grad_norm": 0.27211499214172363, "learning_rate": 9.32468792092174e-05, "loss": 0.0263, "step": 20900 }, { "grad_norm": 0.2530007064342499, "learning_rate": 9.323857841954803e-05, "loss": 0.0265, "step": 20910 }, { "grad_norm": 0.2860144376754761, "learning_rate": 9.323027290137276e-05, "loss": 0.0325, "step": 20920 }, { "grad_norm": 0.3459654450416565, "learning_rate": 9.322196265559981e-05, "loss": 0.0274, "step": 20930 }, { "grad_norm": 0.37810951471328735, "learning_rate": 9.321364768313803e-05, "loss": 0.0315, "step": 20940 }, { "grad_norm": 0.20093218982219696, "learning_rate": 9.32053279848967e-05, "loss": 0.0319, "step": 20950 }, { "grad_norm": 0.23770244419574738, "learning_rate": 9.319700356178567e-05, "loss": 0.032, "step": 20960 }, { "grad_norm": 0.20442110300064087, "learning_rate": 9.318867441471527e-05, "loss": 0.0273, "step": 20970 }, { "grad_norm": 0.2657901346683502, "learning_rate": 9.318034054459637e-05, "loss": 0.0298, "step": 20980 }, { "grad_norm": 0.2531452178955078, "learning_rate": 9.317200195234034e-05, "loss": 0.0297, "step": 20990 }, { "grad_norm": 0.270180881023407, "learning_rate": 9.316365863885909e-05, "loss": 0.0277, "step": 21000 }, { "grad_norm": 0.3200964629650116, "learning_rate": 9.315531060506502e-05, "loss": 0.0323, "step": 21010 }, { "grad_norm": 0.2735401690006256, "learning_rate": 9.314695785187108e-05, "loss": 0.0316, "step": 21020 }, { "grad_norm": 0.302649587392807, "learning_rate": 9.313860038019069e-05, "loss": 0.0331, "step": 21030 }, { "grad_norm": 0.29765981435775757, "learning_rate": 9.313023819093782e-05, "loss": 0.0334, "step": 21040 }, { "grad_norm": 0.3434896767139435, "learning_rate": 9.312187128502695e-05, "loss": 0.0338, "step": 21050 }, { "grad_norm": 0.3022531569004059, "learning_rate": 9.311349966337307e-05, "loss": 0.0315, "step": 21060 }, { "grad_norm": 0.335083544254303, "learning_rate": 9.310512332689169e-05, "loss": 0.0346, "step": 21070 }, { "grad_norm": 0.2736041843891144, "learning_rate": 9.309674227649883e-05, "loss": 0.0313, "step": 21080 }, { "grad_norm": 0.29500794410705566, "learning_rate": 9.308835651311103e-05, "loss": 0.0347, "step": 21090 }, { "grad_norm": 0.28301912546157837, "learning_rate": 9.307996603764533e-05, "loss": 0.0312, "step": 21100 }, { "grad_norm": 0.33758223056793213, "learning_rate": 9.307157085101932e-05, "loss": 0.0315, "step": 21110 }, { "grad_norm": 0.25783562660217285, "learning_rate": 9.306317095415109e-05, "loss": 0.0313, "step": 21120 }, { "grad_norm": 0.25222161412239075, "learning_rate": 9.305476634795922e-05, "loss": 0.0372, "step": 21130 }, { "grad_norm": 0.24614877998828888, "learning_rate": 9.304635703336284e-05, "loss": 0.0266, "step": 21140 }, { "grad_norm": 0.2574382722377777, "learning_rate": 9.303794301128157e-05, "loss": 0.0273, "step": 21150 }, { "grad_norm": 0.20372308790683746, "learning_rate": 9.302952428263555e-05, "loss": 0.0259, "step": 21160 }, { "grad_norm": 0.2852362096309662, "learning_rate": 9.302110084834545e-05, "loss": 0.0336, "step": 21170 }, { "grad_norm": 0.19171345233917236, "learning_rate": 9.301267270933245e-05, "loss": 0.0262, "step": 21180 }, { "grad_norm": 0.28693145513534546, "learning_rate": 9.300423986651823e-05, "loss": 0.031, "step": 21190 }, { "grad_norm": 0.3227458894252777, "learning_rate": 9.299580232082501e-05, "loss": 0.0276, "step": 21200 }, { "grad_norm": 0.2325742393732071, "learning_rate": 9.298736007317547e-05, "loss": 0.0329, "step": 21210 }, { "grad_norm": 0.17734570801258087, "learning_rate": 9.297891312449288e-05, "loss": 0.0278, "step": 21220 }, { "grad_norm": 0.25510600209236145, "learning_rate": 9.297046147570094e-05, "loss": 0.0297, "step": 21230 }, { "grad_norm": 0.22938436269760132, "learning_rate": 9.296200512772396e-05, "loss": 0.0313, "step": 21240 }, { "grad_norm": 0.28318989276885986, "learning_rate": 9.295354408148668e-05, "loss": 0.0243, "step": 21250 }, { "grad_norm": 0.2561521530151367, "learning_rate": 9.294507833791441e-05, "loss": 0.0285, "step": 21260 }, { "grad_norm": 0.3837357759475708, "learning_rate": 9.293660789793295e-05, "loss": 0.0289, "step": 21270 }, { "grad_norm": 0.34072044491767883, "learning_rate": 9.292813276246858e-05, "loss": 0.035, "step": 21280 }, { "grad_norm": 0.2758781909942627, "learning_rate": 9.291965293244816e-05, "loss": 0.0317, "step": 21290 }, { "grad_norm": 0.28722164034843445, "learning_rate": 9.291116840879904e-05, "loss": 0.0292, "step": 21300 }, { "grad_norm": 0.3062759041786194, "learning_rate": 9.290267919244904e-05, "loss": 0.0331, "step": 21310 }, { "grad_norm": 0.2988819479942322, "learning_rate": 9.289418528432655e-05, "loss": 0.0285, "step": 21320 }, { "grad_norm": 0.27300122380256653, "learning_rate": 9.288568668536045e-05, "loss": 0.0292, "step": 21330 }, { "grad_norm": 0.26179859042167664, "learning_rate": 9.287718339648013e-05, "loss": 0.0243, "step": 21340 }, { "grad_norm": 0.20109528303146362, "learning_rate": 9.28686754186155e-05, "loss": 0.0246, "step": 21350 }, { "grad_norm": 0.21241959929466248, "learning_rate": 9.286016275269698e-05, "loss": 0.0219, "step": 21360 }, { "grad_norm": 0.18808789551258087, "learning_rate": 9.285164539965551e-05, "loss": 0.0241, "step": 21370 }, { "grad_norm": 0.21779119968414307, "learning_rate": 9.284312336042251e-05, "loss": 0.0316, "step": 21380 }, { "grad_norm": 0.22215895354747772, "learning_rate": 9.283459663592996e-05, "loss": 0.0293, "step": 21390 }, { "grad_norm": 0.2762089967727661, "learning_rate": 9.282606522711033e-05, "loss": 0.0326, "step": 21400 }, { "grad_norm": 0.29810017347335815, "learning_rate": 9.281752913489657e-05, "loss": 0.0285, "step": 21410 }, { "grad_norm": 0.27075809240341187, "learning_rate": 9.280898836022222e-05, "loss": 0.026, "step": 21420 }, { "grad_norm": 0.25352153182029724, "learning_rate": 9.280044290402126e-05, "loss": 0.0332, "step": 21430 }, { "grad_norm": 0.20105375349521637, "learning_rate": 9.279189276722821e-05, "loss": 0.0308, "step": 21440 }, { "grad_norm": 0.22613924741744995, "learning_rate": 9.278333795077812e-05, "loss": 0.0265, "step": 21450 }, { "grad_norm": 0.3341808319091797, "learning_rate": 9.27747784556065e-05, "loss": 0.0302, "step": 21460 }, { "grad_norm": 0.2711848318576813, "learning_rate": 9.276621428264942e-05, "loss": 0.0239, "step": 21470 }, { "grad_norm": 0.24297328293323517, "learning_rate": 9.275764543284345e-05, "loss": 0.028, "step": 21480 }, { "grad_norm": 0.31773868203163147, "learning_rate": 9.274907190712566e-05, "loss": 0.0262, "step": 21490 }, { "grad_norm": 0.3839193284511566, "learning_rate": 9.274049370643363e-05, "loss": 0.0301, "step": 21500 }, { "grad_norm": 0.30279698967933655, "learning_rate": 9.273191083170547e-05, "loss": 0.0283, "step": 21510 }, { "grad_norm": 0.20327363908290863, "learning_rate": 9.27233232838798e-05, "loss": 0.0294, "step": 21520 }, { "grad_norm": 0.2805038392543793, "learning_rate": 9.27147310638957e-05, "loss": 0.0259, "step": 21530 }, { "grad_norm": 0.3389394283294678, "learning_rate": 9.270613417269286e-05, "loss": 0.0306, "step": 21540 }, { "grad_norm": 0.2464696764945984, "learning_rate": 9.269753261121138e-05, "loss": 0.0262, "step": 21550 }, { "grad_norm": 0.24404004216194153, "learning_rate": 9.268892638039194e-05, "loss": 0.0239, "step": 21560 }, { "grad_norm": 0.230153888463974, "learning_rate": 9.268031548117569e-05, "loss": 0.0309, "step": 21570 }, { "grad_norm": 0.35612860321998596, "learning_rate": 9.26716999145043e-05, "loss": 0.0304, "step": 21580 }, { "grad_norm": 0.3605886697769165, "learning_rate": 9.266307968131998e-05, "loss": 0.0242, "step": 21590 }, { "grad_norm": 0.21823805570602417, "learning_rate": 9.26544547825654e-05, "loss": 0.033, "step": 21600 }, { "grad_norm": 0.20871353149414062, "learning_rate": 9.264582521918376e-05, "loss": 0.0285, "step": 21610 }, { "grad_norm": 0.23010556399822235, "learning_rate": 9.263719099211881e-05, "loss": 0.0228, "step": 21620 }, { "grad_norm": 0.24129793047904968, "learning_rate": 9.262855210231476e-05, "loss": 0.0269, "step": 21630 }, { "grad_norm": 0.22370518743991852, "learning_rate": 9.261990855071633e-05, "loss": 0.0271, "step": 21640 }, { "grad_norm": 0.35314878821372986, "learning_rate": 9.261126033826878e-05, "loss": 0.0274, "step": 21650 }, { "grad_norm": 0.19329074025154114, "learning_rate": 9.260260746591786e-05, "loss": 0.0303, "step": 21660 }, { "grad_norm": 0.3115291893482208, "learning_rate": 9.259394993460985e-05, "loss": 0.0276, "step": 21670 }, { "grad_norm": 0.28319868445396423, "learning_rate": 9.258528774529151e-05, "loss": 0.0321, "step": 21680 }, { "grad_norm": 0.24083228409290314, "learning_rate": 9.257662089891013e-05, "loss": 0.0306, "step": 21690 }, { "grad_norm": 0.21828609704971313, "learning_rate": 9.25679493964135e-05, "loss": 0.0219, "step": 21700 }, { "grad_norm": 0.2572146952152252, "learning_rate": 9.255927323874994e-05, "loss": 0.0321, "step": 21710 }, { "grad_norm": 0.2304946631193161, "learning_rate": 9.255059242686822e-05, "loss": 0.0281, "step": 21720 }, { "grad_norm": 0.26548388600349426, "learning_rate": 9.254190696171769e-05, "loss": 0.0276, "step": 21730 }, { "grad_norm": 0.31882888078689575, "learning_rate": 9.25332168442482e-05, "loss": 0.0306, "step": 21740 }, { "grad_norm": 0.27736517786979675, "learning_rate": 9.252452207541004e-05, "loss": 0.0289, "step": 21750 }, { "grad_norm": 0.30901896953582764, "learning_rate": 9.251582265615409e-05, "loss": 0.0313, "step": 21760 }, { "grad_norm": 0.2817701995372772, "learning_rate": 9.250711858743169e-05, "loss": 0.0298, "step": 21770 }, { "grad_norm": 0.298562228679657, "learning_rate": 9.24984098701947e-05, "loss": 0.0338, "step": 21780 }, { "grad_norm": 0.20683850347995758, "learning_rate": 9.248969650539552e-05, "loss": 0.0233, "step": 21790 }, { "grad_norm": 0.21428583562374115, "learning_rate": 9.2480978493987e-05, "loss": 0.0269, "step": 21800 }, { "grad_norm": 0.22271977365016937, "learning_rate": 9.247225583692256e-05, "loss": 0.0293, "step": 21810 }, { "grad_norm": 0.36668992042541504, "learning_rate": 9.246352853515607e-05, "loss": 0.0298, "step": 21820 }, { "grad_norm": 0.20362582802772522, "learning_rate": 9.245479658964194e-05, "loss": 0.0244, "step": 21830 }, { "grad_norm": 0.28010183572769165, "learning_rate": 9.244606000133507e-05, "loss": 0.0296, "step": 21840 }, { "grad_norm": 0.23770859837532043, "learning_rate": 9.24373187711909e-05, "loss": 0.0271, "step": 21850 }, { "grad_norm": 0.28314539790153503, "learning_rate": 9.242857290016537e-05, "loss": 0.0276, "step": 21860 }, { "grad_norm": 0.2409338504076004, "learning_rate": 9.241982238921488e-05, "loss": 0.0243, "step": 21870 }, { "grad_norm": 0.2687981128692627, "learning_rate": 9.24110672392964e-05, "loss": 0.0276, "step": 21880 }, { "grad_norm": 0.2047022581100464, "learning_rate": 9.240230745136737e-05, "loss": 0.0221, "step": 21890 }, { "grad_norm": 0.24030283093452454, "learning_rate": 9.239354302638575e-05, "loss": 0.0241, "step": 21900 }, { "grad_norm": 0.2928178906440735, "learning_rate": 9.238477396531e-05, "loss": 0.023, "step": 21910 }, { "grad_norm": 0.2894355058670044, "learning_rate": 9.23760002690991e-05, "loss": 0.0269, "step": 21920 }, { "grad_norm": 0.27117571234703064, "learning_rate": 9.236722193871252e-05, "loss": 0.0245, "step": 21930 }, { "grad_norm": 0.23666146397590637, "learning_rate": 9.235843897511023e-05, "loss": 0.0216, "step": 21940 }, { "grad_norm": 0.2131749391555786, "learning_rate": 9.234965137925276e-05, "loss": 0.0292, "step": 21950 }, { "grad_norm": 0.23372182250022888, "learning_rate": 9.234085915210108e-05, "loss": 0.0265, "step": 21960 }, { "grad_norm": 0.28385379910469055, "learning_rate": 9.23320622946167e-05, "loss": 0.0274, "step": 21970 }, { "grad_norm": 0.28301241993904114, "learning_rate": 9.232326080776163e-05, "loss": 0.0269, "step": 21980 }, { "grad_norm": 0.2169645130634308, "learning_rate": 9.23144546924984e-05, "loss": 0.0286, "step": 21990 }, { "grad_norm": 0.2572104334831238, "learning_rate": 9.230564394979e-05, "loss": 0.0282, "step": 22000 }, { "grad_norm": 0.22296392917633057, "learning_rate": 9.22968285806e-05, "loss": 0.0242, "step": 22010 }, { "grad_norm": 0.22255714237689972, "learning_rate": 9.228800858589242e-05, "loss": 0.0279, "step": 22020 }, { "grad_norm": 0.27792513370513916, "learning_rate": 9.227918396663179e-05, "loss": 0.0271, "step": 22030 }, { "grad_norm": 0.256059855222702, "learning_rate": 9.227035472378319e-05, "loss": 0.0265, "step": 22040 }, { "grad_norm": 0.2216314822435379, "learning_rate": 9.226152085831213e-05, "loss": 0.0283, "step": 22050 }, { "grad_norm": 0.22939911484718323, "learning_rate": 9.22526823711847e-05, "loss": 0.0307, "step": 22060 }, { "grad_norm": 0.2330818921327591, "learning_rate": 9.224383926336745e-05, "loss": 0.0259, "step": 22070 }, { "grad_norm": 0.2400883585214615, "learning_rate": 9.223499153582744e-05, "loss": 0.025, "step": 22080 }, { "grad_norm": 0.22946162521839142, "learning_rate": 9.222613918953226e-05, "loss": 0.0267, "step": 22090 }, { "grad_norm": 0.26231878995895386, "learning_rate": 9.221728222544999e-05, "loss": 0.0264, "step": 22100 }, { "grad_norm": 0.24164974689483643, "learning_rate": 9.22084206445492e-05, "loss": 0.0307, "step": 22110 }, { "grad_norm": 0.2937091588973999, "learning_rate": 9.2199554447799e-05, "loss": 0.0242, "step": 22120 }, { "grad_norm": 0.2771403193473816, "learning_rate": 9.219068363616897e-05, "loss": 0.0258, "step": 22130 }, { "grad_norm": 0.2906996011734009, "learning_rate": 9.218180821062919e-05, "loss": 0.0257, "step": 22140 }, { "grad_norm": 0.23925648629665375, "learning_rate": 9.21729281721503e-05, "loss": 0.0266, "step": 22150 }, { "grad_norm": 0.2558637857437134, "learning_rate": 9.216404352170339e-05, "loss": 0.032, "step": 22160 }, { "grad_norm": 0.19619205594062805, "learning_rate": 9.215515426026007e-05, "loss": 0.0266, "step": 22170 }, { "grad_norm": 0.18450886011123657, "learning_rate": 9.214626038879246e-05, "loss": 0.0238, "step": 22180 }, { "grad_norm": 0.21216057240962982, "learning_rate": 9.21373619082732e-05, "loss": 0.0227, "step": 22190 }, { "grad_norm": 0.24806684255599976, "learning_rate": 9.212845881967535e-05, "loss": 0.0258, "step": 22200 }, { "grad_norm": 0.20194201171398163, "learning_rate": 9.211955112397262e-05, "loss": 0.0256, "step": 22210 }, { "grad_norm": 0.2287381887435913, "learning_rate": 9.211063882213909e-05, "loss": 0.024, "step": 22220 }, { "grad_norm": 0.343309223651886, "learning_rate": 9.210172191514942e-05, "loss": 0.0271, "step": 22230 }, { "grad_norm": 0.23478065431118011, "learning_rate": 9.209280040397874e-05, "loss": 0.0247, "step": 22240 }, { "grad_norm": 0.48683568835258484, "learning_rate": 9.208387428960268e-05, "loss": 0.0257, "step": 22250 }, { "grad_norm": 0.18700595200061798, "learning_rate": 9.20749435729974e-05, "loss": 0.0256, "step": 22260 }, { "grad_norm": 0.229469433426857, "learning_rate": 9.206600825513957e-05, "loss": 0.0333, "step": 22270 }, { "grad_norm": 0.23496443033218384, "learning_rate": 9.20570683370063e-05, "loss": 0.0252, "step": 22280 }, { "grad_norm": 0.34916871786117554, "learning_rate": 9.204812381957528e-05, "loss": 0.0324, "step": 22290 }, { "grad_norm": 0.2094596028327942, "learning_rate": 9.203917470382465e-05, "loss": 0.026, "step": 22300 }, { "grad_norm": 0.24922508001327515, "learning_rate": 9.203022099073309e-05, "loss": 0.0216, "step": 22310 }, { "grad_norm": 0.2438092678785324, "learning_rate": 9.202126268127976e-05, "loss": 0.0278, "step": 22320 }, { "grad_norm": 0.25304949283599854, "learning_rate": 9.20122997764443e-05, "loss": 0.0317, "step": 22330 }, { "grad_norm": 0.34641245007514954, "learning_rate": 9.200333227720692e-05, "loss": 0.0323, "step": 22340 }, { "grad_norm": 0.36443448066711426, "learning_rate": 9.199436018454826e-05, "loss": 0.0363, "step": 22350 }, { "grad_norm": 0.274880588054657, "learning_rate": 9.198538349944952e-05, "loss": 0.0264, "step": 22360 }, { "grad_norm": 0.256760835647583, "learning_rate": 9.197640222289234e-05, "loss": 0.0262, "step": 22370 }, { "grad_norm": 0.3113648593425751, "learning_rate": 9.196741635585895e-05, "loss": 0.0306, "step": 22380 }, { "grad_norm": 0.3006621301174164, "learning_rate": 9.195842589933199e-05, "loss": 0.0249, "step": 22390 }, { "grad_norm": 0.20360282063484192, "learning_rate": 9.194943085429466e-05, "loss": 0.0289, "step": 22400 }, { "grad_norm": 0.24004991352558136, "learning_rate": 9.194043122173065e-05, "loss": 0.0311, "step": 22410 }, { "grad_norm": 0.23262476921081543, "learning_rate": 9.193142700262413e-05, "loss": 0.0245, "step": 22420 }, { "grad_norm": 0.21977612376213074, "learning_rate": 9.192241819795979e-05, "loss": 0.0243, "step": 22430 }, { "grad_norm": 0.3107401132583618, "learning_rate": 9.191340480872284e-05, "loss": 0.027, "step": 22440 }, { "grad_norm": 0.21085889637470245, "learning_rate": 9.190438683589895e-05, "loss": 0.0252, "step": 22450 }, { "grad_norm": 0.27728503942489624, "learning_rate": 9.189536428047432e-05, "loss": 0.0272, "step": 22460 }, { "grad_norm": 0.2509302794933319, "learning_rate": 9.188633714343564e-05, "loss": 0.0251, "step": 22470 }, { "grad_norm": 0.2559298574924469, "learning_rate": 9.18773054257701e-05, "loss": 0.0244, "step": 22480 }, { "grad_norm": 0.2733699083328247, "learning_rate": 9.18682691284654e-05, "loss": 0.0255, "step": 22490 }, { "grad_norm": 0.3113766312599182, "learning_rate": 9.185922825250974e-05, "loss": 0.0277, "step": 22500 }, { "grad_norm": 0.26873543858528137, "learning_rate": 9.185018279889181e-05, "loss": 0.0298, "step": 22510 }, { "grad_norm": 0.3207017481327057, "learning_rate": 9.184113276860082e-05, "loss": 0.0324, "step": 22520 }, { "grad_norm": 0.19374069571495056, "learning_rate": 9.183207816262645e-05, "loss": 0.0303, "step": 22530 }, { "grad_norm": 0.24782098829746246, "learning_rate": 9.182301898195891e-05, "loss": 0.0324, "step": 22540 }, { "grad_norm": 0.26053857803344727, "learning_rate": 9.181395522758889e-05, "loss": 0.0235, "step": 22550 }, { "grad_norm": 0.23393678665161133, "learning_rate": 9.180488690050759e-05, "loss": 0.0269, "step": 22560 }, { "grad_norm": 0.2254163771867752, "learning_rate": 9.179581400170671e-05, "loss": 0.0263, "step": 22570 }, { "grad_norm": 0.2555786371231079, "learning_rate": 9.178673653217845e-05, "loss": 0.027, "step": 22580 }, { "grad_norm": 0.15707966685295105, "learning_rate": 9.177765449291551e-05, "loss": 0.0263, "step": 22590 }, { "grad_norm": 0.3781054615974426, "learning_rate": 9.176856788491109e-05, "loss": 0.0263, "step": 22600 }, { "grad_norm": 0.22095812857151031, "learning_rate": 9.175947670915887e-05, "loss": 0.0272, "step": 22610 }, { "grad_norm": 0.2616788148880005, "learning_rate": 9.175038096665309e-05, "loss": 0.0297, "step": 22620 }, { "grad_norm": 0.23140321671962738, "learning_rate": 9.17412806583884e-05, "loss": 0.0264, "step": 22630 }, { "grad_norm": 0.24061758816242218, "learning_rate": 9.173217578536002e-05, "loss": 0.0267, "step": 22640 }, { "grad_norm": 0.25724413990974426, "learning_rate": 9.172306634856362e-05, "loss": 0.0294, "step": 22650 }, { "grad_norm": 0.21493706107139587, "learning_rate": 9.171395234899545e-05, "loss": 0.0274, "step": 22660 }, { "grad_norm": 0.27092066407203674, "learning_rate": 9.170483378765214e-05, "loss": 0.0298, "step": 22670 }, { "grad_norm": 0.4448724091053009, "learning_rate": 9.169571066553091e-05, "loss": 0.0355, "step": 22680 }, { "grad_norm": 0.252501517534256, "learning_rate": 9.168658298362946e-05, "loss": 0.0328, "step": 22690 }, { "grad_norm": 0.23062755167484283, "learning_rate": 9.167745074294598e-05, "loss": 0.0336, "step": 22700 }, { "grad_norm": 0.29357239603996277, "learning_rate": 9.166831394447913e-05, "loss": 0.0267, "step": 22710 }, { "grad_norm": 0.24703501164913177, "learning_rate": 9.165917258922812e-05, "loss": 0.029, "step": 22720 }, { "grad_norm": 0.27975955605506897, "learning_rate": 9.165002667819262e-05, "loss": 0.0247, "step": 22730 }, { "grad_norm": 0.2320530116558075, "learning_rate": 9.164087621237282e-05, "loss": 0.0292, "step": 22740 }, { "grad_norm": 0.28063246607780457, "learning_rate": 9.163172119276942e-05, "loss": 0.0283, "step": 22750 }, { "grad_norm": 0.3450367748737335, "learning_rate": 9.162256162038358e-05, "loss": 0.0336, "step": 22760 }, { "grad_norm": 0.23709982633590698, "learning_rate": 9.161339749621698e-05, "loss": 0.033, "step": 22770 }, { "grad_norm": 0.20926204323768616, "learning_rate": 9.160422882127177e-05, "loss": 0.0238, "step": 22780 }, { "grad_norm": 0.3056942820549011, "learning_rate": 9.159505559655069e-05, "loss": 0.0286, "step": 22790 }, { "grad_norm": 0.24207523465156555, "learning_rate": 9.158587782305684e-05, "loss": 0.0273, "step": 22800 }, { "grad_norm": 0.24670644104480743, "learning_rate": 9.157669550179391e-05, "loss": 0.0249, "step": 22810 }, { "grad_norm": 0.2683890163898468, "learning_rate": 9.156750863376609e-05, "loss": 0.0319, "step": 22820 }, { "grad_norm": 0.2483503669500351, "learning_rate": 9.155831721997801e-05, "loss": 0.0312, "step": 22830 }, { "grad_norm": 0.29606983065605164, "learning_rate": 9.154912126143484e-05, "loss": 0.0367, "step": 22840 }, { "grad_norm": 0.26137465238571167, "learning_rate": 9.153992075914224e-05, "loss": 0.0363, "step": 22850 }, { "grad_norm": 0.22524084150791168, "learning_rate": 9.153071571410635e-05, "loss": 0.0256, "step": 22860 }, { "grad_norm": 0.2538131773471832, "learning_rate": 9.152150612733384e-05, "loss": 0.0286, "step": 22870 }, { "grad_norm": 0.33308011293411255, "learning_rate": 9.151229199983184e-05, "loss": 0.0317, "step": 22880 }, { "grad_norm": 0.22307926416397095, "learning_rate": 9.150307333260802e-05, "loss": 0.0252, "step": 22890 }, { "grad_norm": 0.25370851159095764, "learning_rate": 9.149385012667048e-05, "loss": 0.0261, "step": 22900 }, { "grad_norm": 0.3230663537979126, "learning_rate": 9.148462238302788e-05, "loss": 0.0304, "step": 22910 }, { "grad_norm": 0.2515901029109955, "learning_rate": 9.147539010268936e-05, "loss": 0.026, "step": 22920 }, { "grad_norm": 0.2617718577384949, "learning_rate": 9.14661532866645e-05, "loss": 0.0249, "step": 22930 }, { "grad_norm": 0.2516167461872101, "learning_rate": 9.145691193596348e-05, "loss": 0.0292, "step": 22940 }, { "grad_norm": 0.20932695269584656, "learning_rate": 9.144766605159691e-05, "loss": 0.0278, "step": 22950 }, { "grad_norm": 0.2341499775648117, "learning_rate": 9.14384156345759e-05, "loss": 0.0223, "step": 22960 }, { "grad_norm": 0.21461725234985352, "learning_rate": 9.142916068591204e-05, "loss": 0.027, "step": 22970 }, { "grad_norm": 0.24125435948371887, "learning_rate": 9.141990120661746e-05, "loss": 0.0236, "step": 22980 }, { "grad_norm": 0.26324108242988586, "learning_rate": 9.141063719770475e-05, "loss": 0.0317, "step": 22990 }, { "grad_norm": 0.2628893554210663, "learning_rate": 9.140136866018704e-05, "loss": 0.0243, "step": 23000 }, { "grad_norm": 0.22836489975452423, "learning_rate": 9.139209559507788e-05, "loss": 0.0261, "step": 23010 }, { "grad_norm": 0.2626473903656006, "learning_rate": 9.13828180033914e-05, "loss": 0.0262, "step": 23020 }, { "grad_norm": 0.2539047598838806, "learning_rate": 9.137353588614212e-05, "loss": 0.0282, "step": 23030 }, { "grad_norm": 0.2910510301589966, "learning_rate": 9.136424924434519e-05, "loss": 0.0309, "step": 23040 }, { "grad_norm": 0.27852940559387207, "learning_rate": 9.135495807901615e-05, "loss": 0.0255, "step": 23050 }, { "grad_norm": 0.30510494112968445, "learning_rate": 9.134566239117108e-05, "loss": 0.0311, "step": 23060 }, { "grad_norm": 0.23155243694782257, "learning_rate": 9.13363621818265e-05, "loss": 0.0241, "step": 23070 }, { "grad_norm": 0.29252856969833374, "learning_rate": 9.132705745199953e-05, "loss": 0.0321, "step": 23080 }, { "grad_norm": 0.2382681667804718, "learning_rate": 9.131774820270768e-05, "loss": 0.0197, "step": 23090 }, { "grad_norm": 0.22560983896255493, "learning_rate": 9.130843443496901e-05, "loss": 0.0268, "step": 23100 }, { "grad_norm": 0.24820132553577423, "learning_rate": 9.129911614980206e-05, "loss": 0.0301, "step": 23110 }, { "grad_norm": 0.19387660920619965, "learning_rate": 9.128979334822584e-05, "loss": 0.0249, "step": 23120 }, { "grad_norm": 0.30269473791122437, "learning_rate": 9.128046603125992e-05, "loss": 0.0324, "step": 23130 }, { "grad_norm": 0.2737903892993927, "learning_rate": 9.12711341999243e-05, "loss": 0.0283, "step": 23140 }, { "grad_norm": 0.23269252479076385, "learning_rate": 9.12617978552395e-05, "loss": 0.0228, "step": 23150 }, { "grad_norm": 0.23127174377441406, "learning_rate": 9.12524569982265e-05, "loss": 0.024, "step": 23160 }, { "grad_norm": 0.21736262738704681, "learning_rate": 9.124311162990684e-05, "loss": 0.0239, "step": 23170 }, { "grad_norm": 0.2193843424320221, "learning_rate": 9.12337617513025e-05, "loss": 0.0217, "step": 23180 }, { "grad_norm": 0.2957910895347595, "learning_rate": 9.122440736343596e-05, "loss": 0.0213, "step": 23190 }, { "grad_norm": 0.28509631752967834, "learning_rate": 9.12150484673302e-05, "loss": 0.021, "step": 23200 }, { "grad_norm": 0.21144233644008636, "learning_rate": 9.120568506400873e-05, "loss": 0.0252, "step": 23210 }, { "grad_norm": 0.2754725515842438, "learning_rate": 9.119631715449548e-05, "loss": 0.0247, "step": 23220 }, { "grad_norm": 0.32592180371284485, "learning_rate": 9.118694473981493e-05, "loss": 0.0281, "step": 23230 }, { "grad_norm": 0.26887190341949463, "learning_rate": 9.117756782099203e-05, "loss": 0.0273, "step": 23240 }, { "grad_norm": 0.2707168459892273, "learning_rate": 9.11681863990522e-05, "loss": 0.0218, "step": 23250 }, { "grad_norm": 0.2517959475517273, "learning_rate": 9.115880047502142e-05, "loss": 0.0229, "step": 23260 }, { "grad_norm": 0.2031855285167694, "learning_rate": 9.114941004992609e-05, "loss": 0.0229, "step": 23270 }, { "grad_norm": 0.20693863928318024, "learning_rate": 9.114001512479317e-05, "loss": 0.0255, "step": 23280 }, { "grad_norm": 0.217667818069458, "learning_rate": 9.113061570065003e-05, "loss": 0.0244, "step": 23290 }, { "grad_norm": 0.2533888816833496, "learning_rate": 9.112121177852459e-05, "loss": 0.0253, "step": 23300 }, { "grad_norm": 0.2606852650642395, "learning_rate": 9.111180335944527e-05, "loss": 0.0252, "step": 23310 }, { "grad_norm": 0.25601938366889954, "learning_rate": 9.110239044444093e-05, "loss": 0.023, "step": 23320 }, { "grad_norm": 0.1881880909204483, "learning_rate": 9.109297303454099e-05, "loss": 0.0201, "step": 23330 }, { "grad_norm": 0.22487632930278778, "learning_rate": 9.108355113077526e-05, "loss": 0.0233, "step": 23340 }, { "grad_norm": 0.26319196820259094, "learning_rate": 9.107412473417419e-05, "loss": 0.023, "step": 23350 }, { "grad_norm": 0.18185974657535553, "learning_rate": 9.106469384576858e-05, "loss": 0.0254, "step": 23360 }, { "grad_norm": 0.21466083824634552, "learning_rate": 9.105525846658978e-05, "loss": 0.023, "step": 23370 }, { "grad_norm": 0.31709063053131104, "learning_rate": 9.104581859766965e-05, "loss": 0.0338, "step": 23380 }, { "grad_norm": 0.26641154289245605, "learning_rate": 9.10363742400405e-05, "loss": 0.0285, "step": 23390 }, { "grad_norm": 0.25156041979789734, "learning_rate": 9.102692539473518e-05, "loss": 0.0231, "step": 23400 }, { "grad_norm": 0.2498154640197754, "learning_rate": 9.101747206278697e-05, "loss": 0.0275, "step": 23410 }, { "grad_norm": 0.23739352822303772, "learning_rate": 9.100801424522968e-05, "loss": 0.0265, "step": 23420 }, { "grad_norm": 0.19586971402168274, "learning_rate": 9.099855194309762e-05, "loss": 0.0272, "step": 23430 }, { "grad_norm": 0.2996586263179779, "learning_rate": 9.098908515742554e-05, "loss": 0.0277, "step": 23440 }, { "grad_norm": 0.20836077630519867, "learning_rate": 9.097961388924873e-05, "loss": 0.0225, "step": 23450 }, { "grad_norm": 0.253978431224823, "learning_rate": 9.097013813960298e-05, "loss": 0.0227, "step": 23460 }, { "grad_norm": 0.2020399421453476, "learning_rate": 9.09606579095245e-05, "loss": 0.0226, "step": 23470 }, { "grad_norm": 0.2750151753425598, "learning_rate": 9.095117320005008e-05, "loss": 0.026, "step": 23480 }, { "grad_norm": 0.20786796510219574, "learning_rate": 9.094168401221691e-05, "loss": 0.0241, "step": 23490 }, { "grad_norm": 0.16243582963943481, "learning_rate": 9.093219034706273e-05, "loss": 0.0246, "step": 23500 }, { "grad_norm": 0.236239954829216, "learning_rate": 9.092269220562577e-05, "loss": 0.0248, "step": 23510 }, { "grad_norm": 0.18124382197856903, "learning_rate": 9.09131895889447e-05, "loss": 0.0263, "step": 23520 }, { "grad_norm": 0.2094765603542328, "learning_rate": 9.090368249805873e-05, "loss": 0.0233, "step": 23530 }, { "grad_norm": 0.3811764419078827, "learning_rate": 9.089417093400754e-05, "loss": 0.026, "step": 23540 }, { "grad_norm": 0.27044495940208435, "learning_rate": 9.088465489783131e-05, "loss": 0.0278, "step": 23550 }, { "grad_norm": 0.1922287493944168, "learning_rate": 9.087513439057068e-05, "loss": 0.0313, "step": 23560 }, { "grad_norm": 0.2608116567134857, "learning_rate": 9.08656094132668e-05, "loss": 0.0253, "step": 23570 }, { "grad_norm": 0.23142565786838531, "learning_rate": 9.085607996696134e-05, "loss": 0.0245, "step": 23580 }, { "grad_norm": 0.2605428993701935, "learning_rate": 9.084654605269639e-05, "loss": 0.028, "step": 23590 }, { "grad_norm": 0.23743607103824615, "learning_rate": 9.083700767151457e-05, "loss": 0.0313, "step": 23600 }, { "grad_norm": 0.24712641537189484, "learning_rate": 9.082746482445898e-05, "loss": 0.0238, "step": 23610 }, { "grad_norm": 0.2153167575597763, "learning_rate": 9.081791751257325e-05, "loss": 0.0212, "step": 23620 }, { "grad_norm": 0.22472263872623444, "learning_rate": 9.080836573690142e-05, "loss": 0.0289, "step": 23630 }, { "grad_norm": 0.2539375424385071, "learning_rate": 9.079880949848805e-05, "loss": 0.032, "step": 23640 }, { "grad_norm": 0.2273697555065155, "learning_rate": 9.078924879837822e-05, "loss": 0.0266, "step": 23650 }, { "grad_norm": 0.26218485832214355, "learning_rate": 9.077968363761747e-05, "loss": 0.0303, "step": 23660 }, { "grad_norm": 0.2129841446876526, "learning_rate": 9.077011401725182e-05, "loss": 0.0282, "step": 23670 }, { "grad_norm": 0.33236992359161377, "learning_rate": 9.07605399383278e-05, "loss": 0.0319, "step": 23680 }, { "grad_norm": 0.3351726830005646, "learning_rate": 9.075096140189243e-05, "loss": 0.0355, "step": 23690 }, { "grad_norm": 0.511349618434906, "learning_rate": 9.074137840899318e-05, "loss": 0.0215, "step": 23700 }, { "grad_norm": 0.24803727865219116, "learning_rate": 9.073179096067804e-05, "loss": 0.0344, "step": 23710 }, { "grad_norm": 0.2549594044685364, "learning_rate": 9.072219905799549e-05, "loss": 0.0271, "step": 23720 }, { "grad_norm": 0.28718647360801697, "learning_rate": 9.071260270199447e-05, "loss": 0.0355, "step": 23730 }, { "grad_norm": 0.29090797901153564, "learning_rate": 9.070300189372441e-05, "loss": 0.0312, "step": 23740 }, { "grad_norm": 0.31039613485336304, "learning_rate": 9.069339663423528e-05, "loss": 0.0268, "step": 23750 }, { "grad_norm": 0.2844294607639313, "learning_rate": 9.068378692457747e-05, "loss": 0.026, "step": 23760 }, { "grad_norm": 0.2771057188510895, "learning_rate": 9.067417276580189e-05, "loss": 0.0244, "step": 23770 }, { "grad_norm": 0.24190041422843933, "learning_rate": 9.066455415895993e-05, "loss": 0.0271, "step": 23780 }, { "grad_norm": 0.2895223796367645, "learning_rate": 9.065493110510346e-05, "loss": 0.0309, "step": 23790 }, { "grad_norm": 0.3130161464214325, "learning_rate": 9.064530360528484e-05, "loss": 0.0401, "step": 23800 }, { "grad_norm": 0.21180909872055054, "learning_rate": 9.063567166055695e-05, "loss": 0.0275, "step": 23810 }, { "grad_norm": 0.2344846874475479, "learning_rate": 9.062603527197308e-05, "loss": 0.0261, "step": 23820 }, { "grad_norm": 0.1687796711921692, "learning_rate": 9.06163944405871e-05, "loss": 0.0247, "step": 23830 }, { "grad_norm": 0.2857750654220581, "learning_rate": 9.060674916745327e-05, "loss": 0.0299, "step": 23840 }, { "grad_norm": 0.22744202613830566, "learning_rate": 9.05970994536264e-05, "loss": 0.0254, "step": 23850 }, { "grad_norm": 0.2618624269962311, "learning_rate": 9.05874453001618e-05, "loss": 0.0282, "step": 23860 }, { "grad_norm": 0.22950157523155212, "learning_rate": 9.057778670811517e-05, "loss": 0.0237, "step": 23870 }, { "grad_norm": 0.19770920276641846, "learning_rate": 9.056812367854281e-05, "loss": 0.0239, "step": 23880 }, { "grad_norm": 0.3476702570915222, "learning_rate": 9.055845621250143e-05, "loss": 0.0328, "step": 23890 }, { "grad_norm": 0.1976310908794403, "learning_rate": 9.054878431104825e-05, "loss": 0.0298, "step": 23900 }, { "grad_norm": 0.25654730200767517, "learning_rate": 9.0539107975241e-05, "loss": 0.0298, "step": 23910 }, { "grad_norm": 0.22665461897850037, "learning_rate": 9.052942720613784e-05, "loss": 0.0274, "step": 23920 }, { "grad_norm": 0.23570038378238678, "learning_rate": 9.051974200479745e-05, "loss": 0.0269, "step": 23930 }, { "grad_norm": 0.28088003396987915, "learning_rate": 9.051005237227901e-05, "loss": 0.0266, "step": 23940 }, { "grad_norm": 0.23495441675186157, "learning_rate": 9.050035830964215e-05, "loss": 0.0274, "step": 23950 }, { "grad_norm": 0.2076943814754486, "learning_rate": 9.049065981794698e-05, "loss": 0.0337, "step": 23960 }, { "grad_norm": 0.2284659445285797, "learning_rate": 9.048095689825414e-05, "loss": 0.027, "step": 23970 }, { "grad_norm": 0.2901720106601715, "learning_rate": 9.047124955162472e-05, "loss": 0.0304, "step": 23980 }, { "grad_norm": 0.2260178178548813, "learning_rate": 9.046153777912028e-05, "loss": 0.0285, "step": 23990 }, { "grad_norm": 0.25037872791290283, "learning_rate": 9.045182158180292e-05, "loss": 0.0275, "step": 24000 }, { "grad_norm": 0.2379276007413864, "learning_rate": 9.044210096073516e-05, "loss": 0.028, "step": 24010 }, { "grad_norm": 0.2747325599193573, "learning_rate": 9.043237591698004e-05, "loss": 0.0271, "step": 24020 }, { "grad_norm": 0.29457908868789673, "learning_rate": 9.04226464516011e-05, "loss": 0.0252, "step": 24030 }, { "grad_norm": 0.245807945728302, "learning_rate": 9.041291256566229e-05, "loss": 0.0286, "step": 24040 }, { "grad_norm": 0.2012534886598587, "learning_rate": 9.040317426022814e-05, "loss": 0.0251, "step": 24050 }, { "grad_norm": 0.22570011019706726, "learning_rate": 9.03934315363636e-05, "loss": 0.0297, "step": 24060 }, { "grad_norm": 0.2029499113559723, "learning_rate": 9.038368439513409e-05, "loss": 0.0283, "step": 24070 }, { "grad_norm": 0.22179140150547028, "learning_rate": 9.03739328376056e-05, "loss": 0.0283, "step": 24080 }, { "grad_norm": 0.2907954156398773, "learning_rate": 9.036417686484451e-05, "loss": 0.0245, "step": 24090 }, { "grad_norm": 0.20993083715438843, "learning_rate": 9.035441647791773e-05, "loss": 0.024, "step": 24100 }, { "grad_norm": 0.23788012564182281, "learning_rate": 9.034465167789263e-05, "loss": 0.0277, "step": 24110 }, { "grad_norm": 0.16656653583049774, "learning_rate": 9.033488246583706e-05, "loss": 0.0193, "step": 24120 }, { "grad_norm": 0.2581877112388611, "learning_rate": 9.032510884281941e-05, "loss": 0.022, "step": 24130 }, { "grad_norm": 0.17105522751808167, "learning_rate": 9.031533080990848e-05, "loss": 0.0274, "step": 24140 }, { "grad_norm": 0.21791386604309082, "learning_rate": 9.030554836817358e-05, "loss": 0.0259, "step": 24150 }, { "grad_norm": 0.273483544588089, "learning_rate": 9.029576151868451e-05, "loss": 0.0298, "step": 24160 }, { "grad_norm": 0.20481762290000916, "learning_rate": 9.028597026251155e-05, "loss": 0.0221, "step": 24170 }, { "grad_norm": 0.20681846141815186, "learning_rate": 9.027617460072547e-05, "loss": 0.0251, "step": 24180 }, { "grad_norm": 0.17026546597480774, "learning_rate": 9.026637453439745e-05, "loss": 0.03, "step": 24190 }, { "grad_norm": 0.23663939535617828, "learning_rate": 9.025657006459927e-05, "loss": 0.0209, "step": 24200 }, { "grad_norm": 0.20742586255073547, "learning_rate": 9.024676119240311e-05, "loss": 0.0272, "step": 24210 }, { "grad_norm": 0.1807885318994522, "learning_rate": 9.023694791888166e-05, "loss": 0.0222, "step": 24220 }, { "grad_norm": 0.18968547880649567, "learning_rate": 9.022713024510808e-05, "loss": 0.0231, "step": 24230 }, { "grad_norm": 0.22376592457294464, "learning_rate": 9.021730817215601e-05, "loss": 0.0281, "step": 24240 }, { "grad_norm": 0.2275800257921219, "learning_rate": 9.02074817010996e-05, "loss": 0.0297, "step": 24250 }, { "grad_norm": 0.304021954536438, "learning_rate": 9.019765083301342e-05, "loss": 0.0241, "step": 24260 }, { "grad_norm": 0.2689116597175598, "learning_rate": 9.01878155689726e-05, "loss": 0.0236, "step": 24270 }, { "grad_norm": 0.22346867620944977, "learning_rate": 9.017797591005268e-05, "loss": 0.023, "step": 24280 }, { "grad_norm": 0.246363565325737, "learning_rate": 9.016813185732972e-05, "loss": 0.0241, "step": 24290 }, { "grad_norm": 0.24212047457695007, "learning_rate": 9.015828341188027e-05, "loss": 0.0251, "step": 24300 }, { "grad_norm": 0.18594112992286682, "learning_rate": 9.01484305747813e-05, "loss": 0.0263, "step": 24310 }, { "grad_norm": 0.25416871905326843, "learning_rate": 9.013857334711033e-05, "loss": 0.0261, "step": 24320 }, { "grad_norm": 0.23088976740837097, "learning_rate": 9.012871172994534e-05, "loss": 0.0219, "step": 24330 }, { "grad_norm": 0.23394711315631866, "learning_rate": 9.011884572436476e-05, "loss": 0.0262, "step": 24340 }, { "grad_norm": 0.2553008794784546, "learning_rate": 9.010897533144754e-05, "loss": 0.0226, "step": 24350 }, { "grad_norm": 0.1802748590707779, "learning_rate": 9.009910055227306e-05, "loss": 0.0217, "step": 24360 }, { "grad_norm": 0.23627729713916779, "learning_rate": 9.008922138792124e-05, "loss": 0.0236, "step": 24370 }, { "grad_norm": 0.2970963716506958, "learning_rate": 9.007933783947244e-05, "loss": 0.0255, "step": 24380 }, { "grad_norm": 0.3157614469528198, "learning_rate": 9.006944990800752e-05, "loss": 0.029, "step": 24390 }, { "grad_norm": 0.25485357642173767, "learning_rate": 9.005955759460779e-05, "loss": 0.0291, "step": 24400 }, { "grad_norm": 0.2573937177658081, "learning_rate": 9.004966090035508e-05, "loss": 0.024, "step": 24410 }, { "grad_norm": 0.24521209299564362, "learning_rate": 9.003975982633166e-05, "loss": 0.0228, "step": 24420 }, { "grad_norm": 0.22740234434604645, "learning_rate": 9.00298543736203e-05, "loss": 0.0205, "step": 24430 }, { "grad_norm": 0.2113770991563797, "learning_rate": 9.001994454330427e-05, "loss": 0.0261, "step": 24440 }, { "grad_norm": 0.17987172305583954, "learning_rate": 9.001003033646727e-05, "loss": 0.0209, "step": 24450 }, { "grad_norm": 0.2173251211643219, "learning_rate": 9.00001117541935e-05, "loss": 0.0204, "step": 24460 }, { "grad_norm": 0.18342773616313934, "learning_rate": 8.999018879756764e-05, "loss": 0.0191, "step": 24470 }, { "grad_norm": 0.29334190487861633, "learning_rate": 8.998026146767487e-05, "loss": 0.0253, "step": 24480 }, { "grad_norm": 0.2469930499792099, "learning_rate": 8.99703297656008e-05, "loss": 0.0262, "step": 24490 }, { "grad_norm": 0.2363681197166443, "learning_rate": 8.996039369243156e-05, "loss": 0.027, "step": 24500 }, { "grad_norm": 0.20209050178527832, "learning_rate": 8.995045324925378e-05, "loss": 0.0221, "step": 24510 }, { "grad_norm": 0.2874454855918884, "learning_rate": 8.994050843715448e-05, "loss": 0.0263, "step": 24520 }, { "grad_norm": 0.25697362422943115, "learning_rate": 8.993055925722121e-05, "loss": 0.0297, "step": 24530 }, { "grad_norm": 0.2308020293712616, "learning_rate": 8.992060571054202e-05, "loss": 0.0267, "step": 24540 }, { "grad_norm": 0.24851515889167786, "learning_rate": 8.991064779820542e-05, "loss": 0.0248, "step": 24550 }, { "grad_norm": 0.285663902759552, "learning_rate": 8.990068552130036e-05, "loss": 0.0278, "step": 24560 }, { "grad_norm": 0.2720175087451935, "learning_rate": 8.989071888091634e-05, "loss": 0.0258, "step": 24570 }, { "grad_norm": 0.21361660957336426, "learning_rate": 8.988074787814329e-05, "loss": 0.0267, "step": 24580 }, { "grad_norm": 0.2871790826320648, "learning_rate": 8.987077251407158e-05, "loss": 0.0285, "step": 24590 }, { "grad_norm": 0.4446776807308197, "learning_rate": 8.986079278979216e-05, "loss": 0.0244, "step": 24600 }, { "grad_norm": 0.21046806871891022, "learning_rate": 8.985080870639635e-05, "loss": 0.0296, "step": 24610 }, { "grad_norm": 0.3329572081565857, "learning_rate": 8.984082026497603e-05, "loss": 0.0264, "step": 24620 }, { "grad_norm": 0.17687901854515076, "learning_rate": 8.98308274666235e-05, "loss": 0.0231, "step": 24630 }, { "grad_norm": 0.2132536917924881, "learning_rate": 8.982083031243155e-05, "loss": 0.0247, "step": 24640 }, { "grad_norm": 0.25235840678215027, "learning_rate": 8.98108288034935e-05, "loss": 0.0247, "step": 24650 }, { "grad_norm": 0.21253110468387604, "learning_rate": 8.980082294090305e-05, "loss": 0.023, "step": 24660 }, { "grad_norm": 0.22789892554283142, "learning_rate": 8.979081272575443e-05, "loss": 0.0219, "step": 24670 }, { "grad_norm": 0.28951939940452576, "learning_rate": 8.978079815914236e-05, "loss": 0.0302, "step": 24680 }, { "grad_norm": 0.27546706795692444, "learning_rate": 8.977077924216202e-05, "loss": 0.0267, "step": 24690 }, { "grad_norm": 0.3566857576370239, "learning_rate": 8.976075597590905e-05, "loss": 0.0276, "step": 24700 }, { "grad_norm": 0.28047239780426025, "learning_rate": 8.975072836147958e-05, "loss": 0.0279, "step": 24710 }, { "grad_norm": 0.23141801357269287, "learning_rate": 8.974069639997025e-05, "loss": 0.0316, "step": 24720 }, { "grad_norm": 0.17441432178020477, "learning_rate": 8.973066009247808e-05, "loss": 0.0252, "step": 24730 }, { "grad_norm": 0.24658116698265076, "learning_rate": 8.972061944010066e-05, "loss": 0.028, "step": 24740 }, { "grad_norm": 0.16921839118003845, "learning_rate": 8.971057444393603e-05, "loss": 0.0262, "step": 24750 }, { "grad_norm": 0.2127569317817688, "learning_rate": 8.970052510508268e-05, "loss": 0.0229, "step": 24760 }, { "grad_norm": 0.25576433539390564, "learning_rate": 8.969047142463959e-05, "loss": 0.0202, "step": 24770 }, { "grad_norm": 0.2122911661863327, "learning_rate": 8.968041340370621e-05, "loss": 0.0253, "step": 24780 }, { "grad_norm": 0.34924808144569397, "learning_rate": 8.96703510433825e-05, "loss": 0.0308, "step": 24790 }, { "grad_norm": 0.3363281190395355, "learning_rate": 8.966028434476883e-05, "loss": 0.0292, "step": 24800 }, { "grad_norm": 0.23795074224472046, "learning_rate": 8.96502133089661e-05, "loss": 0.0228, "step": 24810 }, { "grad_norm": 0.23289769887924194, "learning_rate": 8.964013793707564e-05, "loss": 0.0258, "step": 24820 }, { "grad_norm": 0.2758236825466156, "learning_rate": 8.963005823019932e-05, "loss": 0.0242, "step": 24830 }, { "grad_norm": 0.2695900797843933, "learning_rate": 8.961997418943939e-05, "loss": 0.029, "step": 24840 }, { "grad_norm": 0.2837776839733124, "learning_rate": 8.960988581589865e-05, "loss": 0.0216, "step": 24850 }, { "grad_norm": 0.3033598065376282, "learning_rate": 8.959979311068037e-05, "loss": 0.0242, "step": 24860 }, { "grad_norm": 0.2781634032726288, "learning_rate": 8.958969607488823e-05, "loss": 0.0217, "step": 24870 }, { "grad_norm": 0.21695587038993835, "learning_rate": 8.957959470962647e-05, "loss": 0.0243, "step": 24880 }, { "grad_norm": 0.23863258957862854, "learning_rate": 8.956948901599971e-05, "loss": 0.0311, "step": 24890 }, { "grad_norm": 0.2573741674423218, "learning_rate": 8.955937899511315e-05, "loss": 0.0309, "step": 24900 }, { "grad_norm": 0.28494521975517273, "learning_rate": 8.954926464807238e-05, "loss": 0.0252, "step": 24910 }, { "grad_norm": 0.2660313844680786, "learning_rate": 8.953914597598347e-05, "loss": 0.0236, "step": 24920 }, { "grad_norm": 0.20255768299102783, "learning_rate": 8.952902297995303e-05, "loss": 0.0266, "step": 24930 }, { "grad_norm": 0.22789622843265533, "learning_rate": 8.951889566108804e-05, "loss": 0.03, "step": 24940 }, { "grad_norm": 0.3345245122909546, "learning_rate": 8.950876402049606e-05, "loss": 0.0251, "step": 24950 }, { "grad_norm": 0.22005820274353027, "learning_rate": 8.949862805928504e-05, "loss": 0.0222, "step": 24960 }, { "grad_norm": 0.20951120555400848, "learning_rate": 8.948848777856343e-05, "loss": 0.0217, "step": 24970 }, { "grad_norm": 0.30564072728157043, "learning_rate": 8.947834317944017e-05, "loss": 0.025, "step": 24980 }, { "grad_norm": 0.3307311534881592, "learning_rate": 8.946819426302466e-05, "loss": 0.0231, "step": 24990 }, { "grad_norm": 0.21730396151542664, "learning_rate": 8.945804103042676e-05, "loss": 0.0246, "step": 25000 }, { "grad_norm": 0.282703161239624, "learning_rate": 8.944788348275681e-05, "loss": 0.0267, "step": 25010 }, { "grad_norm": 0.34060075879096985, "learning_rate": 8.943772162112565e-05, "loss": 0.0263, "step": 25020 }, { "grad_norm": 0.2982841432094574, "learning_rate": 8.942755544664454e-05, "loss": 0.0234, "step": 25030 }, { "grad_norm": 0.22359764575958252, "learning_rate": 8.941738496042525e-05, "loss": 0.0266, "step": 25040 }, { "grad_norm": 0.23346765339374542, "learning_rate": 8.940721016357999e-05, "loss": 0.025, "step": 25050 }, { "grad_norm": 0.2624254524707794, "learning_rate": 8.939703105722148e-05, "loss": 0.0309, "step": 25060 }, { "grad_norm": 0.2623257040977478, "learning_rate": 8.93868476424629e-05, "loss": 0.0266, "step": 25070 }, { "grad_norm": 0.24591311812400818, "learning_rate": 8.937665992041786e-05, "loss": 0.0252, "step": 25080 }, { "grad_norm": 0.26630252599716187, "learning_rate": 8.93664678922005e-05, "loss": 0.0291, "step": 25090 }, { "grad_norm": 0.30315425992012024, "learning_rate": 8.93562715589254e-05, "loss": 0.026, "step": 25100 }, { "grad_norm": 0.26476791501045227, "learning_rate": 8.934607092170762e-05, "loss": 0.025, "step": 25110 }, { "grad_norm": 0.23763899505138397, "learning_rate": 8.933586598166266e-05, "loss": 0.0238, "step": 25120 }, { "grad_norm": 0.2494320422410965, "learning_rate": 8.932565673990655e-05, "loss": 0.0249, "step": 25130 }, { "grad_norm": 0.24660532176494598, "learning_rate": 8.931544319755574e-05, "loss": 0.0244, "step": 25140 }, { "grad_norm": 0.24401596188545227, "learning_rate": 8.930522535572718e-05, "loss": 0.0234, "step": 25150 }, { "grad_norm": 0.28530240058898926, "learning_rate": 8.929500321553826e-05, "loss": 0.0298, "step": 25160 }, { "grad_norm": 0.25303027033805847, "learning_rate": 8.928477677810686e-05, "loss": 0.0261, "step": 25170 }, { "grad_norm": 0.2823924422264099, "learning_rate": 8.927454604455137e-05, "loss": 0.026, "step": 25180 }, { "grad_norm": 0.27933457493782043, "learning_rate": 8.926431101599053e-05, "loss": 0.0253, "step": 25190 }, { "grad_norm": 0.2715589106082916, "learning_rate": 8.925407169354369e-05, "loss": 0.027, "step": 25200 }, { "grad_norm": 0.27941516041755676, "learning_rate": 8.92438280783306e-05, "loss": 0.0226, "step": 25210 }, { "grad_norm": 0.2369280308485031, "learning_rate": 8.923358017147146e-05, "loss": 0.0353, "step": 25220 }, { "grad_norm": 0.22982047498226166, "learning_rate": 8.922332797408697e-05, "loss": 0.0306, "step": 25230 }, { "grad_norm": 0.2467857301235199, "learning_rate": 8.921307148729831e-05, "loss": 0.0259, "step": 25240 }, { "grad_norm": 0.2687373757362366, "learning_rate": 8.920281071222712e-05, "loss": 0.0263, "step": 25250 }, { "grad_norm": 0.24645832180976868, "learning_rate": 8.919254564999548e-05, "loss": 0.0221, "step": 25260 }, { "grad_norm": 0.24778765439987183, "learning_rate": 8.918227630172598e-05, "loss": 0.0221, "step": 25270 }, { "grad_norm": 0.26608312129974365, "learning_rate": 8.917200266854165e-05, "loss": 0.0256, "step": 25280 }, { "grad_norm": 0.3493694067001343, "learning_rate": 8.9161724751566e-05, "loss": 0.0261, "step": 25290 }, { "grad_norm": 0.19717159867286682, "learning_rate": 8.915144255192302e-05, "loss": 0.0213, "step": 25300 }, { "grad_norm": 0.2988671064376831, "learning_rate": 8.914115607073714e-05, "loss": 0.0215, "step": 25310 }, { "grad_norm": 0.22219203412532806, "learning_rate": 8.913086530913327e-05, "loss": 0.0197, "step": 25320 }, { "grad_norm": 0.25469252467155457, "learning_rate": 8.912057026823681e-05, "loss": 0.0237, "step": 25330 }, { "grad_norm": 0.1932705044746399, "learning_rate": 8.91102709491736e-05, "loss": 0.0227, "step": 25340 }, { "grad_norm": 0.3132217526435852, "learning_rate": 8.909996735306996e-05, "loss": 0.0216, "step": 25350 }, { "grad_norm": 0.2421899139881134, "learning_rate": 8.908965948105268e-05, "loss": 0.0235, "step": 25360 }, { "grad_norm": 0.23323597013950348, "learning_rate": 8.907934733424901e-05, "loss": 0.0286, "step": 25370 }, { "grad_norm": 0.24472205340862274, "learning_rate": 8.906903091378666e-05, "loss": 0.0237, "step": 25380 }, { "grad_norm": 0.2545009255409241, "learning_rate": 8.905871022079384e-05, "loss": 0.0257, "step": 25390 }, { "grad_norm": 0.23308369517326355, "learning_rate": 8.90483852563992e-05, "loss": 0.0202, "step": 25400 }, { "grad_norm": 0.22153256833553314, "learning_rate": 8.903805602173185e-05, "loss": 0.0223, "step": 25410 }, { "grad_norm": 0.2456432282924652, "learning_rate": 8.902772251792137e-05, "loss": 0.0234, "step": 25420 }, { "grad_norm": 0.2697489261627197, "learning_rate": 8.901738474609786e-05, "loss": 0.0234, "step": 25430 }, { "grad_norm": 0.25266698002815247, "learning_rate": 8.900704270739179e-05, "loss": 0.0206, "step": 25440 }, { "grad_norm": 0.21673695743083954, "learning_rate": 8.89966964029342e-05, "loss": 0.0237, "step": 25450 }, { "grad_norm": 0.19278112053871155, "learning_rate": 8.898634583385652e-05, "loss": 0.0247, "step": 25460 }, { "grad_norm": 0.2141353040933609, "learning_rate": 8.897599100129065e-05, "loss": 0.0257, "step": 25470 }, { "grad_norm": 0.188384011387825, "learning_rate": 8.896563190636903e-05, "loss": 0.0241, "step": 25480 }, { "grad_norm": 0.20619705319404602, "learning_rate": 8.895526855022448e-05, "loss": 0.0244, "step": 25490 }, { "grad_norm": 0.21024765074253082, "learning_rate": 8.894490093399033e-05, "loss": 0.0238, "step": 25500 }, { "grad_norm": 0.21740907430648804, "learning_rate": 8.893452905880035e-05, "loss": 0.0267, "step": 25510 }, { "grad_norm": 0.23590593039989471, "learning_rate": 8.892415292578883e-05, "loss": 0.0237, "step": 25520 }, { "grad_norm": 0.2820433974266052, "learning_rate": 8.891377253609046e-05, "loss": 0.0265, "step": 25530 }, { "grad_norm": 0.3136062026023865, "learning_rate": 8.890338789084043e-05, "loss": 0.032, "step": 25540 }, { "grad_norm": 0.27933552861213684, "learning_rate": 8.88929989911744e-05, "loss": 0.0291, "step": 25550 }, { "grad_norm": 0.2601495087146759, "learning_rate": 8.888260583822847e-05, "loss": 0.0322, "step": 25560 }, { "grad_norm": 0.24353712797164917, "learning_rate": 8.887220843313921e-05, "loss": 0.0234, "step": 25570 }, { "grad_norm": 0.26681944727897644, "learning_rate": 8.88618067770437e-05, "loss": 0.0233, "step": 25580 }, { "grad_norm": 0.195066899061203, "learning_rate": 8.885140087107942e-05, "loss": 0.0254, "step": 25590 }, { "grad_norm": 0.23650459945201874, "learning_rate": 8.884099071638436e-05, "loss": 0.0235, "step": 25600 }, { "grad_norm": 0.2861698865890503, "learning_rate": 8.883057631409695e-05, "loss": 0.0218, "step": 25610 }, { "grad_norm": 0.28606531023979187, "learning_rate": 8.882015766535608e-05, "loss": 0.0225, "step": 25620 }, { "grad_norm": 0.24262982606887817, "learning_rate": 8.880973477130115e-05, "loss": 0.0278, "step": 25630 }, { "grad_norm": 0.3085731267929077, "learning_rate": 8.879930763307197e-05, "loss": 0.0271, "step": 25640 }, { "grad_norm": 0.2528150677680969, "learning_rate": 8.878887625180884e-05, "loss": 0.0218, "step": 25650 }, { "grad_norm": 0.23500101268291473, "learning_rate": 8.877844062865253e-05, "loss": 0.0262, "step": 25660 }, { "grad_norm": 0.2191394567489624, "learning_rate": 8.876800076474424e-05, "loss": 0.0239, "step": 25670 }, { "grad_norm": 0.24451003968715668, "learning_rate": 8.875755666122568e-05, "loss": 0.0219, "step": 25680 }, { "grad_norm": 0.22906582057476044, "learning_rate": 8.8747108319239e-05, "loss": 0.0211, "step": 25690 }, { "grad_norm": 0.25431007146835327, "learning_rate": 8.87366557399268e-05, "loss": 0.0249, "step": 25700 }, { "grad_norm": 0.26867765188217163, "learning_rate": 8.872619892443217e-05, "loss": 0.0309, "step": 25710 }, { "grad_norm": 0.30928459763526917, "learning_rate": 8.871573787389865e-05, "loss": 0.025, "step": 25720 }, { "grad_norm": 0.20036479830741882, "learning_rate": 8.870527258947024e-05, "loss": 0.0213, "step": 25730 }, { "grad_norm": 0.24387750029563904, "learning_rate": 8.869480307229143e-05, "loss": 0.0258, "step": 25740 }, { "grad_norm": 0.2681901752948761, "learning_rate": 8.868432932350712e-05, "loss": 0.0239, "step": 25750 }, { "grad_norm": 0.259796679019928, "learning_rate": 8.867385134426272e-05, "loss": 0.024, "step": 25760 }, { "grad_norm": 0.23856650292873383, "learning_rate": 8.866336913570407e-05, "loss": 0.0236, "step": 25770 }, { "grad_norm": 0.2412078082561493, "learning_rate": 8.865288269897751e-05, "loss": 0.0236, "step": 25780 }, { "grad_norm": 0.21179413795471191, "learning_rate": 8.864239203522981e-05, "loss": 0.0229, "step": 25790 }, { "grad_norm": 0.24679060280323029, "learning_rate": 8.863189714560822e-05, "loss": 0.0261, "step": 25800 }, { "grad_norm": 0.2941243350505829, "learning_rate": 8.862139803126043e-05, "loss": 0.0234, "step": 25810 }, { "grad_norm": 0.22430895268917084, "learning_rate": 8.861089469333463e-05, "loss": 0.0199, "step": 25820 }, { "grad_norm": 0.2734968662261963, "learning_rate": 8.860038713297944e-05, "loss": 0.0241, "step": 25830 }, { "grad_norm": 0.20651094615459442, "learning_rate": 8.858987535134394e-05, "loss": 0.0249, "step": 25840 }, { "grad_norm": 0.30204665660858154, "learning_rate": 8.857935934957769e-05, "loss": 0.026, "step": 25850 }, { "grad_norm": 0.2248314917087555, "learning_rate": 8.856883912883071e-05, "loss": 0.0255, "step": 25860 }, { "grad_norm": 0.24197891354560852, "learning_rate": 8.855831469025346e-05, "loss": 0.0272, "step": 25870 }, { "grad_norm": 0.2786908447742462, "learning_rate": 8.854778603499689e-05, "loss": 0.028, "step": 25880 }, { "grad_norm": 0.2082033008337021, "learning_rate": 8.85372531642124e-05, "loss": 0.0216, "step": 25890 }, { "grad_norm": 0.2931627631187439, "learning_rate": 8.852671607905185e-05, "loss": 0.0237, "step": 25900 }, { "grad_norm": 0.30589625239372253, "learning_rate": 8.851617478066754e-05, "loss": 0.0251, "step": 25910 }, { "grad_norm": 0.2684299647808075, "learning_rate": 8.850562927021227e-05, "loss": 0.0271, "step": 25920 }, { "grad_norm": 0.21389488875865936, "learning_rate": 8.849507954883928e-05, "loss": 0.0254, "step": 25930 }, { "grad_norm": 0.21090154349803925, "learning_rate": 8.848452561770226e-05, "loss": 0.0213, "step": 25940 }, { "grad_norm": 0.2181868553161621, "learning_rate": 8.847396747795538e-05, "loss": 0.0225, "step": 25950 }, { "grad_norm": 0.2329837679862976, "learning_rate": 8.846340513075327e-05, "loss": 0.028, "step": 25960 }, { "grad_norm": 0.22816120088100433, "learning_rate": 8.845283857725099e-05, "loss": 0.0358, "step": 25970 }, { "grad_norm": 0.2367270141839981, "learning_rate": 8.844226781860409e-05, "loss": 0.0331, "step": 25980 }, { "grad_norm": 0.20058999955654144, "learning_rate": 8.84316928559686e-05, "loss": 0.0238, "step": 25990 }, { "grad_norm": 0.21960347890853882, "learning_rate": 8.842111369050094e-05, "loss": 0.0262, "step": 26000 }, { "grad_norm": 0.28998422622680664, "learning_rate": 8.841053032335808e-05, "loss": 0.0232, "step": 26010 }, { "grad_norm": 0.2506687045097351, "learning_rate": 8.839994275569735e-05, "loss": 0.0278, "step": 26020 }, { "grad_norm": 0.22620581090450287, "learning_rate": 8.838935098867662e-05, "loss": 0.0227, "step": 26030 }, { "grad_norm": 0.24225540459156036, "learning_rate": 8.837875502345418e-05, "loss": 0.029, "step": 26040 }, { "grad_norm": 0.24316103756427765, "learning_rate": 8.83681548611888e-05, "loss": 0.0223, "step": 26050 }, { "grad_norm": 0.3448779582977295, "learning_rate": 8.835755050303969e-05, "loss": 0.024, "step": 26060 }, { "grad_norm": 0.27910369634628296, "learning_rate": 8.834694195016653e-05, "loss": 0.0269, "step": 26070 }, { "grad_norm": 0.2483695149421692, "learning_rate": 8.833632920372942e-05, "loss": 0.0267, "step": 26080 }, { "grad_norm": 0.20611396431922913, "learning_rate": 8.832571226488903e-05, "loss": 0.0272, "step": 26090 }, { "grad_norm": 0.2282133549451828, "learning_rate": 8.831509113480634e-05, "loss": 0.0241, "step": 26100 }, { "grad_norm": 0.24176564812660217, "learning_rate": 8.83044658146429e-05, "loss": 0.0232, "step": 26110 }, { "grad_norm": 0.2232224941253662, "learning_rate": 8.829383630556067e-05, "loss": 0.0206, "step": 26120 }, { "grad_norm": 0.22383709251880646, "learning_rate": 8.828320260872207e-05, "loss": 0.0201, "step": 26130 }, { "grad_norm": 0.21579158306121826, "learning_rate": 8.827256472529e-05, "loss": 0.0265, "step": 26140 }, { "grad_norm": 0.20482347905635834, "learning_rate": 8.826192265642778e-05, "loss": 0.0243, "step": 26150 }, { "grad_norm": 0.31121963262557983, "learning_rate": 8.825127640329923e-05, "loss": 0.0245, "step": 26160 }, { "grad_norm": 0.24016845226287842, "learning_rate": 8.824062596706861e-05, "loss": 0.0249, "step": 26170 }, { "grad_norm": 0.2281734198331833, "learning_rate": 8.822997134890062e-05, "loss": 0.0244, "step": 26180 }, { "grad_norm": 0.21396346390247345, "learning_rate": 8.821931254996044e-05, "loss": 0.0182, "step": 26190 }, { "grad_norm": 0.24412478506565094, "learning_rate": 8.82086495714137e-05, "loss": 0.0222, "step": 26200 }, { "grad_norm": 0.21538734436035156, "learning_rate": 8.81979824144265e-05, "loss": 0.0236, "step": 26210 }, { "grad_norm": 0.2222774475812912, "learning_rate": 8.818731108016536e-05, "loss": 0.026, "step": 26220 }, { "grad_norm": 0.18479536473751068, "learning_rate": 8.81766355697973e-05, "loss": 0.0216, "step": 26230 }, { "grad_norm": 0.19251756370067596, "learning_rate": 8.816595588448977e-05, "loss": 0.0207, "step": 26240 }, { "grad_norm": 0.25651833415031433, "learning_rate": 8.81552720254107e-05, "loss": 0.0236, "step": 26250 }, { "grad_norm": 0.2751014232635498, "learning_rate": 8.814458399372842e-05, "loss": 0.0331, "step": 26260 }, { "grad_norm": 0.3503093123435974, "learning_rate": 8.813389179061181e-05, "loss": 0.0301, "step": 26270 }, { "grad_norm": 0.2659786641597748, "learning_rate": 8.812319541723012e-05, "loss": 0.0234, "step": 26280 }, { "grad_norm": 0.2733195722103119, "learning_rate": 8.811249487475309e-05, "loss": 0.025, "step": 26290 }, { "grad_norm": 0.32179075479507446, "learning_rate": 8.810179016435092e-05, "loss": 0.0281, "step": 26300 }, { "grad_norm": 0.26258257031440735, "learning_rate": 8.809108128719428e-05, "loss": 0.0241, "step": 26310 }, { "grad_norm": 0.32499369978904724, "learning_rate": 8.808036824445424e-05, "loss": 0.0229, "step": 26320 }, { "grad_norm": 0.22089718282222748, "learning_rate": 8.806965103730238e-05, "loss": 0.0255, "step": 26330 }, { "grad_norm": 0.19385473430156708, "learning_rate": 8.805892966691074e-05, "loss": 0.0272, "step": 26340 }, { "grad_norm": 0.21592071652412415, "learning_rate": 8.804820413445175e-05, "loss": 0.0255, "step": 26350 }, { "grad_norm": 0.37140148878097534, "learning_rate": 8.803747444109837e-05, "loss": 0.0253, "step": 26360 }, { "grad_norm": 0.24291981756687164, "learning_rate": 8.802674058802399e-05, "loss": 0.0291, "step": 26370 }, { "grad_norm": 0.18048563599586487, "learning_rate": 8.801600257640241e-05, "loss": 0.0223, "step": 26380 }, { "grad_norm": 0.22733117640018463, "learning_rate": 8.800526040740795e-05, "loss": 0.0243, "step": 26390 }, { "grad_norm": 0.22782883048057556, "learning_rate": 8.799451408221535e-05, "loss": 0.0221, "step": 26400 }, { "grad_norm": 0.27642926573753357, "learning_rate": 8.798376360199982e-05, "loss": 0.0256, "step": 26410 }, { "grad_norm": 0.21729695796966553, "learning_rate": 8.797300896793701e-05, "loss": 0.0204, "step": 26420 }, { "grad_norm": 0.26602670550346375, "learning_rate": 8.796225018120302e-05, "loss": 0.0262, "step": 26430 }, { "grad_norm": 0.24327418208122253, "learning_rate": 8.795148724297444e-05, "loss": 0.0245, "step": 26440 }, { "grad_norm": 0.22313405573368073, "learning_rate": 8.794072015442825e-05, "loss": 0.0299, "step": 26450 }, { "grad_norm": 0.22922196984291077, "learning_rate": 8.792994891674198e-05, "loss": 0.0241, "step": 26460 }, { "grad_norm": 0.29742148518562317, "learning_rate": 8.79191735310935e-05, "loss": 0.03, "step": 26470 }, { "grad_norm": 0.2359367460012436, "learning_rate": 8.790839399866122e-05, "loss": 0.0282, "step": 26480 }, { "grad_norm": 0.2512698173522949, "learning_rate": 8.789761032062397e-05, "loss": 0.0226, "step": 26490 }, { "grad_norm": 0.22327888011932373, "learning_rate": 8.788682249816103e-05, "loss": 0.0246, "step": 26500 }, { "grad_norm": 0.22686906158924103, "learning_rate": 8.787603053245215e-05, "loss": 0.0244, "step": 26510 }, { "grad_norm": 0.17620526254177094, "learning_rate": 8.78652344246775e-05, "loss": 0.0198, "step": 26520 }, { "grad_norm": 0.22139830887317657, "learning_rate": 8.785443417601776e-05, "loss": 0.0223, "step": 26530 }, { "grad_norm": 0.16238880157470703, "learning_rate": 8.784362978765401e-05, "loss": 0.0176, "step": 26540 }, { "grad_norm": 0.18946844339370728, "learning_rate": 8.783282126076779e-05, "loss": 0.0216, "step": 26550 }, { "grad_norm": 0.2918705940246582, "learning_rate": 8.782200859654112e-05, "loss": 0.0253, "step": 26560 }, { "grad_norm": 0.17579251527786255, "learning_rate": 8.781119179615646e-05, "loss": 0.0215, "step": 26570 }, { "grad_norm": 0.23068128526210785, "learning_rate": 8.780037086079674e-05, "loss": 0.0243, "step": 26580 }, { "grad_norm": 0.26697516441345215, "learning_rate": 8.778954579164527e-05, "loss": 0.0271, "step": 26590 }, { "grad_norm": 0.3220677673816681, "learning_rate": 8.777871658988588e-05, "loss": 0.0264, "step": 26600 }, { "grad_norm": 0.18635641038417816, "learning_rate": 8.776788325670285e-05, "loss": 0.0229, "step": 26610 }, { "grad_norm": 0.2076922208070755, "learning_rate": 8.775704579328089e-05, "loss": 0.0243, "step": 26620 }, { "grad_norm": 0.20544320344924927, "learning_rate": 8.774620420080517e-05, "loss": 0.0242, "step": 26630 }, { "grad_norm": 0.2369968295097351, "learning_rate": 8.773535848046131e-05, "loss": 0.021, "step": 26640 }, { "grad_norm": 0.2590203285217285, "learning_rate": 8.772450863343538e-05, "loss": 0.0257, "step": 26650 }, { "grad_norm": 0.2065986543893814, "learning_rate": 8.77136546609139e-05, "loss": 0.0269, "step": 26660 }, { "grad_norm": 0.19320058822631836, "learning_rate": 8.770279656408385e-05, "loss": 0.0269, "step": 26670 }, { "grad_norm": 0.2443171888589859, "learning_rate": 8.769193434413265e-05, "loss": 0.0256, "step": 26680 }, { "grad_norm": 0.1944308876991272, "learning_rate": 8.76810680022482e-05, "loss": 0.0253, "step": 26690 }, { "grad_norm": 0.24429798126220703, "learning_rate": 8.767019753961878e-05, "loss": 0.0268, "step": 26700 }, { "grad_norm": 0.27939432859420776, "learning_rate": 8.765932295743321e-05, "loss": 0.0282, "step": 26710 }, { "grad_norm": 0.2572264075279236, "learning_rate": 8.764844425688068e-05, "loss": 0.0245, "step": 26720 }, { "grad_norm": 0.3080441355705261, "learning_rate": 8.763756143915092e-05, "loss": 0.0261, "step": 26730 }, { "grad_norm": 0.32479122281074524, "learning_rate": 8.7626674505434e-05, "loss": 0.0261, "step": 26740 }, { "grad_norm": 0.20069943368434906, "learning_rate": 8.761578345692053e-05, "loss": 0.0253, "step": 26750 }, { "grad_norm": 0.2707664668560028, "learning_rate": 8.760488829480156e-05, "loss": 0.0255, "step": 26760 }, { "grad_norm": 0.20305131375789642, "learning_rate": 8.759398902026854e-05, "loss": 0.0215, "step": 26770 }, { "grad_norm": 0.23256148397922516, "learning_rate": 8.758308563451339e-05, "loss": 0.021, "step": 26780 }, { "grad_norm": 0.23403751850128174, "learning_rate": 8.75721781387285e-05, "loss": 0.0247, "step": 26790 }, { "grad_norm": 0.2569277286529541, "learning_rate": 8.75612665341067e-05, "loss": 0.0233, "step": 26800 }, { "grad_norm": 0.2307211458683014, "learning_rate": 8.755035082184126e-05, "loss": 0.0245, "step": 26810 }, { "grad_norm": 0.31543809175491333, "learning_rate": 8.753943100312592e-05, "loss": 0.0212, "step": 26820 }, { "grad_norm": 0.2699650228023529, "learning_rate": 8.752850707915484e-05, "loss": 0.023, "step": 26830 }, { "grad_norm": 0.22464220225811005, "learning_rate": 8.751757905112264e-05, "loss": 0.0275, "step": 26840 }, { "grad_norm": 0.28812944889068604, "learning_rate": 8.75066469202244e-05, "loss": 0.0288, "step": 26850 }, { "grad_norm": 0.29008227586746216, "learning_rate": 8.749571068765567e-05, "loss": 0.0264, "step": 26860 }, { "grad_norm": 0.2215466946363449, "learning_rate": 8.748477035461238e-05, "loss": 0.0288, "step": 26870 }, { "grad_norm": 0.28768390417099, "learning_rate": 8.747382592229095e-05, "loss": 0.0331, "step": 26880 }, { "grad_norm": 0.28738221526145935, "learning_rate": 8.746287739188828e-05, "loss": 0.022, "step": 26890 }, { "grad_norm": 0.28612807393074036, "learning_rate": 8.745192476460165e-05, "loss": 0.0284, "step": 26900 }, { "grad_norm": 0.31822097301483154, "learning_rate": 8.744096804162882e-05, "loss": 0.027, "step": 26910 }, { "grad_norm": 0.20163744688034058, "learning_rate": 8.743000722416804e-05, "loss": 0.0274, "step": 26920 }, { "grad_norm": 0.2523147761821747, "learning_rate": 8.741904231341793e-05, "loss": 0.0288, "step": 26930 }, { "grad_norm": 0.24978166818618774, "learning_rate": 8.740807331057762e-05, "loss": 0.0272, "step": 26940 }, { "grad_norm": 0.21352902054786682, "learning_rate": 8.739710021684667e-05, "loss": 0.0235, "step": 26950 }, { "grad_norm": 0.21133652329444885, "learning_rate": 8.738612303342503e-05, "loss": 0.026, "step": 26960 }, { "grad_norm": 0.23874874413013458, "learning_rate": 8.73751417615132e-05, "loss": 0.0264, "step": 26970 }, { "grad_norm": 0.26472172141075134, "learning_rate": 8.736415640231208e-05, "loss": 0.0261, "step": 26980 }, { "grad_norm": 0.2837572395801544, "learning_rate": 8.735316695702297e-05, "loss": 0.0274, "step": 26990 }, { "grad_norm": 0.27683255076408386, "learning_rate": 8.734217342684769e-05, "loss": 0.0216, "step": 27000 }, { "grad_norm": 0.21277156472206116, "learning_rate": 8.733117581298847e-05, "loss": 0.0278, "step": 27010 }, { "grad_norm": 0.23345832526683807, "learning_rate": 8.732017411664796e-05, "loss": 0.0258, "step": 27020 }, { "grad_norm": 0.23190435767173767, "learning_rate": 8.730916833902936e-05, "loss": 0.0219, "step": 27030 }, { "grad_norm": 0.31816837191581726, "learning_rate": 8.729815848133618e-05, "loss": 0.0238, "step": 27040 }, { "grad_norm": 0.1549568623304367, "learning_rate": 8.728714454477247e-05, "loss": 0.0186, "step": 27050 }, { "grad_norm": 0.2813495993614197, "learning_rate": 8.727612653054269e-05, "loss": 0.0243, "step": 27060 }, { "grad_norm": 0.15511436760425568, "learning_rate": 8.726510443985176e-05, "loss": 0.0194, "step": 27070 }, { "grad_norm": 0.2561745345592499, "learning_rate": 8.725407827390503e-05, "loss": 0.0218, "step": 27080 }, { "grad_norm": 0.1684851348400116, "learning_rate": 8.724304803390833e-05, "loss": 0.0191, "step": 27090 }, { "grad_norm": 0.22071895003318787, "learning_rate": 8.723201372106788e-05, "loss": 0.0222, "step": 27100 }, { "grad_norm": 0.20553910732269287, "learning_rate": 8.722097533659038e-05, "loss": 0.0266, "step": 27110 }, { "grad_norm": 0.2986704111099243, "learning_rate": 8.720993288168299e-05, "loss": 0.0343, "step": 27120 }, { "grad_norm": 0.18568527698516846, "learning_rate": 8.719888635755327e-05, "loss": 0.0227, "step": 27130 }, { "grad_norm": 0.24249807000160217, "learning_rate": 8.718783576540928e-05, "loss": 0.0241, "step": 27140 }, { "grad_norm": 0.17735575139522552, "learning_rate": 8.717678110645948e-05, "loss": 0.0194, "step": 27150 }, { "grad_norm": 0.23411588370800018, "learning_rate": 8.716572238191279e-05, "loss": 0.0213, "step": 27160 }, { "grad_norm": 0.1922314465045929, "learning_rate": 8.715465959297857e-05, "loss": 0.0194, "step": 27170 }, { "grad_norm": 0.22603407502174377, "learning_rate": 8.714359274086665e-05, "loss": 0.0269, "step": 27180 }, { "grad_norm": 0.24605199694633484, "learning_rate": 8.713252182678726e-05, "loss": 0.0211, "step": 27190 }, { "grad_norm": 0.3023376166820526, "learning_rate": 8.712144685195112e-05, "loss": 0.0207, "step": 27200 }, { "grad_norm": 0.19866162538528442, "learning_rate": 8.711036781756936e-05, "loss": 0.0214, "step": 27210 }, { "grad_norm": 0.20253530144691467, "learning_rate": 8.709928472485357e-05, "loss": 0.0249, "step": 27220 }, { "grad_norm": 0.23365597426891327, "learning_rate": 8.708819757501579e-05, "loss": 0.0209, "step": 27230 }, { "grad_norm": 0.23182903230190277, "learning_rate": 8.707710636926846e-05, "loss": 0.0239, "step": 27240 }, { "grad_norm": 0.23338697850704193, "learning_rate": 8.706601110882455e-05, "loss": 0.0239, "step": 27250 }, { "grad_norm": 0.2283831536769867, "learning_rate": 8.705491179489738e-05, "loss": 0.0222, "step": 27260 }, { "grad_norm": 0.24647514522075653, "learning_rate": 8.704380842870077e-05, "loss": 0.0224, "step": 27270 }, { "grad_norm": 0.1912824511528015, "learning_rate": 8.703270101144895e-05, "loss": 0.022, "step": 27280 }, { "grad_norm": 0.2396916300058365, "learning_rate": 8.702158954435664e-05, "loss": 0.0262, "step": 27290 }, { "grad_norm": 0.3253946304321289, "learning_rate": 8.701047402863896e-05, "loss": 0.0222, "step": 27300 }, { "grad_norm": 0.2523268759250641, "learning_rate": 8.699935446551148e-05, "loss": 0.0197, "step": 27310 }, { "grad_norm": 0.3299647867679596, "learning_rate": 8.698823085619022e-05, "loss": 0.0217, "step": 27320 }, { "grad_norm": 0.2201744019985199, "learning_rate": 8.697710320189166e-05, "loss": 0.0198, "step": 27330 }, { "grad_norm": 0.21788406372070312, "learning_rate": 8.696597150383268e-05, "loss": 0.0265, "step": 27340 }, { "grad_norm": 0.2225555181503296, "learning_rate": 8.695483576323063e-05, "loss": 0.0203, "step": 27350 }, { "grad_norm": 0.22787973284721375, "learning_rate": 8.69436959813033e-05, "loss": 0.0241, "step": 27360 }, { "grad_norm": 0.2188078612089157, "learning_rate": 8.693255215926892e-05, "loss": 0.0236, "step": 27370 }, { "grad_norm": 0.31585249304771423, "learning_rate": 8.692140429834617e-05, "loss": 0.0257, "step": 27380 }, { "grad_norm": 0.28327134251594543, "learning_rate": 8.691025239975415e-05, "loss": 0.0221, "step": 27390 }, { "grad_norm": 0.20178548991680145, "learning_rate": 8.689909646471243e-05, "loss": 0.0224, "step": 27400 }, { "grad_norm": 0.2830538749694824, "learning_rate": 8.688793649444099e-05, "loss": 0.0288, "step": 27410 }, { "grad_norm": 0.23385939002037048, "learning_rate": 8.687677249016029e-05, "loss": 0.0249, "step": 27420 }, { "grad_norm": 0.2853049635887146, "learning_rate": 8.686560445309118e-05, "loss": 0.027, "step": 27430 }, { "grad_norm": 0.25355401635169983, "learning_rate": 8.685443238445499e-05, "loss": 0.0254, "step": 27440 }, { "grad_norm": 0.2584156394004822, "learning_rate": 8.68432562854735e-05, "loss": 0.0239, "step": 27450 }, { "grad_norm": 0.23636309802532196, "learning_rate": 8.683207615736887e-05, "loss": 0.0243, "step": 27460 }, { "grad_norm": 0.35825902223587036, "learning_rate": 8.682089200136379e-05, "loss": 0.0327, "step": 27470 }, { "grad_norm": 0.20604656636714935, "learning_rate": 8.680970381868132e-05, "loss": 0.023, "step": 27480 }, { "grad_norm": 0.2314099669456482, "learning_rate": 8.679851161054498e-05, "loss": 0.0221, "step": 27490 }, { "grad_norm": 0.2768557071685791, "learning_rate": 8.678731537817873e-05, "loss": 0.0211, "step": 27500 }, { "grad_norm": 0.22920571267604828, "learning_rate": 8.677611512280697e-05, "loss": 0.0206, "step": 27510 }, { "grad_norm": 0.2079850137233734, "learning_rate": 8.676491084565457e-05, "loss": 0.0218, "step": 27520 }, { "grad_norm": 0.3125520646572113, "learning_rate": 8.675370254794678e-05, "loss": 0.0219, "step": 27530 }, { "grad_norm": 0.2057381570339203, "learning_rate": 8.674249023090935e-05, "loss": 0.0232, "step": 27540 }, { "grad_norm": 0.2502601146697998, "learning_rate": 8.673127389576843e-05, "loss": 0.0284, "step": 27550 }, { "grad_norm": 0.22827765345573425, "learning_rate": 8.67200535437506e-05, "loss": 0.0273, "step": 27560 }, { "grad_norm": 0.22707737982273102, "learning_rate": 8.670882917608296e-05, "loss": 0.0226, "step": 27570 }, { "grad_norm": 0.22573009133338928, "learning_rate": 8.669760079399292e-05, "loss": 0.0244, "step": 27580 }, { "grad_norm": 0.2040591686964035, "learning_rate": 8.668636839870845e-05, "loss": 0.0222, "step": 27590 }, { "grad_norm": 0.2424248307943344, "learning_rate": 8.667513199145789e-05, "loss": 0.0235, "step": 27600 }, { "grad_norm": 0.28643226623535156, "learning_rate": 8.666389157347002e-05, "loss": 0.0214, "step": 27610 }, { "grad_norm": 0.24052736163139343, "learning_rate": 8.66526471459741e-05, "loss": 0.0247, "step": 27620 }, { "grad_norm": 0.24634963274002075, "learning_rate": 8.66413987101998e-05, "loss": 0.0224, "step": 27630 }, { "grad_norm": 0.21327942609786987, "learning_rate": 8.663014626737723e-05, "loss": 0.0222, "step": 27640 }, { "grad_norm": 0.18220558762550354, "learning_rate": 8.661888981873691e-05, "loss": 0.0232, "step": 27650 }, { "grad_norm": 0.1956372708082199, "learning_rate": 8.660762936550988e-05, "loss": 0.0205, "step": 27660 }, { "grad_norm": 0.1728818714618683, "learning_rate": 8.659636490892753e-05, "loss": 0.0194, "step": 27670 }, { "grad_norm": 0.22056441009044647, "learning_rate": 8.658509645022174e-05, "loss": 0.0231, "step": 27680 }, { "grad_norm": 0.27872300148010254, "learning_rate": 8.657382399062481e-05, "loss": 0.0228, "step": 27690 }, { "grad_norm": 0.23357334733009338, "learning_rate": 8.656254753136946e-05, "loss": 0.0255, "step": 27700 }, { "grad_norm": 0.28371548652648926, "learning_rate": 8.655126707368891e-05, "loss": 0.025, "step": 27710 }, { "grad_norm": 0.23145614564418793, "learning_rate": 8.653998261881672e-05, "loss": 0.025, "step": 27720 }, { "grad_norm": 0.2630898356437683, "learning_rate": 8.652869416798699e-05, "loss": 0.0285, "step": 27730 }, { "grad_norm": 0.26275429129600525, "learning_rate": 8.651740172243417e-05, "loss": 0.02, "step": 27740 }, { "grad_norm": 0.2682112157344818, "learning_rate": 8.65061052833932e-05, "loss": 0.0251, "step": 27750 }, { "grad_norm": 0.219699889421463, "learning_rate": 8.649480485209945e-05, "loss": 0.0219, "step": 27760 }, { "grad_norm": 0.18491609394550323, "learning_rate": 8.64835004297887e-05, "loss": 0.0189, "step": 27770 }, { "grad_norm": 0.2357862889766693, "learning_rate": 8.64721920176972e-05, "loss": 0.0245, "step": 27780 }, { "grad_norm": 0.2550540864467621, "learning_rate": 8.646087961706164e-05, "loss": 0.0286, "step": 27790 }, { "grad_norm": 0.7126806378364563, "learning_rate": 8.644956322911908e-05, "loss": 0.0237, "step": 27800 }, { "grad_norm": 0.25878241658210754, "learning_rate": 8.643824285510709e-05, "loss": 0.0235, "step": 27810 }, { "grad_norm": 0.30857956409454346, "learning_rate": 8.642691849626364e-05, "loss": 0.0244, "step": 27820 }, { "grad_norm": 1.8094618320465088, "learning_rate": 8.641559015382717e-05, "loss": 0.0276, "step": 27830 }, { "grad_norm": 0.2590309679508209, "learning_rate": 8.640425782903649e-05, "loss": 0.0279, "step": 27840 }, { "grad_norm": 0.20961473882198334, "learning_rate": 8.639292152313091e-05, "loss": 0.0261, "step": 27850 }, { "grad_norm": 0.19401124119758606, "learning_rate": 8.638158123735015e-05, "loss": 0.0284, "step": 27860 }, { "grad_norm": 0.2830081880092621, "learning_rate": 8.637023697293436e-05, "loss": 0.0236, "step": 27870 }, { "grad_norm": 0.22597283124923706, "learning_rate": 8.635888873112414e-05, "loss": 0.0291, "step": 27880 }, { "grad_norm": 0.244629368185997, "learning_rate": 8.634753651316052e-05, "loss": 0.0252, "step": 27890 }, { "grad_norm": 0.20160561800003052, "learning_rate": 8.633618032028496e-05, "loss": 0.0264, "step": 27900 }, { "grad_norm": 0.23262545466423035, "learning_rate": 8.632482015373934e-05, "loss": 0.0196, "step": 27910 }, { "grad_norm": 0.21964868903160095, "learning_rate": 8.6313456014766e-05, "loss": 0.0227, "step": 27920 }, { "grad_norm": 0.28975507616996765, "learning_rate": 8.630208790460771e-05, "loss": 0.0218, "step": 27930 }, { "grad_norm": 0.2694137692451477, "learning_rate": 8.629071582450768e-05, "loss": 0.0244, "step": 27940 }, { "grad_norm": 0.17323894798755646, "learning_rate": 8.62793397757095e-05, "loss": 0.0224, "step": 27950 }, { "grad_norm": 0.18460820615291595, "learning_rate": 8.626795975945729e-05, "loss": 0.0262, "step": 27960 }, { "grad_norm": 0.20833393931388855, "learning_rate": 8.625657577699551e-05, "loss": 0.0248, "step": 27970 }, { "grad_norm": 0.21624061465263367, "learning_rate": 8.624518782956914e-05, "loss": 0.0229, "step": 27980 }, { "grad_norm": 0.2952490448951721, "learning_rate": 8.62337959184235e-05, "loss": 0.0239, "step": 27990 }, { "grad_norm": 0.23223774135112762, "learning_rate": 8.622240004480441e-05, "loss": 0.0311, "step": 28000 }, { "grad_norm": 0.185220867395401, "learning_rate": 8.621100020995814e-05, "loss": 0.0215, "step": 28010 }, { "grad_norm": 0.23604312539100647, "learning_rate": 8.619959641513132e-05, "loss": 0.0242, "step": 28020 }, { "grad_norm": 0.24316808581352234, "learning_rate": 8.618818866157105e-05, "loss": 0.0245, "step": 28030 }, { "grad_norm": 0.1590012162923813, "learning_rate": 8.617677695052487e-05, "loss": 0.0202, "step": 28040 }, { "grad_norm": 0.18574896454811096, "learning_rate": 8.616536128324078e-05, "loss": 0.0186, "step": 28050 }, { "grad_norm": 0.20404289662837982, "learning_rate": 8.615394166096712e-05, "loss": 0.0262, "step": 28060 }, { "grad_norm": 0.2273361086845398, "learning_rate": 8.614251808495279e-05, "loss": 0.0206, "step": 28070 }, { "grad_norm": 0.24207058548927307, "learning_rate": 8.6131090556447e-05, "loss": 0.0241, "step": 28080 }, { "grad_norm": 0.22971494495868683, "learning_rate": 8.611965907669947e-05, "loss": 0.0197, "step": 28090 }, { "grad_norm": 0.23996901512145996, "learning_rate": 8.610822364696034e-05, "loss": 0.0289, "step": 28100 }, { "grad_norm": 0.22218100726604462, "learning_rate": 8.609678426848015e-05, "loss": 0.0257, "step": 28110 }, { "grad_norm": 0.25259751081466675, "learning_rate": 8.60853409425099e-05, "loss": 0.0212, "step": 28120 }, { "grad_norm": 0.21580199897289276, "learning_rate": 8.607389367030104e-05, "loss": 0.0237, "step": 28130 }, { "grad_norm": 0.35977908968925476, "learning_rate": 8.606244245310538e-05, "loss": 0.0241, "step": 28140 }, { "grad_norm": 0.24297291040420532, "learning_rate": 8.605098729217525e-05, "loss": 0.0216, "step": 28150 }, { "grad_norm": 0.2015523761510849, "learning_rate": 8.603952818876335e-05, "loss": 0.0247, "step": 28160 }, { "grad_norm": 0.3075808584690094, "learning_rate": 8.602806514412281e-05, "loss": 0.0249, "step": 28170 }, { "grad_norm": 0.21965943276882172, "learning_rate": 8.601659815950726e-05, "loss": 0.0226, "step": 28180 }, { "grad_norm": 0.29258057475090027, "learning_rate": 8.600512723617067e-05, "loss": 0.0217, "step": 28190 }, { "grad_norm": 0.2268304079771042, "learning_rate": 8.59936523753675e-05, "loss": 0.0247, "step": 28200 }, { "grad_norm": 0.23270857334136963, "learning_rate": 8.598217357835264e-05, "loss": 0.0239, "step": 28210 }, { "grad_norm": 0.25424662232398987, "learning_rate": 8.597069084638135e-05, "loss": 0.0203, "step": 28220 }, { "grad_norm": 0.16668885946273804, "learning_rate": 8.595920418070939e-05, "loss": 0.0195, "step": 28230 }, { "grad_norm": 0.24378074705600739, "learning_rate": 8.594771358259295e-05, "loss": 0.0239, "step": 28240 }, { "grad_norm": 0.1859445571899414, "learning_rate": 8.593621905328858e-05, "loss": 0.017, "step": 28250 }, { "grad_norm": 0.22518852353096008, "learning_rate": 8.592472059405333e-05, "loss": 0.0225, "step": 28260 }, { "grad_norm": 0.21653932332992554, "learning_rate": 8.591321820614464e-05, "loss": 0.0203, "step": 28270 }, { "grad_norm": 0.1901414841413498, "learning_rate": 8.590171189082041e-05, "loss": 0.0227, "step": 28280 }, { "grad_norm": 0.2068263292312622, "learning_rate": 8.589020164933894e-05, "loss": 0.0223, "step": 28290 }, { "grad_norm": 0.2345043569803238, "learning_rate": 8.587868748295898e-05, "loss": 0.0231, "step": 28300 }, { "grad_norm": 0.2495054006576538, "learning_rate": 8.586716939293971e-05, "loss": 0.0262, "step": 28310 }, { "grad_norm": 0.2606671452522278, "learning_rate": 8.58556473805407e-05, "loss": 0.0213, "step": 28320 }, { "grad_norm": 0.21804460883140564, "learning_rate": 8.584412144702202e-05, "loss": 0.0222, "step": 28330 }, { "grad_norm": 0.2942163348197937, "learning_rate": 8.58325915936441e-05, "loss": 0.023, "step": 28340 }, { "grad_norm": 0.186431884765625, "learning_rate": 8.582105782166783e-05, "loss": 0.0262, "step": 28350 }, { "grad_norm": 0.20220209658145905, "learning_rate": 8.580952013235455e-05, "loss": 0.0283, "step": 28360 }, { "grad_norm": 0.2256481796503067, "learning_rate": 8.579797852696596e-05, "loss": 0.0262, "step": 28370 }, { "grad_norm": 0.2751723527908325, "learning_rate": 8.578643300676428e-05, "loss": 0.0223, "step": 28380 }, { "grad_norm": 0.21473568677902222, "learning_rate": 8.577488357301209e-05, "loss": 0.0281, "step": 28390 }, { "grad_norm": 0.27157101035118103, "learning_rate": 8.576333022697242e-05, "loss": 0.0301, "step": 28400 }, { "grad_norm": 0.26627957820892334, "learning_rate": 8.575177296990873e-05, "loss": 0.022, "step": 28410 }, { "grad_norm": 0.19337551295757294, "learning_rate": 8.574021180308489e-05, "loss": 0.0212, "step": 28420 }, { "grad_norm": 0.21646034717559814, "learning_rate": 8.572864672776523e-05, "loss": 0.0216, "step": 28430 }, { "grad_norm": 0.2863789200782776, "learning_rate": 8.571707774521447e-05, "loss": 0.0209, "step": 28440 }, { "grad_norm": 0.2216871827840805, "learning_rate": 8.57055048566978e-05, "loss": 0.0236, "step": 28450 }, { "grad_norm": 0.21980448067188263, "learning_rate": 8.569392806348078e-05, "loss": 0.0192, "step": 28460 }, { "grad_norm": 0.21268165111541748, "learning_rate": 8.568234736682947e-05, "loss": 0.0208, "step": 28470 }, { "grad_norm": 0.2858617603778839, "learning_rate": 8.567076276801029e-05, "loss": 0.0278, "step": 28480 }, { "grad_norm": 0.23822788894176483, "learning_rate": 8.565917426829013e-05, "loss": 0.0226, "step": 28490 }, { "grad_norm": 0.23784774541854858, "learning_rate": 8.564758186893628e-05, "loss": 0.0207, "step": 28500 }, { "grad_norm": 0.20514680445194244, "learning_rate": 8.563598557121649e-05, "loss": 0.0244, "step": 28510 }, { "grad_norm": 0.21370935440063477, "learning_rate": 8.562438537639888e-05, "loss": 0.0212, "step": 28520 }, { "grad_norm": 0.21184664964675903, "learning_rate": 8.561278128575206e-05, "loss": 0.0253, "step": 28530 }, { "grad_norm": 0.2930290400981903, "learning_rate": 8.5601173300545e-05, "loss": 0.022, "step": 28540 }, { "grad_norm": 0.2771775722503662, "learning_rate": 8.558956142204717e-05, "loss": 0.0237, "step": 28550 }, { "grad_norm": 0.25550904870033264, "learning_rate": 8.55779456515284e-05, "loss": 0.0243, "step": 28560 }, { "grad_norm": 0.20709407329559326, "learning_rate": 8.556632599025898e-05, "loss": 0.0209, "step": 28570 }, { "grad_norm": 0.2847048044204712, "learning_rate": 8.555470243950964e-05, "loss": 0.0252, "step": 28580 }, { "grad_norm": 0.2653869390487671, "learning_rate": 8.554307500055148e-05, "loss": 0.0264, "step": 28590 }, { "grad_norm": 0.21699300408363342, "learning_rate": 8.553144367465609e-05, "loss": 0.0255, "step": 28600 }, { "grad_norm": 0.20886050164699554, "learning_rate": 8.551980846309544e-05, "loss": 0.025, "step": 28610 }, { "grad_norm": 0.19499507546424866, "learning_rate": 8.550816936714193e-05, "loss": 0.0224, "step": 28620 }, { "grad_norm": 0.26048004627227783, "learning_rate": 8.549652638806841e-05, "loss": 0.024, "step": 28630 }, { "grad_norm": 0.23880265653133392, "learning_rate": 8.548487952714812e-05, "loss": 0.0241, "step": 28640 }, { "grad_norm": 0.23061782121658325, "learning_rate": 8.547322878565478e-05, "loss": 0.0218, "step": 28650 }, { "grad_norm": 0.26042479276657104, "learning_rate": 8.546157416486245e-05, "loss": 0.026, "step": 28660 }, { "grad_norm": 0.24654258787631989, "learning_rate": 8.54499156660457e-05, "loss": 0.0213, "step": 28670 }, { "grad_norm": 0.18656161427497864, "learning_rate": 8.543825329047947e-05, "loss": 0.0225, "step": 28680 }, { "grad_norm": 0.29544103145599365, "learning_rate": 8.542658703943913e-05, "loss": 0.0255, "step": 28690 }, { "grad_norm": 0.2412070780992508, "learning_rate": 8.541491691420051e-05, "loss": 0.0215, "step": 28700 }, { "grad_norm": 0.2334507554769516, "learning_rate": 8.54032429160398e-05, "loss": 0.0219, "step": 28710 }, { "grad_norm": 0.23576056957244873, "learning_rate": 8.539156504623369e-05, "loss": 0.0216, "step": 28720 }, { "grad_norm": 0.19927439093589783, "learning_rate": 8.537988330605923e-05, "loss": 0.0197, "step": 28730 }, { "grad_norm": 0.18481846153736115, "learning_rate": 8.536819769679393e-05, "loss": 0.0216, "step": 28740 }, { "grad_norm": 0.22330570220947266, "learning_rate": 8.53565082197157e-05, "loss": 0.0218, "step": 28750 }, { "grad_norm": 0.19266103208065033, "learning_rate": 8.534481487610289e-05, "loss": 0.0208, "step": 28760 }, { "grad_norm": 0.16978512704372406, "learning_rate": 8.533311766723428e-05, "loss": 0.0232, "step": 28770 }, { "grad_norm": 0.2495468407869339, "learning_rate": 8.532141659438901e-05, "loss": 0.0222, "step": 28780 }, { "grad_norm": 0.2859927713871002, "learning_rate": 8.530971165884675e-05, "loss": 0.0247, "step": 28790 }, { "grad_norm": 0.2268660068511963, "learning_rate": 8.529800286188752e-05, "loss": 0.0199, "step": 28800 }, { "grad_norm": 0.1529732197523117, "learning_rate": 8.528629020479175e-05, "loss": 0.022, "step": 28810 }, { "grad_norm": 0.20275798439979553, "learning_rate": 8.527457368884033e-05, "loss": 0.0251, "step": 28820 }, { "grad_norm": 0.16803337633609772, "learning_rate": 8.526285331531458e-05, "loss": 0.0247, "step": 28830 }, { "grad_norm": 0.19699017703533173, "learning_rate": 8.525112908549621e-05, "loss": 0.0188, "step": 28840 }, { "grad_norm": 0.3143371343612671, "learning_rate": 8.523940100066735e-05, "loss": 0.0219, "step": 28850 }, { "grad_norm": 0.22494232654571533, "learning_rate": 8.52276690621106e-05, "loss": 0.021, "step": 28860 }, { "grad_norm": 0.17788995802402496, "learning_rate": 8.521593327110889e-05, "loss": 0.0214, "step": 28870 }, { "grad_norm": 0.23266445100307465, "learning_rate": 8.520419362894569e-05, "loss": 0.0213, "step": 28880 }, { "grad_norm": 0.2125825434923172, "learning_rate": 8.51924501369048e-05, "loss": 0.0215, "step": 28890 }, { "grad_norm": 0.19547998905181885, "learning_rate": 8.518070279627047e-05, "loss": 0.0228, "step": 28900 }, { "grad_norm": 0.3157593011856079, "learning_rate": 8.516895160832737e-05, "loss": 0.0212, "step": 28910 }, { "grad_norm": 0.1427207589149475, "learning_rate": 8.515719657436061e-05, "loss": 0.0218, "step": 28920 }, { "grad_norm": 0.1864393949508667, "learning_rate": 8.514543769565568e-05, "loss": 0.024, "step": 28930 }, { "grad_norm": 0.24603953957557678, "learning_rate": 8.513367497349853e-05, "loss": 0.0249, "step": 28940 }, { "grad_norm": 0.2574710547924042, "learning_rate": 8.51219084091755e-05, "loss": 0.0206, "step": 28950 }, { "grad_norm": 0.2585103213787079, "learning_rate": 8.511013800397338e-05, "loss": 0.0214, "step": 28960 }, { "grad_norm": 0.2599208354949951, "learning_rate": 8.509836375917937e-05, "loss": 0.0252, "step": 28970 }, { "grad_norm": 0.20570950210094452, "learning_rate": 8.508658567608104e-05, "loss": 0.0272, "step": 28980 }, { "grad_norm": 0.3145768940448761, "learning_rate": 8.507480375596647e-05, "loss": 0.0185, "step": 28990 }, { "grad_norm": 0.201312854886055, "learning_rate": 8.506301800012408e-05, "loss": 0.0252, "step": 29000 }, { "grad_norm": 0.2001335620880127, "learning_rate": 8.505122840984278e-05, "loss": 0.0225, "step": 29010 }, { "grad_norm": 0.24061380326747894, "learning_rate": 8.503943498641182e-05, "loss": 0.0215, "step": 29020 }, { "grad_norm": 0.28508657217025757, "learning_rate": 8.502763773112095e-05, "loss": 0.0178, "step": 29030 }, { "grad_norm": 0.17662551999092102, "learning_rate": 8.501583664526026e-05, "loss": 0.022, "step": 29040 }, { "grad_norm": 0.2718694508075714, "learning_rate": 8.500403173012032e-05, "loss": 0.0222, "step": 29050 }, { "grad_norm": 0.21694643795490265, "learning_rate": 8.499222298699211e-05, "loss": 0.0202, "step": 29060 }, { "grad_norm": 0.3444920778274536, "learning_rate": 8.498041041716701e-05, "loss": 0.0245, "step": 29070 }, { "grad_norm": 0.24134904146194458, "learning_rate": 8.496859402193681e-05, "loss": 0.0252, "step": 29080 }, { "grad_norm": 0.1767842024564743, "learning_rate": 8.495677380259374e-05, "loss": 0.0209, "step": 29090 }, { "grad_norm": 0.19227740168571472, "learning_rate": 8.494494976043045e-05, "loss": 0.0226, "step": 29100 }, { "grad_norm": 0.40897542238235474, "learning_rate": 8.493312189673998e-05, "loss": 0.0232, "step": 29110 }, { "grad_norm": 0.2179865688085556, "learning_rate": 8.492129021281584e-05, "loss": 0.0225, "step": 29120 }, { "grad_norm": 0.19834622740745544, "learning_rate": 8.490945470995188e-05, "loss": 0.0241, "step": 29130 }, { "grad_norm": 0.23792590200901031, "learning_rate": 8.489761538944247e-05, "loss": 0.0237, "step": 29140 }, { "grad_norm": 0.17708024382591248, "learning_rate": 8.48857722525823e-05, "loss": 0.0213, "step": 29150 }, { "grad_norm": 0.2635959982872009, "learning_rate": 8.487392530066652e-05, "loss": 0.0215, "step": 29160 }, { "grad_norm": 0.2196030616760254, "learning_rate": 8.486207453499069e-05, "loss": 0.0201, "step": 29170 }, { "grad_norm": 0.21147523820400238, "learning_rate": 8.485021995685082e-05, "loss": 0.0257, "step": 29180 }, { "grad_norm": 0.1995924562215805, "learning_rate": 8.483836156754328e-05, "loss": 0.0212, "step": 29190 }, { "grad_norm": 0.21250827610492706, "learning_rate": 8.482649936836491e-05, "loss": 0.0219, "step": 29200 }, { "grad_norm": 0.2581084668636322, "learning_rate": 8.481463336061293e-05, "loss": 0.024, "step": 29210 }, { "grad_norm": 0.189692884683609, "learning_rate": 8.480276354558496e-05, "loss": 0.0193, "step": 29220 }, { "grad_norm": 0.20180770754814148, "learning_rate": 8.479088992457913e-05, "loss": 0.0223, "step": 29230 }, { "grad_norm": 0.279131680727005, "learning_rate": 8.477901249889387e-05, "loss": 0.0256, "step": 29240 }, { "grad_norm": 0.2070462703704834, "learning_rate": 8.47671312698281e-05, "loss": 0.0226, "step": 29250 }, { "grad_norm": 0.22324776649475098, "learning_rate": 8.475524623868112e-05, "loss": 0.0218, "step": 29260 }, { "grad_norm": 0.22478356957435608, "learning_rate": 8.474335740675266e-05, "loss": 0.0226, "step": 29270 }, { "grad_norm": 0.2325558066368103, "learning_rate": 8.473146477534289e-05, "loss": 0.0197, "step": 29280 }, { "grad_norm": 0.23112456500530243, "learning_rate": 8.471956834575232e-05, "loss": 0.0162, "step": 29290 }, { "grad_norm": 0.23955611884593964, "learning_rate": 8.470766811928197e-05, "loss": 0.0213, "step": 29300 }, { "grad_norm": 0.3135509490966797, "learning_rate": 8.469576409723323e-05, "loss": 0.0297, "step": 29310 }, { "grad_norm": 0.2549801766872406, "learning_rate": 8.468385628090788e-05, "loss": 0.0211, "step": 29320 }, { "grad_norm": 0.25373613834381104, "learning_rate": 8.467194467160815e-05, "loss": 0.022, "step": 29330 }, { "grad_norm": 0.25299373269081116, "learning_rate": 8.466002927063667e-05, "loss": 0.0249, "step": 29340 }, { "grad_norm": 0.2213621884584427, "learning_rate": 8.464811007929651e-05, "loss": 0.0214, "step": 29350 }, { "grad_norm": 0.19920571148395538, "learning_rate": 8.463618709889114e-05, "loss": 0.0237, "step": 29360 }, { "grad_norm": 0.19209806621074677, "learning_rate": 8.462426033072442e-05, "loss": 0.0209, "step": 29370 }, { "grad_norm": 0.17395174503326416, "learning_rate": 8.461232977610061e-05, "loss": 0.0179, "step": 29380 }, { "grad_norm": 0.2131962925195694, "learning_rate": 8.46003954363245e-05, "loss": 0.021, "step": 29390 }, { "grad_norm": 0.22789190709590912, "learning_rate": 8.458845731270115e-05, "loss": 0.0226, "step": 29400 }, { "grad_norm": 0.23674112558364868, "learning_rate": 8.45765154065361e-05, "loss": 0.0229, "step": 29410 }, { "grad_norm": 0.229502871632576, "learning_rate": 8.456456971913532e-05, "loss": 0.0188, "step": 29420 }, { "grad_norm": 0.21998699009418488, "learning_rate": 8.455262025180517e-05, "loss": 0.0229, "step": 29430 }, { "grad_norm": 0.2587198317050934, "learning_rate": 8.454066700585242e-05, "loss": 0.0217, "step": 29440 }, { "grad_norm": 0.2560408413410187, "learning_rate": 8.452870998258423e-05, "loss": 0.0248, "step": 29450 }, { "grad_norm": 0.2407578080892563, "learning_rate": 8.451674918330825e-05, "loss": 0.0215, "step": 29460 }, { "grad_norm": 0.24180066585540771, "learning_rate": 8.450478460933246e-05, "loss": 0.0236, "step": 29470 }, { "grad_norm": 0.20558564364910126, "learning_rate": 8.449281626196532e-05, "loss": 0.0282, "step": 29480 }, { "grad_norm": 0.1876247227191925, "learning_rate": 8.448084414251564e-05, "loss": 0.0228, "step": 29490 }, { "grad_norm": 0.21596172451972961, "learning_rate": 8.446886825229271e-05, "loss": 0.0195, "step": 29500 }, { "grad_norm": 0.18258698284626007, "learning_rate": 8.445688859260615e-05, "loss": 0.0173, "step": 29510 }, { "grad_norm": 0.26938262581825256, "learning_rate": 8.444490516476606e-05, "loss": 0.0201, "step": 29520 }, { "grad_norm": 0.2272406816482544, "learning_rate": 8.443291797008293e-05, "loss": 0.0196, "step": 29530 }, { "grad_norm": 0.19902345538139343, "learning_rate": 8.442092700986765e-05, "loss": 0.0219, "step": 29540 }, { "grad_norm": 0.14475832879543304, "learning_rate": 8.440893228543156e-05, "loss": 0.0193, "step": 29550 }, { "grad_norm": 0.22833910584449768, "learning_rate": 8.439693379808638e-05, "loss": 0.0189, "step": 29560 }, { "grad_norm": 0.2982430160045624, "learning_rate": 8.43849315491442e-05, "loss": 0.0213, "step": 29570 }, { "grad_norm": 0.2445681244134903, "learning_rate": 8.437292553991763e-05, "loss": 0.027, "step": 29580 }, { "grad_norm": 0.29385143518447876, "learning_rate": 8.436091577171959e-05, "loss": 0.0199, "step": 29590 }, { "grad_norm": 0.20620478689670563, "learning_rate": 8.434890224586347e-05, "loss": 0.0245, "step": 29600 }, { "grad_norm": 0.20476455986499786, "learning_rate": 8.433688496366303e-05, "loss": 0.0184, "step": 29610 }, { "grad_norm": 0.20481428503990173, "learning_rate": 8.432486392643248e-05, "loss": 0.0209, "step": 29620 }, { "grad_norm": 0.16922695934772491, "learning_rate": 8.431283913548643e-05, "loss": 0.0205, "step": 29630 }, { "grad_norm": 0.20523343980312347, "learning_rate": 8.430081059213985e-05, "loss": 0.0266, "step": 29640 }, { "grad_norm": 0.272737979888916, "learning_rate": 8.428877829770823e-05, "loss": 0.0237, "step": 29650 }, { "grad_norm": 0.25576260685920715, "learning_rate": 8.427674225350735e-05, "loss": 0.0208, "step": 29660 }, { "grad_norm": 0.21872785687446594, "learning_rate": 8.426470246085347e-05, "loss": 0.022, "step": 29670 }, { "grad_norm": 0.2053612917661667, "learning_rate": 8.425265892106324e-05, "loss": 0.0216, "step": 29680 }, { "grad_norm": 0.2822767198085785, "learning_rate": 8.424061163545374e-05, "loss": 0.0205, "step": 29690 }, { "grad_norm": 0.21779192984104156, "learning_rate": 8.422856060534243e-05, "loss": 0.0208, "step": 29700 }, { "grad_norm": 0.18141034245491028, "learning_rate": 8.421650583204718e-05, "loss": 0.0153, "step": 29710 }, { "grad_norm": 0.20937049388885498, "learning_rate": 8.420444731688633e-05, "loss": 0.018, "step": 29720 }, { "grad_norm": 0.24704071879386902, "learning_rate": 8.419238506117852e-05, "loss": 0.0184, "step": 29730 }, { "grad_norm": 0.20673324167728424, "learning_rate": 8.418031906624289e-05, "loss": 0.0191, "step": 29740 }, { "grad_norm": 0.19040367007255554, "learning_rate": 8.416824933339898e-05, "loss": 0.0246, "step": 29750 }, { "grad_norm": 0.22994646430015564, "learning_rate": 8.415617586396667e-05, "loss": 0.0248, "step": 29760 }, { "grad_norm": 0.3160957992076874, "learning_rate": 8.414409865926632e-05, "loss": 0.0251, "step": 29770 }, { "grad_norm": 0.2235967367887497, "learning_rate": 8.413201772061867e-05, "loss": 0.0234, "step": 29780 }, { "grad_norm": 0.22793874144554138, "learning_rate": 8.411993304934488e-05, "loss": 0.0227, "step": 29790 }, { "grad_norm": 0.25589892268180847, "learning_rate": 8.410784464676654e-05, "loss": 0.0209, "step": 29800 }, { "grad_norm": 0.18888232111930847, "learning_rate": 8.409575251420556e-05, "loss": 0.0192, "step": 29810 }, { "grad_norm": 0.17559044063091278, "learning_rate": 8.408365665298435e-05, "loss": 0.0211, "step": 29820 }, { "grad_norm": 0.27392905950546265, "learning_rate": 8.40715570644257e-05, "loss": 0.0245, "step": 29830 }, { "grad_norm": 0.24403835833072662, "learning_rate": 8.40594537498528e-05, "loss": 0.0234, "step": 29840 }, { "grad_norm": 0.23105449974536896, "learning_rate": 8.404734671058924e-05, "loss": 0.0191, "step": 29850 }, { "grad_norm": 0.2308192253112793, "learning_rate": 8.403523594795902e-05, "loss": 0.0267, "step": 29860 }, { "grad_norm": 0.25301769375801086, "learning_rate": 8.402312146328659e-05, "loss": 0.0201, "step": 29870 }, { "grad_norm": 0.31924110651016235, "learning_rate": 8.401100325789675e-05, "loss": 0.0287, "step": 29880 }, { "grad_norm": 0.2294704169034958, "learning_rate": 8.399888133311472e-05, "loss": 0.0235, "step": 29890 }, { "grad_norm": 0.29345181584358215, "learning_rate": 8.398675569026613e-05, "loss": 0.0209, "step": 29900 }, { "grad_norm": 0.22756832838058472, "learning_rate": 8.397462633067705e-05, "loss": 0.0233, "step": 29910 }, { "grad_norm": 0.20956945419311523, "learning_rate": 8.396249325567392e-05, "loss": 0.0185, "step": 29920 }, { "grad_norm": 0.19018560647964478, "learning_rate": 8.395035646658357e-05, "loss": 0.0319, "step": 29930 }, { "grad_norm": 0.2647618055343628, "learning_rate": 8.39382159647333e-05, "loss": 0.0246, "step": 29940 }, { "grad_norm": 0.2675296366214752, "learning_rate": 8.392607175145075e-05, "loss": 0.0231, "step": 29950 }, { "grad_norm": 0.20255747437477112, "learning_rate": 8.3913923828064e-05, "loss": 0.0211, "step": 29960 }, { "grad_norm": 0.20610861480236053, "learning_rate": 8.390177219590152e-05, "loss": 0.0216, "step": 29970 }, { "grad_norm": 0.22633865475654602, "learning_rate": 8.388961685629222e-05, "loss": 0.0192, "step": 29980 }, { "grad_norm": 0.25563886761665344, "learning_rate": 8.387745781056536e-05, "loss": 0.028, "step": 29990 }, { "grad_norm": 0.22439785301685333, "learning_rate": 8.386529506005065e-05, "loss": 0.0244, "step": 30000 }, { "grad_norm": 0.318238765001297, "learning_rate": 8.38531286060782e-05, "loss": 0.0228, "step": 30010 }, { "grad_norm": 0.23299375176429749, "learning_rate": 8.384095844997849e-05, "loss": 0.0229, "step": 30020 }, { "grad_norm": 0.2339622676372528, "learning_rate": 8.382878459308245e-05, "loss": 0.0222, "step": 30030 }, { "grad_norm": 0.24353963136672974, "learning_rate": 8.381660703672138e-05, "loss": 0.021, "step": 30040 }, { "grad_norm": 0.29206332564353943, "learning_rate": 8.380442578222702e-05, "loss": 0.0205, "step": 30050 }, { "grad_norm": 0.3125546872615814, "learning_rate": 8.379224083093146e-05, "loss": 0.021, "step": 30060 }, { "grad_norm": 0.22696948051452637, "learning_rate": 8.378005218416727e-05, "loss": 0.0195, "step": 30070 }, { "grad_norm": 0.2747948467731476, "learning_rate": 8.376785984326735e-05, "loss": 0.0211, "step": 30080 }, { "grad_norm": 0.30275237560272217, "learning_rate": 8.375566380956506e-05, "loss": 0.0218, "step": 30090 }, { "grad_norm": 0.29729074239730835, "learning_rate": 8.374346408439411e-05, "loss": 0.0316, "step": 30100 }, { "grad_norm": 0.19971302151679993, "learning_rate": 8.373126066908868e-05, "loss": 0.0252, "step": 30110 }, { "grad_norm": 0.39695239067077637, "learning_rate": 8.371905356498326e-05, "loss": 0.0248, "step": 30120 }, { "grad_norm": 0.3190678656101227, "learning_rate": 8.370684277341288e-05, "loss": 0.0275, "step": 30130 }, { "grad_norm": 0.2694339454174042, "learning_rate": 8.369462829571282e-05, "loss": 0.0231, "step": 30140 }, { "grad_norm": 0.22593972086906433, "learning_rate": 8.36824101332189e-05, "loss": 0.0265, "step": 30150 }, { "grad_norm": 0.21906481683254242, "learning_rate": 8.367018828726721e-05, "loss": 0.0245, "step": 30160 }, { "grad_norm": 0.15457826852798462, "learning_rate": 8.365796275919438e-05, "loss": 0.0213, "step": 30170 }, { "grad_norm": 0.2232976108789444, "learning_rate": 8.364573355033734e-05, "loss": 0.0257, "step": 30180 }, { "grad_norm": 0.3069652020931244, "learning_rate": 8.363350066203346e-05, "loss": 0.0257, "step": 30190 }, { "grad_norm": 0.27454283833503723, "learning_rate": 8.362126409562053e-05, "loss": 0.0255, "step": 30200 }, { "grad_norm": 0.3240944743156433, "learning_rate": 8.360902385243667e-05, "loss": 0.0255, "step": 30210 }, { "grad_norm": 0.24627947807312012, "learning_rate": 8.359677993382052e-05, "loss": 0.0274, "step": 30220 }, { "grad_norm": 0.1812216341495514, "learning_rate": 8.358453234111103e-05, "loss": 0.0213, "step": 30230 }, { "grad_norm": 0.21796640753746033, "learning_rate": 8.357228107564756e-05, "loss": 0.0279, "step": 30240 }, { "grad_norm": 0.16790606081485748, "learning_rate": 8.356002613876993e-05, "loss": 0.0243, "step": 30250 }, { "grad_norm": 0.196576789021492, "learning_rate": 8.35477675318183e-05, "loss": 0.0244, "step": 30260 }, { "grad_norm": 0.2619514465332031, "learning_rate": 8.353550525613323e-05, "loss": 0.0278, "step": 30270 }, { "grad_norm": 0.26098760962486267, "learning_rate": 8.352323931305572e-05, "loss": 0.0251, "step": 30280 }, { "grad_norm": 0.222148597240448, "learning_rate": 8.351096970392717e-05, "loss": 0.0242, "step": 30290 }, { "grad_norm": 0.27632656693458557, "learning_rate": 8.349869643008937e-05, "loss": 0.0267, "step": 30300 }, { "grad_norm": 0.219009131193161, "learning_rate": 8.348641949288449e-05, "loss": 0.0211, "step": 30310 }, { "grad_norm": 0.20300006866455078, "learning_rate": 8.34741388936551e-05, "loss": 0.0241, "step": 30320 }, { "grad_norm": 0.21246029436588287, "learning_rate": 8.346185463374423e-05, "loss": 0.0245, "step": 30330 }, { "grad_norm": 0.28277215361595154, "learning_rate": 8.344956671449524e-05, "loss": 0.0204, "step": 30340 }, { "grad_norm": 0.21814358234405518, "learning_rate": 8.343727513725192e-05, "loss": 0.0216, "step": 30350 }, { "grad_norm": 0.23796914517879486, "learning_rate": 8.342497990335847e-05, "loss": 0.02, "step": 30360 }, { "grad_norm": 0.2147505134344101, "learning_rate": 8.341268101415946e-05, "loss": 0.02, "step": 30370 }, { "grad_norm": 0.32629767060279846, "learning_rate": 8.34003784709999e-05, "loss": 0.0264, "step": 30380 }, { "grad_norm": 0.23016679286956787, "learning_rate": 8.338807227522516e-05, "loss": 0.0201, "step": 30390 }, { "grad_norm": 0.24064892530441284, "learning_rate": 8.337576242818103e-05, "loss": 0.0216, "step": 30400 }, { "grad_norm": 0.1971847116947174, "learning_rate": 8.336344893121372e-05, "loss": 0.0216, "step": 30410 }, { "grad_norm": 0.20785892009735107, "learning_rate": 8.335113178566977e-05, "loss": 0.0237, "step": 30420 }, { "grad_norm": 0.21111755073070526, "learning_rate": 8.33388109928962e-05, "loss": 0.0234, "step": 30430 }, { "grad_norm": 0.32258379459381104, "learning_rate": 8.33264865542404e-05, "loss": 0.0228, "step": 30440 }, { "grad_norm": 0.21860231459140778, "learning_rate": 8.331415847105013e-05, "loss": 0.0222, "step": 30450 }, { "grad_norm": 0.2174740880727768, "learning_rate": 8.330182674467357e-05, "loss": 0.0236, "step": 30460 }, { "grad_norm": 0.2377578616142273, "learning_rate": 8.32894913764593e-05, "loss": 0.0203, "step": 30470 }, { "grad_norm": 0.24763236939907074, "learning_rate": 8.327715236775633e-05, "loss": 0.0219, "step": 30480 }, { "grad_norm": 0.1787666231393814, "learning_rate": 8.326480971991398e-05, "loss": 0.0181, "step": 30490 }, { "grad_norm": 0.21093539893627167, "learning_rate": 8.325246343428206e-05, "loss": 0.0218, "step": 30500 }, { "grad_norm": 0.19989624619483948, "learning_rate": 8.324011351221072e-05, "loss": 0.0185, "step": 30510 }, { "grad_norm": 0.2703414559364319, "learning_rate": 8.322775995505057e-05, "loss": 0.0215, "step": 30520 }, { "grad_norm": 0.16527175903320312, "learning_rate": 8.321540276415254e-05, "loss": 0.0189, "step": 30530 }, { "grad_norm": 0.14915527403354645, "learning_rate": 8.320304194086798e-05, "loss": 0.0211, "step": 30540 }, { "grad_norm": 0.2644028663635254, "learning_rate": 8.31906774865487e-05, "loss": 0.024, "step": 30550 }, { "grad_norm": 0.28782621026039124, "learning_rate": 8.317830940254682e-05, "loss": 0.0213, "step": 30560 }, { "grad_norm": 0.29967886209487915, "learning_rate": 8.316593769021491e-05, "loss": 0.022, "step": 30570 }, { "grad_norm": 0.2628825604915619, "learning_rate": 8.315356235090592e-05, "loss": 0.0216, "step": 30580 }, { "grad_norm": 0.22550803422927856, "learning_rate": 8.314118338597319e-05, "loss": 0.0208, "step": 30590 }, { "grad_norm": 0.2737046778202057, "learning_rate": 8.312880079677048e-05, "loss": 0.0199, "step": 30600 }, { "grad_norm": 0.22785623371601105, "learning_rate": 8.311641458465191e-05, "loss": 0.0205, "step": 30610 }, { "grad_norm": 0.20429863035678864, "learning_rate": 8.310402475097205e-05, "loss": 0.0176, "step": 30620 }, { "grad_norm": 0.25516632199287415, "learning_rate": 8.309163129708581e-05, "loss": 0.0191, "step": 30630 }, { "grad_norm": 0.17379088699817657, "learning_rate": 8.307923422434852e-05, "loss": 0.0202, "step": 30640 }, { "grad_norm": 0.28880634903907776, "learning_rate": 8.30668335341159e-05, "loss": 0.0198, "step": 30650 }, { "grad_norm": 0.2354874312877655, "learning_rate": 8.305442922774408e-05, "loss": 0.021, "step": 30660 }, { "grad_norm": 0.27284035086631775, "learning_rate": 8.304202130658959e-05, "loss": 0.0209, "step": 30670 }, { "grad_norm": 0.20087790489196777, "learning_rate": 8.302960977200931e-05, "loss": 0.0234, "step": 30680 }, { "grad_norm": 0.19563306868076324, "learning_rate": 8.301719462536058e-05, "loss": 0.0199, "step": 30690 }, { "grad_norm": 0.2253398895263672, "learning_rate": 8.300477586800108e-05, "loss": 0.0205, "step": 30700 }, { "grad_norm": 0.16880916059017181, "learning_rate": 8.299235350128892e-05, "loss": 0.0201, "step": 30710 }, { "grad_norm": 0.19277295470237732, "learning_rate": 8.297992752658258e-05, "loss": 0.0216, "step": 30720 }, { "grad_norm": 0.249139204621315, "learning_rate": 8.296749794524094e-05, "loss": 0.0266, "step": 30730 }, { "grad_norm": 0.4901202619075775, "learning_rate": 8.295506475862332e-05, "loss": 0.0219, "step": 30740 }, { "grad_norm": 0.2051498144865036, "learning_rate": 8.294262796808933e-05, "loss": 0.0247, "step": 30750 }, { "grad_norm": 0.22205789387226105, "learning_rate": 8.293018757499909e-05, "loss": 0.0274, "step": 30760 }, { "grad_norm": 0.21664968132972717, "learning_rate": 8.291774358071305e-05, "loss": 0.0221, "step": 30770 }, { "grad_norm": 0.19644026458263397, "learning_rate": 8.290529598659205e-05, "loss": 0.0195, "step": 30780 }, { "grad_norm": 0.27847275137901306, "learning_rate": 8.289284479399738e-05, "loss": 0.0233, "step": 30790 }, { "grad_norm": 0.2348603755235672, "learning_rate": 8.288039000429064e-05, "loss": 0.0233, "step": 30800 }, { "grad_norm": 0.1930682510137558, "learning_rate": 8.286793161883388e-05, "loss": 0.0236, "step": 30810 }, { "grad_norm": 0.23807977139949799, "learning_rate": 8.285546963898954e-05, "loss": 0.0224, "step": 30820 }, { "grad_norm": 0.2806253135204315, "learning_rate": 8.284300406612044e-05, "loss": 0.0227, "step": 30830 }, { "grad_norm": 0.26380452513694763, "learning_rate": 8.283053490158978e-05, "loss": 0.0181, "step": 30840 }, { "grad_norm": 0.18230117857456207, "learning_rate": 8.28180621467612e-05, "loss": 0.0196, "step": 30850 }, { "grad_norm": 0.16113431751728058, "learning_rate": 8.280558580299867e-05, "loss": 0.0216, "step": 30860 }, { "grad_norm": 0.2155708372592926, "learning_rate": 8.279310587166661e-05, "loss": 0.0242, "step": 30870 }, { "grad_norm": 0.2623702585697174, "learning_rate": 8.278062235412978e-05, "loss": 0.0208, "step": 30880 }, { "grad_norm": 0.19819258153438568, "learning_rate": 8.276813525175339e-05, "loss": 0.0195, "step": 30890 }, { "grad_norm": 0.17564454674720764, "learning_rate": 8.2755644565903e-05, "loss": 0.0201, "step": 30900 }, { "grad_norm": 0.21448025107383728, "learning_rate": 8.274315029794454e-05, "loss": 0.0261, "step": 30910 }, { "grad_norm": 0.21766901016235352, "learning_rate": 8.273065244924443e-05, "loss": 0.0171, "step": 30920 }, { "grad_norm": 0.2367875874042511, "learning_rate": 8.271815102116936e-05, "loss": 0.025, "step": 30930 }, { "grad_norm": 0.23434682190418243, "learning_rate": 8.270564601508648e-05, "loss": 0.0234, "step": 30940 }, { "grad_norm": 0.24240325391292572, "learning_rate": 8.269313743236333e-05, "loss": 0.0236, "step": 30950 }, { "grad_norm": 0.20708879828453064, "learning_rate": 8.268062527436783e-05, "loss": 0.0196, "step": 30960 }, { "grad_norm": 0.2317342460155487, "learning_rate": 8.266810954246828e-05, "loss": 0.0263, "step": 30970 }, { "grad_norm": 0.16471602022647858, "learning_rate": 8.265559023803338e-05, "loss": 0.0229, "step": 30980 }, { "grad_norm": 0.20005658268928528, "learning_rate": 8.264306736243223e-05, "loss": 0.0229, "step": 30990 }, { "grad_norm": 0.19571608304977417, "learning_rate": 8.263054091703432e-05, "loss": 0.0256, "step": 31000 }, { "grad_norm": 0.22213976085186005, "learning_rate": 8.26180109032095e-05, "loss": 0.0223, "step": 31010 }, { "grad_norm": 0.22773787379264832, "learning_rate": 8.260547732232807e-05, "loss": 0.0254, "step": 31020 }, { "grad_norm": 0.19475291669368744, "learning_rate": 8.259294017576063e-05, "loss": 0.0196, "step": 31030 }, { "grad_norm": 0.20678462088108063, "learning_rate": 8.258039946487828e-05, "loss": 0.0212, "step": 31040 }, { "grad_norm": 0.1676146686077118, "learning_rate": 8.256785519105241e-05, "loss": 0.0202, "step": 31050 }, { "grad_norm": 0.22878605127334595, "learning_rate": 8.255530735565488e-05, "loss": 0.0185, "step": 31060 }, { "grad_norm": 0.2233233004808426, "learning_rate": 8.254275596005787e-05, "loss": 0.0241, "step": 31070 }, { "grad_norm": 0.2429923117160797, "learning_rate": 8.2530201005634e-05, "loss": 0.0196, "step": 31080 }, { "grad_norm": 0.20280759036540985, "learning_rate": 8.251764249375626e-05, "loss": 0.0221, "step": 31090 }, { "grad_norm": 0.2286035716533661, "learning_rate": 8.250508042579803e-05, "loss": 0.0215, "step": 31100 }, { "grad_norm": 0.20683690905570984, "learning_rate": 8.249251480313307e-05, "loss": 0.0196, "step": 31110 }, { "grad_norm": 0.18527080118656158, "learning_rate": 8.247994562713555e-05, "loss": 0.0201, "step": 31120 }, { "grad_norm": 0.2963637411594391, "learning_rate": 8.246737289917998e-05, "loss": 0.019, "step": 31130 }, { "grad_norm": 0.24056002497673035, "learning_rate": 8.245479662064135e-05, "loss": 0.0211, "step": 31140 }, { "grad_norm": 0.19870516657829285, "learning_rate": 8.244221679289496e-05, "loss": 0.0174, "step": 31150 }, { "grad_norm": 0.17230026423931122, "learning_rate": 8.242963341731652e-05, "loss": 0.0182, "step": 31160 }, { "grad_norm": 0.1841433048248291, "learning_rate": 8.24170464952821e-05, "loss": 0.0196, "step": 31170 }, { "grad_norm": 0.1973962038755417, "learning_rate": 8.240445602816824e-05, "loss": 0.021, "step": 31180 }, { "grad_norm": 0.19313277304172516, "learning_rate": 8.239186201735179e-05, "loss": 0.0231, "step": 31190 }, { "grad_norm": 0.1951691210269928, "learning_rate": 8.237926446420998e-05, "loss": 0.0171, "step": 31200 }, { "grad_norm": 0.20933736860752106, "learning_rate": 8.236666337012052e-05, "loss": 0.019, "step": 31210 }, { "grad_norm": 0.18574368953704834, "learning_rate": 8.23540587364614e-05, "loss": 0.0204, "step": 31220 }, { "grad_norm": 0.17421317100524902, "learning_rate": 8.234145056461107e-05, "loss": 0.0206, "step": 31230 }, { "grad_norm": 0.1598932147026062, "learning_rate": 8.232883885594832e-05, "loss": 0.0175, "step": 31240 }, { "grad_norm": 0.29071953892707825, "learning_rate": 8.231622361185236e-05, "loss": 0.0214, "step": 31250 }, { "grad_norm": 0.2252594381570816, "learning_rate": 8.230360483370278e-05, "loss": 0.0203, "step": 31260 }, { "grad_norm": 0.28133487701416016, "learning_rate": 8.229098252287953e-05, "loss": 0.0187, "step": 31270 }, { "grad_norm": 0.2642437815666199, "learning_rate": 8.2278356680763e-05, "loss": 0.0245, "step": 31280 }, { "grad_norm": 0.2186388522386551, "learning_rate": 8.22657273087339e-05, "loss": 0.0213, "step": 31290 }, { "grad_norm": 0.2267298847436905, "learning_rate": 8.225309440817336e-05, "loss": 0.0212, "step": 31300 }, { "grad_norm": 0.2882141172885895, "learning_rate": 8.224045798046293e-05, "loss": 0.0231, "step": 31310 }, { "grad_norm": 0.19076064229011536, "learning_rate": 8.22278180269845e-05, "loss": 0.0198, "step": 31320 }, { "grad_norm": 0.23246066272258759, "learning_rate": 8.221517454912031e-05, "loss": 0.0223, "step": 31330 }, { "grad_norm": 0.16322380304336548, "learning_rate": 8.220252754825308e-05, "loss": 0.0199, "step": 31340 }, { "grad_norm": 0.21992406249046326, "learning_rate": 8.218987702576586e-05, "loss": 0.0185, "step": 31350 }, { "grad_norm": 0.18576399981975555, "learning_rate": 8.217722298304207e-05, "loss": 0.0232, "step": 31360 }, { "grad_norm": 0.14289337396621704, "learning_rate": 8.216456542146557e-05, "loss": 0.0197, "step": 31370 }, { "grad_norm": 0.21647007763385773, "learning_rate": 8.215190434242055e-05, "loss": 0.0193, "step": 31380 }, { "grad_norm": 0.1968323290348053, "learning_rate": 8.213923974729161e-05, "loss": 0.0202, "step": 31390 }, { "grad_norm": 0.22126775979995728, "learning_rate": 8.212657163746373e-05, "loss": 0.0165, "step": 31400 }, { "grad_norm": 0.16649238765239716, "learning_rate": 8.211390001432227e-05, "loss": 0.0189, "step": 31410 }, { "grad_norm": 0.1833125799894333, "learning_rate": 8.210122487925297e-05, "loss": 0.0203, "step": 31420 }, { "grad_norm": 0.22337135672569275, "learning_rate": 8.208854623364202e-05, "loss": 0.0249, "step": 31430 }, { "grad_norm": 0.24251915514469147, "learning_rate": 8.207586407887589e-05, "loss": 0.0208, "step": 31440 }, { "grad_norm": 0.2142203450202942, "learning_rate": 8.206317841634148e-05, "loss": 0.0193, "step": 31450 }, { "grad_norm": 0.3382686674594879, "learning_rate": 8.205048924742609e-05, "loss": 0.0256, "step": 31460 }, { "grad_norm": 0.20039105415344238, "learning_rate": 8.203779657351738e-05, "loss": 0.0242, "step": 31470 }, { "grad_norm": 0.21219295263290405, "learning_rate": 8.20251003960034e-05, "loss": 0.0227, "step": 31480 }, { "grad_norm": 0.20966552197933197, "learning_rate": 8.201240071627258e-05, "loss": 0.0214, "step": 31490 }, { "grad_norm": 0.2174658477306366, "learning_rate": 8.199969753571377e-05, "loss": 0.0232, "step": 31500 }, { "grad_norm": 0.1600358486175537, "learning_rate": 8.198699085571615e-05, "loss": 0.018, "step": 31510 }, { "grad_norm": 0.21166376769542694, "learning_rate": 8.197428067766928e-05, "loss": 0.0256, "step": 31520 }, { "grad_norm": 0.18767744302749634, "learning_rate": 8.196156700296316e-05, "loss": 0.0215, "step": 31530 }, { "grad_norm": 0.27048537135124207, "learning_rate": 8.194884983298814e-05, "loss": 0.026, "step": 31540 }, { "grad_norm": 0.29442062973976135, "learning_rate": 8.193612916913491e-05, "loss": 0.0219, "step": 31550 }, { "grad_norm": 0.2861234247684479, "learning_rate": 8.192340501279463e-05, "loss": 0.026, "step": 31560 }, { "grad_norm": 0.1975843459367752, "learning_rate": 8.191067736535876e-05, "loss": 0.021, "step": 31570 }, { "grad_norm": 0.24222606420516968, "learning_rate": 8.18979462282192e-05, "loss": 0.0273, "step": 31580 }, { "grad_norm": 0.2728392481803894, "learning_rate": 8.188521160276819e-05, "loss": 0.0235, "step": 31590 }, { "grad_norm": 0.2168155461549759, "learning_rate": 8.187247349039837e-05, "loss": 0.0194, "step": 31600 }, { "grad_norm": 0.2274656593799591, "learning_rate": 8.185973189250278e-05, "loss": 0.0188, "step": 31610 }, { "grad_norm": 0.20151758193969727, "learning_rate": 8.184698681047482e-05, "loss": 0.0184, "step": 31620 }, { "grad_norm": 0.24731627106666565, "learning_rate": 8.183423824570827e-05, "loss": 0.0295, "step": 31630 }, { "grad_norm": 0.1299542486667633, "learning_rate": 8.182148619959725e-05, "loss": 0.0187, "step": 31640 }, { "grad_norm": 0.2197977751493454, "learning_rate": 8.180873067353636e-05, "loss": 0.0185, "step": 31650 }, { "grad_norm": 0.21641579270362854, "learning_rate": 8.179597166892052e-05, "loss": 0.0189, "step": 31660 }, { "grad_norm": 0.2177867740392685, "learning_rate": 8.178320918714501e-05, "loss": 0.0292, "step": 31670 }, { "grad_norm": 0.23976269364356995, "learning_rate": 8.177044322960554e-05, "loss": 0.0208, "step": 31680 }, { "grad_norm": 0.24636557698249817, "learning_rate": 8.175767379769816e-05, "loss": 0.0244, "step": 31690 }, { "grad_norm": 0.2435697317123413, "learning_rate": 8.174490089281932e-05, "loss": 0.0185, "step": 31700 }, { "grad_norm": 0.2266402691602707, "learning_rate": 8.173212451636584e-05, "loss": 0.0217, "step": 31710 }, { "grad_norm": 0.20431002974510193, "learning_rate": 8.171934466973493e-05, "loss": 0.0231, "step": 31720 }, { "grad_norm": 0.22732959687709808, "learning_rate": 8.170656135432418e-05, "loss": 0.021, "step": 31730 }, { "grad_norm": 0.2666107714176178, "learning_rate": 8.169377457153155e-05, "loss": 0.0213, "step": 31740 }, { "grad_norm": 0.1965900957584381, "learning_rate": 8.168098432275539e-05, "loss": 0.0284, "step": 31750 }, { "grad_norm": 0.2792988717556, "learning_rate": 8.166819060939442e-05, "loss": 0.0235, "step": 31760 }, { "grad_norm": 0.243368998169899, "learning_rate": 8.165539343284772e-05, "loss": 0.0272, "step": 31770 }, { "grad_norm": 0.26313158869743347, "learning_rate": 8.16425927945148e-05, "loss": 0.0219, "step": 31780 }, { "grad_norm": 0.22693437337875366, "learning_rate": 8.162978869579551e-05, "loss": 0.025, "step": 31790 }, { "grad_norm": 0.23369631171226501, "learning_rate": 8.161698113809007e-05, "loss": 0.0263, "step": 31800 }, { "grad_norm": 0.18253156542778015, "learning_rate": 8.160417012279911e-05, "loss": 0.0248, "step": 31810 }, { "grad_norm": 0.1678922474384308, "learning_rate": 8.159135565132363e-05, "loss": 0.0219, "step": 31820 }, { "grad_norm": 0.2779473066329956, "learning_rate": 8.157853772506498e-05, "loss": 0.0252, "step": 31830 }, { "grad_norm": 0.2909264862537384, "learning_rate": 8.156571634542494e-05, "loss": 0.0274, "step": 31840 }, { "grad_norm": 0.23555594682693481, "learning_rate": 8.15528915138056e-05, "loss": 0.023, "step": 31850 }, { "grad_norm": 0.19871719181537628, "learning_rate": 8.154006323160949e-05, "loss": 0.0233, "step": 31860 }, { "grad_norm": 0.28647878766059875, "learning_rate": 8.152723150023949e-05, "loss": 0.0225, "step": 31870 }, { "grad_norm": 0.29744312167167664, "learning_rate": 8.151439632109886e-05, "loss": 0.0227, "step": 31880 }, { "grad_norm": 0.24099774658679962, "learning_rate": 8.150155769559122e-05, "loss": 0.0255, "step": 31890 }, { "grad_norm": 0.22589533030986786, "learning_rate": 8.148871562512058e-05, "loss": 0.0255, "step": 31900 }, { "grad_norm": 0.6185511946678162, "learning_rate": 8.147587011109136e-05, "loss": 0.0259, "step": 31910 }, { "grad_norm": 0.2372545748949051, "learning_rate": 8.14630211549083e-05, "loss": 0.0198, "step": 31920 }, { "grad_norm": 0.258216917514801, "learning_rate": 8.145016875797655e-05, "loss": 0.023, "step": 31930 }, { "grad_norm": 0.1573205441236496, "learning_rate": 8.143731292170164e-05, "loss": 0.019, "step": 31940 }, { "grad_norm": 0.20291149616241455, "learning_rate": 8.142445364748944e-05, "loss": 0.0222, "step": 31950 }, { "grad_norm": 0.35815930366516113, "learning_rate": 8.141159093674624e-05, "loss": 0.0235, "step": 31960 }, { "grad_norm": 0.206199049949646, "learning_rate": 8.139872479087869e-05, "loss": 0.0209, "step": 31970 }, { "grad_norm": 0.2884787917137146, "learning_rate": 8.13858552112938e-05, "loss": 0.0211, "step": 31980 }, { "grad_norm": 0.21580615639686584, "learning_rate": 8.137298219939895e-05, "loss": 0.0243, "step": 31990 }, { "grad_norm": 0.19325833022594452, "learning_rate": 8.136010575660196e-05, "loss": 0.0221, "step": 32000 }, { "grad_norm": 0.19787795841693878, "learning_rate": 8.134722588431095e-05, "loss": 0.0225, "step": 32010 }, { "grad_norm": 0.2124672532081604, "learning_rate": 8.133434258393444e-05, "loss": 0.0248, "step": 32020 }, { "grad_norm": 0.24223259091377258, "learning_rate": 8.132145585688134e-05, "loss": 0.0235, "step": 32030 }, { "grad_norm": 0.21433323621749878, "learning_rate": 8.130856570456093e-05, "loss": 0.0235, "step": 32040 }, { "grad_norm": 0.3140909671783447, "learning_rate": 8.129567212838283e-05, "loss": 0.0227, "step": 32050 }, { "grad_norm": 0.16088126599788666, "learning_rate": 8.128277512975708e-05, "loss": 0.0188, "step": 32060 }, { "grad_norm": 0.1984875500202179, "learning_rate": 8.126987471009408e-05, "loss": 0.0204, "step": 32070 }, { "grad_norm": 0.2005012333393097, "learning_rate": 8.125697087080459e-05, "loss": 0.0199, "step": 32080 }, { "grad_norm": 0.24235296249389648, "learning_rate": 8.124406361329976e-05, "loss": 0.0194, "step": 32090 }, { "grad_norm": 0.24918639659881592, "learning_rate": 8.123115293899111e-05, "loss": 0.021, "step": 32100 }, { "grad_norm": 0.2742067873477936, "learning_rate": 8.121823884929055e-05, "loss": 0.0231, "step": 32110 }, { "grad_norm": 0.20062683522701263, "learning_rate": 8.12053213456103e-05, "loss": 0.0219, "step": 32120 }, { "grad_norm": 0.17913813889026642, "learning_rate": 8.119240042936303e-05, "loss": 0.0203, "step": 32130 }, { "grad_norm": 0.21490153670310974, "learning_rate": 8.117947610196175e-05, "loss": 0.0181, "step": 32140 }, { "grad_norm": 0.21683834493160248, "learning_rate": 8.116654836481982e-05, "loss": 0.0205, "step": 32150 }, { "grad_norm": 0.23283450305461884, "learning_rate": 8.115361721935106e-05, "loss": 0.0231, "step": 32160 }, { "grad_norm": 0.2209482640028, "learning_rate": 8.114068266696953e-05, "loss": 0.0196, "step": 32170 }, { "grad_norm": 0.3198429346084595, "learning_rate": 8.112774470908978e-05, "loss": 0.0214, "step": 32180 }, { "grad_norm": 0.32328394055366516, "learning_rate": 8.111480334712665e-05, "loss": 0.0217, "step": 32190 }, { "grad_norm": 0.36830267310142517, "learning_rate": 8.110185858249542e-05, "loss": 0.023, "step": 32200 }, { "grad_norm": 0.2717682719230652, "learning_rate": 8.108891041661168e-05, "loss": 0.0212, "step": 32210 }, { "grad_norm": 0.22226467728614807, "learning_rate": 8.107595885089146e-05, "loss": 0.0205, "step": 32220 }, { "grad_norm": 0.1786375343799591, "learning_rate": 8.106300388675108e-05, "loss": 0.0211, "step": 32230 }, { "grad_norm": 0.5421756505966187, "learning_rate": 8.105004552560731e-05, "loss": 0.0202, "step": 32240 }, { "grad_norm": 0.19684725999832153, "learning_rate": 8.103708376887724e-05, "loss": 0.0238, "step": 32250 }, { "grad_norm": 0.17070166766643524, "learning_rate": 8.102411861797836e-05, "loss": 0.0211, "step": 32260 }, { "grad_norm": 0.1771279126405716, "learning_rate": 8.101115007432851e-05, "loss": 0.0207, "step": 32270 }, { "grad_norm": 0.23467884957790375, "learning_rate": 8.09981781393459e-05, "loss": 0.0219, "step": 32280 }, { "grad_norm": 0.24418166279792786, "learning_rate": 8.098520281444915e-05, "loss": 0.0248, "step": 32290 }, { "grad_norm": 0.28696614503860474, "learning_rate": 8.09722241010572e-05, "loss": 0.0227, "step": 32300 }, { "grad_norm": 0.26177725195884705, "learning_rate": 8.095924200058939e-05, "loss": 0.0234, "step": 32310 }, { "grad_norm": 0.2643950879573822, "learning_rate": 8.094625651446541e-05, "loss": 0.0254, "step": 32320 }, { "grad_norm": 0.20150044560432434, "learning_rate": 8.093326764410536e-05, "loss": 0.0217, "step": 32330 }, { "grad_norm": 0.1613304615020752, "learning_rate": 8.092027539092966e-05, "loss": 0.0227, "step": 32340 }, { "grad_norm": 0.16460424661636353, "learning_rate": 8.090727975635913e-05, "loss": 0.0234, "step": 32350 }, { "grad_norm": 0.21205702424049377, "learning_rate": 8.089428074181497e-05, "loss": 0.0207, "step": 32360 }, { "grad_norm": 0.18315833806991577, "learning_rate": 8.088127834871871e-05, "loss": 0.0188, "step": 32370 }, { "grad_norm": 0.17350314557552338, "learning_rate": 8.086827257849226e-05, "loss": 0.0214, "step": 32380 }, { "grad_norm": 0.1630532294511795, "learning_rate": 8.085526343255795e-05, "loss": 0.0238, "step": 32390 }, { "grad_norm": 0.24890756607055664, "learning_rate": 8.084225091233842e-05, "loss": 0.0245, "step": 32400 }, { "grad_norm": 0.14349254965782166, "learning_rate": 8.082923501925668e-05, "loss": 0.0221, "step": 32410 }, { "grad_norm": 0.1701931357383728, "learning_rate": 8.081621575473617e-05, "loss": 0.0234, "step": 32420 }, { "grad_norm": 0.19062209129333496, "learning_rate": 8.080319312020064e-05, "loss": 0.0249, "step": 32430 }, { "grad_norm": 0.20103782415390015, "learning_rate": 8.079016711707421e-05, "loss": 0.0209, "step": 32440 }, { "grad_norm": 0.23358772695064545, "learning_rate": 8.077713774678139e-05, "loss": 0.0189, "step": 32450 }, { "grad_norm": 0.2543669044971466, "learning_rate": 8.076410501074707e-05, "loss": 0.0199, "step": 32460 }, { "grad_norm": 0.26227763295173645, "learning_rate": 8.075106891039647e-05, "loss": 0.0231, "step": 32470 }, { "grad_norm": 0.23543742299079895, "learning_rate": 8.073802944715523e-05, "loss": 0.021, "step": 32480 }, { "grad_norm": 0.21809357404708862, "learning_rate": 8.072498662244929e-05, "loss": 0.0206, "step": 32490 }, { "grad_norm": 0.2127608358860016, "learning_rate": 8.0711940437705e-05, "loss": 0.0173, "step": 32500 }, { "grad_norm": 0.1794544756412506, "learning_rate": 8.06988908943491e-05, "loss": 0.0206, "step": 32510 }, { "grad_norm": 0.19316169619560242, "learning_rate": 8.068583799380863e-05, "loss": 0.0272, "step": 32520 }, { "grad_norm": 0.22027449309825897, "learning_rate": 8.067278173751104e-05, "loss": 0.0218, "step": 32530 }, { "grad_norm": 0.21820563077926636, "learning_rate": 8.065972212688417e-05, "loss": 0.0189, "step": 32540 }, { "grad_norm": 0.24514277279376984, "learning_rate": 8.064665916335618e-05, "loss": 0.0182, "step": 32550 }, { "grad_norm": 0.24177375435829163, "learning_rate": 8.063359284835564e-05, "loss": 0.0273, "step": 32560 }, { "grad_norm": 0.21408429741859436, "learning_rate": 8.062052318331142e-05, "loss": 0.0203, "step": 32570 }, { "grad_norm": 0.1677471250295639, "learning_rate": 8.060745016965283e-05, "loss": 0.0157, "step": 32580 }, { "grad_norm": 0.21932567656040192, "learning_rate": 8.059437380880952e-05, "loss": 0.022, "step": 32590 }, { "grad_norm": 0.28914159536361694, "learning_rate": 8.058129410221146e-05, "loss": 0.0233, "step": 32600 }, { "grad_norm": 0.2163880616426468, "learning_rate": 8.056821105128908e-05, "loss": 0.0219, "step": 32610 }, { "grad_norm": 0.220377579331398, "learning_rate": 8.05551246574731e-05, "loss": 0.0248, "step": 32620 }, { "grad_norm": 0.2363770753145218, "learning_rate": 8.05420349221946e-05, "loss": 0.023, "step": 32630 }, { "grad_norm": 0.15779583156108856, "learning_rate": 8.05289418468851e-05, "loss": 0.0209, "step": 32640 }, { "grad_norm": 0.219403937458992, "learning_rate": 8.051584543297642e-05, "loss": 0.0197, "step": 32650 }, { "grad_norm": 0.2223692685365677, "learning_rate": 8.050274568190074e-05, "loss": 0.0183, "step": 32660 }, { "grad_norm": 0.21243219077587128, "learning_rate": 8.048964259509067e-05, "loss": 0.02, "step": 32670 }, { "grad_norm": 0.22364288568496704, "learning_rate": 8.047653617397914e-05, "loss": 0.0176, "step": 32680 }, { "grad_norm": 0.3407200276851654, "learning_rate": 8.046342641999941e-05, "loss": 0.0232, "step": 32690 }, { "grad_norm": 0.24452577531337738, "learning_rate": 8.045031333458517e-05, "loss": 0.0215, "step": 32700 }, { "grad_norm": 0.188963383436203, "learning_rate": 8.043719691917047e-05, "loss": 0.0225, "step": 32710 }, { "grad_norm": 0.18393553793430328, "learning_rate": 8.042407717518966e-05, "loss": 0.0196, "step": 32720 }, { "grad_norm": 0.20775939524173737, "learning_rate": 8.041095410407751e-05, "loss": 0.0211, "step": 32730 }, { "grad_norm": 0.20363479852676392, "learning_rate": 8.039782770726913e-05, "loss": 0.0185, "step": 32740 }, { "grad_norm": 0.21143190562725067, "learning_rate": 8.038469798620004e-05, "loss": 0.018, "step": 32750 }, { "grad_norm": 0.19414566457271576, "learning_rate": 8.037156494230604e-05, "loss": 0.0176, "step": 32760 }, { "grad_norm": 0.17538179457187653, "learning_rate": 8.035842857702338e-05, "loss": 0.0193, "step": 32770 }, { "grad_norm": 0.24936510622501373, "learning_rate": 8.03452888917886e-05, "loss": 0.0216, "step": 32780 }, { "grad_norm": 0.33387935161590576, "learning_rate": 8.033214588803866e-05, "loss": 0.025, "step": 32790 }, { "grad_norm": 0.24408403038978577, "learning_rate": 8.031899956721083e-05, "loss": 0.0218, "step": 32800 }, { "grad_norm": 0.2407623529434204, "learning_rate": 8.030584993074282e-05, "loss": 0.022, "step": 32810 }, { "grad_norm": 0.43945279717445374, "learning_rate": 8.02926969800726e-05, "loss": 0.0194, "step": 32820 }, { "grad_norm": 0.2307823896408081, "learning_rate": 8.027954071663859e-05, "loss": 0.0234, "step": 32830 }, { "grad_norm": 0.3144441545009613, "learning_rate": 8.026638114187954e-05, "loss": 0.0232, "step": 32840 }, { "grad_norm": 0.168439120054245, "learning_rate": 8.025321825723456e-05, "loss": 0.0177, "step": 32850 }, { "grad_norm": 0.22446982562541962, "learning_rate": 8.02400520641431e-05, "loss": 0.0197, "step": 32860 }, { "grad_norm": 0.2594309151172638, "learning_rate": 8.022688256404501e-05, "loss": 0.018, "step": 32870 }, { "grad_norm": 0.20547249913215637, "learning_rate": 8.02137097583805e-05, "loss": 0.0163, "step": 32880 }, { "grad_norm": 0.22431960701942444, "learning_rate": 8.02005336485901e-05, "loss": 0.0228, "step": 32890 }, { "grad_norm": 0.19889965653419495, "learning_rate": 8.018735423611476e-05, "loss": 0.0228, "step": 32900 }, { "grad_norm": 0.26813098788261414, "learning_rate": 8.017417152239574e-05, "loss": 0.0208, "step": 32910 }, { "grad_norm": 0.19355428218841553, "learning_rate": 8.01609855088747e-05, "loss": 0.0217, "step": 32920 }, { "grad_norm": 0.1811547875404358, "learning_rate": 8.014779619699362e-05, "loss": 0.0207, "step": 32930 }, { "grad_norm": 0.20144116878509521, "learning_rate": 8.013460358819489e-05, "loss": 0.0158, "step": 32940 }, { "grad_norm": 0.24786394834518433, "learning_rate": 8.01214076839212e-05, "loss": 0.0228, "step": 32950 }, { "grad_norm": 0.206246018409729, "learning_rate": 8.010820848561565e-05, "loss": 0.0179, "step": 32960 }, { "grad_norm": 0.22862975299358368, "learning_rate": 8.009500599472171e-05, "loss": 0.0212, "step": 32970 }, { "grad_norm": 0.23809696733951569, "learning_rate": 8.008180021268314e-05, "loss": 0.0174, "step": 32980 }, { "grad_norm": 0.2064894735813141, "learning_rate": 8.006859114094414e-05, "loss": 0.0201, "step": 32990 }, { "grad_norm": 0.21621173620224, "learning_rate": 8.005537878094921e-05, "loss": 0.0219, "step": 33000 }, { "grad_norm": 0.2725827991962433, "learning_rate": 8.004216313414323e-05, "loss": 0.0188, "step": 33010 }, { "grad_norm": 0.27193355560302734, "learning_rate": 8.002894420197149e-05, "loss": 0.0222, "step": 33020 }, { "grad_norm": 0.24092574417591095, "learning_rate": 8.001572198587954e-05, "loss": 0.021, "step": 33030 }, { "grad_norm": 0.20697353780269623, "learning_rate": 8.000249648731338e-05, "loss": 0.0237, "step": 33040 }, { "grad_norm": 0.24738122522830963, "learning_rate": 7.998926770771928e-05, "loss": 0.024, "step": 33050 }, { "grad_norm": 0.1701715737581253, "learning_rate": 7.997603564854397e-05, "loss": 0.0206, "step": 33060 }, { "grad_norm": 0.15589530766010284, "learning_rate": 7.996280031123448e-05, "loss": 0.0193, "step": 33070 }, { "grad_norm": 0.2381712794303894, "learning_rate": 7.994956169723818e-05, "loss": 0.0201, "step": 33080 }, { "grad_norm": 0.23636750876903534, "learning_rate": 7.993631980800285e-05, "loss": 0.0185, "step": 33090 }, { "grad_norm": 0.2490874081850052, "learning_rate": 7.992307464497659e-05, "loss": 0.0184, "step": 33100 }, { "grad_norm": 0.22168610990047455, "learning_rate": 7.990982620960787e-05, "loss": 0.0201, "step": 33110 }, { "grad_norm": 0.144709974527359, "learning_rate": 7.989657450334554e-05, "loss": 0.0227, "step": 33120 }, { "grad_norm": 0.3626977205276489, "learning_rate": 7.988331952763877e-05, "loss": 0.0239, "step": 33130 }, { "grad_norm": 0.1998131275177002, "learning_rate": 7.987006128393709e-05, "loss": 0.0208, "step": 33140 }, { "grad_norm": 0.20817910134792328, "learning_rate": 7.985679977369043e-05, "loss": 0.0224, "step": 33150 }, { "grad_norm": 0.1893274188041687, "learning_rate": 7.984353499834902e-05, "loss": 0.0181, "step": 33160 }, { "grad_norm": 0.2170882225036621, "learning_rate": 7.983026695936351e-05, "loss": 0.0217, "step": 33170 }, { "grad_norm": 0.19246463477611542, "learning_rate": 7.981699565818486e-05, "loss": 0.0186, "step": 33180 }, { "grad_norm": 0.25310543179512024, "learning_rate": 7.980372109626437e-05, "loss": 0.0188, "step": 33190 }, { "grad_norm": 0.19502027332782745, "learning_rate": 7.979044327505375e-05, "loss": 0.0175, "step": 33200 }, { "grad_norm": 0.26382914185523987, "learning_rate": 7.977716219600506e-05, "loss": 0.0193, "step": 33210 }, { "grad_norm": 0.14434270560741425, "learning_rate": 7.97638778605707e-05, "loss": 0.0161, "step": 33220 }, { "grad_norm": 0.2205275148153305, "learning_rate": 7.975059027020338e-05, "loss": 0.0198, "step": 33230 }, { "grad_norm": 0.22541387379169464, "learning_rate": 7.973729942635623e-05, "loss": 0.0206, "step": 33240 }, { "grad_norm": 0.24987423419952393, "learning_rate": 7.972400533048273e-05, "loss": 0.024, "step": 33250 }, { "grad_norm": 0.2311227172613144, "learning_rate": 7.97107079840367e-05, "loss": 0.0196, "step": 33260 }, { "grad_norm": 0.1963399350643158, "learning_rate": 7.969740738847231e-05, "loss": 0.0217, "step": 33270 }, { "grad_norm": 0.3181057274341583, "learning_rate": 7.968410354524411e-05, "loss": 0.0195, "step": 33280 }, { "grad_norm": 0.23174162209033966, "learning_rate": 7.967079645580697e-05, "loss": 0.0225, "step": 33290 }, { "grad_norm": 0.22624866664409637, "learning_rate": 7.965748612161612e-05, "loss": 0.0261, "step": 33300 }, { "grad_norm": 0.24723312258720398, "learning_rate": 7.96441725441272e-05, "loss": 0.021, "step": 33310 }, { "grad_norm": 0.2954545319080353, "learning_rate": 7.963085572479614e-05, "loss": 0.0207, "step": 33320 }, { "grad_norm": 0.19425775110721588, "learning_rate": 7.961753566507924e-05, "loss": 0.0203, "step": 33330 }, { "grad_norm": 0.23400762677192688, "learning_rate": 7.960421236643316e-05, "loss": 0.027, "step": 33340 }, { "grad_norm": 0.1797182708978653, "learning_rate": 7.959088583031496e-05, "loss": 0.0181, "step": 33350 }, { "grad_norm": 0.1824471801519394, "learning_rate": 7.957755605818194e-05, "loss": 0.0203, "step": 33360 }, { "grad_norm": 0.23452632129192352, "learning_rate": 7.956422305149185e-05, "loss": 0.0171, "step": 33370 }, { "grad_norm": 0.2280987948179245, "learning_rate": 7.95508868117028e-05, "loss": 0.0203, "step": 33380 }, { "grad_norm": 0.22345204651355743, "learning_rate": 7.953754734027318e-05, "loss": 0.0176, "step": 33390 }, { "grad_norm": 0.24297413229942322, "learning_rate": 7.952420463866182e-05, "loss": 0.018, "step": 33400 }, { "grad_norm": 0.2979193925857544, "learning_rate": 7.951085870832782e-05, "loss": 0.0185, "step": 33410 }, { "grad_norm": 0.19541387259960175, "learning_rate": 7.949750955073067e-05, "loss": 0.0191, "step": 33420 }, { "grad_norm": 0.28088483214378357, "learning_rate": 7.948415716733022e-05, "loss": 0.02, "step": 33430 }, { "grad_norm": 0.18176503479480743, "learning_rate": 7.947080155958669e-05, "loss": 0.0171, "step": 33440 }, { "grad_norm": 0.2516314387321472, "learning_rate": 7.94574427289606e-05, "loss": 0.0204, "step": 33450 }, { "grad_norm": 0.16781948506832123, "learning_rate": 7.944408067691284e-05, "loss": 0.0166, "step": 33460 }, { "grad_norm": 0.2722166180610657, "learning_rate": 7.943071540490473e-05, "loss": 0.0208, "step": 33470 }, { "grad_norm": 0.28219321370124817, "learning_rate": 7.94173469143978e-05, "loss": 0.024, "step": 33480 }, { "grad_norm": 0.24130234122276306, "learning_rate": 7.940397520685406e-05, "loss": 0.0251, "step": 33490 }, { "grad_norm": 0.20937901735305786, "learning_rate": 7.939060028373577e-05, "loss": 0.0195, "step": 33500 }, { "grad_norm": 0.20697198808193207, "learning_rate": 7.937722214650565e-05, "loss": 0.0176, "step": 33510 }, { "grad_norm": 0.24826879799365997, "learning_rate": 7.936384079662666e-05, "loss": 0.0196, "step": 33520 }, { "grad_norm": 0.20537379384040833, "learning_rate": 7.93504562355622e-05, "loss": 0.0223, "step": 33530 }, { "grad_norm": 0.15995612740516663, "learning_rate": 7.933706846477599e-05, "loss": 0.0206, "step": 33540 }, { "grad_norm": 0.18584905564785004, "learning_rate": 7.932367748573206e-05, "loss": 0.0228, "step": 33550 }, { "grad_norm": 0.22607441246509552, "learning_rate": 7.931028329989485e-05, "loss": 0.0206, "step": 33560 }, { "grad_norm": 0.3193088173866272, "learning_rate": 7.929688590872913e-05, "loss": 0.0208, "step": 33570 }, { "grad_norm": 0.1706942766904831, "learning_rate": 7.928348531370003e-05, "loss": 0.0155, "step": 33580 }, { "grad_norm": 0.23710598051548004, "learning_rate": 7.927008151627297e-05, "loss": 0.019, "step": 33590 }, { "grad_norm": 0.16179552674293518, "learning_rate": 7.925667451791383e-05, "loss": 0.0211, "step": 33600 }, { "grad_norm": 0.20182368159294128, "learning_rate": 7.924326432008874e-05, "loss": 0.0156, "step": 33610 }, { "grad_norm": 0.17870071530342102, "learning_rate": 7.922985092426422e-05, "loss": 0.0155, "step": 33620 }, { "grad_norm": 0.25486132502555847, "learning_rate": 7.921643433190717e-05, "loss": 0.0252, "step": 33630 }, { "grad_norm": 0.20986321568489075, "learning_rate": 7.920301454448478e-05, "loss": 0.0203, "step": 33640 }, { "grad_norm": 0.22263850271701813, "learning_rate": 7.918959156346461e-05, "loss": 0.0169, "step": 33650 }, { "grad_norm": 0.23747088015079498, "learning_rate": 7.91761653903146e-05, "loss": 0.0213, "step": 33660 }, { "grad_norm": 0.19611044228076935, "learning_rate": 7.916273602650302e-05, "loss": 0.0213, "step": 33670 }, { "grad_norm": 0.1869027465581894, "learning_rate": 7.914930347349847e-05, "loss": 0.018, "step": 33680 }, { "grad_norm": 0.31173989176750183, "learning_rate": 7.913586773276992e-05, "loss": 0.0208, "step": 33690 }, { "grad_norm": 0.2736964821815491, "learning_rate": 7.912242880578667e-05, "loss": 0.0218, "step": 33700 }, { "grad_norm": 0.2002171128988266, "learning_rate": 7.910898669401839e-05, "loss": 0.0195, "step": 33710 }, { "grad_norm": 0.20520807802677155, "learning_rate": 7.909554139893511e-05, "loss": 0.0199, "step": 33720 }, { "grad_norm": 0.19360487163066864, "learning_rate": 7.908209292200715e-05, "loss": 0.0203, "step": 33730 }, { "grad_norm": 0.21115157008171082, "learning_rate": 7.906864126470523e-05, "loss": 0.0219, "step": 33740 }, { "grad_norm": 0.25859537720680237, "learning_rate": 7.905518642850041e-05, "loss": 0.0196, "step": 33750 }, { "grad_norm": 0.18934904038906097, "learning_rate": 7.904172841486409e-05, "loss": 0.0197, "step": 33760 }, { "grad_norm": 0.17312012612819672, "learning_rate": 7.902826722526801e-05, "loss": 0.025, "step": 33770 }, { "grad_norm": 0.14571277797222137, "learning_rate": 7.901480286118427e-05, "loss": 0.0149, "step": 33780 }, { "grad_norm": 0.16766946017742157, "learning_rate": 7.900133532408531e-05, "loss": 0.0191, "step": 33790 }, { "grad_norm": 0.20517809689044952, "learning_rate": 7.898786461544395e-05, "loss": 0.0187, "step": 33800 }, { "grad_norm": 0.19513121247291565, "learning_rate": 7.897439073673325e-05, "loss": 0.0166, "step": 33810 }, { "grad_norm": 0.19682501256465912, "learning_rate": 7.896091368942677e-05, "loss": 0.0192, "step": 33820 }, { "grad_norm": 0.19345425069332123, "learning_rate": 7.894743347499832e-05, "loss": 0.0232, "step": 33830 }, { "grad_norm": 0.22241449356079102, "learning_rate": 7.893395009492203e-05, "loss": 0.0205, "step": 33840 }, { "grad_norm": 0.24797379970550537, "learning_rate": 7.892046355067248e-05, "loss": 0.0196, "step": 33850 }, { "grad_norm": 0.22000668942928314, "learning_rate": 7.890697384372451e-05, "loss": 0.017, "step": 33860 }, { "grad_norm": 0.1953284740447998, "learning_rate": 7.889348097555336e-05, "loss": 0.019, "step": 33870 }, { "grad_norm": 0.18072766065597534, "learning_rate": 7.887998494763455e-05, "loss": 0.0205, "step": 33880 }, { "grad_norm": 0.2271917462348938, "learning_rate": 7.886648576144404e-05, "loss": 0.0214, "step": 33890 }, { "grad_norm": 0.19611704349517822, "learning_rate": 7.885298341845802e-05, "loss": 0.0242, "step": 33900 }, { "grad_norm": 0.22892406582832336, "learning_rate": 7.883947792015311e-05, "loss": 0.0198, "step": 33910 }, { "grad_norm": 0.25440385937690735, "learning_rate": 7.882596926800628e-05, "loss": 0.0263, "step": 33920 }, { "grad_norm": 0.2370610535144806, "learning_rate": 7.881245746349477e-05, "loss": 0.0234, "step": 33930 }, { "grad_norm": 0.2700413465499878, "learning_rate": 7.879894250809623e-05, "loss": 0.0249, "step": 33940 }, { "grad_norm": 0.21733196079730988, "learning_rate": 7.878542440328865e-05, "loss": 0.0253, "step": 33950 }, { "grad_norm": 0.26468753814697266, "learning_rate": 7.877190315055031e-05, "loss": 0.0229, "step": 33960 }, { "grad_norm": 0.22786623239517212, "learning_rate": 7.875837875135991e-05, "loss": 0.0191, "step": 33970 }, { "grad_norm": 0.2438778281211853, "learning_rate": 7.874485120719646e-05, "loss": 0.0191, "step": 33980 }, { "grad_norm": 0.2399483323097229, "learning_rate": 7.873132051953928e-05, "loss": 0.0203, "step": 33990 }, { "grad_norm": 0.252213716506958, "learning_rate": 7.87177866898681e-05, "loss": 0.0215, "step": 34000 }, { "grad_norm": 0.19604283571243286, "learning_rate": 7.870424971966294e-05, "loss": 0.0251, "step": 34010 }, { "grad_norm": 0.2321118265390396, "learning_rate": 7.869070961040419e-05, "loss": 0.0239, "step": 34020 }, { "grad_norm": 0.6836498379707336, "learning_rate": 7.867716636357257e-05, "loss": 0.0205, "step": 34030 }, { "grad_norm": 0.19585463404655457, "learning_rate": 7.866361998064915e-05, "loss": 0.021, "step": 34040 }, { "grad_norm": 0.3839249610900879, "learning_rate": 7.865007046311534e-05, "loss": 0.0225, "step": 34050 }, { "grad_norm": 0.2772027254104614, "learning_rate": 7.86365178124529e-05, "loss": 0.0266, "step": 34060 }, { "grad_norm": 0.1958121359348297, "learning_rate": 7.862296203014394e-05, "loss": 0.0246, "step": 34070 }, { "grad_norm": 0.19216415286064148, "learning_rate": 7.860940311767088e-05, "loss": 0.0259, "step": 34080 }, { "grad_norm": 0.41831377148628235, "learning_rate": 7.85958410765165e-05, "loss": 0.0282, "step": 34090 }, { "grad_norm": 0.26842257380485535, "learning_rate": 7.858227590816394e-05, "loss": 0.0199, "step": 34100 }, { "grad_norm": 0.18758489191532135, "learning_rate": 7.856870761409664e-05, "loss": 0.0201, "step": 34110 }, { "grad_norm": 0.381577730178833, "learning_rate": 7.855513619579846e-05, "loss": 0.0277, "step": 34120 }, { "grad_norm": 0.2509455978870392, "learning_rate": 7.85415616547535e-05, "loss": 0.0247, "step": 34130 }, { "grad_norm": 0.21574178338050842, "learning_rate": 7.852798399244627e-05, "loss": 0.02, "step": 34140 }, { "grad_norm": 0.22977638244628906, "learning_rate": 7.851440321036161e-05, "loss": 0.0171, "step": 34150 }, { "grad_norm": 0.1807182878255844, "learning_rate": 7.850081930998468e-05, "loss": 0.0195, "step": 34160 }, { "grad_norm": 0.25418299436569214, "learning_rate": 7.8487232292801e-05, "loss": 0.0206, "step": 34170 }, { "grad_norm": 0.20419667661190033, "learning_rate": 7.847364216029642e-05, "loss": 0.0203, "step": 34180 }, { "grad_norm": 0.23885099589824677, "learning_rate": 7.846004891395716e-05, "loss": 0.0184, "step": 34190 }, { "grad_norm": 0.23131659626960754, "learning_rate": 7.844645255526972e-05, "loss": 0.024, "step": 34200 }, { "grad_norm": 0.24433696269989014, "learning_rate": 7.843285308572101e-05, "loss": 0.0233, "step": 34210 }, { "grad_norm": 0.24220232665538788, "learning_rate": 7.841925050679823e-05, "loss": 0.0188, "step": 34220 }, { "grad_norm": 0.269661545753479, "learning_rate": 7.840564481998895e-05, "loss": 0.0217, "step": 34230 }, { "grad_norm": 0.24252869188785553, "learning_rate": 7.839203602678105e-05, "loss": 0.0225, "step": 34240 }, { "grad_norm": 0.24114298820495605, "learning_rate": 7.837842412866279e-05, "loss": 0.0182, "step": 34250 }, { "grad_norm": 0.25191447138786316, "learning_rate": 7.836480912712272e-05, "loss": 0.0187, "step": 34260 }, { "grad_norm": 0.20429091155529022, "learning_rate": 7.835119102364976e-05, "loss": 0.0239, "step": 34270 }, { "grad_norm": 0.1918024867773056, "learning_rate": 7.83375698197332e-05, "loss": 0.0162, "step": 34280 }, { "grad_norm": 0.2947743237018585, "learning_rate": 7.83239455168626e-05, "loss": 0.0208, "step": 34290 }, { "grad_norm": 0.277107298374176, "learning_rate": 7.83103181165279e-05, "loss": 0.0232, "step": 34300 }, { "grad_norm": 0.18067345023155212, "learning_rate": 7.829668762021937e-05, "loss": 0.0234, "step": 34310 }, { "grad_norm": 0.1941344141960144, "learning_rate": 7.828305402942764e-05, "loss": 0.0199, "step": 34320 }, { "grad_norm": 0.2402196228504181, "learning_rate": 7.826941734564363e-05, "loss": 0.0164, "step": 34330 }, { "grad_norm": 0.24026736617088318, "learning_rate": 7.825577757035865e-05, "loss": 0.0185, "step": 34340 }, { "grad_norm": 0.23533794283866882, "learning_rate": 7.824213470506431e-05, "loss": 0.0194, "step": 34350 }, { "grad_norm": 0.20814983546733856, "learning_rate": 7.822848875125257e-05, "loss": 0.0182, "step": 34360 }, { "grad_norm": 0.2660972476005554, "learning_rate": 7.821483971041576e-05, "loss": 0.0232, "step": 34370 }, { "grad_norm": 0.22360649704933167, "learning_rate": 7.820118758404649e-05, "loss": 0.0223, "step": 34380 }, { "grad_norm": 0.21100768446922302, "learning_rate": 7.818753237363776e-05, "loss": 0.0184, "step": 34390 }, { "grad_norm": 0.22421520948410034, "learning_rate": 7.817387408068286e-05, "loss": 0.0187, "step": 34400 }, { "grad_norm": 0.3177177906036377, "learning_rate": 7.816021270667544e-05, "loss": 0.0218, "step": 34410 }, { "grad_norm": 0.22651876509189606, "learning_rate": 7.81465482531095e-05, "loss": 0.02, "step": 34420 }, { "grad_norm": 0.27380767464637756, "learning_rate": 7.813288072147938e-05, "loss": 0.0233, "step": 34430 }, { "grad_norm": 0.29703041911125183, "learning_rate": 7.811921011327972e-05, "loss": 0.0257, "step": 34440 }, { "grad_norm": 0.24099166691303253, "learning_rate": 7.810553643000549e-05, "loss": 0.0246, "step": 34450 }, { "grad_norm": 0.2925977110862732, "learning_rate": 7.809185967315206e-05, "loss": 0.0244, "step": 34460 }, { "grad_norm": 0.2338685691356659, "learning_rate": 7.80781798442151e-05, "loss": 0.0247, "step": 34470 }, { "grad_norm": 0.24643298983573914, "learning_rate": 7.806449694469061e-05, "loss": 0.023, "step": 34480 }, { "grad_norm": 0.23322492837905884, "learning_rate": 7.805081097607492e-05, "loss": 0.0179, "step": 34490 }, { "grad_norm": 0.18393319845199585, "learning_rate": 7.803712193986474e-05, "loss": 0.0198, "step": 34500 }, { "grad_norm": 0.3713056743144989, "learning_rate": 7.802342983755702e-05, "loss": 0.0219, "step": 34510 }, { "grad_norm": 0.23854000866413116, "learning_rate": 7.800973467064918e-05, "loss": 0.0207, "step": 34520 }, { "grad_norm": 0.22350861132144928, "learning_rate": 7.799603644063884e-05, "loss": 0.0215, "step": 34530 }, { "grad_norm": 0.21825188398361206, "learning_rate": 7.798233514902405e-05, "loss": 0.0219, "step": 34540 }, { "grad_norm": 0.1869673877954483, "learning_rate": 7.796863079730318e-05, "loss": 0.0197, "step": 34550 }, { "grad_norm": 0.14928966760635376, "learning_rate": 7.795492338697488e-05, "loss": 0.0173, "step": 34560 }, { "grad_norm": 0.2116921842098236, "learning_rate": 7.794121291953819e-05, "loss": 0.0181, "step": 34570 }, { "grad_norm": 0.17546969652175903, "learning_rate": 7.792749939649246e-05, "loss": 0.0179, "step": 34580 }, { "grad_norm": 0.2525438070297241, "learning_rate": 7.79137828193374e-05, "loss": 0.0199, "step": 34590 }, { "grad_norm": 0.19920940697193146, "learning_rate": 7.790006318957301e-05, "loss": 0.0199, "step": 34600 }, { "grad_norm": 0.1966814547777176, "learning_rate": 7.788634050869965e-05, "loss": 0.0175, "step": 34610 }, { "grad_norm": 0.21775339543819427, "learning_rate": 7.787261477821803e-05, "loss": 0.0192, "step": 34620 }, { "grad_norm": 0.15778909623622894, "learning_rate": 7.785888599962916e-05, "loss": 0.0199, "step": 34630 }, { "grad_norm": 0.205742746591568, "learning_rate": 7.784515417443439e-05, "loss": 0.0167, "step": 34640 }, { "grad_norm": 0.24066531658172607, "learning_rate": 7.783141930413545e-05, "loss": 0.0213, "step": 34650 }, { "grad_norm": 0.24005749821662903, "learning_rate": 7.78176813902343e-05, "loss": 0.0182, "step": 34660 }, { "grad_norm": 0.2565382719039917, "learning_rate": 7.780394043423336e-05, "loss": 0.0207, "step": 34670 }, { "grad_norm": 0.13424651324748993, "learning_rate": 7.77901964376353e-05, "loss": 0.0167, "step": 34680 }, { "grad_norm": 0.21734817326068878, "learning_rate": 7.777644940194316e-05, "loss": 0.0192, "step": 34690 }, { "grad_norm": 0.2123565971851349, "learning_rate": 7.776269932866023e-05, "loss": 0.0215, "step": 34700 }, { "grad_norm": 0.20541885495185852, "learning_rate": 7.774894621929026e-05, "loss": 0.0201, "step": 34710 }, { "grad_norm": 0.264737993478775, "learning_rate": 7.773519007533725e-05, "loss": 0.0164, "step": 34720 }, { "grad_norm": 0.2043013572692871, "learning_rate": 7.772143089830556e-05, "loss": 0.019, "step": 34730 }, { "grad_norm": 0.17455072700977325, "learning_rate": 7.770766868969985e-05, "loss": 0.0161, "step": 34740 }, { "grad_norm": 0.17772230505943298, "learning_rate": 7.769390345102518e-05, "loss": 0.0193, "step": 34750 }, { "grad_norm": 0.1720583736896515, "learning_rate": 7.768013518378683e-05, "loss": 0.0155, "step": 34760 }, { "grad_norm": 0.16627320647239685, "learning_rate": 7.766636388949053e-05, "loss": 0.021, "step": 34770 }, { "grad_norm": 0.2031630277633667, "learning_rate": 7.765258956964229e-05, "loss": 0.0178, "step": 34780 }, { "grad_norm": 0.17411139607429504, "learning_rate": 7.76388122257484e-05, "loss": 0.0204, "step": 34790 }, { "grad_norm": 0.19016999006271362, "learning_rate": 7.762503185931558e-05, "loss": 0.0253, "step": 34800 }, { "grad_norm": 0.1666605919599533, "learning_rate": 7.76112484718508e-05, "loss": 0.0183, "step": 34810 }, { "grad_norm": 0.2924649119377136, "learning_rate": 7.75974620648614e-05, "loss": 0.0212, "step": 34820 }, { "grad_norm": 0.23337389528751373, "learning_rate": 7.758367263985503e-05, "loss": 0.0201, "step": 34830 }, { "grad_norm": 0.2083926647901535, "learning_rate": 7.75698801983397e-05, "loss": 0.0175, "step": 34840 }, { "grad_norm": 0.17402122914791107, "learning_rate": 7.755608474182372e-05, "loss": 0.0154, "step": 34850 }, { "grad_norm": 0.24578963220119476, "learning_rate": 7.754228627181574e-05, "loss": 0.0224, "step": 34860 }, { "grad_norm": 0.16300402581691742, "learning_rate": 7.752848478982476e-05, "loss": 0.0189, "step": 34870 }, { "grad_norm": 0.2466905266046524, "learning_rate": 7.751468029736006e-05, "loss": 0.0183, "step": 34880 }, { "grad_norm": 0.20107652246952057, "learning_rate": 7.750087279593129e-05, "loss": 0.016, "step": 34890 }, { "grad_norm": 0.2937315106391907, "learning_rate": 7.748706228704843e-05, "loss": 0.0264, "step": 34900 }, { "grad_norm": 0.22388771176338196, "learning_rate": 7.747324877222176e-05, "loss": 0.0179, "step": 34910 }, { "grad_norm": 0.25394776463508606, "learning_rate": 7.745943225296188e-05, "loss": 0.0179, "step": 34920 }, { "grad_norm": 0.2214740663766861, "learning_rate": 7.744561273077981e-05, "loss": 0.0185, "step": 34930 }, { "grad_norm": 0.2081240862607956, "learning_rate": 7.743179020718678e-05, "loss": 0.0245, "step": 34940 }, { "grad_norm": 0.19053398072719574, "learning_rate": 7.741796468369443e-05, "loss": 0.0219, "step": 34950 }, { "grad_norm": 0.2210371196269989, "learning_rate": 7.740413616181466e-05, "loss": 0.0185, "step": 34960 }, { "grad_norm": 0.2381468415260315, "learning_rate": 7.739030464305978e-05, "loss": 0.0218, "step": 34970 }, { "grad_norm": 0.20037251710891724, "learning_rate": 7.737647012894235e-05, "loss": 0.017, "step": 34980 }, { "grad_norm": 0.15824846923351288, "learning_rate": 7.736263262097532e-05, "loss": 0.0162, "step": 34990 }, { "grad_norm": 0.2880193889141083, "learning_rate": 7.734879212067192e-05, "loss": 0.0203, "step": 35000 }, { "grad_norm": 0.1855764091014862, "learning_rate": 7.733494862954573e-05, "loss": 0.0179, "step": 35010 }, { "grad_norm": 0.18261563777923584, "learning_rate": 7.732110214911066e-05, "loss": 0.018, "step": 35020 }, { "grad_norm": 0.17833343148231506, "learning_rate": 7.730725268088094e-05, "loss": 0.0149, "step": 35030 }, { "grad_norm": 0.21425554156303406, "learning_rate": 7.729340022637112e-05, "loss": 0.0154, "step": 35040 }, { "grad_norm": 0.25197121500968933, "learning_rate": 7.727954478709607e-05, "loss": 0.0196, "step": 35050 }, { "grad_norm": 0.2246478945016861, "learning_rate": 7.726568636457103e-05, "loss": 0.0191, "step": 35060 }, { "grad_norm": 0.21479091048240662, "learning_rate": 7.725182496031153e-05, "loss": 0.0167, "step": 35070 }, { "grad_norm": 0.1904950886964798, "learning_rate": 7.723796057583341e-05, "loss": 0.0192, "step": 35080 }, { "grad_norm": 0.3534807562828064, "learning_rate": 7.722409321265287e-05, "loss": 0.0179, "step": 35090 }, { "grad_norm": 0.18002711236476898, "learning_rate": 7.721022287228645e-05, "loss": 0.0185, "step": 35100 }, { "grad_norm": 0.20732776820659637, "learning_rate": 7.719634955625094e-05, "loss": 0.0211, "step": 35110 }, { "grad_norm": 0.25592541694641113, "learning_rate": 7.718247326606355e-05, "loss": 0.016, "step": 35120 }, { "grad_norm": 0.19598600268363953, "learning_rate": 7.716859400324175e-05, "loss": 0.0202, "step": 35130 }, { "grad_norm": 0.1872100681066513, "learning_rate": 7.715471176930335e-05, "loss": 0.02, "step": 35140 }, { "grad_norm": 0.20215201377868652, "learning_rate": 7.714082656576651e-05, "loss": 0.0191, "step": 35150 }, { "grad_norm": 0.257097989320755, "learning_rate": 7.712693839414968e-05, "loss": 0.0191, "step": 35160 }, { "grad_norm": 0.24721582233905792, "learning_rate": 7.711304725597164e-05, "loss": 0.0213, "step": 35170 }, { "grad_norm": 0.15610267221927643, "learning_rate": 7.709915315275151e-05, "loss": 0.017, "step": 35180 }, { "grad_norm": 0.24463950097560883, "learning_rate": 7.708525608600876e-05, "loss": 0.0164, "step": 35190 }, { "grad_norm": 0.24299263954162598, "learning_rate": 7.707135605726311e-05, "loss": 0.0189, "step": 35200 }, { "grad_norm": 0.24705548584461212, "learning_rate": 7.705745306803466e-05, "loss": 0.0208, "step": 35210 }, { "grad_norm": 0.17170563340187073, "learning_rate": 7.704354711984383e-05, "loss": 0.019, "step": 35220 }, { "grad_norm": 0.2371140420436859, "learning_rate": 7.702963821421133e-05, "loss": 0.0169, "step": 35230 }, { "grad_norm": 0.24908065795898438, "learning_rate": 7.701572635265826e-05, "loss": 0.0193, "step": 35240 }, { "grad_norm": 0.1821272373199463, "learning_rate": 7.700181153670596e-05, "loss": 0.0184, "step": 35250 }, { "grad_norm": 0.2189442664384842, "learning_rate": 7.698789376787616e-05, "loss": 0.0249, "step": 35260 }, { "grad_norm": 0.4101484417915344, "learning_rate": 7.697397304769085e-05, "loss": 0.0254, "step": 35270 }, { "grad_norm": 0.20956499874591827, "learning_rate": 7.696004937767241e-05, "loss": 0.0186, "step": 35280 }, { "grad_norm": 0.21105726063251495, "learning_rate": 7.694612275934352e-05, "loss": 0.0168, "step": 35290 }, { "grad_norm": 0.23384056985378265, "learning_rate": 7.693219319422714e-05, "loss": 0.018, "step": 35300 }, { "grad_norm": 0.20123884081840515, "learning_rate": 7.69182606838466e-05, "loss": 0.017, "step": 35310 }, { "grad_norm": 0.46082305908203125, "learning_rate": 7.690432522972558e-05, "loss": 0.0163, "step": 35320 }, { "grad_norm": 0.21487607061862946, "learning_rate": 7.689038683338796e-05, "loss": 0.0227, "step": 35330 }, { "grad_norm": 0.23768578469753265, "learning_rate": 7.687644549635808e-05, "loss": 0.0164, "step": 35340 }, { "grad_norm": 0.18509645760059357, "learning_rate": 7.686250122016053e-05, "loss": 0.0181, "step": 35350 }, { "grad_norm": 0.20020069181919098, "learning_rate": 7.684855400632022e-05, "loss": 0.0238, "step": 35360 }, { "grad_norm": 0.1783236563205719, "learning_rate": 7.683460385636243e-05, "loss": 0.0213, "step": 35370 }, { "grad_norm": 0.26149481534957886, "learning_rate": 7.68206507718127e-05, "loss": 0.024, "step": 35380 }, { "grad_norm": 0.23314417898654938, "learning_rate": 7.680669475419693e-05, "loss": 0.0211, "step": 35390 }, { "grad_norm": 0.22010144591331482, "learning_rate": 7.679273580504132e-05, "loss": 0.0253, "step": 35400 }, { "grad_norm": 0.24032485485076904, "learning_rate": 7.67787739258724e-05, "loss": 0.0205, "step": 35410 }, { "grad_norm": 0.32674795389175415, "learning_rate": 7.676480911821705e-05, "loss": 0.0188, "step": 35420 }, { "grad_norm": 0.2685925364494324, "learning_rate": 7.675084138360238e-05, "loss": 0.0205, "step": 35430 }, { "grad_norm": 0.19900737702846527, "learning_rate": 7.673687072355592e-05, "loss": 0.0201, "step": 35440 }, { "grad_norm": 0.17598044872283936, "learning_rate": 7.67228971396055e-05, "loss": 0.0192, "step": 35450 }, { "grad_norm": 0.2534797787666321, "learning_rate": 7.670892063327922e-05, "loss": 0.0185, "step": 35460 }, { "grad_norm": 0.18689964711666107, "learning_rate": 7.669494120610552e-05, "loss": 0.0237, "step": 35470 }, { "grad_norm": 0.20413890480995178, "learning_rate": 7.66809588596132e-05, "loss": 0.0188, "step": 35480 }, { "grad_norm": 0.15151557326316833, "learning_rate": 7.666697359533132e-05, "loss": 0.0186, "step": 35490 }, { "grad_norm": 0.2992177903652191, "learning_rate": 7.665298541478932e-05, "loss": 0.0191, "step": 35500 }, { "grad_norm": 0.21627911925315857, "learning_rate": 7.66389943195169e-05, "loss": 0.0211, "step": 35510 }, { "grad_norm": 0.24830429255962372, "learning_rate": 7.66250003110441e-05, "loss": 0.0256, "step": 35520 }, { "grad_norm": 0.2648199200630188, "learning_rate": 7.66110033909013e-05, "loss": 0.0217, "step": 35530 }, { "grad_norm": 0.16699951887130737, "learning_rate": 7.659700356061918e-05, "loss": 0.0179, "step": 35540 }, { "grad_norm": 0.16667476296424866, "learning_rate": 7.658300082172875e-05, "loss": 0.0186, "step": 35550 }, { "grad_norm": 0.2290891706943512, "learning_rate": 7.65689951757613e-05, "loss": 0.0178, "step": 35560 }, { "grad_norm": 0.20985209941864014, "learning_rate": 7.655498662424849e-05, "loss": 0.0177, "step": 35570 }, { "grad_norm": 0.2818094491958618, "learning_rate": 7.654097516872227e-05, "loss": 0.0244, "step": 35580 }, { "grad_norm": 0.2128468155860901, "learning_rate": 7.65269608107149e-05, "loss": 0.0165, "step": 35590 }, { "grad_norm": 0.24326708912849426, "learning_rate": 7.6512943551759e-05, "loss": 0.0248, "step": 35600 }, { "grad_norm": 0.3298213481903076, "learning_rate": 7.649892339338743e-05, "loss": 0.0244, "step": 35610 }, { "grad_norm": 0.19935211539268494, "learning_rate": 7.648490033713344e-05, "loss": 0.023, "step": 35620 }, { "grad_norm": 0.20840443670749664, "learning_rate": 7.647087438453058e-05, "loss": 0.0233, "step": 35630 }, { "grad_norm": 0.20412476360797882, "learning_rate": 7.64568455371127e-05, "loss": 0.0196, "step": 35640 }, { "grad_norm": 0.21707960963249207, "learning_rate": 7.644281379641396e-05, "loss": 0.0175, "step": 35650 }, { "grad_norm": 0.17007829248905182, "learning_rate": 7.642877916396887e-05, "loss": 0.0161, "step": 35660 }, { "grad_norm": 0.15857602655887604, "learning_rate": 7.641474164131221e-05, "loss": 0.0223, "step": 35670 }, { "grad_norm": 0.19830434024333954, "learning_rate": 7.640070122997913e-05, "loss": 0.0176, "step": 35680 }, { "grad_norm": 0.22618697583675385, "learning_rate": 7.638665793150505e-05, "loss": 0.0207, "step": 35690 }, { "grad_norm": 0.25436845421791077, "learning_rate": 7.637261174742574e-05, "loss": 0.0173, "step": 35700 }, { "grad_norm": 0.21918214857578278, "learning_rate": 7.635856267927725e-05, "loss": 0.0207, "step": 35710 }, { "grad_norm": 0.21671436727046967, "learning_rate": 7.634451072859597e-05, "loss": 0.02, "step": 35720 }, { "grad_norm": 0.20167770981788635, "learning_rate": 7.633045589691863e-05, "loss": 0.0202, "step": 35730 }, { "grad_norm": 0.18171970546245575, "learning_rate": 7.63163981857822e-05, "loss": 0.0208, "step": 35740 }, { "grad_norm": 0.15614689886569977, "learning_rate": 7.630233759672403e-05, "loss": 0.0168, "step": 35750 }, { "grad_norm": 0.19454185664653778, "learning_rate": 7.628827413128177e-05, "loss": 0.0198, "step": 35760 }, { "grad_norm": 0.16972900927066803, "learning_rate": 7.627420779099338e-05, "loss": 0.0203, "step": 35770 }, { "grad_norm": 0.24505984783172607, "learning_rate": 7.626013857739711e-05, "loss": 0.026, "step": 35780 }, { "grad_norm": 0.2765648663043976, "learning_rate": 7.624606649203158e-05, "loss": 0.0235, "step": 35790 }, { "grad_norm": 0.21279531717300415, "learning_rate": 7.623199153643569e-05, "loss": 0.0235, "step": 35800 }, { "grad_norm": 0.24112732708454132, "learning_rate": 7.621791371214863e-05, "loss": 0.0227, "step": 35810 }, { "grad_norm": 0.23709547519683838, "learning_rate": 7.620383302070995e-05, "loss": 0.0238, "step": 35820 }, { "grad_norm": 0.19789835810661316, "learning_rate": 7.61897494636595e-05, "loss": 0.022, "step": 35830 }, { "grad_norm": 0.20882174372673035, "learning_rate": 7.617566304253739e-05, "loss": 0.0226, "step": 35840 }, { "grad_norm": 0.38286563754081726, "learning_rate": 7.616157375888416e-05, "loss": 0.0178, "step": 35850 }, { "grad_norm": 0.20338693261146545, "learning_rate": 7.614748161424053e-05, "loss": 0.0165, "step": 35860 }, { "grad_norm": 0.28161031007766724, "learning_rate": 7.613338661014763e-05, "loss": 0.0228, "step": 35870 }, { "grad_norm": 0.21958927810192108, "learning_rate": 7.611928874814686e-05, "loss": 0.0203, "step": 35880 }, { "grad_norm": 0.19148044288158417, "learning_rate": 7.610518802977996e-05, "loss": 0.0177, "step": 35890 }, { "grad_norm": 0.22886642813682556, "learning_rate": 7.609108445658893e-05, "loss": 0.0161, "step": 35900 }, { "grad_norm": 0.19217684864997864, "learning_rate": 7.607697803011612e-05, "loss": 0.0181, "step": 35910 }, { "grad_norm": 0.27592039108276367, "learning_rate": 7.606286875190421e-05, "loss": 0.0258, "step": 35920 }, { "grad_norm": 0.14223234355449677, "learning_rate": 7.604875662349617e-05, "loss": 0.0207, "step": 35930 }, { "grad_norm": 0.23271456360816956, "learning_rate": 7.603464164643525e-05, "loss": 0.0182, "step": 35940 }, { "grad_norm": 0.26985543966293335, "learning_rate": 7.602052382226507e-05, "loss": 0.0205, "step": 35950 }, { "grad_norm": 0.24991989135742188, "learning_rate": 7.600640315252954e-05, "loss": 0.0196, "step": 35960 }, { "grad_norm": 0.1835906058549881, "learning_rate": 7.599227963877284e-05, "loss": 0.0197, "step": 35970 }, { "grad_norm": 0.19961321353912354, "learning_rate": 7.597815328253954e-05, "loss": 0.0221, "step": 35980 }, { "grad_norm": 0.19280537962913513, "learning_rate": 7.596402408537443e-05, "loss": 0.0189, "step": 35990 }, { "grad_norm": 0.22413763403892517, "learning_rate": 7.59498920488227e-05, "loss": 0.0181, "step": 36000 }, { "grad_norm": 0.16981512308120728, "learning_rate": 7.593575717442979e-05, "loss": 0.0225, "step": 36010 }, { "grad_norm": 0.29346561431884766, "learning_rate": 7.592161946374147e-05, "loss": 0.0196, "step": 36020 }, { "grad_norm": 0.21753260493278503, "learning_rate": 7.590747891830381e-05, "loss": 0.0194, "step": 36030 }, { "grad_norm": 0.2193840891122818, "learning_rate": 7.58933355396632e-05, "loss": 0.0165, "step": 36040 }, { "grad_norm": 0.1624414473772049, "learning_rate": 7.587918932936636e-05, "loss": 0.0155, "step": 36050 }, { "grad_norm": 0.22085991501808167, "learning_rate": 7.586504028896028e-05, "loss": 0.0198, "step": 36060 }, { "grad_norm": 0.20390601456165314, "learning_rate": 7.585088841999228e-05, "loss": 0.0185, "step": 36070 }, { "grad_norm": 0.18581008911132812, "learning_rate": 7.583673372400999e-05, "loss": 0.0159, "step": 36080 }, { "grad_norm": 0.3554868698120117, "learning_rate": 7.582257620256134e-05, "loss": 0.0246, "step": 36090 }, { "grad_norm": 0.1894441694021225, "learning_rate": 7.580841585719458e-05, "loss": 0.0213, "step": 36100 }, { "grad_norm": 0.1580652892589569, "learning_rate": 7.579425268945825e-05, "loss": 0.0204, "step": 36110 }, { "grad_norm": 0.1511404812335968, "learning_rate": 7.578008670090127e-05, "loss": 0.0215, "step": 36120 }, { "grad_norm": 0.18365339934825897, "learning_rate": 7.576591789307272e-05, "loss": 0.0237, "step": 36130 }, { "grad_norm": 0.18290869891643524, "learning_rate": 7.575174626752216e-05, "loss": 0.0195, "step": 36140 }, { "grad_norm": 0.2159486562013626, "learning_rate": 7.573757182579934e-05, "loss": 0.0173, "step": 36150 }, { "grad_norm": 0.21725495159626007, "learning_rate": 7.572339456945435e-05, "loss": 0.0179, "step": 36160 }, { "grad_norm": 0.19046491384506226, "learning_rate": 7.570921450003762e-05, "loss": 0.0204, "step": 36170 }, { "grad_norm": 0.1379350870847702, "learning_rate": 7.569503161909983e-05, "loss": 0.0184, "step": 36180 }, { "grad_norm": 0.17765910923480988, "learning_rate": 7.568084592819202e-05, "loss": 0.0183, "step": 36190 }, { "grad_norm": 0.18782705068588257, "learning_rate": 7.566665742886551e-05, "loss": 0.0202, "step": 36200 }, { "grad_norm": 0.2985754609107971, "learning_rate": 7.565246612267194e-05, "loss": 0.0201, "step": 36210 }, { "grad_norm": 0.24722497165203094, "learning_rate": 7.563827201116325e-05, "loss": 0.0144, "step": 36220 }, { "grad_norm": 0.2431810200214386, "learning_rate": 7.562407509589164e-05, "loss": 0.0226, "step": 36230 }, { "grad_norm": 0.22406312823295593, "learning_rate": 7.560987537840973e-05, "loss": 0.0241, "step": 36240 }, { "grad_norm": 0.21325427293777466, "learning_rate": 7.559567286027036e-05, "loss": 0.0211, "step": 36250 }, { "grad_norm": 0.26489508152008057, "learning_rate": 7.558146754302668e-05, "loss": 0.0183, "step": 36260 }, { "grad_norm": 0.1899678111076355, "learning_rate": 7.556725942823217e-05, "loss": 0.0159, "step": 36270 }, { "grad_norm": 0.1736060082912445, "learning_rate": 7.555304851744061e-05, "loss": 0.0197, "step": 36280 }, { "grad_norm": 0.18302638828754425, "learning_rate": 7.553883481220608e-05, "loss": 0.0167, "step": 36290 }, { "grad_norm": 0.20390720665454865, "learning_rate": 7.552461831408298e-05, "loss": 0.0166, "step": 36300 }, { "grad_norm": 0.1745801419019699, "learning_rate": 7.551039902462599e-05, "loss": 0.0188, "step": 36310 }, { "grad_norm": 0.2321675568819046, "learning_rate": 7.549617694539011e-05, "loss": 0.017, "step": 36320 }, { "grad_norm": 0.2122301459312439, "learning_rate": 7.548195207793066e-05, "loss": 0.0165, "step": 36330 }, { "grad_norm": 0.17256225645542145, "learning_rate": 7.546772442380323e-05, "loss": 0.0199, "step": 36340 }, { "grad_norm": 0.1993880718946457, "learning_rate": 7.545349398456376e-05, "loss": 0.0182, "step": 36350 }, { "grad_norm": 0.1920551210641861, "learning_rate": 7.543926076176845e-05, "loss": 0.0192, "step": 36360 }, { "grad_norm": 0.20198306441307068, "learning_rate": 7.542502475697385e-05, "loss": 0.0214, "step": 36370 }, { "grad_norm": 0.1982766091823578, "learning_rate": 7.541078597173675e-05, "loss": 0.0237, "step": 36380 }, { "grad_norm": 0.29861393570899963, "learning_rate": 7.539654440761431e-05, "loss": 0.0213, "step": 36390 }, { "grad_norm": 0.1894652098417282, "learning_rate": 7.538230006616395e-05, "loss": 0.0208, "step": 36400 }, { "grad_norm": 0.1786317229270935, "learning_rate": 7.536805294894342e-05, "loss": 0.0158, "step": 36410 }, { "grad_norm": 0.280063271522522, "learning_rate": 7.535380305751076e-05, "loss": 0.0197, "step": 36420 }, { "grad_norm": 0.1741894781589508, "learning_rate": 7.533955039342431e-05, "loss": 0.0182, "step": 36430 }, { "grad_norm": 0.23558922111988068, "learning_rate": 7.532529495824274e-05, "loss": 0.0164, "step": 36440 }, { "grad_norm": 0.25985410809516907, "learning_rate": 7.531103675352497e-05, "loss": 0.0194, "step": 36450 }, { "grad_norm": 0.24686099588871002, "learning_rate": 7.529677578083029e-05, "loss": 0.0227, "step": 36460 }, { "grad_norm": 0.24327456951141357, "learning_rate": 7.528251204171823e-05, "loss": 0.0207, "step": 36470 }, { "grad_norm": 0.17070318758487701, "learning_rate": 7.526824553774867e-05, "loss": 0.0231, "step": 36480 }, { "grad_norm": 0.22453339397907257, "learning_rate": 7.525397627048178e-05, "loss": 0.023, "step": 36490 }, { "grad_norm": 0.2446925938129425, "learning_rate": 7.523970424147802e-05, "loss": 0.0173, "step": 36500 }, { "grad_norm": 0.2355601042509079, "learning_rate": 7.522542945229813e-05, "loss": 0.0219, "step": 36510 }, { "grad_norm": 0.18788740038871765, "learning_rate": 7.521115190450321e-05, "loss": 0.0175, "step": 36520 }, { "grad_norm": 0.23703765869140625, "learning_rate": 7.519687159965462e-05, "loss": 0.0224, "step": 36530 }, { "grad_norm": 0.21539005637168884, "learning_rate": 7.518258853931403e-05, "loss": 0.0176, "step": 36540 }, { "grad_norm": 0.2363535463809967, "learning_rate": 7.516830272504342e-05, "loss": 0.0184, "step": 36550 }, { "grad_norm": 0.23134057223796844, "learning_rate": 7.515401415840509e-05, "loss": 0.0214, "step": 36560 }, { "grad_norm": 0.2728738486766815, "learning_rate": 7.513972284096155e-05, "loss": 0.0192, "step": 36570 }, { "grad_norm": 0.17925818264484406, "learning_rate": 7.512542877427574e-05, "loss": 0.0156, "step": 36580 }, { "grad_norm": 0.19829203188419342, "learning_rate": 7.51111319599108e-05, "loss": 0.0162, "step": 36590 }, { "grad_norm": 0.16651217639446259, "learning_rate": 7.509683239943024e-05, "loss": 0.0145, "step": 36600 }, { "grad_norm": 0.2320668250322342, "learning_rate": 7.50825300943978e-05, "loss": 0.0245, "step": 36610 }, { "grad_norm": 0.2421301156282425, "learning_rate": 7.50682250463776e-05, "loss": 0.0175, "step": 36620 }, { "grad_norm": 0.23515699803829193, "learning_rate": 7.5053917256934e-05, "loss": 0.0192, "step": 36630 }, { "grad_norm": 0.17024414241313934, "learning_rate": 7.503960672763166e-05, "loss": 0.0198, "step": 36640 }, { "grad_norm": 0.1862565129995346, "learning_rate": 7.502529346003559e-05, "loss": 0.0226, "step": 36650 }, { "grad_norm": 0.2560393214225769, "learning_rate": 7.501097745571106e-05, "loss": 0.0254, "step": 36660 }, { "grad_norm": 0.22400148212909698, "learning_rate": 7.499665871622361e-05, "loss": 0.0195, "step": 36670 }, { "grad_norm": 0.20193342864513397, "learning_rate": 7.498233724313917e-05, "loss": 0.0196, "step": 36680 }, { "grad_norm": 0.21453502774238586, "learning_rate": 7.496801303802387e-05, "loss": 0.0209, "step": 36690 }, { "grad_norm": 0.17219528555870056, "learning_rate": 7.495368610244423e-05, "loss": 0.0141, "step": 36700 }, { "grad_norm": 0.17580972611904144, "learning_rate": 7.493935643796697e-05, "loss": 0.0149, "step": 36710 }, { "grad_norm": 0.23789164423942566, "learning_rate": 7.49250240461592e-05, "loss": 0.0175, "step": 36720 }, { "grad_norm": 0.21189986169338226, "learning_rate": 7.491068892858826e-05, "loss": 0.0158, "step": 36730 }, { "grad_norm": 0.24263182282447815, "learning_rate": 7.489635108682184e-05, "loss": 0.0174, "step": 36740 }, { "grad_norm": 0.199827179312706, "learning_rate": 7.48820105224279e-05, "loss": 0.0165, "step": 36750 }, { "grad_norm": 0.26170089840888977, "learning_rate": 7.48676672369747e-05, "loss": 0.0207, "step": 36760 }, { "grad_norm": 0.2377106100320816, "learning_rate": 7.485332123203079e-05, "loss": 0.018, "step": 36770 }, { "grad_norm": 0.21909528970718384, "learning_rate": 7.483897250916504e-05, "loss": 0.0159, "step": 36780 }, { "grad_norm": 0.2235240787267685, "learning_rate": 7.48246210699466e-05, "loss": 0.0195, "step": 36790 }, { "grad_norm": 0.18926705420017242, "learning_rate": 7.481026691594492e-05, "loss": 0.019, "step": 36800 }, { "grad_norm": 0.17607733607292175, "learning_rate": 7.479591004872975e-05, "loss": 0.0179, "step": 36810 }, { "grad_norm": 0.23270751535892487, "learning_rate": 7.478155046987114e-05, "loss": 0.0189, "step": 36820 }, { "grad_norm": 0.16901275515556335, "learning_rate": 7.476718818093944e-05, "loss": 0.0171, "step": 36830 }, { "grad_norm": 0.13933098316192627, "learning_rate": 7.475282318350527e-05, "loss": 0.0177, "step": 36840 }, { "grad_norm": 0.22070881724357605, "learning_rate": 7.473845547913957e-05, "loss": 0.0177, "step": 36850 }, { "grad_norm": 0.27094903588294983, "learning_rate": 7.472408506941357e-05, "loss": 0.0207, "step": 36860 }, { "grad_norm": 0.19007031619548798, "learning_rate": 7.470971195589881e-05, "loss": 0.0221, "step": 36870 }, { "grad_norm": 0.1928042620420456, "learning_rate": 7.46953361401671e-05, "loss": 0.0177, "step": 36880 }, { "grad_norm": 0.2500513792037964, "learning_rate": 7.468095762379055e-05, "loss": 0.0197, "step": 36890 }, { "grad_norm": 0.208116814494133, "learning_rate": 7.466657640834158e-05, "loss": 0.0203, "step": 36900 }, { "grad_norm": 0.21876145899295807, "learning_rate": 7.46521924953929e-05, "loss": 0.016, "step": 36910 }, { "grad_norm": 0.21555060148239136, "learning_rate": 7.463780588651752e-05, "loss": 0.0159, "step": 36920 }, { "grad_norm": 0.2075812965631485, "learning_rate": 7.462341658328873e-05, "loss": 0.0198, "step": 36930 }, { "grad_norm": 0.16417576372623444, "learning_rate": 7.460902458728012e-05, "loss": 0.0152, "step": 36940 }, { "grad_norm": 0.1538412719964981, "learning_rate": 7.459462990006558e-05, "loss": 0.0142, "step": 36950 }, { "grad_norm": 0.17576700448989868, "learning_rate": 7.458023252321929e-05, "loss": 0.0166, "step": 36960 }, { "grad_norm": 0.23516789078712463, "learning_rate": 7.456583245831573e-05, "loss": 0.0205, "step": 36970 }, { "grad_norm": 0.216634139418602, "learning_rate": 7.455142970692968e-05, "loss": 0.0193, "step": 36980 }, { "grad_norm": 0.1714753806591034, "learning_rate": 7.453702427063615e-05, "loss": 0.0175, "step": 36990 }, { "grad_norm": 0.16177479922771454, "learning_rate": 7.452261615101057e-05, "loss": 0.0188, "step": 37000 }, { "grad_norm": 0.21911916136741638, "learning_rate": 7.450820534962856e-05, "loss": 0.0199, "step": 37010 }, { "grad_norm": 0.17097711563110352, "learning_rate": 7.449379186806604e-05, "loss": 0.0186, "step": 37020 }, { "grad_norm": 0.2663766145706177, "learning_rate": 7.44793757078993e-05, "loss": 0.0185, "step": 37030 }, { "grad_norm": 0.20995616912841797, "learning_rate": 7.44649568707048e-05, "loss": 0.0184, "step": 37040 }, { "grad_norm": 0.21304990351200104, "learning_rate": 7.445053535805942e-05, "loss": 0.0172, "step": 37050 }, { "grad_norm": 0.21423505246639252, "learning_rate": 7.443611117154024e-05, "loss": 0.0189, "step": 37060 }, { "grad_norm": 0.1761830896139145, "learning_rate": 7.442168431272469e-05, "loss": 0.017, "step": 37070 }, { "grad_norm": 0.16991981863975525, "learning_rate": 7.440725478319044e-05, "loss": 0.0219, "step": 37080 }, { "grad_norm": 0.20682860910892487, "learning_rate": 7.439282258451553e-05, "loss": 0.0196, "step": 37090 }, { "grad_norm": 0.2142188847064972, "learning_rate": 7.43783877182782e-05, "loss": 0.0213, "step": 37100 }, { "grad_norm": 0.18390174210071564, "learning_rate": 7.436395018605707e-05, "loss": 0.0172, "step": 37110 }, { "grad_norm": 0.1990918219089508, "learning_rate": 7.434950998943094e-05, "loss": 0.0191, "step": 37120 }, { "grad_norm": 0.22981944680213928, "learning_rate": 7.433506712997902e-05, "loss": 0.017, "step": 37130 }, { "grad_norm": 0.20084498822689056, "learning_rate": 7.432062160928077e-05, "loss": 0.018, "step": 37140 }, { "grad_norm": 0.20459237694740295, "learning_rate": 7.430617342891588e-05, "loss": 0.0173, "step": 37150 }, { "grad_norm": 0.24406161904335022, "learning_rate": 7.429172259046442e-05, "loss": 0.0166, "step": 37160 }, { "grad_norm": 0.21450795233249664, "learning_rate": 7.42772690955067e-05, "loss": 0.0155, "step": 37170 }, { "grad_norm": 0.2257349193096161, "learning_rate": 7.426281294562332e-05, "loss": 0.0205, "step": 37180 }, { "grad_norm": 0.17847765982151031, "learning_rate": 7.424835414239519e-05, "loss": 0.0147, "step": 37190 }, { "grad_norm": 0.17440322041511536, "learning_rate": 7.423389268740352e-05, "loss": 0.0169, "step": 37200 }, { "grad_norm": 0.1975603848695755, "learning_rate": 7.42194285822298e-05, "loss": 0.0193, "step": 37210 }, { "grad_norm": 0.2629759907722473, "learning_rate": 7.420496182845576e-05, "loss": 0.0194, "step": 37220 }, { "grad_norm": 0.2407800406217575, "learning_rate": 7.419049242766348e-05, "loss": 0.0205, "step": 37230 }, { "grad_norm": 0.2706225514411926, "learning_rate": 7.417602038143533e-05, "loss": 0.0194, "step": 37240 }, { "grad_norm": 0.2200305312871933, "learning_rate": 7.416154569135393e-05, "loss": 0.0198, "step": 37250 }, { "grad_norm": 0.20527413487434387, "learning_rate": 7.414706835900222e-05, "loss": 0.0192, "step": 37260 }, { "grad_norm": 0.15057624876499176, "learning_rate": 7.413258838596345e-05, "loss": 0.0162, "step": 37270 }, { "grad_norm": 0.15111026167869568, "learning_rate": 7.411810577382106e-05, "loss": 0.0169, "step": 37280 }, { "grad_norm": 0.1698867380619049, "learning_rate": 7.410362052415889e-05, "loss": 0.0168, "step": 37290 }, { "grad_norm": 0.18437708914279938, "learning_rate": 7.408913263856102e-05, "loss": 0.0135, "step": 37300 }, { "grad_norm": 0.20551209151744843, "learning_rate": 7.407464211861183e-05, "loss": 0.0203, "step": 37310 }, { "grad_norm": 0.1675131916999817, "learning_rate": 7.406014896589597e-05, "loss": 0.0154, "step": 37320 }, { "grad_norm": 0.17610642313957214, "learning_rate": 7.404565318199838e-05, "loss": 0.0154, "step": 37330 }, { "grad_norm": 0.19637376070022583, "learning_rate": 7.403115476850432e-05, "loss": 0.0179, "step": 37340 }, { "grad_norm": 0.2480379343032837, "learning_rate": 7.401665372699932e-05, "loss": 0.0238, "step": 37350 }, { "grad_norm": 0.280793696641922, "learning_rate": 7.400215005906916e-05, "loss": 0.0253, "step": 37360 }, { "grad_norm": 0.181918665766716, "learning_rate": 7.398764376629996e-05, "loss": 0.0227, "step": 37370 }, { "grad_norm": 0.29929834604263306, "learning_rate": 7.39731348502781e-05, "loss": 0.0204, "step": 37380 }, { "grad_norm": 0.21741560101509094, "learning_rate": 7.395862331259028e-05, "loss": 0.0222, "step": 37390 }, { "grad_norm": 0.1699610948562622, "learning_rate": 7.394410915482341e-05, "loss": 0.0176, "step": 37400 }, { "grad_norm": 0.20705027878284454, "learning_rate": 7.392959237856476e-05, "loss": 0.0194, "step": 37410 }, { "grad_norm": 0.178043395280838, "learning_rate": 7.391507298540187e-05, "loss": 0.0171, "step": 37420 }, { "grad_norm": 0.16989901661872864, "learning_rate": 7.390055097692257e-05, "loss": 0.0197, "step": 37430 }, { "grad_norm": 0.18223601579666138, "learning_rate": 7.388602635471494e-05, "loss": 0.0155, "step": 37440 }, { "grad_norm": 0.2401726096868515, "learning_rate": 7.387149912036738e-05, "loss": 0.0203, "step": 37450 }, { "grad_norm": 0.23352311551570892, "learning_rate": 7.385696927546854e-05, "loss": 0.0242, "step": 37460 }, { "grad_norm": 0.2592763602733612, "learning_rate": 7.384243682160743e-05, "loss": 0.0225, "step": 37470 }, { "grad_norm": 0.14510445296764374, "learning_rate": 7.382790176037329e-05, "loss": 0.0165, "step": 37480 }, { "grad_norm": 0.20221038162708282, "learning_rate": 7.381336409335562e-05, "loss": 0.0213, "step": 37490 }, { "grad_norm": 0.1471138596534729, "learning_rate": 7.379882382214426e-05, "loss": 0.0168, "step": 37500 }, { "grad_norm": 0.18015703558921814, "learning_rate": 7.378428094832931e-05, "loss": 0.0156, "step": 37510 }, { "grad_norm": 0.2352219820022583, "learning_rate": 7.376973547350114e-05, "loss": 0.0161, "step": 37520 }, { "grad_norm": 0.2136382758617401, "learning_rate": 7.375518739925046e-05, "loss": 0.0169, "step": 37530 }, { "grad_norm": 0.24598929286003113, "learning_rate": 7.374063672716818e-05, "loss": 0.0167, "step": 37540 }, { "grad_norm": 0.2298891395330429, "learning_rate": 7.372608345884558e-05, "loss": 0.0188, "step": 37550 }, { "grad_norm": 0.16756680607795715, "learning_rate": 7.371152759587415e-05, "loss": 0.0159, "step": 37560 }, { "grad_norm": 0.3415600657463074, "learning_rate": 7.369696913984572e-05, "loss": 0.0177, "step": 37570 }, { "grad_norm": 0.21260225772857666, "learning_rate": 7.368240809235237e-05, "loss": 0.0194, "step": 37580 }, { "grad_norm": 0.2173536717891693, "learning_rate": 7.366784445498649e-05, "loss": 0.0168, "step": 37590 }, { "grad_norm": 0.2504984736442566, "learning_rate": 7.36532782293407e-05, "loss": 0.0215, "step": 37600 }, { "grad_norm": 0.19637586176395416, "learning_rate": 7.363870941700797e-05, "loss": 0.0168, "step": 37610 }, { "grad_norm": 0.18635965883731842, "learning_rate": 7.362413801958152e-05, "loss": 0.0167, "step": 37620 }, { "grad_norm": 0.16800068318843842, "learning_rate": 7.360956403865486e-05, "loss": 0.02, "step": 37630 }, { "grad_norm": 0.20050881803035736, "learning_rate": 7.359498747582177e-05, "loss": 0.0231, "step": 37640 }, { "grad_norm": 0.16800542175769806, "learning_rate": 7.358040833267634e-05, "loss": 0.0194, "step": 37650 }, { "grad_norm": 0.1850154846906662, "learning_rate": 7.356582661081288e-05, "loss": 0.0191, "step": 37660 }, { "grad_norm": 0.1654950976371765, "learning_rate": 7.355124231182607e-05, "loss": 0.0162, "step": 37670 }, { "grad_norm": 0.16670145094394684, "learning_rate": 7.353665543731083e-05, "loss": 0.0171, "step": 37680 }, { "grad_norm": 0.16375979781150818, "learning_rate": 7.352206598886229e-05, "loss": 0.0159, "step": 37690 }, { "grad_norm": 0.23775692284107208, "learning_rate": 7.350747396807601e-05, "loss": 0.0163, "step": 37700 }, { "grad_norm": 0.18242265284061432, "learning_rate": 7.349287937654772e-05, "loss": 0.0155, "step": 37710 }, { "grad_norm": 0.17696446180343628, "learning_rate": 7.347828221587345e-05, "loss": 0.0153, "step": 37720 }, { "grad_norm": 0.21786534786224365, "learning_rate": 7.346368248764955e-05, "loss": 0.0159, "step": 37730 }, { "grad_norm": 0.2611207365989685, "learning_rate": 7.34490801934726e-05, "loss": 0.0156, "step": 37740 }, { "grad_norm": 0.2031465321779251, "learning_rate": 7.343447533493947e-05, "loss": 0.0181, "step": 37750 }, { "grad_norm": 0.17836543917655945, "learning_rate": 7.341986791364739e-05, "loss": 0.0171, "step": 37760 }, { "grad_norm": 0.2309107631444931, "learning_rate": 7.340525793119374e-05, "loss": 0.0241, "step": 37770 }, { "grad_norm": 0.20516718924045563, "learning_rate": 7.33906453891763e-05, "loss": 0.0169, "step": 37780 }, { "grad_norm": 0.16831600666046143, "learning_rate": 7.337603028919303e-05, "loss": 0.0165, "step": 37790 }, { "grad_norm": 0.19044123589992523, "learning_rate": 7.336141263284225e-05, "loss": 0.0186, "step": 37800 }, { "grad_norm": 0.24488668143749237, "learning_rate": 7.334679242172249e-05, "loss": 0.017, "step": 37810 }, { "grad_norm": 0.25839459896087646, "learning_rate": 7.33321696574326e-05, "loss": 0.0191, "step": 37820 }, { "grad_norm": 0.21801002323627472, "learning_rate": 7.331754434157174e-05, "loss": 0.0173, "step": 37830 }, { "grad_norm": 0.2091277688741684, "learning_rate": 7.330291647573928e-05, "loss": 0.0205, "step": 37840 }, { "grad_norm": 0.22321079671382904, "learning_rate": 7.32882860615349e-05, "loss": 0.0169, "step": 37850 }, { "grad_norm": 0.21517333388328552, "learning_rate": 7.327365310055858e-05, "loss": 0.019, "step": 37860 }, { "grad_norm": 0.18450772762298584, "learning_rate": 7.325901759441055e-05, "loss": 0.0158, "step": 37870 }, { "grad_norm": 0.18366988003253937, "learning_rate": 7.324437954469132e-05, "loss": 0.0191, "step": 37880 }, { "grad_norm": 0.17292046546936035, "learning_rate": 7.32297389530017e-05, "loss": 0.0188, "step": 37890 }, { "grad_norm": 0.18662291765213013, "learning_rate": 7.321509582094274e-05, "loss": 0.0171, "step": 37900 }, { "grad_norm": 0.18594147264957428, "learning_rate": 7.320045015011582e-05, "loss": 0.0163, "step": 37910 }, { "grad_norm": 0.22166046500205994, "learning_rate": 7.318580194212253e-05, "loss": 0.0204, "step": 37920 }, { "grad_norm": 0.29079470038414, "learning_rate": 7.317115119856483e-05, "loss": 0.0178, "step": 37930 }, { "grad_norm": 0.17719075083732605, "learning_rate": 7.315649792104485e-05, "loss": 0.0193, "step": 37940 }, { "grad_norm": 0.14109154045581818, "learning_rate": 7.314184211116507e-05, "loss": 0.021, "step": 37950 }, { "grad_norm": 0.1603740006685257, "learning_rate": 7.312718377052823e-05, "loss": 0.0178, "step": 37960 }, { "grad_norm": 0.1738981455564499, "learning_rate": 7.311252290073737e-05, "loss": 0.0221, "step": 37970 }, { "grad_norm": 0.2553381025791168, "learning_rate": 7.309785950339572e-05, "loss": 0.0196, "step": 37980 }, { "grad_norm": 0.2923474907875061, "learning_rate": 7.308319358010692e-05, "loss": 0.0263, "step": 37990 }, { "grad_norm": 0.28924912214279175, "learning_rate": 7.306852513247474e-05, "loss": 0.0222, "step": 38000 }, { "grad_norm": 0.22695967555046082, "learning_rate": 7.305385416210336e-05, "loss": 0.0195, "step": 38010 }, { "grad_norm": 0.1567113846540451, "learning_rate": 7.303918067059711e-05, "loss": 0.0214, "step": 38020 }, { "grad_norm": 0.1741839349269867, "learning_rate": 7.302450465956075e-05, "loss": 0.0219, "step": 38030 }, { "grad_norm": 0.16123290359973907, "learning_rate": 7.300982613059914e-05, "loss": 0.0178, "step": 38040 }, { "grad_norm": 0.22101320326328278, "learning_rate": 7.299514508531757e-05, "loss": 0.0202, "step": 38050 }, { "grad_norm": 0.18454398214817047, "learning_rate": 7.298046152532148e-05, "loss": 0.0185, "step": 38060 }, { "grad_norm": 0.1631760150194168, "learning_rate": 7.296577545221668e-05, "loss": 0.017, "step": 38070 }, { "grad_norm": 0.14403878152370453, "learning_rate": 7.29510868676092e-05, "loss": 0.0139, "step": 38080 }, { "grad_norm": 0.1565152108669281, "learning_rate": 7.293639577310538e-05, "loss": 0.015, "step": 38090 }, { "grad_norm": 0.19650483131408691, "learning_rate": 7.292170217031179e-05, "loss": 0.0173, "step": 38100 }, { "grad_norm": 0.15791942179203033, "learning_rate": 7.290700606083532e-05, "loss": 0.0172, "step": 38110 }, { "grad_norm": 0.1555555760860443, "learning_rate": 7.28923074462831e-05, "loss": 0.0157, "step": 38120 }, { "grad_norm": 0.1888117790222168, "learning_rate": 7.287760632826255e-05, "loss": 0.0187, "step": 38130 }, { "grad_norm": 0.20941713452339172, "learning_rate": 7.286290270838139e-05, "loss": 0.0199, "step": 38140 }, { "grad_norm": 0.21551918983459473, "learning_rate": 7.284819658824756e-05, "loss": 0.0214, "step": 38150 }, { "grad_norm": 0.21591536700725555, "learning_rate": 7.283348796946929e-05, "loss": 0.02, "step": 38160 }, { "grad_norm": 0.2868172228336334, "learning_rate": 7.28187768536551e-05, "loss": 0.0205, "step": 38170 }, { "grad_norm": 0.15845178067684174, "learning_rate": 7.280406324241379e-05, "loss": 0.0147, "step": 38180 }, { "grad_norm": 0.2064000964164734, "learning_rate": 7.278934713735441e-05, "loss": 0.0214, "step": 38190 }, { "grad_norm": 0.21567219495773315, "learning_rate": 7.277462854008629e-05, "loss": 0.0145, "step": 38200 }, { "grad_norm": 0.19733017683029175, "learning_rate": 7.275990745221904e-05, "loss": 0.0135, "step": 38210 }, { "grad_norm": 0.21473729610443115, "learning_rate": 7.274518387536252e-05, "loss": 0.0181, "step": 38220 }, { "grad_norm": 0.2438688427209854, "learning_rate": 7.273045781112689e-05, "loss": 0.0193, "step": 38230 }, { "grad_norm": 0.15853066742420197, "learning_rate": 7.271572926112257e-05, "loss": 0.0184, "step": 38240 }, { "grad_norm": 0.2169140726327896, "learning_rate": 7.270099822696024e-05, "loss": 0.0163, "step": 38250 }, { "grad_norm": 0.22069178521633148, "learning_rate": 7.26862647102509e-05, "loss": 0.0152, "step": 38260 }, { "grad_norm": 0.17367979884147644, "learning_rate": 7.267152871260573e-05, "loss": 0.0155, "step": 38270 }, { "grad_norm": 0.1319480538368225, "learning_rate": 7.265679023563629e-05, "loss": 0.017, "step": 38280 }, { "grad_norm": 0.1940174698829651, "learning_rate": 7.264204928095433e-05, "loss": 0.019, "step": 38290 }, { "grad_norm": 0.22536833584308624, "learning_rate": 7.262730585017188e-05, "loss": 0.0168, "step": 38300 }, { "grad_norm": 0.16345319151878357, "learning_rate": 7.261255994490131e-05, "loss": 0.0149, "step": 38310 }, { "grad_norm": 0.24100251495838165, "learning_rate": 7.259781156675517e-05, "loss": 0.0164, "step": 38320 }, { "grad_norm": 0.18954826891422272, "learning_rate": 7.258306071734633e-05, "loss": 0.0152, "step": 38330 }, { "grad_norm": 0.17903850972652435, "learning_rate": 7.256830739828792e-05, "loss": 0.0154, "step": 38340 }, { "grad_norm": 0.19192282855510712, "learning_rate": 7.255355161119336e-05, "loss": 0.0218, "step": 38350 }, { "grad_norm": 0.23685860633850098, "learning_rate": 7.253879335767628e-05, "loss": 0.0172, "step": 38360 }, { "grad_norm": 0.23912860453128815, "learning_rate": 7.252403263935067e-05, "loss": 0.0197, "step": 38370 }, { "grad_norm": 0.19677278399467468, "learning_rate": 7.25092694578307e-05, "loss": 0.0166, "step": 38380 }, { "grad_norm": 0.1895841360092163, "learning_rate": 7.249450381473087e-05, "loss": 0.0175, "step": 38390 }, { "grad_norm": 0.25078126788139343, "learning_rate": 7.247973571166593e-05, "loss": 0.0175, "step": 38400 }, { "grad_norm": 0.3126688003540039, "learning_rate": 7.246496515025089e-05, "loss": 0.0187, "step": 38410 }, { "grad_norm": 0.27420198917388916, "learning_rate": 7.245019213210101e-05, "loss": 0.0184, "step": 38420 }, { "grad_norm": 0.2113153487443924, "learning_rate": 7.24354166588319e-05, "loss": 0.0171, "step": 38430 }, { "grad_norm": 0.1968507170677185, "learning_rate": 7.242063873205935e-05, "loss": 0.0137, "step": 38440 }, { "grad_norm": 0.1842799186706543, "learning_rate": 7.240585835339946e-05, "loss": 0.0175, "step": 38450 }, { "grad_norm": 0.15028734505176544, "learning_rate": 7.239107552446857e-05, "loss": 0.0149, "step": 38460 }, { "grad_norm": 0.20430107414722443, "learning_rate": 7.237629024688334e-05, "loss": 0.0156, "step": 38470 }, { "grad_norm": 0.15990319848060608, "learning_rate": 7.236150252226064e-05, "loss": 0.0174, "step": 38480 }, { "grad_norm": 0.19473706185817719, "learning_rate": 7.234671235221765e-05, "loss": 0.0204, "step": 38490 }, { "grad_norm": 0.18985088169574738, "learning_rate": 7.233191973837179e-05, "loss": 0.0223, "step": 38500 }, { "grad_norm": 0.18633811175823212, "learning_rate": 7.231712468234078e-05, "loss": 0.0181, "step": 38510 }, { "grad_norm": 0.18857863545417786, "learning_rate": 7.230232718574254e-05, "loss": 0.0164, "step": 38520 }, { "grad_norm": 0.2683728039264679, "learning_rate": 7.228752725019535e-05, "loss": 0.0175, "step": 38530 }, { "grad_norm": 0.23857830464839935, "learning_rate": 7.227272487731769e-05, "loss": 0.0157, "step": 38540 }, { "grad_norm": 0.15587922930717468, "learning_rate": 7.225792006872831e-05, "loss": 0.0188, "step": 38550 }, { "grad_norm": 0.264207124710083, "learning_rate": 7.224311282604628e-05, "loss": 0.0181, "step": 38560 }, { "grad_norm": 0.13823527097702026, "learning_rate": 7.222830315089085e-05, "loss": 0.0175, "step": 38570 }, { "grad_norm": 0.16305671632289886, "learning_rate": 7.22134910448816e-05, "loss": 0.0162, "step": 38580 }, { "grad_norm": 0.19033311307430267, "learning_rate": 7.219867650963839e-05, "loss": 0.0141, "step": 38590 }, { "grad_norm": 0.2016158252954483, "learning_rate": 7.21838595467813e-05, "loss": 0.0149, "step": 38600 }, { "grad_norm": 0.20945954322814941, "learning_rate": 7.216904015793067e-05, "loss": 0.0145, "step": 38610 }, { "grad_norm": 0.2171376645565033, "learning_rate": 7.215421834470713e-05, "loss": 0.0162, "step": 38620 }, { "grad_norm": 0.18070627748966217, "learning_rate": 7.213939410873158e-05, "loss": 0.0198, "step": 38630 }, { "grad_norm": 0.15456433594226837, "learning_rate": 7.212456745162519e-05, "loss": 0.017, "step": 38640 }, { "grad_norm": 0.20501255989074707, "learning_rate": 7.210973837500937e-05, "loss": 0.0213, "step": 38650 }, { "grad_norm": 0.17948956787586212, "learning_rate": 7.209490688050578e-05, "loss": 0.0173, "step": 38660 }, { "grad_norm": 0.16679787635803223, "learning_rate": 7.208007296973641e-05, "loss": 0.0185, "step": 38670 }, { "grad_norm": 0.15667660534381866, "learning_rate": 7.206523664432345e-05, "loss": 0.0155, "step": 38680 }, { "grad_norm": 0.20815877616405487, "learning_rate": 7.205039790588939e-05, "loss": 0.0162, "step": 38690 }, { "grad_norm": 0.1619185209274292, "learning_rate": 7.203555675605697e-05, "loss": 0.0139, "step": 38700 }, { "grad_norm": 0.14372077584266663, "learning_rate": 7.202071319644917e-05, "loss": 0.0151, "step": 38710 }, { "grad_norm": 0.15605148673057556, "learning_rate": 7.200586722868932e-05, "loss": 0.0198, "step": 38720 }, { "grad_norm": 0.2702412009239197, "learning_rate": 7.199101885440088e-05, "loss": 0.0176, "step": 38730 }, { "grad_norm": 0.1635785698890686, "learning_rate": 7.197616807520771e-05, "loss": 0.0201, "step": 38740 }, { "grad_norm": 0.1953079104423523, "learning_rate": 7.196131489273381e-05, "loss": 0.019, "step": 38750 }, { "grad_norm": 0.22527626156806946, "learning_rate": 7.194645930860355e-05, "loss": 0.0184, "step": 38760 }, { "grad_norm": 0.18937499821186066, "learning_rate": 7.19316013244415e-05, "loss": 0.0145, "step": 38770 }, { "grad_norm": 0.21563275158405304, "learning_rate": 7.191674094187248e-05, "loss": 0.0218, "step": 38780 }, { "grad_norm": 0.2395145297050476, "learning_rate": 7.190187816252165e-05, "loss": 0.0195, "step": 38790 }, { "grad_norm": 0.20795099437236786, "learning_rate": 7.188701298801435e-05, "loss": 0.0169, "step": 38800 }, { "grad_norm": 0.18552479147911072, "learning_rate": 7.18721454199762e-05, "loss": 0.0216, "step": 38810 }, { "grad_norm": 0.2143174111843109, "learning_rate": 7.185727546003312e-05, "loss": 0.0158, "step": 38820 }, { "grad_norm": 0.22131547331809998, "learning_rate": 7.184240310981126e-05, "loss": 0.0169, "step": 38830 }, { "grad_norm": 0.17015670239925385, "learning_rate": 7.182752837093704e-05, "loss": 0.0141, "step": 38840 }, { "grad_norm": 0.2613038718700409, "learning_rate": 7.181265124503711e-05, "loss": 0.0187, "step": 38850 }, { "grad_norm": 0.21200381219387054, "learning_rate": 7.179777173373847e-05, "loss": 0.0156, "step": 38860 }, { "grad_norm": 0.1579909771680832, "learning_rate": 7.178288983866826e-05, "loss": 0.0161, "step": 38870 }, { "grad_norm": 0.1664726883172989, "learning_rate": 7.176800556145397e-05, "loss": 0.0144, "step": 38880 }, { "grad_norm": 0.18460918962955475, "learning_rate": 7.175311890372334e-05, "loss": 0.0145, "step": 38890 }, { "grad_norm": 0.2386826127767563, "learning_rate": 7.17382298671043e-05, "loss": 0.0168, "step": 38900 }, { "grad_norm": 0.34191015362739563, "learning_rate": 7.172333845322515e-05, "loss": 0.0159, "step": 38910 }, { "grad_norm": 0.226569265127182, "learning_rate": 7.170844466371436e-05, "loss": 0.0169, "step": 38920 }, { "grad_norm": 0.26451000571250916, "learning_rate": 7.16935485002007e-05, "loss": 0.0237, "step": 38930 }, { "grad_norm": 0.20316681265830994, "learning_rate": 7.167864996431319e-05, "loss": 0.0184, "step": 38940 }, { "grad_norm": 0.19602511823177338, "learning_rate": 7.166374905768111e-05, "loss": 0.0197, "step": 38950 }, { "grad_norm": 0.21566340327262878, "learning_rate": 7.164884578193401e-05, "loss": 0.0217, "step": 38960 }, { "grad_norm": 0.17488716542720795, "learning_rate": 7.163394013870168e-05, "loss": 0.0174, "step": 38970 }, { "grad_norm": 0.22826753556728363, "learning_rate": 7.161903212961419e-05, "loss": 0.0199, "step": 38980 }, { "grad_norm": 0.17461957037448883, "learning_rate": 7.160412175630183e-05, "loss": 0.0156, "step": 38990 }, { "grad_norm": 0.26125141978263855, "learning_rate": 7.158920902039521e-05, "loss": 0.0148, "step": 39000 }, { "grad_norm": 0.14734694361686707, "learning_rate": 7.157429392352514e-05, "loss": 0.0198, "step": 39010 }, { "grad_norm": 0.2012835144996643, "learning_rate": 7.155937646732275e-05, "loss": 0.021, "step": 39020 }, { "grad_norm": 0.16994532942771912, "learning_rate": 7.154445665341933e-05, "loss": 0.0176, "step": 39030 }, { "grad_norm": 0.1883516013622284, "learning_rate": 7.152953448344654e-05, "loss": 0.0155, "step": 39040 }, { "grad_norm": 0.20697206258773804, "learning_rate": 7.151460995903624e-05, "loss": 0.0205, "step": 39050 }, { "grad_norm": 0.19681155681610107, "learning_rate": 7.149968308182052e-05, "loss": 0.0178, "step": 39060 }, { "grad_norm": 0.18986637890338898, "learning_rate": 7.14847538534318e-05, "loss": 0.015, "step": 39070 }, { "grad_norm": 0.23862947523593903, "learning_rate": 7.14698222755027e-05, "loss": 0.0199, "step": 39080 }, { "grad_norm": 0.21141347289085388, "learning_rate": 7.14548883496661e-05, "loss": 0.0185, "step": 39090 }, { "grad_norm": 0.1685146987438202, "learning_rate": 7.143995207755517e-05, "loss": 0.0211, "step": 39100 }, { "grad_norm": 0.1734207570552826, "learning_rate": 7.142501346080333e-05, "loss": 0.0178, "step": 39110 }, { "grad_norm": 0.23086649179458618, "learning_rate": 7.141007250104421e-05, "loss": 0.0163, "step": 39120 }, { "grad_norm": 0.18614332377910614, "learning_rate": 7.139512919991176e-05, "loss": 0.019, "step": 39130 }, { "grad_norm": 0.230429008603096, "learning_rate": 7.138018355904014e-05, "loss": 0.0189, "step": 39140 }, { "grad_norm": 0.21146591007709503, "learning_rate": 7.13652355800638e-05, "loss": 0.0164, "step": 39150 }, { "grad_norm": 0.21492397785186768, "learning_rate": 7.135028526461743e-05, "loss": 0.0147, "step": 39160 }, { "grad_norm": 0.19660398364067078, "learning_rate": 7.133533261433594e-05, "loss": 0.017, "step": 39170 }, { "grad_norm": 0.2225688099861145, "learning_rate": 7.132037763085457e-05, "loss": 0.0187, "step": 39180 }, { "grad_norm": 0.20389412343502045, "learning_rate": 7.130542031580875e-05, "loss": 0.0179, "step": 39190 }, { "grad_norm": 0.20322979986667633, "learning_rate": 7.12904606708342e-05, "loss": 0.0184, "step": 39200 }, { "grad_norm": 0.1996956765651703, "learning_rate": 7.127549869756687e-05, "loss": 0.017, "step": 39210 }, { "grad_norm": 0.19051553308963776, "learning_rate": 7.126053439764299e-05, "loss": 0.0151, "step": 39220 }, { "grad_norm": 0.19844770431518555, "learning_rate": 7.124556777269904e-05, "loss": 0.0191, "step": 39230 }, { "grad_norm": 0.23479540646076202, "learning_rate": 7.123059882437174e-05, "loss": 0.0224, "step": 39240 }, { "grad_norm": 0.1970800757408142, "learning_rate": 7.121562755429807e-05, "loss": 0.0163, "step": 39250 }, { "grad_norm": 0.16243614256381989, "learning_rate": 7.120065396411528e-05, "loss": 0.0147, "step": 39260 }, { "grad_norm": 0.24105140566825867, "learning_rate": 7.118567805546084e-05, "loss": 0.0198, "step": 39270 }, { "grad_norm": 0.19706133008003235, "learning_rate": 7.117069982997248e-05, "loss": 0.0196, "step": 39280 }, { "grad_norm": 0.20855721831321716, "learning_rate": 7.115571928928825e-05, "loss": 0.0171, "step": 39290 }, { "grad_norm": 0.17082272469997406, "learning_rate": 7.114073643504635e-05, "loss": 0.0159, "step": 39300 }, { "grad_norm": 0.15264073014259338, "learning_rate": 7.11257512688853e-05, "loss": 0.017, "step": 39310 }, { "grad_norm": 0.2310570776462555, "learning_rate": 7.111076379244384e-05, "loss": 0.0176, "step": 39320 }, { "grad_norm": 0.2074611932039261, "learning_rate": 7.109577400736101e-05, "loss": 0.0136, "step": 39330 }, { "grad_norm": 0.20179426670074463, "learning_rate": 7.108078191527605e-05, "loss": 0.0212, "step": 39340 }, { "grad_norm": 0.16106724739074707, "learning_rate": 7.106578751782847e-05, "loss": 0.0212, "step": 39350 }, { "grad_norm": 0.1576327234506607, "learning_rate": 7.105079081665803e-05, "loss": 0.0176, "step": 39360 }, { "grad_norm": 0.1670384407043457, "learning_rate": 7.103579181340476e-05, "loss": 0.0168, "step": 39370 }, { "grad_norm": 0.17830199003219604, "learning_rate": 7.102079050970893e-05, "loss": 0.0155, "step": 39380 }, { "grad_norm": 0.2826143205165863, "learning_rate": 7.100578690721105e-05, "loss": 0.0137, "step": 39390 }, { "grad_norm": 0.24851375818252563, "learning_rate": 7.09907810075519e-05, "loss": 0.0167, "step": 39400 }, { "grad_norm": 0.21123449504375458, "learning_rate": 7.097577281237249e-05, "loss": 0.0199, "step": 39410 }, { "grad_norm": 0.17152279615402222, "learning_rate": 7.09607623233141e-05, "loss": 0.0157, "step": 39420 }, { "grad_norm": 0.19391265511512756, "learning_rate": 7.094574954201827e-05, "loss": 0.0146, "step": 39430 }, { "grad_norm": 0.21458221971988678, "learning_rate": 7.093073447012675e-05, "loss": 0.0171, "step": 39440 }, { "grad_norm": 0.16463865339756012, "learning_rate": 7.09157171092816e-05, "loss": 0.0208, "step": 39450 }, { "grad_norm": 0.2573408782482147, "learning_rate": 7.090069746112504e-05, "loss": 0.0189, "step": 39460 }, { "grad_norm": 0.20307734608650208, "learning_rate": 7.088567552729965e-05, "loss": 0.0195, "step": 39470 }, { "grad_norm": 0.18261681497097015, "learning_rate": 7.087065130944818e-05, "loss": 0.0172, "step": 39480 }, { "grad_norm": 0.14381273090839386, "learning_rate": 7.085562480921366e-05, "loss": 0.0167, "step": 39490 }, { "grad_norm": 0.1368923783302307, "learning_rate": 7.084059602823937e-05, "loss": 0.0163, "step": 39500 }, { "grad_norm": 0.2004372775554657, "learning_rate": 7.082556496816882e-05, "loss": 0.0143, "step": 39510 }, { "grad_norm": 0.3064517378807068, "learning_rate": 7.081053163064582e-05, "loss": 0.0203, "step": 39520 }, { "grad_norm": 0.1631784737110138, "learning_rate": 7.079549601731434e-05, "loss": 0.0257, "step": 39530 }, { "grad_norm": 0.21375742554664612, "learning_rate": 7.07804581298187e-05, "loss": 0.0237, "step": 39540 }, { "grad_norm": 0.23478178679943085, "learning_rate": 7.07654179698034e-05, "loss": 0.0198, "step": 39550 }, { "grad_norm": 0.20184896886348724, "learning_rate": 7.075037553891321e-05, "loss": 0.0195, "step": 39560 }, { "grad_norm": 0.18685701489448547, "learning_rate": 7.073533083879315e-05, "loss": 0.0169, "step": 39570 }, { "grad_norm": 0.24654734134674072, "learning_rate": 7.072028387108849e-05, "loss": 0.0183, "step": 39580 }, { "grad_norm": 0.23837576806545258, "learning_rate": 7.070523463744472e-05, "loss": 0.0146, "step": 39590 }, { "grad_norm": 0.18408255279064178, "learning_rate": 7.069018313950763e-05, "loss": 0.0149, "step": 39600 }, { "grad_norm": 0.19958007335662842, "learning_rate": 7.067512937892322e-05, "loss": 0.0177, "step": 39610 }, { "grad_norm": 0.15838012099266052, "learning_rate": 7.066007335733775e-05, "loss": 0.0167, "step": 39620 }, { "grad_norm": 0.17722134292125702, "learning_rate": 7.064501507639772e-05, "loss": 0.0182, "step": 39630 }, { "grad_norm": 0.2384863793849945, "learning_rate": 7.062995453774987e-05, "loss": 0.0178, "step": 39640 }, { "grad_norm": 0.3002943694591522, "learning_rate": 7.061489174304121e-05, "loss": 0.0188, "step": 39650 }, { "grad_norm": 0.2207108736038208, "learning_rate": 7.0599826693919e-05, "loss": 0.0197, "step": 39660 }, { "grad_norm": 0.17951209843158722, "learning_rate": 7.05847593920307e-05, "loss": 0.0158, "step": 39670 }, { "grad_norm": 0.19843845069408417, "learning_rate": 7.056968983902406e-05, "loss": 0.0196, "step": 39680 }, { "grad_norm": 0.17439274489879608, "learning_rate": 7.055461803654706e-05, "loss": 0.015, "step": 39690 }, { "grad_norm": 0.16265930235385895, "learning_rate": 7.053954398624794e-05, "loss": 0.0183, "step": 39700 }, { "grad_norm": 0.23856128752231598, "learning_rate": 7.052446768977518e-05, "loss": 0.0192, "step": 39710 }, { "grad_norm": 0.2557302713394165, "learning_rate": 7.050938914877748e-05, "loss": 0.0187, "step": 39720 }, { "grad_norm": 0.19875378906726837, "learning_rate": 7.049430836490382e-05, "loss": 0.0175, "step": 39730 }, { "grad_norm": 0.1561630666255951, "learning_rate": 7.047922533980342e-05, "loss": 0.0197, "step": 39740 }, { "grad_norm": 0.19008325040340424, "learning_rate": 7.046414007512571e-05, "loss": 0.019, "step": 39750 }, { "grad_norm": 0.24376855790615082, "learning_rate": 7.044905257252042e-05, "loss": 0.0172, "step": 39760 }, { "grad_norm": 0.15357840061187744, "learning_rate": 7.04339628336375e-05, "loss": 0.0149, "step": 39770 }, { "grad_norm": 0.1800292432308197, "learning_rate": 7.041887086012711e-05, "loss": 0.0203, "step": 39780 }, { "grad_norm": 0.2166503369808197, "learning_rate": 7.040377665363969e-05, "loss": 0.0178, "step": 39790 }, { "grad_norm": 0.23585617542266846, "learning_rate": 7.038868021582594e-05, "loss": 0.0195, "step": 39800 }, { "grad_norm": 0.26767826080322266, "learning_rate": 7.037358154833679e-05, "loss": 0.0163, "step": 39810 }, { "grad_norm": 0.23942233622074127, "learning_rate": 7.035848065282339e-05, "loss": 0.0232, "step": 39820 }, { "grad_norm": 0.2031271606683731, "learning_rate": 7.034337753093714e-05, "loss": 0.0175, "step": 39830 }, { "grad_norm": 0.19067013263702393, "learning_rate": 7.032827218432972e-05, "loss": 0.0146, "step": 39840 }, { "grad_norm": 0.20101429522037506, "learning_rate": 7.031316461465302e-05, "loss": 0.0147, "step": 39850 }, { "grad_norm": 0.22558680176734924, "learning_rate": 7.029805482355915e-05, "loss": 0.0159, "step": 39860 }, { "grad_norm": 0.17660096287727356, "learning_rate": 7.028294281270055e-05, "loss": 0.0166, "step": 39870 }, { "grad_norm": 0.17226329445838928, "learning_rate": 7.026782858372982e-05, "loss": 0.0183, "step": 39880 }, { "grad_norm": 0.18622608482837677, "learning_rate": 7.025271213829982e-05, "loss": 0.0137, "step": 39890 }, { "grad_norm": 0.1939893662929535, "learning_rate": 7.023759347806366e-05, "loss": 0.0169, "step": 39900 }, { "grad_norm": 0.17867442965507507, "learning_rate": 7.022247260467472e-05, "loss": 0.0176, "step": 39910 }, { "grad_norm": 0.20489223301410675, "learning_rate": 7.020734951978658e-05, "loss": 0.0181, "step": 39920 }, { "grad_norm": 0.15998607873916626, "learning_rate": 7.019222422505307e-05, "loss": 0.0164, "step": 39930 }, { "grad_norm": 0.1732274740934372, "learning_rate": 7.017709672212827e-05, "loss": 0.0196, "step": 39940 }, { "grad_norm": 0.19713209569454193, "learning_rate": 7.016196701266652e-05, "loss": 0.016, "step": 39950 }, { "grad_norm": 0.20484943687915802, "learning_rate": 7.014683509832238e-05, "loss": 0.0166, "step": 39960 }, { "grad_norm": 0.2793322503566742, "learning_rate": 7.013170098075063e-05, "loss": 0.0153, "step": 39970 }, { "grad_norm": 0.17632748186588287, "learning_rate": 7.011656466160632e-05, "loss": 0.0161, "step": 39980 }, { "grad_norm": 0.20772318542003632, "learning_rate": 7.010142614254475e-05, "loss": 0.0184, "step": 39990 }, { "grad_norm": 0.21468877792358398, "learning_rate": 7.008628542522147e-05, "loss": 0.0171, "step": 40000 }, { "grad_norm": 0.18295851349830627, "learning_rate": 7.007114251129218e-05, "loss": 0.0215, "step": 40010 }, { "grad_norm": 0.20275592803955078, "learning_rate": 7.005599740241293e-05, "loss": 0.0171, "step": 40020 }, { "grad_norm": 0.2514416575431824, "learning_rate": 7.004085010023996e-05, "loss": 0.0164, "step": 40030 }, { "grad_norm": 0.23401984572410583, "learning_rate": 7.002570060642976e-05, "loss": 0.0201, "step": 40040 }, { "grad_norm": 0.2602050304412842, "learning_rate": 7.001054892263903e-05, "loss": 0.0184, "step": 40050 }, { "grad_norm": 0.23449993133544922, "learning_rate": 6.999539505052477e-05, "loss": 0.016, "step": 40060 }, { "grad_norm": 0.18456433713436127, "learning_rate": 6.998023899174418e-05, "loss": 0.0154, "step": 40070 }, { "grad_norm": 0.2868987023830414, "learning_rate": 6.996508074795467e-05, "loss": 0.0163, "step": 40080 }, { "grad_norm": 0.22939932346343994, "learning_rate": 6.994992032081396e-05, "loss": 0.0159, "step": 40090 }, { "grad_norm": 0.18988659977912903, "learning_rate": 6.993475771197995e-05, "loss": 0.016, "step": 40100 }, { "grad_norm": 0.2262599915266037, "learning_rate": 6.991959292311082e-05, "loss": 0.0146, "step": 40110 }, { "grad_norm": 0.2337181121110916, "learning_rate": 6.990442595586495e-05, "loss": 0.0195, "step": 40120 }, { "grad_norm": 0.2760540843009949, "learning_rate": 6.988925681190098e-05, "loss": 0.0226, "step": 40130 }, { "grad_norm": 0.28096258640289307, "learning_rate": 6.987408549287778e-05, "loss": 0.0244, "step": 40140 }, { "grad_norm": 0.22438915073871613, "learning_rate": 6.985891200045449e-05, "loss": 0.0189, "step": 40150 }, { "grad_norm": 0.3037337064743042, "learning_rate": 6.984373633629045e-05, "loss": 0.0171, "step": 40160 }, { "grad_norm": 0.23046082258224487, "learning_rate": 6.98285585020452e-05, "loss": 0.0213, "step": 40170 }, { "grad_norm": 0.2174423635005951, "learning_rate": 6.981337849937864e-05, "loss": 0.0219, "step": 40180 }, { "grad_norm": 0.18549343943595886, "learning_rate": 6.979819632995078e-05, "loss": 0.0202, "step": 40190 }, { "grad_norm": 0.3041544258594513, "learning_rate": 6.978301199542193e-05, "loss": 0.0242, "step": 40200 }, { "grad_norm": 0.18175283074378967, "learning_rate": 6.976782549745263e-05, "loss": 0.0213, "step": 40210 }, { "grad_norm": 0.2467627078294754, "learning_rate": 6.975263683770367e-05, "loss": 0.0208, "step": 40220 }, { "grad_norm": 0.2162395417690277, "learning_rate": 6.973744601783604e-05, "loss": 0.0187, "step": 40230 }, { "grad_norm": 0.25887003540992737, "learning_rate": 6.972225303951097e-05, "loss": 0.023, "step": 40240 }, { "grad_norm": 0.1935572624206543, "learning_rate": 6.970705790438998e-05, "loss": 0.0188, "step": 40250 }, { "grad_norm": 0.19840237498283386, "learning_rate": 6.969186061413477e-05, "loss": 0.0205, "step": 40260 }, { "grad_norm": 0.21921192109584808, "learning_rate": 6.967666117040727e-05, "loss": 0.0208, "step": 40270 }, { "grad_norm": 0.2735806107521057, "learning_rate": 6.966145957486972e-05, "loss": 0.0172, "step": 40280 }, { "grad_norm": 0.2685406506061554, "learning_rate": 6.964625582918449e-05, "loss": 0.016, "step": 40290 }, { "grad_norm": 0.24378541111946106, "learning_rate": 6.963104993501425e-05, "loss": 0.0264, "step": 40300 }, { "grad_norm": 0.14662781357765198, "learning_rate": 6.961584189402192e-05, "loss": 0.0163, "step": 40310 }, { "grad_norm": 0.25253763794898987, "learning_rate": 6.96006317078706e-05, "loss": 0.0192, "step": 40320 }, { "grad_norm": 0.2011013627052307, "learning_rate": 6.958541937822367e-05, "loss": 0.0153, "step": 40330 }, { "grad_norm": 0.21329987049102783, "learning_rate": 6.957020490674473e-05, "loss": 0.0172, "step": 40340 }, { "grad_norm": 0.2020692080259323, "learning_rate": 6.95549882950976e-05, "loss": 0.0146, "step": 40350 }, { "grad_norm": 0.11885959655046463, "learning_rate": 6.953976954494635e-05, "loss": 0.0153, "step": 40360 }, { "grad_norm": 0.15886645019054413, "learning_rate": 6.952454865795528e-05, "loss": 0.0192, "step": 40370 }, { "grad_norm": 0.1584453135728836, "learning_rate": 6.950932563578892e-05, "loss": 0.0177, "step": 40380 }, { "grad_norm": 0.25282618403434753, "learning_rate": 6.949410048011206e-05, "loss": 0.0174, "step": 40390 }, { "grad_norm": 0.14616410434246063, "learning_rate": 6.947887319258966e-05, "loss": 0.0134, "step": 40400 }, { "grad_norm": 0.20320004224777222, "learning_rate": 6.946364377488696e-05, "loss": 0.0177, "step": 40410 }, { "grad_norm": 0.18556761741638184, "learning_rate": 6.944841222866947e-05, "loss": 0.0151, "step": 40420 }, { "grad_norm": 0.17014998197555542, "learning_rate": 6.943317855560284e-05, "loss": 0.015, "step": 40430 }, { "grad_norm": 0.17083388566970825, "learning_rate": 6.941794275735302e-05, "loss": 0.0212, "step": 40440 }, { "grad_norm": 0.20023876428604126, "learning_rate": 6.94027048355862e-05, "loss": 0.0148, "step": 40450 }, { "grad_norm": 0.15054914355278015, "learning_rate": 6.938746479196871e-05, "loss": 0.0143, "step": 40460 }, { "grad_norm": 0.18284687399864197, "learning_rate": 6.937222262816724e-05, "loss": 0.0124, "step": 40470 }, { "grad_norm": 0.2420550435781479, "learning_rate": 6.935697834584865e-05, "loss": 0.0176, "step": 40480 }, { "grad_norm": 0.18980517983436584, "learning_rate": 6.934173194667998e-05, "loss": 0.0195, "step": 40490 }, { "grad_norm": 0.14706619083881378, "learning_rate": 6.93264834323286e-05, "loss": 0.0166, "step": 40500 }, { "grad_norm": 0.19980840384960175, "learning_rate": 6.931123280446206e-05, "loss": 0.0183, "step": 40510 }, { "grad_norm": 0.21619726717472076, "learning_rate": 6.929598006474811e-05, "loss": 0.0171, "step": 40520 }, { "grad_norm": 0.23811742663383484, "learning_rate": 6.92807252148548e-05, "loss": 0.0177, "step": 40530 }, { "grad_norm": 0.252111554145813, "learning_rate": 6.926546825645037e-05, "loss": 0.0161, "step": 40540 }, { "grad_norm": 0.22782136499881744, "learning_rate": 6.92502091912033e-05, "loss": 0.0173, "step": 40550 }, { "grad_norm": 0.22793354094028473, "learning_rate": 6.923494802078229e-05, "loss": 0.0176, "step": 40560 }, { "grad_norm": 0.2098301202058792, "learning_rate": 6.921968474685628e-05, "loss": 0.0161, "step": 40570 }, { "grad_norm": 0.19742083549499512, "learning_rate": 6.920441937109445e-05, "loss": 0.017, "step": 40580 }, { "grad_norm": 0.22746416926383972, "learning_rate": 6.91891518951662e-05, "loss": 0.0171, "step": 40590 }, { "grad_norm": 0.16335277259349823, "learning_rate": 6.917388232074114e-05, "loss": 0.017, "step": 40600 }, { "grad_norm": 0.1747017651796341, "learning_rate": 6.915861064948914e-05, "loss": 0.0197, "step": 40610 }, { "grad_norm": 0.20015455782413483, "learning_rate": 6.914333688308025e-05, "loss": 0.0217, "step": 40620 }, { "grad_norm": 0.2087722271680832, "learning_rate": 6.912806102318487e-05, "loss": 0.0151, "step": 40630 }, { "grad_norm": 0.24294811487197876, "learning_rate": 6.911278307147347e-05, "loss": 0.017, "step": 40640 }, { "grad_norm": 0.1781114786863327, "learning_rate": 6.909750302961684e-05, "loss": 0.0167, "step": 40650 }, { "grad_norm": 0.17684321105480194, "learning_rate": 6.9082220899286e-05, "loss": 0.016, "step": 40660 }, { "grad_norm": 0.20391395688056946, "learning_rate": 6.906693668215219e-05, "loss": 0.017, "step": 40670 }, { "grad_norm": 0.1410200297832489, "learning_rate": 6.905165037988683e-05, "loss": 0.0139, "step": 40680 }, { "grad_norm": 0.17491447925567627, "learning_rate": 6.903636199416164e-05, "loss": 0.0149, "step": 40690 }, { "grad_norm": 0.23639358580112457, "learning_rate": 6.902107152664851e-05, "loss": 0.02, "step": 40700 }, { "grad_norm": 0.23080337047576904, "learning_rate": 6.90057789790196e-05, "loss": 0.0175, "step": 40710 }, { "grad_norm": 0.2383696436882019, "learning_rate": 6.899048435294728e-05, "loss": 0.0179, "step": 40720 }, { "grad_norm": 0.13670377433300018, "learning_rate": 6.897518765010415e-05, "loss": 0.0164, "step": 40730 }, { "grad_norm": 0.2162369340658188, "learning_rate": 6.895988887216302e-05, "loss": 0.018, "step": 40740 }, { "grad_norm": 0.18776558339595795, "learning_rate": 6.894458802079694e-05, "loss": 0.015, "step": 40750 }, { "grad_norm": 0.19745981693267822, "learning_rate": 6.892928509767922e-05, "loss": 0.0195, "step": 40760 }, { "grad_norm": 0.20370303094387054, "learning_rate": 6.891398010448333e-05, "loss": 0.0172, "step": 40770 }, { "grad_norm": 0.2687881588935852, "learning_rate": 6.8898673042883e-05, "loss": 0.0194, "step": 40780 }, { "grad_norm": 0.16521598398685455, "learning_rate": 6.888336391455222e-05, "loss": 0.0141, "step": 40790 }, { "grad_norm": 0.19829951226711273, "learning_rate": 6.886805272116513e-05, "loss": 0.0227, "step": 40800 }, { "grad_norm": 0.2016734778881073, "learning_rate": 6.885273946439617e-05, "loss": 0.0185, "step": 40810 }, { "grad_norm": 0.15010195970535278, "learning_rate": 6.883742414591998e-05, "loss": 0.0155, "step": 40820 }, { "grad_norm": 0.2887018620967865, "learning_rate": 6.88221067674114e-05, "loss": 0.0145, "step": 40830 }, { "grad_norm": 0.14770160615444183, "learning_rate": 6.88067873305455e-05, "loss": 0.0136, "step": 40840 }, { "grad_norm": 0.1746395230293274, "learning_rate": 6.879146583699765e-05, "loss": 0.014, "step": 40850 }, { "grad_norm": 0.1529594212770462, "learning_rate": 6.877614228844334e-05, "loss": 0.014, "step": 40860 }, { "grad_norm": 0.1809457540512085, "learning_rate": 6.876081668655832e-05, "loss": 0.0144, "step": 40870 }, { "grad_norm": 0.15114463865756989, "learning_rate": 6.874548903301861e-05, "loss": 0.0196, "step": 40880 }, { "grad_norm": 0.24162203073501587, "learning_rate": 6.873015932950039e-05, "loss": 0.0161, "step": 40890 }, { "grad_norm": 0.19524338841438293, "learning_rate": 6.871482757768012e-05, "loss": 0.0153, "step": 40900 }, { "grad_norm": 0.2610142230987549, "learning_rate": 6.869949377923442e-05, "loss": 0.0205, "step": 40910 }, { "grad_norm": 0.18015822768211365, "learning_rate": 6.868415793584022e-05, "loss": 0.0169, "step": 40920 }, { "grad_norm": 0.1888425201177597, "learning_rate": 6.86688200491746e-05, "loss": 0.0192, "step": 40930 }, { "grad_norm": 0.18657371401786804, "learning_rate": 6.865348012091488e-05, "loss": 0.0174, "step": 40940 }, { "grad_norm": 0.2228395789861679, "learning_rate": 6.86381381527386e-05, "loss": 0.0206, "step": 40950 }, { "grad_norm": 0.16488493978977203, "learning_rate": 6.862279414632359e-05, "loss": 0.0135, "step": 40960 }, { "grad_norm": 0.2557877004146576, "learning_rate": 6.860744810334778e-05, "loss": 0.0212, "step": 40970 }, { "grad_norm": 0.20149415731430054, "learning_rate": 6.859210002548943e-05, "loss": 0.0166, "step": 40980 }, { "grad_norm": 0.20110999047756195, "learning_rate": 6.8576749914427e-05, "loss": 0.0173, "step": 40990 }, { "grad_norm": 0.1719086766242981, "learning_rate": 6.85613977718391e-05, "loss": 0.0156, "step": 41000 }, { "grad_norm": 0.14989741146564484, "learning_rate": 6.854604359940464e-05, "loss": 0.0149, "step": 41010 }, { "grad_norm": 0.1391681730747223, "learning_rate": 6.853068739880276e-05, "loss": 0.015, "step": 41020 }, { "grad_norm": 0.24814602732658386, "learning_rate": 6.851532917171276e-05, "loss": 0.0238, "step": 41030 }, { "grad_norm": 0.16483117640018463, "learning_rate": 6.849996891981418e-05, "loss": 0.0129, "step": 41040 }, { "grad_norm": 0.19447635114192963, "learning_rate": 6.848460664478684e-05, "loss": 0.0192, "step": 41050 }, { "grad_norm": 0.18414172530174255, "learning_rate": 6.846924234831069e-05, "loss": 0.017, "step": 41060 }, { "grad_norm": 0.21126030385494232, "learning_rate": 6.845387603206597e-05, "loss": 0.0148, "step": 41070 }, { "grad_norm": 0.15274280309677124, "learning_rate": 6.843850769773312e-05, "loss": 0.0119, "step": 41080 }, { "grad_norm": 0.15466465055942535, "learning_rate": 6.842313734699278e-05, "loss": 0.0134, "step": 41090 }, { "grad_norm": 0.20191338658332825, "learning_rate": 6.840776498152584e-05, "loss": 0.0191, "step": 41100 }, { "grad_norm": 0.2042243778705597, "learning_rate": 6.83923906030134e-05, "loss": 0.0162, "step": 41110 }, { "grad_norm": 0.167095348238945, "learning_rate": 6.837701421313677e-05, "loss": 0.0187, "step": 41120 }, { "grad_norm": 0.21614401042461395, "learning_rate": 6.83616358135775e-05, "loss": 0.0182, "step": 41130 }, { "grad_norm": 0.15960413217544556, "learning_rate": 6.834625540601734e-05, "loss": 0.0174, "step": 41140 }, { "grad_norm": 0.19655950367450714, "learning_rate": 6.833087299213829e-05, "loss": 0.0151, "step": 41150 }, { "grad_norm": 0.29980382323265076, "learning_rate": 6.83154885736225e-05, "loss": 0.0242, "step": 41160 }, { "grad_norm": 0.19060330092906952, "learning_rate": 6.830010215215244e-05, "loss": 0.0162, "step": 41170 }, { "grad_norm": 0.18280698359012604, "learning_rate": 6.828471372941072e-05, "loss": 0.0149, "step": 41180 }, { "grad_norm": 0.15344741940498352, "learning_rate": 6.826932330708018e-05, "loss": 0.0149, "step": 41190 }, { "grad_norm": 0.17695510387420654, "learning_rate": 6.825393088684393e-05, "loss": 0.015, "step": 41200 }, { "grad_norm": 0.2174452841281891, "learning_rate": 6.823853647038524e-05, "loss": 0.019, "step": 41210 }, { "grad_norm": 0.4431777000427246, "learning_rate": 6.822314005938762e-05, "loss": 0.0155, "step": 41220 }, { "grad_norm": 0.21307097375392914, "learning_rate": 6.820774165553479e-05, "loss": 0.0148, "step": 41230 }, { "grad_norm": 0.19140349328517914, "learning_rate": 6.819234126051072e-05, "loss": 0.0142, "step": 41240 }, { "grad_norm": 0.2009289711713791, "learning_rate": 6.817693887599956e-05, "loss": 0.0159, "step": 41250 }, { "grad_norm": 0.20548929274082184, "learning_rate": 6.816153450368568e-05, "loss": 0.02, "step": 41260 }, { "grad_norm": 0.1799306422472, "learning_rate": 6.814612814525372e-05, "loss": 0.0182, "step": 41270 }, { "grad_norm": 0.2095482498407364, "learning_rate": 6.813071980238846e-05, "loss": 0.0146, "step": 41280 }, { "grad_norm": 0.22977802157402039, "learning_rate": 6.811530947677492e-05, "loss": 0.0152, "step": 41290 }, { "grad_norm": 0.1939115673303604, "learning_rate": 6.809989717009839e-05, "loss": 0.02, "step": 41300 }, { "grad_norm": 0.1269001066684723, "learning_rate": 6.808448288404431e-05, "loss": 0.0172, "step": 41310 }, { "grad_norm": 0.19089901447296143, "learning_rate": 6.806906662029838e-05, "loss": 0.0175, "step": 41320 }, { "grad_norm": 0.2200208306312561, "learning_rate": 6.805364838054647e-05, "loss": 0.021, "step": 41330 }, { "grad_norm": 0.22612953186035156, "learning_rate": 6.803822816647474e-05, "loss": 0.0157, "step": 41340 }, { "grad_norm": 0.16478659212589264, "learning_rate": 6.802280597976949e-05, "loss": 0.0143, "step": 41350 }, { "grad_norm": 0.19884032011032104, "learning_rate": 6.800738182211727e-05, "loss": 0.0193, "step": 41360 }, { "grad_norm": 0.16161580383777618, "learning_rate": 6.799195569520485e-05, "loss": 0.0184, "step": 41370 }, { "grad_norm": 0.1736745685338974, "learning_rate": 6.79765276007192e-05, "loss": 0.0157, "step": 41380 }, { "grad_norm": 0.1658681184053421, "learning_rate": 6.796109754034753e-05, "loss": 0.019, "step": 41390 }, { "grad_norm": 0.21965017914772034, "learning_rate": 6.794566551577724e-05, "loss": 0.0156, "step": 41400 }, { "grad_norm": 0.2079160213470459, "learning_rate": 6.793023152869594e-05, "loss": 0.0141, "step": 41410 }, { "grad_norm": 0.18286362290382385, "learning_rate": 6.791479558079147e-05, "loss": 0.016, "step": 41420 }, { "grad_norm": 0.19134065508842468, "learning_rate": 6.789935767375191e-05, "loss": 0.0185, "step": 41430 }, { "grad_norm": 0.21435892581939697, "learning_rate": 6.78839178092655e-05, "loss": 0.0161, "step": 41440 }, { "grad_norm": 0.18327316641807556, "learning_rate": 6.786847598902072e-05, "loss": 0.0187, "step": 41450 }, { "grad_norm": 0.2095876932144165, "learning_rate": 6.785303221470627e-05, "loss": 0.018, "step": 41460 }, { "grad_norm": 0.21562261879444122, "learning_rate": 6.783758648801108e-05, "loss": 0.0167, "step": 41470 }, { "grad_norm": 0.19448496401309967, "learning_rate": 6.782213881062422e-05, "loss": 0.0183, "step": 41480 }, { "grad_norm": 0.19522397220134735, "learning_rate": 6.780668918423508e-05, "loss": 0.0169, "step": 41490 }, { "grad_norm": 0.23406387865543365, "learning_rate": 6.779123761053317e-05, "loss": 0.0145, "step": 41500 }, { "grad_norm": 0.194196417927742, "learning_rate": 6.777578409120826e-05, "loss": 0.019, "step": 41510 }, { "grad_norm": 0.23368752002716064, "learning_rate": 6.776032862795033e-05, "loss": 0.018, "step": 41520 }, { "grad_norm": 0.22792139649391174, "learning_rate": 6.774487122244956e-05, "loss": 0.0241, "step": 41530 }, { "grad_norm": 0.22873744368553162, "learning_rate": 6.772941187639637e-05, "loss": 0.0187, "step": 41540 }, { "grad_norm": 0.21847939491271973, "learning_rate": 6.771395059148134e-05, "loss": 0.0214, "step": 41550 }, { "grad_norm": 0.19318191707134247, "learning_rate": 6.76984873693953e-05, "loss": 0.0189, "step": 41560 }, { "grad_norm": 0.1569042205810547, "learning_rate": 6.768302221182931e-05, "loss": 0.018, "step": 41570 }, { "grad_norm": 0.23716242611408234, "learning_rate": 6.766755512047457e-05, "loss": 0.0176, "step": 41580 }, { "grad_norm": 0.17265936732292175, "learning_rate": 6.765208609702259e-05, "loss": 0.0179, "step": 41590 }, { "grad_norm": 0.21860481798648834, "learning_rate": 6.763661514316499e-05, "loss": 0.0173, "step": 41600 }, { "grad_norm": 0.12711665034294128, "learning_rate": 6.76211422605937e-05, "loss": 0.0139, "step": 41610 }, { "grad_norm": 0.14200784265995026, "learning_rate": 6.760566745100076e-05, "loss": 0.017, "step": 41620 }, { "grad_norm": 0.18141208589076996, "learning_rate": 6.75901907160785e-05, "loss": 0.0195, "step": 41630 }, { "grad_norm": 0.14965863525867462, "learning_rate": 6.757471205751943e-05, "loss": 0.0132, "step": 41640 }, { "grad_norm": 0.22185321152210236, "learning_rate": 6.75592314770163e-05, "loss": 0.0163, "step": 41650 }, { "grad_norm": 0.16152077913284302, "learning_rate": 6.7543748976262e-05, "loss": 0.0158, "step": 41660 }, { "grad_norm": 0.213151216506958, "learning_rate": 6.752826455694968e-05, "loss": 0.0159, "step": 41670 }, { "grad_norm": 0.1602344810962677, "learning_rate": 6.751277822077271e-05, "loss": 0.0151, "step": 41680 }, { "grad_norm": 0.11140687018632889, "learning_rate": 6.749728996942465e-05, "loss": 0.0161, "step": 41690 }, { "grad_norm": 0.24120262265205383, "learning_rate": 6.748179980459924e-05, "loss": 0.0164, "step": 41700 }, { "grad_norm": 0.16227763891220093, "learning_rate": 6.746630772799052e-05, "loss": 0.0177, "step": 41710 }, { "grad_norm": 0.6235761046409607, "learning_rate": 6.745081374129262e-05, "loss": 0.0187, "step": 41720 }, { "grad_norm": 0.1501588523387909, "learning_rate": 6.743531784619998e-05, "loss": 0.0147, "step": 41730 }, { "grad_norm": 0.17430061101913452, "learning_rate": 6.741982004440719e-05, "loss": 0.0175, "step": 41740 }, { "grad_norm": 0.16159436106681824, "learning_rate": 6.740432033760907e-05, "loss": 0.0156, "step": 41750 }, { "grad_norm": 0.19233158230781555, "learning_rate": 6.738881872750066e-05, "loss": 0.0192, "step": 41760 }, { "grad_norm": 0.20109066367149353, "learning_rate": 6.737331521577715e-05, "loss": 0.0149, "step": 41770 }, { "grad_norm": 0.18657268583774567, "learning_rate": 6.735780980413403e-05, "loss": 0.0166, "step": 41780 }, { "grad_norm": 0.19434529542922974, "learning_rate": 6.734230249426692e-05, "loss": 0.0153, "step": 41790 }, { "grad_norm": 0.19658608734607697, "learning_rate": 6.732679328787168e-05, "loss": 0.0162, "step": 41800 }, { "grad_norm": 0.21731902658939362, "learning_rate": 6.731128218664438e-05, "loss": 0.0185, "step": 41810 }, { "grad_norm": 0.20028744637966156, "learning_rate": 6.729576919228129e-05, "loss": 0.0136, "step": 41820 }, { "grad_norm": 0.1983109712600708, "learning_rate": 6.728025430647888e-05, "loss": 0.0169, "step": 41830 }, { "grad_norm": 0.1974511593580246, "learning_rate": 6.726473753093383e-05, "loss": 0.0152, "step": 41840 }, { "grad_norm": 0.20618420839309692, "learning_rate": 6.724921886734305e-05, "loss": 0.0164, "step": 41850 }, { "grad_norm": 0.21274533867835999, "learning_rate": 6.723369831740363e-05, "loss": 0.0155, "step": 41860 }, { "grad_norm": 0.21237275004386902, "learning_rate": 6.721817588281288e-05, "loss": 0.0213, "step": 41870 }, { "grad_norm": 0.19783127307891846, "learning_rate": 6.720265156526828e-05, "loss": 0.0158, "step": 41880 }, { "grad_norm": 0.19617140293121338, "learning_rate": 6.718712536646758e-05, "loss": 0.0152, "step": 41890 }, { "grad_norm": 0.19647632539272308, "learning_rate": 6.71715972881087e-05, "loss": 0.015, "step": 41900 }, { "grad_norm": 0.1787106990814209, "learning_rate": 6.715606733188974e-05, "loss": 0.0146, "step": 41910 }, { "grad_norm": 0.1670295000076294, "learning_rate": 6.714053549950909e-05, "loss": 0.0154, "step": 41920 }, { "grad_norm": 0.1912410855293274, "learning_rate": 6.71250017926652e-05, "loss": 0.0184, "step": 41930 }, { "grad_norm": 0.21979328989982605, "learning_rate": 6.71094662130569e-05, "loss": 0.0156, "step": 41940 }, { "grad_norm": 0.30992498993873596, "learning_rate": 6.709392876238307e-05, "loss": 0.0153, "step": 41950 }, { "grad_norm": 0.24441300332546234, "learning_rate": 6.707838944234289e-05, "loss": 0.0173, "step": 41960 }, { "grad_norm": 0.20464150607585907, "learning_rate": 6.706284825463574e-05, "loss": 0.0142, "step": 41970 }, { "grad_norm": 0.20779262483119965, "learning_rate": 6.704730520096114e-05, "loss": 0.0158, "step": 41980 }, { "grad_norm": 0.176188126206398, "learning_rate": 6.703176028301888e-05, "loss": 0.022, "step": 41990 }, { "grad_norm": 0.2221212089061737, "learning_rate": 6.701621350250892e-05, "loss": 0.0184, "step": 42000 }, { "grad_norm": 0.18931317329406738, "learning_rate": 6.700066486113144e-05, "loss": 0.0162, "step": 42010 }, { "grad_norm": 0.2078658640384674, "learning_rate": 6.69851143605868e-05, "loss": 0.0141, "step": 42020 }, { "grad_norm": 0.16458414494991302, "learning_rate": 6.696956200257561e-05, "loss": 0.0151, "step": 42030 }, { "grad_norm": 0.16772346198558807, "learning_rate": 6.695400778879863e-05, "loss": 0.0206, "step": 42040 }, { "grad_norm": 0.15805663168430328, "learning_rate": 6.693845172095683e-05, "loss": 0.0173, "step": 42050 }, { "grad_norm": 0.15235623717308044, "learning_rate": 6.692289380075142e-05, "loss": 0.0129, "step": 42060 }, { "grad_norm": 0.17836633324623108, "learning_rate": 6.690733402988379e-05, "loss": 0.0158, "step": 42070 }, { "grad_norm": 0.21074780821800232, "learning_rate": 6.689177241005553e-05, "loss": 0.0161, "step": 42080 }, { "grad_norm": 0.22981631755828857, "learning_rate": 6.687620894296844e-05, "loss": 0.0179, "step": 42090 }, { "grad_norm": 0.21797871589660645, "learning_rate": 6.686064363032451e-05, "loss": 0.0204, "step": 42100 }, { "grad_norm": 0.1458207070827484, "learning_rate": 6.684507647382596e-05, "loss": 0.0115, "step": 42110 }, { "grad_norm": 0.2459738701581955, "learning_rate": 6.682950747517513e-05, "loss": 0.0157, "step": 42120 }, { "grad_norm": 0.19565057754516602, "learning_rate": 6.68139366360747e-05, "loss": 0.0209, "step": 42130 }, { "grad_norm": 0.132188618183136, "learning_rate": 6.679836395822744e-05, "loss": 0.0151, "step": 42140 }, { "grad_norm": 0.17269223928451538, "learning_rate": 6.678278944333633e-05, "loss": 0.0182, "step": 42150 }, { "grad_norm": 0.1406688094139099, "learning_rate": 6.676721309310462e-05, "loss": 0.0157, "step": 42160 }, { "grad_norm": 0.1839502602815628, "learning_rate": 6.67516349092357e-05, "loss": 0.015, "step": 42170 }, { "grad_norm": 0.2167157530784607, "learning_rate": 6.673605489343315e-05, "loss": 0.0157, "step": 42180 }, { "grad_norm": 0.1442970484495163, "learning_rate": 6.672047304740081e-05, "loss": 0.0144, "step": 42190 }, { "grad_norm": 0.14300128817558289, "learning_rate": 6.670488937284268e-05, "loss": 0.0158, "step": 42200 }, { "grad_norm": 0.18801279366016388, "learning_rate": 6.668930387146294e-05, "loss": 0.0148, "step": 42210 }, { "grad_norm": 0.19972947239875793, "learning_rate": 6.667371654496605e-05, "loss": 0.0226, "step": 42220 }, { "grad_norm": 0.16932152211666107, "learning_rate": 6.665812739505659e-05, "loss": 0.0168, "step": 42230 }, { "grad_norm": 0.24369274079799652, "learning_rate": 6.664253642343935e-05, "loss": 0.0133, "step": 42240 }, { "grad_norm": 0.19320376217365265, "learning_rate": 6.662694363181935e-05, "loss": 0.0191, "step": 42250 }, { "grad_norm": 0.21764373779296875, "learning_rate": 6.66113490219018e-05, "loss": 0.0172, "step": 42260 }, { "grad_norm": 0.207135871052742, "learning_rate": 6.659575259539207e-05, "loss": 0.0141, "step": 42270 }, { "grad_norm": 0.18415763974189758, "learning_rate": 6.65801543539958e-05, "loss": 0.0201, "step": 42280 }, { "grad_norm": 0.18161357939243317, "learning_rate": 6.65645542994188e-05, "loss": 0.0189, "step": 42290 }, { "grad_norm": 0.25524571537971497, "learning_rate": 6.654895243336702e-05, "loss": 0.017, "step": 42300 }, { "grad_norm": 0.16110777854919434, "learning_rate": 6.65333487575467e-05, "loss": 0.0175, "step": 42310 }, { "grad_norm": 0.14858417212963104, "learning_rate": 6.651774327366422e-05, "loss": 0.0162, "step": 42320 }, { "grad_norm": 0.17690502107143402, "learning_rate": 6.650213598342613e-05, "loss": 0.0154, "step": 42330 }, { "grad_norm": 0.11473582684993744, "learning_rate": 6.648652688853931e-05, "loss": 0.0158, "step": 42340 }, { "grad_norm": 0.18141010403633118, "learning_rate": 6.647091599071066e-05, "loss": 0.0188, "step": 42350 }, { "grad_norm": 0.20724451541900635, "learning_rate": 6.645530329164743e-05, "loss": 0.0162, "step": 42360 }, { "grad_norm": 0.2597137987613678, "learning_rate": 6.643968879305694e-05, "loss": 0.0235, "step": 42370 }, { "grad_norm": 0.21922993659973145, "learning_rate": 6.642407249664683e-05, "loss": 0.0167, "step": 42380 }, { "grad_norm": 0.163998544216156, "learning_rate": 6.640845440412483e-05, "loss": 0.0147, "step": 42390 }, { "grad_norm": 0.1695224642753601, "learning_rate": 6.639283451719893e-05, "loss": 0.0174, "step": 42400 }, { "grad_norm": 0.16901031136512756, "learning_rate": 6.637721283757729e-05, "loss": 0.0149, "step": 42410 }, { "grad_norm": 0.2378845363855362, "learning_rate": 6.636158936696828e-05, "loss": 0.0167, "step": 42420 }, { "grad_norm": 0.20877885818481445, "learning_rate": 6.634596410708047e-05, "loss": 0.0187, "step": 42430 }, { "grad_norm": 0.22676415741443634, "learning_rate": 6.633033705962257e-05, "loss": 0.0158, "step": 42440 }, { "grad_norm": 0.1908356249332428, "learning_rate": 6.631470822630359e-05, "loss": 0.0168, "step": 42450 }, { "grad_norm": 0.276064395904541, "learning_rate": 6.629907760883263e-05, "loss": 0.0191, "step": 42460 }, { "grad_norm": 0.23406153917312622, "learning_rate": 6.628344520891905e-05, "loss": 0.0145, "step": 42470 }, { "grad_norm": 0.19676530361175537, "learning_rate": 6.62678110282724e-05, "loss": 0.0176, "step": 42480 }, { "grad_norm": 0.17071853578090668, "learning_rate": 6.625217506860237e-05, "loss": 0.0163, "step": 42490 }, { "grad_norm": 0.16070397198200226, "learning_rate": 6.623653733161892e-05, "loss": 0.0152, "step": 42500 }, { "grad_norm": 0.3088975250720978, "learning_rate": 6.622089781903217e-05, "loss": 0.0159, "step": 42510 }, { "grad_norm": 0.20132692158222198, "learning_rate": 6.620525653255241e-05, "loss": 0.0159, "step": 42520 }, { "grad_norm": 0.22682934999465942, "learning_rate": 6.618961347389015e-05, "loss": 0.0202, "step": 42530 }, { "grad_norm": 0.28349095582962036, "learning_rate": 6.617396864475613e-05, "loss": 0.0219, "step": 42540 }, { "grad_norm": 0.2563677430152893, "learning_rate": 6.61583220468612e-05, "loss": 0.0183, "step": 42550 }, { "grad_norm": 0.22137382626533508, "learning_rate": 6.614267368191645e-05, "loss": 0.0194, "step": 42560 }, { "grad_norm": 0.23275169730186462, "learning_rate": 6.61270235516332e-05, "loss": 0.0165, "step": 42570 }, { "grad_norm": 0.16230115294456482, "learning_rate": 6.611137165772287e-05, "loss": 0.0181, "step": 42580 }, { "grad_norm": 0.15811225771903992, "learning_rate": 6.60957180018972e-05, "loss": 0.0161, "step": 42590 }, { "grad_norm": 0.22832044959068298, "learning_rate": 6.608006258586797e-05, "loss": 0.0149, "step": 42600 }, { "grad_norm": 0.22572776675224304, "learning_rate": 6.606440541134731e-05, "loss": 0.0189, "step": 42610 }, { "grad_norm": 0.21715618669986725, "learning_rate": 6.60487464800474e-05, "loss": 0.0182, "step": 42620 }, { "grad_norm": 0.17969444394111633, "learning_rate": 6.603308579368071e-05, "loss": 0.0175, "step": 42630 }, { "grad_norm": 0.16040591895580292, "learning_rate": 6.601742335395987e-05, "loss": 0.0191, "step": 42640 }, { "grad_norm": 0.17932607233524323, "learning_rate": 6.600175916259769e-05, "loss": 0.0154, "step": 42650 }, { "grad_norm": 0.14059939980506897, "learning_rate": 6.598609322130718e-05, "loss": 0.0134, "step": 42660 }, { "grad_norm": 0.1788853108882904, "learning_rate": 6.597042553180155e-05, "loss": 0.0177, "step": 42670 }, { "grad_norm": 0.18506775796413422, "learning_rate": 6.59547560957942e-05, "loss": 0.0123, "step": 42680 }, { "grad_norm": 0.18940889835357666, "learning_rate": 6.59390849149987e-05, "loss": 0.0143, "step": 42690 }, { "grad_norm": 0.2823818027973175, "learning_rate": 6.592341199112886e-05, "loss": 0.019, "step": 42700 }, { "grad_norm": 0.20293337106704712, "learning_rate": 6.590773732589861e-05, "loss": 0.0149, "step": 42710 }, { "grad_norm": 0.1922113299369812, "learning_rate": 6.589206092102213e-05, "loss": 0.0156, "step": 42720 }, { "grad_norm": 0.1589195728302002, "learning_rate": 6.587638277821375e-05, "loss": 0.0143, "step": 42730 }, { "grad_norm": 0.16032835841178894, "learning_rate": 6.586070289918803e-05, "loss": 0.0162, "step": 42740 }, { "grad_norm": 0.21618033945560455, "learning_rate": 6.584502128565968e-05, "loss": 0.0142, "step": 42750 }, { "grad_norm": 0.1682760864496231, "learning_rate": 6.582933793934363e-05, "loss": 0.0124, "step": 42760 }, { "grad_norm": 0.23437102138996124, "learning_rate": 6.581365286195498e-05, "loss": 0.019, "step": 42770 }, { "grad_norm": 0.16348165273666382, "learning_rate": 6.579796605520903e-05, "loss": 0.0159, "step": 42780 }, { "grad_norm": 0.16602922976016998, "learning_rate": 6.578227752082127e-05, "loss": 0.0167, "step": 42790 }, { "grad_norm": 0.17291712760925293, "learning_rate": 6.576658726050735e-05, "loss": 0.0162, "step": 42800 }, { "grad_norm": 0.46956315636634827, "learning_rate": 6.575089527598316e-05, "loss": 0.0171, "step": 42810 }, { "grad_norm": 0.16154862940311432, "learning_rate": 6.573520156896476e-05, "loss": 0.0147, "step": 42820 }, { "grad_norm": 0.16761085391044617, "learning_rate": 6.571950614116835e-05, "loss": 0.0187, "step": 42830 }, { "grad_norm": 0.27345412969589233, "learning_rate": 6.570380899431039e-05, "loss": 0.0158, "step": 42840 }, { "grad_norm": 0.2251947671175003, "learning_rate": 6.568811013010749e-05, "loss": 0.0156, "step": 42850 }, { "grad_norm": 0.2633896768093109, "learning_rate": 6.567240955027645e-05, "loss": 0.0166, "step": 42860 }, { "grad_norm": 0.2334805279970169, "learning_rate": 6.565670725653427e-05, "loss": 0.0207, "step": 42870 }, { "grad_norm": 0.18128295242786407, "learning_rate": 6.564100325059812e-05, "loss": 0.0159, "step": 42880 }, { "grad_norm": 0.2425989806652069, "learning_rate": 6.562529753418538e-05, "loss": 0.0212, "step": 42890 }, { "grad_norm": 0.17900636792182922, "learning_rate": 6.56095901090136e-05, "loss": 0.0196, "step": 42900 }, { "grad_norm": 0.17615658044815063, "learning_rate": 6.55938809768005e-05, "loss": 0.0168, "step": 42910 }, { "grad_norm": 0.20883305370807648, "learning_rate": 6.557817013926402e-05, "loss": 0.0132, "step": 42920 }, { "grad_norm": 0.17589452862739563, "learning_rate": 6.556245759812227e-05, "loss": 0.014, "step": 42930 }, { "grad_norm": 0.17424383759498596, "learning_rate": 6.554674335509357e-05, "loss": 0.0145, "step": 42940 }, { "grad_norm": 0.2590521574020386, "learning_rate": 6.553102741189638e-05, "loss": 0.0163, "step": 42950 }, { "grad_norm": 0.23417596518993378, "learning_rate": 6.55153097702494e-05, "loss": 0.0157, "step": 42960 }, { "grad_norm": 0.16798986494541168, "learning_rate": 6.549959043187146e-05, "loss": 0.0169, "step": 42970 }, { "grad_norm": 0.1566605269908905, "learning_rate": 6.54838693984816e-05, "loss": 0.0146, "step": 42980 }, { "grad_norm": 0.17462283372879028, "learning_rate": 6.546814667179909e-05, "loss": 0.0198, "step": 42990 }, { "grad_norm": 0.1832127422094345, "learning_rate": 6.545242225354328e-05, "loss": 0.0159, "step": 43000 }, { "grad_norm": 0.19463975727558136, "learning_rate": 6.543669614543383e-05, "loss": 0.0146, "step": 43010 }, { "grad_norm": 0.1748269647359848, "learning_rate": 6.542096834919049e-05, "loss": 0.0137, "step": 43020 }, { "grad_norm": 0.22325940430164337, "learning_rate": 6.540523886653324e-05, "loss": 0.0159, "step": 43030 }, { "grad_norm": 0.20933300256729126, "learning_rate": 6.538950769918222e-05, "loss": 0.0187, "step": 43040 }, { "grad_norm": 0.18122373521327972, "learning_rate": 6.537377484885779e-05, "loss": 0.0186, "step": 43050 }, { "grad_norm": 0.20695997774600983, "learning_rate": 6.535804031728045e-05, "loss": 0.0153, "step": 43060 }, { "grad_norm": 0.23385944962501526, "learning_rate": 6.534230410617092e-05, "loss": 0.0154, "step": 43070 }, { "grad_norm": 0.16987988352775574, "learning_rate": 6.532656621725007e-05, "loss": 0.0187, "step": 43080 }, { "grad_norm": 0.1778360903263092, "learning_rate": 6.531082665223898e-05, "loss": 0.0167, "step": 43090 }, { "grad_norm": 0.27369025349617004, "learning_rate": 6.529508541285889e-05, "loss": 0.0157, "step": 43100 }, { "grad_norm": 0.22124888002872467, "learning_rate": 6.527934250083128e-05, "loss": 0.0154, "step": 43110 }, { "grad_norm": 0.1834101527929306, "learning_rate": 6.526359791787775e-05, "loss": 0.017, "step": 43120 }, { "grad_norm": 0.20031431317329407, "learning_rate": 6.524785166572009e-05, "loss": 0.0153, "step": 43130 }, { "grad_norm": 0.16471512615680695, "learning_rate": 6.523210374608029e-05, "loss": 0.0138, "step": 43140 }, { "grad_norm": 0.17941056191921234, "learning_rate": 6.521635416068054e-05, "loss": 0.0204, "step": 43150 }, { "grad_norm": 0.2106553018093109, "learning_rate": 6.520060291124317e-05, "loss": 0.0166, "step": 43160 }, { "grad_norm": 0.1560002863407135, "learning_rate": 6.518484999949072e-05, "loss": 0.0147, "step": 43170 }, { "grad_norm": 0.18298973143100739, "learning_rate": 6.516909542714591e-05, "loss": 0.0201, "step": 43180 }, { "grad_norm": 0.17460189759731293, "learning_rate": 6.515333919593161e-05, "loss": 0.0193, "step": 43190 }, { "grad_norm": 0.18365563452243805, "learning_rate": 6.513758130757094e-05, "loss": 0.0164, "step": 43200 }, { "grad_norm": 0.14626683294773102, "learning_rate": 6.512182176378713e-05, "loss": 0.0164, "step": 43210 }, { "grad_norm": 0.204971581697464, "learning_rate": 6.510606056630362e-05, "loss": 0.0198, "step": 43220 }, { "grad_norm": 0.17933781445026398, "learning_rate": 6.509029771684405e-05, "loss": 0.018, "step": 43230 }, { "grad_norm": 0.14922946691513062, "learning_rate": 6.50745332171322e-05, "loss": 0.0187, "step": 43240 }, { "grad_norm": 0.20224836468696594, "learning_rate": 6.505876706889207e-05, "loss": 0.0159, "step": 43250 }, { "grad_norm": 0.2411995679140091, "learning_rate": 6.504299927384781e-05, "loss": 0.0168, "step": 43260 }, { "grad_norm": 0.22467055916786194, "learning_rate": 6.502722983372378e-05, "loss": 0.0157, "step": 43270 }, { "grad_norm": 0.17725756764411926, "learning_rate": 6.501145875024446e-05, "loss": 0.0137, "step": 43280 }, { "grad_norm": 0.19021232426166534, "learning_rate": 6.499568602513459e-05, "loss": 0.0139, "step": 43290 }, { "grad_norm": 0.17229825258255005, "learning_rate": 6.497991166011903e-05, "loss": 0.0113, "step": 43300 }, { "grad_norm": 0.12910908460617065, "learning_rate": 6.49641356569229e-05, "loss": 0.0119, "step": 43310 }, { "grad_norm": 0.20567987859249115, "learning_rate": 6.494835801727135e-05, "loss": 0.0184, "step": 43320 }, { "grad_norm": 0.14318890869617462, "learning_rate": 6.493257874288986e-05, "loss": 0.0146, "step": 43330 }, { "grad_norm": 0.17504645884037018, "learning_rate": 6.491679783550399e-05, "loss": 0.0165, "step": 43340 }, { "grad_norm": 0.21572358906269073, "learning_rate": 6.490101529683957e-05, "loss": 0.0187, "step": 43350 }, { "grad_norm": 0.14793328940868378, "learning_rate": 6.488523112862248e-05, "loss": 0.0124, "step": 43360 }, { "grad_norm": 0.19544415175914764, "learning_rate": 6.486944533257891e-05, "loss": 0.015, "step": 43370 }, { "grad_norm": 0.17196257412433624, "learning_rate": 6.485365791043517e-05, "loss": 0.0169, "step": 43380 }, { "grad_norm": 0.20047588646411896, "learning_rate": 6.483786886391772e-05, "loss": 0.0177, "step": 43390 }, { "grad_norm": 0.23088490962982178, "learning_rate": 6.482207819475323e-05, "loss": 0.0188, "step": 43400 }, { "grad_norm": 0.2148372232913971, "learning_rate": 6.480628590466857e-05, "loss": 0.0192, "step": 43410 }, { "grad_norm": 0.20173032581806183, "learning_rate": 6.479049199539074e-05, "loss": 0.0132, "step": 43420 }, { "grad_norm": 0.23350025713443756, "learning_rate": 6.477469646864693e-05, "loss": 0.0207, "step": 43430 }, { "grad_norm": 0.1419946700334549, "learning_rate": 6.475889932616454e-05, "loss": 0.0205, "step": 43440 }, { "grad_norm": 0.1484750509262085, "learning_rate": 6.474310056967111e-05, "loss": 0.0164, "step": 43450 }, { "grad_norm": 0.17575064301490784, "learning_rate": 6.472730020089437e-05, "loss": 0.0206, "step": 43460 }, { "grad_norm": 0.15398526191711426, "learning_rate": 6.471149822156222e-05, "loss": 0.0155, "step": 43470 }, { "grad_norm": 0.15272927284240723, "learning_rate": 6.469569463340275e-05, "loss": 0.0132, "step": 43480 }, { "grad_norm": 0.14867015182971954, "learning_rate": 6.467988943814418e-05, "loss": 0.0171, "step": 43490 }, { "grad_norm": 0.15244680643081665, "learning_rate": 6.4664082637515e-05, "loss": 0.0143, "step": 43500 }, { "grad_norm": 0.15996693074703217, "learning_rate": 6.464827423324377e-05, "loss": 0.0154, "step": 43510 }, { "grad_norm": 0.19384856522083282, "learning_rate": 6.463246422705931e-05, "loss": 0.0168, "step": 43520 }, { "grad_norm": 0.1569308191537857, "learning_rate": 6.461665262069055e-05, "loss": 0.0166, "step": 43530 }, { "grad_norm": 0.19778555631637573, "learning_rate": 6.460083941586665e-05, "loss": 0.0198, "step": 43540 }, { "grad_norm": 0.15356425940990448, "learning_rate": 6.45850246143169e-05, "loss": 0.0131, "step": 43550 }, { "grad_norm": 0.257226824760437, "learning_rate": 6.456920821777077e-05, "loss": 0.0176, "step": 43560 }, { "grad_norm": 0.21005339920520782, "learning_rate": 6.455339022795796e-05, "loss": 0.0167, "step": 43570 }, { "grad_norm": 0.2610490918159485, "learning_rate": 6.453757064660825e-05, "loss": 0.0147, "step": 43580 }, { "grad_norm": 0.17932839691638947, "learning_rate": 6.452174947545169e-05, "loss": 0.0134, "step": 43590 }, { "grad_norm": 0.21081550419330597, "learning_rate": 6.450592671621842e-05, "loss": 0.0142, "step": 43600 }, { "grad_norm": 0.2137763798236847, "learning_rate": 6.449010237063882e-05, "loss": 0.0165, "step": 43610 }, { "grad_norm": 0.3084251880645752, "learning_rate": 6.447427644044343e-05, "loss": 0.0153, "step": 43620 }, { "grad_norm": 0.18407008051872253, "learning_rate": 6.44584489273629e-05, "loss": 0.0174, "step": 43630 }, { "grad_norm": 0.1781773716211319, "learning_rate": 6.444261983312817e-05, "loss": 0.0135, "step": 43640 }, { "grad_norm": 0.14888881146907806, "learning_rate": 6.442678915947023e-05, "loss": 0.0159, "step": 43650 }, { "grad_norm": 0.19891661405563354, "learning_rate": 6.441095690812032e-05, "loss": 0.0135, "step": 43660 }, { "grad_norm": 0.1483849138021469, "learning_rate": 6.439512308080982e-05, "loss": 0.0171, "step": 43670 }, { "grad_norm": 0.17449472844600677, "learning_rate": 6.43792876792703e-05, "loss": 0.0152, "step": 43680 }, { "grad_norm": 0.18174029886722565, "learning_rate": 6.43634507052335e-05, "loss": 0.0174, "step": 43690 }, { "grad_norm": 0.18052466213703156, "learning_rate": 6.434761216043135e-05, "loss": 0.0172, "step": 43700 }, { "grad_norm": 0.13597509264945984, "learning_rate": 6.433177204659589e-05, "loss": 0.0159, "step": 43710 }, { "grad_norm": 0.15080620348453522, "learning_rate": 6.431593036545938e-05, "loss": 0.0144, "step": 43720 }, { "grad_norm": 0.16657361388206482, "learning_rate": 6.430008711875426e-05, "loss": 0.0168, "step": 43730 }, { "grad_norm": 0.1852494329214096, "learning_rate": 6.42842423082131e-05, "loss": 0.0181, "step": 43740 }, { "grad_norm": 0.23354792594909668, "learning_rate": 6.426839593556868e-05, "loss": 0.0158, "step": 43750 }, { "grad_norm": 0.25858205556869507, "learning_rate": 6.425254800255392e-05, "loss": 0.0127, "step": 43760 }, { "grad_norm": 0.2082333266735077, "learning_rate": 6.423669851090196e-05, "loss": 0.0138, "step": 43770 }, { "grad_norm": 0.13862766325473785, "learning_rate": 6.422084746234604e-05, "loss": 0.0173, "step": 43780 }, { "grad_norm": 0.19378064572811127, "learning_rate": 6.420499485861965e-05, "loss": 0.0153, "step": 43790 }, { "grad_norm": 0.188792422413826, "learning_rate": 6.418914070145636e-05, "loss": 0.0163, "step": 43800 }, { "grad_norm": 0.1522528976202011, "learning_rate": 6.417328499258996e-05, "loss": 0.0122, "step": 43810 }, { "grad_norm": 0.20736588537693024, "learning_rate": 6.415742773375446e-05, "loss": 0.0152, "step": 43820 }, { "grad_norm": 0.19620637595653534, "learning_rate": 6.414156892668393e-05, "loss": 0.0139, "step": 43830 }, { "grad_norm": 0.1567593216896057, "learning_rate": 6.412570857311267e-05, "loss": 0.0156, "step": 43840 }, { "grad_norm": 0.23637965321540833, "learning_rate": 6.410984667477518e-05, "loss": 0.0144, "step": 43850 }, { "grad_norm": 0.15841332077980042, "learning_rate": 6.409398323340607e-05, "loss": 0.0158, "step": 43860 }, { "grad_norm": 0.19190649688243866, "learning_rate": 6.407811825074012e-05, "loss": 0.0168, "step": 43870 }, { "grad_norm": 0.2099636048078537, "learning_rate": 6.406225172851234e-05, "loss": 0.016, "step": 43880 }, { "grad_norm": 0.1694609522819519, "learning_rate": 6.404638366845786e-05, "loss": 0.0145, "step": 43890 }, { "grad_norm": 0.1777229756116867, "learning_rate": 6.403051407231196e-05, "loss": 0.013, "step": 43900 }, { "grad_norm": 0.22324365377426147, "learning_rate": 6.401464294181016e-05, "loss": 0.0129, "step": 43910 }, { "grad_norm": 0.11215650290250778, "learning_rate": 6.399877027868808e-05, "loss": 0.0161, "step": 43920 }, { "grad_norm": 0.2534016966819763, "learning_rate": 6.39828960846815e-05, "loss": 0.0184, "step": 43930 }, { "grad_norm": 0.18957769870758057, "learning_rate": 6.396702036152644e-05, "loss": 0.0157, "step": 43940 }, { "grad_norm": 0.20891022682189941, "learning_rate": 6.395114311095905e-05, "loss": 0.0181, "step": 43950 }, { "grad_norm": 0.23714463412761688, "learning_rate": 6.393526433471559e-05, "loss": 0.0172, "step": 43960 }, { "grad_norm": 0.20172151923179626, "learning_rate": 6.391938403453262e-05, "loss": 0.0147, "step": 43970 }, { "grad_norm": 0.22411580383777618, "learning_rate": 6.390350221214671e-05, "loss": 0.0172, "step": 43980 }, { "grad_norm": 0.1460818201303482, "learning_rate": 6.38876188692947e-05, "loss": 0.0121, "step": 43990 }, { "grad_norm": 0.2544213533401489, "learning_rate": 6.387173400771358e-05, "loss": 0.0141, "step": 44000 }, { "grad_norm": 0.2073182612657547, "learning_rate": 6.385584762914047e-05, "loss": 0.0163, "step": 44010 }, { "grad_norm": 0.16170938313007355, "learning_rate": 6.38399597353127e-05, "loss": 0.0177, "step": 44020 }, { "grad_norm": 0.2104233056306839, "learning_rate": 6.382407032796775e-05, "loss": 0.0203, "step": 44030 }, { "grad_norm": 0.21966801583766937, "learning_rate": 6.380817940884325e-05, "loss": 0.0168, "step": 44040 }, { "grad_norm": 0.2235807180404663, "learning_rate": 6.379228697967702e-05, "loss": 0.0165, "step": 44050 }, { "grad_norm": 0.19170494377613068, "learning_rate": 6.3776393042207e-05, "loss": 0.0163, "step": 44060 }, { "grad_norm": 0.18836988508701324, "learning_rate": 6.376049759817137e-05, "loss": 0.016, "step": 44070 }, { "grad_norm": 0.23797252774238586, "learning_rate": 6.37446006493084e-05, "loss": 0.0163, "step": 44080 }, { "grad_norm": 0.18645784258842468, "learning_rate": 6.372870219735656e-05, "loss": 0.0137, "step": 44090 }, { "grad_norm": 0.20059911906719208, "learning_rate": 6.371280224405452e-05, "loss": 0.0144, "step": 44100 }, { "grad_norm": 0.14268100261688232, "learning_rate": 6.369690079114101e-05, "loss": 0.0128, "step": 44110 }, { "grad_norm": 0.17429806292057037, "learning_rate": 6.368099784035504e-05, "loss": 0.0138, "step": 44120 }, { "grad_norm": 0.21090167760849, "learning_rate": 6.366509339343572e-05, "loss": 0.0177, "step": 44130 }, { "grad_norm": 0.25662961602211, "learning_rate": 6.364918745212232e-05, "loss": 0.0127, "step": 44140 }, { "grad_norm": 0.15368731319904327, "learning_rate": 6.363328001815431e-05, "loss": 0.0149, "step": 44150 }, { "grad_norm": 0.24383488297462463, "learning_rate": 6.361737109327128e-05, "loss": 0.0141, "step": 44160 }, { "grad_norm": 0.20241616666316986, "learning_rate": 6.360146067921301e-05, "loss": 0.0146, "step": 44170 }, { "grad_norm": 0.17842935025691986, "learning_rate": 6.35855487777195e-05, "loss": 0.0158, "step": 44180 }, { "grad_norm": 0.2266308069229126, "learning_rate": 6.356963539053076e-05, "loss": 0.0171, "step": 44190 }, { "grad_norm": 0.2535260021686554, "learning_rate": 6.355372051938711e-05, "loss": 0.017, "step": 44200 }, { "grad_norm": 0.16731806099414825, "learning_rate": 6.353780416602894e-05, "loss": 0.0129, "step": 44210 }, { "grad_norm": 0.1645086258649826, "learning_rate": 6.352188633219689e-05, "loss": 0.0162, "step": 44220 }, { "grad_norm": 0.1577925682067871, "learning_rate": 6.350596701963166e-05, "loss": 0.0148, "step": 44230 }, { "grad_norm": 0.1606854796409607, "learning_rate": 6.349004623007419e-05, "loss": 0.0158, "step": 44240 }, { "grad_norm": 0.1588023602962494, "learning_rate": 6.347412396526555e-05, "loss": 0.0139, "step": 44250 }, { "grad_norm": 0.19622689485549927, "learning_rate": 6.345820022694696e-05, "loss": 0.0218, "step": 44260 }, { "grad_norm": 0.21828152239322662, "learning_rate": 6.344227501685984e-05, "loss": 0.0148, "step": 44270 }, { "grad_norm": 0.2288447916507721, "learning_rate": 6.342634833674572e-05, "loss": 0.0152, "step": 44280 }, { "grad_norm": 0.1620863825082779, "learning_rate": 6.341042018834635e-05, "loss": 0.015, "step": 44290 }, { "grad_norm": 0.22913657128810883, "learning_rate": 6.339449057340359e-05, "loss": 0.0127, "step": 44300 }, { "grad_norm": 0.21322590112686157, "learning_rate": 6.337855949365945e-05, "loss": 0.017, "step": 44310 }, { "grad_norm": 0.17398694157600403, "learning_rate": 6.336262695085619e-05, "loss": 0.0132, "step": 44320 }, { "grad_norm": 0.1711401641368866, "learning_rate": 6.334669294673612e-05, "loss": 0.0151, "step": 44330 }, { "grad_norm": 0.1957884281873703, "learning_rate": 6.333075748304179e-05, "loss": 0.0152, "step": 44340 }, { "grad_norm": 0.2590726613998413, "learning_rate": 6.331482056151585e-05, "loss": 0.0136, "step": 44350 }, { "grad_norm": 0.20766988396644592, "learning_rate": 6.329888218390117e-05, "loss": 0.0195, "step": 44360 }, { "grad_norm": 0.1500924676656723, "learning_rate": 6.328294235194072e-05, "loss": 0.0164, "step": 44370 }, { "grad_norm": 0.20771776139736176, "learning_rate": 6.326700106737765e-05, "loss": 0.0145, "step": 44380 }, { "grad_norm": 0.21060964465141296, "learning_rate": 6.325105833195531e-05, "loss": 0.0154, "step": 44390 }, { "grad_norm": 0.16950629651546478, "learning_rate": 6.323511414741715e-05, "loss": 0.0144, "step": 44400 }, { "grad_norm": 0.1546814739704132, "learning_rate": 6.321916851550678e-05, "loss": 0.015, "step": 44410 }, { "grad_norm": 0.23041662573814392, "learning_rate": 6.320322143796806e-05, "loss": 0.0144, "step": 44420 }, { "grad_norm": 0.22193564474582672, "learning_rate": 6.318727291654488e-05, "loss": 0.0171, "step": 44430 }, { "grad_norm": 0.24591578543186188, "learning_rate": 6.317132295298134e-05, "loss": 0.0171, "step": 44440 }, { "grad_norm": 0.176434725522995, "learning_rate": 6.315537154902173e-05, "loss": 0.0136, "step": 44450 }, { "grad_norm": 0.14555634558200836, "learning_rate": 6.313941870641048e-05, "loss": 0.0144, "step": 44460 }, { "grad_norm": 0.1742822527885437, "learning_rate": 6.312346442689214e-05, "loss": 0.0143, "step": 44470 }, { "grad_norm": 0.2749587297439575, "learning_rate": 6.310750871221147e-05, "loss": 0.0211, "step": 44480 }, { "grad_norm": 0.20421117544174194, "learning_rate": 6.309155156411335e-05, "loss": 0.0163, "step": 44490 }, { "grad_norm": 0.17127230763435364, "learning_rate": 6.307559298434284e-05, "loss": 0.0251, "step": 44500 }, { "grad_norm": 0.1779218167066574, "learning_rate": 6.305963297464512e-05, "loss": 0.0206, "step": 44510 }, { "grad_norm": 0.20283068716526031, "learning_rate": 6.304367153676561e-05, "loss": 0.0136, "step": 44520 }, { "grad_norm": 0.1916365623474121, "learning_rate": 6.302770867244978e-05, "loss": 0.0145, "step": 44530 }, { "grad_norm": 0.1964436173439026, "learning_rate": 6.301174438344328e-05, "loss": 0.013, "step": 44540 }, { "grad_norm": 0.20441479980945587, "learning_rate": 6.299577867149202e-05, "loss": 0.0135, "step": 44550 }, { "grad_norm": 0.17213128507137299, "learning_rate": 6.297981153834192e-05, "loss": 0.0157, "step": 44560 }, { "grad_norm": 0.3346583843231201, "learning_rate": 6.296384298573916e-05, "loss": 0.0181, "step": 44570 }, { "grad_norm": 0.14492779970169067, "learning_rate": 6.294787301543001e-05, "loss": 0.0122, "step": 44580 }, { "grad_norm": 0.2153502106666565, "learning_rate": 6.293190162916095e-05, "loss": 0.0153, "step": 44590 }, { "grad_norm": 0.16746006906032562, "learning_rate": 6.291592882867855e-05, "loss": 0.013, "step": 44600 }, { "grad_norm": 0.2901468276977539, "learning_rate": 6.28999546157296e-05, "loss": 0.019, "step": 44610 }, { "grad_norm": 0.2099168747663498, "learning_rate": 6.288397899206102e-05, "loss": 0.0158, "step": 44620 }, { "grad_norm": 0.17998124659061432, "learning_rate": 6.286800195941984e-05, "loss": 0.0135, "step": 44630 }, { "grad_norm": 0.200722336769104, "learning_rate": 6.285202351955334e-05, "loss": 0.0141, "step": 44640 }, { "grad_norm": 0.16506515443325043, "learning_rate": 6.283604367420887e-05, "loss": 0.0181, "step": 44650 }, { "grad_norm": 0.19457969069480896, "learning_rate": 6.282006242513394e-05, "loss": 0.0204, "step": 44660 }, { "grad_norm": 0.24072107672691345, "learning_rate": 6.280407977407628e-05, "loss": 0.0158, "step": 44670 }, { "grad_norm": 0.1849573254585266, "learning_rate": 6.27880957227837e-05, "loss": 0.0157, "step": 44680 }, { "grad_norm": 0.2142973393201828, "learning_rate": 6.27721102730042e-05, "loss": 0.0142, "step": 44690 }, { "grad_norm": 0.18118031322956085, "learning_rate": 6.275612342648592e-05, "loss": 0.0176, "step": 44700 }, { "grad_norm": 0.13437440991401672, "learning_rate": 6.274013518497716e-05, "loss": 0.0132, "step": 44710 }, { "grad_norm": 0.13315753638744354, "learning_rate": 6.272414555022636e-05, "loss": 0.0176, "step": 44720 }, { "grad_norm": 0.19640058279037476, "learning_rate": 6.270815452398215e-05, "loss": 0.0136, "step": 44730 }, { "grad_norm": 0.19390851259231567, "learning_rate": 6.269216210799326e-05, "loss": 0.0148, "step": 44740 }, { "grad_norm": 0.18576806783676147, "learning_rate": 6.26761683040086e-05, "loss": 0.0154, "step": 44750 }, { "grad_norm": 0.12920339405536652, "learning_rate": 6.266017311377723e-05, "loss": 0.0134, "step": 44760 }, { "grad_norm": 0.18482939898967743, "learning_rate": 6.264417653904839e-05, "loss": 0.0173, "step": 44770 }, { "grad_norm": 0.25610703229904175, "learning_rate": 6.262817858157139e-05, "loss": 0.0171, "step": 44780 }, { "grad_norm": 0.2142440676689148, "learning_rate": 6.261217924309576e-05, "loss": 0.017, "step": 44790 }, { "grad_norm": 0.14748215675354004, "learning_rate": 6.259617852537118e-05, "loss": 0.018, "step": 44800 }, { "grad_norm": 0.1712256669998169, "learning_rate": 6.258017643014747e-05, "loss": 0.0159, "step": 44810 }, { "grad_norm": 0.19336944818496704, "learning_rate": 6.256417295917456e-05, "loss": 0.0147, "step": 44820 }, { "grad_norm": 0.18455295264720917, "learning_rate": 6.254816811420258e-05, "loss": 0.0216, "step": 44830 }, { "grad_norm": 0.20955875515937805, "learning_rate": 6.253216189698183e-05, "loss": 0.0167, "step": 44840 }, { "grad_norm": 0.17104367911815643, "learning_rate": 6.251615430926267e-05, "loss": 0.0161, "step": 44850 }, { "grad_norm": 0.1904229372739792, "learning_rate": 6.25001453527957e-05, "loss": 0.0184, "step": 44860 }, { "grad_norm": 0.16758114099502563, "learning_rate": 6.248413502933164e-05, "loss": 0.0176, "step": 44870 }, { "grad_norm": 0.1625310480594635, "learning_rate": 6.246812334062133e-05, "loss": 0.0144, "step": 44880 }, { "grad_norm": 0.20753653347492218, "learning_rate": 6.245211028841579e-05, "loss": 0.0153, "step": 44890 }, { "grad_norm": 0.15755301713943481, "learning_rate": 6.24360958744662e-05, "loss": 0.013, "step": 44900 }, { "grad_norm": 0.19258996844291687, "learning_rate": 6.242008010052387e-05, "loss": 0.0129, "step": 44910 }, { "grad_norm": 0.19382014870643616, "learning_rate": 6.240406296834024e-05, "loss": 0.0148, "step": 44920 }, { "grad_norm": 0.2442871481180191, "learning_rate": 6.238804447966694e-05, "loss": 0.0162, "step": 44930 }, { "grad_norm": 0.18020597100257874, "learning_rate": 6.237202463625573e-05, "loss": 0.0165, "step": 44940 }, { "grad_norm": 0.19254639744758606, "learning_rate": 6.235600343985848e-05, "loss": 0.0164, "step": 44950 }, { "grad_norm": 0.16623546183109283, "learning_rate": 6.233998089222729e-05, "loss": 0.0146, "step": 44960 }, { "grad_norm": 0.15286706387996674, "learning_rate": 6.232395699511433e-05, "loss": 0.0134, "step": 44970 }, { "grad_norm": 0.1310833990573883, "learning_rate": 6.230793175027199e-05, "loss": 0.0136, "step": 44980 }, { "grad_norm": 0.18570701777935028, "learning_rate": 6.22919051594527e-05, "loss": 0.0181, "step": 44990 }, { "grad_norm": 0.1661449372768402, "learning_rate": 6.227587722440917e-05, "loss": 0.0109, "step": 45000 }, { "grad_norm": 0.16355623304843903, "learning_rate": 6.225984794689414e-05, "loss": 0.0169, "step": 45010 }, { "grad_norm": 0.201935276389122, "learning_rate": 6.22438173286606e-05, "loss": 0.0125, "step": 45020 }, { "grad_norm": 0.24892497062683105, "learning_rate": 6.222778537146159e-05, "loss": 0.0174, "step": 45030 }, { "grad_norm": 0.14788566529750824, "learning_rate": 6.221175207705037e-05, "loss": 0.0116, "step": 45040 }, { "grad_norm": 0.17215245962142944, "learning_rate": 6.219571744718027e-05, "loss": 0.013, "step": 45050 }, { "grad_norm": 0.19433654844760895, "learning_rate": 6.217968148360487e-05, "loss": 0.0153, "step": 45060 }, { "grad_norm": 0.15707822144031525, "learning_rate": 6.216364418807782e-05, "loss": 0.0179, "step": 45070 }, { "grad_norm": 0.1661730855703354, "learning_rate": 6.214760556235292e-05, "loss": 0.0139, "step": 45080 }, { "grad_norm": 0.20160703361034393, "learning_rate": 6.213156560818414e-05, "loss": 0.0166, "step": 45090 }, { "grad_norm": 0.1715848296880722, "learning_rate": 6.211552432732559e-05, "loss": 0.0144, "step": 45100 }, { "grad_norm": 0.22441406548023224, "learning_rate": 6.20994817215315e-05, "loss": 0.015, "step": 45110 }, { "grad_norm": 0.18290166556835175, "learning_rate": 6.208343779255628e-05, "loss": 0.0164, "step": 45120 }, { "grad_norm": 0.21047000586986542, "learning_rate": 6.206739254215449e-05, "loss": 0.0137, "step": 45130 }, { "grad_norm": 0.18780915439128876, "learning_rate": 6.205134597208077e-05, "loss": 0.0129, "step": 45140 }, { "grad_norm": 0.18547120690345764, "learning_rate": 6.203529808409e-05, "loss": 0.0114, "step": 45150 }, { "grad_norm": 0.14892670512199402, "learning_rate": 6.201924887993712e-05, "loss": 0.0116, "step": 45160 }, { "grad_norm": 0.21053417026996613, "learning_rate": 6.200319836137725e-05, "loss": 0.0128, "step": 45170 }, { "grad_norm": 0.1344286948442459, "learning_rate": 6.198714653016565e-05, "loss": 0.0137, "step": 45180 }, { "grad_norm": 0.18297162652015686, "learning_rate": 6.197109338805774e-05, "loss": 0.0127, "step": 45190 }, { "grad_norm": 0.3045211732387543, "learning_rate": 6.195503893680903e-05, "loss": 0.0168, "step": 45200 }, { "grad_norm": 0.22846706211566925, "learning_rate": 6.193898317817524e-05, "loss": 0.0129, "step": 45210 }, { "grad_norm": 0.14019125699996948, "learning_rate": 6.192292611391221e-05, "loss": 0.0139, "step": 45220 }, { "grad_norm": 0.1704884171485901, "learning_rate": 6.190686774577591e-05, "loss": 0.0138, "step": 45230 }, { "grad_norm": 0.20703452825546265, "learning_rate": 6.189080807552245e-05, "loss": 0.0127, "step": 45240 }, { "grad_norm": 0.1774997115135193, "learning_rate": 6.187474710490809e-05, "loss": 0.0135, "step": 45250 }, { "grad_norm": 0.13655324280261993, "learning_rate": 6.185868483568926e-05, "loss": 0.0157, "step": 45260 }, { "grad_norm": 0.19929783046245575, "learning_rate": 6.184262126962245e-05, "loss": 0.0149, "step": 45270 }, { "grad_norm": 0.19665761291980743, "learning_rate": 6.182655640846442e-05, "loss": 0.0159, "step": 45280 }, { "grad_norm": 0.2560461163520813, "learning_rate": 6.181049025397196e-05, "loss": 0.0205, "step": 45290 }, { "grad_norm": 0.13924582302570343, "learning_rate": 6.179442280790202e-05, "loss": 0.0156, "step": 45300 }, { "grad_norm": 0.13370409607887268, "learning_rate": 6.177835407201174e-05, "loss": 0.018, "step": 45310 }, { "grad_norm": 0.20780572295188904, "learning_rate": 6.176228404805839e-05, "loss": 0.0118, "step": 45320 }, { "grad_norm": 0.19747458398342133, "learning_rate": 6.174621273779932e-05, "loss": 0.0182, "step": 45330 }, { "grad_norm": 0.187955841422081, "learning_rate": 6.17301401429921e-05, "loss": 0.0191, "step": 45340 }, { "grad_norm": 0.3285467028617859, "learning_rate": 6.171406626539436e-05, "loss": 0.0156, "step": 45350 }, { "grad_norm": 0.18451032042503357, "learning_rate": 6.169799110676398e-05, "loss": 0.0182, "step": 45360 }, { "grad_norm": 0.20008568465709686, "learning_rate": 6.168191466885885e-05, "loss": 0.0216, "step": 45370 }, { "grad_norm": 0.2482967972755432, "learning_rate": 6.166583695343711e-05, "loss": 0.0198, "step": 45380 }, { "grad_norm": 0.2440168708562851, "learning_rate": 6.164975796225698e-05, "loss": 0.0196, "step": 45390 }, { "grad_norm": 0.16871508955955505, "learning_rate": 6.163367769707683e-05, "loss": 0.0165, "step": 45400 }, { "grad_norm": 0.1975528746843338, "learning_rate": 6.161759615965519e-05, "loss": 0.0158, "step": 45410 }, { "grad_norm": 0.18313708901405334, "learning_rate": 6.16015133517507e-05, "loss": 0.0185, "step": 45420 }, { "grad_norm": 0.19167335331439972, "learning_rate": 6.158542927512214e-05, "loss": 0.0169, "step": 45430 }, { "grad_norm": 0.1619265079498291, "learning_rate": 6.156934393152846e-05, "loss": 0.0114, "step": 45440 }, { "grad_norm": 0.15534666180610657, "learning_rate": 6.15532573227287e-05, "loss": 0.0132, "step": 45450 }, { "grad_norm": 0.23586224019527435, "learning_rate": 6.153716945048212e-05, "loss": 0.0135, "step": 45460 }, { "grad_norm": 0.19603194296360016, "learning_rate": 6.152108031654802e-05, "loss": 0.0152, "step": 45470 }, { "grad_norm": 0.16342474520206451, "learning_rate": 6.15049899226859e-05, "loss": 0.0172, "step": 45480 }, { "grad_norm": 0.1529616415500641, "learning_rate": 6.148889827065537e-05, "loss": 0.0193, "step": 45490 }, { "grad_norm": 0.24945898354053497, "learning_rate": 6.147280536221622e-05, "loss": 0.0165, "step": 45500 }, { "grad_norm": 0.18766479194164276, "learning_rate": 6.145671119912832e-05, "loss": 0.0134, "step": 45510 }, { "grad_norm": 0.1790674328804016, "learning_rate": 6.144061578315169e-05, "loss": 0.0174, "step": 45520 }, { "grad_norm": 0.2684929668903351, "learning_rate": 6.142451911604654e-05, "loss": 0.021, "step": 45530 }, { "grad_norm": 0.13466030359268188, "learning_rate": 6.140842119957315e-05, "loss": 0.0141, "step": 45540 }, { "grad_norm": 0.17630396783351898, "learning_rate": 6.139232203549197e-05, "loss": 0.0164, "step": 45550 }, { "grad_norm": 0.16890858113765717, "learning_rate": 6.137622162556357e-05, "loss": 0.0161, "step": 45560 }, { "grad_norm": 0.1750210076570511, "learning_rate": 6.13601199715487e-05, "loss": 0.0155, "step": 45570 }, { "grad_norm": 0.1445605307817459, "learning_rate": 6.134401707520816e-05, "loss": 0.0147, "step": 45580 }, { "grad_norm": 0.17520591616630554, "learning_rate": 6.132791293830298e-05, "loss": 0.0135, "step": 45590 }, { "grad_norm": 0.20090675354003906, "learning_rate": 6.131180756259428e-05, "loss": 0.0155, "step": 45600 }, { "grad_norm": 0.23254191875457764, "learning_rate": 6.129570094984331e-05, "loss": 0.0201, "step": 45610 }, { "grad_norm": 0.2008649855852127, "learning_rate": 6.127959310181145e-05, "loss": 0.0183, "step": 45620 }, { "grad_norm": 0.1926192045211792, "learning_rate": 6.126348402026026e-05, "loss": 0.0133, "step": 45630 }, { "grad_norm": 0.18589438498020172, "learning_rate": 6.12473737069514e-05, "loss": 0.0122, "step": 45640 }, { "grad_norm": 0.23866450786590576, "learning_rate": 6.123126216364665e-05, "loss": 0.0152, "step": 45650 }, { "grad_norm": 0.15322333574295044, "learning_rate": 6.121514939210797e-05, "loss": 0.0158, "step": 45660 }, { "grad_norm": 0.15887364745140076, "learning_rate": 6.119903539409741e-05, "loss": 0.0113, "step": 45670 }, { "grad_norm": 0.1856045424938202, "learning_rate": 6.118292017137716e-05, "loss": 0.0141, "step": 45680 }, { "grad_norm": 0.12552708387374878, "learning_rate": 6.116680372570959e-05, "loss": 0.0137, "step": 45690 }, { "grad_norm": 0.1586979329586029, "learning_rate": 6.115068605885713e-05, "loss": 0.0171, "step": 45700 }, { "grad_norm": 0.18473570048809052, "learning_rate": 6.113456717258243e-05, "loss": 0.0134, "step": 45710 }, { "grad_norm": 0.16364434361457825, "learning_rate": 6.11184470686482e-05, "loss": 0.0181, "step": 45720 }, { "grad_norm": 0.18680493533611298, "learning_rate": 6.110232574881731e-05, "loss": 0.013, "step": 45730 }, { "grad_norm": 0.19516107439994812, "learning_rate": 6.108620321485277e-05, "loss": 0.016, "step": 45740 }, { "grad_norm": 0.16370078921318054, "learning_rate": 6.107007946851773e-05, "loss": 0.0138, "step": 45750 }, { "grad_norm": 0.15929396450519562, "learning_rate": 6.105395451157542e-05, "loss": 0.0149, "step": 45760 }, { "grad_norm": 0.14553463459014893, "learning_rate": 6.103782834578928e-05, "loss": 0.0152, "step": 45770 }, { "grad_norm": 0.21504054963588715, "learning_rate": 6.102170097292281e-05, "loss": 0.0156, "step": 45780 }, { "grad_norm": 0.20413093268871307, "learning_rate": 6.1005572394739716e-05, "loss": 0.0156, "step": 45790 }, { "grad_norm": 0.23186644911766052, "learning_rate": 6.0989442613003765e-05, "loss": 0.0184, "step": 45800 }, { "grad_norm": 0.14912384748458862, "learning_rate": 6.097331162947888e-05, "loss": 0.017, "step": 45810 }, { "grad_norm": 0.16160432994365692, "learning_rate": 6.095717944592914e-05, "loss": 0.0147, "step": 45820 }, { "grad_norm": 0.1471090465784073, "learning_rate": 6.094104606411873e-05, "loss": 0.0146, "step": 45830 }, { "grad_norm": 0.15371818840503693, "learning_rate": 6.0924911485811966e-05, "loss": 0.0154, "step": 45840 }, { "grad_norm": 0.12666362524032593, "learning_rate": 6.090877571277331e-05, "loss": 0.0142, "step": 45850 }, { "grad_norm": 0.17726948857307434, "learning_rate": 6.089263874676736e-05, "loss": 0.0157, "step": 45860 }, { "grad_norm": 0.215814009308815, "learning_rate": 6.0876500589558796e-05, "loss": 0.0158, "step": 45870 }, { "grad_norm": 0.2531566917896271, "learning_rate": 6.086036124291248e-05, "loss": 0.0173, "step": 45880 }, { "grad_norm": 0.19110971689224243, "learning_rate": 6.084422070859339e-05, "loss": 0.0158, "step": 45890 }, { "grad_norm": 0.15152521431446075, "learning_rate": 6.082807898836663e-05, "loss": 0.0171, "step": 45900 }, { "grad_norm": 0.3169565200805664, "learning_rate": 6.081193608399742e-05, "loss": 0.017, "step": 45910 }, { "grad_norm": 0.17373065650463104, "learning_rate": 6.0795791997251164e-05, "loss": 0.0157, "step": 45920 }, { "grad_norm": 0.2055959403514862, "learning_rate": 6.0779646729893294e-05, "loss": 0.0166, "step": 45930 }, { "grad_norm": 0.218438521027565, "learning_rate": 6.0763500283689476e-05, "loss": 0.0173, "step": 45940 }, { "grad_norm": 0.20285776257514954, "learning_rate": 6.0747352660405455e-05, "loss": 0.0154, "step": 45950 }, { "grad_norm": 0.21419990062713623, "learning_rate": 6.073120386180709e-05, "loss": 0.0136, "step": 45960 }, { "grad_norm": 0.2479621171951294, "learning_rate": 6.0715053889660425e-05, "loss": 0.0144, "step": 45970 }, { "grad_norm": 0.1879383623600006, "learning_rate": 6.069890274573157e-05, "loss": 0.0119, "step": 45980 }, { "grad_norm": 0.1787673383951187, "learning_rate": 6.068275043178679e-05, "loss": 0.016, "step": 45990 }, { "grad_norm": 0.17581196129322052, "learning_rate": 6.066659694959248e-05, "loss": 0.0143, "step": 46000 }, { "grad_norm": 0.1341915875673294, "learning_rate": 6.0650442300915176e-05, "loss": 0.0134, "step": 46010 }, { "grad_norm": 0.15649154782295227, "learning_rate": 6.063428648752152e-05, "loss": 0.0137, "step": 46020 }, { "grad_norm": 0.4129559397697449, "learning_rate": 6.0618129511178266e-05, "loss": 0.0162, "step": 46030 }, { "grad_norm": 0.20320165157318115, "learning_rate": 6.060197137365234e-05, "loss": 0.014, "step": 46040 }, { "grad_norm": 0.1597813069820404, "learning_rate": 6.058581207671077e-05, "loss": 0.0125, "step": 46050 }, { "grad_norm": 0.11539319157600403, "learning_rate": 6.056965162212072e-05, "loss": 0.0113, "step": 46060 }, { "grad_norm": 0.19797077775001526, "learning_rate": 6.055349001164943e-05, "loss": 0.0159, "step": 46070 }, { "grad_norm": 0.19773712754249573, "learning_rate": 6.0537327247064347e-05, "loss": 0.0151, "step": 46080 }, { "grad_norm": 0.20201334357261658, "learning_rate": 6.0521163330133014e-05, "loss": 0.0133, "step": 46090 }, { "grad_norm": 0.1723761409521103, "learning_rate": 6.050499826262306e-05, "loss": 0.0156, "step": 46100 }, { "grad_norm": 0.1449069380760193, "learning_rate": 6.0488832046302294e-05, "loss": 0.0113, "step": 46110 }, { "grad_norm": 0.20408345758914948, "learning_rate": 6.0472664682938626e-05, "loss": 0.0181, "step": 46120 }, { "grad_norm": 0.24570822715759277, "learning_rate": 6.045649617430009e-05, "loss": 0.016, "step": 46130 }, { "grad_norm": 0.2126595824956894, "learning_rate": 6.0440326522154866e-05, "loss": 0.0132, "step": 46140 }, { "grad_norm": 0.17097052931785583, "learning_rate": 6.0424155728271224e-05, "loss": 0.0129, "step": 46150 }, { "grad_norm": 0.2062021642923355, "learning_rate": 6.040798379441758e-05, "loss": 0.0183, "step": 46160 }, { "grad_norm": 0.1716424971818924, "learning_rate": 6.0391810722362485e-05, "loss": 0.0165, "step": 46170 }, { "grad_norm": 0.2023455947637558, "learning_rate": 6.037563651387458e-05, "loss": 0.0145, "step": 46180 }, { "grad_norm": 0.18393878638744354, "learning_rate": 6.0359461170722666e-05, "loss": 0.0133, "step": 46190 }, { "grad_norm": 0.2063499391078949, "learning_rate": 6.034328469467566e-05, "loss": 0.0147, "step": 46200 }, { "grad_norm": 0.17827069759368896, "learning_rate": 6.0327107087502596e-05, "loss": 0.013, "step": 46210 }, { "grad_norm": 0.1696964055299759, "learning_rate": 6.031092835097262e-05, "loss": 0.0141, "step": 46220 }, { "grad_norm": 0.20269900560379028, "learning_rate": 6.0294748486855024e-05, "loss": 0.0123, "step": 46230 }, { "grad_norm": 0.15258491039276123, "learning_rate": 6.0278567496919216e-05, "loss": 0.0132, "step": 46240 }, { "grad_norm": 0.18141910433769226, "learning_rate": 6.026238538293472e-05, "loss": 0.0146, "step": 46250 }, { "grad_norm": 0.17336013913154602, "learning_rate": 6.024620214667118e-05, "loss": 0.0124, "step": 46260 }, { "grad_norm": 0.26788875460624695, "learning_rate": 6.0230017789898384e-05, "loss": 0.0149, "step": 46270 }, { "grad_norm": 0.14518806338310242, "learning_rate": 6.021383231438622e-05, "loss": 0.0132, "step": 46280 }, { "grad_norm": 0.1538056880235672, "learning_rate": 6.0197645721904704e-05, "loss": 0.0121, "step": 46290 }, { "grad_norm": 0.2537967264652252, "learning_rate": 6.0181458014224e-05, "loss": 0.0182, "step": 46300 }, { "grad_norm": 0.163704514503479, "learning_rate": 6.016526919311435e-05, "loss": 0.0133, "step": 46310 }, { "grad_norm": 0.197994202375412, "learning_rate": 6.014907926034613e-05, "loss": 0.0149, "step": 46320 }, { "grad_norm": 0.17862355709075928, "learning_rate": 6.0132888217689875e-05, "loss": 0.0148, "step": 46330 }, { "grad_norm": 0.19130676984786987, "learning_rate": 6.01166960669162e-05, "loss": 0.0109, "step": 46340 }, { "grad_norm": 0.189164936542511, "learning_rate": 6.0100502809795845e-05, "loss": 0.0147, "step": 46350 }, { "grad_norm": 0.17162813246250153, "learning_rate": 6.00843084480997e-05, "loss": 0.0131, "step": 46360 }, { "grad_norm": 0.11613672971725464, "learning_rate": 6.0068112983598736e-05, "loss": 0.0154, "step": 46370 }, { "grad_norm": 0.16716277599334717, "learning_rate": 6.005191641806407e-05, "loss": 0.0155, "step": 46380 }, { "grad_norm": 0.22751951217651367, "learning_rate": 6.003571875326694e-05, "loss": 0.0181, "step": 46390 }, { "grad_norm": 0.3069222867488861, "learning_rate": 6.00195199909787e-05, "loss": 0.0197, "step": 46400 }, { "grad_norm": 0.19911539554595947, "learning_rate": 6.000332013297082e-05, "loss": 0.0131, "step": 46410 }, { "grad_norm": 0.2159396857023239, "learning_rate": 5.998711918101487e-05, "loss": 0.015, "step": 46420 }, { "grad_norm": 0.12843573093414307, "learning_rate": 5.997091713688261e-05, "loss": 0.0157, "step": 46430 }, { "grad_norm": 0.25626567006111145, "learning_rate": 5.9954714002345836e-05, "loss": 0.011, "step": 46440 }, { "grad_norm": 0.18675436079502106, "learning_rate": 5.993850977917649e-05, "loss": 0.0161, "step": 46450 }, { "grad_norm": 0.16906794905662537, "learning_rate": 5.992230446914667e-05, "loss": 0.0144, "step": 46460 }, { "grad_norm": 0.28446975350379944, "learning_rate": 5.990609807402855e-05, "loss": 0.0146, "step": 46470 }, { "grad_norm": 0.1748630553483963, "learning_rate": 5.988989059559443e-05, "loss": 0.0119, "step": 46480 }, { "grad_norm": 0.2145053595304489, "learning_rate": 5.987368203561675e-05, "loss": 0.0138, "step": 46490 }, { "grad_norm": 0.1802876889705658, "learning_rate": 5.9857472395868055e-05, "loss": 0.0141, "step": 46500 }, { "grad_norm": 0.12419282644987106, "learning_rate": 5.9841261678120983e-05, "loss": 0.0132, "step": 46510 }, { "grad_norm": 0.22529488801956177, "learning_rate": 5.982504988414834e-05, "loss": 0.0148, "step": 46520 }, { "grad_norm": 0.16787981986999512, "learning_rate": 5.9808837015723015e-05, "loss": 0.0131, "step": 46530 }, { "grad_norm": 0.14957331120967865, "learning_rate": 5.9792623074618016e-05, "loss": 0.0157, "step": 46540 }, { "grad_norm": 0.21171744167804718, "learning_rate": 5.977640806260648e-05, "loss": 0.0131, "step": 46550 }, { "grad_norm": 0.25968533754348755, "learning_rate": 5.976019198146166e-05, "loss": 0.0173, "step": 46560 }, { "grad_norm": 0.15999606251716614, "learning_rate": 5.974397483295692e-05, "loss": 0.0161, "step": 46570 }, { "grad_norm": 0.26222893595695496, "learning_rate": 5.972775661886575e-05, "loss": 0.0158, "step": 46580 }, { "grad_norm": 0.1657325178384781, "learning_rate": 5.971153734096173e-05, "loss": 0.0133, "step": 46590 }, { "grad_norm": 0.13211238384246826, "learning_rate": 5.969531700101859e-05, "loss": 0.0129, "step": 46600 }, { "grad_norm": 0.1827368438243866, "learning_rate": 5.9679095600810155e-05, "loss": 0.0123, "step": 46610 }, { "grad_norm": 0.2280651330947876, "learning_rate": 5.9662873142110384e-05, "loss": 0.0162, "step": 46620 }, { "grad_norm": 0.1566794514656067, "learning_rate": 5.964664962669333e-05, "loss": 0.014, "step": 46630 }, { "grad_norm": 0.17957261204719543, "learning_rate": 5.9630425056333186e-05, "loss": 0.0162, "step": 46640 }, { "grad_norm": 0.19000710546970367, "learning_rate": 5.961419943280422e-05, "loss": 0.0142, "step": 46650 }, { "grad_norm": 0.23747481405735016, "learning_rate": 5.959797275788087e-05, "loss": 0.0154, "step": 46660 }, { "grad_norm": 0.1855771690607071, "learning_rate": 5.958174503333765e-05, "loss": 0.0141, "step": 46670 }, { "grad_norm": 0.2279796153306961, "learning_rate": 5.9565516260949195e-05, "loss": 0.0151, "step": 46680 }, { "grad_norm": 0.23804514110088348, "learning_rate": 5.9549286442490273e-05, "loss": 0.0163, "step": 46690 }, { "grad_norm": 0.16248318552970886, "learning_rate": 5.953305557973572e-05, "loss": 0.0164, "step": 46700 }, { "grad_norm": 0.19575975835323334, "learning_rate": 5.9516823674460564e-05, "loss": 0.018, "step": 46710 }, { "grad_norm": 0.178264781832695, "learning_rate": 5.9500590728439875e-05, "loss": 0.0158, "step": 46720 }, { "grad_norm": 0.25232362747192383, "learning_rate": 5.948435674344886e-05, "loss": 0.0161, "step": 46730 }, { "grad_norm": 0.16724511981010437, "learning_rate": 5.946812172126285e-05, "loss": 0.0149, "step": 46740 }, { "grad_norm": 0.18376772105693817, "learning_rate": 5.945188566365729e-05, "loss": 0.0178, "step": 46750 }, { "grad_norm": 0.1442110389471054, "learning_rate": 5.943564857240773e-05, "loss": 0.0151, "step": 46760 }, { "grad_norm": 0.1833178550004959, "learning_rate": 5.941941044928981e-05, "loss": 0.0158, "step": 46770 }, { "grad_norm": 0.1911703646183014, "learning_rate": 5.940317129607935e-05, "loss": 0.0148, "step": 46780 }, { "grad_norm": 0.1590650975704193, "learning_rate": 5.9386931114552204e-05, "loss": 0.0125, "step": 46790 }, { "grad_norm": 0.23556819558143616, "learning_rate": 5.937068990648438e-05, "loss": 0.0158, "step": 46800 }, { "grad_norm": 0.26857274770736694, "learning_rate": 5.935444767365199e-05, "loss": 0.0161, "step": 46810 }, { "grad_norm": 0.19492784142494202, "learning_rate": 5.933820441783129e-05, "loss": 0.0152, "step": 46820 }, { "grad_norm": 0.14422889053821564, "learning_rate": 5.932196014079857e-05, "loss": 0.0169, "step": 46830 }, { "grad_norm": 0.2921183109283447, "learning_rate": 5.930571484433032e-05, "loss": 0.0195, "step": 46840 }, { "grad_norm": 0.16448533535003662, "learning_rate": 5.928946853020309e-05, "loss": 0.0185, "step": 46850 }, { "grad_norm": 0.21868155896663666, "learning_rate": 5.927322120019352e-05, "loss": 0.017, "step": 46860 }, { "grad_norm": 0.14260606467723846, "learning_rate": 5.925697285607844e-05, "loss": 0.0131, "step": 46870 }, { "grad_norm": 0.17785413563251495, "learning_rate": 5.924072349963472e-05, "loss": 0.0127, "step": 46880 }, { "grad_norm": 0.21361711621284485, "learning_rate": 5.922447313263938e-05, "loss": 0.0183, "step": 46890 }, { "grad_norm": 0.20013418793678284, "learning_rate": 5.920822175686952e-05, "loss": 0.0138, "step": 46900 }, { "grad_norm": 0.16423162817955017, "learning_rate": 5.919196937410239e-05, "loss": 0.0173, "step": 46910 }, { "grad_norm": 0.20813021063804626, "learning_rate": 5.917571598611529e-05, "loss": 0.0164, "step": 46920 }, { "grad_norm": 0.2503338158130646, "learning_rate": 5.9159461594685686e-05, "loss": 0.0168, "step": 46930 }, { "grad_norm": 0.1783839762210846, "learning_rate": 5.914320620159114e-05, "loss": 0.0161, "step": 46940 }, { "grad_norm": 0.18175894021987915, "learning_rate": 5.912694980860931e-05, "loss": 0.0137, "step": 46950 }, { "grad_norm": 0.19123584032058716, "learning_rate": 5.9110692417517964e-05, "loss": 0.0134, "step": 46960 }, { "grad_norm": 0.17377915978431702, "learning_rate": 5.9094434030095e-05, "loss": 0.0115, "step": 46970 }, { "grad_norm": 0.15944191813468933, "learning_rate": 5.907817464811841e-05, "loss": 0.0124, "step": 46980 }, { "grad_norm": 0.17998069524765015, "learning_rate": 5.906191427336627e-05, "loss": 0.0161, "step": 46990 }, { "grad_norm": 0.19764266908168793, "learning_rate": 5.9045652907616835e-05, "loss": 0.0172, "step": 47000 }, { "grad_norm": 0.16403448581695557, "learning_rate": 5.902939055264838e-05, "loss": 0.0148, "step": 47010 }, { "grad_norm": 0.24456064403057098, "learning_rate": 5.901312721023935e-05, "loss": 0.0181, "step": 47020 }, { "grad_norm": 0.12470172345638275, "learning_rate": 5.8996862882168294e-05, "loss": 0.013, "step": 47030 }, { "grad_norm": 0.2149621695280075, "learning_rate": 5.8980597570213826e-05, "loss": 0.0163, "step": 47040 }, { "grad_norm": 0.1497802436351776, "learning_rate": 5.896433127615471e-05, "loss": 0.0145, "step": 47050 }, { "grad_norm": 0.18362176418304443, "learning_rate": 5.894806400176981e-05, "loss": 0.0182, "step": 47060 }, { "grad_norm": 0.136003777384758, "learning_rate": 5.893179574883808e-05, "loss": 0.0192, "step": 47070 }, { "grad_norm": 0.1534743756055832, "learning_rate": 5.8915526519138585e-05, "loss": 0.0132, "step": 47080 }, { "grad_norm": 0.16582849621772766, "learning_rate": 5.889925631445053e-05, "loss": 0.0161, "step": 47090 }, { "grad_norm": 0.17303067445755005, "learning_rate": 5.888298513655318e-05, "loss": 0.0146, "step": 47100 }, { "grad_norm": 0.2024381011724472, "learning_rate": 5.886671298722592e-05, "loss": 0.0125, "step": 47110 }, { "grad_norm": 0.16650916635990143, "learning_rate": 5.885043986824828e-05, "loss": 0.0129, "step": 47120 }, { "grad_norm": 0.17762500047683716, "learning_rate": 5.883416578139982e-05, "loss": 0.0153, "step": 47130 }, { "grad_norm": 0.18371622264385223, "learning_rate": 5.881789072846029e-05, "loss": 0.0145, "step": 47140 }, { "grad_norm": 0.13257819414138794, "learning_rate": 5.8801614711209474e-05, "loss": 0.0162, "step": 47150 }, { "grad_norm": 0.17377689480781555, "learning_rate": 5.878533773142732e-05, "loss": 0.0183, "step": 47160 }, { "grad_norm": 0.18705639243125916, "learning_rate": 5.8769059790893834e-05, "loss": 0.0154, "step": 47170 }, { "grad_norm": 0.15777938067913055, "learning_rate": 5.8752780891389145e-05, "loss": 0.0141, "step": 47180 }, { "grad_norm": 0.11543645709753036, "learning_rate": 5.873650103469351e-05, "loss": 0.014, "step": 47190 }, { "grad_norm": 0.1854332685470581, "learning_rate": 5.872022022258725e-05, "loss": 0.0122, "step": 47200 }, { "grad_norm": 0.13291846215724945, "learning_rate": 5.8703938456850814e-05, "loss": 0.0133, "step": 47210 }, { "grad_norm": 0.18112100660800934, "learning_rate": 5.868765573926477e-05, "loss": 0.0136, "step": 47220 }, { "grad_norm": 0.24697323143482208, "learning_rate": 5.867137207160974e-05, "loss": 0.0125, "step": 47230 }, { "grad_norm": 0.17886945605278015, "learning_rate": 5.86550874556665e-05, "loss": 0.0141, "step": 47240 }, { "grad_norm": 0.17621344327926636, "learning_rate": 5.863880189321592e-05, "loss": 0.0153, "step": 47250 }, { "grad_norm": 0.17029890418052673, "learning_rate": 5.862251538603894e-05, "loss": 0.0139, "step": 47260 }, { "grad_norm": 0.21442502737045288, "learning_rate": 5.8606227935916656e-05, "loss": 0.0156, "step": 47270 }, { "grad_norm": 0.1860024780035019, "learning_rate": 5.858993954463021e-05, "loss": 0.013, "step": 47280 }, { "grad_norm": 0.2323920726776123, "learning_rate": 5.8573650213960896e-05, "loss": 0.0137, "step": 47290 }, { "grad_norm": 0.19946397840976715, "learning_rate": 5.8557359945690084e-05, "loss": 0.013, "step": 47300 }, { "grad_norm": 0.2094903588294983, "learning_rate": 5.8541068741599246e-05, "loss": 0.0118, "step": 47310 }, { "grad_norm": 0.17418718338012695, "learning_rate": 5.8524776603469985e-05, "loss": 0.015, "step": 47320 }, { "grad_norm": 0.16717511415481567, "learning_rate": 5.850848353308397e-05, "loss": 0.0138, "step": 47330 }, { "grad_norm": 0.24377278983592987, "learning_rate": 5.849218953222297e-05, "loss": 0.0143, "step": 47340 }, { "grad_norm": 0.15678410232067108, "learning_rate": 5.847589460266891e-05, "loss": 0.0127, "step": 47350 }, { "grad_norm": 0.1244518980383873, "learning_rate": 5.845959874620375e-05, "loss": 0.0108, "step": 47360 }, { "grad_norm": 0.18701893091201782, "learning_rate": 5.84433019646096e-05, "loss": 0.0133, "step": 47370 }, { "grad_norm": 0.3021922707557678, "learning_rate": 5.842700425966863e-05, "loss": 0.0134, "step": 47380 }, { "grad_norm": 0.111005999147892, "learning_rate": 5.841070563316315e-05, "loss": 0.013, "step": 47390 }, { "grad_norm": 0.18243615329265594, "learning_rate": 5.839440608687554e-05, "loss": 0.0151, "step": 47400 }, { "grad_norm": 0.21161141991615295, "learning_rate": 5.837810562258831e-05, "loss": 0.0181, "step": 47410 }, { "grad_norm": 0.16514132916927338, "learning_rate": 5.836180424208405e-05, "loss": 0.013, "step": 47420 }, { "grad_norm": 0.22666169703006744, "learning_rate": 5.834550194714543e-05, "loss": 0.0171, "step": 47430 }, { "grad_norm": 0.16329918801784515, "learning_rate": 5.832919873955528e-05, "loss": 0.0148, "step": 47440 }, { "grad_norm": 0.15334296226501465, "learning_rate": 5.831289462109648e-05, "loss": 0.0132, "step": 47450 }, { "grad_norm": 0.1969919055700302, "learning_rate": 5.829658959355202e-05, "loss": 0.0141, "step": 47460 }, { "grad_norm": 0.16955192387104034, "learning_rate": 5.8280283658704995e-05, "loss": 0.0125, "step": 47470 }, { "grad_norm": 0.11407901346683502, "learning_rate": 5.826397681833859e-05, "loss": 0.0164, "step": 47480 }, { "grad_norm": 0.2055823802947998, "learning_rate": 5.824766907423612e-05, "loss": 0.0137, "step": 47490 }, { "grad_norm": 0.15560445189476013, "learning_rate": 5.8231360428180945e-05, "loss": 0.0133, "step": 47500 }, { "grad_norm": 0.1792769730091095, "learning_rate": 5.821505088195658e-05, "loss": 0.013, "step": 47510 }, { "grad_norm": 0.21839672327041626, "learning_rate": 5.819874043734661e-05, "loss": 0.0171, "step": 47520 }, { "grad_norm": 0.2480907291173935, "learning_rate": 5.8182429096134695e-05, "loss": 0.0152, "step": 47530 }, { "grad_norm": 0.22622661292552948, "learning_rate": 5.816611686010465e-05, "loss": 0.0142, "step": 47540 }, { "grad_norm": 0.20926886796951294, "learning_rate": 5.814980373104033e-05, "loss": 0.0126, "step": 47550 }, { "grad_norm": 0.16020609438419342, "learning_rate": 5.813348971072572e-05, "loss": 0.0158, "step": 47560 }, { "grad_norm": 0.1441412717103958, "learning_rate": 5.811717480094492e-05, "loss": 0.0148, "step": 47570 }, { "grad_norm": 0.16811256110668182, "learning_rate": 5.810085900348209e-05, "loss": 0.0159, "step": 47580 }, { "grad_norm": 0.18787963688373566, "learning_rate": 5.8084542320121483e-05, "loss": 0.0149, "step": 47590 }, { "grad_norm": 0.18855564296245575, "learning_rate": 5.8068224752647497e-05, "loss": 0.0158, "step": 47600 }, { "grad_norm": 0.2204955667257309, "learning_rate": 5.805190630284457e-05, "loss": 0.0137, "step": 47610 }, { "grad_norm": 0.15016186237335205, "learning_rate": 5.803558697249729e-05, "loss": 0.0131, "step": 47620 }, { "grad_norm": 0.21928656101226807, "learning_rate": 5.8019266763390276e-05, "loss": 0.0156, "step": 47630 }, { "grad_norm": 0.28717970848083496, "learning_rate": 5.800294567730833e-05, "loss": 0.0176, "step": 47640 }, { "grad_norm": 0.2349315732717514, "learning_rate": 5.798662371603628e-05, "loss": 0.0128, "step": 47650 }, { "grad_norm": 0.16668222844600677, "learning_rate": 5.797030088135904e-05, "loss": 0.0135, "step": 47660 }, { "grad_norm": 0.20284266769886017, "learning_rate": 5.7953977175061705e-05, "loss": 0.0131, "step": 47670 }, { "grad_norm": 0.19254936277866364, "learning_rate": 5.793765259892938e-05, "loss": 0.0192, "step": 47680 }, { "grad_norm": 0.22398832440376282, "learning_rate": 5.792132715474729e-05, "loss": 0.0181, "step": 47690 }, { "grad_norm": 0.1733344942331314, "learning_rate": 5.790500084430078e-05, "loss": 0.0143, "step": 47700 }, { "grad_norm": 0.14701023697853088, "learning_rate": 5.7888673669375274e-05, "loss": 0.0139, "step": 47710 }, { "grad_norm": 0.19500954449176788, "learning_rate": 5.787234563175625e-05, "loss": 0.0138, "step": 47720 }, { "grad_norm": 0.15139229595661163, "learning_rate": 5.7856016733229355e-05, "loss": 0.0148, "step": 47730 }, { "grad_norm": 0.17237533628940582, "learning_rate": 5.7839686975580297e-05, "loss": 0.0142, "step": 47740 }, { "grad_norm": 0.23190051317214966, "learning_rate": 5.782335636059484e-05, "loss": 0.0143, "step": 47750 }, { "grad_norm": 0.2358352243900299, "learning_rate": 5.780702489005889e-05, "loss": 0.0132, "step": 47760 }, { "grad_norm": 0.1604769378900528, "learning_rate": 5.779069256575845e-05, "loss": 0.0148, "step": 47770 }, { "grad_norm": 0.212010458111763, "learning_rate": 5.7774359389479574e-05, "loss": 0.0155, "step": 47780 }, { "grad_norm": 0.18389186263084412, "learning_rate": 5.775802536300845e-05, "loss": 0.0149, "step": 47790 }, { "grad_norm": 0.15077334642410278, "learning_rate": 5.774169048813134e-05, "loss": 0.0157, "step": 47800 }, { "grad_norm": 0.17522557079792023, "learning_rate": 5.77253547666346e-05, "loss": 0.0143, "step": 47810 }, { "grad_norm": 0.17195184528827667, "learning_rate": 5.770901820030465e-05, "loss": 0.0134, "step": 47820 }, { "grad_norm": 0.14643903076648712, "learning_rate": 5.769268079092809e-05, "loss": 0.0124, "step": 47830 }, { "grad_norm": 0.19388115406036377, "learning_rate": 5.767634254029151e-05, "loss": 0.0164, "step": 47840 }, { "grad_norm": 0.2155720591545105, "learning_rate": 5.7660003450181655e-05, "loss": 0.0143, "step": 47850 }, { "grad_norm": 0.15426261723041534, "learning_rate": 5.764366352238534e-05, "loss": 0.0124, "step": 47860 }, { "grad_norm": 0.15994475781917572, "learning_rate": 5.7627322758689474e-05, "loss": 0.014, "step": 47870 }, { "grad_norm": 0.20088349282741547, "learning_rate": 5.761098116088105e-05, "loss": 0.0123, "step": 47880 }, { "grad_norm": 0.1865849643945694, "learning_rate": 5.759463873074717e-05, "loss": 0.0124, "step": 47890 }, { "grad_norm": 0.19047996401786804, "learning_rate": 5.757829547007504e-05, "loss": 0.0143, "step": 47900 }, { "grad_norm": 0.17610107362270355, "learning_rate": 5.756195138065189e-05, "loss": 0.0163, "step": 47910 }, { "grad_norm": 0.1316964328289032, "learning_rate": 5.754560646426511e-05, "loss": 0.0132, "step": 47920 }, { "grad_norm": 0.12027892470359802, "learning_rate": 5.752926072270216e-05, "loss": 0.0128, "step": 47930 }, { "grad_norm": 0.15155726671218872, "learning_rate": 5.7512914157750563e-05, "loss": 0.0108, "step": 47940 }, { "grad_norm": 0.11552424728870392, "learning_rate": 5.749656677119798e-05, "loss": 0.0122, "step": 47950 }, { "grad_norm": 0.19057846069335938, "learning_rate": 5.7480218564832125e-05, "loss": 0.0111, "step": 47960 }, { "grad_norm": 0.1493980586528778, "learning_rate": 5.746386954044082e-05, "loss": 0.0129, "step": 47970 }, { "grad_norm": 0.16360725462436676, "learning_rate": 5.744751969981195e-05, "loss": 0.0123, "step": 47980 }, { "grad_norm": 0.20424087345600128, "learning_rate": 5.7431169044733526e-05, "loss": 0.0141, "step": 47990 }, { "grad_norm": 0.18597343564033508, "learning_rate": 5.741481757699364e-05, "loss": 0.0151, "step": 48000 }, { "grad_norm": 0.18956324458122253, "learning_rate": 5.7398465298380434e-05, "loss": 0.0184, "step": 48010 }, { "grad_norm": 0.14108026027679443, "learning_rate": 5.7382112210682193e-05, "loss": 0.0112, "step": 48020 }, { "grad_norm": 0.1851630061864853, "learning_rate": 5.7365758315687266e-05, "loss": 0.0165, "step": 48030 }, { "grad_norm": 0.15061064064502716, "learning_rate": 5.734940361518407e-05, "loss": 0.0131, "step": 48040 }, { "grad_norm": 0.21043072640895844, "learning_rate": 5.733304811096116e-05, "loss": 0.0142, "step": 48050 }, { "grad_norm": 0.21238121390342712, "learning_rate": 5.731669180480713e-05, "loss": 0.0129, "step": 48060 }, { "grad_norm": 0.14647014439105988, "learning_rate": 5.730033469851067e-05, "loss": 0.0144, "step": 48070 }, { "grad_norm": 0.2088271677494049, "learning_rate": 5.7283976793860607e-05, "loss": 0.0156, "step": 48080 }, { "grad_norm": 0.19046761095523834, "learning_rate": 5.726761809264577e-05, "loss": 0.0165, "step": 48090 }, { "grad_norm": 0.16937334835529327, "learning_rate": 5.7251258596655155e-05, "loss": 0.015, "step": 48100 }, { "grad_norm": 0.1988595575094223, "learning_rate": 5.72348983076778e-05, "loss": 0.0151, "step": 48110 }, { "grad_norm": 0.14006143808364868, "learning_rate": 5.7218537227502854e-05, "loss": 0.0167, "step": 48120 }, { "grad_norm": 0.16428276896476746, "learning_rate": 5.720217535791951e-05, "loss": 0.0136, "step": 48130 }, { "grad_norm": 0.21142610907554626, "learning_rate": 5.718581270071711e-05, "loss": 0.0133, "step": 48140 }, { "grad_norm": 0.16036120057106018, "learning_rate": 5.716944925768505e-05, "loss": 0.0128, "step": 48150 }, { "grad_norm": 0.18112283945083618, "learning_rate": 5.7153085030612786e-05, "loss": 0.0116, "step": 48160 }, { "grad_norm": 0.1456008404493332, "learning_rate": 5.713672002128989e-05, "loss": 0.0161, "step": 48170 }, { "grad_norm": 0.16275185346603394, "learning_rate": 5.712035423150602e-05, "loss": 0.0149, "step": 48180 }, { "grad_norm": 0.1798987090587616, "learning_rate": 5.710398766305094e-05, "loss": 0.0144, "step": 48190 }, { "grad_norm": 0.15940633416175842, "learning_rate": 5.708762031771442e-05, "loss": 0.0158, "step": 48200 }, { "grad_norm": 0.14244496822357178, "learning_rate": 5.707125219728642e-05, "loss": 0.0121, "step": 48210 }, { "grad_norm": 0.3342890441417694, "learning_rate": 5.7054883303556905e-05, "loss": 0.0169, "step": 48220 }, { "grad_norm": 0.12622202932834625, "learning_rate": 5.703851363831595e-05, "loss": 0.0126, "step": 48230 }, { "grad_norm": 0.21141697466373444, "learning_rate": 5.702214320335374e-05, "loss": 0.0183, "step": 48240 }, { "grad_norm": 0.20734313130378723, "learning_rate": 5.7005772000460514e-05, "loss": 0.0141, "step": 48250 }, { "grad_norm": 0.1271897703409195, "learning_rate": 5.698940003142656e-05, "loss": 0.0119, "step": 48260 }, { "grad_norm": 0.18341164290905, "learning_rate": 5.697302729804236e-05, "loss": 0.0116, "step": 48270 }, { "grad_norm": 0.1665194034576416, "learning_rate": 5.695665380209837e-05, "loss": 0.0149, "step": 48280 }, { "grad_norm": 0.20726126432418823, "learning_rate": 5.6940279545385165e-05, "loss": 0.0123, "step": 48290 }, { "grad_norm": 0.11201166361570358, "learning_rate": 5.692390452969344e-05, "loss": 0.0133, "step": 48300 }, { "grad_norm": 0.16543728113174438, "learning_rate": 5.690752875681392e-05, "loss": 0.0155, "step": 48310 }, { "grad_norm": 0.15493634343147278, "learning_rate": 5.6891152228537435e-05, "loss": 0.0141, "step": 48320 }, { "grad_norm": 0.23386669158935547, "learning_rate": 5.687477494665492e-05, "loss": 0.011, "step": 48330 }, { "grad_norm": 0.1939079761505127, "learning_rate": 5.685839691295733e-05, "loss": 0.0127, "step": 48340 }, { "grad_norm": 0.2727333605289459, "learning_rate": 5.6842018129235786e-05, "loss": 0.0159, "step": 48350 }, { "grad_norm": 0.20760555565357208, "learning_rate": 5.6825638597281404e-05, "loss": 0.0184, "step": 48360 }, { "grad_norm": 0.1828320473432541, "learning_rate": 5.680925831888546e-05, "loss": 0.019, "step": 48370 }, { "grad_norm": 0.17729489505290985, "learning_rate": 5.6792877295839274e-05, "loss": 0.0126, "step": 48380 }, { "grad_norm": 0.17176081240177155, "learning_rate": 5.6776495529934224e-05, "loss": 0.0144, "step": 48390 }, { "grad_norm": 0.1780461072921753, "learning_rate": 5.6760113022961824e-05, "loss": 0.0116, "step": 48400 }, { "grad_norm": 0.20644931495189667, "learning_rate": 5.6743729776713617e-05, "loss": 0.0145, "step": 48410 }, { "grad_norm": 0.22018054127693176, "learning_rate": 5.672734579298126e-05, "loss": 0.0139, "step": 48420 }, { "grad_norm": 0.14996765553951263, "learning_rate": 5.671096107355649e-05, "loss": 0.0136, "step": 48430 }, { "grad_norm": 0.23213165998458862, "learning_rate": 5.66945756202311e-05, "loss": 0.0199, "step": 48440 }, { "grad_norm": 0.1912783831357956, "learning_rate": 5.667818943479699e-05, "loss": 0.0165, "step": 48450 }, { "grad_norm": 0.1484157145023346, "learning_rate": 5.666180251904612e-05, "loss": 0.0131, "step": 48460 }, { "grad_norm": 0.17469476163387299, "learning_rate": 5.6645414874770555e-05, "loss": 0.0132, "step": 48470 }, { "grad_norm": 0.14823494851589203, "learning_rate": 5.66290265037624e-05, "loss": 0.0123, "step": 48480 }, { "grad_norm": 0.17144396901130676, "learning_rate": 5.661263740781386e-05, "loss": 0.0135, "step": 48490 }, { "grad_norm": 0.15034301578998566, "learning_rate": 5.6596247588717254e-05, "loss": 0.0149, "step": 48500 }, { "grad_norm": 0.19088377058506012, "learning_rate": 5.6579857048264926e-05, "loss": 0.0145, "step": 48510 }, { "grad_norm": 0.162791907787323, "learning_rate": 5.6563465788249314e-05, "loss": 0.0141, "step": 48520 }, { "grad_norm": 0.18091212213039398, "learning_rate": 5.6547073810462956e-05, "loss": 0.0128, "step": 48530 }, { "grad_norm": 0.179407000541687, "learning_rate": 5.653068111669846e-05, "loss": 0.0127, "step": 48540 }, { "grad_norm": 0.13842880725860596, "learning_rate": 5.651428770874848e-05, "loss": 0.0152, "step": 48550 }, { "grad_norm": 0.1270778775215149, "learning_rate": 5.64978935884058e-05, "loss": 0.0115, "step": 48560 }, { "grad_norm": 0.25214213132858276, "learning_rate": 5.6481498757463244e-05, "loss": 0.017, "step": 48570 }, { "grad_norm": 0.20722202956676483, "learning_rate": 5.646510321771373e-05, "loss": 0.0152, "step": 48580 }, { "grad_norm": 0.19059576094150543, "learning_rate": 5.644870697095024e-05, "loss": 0.0139, "step": 48590 }, { "grad_norm": 0.14899942278862, "learning_rate": 5.643231001896586e-05, "loss": 0.0107, "step": 48600 }, { "grad_norm": 0.16679395735263824, "learning_rate": 5.6415912363553726e-05, "loss": 0.0122, "step": 48610 }, { "grad_norm": 0.17606277763843536, "learning_rate": 5.639951400650706e-05, "loss": 0.0142, "step": 48620 }, { "grad_norm": 0.18360282480716705, "learning_rate": 5.6383114949619165e-05, "loss": 0.0104, "step": 48630 }, { "grad_norm": 0.1842501312494278, "learning_rate": 5.636671519468342e-05, "loss": 0.0139, "step": 48640 }, { "grad_norm": 0.18899179995059967, "learning_rate": 5.635031474349327e-05, "loss": 0.0118, "step": 48650 }, { "grad_norm": 0.1679547131061554, "learning_rate": 5.6333913597842246e-05, "loss": 0.0136, "step": 48660 }, { "grad_norm": 0.18957391381263733, "learning_rate": 5.6317511759523955e-05, "loss": 0.0135, "step": 48670 }, { "grad_norm": 0.12793131172657013, "learning_rate": 5.630110923033207e-05, "loss": 0.0124, "step": 48680 }, { "grad_norm": 0.1564183384180069, "learning_rate": 5.628470601206036e-05, "loss": 0.0126, "step": 48690 }, { "grad_norm": 0.3086400032043457, "learning_rate": 5.626830210650266e-05, "loss": 0.0125, "step": 48700 }, { "grad_norm": 0.18185876309871674, "learning_rate": 5.625189751545285e-05, "loss": 0.016, "step": 48710 }, { "grad_norm": 0.16778409481048584, "learning_rate": 5.6235492240704936e-05, "loss": 0.014, "step": 48720 }, { "grad_norm": 0.22586549818515778, "learning_rate": 5.621908628405296e-05, "loss": 0.0126, "step": 48730 }, { "grad_norm": 0.16750620305538177, "learning_rate": 5.620267964729106e-05, "loss": 0.0188, "step": 48740 }, { "grad_norm": 0.1858324259519577, "learning_rate": 5.618627233221344e-05, "loss": 0.0141, "step": 48750 }, { "grad_norm": 0.36365360021591187, "learning_rate": 5.616986434061438e-05, "loss": 0.0134, "step": 48760 }, { "grad_norm": 0.20502737164497375, "learning_rate": 5.615345567428822e-05, "loss": 0.0151, "step": 48770 }, { "grad_norm": 0.1750248670578003, "learning_rate": 5.613704633502941e-05, "loss": 0.0144, "step": 48780 }, { "grad_norm": 0.12550750374794006, "learning_rate": 5.612063632463245e-05, "loss": 0.0144, "step": 48790 }, { "grad_norm": 0.17384815216064453, "learning_rate": 5.610422564489188e-05, "loss": 0.0131, "step": 48800 }, { "grad_norm": 0.22272446751594543, "learning_rate": 5.608781429760239e-05, "loss": 0.0127, "step": 48810 }, { "grad_norm": 0.20039428770542145, "learning_rate": 5.607140228455866e-05, "loss": 0.0152, "step": 48820 }, { "grad_norm": 0.210164412856102, "learning_rate": 5.605498960755553e-05, "loss": 0.0142, "step": 48830 }, { "grad_norm": 0.21182674169540405, "learning_rate": 5.603857626838782e-05, "loss": 0.014, "step": 48840 }, { "grad_norm": 0.16399286687374115, "learning_rate": 5.60221622688505e-05, "loss": 0.0145, "step": 48850 }, { "grad_norm": 0.3741653561592102, "learning_rate": 5.6005747610738565e-05, "loss": 0.0132, "step": 48860 }, { "grad_norm": 0.1267252117395401, "learning_rate": 5.59893322958471e-05, "loss": 0.0144, "step": 48870 }, { "grad_norm": 0.24187397956848145, "learning_rate": 5.5972916325971256e-05, "loss": 0.018, "step": 48880 }, { "grad_norm": 0.1353297084569931, "learning_rate": 5.595649970290628e-05, "loss": 0.0134, "step": 48890 }, { "grad_norm": 0.17345568537712097, "learning_rate": 5.5940082428447426e-05, "loss": 0.0166, "step": 48900 }, { "grad_norm": 0.23990051448345184, "learning_rate": 5.592366450439012e-05, "loss": 0.0139, "step": 48910 }, { "grad_norm": 0.20788438618183136, "learning_rate": 5.590724593252975e-05, "loss": 0.0142, "step": 48920 }, { "grad_norm": 0.18568049371242523, "learning_rate": 5.589082671466184e-05, "loss": 0.0156, "step": 48930 }, { "grad_norm": 0.19264529645442963, "learning_rate": 5.587440685258199e-05, "loss": 0.0128, "step": 48940 }, { "grad_norm": 0.16258971393108368, "learning_rate": 5.585798634808583e-05, "loss": 0.0144, "step": 48950 }, { "grad_norm": 0.1819021999835968, "learning_rate": 5.584156520296909e-05, "loss": 0.0124, "step": 48960 }, { "grad_norm": 0.20600298047065735, "learning_rate": 5.582514341902757e-05, "loss": 0.017, "step": 48970 }, { "grad_norm": 0.20193852484226227, "learning_rate": 5.580872099805713e-05, "loss": 0.0151, "step": 48980 }, { "grad_norm": 0.2189759761095047, "learning_rate": 5.5792297941853674e-05, "loss": 0.0158, "step": 48990 }, { "grad_norm": 0.20483867824077606, "learning_rate": 5.5775874252213247e-05, "loss": 0.0191, "step": 49000 }, { "grad_norm": 0.2224719524383545, "learning_rate": 5.575944993093189e-05, "loss": 0.0152, "step": 49010 }, { "grad_norm": 0.13664510846138, "learning_rate": 5.574302497980574e-05, "loss": 0.0142, "step": 49020 }, { "grad_norm": 0.1792525053024292, "learning_rate": 5.5726599400631005e-05, "loss": 0.0138, "step": 49030 }, { "grad_norm": 0.15971606969833374, "learning_rate": 5.571017319520399e-05, "loss": 0.0143, "step": 49040 }, { "grad_norm": 0.1220245510339737, "learning_rate": 5.5693746365320984e-05, "loss": 0.0134, "step": 49050 }, { "grad_norm": 0.18632496893405914, "learning_rate": 5.567731891277848e-05, "loss": 0.0147, "step": 49060 }, { "grad_norm": 0.17619270086288452, "learning_rate": 5.566089083937288e-05, "loss": 0.0167, "step": 49070 }, { "grad_norm": 0.22282452881336212, "learning_rate": 5.564446214690079e-05, "loss": 0.0168, "step": 49080 }, { "grad_norm": 0.16433586180210114, "learning_rate": 5.56280328371588e-05, "loss": 0.0163, "step": 49090 }, { "grad_norm": 0.18347252905368805, "learning_rate": 5.5611602911943596e-05, "loss": 0.0134, "step": 49100 }, { "grad_norm": 0.16217198967933655, "learning_rate": 5.559517237305194e-05, "loss": 0.0147, "step": 49110 }, { "grad_norm": 0.1354481726884842, "learning_rate": 5.557874122228064e-05, "loss": 0.0127, "step": 49120 }, { "grad_norm": 0.19047054648399353, "learning_rate": 5.55623094614266e-05, "loss": 0.0183, "step": 49130 }, { "grad_norm": 0.1574583798646927, "learning_rate": 5.5545877092286743e-05, "loss": 0.0153, "step": 49140 }, { "grad_norm": 0.14161960780620575, "learning_rate": 5.552944411665811e-05, "loss": 0.0135, "step": 49150 }, { "grad_norm": 0.2134132832288742, "learning_rate": 5.5513010536337795e-05, "loss": 0.0116, "step": 49160 }, { "grad_norm": 0.1775631606578827, "learning_rate": 5.5496576353122934e-05, "loss": 0.0149, "step": 49170 }, { "grad_norm": 0.15655216574668884, "learning_rate": 5.548014156881074e-05, "loss": 0.0119, "step": 49180 }, { "grad_norm": 0.2454865723848343, "learning_rate": 5.5463706185198494e-05, "loss": 0.0184, "step": 49190 }, { "grad_norm": 0.19967612624168396, "learning_rate": 5.544727020408358e-05, "loss": 0.0134, "step": 49200 }, { "grad_norm": 0.2214481234550476, "learning_rate": 5.5430833627263366e-05, "loss": 0.0137, "step": 49210 }, { "grad_norm": 0.19199608266353607, "learning_rate": 5.5414396456535354e-05, "loss": 0.0129, "step": 49220 }, { "grad_norm": 0.21926569938659668, "learning_rate": 5.539795869369709e-05, "loss": 0.016, "step": 49230 }, { "grad_norm": 0.19176727533340454, "learning_rate": 5.538152034054618e-05, "loss": 0.0146, "step": 49240 }, { "grad_norm": 0.11391539126634598, "learning_rate": 5.536508139888028e-05, "loss": 0.0115, "step": 49250 }, { "grad_norm": 0.15146352350711823, "learning_rate": 5.534864187049716e-05, "loss": 0.0127, "step": 49260 }, { "grad_norm": 0.1501329392194748, "learning_rate": 5.5332201757194615e-05, "loss": 0.0107, "step": 49270 }, { "grad_norm": 0.14552408456802368, "learning_rate": 5.531576106077049e-05, "loss": 0.0115, "step": 49280 }, { "grad_norm": 0.15470604598522186, "learning_rate": 5.529931978302272e-05, "loss": 0.0128, "step": 49290 }, { "grad_norm": 0.24958011507987976, "learning_rate": 5.528287792574932e-05, "loss": 0.0161, "step": 49300 }, { "grad_norm": 0.21626776456832886, "learning_rate": 5.5266435490748324e-05, "loss": 0.016, "step": 49310 }, { "grad_norm": 0.1623249053955078, "learning_rate": 5.524999247981787e-05, "loss": 0.0138, "step": 49320 }, { "grad_norm": 0.23790717124938965, "learning_rate": 5.523354889475613e-05, "loss": 0.0148, "step": 49330 }, { "grad_norm": 0.21375672519207, "learning_rate": 5.521710473736134e-05, "loss": 0.0149, "step": 49340 }, { "grad_norm": 0.1781850904226303, "learning_rate": 5.5200660009431826e-05, "loss": 0.0151, "step": 49350 }, { "grad_norm": 0.15555083751678467, "learning_rate": 5.518421471276596e-05, "loss": 0.0157, "step": 49360 }, { "grad_norm": 0.23405712842941284, "learning_rate": 5.5167768849162174e-05, "loss": 0.0174, "step": 49370 }, { "grad_norm": 0.1974465698003769, "learning_rate": 5.515132242041893e-05, "loss": 0.0132, "step": 49380 }, { "grad_norm": 0.20178896188735962, "learning_rate": 5.513487542833483e-05, "loss": 0.0153, "step": 49390 }, { "grad_norm": 0.2217586487531662, "learning_rate": 5.5118427874708466e-05, "loss": 0.0134, "step": 49400 }, { "grad_norm": 0.1988717019557953, "learning_rate": 5.510197976133853e-05, "loss": 0.0128, "step": 49410 }, { "grad_norm": 0.19447435438632965, "learning_rate": 5.508553109002376e-05, "loss": 0.0135, "step": 49420 }, { "grad_norm": 0.29633691906929016, "learning_rate": 5.5069081862562957e-05, "loss": 0.0152, "step": 49430 }, { "grad_norm": 0.21946032345294952, "learning_rate": 5.5052632080754965e-05, "loss": 0.0178, "step": 49440 }, { "grad_norm": 0.29504385590553284, "learning_rate": 5.503618174639874e-05, "loss": 0.015, "step": 49450 }, { "grad_norm": 0.20238898694515228, "learning_rate": 5.501973086129325e-05, "loss": 0.0149, "step": 49460 }, { "grad_norm": 0.17225666344165802, "learning_rate": 5.500327942723753e-05, "loss": 0.0123, "step": 49470 }, { "grad_norm": 0.1727956384420395, "learning_rate": 5.498682744603071e-05, "loss": 0.0135, "step": 49480 }, { "grad_norm": 0.17476896941661835, "learning_rate": 5.497037491947194e-05, "loss": 0.0149, "step": 49490 }, { "grad_norm": 0.20505309104919434, "learning_rate": 5.4953921849360424e-05, "loss": 0.0121, "step": 49500 }, { "grad_norm": 0.13811184465885162, "learning_rate": 5.493746823749547e-05, "loss": 0.0114, "step": 49510 }, { "grad_norm": 0.1856851726770401, "learning_rate": 5.49210140856764e-05, "loss": 0.0181, "step": 49520 }, { "grad_norm": 0.16723297536373138, "learning_rate": 5.4904559395702647e-05, "loss": 0.0109, "step": 49530 }, { "grad_norm": 0.13963189721107483, "learning_rate": 5.488810416937364e-05, "loss": 0.0111, "step": 49540 }, { "grad_norm": 0.18266639113426208, "learning_rate": 5.4871648408488926e-05, "loss": 0.0112, "step": 49550 }, { "grad_norm": 0.14838773012161255, "learning_rate": 5.485519211484807e-05, "loss": 0.0133, "step": 49560 }, { "grad_norm": 0.19190452992916107, "learning_rate": 5.4838735290250676e-05, "loss": 0.0144, "step": 49570 }, { "grad_norm": 0.1743556708097458, "learning_rate": 5.482227793649649e-05, "loss": 0.0144, "step": 49580 }, { "grad_norm": 0.17846739292144775, "learning_rate": 5.480582005538524e-05, "loss": 0.0126, "step": 49590 }, { "grad_norm": 0.2492997646331787, "learning_rate": 5.478936164871671e-05, "loss": 0.0157, "step": 49600 }, { "grad_norm": 0.1439845710992813, "learning_rate": 5.477290271829081e-05, "loss": 0.0177, "step": 49610 }, { "grad_norm": 0.18836289644241333, "learning_rate": 5.475644326590744e-05, "loss": 0.0146, "step": 49620 }, { "grad_norm": 0.17256875336170197, "learning_rate": 5.473998329336658e-05, "loss": 0.0143, "step": 49630 }, { "grad_norm": 0.18067875504493713, "learning_rate": 5.4723522802468286e-05, "loss": 0.0098, "step": 49640 }, { "grad_norm": 0.1796674132347107, "learning_rate": 5.470706179501264e-05, "loss": 0.0104, "step": 49650 }, { "grad_norm": 0.24013753235340118, "learning_rate": 5.4690600272799776e-05, "loss": 0.013, "step": 49660 }, { "grad_norm": 0.1410800814628601, "learning_rate": 5.467413823762993e-05, "loss": 0.0122, "step": 49670 }, { "grad_norm": 0.2443818598985672, "learning_rate": 5.465767569130335e-05, "loss": 0.0151, "step": 49680 }, { "grad_norm": 0.13953383266925812, "learning_rate": 5.464121263562036e-05, "loss": 0.012, "step": 49690 }, { "grad_norm": 0.1813509613275528, "learning_rate": 5.4624749072381343e-05, "loss": 0.016, "step": 49700 }, { "grad_norm": 0.16795459389686584, "learning_rate": 5.460828500338672e-05, "loss": 0.0141, "step": 49710 }, { "grad_norm": 0.1679478883743286, "learning_rate": 5.459182043043698e-05, "loss": 0.0118, "step": 49720 }, { "grad_norm": 0.16728626191616058, "learning_rate": 5.457535535533265e-05, "loss": 0.0132, "step": 49730 }, { "grad_norm": 0.15727804601192474, "learning_rate": 5.4558889779874334e-05, "loss": 0.0155, "step": 49740 }, { "grad_norm": 0.1671067625284195, "learning_rate": 5.454242370586269e-05, "loss": 0.0158, "step": 49750 }, { "grad_norm": 0.19161762297153473, "learning_rate": 5.452595713509843e-05, "loss": 0.0123, "step": 49760 }, { "grad_norm": 0.2306191772222519, "learning_rate": 5.4509490069382274e-05, "loss": 0.0109, "step": 49770 }, { "grad_norm": 0.13634797930717468, "learning_rate": 5.449302251051509e-05, "loss": 0.0108, "step": 49780 }, { "grad_norm": 0.24575668573379517, "learning_rate": 5.44765544602977e-05, "loss": 0.012, "step": 49790 }, { "grad_norm": 0.16938616335391998, "learning_rate": 5.446008592053107e-05, "loss": 0.0126, "step": 49800 }, { "grad_norm": 0.17498406767845154, "learning_rate": 5.4443616893016135e-05, "loss": 0.0145, "step": 49810 }, { "grad_norm": 0.13554729521274567, "learning_rate": 5.4427147379553934e-05, "loss": 0.0131, "step": 49820 }, { "grad_norm": 0.25995126366615295, "learning_rate": 5.441067738194556e-05, "loss": 0.0136, "step": 49830 }, { "grad_norm": 0.13582757115364075, "learning_rate": 5.439420690199214e-05, "loss": 0.0176, "step": 49840 }, { "grad_norm": 0.14196063578128815, "learning_rate": 5.437773594149484e-05, "loss": 0.0117, "step": 49850 }, { "grad_norm": 0.2397577464580536, "learning_rate": 5.436126450225495e-05, "loss": 0.0149, "step": 49860 }, { "grad_norm": 0.12713883817195892, "learning_rate": 5.434479258607371e-05, "loss": 0.0123, "step": 49870 }, { "grad_norm": 0.12409593909978867, "learning_rate": 5.43283201947525e-05, "loss": 0.0143, "step": 49880 }, { "grad_norm": 0.16688185930252075, "learning_rate": 5.431184733009268e-05, "loss": 0.0159, "step": 49890 }, { "grad_norm": 0.14379943907260895, "learning_rate": 5.4295373993895736e-05, "loss": 0.0111, "step": 49900 }, { "grad_norm": 0.1785423457622528, "learning_rate": 5.4278900187963157e-05, "loss": 0.0162, "step": 49910 }, { "grad_norm": 0.12383003532886505, "learning_rate": 5.426242591409646e-05, "loss": 0.0143, "step": 49920 }, { "grad_norm": 0.18458621203899384, "learning_rate": 5.42459511740973e-05, "loss": 0.0137, "step": 49930 }, { "grad_norm": 0.19367577135562897, "learning_rate": 5.422947596976729e-05, "loss": 0.0111, "step": 49940 }, { "grad_norm": 0.2162303328514099, "learning_rate": 5.4213000302908134e-05, "loss": 0.0138, "step": 49950 }, { "grad_norm": 0.22154171764850616, "learning_rate": 5.419652417532162e-05, "loss": 0.0191, "step": 49960 }, { "grad_norm": 0.18359866738319397, "learning_rate": 5.4180047588809534e-05, "loss": 0.0141, "step": 49970 }, { "grad_norm": 0.16090764105319977, "learning_rate": 5.4163570545173704e-05, "loss": 0.016, "step": 49980 }, { "grad_norm": 0.15038785338401794, "learning_rate": 5.414709304621608e-05, "loss": 0.0144, "step": 49990 }, { "grad_norm": 0.1999405026435852, "learning_rate": 5.413061509373858e-05, "loss": 0.0146, "step": 50000 }, { "grad_norm": 0.13185667991638184, "learning_rate": 5.411413668954324e-05, "loss": 0.0128, "step": 50010 }, { "grad_norm": 0.15974335372447968, "learning_rate": 5.4097657835432083e-05, "loss": 0.0137, "step": 50020 }, { "grad_norm": 0.1712745577096939, "learning_rate": 5.408117853320723e-05, "loss": 0.0127, "step": 50030 }, { "grad_norm": 0.17531977593898773, "learning_rate": 5.406469878467084e-05, "loss": 0.0144, "step": 50040 }, { "grad_norm": 0.18632382154464722, "learning_rate": 5.404821859162509e-05, "loss": 0.0159, "step": 50050 }, { "grad_norm": 0.20714670419692993, "learning_rate": 5.403173795587225e-05, "loss": 0.0165, "step": 50060 }, { "grad_norm": 0.14679460227489471, "learning_rate": 5.4015256879214606e-05, "loss": 0.0132, "step": 50070 }, { "grad_norm": 0.16844475269317627, "learning_rate": 5.39987753634545e-05, "loss": 0.0124, "step": 50080 }, { "grad_norm": 0.14607886970043182, "learning_rate": 5.3982293410394335e-05, "loss": 0.014, "step": 50090 }, { "grad_norm": 0.1594051867723465, "learning_rate": 5.396581102183655e-05, "loss": 0.0116, "step": 50100 }, { "grad_norm": 0.15136636793613434, "learning_rate": 5.3949328199583634e-05, "loss": 0.012, "step": 50110 }, { "grad_norm": 0.14705152809619904, "learning_rate": 5.393284494543813e-05, "loss": 0.0164, "step": 50120 }, { "grad_norm": 0.1735532134771347, "learning_rate": 5.391636126120262e-05, "loss": 0.013, "step": 50130 }, { "grad_norm": 0.16290298104286194, "learning_rate": 5.389987714867971e-05, "loss": 0.0172, "step": 50140 }, { "grad_norm": 0.18009130656719208, "learning_rate": 5.3883392609672114e-05, "loss": 0.0115, "step": 50150 }, { "grad_norm": 0.1873912662267685, "learning_rate": 5.3866907645982545e-05, "loss": 0.0128, "step": 50160 }, { "grad_norm": 0.34050217270851135, "learning_rate": 5.385042225941375e-05, "loss": 0.0136, "step": 50170 }, { "grad_norm": 0.1752423793077469, "learning_rate": 5.3833936451768585e-05, "loss": 0.0127, "step": 50180 }, { "grad_norm": 0.20793521404266357, "learning_rate": 5.3817450224849884e-05, "loss": 0.0159, "step": 50190 }, { "grad_norm": 0.17010340094566345, "learning_rate": 5.380096358046056e-05, "loss": 0.0124, "step": 50200 }, { "grad_norm": 0.10972204059362411, "learning_rate": 5.378447652040359e-05, "loss": 0.0117, "step": 50210 }, { "grad_norm": 0.15416134893894196, "learning_rate": 5.376798904648195e-05, "loss": 0.0149, "step": 50220 }, { "grad_norm": 0.18013450503349304, "learning_rate": 5.375150116049869e-05, "loss": 0.018, "step": 50230 }, { "grad_norm": 0.1960536688566208, "learning_rate": 5.373501286425691e-05, "loss": 0.0146, "step": 50240 }, { "grad_norm": 0.19497960805892944, "learning_rate": 5.3718524159559726e-05, "loss": 0.0138, "step": 50250 }, { "grad_norm": 0.19062519073486328, "learning_rate": 5.370203504821034e-05, "loss": 0.0151, "step": 50260 }, { "grad_norm": 0.2099740356206894, "learning_rate": 5.368554553201196e-05, "loss": 0.018, "step": 50270 }, { "grad_norm": 0.15277494490146637, "learning_rate": 5.3669055612767874e-05, "loss": 0.0139, "step": 50280 }, { "grad_norm": 0.1661359667778015, "learning_rate": 5.3652565292281376e-05, "loss": 0.0155, "step": 50290 }, { "grad_norm": 0.21819676458835602, "learning_rate": 5.3636074572355824e-05, "loss": 0.0123, "step": 50300 }, { "grad_norm": 0.16434259712696075, "learning_rate": 5.361958345479463e-05, "loss": 0.0129, "step": 50310 }, { "grad_norm": 0.17003227770328522, "learning_rate": 5.360309194140123e-05, "loss": 0.0121, "step": 50320 }, { "grad_norm": 0.14580437541007996, "learning_rate": 5.3586600033979106e-05, "loss": 0.0122, "step": 50330 }, { "grad_norm": 0.20204028487205505, "learning_rate": 5.357010773433181e-05, "loss": 0.0114, "step": 50340 }, { "grad_norm": 0.21454420685768127, "learning_rate": 5.3553615044262895e-05, "loss": 0.013, "step": 50350 }, { "grad_norm": 0.1613638550043106, "learning_rate": 5.3537121965575964e-05, "loss": 0.0132, "step": 50360 }, { "grad_norm": 0.15198075771331787, "learning_rate": 5.352062850007471e-05, "loss": 0.0096, "step": 50370 }, { "grad_norm": 0.20730279386043549, "learning_rate": 5.350413464956282e-05, "loss": 0.0151, "step": 50380 }, { "grad_norm": 0.14461104571819305, "learning_rate": 5.348764041584403e-05, "loss": 0.0105, "step": 50390 }, { "grad_norm": 0.11555282026529312, "learning_rate": 5.347114580072212e-05, "loss": 0.0111, "step": 50400 }, { "grad_norm": 0.15683573484420776, "learning_rate": 5.345465080600093e-05, "loss": 0.0145, "step": 50410 }, { "grad_norm": 0.20243977010250092, "learning_rate": 5.3438155433484326e-05, "loss": 0.0113, "step": 50420 }, { "grad_norm": 0.2538209855556488, "learning_rate": 5.3421659684976197e-05, "loss": 0.0168, "step": 50430 }, { "grad_norm": 0.1378522366285324, "learning_rate": 5.340516356228052e-05, "loss": 0.0147, "step": 50440 }, { "grad_norm": 0.1986742913722992, "learning_rate": 5.338866706720128e-05, "loss": 0.017, "step": 50450 }, { "grad_norm": 0.13390915095806122, "learning_rate": 5.337217020154249e-05, "loss": 0.0122, "step": 50460 }, { "grad_norm": 0.15839959681034088, "learning_rate": 5.335567296710825e-05, "loss": 0.0128, "step": 50470 }, { "grad_norm": 0.13482114672660828, "learning_rate": 5.333917536570265e-05, "loss": 0.0121, "step": 50480 }, { "grad_norm": 0.20967786014080048, "learning_rate": 5.332267739912986e-05, "loss": 0.0129, "step": 50490 }, { "grad_norm": 0.1753579080104828, "learning_rate": 5.330617906919405e-05, "loss": 0.0105, "step": 50500 }, { "grad_norm": 0.11499416828155518, "learning_rate": 5.328968037769949e-05, "loss": 0.0111, "step": 50510 }, { "grad_norm": 0.16904982924461365, "learning_rate": 5.327318132645042e-05, "loss": 0.0118, "step": 50520 }, { "grad_norm": 0.1595100462436676, "learning_rate": 5.325668191725116e-05, "loss": 0.0151, "step": 50530 }, { "grad_norm": 0.14285659790039062, "learning_rate": 5.324018215190606e-05, "loss": 0.0109, "step": 50540 }, { "grad_norm": 0.18958108127117157, "learning_rate": 5.3223682032219515e-05, "loss": 0.0128, "step": 50550 }, { "grad_norm": 0.24109390377998352, "learning_rate": 5.320718155999595e-05, "loss": 0.0145, "step": 50560 }, { "grad_norm": 0.2177293300628662, "learning_rate": 5.3190680737039835e-05, "loss": 0.0166, "step": 50570 }, { "grad_norm": 0.12935543060302734, "learning_rate": 5.317417956515567e-05, "loss": 0.0129, "step": 50580 }, { "grad_norm": 0.17600376904010773, "learning_rate": 5.315767804614798e-05, "loss": 0.0158, "step": 50590 }, { "grad_norm": 0.1641702950000763, "learning_rate": 5.3141176181821395e-05, "loss": 0.0117, "step": 50600 }, { "grad_norm": 0.15559810400009155, "learning_rate": 5.31246739739805e-05, "loss": 0.0176, "step": 50610 }, { "grad_norm": 0.16761457920074463, "learning_rate": 5.310817142442994e-05, "loss": 0.0132, "step": 50620 }, { "grad_norm": 0.16843262314796448, "learning_rate": 5.309166853497445e-05, "loss": 0.0138, "step": 50630 }, { "grad_norm": 0.16287827491760254, "learning_rate": 5.307516530741873e-05, "loss": 0.0127, "step": 50640 }, { "grad_norm": 0.14225460588932037, "learning_rate": 5.305866174356754e-05, "loss": 0.0105, "step": 50650 }, { "grad_norm": 0.1831848919391632, "learning_rate": 5.304215784522571e-05, "loss": 0.0139, "step": 50660 }, { "grad_norm": 0.17064887285232544, "learning_rate": 5.302565361419808e-05, "loss": 0.0121, "step": 50670 }, { "grad_norm": 0.17343007028102875, "learning_rate": 5.3009149052289507e-05, "loss": 0.0125, "step": 50680 }, { "grad_norm": 0.19834987819194794, "learning_rate": 5.299264416130493e-05, "loss": 0.0131, "step": 50690 }, { "grad_norm": 0.23182731866836548, "learning_rate": 5.297613894304928e-05, "loss": 0.0127, "step": 50700 }, { "grad_norm": 0.18028207123279572, "learning_rate": 5.2959633399327534e-05, "loss": 0.0116, "step": 50710 }, { "grad_norm": 0.21647779643535614, "learning_rate": 5.294312753194476e-05, "loss": 0.0136, "step": 50720 }, { "grad_norm": 0.20073522627353668, "learning_rate": 5.292662134270596e-05, "loss": 0.0113, "step": 50730 }, { "grad_norm": 0.1542368233203888, "learning_rate": 5.291011483341626e-05, "loss": 0.0175, "step": 50740 }, { "grad_norm": 0.2096932828426361, "learning_rate": 5.2893608005880767e-05, "loss": 0.0145, "step": 50750 }, { "grad_norm": 0.15775834023952484, "learning_rate": 5.287710086190467e-05, "loss": 0.0152, "step": 50760 }, { "grad_norm": 0.18783557415008545, "learning_rate": 5.286059340329314e-05, "loss": 0.0165, "step": 50770 }, { "grad_norm": 0.12805616855621338, "learning_rate": 5.284408563185141e-05, "loss": 0.0128, "step": 50780 }, { "grad_norm": 0.13789886236190796, "learning_rate": 5.282757754938476e-05, "loss": 0.0159, "step": 50790 }, { "grad_norm": 0.16860105097293854, "learning_rate": 5.281106915769849e-05, "loss": 0.0125, "step": 50800 }, { "grad_norm": 0.2158210426568985, "learning_rate": 5.2794560458597897e-05, "loss": 0.019, "step": 50810 }, { "grad_norm": 0.13538306951522827, "learning_rate": 5.27780514538884e-05, "loss": 0.0151, "step": 50820 }, { "grad_norm": 0.1557222455739975, "learning_rate": 5.2761542145375365e-05, "loss": 0.0146, "step": 50830 }, { "grad_norm": 0.16496042907238007, "learning_rate": 5.274503253486421e-05, "loss": 0.0154, "step": 50840 }, { "grad_norm": 0.19103828072547913, "learning_rate": 5.272852262416046e-05, "loss": 0.0121, "step": 50850 }, { "grad_norm": 0.19362425804138184, "learning_rate": 5.2712012415069555e-05, "loss": 0.0137, "step": 50860 }, { "grad_norm": 0.1585603654384613, "learning_rate": 5.269550190939705e-05, "loss": 0.0156, "step": 50870 }, { "grad_norm": 0.1632416993379593, "learning_rate": 5.267899110894852e-05, "loss": 0.0145, "step": 50880 }, { "grad_norm": 0.16089430451393127, "learning_rate": 5.266248001552955e-05, "loss": 0.0126, "step": 50890 }, { "grad_norm": 0.19434140622615814, "learning_rate": 5.264596863094575e-05, "loss": 0.0122, "step": 50900 }, { "grad_norm": 0.18716254830360413, "learning_rate": 5.262945695700282e-05, "loss": 0.017, "step": 50910 }, { "grad_norm": 0.14537686109542847, "learning_rate": 5.261294499550643e-05, "loss": 0.0127, "step": 50920 }, { "grad_norm": 0.1288740038871765, "learning_rate": 5.25964327482623e-05, "loss": 0.0194, "step": 50930 }, { "grad_norm": 0.15435945987701416, "learning_rate": 5.257992021707617e-05, "loss": 0.0169, "step": 50940 }, { "grad_norm": 0.21567021310329437, "learning_rate": 5.256340740375387e-05, "loss": 0.0197, "step": 50950 }, { "grad_norm": 0.15216004848480225, "learning_rate": 5.254689431010117e-05, "loss": 0.0137, "step": 50960 }, { "grad_norm": 0.16846638917922974, "learning_rate": 5.253038093792395e-05, "loss": 0.0145, "step": 50970 }, { "grad_norm": 0.14409475028514862, "learning_rate": 5.251386728902806e-05, "loss": 0.0138, "step": 50980 }, { "grad_norm": 0.17858724296092987, "learning_rate": 5.2497353365219446e-05, "loss": 0.0128, "step": 50990 }, { "grad_norm": 0.21090947091579437, "learning_rate": 5.2480839168304e-05, "loss": 0.0127, "step": 51000 }, { "grad_norm": 0.2020818293094635, "learning_rate": 5.246432470008773e-05, "loss": 0.0143, "step": 51010 }, { "grad_norm": 0.12875604629516602, "learning_rate": 5.2447809962376616e-05, "loss": 0.0117, "step": 51020 }, { "grad_norm": 0.13307249546051025, "learning_rate": 5.243129495697666e-05, "loss": 0.0141, "step": 51030 }, { "grad_norm": 0.19404153525829315, "learning_rate": 5.241477968569397e-05, "loss": 0.0178, "step": 51040 }, { "grad_norm": 0.1987079679965973, "learning_rate": 5.23982641503346e-05, "loss": 0.0142, "step": 51050 }, { "grad_norm": 0.136204332113266, "learning_rate": 5.238174835270465e-05, "loss": 0.0178, "step": 51060 }, { "grad_norm": 0.6674619913101196, "learning_rate": 5.23652322946103e-05, "loss": 0.012, "step": 51070 }, { "grad_norm": 0.20018377900123596, "learning_rate": 5.234871597785771e-05, "loss": 0.0155, "step": 51080 }, { "grad_norm": 0.16655239462852478, "learning_rate": 5.233219940425306e-05, "loss": 0.0151, "step": 51090 }, { "grad_norm": 0.15271911025047302, "learning_rate": 5.2315682575602576e-05, "loss": 0.0114, "step": 51100 }, { "grad_norm": 0.18698750436306, "learning_rate": 5.229916549371254e-05, "loss": 0.0139, "step": 51110 }, { "grad_norm": 0.15727585554122925, "learning_rate": 5.228264816038924e-05, "loss": 0.0093, "step": 51120 }, { "grad_norm": 0.1738850325345993, "learning_rate": 5.2266130577438935e-05, "loss": 0.0136, "step": 51130 }, { "grad_norm": 0.18788810074329376, "learning_rate": 5.224961274666802e-05, "loss": 0.011, "step": 51140 }, { "grad_norm": 0.14454837143421173, "learning_rate": 5.223309466988283e-05, "loss": 0.0135, "step": 51150 }, { "grad_norm": 0.22097942233085632, "learning_rate": 5.221657634888976e-05, "loss": 0.0151, "step": 51160 }, { "grad_norm": 0.15041786432266235, "learning_rate": 5.2200057785495235e-05, "loss": 0.0122, "step": 51170 }, { "grad_norm": 0.18125973641872406, "learning_rate": 5.2183538981505706e-05, "loss": 0.0189, "step": 51180 }, { "grad_norm": 0.17351388931274414, "learning_rate": 5.216701993872762e-05, "loss": 0.0154, "step": 51190 }, { "grad_norm": 0.1275700032711029, "learning_rate": 5.215050065896752e-05, "loss": 0.0127, "step": 51200 }, { "grad_norm": 0.19164781272411346, "learning_rate": 5.2133981144031875e-05, "loss": 0.0127, "step": 51210 }, { "grad_norm": 0.1069217324256897, "learning_rate": 5.211746139572727e-05, "loss": 0.0139, "step": 51220 }, { "grad_norm": 0.20505952835083008, "learning_rate": 5.210094141586026e-05, "loss": 0.0142, "step": 51230 }, { "grad_norm": 0.15621478855609894, "learning_rate": 5.2084421206237476e-05, "loss": 0.0126, "step": 51240 }, { "grad_norm": 0.15345966815948486, "learning_rate": 5.20679007686655e-05, "loss": 0.0137, "step": 51250 }, { "grad_norm": 0.15599702298641205, "learning_rate": 5.205138010495102e-05, "loss": 0.0145, "step": 51260 }, { "grad_norm": 0.20080919563770294, "learning_rate": 5.20348592169007e-05, "loss": 0.0134, "step": 51270 }, { "grad_norm": 0.164631187915802, "learning_rate": 5.201833810632123e-05, "loss": 0.0125, "step": 51280 }, { "grad_norm": 0.20123939216136932, "learning_rate": 5.200181677501933e-05, "loss": 0.0111, "step": 51290 }, { "grad_norm": 0.21280857920646667, "learning_rate": 5.1985295224801777e-05, "loss": 0.0144, "step": 51300 }, { "grad_norm": 0.13181915879249573, "learning_rate": 5.196877345747531e-05, "loss": 0.0146, "step": 51310 }, { "grad_norm": 0.18763940036296844, "learning_rate": 5.195225147484674e-05, "loss": 0.0123, "step": 51320 }, { "grad_norm": 0.1910463124513626, "learning_rate": 5.193572927872288e-05, "loss": 0.019, "step": 51330 }, { "grad_norm": 0.1620909869670868, "learning_rate": 5.191920687091059e-05, "loss": 0.0161, "step": 51340 }, { "grad_norm": 0.14672300219535828, "learning_rate": 5.1902684253216695e-05, "loss": 0.0153, "step": 51350 }, { "grad_norm": 0.159402534365654, "learning_rate": 5.1886161427448134e-05, "loss": 0.016, "step": 51360 }, { "grad_norm": 0.18112437427043915, "learning_rate": 5.186963839541179e-05, "loss": 0.0111, "step": 51370 }, { "grad_norm": 0.12840020656585693, "learning_rate": 5.185311515891459e-05, "loss": 0.0106, "step": 51380 }, { "grad_norm": 0.19984333217144012, "learning_rate": 5.1836591719763516e-05, "loss": 0.0153, "step": 51390 }, { "grad_norm": 0.1486898958683014, "learning_rate": 5.182006807976554e-05, "loss": 0.0125, "step": 51400 }, { "grad_norm": 0.12621250748634338, "learning_rate": 5.180354424072763e-05, "loss": 0.0133, "step": 51410 }, { "grad_norm": 0.16130435466766357, "learning_rate": 5.178702020445686e-05, "loss": 0.0161, "step": 51420 }, { "grad_norm": 0.14377419650554657, "learning_rate": 5.177049597276024e-05, "loss": 0.0101, "step": 51430 }, { "grad_norm": 0.14967583119869232, "learning_rate": 5.175397154744482e-05, "loss": 0.0121, "step": 51440 }, { "grad_norm": 0.22146500647068024, "learning_rate": 5.1737446930317736e-05, "loss": 0.0144, "step": 51450 }, { "grad_norm": 0.1450432538986206, "learning_rate": 5.172092212318604e-05, "loss": 0.0124, "step": 51460 }, { "grad_norm": 0.17749230563640594, "learning_rate": 5.170439712785692e-05, "loss": 0.0131, "step": 51470 }, { "grad_norm": 0.23007427155971527, "learning_rate": 5.1687871946137467e-05, "loss": 0.0142, "step": 51480 }, { "grad_norm": 0.17124241590499878, "learning_rate": 5.16713465798349e-05, "loss": 0.0108, "step": 51490 }, { "grad_norm": 0.1845405250787735, "learning_rate": 5.165482103075637e-05, "loss": 0.0146, "step": 51500 }, { "grad_norm": 0.23027703166007996, "learning_rate": 5.163829530070909e-05, "loss": 0.0159, "step": 51510 }, { "grad_norm": 0.15882708132266998, "learning_rate": 5.162176939150032e-05, "loss": 0.0164, "step": 51520 }, { "grad_norm": 0.17155210673809052, "learning_rate": 5.160524330493728e-05, "loss": 0.0135, "step": 51530 }, { "grad_norm": 0.13734003901481628, "learning_rate": 5.158871704282724e-05, "loss": 0.0114, "step": 51540 }, { "grad_norm": 0.1768231838941574, "learning_rate": 5.15721906069775e-05, "loss": 0.0114, "step": 51550 }, { "grad_norm": 0.13615654408931732, "learning_rate": 5.155566399919536e-05, "loss": 0.0111, "step": 51560 }, { "grad_norm": 0.12827281653881073, "learning_rate": 5.153913722128813e-05, "loss": 0.011, "step": 51570 }, { "grad_norm": 0.17597393691539764, "learning_rate": 5.1522610275063176e-05, "loss": 0.0126, "step": 51580 }, { "grad_norm": 0.1622685194015503, "learning_rate": 5.150608316232786e-05, "loss": 0.0115, "step": 51590 }, { "grad_norm": 0.12995177507400513, "learning_rate": 5.148955588488955e-05, "loss": 0.0111, "step": 51600 }, { "grad_norm": 0.17633017897605896, "learning_rate": 5.147302844455564e-05, "loss": 0.0126, "step": 51610 }, { "grad_norm": 0.11877796053886414, "learning_rate": 5.145650084313356e-05, "loss": 0.0109, "step": 51620 }, { "grad_norm": 0.16426964104175568, "learning_rate": 5.1439973082430734e-05, "loss": 0.0103, "step": 51630 }, { "grad_norm": 0.20546075701713562, "learning_rate": 5.142344516425461e-05, "loss": 0.0158, "step": 51640 }, { "grad_norm": 0.1489783078432083, "learning_rate": 5.1406917090412673e-05, "loss": 0.0173, "step": 51650 }, { "grad_norm": 0.14833644032478333, "learning_rate": 5.1390388862712404e-05, "loss": 0.0132, "step": 51660 }, { "grad_norm": 0.19238583743572235, "learning_rate": 5.137386048296129e-05, "loss": 0.0151, "step": 51670 }, { "grad_norm": 0.15226031839847565, "learning_rate": 5.135733195296686e-05, "loss": 0.0125, "step": 51680 }, { "grad_norm": 0.1297631710767746, "learning_rate": 5.134080327453664e-05, "loss": 0.0122, "step": 51690 }, { "grad_norm": 0.11342734843492508, "learning_rate": 5.1324274449478204e-05, "loss": 0.0142, "step": 51700 }, { "grad_norm": 0.34225478768348694, "learning_rate": 5.130774547959909e-05, "loss": 0.0187, "step": 51710 }, { "grad_norm": 0.21207301318645477, "learning_rate": 5.129121636670692e-05, "loss": 0.0152, "step": 51720 }, { "grad_norm": 0.14967045187950134, "learning_rate": 5.1274687112609246e-05, "loss": 0.0109, "step": 51730 }, { "grad_norm": 0.1737213283777237, "learning_rate": 5.125815771911372e-05, "loss": 0.0103, "step": 51740 }, { "grad_norm": 0.1130913496017456, "learning_rate": 5.1241628188027956e-05, "loss": 0.0149, "step": 51750 }, { "grad_norm": 0.10612550377845764, "learning_rate": 5.1225098521159585e-05, "loss": 0.0135, "step": 51760 }, { "grad_norm": 0.17902182042598724, "learning_rate": 5.12085687203163e-05, "loss": 0.0122, "step": 51770 }, { "grad_norm": 0.1205497533082962, "learning_rate": 5.119203878730575e-05, "loss": 0.0128, "step": 51780 }, { "grad_norm": 0.15385794639587402, "learning_rate": 5.117550872393564e-05, "loss": 0.0109, "step": 51790 }, { "grad_norm": 0.1627599596977234, "learning_rate": 5.1158978532013645e-05, "loss": 0.0111, "step": 51800 }, { "grad_norm": 0.1474589854478836, "learning_rate": 5.114244821334751e-05, "loss": 0.0128, "step": 51810 }, { "grad_norm": 0.23982703685760498, "learning_rate": 5.112591776974497e-05, "loss": 0.0182, "step": 51820 }, { "grad_norm": 0.16461658477783203, "learning_rate": 5.110938720301374e-05, "loss": 0.0149, "step": 51830 }, { "grad_norm": 0.17379674315452576, "learning_rate": 5.1092856514961595e-05, "loss": 0.0102, "step": 51840 }, { "grad_norm": 0.1820434033870697, "learning_rate": 5.107632570739631e-05, "loss": 0.0108, "step": 51850 }, { "grad_norm": 0.17581325769424438, "learning_rate": 5.1059794782125645e-05, "loss": 0.0114, "step": 51860 }, { "grad_norm": 0.10907145589590073, "learning_rate": 5.104326374095744e-05, "loss": 0.0103, "step": 51870 }, { "grad_norm": 0.17383639514446259, "learning_rate": 5.102673258569947e-05, "loss": 0.0119, "step": 51880 }, { "grad_norm": 0.18717901408672333, "learning_rate": 5.101020131815956e-05, "loss": 0.0138, "step": 51890 }, { "grad_norm": 0.19297295808792114, "learning_rate": 5.099366994014556e-05, "loss": 0.0115, "step": 51900 }, { "grad_norm": 0.21642808616161346, "learning_rate": 5.097713845346532e-05, "loss": 0.0096, "step": 51910 }, { "grad_norm": 0.14389637112617493, "learning_rate": 5.096060685992668e-05, "loss": 0.0108, "step": 51920 }, { "grad_norm": 0.13793297111988068, "learning_rate": 5.0944075161337514e-05, "loss": 0.0115, "step": 51930 }, { "grad_norm": 0.14513909816741943, "learning_rate": 5.09275433595057e-05, "loss": 0.0094, "step": 51940 }, { "grad_norm": 0.15570159256458282, "learning_rate": 5.0911011456239157e-05, "loss": 0.0128, "step": 51950 }, { "grad_norm": 0.15434664487838745, "learning_rate": 5.089447945334574e-05, "loss": 0.0097, "step": 51960 }, { "grad_norm": 0.20258846879005432, "learning_rate": 5.087794735263343e-05, "loss": 0.0136, "step": 51970 }, { "grad_norm": 0.19454139471054077, "learning_rate": 5.086141515591011e-05, "loss": 0.0121, "step": 51980 }, { "grad_norm": 0.1225835308432579, "learning_rate": 5.084488286498371e-05, "loss": 0.0114, "step": 51990 }, { "grad_norm": 0.15653784573078156, "learning_rate": 5.08283504816622e-05, "loss": 0.0106, "step": 52000 }, { "grad_norm": 0.1311642974615097, "learning_rate": 5.081181800775353e-05, "loss": 0.0133, "step": 52010 }, { "grad_norm": 0.21553176641464233, "learning_rate": 5.079528544506565e-05, "loss": 0.019, "step": 52020 }, { "grad_norm": 0.18619970977306366, "learning_rate": 5.077875279540657e-05, "loss": 0.0122, "step": 52030 }, { "grad_norm": 0.16534574329853058, "learning_rate": 5.076222006058424e-05, "loss": 0.0115, "step": 52040 }, { "grad_norm": 0.16235314309597015, "learning_rate": 5.074568724240667e-05, "loss": 0.013, "step": 52050 }, { "grad_norm": 0.16628186404705048, "learning_rate": 5.072915434268187e-05, "loss": 0.016, "step": 52060 }, { "grad_norm": 0.1913149356842041, "learning_rate": 5.071262136321785e-05, "loss": 0.0118, "step": 52070 }, { "grad_norm": 0.2581910490989685, "learning_rate": 5.069608830582262e-05, "loss": 0.0123, "step": 52080 }, { "grad_norm": 0.14504463970661163, "learning_rate": 5.067955517230423e-05, "loss": 0.0118, "step": 52090 }, { "grad_norm": 0.1790078580379486, "learning_rate": 5.06630219644707e-05, "loss": 0.0113, "step": 52100 }, { "grad_norm": 0.16177746653556824, "learning_rate": 5.064648868413008e-05, "loss": 0.0103, "step": 52110 }, { "grad_norm": 0.1647004783153534, "learning_rate": 5.062995533309044e-05, "loss": 0.0146, "step": 52120 }, { "grad_norm": 0.15096251666545868, "learning_rate": 5.061342191315983e-05, "loss": 0.0121, "step": 52130 }, { "grad_norm": 0.11093747615814209, "learning_rate": 5.0596888426146325e-05, "loss": 0.0128, "step": 52140 }, { "grad_norm": 0.13621936738491058, "learning_rate": 5.058035487385798e-05, "loss": 0.0153, "step": 52150 }, { "grad_norm": 0.20051363110542297, "learning_rate": 5.0563821258102915e-05, "loss": 0.0126, "step": 52160 }, { "grad_norm": 0.14898359775543213, "learning_rate": 5.05472875806892e-05, "loss": 0.0115, "step": 52170 }, { "grad_norm": 0.22490128874778748, "learning_rate": 5.053075384342494e-05, "loss": 0.0158, "step": 52180 }, { "grad_norm": 0.21715722978115082, "learning_rate": 5.0514220048118234e-05, "loss": 0.0143, "step": 52190 }, { "grad_norm": 0.19083702564239502, "learning_rate": 5.049768619657721e-05, "loss": 0.0115, "step": 52200 }, { "grad_norm": 0.25781092047691345, "learning_rate": 5.048115229060996e-05, "loss": 0.0151, "step": 52210 }, { "grad_norm": 0.18781092762947083, "learning_rate": 5.046461833202462e-05, "loss": 0.0128, "step": 52220 }, { "grad_norm": 0.10446106642484665, "learning_rate": 5.044808432262932e-05, "loss": 0.011, "step": 52230 }, { "grad_norm": 0.1583532989025116, "learning_rate": 5.043155026423219e-05, "loss": 0.0101, "step": 52240 }, { "grad_norm": 0.17115503549575806, "learning_rate": 5.041501615864137e-05, "loss": 0.0133, "step": 52250 }, { "grad_norm": 0.13878940045833588, "learning_rate": 5.039848200766502e-05, "loss": 0.0132, "step": 52260 }, { "grad_norm": 0.13231413066387177, "learning_rate": 5.038194781311125e-05, "loss": 0.0129, "step": 52270 }, { "grad_norm": 0.1688891053199768, "learning_rate": 5.0365413576788264e-05, "loss": 0.013, "step": 52280 }, { "grad_norm": 0.16021956503391266, "learning_rate": 5.0348879300504194e-05, "loss": 0.014, "step": 52290 }, { "grad_norm": 0.14242883026599884, "learning_rate": 5.033234498606721e-05, "loss": 0.0132, "step": 52300 }, { "grad_norm": 0.16126464307308197, "learning_rate": 5.0315810635285476e-05, "loss": 0.0123, "step": 52310 }, { "grad_norm": 0.19915108382701874, "learning_rate": 5.0299276249967164e-05, "loss": 0.0111, "step": 52320 }, { "grad_norm": 0.11839261651039124, "learning_rate": 5.0282741831920454e-05, "loss": 0.0148, "step": 52330 }, { "grad_norm": 0.2393546998500824, "learning_rate": 5.026620738295351e-05, "loss": 0.0133, "step": 52340 }, { "grad_norm": 0.14955301582813263, "learning_rate": 5.024967290487455e-05, "loss": 0.0112, "step": 52350 }, { "grad_norm": 0.10516903549432755, "learning_rate": 5.023313839949172e-05, "loss": 0.0088, "step": 52360 }, { "grad_norm": 0.15907399356365204, "learning_rate": 5.0216603868613224e-05, "loss": 0.0117, "step": 52370 }, { "grad_norm": 0.19379937648773193, "learning_rate": 5.020006931404726e-05, "loss": 0.012, "step": 52380 }, { "grad_norm": 0.16570161283016205, "learning_rate": 5.0183534737602014e-05, "loss": 0.0142, "step": 52390 }, { "grad_norm": 0.20043721795082092, "learning_rate": 5.0167000141085686e-05, "loss": 0.0123, "step": 52400 }, { "grad_norm": 0.1446424424648285, "learning_rate": 5.015046552630647e-05, "loss": 0.0138, "step": 52410 }, { "grad_norm": 0.18003445863723755, "learning_rate": 5.013393089507258e-05, "loss": 0.0148, "step": 52420 }, { "grad_norm": 0.1461237668991089, "learning_rate": 5.011739624919219e-05, "loss": 0.0128, "step": 52430 }, { "grad_norm": 0.16662372648715973, "learning_rate": 5.0100861590473545e-05, "loss": 0.0122, "step": 52440 }, { "grad_norm": 0.146488755941391, "learning_rate": 5.008432692072482e-05, "loss": 0.0183, "step": 52450 }, { "grad_norm": 0.19407697021961212, "learning_rate": 5.006779224175423e-05, "loss": 0.0146, "step": 52460 }, { "grad_norm": 0.16618512570858002, "learning_rate": 5.005125755536998e-05, "loss": 0.0127, "step": 52470 }, { "grad_norm": 0.29205554723739624, "learning_rate": 5.003472286338029e-05, "loss": 0.0118, "step": 52480 }, { "grad_norm": 0.13978764414787292, "learning_rate": 5.001818816759335e-05, "loss": 0.0116, "step": 52490 }, { "grad_norm": 0.16582046449184418, "learning_rate": 5.000165346981738e-05, "loss": 0.0121, "step": 52500 }, { "grad_norm": 0.16174787282943726, "learning_rate": 4.9985118771860595e-05, "loss": 0.0134, "step": 52510 }, { "grad_norm": 0.14508946239948273, "learning_rate": 4.996858407553119e-05, "loss": 0.0095, "step": 52520 }, { "grad_norm": 0.13249747455120087, "learning_rate": 4.99520493826374e-05, "loss": 0.0105, "step": 52530 }, { "grad_norm": 0.1807975322008133, "learning_rate": 4.9935514694987384e-05, "loss": 0.0115, "step": 52540 }, { "grad_norm": 0.14460188150405884, "learning_rate": 4.991898001438941e-05, "loss": 0.0098, "step": 52550 }, { "grad_norm": 0.20897389948368073, "learning_rate": 4.990244534265166e-05, "loss": 0.0092, "step": 52560 }, { "grad_norm": 0.1662822812795639, "learning_rate": 4.988591068158234e-05, "loss": 0.0109, "step": 52570 }, { "grad_norm": 0.15640586614608765, "learning_rate": 4.986937603298965e-05, "loss": 0.0105, "step": 52580 }, { "grad_norm": 0.1825564205646515, "learning_rate": 4.985284139868181e-05, "loss": 0.0127, "step": 52590 }, { "grad_norm": 0.11581560224294662, "learning_rate": 4.9836306780467e-05, "loss": 0.0131, "step": 52600 }, { "grad_norm": 0.1674671620130539, "learning_rate": 4.981977218015343e-05, "loss": 0.0103, "step": 52610 }, { "grad_norm": 0.13441458344459534, "learning_rate": 4.980323759954932e-05, "loss": 0.0095, "step": 52620 }, { "grad_norm": 0.14044424891471863, "learning_rate": 4.978670304046284e-05, "loss": 0.0116, "step": 52630 }, { "grad_norm": 0.15650399029254913, "learning_rate": 4.9770168504702195e-05, "loss": 0.0119, "step": 52640 }, { "grad_norm": 0.1352764368057251, "learning_rate": 4.975363399407559e-05, "loss": 0.0127, "step": 52650 }, { "grad_norm": 0.18554897606372833, "learning_rate": 4.9737099510391194e-05, "loss": 0.013, "step": 52660 }, { "grad_norm": 0.19108733534812927, "learning_rate": 4.972056505545719e-05, "loss": 0.0141, "step": 52670 }, { "grad_norm": 0.16016675531864166, "learning_rate": 4.9704030631081775e-05, "loss": 0.0136, "step": 52680 }, { "grad_norm": 0.16406139731407166, "learning_rate": 4.968749623907313e-05, "loss": 0.0178, "step": 52690 }, { "grad_norm": 0.13418380916118622, "learning_rate": 4.967096188123943e-05, "loss": 0.0127, "step": 52700 }, { "grad_norm": 0.1255991905927658, "learning_rate": 4.965442755938884e-05, "loss": 0.0105, "step": 52710 }, { "grad_norm": 0.18475542962551117, "learning_rate": 4.963789327532954e-05, "loss": 0.0143, "step": 52720 }, { "grad_norm": 0.1482749730348587, "learning_rate": 4.962135903086968e-05, "loss": 0.012, "step": 52730 }, { "grad_norm": 0.2594817578792572, "learning_rate": 4.960482482781741e-05, "loss": 0.0115, "step": 52740 }, { "grad_norm": 0.16308742761611938, "learning_rate": 4.958829066798092e-05, "loss": 0.013, "step": 52750 }, { "grad_norm": 0.153307244181633, "learning_rate": 4.957175655316835e-05, "loss": 0.016, "step": 52760 }, { "grad_norm": 0.1747117042541504, "learning_rate": 4.9555222485187834e-05, "loss": 0.0134, "step": 52770 }, { "grad_norm": 0.18852752447128296, "learning_rate": 4.953868846584752e-05, "loss": 0.0164, "step": 52780 }, { "grad_norm": 0.17912515997886658, "learning_rate": 4.9522154496955544e-05, "loss": 0.0124, "step": 52790 }, { "grad_norm": 0.16719180345535278, "learning_rate": 4.950562058032001e-05, "loss": 0.0158, "step": 52800 }, { "grad_norm": 0.38731274008750916, "learning_rate": 4.948908671774909e-05, "loss": 0.016, "step": 52810 }, { "grad_norm": 0.1587001234292984, "learning_rate": 4.947255291105087e-05, "loss": 0.0147, "step": 52820 }, { "grad_norm": 0.16755834221839905, "learning_rate": 4.945601916203348e-05, "loss": 0.0157, "step": 52830 }, { "grad_norm": 0.1514049470424652, "learning_rate": 4.9439485472505016e-05, "loss": 0.0124, "step": 52840 }, { "grad_norm": 0.18335551023483276, "learning_rate": 4.942295184427356e-05, "loss": 0.0128, "step": 52850 }, { "grad_norm": 0.2031712830066681, "learning_rate": 4.9406418279147234e-05, "loss": 0.0133, "step": 52860 }, { "grad_norm": 0.18694975972175598, "learning_rate": 4.9389884778934116e-05, "loss": 0.0127, "step": 52870 }, { "grad_norm": 0.1450231969356537, "learning_rate": 4.937335134544228e-05, "loss": 0.0136, "step": 52880 }, { "grad_norm": 0.220389723777771, "learning_rate": 4.935681798047979e-05, "loss": 0.0119, "step": 52890 }, { "grad_norm": 0.1567108929157257, "learning_rate": 4.934028468585473e-05, "loss": 0.0112, "step": 52900 }, { "grad_norm": 0.21131263673305511, "learning_rate": 4.932375146337514e-05, "loss": 0.0146, "step": 52910 }, { "grad_norm": 0.17379668354988098, "learning_rate": 4.930721831484908e-05, "loss": 0.0101, "step": 52920 }, { "grad_norm": 0.13543745875358582, "learning_rate": 4.929068524208456e-05, "loss": 0.0134, "step": 52930 }, { "grad_norm": 0.18895314633846283, "learning_rate": 4.927415224688965e-05, "loss": 0.0117, "step": 52940 }, { "grad_norm": 0.15149906277656555, "learning_rate": 4.9257619331072375e-05, "loss": 0.0135, "step": 52950 }, { "grad_norm": 0.26097866892814636, "learning_rate": 4.924108649644072e-05, "loss": 0.0133, "step": 52960 }, { "grad_norm": 0.458957701921463, "learning_rate": 4.922455374480271e-05, "loss": 0.0136, "step": 52970 }, { "grad_norm": 0.20017719268798828, "learning_rate": 4.920802107796634e-05, "loss": 0.0141, "step": 52980 }, { "grad_norm": 0.14626052975654602, "learning_rate": 4.9191488497739564e-05, "loss": 0.0114, "step": 52990 }, { "grad_norm": 0.23738087713718414, "learning_rate": 4.917495600593042e-05, "loss": 0.0136, "step": 53000 }, { "grad_norm": 0.23826877772808075, "learning_rate": 4.915842360434684e-05, "loss": 0.0135, "step": 53010 }, { "grad_norm": 0.13957329094409943, "learning_rate": 4.914189129479679e-05, "loss": 0.0129, "step": 53020 }, { "grad_norm": 0.16301439702510834, "learning_rate": 4.912535907908822e-05, "loss": 0.0127, "step": 53030 }, { "grad_norm": 0.19694660604000092, "learning_rate": 4.910882695902907e-05, "loss": 0.0135, "step": 53040 }, { "grad_norm": 0.16407854855060577, "learning_rate": 4.9092294936427234e-05, "loss": 0.0135, "step": 53050 }, { "grad_norm": 0.13865213096141815, "learning_rate": 4.9075763013090685e-05, "loss": 0.0132, "step": 53060 }, { "grad_norm": 0.22776077687740326, "learning_rate": 4.9059231190827305e-05, "loss": 0.0117, "step": 53070 }, { "grad_norm": 0.18003661930561066, "learning_rate": 4.904269947144499e-05, "loss": 0.0182, "step": 53080 }, { "grad_norm": 0.15484833717346191, "learning_rate": 4.902616785675161e-05, "loss": 0.0135, "step": 53090 }, { "grad_norm": 0.2082609087228775, "learning_rate": 4.900963634855505e-05, "loss": 0.0131, "step": 53100 }, { "grad_norm": 0.18875858187675476, "learning_rate": 4.8993104948663173e-05, "loss": 0.0129, "step": 53110 }, { "grad_norm": 0.17687126994132996, "learning_rate": 4.8976573658883823e-05, "loss": 0.0126, "step": 53120 }, { "grad_norm": 0.13921087980270386, "learning_rate": 4.896004248102484e-05, "loss": 0.0113, "step": 53130 }, { "grad_norm": 0.15015536546707153, "learning_rate": 4.8943511416894045e-05, "loss": 0.0172, "step": 53140 }, { "grad_norm": 0.15188796818256378, "learning_rate": 4.892698046829926e-05, "loss": 0.0144, "step": 53150 }, { "grad_norm": 0.12069714069366455, "learning_rate": 4.8910449637048275e-05, "loss": 0.0112, "step": 53160 }, { "grad_norm": 0.1192760095000267, "learning_rate": 4.8893918924948884e-05, "loss": 0.0123, "step": 53170 }, { "grad_norm": 0.12224193662405014, "learning_rate": 4.8877388333808835e-05, "loss": 0.0119, "step": 53180 }, { "grad_norm": 0.1782212108373642, "learning_rate": 4.8860857865435925e-05, "loss": 0.0094, "step": 53190 }, { "grad_norm": 0.16065560281276703, "learning_rate": 4.884432752163789e-05, "loss": 0.0147, "step": 53200 }, { "grad_norm": 0.161984384059906, "learning_rate": 4.8827797304222466e-05, "loss": 0.0147, "step": 53210 }, { "grad_norm": 0.20248156785964966, "learning_rate": 4.8811267214997366e-05, "loss": 0.014, "step": 53220 }, { "grad_norm": 0.17971622943878174, "learning_rate": 4.879473725577029e-05, "loss": 0.0167, "step": 53230 }, { "grad_norm": 0.16261951625347137, "learning_rate": 4.877820742834891e-05, "loss": 0.0105, "step": 53240 }, { "grad_norm": 0.19540473818778992, "learning_rate": 4.8761677734540956e-05, "loss": 0.0114, "step": 53250 }, { "grad_norm": 0.23690727353096008, "learning_rate": 4.874514817615407e-05, "loss": 0.0138, "step": 53260 }, { "grad_norm": 0.1718936413526535, "learning_rate": 4.872861875499588e-05, "loss": 0.01, "step": 53270 }, { "grad_norm": 0.1826881468296051, "learning_rate": 4.871208947287404e-05, "loss": 0.0138, "step": 53280 }, { "grad_norm": 0.17518040537834167, "learning_rate": 4.869556033159615e-05, "loss": 0.0103, "step": 53290 }, { "grad_norm": 0.17204362154006958, "learning_rate": 4.8679031332969835e-05, "loss": 0.0132, "step": 53300 }, { "grad_norm": 0.1425798237323761, "learning_rate": 4.8662502478802635e-05, "loss": 0.0126, "step": 53310 }, { "grad_norm": 0.13723918795585632, "learning_rate": 4.8645973770902176e-05, "loss": 0.0121, "step": 53320 }, { "grad_norm": 0.20868194103240967, "learning_rate": 4.862944521107599e-05, "loss": 0.0149, "step": 53330 }, { "grad_norm": 0.1653442084789276, "learning_rate": 4.861291680113162e-05, "loss": 0.0104, "step": 53340 }, { "grad_norm": 0.162943035364151, "learning_rate": 4.8596388542876563e-05, "loss": 0.0112, "step": 53350 }, { "grad_norm": 0.12013374269008636, "learning_rate": 4.857986043811836e-05, "loss": 0.0095, "step": 53360 }, { "grad_norm": 0.13303261995315552, "learning_rate": 4.856333248866448e-05, "loss": 0.0132, "step": 53370 }, { "grad_norm": 0.18999314308166504, "learning_rate": 4.85468046963224e-05, "loss": 0.0108, "step": 53380 }, { "grad_norm": 0.18513815104961395, "learning_rate": 4.853027706289957e-05, "loss": 0.0113, "step": 53390 }, { "grad_norm": 0.15674227476119995, "learning_rate": 4.851374959020344e-05, "loss": 0.0119, "step": 53400 }, { "grad_norm": 0.11562637239694595, "learning_rate": 4.849722228004141e-05, "loss": 0.0141, "step": 53410 }, { "grad_norm": 0.13172993063926697, "learning_rate": 4.848069513422091e-05, "loss": 0.0132, "step": 53420 }, { "grad_norm": 0.1182851567864418, "learning_rate": 4.8464168154549294e-05, "loss": 0.0108, "step": 53430 }, { "grad_norm": 0.16140839457511902, "learning_rate": 4.844764134283392e-05, "loss": 0.0106, "step": 53440 }, { "grad_norm": 0.1738448292016983, "learning_rate": 4.843111470088218e-05, "loss": 0.0124, "step": 53450 }, { "grad_norm": 0.1502045840024948, "learning_rate": 4.841458823050137e-05, "loss": 0.0101, "step": 53460 }, { "grad_norm": 0.23620635271072388, "learning_rate": 4.8398061933498816e-05, "loss": 0.0115, "step": 53470 }, { "grad_norm": 0.10124724358320236, "learning_rate": 4.83815358116818e-05, "loss": 0.0102, "step": 53480 }, { "grad_norm": 0.13269418478012085, "learning_rate": 4.8365009866857584e-05, "loss": 0.0117, "step": 53490 }, { "grad_norm": 0.23258040845394135, "learning_rate": 4.834848410083342e-05, "loss": 0.0113, "step": 53500 }, { "grad_norm": 0.1666388362646103, "learning_rate": 4.833195851541657e-05, "loss": 0.0121, "step": 53510 }, { "grad_norm": 0.15028540790081024, "learning_rate": 4.831543311241423e-05, "loss": 0.0144, "step": 53520 }, { "grad_norm": 0.12956443428993225, "learning_rate": 4.829890789363359e-05, "loss": 0.0111, "step": 53530 }, { "grad_norm": 0.20816238224506378, "learning_rate": 4.828238286088182e-05, "loss": 0.0123, "step": 53540 }, { "grad_norm": 0.2101919949054718, "learning_rate": 4.826585801596609e-05, "loss": 0.0149, "step": 53550 }, { "grad_norm": 0.2196059376001358, "learning_rate": 4.8249333360693486e-05, "loss": 0.0107, "step": 53560 }, { "grad_norm": 0.11773776262998581, "learning_rate": 4.8232808896871185e-05, "loss": 0.0107, "step": 53570 }, { "grad_norm": 0.14732301235198975, "learning_rate": 4.821628462630624e-05, "loss": 0.0098, "step": 53580 }, { "grad_norm": 0.14938268065452576, "learning_rate": 4.8199760550805735e-05, "loss": 0.0103, "step": 53590 }, { "grad_norm": 0.13438087701797485, "learning_rate": 4.818323667217669e-05, "loss": 0.0155, "step": 53600 }, { "grad_norm": 0.1662650853395462, "learning_rate": 4.8166712992226165e-05, "loss": 0.0129, "step": 53610 }, { "grad_norm": 0.1118052527308464, "learning_rate": 4.8150189512761154e-05, "loss": 0.0095, "step": 53620 }, { "grad_norm": 0.10752695053815842, "learning_rate": 4.813366623558862e-05, "loss": 0.0133, "step": 53630 }, { "grad_norm": 0.13198073208332062, "learning_rate": 4.811714316251554e-05, "loss": 0.0109, "step": 53640 }, { "grad_norm": 0.15175989270210266, "learning_rate": 4.810062029534886e-05, "loss": 0.0139, "step": 53650 }, { "grad_norm": 0.19621962308883667, "learning_rate": 4.8084097635895505e-05, "loss": 0.0144, "step": 53660 }, { "grad_norm": 0.22477015852928162, "learning_rate": 4.8067575185962335e-05, "loss": 0.0125, "step": 53670 }, { "grad_norm": 0.1752595752477646, "learning_rate": 4.805105294735625e-05, "loss": 0.013, "step": 53680 }, { "grad_norm": 0.13437259197235107, "learning_rate": 4.803453092188406e-05, "loss": 0.0104, "step": 53690 }, { "grad_norm": 0.20005352795124054, "learning_rate": 4.801800911135263e-05, "loss": 0.0129, "step": 53700 }, { "grad_norm": 0.12242551892995834, "learning_rate": 4.8001487517568744e-05, "loss": 0.0116, "step": 53710 }, { "grad_norm": 0.2351740449666977, "learning_rate": 4.798496614233918e-05, "loss": 0.0122, "step": 53720 }, { "grad_norm": 0.2635844051837921, "learning_rate": 4.796844498747069e-05, "loss": 0.0158, "step": 53730 }, { "grad_norm": 0.18470996618270874, "learning_rate": 4.795192405477e-05, "loss": 0.0109, "step": 53740 }, { "grad_norm": 0.2288573831319809, "learning_rate": 4.79354033460438e-05, "loss": 0.0166, "step": 53750 }, { "grad_norm": 0.19290153682231903, "learning_rate": 4.79188828630988e-05, "loss": 0.0126, "step": 53760 }, { "grad_norm": 0.18921014666557312, "learning_rate": 4.790236260774165e-05, "loss": 0.0104, "step": 53770 }, { "grad_norm": 0.17204256355762482, "learning_rate": 4.788584258177896e-05, "loss": 0.0176, "step": 53780 }, { "grad_norm": 0.16404010355472565, "learning_rate": 4.7869322787017356e-05, "loss": 0.0106, "step": 53790 }, { "grad_norm": 0.13694927096366882, "learning_rate": 4.78528032252634e-05, "loss": 0.0142, "step": 53800 }, { "grad_norm": 0.15716329216957092, "learning_rate": 4.783628389832366e-05, "loss": 0.0103, "step": 53810 }, { "grad_norm": 0.17081625759601593, "learning_rate": 4.7819764808004644e-05, "loss": 0.0154, "step": 53820 }, { "grad_norm": 0.16080830991268158, "learning_rate": 4.780324595611289e-05, "loss": 0.0127, "step": 53830 }, { "grad_norm": 0.18634247779846191, "learning_rate": 4.7786727344454854e-05, "loss": 0.0121, "step": 53840 }, { "grad_norm": 0.17137005925178528, "learning_rate": 4.777020897483697e-05, "loss": 0.0126, "step": 53850 }, { "grad_norm": 0.20622402429580688, "learning_rate": 4.775369084906569e-05, "loss": 0.0135, "step": 53860 }, { "grad_norm": 0.1962958127260208, "learning_rate": 4.773717296894742e-05, "loss": 0.0126, "step": 53870 }, { "grad_norm": 0.20923848450183868, "learning_rate": 4.772065533628848e-05, "loss": 0.0141, "step": 53880 }, { "grad_norm": 0.15347521007061005, "learning_rate": 4.7704137952895265e-05, "loss": 0.0154, "step": 53890 }, { "grad_norm": 0.12674827873706818, "learning_rate": 4.768762082057407e-05, "loss": 0.0137, "step": 53900 }, { "grad_norm": 0.13939520716667175, "learning_rate": 4.7671103941131186e-05, "loss": 0.0124, "step": 53910 }, { "grad_norm": 0.1905445009469986, "learning_rate": 4.765458731637287e-05, "loss": 0.0167, "step": 53920 }, { "grad_norm": 0.17117953300476074, "learning_rate": 4.763807094810537e-05, "loss": 0.0108, "step": 53930 }, { "grad_norm": 0.23695829510688782, "learning_rate": 4.762155483813487e-05, "loss": 0.0137, "step": 53940 }, { "grad_norm": 0.19108977913856506, "learning_rate": 4.760503898826754e-05, "loss": 0.0128, "step": 53950 }, { "grad_norm": 0.19309645891189575, "learning_rate": 4.758852340030957e-05, "loss": 0.0131, "step": 53960 }, { "grad_norm": 0.17325985431671143, "learning_rate": 4.757200807606705e-05, "loss": 0.0126, "step": 53970 }, { "grad_norm": 0.1391587108373642, "learning_rate": 4.755549301734607e-05, "loss": 0.0131, "step": 53980 }, { "grad_norm": 0.1566339135169983, "learning_rate": 4.753897822595271e-05, "loss": 0.0111, "step": 53990 }, { "grad_norm": 0.21259278059005737, "learning_rate": 4.7522463703692994e-05, "loss": 0.0114, "step": 54000 }, { "grad_norm": 0.15865103900432587, "learning_rate": 4.75059494523729e-05, "loss": 0.012, "step": 54010 }, { "grad_norm": 0.14496968686580658, "learning_rate": 4.748943547379843e-05, "loss": 0.0134, "step": 54020 }, { "grad_norm": 0.14634037017822266, "learning_rate": 4.7472921769775535e-05, "loss": 0.0128, "step": 54030 }, { "grad_norm": 0.14685124158859253, "learning_rate": 4.745640834211011e-05, "loss": 0.0128, "step": 54040 }, { "grad_norm": 0.1413058638572693, "learning_rate": 4.7439895192608054e-05, "loss": 0.0111, "step": 54050 }, { "grad_norm": 0.13847897946834564, "learning_rate": 4.7423382323075186e-05, "loss": 0.0117, "step": 54060 }, { "grad_norm": 0.1857682317495346, "learning_rate": 4.740686973531737e-05, "loss": 0.0102, "step": 54070 }, { "grad_norm": 0.17741252481937408, "learning_rate": 4.7390357431140375e-05, "loss": 0.012, "step": 54080 }, { "grad_norm": 0.18002909421920776, "learning_rate": 4.737384541234998e-05, "loss": 0.0098, "step": 54090 }, { "grad_norm": 0.4544377624988556, "learning_rate": 4.735733368075189e-05, "loss": 0.0122, "step": 54100 }, { "grad_norm": 0.20162765681743622, "learning_rate": 4.734082223815182e-05, "loss": 0.0142, "step": 54110 }, { "grad_norm": 0.13200367987155914, "learning_rate": 4.7324311086355436e-05, "loss": 0.0143, "step": 54120 }, { "grad_norm": 0.16016609966754913, "learning_rate": 4.730780022716836e-05, "loss": 0.0169, "step": 54130 }, { "grad_norm": 0.2114652693271637, "learning_rate": 4.729128966239619e-05, "loss": 0.009, "step": 54140 }, { "grad_norm": 0.17091761529445648, "learning_rate": 4.7274779393844536e-05, "loss": 0.0121, "step": 54150 }, { "grad_norm": 0.1655440628528595, "learning_rate": 4.725826942331891e-05, "loss": 0.0138, "step": 54160 }, { "grad_norm": 0.13722491264343262, "learning_rate": 4.724175975262481e-05, "loss": 0.011, "step": 54170 }, { "grad_norm": 0.16097091138362885, "learning_rate": 4.722525038356772e-05, "loss": 0.0089, "step": 54180 }, { "grad_norm": 0.14090104401111603, "learning_rate": 4.7208741317953075e-05, "loss": 0.0105, "step": 54190 }, { "grad_norm": 0.30451610684394836, "learning_rate": 4.7192232557586264e-05, "loss": 0.0129, "step": 54200 }, { "grad_norm": 0.18386003375053406, "learning_rate": 4.717572410427269e-05, "loss": 0.0109, "step": 54210 }, { "grad_norm": 0.17226354777812958, "learning_rate": 4.71592159598177e-05, "loss": 0.0113, "step": 54220 }, { "grad_norm": 0.16496409475803375, "learning_rate": 4.714270812602657e-05, "loss": 0.0117, "step": 54230 }, { "grad_norm": 0.18222938477993011, "learning_rate": 4.712620060470458e-05, "loss": 0.0118, "step": 54240 }, { "grad_norm": 0.18986980617046356, "learning_rate": 4.7109693397656986e-05, "loss": 0.013, "step": 54250 }, { "grad_norm": 0.19677147269248962, "learning_rate": 4.709318650668894e-05, "loss": 0.0144, "step": 54260 }, { "grad_norm": 0.1521080732345581, "learning_rate": 4.7076679933605676e-05, "loss": 0.0114, "step": 54270 }, { "grad_norm": 0.1806333214044571, "learning_rate": 4.70601736802123e-05, "loss": 0.0094, "step": 54280 }, { "grad_norm": 0.18829800188541412, "learning_rate": 4.704366774831391e-05, "loss": 0.0182, "step": 54290 }, { "grad_norm": 0.1548602133989334, "learning_rate": 4.702716213971557e-05, "loss": 0.0097, "step": 54300 }, { "grad_norm": 0.1886722594499588, "learning_rate": 4.70106568562223e-05, "loss": 0.0166, "step": 54310 }, { "grad_norm": 0.13961748778820038, "learning_rate": 4.6994151899639105e-05, "loss": 0.0108, "step": 54320 }, { "grad_norm": 0.19624044001102448, "learning_rate": 4.697764727177093e-05, "loss": 0.014, "step": 54330 }, { "grad_norm": 0.1615694761276245, "learning_rate": 4.696114297442272e-05, "loss": 0.0145, "step": 54340 }, { "grad_norm": 0.18809638917446136, "learning_rate": 4.694463900939933e-05, "loss": 0.0161, "step": 54350 }, { "grad_norm": 0.11306034028530121, "learning_rate": 4.692813537850564e-05, "loss": 0.0077, "step": 54360 }, { "grad_norm": 0.11504241824150085, "learning_rate": 4.6911632083546445e-05, "loss": 0.0135, "step": 54370 }, { "grad_norm": 0.11182238161563873, "learning_rate": 4.6895129126326525e-05, "loss": 0.0143, "step": 54380 }, { "grad_norm": 0.37002384662628174, "learning_rate": 4.687862650865059e-05, "loss": 0.0125, "step": 54390 }, { "grad_norm": 0.21231049299240112, "learning_rate": 4.686212423232339e-05, "loss": 0.0104, "step": 54400 }, { "grad_norm": 0.20767198503017426, "learning_rate": 4.6845622299149565e-05, "loss": 0.0129, "step": 54410 }, { "grad_norm": 0.16089721024036407, "learning_rate": 4.682912071093374e-05, "loss": 0.0103, "step": 54420 }, { "grad_norm": 0.18646186590194702, "learning_rate": 4.681261946948052e-05, "loss": 0.008, "step": 54430 }, { "grad_norm": 0.309730589389801, "learning_rate": 4.679611857659443e-05, "loss": 0.0149, "step": 54440 }, { "grad_norm": 0.1569276601076126, "learning_rate": 4.6779618034079975e-05, "loss": 0.0096, "step": 54450 }, { "grad_norm": 0.17666001617908478, "learning_rate": 4.6763117843741675e-05, "loss": 0.011, "step": 54460 }, { "grad_norm": 0.355630487203598, "learning_rate": 4.674661800738394e-05, "loss": 0.0116, "step": 54470 }, { "grad_norm": 0.1340467780828476, "learning_rate": 4.673011852681117e-05, "loss": 0.0109, "step": 54480 }, { "grad_norm": 0.18980185687541962, "learning_rate": 4.671361940382772e-05, "loss": 0.0113, "step": 54490 }, { "grad_norm": 0.11974900960922241, "learning_rate": 4.669712064023791e-05, "loss": 0.0113, "step": 54500 }, { "grad_norm": 0.16840490698814392, "learning_rate": 4.668062223784602e-05, "loss": 0.0113, "step": 54510 }, { "grad_norm": 0.12152836471796036, "learning_rate": 4.6664124198456274e-05, "loss": 0.0115, "step": 54520 }, { "grad_norm": 0.18825778365135193, "learning_rate": 4.6647626523872905e-05, "loss": 0.0107, "step": 54530 }, { "grad_norm": 0.1491607427597046, "learning_rate": 4.6631129215900066e-05, "loss": 0.0096, "step": 54540 }, { "grad_norm": 0.20457947254180908, "learning_rate": 4.661463227634187e-05, "loss": 0.0128, "step": 54550 }, { "grad_norm": 0.14755879342556, "learning_rate": 4.659813570700238e-05, "loss": 0.0089, "step": 54560 }, { "grad_norm": 0.10384491831064224, "learning_rate": 4.658163950968568e-05, "loss": 0.0107, "step": 54570 }, { "grad_norm": 0.19635909795761108, "learning_rate": 4.6565143686195726e-05, "loss": 0.0135, "step": 54580 }, { "grad_norm": 0.21360215544700623, "learning_rate": 4.654864823833649e-05, "loss": 0.0126, "step": 54590 }, { "grad_norm": 0.2297411561012268, "learning_rate": 4.65321531679119e-05, "loss": 0.0145, "step": 54600 }, { "grad_norm": 0.1584673374891281, "learning_rate": 4.6515658476725834e-05, "loss": 0.0151, "step": 54610 }, { "grad_norm": 0.18354777991771698, "learning_rate": 4.649916416658212e-05, "loss": 0.0134, "step": 54620 }, { "grad_norm": 0.14968380331993103, "learning_rate": 4.648267023928455e-05, "loss": 0.0138, "step": 54630 }, { "grad_norm": 0.16239726543426514, "learning_rate": 4.646617669663687e-05, "loss": 0.0146, "step": 54640 }, { "grad_norm": 0.12854789197444916, "learning_rate": 4.644968354044277e-05, "loss": 0.0102, "step": 54650 }, { "grad_norm": 0.16309796273708344, "learning_rate": 4.643319077250596e-05, "loss": 0.0117, "step": 54660 }, { "grad_norm": 0.17174005508422852, "learning_rate": 4.6416698394630055e-05, "loss": 0.01, "step": 54670 }, { "grad_norm": 0.7184034585952759, "learning_rate": 4.640020640861862e-05, "loss": 0.0152, "step": 54680 }, { "grad_norm": 0.1856880486011505, "learning_rate": 4.6383714816275195e-05, "loss": 0.0126, "step": 54690 }, { "grad_norm": 0.13287270069122314, "learning_rate": 4.636722361940329e-05, "loss": 0.0124, "step": 54700 }, { "grad_norm": 0.1311684250831604, "learning_rate": 4.635073281980632e-05, "loss": 0.0122, "step": 54710 }, { "grad_norm": 0.18359045684337616, "learning_rate": 4.633424241928775e-05, "loss": 0.0101, "step": 54720 }, { "grad_norm": 0.13147982954978943, "learning_rate": 4.6317752419650914e-05, "loss": 0.0129, "step": 54730 }, { "grad_norm": 0.17008014023303986, "learning_rate": 4.630126282269914e-05, "loss": 0.0132, "step": 54740 }, { "grad_norm": 0.1708826720714569, "learning_rate": 4.628477363023569e-05, "loss": 0.0116, "step": 54750 }, { "grad_norm": 0.22676271200180054, "learning_rate": 4.626828484406383e-05, "loss": 0.0146, "step": 54760 }, { "grad_norm": 0.18895752727985382, "learning_rate": 4.625179646598669e-05, "loss": 0.0117, "step": 54770 }, { "grad_norm": 0.10900531709194183, "learning_rate": 4.6235308497807487e-05, "loss": 0.0113, "step": 54780 }, { "grad_norm": 0.13358955085277557, "learning_rate": 4.6218820941329276e-05, "loss": 0.0106, "step": 54790 }, { "grad_norm": 0.16698230803012848, "learning_rate": 4.620233379835513e-05, "loss": 0.0119, "step": 54800 }, { "grad_norm": 0.1590796411037445, "learning_rate": 4.618584707068804e-05, "loss": 0.0106, "step": 54810 }, { "grad_norm": 0.12212841212749481, "learning_rate": 4.616936076013099e-05, "loss": 0.0125, "step": 54820 }, { "grad_norm": 0.15242621302604675, "learning_rate": 4.615287486848688e-05, "loss": 0.012, "step": 54830 }, { "grad_norm": 0.12413479387760162, "learning_rate": 4.6136389397558584e-05, "loss": 0.0128, "step": 54840 }, { "grad_norm": 0.2167114019393921, "learning_rate": 4.6119904349148935e-05, "loss": 0.0096, "step": 54850 }, { "grad_norm": 0.15991906821727753, "learning_rate": 4.6103419725060726e-05, "loss": 0.0173, "step": 54860 }, { "grad_norm": 0.190608948469162, "learning_rate": 4.608693552709667e-05, "loss": 0.0143, "step": 54870 }, { "grad_norm": 0.2627006471157074, "learning_rate": 4.607045175705947e-05, "loss": 0.0161, "step": 54880 }, { "grad_norm": 0.20099253952503204, "learning_rate": 4.605396841675176e-05, "loss": 0.0139, "step": 54890 }, { "grad_norm": 0.23356063663959503, "learning_rate": 4.6037485507976094e-05, "loss": 0.0151, "step": 54900 }, { "grad_norm": 0.15642234683036804, "learning_rate": 4.602100303253509e-05, "loss": 0.0111, "step": 54910 }, { "grad_norm": 0.21618478000164032, "learning_rate": 4.600452099223122e-05, "loss": 0.0131, "step": 54920 }, { "grad_norm": 0.21063943207263947, "learning_rate": 4.598803938886691e-05, "loss": 0.0144, "step": 54930 }, { "grad_norm": 0.1334783136844635, "learning_rate": 4.597155822424459e-05, "loss": 0.0139, "step": 54940 }, { "grad_norm": 0.16515520215034485, "learning_rate": 4.59550775001666e-05, "loss": 0.0122, "step": 54950 }, { "grad_norm": 0.14103472232818604, "learning_rate": 4.5938597218435234e-05, "loss": 0.0116, "step": 54960 }, { "grad_norm": 0.15471675992012024, "learning_rate": 4.592211738085279e-05, "loss": 0.01, "step": 54970 }, { "grad_norm": 0.100080206990242, "learning_rate": 4.590563798922146e-05, "loss": 0.0099, "step": 54980 }, { "grad_norm": 0.18714670836925507, "learning_rate": 4.5889159045343404e-05, "loss": 0.012, "step": 54990 }, { "grad_norm": 0.12070835381746292, "learning_rate": 4.5872680551020734e-05, "loss": 0.011, "step": 55000 }, { "grad_norm": 0.15203271806240082, "learning_rate": 4.585620250805552e-05, "loss": 0.0134, "step": 55010 }, { "grad_norm": 0.16758669912815094, "learning_rate": 4.583972491824975e-05, "loss": 0.0101, "step": 55020 }, { "grad_norm": 0.14660826325416565, "learning_rate": 4.582324778340541e-05, "loss": 0.0136, "step": 55030 }, { "grad_norm": 0.11981703341007233, "learning_rate": 4.5806771105324424e-05, "loss": 0.0118, "step": 55040 }, { "grad_norm": 0.13987930119037628, "learning_rate": 4.579029488580865e-05, "loss": 0.0102, "step": 55050 }, { "grad_norm": 0.21763209998607635, "learning_rate": 4.5773819126659876e-05, "loss": 0.0114, "step": 55060 }, { "grad_norm": 0.1718338429927826, "learning_rate": 4.575734382967991e-05, "loss": 0.012, "step": 55070 }, { "grad_norm": 0.2548646628856659, "learning_rate": 4.574086899667044e-05, "loss": 0.0144, "step": 55080 }, { "grad_norm": 0.17631646990776062, "learning_rate": 4.5724394629433106e-05, "loss": 0.0114, "step": 55090 }, { "grad_norm": 0.12551461160182953, "learning_rate": 4.570792072976958e-05, "loss": 0.0122, "step": 55100 }, { "grad_norm": 0.1312611997127533, "learning_rate": 4.569144729948138e-05, "loss": 0.0103, "step": 55110 }, { "grad_norm": 0.13045091927051544, "learning_rate": 4.5674974340370033e-05, "loss": 0.0097, "step": 55120 }, { "grad_norm": 0.2293737679719925, "learning_rate": 4.565850185423699e-05, "loss": 0.0127, "step": 55130 }, { "grad_norm": 0.2361847460269928, "learning_rate": 4.564202984288365e-05, "loss": 0.0115, "step": 55140 }, { "grad_norm": 0.23394381999969482, "learning_rate": 4.5625558308111356e-05, "loss": 0.0109, "step": 55150 }, { "grad_norm": 0.16983892023563385, "learning_rate": 4.560908725172145e-05, "loss": 0.0127, "step": 55160 }, { "grad_norm": 0.13102656602859497, "learning_rate": 4.559261667551516e-05, "loss": 0.0122, "step": 55170 }, { "grad_norm": 0.17493456602096558, "learning_rate": 4.5576146581293685e-05, "loss": 0.0141, "step": 55180 }, { "grad_norm": 0.1465204805135727, "learning_rate": 4.5559676970858164e-05, "loss": 0.0108, "step": 55190 }, { "grad_norm": 0.12182027101516724, "learning_rate": 4.5543207846009705e-05, "loss": 0.0116, "step": 55200 }, { "grad_norm": 0.2141096591949463, "learning_rate": 4.552673920854933e-05, "loss": 0.0156, "step": 55210 }, { "grad_norm": 0.16299842298030853, "learning_rate": 4.5510271060278006e-05, "loss": 0.0106, "step": 55220 }, { "grad_norm": 0.17037542164325714, "learning_rate": 4.5493803402996704e-05, "loss": 0.0143, "step": 55230 }, { "grad_norm": 0.16624942421913147, "learning_rate": 4.5477336238506295e-05, "loss": 0.0158, "step": 55240 }, { "grad_norm": 0.15756964683532715, "learning_rate": 4.546086956860759e-05, "loss": 0.0114, "step": 55250 }, { "grad_norm": 0.2554906904697418, "learning_rate": 4.5444403395101374e-05, "loss": 0.0137, "step": 55260 }, { "grad_norm": 0.1809055656194687, "learning_rate": 4.542793771978834e-05, "loss": 0.0139, "step": 55270 }, { "grad_norm": 0.24698105454444885, "learning_rate": 4.541147254446916e-05, "loss": 0.0125, "step": 55280 }, { "grad_norm": 0.1907879114151001, "learning_rate": 4.539500787094445e-05, "loss": 0.014, "step": 55290 }, { "grad_norm": 0.27903786301612854, "learning_rate": 4.537854370101475e-05, "loss": 0.0144, "step": 55300 }, { "grad_norm": 0.19168180227279663, "learning_rate": 4.5362080036480566e-05, "loss": 0.013, "step": 55310 }, { "grad_norm": 0.28309062123298645, "learning_rate": 4.534561687914235e-05, "loss": 0.0136, "step": 55320 }, { "grad_norm": 0.16675612330436707, "learning_rate": 4.5329154230800466e-05, "loss": 0.0117, "step": 55330 }, { "grad_norm": 0.19183629751205444, "learning_rate": 4.531269209325525e-05, "loss": 0.0109, "step": 55340 }, { "grad_norm": 0.19313554465770721, "learning_rate": 4.529623046830696e-05, "loss": 0.0111, "step": 55350 }, { "grad_norm": 0.1583234965801239, "learning_rate": 4.5279769357755866e-05, "loss": 0.0108, "step": 55360 }, { "grad_norm": 0.15683449804782867, "learning_rate": 4.5263308763402084e-05, "loss": 0.0136, "step": 55370 }, { "grad_norm": 0.1417814940214157, "learning_rate": 4.524684868704574e-05, "loss": 0.0104, "step": 55380 }, { "grad_norm": 0.17282995581626892, "learning_rate": 4.523038913048688e-05, "loss": 0.0144, "step": 55390 }, { "grad_norm": 0.35760119557380676, "learning_rate": 4.521393009552549e-05, "loss": 0.0134, "step": 55400 }, { "grad_norm": 0.14454704523086548, "learning_rate": 4.5197471583961484e-05, "loss": 0.0127, "step": 55410 }, { "grad_norm": 0.1804657280445099, "learning_rate": 4.51810135975948e-05, "loss": 0.0117, "step": 55420 }, { "grad_norm": 0.5533601641654968, "learning_rate": 4.516455613822521e-05, "loss": 0.0107, "step": 55430 }, { "grad_norm": 0.15475055575370789, "learning_rate": 4.514809920765249e-05, "loss": 0.0108, "step": 55440 }, { "grad_norm": 0.1844698041677475, "learning_rate": 4.5131642807676356e-05, "loss": 0.011, "step": 55450 }, { "grad_norm": 0.10615742206573486, "learning_rate": 4.511518694009644e-05, "loss": 0.0109, "step": 55460 }, { "grad_norm": 0.15939386188983917, "learning_rate": 4.509873160671231e-05, "loss": 0.0104, "step": 55470 }, { "grad_norm": 0.18245281279087067, "learning_rate": 4.508227680932356e-05, "loss": 0.0121, "step": 55480 }, { "grad_norm": 0.20707352459430695, "learning_rate": 4.5065822549729605e-05, "loss": 0.0147, "step": 55490 }, { "grad_norm": 0.13841478526592255, "learning_rate": 4.504936882972989e-05, "loss": 0.0139, "step": 55500 }, { "grad_norm": 0.13168412446975708, "learning_rate": 4.503291565112374e-05, "loss": 0.0098, "step": 55510 }, { "grad_norm": 0.20001648366451263, "learning_rate": 4.501646301571048e-05, "loss": 0.014, "step": 55520 }, { "grad_norm": 0.1977168619632721, "learning_rate": 4.500001092528933e-05, "loss": 0.0145, "step": 55530 }, { "grad_norm": 0.26066336035728455, "learning_rate": 4.4983559381659455e-05, "loss": 0.0125, "step": 55540 }, { "grad_norm": 0.2388356328010559, "learning_rate": 4.496710838661999e-05, "loss": 0.0103, "step": 55550 }, { "grad_norm": 0.16376449167728424, "learning_rate": 4.495065794196999e-05, "loss": 0.0109, "step": 55560 }, { "grad_norm": 0.14757145941257477, "learning_rate": 4.4934208049508455e-05, "loss": 0.011, "step": 55570 }, { "grad_norm": 0.19400878250598907, "learning_rate": 4.4917758711034303e-05, "loss": 0.0117, "step": 55580 }, { "grad_norm": 0.10643672198057175, "learning_rate": 4.490130992834642e-05, "loss": 0.0149, "step": 55590 }, { "grad_norm": 0.13368846476078033, "learning_rate": 4.488486170324359e-05, "loss": 0.0104, "step": 55600 }, { "grad_norm": 0.19160424172878265, "learning_rate": 4.486841403752463e-05, "loss": 0.0116, "step": 55610 }, { "grad_norm": 0.18706095218658447, "learning_rate": 4.4851966932988174e-05, "loss": 0.0128, "step": 55620 }, { "grad_norm": 0.18221542239189148, "learning_rate": 4.4835520391432894e-05, "loss": 0.0167, "step": 55630 }, { "grad_norm": 0.1658409684896469, "learning_rate": 4.481907441465733e-05, "loss": 0.0111, "step": 55640 }, { "grad_norm": 0.22030004858970642, "learning_rate": 4.480262900446e-05, "loss": 0.0126, "step": 55650 }, { "grad_norm": 0.17755164206027985, "learning_rate": 4.478618416263933e-05, "loss": 0.0102, "step": 55660 }, { "grad_norm": 0.1421848088502884, "learning_rate": 4.476973989099373e-05, "loss": 0.0118, "step": 55670 }, { "grad_norm": 0.11866439878940582, "learning_rate": 4.475329619132153e-05, "loss": 0.0097, "step": 55680 }, { "grad_norm": 0.1356373280286789, "learning_rate": 4.4736853065420956e-05, "loss": 0.0124, "step": 55690 }, { "grad_norm": 0.20733004808425903, "learning_rate": 4.472041051509023e-05, "loss": 0.0109, "step": 55700 }, { "grad_norm": 0.24321389198303223, "learning_rate": 4.470396854212746e-05, "loss": 0.0133, "step": 55710 }, { "grad_norm": 0.1305219680070877, "learning_rate": 4.468752714833072e-05, "loss": 0.0109, "step": 55720 }, { "grad_norm": 0.16218224167823792, "learning_rate": 4.467108633549803e-05, "loss": 0.0138, "step": 55730 }, { "grad_norm": 0.14582626521587372, "learning_rate": 4.4654646105427335e-05, "loss": 0.0113, "step": 55740 }, { "grad_norm": 0.15809690952301025, "learning_rate": 4.463820645991651e-05, "loss": 0.0135, "step": 55750 }, { "grad_norm": 0.14061345160007477, "learning_rate": 4.462176740076334e-05, "loss": 0.0124, "step": 55760 }, { "grad_norm": 0.2252613753080368, "learning_rate": 4.4605328929765624e-05, "loss": 0.0137, "step": 55770 }, { "grad_norm": 0.16128002107143402, "learning_rate": 4.458889104872102e-05, "loss": 0.0136, "step": 55780 }, { "grad_norm": 0.16193799674510956, "learning_rate": 4.457245375942715e-05, "loss": 0.0107, "step": 55790 }, { "grad_norm": 0.44518494606018066, "learning_rate": 4.455601706368157e-05, "loss": 0.0135, "step": 55800 }, { "grad_norm": 0.1439393013715744, "learning_rate": 4.453958096328179e-05, "loss": 0.0119, "step": 55810 }, { "grad_norm": 0.13897372782230377, "learning_rate": 4.452314546002523e-05, "loss": 0.0168, "step": 55820 }, { "grad_norm": 0.16534003615379333, "learning_rate": 4.4506710555709246e-05, "loss": 0.0115, "step": 55830 }, { "grad_norm": 0.20036061108112335, "learning_rate": 4.449027625213114e-05, "loss": 0.0102, "step": 55840 }, { "grad_norm": 0.18332861363887787, "learning_rate": 4.4473842551088135e-05, "loss": 0.0146, "step": 55850 }, { "grad_norm": 0.12895233929157257, "learning_rate": 4.445740945437738e-05, "loss": 0.0126, "step": 55860 }, { "grad_norm": 0.1280367374420166, "learning_rate": 4.444097696379602e-05, "loss": 0.0105, "step": 55870 }, { "grad_norm": 0.1377767026424408, "learning_rate": 4.442454508114106e-05, "loss": 0.0114, "step": 55880 }, { "grad_norm": 0.2764202654361725, "learning_rate": 4.4408113808209465e-05, "loss": 0.0158, "step": 55890 }, { "grad_norm": 0.16660109162330627, "learning_rate": 4.439168314679813e-05, "loss": 0.0163, "step": 55900 }, { "grad_norm": 0.11656251549720764, "learning_rate": 4.43752530987039e-05, "loss": 0.0104, "step": 55910 }, { "grad_norm": 0.19796957075595856, "learning_rate": 4.4358823665723515e-05, "loss": 0.0112, "step": 55920 }, { "grad_norm": 0.24031244218349457, "learning_rate": 4.434239484965371e-05, "loss": 0.0132, "step": 55930 }, { "grad_norm": 0.1322421282529831, "learning_rate": 4.4325966652291103e-05, "loss": 0.0119, "step": 55940 }, { "grad_norm": 0.13933412730693817, "learning_rate": 4.430953907543225e-05, "loss": 0.011, "step": 55950 }, { "grad_norm": 0.12859323620796204, "learning_rate": 4.4293112120873645e-05, "loss": 0.0107, "step": 55960 }, { "grad_norm": 0.12683865427970886, "learning_rate": 4.427668579041172e-05, "loss": 0.0129, "step": 55970 }, { "grad_norm": 0.14324462413787842, "learning_rate": 4.4260260085842816e-05, "loss": 0.013, "step": 55980 }, { "grad_norm": 0.14504091441631317, "learning_rate": 4.4243835008963256e-05, "loss": 0.0093, "step": 55990 }, { "grad_norm": 0.16004911065101624, "learning_rate": 4.422741056156925e-05, "loss": 0.011, "step": 56000 }, { "grad_norm": 0.1467774510383606, "learning_rate": 4.421098674545693e-05, "loss": 0.0116, "step": 56010 }, { "grad_norm": 0.17355789244174957, "learning_rate": 4.419456356242241e-05, "loss": 0.0139, "step": 56020 }, { "grad_norm": 0.17339932918548584, "learning_rate": 4.4178141014261685e-05, "loss": 0.0113, "step": 56030 }, { "grad_norm": 0.14908495545387268, "learning_rate": 4.416171910277071e-05, "loss": 0.0125, "step": 56040 }, { "grad_norm": 0.18184109032154083, "learning_rate": 4.414529782974533e-05, "loss": 0.0141, "step": 56050 }, { "grad_norm": 0.14400102198123932, "learning_rate": 4.41288771969814e-05, "loss": 0.0107, "step": 56060 }, { "grad_norm": 0.14237932860851288, "learning_rate": 4.411245720627464e-05, "loss": 0.0103, "step": 56070 }, { "grad_norm": 0.14926017820835114, "learning_rate": 4.40960378594207e-05, "loss": 0.01, "step": 56080 }, { "grad_norm": 0.18705564737319946, "learning_rate": 4.407961915821519e-05, "loss": 0.0145, "step": 56090 }, { "grad_norm": 0.1557990312576294, "learning_rate": 4.406320110445362e-05, "loss": 0.0109, "step": 56100 }, { "grad_norm": 0.1774042695760727, "learning_rate": 4.404678369993144e-05, "loss": 0.0112, "step": 56110 }, { "grad_norm": 0.16385257244110107, "learning_rate": 4.403036694644406e-05, "loss": 0.0106, "step": 56120 }, { "grad_norm": 0.1828833669424057, "learning_rate": 4.4013950845786764e-05, "loss": 0.0093, "step": 56130 }, { "grad_norm": 0.12699362635612488, "learning_rate": 4.399753539975482e-05, "loss": 0.012, "step": 56140 }, { "grad_norm": 0.11987614631652832, "learning_rate": 4.398112061014337e-05, "loss": 0.0076, "step": 56150 }, { "grad_norm": 0.19630973041057587, "learning_rate": 4.396470647874753e-05, "loss": 0.0137, "step": 56160 }, { "grad_norm": 0.13987508416175842, "learning_rate": 4.394829300736229e-05, "loss": 0.0099, "step": 56170 }, { "grad_norm": 0.141071155667305, "learning_rate": 4.393188019778265e-05, "loss": 0.0086, "step": 56180 }, { "grad_norm": 0.13091464340686798, "learning_rate": 4.391546805180347e-05, "loss": 0.0108, "step": 56190 }, { "grad_norm": 0.2016482949256897, "learning_rate": 4.389905657121955e-05, "loss": 0.0159, "step": 56200 }, { "grad_norm": 0.12700854241847992, "learning_rate": 4.3882645757825635e-05, "loss": 0.0083, "step": 56210 }, { "grad_norm": 0.15222710371017456, "learning_rate": 4.386623561341637e-05, "loss": 0.0102, "step": 56220 }, { "grad_norm": 0.18347007036209106, "learning_rate": 4.384982613978637e-05, "loss": 0.0102, "step": 56230 }, { "grad_norm": 0.157422736287117, "learning_rate": 4.383341733873012e-05, "loss": 0.0091, "step": 56240 }, { "grad_norm": 0.14724911749362946, "learning_rate": 4.381700921204209e-05, "loss": 0.0101, "step": 56250 }, { "grad_norm": 0.16473416984081268, "learning_rate": 4.380060176151663e-05, "loss": 0.0102, "step": 56260 }, { "grad_norm": 0.23724955320358276, "learning_rate": 4.378419498894805e-05, "loss": 0.0107, "step": 56270 }, { "grad_norm": 0.19184471666812897, "learning_rate": 4.376778889613056e-05, "loss": 0.0117, "step": 56280 }, { "grad_norm": 0.19457006454467773, "learning_rate": 4.375138348485831e-05, "loss": 0.0104, "step": 56290 }, { "grad_norm": 0.1497393697500229, "learning_rate": 4.3734978756925346e-05, "loss": 0.0092, "step": 56300 }, { "grad_norm": 0.18227452039718628, "learning_rate": 4.3718574714125706e-05, "loss": 0.0101, "step": 56310 }, { "grad_norm": 0.12599897384643555, "learning_rate": 4.370217135825329e-05, "loss": 0.0115, "step": 56320 }, { "grad_norm": 0.20377221703529358, "learning_rate": 4.368576869110194e-05, "loss": 0.0173, "step": 56330 }, { "grad_norm": 0.13020780682563782, "learning_rate": 4.366936671446544e-05, "loss": 0.0104, "step": 56340 }, { "grad_norm": 0.1643131971359253, "learning_rate": 4.3652965430137474e-05, "loss": 0.0102, "step": 56350 }, { "grad_norm": 0.12549874186515808, "learning_rate": 4.3636564839911646e-05, "loss": 0.0102, "step": 56360 }, { "grad_norm": 0.1509050875902176, "learning_rate": 4.3620164945581545e-05, "loss": 0.0128, "step": 56370 }, { "grad_norm": 0.16161659359931946, "learning_rate": 4.360376574894061e-05, "loss": 0.0112, "step": 56380 }, { "grad_norm": 0.1487365961074829, "learning_rate": 4.358736725178224e-05, "loss": 0.01, "step": 56390 }, { "grad_norm": 0.1434343308210373, "learning_rate": 4.357096945589974e-05, "loss": 0.0124, "step": 56400 }, { "grad_norm": 0.1529948115348816, "learning_rate": 4.355457236308636e-05, "loss": 0.0117, "step": 56410 }, { "grad_norm": 0.1819460093975067, "learning_rate": 4.353817597513526e-05, "loss": 0.0117, "step": 56420 }, { "grad_norm": 0.1619623601436615, "learning_rate": 4.352178029383948e-05, "loss": 0.0124, "step": 56430 }, { "grad_norm": 0.1186174601316452, "learning_rate": 4.3505385320992105e-05, "loss": 0.0115, "step": 56440 }, { "grad_norm": 0.15288197994232178, "learning_rate": 4.348899105838602e-05, "loss": 0.0118, "step": 56450 }, { "grad_norm": 0.12185563147068024, "learning_rate": 4.3472597507814087e-05, "loss": 0.0104, "step": 56460 }, { "grad_norm": 0.15234488248825073, "learning_rate": 4.3456204671069066e-05, "loss": 0.009, "step": 56470 }, { "grad_norm": 0.17290300130844116, "learning_rate": 4.343981254994367e-05, "loss": 0.012, "step": 56480 }, { "grad_norm": 0.19725321233272552, "learning_rate": 4.34234211462305e-05, "loss": 0.0086, "step": 56490 }, { "grad_norm": 0.2094973623752594, "learning_rate": 4.340703046172213e-05, "loss": 0.0131, "step": 56500 }, { "grad_norm": 0.20060193538665771, "learning_rate": 4.339064049821097e-05, "loss": 0.0152, "step": 56510 }, { "grad_norm": 0.1694125086069107, "learning_rate": 4.3374251257489446e-05, "loss": 0.0112, "step": 56520 }, { "grad_norm": 0.18166904151439667, "learning_rate": 4.335786274134984e-05, "loss": 0.01, "step": 56530 }, { "grad_norm": 0.15592510998249054, "learning_rate": 4.334147495158438e-05, "loss": 0.0095, "step": 56540 }, { "grad_norm": 0.15686041116714478, "learning_rate": 4.33250878899852e-05, "loss": 0.0125, "step": 56550 }, { "grad_norm": 0.1544775813817978, "learning_rate": 4.330870155834437e-05, "loss": 0.0129, "step": 56560 }, { "grad_norm": 0.14080467820167542, "learning_rate": 4.3292315958453884e-05, "loss": 0.0087, "step": 56570 }, { "grad_norm": 0.16545788943767548, "learning_rate": 4.3275931092105646e-05, "loss": 0.0108, "step": 56580 }, { "grad_norm": 0.1630956083536148, "learning_rate": 4.325954696109147e-05, "loss": 0.0122, "step": 56590 }, { "grad_norm": 0.14593735337257385, "learning_rate": 4.32431635672031e-05, "loss": 0.0106, "step": 56600 }, { "grad_norm": 0.1479644924402237, "learning_rate": 4.3226780912232214e-05, "loss": 0.0114, "step": 56610 }, { "grad_norm": 0.21333099901676178, "learning_rate": 4.321039899797036e-05, "loss": 0.011, "step": 56620 }, { "grad_norm": 0.1680959314107895, "learning_rate": 4.319401782620908e-05, "loss": 0.0103, "step": 56630 }, { "grad_norm": 0.16102191805839539, "learning_rate": 4.317763739873978e-05, "loss": 0.0128, "step": 56640 }, { "grad_norm": 0.15726658701896667, "learning_rate": 4.31612577173538e-05, "loss": 0.009, "step": 56650 }, { "grad_norm": 0.18772660195827484, "learning_rate": 4.314487878384239e-05, "loss": 0.0104, "step": 56660 }, { "grad_norm": 0.16281729936599731, "learning_rate": 4.3128500599996736e-05, "loss": 0.0109, "step": 56670 }, { "grad_norm": 0.17241418361663818, "learning_rate": 4.31121231676079e-05, "loss": 0.0094, "step": 56680 }, { "grad_norm": 0.18911337852478027, "learning_rate": 4.309574648846694e-05, "loss": 0.0096, "step": 56690 }, { "grad_norm": 0.1625010222196579, "learning_rate": 4.3079370564364755e-05, "loss": 0.0115, "step": 56700 }, { "grad_norm": 0.161275252699852, "learning_rate": 4.306299539709221e-05, "loss": 0.0101, "step": 56710 }, { "grad_norm": 0.1716528832912445, "learning_rate": 4.3046620988440046e-05, "loss": 0.0097, "step": 56720 }, { "grad_norm": 0.11380474269390106, "learning_rate": 4.303024734019897e-05, "loss": 0.0097, "step": 56730 }, { "grad_norm": 0.16539274156093597, "learning_rate": 4.3013874454159565e-05, "loss": 0.0101, "step": 56740 }, { "grad_norm": 0.18399225175380707, "learning_rate": 4.299750233211233e-05, "loss": 0.01, "step": 56750 }, { "grad_norm": 0.09383274614810944, "learning_rate": 4.2981130975847715e-05, "loss": 0.0105, "step": 56760 }, { "grad_norm": 0.1341027021408081, "learning_rate": 4.2964760387156076e-05, "loss": 0.0118, "step": 56770 }, { "grad_norm": 0.19627787172794342, "learning_rate": 4.294839056782767e-05, "loss": 0.0129, "step": 56780 }, { "grad_norm": 0.15844237804412842, "learning_rate": 4.293202151965266e-05, "loss": 0.0087, "step": 56790 }, { "grad_norm": 0.15618234872817993, "learning_rate": 4.291565324442115e-05, "loss": 0.0097, "step": 56800 }, { "grad_norm": 0.16542267799377441, "learning_rate": 4.2899285743923135e-05, "loss": 0.016, "step": 56810 }, { "grad_norm": 0.15892206132411957, "learning_rate": 4.288291901994857e-05, "loss": 0.0103, "step": 56820 }, { "grad_norm": 0.17549097537994385, "learning_rate": 4.286655307428729e-05, "loss": 0.0123, "step": 56830 }, { "grad_norm": 0.11414042860269547, "learning_rate": 4.285018790872903e-05, "loss": 0.0093, "step": 56840 }, { "grad_norm": 0.17080767452716827, "learning_rate": 4.283382352506347e-05, "loss": 0.01, "step": 56850 }, { "grad_norm": 0.1231452077627182, "learning_rate": 4.2817459925080205e-05, "loss": 0.0106, "step": 56860 }, { "grad_norm": 0.22839096188545227, "learning_rate": 4.28010971105687e-05, "loss": 0.0133, "step": 56870 }, { "grad_norm": 0.16874010860919952, "learning_rate": 4.278473508331841e-05, "loss": 0.0095, "step": 56880 }, { "grad_norm": 0.17539401352405548, "learning_rate": 4.276837384511864e-05, "loss": 0.0116, "step": 56890 }, { "grad_norm": 0.2074841558933258, "learning_rate": 4.275201339775864e-05, "loss": 0.012, "step": 56900 }, { "grad_norm": 0.16001641750335693, "learning_rate": 4.273565374302756e-05, "loss": 0.0093, "step": 56910 }, { "grad_norm": 0.20419983565807343, "learning_rate": 4.2719294882714454e-05, "loss": 0.0124, "step": 56920 }, { "grad_norm": 0.18110232055187225, "learning_rate": 4.270293681860831e-05, "loss": 0.0149, "step": 56930 }, { "grad_norm": 0.17711995542049408, "learning_rate": 4.268657955249802e-05, "loss": 0.013, "step": 56940 }, { "grad_norm": 0.26763978600502014, "learning_rate": 4.267022308617241e-05, "loss": 0.0133, "step": 56950 }, { "grad_norm": 0.1925099492073059, "learning_rate": 4.265386742142018e-05, "loss": 0.0151, "step": 56960 }, { "grad_norm": 0.17740409076213837, "learning_rate": 4.263751256002996e-05, "loss": 0.0097, "step": 56970 }, { "grad_norm": 0.14179547131061554, "learning_rate": 4.26211585037903e-05, "loss": 0.0129, "step": 56980 }, { "grad_norm": 0.17904260754585266, "learning_rate": 4.2604805254489656e-05, "loss": 0.0113, "step": 56990 }, { "grad_norm": 0.11464165896177292, "learning_rate": 4.258845281391638e-05, "loss": 0.011, "step": 57000 }, { "grad_norm": 0.16661633551120758, "learning_rate": 4.257210118385876e-05, "loss": 0.0089, "step": 57010 }, { "grad_norm": 0.14233140647411346, "learning_rate": 4.2555750366104994e-05, "loss": 0.0092, "step": 57020 }, { "grad_norm": 0.12232402712106705, "learning_rate": 4.2539400362443184e-05, "loss": 0.0097, "step": 57030 }, { "grad_norm": 0.17649832367897034, "learning_rate": 4.252305117466133e-05, "loss": 0.0119, "step": 57040 }, { "grad_norm": 0.12488862872123718, "learning_rate": 4.2506702804547356e-05, "loss": 0.0103, "step": 57050 }, { "grad_norm": 0.17238003015518188, "learning_rate": 4.249035525388907e-05, "loss": 0.0134, "step": 57060 }, { "grad_norm": 0.1444230079650879, "learning_rate": 4.247400852447427e-05, "loss": 0.011, "step": 57070 }, { "grad_norm": 0.21791601181030273, "learning_rate": 4.245766261809059e-05, "loss": 0.0109, "step": 57080 }, { "grad_norm": 0.1869509071111679, "learning_rate": 4.244131753652558e-05, "loss": 0.0127, "step": 57090 }, { "grad_norm": 0.18242689967155457, "learning_rate": 4.242497328156672e-05, "loss": 0.0093, "step": 57100 }, { "grad_norm": 0.19190263748168945, "learning_rate": 4.240862985500139e-05, "loss": 0.0135, "step": 57110 }, { "grad_norm": 0.1973433643579483, "learning_rate": 4.2392287258616894e-05, "loss": 0.0115, "step": 57120 }, { "grad_norm": 0.1573527604341507, "learning_rate": 4.2375945494200396e-05, "loss": 0.0105, "step": 57130 }, { "grad_norm": 0.13590697944164276, "learning_rate": 4.2359604563539054e-05, "loss": 0.013, "step": 57140 }, { "grad_norm": 0.135468989610672, "learning_rate": 4.2343264468419885e-05, "loss": 0.0122, "step": 57150 }, { "grad_norm": 0.10504772514104843, "learning_rate": 4.232692521062979e-05, "loss": 0.0102, "step": 57160 }, { "grad_norm": 0.19997809827327728, "learning_rate": 4.231058679195562e-05, "loss": 0.0129, "step": 57170 }, { "grad_norm": 0.24456727504730225, "learning_rate": 4.229424921418411e-05, "loss": 0.0099, "step": 57180 }, { "grad_norm": 0.10708533972501755, "learning_rate": 4.2277912479101916e-05, "loss": 0.0099, "step": 57190 }, { "grad_norm": 0.15767532587051392, "learning_rate": 4.226157658849562e-05, "loss": 0.0115, "step": 57200 }, { "grad_norm": 0.12258069217205048, "learning_rate": 4.2245241544151664e-05, "loss": 0.0092, "step": 57210 }, { "grad_norm": 0.17565801739692688, "learning_rate": 4.2228907347856425e-05, "loss": 0.0108, "step": 57220 }, { "grad_norm": 0.11529050767421722, "learning_rate": 4.2212574001396214e-05, "loss": 0.0138, "step": 57230 }, { "grad_norm": 0.09903685748577118, "learning_rate": 4.21962415065572e-05, "loss": 0.0089, "step": 57240 }, { "grad_norm": 0.19144970178604126, "learning_rate": 4.217990986512548e-05, "loss": 0.0146, "step": 57250 }, { "grad_norm": 0.18269400298595428, "learning_rate": 4.216357907888704e-05, "loss": 0.0101, "step": 57260 }, { "grad_norm": 0.17169371247291565, "learning_rate": 4.2147249149627824e-05, "loss": 0.0101, "step": 57270 }, { "grad_norm": 0.16928471624851227, "learning_rate": 4.213092007913364e-05, "loss": 0.0103, "step": 57280 }, { "grad_norm": 0.1513407677412033, "learning_rate": 4.2114591869190205e-05, "loss": 0.0121, "step": 57290 }, { "grad_norm": 0.1683441400527954, "learning_rate": 4.209826452158315e-05, "loss": 0.0106, "step": 57300 }, { "grad_norm": 0.14953331649303436, "learning_rate": 4.2081938038098e-05, "loss": 0.0099, "step": 57310 }, { "grad_norm": 0.1398017555475235, "learning_rate": 4.2065612420520185e-05, "loss": 0.0087, "step": 57320 }, { "grad_norm": 0.1382317692041397, "learning_rate": 4.204928767063509e-05, "loss": 0.0096, "step": 57330 }, { "grad_norm": 0.18400970101356506, "learning_rate": 4.2032963790227944e-05, "loss": 0.0106, "step": 57340 }, { "grad_norm": 0.20769450068473816, "learning_rate": 4.2016640781083887e-05, "loss": 0.0129, "step": 57350 }, { "grad_norm": 0.13720346987247467, "learning_rate": 4.2000318644988004e-05, "loss": 0.0128, "step": 57360 }, { "grad_norm": 0.14009802043437958, "learning_rate": 4.198399738372524e-05, "loss": 0.0107, "step": 57370 }, { "grad_norm": 0.15077361464500427, "learning_rate": 4.196767699908045e-05, "loss": 0.0123, "step": 57380 }, { "grad_norm": 0.14495907723903656, "learning_rate": 4.195135749283845e-05, "loss": 0.0141, "step": 57390 }, { "grad_norm": 0.15708895027637482, "learning_rate": 4.1935038866783894e-05, "loss": 0.0115, "step": 57400 }, { "grad_norm": 0.2018575370311737, "learning_rate": 4.1918721122701355e-05, "loss": 0.012, "step": 57410 }, { "grad_norm": 0.15088070929050446, "learning_rate": 4.190240426237533e-05, "loss": 0.0126, "step": 57420 }, { "grad_norm": 0.1712263524532318, "learning_rate": 4.188608828759019e-05, "loss": 0.0108, "step": 57430 }, { "grad_norm": 0.2156255543231964, "learning_rate": 4.1869773200130246e-05, "loss": 0.0107, "step": 57440 }, { "grad_norm": 0.1305922269821167, "learning_rate": 4.185345900177967e-05, "loss": 0.0137, "step": 57450 }, { "grad_norm": 0.13066831231117249, "learning_rate": 4.183714569432258e-05, "loss": 0.0115, "step": 57460 }, { "grad_norm": 0.2170066386461258, "learning_rate": 4.182083327954295e-05, "loss": 0.0123, "step": 57470 }, { "grad_norm": 0.103436179459095, "learning_rate": 4.1804521759224714e-05, "loss": 0.0094, "step": 57480 }, { "grad_norm": 0.19626213610172272, "learning_rate": 4.178821113515165e-05, "loss": 0.0125, "step": 57490 }, { "grad_norm": 0.22264228761196136, "learning_rate": 4.1771901409107476e-05, "loss": 0.0112, "step": 57500 }, { "grad_norm": 0.18205489218235016, "learning_rate": 4.175559258287577e-05, "loss": 0.0117, "step": 57510 }, { "grad_norm": 0.12006117403507233, "learning_rate": 4.173928465824008e-05, "loss": 0.0098, "step": 57520 }, { "grad_norm": 0.16061992943286896, "learning_rate": 4.17229776369838e-05, "loss": 0.0143, "step": 57530 }, { "grad_norm": 0.14421004056930542, "learning_rate": 4.170667152089025e-05, "loss": 0.0083, "step": 57540 }, { "grad_norm": 0.1167537048459053, "learning_rate": 4.1690366311742626e-05, "loss": 0.0119, "step": 57550 }, { "grad_norm": 0.12181650102138519, "learning_rate": 4.1674062011324055e-05, "loss": 0.0129, "step": 57560 }, { "grad_norm": 0.11957395076751709, "learning_rate": 4.165775862141752e-05, "loss": 0.0116, "step": 57570 }, { "grad_norm": 0.15440627932548523, "learning_rate": 4.164145614380598e-05, "loss": 0.0104, "step": 57580 }, { "grad_norm": 0.17114916443824768, "learning_rate": 4.162515458027223e-05, "loss": 0.0102, "step": 57590 }, { "grad_norm": 0.11270193010568619, "learning_rate": 4.160885393259899e-05, "loss": 0.0114, "step": 57600 }, { "grad_norm": 0.12213089317083359, "learning_rate": 4.159255420256886e-05, "loss": 0.0083, "step": 57610 }, { "grad_norm": 0.1911809742450714, "learning_rate": 4.157625539196436e-05, "loss": 0.0114, "step": 57620 }, { "grad_norm": 0.24015215039253235, "learning_rate": 4.155995750256791e-05, "loss": 0.0137, "step": 57630 }, { "grad_norm": 0.1507209837436676, "learning_rate": 4.15436605361618e-05, "loss": 0.013, "step": 57640 }, { "grad_norm": 0.18749390542507172, "learning_rate": 4.152736449452827e-05, "loss": 0.0108, "step": 57650 }, { "grad_norm": 0.15288889408111572, "learning_rate": 4.151106937944943e-05, "loss": 0.0113, "step": 57660 }, { "grad_norm": 0.09741771966218948, "learning_rate": 4.149477519270727e-05, "loss": 0.0141, "step": 57670 }, { "grad_norm": 0.1637604534626007, "learning_rate": 4.147848193608371e-05, "loss": 0.0097, "step": 57680 }, { "grad_norm": 0.143344447016716, "learning_rate": 4.146218961136056e-05, "loss": 0.0096, "step": 57690 }, { "grad_norm": 0.15693901479244232, "learning_rate": 4.14458982203195e-05, "loss": 0.0101, "step": 57700 }, { "grad_norm": 0.10942205786705017, "learning_rate": 4.1429607764742165e-05, "loss": 0.01, "step": 57710 }, { "grad_norm": 0.18926812708377838, "learning_rate": 4.141331824641003e-05, "loss": 0.0107, "step": 57720 }, { "grad_norm": 0.14112909138202667, "learning_rate": 4.139702966710452e-05, "loss": 0.0111, "step": 57730 }, { "grad_norm": 0.12431848049163818, "learning_rate": 4.13807420286069e-05, "loss": 0.0095, "step": 57740 }, { "grad_norm": 0.17735230922698975, "learning_rate": 4.136445533269837e-05, "loss": 0.0106, "step": 57750 }, { "grad_norm": 0.12189647555351257, "learning_rate": 4.134816958116003e-05, "loss": 0.009, "step": 57760 }, { "grad_norm": 0.16524069011211395, "learning_rate": 4.133188477577283e-05, "loss": 0.0082, "step": 57770 }, { "grad_norm": 0.15927855670452118, "learning_rate": 4.131560091831771e-05, "loss": 0.0112, "step": 57780 }, { "grad_norm": 0.13795843720436096, "learning_rate": 4.1299318010575405e-05, "loss": 0.012, "step": 57790 }, { "grad_norm": 0.12549340724945068, "learning_rate": 4.128303605432661e-05, "loss": 0.012, "step": 57800 }, { "grad_norm": 0.1738080084323883, "learning_rate": 4.126675505135188e-05, "loss": 0.0115, "step": 57810 }, { "grad_norm": 0.13647447526454926, "learning_rate": 4.125047500343169e-05, "loss": 0.0092, "step": 57820 }, { "grad_norm": 0.10724908858537674, "learning_rate": 4.1234195912346376e-05, "loss": 0.0091, "step": 57830 }, { "grad_norm": 0.13376307487487793, "learning_rate": 4.1217917779876235e-05, "loss": 0.0089, "step": 57840 }, { "grad_norm": 0.10314489156007767, "learning_rate": 4.120164060780141e-05, "loss": 0.013, "step": 57850 }, { "grad_norm": 0.11998653411865234, "learning_rate": 4.1185364397901935e-05, "loss": 0.0126, "step": 57860 }, { "grad_norm": 0.18132424354553223, "learning_rate": 4.116908915195776e-05, "loss": 0.0129, "step": 57870 }, { "grad_norm": 0.21172849833965302, "learning_rate": 4.115281487174872e-05, "loss": 0.0087, "step": 57880 }, { "grad_norm": 0.17496523261070251, "learning_rate": 4.113654155905452e-05, "loss": 0.0117, "step": 57890 }, { "grad_norm": 0.14408445358276367, "learning_rate": 4.112026921565483e-05, "loss": 0.0089, "step": 57900 }, { "grad_norm": 0.1723979413509369, "learning_rate": 4.1103997843329147e-05, "loss": 0.0121, "step": 57910 }, { "grad_norm": 0.24903354048728943, "learning_rate": 4.108772744385689e-05, "loss": 0.01, "step": 57920 }, { "grad_norm": 0.1721310019493103, "learning_rate": 4.1071458019017346e-05, "loss": 0.0099, "step": 57930 }, { "grad_norm": 0.15639063715934753, "learning_rate": 4.1055189570589734e-05, "loss": 0.0103, "step": 57940 }, { "grad_norm": 0.14616557955741882, "learning_rate": 4.103892210035315e-05, "loss": 0.0113, "step": 57950 }, { "grad_norm": 0.1414787769317627, "learning_rate": 4.102265561008656e-05, "loss": 0.0116, "step": 57960 }, { "grad_norm": 0.1952463984489441, "learning_rate": 4.100639010156885e-05, "loss": 0.0109, "step": 57970 }, { "grad_norm": 0.1651594638824463, "learning_rate": 4.099012557657882e-05, "loss": 0.0117, "step": 57980 }, { "grad_norm": 0.14458785951137543, "learning_rate": 4.09738620368951e-05, "loss": 0.0105, "step": 57990 }, { "grad_norm": 0.17824317514896393, "learning_rate": 4.095759948429626e-05, "loss": 0.0119, "step": 58000 }, { "grad_norm": 0.1769111603498459, "learning_rate": 4.094133792056075e-05, "loss": 0.0109, "step": 58010 }, { "grad_norm": 0.1724306046962738, "learning_rate": 4.092507734746688e-05, "loss": 0.0116, "step": 58020 }, { "grad_norm": 0.15764367580413818, "learning_rate": 4.090881776679293e-05, "loss": 0.0098, "step": 58030 }, { "grad_norm": 0.17640554904937744, "learning_rate": 4.0892559180316994e-05, "loss": 0.0089, "step": 58040 }, { "grad_norm": 0.13196419179439545, "learning_rate": 4.08763015898171e-05, "loss": 0.0101, "step": 58050 }, { "grad_norm": 0.1889871507883072, "learning_rate": 4.0860044997071145e-05, "loss": 0.0093, "step": 58060 }, { "grad_norm": 0.21918180584907532, "learning_rate": 4.084378940385693e-05, "loss": 0.0138, "step": 58070 }, { "grad_norm": 0.19280825555324554, "learning_rate": 4.082753481195211e-05, "loss": 0.0132, "step": 58080 }, { "grad_norm": 0.14945395290851593, "learning_rate": 4.0811281223134324e-05, "loss": 0.0087, "step": 58090 }, { "grad_norm": 0.14487317204475403, "learning_rate": 4.0795028639181e-05, "loss": 0.0082, "step": 58100 }, { "grad_norm": 0.17987126111984253, "learning_rate": 4.077877706186951e-05, "loss": 0.0116, "step": 58110 }, { "grad_norm": 0.13810953497886658, "learning_rate": 4.076252649297709e-05, "loss": 0.0082, "step": 58120 }, { "grad_norm": 0.15716521441936493, "learning_rate": 4.07462769342809e-05, "loss": 0.0118, "step": 58130 }, { "grad_norm": 0.16909807920455933, "learning_rate": 4.073002838755793e-05, "loss": 0.0102, "step": 58140 }, { "grad_norm": 0.11414936929941177, "learning_rate": 4.0713780854585115e-05, "loss": 0.0127, "step": 58150 }, { "grad_norm": 0.12720152735710144, "learning_rate": 4.069753433713927e-05, "loss": 0.0128, "step": 58160 }, { "grad_norm": 0.20466628670692444, "learning_rate": 4.068128883699709e-05, "loss": 0.0119, "step": 58170 }, { "grad_norm": 0.14463494718074799, "learning_rate": 4.0665044355935155e-05, "loss": 0.0113, "step": 58180 }, { "grad_norm": 0.16098591685295105, "learning_rate": 4.064880089572993e-05, "loss": 0.0127, "step": 58190 }, { "grad_norm": 0.17647846043109894, "learning_rate": 4.063255845815779e-05, "loss": 0.0104, "step": 58200 }, { "grad_norm": 0.1478762924671173, "learning_rate": 4.0616317044994945e-05, "loss": 0.0122, "step": 58210 }, { "grad_norm": 0.13393200933933258, "learning_rate": 4.0600076658017585e-05, "loss": 0.0084, "step": 58220 }, { "grad_norm": 0.1812731772661209, "learning_rate": 4.0583837299001706e-05, "loss": 0.0109, "step": 58230 }, { "grad_norm": 0.19236557185649872, "learning_rate": 4.056759896972322e-05, "loss": 0.0109, "step": 58240 }, { "grad_norm": 0.20899076759815216, "learning_rate": 4.055136167195793e-05, "loss": 0.0136, "step": 58250 }, { "grad_norm": 0.1301155388355255, "learning_rate": 4.053512540748152e-05, "loss": 0.0092, "step": 58260 }, { "grad_norm": 0.1466817557811737, "learning_rate": 4.051889017806955e-05, "loss": 0.0124, "step": 58270 }, { "grad_norm": 0.11643002182245255, "learning_rate": 4.0502655985497516e-05, "loss": 0.0111, "step": 58280 }, { "grad_norm": 0.1532406210899353, "learning_rate": 4.048642283154074e-05, "loss": 0.0119, "step": 58290 }, { "grad_norm": 0.20057646930217743, "learning_rate": 4.0470190717974464e-05, "loss": 0.0122, "step": 58300 }, { "grad_norm": 0.16681578755378723, "learning_rate": 4.045395964657381e-05, "loss": 0.0127, "step": 58310 }, { "grad_norm": 0.1616484522819519, "learning_rate": 4.043772961911376e-05, "loss": 0.0126, "step": 58320 }, { "grad_norm": 0.14091846346855164, "learning_rate": 4.042150063736924e-05, "loss": 0.0112, "step": 58330 }, { "grad_norm": 0.1282404214143753, "learning_rate": 4.0405272703114976e-05, "loss": 0.0109, "step": 58340 }, { "grad_norm": 0.1271274983882904, "learning_rate": 4.038904581812569e-05, "loss": 0.0094, "step": 58350 }, { "grad_norm": 0.1638338267803192, "learning_rate": 4.03728199841759e-05, "loss": 0.0114, "step": 58360 }, { "grad_norm": 0.1522224396467209, "learning_rate": 4.0356595203040044e-05, "loss": 0.0094, "step": 58370 }, { "grad_norm": 0.14735479652881622, "learning_rate": 4.0340371476492426e-05, "loss": 0.0104, "step": 58380 }, { "grad_norm": 0.19647109508514404, "learning_rate": 4.032414880630727e-05, "loss": 0.0093, "step": 58390 }, { "grad_norm": 0.12796388566493988, "learning_rate": 4.0307927194258646e-05, "loss": 0.0094, "step": 58400 }, { "grad_norm": 0.11349471658468246, "learning_rate": 4.029170664212054e-05, "loss": 0.009, "step": 58410 }, { "grad_norm": 0.18431714177131653, "learning_rate": 4.0275487151666794e-05, "loss": 0.011, "step": 58420 }, { "grad_norm": 0.11290198564529419, "learning_rate": 4.025926872467116e-05, "loss": 0.0133, "step": 58430 }, { "grad_norm": 0.1358090192079544, "learning_rate": 4.024305136290725e-05, "loss": 0.0078, "step": 58440 }, { "grad_norm": 0.07687632739543915, "learning_rate": 4.022683506814858e-05, "loss": 0.0096, "step": 58450 }, { "grad_norm": 0.15778693556785583, "learning_rate": 4.021061984216853e-05, "loss": 0.0116, "step": 58460 }, { "grad_norm": 0.15856054425239563, "learning_rate": 4.019440568674035e-05, "loss": 0.0118, "step": 58470 }, { "grad_norm": 0.15324874222278595, "learning_rate": 4.017819260363724e-05, "loss": 0.0118, "step": 58480 }, { "grad_norm": 0.2511698603630066, "learning_rate": 4.016198059463223e-05, "loss": 0.0143, "step": 58490 }, { "grad_norm": 0.10613793134689331, "learning_rate": 4.0145769661498233e-05, "loss": 0.009, "step": 58500 }, { "grad_norm": 0.13151915371418, "learning_rate": 4.012955980600804e-05, "loss": 0.0104, "step": 58510 }, { "grad_norm": 0.13411231338977814, "learning_rate": 4.0113351029934355e-05, "loss": 0.0096, "step": 58520 }, { "grad_norm": 0.1692076027393341, "learning_rate": 4.0097143335049714e-05, "loss": 0.0104, "step": 58530 }, { "grad_norm": 0.17603804171085358, "learning_rate": 4.0080936723126606e-05, "loss": 0.0093, "step": 58540 }, { "grad_norm": 0.14744330942630768, "learning_rate": 4.006473119593735e-05, "loss": 0.009, "step": 58550 }, { "grad_norm": 0.15029096603393555, "learning_rate": 4.004852675525415e-05, "loss": 0.0127, "step": 58560 }, { "grad_norm": 0.16957731544971466, "learning_rate": 4.00323234028491e-05, "loss": 0.0115, "step": 58570 }, { "grad_norm": 0.21453861892223358, "learning_rate": 4.0016121140494186e-05, "loss": 0.011, "step": 58580 }, { "grad_norm": 0.19043368101119995, "learning_rate": 3.999991996996122e-05, "loss": 0.0108, "step": 58590 }, { "grad_norm": 0.18899881839752197, "learning_rate": 3.9983719893022e-05, "loss": 0.0124, "step": 58600 }, { "grad_norm": 0.197395920753479, "learning_rate": 3.996752091144812e-05, "loss": 0.0085, "step": 58610 }, { "grad_norm": 0.16767995059490204, "learning_rate": 3.9951323027011055e-05, "loss": 0.0105, "step": 58620 }, { "grad_norm": 0.17061655223369598, "learning_rate": 3.993512624148219e-05, "loss": 0.0108, "step": 58630 }, { "grad_norm": 0.1429717093706131, "learning_rate": 3.99189305566328e-05, "loss": 0.0089, "step": 58640 }, { "grad_norm": 0.1485881358385086, "learning_rate": 3.990273597423401e-05, "loss": 0.0138, "step": 58650 }, { "grad_norm": 0.14453035593032837, "learning_rate": 3.988654249605681e-05, "loss": 0.0106, "step": 58660 }, { "grad_norm": 0.15575197339057922, "learning_rate": 3.987035012387213e-05, "loss": 0.0125, "step": 58670 }, { "grad_norm": 0.1429060995578766, "learning_rate": 3.9854158859450735e-05, "loss": 0.0087, "step": 58680 }, { "grad_norm": 0.16536957025527954, "learning_rate": 3.983796870456328e-05, "loss": 0.0126, "step": 58690 }, { "grad_norm": 0.15348273515701294, "learning_rate": 3.982177966098028e-05, "loss": 0.0116, "step": 58700 }, { "grad_norm": 0.1523474156856537, "learning_rate": 3.980559173047216e-05, "loss": 0.0122, "step": 58710 }, { "grad_norm": 0.2299910932779312, "learning_rate": 3.9789404914809176e-05, "loss": 0.0133, "step": 58720 }, { "grad_norm": 0.14915144443511963, "learning_rate": 3.9773219215761544e-05, "loss": 0.0123, "step": 58730 }, { "grad_norm": 0.16685469448566437, "learning_rate": 3.975703463509928e-05, "loss": 0.0121, "step": 58740 }, { "grad_norm": 0.19031360745429993, "learning_rate": 3.974085117459231e-05, "loss": 0.0116, "step": 58750 }, { "grad_norm": 0.15733017027378082, "learning_rate": 3.972466883601043e-05, "loss": 0.0107, "step": 58760 }, { "grad_norm": 0.13041487336158752, "learning_rate": 3.9708487621123316e-05, "loss": 0.0092, "step": 58770 }, { "grad_norm": 0.13975714147090912, "learning_rate": 3.969230753170051e-05, "loss": 0.0104, "step": 58780 }, { "grad_norm": 0.1933184117078781, "learning_rate": 3.967612856951146e-05, "loss": 0.011, "step": 58790 }, { "grad_norm": 0.13886654376983643, "learning_rate": 3.9659950736325476e-05, "loss": 0.0097, "step": 58800 }, { "grad_norm": 0.13576947152614594, "learning_rate": 3.964377403391174e-05, "loss": 0.0104, "step": 58810 }, { "grad_norm": 0.18007031083106995, "learning_rate": 3.9627598464039303e-05, "loss": 0.0103, "step": 58820 }, { "grad_norm": 0.11667907238006592, "learning_rate": 3.9611424028477096e-05, "loss": 0.0096, "step": 58830 }, { "grad_norm": 0.16200976073741913, "learning_rate": 3.959525072899394e-05, "loss": 0.0086, "step": 58840 }, { "grad_norm": 0.130319282412529, "learning_rate": 3.9579078567358506e-05, "loss": 0.0115, "step": 58850 }, { "grad_norm": 0.13725483417510986, "learning_rate": 3.956290754533939e-05, "loss": 0.0129, "step": 58860 }, { "grad_norm": 0.21203675866127014, "learning_rate": 3.954673766470501e-05, "loss": 0.0099, "step": 58870 }, { "grad_norm": 0.15603339672088623, "learning_rate": 3.9530568927223674e-05, "loss": 0.0102, "step": 58880 }, { "grad_norm": 0.17345555126667023, "learning_rate": 3.95144013346636e-05, "loss": 0.0107, "step": 58890 }, { "grad_norm": 0.16651767492294312, "learning_rate": 3.949823488879282e-05, "loss": 0.0087, "step": 58900 }, { "grad_norm": 0.14763806760311127, "learning_rate": 3.9482069591379276e-05, "loss": 0.0127, "step": 58910 }, { "grad_norm": 0.41891661286354065, "learning_rate": 3.9465905444190784e-05, "loss": 0.0108, "step": 58920 }, { "grad_norm": 0.15548256039619446, "learning_rate": 3.944974244899505e-05, "loss": 0.0094, "step": 58930 }, { "grad_norm": 0.14899815618991852, "learning_rate": 3.9433580607559616e-05, "loss": 0.009, "step": 58940 }, { "grad_norm": 0.20160020887851715, "learning_rate": 3.941741992165193e-05, "loss": 0.0112, "step": 58950 }, { "grad_norm": 0.1453791707754135, "learning_rate": 3.940126039303928e-05, "loss": 0.0102, "step": 58960 }, { "grad_norm": 0.18719932436943054, "learning_rate": 3.938510202348886e-05, "loss": 0.0092, "step": 58970 }, { "grad_norm": 0.1514645665884018, "learning_rate": 3.93689448147677e-05, "loss": 0.0098, "step": 58980 }, { "grad_norm": 0.17391884326934814, "learning_rate": 3.935278876864278e-05, "loss": 0.0085, "step": 58990 }, { "grad_norm": 0.1617852747440338, "learning_rate": 3.933663388688087e-05, "loss": 0.0123, "step": 59000 }, { "grad_norm": 0.22346417605876923, "learning_rate": 3.9320480171248645e-05, "loss": 0.0122, "step": 59010 }, { "grad_norm": 0.17986156046390533, "learning_rate": 3.930432762351265e-05, "loss": 0.0112, "step": 59020 }, { "grad_norm": 0.16929052770137787, "learning_rate": 3.928817624543931e-05, "loss": 0.0093, "step": 59030 }, { "grad_norm": 0.1901577115058899, "learning_rate": 3.9272026038794885e-05, "loss": 0.0118, "step": 59040 }, { "grad_norm": 0.17374949157238007, "learning_rate": 3.9255877005345586e-05, "loss": 0.0111, "step": 59050 }, { "grad_norm": 0.15867206454277039, "learning_rate": 3.923972914685743e-05, "loss": 0.0089, "step": 59060 }, { "grad_norm": 0.19388489425182343, "learning_rate": 3.922358246509631e-05, "loss": 0.0118, "step": 59070 }, { "grad_norm": 0.15550485253334045, "learning_rate": 3.920743696182801e-05, "loss": 0.0147, "step": 59080 }, { "grad_norm": 0.12320073693990707, "learning_rate": 3.9191292638818164e-05, "loss": 0.0095, "step": 59090 }, { "grad_norm": 0.11895576864480972, "learning_rate": 3.9175149497832316e-05, "loss": 0.0086, "step": 59100 }, { "grad_norm": 0.10466600209474564, "learning_rate": 3.915900754063584e-05, "loss": 0.0083, "step": 59110 }, { "grad_norm": 0.18247303366661072, "learning_rate": 3.9142866768994014e-05, "loss": 0.0118, "step": 59120 }, { "grad_norm": 0.17092746496200562, "learning_rate": 3.912672718467193e-05, "loss": 0.0092, "step": 59130 }, { "grad_norm": 0.13384531438350677, "learning_rate": 3.911058878943463e-05, "loss": 0.0092, "step": 59140 }, { "grad_norm": 0.15228691697120667, "learning_rate": 3.9094451585046974e-05, "loss": 0.0094, "step": 59150 }, { "grad_norm": 0.15403316915035248, "learning_rate": 3.90783155732737e-05, "loss": 0.0088, "step": 59160 }, { "grad_norm": 0.17944389581680298, "learning_rate": 3.9062180755879373e-05, "loss": 0.0118, "step": 59170 }, { "grad_norm": 0.19469092786312103, "learning_rate": 3.904604713462855e-05, "loss": 0.0098, "step": 59180 }, { "grad_norm": 0.1564628928899765, "learning_rate": 3.902991471128554e-05, "loss": 0.0111, "step": 59190 }, { "grad_norm": 0.14899756014347076, "learning_rate": 3.901378348761456e-05, "loss": 0.0114, "step": 59200 }, { "grad_norm": 0.20606765151023865, "learning_rate": 3.899765346537971e-05, "loss": 0.0118, "step": 59210 }, { "grad_norm": 0.16321907937526703, "learning_rate": 3.898152464634493e-05, "loss": 0.0183, "step": 59220 }, { "grad_norm": 0.15269459784030914, "learning_rate": 3.896539703227402e-05, "loss": 0.0126, "step": 59230 }, { "grad_norm": 0.17071698606014252, "learning_rate": 3.894927062493073e-05, "loss": 0.0091, "step": 59240 }, { "grad_norm": 0.1258946806192398, "learning_rate": 3.893314542607859e-05, "loss": 0.0085, "step": 59250 }, { "grad_norm": 0.13026010990142822, "learning_rate": 3.891702143748103e-05, "loss": 0.0093, "step": 59260 }, { "grad_norm": 0.21105919778347015, "learning_rate": 3.890089866090133e-05, "loss": 0.0099, "step": 59270 }, { "grad_norm": 0.13892969489097595, "learning_rate": 3.888477709810267e-05, "loss": 0.0092, "step": 59280 }, { "grad_norm": 0.4795178174972534, "learning_rate": 3.886865675084806e-05, "loss": 0.0108, "step": 59290 }, { "grad_norm": 0.15434320271015167, "learning_rate": 3.885253762090043e-05, "loss": 0.0092, "step": 59300 }, { "grad_norm": 0.14233575761318207, "learning_rate": 3.8836419710022534e-05, "loss": 0.0085, "step": 59310 }, { "grad_norm": 0.2228865772485733, "learning_rate": 3.882030301997698e-05, "loss": 0.0121, "step": 59320 }, { "grad_norm": 0.14357014000415802, "learning_rate": 3.880418755252629e-05, "loss": 0.01, "step": 59330 }, { "grad_norm": 0.11514213681221008, "learning_rate": 3.87880733094328e-05, "loss": 0.0106, "step": 59340 }, { "grad_norm": 0.16971513628959656, "learning_rate": 3.877196029245877e-05, "loss": 0.0102, "step": 59350 }, { "grad_norm": 0.13694317638874054, "learning_rate": 3.875584850336627e-05, "loss": 0.0084, "step": 59360 }, { "grad_norm": 0.16473592817783356, "learning_rate": 3.873973794391728e-05, "loss": 0.0129, "step": 59370 }, { "grad_norm": 0.16617967188358307, "learning_rate": 3.872362861587361e-05, "loss": 0.0133, "step": 59380 }, { "grad_norm": 0.12914973497390747, "learning_rate": 3.870752052099698e-05, "loss": 0.0104, "step": 59390 }, { "grad_norm": 0.2601014971733093, "learning_rate": 3.8691413661048916e-05, "loss": 0.0121, "step": 59400 }, { "grad_norm": 0.12310182303190231, "learning_rate": 3.867530803779085e-05, "loss": 0.0132, "step": 59410 }, { "grad_norm": 0.15512506663799286, "learning_rate": 3.8659203652984054e-05, "loss": 0.0104, "step": 59420 }, { "grad_norm": 0.15021508932113647, "learning_rate": 3.864310050838972e-05, "loss": 0.0106, "step": 59430 }, { "grad_norm": 0.1667003631591797, "learning_rate": 3.8626998605768835e-05, "loss": 0.0116, "step": 59440 }, { "grad_norm": 0.22446995973587036, "learning_rate": 3.8610897946882294e-05, "loss": 0.0176, "step": 59450 }, { "grad_norm": 0.14950866997241974, "learning_rate": 3.859479853349082e-05, "loss": 0.0077, "step": 59460 }, { "grad_norm": 0.1648387312889099, "learning_rate": 3.857870036735503e-05, "loss": 0.0074, "step": 59470 }, { "grad_norm": 0.14641201496124268, "learning_rate": 3.856260345023538e-05, "loss": 0.0104, "step": 59480 }, { "grad_norm": 0.13657109439373016, "learning_rate": 3.854650778389224e-05, "loss": 0.0137, "step": 59490 }, { "grad_norm": 0.16104060411453247, "learning_rate": 3.8530413370085786e-05, "loss": 0.0104, "step": 59500 }, { "grad_norm": 0.12752531468868256, "learning_rate": 3.851432021057608e-05, "loss": 0.0087, "step": 59510 }, { "grad_norm": 0.1537959724664688, "learning_rate": 3.849822830712306e-05, "loss": 0.0128, "step": 59520 }, { "grad_norm": 0.12096056342124939, "learning_rate": 3.848213766148649e-05, "loss": 0.011, "step": 59530 }, { "grad_norm": 0.12910492718219757, "learning_rate": 3.846604827542602e-05, "loss": 0.0099, "step": 59540 }, { "grad_norm": 0.15177910029888153, "learning_rate": 3.844996015070117e-05, "loss": 0.0095, "step": 59550 }, { "grad_norm": 0.15512502193450928, "learning_rate": 3.843387328907132e-05, "loss": 0.0107, "step": 59560 }, { "grad_norm": 0.14374352991580963, "learning_rate": 3.8417787692295696e-05, "loss": 0.0109, "step": 59570 }, { "grad_norm": 0.1613045036792755, "learning_rate": 3.84017033621334e-05, "loss": 0.0088, "step": 59580 }, { "grad_norm": 0.14458942413330078, "learning_rate": 3.838562030034338e-05, "loss": 0.0129, "step": 59590 }, { "grad_norm": 0.18233659863471985, "learning_rate": 3.836953850868447e-05, "loss": 0.0134, "step": 59600 }, { "grad_norm": 0.13878317177295685, "learning_rate": 3.835345798891533e-05, "loss": 0.0086, "step": 59610 }, { "grad_norm": 0.1763858050107956, "learning_rate": 3.833737874279452e-05, "loss": 0.0086, "step": 59620 }, { "grad_norm": 0.1602369099855423, "learning_rate": 3.8321300772080433e-05, "loss": 0.0118, "step": 59630 }, { "grad_norm": 0.1455760896205902, "learning_rate": 3.8305224078531345e-05, "loss": 0.0081, "step": 59640 }, { "grad_norm": 0.22647838294506073, "learning_rate": 3.828914866390537e-05, "loss": 0.0121, "step": 59650 }, { "grad_norm": 0.1609407663345337, "learning_rate": 3.827307452996048e-05, "loss": 0.01, "step": 59660 }, { "grad_norm": 0.12655574083328247, "learning_rate": 3.825700167845454e-05, "loss": 0.0109, "step": 59670 }, { "grad_norm": 0.16413302719593048, "learning_rate": 3.8240930111145214e-05, "loss": 0.0089, "step": 59680 }, { "grad_norm": 0.1829773336648941, "learning_rate": 3.8224859829790116e-05, "loss": 0.0084, "step": 59690 }, { "grad_norm": 0.15140585601329803, "learning_rate": 3.820879083614664e-05, "loss": 0.014, "step": 59700 }, { "grad_norm": 0.2324744164943695, "learning_rate": 3.819272313197208e-05, "loss": 0.011, "step": 59710 }, { "grad_norm": 0.15624617040157318, "learning_rate": 3.817665671902355e-05, "loss": 0.013, "step": 59720 }, { "grad_norm": 0.13230827450752258, "learning_rate": 3.816059159905808e-05, "loss": 0.0111, "step": 59730 }, { "grad_norm": 0.12570852041244507, "learning_rate": 3.8144527773832476e-05, "loss": 0.0093, "step": 59740 }, { "grad_norm": 0.19873684644699097, "learning_rate": 3.812846524510352e-05, "loss": 0.0133, "step": 59750 }, { "grad_norm": 0.13854405283927917, "learning_rate": 3.811240401462775e-05, "loss": 0.012, "step": 59760 }, { "grad_norm": 0.16014696657657623, "learning_rate": 3.809634408416162e-05, "loss": 0.0136, "step": 59770 }, { "grad_norm": 0.16975969076156616, "learning_rate": 3.8080285455461394e-05, "loss": 0.0102, "step": 59780 }, { "grad_norm": 0.13988499343395233, "learning_rate": 3.806422813028323e-05, "loss": 0.0104, "step": 59790 }, { "grad_norm": 0.2852894365787506, "learning_rate": 3.8048172110383114e-05, "loss": 0.009, "step": 59800 }, { "grad_norm": 0.18524689972400665, "learning_rate": 3.8032117397516944e-05, "loss": 0.0095, "step": 59810 }, { "grad_norm": 0.16217651963233948, "learning_rate": 3.801606399344042e-05, "loss": 0.0127, "step": 59820 }, { "grad_norm": 0.20613734424114227, "learning_rate": 3.800001189990913e-05, "loss": 0.0119, "step": 59830 }, { "grad_norm": 0.08502611517906189, "learning_rate": 3.798396111867847e-05, "loss": 0.0114, "step": 59840 }, { "grad_norm": 0.20293337106704712, "learning_rate": 3.796791165150378e-05, "loss": 0.0152, "step": 59850 }, { "grad_norm": 0.15533201396465302, "learning_rate": 3.795186350014018e-05, "loss": 0.0105, "step": 59860 }, { "grad_norm": 0.1701236218214035, "learning_rate": 3.793581666634266e-05, "loss": 0.01, "step": 59870 }, { "grad_norm": 0.12372900545597076, "learning_rate": 3.791977115186609e-05, "loss": 0.0117, "step": 59880 }, { "grad_norm": 0.10834203660488129, "learning_rate": 3.79037269584652e-05, "loss": 0.0119, "step": 59890 }, { "grad_norm": 0.14162641763687134, "learning_rate": 3.7887684087894545e-05, "loss": 0.0106, "step": 59900 }, { "grad_norm": 0.15376824140548706, "learning_rate": 3.787164254190854e-05, "loss": 0.007, "step": 59910 }, { "grad_norm": 0.12216700613498688, "learning_rate": 3.7855602322261473e-05, "loss": 0.0088, "step": 59920 }, { "grad_norm": 0.13031446933746338, "learning_rate": 3.783956343070746e-05, "loss": 0.0089, "step": 59930 }, { "grad_norm": 0.12398822605609894, "learning_rate": 3.7823525869000526e-05, "loss": 0.0087, "step": 59940 }, { "grad_norm": 0.1821839064359665, "learning_rate": 3.78074896388945e-05, "loss": 0.0093, "step": 59950 }, { "grad_norm": 0.1525343507528305, "learning_rate": 3.7791454742143075e-05, "loss": 0.0078, "step": 59960 }, { "grad_norm": 0.16520732641220093, "learning_rate": 3.7775421180499795e-05, "loss": 0.0074, "step": 59970 }, { "grad_norm": 0.20115859806537628, "learning_rate": 3.7759388955718086e-05, "loss": 0.011, "step": 59980 }, { "grad_norm": 0.1401427835226059, "learning_rate": 3.774335806955117e-05, "loss": 0.008, "step": 59990 }, { "grad_norm": 0.14385944604873657, "learning_rate": 3.7727328523752215e-05, "loss": 0.0098, "step": 60000 } ], "logging_steps": 10, "max_steps": 100000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 10000, "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": 24, "trial_name": null, "trial_params": null }