{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 100000, "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 }, { "grad_norm": 0.1335127055644989, "learning_rate": 3.771130032007415e-05, "loss": 0.0082, "step": 60010 }, { "grad_norm": 0.1340048909187317, "learning_rate": 3.769527346026982e-05, "loss": 0.0095, "step": 60020 }, { "grad_norm": 0.22054123878479004, "learning_rate": 3.767924794609188e-05, "loss": 0.0125, "step": 60030 }, { "grad_norm": 0.11846527457237244, "learning_rate": 3.7663223779292854e-05, "loss": 0.0116, "step": 60040 }, { "grad_norm": 0.14558298885822296, "learning_rate": 3.764720096162512e-05, "loss": 0.0102, "step": 60050 }, { "grad_norm": 0.18143846094608307, "learning_rate": 3.763117949484092e-05, "loss": 0.0096, "step": 60060 }, { "grad_norm": 0.14561504125595093, "learning_rate": 3.7615159380692344e-05, "loss": 0.0088, "step": 60070 }, { "grad_norm": 0.16367170214653015, "learning_rate": 3.759914062093132e-05, "loss": 0.0084, "step": 60080 }, { "grad_norm": 0.16034142673015594, "learning_rate": 3.7583123217309615e-05, "loss": 0.0103, "step": 60090 }, { "grad_norm": 0.13069818913936615, "learning_rate": 3.7567107171578904e-05, "loss": 0.011, "step": 60100 }, { "grad_norm": 0.1733730286359787, "learning_rate": 3.755109248549066e-05, "loss": 0.0109, "step": 60110 }, { "grad_norm": 0.19744405150413513, "learning_rate": 3.75350791607962e-05, "loss": 0.0128, "step": 60120 }, { "grad_norm": 0.15747906267642975, "learning_rate": 3.751906719924676e-05, "loss": 0.011, "step": 60130 }, { "grad_norm": 0.17248764634132385, "learning_rate": 3.750305660259337e-05, "loss": 0.0095, "step": 60140 }, { "grad_norm": 0.12980276346206665, "learning_rate": 3.748704737258693e-05, "loss": 0.0079, "step": 60150 }, { "grad_norm": 0.11746487021446228, "learning_rate": 3.747103951097816e-05, "loss": 0.0092, "step": 60160 }, { "grad_norm": 0.1315392404794693, "learning_rate": 3.745503301951768e-05, "loss": 0.0116, "step": 60170 }, { "grad_norm": 0.12753675878047943, "learning_rate": 3.743902789995592e-05, "loss": 0.0097, "step": 60180 }, { "grad_norm": 0.12809233367443085, "learning_rate": 3.7423024154043183e-05, "loss": 0.0102, "step": 60190 }, { "grad_norm": 0.1507616937160492, "learning_rate": 3.740702178352963e-05, "loss": 0.0119, "step": 60200 }, { "grad_norm": 0.3335401713848114, "learning_rate": 3.739102079016523e-05, "loss": 0.0143, "step": 60210 }, { "grad_norm": 0.18236511945724487, "learning_rate": 3.7375021175699846e-05, "loss": 0.0114, "step": 60220 }, { "grad_norm": 0.17383693158626556, "learning_rate": 3.735902294188317e-05, "loss": 0.0104, "step": 60230 }, { "grad_norm": 0.14663375914096832, "learning_rate": 3.734302609046473e-05, "loss": 0.0102, "step": 60240 }, { "grad_norm": 0.22541800141334534, "learning_rate": 3.732703062319391e-05, "loss": 0.0102, "step": 60250 }, { "grad_norm": 0.17343731224536896, "learning_rate": 3.731103654181999e-05, "loss": 0.0084, "step": 60260 }, { "grad_norm": 0.1247483566403389, "learning_rate": 3.729504384809203e-05, "loss": 0.0076, "step": 60270 }, { "grad_norm": 0.1606508046388626, "learning_rate": 3.727905254375898e-05, "loss": 0.0162, "step": 60280 }, { "grad_norm": 0.17295774817466736, "learning_rate": 3.7263062630569614e-05, "loss": 0.0106, "step": 60290 }, { "grad_norm": 0.13996608555316925, "learning_rate": 3.724707411027256e-05, "loss": 0.0098, "step": 60300 }, { "grad_norm": 0.1228455975651741, "learning_rate": 3.723108698461631e-05, "loss": 0.0093, "step": 60310 }, { "grad_norm": 0.14798696339130402, "learning_rate": 3.72151012553492e-05, "loss": 0.0115, "step": 60320 }, { "grad_norm": 0.12263181060552597, "learning_rate": 3.719911692421939e-05, "loss": 0.0096, "step": 60330 }, { "grad_norm": 0.14047707617282867, "learning_rate": 3.71831339929749e-05, "loss": 0.0153, "step": 60340 }, { "grad_norm": 0.1357598453760147, "learning_rate": 3.716715246336362e-05, "loss": 0.0089, "step": 60350 }, { "grad_norm": 0.18810777366161346, "learning_rate": 3.715117233713324e-05, "loss": 0.0123, "step": 60360 }, { "grad_norm": 0.21195676922798157, "learning_rate": 3.713519361603135e-05, "loss": 0.0114, "step": 60370 }, { "grad_norm": 0.15395335853099823, "learning_rate": 3.711921630180532e-05, "loss": 0.0098, "step": 60380 }, { "grad_norm": 0.12314858287572861, "learning_rate": 3.710324039620245e-05, "loss": 0.011, "step": 60390 }, { "grad_norm": 0.1642114669084549, "learning_rate": 3.708726590096982e-05, "loss": 0.0138, "step": 60400 }, { "grad_norm": 0.17818517982959747, "learning_rate": 3.707129281785437e-05, "loss": 0.0122, "step": 60410 }, { "grad_norm": 0.3282080292701721, "learning_rate": 3.705532114860291e-05, "loss": 0.0098, "step": 60420 }, { "grad_norm": 0.12509116530418396, "learning_rate": 3.703935089496207e-05, "loss": 0.0101, "step": 60430 }, { "grad_norm": 0.13403767347335815, "learning_rate": 3.702338205867829e-05, "loss": 0.0148, "step": 60440 }, { "grad_norm": 0.17445813119411469, "learning_rate": 3.7007414641497976e-05, "loss": 0.0112, "step": 60450 }, { "grad_norm": 0.1674208641052246, "learning_rate": 3.699144864516726e-05, "loss": 0.0077, "step": 60460 }, { "grad_norm": 0.1700284630060196, "learning_rate": 3.6975484071432153e-05, "loss": 0.0108, "step": 60470 }, { "grad_norm": 0.1735779047012329, "learning_rate": 3.695952092203854e-05, "loss": 0.0108, "step": 60480 }, { "grad_norm": 0.1464967280626297, "learning_rate": 3.69435591987321e-05, "loss": 0.0099, "step": 60490 }, { "grad_norm": 0.12397091090679169, "learning_rate": 3.6927598903258374e-05, "loss": 0.018, "step": 60500 }, { "grad_norm": 0.15436305105686188, "learning_rate": 3.691164003736281e-05, "loss": 0.0087, "step": 60510 }, { "grad_norm": 0.11498961597681046, "learning_rate": 3.689568260279061e-05, "loss": 0.0128, "step": 60520 }, { "grad_norm": 0.11064460128545761, "learning_rate": 3.687972660128686e-05, "loss": 0.0147, "step": 60530 }, { "grad_norm": 0.09854669123888016, "learning_rate": 3.686377203459648e-05, "loss": 0.0095, "step": 60540 }, { "grad_norm": 0.13809095323085785, "learning_rate": 3.6847818904464226e-05, "loss": 0.0166, "step": 60550 }, { "grad_norm": 0.15451325476169586, "learning_rate": 3.683186721263474e-05, "loss": 0.0099, "step": 60560 }, { "grad_norm": 0.1898036003112793, "learning_rate": 3.6815916960852435e-05, "loss": 0.0135, "step": 60570 }, { "grad_norm": 0.13273441791534424, "learning_rate": 3.679996815086165e-05, "loss": 0.0102, "step": 60580 }, { "grad_norm": 0.14347633719444275, "learning_rate": 3.678402078440649e-05, "loss": 0.0093, "step": 60590 }, { "grad_norm": 0.21030285954475403, "learning_rate": 3.676807486323096e-05, "loss": 0.0099, "step": 60600 }, { "grad_norm": 0.2955179810523987, "learning_rate": 3.6752130389078864e-05, "loss": 0.0096, "step": 60610 }, { "grad_norm": 0.12851521372795105, "learning_rate": 3.673618736369388e-05, "loss": 0.0081, "step": 60620 }, { "grad_norm": 0.13398882746696472, "learning_rate": 3.6720245788819474e-05, "loss": 0.0103, "step": 60630 }, { "grad_norm": 0.14572902023792267, "learning_rate": 3.6704305666199044e-05, "loss": 0.012, "step": 60640 }, { "grad_norm": 0.21140210330486298, "learning_rate": 3.668836699757576e-05, "loss": 0.0122, "step": 60650 }, { "grad_norm": 0.12876936793327332, "learning_rate": 3.667242978469265e-05, "loss": 0.0098, "step": 60660 }, { "grad_norm": 0.1141122579574585, "learning_rate": 3.6656494029292575e-05, "loss": 0.0099, "step": 60670 }, { "grad_norm": 0.12387625873088837, "learning_rate": 3.664055973311825e-05, "loss": 0.0113, "step": 60680 }, { "grad_norm": 0.1546521931886673, "learning_rate": 3.662462689791221e-05, "loss": 0.0091, "step": 60690 }, { "grad_norm": 0.19698737561702728, "learning_rate": 3.660869552541689e-05, "loss": 0.0115, "step": 60700 }, { "grad_norm": 0.17799459397792816, "learning_rate": 3.659276561737448e-05, "loss": 0.0117, "step": 60710 }, { "grad_norm": 0.1674104928970337, "learning_rate": 3.657683717552708e-05, "loss": 0.0118, "step": 60720 }, { "grad_norm": 0.15156756341457367, "learning_rate": 3.6560910201616574e-05, "loss": 0.0109, "step": 60730 }, { "grad_norm": 0.15368223190307617, "learning_rate": 3.654498469738472e-05, "loss": 0.0093, "step": 60740 }, { "grad_norm": 0.15361034870147705, "learning_rate": 3.6529060664573126e-05, "loss": 0.0098, "step": 60750 }, { "grad_norm": 0.11168870329856873, "learning_rate": 3.6513138104923176e-05, "loss": 0.0093, "step": 60760 }, { "grad_norm": 0.12519405782222748, "learning_rate": 3.6497217020176176e-05, "loss": 0.0104, "step": 60770 }, { "grad_norm": 0.24677039682865143, "learning_rate": 3.648129741207323e-05, "loss": 0.0105, "step": 60780 }, { "grad_norm": 0.11153345555067062, "learning_rate": 3.646537928235527e-05, "loss": 0.0077, "step": 60790 }, { "grad_norm": 0.18846599757671356, "learning_rate": 3.6449462632763075e-05, "loss": 0.0095, "step": 60800 }, { "grad_norm": 0.1896664947271347, "learning_rate": 3.643354746503729e-05, "loss": 0.0104, "step": 60810 }, { "grad_norm": 0.15756289660930634, "learning_rate": 3.6417633780918345e-05, "loss": 0.0094, "step": 60820 }, { "grad_norm": 0.10731121152639389, "learning_rate": 3.640172158214655e-05, "loss": 0.0108, "step": 60830 }, { "grad_norm": 0.1495252549648285, "learning_rate": 3.6385810870462035e-05, "loss": 0.0095, "step": 60840 }, { "grad_norm": 0.15526391565799713, "learning_rate": 3.636990164760479e-05, "loss": 0.007, "step": 60850 }, { "grad_norm": 0.1355343908071518, "learning_rate": 3.6353993915314596e-05, "loss": 0.007, "step": 60860 }, { "grad_norm": 0.19195109605789185, "learning_rate": 3.6338087675331124e-05, "loss": 0.0077, "step": 60870 }, { "grad_norm": 0.13252247869968414, "learning_rate": 3.632218292939383e-05, "loss": 0.0089, "step": 60880 }, { "grad_norm": 0.10023027658462524, "learning_rate": 3.6306279679242033e-05, "loss": 0.0083, "step": 60890 }, { "grad_norm": 0.1654358059167862, "learning_rate": 3.629037792661492e-05, "loss": 0.0079, "step": 60900 }, { "grad_norm": 0.12736232578754425, "learning_rate": 3.627447767325147e-05, "loss": 0.0078, "step": 60910 }, { "grad_norm": 0.15546977519989014, "learning_rate": 3.62585789208905e-05, "loss": 0.0096, "step": 60920 }, { "grad_norm": 0.14530789852142334, "learning_rate": 3.624268167127068e-05, "loss": 0.0095, "step": 60930 }, { "grad_norm": 0.2858467698097229, "learning_rate": 3.622678592613051e-05, "loss": 0.0104, "step": 60940 }, { "grad_norm": 0.16849668323993683, "learning_rate": 3.6210891687208304e-05, "loss": 0.0116, "step": 60950 }, { "grad_norm": 0.1598639041185379, "learning_rate": 3.619499895624228e-05, "loss": 0.0081, "step": 60960 }, { "grad_norm": 0.12854903936386108, "learning_rate": 3.61791077349704e-05, "loss": 0.0097, "step": 60970 }, { "grad_norm": 0.14970539510250092, "learning_rate": 3.616321802513053e-05, "loss": 0.0108, "step": 60980 }, { "grad_norm": 0.12796346843242645, "learning_rate": 3.6147329828460325e-05, "loss": 0.0102, "step": 60990 }, { "grad_norm": 0.1655004918575287, "learning_rate": 3.6131443146697306e-05, "loss": 0.0136, "step": 61000 }, { "grad_norm": 0.18174229562282562, "learning_rate": 3.611555798157879e-05, "loss": 0.0119, "step": 61010 }, { "grad_norm": 0.22256731986999512, "learning_rate": 3.6099674334842e-05, "loss": 0.0099, "step": 61020 }, { "grad_norm": 0.12572534382343292, "learning_rate": 3.6083792208223934e-05, "loss": 0.0106, "step": 61030 }, { "grad_norm": 0.17724885046482086, "learning_rate": 3.606791160346142e-05, "loss": 0.0102, "step": 61040 }, { "grad_norm": 0.1674247831106186, "learning_rate": 3.6052032522291134e-05, "loss": 0.0102, "step": 61050 }, { "grad_norm": 0.1676291674375534, "learning_rate": 3.603615496644962e-05, "loss": 0.0112, "step": 61060 }, { "grad_norm": 0.1265888214111328, "learning_rate": 3.60202789376732e-05, "loss": 0.0072, "step": 61070 }, { "grad_norm": 0.14629192650318146, "learning_rate": 3.6004404437698036e-05, "loss": 0.0138, "step": 61080 }, { "grad_norm": 0.09985940903425217, "learning_rate": 3.5988531468260184e-05, "loss": 0.0094, "step": 61090 }, { "grad_norm": 0.2505533695220947, "learning_rate": 3.597266003109546e-05, "loss": 0.0084, "step": 61100 }, { "grad_norm": 0.13108788430690765, "learning_rate": 3.5956790127939557e-05, "loss": 0.0127, "step": 61110 }, { "grad_norm": 0.15213543176651, "learning_rate": 3.594092176052796e-05, "loss": 0.0115, "step": 61120 }, { "grad_norm": 0.15287061035633087, "learning_rate": 3.592505493059603e-05, "loss": 0.0091, "step": 61130 }, { "grad_norm": 0.18450647592544556, "learning_rate": 3.5909189639878905e-05, "loss": 0.0165, "step": 61140 }, { "grad_norm": 0.18762357532978058, "learning_rate": 3.5893325890111626e-05, "loss": 0.0105, "step": 61150 }, { "grad_norm": 0.12263856083154678, "learning_rate": 3.5877463683029034e-05, "loss": 0.0146, "step": 61160 }, { "grad_norm": 0.15892986953258514, "learning_rate": 3.586160302036578e-05, "loss": 0.0089, "step": 61170 }, { "grad_norm": 0.12149752676486969, "learning_rate": 3.584574390385636e-05, "loss": 0.0097, "step": 61180 }, { "grad_norm": 0.14832043647766113, "learning_rate": 3.582988633523511e-05, "loss": 0.0092, "step": 61190 }, { "grad_norm": 0.1683570146560669, "learning_rate": 3.581403031623616e-05, "loss": 0.0093, "step": 61200 }, { "grad_norm": 0.10496657341718674, "learning_rate": 3.579817584859354e-05, "loss": 0.0112, "step": 61210 }, { "grad_norm": 0.1765417754650116, "learning_rate": 3.578232293404107e-05, "loss": 0.0106, "step": 61220 }, { "grad_norm": 0.11300667375326157, "learning_rate": 3.5766471574312385e-05, "loss": 0.0076, "step": 61230 }, { "grad_norm": 0.13065524399280548, "learning_rate": 3.5750621771140964e-05, "loss": 0.008, "step": 61240 }, { "grad_norm": 0.13739360868930817, "learning_rate": 3.5734773526260115e-05, "loss": 0.0105, "step": 61250 }, { "grad_norm": 0.13654963672161102, "learning_rate": 3.5718926841402996e-05, "loss": 0.0127, "step": 61260 }, { "grad_norm": 0.15857060253620148, "learning_rate": 3.5703081718302555e-05, "loss": 0.0114, "step": 61270 }, { "grad_norm": 0.19427692890167236, "learning_rate": 3.568723815869162e-05, "loss": 0.0123, "step": 61280 }, { "grad_norm": 0.14882716536521912, "learning_rate": 3.5671396164302777e-05, "loss": 0.0098, "step": 61290 }, { "grad_norm": 0.15829381346702576, "learning_rate": 3.565555573686853e-05, "loss": 0.0092, "step": 61300 }, { "grad_norm": 0.18023784458637238, "learning_rate": 3.563971687812113e-05, "loss": 0.013, "step": 61310 }, { "grad_norm": 0.13155993819236755, "learning_rate": 3.5623879589792706e-05, "loss": 0.0085, "step": 61320 }, { "grad_norm": 0.18100760877132416, "learning_rate": 3.5608043873615185e-05, "loss": 0.009, "step": 61330 }, { "grad_norm": 0.1598932147026062, "learning_rate": 3.559220973132035e-05, "loss": 0.0097, "step": 61340 }, { "grad_norm": 0.13728907704353333, "learning_rate": 3.557637716463981e-05, "loss": 0.0083, "step": 61350 }, { "grad_norm": 0.17300450801849365, "learning_rate": 3.556054617530497e-05, "loss": 0.0098, "step": 61360 }, { "grad_norm": 0.17459803819656372, "learning_rate": 3.5544716765047106e-05, "loss": 0.0121, "step": 61370 }, { "grad_norm": 0.20748856663703918, "learning_rate": 3.5528888935597274e-05, "loss": 0.009, "step": 61380 }, { "grad_norm": 0.14732322096824646, "learning_rate": 3.551306268868636e-05, "loss": 0.0088, "step": 61390 }, { "grad_norm": 0.19051745533943176, "learning_rate": 3.5497238026045166e-05, "loss": 0.0115, "step": 61400 }, { "grad_norm": 0.14374999701976776, "learning_rate": 3.548141494940421e-05, "loss": 0.0124, "step": 61410 }, { "grad_norm": 0.17022328078746796, "learning_rate": 3.546559346049388e-05, "loss": 0.0102, "step": 61420 }, { "grad_norm": 0.11314056068658829, "learning_rate": 3.544977356104441e-05, "loss": 0.0083, "step": 61430 }, { "grad_norm": 0.19412946701049805, "learning_rate": 3.543395525278582e-05, "loss": 0.0111, "step": 61440 }, { "grad_norm": 0.15017634630203247, "learning_rate": 3.5418138537447985e-05, "loss": 0.0121, "step": 61450 }, { "grad_norm": 0.12486223131418228, "learning_rate": 3.5402323416760574e-05, "loss": 0.0107, "step": 61460 }, { "grad_norm": 0.1253863424062729, "learning_rate": 3.538650989245316e-05, "loss": 0.0114, "step": 61470 }, { "grad_norm": 0.09712029993534088, "learning_rate": 3.537069796625504e-05, "loss": 0.0086, "step": 61480 }, { "grad_norm": 0.18045659363269806, "learning_rate": 3.53548876398954e-05, "loss": 0.0087, "step": 61490 }, { "grad_norm": 0.12593699991703033, "learning_rate": 3.5339078915103214e-05, "loss": 0.0112, "step": 61500 }, { "grad_norm": 0.1473025679588318, "learning_rate": 3.532327179360733e-05, "loss": 0.0108, "step": 61510 }, { "grad_norm": 0.13789303600788116, "learning_rate": 3.530746627713636e-05, "loss": 0.0094, "step": 61520 }, { "grad_norm": 0.11046163737773895, "learning_rate": 3.5291662367418805e-05, "loss": 0.0073, "step": 61530 }, { "grad_norm": 0.1199910119175911, "learning_rate": 3.527586006618293e-05, "loss": 0.0105, "step": 61540 }, { "grad_norm": 0.14892101287841797, "learning_rate": 3.526005937515687e-05, "loss": 0.0135, "step": 61550 }, { "grad_norm": 0.11949222534894943, "learning_rate": 3.524426029606856e-05, "loss": 0.0097, "step": 61560 }, { "grad_norm": 0.14344601333141327, "learning_rate": 3.5228462830645746e-05, "loss": 0.0079, "step": 61570 }, { "grad_norm": 0.19313181936740875, "learning_rate": 3.5212666980616036e-05, "loss": 0.0098, "step": 61580 }, { "grad_norm": 0.1357339322566986, "learning_rate": 3.519687274770681e-05, "loss": 0.009, "step": 61590 }, { "grad_norm": 0.1237134039402008, "learning_rate": 3.518108013364535e-05, "loss": 0.0117, "step": 61600 }, { "grad_norm": 0.14595940709114075, "learning_rate": 3.516528914015868e-05, "loss": 0.0125, "step": 61610 }, { "grad_norm": 0.16849234700202942, "learning_rate": 3.5149499768973694e-05, "loss": 0.0107, "step": 61620 }, { "grad_norm": 0.15896832942962646, "learning_rate": 3.513371202181708e-05, "loss": 0.0108, "step": 61630 }, { "grad_norm": 0.12114760279655457, "learning_rate": 3.511792590041537e-05, "loss": 0.0108, "step": 61640 }, { "grad_norm": 0.19606268405914307, "learning_rate": 3.510214140649488e-05, "loss": 0.0108, "step": 61650 }, { "grad_norm": 0.1495964080095291, "learning_rate": 3.508635854178184e-05, "loss": 0.013, "step": 61660 }, { "grad_norm": 0.129526287317276, "learning_rate": 3.50705773080022e-05, "loss": 0.0098, "step": 61670 }, { "grad_norm": 0.1482258290052414, "learning_rate": 3.505479770688178e-05, "loss": 0.0078, "step": 61680 }, { "grad_norm": 0.15818002820014954, "learning_rate": 3.503901974014622e-05, "loss": 0.0117, "step": 61690 }, { "grad_norm": 0.1799362152814865, "learning_rate": 3.502324340952095e-05, "loss": 0.0101, "step": 61700 }, { "grad_norm": 0.21691972017288208, "learning_rate": 3.500746871673126e-05, "loss": 0.0099, "step": 61710 }, { "grad_norm": 0.16161392629146576, "learning_rate": 3.499169566350227e-05, "loss": 0.0126, "step": 61720 }, { "grad_norm": 0.11746234446763992, "learning_rate": 3.4975924251558876e-05, "loss": 0.0104, "step": 61730 }, { "grad_norm": 0.25608426332473755, "learning_rate": 3.496015448262582e-05, "loss": 0.0149, "step": 61740 }, { "grad_norm": 0.131087526679039, "learning_rate": 3.494438635842764e-05, "loss": 0.0114, "step": 61750 }, { "grad_norm": 0.11958620697259903, "learning_rate": 3.492861988068874e-05, "loss": 0.0091, "step": 61760 }, { "grad_norm": 0.15430599451065063, "learning_rate": 3.491285505113332e-05, "loss": 0.0102, "step": 61770 }, { "grad_norm": 0.15728066861629486, "learning_rate": 3.4897091871485376e-05, "loss": 0.0098, "step": 61780 }, { "grad_norm": 0.17418678104877472, "learning_rate": 3.488133034346875e-05, "loss": 0.0102, "step": 61790 }, { "grad_norm": 0.15515829622745514, "learning_rate": 3.486557046880713e-05, "loss": 0.0094, "step": 61800 }, { "grad_norm": 0.1419200599193573, "learning_rate": 3.484981224922395e-05, "loss": 0.012, "step": 61810 }, { "grad_norm": 0.1344301551580429, "learning_rate": 3.4834055686442544e-05, "loss": 0.0085, "step": 61820 }, { "grad_norm": 0.12728066742420197, "learning_rate": 3.4818300782186e-05, "loss": 0.0084, "step": 61830 }, { "grad_norm": 0.12399087101221085, "learning_rate": 3.4802547538177236e-05, "loss": 0.0111, "step": 61840 }, { "grad_norm": 0.1793571412563324, "learning_rate": 3.478679595613905e-05, "loss": 0.0118, "step": 61850 }, { "grad_norm": 0.10472247004508972, "learning_rate": 3.4771046037793975e-05, "loss": 0.0064, "step": 61860 }, { "grad_norm": 0.15617376565933228, "learning_rate": 3.4755297784864416e-05, "loss": 0.0089, "step": 61870 }, { "grad_norm": 0.1442759484052658, "learning_rate": 3.473955119907257e-05, "loss": 0.009, "step": 61880 }, { "grad_norm": 0.1104407086968422, "learning_rate": 3.472380628214046e-05, "loss": 0.008, "step": 61890 }, { "grad_norm": 0.13990485668182373, "learning_rate": 3.4708063035789906e-05, "loss": 0.0099, "step": 61900 }, { "grad_norm": 0.1680605113506317, "learning_rate": 3.4692321461742614e-05, "loss": 0.0085, "step": 61910 }, { "grad_norm": 0.12206948548555374, "learning_rate": 3.4676581561720023e-05, "loss": 0.0107, "step": 61920 }, { "grad_norm": 0.1409778743982315, "learning_rate": 3.466084333744344e-05, "loss": 0.0077, "step": 61930 }, { "grad_norm": 0.11569545418024063, "learning_rate": 3.4645106790633964e-05, "loss": 0.0112, "step": 61940 }, { "grad_norm": 0.11114251613616943, "learning_rate": 3.4629371923012526e-05, "loss": 0.0095, "step": 61950 }, { "grad_norm": 0.14462745189666748, "learning_rate": 3.4613638736299856e-05, "loss": 0.0096, "step": 61960 }, { "grad_norm": 0.12777280807495117, "learning_rate": 3.459790723221652e-05, "loss": 0.0099, "step": 61970 }, { "grad_norm": 0.14800803363323212, "learning_rate": 3.45821774124829e-05, "loss": 0.0131, "step": 61980 }, { "grad_norm": 0.15700751543045044, "learning_rate": 3.456644927881918e-05, "loss": 0.0086, "step": 61990 }, { "grad_norm": 0.14160175621509552, "learning_rate": 3.4550722832945354e-05, "loss": 0.0092, "step": 62000 }, { "grad_norm": 0.13229548931121826, "learning_rate": 3.453499807658126e-05, "loss": 0.0103, "step": 62010 }, { "grad_norm": 0.15488198399543762, "learning_rate": 3.451927501144653e-05, "loss": 0.0101, "step": 62020 }, { "grad_norm": 0.19220630824565887, "learning_rate": 3.450355363926059e-05, "loss": 0.0097, "step": 62030 }, { "grad_norm": 0.14358346164226532, "learning_rate": 3.448783396174273e-05, "loss": 0.0095, "step": 62040 }, { "grad_norm": 0.17330673336982727, "learning_rate": 3.4472115980612034e-05, "loss": 0.0093, "step": 62050 }, { "grad_norm": 0.13343079388141632, "learning_rate": 3.4456399697587396e-05, "loss": 0.0088, "step": 62060 }, { "grad_norm": 0.18589328229427338, "learning_rate": 3.4440685114387506e-05, "loss": 0.0125, "step": 62070 }, { "grad_norm": 0.19384482502937317, "learning_rate": 3.442497223273091e-05, "loss": 0.0124, "step": 62080 }, { "grad_norm": 0.13857926428318024, "learning_rate": 3.4409261054335926e-05, "loss": 0.0101, "step": 62090 }, { "grad_norm": 0.1259365826845169, "learning_rate": 3.439355158092069e-05, "loss": 0.0106, "step": 62100 }, { "grad_norm": 0.1397615522146225, "learning_rate": 3.437784381420322e-05, "loss": 0.0092, "step": 62110 }, { "grad_norm": 0.11935430765151978, "learning_rate": 3.436213775590126e-05, "loss": 0.0083, "step": 62120 }, { "grad_norm": 0.1323608160018921, "learning_rate": 3.4346433407732405e-05, "loss": 0.0089, "step": 62130 }, { "grad_norm": 0.13802459836006165, "learning_rate": 3.433073077141405e-05, "loss": 0.0082, "step": 62140 }, { "grad_norm": 0.11988148093223572, "learning_rate": 3.431502984866343e-05, "loss": 0.0126, "step": 62150 }, { "grad_norm": 0.1165241152048111, "learning_rate": 3.429933064119753e-05, "loss": 0.0084, "step": 62160 }, { "grad_norm": 0.12379521131515503, "learning_rate": 3.4283633150733255e-05, "loss": 0.0088, "step": 62170 }, { "grad_norm": 0.1120690181851387, "learning_rate": 3.4267937378987225e-05, "loss": 0.0076, "step": 62180 }, { "grad_norm": 0.19184444844722748, "learning_rate": 3.425224332767592e-05, "loss": 0.0105, "step": 62190 }, { "grad_norm": 0.14317554235458374, "learning_rate": 3.4236550998515595e-05, "loss": 0.0098, "step": 62200 }, { "grad_norm": 0.18687552213668823, "learning_rate": 3.422086039322235e-05, "loss": 0.0088, "step": 62210 }, { "grad_norm": 0.17627157270908356, "learning_rate": 3.420517151351209e-05, "loss": 0.0123, "step": 62220 }, { "grad_norm": 0.1359926164150238, "learning_rate": 3.418948436110053e-05, "loss": 0.0107, "step": 62230 }, { "grad_norm": 0.2083519995212555, "learning_rate": 3.417379893770319e-05, "loss": 0.0094, "step": 62240 }, { "grad_norm": 0.1405269056558609, "learning_rate": 3.41581152450354e-05, "loss": 0.0083, "step": 62250 }, { "grad_norm": 0.1443355232477188, "learning_rate": 3.414243328481231e-05, "loss": 0.0098, "step": 62260 }, { "grad_norm": 0.13408370316028595, "learning_rate": 3.412675305874888e-05, "loss": 0.0092, "step": 62270 }, { "grad_norm": 0.16587507724761963, "learning_rate": 3.4111074568559876e-05, "loss": 0.0101, "step": 62280 }, { "grad_norm": 0.1399962604045868, "learning_rate": 3.4095397815959846e-05, "loss": 0.0092, "step": 62290 }, { "grad_norm": 0.14799077808856964, "learning_rate": 3.407972280266322e-05, "loss": 0.0101, "step": 62300 }, { "grad_norm": 0.1077803447842598, "learning_rate": 3.4064049530384166e-05, "loss": 0.0114, "step": 62310 }, { "grad_norm": 0.25321945548057556, "learning_rate": 3.40483780008367e-05, "loss": 0.0107, "step": 62320 }, { "grad_norm": 0.13374063372612, "learning_rate": 3.4032708215734635e-05, "loss": 0.01, "step": 62330 }, { "grad_norm": 0.18959975242614746, "learning_rate": 3.4017040176791584e-05, "loss": 0.0092, "step": 62340 }, { "grad_norm": 0.11483656615018845, "learning_rate": 3.4001373885720976e-05, "loss": 0.0081, "step": 62350 }, { "grad_norm": 0.13321992754936218, "learning_rate": 3.398570934423608e-05, "loss": 0.0114, "step": 62360 }, { "grad_norm": 0.14857645332813263, "learning_rate": 3.397004655404994e-05, "loss": 0.0091, "step": 62370 }, { "grad_norm": 0.14765995740890503, "learning_rate": 3.395438551687541e-05, "loss": 0.0076, "step": 62380 }, { "grad_norm": 0.13530735671520233, "learning_rate": 3.393872623442515e-05, "loss": 0.0067, "step": 62390 }, { "grad_norm": 0.14613239467144012, "learning_rate": 3.392306870841164e-05, "loss": 0.01, "step": 62400 }, { "grad_norm": 0.16468928754329681, "learning_rate": 3.390741294054715e-05, "loss": 0.0115, "step": 62410 }, { "grad_norm": 0.16883999109268188, "learning_rate": 3.38917589325438e-05, "loss": 0.0096, "step": 62420 }, { "grad_norm": 0.13016937673091888, "learning_rate": 3.387610668611349e-05, "loss": 0.008, "step": 62430 }, { "grad_norm": 0.16419748961925507, "learning_rate": 3.38604562029679e-05, "loss": 0.0139, "step": 62440 }, { "grad_norm": 0.12434813380241394, "learning_rate": 3.384480748481855e-05, "loss": 0.0112, "step": 62450 }, { "grad_norm": 0.16122649610042572, "learning_rate": 3.382916053337676e-05, "loss": 0.0091, "step": 62460 }, { "grad_norm": 0.10558974742889404, "learning_rate": 3.381351535035366e-05, "loss": 0.0135, "step": 62470 }, { "grad_norm": 0.108008973300457, "learning_rate": 3.379787193746019e-05, "loss": 0.0105, "step": 62480 }, { "grad_norm": 0.12937963008880615, "learning_rate": 3.378223029640709e-05, "loss": 0.0101, "step": 62490 }, { "grad_norm": 0.10638970881700516, "learning_rate": 3.376659042890489e-05, "loss": 0.0094, "step": 62500 }, { "grad_norm": 0.16582967340946198, "learning_rate": 3.375095233666397e-05, "loss": 0.0096, "step": 62510 }, { "grad_norm": 0.12818317115306854, "learning_rate": 3.373531602139447e-05, "loss": 0.0073, "step": 62520 }, { "grad_norm": 0.10378482937812805, "learning_rate": 3.371968148480635e-05, "loss": 0.0078, "step": 62530 }, { "grad_norm": 0.14309515058994293, "learning_rate": 3.370404872860937e-05, "loss": 0.0121, "step": 62540 }, { "grad_norm": 0.10600723326206207, "learning_rate": 3.3688417754513145e-05, "loss": 0.0081, "step": 62550 }, { "grad_norm": 0.17741243541240692, "learning_rate": 3.367278856422703e-05, "loss": 0.0174, "step": 62560 }, { "grad_norm": 0.16010713577270508, "learning_rate": 3.365716115946021e-05, "loss": 0.012, "step": 62570 }, { "grad_norm": 0.14922909438610077, "learning_rate": 3.3641535541921666e-05, "loss": 0.011, "step": 62580 }, { "grad_norm": 0.1478927880525589, "learning_rate": 3.36259117133202e-05, "loss": 0.0121, "step": 62590 }, { "grad_norm": 0.17451247572898865, "learning_rate": 3.36102896753644e-05, "loss": 0.0119, "step": 62600 }, { "grad_norm": 0.123294897377491, "learning_rate": 3.35946694297627e-05, "loss": 0.0096, "step": 62610 }, { "grad_norm": 0.11343201994895935, "learning_rate": 3.357905097822328e-05, "loss": 0.0077, "step": 62620 }, { "grad_norm": 0.22223682701587677, "learning_rate": 3.356343432245416e-05, "loss": 0.0098, "step": 62630 }, { "grad_norm": 0.11658748239278793, "learning_rate": 3.354781946416315e-05, "loss": 0.0095, "step": 62640 }, { "grad_norm": 0.11851099133491516, "learning_rate": 3.3532206405057864e-05, "loss": 0.009, "step": 62650 }, { "grad_norm": 0.13256551325321198, "learning_rate": 3.351659514684573e-05, "loss": 0.0095, "step": 62660 }, { "grad_norm": 0.12358596920967102, "learning_rate": 3.350098569123395e-05, "loss": 0.0107, "step": 62670 }, { "grad_norm": 0.1793200671672821, "learning_rate": 3.348537803992958e-05, "loss": 0.0094, "step": 62680 }, { "grad_norm": 0.19309167563915253, "learning_rate": 3.346977219463945e-05, "loss": 0.0132, "step": 62690 }, { "grad_norm": 0.14425845444202423, "learning_rate": 3.345416815707017e-05, "loss": 0.0093, "step": 62700 }, { "grad_norm": 0.13995057344436646, "learning_rate": 3.3438565928928185e-05, "loss": 0.0111, "step": 62710 }, { "grad_norm": 0.2154327929019928, "learning_rate": 3.3422965511919736e-05, "loss": 0.0121, "step": 62720 }, { "grad_norm": 0.18338307738304138, "learning_rate": 3.3407366907750846e-05, "loss": 0.0107, "step": 62730 }, { "grad_norm": 0.16699764132499695, "learning_rate": 3.339177011812737e-05, "loss": 0.0122, "step": 62740 }, { "grad_norm": 0.16453948616981506, "learning_rate": 3.3376175144754935e-05, "loss": 0.0097, "step": 62750 }, { "grad_norm": 0.1541115641593933, "learning_rate": 3.3360581989339004e-05, "loss": 0.0116, "step": 62760 }, { "grad_norm": 0.2358064353466034, "learning_rate": 3.3344990653584804e-05, "loss": 0.0133, "step": 62770 }, { "grad_norm": 0.15767067670822144, "learning_rate": 3.332940113919739e-05, "loss": 0.0132, "step": 62780 }, { "grad_norm": 0.19919143617153168, "learning_rate": 3.331381344788161e-05, "loss": 0.0133, "step": 62790 }, { "grad_norm": 0.15591907501220703, "learning_rate": 3.3298227581342076e-05, "loss": 0.0084, "step": 62800 }, { "grad_norm": 0.1752486377954483, "learning_rate": 3.328264354128329e-05, "loss": 0.0093, "step": 62810 }, { "grad_norm": 0.13839900493621826, "learning_rate": 3.3267061329409465e-05, "loss": 0.011, "step": 62820 }, { "grad_norm": 0.1608147770166397, "learning_rate": 3.325148094742466e-05, "loss": 0.0093, "step": 62830 }, { "grad_norm": 0.13852691650390625, "learning_rate": 3.323590239703271e-05, "loss": 0.0147, "step": 62840 }, { "grad_norm": 0.16552069783210754, "learning_rate": 3.322032567993728e-05, "loss": 0.0097, "step": 62850 }, { "grad_norm": 0.16148358583450317, "learning_rate": 3.3204750797841774e-05, "loss": 0.0093, "step": 62860 }, { "grad_norm": 0.10991569608449936, "learning_rate": 3.31891777524495e-05, "loss": 0.009, "step": 62870 }, { "grad_norm": 0.21120861172676086, "learning_rate": 3.317360654546347e-05, "loss": 0.0125, "step": 62880 }, { "grad_norm": 0.10844092816114426, "learning_rate": 3.3158037178586534e-05, "loss": 0.0081, "step": 62890 }, { "grad_norm": 0.180018350481987, "learning_rate": 3.3142469653521316e-05, "loss": 0.0097, "step": 62900 }, { "grad_norm": 0.11199767142534256, "learning_rate": 3.3126903971970284e-05, "loss": 0.0095, "step": 62910 }, { "grad_norm": 0.14873096346855164, "learning_rate": 3.311134013563564e-05, "loss": 0.0112, "step": 62920 }, { "grad_norm": 0.17920619249343872, "learning_rate": 3.309577814621947e-05, "loss": 0.0092, "step": 62930 }, { "grad_norm": 0.1592942625284195, "learning_rate": 3.308021800542359e-05, "loss": 0.0102, "step": 62940 }, { "grad_norm": 0.1354142725467682, "learning_rate": 3.306465971494963e-05, "loss": 0.0099, "step": 62950 }, { "grad_norm": 0.11552229523658752, "learning_rate": 3.304910327649901e-05, "loss": 0.0085, "step": 62960 }, { "grad_norm": 0.15426722168922424, "learning_rate": 3.303354869177297e-05, "loss": 0.0103, "step": 62970 }, { "grad_norm": 0.1369110643863678, "learning_rate": 3.301799596247255e-05, "loss": 0.0098, "step": 62980 }, { "grad_norm": 0.15900079905986786, "learning_rate": 3.300244509029854e-05, "loss": 0.0087, "step": 62990 }, { "grad_norm": 0.14811182022094727, "learning_rate": 3.2986896076951585e-05, "loss": 0.0116, "step": 63000 }, { "grad_norm": 0.15612907707691193, "learning_rate": 3.297134892413211e-05, "loss": 0.0088, "step": 63010 }, { "grad_norm": 0.14730732142925262, "learning_rate": 3.295580363354031e-05, "loss": 0.0096, "step": 63020 }, { "grad_norm": 0.15788483619689941, "learning_rate": 3.2940260206876194e-05, "loss": 0.0075, "step": 63030 }, { "grad_norm": 0.12592267990112305, "learning_rate": 3.292471864583958e-05, "loss": 0.0085, "step": 63040 }, { "grad_norm": 0.14478744566440582, "learning_rate": 3.290917895213004e-05, "loss": 0.0084, "step": 63050 }, { "grad_norm": 0.14025330543518066, "learning_rate": 3.289364112744701e-05, "loss": 0.0096, "step": 63060 }, { "grad_norm": 0.2186822146177292, "learning_rate": 3.287810517348967e-05, "loss": 0.013, "step": 63070 }, { "grad_norm": 0.14886076748371124, "learning_rate": 3.2862571091957014e-05, "loss": 0.0102, "step": 63080 }, { "grad_norm": 0.2891063392162323, "learning_rate": 3.284703888454781e-05, "loss": 0.0125, "step": 63090 }, { "grad_norm": 0.2125525176525116, "learning_rate": 3.2831508552960646e-05, "loss": 0.0119, "step": 63100 }, { "grad_norm": 0.12792260944843292, "learning_rate": 3.281598009889387e-05, "loss": 0.0096, "step": 63110 }, { "grad_norm": 0.14745743572711945, "learning_rate": 3.28004535240457e-05, "loss": 0.0086, "step": 63120 }, { "grad_norm": 0.13801580667495728, "learning_rate": 3.278492883011407e-05, "loss": 0.0071, "step": 63130 }, { "grad_norm": 0.14410071074962616, "learning_rate": 3.276940601879675e-05, "loss": 0.008, "step": 63140 }, { "grad_norm": 0.17224375903606415, "learning_rate": 3.275388509179127e-05, "loss": 0.0088, "step": 63150 }, { "grad_norm": 0.11053384095430374, "learning_rate": 3.2738366050794985e-05, "loss": 0.0054, "step": 63160 }, { "grad_norm": 0.2037811279296875, "learning_rate": 3.2722848897505034e-05, "loss": 0.0084, "step": 63170 }, { "grad_norm": 0.16103994846343994, "learning_rate": 3.2707333633618356e-05, "loss": 0.0085, "step": 63180 }, { "grad_norm": 0.12807543575763702, "learning_rate": 3.269182026083167e-05, "loss": 0.0072, "step": 63190 }, { "grad_norm": 0.13828688859939575, "learning_rate": 3.2676308780841516e-05, "loss": 0.0096, "step": 63200 }, { "grad_norm": 0.17561344802379608, "learning_rate": 3.266079919534416e-05, "loss": 0.0098, "step": 63210 }, { "grad_norm": 0.17954233288764954, "learning_rate": 3.264529150603575e-05, "loss": 0.0091, "step": 63220 }, { "grad_norm": 0.15097588300704956, "learning_rate": 3.262978571461217e-05, "loss": 0.0102, "step": 63230 }, { "grad_norm": 0.15834493935108185, "learning_rate": 3.26142818227691e-05, "loss": 0.0139, "step": 63240 }, { "grad_norm": 0.21638444066047668, "learning_rate": 3.259877983220203e-05, "loss": 0.0085, "step": 63250 }, { "grad_norm": 0.1795503944158554, "learning_rate": 3.2583279744606246e-05, "loss": 0.0089, "step": 63260 }, { "grad_norm": 0.1611493080854416, "learning_rate": 3.25677815616768e-05, "loss": 0.0104, "step": 63270 }, { "grad_norm": 0.1323973536491394, "learning_rate": 3.255228528510856e-05, "loss": 0.0111, "step": 63280 }, { "grad_norm": 0.1300734430551529, "learning_rate": 3.253679091659617e-05, "loss": 0.0108, "step": 63290 }, { "grad_norm": 0.2170945256948471, "learning_rate": 3.252129845783404e-05, "loss": 0.0112, "step": 63300 }, { "grad_norm": 0.18132376670837402, "learning_rate": 3.2505807910516456e-05, "loss": 0.0088, "step": 63310 }, { "grad_norm": 0.1571640819311142, "learning_rate": 3.249031927633742e-05, "loss": 0.0127, "step": 63320 }, { "grad_norm": 0.1781836599111557, "learning_rate": 3.2474832556990737e-05, "loss": 0.0083, "step": 63330 }, { "grad_norm": 0.23299559950828552, "learning_rate": 3.245934775417002e-05, "loss": 0.0124, "step": 63340 }, { "grad_norm": 0.16404613852500916, "learning_rate": 3.244386486956866e-05, "loss": 0.008, "step": 63350 }, { "grad_norm": 0.14140093326568604, "learning_rate": 3.2428383904879846e-05, "loss": 0.0116, "step": 63360 }, { "grad_norm": 0.18239933252334595, "learning_rate": 3.241290486179652e-05, "loss": 0.0069, "step": 63370 }, { "grad_norm": 0.1715167760848999, "learning_rate": 3.239742774201151e-05, "loss": 0.0098, "step": 63380 }, { "grad_norm": 0.125681072473526, "learning_rate": 3.238195254721733e-05, "loss": 0.0104, "step": 63390 }, { "grad_norm": 0.1650896519422531, "learning_rate": 3.236647927910633e-05, "loss": 0.0096, "step": 63400 }, { "grad_norm": 0.22695329785346985, "learning_rate": 3.2351007939370646e-05, "loss": 0.009, "step": 63410 }, { "grad_norm": 0.14351962506771088, "learning_rate": 3.2335538529702184e-05, "loss": 0.0087, "step": 63420 }, { "grad_norm": 0.2427147477865219, "learning_rate": 3.232007105179268e-05, "loss": 0.0107, "step": 63430 }, { "grad_norm": 0.14274844527244568, "learning_rate": 3.230460550733363e-05, "loss": 0.0074, "step": 63440 }, { "grad_norm": 0.1699095368385315, "learning_rate": 3.2289141898016315e-05, "loss": 0.0088, "step": 63450 }, { "grad_norm": 0.16114965081214905, "learning_rate": 3.227368022553181e-05, "loss": 0.0103, "step": 63460 }, { "grad_norm": 0.17459644377231598, "learning_rate": 3.225822049157099e-05, "loss": 0.0123, "step": 63470 }, { "grad_norm": 0.12802532315254211, "learning_rate": 3.2242762697824494e-05, "loss": 0.0089, "step": 63480 }, { "grad_norm": 0.10856039077043533, "learning_rate": 3.222730684598279e-05, "loss": 0.0102, "step": 63490 }, { "grad_norm": 0.09290441125631332, "learning_rate": 3.221185293773605e-05, "loss": 0.012, "step": 63500 }, { "grad_norm": 0.15142515301704407, "learning_rate": 3.2196400974774354e-05, "loss": 0.0089, "step": 63510 }, { "grad_norm": 0.14203119277954102, "learning_rate": 3.2180950958787485e-05, "loss": 0.0107, "step": 63520 }, { "grad_norm": 0.14773467183113098, "learning_rate": 3.2165502891465024e-05, "loss": 0.0111, "step": 63530 }, { "grad_norm": 0.1965627670288086, "learning_rate": 3.215005677449636e-05, "loss": 0.0105, "step": 63540 }, { "grad_norm": 0.15034857392311096, "learning_rate": 3.2134612609570644e-05, "loss": 0.0085, "step": 63550 }, { "grad_norm": 0.14574213325977325, "learning_rate": 3.211917039837682e-05, "loss": 0.0093, "step": 63560 }, { "grad_norm": 0.13796579837799072, "learning_rate": 3.210373014260365e-05, "loss": 0.0091, "step": 63570 }, { "grad_norm": 0.21303905546665192, "learning_rate": 3.208829184393965e-05, "loss": 0.0102, "step": 63580 }, { "grad_norm": 0.12354428321123123, "learning_rate": 3.207285550407312e-05, "loss": 0.0086, "step": 63590 }, { "grad_norm": 0.12757550179958344, "learning_rate": 3.205742112469217e-05, "loss": 0.0111, "step": 63600 }, { "grad_norm": 0.19158963859081268, "learning_rate": 3.204198870748465e-05, "loss": 0.0111, "step": 63610 }, { "grad_norm": 0.09814387559890747, "learning_rate": 3.202655825413824e-05, "loss": 0.0093, "step": 63620 }, { "grad_norm": 0.150206059217453, "learning_rate": 3.2011129766340416e-05, "loss": 0.008, "step": 63630 }, { "grad_norm": 0.155390664935112, "learning_rate": 3.1995703245778384e-05, "loss": 0.0071, "step": 63640 }, { "grad_norm": 0.14245980978012085, "learning_rate": 3.1980278694139186e-05, "loss": 0.0086, "step": 63650 }, { "grad_norm": 0.11221477389335632, "learning_rate": 3.196485611310962e-05, "loss": 0.0075, "step": 63660 }, { "grad_norm": 0.09772608429193497, "learning_rate": 3.1949435504376257e-05, "loss": 0.0081, "step": 63670 }, { "grad_norm": 0.11458399891853333, "learning_rate": 3.1934016869625505e-05, "loss": 0.0078, "step": 63680 }, { "grad_norm": 0.12564408779144287, "learning_rate": 3.191860021054349e-05, "loss": 0.0102, "step": 63690 }, { "grad_norm": 0.16598325967788696, "learning_rate": 3.190318552881618e-05, "loss": 0.0099, "step": 63700 }, { "grad_norm": 0.16790135204792023, "learning_rate": 3.1887772826129285e-05, "loss": 0.0072, "step": 63710 }, { "grad_norm": 0.12455978989601135, "learning_rate": 3.1872362104168335e-05, "loss": 0.0088, "step": 63720 }, { "grad_norm": 0.15393385291099548, "learning_rate": 3.185695336461861e-05, "loss": 0.0107, "step": 63730 }, { "grad_norm": 0.13272933661937714, "learning_rate": 3.184154660916519e-05, "loss": 0.0095, "step": 63740 }, { "grad_norm": 0.12198866903781891, "learning_rate": 3.1826141839492906e-05, "loss": 0.0099, "step": 63750 }, { "grad_norm": 0.12935100495815277, "learning_rate": 3.181073905728644e-05, "loss": 0.0078, "step": 63760 }, { "grad_norm": 0.23441308736801147, "learning_rate": 3.1795338264230214e-05, "loss": 0.0081, "step": 63770 }, { "grad_norm": 0.09537330269813538, "learning_rate": 3.1779939462008426e-05, "loss": 0.0073, "step": 63780 }, { "grad_norm": 0.11154188215732574, "learning_rate": 3.1764542652305066e-05, "loss": 0.0098, "step": 63790 }, { "grad_norm": 0.1658981442451477, "learning_rate": 3.17491478368039e-05, "loss": 0.0094, "step": 63800 }, { "grad_norm": 0.15160506963729858, "learning_rate": 3.173375501718847e-05, "loss": 0.0099, "step": 63810 }, { "grad_norm": 0.12513409554958344, "learning_rate": 3.171836419514215e-05, "loss": 0.0091, "step": 63820 }, { "grad_norm": 0.15296325087547302, "learning_rate": 3.170297537234803e-05, "loss": 0.0093, "step": 63830 }, { "grad_norm": 0.16611847281455994, "learning_rate": 3.168758855048901e-05, "loss": 0.0086, "step": 63840 }, { "grad_norm": 0.13054150342941284, "learning_rate": 3.1672203731247785e-05, "loss": 0.0095, "step": 63850 }, { "grad_norm": 0.1824042648077011, "learning_rate": 3.16568209163068e-05, "loss": 0.009, "step": 63860 }, { "grad_norm": 0.16356833279132843, "learning_rate": 3.1641440107348295e-05, "loss": 0.0126, "step": 63870 }, { "grad_norm": 0.18556258082389832, "learning_rate": 3.162606130605429e-05, "loss": 0.0127, "step": 63880 }, { "grad_norm": 0.1652880609035492, "learning_rate": 3.16106845141066e-05, "loss": 0.0103, "step": 63890 }, { "grad_norm": 0.1859140545129776, "learning_rate": 3.159530973318682e-05, "loss": 0.0105, "step": 63900 }, { "grad_norm": 0.15133190155029297, "learning_rate": 3.1579936964976286e-05, "loss": 0.0144, "step": 63910 }, { "grad_norm": 0.11278683692216873, "learning_rate": 3.156456621115615e-05, "loss": 0.0091, "step": 63920 }, { "grad_norm": 0.1322164386510849, "learning_rate": 3.154919747340734e-05, "loss": 0.0081, "step": 63930 }, { "grad_norm": 0.13158291578292847, "learning_rate": 3.153383075341054e-05, "loss": 0.0077, "step": 63940 }, { "grad_norm": 0.14319010078907013, "learning_rate": 3.1518466052846264e-05, "loss": 0.0108, "step": 63950 }, { "grad_norm": 0.14741209149360657, "learning_rate": 3.150310337339474e-05, "loss": 0.01, "step": 63960 }, { "grad_norm": 0.13601887226104736, "learning_rate": 3.148774271673603e-05, "loss": 0.0084, "step": 63970 }, { "grad_norm": 0.09323431551456451, "learning_rate": 3.147238408454995e-05, "loss": 0.0079, "step": 63980 }, { "grad_norm": 0.08261194825172424, "learning_rate": 3.1457027478516086e-05, "loss": 0.0073, "step": 63990 }, { "grad_norm": 0.13929212093353271, "learning_rate": 3.144167290031382e-05, "loss": 0.0106, "step": 64000 }, { "grad_norm": 0.16172078251838684, "learning_rate": 3.142632035162227e-05, "loss": 0.0108, "step": 64010 }, { "grad_norm": 0.16467484831809998, "learning_rate": 3.141096983412044e-05, "loss": 0.0122, "step": 64020 }, { "grad_norm": 0.17588983476161957, "learning_rate": 3.139562134948699e-05, "loss": 0.0104, "step": 64030 }, { "grad_norm": 0.1671774536371231, "learning_rate": 3.1380274899400424e-05, "loss": 0.0109, "step": 64040 }, { "grad_norm": 0.19473421573638916, "learning_rate": 3.1364930485539e-05, "loss": 0.0098, "step": 64050 }, { "grad_norm": 0.203559011220932, "learning_rate": 3.134958810958076e-05, "loss": 0.0104, "step": 64060 }, { "grad_norm": 0.16483734548091888, "learning_rate": 3.13342477732035e-05, "loss": 0.0083, "step": 64070 }, { "grad_norm": 0.1881406456232071, "learning_rate": 3.131890947808487e-05, "loss": 0.0111, "step": 64080 }, { "grad_norm": 0.13521680235862732, "learning_rate": 3.13035732259022e-05, "loss": 0.0097, "step": 64090 }, { "grad_norm": 0.25101661682128906, "learning_rate": 3.1288239018332656e-05, "loss": 0.0078, "step": 64100 }, { "grad_norm": 0.13175858557224274, "learning_rate": 3.1272906857053164e-05, "loss": 0.0081, "step": 64110 }, { "grad_norm": 0.1414191573858261, "learning_rate": 3.125757674374042e-05, "loss": 0.0105, "step": 64120 }, { "grad_norm": 0.10331543534994125, "learning_rate": 3.124224868007088e-05, "loss": 0.008, "step": 64130 }, { "grad_norm": 0.0985066220164299, "learning_rate": 3.1226922667720854e-05, "loss": 0.0082, "step": 64140 }, { "grad_norm": 0.13289880752563477, "learning_rate": 3.121159870836633e-05, "loss": 0.0122, "step": 64150 }, { "grad_norm": 0.18903875350952148, "learning_rate": 3.119627680368312e-05, "loss": 0.0101, "step": 64160 }, { "grad_norm": 0.1303219497203827, "learning_rate": 3.1180956955346816e-05, "loss": 0.0109, "step": 64170 }, { "grad_norm": 0.09052563458681107, "learning_rate": 3.116563916503276e-05, "loss": 0.0074, "step": 64180 }, { "grad_norm": 0.15682528913021088, "learning_rate": 3.1150323434416095e-05, "loss": 0.0077, "step": 64190 }, { "grad_norm": 0.1491178572177887, "learning_rate": 3.1135009765171696e-05, "loss": 0.0079, "step": 64200 }, { "grad_norm": 0.16058214008808136, "learning_rate": 3.111969815897429e-05, "loss": 0.0092, "step": 64210 }, { "grad_norm": 0.12144366651773453, "learning_rate": 3.1104388617498295e-05, "loss": 0.0083, "step": 64220 }, { "grad_norm": 0.09632952511310577, "learning_rate": 3.1089081142417966e-05, "loss": 0.0093, "step": 64230 }, { "grad_norm": 0.20273306965827942, "learning_rate": 3.1073775735407285e-05, "loss": 0.0094, "step": 64240 }, { "grad_norm": 0.16355477273464203, "learning_rate": 3.1058472398140035e-05, "loss": 0.0092, "step": 64250 }, { "grad_norm": 0.1868838220834732, "learning_rate": 3.1043171132289736e-05, "loss": 0.0083, "step": 64260 }, { "grad_norm": 0.14895778894424438, "learning_rate": 3.1027871939529774e-05, "loss": 0.0117, "step": 64270 }, { "grad_norm": 0.14518819749355316, "learning_rate": 3.1012574821533214e-05, "loss": 0.0074, "step": 64280 }, { "grad_norm": 0.15796148777008057, "learning_rate": 3.099727977997291e-05, "loss": 0.0114, "step": 64290 }, { "grad_norm": 0.09609119594097137, "learning_rate": 3.098198681652154e-05, "loss": 0.0063, "step": 64300 }, { "grad_norm": 0.1332763135433197, "learning_rate": 3.0966695932851475e-05, "loss": 0.0071, "step": 64310 }, { "grad_norm": 0.11845073103904724, "learning_rate": 3.0951407130634926e-05, "loss": 0.008, "step": 64320 }, { "grad_norm": 0.12024563550949097, "learning_rate": 3.0936120411543864e-05, "loss": 0.0084, "step": 64330 }, { "grad_norm": 0.12430166453123093, "learning_rate": 3.0920835777250015e-05, "loss": 0.0089, "step": 64340 }, { "grad_norm": 0.1455768197774887, "learning_rate": 3.090555322942488e-05, "loss": 0.0111, "step": 64350 }, { "grad_norm": 0.19105006754398346, "learning_rate": 3.089027276973973e-05, "loss": 0.0106, "step": 64360 }, { "grad_norm": 0.1397116780281067, "learning_rate": 3.0874994399865615e-05, "loss": 0.0072, "step": 64370 }, { "grad_norm": 0.09760534018278122, "learning_rate": 3.085971812147337e-05, "loss": 0.0074, "step": 64380 }, { "grad_norm": 0.11156536638736725, "learning_rate": 3.0844443936233557e-05, "loss": 0.0091, "step": 64390 }, { "grad_norm": 0.15504948794841766, "learning_rate": 3.082917184581657e-05, "loss": 0.0077, "step": 64400 }, { "grad_norm": 0.1929614245891571, "learning_rate": 3.0813901851892513e-05, "loss": 0.0146, "step": 64410 }, { "grad_norm": 0.13592971861362457, "learning_rate": 3.0798633956131316e-05, "loss": 0.0108, "step": 64420 }, { "grad_norm": 0.10701218992471695, "learning_rate": 3.078336816020264e-05, "loss": 0.0072, "step": 64430 }, { "grad_norm": 0.12596508860588074, "learning_rate": 3.0768104465775935e-05, "loss": 0.009, "step": 64440 }, { "grad_norm": 0.15808792412281036, "learning_rate": 3.0752842874520396e-05, "loss": 0.0095, "step": 64450 }, { "grad_norm": 0.15126203000545502, "learning_rate": 3.073758338810503e-05, "loss": 0.0101, "step": 64460 }, { "grad_norm": 0.12034881860017776, "learning_rate": 3.072232600819859e-05, "loss": 0.0099, "step": 64470 }, { "grad_norm": 0.12484486401081085, "learning_rate": 3.070707073646959e-05, "loss": 0.0094, "step": 64480 }, { "grad_norm": 0.2076112926006317, "learning_rate": 3.069181757458633e-05, "loss": 0.0082, "step": 64490 }, { "grad_norm": 0.13109175860881805, "learning_rate": 3.067656652421687e-05, "loss": 0.0082, "step": 64500 }, { "grad_norm": 0.2531459331512451, "learning_rate": 3.0661317587029025e-05, "loss": 0.0132, "step": 64510 }, { "grad_norm": 0.1359851360321045, "learning_rate": 3.064607076469043e-05, "loss": 0.009, "step": 64520 }, { "grad_norm": 0.10559068620204926, "learning_rate": 3.063082605886845e-05, "loss": 0.0114, "step": 64530 }, { "grad_norm": 0.17521560192108154, "learning_rate": 3.061558347123019e-05, "loss": 0.0082, "step": 64540 }, { "grad_norm": 0.1259729564189911, "learning_rate": 3.060034300344259e-05, "loss": 0.0107, "step": 64550 }, { "grad_norm": 0.1736188530921936, "learning_rate": 3.058510465717231e-05, "loss": 0.0092, "step": 64560 }, { "grad_norm": 0.172332301735878, "learning_rate": 3.0569868434085804e-05, "loss": 0.009, "step": 64570 }, { "grad_norm": 0.16491299867630005, "learning_rate": 3.055463433584924e-05, "loss": 0.0122, "step": 64580 }, { "grad_norm": 0.12058111280202866, "learning_rate": 3.0539402364128656e-05, "loss": 0.0084, "step": 64590 }, { "grad_norm": 0.109617680311203, "learning_rate": 3.052417252058977e-05, "loss": 0.0108, "step": 64600 }, { "grad_norm": 0.11088574677705765, "learning_rate": 3.0508944806898092e-05, "loss": 0.0071, "step": 64610 }, { "grad_norm": 0.21040043234825134, "learning_rate": 3.04937192247189e-05, "loss": 0.0107, "step": 64620 }, { "grad_norm": 0.09440628439188004, "learning_rate": 3.047849577571725e-05, "loss": 0.0079, "step": 64630 }, { "grad_norm": 0.09350477159023285, "learning_rate": 3.046327446155795e-05, "loss": 0.0072, "step": 64640 }, { "grad_norm": 0.14105968177318573, "learning_rate": 3.0448055283905587e-05, "loss": 0.0089, "step": 64650 }, { "grad_norm": 0.14124487340450287, "learning_rate": 3.0432838244424494e-05, "loss": 0.009, "step": 64660 }, { "grad_norm": 0.15986524522304535, "learning_rate": 3.04176233447788e-05, "loss": 0.0093, "step": 64670 }, { "grad_norm": 0.13386350870132446, "learning_rate": 3.040241058663238e-05, "loss": 0.0075, "step": 64680 }, { "grad_norm": 0.15333732962608337, "learning_rate": 3.038719997164887e-05, "loss": 0.01, "step": 64690 }, { "grad_norm": 0.14472399652004242, "learning_rate": 3.0371991501491683e-05, "loss": 0.0083, "step": 64700 }, { "grad_norm": 0.129332035779953, "learning_rate": 3.0356785177823975e-05, "loss": 0.0085, "step": 64710 }, { "grad_norm": 0.09564579278230667, "learning_rate": 3.0341581002308726e-05, "loss": 0.009, "step": 64720 }, { "grad_norm": 0.13893520832061768, "learning_rate": 3.0326378976608627e-05, "loss": 0.0109, "step": 64730 }, { "grad_norm": 0.12357960641384125, "learning_rate": 3.0311179102386146e-05, "loss": 0.0088, "step": 64740 }, { "grad_norm": 0.12446137517690659, "learning_rate": 3.0295981381303513e-05, "loss": 0.0082, "step": 64750 }, { "grad_norm": 0.1542859673500061, "learning_rate": 3.0280785815022727e-05, "loss": 0.0089, "step": 64760 }, { "grad_norm": 0.13880446553230286, "learning_rate": 3.0265592405205544e-05, "loss": 0.0104, "step": 64770 }, { "grad_norm": 0.2057453691959381, "learning_rate": 3.0250401153513517e-05, "loss": 0.0127, "step": 64780 }, { "grad_norm": 0.17746120691299438, "learning_rate": 3.0235212061607936e-05, "loss": 0.0095, "step": 64790 }, { "grad_norm": 0.17930027842521667, "learning_rate": 3.0220025131149842e-05, "loss": 0.0108, "step": 64800 }, { "grad_norm": 0.12045265734195709, "learning_rate": 3.020484036380007e-05, "loss": 0.015, "step": 64810 }, { "grad_norm": 0.15965485572814941, "learning_rate": 3.0189657761219188e-05, "loss": 0.0094, "step": 64820 }, { "grad_norm": 0.1306329369544983, "learning_rate": 3.017447732506753e-05, "loss": 0.0091, "step": 64830 }, { "grad_norm": 0.13381756842136383, "learning_rate": 3.015929905700524e-05, "loss": 0.0104, "step": 64840 }, { "grad_norm": 0.10385075956583023, "learning_rate": 3.014412295869218e-05, "loss": 0.0078, "step": 64850 }, { "grad_norm": 0.18329280614852905, "learning_rate": 3.0128949031787977e-05, "loss": 0.01, "step": 64860 }, { "grad_norm": 0.1670812964439392, "learning_rate": 3.0113777277952022e-05, "loss": 0.0084, "step": 64870 }, { "grad_norm": 0.11820483952760696, "learning_rate": 3.0098607698843496e-05, "loss": 0.0092, "step": 64880 }, { "grad_norm": 0.14817938208580017, "learning_rate": 3.0083440296121312e-05, "loss": 0.0132, "step": 64890 }, { "grad_norm": 0.1325322836637497, "learning_rate": 3.0068275071444142e-05, "loss": 0.008, "step": 64900 }, { "grad_norm": 0.19247469305992126, "learning_rate": 3.0053112026470442e-05, "loss": 0.0099, "step": 64910 }, { "grad_norm": 0.13636033236980438, "learning_rate": 3.0037951162858434e-05, "loss": 0.01, "step": 64920 }, { "grad_norm": 0.11772819608449936, "learning_rate": 3.002279248226607e-05, "loss": 0.0103, "step": 64930 }, { "grad_norm": 0.15390434861183167, "learning_rate": 3.000763598635109e-05, "loss": 0.0104, "step": 64940 }, { "grad_norm": 0.2841349244117737, "learning_rate": 2.999248167677098e-05, "loss": 0.011, "step": 64950 }, { "grad_norm": 0.17824018001556396, "learning_rate": 2.9977329555182976e-05, "loss": 0.0098, "step": 64960 }, { "grad_norm": 0.16827787458896637, "learning_rate": 2.9962179623244124e-05, "loss": 0.0089, "step": 64970 }, { "grad_norm": 0.14112010598182678, "learning_rate": 2.9947031882611175e-05, "loss": 0.0085, "step": 64980 }, { "grad_norm": 0.15783028304576874, "learning_rate": 2.9931886334940685e-05, "loss": 0.0096, "step": 64990 }, { "grad_norm": 0.13776427507400513, "learning_rate": 2.9916742981888923e-05, "loss": 0.007, "step": 65000 }, { "grad_norm": 0.16756336390972137, "learning_rate": 2.990160182511196e-05, "loss": 0.009, "step": 65010 }, { "grad_norm": 0.16484138369560242, "learning_rate": 2.9886462866265574e-05, "loss": 0.0082, "step": 65020 }, { "grad_norm": 0.20867682993412018, "learning_rate": 2.9871326107005392e-05, "loss": 0.008, "step": 65030 }, { "grad_norm": 0.21549344062805176, "learning_rate": 2.9856191548986735e-05, "loss": 0.0117, "step": 65040 }, { "grad_norm": 0.14829181134700775, "learning_rate": 2.9841059193864672e-05, "loss": 0.0095, "step": 65050 }, { "grad_norm": 0.12569810450077057, "learning_rate": 2.982592904329407e-05, "loss": 0.0089, "step": 65060 }, { "grad_norm": 0.1546565443277359, "learning_rate": 2.981080109892954e-05, "loss": 0.0149, "step": 65070 }, { "grad_norm": 0.16544944047927856, "learning_rate": 2.9795675362425434e-05, "loss": 0.0103, "step": 65080 }, { "grad_norm": 0.15953461825847626, "learning_rate": 2.97805518354359e-05, "loss": 0.0098, "step": 65090 }, { "grad_norm": 0.16004091501235962, "learning_rate": 2.9765430519614825e-05, "loss": 0.008, "step": 65100 }, { "grad_norm": 0.16061444580554962, "learning_rate": 2.975031141661584e-05, "loss": 0.0083, "step": 65110 }, { "grad_norm": 0.11892528086900711, "learning_rate": 2.9735194528092347e-05, "loss": 0.0083, "step": 65120 }, { "grad_norm": 0.16442781686782837, "learning_rate": 2.9720079855697525e-05, "loss": 0.0075, "step": 65130 }, { "grad_norm": 0.16107501089572906, "learning_rate": 2.970496740108427e-05, "loss": 0.0101, "step": 65140 }, { "grad_norm": 0.17353279888629913, "learning_rate": 2.968985716590527e-05, "loss": 0.0108, "step": 65150 }, { "grad_norm": 0.16844451427459717, "learning_rate": 2.9674749151812953e-05, "loss": 0.0111, "step": 65160 }, { "grad_norm": 0.1449369341135025, "learning_rate": 2.965964336045951e-05, "loss": 0.0082, "step": 65170 }, { "grad_norm": 0.12760813534259796, "learning_rate": 2.9644539793496894e-05, "loss": 0.0127, "step": 65180 }, { "grad_norm": 0.12732148170471191, "learning_rate": 2.9629438452576807e-05, "loss": 0.0113, "step": 65190 }, { "grad_norm": 0.10494785010814667, "learning_rate": 2.9614339339350698e-05, "loss": 0.0118, "step": 65200 }, { "grad_norm": 0.1484517604112625, "learning_rate": 2.959924245546979e-05, "loss": 0.0076, "step": 65210 }, { "grad_norm": 0.11544414609670639, "learning_rate": 2.9584147802585038e-05, "loss": 0.0066, "step": 65220 }, { "grad_norm": 0.165392205119133, "learning_rate": 2.956905538234721e-05, "loss": 0.0077, "step": 65230 }, { "grad_norm": 0.1631244570016861, "learning_rate": 2.9553965196406764e-05, "loss": 0.0065, "step": 65240 }, { "grad_norm": 0.1943434774875641, "learning_rate": 2.9538877246413943e-05, "loss": 0.0094, "step": 65250 }, { "grad_norm": 0.13457722961902618, "learning_rate": 2.952379153401874e-05, "loss": 0.0115, "step": 65260 }, { "grad_norm": 0.15494191646575928, "learning_rate": 2.9508708060870914e-05, "loss": 0.0078, "step": 65270 }, { "grad_norm": 0.12516213953495026, "learning_rate": 2.9493626828619948e-05, "loss": 0.0078, "step": 65280 }, { "grad_norm": 0.10284898430109024, "learning_rate": 2.9478547838915137e-05, "loss": 0.0061, "step": 65290 }, { "grad_norm": 0.14895789325237274, "learning_rate": 2.946347109340548e-05, "loss": 0.0063, "step": 65300 }, { "grad_norm": 0.09794692695140839, "learning_rate": 2.944839659373976e-05, "loss": 0.0085, "step": 65310 }, { "grad_norm": 0.13824696838855743, "learning_rate": 2.943332434156647e-05, "loss": 0.0101, "step": 65320 }, { "grad_norm": 0.18012918531894684, "learning_rate": 2.9418254338533912e-05, "loss": 0.0094, "step": 65330 }, { "grad_norm": 0.18109238147735596, "learning_rate": 2.940318658629011e-05, "loss": 0.0087, "step": 65340 }, { "grad_norm": 0.15880940854549408, "learning_rate": 2.938812108648287e-05, "loss": 0.0117, "step": 65350 }, { "grad_norm": 0.3554462492465973, "learning_rate": 2.937305784075972e-05, "loss": 0.0113, "step": 65360 }, { "grad_norm": 0.13147376477718353, "learning_rate": 2.9357996850767937e-05, "loss": 0.0099, "step": 65370 }, { "grad_norm": 0.15769679844379425, "learning_rate": 2.9342938118154595e-05, "loss": 0.007, "step": 65380 }, { "grad_norm": 0.1465953290462494, "learning_rate": 2.9327881644566485e-05, "loss": 0.0077, "step": 65390 }, { "grad_norm": 0.12099028378725052, "learning_rate": 2.9312827431650157e-05, "loss": 0.011, "step": 65400 }, { "grad_norm": 0.10284357517957687, "learning_rate": 2.92977754810519e-05, "loss": 0.0074, "step": 65410 }, { "grad_norm": 0.180915966629982, "learning_rate": 2.928272579441781e-05, "loss": 0.0072, "step": 65420 }, { "grad_norm": 0.14432533085346222, "learning_rate": 2.9267678373393676e-05, "loss": 0.0086, "step": 65430 }, { "grad_norm": 0.17918190360069275, "learning_rate": 2.925263321962507e-05, "loss": 0.0086, "step": 65440 }, { "grad_norm": 0.11554815620183945, "learning_rate": 2.9237590334757292e-05, "loss": 0.0086, "step": 65450 }, { "grad_norm": 0.15732260048389435, "learning_rate": 2.922254972043543e-05, "loss": 0.0112, "step": 65460 }, { "grad_norm": 0.08588408678770065, "learning_rate": 2.9207511378304265e-05, "loss": 0.0066, "step": 65470 }, { "grad_norm": 0.17786520719528198, "learning_rate": 2.9192475310008428e-05, "loss": 0.0075, "step": 65480 }, { "grad_norm": 0.1260496824979782, "learning_rate": 2.917744151719219e-05, "loss": 0.0088, "step": 65490 }, { "grad_norm": 0.11798039078712463, "learning_rate": 2.916241000149964e-05, "loss": 0.0084, "step": 65500 }, { "grad_norm": 0.1637057662010193, "learning_rate": 2.9147380764574622e-05, "loss": 0.0082, "step": 65510 }, { "grad_norm": 0.1182316392660141, "learning_rate": 2.9132353808060674e-05, "loss": 0.009, "step": 65520 }, { "grad_norm": 0.12636876106262207, "learning_rate": 2.9117329133601134e-05, "loss": 0.0108, "step": 65530 }, { "grad_norm": 0.14054031670093536, "learning_rate": 2.9102306742839093e-05, "loss": 0.0102, "step": 65540 }, { "grad_norm": 0.1729210466146469, "learning_rate": 2.9087286637417375e-05, "loss": 0.011, "step": 65550 }, { "grad_norm": 0.1441061645746231, "learning_rate": 2.9072268818978537e-05, "loss": 0.0076, "step": 65560 }, { "grad_norm": 0.13900025188922882, "learning_rate": 2.905725328916493e-05, "loss": 0.0063, "step": 65570 }, { "grad_norm": 0.14384527504444122, "learning_rate": 2.9042240049618608e-05, "loss": 0.0091, "step": 65580 }, { "grad_norm": 0.20404095947742462, "learning_rate": 2.9027229101981413e-05, "loss": 0.0099, "step": 65590 }, { "grad_norm": 0.11090390384197235, "learning_rate": 2.901222044789488e-05, "loss": 0.0075, "step": 65600 }, { "grad_norm": 0.17392227053642273, "learning_rate": 2.89972140890004e-05, "loss": 0.0069, "step": 65610 }, { "grad_norm": 0.16219310462474823, "learning_rate": 2.8982210026938988e-05, "loss": 0.0093, "step": 65620 }, { "grad_norm": 0.13795757293701172, "learning_rate": 2.896720826335151e-05, "loss": 0.0089, "step": 65630 }, { "grad_norm": 0.11152727901935577, "learning_rate": 2.8952208799878493e-05, "loss": 0.0069, "step": 65640 }, { "grad_norm": 0.13904070854187012, "learning_rate": 2.89372116381603e-05, "loss": 0.0122, "step": 65650 }, { "grad_norm": 0.15784218907356262, "learning_rate": 2.8922216779836932e-05, "loss": 0.0107, "step": 65660 }, { "grad_norm": 0.10826791077852249, "learning_rate": 2.8907224226548285e-05, "loss": 0.0079, "step": 65670 }, { "grad_norm": 0.10114115476608276, "learning_rate": 2.8892233979933868e-05, "loss": 0.0067, "step": 65680 }, { "grad_norm": 0.1434120237827301, "learning_rate": 2.8877246041633026e-05, "loss": 0.0077, "step": 65690 }, { "grad_norm": 0.10800506174564362, "learning_rate": 2.886226041328478e-05, "loss": 0.0088, "step": 65700 }, { "grad_norm": 0.12699873745441437, "learning_rate": 2.8847277096527957e-05, "loss": 0.0094, "step": 65710 }, { "grad_norm": 0.16088731586933136, "learning_rate": 2.8832296093001105e-05, "loss": 0.0062, "step": 65720 }, { "grad_norm": 0.13473132252693176, "learning_rate": 2.8817317404342547e-05, "loss": 0.0092, "step": 65730 }, { "grad_norm": 0.10922374576330185, "learning_rate": 2.880234103219029e-05, "loss": 0.0078, "step": 65740 }, { "grad_norm": 0.12626689672470093, "learning_rate": 2.878736697818215e-05, "loss": 0.0091, "step": 65750 }, { "grad_norm": 0.08090092986822128, "learning_rate": 2.877239524395568e-05, "loss": 0.0089, "step": 65760 }, { "grad_norm": 0.11393196135759354, "learning_rate": 2.8757425831148134e-05, "loss": 0.0081, "step": 65770 }, { "grad_norm": 0.4529117941856384, "learning_rate": 2.874245874139658e-05, "loss": 0.0078, "step": 65780 }, { "grad_norm": 0.12660522758960724, "learning_rate": 2.8727493976337737e-05, "loss": 0.0086, "step": 65790 }, { "grad_norm": 0.14656686782836914, "learning_rate": 2.8712531537608218e-05, "loss": 0.0089, "step": 65800 }, { "grad_norm": 0.10683676600456238, "learning_rate": 2.8697571426844217e-05, "loss": 0.0061, "step": 65810 }, { "grad_norm": 0.21280743181705475, "learning_rate": 2.86826136456818e-05, "loss": 0.0068, "step": 65820 }, { "grad_norm": 0.14691978693008423, "learning_rate": 2.8667658195756685e-05, "loss": 0.0089, "step": 65830 }, { "grad_norm": 0.11490832269191742, "learning_rate": 2.8652705078704412e-05, "loss": 0.0089, "step": 65840 }, { "grad_norm": 0.13834701478481293, "learning_rate": 2.863775429616018e-05, "loss": 0.0077, "step": 65850 }, { "grad_norm": 0.13936972618103027, "learning_rate": 2.862280584975906e-05, "loss": 0.0078, "step": 65860 }, { "grad_norm": 0.18639147281646729, "learning_rate": 2.8607859741135735e-05, "loss": 0.0092, "step": 65870 }, { "grad_norm": 0.2008131742477417, "learning_rate": 2.8592915971924723e-05, "loss": 0.012, "step": 65880 }, { "grad_norm": 0.14614495635032654, "learning_rate": 2.8577974543760212e-05, "loss": 0.0077, "step": 65890 }, { "grad_norm": 0.12533259391784668, "learning_rate": 2.8563035458276222e-05, "loss": 0.0095, "step": 65900 }, { "grad_norm": 0.09790971875190735, "learning_rate": 2.8548098717106424e-05, "loss": 0.0081, "step": 65910 }, { "grad_norm": 0.12252523005008698, "learning_rate": 2.85331643218843e-05, "loss": 0.0065, "step": 65920 }, { "grad_norm": 0.14461836218833923, "learning_rate": 2.8518232274243046e-05, "loss": 0.0064, "step": 65930 }, { "grad_norm": 0.15486203134059906, "learning_rate": 2.8503302575815638e-05, "loss": 0.0072, "step": 65940 }, { "grad_norm": 0.19845417141914368, "learning_rate": 2.8488375228234716e-05, "loss": 0.01, "step": 65950 }, { "grad_norm": 0.25405722856521606, "learning_rate": 2.8473450233132738e-05, "loss": 0.0076, "step": 65960 }, { "grad_norm": 0.2466239035129547, "learning_rate": 2.8458527592141898e-05, "loss": 0.0102, "step": 65970 }, { "grad_norm": 0.15539872646331787, "learning_rate": 2.8443607306894073e-05, "loss": 0.0084, "step": 65980 }, { "grad_norm": 0.12950414419174194, "learning_rate": 2.8428689379020946e-05, "loss": 0.0109, "step": 65990 }, { "grad_norm": 0.18104349076747894, "learning_rate": 2.841377381015391e-05, "loss": 0.009, "step": 66000 }, { "grad_norm": 0.12837782502174377, "learning_rate": 2.8398860601924144e-05, "loss": 0.0063, "step": 66010 }, { "grad_norm": 0.1481981873512268, "learning_rate": 2.838394975596248e-05, "loss": 0.0099, "step": 66020 }, { "grad_norm": 0.1945754438638687, "learning_rate": 2.83690412738996e-05, "loss": 0.009, "step": 66030 }, { "grad_norm": 0.1591600626707077, "learning_rate": 2.835413515736579e-05, "loss": 0.009, "step": 66040 }, { "grad_norm": 0.1773996651172638, "learning_rate": 2.8339231407991267e-05, "loss": 0.0152, "step": 66050 }, { "grad_norm": 0.1469523012638092, "learning_rate": 2.832433002740581e-05, "loss": 0.0089, "step": 66060 }, { "grad_norm": 0.14464867115020752, "learning_rate": 2.8309431017239047e-05, "loss": 0.0095, "step": 66070 }, { "grad_norm": 0.11252051591873169, "learning_rate": 2.829453437912028e-05, "loss": 0.0105, "step": 66080 }, { "grad_norm": 0.1588548719882965, "learning_rate": 2.8279640114678617e-05, "loss": 0.0097, "step": 66090 }, { "grad_norm": 0.10792285948991776, "learning_rate": 2.8264748225542836e-05, "loss": 0.0088, "step": 66100 }, { "grad_norm": 0.12862248718738556, "learning_rate": 2.8249858713341516e-05, "loss": 0.0112, "step": 66110 }, { "grad_norm": 0.1315593719482422, "learning_rate": 2.823497157970294e-05, "loss": 0.0092, "step": 66120 }, { "grad_norm": 0.15309825539588928, "learning_rate": 2.8220086826255164e-05, "loss": 0.0108, "step": 66130 }, { "grad_norm": 0.15885688364505768, "learning_rate": 2.8205204454625934e-05, "loss": 0.0069, "step": 66140 }, { "grad_norm": 0.15249566733837128, "learning_rate": 2.8190324466442798e-05, "loss": 0.0079, "step": 66150 }, { "grad_norm": 0.09177353233098984, "learning_rate": 2.8175446863332965e-05, "loss": 0.0128, "step": 66160 }, { "grad_norm": 0.11944455653429031, "learning_rate": 2.816057164692345e-05, "loss": 0.0094, "step": 66170 }, { "grad_norm": 0.1457401067018509, "learning_rate": 2.8145698818840992e-05, "loss": 0.0113, "step": 66180 }, { "grad_norm": 0.12667016685009003, "learning_rate": 2.8130828380712065e-05, "loss": 0.0099, "step": 66190 }, { "grad_norm": 0.14564338326454163, "learning_rate": 2.811596033416285e-05, "loss": 0.0085, "step": 66200 }, { "grad_norm": 0.1359904408454895, "learning_rate": 2.8101094680819307e-05, "loss": 0.0099, "step": 66210 }, { "grad_norm": 0.14041322469711304, "learning_rate": 2.808623142230715e-05, "loss": 0.0093, "step": 66220 }, { "grad_norm": 0.09296280145645142, "learning_rate": 2.8071370560251757e-05, "loss": 0.0074, "step": 66230 }, { "grad_norm": 0.14810267090797424, "learning_rate": 2.805651209627831e-05, "loss": 0.0079, "step": 66240 }, { "grad_norm": 0.15698908269405365, "learning_rate": 2.804165603201171e-05, "loss": 0.0103, "step": 66250 }, { "grad_norm": 0.12425313144922256, "learning_rate": 2.802680236907661e-05, "loss": 0.006, "step": 66260 }, { "grad_norm": 0.11920619755983353, "learning_rate": 2.8011951109097344e-05, "loss": 0.0088, "step": 66270 }, { "grad_norm": 0.11639617383480072, "learning_rate": 2.7997102253698072e-05, "loss": 0.0134, "step": 66280 }, { "grad_norm": 0.1375483274459839, "learning_rate": 2.7982255804502588e-05, "loss": 0.0082, "step": 66290 }, { "grad_norm": 0.15481488406658173, "learning_rate": 2.7967411763134516e-05, "loss": 0.0097, "step": 66300 }, { "grad_norm": 0.14841587841510773, "learning_rate": 2.7952570131217166e-05, "loss": 0.0076, "step": 66310 }, { "grad_norm": 0.18710120022296906, "learning_rate": 2.7937730910373618e-05, "loss": 0.0104, "step": 66320 }, { "grad_norm": 0.13332027196884155, "learning_rate": 2.7922894102226628e-05, "loss": 0.0092, "step": 66330 }, { "grad_norm": 0.11296196281909943, "learning_rate": 2.790805970839876e-05, "loss": 0.0104, "step": 66340 }, { "grad_norm": 0.15567100048065186, "learning_rate": 2.789322773051225e-05, "loss": 0.0084, "step": 66350 }, { "grad_norm": 0.11036193370819092, "learning_rate": 2.7878398170189124e-05, "loss": 0.01, "step": 66360 }, { "grad_norm": 0.11416412889957428, "learning_rate": 2.7863571029051112e-05, "loss": 0.0086, "step": 66370 }, { "grad_norm": 0.1792123168706894, "learning_rate": 2.784874630871971e-05, "loss": 0.0084, "step": 66380 }, { "grad_norm": 0.18897642195224762, "learning_rate": 2.7833924010816086e-05, "loss": 0.007, "step": 66390 }, { "grad_norm": 0.15118779242038727, "learning_rate": 2.7819104136961226e-05, "loss": 0.0099, "step": 66400 }, { "grad_norm": 0.12157925963401794, "learning_rate": 2.780428668877577e-05, "loss": 0.0118, "step": 66410 }, { "grad_norm": 0.16677623987197876, "learning_rate": 2.7789471667880152e-05, "loss": 0.0098, "step": 66420 }, { "grad_norm": 0.1931968778371811, "learning_rate": 2.7774659075894523e-05, "loss": 0.0075, "step": 66430 }, { "grad_norm": 0.1298021823167801, "learning_rate": 2.7759848914438778e-05, "loss": 0.0082, "step": 66440 }, { "grad_norm": 0.1855933666229248, "learning_rate": 2.7745041185132496e-05, "loss": 0.0087, "step": 66450 }, { "grad_norm": 0.17526797950267792, "learning_rate": 2.773023588959506e-05, "loss": 0.0097, "step": 66460 }, { "grad_norm": 0.14876416325569153, "learning_rate": 2.7715433029445555e-05, "loss": 0.0096, "step": 66470 }, { "grad_norm": 0.12131627649068832, "learning_rate": 2.7700632606302766e-05, "loss": 0.0085, "step": 66480 }, { "grad_norm": 0.20395272970199585, "learning_rate": 2.768583462178528e-05, "loss": 0.0081, "step": 66490 }, { "grad_norm": 0.09928187727928162, "learning_rate": 2.767103907751136e-05, "loss": 0.0109, "step": 66500 }, { "grad_norm": 0.11224886775016785, "learning_rate": 2.7656245975099065e-05, "loss": 0.0136, "step": 66510 }, { "grad_norm": 0.14372466504573822, "learning_rate": 2.7641455316166093e-05, "loss": 0.0086, "step": 66520 }, { "grad_norm": 0.10990386456251144, "learning_rate": 2.7626667102329973e-05, "loss": 0.006, "step": 66530 }, { "grad_norm": 0.13145999610424042, "learning_rate": 2.761188133520788e-05, "loss": 0.0099, "step": 66540 }, { "grad_norm": 0.14082805812358856, "learning_rate": 2.7597098016416784e-05, "loss": 0.0087, "step": 66550 }, { "grad_norm": 0.10654477775096893, "learning_rate": 2.7582317147573368e-05, "loss": 0.0114, "step": 66560 }, { "grad_norm": 0.16619785130023956, "learning_rate": 2.7567538730294064e-05, "loss": 0.0101, "step": 66570 }, { "grad_norm": 0.14461158215999603, "learning_rate": 2.7552762766194973e-05, "loss": 0.0093, "step": 66580 }, { "grad_norm": 0.1224270910024643, "learning_rate": 2.753798925689202e-05, "loss": 0.0093, "step": 66590 }, { "grad_norm": 0.1703302562236786, "learning_rate": 2.7523218204000774e-05, "loss": 0.0111, "step": 66600 }, { "grad_norm": 0.18664860725402832, "learning_rate": 2.7508449609136612e-05, "loss": 0.0109, "step": 66610 }, { "grad_norm": 0.10823577642440796, "learning_rate": 2.749368347391454e-05, "loss": 0.0088, "step": 66620 }, { "grad_norm": 0.17013856768608093, "learning_rate": 2.7478919799949443e-05, "loss": 0.009, "step": 66630 }, { "grad_norm": 0.15396620333194733, "learning_rate": 2.7464158588855803e-05, "loss": 0.0075, "step": 66640 }, { "grad_norm": 0.10129011422395706, "learning_rate": 2.7449399842247918e-05, "loss": 0.0135, "step": 66650 }, { "grad_norm": 0.14547099173069, "learning_rate": 2.7434643561739736e-05, "loss": 0.0102, "step": 66660 }, { "grad_norm": 0.08005566149950027, "learning_rate": 2.7419889748945016e-05, "loss": 0.0085, "step": 66670 }, { "grad_norm": 0.12709462642669678, "learning_rate": 2.7405138405477203e-05, "loss": 0.0075, "step": 66680 }, { "grad_norm": 0.1655554324388504, "learning_rate": 2.73903895329495e-05, "loss": 0.0079, "step": 66690 }, { "grad_norm": 0.1529843956232071, "learning_rate": 2.737564313297478e-05, "loss": 0.0104, "step": 66700 }, { "grad_norm": 0.17678038775920868, "learning_rate": 2.736089920716572e-05, "loss": 0.0086, "step": 66710 }, { "grad_norm": 0.18419669568538666, "learning_rate": 2.7346157757134706e-05, "loss": 0.0085, "step": 66720 }, { "grad_norm": 0.10489682853221893, "learning_rate": 2.733141878449379e-05, "loss": 0.0088, "step": 66730 }, { "grad_norm": 0.14694127440452576, "learning_rate": 2.7316682290854844e-05, "loss": 0.0082, "step": 66740 }, { "grad_norm": 0.11234636604785919, "learning_rate": 2.7301948277829414e-05, "loss": 0.0077, "step": 66750 }, { "grad_norm": 0.11964932829141617, "learning_rate": 2.7287216747028814e-05, "loss": 0.0106, "step": 66760 }, { "grad_norm": 0.09678163379430771, "learning_rate": 2.7272487700064025e-05, "loss": 0.0106, "step": 66770 }, { "grad_norm": 0.11578943580389023, "learning_rate": 2.725776113854583e-05, "loss": 0.0076, "step": 66780 }, { "grad_norm": 0.2617177963256836, "learning_rate": 2.7243037064084664e-05, "loss": 0.009, "step": 66790 }, { "grad_norm": 0.15173538029193878, "learning_rate": 2.7228315478290766e-05, "loss": 0.0089, "step": 66800 }, { "grad_norm": 0.11665785312652588, "learning_rate": 2.721359638277402e-05, "loss": 0.0099, "step": 66810 }, { "grad_norm": 0.3206346929073334, "learning_rate": 2.7198879779144148e-05, "loss": 0.0088, "step": 66820 }, { "grad_norm": 0.09792827069759369, "learning_rate": 2.7184165669010485e-05, "loss": 0.0089, "step": 66830 }, { "grad_norm": 0.18667031824588776, "learning_rate": 2.716945405398218e-05, "loss": 0.0078, "step": 66840 }, { "grad_norm": 0.12960240244865417, "learning_rate": 2.715474493566803e-05, "loss": 0.008, "step": 66850 }, { "grad_norm": 0.2721397876739502, "learning_rate": 2.7140038315676652e-05, "loss": 0.0093, "step": 66860 }, { "grad_norm": 0.13594914972782135, "learning_rate": 2.7125334195616275e-05, "loss": 0.012, "step": 66870 }, { "grad_norm": 0.14278683066368103, "learning_rate": 2.7110632577094997e-05, "loss": 0.0112, "step": 66880 }, { "grad_norm": 0.146107017993927, "learning_rate": 2.709593346172051e-05, "loss": 0.0093, "step": 66890 }, { "grad_norm": 0.08238916844129562, "learning_rate": 2.708123685110032e-05, "loss": 0.0088, "step": 66900 }, { "grad_norm": 0.1471003144979477, "learning_rate": 2.7066542746841588e-05, "loss": 0.0082, "step": 66910 }, { "grad_norm": 0.1149807721376419, "learning_rate": 2.7051851150551267e-05, "loss": 0.0067, "step": 66920 }, { "grad_norm": 0.1780194342136383, "learning_rate": 2.7037162063835996e-05, "loss": 0.0094, "step": 66930 }, { "grad_norm": 0.13345162570476532, "learning_rate": 2.702247548830218e-05, "loss": 0.0089, "step": 66940 }, { "grad_norm": 0.15093758702278137, "learning_rate": 2.7007791425555883e-05, "loss": 0.0145, "step": 66950 }, { "grad_norm": 0.14304448664188385, "learning_rate": 2.6993109877202945e-05, "loss": 0.0086, "step": 66960 }, { "grad_norm": 0.10886260122060776, "learning_rate": 2.6978430844848935e-05, "loss": 0.0103, "step": 66970 }, { "grad_norm": 0.09831036627292633, "learning_rate": 2.69637543300991e-05, "loss": 0.0067, "step": 66980 }, { "grad_norm": 0.1344635784626007, "learning_rate": 2.694908033455848e-05, "loss": 0.0076, "step": 66990 }, { "grad_norm": 0.13141894340515137, "learning_rate": 2.6934408859831733e-05, "loss": 0.0079, "step": 67000 }, { "grad_norm": 0.15675756335258484, "learning_rate": 2.69197399075234e-05, "loss": 0.0104, "step": 67010 }, { "grad_norm": 0.14326724410057068, "learning_rate": 2.6905073479237584e-05, "loss": 0.0124, "step": 67020 }, { "grad_norm": 0.11309656500816345, "learning_rate": 2.6890409576578235e-05, "loss": 0.0081, "step": 67030 }, { "grad_norm": 0.1405564844608307, "learning_rate": 2.687574820114892e-05, "loss": 0.0093, "step": 67040 }, { "grad_norm": 0.19104135036468506, "learning_rate": 2.686108935455305e-05, "loss": 0.0113, "step": 67050 }, { "grad_norm": 0.17237813770771027, "learning_rate": 2.6846433038393603e-05, "loss": 0.011, "step": 67060 }, { "grad_norm": 0.15598993003368378, "learning_rate": 2.6831779254273482e-05, "loss": 0.0081, "step": 67070 }, { "grad_norm": 0.11977032572031021, "learning_rate": 2.6817128003795122e-05, "loss": 0.0071, "step": 67080 }, { "grad_norm": 0.13611994683742523, "learning_rate": 2.680247928856081e-05, "loss": 0.0087, "step": 67090 }, { "grad_norm": 0.13126949965953827, "learning_rate": 2.6787833110172466e-05, "loss": 0.0086, "step": 67100 }, { "grad_norm": 0.12868717312812805, "learning_rate": 2.6773189470231807e-05, "loss": 0.0087, "step": 67110 }, { "grad_norm": 0.11397744715213776, "learning_rate": 2.675854837034022e-05, "loss": 0.0094, "step": 67120 }, { "grad_norm": 0.11511073261499405, "learning_rate": 2.6743909812098833e-05, "loss": 0.0059, "step": 67130 }, { "grad_norm": 0.15160533785820007, "learning_rate": 2.67292737971085e-05, "loss": 0.0099, "step": 67140 }, { "grad_norm": 0.13686908781528473, "learning_rate": 2.671464032696982e-05, "loss": 0.0102, "step": 67150 }, { "grad_norm": 0.15416152775287628, "learning_rate": 2.6700009403283055e-05, "loss": 0.0096, "step": 67160 }, { "grad_norm": 0.16539470851421356, "learning_rate": 2.6685381027648214e-05, "loss": 0.0088, "step": 67170 }, { "grad_norm": 0.1334080994129181, "learning_rate": 2.6670755201665086e-05, "loss": 0.0081, "step": 67180 }, { "grad_norm": 0.1521933674812317, "learning_rate": 2.6656131926933066e-05, "loss": 0.0082, "step": 67190 }, { "grad_norm": 0.10724464058876038, "learning_rate": 2.664151120505136e-05, "loss": 0.007, "step": 67200 }, { "grad_norm": 0.10901376605033875, "learning_rate": 2.6626893037618873e-05, "loss": 0.0069, "step": 67210 }, { "grad_norm": 0.19457851350307465, "learning_rate": 2.6612277426234243e-05, "loss": 0.0122, "step": 67220 }, { "grad_norm": 0.11795607209205627, "learning_rate": 2.659766437249577e-05, "loss": 0.0089, "step": 67230 }, { "grad_norm": 0.12145402282476425, "learning_rate": 2.6583053878001557e-05, "loss": 0.008, "step": 67240 }, { "grad_norm": 0.13698384165763855, "learning_rate": 2.6568445944349323e-05, "loss": 0.0102, "step": 67250 }, { "grad_norm": 0.15183894336223602, "learning_rate": 2.6553840573136657e-05, "loss": 0.0079, "step": 67260 }, { "grad_norm": 0.13327142596244812, "learning_rate": 2.6539237765960706e-05, "loss": 0.0087, "step": 67270 }, { "grad_norm": 0.13045428693294525, "learning_rate": 2.6524637524418465e-05, "loss": 0.0087, "step": 67280 }, { "grad_norm": 0.09022659808397293, "learning_rate": 2.651003985010655e-05, "loss": 0.0102, "step": 67290 }, { "grad_norm": 0.16442304849624634, "learning_rate": 2.649544474462138e-05, "loss": 0.0097, "step": 67300 }, { "grad_norm": 0.09922542423009872, "learning_rate": 2.6480852209559015e-05, "loss": 0.007, "step": 67310 }, { "grad_norm": 0.12521643936634064, "learning_rate": 2.6466262246515296e-05, "loss": 0.0116, "step": 67320 }, { "grad_norm": 0.13344615697860718, "learning_rate": 2.6451674857085746e-05, "loss": 0.0069, "step": 67330 }, { "grad_norm": 0.17756545543670654, "learning_rate": 2.6437090042865655e-05, "loss": 0.0097, "step": 67340 }, { "grad_norm": 0.1434076726436615, "learning_rate": 2.6422507805449947e-05, "loss": 0.0082, "step": 67350 }, { "grad_norm": 0.10832897573709488, "learning_rate": 2.6407928146433357e-05, "loss": 0.0067, "step": 67360 }, { "grad_norm": 0.14970695972442627, "learning_rate": 2.639335106741026e-05, "loss": 0.0144, "step": 67370 }, { "grad_norm": 0.13484226167201996, "learning_rate": 2.6378776569974796e-05, "loss": 0.0074, "step": 67380 }, { "grad_norm": 0.1381731927394867, "learning_rate": 2.6364204655720814e-05, "loss": 0.0087, "step": 67390 }, { "grad_norm": 0.13998937606811523, "learning_rate": 2.6349635326241894e-05, "loss": 0.0062, "step": 67400 }, { "grad_norm": 0.1684977263212204, "learning_rate": 2.633506858313127e-05, "loss": 0.0086, "step": 67410 }, { "grad_norm": 0.15100154280662537, "learning_rate": 2.6320504427981975e-05, "loss": 0.0097, "step": 67420 }, { "grad_norm": 0.14544092118740082, "learning_rate": 2.6305942862386734e-05, "loss": 0.0053, "step": 67430 }, { "grad_norm": 0.21443629264831543, "learning_rate": 2.629138388793793e-05, "loss": 0.0122, "step": 67440 }, { "grad_norm": 0.10208764672279358, "learning_rate": 2.6276827506227743e-05, "loss": 0.0081, "step": 67450 }, { "grad_norm": 0.16652259230613708, "learning_rate": 2.6262273718848037e-05, "loss": 0.0077, "step": 67460 }, { "grad_norm": 0.1257946640253067, "learning_rate": 2.6247722527390407e-05, "loss": 0.0066, "step": 67470 }, { "grad_norm": 0.12587031722068787, "learning_rate": 2.6233173933446104e-05, "loss": 0.0087, "step": 67480 }, { "grad_norm": 0.1998613476753235, "learning_rate": 2.6218627938606188e-05, "loss": 0.0096, "step": 67490 }, { "grad_norm": 0.09608402103185654, "learning_rate": 2.620408454446135e-05, "loss": 0.0113, "step": 67500 }, { "grad_norm": 0.08955010771751404, "learning_rate": 2.6189543752602046e-05, "loss": 0.0082, "step": 67510 }, { "grad_norm": 0.12341862916946411, "learning_rate": 2.6175005564618444e-05, "loss": 0.0061, "step": 67520 }, { "grad_norm": 0.17277070879936218, "learning_rate": 2.6160469982100428e-05, "loss": 0.0089, "step": 67530 }, { "grad_norm": 0.11216823756694794, "learning_rate": 2.6145937006637555e-05, "loss": 0.0085, "step": 67540 }, { "grad_norm": 0.1588517129421234, "learning_rate": 2.6131406639819166e-05, "loss": 0.009, "step": 67550 }, { "grad_norm": 0.18088477849960327, "learning_rate": 2.611687888323424e-05, "loss": 0.0089, "step": 67560 }, { "grad_norm": 0.13966399431228638, "learning_rate": 2.610235373847153e-05, "loss": 0.0083, "step": 67570 }, { "grad_norm": 0.13435260951519012, "learning_rate": 2.6087831207119494e-05, "loss": 0.009, "step": 67580 }, { "grad_norm": 0.20642369985580444, "learning_rate": 2.60733112907663e-05, "loss": 0.0116, "step": 67590 }, { "grad_norm": 0.11493714153766632, "learning_rate": 2.6058793990999786e-05, "loss": 0.0071, "step": 67600 }, { "grad_norm": 0.1325707882642746, "learning_rate": 2.6044279309407593e-05, "loss": 0.0106, "step": 67610 }, { "grad_norm": 0.16624735295772552, "learning_rate": 2.6029767247576976e-05, "loss": 0.0086, "step": 67620 }, { "grad_norm": 0.11456625163555145, "learning_rate": 2.601525780709497e-05, "loss": 0.0064, "step": 67630 }, { "grad_norm": 0.09853058308362961, "learning_rate": 2.6000750989548318e-05, "loss": 0.0063, "step": 67640 }, { "grad_norm": 0.17076963186264038, "learning_rate": 2.5986246796523462e-05, "loss": 0.009, "step": 67650 }, { "grad_norm": 0.11414080858230591, "learning_rate": 2.5971745229606546e-05, "loss": 0.01, "step": 67660 }, { "grad_norm": 0.12976685166358948, "learning_rate": 2.5957246290383442e-05, "loss": 0.0074, "step": 67670 }, { "grad_norm": 0.10106737911701202, "learning_rate": 2.594274998043975e-05, "loss": 0.0069, "step": 67680 }, { "grad_norm": 0.14702238142490387, "learning_rate": 2.5928256301360733e-05, "loss": 0.0084, "step": 67690 }, { "grad_norm": 0.18451745808124542, "learning_rate": 2.591376525473142e-05, "loss": 0.0093, "step": 67700 }, { "grad_norm": 0.15605442225933075, "learning_rate": 2.5899276842136523e-05, "loss": 0.0068, "step": 67710 }, { "grad_norm": 0.10055530071258545, "learning_rate": 2.5884791065160495e-05, "loss": 0.0073, "step": 67720 }, { "grad_norm": 0.09197118878364563, "learning_rate": 2.5870307925387436e-05, "loss": 0.009, "step": 67730 }, { "grad_norm": 0.1563023030757904, "learning_rate": 2.585582742440125e-05, "loss": 0.0068, "step": 67740 }, { "grad_norm": 0.12825079262256622, "learning_rate": 2.5841349563785455e-05, "loss": 0.0109, "step": 67750 }, { "grad_norm": 0.16693052649497986, "learning_rate": 2.5826874345123355e-05, "loss": 0.008, "step": 67760 }, { "grad_norm": 0.15868040919303894, "learning_rate": 2.581240176999792e-05, "loss": 0.0095, "step": 67770 }, { "grad_norm": 0.16214461624622345, "learning_rate": 2.579793183999189e-05, "loss": 0.0094, "step": 67780 }, { "grad_norm": 0.19224482774734497, "learning_rate": 2.5783464556687627e-05, "loss": 0.0103, "step": 67790 }, { "grad_norm": 0.12025659531354904, "learning_rate": 2.576899992166729e-05, "loss": 0.0113, "step": 67800 }, { "grad_norm": 0.2738710343837738, "learning_rate": 2.5754537936512668e-05, "loss": 0.0074, "step": 67810 }, { "grad_norm": 0.11167591065168381, "learning_rate": 2.574007860280535e-05, "loss": 0.0079, "step": 67820 }, { "grad_norm": 0.12482950836420059, "learning_rate": 2.5725621922126518e-05, "loss": 0.0077, "step": 67830 }, { "grad_norm": 0.08111289143562317, "learning_rate": 2.571116789605722e-05, "loss": 0.0081, "step": 67840 }, { "grad_norm": 0.14673756062984467, "learning_rate": 2.5696716526178054e-05, "loss": 0.0124, "step": 67850 }, { "grad_norm": 0.11006853729486465, "learning_rate": 2.568226781406946e-05, "loss": 0.0087, "step": 67860 }, { "grad_norm": 0.15048857033252716, "learning_rate": 2.5667821761311474e-05, "loss": 0.0097, "step": 67870 }, { "grad_norm": 0.10119593143463135, "learning_rate": 2.5653378369483917e-05, "loss": 0.0071, "step": 67880 }, { "grad_norm": 0.10946203768253326, "learning_rate": 2.5638937640166292e-05, "loss": 0.0071, "step": 67890 }, { "grad_norm": 0.13339059054851532, "learning_rate": 2.5624499574937842e-05, "loss": 0.0075, "step": 67900 }, { "grad_norm": 0.13792425394058228, "learning_rate": 2.5610064175377456e-05, "loss": 0.0109, "step": 67910 }, { "grad_norm": 0.13061007857322693, "learning_rate": 2.5595631443063782e-05, "loss": 0.0084, "step": 67920 }, { "grad_norm": 0.19544421136379242, "learning_rate": 2.5581201379575177e-05, "loss": 0.0089, "step": 67930 }, { "grad_norm": 0.19104540348052979, "learning_rate": 2.5566773986489655e-05, "loss": 0.0089, "step": 67940 }, { "grad_norm": 0.17846454679965973, "learning_rate": 2.5552349265385e-05, "loss": 0.0099, "step": 67950 }, { "grad_norm": 0.12166790664196014, "learning_rate": 2.5537927217838675e-05, "loss": 0.0094, "step": 67960 }, { "grad_norm": 0.14168764650821686, "learning_rate": 2.5523507845427864e-05, "loss": 0.0091, "step": 67970 }, { "grad_norm": 0.17704756557941437, "learning_rate": 2.5509091149729413e-05, "loss": 0.009, "step": 67980 }, { "grad_norm": 0.1279565542936325, "learning_rate": 2.5494677132319962e-05, "loss": 0.0087, "step": 67990 }, { "grad_norm": 0.08530221879482269, "learning_rate": 2.548026579477575e-05, "loss": 0.0098, "step": 68000 }, { "grad_norm": 0.13330326974391937, "learning_rate": 2.5465857138672834e-05, "loss": 0.0061, "step": 68010 }, { "grad_norm": 0.10868439823389053, "learning_rate": 2.5451451165586848e-05, "loss": 0.0093, "step": 68020 }, { "grad_norm": 0.12662146985530853, "learning_rate": 2.543704787709329e-05, "loss": 0.0073, "step": 68030 }, { "grad_norm": 0.22660717368125916, "learning_rate": 2.542264727476723e-05, "loss": 0.0103, "step": 68040 }, { "grad_norm": 0.134850412607193, "learning_rate": 2.5408249360183535e-05, "loss": 0.008, "step": 68050 }, { "grad_norm": 0.13945576548576355, "learning_rate": 2.5393854134916694e-05, "loss": 0.0111, "step": 68060 }, { "grad_norm": 0.12357434630393982, "learning_rate": 2.5379461600540993e-05, "loss": 0.0094, "step": 68070 }, { "grad_norm": 0.08662696182727814, "learning_rate": 2.5365071758630317e-05, "loss": 0.0077, "step": 68080 }, { "grad_norm": 0.1266249567270279, "learning_rate": 2.535068461075839e-05, "loss": 0.0094, "step": 68090 }, { "grad_norm": 0.15370258688926697, "learning_rate": 2.5336300158498516e-05, "loss": 0.0074, "step": 68100 }, { "grad_norm": 0.1839764416217804, "learning_rate": 2.5321918403423793e-05, "loss": 0.0122, "step": 68110 }, { "grad_norm": 0.15837706625461578, "learning_rate": 2.5307539347106957e-05, "loss": 0.0082, "step": 68120 }, { "grad_norm": 0.10191306471824646, "learning_rate": 2.529316299112049e-05, "loss": 0.0072, "step": 68130 }, { "grad_norm": 0.12983271479606628, "learning_rate": 2.5278789337036567e-05, "loss": 0.0104, "step": 68140 }, { "grad_norm": 0.20587778091430664, "learning_rate": 2.52644183864271e-05, "loss": 0.0083, "step": 68150 }, { "grad_norm": 0.11585800349712372, "learning_rate": 2.525005014086363e-05, "loss": 0.0085, "step": 68160 }, { "grad_norm": 0.12354518473148346, "learning_rate": 2.5235684601917465e-05, "loss": 0.0084, "step": 68170 }, { "grad_norm": 0.13271985948085785, "learning_rate": 2.5221321771159618e-05, "loss": 0.0106, "step": 68180 }, { "grad_norm": 0.0948663204908371, "learning_rate": 2.5206961650160743e-05, "loss": 0.007, "step": 68190 }, { "grad_norm": 0.09622843563556671, "learning_rate": 2.519260424049129e-05, "loss": 0.0068, "step": 68200 }, { "grad_norm": 0.09531483054161072, "learning_rate": 2.5178249543721304e-05, "loss": 0.0076, "step": 68210 }, { "grad_norm": 0.1349189281463623, "learning_rate": 2.5163897561420668e-05, "loss": 0.0124, "step": 68220 }, { "grad_norm": 0.21122026443481445, "learning_rate": 2.514954829515883e-05, "loss": 0.009, "step": 68230 }, { "grad_norm": 0.17402073740959167, "learning_rate": 2.513520174650505e-05, "loss": 0.0096, "step": 68240 }, { "grad_norm": 0.1350487917661667, "learning_rate": 2.5120857917028207e-05, "loss": 0.0099, "step": 68250 }, { "grad_norm": 0.16511313617229462, "learning_rate": 2.5106516808296953e-05, "loss": 0.0084, "step": 68260 }, { "grad_norm": 0.13889886438846588, "learning_rate": 2.5092178421879552e-05, "loss": 0.0073, "step": 68270 }, { "grad_norm": 0.17134171724319458, "learning_rate": 2.5077842759344116e-05, "loss": 0.0071, "step": 68280 }, { "grad_norm": 0.15267260372638702, "learning_rate": 2.5063509822258304e-05, "loss": 0.0089, "step": 68290 }, { "grad_norm": 0.135330468416214, "learning_rate": 2.5049179612189576e-05, "loss": 0.0092, "step": 68300 }, { "grad_norm": 0.1518053263425827, "learning_rate": 2.5034852130705043e-05, "loss": 0.0088, "step": 68310 }, { "grad_norm": 0.14388175308704376, "learning_rate": 2.5020527379371572e-05, "loss": 0.0066, "step": 68320 }, { "grad_norm": 0.1358547806739807, "learning_rate": 2.500620535975564e-05, "loss": 0.0082, "step": 68330 }, { "grad_norm": 0.15635275840759277, "learning_rate": 2.4991886073423515e-05, "loss": 0.0082, "step": 68340 }, { "grad_norm": 0.13071836531162262, "learning_rate": 2.4977569521941134e-05, "loss": 0.0082, "step": 68350 }, { "grad_norm": 0.1520875096321106, "learning_rate": 2.4963255706874143e-05, "loss": 0.0078, "step": 68360 }, { "grad_norm": 0.12947902083396912, "learning_rate": 2.4948944629787853e-05, "loss": 0.0074, "step": 68370 }, { "grad_norm": 0.1060781478881836, "learning_rate": 2.493463629224731e-05, "loss": 0.0082, "step": 68380 }, { "grad_norm": 0.1536587029695511, "learning_rate": 2.492033069581728e-05, "loss": 0.0066, "step": 68390 }, { "grad_norm": 0.1917090117931366, "learning_rate": 2.4906027842062137e-05, "loss": 0.0078, "step": 68400 }, { "grad_norm": 0.11564439535140991, "learning_rate": 2.48917277325461e-05, "loss": 0.0062, "step": 68410 }, { "grad_norm": 0.09829097241163254, "learning_rate": 2.487743036883295e-05, "loss": 0.0066, "step": 68420 }, { "grad_norm": 0.18074184656143188, "learning_rate": 2.486313575248626e-05, "loss": 0.0085, "step": 68430 }, { "grad_norm": 0.168065145611763, "learning_rate": 2.4848843885069235e-05, "loss": 0.0066, "step": 68440 }, { "grad_norm": 0.14121054112911224, "learning_rate": 2.483455476814484e-05, "loss": 0.0077, "step": 68450 }, { "grad_norm": 0.13449420034885406, "learning_rate": 2.4820268403275672e-05, "loss": 0.0092, "step": 68460 }, { "grad_norm": 0.1428985893726349, "learning_rate": 2.480598479202413e-05, "loss": 0.0087, "step": 68470 }, { "grad_norm": 0.19775155186653137, "learning_rate": 2.4791703935952193e-05, "loss": 0.009, "step": 68480 }, { "grad_norm": 0.14905378222465515, "learning_rate": 2.477742583662163e-05, "loss": 0.0083, "step": 68490 }, { "grad_norm": 0.09248929470777512, "learning_rate": 2.476315049559384e-05, "loss": 0.0092, "step": 68500 }, { "grad_norm": 0.1403735876083374, "learning_rate": 2.4748877914429987e-05, "loss": 0.009, "step": 68510 }, { "grad_norm": 0.1740359514951706, "learning_rate": 2.4734608094690857e-05, "loss": 0.0091, "step": 68520 }, { "grad_norm": 0.14580807089805603, "learning_rate": 2.4720341037937006e-05, "loss": 0.0078, "step": 68530 }, { "grad_norm": 0.11300137639045715, "learning_rate": 2.4706076745728645e-05, "loss": 0.0084, "step": 68540 }, { "grad_norm": 0.18190209567546844, "learning_rate": 2.469181521962573e-05, "loss": 0.007, "step": 68550 }, { "grad_norm": 0.14599047601222992, "learning_rate": 2.467755646118783e-05, "loss": 0.0104, "step": 68560 }, { "grad_norm": 0.14992044866085052, "learning_rate": 2.46633004719743e-05, "loss": 0.0084, "step": 68570 }, { "grad_norm": 0.21747714281082153, "learning_rate": 2.4649047253544123e-05, "loss": 0.0084, "step": 68580 }, { "grad_norm": 0.10332705825567245, "learning_rate": 2.4634796807456016e-05, "loss": 0.0066, "step": 68590 }, { "grad_norm": 0.0992494449019432, "learning_rate": 2.46205491352684e-05, "loss": 0.0067, "step": 68600 }, { "grad_norm": 0.13165508210659027, "learning_rate": 2.4606304238539375e-05, "loss": 0.0064, "step": 68610 }, { "grad_norm": 0.1416667103767395, "learning_rate": 2.4592062118826753e-05, "loss": 0.0077, "step": 68620 }, { "grad_norm": 0.18980775773525238, "learning_rate": 2.4577822777688004e-05, "loss": 0.0099, "step": 68630 }, { "grad_norm": 0.20872218906879425, "learning_rate": 2.4563586216680347e-05, "loss": 0.0091, "step": 68640 }, { "grad_norm": 0.10482708364725113, "learning_rate": 2.4549352437360622e-05, "loss": 0.0078, "step": 68650 }, { "grad_norm": 0.13131548464298248, "learning_rate": 2.4535121441285493e-05, "loss": 0.0075, "step": 68660 }, { "grad_norm": 0.09770316630601883, "learning_rate": 2.4520893230011172e-05, "loss": 0.0069, "step": 68670 }, { "grad_norm": 0.07716378569602966, "learning_rate": 2.450666780509368e-05, "loss": 0.0081, "step": 68680 }, { "grad_norm": 0.15542981028556824, "learning_rate": 2.4492445168088656e-05, "loss": 0.0084, "step": 68690 }, { "grad_norm": 0.11822858452796936, "learning_rate": 2.447822532055149e-05, "loss": 0.0074, "step": 68700 }, { "grad_norm": 0.11210735142230988, "learning_rate": 2.4464008264037212e-05, "loss": 0.0079, "step": 68710 }, { "grad_norm": 0.16616804897785187, "learning_rate": 2.4449794000100605e-05, "loss": 0.009, "step": 68720 }, { "grad_norm": 0.15348614752292633, "learning_rate": 2.4435582530296108e-05, "loss": 0.0086, "step": 68730 }, { "grad_norm": 0.1614333987236023, "learning_rate": 2.4421373856177887e-05, "loss": 0.0076, "step": 68740 }, { "grad_norm": 0.10249649733304977, "learning_rate": 2.440716797929975e-05, "loss": 0.0083, "step": 68750 }, { "grad_norm": 0.12242802232503891, "learning_rate": 2.439296490121526e-05, "loss": 0.0086, "step": 68760 }, { "grad_norm": 0.14711476862430573, "learning_rate": 2.4378764623477608e-05, "loss": 0.0141, "step": 68770 }, { "grad_norm": 0.1559850126504898, "learning_rate": 2.436456714763974e-05, "loss": 0.0097, "step": 68780 }, { "grad_norm": 0.21596835553646088, "learning_rate": 2.4350372475254262e-05, "loss": 0.0089, "step": 68790 }, { "grad_norm": 0.14083647727966309, "learning_rate": 2.433618060787351e-05, "loss": 0.0084, "step": 68800 }, { "grad_norm": 0.07069985568523407, "learning_rate": 2.432199154704944e-05, "loss": 0.0089, "step": 68810 }, { "grad_norm": 0.10350879281759262, "learning_rate": 2.430780529433377e-05, "loss": 0.0057, "step": 68820 }, { "grad_norm": 0.10678748786449432, "learning_rate": 2.42936218512779e-05, "loss": 0.0072, "step": 68830 }, { "grad_norm": 0.11038468033075333, "learning_rate": 2.427944121943288e-05, "loss": 0.0071, "step": 68840 }, { "grad_norm": 0.15189696848392487, "learning_rate": 2.4265263400349497e-05, "loss": 0.0078, "step": 68850 }, { "grad_norm": 0.14343959093093872, "learning_rate": 2.4251088395578213e-05, "loss": 0.0086, "step": 68860 }, { "grad_norm": 0.3402525782585144, "learning_rate": 2.4236916206669214e-05, "loss": 0.0089, "step": 68870 }, { "grad_norm": 0.11562664806842804, "learning_rate": 2.4222746835172304e-05, "loss": 0.0062, "step": 68880 }, { "grad_norm": 0.11322557926177979, "learning_rate": 2.4208580282637065e-05, "loss": 0.0099, "step": 68890 }, { "grad_norm": 0.15784168243408203, "learning_rate": 2.4194416550612687e-05, "loss": 0.0089, "step": 68900 }, { "grad_norm": 0.24663101136684418, "learning_rate": 2.418025564064812e-05, "loss": 0.0093, "step": 68910 }, { "grad_norm": 0.3383356034755707, "learning_rate": 2.4166097554291978e-05, "loss": 0.0092, "step": 68920 }, { "grad_norm": 0.149814635515213, "learning_rate": 2.4151942293092583e-05, "loss": 0.0086, "step": 68930 }, { "grad_norm": 0.1396271288394928, "learning_rate": 2.41377898585979e-05, "loss": 0.0082, "step": 68940 }, { "grad_norm": 0.1427379995584488, "learning_rate": 2.4123640252355652e-05, "loss": 0.0102, "step": 68950 }, { "grad_norm": 0.12039556354284286, "learning_rate": 2.410949347591318e-05, "loss": 0.0092, "step": 68960 }, { "grad_norm": 0.11694806069135666, "learning_rate": 2.4095349530817578e-05, "loss": 0.0067, "step": 68970 }, { "grad_norm": 0.10406401008367538, "learning_rate": 2.4081208418615603e-05, "loss": 0.0086, "step": 68980 }, { "grad_norm": 0.11206002533435822, "learning_rate": 2.4067070140853725e-05, "loss": 0.0105, "step": 68990 }, { "grad_norm": 0.15288826823234558, "learning_rate": 2.405293469907805e-05, "loss": 0.0077, "step": 69000 }, { "grad_norm": 0.1561686247587204, "learning_rate": 2.4038802094834446e-05, "loss": 0.0096, "step": 69010 }, { "grad_norm": 0.09365659207105637, "learning_rate": 2.4024672329668395e-05, "loss": 0.0096, "step": 69020 }, { "grad_norm": 0.11475707590579987, "learning_rate": 2.4010545405125125e-05, "loss": 0.0108, "step": 69030 }, { "grad_norm": 0.09044265002012253, "learning_rate": 2.3996421322749528e-05, "loss": 0.007, "step": 69040 }, { "grad_norm": 0.15742354094982147, "learning_rate": 2.3982300084086224e-05, "loss": 0.0079, "step": 69050 }, { "grad_norm": 0.09929978102445602, "learning_rate": 2.396818169067944e-05, "loss": 0.0093, "step": 69060 }, { "grad_norm": 0.11417603492736816, "learning_rate": 2.3954066144073178e-05, "loss": 0.008, "step": 69070 }, { "grad_norm": 0.10781678557395935, "learning_rate": 2.39399534458111e-05, "loss": 0.0085, "step": 69080 }, { "grad_norm": 0.15182405710220337, "learning_rate": 2.392584359743651e-05, "loss": 0.0078, "step": 69090 }, { "grad_norm": 0.10288049280643463, "learning_rate": 2.3911736600492463e-05, "loss": 0.0072, "step": 69100 }, { "grad_norm": 0.10343436896800995, "learning_rate": 2.3897632456521672e-05, "loss": 0.0076, "step": 69110 }, { "grad_norm": 0.1658540517091751, "learning_rate": 2.388353116706658e-05, "loss": 0.0088, "step": 69120 }, { "grad_norm": 0.1558593362569809, "learning_rate": 2.3869432733669232e-05, "loss": 0.0101, "step": 69130 }, { "grad_norm": 0.15674827992916107, "learning_rate": 2.385533715787145e-05, "loss": 0.0077, "step": 69140 }, { "grad_norm": 0.0994291752576828, "learning_rate": 2.3841244441214673e-05, "loss": 0.007, "step": 69150 }, { "grad_norm": 0.09830906987190247, "learning_rate": 2.3827154585240073e-05, "loss": 0.0116, "step": 69160 }, { "grad_norm": 0.14731380343437195, "learning_rate": 2.3813067591488498e-05, "loss": 0.0082, "step": 69170 }, { "grad_norm": 0.13796702027320862, "learning_rate": 2.3798983461500494e-05, "loss": 0.0083, "step": 69180 }, { "grad_norm": 0.18829600512981415, "learning_rate": 2.3784902196816256e-05, "loss": 0.0078, "step": 69190 }, { "grad_norm": 0.11220989376306534, "learning_rate": 2.3770823798975716e-05, "loss": 0.0076, "step": 69200 }, { "grad_norm": 0.13077500462532043, "learning_rate": 2.3756748269518424e-05, "loss": 0.0109, "step": 69210 }, { "grad_norm": 0.09433246403932571, "learning_rate": 2.374267560998371e-05, "loss": 0.0054, "step": 69220 }, { "grad_norm": 0.13729433715343475, "learning_rate": 2.3728605821910477e-05, "loss": 0.0078, "step": 69230 }, { "grad_norm": 0.14255240559577942, "learning_rate": 2.3714538906837453e-05, "loss": 0.0097, "step": 69240 }, { "grad_norm": 0.14580784738063812, "learning_rate": 2.3700474866302913e-05, "loss": 0.0077, "step": 69250 }, { "grad_norm": 0.10724546015262604, "learning_rate": 2.3686413701844917e-05, "loss": 0.0093, "step": 69260 }, { "grad_norm": 0.14857690036296844, "learning_rate": 2.367235541500114e-05, "loss": 0.0086, "step": 69270 }, { "grad_norm": 0.11441953480243683, "learning_rate": 2.3658300007308993e-05, "loss": 0.0078, "step": 69280 }, { "grad_norm": 0.11256945878267288, "learning_rate": 2.364424748030555e-05, "loss": 0.0104, "step": 69290 }, { "grad_norm": 0.156716987490654, "learning_rate": 2.3630197835527595e-05, "loss": 0.0094, "step": 69300 }, { "grad_norm": 0.119113989174366, "learning_rate": 2.3616151074511546e-05, "loss": 0.0073, "step": 69310 }, { "grad_norm": 0.12796734273433685, "learning_rate": 2.3602107198793545e-05, "loss": 0.0074, "step": 69320 }, { "grad_norm": 0.13818515837192535, "learning_rate": 2.3588066209909427e-05, "loss": 0.0068, "step": 69330 }, { "grad_norm": 0.1513165533542633, "learning_rate": 2.3574028109394662e-05, "loss": 0.0086, "step": 69340 }, { "grad_norm": 0.13164320588111877, "learning_rate": 2.3559992898784454e-05, "loss": 0.0072, "step": 69350 }, { "grad_norm": 0.16046705842018127, "learning_rate": 2.354596057961367e-05, "loss": 0.0106, "step": 69360 }, { "grad_norm": 0.16015905141830444, "learning_rate": 2.353193115341688e-05, "loss": 0.0082, "step": 69370 }, { "grad_norm": 0.1349056214094162, "learning_rate": 2.3517904621728294e-05, "loss": 0.0072, "step": 69380 }, { "grad_norm": 0.13081268966197968, "learning_rate": 2.3503880986081855e-05, "loss": 0.0086, "step": 69390 }, { "grad_norm": 0.13561488687992096, "learning_rate": 2.3489860248011136e-05, "loss": 0.0089, "step": 69400 }, { "grad_norm": 0.13814106583595276, "learning_rate": 2.3475842409049464e-05, "loss": 0.0058, "step": 69410 }, { "grad_norm": 0.10636857897043228, "learning_rate": 2.346182747072975e-05, "loss": 0.0099, "step": 69420 }, { "grad_norm": 0.11450584977865219, "learning_rate": 2.3447815434584718e-05, "loss": 0.007, "step": 69430 }, { "grad_norm": 0.14652420580387115, "learning_rate": 2.3433806302146656e-05, "loss": 0.0076, "step": 69440 }, { "grad_norm": 0.15112164616584778, "learning_rate": 2.3419800074947612e-05, "loss": 0.0075, "step": 69450 }, { "grad_norm": 0.17300397157669067, "learning_rate": 2.3405796754519248e-05, "loss": 0.008, "step": 69460 }, { "grad_norm": 0.14077065885066986, "learning_rate": 2.3391796342392992e-05, "loss": 0.0081, "step": 69470 }, { "grad_norm": 0.1143968254327774, "learning_rate": 2.337779884009984e-05, "loss": 0.0079, "step": 69480 }, { "grad_norm": 0.13618122041225433, "learning_rate": 2.3363804249170624e-05, "loss": 0.01, "step": 69490 }, { "grad_norm": 0.1093304306268692, "learning_rate": 2.3349812571135703e-05, "loss": 0.008, "step": 69500 }, { "grad_norm": 0.10415983200073242, "learning_rate": 2.333582380752523e-05, "loss": 0.0067, "step": 69510 }, { "grad_norm": 0.10408470034599304, "learning_rate": 2.3321837959868963e-05, "loss": 0.0079, "step": 69520 }, { "grad_norm": 0.12180820852518082, "learning_rate": 2.3307855029696383e-05, "loss": 0.0067, "step": 69530 }, { "grad_norm": 0.1401568055152893, "learning_rate": 2.3293875018536648e-05, "loss": 0.0078, "step": 69540 }, { "grad_norm": 0.1249924823641777, "learning_rate": 2.327989792791861e-05, "loss": 0.0075, "step": 69550 }, { "grad_norm": 0.14630475640296936, "learning_rate": 2.326592375937074e-05, "loss": 0.0065, "step": 69560 }, { "grad_norm": 0.2203906774520874, "learning_rate": 2.3251952514421248e-05, "loss": 0.0079, "step": 69570 }, { "grad_norm": 0.13818903267383575, "learning_rate": 2.323798419459804e-05, "loss": 0.0075, "step": 69580 }, { "grad_norm": 0.22729872167110443, "learning_rate": 2.3224018801428625e-05, "loss": 0.0075, "step": 69590 }, { "grad_norm": 0.10809841006994247, "learning_rate": 2.3210056336440273e-05, "loss": 0.0073, "step": 69600 }, { "grad_norm": 0.16164878010749817, "learning_rate": 2.3196096801159844e-05, "loss": 0.0086, "step": 69610 }, { "grad_norm": 0.1868763118982315, "learning_rate": 2.3182140197114012e-05, "loss": 0.0095, "step": 69620 }, { "grad_norm": 0.10677538067102432, "learning_rate": 2.3168186525828985e-05, "loss": 0.0087, "step": 69630 }, { "grad_norm": 0.17952211201190948, "learning_rate": 2.3154235788830757e-05, "loss": 0.008, "step": 69640 }, { "grad_norm": 0.11711857467889786, "learning_rate": 2.3140287987644927e-05, "loss": 0.0103, "step": 69650 }, { "grad_norm": 0.1313892900943756, "learning_rate": 2.3126343123796833e-05, "loss": 0.0114, "step": 69660 }, { "grad_norm": 0.1800781786441803, "learning_rate": 2.3112401198811418e-05, "loss": 0.0092, "step": 69670 }, { "grad_norm": 0.14440114796161652, "learning_rate": 2.3098462214213418e-05, "loss": 0.0079, "step": 69680 }, { "grad_norm": 0.1404709368944168, "learning_rate": 2.308452617152712e-05, "loss": 0.0078, "step": 69690 }, { "grad_norm": 0.11628662794828415, "learning_rate": 2.3070593072276603e-05, "loss": 0.0071, "step": 69700 }, { "grad_norm": 0.13554809987545013, "learning_rate": 2.3056662917985518e-05, "loss": 0.0075, "step": 69710 }, { "grad_norm": 0.11237841844558716, "learning_rate": 2.3042735710177283e-05, "loss": 0.0052, "step": 69720 }, { "grad_norm": 0.16468198597431183, "learning_rate": 2.302881145037492e-05, "loss": 0.0103, "step": 69730 }, { "grad_norm": 0.13618460297584534, "learning_rate": 2.3014890140101192e-05, "loss": 0.0084, "step": 69740 }, { "grad_norm": 0.12148761749267578, "learning_rate": 2.30009717808785e-05, "loss": 0.0099, "step": 69750 }, { "grad_norm": 0.14365284144878387, "learning_rate": 2.2987056374228967e-05, "loss": 0.0066, "step": 69760 }, { "grad_norm": 0.16375620663166046, "learning_rate": 2.297314392167431e-05, "loss": 0.0102, "step": 69770 }, { "grad_norm": 0.15816658735275269, "learning_rate": 2.295923442473601e-05, "loss": 0.0084, "step": 69780 }, { "grad_norm": 0.23368071019649506, "learning_rate": 2.294532788493519e-05, "loss": 0.0066, "step": 69790 }, { "grad_norm": 0.12953297793865204, "learning_rate": 2.293142430379262e-05, "loss": 0.008, "step": 69800 }, { "grad_norm": 0.09506472945213318, "learning_rate": 2.291752368282879e-05, "loss": 0.0063, "step": 69810 }, { "grad_norm": 0.23654112219810486, "learning_rate": 2.290362602356385e-05, "loss": 0.0086, "step": 69820 }, { "grad_norm": 0.14471933245658875, "learning_rate": 2.288973132751765e-05, "loss": 0.0065, "step": 69830 }, { "grad_norm": 0.1737174242734909, "learning_rate": 2.287583959620965e-05, "loss": 0.0089, "step": 69840 }, { "grad_norm": 0.15382139384746552, "learning_rate": 2.2861950831159074e-05, "loss": 0.0087, "step": 69850 }, { "grad_norm": 0.15148715674877167, "learning_rate": 2.284806503388472e-05, "loss": 0.0075, "step": 69860 }, { "grad_norm": 0.10890062153339386, "learning_rate": 2.283418220590518e-05, "loss": 0.0107, "step": 69870 }, { "grad_norm": 0.11557791382074356, "learning_rate": 2.282030234873862e-05, "loss": 0.0071, "step": 69880 }, { "grad_norm": 0.1673569530248642, "learning_rate": 2.2806425463902948e-05, "loss": 0.0071, "step": 69890 }, { "grad_norm": 0.14393846690654755, "learning_rate": 2.279255155291568e-05, "loss": 0.0096, "step": 69900 }, { "grad_norm": 0.09707470238208771, "learning_rate": 2.277868061729409e-05, "loss": 0.0066, "step": 69910 }, { "grad_norm": 0.10331770777702332, "learning_rate": 2.276481265855504e-05, "loss": 0.0091, "step": 69920 }, { "grad_norm": 0.14242836833000183, "learning_rate": 2.2750947678215128e-05, "loss": 0.0087, "step": 69930 }, { "grad_norm": 0.08468464761972427, "learning_rate": 2.2737085677790616e-05, "loss": 0.0068, "step": 69940 }, { "grad_norm": 0.10539644956588745, "learning_rate": 2.2723226658797443e-05, "loss": 0.0079, "step": 69950 }, { "grad_norm": 0.12695415318012238, "learning_rate": 2.270937062275117e-05, "loss": 0.0085, "step": 69960 }, { "grad_norm": 0.12747395038604736, "learning_rate": 2.2695517571167113e-05, "loss": 0.0094, "step": 69970 }, { "grad_norm": 0.11791687458753586, "learning_rate": 2.268166750556019e-05, "loss": 0.0076, "step": 69980 }, { "grad_norm": 0.10675033181905746, "learning_rate": 2.266782042744503e-05, "loss": 0.0085, "step": 69990 }, { "grad_norm": 0.15204298496246338, "learning_rate": 2.2653976338335935e-05, "loss": 0.0067, "step": 70000 }, { "grad_norm": 0.13702207803726196, "learning_rate": 2.2640135239746895e-05, "loss": 0.0105, "step": 70010 }, { "grad_norm": 0.19290608167648315, "learning_rate": 2.2626297133191515e-05, "loss": 0.0107, "step": 70020 }, { "grad_norm": 0.13756205141544342, "learning_rate": 2.261246202018312e-05, "loss": 0.0084, "step": 70030 }, { "grad_norm": 0.15582263469696045, "learning_rate": 2.2598629902234718e-05, "loss": 0.0084, "step": 70040 }, { "grad_norm": 0.1436137706041336, "learning_rate": 2.258480078085894e-05, "loss": 0.0083, "step": 70050 }, { "grad_norm": 0.10972902178764343, "learning_rate": 2.2570974657568126e-05, "loss": 0.0082, "step": 70060 }, { "grad_norm": 0.1602378785610199, "learning_rate": 2.2557151533874286e-05, "loss": 0.0058, "step": 70070 }, { "grad_norm": 0.11237749457359314, "learning_rate": 2.2543331411289116e-05, "loss": 0.009, "step": 70080 }, { "grad_norm": 0.11119262874126434, "learning_rate": 2.252951429132392e-05, "loss": 0.008, "step": 70090 }, { "grad_norm": 0.30248528718948364, "learning_rate": 2.2515700175489762e-05, "loss": 0.0072, "step": 70100 }, { "grad_norm": 0.12467477470636368, "learning_rate": 2.2501889065297288e-05, "loss": 0.0084, "step": 70110 }, { "grad_norm": 0.09335489571094513, "learning_rate": 2.2488080962256887e-05, "loss": 0.0067, "step": 70120 }, { "grad_norm": 0.10477755963802338, "learning_rate": 2.2474275867878585e-05, "loss": 0.0069, "step": 70130 }, { "grad_norm": 0.12112848460674286, "learning_rate": 2.246047378367211e-05, "loss": 0.0091, "step": 70140 }, { "grad_norm": 0.1586175113916397, "learning_rate": 2.2446674711146798e-05, "loss": 0.0115, "step": 70150 }, { "grad_norm": 0.1430826187133789, "learning_rate": 2.2432878651811734e-05, "loss": 0.011, "step": 70160 }, { "grad_norm": 0.15168306231498718, "learning_rate": 2.2419085607175594e-05, "loss": 0.0098, "step": 70170 }, { "grad_norm": 0.16033273935317993, "learning_rate": 2.240529557874678e-05, "loss": 0.008, "step": 70180 }, { "grad_norm": 0.159272238612175, "learning_rate": 2.2391508568033358e-05, "loss": 0.0078, "step": 70190 }, { "grad_norm": 0.14032414555549622, "learning_rate": 2.237772457654307e-05, "loss": 0.0105, "step": 70200 }, { "grad_norm": 0.1316506415605545, "learning_rate": 2.236394360578327e-05, "loss": 0.0089, "step": 70210 }, { "grad_norm": 0.2295970767736435, "learning_rate": 2.2350165657261073e-05, "loss": 0.0122, "step": 70220 }, { "grad_norm": 0.13022319972515106, "learning_rate": 2.2336390732483164e-05, "loss": 0.0086, "step": 70230 }, { "grad_norm": 0.24072407186031342, "learning_rate": 2.2322618832955976e-05, "loss": 0.0117, "step": 70240 }, { "grad_norm": 0.20775435864925385, "learning_rate": 2.2308849960185578e-05, "loss": 0.0078, "step": 70250 }, { "grad_norm": 0.14459648728370667, "learning_rate": 2.229508411567774e-05, "loss": 0.008, "step": 70260 }, { "grad_norm": 0.10750298202037811, "learning_rate": 2.2281321300937823e-05, "loss": 0.0072, "step": 70270 }, { "grad_norm": 0.12325073033571243, "learning_rate": 2.2267561517470937e-05, "loss": 0.0058, "step": 70280 }, { "grad_norm": 0.12439339607954025, "learning_rate": 2.2253804766781845e-05, "loss": 0.0057, "step": 70290 }, { "grad_norm": 0.22563564777374268, "learning_rate": 2.2240051050374933e-05, "loss": 0.0099, "step": 70300 }, { "grad_norm": 0.1421373039484024, "learning_rate": 2.22263003697543e-05, "loss": 0.0067, "step": 70310 }, { "grad_norm": 0.20044992864131927, "learning_rate": 2.2212552726423702e-05, "loss": 0.0079, "step": 70320 }, { "grad_norm": 0.11581553518772125, "learning_rate": 2.2198808121886578e-05, "loss": 0.0076, "step": 70330 }, { "grad_norm": 0.10697264224290848, "learning_rate": 2.218506655764598e-05, "loss": 0.0079, "step": 70340 }, { "grad_norm": 0.14963285624980927, "learning_rate": 2.2171328035204704e-05, "loss": 0.0102, "step": 70350 }, { "grad_norm": 0.11269992589950562, "learning_rate": 2.2157592556065133e-05, "loss": 0.0095, "step": 70360 }, { "grad_norm": 0.15877795219421387, "learning_rate": 2.2143860121729382e-05, "loss": 0.0102, "step": 70370 }, { "grad_norm": 0.13263878226280212, "learning_rate": 2.2130130733699206e-05, "loss": 0.0095, "step": 70380 }, { "grad_norm": 0.16199862957000732, "learning_rate": 2.2116404393476052e-05, "loss": 0.0066, "step": 70390 }, { "grad_norm": 0.17110607028007507, "learning_rate": 2.2102681102560967e-05, "loss": 0.0152, "step": 70400 }, { "grad_norm": 0.15948183834552765, "learning_rate": 2.208896086245476e-05, "loss": 0.0084, "step": 70410 }, { "grad_norm": 0.13725298643112183, "learning_rate": 2.2075243674657808e-05, "loss": 0.0075, "step": 70420 }, { "grad_norm": 0.10691259056329727, "learning_rate": 2.206152954067024e-05, "loss": 0.0062, "step": 70430 }, { "grad_norm": 0.17822803556919098, "learning_rate": 2.2047818461991758e-05, "loss": 0.0081, "step": 70440 }, { "grad_norm": 0.12374747544527054, "learning_rate": 2.2034110440121864e-05, "loss": 0.0058, "step": 70450 }, { "grad_norm": 0.10444685071706772, "learning_rate": 2.2020405476559585e-05, "loss": 0.0087, "step": 70460 }, { "grad_norm": 0.09540476649999619, "learning_rate": 2.2006703572803722e-05, "loss": 0.0077, "step": 70470 }, { "grad_norm": 0.12330307066440582, "learning_rate": 2.1993004730352646e-05, "loss": 0.0077, "step": 70480 }, { "grad_norm": 0.13372763991355896, "learning_rate": 2.1979308950704464e-05, "loss": 0.0069, "step": 70490 }, { "grad_norm": 0.0984203889966011, "learning_rate": 2.1965616235356923e-05, "loss": 0.0067, "step": 70500 }, { "grad_norm": 0.14268817007541656, "learning_rate": 2.1951926585807465e-05, "loss": 0.0094, "step": 70510 }, { "grad_norm": 0.10401427745819092, "learning_rate": 2.193824000355313e-05, "loss": 0.0071, "step": 70520 }, { "grad_norm": 0.15738356113433838, "learning_rate": 2.192455649009067e-05, "loss": 0.0069, "step": 70530 }, { "grad_norm": 0.10265757888555527, "learning_rate": 2.1910876046916523e-05, "loss": 0.0089, "step": 70540 }, { "grad_norm": 0.13420909643173218, "learning_rate": 2.1897198675526725e-05, "loss": 0.0054, "step": 70550 }, { "grad_norm": 0.12562957406044006, "learning_rate": 2.1883524377417024e-05, "loss": 0.0062, "step": 70560 }, { "grad_norm": 0.10969071090221405, "learning_rate": 2.1869853154082825e-05, "loss": 0.0067, "step": 70570 }, { "grad_norm": 0.10739029198884964, "learning_rate": 2.1856185007019215e-05, "loss": 0.0108, "step": 70580 }, { "grad_norm": 0.14630651473999023, "learning_rate": 2.1842519937720874e-05, "loss": 0.0092, "step": 70590 }, { "grad_norm": 0.09082668274641037, "learning_rate": 2.182885794768224e-05, "loss": 0.0081, "step": 70600 }, { "grad_norm": 0.1775277853012085, "learning_rate": 2.1815199038397326e-05, "loss": 0.0091, "step": 70610 }, { "grad_norm": 0.13816730678081512, "learning_rate": 2.1801543211359893e-05, "loss": 0.0087, "step": 70620 }, { "grad_norm": 0.13229776918888092, "learning_rate": 2.178789046806326e-05, "loss": 0.0085, "step": 70630 }, { "grad_norm": 0.12407439202070236, "learning_rate": 2.1774240810000547e-05, "loss": 0.0086, "step": 70640 }, { "grad_norm": 0.10892750322818756, "learning_rate": 2.17605942386644e-05, "loss": 0.0079, "step": 70650 }, { "grad_norm": 0.154094859957695, "learning_rate": 2.174695075554723e-05, "loss": 0.0111, "step": 70660 }, { "grad_norm": 0.15861044824123383, "learning_rate": 2.173331036214103e-05, "loss": 0.0105, "step": 70670 }, { "grad_norm": 0.13544003665447235, "learning_rate": 2.171967305993753e-05, "loss": 0.0087, "step": 70680 }, { "grad_norm": 0.20807699859142303, "learning_rate": 2.1706038850428024e-05, "loss": 0.0111, "step": 70690 }, { "grad_norm": 0.08778616040945053, "learning_rate": 2.169240773510361e-05, "loss": 0.0055, "step": 70700 }, { "grad_norm": 0.10986320674419403, "learning_rate": 2.1678779715454906e-05, "loss": 0.0071, "step": 70710 }, { "grad_norm": 0.10753143578767776, "learning_rate": 2.166515479297229e-05, "loss": 0.0064, "step": 70720 }, { "grad_norm": 0.1557237207889557, "learning_rate": 2.1651532969145732e-05, "loss": 0.0132, "step": 70730 }, { "grad_norm": 0.1348300278186798, "learning_rate": 2.1637914245464903e-05, "loss": 0.0082, "step": 70740 }, { "grad_norm": 0.11083085089921951, "learning_rate": 2.1624298623419133e-05, "loss": 0.0063, "step": 70750 }, { "grad_norm": 0.12389275431632996, "learning_rate": 2.1610686104497413e-05, "loss": 0.0056, "step": 70760 }, { "grad_norm": 0.11521996557712555, "learning_rate": 2.1597076690188366e-05, "loss": 0.0116, "step": 70770 }, { "grad_norm": 0.14616568386554718, "learning_rate": 2.1583470381980315e-05, "loss": 0.0054, "step": 70780 }, { "grad_norm": 0.13463720679283142, "learning_rate": 2.156986718136123e-05, "loss": 0.0085, "step": 70790 }, { "grad_norm": 0.15704302489757538, "learning_rate": 2.155626708981871e-05, "loss": 0.0089, "step": 70800 }, { "grad_norm": 0.12187696993350983, "learning_rate": 2.1542670108840075e-05, "loss": 0.0086, "step": 70810 }, { "grad_norm": 0.20693375170230865, "learning_rate": 2.152907623991221e-05, "loss": 0.0059, "step": 70820 }, { "grad_norm": 0.20378755033016205, "learning_rate": 2.1515485484521802e-05, "loss": 0.0084, "step": 70830 }, { "grad_norm": 0.15758861601352692, "learning_rate": 2.150189784415506e-05, "loss": 0.0109, "step": 70840 }, { "grad_norm": 0.10282571613788605, "learning_rate": 2.148831332029794e-05, "loss": 0.0095, "step": 70850 }, { "grad_norm": 0.08077589422464371, "learning_rate": 2.147473191443599e-05, "loss": 0.0067, "step": 70860 }, { "grad_norm": 0.16237910091876984, "learning_rate": 2.1461153628054492e-05, "loss": 0.0116, "step": 70870 }, { "grad_norm": 0.2064485102891922, "learning_rate": 2.1447578462638284e-05, "loss": 0.0078, "step": 70880 }, { "grad_norm": 0.09369370341300964, "learning_rate": 2.143400641967201e-05, "loss": 0.0082, "step": 70890 }, { "grad_norm": 0.18955963850021362, "learning_rate": 2.1420437500639823e-05, "loss": 0.0072, "step": 70900 }, { "grad_norm": 0.10944400727748871, "learning_rate": 2.1406871707025646e-05, "loss": 0.0052, "step": 70910 }, { "grad_norm": 0.1099577397108078, "learning_rate": 2.1393309040312963e-05, "loss": 0.0107, "step": 70920 }, { "grad_norm": 0.20796102285385132, "learning_rate": 2.137974950198502e-05, "loss": 0.0103, "step": 70930 }, { "grad_norm": 0.15563564002513885, "learning_rate": 2.1366193093524617e-05, "loss": 0.008, "step": 70940 }, { "grad_norm": 0.10673739016056061, "learning_rate": 2.1352639816414288e-05, "loss": 0.0084, "step": 70950 }, { "grad_norm": 0.14662520587444305, "learning_rate": 2.1339089672136202e-05, "loss": 0.0094, "step": 70960 }, { "grad_norm": 0.11707782000303268, "learning_rate": 2.1325542662172193e-05, "loss": 0.0067, "step": 70970 }, { "grad_norm": 0.14192532002925873, "learning_rate": 2.1311998788003712e-05, "loss": 0.0064, "step": 70980 }, { "grad_norm": 0.08440794795751572, "learning_rate": 2.1298458051111907e-05, "loss": 0.0124, "step": 70990 }, { "grad_norm": 0.16774910688400269, "learning_rate": 2.1284920452977602e-05, "loss": 0.0076, "step": 71000 }, { "grad_norm": 0.1637561023235321, "learning_rate": 2.12713859950812e-05, "loss": 0.013, "step": 71010 }, { "grad_norm": 0.13655558228492737, "learning_rate": 2.1257854678902832e-05, "loss": 0.0097, "step": 71020 }, { "grad_norm": 0.1797245293855667, "learning_rate": 2.124432650592227e-05, "loss": 0.0088, "step": 71030 }, { "grad_norm": 0.18315811455249786, "learning_rate": 2.1230801477618937e-05, "loss": 0.0077, "step": 71040 }, { "grad_norm": 0.1226547509431839, "learning_rate": 2.121727959547189e-05, "loss": 0.0084, "step": 71050 }, { "grad_norm": 0.19589945673942566, "learning_rate": 2.1203760860959886e-05, "loss": 0.0086, "step": 71060 }, { "grad_norm": 0.14389942586421967, "learning_rate": 2.119024527556127e-05, "loss": 0.0087, "step": 71070 }, { "grad_norm": 0.16130410134792328, "learning_rate": 2.117673284075415e-05, "loss": 0.0079, "step": 71080 }, { "grad_norm": 0.1878150999546051, "learning_rate": 2.1163223558016165e-05, "loss": 0.0081, "step": 71090 }, { "grad_norm": 0.12046666443347931, "learning_rate": 2.1149717428824722e-05, "loss": 0.007, "step": 71100 }, { "grad_norm": 0.14755438268184662, "learning_rate": 2.1136214454656784e-05, "loss": 0.0092, "step": 71110 }, { "grad_norm": 0.10425413399934769, "learning_rate": 2.1122714636989056e-05, "loss": 0.0068, "step": 71120 }, { "grad_norm": 0.15051400661468506, "learning_rate": 2.1109217977297823e-05, "loss": 0.0098, "step": 71130 }, { "grad_norm": 0.15224668383598328, "learning_rate": 2.1095724477059077e-05, "loss": 0.0081, "step": 71140 }, { "grad_norm": 0.1081179529428482, "learning_rate": 2.1082234137748434e-05, "loss": 0.007, "step": 71150 }, { "grad_norm": 0.13052019476890564, "learning_rate": 2.106874696084122e-05, "loss": 0.008, "step": 71160 }, { "grad_norm": 0.14763137698173523, "learning_rate": 2.1055262947812315e-05, "loss": 0.0086, "step": 71170 }, { "grad_norm": 0.12130986899137497, "learning_rate": 2.104178210013636e-05, "loss": 0.0083, "step": 71180 }, { "grad_norm": 0.15728294849395752, "learning_rate": 2.1028304419287554e-05, "loss": 0.0079, "step": 71190 }, { "grad_norm": 0.14247079193592072, "learning_rate": 2.101482990673982e-05, "loss": 0.0078, "step": 71200 }, { "grad_norm": 0.15268193185329437, "learning_rate": 2.1001358563966707e-05, "loss": 0.0067, "step": 71210 }, { "grad_norm": 0.2166484296321869, "learning_rate": 2.098789039244144e-05, "loss": 0.0089, "step": 71220 }, { "grad_norm": 0.152792289853096, "learning_rate": 2.097442539363685e-05, "loss": 0.0088, "step": 71230 }, { "grad_norm": 0.11647035926580429, "learning_rate": 2.0960963569025448e-05, "loss": 0.0077, "step": 71240 }, { "grad_norm": 0.12285827845335007, "learning_rate": 2.0947504920079435e-05, "loss": 0.0079, "step": 71250 }, { "grad_norm": 0.13234412670135498, "learning_rate": 2.093404944827059e-05, "loss": 0.0083, "step": 71260 }, { "grad_norm": 0.14266261458396912, "learning_rate": 2.0920597155070394e-05, "loss": 0.0082, "step": 71270 }, { "grad_norm": 0.21530623733997345, "learning_rate": 2.090714804194997e-05, "loss": 0.0111, "step": 71280 }, { "grad_norm": 0.11306190490722656, "learning_rate": 2.0893702110380114e-05, "loss": 0.006, "step": 71290 }, { "grad_norm": 0.11124326288700104, "learning_rate": 2.0880259361831216e-05, "loss": 0.0065, "step": 71300 }, { "grad_norm": 0.11497551202774048, "learning_rate": 2.0866819797773397e-05, "loss": 0.0076, "step": 71310 }, { "grad_norm": 0.13233202695846558, "learning_rate": 2.085338341967634e-05, "loss": 0.0084, "step": 71320 }, { "grad_norm": 0.1820402294397354, "learning_rate": 2.083995022900946e-05, "loss": 0.0064, "step": 71330 }, { "grad_norm": 0.08793649077415466, "learning_rate": 2.0826520227241768e-05, "loss": 0.0094, "step": 71340 }, { "grad_norm": 0.17233429849147797, "learning_rate": 2.081309341584199e-05, "loss": 0.0088, "step": 71350 }, { "grad_norm": 0.16842786967754364, "learning_rate": 2.0799669796278413e-05, "loss": 0.0072, "step": 71360 }, { "grad_norm": 0.2000318020582199, "learning_rate": 2.078624937001906e-05, "loss": 0.0106, "step": 71370 }, { "grad_norm": 0.2527008354663849, "learning_rate": 2.0772832138531544e-05, "loss": 0.0094, "step": 71380 }, { "grad_norm": 0.23782262206077576, "learning_rate": 2.0759418103283158e-05, "loss": 0.0088, "step": 71390 }, { "grad_norm": 0.13537704944610596, "learning_rate": 2.0746007265740846e-05, "loss": 0.0086, "step": 71400 }, { "grad_norm": 0.2188376486301422, "learning_rate": 2.0732599627371215e-05, "loss": 0.0078, "step": 71410 }, { "grad_norm": 0.11641319841146469, "learning_rate": 2.0719195189640468e-05, "loss": 0.0073, "step": 71420 }, { "grad_norm": 0.2010340392589569, "learning_rate": 2.0705793954014528e-05, "loss": 0.014, "step": 71430 }, { "grad_norm": 0.11636196076869965, "learning_rate": 2.06923959219589e-05, "loss": 0.0076, "step": 71440 }, { "grad_norm": 0.12556767463684082, "learning_rate": 2.0679001094938793e-05, "loss": 0.0088, "step": 71450 }, { "grad_norm": 0.18475820124149323, "learning_rate": 2.066560947441904e-05, "loss": 0.0109, "step": 71460 }, { "grad_norm": 0.09969884902238846, "learning_rate": 2.0652221061864152e-05, "loss": 0.007, "step": 71470 }, { "grad_norm": 0.13679271936416626, "learning_rate": 2.0638835858738226e-05, "loss": 0.0075, "step": 71480 }, { "grad_norm": 0.10043603181838989, "learning_rate": 2.0625453866505068e-05, "loss": 0.0061, "step": 71490 }, { "grad_norm": 0.14395256340503693, "learning_rate": 2.0612075086628123e-05, "loss": 0.0087, "step": 71500 }, { "grad_norm": 0.14961287379264832, "learning_rate": 2.059869952057045e-05, "loss": 0.0082, "step": 71510 }, { "grad_norm": 0.1046438217163086, "learning_rate": 2.0585327169794793e-05, "loss": 0.0069, "step": 71520 }, { "grad_norm": 0.1290416568517685, "learning_rate": 2.0571958035763534e-05, "loss": 0.0069, "step": 71530 }, { "grad_norm": 0.12881141901016235, "learning_rate": 2.0558592119938715e-05, "loss": 0.0105, "step": 71540 }, { "grad_norm": 0.09812600910663605, "learning_rate": 2.054522942378198e-05, "loss": 0.0061, "step": 71550 }, { "grad_norm": 0.22258034348487854, "learning_rate": 2.053186994875469e-05, "loss": 0.0072, "step": 71560 }, { "grad_norm": 0.08388128131628036, "learning_rate": 2.0518513696317775e-05, "loss": 0.0072, "step": 71570 }, { "grad_norm": 0.1345667541027069, "learning_rate": 2.050516066793188e-05, "loss": 0.006, "step": 71580 }, { "grad_norm": 0.17372608184814453, "learning_rate": 2.0491810865057276e-05, "loss": 0.0086, "step": 71590 }, { "grad_norm": 0.1330404132604599, "learning_rate": 2.0478464289153882e-05, "loss": 0.0078, "step": 71600 }, { "grad_norm": 0.12504322826862335, "learning_rate": 2.0465120941681232e-05, "loss": 0.0076, "step": 71610 }, { "grad_norm": 0.12250372767448425, "learning_rate": 2.0451780824098566e-05, "loss": 0.0092, "step": 71620 }, { "grad_norm": 0.14120715856552124, "learning_rate": 2.043844393786471e-05, "loss": 0.0068, "step": 71630 }, { "grad_norm": 0.09944300353527069, "learning_rate": 2.0425110284438193e-05, "loss": 0.0067, "step": 71640 }, { "grad_norm": 0.11266268044710159, "learning_rate": 2.041177986527712e-05, "loss": 0.0077, "step": 71650 }, { "grad_norm": 0.08933315426111221, "learning_rate": 2.0398452681839347e-05, "loss": 0.007, "step": 71660 }, { "grad_norm": 0.13393881916999817, "learning_rate": 2.0385128735582272e-05, "loss": 0.0076, "step": 71670 }, { "grad_norm": 0.13667066395282745, "learning_rate": 2.0371808027963003e-05, "loss": 0.0095, "step": 71680 }, { "grad_norm": 0.07496407628059387, "learning_rate": 2.035849056043825e-05, "loss": 0.0085, "step": 71690 }, { "grad_norm": 0.143389031291008, "learning_rate": 2.03451763344644e-05, "loss": 0.0062, "step": 71700 }, { "grad_norm": 0.1372164934873581, "learning_rate": 2.0331865351497482e-05, "loss": 0.0063, "step": 71710 }, { "grad_norm": 0.1686871200799942, "learning_rate": 2.0318557612993185e-05, "loss": 0.0068, "step": 71720 }, { "grad_norm": 0.17054112255573273, "learning_rate": 2.0305253120406782e-05, "loss": 0.0091, "step": 71730 }, { "grad_norm": 0.11693807691335678, "learning_rate": 2.029195187519325e-05, "loss": 0.0094, "step": 71740 }, { "grad_norm": 0.08640683442354202, "learning_rate": 2.0278653878807217e-05, "loss": 0.0064, "step": 71750 }, { "grad_norm": 0.14622265100479126, "learning_rate": 2.0265359132702892e-05, "loss": 0.0059, "step": 71760 }, { "grad_norm": 0.12178654968738556, "learning_rate": 2.0252067638334182e-05, "loss": 0.0056, "step": 71770 }, { "grad_norm": 0.15800635516643524, "learning_rate": 2.023877939715464e-05, "loss": 0.0084, "step": 71780 }, { "grad_norm": 0.08850040286779404, "learning_rate": 2.0225494410617448e-05, "loss": 0.0068, "step": 71790 }, { "grad_norm": 0.15151219069957733, "learning_rate": 2.02122126801754e-05, "loss": 0.0061, "step": 71800 }, { "grad_norm": 0.12856635451316833, "learning_rate": 2.0198934207281018e-05, "loss": 0.0092, "step": 71810 }, { "grad_norm": 0.12710422277450562, "learning_rate": 2.0185658993386357e-05, "loss": 0.008, "step": 71820 }, { "grad_norm": 0.10156207531690598, "learning_rate": 2.0172387039943226e-05, "loss": 0.0062, "step": 71830 }, { "grad_norm": 0.20304536819458008, "learning_rate": 2.0159118348402968e-05, "loss": 0.0081, "step": 71840 }, { "grad_norm": 0.1584852635860443, "learning_rate": 2.0145852920216697e-05, "loss": 0.0063, "step": 71850 }, { "grad_norm": 0.12316646426916122, "learning_rate": 2.0132590756835045e-05, "loss": 0.0061, "step": 71860 }, { "grad_norm": 0.13352039456367493, "learning_rate": 2.0119331859708386e-05, "loss": 0.0071, "step": 71870 }, { "grad_norm": 0.11530547589063644, "learning_rate": 2.0106076230286654e-05, "loss": 0.0067, "step": 71880 }, { "grad_norm": 0.1256868839263916, "learning_rate": 2.0092823870019488e-05, "loss": 0.0082, "step": 71890 }, { "grad_norm": 0.1245754137635231, "learning_rate": 2.0079574780356115e-05, "loss": 0.0068, "step": 71900 }, { "grad_norm": 0.08974639326334, "learning_rate": 2.0066328962745486e-05, "loss": 0.0065, "step": 71910 }, { "grad_norm": 0.10955610126256943, "learning_rate": 2.00530864186361e-05, "loss": 0.0068, "step": 71920 }, { "grad_norm": 0.1431102603673935, "learning_rate": 2.003984714947618e-05, "loss": 0.0074, "step": 71930 }, { "grad_norm": 0.1326378732919693, "learning_rate": 2.0026611156713515e-05, "loss": 0.0074, "step": 71940 }, { "grad_norm": 0.0856412723660469, "learning_rate": 2.0013378441795584e-05, "loss": 0.008, "step": 71950 }, { "grad_norm": 0.1758226752281189, "learning_rate": 2.0000149006169506e-05, "loss": 0.0069, "step": 71960 }, { "grad_norm": 0.16101327538490295, "learning_rate": 1.9986922851282048e-05, "loss": 0.0104, "step": 71970 }, { "grad_norm": 0.14328759908676147, "learning_rate": 1.997369997857956e-05, "loss": 0.008, "step": 71980 }, { "grad_norm": 0.12504124641418457, "learning_rate": 1.9960480389508103e-05, "loss": 0.0072, "step": 71990 }, { "grad_norm": 0.18895795941352844, "learning_rate": 1.9947264085513363e-05, "loss": 0.0111, "step": 72000 }, { "grad_norm": 0.08085764199495316, "learning_rate": 1.993405106804062e-05, "loss": 0.0083, "step": 72010 }, { "grad_norm": 0.16767509281635284, "learning_rate": 1.9920841338534867e-05, "loss": 0.0074, "step": 72020 }, { "grad_norm": 0.08578360825777054, "learning_rate": 1.9907634898440646e-05, "loss": 0.0058, "step": 72030 }, { "grad_norm": 0.13603699207305908, "learning_rate": 1.989443174920227e-05, "loss": 0.006, "step": 72040 }, { "grad_norm": 0.1418984830379486, "learning_rate": 1.988123189226355e-05, "loss": 0.0083, "step": 72050 }, { "grad_norm": 0.15678413212299347, "learning_rate": 1.986803532906805e-05, "loss": 0.0109, "step": 72060 }, { "grad_norm": 0.18029415607452393, "learning_rate": 1.985484206105889e-05, "loss": 0.0072, "step": 72070 }, { "grad_norm": 0.10527034848928452, "learning_rate": 1.9841652089678887e-05, "loss": 0.0094, "step": 72080 }, { "grad_norm": 0.09927666932344437, "learning_rate": 1.9828465416370435e-05, "loss": 0.0083, "step": 72090 }, { "grad_norm": 0.16314436495304108, "learning_rate": 1.9815282042575685e-05, "loss": 0.0066, "step": 72100 }, { "grad_norm": 0.12957681715488434, "learning_rate": 1.9802101969736285e-05, "loss": 0.0085, "step": 72110 }, { "grad_norm": 0.10282289236783981, "learning_rate": 1.9788925199293633e-05, "loss": 0.007, "step": 72120 }, { "grad_norm": 0.12121088802814484, "learning_rate": 1.977575173268868e-05, "loss": 0.0076, "step": 72130 }, { "grad_norm": 0.15577352046966553, "learning_rate": 1.9762581571362094e-05, "loss": 0.0082, "step": 72140 }, { "grad_norm": 0.10437142848968506, "learning_rate": 1.9749414716754112e-05, "loss": 0.0098, "step": 72150 }, { "grad_norm": 0.15862569212913513, "learning_rate": 1.973625117030465e-05, "loss": 0.0096, "step": 72160 }, { "grad_norm": 0.09340568631887436, "learning_rate": 1.972309093345326e-05, "loss": 0.0091, "step": 72170 }, { "grad_norm": 0.12188761681318283, "learning_rate": 1.970993400763915e-05, "loss": 0.0094, "step": 72180 }, { "grad_norm": 0.13651752471923828, "learning_rate": 1.969678039430109e-05, "loss": 0.0081, "step": 72190 }, { "grad_norm": 0.10381461679935455, "learning_rate": 1.9683630094877564e-05, "loss": 0.0093, "step": 72200 }, { "grad_norm": 0.10671503841876984, "learning_rate": 1.9670483110806696e-05, "loss": 0.0086, "step": 72210 }, { "grad_norm": 0.08554209768772125, "learning_rate": 1.9657339443526172e-05, "loss": 0.0059, "step": 72220 }, { "grad_norm": 0.13069039583206177, "learning_rate": 1.964419909447339e-05, "loss": 0.0059, "step": 72230 }, { "grad_norm": 0.15114392340183258, "learning_rate": 1.963106206508536e-05, "loss": 0.0081, "step": 72240 }, { "grad_norm": 0.1264341026544571, "learning_rate": 1.9617928356798738e-05, "loss": 0.0067, "step": 72250 }, { "grad_norm": 0.2019815593957901, "learning_rate": 1.960479797104977e-05, "loss": 0.0065, "step": 72260 }, { "grad_norm": 0.12677468359470367, "learning_rate": 1.9591670909274418e-05, "loss": 0.007, "step": 72270 }, { "grad_norm": 0.09070975333452225, "learning_rate": 1.9578547172908184e-05, "loss": 0.0062, "step": 72280 }, { "grad_norm": 0.1288401037454605, "learning_rate": 1.9565426763386325e-05, "loss": 0.0073, "step": 72290 }, { "grad_norm": 0.12847736477851868, "learning_rate": 1.9552309682143622e-05, "loss": 0.0076, "step": 72300 }, { "grad_norm": 0.1585085093975067, "learning_rate": 1.9539195930614574e-05, "loss": 0.0099, "step": 72310 }, { "grad_norm": 0.11467245221138, "learning_rate": 1.952608551023324e-05, "loss": 0.0054, "step": 72320 }, { "grad_norm": 0.19096563756465912, "learning_rate": 1.9512978422433397e-05, "loss": 0.0114, "step": 72330 }, { "grad_norm": 0.11512588709592819, "learning_rate": 1.9499874668648376e-05, "loss": 0.0068, "step": 72340 }, { "grad_norm": 0.10639552772045135, "learning_rate": 1.9486774250311213e-05, "loss": 0.0077, "step": 72350 }, { "grad_norm": 0.11252257227897644, "learning_rate": 1.9473677168854537e-05, "loss": 0.0076, "step": 72360 }, { "grad_norm": 0.08574293553829193, "learning_rate": 1.946058342571065e-05, "loss": 0.0077, "step": 72370 }, { "grad_norm": 0.10500593483448029, "learning_rate": 1.9447493022311424e-05, "loss": 0.0069, "step": 72380 }, { "grad_norm": 0.11773674935102463, "learning_rate": 1.9434405960088442e-05, "loss": 0.0064, "step": 72390 }, { "grad_norm": 0.11352776736021042, "learning_rate": 1.9421322240472857e-05, "loss": 0.0071, "step": 72400 }, { "grad_norm": 0.15154197812080383, "learning_rate": 1.9408241864895494e-05, "loss": 0.0073, "step": 72410 }, { "grad_norm": 0.1431552618741989, "learning_rate": 1.939516483478681e-05, "loss": 0.0074, "step": 72420 }, { "grad_norm": 0.11923334002494812, "learning_rate": 1.9382091151576902e-05, "loss": 0.0064, "step": 72430 }, { "grad_norm": 0.17909382283687592, "learning_rate": 1.9369020816695454e-05, "loss": 0.0065, "step": 72440 }, { "grad_norm": 0.17649173736572266, "learning_rate": 1.935595383157184e-05, "loss": 0.008, "step": 72450 }, { "grad_norm": 0.12461019307374954, "learning_rate": 1.9342890197635057e-05, "loss": 0.0093, "step": 72460 }, { "grad_norm": 0.16975931823253632, "learning_rate": 1.9329829916313685e-05, "loss": 0.0079, "step": 72470 }, { "grad_norm": 0.1621045023202896, "learning_rate": 1.9316772989036013e-05, "loss": 0.0089, "step": 72480 }, { "grad_norm": 0.07732554525136948, "learning_rate": 1.930371941722991e-05, "loss": 0.0079, "step": 72490 }, { "grad_norm": 0.12346649914979935, "learning_rate": 1.929066920232292e-05, "loss": 0.0071, "step": 72500 }, { "grad_norm": 0.12840087711811066, "learning_rate": 1.9277622345742152e-05, "loss": 0.0064, "step": 72510 }, { "grad_norm": 0.12224914133548737, "learning_rate": 1.926457884891444e-05, "loss": 0.0089, "step": 72520 }, { "grad_norm": 0.1332883983850479, "learning_rate": 1.9251538713266147e-05, "loss": 0.0079, "step": 72530 }, { "grad_norm": 0.08825141191482544, "learning_rate": 1.9238501940223358e-05, "loss": 0.0091, "step": 72540 }, { "grad_norm": 0.1251579225063324, "learning_rate": 1.922546853121175e-05, "loss": 0.0079, "step": 72550 }, { "grad_norm": 0.1449860781431198, "learning_rate": 1.9212438487656647e-05, "loss": 0.0083, "step": 72560 }, { "grad_norm": 0.1140928864479065, "learning_rate": 1.9199411810982965e-05, "loss": 0.0064, "step": 72570 }, { "grad_norm": 0.13581852614879608, "learning_rate": 1.918638850261532e-05, "loss": 0.0085, "step": 72580 }, { "grad_norm": 0.1206030547618866, "learning_rate": 1.9173368563977885e-05, "loss": 0.008, "step": 72590 }, { "grad_norm": 0.12958119809627533, "learning_rate": 1.9160351996494513e-05, "loss": 0.0071, "step": 72600 }, { "grad_norm": 0.13537068665027618, "learning_rate": 1.9147338801588677e-05, "loss": 0.0068, "step": 72610 }, { "grad_norm": 0.18734262883663177, "learning_rate": 1.913432898068351e-05, "loss": 0.0071, "step": 72620 }, { "grad_norm": 0.12275472283363342, "learning_rate": 1.912132253520169e-05, "loss": 0.007, "step": 72630 }, { "grad_norm": 0.188710018992424, "learning_rate": 1.910831946656564e-05, "loss": 0.0071, "step": 72640 }, { "grad_norm": 0.10163525491952896, "learning_rate": 1.909531977619731e-05, "loss": 0.0068, "step": 72650 }, { "grad_norm": 0.09738617390394211, "learning_rate": 1.908232346551834e-05, "loss": 0.0097, "step": 72660 }, { "grad_norm": 0.14639727771282196, "learning_rate": 1.9069330535949993e-05, "loss": 0.0083, "step": 72670 }, { "grad_norm": 0.10984595119953156, "learning_rate": 1.9056340988913167e-05, "loss": 0.0082, "step": 72680 }, { "grad_norm": 0.12695099413394928, "learning_rate": 1.9043354825828357e-05, "loss": 0.0052, "step": 72690 }, { "grad_norm": 0.1541469246149063, "learning_rate": 1.9030372048115724e-05, "loss": 0.0065, "step": 72700 }, { "grad_norm": 0.13309790194034576, "learning_rate": 1.9017392657195056e-05, "loss": 0.0081, "step": 72710 }, { "grad_norm": 0.12866997718811035, "learning_rate": 1.9004416654485724e-05, "loss": 0.0049, "step": 72720 }, { "grad_norm": 0.08080074191093445, "learning_rate": 1.8991444041406785e-05, "loss": 0.0074, "step": 72730 }, { "grad_norm": 0.1142958551645279, "learning_rate": 1.8978474819376912e-05, "loss": 0.0072, "step": 72740 }, { "grad_norm": 0.09339091181755066, "learning_rate": 1.8965508989814406e-05, "loss": 0.0079, "step": 72750 }, { "grad_norm": 0.08710122108459473, "learning_rate": 1.8952546554137158e-05, "loss": 0.0094, "step": 72760 }, { "grad_norm": 0.09393121302127838, "learning_rate": 1.8939587513762757e-05, "loss": 0.0088, "step": 72770 }, { "grad_norm": 0.11774710565805435, "learning_rate": 1.892663187010835e-05, "loss": 0.0089, "step": 72780 }, { "grad_norm": 0.12716512382030487, "learning_rate": 1.891367962459077e-05, "loss": 0.0066, "step": 72790 }, { "grad_norm": 0.10495226830244064, "learning_rate": 1.890073077862644e-05, "loss": 0.0069, "step": 72800 }, { "grad_norm": 0.12369215488433838, "learning_rate": 1.888778533363146e-05, "loss": 0.0066, "step": 72810 }, { "grad_norm": 0.13864248991012573, "learning_rate": 1.8874843291021478e-05, "loss": 0.0073, "step": 72820 }, { "grad_norm": 0.14620284736156464, "learning_rate": 1.8861904652211853e-05, "loss": 0.0053, "step": 72830 }, { "grad_norm": 0.1178869679570198, "learning_rate": 1.8848969418617507e-05, "loss": 0.0074, "step": 72840 }, { "grad_norm": 0.08848705142736435, "learning_rate": 1.8836037591653044e-05, "loss": 0.0072, "step": 72850 }, { "grad_norm": 0.08883080631494522, "learning_rate": 1.882310917273262e-05, "loss": 0.0085, "step": 72860 }, { "grad_norm": 0.13353346288204193, "learning_rate": 1.881018416327014e-05, "loss": 0.0062, "step": 72870 }, { "grad_norm": 0.168624147772789, "learning_rate": 1.8797262564679014e-05, "loss": 0.0084, "step": 72880 }, { "grad_norm": 0.1256076842546463, "learning_rate": 1.8784344378372355e-05, "loss": 0.0072, "step": 72890 }, { "grad_norm": 0.16408412158489227, "learning_rate": 1.8771429605762843e-05, "loss": 0.0081, "step": 72900 }, { "grad_norm": 0.1670767068862915, "learning_rate": 1.8758518248262835e-05, "loss": 0.0081, "step": 72910 }, { "grad_norm": 0.10941266268491745, "learning_rate": 1.8745610307284305e-05, "loss": 0.0058, "step": 72920 }, { "grad_norm": 0.16096442937850952, "learning_rate": 1.8732705784238862e-05, "loss": 0.0092, "step": 72930 }, { "grad_norm": 0.12381988763809204, "learning_rate": 1.8719804680537682e-05, "loss": 0.0065, "step": 72940 }, { "grad_norm": 0.10186226665973663, "learning_rate": 1.870690699759164e-05, "loss": 0.0053, "step": 72950 }, { "grad_norm": 0.09298716485500336, "learning_rate": 1.8694012736811213e-05, "loss": 0.0087, "step": 72960 }, { "grad_norm": 0.11646966636180878, "learning_rate": 1.8681121899606463e-05, "loss": 0.0071, "step": 72970 }, { "grad_norm": 0.13559037446975708, "learning_rate": 1.8668234487387144e-05, "loss": 0.0074, "step": 72980 }, { "grad_norm": 0.13464710116386414, "learning_rate": 1.8655350501562592e-05, "loss": 0.0057, "step": 72990 }, { "grad_norm": 0.20547299087047577, "learning_rate": 1.8642469943541796e-05, "loss": 0.0078, "step": 73000 }, { "grad_norm": 0.15164439380168915, "learning_rate": 1.8629592814733328e-05, "loss": 0.0054, "step": 73010 }, { "grad_norm": 0.10482706874608994, "learning_rate": 1.8616719116545438e-05, "loss": 0.0054, "step": 73020 }, { "grad_norm": 0.11297681927680969, "learning_rate": 1.8603848850385952e-05, "loss": 0.0071, "step": 73030 }, { "grad_norm": 0.1672246903181076, "learning_rate": 1.859098201766236e-05, "loss": 0.0047, "step": 73040 }, { "grad_norm": 0.12361402809619904, "learning_rate": 1.8578118619781726e-05, "loss": 0.006, "step": 73050 }, { "grad_norm": 0.11182169616222382, "learning_rate": 1.856525865815083e-05, "loss": 0.007, "step": 73060 }, { "grad_norm": 0.07608571648597717, "learning_rate": 1.8552402134175966e-05, "loss": 0.0089, "step": 73070 }, { "grad_norm": 0.16963718831539154, "learning_rate": 1.8539549049263143e-05, "loss": 0.0068, "step": 73080 }, { "grad_norm": 0.09530201554298401, "learning_rate": 1.8526699404817917e-05, "loss": 0.0058, "step": 73090 }, { "grad_norm": 0.1064385324716568, "learning_rate": 1.8513853202245546e-05, "loss": 0.0068, "step": 73100 }, { "grad_norm": 0.1427047997713089, "learning_rate": 1.850101044295081e-05, "loss": 0.0058, "step": 73110 }, { "grad_norm": 0.12348852306604385, "learning_rate": 1.848817112833825e-05, "loss": 0.0062, "step": 73120 }, { "grad_norm": 0.1447162926197052, "learning_rate": 1.8475335259811905e-05, "loss": 0.0085, "step": 73130 }, { "grad_norm": 0.13667088747024536, "learning_rate": 1.8462502838775508e-05, "loss": 0.0065, "step": 73140 }, { "grad_norm": 0.08456220477819443, "learning_rate": 1.8449673866632378e-05, "loss": 0.0113, "step": 73150 }, { "grad_norm": 0.13250857591629028, "learning_rate": 1.8436848344785473e-05, "loss": 0.009, "step": 73160 }, { "grad_norm": 0.20669926702976227, "learning_rate": 1.8424026274637378e-05, "loss": 0.0093, "step": 73170 }, { "grad_norm": 0.10436326265335083, "learning_rate": 1.8411207657590312e-05, "loss": 0.0082, "step": 73180 }, { "grad_norm": 0.15075653791427612, "learning_rate": 1.839839249504607e-05, "loss": 0.0098, "step": 73190 }, { "grad_norm": 0.10517194867134094, "learning_rate": 1.8385580788406108e-05, "loss": 0.0064, "step": 73200 }, { "grad_norm": 0.12595970928668976, "learning_rate": 1.837277253907152e-05, "loss": 0.0091, "step": 73210 }, { "grad_norm": 0.13437142968177795, "learning_rate": 1.8359967748442957e-05, "loss": 0.0088, "step": 73220 }, { "grad_norm": 0.14795708656311035, "learning_rate": 1.8347166417920774e-05, "loss": 0.009, "step": 73230 }, { "grad_norm": 0.13577885925769806, "learning_rate": 1.833436854890484e-05, "loss": 0.0072, "step": 73240 }, { "grad_norm": 0.11042319983243942, "learning_rate": 1.8321574142794796e-05, "loss": 0.0073, "step": 73250 }, { "grad_norm": 0.1354992836713791, "learning_rate": 1.8308783200989748e-05, "loss": 0.0065, "step": 73260 }, { "grad_norm": 0.10850631445646286, "learning_rate": 1.829599572488855e-05, "loss": 0.0073, "step": 73270 }, { "grad_norm": 0.23344473540782928, "learning_rate": 1.828321171588957e-05, "loss": 0.0067, "step": 73280 }, { "grad_norm": 0.13434885442256927, "learning_rate": 1.827043117539089e-05, "loss": 0.0089, "step": 73290 }, { "grad_norm": 0.10518018901348114, "learning_rate": 1.8257654104790116e-05, "loss": 0.0076, "step": 73300 }, { "grad_norm": 0.08811403810977936, "learning_rate": 1.82448805054846e-05, "loss": 0.0059, "step": 73310 }, { "grad_norm": 0.0886741653084755, "learning_rate": 1.823211037887119e-05, "loss": 0.0062, "step": 73320 }, { "grad_norm": 0.11765705049037933, "learning_rate": 1.821934372634645e-05, "loss": 0.0056, "step": 73330 }, { "grad_norm": 0.13277782499790192, "learning_rate": 1.8206580549306472e-05, "loss": 0.0073, "step": 73340 }, { "grad_norm": 0.15411390364170074, "learning_rate": 1.819382084914707e-05, "loss": 0.0078, "step": 73350 }, { "grad_norm": 0.1189877986907959, "learning_rate": 1.8181064627263577e-05, "loss": 0.0065, "step": 73360 }, { "grad_norm": 0.1408795416355133, "learning_rate": 1.8168311885051022e-05, "loss": 0.0078, "step": 73370 }, { "grad_norm": 0.15768621861934662, "learning_rate": 1.815556262390402e-05, "loss": 0.0091, "step": 73380 }, { "grad_norm": 0.08610675483942032, "learning_rate": 1.8142816845216826e-05, "loss": 0.0087, "step": 73390 }, { "grad_norm": 0.19990237057209015, "learning_rate": 1.8130074550383276e-05, "loss": 0.0092, "step": 73400 }, { "grad_norm": 0.13226136565208435, "learning_rate": 1.8117335740796848e-05, "loss": 0.006, "step": 73410 }, { "grad_norm": 0.09893333911895752, "learning_rate": 1.810460041785067e-05, "loss": 0.0072, "step": 73420 }, { "grad_norm": 0.09865275770425797, "learning_rate": 1.809186858293742e-05, "loss": 0.0069, "step": 73430 }, { "grad_norm": 0.10702857375144958, "learning_rate": 1.8079140237449448e-05, "loss": 0.0084, "step": 73440 }, { "grad_norm": 0.15690714120864868, "learning_rate": 1.8066415382778707e-05, "loss": 0.0081, "step": 73450 }, { "grad_norm": 0.16569125652313232, "learning_rate": 1.8053694020316786e-05, "loss": 0.0063, "step": 73460 }, { "grad_norm": 0.13207627832889557, "learning_rate": 1.8040976151454837e-05, "loss": 0.0064, "step": 73470 }, { "grad_norm": 0.15610964596271515, "learning_rate": 1.8028261777583704e-05, "loss": 0.0065, "step": 73480 }, { "grad_norm": 0.1664264053106308, "learning_rate": 1.8015550900093758e-05, "loss": 0.0095, "step": 73490 }, { "grad_norm": 0.09967204928398132, "learning_rate": 1.800284352037512e-05, "loss": 0.0089, "step": 73500 }, { "grad_norm": 0.11346296966075897, "learning_rate": 1.799013963981738e-05, "loss": 0.0073, "step": 73510 }, { "grad_norm": 0.1493850201368332, "learning_rate": 1.797743925980987e-05, "loss": 0.0105, "step": 73520 }, { "grad_norm": 0.1334487795829773, "learning_rate": 1.7964742381741435e-05, "loss": 0.0044, "step": 73530 }, { "grad_norm": 0.13273519277572632, "learning_rate": 1.7952049007000627e-05, "loss": 0.0073, "step": 73540 }, { "grad_norm": 0.08886891603469849, "learning_rate": 1.7939359136975535e-05, "loss": 0.0072, "step": 73550 }, { "grad_norm": 0.11967509984970093, "learning_rate": 1.7926672773053932e-05, "loss": 0.0074, "step": 73560 }, { "grad_norm": 0.09494094550609589, "learning_rate": 1.791398991662317e-05, "loss": 0.0075, "step": 73570 }, { "grad_norm": 0.12373790889978409, "learning_rate": 1.7901310569070244e-05, "loss": 0.0062, "step": 73580 }, { "grad_norm": 0.1521085798740387, "learning_rate": 1.7888634731781724e-05, "loss": 0.0076, "step": 73590 }, { "grad_norm": 0.17246262729167938, "learning_rate": 1.7875962406143844e-05, "loss": 0.0098, "step": 73600 }, { "grad_norm": 0.13430866599082947, "learning_rate": 1.7863293593542403e-05, "loss": 0.0069, "step": 73610 }, { "grad_norm": 0.1114688590168953, "learning_rate": 1.7850628295362853e-05, "loss": 0.0057, "step": 73620 }, { "grad_norm": 0.08099501579999924, "learning_rate": 1.7837966512990263e-05, "loss": 0.0073, "step": 73630 }, { "grad_norm": 0.12025243788957596, "learning_rate": 1.782530824780931e-05, "loss": 0.0079, "step": 73640 }, { "grad_norm": 0.15046681463718414, "learning_rate": 1.7812653501204263e-05, "loss": 0.0076, "step": 73650 }, { "grad_norm": 0.1769387573003769, "learning_rate": 1.780000227455903e-05, "loss": 0.0089, "step": 73660 }, { "grad_norm": 0.17532561719417572, "learning_rate": 1.778735456925716e-05, "loss": 0.0072, "step": 73670 }, { "grad_norm": 0.14158454537391663, "learning_rate": 1.7774710386681737e-05, "loss": 0.0063, "step": 73680 }, { "grad_norm": 0.10999783128499985, "learning_rate": 1.7762069728215545e-05, "loss": 0.0091, "step": 73690 }, { "grad_norm": 0.11850445717573166, "learning_rate": 1.7749432595240932e-05, "loss": 0.0068, "step": 73700 }, { "grad_norm": 0.14968536794185638, "learning_rate": 1.773679898913991e-05, "loss": 0.008, "step": 73710 }, { "grad_norm": 0.1479235291481018, "learning_rate": 1.7724168911294015e-05, "loss": 0.009, "step": 73720 }, { "grad_norm": 0.1140771433711052, "learning_rate": 1.7711542363084504e-05, "loss": 0.0068, "step": 73730 }, { "grad_norm": 0.10876944661140442, "learning_rate": 1.7698919345892158e-05, "loss": 0.0068, "step": 73740 }, { "grad_norm": 0.1844380646944046, "learning_rate": 1.768629986109743e-05, "loss": 0.0077, "step": 73750 }, { "grad_norm": 0.13387984037399292, "learning_rate": 1.7673683910080357e-05, "loss": 0.0096, "step": 73760 }, { "grad_norm": 0.12877210974693298, "learning_rate": 1.7661071494220638e-05, "loss": 0.0058, "step": 73770 }, { "grad_norm": 0.17563128471374512, "learning_rate": 1.7648462614897498e-05, "loss": 0.0064, "step": 73780 }, { "grad_norm": 0.11022286117076874, "learning_rate": 1.7635857273489863e-05, "loss": 0.0075, "step": 73790 }, { "grad_norm": 0.12149521708488464, "learning_rate": 1.7623255471376198e-05, "loss": 0.0074, "step": 73800 }, { "grad_norm": 0.1662353277206421, "learning_rate": 1.7610657209934646e-05, "loss": 0.0091, "step": 73810 }, { "grad_norm": 0.18613441288471222, "learning_rate": 1.7598062490542915e-05, "loss": 0.0103, "step": 73820 }, { "grad_norm": 0.18915344774723053, "learning_rate": 1.7585471314578377e-05, "loss": 0.0064, "step": 73830 }, { "grad_norm": 0.11197803914546967, "learning_rate": 1.7572883683417945e-05, "loss": 0.0068, "step": 73840 }, { "grad_norm": 0.1344156563282013, "learning_rate": 1.7560299598438212e-05, "loss": 0.0066, "step": 73850 }, { "grad_norm": 0.08204077184200287, "learning_rate": 1.7547719061015334e-05, "loss": 0.0065, "step": 73860 }, { "grad_norm": 0.12990526854991913, "learning_rate": 1.75351420725251e-05, "loss": 0.0068, "step": 73870 }, { "grad_norm": 0.15454480051994324, "learning_rate": 1.7522568634342923e-05, "loss": 0.0096, "step": 73880 }, { "grad_norm": 0.23250330984592438, "learning_rate": 1.7509998747843825e-05, "loss": 0.0103, "step": 73890 }, { "grad_norm": 0.1259104162454605, "learning_rate": 1.7497432414402403e-05, "loss": 0.0085, "step": 73900 }, { "grad_norm": 0.14294233918190002, "learning_rate": 1.7484869635392903e-05, "loss": 0.0117, "step": 73910 }, { "grad_norm": 0.0767148956656456, "learning_rate": 1.747231041218919e-05, "loss": 0.0105, "step": 73920 }, { "grad_norm": 0.11707863956689835, "learning_rate": 1.7459754746164692e-05, "loss": 0.0078, "step": 73930 }, { "grad_norm": 0.0893203392624855, "learning_rate": 1.744720263869249e-05, "loss": 0.0083, "step": 73940 }, { "grad_norm": 0.17052432894706726, "learning_rate": 1.7434654091145263e-05, "loss": 0.0081, "step": 73950 }, { "grad_norm": 0.11548265069723129, "learning_rate": 1.742210910489532e-05, "loss": 0.0068, "step": 73960 }, { "grad_norm": 0.16596265137195587, "learning_rate": 1.7409567681314532e-05, "loss": 0.0061, "step": 73970 }, { "grad_norm": 0.12668444216251373, "learning_rate": 1.7397029821774434e-05, "loss": 0.0078, "step": 73980 }, { "grad_norm": 0.09392665326595306, "learning_rate": 1.7384495527646126e-05, "loss": 0.0114, "step": 73990 }, { "grad_norm": 0.11155933141708374, "learning_rate": 1.737196480030035e-05, "loss": 0.0059, "step": 74000 }, { "grad_norm": 0.1069071963429451, "learning_rate": 1.7359437641107447e-05, "loss": 0.0078, "step": 74010 }, { "grad_norm": 0.10666315257549286, "learning_rate": 1.7346914051437387e-05, "loss": 0.008, "step": 74020 }, { "grad_norm": 0.13307130336761475, "learning_rate": 1.733439403265969e-05, "loss": 0.0067, "step": 74030 }, { "grad_norm": 0.06117004156112671, "learning_rate": 1.732187758614357e-05, "loss": 0.0052, "step": 74040 }, { "grad_norm": 0.15169596672058105, "learning_rate": 1.730936471325777e-05, "loss": 0.0088, "step": 74050 }, { "grad_norm": 0.1178455576300621, "learning_rate": 1.7296855415370718e-05, "loss": 0.005, "step": 74060 }, { "grad_norm": 0.19065023958683014, "learning_rate": 1.728434969385035e-05, "loss": 0.0105, "step": 74070 }, { "grad_norm": 0.1501539796590805, "learning_rate": 1.7271847550064346e-05, "loss": 0.0067, "step": 74080 }, { "grad_norm": 0.13871093094348907, "learning_rate": 1.725934898537987e-05, "loss": 0.0094, "step": 74090 }, { "grad_norm": 0.09612007439136505, "learning_rate": 1.7246854001163783e-05, "loss": 0.0067, "step": 74100 }, { "grad_norm": 0.09006505459547043, "learning_rate": 1.7234362598782476e-05, "loss": 0.0076, "step": 74110 }, { "grad_norm": 0.14936524629592896, "learning_rate": 1.722187477960202e-05, "loss": 0.0092, "step": 74120 }, { "grad_norm": 0.08484780043363571, "learning_rate": 1.7209390544988057e-05, "loss": 0.0064, "step": 74130 }, { "grad_norm": 0.13092206418514252, "learning_rate": 1.7196909896305856e-05, "loss": 0.007, "step": 74140 }, { "grad_norm": 0.14730100333690643, "learning_rate": 1.718443283492026e-05, "loss": 0.0062, "step": 74150 }, { "grad_norm": 0.12003780901432037, "learning_rate": 1.7171959362195746e-05, "loss": 0.0081, "step": 74160 }, { "grad_norm": 0.15075397491455078, "learning_rate": 1.715948947949642e-05, "loss": 0.0073, "step": 74170 }, { "grad_norm": 0.1136869415640831, "learning_rate": 1.714702318818593e-05, "loss": 0.0095, "step": 74180 }, { "grad_norm": 0.10827882587909698, "learning_rate": 1.7134560489627592e-05, "loss": 0.0064, "step": 74190 }, { "grad_norm": 0.08672614395618439, "learning_rate": 1.712210138518431e-05, "loss": 0.0089, "step": 74200 }, { "grad_norm": 0.1245235726237297, "learning_rate": 1.71096458762186e-05, "loss": 0.0076, "step": 74210 }, { "grad_norm": 0.11686331033706665, "learning_rate": 1.7097193964092552e-05, "loss": 0.005, "step": 74220 }, { "grad_norm": 0.08542244136333466, "learning_rate": 1.708474565016792e-05, "loss": 0.0048, "step": 74230 }, { "grad_norm": 0.10809355229139328, "learning_rate": 1.7072300935806e-05, "loss": 0.007, "step": 74240 }, { "grad_norm": 0.10475395619869232, "learning_rate": 1.705985982236776e-05, "loss": 0.0075, "step": 74250 }, { "grad_norm": 0.1355685591697693, "learning_rate": 1.7047422311213697e-05, "loss": 0.0072, "step": 74260 }, { "grad_norm": 0.0969601646065712, "learning_rate": 1.703498840370401e-05, "loss": 0.0077, "step": 74270 }, { "grad_norm": 0.0702921599149704, "learning_rate": 1.702255810119842e-05, "loss": 0.0043, "step": 74280 }, { "grad_norm": 0.10735980421304703, "learning_rate": 1.7010131405056305e-05, "loss": 0.0062, "step": 74290 }, { "grad_norm": 0.12958809733390808, "learning_rate": 1.6997708316636613e-05, "loss": 0.0077, "step": 74300 }, { "grad_norm": 0.16885431110858917, "learning_rate": 1.6985288837297932e-05, "loss": 0.0053, "step": 74310 }, { "grad_norm": 0.17833007872104645, "learning_rate": 1.6972872968398394e-05, "loss": 0.0071, "step": 74320 }, { "grad_norm": 0.10683642327785492, "learning_rate": 1.6960460711295845e-05, "loss": 0.0082, "step": 74330 }, { "grad_norm": 0.10209938138723373, "learning_rate": 1.6948052067347623e-05, "loss": 0.0076, "step": 74340 }, { "grad_norm": 0.11312052607536316, "learning_rate": 1.693564703791075e-05, "loss": 0.0082, "step": 74350 }, { "grad_norm": 0.08163610845804214, "learning_rate": 1.6923245624341793e-05, "loss": 0.0062, "step": 74360 }, { "grad_norm": 0.1371280401945114, "learning_rate": 1.6910847827996962e-05, "loss": 0.0095, "step": 74370 }, { "grad_norm": 0.46673309803009033, "learning_rate": 1.689845365023206e-05, "loss": 0.011, "step": 74380 }, { "grad_norm": 0.08888686448335648, "learning_rate": 1.6886063092402505e-05, "loss": 0.0076, "step": 74390 }, { "grad_norm": 0.09544304758310318, "learning_rate": 1.6873676155863317e-05, "loss": 0.008, "step": 74400 }, { "grad_norm": 0.11759490519762039, "learning_rate": 1.6861292841969085e-05, "loss": 0.0076, "step": 74410 }, { "grad_norm": 0.13276702165603638, "learning_rate": 1.684891315207406e-05, "loss": 0.0096, "step": 74420 }, { "grad_norm": 0.16019625961780548, "learning_rate": 1.6836537087532034e-05, "loss": 0.008, "step": 74430 }, { "grad_norm": 0.10369027405977249, "learning_rate": 1.6824164649696465e-05, "loss": 0.0073, "step": 74440 }, { "grad_norm": 0.18201521039009094, "learning_rate": 1.681179583992034e-05, "loss": 0.0117, "step": 74450 }, { "grad_norm": 0.11802859604358673, "learning_rate": 1.679943065955636e-05, "loss": 0.0061, "step": 74460 }, { "grad_norm": 0.14952945709228516, "learning_rate": 1.678706910995671e-05, "loss": 0.0069, "step": 74470 }, { "grad_norm": 0.10474443435668945, "learning_rate": 1.6774711192473262e-05, "loss": 0.0055, "step": 74480 }, { "grad_norm": 0.16125687956809998, "learning_rate": 1.676235690845743e-05, "loss": 0.0055, "step": 74490 }, { "grad_norm": 0.064154714345932, "learning_rate": 1.6750006259260285e-05, "loss": 0.0072, "step": 74500 }, { "grad_norm": 0.13061730563640594, "learning_rate": 1.673765924623244e-05, "loss": 0.0065, "step": 74510 }, { "grad_norm": 0.11208469420671463, "learning_rate": 1.67253158707242e-05, "loss": 0.007, "step": 74520 }, { "grad_norm": 0.1354355663061142, "learning_rate": 1.671297613408537e-05, "loss": 0.0059, "step": 74530 }, { "grad_norm": 0.1259709596633911, "learning_rate": 1.6700640037665444e-05, "loss": 0.0071, "step": 74540 }, { "grad_norm": 0.12973105907440186, "learning_rate": 1.668830758281344e-05, "loss": 0.0077, "step": 74550 }, { "grad_norm": 0.1666598618030548, "learning_rate": 1.6675978770878052e-05, "loss": 0.0096, "step": 74560 }, { "grad_norm": 0.09533058851957321, "learning_rate": 1.6663653603207506e-05, "loss": 0.0059, "step": 74570 }, { "grad_norm": 0.12246839702129364, "learning_rate": 1.6651332081149685e-05, "loss": 0.0059, "step": 74580 }, { "grad_norm": 0.10090389847755432, "learning_rate": 1.6639014206052046e-05, "loss": 0.0055, "step": 74590 }, { "grad_norm": 0.11790087819099426, "learning_rate": 1.6626699979261657e-05, "loss": 0.0102, "step": 74600 }, { "grad_norm": 0.10339166969060898, "learning_rate": 1.6614389402125197e-05, "loss": 0.0059, "step": 74610 }, { "grad_norm": 0.511155903339386, "learning_rate": 1.66020824759889e-05, "loss": 0.0067, "step": 74620 }, { "grad_norm": 0.10254920274019241, "learning_rate": 1.6589779202198675e-05, "loss": 0.0065, "step": 74630 }, { "grad_norm": 0.08512672781944275, "learning_rate": 1.6577479582099924e-05, "loss": 0.0124, "step": 74640 }, { "grad_norm": 0.18971899151802063, "learning_rate": 1.6565183617037788e-05, "loss": 0.0093, "step": 74650 }, { "grad_norm": 0.16856937110424042, "learning_rate": 1.6552891308356894e-05, "loss": 0.0068, "step": 74660 }, { "grad_norm": 0.11191107332706451, "learning_rate": 1.6540602657401526e-05, "loss": 0.0095, "step": 74670 }, { "grad_norm": 0.11868835240602493, "learning_rate": 1.6528317665515537e-05, "loss": 0.0069, "step": 74680 }, { "grad_norm": 0.11158763617277145, "learning_rate": 1.6516036334042416e-05, "loss": 0.0072, "step": 74690 }, { "grad_norm": 0.09836241602897644, "learning_rate": 1.650375866432519e-05, "loss": 0.0073, "step": 74700 }, { "grad_norm": 0.19157591462135315, "learning_rate": 1.649148465770659e-05, "loss": 0.0076, "step": 74710 }, { "grad_norm": 0.09288832545280457, "learning_rate": 1.647921431552883e-05, "loss": 0.0086, "step": 74720 }, { "grad_norm": 0.1564878672361374, "learning_rate": 1.646694763913381e-05, "loss": 0.0067, "step": 74730 }, { "grad_norm": 0.15026745200157166, "learning_rate": 1.6454684629862966e-05, "loss": 0.0057, "step": 74740 }, { "grad_norm": 0.12618137896060944, "learning_rate": 1.6442425289057388e-05, "loss": 0.0062, "step": 74750 }, { "grad_norm": 0.1320549100637436, "learning_rate": 1.6430169618057718e-05, "loss": 0.0068, "step": 74760 }, { "grad_norm": 0.16280536353588104, "learning_rate": 1.641791761820422e-05, "loss": 0.0097, "step": 74770 }, { "grad_norm": 0.09232906252145767, "learning_rate": 1.640566929083676e-05, "loss": 0.0075, "step": 74780 }, { "grad_norm": 0.08500295132398605, "learning_rate": 1.6393424637294823e-05, "loss": 0.0068, "step": 74790 }, { "grad_norm": 0.09158360958099365, "learning_rate": 1.638118365891742e-05, "loss": 0.0089, "step": 74800 }, { "grad_norm": 0.11857333034276962, "learning_rate": 1.636894635704323e-05, "loss": 0.0089, "step": 74810 }, { "grad_norm": 0.11236618459224701, "learning_rate": 1.6356712733010517e-05, "loss": 0.0093, "step": 74820 }, { "grad_norm": 0.07915783673524857, "learning_rate": 1.6344482788157096e-05, "loss": 0.0065, "step": 74830 }, { "grad_norm": 0.10972213745117188, "learning_rate": 1.6332256523820444e-05, "loss": 0.007, "step": 74840 }, { "grad_norm": 0.10979589074850082, "learning_rate": 1.6320033941337597e-05, "loss": 0.0059, "step": 74850 }, { "grad_norm": 0.20426033437252045, "learning_rate": 1.6307815042045216e-05, "loss": 0.0059, "step": 74860 }, { "grad_norm": 0.11931558698415756, "learning_rate": 1.6295599827279513e-05, "loss": 0.0109, "step": 74870 }, { "grad_norm": 0.14046959578990936, "learning_rate": 1.6283388298376356e-05, "loss": 0.0054, "step": 74880 }, { "grad_norm": 0.23655934631824493, "learning_rate": 1.6271180456671125e-05, "loss": 0.0071, "step": 74890 }, { "grad_norm": 0.11188941448926926, "learning_rate": 1.625897630349893e-05, "loss": 0.0068, "step": 74900 }, { "grad_norm": 0.24355773627758026, "learning_rate": 1.6246775840194335e-05, "loss": 0.0113, "step": 74910 }, { "grad_norm": 0.09932323545217514, "learning_rate": 1.6234579068091604e-05, "loss": 0.0069, "step": 74920 }, { "grad_norm": 0.09306015074253082, "learning_rate": 1.6222385988524524e-05, "loss": 0.0058, "step": 74930 }, { "grad_norm": 0.1278708428144455, "learning_rate": 1.6210196602826544e-05, "loss": 0.0107, "step": 74940 }, { "grad_norm": 0.11683652549982071, "learning_rate": 1.6198010912330646e-05, "loss": 0.007, "step": 74950 }, { "grad_norm": 0.18448030948638916, "learning_rate": 1.6185828918369456e-05, "loss": 0.0067, "step": 74960 }, { "grad_norm": 0.10069161653518677, "learning_rate": 1.617365062227517e-05, "loss": 0.006, "step": 74970 }, { "grad_norm": 0.0943572148680687, "learning_rate": 1.616147602537962e-05, "loss": 0.0072, "step": 74980 }, { "grad_norm": 0.2104211300611496, "learning_rate": 1.614930512901415e-05, "loss": 0.0056, "step": 74990 }, { "grad_norm": 0.11733515560626984, "learning_rate": 1.6137137934509804e-05, "loss": 0.0089, "step": 75000 }, { "grad_norm": 0.1717144101858139, "learning_rate": 1.6124974443197123e-05, "loss": 0.007, "step": 75010 }, { "grad_norm": 0.11922916769981384, "learning_rate": 1.6112814656406306e-05, "loss": 0.0066, "step": 75020 }, { "grad_norm": 0.15156161785125732, "learning_rate": 1.6100658575467125e-05, "loss": 0.0125, "step": 75030 }, { "grad_norm": 0.09132544696331024, "learning_rate": 1.6088506201708974e-05, "loss": 0.006, "step": 75040 }, { "grad_norm": 0.09768914431333542, "learning_rate": 1.6076357536460785e-05, "loss": 0.0057, "step": 75050 }, { "grad_norm": 0.15238332748413086, "learning_rate": 1.606421258105113e-05, "loss": 0.0088, "step": 75060 }, { "grad_norm": 0.17087039351463318, "learning_rate": 1.6052071336808183e-05, "loss": 0.0086, "step": 75070 }, { "grad_norm": 0.1113179624080658, "learning_rate": 1.6039933805059658e-05, "loss": 0.0066, "step": 75080 }, { "grad_norm": 0.16767272353172302, "learning_rate": 1.6027799987132915e-05, "loss": 0.0074, "step": 75090 }, { "grad_norm": 0.1531735062599182, "learning_rate": 1.6015669884354888e-05, "loss": 0.0063, "step": 75100 }, { "grad_norm": 0.09845110774040222, "learning_rate": 1.6003543498052125e-05, "loss": 0.0072, "step": 75110 }, { "grad_norm": 0.08379782736301422, "learning_rate": 1.5991420829550715e-05, "loss": 0.0072, "step": 75120 }, { "grad_norm": 0.13255047798156738, "learning_rate": 1.5979301880176405e-05, "loss": 0.0089, "step": 75130 }, { "grad_norm": 0.1113971471786499, "learning_rate": 1.5967186651254485e-05, "loss": 0.0081, "step": 75140 }, { "grad_norm": 0.11409719288349152, "learning_rate": 1.5955075144109853e-05, "loss": 0.0053, "step": 75150 }, { "grad_norm": 0.13570018112659454, "learning_rate": 1.594296736006703e-05, "loss": 0.0072, "step": 75160 }, { "grad_norm": 0.06956417858600616, "learning_rate": 1.59308633004501e-05, "loss": 0.0058, "step": 75170 }, { "grad_norm": 0.154883474111557, "learning_rate": 1.5918762966582724e-05, "loss": 0.0063, "step": 75180 }, { "grad_norm": 0.08835150301456451, "learning_rate": 1.5906666359788203e-05, "loss": 0.0052, "step": 75190 }, { "grad_norm": 0.10002852231264114, "learning_rate": 1.589457348138938e-05, "loss": 0.0059, "step": 75200 }, { "grad_norm": 0.0937049388885498, "learning_rate": 1.588248433270872e-05, "loss": 0.0055, "step": 75210 }, { "grad_norm": 0.11552523076534271, "learning_rate": 1.587039891506828e-05, "loss": 0.0071, "step": 75220 }, { "grad_norm": 0.10138910263776779, "learning_rate": 1.5858317229789715e-05, "loss": 0.0051, "step": 75230 }, { "grad_norm": 0.1255055069923401, "learning_rate": 1.584623927819423e-05, "loss": 0.0076, "step": 75240 }, { "grad_norm": 0.09649979323148727, "learning_rate": 1.583416506160269e-05, "loss": 0.0061, "step": 75250 }, { "grad_norm": 0.18813462555408478, "learning_rate": 1.5822094581335472e-05, "loss": 0.0066, "step": 75260 }, { "grad_norm": 0.16844964027404785, "learning_rate": 1.5810027838712604e-05, "loss": 0.0066, "step": 75270 }, { "grad_norm": 0.18041929602622986, "learning_rate": 1.5797964835053687e-05, "loss": 0.0065, "step": 75280 }, { "grad_norm": 0.1973792165517807, "learning_rate": 1.5785905571677933e-05, "loss": 0.0084, "step": 75290 }, { "grad_norm": 0.18518777191638947, "learning_rate": 1.5773850049904086e-05, "loss": 0.0074, "step": 75300 }, { "grad_norm": 0.16791567206382751, "learning_rate": 1.5761798271050543e-05, "loss": 0.0068, "step": 75310 }, { "grad_norm": 0.14168301224708557, "learning_rate": 1.5749750236435277e-05, "loss": 0.0081, "step": 75320 }, { "grad_norm": 0.17239345610141754, "learning_rate": 1.5737705947375815e-05, "loss": 0.009, "step": 75330 }, { "grad_norm": 0.11849561333656311, "learning_rate": 1.5725665405189322e-05, "loss": 0.0083, "step": 75340 }, { "grad_norm": 0.12393990904092789, "learning_rate": 1.571362861119253e-05, "loss": 0.0076, "step": 75350 }, { "grad_norm": 0.14990870654582977, "learning_rate": 1.570159556670177e-05, "loss": 0.0113, "step": 75360 }, { "grad_norm": 0.10606636852025986, "learning_rate": 1.5689566273032946e-05, "loss": 0.0076, "step": 75370 }, { "grad_norm": 0.125232994556427, "learning_rate": 1.5677540731501583e-05, "loss": 0.0084, "step": 75380 }, { "grad_norm": 0.15609173476696014, "learning_rate": 1.5665518943422746e-05, "loss": 0.0078, "step": 75390 }, { "grad_norm": 0.11173397302627563, "learning_rate": 1.5653500910111134e-05, "loss": 0.005, "step": 75400 }, { "grad_norm": 0.11804370582103729, "learning_rate": 1.5641486632881023e-05, "loss": 0.0072, "step": 75410 }, { "grad_norm": 0.12199630588293076, "learning_rate": 1.56294761130463e-05, "loss": 0.0068, "step": 75420 }, { "grad_norm": 0.09403600543737411, "learning_rate": 1.5617469351920373e-05, "loss": 0.0071, "step": 75430 }, { "grad_norm": 0.18726080656051636, "learning_rate": 1.5605466350816315e-05, "loss": 0.0071, "step": 75440 }, { "grad_norm": 0.13397005200386047, "learning_rate": 1.559346711104674e-05, "loss": 0.0065, "step": 75450 }, { "grad_norm": 0.08845315128564835, "learning_rate": 1.5581471633923883e-05, "loss": 0.0057, "step": 75460 }, { "grad_norm": 0.11976104229688644, "learning_rate": 1.5569479920759505e-05, "loss": 0.0063, "step": 75470 }, { "grad_norm": 0.07480484992265701, "learning_rate": 1.5557491972865072e-05, "loss": 0.0071, "step": 75480 }, { "grad_norm": 0.141788512468338, "learning_rate": 1.554550779155152e-05, "loss": 0.009, "step": 75490 }, { "grad_norm": 0.10286196321249008, "learning_rate": 1.5533527378129453e-05, "loss": 0.0058, "step": 75500 }, { "grad_norm": 0.2508520185947418, "learning_rate": 1.552155073390899e-05, "loss": 0.0068, "step": 75510 }, { "grad_norm": 0.10063596814870834, "learning_rate": 1.5509577860199903e-05, "loss": 0.0066, "step": 75520 }, { "grad_norm": 0.11981602758169174, "learning_rate": 1.5497608758311525e-05, "loss": 0.0065, "step": 75530 }, { "grad_norm": 0.11308488249778748, "learning_rate": 1.5485643429552797e-05, "loss": 0.0064, "step": 75540 }, { "grad_norm": 0.10089502483606339, "learning_rate": 1.5473681875232197e-05, "loss": 0.0065, "step": 75550 }, { "grad_norm": 0.13807564973831177, "learning_rate": 1.5461724096657837e-05, "loss": 0.0081, "step": 75560 }, { "grad_norm": 0.1422889083623886, "learning_rate": 1.5449770095137416e-05, "loss": 0.0066, "step": 75570 }, { "grad_norm": 0.14498835802078247, "learning_rate": 1.5437819871978178e-05, "loss": 0.0058, "step": 75580 }, { "grad_norm": 0.11220645904541016, "learning_rate": 1.5425873428486992e-05, "loss": 0.0059, "step": 75590 }, { "grad_norm": 0.16317445039749146, "learning_rate": 1.5413930765970312e-05, "loss": 0.0068, "step": 75600 }, { "grad_norm": 0.12516342103481293, "learning_rate": 1.5401991885734175e-05, "loss": 0.0092, "step": 75610 }, { "grad_norm": 0.0957755446434021, "learning_rate": 1.5390056789084174e-05, "loss": 0.0067, "step": 75620 }, { "grad_norm": 0.13148367404937744, "learning_rate": 1.5378125477325545e-05, "loss": 0.0068, "step": 75630 }, { "grad_norm": 0.09456807374954224, "learning_rate": 1.5366197951763046e-05, "loss": 0.0085, "step": 75640 }, { "grad_norm": 0.09358867257833481, "learning_rate": 1.5354274213701075e-05, "loss": 0.0054, "step": 75650 }, { "grad_norm": 0.1461537927389145, "learning_rate": 1.5342354264443566e-05, "loss": 0.0058, "step": 75660 }, { "grad_norm": 0.11692751199007034, "learning_rate": 1.533043810529411e-05, "loss": 0.0055, "step": 75670 }, { "grad_norm": 0.09646911174058914, "learning_rate": 1.5318525737555807e-05, "loss": 0.0096, "step": 75680 }, { "grad_norm": 0.2009841948747635, "learning_rate": 1.53066171625314e-05, "loss": 0.0085, "step": 75690 }, { "grad_norm": 0.08920308202505112, "learning_rate": 1.529471238152317e-05, "loss": 0.0082, "step": 75700 }, { "grad_norm": 0.1146097183227539, "learning_rate": 1.5282811395833023e-05, "loss": 0.007, "step": 75710 }, { "grad_norm": 0.10249140113592148, "learning_rate": 1.5270914206762397e-05, "loss": 0.0053, "step": 75720 }, { "grad_norm": 0.12570184469223022, "learning_rate": 1.525902081561241e-05, "loss": 0.0056, "step": 75730 }, { "grad_norm": 0.10841191560029984, "learning_rate": 1.5247131223683664e-05, "loss": 0.0054, "step": 75740 }, { "grad_norm": 0.12136969715356827, "learning_rate": 1.5235245432276408e-05, "loss": 0.0064, "step": 75750 }, { "grad_norm": 0.09754324704408646, "learning_rate": 1.5223363442690424e-05, "loss": 0.0072, "step": 75760 }, { "grad_norm": 0.13423483073711395, "learning_rate": 1.5211485256225133e-05, "loss": 0.0067, "step": 75770 }, { "grad_norm": 0.14890331029891968, "learning_rate": 1.519961087417951e-05, "loss": 0.0059, "step": 75780 }, { "grad_norm": 0.08771727979183197, "learning_rate": 1.5187740297852137e-05, "loss": 0.005, "step": 75790 }, { "grad_norm": 0.07689293473958969, "learning_rate": 1.517587352854113e-05, "loss": 0.0058, "step": 75800 }, { "grad_norm": 0.14611627161502838, "learning_rate": 1.516401056754423e-05, "loss": 0.0063, "step": 75810 }, { "grad_norm": 0.11302284896373749, "learning_rate": 1.5152151416158783e-05, "loss": 0.0045, "step": 75820 }, { "grad_norm": 0.08015333116054535, "learning_rate": 1.5140296075681642e-05, "loss": 0.0054, "step": 75830 }, { "grad_norm": 0.1378035992383957, "learning_rate": 1.5128444547409331e-05, "loss": 0.0063, "step": 75840 }, { "grad_norm": 0.14016443490982056, "learning_rate": 1.5116596832637864e-05, "loss": 0.008, "step": 75850 }, { "grad_norm": 0.10670270025730133, "learning_rate": 1.5104752932662952e-05, "loss": 0.0073, "step": 75860 }, { "grad_norm": 0.14131160080432892, "learning_rate": 1.5092912848779778e-05, "loss": 0.0054, "step": 75870 }, { "grad_norm": 0.11557232588529587, "learning_rate": 1.5081076582283194e-05, "loss": 0.0066, "step": 75880 }, { "grad_norm": 0.2003435492515564, "learning_rate": 1.5069244134467553e-05, "loss": 0.0062, "step": 75890 }, { "grad_norm": 0.12002480775117874, "learning_rate": 1.5057415506626876e-05, "loss": 0.0077, "step": 75900 }, { "grad_norm": 0.09236105531454086, "learning_rate": 1.5045590700054668e-05, "loss": 0.0088, "step": 75910 }, { "grad_norm": 0.07840932905673981, "learning_rate": 1.5033769716044144e-05, "loss": 0.0055, "step": 75920 }, { "grad_norm": 0.12865076959133148, "learning_rate": 1.5021952555887975e-05, "loss": 0.0071, "step": 75930 }, { "grad_norm": 0.0868668258190155, "learning_rate": 1.50101392208785e-05, "loss": 0.0074, "step": 75940 }, { "grad_norm": 0.1294633001089096, "learning_rate": 1.4998329712307579e-05, "loss": 0.0069, "step": 75950 }, { "grad_norm": 0.18573680520057678, "learning_rate": 1.4986524031466708e-05, "loss": 0.009, "step": 75960 }, { "grad_norm": 0.10331106185913086, "learning_rate": 1.4974722179646916e-05, "loss": 0.0059, "step": 75970 }, { "grad_norm": 0.08560045063495636, "learning_rate": 1.4962924158138847e-05, "loss": 0.0075, "step": 75980 }, { "grad_norm": 0.1655140519142151, "learning_rate": 1.4951129968232714e-05, "loss": 0.0098, "step": 75990 }, { "grad_norm": 0.1560094952583313, "learning_rate": 1.4939339611218329e-05, "loss": 0.0104, "step": 76000 }, { "grad_norm": 0.11331095546483994, "learning_rate": 1.4927553088385032e-05, "loss": 0.0067, "step": 76010 }, { "grad_norm": 0.11633000522851944, "learning_rate": 1.4915770401021801e-05, "loss": 0.007, "step": 76020 }, { "grad_norm": 0.1209258884191513, "learning_rate": 1.4903991550417185e-05, "loss": 0.0055, "step": 76030 }, { "grad_norm": 0.1658443808555603, "learning_rate": 1.4892216537859277e-05, "loss": 0.0081, "step": 76040 }, { "grad_norm": 0.1415838748216629, "learning_rate": 1.4880445364635775e-05, "loss": 0.0073, "step": 76050 }, { "grad_norm": 0.12753233313560486, "learning_rate": 1.4868678032033973e-05, "loss": 0.0055, "step": 76060 }, { "grad_norm": 0.17414934933185577, "learning_rate": 1.4856914541340739e-05, "loss": 0.0062, "step": 76070 }, { "grad_norm": 0.10544916242361069, "learning_rate": 1.4845154893842472e-05, "loss": 0.0071, "step": 76080 }, { "grad_norm": 0.12326522916555405, "learning_rate": 1.4833399090825234e-05, "loss": 0.0053, "step": 76090 }, { "grad_norm": 0.15170393884181976, "learning_rate": 1.4821647133574562e-05, "loss": 0.0078, "step": 76100 }, { "grad_norm": 0.12670397758483887, "learning_rate": 1.4809899023375706e-05, "loss": 0.0065, "step": 76110 }, { "grad_norm": 0.15158718824386597, "learning_rate": 1.4798154761513366e-05, "loss": 0.0069, "step": 76120 }, { "grad_norm": 0.12154767662286758, "learning_rate": 1.4786414349271916e-05, "loss": 0.0053, "step": 76130 }, { "grad_norm": 0.11261295527219772, "learning_rate": 1.4774677787935226e-05, "loss": 0.0056, "step": 76140 }, { "grad_norm": 0.0915670171380043, "learning_rate": 1.4762945078786838e-05, "loss": 0.0066, "step": 76150 }, { "grad_norm": 0.13721445202827454, "learning_rate": 1.4751216223109777e-05, "loss": 0.0069, "step": 76160 }, { "grad_norm": 0.17881712317466736, "learning_rate": 1.4739491222186714e-05, "loss": 0.0088, "step": 76170 }, { "grad_norm": 0.10533852875232697, "learning_rate": 1.4727770077299879e-05, "loss": 0.0077, "step": 76180 }, { "grad_norm": 0.16110116243362427, "learning_rate": 1.4716052789731094e-05, "loss": 0.0103, "step": 76190 }, { "grad_norm": 0.2007216215133667, "learning_rate": 1.470433936076171e-05, "loss": 0.0074, "step": 76200 }, { "grad_norm": 0.14602799713611603, "learning_rate": 1.4692629791672719e-05, "loss": 0.0081, "step": 76210 }, { "grad_norm": 0.12633925676345825, "learning_rate": 1.468092408374464e-05, "loss": 0.0061, "step": 76220 }, { "grad_norm": 0.16585209965705872, "learning_rate": 1.4669222238257602e-05, "loss": 0.0069, "step": 76230 }, { "grad_norm": 0.12058930099010468, "learning_rate": 1.4657524256491301e-05, "loss": 0.008, "step": 76240 }, { "grad_norm": 0.13650664687156677, "learning_rate": 1.464583013972502e-05, "loss": 0.008, "step": 76250 }, { "grad_norm": 0.1185469776391983, "learning_rate": 1.4634139889237591e-05, "loss": 0.0085, "step": 76260 }, { "grad_norm": 0.0874568447470665, "learning_rate": 1.4622453506307448e-05, "loss": 0.0061, "step": 76270 }, { "grad_norm": 0.1053338274359703, "learning_rate": 1.461077099221262e-05, "loss": 0.0054, "step": 76280 }, { "grad_norm": 0.24641555547714233, "learning_rate": 1.4599092348230652e-05, "loss": 0.0062, "step": 76290 }, { "grad_norm": 0.1588539183139801, "learning_rate": 1.4587417575638724e-05, "loss": 0.007, "step": 76300 }, { "grad_norm": 0.14301860332489014, "learning_rate": 1.4575746675713563e-05, "loss": 0.0065, "step": 76310 }, { "grad_norm": 0.10508401691913605, "learning_rate": 1.4564079649731504e-05, "loss": 0.0072, "step": 76320 }, { "grad_norm": 0.15085619688034058, "learning_rate": 1.4552416498968402e-05, "loss": 0.0078, "step": 76330 }, { "grad_norm": 0.12788322567939758, "learning_rate": 1.4540757224699747e-05, "loss": 0.0097, "step": 76340 }, { "grad_norm": 0.13906893134117126, "learning_rate": 1.452910182820056e-05, "loss": 0.0055, "step": 76350 }, { "grad_norm": 0.13295748829841614, "learning_rate": 1.4517450310745468e-05, "loss": 0.0056, "step": 76360 }, { "grad_norm": 0.10819552838802338, "learning_rate": 1.450580267360866e-05, "loss": 0.0064, "step": 76370 }, { "grad_norm": 0.14410161972045898, "learning_rate": 1.4494158918063922e-05, "loss": 0.0075, "step": 76380 }, { "grad_norm": 0.09447191655635834, "learning_rate": 1.4482519045384568e-05, "loss": 0.0058, "step": 76390 }, { "grad_norm": 0.10565980523824692, "learning_rate": 1.4470883056843548e-05, "loss": 0.0065, "step": 76400 }, { "grad_norm": 0.10874713957309723, "learning_rate": 1.4459250953713322e-05, "loss": 0.0064, "step": 76410 }, { "grad_norm": 0.07917218655347824, "learning_rate": 1.444762273726598e-05, "loss": 0.0074, "step": 76420 }, { "grad_norm": 0.1096063181757927, "learning_rate": 1.443599840877316e-05, "loss": 0.0056, "step": 76430 }, { "grad_norm": 0.08802919089794159, "learning_rate": 1.4424377969506103e-05, "loss": 0.0088, "step": 76440 }, { "grad_norm": 0.1252574771642685, "learning_rate": 1.4412761420735565e-05, "loss": 0.0099, "step": 76450 }, { "grad_norm": 0.14202550053596497, "learning_rate": 1.4401148763731953e-05, "loss": 0.0065, "step": 76460 }, { "grad_norm": 0.1648140549659729, "learning_rate": 1.4389539999765173e-05, "loss": 0.0074, "step": 76470 }, { "grad_norm": 0.10037726163864136, "learning_rate": 1.4377935130104758e-05, "loss": 0.0067, "step": 76480 }, { "grad_norm": 0.0896868035197258, "learning_rate": 1.4366334156019801e-05, "loss": 0.0078, "step": 76490 }, { "grad_norm": 0.13010182976722717, "learning_rate": 1.435473707877898e-05, "loss": 0.0062, "step": 76500 }, { "grad_norm": 0.20313364267349243, "learning_rate": 1.4343143899650508e-05, "loss": 0.0081, "step": 76510 }, { "grad_norm": 0.13037115335464478, "learning_rate": 1.4331554619902204e-05, "loss": 0.0065, "step": 76520 }, { "grad_norm": 0.13716137409210205, "learning_rate": 1.4319969240801484e-05, "loss": 0.0059, "step": 76530 }, { "grad_norm": 0.10247395932674408, "learning_rate": 1.430838776361526e-05, "loss": 0.0065, "step": 76540 }, { "grad_norm": 0.2088966816663742, "learning_rate": 1.4296810189610093e-05, "loss": 0.0063, "step": 76550 }, { "grad_norm": 0.14034762978553772, "learning_rate": 1.4285236520052087e-05, "loss": 0.0058, "step": 76560 }, { "grad_norm": 0.14422672986984253, "learning_rate": 1.4273666756206933e-05, "loss": 0.007, "step": 76570 }, { "grad_norm": 0.1949690729379654, "learning_rate": 1.4262100899339848e-05, "loss": 0.0077, "step": 76580 }, { "grad_norm": 0.15737631916999817, "learning_rate": 1.4250538950715696e-05, "loss": 0.0059, "step": 76590 }, { "grad_norm": 0.10078702121973038, "learning_rate": 1.4238980911598842e-05, "loss": 0.0074, "step": 76600 }, { "grad_norm": 0.11915234476327896, "learning_rate": 1.4227426783253267e-05, "loss": 0.0055, "step": 76610 }, { "grad_norm": 0.13517926633358002, "learning_rate": 1.4215876566942521e-05, "loss": 0.008, "step": 76620 }, { "grad_norm": 0.10319963842630386, "learning_rate": 1.4204330263929721e-05, "loss": 0.007, "step": 76630 }, { "grad_norm": 0.12015853822231293, "learning_rate": 1.4192787875477537e-05, "loss": 0.0054, "step": 76640 }, { "grad_norm": 0.10556609183549881, "learning_rate": 1.4181249402848246e-05, "loss": 0.0066, "step": 76650 }, { "grad_norm": 0.1552116423845291, "learning_rate": 1.4169714847303656e-05, "loss": 0.008, "step": 76660 }, { "grad_norm": 0.250754177570343, "learning_rate": 1.4158184210105197e-05, "loss": 0.0135, "step": 76670 }, { "grad_norm": 0.13014793395996094, "learning_rate": 1.4146657492513787e-05, "loss": 0.0068, "step": 76680 }, { "grad_norm": 0.09197919070720673, "learning_rate": 1.4135134695790047e-05, "loss": 0.0068, "step": 76690 }, { "grad_norm": 0.08258669823408127, "learning_rate": 1.4123615821194037e-05, "loss": 0.0063, "step": 76700 }, { "grad_norm": 0.10971236974000931, "learning_rate": 1.4112100869985473e-05, "loss": 0.0071, "step": 76710 }, { "grad_norm": 0.17215216159820557, "learning_rate": 1.4100589843423589e-05, "loss": 0.0061, "step": 76720 }, { "grad_norm": 0.0857354924082756, "learning_rate": 1.408908274276723e-05, "loss": 0.0065, "step": 76730 }, { "grad_norm": 0.21135178208351135, "learning_rate": 1.4077579569274786e-05, "loss": 0.0086, "step": 76740 }, { "grad_norm": 0.1674167364835739, "learning_rate": 1.4066080324204245e-05, "loss": 0.0065, "step": 76750 }, { "grad_norm": 0.17803378403186798, "learning_rate": 1.405458500881312e-05, "loss": 0.0067, "step": 76760 }, { "grad_norm": 0.12837672233581543, "learning_rate": 1.4043093624358538e-05, "loss": 0.005, "step": 76770 }, { "grad_norm": 0.19337962567806244, "learning_rate": 1.4031606172097189e-05, "loss": 0.0068, "step": 76780 }, { "grad_norm": 0.09396139532327652, "learning_rate": 1.40201226532853e-05, "loss": 0.0061, "step": 76790 }, { "grad_norm": 0.11839121580123901, "learning_rate": 1.4008643069178701e-05, "loss": 0.0057, "step": 76800 }, { "grad_norm": 0.25293204188346863, "learning_rate": 1.399716742103278e-05, "loss": 0.0062, "step": 76810 }, { "grad_norm": 0.1705315113067627, "learning_rate": 1.398569571010252e-05, "loss": 0.0096, "step": 76820 }, { "grad_norm": 0.1123824417591095, "learning_rate": 1.3974227937642414e-05, "loss": 0.0093, "step": 76830 }, { "grad_norm": 0.16964584589004517, "learning_rate": 1.3962764104906595e-05, "loss": 0.0061, "step": 76840 }, { "grad_norm": 0.14910171926021576, "learning_rate": 1.3951304213148696e-05, "loss": 0.0072, "step": 76850 }, { "grad_norm": 0.14705996215343475, "learning_rate": 1.3939848263621979e-05, "loss": 0.0062, "step": 76860 }, { "grad_norm": 0.10767832398414612, "learning_rate": 1.3928396257579219e-05, "loss": 0.006, "step": 76870 }, { "grad_norm": 0.18153485655784607, "learning_rate": 1.3916948196272833e-05, "loss": 0.0066, "step": 76880 }, { "grad_norm": 0.0891849547624588, "learning_rate": 1.3905504080954734e-05, "loss": 0.006, "step": 76890 }, { "grad_norm": 0.15050271153450012, "learning_rate": 1.3894063912876454e-05, "loss": 0.0064, "step": 76900 }, { "grad_norm": 0.11667495220899582, "learning_rate": 1.3882627693289046e-05, "loss": 0.0077, "step": 76910 }, { "grad_norm": 0.14236027002334595, "learning_rate": 1.3871195423443183e-05, "loss": 0.0084, "step": 76920 }, { "grad_norm": 0.1401020586490631, "learning_rate": 1.3859767104589044e-05, "loss": 0.0074, "step": 76930 }, { "grad_norm": 0.10280446708202362, "learning_rate": 1.3848342737976467e-05, "loss": 0.008, "step": 76940 }, { "grad_norm": 0.11540110409259796, "learning_rate": 1.3836922324854757e-05, "loss": 0.0076, "step": 76950 }, { "grad_norm": 0.09645754098892212, "learning_rate": 1.382550586647287e-05, "loss": 0.0066, "step": 76960 }, { "grad_norm": 0.09946906566619873, "learning_rate": 1.3814093364079257e-05, "loss": 0.0074, "step": 76970 }, { "grad_norm": 0.1348849982023239, "learning_rate": 1.3802684818921991e-05, "loss": 0.0082, "step": 76980 }, { "grad_norm": 0.1764698475599289, "learning_rate": 1.3791280232248687e-05, "loss": 0.0117, "step": 76990 }, { "grad_norm": 0.17626312375068665, "learning_rate": 1.3779879605306555e-05, "loss": 0.0049, "step": 77000 }, { "grad_norm": 0.17318660020828247, "learning_rate": 1.376848293934232e-05, "loss": 0.008, "step": 77010 }, { "grad_norm": 0.15572930872440338, "learning_rate": 1.3757090235602316e-05, "loss": 0.0075, "step": 77020 }, { "grad_norm": 0.09446895867586136, "learning_rate": 1.3745701495332447e-05, "loss": 0.0055, "step": 77030 }, { "grad_norm": 0.08615528792142868, "learning_rate": 1.3734316719778145e-05, "loss": 0.0062, "step": 77040 }, { "grad_norm": 0.13150480389595032, "learning_rate": 1.3722935910184454e-05, "loss": 0.0066, "step": 77050 }, { "grad_norm": 0.08228576928377151, "learning_rate": 1.3711559067795921e-05, "loss": 0.0073, "step": 77060 }, { "grad_norm": 0.12397850304841995, "learning_rate": 1.3700186193856756e-05, "loss": 0.0083, "step": 77070 }, { "grad_norm": 0.12595979869365692, "learning_rate": 1.368881728961064e-05, "loss": 0.0063, "step": 77080 }, { "grad_norm": 0.09865786135196686, "learning_rate": 1.3677452356300885e-05, "loss": 0.0112, "step": 77090 }, { "grad_norm": 0.17104178667068481, "learning_rate": 1.3666091395170321e-05, "loss": 0.0063, "step": 77100 }, { "grad_norm": 0.1649198830127716, "learning_rate": 1.3654734407461384e-05, "loss": 0.0064, "step": 77110 }, { "grad_norm": 0.09180565923452377, "learning_rate": 1.3643381394416022e-05, "loss": 0.0058, "step": 77120 }, { "grad_norm": 0.107086680829525, "learning_rate": 1.3632032357275838e-05, "loss": 0.0056, "step": 77130 }, { "grad_norm": 0.1364041566848755, "learning_rate": 1.3620687297281898e-05, "loss": 0.0067, "step": 77140 }, { "grad_norm": 0.17594729363918304, "learning_rate": 1.3609346215674917e-05, "loss": 0.0091, "step": 77150 }, { "grad_norm": 0.1252983659505844, "learning_rate": 1.3598009113695109e-05, "loss": 0.0061, "step": 77160 }, { "grad_norm": 0.11339402198791504, "learning_rate": 1.358667599258231e-05, "loss": 0.0069, "step": 77170 }, { "grad_norm": 0.11267364770174026, "learning_rate": 1.357534685357586e-05, "loss": 0.0096, "step": 77180 }, { "grad_norm": 0.0923636183142662, "learning_rate": 1.3564021697914715e-05, "loss": 0.0072, "step": 77190 }, { "grad_norm": 0.16150259971618652, "learning_rate": 1.3552700526837381e-05, "loss": 0.0083, "step": 77200 }, { "grad_norm": 0.16104207932949066, "learning_rate": 1.3541383341581925e-05, "loss": 0.0081, "step": 77210 }, { "grad_norm": 0.09988360106945038, "learning_rate": 1.3530070143385965e-05, "loss": 0.0068, "step": 77220 }, { "grad_norm": 0.10584044456481934, "learning_rate": 1.3518760933486702e-05, "loss": 0.009, "step": 77230 }, { "grad_norm": 0.09405917674303055, "learning_rate": 1.3507455713120908e-05, "loss": 0.0073, "step": 77240 }, { "grad_norm": 0.13197296857833862, "learning_rate": 1.3496154483524876e-05, "loss": 0.0053, "step": 77250 }, { "grad_norm": 0.24712319672107697, "learning_rate": 1.3484857245934506e-05, "loss": 0.0073, "step": 77260 }, { "grad_norm": 0.1862587183713913, "learning_rate": 1.3473564001585248e-05, "loss": 0.006, "step": 77270 }, { "grad_norm": 0.1240641176700592, "learning_rate": 1.3462274751712129e-05, "loss": 0.0075, "step": 77280 }, { "grad_norm": 0.06076739355921745, "learning_rate": 1.3450989497549698e-05, "loss": 0.0078, "step": 77290 }, { "grad_norm": 0.16911736130714417, "learning_rate": 1.3439708240332115e-05, "loss": 0.0063, "step": 77300 }, { "grad_norm": 0.15362860262393951, "learning_rate": 1.3428430981293043e-05, "loss": 0.0069, "step": 77310 }, { "grad_norm": 0.15213458240032196, "learning_rate": 1.3417157721665808e-05, "loss": 0.0072, "step": 77320 }, { "grad_norm": 0.19380252063274384, "learning_rate": 1.3405888462683186e-05, "loss": 0.006, "step": 77330 }, { "grad_norm": 0.10009908676147461, "learning_rate": 1.3394623205577595e-05, "loss": 0.0059, "step": 77340 }, { "grad_norm": 0.109915591776371, "learning_rate": 1.3383361951580963e-05, "loss": 0.0097, "step": 77350 }, { "grad_norm": 0.11905417591333389, "learning_rate": 1.337210470192483e-05, "loss": 0.0055, "step": 77360 }, { "grad_norm": 0.09226179122924805, "learning_rate": 1.3360851457840245e-05, "loss": 0.0053, "step": 77370 }, { "grad_norm": 0.11734987050294876, "learning_rate": 1.3349602220557856e-05, "loss": 0.0071, "step": 77380 }, { "grad_norm": 0.11769696325063705, "learning_rate": 1.3338356991307865e-05, "loss": 0.0063, "step": 77390 }, { "grad_norm": 0.18047796189785004, "learning_rate": 1.3327115771320048e-05, "loss": 0.0069, "step": 77400 }, { "grad_norm": 0.08058606833219528, "learning_rate": 1.3315878561823697e-05, "loss": 0.0071, "step": 77410 }, { "grad_norm": 0.13384698331356049, "learning_rate": 1.330464536404773e-05, "loss": 0.0062, "step": 77420 }, { "grad_norm": 0.05886880308389664, "learning_rate": 1.3293416179220563e-05, "loss": 0.0072, "step": 77430 }, { "grad_norm": 0.12745383381843567, "learning_rate": 1.328219100857021e-05, "loss": 0.0083, "step": 77440 }, { "grad_norm": 0.14479093253612518, "learning_rate": 1.327096985332425e-05, "loss": 0.0064, "step": 77450 }, { "grad_norm": 0.1237097904086113, "learning_rate": 1.3259752714709816e-05, "loss": 0.0075, "step": 77460 }, { "grad_norm": 0.12917745113372803, "learning_rate": 1.3248539593953569e-05, "loss": 0.0067, "step": 77470 }, { "grad_norm": 0.14463604986667633, "learning_rate": 1.323733049228178e-05, "loss": 0.0094, "step": 77480 }, { "grad_norm": 0.12692375481128693, "learning_rate": 1.3226125410920276e-05, "loss": 0.0076, "step": 77490 }, { "grad_norm": 0.15615110099315643, "learning_rate": 1.3214924351094388e-05, "loss": 0.0083, "step": 77500 }, { "grad_norm": 0.22825047373771667, "learning_rate": 1.3203727314029069e-05, "loss": 0.0058, "step": 77510 }, { "grad_norm": 0.2209206372499466, "learning_rate": 1.3192534300948811e-05, "loss": 0.0065, "step": 77520 }, { "grad_norm": 0.08285178244113922, "learning_rate": 1.3181345313077675e-05, "loss": 0.0048, "step": 77530 }, { "grad_norm": 0.24545519053936005, "learning_rate": 1.3170160351639243e-05, "loss": 0.0099, "step": 77540 }, { "grad_norm": 0.12669338285923004, "learning_rate": 1.3158979417856721e-05, "loss": 0.0072, "step": 77550 }, { "grad_norm": 0.09493110328912735, "learning_rate": 1.31478025129528e-05, "loss": 0.0053, "step": 77560 }, { "grad_norm": 0.10558725148439407, "learning_rate": 1.3136629638149794e-05, "loss": 0.0052, "step": 77570 }, { "grad_norm": 0.0875847116112709, "learning_rate": 1.3125460794669547e-05, "loss": 0.0066, "step": 77580 }, { "grad_norm": 0.14771804213523865, "learning_rate": 1.3114295983733483e-05, "loss": 0.0083, "step": 77590 }, { "grad_norm": 0.15768134593963623, "learning_rate": 1.3103135206562534e-05, "loss": 0.0073, "step": 77600 }, { "grad_norm": 0.20192289352416992, "learning_rate": 1.3091978464377264e-05, "loss": 0.0055, "step": 77610 }, { "grad_norm": 0.156413272023201, "learning_rate": 1.3080825758397724e-05, "loss": 0.0066, "step": 77620 }, { "grad_norm": 0.11021362245082855, "learning_rate": 1.3069677089843568e-05, "loss": 0.0059, "step": 77630 }, { "grad_norm": 0.06982234865427017, "learning_rate": 1.305853245993401e-05, "loss": 0.008, "step": 77640 }, { "grad_norm": 0.10752392560243607, "learning_rate": 1.3047391869887809e-05, "loss": 0.0079, "step": 77650 }, { "grad_norm": 0.19130375981330872, "learning_rate": 1.3036255320923263e-05, "loss": 0.0071, "step": 77660 }, { "grad_norm": 0.10748905688524246, "learning_rate": 1.3025122814258273e-05, "loss": 0.0072, "step": 77670 }, { "grad_norm": 0.07465071976184845, "learning_rate": 1.3013994351110249e-05, "loss": 0.0058, "step": 77680 }, { "grad_norm": 0.14482054114341736, "learning_rate": 1.3002869932696194e-05, "loss": 0.0064, "step": 77690 }, { "grad_norm": 0.1139601618051529, "learning_rate": 1.2991749560232664e-05, "loss": 0.0062, "step": 77700 }, { "grad_norm": 0.0716744214296341, "learning_rate": 1.2980633234935768e-05, "loss": 0.0076, "step": 77710 }, { "grad_norm": 0.08735958486795425, "learning_rate": 1.2969520958021153e-05, "loss": 0.0099, "step": 77720 }, { "grad_norm": 0.12689848244190216, "learning_rate": 1.295841273070405e-05, "loss": 0.0057, "step": 77730 }, { "grad_norm": 0.13842977583408356, "learning_rate": 1.2947308554199255e-05, "loss": 0.0063, "step": 77740 }, { "grad_norm": 0.18304653465747833, "learning_rate": 1.2936208429721075e-05, "loss": 0.0073, "step": 77750 }, { "grad_norm": 0.11820518225431442, "learning_rate": 1.2925112358483416e-05, "loss": 0.0064, "step": 77760 }, { "grad_norm": 0.14186562597751617, "learning_rate": 1.2914020341699728e-05, "loss": 0.0058, "step": 77770 }, { "grad_norm": 0.11554202437400818, "learning_rate": 1.2902932380583032e-05, "loss": 0.0062, "step": 77780 }, { "grad_norm": 0.1210804358124733, "learning_rate": 1.2891848476345863e-05, "loss": 0.0075, "step": 77790 }, { "grad_norm": 0.09446676820516586, "learning_rate": 1.2880768630200368e-05, "loss": 0.0075, "step": 77800 }, { "grad_norm": 0.1573287546634674, "learning_rate": 1.286969284335819e-05, "loss": 0.0072, "step": 77810 }, { "grad_norm": 0.09186718612909317, "learning_rate": 1.2858621117030578e-05, "loss": 0.0063, "step": 77820 }, { "grad_norm": 0.11506098508834839, "learning_rate": 1.2847553452428324e-05, "loss": 0.0064, "step": 77830 }, { "grad_norm": 0.10649005323648453, "learning_rate": 1.283648985076178e-05, "loss": 0.0058, "step": 77840 }, { "grad_norm": 0.10076773166656494, "learning_rate": 1.2825430313240817e-05, "loss": 0.0064, "step": 77850 }, { "grad_norm": 0.07693082094192505, "learning_rate": 1.281437484107491e-05, "loss": 0.0052, "step": 77860 }, { "grad_norm": 0.11291130632162094, "learning_rate": 1.2803323435473058e-05, "loss": 0.0053, "step": 77870 }, { "grad_norm": 0.0761977881193161, "learning_rate": 1.2792276097643841e-05, "loss": 0.0067, "step": 77880 }, { "grad_norm": 0.13588641583919525, "learning_rate": 1.2781232828795342e-05, "loss": 0.0077, "step": 77890 }, { "grad_norm": 0.1672632098197937, "learning_rate": 1.2770193630135292e-05, "loss": 0.0084, "step": 77900 }, { "grad_norm": 0.21284902095794678, "learning_rate": 1.2759158502870877e-05, "loss": 0.0061, "step": 77910 }, { "grad_norm": 0.14296627044677734, "learning_rate": 1.2748127448208913e-05, "loss": 0.0062, "step": 77920 }, { "grad_norm": 0.10617059469223022, "learning_rate": 1.2737100467355706e-05, "loss": 0.0051, "step": 77930 }, { "grad_norm": 0.10821134597063065, "learning_rate": 1.2726077561517169e-05, "loss": 0.008, "step": 77940 }, { "grad_norm": 0.09782911092042923, "learning_rate": 1.2715058731898755e-05, "loss": 0.0072, "step": 77950 }, { "grad_norm": 0.09213732182979584, "learning_rate": 1.2704043979705471e-05, "loss": 0.0047, "step": 77960 }, { "grad_norm": 0.1500069797039032, "learning_rate": 1.269303330614185e-05, "loss": 0.0094, "step": 77970 }, { "grad_norm": 0.08838927000761032, "learning_rate": 1.2682026712412015e-05, "loss": 0.0046, "step": 77980 }, { "grad_norm": 0.12972262501716614, "learning_rate": 1.2671024199719644e-05, "loss": 0.0049, "step": 77990 }, { "grad_norm": 0.1738046407699585, "learning_rate": 1.2660025769267936e-05, "loss": 0.0056, "step": 78000 }, { "grad_norm": 0.11675412207841873, "learning_rate": 1.2649031422259667e-05, "loss": 0.0093, "step": 78010 }, { "grad_norm": 0.12087652087211609, "learning_rate": 1.2638041159897163e-05, "loss": 0.0064, "step": 78020 }, { "grad_norm": 0.09068608283996582, "learning_rate": 1.262705498338232e-05, "loss": 0.0051, "step": 78030 }, { "grad_norm": 0.12069369852542877, "learning_rate": 1.2616072893916542e-05, "loss": 0.0059, "step": 78040 }, { "grad_norm": 0.10429266840219498, "learning_rate": 1.2605094892700836e-05, "loss": 0.0062, "step": 78050 }, { "grad_norm": 0.11147274821996689, "learning_rate": 1.259412098093572e-05, "loss": 0.0074, "step": 78060 }, { "grad_norm": 0.1330648511648178, "learning_rate": 1.258315115982131e-05, "loss": 0.0062, "step": 78070 }, { "grad_norm": 0.1757335364818573, "learning_rate": 1.2572185430557203e-05, "loss": 0.009, "step": 78080 }, { "grad_norm": 0.09846710413694382, "learning_rate": 1.2561223794342659e-05, "loss": 0.0053, "step": 78090 }, { "grad_norm": 0.1709536910057068, "learning_rate": 1.255026625237637e-05, "loss": 0.0056, "step": 78100 }, { "grad_norm": 0.1092587560415268, "learning_rate": 1.2539312805856678e-05, "loss": 0.0099, "step": 78110 }, { "grad_norm": 0.10141913592815399, "learning_rate": 1.25283634559814e-05, "loss": 0.0053, "step": 78120 }, { "grad_norm": 0.18690595030784607, "learning_rate": 1.2517418203947972e-05, "loss": 0.0054, "step": 78130 }, { "grad_norm": 0.16858460009098053, "learning_rate": 1.2506477050953308e-05, "loss": 0.0089, "step": 78140 }, { "grad_norm": 0.09424524754285812, "learning_rate": 1.2495539998193972e-05, "loss": 0.0048, "step": 78150 }, { "grad_norm": 0.15238924324512482, "learning_rate": 1.2484607046865975e-05, "loss": 0.0046, "step": 78160 }, { "grad_norm": 0.1829279661178589, "learning_rate": 1.2473678198164968e-05, "loss": 0.0076, "step": 78170 }, { "grad_norm": 0.1107524037361145, "learning_rate": 1.2462753453286075e-05, "loss": 0.007, "step": 78180 }, { "grad_norm": 0.1035054475069046, "learning_rate": 1.2451832813424031e-05, "loss": 0.0063, "step": 78190 }, { "grad_norm": 0.11892760545015335, "learning_rate": 1.2440916279773095e-05, "loss": 0.0065, "step": 78200 }, { "grad_norm": 0.12885703146457672, "learning_rate": 1.2430003853527101e-05, "loss": 0.0055, "step": 78210 }, { "grad_norm": 0.16233736276626587, "learning_rate": 1.2419095535879383e-05, "loss": 0.0056, "step": 78220 }, { "grad_norm": 0.11436309665441513, "learning_rate": 1.2408191328022878e-05, "loss": 0.0081, "step": 78230 }, { "grad_norm": 0.10735559463500977, "learning_rate": 1.2397291231150066e-05, "loss": 0.0061, "step": 78240 }, { "grad_norm": 0.17895658314228058, "learning_rate": 1.2386395246452937e-05, "loss": 0.0094, "step": 78250 }, { "grad_norm": 0.10659729689359665, "learning_rate": 1.2375503375123083e-05, "loss": 0.0064, "step": 78260 }, { "grad_norm": 0.1478453278541565, "learning_rate": 1.2364615618351583e-05, "loss": 0.0087, "step": 78270 }, { "grad_norm": 0.2333257794380188, "learning_rate": 1.2353731977329169e-05, "loss": 0.0057, "step": 78280 }, { "grad_norm": 0.12901099026203156, "learning_rate": 1.2342852453246007e-05, "loss": 0.0054, "step": 78290 }, { "grad_norm": 0.10372601449489594, "learning_rate": 1.2331977047291898e-05, "loss": 0.0049, "step": 78300 }, { "grad_norm": 0.05906422808766365, "learning_rate": 1.2321105760656137e-05, "loss": 0.0065, "step": 78310 }, { "grad_norm": 0.13655917346477509, "learning_rate": 1.2310238594527613e-05, "loss": 0.0062, "step": 78320 }, { "grad_norm": 0.10181759297847748, "learning_rate": 1.2299375550094693e-05, "loss": 0.0059, "step": 78330 }, { "grad_norm": 0.0861118733882904, "learning_rate": 1.2288516628545421e-05, "loss": 0.0061, "step": 78340 }, { "grad_norm": 0.24688875675201416, "learning_rate": 1.2277661831067255e-05, "loss": 0.0062, "step": 78350 }, { "grad_norm": 0.09769147634506226, "learning_rate": 1.2266811158847286e-05, "loss": 0.0054, "step": 78360 }, { "grad_norm": 0.09135814011096954, "learning_rate": 1.2255964613072107e-05, "loss": 0.0097, "step": 78370 }, { "grad_norm": 0.09784357994794846, "learning_rate": 1.2245122194927905e-05, "loss": 0.0062, "step": 78380 }, { "grad_norm": 0.08802465349435806, "learning_rate": 1.2234283905600364e-05, "loss": 0.0057, "step": 78390 }, { "grad_norm": 0.12301691621541977, "learning_rate": 1.222344974627475e-05, "loss": 0.0052, "step": 78400 }, { "grad_norm": 0.15870077908039093, "learning_rate": 1.2212619718135875e-05, "loss": 0.006, "step": 78410 }, { "grad_norm": 0.11726021021604538, "learning_rate": 1.2201793822368118e-05, "loss": 0.008, "step": 78420 }, { "grad_norm": 0.11724685877561569, "learning_rate": 1.2190972060155337e-05, "loss": 0.0059, "step": 78430 }, { "grad_norm": 0.16077326238155365, "learning_rate": 1.2180154432681013e-05, "loss": 0.006, "step": 78440 }, { "grad_norm": 0.10204274952411652, "learning_rate": 1.2169340941128148e-05, "loss": 0.0065, "step": 78450 }, { "grad_norm": 0.1228993833065033, "learning_rate": 1.2158531586679267e-05, "loss": 0.0068, "step": 78460 }, { "grad_norm": 0.10348086059093475, "learning_rate": 1.2147726370516477e-05, "loss": 0.0074, "step": 78470 }, { "grad_norm": 0.10013141483068466, "learning_rate": 1.2136925293821422e-05, "loss": 0.0103, "step": 78480 }, { "grad_norm": 0.14230704307556152, "learning_rate": 1.2126128357775306e-05, "loss": 0.0055, "step": 78490 }, { "grad_norm": 0.1118437871336937, "learning_rate": 1.2115335563558838e-05, "loss": 0.0047, "step": 78500 }, { "grad_norm": 0.15369707345962524, "learning_rate": 1.2104546912352327e-05, "loss": 0.0056, "step": 78510 }, { "grad_norm": 0.1221206933259964, "learning_rate": 1.2093762405335557e-05, "loss": 0.005, "step": 78520 }, { "grad_norm": 0.0728394165635109, "learning_rate": 1.2082982043687974e-05, "loss": 0.0052, "step": 78530 }, { "grad_norm": 0.2348700314760208, "learning_rate": 1.207220582858845e-05, "loss": 0.0055, "step": 78540 }, { "grad_norm": 0.11531601846218109, "learning_rate": 1.206143376121549e-05, "loss": 0.0059, "step": 78550 }, { "grad_norm": 0.18773505091667175, "learning_rate": 1.2050665842747078e-05, "loss": 0.0109, "step": 78560 }, { "grad_norm": 0.12127482146024704, "learning_rate": 1.2039902074360804e-05, "loss": 0.0065, "step": 78570 }, { "grad_norm": 0.1297021359205246, "learning_rate": 1.2029142457233755e-05, "loss": 0.0094, "step": 78580 }, { "grad_norm": 0.12058784067630768, "learning_rate": 1.2018386992542601e-05, "loss": 0.0066, "step": 78590 }, { "grad_norm": 0.14047586917877197, "learning_rate": 1.2007635681463542e-05, "loss": 0.0069, "step": 78600 }, { "grad_norm": 0.13534080982208252, "learning_rate": 1.1996888525172334e-05, "loss": 0.0073, "step": 78610 }, { "grad_norm": 0.10246355831623077, "learning_rate": 1.198614552484425e-05, "loss": 0.0074, "step": 78620 }, { "grad_norm": 0.15707041323184967, "learning_rate": 1.197540668165415e-05, "loss": 0.0063, "step": 78630 }, { "grad_norm": 0.13662421703338623, "learning_rate": 1.1964671996776395e-05, "loss": 0.0063, "step": 78640 }, { "grad_norm": 0.08941879868507385, "learning_rate": 1.1953941471384922e-05, "loss": 0.0086, "step": 78650 }, { "grad_norm": 0.0859532579779625, "learning_rate": 1.1943215106653205e-05, "loss": 0.0068, "step": 78660 }, { "grad_norm": 0.26127302646636963, "learning_rate": 1.193249290375429e-05, "loss": 0.0053, "step": 78670 }, { "grad_norm": 0.07859904319047928, "learning_rate": 1.1921774863860696e-05, "loss": 0.0054, "step": 78680 }, { "grad_norm": 0.1431523859500885, "learning_rate": 1.1911060988144556e-05, "loss": 0.0083, "step": 78690 }, { "grad_norm": 0.19849705696105957, "learning_rate": 1.1900351277777533e-05, "loss": 0.0079, "step": 78700 }, { "grad_norm": 0.15667150914669037, "learning_rate": 1.1889645733930798e-05, "loss": 0.0078, "step": 78710 }, { "grad_norm": 0.11156527698040009, "learning_rate": 1.1878944357775107e-05, "loss": 0.0065, "step": 78720 }, { "grad_norm": 0.10369502007961273, "learning_rate": 1.1868247150480749e-05, "loss": 0.0069, "step": 78730 }, { "grad_norm": 0.1740146279335022, "learning_rate": 1.1857554113217567e-05, "loss": 0.0055, "step": 78740 }, { "grad_norm": 0.9627758264541626, "learning_rate": 1.1846865247154903e-05, "loss": 0.0067, "step": 78750 }, { "grad_norm": 0.1293146014213562, "learning_rate": 1.1836180553461711e-05, "loss": 0.0051, "step": 78760 }, { "grad_norm": 0.10293763875961304, "learning_rate": 1.1825500033306419e-05, "loss": 0.0061, "step": 78770 }, { "grad_norm": 0.12055690586566925, "learning_rate": 1.1814823687857052e-05, "loss": 0.0078, "step": 78780 }, { "grad_norm": 0.08936480432748795, "learning_rate": 1.1804151518281153e-05, "loss": 0.0045, "step": 78790 }, { "grad_norm": 0.10438156872987747, "learning_rate": 1.179348352574583e-05, "loss": 0.0075, "step": 78800 }, { "grad_norm": 0.07729631662368774, "learning_rate": 1.1782819711417697e-05, "loss": 0.0051, "step": 78810 }, { "grad_norm": 0.1327306479215622, "learning_rate": 1.1772160076462951e-05, "loss": 0.005, "step": 78820 }, { "grad_norm": 0.13107706606388092, "learning_rate": 1.1761504622047293e-05, "loss": 0.0077, "step": 78830 }, { "grad_norm": 0.18154917657375336, "learning_rate": 1.1750853349335999e-05, "loss": 0.008, "step": 78840 }, { "grad_norm": 0.14949077367782593, "learning_rate": 1.1740206259493874e-05, "loss": 0.006, "step": 78850 }, { "grad_norm": 0.2526397705078125, "learning_rate": 1.1729563353685286e-05, "loss": 0.0081, "step": 78860 }, { "grad_norm": 0.0840618759393692, "learning_rate": 1.1718924633074091e-05, "loss": 0.0093, "step": 78870 }, { "grad_norm": 0.11936761438846588, "learning_rate": 1.1708290098823765e-05, "loss": 0.0057, "step": 78880 }, { "grad_norm": 0.1952802687883377, "learning_rate": 1.1697659752097245e-05, "loss": 0.0091, "step": 78890 }, { "grad_norm": 0.2874143123626709, "learning_rate": 1.1687033594057068e-05, "loss": 0.0045, "step": 78900 }, { "grad_norm": 0.11121420562267303, "learning_rate": 1.1676411625865297e-05, "loss": 0.0058, "step": 78910 }, { "grad_norm": 0.18361878395080566, "learning_rate": 1.1665793848683542e-05, "loss": 0.0069, "step": 78920 }, { "grad_norm": 0.16677841544151306, "learning_rate": 1.1655180263672927e-05, "loss": 0.006, "step": 78930 }, { "grad_norm": 0.14851684868335724, "learning_rate": 1.1644570871994142e-05, "loss": 0.0062, "step": 78940 }, { "grad_norm": 0.11763569712638855, "learning_rate": 1.163396567480744e-05, "loss": 0.0048, "step": 78950 }, { "grad_norm": 0.10099469870328903, "learning_rate": 1.1623364673272552e-05, "loss": 0.0057, "step": 78960 }, { "grad_norm": 0.12395799905061722, "learning_rate": 1.1612767868548807e-05, "loss": 0.0058, "step": 78970 }, { "grad_norm": 0.10407814383506775, "learning_rate": 1.1602175261795057e-05, "loss": 0.0078, "step": 78980 }, { "grad_norm": 0.15793702006340027, "learning_rate": 1.1591586854169696e-05, "loss": 0.0053, "step": 78990 }, { "grad_norm": 0.06434907764196396, "learning_rate": 1.158100264683064e-05, "loss": 0.0058, "step": 79000 }, { "grad_norm": 0.12633581459522247, "learning_rate": 1.1570422640935386e-05, "loss": 0.0108, "step": 79010 }, { "grad_norm": 0.15463787317276, "learning_rate": 1.155984683764092e-05, "loss": 0.008, "step": 79020 }, { "grad_norm": 0.1158856526017189, "learning_rate": 1.1549275238103812e-05, "loss": 0.0106, "step": 79030 }, { "grad_norm": 0.16972893476486206, "learning_rate": 1.1538707843480152e-05, "loss": 0.0076, "step": 79040 }, { "grad_norm": 0.11628260463476181, "learning_rate": 1.1528144654925593e-05, "loss": 0.0045, "step": 79050 }, { "grad_norm": 0.08971519023180008, "learning_rate": 1.1517585673595277e-05, "loss": 0.006, "step": 79060 }, { "grad_norm": 0.10697636753320694, "learning_rate": 1.150703090064395e-05, "loss": 0.0072, "step": 79070 }, { "grad_norm": 0.15579243004322052, "learning_rate": 1.1496480337225835e-05, "loss": 0.0053, "step": 79080 }, { "grad_norm": 0.10193843394517899, "learning_rate": 1.1485933984494751e-05, "loss": 0.0054, "step": 79090 }, { "grad_norm": 0.10014978051185608, "learning_rate": 1.1475391843604e-05, "loss": 0.0052, "step": 79100 }, { "grad_norm": 0.10322900116443634, "learning_rate": 1.1464853915706503e-05, "loss": 0.0051, "step": 79110 }, { "grad_norm": 0.10967306047677994, "learning_rate": 1.1454320201954626e-05, "loss": 0.005, "step": 79120 }, { "grad_norm": 0.10928413271903992, "learning_rate": 1.1443790703500356e-05, "loss": 0.0054, "step": 79130 }, { "grad_norm": 0.06462998688220978, "learning_rate": 1.1433265421495154e-05, "loss": 0.0063, "step": 79140 }, { "grad_norm": 0.10098123550415039, "learning_rate": 1.1422744357090065e-05, "loss": 0.0051, "step": 79150 }, { "grad_norm": 0.1720433533191681, "learning_rate": 1.1412227511435647e-05, "loss": 0.0107, "step": 79160 }, { "grad_norm": 0.07540559023618698, "learning_rate": 1.1401714885682025e-05, "loss": 0.0047, "step": 79170 }, { "grad_norm": 0.09828241169452667, "learning_rate": 1.1391206480978823e-05, "loss": 0.01, "step": 79180 }, { "grad_norm": 0.15398815274238586, "learning_rate": 1.1380702298475227e-05, "loss": 0.0076, "step": 79190 }, { "grad_norm": 0.1085420623421669, "learning_rate": 1.1370202339319985e-05, "loss": 0.0057, "step": 79200 }, { "grad_norm": 0.12191423028707504, "learning_rate": 1.1359706604661319e-05, "loss": 0.0088, "step": 79210 }, { "grad_norm": 0.2086804360151291, "learning_rate": 1.1349215095647042e-05, "loss": 0.0058, "step": 79220 }, { "grad_norm": 0.0880054458975792, "learning_rate": 1.1338727813424493e-05, "loss": 0.0051, "step": 79230 }, { "grad_norm": 0.1582484096288681, "learning_rate": 1.132824475914056e-05, "loss": 0.005, "step": 79240 }, { "grad_norm": 0.14005716145038605, "learning_rate": 1.131776593394162e-05, "loss": 0.0076, "step": 79250 }, { "grad_norm": 0.15024517476558685, "learning_rate": 1.1307291338973657e-05, "loss": 0.0051, "step": 79260 }, { "grad_norm": 0.17472760379314423, "learning_rate": 1.1296820975382121e-05, "loss": 0.006, "step": 79270 }, { "grad_norm": 0.1324390321969986, "learning_rate": 1.1286354844312074e-05, "loss": 0.0055, "step": 79280 }, { "grad_norm": 0.1236453503370285, "learning_rate": 1.1275892946908023e-05, "loss": 0.0061, "step": 79290 }, { "grad_norm": 0.11793410032987595, "learning_rate": 1.126543528431413e-05, "loss": 0.0047, "step": 79300 }, { "grad_norm": 0.15051499009132385, "learning_rate": 1.125498185767398e-05, "loss": 0.0061, "step": 79310 }, { "grad_norm": 0.11331293731927872, "learning_rate": 1.1244532668130781e-05, "loss": 0.0069, "step": 79320 }, { "grad_norm": 0.16667823493480682, "learning_rate": 1.1234087716827203e-05, "loss": 0.0076, "step": 79330 }, { "grad_norm": 0.18627400696277618, "learning_rate": 1.1223647004905524e-05, "loss": 0.0064, "step": 79340 }, { "grad_norm": 0.13455595076084137, "learning_rate": 1.1213210533507485e-05, "loss": 0.0075, "step": 79350 }, { "grad_norm": 0.0962454080581665, "learning_rate": 1.1202778303774447e-05, "loss": 0.0053, "step": 79360 }, { "grad_norm": 0.11715048551559448, "learning_rate": 1.1192350316847234e-05, "loss": 0.006, "step": 79370 }, { "grad_norm": 0.11866523325443268, "learning_rate": 1.1181926573866258e-05, "loss": 0.0062, "step": 79380 }, { "grad_norm": 0.11350194364786148, "learning_rate": 1.1171507075971416e-05, "loss": 0.0079, "step": 79390 }, { "grad_norm": 0.12876807153224945, "learning_rate": 1.1161091824302177e-05, "loss": 0.0064, "step": 79400 }, { "grad_norm": 0.0689568966627121, "learning_rate": 1.115068081999755e-05, "loss": 0.0081, "step": 79410 }, { "grad_norm": 0.11580656468868256, "learning_rate": 1.1140274064196071e-05, "loss": 0.0067, "step": 79420 }, { "grad_norm": 0.1149725541472435, "learning_rate": 1.112987155803578e-05, "loss": 0.0073, "step": 79430 }, { "grad_norm": 0.08316157758235931, "learning_rate": 1.1119473302654298e-05, "loss": 0.006, "step": 79440 }, { "grad_norm": 0.09295704960823059, "learning_rate": 1.1109079299188769e-05, "loss": 0.007, "step": 79450 }, { "grad_norm": 0.21603326499462128, "learning_rate": 1.1098689548775847e-05, "loss": 0.0074, "step": 79460 }, { "grad_norm": 0.10918904840946198, "learning_rate": 1.1088304052551762e-05, "loss": 0.0061, "step": 79470 }, { "grad_norm": 0.11520873755216599, "learning_rate": 1.1077922811652215e-05, "loss": 0.0109, "step": 79480 }, { "grad_norm": 0.10501696169376373, "learning_rate": 1.1067545827212534e-05, "loss": 0.0047, "step": 79490 }, { "grad_norm": 0.09153499454259872, "learning_rate": 1.1057173100367496e-05, "loss": 0.0053, "step": 79500 }, { "grad_norm": 0.18949156999588013, "learning_rate": 1.1046804632251474e-05, "loss": 0.0054, "step": 79510 }, { "grad_norm": 0.21136322617530823, "learning_rate": 1.1036440423998318e-05, "loss": 0.0074, "step": 79520 }, { "grad_norm": 0.16329924762248993, "learning_rate": 1.1026080476741469e-05, "loss": 0.0061, "step": 79530 }, { "grad_norm": 0.1785728633403778, "learning_rate": 1.1015724791613836e-05, "loss": 0.0059, "step": 79540 }, { "grad_norm": 0.10553419589996338, "learning_rate": 1.1005373369747951e-05, "loss": 0.0058, "step": 79550 }, { "grad_norm": 0.08353134244680405, "learning_rate": 1.09950262122758e-05, "loss": 0.0064, "step": 79560 }, { "grad_norm": 0.08155523240566254, "learning_rate": 1.0984683320328948e-05, "loss": 0.0076, "step": 79570 }, { "grad_norm": 0.11803784966468811, "learning_rate": 1.0974344695038458e-05, "loss": 0.0068, "step": 79580 }, { "grad_norm": 0.16321209073066711, "learning_rate": 1.0964010337534968e-05, "loss": 0.0065, "step": 79590 }, { "grad_norm": 0.143156498670578, "learning_rate": 1.0953680248948611e-05, "loss": 0.008, "step": 79600 }, { "grad_norm": 0.1917591392993927, "learning_rate": 1.0943354430409069e-05, "loss": 0.0087, "step": 79610 }, { "grad_norm": 0.07943940907716751, "learning_rate": 1.0933032883045574e-05, "loss": 0.0055, "step": 79620 }, { "grad_norm": 0.24212463200092316, "learning_rate": 1.0922715607986872e-05, "loss": 0.0069, "step": 79630 }, { "grad_norm": 0.07450113445520401, "learning_rate": 1.0912402606361234e-05, "loss": 0.0089, "step": 79640 }, { "grad_norm": 0.14254774153232574, "learning_rate": 1.090209387929647e-05, "loss": 0.0051, "step": 79650 }, { "grad_norm": 0.13166861236095428, "learning_rate": 1.0891789427919957e-05, "loss": 0.0062, "step": 79660 }, { "grad_norm": 0.10669813305139542, "learning_rate": 1.088148925335854e-05, "loss": 0.0067, "step": 79670 }, { "grad_norm": 0.12010017782449722, "learning_rate": 1.087119335673864e-05, "loss": 0.0076, "step": 79680 }, { "grad_norm": 0.2791071832180023, "learning_rate": 1.0860901739186208e-05, "loss": 0.0069, "step": 79690 }, { "grad_norm": 0.148931622505188, "learning_rate": 1.085061440182673e-05, "loss": 0.0089, "step": 79700 }, { "grad_norm": 0.1043446809053421, "learning_rate": 1.0840331345785193e-05, "loss": 0.0073, "step": 79710 }, { "grad_norm": 0.1045607253909111, "learning_rate": 1.083005257218615e-05, "loss": 0.0058, "step": 79720 }, { "grad_norm": 0.19828981161117554, "learning_rate": 1.0819778082153642e-05, "loss": 0.0066, "step": 79730 }, { "grad_norm": 0.10740478336811066, "learning_rate": 1.0809507876811325e-05, "loss": 0.0052, "step": 79740 }, { "grad_norm": 0.1668630987405777, "learning_rate": 1.0799241957282292e-05, "loss": 0.0066, "step": 79750 }, { "grad_norm": 0.11085031181573868, "learning_rate": 1.078898032468923e-05, "loss": 0.0049, "step": 79760 }, { "grad_norm": 0.11213447153568268, "learning_rate": 1.0778722980154315e-05, "loss": 0.0073, "step": 79770 }, { "grad_norm": 0.18155701458454132, "learning_rate": 1.0768469924799302e-05, "loss": 0.0093, "step": 79780 }, { "grad_norm": 0.18986929953098297, "learning_rate": 1.0758221159745419e-05, "loss": 0.0071, "step": 79790 }, { "grad_norm": 0.2279922515153885, "learning_rate": 1.074797668611347e-05, "loss": 0.0083, "step": 79800 }, { "grad_norm": 0.119014672935009, "learning_rate": 1.0737736505023777e-05, "loss": 0.0072, "step": 79810 }, { "grad_norm": 0.08294833451509476, "learning_rate": 1.0727500617596203e-05, "loss": 0.0051, "step": 79820 }, { "grad_norm": 0.127372145652771, "learning_rate": 1.0717269024950104e-05, "loss": 0.0075, "step": 79830 }, { "grad_norm": 0.08790972083806992, "learning_rate": 1.070704172820442e-05, "loss": 0.0077, "step": 79840 }, { "grad_norm": 0.06432918459177017, "learning_rate": 1.0696818728477559e-05, "loss": 0.0067, "step": 79850 }, { "grad_norm": 0.1491575986146927, "learning_rate": 1.0686600026887511e-05, "loss": 0.0092, "step": 79860 }, { "grad_norm": 0.08351622521877289, "learning_rate": 1.0676385624551777e-05, "loss": 0.0068, "step": 79870 }, { "grad_norm": 0.1099674254655838, "learning_rate": 1.0666175522587402e-05, "loss": 0.0054, "step": 79880 }, { "grad_norm": 0.1097417026758194, "learning_rate": 1.0655969722110925e-05, "loss": 0.0055, "step": 79890 }, { "grad_norm": 0.08895834535360336, "learning_rate": 1.0645768224238439e-05, "loss": 0.0045, "step": 79900 }, { "grad_norm": 0.17869991064071655, "learning_rate": 1.063557103008559e-05, "loss": 0.0086, "step": 79910 }, { "grad_norm": 0.1693403273820877, "learning_rate": 1.0625378140767494e-05, "loss": 0.006, "step": 79920 }, { "grad_norm": 0.08967672288417816, "learning_rate": 1.0615189557398846e-05, "loss": 0.0055, "step": 79930 }, { "grad_norm": 0.06736291199922562, "learning_rate": 1.060500528109386e-05, "loss": 0.0043, "step": 79940 }, { "grad_norm": 0.10492938756942749, "learning_rate": 1.0594825312966272e-05, "loss": 0.0071, "step": 79950 }, { "grad_norm": 0.09487180411815643, "learning_rate": 1.0584649654129336e-05, "loss": 0.0052, "step": 79960 }, { "grad_norm": 0.1176060140132904, "learning_rate": 1.0574478305695863e-05, "loss": 0.0055, "step": 79970 }, { "grad_norm": 0.12045898288488388, "learning_rate": 1.0564311268778154e-05, "loss": 0.0054, "step": 79980 }, { "grad_norm": 0.08617271482944489, "learning_rate": 1.0554148544488073e-05, "loss": 0.0053, "step": 79990 }, { "grad_norm": 0.13303183019161224, "learning_rate": 1.0543990133937004e-05, "loss": 0.0044, "step": 80000 }, { "grad_norm": 0.13240332901477814, "learning_rate": 1.0533836038235868e-05, "loss": 0.0066, "step": 80010 }, { "grad_norm": 0.14371182024478912, "learning_rate": 1.052368625849507e-05, "loss": 0.008, "step": 80020 }, { "grad_norm": 0.08466688543558121, "learning_rate": 1.05135407958246e-05, "loss": 0.0068, "step": 80030 }, { "grad_norm": 0.13126100599765778, "learning_rate": 1.0503399651333934e-05, "loss": 0.0079, "step": 80040 }, { "grad_norm": 0.10714305937290192, "learning_rate": 1.04932628261321e-05, "loss": 0.0061, "step": 80050 }, { "grad_norm": 0.09414070844650269, "learning_rate": 1.0483130321327644e-05, "loss": 0.0069, "step": 80060 }, { "grad_norm": 0.1891670525074005, "learning_rate": 1.0473002138028654e-05, "loss": 0.0062, "step": 80070 }, { "grad_norm": 0.11516822129487991, "learning_rate": 1.0462878277342713e-05, "loss": 0.005, "step": 80080 }, { "grad_norm": 0.06323281675577164, "learning_rate": 1.0452758740376972e-05, "loss": 0.0077, "step": 80090 }, { "grad_norm": 0.1730279177427292, "learning_rate": 1.0442643528238067e-05, "loss": 0.0051, "step": 80100 }, { "grad_norm": 0.10347907990217209, "learning_rate": 1.0432532642032188e-05, "loss": 0.0062, "step": 80110 }, { "grad_norm": 0.12029337882995605, "learning_rate": 1.042242608286505e-05, "loss": 0.0068, "step": 80120 }, { "grad_norm": 0.10588523000478745, "learning_rate": 1.041232385184191e-05, "loss": 0.0052, "step": 80130 }, { "grad_norm": 0.1192200779914856, "learning_rate": 1.0402225950067495e-05, "loss": 0.0073, "step": 80140 }, { "grad_norm": 0.13161227107048035, "learning_rate": 1.0392132378646119e-05, "loss": 0.0059, "step": 80150 }, { "grad_norm": 0.0941484346985817, "learning_rate": 1.0382043138681607e-05, "loss": 0.0066, "step": 80160 }, { "grad_norm": 0.1454339474439621, "learning_rate": 1.0371958231277284e-05, "loss": 0.0085, "step": 80170 }, { "grad_norm": 0.22120678424835205, "learning_rate": 1.0361877657536024e-05, "loss": 0.0084, "step": 80180 }, { "grad_norm": 0.3851030170917511, "learning_rate": 1.035180141856023e-05, "loss": 0.0077, "step": 80190 }, { "grad_norm": 0.1601240634918213, "learning_rate": 1.0341729515451836e-05, "loss": 0.0068, "step": 80200 }, { "grad_norm": 0.10293278843164444, "learning_rate": 1.0331661949312265e-05, "loss": 0.0079, "step": 80210 }, { "grad_norm": 0.1395479291677475, "learning_rate": 1.0321598721242514e-05, "loss": 0.005, "step": 80220 }, { "grad_norm": 0.10342356562614441, "learning_rate": 1.0311539832343064e-05, "loss": 0.0053, "step": 80230 }, { "grad_norm": 0.1627812534570694, "learning_rate": 1.0301485283713942e-05, "loss": 0.0114, "step": 80240 }, { "grad_norm": 0.12873826920986176, "learning_rate": 1.0291435076454703e-05, "loss": 0.0101, "step": 80250 }, { "grad_norm": 0.06703665852546692, "learning_rate": 1.028138921166444e-05, "loss": 0.0068, "step": 80260 }, { "grad_norm": 0.1435099095106125, "learning_rate": 1.0271347690441719e-05, "loss": 0.0092, "step": 80270 }, { "grad_norm": 0.08244679123163223, "learning_rate": 1.0261310513884698e-05, "loss": 0.0075, "step": 80280 }, { "grad_norm": 0.128352090716362, "learning_rate": 1.0251277683090998e-05, "loss": 0.0062, "step": 80290 }, { "grad_norm": 0.10113553702831268, "learning_rate": 1.0241249199157826e-05, "loss": 0.0055, "step": 80300 }, { "grad_norm": 0.12149869650602341, "learning_rate": 1.0231225063181837e-05, "loss": 0.0059, "step": 80310 }, { "grad_norm": 0.11854042112827301, "learning_rate": 1.0221205276259316e-05, "loss": 0.0073, "step": 80320 }, { "grad_norm": 0.3614053428173065, "learning_rate": 1.0211189839485958e-05, "loss": 0.0093, "step": 80330 }, { "grad_norm": 0.15985654294490814, "learning_rate": 1.0201178753957075e-05, "loss": 0.0082, "step": 80340 }, { "grad_norm": 0.07251385599374771, "learning_rate": 1.0191172020767432e-05, "loss": 0.0054, "step": 80350 }, { "grad_norm": 0.1317775845527649, "learning_rate": 1.0181169641011362e-05, "loss": 0.0067, "step": 80360 }, { "grad_norm": 0.11282441765069962, "learning_rate": 1.0171171615782721e-05, "loss": 0.007, "step": 80370 }, { "grad_norm": 0.10870306938886642, "learning_rate": 1.0161177946174866e-05, "loss": 0.0061, "step": 80380 }, { "grad_norm": 0.09242063015699387, "learning_rate": 1.0151188633280706e-05, "loss": 0.0045, "step": 80390 }, { "grad_norm": 0.1701112985610962, "learning_rate": 1.0141203678192635e-05, "loss": 0.0077, "step": 80400 }, { "grad_norm": 0.14710429310798645, "learning_rate": 1.0131223082002612e-05, "loss": 0.0066, "step": 80410 }, { "grad_norm": 0.15258227288722992, "learning_rate": 1.012124684580208e-05, "loss": 0.0053, "step": 80420 }, { "grad_norm": 0.14883355796337128, "learning_rate": 1.0111274970682039e-05, "loss": 0.0041, "step": 80430 }, { "grad_norm": 0.06809093803167343, "learning_rate": 1.010130745773299e-05, "loss": 0.0071, "step": 80440 }, { "grad_norm": 0.07312040030956268, "learning_rate": 1.0091344308044986e-05, "loss": 0.0049, "step": 80450 }, { "grad_norm": 0.18388037383556366, "learning_rate": 1.0081385522707554e-05, "loss": 0.0083, "step": 80460 }, { "grad_norm": 0.15416331589221954, "learning_rate": 1.0071431102809803e-05, "loss": 0.0067, "step": 80470 }, { "grad_norm": 0.20004670321941376, "learning_rate": 1.0061481049440297e-05, "loss": 0.0049, "step": 80480 }, { "grad_norm": 0.07838930189609528, "learning_rate": 1.0051535363687187e-05, "loss": 0.0047, "step": 80490 }, { "grad_norm": 0.11814796179533005, "learning_rate": 1.0041594046638087e-05, "loss": 0.0041, "step": 80500 }, { "grad_norm": 0.19133348762989044, "learning_rate": 1.0031657099380204e-05, "loss": 0.0059, "step": 80510 }, { "grad_norm": 0.14525935053825378, "learning_rate": 1.0021724523000204e-05, "loss": 0.0078, "step": 80520 }, { "grad_norm": 0.08334145694971085, "learning_rate": 1.001179631858431e-05, "loss": 0.0056, "step": 80530 }, { "grad_norm": 0.07904329895973206, "learning_rate": 1.0001872487218245e-05, "loss": 0.0052, "step": 80540 }, { "grad_norm": 0.16963614523410797, "learning_rate": 9.991953029987272e-06, "loss": 0.0068, "step": 80550 }, { "grad_norm": 0.2082546055316925, "learning_rate": 9.982037947976147e-06, "loss": 0.0086, "step": 80560 }, { "grad_norm": 0.1002274826169014, "learning_rate": 9.972127242269208e-06, "loss": 0.0048, "step": 80570 }, { "grad_norm": 0.18195372819900513, "learning_rate": 9.962220913950238e-06, "loss": 0.0077, "step": 80580 }, { "grad_norm": 0.10568136721849442, "learning_rate": 9.952318964102591e-06, "loss": 0.0043, "step": 80590 }, { "grad_norm": 0.24422454833984375, "learning_rate": 9.942421393809148e-06, "loss": 0.0061, "step": 80600 }, { "grad_norm": 0.07576759904623032, "learning_rate": 9.932528204152264e-06, "loss": 0.0069, "step": 80610 }, { "grad_norm": 0.0970507562160492, "learning_rate": 9.922639396213856e-06, "loss": 0.0059, "step": 80620 }, { "grad_norm": 0.16141842305660248, "learning_rate": 9.912754971075344e-06, "loss": 0.0082, "step": 80630 }, { "grad_norm": 0.0920368954539299, "learning_rate": 9.902874929817696e-06, "loss": 0.0068, "step": 80640 }, { "grad_norm": 0.10838496685028076, "learning_rate": 9.892999273521342e-06, "loss": 0.0097, "step": 80650 }, { "grad_norm": 0.18243756890296936, "learning_rate": 9.883128003266302e-06, "loss": 0.0074, "step": 80660 }, { "grad_norm": 0.07271163165569305, "learning_rate": 9.87326112013206e-06, "loss": 0.004, "step": 80670 }, { "grad_norm": 0.13466452062129974, "learning_rate": 9.863398625197662e-06, "loss": 0.0066, "step": 80680 }, { "grad_norm": 0.08568654954433441, "learning_rate": 9.853540519541615e-06, "loss": 0.0058, "step": 80690 }, { "grad_norm": 0.10915976017713547, "learning_rate": 9.843686804242053e-06, "loss": 0.0049, "step": 80700 }, { "grad_norm": 0.12513607740402222, "learning_rate": 9.833837480376506e-06, "loss": 0.0086, "step": 80710 }, { "grad_norm": 0.1191030964255333, "learning_rate": 9.82399254902212e-06, "loss": 0.0102, "step": 80720 }, { "grad_norm": 0.11756537854671478, "learning_rate": 9.814152011255494e-06, "loss": 0.006, "step": 80730 }, { "grad_norm": 0.129184752702713, "learning_rate": 9.804315868152797e-06, "loss": 0.0058, "step": 80740 }, { "grad_norm": 0.13376040756702423, "learning_rate": 9.79448412078966e-06, "loss": 0.0055, "step": 80750 }, { "grad_norm": 0.09860693663358688, "learning_rate": 9.784656770241312e-06, "loss": 0.0064, "step": 80760 }, { "grad_norm": 0.24013015627861023, "learning_rate": 9.77483381758243e-06, "loss": 0.0052, "step": 80770 }, { "grad_norm": 0.1044684648513794, "learning_rate": 9.76501526388725e-06, "loss": 0.0074, "step": 80780 }, { "grad_norm": 0.1032460629940033, "learning_rate": 9.7552011102295e-06, "loss": 0.0064, "step": 80790 }, { "grad_norm": 0.16274841129779816, "learning_rate": 9.745391357682449e-06, "loss": 0.0067, "step": 80800 }, { "grad_norm": 0.13198347389698029, "learning_rate": 9.735586007318875e-06, "loss": 0.0056, "step": 80810 }, { "grad_norm": 0.09491531550884247, "learning_rate": 9.725785060211096e-06, "loss": 0.0061, "step": 80820 }, { "grad_norm": 0.21785090863704681, "learning_rate": 9.715988517430896e-06, "loss": 0.006, "step": 80830 }, { "grad_norm": 0.14507663249969482, "learning_rate": 9.706196380049625e-06, "loss": 0.0055, "step": 80840 }, { "grad_norm": 0.1254207193851471, "learning_rate": 9.69640864913815e-06, "loss": 0.006, "step": 80850 }, { "grad_norm": 0.07416517287492752, "learning_rate": 9.686625325766818e-06, "loss": 0.008, "step": 80860 }, { "grad_norm": 0.1454884111881256, "learning_rate": 9.67684641100554e-06, "loss": 0.0063, "step": 80870 }, { "grad_norm": 0.0945386067032814, "learning_rate": 9.667071905923692e-06, "loss": 0.0084, "step": 80880 }, { "grad_norm": 0.07482314109802246, "learning_rate": 9.657301811590242e-06, "loss": 0.0073, "step": 80890 }, { "grad_norm": 0.07259352505207062, "learning_rate": 9.647536129073603e-06, "loss": 0.0055, "step": 80900 }, { "grad_norm": 0.16207849979400635, "learning_rate": 9.637774859441751e-06, "loss": 0.0112, "step": 80910 }, { "grad_norm": 0.10081051290035248, "learning_rate": 9.628018003762152e-06, "loss": 0.0063, "step": 80920 }, { "grad_norm": 0.21810396015644073, "learning_rate": 9.618265563101813e-06, "loss": 0.0054, "step": 80930 }, { "grad_norm": 0.0967821255326271, "learning_rate": 9.608517538527211e-06, "loss": 0.0054, "step": 80940 }, { "grad_norm": 0.10500714927911758, "learning_rate": 9.598773931104431e-06, "loss": 0.006, "step": 80950 }, { "grad_norm": 0.11296099424362183, "learning_rate": 9.589034741898983e-06, "loss": 0.0058, "step": 80960 }, { "grad_norm": 0.19591496884822845, "learning_rate": 9.579299971975947e-06, "loss": 0.0052, "step": 80970 }, { "grad_norm": 0.10594936460256577, "learning_rate": 9.569569622399882e-06, "loss": 0.0055, "step": 80980 }, { "grad_norm": 0.07710713148117065, "learning_rate": 9.559843694234916e-06, "loss": 0.0064, "step": 80990 }, { "grad_norm": 0.0809393972158432, "learning_rate": 9.550122188544625e-06, "loss": 0.0054, "step": 81000 }, { "grad_norm": 0.1652802675962448, "learning_rate": 9.540405106392152e-06, "loss": 0.0061, "step": 81010 }, { "grad_norm": 0.10641292482614517, "learning_rate": 9.53069244884015e-06, "loss": 0.0051, "step": 81020 }, { "grad_norm": 0.1304953545331955, "learning_rate": 9.52098421695079e-06, "loss": 0.0068, "step": 81030 }, { "grad_norm": 0.06262090802192688, "learning_rate": 9.511280411785728e-06, "loss": 0.0062, "step": 81040 }, { "grad_norm": 0.14389927685260773, "learning_rate": 9.50158103440616e-06, "loss": 0.0058, "step": 81050 }, { "grad_norm": 0.14400392770767212, "learning_rate": 9.491886085872815e-06, "loss": 0.0086, "step": 81060 }, { "grad_norm": 0.22559134662151337, "learning_rate": 9.482195567245895e-06, "loss": 0.0128, "step": 81070 }, { "grad_norm": 0.10220479965209961, "learning_rate": 9.472509479585151e-06, "loss": 0.0064, "step": 81080 }, { "grad_norm": 0.14038679003715515, "learning_rate": 9.462827823949833e-06, "loss": 0.01, "step": 81090 }, { "grad_norm": 0.07595165818929672, "learning_rate": 9.453150601398735e-06, "loss": 0.0059, "step": 81100 }, { "grad_norm": 0.09603379666805267, "learning_rate": 9.44347781299011e-06, "loss": 0.0074, "step": 81110 }, { "grad_norm": 0.14112034440040588, "learning_rate": 9.43380945978179e-06, "loss": 0.0075, "step": 81120 }, { "grad_norm": 0.10495361685752869, "learning_rate": 9.424145542831047e-06, "loss": 0.0051, "step": 81130 }, { "grad_norm": 0.1457376927137375, "learning_rate": 9.414486063194778e-06, "loss": 0.0055, "step": 81140 }, { "grad_norm": 0.1441269963979721, "learning_rate": 9.404831021929273e-06, "loss": 0.0063, "step": 81150 }, { "grad_norm": 0.11617489159107208, "learning_rate": 9.395180420090427e-06, "loss": 0.0051, "step": 81160 }, { "grad_norm": 0.12155179679393768, "learning_rate": 9.385534258733591e-06, "loss": 0.0047, "step": 81170 }, { "grad_norm": 0.13033443689346313, "learning_rate": 9.375892538913677e-06, "loss": 0.006, "step": 81180 }, { "grad_norm": 0.12023863941431046, "learning_rate": 9.36625526168507e-06, "loss": 0.0068, "step": 81190 }, { "grad_norm": 0.09526699036359787, "learning_rate": 9.356622428101686e-06, "loss": 0.0065, "step": 81200 }, { "grad_norm": 0.13388828933238983, "learning_rate": 9.346994039216972e-06, "loss": 0.0059, "step": 81210 }, { "grad_norm": 0.09038496017456055, "learning_rate": 9.337370096083886e-06, "loss": 0.0068, "step": 81220 }, { "grad_norm": 0.12405301630496979, "learning_rate": 9.327750599754854e-06, "loss": 0.0051, "step": 81230 }, { "grad_norm": 0.09163161367177963, "learning_rate": 9.318135551281875e-06, "loss": 0.005, "step": 81240 }, { "grad_norm": 0.06623288989067078, "learning_rate": 9.308524951716424e-06, "loss": 0.0051, "step": 81250 }, { "grad_norm": 0.12023226171731949, "learning_rate": 9.298918802109497e-06, "loss": 0.0065, "step": 81260 }, { "grad_norm": 0.1551571488380432, "learning_rate": 9.289317103511625e-06, "loss": 0.0052, "step": 81270 }, { "grad_norm": 0.19642940163612366, "learning_rate": 9.279719856972835e-06, "loss": 0.0055, "step": 81280 }, { "grad_norm": 0.10077565908432007, "learning_rate": 9.270127063542645e-06, "loss": 0.0069, "step": 81290 }, { "grad_norm": 0.15177200734615326, "learning_rate": 9.260538724270124e-06, "loss": 0.0052, "step": 81300 }, { "grad_norm": 0.1768132746219635, "learning_rate": 9.250954840203852e-06, "loss": 0.0065, "step": 81310 }, { "grad_norm": 0.1366969496011734, "learning_rate": 9.241375412391878e-06, "loss": 0.0052, "step": 81320 }, { "grad_norm": 0.19666516780853271, "learning_rate": 9.23180044188181e-06, "loss": 0.0038, "step": 81330 }, { "grad_norm": 0.09761671721935272, "learning_rate": 9.222229929720755e-06, "loss": 0.0047, "step": 81340 }, { "grad_norm": 0.1502099633216858, "learning_rate": 9.212663876955336e-06, "loss": 0.0056, "step": 81350 }, { "grad_norm": 0.0823688879609108, "learning_rate": 9.20310228463166e-06, "loss": 0.0059, "step": 81360 }, { "grad_norm": 0.1342718005180359, "learning_rate": 9.19354515379539e-06, "loss": 0.0099, "step": 81370 }, { "grad_norm": 0.10467244684696198, "learning_rate": 9.183992485491666e-06, "loss": 0.0055, "step": 81380 }, { "grad_norm": 0.09381935745477676, "learning_rate": 9.174444280765154e-06, "loss": 0.0041, "step": 81390 }, { "grad_norm": 0.15748122334480286, "learning_rate": 9.164900540660032e-06, "loss": 0.0062, "step": 81400 }, { "grad_norm": 0.12666188180446625, "learning_rate": 9.155361266220008e-06, "loss": 0.0063, "step": 81410 }, { "grad_norm": 0.10909239202737808, "learning_rate": 9.145826458488254e-06, "loss": 0.0062, "step": 81420 }, { "grad_norm": 0.10730844736099243, "learning_rate": 9.136296118507509e-06, "loss": 0.0056, "step": 81430 }, { "grad_norm": 0.18910863995552063, "learning_rate": 9.126770247319966e-06, "loss": 0.0086, "step": 81440 }, { "grad_norm": 0.1736801415681839, "learning_rate": 9.11724884596738e-06, "loss": 0.0059, "step": 81450 }, { "grad_norm": 0.08600934594869614, "learning_rate": 9.107731915490996e-06, "loss": 0.0054, "step": 81460 }, { "grad_norm": 0.11904904991388321, "learning_rate": 9.098219456931579e-06, "loss": 0.0069, "step": 81470 }, { "grad_norm": 0.18070726096630096, "learning_rate": 9.088711471329376e-06, "loss": 0.0056, "step": 81480 }, { "grad_norm": 0.11797056347131729, "learning_rate": 9.079207959724184e-06, "loss": 0.0066, "step": 81490 }, { "grad_norm": 0.12385400384664536, "learning_rate": 9.069708923155279e-06, "loss": 0.0053, "step": 81500 }, { "grad_norm": 0.19862203299999237, "learning_rate": 9.060214362661468e-06, "loss": 0.0068, "step": 81510 }, { "grad_norm": 0.14194563031196594, "learning_rate": 9.050724279281064e-06, "loss": 0.0064, "step": 81520 }, { "grad_norm": 0.06365935504436493, "learning_rate": 9.041238674051894e-06, "loss": 0.0033, "step": 81530 }, { "grad_norm": 0.11322346329689026, "learning_rate": 9.031757548011277e-06, "loss": 0.0046, "step": 81540 }, { "grad_norm": 0.07243753969669342, "learning_rate": 9.022280902196051e-06, "loss": 0.0069, "step": 81550 }, { "grad_norm": 0.1506889909505844, "learning_rate": 9.012808737642592e-06, "loss": 0.0058, "step": 81560 }, { "grad_norm": 0.04914363846182823, "learning_rate": 9.003341055386739e-06, "loss": 0.0057, "step": 81570 }, { "grad_norm": 0.20554685592651367, "learning_rate": 8.993877856463862e-06, "loss": 0.0055, "step": 81580 }, { "grad_norm": 0.08256776630878448, "learning_rate": 8.984419141908856e-06, "loss": 0.007, "step": 81590 }, { "grad_norm": 0.1401854157447815, "learning_rate": 8.974964912756117e-06, "loss": 0.0078, "step": 81600 }, { "grad_norm": 0.11547761410474777, "learning_rate": 8.965515170039523e-06, "loss": 0.0076, "step": 81610 }, { "grad_norm": 0.11158394813537598, "learning_rate": 8.956069914792514e-06, "loss": 0.006, "step": 81620 }, { "grad_norm": 0.10600466281175613, "learning_rate": 8.946629148047969e-06, "loss": 0.0063, "step": 81630 }, { "grad_norm": 0.066804438829422, "learning_rate": 8.93719287083834e-06, "loss": 0.0069, "step": 81640 }, { "grad_norm": 0.17671851813793182, "learning_rate": 8.927761084195569e-06, "loss": 0.0064, "step": 81650 }, { "grad_norm": 0.08139611035585403, "learning_rate": 8.918333789151096e-06, "loss": 0.0052, "step": 81660 }, { "grad_norm": 0.10890740156173706, "learning_rate": 8.90891098673587e-06, "loss": 0.0068, "step": 81670 }, { "grad_norm": 0.15927958488464355, "learning_rate": 8.89949267798037e-06, "loss": 0.0047, "step": 81680 }, { "grad_norm": 0.17229625582695007, "learning_rate": 8.890078863914542e-06, "loss": 0.0048, "step": 81690 }, { "grad_norm": 0.12107148766517639, "learning_rate": 8.880669545567894e-06, "loss": 0.0055, "step": 81700 }, { "grad_norm": 0.08945160359144211, "learning_rate": 8.871264723969381e-06, "loss": 0.0044, "step": 81710 }, { "grad_norm": 0.10719545185565948, "learning_rate": 8.86186440014754e-06, "loss": 0.0044, "step": 81720 }, { "grad_norm": 0.15663164854049683, "learning_rate": 8.852468575130351e-06, "loss": 0.0046, "step": 81730 }, { "grad_norm": 0.13308055698871613, "learning_rate": 8.843077249945347e-06, "loss": 0.0054, "step": 81740 }, { "grad_norm": 0.08201316744089127, "learning_rate": 8.833690425619523e-06, "loss": 0.0044, "step": 81750 }, { "grad_norm": 0.1272534728050232, "learning_rate": 8.82430810317942e-06, "loss": 0.0063, "step": 81760 }, { "grad_norm": 0.10163339227437973, "learning_rate": 8.814930283651074e-06, "loss": 0.0055, "step": 81770 }, { "grad_norm": 0.0805610939860344, "learning_rate": 8.805556968060048e-06, "loss": 0.0069, "step": 81780 }, { "grad_norm": 0.10125303268432617, "learning_rate": 8.796188157431362e-06, "loss": 0.0065, "step": 81790 }, { "grad_norm": 0.08003664761781693, "learning_rate": 8.786823852789595e-06, "loss": 0.0075, "step": 81800 }, { "grad_norm": 0.15456710755825043, "learning_rate": 8.777464055158813e-06, "loss": 0.0067, "step": 81810 }, { "grad_norm": 0.10456077009439468, "learning_rate": 8.768108765562578e-06, "loss": 0.0063, "step": 81820 }, { "grad_norm": 0.15510715544223785, "learning_rate": 8.758757985023975e-06, "loss": 0.0061, "step": 81830 }, { "grad_norm": 0.08926473557949066, "learning_rate": 8.749411714565598e-06, "loss": 0.0048, "step": 81840 }, { "grad_norm": 0.16897810995578766, "learning_rate": 8.74006995520954e-06, "loss": 0.0059, "step": 81850 }, { "grad_norm": 0.15260830521583557, "learning_rate": 8.730732707977397e-06, "loss": 0.0059, "step": 81860 }, { "grad_norm": 0.15122495591640472, "learning_rate": 8.721399973890281e-06, "loss": 0.007, "step": 81870 }, { "grad_norm": 0.20409536361694336, "learning_rate": 8.712071753968792e-06, "loss": 0.006, "step": 81880 }, { "grad_norm": 0.2951411306858063, "learning_rate": 8.702748049233072e-06, "loss": 0.0078, "step": 81890 }, { "grad_norm": 0.09985551983118057, "learning_rate": 8.693428860702718e-06, "loss": 0.0058, "step": 81900 }, { "grad_norm": 0.162896066904068, "learning_rate": 8.684114189396898e-06, "loss": 0.0058, "step": 81910 }, { "grad_norm": 0.13038194179534912, "learning_rate": 8.674804036334222e-06, "loss": 0.0068, "step": 81920 }, { "grad_norm": 0.09598852694034576, "learning_rate": 8.665498402532857e-06, "loss": 0.0074, "step": 81930 }, { "grad_norm": 0.18186740577220917, "learning_rate": 8.656197289010432e-06, "loss": 0.0077, "step": 81940 }, { "grad_norm": 0.16708877682685852, "learning_rate": 8.646900696784122e-06, "loss": 0.0044, "step": 81950 }, { "grad_norm": 0.10034887492656708, "learning_rate": 8.637608626870553e-06, "loss": 0.0047, "step": 81960 }, { "grad_norm": 0.07793419063091278, "learning_rate": 8.628321080285944e-06, "loss": 0.0082, "step": 81970 }, { "grad_norm": 0.1758103370666504, "learning_rate": 8.619038058045925e-06, "loss": 0.0068, "step": 81980 }, { "grad_norm": 0.09271324425935745, "learning_rate": 8.609759561165704e-06, "loss": 0.0079, "step": 81990 }, { "grad_norm": 0.07145044952630997, "learning_rate": 8.600485590659934e-06, "loss": 0.0048, "step": 82000 }, { "grad_norm": 0.11957128345966339, "learning_rate": 8.591216147542813e-06, "loss": 0.0053, "step": 82010 }, { "grad_norm": 0.07793360203504562, "learning_rate": 8.581951232828045e-06, "loss": 0.0062, "step": 82020 }, { "grad_norm": 0.07205578684806824, "learning_rate": 8.572690847528825e-06, "loss": 0.0044, "step": 82030 }, { "grad_norm": 0.1015315130352974, "learning_rate": 8.563434992657838e-06, "loss": 0.0053, "step": 82040 }, { "grad_norm": 0.0799766331911087, "learning_rate": 8.554183669227307e-06, "loss": 0.0058, "step": 82050 }, { "grad_norm": 0.09184824675321579, "learning_rate": 8.544936878248937e-06, "loss": 0.0054, "step": 82060 }, { "grad_norm": 0.13265199959278107, "learning_rate": 8.535694620733942e-06, "loss": 0.0058, "step": 82070 }, { "grad_norm": 0.1445591002702713, "learning_rate": 8.526456897693053e-06, "loss": 0.0049, "step": 82080 }, { "grad_norm": 0.09105589985847473, "learning_rate": 8.517223710136456e-06, "loss": 0.0053, "step": 82090 }, { "grad_norm": 0.10040202736854553, "learning_rate": 8.507995059073931e-06, "loss": 0.0059, "step": 82100 }, { "grad_norm": 0.29599541425704956, "learning_rate": 8.498770945514673e-06, "loss": 0.0048, "step": 82110 }, { "grad_norm": 0.09019531309604645, "learning_rate": 8.489551370467436e-06, "loss": 0.0058, "step": 82120 }, { "grad_norm": 0.14433430135250092, "learning_rate": 8.480336334940442e-06, "loss": 0.0062, "step": 82130 }, { "grad_norm": 0.11518745124340057, "learning_rate": 8.471125839941458e-06, "loss": 0.005, "step": 82140 }, { "grad_norm": 0.1218668520450592, "learning_rate": 8.461919886477682e-06, "loss": 0.0067, "step": 82150 }, { "grad_norm": 0.07073600590229034, "learning_rate": 8.452718475555927e-06, "loss": 0.0061, "step": 82160 }, { "grad_norm": 0.1592463105916977, "learning_rate": 8.443521608182398e-06, "loss": 0.0075, "step": 82170 }, { "grad_norm": 0.12057983130216599, "learning_rate": 8.43432928536288e-06, "loss": 0.0043, "step": 82180 }, { "grad_norm": 0.1021459773182869, "learning_rate": 8.425141508102607e-06, "loss": 0.0053, "step": 82190 }, { "grad_norm": 0.1009269654750824, "learning_rate": 8.415958277406365e-06, "loss": 0.0064, "step": 82200 }, { "grad_norm": 0.10045124590396881, "learning_rate": 8.406779594278392e-06, "loss": 0.0045, "step": 82210 }, { "grad_norm": 0.08371302485466003, "learning_rate": 8.397605459722468e-06, "loss": 0.0061, "step": 82220 }, { "grad_norm": 0.09541463851928711, "learning_rate": 8.388435874741863e-06, "loss": 0.0043, "step": 82230 }, { "grad_norm": 0.12216462194919586, "learning_rate": 8.37927084033936e-06, "loss": 0.0047, "step": 82240 }, { "grad_norm": 0.1250700205564499, "learning_rate": 8.370110357517208e-06, "loss": 0.0054, "step": 82250 }, { "grad_norm": 0.11372096091508865, "learning_rate": 8.360954427277201e-06, "loss": 0.0059, "step": 82260 }, { "grad_norm": 0.16236603260040283, "learning_rate": 8.351803050620627e-06, "loss": 0.0062, "step": 82270 }, { "grad_norm": 0.14363794028759003, "learning_rate": 8.34265622854824e-06, "loss": 0.0053, "step": 82280 }, { "grad_norm": 0.1270972341299057, "learning_rate": 8.333513962060335e-06, "loss": 0.0056, "step": 82290 }, { "grad_norm": 0.14395077526569366, "learning_rate": 8.324376252156701e-06, "loss": 0.0052, "step": 82300 }, { "grad_norm": 0.08877075463533401, "learning_rate": 8.315243099836633e-06, "loss": 0.0056, "step": 82310 }, { "grad_norm": 0.16110822558403015, "learning_rate": 8.3061145060989e-06, "loss": 0.0047, "step": 82320 }, { "grad_norm": 0.21050195395946503, "learning_rate": 8.296990471941806e-06, "loss": 0.0071, "step": 82330 }, { "grad_norm": 0.19351360201835632, "learning_rate": 8.287870998363112e-06, "loss": 0.0053, "step": 82340 }, { "grad_norm": 0.09233561158180237, "learning_rate": 8.278756086360156e-06, "loss": 0.0046, "step": 82350 }, { "grad_norm": 0.23538567125797272, "learning_rate": 8.269645736929693e-06, "loss": 0.0062, "step": 82360 }, { "grad_norm": 0.09091581404209137, "learning_rate": 8.260539951068048e-06, "loss": 0.0093, "step": 82370 }, { "grad_norm": 0.1643035113811493, "learning_rate": 8.251438729770983e-06, "loss": 0.0061, "step": 82380 }, { "grad_norm": 0.08010862022638321, "learning_rate": 8.242342074033826e-06, "loss": 0.0091, "step": 82390 }, { "grad_norm": 0.1202750876545906, "learning_rate": 8.233249984851344e-06, "loss": 0.0087, "step": 82400 }, { "grad_norm": 0.11406345665454865, "learning_rate": 8.224162463217849e-06, "loss": 0.0038, "step": 82410 }, { "grad_norm": 0.18023797869682312, "learning_rate": 8.215079510127139e-06, "loss": 0.0066, "step": 82420 }, { "grad_norm": 0.15115395188331604, "learning_rate": 8.206001126572521e-06, "loss": 0.0058, "step": 82430 }, { "grad_norm": 0.12799601256847382, "learning_rate": 8.196927313546776e-06, "loss": 0.0059, "step": 82440 }, { "grad_norm": 0.15075454115867615, "learning_rate": 8.18785807204222e-06, "loss": 0.0051, "step": 82450 }, { "grad_norm": 0.09738529473543167, "learning_rate": 8.178793403050627e-06, "loss": 0.004, "step": 82460 }, { "grad_norm": 0.07346034049987793, "learning_rate": 8.16973330756331e-06, "loss": 0.0068, "step": 82470 }, { "grad_norm": 0.14770939946174622, "learning_rate": 8.16067778657107e-06, "loss": 0.0057, "step": 82480 }, { "grad_norm": 0.1502438187599182, "learning_rate": 8.151626841064214e-06, "loss": 0.0087, "step": 82490 }, { "grad_norm": 0.08416976779699326, "learning_rate": 8.142580472032518e-06, "loss": 0.0052, "step": 82500 }, { "grad_norm": 0.13340647518634796, "learning_rate": 8.133538680465285e-06, "loss": 0.0059, "step": 82510 }, { "grad_norm": 0.1279209405183792, "learning_rate": 8.124501467351336e-06, "loss": 0.0072, "step": 82520 }, { "grad_norm": 0.15226519107818604, "learning_rate": 8.11546883367893e-06, "loss": 0.0043, "step": 82530 }, { "grad_norm": 0.1428391933441162, "learning_rate": 8.106440780435882e-06, "loss": 0.0092, "step": 82540 }, { "grad_norm": 0.4342922270298004, "learning_rate": 8.097417308609485e-06, "loss": 0.0091, "step": 82550 }, { "grad_norm": 0.09135505557060242, "learning_rate": 8.088398419186544e-06, "loss": 0.0061, "step": 82560 }, { "grad_norm": 0.14741292595863342, "learning_rate": 8.079384113153321e-06, "loss": 0.0049, "step": 82570 }, { "grad_norm": 0.08839478343725204, "learning_rate": 8.070374391495644e-06, "loss": 0.0041, "step": 82580 }, { "grad_norm": 0.12523885071277618, "learning_rate": 8.061369255198764e-06, "loss": 0.0042, "step": 82590 }, { "grad_norm": 0.16875523328781128, "learning_rate": 8.052368705247493e-06, "loss": 0.0064, "step": 82600 }, { "grad_norm": 0.1398337483406067, "learning_rate": 8.043372742626114e-06, "loss": 0.0051, "step": 82610 }, { "grad_norm": 0.14103654026985168, "learning_rate": 8.034381368318422e-06, "loss": 0.005, "step": 82620 }, { "grad_norm": 0.1107606366276741, "learning_rate": 8.025394583307677e-06, "loss": 0.0066, "step": 82630 }, { "grad_norm": 0.1801975816488266, "learning_rate": 8.016412388576683e-06, "loss": 0.0065, "step": 82640 }, { "grad_norm": 0.15313436090946198, "learning_rate": 8.007434785107699e-06, "loss": 0.0063, "step": 82650 }, { "grad_norm": 0.05389544367790222, "learning_rate": 7.998461773882509e-06, "loss": 0.0058, "step": 82660 }, { "grad_norm": 0.08981457352638245, "learning_rate": 7.989493355882393e-06, "loss": 0.0063, "step": 82670 }, { "grad_norm": 0.07525652647018433, "learning_rate": 7.980529532088137e-06, "loss": 0.0063, "step": 82680 }, { "grad_norm": 0.1256159394979477, "learning_rate": 7.97157030347998e-06, "loss": 0.0064, "step": 82690 }, { "grad_norm": 0.10750018060207367, "learning_rate": 7.96261567103772e-06, "loss": 0.0056, "step": 82700 }, { "grad_norm": 0.10085532069206238, "learning_rate": 7.953665635740593e-06, "loss": 0.0066, "step": 82710 }, { "grad_norm": 0.13006620109081268, "learning_rate": 7.94472019856738e-06, "loss": 0.0051, "step": 82720 }, { "grad_norm": 0.08760608732700348, "learning_rate": 7.935779360496337e-06, "loss": 0.004, "step": 82730 }, { "grad_norm": 0.09901577234268188, "learning_rate": 7.92684312250523e-06, "loss": 0.006, "step": 82740 }, { "grad_norm": 0.16212576627731323, "learning_rate": 7.917911485571289e-06, "loss": 0.0073, "step": 82750 }, { "grad_norm": 0.09238114207983017, "learning_rate": 7.90898445067128e-06, "loss": 0.0041, "step": 82760 }, { "grad_norm": 0.08227362483739853, "learning_rate": 7.900062018781462e-06, "loss": 0.0055, "step": 82770 }, { "grad_norm": 0.2017553597688675, "learning_rate": 7.891144190877553e-06, "loss": 0.0093, "step": 82780 }, { "grad_norm": 0.07524149864912033, "learning_rate": 7.882230967934806e-06, "loss": 0.0053, "step": 82790 }, { "grad_norm": 0.10015721619129181, "learning_rate": 7.873322350927959e-06, "loss": 0.0044, "step": 82800 }, { "grad_norm": 0.11023407429456711, "learning_rate": 7.864418340831253e-06, "loss": 0.0066, "step": 82810 }, { "grad_norm": 0.1307540386915207, "learning_rate": 7.855518938618395e-06, "loss": 0.0048, "step": 82820 }, { "grad_norm": 0.10376781225204468, "learning_rate": 7.846624145262633e-06, "loss": 0.0064, "step": 82830 }, { "grad_norm": 0.09273448586463928, "learning_rate": 7.837733961736672e-06, "loss": 0.0042, "step": 82840 }, { "grad_norm": 0.13058890402317047, "learning_rate": 7.828848389012727e-06, "loss": 0.0035, "step": 82850 }, { "grad_norm": 0.13227756321430206, "learning_rate": 7.819967428062524e-06, "loss": 0.0066, "step": 82860 }, { "grad_norm": 0.13399560749530792, "learning_rate": 7.811091079857274e-06, "loss": 0.0082, "step": 82870 }, { "grad_norm": 0.11288225650787354, "learning_rate": 7.802219345367667e-06, "loss": 0.0052, "step": 82880 }, { "grad_norm": 0.1558918058872223, "learning_rate": 7.793352225563916e-06, "loss": 0.0063, "step": 82890 }, { "grad_norm": 0.1107856035232544, "learning_rate": 7.7844897214157e-06, "loss": 0.0059, "step": 82900 }, { "grad_norm": 0.1067328155040741, "learning_rate": 7.775631833892228e-06, "loss": 0.0065, "step": 82910 }, { "grad_norm": 0.08366771787405014, "learning_rate": 7.76677856396215e-06, "loss": 0.0042, "step": 82920 }, { "grad_norm": 0.16082867980003357, "learning_rate": 7.757929912593697e-06, "loss": 0.0079, "step": 82930 }, { "grad_norm": 0.13902416825294495, "learning_rate": 7.749085880754509e-06, "loss": 0.0053, "step": 82940 }, { "grad_norm": 0.06432578712701797, "learning_rate": 7.740246469411777e-06, "loss": 0.0062, "step": 82950 }, { "grad_norm": 0.15876416862010956, "learning_rate": 7.731411679532142e-06, "loss": 0.0059, "step": 82960 }, { "grad_norm": 0.07309271395206451, "learning_rate": 7.722581512081779e-06, "loss": 0.0063, "step": 82970 }, { "grad_norm": 0.15316906571388245, "learning_rate": 7.713755968026331e-06, "loss": 0.0062, "step": 82980 }, { "grad_norm": 0.07854367047548294, "learning_rate": 7.704935048330969e-06, "loss": 0.0059, "step": 82990 }, { "grad_norm": 0.10199950635433197, "learning_rate": 7.696118753960313e-06, "loss": 0.0059, "step": 83000 }, { "grad_norm": 0.11440114676952362, "learning_rate": 7.687307085878503e-06, "loss": 0.0051, "step": 83010 }, { "grad_norm": 0.09626605361700058, "learning_rate": 7.678500045049186e-06, "loss": 0.0055, "step": 83020 }, { "grad_norm": 0.09458809345960617, "learning_rate": 7.669697632435463e-06, "loss": 0.0054, "step": 83030 }, { "grad_norm": 0.08929648250341415, "learning_rate": 7.660899848999963e-06, "loss": 0.0047, "step": 83040 }, { "grad_norm": 0.15295059978961945, "learning_rate": 7.652106695704803e-06, "loss": 0.0072, "step": 83050 }, { "grad_norm": 0.08346293121576309, "learning_rate": 7.643318173511598e-06, "loss": 0.0039, "step": 83060 }, { "grad_norm": 0.13879279792308807, "learning_rate": 7.634534283381423e-06, "loss": 0.0072, "step": 83070 }, { "grad_norm": 0.09090404957532883, "learning_rate": 7.625755026274889e-06, "loss": 0.0049, "step": 83080 }, { "grad_norm": 0.2092554122209549, "learning_rate": 7.616980403152069e-06, "loss": 0.0072, "step": 83090 }, { "grad_norm": 0.09774854779243469, "learning_rate": 7.608210414972561e-06, "loss": 0.0061, "step": 83100 }, { "grad_norm": 0.11758571863174438, "learning_rate": 7.599445062695404e-06, "loss": 0.0057, "step": 83110 }, { "grad_norm": 0.1019887700676918, "learning_rate": 7.590684347279214e-06, "loss": 0.0051, "step": 83120 }, { "grad_norm": 0.1699976921081543, "learning_rate": 7.581928269682004e-06, "loss": 0.0078, "step": 83130 }, { "grad_norm": 0.2827055752277374, "learning_rate": 7.573176830861361e-06, "loss": 0.0058, "step": 83140 }, { "grad_norm": 0.13668811321258545, "learning_rate": 7.5644300317743e-06, "loss": 0.0064, "step": 83150 }, { "grad_norm": 0.1316378265619278, "learning_rate": 7.555687873377376e-06, "loss": 0.0068, "step": 83160 }, { "grad_norm": 0.1922621726989746, "learning_rate": 7.546950356626592e-06, "loss": 0.0095, "step": 83170 }, { "grad_norm": 0.1400977075099945, "learning_rate": 7.538217482477516e-06, "loss": 0.0057, "step": 83180 }, { "grad_norm": 0.09390247613191605, "learning_rate": 7.529489251885119e-06, "loss": 0.0062, "step": 83190 }, { "grad_norm": 0.07672475278377533, "learning_rate": 7.520765665803941e-06, "loss": 0.0065, "step": 83200 }, { "grad_norm": 0.06321868300437927, "learning_rate": 7.512046725187949e-06, "loss": 0.0059, "step": 83210 }, { "grad_norm": 0.13023175299167633, "learning_rate": 7.503332430990645e-06, "loss": 0.0053, "step": 83220 }, { "grad_norm": 0.11055686324834824, "learning_rate": 7.494622784165017e-06, "loss": 0.0061, "step": 83230 }, { "grad_norm": 0.08677616715431213, "learning_rate": 7.485917785663543e-06, "loss": 0.0108, "step": 83240 }, { "grad_norm": 0.13060136139392853, "learning_rate": 7.477217436438172e-06, "loss": 0.0042, "step": 83250 }, { "grad_norm": 0.15355284512043, "learning_rate": 7.46852173744037e-06, "loss": 0.0052, "step": 83260 }, { "grad_norm": 0.07763225585222244, "learning_rate": 7.4598306896211004e-06, "loss": 0.0062, "step": 83270 }, { "grad_norm": 0.10060293972492218, "learning_rate": 7.451144293930773e-06, "loss": 0.0057, "step": 83280 }, { "grad_norm": 0.09251877665519714, "learning_rate": 7.4424625513193455e-06, "loss": 0.0046, "step": 83290 }, { "grad_norm": 0.08459474891424179, "learning_rate": 7.433785462736209e-06, "loss": 0.0048, "step": 83300 }, { "grad_norm": 0.14418593049049377, "learning_rate": 7.425113029130315e-06, "loss": 0.005, "step": 83310 }, { "grad_norm": 0.14229130744934082, "learning_rate": 7.416445251450044e-06, "loss": 0.0042, "step": 83320 }, { "grad_norm": 0.14323560893535614, "learning_rate": 7.4077821306433035e-06, "loss": 0.0084, "step": 83330 }, { "grad_norm": 0.10182500630617142, "learning_rate": 7.399123667657465e-06, "loss": 0.006, "step": 83340 }, { "grad_norm": 0.10830623656511307, "learning_rate": 7.390469863439426e-06, "loss": 0.0046, "step": 83350 }, { "grad_norm": 0.13181623816490173, "learning_rate": 7.381820718935512e-06, "loss": 0.0045, "step": 83360 }, { "grad_norm": 0.1743851602077484, "learning_rate": 7.373176235091633e-06, "loss": 0.0081, "step": 83370 }, { "grad_norm": 0.19894298911094666, "learning_rate": 7.364536412853101e-06, "loss": 0.0059, "step": 83380 }, { "grad_norm": 0.12913118302822113, "learning_rate": 7.355901253164776e-06, "loss": 0.0064, "step": 83390 }, { "grad_norm": 0.1253805011510849, "learning_rate": 7.34727075697097e-06, "loss": 0.0062, "step": 83400 }, { "grad_norm": 0.0907372459769249, "learning_rate": 7.3386449252155095e-06, "loss": 0.0047, "step": 83410 }, { "grad_norm": 0.12183833122253418, "learning_rate": 7.330023758841692e-06, "loss": 0.0061, "step": 83420 }, { "grad_norm": 0.09597063809633255, "learning_rate": 7.321407258792323e-06, "loss": 0.0073, "step": 83430 }, { "grad_norm": 0.12986496090888977, "learning_rate": 7.312795426009694e-06, "loss": 0.0049, "step": 83440 }, { "grad_norm": 0.10194948315620422, "learning_rate": 7.304188261435585e-06, "loss": 0.0056, "step": 83450 }, { "grad_norm": 0.14601822197437286, "learning_rate": 7.295585766011242e-06, "loss": 0.0046, "step": 83460 }, { "grad_norm": 0.13776224851608276, "learning_rate": 7.286987940677436e-06, "loss": 0.0045, "step": 83470 }, { "grad_norm": 0.13556212186813354, "learning_rate": 7.278394786374426e-06, "loss": 0.0073, "step": 83480 }, { "grad_norm": 0.11476557701826096, "learning_rate": 7.269806304041915e-06, "loss": 0.0047, "step": 83490 }, { "grad_norm": 0.08440499752759933, "learning_rate": 7.261222494619147e-06, "loss": 0.0063, "step": 83500 }, { "grad_norm": 0.09639399498701096, "learning_rate": 7.252643359044825e-06, "loss": 0.0057, "step": 83510 }, { "grad_norm": 0.110355906188488, "learning_rate": 7.244068898257172e-06, "loss": 0.0054, "step": 83520 }, { "grad_norm": 0.07941443473100662, "learning_rate": 7.2354991131938495e-06, "loss": 0.0064, "step": 83530 }, { "grad_norm": 0.11974616348743439, "learning_rate": 7.2269340047920574e-06, "loss": 0.008, "step": 83540 }, { "grad_norm": 0.1528206467628479, "learning_rate": 7.218373573988435e-06, "loss": 0.007, "step": 83550 }, { "grad_norm": 0.1904144138097763, "learning_rate": 7.209817821719178e-06, "loss": 0.0058, "step": 83560 }, { "grad_norm": 0.18388479948043823, "learning_rate": 7.201266748919894e-06, "loss": 0.0086, "step": 83570 }, { "grad_norm": 0.08833631128072739, "learning_rate": 7.19272035652575e-06, "loss": 0.0053, "step": 83580 }, { "grad_norm": 0.10980713367462158, "learning_rate": 7.184178645471334e-06, "loss": 0.0064, "step": 83590 }, { "grad_norm": 0.15445473790168762, "learning_rate": 7.175641616690775e-06, "loss": 0.0063, "step": 83600 }, { "grad_norm": 0.17811137437820435, "learning_rate": 7.167109271117661e-06, "loss": 0.0053, "step": 83610 }, { "grad_norm": 0.1451738029718399, "learning_rate": 7.158581609685072e-06, "loss": 0.0054, "step": 83620 }, { "grad_norm": 0.15207083523273468, "learning_rate": 7.150058633325591e-06, "loss": 0.006, "step": 83630 }, { "grad_norm": 0.1291000097990036, "learning_rate": 7.141540342971287e-06, "loss": 0.0067, "step": 83640 }, { "grad_norm": 0.13703933358192444, "learning_rate": 7.133026739553677e-06, "loss": 0.0073, "step": 83650 }, { "grad_norm": 0.0711144283413887, "learning_rate": 7.124517824003829e-06, "loss": 0.0057, "step": 83660 }, { "grad_norm": 0.08919244259595871, "learning_rate": 7.116013597252236e-06, "loss": 0.0035, "step": 83670 }, { "grad_norm": 0.10320325940847397, "learning_rate": 7.107514060228921e-06, "loss": 0.0043, "step": 83680 }, { "grad_norm": 0.09972696006298065, "learning_rate": 7.099019213863378e-06, "loss": 0.0073, "step": 83690 }, { "grad_norm": 0.18174207210540771, "learning_rate": 7.090529059084611e-06, "loss": 0.0067, "step": 83700 }, { "grad_norm": 0.11699993163347244, "learning_rate": 7.082043596821058e-06, "loss": 0.0054, "step": 83710 }, { "grad_norm": 0.16341130435466766, "learning_rate": 7.07356282800069e-06, "loss": 0.0057, "step": 83720 }, { "grad_norm": 0.10453535616397858, "learning_rate": 7.0650867535509645e-06, "loss": 0.0059, "step": 83730 }, { "grad_norm": 0.09461908042430878, "learning_rate": 7.0566153743987895e-06, "loss": 0.0045, "step": 83740 }, { "grad_norm": 0.11029776185750961, "learning_rate": 7.04814869147059e-06, "loss": 0.0079, "step": 83750 }, { "grad_norm": 0.13639356195926666, "learning_rate": 7.0396867056922725e-06, "loss": 0.0086, "step": 83760 }, { "grad_norm": 0.07754992693662643, "learning_rate": 7.031229417989244e-06, "loss": 0.0046, "step": 83770 }, { "grad_norm": 0.15408478677272797, "learning_rate": 7.022776829286348e-06, "loss": 0.0052, "step": 83780 }, { "grad_norm": 0.12952657043933868, "learning_rate": 7.014328940507975e-06, "loss": 0.005, "step": 83790 }, { "grad_norm": 0.07092409580945969, "learning_rate": 7.005885752577945e-06, "loss": 0.0052, "step": 83800 }, { "grad_norm": 0.20048044621944427, "learning_rate": 6.997447266419615e-06, "loss": 0.0102, "step": 83810 }, { "grad_norm": 0.09466667473316193, "learning_rate": 6.989013482955797e-06, "loss": 0.0047, "step": 83820 }, { "grad_norm": 0.09481306374073029, "learning_rate": 6.980584403108808e-06, "loss": 0.006, "step": 83830 }, { "grad_norm": 0.11728940904140472, "learning_rate": 6.972160027800417e-06, "loss": 0.0057, "step": 83840 }, { "grad_norm": 0.16554617881774902, "learning_rate": 6.963740357951931e-06, "loss": 0.0066, "step": 83850 }, { "grad_norm": 0.13690166175365448, "learning_rate": 6.95532539448408e-06, "loss": 0.005, "step": 83860 }, { "grad_norm": 0.14281520247459412, "learning_rate": 6.946915138317128e-06, "loss": 0.0052, "step": 83870 }, { "grad_norm": 0.09856350719928741, "learning_rate": 6.9385095903708065e-06, "loss": 0.005, "step": 83880 }, { "grad_norm": 0.15604113042354584, "learning_rate": 6.930108751564346e-06, "loss": 0.0058, "step": 83890 }, { "grad_norm": 0.16671152412891388, "learning_rate": 6.921712622816428e-06, "loss": 0.0057, "step": 83900 }, { "grad_norm": 0.12668398022651672, "learning_rate": 6.913321205045259e-06, "loss": 0.0064, "step": 83910 }, { "grad_norm": 0.15255586802959442, "learning_rate": 6.904934499168491e-06, "loss": 0.005, "step": 83920 }, { "grad_norm": 0.097898930311203, "learning_rate": 6.896552506103299e-06, "loss": 0.0046, "step": 83930 }, { "grad_norm": 0.10192513465881348, "learning_rate": 6.888175226766313e-06, "loss": 0.0058, "step": 83940 }, { "grad_norm": 0.08197004348039627, "learning_rate": 6.879802662073686e-06, "loss": 0.0062, "step": 83950 }, { "grad_norm": 0.09515763819217682, "learning_rate": 6.871434812940991e-06, "loss": 0.0051, "step": 83960 }, { "grad_norm": 0.10791058093309402, "learning_rate": 6.863071680283345e-06, "loss": 0.0051, "step": 83970 }, { "grad_norm": 0.11523228138685226, "learning_rate": 6.854713265015339e-06, "loss": 0.0065, "step": 83980 }, { "grad_norm": 0.08549429476261139, "learning_rate": 6.8463595680510025e-06, "loss": 0.0057, "step": 83990 }, { "grad_norm": 0.1376272290945053, "learning_rate": 6.838010590303911e-06, "loss": 0.0042, "step": 84000 }, { "grad_norm": 0.12683464586734772, "learning_rate": 6.829666332687085e-06, "loss": 0.0066, "step": 84010 }, { "grad_norm": 0.1771245300769806, "learning_rate": 6.821326796113059e-06, "loss": 0.0066, "step": 84020 }, { "grad_norm": 0.14003075659275055, "learning_rate": 6.812991981493799e-06, "loss": 0.0046, "step": 84030 }, { "grad_norm": 0.13207025825977325, "learning_rate": 6.804661889740821e-06, "loss": 0.0056, "step": 84040 }, { "grad_norm": 0.06480499356985092, "learning_rate": 6.7963365217650625e-06, "loss": 0.0044, "step": 84050 }, { "grad_norm": 0.10674874484539032, "learning_rate": 6.788015878476983e-06, "loss": 0.008, "step": 84060 }, { "grad_norm": 0.08173665404319763, "learning_rate": 6.779699960786523e-06, "loss": 0.0054, "step": 84070 }, { "grad_norm": 0.11009828001260757, "learning_rate": 6.771388769603104e-06, "loss": 0.0043, "step": 84080 }, { "grad_norm": 0.08705952018499374, "learning_rate": 6.763082305835605e-06, "loss": 0.0048, "step": 84090 }, { "grad_norm": 0.09060489386320114, "learning_rate": 6.754780570392433e-06, "loss": 0.0049, "step": 84100 }, { "grad_norm": 0.12472148984670639, "learning_rate": 6.7464835641814215e-06, "loss": 0.0059, "step": 84110 }, { "grad_norm": 0.07829757034778595, "learning_rate": 6.7381912881099515e-06, "loss": 0.0077, "step": 84120 }, { "grad_norm": 0.09444867819547653, "learning_rate": 6.72990374308482e-06, "loss": 0.0053, "step": 84130 }, { "grad_norm": 0.14294058084487915, "learning_rate": 6.721620930012379e-06, "loss": 0.0055, "step": 84140 }, { "grad_norm": 0.08441556990146637, "learning_rate": 6.713342849798393e-06, "loss": 0.0058, "step": 84150 }, { "grad_norm": 0.08681430667638779, "learning_rate": 6.705069503348166e-06, "loss": 0.0048, "step": 84160 }, { "grad_norm": 0.0839085578918457, "learning_rate": 6.696800891566435e-06, "loss": 0.0052, "step": 84170 }, { "grad_norm": 0.11702141910791397, "learning_rate": 6.688537015357449e-06, "loss": 0.0051, "step": 84180 }, { "grad_norm": 0.07186945527791977, "learning_rate": 6.680277875624941e-06, "loss": 0.0051, "step": 84190 }, { "grad_norm": 0.17042221128940582, "learning_rate": 6.672023473272121e-06, "loss": 0.007, "step": 84200 }, { "grad_norm": 0.0855325311422348, "learning_rate": 6.663773809201667e-06, "loss": 0.0048, "step": 84210 }, { "grad_norm": 0.202472984790802, "learning_rate": 6.655528884315748e-06, "loss": 0.0059, "step": 84220 }, { "grad_norm": 0.1200949102640152, "learning_rate": 6.647288699516036e-06, "loss": 0.005, "step": 84230 }, { "grad_norm": 0.12043110281229019, "learning_rate": 6.639053255703642e-06, "loss": 0.0043, "step": 84240 }, { "grad_norm": 0.11404674500226974, "learning_rate": 6.6308225537791925e-06, "loss": 0.005, "step": 84250 }, { "grad_norm": 0.05312467738986015, "learning_rate": 6.622596594642777e-06, "loss": 0.0058, "step": 84260 }, { "grad_norm": 0.15636099874973297, "learning_rate": 6.614375379193999e-06, "loss": 0.0044, "step": 84270 }, { "grad_norm": 0.08152931183576584, "learning_rate": 6.606158908331889e-06, "loss": 0.0051, "step": 84280 }, { "grad_norm": 0.15456482768058777, "learning_rate": 6.597947182955005e-06, "loss": 0.0064, "step": 84290 }, { "grad_norm": 0.1451176106929779, "learning_rate": 6.589740203961359e-06, "loss": 0.0059, "step": 84300 }, { "grad_norm": 0.09596553444862366, "learning_rate": 6.581537972248464e-06, "loss": 0.0063, "step": 84310 }, { "grad_norm": 0.09958570450544357, "learning_rate": 6.573340488713275e-06, "loss": 0.0067, "step": 84320 }, { "grad_norm": 0.11488307267427444, "learning_rate": 6.565147754252304e-06, "loss": 0.0048, "step": 84330 }, { "grad_norm": 0.1495635211467743, "learning_rate": 6.556959769761462e-06, "loss": 0.0059, "step": 84340 }, { "grad_norm": 0.10753803700208664, "learning_rate": 6.548776536136192e-06, "loss": 0.0073, "step": 84350 }, { "grad_norm": 0.6604401469230652, "learning_rate": 6.54059805427138e-06, "loss": 0.0057, "step": 84360 }, { "grad_norm": 0.11978272348642349, "learning_rate": 6.532424325061437e-06, "loss": 0.0055, "step": 84370 }, { "grad_norm": 0.11491260677576065, "learning_rate": 6.5242553494001935e-06, "loss": 0.0068, "step": 84380 }, { "grad_norm": 0.13478592038154602, "learning_rate": 6.516091128181046e-06, "loss": 0.0046, "step": 84390 }, { "grad_norm": 0.08385500311851501, "learning_rate": 6.507931662296785e-06, "loss": 0.007, "step": 84400 }, { "grad_norm": 0.13005854189395905, "learning_rate": 6.4997769526397354e-06, "loss": 0.0065, "step": 84410 }, { "grad_norm": 0.14387303590774536, "learning_rate": 6.491627000101663e-06, "loss": 0.005, "step": 84420 }, { "grad_norm": 0.07786235213279724, "learning_rate": 6.4834818055738545e-06, "loss": 0.0042, "step": 84430 }, { "grad_norm": 0.09803888201713562, "learning_rate": 6.4753413699470465e-06, "loss": 0.0051, "step": 84440 }, { "grad_norm": 0.09807974845170975, "learning_rate": 6.467205694111478e-06, "loss": 0.0064, "step": 84450 }, { "grad_norm": 0.1332712024450302, "learning_rate": 6.459074778956836e-06, "loss": 0.0066, "step": 84460 }, { "grad_norm": 0.09966450929641724, "learning_rate": 6.450948625372311e-06, "loss": 0.0048, "step": 84470 }, { "grad_norm": 0.12698513269424438, "learning_rate": 6.442827234246585e-06, "loss": 0.0052, "step": 84480 }, { "grad_norm": 0.0678626224398613, "learning_rate": 6.43471060646777e-06, "loss": 0.0061, "step": 84490 }, { "grad_norm": 0.20330661535263062, "learning_rate": 6.426598742923512e-06, "loss": 0.0057, "step": 84500 }, { "grad_norm": 0.09784375131130219, "learning_rate": 6.418491644500885e-06, "loss": 0.0046, "step": 84510 }, { "grad_norm": 0.09061486274003983, "learning_rate": 6.410389312086512e-06, "loss": 0.0055, "step": 84520 }, { "grad_norm": 0.09620289504528046, "learning_rate": 6.4022917465664135e-06, "loss": 0.0062, "step": 84530 }, { "grad_norm": 0.2426600307226181, "learning_rate": 6.3941989488261575e-06, "loss": 0.0082, "step": 84540 }, { "grad_norm": 0.09829603880643845, "learning_rate": 6.386110919750726e-06, "loss": 0.0059, "step": 84550 }, { "grad_norm": 0.09606681764125824, "learning_rate": 6.378027660224651e-06, "loss": 0.0047, "step": 84560 }, { "grad_norm": 0.13997694849967957, "learning_rate": 6.369949171131856e-06, "loss": 0.0049, "step": 84570 }, { "grad_norm": 0.12011747807264328, "learning_rate": 6.361875453355848e-06, "loss": 0.0048, "step": 84580 }, { "grad_norm": 0.10863228887319565, "learning_rate": 6.353806507779525e-06, "loss": 0.0054, "step": 84590 }, { "grad_norm": 0.10663546621799469, "learning_rate": 6.3457423352853044e-06, "loss": 0.0075, "step": 84600 }, { "grad_norm": 0.12064844369888306, "learning_rate": 6.337682936755062e-06, "loss": 0.0053, "step": 84610 }, { "grad_norm": 0.10351090878248215, "learning_rate": 6.329628313070179e-06, "loss": 0.0052, "step": 84620 }, { "grad_norm": 0.19061356782913208, "learning_rate": 6.321578465111477e-06, "loss": 0.0098, "step": 84630 }, { "grad_norm": 0.10478556156158447, "learning_rate": 6.31353339375928e-06, "loss": 0.0053, "step": 84640 }, { "grad_norm": 0.06751630455255508, "learning_rate": 6.305493099893395e-06, "loss": 0.0065, "step": 84650 }, { "grad_norm": 0.15926603972911835, "learning_rate": 6.297457584393096e-06, "loss": 0.0055, "step": 84660 }, { "grad_norm": 0.07175198942422867, "learning_rate": 6.289426848137126e-06, "loss": 0.0063, "step": 84670 }, { "grad_norm": 0.1022363230586052, "learning_rate": 6.281400892003713e-06, "loss": 0.0042, "step": 84680 }, { "grad_norm": 0.12214089184999466, "learning_rate": 6.273379716870581e-06, "loss": 0.0064, "step": 84690 }, { "grad_norm": 0.08818662911653519, "learning_rate": 6.265363323614887e-06, "loss": 0.0066, "step": 84700 }, { "grad_norm": 0.13299956917762756, "learning_rate": 6.257351713113308e-06, "loss": 0.0052, "step": 84710 }, { "grad_norm": 0.08963971585035324, "learning_rate": 6.249344886241981e-06, "loss": 0.0057, "step": 84720 }, { "grad_norm": 0.1167282685637474, "learning_rate": 6.2413428438765295e-06, "loss": 0.007, "step": 84730 }, { "grad_norm": 0.16691094636917114, "learning_rate": 6.233345586892026e-06, "loss": 0.0055, "step": 84740 }, { "grad_norm": 0.10851480811834335, "learning_rate": 6.225353116163052e-06, "loss": 0.0076, "step": 84750 }, { "grad_norm": 0.08134651929140091, "learning_rate": 6.217365432563627e-06, "loss": 0.0061, "step": 84760 }, { "grad_norm": 0.1508580446243286, "learning_rate": 6.209382536967318e-06, "loss": 0.0058, "step": 84770 }, { "grad_norm": 0.16788449883460999, "learning_rate": 6.201404430247082e-06, "loss": 0.008, "step": 84780 }, { "grad_norm": 0.08349442481994629, "learning_rate": 6.1934311132754165e-06, "loss": 0.0057, "step": 84790 }, { "grad_norm": 0.09140469133853912, "learning_rate": 6.185462586924257e-06, "loss": 0.004, "step": 84800 }, { "grad_norm": 0.10691355913877487, "learning_rate": 6.177498852065039e-06, "loss": 0.004, "step": 84810 }, { "grad_norm": 0.14665432274341583, "learning_rate": 6.169539909568655e-06, "loss": 0.0073, "step": 84820 }, { "grad_norm": 0.09864624589681625, "learning_rate": 6.161585760305488e-06, "loss": 0.0049, "step": 84830 }, { "grad_norm": 0.13753177225589752, "learning_rate": 6.153636405145391e-06, "loss": 0.0069, "step": 84840 }, { "grad_norm": 0.13090185821056366, "learning_rate": 6.145691844957702e-06, "loss": 0.0041, "step": 84850 }, { "grad_norm": 0.1088581383228302, "learning_rate": 6.1377520806112145e-06, "loss": 0.0105, "step": 84860 }, { "grad_norm": 0.07561879605054855, "learning_rate": 6.129817112974218e-06, "loss": 0.0053, "step": 84870 }, { "grad_norm": 0.059193991124629974, "learning_rate": 6.121886942914457e-06, "loss": 0.0041, "step": 84880 }, { "grad_norm": 0.1081901267170906, "learning_rate": 6.11396157129917e-06, "loss": 0.0053, "step": 84890 }, { "grad_norm": 0.0966096743941307, "learning_rate": 6.106040998995066e-06, "loss": 0.0046, "step": 84900 }, { "grad_norm": 0.09367072582244873, "learning_rate": 6.098125226868329e-06, "loss": 0.0073, "step": 84910 }, { "grad_norm": 0.09843196719884872, "learning_rate": 6.090214255784604e-06, "loss": 0.0059, "step": 84920 }, { "grad_norm": 0.15377308428287506, "learning_rate": 6.0823080866090346e-06, "loss": 0.0077, "step": 84930 }, { "grad_norm": 0.10035168379545212, "learning_rate": 6.074406720206233e-06, "loss": 0.0053, "step": 84940 }, { "grad_norm": 0.1589987874031067, "learning_rate": 6.066510157440259e-06, "loss": 0.0058, "step": 84950 }, { "grad_norm": 0.10792066901922226, "learning_rate": 6.058618399174676e-06, "loss": 0.0068, "step": 84960 }, { "grad_norm": 0.06211809441447258, "learning_rate": 6.050731446272528e-06, "loss": 0.0044, "step": 84970 }, { "grad_norm": 0.10581846535205841, "learning_rate": 6.042849299596321e-06, "loss": 0.0065, "step": 84980 }, { "grad_norm": 0.06342089176177979, "learning_rate": 6.034971960008013e-06, "loss": 0.0068, "step": 84990 }, { "grad_norm": 0.1005084365606308, "learning_rate": 6.027099428369082e-06, "loss": 0.0071, "step": 85000 }, { "grad_norm": 0.07163078337907791, "learning_rate": 6.019231705540435e-06, "loss": 0.0054, "step": 85010 }, { "grad_norm": 0.20580177009105682, "learning_rate": 6.011368792382488e-06, "loss": 0.0069, "step": 85020 }, { "grad_norm": 0.14148005843162537, "learning_rate": 6.00351068975511e-06, "loss": 0.0081, "step": 85030 }, { "grad_norm": 0.09351636469364166, "learning_rate": 5.9956573985176674e-06, "loss": 0.0071, "step": 85040 }, { "grad_norm": 0.08029378950595856, "learning_rate": 5.987808919528959e-06, "loss": 0.0052, "step": 85050 }, { "grad_norm": 0.13046440482139587, "learning_rate": 5.979965253647307e-06, "loss": 0.0055, "step": 85060 }, { "grad_norm": 0.16070732474327087, "learning_rate": 5.9721264017304655e-06, "loss": 0.0059, "step": 85070 }, { "grad_norm": 0.07043621689081192, "learning_rate": 5.964292364635682e-06, "loss": 0.0057, "step": 85080 }, { "grad_norm": 0.07941088080406189, "learning_rate": 5.956463143219682e-06, "loss": 0.0056, "step": 85090 }, { "grad_norm": 0.08622319251298904, "learning_rate": 5.948638738338663e-06, "loss": 0.0045, "step": 85100 }, { "grad_norm": 0.12306904792785645, "learning_rate": 5.940819150848271e-06, "loss": 0.0054, "step": 85110 }, { "grad_norm": 0.08393708616495132, "learning_rate": 5.933004381603663e-06, "loss": 0.0042, "step": 85120 }, { "grad_norm": 0.12986142933368683, "learning_rate": 5.92519443145943e-06, "loss": 0.0052, "step": 85130 }, { "grad_norm": 0.08214041590690613, "learning_rate": 5.917389301269671e-06, "loss": 0.0041, "step": 85140 }, { "grad_norm": 0.1608695685863495, "learning_rate": 5.9095889918879435e-06, "loss": 0.0061, "step": 85150 }, { "grad_norm": 0.12464332580566406, "learning_rate": 5.901793504167286e-06, "loss": 0.0042, "step": 85160 }, { "grad_norm": 0.15519271790981293, "learning_rate": 5.894002838960172e-06, "loss": 0.0056, "step": 85170 }, { "grad_norm": 0.18268631398677826, "learning_rate": 5.886216997118604e-06, "loss": 0.0055, "step": 85180 }, { "grad_norm": 0.2204969823360443, "learning_rate": 5.878435979494023e-06, "loss": 0.0107, "step": 85190 }, { "grad_norm": 0.1101856678724289, "learning_rate": 5.870659786937344e-06, "loss": 0.0054, "step": 85200 }, { "grad_norm": 0.16268059611320496, "learning_rate": 5.862888420298962e-06, "loss": 0.0059, "step": 85210 }, { "grad_norm": 0.11517437547445297, "learning_rate": 5.855121880428743e-06, "loss": 0.0041, "step": 85220 }, { "grad_norm": 0.09646417945623398, "learning_rate": 5.84736016817603e-06, "loss": 0.0049, "step": 85230 }, { "grad_norm": 0.1276124119758606, "learning_rate": 5.839603284389622e-06, "loss": 0.0062, "step": 85240 }, { "grad_norm": 0.08219454437494278, "learning_rate": 5.831851229917812e-06, "loss": 0.0079, "step": 85250 }, { "grad_norm": 0.1320389062166214, "learning_rate": 5.824104005608338e-06, "loss": 0.0061, "step": 85260 }, { "grad_norm": 0.1363169401884079, "learning_rate": 5.81636161230843e-06, "loss": 0.0055, "step": 85270 }, { "grad_norm": 0.09561403095722198, "learning_rate": 5.808624050864791e-06, "loss": 0.006, "step": 85280 }, { "grad_norm": 0.09543333202600479, "learning_rate": 5.800891322123592e-06, "loss": 0.0042, "step": 85290 }, { "grad_norm": 0.24765703082084656, "learning_rate": 5.793163426930454e-06, "loss": 0.0079, "step": 85300 }, { "grad_norm": 0.11503114551305771, "learning_rate": 5.785440366130512e-06, "loss": 0.0051, "step": 85310 }, { "grad_norm": 0.06814319640398026, "learning_rate": 5.777722140568326e-06, "loss": 0.0053, "step": 85320 }, { "grad_norm": 0.0672900602221489, "learning_rate": 5.7700087510879685e-06, "loss": 0.0055, "step": 85330 }, { "grad_norm": 0.13867071270942688, "learning_rate": 5.762300198532933e-06, "loss": 0.0057, "step": 85340 }, { "grad_norm": 0.1003708466887474, "learning_rate": 5.7545964837462606e-06, "loss": 0.005, "step": 85350 }, { "grad_norm": 0.12542276084423065, "learning_rate": 5.746897607570384e-06, "loss": 0.0046, "step": 85360 }, { "grad_norm": 0.11241419613361359, "learning_rate": 5.7392035708472626e-06, "loss": 0.0045, "step": 85370 }, { "grad_norm": 0.11108580231666565, "learning_rate": 5.731514374418278e-06, "loss": 0.0078, "step": 85380 }, { "grad_norm": 0.1803896427154541, "learning_rate": 5.7238300191243325e-06, "loss": 0.0066, "step": 85390 }, { "grad_norm": 0.15022648870944977, "learning_rate": 5.716150505805762e-06, "loss": 0.0061, "step": 85400 }, { "grad_norm": 0.12830618023872375, "learning_rate": 5.708475835302407e-06, "loss": 0.0064, "step": 85410 }, { "grad_norm": 0.14813823997974396, "learning_rate": 5.70080600845353e-06, "loss": 0.0061, "step": 85420 }, { "grad_norm": 0.08654434233903885, "learning_rate": 5.693141026097909e-06, "loss": 0.0051, "step": 85430 }, { "grad_norm": 0.07691723108291626, "learning_rate": 5.6854808890737745e-06, "loss": 0.0063, "step": 85440 }, { "grad_norm": 0.12512150406837463, "learning_rate": 5.6778255982188225e-06, "loss": 0.0055, "step": 85450 }, { "grad_norm": 0.34900906682014465, "learning_rate": 5.670175154370217e-06, "loss": 0.006, "step": 85460 }, { "grad_norm": 0.07998432219028473, "learning_rate": 5.662529558364616e-06, "loss": 0.0049, "step": 85470 }, { "grad_norm": 0.08769143372774124, "learning_rate": 5.654888811038128e-06, "loss": 0.0067, "step": 85480 }, { "grad_norm": 0.08049331605434418, "learning_rate": 5.647252913226314e-06, "loss": 0.004, "step": 85490 }, { "grad_norm": 0.09092755615711212, "learning_rate": 5.639621865764255e-06, "loss": 0.0048, "step": 85500 }, { "grad_norm": 0.07801056653261185, "learning_rate": 5.63199566948644e-06, "loss": 0.0071, "step": 85510 }, { "grad_norm": 0.05648201331496239, "learning_rate": 5.6243743252268835e-06, "loss": 0.0043, "step": 85520 }, { "grad_norm": 0.12759310007095337, "learning_rate": 5.616757833819009e-06, "loss": 0.0039, "step": 85530 }, { "grad_norm": 0.08847469836473465, "learning_rate": 5.609146196095794e-06, "loss": 0.0052, "step": 85540 }, { "grad_norm": 0.10054440051317215, "learning_rate": 5.6015394128896005e-06, "loss": 0.0069, "step": 85550 }, { "grad_norm": 0.1157984808087349, "learning_rate": 5.593937485032308e-06, "loss": 0.0061, "step": 85560 }, { "grad_norm": 0.08524840325117111, "learning_rate": 5.586340413355246e-06, "loss": 0.0047, "step": 85570 }, { "grad_norm": 0.16746653616428375, "learning_rate": 5.578748198689227e-06, "loss": 0.0054, "step": 85580 }, { "grad_norm": 0.13319775462150574, "learning_rate": 5.571160841864504e-06, "loss": 0.0052, "step": 85590 }, { "grad_norm": 0.13548599183559418, "learning_rate": 5.563578343710846e-06, "loss": 0.0069, "step": 85600 }, { "grad_norm": 0.10536949336528778, "learning_rate": 5.556000705057446e-06, "loss": 0.0048, "step": 85610 }, { "grad_norm": 0.1306893676519394, "learning_rate": 5.548427926732996e-06, "loss": 0.0043, "step": 85620 }, { "grad_norm": 0.12270431965589523, "learning_rate": 5.540860009565629e-06, "loss": 0.01, "step": 85630 }, { "grad_norm": 0.0707523301243782, "learning_rate": 5.533296954382966e-06, "loss": 0.0055, "step": 85640 }, { "grad_norm": 0.12285108864307404, "learning_rate": 5.525738762012095e-06, "loss": 0.0048, "step": 85650 }, { "grad_norm": 0.11607484519481659, "learning_rate": 5.518185433279577e-06, "loss": 0.006, "step": 85660 }, { "grad_norm": 0.09072598069906235, "learning_rate": 5.510636969011407e-06, "loss": 0.0056, "step": 85670 }, { "grad_norm": 0.05803914740681648, "learning_rate": 5.503093370033096e-06, "loss": 0.0053, "step": 85680 }, { "grad_norm": 0.1398344188928604, "learning_rate": 5.495554637169603e-06, "loss": 0.0053, "step": 85690 }, { "grad_norm": 0.12324930727481842, "learning_rate": 5.488020771245328e-06, "loss": 0.0055, "step": 85700 }, { "grad_norm": 0.15304633975028992, "learning_rate": 5.480491773084195e-06, "loss": 0.0051, "step": 85710 }, { "grad_norm": 0.11463551223278046, "learning_rate": 5.472967643509525e-06, "loss": 0.0065, "step": 85720 }, { "grad_norm": 0.11534274369478226, "learning_rate": 5.46544838334419e-06, "loss": 0.0054, "step": 85730 }, { "grad_norm": 0.1493830680847168, "learning_rate": 5.4579339934104525e-06, "loss": 0.0055, "step": 85740 }, { "grad_norm": 0.1377163827419281, "learning_rate": 5.450424474530091e-06, "loss": 0.0051, "step": 85750 }, { "grad_norm": 0.10280955582857132, "learning_rate": 5.442919827524323e-06, "loss": 0.009, "step": 85760 }, { "grad_norm": 0.08385000377893448, "learning_rate": 5.435420053213863e-06, "loss": 0.005, "step": 85770 }, { "grad_norm": 0.11915212869644165, "learning_rate": 5.427925152418845e-06, "loss": 0.0055, "step": 85780 }, { "grad_norm": 0.11529836803674698, "learning_rate": 5.420435125958934e-06, "loss": 0.0057, "step": 85790 }, { "grad_norm": 0.1176796406507492, "learning_rate": 5.412949974653209e-06, "loss": 0.0066, "step": 85800 }, { "grad_norm": 0.10419411212205887, "learning_rate": 5.405469699320248e-06, "loss": 0.0054, "step": 85810 }, { "grad_norm": 0.14193189144134521, "learning_rate": 5.397994300778064e-06, "loss": 0.0059, "step": 85820 }, { "grad_norm": 0.13887962698936462, "learning_rate": 5.39052377984417e-06, "loss": 0.0061, "step": 85830 }, { "grad_norm": 0.058149732649326324, "learning_rate": 5.383058137335523e-06, "loss": 0.005, "step": 85840 }, { "grad_norm": 0.08091852813959122, "learning_rate": 5.375597374068553e-06, "loss": 0.0057, "step": 85850 }, { "grad_norm": 0.10839652270078659, "learning_rate": 5.368141490859163e-06, "loss": 0.0064, "step": 85860 }, { "grad_norm": 0.13756109774112701, "learning_rate": 5.360690488522729e-06, "loss": 0.0048, "step": 85870 }, { "grad_norm": 0.1619599461555481, "learning_rate": 5.353244367874055e-06, "loss": 0.0076, "step": 85880 }, { "grad_norm": 0.12116116285324097, "learning_rate": 5.345803129727456e-06, "loss": 0.0038, "step": 85890 }, { "grad_norm": 0.10946542024612427, "learning_rate": 5.338366774896697e-06, "loss": 0.0049, "step": 85900 }, { "grad_norm": 0.11452748626470566, "learning_rate": 5.330935304194995e-06, "loss": 0.0072, "step": 85910 }, { "grad_norm": 0.10302946716547012, "learning_rate": 5.323508718435044e-06, "loss": 0.0058, "step": 85920 }, { "grad_norm": 0.14623650908470154, "learning_rate": 5.316087018429011e-06, "loss": 0.0067, "step": 85930 }, { "grad_norm": 0.08317901194095612, "learning_rate": 5.30867020498853e-06, "loss": 0.005, "step": 85940 }, { "grad_norm": 0.09450103342533112, "learning_rate": 5.30125827892467e-06, "loss": 0.0078, "step": 85950 }, { "grad_norm": 0.07531613111495972, "learning_rate": 5.293851241048015e-06, "loss": 0.005, "step": 85960 }, { "grad_norm": 0.11054614931344986, "learning_rate": 5.286449092168555e-06, "loss": 0.0048, "step": 85970 }, { "grad_norm": 0.18037937581539154, "learning_rate": 5.2790518330958105e-06, "loss": 0.0045, "step": 85980 }, { "grad_norm": 0.08617829531431198, "learning_rate": 5.271659464638717e-06, "loss": 0.005, "step": 85990 }, { "grad_norm": 0.1461467146873474, "learning_rate": 5.264271987605701e-06, "loss": 0.0066, "step": 86000 }, { "grad_norm": 0.12922924757003784, "learning_rate": 5.256889402804632e-06, "loss": 0.0062, "step": 86010 }, { "grad_norm": 0.11167238652706146, "learning_rate": 5.249511711042881e-06, "loss": 0.0051, "step": 86020 }, { "grad_norm": 0.15973341464996338, "learning_rate": 5.242138913127237e-06, "loss": 0.0054, "step": 86030 }, { "grad_norm": 0.11416463553905487, "learning_rate": 5.234771009863987e-06, "loss": 0.0049, "step": 86040 }, { "grad_norm": 0.11908527463674545, "learning_rate": 5.227408002058881e-06, "loss": 0.0046, "step": 86050 }, { "grad_norm": 0.08928661048412323, "learning_rate": 5.220049890517126e-06, "loss": 0.005, "step": 86060 }, { "grad_norm": 0.11055435985326767, "learning_rate": 5.212696676043377e-06, "loss": 0.0057, "step": 86070 }, { "grad_norm": 0.08622459322214127, "learning_rate": 5.205348359441798e-06, "loss": 0.0058, "step": 86080 }, { "grad_norm": 0.06928566843271255, "learning_rate": 5.198004941515966e-06, "loss": 0.0047, "step": 86090 }, { "grad_norm": 0.11653047055006027, "learning_rate": 5.190666423068952e-06, "loss": 0.006, "step": 86100 }, { "grad_norm": 0.083541139960289, "learning_rate": 5.183332804903284e-06, "loss": 0.0044, "step": 86110 }, { "grad_norm": 0.19923347234725952, "learning_rate": 5.176004087820979e-06, "loss": 0.0053, "step": 86120 }, { "grad_norm": 0.15467774868011475, "learning_rate": 5.168680272623455e-06, "loss": 0.0091, "step": 86130 }, { "grad_norm": 0.20862475037574768, "learning_rate": 5.161361360111661e-06, "loss": 0.0077, "step": 86140 }, { "grad_norm": 0.10411769151687622, "learning_rate": 5.154047351085984e-06, "loss": 0.005, "step": 86150 }, { "grad_norm": 0.11139943450689316, "learning_rate": 5.1467382463462385e-06, "loss": 0.0049, "step": 86160 }, { "grad_norm": 0.22217997908592224, "learning_rate": 5.139434046691788e-06, "loss": 0.0087, "step": 86170 }, { "grad_norm": 0.07899452745914459, "learning_rate": 5.13213475292137e-06, "loss": 0.0042, "step": 86180 }, { "grad_norm": 0.12043342739343643, "learning_rate": 5.12484036583325e-06, "loss": 0.0042, "step": 86190 }, { "grad_norm": 0.09024585783481598, "learning_rate": 5.117550886225109e-06, "loss": 0.0048, "step": 86200 }, { "grad_norm": 0.06884284317493439, "learning_rate": 5.11026631489413e-06, "loss": 0.0053, "step": 86210 }, { "grad_norm": 0.0865381509065628, "learning_rate": 5.102986652636926e-06, "loss": 0.0053, "step": 86220 }, { "grad_norm": 0.16794803738594055, "learning_rate": 5.095711900249606e-06, "loss": 0.0067, "step": 86230 }, { "grad_norm": 0.10068754851818085, "learning_rate": 5.088442058527715e-06, "loss": 0.0046, "step": 86240 }, { "grad_norm": 0.07848448306322098, "learning_rate": 5.08117712826629e-06, "loss": 0.0081, "step": 86250 }, { "grad_norm": 0.1276603788137436, "learning_rate": 5.073917110259785e-06, "loss": 0.0043, "step": 86260 }, { "grad_norm": 0.05670863389968872, "learning_rate": 5.066662005302175e-06, "loss": 0.0053, "step": 86270 }, { "grad_norm": 0.08027155697345734, "learning_rate": 5.059411814186838e-06, "loss": 0.0037, "step": 86280 }, { "grad_norm": 0.14292433857917786, "learning_rate": 5.05216653770666e-06, "loss": 0.0068, "step": 86290 }, { "grad_norm": 0.05818893387913704, "learning_rate": 5.044926176653969e-06, "loss": 0.0053, "step": 86300 }, { "grad_norm": 0.06453665345907211, "learning_rate": 5.037690731820571e-06, "loss": 0.0037, "step": 86310 }, { "grad_norm": 0.1189759150147438, "learning_rate": 5.030460203997711e-06, "loss": 0.0046, "step": 86320 }, { "grad_norm": 0.057845454663038254, "learning_rate": 5.0232345939761174e-06, "loss": 0.0055, "step": 86330 }, { "grad_norm": 0.17703357338905334, "learning_rate": 5.016013902545957e-06, "loss": 0.0069, "step": 86340 }, { "grad_norm": 0.25000378489494324, "learning_rate": 5.008798130496889e-06, "loss": 0.0065, "step": 86350 }, { "grad_norm": 0.11404381692409515, "learning_rate": 5.001587278618008e-06, "loss": 0.0046, "step": 86360 }, { "grad_norm": 0.12081750482320786, "learning_rate": 4.994381347697891e-06, "loss": 0.005, "step": 86370 }, { "grad_norm": 0.18866509199142456, "learning_rate": 4.987180338524572e-06, "loss": 0.0067, "step": 86380 }, { "grad_norm": 0.1403425633907318, "learning_rate": 4.9799842518855275e-06, "loss": 0.0048, "step": 86390 }, { "grad_norm": 0.11530708521604538, "learning_rate": 4.9727930885677285e-06, "loss": 0.0054, "step": 86400 }, { "grad_norm": 0.10101468861103058, "learning_rate": 4.965606849357573e-06, "loss": 0.0056, "step": 86410 }, { "grad_norm": 0.10676029324531555, "learning_rate": 4.9584255350409445e-06, "loss": 0.0056, "step": 86420 }, { "grad_norm": 0.11195860803127289, "learning_rate": 4.951249146403175e-06, "loss": 0.0038, "step": 86430 }, { "grad_norm": 0.18568465113639832, "learning_rate": 4.9440776842290815e-06, "loss": 0.01, "step": 86440 }, { "grad_norm": 0.100306935608387, "learning_rate": 4.936911149302903e-06, "loss": 0.0048, "step": 86450 }, { "grad_norm": 0.08285939693450928, "learning_rate": 4.929749542408379e-06, "loss": 0.0111, "step": 86460 }, { "grad_norm": 0.10006241500377655, "learning_rate": 4.922592864328673e-06, "loss": 0.0058, "step": 86470 }, { "grad_norm": 0.08745522797107697, "learning_rate": 4.915441115846436e-06, "loss": 0.0064, "step": 86480 }, { "grad_norm": 0.13288825750350952, "learning_rate": 4.908294297743782e-06, "loss": 0.0051, "step": 86490 }, { "grad_norm": 0.23276017606258392, "learning_rate": 4.901152410802273e-06, "loss": 0.0076, "step": 86500 }, { "grad_norm": 0.06513789296150208, "learning_rate": 4.894015455802919e-06, "loss": 0.0058, "step": 86510 }, { "grad_norm": 0.1024528220295906, "learning_rate": 4.8868834335262346e-06, "loss": 0.0047, "step": 86520 }, { "grad_norm": 0.13848604261875153, "learning_rate": 4.87975634475214e-06, "loss": 0.0059, "step": 86530 }, { "grad_norm": 0.09278055280447006, "learning_rate": 4.8726341902600614e-06, "loss": 0.0045, "step": 86540 }, { "grad_norm": 0.08032041788101196, "learning_rate": 4.865516970828843e-06, "loss": 0.0044, "step": 86550 }, { "grad_norm": 0.11453988403081894, "learning_rate": 4.85840468723685e-06, "loss": 0.0052, "step": 86560 }, { "grad_norm": 0.141050785779953, "learning_rate": 4.85129734026184e-06, "loss": 0.0055, "step": 86570 }, { "grad_norm": 0.12228917330503464, "learning_rate": 4.844194930681073e-06, "loss": 0.0051, "step": 86580 }, { "grad_norm": 0.1769857108592987, "learning_rate": 4.837097459271267e-06, "loss": 0.0062, "step": 86590 }, { "grad_norm": 0.07534031569957733, "learning_rate": 4.830004926808573e-06, "loss": 0.0062, "step": 86600 }, { "grad_norm": 0.1541818082332611, "learning_rate": 4.822917334068627e-06, "loss": 0.005, "step": 86610 }, { "grad_norm": 0.19236403703689575, "learning_rate": 4.81583468182652e-06, "loss": 0.0054, "step": 86620 }, { "grad_norm": 0.09119649231433868, "learning_rate": 4.808756970856809e-06, "loss": 0.004, "step": 86630 }, { "grad_norm": 0.08982902020215988, "learning_rate": 4.801684201933476e-06, "loss": 0.0056, "step": 86640 }, { "grad_norm": 0.24406179785728455, "learning_rate": 4.794616375830019e-06, "loss": 0.0087, "step": 86650 }, { "grad_norm": 0.17250271141529083, "learning_rate": 4.787553493319341e-06, "loss": 0.0067, "step": 86660 }, { "grad_norm": 0.08015795052051544, "learning_rate": 4.780495555173836e-06, "loss": 0.005, "step": 86670 }, { "grad_norm": 0.12210648506879807, "learning_rate": 4.7734425621653475e-06, "loss": 0.0074, "step": 86680 }, { "grad_norm": 0.10622753202915192, "learning_rate": 4.766394515065198e-06, "loss": 0.0049, "step": 86690 }, { "grad_norm": 0.13541939854621887, "learning_rate": 4.75935141464412e-06, "loss": 0.0076, "step": 86700 }, { "grad_norm": 0.1380060315132141, "learning_rate": 4.75231326167237e-06, "loss": 0.0049, "step": 86710 }, { "grad_norm": 0.1162339299917221, "learning_rate": 4.745280056919599e-06, "loss": 0.0072, "step": 86720 }, { "grad_norm": 0.06590885668992996, "learning_rate": 4.738251801154975e-06, "loss": 0.0065, "step": 86730 }, { "grad_norm": 0.2733510136604309, "learning_rate": 4.731228495147061e-06, "loss": 0.0051, "step": 86740 }, { "grad_norm": 0.08111369609832764, "learning_rate": 4.7242101396639595e-06, "loss": 0.0069, "step": 86750 }, { "grad_norm": 0.08455152809619904, "learning_rate": 4.717196735473151e-06, "loss": 0.0046, "step": 86760 }, { "grad_norm": 0.10295772552490234, "learning_rate": 4.710188283341643e-06, "loss": 0.0083, "step": 86770 }, { "grad_norm": 0.12724098563194275, "learning_rate": 4.703184784035835e-06, "loss": 0.0044, "step": 86780 }, { "grad_norm": 0.08124963194131851, "learning_rate": 4.696186238321642e-06, "loss": 0.0075, "step": 86790 }, { "grad_norm": 0.07191993296146393, "learning_rate": 4.689192646964408e-06, "loss": 0.0056, "step": 86800 }, { "grad_norm": 0.12992389500141144, "learning_rate": 4.682204010728952e-06, "loss": 0.0071, "step": 86810 }, { "grad_norm": 0.09123285859823227, "learning_rate": 4.67522033037952e-06, "loss": 0.0055, "step": 86820 }, { "grad_norm": 0.18835070729255676, "learning_rate": 4.668241606679846e-06, "loss": 0.0087, "step": 86830 }, { "grad_norm": 0.0720619335770607, "learning_rate": 4.66126784039313e-06, "loss": 0.0085, "step": 86840 }, { "grad_norm": 0.08621066808700562, "learning_rate": 4.6542990322819836e-06, "loss": 0.0067, "step": 86850 }, { "grad_norm": 0.10568788647651672, "learning_rate": 4.64733518310852e-06, "loss": 0.0056, "step": 86860 }, { "grad_norm": 0.12458032369613647, "learning_rate": 4.640376293634291e-06, "loss": 0.0047, "step": 86870 }, { "grad_norm": 0.09672340005636215, "learning_rate": 4.633422364620327e-06, "loss": 0.0064, "step": 86880 }, { "grad_norm": 0.23651844263076782, "learning_rate": 4.6264733968270745e-06, "loss": 0.006, "step": 86890 }, { "grad_norm": 0.39715972542762756, "learning_rate": 4.619529391014482e-06, "loss": 0.005, "step": 86900 }, { "grad_norm": 0.13651368021965027, "learning_rate": 4.61259034794192e-06, "loss": 0.0036, "step": 86910 }, { "grad_norm": 0.2129959613084793, "learning_rate": 4.605656268368247e-06, "loss": 0.007, "step": 86920 }, { "grad_norm": 0.47182753682136536, "learning_rate": 4.598727153051735e-06, "loss": 0.0046, "step": 86930 }, { "grad_norm": 0.08652546256780624, "learning_rate": 4.591803002750178e-06, "loss": 0.006, "step": 86940 }, { "grad_norm": 0.11176470667123795, "learning_rate": 4.58488381822077e-06, "loss": 0.0057, "step": 86950 }, { "grad_norm": 0.05277964845299721, "learning_rate": 4.5779696002201946e-06, "loss": 0.0068, "step": 86960 }, { "grad_norm": 0.12151270359754562, "learning_rate": 4.571060349504563e-06, "loss": 0.0053, "step": 86970 }, { "grad_norm": 0.11180000752210617, "learning_rate": 4.564156066829478e-06, "loss": 0.0049, "step": 86980 }, { "grad_norm": 0.25682902336120605, "learning_rate": 4.557256752949957e-06, "loss": 0.0056, "step": 86990 }, { "grad_norm": 0.13006536662578583, "learning_rate": 4.550362408620534e-06, "loss": 0.006, "step": 87000 }, { "grad_norm": 0.06748034060001373, "learning_rate": 4.543473034595136e-06, "loss": 0.005, "step": 87010 }, { "grad_norm": 0.05856294557452202, "learning_rate": 4.536588631627192e-06, "loss": 0.0038, "step": 87020 }, { "grad_norm": 0.07035387307405472, "learning_rate": 4.5297092004695565e-06, "loss": 0.0043, "step": 87030 }, { "grad_norm": 0.18170857429504395, "learning_rate": 4.522834741874555e-06, "loss": 0.0075, "step": 87040 }, { "grad_norm": 0.10936816036701202, "learning_rate": 4.515965256593974e-06, "loss": 0.0051, "step": 87050 }, { "grad_norm": 0.12916921079158783, "learning_rate": 4.509100745379058e-06, "loss": 0.0065, "step": 87060 }, { "grad_norm": 0.1599046140909195, "learning_rate": 4.502241208980479e-06, "loss": 0.0065, "step": 87070 }, { "grad_norm": 0.29660314321517944, "learning_rate": 4.495386648148398e-06, "loss": 0.004, "step": 87080 }, { "grad_norm": 0.09891242533922195, "learning_rate": 4.488537063632431e-06, "loss": 0.0042, "step": 87090 }, { "grad_norm": 0.08078323304653168, "learning_rate": 4.481692456181608e-06, "loss": 0.0054, "step": 87100 }, { "grad_norm": 0.14242395758628845, "learning_rate": 4.474852826544479e-06, "loss": 0.0063, "step": 87110 }, { "grad_norm": 0.08193130046129227, "learning_rate": 4.468018175468974e-06, "loss": 0.0048, "step": 87120 }, { "grad_norm": 0.08514546602964401, "learning_rate": 4.461188503702568e-06, "loss": 0.005, "step": 87130 }, { "grad_norm": 0.12174724042415619, "learning_rate": 4.454363811992107e-06, "loss": 0.0057, "step": 87140 }, { "grad_norm": 0.07379594445228577, "learning_rate": 4.447544101083956e-06, "loss": 0.0045, "step": 87150 }, { "grad_norm": 0.1927379071712494, "learning_rate": 4.440729371723884e-06, "loss": 0.0051, "step": 87160 }, { "grad_norm": 0.25734224915504456, "learning_rate": 4.433919624657162e-06, "loss": 0.0053, "step": 87170 }, { "grad_norm": 0.28938573598861694, "learning_rate": 4.427114860628462e-06, "loss": 0.0051, "step": 87180 }, { "grad_norm": 0.10550493746995926, "learning_rate": 4.420315080381982e-06, "loss": 0.0062, "step": 87190 }, { "grad_norm": 0.10540855675935745, "learning_rate": 4.4135202846613045e-06, "loss": 0.0066, "step": 87200 }, { "grad_norm": 0.2913520038127899, "learning_rate": 4.406730474209525e-06, "loss": 0.0062, "step": 87210 }, { "grad_norm": 0.0715063139796257, "learning_rate": 4.399945649769138e-06, "loss": 0.0051, "step": 87220 }, { "grad_norm": 0.08908440917730331, "learning_rate": 4.393165812082151e-06, "loss": 0.0046, "step": 87230 }, { "grad_norm": 0.07625389099121094, "learning_rate": 4.386390961889969e-06, "loss": 0.0054, "step": 87240 }, { "grad_norm": 0.14000675082206726, "learning_rate": 4.3796210999334895e-06, "loss": 0.0047, "step": 87250 }, { "grad_norm": 0.09799028933048248, "learning_rate": 4.372856226953065e-06, "loss": 0.0043, "step": 87260 }, { "grad_norm": 0.10047052055597305, "learning_rate": 4.366096343688486e-06, "loss": 0.0056, "step": 87270 }, { "grad_norm": 0.06629251688718796, "learning_rate": 4.359341450878995e-06, "loss": 0.006, "step": 87280 }, { "grad_norm": 0.10022755712270737, "learning_rate": 4.352591549263302e-06, "loss": 0.0052, "step": 87290 }, { "grad_norm": 0.09331239759922028, "learning_rate": 4.345846639579582e-06, "loss": 0.0068, "step": 87300 }, { "grad_norm": 0.08068350702524185, "learning_rate": 4.339106722565417e-06, "loss": 0.0041, "step": 87310 }, { "grad_norm": 0.0792110413312912, "learning_rate": 4.332371798957896e-06, "loss": 0.0067, "step": 87320 }, { "grad_norm": 0.08974834531545639, "learning_rate": 4.3256418694935295e-06, "loss": 0.0045, "step": 87330 }, { "grad_norm": 0.0692405179142952, "learning_rate": 4.3189169349083125e-06, "loss": 0.0051, "step": 87340 }, { "grad_norm": 0.11371210217475891, "learning_rate": 4.312196995937645e-06, "loss": 0.0058, "step": 87350 }, { "grad_norm": 0.09169352799654007, "learning_rate": 4.305482053316434e-06, "loss": 0.0047, "step": 87360 }, { "grad_norm": 0.4404505789279938, "learning_rate": 4.2987721077789925e-06, "loss": 0.0076, "step": 87370 }, { "grad_norm": 0.13323168456554413, "learning_rate": 4.2920671600591325e-06, "loss": 0.0053, "step": 87380 }, { "grad_norm": 0.0674438625574112, "learning_rate": 4.285367210890084e-06, "loss": 0.0056, "step": 87390 }, { "grad_norm": 0.10900933295488358, "learning_rate": 4.278672261004551e-06, "loss": 0.007, "step": 87400 }, { "grad_norm": 0.08634766936302185, "learning_rate": 4.2719823111346704e-06, "loss": 0.006, "step": 87410 }, { "grad_norm": 0.1307460069656372, "learning_rate": 4.265297362012061e-06, "loss": 0.0066, "step": 87420 }, { "grad_norm": 0.10730817914009094, "learning_rate": 4.258617414367766e-06, "loss": 0.0055, "step": 87430 }, { "grad_norm": 0.1377980262041092, "learning_rate": 4.251942468932296e-06, "loss": 0.006, "step": 87440 }, { "grad_norm": 0.06694384664297104, "learning_rate": 4.2452725264356175e-06, "loss": 0.0038, "step": 87450 }, { "grad_norm": 0.10506238043308258, "learning_rate": 4.2386075876071575e-06, "loss": 0.0046, "step": 87460 }, { "grad_norm": 0.15094386041164398, "learning_rate": 4.231947653175755e-06, "loss": 0.0047, "step": 87470 }, { "grad_norm": 0.13833685219287872, "learning_rate": 4.225292723869761e-06, "loss": 0.0054, "step": 87480 }, { "grad_norm": 0.04351940006017685, "learning_rate": 4.218642800416916e-06, "loss": 0.0046, "step": 87490 }, { "grad_norm": 0.12123885750770569, "learning_rate": 4.211997883544472e-06, "loss": 0.0044, "step": 87500 }, { "grad_norm": 0.1021096333861351, "learning_rate": 4.205357973979096e-06, "loss": 0.0042, "step": 87510 }, { "grad_norm": 0.26933085918426514, "learning_rate": 4.198723072446925e-06, "loss": 0.0057, "step": 87520 }, { "grad_norm": 0.08920569717884064, "learning_rate": 4.192093179673534e-06, "loss": 0.0043, "step": 87530 }, { "grad_norm": 0.10514718294143677, "learning_rate": 4.185468296383959e-06, "loss": 0.004, "step": 87540 }, { "grad_norm": 0.15547893941402435, "learning_rate": 4.1788484233027005e-06, "loss": 0.0056, "step": 87550 }, { "grad_norm": 0.10904590785503387, "learning_rate": 4.172233561153677e-06, "loss": 0.0061, "step": 87560 }, { "grad_norm": 0.11667252331972122, "learning_rate": 4.165623710660288e-06, "loss": 0.0058, "step": 87570 }, { "grad_norm": 0.1047414243221283, "learning_rate": 4.1590188725453835e-06, "loss": 0.0056, "step": 87580 }, { "grad_norm": 0.13345849514007568, "learning_rate": 4.152419047531259e-06, "loss": 0.0048, "step": 87590 }, { "grad_norm": 0.07812352478504181, "learning_rate": 4.145824236339651e-06, "loss": 0.0068, "step": 87600 }, { "grad_norm": 0.14010463654994965, "learning_rate": 4.1392344396917695e-06, "loss": 0.0036, "step": 87610 }, { "grad_norm": 0.134598046541214, "learning_rate": 4.132649658308252e-06, "loss": 0.0039, "step": 87620 }, { "grad_norm": 0.10541100054979324, "learning_rate": 4.126069892909213e-06, "loss": 0.007, "step": 87630 }, { "grad_norm": 0.07910291105508804, "learning_rate": 4.119495144214191e-06, "loss": 0.0057, "step": 87640 }, { "grad_norm": 0.13760003447532654, "learning_rate": 4.112925412942214e-06, "loss": 0.0047, "step": 87650 }, { "grad_norm": 0.09615472704172134, "learning_rate": 4.106360699811712e-06, "loss": 0.0055, "step": 87660 }, { "grad_norm": 0.06664178520441055, "learning_rate": 4.099801005540616e-06, "loss": 0.0052, "step": 87670 }, { "grad_norm": 0.12063942104578018, "learning_rate": 4.093246330846256e-06, "loss": 0.0047, "step": 87680 }, { "grad_norm": 0.13728603720664978, "learning_rate": 4.086696676445467e-06, "loss": 0.0055, "step": 87690 }, { "grad_norm": 0.08692807704210281, "learning_rate": 4.0801520430544914e-06, "loss": 0.0042, "step": 87700 }, { "grad_norm": 0.09317969530820847, "learning_rate": 4.073612431389062e-06, "loss": 0.0062, "step": 87710 }, { "grad_norm": 0.10325734317302704, "learning_rate": 4.067077842164319e-06, "loss": 0.006, "step": 87720 }, { "grad_norm": 0.12032417207956314, "learning_rate": 4.0605482760948945e-06, "loss": 0.0065, "step": 87730 }, { "grad_norm": 0.10149934887886047, "learning_rate": 4.054023733894829e-06, "loss": 0.0061, "step": 87740 }, { "grad_norm": 0.10438293218612671, "learning_rate": 4.047504216277653e-06, "loss": 0.0058, "step": 87750 }, { "grad_norm": 0.11717937886714935, "learning_rate": 4.040989723956323e-06, "loss": 0.0065, "step": 87760 }, { "grad_norm": 0.0869583711028099, "learning_rate": 4.03448025764327e-06, "loss": 0.0046, "step": 87770 }, { "grad_norm": 0.16517281532287598, "learning_rate": 4.027975818050339e-06, "loss": 0.0036, "step": 87780 }, { "grad_norm": 0.1406865268945694, "learning_rate": 4.021476405888858e-06, "loss": 0.0058, "step": 87790 }, { "grad_norm": 0.12448229640722275, "learning_rate": 4.01498202186959e-06, "loss": 0.0055, "step": 87800 }, { "grad_norm": 0.09968321025371552, "learning_rate": 4.008492666702751e-06, "loss": 0.0063, "step": 87810 }, { "grad_norm": 0.12316089123487473, "learning_rate": 4.002008341098001e-06, "loss": 0.0039, "step": 87820 }, { "grad_norm": 0.12446888536214828, "learning_rate": 3.995529045764462e-06, "loss": 0.0046, "step": 87830 }, { "grad_norm": 0.16351400315761566, "learning_rate": 3.989054781410712e-06, "loss": 0.0049, "step": 87840 }, { "grad_norm": 0.12064406275749207, "learning_rate": 3.982585548744744e-06, "loss": 0.0048, "step": 87850 }, { "grad_norm": 0.12210641801357269, "learning_rate": 3.9761213484740435e-06, "loss": 0.0039, "step": 87860 }, { "grad_norm": 0.10565680265426636, "learning_rate": 3.9696621813055055e-06, "loss": 0.0061, "step": 87870 }, { "grad_norm": 0.12210027128458023, "learning_rate": 3.9632080479455044e-06, "loss": 0.005, "step": 87880 }, { "grad_norm": 0.08795889467000961, "learning_rate": 3.956758949099859e-06, "loss": 0.0083, "step": 87890 }, { "grad_norm": 0.1243944987654686, "learning_rate": 3.9503148854738385e-06, "loss": 0.0071, "step": 87900 }, { "grad_norm": 0.0910310223698616, "learning_rate": 3.943875857772139e-06, "loss": 0.0062, "step": 87910 }, { "grad_norm": 0.11715731024742126, "learning_rate": 3.9374418666989375e-06, "loss": 0.0044, "step": 87920 }, { "grad_norm": 0.10695594549179077, "learning_rate": 3.931012912957838e-06, "loss": 0.0046, "step": 87930 }, { "grad_norm": 0.10703305900096893, "learning_rate": 3.9245889972519045e-06, "loss": 0.0038, "step": 87940 }, { "grad_norm": 0.07276339828968048, "learning_rate": 3.9181701202836265e-06, "loss": 0.0063, "step": 87950 }, { "grad_norm": 0.11092645674943924, "learning_rate": 3.911756282755003e-06, "loss": 0.005, "step": 87960 }, { "grad_norm": 0.07523337751626968, "learning_rate": 3.905347485367411e-06, "loss": 0.0071, "step": 87970 }, { "grad_norm": 0.09841467440128326, "learning_rate": 3.898943728821725e-06, "loss": 0.0051, "step": 87980 }, { "grad_norm": 0.07825246453285217, "learning_rate": 3.892545013818227e-06, "loss": 0.0041, "step": 87990 }, { "grad_norm": 0.14018608629703522, "learning_rate": 3.886151341056687e-06, "loss": 0.0065, "step": 88000 }, { "grad_norm": 0.09602867066860199, "learning_rate": 3.8797627112363066e-06, "loss": 0.0042, "step": 88010 }, { "grad_norm": 0.11407681554555893, "learning_rate": 3.873379125055743e-06, "loss": 0.0058, "step": 88020 }, { "grad_norm": 0.17736248672008514, "learning_rate": 3.867000583213082e-06, "loss": 0.0055, "step": 88030 }, { "grad_norm": 0.12806399166584015, "learning_rate": 3.860627086405871e-06, "loss": 0.0054, "step": 88040 }, { "grad_norm": 0.091669462621212, "learning_rate": 3.854258635331126e-06, "loss": 0.004, "step": 88050 }, { "grad_norm": 0.12764671444892883, "learning_rate": 3.8478952306852665e-06, "loss": 0.0055, "step": 88060 }, { "grad_norm": 0.05510778725147247, "learning_rate": 3.841536873164198e-06, "loss": 0.0052, "step": 88070 }, { "grad_norm": 0.06547575443983078, "learning_rate": 3.8351835634632626e-06, "loss": 0.0056, "step": 88080 }, { "grad_norm": 0.1519266664981842, "learning_rate": 3.828835302277245e-06, "loss": 0.0048, "step": 88090 }, { "grad_norm": 0.14419688284397125, "learning_rate": 3.822492090300378e-06, "loss": 0.0047, "step": 88100 }, { "grad_norm": 0.12869380414485931, "learning_rate": 3.8161539282263566e-06, "loss": 0.0059, "step": 88110 }, { "grad_norm": 0.0946546420454979, "learning_rate": 3.8098208167482995e-06, "loss": 0.0042, "step": 88120 }, { "grad_norm": 0.07030851393938065, "learning_rate": 3.803492756558796e-06, "loss": 0.006, "step": 88130 }, { "grad_norm": 0.08121725916862488, "learning_rate": 3.797169748349855e-06, "loss": 0.0058, "step": 88140 }, { "grad_norm": 0.07540949434041977, "learning_rate": 3.790851792812977e-06, "loss": 0.0055, "step": 88150 }, { "grad_norm": 0.11855149269104004, "learning_rate": 3.784538890639072e-06, "loss": 0.0046, "step": 88160 }, { "grad_norm": 0.08129053562879562, "learning_rate": 3.7782310425185153e-06, "loss": 0.0042, "step": 88170 }, { "grad_norm": 0.08415286988019943, "learning_rate": 3.77192824914111e-06, "loss": 0.0041, "step": 88180 }, { "grad_norm": 0.13076144456863403, "learning_rate": 3.7656305111961333e-06, "loss": 0.0052, "step": 88190 }, { "grad_norm": 0.12465421110391617, "learning_rate": 3.7593378293722782e-06, "loss": 0.005, "step": 88200 }, { "grad_norm": 0.09690627455711365, "learning_rate": 3.7530502043577333e-06, "loss": 0.006, "step": 88210 }, { "grad_norm": 0.19503343105316162, "learning_rate": 3.746767636840076e-06, "loss": 0.0058, "step": 88220 }, { "grad_norm": 0.09712288528680801, "learning_rate": 3.74049012750638e-06, "loss": 0.0061, "step": 88230 }, { "grad_norm": 0.07891960442066193, "learning_rate": 3.734217677043128e-06, "loss": 0.0046, "step": 88240 }, { "grad_norm": 0.13214118778705597, "learning_rate": 3.727950286136267e-06, "loss": 0.0057, "step": 88250 }, { "grad_norm": 0.09467517584562302, "learning_rate": 3.7216879554711925e-06, "loss": 0.0032, "step": 88260 }, { "grad_norm": 0.08821222186088562, "learning_rate": 3.715430685732757e-06, "loss": 0.0047, "step": 88270 }, { "grad_norm": 0.15841904282569885, "learning_rate": 3.7091784776052243e-06, "loss": 0.0047, "step": 88280 }, { "grad_norm": 0.11886288970708847, "learning_rate": 3.7029313317723356e-06, "loss": 0.0059, "step": 88290 }, { "grad_norm": 0.10232797265052795, "learning_rate": 3.6966892489172787e-06, "loss": 0.0043, "step": 88300 }, { "grad_norm": 0.0943203940987587, "learning_rate": 3.6904522297226575e-06, "loss": 0.0042, "step": 88310 }, { "grad_norm": 0.07467324286699295, "learning_rate": 3.6842202748705647e-06, "loss": 0.0042, "step": 88320 }, { "grad_norm": 0.08894191682338715, "learning_rate": 3.677993385042494e-06, "loss": 0.005, "step": 88330 }, { "grad_norm": 0.10399052500724792, "learning_rate": 3.67177156091944e-06, "loss": 0.0072, "step": 88340 }, { "grad_norm": 0.05920887365937233, "learning_rate": 3.6655548031817866e-06, "loss": 0.0046, "step": 88350 }, { "grad_norm": 0.09476706385612488, "learning_rate": 3.6593431125094e-06, "loss": 0.0049, "step": 88360 }, { "grad_norm": 0.13693732023239136, "learning_rate": 3.6531364895815766e-06, "loss": 0.0047, "step": 88370 }, { "grad_norm": 0.12260263413190842, "learning_rate": 3.6469349350770677e-06, "loss": 0.0043, "step": 88380 }, { "grad_norm": 0.11840196698904037, "learning_rate": 3.6407384496740525e-06, "loss": 0.0056, "step": 88390 }, { "grad_norm": 0.046798478811979294, "learning_rate": 3.6345470340501954e-06, "loss": 0.0054, "step": 88400 }, { "grad_norm": 0.06836345791816711, "learning_rate": 3.6283606888825537e-06, "loss": 0.0047, "step": 88410 }, { "grad_norm": 0.14497224986553192, "learning_rate": 3.622179414847676e-06, "loss": 0.0046, "step": 88420 }, { "grad_norm": 0.09604418277740479, "learning_rate": 3.616003212621527e-06, "loss": 0.0051, "step": 88430 }, { "grad_norm": 0.10995197296142578, "learning_rate": 3.609832082879533e-06, "loss": 0.0034, "step": 88440 }, { "grad_norm": 0.08943119645118713, "learning_rate": 3.6036660262965485e-06, "loss": 0.0068, "step": 88450 }, { "grad_norm": 0.09057488292455673, "learning_rate": 3.597505043546895e-06, "loss": 0.0081, "step": 88460 }, { "grad_norm": 0.12804444134235382, "learning_rate": 3.5913491353043227e-06, "loss": 0.005, "step": 88470 }, { "grad_norm": 0.15781739354133606, "learning_rate": 3.5851983022420375e-06, "loss": 0.0044, "step": 88480 }, { "grad_norm": 0.12779316306114197, "learning_rate": 3.5790525450326797e-06, "loss": 0.0066, "step": 88490 }, { "grad_norm": 0.2031072974205017, "learning_rate": 3.572911864348344e-06, "loss": 0.0061, "step": 88500 }, { "grad_norm": 0.10332266986370087, "learning_rate": 3.5667762608605717e-06, "loss": 0.0045, "step": 88510 }, { "grad_norm": 0.11355938762426376, "learning_rate": 3.5606457352403255e-06, "loss": 0.0072, "step": 88520 }, { "grad_norm": 0.09890627861022949, "learning_rate": 3.554520288158042e-06, "loss": 0.0047, "step": 88530 }, { "grad_norm": 0.14614461362361908, "learning_rate": 3.5483999202835903e-06, "loss": 0.0066, "step": 88540 }, { "grad_norm": 0.12815013527870178, "learning_rate": 3.5422846322862913e-06, "loss": 0.0075, "step": 88550 }, { "grad_norm": 0.13736172020435333, "learning_rate": 3.5361744248348925e-06, "loss": 0.0043, "step": 88560 }, { "grad_norm": 0.09277227520942688, "learning_rate": 3.5300692985976047e-06, "loss": 0.0082, "step": 88570 }, { "grad_norm": 0.10316198319196701, "learning_rate": 3.52396925424206e-06, "loss": 0.0045, "step": 88580 }, { "grad_norm": 0.08051271736621857, "learning_rate": 3.5178742924353747e-06, "loss": 0.004, "step": 88590 }, { "grad_norm": 0.09891857206821442, "learning_rate": 3.511784413844066e-06, "loss": 0.0049, "step": 88600 }, { "grad_norm": 0.22535020112991333, "learning_rate": 3.5056996191341287e-06, "loss": 0.0061, "step": 88610 }, { "grad_norm": 0.08923370391130447, "learning_rate": 3.4996199089709692e-06, "loss": 0.0044, "step": 88620 }, { "grad_norm": 0.09954900294542313, "learning_rate": 3.4935452840194725e-06, "loss": 0.0054, "step": 88630 }, { "grad_norm": 0.14249718189239502, "learning_rate": 3.48747574494393e-06, "loss": 0.008, "step": 88640 }, { "grad_norm": 0.09870780259370804, "learning_rate": 3.4814112924081154e-06, "loss": 0.0063, "step": 88650 }, { "grad_norm": 0.08371616154909134, "learning_rate": 3.475351927075221e-06, "loss": 0.0048, "step": 88660 }, { "grad_norm": 0.1723753660917282, "learning_rate": 3.469297649607894e-06, "loss": 0.008, "step": 88670 }, { "grad_norm": 0.15040229260921478, "learning_rate": 3.463248460668217e-06, "loss": 0.0048, "step": 88680 }, { "grad_norm": 0.06550882011651993, "learning_rate": 3.457204360917726e-06, "loss": 0.0047, "step": 88690 }, { "grad_norm": 0.11514536291360855, "learning_rate": 3.4511653510173827e-06, "loss": 0.0061, "step": 88700 }, { "grad_norm": 0.0880901888012886, "learning_rate": 3.4451314316276074e-06, "loss": 0.0053, "step": 88710 }, { "grad_norm": 0.04522491991519928, "learning_rate": 3.4391026034082674e-06, "loss": 0.0055, "step": 88720 }, { "grad_norm": 0.07843457907438278, "learning_rate": 3.433078867018674e-06, "loss": 0.003, "step": 88730 }, { "grad_norm": 0.11056909710168839, "learning_rate": 3.427060223117551e-06, "loss": 0.0046, "step": 88740 }, { "grad_norm": 0.34721025824546814, "learning_rate": 3.4210466723631042e-06, "loss": 0.006, "step": 88750 }, { "grad_norm": 0.13285548985004425, "learning_rate": 3.415038215412969e-06, "loss": 0.005, "step": 88760 }, { "grad_norm": 0.04829656332731247, "learning_rate": 3.4090348529242088e-06, "loss": 0.0064, "step": 88770 }, { "grad_norm": 0.15265408158302307, "learning_rate": 3.4030365855533487e-06, "loss": 0.0067, "step": 88780 }, { "grad_norm": 0.07821979373693466, "learning_rate": 3.397043413956347e-06, "loss": 0.0039, "step": 88790 }, { "grad_norm": 0.10296043753623962, "learning_rate": 3.3910553387886237e-06, "loss": 0.0046, "step": 88800 }, { "grad_norm": 0.10674064606428146, "learning_rate": 3.3850723607049996e-06, "loss": 0.0074, "step": 88810 }, { "grad_norm": 0.06036478281021118, "learning_rate": 3.3790944803597902e-06, "loss": 0.0054, "step": 88820 }, { "grad_norm": 0.19045913219451904, "learning_rate": 3.373121698406706e-06, "loss": 0.0064, "step": 88830 }, { "grad_norm": 0.09716162085533142, "learning_rate": 3.3671540154989302e-06, "loss": 0.0051, "step": 88840 }, { "grad_norm": 0.17133118212223053, "learning_rate": 3.361191432289079e-06, "loss": 0.0051, "step": 88850 }, { "grad_norm": 0.054482389241456985, "learning_rate": 3.3552339494292207e-06, "loss": 0.0057, "step": 88860 }, { "grad_norm": 0.08994857966899872, "learning_rate": 3.349281567570839e-06, "loss": 0.0063, "step": 88870 }, { "grad_norm": 0.10760333389043808, "learning_rate": 3.3433342873648965e-06, "loss": 0.0051, "step": 88880 }, { "grad_norm": 0.08613234758377075, "learning_rate": 3.337392109461762e-06, "loss": 0.008, "step": 88890 }, { "grad_norm": 0.07212766259908676, "learning_rate": 3.331455034511266e-06, "loss": 0.0041, "step": 88900 }, { "grad_norm": 0.09392213076353073, "learning_rate": 3.325523063162689e-06, "loss": 0.004, "step": 88910 }, { "grad_norm": 0.1288374811410904, "learning_rate": 3.3195961960647393e-06, "loss": 0.0058, "step": 88920 }, { "grad_norm": 0.08197011053562164, "learning_rate": 3.31367443386556e-06, "loss": 0.0072, "step": 88930 }, { "grad_norm": 0.10227207839488983, "learning_rate": 3.3077577772127556e-06, "loss": 0.0058, "step": 88940 }, { "grad_norm": 0.12475673109292984, "learning_rate": 3.301846226753358e-06, "loss": 0.005, "step": 88950 }, { "grad_norm": 0.16097591817378998, "learning_rate": 3.2959397831338444e-06, "loss": 0.0054, "step": 88960 }, { "grad_norm": 0.05331679433584213, "learning_rate": 3.2900384470001432e-06, "loss": 0.0041, "step": 88970 }, { "grad_norm": 0.08469480276107788, "learning_rate": 3.28414221899761e-06, "loss": 0.004, "step": 88980 }, { "grad_norm": 0.22127707302570343, "learning_rate": 3.278251099771046e-06, "loss": 0.0048, "step": 88990 }, { "grad_norm": 0.12881842255592346, "learning_rate": 3.2723650899646906e-06, "loss": 0.0044, "step": 89000 }, { "grad_norm": 0.12047366797924042, "learning_rate": 3.2664841902222466e-06, "loss": 0.0041, "step": 89010 }, { "grad_norm": 0.07872378081083298, "learning_rate": 3.2606084011868208e-06, "loss": 0.0039, "step": 89020 }, { "grad_norm": 0.13011109828948975, "learning_rate": 3.254737723500989e-06, "loss": 0.0047, "step": 89030 }, { "grad_norm": 0.1331816017627716, "learning_rate": 3.2488721578067595e-06, "loss": 0.0044, "step": 89040 }, { "grad_norm": 0.09229686111211777, "learning_rate": 3.243011704745591e-06, "loss": 0.0054, "step": 89050 }, { "grad_norm": 0.12084998190402985, "learning_rate": 3.237156364958355e-06, "loss": 0.0042, "step": 89060 }, { "grad_norm": 0.08135716617107391, "learning_rate": 3.231306139085394e-06, "loss": 0.0043, "step": 89070 }, { "grad_norm": 0.19382435083389282, "learning_rate": 3.2254610277664797e-06, "loss": 0.0072, "step": 89080 }, { "grad_norm": 0.10105372965335846, "learning_rate": 3.2196210316408183e-06, "loss": 0.0048, "step": 89090 }, { "grad_norm": 0.10018175095319748, "learning_rate": 3.2137861513470644e-06, "loss": 0.0054, "step": 89100 }, { "grad_norm": 0.10269100964069366, "learning_rate": 3.2079563875233252e-06, "loss": 0.0056, "step": 89110 }, { "grad_norm": 0.0750313401222229, "learning_rate": 3.2021317408071184e-06, "loss": 0.006, "step": 89120 }, { "grad_norm": 0.09368234127759933, "learning_rate": 3.196312211835434e-06, "loss": 0.0045, "step": 89130 }, { "grad_norm": 0.097596175968647, "learning_rate": 3.1904978012446686e-06, "loss": 0.0053, "step": 89140 }, { "grad_norm": 0.13533103466033936, "learning_rate": 3.184688509670691e-06, "loss": 0.0069, "step": 89150 }, { "grad_norm": 0.07593703269958496, "learning_rate": 3.178884337748783e-06, "loss": 0.0062, "step": 89160 }, { "grad_norm": 0.14655308425426483, "learning_rate": 3.173085286113697e-06, "loss": 0.0055, "step": 89170 }, { "grad_norm": 0.09082010388374329, "learning_rate": 3.1672913553995986e-06, "loss": 0.0062, "step": 89180 }, { "grad_norm": 0.2094985991716385, "learning_rate": 3.1615025462401138e-06, "loss": 0.0056, "step": 89190 }, { "grad_norm": 0.1182185560464859, "learning_rate": 3.1557188592682808e-06, "loss": 0.0043, "step": 89200 }, { "grad_norm": 0.060026392340660095, "learning_rate": 3.1499402951166045e-06, "loss": 0.0062, "step": 89210 }, { "grad_norm": 0.06510094553232193, "learning_rate": 3.144166854417019e-06, "loss": 0.0045, "step": 89220 }, { "grad_norm": 0.0754639208316803, "learning_rate": 3.1383985378009073e-06, "loss": 0.0059, "step": 89230 }, { "grad_norm": 0.19923055171966553, "learning_rate": 3.1326353458990654e-06, "loss": 0.0063, "step": 89240 }, { "grad_norm": 0.1608317494392395, "learning_rate": 3.126877279341761e-06, "loss": 0.0048, "step": 89250 }, { "grad_norm": 0.18251734972000122, "learning_rate": 3.1211243387586964e-06, "loss": 0.0055, "step": 89260 }, { "grad_norm": 0.18377478420734406, "learning_rate": 3.115376524778979e-06, "loss": 0.0064, "step": 89270 }, { "grad_norm": 0.10598146170377731, "learning_rate": 3.109633838031195e-06, "loss": 0.0042, "step": 89280 }, { "grad_norm": 0.14114056527614594, "learning_rate": 3.103896279143359e-06, "loss": 0.0057, "step": 89290 }, { "grad_norm": 0.12672440707683563, "learning_rate": 3.098163848742924e-06, "loss": 0.0051, "step": 89300 }, { "grad_norm": 0.1696566939353943, "learning_rate": 3.0924365474567675e-06, "loss": 0.0072, "step": 89310 }, { "grad_norm": 0.09047320485115051, "learning_rate": 3.086714375911237e-06, "loss": 0.0058, "step": 89320 }, { "grad_norm": 0.10897676646709442, "learning_rate": 3.0809973347320775e-06, "loss": 0.0053, "step": 89330 }, { "grad_norm": 0.1218068078160286, "learning_rate": 3.075285424544516e-06, "loss": 0.0057, "step": 89340 }, { "grad_norm": 0.10929958522319794, "learning_rate": 3.069578645973176e-06, "loss": 0.0058, "step": 89350 }, { "grad_norm": 0.14012491703033447, "learning_rate": 3.0638769996421735e-06, "loss": 0.0066, "step": 89360 }, { "grad_norm": 0.06682701408863068, "learning_rate": 3.058180486175005e-06, "loss": 0.0049, "step": 89370 }, { "grad_norm": 0.1333913803100586, "learning_rate": 3.052489106194645e-06, "loss": 0.0058, "step": 89380 }, { "grad_norm": 0.10546836256980896, "learning_rate": 3.0468028603234946e-06, "loss": 0.0057, "step": 89390 }, { "grad_norm": 0.1079997718334198, "learning_rate": 3.04112174918339e-06, "loss": 0.0061, "step": 89400 }, { "grad_norm": 0.06716286391019821, "learning_rate": 3.0354457733955954e-06, "loss": 0.007, "step": 89410 }, { "grad_norm": 0.18691127002239227, "learning_rate": 3.029774933580859e-06, "loss": 0.0064, "step": 89420 }, { "grad_norm": 0.09257865697145462, "learning_rate": 3.0241092303593065e-06, "loss": 0.0051, "step": 89430 }, { "grad_norm": 0.09867848455905914, "learning_rate": 3.018448664350554e-06, "loss": 0.006, "step": 89440 }, { "grad_norm": 0.12783175706863403, "learning_rate": 3.0127932361736055e-06, "loss": 0.0048, "step": 89450 }, { "grad_norm": 0.23008129000663757, "learning_rate": 3.00714294644695e-06, "loss": 0.0058, "step": 89460 }, { "grad_norm": 0.08193869143724442, "learning_rate": 3.0014977957884828e-06, "loss": 0.0065, "step": 89470 }, { "grad_norm": 0.07649917155504227, "learning_rate": 2.99585778481557e-06, "loss": 0.0048, "step": 89480 }, { "grad_norm": 0.11567423492670059, "learning_rate": 2.990222914144969e-06, "loss": 0.0041, "step": 89490 }, { "grad_norm": 0.11549244076013565, "learning_rate": 2.984593184392914e-06, "loss": 0.0067, "step": 89500 }, { "grad_norm": 0.11735192686319351, "learning_rate": 2.978968596175069e-06, "loss": 0.0043, "step": 89510 }, { "grad_norm": 0.08544634282588959, "learning_rate": 2.9733491501065137e-06, "loss": 0.0042, "step": 89520 }, { "grad_norm": 0.07672715932130814, "learning_rate": 2.9677348468018005e-06, "loss": 0.0054, "step": 89530 }, { "grad_norm": 0.1262577772140503, "learning_rate": 2.9621256868748726e-06, "loss": 0.005, "step": 89540 }, { "grad_norm": 0.09677986055612564, "learning_rate": 2.956521670939183e-06, "loss": 0.0074, "step": 89550 }, { "grad_norm": 0.09602265059947968, "learning_rate": 2.950922799607542e-06, "loss": 0.005, "step": 89560 }, { "grad_norm": 0.07694683969020844, "learning_rate": 2.9453290734922535e-06, "loss": 0.004, "step": 89570 }, { "grad_norm": 0.07748493552207947, "learning_rate": 2.939740493205029e-06, "loss": 0.005, "step": 89580 }, { "grad_norm": 0.1052853986620903, "learning_rate": 2.9341570593570345e-06, "loss": 0.0044, "step": 89590 }, { "grad_norm": 0.1318482905626297, "learning_rate": 2.928578772558843e-06, "loss": 0.0057, "step": 89600 }, { "grad_norm": 0.07933087646961212, "learning_rate": 2.923005633420528e-06, "loss": 0.0053, "step": 89610 }, { "grad_norm": 0.12008210271596909, "learning_rate": 2.9174376425515237e-06, "loss": 0.0048, "step": 89620 }, { "grad_norm": 0.1761518269777298, "learning_rate": 2.91187480056076e-06, "loss": 0.0058, "step": 89630 }, { "grad_norm": 0.06077269837260246, "learning_rate": 2.9063171080565664e-06, "loss": 0.0041, "step": 89640 }, { "grad_norm": 0.06779218465089798, "learning_rate": 2.900764565646741e-06, "loss": 0.007, "step": 89650 }, { "grad_norm": 0.29697033762931824, "learning_rate": 2.895217173938475e-06, "loss": 0.0063, "step": 89660 }, { "grad_norm": 0.07652925699949265, "learning_rate": 2.8896749335384445e-06, "loss": 0.0057, "step": 89670 }, { "grad_norm": 0.12684276700019836, "learning_rate": 2.884137845052731e-06, "loss": 0.0064, "step": 89680 }, { "grad_norm": 0.1439298689365387, "learning_rate": 2.8786059090868734e-06, "loss": 0.0047, "step": 89690 }, { "grad_norm": 0.13770617544651031, "learning_rate": 2.87307912624582e-06, "loss": 0.0051, "step": 89700 }, { "grad_norm": 0.09575104713439941, "learning_rate": 2.867557497133977e-06, "loss": 0.0073, "step": 89710 }, { "grad_norm": 0.19196318089962006, "learning_rate": 2.862041022355194e-06, "loss": 0.0055, "step": 89720 }, { "grad_norm": 0.07578575611114502, "learning_rate": 2.8565297025127226e-06, "loss": 0.0051, "step": 89730 }, { "grad_norm": 0.12440618872642517, "learning_rate": 2.8510235382092853e-06, "loss": 0.0069, "step": 89740 }, { "grad_norm": 0.1399124711751938, "learning_rate": 2.845522530047029e-06, "loss": 0.0054, "step": 89750 }, { "grad_norm": 0.12357067316770554, "learning_rate": 2.8400266786275387e-06, "loss": 0.0049, "step": 89760 }, { "grad_norm": 0.0831754133105278, "learning_rate": 2.834535984551817e-06, "loss": 0.0043, "step": 89770 }, { "grad_norm": 0.10219747573137283, "learning_rate": 2.8290504484203397e-06, "loss": 0.0045, "step": 89780 }, { "grad_norm": 0.11145316064357758, "learning_rate": 2.8235700708329705e-06, "loss": 0.0052, "step": 89790 }, { "grad_norm": 0.20261314511299133, "learning_rate": 2.818094852389058e-06, "loss": 0.004, "step": 89800 }, { "grad_norm": 0.058856379240751266, "learning_rate": 2.8126247936873516e-06, "loss": 0.0066, "step": 89810 }, { "grad_norm": 0.1703905314207077, "learning_rate": 2.8071598953260614e-06, "loss": 0.0055, "step": 89820 }, { "grad_norm": 0.07191277295351028, "learning_rate": 2.8017001579028035e-06, "loss": 0.0035, "step": 89830 }, { "grad_norm": 0.07932549715042114, "learning_rate": 2.7962455820146617e-06, "loss": 0.005, "step": 89840 }, { "grad_norm": 0.1811380237340927, "learning_rate": 2.7907961682581253e-06, "loss": 0.0084, "step": 89850 }, { "grad_norm": 0.17423036694526672, "learning_rate": 2.7853519172291453e-06, "loss": 0.0074, "step": 89860 }, { "grad_norm": 0.18409278988838196, "learning_rate": 2.7799128295230904e-06, "loss": 0.0058, "step": 89870 }, { "grad_norm": 0.10363984853029251, "learning_rate": 2.774478905734784e-06, "loss": 0.0049, "step": 89880 }, { "grad_norm": 0.14439120888710022, "learning_rate": 2.7690501464584516e-06, "loss": 0.0067, "step": 89890 }, { "grad_norm": 0.06931116431951523, "learning_rate": 2.763626552287796e-06, "loss": 0.005, "step": 89900 }, { "grad_norm": 0.07949219644069672, "learning_rate": 2.7582081238159083e-06, "loss": 0.0044, "step": 89910 }, { "grad_norm": 0.08615421503782272, "learning_rate": 2.7527948616353548e-06, "loss": 0.0054, "step": 89920 }, { "grad_norm": 0.10346674174070358, "learning_rate": 2.7473867663381226e-06, "loss": 0.004, "step": 89930 }, { "grad_norm": 0.07778717577457428, "learning_rate": 2.7419838385156383e-06, "loss": 0.0039, "step": 89940 }, { "grad_norm": 0.07945632934570312, "learning_rate": 2.7365860787587407e-06, "loss": 0.0067, "step": 89950 }, { "grad_norm": 0.0783858448266983, "learning_rate": 2.7311934876577295e-06, "loss": 0.0052, "step": 89960 }, { "grad_norm": 0.07446582615375519, "learning_rate": 2.725806065802339e-06, "loss": 0.0067, "step": 89970 }, { "grad_norm": 0.09037021547555923, "learning_rate": 2.720423813781714e-06, "loss": 0.0062, "step": 89980 }, { "grad_norm": 0.06954097747802734, "learning_rate": 2.715046732184462e-06, "loss": 0.0045, "step": 89990 }, { "grad_norm": 0.12512870132923126, "learning_rate": 2.7096748215986013e-06, "loss": 0.005, "step": 90000 }, { "grad_norm": 0.09010086953639984, "learning_rate": 2.704308082611612e-06, "loss": 0.0043, "step": 90010 }, { "grad_norm": 0.1487579345703125, "learning_rate": 2.698946515810374e-06, "loss": 0.0048, "step": 90020 }, { "grad_norm": 0.09124664962291718, "learning_rate": 2.6935901217812367e-06, "loss": 0.006, "step": 90030 }, { "grad_norm": 0.16641239821910858, "learning_rate": 2.688238901109952e-06, "loss": 0.0043, "step": 90040 }, { "grad_norm": 0.13407360017299652, "learning_rate": 2.68289285438173e-06, "loss": 0.0046, "step": 90050 }, { "grad_norm": 0.14229577779769897, "learning_rate": 2.6775519821812033e-06, "loss": 0.0053, "step": 90060 }, { "grad_norm": 0.17300820350646973, "learning_rate": 2.672216285092449e-06, "loss": 0.0046, "step": 90070 }, { "grad_norm": 0.1587972193956375, "learning_rate": 2.666885763698962e-06, "loss": 0.0067, "step": 90080 }, { "grad_norm": 0.22149518132209778, "learning_rate": 2.661560418583692e-06, "loss": 0.0068, "step": 90090 }, { "grad_norm": 0.133439838886261, "learning_rate": 2.656240250328995e-06, "loss": 0.0047, "step": 90100 }, { "grad_norm": 0.1443481743335724, "learning_rate": 2.650925259516679e-06, "loss": 0.0047, "step": 90110 }, { "grad_norm": 0.0817752406001091, "learning_rate": 2.645615446727995e-06, "loss": 0.0052, "step": 90120 }, { "grad_norm": 0.11238790303468704, "learning_rate": 2.6403108125436115e-06, "loss": 0.0049, "step": 90130 }, { "grad_norm": 0.060663215816020966, "learning_rate": 2.6350113575436264e-06, "loss": 0.0044, "step": 90140 }, { "grad_norm": 0.08610903471708298, "learning_rate": 2.629717082307598e-06, "loss": 0.0061, "step": 90150 }, { "grad_norm": 0.100364550948143, "learning_rate": 2.6244279874144793e-06, "loss": 0.0058, "step": 90160 }, { "grad_norm": 0.20840966701507568, "learning_rate": 2.6191440734426862e-06, "loss": 0.005, "step": 90170 }, { "grad_norm": 0.1070779487490654, "learning_rate": 2.613865340970062e-06, "loss": 0.0047, "step": 90180 }, { "grad_norm": 0.2697802186012268, "learning_rate": 2.608591790573889e-06, "loss": 0.0054, "step": 90190 }, { "grad_norm": 0.10116687417030334, "learning_rate": 2.603323422830861e-06, "loss": 0.0059, "step": 90200 }, { "grad_norm": 0.11112944036722183, "learning_rate": 2.5980602383171183e-06, "loss": 0.0051, "step": 90210 }, { "grad_norm": 0.0689190998673439, "learning_rate": 2.5928022376082496e-06, "loss": 0.0043, "step": 90220 }, { "grad_norm": 0.1048085018992424, "learning_rate": 2.5875494212792505e-06, "loss": 0.0057, "step": 90230 }, { "grad_norm": 0.10000137239694595, "learning_rate": 2.5823017899045565e-06, "loss": 0.0047, "step": 90240 }, { "grad_norm": 0.11441182345151901, "learning_rate": 2.577059344058047e-06, "loss": 0.005, "step": 90250 }, { "grad_norm": 0.15520377457141876, "learning_rate": 2.5718220843130415e-06, "loss": 0.0045, "step": 90260 }, { "grad_norm": 0.09009388834238052, "learning_rate": 2.5665900112422535e-06, "loss": 0.005, "step": 90270 }, { "grad_norm": 0.11321365088224411, "learning_rate": 2.561363125417876e-06, "loss": 0.007, "step": 90280 }, { "grad_norm": 0.08199279755353928, "learning_rate": 2.556141427411496e-06, "loss": 0.0045, "step": 90290 }, { "grad_norm": 0.18109840154647827, "learning_rate": 2.550924917794156e-06, "loss": 0.007, "step": 90300 }, { "grad_norm": 0.14544124901294708, "learning_rate": 2.545713597136329e-06, "loss": 0.0052, "step": 90310 }, { "grad_norm": 0.08993775397539139, "learning_rate": 2.5405074660079298e-06, "loss": 0.007, "step": 90320 }, { "grad_norm": 0.06811091303825378, "learning_rate": 2.5353065249782647e-06, "loss": 0.0046, "step": 90330 }, { "grad_norm": 0.3211016058921814, "learning_rate": 2.530110774616129e-06, "loss": 0.0065, "step": 90340 }, { "grad_norm": 0.1123778223991394, "learning_rate": 2.524920215489701e-06, "loss": 0.0076, "step": 90350 }, { "grad_norm": 0.21799714863300323, "learning_rate": 2.519734848166627e-06, "loss": 0.0054, "step": 90360 }, { "grad_norm": 0.1809682548046112, "learning_rate": 2.514554673213948e-06, "loss": 0.0057, "step": 90370 }, { "grad_norm": 0.18197713792324066, "learning_rate": 2.5093796911981936e-06, "loss": 0.0047, "step": 90380 }, { "grad_norm": 0.09076027572154999, "learning_rate": 2.5042099026852672e-06, "loss": 0.0059, "step": 90390 }, { "grad_norm": 0.21386025846004486, "learning_rate": 2.499045308240544e-06, "loss": 0.0065, "step": 90400 }, { "grad_norm": 0.06475617736577988, "learning_rate": 2.4938859084288003e-06, "loss": 0.0048, "step": 90410 }, { "grad_norm": 0.07085022330284119, "learning_rate": 2.4887317038142734e-06, "loss": 0.0043, "step": 90420 }, { "grad_norm": 0.11458883434534073, "learning_rate": 2.4835826949606124e-06, "loss": 0.0044, "step": 90430 }, { "grad_norm": 0.10222896188497543, "learning_rate": 2.4784388824309167e-06, "loss": 0.0041, "step": 90440 }, { "grad_norm": 0.09294336289167404, "learning_rate": 2.473300266787687e-06, "loss": 0.0038, "step": 90450 }, { "grad_norm": 0.22559046745300293, "learning_rate": 2.46816684859289e-06, "loss": 0.0052, "step": 90460 }, { "grad_norm": 0.12415118515491486, "learning_rate": 2.4630386284079043e-06, "loss": 0.0052, "step": 90470 }, { "grad_norm": 0.1277211755514145, "learning_rate": 2.4579156067935427e-06, "loss": 0.0038, "step": 90480 }, { "grad_norm": 0.5831364393234253, "learning_rate": 2.452797784310046e-06, "loss": 0.0047, "step": 90490 }, { "grad_norm": 0.05754515528678894, "learning_rate": 2.447685161517105e-06, "loss": 0.0095, "step": 90500 }, { "grad_norm": 0.0834830179810524, "learning_rate": 2.4425777389738224e-06, "loss": 0.0051, "step": 90510 }, { "grad_norm": 0.09116460382938385, "learning_rate": 2.43747551723873e-06, "loss": 0.006, "step": 90520 }, { "grad_norm": 0.1465686857700348, "learning_rate": 2.4323784968698136e-06, "loss": 0.0051, "step": 90530 }, { "grad_norm": 0.09143280982971191, "learning_rate": 2.4272866784244565e-06, "loss": 0.0055, "step": 90540 }, { "grad_norm": 0.14226393401622772, "learning_rate": 2.422200062459518e-06, "loss": 0.0048, "step": 90550 }, { "grad_norm": 0.16800959408283234, "learning_rate": 2.41711864953123e-06, "loss": 0.0057, "step": 90560 }, { "grad_norm": 0.056909870356321335, "learning_rate": 2.412042440195322e-06, "loss": 0.0046, "step": 90570 }, { "grad_norm": 0.10532179474830627, "learning_rate": 2.4069714350068993e-06, "loss": 0.0055, "step": 90580 }, { "grad_norm": 0.1333203911781311, "learning_rate": 2.4019056345205294e-06, "loss": 0.0053, "step": 90590 }, { "grad_norm": 0.09476164728403091, "learning_rate": 2.396845039290191e-06, "loss": 0.0072, "step": 90600 }, { "grad_norm": 0.12709124386310577, "learning_rate": 2.3917896498693194e-06, "loss": 0.0051, "step": 90610 }, { "grad_norm": 0.08375664800405502, "learning_rate": 2.3867394668107335e-06, "loss": 0.0051, "step": 90620 }, { "grad_norm": 0.059226226061582565, "learning_rate": 2.3816944906667527e-06, "loss": 0.0034, "step": 90630 }, { "grad_norm": 0.07186202704906464, "learning_rate": 2.3766547219890634e-06, "loss": 0.0045, "step": 90640 }, { "grad_norm": 0.126424640417099, "learning_rate": 2.371620161328819e-06, "loss": 0.0067, "step": 90650 }, { "grad_norm": 0.06549938023090363, "learning_rate": 2.366590809236574e-06, "loss": 0.0043, "step": 90660 }, { "grad_norm": 0.09877213835716248, "learning_rate": 2.3615666662623494e-06, "loss": 0.0042, "step": 90670 }, { "grad_norm": 0.12835027277469635, "learning_rate": 2.3565477329555664e-06, "loss": 0.006, "step": 90680 }, { "grad_norm": 0.09977696090936661, "learning_rate": 2.3515340098650975e-06, "loss": 0.005, "step": 90690 }, { "grad_norm": 0.10796666145324707, "learning_rate": 2.3465254975392258e-06, "loss": 0.0055, "step": 90700 }, { "grad_norm": 0.10962891578674316, "learning_rate": 2.3415221965256805e-06, "loss": 0.0047, "step": 90710 }, { "grad_norm": 0.3609377443790436, "learning_rate": 2.3365241073716237e-06, "loss": 0.0068, "step": 90720 }, { "grad_norm": 0.14209024608135223, "learning_rate": 2.3315312306236183e-06, "loss": 0.0052, "step": 90730 }, { "grad_norm": 0.13625359535217285, "learning_rate": 2.326543566827699e-06, "loss": 0.0071, "step": 90740 }, { "grad_norm": 0.124146468937397, "learning_rate": 2.3215611165292862e-06, "loss": 0.0069, "step": 90750 }, { "grad_norm": 0.11350266635417938, "learning_rate": 2.3165838802732765e-06, "loss": 0.0052, "step": 90760 }, { "grad_norm": 0.10435892641544342, "learning_rate": 2.311611858603957e-06, "loss": 0.0046, "step": 90770 }, { "grad_norm": 0.11543074995279312, "learning_rate": 2.3066450520650763e-06, "loss": 0.0065, "step": 90780 }, { "grad_norm": 0.0796801894903183, "learning_rate": 2.3016834611997773e-06, "loss": 0.0085, "step": 90790 }, { "grad_norm": 0.07823674380779266, "learning_rate": 2.296727086550665e-06, "loss": 0.0065, "step": 90800 }, { "grad_norm": 0.16635723412036896, "learning_rate": 2.2917759286597507e-06, "loss": 0.0064, "step": 90810 }, { "grad_norm": 0.07400718331336975, "learning_rate": 2.2868299880685006e-06, "loss": 0.0041, "step": 90820 }, { "grad_norm": 0.07384564727544785, "learning_rate": 2.2818892653177827e-06, "loss": 0.0048, "step": 90830 }, { "grad_norm": 0.18168315291404724, "learning_rate": 2.2769537609479196e-06, "loss": 0.0045, "step": 90840 }, { "grad_norm": 0.08986115455627441, "learning_rate": 2.2720234754986357e-06, "loss": 0.0036, "step": 90850 }, { "grad_norm": 0.16588440537452698, "learning_rate": 2.2670984095091108e-06, "loss": 0.0036, "step": 90860 }, { "grad_norm": 0.21387648582458496, "learning_rate": 2.262178563517936e-06, "loss": 0.0063, "step": 90870 }, { "grad_norm": 0.1893904209136963, "learning_rate": 2.257263938063137e-06, "loss": 0.0062, "step": 90880 }, { "grad_norm": 0.13863880932331085, "learning_rate": 2.2523545336821715e-06, "loss": 0.0093, "step": 90890 }, { "grad_norm": 0.10111382603645325, "learning_rate": 2.247450350911939e-06, "loss": 0.0041, "step": 90900 }, { "grad_norm": 0.09331833571195602, "learning_rate": 2.2425513902887328e-06, "loss": 0.0043, "step": 90910 }, { "grad_norm": 0.13178661465644836, "learning_rate": 2.2376576523483015e-06, "loss": 0.005, "step": 90920 }, { "grad_norm": 0.16807158291339874, "learning_rate": 2.232769137625829e-06, "loss": 0.0053, "step": 90930 }, { "grad_norm": 0.06339607387781143, "learning_rate": 2.227885846655903e-06, "loss": 0.0055, "step": 90940 }, { "grad_norm": 0.1487269103527069, "learning_rate": 2.2230077799725523e-06, "loss": 0.0064, "step": 90950 }, { "grad_norm": 0.10217871516942978, "learning_rate": 2.2181349381092396e-06, "loss": 0.0048, "step": 90960 }, { "grad_norm": 0.13200004398822784, "learning_rate": 2.2132673215988552e-06, "loss": 0.0054, "step": 90970 }, { "grad_norm": 0.1634935736656189, "learning_rate": 2.2084049309737065e-06, "loss": 0.0063, "step": 90980 }, { "grad_norm": 0.05554774031043053, "learning_rate": 2.2035477667655513e-06, "loss": 0.0042, "step": 90990 }, { "grad_norm": 0.12373173981904984, "learning_rate": 2.1986958295055315e-06, "loss": 0.0041, "step": 91000 }, { "grad_norm": 0.07800349593162537, "learning_rate": 2.193849119724284e-06, "loss": 0.005, "step": 91010 }, { "grad_norm": 0.11741983890533447, "learning_rate": 2.1890076379518175e-06, "loss": 0.0051, "step": 91020 }, { "grad_norm": 0.09282582998275757, "learning_rate": 2.1841713847175925e-06, "loss": 0.0052, "step": 91030 }, { "grad_norm": 0.2775152623653412, "learning_rate": 2.1793403605504968e-06, "loss": 0.0078, "step": 91040 }, { "grad_norm": 0.09567606449127197, "learning_rate": 2.1745145659788413e-06, "loss": 0.005, "step": 91050 }, { "grad_norm": 0.11916683614253998, "learning_rate": 2.1696940015303645e-06, "loss": 0.0036, "step": 91060 }, { "grad_norm": 0.11778310686349869, "learning_rate": 2.1648786677322397e-06, "loss": 0.0064, "step": 91070 }, { "grad_norm": 0.17921125888824463, "learning_rate": 2.1600685651110673e-06, "loss": 0.0063, "step": 91080 }, { "grad_norm": 0.10750935971736908, "learning_rate": 2.1552636941928717e-06, "loss": 0.004, "step": 91090 }, { "grad_norm": 0.0992858037352562, "learning_rate": 2.1504640555031043e-06, "loss": 0.0053, "step": 91100 }, { "grad_norm": 0.10369238257408142, "learning_rate": 2.1456696495666506e-06, "loss": 0.0057, "step": 91110 }, { "grad_norm": 0.12868067622184753, "learning_rate": 2.1408804769078084e-06, "loss": 0.0051, "step": 91120 }, { "grad_norm": 0.17444124817848206, "learning_rate": 2.136096538050325e-06, "loss": 0.0084, "step": 91130 }, { "grad_norm": 0.06695950031280518, "learning_rate": 2.13131783351736e-06, "loss": 0.0043, "step": 91140 }, { "grad_norm": 0.08644452691078186, "learning_rate": 2.126544363831512e-06, "loss": 0.0047, "step": 91150 }, { "grad_norm": 0.15590669214725494, "learning_rate": 2.121776129514791e-06, "loss": 0.0067, "step": 91160 }, { "grad_norm": 0.12093555182218552, "learning_rate": 2.1170131310886523e-06, "loss": 0.0042, "step": 91170 }, { "grad_norm": 0.10420595109462738, "learning_rate": 2.112255369073968e-06, "loss": 0.0077, "step": 91180 }, { "grad_norm": 0.8950021266937256, "learning_rate": 2.1075028439910328e-06, "loss": 0.0064, "step": 91190 }, { "grad_norm": 0.1210622489452362, "learning_rate": 2.1027555563595813e-06, "loss": 0.0072, "step": 91200 }, { "grad_norm": 0.10883519798517227, "learning_rate": 2.0980135066987706e-06, "loss": 0.0063, "step": 91210 }, { "grad_norm": 0.09173357486724854, "learning_rate": 2.093276695527191e-06, "loss": 0.0046, "step": 91220 }, { "grad_norm": 0.07298491150140762, "learning_rate": 2.0885451233628396e-06, "loss": 0.0057, "step": 91230 }, { "grad_norm": 0.2799190878868103, "learning_rate": 2.0838187907231633e-06, "loss": 0.0054, "step": 91240 }, { "grad_norm": 0.05004051700234413, "learning_rate": 2.0790976981250153e-06, "loss": 0.0044, "step": 91250 }, { "grad_norm": 0.11917325854301453, "learning_rate": 2.0743818460847e-06, "loss": 0.0042, "step": 91260 }, { "grad_norm": 0.04633372277021408, "learning_rate": 2.069671235117926e-06, "loss": 0.0051, "step": 91270 }, { "grad_norm": 0.16952216625213623, "learning_rate": 2.0649658657398484e-06, "loss": 0.005, "step": 91280 }, { "grad_norm": 0.1718987226486206, "learning_rate": 2.060265738465034e-06, "loss": 0.0055, "step": 91290 }, { "grad_norm": 0.13295549154281616, "learning_rate": 2.055570853807487e-06, "loss": 0.0037, "step": 91300 }, { "grad_norm": 0.08907502144575119, "learning_rate": 2.0508812122806196e-06, "loss": 0.0053, "step": 91310 }, { "grad_norm": 0.1314365267753601, "learning_rate": 2.046196814397294e-06, "loss": 0.0047, "step": 91320 }, { "grad_norm": 0.054105859249830246, "learning_rate": 2.0415176606697896e-06, "loss": 0.0054, "step": 91330 }, { "grad_norm": 0.13360360264778137, "learning_rate": 2.0368437516098126e-06, "loss": 0.0051, "step": 91340 }, { "grad_norm": 0.1686224639415741, "learning_rate": 2.032175087728483e-06, "loss": 0.0054, "step": 91350 }, { "grad_norm": 0.11108633130788803, "learning_rate": 2.027511669536375e-06, "loss": 0.0055, "step": 91360 }, { "grad_norm": 0.06751824915409088, "learning_rate": 2.0228534975434643e-06, "loss": 0.0044, "step": 91370 }, { "grad_norm": 0.12440290302038193, "learning_rate": 2.0182005722591548e-06, "loss": 0.0061, "step": 91380 }, { "grad_norm": 0.12711913883686066, "learning_rate": 2.0135528941922944e-06, "loss": 0.0047, "step": 91390 }, { "grad_norm": 0.061366256326436996, "learning_rate": 2.008910463851149e-06, "loss": 0.004, "step": 91400 }, { "grad_norm": 0.11968400329351425, "learning_rate": 2.0042732817433963e-06, "loss": 0.0072, "step": 91410 }, { "grad_norm": 0.1592041254043579, "learning_rate": 1.999641348376158e-06, "loss": 0.0059, "step": 91420 }, { "grad_norm": 0.07295209914445877, "learning_rate": 1.9950146642559787e-06, "loss": 0.0052, "step": 91430 }, { "grad_norm": 0.13269373774528503, "learning_rate": 1.990393229888815e-06, "loss": 0.0036, "step": 91440 }, { "grad_norm": 0.23078468441963196, "learning_rate": 1.9857770457800684e-06, "loss": 0.0083, "step": 91450 }, { "grad_norm": 0.07266281545162201, "learning_rate": 1.9811661124345517e-06, "loss": 0.0044, "step": 91460 }, { "grad_norm": 0.07021720707416534, "learning_rate": 1.976560430356522e-06, "loss": 0.0053, "step": 91470 }, { "grad_norm": 0.12323281168937683, "learning_rate": 1.9719600000496383e-06, "loss": 0.0062, "step": 91480 }, { "grad_norm": 0.18759214878082275, "learning_rate": 1.967364822017004e-06, "loss": 0.0067, "step": 91490 }, { "grad_norm": 0.08744405210018158, "learning_rate": 1.962774896761127e-06, "loss": 0.0075, "step": 91500 }, { "grad_norm": 0.10887906700372696, "learning_rate": 1.958190224783968e-06, "loss": 0.0069, "step": 91510 }, { "grad_norm": 0.1050431877374649, "learning_rate": 1.9536108065868975e-06, "loss": 0.0047, "step": 91520 }, { "grad_norm": 0.12588021159172058, "learning_rate": 1.949036642670715e-06, "loss": 0.0079, "step": 91530 }, { "grad_norm": 0.077739417552948, "learning_rate": 1.9444677335356377e-06, "loss": 0.0064, "step": 91540 }, { "grad_norm": 0.06868235766887665, "learning_rate": 1.939904079681326e-06, "loss": 0.0046, "step": 91550 }, { "grad_norm": 0.1374119222164154, "learning_rate": 1.935345681606837e-06, "loss": 0.0065, "step": 91560 }, { "grad_norm": 0.09846065193414688, "learning_rate": 1.930792539810694e-06, "loss": 0.0046, "step": 91570 }, { "grad_norm": 0.13445913791656494, "learning_rate": 1.9262446547907877e-06, "loss": 0.0055, "step": 91580 }, { "grad_norm": 0.2019723653793335, "learning_rate": 1.921702027044503e-06, "loss": 0.0057, "step": 91590 }, { "grad_norm": 0.1437223255634308, "learning_rate": 1.917164657068593e-06, "loss": 0.01, "step": 91600 }, { "grad_norm": 0.15888948738574982, "learning_rate": 1.912632545359272e-06, "loss": 0.0047, "step": 91610 }, { "grad_norm": 0.12605158984661102, "learning_rate": 1.9081056924121488e-06, "loss": 0.0052, "step": 91620 }, { "grad_norm": 0.13986822962760925, "learning_rate": 1.9035840987222887e-06, "loss": 0.0061, "step": 91630 }, { "grad_norm": 0.39643967151641846, "learning_rate": 1.8990677647841515e-06, "loss": 0.0044, "step": 91640 }, { "grad_norm": 0.09988036751747131, "learning_rate": 1.8945566910916534e-06, "loss": 0.0049, "step": 91650 }, { "grad_norm": 0.14961287379264832, "learning_rate": 1.8900508781381054e-06, "loss": 0.0051, "step": 91660 }, { "grad_norm": 0.1062810868024826, "learning_rate": 1.8855503264162633e-06, "loss": 0.0043, "step": 91670 }, { "grad_norm": 0.220185786485672, "learning_rate": 1.8810550364182998e-06, "loss": 0.0074, "step": 91680 }, { "grad_norm": 0.14203673601150513, "learning_rate": 1.8765650086358112e-06, "loss": 0.0053, "step": 91690 }, { "grad_norm": 0.08559176325798035, "learning_rate": 1.8720802435598151e-06, "loss": 0.0052, "step": 91700 }, { "grad_norm": 0.09665771573781967, "learning_rate": 1.8676007416807694e-06, "loss": 0.0035, "step": 91710 }, { "grad_norm": 0.16912217438220978, "learning_rate": 1.8631265034885437e-06, "loss": 0.0056, "step": 91720 }, { "grad_norm": 0.11013130843639374, "learning_rate": 1.858657529472424e-06, "loss": 0.0066, "step": 91730 }, { "grad_norm": 0.09632819145917892, "learning_rate": 1.8541938201211473e-06, "loss": 0.0076, "step": 91740 }, { "grad_norm": 0.07675432413816452, "learning_rate": 1.8497353759228398e-06, "loss": 0.0042, "step": 91750 }, { "grad_norm": 0.07889509946107864, "learning_rate": 1.8452821973650835e-06, "loss": 0.0062, "step": 91760 }, { "grad_norm": 0.12518014013767242, "learning_rate": 1.840834284934856e-06, "loss": 0.0033, "step": 91770 }, { "grad_norm": 0.13998211920261383, "learning_rate": 1.8363916391185954e-06, "loss": 0.0055, "step": 91780 }, { "grad_norm": 0.09789878875017166, "learning_rate": 1.8319542604021245e-06, "loss": 0.004, "step": 91790 }, { "grad_norm": 0.055361032485961914, "learning_rate": 1.8275221492707273e-06, "loss": 0.0037, "step": 91800 }, { "grad_norm": 0.06990678608417511, "learning_rate": 1.8230953062090717e-06, "loss": 0.005, "step": 91810 }, { "grad_norm": 0.10729849338531494, "learning_rate": 1.8186737317012926e-06, "loss": 0.004, "step": 91820 }, { "grad_norm": 0.077479287981987, "learning_rate": 1.8142574262308977e-06, "loss": 0.004, "step": 91830 }, { "grad_norm": 0.09443369507789612, "learning_rate": 1.8098463902808727e-06, "loss": 0.0061, "step": 91840 }, { "grad_norm": 0.0803578644990921, "learning_rate": 1.805440624333593e-06, "loss": 0.0042, "step": 91850 }, { "grad_norm": 0.14953362941741943, "learning_rate": 1.8010401288708734e-06, "loss": 0.0063, "step": 91860 }, { "grad_norm": 0.11978531628847122, "learning_rate": 1.7966449043739343e-06, "loss": 0.0043, "step": 91870 }, { "grad_norm": 0.1495368480682373, "learning_rate": 1.79225495132343e-06, "loss": 0.0047, "step": 91880 }, { "grad_norm": 0.09824954718351364, "learning_rate": 1.7878702701994487e-06, "loss": 0.005, "step": 91890 }, { "grad_norm": 0.187781423330307, "learning_rate": 1.783490861481496e-06, "loss": 0.0047, "step": 91900 }, { "grad_norm": 0.28371331095695496, "learning_rate": 1.7791167256484832e-06, "loss": 0.0042, "step": 91910 }, { "grad_norm": 0.16466622054576874, "learning_rate": 1.774747863178766e-06, "loss": 0.0071, "step": 91920 }, { "grad_norm": 0.0946974903345108, "learning_rate": 1.7703842745501186e-06, "loss": 0.006, "step": 91930 }, { "grad_norm": 0.06045278534293175, "learning_rate": 1.7660259602397366e-06, "loss": 0.0042, "step": 91940 }, { "grad_norm": 0.09925368428230286, "learning_rate": 1.7616729207242388e-06, "loss": 0.0052, "step": 91950 }, { "grad_norm": 0.0865471363067627, "learning_rate": 1.7573251564796556e-06, "loss": 0.0046, "step": 91960 }, { "grad_norm": 0.08838066458702087, "learning_rate": 1.7529826679814676e-06, "loss": 0.0053, "step": 91970 }, { "grad_norm": 0.09708571434020996, "learning_rate": 1.7486454557045562e-06, "loss": 0.0049, "step": 91980 }, { "grad_norm": 0.200869619846344, "learning_rate": 1.7443135201232364e-06, "loss": 0.0104, "step": 91990 }, { "grad_norm": 0.14162643253803253, "learning_rate": 1.7399868617112347e-06, "loss": 0.0052, "step": 92000 }, { "grad_norm": 0.1129709854722023, "learning_rate": 1.7356654809417171e-06, "loss": 0.005, "step": 92010 }, { "grad_norm": 0.1080775186419487, "learning_rate": 1.7313493782872448e-06, "loss": 0.0047, "step": 92020 }, { "grad_norm": 0.06084000691771507, "learning_rate": 1.7270385542198453e-06, "loss": 0.0047, "step": 92030 }, { "grad_norm": 0.08088790625333786, "learning_rate": 1.7227330092109307e-06, "loss": 0.0045, "step": 92040 }, { "grad_norm": 0.049856435507535934, "learning_rate": 1.718432743731352e-06, "loss": 0.0044, "step": 92050 }, { "grad_norm": 0.12417711317539215, "learning_rate": 1.7141377582513718e-06, "loss": 0.0045, "step": 92060 }, { "grad_norm": 0.09872523695230484, "learning_rate": 1.709848053240698e-06, "loss": 0.0052, "step": 92070 }, { "grad_norm": 0.07720539718866348, "learning_rate": 1.7055636291684328e-06, "loss": 0.0076, "step": 92080 }, { "grad_norm": 0.061281170696020126, "learning_rate": 1.7012844865031185e-06, "loss": 0.0081, "step": 92090 }, { "grad_norm": 0.1747591346502304, "learning_rate": 1.6970106257127138e-06, "loss": 0.0043, "step": 92100 }, { "grad_norm": 0.07764153182506561, "learning_rate": 1.6927420472646117e-06, "loss": 0.0063, "step": 92110 }, { "grad_norm": 0.06647689640522003, "learning_rate": 1.6884787516256106e-06, "loss": 0.0056, "step": 92120 }, { "grad_norm": 0.1299452930688858, "learning_rate": 1.6842207392619325e-06, "loss": 0.0059, "step": 92130 }, { "grad_norm": 0.16958607733249664, "learning_rate": 1.679968010639238e-06, "loss": 0.0049, "step": 92140 }, { "grad_norm": 0.17646630108356476, "learning_rate": 1.6757205662225829e-06, "loss": 0.0045, "step": 92150 }, { "grad_norm": 0.14636175334453583, "learning_rate": 1.6714784064764844e-06, "loss": 0.0054, "step": 92160 }, { "grad_norm": 0.10679754614830017, "learning_rate": 1.6672415318648383e-06, "loss": 0.004, "step": 92170 }, { "grad_norm": 0.1679804027080536, "learning_rate": 1.6630099428510016e-06, "loss": 0.0052, "step": 92180 }, { "grad_norm": 0.11646803468465805, "learning_rate": 1.6587836398977153e-06, "loss": 0.0039, "step": 92190 }, { "grad_norm": 0.1850917488336563, "learning_rate": 1.6545626234671762e-06, "loss": 0.0061, "step": 92200 }, { "grad_norm": 0.08586125820875168, "learning_rate": 1.650346894020971e-06, "loss": 0.0047, "step": 92210 }, { "grad_norm": 0.0780746191740036, "learning_rate": 1.6461364520201472e-06, "loss": 0.0052, "step": 92220 }, { "grad_norm": 0.10739535838365555, "learning_rate": 1.6419312979251366e-06, "loss": 0.0069, "step": 92230 }, { "grad_norm": 0.1165233850479126, "learning_rate": 1.6377314321958214e-06, "loss": 0.0073, "step": 92240 }, { "grad_norm": 0.09576370567083359, "learning_rate": 1.6335368552914787e-06, "loss": 0.0045, "step": 92250 }, { "grad_norm": 0.16869641840457916, "learning_rate": 1.6293475676708358e-06, "loss": 0.0085, "step": 92260 }, { "grad_norm": 0.08789016306400299, "learning_rate": 1.6251635697920097e-06, "loss": 0.0048, "step": 92270 }, { "grad_norm": 0.1111140251159668, "learning_rate": 1.6209848621125624e-06, "loss": 0.005, "step": 92280 }, { "grad_norm": 0.1025606021285057, "learning_rate": 1.6168114450894778e-06, "loss": 0.0047, "step": 92290 }, { "grad_norm": 0.12379976361989975, "learning_rate": 1.612643319179158e-06, "loss": 0.0047, "step": 92300 }, { "grad_norm": 0.08372961729764938, "learning_rate": 1.6084804848374046e-06, "loss": 0.006, "step": 92310 }, { "grad_norm": 0.0819060206413269, "learning_rate": 1.6043229425194817e-06, "loss": 0.0041, "step": 92320 }, { "grad_norm": 0.17568987607955933, "learning_rate": 1.6001706926800308e-06, "loss": 0.0042, "step": 92330 }, { "grad_norm": 0.09414684772491455, "learning_rate": 1.5960237357731444e-06, "loss": 0.0061, "step": 92340 }, { "grad_norm": 0.0701872706413269, "learning_rate": 1.5918820722523264e-06, "loss": 0.0037, "step": 92350 }, { "grad_norm": 0.07372226566076279, "learning_rate": 1.587745702570509e-06, "loss": 0.0049, "step": 92360 }, { "grad_norm": 0.07990241050720215, "learning_rate": 1.5836146271800356e-06, "loss": 0.0067, "step": 92370 }, { "grad_norm": 0.11615332216024399, "learning_rate": 1.5794888465326728e-06, "loss": 0.0058, "step": 92380 }, { "grad_norm": 0.08295176923274994, "learning_rate": 1.5753683610796155e-06, "loss": 0.0048, "step": 92390 }, { "grad_norm": 0.10860440135002136, "learning_rate": 1.5712531712714528e-06, "loss": 0.0058, "step": 92400 }, { "grad_norm": 0.10655976831912994, "learning_rate": 1.5671432775582474e-06, "loss": 0.0051, "step": 92410 }, { "grad_norm": 0.07901829481124878, "learning_rate": 1.563038680389428e-06, "loss": 0.0045, "step": 92420 }, { "grad_norm": 0.16776785254478455, "learning_rate": 1.5589393802138808e-06, "loss": 0.0043, "step": 92430 }, { "grad_norm": 0.10069624334573746, "learning_rate": 1.554845377479891e-06, "loss": 0.0034, "step": 92440 }, { "grad_norm": 0.1066424697637558, "learning_rate": 1.550756672635173e-06, "loss": 0.0041, "step": 92450 }, { "grad_norm": 0.1565665751695633, "learning_rate": 1.5466732661268635e-06, "loss": 0.0055, "step": 92460 }, { "grad_norm": 0.09513524174690247, "learning_rate": 1.5425951584015163e-06, "loss": 0.0048, "step": 92470 }, { "grad_norm": 0.12601566314697266, "learning_rate": 1.5385223499051026e-06, "loss": 0.0038, "step": 92480 }, { "grad_norm": 0.06361138820648193, "learning_rate": 1.5344548410830328e-06, "loss": 0.0045, "step": 92490 }, { "grad_norm": 0.058011744171381, "learning_rate": 1.5303926323801121e-06, "loss": 0.0044, "step": 92500 }, { "grad_norm": 0.15066979825496674, "learning_rate": 1.5263357242405795e-06, "loss": 0.0046, "step": 92510 }, { "grad_norm": 0.10025262087583542, "learning_rate": 1.5222841171080915e-06, "loss": 0.0045, "step": 92520 }, { "grad_norm": 0.08282008767127991, "learning_rate": 1.5182378114257268e-06, "loss": 0.0053, "step": 92530 }, { "grad_norm": 0.09696304053068161, "learning_rate": 1.5141968076359869e-06, "loss": 0.0045, "step": 92540 }, { "grad_norm": 0.10409887880086899, "learning_rate": 1.5101611061807908e-06, "loss": 0.0037, "step": 92550 }, { "grad_norm": 0.09555037319660187, "learning_rate": 1.506130707501463e-06, "loss": 0.0043, "step": 92560 }, { "grad_norm": 0.058751072734594345, "learning_rate": 1.5021056120387734e-06, "loss": 0.0056, "step": 92570 }, { "grad_norm": 0.08014004677534103, "learning_rate": 1.4980858202329084e-06, "loss": 0.005, "step": 92580 }, { "grad_norm": 0.11159226298332214, "learning_rate": 1.4940713325234445e-06, "loss": 0.0054, "step": 92590 }, { "grad_norm": 0.09266047179698944, "learning_rate": 1.4900621493494137e-06, "loss": 0.0052, "step": 92600 }, { "grad_norm": 0.14159534871578217, "learning_rate": 1.4860582711492544e-06, "loss": 0.0061, "step": 92610 }, { "grad_norm": 0.11809170246124268, "learning_rate": 1.4820596983608325e-06, "loss": 0.006, "step": 92620 }, { "grad_norm": 0.1063227429986, "learning_rate": 1.4780664314214044e-06, "loss": 0.0057, "step": 92630 }, { "grad_norm": 0.05154542997479439, "learning_rate": 1.474078470767687e-06, "loss": 0.0051, "step": 92640 }, { "grad_norm": 0.06539548188447952, "learning_rate": 1.470095816835787e-06, "loss": 0.0048, "step": 92650 }, { "grad_norm": 0.10306786745786667, "learning_rate": 1.4661184700612395e-06, "loss": 0.0044, "step": 92660 }, { "grad_norm": 0.17421434819698334, "learning_rate": 1.4621464308790133e-06, "loss": 0.0073, "step": 92670 }, { "grad_norm": 0.10185325890779495, "learning_rate": 1.4581796997234775e-06, "loss": 0.0063, "step": 92680 }, { "grad_norm": 0.1105353906750679, "learning_rate": 1.4542182770284297e-06, "loss": 0.0086, "step": 92690 }, { "grad_norm": 0.12262455374002457, "learning_rate": 1.450262163227084e-06, "loss": 0.0039, "step": 92700 }, { "grad_norm": 0.16472849249839783, "learning_rate": 1.4463113587520671e-06, "loss": 0.0048, "step": 92710 }, { "grad_norm": 0.06025064364075661, "learning_rate": 1.4423658640354443e-06, "loss": 0.0047, "step": 92720 }, { "grad_norm": 0.10785244405269623, "learning_rate": 1.438425679508687e-06, "loss": 0.0062, "step": 92730 }, { "grad_norm": 0.11859311163425446, "learning_rate": 1.4344908056026895e-06, "loss": 0.0046, "step": 92740 }, { "grad_norm": 0.15385319292545319, "learning_rate": 1.4305612427477521e-06, "loss": 0.0042, "step": 92750 }, { "grad_norm": 0.1536693572998047, "learning_rate": 1.42663699137362e-06, "loss": 0.0056, "step": 92760 }, { "grad_norm": 0.29468637704849243, "learning_rate": 1.4227180519094386e-06, "loss": 0.0082, "step": 92770 }, { "grad_norm": 0.0793660432100296, "learning_rate": 1.4188044247837706e-06, "loss": 0.0045, "step": 92780 }, { "grad_norm": 0.08915164321660995, "learning_rate": 1.4148961104246128e-06, "loss": 0.0044, "step": 92790 }, { "grad_norm": 0.13403619825839996, "learning_rate": 1.4109931092593731e-06, "loss": 0.0049, "step": 92800 }, { "grad_norm": 0.13400834798812866, "learning_rate": 1.4070954217148658e-06, "loss": 0.0072, "step": 92810 }, { "grad_norm": 0.11385687440633774, "learning_rate": 1.4032030482173497e-06, "loss": 0.0055, "step": 92820 }, { "grad_norm": 0.08532042056322098, "learning_rate": 1.3993159891924901e-06, "loss": 0.0051, "step": 92830 }, { "grad_norm": 0.07984241098165512, "learning_rate": 1.3954342450653524e-06, "loss": 0.0056, "step": 92840 }, { "grad_norm": 0.16376635432243347, "learning_rate": 1.3915578162604582e-06, "loss": 0.0052, "step": 92850 }, { "grad_norm": 0.1365293264389038, "learning_rate": 1.3876867032017126e-06, "loss": 0.0069, "step": 92860 }, { "grad_norm": 0.06269379705190659, "learning_rate": 1.383820906312472e-06, "loss": 0.0041, "step": 92870 }, { "grad_norm": 0.099045529961586, "learning_rate": 1.3799604260154752e-06, "loss": 0.0062, "step": 92880 }, { "grad_norm": 0.0896124318242073, "learning_rate": 1.376105262732913e-06, "loss": 0.0052, "step": 92890 }, { "grad_norm": 0.06359021365642548, "learning_rate": 1.37225541688637e-06, "loss": 0.005, "step": 92900 }, { "grad_norm": 0.22171208262443542, "learning_rate": 1.3684108888968594e-06, "loss": 0.0029, "step": 92910 }, { "grad_norm": 0.0959465354681015, "learning_rate": 1.364571679184823e-06, "loss": 0.0053, "step": 92920 }, { "grad_norm": 0.21689897775650024, "learning_rate": 1.3607377881701077e-06, "loss": 0.0051, "step": 92930 }, { "grad_norm": 0.1646483838558197, "learning_rate": 1.3569092162719787e-06, "loss": 0.0039, "step": 92940 }, { "grad_norm": 0.05504989251494408, "learning_rate": 1.353085963909123e-06, "loss": 0.0067, "step": 92950 }, { "grad_norm": 0.04271689057350159, "learning_rate": 1.3492680314996454e-06, "loss": 0.0048, "step": 92960 }, { "grad_norm": 0.1644197255373001, "learning_rate": 1.345455419461078e-06, "loss": 0.0052, "step": 92970 }, { "grad_norm": 0.040890540927648544, "learning_rate": 1.3416481282103431e-06, "loss": 0.0045, "step": 92980 }, { "grad_norm": 0.11245812475681305, "learning_rate": 1.337846158163819e-06, "loss": 0.004, "step": 92990 }, { "grad_norm": 0.10001321882009506, "learning_rate": 1.334049509737273e-06, "loss": 0.0037, "step": 93000 }, { "grad_norm": 0.11747227609157562, "learning_rate": 1.3302581833459116e-06, "loss": 0.006, "step": 93010 }, { "grad_norm": 0.1883627027273178, "learning_rate": 1.3264721794043366e-06, "loss": 0.0061, "step": 93020 }, { "grad_norm": 0.10419096052646637, "learning_rate": 1.3226914983265836e-06, "loss": 0.0044, "step": 93030 }, { "grad_norm": 0.10150185227394104, "learning_rate": 1.3189161405260997e-06, "loss": 0.0043, "step": 93040 }, { "grad_norm": 0.16057713329792023, "learning_rate": 1.3151461064157656e-06, "loss": 0.0059, "step": 93050 }, { "grad_norm": 0.10478012263774872, "learning_rate": 1.3113813964078459e-06, "loss": 0.0041, "step": 93060 }, { "grad_norm": 0.12434237450361252, "learning_rate": 1.3076220109140614e-06, "loss": 0.0042, "step": 93070 }, { "grad_norm": 0.10088886320590973, "learning_rate": 1.3038679503455275e-06, "loss": 0.0038, "step": 93080 }, { "grad_norm": 0.08695521205663681, "learning_rate": 1.3001192151127773e-06, "loss": 0.0066, "step": 93090 }, { "grad_norm": 0.10846874862909317, "learning_rate": 1.2963758056257657e-06, "loss": 0.006, "step": 93100 }, { "grad_norm": 0.14146779477596283, "learning_rate": 1.2926377222938768e-06, "loss": 0.0048, "step": 93110 }, { "grad_norm": 0.13041375577449799, "learning_rate": 1.2889049655259e-06, "loss": 0.0044, "step": 93120 }, { "grad_norm": 0.26490703225135803, "learning_rate": 1.2851775357300312e-06, "loss": 0.0044, "step": 93130 }, { "grad_norm": 0.10756327211856842, "learning_rate": 1.281455433313916e-06, "loss": 0.0055, "step": 93140 }, { "grad_norm": 0.10701080411672592, "learning_rate": 1.2777386586845796e-06, "loss": 0.0058, "step": 93150 }, { "grad_norm": 0.22023002803325653, "learning_rate": 1.2740272122484964e-06, "loss": 0.0054, "step": 93160 }, { "grad_norm": 0.13452407717704773, "learning_rate": 1.2703210944115308e-06, "loss": 0.0044, "step": 93170 }, { "grad_norm": 0.11342748999595642, "learning_rate": 1.2666203055789916e-06, "loss": 0.0065, "step": 93180 }, { "grad_norm": 0.07829087972640991, "learning_rate": 1.2629248461555832e-06, "loss": 0.0044, "step": 93190 }, { "grad_norm": 0.10525622963905334, "learning_rate": 1.2592347165454431e-06, "loss": 0.0047, "step": 93200 }, { "grad_norm": 0.08403483033180237, "learning_rate": 1.2555499171521102e-06, "loss": 0.005, "step": 93210 }, { "grad_norm": 0.11583579331636429, "learning_rate": 1.251870448378556e-06, "loss": 0.0058, "step": 93220 }, { "grad_norm": 0.06335458159446716, "learning_rate": 1.2481963106271478e-06, "loss": 0.0041, "step": 93230 }, { "grad_norm": 0.06267756968736649, "learning_rate": 1.2445275042997084e-06, "loss": 0.0093, "step": 93240 }, { "grad_norm": 0.18222779035568237, "learning_rate": 1.2408640297974283e-06, "loss": 0.0051, "step": 93250 }, { "grad_norm": 0.259690523147583, "learning_rate": 1.2372058875209535e-06, "loss": 0.0058, "step": 93260 }, { "grad_norm": 0.05834018811583519, "learning_rate": 1.2335530778703252e-06, "loss": 0.0036, "step": 93270 }, { "grad_norm": 0.15020035207271576, "learning_rate": 1.2299056012450182e-06, "loss": 0.0051, "step": 93280 }, { "grad_norm": 0.059095416218042374, "learning_rate": 1.2262634580439026e-06, "loss": 0.0048, "step": 93290 }, { "grad_norm": 0.2527061998844147, "learning_rate": 1.2226266486652926e-06, "loss": 0.0056, "step": 93300 }, { "grad_norm": 0.13587428629398346, "learning_rate": 1.2189951735068927e-06, "loss": 0.0041, "step": 93310 }, { "grad_norm": 0.1209862232208252, "learning_rate": 1.2153690329658407e-06, "loss": 0.0057, "step": 93320 }, { "grad_norm": 0.11210745573043823, "learning_rate": 1.2117482274386859e-06, "loss": 0.0058, "step": 93330 }, { "grad_norm": 0.09648752957582474, "learning_rate": 1.2081327573213953e-06, "loss": 0.0041, "step": 93340 }, { "grad_norm": 0.09776918590068817, "learning_rate": 1.2045226230093465e-06, "loss": 0.0048, "step": 93350 }, { "grad_norm": 0.2476254403591156, "learning_rate": 1.2009178248973296e-06, "loss": 0.0048, "step": 93360 }, { "grad_norm": 0.07957419008016586, "learning_rate": 1.1973183633795849e-06, "loss": 0.0044, "step": 93370 }, { "grad_norm": 0.05972859635949135, "learning_rate": 1.193724238849725e-06, "loss": 0.0066, "step": 93380 }, { "grad_norm": 0.24863292276859283, "learning_rate": 1.190135451700808e-06, "loss": 0.0092, "step": 93390 }, { "grad_norm": 0.07271896302700043, "learning_rate": 1.186552002325292e-06, "loss": 0.0051, "step": 93400 }, { "grad_norm": 0.14880666136741638, "learning_rate": 1.1829738911150578e-06, "loss": 0.0056, "step": 93410 }, { "grad_norm": 0.09640047699213028, "learning_rate": 1.1794011184613985e-06, "loss": 0.0054, "step": 93420 }, { "grad_norm": 0.08871536701917648, "learning_rate": 1.17583368475504e-06, "loss": 0.0063, "step": 93430 }, { "grad_norm": 0.10877107083797455, "learning_rate": 1.1722715903860982e-06, "loss": 0.0047, "step": 93440 }, { "grad_norm": 0.07538524270057678, "learning_rate": 1.168714835744128e-06, "loss": 0.0068, "step": 93450 }, { "grad_norm": 0.07526955008506775, "learning_rate": 1.165163421218085e-06, "loss": 0.0056, "step": 93460 }, { "grad_norm": 0.13052205741405487, "learning_rate": 1.1616173471963477e-06, "loss": 0.0042, "step": 93470 }, { "grad_norm": 0.09319079667329788, "learning_rate": 1.1580766140667109e-06, "loss": 0.005, "step": 93480 }, { "grad_norm": 0.09220407903194427, "learning_rate": 1.1545412222163764e-06, "loss": 0.0059, "step": 93490 }, { "grad_norm": 0.11661948263645172, "learning_rate": 1.1510111720319795e-06, "loss": 0.0057, "step": 93500 }, { "grad_norm": 0.07199323922395706, "learning_rate": 1.1474864638995663e-06, "loss": 0.0045, "step": 93510 }, { "grad_norm": 0.22855912148952484, "learning_rate": 1.1439670982045735e-06, "loss": 0.0046, "step": 93520 }, { "grad_norm": 0.07452395558357239, "learning_rate": 1.1404530753318877e-06, "loss": 0.005, "step": 93530 }, { "grad_norm": 0.23674379289150238, "learning_rate": 1.136944395665801e-06, "loss": 0.005, "step": 93540 }, { "grad_norm": 0.052943482995033264, "learning_rate": 1.1334410595900014e-06, "loss": 0.0052, "step": 93550 }, { "grad_norm": 0.09948223829269409, "learning_rate": 1.129943067487621e-06, "loss": 0.004, "step": 93560 }, { "grad_norm": 0.05539017915725708, "learning_rate": 1.126450419741193e-06, "loss": 0.0054, "step": 93570 }, { "grad_norm": 0.14960235357284546, "learning_rate": 1.1229631167326671e-06, "loss": 0.0042, "step": 93580 }, { "grad_norm": 0.13911749422550201, "learning_rate": 1.1194811588434052e-06, "loss": 0.0062, "step": 93590 }, { "grad_norm": 0.5041880011558533, "learning_rate": 1.1160045464542023e-06, "loss": 0.0063, "step": 93600 }, { "grad_norm": 0.08889248967170715, "learning_rate": 1.1125332799452325e-06, "loss": 0.004, "step": 93610 }, { "grad_norm": 0.07066291570663452, "learning_rate": 1.109067359696131e-06, "loss": 0.0055, "step": 93620 }, { "grad_norm": 0.334885835647583, "learning_rate": 1.1056067860859166e-06, "loss": 0.0045, "step": 93630 }, { "grad_norm": 0.06390806287527084, "learning_rate": 1.102151559493031e-06, "loss": 0.0053, "step": 93640 }, { "grad_norm": 0.11455720663070679, "learning_rate": 1.0987016802953332e-06, "loss": 0.0044, "step": 93650 }, { "grad_norm": 0.09570293873548508, "learning_rate": 1.0952571488701048e-06, "loss": 0.005, "step": 93660 }, { "grad_norm": 0.16122573614120483, "learning_rate": 1.0918179655940163e-06, "loss": 0.0049, "step": 93670 }, { "grad_norm": 0.1128150224685669, "learning_rate": 1.088384130843184e-06, "loss": 0.004, "step": 93680 }, { "grad_norm": 0.10692022740840912, "learning_rate": 1.0849556449931296e-06, "loss": 0.0039, "step": 93690 }, { "grad_norm": 0.09275901317596436, "learning_rate": 1.0815325084187866e-06, "loss": 0.0037, "step": 93700 }, { "grad_norm": 0.07620471715927124, "learning_rate": 1.0781147214944943e-06, "loss": 0.0069, "step": 93710 }, { "grad_norm": 0.15523917973041534, "learning_rate": 1.0747022845940313e-06, "loss": 0.0047, "step": 93720 }, { "grad_norm": 0.1456558257341385, "learning_rate": 1.0712951980905605e-06, "loss": 0.0039, "step": 93730 }, { "grad_norm": 0.14855241775512695, "learning_rate": 1.0678934623566839e-06, "loss": 0.0048, "step": 93740 }, { "grad_norm": 0.06449158489704132, "learning_rate": 1.0644970777644091e-06, "loss": 0.0061, "step": 93750 }, { "grad_norm": 0.10118864476680756, "learning_rate": 1.0611060446851673e-06, "loss": 0.0059, "step": 93760 }, { "grad_norm": 0.09972354769706726, "learning_rate": 1.0577203634897836e-06, "loss": 0.0033, "step": 93770 }, { "grad_norm": 0.08705025911331177, "learning_rate": 1.0543400345485178e-06, "loss": 0.0038, "step": 93780 }, { "grad_norm": 0.09001243859529495, "learning_rate": 1.0509650582310404e-06, "loss": 0.0042, "step": 93790 }, { "grad_norm": 0.10569191724061966, "learning_rate": 1.0475954349064287e-06, "loss": 0.0043, "step": 93800 }, { "grad_norm": 0.11953893303871155, "learning_rate": 1.0442311649431825e-06, "loss": 0.004, "step": 93810 }, { "grad_norm": 0.1353771686553955, "learning_rate": 1.0408722487092071e-06, "loss": 0.004, "step": 93820 }, { "grad_norm": 0.098187156021595, "learning_rate": 1.0375186865718424e-06, "loss": 0.0037, "step": 93830 }, { "grad_norm": 0.16417190432548523, "learning_rate": 1.0341704788978113e-06, "loss": 0.0041, "step": 93840 }, { "grad_norm": 0.09170478582382202, "learning_rate": 1.0308276260532879e-06, "loss": 0.0053, "step": 93850 }, { "grad_norm": 0.05424027517437935, "learning_rate": 1.0274901284038241e-06, "loss": 0.0055, "step": 93860 }, { "grad_norm": 0.12769146263599396, "learning_rate": 1.0241579863144114e-06, "loss": 0.0058, "step": 93870 }, { "grad_norm": 0.11261298507452011, "learning_rate": 1.0208312001494413e-06, "loss": 0.0045, "step": 93880 }, { "grad_norm": 0.13559265434741974, "learning_rate": 1.0175097702727453e-06, "loss": 0.0048, "step": 93890 }, { "grad_norm": 0.06156466528773308, "learning_rate": 1.0141936970475275e-06, "loss": 0.0073, "step": 93900 }, { "grad_norm": 0.10094882547855377, "learning_rate": 1.0108829808364417e-06, "loss": 0.0058, "step": 93910 }, { "grad_norm": 0.07820933312177658, "learning_rate": 1.0075776220015376e-06, "loss": 0.0051, "step": 93920 }, { "grad_norm": 0.1596112698316574, "learning_rate": 1.0042776209042816e-06, "loss": 0.0047, "step": 93930 }, { "grad_norm": 0.08146177232265472, "learning_rate": 1.000982977905568e-06, "loss": 0.0047, "step": 93940 }, { "grad_norm": 0.13466282188892365, "learning_rate": 9.976936933656867e-07, "loss": 0.0065, "step": 93950 }, { "grad_norm": 0.08946221321821213, "learning_rate": 9.944097676443442e-07, "loss": 0.007, "step": 93960 }, { "grad_norm": 0.10812084376811981, "learning_rate": 9.911312011006757e-07, "loss": 0.0043, "step": 93970 }, { "grad_norm": 0.08287100493907928, "learning_rate": 9.878579940932109e-07, "loss": 0.0038, "step": 93980 }, { "grad_norm": 0.05533892288804054, "learning_rate": 9.84590146979908e-07, "loss": 0.0032, "step": 93990 }, { "grad_norm": 0.08460884541273117, "learning_rate": 9.813276601181365e-07, "loss": 0.0034, "step": 94000 }, { "grad_norm": 0.0862787663936615, "learning_rate": 9.780705338646724e-07, "loss": 0.0048, "step": 94010 }, { "grad_norm": 0.10341894626617432, "learning_rate": 9.748187685757083e-07, "loss": 0.0046, "step": 94020 }, { "grad_norm": 0.08159507066011429, "learning_rate": 9.715723646068543e-07, "loss": 0.0043, "step": 94030 }, { "grad_norm": 0.11472996324300766, "learning_rate": 9.68331322313143e-07, "loss": 0.0046, "step": 94040 }, { "grad_norm": 0.05974862724542618, "learning_rate": 9.650956420489909e-07, "loss": 0.0043, "step": 94050 }, { "grad_norm": 0.10182969272136688, "learning_rate": 9.618653241682595e-07, "loss": 0.0053, "step": 94060 }, { "grad_norm": 0.17569610476493835, "learning_rate": 9.586403690242053e-07, "loss": 0.01, "step": 94070 }, { "grad_norm": 0.0971718356013298, "learning_rate": 9.554207769695122e-07, "loss": 0.0054, "step": 94080 }, { "grad_norm": 0.14412029087543488, "learning_rate": 9.522065483562659e-07, "loss": 0.0051, "step": 94090 }, { "grad_norm": 0.1108335331082344, "learning_rate": 9.489976835359681e-07, "loss": 0.0052, "step": 94100 }, { "grad_norm": 0.1723320037126541, "learning_rate": 9.457941828595329e-07, "loss": 0.006, "step": 94110 }, { "grad_norm": 0.1120363399386406, "learning_rate": 9.425960466772965e-07, "loss": 0.0047, "step": 94120 }, { "grad_norm": 0.07987991720438004, "learning_rate": 9.394032753390014e-07, "loss": 0.0049, "step": 94130 }, { "grad_norm": 0.08906961977481842, "learning_rate": 9.362158691938017e-07, "loss": 0.0058, "step": 94140 }, { "grad_norm": 0.13817909359931946, "learning_rate": 9.330338285902685e-07, "loss": 0.0043, "step": 94150 }, { "grad_norm": 0.17733459174633026, "learning_rate": 9.298571538763845e-07, "loss": 0.0042, "step": 94160 }, { "grad_norm": 0.19306069612503052, "learning_rate": 9.266858453995441e-07, "loss": 0.0052, "step": 94170 }, { "grad_norm": 0.06850971281528473, "learning_rate": 9.235199035065645e-07, "loss": 0.0054, "step": 94180 }, { "grad_norm": 0.054780565202236176, "learning_rate": 9.203593285436574e-07, "loss": 0.0057, "step": 94190 }, { "grad_norm": 0.08103136718273163, "learning_rate": 9.172041208564686e-07, "loss": 0.0061, "step": 94200 }, { "grad_norm": 0.10969987511634827, "learning_rate": 9.140542807900332e-07, "loss": 0.007, "step": 94210 }, { "grad_norm": 0.07494734227657318, "learning_rate": 9.109098086888313e-07, "loss": 0.0049, "step": 94220 }, { "grad_norm": 0.1466396450996399, "learning_rate": 9.077707048967265e-07, "loss": 0.0057, "step": 94230 }, { "grad_norm": 0.06046094372868538, "learning_rate": 9.046369697570057e-07, "loss": 0.0045, "step": 94240 }, { "grad_norm": 0.06423381716012955, "learning_rate": 9.015086036123721e-07, "loss": 0.0048, "step": 94250 }, { "grad_norm": 0.10073884576559067, "learning_rate": 8.983856068049413e-07, "loss": 0.0051, "step": 94260 }, { "grad_norm": 0.1169290617108345, "learning_rate": 8.952679796762398e-07, "loss": 0.0055, "step": 94270 }, { "grad_norm": 0.099754199385643, "learning_rate": 8.921557225671951e-07, "loss": 0.0043, "step": 94280 }, { "grad_norm": 0.07803881168365479, "learning_rate": 8.890488358181792e-07, "loss": 0.005, "step": 94290 }, { "grad_norm": 0.04746174067258835, "learning_rate": 8.859473197689372e-07, "loss": 0.0057, "step": 94300 }, { "grad_norm": 0.07117242366075516, "learning_rate": 8.828511747586588e-07, "loss": 0.006, "step": 94310 }, { "grad_norm": 0.05156068503856659, "learning_rate": 8.797604011259287e-07, "loss": 0.0049, "step": 94320 }, { "grad_norm": 0.0901227667927742, "learning_rate": 8.766749992087487e-07, "loss": 0.0042, "step": 94330 }, { "grad_norm": 0.09938153624534607, "learning_rate": 8.735949693445378e-07, "loss": 0.0038, "step": 94340 }, { "grad_norm": 0.13733331859111786, "learning_rate": 8.705203118701211e-07, "loss": 0.0072, "step": 94350 }, { "grad_norm": 0.12574662268161774, "learning_rate": 8.674510271217295e-07, "loss": 0.0051, "step": 94360 }, { "grad_norm": 0.08943372964859009, "learning_rate": 8.643871154350336e-07, "loss": 0.0045, "step": 94370 }, { "grad_norm": 0.1449301689863205, "learning_rate": 8.613285771450818e-07, "loss": 0.0061, "step": 94380 }, { "grad_norm": 0.10976339131593704, "learning_rate": 8.582754125863623e-07, "loss": 0.005, "step": 94390 }, { "grad_norm": 0.12218711525201797, "learning_rate": 8.55227622092758e-07, "loss": 0.0081, "step": 94400 }, { "grad_norm": 0.07723911106586456, "learning_rate": 8.521852059975743e-07, "loss": 0.0072, "step": 94410 }, { "grad_norm": 0.08372408151626587, "learning_rate": 8.491481646335231e-07, "loss": 0.0052, "step": 94420 }, { "grad_norm": 0.14839506149291992, "learning_rate": 8.46116498332733e-07, "loss": 0.0063, "step": 94430 }, { "grad_norm": 0.10165132582187653, "learning_rate": 8.430902074267333e-07, "loss": 0.0045, "step": 94440 }, { "grad_norm": 0.24484676122665405, "learning_rate": 8.400692922464925e-07, "loss": 0.0041, "step": 94450 }, { "grad_norm": 0.06931230425834656, "learning_rate": 8.370537531223577e-07, "loss": 0.0061, "step": 94460 }, { "grad_norm": 0.12899956107139587, "learning_rate": 8.340435903841092e-07, "loss": 0.0049, "step": 94470 }, { "grad_norm": 0.15575183928012848, "learning_rate": 8.310388043609341e-07, "loss": 0.005, "step": 94480 }, { "grad_norm": 0.1429363340139389, "learning_rate": 8.280393953814303e-07, "loss": 0.0067, "step": 94490 }, { "grad_norm": 0.12004587799310684, "learning_rate": 8.250453637736077e-07, "loss": 0.004, "step": 94500 }, { "grad_norm": 0.1124381273984909, "learning_rate": 8.22056709864899e-07, "loss": 0.0056, "step": 94510 }, { "grad_norm": 0.15169201791286469, "learning_rate": 8.190734339821204e-07, "loss": 0.0041, "step": 94520 }, { "grad_norm": 0.10214697569608688, "learning_rate": 8.160955364515333e-07, "loss": 0.0057, "step": 94530 }, { "grad_norm": 0.22646327316761017, "learning_rate": 8.131230175987936e-07, "loss": 0.0065, "step": 94540 }, { "grad_norm": 0.07591002434492111, "learning_rate": 8.101558777489693e-07, "loss": 0.0039, "step": 94550 }, { "grad_norm": 0.13491350412368774, "learning_rate": 8.071941172265396e-07, "loss": 0.0038, "step": 94560 }, { "grad_norm": 0.0829864889383316, "learning_rate": 8.042377363553955e-07, "loss": 0.005, "step": 94570 }, { "grad_norm": 0.0968816876411438, "learning_rate": 8.01286735458856e-07, "loss": 0.0062, "step": 94580 }, { "grad_norm": 0.07966436445713043, "learning_rate": 7.983411148596298e-07, "loss": 0.0036, "step": 94590 }, { "grad_norm": 0.09945443272590637, "learning_rate": 7.954008748798425e-07, "loss": 0.0041, "step": 94600 }, { "grad_norm": 0.14598913490772247, "learning_rate": 7.924660158410369e-07, "loss": 0.0046, "step": 94610 }, { "grad_norm": 0.06754540652036667, "learning_rate": 7.895365380641728e-07, "loss": 0.0045, "step": 94620 }, { "grad_norm": 0.149624764919281, "learning_rate": 7.866124418695941e-07, "loss": 0.0058, "step": 94630 }, { "grad_norm": 0.10603706538677216, "learning_rate": 7.83693727577095e-07, "loss": 0.0049, "step": 94640 }, { "grad_norm": 0.11421508342027664, "learning_rate": 7.807803955058535e-07, "loss": 0.0041, "step": 94650 }, { "grad_norm": 0.12209754437208176, "learning_rate": 7.778724459744702e-07, "loss": 0.0034, "step": 94660 }, { "grad_norm": 0.09311727434396744, "learning_rate": 7.749698793009464e-07, "loss": 0.0043, "step": 94670 }, { "grad_norm": 0.10957911610603333, "learning_rate": 7.720726958027169e-07, "loss": 0.0043, "step": 94680 }, { "grad_norm": 0.07350169122219086, "learning_rate": 7.691808957965951e-07, "loss": 0.0031, "step": 94690 }, { "grad_norm": 0.1586480289697647, "learning_rate": 7.662944795988336e-07, "loss": 0.0092, "step": 94700 }, { "grad_norm": 0.07289449870586395, "learning_rate": 7.634134475250854e-07, "loss": 0.0051, "step": 94710 }, { "grad_norm": 0.08564750850200653, "learning_rate": 7.605377998904262e-07, "loss": 0.0063, "step": 94720 }, { "grad_norm": 0.08770908415317535, "learning_rate": 7.5766753700931e-07, "loss": 0.0048, "step": 94730 }, { "grad_norm": 0.0837954506278038, "learning_rate": 7.54802659195647e-07, "loss": 0.0041, "step": 94740 }, { "grad_norm": 0.08744194358587265, "learning_rate": 7.519431667627253e-07, "loss": 0.0064, "step": 94750 }, { "grad_norm": 0.10169067978858948, "learning_rate": 7.490890600232503e-07, "loss": 0.0045, "step": 94760 }, { "grad_norm": 0.09943287819623947, "learning_rate": 7.462403392893502e-07, "loss": 0.0034, "step": 94770 }, { "grad_norm": 0.09096793830394745, "learning_rate": 7.433970048725536e-07, "loss": 0.0065, "step": 94780 }, { "grad_norm": 0.15965667366981506, "learning_rate": 7.405590570838062e-07, "loss": 0.0039, "step": 94790 }, { "grad_norm": 0.09751933068037033, "learning_rate": 7.377264962334596e-07, "loss": 0.0057, "step": 94800 }, { "grad_norm": 0.057911377400159836, "learning_rate": 7.348993226312773e-07, "loss": 0.0042, "step": 94810 }, { "grad_norm": 0.1861106902360916, "learning_rate": 7.32077536586434e-07, "loss": 0.0075, "step": 94820 }, { "grad_norm": 0.08396237343549728, "learning_rate": 7.292611384075221e-07, "loss": 0.0037, "step": 94830 }, { "grad_norm": 0.06927555799484253, "learning_rate": 7.264501284025338e-07, "loss": 0.0046, "step": 94840 }, { "grad_norm": 0.09240572899580002, "learning_rate": 7.236445068788789e-07, "loss": 0.0042, "step": 94850 }, { "grad_norm": 0.07712265104055405, "learning_rate": 7.20844274143373e-07, "loss": 0.0081, "step": 94860 }, { "grad_norm": 0.08969856798648834, "learning_rate": 7.180494305022489e-07, "loss": 0.0043, "step": 94870 }, { "grad_norm": 0.13234776258468628, "learning_rate": 7.152599762611456e-07, "loss": 0.0057, "step": 94880 }, { "grad_norm": 0.06376846134662628, "learning_rate": 7.124759117251079e-07, "loss": 0.0053, "step": 94890 }, { "grad_norm": 0.05538846552371979, "learning_rate": 7.096972371986033e-07, "loss": 0.005, "step": 94900 }, { "grad_norm": 0.12460027635097504, "learning_rate": 7.069239529855054e-07, "loss": 0.0048, "step": 94910 }, { "grad_norm": 0.14644385874271393, "learning_rate": 7.041560593890939e-07, "loss": 0.0072, "step": 94920 }, { "grad_norm": 0.07380053400993347, "learning_rate": 7.013935567120599e-07, "loss": 0.0043, "step": 94930 }, { "grad_norm": 0.07418476045131683, "learning_rate": 6.986364452565009e-07, "loss": 0.0037, "step": 94940 }, { "grad_norm": 0.08555971086025238, "learning_rate": 6.95884725323942e-07, "loss": 0.0057, "step": 94950 }, { "grad_norm": 0.18508656322956085, "learning_rate": 6.931383972152982e-07, "loss": 0.0058, "step": 94960 }, { "grad_norm": 0.07948920875787735, "learning_rate": 6.903974612309127e-07, "loss": 0.0046, "step": 94970 }, { "grad_norm": 0.18788397312164307, "learning_rate": 6.876619176705179e-07, "loss": 0.0043, "step": 94980 }, { "grad_norm": 0.0885678082704544, "learning_rate": 6.8493176683328e-07, "loss": 0.0044, "step": 94990 }, { "grad_norm": 0.0615290030837059, "learning_rate": 6.822070090177657e-07, "loss": 0.0049, "step": 95000 }, { "grad_norm": 0.1124802976846695, "learning_rate": 6.794876445219367e-07, "loss": 0.0044, "step": 95010 }, { "grad_norm": 0.131581649184227, "learning_rate": 6.767736736431885e-07, "loss": 0.0067, "step": 95020 }, { "grad_norm": 0.1473478376865387, "learning_rate": 6.740650966783113e-07, "loss": 0.004, "step": 95030 }, { "grad_norm": 0.07700604945421219, "learning_rate": 6.713619139235239e-07, "loss": 0.0043, "step": 95040 }, { "grad_norm": 0.17739900946617126, "learning_rate": 6.686641256744286e-07, "loss": 0.0041, "step": 95050 }, { "grad_norm": 0.07033514976501465, "learning_rate": 6.659717322260562e-07, "loss": 0.0045, "step": 95060 }, { "grad_norm": 0.09769029170274734, "learning_rate": 6.632847338728377e-07, "loss": 0.0062, "step": 95070 }, { "grad_norm": 0.1468537598848343, "learning_rate": 6.60603130908627e-07, "loss": 0.0051, "step": 95080 }, { "grad_norm": 0.08721039444208145, "learning_rate": 6.579269236266783e-07, "loss": 0.0051, "step": 95090 }, { "grad_norm": 0.10683049261569977, "learning_rate": 6.552561123196632e-07, "loss": 0.0046, "step": 95100 }, { "grad_norm": 0.12776288390159607, "learning_rate": 6.525906972796425e-07, "loss": 0.0055, "step": 95110 }, { "grad_norm": 0.16172492504119873, "learning_rate": 6.499306787981108e-07, "loss": 0.0038, "step": 95120 }, { "grad_norm": 0.0907851904630661, "learning_rate": 6.472760571659631e-07, "loss": 0.0051, "step": 95130 }, { "grad_norm": 0.17844228446483612, "learning_rate": 6.446268326735061e-07, "loss": 0.005, "step": 95140 }, { "grad_norm": 0.06807667762041092, "learning_rate": 6.41983005610458e-07, "loss": 0.0047, "step": 95150 }, { "grad_norm": 0.10766439139842987, "learning_rate": 6.393445762659378e-07, "loss": 0.0065, "step": 95160 }, { "grad_norm": 0.08436866849660873, "learning_rate": 6.36711544928481e-07, "loss": 0.0047, "step": 95170 }, { "grad_norm": 0.10855499655008316, "learning_rate": 6.340839118860298e-07, "loss": 0.0044, "step": 95180 }, { "grad_norm": 0.19122439622879028, "learning_rate": 6.314616774259485e-07, "loss": 0.0044, "step": 95190 }, { "grad_norm": 0.06689844280481339, "learning_rate": 6.288448418349858e-07, "loss": 0.0037, "step": 95200 }, { "grad_norm": 0.06685864180326462, "learning_rate": 6.262334053993235e-07, "loss": 0.0095, "step": 95210 }, { "grad_norm": 0.13959476351737976, "learning_rate": 6.236273684045501e-07, "loss": 0.0041, "step": 95220 }, { "grad_norm": 0.07589404284954071, "learning_rate": 6.210267311356488e-07, "loss": 0.0066, "step": 95230 }, { "grad_norm": 0.14812076091766357, "learning_rate": 6.184314938770197e-07, "loss": 0.0041, "step": 95240 }, { "grad_norm": 0.0938694030046463, "learning_rate": 6.158416569124858e-07, "loss": 0.0038, "step": 95250 }, { "grad_norm": 0.08824033290147781, "learning_rate": 6.132572205252541e-07, "loss": 0.0042, "step": 95260 }, { "grad_norm": 0.1063726618885994, "learning_rate": 6.106781849979648e-07, "loss": 0.0049, "step": 95270 }, { "grad_norm": 0.08416394144296646, "learning_rate": 6.081045506126482e-07, "loss": 0.0041, "step": 95280 }, { "grad_norm": 0.10305853188037872, "learning_rate": 6.055363176507678e-07, "loss": 0.004, "step": 95290 }, { "grad_norm": 0.0770421251654625, "learning_rate": 6.029734863931602e-07, "loss": 0.0055, "step": 95300 }, { "grad_norm": 0.09403622895479202, "learning_rate": 6.004160571201179e-07, "loss": 0.0052, "step": 95310 }, { "grad_norm": 0.08669286966323853, "learning_rate": 5.978640301112947e-07, "loss": 0.0042, "step": 95320 }, { "grad_norm": 0.07350341975688934, "learning_rate": 5.9531740564579e-07, "loss": 0.0048, "step": 95330 }, { "grad_norm": 0.11447666585445404, "learning_rate": 5.927761840020918e-07, "loss": 0.0055, "step": 95340 }, { "grad_norm": 0.12982265651226044, "learning_rate": 5.902403654581113e-07, "loss": 0.0055, "step": 95350 }, { "grad_norm": 0.1295916736125946, "learning_rate": 5.877099502911542e-07, "loss": 0.0054, "step": 95360 }, { "grad_norm": 0.1544453501701355, "learning_rate": 5.851849387779496e-07, "loss": 0.004, "step": 95370 }, { "grad_norm": 0.10358079522848129, "learning_rate": 5.826653311946262e-07, "loss": 0.0061, "step": 95380 }, { "grad_norm": 0.1082991287112236, "learning_rate": 5.801511278167249e-07, "loss": 0.0041, "step": 95390 }, { "grad_norm": 0.10376475751399994, "learning_rate": 5.776423289191868e-07, "loss": 0.0044, "step": 95400 }, { "grad_norm": 0.16276215016841888, "learning_rate": 5.751389347763869e-07, "loss": 0.0066, "step": 95410 }, { "grad_norm": 0.07208844274282455, "learning_rate": 5.726409456620841e-07, "loss": 0.0047, "step": 95420 }, { "grad_norm": 0.14875487983226776, "learning_rate": 5.701483618494541e-07, "loss": 0.0047, "step": 95430 }, { "grad_norm": 0.11055472493171692, "learning_rate": 5.67661183611079e-07, "loss": 0.0037, "step": 95440 }, { "grad_norm": 0.11115901917219162, "learning_rate": 5.651794112189579e-07, "loss": 0.0034, "step": 95450 }, { "grad_norm": 0.07860279828310013, "learning_rate": 5.627030449444903e-07, "loss": 0.0049, "step": 95460 }, { "grad_norm": 0.13822022080421448, "learning_rate": 5.60232085058493e-07, "loss": 0.0065, "step": 95470 }, { "grad_norm": 0.10340330749750137, "learning_rate": 5.577665318311776e-07, "loss": 0.0056, "step": 95480 }, { "grad_norm": 0.05314812436699867, "learning_rate": 5.553063855321838e-07, "loss": 0.0035, "step": 95490 }, { "grad_norm": 0.0859929695725441, "learning_rate": 5.528516464305467e-07, "loss": 0.0099, "step": 95500 }, { "grad_norm": 0.08342219889163971, "learning_rate": 5.504023147947068e-07, "loss": 0.0041, "step": 95510 }, { "grad_norm": 0.08816295862197876, "learning_rate": 5.479583908925279e-07, "loss": 0.0052, "step": 95520 }, { "grad_norm": 0.12391914427280426, "learning_rate": 5.455198749912626e-07, "loss": 0.0041, "step": 95530 }, { "grad_norm": 0.07494959980249405, "learning_rate": 5.430867673575979e-07, "loss": 0.0043, "step": 95540 }, { "grad_norm": 0.09837178140878677, "learning_rate": 5.406590682575985e-07, "loss": 0.0033, "step": 95550 }, { "grad_norm": 0.08668597787618637, "learning_rate": 5.382367779567687e-07, "loss": 0.0063, "step": 95560 }, { "grad_norm": 0.1703708916902542, "learning_rate": 5.358198967200023e-07, "loss": 0.005, "step": 95570 }, { "grad_norm": 0.08824162930250168, "learning_rate": 5.334084248115989e-07, "loss": 0.0048, "step": 95580 }, { "grad_norm": 0.21788731217384338, "learning_rate": 5.310023624952754e-07, "loss": 0.007, "step": 95590 }, { "grad_norm": 0.11549638956785202, "learning_rate": 5.286017100341656e-07, "loss": 0.0042, "step": 95600 }, { "grad_norm": 0.1158066913485527, "learning_rate": 5.262064676907874e-07, "loss": 0.0039, "step": 95610 }, { "grad_norm": 0.16944296658039093, "learning_rate": 5.238166357270924e-07, "loss": 0.0065, "step": 95620 }, { "grad_norm": 0.07698851078748703, "learning_rate": 5.214322144044215e-07, "loss": 0.0041, "step": 95630 }, { "grad_norm": 0.1651913821697235, "learning_rate": 5.190532039835327e-07, "loss": 0.007, "step": 95640 }, { "grad_norm": 0.1020788699388504, "learning_rate": 5.166796047245903e-07, "loss": 0.0057, "step": 95650 }, { "grad_norm": 0.11256466805934906, "learning_rate": 5.143114168871699e-07, "loss": 0.0045, "step": 95660 }, { "grad_norm": 0.07826019823551178, "learning_rate": 5.119486407302532e-07, "loss": 0.0037, "step": 95670 }, { "grad_norm": 0.11507275700569153, "learning_rate": 5.095912765122335e-07, "loss": 0.0052, "step": 95680 }, { "grad_norm": 0.11947370320558548, "learning_rate": 5.07239324490899e-07, "loss": 0.0041, "step": 95690 }, { "grad_norm": 0.0679604634642601, "learning_rate": 5.048927849234608e-07, "loss": 0.005, "step": 95700 }, { "grad_norm": 0.17032311856746674, "learning_rate": 5.025516580665358e-07, "loss": 0.0053, "step": 95710 }, { "grad_norm": 0.0675937682390213, "learning_rate": 5.002159441761412e-07, "loss": 0.004, "step": 95720 }, { "grad_norm": 0.08220444619655609, "learning_rate": 4.978856435077062e-07, "loss": 0.0063, "step": 95730 }, { "grad_norm": 0.11779296398162842, "learning_rate": 4.955607563160769e-07, "loss": 0.0043, "step": 95740 }, { "grad_norm": 0.14403727650642395, "learning_rate": 4.932412828554944e-07, "loss": 0.0042, "step": 95750 }, { "grad_norm": 0.1615452617406845, "learning_rate": 4.909272233796114e-07, "loss": 0.0077, "step": 95760 }, { "grad_norm": 0.051551297307014465, "learning_rate": 4.88618578141492e-07, "loss": 0.005, "step": 95770 }, { "grad_norm": 0.19921036064624786, "learning_rate": 4.863153473936011e-07, "loss": 0.0058, "step": 95780 }, { "grad_norm": 0.05777869001030922, "learning_rate": 4.840175313878259e-07, "loss": 0.005, "step": 95790 }, { "grad_norm": 0.09259756654500961, "learning_rate": 4.817251303754433e-07, "loss": 0.0051, "step": 95800 }, { "grad_norm": 0.07462966442108154, "learning_rate": 4.794381446071583e-07, "loss": 0.0048, "step": 95810 }, { "grad_norm": 0.10492875427007675, "learning_rate": 4.771565743330597e-07, "loss": 0.0042, "step": 95820 }, { "grad_norm": 0.15222305059432983, "learning_rate": 4.7488041980266463e-07, "loss": 0.0051, "step": 95830 }, { "grad_norm": 0.09503186494112015, "learning_rate": 4.726096812648795e-07, "loss": 0.0063, "step": 95840 }, { "grad_norm": 0.08569882065057755, "learning_rate": 4.703443589680445e-07, "loss": 0.0047, "step": 95850 }, { "grad_norm": 0.11046037822961807, "learning_rate": 4.6808445315987826e-07, "loss": 0.007, "step": 95860 }, { "grad_norm": 0.08251380920410156, "learning_rate": 4.658299640875274e-07, "loss": 0.0055, "step": 95870 }, { "grad_norm": 0.03907545283436775, "learning_rate": 4.635808919975393e-07, "loss": 0.0042, "step": 95880 }, { "grad_norm": 0.05831735581159592, "learning_rate": 4.6133723713586706e-07, "loss": 0.0035, "step": 95890 }, { "grad_norm": 0.09769577533006668, "learning_rate": 4.590989997478756e-07, "loss": 0.0064, "step": 95900 }, { "grad_norm": 0.15546931326389313, "learning_rate": 4.5686618007833027e-07, "loss": 0.0057, "step": 95910 }, { "grad_norm": 0.125044584274292, "learning_rate": 4.54638778371419e-07, "loss": 0.0047, "step": 95920 }, { "grad_norm": 0.24921096861362457, "learning_rate": 4.524167948707192e-07, "loss": 0.0083, "step": 95930 }, { "grad_norm": 0.08156516402959824, "learning_rate": 4.5020022981921986e-07, "loss": 0.0049, "step": 95940 }, { "grad_norm": 0.07544062286615372, "learning_rate": 4.47989083459327e-07, "loss": 0.0051, "step": 95950 }, { "grad_norm": 0.06461337208747864, "learning_rate": 4.457833560328528e-07, "loss": 0.0049, "step": 95960 }, { "grad_norm": 0.09157416224479675, "learning_rate": 4.4358304778100434e-07, "loss": 0.0042, "step": 95970 }, { "grad_norm": 0.2978491187095642, "learning_rate": 4.413881589444058e-07, "loss": 0.0072, "step": 95980 }, { "grad_norm": 0.14270815253257751, "learning_rate": 4.391986897630873e-07, "loss": 0.0062, "step": 95990 }, { "grad_norm": 0.07687776535749435, "learning_rate": 4.370146404764908e-07, "loss": 0.0046, "step": 96000 }, { "grad_norm": 0.10956365615129471, "learning_rate": 4.3483601132345285e-07, "loss": 0.0053, "step": 96010 }, { "grad_norm": 0.10129958391189575, "learning_rate": 4.32662802542233e-07, "loss": 0.0033, "step": 96020 }, { "grad_norm": 0.10749797523021698, "learning_rate": 4.3049501437047444e-07, "loss": 0.0048, "step": 96030 }, { "grad_norm": 0.058433033525943756, "learning_rate": 4.283326470452653e-07, "loss": 0.0056, "step": 96040 }, { "grad_norm": 0.06158330664038658, "learning_rate": 4.2617570080306647e-07, "loss": 0.0039, "step": 96050 }, { "grad_norm": 0.042213499546051025, "learning_rate": 4.2402417587975585e-07, "loss": 0.005, "step": 96060 }, { "grad_norm": 0.09204650670289993, "learning_rate": 4.218780725106286e-07, "loss": 0.0052, "step": 96070 }, { "grad_norm": 0.07573073357343674, "learning_rate": 4.197373909303748e-07, "loss": 0.0049, "step": 96080 }, { "grad_norm": 0.146663099527359, "learning_rate": 4.17602131373096e-07, "loss": 0.0059, "step": 96090 }, { "grad_norm": 0.13467293977737427, "learning_rate": 4.154722940723055e-07, "loss": 0.0032, "step": 96100 }, { "grad_norm": 0.06375335156917572, "learning_rate": 4.1334787926091136e-07, "loss": 0.0055, "step": 96110 }, { "grad_norm": 0.06448429822921753, "learning_rate": 4.1122888717124444e-07, "loss": 0.0049, "step": 96120 }, { "grad_norm": 0.044665560126304626, "learning_rate": 4.091153180350249e-07, "loss": 0.0047, "step": 96130 }, { "grad_norm": 0.2265329360961914, "learning_rate": 4.0700717208340125e-07, "loss": 0.0055, "step": 96140 }, { "grad_norm": 0.1498759239912033, "learning_rate": 4.049044495469112e-07, "loss": 0.0044, "step": 96150 }, { "grad_norm": 0.4270683825016022, "learning_rate": 4.0280715065549866e-07, "loss": 0.0043, "step": 96160 }, { "grad_norm": 0.24139724671840668, "learning_rate": 4.0071527563853575e-07, "loss": 0.0085, "step": 96170 }, { "grad_norm": 0.09169404953718185, "learning_rate": 3.9862882472477823e-07, "loss": 0.0052, "step": 96180 }, { "grad_norm": 0.11581551283597946, "learning_rate": 3.9654779814239375e-07, "loss": 0.004, "step": 96190 }, { "grad_norm": 0.1013563722372055, "learning_rate": 3.944721961189668e-07, "loss": 0.008, "step": 96200 }, { "grad_norm": 0.08896712958812714, "learning_rate": 3.9240201888147697e-07, "loss": 0.004, "step": 96210 }, { "grad_norm": 0.16301238536834717, "learning_rate": 3.9033726665632096e-07, "loss": 0.0057, "step": 96220 }, { "grad_norm": 0.17744676768779755, "learning_rate": 3.882779396692959e-07, "loss": 0.005, "step": 96230 }, { "grad_norm": 0.22425180673599243, "learning_rate": 3.8622403814559947e-07, "loss": 0.0037, "step": 96240 }, { "grad_norm": 0.18415792286396027, "learning_rate": 3.841755623098575e-07, "loss": 0.0071, "step": 96250 }, { "grad_norm": 0.29413503408432007, "learning_rate": 3.821325123860742e-07, "loss": 0.0037, "step": 96260 }, { "grad_norm": 0.4460933804512024, "learning_rate": 3.800948885976818e-07, "loss": 0.0059, "step": 96270 }, { "grad_norm": 0.11787223070859909, "learning_rate": 3.780626911675078e-07, "loss": 0.0046, "step": 96280 }, { "grad_norm": 0.106745645403862, "learning_rate": 3.76035920317791e-07, "loss": 0.0066, "step": 96290 }, { "grad_norm": 0.15666118264198303, "learning_rate": 3.740145762701819e-07, "loss": 0.005, "step": 96300 }, { "grad_norm": 0.10318976640701294, "learning_rate": 3.7199865924572585e-07, "loss": 0.0044, "step": 96310 }, { "grad_norm": 0.06029180809855461, "learning_rate": 3.699881694648799e-07, "loss": 0.0045, "step": 96320 }, { "grad_norm": 0.17631536722183228, "learning_rate": 3.679831071475126e-07, "loss": 0.0054, "step": 96330 }, { "grad_norm": 0.17871220409870148, "learning_rate": 3.65983472512893e-07, "loss": 0.004, "step": 96340 }, { "grad_norm": 0.1376560479402542, "learning_rate": 3.6398926577969616e-07, "loss": 0.0038, "step": 96350 }, { "grad_norm": 0.06768731772899628, "learning_rate": 3.620004871660032e-07, "loss": 0.0058, "step": 96360 }, { "grad_norm": 0.07256264239549637, "learning_rate": 3.600171368893124e-07, "loss": 0.0044, "step": 96370 }, { "grad_norm": 0.1349891573190689, "learning_rate": 3.580392151665113e-07, "loss": 0.004, "step": 96380 }, { "grad_norm": 0.03925703838467598, "learning_rate": 3.5606672221391024e-07, "loss": 0.0055, "step": 96390 }, { "grad_norm": 0.07978955656290054, "learning_rate": 3.540996582472089e-07, "loss": 0.0037, "step": 96400 }, { "grad_norm": 0.09331005066633224, "learning_rate": 3.521380234815297e-07, "loss": 0.0056, "step": 96410 }, { "grad_norm": 0.07577851414680481, "learning_rate": 3.5018181813138985e-07, "loss": 0.0046, "step": 96420 }, { "grad_norm": 0.15015292167663574, "learning_rate": 3.482310424107238e-07, "loss": 0.0039, "step": 96430 }, { "grad_norm": 0.10179344564676285, "learning_rate": 3.4628569653285537e-07, "loss": 0.0047, "step": 96440 }, { "grad_norm": 0.1668735146522522, "learning_rate": 3.44345780710531e-07, "loss": 0.0061, "step": 96450 }, { "grad_norm": 0.12322396039962769, "learning_rate": 3.424112951558978e-07, "loss": 0.0059, "step": 96460 }, { "grad_norm": 0.0712057501077652, "learning_rate": 3.40482240080503e-07, "loss": 0.0049, "step": 96470 }, { "grad_norm": 0.1427677869796753, "learning_rate": 3.3855861569531133e-07, "loss": 0.0064, "step": 96480 }, { "grad_norm": 0.09849180281162262, "learning_rate": 3.3664042221068226e-07, "loss": 0.0052, "step": 96490 }, { "grad_norm": 0.9893863797187805, "learning_rate": 3.347276598363924e-07, "loss": 0.0071, "step": 96500 }, { "grad_norm": 0.09882312268018723, "learning_rate": 3.328203287816134e-07, "loss": 0.005, "step": 96510 }, { "grad_norm": 0.08712570369243622, "learning_rate": 3.3091842925493386e-07, "loss": 0.0053, "step": 96520 }, { "grad_norm": 0.11646828800439835, "learning_rate": 3.2902196146433196e-07, "loss": 0.0043, "step": 96530 }, { "grad_norm": 0.11229635775089264, "learning_rate": 3.2713092561720836e-07, "loss": 0.0057, "step": 96540 }, { "grad_norm": 0.13953061401844025, "learning_rate": 3.2524532192036996e-07, "loss": 0.0052, "step": 96550 }, { "grad_norm": 0.08954524993896484, "learning_rate": 3.2336515058001836e-07, "loss": 0.004, "step": 96560 }, { "grad_norm": 0.11327022314071655, "learning_rate": 3.2149041180176143e-07, "loss": 0.0043, "step": 96570 }, { "grad_norm": 0.16684171557426453, "learning_rate": 3.1962110579062954e-07, "loss": 0.0048, "step": 96580 }, { "grad_norm": 0.10754604637622833, "learning_rate": 3.177572327510314e-07, "loss": 0.0056, "step": 96590 }, { "grad_norm": 0.1355927735567093, "learning_rate": 3.158987928868151e-07, "loss": 0.0064, "step": 96600 }, { "grad_norm": 0.11810968816280365, "learning_rate": 3.1404578640120144e-07, "loss": 0.0052, "step": 96610 }, { "grad_norm": 0.1664714515209198, "learning_rate": 3.1219821349683933e-07, "loss": 0.0066, "step": 96620 }, { "grad_norm": 0.14098767936229706, "learning_rate": 3.103560743757783e-07, "loss": 0.0048, "step": 96630 }, { "grad_norm": 0.05214807018637657, "learning_rate": 3.0851936923946834e-07, "loss": 0.0078, "step": 96640 }, { "grad_norm": 0.07761669903993607, "learning_rate": 3.06688098288771e-07, "loss": 0.005, "step": 96650 }, { "grad_norm": 0.10960681736469269, "learning_rate": 3.0486226172394825e-07, "loss": 0.0045, "step": 96660 }, { "grad_norm": 0.2229408621788025, "learning_rate": 3.030418597446738e-07, "loss": 0.0045, "step": 96670 }, { "grad_norm": 0.08803398907184601, "learning_rate": 3.012268925500272e-07, "loss": 0.0049, "step": 96680 }, { "grad_norm": 0.10549075901508331, "learning_rate": 2.994173603384831e-07, "loss": 0.0048, "step": 96690 }, { "grad_norm": 0.09191040694713593, "learning_rate": 2.9761326330793317e-07, "loss": 0.0041, "step": 96700 }, { "grad_norm": 0.08234913647174835, "learning_rate": 2.958146016556751e-07, "loss": 0.0052, "step": 96710 }, { "grad_norm": 0.07079172879457474, "learning_rate": 2.94021375578396e-07, "loss": 0.0066, "step": 96720 }, { "grad_norm": 0.19568820297718048, "learning_rate": 2.9223358527221135e-07, "loss": 0.0073, "step": 96730 }, { "grad_norm": 0.10555003583431244, "learning_rate": 2.904512309326257e-07, "loss": 0.0047, "step": 96740 }, { "grad_norm": 0.09183283150196075, "learning_rate": 2.8867431275455545e-07, "loss": 0.0051, "step": 96750 }, { "grad_norm": 0.1625087410211563, "learning_rate": 2.86902830932323e-07, "loss": 0.0056, "step": 96760 }, { "grad_norm": 0.16931194067001343, "learning_rate": 2.851367856596565e-07, "loss": 0.0059, "step": 96770 }, { "grad_norm": 0.0634470283985138, "learning_rate": 2.83376177129685e-07, "loss": 0.0032, "step": 96780 }, { "grad_norm": 0.10441368818283081, "learning_rate": 2.8162100553494884e-07, "loss": 0.004, "step": 96790 }, { "grad_norm": 0.09025257080793381, "learning_rate": 2.798712710673834e-07, "loss": 0.005, "step": 96800 }, { "grad_norm": 0.061682309955358505, "learning_rate": 2.7812697391834676e-07, "loss": 0.005, "step": 96810 }, { "grad_norm": 0.0849299430847168, "learning_rate": 2.7638811427858627e-07, "loss": 0.0042, "step": 96820 }, { "grad_norm": 0.06600485742092133, "learning_rate": 2.746546923382665e-07, "loss": 0.0047, "step": 96830 }, { "grad_norm": 0.09015851467847824, "learning_rate": 2.729267082869469e-07, "loss": 0.0042, "step": 96840 }, { "grad_norm": 0.08373480290174484, "learning_rate": 2.712041623135986e-07, "loss": 0.0048, "step": 96850 }, { "grad_norm": 0.15878352522850037, "learning_rate": 2.694870546065931e-07, "loss": 0.0044, "step": 96860 }, { "grad_norm": 0.19072353839874268, "learning_rate": 2.677753853537246e-07, "loss": 0.0053, "step": 96870 }, { "grad_norm": 0.08347907662391663, "learning_rate": 2.660691547421601e-07, "loss": 0.0057, "step": 96880 }, { "grad_norm": 0.07453572005033493, "learning_rate": 2.6436836295850033e-07, "loss": 0.0043, "step": 96890 }, { "grad_norm": 0.1051492989063263, "learning_rate": 2.6267301018874093e-07, "loss": 0.0039, "step": 96900 }, { "grad_norm": 0.11599200963973999, "learning_rate": 2.6098309661828355e-07, "loss": 0.0051, "step": 96910 }, { "grad_norm": 0.08201298862695694, "learning_rate": 2.592986224319305e-07, "loss": 0.0052, "step": 96920 }, { "grad_norm": 0.08306443691253662, "learning_rate": 2.576195878139009e-07, "loss": 0.0045, "step": 96930 }, { "grad_norm": 0.0748552605509758, "learning_rate": 2.5594599294780364e-07, "loss": 0.0028, "step": 96940 }, { "grad_norm": 0.07773122936487198, "learning_rate": 2.542778380166644e-07, "loss": 0.004, "step": 96950 }, { "grad_norm": 0.09713815897703171, "learning_rate": 2.526151232029095e-07, "loss": 0.0046, "step": 96960 }, { "grad_norm": 0.11970160901546478, "learning_rate": 2.5095784868836567e-07, "loss": 0.0044, "step": 96970 }, { "grad_norm": 0.1532006561756134, "learning_rate": 2.493060146542825e-07, "loss": 0.0068, "step": 96980 }, { "grad_norm": 0.11927391588687897, "learning_rate": 2.476596212812876e-07, "loss": 0.0066, "step": 96990 }, { "grad_norm": 0.10648065060377121, "learning_rate": 2.4601866874943703e-07, "loss": 0.0046, "step": 97000 }, { "grad_norm": 0.052142322063446045, "learning_rate": 2.4438315723818163e-07, "loss": 0.0068, "step": 97010 }, { "grad_norm": 0.09245534986257553, "learning_rate": 2.4275308692637833e-07, "loss": 0.0048, "step": 97020 }, { "grad_norm": 0.10436751693487167, "learning_rate": 2.411284579922901e-07, "loss": 0.0053, "step": 97030 }, { "grad_norm": 0.2744296193122864, "learning_rate": 2.395092706135804e-07, "loss": 0.0095, "step": 97040 }, { "grad_norm": 0.1696530282497406, "learning_rate": 2.3789552496731872e-07, "loss": 0.0064, "step": 97050 }, { "grad_norm": 0.0976724699139595, "learning_rate": 2.3628722122999158e-07, "loss": 0.0069, "step": 97060 }, { "grad_norm": 0.11824879050254822, "learning_rate": 2.3468435957747503e-07, "loss": 0.0056, "step": 97070 }, { "grad_norm": 0.11778784543275833, "learning_rate": 2.330869401850566e-07, "loss": 0.0036, "step": 97080 }, { "grad_norm": 0.24609173834323883, "learning_rate": 2.3149496322742437e-07, "loss": 0.0045, "step": 97090 }, { "grad_norm": 0.08259329944849014, "learning_rate": 2.2990842887868346e-07, "loss": 0.0047, "step": 97100 }, { "grad_norm": 0.09937389194965363, "learning_rate": 2.2832733731232292e-07, "loss": 0.0057, "step": 97110 }, { "grad_norm": 0.16299273073673248, "learning_rate": 2.2675168870124886e-07, "loss": 0.0071, "step": 97120 }, { "grad_norm": 0.06479356437921524, "learning_rate": 2.2518148321778455e-07, "loss": 0.0035, "step": 97130 }, { "grad_norm": 0.06896024942398071, "learning_rate": 2.2361672103363706e-07, "loss": 0.0043, "step": 97140 }, { "grad_norm": 0.10914912074804306, "learning_rate": 2.220574023199251e-07, "loss": 0.0057, "step": 97150 }, { "grad_norm": 0.08530737459659576, "learning_rate": 2.2050352724717894e-07, "loss": 0.0048, "step": 97160 }, { "grad_norm": 0.06812956929206848, "learning_rate": 2.1895509598532372e-07, "loss": 0.005, "step": 97170 }, { "grad_norm": 0.06928594410419464, "learning_rate": 2.1741210870369066e-07, "loss": 0.0059, "step": 97180 }, { "grad_norm": 0.10164078325033188, "learning_rate": 2.1587456557102814e-07, "loss": 0.0059, "step": 97190 }, { "grad_norm": 0.0906270295381546, "learning_rate": 2.143424667554683e-07, "loss": 0.0049, "step": 97200 }, { "grad_norm": 0.1044503003358841, "learning_rate": 2.128158124245716e-07, "loss": 0.0072, "step": 97210 }, { "grad_norm": 0.15386496484279633, "learning_rate": 2.1129460274527668e-07, "loss": 0.0049, "step": 97220 }, { "grad_norm": 0.09844045341014862, "learning_rate": 2.0977883788395046e-07, "loss": 0.0052, "step": 97230 }, { "grad_norm": 0.16361606121063232, "learning_rate": 2.0826851800634928e-07, "loss": 0.0049, "step": 97240 }, { "grad_norm": 0.08692161738872528, "learning_rate": 2.0676364327764653e-07, "loss": 0.0038, "step": 97250 }, { "grad_norm": 0.08135252445936203, "learning_rate": 2.0526421386240502e-07, "loss": 0.0055, "step": 97260 }, { "grad_norm": 0.09754745662212372, "learning_rate": 2.0377022992461025e-07, "loss": 0.0048, "step": 97270 }, { "grad_norm": 0.24395360052585602, "learning_rate": 2.0228169162763156e-07, "loss": 0.0047, "step": 97280 }, { "grad_norm": 0.11079324036836624, "learning_rate": 2.0079859913425536e-07, "loss": 0.0068, "step": 97290 }, { "grad_norm": 0.23557746410369873, "learning_rate": 1.9932095260667417e-07, "loss": 0.0078, "step": 97300 }, { "grad_norm": 0.07636772841215134, "learning_rate": 1.9784875220648092e-07, "loss": 0.0035, "step": 97310 }, { "grad_norm": 0.08464863896369934, "learning_rate": 1.9638199809466907e-07, "loss": 0.0055, "step": 97320 }, { "grad_norm": 0.09357598423957825, "learning_rate": 1.9492069043164364e-07, "loss": 0.0054, "step": 97330 }, { "grad_norm": 0.08447600901126862, "learning_rate": 1.9346482937721567e-07, "loss": 0.0057, "step": 97340 }, { "grad_norm": 0.07401682436466217, "learning_rate": 1.920144150905856e-07, "loss": 0.0037, "step": 97350 }, { "grad_norm": 0.08428420126438141, "learning_rate": 1.9056944773037655e-07, "loss": 0.0056, "step": 97360 }, { "grad_norm": 0.14090117812156677, "learning_rate": 1.8912992745460102e-07, "loss": 0.0053, "step": 97370 }, { "grad_norm": 0.09687214344739914, "learning_rate": 1.876958544206886e-07, "loss": 0.0047, "step": 97380 }, { "grad_norm": 0.06228528544306755, "learning_rate": 1.8626722878546942e-07, "loss": 0.007, "step": 97390 }, { "grad_norm": 0.0967174842953682, "learning_rate": 1.8484405070516851e-07, "loss": 0.006, "step": 97400 }, { "grad_norm": 0.12324681878089905, "learning_rate": 1.8342632033542805e-07, "loss": 0.0057, "step": 97410 }, { "grad_norm": 0.131325826048851, "learning_rate": 1.8201403783129066e-07, "loss": 0.0044, "step": 97420 }, { "grad_norm": 0.07625596225261688, "learning_rate": 1.806072033471884e-07, "loss": 0.0038, "step": 97430 }, { "grad_norm": 0.0937357172369957, "learning_rate": 1.79205817036987e-07, "loss": 0.0056, "step": 97440 }, { "grad_norm": 0.08864285051822662, "learning_rate": 1.7780987905393065e-07, "loss": 0.0053, "step": 97450 }, { "grad_norm": 0.23825328052043915, "learning_rate": 1.764193895506805e-07, "loss": 0.0058, "step": 97460 }, { "grad_norm": 0.0763416513800621, "learning_rate": 1.7503434867929824e-07, "loss": 0.0043, "step": 97470 }, { "grad_norm": 0.07556019723415375, "learning_rate": 1.7365475659125165e-07, "loss": 0.004, "step": 97480 }, { "grad_norm": 0.07878085225820541, "learning_rate": 1.7228061343740332e-07, "loss": 0.0051, "step": 97490 }, { "grad_norm": 0.1444895714521408, "learning_rate": 1.7091191936803308e-07, "loss": 0.0042, "step": 97500 }, { "grad_norm": 0.126768097281456, "learning_rate": 1.6954867453281563e-07, "loss": 0.0041, "step": 97510 }, { "grad_norm": 0.11425741761922836, "learning_rate": 1.6819087908084285e-07, "loss": 0.0051, "step": 97520 }, { "grad_norm": 0.07231871783733368, "learning_rate": 1.6683853316058485e-07, "loss": 0.0066, "step": 97530 }, { "grad_norm": 0.07368120551109314, "learning_rate": 1.6549163691995106e-07, "loss": 0.0039, "step": 97540 }, { "grad_norm": 0.08984989672899246, "learning_rate": 1.641501905062237e-07, "loss": 0.0048, "step": 97550 }, { "grad_norm": 0.04586271941661835, "learning_rate": 1.628141940661021e-07, "loss": 0.0057, "step": 97560 }, { "grad_norm": 0.1826307624578476, "learning_rate": 1.6148364774569158e-07, "loss": 0.0057, "step": 97570 }, { "grad_norm": 0.16273920238018036, "learning_rate": 1.6015855169050355e-07, "loss": 0.0049, "step": 97580 }, { "grad_norm": 0.3242436349391937, "learning_rate": 1.5883890604543873e-07, "loss": 0.0035, "step": 97590 }, { "grad_norm": 0.09346702694892883, "learning_rate": 1.5752471095482057e-07, "loss": 0.0047, "step": 97600 }, { "grad_norm": 0.11305903643369675, "learning_rate": 1.5621596656235638e-07, "loss": 0.0043, "step": 97610 }, { "grad_norm": 0.11693814396858215, "learning_rate": 1.5491267301117608e-07, "loss": 0.0034, "step": 97620 }, { "grad_norm": 0.06689152866601944, "learning_rate": 1.536148304438101e-07, "loss": 0.0036, "step": 97630 }, { "grad_norm": 0.07833388447761536, "learning_rate": 1.5232243900217823e-07, "loss": 0.0057, "step": 97640 }, { "grad_norm": 0.20050081610679626, "learning_rate": 1.51035498827623e-07, "loss": 0.0064, "step": 97650 }, { "grad_norm": 0.13711005449295044, "learning_rate": 1.497540100608763e-07, "loss": 0.0041, "step": 97660 }, { "grad_norm": 0.06233340874314308, "learning_rate": 1.4847797284208153e-07, "loss": 0.0085, "step": 97670 }, { "grad_norm": 0.1396171748638153, "learning_rate": 1.472073873107882e-07, "loss": 0.0051, "step": 97680 }, { "grad_norm": 0.2111409604549408, "learning_rate": 1.459422536059407e-07, "loss": 0.0085, "step": 97690 }, { "grad_norm": 0.09965581446886063, "learning_rate": 1.4468257186589508e-07, "loss": 0.0039, "step": 97700 }, { "grad_norm": 0.09957968443632126, "learning_rate": 1.4342834222840217e-07, "loss": 0.0062, "step": 97710 }, { "grad_norm": 0.11848373711109161, "learning_rate": 1.4217956483063566e-07, "loss": 0.0037, "step": 97720 }, { "grad_norm": 0.06488437950611115, "learning_rate": 1.409362398091474e-07, "loss": 0.0038, "step": 97730 }, { "grad_norm": 0.12511618435382843, "learning_rate": 1.3969836729990638e-07, "loss": 0.0078, "step": 97740 }, { "grad_norm": 0.14561469852924347, "learning_rate": 1.384659474382932e-07, "loss": 0.005, "step": 97750 }, { "grad_norm": 0.13312023878097534, "learning_rate": 1.372389803590779e-07, "loss": 0.0039, "step": 97760 }, { "grad_norm": 0.07487239688634872, "learning_rate": 1.360174661964364e-07, "loss": 0.005, "step": 97770 }, { "grad_norm": 0.2881588041782379, "learning_rate": 1.3480140508396188e-07, "loss": 0.0044, "step": 97780 }, { "grad_norm": 0.23503541946411133, "learning_rate": 1.335907971546313e-07, "loss": 0.006, "step": 97790 }, { "grad_norm": 0.08991523832082748, "learning_rate": 1.3238564254083875e-07, "loss": 0.0052, "step": 97800 }, { "grad_norm": 0.16429449617862701, "learning_rate": 1.3118594137437323e-07, "loss": 0.0038, "step": 97810 }, { "grad_norm": 0.08396050333976746, "learning_rate": 1.2999169378644094e-07, "loss": 0.0037, "step": 97820 }, { "grad_norm": 0.0838082805275917, "learning_rate": 1.2880289990763738e-07, "loss": 0.0047, "step": 97830 }, { "grad_norm": 0.1012682095170021, "learning_rate": 1.2761955986796968e-07, "loss": 0.0038, "step": 97840 }, { "grad_norm": 0.16841581463813782, "learning_rate": 1.264416737968399e-07, "loss": 0.0049, "step": 97850 }, { "grad_norm": 0.12643350660800934, "learning_rate": 1.2526924182307275e-07, "loss": 0.0041, "step": 97860 }, { "grad_norm": 0.14657030999660492, "learning_rate": 1.2410226407487124e-07, "loss": 0.0064, "step": 97870 }, { "grad_norm": 0.2168341428041458, "learning_rate": 1.229407406798555e-07, "loss": 0.0075, "step": 97880 }, { "grad_norm": 0.13939028978347778, "learning_rate": 1.2178467176505726e-07, "loss": 0.0069, "step": 97890 }, { "grad_norm": 0.12406173348426819, "learning_rate": 1.2063405745689205e-07, "loss": 0.0045, "step": 97900 }, { "grad_norm": 0.16850076615810394, "learning_rate": 1.1948889788119812e-07, "loss": 0.0038, "step": 97910 }, { "grad_norm": 0.05938919633626938, "learning_rate": 1.1834919316320303e-07, "loss": 0.0038, "step": 97920 }, { "grad_norm": 0.06697746366262436, "learning_rate": 1.1721494342754048e-07, "loss": 0.0052, "step": 97930 }, { "grad_norm": 0.060667555779218674, "learning_rate": 1.1608614879825563e-07, "loss": 0.0039, "step": 97940 }, { "grad_norm": 0.1914956271648407, "learning_rate": 1.1496280939879422e-07, "loss": 0.005, "step": 97950 }, { "grad_norm": 0.09144917130470276, "learning_rate": 1.1384492535199687e-07, "loss": 0.0054, "step": 97960 }, { "grad_norm": 0.13886114954948425, "learning_rate": 1.1273249678011578e-07, "loss": 0.0041, "step": 97970 }, { "grad_norm": 0.072356678545475, "learning_rate": 1.1162552380480362e-07, "loss": 0.0033, "step": 97980 }, { "grad_norm": 0.10996883362531662, "learning_rate": 1.1052400654711359e-07, "loss": 0.0045, "step": 97990 }, { "grad_norm": 0.17386668920516968, "learning_rate": 1.0942794512751597e-07, "loss": 0.0072, "step": 98000 }, { "grad_norm": 0.10519369691610336, "learning_rate": 1.0833733966587045e-07, "loss": 0.0051, "step": 98010 }, { "grad_norm": 0.08063217997550964, "learning_rate": 1.0725219028143718e-07, "loss": 0.0046, "step": 98020 }, { "grad_norm": 0.07774881273508072, "learning_rate": 1.0617249709289345e-07, "loss": 0.004, "step": 98030 }, { "grad_norm": 0.13132885098457336, "learning_rate": 1.0509826021831703e-07, "loss": 0.0045, "step": 98040 }, { "grad_norm": 0.10035628080368042, "learning_rate": 1.0402947977517508e-07, "loss": 0.0059, "step": 98050 }, { "grad_norm": 0.08402828872203827, "learning_rate": 1.0296615588035185e-07, "loss": 0.0064, "step": 98060 }, { "grad_norm": 0.12195593118667603, "learning_rate": 1.0190828865012659e-07, "loss": 0.0066, "step": 98070 }, { "grad_norm": 0.13148784637451172, "learning_rate": 1.008558782001956e-07, "loss": 0.0049, "step": 98080 }, { "grad_norm": 0.11098430305719376, "learning_rate": 9.980892464563906e-08, "loss": 0.0038, "step": 98090 }, { "grad_norm": 0.12397134304046631, "learning_rate": 9.876742810095985e-08, "loss": 0.0071, "step": 98100 }, { "grad_norm": 0.0634768009185791, "learning_rate": 9.77313886800446e-08, "loss": 0.0047, "step": 98110 }, { "grad_norm": 0.08303961157798767, "learning_rate": 9.670080649619717e-08, "loss": 0.0029, "step": 98120 }, { "grad_norm": 0.11361133307218552, "learning_rate": 9.567568166212737e-08, "loss": 0.0039, "step": 98130 }, { "grad_norm": 0.0847926139831543, "learning_rate": 9.465601428992888e-08, "loss": 0.0042, "step": 98140 }, { "grad_norm": 0.12650984525680542, "learning_rate": 9.364180449111803e-08, "loss": 0.0057, "step": 98150 }, { "grad_norm": 0.11565900593996048, "learning_rate": 9.263305237661169e-08, "loss": 0.0121, "step": 98160 }, { "grad_norm": 0.06755612045526505, "learning_rate": 9.162975805671603e-08, "loss": 0.0044, "step": 98170 }, { "grad_norm": 0.07130073010921478, "learning_rate": 9.063192164115442e-08, "loss": 0.0042, "step": 98180 }, { "grad_norm": 0.08997868001461029, "learning_rate": 8.963954323904511e-08, "loss": 0.0041, "step": 98190 }, { "grad_norm": 0.17138390243053436, "learning_rate": 8.865262295891796e-08, "loss": 0.0043, "step": 98200 }, { "grad_norm": 0.10297229886054993, "learning_rate": 8.767116090870331e-08, "loss": 0.0044, "step": 98210 }, { "grad_norm": 0.2815564274787903, "learning_rate": 8.669515719572086e-08, "loss": 0.0092, "step": 98220 }, { "grad_norm": 0.07480955868959427, "learning_rate": 8.572461192671855e-08, "loss": 0.0034, "step": 98230 }, { "grad_norm": 0.07966705411672592, "learning_rate": 8.475952520782815e-08, "loss": 0.0036, "step": 98240 }, { "grad_norm": 0.07364852726459503, "learning_rate": 8.379989714458747e-08, "loss": 0.0032, "step": 98250 }, { "grad_norm": 0.06052424758672714, "learning_rate": 8.284572784194034e-08, "loss": 0.0048, "step": 98260 }, { "grad_norm": 0.11640218645334244, "learning_rate": 8.189701740424216e-08, "loss": 0.0058, "step": 98270 }, { "grad_norm": 0.07090295851230621, "learning_rate": 8.095376593522663e-08, "loss": 0.0056, "step": 98280 }, { "grad_norm": 0.11227474361658096, "learning_rate": 8.001597353806123e-08, "loss": 0.0046, "step": 98290 }, { "grad_norm": 0.08336787670850754, "learning_rate": 7.90836403152917e-08, "loss": 0.0057, "step": 98300 }, { "grad_norm": 0.1481373906135559, "learning_rate": 7.815676636888092e-08, "loss": 0.0056, "step": 98310 }, { "grad_norm": 0.150326207280159, "learning_rate": 7.723535180019226e-08, "loss": 0.006, "step": 98320 }, { "grad_norm": 0.06461042165756226, "learning_rate": 7.6319396709984e-08, "loss": 0.005, "step": 98330 }, { "grad_norm": 0.08675245195627213, "learning_rate": 7.540890119843158e-08, "loss": 0.0054, "step": 98340 }, { "grad_norm": 0.3609199821949005, "learning_rate": 7.450386536509979e-08, "loss": 0.005, "step": 98350 }, { "grad_norm": 0.08388006687164307, "learning_rate": 7.360428930895947e-08, "loss": 0.0041, "step": 98360 }, { "grad_norm": 0.05732740834355354, "learning_rate": 7.271017312839302e-08, "loss": 0.0058, "step": 98370 }, { "grad_norm": 0.1421276181936264, "learning_rate": 7.182151692118333e-08, "loss": 0.0052, "step": 98380 }, { "grad_norm": 0.08561436086893082, "learning_rate": 7.093832078449713e-08, "loss": 0.004, "step": 98390 }, { "grad_norm": 0.15267592668533325, "learning_rate": 7.006058481493493e-08, "loss": 0.0091, "step": 98400 }, { "grad_norm": 0.07415146380662918, "learning_rate": 6.918830910847552e-08, "loss": 0.0051, "step": 98410 }, { "grad_norm": 0.12204684317111969, "learning_rate": 6.832149376052033e-08, "loss": 0.0046, "step": 98420 }, { "grad_norm": 0.09603048115968704, "learning_rate": 6.746013886584912e-08, "loss": 0.004, "step": 98430 }, { "grad_norm": 0.16478092968463898, "learning_rate": 6.660424451866431e-08, "loss": 0.0038, "step": 98440 }, { "grad_norm": 0.12946081161499023, "learning_rate": 6.575381081256327e-08, "loss": 0.0053, "step": 98450 }, { "grad_norm": 0.22705675661563873, "learning_rate": 6.49088378405549e-08, "loss": 0.0043, "step": 98460 }, { "grad_norm": 0.12201090902090073, "learning_rate": 6.406932569503755e-08, "loss": 0.0051, "step": 98470 }, { "grad_norm": 0.08197092264890671, "learning_rate": 6.323527446782662e-08, "loss": 0.0053, "step": 98480 }, { "grad_norm": 0.10132572054862976, "learning_rate": 6.240668425012142e-08, "loss": 0.004, "step": 98490 }, { "grad_norm": 0.16192524135112762, "learning_rate": 6.158355513254388e-08, "loss": 0.0065, "step": 98500 }, { "grad_norm": 0.11435447633266449, "learning_rate": 6.076588720510534e-08, "loss": 0.0041, "step": 98510 }, { "grad_norm": 0.14979492127895355, "learning_rate": 5.995368055722316e-08, "loss": 0.0053, "step": 98520 }, { "grad_norm": 0.1954156905412674, "learning_rate": 5.914693527773185e-08, "loss": 0.0044, "step": 98530 }, { "grad_norm": 0.25169533491134644, "learning_rate": 5.834565145483861e-08, "loss": 0.0072, "step": 98540 }, { "grad_norm": 0.11333798617124557, "learning_rate": 5.754982917618446e-08, "loss": 0.004, "step": 98550 }, { "grad_norm": 0.09594418108463287, "learning_rate": 5.675946852878866e-08, "loss": 0.0053, "step": 98560 }, { "grad_norm": 0.1848229020833969, "learning_rate": 5.59745695990932e-08, "loss": 0.0053, "step": 98570 }, { "grad_norm": 0.04431642219424248, "learning_rate": 5.51951324729294e-08, "loss": 0.0038, "step": 98580 }, { "grad_norm": 0.1399596482515335, "learning_rate": 5.4421157235529093e-08, "loss": 0.0051, "step": 98590 }, { "grad_norm": 0.09868916124105453, "learning_rate": 5.365264397154679e-08, "loss": 0.0046, "step": 98600 }, { "grad_norm": 0.12303376197814941, "learning_rate": 5.288959276501526e-08, "loss": 0.0041, "step": 98610 }, { "grad_norm": 0.07312256097793579, "learning_rate": 5.2132003699378876e-08, "loss": 0.0043, "step": 98620 }, { "grad_norm": 0.1556726098060608, "learning_rate": 5.137987685749912e-08, "loss": 0.0044, "step": 98630 }, { "grad_norm": 0.05523701384663582, "learning_rate": 5.0633212321610227e-08, "loss": 0.0033, "step": 98640 }, { "grad_norm": 0.045811254531145096, "learning_rate": 4.989201017338019e-08, "loss": 0.0036, "step": 98650 }, { "grad_norm": 0.123001329600811, "learning_rate": 4.9156270493860846e-08, "loss": 0.0048, "step": 98660 }, { "grad_norm": 0.10132615268230438, "learning_rate": 4.842599336351561e-08, "loss": 0.0067, "step": 98670 }, { "grad_norm": 0.09142711013555527, "learning_rate": 4.7701178862197274e-08, "loss": 0.0046, "step": 98680 }, { "grad_norm": 0.16836665570735931, "learning_rate": 4.698182706918131e-08, "loss": 0.0058, "step": 98690 }, { "grad_norm": 0.18019209802150726, "learning_rate": 4.6267938063121466e-08, "loss": 0.0037, "step": 98700 }, { "grad_norm": 0.17771685123443604, "learning_rate": 4.555951192209973e-08, "loss": 0.005, "step": 98710 }, { "grad_norm": 0.10760564357042313, "learning_rate": 4.485654872358747e-08, "loss": 0.0052, "step": 98720 }, { "grad_norm": 0.05504203960299492, "learning_rate": 4.415904854445097e-08, "loss": 0.0045, "step": 98730 }, { "grad_norm": 0.09621472656726837, "learning_rate": 4.34670114609792e-08, "loss": 0.0045, "step": 98740 }, { "grad_norm": 0.061262890696525574, "learning_rate": 4.278043754884498e-08, "loss": 0.0044, "step": 98750 }, { "grad_norm": 0.12994307279586792, "learning_rate": 4.209932688312712e-08, "loss": 0.0071, "step": 98760 }, { "grad_norm": 0.1741643100976944, "learning_rate": 4.14236795383216e-08, "loss": 0.0077, "step": 98770 }, { "grad_norm": 0.13581353425979614, "learning_rate": 4.075349558830821e-08, "loss": 0.0045, "step": 98780 }, { "grad_norm": 0.0861257016658783, "learning_rate": 4.008877510638387e-08, "loss": 0.0034, "step": 98790 }, { "grad_norm": 0.10285436362028122, "learning_rate": 3.942951816523488e-08, "loss": 0.0045, "step": 98800 }, { "grad_norm": 0.053450897336006165, "learning_rate": 3.8775724836959125e-08, "loss": 0.006, "step": 98810 }, { "grad_norm": 0.08014117181301117, "learning_rate": 3.812739519305497e-08, "loss": 0.006, "step": 98820 }, { "grad_norm": 0.09271253645420074, "learning_rate": 3.748452930442126e-08, "loss": 0.0069, "step": 98830 }, { "grad_norm": 0.13631987571716309, "learning_rate": 3.6847127241362855e-08, "loss": 0.006, "step": 98840 }, { "grad_norm": 0.10257894545793533, "learning_rate": 3.621518907358512e-08, "loss": 0.0049, "step": 98850 }, { "grad_norm": 0.160654217004776, "learning_rate": 3.558871487019388e-08, "loss": 0.0054, "step": 98860 }, { "grad_norm": 0.08519493788480759, "learning_rate": 3.4967704699701007e-08, "loss": 0.0042, "step": 98870 }, { "grad_norm": 0.13523997366428375, "learning_rate": 3.435215863001884e-08, "loss": 0.0041, "step": 98880 }, { "grad_norm": 0.07184949517250061, "learning_rate": 3.374207672846019e-08, "loss": 0.0038, "step": 98890 }, { "grad_norm": 0.06642758101224899, "learning_rate": 3.313745906174392e-08, "loss": 0.0043, "step": 98900 }, { "grad_norm": 0.11620714515447617, "learning_rate": 3.253830569599492e-08, "loss": 0.0062, "step": 98910 }, { "grad_norm": 0.06397780030965805, "learning_rate": 3.1944616696727436e-08, "loss": 0.0049, "step": 98920 }, { "grad_norm": 0.14018455147743225, "learning_rate": 3.135639212887287e-08, "loss": 0.0038, "step": 98930 }, { "grad_norm": 0.06587935239076614, "learning_rate": 3.077363205675754e-08, "loss": 0.0042, "step": 98940 }, { "grad_norm": 0.14729461073875427, "learning_rate": 3.0196336544113804e-08, "loss": 0.005, "step": 98950 }, { "grad_norm": 0.07831515371799469, "learning_rate": 2.9624505654063383e-08, "loss": 0.0041, "step": 98960 }, { "grad_norm": 0.07057365775108337, "learning_rate": 2.90581394491507e-08, "loss": 0.0046, "step": 98970 }, { "grad_norm": 0.07685457170009613, "learning_rate": 2.8497237991309545e-08, "loss": 0.0063, "step": 98980 }, { "grad_norm": 0.090079665184021, "learning_rate": 2.7941801341879735e-08, "loss": 0.0049, "step": 98990 }, { "grad_norm": 0.08219407498836517, "learning_rate": 2.7391829561601578e-08, "loss": 0.0074, "step": 99000 }, { "grad_norm": 0.22211295366287231, "learning_rate": 2.6847322710621403e-08, "loss": 0.0055, "step": 99010 }, { "grad_norm": 0.14274993538856506, "learning_rate": 2.6308280848486022e-08, "loss": 0.0052, "step": 99020 }, { "grad_norm": 0.14009235799312592, "learning_rate": 2.5774704034137177e-08, "loss": 0.0041, "step": 99030 }, { "grad_norm": 0.10589028149843216, "learning_rate": 2.524659232593929e-08, "loss": 0.004, "step": 99040 }, { "grad_norm": 0.13714514672756195, "learning_rate": 2.472394578163506e-08, "loss": 0.005, "step": 99050 }, { "grad_norm": 0.1309465914964676, "learning_rate": 2.4206764458378772e-08, "loss": 0.0048, "step": 99060 }, { "grad_norm": 0.06922776252031326, "learning_rate": 2.3695048412736285e-08, "loss": 0.0045, "step": 99070 }, { "grad_norm": 0.15037770569324493, "learning_rate": 2.3188797700668395e-08, "loss": 0.0057, "step": 99080 }, { "grad_norm": 0.10312596708536148, "learning_rate": 2.268801237753082e-08, "loss": 0.0048, "step": 99090 }, { "grad_norm": 0.15877573192119598, "learning_rate": 2.2192692498090862e-08, "loss": 0.0043, "step": 99100 }, { "grad_norm": 0.16308796405792236, "learning_rate": 2.170283811652185e-08, "loss": 0.0041, "step": 99110 }, { "grad_norm": 0.05212241783738136, "learning_rate": 2.12184492863865e-08, "loss": 0.0036, "step": 99120 }, { "grad_norm": 0.11720877140760422, "learning_rate": 2.073952606066465e-08, "loss": 0.0055, "step": 99130 }, { "grad_norm": 0.11233994364738464, "learning_rate": 2.026606849171997e-08, "loss": 0.0054, "step": 99140 }, { "grad_norm": 0.0998927429318428, "learning_rate": 1.9798076631333262e-08, "loss": 0.005, "step": 99150 }, { "grad_norm": 0.11204418540000916, "learning_rate": 1.933555053069136e-08, "loss": 0.0052, "step": 99160 }, { "grad_norm": 0.091915562748909, "learning_rate": 1.887849024036492e-08, "loss": 0.0067, "step": 99170 }, { "grad_norm": 0.12467651069164276, "learning_rate": 1.8426895810341736e-08, "loss": 0.0099, "step": 99180 }, { "grad_norm": 0.07632312178611755, "learning_rate": 1.798076729000453e-08, "loss": 0.0044, "step": 99190 }, { "grad_norm": 0.10486740618944168, "learning_rate": 1.75401047281476e-08, "loss": 0.0065, "step": 99200 }, { "grad_norm": 0.10061022639274597, "learning_rate": 1.7104908172954626e-08, "loss": 0.0063, "step": 99210 }, { "grad_norm": 0.12816482782363892, "learning_rate": 1.6675177672020868e-08, "loss": 0.0034, "step": 99220 }, { "grad_norm": 0.062482547014951706, "learning_rate": 1.6250913272342073e-08, "loss": 0.0037, "step": 99230 }, { "grad_norm": 0.1240556463599205, "learning_rate": 1.5832115020314453e-08, "loss": 0.0052, "step": 99240 }, { "grad_norm": 0.19637872278690338, "learning_rate": 1.5418782961734712e-08, "loss": 0.0049, "step": 99250 }, { "grad_norm": 0.1581990271806717, "learning_rate": 1.501091714181113e-08, "loss": 0.005, "step": 99260 }, { "grad_norm": 0.3041532337665558, "learning_rate": 1.460851760513582e-08, "loss": 0.0039, "step": 99270 }, { "grad_norm": 0.07934746891260147, "learning_rate": 1.4211584395723566e-08, "loss": 0.0057, "step": 99280 }, { "grad_norm": 0.204880952835083, "learning_rate": 1.382011755697854e-08, "loss": 0.0066, "step": 99290 }, { "grad_norm": 0.13053064048290253, "learning_rate": 1.3434117131716494e-08, "loss": 0.0055, "step": 99300 }, { "grad_norm": 0.06403068453073502, "learning_rate": 1.3053583162142557e-08, "loss": 0.004, "step": 99310 }, { "grad_norm": 0.08312426507472992, "learning_rate": 1.2678515689873438e-08, "loss": 0.0042, "step": 99320 }, { "grad_norm": 0.09508290141820908, "learning_rate": 1.2308914755931877e-08, "loss": 0.0042, "step": 99330 }, { "grad_norm": 0.10518596321344376, "learning_rate": 1.1944780400729993e-08, "loss": 0.0078, "step": 99340 }, { "grad_norm": 0.11128129065036774, "learning_rate": 1.1586112664085935e-08, "loss": 0.0056, "step": 99350 }, { "grad_norm": 0.056792065501213074, "learning_rate": 1.1232911585234984e-08, "loss": 0.0054, "step": 99360 }, { "grad_norm": 0.11510232836008072, "learning_rate": 1.0885177202790697e-08, "loss": 0.004, "step": 99370 }, { "grad_norm": 0.1095520630478859, "learning_rate": 1.0542909554789315e-08, "loss": 0.0065, "step": 99380 }, { "grad_norm": 0.10393520444631577, "learning_rate": 1.0206108678656456e-08, "loss": 0.0047, "step": 99390 }, { "grad_norm": 0.1862335205078125, "learning_rate": 9.874774611223769e-09, "loss": 0.0054, "step": 99400 }, { "grad_norm": 0.04472799971699715, "learning_rate": 9.548907388728933e-09, "loss": 0.0051, "step": 99410 }, { "grad_norm": 0.1791868507862091, "learning_rate": 9.228507046804557e-09, "loss": 0.0057, "step": 99420 }, { "grad_norm": 0.16382096707820892, "learning_rate": 8.91357362048928e-09, "loss": 0.0051, "step": 99430 }, { "grad_norm": 0.10569902509450912, "learning_rate": 8.604107144227769e-09, "loss": 0.0067, "step": 99440 }, { "grad_norm": 0.1280544400215149, "learning_rate": 8.300107651859623e-09, "loss": 0.0052, "step": 99450 }, { "grad_norm": 0.11859682947397232, "learning_rate": 8.001575176630472e-09, "loss": 0.0061, "step": 99460 }, { "grad_norm": 0.13035616278648376, "learning_rate": 7.708509751186422e-09, "loss": 0.0081, "step": 99470 }, { "grad_norm": 0.0631442591547966, "learning_rate": 7.420911407579611e-09, "loss": 0.0045, "step": 99480 }, { "grad_norm": 0.043004393577575684, "learning_rate": 7.13878017725711e-09, "loss": 0.0047, "step": 99490 }, { "grad_norm": 0.06424744427204132, "learning_rate": 6.8621160910720125e-09, "loss": 0.0043, "step": 99500 }, { "grad_norm": 0.08281582593917847, "learning_rate": 6.59091917928345e-09, "loss": 0.0066, "step": 99510 }, { "grad_norm": 0.08984939754009247, "learning_rate": 6.32518947155103e-09, "loss": 0.0044, "step": 99520 }, { "grad_norm": 0.17312827706336975, "learning_rate": 6.064926996929288e-09, "loss": 0.0036, "step": 99530 }, { "grad_norm": 0.08533716946840286, "learning_rate": 5.810131783884343e-09, "loss": 0.0054, "step": 99540 }, { "grad_norm": 0.056642722338438034, "learning_rate": 5.5608038602772415e-09, "loss": 0.0041, "step": 99550 }, { "grad_norm": 0.09449274092912674, "learning_rate": 5.316943253375062e-09, "loss": 0.0042, "step": 99560 }, { "grad_norm": 0.10641611367464066, "learning_rate": 5.07854998984536e-09, "loss": 0.0047, "step": 99570 }, { "grad_norm": 0.2670409083366394, "learning_rate": 4.845624095756174e-09, "loss": 0.005, "step": 99580 }, { "grad_norm": 0.2766413688659668, "learning_rate": 4.618165596587121e-09, "loss": 0.0049, "step": 99590 }, { "grad_norm": 0.11702876538038254, "learning_rate": 4.396174517207197e-09, "loss": 0.0049, "step": 99600 }, { "grad_norm": 0.13561084866523743, "learning_rate": 4.179650881896979e-09, "loss": 0.0048, "step": 99610 }, { "grad_norm": 0.07464014738798141, "learning_rate": 3.968594714331975e-09, "loss": 0.0037, "step": 99620 }, { "grad_norm": 0.08428742736577988, "learning_rate": 3.763006037593719e-09, "loss": 0.0047, "step": 99630 }, { "grad_norm": 0.13726474344730377, "learning_rate": 3.5628848741586786e-09, "loss": 0.0043, "step": 99640 }, { "grad_norm": 0.24468167126178741, "learning_rate": 3.368231245926001e-09, "loss": 0.0051, "step": 99650 }, { "grad_norm": 0.08566653728485107, "learning_rate": 3.179045174167561e-09, "loss": 0.0045, "step": 99660 }, { "grad_norm": 0.1463620364665985, "learning_rate": 2.9953266795834657e-09, "loss": 0.005, "step": 99670 }, { "grad_norm": 0.07956513017416, "learning_rate": 2.817075782263201e-09, "loss": 0.0041, "step": 99680 }, { "grad_norm": 0.10278591513633728, "learning_rate": 2.6442925016967325e-09, "loss": 0.0038, "step": 99690 }, { "grad_norm": 0.050185345113277435, "learning_rate": 2.4769768567800555e-09, "loss": 0.0072, "step": 99700 }, { "grad_norm": 0.10319896787405014, "learning_rate": 2.315128865809646e-09, "loss": 0.0042, "step": 99710 }, { "grad_norm": 0.15073050558567047, "learning_rate": 2.1587485464880097e-09, "loss": 0.0046, "step": 99720 }, { "grad_norm": 0.11160267144441605, "learning_rate": 2.0078359159125813e-09, "loss": 0.005, "step": 99730 }, { "grad_norm": 0.1427803635597229, "learning_rate": 1.8623909905923776e-09, "loss": 0.0059, "step": 99740 }, { "grad_norm": 0.06351056694984436, "learning_rate": 1.7224137864257916e-09, "loss": 0.0039, "step": 99750 }, { "grad_norm": 0.060781922191381454, "learning_rate": 1.5879043187283505e-09, "loss": 0.0057, "step": 99760 }, { "grad_norm": 0.12918691337108612, "learning_rate": 1.458862602204958e-09, "loss": 0.0041, "step": 99770 }, { "grad_norm": 0.07339490205049515, "learning_rate": 1.3352886509720996e-09, "loss": 0.0071, "step": 99780 }, { "grad_norm": 0.0442308746278286, "learning_rate": 1.2171824785356389e-09, "loss": 0.0056, "step": 99790 }, { "grad_norm": 0.08976515382528305, "learning_rate": 1.1045440978185718e-09, "loss": 0.0042, "step": 99800 }, { "grad_norm": 0.18764810264110565, "learning_rate": 9.973735211332714e-10, "loss": 0.0054, "step": 99810 }, { "grad_norm": 0.0631922036409378, "learning_rate": 8.956707602036929e-10, "loss": 0.0043, "step": 99820 }, { "grad_norm": 0.16033051908016205, "learning_rate": 7.994358261542712e-10, "loss": 0.0047, "step": 99830 }, { "grad_norm": 0.07609780132770538, "learning_rate": 7.086687295043692e-10, "loss": 0.0055, "step": 99840 }, { "grad_norm": 0.13978511095046997, "learning_rate": 6.23369480179381e-10, "loss": 0.0079, "step": 99850 }, { "grad_norm": 0.21057367324829102, "learning_rate": 5.43538087510731e-10, "loss": 0.0061, "step": 99860 }, { "grad_norm": 0.09435256570577621, "learning_rate": 4.691745602303233e-10, "loss": 0.0052, "step": 99870 }, { "grad_norm": 0.09126381576061249, "learning_rate": 4.0027890646499034e-10, "loss": 0.0049, "step": 99880 }, { "grad_norm": 0.10653437674045563, "learning_rate": 3.368511337531466e-10, "loss": 0.0052, "step": 99890 }, { "grad_norm": 0.12321631610393524, "learning_rate": 2.788912490281348e-10, "loss": 0.0046, "step": 99900 }, { "grad_norm": 0.0691199079155922, "learning_rate": 2.2639925863487954e-10, "loss": 0.0049, "step": 99910 }, { "grad_norm": 0.07707135379314423, "learning_rate": 1.7937516830213163e-10, "loss": 0.0041, "step": 99920 }, { "grad_norm": 0.0881899818778038, "learning_rate": 1.3781898318687702e-10, "loss": 0.005, "step": 99930 }, { "grad_norm": 0.08916225284337997, "learning_rate": 1.0173070781882566e-10, "loss": 0.0034, "step": 99940 }, { "grad_norm": 0.1272718459367752, "learning_rate": 7.111034615592261e-11, "loss": 0.0053, "step": 99950 }, { "grad_norm": 0.1527332067489624, "learning_rate": 4.595790153993918e-11, "loss": 0.0045, "step": 99960 }, { "grad_norm": 0.17845052480697632, "learning_rate": 2.6273376729779585e-11, "loss": 0.005, "step": 99970 }, { "grad_norm": 0.12026455253362656, "learning_rate": 1.2056773868174276e-11, "loss": 0.0063, "step": 99980 }, { "grad_norm": 0.08542952686548233, "learning_rate": 3.3080945149865926e-12, "loss": 0.0073, "step": 99990 }, { "grad_norm": 0.11924929171800613, "learning_rate": 2.7339625008337977e-14, "loss": 0.0053, "step": 100000 } ], "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": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 24, "trial_name": null, "trial_params": null }