{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.9, "eval_steps": 500, "global_step": 90000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 0.7148909568786621, "learning_rate": 1.8e-07, "loss": 1.3553, "step": 10 }, { "grad_norm": 0.7733553647994995, "learning_rate": 3.8e-07, "loss": 1.348, "step": 20 }, { "grad_norm": 0.5662791728973389, "learning_rate": 5.8e-07, "loss": 1.3537, "step": 30 }, { "grad_norm": 0.4940169155597687, "learning_rate": 7.8e-07, "loss": 1.3585, "step": 40 }, { "grad_norm": 0.5199644565582275, "learning_rate": 9.8e-07, "loss": 1.3508, "step": 50 }, { "grad_norm": 0.46559250354766846, "learning_rate": 1.18e-06, "loss": 1.3393, "step": 60 }, { "grad_norm": 1.0874249935150146, "learning_rate": 1.3800000000000001e-06, "loss": 1.3462, "step": 70 }, { "grad_norm": 1.1132304668426514, "learning_rate": 1.5800000000000003e-06, "loss": 1.3245, "step": 80 }, { "grad_norm": 1.5403821468353271, "learning_rate": 1.7800000000000001e-06, "loss": 1.3074, "step": 90 }, { "grad_norm": 1.193524956703186, "learning_rate": 1.98e-06, "loss": 1.2226, "step": 100 }, { "grad_norm": 1.9003077745437622, "learning_rate": 2.1800000000000003e-06, "loss": 1.1826, "step": 110 }, { "grad_norm": 1.0355403423309326, "learning_rate": 2.38e-06, "loss": 1.1711, "step": 120 }, { "grad_norm": 1.3969544172286987, "learning_rate": 2.5800000000000003e-06, "loss": 1.1278, "step": 130 }, { "grad_norm": 1.4826934337615967, "learning_rate": 2.78e-06, "loss": 1.1125, "step": 140 }, { "grad_norm": 1.8849164247512817, "learning_rate": 2.9800000000000003e-06, "loss": 1.119, "step": 150 }, { "grad_norm": 1.1387088298797607, "learning_rate": 3.1800000000000005e-06, "loss": 1.1079, "step": 160 }, { "grad_norm": 1.6981850862503052, "learning_rate": 3.38e-06, "loss": 1.1232, "step": 170 }, { "grad_norm": 0.9863775968551636, "learning_rate": 3.58e-06, "loss": 1.1154, "step": 180 }, { "grad_norm": 1.4443426132202148, "learning_rate": 3.7800000000000002e-06, "loss": 1.1073, "step": 190 }, { "grad_norm": 1.4970171451568604, "learning_rate": 3.98e-06, "loss": 1.1165, "step": 200 }, { "grad_norm": 1.5880851745605469, "learning_rate": 4.18e-06, "loss": 1.1025, "step": 210 }, { "grad_norm": 1.831956386566162, "learning_rate": 4.38e-06, "loss": 1.1139, "step": 220 }, { "grad_norm": 1.7149401903152466, "learning_rate": 4.58e-06, "loss": 1.0847, "step": 230 }, { "grad_norm": 1.6966031789779663, "learning_rate": 4.780000000000001e-06, "loss": 1.0782, "step": 240 }, { "grad_norm": 1.23192298412323, "learning_rate": 4.98e-06, "loss": 1.1168, "step": 250 }, { "grad_norm": 1.1089489459991455, "learning_rate": 5.18e-06, "loss": 1.1039, "step": 260 }, { "grad_norm": 2.828932523727417, "learning_rate": 5.38e-06, "loss": 1.0801, "step": 270 }, { "grad_norm": 0.8576074838638306, "learning_rate": 5.580000000000001e-06, "loss": 1.093, "step": 280 }, { "grad_norm": 1.1523319482803345, "learning_rate": 5.78e-06, "loss": 1.0529, "step": 290 }, { "grad_norm": 2.026623249053955, "learning_rate": 5.98e-06, "loss": 1.08, "step": 300 }, { "grad_norm": 2.504955291748047, "learning_rate": 6.18e-06, "loss": 1.0622, "step": 310 }, { "grad_norm": 1.6328591108322144, "learning_rate": 6.38e-06, "loss": 1.0585, "step": 320 }, { "grad_norm": 2.065979242324829, "learning_rate": 6.58e-06, "loss": 1.0391, "step": 330 }, { "grad_norm": 2.0169336795806885, "learning_rate": 6.78e-06, "loss": 1.0072, "step": 340 }, { "grad_norm": 2.43088698387146, "learning_rate": 6.98e-06, "loss": 0.9913, "step": 350 }, { "grad_norm": 2.630277633666992, "learning_rate": 7.180000000000001e-06, "loss": 0.9812, "step": 360 }, { "grad_norm": 2.9819109439849854, "learning_rate": 7.3800000000000005e-06, "loss": 0.9512, "step": 370 }, { "grad_norm": 2.32175350189209, "learning_rate": 7.580000000000001e-06, "loss": 0.8875, "step": 380 }, { "grad_norm": 2.8873038291931152, "learning_rate": 7.78e-06, "loss": 0.8379, "step": 390 }, { "grad_norm": 2.973228693008423, "learning_rate": 7.98e-06, "loss": 0.7683, "step": 400 }, { "grad_norm": 3.0076844692230225, "learning_rate": 8.18e-06, "loss": 0.6866, "step": 410 }, { "grad_norm": 2.9163527488708496, "learning_rate": 8.380000000000001e-06, "loss": 0.609, "step": 420 }, { "grad_norm": 3.2866177558898926, "learning_rate": 8.580000000000001e-06, "loss": 0.568, "step": 430 }, { "grad_norm": 3.3103508949279785, "learning_rate": 8.78e-06, "loss": 0.5343, "step": 440 }, { "grad_norm": 3.0231640338897705, "learning_rate": 8.98e-06, "loss": 0.447, "step": 450 }, { "grad_norm": 4.8128228187561035, "learning_rate": 9.180000000000002e-06, "loss": 0.4235, "step": 460 }, { "grad_norm": 3.48710298538208, "learning_rate": 9.38e-06, "loss": 0.3702, "step": 470 }, { "grad_norm": 3.650515079498291, "learning_rate": 9.58e-06, "loss": 0.3416, "step": 480 }, { "grad_norm": 3.3632960319519043, "learning_rate": 9.78e-06, "loss": 0.3375, "step": 490 }, { "grad_norm": 3.674513578414917, "learning_rate": 9.980000000000001e-06, "loss": 0.2905, "step": 500 }, { "grad_norm": 3.0074410438537598, "learning_rate": 1.018e-05, "loss": 0.3001, "step": 510 }, { "grad_norm": 2.2805633544921875, "learning_rate": 1.038e-05, "loss": 0.2653, "step": 520 }, { "grad_norm": 3.3534135818481445, "learning_rate": 1.058e-05, "loss": 0.2672, "step": 530 }, { "grad_norm": 3.115891218185425, "learning_rate": 1.0780000000000002e-05, "loss": 0.2489, "step": 540 }, { "grad_norm": 3.4248104095458984, "learning_rate": 1.098e-05, "loss": 0.2695, "step": 550 }, { "grad_norm": 2.8082008361816406, "learning_rate": 1.118e-05, "loss": 0.2304, "step": 560 }, { "grad_norm": 3.221757173538208, "learning_rate": 1.1380000000000001e-05, "loss": 0.2461, "step": 570 }, { "grad_norm": 3.737562417984009, "learning_rate": 1.1580000000000001e-05, "loss": 0.2285, "step": 580 }, { "grad_norm": 3.7301676273345947, "learning_rate": 1.178e-05, "loss": 0.2339, "step": 590 }, { "grad_norm": 2.666806936264038, "learning_rate": 1.198e-05, "loss": 0.2263, "step": 600 }, { "grad_norm": 3.5371651649475098, "learning_rate": 1.2180000000000002e-05, "loss": 0.2213, "step": 610 }, { "grad_norm": 3.0354127883911133, "learning_rate": 1.238e-05, "loss": 0.2299, "step": 620 }, { "grad_norm": 2.3548262119293213, "learning_rate": 1.258e-05, "loss": 0.1902, "step": 630 }, { "grad_norm": 2.843712329864502, "learning_rate": 1.278e-05, "loss": 0.2161, "step": 640 }, { "grad_norm": 3.040304183959961, "learning_rate": 1.2980000000000001e-05, "loss": 0.1909, "step": 650 }, { "grad_norm": 2.5794312953948975, "learning_rate": 1.3180000000000001e-05, "loss": 0.2041, "step": 660 }, { "grad_norm": 2.928281784057617, "learning_rate": 1.338e-05, "loss": 0.2001, "step": 670 }, { "grad_norm": 2.4542737007141113, "learning_rate": 1.358e-05, "loss": 0.1857, "step": 680 }, { "grad_norm": 3.1455600261688232, "learning_rate": 1.3780000000000002e-05, "loss": 0.1867, "step": 690 }, { "grad_norm": 2.8499414920806885, "learning_rate": 1.3980000000000002e-05, "loss": 0.196, "step": 700 }, { "grad_norm": 3.0325703620910645, "learning_rate": 1.4180000000000001e-05, "loss": 0.1948, "step": 710 }, { "grad_norm": 2.5370843410491943, "learning_rate": 1.4380000000000001e-05, "loss": 0.1861, "step": 720 }, { "grad_norm": 2.966275215148926, "learning_rate": 1.4580000000000003e-05, "loss": 0.1899, "step": 730 }, { "grad_norm": 1.8963969945907593, "learning_rate": 1.4779999999999999e-05, "loss": 0.1733, "step": 740 }, { "grad_norm": 2.3449273109436035, "learning_rate": 1.4979999999999999e-05, "loss": 0.1655, "step": 750 }, { "grad_norm": 2.4902689456939697, "learning_rate": 1.518e-05, "loss": 0.1569, "step": 760 }, { "grad_norm": 1.896379828453064, "learning_rate": 1.538e-05, "loss": 0.1649, "step": 770 }, { "grad_norm": 1.7662169933319092, "learning_rate": 1.558e-05, "loss": 0.1529, "step": 780 }, { "grad_norm": 2.2506911754608154, "learning_rate": 1.578e-05, "loss": 0.1674, "step": 790 }, { "grad_norm": 2.180724620819092, "learning_rate": 1.598e-05, "loss": 0.169, "step": 800 }, { "grad_norm": 3.0045385360717773, "learning_rate": 1.618e-05, "loss": 0.1896, "step": 810 }, { "grad_norm": 1.8347951173782349, "learning_rate": 1.6380000000000002e-05, "loss": 0.1651, "step": 820 }, { "grad_norm": 2.6088662147521973, "learning_rate": 1.658e-05, "loss": 0.1582, "step": 830 }, { "grad_norm": 2.142300605773926, "learning_rate": 1.6780000000000002e-05, "loss": 0.155, "step": 840 }, { "grad_norm": 1.8997673988342285, "learning_rate": 1.698e-05, "loss": 0.1429, "step": 850 }, { "grad_norm": 2.5325191020965576, "learning_rate": 1.718e-05, "loss": 0.1472, "step": 860 }, { "grad_norm": 3.250920534133911, "learning_rate": 1.7380000000000003e-05, "loss": 0.159, "step": 870 }, { "grad_norm": 2.424344062805176, "learning_rate": 1.758e-05, "loss": 0.1505, "step": 880 }, { "grad_norm": 2.2986509799957275, "learning_rate": 1.7780000000000003e-05, "loss": 0.1556, "step": 890 }, { "grad_norm": 2.0581161975860596, "learning_rate": 1.798e-05, "loss": 0.1534, "step": 900 }, { "grad_norm": 1.8022172451019287, "learning_rate": 1.818e-05, "loss": 0.1588, "step": 910 }, { "grad_norm": 2.5716609954833984, "learning_rate": 1.838e-05, "loss": 0.1427, "step": 920 }, { "grad_norm": 2.1608235836029053, "learning_rate": 1.858e-05, "loss": 0.1646, "step": 930 }, { "grad_norm": 2.5065858364105225, "learning_rate": 1.878e-05, "loss": 0.1614, "step": 940 }, { "grad_norm": 1.799443244934082, "learning_rate": 1.898e-05, "loss": 0.1476, "step": 950 }, { "grad_norm": 1.2547683715820312, "learning_rate": 1.918e-05, "loss": 0.1406, "step": 960 }, { "grad_norm": 1.9115350246429443, "learning_rate": 1.938e-05, "loss": 0.1369, "step": 970 }, { "grad_norm": 1.7809206247329712, "learning_rate": 1.9580000000000002e-05, "loss": 0.1381, "step": 980 }, { "grad_norm": 1.6764086484909058, "learning_rate": 1.978e-05, "loss": 0.1355, "step": 990 }, { "grad_norm": 1.9880117177963257, "learning_rate": 1.9980000000000002e-05, "loss": 0.1501, "step": 1000 }, { "grad_norm": 1.5587902069091797, "learning_rate": 2.0180000000000003e-05, "loss": 0.1267, "step": 1010 }, { "grad_norm": 2.183574914932251, "learning_rate": 2.038e-05, "loss": 0.139, "step": 1020 }, { "grad_norm": 1.8879772424697876, "learning_rate": 2.0580000000000003e-05, "loss": 0.1419, "step": 1030 }, { "grad_norm": 1.5179930925369263, "learning_rate": 2.078e-05, "loss": 0.1308, "step": 1040 }, { "grad_norm": 2.211318016052246, "learning_rate": 2.098e-05, "loss": 0.1345, "step": 1050 }, { "grad_norm": 2.3245046138763428, "learning_rate": 2.118e-05, "loss": 0.1477, "step": 1060 }, { "grad_norm": 1.3649100065231323, "learning_rate": 2.138e-05, "loss": 0.1361, "step": 1070 }, { "grad_norm": 1.7131909132003784, "learning_rate": 2.158e-05, "loss": 0.1422, "step": 1080 }, { "grad_norm": 1.6704899072647095, "learning_rate": 2.178e-05, "loss": 0.1385, "step": 1090 }, { "grad_norm": 1.5453004837036133, "learning_rate": 2.198e-05, "loss": 0.1133, "step": 1100 }, { "grad_norm": 2.3557870388031006, "learning_rate": 2.218e-05, "loss": 0.1376, "step": 1110 }, { "grad_norm": 1.2919601202011108, "learning_rate": 2.2380000000000003e-05, "loss": 0.1216, "step": 1120 }, { "grad_norm": 1.680527925491333, "learning_rate": 2.258e-05, "loss": 0.1407, "step": 1130 }, { "grad_norm": 1.6557466983795166, "learning_rate": 2.2780000000000002e-05, "loss": 0.1264, "step": 1140 }, { "grad_norm": 2.0247550010681152, "learning_rate": 2.298e-05, "loss": 0.1394, "step": 1150 }, { "grad_norm": 1.8177584409713745, "learning_rate": 2.318e-05, "loss": 0.1334, "step": 1160 }, { "grad_norm": 2.4169247150421143, "learning_rate": 2.3380000000000003e-05, "loss": 0.1328, "step": 1170 }, { "grad_norm": 1.5977981090545654, "learning_rate": 2.358e-05, "loss": 0.1227, "step": 1180 }, { "grad_norm": 1.7045165300369263, "learning_rate": 2.3780000000000003e-05, "loss": 0.1248, "step": 1190 }, { "grad_norm": 1.4435311555862427, "learning_rate": 2.398e-05, "loss": 0.1369, "step": 1200 }, { "grad_norm": 1.2464182376861572, "learning_rate": 2.418e-05, "loss": 0.1131, "step": 1210 }, { "grad_norm": 1.5733189582824707, "learning_rate": 2.438e-05, "loss": 0.1178, "step": 1220 }, { "grad_norm": 1.431103229522705, "learning_rate": 2.4580000000000002e-05, "loss": 0.1275, "step": 1230 }, { "grad_norm": 1.6488111019134521, "learning_rate": 2.478e-05, "loss": 0.1156, "step": 1240 }, { "grad_norm": 1.1241775751113892, "learning_rate": 2.498e-05, "loss": 0.1151, "step": 1250 }, { "grad_norm": 1.3198717832565308, "learning_rate": 2.5180000000000003e-05, "loss": 0.1314, "step": 1260 }, { "grad_norm": 1.4666950702667236, "learning_rate": 2.5380000000000004e-05, "loss": 0.126, "step": 1270 }, { "grad_norm": 1.079308271408081, "learning_rate": 2.5580000000000002e-05, "loss": 0.1101, "step": 1280 }, { "grad_norm": 1.5161341428756714, "learning_rate": 2.5779999999999997e-05, "loss": 0.1152, "step": 1290 }, { "grad_norm": 1.3645700216293335, "learning_rate": 2.598e-05, "loss": 0.116, "step": 1300 }, { "grad_norm": 1.6512712240219116, "learning_rate": 2.618e-05, "loss": 0.1085, "step": 1310 }, { "grad_norm": 1.21893310546875, "learning_rate": 2.6379999999999998e-05, "loss": 0.1398, "step": 1320 }, { "grad_norm": 1.6886892318725586, "learning_rate": 2.658e-05, "loss": 0.1288, "step": 1330 }, { "grad_norm": 1.0771760940551758, "learning_rate": 2.678e-05, "loss": 0.1159, "step": 1340 }, { "grad_norm": 1.5079783201217651, "learning_rate": 2.698e-05, "loss": 0.1147, "step": 1350 }, { "grad_norm": 1.2618318796157837, "learning_rate": 2.718e-05, "loss": 0.1194, "step": 1360 }, { "grad_norm": 1.3514701128005981, "learning_rate": 2.738e-05, "loss": 0.1155, "step": 1370 }, { "grad_norm": 1.337638258934021, "learning_rate": 2.758e-05, "loss": 0.1209, "step": 1380 }, { "grad_norm": 1.1854989528656006, "learning_rate": 2.778e-05, "loss": 0.118, "step": 1390 }, { "grad_norm": 1.2821794748306274, "learning_rate": 2.798e-05, "loss": 0.1069, "step": 1400 }, { "grad_norm": 1.4120070934295654, "learning_rate": 2.818e-05, "loss": 0.1293, "step": 1410 }, { "grad_norm": 1.4999290704727173, "learning_rate": 2.8380000000000003e-05, "loss": 0.1285, "step": 1420 }, { "grad_norm": 1.800325870513916, "learning_rate": 2.858e-05, "loss": 0.1165, "step": 1430 }, { "grad_norm": 1.403993844985962, "learning_rate": 2.8780000000000002e-05, "loss": 0.1167, "step": 1440 }, { "grad_norm": 0.8991517424583435, "learning_rate": 2.898e-05, "loss": 0.1046, "step": 1450 }, { "grad_norm": 1.4236302375793457, "learning_rate": 2.9180000000000002e-05, "loss": 0.1106, "step": 1460 }, { "grad_norm": 1.2315449714660645, "learning_rate": 2.9380000000000003e-05, "loss": 0.1101, "step": 1470 }, { "grad_norm": 1.5821279287338257, "learning_rate": 2.958e-05, "loss": 0.1204, "step": 1480 }, { "grad_norm": 1.348388671875, "learning_rate": 2.9780000000000003e-05, "loss": 0.121, "step": 1490 }, { "grad_norm": 1.390174388885498, "learning_rate": 2.998e-05, "loss": 0.1114, "step": 1500 }, { "grad_norm": 1.423020601272583, "learning_rate": 3.0180000000000002e-05, "loss": 0.1163, "step": 1510 }, { "grad_norm": 1.3241568803787231, "learning_rate": 3.0380000000000004e-05, "loss": 0.1264, "step": 1520 }, { "grad_norm": 1.365360975265503, "learning_rate": 3.058e-05, "loss": 0.1183, "step": 1530 }, { "grad_norm": 1.507880687713623, "learning_rate": 3.078e-05, "loss": 0.1032, "step": 1540 }, { "grad_norm": 1.018844723701477, "learning_rate": 3.0980000000000005e-05, "loss": 0.1212, "step": 1550 }, { "grad_norm": 1.2256401777267456, "learning_rate": 3.118e-05, "loss": 0.1252, "step": 1560 }, { "grad_norm": 1.1648406982421875, "learning_rate": 3.138e-05, "loss": 0.115, "step": 1570 }, { "grad_norm": 1.4266129732131958, "learning_rate": 3.1580000000000006e-05, "loss": 0.1156, "step": 1580 }, { "grad_norm": 1.3068782091140747, "learning_rate": 3.1780000000000004e-05, "loss": 0.1149, "step": 1590 }, { "grad_norm": 0.829749345779419, "learning_rate": 3.198e-05, "loss": 0.1182, "step": 1600 }, { "grad_norm": 1.1183444261550903, "learning_rate": 3.218e-05, "loss": 0.1146, "step": 1610 }, { "grad_norm": 0.9819296002388, "learning_rate": 3.238e-05, "loss": 0.1144, "step": 1620 }, { "grad_norm": 1.2433115243911743, "learning_rate": 3.2579999999999996e-05, "loss": 0.1044, "step": 1630 }, { "grad_norm": 1.3647418022155762, "learning_rate": 3.278e-05, "loss": 0.1067, "step": 1640 }, { "grad_norm": 1.0699095726013184, "learning_rate": 3.298e-05, "loss": 0.1037, "step": 1650 }, { "grad_norm": 1.4841372966766357, "learning_rate": 3.318e-05, "loss": 0.1177, "step": 1660 }, { "grad_norm": 0.9250832200050354, "learning_rate": 3.338e-05, "loss": 0.1089, "step": 1670 }, { "grad_norm": 1.1990487575531006, "learning_rate": 3.358e-05, "loss": 0.0992, "step": 1680 }, { "grad_norm": 1.1159650087356567, "learning_rate": 3.378e-05, "loss": 0.1223, "step": 1690 }, { "grad_norm": 1.066735863685608, "learning_rate": 3.398e-05, "loss": 0.107, "step": 1700 }, { "grad_norm": 1.1770398616790771, "learning_rate": 3.418e-05, "loss": 0.1269, "step": 1710 }, { "grad_norm": 0.9545903205871582, "learning_rate": 3.438e-05, "loss": 0.1144, "step": 1720 }, { "grad_norm": 1.1905124187469482, "learning_rate": 3.4580000000000004e-05, "loss": 0.1043, "step": 1730 }, { "grad_norm": 1.0317710638046265, "learning_rate": 3.478e-05, "loss": 0.117, "step": 1740 }, { "grad_norm": 1.1006184816360474, "learning_rate": 3.498e-05, "loss": 0.1157, "step": 1750 }, { "grad_norm": 0.9281287789344788, "learning_rate": 3.518e-05, "loss": 0.1167, "step": 1760 }, { "grad_norm": 0.8545994162559509, "learning_rate": 3.5380000000000003e-05, "loss": 0.1076, "step": 1770 }, { "grad_norm": 0.860276997089386, "learning_rate": 3.558e-05, "loss": 0.1111, "step": 1780 }, { "grad_norm": 1.1597980260849, "learning_rate": 3.578e-05, "loss": 0.1029, "step": 1790 }, { "grad_norm": 1.2760350704193115, "learning_rate": 3.5980000000000004e-05, "loss": 0.1056, "step": 1800 }, { "grad_norm": 1.228945255279541, "learning_rate": 3.618e-05, "loss": 0.1139, "step": 1810 }, { "grad_norm": 0.9193392395973206, "learning_rate": 3.638e-05, "loss": 0.1149, "step": 1820 }, { "grad_norm": 1.246351957321167, "learning_rate": 3.6580000000000006e-05, "loss": 0.1056, "step": 1830 }, { "grad_norm": 1.0556213855743408, "learning_rate": 3.6780000000000004e-05, "loss": 0.11, "step": 1840 }, { "grad_norm": 0.8405678272247314, "learning_rate": 3.698e-05, "loss": 0.0976, "step": 1850 }, { "grad_norm": 1.4897795915603638, "learning_rate": 3.7180000000000007e-05, "loss": 0.1175, "step": 1860 }, { "grad_norm": 0.7007404565811157, "learning_rate": 3.7380000000000005e-05, "loss": 0.1066, "step": 1870 }, { "grad_norm": 1.3516809940338135, "learning_rate": 3.758e-05, "loss": 0.0979, "step": 1880 }, { "grad_norm": 1.3927785158157349, "learning_rate": 3.778000000000001e-05, "loss": 0.1102, "step": 1890 }, { "grad_norm": 0.9046825170516968, "learning_rate": 3.7980000000000006e-05, "loss": 0.1165, "step": 1900 }, { "grad_norm": 0.9050040245056152, "learning_rate": 3.818e-05, "loss": 0.1018, "step": 1910 }, { "grad_norm": 0.7670705914497375, "learning_rate": 3.838e-05, "loss": 0.0968, "step": 1920 }, { "grad_norm": 1.0595147609710693, "learning_rate": 3.858e-05, "loss": 0.1016, "step": 1930 }, { "grad_norm": 1.0906118154525757, "learning_rate": 3.878e-05, "loss": 0.1028, "step": 1940 }, { "grad_norm": 0.9248902201652527, "learning_rate": 3.898e-05, "loss": 0.1056, "step": 1950 }, { "grad_norm": 1.0650578737258911, "learning_rate": 3.918e-05, "loss": 0.1127, "step": 1960 }, { "grad_norm": 0.8802879452705383, "learning_rate": 3.938e-05, "loss": 0.1193, "step": 1970 }, { "grad_norm": 0.9944506287574768, "learning_rate": 3.958e-05, "loss": 0.1086, "step": 1980 }, { "grad_norm": 1.110724925994873, "learning_rate": 3.978e-05, "loss": 0.113, "step": 1990 }, { "grad_norm": 0.9517561793327332, "learning_rate": 3.998e-05, "loss": 0.1183, "step": 2000 }, { "grad_norm": 1.0140929222106934, "learning_rate": 4.018e-05, "loss": 0.1038, "step": 2010 }, { "grad_norm": 0.8025938272476196, "learning_rate": 4.038e-05, "loss": 0.1071, "step": 2020 }, { "grad_norm": 1.0816075801849365, "learning_rate": 4.058e-05, "loss": 0.1181, "step": 2030 }, { "grad_norm": 1.1138700246810913, "learning_rate": 4.078e-05, "loss": 0.1123, "step": 2040 }, { "grad_norm": 1.0064568519592285, "learning_rate": 4.0980000000000004e-05, "loss": 0.115, "step": 2050 }, { "grad_norm": 0.9331321716308594, "learning_rate": 4.118e-05, "loss": 0.1116, "step": 2060 }, { "grad_norm": 1.1795645952224731, "learning_rate": 4.138e-05, "loss": 0.1076, "step": 2070 }, { "grad_norm": 0.8831111192703247, "learning_rate": 4.1580000000000005e-05, "loss": 0.1006, "step": 2080 }, { "grad_norm": 0.8879129886627197, "learning_rate": 4.178e-05, "loss": 0.1037, "step": 2090 }, { "grad_norm": 1.222538948059082, "learning_rate": 4.198e-05, "loss": 0.1023, "step": 2100 }, { "grad_norm": 0.876069962978363, "learning_rate": 4.2180000000000006e-05, "loss": 0.1085, "step": 2110 }, { "grad_norm": 0.8799285292625427, "learning_rate": 4.2380000000000004e-05, "loss": 0.1135, "step": 2120 }, { "grad_norm": 1.2035514116287231, "learning_rate": 4.258e-05, "loss": 0.0989, "step": 2130 }, { "grad_norm": 0.8126112818717957, "learning_rate": 4.278e-05, "loss": 0.0935, "step": 2140 }, { "grad_norm": 0.9505549073219299, "learning_rate": 4.2980000000000005e-05, "loss": 0.1041, "step": 2150 }, { "grad_norm": 0.7023047804832458, "learning_rate": 4.318e-05, "loss": 0.1023, "step": 2160 }, { "grad_norm": 0.9504386186599731, "learning_rate": 4.338e-05, "loss": 0.1033, "step": 2170 }, { "grad_norm": 1.1097073554992676, "learning_rate": 4.3580000000000006e-05, "loss": 0.1042, "step": 2180 }, { "grad_norm": 1.2179919481277466, "learning_rate": 4.3780000000000004e-05, "loss": 0.108, "step": 2190 }, { "grad_norm": 1.0617380142211914, "learning_rate": 4.398e-05, "loss": 0.1133, "step": 2200 }, { "grad_norm": 0.7858248353004456, "learning_rate": 4.418000000000001e-05, "loss": 0.0973, "step": 2210 }, { "grad_norm": 1.0753215551376343, "learning_rate": 4.438e-05, "loss": 0.0996, "step": 2220 }, { "grad_norm": 0.8307908177375793, "learning_rate": 4.458e-05, "loss": 0.0971, "step": 2230 }, { "grad_norm": 1.127766489982605, "learning_rate": 4.478e-05, "loss": 0.1059, "step": 2240 }, { "grad_norm": 0.7246810793876648, "learning_rate": 4.498e-05, "loss": 0.105, "step": 2250 }, { "grad_norm": 0.871928870677948, "learning_rate": 4.518e-05, "loss": 0.1079, "step": 2260 }, { "grad_norm": 0.9735716581344604, "learning_rate": 4.538e-05, "loss": 0.1006, "step": 2270 }, { "grad_norm": 1.0605742931365967, "learning_rate": 4.558e-05, "loss": 0.1014, "step": 2280 }, { "grad_norm": 0.9977045655250549, "learning_rate": 4.578e-05, "loss": 0.0975, "step": 2290 }, { "grad_norm": 0.8908712267875671, "learning_rate": 4.5980000000000004e-05, "loss": 0.1, "step": 2300 }, { "grad_norm": 0.7256587743759155, "learning_rate": 4.618e-05, "loss": 0.1009, "step": 2310 }, { "grad_norm": 0.7081485390663147, "learning_rate": 4.638e-05, "loss": 0.0853, "step": 2320 }, { "grad_norm": 0.8117384910583496, "learning_rate": 4.6580000000000005e-05, "loss": 0.0938, "step": 2330 }, { "grad_norm": 0.911388099193573, "learning_rate": 4.678e-05, "loss": 0.0983, "step": 2340 }, { "grad_norm": 0.71697598695755, "learning_rate": 4.698e-05, "loss": 0.1007, "step": 2350 }, { "grad_norm": 0.855356752872467, "learning_rate": 4.718e-05, "loss": 0.1093, "step": 2360 }, { "grad_norm": 0.8690946698188782, "learning_rate": 4.7380000000000004e-05, "loss": 0.1042, "step": 2370 }, { "grad_norm": 0.6823745369911194, "learning_rate": 4.758e-05, "loss": 0.0975, "step": 2380 }, { "grad_norm": 0.728854775428772, "learning_rate": 4.778e-05, "loss": 0.1032, "step": 2390 }, { "grad_norm": 0.8282954096794128, "learning_rate": 4.7980000000000005e-05, "loss": 0.0876, "step": 2400 }, { "grad_norm": 0.9815107583999634, "learning_rate": 4.818e-05, "loss": 0.1006, "step": 2410 }, { "grad_norm": 0.8056247234344482, "learning_rate": 4.838e-05, "loss": 0.097, "step": 2420 }, { "grad_norm": 0.7111151218414307, "learning_rate": 4.8580000000000006e-05, "loss": 0.0976, "step": 2430 }, { "grad_norm": 0.5682677030563354, "learning_rate": 4.8780000000000004e-05, "loss": 0.0984, "step": 2440 }, { "grad_norm": 0.6743620038032532, "learning_rate": 4.898e-05, "loss": 0.0915, "step": 2450 }, { "grad_norm": 0.8257128000259399, "learning_rate": 4.918000000000001e-05, "loss": 0.0898, "step": 2460 }, { "grad_norm": 0.9174513816833496, "learning_rate": 4.9380000000000005e-05, "loss": 0.0882, "step": 2470 }, { "grad_norm": 0.8507484197616577, "learning_rate": 4.958e-05, "loss": 0.1023, "step": 2480 }, { "grad_norm": 0.8523046970367432, "learning_rate": 4.978e-05, "loss": 0.0931, "step": 2490 }, { "grad_norm": 0.8273335695266724, "learning_rate": 4.9980000000000006e-05, "loss": 0.0969, "step": 2500 }, { "grad_norm": 0.9451372623443604, "learning_rate": 5.0180000000000004e-05, "loss": 0.085, "step": 2510 }, { "grad_norm": 0.6669540405273438, "learning_rate": 5.038e-05, "loss": 0.0935, "step": 2520 }, { "grad_norm": 1.0917842388153076, "learning_rate": 5.058000000000001e-05, "loss": 0.109, "step": 2530 }, { "grad_norm": 0.7812323570251465, "learning_rate": 5.0780000000000005e-05, "loss": 0.1129, "step": 2540 }, { "grad_norm": 0.8318679928779602, "learning_rate": 5.098e-05, "loss": 0.0979, "step": 2550 }, { "grad_norm": 0.8776370286941528, "learning_rate": 5.118000000000001e-05, "loss": 0.0935, "step": 2560 }, { "grad_norm": 1.003862738609314, "learning_rate": 5.1380000000000006e-05, "loss": 0.0917, "step": 2570 }, { "grad_norm": 0.742704451084137, "learning_rate": 5.1580000000000004e-05, "loss": 0.099, "step": 2580 }, { "grad_norm": 0.7036648988723755, "learning_rate": 5.178000000000001e-05, "loss": 0.0884, "step": 2590 }, { "grad_norm": 0.7683124542236328, "learning_rate": 5.198000000000001e-05, "loss": 0.0939, "step": 2600 }, { "grad_norm": 0.9644230604171753, "learning_rate": 5.2180000000000005e-05, "loss": 0.0844, "step": 2610 }, { "grad_norm": 0.6866705417633057, "learning_rate": 5.238000000000001e-05, "loss": 0.0933, "step": 2620 }, { "grad_norm": 0.8024722337722778, "learning_rate": 5.258000000000001e-05, "loss": 0.1068, "step": 2630 }, { "grad_norm": 1.0606228113174438, "learning_rate": 5.2780000000000006e-05, "loss": 0.1188, "step": 2640 }, { "grad_norm": 0.8449191451072693, "learning_rate": 5.2980000000000004e-05, "loss": 0.098, "step": 2650 }, { "grad_norm": 0.7770112156867981, "learning_rate": 5.318000000000001e-05, "loss": 0.1027, "step": 2660 }, { "grad_norm": 0.7105051875114441, "learning_rate": 5.338000000000001e-05, "loss": 0.0818, "step": 2670 }, { "grad_norm": 1.040837049484253, "learning_rate": 5.3580000000000005e-05, "loss": 0.0854, "step": 2680 }, { "grad_norm": 0.6928813457489014, "learning_rate": 5.378e-05, "loss": 0.083, "step": 2690 }, { "grad_norm": 0.7051820755004883, "learning_rate": 5.3979999999999995e-05, "loss": 0.1006, "step": 2700 }, { "grad_norm": 1.0461666584014893, "learning_rate": 5.418e-05, "loss": 0.1004, "step": 2710 }, { "grad_norm": 0.6785884499549866, "learning_rate": 5.438e-05, "loss": 0.0917, "step": 2720 }, { "grad_norm": 0.8983668684959412, "learning_rate": 5.4579999999999996e-05, "loss": 0.0922, "step": 2730 }, { "grad_norm": 0.8253235220909119, "learning_rate": 5.478e-05, "loss": 0.0833, "step": 2740 }, { "grad_norm": 0.7489354014396667, "learning_rate": 5.498e-05, "loss": 0.0858, "step": 2750 }, { "grad_norm": 1.0161556005477905, "learning_rate": 5.518e-05, "loss": 0.0991, "step": 2760 }, { "grad_norm": 0.9255701899528503, "learning_rate": 5.538e-05, "loss": 0.0906, "step": 2770 }, { "grad_norm": 0.8022998571395874, "learning_rate": 5.558e-05, "loss": 0.0852, "step": 2780 }, { "grad_norm": 0.780035138130188, "learning_rate": 5.578e-05, "loss": 0.0917, "step": 2790 }, { "grad_norm": 1.0644786357879639, "learning_rate": 5.5979999999999996e-05, "loss": 0.0913, "step": 2800 }, { "grad_norm": 0.5021172165870667, "learning_rate": 5.618e-05, "loss": 0.0822, "step": 2810 }, { "grad_norm": 0.7204877138137817, "learning_rate": 5.638e-05, "loss": 0.091, "step": 2820 }, { "grad_norm": 0.6375316977500916, "learning_rate": 5.658e-05, "loss": 0.0866, "step": 2830 }, { "grad_norm": 0.7513657808303833, "learning_rate": 5.678e-05, "loss": 0.0898, "step": 2840 }, { "grad_norm": 0.6159160733222961, "learning_rate": 5.698e-05, "loss": 0.0879, "step": 2850 }, { "grad_norm": 0.7915671467781067, "learning_rate": 5.718e-05, "loss": 0.0967, "step": 2860 }, { "grad_norm": 0.7962520718574524, "learning_rate": 5.738e-05, "loss": 0.0784, "step": 2870 }, { "grad_norm": 0.7273536920547485, "learning_rate": 5.758e-05, "loss": 0.0902, "step": 2880 }, { "grad_norm": 0.7342036366462708, "learning_rate": 5.778e-05, "loss": 0.0824, "step": 2890 }, { "grad_norm": 0.6373608708381653, "learning_rate": 5.7980000000000004e-05, "loss": 0.0829, "step": 2900 }, { "grad_norm": 0.687831461429596, "learning_rate": 5.818e-05, "loss": 0.0918, "step": 2910 }, { "grad_norm": 0.5794326663017273, "learning_rate": 5.838e-05, "loss": 0.0832, "step": 2920 }, { "grad_norm": 0.9133846759796143, "learning_rate": 5.858e-05, "loss": 0.0833, "step": 2930 }, { "grad_norm": 0.7872337698936462, "learning_rate": 5.878e-05, "loss": 0.0832, "step": 2940 }, { "grad_norm": 1.0088694095611572, "learning_rate": 5.898e-05, "loss": 0.0934, "step": 2950 }, { "grad_norm": 0.5904010534286499, "learning_rate": 5.918e-05, "loss": 0.0908, "step": 2960 }, { "grad_norm": 0.7214100360870361, "learning_rate": 5.9380000000000004e-05, "loss": 0.0915, "step": 2970 }, { "grad_norm": 0.7613206505775452, "learning_rate": 5.958e-05, "loss": 0.0876, "step": 2980 }, { "grad_norm": 0.6639831066131592, "learning_rate": 5.978e-05, "loss": 0.0777, "step": 2990 }, { "grad_norm": 0.6469506025314331, "learning_rate": 5.9980000000000005e-05, "loss": 0.0981, "step": 3000 }, { "grad_norm": 0.676604688167572, "learning_rate": 6.018e-05, "loss": 0.0897, "step": 3010 }, { "grad_norm": 0.6771364808082581, "learning_rate": 6.038e-05, "loss": 0.0883, "step": 3020 }, { "grad_norm": 0.7683243155479431, "learning_rate": 6.0580000000000006e-05, "loss": 0.0895, "step": 3030 }, { "grad_norm": 0.4868093729019165, "learning_rate": 6.0780000000000004e-05, "loss": 0.0878, "step": 3040 }, { "grad_norm": 0.5788745284080505, "learning_rate": 6.098e-05, "loss": 0.0913, "step": 3050 }, { "grad_norm": 0.6766509413719177, "learning_rate": 6.118000000000001e-05, "loss": 0.081, "step": 3060 }, { "grad_norm": 0.8000262975692749, "learning_rate": 6.138e-05, "loss": 0.0872, "step": 3070 }, { "grad_norm": 0.62906813621521, "learning_rate": 6.158e-05, "loss": 0.096, "step": 3080 }, { "grad_norm": 0.7612301707267761, "learning_rate": 6.178000000000001e-05, "loss": 0.0809, "step": 3090 }, { "grad_norm": 0.7531114816665649, "learning_rate": 6.198e-05, "loss": 0.1, "step": 3100 }, { "grad_norm": 0.6424494981765747, "learning_rate": 6.218e-05, "loss": 0.0862, "step": 3110 }, { "grad_norm": 0.700787365436554, "learning_rate": 6.238000000000001e-05, "loss": 0.0864, "step": 3120 }, { "grad_norm": 0.7568230628967285, "learning_rate": 6.258e-05, "loss": 0.0854, "step": 3130 }, { "grad_norm": 1.0745251178741455, "learning_rate": 6.278e-05, "loss": 0.1095, "step": 3140 }, { "grad_norm": 0.9170243740081787, "learning_rate": 6.298000000000001e-05, "loss": 0.0958, "step": 3150 }, { "grad_norm": 0.7342617511749268, "learning_rate": 6.318e-05, "loss": 0.0985, "step": 3160 }, { "grad_norm": 0.6538165807723999, "learning_rate": 6.338e-05, "loss": 0.086, "step": 3170 }, { "grad_norm": 0.6960799098014832, "learning_rate": 6.358000000000001e-05, "loss": 0.093, "step": 3180 }, { "grad_norm": 0.665551483631134, "learning_rate": 6.378e-05, "loss": 0.0957, "step": 3190 }, { "grad_norm": 0.7369246482849121, "learning_rate": 6.398000000000001e-05, "loss": 0.0918, "step": 3200 }, { "grad_norm": 0.8165823817253113, "learning_rate": 6.418000000000001e-05, "loss": 0.0859, "step": 3210 }, { "grad_norm": 0.8036335706710815, "learning_rate": 6.438e-05, "loss": 0.0939, "step": 3220 }, { "grad_norm": 0.7371277809143066, "learning_rate": 6.458000000000001e-05, "loss": 0.0874, "step": 3230 }, { "grad_norm": 0.7164750099182129, "learning_rate": 6.478000000000001e-05, "loss": 0.0818, "step": 3240 }, { "grad_norm": 0.6649787425994873, "learning_rate": 6.498e-05, "loss": 0.0815, "step": 3250 }, { "grad_norm": 0.4931851625442505, "learning_rate": 6.518000000000001e-05, "loss": 0.091, "step": 3260 }, { "grad_norm": 0.7027093768119812, "learning_rate": 6.538000000000001e-05, "loss": 0.0843, "step": 3270 }, { "grad_norm": 0.6421675086021423, "learning_rate": 6.558e-05, "loss": 0.0896, "step": 3280 }, { "grad_norm": 0.6558473706245422, "learning_rate": 6.578000000000001e-05, "loss": 0.0882, "step": 3290 }, { "grad_norm": 0.684983491897583, "learning_rate": 6.598e-05, "loss": 0.082, "step": 3300 }, { "grad_norm": 0.610790491104126, "learning_rate": 6.618e-05, "loss": 0.0861, "step": 3310 }, { "grad_norm": 0.6636892557144165, "learning_rate": 6.638e-05, "loss": 0.0834, "step": 3320 }, { "grad_norm": 0.5751680135726929, "learning_rate": 6.658e-05, "loss": 0.0876, "step": 3330 }, { "grad_norm": 0.7362626194953918, "learning_rate": 6.678e-05, "loss": 0.0862, "step": 3340 }, { "grad_norm": 0.6782563924789429, "learning_rate": 6.698e-05, "loss": 0.087, "step": 3350 }, { "grad_norm": 0.761106014251709, "learning_rate": 6.718e-05, "loss": 0.0863, "step": 3360 }, { "grad_norm": 0.7741901278495789, "learning_rate": 6.738e-05, "loss": 0.0813, "step": 3370 }, { "grad_norm": 0.7764400839805603, "learning_rate": 6.758e-05, "loss": 0.0919, "step": 3380 }, { "grad_norm": 1.068143606185913, "learning_rate": 6.778e-05, "loss": 0.0899, "step": 3390 }, { "grad_norm": 0.7449162006378174, "learning_rate": 6.798e-05, "loss": 0.1041, "step": 3400 }, { "grad_norm": 0.7341086268424988, "learning_rate": 6.818e-05, "loss": 0.0968, "step": 3410 }, { "grad_norm": 0.5481348633766174, "learning_rate": 6.838e-05, "loss": 0.0929, "step": 3420 }, { "grad_norm": 0.5736236572265625, "learning_rate": 6.858e-05, "loss": 0.0934, "step": 3430 }, { "grad_norm": 0.6417604684829712, "learning_rate": 6.878e-05, "loss": 0.0901, "step": 3440 }, { "grad_norm": 0.785528302192688, "learning_rate": 6.898e-05, "loss": 0.0977, "step": 3450 }, { "grad_norm": 0.707271158695221, "learning_rate": 6.918e-05, "loss": 0.0849, "step": 3460 }, { "grad_norm": 0.5516840219497681, "learning_rate": 6.938e-05, "loss": 0.0923, "step": 3470 }, { "grad_norm": 0.5218857526779175, "learning_rate": 6.958e-05, "loss": 0.0831, "step": 3480 }, { "grad_norm": 0.6308732032775879, "learning_rate": 6.978e-05, "loss": 0.0875, "step": 3490 }, { "grad_norm": 0.644157886505127, "learning_rate": 6.998e-05, "loss": 0.0783, "step": 3500 }, { "grad_norm": 0.6537100672721863, "learning_rate": 7.018e-05, "loss": 0.0831, "step": 3510 }, { "grad_norm": 0.7253512740135193, "learning_rate": 7.038e-05, "loss": 0.0773, "step": 3520 }, { "grad_norm": 0.6004522442817688, "learning_rate": 7.058e-05, "loss": 0.0842, "step": 3530 }, { "grad_norm": 0.599434494972229, "learning_rate": 7.078e-05, "loss": 0.1022, "step": 3540 }, { "grad_norm": 0.682163655757904, "learning_rate": 7.098e-05, "loss": 0.0907, "step": 3550 }, { "grad_norm": 0.5491788983345032, "learning_rate": 7.118e-05, "loss": 0.0897, "step": 3560 }, { "grad_norm": 0.5604795217514038, "learning_rate": 7.138e-05, "loss": 0.0934, "step": 3570 }, { "grad_norm": 0.6893128156661987, "learning_rate": 7.158e-05, "loss": 0.0875, "step": 3580 }, { "grad_norm": 0.7632449269294739, "learning_rate": 7.178000000000001e-05, "loss": 0.0916, "step": 3590 }, { "grad_norm": 0.6013221740722656, "learning_rate": 7.198e-05, "loss": 0.0849, "step": 3600 }, { "grad_norm": 0.7641456723213196, "learning_rate": 7.218e-05, "loss": 0.0856, "step": 3610 }, { "grad_norm": 0.701860249042511, "learning_rate": 7.238000000000001e-05, "loss": 0.0851, "step": 3620 }, { "grad_norm": 0.7968350052833557, "learning_rate": 7.258e-05, "loss": 0.0906, "step": 3630 }, { "grad_norm": 0.4983248710632324, "learning_rate": 7.278e-05, "loss": 0.0792, "step": 3640 }, { "grad_norm": 0.7856201529502869, "learning_rate": 7.298000000000001e-05, "loss": 0.0992, "step": 3650 }, { "grad_norm": 0.6614296436309814, "learning_rate": 7.318e-05, "loss": 0.0902, "step": 3660 }, { "grad_norm": 0.5820555090904236, "learning_rate": 7.338e-05, "loss": 0.0814, "step": 3670 }, { "grad_norm": 0.7379239797592163, "learning_rate": 7.358000000000001e-05, "loss": 0.0811, "step": 3680 }, { "grad_norm": 0.6139934062957764, "learning_rate": 7.378e-05, "loss": 0.089, "step": 3690 }, { "grad_norm": 0.7443228363990784, "learning_rate": 7.398e-05, "loss": 0.0845, "step": 3700 }, { "grad_norm": 0.7788676619529724, "learning_rate": 7.418000000000001e-05, "loss": 0.0936, "step": 3710 }, { "grad_norm": 0.6585743427276611, "learning_rate": 7.438e-05, "loss": 0.0792, "step": 3720 }, { "grad_norm": 0.6545817852020264, "learning_rate": 7.458000000000001e-05, "loss": 0.0892, "step": 3730 }, { "grad_norm": 0.7965897917747498, "learning_rate": 7.478e-05, "loss": 0.095, "step": 3740 }, { "grad_norm": 0.6615657806396484, "learning_rate": 7.498e-05, "loss": 0.0824, "step": 3750 }, { "grad_norm": 0.8216867446899414, "learning_rate": 7.518000000000001e-05, "loss": 0.0821, "step": 3760 }, { "grad_norm": 0.5553804039955139, "learning_rate": 7.538e-05, "loss": 0.0953, "step": 3770 }, { "grad_norm": 0.6225300431251526, "learning_rate": 7.558e-05, "loss": 0.0851, "step": 3780 }, { "grad_norm": 0.5993133187294006, "learning_rate": 7.578000000000001e-05, "loss": 0.0858, "step": 3790 }, { "grad_norm": 0.7722537517547607, "learning_rate": 7.598e-05, "loss": 0.0892, "step": 3800 }, { "grad_norm": 0.7686857581138611, "learning_rate": 7.618e-05, "loss": 0.0913, "step": 3810 }, { "grad_norm": 0.762672483921051, "learning_rate": 7.638000000000001e-05, "loss": 0.0966, "step": 3820 }, { "grad_norm": 0.6061403155326843, "learning_rate": 7.658e-05, "loss": 0.0885, "step": 3830 }, { "grad_norm": 0.5104354023933411, "learning_rate": 7.678000000000001e-05, "loss": 0.0872, "step": 3840 }, { "grad_norm": 0.8424704074859619, "learning_rate": 7.698000000000001e-05, "loss": 0.0903, "step": 3850 }, { "grad_norm": 0.7107453346252441, "learning_rate": 7.718e-05, "loss": 0.087, "step": 3860 }, { "grad_norm": 0.6500034928321838, "learning_rate": 7.738000000000001e-05, "loss": 0.0833, "step": 3870 }, { "grad_norm": 0.4402385354042053, "learning_rate": 7.758000000000001e-05, "loss": 0.0932, "step": 3880 }, { "grad_norm": 0.4190540611743927, "learning_rate": 7.778e-05, "loss": 0.0779, "step": 3890 }, { "grad_norm": 0.6420472264289856, "learning_rate": 7.798000000000001e-05, "loss": 0.0928, "step": 3900 }, { "grad_norm": 0.8854533433914185, "learning_rate": 7.818000000000001e-05, "loss": 0.0843, "step": 3910 }, { "grad_norm": 0.5229454636573792, "learning_rate": 7.838e-05, "loss": 0.0855, "step": 3920 }, { "grad_norm": 0.5419136881828308, "learning_rate": 7.858000000000001e-05, "loss": 0.0782, "step": 3930 }, { "grad_norm": 0.7544001936912537, "learning_rate": 7.878e-05, "loss": 0.0894, "step": 3940 }, { "grad_norm": 0.5005016326904297, "learning_rate": 7.897999999999999e-05, "loss": 0.0969, "step": 3950 }, { "grad_norm": 0.6436410546302795, "learning_rate": 7.918e-05, "loss": 0.0949, "step": 3960 }, { "grad_norm": 0.6740548610687256, "learning_rate": 7.938e-05, "loss": 0.0876, "step": 3970 }, { "grad_norm": 0.72915118932724, "learning_rate": 7.958e-05, "loss": 0.0898, "step": 3980 }, { "grad_norm": 0.5677655339241028, "learning_rate": 7.978e-05, "loss": 0.074, "step": 3990 }, { "grad_norm": 0.7627431154251099, "learning_rate": 7.998e-05, "loss": 0.091, "step": 4000 }, { "grad_norm": 0.7017573118209839, "learning_rate": 8.018e-05, "loss": 0.0899, "step": 4010 }, { "grad_norm": 0.7081181406974792, "learning_rate": 8.038e-05, "loss": 0.0928, "step": 4020 }, { "grad_norm": 0.4589022696018219, "learning_rate": 8.058e-05, "loss": 0.0846, "step": 4030 }, { "grad_norm": 0.6382048726081848, "learning_rate": 8.078e-05, "loss": 0.0833, "step": 4040 }, { "grad_norm": 0.6544201970100403, "learning_rate": 8.098e-05, "loss": 0.0918, "step": 4050 }, { "grad_norm": 0.595919668674469, "learning_rate": 8.118e-05, "loss": 0.0928, "step": 4060 }, { "grad_norm": 0.6490733623504639, "learning_rate": 8.138e-05, "loss": 0.098, "step": 4070 }, { "grad_norm": 0.6665124297142029, "learning_rate": 8.158e-05, "loss": 0.078, "step": 4080 }, { "grad_norm": 0.47993531823158264, "learning_rate": 8.178e-05, "loss": 0.093, "step": 4090 }, { "grad_norm": 0.571739912033081, "learning_rate": 8.198e-05, "loss": 0.0894, "step": 4100 }, { "grad_norm": 0.5187159180641174, "learning_rate": 8.218e-05, "loss": 0.08, "step": 4110 }, { "grad_norm": 0.7457250952720642, "learning_rate": 8.238000000000001e-05, "loss": 0.0922, "step": 4120 }, { "grad_norm": 0.4973583519458771, "learning_rate": 8.258e-05, "loss": 0.0895, "step": 4130 }, { "grad_norm": 0.5503352284431458, "learning_rate": 8.278e-05, "loss": 0.0929, "step": 4140 }, { "grad_norm": 0.6506848335266113, "learning_rate": 8.298000000000001e-05, "loss": 0.0901, "step": 4150 }, { "grad_norm": 0.4872720241546631, "learning_rate": 8.318e-05, "loss": 0.0777, "step": 4160 }, { "grad_norm": 0.5342832207679749, "learning_rate": 8.338e-05, "loss": 0.0803, "step": 4170 }, { "grad_norm": 0.6546953916549683, "learning_rate": 8.358e-05, "loss": 0.083, "step": 4180 }, { "grad_norm": 0.5896704196929932, "learning_rate": 8.378e-05, "loss": 0.0871, "step": 4190 }, { "grad_norm": 0.6374898552894592, "learning_rate": 8.398e-05, "loss": 0.0896, "step": 4200 }, { "grad_norm": 0.7480771541595459, "learning_rate": 8.418e-05, "loss": 0.0785, "step": 4210 }, { "grad_norm": 0.6379032731056213, "learning_rate": 8.438e-05, "loss": 0.0989, "step": 4220 }, { "grad_norm": 0.5488951206207275, "learning_rate": 8.458e-05, "loss": 0.0874, "step": 4230 }, { "grad_norm": 0.6795976161956787, "learning_rate": 8.478e-05, "loss": 0.0824, "step": 4240 }, { "grad_norm": 0.6537259221076965, "learning_rate": 8.498e-05, "loss": 0.0791, "step": 4250 }, { "grad_norm": 0.518762469291687, "learning_rate": 8.518000000000001e-05, "loss": 0.0833, "step": 4260 }, { "grad_norm": 0.8005845546722412, "learning_rate": 8.538e-05, "loss": 0.0893, "step": 4270 }, { "grad_norm": 0.5243597626686096, "learning_rate": 8.558e-05, "loss": 0.0813, "step": 4280 }, { "grad_norm": 0.5144310593605042, "learning_rate": 8.578000000000001e-05, "loss": 0.086, "step": 4290 }, { "grad_norm": 0.5731711983680725, "learning_rate": 8.598e-05, "loss": 0.0922, "step": 4300 }, { "grad_norm": 0.7044820785522461, "learning_rate": 8.618e-05, "loss": 0.0782, "step": 4310 }, { "grad_norm": 0.7028088569641113, "learning_rate": 8.638000000000001e-05, "loss": 0.0852, "step": 4320 }, { "grad_norm": 0.5933824181556702, "learning_rate": 8.658e-05, "loss": 0.0811, "step": 4330 }, { "grad_norm": 0.5878687500953674, "learning_rate": 8.678e-05, "loss": 0.093, "step": 4340 }, { "grad_norm": 0.5306347608566284, "learning_rate": 8.698000000000001e-05, "loss": 0.0829, "step": 4350 }, { "grad_norm": 0.5140297412872314, "learning_rate": 8.718e-05, "loss": 0.0815, "step": 4360 }, { "grad_norm": 0.7154054641723633, "learning_rate": 8.738000000000001e-05, "loss": 0.0782, "step": 4370 }, { "grad_norm": 0.6531661748886108, "learning_rate": 8.758000000000001e-05, "loss": 0.0827, "step": 4380 }, { "grad_norm": 0.7498372793197632, "learning_rate": 8.778e-05, "loss": 0.0871, "step": 4390 }, { "grad_norm": 0.47136253118515015, "learning_rate": 8.798000000000001e-05, "loss": 0.0841, "step": 4400 }, { "grad_norm": 0.4963436722755432, "learning_rate": 8.818000000000001e-05, "loss": 0.086, "step": 4410 }, { "grad_norm": 0.6920421719551086, "learning_rate": 8.838e-05, "loss": 0.0926, "step": 4420 }, { "grad_norm": 0.669862687587738, "learning_rate": 8.858000000000001e-05, "loss": 0.0823, "step": 4430 }, { "grad_norm": 0.6312661170959473, "learning_rate": 8.878000000000001e-05, "loss": 0.0847, "step": 4440 }, { "grad_norm": 0.5755856037139893, "learning_rate": 8.898e-05, "loss": 0.08, "step": 4450 }, { "grad_norm": 0.484380841255188, "learning_rate": 8.918000000000001e-05, "loss": 0.0819, "step": 4460 }, { "grad_norm": 0.6658583879470825, "learning_rate": 8.938e-05, "loss": 0.0981, "step": 4470 }, { "grad_norm": 0.6115947961807251, "learning_rate": 8.958e-05, "loss": 0.0885, "step": 4480 }, { "grad_norm": 0.7429367303848267, "learning_rate": 8.978000000000001e-05, "loss": 0.1027, "step": 4490 }, { "grad_norm": 0.5831069946289062, "learning_rate": 8.998e-05, "loss": 0.088, "step": 4500 }, { "grad_norm": 0.523047149181366, "learning_rate": 9.018000000000001e-05, "loss": 0.0896, "step": 4510 }, { "grad_norm": 0.49259570240974426, "learning_rate": 9.038000000000001e-05, "loss": 0.0849, "step": 4520 }, { "grad_norm": 0.450690895318985, "learning_rate": 9.058e-05, "loss": 0.0839, "step": 4530 }, { "grad_norm": 0.6901940107345581, "learning_rate": 9.078000000000001e-05, "loss": 0.0914, "step": 4540 }, { "grad_norm": 0.656805157661438, "learning_rate": 9.098000000000001e-05, "loss": 0.0833, "step": 4550 }, { "grad_norm": 0.5382756590843201, "learning_rate": 9.118e-05, "loss": 0.0801, "step": 4560 }, { "grad_norm": 0.4026874005794525, "learning_rate": 9.138e-05, "loss": 0.078, "step": 4570 }, { "grad_norm": 0.5597079992294312, "learning_rate": 9.158e-05, "loss": 0.0767, "step": 4580 }, { "grad_norm": 0.45457717776298523, "learning_rate": 9.178e-05, "loss": 0.0901, "step": 4590 }, { "grad_norm": 0.5089876651763916, "learning_rate": 9.198e-05, "loss": 0.0757, "step": 4600 }, { "grad_norm": 0.6574233174324036, "learning_rate": 9.218e-05, "loss": 0.0878, "step": 4610 }, { "grad_norm": 0.4635625183582306, "learning_rate": 9.238e-05, "loss": 0.0767, "step": 4620 }, { "grad_norm": 0.8811632394790649, "learning_rate": 9.258e-05, "loss": 0.0922, "step": 4630 }, { "grad_norm": 0.49447163939476013, "learning_rate": 9.278e-05, "loss": 0.0798, "step": 4640 }, { "grad_norm": 0.7198284864425659, "learning_rate": 9.298e-05, "loss": 0.0804, "step": 4650 }, { "grad_norm": 0.6965271234512329, "learning_rate": 9.318e-05, "loss": 0.1013, "step": 4660 }, { "grad_norm": 0.4042443037033081, "learning_rate": 9.338e-05, "loss": 0.0863, "step": 4670 }, { "grad_norm": 0.5067155361175537, "learning_rate": 9.358e-05, "loss": 0.0902, "step": 4680 }, { "grad_norm": 0.6044760346412659, "learning_rate": 9.378e-05, "loss": 0.0953, "step": 4690 }, { "grad_norm": 0.8051416873931885, "learning_rate": 9.398e-05, "loss": 0.0865, "step": 4700 }, { "grad_norm": 0.6468337774276733, "learning_rate": 9.418e-05, "loss": 0.0879, "step": 4710 }, { "grad_norm": 0.544281542301178, "learning_rate": 9.438e-05, "loss": 0.0887, "step": 4720 }, { "grad_norm": 0.7443307042121887, "learning_rate": 9.458e-05, "loss": 0.0836, "step": 4730 }, { "grad_norm": 0.5230416655540466, "learning_rate": 9.478e-05, "loss": 0.0882, "step": 4740 }, { "grad_norm": 0.6015952229499817, "learning_rate": 9.498e-05, "loss": 0.088, "step": 4750 }, { "grad_norm": 0.44677814841270447, "learning_rate": 9.518000000000001e-05, "loss": 0.0837, "step": 4760 }, { "grad_norm": 0.5237294435501099, "learning_rate": 9.538e-05, "loss": 0.084, "step": 4770 }, { "grad_norm": 0.6377003192901611, "learning_rate": 9.558e-05, "loss": 0.0808, "step": 4780 }, { "grad_norm": 0.6529236435890198, "learning_rate": 9.578000000000001e-05, "loss": 0.0868, "step": 4790 }, { "grad_norm": 0.6263142228126526, "learning_rate": 9.598e-05, "loss": 0.085, "step": 4800 }, { "grad_norm": 0.6236782670021057, "learning_rate": 9.618e-05, "loss": 0.0766, "step": 4810 }, { "grad_norm": 0.5892521739006042, "learning_rate": 9.638000000000001e-05, "loss": 0.0828, "step": 4820 }, { "grad_norm": 0.7297564148902893, "learning_rate": 9.658e-05, "loss": 0.0771, "step": 4830 }, { "grad_norm": 0.9400388598442078, "learning_rate": 9.678e-05, "loss": 0.0867, "step": 4840 }, { "grad_norm": 0.5793445706367493, "learning_rate": 9.698000000000001e-05, "loss": 0.0907, "step": 4850 }, { "grad_norm": 0.7277482151985168, "learning_rate": 9.718e-05, "loss": 0.0893, "step": 4860 }, { "grad_norm": 0.42602190375328064, "learning_rate": 9.738e-05, "loss": 0.0857, "step": 4870 }, { "grad_norm": 0.426835298538208, "learning_rate": 9.758000000000001e-05, "loss": 0.0748, "step": 4880 }, { "grad_norm": 0.4591825604438782, "learning_rate": 9.778e-05, "loss": 0.0775, "step": 4890 }, { "grad_norm": 0.5181707739830017, "learning_rate": 9.798000000000001e-05, "loss": 0.0863, "step": 4900 }, { "grad_norm": 0.5256882309913635, "learning_rate": 9.818000000000001e-05, "loss": 0.0955, "step": 4910 }, { "grad_norm": 0.6595622301101685, "learning_rate": 9.838e-05, "loss": 0.0942, "step": 4920 }, { "grad_norm": 0.5943392515182495, "learning_rate": 9.858000000000001e-05, "loss": 0.0841, "step": 4930 }, { "grad_norm": 0.6031759977340698, "learning_rate": 9.878e-05, "loss": 0.0827, "step": 4940 }, { "grad_norm": 0.5973065495491028, "learning_rate": 9.898e-05, "loss": 0.1035, "step": 4950 }, { "grad_norm": 0.578787624835968, "learning_rate": 9.918000000000001e-05, "loss": 0.0827, "step": 4960 }, { "grad_norm": 0.49584248661994934, "learning_rate": 9.938e-05, "loss": 0.0827, "step": 4970 }, { "grad_norm": 0.4904741644859314, "learning_rate": 9.958e-05, "loss": 0.0859, "step": 4980 }, { "grad_norm": 0.7292605638504028, "learning_rate": 9.978000000000001e-05, "loss": 0.0985, "step": 4990 }, { "grad_norm": 0.6313908696174622, "learning_rate": 9.998e-05, "loss": 0.098, "step": 5000 }, { "grad_norm": 0.5548058152198792, "learning_rate": 9.999999778549045e-05, "loss": 0.0862, "step": 5010 }, { "grad_norm": 0.6946904063224792, "learning_rate": 9.999999013039593e-05, "loss": 0.1015, "step": 5020 }, { "grad_norm": 0.8209462761878967, "learning_rate": 9.999997700737766e-05, "loss": 0.0899, "step": 5030 }, { "grad_norm": 0.6223562359809875, "learning_rate": 9.999995841643709e-05, "loss": 0.0858, "step": 5040 }, { "grad_norm": 0.5321465134620667, "learning_rate": 9.999993435757623e-05, "loss": 0.0855, "step": 5050 }, { "grad_norm": 0.6648877859115601, "learning_rate": 9.999990483079773e-05, "loss": 0.0882, "step": 5060 }, { "grad_norm": 0.4857560992240906, "learning_rate": 9.999986983610481e-05, "loss": 0.0882, "step": 5070 }, { "grad_norm": 0.47842395305633545, "learning_rate": 9.99998293735013e-05, "loss": 0.0813, "step": 5080 }, { "grad_norm": 0.7881121039390564, "learning_rate": 9.999978344299161e-05, "loss": 0.0999, "step": 5090 }, { "grad_norm": 0.5311048030853271, "learning_rate": 9.99997320445808e-05, "loss": 0.0892, "step": 5100 }, { "grad_norm": 0.5943753123283386, "learning_rate": 9.999967517827444e-05, "loss": 0.0861, "step": 5110 }, { "grad_norm": 0.5790807604789734, "learning_rate": 9.999961284407879e-05, "loss": 0.0917, "step": 5120 }, { "grad_norm": 0.610430121421814, "learning_rate": 9.999954504200067e-05, "loss": 0.086, "step": 5130 }, { "grad_norm": 0.7748014330863953, "learning_rate": 9.999947177204744e-05, "loss": 0.0856, "step": 5140 }, { "grad_norm": 0.4909389913082123, "learning_rate": 9.999939303422718e-05, "loss": 0.0913, "step": 5150 }, { "grad_norm": 0.5167636871337891, "learning_rate": 9.999930882854847e-05, "loss": 0.0774, "step": 5160 }, { "grad_norm": 0.7121630311012268, "learning_rate": 9.999921915502051e-05, "loss": 0.0865, "step": 5170 }, { "grad_norm": 0.5098788142204285, "learning_rate": 9.99991240136531e-05, "loss": 0.0776, "step": 5180 }, { "grad_norm": 0.6412806510925293, "learning_rate": 9.999902340445668e-05, "loss": 0.0851, "step": 5190 }, { "grad_norm": 0.5652850270271301, "learning_rate": 9.999891732744224e-05, "loss": 0.0763, "step": 5200 }, { "grad_norm": 0.5045563578605652, "learning_rate": 9.999880578262135e-05, "loss": 0.0833, "step": 5210 }, { "grad_norm": 0.7559992671012878, "learning_rate": 9.999868877000624e-05, "loss": 0.0846, "step": 5220 }, { "grad_norm": 0.5905073881149292, "learning_rate": 9.99985662896097e-05, "loss": 0.0838, "step": 5230 }, { "grad_norm": 0.40594547986984253, "learning_rate": 9.999843834144513e-05, "loss": 0.0846, "step": 5240 }, { "grad_norm": 0.6155417561531067, "learning_rate": 9.99983049255265e-05, "loss": 0.0928, "step": 5250 }, { "grad_norm": 0.48051297664642334, "learning_rate": 9.999816604186843e-05, "loss": 0.0941, "step": 5260 }, { "grad_norm": 0.5510441660881042, "learning_rate": 9.999802169048609e-05, "loss": 0.0894, "step": 5270 }, { "grad_norm": 0.5540792346000671, "learning_rate": 9.999787187139527e-05, "loss": 0.0792, "step": 5280 }, { "grad_norm": 0.6002240777015686, "learning_rate": 9.999771658461234e-05, "loss": 0.088, "step": 5290 }, { "grad_norm": 0.6350477337837219, "learning_rate": 9.999755583015431e-05, "loss": 0.0894, "step": 5300 }, { "grad_norm": 0.5614681839942932, "learning_rate": 9.999738960803874e-05, "loss": 0.0813, "step": 5310 }, { "grad_norm": 0.5331659317016602, "learning_rate": 9.99972179182838e-05, "loss": 0.0801, "step": 5320 }, { "grad_norm": 0.5910619497299194, "learning_rate": 9.99970407609083e-05, "loss": 0.0812, "step": 5330 }, { "grad_norm": 0.5775790810585022, "learning_rate": 9.999685813593159e-05, "loss": 0.082, "step": 5340 }, { "grad_norm": 0.5460905432701111, "learning_rate": 9.999667004337362e-05, "loss": 0.0802, "step": 5350 }, { "grad_norm": 0.46129581332206726, "learning_rate": 9.9996476483255e-05, "loss": 0.0826, "step": 5360 }, { "grad_norm": 0.4847520887851715, "learning_rate": 9.999627745559688e-05, "loss": 0.0759, "step": 5370 }, { "grad_norm": 0.41289111971855164, "learning_rate": 9.999607296042101e-05, "loss": 0.0819, "step": 5380 }, { "grad_norm": 0.5231015086174011, "learning_rate": 9.99958629977498e-05, "loss": 0.0846, "step": 5390 }, { "grad_norm": 0.5782631635665894, "learning_rate": 9.999564756760615e-05, "loss": 0.073, "step": 5400 }, { "grad_norm": 0.570857584476471, "learning_rate": 9.999542667001366e-05, "loss": 0.0846, "step": 5410 }, { "grad_norm": 0.5019176006317139, "learning_rate": 9.999520030499647e-05, "loss": 0.0858, "step": 5420 }, { "grad_norm": 0.5333186984062195, "learning_rate": 9.999496847257936e-05, "loss": 0.085, "step": 5430 }, { "grad_norm": 0.5639175772666931, "learning_rate": 9.999473117278764e-05, "loss": 0.0863, "step": 5440 }, { "grad_norm": 0.5260929465293884, "learning_rate": 9.999448840564731e-05, "loss": 0.0886, "step": 5450 }, { "grad_norm": 0.42551615834236145, "learning_rate": 9.999424017118488e-05, "loss": 0.0886, "step": 5460 }, { "grad_norm": 0.5432234406471252, "learning_rate": 9.999398646942751e-05, "loss": 0.0815, "step": 5470 }, { "grad_norm": 0.518031895160675, "learning_rate": 9.999372730040296e-05, "loss": 0.0841, "step": 5480 }, { "grad_norm": 0.43700599670410156, "learning_rate": 9.999346266413953e-05, "loss": 0.0864, "step": 5490 }, { "grad_norm": 0.4734118580818176, "learning_rate": 9.99931925606662e-05, "loss": 0.0885, "step": 5500 }, { "grad_norm": 0.6726197004318237, "learning_rate": 9.99929169900125e-05, "loss": 0.082, "step": 5510 }, { "grad_norm": 0.8004509210586548, "learning_rate": 9.999263595220855e-05, "loss": 0.0838, "step": 5520 }, { "grad_norm": 0.6635643243789673, "learning_rate": 9.99923494472851e-05, "loss": 0.0894, "step": 5530 }, { "grad_norm": 0.6901882886886597, "learning_rate": 9.999205747527348e-05, "loss": 0.0822, "step": 5540 }, { "grad_norm": 0.45049038529396057, "learning_rate": 9.999176003620561e-05, "loss": 0.0864, "step": 5550 }, { "grad_norm": 0.49200505018234253, "learning_rate": 9.999145713011405e-05, "loss": 0.0899, "step": 5560 }, { "grad_norm": 0.5437648892402649, "learning_rate": 9.999114875703186e-05, "loss": 0.0738, "step": 5570 }, { "grad_norm": 0.5352596044540405, "learning_rate": 9.999083491699281e-05, "loss": 0.0873, "step": 5580 }, { "grad_norm": 0.4270751476287842, "learning_rate": 9.999051561003123e-05, "loss": 0.0922, "step": 5590 }, { "grad_norm": 0.6928128004074097, "learning_rate": 9.999019083618202e-05, "loss": 0.0884, "step": 5600 }, { "grad_norm": 0.6088253855705261, "learning_rate": 9.99898605954807e-05, "loss": 0.0852, "step": 5610 }, { "grad_norm": 0.5970381498336792, "learning_rate": 9.998952488796338e-05, "loss": 0.0879, "step": 5620 }, { "grad_norm": 0.6546837687492371, "learning_rate": 9.998918371366676e-05, "loss": 0.0824, "step": 5630 }, { "grad_norm": 0.637392520904541, "learning_rate": 9.99888370726282e-05, "loss": 0.0801, "step": 5640 }, { "grad_norm": 0.5175595879554749, "learning_rate": 9.998848496488556e-05, "loss": 0.0805, "step": 5650 }, { "grad_norm": 0.6038771271705627, "learning_rate": 9.998812739047736e-05, "loss": 0.0806, "step": 5660 }, { "grad_norm": 0.5419846773147583, "learning_rate": 9.99877643494427e-05, "loss": 0.0896, "step": 5670 }, { "grad_norm": 0.5076566934585571, "learning_rate": 9.998739584182128e-05, "loss": 0.0902, "step": 5680 }, { "grad_norm": 0.5021442770957947, "learning_rate": 9.998702186765342e-05, "loss": 0.0937, "step": 5690 }, { "grad_norm": 0.5743950009346008, "learning_rate": 9.998664242698e-05, "loss": 0.0801, "step": 5700 }, { "grad_norm": 0.458979070186615, "learning_rate": 9.998625751984251e-05, "loss": 0.0851, "step": 5710 }, { "grad_norm": 0.5086292624473572, "learning_rate": 9.998586714628307e-05, "loss": 0.0781, "step": 5720 }, { "grad_norm": 0.5746293663978577, "learning_rate": 9.998547130634432e-05, "loss": 0.0861, "step": 5730 }, { "grad_norm": 0.46278929710388184, "learning_rate": 9.99850700000696e-05, "loss": 0.0804, "step": 5740 }, { "grad_norm": 0.8056983351707458, "learning_rate": 9.998466322750278e-05, "loss": 0.0977, "step": 5750 }, { "grad_norm": 0.515885055065155, "learning_rate": 9.998425098868834e-05, "loss": 0.087, "step": 5760 }, { "grad_norm": 0.5330471396446228, "learning_rate": 9.998383328367136e-05, "loss": 0.0851, "step": 5770 }, { "grad_norm": 0.49575456976890564, "learning_rate": 9.99834101124975e-05, "loss": 0.0831, "step": 5780 }, { "grad_norm": 0.48014143109321594, "learning_rate": 9.998298147521309e-05, "loss": 0.0837, "step": 5790 }, { "grad_norm": 0.6617678999900818, "learning_rate": 9.998254737186496e-05, "loss": 0.0856, "step": 5800 }, { "grad_norm": 0.44076454639434814, "learning_rate": 9.99821078025006e-05, "loss": 0.0753, "step": 5810 }, { "grad_norm": 0.5433362722396851, "learning_rate": 9.998166276716807e-05, "loss": 0.0959, "step": 5820 }, { "grad_norm": 0.4911688566207886, "learning_rate": 9.998121226591606e-05, "loss": 0.0868, "step": 5830 }, { "grad_norm": 0.4067508280277252, "learning_rate": 9.998075629879382e-05, "loss": 0.0841, "step": 5840 }, { "grad_norm": 0.6109286546707153, "learning_rate": 9.99802948658512e-05, "loss": 0.0885, "step": 5850 }, { "grad_norm": 0.48330920934677124, "learning_rate": 9.99798279671387e-05, "loss": 0.0864, "step": 5860 }, { "grad_norm": 0.5586318373680115, "learning_rate": 9.997935560270734e-05, "loss": 0.0972, "step": 5870 }, { "grad_norm": 0.733757734298706, "learning_rate": 9.997887777260879e-05, "loss": 0.0933, "step": 5880 }, { "grad_norm": 0.49745458364486694, "learning_rate": 9.997839447689532e-05, "loss": 0.0906, "step": 5890 }, { "grad_norm": 0.5853675603866577, "learning_rate": 9.997790571561978e-05, "loss": 0.0874, "step": 5900 }, { "grad_norm": 0.63050377368927, "learning_rate": 9.99774114888356e-05, "loss": 0.0837, "step": 5910 }, { "grad_norm": 0.5059924721717834, "learning_rate": 9.997691179659684e-05, "loss": 0.0984, "step": 5920 }, { "grad_norm": 0.5456470251083374, "learning_rate": 9.997640663895815e-05, "loss": 0.0818, "step": 5930 }, { "grad_norm": 0.4589748680591583, "learning_rate": 9.997589601597477e-05, "loss": 0.0796, "step": 5940 }, { "grad_norm": 0.5481772422790527, "learning_rate": 9.997537992770252e-05, "loss": 0.0842, "step": 5950 }, { "grad_norm": 0.3683588206768036, "learning_rate": 9.997485837419788e-05, "loss": 0.0894, "step": 5960 }, { "grad_norm": 0.5539022088050842, "learning_rate": 9.997433135551786e-05, "loss": 0.0855, "step": 5970 }, { "grad_norm": 0.7320805191993713, "learning_rate": 9.997379887172009e-05, "loss": 0.0907, "step": 5980 }, { "grad_norm": 0.4183509051799774, "learning_rate": 9.997326092286281e-05, "loss": 0.0836, "step": 5990 }, { "grad_norm": 0.5542795658111572, "learning_rate": 9.997271750900486e-05, "loss": 0.0814, "step": 6000 }, { "grad_norm": 0.5426478385925293, "learning_rate": 9.997216863020565e-05, "loss": 0.0874, "step": 6010 }, { "grad_norm": 0.5256152153015137, "learning_rate": 9.99716142865252e-05, "loss": 0.0801, "step": 6020 }, { "grad_norm": 0.5700957775115967, "learning_rate": 9.997105447802415e-05, "loss": 0.0829, "step": 6030 }, { "grad_norm": 1.4146885871887207, "learning_rate": 9.997048920476373e-05, "loss": 0.0871, "step": 6040 }, { "grad_norm": 0.45576998591423035, "learning_rate": 9.996991846680572e-05, "loss": 0.0791, "step": 6050 }, { "grad_norm": 0.5058749914169312, "learning_rate": 9.996934226421257e-05, "loss": 0.0744, "step": 6060 }, { "grad_norm": 0.547542154788971, "learning_rate": 9.996876059704726e-05, "loss": 0.0863, "step": 6070 }, { "grad_norm": 0.49035775661468506, "learning_rate": 9.996817346537343e-05, "loss": 0.0806, "step": 6080 }, { "grad_norm": 0.40619009733200073, "learning_rate": 9.996758086925526e-05, "loss": 0.0843, "step": 6090 }, { "grad_norm": 0.5376760363578796, "learning_rate": 9.996698280875759e-05, "loss": 0.073, "step": 6100 }, { "grad_norm": 0.566310703754425, "learning_rate": 9.99663792839458e-05, "loss": 0.0762, "step": 6110 }, { "grad_norm": 0.6435236930847168, "learning_rate": 9.99657702948859e-05, "loss": 0.0778, "step": 6120 }, { "grad_norm": 0.7869743704795837, "learning_rate": 9.996515584164448e-05, "loss": 0.0824, "step": 6130 }, { "grad_norm": 0.6207780838012695, "learning_rate": 9.996453592428873e-05, "loss": 0.0859, "step": 6140 }, { "grad_norm": 0.5571346282958984, "learning_rate": 9.996391054288646e-05, "loss": 0.0863, "step": 6150 }, { "grad_norm": 0.6270369291305542, "learning_rate": 9.996327969750605e-05, "loss": 0.0811, "step": 6160 }, { "grad_norm": 0.6243158578872681, "learning_rate": 9.996264338821649e-05, "loss": 0.0862, "step": 6170 }, { "grad_norm": 0.4829599857330322, "learning_rate": 9.996200161508735e-05, "loss": 0.0858, "step": 6180 }, { "grad_norm": 0.496533066034317, "learning_rate": 9.996135437818885e-05, "loss": 0.0839, "step": 6190 }, { "grad_norm": 0.508467435836792, "learning_rate": 9.996070167759175e-05, "loss": 0.0867, "step": 6200 }, { "grad_norm": 0.786100447177887, "learning_rate": 9.996004351336743e-05, "loss": 0.0918, "step": 6210 }, { "grad_norm": 0.7000240087509155, "learning_rate": 9.995937988558785e-05, "loss": 0.081, "step": 6220 }, { "grad_norm": 0.6008635759353638, "learning_rate": 9.995871079432561e-05, "loss": 0.0847, "step": 6230 }, { "grad_norm": 0.4137488603591919, "learning_rate": 9.995803623965389e-05, "loss": 0.0749, "step": 6240 }, { "grad_norm": 0.44527262449264526, "learning_rate": 9.995735622164641e-05, "loss": 0.0917, "step": 6250 }, { "grad_norm": 0.6068211793899536, "learning_rate": 9.995667074037758e-05, "loss": 0.0857, "step": 6260 }, { "grad_norm": 0.4996461272239685, "learning_rate": 9.995597979592232e-05, "loss": 0.0755, "step": 6270 }, { "grad_norm": 0.6006132960319519, "learning_rate": 9.995528338835625e-05, "loss": 0.0771, "step": 6280 }, { "grad_norm": 0.5855443477630615, "learning_rate": 9.995458151775547e-05, "loss": 0.0882, "step": 6290 }, { "grad_norm": 0.7313621640205383, "learning_rate": 9.995387418419677e-05, "loss": 0.0821, "step": 6300 }, { "grad_norm": 0.4381471872329712, "learning_rate": 9.99531613877575e-05, "loss": 0.0823, "step": 6310 }, { "grad_norm": 0.4177899658679962, "learning_rate": 9.995244312851559e-05, "loss": 0.073, "step": 6320 }, { "grad_norm": 0.47939059138298035, "learning_rate": 9.995171940654961e-05, "loss": 0.0785, "step": 6330 }, { "grad_norm": 0.362014502286911, "learning_rate": 9.995099022193871e-05, "loss": 0.0798, "step": 6340 }, { "grad_norm": 0.47691744565963745, "learning_rate": 9.995025557476261e-05, "loss": 0.0832, "step": 6350 }, { "grad_norm": 0.40387120842933655, "learning_rate": 9.994951546510165e-05, "loss": 0.0774, "step": 6360 }, { "grad_norm": 0.5232626795768738, "learning_rate": 9.994876989303679e-05, "loss": 0.0856, "step": 6370 }, { "grad_norm": 0.542037308216095, "learning_rate": 9.994801885864955e-05, "loss": 0.0871, "step": 6380 }, { "grad_norm": 0.5285590887069702, "learning_rate": 9.994726236202205e-05, "loss": 0.0795, "step": 6390 }, { "grad_norm": 0.5527752041816711, "learning_rate": 9.994650040323704e-05, "loss": 0.0845, "step": 6400 }, { "grad_norm": 0.5194333791732788, "learning_rate": 9.994573298237784e-05, "loss": 0.0765, "step": 6410 }, { "grad_norm": 0.42470282316207886, "learning_rate": 9.994496009952837e-05, "loss": 0.073, "step": 6420 }, { "grad_norm": 0.7283902764320374, "learning_rate": 9.994418175477316e-05, "loss": 0.0827, "step": 6430 }, { "grad_norm": 0.5490550398826599, "learning_rate": 9.994339794819733e-05, "loss": 0.0812, "step": 6440 }, { "grad_norm": 0.6404784917831421, "learning_rate": 9.994260867988658e-05, "loss": 0.0947, "step": 6450 }, { "grad_norm": 0.5370270609855652, "learning_rate": 9.994181394992723e-05, "loss": 0.073, "step": 6460 }, { "grad_norm": 0.46347957849502563, "learning_rate": 9.994101375840618e-05, "loss": 0.0786, "step": 6470 }, { "grad_norm": 0.5189306735992432, "learning_rate": 9.994020810541098e-05, "loss": 0.0805, "step": 6480 }, { "grad_norm": 0.53156578540802, "learning_rate": 9.99393969910297e-05, "loss": 0.0788, "step": 6490 }, { "grad_norm": 0.5813888311386108, "learning_rate": 9.993858041535104e-05, "loss": 0.0861, "step": 6500 }, { "grad_norm": 0.4281483292579651, "learning_rate": 9.99377583784643e-05, "loss": 0.0712, "step": 6510 }, { "grad_norm": 0.5015400052070618, "learning_rate": 9.993693088045939e-05, "loss": 0.0857, "step": 6520 }, { "grad_norm": 0.520889163017273, "learning_rate": 9.99360979214268e-05, "loss": 0.0764, "step": 6530 }, { "grad_norm": 0.5332992672920227, "learning_rate": 9.99352595014576e-05, "loss": 0.0762, "step": 6540 }, { "grad_norm": 0.5416436195373535, "learning_rate": 9.993441562064354e-05, "loss": 0.0755, "step": 6550 }, { "grad_norm": 0.45187926292419434, "learning_rate": 9.993356627907685e-05, "loss": 0.074, "step": 6560 }, { "grad_norm": 0.48361867666244507, "learning_rate": 9.99327114768504e-05, "loss": 0.0881, "step": 6570 }, { "grad_norm": 0.5985421538352966, "learning_rate": 9.99318512140577e-05, "loss": 0.0815, "step": 6580 }, { "grad_norm": 0.6392854452133179, "learning_rate": 9.993098549079284e-05, "loss": 0.0764, "step": 6590 }, { "grad_norm": 0.42814749479293823, "learning_rate": 9.993011430715047e-05, "loss": 0.0833, "step": 6600 }, { "grad_norm": 0.48547685146331787, "learning_rate": 9.992923766322586e-05, "loss": 0.0738, "step": 6610 }, { "grad_norm": 0.4098338484764099, "learning_rate": 9.99283555591149e-05, "loss": 0.0815, "step": 6620 }, { "grad_norm": 0.5467835068702698, "learning_rate": 9.992746799491404e-05, "loss": 0.0814, "step": 6630 }, { "grad_norm": 0.41236698627471924, "learning_rate": 9.992657497072033e-05, "loss": 0.0887, "step": 6640 }, { "grad_norm": 0.41531017422676086, "learning_rate": 9.992567648663147e-05, "loss": 0.0838, "step": 6650 }, { "grad_norm": 0.6690495610237122, "learning_rate": 9.992477254274568e-05, "loss": 0.0833, "step": 6660 }, { "grad_norm": 0.4925600588321686, "learning_rate": 9.992386313916183e-05, "loss": 0.0855, "step": 6670 }, { "grad_norm": 0.4297494888305664, "learning_rate": 9.992294827597934e-05, "loss": 0.0849, "step": 6680 }, { "grad_norm": 0.6542667150497437, "learning_rate": 9.992202795329831e-05, "loss": 0.0841, "step": 6690 }, { "grad_norm": 0.5215906500816345, "learning_rate": 9.992110217121936e-05, "loss": 0.077, "step": 6700 }, { "grad_norm": 0.4419653117656708, "learning_rate": 9.992017092984372e-05, "loss": 0.0789, "step": 6710 }, { "grad_norm": 0.5207825303077698, "learning_rate": 9.991923422927326e-05, "loss": 0.0868, "step": 6720 }, { "grad_norm": 0.5645508170127869, "learning_rate": 9.991829206961037e-05, "loss": 0.0719, "step": 6730 }, { "grad_norm": 0.3228447139263153, "learning_rate": 9.991734445095813e-05, "loss": 0.0744, "step": 6740 }, { "grad_norm": 0.422719269990921, "learning_rate": 9.991639137342015e-05, "loss": 0.0781, "step": 6750 }, { "grad_norm": 0.48791489005088806, "learning_rate": 9.991543283710064e-05, "loss": 0.0743, "step": 6760 }, { "grad_norm": 0.7438668608665466, "learning_rate": 9.991446884210445e-05, "loss": 0.0861, "step": 6770 }, { "grad_norm": 0.38218390941619873, "learning_rate": 9.9913499388537e-05, "loss": 0.0771, "step": 6780 }, { "grad_norm": 0.5097927451133728, "learning_rate": 9.99125244765043e-05, "loss": 0.088, "step": 6790 }, { "grad_norm": 0.49567893147468567, "learning_rate": 9.991154410611296e-05, "loss": 0.0845, "step": 6800 }, { "grad_norm": 0.5316175222396851, "learning_rate": 9.99105582774702e-05, "loss": 0.0791, "step": 6810 }, { "grad_norm": 0.6751384139060974, "learning_rate": 9.990956699068384e-05, "loss": 0.0972, "step": 6820 }, { "grad_norm": 0.49787798523902893, "learning_rate": 9.990857024586224e-05, "loss": 0.0814, "step": 6830 }, { "grad_norm": 0.4652940332889557, "learning_rate": 9.990756804311446e-05, "loss": 0.0836, "step": 6840 }, { "grad_norm": 0.5234807729721069, "learning_rate": 9.990656038255006e-05, "loss": 0.0741, "step": 6850 }, { "grad_norm": 0.42974573373794556, "learning_rate": 9.990554726427926e-05, "loss": 0.075, "step": 6860 }, { "grad_norm": 0.4760657250881195, "learning_rate": 9.990452868841284e-05, "loss": 0.0752, "step": 6870 }, { "grad_norm": 0.47619912028312683, "learning_rate": 9.99035046550622e-05, "loss": 0.0775, "step": 6880 }, { "grad_norm": 0.4681600332260132, "learning_rate": 9.99024751643393e-05, "loss": 0.0681, "step": 6890 }, { "grad_norm": 0.4939647912979126, "learning_rate": 9.990144021635677e-05, "loss": 0.0743, "step": 6900 }, { "grad_norm": 0.6140620112419128, "learning_rate": 9.990039981122775e-05, "loss": 0.0752, "step": 6910 }, { "grad_norm": 0.5511165857315063, "learning_rate": 9.989935394906602e-05, "loss": 0.0851, "step": 6920 }, { "grad_norm": 0.5245757699012756, "learning_rate": 9.989830262998598e-05, "loss": 0.0851, "step": 6930 }, { "grad_norm": 0.5324094891548157, "learning_rate": 9.989724585410259e-05, "loss": 0.0844, "step": 6940 }, { "grad_norm": 0.39931151270866394, "learning_rate": 9.989618362153139e-05, "loss": 0.0764, "step": 6950 }, { "grad_norm": 0.577885627746582, "learning_rate": 9.989511593238859e-05, "loss": 0.077, "step": 6960 }, { "grad_norm": 0.4455796480178833, "learning_rate": 9.98940427867909e-05, "loss": 0.0749, "step": 6970 }, { "grad_norm": 0.3729593753814697, "learning_rate": 9.989296418485573e-05, "loss": 0.069, "step": 6980 }, { "grad_norm": 0.4280499517917633, "learning_rate": 9.989188012670101e-05, "loss": 0.0764, "step": 6990 }, { "grad_norm": 0.39667460322380066, "learning_rate": 9.989079061244528e-05, "loss": 0.0671, "step": 7000 }, { "grad_norm": 0.47458603978157043, "learning_rate": 9.988969564220769e-05, "loss": 0.0703, "step": 7010 }, { "grad_norm": 0.4228031635284424, "learning_rate": 9.988859521610801e-05, "loss": 0.0825, "step": 7020 }, { "grad_norm": 0.6685456037521362, "learning_rate": 9.988748933426656e-05, "loss": 0.0737, "step": 7030 }, { "grad_norm": 0.6172575950622559, "learning_rate": 9.988637799680428e-05, "loss": 0.084, "step": 7040 }, { "grad_norm": 0.3835444450378418, "learning_rate": 9.98852612038427e-05, "loss": 0.0739, "step": 7050 }, { "grad_norm": 0.5088086128234863, "learning_rate": 9.988413895550397e-05, "loss": 0.0812, "step": 7060 }, { "grad_norm": 0.42396047711372375, "learning_rate": 9.98830112519108e-05, "loss": 0.0809, "step": 7070 }, { "grad_norm": 0.45689424872398376, "learning_rate": 9.98818780931865e-05, "loss": 0.0663, "step": 7080 }, { "grad_norm": 0.4011598825454712, "learning_rate": 9.988073947945502e-05, "loss": 0.0775, "step": 7090 }, { "grad_norm": 0.5146798491477966, "learning_rate": 9.987959541084087e-05, "loss": 0.071, "step": 7100 }, { "grad_norm": 0.3506007194519043, "learning_rate": 9.987844588746915e-05, "loss": 0.0692, "step": 7110 }, { "grad_norm": 0.608523964881897, "learning_rate": 9.987729090946558e-05, "loss": 0.0814, "step": 7120 }, { "grad_norm": 0.3821358382701874, "learning_rate": 9.987613047695647e-05, "loss": 0.0758, "step": 7130 }, { "grad_norm": 0.4911912977695465, "learning_rate": 9.987496459006871e-05, "loss": 0.0895, "step": 7140 }, { "grad_norm": 0.6094714403152466, "learning_rate": 9.987379324892982e-05, "loss": 0.0769, "step": 7150 }, { "grad_norm": 0.5425605177879333, "learning_rate": 9.987261645366788e-05, "loss": 0.0781, "step": 7160 }, { "grad_norm": 0.5450642704963684, "learning_rate": 9.987143420441158e-05, "loss": 0.0736, "step": 7170 }, { "grad_norm": 0.4851204752922058, "learning_rate": 9.987024650129022e-05, "loss": 0.0718, "step": 7180 }, { "grad_norm": 0.37704119086265564, "learning_rate": 9.986905334443368e-05, "loss": 0.0711, "step": 7190 }, { "grad_norm": 0.480328232049942, "learning_rate": 9.986785473397245e-05, "loss": 0.0755, "step": 7200 }, { "grad_norm": 0.4144367575645447, "learning_rate": 9.98666506700376e-05, "loss": 0.0748, "step": 7210 }, { "grad_norm": 0.5680968165397644, "learning_rate": 9.986544115276081e-05, "loss": 0.0849, "step": 7220 }, { "grad_norm": 0.4509221017360687, "learning_rate": 9.986422618227433e-05, "loss": 0.0855, "step": 7230 }, { "grad_norm": 0.47118547558784485, "learning_rate": 9.986300575871106e-05, "loss": 0.0852, "step": 7240 }, { "grad_norm": 0.5161157846450806, "learning_rate": 9.986177988220444e-05, "loss": 0.0746, "step": 7250 }, { "grad_norm": 0.44233354926109314, "learning_rate": 9.986054855288856e-05, "loss": 0.0887, "step": 7260 }, { "grad_norm": 0.4465921223163605, "learning_rate": 9.985931177089802e-05, "loss": 0.0768, "step": 7270 }, { "grad_norm": 0.4770476222038269, "learning_rate": 9.985806953636814e-05, "loss": 0.0711, "step": 7280 }, { "grad_norm": 0.6497151255607605, "learning_rate": 9.985682184943471e-05, "loss": 0.0829, "step": 7290 }, { "grad_norm": 0.4587714374065399, "learning_rate": 9.98555687102342e-05, "loss": 0.0648, "step": 7300 }, { "grad_norm": 0.6366738677024841, "learning_rate": 9.985431011890367e-05, "loss": 0.0743, "step": 7310 }, { "grad_norm": 0.5507394075393677, "learning_rate": 9.985304607558075e-05, "loss": 0.0678, "step": 7320 }, { "grad_norm": 0.4284402132034302, "learning_rate": 9.985177658040364e-05, "loss": 0.0692, "step": 7330 }, { "grad_norm": 0.46756696701049805, "learning_rate": 9.985050163351119e-05, "loss": 0.0676, "step": 7340 }, { "grad_norm": 0.5105076432228088, "learning_rate": 9.984922123504286e-05, "loss": 0.0739, "step": 7350 }, { "grad_norm": 0.40787458419799805, "learning_rate": 9.984793538513862e-05, "loss": 0.0667, "step": 7360 }, { "grad_norm": 0.35816577076911926, "learning_rate": 9.984664408393912e-05, "loss": 0.083, "step": 7370 }, { "grad_norm": 0.6029694676399231, "learning_rate": 9.984534733158556e-05, "loss": 0.0799, "step": 7380 }, { "grad_norm": 0.5106503963470459, "learning_rate": 9.984404512821977e-05, "loss": 0.0763, "step": 7390 }, { "grad_norm": 0.5725816488265991, "learning_rate": 9.984273747398411e-05, "loss": 0.0714, "step": 7400 }, { "grad_norm": 0.6656414270401001, "learning_rate": 9.984142436902165e-05, "loss": 0.0764, "step": 7410 }, { "grad_norm": 0.5863409042358398, "learning_rate": 9.984010581347596e-05, "loss": 0.0879, "step": 7420 }, { "grad_norm": 0.5056666135787964, "learning_rate": 9.983878180749121e-05, "loss": 0.0724, "step": 7430 }, { "grad_norm": 0.42394503951072693, "learning_rate": 9.983745235121222e-05, "loss": 0.0676, "step": 7440 }, { "grad_norm": 0.5748369693756104, "learning_rate": 9.983611744478438e-05, "loss": 0.0765, "step": 7450 }, { "grad_norm": 0.48565933108329773, "learning_rate": 9.983477708835365e-05, "loss": 0.0751, "step": 7460 }, { "grad_norm": 0.5332090854644775, "learning_rate": 9.983343128206664e-05, "loss": 0.0839, "step": 7470 }, { "grad_norm": 0.5681036710739136, "learning_rate": 9.983208002607049e-05, "loss": 0.0714, "step": 7480 }, { "grad_norm": 0.45702189207077026, "learning_rate": 9.9830723320513e-05, "loss": 0.0744, "step": 7490 }, { "grad_norm": 0.4578714966773987, "learning_rate": 9.982936116554254e-05, "loss": 0.0761, "step": 7500 }, { "grad_norm": 0.4173124432563782, "learning_rate": 9.982799356130803e-05, "loss": 0.0776, "step": 7510 }, { "grad_norm": 0.44988399744033813, "learning_rate": 9.982662050795908e-05, "loss": 0.0837, "step": 7520 }, { "grad_norm": 0.42955338954925537, "learning_rate": 9.982524200564583e-05, "loss": 0.0729, "step": 7530 }, { "grad_norm": 0.4394908547401428, "learning_rate": 9.982385805451901e-05, "loss": 0.0738, "step": 7540 }, { "grad_norm": 0.38923269510269165, "learning_rate": 9.982246865472998e-05, "loss": 0.0763, "step": 7550 }, { "grad_norm": 0.5336036086082458, "learning_rate": 9.982107380643069e-05, "loss": 0.0722, "step": 7560 }, { "grad_norm": 0.5792656540870667, "learning_rate": 9.981967350977368e-05, "loss": 0.0782, "step": 7570 }, { "grad_norm": 0.47578561305999756, "learning_rate": 9.981826776491208e-05, "loss": 0.083, "step": 7580 }, { "grad_norm": 0.43515509366989136, "learning_rate": 9.98168565719996e-05, "loss": 0.0865, "step": 7590 }, { "grad_norm": 0.5296784043312073, "learning_rate": 9.98154399311906e-05, "loss": 0.0697, "step": 7600 }, { "grad_norm": 0.4669059216976166, "learning_rate": 9.981401784263997e-05, "loss": 0.0785, "step": 7610 }, { "grad_norm": 0.4169544279575348, "learning_rate": 9.981259030650326e-05, "loss": 0.0789, "step": 7620 }, { "grad_norm": 0.4685817360877991, "learning_rate": 9.981115732293655e-05, "loss": 0.0739, "step": 7630 }, { "grad_norm": 0.49235087633132935, "learning_rate": 9.980971889209659e-05, "loss": 0.0863, "step": 7640 }, { "grad_norm": 0.5085216760635376, "learning_rate": 9.980827501414064e-05, "loss": 0.068, "step": 7650 }, { "grad_norm": 0.6615095734596252, "learning_rate": 9.980682568922663e-05, "loss": 0.0743, "step": 7660 }, { "grad_norm": 0.4495551884174347, "learning_rate": 9.980537091751304e-05, "loss": 0.0754, "step": 7670 }, { "grad_norm": 0.5344926714897156, "learning_rate": 9.980391069915897e-05, "loss": 0.0788, "step": 7680 }, { "grad_norm": 0.43705588579177856, "learning_rate": 9.98024450343241e-05, "loss": 0.0743, "step": 7690 }, { "grad_norm": 0.38209792971611023, "learning_rate": 9.980097392316872e-05, "loss": 0.0772, "step": 7700 }, { "grad_norm": 0.42798930406570435, "learning_rate": 9.97994973658537e-05, "loss": 0.0719, "step": 7710 }, { "grad_norm": 0.6041901111602783, "learning_rate": 9.979801536254054e-05, "loss": 0.0707, "step": 7720 }, { "grad_norm": 0.5507277250289917, "learning_rate": 9.979652791339127e-05, "loss": 0.0718, "step": 7730 }, { "grad_norm": 0.4180663526058197, "learning_rate": 9.97950350185686e-05, "loss": 0.0661, "step": 7740 }, { "grad_norm": 0.5728918313980103, "learning_rate": 9.979353667823574e-05, "loss": 0.073, "step": 7750 }, { "grad_norm": 0.672258198261261, "learning_rate": 9.979203289255658e-05, "loss": 0.0699, "step": 7760 }, { "grad_norm": 0.3726671040058136, "learning_rate": 9.979052366169557e-05, "loss": 0.0748, "step": 7770 }, { "grad_norm": 0.4946167767047882, "learning_rate": 9.978900898581775e-05, "loss": 0.076, "step": 7780 }, { "grad_norm": 0.33220407366752625, "learning_rate": 9.978748886508875e-05, "loss": 0.0755, "step": 7790 }, { "grad_norm": 0.4006000757217407, "learning_rate": 9.978596329967484e-05, "loss": 0.0821, "step": 7800 }, { "grad_norm": 0.41570863127708435, "learning_rate": 9.978443228974284e-05, "loss": 0.0725, "step": 7810 }, { "grad_norm": 0.29796338081359863, "learning_rate": 9.978289583546015e-05, "loss": 0.0656, "step": 7820 }, { "grad_norm": 0.4300006330013275, "learning_rate": 9.978135393699484e-05, "loss": 0.0685, "step": 7830 }, { "grad_norm": 0.4263463020324707, "learning_rate": 9.977980659451548e-05, "loss": 0.0927, "step": 7840 }, { "grad_norm": 0.43094298243522644, "learning_rate": 9.977825380819135e-05, "loss": 0.0747, "step": 7850 }, { "grad_norm": 0.48073288798332214, "learning_rate": 9.97766955781922e-05, "loss": 0.0744, "step": 7860 }, { "grad_norm": 0.341185986995697, "learning_rate": 9.977513190468848e-05, "loss": 0.0774, "step": 7870 }, { "grad_norm": 0.35896676778793335, "learning_rate": 9.977356278785116e-05, "loss": 0.0706, "step": 7880 }, { "grad_norm": 0.5352097153663635, "learning_rate": 9.977198822785184e-05, "loss": 0.076, "step": 7890 }, { "grad_norm": 0.4418313801288605, "learning_rate": 9.977040822486273e-05, "loss": 0.077, "step": 7900 }, { "grad_norm": 0.44698765873908997, "learning_rate": 9.97688227790566e-05, "loss": 0.0765, "step": 7910 }, { "grad_norm": 0.505185067653656, "learning_rate": 9.976723189060684e-05, "loss": 0.0779, "step": 7920 }, { "grad_norm": 0.42256948351860046, "learning_rate": 9.976563555968742e-05, "loss": 0.0708, "step": 7930 }, { "grad_norm": 0.40005385875701904, "learning_rate": 9.976403378647292e-05, "loss": 0.0706, "step": 7940 }, { "grad_norm": 0.3780174255371094, "learning_rate": 9.97624265711385e-05, "loss": 0.0631, "step": 7950 }, { "grad_norm": 0.3970928192138672, "learning_rate": 9.976081391385993e-05, "loss": 0.0618, "step": 7960 }, { "grad_norm": 0.31584933400154114, "learning_rate": 9.975919581481356e-05, "loss": 0.0642, "step": 7970 }, { "grad_norm": 0.43062520027160645, "learning_rate": 9.975757227417634e-05, "loss": 0.0673, "step": 7980 }, { "grad_norm": 0.4625686705112457, "learning_rate": 9.975594329212586e-05, "loss": 0.0687, "step": 7990 }, { "grad_norm": 0.5338881611824036, "learning_rate": 9.97543088688402e-05, "loss": 0.067, "step": 8000 }, { "grad_norm": 0.41912707686424255, "learning_rate": 9.975266900449814e-05, "loss": 0.0742, "step": 8010 }, { "grad_norm": 0.6090839505195618, "learning_rate": 9.975102369927898e-05, "loss": 0.0831, "step": 8020 }, { "grad_norm": 0.39209944009780884, "learning_rate": 9.974937295336269e-05, "loss": 0.0687, "step": 8030 }, { "grad_norm": 0.4037485420703888, "learning_rate": 9.974771676692975e-05, "loss": 0.0783, "step": 8040 }, { "grad_norm": 0.46380773186683655, "learning_rate": 9.974605514016131e-05, "loss": 0.0871, "step": 8050 }, { "grad_norm": 0.5991764664649963, "learning_rate": 9.974438807323907e-05, "loss": 0.0734, "step": 8060 }, { "grad_norm": 0.41150471568107605, "learning_rate": 9.974271556634535e-05, "loss": 0.0641, "step": 8070 }, { "grad_norm": 0.4114815294742584, "learning_rate": 9.974103761966302e-05, "loss": 0.0701, "step": 8080 }, { "grad_norm": 0.8531020879745483, "learning_rate": 9.973935423337563e-05, "loss": 0.0757, "step": 8090 }, { "grad_norm": 0.32709449529647827, "learning_rate": 9.973766540766722e-05, "loss": 0.0724, "step": 8100 }, { "grad_norm": 0.42338213324546814, "learning_rate": 9.97359711427225e-05, "loss": 0.0677, "step": 8110 }, { "grad_norm": 0.3941614329814911, "learning_rate": 9.973427143872677e-05, "loss": 0.0851, "step": 8120 }, { "grad_norm": 0.4761667847633362, "learning_rate": 9.973256629586589e-05, "loss": 0.0673, "step": 8130 }, { "grad_norm": 0.38195934891700745, "learning_rate": 9.973085571432632e-05, "loss": 0.0669, "step": 8140 }, { "grad_norm": 0.4717448949813843, "learning_rate": 9.972913969429513e-05, "loss": 0.0742, "step": 8150 }, { "grad_norm": 0.4908832013607025, "learning_rate": 9.972741823596e-05, "loss": 0.075, "step": 8160 }, { "grad_norm": 0.4805298149585724, "learning_rate": 9.972569133950917e-05, "loss": 0.0758, "step": 8170 }, { "grad_norm": 0.4009990990161896, "learning_rate": 9.972395900513151e-05, "loss": 0.0753, "step": 8180 }, { "grad_norm": 0.40659669041633606, "learning_rate": 9.972222123301645e-05, "loss": 0.0675, "step": 8190 }, { "grad_norm": 0.39069658517837524, "learning_rate": 9.972047802335403e-05, "loss": 0.0695, "step": 8200 }, { "grad_norm": 0.4119361639022827, "learning_rate": 9.971872937633488e-05, "loss": 0.0714, "step": 8210 }, { "grad_norm": 0.7243053913116455, "learning_rate": 9.971697529215024e-05, "loss": 0.0688, "step": 8220 }, { "grad_norm": 0.5469810366630554, "learning_rate": 9.971521577099192e-05, "loss": 0.0845, "step": 8230 }, { "grad_norm": 0.5755481123924255, "learning_rate": 9.971345081305236e-05, "loss": 0.0798, "step": 8240 }, { "grad_norm": 0.40567466616630554, "learning_rate": 9.971168041852456e-05, "loss": 0.0655, "step": 8250 }, { "grad_norm": 0.5257545113563538, "learning_rate": 9.970990458760215e-05, "loss": 0.0691, "step": 8260 }, { "grad_norm": 0.5602033138275146, "learning_rate": 9.970812332047929e-05, "loss": 0.0667, "step": 8270 }, { "grad_norm": 0.47924333810806274, "learning_rate": 9.97063366173508e-05, "loss": 0.0764, "step": 8280 }, { "grad_norm": 0.4295230805873871, "learning_rate": 9.970454447841207e-05, "loss": 0.0694, "step": 8290 }, { "grad_norm": 0.5392473936080933, "learning_rate": 9.970274690385909e-05, "loss": 0.0742, "step": 8300 }, { "grad_norm": 0.48157384991645813, "learning_rate": 9.970094389388844e-05, "loss": 0.0652, "step": 8310 }, { "grad_norm": 0.5230880975723267, "learning_rate": 9.969913544869728e-05, "loss": 0.0728, "step": 8320 }, { "grad_norm": 0.45588797330856323, "learning_rate": 9.96973215684834e-05, "loss": 0.0702, "step": 8330 }, { "grad_norm": 0.6298514008522034, "learning_rate": 9.969550225344513e-05, "loss": 0.0753, "step": 8340 }, { "grad_norm": 0.5892630219459534, "learning_rate": 9.969367750378147e-05, "loss": 0.0713, "step": 8350 }, { "grad_norm": 0.3359255790710449, "learning_rate": 9.969184731969194e-05, "loss": 0.0737, "step": 8360 }, { "grad_norm": 0.6192734837532043, "learning_rate": 9.96900117013767e-05, "loss": 0.0701, "step": 8370 }, { "grad_norm": 0.5867041945457458, "learning_rate": 9.96881706490365e-05, "loss": 0.0771, "step": 8380 }, { "grad_norm": 0.4899236857891083, "learning_rate": 9.968632416287265e-05, "loss": 0.0673, "step": 8390 }, { "grad_norm": 0.4060177803039551, "learning_rate": 9.96844722430871e-05, "loss": 0.0829, "step": 8400 }, { "grad_norm": 0.5468680262565613, "learning_rate": 9.968261488988235e-05, "loss": 0.0817, "step": 8410 }, { "grad_norm": 0.42730575799942017, "learning_rate": 9.968075210346155e-05, "loss": 0.0675, "step": 8420 }, { "grad_norm": 0.5341565012931824, "learning_rate": 9.967888388402839e-05, "loss": 0.0751, "step": 8430 }, { "grad_norm": 0.5821292996406555, "learning_rate": 9.967701023178717e-05, "loss": 0.0769, "step": 8440 }, { "grad_norm": 0.40393370389938354, "learning_rate": 9.967513114694282e-05, "loss": 0.0653, "step": 8450 }, { "grad_norm": 0.44113802909851074, "learning_rate": 9.967324662970079e-05, "loss": 0.0751, "step": 8460 }, { "grad_norm": 0.3776986002922058, "learning_rate": 9.96713566802672e-05, "loss": 0.0666, "step": 8470 }, { "grad_norm": 0.4532697796821594, "learning_rate": 9.966946129884873e-05, "loss": 0.0698, "step": 8480 }, { "grad_norm": 0.40877407789230347, "learning_rate": 9.966756048565265e-05, "loss": 0.0801, "step": 8490 }, { "grad_norm": 0.5046908259391785, "learning_rate": 9.966565424088681e-05, "loss": 0.071, "step": 8500 }, { "grad_norm": 0.7039467692375183, "learning_rate": 9.96637425647597e-05, "loss": 0.076, "step": 8510 }, { "grad_norm": 0.5690678358078003, "learning_rate": 9.966182545748038e-05, "loss": 0.0731, "step": 8520 }, { "grad_norm": 0.4588729441165924, "learning_rate": 9.96599029192585e-05, "loss": 0.0835, "step": 8530 }, { "grad_norm": 0.5211263298988342, "learning_rate": 9.965797495030428e-05, "loss": 0.0752, "step": 8540 }, { "grad_norm": 0.5155799388885498, "learning_rate": 9.96560415508286e-05, "loss": 0.0755, "step": 8550 }, { "grad_norm": 0.6061550974845886, "learning_rate": 9.965410272104286e-05, "loss": 0.0722, "step": 8560 }, { "grad_norm": 0.4772128760814667, "learning_rate": 9.96521584611591e-05, "loss": 0.0731, "step": 8570 }, { "grad_norm": 0.5566552877426147, "learning_rate": 9.965020877138994e-05, "loss": 0.0761, "step": 8580 }, { "grad_norm": 0.5767468214035034, "learning_rate": 9.964825365194861e-05, "loss": 0.0743, "step": 8590 }, { "grad_norm": 0.5303727388381958, "learning_rate": 9.96462931030489e-05, "loss": 0.0704, "step": 8600 }, { "grad_norm": 0.5846579074859619, "learning_rate": 9.96443271249052e-05, "loss": 0.0703, "step": 8610 }, { "grad_norm": 0.3858364224433899, "learning_rate": 9.964235571773255e-05, "loss": 0.0703, "step": 8620 }, { "grad_norm": 0.3397829532623291, "learning_rate": 9.96403788817465e-05, "loss": 0.0836, "step": 8630 }, { "grad_norm": 0.39111560583114624, "learning_rate": 9.963839661716325e-05, "loss": 0.0775, "step": 8640 }, { "grad_norm": 0.45936644077301025, "learning_rate": 9.963640892419958e-05, "loss": 0.067, "step": 8650 }, { "grad_norm": 0.456405907869339, "learning_rate": 9.963441580307286e-05, "loss": 0.079, "step": 8660 }, { "grad_norm": 0.597902774810791, "learning_rate": 9.963241725400104e-05, "loss": 0.0854, "step": 8670 }, { "grad_norm": 0.46849653124809265, "learning_rate": 9.963041327720271e-05, "loss": 0.0752, "step": 8680 }, { "grad_norm": 0.43834131956100464, "learning_rate": 9.962840387289697e-05, "loss": 0.0755, "step": 8690 }, { "grad_norm": 0.49583566188812256, "learning_rate": 9.962638904130363e-05, "loss": 0.081, "step": 8700 }, { "grad_norm": 0.48236140608787537, "learning_rate": 9.962436878264298e-05, "loss": 0.0746, "step": 8710 }, { "grad_norm": 0.34840813279151917, "learning_rate": 9.962234309713598e-05, "loss": 0.0702, "step": 8720 }, { "grad_norm": 0.42713412642478943, "learning_rate": 9.962031198500414e-05, "loss": 0.0725, "step": 8730 }, { "grad_norm": 0.5278006196022034, "learning_rate": 9.961827544646958e-05, "loss": 0.0782, "step": 8740 }, { "grad_norm": 0.5569372773170471, "learning_rate": 9.961623348175501e-05, "loss": 0.0702, "step": 8750 }, { "grad_norm": 0.43391042947769165, "learning_rate": 9.961418609108377e-05, "loss": 0.083, "step": 8760 }, { "grad_norm": 0.4711116850376129, "learning_rate": 9.961213327467971e-05, "loss": 0.0852, "step": 8770 }, { "grad_norm": 0.38861438632011414, "learning_rate": 9.961007503276736e-05, "loss": 0.0633, "step": 8780 }, { "grad_norm": 0.869493842124939, "learning_rate": 9.960801136557179e-05, "loss": 0.0676, "step": 8790 }, { "grad_norm": 0.4536612629890442, "learning_rate": 9.960594227331866e-05, "loss": 0.0706, "step": 8800 }, { "grad_norm": 0.4684891402721405, "learning_rate": 9.960386775623429e-05, "loss": 0.0957, "step": 8810 }, { "grad_norm": 0.5156434178352356, "learning_rate": 9.96017878145455e-05, "loss": 0.0824, "step": 8820 }, { "grad_norm": 0.5168371200561523, "learning_rate": 9.959970244847977e-05, "loss": 0.0737, "step": 8830 }, { "grad_norm": 0.39698222279548645, "learning_rate": 9.959761165826518e-05, "loss": 0.0762, "step": 8840 }, { "grad_norm": 0.5687766075134277, "learning_rate": 9.959551544413033e-05, "loss": 0.0728, "step": 8850 }, { "grad_norm": 0.40203168988227844, "learning_rate": 9.959341380630448e-05, "loss": 0.0676, "step": 8860 }, { "grad_norm": 0.3514251112937927, "learning_rate": 9.959130674501746e-05, "loss": 0.0658, "step": 8870 }, { "grad_norm": 0.5289137363433838, "learning_rate": 9.958919426049968e-05, "loss": 0.0668, "step": 8880 }, { "grad_norm": 0.4651699662208557, "learning_rate": 9.958707635298219e-05, "loss": 0.068, "step": 8890 }, { "grad_norm": 0.4172056019306183, "learning_rate": 9.958495302269657e-05, "loss": 0.0786, "step": 8900 }, { "grad_norm": 0.5674433708190918, "learning_rate": 9.958282426987503e-05, "loss": 0.0719, "step": 8910 }, { "grad_norm": 0.3354799449443817, "learning_rate": 9.95806900947504e-05, "loss": 0.0674, "step": 8920 }, { "grad_norm": 0.3733663260936737, "learning_rate": 9.957855049755604e-05, "loss": 0.0685, "step": 8930 }, { "grad_norm": 0.3949604630470276, "learning_rate": 9.957640547852593e-05, "loss": 0.0651, "step": 8940 }, { "grad_norm": 0.4766322374343872, "learning_rate": 9.957425503789466e-05, "loss": 0.0769, "step": 8950 }, { "grad_norm": 0.7661460638046265, "learning_rate": 9.957209917589738e-05, "loss": 0.0686, "step": 8960 }, { "grad_norm": 0.472709596157074, "learning_rate": 9.956993789276987e-05, "loss": 0.0657, "step": 8970 }, { "grad_norm": 0.4827805459499359, "learning_rate": 9.956777118874847e-05, "loss": 0.0788, "step": 8980 }, { "grad_norm": 0.3519286811351776, "learning_rate": 9.956559906407016e-05, "loss": 0.0784, "step": 8990 }, { "grad_norm": 0.55181884765625, "learning_rate": 9.956342151897245e-05, "loss": 0.0637, "step": 9000 }, { "grad_norm": 0.3952200412750244, "learning_rate": 9.956123855369346e-05, "loss": 0.0687, "step": 9010 }, { "grad_norm": 0.4131239354610443, "learning_rate": 9.955905016847196e-05, "loss": 0.0732, "step": 9020 }, { "grad_norm": 0.3410288691520691, "learning_rate": 9.955685636354723e-05, "loss": 0.0618, "step": 9030 }, { "grad_norm": 0.4007139503955841, "learning_rate": 9.95546571391592e-05, "loss": 0.0718, "step": 9040 }, { "grad_norm": 0.49543020129203796, "learning_rate": 9.955245249554837e-05, "loss": 0.0665, "step": 9050 }, { "grad_norm": 0.43311095237731934, "learning_rate": 9.955024243295582e-05, "loss": 0.072, "step": 9060 }, { "grad_norm": 0.39758509397506714, "learning_rate": 9.954802695162328e-05, "loss": 0.0783, "step": 9070 }, { "grad_norm": 0.49770671129226685, "learning_rate": 9.954580605179302e-05, "loss": 0.0763, "step": 9080 }, { "grad_norm": 0.3823077976703644, "learning_rate": 9.954357973370788e-05, "loss": 0.0768, "step": 9090 }, { "grad_norm": 0.4546571969985962, "learning_rate": 9.954134799761135e-05, "loss": 0.0613, "step": 9100 }, { "grad_norm": 0.5522615313529968, "learning_rate": 9.953911084374748e-05, "loss": 0.0697, "step": 9110 }, { "grad_norm": 0.35819631814956665, "learning_rate": 9.953686827236093e-05, "loss": 0.0657, "step": 9120 }, { "grad_norm": 0.4612543284893036, "learning_rate": 9.953462028369695e-05, "loss": 0.0775, "step": 9130 }, { "grad_norm": 0.4205882251262665, "learning_rate": 9.953236687800136e-05, "loss": 0.0684, "step": 9140 }, { "grad_norm": 0.40412843227386475, "learning_rate": 9.95301080555206e-05, "loss": 0.0705, "step": 9150 }, { "grad_norm": 0.4450394809246063, "learning_rate": 9.952784381650171e-05, "loss": 0.0711, "step": 9160 }, { "grad_norm": 0.40257835388183594, "learning_rate": 9.952557416119226e-05, "loss": 0.0775, "step": 9170 }, { "grad_norm": 0.39759084582328796, "learning_rate": 9.95232990898405e-05, "loss": 0.0714, "step": 9180 }, { "grad_norm": 0.5144035816192627, "learning_rate": 9.95210186026952e-05, "loss": 0.0708, "step": 9190 }, { "grad_norm": 0.40229302644729614, "learning_rate": 9.951873270000576e-05, "loss": 0.0664, "step": 9200 }, { "grad_norm": 0.42717841267585754, "learning_rate": 9.951644138202216e-05, "loss": 0.0681, "step": 9210 }, { "grad_norm": 0.46597525477409363, "learning_rate": 9.951414464899498e-05, "loss": 0.0744, "step": 9220 }, { "grad_norm": 0.5512407422065735, "learning_rate": 9.951184250117538e-05, "loss": 0.071, "step": 9230 }, { "grad_norm": 0.3417923152446747, "learning_rate": 9.950953493881513e-05, "loss": 0.068, "step": 9240 }, { "grad_norm": 0.4009872078895569, "learning_rate": 9.950722196216658e-05, "loss": 0.0743, "step": 9250 }, { "grad_norm": 0.3808711767196655, "learning_rate": 9.950490357148265e-05, "loss": 0.0651, "step": 9260 }, { "grad_norm": 0.5247361063957214, "learning_rate": 9.950257976701692e-05, "loss": 0.0742, "step": 9270 }, { "grad_norm": 0.6160919070243835, "learning_rate": 9.950025054902348e-05, "loss": 0.0723, "step": 9280 }, { "grad_norm": 0.5815402269363403, "learning_rate": 9.949791591775706e-05, "loss": 0.0646, "step": 9290 }, { "grad_norm": 0.5593011379241943, "learning_rate": 9.949557587347298e-05, "loss": 0.0769, "step": 9300 }, { "grad_norm": 0.467936635017395, "learning_rate": 9.949323041642713e-05, "loss": 0.0748, "step": 9310 }, { "grad_norm": 0.612797200679779, "learning_rate": 9.949087954687602e-05, "loss": 0.0825, "step": 9320 }, { "grad_norm": 0.3864544928073883, "learning_rate": 9.948852326507672e-05, "loss": 0.0603, "step": 9330 }, { "grad_norm": 0.534493088722229, "learning_rate": 9.948616157128694e-05, "loss": 0.0774, "step": 9340 }, { "grad_norm": 0.4868377447128296, "learning_rate": 9.948379446576493e-05, "loss": 0.0621, "step": 9350 }, { "grad_norm": 0.5003422498703003, "learning_rate": 9.948142194876952e-05, "loss": 0.0648, "step": 9360 }, { "grad_norm": 0.5284768342971802, "learning_rate": 9.947904402056024e-05, "loss": 0.0683, "step": 9370 }, { "grad_norm": 0.43280085921287537, "learning_rate": 9.947666068139708e-05, "loss": 0.0643, "step": 9380 }, { "grad_norm": 0.5086074471473694, "learning_rate": 9.947427193154071e-05, "loss": 0.0758, "step": 9390 }, { "grad_norm": 0.4872317314147949, "learning_rate": 9.947187777125233e-05, "loss": 0.0783, "step": 9400 }, { "grad_norm": 0.5852182507514954, "learning_rate": 9.946947820079377e-05, "loss": 0.076, "step": 9410 }, { "grad_norm": 0.35115155577659607, "learning_rate": 9.946707322042747e-05, "loss": 0.0721, "step": 9420 }, { "grad_norm": 0.4837195575237274, "learning_rate": 9.94646628304164e-05, "loss": 0.0736, "step": 9430 }, { "grad_norm": 0.5770086646080017, "learning_rate": 9.946224703102418e-05, "loss": 0.0751, "step": 9440 }, { "grad_norm": 0.52701336145401, "learning_rate": 9.945982582251498e-05, "loss": 0.0718, "step": 9450 }, { "grad_norm": 0.4483575224876404, "learning_rate": 9.94573992051536e-05, "loss": 0.0692, "step": 9460 }, { "grad_norm": 0.4336952865123749, "learning_rate": 9.94549671792054e-05, "loss": 0.0748, "step": 9470 }, { "grad_norm": 0.5787900686264038, "learning_rate": 9.945252974493635e-05, "loss": 0.0682, "step": 9480 }, { "grad_norm": 0.5362372994422913, "learning_rate": 9.9450086902613e-05, "loss": 0.0667, "step": 9490 }, { "grad_norm": 0.5410675406455994, "learning_rate": 9.944763865250248e-05, "loss": 0.0719, "step": 9500 }, { "grad_norm": 0.48150816559791565, "learning_rate": 9.944518499487254e-05, "loss": 0.0745, "step": 9510 }, { "grad_norm": 0.4642176330089569, "learning_rate": 9.944272592999151e-05, "loss": 0.0692, "step": 9520 }, { "grad_norm": 0.4985502362251282, "learning_rate": 9.94402614581283e-05, "loss": 0.0738, "step": 9530 }, { "grad_norm": 0.5794498920440674, "learning_rate": 9.943779157955244e-05, "loss": 0.0734, "step": 9540 }, { "grad_norm": 0.41662099957466125, "learning_rate": 9.943531629453403e-05, "loss": 0.0636, "step": 9550 }, { "grad_norm": 0.49380597472190857, "learning_rate": 9.943283560334375e-05, "loss": 0.074, "step": 9560 }, { "grad_norm": 0.5060486197471619, "learning_rate": 9.943034950625288e-05, "loss": 0.0709, "step": 9570 }, { "grad_norm": 0.3091528117656708, "learning_rate": 9.942785800353332e-05, "loss": 0.0769, "step": 9580 }, { "grad_norm": 0.4502407908439636, "learning_rate": 9.942536109545751e-05, "loss": 0.0778, "step": 9590 }, { "grad_norm": 0.38042834401130676, "learning_rate": 9.942285878229853e-05, "loss": 0.0669, "step": 9600 }, { "grad_norm": 0.5833198428153992, "learning_rate": 9.942035106433001e-05, "loss": 0.0699, "step": 9610 }, { "grad_norm": 0.4033151865005493, "learning_rate": 9.94178379418262e-05, "loss": 0.0692, "step": 9620 }, { "grad_norm": 0.5684360861778259, "learning_rate": 9.941531941506194e-05, "loss": 0.0786, "step": 9630 }, { "grad_norm": 0.5191251635551453, "learning_rate": 9.941279548431263e-05, "loss": 0.0667, "step": 9640 }, { "grad_norm": 0.49700358510017395, "learning_rate": 9.941026614985431e-05, "loss": 0.0757, "step": 9650 }, { "grad_norm": 0.40343594551086426, "learning_rate": 9.940773141196357e-05, "loss": 0.0695, "step": 9660 }, { "grad_norm": 0.4825495481491089, "learning_rate": 9.94051912709176e-05, "loss": 0.0678, "step": 9670 }, { "grad_norm": 0.4236222207546234, "learning_rate": 9.940264572699421e-05, "loss": 0.0732, "step": 9680 }, { "grad_norm": 0.5236784219741821, "learning_rate": 9.940009478047174e-05, "loss": 0.0698, "step": 9690 }, { "grad_norm": 0.4317050874233246, "learning_rate": 9.939753843162918e-05, "loss": 0.0763, "step": 9700 }, { "grad_norm": 0.42893165349960327, "learning_rate": 9.939497668074609e-05, "loss": 0.0619, "step": 9710 }, { "grad_norm": 0.45670387148857117, "learning_rate": 9.93924095281026e-05, "loss": 0.0656, "step": 9720 }, { "grad_norm": 0.5329908132553101, "learning_rate": 9.938983697397948e-05, "loss": 0.0739, "step": 9730 }, { "grad_norm": 0.5888304710388184, "learning_rate": 9.938725901865805e-05, "loss": 0.0708, "step": 9740 }, { "grad_norm": 0.4973049461841583, "learning_rate": 9.93846756624202e-05, "loss": 0.0716, "step": 9750 }, { "grad_norm": 0.5508347153663635, "learning_rate": 9.938208690554849e-05, "loss": 0.0707, "step": 9760 }, { "grad_norm": 0.42993462085723877, "learning_rate": 9.9379492748326e-05, "loss": 0.0717, "step": 9770 }, { "grad_norm": 0.5544666051864624, "learning_rate": 9.937689319103641e-05, "loss": 0.0704, "step": 9780 }, { "grad_norm": 0.4417281150817871, "learning_rate": 9.937428823396404e-05, "loss": 0.0685, "step": 9790 }, { "grad_norm": 0.32399341464042664, "learning_rate": 9.937167787739372e-05, "loss": 0.0656, "step": 9800 }, { "grad_norm": 0.4504756033420563, "learning_rate": 9.936906212161095e-05, "loss": 0.0681, "step": 9810 }, { "grad_norm": 0.6160427927970886, "learning_rate": 9.936644096690176e-05, "loss": 0.0735, "step": 9820 }, { "grad_norm": 0.6612057089805603, "learning_rate": 9.936381441355282e-05, "loss": 0.0683, "step": 9830 }, { "grad_norm": 0.4848567545413971, "learning_rate": 9.936118246185136e-05, "loss": 0.0656, "step": 9840 }, { "grad_norm": 0.4999466836452484, "learning_rate": 9.935854511208518e-05, "loss": 0.0821, "step": 9850 }, { "grad_norm": 0.39433181285858154, "learning_rate": 9.935590236454272e-05, "loss": 0.071, "step": 9860 }, { "grad_norm": 0.32659104466438293, "learning_rate": 9.935325421951298e-05, "loss": 0.064, "step": 9870 }, { "grad_norm": 0.44452789425849915, "learning_rate": 9.935060067728557e-05, "loss": 0.0748, "step": 9880 }, { "grad_norm": 0.4176996350288391, "learning_rate": 9.934794173815067e-05, "loss": 0.0677, "step": 9890 }, { "grad_norm": 0.41578030586242676, "learning_rate": 9.934527740239906e-05, "loss": 0.0696, "step": 9900 }, { "grad_norm": 0.39458298683166504, "learning_rate": 9.934260767032209e-05, "loss": 0.0839, "step": 9910 }, { "grad_norm": 0.39296162128448486, "learning_rate": 9.933993254221172e-05, "loss": 0.0654, "step": 9920 }, { "grad_norm": 0.45974257588386536, "learning_rate": 9.933725201836053e-05, "loss": 0.0705, "step": 9930 }, { "grad_norm": 0.5595496296882629, "learning_rate": 9.933456609906162e-05, "loss": 0.0686, "step": 9940 }, { "grad_norm": 0.4263915419578552, "learning_rate": 9.933187478460875e-05, "loss": 0.0773, "step": 9950 }, { "grad_norm": 0.5136075615882874, "learning_rate": 9.93291780752962e-05, "loss": 0.0665, "step": 9960 }, { "grad_norm": 0.566550076007843, "learning_rate": 9.932647597141893e-05, "loss": 0.0711, "step": 9970 }, { "grad_norm": 0.5951512455940247, "learning_rate": 9.932376847327239e-05, "loss": 0.0715, "step": 9980 }, { "grad_norm": 0.43263545632362366, "learning_rate": 9.932105558115268e-05, "loss": 0.0682, "step": 9990 }, { "grad_norm": 0.48652184009552, "learning_rate": 9.931833729535651e-05, "loss": 0.0626, "step": 10000 }, { "grad_norm": 0.4607294201850891, "learning_rate": 9.931561361618111e-05, "loss": 0.0593, "step": 10010 }, { "grad_norm": 0.48242855072021484, "learning_rate": 9.931288454392435e-05, "loss": 0.0784, "step": 10020 }, { "grad_norm": 0.4629051387310028, "learning_rate": 9.931015007888467e-05, "loss": 0.0719, "step": 10030 }, { "grad_norm": 2.5057222843170166, "learning_rate": 9.930741022136112e-05, "loss": 0.0868, "step": 10040 }, { "grad_norm": 0.40990978479385376, "learning_rate": 9.930466497165333e-05, "loss": 0.079, "step": 10050 }, { "grad_norm": 0.48297134041786194, "learning_rate": 9.93019143300615e-05, "loss": 0.0792, "step": 10060 }, { "grad_norm": 0.5167965292930603, "learning_rate": 9.929915829688644e-05, "loss": 0.0784, "step": 10070 }, { "grad_norm": 0.45463165640830994, "learning_rate": 9.929639687242955e-05, "loss": 0.0711, "step": 10080 }, { "grad_norm": 0.6691656112670898, "learning_rate": 9.929363005699281e-05, "loss": 0.0714, "step": 10090 }, { "grad_norm": 0.4858325719833374, "learning_rate": 9.92908578508788e-05, "loss": 0.0799, "step": 10100 }, { "grad_norm": 0.5296667814254761, "learning_rate": 9.928808025439069e-05, "loss": 0.0758, "step": 10110 }, { "grad_norm": 0.44913697242736816, "learning_rate": 9.928529726783223e-05, "loss": 0.0662, "step": 10120 }, { "grad_norm": 0.40211477875709534, "learning_rate": 9.928250889150774e-05, "loss": 0.071, "step": 10130 }, { "grad_norm": 0.466731995344162, "learning_rate": 9.92797151257222e-05, "loss": 0.0657, "step": 10140 }, { "grad_norm": 0.397511750459671, "learning_rate": 9.927691597078108e-05, "loss": 0.0668, "step": 10150 }, { "grad_norm": 0.3769861161708832, "learning_rate": 9.927411142699053e-05, "loss": 0.0721, "step": 10160 }, { "grad_norm": 0.4593099355697632, "learning_rate": 9.927130149465725e-05, "loss": 0.0732, "step": 10170 }, { "grad_norm": 0.5268930196762085, "learning_rate": 9.92684861740885e-05, "loss": 0.0663, "step": 10180 }, { "grad_norm": 0.4148937463760376, "learning_rate": 9.926566546559217e-05, "loss": 0.0627, "step": 10190 }, { "grad_norm": 0.3853703439235687, "learning_rate": 9.926283936947673e-05, "loss": 0.0685, "step": 10200 }, { "grad_norm": 0.6212238669395447, "learning_rate": 9.926000788605126e-05, "loss": 0.0666, "step": 10210 }, { "grad_norm": 0.5006259083747864, "learning_rate": 9.92571710156254e-05, "loss": 0.0721, "step": 10220 }, { "grad_norm": 0.42256268858909607, "learning_rate": 9.925432875850936e-05, "loss": 0.0704, "step": 10230 }, { "grad_norm": 0.4044932425022125, "learning_rate": 9.925148111501396e-05, "loss": 0.0695, "step": 10240 }, { "grad_norm": 0.4436195492744446, "learning_rate": 9.924862808545066e-05, "loss": 0.0601, "step": 10250 }, { "grad_norm": 0.4131959080696106, "learning_rate": 9.924576967013141e-05, "loss": 0.0728, "step": 10260 }, { "grad_norm": 0.4919353425502777, "learning_rate": 9.924290586936887e-05, "loss": 0.0769, "step": 10270 }, { "grad_norm": 0.5394211411476135, "learning_rate": 9.924003668347614e-05, "loss": 0.0718, "step": 10280 }, { "grad_norm": 0.44990235567092896, "learning_rate": 9.923716211276704e-05, "loss": 0.0651, "step": 10290 }, { "grad_norm": 0.4565923511981964, "learning_rate": 9.923428215755594e-05, "loss": 0.0715, "step": 10300 }, { "grad_norm": 0.4094965159893036, "learning_rate": 9.923139681815775e-05, "loss": 0.0671, "step": 10310 }, { "grad_norm": 0.38192975521087646, "learning_rate": 9.922850609488801e-05, "loss": 0.0713, "step": 10320 }, { "grad_norm": 0.4690343737602234, "learning_rate": 9.922560998806287e-05, "loss": 0.0662, "step": 10330 }, { "grad_norm": 0.4828832447528839, "learning_rate": 9.922270849799905e-05, "loss": 0.0612, "step": 10340 }, { "grad_norm": 0.46609267592430115, "learning_rate": 9.92198016250138e-05, "loss": 0.0757, "step": 10350 }, { "grad_norm": 0.4863550066947937, "learning_rate": 9.921688936942506e-05, "loss": 0.0655, "step": 10360 }, { "grad_norm": 0.46058163046836853, "learning_rate": 9.921397173155129e-05, "loss": 0.0793, "step": 10370 }, { "grad_norm": 0.42917266488075256, "learning_rate": 9.921104871171157e-05, "loss": 0.0713, "step": 10380 }, { "grad_norm": 0.3872198760509491, "learning_rate": 9.920812031022554e-05, "loss": 0.0593, "step": 10390 }, { "grad_norm": 0.32539212703704834, "learning_rate": 9.920518652741348e-05, "loss": 0.064, "step": 10400 }, { "grad_norm": 0.47413936257362366, "learning_rate": 9.920224736359618e-05, "loss": 0.0668, "step": 10410 }, { "grad_norm": 0.5273699164390564, "learning_rate": 9.91993028190951e-05, "loss": 0.0752, "step": 10420 }, { "grad_norm": 0.6465321779251099, "learning_rate": 9.919635289423222e-05, "loss": 0.0665, "step": 10430 }, { "grad_norm": 0.40985703468322754, "learning_rate": 9.919339758933015e-05, "loss": 0.0808, "step": 10440 }, { "grad_norm": 0.46415698528289795, "learning_rate": 9.919043690471209e-05, "loss": 0.0676, "step": 10450 }, { "grad_norm": 0.400473415851593, "learning_rate": 9.91874708407018e-05, "loss": 0.0709, "step": 10460 }, { "grad_norm": 0.542919397354126, "learning_rate": 9.918449939762367e-05, "loss": 0.0657, "step": 10470 }, { "grad_norm": 0.5956948399543762, "learning_rate": 9.91815225758026e-05, "loss": 0.0663, "step": 10480 }, { "grad_norm": 0.48112574219703674, "learning_rate": 9.917854037556419e-05, "loss": 0.067, "step": 10490 }, { "grad_norm": 0.3533588647842407, "learning_rate": 9.917555279723454e-05, "loss": 0.0751, "step": 10500 }, { "grad_norm": 0.4803289473056793, "learning_rate": 9.917255984114036e-05, "loss": 0.0745, "step": 10510 }, { "grad_norm": 0.48279494047164917, "learning_rate": 9.916956150760896e-05, "loss": 0.0657, "step": 10520 }, { "grad_norm": 0.4304877519607544, "learning_rate": 9.916655779696826e-05, "loss": 0.0636, "step": 10530 }, { "grad_norm": 0.5080978870391846, "learning_rate": 9.916354870954671e-05, "loss": 0.0682, "step": 10540 }, { "grad_norm": 0.619437575340271, "learning_rate": 9.91605342456734e-05, "loss": 0.0715, "step": 10550 }, { "grad_norm": 0.4172389805316925, "learning_rate": 9.915751440567795e-05, "loss": 0.061, "step": 10560 }, { "grad_norm": 0.5594256520271301, "learning_rate": 9.915448918989066e-05, "loss": 0.0689, "step": 10570 }, { "grad_norm": 0.4916750490665436, "learning_rate": 9.915145859864232e-05, "loss": 0.0759, "step": 10580 }, { "grad_norm": 0.3855871260166168, "learning_rate": 9.914842263226437e-05, "loss": 0.0705, "step": 10590 }, { "grad_norm": 0.3976268172264099, "learning_rate": 9.914538129108882e-05, "loss": 0.0793, "step": 10600 }, { "grad_norm": 0.5309447646141052, "learning_rate": 9.914233457544825e-05, "loss": 0.0715, "step": 10610 }, { "grad_norm": 0.40619075298309326, "learning_rate": 9.913928248567586e-05, "loss": 0.0732, "step": 10620 }, { "grad_norm": 0.4593929052352905, "learning_rate": 9.913622502210542e-05, "loss": 0.071, "step": 10630 }, { "grad_norm": 0.3949621319770813, "learning_rate": 9.913316218507128e-05, "loss": 0.0692, "step": 10640 }, { "grad_norm": 0.47915568947792053, "learning_rate": 9.91300939749084e-05, "loss": 0.0631, "step": 10650 }, { "grad_norm": 0.3980046212673187, "learning_rate": 9.91270203919523e-05, "loss": 0.0665, "step": 10660 }, { "grad_norm": 0.4513587951660156, "learning_rate": 9.912394143653912e-05, "loss": 0.0707, "step": 10670 }, { "grad_norm": 0.5394314527511597, "learning_rate": 9.912085710900555e-05, "loss": 0.0713, "step": 10680 }, { "grad_norm": 0.3990342915058136, "learning_rate": 9.911776740968892e-05, "loss": 0.0736, "step": 10690 }, { "grad_norm": 0.6569010019302368, "learning_rate": 9.911467233892709e-05, "loss": 0.0733, "step": 10700 }, { "grad_norm": 0.3988771140575409, "learning_rate": 9.911157189705853e-05, "loss": 0.0698, "step": 10710 }, { "grad_norm": 0.45654982328414917, "learning_rate": 9.910846608442229e-05, "loss": 0.0608, "step": 10720 }, { "grad_norm": 0.5845324993133545, "learning_rate": 9.910535490135805e-05, "loss": 0.068, "step": 10730 }, { "grad_norm": 0.512550413608551, "learning_rate": 9.910223834820603e-05, "loss": 0.067, "step": 10740 }, { "grad_norm": 0.4913901686668396, "learning_rate": 9.909911642530703e-05, "loss": 0.0671, "step": 10750 }, { "grad_norm": 0.4325422942638397, "learning_rate": 9.909598913300249e-05, "loss": 0.0588, "step": 10760 }, { "grad_norm": 0.49136051535606384, "learning_rate": 9.909285647163438e-05, "loss": 0.0673, "step": 10770 }, { "grad_norm": 0.5553287863731384, "learning_rate": 9.908971844154531e-05, "loss": 0.0612, "step": 10780 }, { "grad_norm": 0.4745882749557495, "learning_rate": 9.908657504307843e-05, "loss": 0.0615, "step": 10790 }, { "grad_norm": 0.5558590292930603, "learning_rate": 9.908342627657751e-05, "loss": 0.0606, "step": 10800 }, { "grad_norm": 0.6977623105049133, "learning_rate": 9.908027214238689e-05, "loss": 0.0764, "step": 10810 }, { "grad_norm": 0.441669762134552, "learning_rate": 9.90771126408515e-05, "loss": 0.0661, "step": 10820 }, { "grad_norm": 0.5635393857955933, "learning_rate": 9.907394777231685e-05, "loss": 0.0782, "step": 10830 }, { "grad_norm": 0.4622465968132019, "learning_rate": 9.907077753712905e-05, "loss": 0.0688, "step": 10840 }, { "grad_norm": 0.5362545251846313, "learning_rate": 9.906760193563482e-05, "loss": 0.0711, "step": 10850 }, { "grad_norm": 0.37059536576271057, "learning_rate": 9.906442096818139e-05, "loss": 0.0784, "step": 10860 }, { "grad_norm": 0.33535638451576233, "learning_rate": 9.906123463511665e-05, "loss": 0.0698, "step": 10870 }, { "grad_norm": 0.500103235244751, "learning_rate": 9.905804293678907e-05, "loss": 0.0669, "step": 10880 }, { "grad_norm": 0.43850401043891907, "learning_rate": 9.905484587354766e-05, "loss": 0.0704, "step": 10890 }, { "grad_norm": 0.6450851559638977, "learning_rate": 9.905164344574205e-05, "loss": 0.0726, "step": 10900 }, { "grad_norm": 0.421535462141037, "learning_rate": 9.904843565372248e-05, "loss": 0.0706, "step": 10910 }, { "grad_norm": 0.565152645111084, "learning_rate": 9.904522249783972e-05, "loss": 0.0657, "step": 10920 }, { "grad_norm": 0.40530288219451904, "learning_rate": 9.904200397844517e-05, "loss": 0.0693, "step": 10930 }, { "grad_norm": 0.4469291567802429, "learning_rate": 9.903878009589078e-05, "loss": 0.0647, "step": 10940 }, { "grad_norm": 0.4723098874092102, "learning_rate": 9.903555085052915e-05, "loss": 0.0692, "step": 10950 }, { "grad_norm": 0.5677735209465027, "learning_rate": 9.903231624271338e-05, "loss": 0.0679, "step": 10960 }, { "grad_norm": 0.5135167837142944, "learning_rate": 9.902907627279724e-05, "loss": 0.0637, "step": 10970 }, { "grad_norm": 0.6426823735237122, "learning_rate": 9.902583094113504e-05, "loss": 0.0705, "step": 10980 }, { "grad_norm": 0.5272765159606934, "learning_rate": 9.902258024808168e-05, "loss": 0.0633, "step": 10990 }, { "grad_norm": 0.5241349339485168, "learning_rate": 9.901932419399264e-05, "loss": 0.0623, "step": 11000 }, { "grad_norm": 0.412709504365921, "learning_rate": 9.9016062779224e-05, "loss": 0.0709, "step": 11010 }, { "grad_norm": 0.45777732133865356, "learning_rate": 9.901279600413242e-05, "loss": 0.0784, "step": 11020 }, { "grad_norm": 0.44411230087280273, "learning_rate": 9.900952386907518e-05, "loss": 0.0754, "step": 11030 }, { "grad_norm": 0.37789186835289, "learning_rate": 9.90062463744101e-05, "loss": 0.07, "step": 11040 }, { "grad_norm": 0.49380555748939514, "learning_rate": 9.900296352049558e-05, "loss": 0.0664, "step": 11050 }, { "grad_norm": 0.6242077350616455, "learning_rate": 9.899967530769065e-05, "loss": 0.064, "step": 11060 }, { "grad_norm": 0.42219188809394836, "learning_rate": 9.899638173635489e-05, "loss": 0.0646, "step": 11070 }, { "grad_norm": 0.4183530807495117, "learning_rate": 9.899308280684849e-05, "loss": 0.0711, "step": 11080 }, { "grad_norm": 0.30128633975982666, "learning_rate": 9.898977851953222e-05, "loss": 0.0707, "step": 11090 }, { "grad_norm": 0.35067349672317505, "learning_rate": 9.898646887476741e-05, "loss": 0.0701, "step": 11100 }, { "grad_norm": 0.39259999990463257, "learning_rate": 9.898315387291603e-05, "loss": 0.0718, "step": 11110 }, { "grad_norm": 0.53420090675354, "learning_rate": 9.89798335143406e-05, "loss": 0.0653, "step": 11120 }, { "grad_norm": 0.517990231513977, "learning_rate": 9.897650779940419e-05, "loss": 0.0668, "step": 11130 }, { "grad_norm": 0.4515586495399475, "learning_rate": 9.897317672847054e-05, "loss": 0.0626, "step": 11140 }, { "grad_norm": 0.4661615490913391, "learning_rate": 9.89698403019039e-05, "loss": 0.067, "step": 11150 }, { "grad_norm": 0.4197767972946167, "learning_rate": 9.896649852006917e-05, "loss": 0.0625, "step": 11160 }, { "grad_norm": 0.46370720863342285, "learning_rate": 9.896315138333177e-05, "loss": 0.0727, "step": 11170 }, { "grad_norm": 0.3261590600013733, "learning_rate": 9.895979889205774e-05, "loss": 0.0586, "step": 11180 }, { "grad_norm": 0.42905619740486145, "learning_rate": 9.895644104661372e-05, "loss": 0.0656, "step": 11190 }, { "grad_norm": 0.6115080118179321, "learning_rate": 9.895307784736691e-05, "loss": 0.0599, "step": 11200 }, { "grad_norm": 0.46171537041664124, "learning_rate": 9.894970929468512e-05, "loss": 0.0594, "step": 11210 }, { "grad_norm": 0.5117322206497192, "learning_rate": 9.89463353889367e-05, "loss": 0.0619, "step": 11220 }, { "grad_norm": 0.5396983623504639, "learning_rate": 9.894295613049065e-05, "loss": 0.0672, "step": 11230 }, { "grad_norm": 0.3362715244293213, "learning_rate": 9.893957151971649e-05, "loss": 0.0612, "step": 11240 }, { "grad_norm": 0.39695993065834045, "learning_rate": 9.893618155698436e-05, "loss": 0.058, "step": 11250 }, { "grad_norm": 0.47230425477027893, "learning_rate": 9.8932786242665e-05, "loss": 0.0673, "step": 11260 }, { "grad_norm": 0.36414945125579834, "learning_rate": 9.89293855771297e-05, "loss": 0.0698, "step": 11270 }, { "grad_norm": 0.38938266038894653, "learning_rate": 9.892597956075036e-05, "loss": 0.064, "step": 11280 }, { "grad_norm": 0.36198773980140686, "learning_rate": 9.892256819389947e-05, "loss": 0.0623, "step": 11290 }, { "grad_norm": 0.4328870177268982, "learning_rate": 9.891915147695006e-05, "loss": 0.0677, "step": 11300 }, { "grad_norm": 0.4356273412704468, "learning_rate": 9.891572941027577e-05, "loss": 0.0816, "step": 11310 }, { "grad_norm": 0.5389734506607056, "learning_rate": 9.89123019942509e-05, "loss": 0.0814, "step": 11320 }, { "grad_norm": 0.3881455659866333, "learning_rate": 9.89088692292502e-05, "loss": 0.061, "step": 11330 }, { "grad_norm": 0.6969891786575317, "learning_rate": 9.89054311156491e-05, "loss": 0.0648, "step": 11340 }, { "grad_norm": 0.33605989813804626, "learning_rate": 9.890198765382357e-05, "loss": 0.0665, "step": 11350 }, { "grad_norm": 0.44363394379615784, "learning_rate": 9.889853884415021e-05, "loss": 0.0612, "step": 11360 }, { "grad_norm": 0.3984759449958801, "learning_rate": 9.889508468700614e-05, "loss": 0.0623, "step": 11370 }, { "grad_norm": 0.5861871838569641, "learning_rate": 9.889162518276915e-05, "loss": 0.0643, "step": 11380 }, { "grad_norm": 0.45428577065467834, "learning_rate": 9.888816033181752e-05, "loss": 0.0698, "step": 11390 }, { "grad_norm": 0.39497870206832886, "learning_rate": 9.888469013453018e-05, "loss": 0.0684, "step": 11400 }, { "grad_norm": 0.5325115323066711, "learning_rate": 9.888121459128663e-05, "loss": 0.0616, "step": 11410 }, { "grad_norm": 0.6778973937034607, "learning_rate": 9.887773370246693e-05, "loss": 0.0728, "step": 11420 }, { "grad_norm": 0.4099316895008087, "learning_rate": 9.887424746845177e-05, "loss": 0.0665, "step": 11430 }, { "grad_norm": 0.3812906742095947, "learning_rate": 9.887075588962239e-05, "loss": 0.0707, "step": 11440 }, { "grad_norm": 0.39993321895599365, "learning_rate": 9.88672589663606e-05, "loss": 0.0619, "step": 11450 }, { "grad_norm": 0.49079573154449463, "learning_rate": 9.886375669904886e-05, "loss": 0.0668, "step": 11460 }, { "grad_norm": 0.4364258050918579, "learning_rate": 9.886024908807014e-05, "loss": 0.0647, "step": 11470 }, { "grad_norm": 0.5505965352058411, "learning_rate": 9.885673613380806e-05, "loss": 0.0653, "step": 11480 }, { "grad_norm": 0.38114920258522034, "learning_rate": 9.885321783664676e-05, "loss": 0.0721, "step": 11490 }, { "grad_norm": 0.4062410295009613, "learning_rate": 9.884969419697101e-05, "loss": 0.0706, "step": 11500 }, { "grad_norm": 0.40405523777008057, "learning_rate": 9.884616521516614e-05, "loss": 0.0639, "step": 11510 }, { "grad_norm": 0.9836177825927734, "learning_rate": 9.88426308916181e-05, "loss": 0.0622, "step": 11520 }, { "grad_norm": 0.5449181795120239, "learning_rate": 9.883909122671335e-05, "loss": 0.0696, "step": 11530 }, { "grad_norm": 0.4582502841949463, "learning_rate": 9.883554622083904e-05, "loss": 0.0766, "step": 11540 }, { "grad_norm": 0.43563589453697205, "learning_rate": 9.88319958743828e-05, "loss": 0.0765, "step": 11550 }, { "grad_norm": 0.4499087333679199, "learning_rate": 9.882844018773291e-05, "loss": 0.0657, "step": 11560 }, { "grad_norm": 0.4854765832424164, "learning_rate": 9.882487916127823e-05, "loss": 0.072, "step": 11570 }, { "grad_norm": 0.5668764114379883, "learning_rate": 9.882131279540815e-05, "loss": 0.0741, "step": 11580 }, { "grad_norm": 0.4403049647808075, "learning_rate": 9.881774109051271e-05, "loss": 0.0604, "step": 11590 }, { "grad_norm": 0.5448663234710693, "learning_rate": 9.881416404698252e-05, "loss": 0.0705, "step": 11600 }, { "grad_norm": 0.4020233750343323, "learning_rate": 9.881058166520873e-05, "loss": 0.0702, "step": 11610 }, { "grad_norm": 0.44705331325531006, "learning_rate": 9.880699394558311e-05, "loss": 0.0731, "step": 11620 }, { "grad_norm": 0.4896434545516968, "learning_rate": 9.880340088849801e-05, "loss": 0.0684, "step": 11630 }, { "grad_norm": 0.39256998896598816, "learning_rate": 9.879980249434637e-05, "loss": 0.0668, "step": 11640 }, { "grad_norm": 0.4587419629096985, "learning_rate": 9.879619876352168e-05, "loss": 0.07, "step": 11650 }, { "grad_norm": 0.5491181015968323, "learning_rate": 9.879258969641809e-05, "loss": 0.0768, "step": 11660 }, { "grad_norm": 0.4028148055076599, "learning_rate": 9.878897529343023e-05, "loss": 0.0672, "step": 11670 }, { "grad_norm": 0.5671609044075012, "learning_rate": 9.878535555495338e-05, "loss": 0.0688, "step": 11680 }, { "grad_norm": 0.5767956972122192, "learning_rate": 9.87817304813834e-05, "loss": 0.0699, "step": 11690 }, { "grad_norm": 0.602375328540802, "learning_rate": 9.877810007311671e-05, "loss": 0.0779, "step": 11700 }, { "grad_norm": 0.6168788075447083, "learning_rate": 9.877446433055035e-05, "loss": 0.073, "step": 11710 }, { "grad_norm": 0.5811372399330139, "learning_rate": 9.877082325408191e-05, "loss": 0.0767, "step": 11720 }, { "grad_norm": 0.39892491698265076, "learning_rate": 9.876717684410954e-05, "loss": 0.0721, "step": 11730 }, { "grad_norm": 0.40543612837791443, "learning_rate": 9.876352510103204e-05, "loss": 0.078, "step": 11740 }, { "grad_norm": 0.4547433853149414, "learning_rate": 9.875986802524875e-05, "loss": 0.0693, "step": 11750 }, { "grad_norm": 0.5166023373603821, "learning_rate": 9.87562056171596e-05, "loss": 0.0683, "step": 11760 }, { "grad_norm": 0.5788986682891846, "learning_rate": 9.875253787716511e-05, "loss": 0.0687, "step": 11770 }, { "grad_norm": 0.5857823491096497, "learning_rate": 9.874886480566637e-05, "loss": 0.0715, "step": 11780 }, { "grad_norm": 0.43610668182373047, "learning_rate": 9.874518640306507e-05, "loss": 0.0688, "step": 11790 }, { "grad_norm": 0.4866793751716614, "learning_rate": 9.874150266976347e-05, "loss": 0.0684, "step": 11800 }, { "grad_norm": 0.47100940346717834, "learning_rate": 9.873781360616443e-05, "loss": 0.0707, "step": 11810 }, { "grad_norm": 0.5033420920372009, "learning_rate": 9.873411921267137e-05, "loss": 0.0684, "step": 11820 }, { "grad_norm": 0.5430794954299927, "learning_rate": 9.873041948968829e-05, "loss": 0.0651, "step": 11830 }, { "grad_norm": 0.3278644382953644, "learning_rate": 9.872671443761981e-05, "loss": 0.0693, "step": 11840 }, { "grad_norm": 0.6554286479949951, "learning_rate": 9.872300405687109e-05, "loss": 0.0675, "step": 11850 }, { "grad_norm": 0.49980324506759644, "learning_rate": 9.871928834784792e-05, "loss": 0.0643, "step": 11860 }, { "grad_norm": 0.5515039563179016, "learning_rate": 9.871556731095661e-05, "loss": 0.0731, "step": 11870 }, { "grad_norm": 0.7402591109275818, "learning_rate": 9.871184094660411e-05, "loss": 0.0658, "step": 11880 }, { "grad_norm": 0.608961284160614, "learning_rate": 9.870810925519791e-05, "loss": 0.0645, "step": 11890 }, { "grad_norm": 0.43403905630111694, "learning_rate": 9.870437223714612e-05, "loss": 0.0626, "step": 11900 }, { "grad_norm": 0.4247284531593323, "learning_rate": 9.87006298928574e-05, "loss": 0.0604, "step": 11910 }, { "grad_norm": 0.4753005802631378, "learning_rate": 9.869688222274103e-05, "loss": 0.0721, "step": 11920 }, { "grad_norm": 0.5217421054840088, "learning_rate": 9.869312922720681e-05, "loss": 0.0668, "step": 11930 }, { "grad_norm": 0.4495072364807129, "learning_rate": 9.868937090666521e-05, "loss": 0.0654, "step": 11940 }, { "grad_norm": 0.445295512676239, "learning_rate": 9.86856072615272e-05, "loss": 0.0709, "step": 11950 }, { "grad_norm": 0.3551023304462433, "learning_rate": 9.868183829220438e-05, "loss": 0.0631, "step": 11960 }, { "grad_norm": 0.3645559847354889, "learning_rate": 9.867806399910893e-05, "loss": 0.0734, "step": 11970 }, { "grad_norm": 0.6009150743484497, "learning_rate": 9.867428438265356e-05, "loss": 0.0638, "step": 11980 }, { "grad_norm": 0.3977097272872925, "learning_rate": 9.867049944325165e-05, "loss": 0.0624, "step": 11990 }, { "grad_norm": 0.4387725591659546, "learning_rate": 9.86667091813171e-05, "loss": 0.0756, "step": 12000 }, { "grad_norm": 0.3599504828453064, "learning_rate": 9.866291359726438e-05, "loss": 0.0583, "step": 12010 }, { "grad_norm": 0.5171496272087097, "learning_rate": 9.865911269150861e-05, "loss": 0.0707, "step": 12020 }, { "grad_norm": 0.5998235940933228, "learning_rate": 9.865530646446544e-05, "loss": 0.0675, "step": 12030 }, { "grad_norm": 0.5691321492195129, "learning_rate": 9.86514949165511e-05, "loss": 0.0686, "step": 12040 }, { "grad_norm": 0.3304007053375244, "learning_rate": 9.864767804818243e-05, "loss": 0.0611, "step": 12050 }, { "grad_norm": 0.3677727282047272, "learning_rate": 9.86438558597768e-05, "loss": 0.0667, "step": 12060 }, { "grad_norm": 0.40550699830055237, "learning_rate": 9.864002835175225e-05, "loss": 0.0623, "step": 12070 }, { "grad_norm": 0.5360087156295776, "learning_rate": 9.863619552452734e-05, "loss": 0.0663, "step": 12080 }, { "grad_norm": 0.4053652584552765, "learning_rate": 9.863235737852119e-05, "loss": 0.0749, "step": 12090 }, { "grad_norm": 0.33762067556381226, "learning_rate": 9.862851391415356e-05, "loss": 0.0695, "step": 12100 }, { "grad_norm": 0.5900738835334778, "learning_rate": 9.862466513184477e-05, "loss": 0.064, "step": 12110 }, { "grad_norm": 0.43325307965278625, "learning_rate": 9.86208110320157e-05, "loss": 0.0711, "step": 12120 }, { "grad_norm": 0.48278534412384033, "learning_rate": 9.861695161508784e-05, "loss": 0.0661, "step": 12130 }, { "grad_norm": 0.42874208092689514, "learning_rate": 9.861308688148324e-05, "loss": 0.0602, "step": 12140 }, { "grad_norm": 0.5179710388183594, "learning_rate": 9.860921683162455e-05, "loss": 0.0786, "step": 12150 }, { "grad_norm": 0.45628973841667175, "learning_rate": 9.860534146593499e-05, "loss": 0.0655, "step": 12160 }, { "grad_norm": 0.42387640476226807, "learning_rate": 9.860146078483836e-05, "loss": 0.0618, "step": 12170 }, { "grad_norm": 0.34170159697532654, "learning_rate": 9.859757478875905e-05, "loss": 0.0706, "step": 12180 }, { "grad_norm": 0.5217971801757812, "learning_rate": 9.859368347812204e-05, "loss": 0.0611, "step": 12190 }, { "grad_norm": 0.3842659890651703, "learning_rate": 9.858978685335285e-05, "loss": 0.0752, "step": 12200 }, { "grad_norm": 0.3122444748878479, "learning_rate": 9.858588491487763e-05, "loss": 0.0646, "step": 12210 }, { "grad_norm": 0.43298423290252686, "learning_rate": 9.858197766312308e-05, "loss": 0.0628, "step": 12220 }, { "grad_norm": 0.5064117312431335, "learning_rate": 9.857806509851649e-05, "loss": 0.0769, "step": 12230 }, { "grad_norm": 0.6011719107627869, "learning_rate": 9.857414722148574e-05, "loss": 0.0739, "step": 12240 }, { "grad_norm": 0.5044312477111816, "learning_rate": 9.857022403245928e-05, "loss": 0.0671, "step": 12250 }, { "grad_norm": 0.5445764064788818, "learning_rate": 9.856629553186615e-05, "loss": 0.0749, "step": 12260 }, { "grad_norm": 0.6686264872550964, "learning_rate": 9.856236172013595e-05, "loss": 0.0759, "step": 12270 }, { "grad_norm": 0.3893170952796936, "learning_rate": 9.85584225976989e-05, "loss": 0.0681, "step": 12280 }, { "grad_norm": 0.4490833878517151, "learning_rate": 9.855447816498575e-05, "loss": 0.0681, "step": 12290 }, { "grad_norm": 0.6338330507278442, "learning_rate": 9.855052842242787e-05, "loss": 0.0729, "step": 12300 }, { "grad_norm": 0.35451823472976685, "learning_rate": 9.85465733704572e-05, "loss": 0.0718, "step": 12310 }, { "grad_norm": 0.3750683665275574, "learning_rate": 9.854261300950624e-05, "loss": 0.0674, "step": 12320 }, { "grad_norm": 0.44083890318870544, "learning_rate": 9.853864734000813e-05, "loss": 0.068, "step": 12330 }, { "grad_norm": 0.5201770663261414, "learning_rate": 9.85346763623965e-05, "loss": 0.0738, "step": 12340 }, { "grad_norm": 0.4626065790653229, "learning_rate": 9.853070007710564e-05, "loss": 0.0696, "step": 12350 }, { "grad_norm": 0.9875331521034241, "learning_rate": 9.85267184845704e-05, "loss": 0.0671, "step": 12360 }, { "grad_norm": 0.526080310344696, "learning_rate": 9.852273158522616e-05, "loss": 0.0566, "step": 12370 }, { "grad_norm": 0.5025755763053894, "learning_rate": 9.851873937950896e-05, "loss": 0.068, "step": 12380 }, { "grad_norm": 0.4985204041004181, "learning_rate": 9.851474186785537e-05, "loss": 0.065, "step": 12390 }, { "grad_norm": 0.5060468316078186, "learning_rate": 9.851073905070254e-05, "loss": 0.0761, "step": 12400 }, { "grad_norm": 0.35571902990341187, "learning_rate": 9.850673092848824e-05, "loss": 0.063, "step": 12410 }, { "grad_norm": 0.418035626411438, "learning_rate": 9.850271750165077e-05, "loss": 0.0621, "step": 12420 }, { "grad_norm": 0.4149559736251831, "learning_rate": 9.849869877062902e-05, "loss": 0.0627, "step": 12430 }, { "grad_norm": 0.522647500038147, "learning_rate": 9.849467473586252e-05, "loss": 0.0641, "step": 12440 }, { "grad_norm": 0.4632730185985565, "learning_rate": 9.849064539779127e-05, "loss": 0.0664, "step": 12450 }, { "grad_norm": 0.47918716073036194, "learning_rate": 9.848661075685594e-05, "loss": 0.0644, "step": 12460 }, { "grad_norm": 0.49777787923812866, "learning_rate": 9.848257081349778e-05, "loss": 0.0678, "step": 12470 }, { "grad_norm": 0.3099057972431183, "learning_rate": 9.847852556815856e-05, "loss": 0.0608, "step": 12480 }, { "grad_norm": 0.5353414416313171, "learning_rate": 9.847447502128067e-05, "loss": 0.0788, "step": 12490 }, { "grad_norm": 0.5212224125862122, "learning_rate": 9.847041917330708e-05, "loss": 0.0705, "step": 12500 }, { "grad_norm": 0.5792914032936096, "learning_rate": 9.846635802468132e-05, "loss": 0.0781, "step": 12510 }, { "grad_norm": 0.44767606258392334, "learning_rate": 9.84622915758475e-05, "loss": 0.0691, "step": 12520 }, { "grad_norm": 0.5001344680786133, "learning_rate": 9.845821982725034e-05, "loss": 0.062, "step": 12530 }, { "grad_norm": 0.34441691637039185, "learning_rate": 9.845414277933514e-05, "loss": 0.0608, "step": 12540 }, { "grad_norm": 0.42942333221435547, "learning_rate": 9.845006043254771e-05, "loss": 0.0605, "step": 12550 }, { "grad_norm": 0.4565560221672058, "learning_rate": 9.844597278733451e-05, "loss": 0.0698, "step": 12560 }, { "grad_norm": 0.3684162497520447, "learning_rate": 9.844187984414259e-05, "loss": 0.0678, "step": 12570 }, { "grad_norm": 0.4062501788139343, "learning_rate": 9.84377816034195e-05, "loss": 0.0614, "step": 12580 }, { "grad_norm": 0.3809932470321655, "learning_rate": 9.843367806561345e-05, "loss": 0.0713, "step": 12590 }, { "grad_norm": 0.37752726674079895, "learning_rate": 9.842956923117317e-05, "loss": 0.0653, "step": 12600 }, { "grad_norm": 0.4324024021625519, "learning_rate": 9.842545510054802e-05, "loss": 0.0695, "step": 12610 }, { "grad_norm": 0.5766841769218445, "learning_rate": 9.842133567418792e-05, "loss": 0.0694, "step": 12620 }, { "grad_norm": 0.5597012639045715, "learning_rate": 9.841721095254333e-05, "loss": 0.0718, "step": 12630 }, { "grad_norm": 0.4439977705478668, "learning_rate": 9.841308093606537e-05, "loss": 0.0661, "step": 12640 }, { "grad_norm": 0.5190489888191223, "learning_rate": 9.840894562520565e-05, "loss": 0.0696, "step": 12650 }, { "grad_norm": 0.4447062611579895, "learning_rate": 9.840480502041642e-05, "loss": 0.0646, "step": 12660 }, { "grad_norm": 0.33920761942863464, "learning_rate": 9.840065912215049e-05, "loss": 0.0681, "step": 12670 }, { "grad_norm": 0.49982887506484985, "learning_rate": 9.839650793086124e-05, "loss": 0.0605, "step": 12680 }, { "grad_norm": 0.558091938495636, "learning_rate": 9.839235144700265e-05, "loss": 0.0721, "step": 12690 }, { "grad_norm": 0.3920080363750458, "learning_rate": 9.838818967102926e-05, "loss": 0.0702, "step": 12700 }, { "grad_norm": 0.33805230259895325, "learning_rate": 9.83840226033962e-05, "loss": 0.0692, "step": 12710 }, { "grad_norm": 0.6406195759773254, "learning_rate": 9.837985024455918e-05, "loss": 0.074, "step": 12720 }, { "grad_norm": 0.43506044149398804, "learning_rate": 9.837567259497447e-05, "loss": 0.0745, "step": 12730 }, { "grad_norm": 0.4061010777950287, "learning_rate": 9.837148965509894e-05, "loss": 0.0644, "step": 12740 }, { "grad_norm": 0.5214067101478577, "learning_rate": 9.836730142539001e-05, "loss": 0.0682, "step": 12750 }, { "grad_norm": 0.3626599609851837, "learning_rate": 9.836310790630574e-05, "loss": 0.0668, "step": 12760 }, { "grad_norm": 0.4252278208732605, "learning_rate": 9.83589090983047e-05, "loss": 0.066, "step": 12770 }, { "grad_norm": 0.4224216938018799, "learning_rate": 9.835470500184605e-05, "loss": 0.0612, "step": 12780 }, { "grad_norm": 0.3478967249393463, "learning_rate": 9.835049561738957e-05, "loss": 0.0535, "step": 12790 }, { "grad_norm": 0.5390461087226868, "learning_rate": 9.834628094539558e-05, "loss": 0.0671, "step": 12800 }, { "grad_norm": 0.5321718454360962, "learning_rate": 9.834206098632499e-05, "loss": 0.0716, "step": 12810 }, { "grad_norm": 0.4239516258239746, "learning_rate": 9.833783574063931e-05, "loss": 0.0745, "step": 12820 }, { "grad_norm": 0.3795435428619385, "learning_rate": 9.833360520880058e-05, "loss": 0.0659, "step": 12830 }, { "grad_norm": 0.43361422419548035, "learning_rate": 9.832936939127144e-05, "loss": 0.0664, "step": 12840 }, { "grad_norm": 0.39543986320495605, "learning_rate": 9.832512828851515e-05, "loss": 0.0669, "step": 12850 }, { "grad_norm": 0.7293070554733276, "learning_rate": 9.832088190099546e-05, "loss": 0.0677, "step": 12860 }, { "grad_norm": 0.42321205139160156, "learning_rate": 9.831663022917679e-05, "loss": 0.0589, "step": 12870 }, { "grad_norm": 0.3822060227394104, "learning_rate": 9.831237327352407e-05, "loss": 0.0668, "step": 12880 }, { "grad_norm": 0.3712329566478729, "learning_rate": 9.830811103450286e-05, "loss": 0.0588, "step": 12890 }, { "grad_norm": 0.7440100908279419, "learning_rate": 9.830384351257924e-05, "loss": 0.0724, "step": 12900 }, { "grad_norm": 0.44049280881881714, "learning_rate": 9.829957070821993e-05, "loss": 0.0634, "step": 12910 }, { "grad_norm": 0.4364216923713684, "learning_rate": 9.829529262189218e-05, "loss": 0.0713, "step": 12920 }, { "grad_norm": 0.38634970784187317, "learning_rate": 9.829100925406385e-05, "loss": 0.0631, "step": 12930 }, { "grad_norm": 0.5322594046592712, "learning_rate": 9.828672060520333e-05, "loss": 0.0659, "step": 12940 }, { "grad_norm": 0.39533039927482605, "learning_rate": 9.828242667577966e-05, "loss": 0.0793, "step": 12950 }, { "grad_norm": 0.41352081298828125, "learning_rate": 9.82781274662624e-05, "loss": 0.0736, "step": 12960 }, { "grad_norm": 0.6418298482894897, "learning_rate": 9.82738229771217e-05, "loss": 0.0663, "step": 12970 }, { "grad_norm": 0.2954692244529724, "learning_rate": 9.826951320882829e-05, "loss": 0.0628, "step": 12980 }, { "grad_norm": 0.5636436343193054, "learning_rate": 9.826519816185351e-05, "loss": 0.0694, "step": 12990 }, { "grad_norm": 0.48157426714897156, "learning_rate": 9.826087783666921e-05, "loss": 0.0655, "step": 13000 }, { "grad_norm": 0.44827455282211304, "learning_rate": 9.825655223374787e-05, "loss": 0.0668, "step": 13010 }, { "grad_norm": 0.39310380816459656, "learning_rate": 9.825222135356253e-05, "loss": 0.0713, "step": 13020 }, { "grad_norm": 0.44324979186058044, "learning_rate": 9.82478851965868e-05, "loss": 0.0692, "step": 13030 }, { "grad_norm": 0.4462292194366455, "learning_rate": 9.82435437632949e-05, "loss": 0.0619, "step": 13040 }, { "grad_norm": 0.44644245505332947, "learning_rate": 9.823919705416158e-05, "loss": 0.0707, "step": 13050 }, { "grad_norm": 0.4919774532318115, "learning_rate": 9.82348450696622e-05, "loss": 0.0559, "step": 13060 }, { "grad_norm": 0.5092033743858337, "learning_rate": 9.823048781027268e-05, "loss": 0.0677, "step": 13070 }, { "grad_norm": 0.7659698128700256, "learning_rate": 9.822612527646953e-05, "loss": 0.0637, "step": 13080 }, { "grad_norm": 0.48619988560676575, "learning_rate": 9.822175746872984e-05, "loss": 0.0628, "step": 13090 }, { "grad_norm": 0.4203673303127289, "learning_rate": 9.821738438753123e-05, "loss": 0.0693, "step": 13100 }, { "grad_norm": 0.3543637692928314, "learning_rate": 9.821300603335196e-05, "loss": 0.0587, "step": 13110 }, { "grad_norm": 0.46084919571876526, "learning_rate": 9.820862240667085e-05, "loss": 0.0684, "step": 13120 }, { "grad_norm": 0.4793654978275299, "learning_rate": 9.820423350796726e-05, "loss": 0.0675, "step": 13130 }, { "grad_norm": 0.4206366539001465, "learning_rate": 9.819983933772118e-05, "loss": 0.0689, "step": 13140 }, { "grad_norm": 0.3855423927307129, "learning_rate": 9.819543989641314e-05, "loss": 0.0685, "step": 13150 }, { "grad_norm": 0.49029701948165894, "learning_rate": 9.819103518452423e-05, "loss": 0.0575, "step": 13160 }, { "grad_norm": 0.5272868871688843, "learning_rate": 9.818662520253618e-05, "loss": 0.0678, "step": 13170 }, { "grad_norm": 0.3670617640018463, "learning_rate": 9.818220995093126e-05, "loss": 0.0654, "step": 13180 }, { "grad_norm": 0.527320146560669, "learning_rate": 9.817778943019228e-05, "loss": 0.0624, "step": 13190 }, { "grad_norm": 0.3871407210826874, "learning_rate": 9.81733636408027e-05, "loss": 0.0668, "step": 13200 }, { "grad_norm": 0.3462892770767212, "learning_rate": 9.816893258324649e-05, "loss": 0.0568, "step": 13210 }, { "grad_norm": 0.6880624294281006, "learning_rate": 9.816449625800823e-05, "loss": 0.0758, "step": 13220 }, { "grad_norm": 0.5465054512023926, "learning_rate": 9.816005466557308e-05, "loss": 0.0633, "step": 13230 }, { "grad_norm": 0.41620850563049316, "learning_rate": 9.815560780642674e-05, "loss": 0.0555, "step": 13240 }, { "grad_norm": 0.45612460374832153, "learning_rate": 9.815115568105555e-05, "loss": 0.0703, "step": 13250 }, { "grad_norm": 0.35622525215148926, "learning_rate": 9.814669828994638e-05, "loss": 0.0629, "step": 13260 }, { "grad_norm": 0.4929277300834656, "learning_rate": 9.814223563358665e-05, "loss": 0.0605, "step": 13270 }, { "grad_norm": 0.4797976016998291, "learning_rate": 9.813776771246443e-05, "loss": 0.0612, "step": 13280 }, { "grad_norm": 0.541360080242157, "learning_rate": 9.813329452706829e-05, "loss": 0.056, "step": 13290 }, { "grad_norm": 1.175209403038025, "learning_rate": 9.812881607788744e-05, "loss": 0.064, "step": 13300 }, { "grad_norm": 0.4088849127292633, "learning_rate": 9.812433236541163e-05, "loss": 0.068, "step": 13310 }, { "grad_norm": 0.404744416475296, "learning_rate": 9.811984339013116e-05, "loss": 0.0608, "step": 13320 }, { "grad_norm": 0.47555047273635864, "learning_rate": 9.811534915253698e-05, "loss": 0.0639, "step": 13330 }, { "grad_norm": 0.314974844455719, "learning_rate": 9.811084965312056e-05, "loss": 0.0609, "step": 13340 }, { "grad_norm": 0.43998613953590393, "learning_rate": 9.810634489237396e-05, "loss": 0.0606, "step": 13350 }, { "grad_norm": 0.3282676339149475, "learning_rate": 9.81018348707898e-05, "loss": 0.0593, "step": 13360 }, { "grad_norm": 0.40648579597473145, "learning_rate": 9.809731958886131e-05, "loss": 0.0636, "step": 13370 }, { "grad_norm": 0.4170915484428406, "learning_rate": 9.809279904708224e-05, "loss": 0.062, "step": 13380 }, { "grad_norm": 0.3718263804912567, "learning_rate": 9.808827324594699e-05, "loss": 0.062, "step": 13390 }, { "grad_norm": 0.3799842596054077, "learning_rate": 9.808374218595046e-05, "loss": 0.0693, "step": 13400 }, { "grad_norm": 0.44434672594070435, "learning_rate": 9.80792058675882e-05, "loss": 0.0677, "step": 13410 }, { "grad_norm": 0.6998065710067749, "learning_rate": 9.807466429135627e-05, "loss": 0.0744, "step": 13420 }, { "grad_norm": 0.455011248588562, "learning_rate": 9.807011745775132e-05, "loss": 0.0646, "step": 13430 }, { "grad_norm": 0.4958494305610657, "learning_rate": 9.806556536727061e-05, "loss": 0.0636, "step": 13440 }, { "grad_norm": 0.5276517271995544, "learning_rate": 9.806100802041193e-05, "loss": 0.0789, "step": 13450 }, { "grad_norm": 0.5166394710540771, "learning_rate": 9.805644541767368e-05, "loss": 0.0711, "step": 13460 }, { "grad_norm": 0.5179041624069214, "learning_rate": 9.805187755955478e-05, "loss": 0.0727, "step": 13470 }, { "grad_norm": 0.38710513710975647, "learning_rate": 9.804730444655483e-05, "loss": 0.0649, "step": 13480 }, { "grad_norm": 0.5532407760620117, "learning_rate": 9.804272607917388e-05, "loss": 0.0647, "step": 13490 }, { "grad_norm": 0.37691745162010193, "learning_rate": 9.803814245791265e-05, "loss": 0.0684, "step": 13500 }, { "grad_norm": 0.4774869680404663, "learning_rate": 9.803355358327239e-05, "loss": 0.0748, "step": 13510 }, { "grad_norm": 0.4126184284687042, "learning_rate": 9.802895945575492e-05, "loss": 0.0651, "step": 13520 }, { "grad_norm": 0.8155035972595215, "learning_rate": 9.802436007586266e-05, "loss": 0.0663, "step": 13530 }, { "grad_norm": 0.5104736089706421, "learning_rate": 9.801975544409858e-05, "loss": 0.061, "step": 13540 }, { "grad_norm": 0.40406185388565063, "learning_rate": 9.801514556096625e-05, "loss": 0.0626, "step": 13550 }, { "grad_norm": 0.3226799964904785, "learning_rate": 9.801053042696977e-05, "loss": 0.0662, "step": 13560 }, { "grad_norm": 0.413339227437973, "learning_rate": 9.800591004261388e-05, "loss": 0.0608, "step": 13570 }, { "grad_norm": 0.3738732635974884, "learning_rate": 9.800128440840385e-05, "loss": 0.0644, "step": 13580 }, { "grad_norm": 0.3981686532497406, "learning_rate": 9.799665352484552e-05, "loss": 0.0602, "step": 13590 }, { "grad_norm": 0.5207033753395081, "learning_rate": 9.799201739244532e-05, "loss": 0.0668, "step": 13600 }, { "grad_norm": 0.49098649621009827, "learning_rate": 9.798737601171025e-05, "loss": 0.0641, "step": 13610 }, { "grad_norm": 0.42051804065704346, "learning_rate": 9.79827293831479e-05, "loss": 0.0622, "step": 13620 }, { "grad_norm": 0.43767601251602173, "learning_rate": 9.797807750726638e-05, "loss": 0.0689, "step": 13630 }, { "grad_norm": 0.4081884026527405, "learning_rate": 9.797342038457446e-05, "loss": 0.0688, "step": 13640 }, { "grad_norm": 0.4729579985141754, "learning_rate": 9.796875801558141e-05, "loss": 0.07, "step": 13650 }, { "grad_norm": 0.4509713053703308, "learning_rate": 9.79640904007971e-05, "loss": 0.0669, "step": 13660 }, { "grad_norm": 0.3380942940711975, "learning_rate": 9.795941754073199e-05, "loss": 0.0612, "step": 13670 }, { "grad_norm": 0.49110716581344604, "learning_rate": 9.795473943589705e-05, "loss": 0.0696, "step": 13680 }, { "grad_norm": 0.6275022625923157, "learning_rate": 9.795005608680394e-05, "loss": 0.0727, "step": 13690 }, { "grad_norm": 0.6162601709365845, "learning_rate": 9.794536749396477e-05, "loss": 0.0821, "step": 13700 }, { "grad_norm": 0.37300732731819153, "learning_rate": 9.79406736578923e-05, "loss": 0.0703, "step": 13710 }, { "grad_norm": 0.4770631790161133, "learning_rate": 9.793597457909984e-05, "loss": 0.0692, "step": 13720 }, { "grad_norm": 0.5923783183097839, "learning_rate": 9.793127025810127e-05, "loss": 0.0651, "step": 13730 }, { "grad_norm": 0.47790664434432983, "learning_rate": 9.792656069541104e-05, "loss": 0.0609, "step": 13740 }, { "grad_norm": 0.47706177830696106, "learning_rate": 9.79218458915442e-05, "loss": 0.0684, "step": 13750 }, { "grad_norm": 0.3628493845462799, "learning_rate": 9.791712584701634e-05, "loss": 0.0744, "step": 13760 }, { "grad_norm": 0.54957115650177, "learning_rate": 9.791240056234364e-05, "loss": 0.0668, "step": 13770 }, { "grad_norm": 0.404175341129303, "learning_rate": 9.790767003804283e-05, "loss": 0.0617, "step": 13780 }, { "grad_norm": 0.3771282136440277, "learning_rate": 9.790293427463126e-05, "loss": 0.0662, "step": 13790 }, { "grad_norm": 0.5618253350257874, "learning_rate": 9.789819327262684e-05, "loss": 0.0737, "step": 13800 }, { "grad_norm": 0.6487759947776794, "learning_rate": 9.7893447032548e-05, "loss": 0.0671, "step": 13810 }, { "grad_norm": 0.3302943706512451, "learning_rate": 9.78886955549138e-05, "loss": 0.0661, "step": 13820 }, { "grad_norm": 0.4533965587615967, "learning_rate": 9.788393884024387e-05, "loss": 0.0679, "step": 13830 }, { "grad_norm": 0.4235398769378662, "learning_rate": 9.787917688905836e-05, "loss": 0.0634, "step": 13840 }, { "grad_norm": 0.4629920721054077, "learning_rate": 9.787440970187807e-05, "loss": 0.0552, "step": 13850 }, { "grad_norm": 0.38609349727630615, "learning_rate": 9.786963727922429e-05, "loss": 0.0642, "step": 13860 }, { "grad_norm": 0.4493330121040344, "learning_rate": 9.786485962161897e-05, "loss": 0.0551, "step": 13870 }, { "grad_norm": 0.40614446997642517, "learning_rate": 9.786007672958455e-05, "loss": 0.0623, "step": 13880 }, { "grad_norm": 0.4204546809196472, "learning_rate": 9.78552886036441e-05, "loss": 0.0636, "step": 13890 }, { "grad_norm": 0.48063400387763977, "learning_rate": 9.785049524432124e-05, "loss": 0.0619, "step": 13900 }, { "grad_norm": 0.4516896605491638, "learning_rate": 9.784569665214016e-05, "loss": 0.0665, "step": 13910 }, { "grad_norm": 0.7505221962928772, "learning_rate": 9.784089282762563e-05, "loss": 0.0648, "step": 13920 }, { "grad_norm": 0.614343523979187, "learning_rate": 9.7836083771303e-05, "loss": 0.0732, "step": 13930 }, { "grad_norm": 0.3806088864803314, "learning_rate": 9.783126948369817e-05, "loss": 0.0587, "step": 13940 }, { "grad_norm": 0.29744312167167664, "learning_rate": 9.78264499653376e-05, "loss": 0.0618, "step": 13950 }, { "grad_norm": 0.5293594598770142, "learning_rate": 9.782162521674838e-05, "loss": 0.0673, "step": 13960 }, { "grad_norm": 0.5151064991950989, "learning_rate": 9.781679523845812e-05, "loss": 0.0613, "step": 13970 }, { "grad_norm": 0.3402988612651825, "learning_rate": 9.781196003099502e-05, "loss": 0.0653, "step": 13980 }, { "grad_norm": 0.4198424220085144, "learning_rate": 9.780711959488786e-05, "loss": 0.0665, "step": 13990 }, { "grad_norm": 0.4330315887928009, "learning_rate": 9.780227393066599e-05, "loss": 0.0734, "step": 14000 }, { "grad_norm": 0.37746161222457886, "learning_rate": 9.77974230388593e-05, "loss": 0.0691, "step": 14010 }, { "grad_norm": 0.4026181995868683, "learning_rate": 9.779256691999829e-05, "loss": 0.066, "step": 14020 }, { "grad_norm": 0.3503836989402771, "learning_rate": 9.778770557461403e-05, "loss": 0.0598, "step": 14030 }, { "grad_norm": 0.43374159932136536, "learning_rate": 9.778283900323812e-05, "loss": 0.0647, "step": 14040 }, { "grad_norm": 0.44663679599761963, "learning_rate": 9.777796720640277e-05, "loss": 0.0561, "step": 14050 }, { "grad_norm": 0.529263436794281, "learning_rate": 9.777309018464078e-05, "loss": 0.0609, "step": 14060 }, { "grad_norm": 0.4279794991016388, "learning_rate": 9.776820793848547e-05, "loss": 0.0568, "step": 14070 }, { "grad_norm": 0.40025749802589417, "learning_rate": 9.776332046847075e-05, "loss": 0.0656, "step": 14080 }, { "grad_norm": 0.496141642332077, "learning_rate": 9.775842777513111e-05, "loss": 0.0631, "step": 14090 }, { "grad_norm": 0.5491344928741455, "learning_rate": 9.775352985900163e-05, "loss": 0.0619, "step": 14100 }, { "grad_norm": 0.38787877559661865, "learning_rate": 9.774862672061791e-05, "loss": 0.062, "step": 14110 }, { "grad_norm": 0.49458742141723633, "learning_rate": 9.774371836051616e-05, "loss": 0.0703, "step": 14120 }, { "grad_norm": 0.4243490695953369, "learning_rate": 9.773880477923315e-05, "loss": 0.0648, "step": 14130 }, { "grad_norm": 0.5103504061698914, "learning_rate": 9.773388597730623e-05, "loss": 0.061, "step": 14140 }, { "grad_norm": 0.4692705571651459, "learning_rate": 9.77289619552733e-05, "loss": 0.069, "step": 14150 }, { "grad_norm": 0.3535493314266205, "learning_rate": 9.772403271367285e-05, "loss": 0.0591, "step": 14160 }, { "grad_norm": 0.39904019236564636, "learning_rate": 9.771909825304396e-05, "loss": 0.0652, "step": 14170 }, { "grad_norm": 0.44244372844696045, "learning_rate": 9.771415857392619e-05, "loss": 0.0615, "step": 14180 }, { "grad_norm": 0.42359963059425354, "learning_rate": 9.770921367685978e-05, "loss": 0.0547, "step": 14190 }, { "grad_norm": 0.617170512676239, "learning_rate": 9.770426356238551e-05, "loss": 0.0591, "step": 14200 }, { "grad_norm": 0.37420856952667236, "learning_rate": 9.769930823104469e-05, "loss": 0.0678, "step": 14210 }, { "grad_norm": 0.4638821482658386, "learning_rate": 9.769434768337926e-05, "loss": 0.0624, "step": 14220 }, { "grad_norm": 0.3755790591239929, "learning_rate": 9.768938191993164e-05, "loss": 0.057, "step": 14230 }, { "grad_norm": 0.3991162180900574, "learning_rate": 9.768441094124494e-05, "loss": 0.0643, "step": 14240 }, { "grad_norm": 0.43916040658950806, "learning_rate": 9.767943474786275e-05, "loss": 0.0666, "step": 14250 }, { "grad_norm": 0.32383111119270325, "learning_rate": 9.767445334032923e-05, "loss": 0.0617, "step": 14260 }, { "grad_norm": 0.4637841582298279, "learning_rate": 9.766946671918919e-05, "loss": 0.0568, "step": 14270 }, { "grad_norm": 0.379219114780426, "learning_rate": 9.766447488498796e-05, "loss": 0.0762, "step": 14280 }, { "grad_norm": 0.3547130823135376, "learning_rate": 9.765947783827139e-05, "loss": 0.0565, "step": 14290 }, { "grad_norm": 0.432909220457077, "learning_rate": 9.765447557958599e-05, "loss": 0.0605, "step": 14300 }, { "grad_norm": 0.4524938762187958, "learning_rate": 9.764946810947879e-05, "loss": 0.0559, "step": 14310 }, { "grad_norm": 0.3926675319671631, "learning_rate": 9.764445542849738e-05, "loss": 0.0622, "step": 14320 }, { "grad_norm": 0.45668554306030273, "learning_rate": 9.763943753718998e-05, "loss": 0.0679, "step": 14330 }, { "grad_norm": 0.654726505279541, "learning_rate": 9.76344144361053e-05, "loss": 0.0573, "step": 14340 }, { "grad_norm": 0.48044174909591675, "learning_rate": 9.762938612579269e-05, "loss": 0.0635, "step": 14350 }, { "grad_norm": 0.41505682468414307, "learning_rate": 9.762435260680202e-05, "loss": 0.0639, "step": 14360 }, { "grad_norm": 1.3296575546264648, "learning_rate": 9.761931387968373e-05, "loss": 0.0652, "step": 14370 }, { "grad_norm": 0.5332413911819458, "learning_rate": 9.76142699449889e-05, "loss": 0.0585, "step": 14380 }, { "grad_norm": 0.3422377109527588, "learning_rate": 9.760922080326908e-05, "loss": 0.0612, "step": 14390 }, { "grad_norm": 1.0600333213806152, "learning_rate": 9.760416645507644e-05, "loss": 0.0659, "step": 14400 }, { "grad_norm": 0.34689387679100037, "learning_rate": 9.759910690096375e-05, "loss": 0.0629, "step": 14410 }, { "grad_norm": 0.42336681485176086, "learning_rate": 9.759404214148429e-05, "loss": 0.0622, "step": 14420 }, { "grad_norm": 0.5038828253746033, "learning_rate": 9.758897217719191e-05, "loss": 0.0661, "step": 14430 }, { "grad_norm": 0.4186626374721527, "learning_rate": 9.758389700864113e-05, "loss": 0.0636, "step": 14440 }, { "grad_norm": 0.6150849461555481, "learning_rate": 9.757881663638688e-05, "loss": 0.0685, "step": 14450 }, { "grad_norm": 0.5049463510513306, "learning_rate": 9.757373106098478e-05, "loss": 0.0706, "step": 14460 }, { "grad_norm": 0.48154687881469727, "learning_rate": 9.756864028299097e-05, "loss": 0.0613, "step": 14470 }, { "grad_norm": 0.5283561944961548, "learning_rate": 9.75635443029622e-05, "loss": 0.0669, "step": 14480 }, { "grad_norm": 0.5047129392623901, "learning_rate": 9.755844312145572e-05, "loss": 0.0622, "step": 14490 }, { "grad_norm": 0.4094431698322296, "learning_rate": 9.755333673902941e-05, "loss": 0.0638, "step": 14500 }, { "grad_norm": 0.4033535122871399, "learning_rate": 9.75482251562417e-05, "loss": 0.0637, "step": 14510 }, { "grad_norm": 0.5143373608589172, "learning_rate": 9.754310837365155e-05, "loss": 0.0585, "step": 14520 }, { "grad_norm": 0.37559473514556885, "learning_rate": 9.753798639181856e-05, "loss": 0.06, "step": 14530 }, { "grad_norm": 0.7173538208007812, "learning_rate": 9.753285921130286e-05, "loss": 0.0644, "step": 14540 }, { "grad_norm": 0.4030689001083374, "learning_rate": 9.752772683266512e-05, "loss": 0.0627, "step": 14550 }, { "grad_norm": 0.5673761963844299, "learning_rate": 9.752258925646665e-05, "loss": 0.0571, "step": 14560 }, { "grad_norm": 0.45504137873649597, "learning_rate": 9.751744648326926e-05, "loss": 0.0703, "step": 14570 }, { "grad_norm": 0.5192091464996338, "learning_rate": 9.751229851363536e-05, "loss": 0.0617, "step": 14580 }, { "grad_norm": 0.44346362352371216, "learning_rate": 9.750714534812793e-05, "loss": 0.0677, "step": 14590 }, { "grad_norm": 0.3706375062465668, "learning_rate": 9.750198698731053e-05, "loss": 0.0581, "step": 14600 }, { "grad_norm": 0.39506974816322327, "learning_rate": 9.749682343174722e-05, "loss": 0.0607, "step": 14610 }, { "grad_norm": 0.38747456669807434, "learning_rate": 9.749165468200272e-05, "loss": 0.0603, "step": 14620 }, { "grad_norm": 0.3809957504272461, "learning_rate": 9.748648073864229e-05, "loss": 0.062, "step": 14630 }, { "grad_norm": 0.4106176197528839, "learning_rate": 9.748130160223168e-05, "loss": 0.058, "step": 14640 }, { "grad_norm": 0.5035122632980347, "learning_rate": 9.747611727333734e-05, "loss": 0.0654, "step": 14650 }, { "grad_norm": 0.37683096528053284, "learning_rate": 9.74709277525262e-05, "loss": 0.0665, "step": 14660 }, { "grad_norm": 0.4772532284259796, "learning_rate": 9.746573304036576e-05, "loss": 0.0636, "step": 14670 }, { "grad_norm": 0.4259271025657654, "learning_rate": 9.746053313742412e-05, "loss": 0.0576, "step": 14680 }, { "grad_norm": 0.40543869137763977, "learning_rate": 9.745532804426994e-05, "loss": 0.0567, "step": 14690 }, { "grad_norm": 0.38261210918426514, "learning_rate": 9.745011776147242e-05, "loss": 0.0708, "step": 14700 }, { "grad_norm": 0.4590033292770386, "learning_rate": 9.744490228960138e-05, "loss": 0.0639, "step": 14710 }, { "grad_norm": 0.44151994585990906, "learning_rate": 9.743968162922713e-05, "loss": 0.0609, "step": 14720 }, { "grad_norm": 0.38339489698410034, "learning_rate": 9.743445578092064e-05, "loss": 0.0685, "step": 14730 }, { "grad_norm": 0.4817928969860077, "learning_rate": 9.742922474525338e-05, "loss": 0.061, "step": 14740 }, { "grad_norm": 0.7890418171882629, "learning_rate": 9.742398852279741e-05, "loss": 0.0645, "step": 14750 }, { "grad_norm": 0.37114813923835754, "learning_rate": 9.741874711412535e-05, "loss": 0.0651, "step": 14760 }, { "grad_norm": 0.5532239675521851, "learning_rate": 9.741350051981042e-05, "loss": 0.0631, "step": 14770 }, { "grad_norm": 0.4119279384613037, "learning_rate": 9.740824874042633e-05, "loss": 0.0665, "step": 14780 }, { "grad_norm": 0.4774208068847656, "learning_rate": 9.740299177654746e-05, "loss": 0.0579, "step": 14790 }, { "grad_norm": 0.4794325530529022, "learning_rate": 9.739772962874867e-05, "loss": 0.0636, "step": 14800 }, { "grad_norm": 0.6404914855957031, "learning_rate": 9.739246229760541e-05, "loss": 0.0708, "step": 14810 }, { "grad_norm": 0.39039525389671326, "learning_rate": 9.738718978369376e-05, "loss": 0.059, "step": 14820 }, { "grad_norm": 0.3482814133167267, "learning_rate": 9.738191208759025e-05, "loss": 0.06, "step": 14830 }, { "grad_norm": 0.4977462589740753, "learning_rate": 9.73766292098721e-05, "loss": 0.0599, "step": 14840 }, { "grad_norm": 0.38099247217178345, "learning_rate": 9.737134115111699e-05, "loss": 0.0689, "step": 14850 }, { "grad_norm": 0.6862410306930542, "learning_rate": 9.736604791190323e-05, "loss": 0.0779, "step": 14860 }, { "grad_norm": 0.39636266231536865, "learning_rate": 9.73607494928097e-05, "loss": 0.0677, "step": 14870 }, { "grad_norm": 0.475962370634079, "learning_rate": 9.735544589441581e-05, "loss": 0.0652, "step": 14880 }, { "grad_norm": 0.4436344802379608, "learning_rate": 9.735013711730154e-05, "loss": 0.0628, "step": 14890 }, { "grad_norm": 0.583672046661377, "learning_rate": 9.734482316204747e-05, "loss": 0.0684, "step": 14900 }, { "grad_norm": 0.46533653140068054, "learning_rate": 9.733950402923473e-05, "loss": 0.0623, "step": 14910 }, { "grad_norm": 0.5131239295005798, "learning_rate": 9.7334179719445e-05, "loss": 0.0688, "step": 14920 }, { "grad_norm": 0.3490702211856842, "learning_rate": 9.732885023326053e-05, "loss": 0.0757, "step": 14930 }, { "grad_norm": 0.4992319643497467, "learning_rate": 9.732351557126418e-05, "loss": 0.0713, "step": 14940 }, { "grad_norm": 0.47701799869537354, "learning_rate": 9.731817573403929e-05, "loss": 0.0627, "step": 14950 }, { "grad_norm": 0.43166816234588623, "learning_rate": 9.731283072216985e-05, "loss": 0.0576, "step": 14960 }, { "grad_norm": 0.3951166868209839, "learning_rate": 9.730748053624039e-05, "loss": 0.0665, "step": 14970 }, { "grad_norm": 0.4174440801143646, "learning_rate": 9.730212517683598e-05, "loss": 0.0685, "step": 14980 }, { "grad_norm": 0.3801470696926117, "learning_rate": 9.729676464454228e-05, "loss": 0.0723, "step": 14990 }, { "grad_norm": 0.3694068491458893, "learning_rate": 9.72913989399455e-05, "loss": 0.0597, "step": 15000 }, { "grad_norm": 0.31176578998565674, "learning_rate": 9.728602806363242e-05, "loss": 0.0679, "step": 15010 }, { "grad_norm": 0.47855156660079956, "learning_rate": 9.728065201619043e-05, "loss": 0.0596, "step": 15020 }, { "grad_norm": 0.34806448221206665, "learning_rate": 9.727527079820742e-05, "loss": 0.0569, "step": 15030 }, { "grad_norm": 0.34796142578125, "learning_rate": 9.726988441027186e-05, "loss": 0.066, "step": 15040 }, { "grad_norm": 0.3954446017742157, "learning_rate": 9.726449285297281e-05, "loss": 0.0594, "step": 15050 }, { "grad_norm": 0.48653489351272583, "learning_rate": 9.72590961268999e-05, "loss": 0.0687, "step": 15060 }, { "grad_norm": 0.5023720264434814, "learning_rate": 9.725369423264328e-05, "loss": 0.0624, "step": 15070 }, { "grad_norm": 0.38895776867866516, "learning_rate": 9.72482871707937e-05, "loss": 0.0669, "step": 15080 }, { "grad_norm": 0.3971521258354187, "learning_rate": 9.724287494194247e-05, "loss": 0.0662, "step": 15090 }, { "grad_norm": 0.2873542308807373, "learning_rate": 9.723745754668147e-05, "loss": 0.0638, "step": 15100 }, { "grad_norm": 0.4210227429866791, "learning_rate": 9.723203498560313e-05, "loss": 0.0614, "step": 15110 }, { "grad_norm": 0.4495655298233032, "learning_rate": 9.722660725930046e-05, "loss": 0.0655, "step": 15120 }, { "grad_norm": 0.4274287819862366, "learning_rate": 9.722117436836702e-05, "loss": 0.0642, "step": 15130 }, { "grad_norm": 0.4419638216495514, "learning_rate": 9.721573631339696e-05, "loss": 0.0588, "step": 15140 }, { "grad_norm": 0.511803925037384, "learning_rate": 9.721029309498494e-05, "loss": 0.0609, "step": 15150 }, { "grad_norm": 0.4175030291080475, "learning_rate": 9.720484471372627e-05, "loss": 0.07, "step": 15160 }, { "grad_norm": 0.4598087668418884, "learning_rate": 9.719939117021673e-05, "loss": 0.0588, "step": 15170 }, { "grad_norm": 0.28739434480667114, "learning_rate": 9.719393246505275e-05, "loss": 0.0603, "step": 15180 }, { "grad_norm": 0.4162970185279846, "learning_rate": 9.718846859883128e-05, "loss": 0.0562, "step": 15190 }, { "grad_norm": 0.5053355693817139, "learning_rate": 9.718299957214982e-05, "loss": 0.0514, "step": 15200 }, { "grad_norm": 0.395190566778183, "learning_rate": 9.717752538560646e-05, "loss": 0.062, "step": 15210 }, { "grad_norm": 0.44335514307022095, "learning_rate": 9.717204603979986e-05, "loss": 0.0604, "step": 15220 }, { "grad_norm": 0.3532601296901703, "learning_rate": 9.716656153532922e-05, "loss": 0.0586, "step": 15230 }, { "grad_norm": 0.3840133845806122, "learning_rate": 9.716107187279434e-05, "loss": 0.0613, "step": 15240 }, { "grad_norm": 0.3804410994052887, "learning_rate": 9.715557705279555e-05, "loss": 0.0562, "step": 15250 }, { "grad_norm": 0.4803486466407776, "learning_rate": 9.715007707593372e-05, "loss": 0.0566, "step": 15260 }, { "grad_norm": 0.4317476749420166, "learning_rate": 9.714457194281036e-05, "loss": 0.0647, "step": 15270 }, { "grad_norm": 0.6744961142539978, "learning_rate": 9.713906165402751e-05, "loss": 0.0692, "step": 15280 }, { "grad_norm": 0.30669930577278137, "learning_rate": 9.713354621018774e-05, "loss": 0.067, "step": 15290 }, { "grad_norm": 0.4138084352016449, "learning_rate": 9.712802561189422e-05, "loss": 0.0611, "step": 15300 }, { "grad_norm": 0.3827626407146454, "learning_rate": 9.712249985975069e-05, "loss": 0.0604, "step": 15310 }, { "grad_norm": 0.33438730239868164, "learning_rate": 9.71169689543614e-05, "loss": 0.0559, "step": 15320 }, { "grad_norm": 0.42044875025749207, "learning_rate": 9.711143289633123e-05, "loss": 0.0647, "step": 15330 }, { "grad_norm": 0.36984983086586, "learning_rate": 9.710589168626561e-05, "loss": 0.057, "step": 15340 }, { "grad_norm": 0.43586868047714233, "learning_rate": 9.710034532477048e-05, "loss": 0.0602, "step": 15350 }, { "grad_norm": 0.3946310877799988, "learning_rate": 9.709479381245239e-05, "loss": 0.0676, "step": 15360 }, { "grad_norm": 0.413074254989624, "learning_rate": 9.708923714991847e-05, "loss": 0.068, "step": 15370 }, { "grad_norm": 0.4185788035392761, "learning_rate": 9.708367533777638e-05, "loss": 0.0559, "step": 15380 }, { "grad_norm": 0.37720057368278503, "learning_rate": 9.707810837663431e-05, "loss": 0.0567, "step": 15390 }, { "grad_norm": 0.6117680072784424, "learning_rate": 9.707253626710113e-05, "loss": 0.0589, "step": 15400 }, { "grad_norm": 0.5187959671020508, "learning_rate": 9.706695900978613e-05, "loss": 0.0547, "step": 15410 }, { "grad_norm": 0.514106273651123, "learning_rate": 9.706137660529926e-05, "loss": 0.0662, "step": 15420 }, { "grad_norm": 0.6279237270355225, "learning_rate": 9.705578905425101e-05, "loss": 0.0522, "step": 15430 }, { "grad_norm": 0.41295939683914185, "learning_rate": 9.705019635725241e-05, "loss": 0.0622, "step": 15440 }, { "grad_norm": 0.3992364704608917, "learning_rate": 9.704459851491508e-05, "loss": 0.0668, "step": 15450 }, { "grad_norm": 0.5148054957389832, "learning_rate": 9.703899552785118e-05, "loss": 0.0609, "step": 15460 }, { "grad_norm": 0.3607728183269501, "learning_rate": 9.703338739667346e-05, "loss": 0.0655, "step": 15470 }, { "grad_norm": 0.5554201006889343, "learning_rate": 9.70277741219952e-05, "loss": 0.0661, "step": 15480 }, { "grad_norm": 0.45937207341194153, "learning_rate": 9.702215570443027e-05, "loss": 0.0669, "step": 15490 }, { "grad_norm": 0.48480910062789917, "learning_rate": 9.701653214459309e-05, "loss": 0.0697, "step": 15500 }, { "grad_norm": 0.42144832015037537, "learning_rate": 9.701090344309865e-05, "loss": 0.0643, "step": 15510 }, { "grad_norm": 0.37328609824180603, "learning_rate": 9.700526960056247e-05, "loss": 0.0577, "step": 15520 }, { "grad_norm": 0.3315489888191223, "learning_rate": 9.699963061760068e-05, "loss": 0.0562, "step": 15530 }, { "grad_norm": 0.43785056471824646, "learning_rate": 9.699398649482997e-05, "loss": 0.0605, "step": 15540 }, { "grad_norm": 0.40197262167930603, "learning_rate": 9.698833723286753e-05, "loss": 0.0601, "step": 15550 }, { "grad_norm": 0.4871131181716919, "learning_rate": 9.698268283233118e-05, "loss": 0.0638, "step": 15560 }, { "grad_norm": 0.38944950699806213, "learning_rate": 9.697702329383929e-05, "loss": 0.0596, "step": 15570 }, { "grad_norm": 0.31329500675201416, "learning_rate": 9.697135861801074e-05, "loss": 0.0556, "step": 15580 }, { "grad_norm": 0.5739109516143799, "learning_rate": 9.696568880546505e-05, "loss": 0.0683, "step": 15590 }, { "grad_norm": 0.3947731852531433, "learning_rate": 9.696001385682223e-05, "loss": 0.0627, "step": 15600 }, { "grad_norm": 0.40440210700035095, "learning_rate": 9.695433377270291e-05, "loss": 0.0704, "step": 15610 }, { "grad_norm": 0.38312211632728577, "learning_rate": 9.694864855372824e-05, "loss": 0.0564, "step": 15620 }, { "grad_norm": 0.31747764348983765, "learning_rate": 9.694295820051995e-05, "loss": 0.0632, "step": 15630 }, { "grad_norm": 0.44796326756477356, "learning_rate": 9.693726271370032e-05, "loss": 0.0584, "step": 15640 }, { "grad_norm": 0.4551287889480591, "learning_rate": 9.693156209389221e-05, "loss": 0.0564, "step": 15650 }, { "grad_norm": 0.4404459297657013, "learning_rate": 9.692585634171905e-05, "loss": 0.06, "step": 15660 }, { "grad_norm": 0.4836992919445038, "learning_rate": 9.692014545780476e-05, "loss": 0.061, "step": 15670 }, { "grad_norm": 0.47238248586654663, "learning_rate": 9.691442944277393e-05, "loss": 0.056, "step": 15680 }, { "grad_norm": 0.31832587718963623, "learning_rate": 9.690870829725162e-05, "loss": 0.0584, "step": 15690 }, { "grad_norm": 0.5701001882553101, "learning_rate": 9.69029820218635e-05, "loss": 0.0622, "step": 15700 }, { "grad_norm": 0.3452346920967102, "learning_rate": 9.689725061723579e-05, "loss": 0.05, "step": 15710 }, { "grad_norm": 0.566537618637085, "learning_rate": 9.689151408399527e-05, "loss": 0.0615, "step": 15720 }, { "grad_norm": 0.3568437397480011, "learning_rate": 9.688577242276924e-05, "loss": 0.0622, "step": 15730 }, { "grad_norm": 0.44915395975112915, "learning_rate": 9.688002563418566e-05, "loss": 0.0639, "step": 15740 }, { "grad_norm": 0.4709109663963318, "learning_rate": 9.687427371887293e-05, "loss": 0.0664, "step": 15750 }, { "grad_norm": 0.4662081003189087, "learning_rate": 9.686851667746012e-05, "loss": 0.0649, "step": 15760 }, { "grad_norm": 0.4353863298892975, "learning_rate": 9.686275451057677e-05, "loss": 0.0615, "step": 15770 }, { "grad_norm": 0.7492551207542419, "learning_rate": 9.685698721885308e-05, "loss": 0.0541, "step": 15780 }, { "grad_norm": 0.3491419851779938, "learning_rate": 9.68512148029197e-05, "loss": 0.06, "step": 15790 }, { "grad_norm": 0.4188554286956787, "learning_rate": 9.684543726340791e-05, "loss": 0.0593, "step": 15800 }, { "grad_norm": 0.4198671877384186, "learning_rate": 9.683965460094952e-05, "loss": 0.0646, "step": 15810 }, { "grad_norm": 0.42883989214897156, "learning_rate": 9.683386681617694e-05, "loss": 0.0638, "step": 15820 }, { "grad_norm": 0.4232170283794403, "learning_rate": 9.68280739097231e-05, "loss": 0.0626, "step": 15830 }, { "grad_norm": 0.35710129141807556, "learning_rate": 9.682227588222148e-05, "loss": 0.0628, "step": 15840 }, { "grad_norm": 0.4513288736343384, "learning_rate": 9.681647273430618e-05, "loss": 0.0613, "step": 15850 }, { "grad_norm": 0.43281984329223633, "learning_rate": 9.681066446661182e-05, "loss": 0.0674, "step": 15860 }, { "grad_norm": 0.4098341464996338, "learning_rate": 9.680485107977357e-05, "loss": 0.0679, "step": 15870 }, { "grad_norm": 0.6490739583969116, "learning_rate": 9.679903257442716e-05, "loss": 0.0623, "step": 15880 }, { "grad_norm": 0.5478858947753906, "learning_rate": 9.679320895120891e-05, "loss": 0.0553, "step": 15890 }, { "grad_norm": 0.438327819108963, "learning_rate": 9.67873802107557e-05, "loss": 0.0617, "step": 15900 }, { "grad_norm": 0.3345422148704529, "learning_rate": 9.67815463537049e-05, "loss": 0.054, "step": 15910 }, { "grad_norm": 0.4513009488582611, "learning_rate": 9.677570738069457e-05, "loss": 0.0619, "step": 15920 }, { "grad_norm": 0.3784838616847992, "learning_rate": 9.676986329236318e-05, "loss": 0.0603, "step": 15930 }, { "grad_norm": 0.5199407935142517, "learning_rate": 9.676401408934987e-05, "loss": 0.0652, "step": 15940 }, { "grad_norm": 0.6228814125061035, "learning_rate": 9.675815977229428e-05, "loss": 0.0683, "step": 15950 }, { "grad_norm": 0.5913652777671814, "learning_rate": 9.675230034183664e-05, "loss": 0.0666, "step": 15960 }, { "grad_norm": 0.5310885310173035, "learning_rate": 9.674643579861773e-05, "loss": 0.0564, "step": 15970 }, { "grad_norm": 0.48606136441230774, "learning_rate": 9.674056614327886e-05, "loss": 0.0654, "step": 15980 }, { "grad_norm": 0.40910086035728455, "learning_rate": 9.673469137646198e-05, "loss": 0.0616, "step": 15990 }, { "grad_norm": 0.513569712638855, "learning_rate": 9.67288114988095e-05, "loss": 0.0559, "step": 16000 }, { "grad_norm": 0.37011706829071045, "learning_rate": 9.672292651096447e-05, "loss": 0.0574, "step": 16010 }, { "grad_norm": 0.48890459537506104, "learning_rate": 9.671703641357042e-05, "loss": 0.0592, "step": 16020 }, { "grad_norm": 0.5683310031890869, "learning_rate": 9.67111412072715e-05, "loss": 0.0736, "step": 16030 }, { "grad_norm": 0.4751211106777191, "learning_rate": 9.670524089271242e-05, "loss": 0.0621, "step": 16040 }, { "grad_norm": 0.4105353057384491, "learning_rate": 9.669933547053842e-05, "loss": 0.0598, "step": 16050 }, { "grad_norm": 0.5052503943443298, "learning_rate": 9.669342494139531e-05, "loss": 0.0603, "step": 16060 }, { "grad_norm": 0.41787415742874146, "learning_rate": 9.668750930592943e-05, "loss": 0.0659, "step": 16070 }, { "grad_norm": 0.49336275458335876, "learning_rate": 9.668158856478775e-05, "loss": 0.066, "step": 16080 }, { "grad_norm": 0.48892277479171753, "learning_rate": 9.66756627186177e-05, "loss": 0.0576, "step": 16090 }, { "grad_norm": 0.5523141622543335, "learning_rate": 9.666973176806737e-05, "loss": 0.0597, "step": 16100 }, { "grad_norm": 0.39955875277519226, "learning_rate": 9.666379571378534e-05, "loss": 0.0649, "step": 16110 }, { "grad_norm": 0.43402525782585144, "learning_rate": 9.665785455642076e-05, "loss": 0.0613, "step": 16120 }, { "grad_norm": 0.39012283086776733, "learning_rate": 9.665190829662337e-05, "loss": 0.0643, "step": 16130 }, { "grad_norm": 0.37951740622520447, "learning_rate": 9.664595693504342e-05, "loss": 0.0555, "step": 16140 }, { "grad_norm": 0.3632109463214874, "learning_rate": 9.664000047233175e-05, "loss": 0.0536, "step": 16150 }, { "grad_norm": 0.4916727542877197, "learning_rate": 9.663403890913976e-05, "loss": 0.0592, "step": 16160 }, { "grad_norm": 0.4763434827327728, "learning_rate": 9.662807224611938e-05, "loss": 0.0628, "step": 16170 }, { "grad_norm": 0.5421189069747925, "learning_rate": 9.662210048392311e-05, "loss": 0.0592, "step": 16180 }, { "grad_norm": 0.5029249787330627, "learning_rate": 9.661612362320405e-05, "loss": 0.0617, "step": 16190 }, { "grad_norm": 0.5276304483413696, "learning_rate": 9.661014166461579e-05, "loss": 0.0642, "step": 16200 }, { "grad_norm": 0.35358279943466187, "learning_rate": 9.66041546088125e-05, "loss": 0.0708, "step": 16210 }, { "grad_norm": 0.33940044045448303, "learning_rate": 9.659816245644895e-05, "loss": 0.0583, "step": 16220 }, { "grad_norm": 0.37191954255104065, "learning_rate": 9.65921652081804e-05, "loss": 0.0599, "step": 16230 }, { "grad_norm": 0.5564417243003845, "learning_rate": 9.658616286466271e-05, "loss": 0.0608, "step": 16240 }, { "grad_norm": 0.4366739094257355, "learning_rate": 9.65801554265523e-05, "loss": 0.0609, "step": 16250 }, { "grad_norm": 0.3524635136127472, "learning_rate": 9.657414289450612e-05, "loss": 0.0557, "step": 16260 }, { "grad_norm": 0.3979763090610504, "learning_rate": 9.656812526918171e-05, "loss": 0.0606, "step": 16270 }, { "grad_norm": 0.4525822401046753, "learning_rate": 9.656210255123712e-05, "loss": 0.0599, "step": 16280 }, { "grad_norm": 0.36327505111694336, "learning_rate": 9.6556074741331e-05, "loss": 0.0515, "step": 16290 }, { "grad_norm": 0.45617443323135376, "learning_rate": 9.655004184012256e-05, "loss": 0.0584, "step": 16300 }, { "grad_norm": 0.3797754943370819, "learning_rate": 9.654400384827152e-05, "loss": 0.06, "step": 16310 }, { "grad_norm": 0.3731183111667633, "learning_rate": 9.653796076643818e-05, "loss": 0.059, "step": 16320 }, { "grad_norm": 0.38389429450035095, "learning_rate": 9.653191259528344e-05, "loss": 0.0634, "step": 16330 }, { "grad_norm": 0.39166390895843506, "learning_rate": 9.65258593354687e-05, "loss": 0.0628, "step": 16340 }, { "grad_norm": 0.4318544864654541, "learning_rate": 9.651980098765591e-05, "loss": 0.0613, "step": 16350 }, { "grad_norm": 0.5263717174530029, "learning_rate": 9.651373755250765e-05, "loss": 0.0569, "step": 16360 }, { "grad_norm": 0.35639140009880066, "learning_rate": 9.650766903068697e-05, "loss": 0.0613, "step": 16370 }, { "grad_norm": 0.46984317898750305, "learning_rate": 9.650159542285753e-05, "loss": 0.0677, "step": 16380 }, { "grad_norm": 0.35541826486587524, "learning_rate": 9.649551672968353e-05, "loss": 0.0588, "step": 16390 }, { "grad_norm": 0.30756065249443054, "learning_rate": 9.648943295182973e-05, "loss": 0.054, "step": 16400 }, { "grad_norm": 0.33783721923828125, "learning_rate": 9.648334408996144e-05, "loss": 0.0549, "step": 16410 }, { "grad_norm": 0.6952974796295166, "learning_rate": 9.647725014474452e-05, "loss": 0.0564, "step": 16420 }, { "grad_norm": 0.36263734102249146, "learning_rate": 9.64711511168454e-05, "loss": 0.0638, "step": 16430 }, { "grad_norm": 0.3211722671985626, "learning_rate": 9.646504700693108e-05, "loss": 0.056, "step": 16440 }, { "grad_norm": 0.3325880169868469, "learning_rate": 9.645893781566907e-05, "loss": 0.0619, "step": 16450 }, { "grad_norm": 0.48596733808517456, "learning_rate": 9.645282354372744e-05, "loss": 0.0587, "step": 16460 }, { "grad_norm": 0.3346984088420868, "learning_rate": 9.644670419177491e-05, "loss": 0.0586, "step": 16470 }, { "grad_norm": 0.40149548649787903, "learning_rate": 9.644057976048062e-05, "loss": 0.0589, "step": 16480 }, { "grad_norm": 0.4171065390110016, "learning_rate": 9.643445025051435e-05, "loss": 0.0585, "step": 16490 }, { "grad_norm": 0.5521630048751831, "learning_rate": 9.642831566254641e-05, "loss": 0.0649, "step": 16500 }, { "grad_norm": 0.5816504955291748, "learning_rate": 9.642217599724769e-05, "loss": 0.0715, "step": 16510 }, { "grad_norm": 0.7581838369369507, "learning_rate": 9.64160312552896e-05, "loss": 0.0634, "step": 16520 }, { "grad_norm": 0.5480811595916748, "learning_rate": 9.64098814373441e-05, "loss": 0.0726, "step": 16530 }, { "grad_norm": 0.3366483151912689, "learning_rate": 9.640372654408374e-05, "loss": 0.064, "step": 16540 }, { "grad_norm": 0.352573961019516, "learning_rate": 9.639756657618162e-05, "loss": 0.0568, "step": 16550 }, { "grad_norm": 0.42120397090911865, "learning_rate": 9.639140153431138e-05, "loss": 0.0568, "step": 16560 }, { "grad_norm": 0.2730524241924286, "learning_rate": 9.638523141914721e-05, "loss": 0.0558, "step": 16570 }, { "grad_norm": 0.5420225262641907, "learning_rate": 9.637905623136388e-05, "loss": 0.0604, "step": 16580 }, { "grad_norm": 0.4366390109062195, "learning_rate": 9.637287597163669e-05, "loss": 0.0587, "step": 16590 }, { "grad_norm": 0.4026181995868683, "learning_rate": 9.63666906406415e-05, "loss": 0.0555, "step": 16600 }, { "grad_norm": 0.5754119753837585, "learning_rate": 9.636050023905473e-05, "loss": 0.0589, "step": 16610 }, { "grad_norm": 0.5465618968009949, "learning_rate": 9.635430476755336e-05, "loss": 0.0619, "step": 16620 }, { "grad_norm": 0.48868876695632935, "learning_rate": 9.63481042268149e-05, "loss": 0.0599, "step": 16630 }, { "grad_norm": 0.4208429455757141, "learning_rate": 9.634189861751745e-05, "loss": 0.0605, "step": 16640 }, { "grad_norm": 0.41698089241981506, "learning_rate": 9.633568794033967e-05, "loss": 0.0616, "step": 16650 }, { "grad_norm": 0.2736318111419678, "learning_rate": 9.63294721959607e-05, "loss": 0.0579, "step": 16660 }, { "grad_norm": 0.4190926253795624, "learning_rate": 9.63232513850603e-05, "loss": 0.0613, "step": 16670 }, { "grad_norm": 0.6459277868270874, "learning_rate": 9.631702550831878e-05, "loss": 0.0565, "step": 16680 }, { "grad_norm": 0.4406355023384094, "learning_rate": 9.631079456641698e-05, "loss": 0.0595, "step": 16690 }, { "grad_norm": 0.4322490394115448, "learning_rate": 9.630455856003632e-05, "loss": 0.0666, "step": 16700 }, { "grad_norm": 0.47930383682250977, "learning_rate": 9.629831748985876e-05, "loss": 0.0613, "step": 16710 }, { "grad_norm": 0.5580645203590393, "learning_rate": 9.629207135656679e-05, "loss": 0.0615, "step": 16720 }, { "grad_norm": 0.4719333350658417, "learning_rate": 9.628582016084353e-05, "loss": 0.0612, "step": 16730 }, { "grad_norm": 0.3637431859970093, "learning_rate": 9.627956390337254e-05, "loss": 0.0643, "step": 16740 }, { "grad_norm": 0.4810536205768585, "learning_rate": 9.627330258483802e-05, "loss": 0.0577, "step": 16750 }, { "grad_norm": 0.5278491973876953, "learning_rate": 9.62670362059247e-05, "loss": 0.0734, "step": 16760 }, { "grad_norm": 0.35862284898757935, "learning_rate": 9.626076476731786e-05, "loss": 0.06, "step": 16770 }, { "grad_norm": 0.3652914762496948, "learning_rate": 9.625448826970336e-05, "loss": 0.0568, "step": 16780 }, { "grad_norm": 0.3773084878921509, "learning_rate": 9.624820671376755e-05, "loss": 0.0583, "step": 16790 }, { "grad_norm": 0.34553641080856323, "learning_rate": 9.62419201001974e-05, "loss": 0.0661, "step": 16800 }, { "grad_norm": 0.4362682104110718, "learning_rate": 9.623562842968037e-05, "loss": 0.0583, "step": 16810 }, { "grad_norm": 0.6021727919578552, "learning_rate": 9.622933170290454e-05, "loss": 0.0569, "step": 16820 }, { "grad_norm": 0.3345867097377777, "learning_rate": 9.622302992055849e-05, "loss": 0.0543, "step": 16830 }, { "grad_norm": 0.6041052937507629, "learning_rate": 9.62167230833314e-05, "loss": 0.0677, "step": 16840 }, { "grad_norm": 0.4019497036933899, "learning_rate": 9.621041119191295e-05, "loss": 0.057, "step": 16850 }, { "grad_norm": 0.5164831876754761, "learning_rate": 9.620409424699342e-05, "loss": 0.0618, "step": 16860 }, { "grad_norm": 0.4876742959022522, "learning_rate": 9.619777224926359e-05, "loss": 0.0578, "step": 16870 }, { "grad_norm": 0.4293515086174011, "learning_rate": 9.619144519941485e-05, "loss": 0.0675, "step": 16880 }, { "grad_norm": 0.4221687316894531, "learning_rate": 9.618511309813912e-05, "loss": 0.0639, "step": 16890 }, { "grad_norm": 0.48922091722488403, "learning_rate": 9.617877594612886e-05, "loss": 0.0637, "step": 16900 }, { "grad_norm": 0.4873340129852295, "learning_rate": 9.617243374407707e-05, "loss": 0.0637, "step": 16910 }, { "grad_norm": 0.5431274771690369, "learning_rate": 9.616608649267736e-05, "loss": 0.0693, "step": 16920 }, { "grad_norm": 0.37688687443733215, "learning_rate": 9.615973419262385e-05, "loss": 0.0687, "step": 16930 }, { "grad_norm": 0.41073206067085266, "learning_rate": 9.615337684461119e-05, "loss": 0.0624, "step": 16940 }, { "grad_norm": 0.6231341361999512, "learning_rate": 9.614701444933465e-05, "loss": 0.0634, "step": 16950 }, { "grad_norm": 0.39668217301368713, "learning_rate": 9.614064700748997e-05, "loss": 0.059, "step": 16960 }, { "grad_norm": 0.360948383808136, "learning_rate": 9.613427451977352e-05, "loss": 0.0548, "step": 16970 }, { "grad_norm": 0.4141702950000763, "learning_rate": 9.612789698688216e-05, "loss": 0.0617, "step": 16980 }, { "grad_norm": 0.3804275691509247, "learning_rate": 9.612151440951334e-05, "loss": 0.056, "step": 16990 }, { "grad_norm": 0.4414640963077545, "learning_rate": 9.611512678836506e-05, "loss": 0.0598, "step": 17000 }, { "grad_norm": 0.39802923798561096, "learning_rate": 9.610873412413584e-05, "loss": 0.0585, "step": 17010 }, { "grad_norm": 0.4058459997177124, "learning_rate": 9.610233641752476e-05, "loss": 0.0594, "step": 17020 }, { "grad_norm": 0.5309897065162659, "learning_rate": 9.609593366923151e-05, "loss": 0.0573, "step": 17030 }, { "grad_norm": 0.38680094480514526, "learning_rate": 9.608952587995625e-05, "loss": 0.0624, "step": 17040 }, { "grad_norm": 0.4336685836315155, "learning_rate": 9.608311305039972e-05, "loss": 0.0655, "step": 17050 }, { "grad_norm": 0.4499099850654602, "learning_rate": 9.607669518126326e-05, "loss": 0.0649, "step": 17060 }, { "grad_norm": 0.4335681200027466, "learning_rate": 9.607027227324866e-05, "loss": 0.0653, "step": 17070 }, { "grad_norm": 0.4344141185283661, "learning_rate": 9.606384432705837e-05, "loss": 0.0629, "step": 17080 }, { "grad_norm": 0.5469728112220764, "learning_rate": 9.60574113433953e-05, "loss": 0.064, "step": 17090 }, { "grad_norm": 0.5944713354110718, "learning_rate": 9.6050973322963e-05, "loss": 0.0555, "step": 17100 }, { "grad_norm": 0.5002816319465637, "learning_rate": 9.604453026646547e-05, "loss": 0.0634, "step": 17110 }, { "grad_norm": 0.3796822428703308, "learning_rate": 9.603808217460735e-05, "loss": 0.0625, "step": 17120 }, { "grad_norm": 0.48392900824546814, "learning_rate": 9.603162904809377e-05, "loss": 0.0621, "step": 17130 }, { "grad_norm": 0.5287402868270874, "learning_rate": 9.602517088763045e-05, "loss": 0.0579, "step": 17140 }, { "grad_norm": 0.4631866216659546, "learning_rate": 9.601870769392365e-05, "loss": 0.0626, "step": 17150 }, { "grad_norm": 0.39890944957733154, "learning_rate": 9.601223946768017e-05, "loss": 0.0632, "step": 17160 }, { "grad_norm": 0.44447043538093567, "learning_rate": 9.600576620960734e-05, "loss": 0.059, "step": 17170 }, { "grad_norm": 0.42908474802970886, "learning_rate": 9.599928792041308e-05, "loss": 0.0545, "step": 17180 }, { "grad_norm": 0.3660754859447479, "learning_rate": 9.599280460080587e-05, "loss": 0.0634, "step": 17190 }, { "grad_norm": 0.3697856068611145, "learning_rate": 9.59863162514947e-05, "loss": 0.0509, "step": 17200 }, { "grad_norm": 0.619385302066803, "learning_rate": 9.597982287318911e-05, "loss": 0.0641, "step": 17210 }, { "grad_norm": 0.35664722323417664, "learning_rate": 9.597332446659923e-05, "loss": 0.0586, "step": 17220 }, { "grad_norm": 0.5721072554588318, "learning_rate": 9.59668210324357e-05, "loss": 0.059, "step": 17230 }, { "grad_norm": 0.3380691409111023, "learning_rate": 9.596031257140974e-05, "loss": 0.0561, "step": 17240 }, { "grad_norm": 0.3701290488243103, "learning_rate": 9.59537990842331e-05, "loss": 0.0496, "step": 17250 }, { "grad_norm": 0.3064689040184021, "learning_rate": 9.594728057161806e-05, "loss": 0.0535, "step": 17260 }, { "grad_norm": 0.43313848972320557, "learning_rate": 9.594075703427752e-05, "loss": 0.0545, "step": 17270 }, { "grad_norm": 0.37686407566070557, "learning_rate": 9.593422847292486e-05, "loss": 0.0623, "step": 17280 }, { "grad_norm": 0.4941509962081909, "learning_rate": 9.592769488827402e-05, "loss": 0.0651, "step": 17290 }, { "grad_norm": 0.40774449706077576, "learning_rate": 9.592115628103952e-05, "loss": 0.0553, "step": 17300 }, { "grad_norm": 0.4137912392616272, "learning_rate": 9.591461265193643e-05, "loss": 0.0599, "step": 17310 }, { "grad_norm": 0.48469144105911255, "learning_rate": 9.590806400168032e-05, "loss": 0.0545, "step": 17320 }, { "grad_norm": 0.4426971673965454, "learning_rate": 9.590151033098735e-05, "loss": 0.0575, "step": 17330 }, { "grad_norm": 0.3039480447769165, "learning_rate": 9.589495164057423e-05, "loss": 0.0713, "step": 17340 }, { "grad_norm": 0.5287808179855347, "learning_rate": 9.58883879311582e-05, "loss": 0.0619, "step": 17350 }, { "grad_norm": 0.4078473150730133, "learning_rate": 9.588181920345705e-05, "loss": 0.0587, "step": 17360 }, { "grad_norm": 0.4066232442855835, "learning_rate": 9.587524545818913e-05, "loss": 0.0593, "step": 17370 }, { "grad_norm": 0.3986911177635193, "learning_rate": 9.586866669607335e-05, "loss": 0.0554, "step": 17380 }, { "grad_norm": 1.6394990682601929, "learning_rate": 9.586208291782915e-05, "loss": 0.0587, "step": 17390 }, { "grad_norm": 0.6868548393249512, "learning_rate": 9.58554941241765e-05, "loss": 0.0708, "step": 17400 }, { "grad_norm": 0.4334648847579956, "learning_rate": 9.584890031583596e-05, "loss": 0.068, "step": 17410 }, { "grad_norm": 0.367317259311676, "learning_rate": 9.584230149352861e-05, "loss": 0.0607, "step": 17420 }, { "grad_norm": 0.46241629123687744, "learning_rate": 9.58356976579761e-05, "loss": 0.061, "step": 17430 }, { "grad_norm": 0.5347324013710022, "learning_rate": 9.58290888099006e-05, "loss": 0.0654, "step": 17440 }, { "grad_norm": 0.5060574412345886, "learning_rate": 9.582247495002486e-05, "loss": 0.0707, "step": 17450 }, { "grad_norm": 0.34265372157096863, "learning_rate": 9.581585607907214e-05, "loss": 0.057, "step": 17460 }, { "grad_norm": 0.413983017206192, "learning_rate": 9.580923219776628e-05, "loss": 0.0588, "step": 17470 }, { "grad_norm": 0.6745619773864746, "learning_rate": 9.580260330683167e-05, "loss": 0.0635, "step": 17480 }, { "grad_norm": 0.367171049118042, "learning_rate": 9.579596940699322e-05, "loss": 0.0637, "step": 17490 }, { "grad_norm": 0.31820183992385864, "learning_rate": 9.578933049897643e-05, "loss": 0.0598, "step": 17500 }, { "grad_norm": 0.3695896863937378, "learning_rate": 9.578268658350728e-05, "loss": 0.0615, "step": 17510 }, { "grad_norm": 0.5227468609809875, "learning_rate": 9.577603766131235e-05, "loss": 0.0586, "step": 17520 }, { "grad_norm": 0.34405311942100525, "learning_rate": 9.576938373311878e-05, "loss": 0.0519, "step": 17530 }, { "grad_norm": 0.5346696972846985, "learning_rate": 9.576272479965421e-05, "loss": 0.0556, "step": 17540 }, { "grad_norm": 0.3974076509475708, "learning_rate": 9.575606086164687e-05, "loss": 0.0573, "step": 17550 }, { "grad_norm": 0.7337425947189331, "learning_rate": 9.57493919198255e-05, "loss": 0.0622, "step": 17560 }, { "grad_norm": 0.4461428225040436, "learning_rate": 9.57427179749194e-05, "loss": 0.0642, "step": 17570 }, { "grad_norm": 0.42605262994766235, "learning_rate": 9.573603902765846e-05, "loss": 0.0606, "step": 17580 }, { "grad_norm": 0.38789626955986023, "learning_rate": 9.572935507877304e-05, "loss": 0.0595, "step": 17590 }, { "grad_norm": 0.31385499238967896, "learning_rate": 9.57226661289941e-05, "loss": 0.0604, "step": 17600 }, { "grad_norm": 0.46848946809768677, "learning_rate": 9.571597217905315e-05, "loss": 0.0589, "step": 17610 }, { "grad_norm": 0.5144411325454712, "learning_rate": 9.57092732296822e-05, "loss": 0.0609, "step": 17620 }, { "grad_norm": 0.4585201144218445, "learning_rate": 9.570256928161385e-05, "loss": 0.0644, "step": 17630 }, { "grad_norm": 0.384966641664505, "learning_rate": 9.569586033558126e-05, "loss": 0.0628, "step": 17640 }, { "grad_norm": 0.4132063388824463, "learning_rate": 9.568914639231807e-05, "loss": 0.0662, "step": 17650 }, { "grad_norm": 0.3886067271232605, "learning_rate": 9.568242745255852e-05, "loss": 0.0698, "step": 17660 }, { "grad_norm": 0.5322851538658142, "learning_rate": 9.567570351703739e-05, "loss": 0.0565, "step": 17670 }, { "grad_norm": 0.41016361117362976, "learning_rate": 9.566897458649001e-05, "loss": 0.062, "step": 17680 }, { "grad_norm": 1.1494522094726562, "learning_rate": 9.566224066165221e-05, "loss": 0.0583, "step": 17690 }, { "grad_norm": 0.36487317085266113, "learning_rate": 9.565550174326043e-05, "loss": 0.0719, "step": 17700 }, { "grad_norm": 0.5080239772796631, "learning_rate": 9.564875783205162e-05, "loss": 0.0579, "step": 17710 }, { "grad_norm": 0.4742264449596405, "learning_rate": 9.564200892876328e-05, "loss": 0.0586, "step": 17720 }, { "grad_norm": 0.4661191999912262, "learning_rate": 9.563525503413348e-05, "loss": 0.0674, "step": 17730 }, { "grad_norm": 0.3770574927330017, "learning_rate": 9.562849614890079e-05, "loss": 0.0618, "step": 17740 }, { "grad_norm": 0.5617015361785889, "learning_rate": 9.562173227380436e-05, "loss": 0.0557, "step": 17750 }, { "grad_norm": 0.4334663152694702, "learning_rate": 9.561496340958389e-05, "loss": 0.0602, "step": 17760 }, { "grad_norm": 0.4108281135559082, "learning_rate": 9.560818955697959e-05, "loss": 0.0677, "step": 17770 }, { "grad_norm": 0.6031604409217834, "learning_rate": 9.560141071673228e-05, "loss": 0.0612, "step": 17780 }, { "grad_norm": 0.4449359178543091, "learning_rate": 9.559462688958323e-05, "loss": 0.0552, "step": 17790 }, { "grad_norm": 0.48625174164772034, "learning_rate": 9.558783807627434e-05, "loss": 0.062, "step": 17800 }, { "grad_norm": 0.3957100212574005, "learning_rate": 9.558104427754801e-05, "loss": 0.0592, "step": 17810 }, { "grad_norm": 0.2847914397716522, "learning_rate": 9.557424549414722e-05, "loss": 0.0542, "step": 17820 }, { "grad_norm": 0.47351789474487305, "learning_rate": 9.556744172681546e-05, "loss": 0.0615, "step": 17830 }, { "grad_norm": 0.4262940585613251, "learning_rate": 9.556063297629677e-05, "loss": 0.0529, "step": 17840 }, { "grad_norm": 0.5410462021827698, "learning_rate": 9.555381924333578e-05, "loss": 0.0563, "step": 17850 }, { "grad_norm": 0.4944166839122772, "learning_rate": 9.554700052867758e-05, "loss": 0.0606, "step": 17860 }, { "grad_norm": 0.37445974349975586, "learning_rate": 9.554017683306789e-05, "loss": 0.0559, "step": 17870 }, { "grad_norm": 0.3165874481201172, "learning_rate": 9.553334815725294e-05, "loss": 0.0568, "step": 17880 }, { "grad_norm": 0.4159030616283417, "learning_rate": 9.552651450197949e-05, "loss": 0.0633, "step": 17890 }, { "grad_norm": 0.4526112377643585, "learning_rate": 9.551967586799486e-05, "loss": 0.0625, "step": 17900 }, { "grad_norm": 0.49869710206985474, "learning_rate": 9.551283225604692e-05, "loss": 0.0615, "step": 17910 }, { "grad_norm": 0.43937328457832336, "learning_rate": 9.550598366688406e-05, "loss": 0.0679, "step": 17920 }, { "grad_norm": 0.47219470143318176, "learning_rate": 9.549913010125526e-05, "loss": 0.0593, "step": 17930 }, { "grad_norm": 0.43084457516670227, "learning_rate": 9.549227155990999e-05, "loss": 0.0575, "step": 17940 }, { "grad_norm": 0.40419238805770874, "learning_rate": 9.548540804359828e-05, "loss": 0.0586, "step": 17950 }, { "grad_norm": 0.39565619826316833, "learning_rate": 9.547853955307077e-05, "loss": 0.0556, "step": 17960 }, { "grad_norm": 0.3852200508117676, "learning_rate": 9.547166608907853e-05, "loss": 0.0546, "step": 17970 }, { "grad_norm": 0.330167680978775, "learning_rate": 9.546478765237326e-05, "loss": 0.0621, "step": 17980 }, { "grad_norm": 0.5133708715438843, "learning_rate": 9.545790424370715e-05, "loss": 0.0686, "step": 17990 }, { "grad_norm": 0.37062856554985046, "learning_rate": 9.5451015863833e-05, "loss": 0.0608, "step": 18000 }, { "grad_norm": 0.4548881947994232, "learning_rate": 9.544412251350408e-05, "loss": 0.0612, "step": 18010 }, { "grad_norm": 0.3687450885772705, "learning_rate": 9.543722419347422e-05, "loss": 0.0597, "step": 18020 }, { "grad_norm": 0.4157063961029053, "learning_rate": 9.543032090449788e-05, "loss": 0.0479, "step": 18030 }, { "grad_norm": 0.49981868267059326, "learning_rate": 9.542341264732992e-05, "loss": 0.0596, "step": 18040 }, { "grad_norm": 0.468997061252594, "learning_rate": 9.541649942272585e-05, "loss": 0.0622, "step": 18050 }, { "grad_norm": 0.872050404548645, "learning_rate": 9.54095812314417e-05, "loss": 0.0716, "step": 18060 }, { "grad_norm": 0.4143069088459015, "learning_rate": 9.540265807423401e-05, "loss": 0.0583, "step": 18070 }, { "grad_norm": 0.42095473408699036, "learning_rate": 9.53957299518599e-05, "loss": 0.059, "step": 18080 }, { "grad_norm": 0.6274218559265137, "learning_rate": 9.5388796865077e-05, "loss": 0.0522, "step": 18090 }, { "grad_norm": 0.4646688401699066, "learning_rate": 9.538185881464353e-05, "loss": 0.0593, "step": 18100 }, { "grad_norm": 0.43354880809783936, "learning_rate": 9.537491580131821e-05, "loss": 0.0663, "step": 18110 }, { "grad_norm": 0.421513170003891, "learning_rate": 9.53679678258603e-05, "loss": 0.057, "step": 18120 }, { "grad_norm": 0.583979606628418, "learning_rate": 9.536101488902966e-05, "loss": 0.0575, "step": 18130 }, { "grad_norm": 0.7855650782585144, "learning_rate": 9.535405699158663e-05, "loss": 0.0569, "step": 18140 }, { "grad_norm": 0.8065924644470215, "learning_rate": 9.53470941342921e-05, "loss": 0.0612, "step": 18150 }, { "grad_norm": 0.35715407133102417, "learning_rate": 9.534012631790756e-05, "loss": 0.0625, "step": 18160 }, { "grad_norm": 0.35093796253204346, "learning_rate": 9.533315354319494e-05, "loss": 0.0639, "step": 18170 }, { "grad_norm": 0.4330601692199707, "learning_rate": 9.532617581091682e-05, "loss": 0.0588, "step": 18180 }, { "grad_norm": 0.6135856509208679, "learning_rate": 9.531919312183629e-05, "loss": 0.0524, "step": 18190 }, { "grad_norm": 0.5952391624450684, "learning_rate": 9.531220547671688e-05, "loss": 0.0611, "step": 18200 }, { "grad_norm": 0.4412406086921692, "learning_rate": 9.530521287632285e-05, "loss": 0.0632, "step": 18210 }, { "grad_norm": 0.48638319969177246, "learning_rate": 9.529821532141884e-05, "loss": 0.0546, "step": 18220 }, { "grad_norm": 0.7653417587280273, "learning_rate": 9.52912128127701e-05, "loss": 0.0625, "step": 18230 }, { "grad_norm": 0.5704032182693481, "learning_rate": 9.528420535114244e-05, "loss": 0.064, "step": 18240 }, { "grad_norm": 0.33664950728416443, "learning_rate": 9.527719293730215e-05, "loss": 0.0598, "step": 18250 }, { "grad_norm": 0.31725332140922546, "learning_rate": 9.527017557201611e-05, "loss": 0.0593, "step": 18260 }, { "grad_norm": 0.57163405418396, "learning_rate": 9.526315325605176e-05, "loss": 0.0552, "step": 18270 }, { "grad_norm": 0.5396992564201355, "learning_rate": 9.525612599017699e-05, "loss": 0.0638, "step": 18280 }, { "grad_norm": 0.39253464341163635, "learning_rate": 9.524909377516033e-05, "loss": 0.0534, "step": 18290 }, { "grad_norm": 0.4169231653213501, "learning_rate": 9.524205661177081e-05, "loss": 0.0531, "step": 18300 }, { "grad_norm": 0.3979591131210327, "learning_rate": 9.523501450077801e-05, "loss": 0.0545, "step": 18310 }, { "grad_norm": 0.37569618225097656, "learning_rate": 9.522796744295202e-05, "loss": 0.0606, "step": 18320 }, { "grad_norm": 0.39191195368766785, "learning_rate": 9.522091543906352e-05, "loss": 0.0605, "step": 18330 }, { "grad_norm": 0.419361412525177, "learning_rate": 9.521385848988369e-05, "loss": 0.0593, "step": 18340 }, { "grad_norm": 0.7037438154220581, "learning_rate": 9.520679659618428e-05, "loss": 0.0621, "step": 18350 }, { "grad_norm": 0.4254302680492401, "learning_rate": 9.519972975873754e-05, "loss": 0.057, "step": 18360 }, { "grad_norm": 0.40317463874816895, "learning_rate": 9.519265797831633e-05, "loss": 0.0521, "step": 18370 }, { "grad_norm": 0.5181717276573181, "learning_rate": 9.518558125569399e-05, "loss": 0.0614, "step": 18380 }, { "grad_norm": 0.40705764293670654, "learning_rate": 9.517849959164442e-05, "loss": 0.0546, "step": 18390 }, { "grad_norm": 0.44924479722976685, "learning_rate": 9.517141298694205e-05, "loss": 0.0612, "step": 18400 }, { "grad_norm": 0.421049565076828, "learning_rate": 9.516432144236188e-05, "loss": 0.0653, "step": 18410 }, { "grad_norm": 0.4191572070121765, "learning_rate": 9.515722495867941e-05, "loss": 0.0624, "step": 18420 }, { "grad_norm": 0.43162715435028076, "learning_rate": 9.515012353667072e-05, "loss": 0.0615, "step": 18430 }, { "grad_norm": 0.4446653425693512, "learning_rate": 9.51430171771124e-05, "loss": 0.0592, "step": 18440 }, { "grad_norm": 0.5137816667556763, "learning_rate": 9.513590588078159e-05, "loss": 0.0574, "step": 18450 }, { "grad_norm": 0.4253771901130676, "learning_rate": 9.512878964845597e-05, "loss": 0.059, "step": 18460 }, { "grad_norm": 0.48977088928222656, "learning_rate": 9.512166848091377e-05, "loss": 0.0575, "step": 18470 }, { "grad_norm": 0.5650104284286499, "learning_rate": 9.511454237893376e-05, "loss": 0.0648, "step": 18480 }, { "grad_norm": 0.39306917786598206, "learning_rate": 9.51074113432952e-05, "loss": 0.0608, "step": 18490 }, { "grad_norm": 0.35441532731056213, "learning_rate": 9.510027537477797e-05, "loss": 0.0597, "step": 18500 }, { "grad_norm": 0.40561267733573914, "learning_rate": 9.509313447416242e-05, "loss": 0.057, "step": 18510 }, { "grad_norm": 0.49405762553215027, "learning_rate": 9.508598864222949e-05, "loss": 0.0631, "step": 18520 }, { "grad_norm": 0.5292892456054688, "learning_rate": 9.507883787976062e-05, "loss": 0.0525, "step": 18530 }, { "grad_norm": 0.39547646045684814, "learning_rate": 9.507168218753781e-05, "loss": 0.0558, "step": 18540 }, { "grad_norm": 0.33773085474967957, "learning_rate": 9.506452156634362e-05, "loss": 0.0623, "step": 18550 }, { "grad_norm": 0.3593496084213257, "learning_rate": 9.505735601696109e-05, "loss": 0.0597, "step": 18560 }, { "grad_norm": 0.3927620053291321, "learning_rate": 9.505018554017385e-05, "loss": 0.053, "step": 18570 }, { "grad_norm": 0.4927971661090851, "learning_rate": 9.504301013676604e-05, "loss": 0.064, "step": 18580 }, { "grad_norm": 0.44170764088630676, "learning_rate": 9.503582980752238e-05, "loss": 0.0664, "step": 18590 }, { "grad_norm": 0.431861013174057, "learning_rate": 9.502864455322809e-05, "loss": 0.0624, "step": 18600 }, { "grad_norm": 0.41543877124786377, "learning_rate": 9.502145437466891e-05, "loss": 0.0526, "step": 18610 }, { "grad_norm": 0.5559726357460022, "learning_rate": 9.501425927263116e-05, "loss": 0.0631, "step": 18620 }, { "grad_norm": 0.3638702929019928, "learning_rate": 9.500705924790172e-05, "loss": 0.0583, "step": 18630 }, { "grad_norm": 0.4521275758743286, "learning_rate": 9.499985430126794e-05, "loss": 0.0557, "step": 18640 }, { "grad_norm": 0.6029384136199951, "learning_rate": 9.499264443351775e-05, "loss": 0.0539, "step": 18650 }, { "grad_norm": 0.36713311076164246, "learning_rate": 9.498542964543961e-05, "loss": 0.0558, "step": 18660 }, { "grad_norm": 0.44600293040275574, "learning_rate": 9.497820993782252e-05, "loss": 0.0508, "step": 18670 }, { "grad_norm": 0.36694836616516113, "learning_rate": 9.497098531145601e-05, "loss": 0.0599, "step": 18680 }, { "grad_norm": 0.5181728601455688, "learning_rate": 9.496375576713017e-05, "loss": 0.0594, "step": 18690 }, { "grad_norm": 0.32753172516822815, "learning_rate": 9.49565213056356e-05, "loss": 0.0556, "step": 18700 }, { "grad_norm": 0.5832945108413696, "learning_rate": 9.494928192776342e-05, "loss": 0.0572, "step": 18710 }, { "grad_norm": 0.3308200538158417, "learning_rate": 9.494203763430538e-05, "loss": 0.0562, "step": 18720 }, { "grad_norm": 0.32742443680763245, "learning_rate": 9.493478842605366e-05, "loss": 0.0534, "step": 18730 }, { "grad_norm": 0.44621068239212036, "learning_rate": 9.492753430380105e-05, "loss": 0.0632, "step": 18740 }, { "grad_norm": 0.4242573082447052, "learning_rate": 9.492027526834083e-05, "loss": 0.0611, "step": 18750 }, { "grad_norm": 0.5156534910202026, "learning_rate": 9.491301132046684e-05, "loss": 0.0645, "step": 18760 }, { "grad_norm": 0.6498997807502747, "learning_rate": 9.490574246097345e-05, "loss": 0.0528, "step": 18770 }, { "grad_norm": 0.3869903087615967, "learning_rate": 9.48984686906556e-05, "loss": 0.0544, "step": 18780 }, { "grad_norm": 0.649190366268158, "learning_rate": 9.489119001030871e-05, "loss": 0.0561, "step": 18790 }, { "grad_norm": 0.43795299530029297, "learning_rate": 9.488390642072878e-05, "loss": 0.0597, "step": 18800 }, { "grad_norm": 0.38436251878738403, "learning_rate": 9.48766179227123e-05, "loss": 0.0516, "step": 18810 }, { "grad_norm": 0.44505682587623596, "learning_rate": 9.486932451705636e-05, "loss": 0.0521, "step": 18820 }, { "grad_norm": 0.5509033799171448, "learning_rate": 9.486202620455857e-05, "loss": 0.0603, "step": 18830 }, { "grad_norm": 0.41587522625923157, "learning_rate": 9.485472298601704e-05, "loss": 0.0622, "step": 18840 }, { "grad_norm": 0.6115548014640808, "learning_rate": 9.484741486223043e-05, "loss": 0.0618, "step": 18850 }, { "grad_norm": 0.321826696395874, "learning_rate": 9.484010183399797e-05, "loss": 0.0585, "step": 18860 }, { "grad_norm": 0.41590040922164917, "learning_rate": 9.483278390211938e-05, "loss": 0.0538, "step": 18870 }, { "grad_norm": 0.34153497219085693, "learning_rate": 9.482546106739496e-05, "loss": 0.0613, "step": 18880 }, { "grad_norm": 0.512162446975708, "learning_rate": 9.48181333306255e-05, "loss": 0.0571, "step": 18890 }, { "grad_norm": 0.6124624609947205, "learning_rate": 9.481080069261237e-05, "loss": 0.0624, "step": 18900 }, { "grad_norm": 0.30260831117630005, "learning_rate": 9.480346315415745e-05, "loss": 0.0549, "step": 18910 }, { "grad_norm": 0.4105908274650574, "learning_rate": 9.479612071606314e-05, "loss": 0.0628, "step": 18920 }, { "grad_norm": 0.7662938237190247, "learning_rate": 9.478877337913244e-05, "loss": 0.0587, "step": 18930 }, { "grad_norm": 0.4757697582244873, "learning_rate": 9.478142114416881e-05, "loss": 0.0584, "step": 18940 }, { "grad_norm": 0.5214835405349731, "learning_rate": 9.47740640119763e-05, "loss": 0.0598, "step": 18950 }, { "grad_norm": 0.38667893409729004, "learning_rate": 9.476670198335947e-05, "loss": 0.0574, "step": 18960 }, { "grad_norm": 0.4274574816226959, "learning_rate": 9.47593350591234e-05, "loss": 0.0634, "step": 18970 }, { "grad_norm": 0.42386025190353394, "learning_rate": 9.475196324007376e-05, "loss": 0.055, "step": 18980 }, { "grad_norm": 0.48689982295036316, "learning_rate": 9.474458652701669e-05, "loss": 0.0555, "step": 18990 }, { "grad_norm": 0.35662463307380676, "learning_rate": 9.473720492075892e-05, "loss": 0.0545, "step": 19000 }, { "grad_norm": 0.4051712155342102, "learning_rate": 9.472981842210768e-05, "loss": 0.0584, "step": 19010 }, { "grad_norm": 0.5358332991600037, "learning_rate": 9.472242703187074e-05, "loss": 0.0566, "step": 19020 }, { "grad_norm": 0.4209580421447754, "learning_rate": 9.471503075085643e-05, "loss": 0.0535, "step": 19030 }, { "grad_norm": 0.5467994213104248, "learning_rate": 9.470762957987359e-05, "loss": 0.0549, "step": 19040 }, { "grad_norm": 0.4965786337852478, "learning_rate": 9.470022351973158e-05, "loss": 0.0515, "step": 19050 }, { "grad_norm": 0.6412773132324219, "learning_rate": 9.469281257124034e-05, "loss": 0.0524, "step": 19060 }, { "grad_norm": 0.4981614649295807, "learning_rate": 9.46853967352103e-05, "loss": 0.0592, "step": 19070 }, { "grad_norm": 0.433228075504303, "learning_rate": 9.467797601245246e-05, "loss": 0.0624, "step": 19080 }, { "grad_norm": 0.34981095790863037, "learning_rate": 9.467055040377834e-05, "loss": 0.0658, "step": 19090 }, { "grad_norm": 0.4062668979167938, "learning_rate": 9.466311990999999e-05, "loss": 0.0626, "step": 19100 }, { "grad_norm": 0.5059645771980286, "learning_rate": 9.465568453193e-05, "loss": 0.0598, "step": 19110 }, { "grad_norm": 0.3984818160533905, "learning_rate": 9.464824427038148e-05, "loss": 0.065, "step": 19120 }, { "grad_norm": 0.39399248361587524, "learning_rate": 9.46407991261681e-05, "loss": 0.0544, "step": 19130 }, { "grad_norm": 0.3551129996776581, "learning_rate": 9.463334910010404e-05, "loss": 0.0599, "step": 19140 }, { "grad_norm": 0.4824344217777252, "learning_rate": 9.462589419300403e-05, "loss": 0.0712, "step": 19150 }, { "grad_norm": 0.3960109055042267, "learning_rate": 9.461843440568333e-05, "loss": 0.0589, "step": 19160 }, { "grad_norm": 0.27233871817588806, "learning_rate": 9.461096973895773e-05, "loss": 0.0548, "step": 19170 }, { "grad_norm": 0.5291441679000854, "learning_rate": 9.460350019364355e-05, "loss": 0.0652, "step": 19180 }, { "grad_norm": 0.4366440773010254, "learning_rate": 9.459602577055764e-05, "loss": 0.0583, "step": 19190 }, { "grad_norm": 0.44121435284614563, "learning_rate": 9.45885464705174e-05, "loss": 0.0525, "step": 19200 }, { "grad_norm": 0.39242497086524963, "learning_rate": 9.458106229434076e-05, "loss": 0.0629, "step": 19210 }, { "grad_norm": 0.4814622104167938, "learning_rate": 9.457357324284617e-05, "loss": 0.0551, "step": 19220 }, { "grad_norm": 0.2878376543521881, "learning_rate": 9.456607931685262e-05, "loss": 0.058, "step": 19230 }, { "grad_norm": 0.3160136044025421, "learning_rate": 9.455858051717965e-05, "loss": 0.0544, "step": 19240 }, { "grad_norm": 0.39687588810920715, "learning_rate": 9.45510768446473e-05, "loss": 0.059, "step": 19250 }, { "grad_norm": 0.4901356101036072, "learning_rate": 9.454356830007618e-05, "loss": 0.062, "step": 19260 }, { "grad_norm": 0.452955037355423, "learning_rate": 9.45360548842874e-05, "loss": 0.0676, "step": 19270 }, { "grad_norm": 0.4356793463230133, "learning_rate": 9.452853659810261e-05, "loss": 0.0528, "step": 19280 }, { "grad_norm": 0.3483065664768219, "learning_rate": 9.452101344234401e-05, "loss": 0.0588, "step": 19290 }, { "grad_norm": 0.517443060874939, "learning_rate": 9.451348541783431e-05, "loss": 0.0635, "step": 19300 }, { "grad_norm": 0.4147803485393524, "learning_rate": 9.450595252539678e-05, "loss": 0.0688, "step": 19310 }, { "grad_norm": 0.4545131027698517, "learning_rate": 9.449841476585518e-05, "loss": 0.0617, "step": 19320 }, { "grad_norm": 0.5326856970787048, "learning_rate": 9.449087214003384e-05, "loss": 0.0535, "step": 19330 }, { "grad_norm": 0.3260642886161804, "learning_rate": 9.448332464875765e-05, "loss": 0.0594, "step": 19340 }, { "grad_norm": 0.4285831153392792, "learning_rate": 9.447577229285192e-05, "loss": 0.0546, "step": 19350 }, { "grad_norm": 0.5416764616966248, "learning_rate": 9.446821507314261e-05, "loss": 0.0542, "step": 19360 }, { "grad_norm": 0.3923390209674835, "learning_rate": 9.446065299045617e-05, "loss": 0.0532, "step": 19370 }, { "grad_norm": 0.4689120650291443, "learning_rate": 9.445308604561955e-05, "loss": 0.0628, "step": 19380 }, { "grad_norm": 0.5193429589271545, "learning_rate": 9.444551423946028e-05, "loss": 0.0629, "step": 19390 }, { "grad_norm": 0.47725382447242737, "learning_rate": 9.443793757280638e-05, "loss": 0.0689, "step": 19400 }, { "grad_norm": 0.3584183156490326, "learning_rate": 9.443035604648646e-05, "loss": 0.058, "step": 19410 }, { "grad_norm": 0.28945106267929077, "learning_rate": 9.44227696613296e-05, "loss": 0.0585, "step": 19420 }, { "grad_norm": 0.463171124458313, "learning_rate": 9.441517841816542e-05, "loss": 0.0608, "step": 19430 }, { "grad_norm": 0.37339168787002563, "learning_rate": 9.440758231782413e-05, "loss": 0.06, "step": 19440 }, { "grad_norm": 0.46366479992866516, "learning_rate": 9.439998136113639e-05, "loss": 0.0574, "step": 19450 }, { "grad_norm": 0.3135530352592468, "learning_rate": 9.439237554893344e-05, "loss": 0.0602, "step": 19460 }, { "grad_norm": 0.37881582975387573, "learning_rate": 9.438476488204705e-05, "loss": 0.0533, "step": 19470 }, { "grad_norm": 0.35618162155151367, "learning_rate": 9.43771493613095e-05, "loss": 0.0512, "step": 19480 }, { "grad_norm": 0.4193100035190582, "learning_rate": 9.436952898755362e-05, "loss": 0.0531, "step": 19490 }, { "grad_norm": 0.5892905592918396, "learning_rate": 9.436190376161276e-05, "loss": 0.0643, "step": 19500 }, { "grad_norm": 0.4238111674785614, "learning_rate": 9.43542736843208e-05, "loss": 0.0613, "step": 19510 }, { "grad_norm": 0.3981034755706787, "learning_rate": 9.434663875651216e-05, "loss": 0.0606, "step": 19520 }, { "grad_norm": 0.6386468410491943, "learning_rate": 9.433899897902177e-05, "loss": 0.0616, "step": 19530 }, { "grad_norm": 0.42628446221351624, "learning_rate": 9.433135435268511e-05, "loss": 0.0545, "step": 19540 }, { "grad_norm": 0.5508812069892883, "learning_rate": 9.432370487833819e-05, "loss": 0.0508, "step": 19550 }, { "grad_norm": 0.35940417647361755, "learning_rate": 9.431605055681756e-05, "loss": 0.0541, "step": 19560 }, { "grad_norm": 0.4227598309516907, "learning_rate": 9.430839138896026e-05, "loss": 0.0581, "step": 19570 }, { "grad_norm": 0.4297493100166321, "learning_rate": 9.43007273756039e-05, "loss": 0.0524, "step": 19580 }, { "grad_norm": 0.3646757900714874, "learning_rate": 9.429305851758658e-05, "loss": 0.052, "step": 19590 }, { "grad_norm": 0.31453877687454224, "learning_rate": 9.428538481574699e-05, "loss": 0.0477, "step": 19600 }, { "grad_norm": 0.341958224773407, "learning_rate": 9.42777062709243e-05, "loss": 0.0708, "step": 19610 }, { "grad_norm": 0.5104029178619385, "learning_rate": 9.427002288395821e-05, "loss": 0.0595, "step": 19620 }, { "grad_norm": 0.6182937622070312, "learning_rate": 9.426233465568898e-05, "loss": 0.0577, "step": 19630 }, { "grad_norm": 0.43157780170440674, "learning_rate": 9.42546415869574e-05, "loss": 0.0565, "step": 19640 }, { "grad_norm": 0.679278552532196, "learning_rate": 9.424694367860473e-05, "loss": 0.0641, "step": 19650 }, { "grad_norm": 0.41356518864631653, "learning_rate": 9.423924093147284e-05, "loss": 0.0533, "step": 19660 }, { "grad_norm": 0.44230765104293823, "learning_rate": 9.423153334640407e-05, "loss": 0.0531, "step": 19670 }, { "grad_norm": 0.35940173268318176, "learning_rate": 9.42238209242413e-05, "loss": 0.057, "step": 19680 }, { "grad_norm": 0.41681525111198425, "learning_rate": 9.421610366582798e-05, "loss": 0.0648, "step": 19690 }, { "grad_norm": 0.38304632902145386, "learning_rate": 9.420838157200803e-05, "loss": 0.0583, "step": 19700 }, { "grad_norm": 0.5558722615242004, "learning_rate": 9.420065464362594e-05, "loss": 0.0612, "step": 19710 }, { "grad_norm": 0.4331056475639343, "learning_rate": 9.419292288152673e-05, "loss": 0.0664, "step": 19720 }, { "grad_norm": 0.5228866338729858, "learning_rate": 9.418518628655588e-05, "loss": 0.0553, "step": 19730 }, { "grad_norm": 0.33105671405792236, "learning_rate": 9.417744485955951e-05, "loss": 0.0553, "step": 19740 }, { "grad_norm": 0.5162004232406616, "learning_rate": 9.41696986013842e-05, "loss": 0.0509, "step": 19750 }, { "grad_norm": 0.39760154485702515, "learning_rate": 9.416194751287705e-05, "loss": 0.0602, "step": 19760 }, { "grad_norm": 0.39983147382736206, "learning_rate": 9.415419159488572e-05, "loss": 0.055, "step": 19770 }, { "grad_norm": 0.48378297686576843, "learning_rate": 9.414643084825837e-05, "loss": 0.0547, "step": 19780 }, { "grad_norm": 0.5567028522491455, "learning_rate": 9.413866527384372e-05, "loss": 0.0613, "step": 19790 }, { "grad_norm": 0.45518213510513306, "learning_rate": 9.4130894872491e-05, "loss": 0.0534, "step": 19800 }, { "grad_norm": 0.3623712360858917, "learning_rate": 9.412311964504998e-05, "loss": 0.0573, "step": 19810 }, { "grad_norm": 0.3499563932418823, "learning_rate": 9.411533959237091e-05, "loss": 0.0532, "step": 19820 }, { "grad_norm": 0.414948970079422, "learning_rate": 9.410755471530464e-05, "loss": 0.0601, "step": 19830 }, { "grad_norm": 0.3348436653614044, "learning_rate": 9.40997650147025e-05, "loss": 0.0554, "step": 19840 }, { "grad_norm": 0.4852510690689087, "learning_rate": 9.409197049141637e-05, "loss": 0.0578, "step": 19850 }, { "grad_norm": 0.4678589105606079, "learning_rate": 9.408417114629863e-05, "loss": 0.0536, "step": 19860 }, { "grad_norm": 0.31124168634414673, "learning_rate": 9.40763669802022e-05, "loss": 0.0499, "step": 19870 }, { "grad_norm": 0.38891980051994324, "learning_rate": 9.406855799398056e-05, "loss": 0.0544, "step": 19880 }, { "grad_norm": 0.4852403700351715, "learning_rate": 9.406074418848767e-05, "loss": 0.0639, "step": 19890 }, { "grad_norm": 0.4470217525959015, "learning_rate": 9.405292556457805e-05, "loss": 0.0585, "step": 19900 }, { "grad_norm": 0.3669516146183014, "learning_rate": 9.404510212310671e-05, "loss": 0.0507, "step": 19910 }, { "grad_norm": 0.3459717631340027, "learning_rate": 9.403727386492924e-05, "loss": 0.0504, "step": 19920 }, { "grad_norm": 0.3566475212574005, "learning_rate": 9.40294407909017e-05, "loss": 0.0536, "step": 19930 }, { "grad_norm": 0.4792861342430115, "learning_rate": 9.40216029018807e-05, "loss": 0.0535, "step": 19940 }, { "grad_norm": 0.4433431923389435, "learning_rate": 9.401376019872338e-05, "loss": 0.0562, "step": 19950 }, { "grad_norm": 0.3088551461696625, "learning_rate": 9.400591268228746e-05, "loss": 0.0532, "step": 19960 }, { "grad_norm": 0.47570809721946716, "learning_rate": 9.399806035343106e-05, "loss": 0.0628, "step": 19970 }, { "grad_norm": 0.48779726028442383, "learning_rate": 9.399020321301294e-05, "loss": 0.06, "step": 19980 }, { "grad_norm": 0.3478095531463623, "learning_rate": 9.398234126189234e-05, "loss": 0.0559, "step": 19990 }, { "grad_norm": 0.30134451389312744, "learning_rate": 9.397447450092902e-05, "loss": 0.0552, "step": 20000 }, { "grad_norm": 0.4704899787902832, "learning_rate": 9.39666029309833e-05, "loss": 0.0602, "step": 20010 }, { "grad_norm": 0.32345494627952576, "learning_rate": 9.395872655291596e-05, "loss": 0.0522, "step": 20020 }, { "grad_norm": 0.5011016130447388, "learning_rate": 9.395084536758838e-05, "loss": 0.047, "step": 20030 }, { "grad_norm": 0.5626744031906128, "learning_rate": 9.394295937586243e-05, "loss": 0.0608, "step": 20040 }, { "grad_norm": 0.39614295959472656, "learning_rate": 9.393506857860052e-05, "loss": 0.0608, "step": 20050 }, { "grad_norm": 0.5476111769676208, "learning_rate": 9.392717297666555e-05, "loss": 0.0632, "step": 20060 }, { "grad_norm": 0.5819023847579956, "learning_rate": 9.391927257092101e-05, "loss": 0.0593, "step": 20070 }, { "grad_norm": 0.49878931045532227, "learning_rate": 9.391136736223085e-05, "loss": 0.0603, "step": 20080 }, { "grad_norm": 0.3873739242553711, "learning_rate": 9.390345735145956e-05, "loss": 0.0495, "step": 20090 }, { "grad_norm": 0.3282679319381714, "learning_rate": 9.389554253947219e-05, "loss": 0.0525, "step": 20100 }, { "grad_norm": 0.47419100999832153, "learning_rate": 9.388762292713428e-05, "loss": 0.0635, "step": 20110 }, { "grad_norm": 0.4388660490512848, "learning_rate": 9.38796985153119e-05, "loss": 0.0601, "step": 20120 }, { "grad_norm": 0.3119131922721863, "learning_rate": 9.387176930487169e-05, "loss": 0.0575, "step": 20130 }, { "grad_norm": 0.43252235651016235, "learning_rate": 9.386383529668072e-05, "loss": 0.0547, "step": 20140 }, { "grad_norm": 0.4138001799583435, "learning_rate": 9.385589649160669e-05, "loss": 0.0677, "step": 20150 }, { "grad_norm": 0.40541014075279236, "learning_rate": 9.384795289051775e-05, "loss": 0.0566, "step": 20160 }, { "grad_norm": 0.325931578874588, "learning_rate": 9.384000449428261e-05, "loss": 0.0608, "step": 20170 }, { "grad_norm": 0.5797897577285767, "learning_rate": 9.383205130377048e-05, "loss": 0.0623, "step": 20180 }, { "grad_norm": 0.44114065170288086, "learning_rate": 9.382409331985114e-05, "loss": 0.0531, "step": 20190 }, { "grad_norm": 0.5164070725440979, "learning_rate": 9.381613054339482e-05, "loss": 0.0553, "step": 20200 }, { "grad_norm": 0.42703935503959656, "learning_rate": 9.380816297527235e-05, "loss": 0.0606, "step": 20210 }, { "grad_norm": 0.3370570242404938, "learning_rate": 9.380019061635506e-05, "loss": 0.0558, "step": 20220 }, { "grad_norm": 0.5855211019515991, "learning_rate": 9.379221346751474e-05, "loss": 0.0535, "step": 20230 }, { "grad_norm": 0.4251437783241272, "learning_rate": 9.378423152962382e-05, "loss": 0.0538, "step": 20240 }, { "grad_norm": 0.3887021243572235, "learning_rate": 9.377624480355517e-05, "loss": 0.0592, "step": 20250 }, { "grad_norm": 0.5532496571540833, "learning_rate": 9.376825329018219e-05, "loss": 0.0688, "step": 20260 }, { "grad_norm": 0.4876416325569153, "learning_rate": 9.376025699037884e-05, "loss": 0.062, "step": 20270 }, { "grad_norm": 0.32053256034851074, "learning_rate": 9.37522559050196e-05, "loss": 0.0571, "step": 20280 }, { "grad_norm": 0.4488588869571686, "learning_rate": 9.37442500349794e-05, "loss": 0.0598, "step": 20290 }, { "grad_norm": 0.8317111134529114, "learning_rate": 9.373623938113381e-05, "loss": 0.064, "step": 20300 }, { "grad_norm": 0.558347761631012, "learning_rate": 9.372822394435883e-05, "loss": 0.0632, "step": 20310 }, { "grad_norm": 0.5389313101768494, "learning_rate": 9.372020372553102e-05, "loss": 0.0594, "step": 20320 }, { "grad_norm": 0.5012449622154236, "learning_rate": 9.371217872552746e-05, "loss": 0.0631, "step": 20330 }, { "grad_norm": 0.4774869978427887, "learning_rate": 9.370414894522576e-05, "loss": 0.0565, "step": 20340 }, { "grad_norm": 0.6094645857810974, "learning_rate": 9.369611438550406e-05, "loss": 0.0563, "step": 20350 }, { "grad_norm": 0.5134311318397522, "learning_rate": 9.368807504724095e-05, "loss": 0.0566, "step": 20360 }, { "grad_norm": 0.7713476419448853, "learning_rate": 9.368003093131565e-05, "loss": 0.0617, "step": 20370 }, { "grad_norm": 0.4748772978782654, "learning_rate": 9.367198203860785e-05, "loss": 0.0606, "step": 20380 }, { "grad_norm": 0.8537551164627075, "learning_rate": 9.366392836999774e-05, "loss": 0.06, "step": 20390 }, { "grad_norm": 0.3044625222682953, "learning_rate": 9.365586992636607e-05, "loss": 0.0498, "step": 20400 }, { "grad_norm": 0.4421145021915436, "learning_rate": 9.364780670859412e-05, "loss": 0.0667, "step": 20410 }, { "grad_norm": 0.40587151050567627, "learning_rate": 9.363973871756364e-05, "loss": 0.0636, "step": 20420 }, { "grad_norm": 0.39137712121009827, "learning_rate": 9.363166595415696e-05, "loss": 0.0691, "step": 20430 }, { "grad_norm": 0.5318548679351807, "learning_rate": 9.362358841925686e-05, "loss": 0.0564, "step": 20440 }, { "grad_norm": 0.4302903115749359, "learning_rate": 9.361550611374674e-05, "loss": 0.0582, "step": 20450 }, { "grad_norm": 0.5484476685523987, "learning_rate": 9.360741903851043e-05, "loss": 0.0572, "step": 20460 }, { "grad_norm": 0.48172760009765625, "learning_rate": 9.359932719443236e-05, "loss": 0.063, "step": 20470 }, { "grad_norm": 0.40850064158439636, "learning_rate": 9.35912305823974e-05, "loss": 0.0568, "step": 20480 }, { "grad_norm": 0.45224708318710327, "learning_rate": 9.358312920329101e-05, "loss": 0.0597, "step": 20490 }, { "grad_norm": 0.3955809772014618, "learning_rate": 9.357502305799914e-05, "loss": 0.0514, "step": 20500 }, { "grad_norm": 0.45544153451919556, "learning_rate": 9.356691214740824e-05, "loss": 0.0536, "step": 20510 }, { "grad_norm": 0.35767489671707153, "learning_rate": 9.355879647240535e-05, "loss": 0.0665, "step": 20520 }, { "grad_norm": 0.37755638360977173, "learning_rate": 9.355067603387798e-05, "loss": 0.0531, "step": 20530 }, { "grad_norm": 0.40558815002441406, "learning_rate": 9.354255083271412e-05, "loss": 0.0566, "step": 20540 }, { "grad_norm": 0.38427597284317017, "learning_rate": 9.353442086980239e-05, "loss": 0.0504, "step": 20550 }, { "grad_norm": 0.3623058795928955, "learning_rate": 9.352628614603185e-05, "loss": 0.0522, "step": 20560 }, { "grad_norm": 0.4467846751213074, "learning_rate": 9.351814666229209e-05, "loss": 0.0462, "step": 20570 }, { "grad_norm": 0.4767412543296814, "learning_rate": 9.351000241947324e-05, "loss": 0.0579, "step": 20580 }, { "grad_norm": 0.3971506655216217, "learning_rate": 9.350185341846594e-05, "loss": 0.0521, "step": 20590 }, { "grad_norm": 0.30865177512168884, "learning_rate": 9.349369966016134e-05, "loss": 0.0518, "step": 20600 }, { "grad_norm": 0.413872092962265, "learning_rate": 9.348554114545117e-05, "loss": 0.0557, "step": 20610 }, { "grad_norm": 0.4732162356376648, "learning_rate": 9.347737787522758e-05, "loss": 0.054, "step": 20620 }, { "grad_norm": 0.5239549279212952, "learning_rate": 9.346920985038332e-05, "loss": 0.0653, "step": 20630 }, { "grad_norm": 0.39418238401412964, "learning_rate": 9.346103707181162e-05, "loss": 0.0499, "step": 20640 }, { "grad_norm": 0.38619840145111084, "learning_rate": 9.345285954040626e-05, "loss": 0.0605, "step": 20650 }, { "grad_norm": 0.3500780463218689, "learning_rate": 9.34446772570615e-05, "loss": 0.0532, "step": 20660 }, { "grad_norm": 0.36868464946746826, "learning_rate": 9.343649022267214e-05, "loss": 0.0557, "step": 20670 }, { "grad_norm": 0.5107765197753906, "learning_rate": 9.342829843813353e-05, "loss": 0.0515, "step": 20680 }, { "grad_norm": 0.375772088766098, "learning_rate": 9.342010190434149e-05, "loss": 0.0605, "step": 20690 }, { "grad_norm": 0.4494647681713104, "learning_rate": 9.34119006221924e-05, "loss": 0.074, "step": 20700 }, { "grad_norm": 0.547175407409668, "learning_rate": 9.340369459258313e-05, "loss": 0.0609, "step": 20710 }, { "grad_norm": 0.4422300159931183, "learning_rate": 9.339548381641106e-05, "loss": 0.0651, "step": 20720 }, { "grad_norm": 0.5071959495544434, "learning_rate": 9.338726829457413e-05, "loss": 0.0507, "step": 20730 }, { "grad_norm": 0.4587486982345581, "learning_rate": 9.337904802797078e-05, "loss": 0.0532, "step": 20740 }, { "grad_norm": 0.37307584285736084, "learning_rate": 9.337082301749993e-05, "loss": 0.0537, "step": 20750 }, { "grad_norm": 0.37573620676994324, "learning_rate": 9.336259326406109e-05, "loss": 0.0516, "step": 20760 }, { "grad_norm": 0.39322376251220703, "learning_rate": 9.335435876855427e-05, "loss": 0.0569, "step": 20770 }, { "grad_norm": 0.33820661902427673, "learning_rate": 9.334611953187994e-05, "loss": 0.0575, "step": 20780 }, { "grad_norm": 0.3066241443157196, "learning_rate": 9.333787555493914e-05, "loss": 0.057, "step": 20790 }, { "grad_norm": 0.4057273268699646, "learning_rate": 9.332962683863345e-05, "loss": 0.0517, "step": 20800 }, { "grad_norm": 0.47438475489616394, "learning_rate": 9.332137338386489e-05, "loss": 0.0527, "step": 20810 }, { "grad_norm": 0.5628864765167236, "learning_rate": 9.33131151915361e-05, "loss": 0.0523, "step": 20820 }, { "grad_norm": 0.36291784048080444, "learning_rate": 9.330485226255012e-05, "loss": 0.0518, "step": 20830 }, { "grad_norm": 0.26233160495758057, "learning_rate": 9.329658459781061e-05, "loss": 0.0478, "step": 20840 }, { "grad_norm": 0.3467245399951935, "learning_rate": 9.328831219822172e-05, "loss": 0.0515, "step": 20850 }, { "grad_norm": 0.47344502806663513, "learning_rate": 9.328003506468808e-05, "loss": 0.0558, "step": 20860 }, { "grad_norm": 0.40989920496940613, "learning_rate": 9.327175319811488e-05, "loss": 0.0523, "step": 20870 }, { "grad_norm": 0.3600519299507141, "learning_rate": 9.326346659940781e-05, "loss": 0.0548, "step": 20880 }, { "grad_norm": 0.3988358676433563, "learning_rate": 9.325517526947308e-05, "loss": 0.0496, "step": 20890 }, { "grad_norm": 0.4483540952205658, "learning_rate": 9.32468792092174e-05, "loss": 0.0552, "step": 20900 }, { "grad_norm": 0.3045106828212738, "learning_rate": 9.323857841954803e-05, "loss": 0.0498, "step": 20910 }, { "grad_norm": 0.3253291845321655, "learning_rate": 9.323027290137276e-05, "loss": 0.0514, "step": 20920 }, { "grad_norm": 0.3109607994556427, "learning_rate": 9.322196265559981e-05, "loss": 0.0442, "step": 20930 }, { "grad_norm": 0.40161845088005066, "learning_rate": 9.321364768313803e-05, "loss": 0.0609, "step": 20940 }, { "grad_norm": 0.4667169749736786, "learning_rate": 9.32053279848967e-05, "loss": 0.0594, "step": 20950 }, { "grad_norm": 0.46179673075675964, "learning_rate": 9.319700356178567e-05, "loss": 0.0578, "step": 20960 }, { "grad_norm": 0.4184376299381256, "learning_rate": 9.318867441471527e-05, "loss": 0.0586, "step": 20970 }, { "grad_norm": 0.36458009481430054, "learning_rate": 9.318034054459637e-05, "loss": 0.0537, "step": 20980 }, { "grad_norm": 0.39576250314712524, "learning_rate": 9.317200195234034e-05, "loss": 0.0589, "step": 20990 }, { "grad_norm": 0.3697263300418854, "learning_rate": 9.316365863885909e-05, "loss": 0.0507, "step": 21000 }, { "grad_norm": 0.2987929880619049, "learning_rate": 9.315531060506502e-05, "loss": 0.046, "step": 21010 }, { "grad_norm": 0.39229944348335266, "learning_rate": 9.314695785187108e-05, "loss": 0.0518, "step": 21020 }, { "grad_norm": 0.7040982246398926, "learning_rate": 9.313860038019069e-05, "loss": 0.0561, "step": 21030 }, { "grad_norm": 0.4406619071960449, "learning_rate": 9.313023819093782e-05, "loss": 0.0599, "step": 21040 }, { "grad_norm": 0.41585996747016907, "learning_rate": 9.312187128502695e-05, "loss": 0.0503, "step": 21050 }, { "grad_norm": 0.48494839668273926, "learning_rate": 9.311349966337307e-05, "loss": 0.0485, "step": 21060 }, { "grad_norm": 0.44904625415802, "learning_rate": 9.310512332689169e-05, "loss": 0.0577, "step": 21070 }, { "grad_norm": 0.368186354637146, "learning_rate": 9.309674227649883e-05, "loss": 0.0542, "step": 21080 }, { "grad_norm": 0.4437543451786041, "learning_rate": 9.308835651311103e-05, "loss": 0.0645, "step": 21090 }, { "grad_norm": 0.3411634862422943, "learning_rate": 9.307996603764533e-05, "loss": 0.0544, "step": 21100 }, { "grad_norm": 0.47682732343673706, "learning_rate": 9.307157085101932e-05, "loss": 0.0536, "step": 21110 }, { "grad_norm": 0.36954113841056824, "learning_rate": 9.306317095415109e-05, "loss": 0.0546, "step": 21120 }, { "grad_norm": 0.5148501396179199, "learning_rate": 9.305476634795922e-05, "loss": 0.058, "step": 21130 }, { "grad_norm": 0.41154271364212036, "learning_rate": 9.304635703336284e-05, "loss": 0.0535, "step": 21140 }, { "grad_norm": 0.5486304759979248, "learning_rate": 9.303794301128157e-05, "loss": 0.0462, "step": 21150 }, { "grad_norm": 0.29528847336769104, "learning_rate": 9.302952428263555e-05, "loss": 0.053, "step": 21160 }, { "grad_norm": 0.38830265402793884, "learning_rate": 9.302110084834545e-05, "loss": 0.0594, "step": 21170 }, { "grad_norm": 0.5118646025657654, "learning_rate": 9.301267270933245e-05, "loss": 0.0579, "step": 21180 }, { "grad_norm": 0.4290807843208313, "learning_rate": 9.300423986651823e-05, "loss": 0.0622, "step": 21190 }, { "grad_norm": 0.30711522698402405, "learning_rate": 9.299580232082501e-05, "loss": 0.0512, "step": 21200 }, { "grad_norm": 0.40312501788139343, "learning_rate": 9.298736007317547e-05, "loss": 0.0582, "step": 21210 }, { "grad_norm": 0.3674461841583252, "learning_rate": 9.297891312449288e-05, "loss": 0.0558, "step": 21220 }, { "grad_norm": 0.38053956627845764, "learning_rate": 9.297046147570094e-05, "loss": 0.0575, "step": 21230 }, { "grad_norm": 0.45127424597740173, "learning_rate": 9.296200512772396e-05, "loss": 0.0589, "step": 21240 }, { "grad_norm": 0.41899773478507996, "learning_rate": 9.295354408148668e-05, "loss": 0.0547, "step": 21250 }, { "grad_norm": 0.4522240459918976, "learning_rate": 9.294507833791441e-05, "loss": 0.0578, "step": 21260 }, { "grad_norm": 0.6580183506011963, "learning_rate": 9.293660789793295e-05, "loss": 0.0561, "step": 21270 }, { "grad_norm": 0.6123034358024597, "learning_rate": 9.292813276246858e-05, "loss": 0.0578, "step": 21280 }, { "grad_norm": 0.4834691286087036, "learning_rate": 9.291965293244816e-05, "loss": 0.0573, "step": 21290 }, { "grad_norm": 0.3705079257488251, "learning_rate": 9.291116840879904e-05, "loss": 0.0547, "step": 21300 }, { "grad_norm": 0.42379334568977356, "learning_rate": 9.290267919244904e-05, "loss": 0.0669, "step": 21310 }, { "grad_norm": 0.34026220440864563, "learning_rate": 9.289418528432655e-05, "loss": 0.0501, "step": 21320 }, { "grad_norm": 0.34200602769851685, "learning_rate": 9.288568668536045e-05, "loss": 0.0511, "step": 21330 }, { "grad_norm": 0.4037744998931885, "learning_rate": 9.287718339648013e-05, "loss": 0.0576, "step": 21340 }, { "grad_norm": 0.41520074009895325, "learning_rate": 9.28686754186155e-05, "loss": 0.0583, "step": 21350 }, { "grad_norm": 0.3603334426879883, "learning_rate": 9.286016275269698e-05, "loss": 0.0505, "step": 21360 }, { "grad_norm": 0.48717373609542847, "learning_rate": 9.285164539965551e-05, "loss": 0.0452, "step": 21370 }, { "grad_norm": 0.3920605480670929, "learning_rate": 9.284312336042251e-05, "loss": 0.0541, "step": 21380 }, { "grad_norm": 0.30831146240234375, "learning_rate": 9.283459663592996e-05, "loss": 0.0576, "step": 21390 }, { "grad_norm": 0.39683908224105835, "learning_rate": 9.282606522711033e-05, "loss": 0.0545, "step": 21400 }, { "grad_norm": 0.37569162249565125, "learning_rate": 9.281752913489657e-05, "loss": 0.0528, "step": 21410 }, { "grad_norm": 0.40246766805648804, "learning_rate": 9.280898836022222e-05, "loss": 0.0485, "step": 21420 }, { "grad_norm": 0.40117812156677246, "learning_rate": 9.280044290402126e-05, "loss": 0.0549, "step": 21430 }, { "grad_norm": 0.37769046425819397, "learning_rate": 9.279189276722821e-05, "loss": 0.054, "step": 21440 }, { "grad_norm": 0.7636445760726929, "learning_rate": 9.278333795077812e-05, "loss": 0.0583, "step": 21450 }, { "grad_norm": 0.5018916130065918, "learning_rate": 9.27747784556065e-05, "loss": 0.0608, "step": 21460 }, { "grad_norm": 0.34372928738594055, "learning_rate": 9.276621428264942e-05, "loss": 0.0525, "step": 21470 }, { "grad_norm": 0.3915766477584839, "learning_rate": 9.275764543284345e-05, "loss": 0.0607, "step": 21480 }, { "grad_norm": 0.39390844106674194, "learning_rate": 9.274907190712566e-05, "loss": 0.0598, "step": 21490 }, { "grad_norm": 0.40775755047798157, "learning_rate": 9.274049370643363e-05, "loss": 0.0627, "step": 21500 }, { "grad_norm": 0.481908917427063, "learning_rate": 9.273191083170547e-05, "loss": 0.0582, "step": 21510 }, { "grad_norm": 1.0770611763000488, "learning_rate": 9.27233232838798e-05, "loss": 0.0551, "step": 21520 }, { "grad_norm": 0.4928950071334839, "learning_rate": 9.27147310638957e-05, "loss": 0.0664, "step": 21530 }, { "grad_norm": 0.3639961779117584, "learning_rate": 9.270613417269286e-05, "loss": 0.0605, "step": 21540 }, { "grad_norm": 0.3980923593044281, "learning_rate": 9.269753261121138e-05, "loss": 0.0531, "step": 21550 }, { "grad_norm": 0.3018175959587097, "learning_rate": 9.268892638039194e-05, "loss": 0.053, "step": 21560 }, { "grad_norm": 0.36135855317115784, "learning_rate": 9.268031548117569e-05, "loss": 0.0564, "step": 21570 }, { "grad_norm": 0.5731819272041321, "learning_rate": 9.26716999145043e-05, "loss": 0.0524, "step": 21580 }, { "grad_norm": 0.4316204786300659, "learning_rate": 9.266307968131998e-05, "loss": 0.0507, "step": 21590 }, { "grad_norm": 0.30617085099220276, "learning_rate": 9.26544547825654e-05, "loss": 0.0577, "step": 21600 }, { "grad_norm": 0.4417891800403595, "learning_rate": 9.264582521918376e-05, "loss": 0.0465, "step": 21610 }, { "grad_norm": 0.37509027123451233, "learning_rate": 9.263719099211881e-05, "loss": 0.0498, "step": 21620 }, { "grad_norm": 0.4136989414691925, "learning_rate": 9.262855210231476e-05, "loss": 0.0535, "step": 21630 }, { "grad_norm": 0.2902127802371979, "learning_rate": 9.261990855071633e-05, "loss": 0.0518, "step": 21640 }, { "grad_norm": 0.5423891544342041, "learning_rate": 9.261126033826878e-05, "loss": 0.0516, "step": 21650 }, { "grad_norm": 0.5232083797454834, "learning_rate": 9.260260746591786e-05, "loss": 0.0601, "step": 21660 }, { "grad_norm": 0.4918079078197479, "learning_rate": 9.259394993460985e-05, "loss": 0.0545, "step": 21670 }, { "grad_norm": 0.2797209322452545, "learning_rate": 9.258528774529151e-05, "loss": 0.0494, "step": 21680 }, { "grad_norm": 0.42265695333480835, "learning_rate": 9.257662089891013e-05, "loss": 0.0576, "step": 21690 }, { "grad_norm": 0.32700005173683167, "learning_rate": 9.25679493964135e-05, "loss": 0.0511, "step": 21700 }, { "grad_norm": 0.5956893563270569, "learning_rate": 9.255927323874994e-05, "loss": 0.0568, "step": 21710 }, { "grad_norm": 0.3722003996372223, "learning_rate": 9.255059242686822e-05, "loss": 0.051, "step": 21720 }, { "grad_norm": 0.38885241746902466, "learning_rate": 9.254190696171769e-05, "loss": 0.0522, "step": 21730 }, { "grad_norm": 0.522602379322052, "learning_rate": 9.25332168442482e-05, "loss": 0.0585, "step": 21740 }, { "grad_norm": 0.33352360129356384, "learning_rate": 9.252452207541004e-05, "loss": 0.0552, "step": 21750 }, { "grad_norm": 0.5152405500411987, "learning_rate": 9.251582265615409e-05, "loss": 0.0474, "step": 21760 }, { "grad_norm": 0.3508208692073822, "learning_rate": 9.250711858743169e-05, "loss": 0.0501, "step": 21770 }, { "grad_norm": 0.48531389236450195, "learning_rate": 9.24984098701947e-05, "loss": 0.0483, "step": 21780 }, { "grad_norm": 0.4049469232559204, "learning_rate": 9.248969650539552e-05, "loss": 0.0456, "step": 21790 }, { "grad_norm": 0.4564661383628845, "learning_rate": 9.2480978493987e-05, "loss": 0.0581, "step": 21800 }, { "grad_norm": 0.5198459029197693, "learning_rate": 9.247225583692256e-05, "loss": 0.0583, "step": 21810 }, { "grad_norm": 0.485309362411499, "learning_rate": 9.246352853515607e-05, "loss": 0.0476, "step": 21820 }, { "grad_norm": 0.37101519107818604, "learning_rate": 9.245479658964194e-05, "loss": 0.0543, "step": 21830 }, { "grad_norm": 0.3585911691188812, "learning_rate": 9.244606000133507e-05, "loss": 0.0588, "step": 21840 }, { "grad_norm": 0.6783564686775208, "learning_rate": 9.24373187711909e-05, "loss": 0.0584, "step": 21850 }, { "grad_norm": 0.4909589886665344, "learning_rate": 9.242857290016537e-05, "loss": 0.0582, "step": 21860 }, { "grad_norm": 0.789391279220581, "learning_rate": 9.241982238921488e-05, "loss": 0.0491, "step": 21870 }, { "grad_norm": 0.35330837965011597, "learning_rate": 9.24110672392964e-05, "loss": 0.0564, "step": 21880 }, { "grad_norm": 0.3337242901325226, "learning_rate": 9.240230745136737e-05, "loss": 0.0491, "step": 21890 }, { "grad_norm": 0.4089243710041046, "learning_rate": 9.239354302638575e-05, "loss": 0.0571, "step": 21900 }, { "grad_norm": 0.4532913267612457, "learning_rate": 9.238477396531e-05, "loss": 0.0515, "step": 21910 }, { "grad_norm": 0.36101382970809937, "learning_rate": 9.23760002690991e-05, "loss": 0.0549, "step": 21920 }, { "grad_norm": 0.3552088439464569, "learning_rate": 9.236722193871252e-05, "loss": 0.0544, "step": 21930 }, { "grad_norm": 0.4067046046257019, "learning_rate": 9.235843897511023e-05, "loss": 0.0466, "step": 21940 }, { "grad_norm": 0.411490797996521, "learning_rate": 9.234965137925276e-05, "loss": 0.0624, "step": 21950 }, { "grad_norm": 0.3650219440460205, "learning_rate": 9.234085915210108e-05, "loss": 0.0522, "step": 21960 }, { "grad_norm": 0.4480881690979004, "learning_rate": 9.23320622946167e-05, "loss": 0.0594, "step": 21970 }, { "grad_norm": 0.3051755130290985, "learning_rate": 9.232326080776163e-05, "loss": 0.0577, "step": 21980 }, { "grad_norm": 0.2633715867996216, "learning_rate": 9.23144546924984e-05, "loss": 0.0498, "step": 21990 }, { "grad_norm": 0.37730273604393005, "learning_rate": 9.230564394979e-05, "loss": 0.0462, "step": 22000 }, { "grad_norm": 0.4803430438041687, "learning_rate": 9.22968285806e-05, "loss": 0.0457, "step": 22010 }, { "grad_norm": 0.3491601347923279, "learning_rate": 9.228800858589242e-05, "loss": 0.0579, "step": 22020 }, { "grad_norm": 0.39407429099082947, "learning_rate": 9.227918396663179e-05, "loss": 0.051, "step": 22030 }, { "grad_norm": 0.38880664110183716, "learning_rate": 9.227035472378319e-05, "loss": 0.0517, "step": 22040 }, { "grad_norm": 0.4842284917831421, "learning_rate": 9.226152085831213e-05, "loss": 0.0629, "step": 22050 }, { "grad_norm": 0.41783368587493896, "learning_rate": 9.22526823711847e-05, "loss": 0.0534, "step": 22060 }, { "grad_norm": 0.356875479221344, "learning_rate": 9.224383926336745e-05, "loss": 0.052, "step": 22070 }, { "grad_norm": 0.4827050566673279, "learning_rate": 9.223499153582744e-05, "loss": 0.0589, "step": 22080 }, { "grad_norm": 0.360006719827652, "learning_rate": 9.222613918953226e-05, "loss": 0.0557, "step": 22090 }, { "grad_norm": 0.6619423031806946, "learning_rate": 9.221728222544999e-05, "loss": 0.0642, "step": 22100 }, { "grad_norm": 0.4607240855693817, "learning_rate": 9.22084206445492e-05, "loss": 0.0596, "step": 22110 }, { "grad_norm": 0.6020370721817017, "learning_rate": 9.2199554447799e-05, "loss": 0.0522, "step": 22120 }, { "grad_norm": 0.4556382894515991, "learning_rate": 9.219068363616897e-05, "loss": 0.0561, "step": 22130 }, { "grad_norm": 0.3500634431838989, "learning_rate": 9.218180821062919e-05, "loss": 0.05, "step": 22140 }, { "grad_norm": 0.4258211851119995, "learning_rate": 9.21729281721503e-05, "loss": 0.0536, "step": 22150 }, { "grad_norm": 0.8063543438911438, "learning_rate": 9.216404352170339e-05, "loss": 0.0557, "step": 22160 }, { "grad_norm": 0.36735883355140686, "learning_rate": 9.215515426026007e-05, "loss": 0.0514, "step": 22170 }, { "grad_norm": 0.5378040671348572, "learning_rate": 9.214626038879246e-05, "loss": 0.058, "step": 22180 }, { "grad_norm": 0.6177704334259033, "learning_rate": 9.21373619082732e-05, "loss": 0.0545, "step": 22190 }, { "grad_norm": 0.4133414328098297, "learning_rate": 9.212845881967535e-05, "loss": 0.054, "step": 22200 }, { "grad_norm": 0.37108370661735535, "learning_rate": 9.211955112397262e-05, "loss": 0.046, "step": 22210 }, { "grad_norm": 0.42189475893974304, "learning_rate": 9.211063882213909e-05, "loss": 0.0458, "step": 22220 }, { "grad_norm": 0.6873897910118103, "learning_rate": 9.210172191514942e-05, "loss": 0.0579, "step": 22230 }, { "grad_norm": 0.5561271905899048, "learning_rate": 9.209280040397874e-05, "loss": 0.0595, "step": 22240 }, { "grad_norm": 0.32203859090805054, "learning_rate": 9.208387428960268e-05, "loss": 0.0576, "step": 22250 }, { "grad_norm": 0.3248468339443207, "learning_rate": 9.20749435729974e-05, "loss": 0.0512, "step": 22260 }, { "grad_norm": 0.3157385289669037, "learning_rate": 9.206600825513957e-05, "loss": 0.0542, "step": 22270 }, { "grad_norm": 0.5297687649726868, "learning_rate": 9.20570683370063e-05, "loss": 0.0547, "step": 22280 }, { "grad_norm": 0.5019785165786743, "learning_rate": 9.204812381957528e-05, "loss": 0.0575, "step": 22290 }, { "grad_norm": 0.3597990870475769, "learning_rate": 9.203917470382465e-05, "loss": 0.0515, "step": 22300 }, { "grad_norm": 0.4949275255203247, "learning_rate": 9.203022099073309e-05, "loss": 0.0523, "step": 22310 }, { "grad_norm": 0.4506891369819641, "learning_rate": 9.202126268127976e-05, "loss": 0.0545, "step": 22320 }, { "grad_norm": 0.42301759123802185, "learning_rate": 9.20122997764443e-05, "loss": 0.0475, "step": 22330 }, { "grad_norm": 0.40368321537971497, "learning_rate": 9.200333227720692e-05, "loss": 0.0518, "step": 22340 }, { "grad_norm": 0.3725053071975708, "learning_rate": 9.199436018454826e-05, "loss": 0.0546, "step": 22350 }, { "grad_norm": 0.24899108707904816, "learning_rate": 9.198538349944952e-05, "loss": 0.0426, "step": 22360 }, { "grad_norm": 0.34320488572120667, "learning_rate": 9.197640222289234e-05, "loss": 0.0527, "step": 22370 }, { "grad_norm": 0.383869469165802, "learning_rate": 9.196741635585895e-05, "loss": 0.0605, "step": 22380 }, { "grad_norm": 0.42788782715797424, "learning_rate": 9.195842589933199e-05, "loss": 0.0479, "step": 22390 }, { "grad_norm": 0.6553761959075928, "learning_rate": 9.194943085429466e-05, "loss": 0.0549, "step": 22400 }, { "grad_norm": 0.37360554933547974, "learning_rate": 9.194043122173065e-05, "loss": 0.0652, "step": 22410 }, { "grad_norm": 0.4327080547809601, "learning_rate": 9.193142700262413e-05, "loss": 0.0535, "step": 22420 }, { "grad_norm": 0.4148058295249939, "learning_rate": 9.192241819795979e-05, "loss": 0.0598, "step": 22430 }, { "grad_norm": 0.5106410384178162, "learning_rate": 9.191340480872284e-05, "loss": 0.0554, "step": 22440 }, { "grad_norm": 0.6556318402290344, "learning_rate": 9.190438683589895e-05, "loss": 0.0578, "step": 22450 }, { "grad_norm": 0.4158819317817688, "learning_rate": 9.189536428047432e-05, "loss": 0.0493, "step": 22460 }, { "grad_norm": 0.3794823884963989, "learning_rate": 9.188633714343564e-05, "loss": 0.0452, "step": 22470 }, { "grad_norm": 0.3469582796096802, "learning_rate": 9.18773054257701e-05, "loss": 0.0568, "step": 22480 }, { "grad_norm": 0.5225470066070557, "learning_rate": 9.18682691284654e-05, "loss": 0.0456, "step": 22490 }, { "grad_norm": 0.4112301468849182, "learning_rate": 9.185922825250974e-05, "loss": 0.0581, "step": 22500 }, { "grad_norm": 0.5243736505508423, "learning_rate": 9.185018279889181e-05, "loss": 0.058, "step": 22510 }, { "grad_norm": 0.5034910440444946, "learning_rate": 9.184113276860082e-05, "loss": 0.0571, "step": 22520 }, { "grad_norm": 0.3507462739944458, "learning_rate": 9.183207816262645e-05, "loss": 0.066, "step": 22530 }, { "grad_norm": 0.41139233112335205, "learning_rate": 9.182301898195891e-05, "loss": 0.0582, "step": 22540 }, { "grad_norm": 0.3833296000957489, "learning_rate": 9.181395522758889e-05, "loss": 0.052, "step": 22550 }, { "grad_norm": 0.4468575716018677, "learning_rate": 9.180488690050759e-05, "loss": 0.0628, "step": 22560 }, { "grad_norm": 0.5201694369316101, "learning_rate": 9.179581400170671e-05, "loss": 0.0491, "step": 22570 }, { "grad_norm": 0.5823922753334045, "learning_rate": 9.178673653217845e-05, "loss": 0.0588, "step": 22580 }, { "grad_norm": 0.514213502407074, "learning_rate": 9.177765449291551e-05, "loss": 0.0561, "step": 22590 }, { "grad_norm": 0.5808743238449097, "learning_rate": 9.176856788491109e-05, "loss": 0.0513, "step": 22600 }, { "grad_norm": 0.3789944648742676, "learning_rate": 9.175947670915887e-05, "loss": 0.0529, "step": 22610 }, { "grad_norm": 0.3777036666870117, "learning_rate": 9.175038096665309e-05, "loss": 0.0538, "step": 22620 }, { "grad_norm": 0.4169272184371948, "learning_rate": 9.17412806583884e-05, "loss": 0.0554, "step": 22630 }, { "grad_norm": 0.39061522483825684, "learning_rate": 9.173217578536002e-05, "loss": 0.0544, "step": 22640 }, { "grad_norm": 1.7497146129608154, "learning_rate": 9.172306634856362e-05, "loss": 0.0557, "step": 22650 }, { "grad_norm": 0.5519297122955322, "learning_rate": 9.171395234899545e-05, "loss": 0.051, "step": 22660 }, { "grad_norm": 0.6606260538101196, "learning_rate": 9.170483378765214e-05, "loss": 0.0617, "step": 22670 }, { "grad_norm": 0.35495394468307495, "learning_rate": 9.169571066553091e-05, "loss": 0.05, "step": 22680 }, { "grad_norm": 0.44604265689849854, "learning_rate": 9.168658298362946e-05, "loss": 0.054, "step": 22690 }, { "grad_norm": 0.3944990038871765, "learning_rate": 9.167745074294598e-05, "loss": 0.0578, "step": 22700 }, { "grad_norm": 0.42654499411582947, "learning_rate": 9.166831394447913e-05, "loss": 0.048, "step": 22710 }, { "grad_norm": 0.5743934512138367, "learning_rate": 9.165917258922812e-05, "loss": 0.0516, "step": 22720 }, { "grad_norm": 0.36831191182136536, "learning_rate": 9.165002667819262e-05, "loss": 0.0461, "step": 22730 }, { "grad_norm": 0.3651168644428253, "learning_rate": 9.164087621237282e-05, "loss": 0.054, "step": 22740 }, { "grad_norm": 0.3545396327972412, "learning_rate": 9.163172119276942e-05, "loss": 0.0501, "step": 22750 }, { "grad_norm": 0.4090024530887604, "learning_rate": 9.162256162038358e-05, "loss": 0.0543, "step": 22760 }, { "grad_norm": 0.43156835436820984, "learning_rate": 9.161339749621698e-05, "loss": 0.0588, "step": 22770 }, { "grad_norm": 0.3255932033061981, "learning_rate": 9.160422882127177e-05, "loss": 0.0537, "step": 22780 }, { "grad_norm": 0.38434281945228577, "learning_rate": 9.159505559655069e-05, "loss": 0.0523, "step": 22790 }, { "grad_norm": 0.5399733781814575, "learning_rate": 9.158587782305684e-05, "loss": 0.0575, "step": 22800 }, { "grad_norm": 0.3974266052246094, "learning_rate": 9.157669550179391e-05, "loss": 0.0484, "step": 22810 }, { "grad_norm": 0.557065486907959, "learning_rate": 9.156750863376609e-05, "loss": 0.0554, "step": 22820 }, { "grad_norm": 0.48998919129371643, "learning_rate": 9.155831721997801e-05, "loss": 0.0582, "step": 22830 }, { "grad_norm": 0.422426700592041, "learning_rate": 9.154912126143484e-05, "loss": 0.0559, "step": 22840 }, { "grad_norm": 0.4745309352874756, "learning_rate": 9.153992075914224e-05, "loss": 0.0584, "step": 22850 }, { "grad_norm": 0.32638996839523315, "learning_rate": 9.153071571410635e-05, "loss": 0.0521, "step": 22860 }, { "grad_norm": 0.3536710739135742, "learning_rate": 9.152150612733384e-05, "loss": 0.0527, "step": 22870 }, { "grad_norm": 0.5844343900680542, "learning_rate": 9.151229199983184e-05, "loss": 0.0538, "step": 22880 }, { "grad_norm": 0.42106199264526367, "learning_rate": 9.150307333260802e-05, "loss": 0.0552, "step": 22890 }, { "grad_norm": 0.41733676195144653, "learning_rate": 9.149385012667048e-05, "loss": 0.0474, "step": 22900 }, { "grad_norm": 0.6328219175338745, "learning_rate": 9.148462238302788e-05, "loss": 0.0537, "step": 22910 }, { "grad_norm": 0.3857824504375458, "learning_rate": 9.147539010268936e-05, "loss": 0.0532, "step": 22920 }, { "grad_norm": 0.4095267951488495, "learning_rate": 9.14661532866645e-05, "loss": 0.0507, "step": 22930 }, { "grad_norm": 0.3408958315849304, "learning_rate": 9.145691193596348e-05, "loss": 0.0519, "step": 22940 }, { "grad_norm": 0.3327902853488922, "learning_rate": 9.144766605159691e-05, "loss": 0.0515, "step": 22950 }, { "grad_norm": 0.42886245250701904, "learning_rate": 9.14384156345759e-05, "loss": 0.052, "step": 22960 }, { "grad_norm": 0.3725050389766693, "learning_rate": 9.142916068591204e-05, "loss": 0.0527, "step": 22970 }, { "grad_norm": 0.4936218559741974, "learning_rate": 9.141990120661746e-05, "loss": 0.0552, "step": 22980 }, { "grad_norm": 0.4110233783721924, "learning_rate": 9.141063719770475e-05, "loss": 0.0574, "step": 22990 }, { "grad_norm": 0.45265793800354004, "learning_rate": 9.140136866018704e-05, "loss": 0.056, "step": 23000 }, { "grad_norm": 0.38002902269363403, "learning_rate": 9.139209559507788e-05, "loss": 0.0539, "step": 23010 }, { "grad_norm": 0.42835351824760437, "learning_rate": 9.13828180033914e-05, "loss": 0.0571, "step": 23020 }, { "grad_norm": 0.37261101603507996, "learning_rate": 9.137353588614212e-05, "loss": 0.0614, "step": 23030 }, { "grad_norm": 0.3790356516838074, "learning_rate": 9.136424924434519e-05, "loss": 0.0606, "step": 23040 }, { "grad_norm": 0.46605899930000305, "learning_rate": 9.135495807901615e-05, "loss": 0.0534, "step": 23050 }, { "grad_norm": 0.3332359194755554, "learning_rate": 9.134566239117108e-05, "loss": 0.0569, "step": 23060 }, { "grad_norm": 0.42071953415870667, "learning_rate": 9.13363621818265e-05, "loss": 0.0512, "step": 23070 }, { "grad_norm": 0.5226825475692749, "learning_rate": 9.132705745199953e-05, "loss": 0.0569, "step": 23080 }, { "grad_norm": 0.3678807020187378, "learning_rate": 9.131774820270768e-05, "loss": 0.052, "step": 23090 }, { "grad_norm": 0.4943200945854187, "learning_rate": 9.130843443496901e-05, "loss": 0.0537, "step": 23100 }, { "grad_norm": 0.36696988344192505, "learning_rate": 9.129911614980206e-05, "loss": 0.0626, "step": 23110 }, { "grad_norm": 0.54903644323349, "learning_rate": 9.128979334822584e-05, "loss": 0.058, "step": 23120 }, { "grad_norm": 0.33289727568626404, "learning_rate": 9.128046603125992e-05, "loss": 0.062, "step": 23130 }, { "grad_norm": 0.475698322057724, "learning_rate": 9.12711341999243e-05, "loss": 0.0557, "step": 23140 }, { "grad_norm": 0.5499624609947205, "learning_rate": 9.12617978552395e-05, "loss": 0.0547, "step": 23150 }, { "grad_norm": 0.4872267544269562, "learning_rate": 9.12524569982265e-05, "loss": 0.045, "step": 23160 }, { "grad_norm": 0.44321122765541077, "learning_rate": 9.124311162990684e-05, "loss": 0.0559, "step": 23170 }, { "grad_norm": 0.704718291759491, "learning_rate": 9.12337617513025e-05, "loss": 0.0513, "step": 23180 }, { "grad_norm": 0.31839826703071594, "learning_rate": 9.122440736343596e-05, "loss": 0.0481, "step": 23190 }, { "grad_norm": 0.4130684435367584, "learning_rate": 9.12150484673302e-05, "loss": 0.0516, "step": 23200 }, { "grad_norm": 0.3433191478252411, "learning_rate": 9.120568506400873e-05, "loss": 0.0508, "step": 23210 }, { "grad_norm": 0.519745409488678, "learning_rate": 9.119631715449548e-05, "loss": 0.0583, "step": 23220 }, { "grad_norm": 0.3248799443244934, "learning_rate": 9.118694473981493e-05, "loss": 0.0576, "step": 23230 }, { "grad_norm": 0.3600580394268036, "learning_rate": 9.117756782099203e-05, "loss": 0.0517, "step": 23240 }, { "grad_norm": 0.3570035696029663, "learning_rate": 9.11681863990522e-05, "loss": 0.0477, "step": 23250 }, { "grad_norm": 0.42365604639053345, "learning_rate": 9.115880047502142e-05, "loss": 0.0497, "step": 23260 }, { "grad_norm": 0.35206547379493713, "learning_rate": 9.114941004992609e-05, "loss": 0.0497, "step": 23270 }, { "grad_norm": 0.6030888557434082, "learning_rate": 9.114001512479317e-05, "loss": 0.0564, "step": 23280 }, { "grad_norm": 0.4364825487136841, "learning_rate": 9.113061570065003e-05, "loss": 0.0495, "step": 23290 }, { "grad_norm": 0.5265839695930481, "learning_rate": 9.112121177852459e-05, "loss": 0.0508, "step": 23300 }, { "grad_norm": 0.4324362576007843, "learning_rate": 9.111180335944527e-05, "loss": 0.0525, "step": 23310 }, { "grad_norm": 0.40252000093460083, "learning_rate": 9.110239044444093e-05, "loss": 0.0477, "step": 23320 }, { "grad_norm": 0.5235879421234131, "learning_rate": 9.109297303454099e-05, "loss": 0.0513, "step": 23330 }, { "grad_norm": 0.2860035002231598, "learning_rate": 9.108355113077526e-05, "loss": 0.0544, "step": 23340 }, { "grad_norm": 0.5331764221191406, "learning_rate": 9.107412473417419e-05, "loss": 0.0464, "step": 23350 }, { "grad_norm": 0.4049032926559448, "learning_rate": 9.106469384576858e-05, "loss": 0.0508, "step": 23360 }, { "grad_norm": 0.44545304775238037, "learning_rate": 9.105525846658978e-05, "loss": 0.0622, "step": 23370 }, { "grad_norm": 0.4317170977592468, "learning_rate": 9.104581859766965e-05, "loss": 0.063, "step": 23380 }, { "grad_norm": 0.33210963010787964, "learning_rate": 9.10363742400405e-05, "loss": 0.0552, "step": 23390 }, { "grad_norm": 0.5147750377655029, "learning_rate": 9.102692539473518e-05, "loss": 0.0524, "step": 23400 }, { "grad_norm": 0.41572582721710205, "learning_rate": 9.101747206278697e-05, "loss": 0.0559, "step": 23410 }, { "grad_norm": 0.39981964230537415, "learning_rate": 9.100801424522968e-05, "loss": 0.0547, "step": 23420 }, { "grad_norm": 0.30624818801879883, "learning_rate": 9.099855194309762e-05, "loss": 0.0508, "step": 23430 }, { "grad_norm": 0.5248603820800781, "learning_rate": 9.098908515742554e-05, "loss": 0.0552, "step": 23440 }, { "grad_norm": 0.29057157039642334, "learning_rate": 9.097961388924873e-05, "loss": 0.0497, "step": 23450 }, { "grad_norm": 0.5242655873298645, "learning_rate": 9.097013813960298e-05, "loss": 0.0548, "step": 23460 }, { "grad_norm": 0.36088332533836365, "learning_rate": 9.09606579095245e-05, "loss": 0.0515, "step": 23470 }, { "grad_norm": 0.41467586159706116, "learning_rate": 9.095117320005008e-05, "loss": 0.0511, "step": 23480 }, { "grad_norm": 0.41067972779273987, "learning_rate": 9.094168401221691e-05, "loss": 0.0438, "step": 23490 }, { "grad_norm": 0.663703441619873, "learning_rate": 9.093219034706273e-05, "loss": 0.0496, "step": 23500 }, { "grad_norm": 0.3727395534515381, "learning_rate": 9.092269220562577e-05, "loss": 0.0518, "step": 23510 }, { "grad_norm": 0.5894927382469177, "learning_rate": 9.09131895889447e-05, "loss": 0.052, "step": 23520 }, { "grad_norm": 0.5833401083946228, "learning_rate": 9.090368249805873e-05, "loss": 0.0579, "step": 23530 }, { "grad_norm": 0.46212032437324524, "learning_rate": 9.089417093400754e-05, "loss": 0.0515, "step": 23540 }, { "grad_norm": 0.30299481749534607, "learning_rate": 9.088465489783131e-05, "loss": 0.0555, "step": 23550 }, { "grad_norm": 0.3670523166656494, "learning_rate": 9.087513439057068e-05, "loss": 0.0607, "step": 23560 }, { "grad_norm": 0.5567654371261597, "learning_rate": 9.08656094132668e-05, "loss": 0.0528, "step": 23570 }, { "grad_norm": 0.40852612257003784, "learning_rate": 9.085607996696134e-05, "loss": 0.0574, "step": 23580 }, { "grad_norm": 0.5525782108306885, "learning_rate": 9.084654605269639e-05, "loss": 0.0605, "step": 23590 }, { "grad_norm": 0.41047564148902893, "learning_rate": 9.083700767151457e-05, "loss": 0.0594, "step": 23600 }, { "grad_norm": 0.4511166214942932, "learning_rate": 9.082746482445898e-05, "loss": 0.053, "step": 23610 }, { "grad_norm": 0.5269184112548828, "learning_rate": 9.081791751257325e-05, "loss": 0.0477, "step": 23620 }, { "grad_norm": 0.48198407888412476, "learning_rate": 9.080836573690142e-05, "loss": 0.0542, "step": 23630 }, { "grad_norm": 0.3451294004917145, "learning_rate": 9.079880949848805e-05, "loss": 0.0497, "step": 23640 }, { "grad_norm": 0.37781524658203125, "learning_rate": 9.078924879837822e-05, "loss": 0.054, "step": 23650 }, { "grad_norm": 0.42940554022789, "learning_rate": 9.077968363761747e-05, "loss": 0.0583, "step": 23660 }, { "grad_norm": 0.3746063709259033, "learning_rate": 9.077011401725182e-05, "loss": 0.0535, "step": 23670 }, { "grad_norm": 0.4754994213581085, "learning_rate": 9.07605399383278e-05, "loss": 0.0594, "step": 23680 }, { "grad_norm": 0.4241623282432556, "learning_rate": 9.075096140189243e-05, "loss": 0.0559, "step": 23690 }, { "grad_norm": 0.41856569051742554, "learning_rate": 9.074137840899318e-05, "loss": 0.049, "step": 23700 }, { "grad_norm": 0.3581172525882721, "learning_rate": 9.073179096067804e-05, "loss": 0.0504, "step": 23710 }, { "grad_norm": 0.4662826955318451, "learning_rate": 9.072219905799549e-05, "loss": 0.0493, "step": 23720 }, { "grad_norm": 0.48766466975212097, "learning_rate": 9.071260270199447e-05, "loss": 0.0506, "step": 23730 }, { "grad_norm": 0.4122667908668518, "learning_rate": 9.070300189372441e-05, "loss": 0.0524, "step": 23740 }, { "grad_norm": 0.5610342025756836, "learning_rate": 9.069339663423528e-05, "loss": 0.0507, "step": 23750 }, { "grad_norm": 0.9328485131263733, "learning_rate": 9.068378692457747e-05, "loss": 0.0537, "step": 23760 }, { "grad_norm": 0.3620521128177643, "learning_rate": 9.067417276580189e-05, "loss": 0.0552, "step": 23770 }, { "grad_norm": 0.34978941082954407, "learning_rate": 9.066455415895993e-05, "loss": 0.0524, "step": 23780 }, { "grad_norm": 0.3788587152957916, "learning_rate": 9.065493110510346e-05, "loss": 0.0567, "step": 23790 }, { "grad_norm": 0.4647175371646881, "learning_rate": 9.064530360528484e-05, "loss": 0.0649, "step": 23800 }, { "grad_norm": 0.6336314678192139, "learning_rate": 9.063567166055695e-05, "loss": 0.0507, "step": 23810 }, { "grad_norm": 0.43561381101608276, "learning_rate": 9.062603527197308e-05, "loss": 0.0488, "step": 23820 }, { "grad_norm": 0.5032839179039001, "learning_rate": 9.06163944405871e-05, "loss": 0.0456, "step": 23830 }, { "grad_norm": 0.4291130304336548, "learning_rate": 9.060674916745327e-05, "loss": 0.0522, "step": 23840 }, { "grad_norm": 0.4244595468044281, "learning_rate": 9.05970994536264e-05, "loss": 0.0459, "step": 23850 }, { "grad_norm": 0.3312079906463623, "learning_rate": 9.05874453001618e-05, "loss": 0.0486, "step": 23860 }, { "grad_norm": 0.34055960178375244, "learning_rate": 9.057778670811517e-05, "loss": 0.0427, "step": 23870 }, { "grad_norm": 0.2873983085155487, "learning_rate": 9.056812367854281e-05, "loss": 0.0497, "step": 23880 }, { "grad_norm": 0.3551613688468933, "learning_rate": 9.055845621250143e-05, "loss": 0.0592, "step": 23890 }, { "grad_norm": 0.29451820254325867, "learning_rate": 9.054878431104825e-05, "loss": 0.0478, "step": 23900 }, { "grad_norm": 0.45773062109947205, "learning_rate": 9.0539107975241e-05, "loss": 0.0547, "step": 23910 }, { "grad_norm": 0.40474891662597656, "learning_rate": 9.052942720613784e-05, "loss": 0.0525, "step": 23920 }, { "grad_norm": 0.40415331721305847, "learning_rate": 9.051974200479745e-05, "loss": 0.0498, "step": 23930 }, { "grad_norm": 0.4267265498638153, "learning_rate": 9.051005237227901e-05, "loss": 0.052, "step": 23940 }, { "grad_norm": 0.5935120582580566, "learning_rate": 9.050035830964215e-05, "loss": 0.0526, "step": 23950 }, { "grad_norm": 0.6727884411811829, "learning_rate": 9.049065981794698e-05, "loss": 0.0554, "step": 23960 }, { "grad_norm": 0.5109757781028748, "learning_rate": 9.048095689825414e-05, "loss": 0.048, "step": 23970 }, { "grad_norm": 0.41825589537620544, "learning_rate": 9.047124955162472e-05, "loss": 0.0477, "step": 23980 }, { "grad_norm": 0.43128228187561035, "learning_rate": 9.046153777912028e-05, "loss": 0.0536, "step": 23990 }, { "grad_norm": 0.39227554202079773, "learning_rate": 9.045182158180292e-05, "loss": 0.0562, "step": 24000 }, { "grad_norm": 0.4519027769565582, "learning_rate": 9.044210096073516e-05, "loss": 0.0562, "step": 24010 }, { "grad_norm": 0.43331098556518555, "learning_rate": 9.043237591698004e-05, "loss": 0.0634, "step": 24020 }, { "grad_norm": 0.31860998272895813, "learning_rate": 9.04226464516011e-05, "loss": 0.0527, "step": 24030 }, { "grad_norm": 0.45475301146507263, "learning_rate": 9.041291256566229e-05, "loss": 0.0563, "step": 24040 }, { "grad_norm": 0.3625566065311432, "learning_rate": 9.040317426022814e-05, "loss": 0.048, "step": 24050 }, { "grad_norm": 0.5180587768554688, "learning_rate": 9.03934315363636e-05, "loss": 0.054, "step": 24060 }, { "grad_norm": 0.36215659976005554, "learning_rate": 9.038368439513409e-05, "loss": 0.0544, "step": 24070 }, { "grad_norm": 0.48591139912605286, "learning_rate": 9.03739328376056e-05, "loss": 0.0541, "step": 24080 }, { "grad_norm": 0.3237672448158264, "learning_rate": 9.036417686484451e-05, "loss": 0.0529, "step": 24090 }, { "grad_norm": 0.32669946551322937, "learning_rate": 9.035441647791773e-05, "loss": 0.0498, "step": 24100 }, { "grad_norm": 0.43306398391723633, "learning_rate": 9.034465167789263e-05, "loss": 0.0546, "step": 24110 }, { "grad_norm": 0.4012815058231354, "learning_rate": 9.033488246583706e-05, "loss": 0.0514, "step": 24120 }, { "grad_norm": 0.4457390606403351, "learning_rate": 9.032510884281941e-05, "loss": 0.0479, "step": 24130 }, { "grad_norm": 0.3685852289199829, "learning_rate": 9.031533080990848e-05, "loss": 0.0487, "step": 24140 }, { "grad_norm": 0.2571549415588379, "learning_rate": 9.030554836817358e-05, "loss": 0.0489, "step": 24150 }, { "grad_norm": 0.31355157494544983, "learning_rate": 9.029576151868451e-05, "loss": 0.0508, "step": 24160 }, { "grad_norm": 0.3853192925453186, "learning_rate": 9.028597026251155e-05, "loss": 0.042, "step": 24170 }, { "grad_norm": 0.545741081237793, "learning_rate": 9.027617460072547e-05, "loss": 0.0553, "step": 24180 }, { "grad_norm": 0.48757076263427734, "learning_rate": 9.026637453439745e-05, "loss": 0.057, "step": 24190 }, { "grad_norm": 0.36193549633026123, "learning_rate": 9.025657006459927e-05, "loss": 0.0425, "step": 24200 }, { "grad_norm": 0.39471399784088135, "learning_rate": 9.024676119240311e-05, "loss": 0.0597, "step": 24210 }, { "grad_norm": 0.45511457324028015, "learning_rate": 9.023694791888166e-05, "loss": 0.052, "step": 24220 }, { "grad_norm": 0.6266444325447083, "learning_rate": 9.022713024510808e-05, "loss": 0.058, "step": 24230 }, { "grad_norm": 0.516043484210968, "learning_rate": 9.021730817215601e-05, "loss": 0.0522, "step": 24240 }, { "grad_norm": 0.3964608907699585, "learning_rate": 9.02074817010996e-05, "loss": 0.0506, "step": 24250 }, { "grad_norm": 0.4371621608734131, "learning_rate": 9.019765083301342e-05, "loss": 0.0406, "step": 24260 }, { "grad_norm": 0.4060054123401642, "learning_rate": 9.01878155689726e-05, "loss": 0.045, "step": 24270 }, { "grad_norm": 0.40656232833862305, "learning_rate": 9.017797591005268e-05, "loss": 0.0464, "step": 24280 }, { "grad_norm": 0.4176592230796814, "learning_rate": 9.016813185732972e-05, "loss": 0.0479, "step": 24290 }, { "grad_norm": 0.39765268564224243, "learning_rate": 9.015828341188027e-05, "loss": 0.051, "step": 24300 }, { "grad_norm": 0.3930553197860718, "learning_rate": 9.01484305747813e-05, "loss": 0.0517, "step": 24310 }, { "grad_norm": 0.3961636424064636, "learning_rate": 9.013857334711033e-05, "loss": 0.0486, "step": 24320 }, { "grad_norm": 0.48097845911979675, "learning_rate": 9.012871172994534e-05, "loss": 0.0505, "step": 24330 }, { "grad_norm": 0.42979171872138977, "learning_rate": 9.011884572436476e-05, "loss": 0.0552, "step": 24340 }, { "grad_norm": 0.6710344552993774, "learning_rate": 9.010897533144754e-05, "loss": 0.052, "step": 24350 }, { "grad_norm": 0.3171534836292267, "learning_rate": 9.009910055227306e-05, "loss": 0.0528, "step": 24360 }, { "grad_norm": 0.32469066977500916, "learning_rate": 9.008922138792124e-05, "loss": 0.0523, "step": 24370 }, { "grad_norm": 0.36438000202178955, "learning_rate": 9.007933783947244e-05, "loss": 0.0536, "step": 24380 }, { "grad_norm": 0.4117169678211212, "learning_rate": 9.006944990800752e-05, "loss": 0.0545, "step": 24390 }, { "grad_norm": 0.43645671010017395, "learning_rate": 9.005955759460779e-05, "loss": 0.0604, "step": 24400 }, { "grad_norm": 0.4845719635486603, "learning_rate": 9.004966090035508e-05, "loss": 0.0509, "step": 24410 }, { "grad_norm": 0.31137615442276, "learning_rate": 9.003975982633166e-05, "loss": 0.052, "step": 24420 }, { "grad_norm": 0.34810715913772583, "learning_rate": 9.00298543736203e-05, "loss": 0.0452, "step": 24430 }, { "grad_norm": 0.38779425621032715, "learning_rate": 9.001994454330427e-05, "loss": 0.0515, "step": 24440 }, { "grad_norm": 0.28535816073417664, "learning_rate": 9.001003033646727e-05, "loss": 0.0507, "step": 24450 }, { "grad_norm": 0.34430205821990967, "learning_rate": 9.00001117541935e-05, "loss": 0.0467, "step": 24460 }, { "grad_norm": 0.5288655757904053, "learning_rate": 8.999018879756764e-05, "loss": 0.049, "step": 24470 }, { "grad_norm": 0.47225821018218994, "learning_rate": 8.998026146767487e-05, "loss": 0.0535, "step": 24480 }, { "grad_norm": 0.4184420108795166, "learning_rate": 8.99703297656008e-05, "loss": 0.0556, "step": 24490 }, { "grad_norm": 0.3720954954624176, "learning_rate": 8.996039369243156e-05, "loss": 0.0536, "step": 24500 }, { "grad_norm": 0.4961066246032715, "learning_rate": 8.995045324925378e-05, "loss": 0.053, "step": 24510 }, { "grad_norm": 0.3839170038700104, "learning_rate": 8.994050843715448e-05, "loss": 0.0535, "step": 24520 }, { "grad_norm": 0.4713021218776703, "learning_rate": 8.993055925722121e-05, "loss": 0.0613, "step": 24530 }, { "grad_norm": 0.3357751667499542, "learning_rate": 8.992060571054202e-05, "loss": 0.0529, "step": 24540 }, { "grad_norm": 0.4491337835788727, "learning_rate": 8.991064779820542e-05, "loss": 0.05, "step": 24550 }, { "grad_norm": 0.38799282908439636, "learning_rate": 8.990068552130036e-05, "loss": 0.0506, "step": 24560 }, { "grad_norm": 0.3411529064178467, "learning_rate": 8.989071888091634e-05, "loss": 0.055, "step": 24570 }, { "grad_norm": 0.451344758272171, "learning_rate": 8.988074787814329e-05, "loss": 0.0497, "step": 24580 }, { "grad_norm": 0.4288961589336395, "learning_rate": 8.987077251407158e-05, "loss": 0.0538, "step": 24590 }, { "grad_norm": 0.4036814570426941, "learning_rate": 8.986079278979216e-05, "loss": 0.0488, "step": 24600 }, { "grad_norm": 0.49563589692115784, "learning_rate": 8.985080870639635e-05, "loss": 0.0533, "step": 24610 }, { "grad_norm": 0.44718143343925476, "learning_rate": 8.984082026497603e-05, "loss": 0.0501, "step": 24620 }, { "grad_norm": 0.4481351375579834, "learning_rate": 8.98308274666235e-05, "loss": 0.0477, "step": 24630 }, { "grad_norm": 0.47707509994506836, "learning_rate": 8.982083031243155e-05, "loss": 0.0469, "step": 24640 }, { "grad_norm": 0.5070051550865173, "learning_rate": 8.98108288034935e-05, "loss": 0.0513, "step": 24650 }, { "grad_norm": 0.43627259135246277, "learning_rate": 8.980082294090305e-05, "loss": 0.0426, "step": 24660 }, { "grad_norm": 0.3367418050765991, "learning_rate": 8.979081272575443e-05, "loss": 0.0543, "step": 24670 }, { "grad_norm": 0.5041260123252869, "learning_rate": 8.978079815914236e-05, "loss": 0.0608, "step": 24680 }, { "grad_norm": 0.3083857297897339, "learning_rate": 8.977077924216202e-05, "loss": 0.0536, "step": 24690 }, { "grad_norm": 0.42035284638404846, "learning_rate": 8.976075597590905e-05, "loss": 0.0548, "step": 24700 }, { "grad_norm": 0.3247222602367401, "learning_rate": 8.975072836147958e-05, "loss": 0.0523, "step": 24710 }, { "grad_norm": 0.37946587800979614, "learning_rate": 8.974069639997025e-05, "loss": 0.0583, "step": 24720 }, { "grad_norm": 0.4052644968032837, "learning_rate": 8.973066009247808e-05, "loss": 0.0513, "step": 24730 }, { "grad_norm": 0.40021568536758423, "learning_rate": 8.972061944010066e-05, "loss": 0.0541, "step": 24740 }, { "grad_norm": 0.34197598695755005, "learning_rate": 8.971057444393603e-05, "loss": 0.0569, "step": 24750 }, { "grad_norm": 0.33789196610450745, "learning_rate": 8.970052510508268e-05, "loss": 0.0508, "step": 24760 }, { "grad_norm": 0.45436790585517883, "learning_rate": 8.969047142463959e-05, "loss": 0.0482, "step": 24770 }, { "grad_norm": 0.39274850487709045, "learning_rate": 8.968041340370621e-05, "loss": 0.0555, "step": 24780 }, { "grad_norm": 0.5131451487541199, "learning_rate": 8.96703510433825e-05, "loss": 0.057, "step": 24790 }, { "grad_norm": 0.3568720519542694, "learning_rate": 8.966028434476883e-05, "loss": 0.0524, "step": 24800 }, { "grad_norm": 0.3825308382511139, "learning_rate": 8.96502133089661e-05, "loss": 0.0516, "step": 24810 }, { "grad_norm": 0.4327036440372467, "learning_rate": 8.964013793707564e-05, "loss": 0.0487, "step": 24820 }, { "grad_norm": 0.3674275875091553, "learning_rate": 8.963005823019932e-05, "loss": 0.0458, "step": 24830 }, { "grad_norm": 0.4152040481567383, "learning_rate": 8.961997418943939e-05, "loss": 0.0499, "step": 24840 }, { "grad_norm": 0.4180203676223755, "learning_rate": 8.960988581589865e-05, "loss": 0.0505, "step": 24850 }, { "grad_norm": 0.7547603249549866, "learning_rate": 8.959979311068037e-05, "loss": 0.0505, "step": 24860 }, { "grad_norm": 0.4529622793197632, "learning_rate": 8.958969607488823e-05, "loss": 0.0502, "step": 24870 }, { "grad_norm": 0.6268004775047302, "learning_rate": 8.957959470962647e-05, "loss": 0.0532, "step": 24880 }, { "grad_norm": 0.44117608666419983, "learning_rate": 8.956948901599971e-05, "loss": 0.0536, "step": 24890 }, { "grad_norm": 0.33712098002433777, "learning_rate": 8.955937899511315e-05, "loss": 0.0621, "step": 24900 }, { "grad_norm": 0.27806556224823, "learning_rate": 8.954926464807238e-05, "loss": 0.0548, "step": 24910 }, { "grad_norm": 0.8361673951148987, "learning_rate": 8.953914597598347e-05, "loss": 0.0553, "step": 24920 }, { "grad_norm": 0.34143877029418945, "learning_rate": 8.952902297995303e-05, "loss": 0.0551, "step": 24930 }, { "grad_norm": 0.555549144744873, "learning_rate": 8.951889566108804e-05, "loss": 0.0637, "step": 24940 }, { "grad_norm": 0.4272335171699524, "learning_rate": 8.950876402049606e-05, "loss": 0.0539, "step": 24950 }, { "grad_norm": 0.41065919399261475, "learning_rate": 8.949862805928504e-05, "loss": 0.0464, "step": 24960 }, { "grad_norm": 0.32477402687072754, "learning_rate": 8.948848777856343e-05, "loss": 0.0515, "step": 24970 }, { "grad_norm": 0.45240750908851624, "learning_rate": 8.947834317944017e-05, "loss": 0.051, "step": 24980 }, { "grad_norm": 0.352425754070282, "learning_rate": 8.946819426302466e-05, "loss": 0.0492, "step": 24990 }, { "grad_norm": 0.4703863859176636, "learning_rate": 8.945804103042676e-05, "loss": 0.0557, "step": 25000 }, { "grad_norm": 0.42959773540496826, "learning_rate": 8.944788348275681e-05, "loss": 0.0458, "step": 25010 }, { "grad_norm": 0.4560847580432892, "learning_rate": 8.943772162112565e-05, "loss": 0.0506, "step": 25020 }, { "grad_norm": 0.4084087312221527, "learning_rate": 8.942755544664454e-05, "loss": 0.0509, "step": 25030 }, { "grad_norm": 0.4970782399177551, "learning_rate": 8.941738496042525e-05, "loss": 0.0516, "step": 25040 }, { "grad_norm": 0.45605695247650146, "learning_rate": 8.940721016357999e-05, "loss": 0.0496, "step": 25050 }, { "grad_norm": 0.2623547613620758, "learning_rate": 8.939703105722148e-05, "loss": 0.051, "step": 25060 }, { "grad_norm": 0.46994829177856445, "learning_rate": 8.93868476424629e-05, "loss": 0.0544, "step": 25070 }, { "grad_norm": 0.3429519236087799, "learning_rate": 8.937665992041786e-05, "loss": 0.0486, "step": 25080 }, { "grad_norm": 0.39670947194099426, "learning_rate": 8.93664678922005e-05, "loss": 0.0512, "step": 25090 }, { "grad_norm": 0.6175389885902405, "learning_rate": 8.93562715589254e-05, "loss": 0.0514, "step": 25100 }, { "grad_norm": 0.7860468029975891, "learning_rate": 8.934607092170762e-05, "loss": 0.0519, "step": 25110 }, { "grad_norm": 0.3537036180496216, "learning_rate": 8.933586598166266e-05, "loss": 0.052, "step": 25120 }, { "grad_norm": 0.3054445683956146, "learning_rate": 8.932565673990655e-05, "loss": 0.0507, "step": 25130 }, { "grad_norm": 0.4736350476741791, "learning_rate": 8.931544319755574e-05, "loss": 0.0516, "step": 25140 }, { "grad_norm": 0.45270177721977234, "learning_rate": 8.930522535572718e-05, "loss": 0.0548, "step": 25150 }, { "grad_norm": 0.4498427212238312, "learning_rate": 8.929500321553826e-05, "loss": 0.052, "step": 25160 }, { "grad_norm": 0.2957780957221985, "learning_rate": 8.928477677810686e-05, "loss": 0.0456, "step": 25170 }, { "grad_norm": 0.8166987895965576, "learning_rate": 8.927454604455137e-05, "loss": 0.0642, "step": 25180 }, { "grad_norm": 0.4307924211025238, "learning_rate": 8.926431101599053e-05, "loss": 0.0524, "step": 25190 }, { "grad_norm": 0.3172207474708557, "learning_rate": 8.925407169354369e-05, "loss": 0.0563, "step": 25200 }, { "grad_norm": 0.315544068813324, "learning_rate": 8.92438280783306e-05, "loss": 0.0429, "step": 25210 }, { "grad_norm": 0.4627106487751007, "learning_rate": 8.923358017147146e-05, "loss": 0.0506, "step": 25220 }, { "grad_norm": 0.34288519620895386, "learning_rate": 8.922332797408697e-05, "loss": 0.0565, "step": 25230 }, { "grad_norm": 0.45221686363220215, "learning_rate": 8.921307148729831e-05, "loss": 0.0523, "step": 25240 }, { "grad_norm": 0.44103291630744934, "learning_rate": 8.920281071222712e-05, "loss": 0.0503, "step": 25250 }, { "grad_norm": 0.39415818452835083, "learning_rate": 8.919254564999548e-05, "loss": 0.0467, "step": 25260 }, { "grad_norm": 0.3809119760990143, "learning_rate": 8.918227630172598e-05, "loss": 0.0484, "step": 25270 }, { "grad_norm": 0.4614782929420471, "learning_rate": 8.917200266854165e-05, "loss": 0.0544, "step": 25280 }, { "grad_norm": 0.4652184247970581, "learning_rate": 8.9161724751566e-05, "loss": 0.057, "step": 25290 }, { "grad_norm": 0.3832756578922272, "learning_rate": 8.915144255192302e-05, "loss": 0.0491, "step": 25300 }, { "grad_norm": 0.36188390851020813, "learning_rate": 8.914115607073714e-05, "loss": 0.0456, "step": 25310 }, { "grad_norm": 0.5374749302864075, "learning_rate": 8.913086530913327e-05, "loss": 0.0535, "step": 25320 }, { "grad_norm": 0.4888797700405121, "learning_rate": 8.912057026823681e-05, "loss": 0.0495, "step": 25330 }, { "grad_norm": 0.4064587354660034, "learning_rate": 8.91102709491736e-05, "loss": 0.0577, "step": 25340 }, { "grad_norm": 0.3856950104236603, "learning_rate": 8.909996735306996e-05, "loss": 0.0498, "step": 25350 }, { "grad_norm": 0.45798587799072266, "learning_rate": 8.908965948105268e-05, "loss": 0.051, "step": 25360 }, { "grad_norm": 0.35292360186576843, "learning_rate": 8.907934733424901e-05, "loss": 0.0523, "step": 25370 }, { "grad_norm": 0.42093995213508606, "learning_rate": 8.906903091378666e-05, "loss": 0.0481, "step": 25380 }, { "grad_norm": 0.5503621101379395, "learning_rate": 8.905871022079384e-05, "loss": 0.046, "step": 25390 }, { "grad_norm": 0.38666415214538574, "learning_rate": 8.90483852563992e-05, "loss": 0.0399, "step": 25400 }, { "grad_norm": 0.2865085005760193, "learning_rate": 8.903805602173185e-05, "loss": 0.0413, "step": 25410 }, { "grad_norm": 0.2798668444156647, "learning_rate": 8.902772251792137e-05, "loss": 0.0459, "step": 25420 }, { "grad_norm": 0.5117411613464355, "learning_rate": 8.901738474609786e-05, "loss": 0.049, "step": 25430 }, { "grad_norm": 0.5940896272659302, "learning_rate": 8.900704270739179e-05, "loss": 0.05, "step": 25440 }, { "grad_norm": 0.40245702862739563, "learning_rate": 8.89966964029342e-05, "loss": 0.0523, "step": 25450 }, { "grad_norm": 0.4402085840702057, "learning_rate": 8.898634583385652e-05, "loss": 0.0504, "step": 25460 }, { "grad_norm": 0.46873265504837036, "learning_rate": 8.897599100129065e-05, "loss": 0.0547, "step": 25470 }, { "grad_norm": 0.4229775369167328, "learning_rate": 8.896563190636903e-05, "loss": 0.0489, "step": 25480 }, { "grad_norm": 0.37998244166374207, "learning_rate": 8.895526855022448e-05, "loss": 0.0504, "step": 25490 }, { "grad_norm": 0.3124900460243225, "learning_rate": 8.894490093399033e-05, "loss": 0.0486, "step": 25500 }, { "grad_norm": 0.43314307928085327, "learning_rate": 8.893452905880035e-05, "loss": 0.0571, "step": 25510 }, { "grad_norm": 0.39520522952079773, "learning_rate": 8.892415292578883e-05, "loss": 0.0462, "step": 25520 }, { "grad_norm": 0.44495823979377747, "learning_rate": 8.891377253609046e-05, "loss": 0.0521, "step": 25530 }, { "grad_norm": 0.5717918276786804, "learning_rate": 8.890338789084043e-05, "loss": 0.05, "step": 25540 }, { "grad_norm": 0.35490062832832336, "learning_rate": 8.88929989911744e-05, "loss": 0.052, "step": 25550 }, { "grad_norm": 0.3986889719963074, "learning_rate": 8.888260583822847e-05, "loss": 0.0601, "step": 25560 }, { "grad_norm": 0.42385613918304443, "learning_rate": 8.887220843313921e-05, "loss": 0.0491, "step": 25570 }, { "grad_norm": 0.40087592601776123, "learning_rate": 8.88618067770437e-05, "loss": 0.0457, "step": 25580 }, { "grad_norm": 0.3526656925678253, "learning_rate": 8.885140087107942e-05, "loss": 0.0472, "step": 25590 }, { "grad_norm": 0.5019330978393555, "learning_rate": 8.884099071638436e-05, "loss": 0.0525, "step": 25600 }, { "grad_norm": 1.4811517000198364, "learning_rate": 8.883057631409695e-05, "loss": 0.0472, "step": 25610 }, { "grad_norm": 0.33916065096855164, "learning_rate": 8.882015766535608e-05, "loss": 0.0434, "step": 25620 }, { "grad_norm": 0.4535596966743469, "learning_rate": 8.880973477130115e-05, "loss": 0.0501, "step": 25630 }, { "grad_norm": 0.37506935000419617, "learning_rate": 8.879930763307197e-05, "loss": 0.0585, "step": 25640 }, { "grad_norm": 0.3870231807231903, "learning_rate": 8.878887625180884e-05, "loss": 0.0483, "step": 25650 }, { "grad_norm": 0.62034672498703, "learning_rate": 8.877844062865253e-05, "loss": 0.0518, "step": 25660 }, { "grad_norm": 0.437406063079834, "learning_rate": 8.876800076474424e-05, "loss": 0.0569, "step": 25670 }, { "grad_norm": 0.4909292757511139, "learning_rate": 8.875755666122568e-05, "loss": 0.0552, "step": 25680 }, { "grad_norm": 0.3338167369365692, "learning_rate": 8.8747108319239e-05, "loss": 0.0509, "step": 25690 }, { "grad_norm": 0.4838374853134155, "learning_rate": 8.87366557399268e-05, "loss": 0.0576, "step": 25700 }, { "grad_norm": 0.5966087579727173, "learning_rate": 8.872619892443217e-05, "loss": 0.0527, "step": 25710 }, { "grad_norm": 0.9637616872787476, "learning_rate": 8.871573787389865e-05, "loss": 0.0561, "step": 25720 }, { "grad_norm": 0.4313136339187622, "learning_rate": 8.870527258947024e-05, "loss": 0.0463, "step": 25730 }, { "grad_norm": 0.3365718722343445, "learning_rate": 8.869480307229143e-05, "loss": 0.0507, "step": 25740 }, { "grad_norm": 0.4111678898334503, "learning_rate": 8.868432932350712e-05, "loss": 0.0523, "step": 25750 }, { "grad_norm": 0.4872879385948181, "learning_rate": 8.867385134426272e-05, "loss": 0.0438, "step": 25760 }, { "grad_norm": 0.42557093501091003, "learning_rate": 8.866336913570407e-05, "loss": 0.0553, "step": 25770 }, { "grad_norm": 0.4120928943157196, "learning_rate": 8.865288269897751e-05, "loss": 0.0519, "step": 25780 }, { "grad_norm": 0.4752865135669708, "learning_rate": 8.864239203522981e-05, "loss": 0.0503, "step": 25790 }, { "grad_norm": 0.354671835899353, "learning_rate": 8.863189714560822e-05, "loss": 0.0563, "step": 25800 }, { "grad_norm": 0.9139666557312012, "learning_rate": 8.862139803126043e-05, "loss": 0.0502, "step": 25810 }, { "grad_norm": 0.4329809248447418, "learning_rate": 8.861089469333463e-05, "loss": 0.0435, "step": 25820 }, { "grad_norm": 0.43380001187324524, "learning_rate": 8.860038713297944e-05, "loss": 0.0454, "step": 25830 }, { "grad_norm": 0.35451388359069824, "learning_rate": 8.858987535134394e-05, "loss": 0.051, "step": 25840 }, { "grad_norm": 0.405202716588974, "learning_rate": 8.857935934957769e-05, "loss": 0.0514, "step": 25850 }, { "grad_norm": 0.644619345664978, "learning_rate": 8.856883912883071e-05, "loss": 0.0553, "step": 25860 }, { "grad_norm": 0.4185960590839386, "learning_rate": 8.855831469025346e-05, "loss": 0.0461, "step": 25870 }, { "grad_norm": 0.5083804130554199, "learning_rate": 8.854778603499689e-05, "loss": 0.0553, "step": 25880 }, { "grad_norm": 0.46028921008110046, "learning_rate": 8.85372531642124e-05, "loss": 0.0506, "step": 25890 }, { "grad_norm": 0.5007681846618652, "learning_rate": 8.852671607905185e-05, "loss": 0.0531, "step": 25900 }, { "grad_norm": 0.38613900542259216, "learning_rate": 8.851617478066754e-05, "loss": 0.0517, "step": 25910 }, { "grad_norm": 0.6731433272361755, "learning_rate": 8.850562927021227e-05, "loss": 0.055, "step": 25920 }, { "grad_norm": 0.423824667930603, "learning_rate": 8.849507954883928e-05, "loss": 0.0544, "step": 25930 }, { "grad_norm": 0.3747345209121704, "learning_rate": 8.848452561770226e-05, "loss": 0.0448, "step": 25940 }, { "grad_norm": 0.31492796540260315, "learning_rate": 8.847396747795538e-05, "loss": 0.0506, "step": 25950 }, { "grad_norm": 0.34799572825431824, "learning_rate": 8.846340513075327e-05, "loss": 0.0515, "step": 25960 }, { "grad_norm": 0.4314271807670593, "learning_rate": 8.845283857725099e-05, "loss": 0.054, "step": 25970 }, { "grad_norm": 0.38699886202812195, "learning_rate": 8.844226781860409e-05, "loss": 0.058, "step": 25980 }, { "grad_norm": 0.3021339178085327, "learning_rate": 8.84316928559686e-05, "loss": 0.0444, "step": 25990 }, { "grad_norm": 0.39517590403556824, "learning_rate": 8.842111369050094e-05, "loss": 0.053, "step": 26000 }, { "grad_norm": 0.36742475628852844, "learning_rate": 8.841053032335808e-05, "loss": 0.0535, "step": 26010 }, { "grad_norm": 0.44105058908462524, "learning_rate": 8.839994275569735e-05, "loss": 0.0592, "step": 26020 }, { "grad_norm": 0.4249357283115387, "learning_rate": 8.838935098867662e-05, "loss": 0.0442, "step": 26030 }, { "grad_norm": 0.5488510131835938, "learning_rate": 8.837875502345418e-05, "loss": 0.0554, "step": 26040 }, { "grad_norm": 0.32535549998283386, "learning_rate": 8.83681548611888e-05, "loss": 0.0521, "step": 26050 }, { "grad_norm": 0.3492363691329956, "learning_rate": 8.835755050303969e-05, "loss": 0.0519, "step": 26060 }, { "grad_norm": 0.5620190501213074, "learning_rate": 8.834694195016653e-05, "loss": 0.0604, "step": 26070 }, { "grad_norm": 0.5712189674377441, "learning_rate": 8.833632920372942e-05, "loss": 0.0578, "step": 26080 }, { "grad_norm": 0.4555111825466156, "learning_rate": 8.832571226488903e-05, "loss": 0.0524, "step": 26090 }, { "grad_norm": 0.3514549732208252, "learning_rate": 8.831509113480634e-05, "loss": 0.0505, "step": 26100 }, { "grad_norm": 0.385150671005249, "learning_rate": 8.83044658146429e-05, "loss": 0.0516, "step": 26110 }, { "grad_norm": 0.48313820362091064, "learning_rate": 8.829383630556067e-05, "loss": 0.0444, "step": 26120 }, { "grad_norm": 0.4566771984100342, "learning_rate": 8.828320260872207e-05, "loss": 0.0442, "step": 26130 }, { "grad_norm": 0.34011998772621155, "learning_rate": 8.827256472529e-05, "loss": 0.0469, "step": 26140 }, { "grad_norm": 0.37407854199409485, "learning_rate": 8.826192265642778e-05, "loss": 0.0483, "step": 26150 }, { "grad_norm": 0.39073824882507324, "learning_rate": 8.825127640329923e-05, "loss": 0.0459, "step": 26160 }, { "grad_norm": 0.3396824300289154, "learning_rate": 8.824062596706861e-05, "loss": 0.0498, "step": 26170 }, { "grad_norm": 0.39889365434646606, "learning_rate": 8.822997134890062e-05, "loss": 0.0538, "step": 26180 }, { "grad_norm": 0.3559979498386383, "learning_rate": 8.821931254996044e-05, "loss": 0.0494, "step": 26190 }, { "grad_norm": 0.4714411199092865, "learning_rate": 8.82086495714137e-05, "loss": 0.048, "step": 26200 }, { "grad_norm": 0.5482955574989319, "learning_rate": 8.81979824144265e-05, "loss": 0.055, "step": 26210 }, { "grad_norm": 0.4227832555770874, "learning_rate": 8.818731108016536e-05, "loss": 0.0554, "step": 26220 }, { "grad_norm": 0.45778757333755493, "learning_rate": 8.81766355697973e-05, "loss": 0.0531, "step": 26230 }, { "grad_norm": 0.5407856702804565, "learning_rate": 8.816595588448977e-05, "loss": 0.0514, "step": 26240 }, { "grad_norm": 0.40700721740722656, "learning_rate": 8.81552720254107e-05, "loss": 0.0553, "step": 26250 }, { "grad_norm": 0.4976534843444824, "learning_rate": 8.814458399372842e-05, "loss": 0.0557, "step": 26260 }, { "grad_norm": 0.4216909110546112, "learning_rate": 8.813389179061181e-05, "loss": 0.0591, "step": 26270 }, { "grad_norm": 0.3325718641281128, "learning_rate": 8.812319541723012e-05, "loss": 0.0561, "step": 26280 }, { "grad_norm": 0.4887401759624481, "learning_rate": 8.811249487475309e-05, "loss": 0.0509, "step": 26290 }, { "grad_norm": 0.44485941529273987, "learning_rate": 8.810179016435092e-05, "loss": 0.0521, "step": 26300 }, { "grad_norm": 0.3272031545639038, "learning_rate": 8.809108128719428e-05, "loss": 0.0541, "step": 26310 }, { "grad_norm": 0.4817565381526947, "learning_rate": 8.808036824445424e-05, "loss": 0.057, "step": 26320 }, { "grad_norm": 0.6070782542228699, "learning_rate": 8.806965103730238e-05, "loss": 0.0547, "step": 26330 }, { "grad_norm": 0.4064006209373474, "learning_rate": 8.805892966691074e-05, "loss": 0.0495, "step": 26340 }, { "grad_norm": 0.4063737690448761, "learning_rate": 8.804820413445175e-05, "loss": 0.0467, "step": 26350 }, { "grad_norm": 0.36844825744628906, "learning_rate": 8.803747444109837e-05, "loss": 0.0445, "step": 26360 }, { "grad_norm": 0.5522575974464417, "learning_rate": 8.802674058802399e-05, "loss": 0.053, "step": 26370 }, { "grad_norm": 0.36794906854629517, "learning_rate": 8.801600257640241e-05, "loss": 0.0371, "step": 26380 }, { "grad_norm": 0.3571677505970001, "learning_rate": 8.800526040740795e-05, "loss": 0.0491, "step": 26390 }, { "grad_norm": 0.34967485070228577, "learning_rate": 8.799451408221535e-05, "loss": 0.0448, "step": 26400 }, { "grad_norm": 0.3769148290157318, "learning_rate": 8.798376360199982e-05, "loss": 0.0554, "step": 26410 }, { "grad_norm": 0.3818672001361847, "learning_rate": 8.797300896793701e-05, "loss": 0.0445, "step": 26420 }, { "grad_norm": 0.4589477777481079, "learning_rate": 8.796225018120302e-05, "loss": 0.0562, "step": 26430 }, { "grad_norm": 0.2667498290538788, "learning_rate": 8.795148724297444e-05, "loss": 0.0433, "step": 26440 }, { "grad_norm": 0.3846791386604309, "learning_rate": 8.794072015442825e-05, "loss": 0.0562, "step": 26450 }, { "grad_norm": 0.40559524297714233, "learning_rate": 8.792994891674198e-05, "loss": 0.0529, "step": 26460 }, { "grad_norm": 0.5425451993942261, "learning_rate": 8.79191735310935e-05, "loss": 0.0527, "step": 26470 }, { "grad_norm": 0.4474685788154602, "learning_rate": 8.790839399866122e-05, "loss": 0.0532, "step": 26480 }, { "grad_norm": 0.3880092203617096, "learning_rate": 8.789761032062397e-05, "loss": 0.0499, "step": 26490 }, { "grad_norm": 0.6138405203819275, "learning_rate": 8.788682249816103e-05, "loss": 0.0523, "step": 26500 }, { "grad_norm": 0.351340115070343, "learning_rate": 8.787603053245215e-05, "loss": 0.0463, "step": 26510 }, { "grad_norm": 0.37460073828697205, "learning_rate": 8.78652344246775e-05, "loss": 0.0511, "step": 26520 }, { "grad_norm": 0.2945215106010437, "learning_rate": 8.785443417601776e-05, "loss": 0.0497, "step": 26530 }, { "grad_norm": 0.3421037793159485, "learning_rate": 8.784362978765401e-05, "loss": 0.0454, "step": 26540 }, { "grad_norm": 0.3907395899295807, "learning_rate": 8.783282126076779e-05, "loss": 0.0503, "step": 26550 }, { "grad_norm": 0.39459922909736633, "learning_rate": 8.782200859654112e-05, "loss": 0.0427, "step": 26560 }, { "grad_norm": 0.39701396226882935, "learning_rate": 8.781119179615646e-05, "loss": 0.0414, "step": 26570 }, { "grad_norm": 0.34914979338645935, "learning_rate": 8.780037086079674e-05, "loss": 0.04, "step": 26580 }, { "grad_norm": 0.34596309065818787, "learning_rate": 8.778954579164527e-05, "loss": 0.0474, "step": 26590 }, { "grad_norm": 0.34584444761276245, "learning_rate": 8.777871658988588e-05, "loss": 0.048, "step": 26600 }, { "grad_norm": 0.3638916313648224, "learning_rate": 8.776788325670285e-05, "loss": 0.0493, "step": 26610 }, { "grad_norm": 0.3908783793449402, "learning_rate": 8.775704579328089e-05, "loss": 0.049, "step": 26620 }, { "grad_norm": 0.3822311460971832, "learning_rate": 8.774620420080517e-05, "loss": 0.0533, "step": 26630 }, { "grad_norm": 0.501423716545105, "learning_rate": 8.773535848046131e-05, "loss": 0.048, "step": 26640 }, { "grad_norm": 0.395134836435318, "learning_rate": 8.772450863343538e-05, "loss": 0.0493, "step": 26650 }, { "grad_norm": 0.29220888018608093, "learning_rate": 8.77136546609139e-05, "loss": 0.0503, "step": 26660 }, { "grad_norm": 0.4034575819969177, "learning_rate": 8.770279656408385e-05, "loss": 0.0464, "step": 26670 }, { "grad_norm": 0.2843516767024994, "learning_rate": 8.769193434413265e-05, "loss": 0.0446, "step": 26680 }, { "grad_norm": 0.8963988423347473, "learning_rate": 8.76810680022482e-05, "loss": 0.0518, "step": 26690 }, { "grad_norm": 0.34458959102630615, "learning_rate": 8.767019753961878e-05, "loss": 0.0517, "step": 26700 }, { "grad_norm": 0.5141101479530334, "learning_rate": 8.765932295743321e-05, "loss": 0.0519, "step": 26710 }, { "grad_norm": 0.4882197976112366, "learning_rate": 8.764844425688068e-05, "loss": 0.0523, "step": 26720 }, { "grad_norm": 0.44412761926651, "learning_rate": 8.763756143915092e-05, "loss": 0.0448, "step": 26730 }, { "grad_norm": 0.3283916413784027, "learning_rate": 8.7626674505434e-05, "loss": 0.0484, "step": 26740 }, { "grad_norm": 0.41154050827026367, "learning_rate": 8.761578345692053e-05, "loss": 0.0528, "step": 26750 }, { "grad_norm": 0.30107763409614563, "learning_rate": 8.760488829480156e-05, "loss": 0.048, "step": 26760 }, { "grad_norm": 0.3831203579902649, "learning_rate": 8.759398902026854e-05, "loss": 0.0428, "step": 26770 }, { "grad_norm": 0.5459074974060059, "learning_rate": 8.758308563451339e-05, "loss": 0.0461, "step": 26780 }, { "grad_norm": 0.2648929953575134, "learning_rate": 8.75721781387285e-05, "loss": 0.0516, "step": 26790 }, { "grad_norm": 0.3607632517814636, "learning_rate": 8.75612665341067e-05, "loss": 0.0525, "step": 26800 }, { "grad_norm": 0.3650318384170532, "learning_rate": 8.755035082184126e-05, "loss": 0.0507, "step": 26810 }, { "grad_norm": 0.3534321188926697, "learning_rate": 8.753943100312592e-05, "loss": 0.047, "step": 26820 }, { "grad_norm": 0.6429713368415833, "learning_rate": 8.752850707915484e-05, "loss": 0.0462, "step": 26830 }, { "grad_norm": 0.42691704630851746, "learning_rate": 8.751757905112264e-05, "loss": 0.0529, "step": 26840 }, { "grad_norm": 0.4542734920978546, "learning_rate": 8.75066469202244e-05, "loss": 0.0531, "step": 26850 }, { "grad_norm": 0.4782114028930664, "learning_rate": 8.749571068765567e-05, "loss": 0.0534, "step": 26860 }, { "grad_norm": 0.4451834559440613, "learning_rate": 8.748477035461238e-05, "loss": 0.0554, "step": 26870 }, { "grad_norm": 0.41152676939964294, "learning_rate": 8.747382592229095e-05, "loss": 0.0522, "step": 26880 }, { "grad_norm": 0.48938825726509094, "learning_rate": 8.746287739188828e-05, "loss": 0.0446, "step": 26890 }, { "grad_norm": 0.5617514848709106, "learning_rate": 8.745192476460165e-05, "loss": 0.0523, "step": 26900 }, { "grad_norm": 0.4682668447494507, "learning_rate": 8.744096804162882e-05, "loss": 0.0542, "step": 26910 }, { "grad_norm": 0.44113433361053467, "learning_rate": 8.743000722416804e-05, "loss": 0.0485, "step": 26920 }, { "grad_norm": 0.39494141936302185, "learning_rate": 8.741904231341793e-05, "loss": 0.0509, "step": 26930 }, { "grad_norm": 0.4355514347553253, "learning_rate": 8.740807331057762e-05, "loss": 0.0568, "step": 26940 }, { "grad_norm": 0.36564376950263977, "learning_rate": 8.739710021684667e-05, "loss": 0.0504, "step": 26950 }, { "grad_norm": 0.4709961414337158, "learning_rate": 8.738612303342503e-05, "loss": 0.0486, "step": 26960 }, { "grad_norm": 0.5475605130195618, "learning_rate": 8.73751417615132e-05, "loss": 0.0518, "step": 26970 }, { "grad_norm": 0.5324465036392212, "learning_rate": 8.736415640231208e-05, "loss": 0.0485, "step": 26980 }, { "grad_norm": 0.4674978256225586, "learning_rate": 8.735316695702297e-05, "loss": 0.0557, "step": 26990 }, { "grad_norm": 0.3920212984085083, "learning_rate": 8.734217342684769e-05, "loss": 0.0451, "step": 27000 }, { "grad_norm": 0.47970202565193176, "learning_rate": 8.733117581298847e-05, "loss": 0.0502, "step": 27010 }, { "grad_norm": 0.43522506952285767, "learning_rate": 8.732017411664796e-05, "loss": 0.0495, "step": 27020 }, { "grad_norm": 0.36879247426986694, "learning_rate": 8.730916833902936e-05, "loss": 0.045, "step": 27030 }, { "grad_norm": 0.32338178157806396, "learning_rate": 8.729815848133618e-05, "loss": 0.0475, "step": 27040 }, { "grad_norm": 0.3051145076751709, "learning_rate": 8.728714454477247e-05, "loss": 0.0378, "step": 27050 }, { "grad_norm": 0.5103029608726501, "learning_rate": 8.727612653054269e-05, "loss": 0.0457, "step": 27060 }, { "grad_norm": 0.4247097373008728, "learning_rate": 8.726510443985176e-05, "loss": 0.0393, "step": 27070 }, { "grad_norm": 0.41955751180648804, "learning_rate": 8.725407827390503e-05, "loss": 0.0501, "step": 27080 }, { "grad_norm": 0.4912891685962677, "learning_rate": 8.724304803390833e-05, "loss": 0.046, "step": 27090 }, { "grad_norm": 0.43272948265075684, "learning_rate": 8.723201372106788e-05, "loss": 0.0479, "step": 27100 }, { "grad_norm": 0.4910946488380432, "learning_rate": 8.722097533659038e-05, "loss": 0.0464, "step": 27110 }, { "grad_norm": 0.5383174419403076, "learning_rate": 8.720993288168299e-05, "loss": 0.0597, "step": 27120 }, { "grad_norm": 0.49964118003845215, "learning_rate": 8.719888635755327e-05, "loss": 0.0492, "step": 27130 }, { "grad_norm": 0.4250996708869934, "learning_rate": 8.718783576540928e-05, "loss": 0.0513, "step": 27140 }, { "grad_norm": 0.42243248224258423, "learning_rate": 8.717678110645948e-05, "loss": 0.0525, "step": 27150 }, { "grad_norm": 0.3469915986061096, "learning_rate": 8.716572238191279e-05, "loss": 0.0466, "step": 27160 }, { "grad_norm": 0.5046111345291138, "learning_rate": 8.715465959297857e-05, "loss": 0.047, "step": 27170 }, { "grad_norm": 0.40317410230636597, "learning_rate": 8.714359274086665e-05, "loss": 0.0548, "step": 27180 }, { "grad_norm": 0.33758097887039185, "learning_rate": 8.713252182678726e-05, "loss": 0.0477, "step": 27190 }, { "grad_norm": 0.347426176071167, "learning_rate": 8.712144685195112e-05, "loss": 0.0444, "step": 27200 }, { "grad_norm": 0.44826412200927734, "learning_rate": 8.711036781756936e-05, "loss": 0.0515, "step": 27210 }, { "grad_norm": 0.40033459663391113, "learning_rate": 8.709928472485357e-05, "loss": 0.0475, "step": 27220 }, { "grad_norm": 0.3249545991420746, "learning_rate": 8.708819757501579e-05, "loss": 0.0436, "step": 27230 }, { "grad_norm": 0.25668075680732727, "learning_rate": 8.707710636926846e-05, "loss": 0.0466, "step": 27240 }, { "grad_norm": 0.36768800020217896, "learning_rate": 8.706601110882455e-05, "loss": 0.0491, "step": 27250 }, { "grad_norm": 0.29607588052749634, "learning_rate": 8.705491179489738e-05, "loss": 0.0438, "step": 27260 }, { "grad_norm": 0.31673043966293335, "learning_rate": 8.704380842870077e-05, "loss": 0.0466, "step": 27270 }, { "grad_norm": 0.33071669936180115, "learning_rate": 8.703270101144895e-05, "loss": 0.0447, "step": 27280 }, { "grad_norm": 0.4802577495574951, "learning_rate": 8.702158954435664e-05, "loss": 0.0467, "step": 27290 }, { "grad_norm": 0.31116101145744324, "learning_rate": 8.701047402863896e-05, "loss": 0.0433, "step": 27300 }, { "grad_norm": 0.5180997848510742, "learning_rate": 8.699935446551148e-05, "loss": 0.0461, "step": 27310 }, { "grad_norm": 0.3758227229118347, "learning_rate": 8.698823085619022e-05, "loss": 0.0418, "step": 27320 }, { "grad_norm": 0.3045603036880493, "learning_rate": 8.697710320189166e-05, "loss": 0.0431, "step": 27330 }, { "grad_norm": 0.5113352537155151, "learning_rate": 8.696597150383268e-05, "loss": 0.0491, "step": 27340 }, { "grad_norm": 0.42262524366378784, "learning_rate": 8.695483576323063e-05, "loss": 0.0472, "step": 27350 }, { "grad_norm": 0.4915921688079834, "learning_rate": 8.69436959813033e-05, "loss": 0.0526, "step": 27360 }, { "grad_norm": 0.3554156720638275, "learning_rate": 8.693255215926892e-05, "loss": 0.0458, "step": 27370 }, { "grad_norm": 0.5235636830329895, "learning_rate": 8.692140429834617e-05, "loss": 0.0453, "step": 27380 }, { "grad_norm": 1.1937553882598877, "learning_rate": 8.691025239975415e-05, "loss": 0.0492, "step": 27390 }, { "grad_norm": 0.29981768131256104, "learning_rate": 8.689909646471243e-05, "loss": 0.0495, "step": 27400 }, { "grad_norm": 0.554132878780365, "learning_rate": 8.688793649444099e-05, "loss": 0.0649, "step": 27410 }, { "grad_norm": 0.6664061546325684, "learning_rate": 8.687677249016029e-05, "loss": 0.0523, "step": 27420 }, { "grad_norm": 0.3513062000274658, "learning_rate": 8.686560445309118e-05, "loss": 0.0535, "step": 27430 }, { "grad_norm": 0.7206132411956787, "learning_rate": 8.685443238445499e-05, "loss": 0.0554, "step": 27440 }, { "grad_norm": 0.4744962155818939, "learning_rate": 8.68432562854735e-05, "loss": 0.051, "step": 27450 }, { "grad_norm": 0.5242881178855896, "learning_rate": 8.683207615736887e-05, "loss": 0.0476, "step": 27460 }, { "grad_norm": 0.6073147058486938, "learning_rate": 8.682089200136379e-05, "loss": 0.0589, "step": 27470 }, { "grad_norm": 0.5308888554573059, "learning_rate": 8.680970381868132e-05, "loss": 0.0468, "step": 27480 }, { "grad_norm": 0.4988384544849396, "learning_rate": 8.679851161054498e-05, "loss": 0.0503, "step": 27490 }, { "grad_norm": 0.4778937101364136, "learning_rate": 8.678731537817873e-05, "loss": 0.052, "step": 27500 }, { "grad_norm": 0.4583592414855957, "learning_rate": 8.677611512280697e-05, "loss": 0.0548, "step": 27510 }, { "grad_norm": 0.36994796991348267, "learning_rate": 8.676491084565457e-05, "loss": 0.0442, "step": 27520 }, { "grad_norm": 0.3765314817428589, "learning_rate": 8.675370254794678e-05, "loss": 0.0513, "step": 27530 }, { "grad_norm": 0.3197827637195587, "learning_rate": 8.674249023090935e-05, "loss": 0.0472, "step": 27540 }, { "grad_norm": 0.37868237495422363, "learning_rate": 8.673127389576843e-05, "loss": 0.0481, "step": 27550 }, { "grad_norm": 0.3687768280506134, "learning_rate": 8.67200535437506e-05, "loss": 0.0515, "step": 27560 }, { "grad_norm": 0.5015280246734619, "learning_rate": 8.670882917608296e-05, "loss": 0.0492, "step": 27570 }, { "grad_norm": 0.4416026473045349, "learning_rate": 8.669760079399292e-05, "loss": 0.057, "step": 27580 }, { "grad_norm": 0.39149147272109985, "learning_rate": 8.668636839870845e-05, "loss": 0.0401, "step": 27590 }, { "grad_norm": 0.45036551356315613, "learning_rate": 8.667513199145789e-05, "loss": 0.0516, "step": 27600 }, { "grad_norm": 0.45815831422805786, "learning_rate": 8.666389157347002e-05, "loss": 0.0432, "step": 27610 }, { "grad_norm": 0.4079248011112213, "learning_rate": 8.66526471459741e-05, "loss": 0.0467, "step": 27620 }, { "grad_norm": 0.37144237756729126, "learning_rate": 8.66413987101998e-05, "loss": 0.0444, "step": 27630 }, { "grad_norm": 0.3312183916568756, "learning_rate": 8.663014626737723e-05, "loss": 0.0507, "step": 27640 }, { "grad_norm": 0.46785324811935425, "learning_rate": 8.661888981873691e-05, "loss": 0.0491, "step": 27650 }, { "grad_norm": 0.4327024519443512, "learning_rate": 8.660762936550988e-05, "loss": 0.0454, "step": 27660 }, { "grad_norm": 0.4451892375946045, "learning_rate": 8.659636490892753e-05, "loss": 0.0503, "step": 27670 }, { "grad_norm": 0.4331243634223938, "learning_rate": 8.658509645022174e-05, "loss": 0.0411, "step": 27680 }, { "grad_norm": 0.4105669856071472, "learning_rate": 8.657382399062481e-05, "loss": 0.0514, "step": 27690 }, { "grad_norm": 0.4476358890533447, "learning_rate": 8.656254753136946e-05, "loss": 0.0498, "step": 27700 }, { "grad_norm": 0.8975613117218018, "learning_rate": 8.655126707368891e-05, "loss": 0.058, "step": 27710 }, { "grad_norm": 0.34799763560295105, "learning_rate": 8.653998261881672e-05, "loss": 0.0529, "step": 27720 }, { "grad_norm": 0.42044597864151, "learning_rate": 8.652869416798699e-05, "loss": 0.0533, "step": 27730 }, { "grad_norm": 1.1996281147003174, "learning_rate": 8.651740172243417e-05, "loss": 0.047, "step": 27740 }, { "grad_norm": 0.48991677165031433, "learning_rate": 8.65061052833932e-05, "loss": 0.0489, "step": 27750 }, { "grad_norm": 0.32716596126556396, "learning_rate": 8.649480485209945e-05, "loss": 0.0432, "step": 27760 }, { "grad_norm": 0.4394090175628662, "learning_rate": 8.64835004297887e-05, "loss": 0.0514, "step": 27770 }, { "grad_norm": 0.48338082432746887, "learning_rate": 8.64721920176972e-05, "loss": 0.0483, "step": 27780 }, { "grad_norm": 0.36895519495010376, "learning_rate": 8.646087961706164e-05, "loss": 0.0582, "step": 27790 }, { "grad_norm": 0.24473118782043457, "learning_rate": 8.644956322911908e-05, "loss": 0.0528, "step": 27800 }, { "grad_norm": 0.3278049826622009, "learning_rate": 8.643824285510709e-05, "loss": 0.0453, "step": 27810 }, { "grad_norm": 0.31249746680259705, "learning_rate": 8.642691849626364e-05, "loss": 0.0427, "step": 27820 }, { "grad_norm": 0.30756673216819763, "learning_rate": 8.641559015382717e-05, "loss": 0.0477, "step": 27830 }, { "grad_norm": 0.4286881387233734, "learning_rate": 8.640425782903649e-05, "loss": 0.0583, "step": 27840 }, { "grad_norm": 0.4779145419597626, "learning_rate": 8.639292152313091e-05, "loss": 0.047, "step": 27850 }, { "grad_norm": 0.421428382396698, "learning_rate": 8.638158123735015e-05, "loss": 0.0498, "step": 27860 }, { "grad_norm": 0.5148417353630066, "learning_rate": 8.637023697293436e-05, "loss": 0.047, "step": 27870 }, { "grad_norm": 0.3053436875343323, "learning_rate": 8.635888873112414e-05, "loss": 0.0453, "step": 27880 }, { "grad_norm": 0.4645369350910187, "learning_rate": 8.634753651316052e-05, "loss": 0.0485, "step": 27890 }, { "grad_norm": 0.3386497497558594, "learning_rate": 8.633618032028496e-05, "loss": 0.0565, "step": 27900 }, { "grad_norm": 0.37916287779808044, "learning_rate": 8.632482015373934e-05, "loss": 0.0483, "step": 27910 }, { "grad_norm": 0.44752901792526245, "learning_rate": 8.6313456014766e-05, "loss": 0.0532, "step": 27920 }, { "grad_norm": 0.453565388917923, "learning_rate": 8.630208790460771e-05, "loss": 0.0488, "step": 27930 }, { "grad_norm": 0.49631521105766296, "learning_rate": 8.629071582450768e-05, "loss": 0.0525, "step": 27940 }, { "grad_norm": 0.4433498680591583, "learning_rate": 8.62793397757095e-05, "loss": 0.0478, "step": 27950 }, { "grad_norm": 0.3866025507450104, "learning_rate": 8.626795975945729e-05, "loss": 0.052, "step": 27960 }, { "grad_norm": 0.3617342710494995, "learning_rate": 8.625657577699551e-05, "loss": 0.0525, "step": 27970 }, { "grad_norm": 0.3032146692276001, "learning_rate": 8.624518782956914e-05, "loss": 0.0529, "step": 27980 }, { "grad_norm": 0.32873135805130005, "learning_rate": 8.62337959184235e-05, "loss": 0.0467, "step": 27990 }, { "grad_norm": 0.29813799262046814, "learning_rate": 8.622240004480441e-05, "loss": 0.0539, "step": 28000 }, { "grad_norm": 0.46924903988838196, "learning_rate": 8.621100020995814e-05, "loss": 0.0391, "step": 28010 }, { "grad_norm": 0.4604223370552063, "learning_rate": 8.619959641513132e-05, "loss": 0.0516, "step": 28020 }, { "grad_norm": 0.34206530451774597, "learning_rate": 8.618818866157105e-05, "loss": 0.05, "step": 28030 }, { "grad_norm": 0.3280125558376312, "learning_rate": 8.617677695052487e-05, "loss": 0.044, "step": 28040 }, { "grad_norm": 0.30592644214630127, "learning_rate": 8.616536128324078e-05, "loss": 0.0445, "step": 28050 }, { "grad_norm": 0.375765323638916, "learning_rate": 8.615394166096712e-05, "loss": 0.0471, "step": 28060 }, { "grad_norm": 0.31974369287490845, "learning_rate": 8.614251808495279e-05, "loss": 0.0473, "step": 28070 }, { "grad_norm": 0.31180351972579956, "learning_rate": 8.6131090556447e-05, "loss": 0.0505, "step": 28080 }, { "grad_norm": 0.3644102215766907, "learning_rate": 8.611965907669947e-05, "loss": 0.0378, "step": 28090 }, { "grad_norm": 0.4396370053291321, "learning_rate": 8.610822364696034e-05, "loss": 0.0535, "step": 28100 }, { "grad_norm": 0.6918260455131531, "learning_rate": 8.609678426848015e-05, "loss": 0.0511, "step": 28110 }, { "grad_norm": 0.3020876348018646, "learning_rate": 8.60853409425099e-05, "loss": 0.0464, "step": 28120 }, { "grad_norm": 0.40975284576416016, "learning_rate": 8.607389367030104e-05, "loss": 0.0559, "step": 28130 }, { "grad_norm": 0.39107486605644226, "learning_rate": 8.606244245310538e-05, "loss": 0.053, "step": 28140 }, { "grad_norm": 0.43536990880966187, "learning_rate": 8.605098729217525e-05, "loss": 0.0522, "step": 28150 }, { "grad_norm": 0.3379436731338501, "learning_rate": 8.603952818876335e-05, "loss": 0.0525, "step": 28160 }, { "grad_norm": 0.4242214262485504, "learning_rate": 8.602806514412281e-05, "loss": 0.0428, "step": 28170 }, { "grad_norm": 0.39187178015708923, "learning_rate": 8.601659815950726e-05, "loss": 0.0489, "step": 28180 }, { "grad_norm": 0.39235371351242065, "learning_rate": 8.600512723617067e-05, "loss": 0.0514, "step": 28190 }, { "grad_norm": 0.41147589683532715, "learning_rate": 8.59936523753675e-05, "loss": 0.0473, "step": 28200 }, { "grad_norm": 0.4039556384086609, "learning_rate": 8.598217357835264e-05, "loss": 0.0505, "step": 28210 }, { "grad_norm": 0.46853122115135193, "learning_rate": 8.597069084638135e-05, "loss": 0.0454, "step": 28220 }, { "grad_norm": 0.4408825635910034, "learning_rate": 8.595920418070939e-05, "loss": 0.0509, "step": 28230 }, { "grad_norm": 0.3795383870601654, "learning_rate": 8.594771358259295e-05, "loss": 0.0487, "step": 28240 }, { "grad_norm": 0.5004359483718872, "learning_rate": 8.593621905328858e-05, "loss": 0.0444, "step": 28250 }, { "grad_norm": 0.48532870411872864, "learning_rate": 8.592472059405333e-05, "loss": 0.0604, "step": 28260 }, { "grad_norm": 0.3166752755641937, "learning_rate": 8.591321820614464e-05, "loss": 0.0464, "step": 28270 }, { "grad_norm": 0.4576345384120941, "learning_rate": 8.590171189082041e-05, "loss": 0.0481, "step": 28280 }, { "grad_norm": 0.30938461422920227, "learning_rate": 8.589020164933894e-05, "loss": 0.0456, "step": 28290 }, { "grad_norm": 0.5042644739151001, "learning_rate": 8.587868748295898e-05, "loss": 0.0523, "step": 28300 }, { "grad_norm": 0.5092750191688538, "learning_rate": 8.586716939293971e-05, "loss": 0.0581, "step": 28310 }, { "grad_norm": 0.4355315864086151, "learning_rate": 8.58556473805407e-05, "loss": 0.0462, "step": 28320 }, { "grad_norm": 0.3043104410171509, "learning_rate": 8.584412144702202e-05, "loss": 0.0492, "step": 28330 }, { "grad_norm": 0.34450268745422363, "learning_rate": 8.58325915936441e-05, "loss": 0.0478, "step": 28340 }, { "grad_norm": 0.45659002661705017, "learning_rate": 8.582105782166783e-05, "loss": 0.0478, "step": 28350 }, { "grad_norm": 0.7245900630950928, "learning_rate": 8.580952013235455e-05, "loss": 0.0555, "step": 28360 }, { "grad_norm": 0.31171488761901855, "learning_rate": 8.579797852696596e-05, "loss": 0.0486, "step": 28370 }, { "grad_norm": 0.2766164541244507, "learning_rate": 8.578643300676428e-05, "loss": 0.0454, "step": 28380 }, { "grad_norm": 0.5893478393554688, "learning_rate": 8.577488357301209e-05, "loss": 0.0537, "step": 28390 }, { "grad_norm": 0.6793841123580933, "learning_rate": 8.576333022697242e-05, "loss": 0.0551, "step": 28400 }, { "grad_norm": 0.35744354128837585, "learning_rate": 8.575177296990873e-05, "loss": 0.0437, "step": 28410 }, { "grad_norm": 0.5461445450782776, "learning_rate": 8.574021180308489e-05, "loss": 0.0544, "step": 28420 }, { "grad_norm": 0.38370418548583984, "learning_rate": 8.572864672776523e-05, "loss": 0.0467, "step": 28430 }, { "grad_norm": 0.5435857176780701, "learning_rate": 8.571707774521447e-05, "loss": 0.0497, "step": 28440 }, { "grad_norm": 0.3369999825954437, "learning_rate": 8.57055048566978e-05, "loss": 0.0512, "step": 28450 }, { "grad_norm": 0.9231253266334534, "learning_rate": 8.569392806348078e-05, "loss": 0.0464, "step": 28460 }, { "grad_norm": 0.49652916193008423, "learning_rate": 8.568234736682947e-05, "loss": 0.0545, "step": 28470 }, { "grad_norm": 0.49733296036720276, "learning_rate": 8.567076276801029e-05, "loss": 0.0498, "step": 28480 }, { "grad_norm": 0.4316411018371582, "learning_rate": 8.565917426829013e-05, "loss": 0.0464, "step": 28490 }, { "grad_norm": 0.43930116295814514, "learning_rate": 8.564758186893628e-05, "loss": 0.0501, "step": 28500 }, { "grad_norm": 0.48394352197647095, "learning_rate": 8.563598557121649e-05, "loss": 0.0451, "step": 28510 }, { "grad_norm": 0.33701765537261963, "learning_rate": 8.562438537639888e-05, "loss": 0.0471, "step": 28520 }, { "grad_norm": 0.3540489971637726, "learning_rate": 8.561278128575206e-05, "loss": 0.0519, "step": 28530 }, { "grad_norm": 0.4148014783859253, "learning_rate": 8.5601173300545e-05, "loss": 0.0468, "step": 28540 }, { "grad_norm": 0.35818812251091003, "learning_rate": 8.558956142204717e-05, "loss": 0.0527, "step": 28550 }, { "grad_norm": 0.4773383140563965, "learning_rate": 8.55779456515284e-05, "loss": 0.0502, "step": 28560 }, { "grad_norm": 0.3984370231628418, "learning_rate": 8.556632599025898e-05, "loss": 0.0522, "step": 28570 }, { "grad_norm": 0.4899783134460449, "learning_rate": 8.555470243950964e-05, "loss": 0.0507, "step": 28580 }, { "grad_norm": 0.41718122363090515, "learning_rate": 8.554307500055148e-05, "loss": 0.0507, "step": 28590 }, { "grad_norm": 0.4148993194103241, "learning_rate": 8.553144367465609e-05, "loss": 0.0502, "step": 28600 }, { "grad_norm": 0.48148733377456665, "learning_rate": 8.551980846309544e-05, "loss": 0.0498, "step": 28610 }, { "grad_norm": 0.2895423173904419, "learning_rate": 8.550816936714193e-05, "loss": 0.0442, "step": 28620 }, { "grad_norm": 0.430608332157135, "learning_rate": 8.549652638806841e-05, "loss": 0.0514, "step": 28630 }, { "grad_norm": 0.4326041042804718, "learning_rate": 8.548487952714812e-05, "loss": 0.0488, "step": 28640 }, { "grad_norm": 0.5023728609085083, "learning_rate": 8.547322878565478e-05, "loss": 0.0525, "step": 28650 }, { "grad_norm": 0.4311089813709259, "learning_rate": 8.546157416486245e-05, "loss": 0.0502, "step": 28660 }, { "grad_norm": 0.5233955383300781, "learning_rate": 8.54499156660457e-05, "loss": 0.0478, "step": 28670 }, { "grad_norm": 0.45324045419692993, "learning_rate": 8.543825329047947e-05, "loss": 0.0506, "step": 28680 }, { "grad_norm": 0.34035009145736694, "learning_rate": 8.542658703943913e-05, "loss": 0.0544, "step": 28690 }, { "grad_norm": 0.538594126701355, "learning_rate": 8.541491691420051e-05, "loss": 0.0419, "step": 28700 }, { "grad_norm": 0.43577125668525696, "learning_rate": 8.54032429160398e-05, "loss": 0.0498, "step": 28710 }, { "grad_norm": 0.44503355026245117, "learning_rate": 8.539156504623369e-05, "loss": 0.0575, "step": 28720 }, { "grad_norm": 0.5683613419532776, "learning_rate": 8.537988330605923e-05, "loss": 0.0533, "step": 28730 }, { "grad_norm": 0.3477596342563629, "learning_rate": 8.536819769679393e-05, "loss": 0.0451, "step": 28740 }, { "grad_norm": 0.30335575342178345, "learning_rate": 8.53565082197157e-05, "loss": 0.0493, "step": 28750 }, { "grad_norm": 0.35682347416877747, "learning_rate": 8.534481487610289e-05, "loss": 0.0517, "step": 28760 }, { "grad_norm": 0.34498924016952515, "learning_rate": 8.533311766723428e-05, "loss": 0.0508, "step": 28770 }, { "grad_norm": 0.47967034578323364, "learning_rate": 8.532141659438901e-05, "loss": 0.0493, "step": 28780 }, { "grad_norm": 0.4097300171852112, "learning_rate": 8.530971165884675e-05, "loss": 0.0454, "step": 28790 }, { "grad_norm": 0.44545310735702515, "learning_rate": 8.529800286188752e-05, "loss": 0.0405, "step": 28800 }, { "grad_norm": 0.4578157663345337, "learning_rate": 8.528629020479175e-05, "loss": 0.0508, "step": 28810 }, { "grad_norm": 0.3941586911678314, "learning_rate": 8.527457368884033e-05, "loss": 0.0481, "step": 28820 }, { "grad_norm": 0.3867120146751404, "learning_rate": 8.526285331531458e-05, "loss": 0.0453, "step": 28830 }, { "grad_norm": 0.36696097254753113, "learning_rate": 8.525112908549621e-05, "loss": 0.0476, "step": 28840 }, { "grad_norm": 0.30575406551361084, "learning_rate": 8.523940100066735e-05, "loss": 0.0416, "step": 28850 }, { "grad_norm": 0.2720337510108948, "learning_rate": 8.52276690621106e-05, "loss": 0.0482, "step": 28860 }, { "grad_norm": 0.9924188256263733, "learning_rate": 8.521593327110889e-05, "loss": 0.0433, "step": 28870 }, { "grad_norm": 0.36989444494247437, "learning_rate": 8.520419362894569e-05, "loss": 0.046, "step": 28880 }, { "grad_norm": 0.43389326333999634, "learning_rate": 8.51924501369048e-05, "loss": 0.0429, "step": 28890 }, { "grad_norm": 0.34129786491394043, "learning_rate": 8.518070279627047e-05, "loss": 0.0464, "step": 28900 }, { "grad_norm": 0.40833398699760437, "learning_rate": 8.516895160832737e-05, "loss": 0.049, "step": 28910 }, { "grad_norm": 0.3984971344470978, "learning_rate": 8.515719657436061e-05, "loss": 0.0472, "step": 28920 }, { "grad_norm": 0.4243762195110321, "learning_rate": 8.514543769565568e-05, "loss": 0.0509, "step": 28930 }, { "grad_norm": 0.5162206888198853, "learning_rate": 8.513367497349853e-05, "loss": 0.0514, "step": 28940 }, { "grad_norm": 0.3615514934062958, "learning_rate": 8.51219084091755e-05, "loss": 0.0463, "step": 28950 }, { "grad_norm": 0.5837903618812561, "learning_rate": 8.511013800397338e-05, "loss": 0.0473, "step": 28960 }, { "grad_norm": 0.3150383532047272, "learning_rate": 8.509836375917937e-05, "loss": 0.0462, "step": 28970 }, { "grad_norm": 0.3194621503353119, "learning_rate": 8.508658567608104e-05, "loss": 0.0454, "step": 28980 }, { "grad_norm": 0.37431231141090393, "learning_rate": 8.507480375596647e-05, "loss": 0.0458, "step": 28990 }, { "grad_norm": 0.3508991301059723, "learning_rate": 8.506301800012408e-05, "loss": 0.0546, "step": 29000 }, { "grad_norm": 0.312324583530426, "learning_rate": 8.505122840984278e-05, "loss": 0.0484, "step": 29010 }, { "grad_norm": 0.4048880934715271, "learning_rate": 8.503943498641182e-05, "loss": 0.0443, "step": 29020 }, { "grad_norm": 0.716925323009491, "learning_rate": 8.502763773112095e-05, "loss": 0.0428, "step": 29030 }, { "grad_norm": 0.33775755763053894, "learning_rate": 8.501583664526026e-05, "loss": 0.0504, "step": 29040 }, { "grad_norm": 0.3109666407108307, "learning_rate": 8.500403173012032e-05, "loss": 0.0484, "step": 29050 }, { "grad_norm": 0.4841243028640747, "learning_rate": 8.499222298699211e-05, "loss": 0.0476, "step": 29060 }, { "grad_norm": 0.41483399271965027, "learning_rate": 8.498041041716701e-05, "loss": 0.0474, "step": 29070 }, { "grad_norm": 0.48106151819229126, "learning_rate": 8.496859402193681e-05, "loss": 0.0529, "step": 29080 }, { "grad_norm": 0.4548751711845398, "learning_rate": 8.495677380259374e-05, "loss": 0.0497, "step": 29090 }, { "grad_norm": 0.6814612150192261, "learning_rate": 8.494494976043045e-05, "loss": 0.0522, "step": 29100 }, { "grad_norm": 0.6429638266563416, "learning_rate": 8.493312189673998e-05, "loss": 0.0529, "step": 29110 }, { "grad_norm": 0.3837997019290924, "learning_rate": 8.492129021281584e-05, "loss": 0.0467, "step": 29120 }, { "grad_norm": 0.39997580647468567, "learning_rate": 8.490945470995188e-05, "loss": 0.0514, "step": 29130 }, { "grad_norm": 0.42170730233192444, "learning_rate": 8.489761538944247e-05, "loss": 0.0477, "step": 29140 }, { "grad_norm": 0.38647815585136414, "learning_rate": 8.48857722525823e-05, "loss": 0.0481, "step": 29150 }, { "grad_norm": 0.44029319286346436, "learning_rate": 8.487392530066652e-05, "loss": 0.0533, "step": 29160 }, { "grad_norm": 0.4328296482563019, "learning_rate": 8.486207453499069e-05, "loss": 0.0481, "step": 29170 }, { "grad_norm": 0.5116914510726929, "learning_rate": 8.485021995685082e-05, "loss": 0.0487, "step": 29180 }, { "grad_norm": 0.3165960907936096, "learning_rate": 8.483836156754328e-05, "loss": 0.048, "step": 29190 }, { "grad_norm": 0.4333420693874359, "learning_rate": 8.482649936836491e-05, "loss": 0.0422, "step": 29200 }, { "grad_norm": 0.496488094329834, "learning_rate": 8.481463336061293e-05, "loss": 0.0539, "step": 29210 }, { "grad_norm": 0.3374953269958496, "learning_rate": 8.480276354558496e-05, "loss": 0.0452, "step": 29220 }, { "grad_norm": 0.3124271035194397, "learning_rate": 8.479088992457913e-05, "loss": 0.0484, "step": 29230 }, { "grad_norm": 0.40687692165374756, "learning_rate": 8.477901249889387e-05, "loss": 0.0458, "step": 29240 }, { "grad_norm": 0.44524556398391724, "learning_rate": 8.47671312698281e-05, "loss": 0.0566, "step": 29250 }, { "grad_norm": 0.6711841821670532, "learning_rate": 8.475524623868112e-05, "loss": 0.0487, "step": 29260 }, { "grad_norm": 0.5385261178016663, "learning_rate": 8.474335740675266e-05, "loss": 0.0479, "step": 29270 }, { "grad_norm": 0.5284820795059204, "learning_rate": 8.473146477534289e-05, "loss": 0.0449, "step": 29280 }, { "grad_norm": 0.3848534822463989, "learning_rate": 8.471956834575232e-05, "loss": 0.0399, "step": 29290 }, { "grad_norm": 0.512983500957489, "learning_rate": 8.470766811928197e-05, "loss": 0.0474, "step": 29300 }, { "grad_norm": 0.3924538791179657, "learning_rate": 8.469576409723323e-05, "loss": 0.0531, "step": 29310 }, { "grad_norm": 0.6258087754249573, "learning_rate": 8.468385628090788e-05, "loss": 0.0449, "step": 29320 }, { "grad_norm": 0.40899714827537537, "learning_rate": 8.467194467160815e-05, "loss": 0.0448, "step": 29330 }, { "grad_norm": 0.518359363079071, "learning_rate": 8.466002927063667e-05, "loss": 0.052, "step": 29340 }, { "grad_norm": 0.47970467805862427, "learning_rate": 8.464811007929651e-05, "loss": 0.0434, "step": 29350 }, { "grad_norm": 0.45955944061279297, "learning_rate": 8.463618709889114e-05, "loss": 0.0496, "step": 29360 }, { "grad_norm": 0.3498273491859436, "learning_rate": 8.462426033072442e-05, "loss": 0.0505, "step": 29370 }, { "grad_norm": 0.8182283639907837, "learning_rate": 8.461232977610061e-05, "loss": 0.0488, "step": 29380 }, { "grad_norm": 0.42760151624679565, "learning_rate": 8.46003954363245e-05, "loss": 0.055, "step": 29390 }, { "grad_norm": 0.37000951170921326, "learning_rate": 8.458845731270115e-05, "loss": 0.0481, "step": 29400 }, { "grad_norm": 0.4007814824581146, "learning_rate": 8.45765154065361e-05, "loss": 0.0547, "step": 29410 }, { "grad_norm": 0.6629284024238586, "learning_rate": 8.456456971913532e-05, "loss": 0.0492, "step": 29420 }, { "grad_norm": 0.5270248055458069, "learning_rate": 8.455262025180517e-05, "loss": 0.0543, "step": 29430 }, { "grad_norm": 0.40235719084739685, "learning_rate": 8.454066700585242e-05, "loss": 0.0476, "step": 29440 }, { "grad_norm": 0.49042657017707825, "learning_rate": 8.452870998258423e-05, "loss": 0.0516, "step": 29450 }, { "grad_norm": 0.271808385848999, "learning_rate": 8.451674918330825e-05, "loss": 0.0482, "step": 29460 }, { "grad_norm": 0.31201112270355225, "learning_rate": 8.450478460933246e-05, "loss": 0.0439, "step": 29470 }, { "grad_norm": 0.33143365383148193, "learning_rate": 8.449281626196532e-05, "loss": 0.0477, "step": 29480 }, { "grad_norm": 0.5223338603973389, "learning_rate": 8.448084414251564e-05, "loss": 0.049, "step": 29490 }, { "grad_norm": 0.5128478407859802, "learning_rate": 8.446886825229271e-05, "loss": 0.0462, "step": 29500 }, { "grad_norm": 0.4074990153312683, "learning_rate": 8.445688859260615e-05, "loss": 0.0449, "step": 29510 }, { "grad_norm": 0.4206814169883728, "learning_rate": 8.444490516476606e-05, "loss": 0.0469, "step": 29520 }, { "grad_norm": 0.4262973666191101, "learning_rate": 8.443291797008293e-05, "loss": 0.0481, "step": 29530 }, { "grad_norm": 0.393145889043808, "learning_rate": 8.442092700986765e-05, "loss": 0.0511, "step": 29540 }, { "grad_norm": 1.2611496448516846, "learning_rate": 8.440893228543156e-05, "loss": 0.0571, "step": 29550 }, { "grad_norm": 0.4319279193878174, "learning_rate": 8.439693379808638e-05, "loss": 0.0379, "step": 29560 }, { "grad_norm": 0.2804116904735565, "learning_rate": 8.43849315491442e-05, "loss": 0.0489, "step": 29570 }, { "grad_norm": 0.40407484769821167, "learning_rate": 8.437292553991763e-05, "loss": 0.0538, "step": 29580 }, { "grad_norm": 0.3242136538028717, "learning_rate": 8.436091577171959e-05, "loss": 0.0456, "step": 29590 }, { "grad_norm": 0.38638803362846375, "learning_rate": 8.434890224586347e-05, "loss": 0.0433, "step": 29600 }, { "grad_norm": 0.3549934923648834, "learning_rate": 8.433688496366303e-05, "loss": 0.0474, "step": 29610 }, { "grad_norm": 0.5439973473548889, "learning_rate": 8.432486392643248e-05, "loss": 0.0547, "step": 29620 }, { "grad_norm": 0.5034233927726746, "learning_rate": 8.431283913548643e-05, "loss": 0.0481, "step": 29630 }, { "grad_norm": 0.3160625100135803, "learning_rate": 8.430081059213985e-05, "loss": 0.0466, "step": 29640 }, { "grad_norm": 0.4314109683036804, "learning_rate": 8.428877829770823e-05, "loss": 0.0485, "step": 29650 }, { "grad_norm": 0.31347909569740295, "learning_rate": 8.427674225350735e-05, "loss": 0.0564, "step": 29660 }, { "grad_norm": 0.35585132241249084, "learning_rate": 8.426470246085347e-05, "loss": 0.0479, "step": 29670 }, { "grad_norm": 0.41526591777801514, "learning_rate": 8.425265892106324e-05, "loss": 0.0448, "step": 29680 }, { "grad_norm": 0.2977835536003113, "learning_rate": 8.424061163545374e-05, "loss": 0.0442, "step": 29690 }, { "grad_norm": 0.4837217926979065, "learning_rate": 8.422856060534243e-05, "loss": 0.0521, "step": 29700 }, { "grad_norm": 0.4309479892253876, "learning_rate": 8.421650583204718e-05, "loss": 0.0465, "step": 29710 }, { "grad_norm": 0.41712889075279236, "learning_rate": 8.420444731688633e-05, "loss": 0.0489, "step": 29720 }, { "grad_norm": 0.4148486256599426, "learning_rate": 8.419238506117852e-05, "loss": 0.0442, "step": 29730 }, { "grad_norm": 0.47585922479629517, "learning_rate": 8.418031906624289e-05, "loss": 0.0445, "step": 29740 }, { "grad_norm": 0.3481978476047516, "learning_rate": 8.416824933339898e-05, "loss": 0.0492, "step": 29750 }, { "grad_norm": 0.5153773427009583, "learning_rate": 8.415617586396667e-05, "loss": 0.052, "step": 29760 }, { "grad_norm": 0.5533387064933777, "learning_rate": 8.414409865926632e-05, "loss": 0.0529, "step": 29770 }, { "grad_norm": 0.5446557402610779, "learning_rate": 8.413201772061867e-05, "loss": 0.0445, "step": 29780 }, { "grad_norm": 0.43531695008277893, "learning_rate": 8.411993304934488e-05, "loss": 0.046, "step": 29790 }, { "grad_norm": 0.42596304416656494, "learning_rate": 8.410784464676654e-05, "loss": 0.0408, "step": 29800 }, { "grad_norm": 0.2817116975784302, "learning_rate": 8.409575251420556e-05, "loss": 0.0447, "step": 29810 }, { "grad_norm": 0.31225964426994324, "learning_rate": 8.408365665298435e-05, "loss": 0.0495, "step": 29820 }, { "grad_norm": 0.42256590723991394, "learning_rate": 8.40715570644257e-05, "loss": 0.0478, "step": 29830 }, { "grad_norm": 0.3095856308937073, "learning_rate": 8.40594537498528e-05, "loss": 0.0442, "step": 29840 }, { "grad_norm": 0.3564204275608063, "learning_rate": 8.404734671058924e-05, "loss": 0.0481, "step": 29850 }, { "grad_norm": 0.36233967542648315, "learning_rate": 8.403523594795902e-05, "loss": 0.0501, "step": 29860 }, { "grad_norm": 0.3401698172092438, "learning_rate": 8.402312146328659e-05, "loss": 0.0512, "step": 29870 }, { "grad_norm": 0.3708573579788208, "learning_rate": 8.401100325789675e-05, "loss": 0.0508, "step": 29880 }, { "grad_norm": 0.4079502820968628, "learning_rate": 8.399888133311472e-05, "loss": 0.0486, "step": 29890 }, { "grad_norm": 0.3856699466705322, "learning_rate": 8.398675569026613e-05, "loss": 0.043, "step": 29900 }, { "grad_norm": 0.27911853790283203, "learning_rate": 8.397462633067705e-05, "loss": 0.0409, "step": 29910 }, { "grad_norm": 0.5081014633178711, "learning_rate": 8.396249325567392e-05, "loss": 0.0464, "step": 29920 }, { "grad_norm": 0.33134523034095764, "learning_rate": 8.395035646658357e-05, "loss": 0.0526, "step": 29930 }, { "grad_norm": 0.3697454333305359, "learning_rate": 8.39382159647333e-05, "loss": 0.053, "step": 29940 }, { "grad_norm": 0.4847826659679413, "learning_rate": 8.392607175145075e-05, "loss": 0.0482, "step": 29950 }, { "grad_norm": 0.5707476735115051, "learning_rate": 8.3913923828064e-05, "loss": 0.046, "step": 29960 }, { "grad_norm": 0.5770355463027954, "learning_rate": 8.390177219590152e-05, "loss": 0.0477, "step": 29970 }, { "grad_norm": 0.3223172724246979, "learning_rate": 8.388961685629222e-05, "loss": 0.0445, "step": 29980 }, { "grad_norm": 0.31060028076171875, "learning_rate": 8.387745781056536e-05, "loss": 0.0578, "step": 29990 }, { "grad_norm": 0.3349703252315521, "learning_rate": 8.386529506005065e-05, "loss": 0.0439, "step": 30000 }, { "grad_norm": 0.49424776434898376, "learning_rate": 8.38531286060782e-05, "loss": 0.0414, "step": 30010 }, { "grad_norm": 0.32234254479408264, "learning_rate": 8.384095844997849e-05, "loss": 0.0508, "step": 30020 }, { "grad_norm": 0.3971530497074127, "learning_rate": 8.382878459308245e-05, "loss": 0.0491, "step": 30030 }, { "grad_norm": 0.3904407024383545, "learning_rate": 8.381660703672138e-05, "loss": 0.0433, "step": 30040 }, { "grad_norm": 0.29763659834861755, "learning_rate": 8.380442578222702e-05, "loss": 0.0387, "step": 30050 }, { "grad_norm": 0.5069296360015869, "learning_rate": 8.379224083093146e-05, "loss": 0.0449, "step": 30060 }, { "grad_norm": 0.5765469670295715, "learning_rate": 8.378005218416727e-05, "loss": 0.0465, "step": 30070 }, { "grad_norm": 0.3786625564098358, "learning_rate": 8.376785984326735e-05, "loss": 0.0442, "step": 30080 }, { "grad_norm": 0.35056018829345703, "learning_rate": 8.375566380956506e-05, "loss": 0.0419, "step": 30090 }, { "grad_norm": 0.6318167448043823, "learning_rate": 8.374346408439411e-05, "loss": 0.0486, "step": 30100 }, { "grad_norm": 0.3973468244075775, "learning_rate": 8.373126066908868e-05, "loss": 0.0409, "step": 30110 }, { "grad_norm": 0.5418396592140198, "learning_rate": 8.371905356498326e-05, "loss": 0.0412, "step": 30120 }, { "grad_norm": 0.3747638463973999, "learning_rate": 8.370684277341288e-05, "loss": 0.0435, "step": 30130 }, { "grad_norm": 0.45416000485420227, "learning_rate": 8.369462829571282e-05, "loss": 0.0446, "step": 30140 }, { "grad_norm": 0.35953307151794434, "learning_rate": 8.36824101332189e-05, "loss": 0.0485, "step": 30150 }, { "grad_norm": 0.3932609558105469, "learning_rate": 8.367018828726721e-05, "loss": 0.0427, "step": 30160 }, { "grad_norm": 0.31313803791999817, "learning_rate": 8.365796275919438e-05, "loss": 0.0435, "step": 30170 }, { "grad_norm": 0.3111817240715027, "learning_rate": 8.364573355033734e-05, "loss": 0.0478, "step": 30180 }, { "grad_norm": 0.42834630608558655, "learning_rate": 8.363350066203346e-05, "loss": 0.0472, "step": 30190 }, { "grad_norm": 0.5150425434112549, "learning_rate": 8.362126409562053e-05, "loss": 0.0495, "step": 30200 }, { "grad_norm": 0.40309420228004456, "learning_rate": 8.360902385243667e-05, "loss": 0.0503, "step": 30210 }, { "grad_norm": 0.43859320878982544, "learning_rate": 8.359677993382052e-05, "loss": 0.0516, "step": 30220 }, { "grad_norm": 0.325191855430603, "learning_rate": 8.358453234111103e-05, "loss": 0.0442, "step": 30230 }, { "grad_norm": 0.3320876359939575, "learning_rate": 8.357228107564756e-05, "loss": 0.0535, "step": 30240 }, { "grad_norm": 0.3971262574195862, "learning_rate": 8.356002613876993e-05, "loss": 0.0518, "step": 30250 }, { "grad_norm": 0.859880268573761, "learning_rate": 8.35477675318183e-05, "loss": 0.0479, "step": 30260 }, { "grad_norm": 0.31782668828964233, "learning_rate": 8.353550525613323e-05, "loss": 0.0525, "step": 30270 }, { "grad_norm": 0.4466150403022766, "learning_rate": 8.352323931305572e-05, "loss": 0.0476, "step": 30280 }, { "grad_norm": 0.6017975807189941, "learning_rate": 8.351096970392717e-05, "loss": 0.0541, "step": 30290 }, { "grad_norm": 0.47307130694389343, "learning_rate": 8.349869643008937e-05, "loss": 0.0444, "step": 30300 }, { "grad_norm": 0.4004734456539154, "learning_rate": 8.348641949288449e-05, "loss": 0.0452, "step": 30310 }, { "grad_norm": 0.40362951159477234, "learning_rate": 8.34741388936551e-05, "loss": 0.0489, "step": 30320 }, { "grad_norm": 0.306202232837677, "learning_rate": 8.346185463374423e-05, "loss": 0.0447, "step": 30330 }, { "grad_norm": 0.5686696767807007, "learning_rate": 8.344956671449524e-05, "loss": 0.048, "step": 30340 }, { "grad_norm": 0.37945306301116943, "learning_rate": 8.343727513725192e-05, "loss": 0.0488, "step": 30350 }, { "grad_norm": 0.2878044843673706, "learning_rate": 8.342497990335847e-05, "loss": 0.0463, "step": 30360 }, { "grad_norm": 1.3792911767959595, "learning_rate": 8.341268101415946e-05, "loss": 0.0586, "step": 30370 }, { "grad_norm": 0.5475983619689941, "learning_rate": 8.34003784709999e-05, "loss": 0.0594, "step": 30380 }, { "grad_norm": 0.45894214510917664, "learning_rate": 8.338807227522516e-05, "loss": 0.0458, "step": 30390 }, { "grad_norm": 0.3194667398929596, "learning_rate": 8.337576242818103e-05, "loss": 0.0516, "step": 30400 }, { "grad_norm": 0.39861199259757996, "learning_rate": 8.336344893121372e-05, "loss": 0.0517, "step": 30410 }, { "grad_norm": 0.31062328815460205, "learning_rate": 8.335113178566977e-05, "loss": 0.0448, "step": 30420 }, { "grad_norm": 0.6568999290466309, "learning_rate": 8.33388109928962e-05, "loss": 0.0437, "step": 30430 }, { "grad_norm": 0.36630162596702576, "learning_rate": 8.33264865542404e-05, "loss": 0.0492, "step": 30440 }, { "grad_norm": 0.5197640657424927, "learning_rate": 8.331415847105013e-05, "loss": 0.0419, "step": 30450 }, { "grad_norm": 0.27719637751579285, "learning_rate": 8.330182674467357e-05, "loss": 0.0453, "step": 30460 }, { "grad_norm": 0.46786877512931824, "learning_rate": 8.32894913764593e-05, "loss": 0.0473, "step": 30470 }, { "grad_norm": 0.5447714924812317, "learning_rate": 8.327715236775633e-05, "loss": 0.0496, "step": 30480 }, { "grad_norm": 0.5463007092475891, "learning_rate": 8.326480971991398e-05, "loss": 0.0448, "step": 30490 }, { "grad_norm": 0.28953948616981506, "learning_rate": 8.325246343428206e-05, "loss": 0.0483, "step": 30500 }, { "grad_norm": 0.49681565165519714, "learning_rate": 8.324011351221072e-05, "loss": 0.043, "step": 30510 }, { "grad_norm": 0.3578607439994812, "learning_rate": 8.322775995505057e-05, "loss": 0.0509, "step": 30520 }, { "grad_norm": 0.6136692762374878, "learning_rate": 8.321540276415254e-05, "loss": 0.0432, "step": 30530 }, { "grad_norm": 0.6115115284919739, "learning_rate": 8.320304194086798e-05, "loss": 0.0487, "step": 30540 }, { "grad_norm": 0.5106978416442871, "learning_rate": 8.31906774865487e-05, "loss": 0.0502, "step": 30550 }, { "grad_norm": 0.5633147358894348, "learning_rate": 8.317830940254682e-05, "loss": 0.0477, "step": 30560 }, { "grad_norm": 0.3992299735546112, "learning_rate": 8.316593769021491e-05, "loss": 0.0473, "step": 30570 }, { "grad_norm": 0.44663840532302856, "learning_rate": 8.315356235090592e-05, "loss": 0.0441, "step": 30580 }, { "grad_norm": 0.3918093144893646, "learning_rate": 8.314118338597319e-05, "loss": 0.0423, "step": 30590 }, { "grad_norm": 0.331563800573349, "learning_rate": 8.312880079677048e-05, "loss": 0.0462, "step": 30600 }, { "grad_norm": 0.24941380321979523, "learning_rate": 8.311641458465191e-05, "loss": 0.0502, "step": 30610 }, { "grad_norm": 0.3900989294052124, "learning_rate": 8.310402475097205e-05, "loss": 0.0481, "step": 30620 }, { "grad_norm": 0.5216914415359497, "learning_rate": 8.309163129708581e-05, "loss": 0.0501, "step": 30630 }, { "grad_norm": 0.4242701828479767, "learning_rate": 8.307923422434852e-05, "loss": 0.0457, "step": 30640 }, { "grad_norm": 0.46181178092956543, "learning_rate": 8.30668335341159e-05, "loss": 0.0487, "step": 30650 }, { "grad_norm": 0.43775713443756104, "learning_rate": 8.305442922774408e-05, "loss": 0.0389, "step": 30660 }, { "grad_norm": 0.3483274281024933, "learning_rate": 8.304202130658959e-05, "loss": 0.0425, "step": 30670 }, { "grad_norm": 0.4923296570777893, "learning_rate": 8.302960977200931e-05, "loss": 0.0491, "step": 30680 }, { "grad_norm": 0.4631172716617584, "learning_rate": 8.301719462536058e-05, "loss": 0.0454, "step": 30690 }, { "grad_norm": 0.48302507400512695, "learning_rate": 8.300477586800108e-05, "loss": 0.0518, "step": 30700 }, { "grad_norm": 0.3646997809410095, "learning_rate": 8.299235350128892e-05, "loss": 0.0442, "step": 30710 }, { "grad_norm": 0.379598468542099, "learning_rate": 8.297992752658258e-05, "loss": 0.0481, "step": 30720 }, { "grad_norm": 0.34621185064315796, "learning_rate": 8.296749794524094e-05, "loss": 0.0532, "step": 30730 }, { "grad_norm": 0.8367431163787842, "learning_rate": 8.295506475862332e-05, "loss": 0.0457, "step": 30740 }, { "grad_norm": 0.4500652253627777, "learning_rate": 8.294262796808933e-05, "loss": 0.0474, "step": 30750 }, { "grad_norm": 0.4200296700000763, "learning_rate": 8.293018757499909e-05, "loss": 0.0518, "step": 30760 }, { "grad_norm": 0.324677973985672, "learning_rate": 8.291774358071305e-05, "loss": 0.0442, "step": 30770 }, { "grad_norm": 0.3714374899864197, "learning_rate": 8.290529598659205e-05, "loss": 0.0459, "step": 30780 }, { "grad_norm": 0.299166738986969, "learning_rate": 8.289284479399738e-05, "loss": 0.0455, "step": 30790 }, { "grad_norm": 0.48782044649124146, "learning_rate": 8.288039000429064e-05, "loss": 0.0492, "step": 30800 }, { "grad_norm": 0.4269440472126007, "learning_rate": 8.286793161883388e-05, "loss": 0.0483, "step": 30810 }, { "grad_norm": 0.4769890606403351, "learning_rate": 8.285546963898954e-05, "loss": 0.0447, "step": 30820 }, { "grad_norm": 0.34296026825904846, "learning_rate": 8.284300406612044e-05, "loss": 0.0475, "step": 30830 }, { "grad_norm": 0.5513113737106323, "learning_rate": 8.283053490158978e-05, "loss": 0.0431, "step": 30840 }, { "grad_norm": 0.39561721682548523, "learning_rate": 8.28180621467612e-05, "loss": 0.0421, "step": 30850 }, { "grad_norm": 0.3047952651977539, "learning_rate": 8.280558580299867e-05, "loss": 0.0413, "step": 30860 }, { "grad_norm": 0.6015605926513672, "learning_rate": 8.279310587166661e-05, "loss": 0.0457, "step": 30870 }, { "grad_norm": 0.39665600657463074, "learning_rate": 8.278062235412978e-05, "loss": 0.0409, "step": 30880 }, { "grad_norm": 0.5752909779548645, "learning_rate": 8.276813525175339e-05, "loss": 0.0506, "step": 30890 }, { "grad_norm": 0.2780332863330841, "learning_rate": 8.2755644565903e-05, "loss": 0.0447, "step": 30900 }, { "grad_norm": 0.5159726738929749, "learning_rate": 8.274315029794454e-05, "loss": 0.0504, "step": 30910 }, { "grad_norm": 0.4079306423664093, "learning_rate": 8.273065244924443e-05, "loss": 0.0408, "step": 30920 }, { "grad_norm": 0.3499327003955841, "learning_rate": 8.271815102116936e-05, "loss": 0.0448, "step": 30930 }, { "grad_norm": 0.4497828185558319, "learning_rate": 8.270564601508648e-05, "loss": 0.0478, "step": 30940 }, { "grad_norm": 0.712355375289917, "learning_rate": 8.269313743236333e-05, "loss": 0.0487, "step": 30950 }, { "grad_norm": 0.2660602033138275, "learning_rate": 8.268062527436783e-05, "loss": 0.0445, "step": 30960 }, { "grad_norm": 0.2634340822696686, "learning_rate": 8.266810954246828e-05, "loss": 0.0536, "step": 30970 }, { "grad_norm": 0.43631139397621155, "learning_rate": 8.265559023803338e-05, "loss": 0.0528, "step": 30980 }, { "grad_norm": 0.3355555236339569, "learning_rate": 8.264306736243223e-05, "loss": 0.0501, "step": 30990 }, { "grad_norm": 0.6181844472885132, "learning_rate": 8.263054091703432e-05, "loss": 0.0481, "step": 31000 }, { "grad_norm": 0.3798772096633911, "learning_rate": 8.26180109032095e-05, "loss": 0.0523, "step": 31010 }, { "grad_norm": 0.4901673495769501, "learning_rate": 8.260547732232807e-05, "loss": 0.0548, "step": 31020 }, { "grad_norm": 0.4029640555381775, "learning_rate": 8.259294017576063e-05, "loss": 0.0451, "step": 31030 }, { "grad_norm": 0.44648829102516174, "learning_rate": 8.258039946487828e-05, "loss": 0.0507, "step": 31040 }, { "grad_norm": 0.44267621636390686, "learning_rate": 8.256785519105241e-05, "loss": 0.0472, "step": 31050 }, { "grad_norm": 0.35385504364967346, "learning_rate": 8.255530735565488e-05, "loss": 0.0524, "step": 31060 }, { "grad_norm": 0.49178096652030945, "learning_rate": 8.254275596005787e-05, "loss": 0.0474, "step": 31070 }, { "grad_norm": 0.36309632658958435, "learning_rate": 8.2530201005634e-05, "loss": 0.0426, "step": 31080 }, { "grad_norm": 0.5867704153060913, "learning_rate": 8.251764249375626e-05, "loss": 0.0515, "step": 31090 }, { "grad_norm": 0.3930840790271759, "learning_rate": 8.250508042579803e-05, "loss": 0.0484, "step": 31100 }, { "grad_norm": 0.3300028145313263, "learning_rate": 8.249251480313307e-05, "loss": 0.046, "step": 31110 }, { "grad_norm": 0.3052823543548584, "learning_rate": 8.247994562713555e-05, "loss": 0.0447, "step": 31120 }, { "grad_norm": 0.3674508333206177, "learning_rate": 8.246737289917998e-05, "loss": 0.0446, "step": 31130 }, { "grad_norm": 0.3538217842578888, "learning_rate": 8.245479662064135e-05, "loss": 0.0456, "step": 31140 }, { "grad_norm": 0.2660070061683655, "learning_rate": 8.244221679289496e-05, "loss": 0.0387, "step": 31150 }, { "grad_norm": 0.37351202964782715, "learning_rate": 8.242963341731652e-05, "loss": 0.0441, "step": 31160 }, { "grad_norm": 0.28318700194358826, "learning_rate": 8.24170464952821e-05, "loss": 0.0439, "step": 31170 }, { "grad_norm": 0.30213460326194763, "learning_rate": 8.240445602816824e-05, "loss": 0.0441, "step": 31180 }, { "grad_norm": 0.35742634534835815, "learning_rate": 8.239186201735179e-05, "loss": 0.0494, "step": 31190 }, { "grad_norm": 0.46169155836105347, "learning_rate": 8.237926446420998e-05, "loss": 0.0433, "step": 31200 }, { "grad_norm": 0.28386378288269043, "learning_rate": 8.236666337012052e-05, "loss": 0.0449, "step": 31210 }, { "grad_norm": 0.5304089784622192, "learning_rate": 8.23540587364614e-05, "loss": 0.0442, "step": 31220 }, { "grad_norm": 0.3575597107410431, "learning_rate": 8.234145056461107e-05, "loss": 0.0485, "step": 31230 }, { "grad_norm": 0.38742953538894653, "learning_rate": 8.232883885594832e-05, "loss": 0.0425, "step": 31240 }, { "grad_norm": 0.3731825053691864, "learning_rate": 8.231622361185236e-05, "loss": 0.0399, "step": 31250 }, { "grad_norm": 0.4780954122543335, "learning_rate": 8.230360483370278e-05, "loss": 0.048, "step": 31260 }, { "grad_norm": 0.35519441962242126, "learning_rate": 8.229098252287953e-05, "loss": 0.0457, "step": 31270 }, { "grad_norm": 0.9519651532173157, "learning_rate": 8.2278356680763e-05, "loss": 0.0462, "step": 31280 }, { "grad_norm": 0.4354012906551361, "learning_rate": 8.22657273087339e-05, "loss": 0.0457, "step": 31290 }, { "grad_norm": 0.4279547929763794, "learning_rate": 8.225309440817336e-05, "loss": 0.055, "step": 31300 }, { "grad_norm": 0.4895695745944977, "learning_rate": 8.224045798046293e-05, "loss": 0.0473, "step": 31310 }, { "grad_norm": 0.3706169128417969, "learning_rate": 8.22278180269845e-05, "loss": 0.0447, "step": 31320 }, { "grad_norm": 0.4431590735912323, "learning_rate": 8.221517454912031e-05, "loss": 0.0428, "step": 31330 }, { "grad_norm": 0.33743828535079956, "learning_rate": 8.220252754825308e-05, "loss": 0.0472, "step": 31340 }, { "grad_norm": 0.4490987956523895, "learning_rate": 8.218987702576586e-05, "loss": 0.0426, "step": 31350 }, { "grad_norm": 0.3039839565753937, "learning_rate": 8.217722298304207e-05, "loss": 0.05, "step": 31360 }, { "grad_norm": 0.38837188482284546, "learning_rate": 8.216456542146557e-05, "loss": 0.0417, "step": 31370 }, { "grad_norm": 0.34201738238334656, "learning_rate": 8.215190434242055e-05, "loss": 0.0483, "step": 31380 }, { "grad_norm": 0.4155837595462799, "learning_rate": 8.213923974729161e-05, "loss": 0.0477, "step": 31390 }, { "grad_norm": 0.39978447556495667, "learning_rate": 8.212657163746373e-05, "loss": 0.0422, "step": 31400 }, { "grad_norm": 0.34894728660583496, "learning_rate": 8.211390001432227e-05, "loss": 0.0444, "step": 31410 }, { "grad_norm": 0.30860793590545654, "learning_rate": 8.210122487925297e-05, "loss": 0.0417, "step": 31420 }, { "grad_norm": 0.46451684832572937, "learning_rate": 8.208854623364202e-05, "loss": 0.0461, "step": 31430 }, { "grad_norm": 0.5203738808631897, "learning_rate": 8.207586407887589e-05, "loss": 0.0452, "step": 31440 }, { "grad_norm": 0.44275468587875366, "learning_rate": 8.206317841634148e-05, "loss": 0.0433, "step": 31450 }, { "grad_norm": 0.379738450050354, "learning_rate": 8.205048924742609e-05, "loss": 0.0537, "step": 31460 }, { "grad_norm": 0.3649391233921051, "learning_rate": 8.203779657351738e-05, "loss": 0.0452, "step": 31470 }, { "grad_norm": 0.3552153706550598, "learning_rate": 8.20251003960034e-05, "loss": 0.0426, "step": 31480 }, { "grad_norm": 0.3966030478477478, "learning_rate": 8.201240071627258e-05, "loss": 0.0496, "step": 31490 }, { "grad_norm": 0.49885839223861694, "learning_rate": 8.199969753571377e-05, "loss": 0.0459, "step": 31500 }, { "grad_norm": 0.7712454795837402, "learning_rate": 8.198699085571615e-05, "loss": 0.041, "step": 31510 }, { "grad_norm": 1.660168170928955, "learning_rate": 8.197428067766928e-05, "loss": 0.0446, "step": 31520 }, { "grad_norm": 0.4046449065208435, "learning_rate": 8.196156700296316e-05, "loss": 0.0458, "step": 31530 }, { "grad_norm": 0.6109291911125183, "learning_rate": 8.194884983298814e-05, "loss": 0.0527, "step": 31540 }, { "grad_norm": 0.6250191926956177, "learning_rate": 8.193612916913491e-05, "loss": 0.047, "step": 31550 }, { "grad_norm": 0.4024165868759155, "learning_rate": 8.192340501279463e-05, "loss": 0.0536, "step": 31560 }, { "grad_norm": 0.2746507525444031, "learning_rate": 8.191067736535876e-05, "loss": 0.0441, "step": 31570 }, { "grad_norm": 0.3841416537761688, "learning_rate": 8.18979462282192e-05, "loss": 0.049, "step": 31580 }, { "grad_norm": 0.3823961913585663, "learning_rate": 8.188521160276819e-05, "loss": 0.0472, "step": 31590 }, { "grad_norm": 0.44736990332603455, "learning_rate": 8.187247349039837e-05, "loss": 0.0413, "step": 31600 }, { "grad_norm": 0.43079566955566406, "learning_rate": 8.185973189250278e-05, "loss": 0.0479, "step": 31610 }, { "grad_norm": 0.45415034890174866, "learning_rate": 8.184698681047482e-05, "loss": 0.0489, "step": 31620 }, { "grad_norm": 0.5113688707351685, "learning_rate": 8.183423824570827e-05, "loss": 0.0511, "step": 31630 }, { "grad_norm": 0.3292008638381958, "learning_rate": 8.182148619959725e-05, "loss": 0.0412, "step": 31640 }, { "grad_norm": 0.4482325315475464, "learning_rate": 8.180873067353636e-05, "loss": 0.0494, "step": 31650 }, { "grad_norm": 0.7049188613891602, "learning_rate": 8.179597166892052e-05, "loss": 0.0422, "step": 31660 }, { "grad_norm": 0.39693647623062134, "learning_rate": 8.178320918714501e-05, "loss": 0.0523, "step": 31670 }, { "grad_norm": 0.44698256254196167, "learning_rate": 8.177044322960554e-05, "loss": 0.05, "step": 31680 }, { "grad_norm": 0.43730175495147705, "learning_rate": 8.175767379769816e-05, "loss": 0.05, "step": 31690 }, { "grad_norm": 0.2770303189754486, "learning_rate": 8.174490089281932e-05, "loss": 0.0494, "step": 31700 }, { "grad_norm": 0.3852122128009796, "learning_rate": 8.173212451636584e-05, "loss": 0.0493, "step": 31710 }, { "grad_norm": 0.46329012513160706, "learning_rate": 8.171934466973493e-05, "loss": 0.0546, "step": 31720 }, { "grad_norm": 0.3024803698062897, "learning_rate": 8.170656135432418e-05, "loss": 0.0468, "step": 31730 }, { "grad_norm": 0.4567078948020935, "learning_rate": 8.169377457153155e-05, "loss": 0.0451, "step": 31740 }, { "grad_norm": 0.3071468472480774, "learning_rate": 8.168098432275539e-05, "loss": 0.0471, "step": 31750 }, { "grad_norm": 0.41292575001716614, "learning_rate": 8.166819060939442e-05, "loss": 0.0476, "step": 31760 }, { "grad_norm": 0.41777193546295166, "learning_rate": 8.165539343284772e-05, "loss": 0.0481, "step": 31770 }, { "grad_norm": 0.48769915103912354, "learning_rate": 8.16425927945148e-05, "loss": 0.0398, "step": 31780 }, { "grad_norm": 0.32181820273399353, "learning_rate": 8.162978869579551e-05, "loss": 0.0446, "step": 31790 }, { "grad_norm": 0.5708882212638855, "learning_rate": 8.161698113809007e-05, "loss": 0.0553, "step": 31800 }, { "grad_norm": 1.8808013200759888, "learning_rate": 8.160417012279911e-05, "loss": 0.0534, "step": 31810 }, { "grad_norm": 0.41876792907714844, "learning_rate": 8.159135565132363e-05, "loss": 0.0465, "step": 31820 }, { "grad_norm": 0.35378697514533997, "learning_rate": 8.157853772506498e-05, "loss": 0.0482, "step": 31830 }, { "grad_norm": 0.3540392518043518, "learning_rate": 8.156571634542494e-05, "loss": 0.0498, "step": 31840 }, { "grad_norm": 0.3803573548793793, "learning_rate": 8.15528915138056e-05, "loss": 0.0465, "step": 31850 }, { "grad_norm": 0.31854692101478577, "learning_rate": 8.154006323160949e-05, "loss": 0.0469, "step": 31860 }, { "grad_norm": 0.43251296877861023, "learning_rate": 8.152723150023949e-05, "loss": 0.0461, "step": 31870 }, { "grad_norm": 0.4369013011455536, "learning_rate": 8.151439632109886e-05, "loss": 0.0413, "step": 31880 }, { "grad_norm": 0.47837117314338684, "learning_rate": 8.150155769559122e-05, "loss": 0.0465, "step": 31890 }, { "grad_norm": 0.42528894543647766, "learning_rate": 8.148871562512058e-05, "loss": 0.05, "step": 31900 }, { "grad_norm": 0.31880488991737366, "learning_rate": 8.147587011109136e-05, "loss": 0.0474, "step": 31910 }, { "grad_norm": 0.36111563444137573, "learning_rate": 8.14630211549083e-05, "loss": 0.0427, "step": 31920 }, { "grad_norm": 0.37655168771743774, "learning_rate": 8.145016875797655e-05, "loss": 0.0447, "step": 31930 }, { "grad_norm": 0.5717474222183228, "learning_rate": 8.143731292170164e-05, "loss": 0.0484, "step": 31940 }, { "grad_norm": 0.31133103370666504, "learning_rate": 8.142445364748944e-05, "loss": 0.0472, "step": 31950 }, { "grad_norm": 0.30309054255485535, "learning_rate": 8.141159093674624e-05, "loss": 0.0454, "step": 31960 }, { "grad_norm": 0.32967159152030945, "learning_rate": 8.139872479087869e-05, "loss": 0.0434, "step": 31970 }, { "grad_norm": 0.3869982361793518, "learning_rate": 8.13858552112938e-05, "loss": 0.0438, "step": 31980 }, { "grad_norm": 0.5878490805625916, "learning_rate": 8.137298219939895e-05, "loss": 0.0524, "step": 31990 }, { "grad_norm": 0.5141135454177856, "learning_rate": 8.136010575660196e-05, "loss": 0.0514, "step": 32000 }, { "grad_norm": 0.4033222496509552, "learning_rate": 8.134722588431095e-05, "loss": 0.0459, "step": 32010 }, { "grad_norm": 0.3884024918079376, "learning_rate": 8.133434258393444e-05, "loss": 0.049, "step": 32020 }, { "grad_norm": 0.42109230160713196, "learning_rate": 8.132145585688134e-05, "loss": 0.052, "step": 32030 }, { "grad_norm": 0.3054525852203369, "learning_rate": 8.130856570456093e-05, "loss": 0.0524, "step": 32040 }, { "grad_norm": 0.4406728744506836, "learning_rate": 8.129567212838283e-05, "loss": 0.0477, "step": 32050 }, { "grad_norm": 0.32026395201683044, "learning_rate": 8.128277512975708e-05, "loss": 0.0438, "step": 32060 }, { "grad_norm": 0.3918361961841583, "learning_rate": 8.126987471009408e-05, "loss": 0.0448, "step": 32070 }, { "grad_norm": 0.37206369638442993, "learning_rate": 8.125697087080459e-05, "loss": 0.0457, "step": 32080 }, { "grad_norm": 0.4051385223865509, "learning_rate": 8.124406361329976e-05, "loss": 0.0519, "step": 32090 }, { "grad_norm": 0.5613278746604919, "learning_rate": 8.123115293899111e-05, "loss": 0.0483, "step": 32100 }, { "grad_norm": 0.4213239252567291, "learning_rate": 8.121823884929055e-05, "loss": 0.044, "step": 32110 }, { "grad_norm": 0.35374173521995544, "learning_rate": 8.12053213456103e-05, "loss": 0.051, "step": 32120 }, { "grad_norm": 0.8223071694374084, "learning_rate": 8.119240042936303e-05, "loss": 0.0439, "step": 32130 }, { "grad_norm": 0.37935665249824524, "learning_rate": 8.117947610196175e-05, "loss": 0.0413, "step": 32140 }, { "grad_norm": 0.3882150948047638, "learning_rate": 8.116654836481982e-05, "loss": 0.0456, "step": 32150 }, { "grad_norm": 0.410501092672348, "learning_rate": 8.115361721935106e-05, "loss": 0.0515, "step": 32160 }, { "grad_norm": 0.4297980070114136, "learning_rate": 8.114068266696953e-05, "loss": 0.0393, "step": 32170 }, { "grad_norm": 0.47208547592163086, "learning_rate": 8.112774470908978e-05, "loss": 0.0487, "step": 32180 }, { "grad_norm": 0.4973609447479248, "learning_rate": 8.111480334712665e-05, "loss": 0.0526, "step": 32190 }, { "grad_norm": 0.5441886782646179, "learning_rate": 8.110185858249542e-05, "loss": 0.0497, "step": 32200 }, { "grad_norm": 0.5971451997756958, "learning_rate": 8.108891041661168e-05, "loss": 0.0519, "step": 32210 }, { "grad_norm": 0.31968510150909424, "learning_rate": 8.107595885089146e-05, "loss": 0.0468, "step": 32220 }, { "grad_norm": 0.4573981761932373, "learning_rate": 8.106300388675108e-05, "loss": 0.0518, "step": 32230 }, { "grad_norm": 0.35175541043281555, "learning_rate": 8.105004552560731e-05, "loss": 0.0483, "step": 32240 }, { "grad_norm": 0.5252546072006226, "learning_rate": 8.103708376887724e-05, "loss": 0.0509, "step": 32250 }, { "grad_norm": 0.30220919847488403, "learning_rate": 8.102411861797836e-05, "loss": 0.0422, "step": 32260 }, { "grad_norm": 0.53425532579422, "learning_rate": 8.101115007432851e-05, "loss": 0.0439, "step": 32270 }, { "grad_norm": 0.45624566078186035, "learning_rate": 8.09981781393459e-05, "loss": 0.0385, "step": 32280 }, { "grad_norm": 0.2986956238746643, "learning_rate": 8.098520281444915e-05, "loss": 0.0499, "step": 32290 }, { "grad_norm": 0.5362474918365479, "learning_rate": 8.09722241010572e-05, "loss": 0.0437, "step": 32300 }, { "grad_norm": 0.958164632320404, "learning_rate": 8.095924200058939e-05, "loss": 0.0438, "step": 32310 }, { "grad_norm": 0.5580820441246033, "learning_rate": 8.094625651446541e-05, "loss": 0.0492, "step": 32320 }, { "grad_norm": 0.49971210956573486, "learning_rate": 8.093326764410536e-05, "loss": 0.0455, "step": 32330 }, { "grad_norm": 0.47218531370162964, "learning_rate": 8.092027539092966e-05, "loss": 0.042, "step": 32340 }, { "grad_norm": 0.3191186785697937, "learning_rate": 8.090727975635913e-05, "loss": 0.0518, "step": 32350 }, { "grad_norm": 0.4244464039802551, "learning_rate": 8.089428074181497e-05, "loss": 0.0463, "step": 32360 }, { "grad_norm": 0.3126187324523926, "learning_rate": 8.088127834871871e-05, "loss": 0.0459, "step": 32370 }, { "grad_norm": 0.4608692526817322, "learning_rate": 8.086827257849226e-05, "loss": 0.0425, "step": 32380 }, { "grad_norm": 0.348264217376709, "learning_rate": 8.085526343255795e-05, "loss": 0.043, "step": 32390 }, { "grad_norm": 0.39451467990875244, "learning_rate": 8.084225091233842e-05, "loss": 0.0431, "step": 32400 }, { "grad_norm": 0.39907917380332947, "learning_rate": 8.082923501925668e-05, "loss": 0.0529, "step": 32410 }, { "grad_norm": 0.33377474546432495, "learning_rate": 8.081621575473617e-05, "loss": 0.0501, "step": 32420 }, { "grad_norm": 0.36680933833122253, "learning_rate": 8.080319312020064e-05, "loss": 0.0461, "step": 32430 }, { "grad_norm": 0.37327373027801514, "learning_rate": 8.079016711707421e-05, "loss": 0.044, "step": 32440 }, { "grad_norm": 0.4037942886352539, "learning_rate": 8.077713774678139e-05, "loss": 0.0418, "step": 32450 }, { "grad_norm": 0.5540410876274109, "learning_rate": 8.076410501074707e-05, "loss": 0.0458, "step": 32460 }, { "grad_norm": 0.45347103476524353, "learning_rate": 8.075106891039647e-05, "loss": 0.0476, "step": 32470 }, { "grad_norm": 0.34677207469940186, "learning_rate": 8.073802944715523e-05, "loss": 0.0447, "step": 32480 }, { "grad_norm": 0.6196181774139404, "learning_rate": 8.072498662244929e-05, "loss": 0.0444, "step": 32490 }, { "grad_norm": 0.5481371879577637, "learning_rate": 8.0711940437705e-05, "loss": 0.049, "step": 32500 }, { "grad_norm": 0.4021889865398407, "learning_rate": 8.06988908943491e-05, "loss": 0.0471, "step": 32510 }, { "grad_norm": 0.3964548408985138, "learning_rate": 8.068583799380863e-05, "loss": 0.0554, "step": 32520 }, { "grad_norm": 0.5109497308731079, "learning_rate": 8.067278173751104e-05, "loss": 0.0551, "step": 32530 }, { "grad_norm": 0.46815887093544006, "learning_rate": 8.065972212688417e-05, "loss": 0.0501, "step": 32540 }, { "grad_norm": 0.4600187838077545, "learning_rate": 8.064665916335618e-05, "loss": 0.0451, "step": 32550 }, { "grad_norm": 0.36821219325065613, "learning_rate": 8.063359284835564e-05, "loss": 0.0508, "step": 32560 }, { "grad_norm": 0.39106854796409607, "learning_rate": 8.062052318331142e-05, "loss": 0.0442, "step": 32570 }, { "grad_norm": 0.43896356225013733, "learning_rate": 8.060745016965283e-05, "loss": 0.0418, "step": 32580 }, { "grad_norm": 0.3466191589832306, "learning_rate": 8.059437380880952e-05, "loss": 0.0431, "step": 32590 }, { "grad_norm": 0.41047999262809753, "learning_rate": 8.058129410221146e-05, "loss": 0.0459, "step": 32600 }, { "grad_norm": 0.2552751898765564, "learning_rate": 8.056821105128908e-05, "loss": 0.0452, "step": 32610 }, { "grad_norm": 0.43830713629722595, "learning_rate": 8.05551246574731e-05, "loss": 0.0444, "step": 32620 }, { "grad_norm": 0.35081908106803894, "learning_rate": 8.05420349221946e-05, "loss": 0.0421, "step": 32630 }, { "grad_norm": 0.3742074966430664, "learning_rate": 8.05289418468851e-05, "loss": 0.0431, "step": 32640 }, { "grad_norm": 0.3593680262565613, "learning_rate": 8.051584543297642e-05, "loss": 0.0439, "step": 32650 }, { "grad_norm": 0.5306724309921265, "learning_rate": 8.050274568190074e-05, "loss": 0.0442, "step": 32660 }, { "grad_norm": 0.4088992774486542, "learning_rate": 8.048964259509067e-05, "loss": 0.044, "step": 32670 }, { "grad_norm": 0.37552082538604736, "learning_rate": 8.047653617397914e-05, "loss": 0.0439, "step": 32680 }, { "grad_norm": 0.32593682408332825, "learning_rate": 8.046342641999941e-05, "loss": 0.0474, "step": 32690 }, { "grad_norm": 0.42575833201408386, "learning_rate": 8.045031333458517e-05, "loss": 0.0477, "step": 32700 }, { "grad_norm": 0.5172208547592163, "learning_rate": 8.043719691917047e-05, "loss": 0.0509, "step": 32710 }, { "grad_norm": 0.4000219404697418, "learning_rate": 8.042407717518966e-05, "loss": 0.0473, "step": 32720 }, { "grad_norm": 0.4419083893299103, "learning_rate": 8.041095410407751e-05, "loss": 0.0405, "step": 32730 }, { "grad_norm": 0.44411057233810425, "learning_rate": 8.039782770726913e-05, "loss": 0.0426, "step": 32740 }, { "grad_norm": 0.37609174847602844, "learning_rate": 8.038469798620004e-05, "loss": 0.0442, "step": 32750 }, { "grad_norm": 0.42664897441864014, "learning_rate": 8.037156494230604e-05, "loss": 0.0458, "step": 32760 }, { "grad_norm": 0.4818483591079712, "learning_rate": 8.035842857702338e-05, "loss": 0.0436, "step": 32770 }, { "grad_norm": 0.462332159280777, "learning_rate": 8.03452888917886e-05, "loss": 0.0458, "step": 32780 }, { "grad_norm": 0.44636401534080505, "learning_rate": 8.033214588803866e-05, "loss": 0.0505, "step": 32790 }, { "grad_norm": 0.4341273903846741, "learning_rate": 8.031899956721083e-05, "loss": 0.0512, "step": 32800 }, { "grad_norm": 0.3629356622695923, "learning_rate": 8.030584993074282e-05, "loss": 0.0457, "step": 32810 }, { "grad_norm": 1.1203701496124268, "learning_rate": 8.02926969800726e-05, "loss": 0.0501, "step": 32820 }, { "grad_norm": 0.31597793102264404, "learning_rate": 8.027954071663859e-05, "loss": 0.0498, "step": 32830 }, { "grad_norm": 0.46724238991737366, "learning_rate": 8.026638114187954e-05, "loss": 0.0513, "step": 32840 }, { "grad_norm": 0.377714067697525, "learning_rate": 8.025321825723456e-05, "loss": 0.0462, "step": 32850 }, { "grad_norm": 0.36446836590766907, "learning_rate": 8.02400520641431e-05, "loss": 0.0433, "step": 32860 }, { "grad_norm": 0.31011930108070374, "learning_rate": 8.022688256404501e-05, "loss": 0.0497, "step": 32870 }, { "grad_norm": 0.9996450543403625, "learning_rate": 8.02137097583805e-05, "loss": 0.0455, "step": 32880 }, { "grad_norm": 0.36790573596954346, "learning_rate": 8.02005336485901e-05, "loss": 0.0508, "step": 32890 }, { "grad_norm": 0.30146634578704834, "learning_rate": 8.018735423611476e-05, "loss": 0.0481, "step": 32900 }, { "grad_norm": 0.351933091878891, "learning_rate": 8.017417152239574e-05, "loss": 0.0415, "step": 32910 }, { "grad_norm": 0.4180256426334381, "learning_rate": 8.01609855088747e-05, "loss": 0.0509, "step": 32920 }, { "grad_norm": 0.4288049042224884, "learning_rate": 8.014779619699362e-05, "loss": 0.045, "step": 32930 }, { "grad_norm": 0.4422962963581085, "learning_rate": 8.013460358819489e-05, "loss": 0.0388, "step": 32940 }, { "grad_norm": 0.3681096136569977, "learning_rate": 8.01214076839212e-05, "loss": 0.0526, "step": 32950 }, { "grad_norm": 0.500246524810791, "learning_rate": 8.010820848561565e-05, "loss": 0.045, "step": 32960 }, { "grad_norm": 0.5432372689247131, "learning_rate": 8.009500599472171e-05, "loss": 0.0475, "step": 32970 }, { "grad_norm": 0.26809629797935486, "learning_rate": 8.008180021268314e-05, "loss": 0.0414, "step": 32980 }, { "grad_norm": 0.4436333477497101, "learning_rate": 8.006859114094414e-05, "loss": 0.0446, "step": 32990 }, { "grad_norm": 0.3318993151187897, "learning_rate": 8.005537878094921e-05, "loss": 0.0425, "step": 33000 }, { "grad_norm": 0.4349215626716614, "learning_rate": 8.004216313414323e-05, "loss": 0.0381, "step": 33010 }, { "grad_norm": 0.4401482045650482, "learning_rate": 8.002894420197149e-05, "loss": 0.042, "step": 33020 }, { "grad_norm": 0.3450430929660797, "learning_rate": 8.001572198587954e-05, "loss": 0.044, "step": 33030 }, { "grad_norm": 0.36682161688804626, "learning_rate": 8.000249648731338e-05, "loss": 0.0534, "step": 33040 }, { "grad_norm": 0.40978121757507324, "learning_rate": 7.998926770771928e-05, "loss": 0.0512, "step": 33050 }, { "grad_norm": 0.3783752918243408, "learning_rate": 7.997603564854397e-05, "loss": 0.0523, "step": 33060 }, { "grad_norm": 0.43425753712654114, "learning_rate": 7.996280031123448e-05, "loss": 0.0398, "step": 33070 }, { "grad_norm": 0.40625736117362976, "learning_rate": 7.994956169723818e-05, "loss": 0.0491, "step": 33080 }, { "grad_norm": 0.41989776492118835, "learning_rate": 7.993631980800285e-05, "loss": 0.0475, "step": 33090 }, { "grad_norm": 0.35207051038742065, "learning_rate": 7.992307464497659e-05, "loss": 0.044, "step": 33100 }, { "grad_norm": 0.30281493067741394, "learning_rate": 7.990982620960787e-05, "loss": 0.0425, "step": 33110 }, { "grad_norm": 0.4853131175041199, "learning_rate": 7.989657450334554e-05, "loss": 0.0466, "step": 33120 }, { "grad_norm": 0.3458344042301178, "learning_rate": 7.988331952763877e-05, "loss": 0.0486, "step": 33130 }, { "grad_norm": 0.29471802711486816, "learning_rate": 7.987006128393709e-05, "loss": 0.0447, "step": 33140 }, { "grad_norm": 0.29764479398727417, "learning_rate": 7.985679977369043e-05, "loss": 0.049, "step": 33150 }, { "grad_norm": 0.693352460861206, "learning_rate": 7.984353499834902e-05, "loss": 0.0412, "step": 33160 }, { "grad_norm": 0.3268052935600281, "learning_rate": 7.983026695936351e-05, "loss": 0.0434, "step": 33170 }, { "grad_norm": 0.3679386079311371, "learning_rate": 7.981699565818486e-05, "loss": 0.0435, "step": 33180 }, { "grad_norm": 0.32079094648361206, "learning_rate": 7.980372109626437e-05, "loss": 0.0453, "step": 33190 }, { "grad_norm": 0.5350250601768494, "learning_rate": 7.979044327505375e-05, "loss": 0.0432, "step": 33200 }, { "grad_norm": 0.42068973183631897, "learning_rate": 7.977716219600506e-05, "loss": 0.0446, "step": 33210 }, { "grad_norm": 0.2584574222564697, "learning_rate": 7.97638778605707e-05, "loss": 0.0385, "step": 33220 }, { "grad_norm": 0.5230417251586914, "learning_rate": 7.975059027020338e-05, "loss": 0.0365, "step": 33230 }, { "grad_norm": 0.35039302706718445, "learning_rate": 7.973729942635623e-05, "loss": 0.0489, "step": 33240 }, { "grad_norm": 0.44211262464523315, "learning_rate": 7.972400533048273e-05, "loss": 0.0518, "step": 33250 }, { "grad_norm": 0.42865195870399475, "learning_rate": 7.97107079840367e-05, "loss": 0.0511, "step": 33260 }, { "grad_norm": 0.44491147994995117, "learning_rate": 7.969740738847231e-05, "loss": 0.0474, "step": 33270 }, { "grad_norm": 0.2979136109352112, "learning_rate": 7.968410354524411e-05, "loss": 0.05, "step": 33280 }, { "grad_norm": 0.4681205153465271, "learning_rate": 7.967079645580697e-05, "loss": 0.0498, "step": 33290 }, { "grad_norm": 0.37692224979400635, "learning_rate": 7.965748612161612e-05, "loss": 0.0512, "step": 33300 }, { "grad_norm": 0.4650694727897644, "learning_rate": 7.96441725441272e-05, "loss": 0.0505, "step": 33310 }, { "grad_norm": 0.288491815328598, "learning_rate": 7.963085572479614e-05, "loss": 0.0398, "step": 33320 }, { "grad_norm": 0.37076810002326965, "learning_rate": 7.961753566507924e-05, "loss": 0.0429, "step": 33330 }, { "grad_norm": 0.35695880651474, "learning_rate": 7.960421236643316e-05, "loss": 0.0483, "step": 33340 }, { "grad_norm": 0.46429452300071716, "learning_rate": 7.959088583031496e-05, "loss": 0.0443, "step": 33350 }, { "grad_norm": 0.434902548789978, "learning_rate": 7.957755605818194e-05, "loss": 0.0502, "step": 33360 }, { "grad_norm": 0.27939844131469727, "learning_rate": 7.956422305149185e-05, "loss": 0.0391, "step": 33370 }, { "grad_norm": 0.7558864951133728, "learning_rate": 7.95508868117028e-05, "loss": 0.05, "step": 33380 }, { "grad_norm": 0.6242010593414307, "learning_rate": 7.953754734027318e-05, "loss": 0.0436, "step": 33390 }, { "grad_norm": 0.4637543261051178, "learning_rate": 7.952420463866182e-05, "loss": 0.0461, "step": 33400 }, { "grad_norm": 0.4334569275379181, "learning_rate": 7.951085870832782e-05, "loss": 0.0449, "step": 33410 }, { "grad_norm": 0.29855963587760925, "learning_rate": 7.949750955073067e-05, "loss": 0.048, "step": 33420 }, { "grad_norm": 0.5487177968025208, "learning_rate": 7.948415716733022e-05, "loss": 0.0497, "step": 33430 }, { "grad_norm": 0.34562045335769653, "learning_rate": 7.947080155958669e-05, "loss": 0.042, "step": 33440 }, { "grad_norm": 0.3826068341732025, "learning_rate": 7.94574427289606e-05, "loss": 0.046, "step": 33450 }, { "grad_norm": 0.916257917881012, "learning_rate": 7.944408067691284e-05, "loss": 0.0446, "step": 33460 }, { "grad_norm": 0.4274612367153168, "learning_rate": 7.943071540490473e-05, "loss": 0.048, "step": 33470 }, { "grad_norm": 0.5154330134391785, "learning_rate": 7.94173469143978e-05, "loss": 0.0471, "step": 33480 }, { "grad_norm": 0.50360506772995, "learning_rate": 7.940397520685406e-05, "loss": 0.0478, "step": 33490 }, { "grad_norm": 0.5601580142974854, "learning_rate": 7.939060028373577e-05, "loss": 0.047, "step": 33500 }, { "grad_norm": 0.45704320073127747, "learning_rate": 7.937722214650565e-05, "loss": 0.0516, "step": 33510 }, { "grad_norm": 0.40876877307891846, "learning_rate": 7.936384079662666e-05, "loss": 0.0431, "step": 33520 }, { "grad_norm": 0.3215845227241516, "learning_rate": 7.93504562355622e-05, "loss": 0.0512, "step": 33530 }, { "grad_norm": 0.40257489681243896, "learning_rate": 7.933706846477599e-05, "loss": 0.0455, "step": 33540 }, { "grad_norm": 0.4313448965549469, "learning_rate": 7.932367748573206e-05, "loss": 0.047, "step": 33550 }, { "grad_norm": 0.3468565046787262, "learning_rate": 7.931028329989485e-05, "loss": 0.0482, "step": 33560 }, { "grad_norm": 0.36317041516304016, "learning_rate": 7.929688590872913e-05, "loss": 0.0459, "step": 33570 }, { "grad_norm": 0.4897274971008301, "learning_rate": 7.928348531370003e-05, "loss": 0.0499, "step": 33580 }, { "grad_norm": 0.48805078864097595, "learning_rate": 7.927008151627297e-05, "loss": 0.0419, "step": 33590 }, { "grad_norm": 0.40950319170951843, "learning_rate": 7.925667451791383e-05, "loss": 0.0476, "step": 33600 }, { "grad_norm": 0.3760738968849182, "learning_rate": 7.924326432008874e-05, "loss": 0.0397, "step": 33610 }, { "grad_norm": 0.6842674016952515, "learning_rate": 7.922985092426422e-05, "loss": 0.0422, "step": 33620 }, { "grad_norm": 0.3155037760734558, "learning_rate": 7.921643433190717e-05, "loss": 0.0442, "step": 33630 }, { "grad_norm": 0.3080408573150635, "learning_rate": 7.920301454448478e-05, "loss": 0.0404, "step": 33640 }, { "grad_norm": 0.39255475997924805, "learning_rate": 7.918959156346461e-05, "loss": 0.0472, "step": 33650 }, { "grad_norm": 0.46951669454574585, "learning_rate": 7.91761653903146e-05, "loss": 0.0514, "step": 33660 }, { "grad_norm": 0.5049958825111389, "learning_rate": 7.916273602650302e-05, "loss": 0.0466, "step": 33670 }, { "grad_norm": 0.5766080617904663, "learning_rate": 7.914930347349847e-05, "loss": 0.0466, "step": 33680 }, { "grad_norm": 0.37032079696655273, "learning_rate": 7.913586773276992e-05, "loss": 0.0471, "step": 33690 }, { "grad_norm": 0.43211206793785095, "learning_rate": 7.912242880578667e-05, "loss": 0.0501, "step": 33700 }, { "grad_norm": 0.4963133633136749, "learning_rate": 7.910898669401839e-05, "loss": 0.0446, "step": 33710 }, { "grad_norm": 0.5230227708816528, "learning_rate": 7.909554139893511e-05, "loss": 0.0423, "step": 33720 }, { "grad_norm": 0.5230244398117065, "learning_rate": 7.908209292200715e-05, "loss": 0.0397, "step": 33730 }, { "grad_norm": 0.5006048679351807, "learning_rate": 7.906864126470523e-05, "loss": 0.0464, "step": 33740 }, { "grad_norm": 0.36573418974876404, "learning_rate": 7.905518642850041e-05, "loss": 0.0431, "step": 33750 }, { "grad_norm": 0.4700218141078949, "learning_rate": 7.904172841486409e-05, "loss": 0.0428, "step": 33760 }, { "grad_norm": 0.49232786893844604, "learning_rate": 7.902826722526801e-05, "loss": 0.0504, "step": 33770 }, { "grad_norm": 0.26551586389541626, "learning_rate": 7.901480286118427e-05, "loss": 0.0371, "step": 33780 }, { "grad_norm": 0.3953196406364441, "learning_rate": 7.900133532408531e-05, "loss": 0.0421, "step": 33790 }, { "grad_norm": 0.4028092324733734, "learning_rate": 7.898786461544395e-05, "loss": 0.0467, "step": 33800 }, { "grad_norm": 0.41663655638694763, "learning_rate": 7.897439073673325e-05, "loss": 0.0395, "step": 33810 }, { "grad_norm": 0.45891132950782776, "learning_rate": 7.896091368942677e-05, "loss": 0.0492, "step": 33820 }, { "grad_norm": 0.280788779258728, "learning_rate": 7.894743347499832e-05, "loss": 0.0461, "step": 33830 }, { "grad_norm": 0.3828178644180298, "learning_rate": 7.893395009492203e-05, "loss": 0.045, "step": 33840 }, { "grad_norm": 0.4610137343406677, "learning_rate": 7.892046355067248e-05, "loss": 0.0457, "step": 33850 }, { "grad_norm": 0.38586392998695374, "learning_rate": 7.890697384372451e-05, "loss": 0.0413, "step": 33860 }, { "grad_norm": 0.3624717891216278, "learning_rate": 7.889348097555336e-05, "loss": 0.0411, "step": 33870 }, { "grad_norm": 0.26649731397628784, "learning_rate": 7.887998494763455e-05, "loss": 0.0483, "step": 33880 }, { "grad_norm": 0.3930051028728485, "learning_rate": 7.886648576144404e-05, "loss": 0.0436, "step": 33890 }, { "grad_norm": 0.4344770014286041, "learning_rate": 7.885298341845802e-05, "loss": 0.0461, "step": 33900 }, { "grad_norm": 0.3151707351207733, "learning_rate": 7.883947792015311e-05, "loss": 0.0436, "step": 33910 }, { "grad_norm": 0.2751193344593048, "learning_rate": 7.882596926800628e-05, "loss": 0.0542, "step": 33920 }, { "grad_norm": 0.4417133331298828, "learning_rate": 7.881245746349477e-05, "loss": 0.0501, "step": 33930 }, { "grad_norm": 0.3476828336715698, "learning_rate": 7.879894250809623e-05, "loss": 0.0405, "step": 33940 }, { "grad_norm": 0.40851831436157227, "learning_rate": 7.878542440328865e-05, "loss": 0.0451, "step": 33950 }, { "grad_norm": 0.5720603466033936, "learning_rate": 7.877190315055031e-05, "loss": 0.0443, "step": 33960 }, { "grad_norm": 0.503239631652832, "learning_rate": 7.875837875135991e-05, "loss": 0.0448, "step": 33970 }, { "grad_norm": 0.49363937973976135, "learning_rate": 7.874485120719646e-05, "loss": 0.0439, "step": 33980 }, { "grad_norm": 0.6280885934829712, "learning_rate": 7.873132051953928e-05, "loss": 0.0418, "step": 33990 }, { "grad_norm": 0.541372537612915, "learning_rate": 7.87177866898681e-05, "loss": 0.0472, "step": 34000 }, { "grad_norm": 0.38095319271087646, "learning_rate": 7.870424971966294e-05, "loss": 0.0524, "step": 34010 }, { "grad_norm": 0.5271213054656982, "learning_rate": 7.869070961040419e-05, "loss": 0.0426, "step": 34020 }, { "grad_norm": 0.39830130338668823, "learning_rate": 7.867716636357257e-05, "loss": 0.0473, "step": 34030 }, { "grad_norm": 0.4053798019886017, "learning_rate": 7.866361998064915e-05, "loss": 0.0451, "step": 34040 }, { "grad_norm": 0.431401789188385, "learning_rate": 7.865007046311534e-05, "loss": 0.0418, "step": 34050 }, { "grad_norm": 0.3630269169807434, "learning_rate": 7.86365178124529e-05, "loss": 0.0443, "step": 34060 }, { "grad_norm": 0.3896406888961792, "learning_rate": 7.862296203014394e-05, "loss": 0.0448, "step": 34070 }, { "grad_norm": 0.5224427580833435, "learning_rate": 7.860940311767088e-05, "loss": 0.0499, "step": 34080 }, { "grad_norm": 0.4602348804473877, "learning_rate": 7.85958410765165e-05, "loss": 0.0581, "step": 34090 }, { "grad_norm": 0.4318336844444275, "learning_rate": 7.858227590816394e-05, "loss": 0.0486, "step": 34100 }, { "grad_norm": 0.38841819763183594, "learning_rate": 7.856870761409664e-05, "loss": 0.0399, "step": 34110 }, { "grad_norm": 0.4630899131298065, "learning_rate": 7.855513619579846e-05, "loss": 0.0465, "step": 34120 }, { "grad_norm": 0.6855912804603577, "learning_rate": 7.85415616547535e-05, "loss": 0.0397, "step": 34130 }, { "grad_norm": 0.40144965052604675, "learning_rate": 7.852798399244627e-05, "loss": 0.047, "step": 34140 }, { "grad_norm": 0.4075039327144623, "learning_rate": 7.851440321036161e-05, "loss": 0.0412, "step": 34150 }, { "grad_norm": 0.5228572487831116, "learning_rate": 7.850081930998468e-05, "loss": 0.0481, "step": 34160 }, { "grad_norm": 0.37460818886756897, "learning_rate": 7.8487232292801e-05, "loss": 0.0489, "step": 34170 }, { "grad_norm": 0.5163187384605408, "learning_rate": 7.847364216029642e-05, "loss": 0.0425, "step": 34180 }, { "grad_norm": 0.3192182183265686, "learning_rate": 7.846004891395716e-05, "loss": 0.0341, "step": 34190 }, { "grad_norm": 0.30454909801483154, "learning_rate": 7.844645255526972e-05, "loss": 0.04, "step": 34200 }, { "grad_norm": 0.31990551948547363, "learning_rate": 7.843285308572101e-05, "loss": 0.0412, "step": 34210 }, { "grad_norm": 0.48446398973464966, "learning_rate": 7.841925050679823e-05, "loss": 0.0455, "step": 34220 }, { "grad_norm": 0.44358429312705994, "learning_rate": 7.840564481998895e-05, "loss": 0.0421, "step": 34230 }, { "grad_norm": 0.5286297798156738, "learning_rate": 7.839203602678105e-05, "loss": 0.0432, "step": 34240 }, { "grad_norm": 0.3252802789211273, "learning_rate": 7.837842412866279e-05, "loss": 0.0463, "step": 34250 }, { "grad_norm": 0.43420323729515076, "learning_rate": 7.836480912712272e-05, "loss": 0.0467, "step": 34260 }, { "grad_norm": 0.27789750695228577, "learning_rate": 7.835119102364976e-05, "loss": 0.0435, "step": 34270 }, { "grad_norm": 0.39906126260757446, "learning_rate": 7.83375698197332e-05, "loss": 0.0435, "step": 34280 }, { "grad_norm": 0.37561318278312683, "learning_rate": 7.83239455168626e-05, "loss": 0.0435, "step": 34290 }, { "grad_norm": 0.3823513686656952, "learning_rate": 7.83103181165279e-05, "loss": 0.0409, "step": 34300 }, { "grad_norm": 0.5671548247337341, "learning_rate": 7.829668762021937e-05, "loss": 0.0469, "step": 34310 }, { "grad_norm": 0.2597391903400421, "learning_rate": 7.828305402942764e-05, "loss": 0.0437, "step": 34320 }, { "grad_norm": 0.4255762994289398, "learning_rate": 7.826941734564363e-05, "loss": 0.0391, "step": 34330 }, { "grad_norm": 0.5456750988960266, "learning_rate": 7.825577757035865e-05, "loss": 0.0469, "step": 34340 }, { "grad_norm": 0.696699321269989, "learning_rate": 7.824213470506431e-05, "loss": 0.0476, "step": 34350 }, { "grad_norm": 0.4621336758136749, "learning_rate": 7.822848875125257e-05, "loss": 0.042, "step": 34360 }, { "grad_norm": 0.5092948079109192, "learning_rate": 7.821483971041576e-05, "loss": 0.0424, "step": 34370 }, { "grad_norm": 0.38242247700691223, "learning_rate": 7.820118758404649e-05, "loss": 0.0542, "step": 34380 }, { "grad_norm": 0.30440378189086914, "learning_rate": 7.818753237363776e-05, "loss": 0.0464, "step": 34390 }, { "grad_norm": 0.292296439409256, "learning_rate": 7.817387408068286e-05, "loss": 0.042, "step": 34400 }, { "grad_norm": 0.48296624422073364, "learning_rate": 7.816021270667544e-05, "loss": 0.0444, "step": 34410 }, { "grad_norm": 0.4018351435661316, "learning_rate": 7.81465482531095e-05, "loss": 0.0409, "step": 34420 }, { "grad_norm": 0.4289415180683136, "learning_rate": 7.813288072147938e-05, "loss": 0.0459, "step": 34430 }, { "grad_norm": 0.3148002326488495, "learning_rate": 7.811921011327972e-05, "loss": 0.0409, "step": 34440 }, { "grad_norm": 0.41730085015296936, "learning_rate": 7.810553643000549e-05, "loss": 0.0539, "step": 34450 }, { "grad_norm": 0.6050211191177368, "learning_rate": 7.809185967315206e-05, "loss": 0.0456, "step": 34460 }, { "grad_norm": 0.4719637334346771, "learning_rate": 7.80781798442151e-05, "loss": 0.0483, "step": 34470 }, { "grad_norm": 0.4339495599269867, "learning_rate": 7.806449694469061e-05, "loss": 0.0482, "step": 34480 }, { "grad_norm": 0.37716275453567505, "learning_rate": 7.805081097607492e-05, "loss": 0.0481, "step": 34490 }, { "grad_norm": 0.4601396918296814, "learning_rate": 7.803712193986474e-05, "loss": 0.0397, "step": 34500 }, { "grad_norm": 0.3120681047439575, "learning_rate": 7.802342983755702e-05, "loss": 0.0448, "step": 34510 }, { "grad_norm": 0.3575383126735687, "learning_rate": 7.800973467064918e-05, "loss": 0.0472, "step": 34520 }, { "grad_norm": 0.4333760142326355, "learning_rate": 7.799603644063884e-05, "loss": 0.0436, "step": 34530 }, { "grad_norm": 0.36266133189201355, "learning_rate": 7.798233514902405e-05, "loss": 0.0416, "step": 34540 }, { "grad_norm": 0.399877667427063, "learning_rate": 7.796863079730318e-05, "loss": 0.0455, "step": 34550 }, { "grad_norm": 0.3321101665496826, "learning_rate": 7.795492338697488e-05, "loss": 0.0366, "step": 34560 }, { "grad_norm": 0.44399675726890564, "learning_rate": 7.794121291953819e-05, "loss": 0.0435, "step": 34570 }, { "grad_norm": 0.3697647452354431, "learning_rate": 7.792749939649246e-05, "loss": 0.0391, "step": 34580 }, { "grad_norm": 0.523083508014679, "learning_rate": 7.79137828193374e-05, "loss": 0.0418, "step": 34590 }, { "grad_norm": 0.3554908335208893, "learning_rate": 7.790006318957301e-05, "loss": 0.0467, "step": 34600 }, { "grad_norm": 0.2992898225784302, "learning_rate": 7.788634050869965e-05, "loss": 0.0428, "step": 34610 }, { "grad_norm": 0.2984796166419983, "learning_rate": 7.787261477821803e-05, "loss": 0.044, "step": 34620 }, { "grad_norm": 0.30497506260871887, "learning_rate": 7.785888599962916e-05, "loss": 0.0472, "step": 34630 }, { "grad_norm": 0.5265032649040222, "learning_rate": 7.784515417443439e-05, "loss": 0.0396, "step": 34640 }, { "grad_norm": 0.39265280961990356, "learning_rate": 7.783141930413545e-05, "loss": 0.0481, "step": 34650 }, { "grad_norm": 0.37714749574661255, "learning_rate": 7.78176813902343e-05, "loss": 0.042, "step": 34660 }, { "grad_norm": 0.467339426279068, "learning_rate": 7.780394043423336e-05, "loss": 0.0404, "step": 34670 }, { "grad_norm": 0.3878691494464874, "learning_rate": 7.77901964376353e-05, "loss": 0.0468, "step": 34680 }, { "grad_norm": 0.41813698410987854, "learning_rate": 7.777644940194316e-05, "loss": 0.0442, "step": 34690 }, { "grad_norm": 0.37336868047714233, "learning_rate": 7.776269932866023e-05, "loss": 0.0473, "step": 34700 }, { "grad_norm": 0.410915732383728, "learning_rate": 7.774894621929026e-05, "loss": 0.0463, "step": 34710 }, { "grad_norm": 0.7005165219306946, "learning_rate": 7.773519007533725e-05, "loss": 0.0413, "step": 34720 }, { "grad_norm": 0.819415807723999, "learning_rate": 7.772143089830556e-05, "loss": 0.0414, "step": 34730 }, { "grad_norm": 0.417085200548172, "learning_rate": 7.770766868969985e-05, "loss": 0.0442, "step": 34740 }, { "grad_norm": 0.4197891056537628, "learning_rate": 7.769390345102518e-05, "loss": 0.0408, "step": 34750 }, { "grad_norm": 0.4106028378009796, "learning_rate": 7.768013518378683e-05, "loss": 0.0381, "step": 34760 }, { "grad_norm": 0.39632412791252136, "learning_rate": 7.766636388949053e-05, "loss": 0.0452, "step": 34770 }, { "grad_norm": 0.4403478801250458, "learning_rate": 7.765258956964229e-05, "loss": 0.047, "step": 34780 }, { "grad_norm": 0.5560224056243896, "learning_rate": 7.76388122257484e-05, "loss": 0.0483, "step": 34790 }, { "grad_norm": 0.37097591161727905, "learning_rate": 7.762503185931558e-05, "loss": 0.0452, "step": 34800 }, { "grad_norm": 0.6908332109451294, "learning_rate": 7.76112484718508e-05, "loss": 0.0476, "step": 34810 }, { "grad_norm": 0.5940592885017395, "learning_rate": 7.75974620648614e-05, "loss": 0.0514, "step": 34820 }, { "grad_norm": 0.33468756079673767, "learning_rate": 7.758367263985503e-05, "loss": 0.0403, "step": 34830 }, { "grad_norm": 0.3234838843345642, "learning_rate": 7.75698801983397e-05, "loss": 0.0503, "step": 34840 }, { "grad_norm": 0.47560158371925354, "learning_rate": 7.755608474182372e-05, "loss": 0.0444, "step": 34850 }, { "grad_norm": 0.36126720905303955, "learning_rate": 7.754228627181574e-05, "loss": 0.0419, "step": 34860 }, { "grad_norm": 0.38462844491004944, "learning_rate": 7.752848478982476e-05, "loss": 0.0416, "step": 34870 }, { "grad_norm": 0.4663386940956116, "learning_rate": 7.751468029736006e-05, "loss": 0.0441, "step": 34880 }, { "grad_norm": 0.4711589515209198, "learning_rate": 7.750087279593129e-05, "loss": 0.0435, "step": 34890 }, { "grad_norm": 0.5656384825706482, "learning_rate": 7.748706228704843e-05, "loss": 0.0465, "step": 34900 }, { "grad_norm": 0.35344910621643066, "learning_rate": 7.747324877222176e-05, "loss": 0.0416, "step": 34910 }, { "grad_norm": 0.3023432493209839, "learning_rate": 7.745943225296188e-05, "loss": 0.0411, "step": 34920 }, { "grad_norm": 0.35875049233436584, "learning_rate": 7.744561273077981e-05, "loss": 0.0434, "step": 34930 }, { "grad_norm": 0.45701298117637634, "learning_rate": 7.743179020718678e-05, "loss": 0.045, "step": 34940 }, { "grad_norm": 0.3299846649169922, "learning_rate": 7.741796468369443e-05, "loss": 0.0412, "step": 34950 }, { "grad_norm": 0.3466973900794983, "learning_rate": 7.740413616181466e-05, "loss": 0.041, "step": 34960 }, { "grad_norm": 0.4027473032474518, "learning_rate": 7.739030464305978e-05, "loss": 0.0466, "step": 34970 }, { "grad_norm": 0.3183833062648773, "learning_rate": 7.737647012894235e-05, "loss": 0.0416, "step": 34980 }, { "grad_norm": 0.25726038217544556, "learning_rate": 7.736263262097532e-05, "loss": 0.0431, "step": 34990 }, { "grad_norm": 0.7694525718688965, "learning_rate": 7.734879212067192e-05, "loss": 0.046, "step": 35000 }, { "grad_norm": 0.3510056436061859, "learning_rate": 7.733494862954573e-05, "loss": 0.0421, "step": 35010 }, { "grad_norm": 0.7782676219940186, "learning_rate": 7.732110214911066e-05, "loss": 0.0461, "step": 35020 }, { "grad_norm": 0.8479032516479492, "learning_rate": 7.730725268088094e-05, "loss": 0.0451, "step": 35030 }, { "grad_norm": 0.6349579691886902, "learning_rate": 7.729340022637112e-05, "loss": 0.0466, "step": 35040 }, { "grad_norm": 0.4640619158744812, "learning_rate": 7.727954478709607e-05, "loss": 0.0442, "step": 35050 }, { "grad_norm": 0.42389440536499023, "learning_rate": 7.726568636457103e-05, "loss": 0.0423, "step": 35060 }, { "grad_norm": 0.41255754232406616, "learning_rate": 7.725182496031153e-05, "loss": 0.0447, "step": 35070 }, { "grad_norm": 0.25845253467559814, "learning_rate": 7.723796057583341e-05, "loss": 0.0464, "step": 35080 }, { "grad_norm": 0.4081452488899231, "learning_rate": 7.722409321265287e-05, "loss": 0.0418, "step": 35090 }, { "grad_norm": 0.590259850025177, "learning_rate": 7.721022287228645e-05, "loss": 0.0437, "step": 35100 }, { "grad_norm": 0.4804774820804596, "learning_rate": 7.719634955625094e-05, "loss": 0.0479, "step": 35110 }, { "grad_norm": 0.5421051383018494, "learning_rate": 7.718247326606355e-05, "loss": 0.0463, "step": 35120 }, { "grad_norm": 0.5319561958312988, "learning_rate": 7.716859400324175e-05, "loss": 0.0485, "step": 35130 }, { "grad_norm": 0.3651076853275299, "learning_rate": 7.715471176930335e-05, "loss": 0.0449, "step": 35140 }, { "grad_norm": 0.41025876998901367, "learning_rate": 7.714082656576651e-05, "loss": 0.0411, "step": 35150 }, { "grad_norm": 0.5030275583267212, "learning_rate": 7.712693839414968e-05, "loss": 0.0494, "step": 35160 }, { "grad_norm": 0.5759290456771851, "learning_rate": 7.711304725597164e-05, "loss": 0.0511, "step": 35170 }, { "grad_norm": 0.4015807509422302, "learning_rate": 7.709915315275151e-05, "loss": 0.0408, "step": 35180 }, { "grad_norm": 0.27709972858428955, "learning_rate": 7.708525608600876e-05, "loss": 0.0434, "step": 35190 }, { "grad_norm": 0.3087311089038849, "learning_rate": 7.707135605726311e-05, "loss": 0.0498, "step": 35200 }, { "grad_norm": 0.3975389897823334, "learning_rate": 7.705745306803466e-05, "loss": 0.0451, "step": 35210 }, { "grad_norm": 0.41038861870765686, "learning_rate": 7.704354711984383e-05, "loss": 0.0459, "step": 35220 }, { "grad_norm": 0.4192773699760437, "learning_rate": 7.702963821421133e-05, "loss": 0.04, "step": 35230 }, { "grad_norm": 0.30340760946273804, "learning_rate": 7.701572635265826e-05, "loss": 0.0413, "step": 35240 }, { "grad_norm": 0.4553314745426178, "learning_rate": 7.700181153670596e-05, "loss": 0.0431, "step": 35250 }, { "grad_norm": 0.36910998821258545, "learning_rate": 7.698789376787616e-05, "loss": 0.0457, "step": 35260 }, { "grad_norm": 0.364375501871109, "learning_rate": 7.697397304769085e-05, "loss": 0.0483, "step": 35270 }, { "grad_norm": 0.7876623272895813, "learning_rate": 7.696004937767241e-05, "loss": 0.043, "step": 35280 }, { "grad_norm": 0.3529497981071472, "learning_rate": 7.694612275934352e-05, "loss": 0.0408, "step": 35290 }, { "grad_norm": 0.43650731444358826, "learning_rate": 7.693219319422714e-05, "loss": 0.0433, "step": 35300 }, { "grad_norm": 0.33103349804878235, "learning_rate": 7.69182606838466e-05, "loss": 0.0397, "step": 35310 }, { "grad_norm": 0.49742376804351807, "learning_rate": 7.690432522972558e-05, "loss": 0.0449, "step": 35320 }, { "grad_norm": 0.5024480819702148, "learning_rate": 7.689038683338796e-05, "loss": 0.0509, "step": 35330 }, { "grad_norm": 0.41264986991882324, "learning_rate": 7.687644549635808e-05, "loss": 0.0448, "step": 35340 }, { "grad_norm": 0.45131734013557434, "learning_rate": 7.686250122016053e-05, "loss": 0.0497, "step": 35350 }, { "grad_norm": 0.6577125191688538, "learning_rate": 7.684855400632022e-05, "loss": 0.042, "step": 35360 }, { "grad_norm": 0.4955448508262634, "learning_rate": 7.683460385636243e-05, "loss": 0.036, "step": 35370 }, { "grad_norm": 0.4186219573020935, "learning_rate": 7.68206507718127e-05, "loss": 0.0496, "step": 35380 }, { "grad_norm": 0.45850858092308044, "learning_rate": 7.680669475419693e-05, "loss": 0.0415, "step": 35390 }, { "grad_norm": 0.34473171830177307, "learning_rate": 7.679273580504132e-05, "loss": 0.0448, "step": 35400 }, { "grad_norm": 0.3067224323749542, "learning_rate": 7.67787739258724e-05, "loss": 0.0457, "step": 35410 }, { "grad_norm": 0.34876808524131775, "learning_rate": 7.676480911821705e-05, "loss": 0.0425, "step": 35420 }, { "grad_norm": 0.4275575876235962, "learning_rate": 7.675084138360238e-05, "loss": 0.0415, "step": 35430 }, { "grad_norm": 0.46643972396850586, "learning_rate": 7.673687072355592e-05, "loss": 0.0485, "step": 35440 }, { "grad_norm": 0.4317604601383209, "learning_rate": 7.67228971396055e-05, "loss": 0.0389, "step": 35450 }, { "grad_norm": 0.4252898097038269, "learning_rate": 7.670892063327922e-05, "loss": 0.0384, "step": 35460 }, { "grad_norm": 0.3037172853946686, "learning_rate": 7.669494120610552e-05, "loss": 0.044, "step": 35470 }, { "grad_norm": 0.4491603374481201, "learning_rate": 7.66809588596132e-05, "loss": 0.0444, "step": 35480 }, { "grad_norm": 0.39878034591674805, "learning_rate": 7.666697359533132e-05, "loss": 0.0434, "step": 35490 }, { "grad_norm": 0.6717430949211121, "learning_rate": 7.665298541478932e-05, "loss": 0.0446, "step": 35500 }, { "grad_norm": 0.45370954275131226, "learning_rate": 7.66389943195169e-05, "loss": 0.0477, "step": 35510 }, { "grad_norm": 0.3904096782207489, "learning_rate": 7.66250003110441e-05, "loss": 0.05, "step": 35520 }, { "grad_norm": 0.3566552996635437, "learning_rate": 7.66110033909013e-05, "loss": 0.0448, "step": 35530 }, { "grad_norm": 0.43265458941459656, "learning_rate": 7.659700356061918e-05, "loss": 0.0449, "step": 35540 }, { "grad_norm": 0.36823949217796326, "learning_rate": 7.658300082172875e-05, "loss": 0.0423, "step": 35550 }, { "grad_norm": 0.34849604964256287, "learning_rate": 7.65689951757613e-05, "loss": 0.043, "step": 35560 }, { "grad_norm": 0.420050710439682, "learning_rate": 7.655498662424849e-05, "loss": 0.0415, "step": 35570 }, { "grad_norm": 0.3196597397327423, "learning_rate": 7.654097516872227e-05, "loss": 0.0416, "step": 35580 }, { "grad_norm": 0.4588514566421509, "learning_rate": 7.65269608107149e-05, "loss": 0.0424, "step": 35590 }, { "grad_norm": 0.4372391700744629, "learning_rate": 7.6512943551759e-05, "loss": 0.0417, "step": 35600 }, { "grad_norm": 0.4652569890022278, "learning_rate": 7.649892339338743e-05, "loss": 0.043, "step": 35610 }, { "grad_norm": 0.420052170753479, "learning_rate": 7.648490033713344e-05, "loss": 0.0488, "step": 35620 }, { "grad_norm": 0.3809655010700226, "learning_rate": 7.647087438453058e-05, "loss": 0.0476, "step": 35630 }, { "grad_norm": 0.4722665846347809, "learning_rate": 7.64568455371127e-05, "loss": 0.0482, "step": 35640 }, { "grad_norm": 0.41597554087638855, "learning_rate": 7.644281379641396e-05, "loss": 0.0402, "step": 35650 }, { "grad_norm": 0.5469096899032593, "learning_rate": 7.642877916396887e-05, "loss": 0.0419, "step": 35660 }, { "grad_norm": 0.27893805503845215, "learning_rate": 7.641474164131221e-05, "loss": 0.0395, "step": 35670 }, { "grad_norm": 0.39246416091918945, "learning_rate": 7.640070122997913e-05, "loss": 0.0396, "step": 35680 }, { "grad_norm": 0.5212214589118958, "learning_rate": 7.638665793150505e-05, "loss": 0.0445, "step": 35690 }, { "grad_norm": 0.3186897039413452, "learning_rate": 7.637261174742574e-05, "loss": 0.0415, "step": 35700 }, { "grad_norm": 0.39485108852386475, "learning_rate": 7.635856267927725e-05, "loss": 0.0434, "step": 35710 }, { "grad_norm": 0.40926656126976013, "learning_rate": 7.634451072859597e-05, "loss": 0.044, "step": 35720 }, { "grad_norm": 0.9631550312042236, "learning_rate": 7.633045589691863e-05, "loss": 0.0467, "step": 35730 }, { "grad_norm": 0.43580004572868347, "learning_rate": 7.63163981857822e-05, "loss": 0.0523, "step": 35740 }, { "grad_norm": 0.9731085300445557, "learning_rate": 7.630233759672403e-05, "loss": 0.0469, "step": 35750 }, { "grad_norm": 0.5040515065193176, "learning_rate": 7.628827413128177e-05, "loss": 0.0463, "step": 35760 }, { "grad_norm": 0.3011516332626343, "learning_rate": 7.627420779099338e-05, "loss": 0.0381, "step": 35770 }, { "grad_norm": 0.31561288237571716, "learning_rate": 7.626013857739711e-05, "loss": 0.0463, "step": 35780 }, { "grad_norm": 0.4242202341556549, "learning_rate": 7.624606649203158e-05, "loss": 0.0445, "step": 35790 }, { "grad_norm": 0.6192983388900757, "learning_rate": 7.623199153643569e-05, "loss": 0.0448, "step": 35800 }, { "grad_norm": 0.5022764205932617, "learning_rate": 7.621791371214863e-05, "loss": 0.0459, "step": 35810 }, { "grad_norm": 0.38538357615470886, "learning_rate": 7.620383302070995e-05, "loss": 0.0457, "step": 35820 }, { "grad_norm": 0.3673400282859802, "learning_rate": 7.61897494636595e-05, "loss": 0.0484, "step": 35830 }, { "grad_norm": 0.30269932746887207, "learning_rate": 7.617566304253739e-05, "loss": 0.0561, "step": 35840 }, { "grad_norm": 0.35940179228782654, "learning_rate": 7.616157375888416e-05, "loss": 0.0426, "step": 35850 }, { "grad_norm": 0.33312082290649414, "learning_rate": 7.614748161424053e-05, "loss": 0.041, "step": 35860 }, { "grad_norm": 0.5403600335121155, "learning_rate": 7.613338661014763e-05, "loss": 0.0493, "step": 35870 }, { "grad_norm": 0.33829623460769653, "learning_rate": 7.611928874814686e-05, "loss": 0.0428, "step": 35880 }, { "grad_norm": 0.5447784662246704, "learning_rate": 7.610518802977996e-05, "loss": 0.0396, "step": 35890 }, { "grad_norm": 0.3122923970222473, "learning_rate": 7.609108445658893e-05, "loss": 0.0348, "step": 35900 }, { "grad_norm": 0.31386515498161316, "learning_rate": 7.607697803011612e-05, "loss": 0.0416, "step": 35910 }, { "grad_norm": 0.32329556345939636, "learning_rate": 7.606286875190421e-05, "loss": 0.0537, "step": 35920 }, { "grad_norm": 0.30464860796928406, "learning_rate": 7.604875662349617e-05, "loss": 0.0419, "step": 35930 }, { "grad_norm": 0.2917879521846771, "learning_rate": 7.603464164643525e-05, "loss": 0.0382, "step": 35940 }, { "grad_norm": 0.4002886414527893, "learning_rate": 7.602052382226507e-05, "loss": 0.0461, "step": 35950 }, { "grad_norm": 0.5592774152755737, "learning_rate": 7.600640315252954e-05, "loss": 0.0479, "step": 35960 }, { "grad_norm": 0.5803223848342896, "learning_rate": 7.599227963877284e-05, "loss": 0.0436, "step": 35970 }, { "grad_norm": 0.38082820177078247, "learning_rate": 7.597815328253954e-05, "loss": 0.0482, "step": 35980 }, { "grad_norm": 0.3071994185447693, "learning_rate": 7.596402408537443e-05, "loss": 0.0398, "step": 35990 }, { "grad_norm": 0.36213067173957825, "learning_rate": 7.59498920488227e-05, "loss": 0.0401, "step": 36000 }, { "grad_norm": 0.4962560534477234, "learning_rate": 7.593575717442979e-05, "loss": 0.0446, "step": 36010 }, { "grad_norm": 0.3283732831478119, "learning_rate": 7.592161946374147e-05, "loss": 0.045, "step": 36020 }, { "grad_norm": 0.3559679090976715, "learning_rate": 7.590747891830381e-05, "loss": 0.0431, "step": 36030 }, { "grad_norm": 0.4198724925518036, "learning_rate": 7.58933355396632e-05, "loss": 0.0429, "step": 36040 }, { "grad_norm": 0.3230794072151184, "learning_rate": 7.587918932936636e-05, "loss": 0.041, "step": 36050 }, { "grad_norm": 0.5455100536346436, "learning_rate": 7.586504028896028e-05, "loss": 0.0496, "step": 36060 }, { "grad_norm": 0.419175386428833, "learning_rate": 7.585088841999228e-05, "loss": 0.0439, "step": 36070 }, { "grad_norm": 0.30076727271080017, "learning_rate": 7.583673372400999e-05, "loss": 0.036, "step": 36080 }, { "grad_norm": 0.8121820688247681, "learning_rate": 7.582257620256134e-05, "loss": 0.0527, "step": 36090 }, { "grad_norm": 0.6581758856773376, "learning_rate": 7.580841585719458e-05, "loss": 0.0489, "step": 36100 }, { "grad_norm": 0.375199556350708, "learning_rate": 7.579425268945825e-05, "loss": 0.0516, "step": 36110 }, { "grad_norm": 0.378082275390625, "learning_rate": 7.578008670090127e-05, "loss": 0.0471, "step": 36120 }, { "grad_norm": 0.48453274369239807, "learning_rate": 7.576591789307272e-05, "loss": 0.0552, "step": 36130 }, { "grad_norm": 0.691239595413208, "learning_rate": 7.575174626752216e-05, "loss": 0.0464, "step": 36140 }, { "grad_norm": 0.5785589814186096, "learning_rate": 7.573757182579934e-05, "loss": 0.0466, "step": 36150 }, { "grad_norm": 0.36132314801216125, "learning_rate": 7.572339456945435e-05, "loss": 0.0475, "step": 36160 }, { "grad_norm": 0.3336104154586792, "learning_rate": 7.570921450003762e-05, "loss": 0.0445, "step": 36170 }, { "grad_norm": 0.3343847692012787, "learning_rate": 7.569503161909983e-05, "loss": 0.0475, "step": 36180 }, { "grad_norm": 0.3807026147842407, "learning_rate": 7.568084592819202e-05, "loss": 0.0477, "step": 36190 }, { "grad_norm": 0.5287894606590271, "learning_rate": 7.566665742886551e-05, "loss": 0.0417, "step": 36200 }, { "grad_norm": 0.30440372228622437, "learning_rate": 7.565246612267194e-05, "loss": 0.0465, "step": 36210 }, { "grad_norm": 0.24432198703289032, "learning_rate": 7.563827201116325e-05, "loss": 0.039, "step": 36220 }, { "grad_norm": 0.7438514828681946, "learning_rate": 7.562407509589164e-05, "loss": 0.0486, "step": 36230 }, { "grad_norm": 0.3855763077735901, "learning_rate": 7.560987537840973e-05, "loss": 0.0444, "step": 36240 }, { "grad_norm": 0.439816951751709, "learning_rate": 7.559567286027036e-05, "loss": 0.0441, "step": 36250 }, { "grad_norm": 0.3294244110584259, "learning_rate": 7.558146754302668e-05, "loss": 0.0428, "step": 36260 }, { "grad_norm": 0.382678359746933, "learning_rate": 7.556725942823217e-05, "loss": 0.0401, "step": 36270 }, { "grad_norm": 0.4936324954032898, "learning_rate": 7.555304851744061e-05, "loss": 0.0407, "step": 36280 }, { "grad_norm": 0.3007042407989502, "learning_rate": 7.553883481220608e-05, "loss": 0.0408, "step": 36290 }, { "grad_norm": 0.5531551837921143, "learning_rate": 7.552461831408298e-05, "loss": 0.041, "step": 36300 }, { "grad_norm": 0.5347194075584412, "learning_rate": 7.551039902462599e-05, "loss": 0.043, "step": 36310 }, { "grad_norm": 0.3875235915184021, "learning_rate": 7.549617694539011e-05, "loss": 0.0425, "step": 36320 }, { "grad_norm": 0.38637033104896545, "learning_rate": 7.548195207793066e-05, "loss": 0.042, "step": 36330 }, { "grad_norm": 0.3709883689880371, "learning_rate": 7.546772442380323e-05, "loss": 0.0437, "step": 36340 }, { "grad_norm": 0.4941703975200653, "learning_rate": 7.545349398456376e-05, "loss": 0.0504, "step": 36350 }, { "grad_norm": 0.3101380467414856, "learning_rate": 7.543926076176845e-05, "loss": 0.0431, "step": 36360 }, { "grad_norm": 0.5309861302375793, "learning_rate": 7.542502475697385e-05, "loss": 0.0446, "step": 36370 }, { "grad_norm": 0.4496811628341675, "learning_rate": 7.541078597173675e-05, "loss": 0.0484, "step": 36380 }, { "grad_norm": 0.3742130994796753, "learning_rate": 7.539654440761431e-05, "loss": 0.0424, "step": 36390 }, { "grad_norm": 0.37878990173339844, "learning_rate": 7.538230006616395e-05, "loss": 0.0492, "step": 36400 }, { "grad_norm": 0.23982654511928558, "learning_rate": 7.536805294894342e-05, "loss": 0.0369, "step": 36410 }, { "grad_norm": 0.4461963176727295, "learning_rate": 7.535380305751076e-05, "loss": 0.0433, "step": 36420 }, { "grad_norm": 0.3409416377544403, "learning_rate": 7.533955039342431e-05, "loss": 0.0416, "step": 36430 }, { "grad_norm": 0.38231152296066284, "learning_rate": 7.532529495824274e-05, "loss": 0.042, "step": 36440 }, { "grad_norm": 0.33073580265045166, "learning_rate": 7.531103675352497e-05, "loss": 0.044, "step": 36450 }, { "grad_norm": 0.44486454129219055, "learning_rate": 7.529677578083029e-05, "loss": 0.0436, "step": 36460 }, { "grad_norm": 0.36513563990592957, "learning_rate": 7.528251204171823e-05, "loss": 0.0448, "step": 36470 }, { "grad_norm": 0.3311263918876648, "learning_rate": 7.526824553774867e-05, "loss": 0.0476, "step": 36480 }, { "grad_norm": 0.3903033137321472, "learning_rate": 7.525397627048178e-05, "loss": 0.0469, "step": 36490 }, { "grad_norm": 0.39998114109039307, "learning_rate": 7.523970424147802e-05, "loss": 0.039, "step": 36500 }, { "grad_norm": 0.4420222043991089, "learning_rate": 7.522542945229813e-05, "loss": 0.0434, "step": 36510 }, { "grad_norm": 0.4395866096019745, "learning_rate": 7.521115190450321e-05, "loss": 0.0405, "step": 36520 }, { "grad_norm": 0.33049309253692627, "learning_rate": 7.519687159965462e-05, "loss": 0.0466, "step": 36530 }, { "grad_norm": 0.2783118486404419, "learning_rate": 7.518258853931403e-05, "loss": 0.0385, "step": 36540 }, { "grad_norm": 0.3539849519729614, "learning_rate": 7.516830272504342e-05, "loss": 0.0388, "step": 36550 }, { "grad_norm": 0.4943154752254486, "learning_rate": 7.515401415840509e-05, "loss": 0.0477, "step": 36560 }, { "grad_norm": 0.4060858488082886, "learning_rate": 7.513972284096155e-05, "loss": 0.0432, "step": 36570 }, { "grad_norm": 0.36471229791641235, "learning_rate": 7.512542877427574e-05, "loss": 0.0409, "step": 36580 }, { "grad_norm": 0.31915581226348877, "learning_rate": 7.51111319599108e-05, "loss": 0.0418, "step": 36590 }, { "grad_norm": 0.29513001441955566, "learning_rate": 7.509683239943024e-05, "loss": 0.0377, "step": 36600 }, { "grad_norm": 0.4278605282306671, "learning_rate": 7.50825300943978e-05, "loss": 0.0532, "step": 36610 }, { "grad_norm": 0.3480810523033142, "learning_rate": 7.50682250463776e-05, "loss": 0.0456, "step": 36620 }, { "grad_norm": 0.5330085158348083, "learning_rate": 7.5053917256934e-05, "loss": 0.045, "step": 36630 }, { "grad_norm": 0.7704862952232361, "learning_rate": 7.503960672763166e-05, "loss": 0.0377, "step": 36640 }, { "grad_norm": 0.46902865171432495, "learning_rate": 7.502529346003559e-05, "loss": 0.0406, "step": 36650 }, { "grad_norm": 0.6639729738235474, "learning_rate": 7.501097745571106e-05, "loss": 0.052, "step": 36660 }, { "grad_norm": 0.35008567571640015, "learning_rate": 7.499665871622361e-05, "loss": 0.0366, "step": 36670 }, { "grad_norm": 0.31511738896369934, "learning_rate": 7.498233724313917e-05, "loss": 0.0409, "step": 36680 }, { "grad_norm": 0.43012720346450806, "learning_rate": 7.496801303802387e-05, "loss": 0.043, "step": 36690 }, { "grad_norm": 0.35829827189445496, "learning_rate": 7.495368610244423e-05, "loss": 0.0416, "step": 36700 }, { "grad_norm": 0.41212865710258484, "learning_rate": 7.493935643796697e-05, "loss": 0.0402, "step": 36710 }, { "grad_norm": 0.42171645164489746, "learning_rate": 7.49250240461592e-05, "loss": 0.0413, "step": 36720 }, { "grad_norm": 0.5390222072601318, "learning_rate": 7.491068892858826e-05, "loss": 0.041, "step": 36730 }, { "grad_norm": 0.350318044424057, "learning_rate": 7.489635108682184e-05, "loss": 0.0436, "step": 36740 }, { "grad_norm": 0.358187735080719, "learning_rate": 7.48820105224279e-05, "loss": 0.0452, "step": 36750 }, { "grad_norm": 0.34192079305648804, "learning_rate": 7.48676672369747e-05, "loss": 0.0423, "step": 36760 }, { "grad_norm": 0.4189211130142212, "learning_rate": 7.485332123203079e-05, "loss": 0.0484, "step": 36770 }, { "grad_norm": 0.3438754081726074, "learning_rate": 7.483897250916504e-05, "loss": 0.0401, "step": 36780 }, { "grad_norm": 0.393251895904541, "learning_rate": 7.48246210699466e-05, "loss": 0.0408, "step": 36790 }, { "grad_norm": 0.4586982727050781, "learning_rate": 7.481026691594492e-05, "loss": 0.0427, "step": 36800 }, { "grad_norm": 0.386851042509079, "learning_rate": 7.479591004872975e-05, "loss": 0.0434, "step": 36810 }, { "grad_norm": 0.2720058560371399, "learning_rate": 7.478155046987114e-05, "loss": 0.0417, "step": 36820 }, { "grad_norm": 0.3766162693500519, "learning_rate": 7.476718818093944e-05, "loss": 0.0414, "step": 36830 }, { "grad_norm": 0.3442586660385132, "learning_rate": 7.475282318350527e-05, "loss": 0.0393, "step": 36840 }, { "grad_norm": 1.6219033002853394, "learning_rate": 7.473845547913957e-05, "loss": 0.0432, "step": 36850 }, { "grad_norm": 0.7924520373344421, "learning_rate": 7.472408506941357e-05, "loss": 0.0527, "step": 36860 }, { "grad_norm": 0.38681724667549133, "learning_rate": 7.470971195589881e-05, "loss": 0.0555, "step": 36870 }, { "grad_norm": 0.5907882452011108, "learning_rate": 7.46953361401671e-05, "loss": 0.0457, "step": 36880 }, { "grad_norm": 0.43725576996803284, "learning_rate": 7.468095762379055e-05, "loss": 0.0456, "step": 36890 }, { "grad_norm": 0.3520839214324951, "learning_rate": 7.466657640834158e-05, "loss": 0.04, "step": 36900 }, { "grad_norm": 0.85029536485672, "learning_rate": 7.46521924953929e-05, "loss": 0.0382, "step": 36910 }, { "grad_norm": 0.33900076150894165, "learning_rate": 7.463780588651752e-05, "loss": 0.0426, "step": 36920 }, { "grad_norm": 0.3410351872444153, "learning_rate": 7.462341658328873e-05, "loss": 0.0435, "step": 36930 }, { "grad_norm": 0.3402842581272125, "learning_rate": 7.460902458728012e-05, "loss": 0.0425, "step": 36940 }, { "grad_norm": 0.35734137892723083, "learning_rate": 7.459462990006558e-05, "loss": 0.0395, "step": 36950 }, { "grad_norm": 0.30693578720092773, "learning_rate": 7.458023252321929e-05, "loss": 0.0384, "step": 36960 }, { "grad_norm": 0.48689159750938416, "learning_rate": 7.456583245831573e-05, "loss": 0.0476, "step": 36970 }, { "grad_norm": 0.4411711096763611, "learning_rate": 7.455142970692968e-05, "loss": 0.0448, "step": 36980 }, { "grad_norm": 0.24478362500667572, "learning_rate": 7.453702427063615e-05, "loss": 0.0457, "step": 36990 }, { "grad_norm": 0.45164260268211365, "learning_rate": 7.452261615101057e-05, "loss": 0.0435, "step": 37000 }, { "grad_norm": 0.48291048407554626, "learning_rate": 7.450820534962856e-05, "loss": 0.0441, "step": 37010 }, { "grad_norm": 0.3943241238594055, "learning_rate": 7.449379186806604e-05, "loss": 0.0431, "step": 37020 }, { "grad_norm": 0.49306201934814453, "learning_rate": 7.44793757078993e-05, "loss": 0.0408, "step": 37030 }, { "grad_norm": 0.34890076518058777, "learning_rate": 7.44649568707048e-05, "loss": 0.0466, "step": 37040 }, { "grad_norm": 0.41063350439071655, "learning_rate": 7.445053535805942e-05, "loss": 0.0445, "step": 37050 }, { "grad_norm": 0.3051064908504486, "learning_rate": 7.443611117154024e-05, "loss": 0.043, "step": 37060 }, { "grad_norm": 0.36919084191322327, "learning_rate": 7.442168431272469e-05, "loss": 0.0422, "step": 37070 }, { "grad_norm": 0.35814541578292847, "learning_rate": 7.440725478319044e-05, "loss": 0.0459, "step": 37080 }, { "grad_norm": 0.6892855167388916, "learning_rate": 7.439282258451553e-05, "loss": 0.0469, "step": 37090 }, { "grad_norm": 0.5287664532661438, "learning_rate": 7.43783877182782e-05, "loss": 0.0503, "step": 37100 }, { "grad_norm": 0.47469690442085266, "learning_rate": 7.436395018605707e-05, "loss": 0.0426, "step": 37110 }, { "grad_norm": 0.2988295257091522, "learning_rate": 7.434950998943094e-05, "loss": 0.0428, "step": 37120 }, { "grad_norm": 0.38160020112991333, "learning_rate": 7.433506712997902e-05, "loss": 0.0454, "step": 37130 }, { "grad_norm": 0.5048115253448486, "learning_rate": 7.432062160928077e-05, "loss": 0.0393, "step": 37140 }, { "grad_norm": 0.25729241967201233, "learning_rate": 7.430617342891588e-05, "loss": 0.0354, "step": 37150 }, { "grad_norm": 0.3414977490901947, "learning_rate": 7.429172259046442e-05, "loss": 0.0408, "step": 37160 }, { "grad_norm": 0.47210952639579773, "learning_rate": 7.42772690955067e-05, "loss": 0.0392, "step": 37170 }, { "grad_norm": 0.29662978649139404, "learning_rate": 7.426281294562332e-05, "loss": 0.0439, "step": 37180 }, { "grad_norm": 0.5339497923851013, "learning_rate": 7.424835414239519e-05, "loss": 0.0415, "step": 37190 }, { "grad_norm": 0.31226256489753723, "learning_rate": 7.423389268740352e-05, "loss": 0.0435, "step": 37200 }, { "grad_norm": 0.4287410080432892, "learning_rate": 7.42194285822298e-05, "loss": 0.0408, "step": 37210 }, { "grad_norm": 0.3133644163608551, "learning_rate": 7.420496182845576e-05, "loss": 0.0419, "step": 37220 }, { "grad_norm": 0.39517244696617126, "learning_rate": 7.419049242766348e-05, "loss": 0.0436, "step": 37230 }, { "grad_norm": 0.40766721963882446, "learning_rate": 7.417602038143533e-05, "loss": 0.039, "step": 37240 }, { "grad_norm": 0.31000831723213196, "learning_rate": 7.416154569135393e-05, "loss": 0.0477, "step": 37250 }, { "grad_norm": 0.5928859710693359, "learning_rate": 7.414706835900222e-05, "loss": 0.0448, "step": 37260 }, { "grad_norm": 0.29854726791381836, "learning_rate": 7.413258838596345e-05, "loss": 0.0435, "step": 37270 }, { "grad_norm": 0.4563325345516205, "learning_rate": 7.411810577382106e-05, "loss": 0.0401, "step": 37280 }, { "grad_norm": 0.4139626622200012, "learning_rate": 7.410362052415889e-05, "loss": 0.0467, "step": 37290 }, { "grad_norm": 0.42838600277900696, "learning_rate": 7.408913263856102e-05, "loss": 0.0368, "step": 37300 }, { "grad_norm": 0.33672401309013367, "learning_rate": 7.407464211861183e-05, "loss": 0.0421, "step": 37310 }, { "grad_norm": 0.3306023180484772, "learning_rate": 7.406014896589597e-05, "loss": 0.0384, "step": 37320 }, { "grad_norm": 0.473427414894104, "learning_rate": 7.404565318199838e-05, "loss": 0.0454, "step": 37330 }, { "grad_norm": 0.3663621246814728, "learning_rate": 7.403115476850432e-05, "loss": 0.0405, "step": 37340 }, { "grad_norm": 0.388094425201416, "learning_rate": 7.401665372699932e-05, "loss": 0.0415, "step": 37350 }, { "grad_norm": 0.49391570687294006, "learning_rate": 7.400215005906916e-05, "loss": 0.0482, "step": 37360 }, { "grad_norm": 0.4035390019416809, "learning_rate": 7.398764376629996e-05, "loss": 0.0428, "step": 37370 }, { "grad_norm": 0.2821062505245209, "learning_rate": 7.39731348502781e-05, "loss": 0.0405, "step": 37380 }, { "grad_norm": 0.3666801452636719, "learning_rate": 7.395862331259028e-05, "loss": 0.0449, "step": 37390 }, { "grad_norm": 0.36063510179519653, "learning_rate": 7.394410915482341e-05, "loss": 0.0381, "step": 37400 }, { "grad_norm": 0.3755360245704651, "learning_rate": 7.392959237856476e-05, "loss": 0.0382, "step": 37410 }, { "grad_norm": 0.5026147365570068, "learning_rate": 7.391507298540187e-05, "loss": 0.0386, "step": 37420 }, { "grad_norm": 0.5029518008232117, "learning_rate": 7.390055097692257e-05, "loss": 0.0384, "step": 37430 }, { "grad_norm": 0.35118481516838074, "learning_rate": 7.388602635471494e-05, "loss": 0.0381, "step": 37440 }, { "grad_norm": 0.26184573769569397, "learning_rate": 7.387149912036738e-05, "loss": 0.0388, "step": 37450 }, { "grad_norm": 0.4422592520713806, "learning_rate": 7.385696927546854e-05, "loss": 0.0446, "step": 37460 }, { "grad_norm": 0.814854085445404, "learning_rate": 7.384243682160743e-05, "loss": 0.0455, "step": 37470 }, { "grad_norm": 0.37349605560302734, "learning_rate": 7.382790176037329e-05, "loss": 0.0388, "step": 37480 }, { "grad_norm": 0.3930959701538086, "learning_rate": 7.381336409335562e-05, "loss": 0.0434, "step": 37490 }, { "grad_norm": 0.33282551169395447, "learning_rate": 7.379882382214426e-05, "loss": 0.04, "step": 37500 }, { "grad_norm": 1.1038366556167603, "learning_rate": 7.378428094832931e-05, "loss": 0.0386, "step": 37510 }, { "grad_norm": 0.5534400343894958, "learning_rate": 7.376973547350114e-05, "loss": 0.0358, "step": 37520 }, { "grad_norm": 0.4371408522129059, "learning_rate": 7.375518739925046e-05, "loss": 0.043, "step": 37530 }, { "grad_norm": 0.32228201627731323, "learning_rate": 7.374063672716818e-05, "loss": 0.0366, "step": 37540 }, { "grad_norm": 0.38245388865470886, "learning_rate": 7.372608345884558e-05, "loss": 0.0483, "step": 37550 }, { "grad_norm": 0.6979755163192749, "learning_rate": 7.371152759587415e-05, "loss": 0.0398, "step": 37560 }, { "grad_norm": 0.48586592078208923, "learning_rate": 7.369696913984572e-05, "loss": 0.0414, "step": 37570 }, { "grad_norm": 0.34830522537231445, "learning_rate": 7.368240809235237e-05, "loss": 0.0367, "step": 37580 }, { "grad_norm": 0.41615232825279236, "learning_rate": 7.366784445498649e-05, "loss": 0.0355, "step": 37590 }, { "grad_norm": 1.2364673614501953, "learning_rate": 7.36532782293407e-05, "loss": 0.0446, "step": 37600 }, { "grad_norm": 0.7232776880264282, "learning_rate": 7.363870941700797e-05, "loss": 0.0433, "step": 37610 }, { "grad_norm": 0.33231309056282043, "learning_rate": 7.362413801958152e-05, "loss": 0.0414, "step": 37620 }, { "grad_norm": 0.26894092559814453, "learning_rate": 7.360956403865486e-05, "loss": 0.0428, "step": 37630 }, { "grad_norm": 0.3356544077396393, "learning_rate": 7.359498747582177e-05, "loss": 0.0463, "step": 37640 }, { "grad_norm": 0.3607504367828369, "learning_rate": 7.358040833267634e-05, "loss": 0.0409, "step": 37650 }, { "grad_norm": 0.4632847309112549, "learning_rate": 7.356582661081288e-05, "loss": 0.0448, "step": 37660 }, { "grad_norm": 0.35537290573120117, "learning_rate": 7.355124231182607e-05, "loss": 0.0419, "step": 37670 }, { "grad_norm": 0.2608882486820221, "learning_rate": 7.353665543731083e-05, "loss": 0.0418, "step": 37680 }, { "grad_norm": 0.4240745007991791, "learning_rate": 7.352206598886229e-05, "loss": 0.0408, "step": 37690 }, { "grad_norm": 0.3384249806404114, "learning_rate": 7.350747396807601e-05, "loss": 0.0385, "step": 37700 }, { "grad_norm": 0.31891950964927673, "learning_rate": 7.349287937654772e-05, "loss": 0.0393, "step": 37710 }, { "grad_norm": 0.5521169900894165, "learning_rate": 7.347828221587345e-05, "loss": 0.0456, "step": 37720 }, { "grad_norm": 0.3650643527507782, "learning_rate": 7.346368248764955e-05, "loss": 0.0402, "step": 37730 }, { "grad_norm": 0.4332800805568695, "learning_rate": 7.34490801934726e-05, "loss": 0.0393, "step": 37740 }, { "grad_norm": 0.39949169754981995, "learning_rate": 7.343447533493947e-05, "loss": 0.0493, "step": 37750 }, { "grad_norm": 0.539074718952179, "learning_rate": 7.341986791364739e-05, "loss": 0.0428, "step": 37760 }, { "grad_norm": 0.5433408617973328, "learning_rate": 7.340525793119374e-05, "loss": 0.0465, "step": 37770 }, { "grad_norm": 0.4041367769241333, "learning_rate": 7.33906453891763e-05, "loss": 0.0455, "step": 37780 }, { "grad_norm": 0.3111532926559448, "learning_rate": 7.337603028919303e-05, "loss": 0.0412, "step": 37790 }, { "grad_norm": 0.3134002089500427, "learning_rate": 7.336141263284225e-05, "loss": 0.042, "step": 37800 }, { "grad_norm": 0.31741610169410706, "learning_rate": 7.334679242172249e-05, "loss": 0.0379, "step": 37810 }, { "grad_norm": 0.2678135335445404, "learning_rate": 7.33321696574326e-05, "loss": 0.0416, "step": 37820 }, { "grad_norm": 0.37454304099082947, "learning_rate": 7.331754434157174e-05, "loss": 0.0409, "step": 37830 }, { "grad_norm": 0.40861186385154724, "learning_rate": 7.330291647573928e-05, "loss": 0.0405, "step": 37840 }, { "grad_norm": 0.35338467359542847, "learning_rate": 7.32882860615349e-05, "loss": 0.0384, "step": 37850 }, { "grad_norm": 0.36878472566604614, "learning_rate": 7.327365310055858e-05, "loss": 0.0452, "step": 37860 }, { "grad_norm": 0.33411917090415955, "learning_rate": 7.325901759441055e-05, "loss": 0.0368, "step": 37870 }, { "grad_norm": 0.3109825551509857, "learning_rate": 7.324437954469132e-05, "loss": 0.0467, "step": 37880 }, { "grad_norm": 0.5614144802093506, "learning_rate": 7.32297389530017e-05, "loss": 0.0444, "step": 37890 }, { "grad_norm": 0.30196893215179443, "learning_rate": 7.321509582094274e-05, "loss": 0.0369, "step": 37900 }, { "grad_norm": 0.40487775206565857, "learning_rate": 7.320045015011582e-05, "loss": 0.041, "step": 37910 }, { "grad_norm": 0.43993982672691345, "learning_rate": 7.318580194212253e-05, "loss": 0.0457, "step": 37920 }, { "grad_norm": 0.7554375529289246, "learning_rate": 7.317115119856483e-05, "loss": 0.0366, "step": 37930 }, { "grad_norm": 0.6537072062492371, "learning_rate": 7.315649792104485e-05, "loss": 0.0464, "step": 37940 }, { "grad_norm": 0.42202088236808777, "learning_rate": 7.314184211116507e-05, "loss": 0.0444, "step": 37950 }, { "grad_norm": 0.36749130487442017, "learning_rate": 7.312718377052823e-05, "loss": 0.0366, "step": 37960 }, { "grad_norm": 0.6100521683692932, "learning_rate": 7.311252290073737e-05, "loss": 0.0449, "step": 37970 }, { "grad_norm": 0.46319493651390076, "learning_rate": 7.309785950339572e-05, "loss": 0.0429, "step": 37980 }, { "grad_norm": 0.38312026858329773, "learning_rate": 7.308319358010692e-05, "loss": 0.0438, "step": 37990 }, { "grad_norm": 0.37902334332466125, "learning_rate": 7.306852513247474e-05, "loss": 0.0407, "step": 38000 }, { "grad_norm": 0.45476454496383667, "learning_rate": 7.305385416210336e-05, "loss": 0.0457, "step": 38010 }, { "grad_norm": 0.40710893273353577, "learning_rate": 7.303918067059711e-05, "loss": 0.0433, "step": 38020 }, { "grad_norm": 0.3185909390449524, "learning_rate": 7.302450465956075e-05, "loss": 0.0394, "step": 38030 }, { "grad_norm": 0.41586798429489136, "learning_rate": 7.300982613059914e-05, "loss": 0.0455, "step": 38040 }, { "grad_norm": 0.31124308705329895, "learning_rate": 7.299514508531757e-05, "loss": 0.0476, "step": 38050 }, { "grad_norm": 0.5125386714935303, "learning_rate": 7.298046152532148e-05, "loss": 0.0453, "step": 38060 }, { "grad_norm": 0.2980075776576996, "learning_rate": 7.296577545221668e-05, "loss": 0.0452, "step": 38070 }, { "grad_norm": 0.38131746649742126, "learning_rate": 7.29510868676092e-05, "loss": 0.0443, "step": 38080 }, { "grad_norm": 0.39451029896736145, "learning_rate": 7.293639577310538e-05, "loss": 0.0434, "step": 38090 }, { "grad_norm": 0.31710827350616455, "learning_rate": 7.292170217031179e-05, "loss": 0.0407, "step": 38100 }, { "grad_norm": 0.44106727838516235, "learning_rate": 7.290700606083532e-05, "loss": 0.0416, "step": 38110 }, { "grad_norm": 0.34926754236221313, "learning_rate": 7.28923074462831e-05, "loss": 0.0368, "step": 38120 }, { "grad_norm": 0.27160847187042236, "learning_rate": 7.287760632826255e-05, "loss": 0.0387, "step": 38130 }, { "grad_norm": 0.3502156138420105, "learning_rate": 7.286290270838139e-05, "loss": 0.0484, "step": 38140 }, { "grad_norm": 0.452069491147995, "learning_rate": 7.284819658824756e-05, "loss": 0.047, "step": 38150 }, { "grad_norm": 0.48846787214279175, "learning_rate": 7.283348796946929e-05, "loss": 0.0463, "step": 38160 }, { "grad_norm": 0.364576518535614, "learning_rate": 7.28187768536551e-05, "loss": 0.0431, "step": 38170 }, { "grad_norm": 0.40136444568634033, "learning_rate": 7.280406324241379e-05, "loss": 0.0398, "step": 38180 }, { "grad_norm": 0.3749907910823822, "learning_rate": 7.278934713735441e-05, "loss": 0.0408, "step": 38190 }, { "grad_norm": 0.3942352533340454, "learning_rate": 7.277462854008629e-05, "loss": 0.0437, "step": 38200 }, { "grad_norm": 0.45693156123161316, "learning_rate": 7.275990745221904e-05, "loss": 0.0373, "step": 38210 }, { "grad_norm": 0.5245581865310669, "learning_rate": 7.274518387536252e-05, "loss": 0.044, "step": 38220 }, { "grad_norm": 0.364086389541626, "learning_rate": 7.273045781112689e-05, "loss": 0.0423, "step": 38230 }, { "grad_norm": 0.3292139768600464, "learning_rate": 7.271572926112257e-05, "loss": 0.0371, "step": 38240 }, { "grad_norm": 0.339446097612381, "learning_rate": 7.270099822696024e-05, "loss": 0.0442, "step": 38250 }, { "grad_norm": 0.3586001396179199, "learning_rate": 7.26862647102509e-05, "loss": 0.0386, "step": 38260 }, { "grad_norm": 0.343843013048172, "learning_rate": 7.267152871260573e-05, "loss": 0.043, "step": 38270 }, { "grad_norm": 0.3761252760887146, "learning_rate": 7.265679023563629e-05, "loss": 0.0468, "step": 38280 }, { "grad_norm": 0.4046306014060974, "learning_rate": 7.264204928095433e-05, "loss": 0.0429, "step": 38290 }, { "grad_norm": 0.45000582933425903, "learning_rate": 7.262730585017188e-05, "loss": 0.0427, "step": 38300 }, { "grad_norm": 0.4055558145046234, "learning_rate": 7.261255994490131e-05, "loss": 0.0418, "step": 38310 }, { "grad_norm": 0.5310707092285156, "learning_rate": 7.259781156675517e-05, "loss": 0.0404, "step": 38320 }, { "grad_norm": 0.33516985177993774, "learning_rate": 7.258306071734633e-05, "loss": 0.0353, "step": 38330 }, { "grad_norm": 0.4245564341545105, "learning_rate": 7.256830739828792e-05, "loss": 0.0436, "step": 38340 }, { "grad_norm": 0.5021425485610962, "learning_rate": 7.255355161119336e-05, "loss": 0.045, "step": 38350 }, { "grad_norm": 0.3629514276981354, "learning_rate": 7.253879335767628e-05, "loss": 0.0444, "step": 38360 }, { "grad_norm": 0.46021074056625366, "learning_rate": 7.252403263935067e-05, "loss": 0.0424, "step": 38370 }, { "grad_norm": 0.40352663397789, "learning_rate": 7.25092694578307e-05, "loss": 0.0389, "step": 38380 }, { "grad_norm": 0.43910863995552063, "learning_rate": 7.249450381473087e-05, "loss": 0.0431, "step": 38390 }, { "grad_norm": 0.3339191973209381, "learning_rate": 7.247973571166593e-05, "loss": 0.0416, "step": 38400 }, { "grad_norm": 0.2788461744785309, "learning_rate": 7.246496515025089e-05, "loss": 0.039, "step": 38410 }, { "grad_norm": 0.43863800168037415, "learning_rate": 7.245019213210101e-05, "loss": 0.0401, "step": 38420 }, { "grad_norm": 0.3265659213066101, "learning_rate": 7.24354166588319e-05, "loss": 0.0388, "step": 38430 }, { "grad_norm": 0.41357555985450745, "learning_rate": 7.242063873205935e-05, "loss": 0.0365, "step": 38440 }, { "grad_norm": 0.39653053879737854, "learning_rate": 7.240585835339946e-05, "loss": 0.0415, "step": 38450 }, { "grad_norm": 0.40457862615585327, "learning_rate": 7.239107552446857e-05, "loss": 0.0409, "step": 38460 }, { "grad_norm": 0.39213064312934875, "learning_rate": 7.237629024688334e-05, "loss": 0.039, "step": 38470 }, { "grad_norm": 0.42940056324005127, "learning_rate": 7.236150252226064e-05, "loss": 0.0393, "step": 38480 }, { "grad_norm": 0.45727112889289856, "learning_rate": 7.234671235221765e-05, "loss": 0.047, "step": 38490 }, { "grad_norm": 0.6249489188194275, "learning_rate": 7.233191973837179e-05, "loss": 0.0447, "step": 38500 }, { "grad_norm": 0.4035446047782898, "learning_rate": 7.231712468234078e-05, "loss": 0.0431, "step": 38510 }, { "grad_norm": 0.5289670825004578, "learning_rate": 7.230232718574254e-05, "loss": 0.0415, "step": 38520 }, { "grad_norm": 0.4586329758167267, "learning_rate": 7.228752725019535e-05, "loss": 0.0489, "step": 38530 }, { "grad_norm": 0.2976156175136566, "learning_rate": 7.227272487731769e-05, "loss": 0.0348, "step": 38540 }, { "grad_norm": 0.37943077087402344, "learning_rate": 7.225792006872831e-05, "loss": 0.0433, "step": 38550 }, { "grad_norm": 0.3379361629486084, "learning_rate": 7.224311282604628e-05, "loss": 0.0439, "step": 38560 }, { "grad_norm": 0.42898187041282654, "learning_rate": 7.222830315089085e-05, "loss": 0.0438, "step": 38570 }, { "grad_norm": 0.48775139451026917, "learning_rate": 7.22134910448816e-05, "loss": 0.0439, "step": 38580 }, { "grad_norm": 0.33680644631385803, "learning_rate": 7.219867650963839e-05, "loss": 0.0402, "step": 38590 }, { "grad_norm": 0.3995473384857178, "learning_rate": 7.21838595467813e-05, "loss": 0.0393, "step": 38600 }, { "grad_norm": 0.4468727111816406, "learning_rate": 7.216904015793067e-05, "loss": 0.0376, "step": 38610 }, { "grad_norm": 0.3577767312526703, "learning_rate": 7.215421834470713e-05, "loss": 0.0391, "step": 38620 }, { "grad_norm": 0.32912102341651917, "learning_rate": 7.213939410873158e-05, "loss": 0.0377, "step": 38630 }, { "grad_norm": 0.4207931160926819, "learning_rate": 7.212456745162519e-05, "loss": 0.0412, "step": 38640 }, { "grad_norm": 0.45655542612075806, "learning_rate": 7.210973837500937e-05, "loss": 0.0402, "step": 38650 }, { "grad_norm": 0.32967206835746765, "learning_rate": 7.209490688050578e-05, "loss": 0.04, "step": 38660 }, { "grad_norm": 0.3594740033149719, "learning_rate": 7.208007296973641e-05, "loss": 0.0457, "step": 38670 }, { "grad_norm": 0.38011112809181213, "learning_rate": 7.206523664432345e-05, "loss": 0.0332, "step": 38680 }, { "grad_norm": 0.30592644214630127, "learning_rate": 7.205039790588939e-05, "loss": 0.0407, "step": 38690 }, { "grad_norm": 0.4632819890975952, "learning_rate": 7.203555675605697e-05, "loss": 0.0392, "step": 38700 }, { "grad_norm": 0.34468090534210205, "learning_rate": 7.202071319644917e-05, "loss": 0.0358, "step": 38710 }, { "grad_norm": 0.37474629282951355, "learning_rate": 7.200586722868932e-05, "loss": 0.045, "step": 38720 }, { "grad_norm": 0.4489440619945526, "learning_rate": 7.199101885440088e-05, "loss": 0.0425, "step": 38730 }, { "grad_norm": 0.33450281620025635, "learning_rate": 7.197616807520771e-05, "loss": 0.0381, "step": 38740 }, { "grad_norm": 0.2971711754798889, "learning_rate": 7.196131489273381e-05, "loss": 0.0469, "step": 38750 }, { "grad_norm": 0.5601000189781189, "learning_rate": 7.194645930860355e-05, "loss": 0.0416, "step": 38760 }, { "grad_norm": 0.3592897355556488, "learning_rate": 7.19316013244415e-05, "loss": 0.0435, "step": 38770 }, { "grad_norm": 0.4132348895072937, "learning_rate": 7.191674094187248e-05, "loss": 0.0538, "step": 38780 }, { "grad_norm": 0.42517468333244324, "learning_rate": 7.190187816252165e-05, "loss": 0.0426, "step": 38790 }, { "grad_norm": 0.5255347490310669, "learning_rate": 7.188701298801435e-05, "loss": 0.0462, "step": 38800 }, { "grad_norm": 0.3712787628173828, "learning_rate": 7.18721454199762e-05, "loss": 0.0473, "step": 38810 }, { "grad_norm": 0.4488218426704407, "learning_rate": 7.185727546003312e-05, "loss": 0.0471, "step": 38820 }, { "grad_norm": 0.4024142324924469, "learning_rate": 7.184240310981126e-05, "loss": 0.036, "step": 38830 }, { "grad_norm": 0.49024567008018494, "learning_rate": 7.182752837093704e-05, "loss": 0.0369, "step": 38840 }, { "grad_norm": 0.4348437786102295, "learning_rate": 7.181265124503711e-05, "loss": 0.0428, "step": 38850 }, { "grad_norm": 0.4733641743659973, "learning_rate": 7.179777173373847e-05, "loss": 0.0373, "step": 38860 }, { "grad_norm": 0.41916587948799133, "learning_rate": 7.178288983866826e-05, "loss": 0.0381, "step": 38870 }, { "grad_norm": 0.3522491753101349, "learning_rate": 7.176800556145397e-05, "loss": 0.0368, "step": 38880 }, { "grad_norm": 0.3335578143596649, "learning_rate": 7.175311890372334e-05, "loss": 0.0397, "step": 38890 }, { "grad_norm": 0.4305399954319, "learning_rate": 7.17382298671043e-05, "loss": 0.0452, "step": 38900 }, { "grad_norm": 0.29349008202552795, "learning_rate": 7.172333845322515e-05, "loss": 0.0414, "step": 38910 }, { "grad_norm": 0.5244922041893005, "learning_rate": 7.170844466371436e-05, "loss": 0.0424, "step": 38920 }, { "grad_norm": 0.32388415932655334, "learning_rate": 7.16935485002007e-05, "loss": 0.0409, "step": 38930 }, { "grad_norm": 0.37361496686935425, "learning_rate": 7.167864996431319e-05, "loss": 0.039, "step": 38940 }, { "grad_norm": 0.35883885622024536, "learning_rate": 7.166374905768111e-05, "loss": 0.0398, "step": 38950 }, { "grad_norm": 0.3776291012763977, "learning_rate": 7.164884578193401e-05, "loss": 0.0442, "step": 38960 }, { "grad_norm": 0.46683499217033386, "learning_rate": 7.163394013870168e-05, "loss": 0.0401, "step": 38970 }, { "grad_norm": 0.3548029661178589, "learning_rate": 7.161903212961419e-05, "loss": 0.0419, "step": 38980 }, { "grad_norm": 0.5019399523735046, "learning_rate": 7.160412175630183e-05, "loss": 0.0393, "step": 38990 }, { "grad_norm": 0.3557341396808624, "learning_rate": 7.158920902039521e-05, "loss": 0.0359, "step": 39000 }, { "grad_norm": 0.30747556686401367, "learning_rate": 7.157429392352514e-05, "loss": 0.0427, "step": 39010 }, { "grad_norm": 0.36133334040641785, "learning_rate": 7.155937646732275e-05, "loss": 0.0475, "step": 39020 }, { "grad_norm": 0.3259209990501404, "learning_rate": 7.154445665341933e-05, "loss": 0.0398, "step": 39030 }, { "grad_norm": 0.38673344254493713, "learning_rate": 7.152953448344654e-05, "loss": 0.0456, "step": 39040 }, { "grad_norm": 0.29500654339790344, "learning_rate": 7.151460995903624e-05, "loss": 0.0412, "step": 39050 }, { "grad_norm": 0.5105418562889099, "learning_rate": 7.149968308182052e-05, "loss": 0.0413, "step": 39060 }, { "grad_norm": 0.3072485029697418, "learning_rate": 7.14847538534318e-05, "loss": 0.0378, "step": 39070 }, { "grad_norm": 0.36869361996650696, "learning_rate": 7.14698222755027e-05, "loss": 0.048, "step": 39080 }, { "grad_norm": 0.34594476222991943, "learning_rate": 7.14548883496661e-05, "loss": 0.0409, "step": 39090 }, { "grad_norm": 0.5202046036720276, "learning_rate": 7.143995207755517e-05, "loss": 0.0431, "step": 39100 }, { "grad_norm": 0.3648332357406616, "learning_rate": 7.142501346080333e-05, "loss": 0.0384, "step": 39110 }, { "grad_norm": 0.37185201048851013, "learning_rate": 7.141007250104421e-05, "loss": 0.0375, "step": 39120 }, { "grad_norm": 0.3641622066497803, "learning_rate": 7.139512919991176e-05, "loss": 0.0378, "step": 39130 }, { "grad_norm": 0.3195621371269226, "learning_rate": 7.138018355904014e-05, "loss": 0.0421, "step": 39140 }, { "grad_norm": 0.4049971103668213, "learning_rate": 7.13652355800638e-05, "loss": 0.0364, "step": 39150 }, { "grad_norm": 0.2766962945461273, "learning_rate": 7.135028526461743e-05, "loss": 0.0401, "step": 39160 }, { "grad_norm": 0.3006713092327118, "learning_rate": 7.133533261433594e-05, "loss": 0.039, "step": 39170 }, { "grad_norm": 0.381283164024353, "learning_rate": 7.132037763085457e-05, "loss": 0.0448, "step": 39180 }, { "grad_norm": 0.32701581716537476, "learning_rate": 7.130542031580875e-05, "loss": 0.0455, "step": 39190 }, { "grad_norm": 0.5014355778694153, "learning_rate": 7.12904606708342e-05, "loss": 0.0442, "step": 39200 }, { "grad_norm": 0.4767032265663147, "learning_rate": 7.127549869756687e-05, "loss": 0.0414, "step": 39210 }, { "grad_norm": 0.49586978554725647, "learning_rate": 7.126053439764299e-05, "loss": 0.0414, "step": 39220 }, { "grad_norm": 0.32809117436408997, "learning_rate": 7.124556777269904e-05, "loss": 0.0407, "step": 39230 }, { "grad_norm": 0.7104776501655579, "learning_rate": 7.123059882437174e-05, "loss": 0.0515, "step": 39240 }, { "grad_norm": 0.4170003831386566, "learning_rate": 7.121562755429807e-05, "loss": 0.0396, "step": 39250 }, { "grad_norm": 0.4054886996746063, "learning_rate": 7.120065396411528e-05, "loss": 0.036, "step": 39260 }, { "grad_norm": 0.48169994354248047, "learning_rate": 7.118567805546084e-05, "loss": 0.0472, "step": 39270 }, { "grad_norm": 0.4958599805831909, "learning_rate": 7.117069982997248e-05, "loss": 0.0489, "step": 39280 }, { "grad_norm": 0.4482018053531647, "learning_rate": 7.115571928928825e-05, "loss": 0.0452, "step": 39290 }, { "grad_norm": 0.5198232531547546, "learning_rate": 7.114073643504635e-05, "loss": 0.0432, "step": 39300 }, { "grad_norm": 0.42601969838142395, "learning_rate": 7.11257512688853e-05, "loss": 0.0432, "step": 39310 }, { "grad_norm": 0.41782286763191223, "learning_rate": 7.111076379244384e-05, "loss": 0.0405, "step": 39320 }, { "grad_norm": 1.297874093055725, "learning_rate": 7.109577400736101e-05, "loss": 0.0403, "step": 39330 }, { "grad_norm": 0.5762768387794495, "learning_rate": 7.108078191527605e-05, "loss": 0.0513, "step": 39340 }, { "grad_norm": 0.7398653030395508, "learning_rate": 7.106578751782847e-05, "loss": 0.0451, "step": 39350 }, { "grad_norm": 1.462012529373169, "learning_rate": 7.105079081665803e-05, "loss": 0.0413, "step": 39360 }, { "grad_norm": 0.3428683578968048, "learning_rate": 7.103579181340476e-05, "loss": 0.0456, "step": 39370 }, { "grad_norm": 0.3524932265281677, "learning_rate": 7.102079050970893e-05, "loss": 0.0439, "step": 39380 }, { "grad_norm": 0.33539965748786926, "learning_rate": 7.100578690721105e-05, "loss": 0.0382, "step": 39390 }, { "grad_norm": 0.3600863516330719, "learning_rate": 7.09907810075519e-05, "loss": 0.0356, "step": 39400 }, { "grad_norm": 0.3767688572406769, "learning_rate": 7.097577281237249e-05, "loss": 0.0402, "step": 39410 }, { "grad_norm": 0.5478448867797852, "learning_rate": 7.09607623233141e-05, "loss": 0.0428, "step": 39420 }, { "grad_norm": 0.32859593629837036, "learning_rate": 7.094574954201827e-05, "loss": 0.0394, "step": 39430 }, { "grad_norm": 0.30237653851509094, "learning_rate": 7.093073447012675e-05, "loss": 0.0425, "step": 39440 }, { "grad_norm": 0.39572227001190186, "learning_rate": 7.09157171092816e-05, "loss": 0.0465, "step": 39450 }, { "grad_norm": 0.3406713902950287, "learning_rate": 7.090069746112504e-05, "loss": 0.0416, "step": 39460 }, { "grad_norm": 0.3637121319770813, "learning_rate": 7.088567552729965e-05, "loss": 0.0407, "step": 39470 }, { "grad_norm": 0.4045642614364624, "learning_rate": 7.087065130944818e-05, "loss": 0.0332, "step": 39480 }, { "grad_norm": 0.43375733494758606, "learning_rate": 7.085562480921366e-05, "loss": 0.0381, "step": 39490 }, { "grad_norm": 0.256450891494751, "learning_rate": 7.084059602823937e-05, "loss": 0.0366, "step": 39500 }, { "grad_norm": 0.356682687997818, "learning_rate": 7.082556496816882e-05, "loss": 0.0343, "step": 39510 }, { "grad_norm": 0.40606051683425903, "learning_rate": 7.081053163064582e-05, "loss": 0.0374, "step": 39520 }, { "grad_norm": 0.34607356786727905, "learning_rate": 7.079549601731434e-05, "loss": 0.0462, "step": 39530 }, { "grad_norm": 0.4598497748374939, "learning_rate": 7.07804581298187e-05, "loss": 0.0378, "step": 39540 }, { "grad_norm": 0.32974013686180115, "learning_rate": 7.07654179698034e-05, "loss": 0.0452, "step": 39550 }, { "grad_norm": 0.24822521209716797, "learning_rate": 7.075037553891321e-05, "loss": 0.0403, "step": 39560 }, { "grad_norm": 0.47163575887680054, "learning_rate": 7.073533083879315e-05, "loss": 0.0402, "step": 39570 }, { "grad_norm": 0.48116031289100647, "learning_rate": 7.072028387108849e-05, "loss": 0.0432, "step": 39580 }, { "grad_norm": 0.46135178208351135, "learning_rate": 7.070523463744472e-05, "loss": 0.0395, "step": 39590 }, { "grad_norm": 0.33716845512390137, "learning_rate": 7.069018313950763e-05, "loss": 0.0358, "step": 39600 }, { "grad_norm": 0.4938228726387024, "learning_rate": 7.067512937892322e-05, "loss": 0.0413, "step": 39610 }, { "grad_norm": 0.2720371186733246, "learning_rate": 7.066007335733775e-05, "loss": 0.0411, "step": 39620 }, { "grad_norm": 0.797909140586853, "learning_rate": 7.064501507639772e-05, "loss": 0.0403, "step": 39630 }, { "grad_norm": 0.28909289836883545, "learning_rate": 7.062995453774987e-05, "loss": 0.0377, "step": 39640 }, { "grad_norm": 0.330617219209671, "learning_rate": 7.061489174304121e-05, "loss": 0.0375, "step": 39650 }, { "grad_norm": 0.5918197631835938, "learning_rate": 7.0599826693919e-05, "loss": 0.0414, "step": 39660 }, { "grad_norm": 0.36662569642066956, "learning_rate": 7.05847593920307e-05, "loss": 0.0445, "step": 39670 }, { "grad_norm": 0.31079062819480896, "learning_rate": 7.056968983902406e-05, "loss": 0.0394, "step": 39680 }, { "grad_norm": 0.4569682776927948, "learning_rate": 7.055461803654706e-05, "loss": 0.0365, "step": 39690 }, { "grad_norm": 0.3172385096549988, "learning_rate": 7.053954398624794e-05, "loss": 0.0432, "step": 39700 }, { "grad_norm": 0.28432390093803406, "learning_rate": 7.052446768977518e-05, "loss": 0.0367, "step": 39710 }, { "grad_norm": 0.4647182524204254, "learning_rate": 7.050938914877748e-05, "loss": 0.0411, "step": 39720 }, { "grad_norm": 0.30227217078208923, "learning_rate": 7.049430836490382e-05, "loss": 0.0427, "step": 39730 }, { "grad_norm": 0.41686686873435974, "learning_rate": 7.047922533980342e-05, "loss": 0.047, "step": 39740 }, { "grad_norm": 0.4196765124797821, "learning_rate": 7.046414007512571e-05, "loss": 0.0453, "step": 39750 }, { "grad_norm": 0.428408145904541, "learning_rate": 7.044905257252042e-05, "loss": 0.0454, "step": 39760 }, { "grad_norm": 0.34202873706817627, "learning_rate": 7.04339628336375e-05, "loss": 0.0366, "step": 39770 }, { "grad_norm": 0.539326012134552, "learning_rate": 7.041887086012711e-05, "loss": 0.0424, "step": 39780 }, { "grad_norm": 0.37918415665626526, "learning_rate": 7.040377665363969e-05, "loss": 0.0358, "step": 39790 }, { "grad_norm": 0.4337323307991028, "learning_rate": 7.038868021582594e-05, "loss": 0.0462, "step": 39800 }, { "grad_norm": 0.3934783339500427, "learning_rate": 7.037358154833679e-05, "loss": 0.0409, "step": 39810 }, { "grad_norm": 0.4622700810432434, "learning_rate": 7.035848065282339e-05, "loss": 0.0542, "step": 39820 }, { "grad_norm": 0.44703859090805054, "learning_rate": 7.034337753093714e-05, "loss": 0.0427, "step": 39830 }, { "grad_norm": 0.64165860414505, "learning_rate": 7.032827218432972e-05, "loss": 0.0405, "step": 39840 }, { "grad_norm": 0.42055413126945496, "learning_rate": 7.031316461465302e-05, "loss": 0.0432, "step": 39850 }, { "grad_norm": 0.3182365894317627, "learning_rate": 7.029805482355915e-05, "loss": 0.0378, "step": 39860 }, { "grad_norm": 0.35507458448410034, "learning_rate": 7.028294281270055e-05, "loss": 0.034, "step": 39870 }, { "grad_norm": 0.3309671878814697, "learning_rate": 7.026782858372982e-05, "loss": 0.0377, "step": 39880 }, { "grad_norm": 0.6500388979911804, "learning_rate": 7.025271213829982e-05, "loss": 0.0353, "step": 39890 }, { "grad_norm": 0.4257603585720062, "learning_rate": 7.023759347806366e-05, "loss": 0.0408, "step": 39900 }, { "grad_norm": 0.7358371019363403, "learning_rate": 7.022247260467472e-05, "loss": 0.0414, "step": 39910 }, { "grad_norm": 0.3839170038700104, "learning_rate": 7.020734951978658e-05, "loss": 0.0394, "step": 39920 }, { "grad_norm": 0.2360369712114334, "learning_rate": 7.019222422505307e-05, "loss": 0.0461, "step": 39930 }, { "grad_norm": 0.4781964421272278, "learning_rate": 7.017709672212827e-05, "loss": 0.0469, "step": 39940 }, { "grad_norm": 0.36366578936576843, "learning_rate": 7.016196701266652e-05, "loss": 0.0399, "step": 39950 }, { "grad_norm": 0.4490617513656616, "learning_rate": 7.014683509832238e-05, "loss": 0.042, "step": 39960 }, { "grad_norm": 0.3933566212654114, "learning_rate": 7.013170098075063e-05, "loss": 0.0463, "step": 39970 }, { "grad_norm": 0.3517151176929474, "learning_rate": 7.011656466160632e-05, "loss": 0.0359, "step": 39980 }, { "grad_norm": 0.32504865527153015, "learning_rate": 7.010142614254475e-05, "loss": 0.0403, "step": 39990 }, { "grad_norm": 2.5022778511047363, "learning_rate": 7.008628542522147e-05, "loss": 0.0395, "step": 40000 }, { "grad_norm": 0.40047913789749146, "learning_rate": 7.007114251129218e-05, "loss": 0.0496, "step": 40010 }, { "grad_norm": 0.5263808965682983, "learning_rate": 7.005599740241293e-05, "loss": 0.0453, "step": 40020 }, { "grad_norm": 0.35914549231529236, "learning_rate": 7.004085010023996e-05, "loss": 0.0488, "step": 40030 }, { "grad_norm": 0.3817315101623535, "learning_rate": 7.002570060642976e-05, "loss": 0.0448, "step": 40040 }, { "grad_norm": 0.39999380707740784, "learning_rate": 7.001054892263903e-05, "loss": 0.0434, "step": 40050 }, { "grad_norm": 0.3842645287513733, "learning_rate": 6.999539505052477e-05, "loss": 0.0415, "step": 40060 }, { "grad_norm": 0.29902708530426025, "learning_rate": 6.998023899174418e-05, "loss": 0.0379, "step": 40070 }, { "grad_norm": 0.4288756549358368, "learning_rate": 6.996508074795467e-05, "loss": 0.0379, "step": 40080 }, { "grad_norm": 0.4167344570159912, "learning_rate": 6.994992032081396e-05, "loss": 0.0428, "step": 40090 }, { "grad_norm": 0.34495672583580017, "learning_rate": 6.993475771197995e-05, "loss": 0.0426, "step": 40100 }, { "grad_norm": 0.5097142457962036, "learning_rate": 6.991959292311082e-05, "loss": 0.0349, "step": 40110 }, { "grad_norm": 0.2783566415309906, "learning_rate": 6.990442595586495e-05, "loss": 0.0373, "step": 40120 }, { "grad_norm": 0.3924000561237335, "learning_rate": 6.988925681190098e-05, "loss": 0.0425, "step": 40130 }, { "grad_norm": 0.3018556833267212, "learning_rate": 6.987408549287778e-05, "loss": 0.042, "step": 40140 }, { "grad_norm": 0.34517401456832886, "learning_rate": 6.985891200045449e-05, "loss": 0.0389, "step": 40150 }, { "grad_norm": 0.4452323615550995, "learning_rate": 6.984373633629045e-05, "loss": 0.0342, "step": 40160 }, { "grad_norm": 0.4575531482696533, "learning_rate": 6.98285585020452e-05, "loss": 0.0484, "step": 40170 }, { "grad_norm": 0.503734827041626, "learning_rate": 6.981337849937864e-05, "loss": 0.0397, "step": 40180 }, { "grad_norm": 0.3881433308124542, "learning_rate": 6.979819632995078e-05, "loss": 0.0407, "step": 40190 }, { "grad_norm": 0.4077278673648834, "learning_rate": 6.978301199542193e-05, "loss": 0.0515, "step": 40200 }, { "grad_norm": 0.37160494923591614, "learning_rate": 6.976782549745263e-05, "loss": 0.049, "step": 40210 }, { "grad_norm": 0.4785003066062927, "learning_rate": 6.975263683770367e-05, "loss": 0.0445, "step": 40220 }, { "grad_norm": 0.4738521873950958, "learning_rate": 6.973744601783604e-05, "loss": 0.042, "step": 40230 }, { "grad_norm": 0.3493054509162903, "learning_rate": 6.972225303951097e-05, "loss": 0.0427, "step": 40240 }, { "grad_norm": 0.32070332765579224, "learning_rate": 6.970705790438998e-05, "loss": 0.0458, "step": 40250 }, { "grad_norm": 0.3665044605731964, "learning_rate": 6.969186061413477e-05, "loss": 0.0394, "step": 40260 }, { "grad_norm": 0.5100756883621216, "learning_rate": 6.967666117040727e-05, "loss": 0.0438, "step": 40270 }, { "grad_norm": 0.33811140060424805, "learning_rate": 6.966145957486972e-05, "loss": 0.0408, "step": 40280 }, { "grad_norm": 0.3568381667137146, "learning_rate": 6.964625582918449e-05, "loss": 0.0372, "step": 40290 }, { "grad_norm": 0.4259260296821594, "learning_rate": 6.963104993501425e-05, "loss": 0.0485, "step": 40300 }, { "grad_norm": 0.4375740885734558, "learning_rate": 6.961584189402192e-05, "loss": 0.0386, "step": 40310 }, { "grad_norm": 0.33958137035369873, "learning_rate": 6.96006317078706e-05, "loss": 0.0449, "step": 40320 }, { "grad_norm": 0.33717969059944153, "learning_rate": 6.958541937822367e-05, "loss": 0.0375, "step": 40330 }, { "grad_norm": 0.4415762424468994, "learning_rate": 6.957020490674473e-05, "loss": 0.0424, "step": 40340 }, { "grad_norm": 0.5191157460212708, "learning_rate": 6.95549882950976e-05, "loss": 0.0354, "step": 40350 }, { "grad_norm": 0.40868043899536133, "learning_rate": 6.953976954494635e-05, "loss": 0.0409, "step": 40360 }, { "grad_norm": 0.2931416630744934, "learning_rate": 6.952454865795528e-05, "loss": 0.0415, "step": 40370 }, { "grad_norm": 0.37295180559158325, "learning_rate": 6.950932563578892e-05, "loss": 0.0395, "step": 40380 }, { "grad_norm": 0.4948538839817047, "learning_rate": 6.949410048011206e-05, "loss": 0.0381, "step": 40390 }, { "grad_norm": 0.3381229639053345, "learning_rate": 6.947887319258966e-05, "loss": 0.0397, "step": 40400 }, { "grad_norm": 1.1839755773544312, "learning_rate": 6.946364377488696e-05, "loss": 0.0409, "step": 40410 }, { "grad_norm": 0.559840977191925, "learning_rate": 6.944841222866947e-05, "loss": 0.0384, "step": 40420 }, { "grad_norm": 0.4400167763233185, "learning_rate": 6.943317855560284e-05, "loss": 0.0402, "step": 40430 }, { "grad_norm": 0.3670082986354828, "learning_rate": 6.941794275735302e-05, "loss": 0.05, "step": 40440 }, { "grad_norm": 0.328407883644104, "learning_rate": 6.94027048355862e-05, "loss": 0.0399, "step": 40450 }, { "grad_norm": 0.3063327670097351, "learning_rate": 6.938746479196871e-05, "loss": 0.0413, "step": 40460 }, { "grad_norm": 0.2931089997291565, "learning_rate": 6.937222262816724e-05, "loss": 0.0322, "step": 40470 }, { "grad_norm": 0.4028562605381012, "learning_rate": 6.935697834584865e-05, "loss": 0.0415, "step": 40480 }, { "grad_norm": 0.5447810292243958, "learning_rate": 6.934173194667998e-05, "loss": 0.044, "step": 40490 }, { "grad_norm": 0.3551020920276642, "learning_rate": 6.93264834323286e-05, "loss": 0.0411, "step": 40500 }, { "grad_norm": 0.34322935342788696, "learning_rate": 6.931123280446206e-05, "loss": 0.0378, "step": 40510 }, { "grad_norm": 0.3667369782924652, "learning_rate": 6.929598006474811e-05, "loss": 0.0396, "step": 40520 }, { "grad_norm": 0.5295559167861938, "learning_rate": 6.92807252148548e-05, "loss": 0.0427, "step": 40530 }, { "grad_norm": 0.36469539999961853, "learning_rate": 6.926546825645037e-05, "loss": 0.0368, "step": 40540 }, { "grad_norm": 0.40452802181243896, "learning_rate": 6.92502091912033e-05, "loss": 0.0353, "step": 40550 }, { "grad_norm": 0.3714604079723358, "learning_rate": 6.923494802078229e-05, "loss": 0.0366, "step": 40560 }, { "grad_norm": 0.5612507462501526, "learning_rate": 6.921968474685628e-05, "loss": 0.0354, "step": 40570 }, { "grad_norm": 0.8121991753578186, "learning_rate": 6.920441937109445e-05, "loss": 0.0406, "step": 40580 }, { "grad_norm": 0.40932196378707886, "learning_rate": 6.91891518951662e-05, "loss": 0.038, "step": 40590 }, { "grad_norm": 0.31991130113601685, "learning_rate": 6.917388232074114e-05, "loss": 0.0461, "step": 40600 }, { "grad_norm": 0.36679574847221375, "learning_rate": 6.915861064948914e-05, "loss": 0.0475, "step": 40610 }, { "grad_norm": 0.2966996133327484, "learning_rate": 6.914333688308025e-05, "loss": 0.0447, "step": 40620 }, { "grad_norm": 0.4232980012893677, "learning_rate": 6.912806102318487e-05, "loss": 0.0421, "step": 40630 }, { "grad_norm": 0.46709564328193665, "learning_rate": 6.911278307147347e-05, "loss": 0.0399, "step": 40640 }, { "grad_norm": 0.49355369806289673, "learning_rate": 6.909750302961684e-05, "loss": 0.0397, "step": 40650 }, { "grad_norm": 0.29956597089767456, "learning_rate": 6.9082220899286e-05, "loss": 0.0378, "step": 40660 }, { "grad_norm": 0.37383273243904114, "learning_rate": 6.906693668215219e-05, "loss": 0.0401, "step": 40670 }, { "grad_norm": 0.40009257197380066, "learning_rate": 6.905165037988683e-05, "loss": 0.0369, "step": 40680 }, { "grad_norm": 0.40890806913375854, "learning_rate": 6.903636199416164e-05, "loss": 0.0387, "step": 40690 }, { "grad_norm": 0.4507765471935272, "learning_rate": 6.902107152664851e-05, "loss": 0.0444, "step": 40700 }, { "grad_norm": 0.4313422441482544, "learning_rate": 6.90057789790196e-05, "loss": 0.0434, "step": 40710 }, { "grad_norm": 0.3914858400821686, "learning_rate": 6.899048435294728e-05, "loss": 0.032, "step": 40720 }, { "grad_norm": 0.3186582326889038, "learning_rate": 6.897518765010415e-05, "loss": 0.036, "step": 40730 }, { "grad_norm": 0.3112695515155792, "learning_rate": 6.895988887216302e-05, "loss": 0.0363, "step": 40740 }, { "grad_norm": 0.4769490659236908, "learning_rate": 6.894458802079694e-05, "loss": 0.0392, "step": 40750 }, { "grad_norm": 0.3257695734500885, "learning_rate": 6.892928509767922e-05, "loss": 0.0465, "step": 40760 }, { "grad_norm": 0.5240976214408875, "learning_rate": 6.891398010448333e-05, "loss": 0.0455, "step": 40770 }, { "grad_norm": 0.39620840549468994, "learning_rate": 6.8898673042883e-05, "loss": 0.0406, "step": 40780 }, { "grad_norm": 0.4594935476779938, "learning_rate": 6.888336391455222e-05, "loss": 0.0335, "step": 40790 }, { "grad_norm": 0.35961979627609253, "learning_rate": 6.886805272116513e-05, "loss": 0.0503, "step": 40800 }, { "grad_norm": 0.680606484413147, "learning_rate": 6.885273946439617e-05, "loss": 0.044, "step": 40810 }, { "grad_norm": 0.8022782802581787, "learning_rate": 6.883742414591998e-05, "loss": 0.0356, "step": 40820 }, { "grad_norm": 0.3138004541397095, "learning_rate": 6.88221067674114e-05, "loss": 0.0354, "step": 40830 }, { "grad_norm": 0.44318699836730957, "learning_rate": 6.88067873305455e-05, "loss": 0.0411, "step": 40840 }, { "grad_norm": 0.38679105043411255, "learning_rate": 6.879146583699765e-05, "loss": 0.0389, "step": 40850 }, { "grad_norm": 0.4679839015007019, "learning_rate": 6.877614228844334e-05, "loss": 0.0459, "step": 40860 }, { "grad_norm": 0.45562514662742615, "learning_rate": 6.876081668655832e-05, "loss": 0.0383, "step": 40870 }, { "grad_norm": 0.41609930992126465, "learning_rate": 6.874548903301861e-05, "loss": 0.0444, "step": 40880 }, { "grad_norm": 0.4777633845806122, "learning_rate": 6.873015932950039e-05, "loss": 0.0447, "step": 40890 }, { "grad_norm": 0.47801798582077026, "learning_rate": 6.871482757768012e-05, "loss": 0.0429, "step": 40900 }, { "grad_norm": 0.3202449381351471, "learning_rate": 6.869949377923442e-05, "loss": 0.0439, "step": 40910 }, { "grad_norm": 0.30482202768325806, "learning_rate": 6.868415793584022e-05, "loss": 0.0434, "step": 40920 }, { "grad_norm": 0.3852873742580414, "learning_rate": 6.86688200491746e-05, "loss": 0.0403, "step": 40930 }, { "grad_norm": 0.4716496169567108, "learning_rate": 6.865348012091488e-05, "loss": 0.04, "step": 40940 }, { "grad_norm": 0.40346425771713257, "learning_rate": 6.86381381527386e-05, "loss": 0.0432, "step": 40950 }, { "grad_norm": 0.3644118905067444, "learning_rate": 6.862279414632359e-05, "loss": 0.0422, "step": 40960 }, { "grad_norm": 0.4780072569847107, "learning_rate": 6.860744810334778e-05, "loss": 0.0474, "step": 40970 }, { "grad_norm": 0.4282360076904297, "learning_rate": 6.859210002548943e-05, "loss": 0.0387, "step": 40980 }, { "grad_norm": 0.36026740074157715, "learning_rate": 6.8576749914427e-05, "loss": 0.0359, "step": 40990 }, { "grad_norm": 0.27681592106819153, "learning_rate": 6.85613977718391e-05, "loss": 0.0403, "step": 41000 }, { "grad_norm": 0.3372633755207062, "learning_rate": 6.854604359940464e-05, "loss": 0.0343, "step": 41010 }, { "grad_norm": 0.4833396375179291, "learning_rate": 6.853068739880276e-05, "loss": 0.0417, "step": 41020 }, { "grad_norm": 0.41820892691612244, "learning_rate": 6.851532917171276e-05, "loss": 0.047, "step": 41030 }, { "grad_norm": 0.5018616318702698, "learning_rate": 6.849996891981418e-05, "loss": 0.0379, "step": 41040 }, { "grad_norm": 0.37112534046173096, "learning_rate": 6.848460664478684e-05, "loss": 0.051, "step": 41050 }, { "grad_norm": 0.4192328453063965, "learning_rate": 6.846924234831069e-05, "loss": 0.0342, "step": 41060 }, { "grad_norm": 0.31552255153656006, "learning_rate": 6.845387603206597e-05, "loss": 0.0404, "step": 41070 }, { "grad_norm": 0.616126298904419, "learning_rate": 6.843850769773312e-05, "loss": 0.0411, "step": 41080 }, { "grad_norm": 0.512521505355835, "learning_rate": 6.842313734699278e-05, "loss": 0.0458, "step": 41090 }, { "grad_norm": 0.3969557583332062, "learning_rate": 6.840776498152584e-05, "loss": 0.0449, "step": 41100 }, { "grad_norm": 0.41056033968925476, "learning_rate": 6.83923906030134e-05, "loss": 0.0397, "step": 41110 }, { "grad_norm": 0.3759356439113617, "learning_rate": 6.837701421313677e-05, "loss": 0.0463, "step": 41120 }, { "grad_norm": 0.29979243874549866, "learning_rate": 6.83616358135775e-05, "loss": 0.0441, "step": 41130 }, { "grad_norm": 0.3834496736526489, "learning_rate": 6.834625540601734e-05, "loss": 0.0406, "step": 41140 }, { "grad_norm": 0.35495099425315857, "learning_rate": 6.833087299213829e-05, "loss": 0.0399, "step": 41150 }, { "grad_norm": 0.992669939994812, "learning_rate": 6.83154885736225e-05, "loss": 0.0466, "step": 41160 }, { "grad_norm": 0.3610515594482422, "learning_rate": 6.830010215215244e-05, "loss": 0.0421, "step": 41170 }, { "grad_norm": 0.5306662321090698, "learning_rate": 6.828471372941072e-05, "loss": 0.0389, "step": 41180 }, { "grad_norm": 0.3711506426334381, "learning_rate": 6.826932330708018e-05, "loss": 0.0384, "step": 41190 }, { "grad_norm": 0.5742736458778381, "learning_rate": 6.825393088684393e-05, "loss": 0.0459, "step": 41200 }, { "grad_norm": 0.3707432746887207, "learning_rate": 6.823853647038524e-05, "loss": 0.0444, "step": 41210 }, { "grad_norm": 0.33446258306503296, "learning_rate": 6.822314005938762e-05, "loss": 0.0368, "step": 41220 }, { "grad_norm": 0.3688582181930542, "learning_rate": 6.820774165553479e-05, "loss": 0.037, "step": 41230 }, { "grad_norm": 0.44666537642478943, "learning_rate": 6.819234126051072e-05, "loss": 0.0438, "step": 41240 }, { "grad_norm": 0.3826209306716919, "learning_rate": 6.817693887599956e-05, "loss": 0.0391, "step": 41250 }, { "grad_norm": 0.311166912317276, "learning_rate": 6.816153450368568e-05, "loss": 0.0414, "step": 41260 }, { "grad_norm": 0.41681969165802, "learning_rate": 6.814612814525372e-05, "loss": 0.0403, "step": 41270 }, { "grad_norm": 0.3517758250236511, "learning_rate": 6.813071980238846e-05, "loss": 0.0478, "step": 41280 }, { "grad_norm": 0.3709564507007599, "learning_rate": 6.811530947677492e-05, "loss": 0.043, "step": 41290 }, { "grad_norm": 0.3444688022136688, "learning_rate": 6.809989717009839e-05, "loss": 0.0432, "step": 41300 }, { "grad_norm": 0.39585641026496887, "learning_rate": 6.808448288404431e-05, "loss": 0.0449, "step": 41310 }, { "grad_norm": 0.3650799095630646, "learning_rate": 6.806906662029838e-05, "loss": 0.0416, "step": 41320 }, { "grad_norm": 0.34434646368026733, "learning_rate": 6.805364838054647e-05, "loss": 0.0464, "step": 41330 }, { "grad_norm": 0.40069758892059326, "learning_rate": 6.803822816647474e-05, "loss": 0.0379, "step": 41340 }, { "grad_norm": 0.37999606132507324, "learning_rate": 6.802280597976949e-05, "loss": 0.0358, "step": 41350 }, { "grad_norm": 0.5521848201751709, "learning_rate": 6.800738182211727e-05, "loss": 0.043, "step": 41360 }, { "grad_norm": 0.48246148228645325, "learning_rate": 6.799195569520485e-05, "loss": 0.0424, "step": 41370 }, { "grad_norm": 0.39718547463417053, "learning_rate": 6.79765276007192e-05, "loss": 0.0454, "step": 41380 }, { "grad_norm": 0.5493281483650208, "learning_rate": 6.796109754034753e-05, "loss": 0.0496, "step": 41390 }, { "grad_norm": 0.532223641872406, "learning_rate": 6.794566551577724e-05, "loss": 0.0429, "step": 41400 }, { "grad_norm": 0.39500510692596436, "learning_rate": 6.793023152869594e-05, "loss": 0.0376, "step": 41410 }, { "grad_norm": 0.3211822807788849, "learning_rate": 6.791479558079147e-05, "loss": 0.0414, "step": 41420 }, { "grad_norm": 0.4293495714664459, "learning_rate": 6.789935767375191e-05, "loss": 0.0424, "step": 41430 }, { "grad_norm": 0.4621383547782898, "learning_rate": 6.78839178092655e-05, "loss": 0.0379, "step": 41440 }, { "grad_norm": 0.5507915616035461, "learning_rate": 6.786847598902072e-05, "loss": 0.0429, "step": 41450 }, { "grad_norm": 0.36262834072113037, "learning_rate": 6.785303221470627e-05, "loss": 0.0398, "step": 41460 }, { "grad_norm": 0.42839667201042175, "learning_rate": 6.783758648801108e-05, "loss": 0.0421, "step": 41470 }, { "grad_norm": 0.3890976905822754, "learning_rate": 6.782213881062422e-05, "loss": 0.0424, "step": 41480 }, { "grad_norm": 0.4328518211841583, "learning_rate": 6.780668918423508e-05, "loss": 0.0412, "step": 41490 }, { "grad_norm": 0.40482327342033386, "learning_rate": 6.779123761053317e-05, "loss": 0.0341, "step": 41500 }, { "grad_norm": 0.3953509032726288, "learning_rate": 6.777578409120826e-05, "loss": 0.0425, "step": 41510 }, { "grad_norm": 0.3739493489265442, "learning_rate": 6.776032862795033e-05, "loss": 0.0436, "step": 41520 }, { "grad_norm": 0.45465198159217834, "learning_rate": 6.774487122244956e-05, "loss": 0.0501, "step": 41530 }, { "grad_norm": 0.44601091742515564, "learning_rate": 6.772941187639637e-05, "loss": 0.0433, "step": 41540 }, { "grad_norm": 0.3375217616558075, "learning_rate": 6.771395059148134e-05, "loss": 0.0372, "step": 41550 }, { "grad_norm": 0.4124508500099182, "learning_rate": 6.76984873693953e-05, "loss": 0.0399, "step": 41560 }, { "grad_norm": 0.41178271174430847, "learning_rate": 6.768302221182931e-05, "loss": 0.0342, "step": 41570 }, { "grad_norm": 0.4176803529262543, "learning_rate": 6.766755512047457e-05, "loss": 0.0343, "step": 41580 }, { "grad_norm": 0.37854158878326416, "learning_rate": 6.765208609702259e-05, "loss": 0.0431, "step": 41590 }, { "grad_norm": 0.4699508547782898, "learning_rate": 6.763661514316499e-05, "loss": 0.0403, "step": 41600 }, { "grad_norm": 0.38030797243118286, "learning_rate": 6.76211422605937e-05, "loss": 0.0389, "step": 41610 }, { "grad_norm": 0.3221041262149811, "learning_rate": 6.760566745100076e-05, "loss": 0.0398, "step": 41620 }, { "grad_norm": 0.3829304575920105, "learning_rate": 6.75901907160785e-05, "loss": 0.0386, "step": 41630 }, { "grad_norm": 0.24640622735023499, "learning_rate": 6.757471205751943e-05, "loss": 0.038, "step": 41640 }, { "grad_norm": 0.31028544902801514, "learning_rate": 6.75592314770163e-05, "loss": 0.0401, "step": 41650 }, { "grad_norm": 0.3734418451786041, "learning_rate": 6.7543748976262e-05, "loss": 0.0416, "step": 41660 }, { "grad_norm": 0.36760634183883667, "learning_rate": 6.752826455694968e-05, "loss": 0.0427, "step": 41670 }, { "grad_norm": 0.38539746403694153, "learning_rate": 6.751277822077271e-05, "loss": 0.0382, "step": 41680 }, { "grad_norm": 0.2612520158290863, "learning_rate": 6.749728996942465e-05, "loss": 0.0395, "step": 41690 }, { "grad_norm": 0.45403608679771423, "learning_rate": 6.748179980459924e-05, "loss": 0.0416, "step": 41700 }, { "grad_norm": 0.6515013575553894, "learning_rate": 6.746630772799052e-05, "loss": 0.0449, "step": 41710 }, { "grad_norm": 0.3885582685470581, "learning_rate": 6.745081374129262e-05, "loss": 0.0374, "step": 41720 }, { "grad_norm": 0.383969783782959, "learning_rate": 6.743531784619998e-05, "loss": 0.0394, "step": 41730 }, { "grad_norm": 0.3778354227542877, "learning_rate": 6.741982004440719e-05, "loss": 0.0431, "step": 41740 }, { "grad_norm": 0.2572868764400482, "learning_rate": 6.740432033760907e-05, "loss": 0.0364, "step": 41750 }, { "grad_norm": 0.33728235960006714, "learning_rate": 6.738881872750066e-05, "loss": 0.0449, "step": 41760 }, { "grad_norm": 0.3603743612766266, "learning_rate": 6.737331521577715e-05, "loss": 0.0374, "step": 41770 }, { "grad_norm": 0.4623841345310211, "learning_rate": 6.735780980413403e-05, "loss": 0.0431, "step": 41780 }, { "grad_norm": 0.4479014575481415, "learning_rate": 6.734230249426692e-05, "loss": 0.0381, "step": 41790 }, { "grad_norm": 0.36651256680488586, "learning_rate": 6.732679328787168e-05, "loss": 0.0374, "step": 41800 }, { "grad_norm": 0.48305371403694153, "learning_rate": 6.731128218664438e-05, "loss": 0.0435, "step": 41810 }, { "grad_norm": 0.41680851578712463, "learning_rate": 6.729576919228129e-05, "loss": 0.0395, "step": 41820 }, { "grad_norm": 0.6182191967964172, "learning_rate": 6.728025430647888e-05, "loss": 0.0479, "step": 41830 }, { "grad_norm": 0.583578884601593, "learning_rate": 6.726473753093383e-05, "loss": 0.0472, "step": 41840 }, { "grad_norm": 0.3530881702899933, "learning_rate": 6.724921886734305e-05, "loss": 0.0425, "step": 41850 }, { "grad_norm": 0.43671661615371704, "learning_rate": 6.723369831740363e-05, "loss": 0.0413, "step": 41860 }, { "grad_norm": 0.4130737781524658, "learning_rate": 6.721817588281288e-05, "loss": 0.0413, "step": 41870 }, { "grad_norm": 0.34514912962913513, "learning_rate": 6.720265156526828e-05, "loss": 0.0444, "step": 41880 }, { "grad_norm": 0.4756743609905243, "learning_rate": 6.718712536646758e-05, "loss": 0.0354, "step": 41890 }, { "grad_norm": 0.3801230788230896, "learning_rate": 6.71715972881087e-05, "loss": 0.0368, "step": 41900 }, { "grad_norm": 0.35418131947517395, "learning_rate": 6.715606733188974e-05, "loss": 0.0409, "step": 41910 }, { "grad_norm": 0.3869868516921997, "learning_rate": 6.714053549950909e-05, "loss": 0.0403, "step": 41920 }, { "grad_norm": 0.4624992609024048, "learning_rate": 6.71250017926652e-05, "loss": 0.0428, "step": 41930 }, { "grad_norm": 0.31791773438453674, "learning_rate": 6.71094662130569e-05, "loss": 0.0341, "step": 41940 }, { "grad_norm": 0.6193731427192688, "learning_rate": 6.709392876238307e-05, "loss": 0.0371, "step": 41950 }, { "grad_norm": 0.3047066330909729, "learning_rate": 6.707838944234289e-05, "loss": 0.041, "step": 41960 }, { "grad_norm": 0.3445495367050171, "learning_rate": 6.706284825463574e-05, "loss": 0.037, "step": 41970 }, { "grad_norm": 0.35437536239624023, "learning_rate": 6.704730520096114e-05, "loss": 0.0379, "step": 41980 }, { "grad_norm": 0.3807898163795471, "learning_rate": 6.703176028301888e-05, "loss": 0.0434, "step": 41990 }, { "grad_norm": 0.406320184469223, "learning_rate": 6.701621350250892e-05, "loss": 0.0423, "step": 42000 }, { "grad_norm": 0.3966005742549896, "learning_rate": 6.700066486113144e-05, "loss": 0.044, "step": 42010 }, { "grad_norm": 0.3658177852630615, "learning_rate": 6.69851143605868e-05, "loss": 0.0384, "step": 42020 }, { "grad_norm": 0.3971155881881714, "learning_rate": 6.696956200257561e-05, "loss": 0.0404, "step": 42030 }, { "grad_norm": 0.6653056740760803, "learning_rate": 6.695400778879863e-05, "loss": 0.0483, "step": 42040 }, { "grad_norm": 0.32487186789512634, "learning_rate": 6.693845172095683e-05, "loss": 0.0476, "step": 42050 }, { "grad_norm": 0.5003751516342163, "learning_rate": 6.692289380075142e-05, "loss": 0.0396, "step": 42060 }, { "grad_norm": 0.4319759011268616, "learning_rate": 6.690733402988379e-05, "loss": 0.039, "step": 42070 }, { "grad_norm": 0.31536009907722473, "learning_rate": 6.689177241005553e-05, "loss": 0.0378, "step": 42080 }, { "grad_norm": 0.3509129583835602, "learning_rate": 6.687620894296844e-05, "loss": 0.0436, "step": 42090 }, { "grad_norm": 0.2860470414161682, "learning_rate": 6.686064363032451e-05, "loss": 0.0467, "step": 42100 }, { "grad_norm": 0.4559667408466339, "learning_rate": 6.684507647382596e-05, "loss": 0.0334, "step": 42110 }, { "grad_norm": 0.3971175253391266, "learning_rate": 6.682950747517513e-05, "loss": 0.0356, "step": 42120 }, { "grad_norm": 0.42582836747169495, "learning_rate": 6.68139366360747e-05, "loss": 0.0426, "step": 42130 }, { "grad_norm": 0.4171934425830841, "learning_rate": 6.679836395822744e-05, "loss": 0.0368, "step": 42140 }, { "grad_norm": 0.45175281167030334, "learning_rate": 6.678278944333633e-05, "loss": 0.0416, "step": 42150 }, { "grad_norm": 0.5597723126411438, "learning_rate": 6.676721309310462e-05, "loss": 0.0334, "step": 42160 }, { "grad_norm": 0.3504655957221985, "learning_rate": 6.67516349092357e-05, "loss": 0.0325, "step": 42170 }, { "grad_norm": 0.3805023729801178, "learning_rate": 6.673605489343315e-05, "loss": 0.0416, "step": 42180 }, { "grad_norm": 0.4191930294036865, "learning_rate": 6.672047304740081e-05, "loss": 0.0356, "step": 42190 }, { "grad_norm": 0.3637944161891937, "learning_rate": 6.670488937284268e-05, "loss": 0.0384, "step": 42200 }, { "grad_norm": 0.31691625714302063, "learning_rate": 6.668930387146294e-05, "loss": 0.0361, "step": 42210 }, { "grad_norm": 0.296924889087677, "learning_rate": 6.667371654496605e-05, "loss": 0.0439, "step": 42220 }, { "grad_norm": 0.3965461254119873, "learning_rate": 6.665812739505659e-05, "loss": 0.0414, "step": 42230 }, { "grad_norm": 0.5951659679412842, "learning_rate": 6.664253642343935e-05, "loss": 0.0408, "step": 42240 }, { "grad_norm": 0.3342944085597992, "learning_rate": 6.662694363181935e-05, "loss": 0.0404, "step": 42250 }, { "grad_norm": 0.42432981729507446, "learning_rate": 6.66113490219018e-05, "loss": 0.0432, "step": 42260 }, { "grad_norm": 0.257863849401474, "learning_rate": 6.659575259539207e-05, "loss": 0.0339, "step": 42270 }, { "grad_norm": 0.4610397219657898, "learning_rate": 6.65801543539958e-05, "loss": 0.0428, "step": 42280 }, { "grad_norm": 0.6269466876983643, "learning_rate": 6.65645542994188e-05, "loss": 0.0472, "step": 42290 }, { "grad_norm": 0.3658585250377655, "learning_rate": 6.654895243336702e-05, "loss": 0.0392, "step": 42300 }, { "grad_norm": 0.3799441456794739, "learning_rate": 6.65333487575467e-05, "loss": 0.0382, "step": 42310 }, { "grad_norm": 0.4429280161857605, "learning_rate": 6.651774327366422e-05, "loss": 0.0381, "step": 42320 }, { "grad_norm": 0.4279770851135254, "learning_rate": 6.650213598342613e-05, "loss": 0.0343, "step": 42330 }, { "grad_norm": 0.5693111419677734, "learning_rate": 6.648652688853931e-05, "loss": 0.0401, "step": 42340 }, { "grad_norm": 0.45925331115722656, "learning_rate": 6.647091599071066e-05, "loss": 0.0436, "step": 42350 }, { "grad_norm": 0.35109877586364746, "learning_rate": 6.645530329164743e-05, "loss": 0.0395, "step": 42360 }, { "grad_norm": 0.3706180453300476, "learning_rate": 6.643968879305694e-05, "loss": 0.0517, "step": 42370 }, { "grad_norm": 0.3458664119243622, "learning_rate": 6.642407249664683e-05, "loss": 0.0423, "step": 42380 }, { "grad_norm": 0.3638364374637604, "learning_rate": 6.640845440412483e-05, "loss": 0.0334, "step": 42390 }, { "grad_norm": 0.2634250521659851, "learning_rate": 6.639283451719893e-05, "loss": 0.0418, "step": 42400 }, { "grad_norm": 0.4020487368106842, "learning_rate": 6.637721283757729e-05, "loss": 0.0449, "step": 42410 }, { "grad_norm": 0.46292638778686523, "learning_rate": 6.636158936696828e-05, "loss": 0.0389, "step": 42420 }, { "grad_norm": 0.40623757243156433, "learning_rate": 6.634596410708047e-05, "loss": 0.0431, "step": 42430 }, { "grad_norm": 0.4099200963973999, "learning_rate": 6.633033705962257e-05, "loss": 0.0379, "step": 42440 }, { "grad_norm": 0.3831731379032135, "learning_rate": 6.631470822630359e-05, "loss": 0.042, "step": 42450 }, { "grad_norm": 0.4905517101287842, "learning_rate": 6.629907760883263e-05, "loss": 0.0432, "step": 42460 }, { "grad_norm": 0.30179914832115173, "learning_rate": 6.628344520891905e-05, "loss": 0.0353, "step": 42470 }, { "grad_norm": 0.38075175881385803, "learning_rate": 6.62678110282724e-05, "loss": 0.0389, "step": 42480 }, { "grad_norm": 0.2700445353984833, "learning_rate": 6.625217506860237e-05, "loss": 0.0375, "step": 42490 }, { "grad_norm": 0.3593177795410156, "learning_rate": 6.623653733161892e-05, "loss": 0.0395, "step": 42500 }, { "grad_norm": 0.3672848343849182, "learning_rate": 6.622089781903217e-05, "loss": 0.0398, "step": 42510 }, { "grad_norm": 0.27551984786987305, "learning_rate": 6.620525653255241e-05, "loss": 0.0379, "step": 42520 }, { "grad_norm": 0.47626060247421265, "learning_rate": 6.618961347389015e-05, "loss": 0.0445, "step": 42530 }, { "grad_norm": 0.49273431301116943, "learning_rate": 6.617396864475613e-05, "loss": 0.042, "step": 42540 }, { "grad_norm": 0.4100935161113739, "learning_rate": 6.61583220468612e-05, "loss": 0.0389, "step": 42550 }, { "grad_norm": 0.3847631812095642, "learning_rate": 6.614267368191645e-05, "loss": 0.0408, "step": 42560 }, { "grad_norm": 0.3742464780807495, "learning_rate": 6.61270235516332e-05, "loss": 0.0333, "step": 42570 }, { "grad_norm": 0.3264952003955841, "learning_rate": 6.611137165772287e-05, "loss": 0.0395, "step": 42580 }, { "grad_norm": 0.3769221305847168, "learning_rate": 6.60957180018972e-05, "loss": 0.0365, "step": 42590 }, { "grad_norm": 0.3466988503932953, "learning_rate": 6.608006258586797e-05, "loss": 0.0389, "step": 42600 }, { "grad_norm": 0.5628595948219299, "learning_rate": 6.606440541134731e-05, "loss": 0.047, "step": 42610 }, { "grad_norm": 0.4787522554397583, "learning_rate": 6.60487464800474e-05, "loss": 0.0399, "step": 42620 }, { "grad_norm": 0.40073174238204956, "learning_rate": 6.603308579368071e-05, "loss": 0.0368, "step": 42630 }, { "grad_norm": 0.3602164685726166, "learning_rate": 6.601742335395987e-05, "loss": 0.0402, "step": 42640 }, { "grad_norm": 0.45692944526672363, "learning_rate": 6.600175916259769e-05, "loss": 0.0388, "step": 42650 }, { "grad_norm": 0.25770559906959534, "learning_rate": 6.598609322130718e-05, "loss": 0.0341, "step": 42660 }, { "grad_norm": 0.41298216581344604, "learning_rate": 6.597042553180155e-05, "loss": 0.0387, "step": 42670 }, { "grad_norm": 0.31661057472229004, "learning_rate": 6.59547560957942e-05, "loss": 0.0348, "step": 42680 }, { "grad_norm": 0.3600796163082123, "learning_rate": 6.59390849149987e-05, "loss": 0.0387, "step": 42690 }, { "grad_norm": 0.44591960310935974, "learning_rate": 6.592341199112886e-05, "loss": 0.0423, "step": 42700 }, { "grad_norm": 0.3339354395866394, "learning_rate": 6.590773732589861e-05, "loss": 0.0394, "step": 42710 }, { "grad_norm": 0.35128211975097656, "learning_rate": 6.589206092102213e-05, "loss": 0.0427, "step": 42720 }, { "grad_norm": 0.4758853018283844, "learning_rate": 6.587638277821375e-05, "loss": 0.0381, "step": 42730 }, { "grad_norm": 0.33781540393829346, "learning_rate": 6.586070289918803e-05, "loss": 0.0343, "step": 42740 }, { "grad_norm": 0.325623482465744, "learning_rate": 6.584502128565968e-05, "loss": 0.0367, "step": 42750 }, { "grad_norm": 0.5671786069869995, "learning_rate": 6.582933793934363e-05, "loss": 0.0404, "step": 42760 }, { "grad_norm": 0.35184210538864136, "learning_rate": 6.581365286195498e-05, "loss": 0.0368, "step": 42770 }, { "grad_norm": 0.39351406693458557, "learning_rate": 6.579796605520903e-05, "loss": 0.0342, "step": 42780 }, { "grad_norm": 0.36892062425613403, "learning_rate": 6.578227752082127e-05, "loss": 0.0384, "step": 42790 }, { "grad_norm": 0.3320011496543884, "learning_rate": 6.576658726050735e-05, "loss": 0.0429, "step": 42800 }, { "grad_norm": 0.3062589168548584, "learning_rate": 6.575089527598316e-05, "loss": 0.038, "step": 42810 }, { "grad_norm": 0.37277835607528687, "learning_rate": 6.573520156896476e-05, "loss": 0.0365, "step": 42820 }, { "grad_norm": 0.22753261029720306, "learning_rate": 6.571950614116835e-05, "loss": 0.0432, "step": 42830 }, { "grad_norm": 0.5260692834854126, "learning_rate": 6.570380899431039e-05, "loss": 0.0382, "step": 42840 }, { "grad_norm": 0.28700852394104004, "learning_rate": 6.568811013010749e-05, "loss": 0.0387, "step": 42850 }, { "grad_norm": 0.43324846029281616, "learning_rate": 6.567240955027645e-05, "loss": 0.0466, "step": 42860 }, { "grad_norm": 0.6697880029678345, "learning_rate": 6.565670725653427e-05, "loss": 0.0464, "step": 42870 }, { "grad_norm": 0.8738181591033936, "learning_rate": 6.564100325059812e-05, "loss": 0.0381, "step": 42880 }, { "grad_norm": 0.671493649482727, "learning_rate": 6.562529753418538e-05, "loss": 0.0414, "step": 42890 }, { "grad_norm": 0.27590569853782654, "learning_rate": 6.56095901090136e-05, "loss": 0.0404, "step": 42900 }, { "grad_norm": 0.4871627688407898, "learning_rate": 6.55938809768005e-05, "loss": 0.0394, "step": 42910 }, { "grad_norm": 0.41905078291893005, "learning_rate": 6.557817013926402e-05, "loss": 0.0344, "step": 42920 }, { "grad_norm": 0.36498120427131653, "learning_rate": 6.556245759812227e-05, "loss": 0.0323, "step": 42930 }, { "grad_norm": 0.30420082807540894, "learning_rate": 6.554674335509357e-05, "loss": 0.0393, "step": 42940 }, { "grad_norm": 0.37631869316101074, "learning_rate": 6.553102741189638e-05, "loss": 0.0397, "step": 42950 }, { "grad_norm": 0.35470661520957947, "learning_rate": 6.55153097702494e-05, "loss": 0.0389, "step": 42960 }, { "grad_norm": 0.4960648715496063, "learning_rate": 6.549959043187146e-05, "loss": 0.0368, "step": 42970 }, { "grad_norm": 0.431328684091568, "learning_rate": 6.54838693984816e-05, "loss": 0.04, "step": 42980 }, { "grad_norm": 0.34525710344314575, "learning_rate": 6.546814667179909e-05, "loss": 0.0461, "step": 42990 }, { "grad_norm": 0.3657677471637726, "learning_rate": 6.545242225354328e-05, "loss": 0.0338, "step": 43000 }, { "grad_norm": 0.47166913747787476, "learning_rate": 6.543669614543383e-05, "loss": 0.0346, "step": 43010 }, { "grad_norm": 0.3601316213607788, "learning_rate": 6.542096834919049e-05, "loss": 0.0416, "step": 43020 }, { "grad_norm": 0.32762855291366577, "learning_rate": 6.540523886653324e-05, "loss": 0.0353, "step": 43030 }, { "grad_norm": 0.8842108249664307, "learning_rate": 6.538950769918222e-05, "loss": 0.0408, "step": 43040 }, { "grad_norm": 0.33120492100715637, "learning_rate": 6.537377484885779e-05, "loss": 0.0446, "step": 43050 }, { "grad_norm": 0.4076310396194458, "learning_rate": 6.535804031728045e-05, "loss": 0.0354, "step": 43060 }, { "grad_norm": 0.374885618686676, "learning_rate": 6.534230410617092e-05, "loss": 0.0379, "step": 43070 }, { "grad_norm": 0.28922176361083984, "learning_rate": 6.532656621725007e-05, "loss": 0.0434, "step": 43080 }, { "grad_norm": 0.44284337759017944, "learning_rate": 6.531082665223898e-05, "loss": 0.034, "step": 43090 }, { "grad_norm": 0.4164057672023773, "learning_rate": 6.529508541285889e-05, "loss": 0.0373, "step": 43100 }, { "grad_norm": 0.3497878611087799, "learning_rate": 6.527934250083128e-05, "loss": 0.0444, "step": 43110 }, { "grad_norm": 0.3689406216144562, "learning_rate": 6.526359791787775e-05, "loss": 0.0378, "step": 43120 }, { "grad_norm": 0.6226054430007935, "learning_rate": 6.524785166572009e-05, "loss": 0.0399, "step": 43130 }, { "grad_norm": 0.5003830790519714, "learning_rate": 6.523210374608029e-05, "loss": 0.0385, "step": 43140 }, { "grad_norm": 0.36531227827072144, "learning_rate": 6.521635416068054e-05, "loss": 0.0402, "step": 43150 }, { "grad_norm": 0.3810792565345764, "learning_rate": 6.520060291124317e-05, "loss": 0.0361, "step": 43160 }, { "grad_norm": 0.24543088674545288, "learning_rate": 6.518484999949072e-05, "loss": 0.0353, "step": 43170 }, { "grad_norm": 0.3709489703178406, "learning_rate": 6.516909542714591e-05, "loss": 0.0417, "step": 43180 }, { "grad_norm": 0.3638562560081482, "learning_rate": 6.515333919593161e-05, "loss": 0.0431, "step": 43190 }, { "grad_norm": 0.31132543087005615, "learning_rate": 6.513758130757094e-05, "loss": 0.041, "step": 43200 }, { "grad_norm": 0.5104261040687561, "learning_rate": 6.512182176378713e-05, "loss": 0.0485, "step": 43210 }, { "grad_norm": 0.42250826954841614, "learning_rate": 6.510606056630362e-05, "loss": 0.0435, "step": 43220 }, { "grad_norm": 0.4193108379840851, "learning_rate": 6.509029771684405e-05, "loss": 0.0372, "step": 43230 }, { "grad_norm": 0.3513339161872864, "learning_rate": 6.50745332171322e-05, "loss": 0.0443, "step": 43240 }, { "grad_norm": 0.48703503608703613, "learning_rate": 6.505876706889207e-05, "loss": 0.0366, "step": 43250 }, { "grad_norm": 0.43072274327278137, "learning_rate": 6.504299927384781e-05, "loss": 0.0404, "step": 43260 }, { "grad_norm": 0.4202333092689514, "learning_rate": 6.502722983372378e-05, "loss": 0.0424, "step": 43270 }, { "grad_norm": 0.3657507002353668, "learning_rate": 6.501145875024446e-05, "loss": 0.0371, "step": 43280 }, { "grad_norm": 0.33852753043174744, "learning_rate": 6.499568602513459e-05, "loss": 0.0406, "step": 43290 }, { "grad_norm": 0.46189558506011963, "learning_rate": 6.497991166011903e-05, "loss": 0.0348, "step": 43300 }, { "grad_norm": 0.3721068799495697, "learning_rate": 6.49641356569229e-05, "loss": 0.0401, "step": 43310 }, { "grad_norm": 0.2637307345867157, "learning_rate": 6.494835801727135e-05, "loss": 0.0431, "step": 43320 }, { "grad_norm": 0.34069088101387024, "learning_rate": 6.493257874288986e-05, "loss": 0.0356, "step": 43330 }, { "grad_norm": 0.43883562088012695, "learning_rate": 6.491679783550399e-05, "loss": 0.0412, "step": 43340 }, { "grad_norm": 0.3651091754436493, "learning_rate": 6.490101529683957e-05, "loss": 0.0412, "step": 43350 }, { "grad_norm": 0.3907252252101898, "learning_rate": 6.488523112862248e-05, "loss": 0.0413, "step": 43360 }, { "grad_norm": 0.4508742392063141, "learning_rate": 6.486944533257891e-05, "loss": 0.037, "step": 43370 }, { "grad_norm": 0.409445196390152, "learning_rate": 6.485365791043517e-05, "loss": 0.0441, "step": 43380 }, { "grad_norm": 0.5655009150505066, "learning_rate": 6.483786886391772e-05, "loss": 0.0373, "step": 43390 }, { "grad_norm": 0.3502785861492157, "learning_rate": 6.482207819475323e-05, "loss": 0.0373, "step": 43400 }, { "grad_norm": 0.30982106924057007, "learning_rate": 6.480628590466857e-05, "loss": 0.038, "step": 43410 }, { "grad_norm": 0.4080727696418762, "learning_rate": 6.479049199539074e-05, "loss": 0.0452, "step": 43420 }, { "grad_norm": 0.4112193286418915, "learning_rate": 6.477469646864693e-05, "loss": 0.0472, "step": 43430 }, { "grad_norm": 0.531117856502533, "learning_rate": 6.475889932616454e-05, "loss": 0.0508, "step": 43440 }, { "grad_norm": 0.3306344747543335, "learning_rate": 6.474310056967111e-05, "loss": 0.0424, "step": 43450 }, { "grad_norm": 0.33027759194374084, "learning_rate": 6.472730020089437e-05, "loss": 0.046, "step": 43460 }, { "grad_norm": 0.9963865876197815, "learning_rate": 6.471149822156222e-05, "loss": 0.0357, "step": 43470 }, { "grad_norm": 0.5552625060081482, "learning_rate": 6.469569463340275e-05, "loss": 0.0382, "step": 43480 }, { "grad_norm": 0.886240541934967, "learning_rate": 6.467988943814418e-05, "loss": 0.0408, "step": 43490 }, { "grad_norm": 0.5622723698616028, "learning_rate": 6.4664082637515e-05, "loss": 0.0405, "step": 43500 }, { "grad_norm": 0.3558625876903534, "learning_rate": 6.464827423324377e-05, "loss": 0.0324, "step": 43510 }, { "grad_norm": 0.38635900616645813, "learning_rate": 6.463246422705931e-05, "loss": 0.0471, "step": 43520 }, { "grad_norm": 0.3284810781478882, "learning_rate": 6.461665262069055e-05, "loss": 0.04, "step": 43530 }, { "grad_norm": 0.6310411095619202, "learning_rate": 6.460083941586665e-05, "loss": 0.0451, "step": 43540 }, { "grad_norm": 0.47819167375564575, "learning_rate": 6.45850246143169e-05, "loss": 0.0362, "step": 43550 }, { "grad_norm": 0.375713974237442, "learning_rate": 6.456920821777077e-05, "loss": 0.0373, "step": 43560 }, { "grad_norm": 0.5162724852561951, "learning_rate": 6.455339022795796e-05, "loss": 0.0366, "step": 43570 }, { "grad_norm": 0.5381721258163452, "learning_rate": 6.453757064660825e-05, "loss": 0.0381, "step": 43580 }, { "grad_norm": 0.334100604057312, "learning_rate": 6.452174947545169e-05, "loss": 0.0379, "step": 43590 }, { "grad_norm": 0.4374145269393921, "learning_rate": 6.450592671621842e-05, "loss": 0.0377, "step": 43600 }, { "grad_norm": 0.540343701839447, "learning_rate": 6.449010237063882e-05, "loss": 0.0423, "step": 43610 }, { "grad_norm": 0.4100266396999359, "learning_rate": 6.447427644044343e-05, "loss": 0.0393, "step": 43620 }, { "grad_norm": 0.3568008840084076, "learning_rate": 6.44584489273629e-05, "loss": 0.0437, "step": 43630 }, { "grad_norm": 0.28721773624420166, "learning_rate": 6.444261983312817e-05, "loss": 0.0345, "step": 43640 }, { "grad_norm": 0.2880375385284424, "learning_rate": 6.442678915947023e-05, "loss": 0.0346, "step": 43650 }, { "grad_norm": 0.33510783314704895, "learning_rate": 6.441095690812032e-05, "loss": 0.0333, "step": 43660 }, { "grad_norm": 0.3861373960971832, "learning_rate": 6.439512308080982e-05, "loss": 0.0416, "step": 43670 }, { "grad_norm": 0.4953594207763672, "learning_rate": 6.43792876792703e-05, "loss": 0.0353, "step": 43680 }, { "grad_norm": 0.41587018966674805, "learning_rate": 6.43634507052335e-05, "loss": 0.0444, "step": 43690 }, { "grad_norm": 0.4444204568862915, "learning_rate": 6.434761216043135e-05, "loss": 0.0429, "step": 43700 }, { "grad_norm": 0.412331223487854, "learning_rate": 6.433177204659589e-05, "loss": 0.0389, "step": 43710 }, { "grad_norm": 0.35283470153808594, "learning_rate": 6.431593036545938e-05, "loss": 0.0394, "step": 43720 }, { "grad_norm": 0.5083481073379517, "learning_rate": 6.430008711875426e-05, "loss": 0.041, "step": 43730 }, { "grad_norm": 0.4100798964500427, "learning_rate": 6.42842423082131e-05, "loss": 0.0418, "step": 43740 }, { "grad_norm": 0.3636026084423065, "learning_rate": 6.426839593556868e-05, "loss": 0.0355, "step": 43750 }, { "grad_norm": 0.3493400812149048, "learning_rate": 6.425254800255392e-05, "loss": 0.0388, "step": 43760 }, { "grad_norm": 0.3686246871948242, "learning_rate": 6.423669851090196e-05, "loss": 0.0394, "step": 43770 }, { "grad_norm": 0.2824319005012512, "learning_rate": 6.422084746234604e-05, "loss": 0.0372, "step": 43780 }, { "grad_norm": 0.30603891611099243, "learning_rate": 6.420499485861965e-05, "loss": 0.0394, "step": 43790 }, { "grad_norm": 0.30226966738700867, "learning_rate": 6.418914070145636e-05, "loss": 0.0389, "step": 43800 }, { "grad_norm": 0.3213949501514435, "learning_rate": 6.417328499258996e-05, "loss": 0.0414, "step": 43810 }, { "grad_norm": 0.4168335199356079, "learning_rate": 6.415742773375446e-05, "loss": 0.0349, "step": 43820 }, { "grad_norm": 0.4387018382549286, "learning_rate": 6.414156892668393e-05, "loss": 0.0374, "step": 43830 }, { "grad_norm": 0.40525147318840027, "learning_rate": 6.412570857311267e-05, "loss": 0.0394, "step": 43840 }, { "grad_norm": 0.32810431718826294, "learning_rate": 6.410984667477518e-05, "loss": 0.0377, "step": 43850 }, { "grad_norm": 0.26866278052330017, "learning_rate": 6.409398323340607e-05, "loss": 0.0336, "step": 43860 }, { "grad_norm": 0.32084497809410095, "learning_rate": 6.407811825074012e-05, "loss": 0.0351, "step": 43870 }, { "grad_norm": 0.2721840739250183, "learning_rate": 6.406225172851234e-05, "loss": 0.0398, "step": 43880 }, { "grad_norm": 1.1195648908615112, "learning_rate": 6.404638366845786e-05, "loss": 0.0365, "step": 43890 }, { "grad_norm": 0.37253835797309875, "learning_rate": 6.403051407231196e-05, "loss": 0.0378, "step": 43900 }, { "grad_norm": 0.37941181659698486, "learning_rate": 6.401464294181016e-05, "loss": 0.0327, "step": 43910 }, { "grad_norm": 0.30783066153526306, "learning_rate": 6.399877027868808e-05, "loss": 0.0397, "step": 43920 }, { "grad_norm": 0.3653622567653656, "learning_rate": 6.39828960846815e-05, "loss": 0.0402, "step": 43930 }, { "grad_norm": 0.2980310916900635, "learning_rate": 6.396702036152644e-05, "loss": 0.044, "step": 43940 }, { "grad_norm": 0.5405564308166504, "learning_rate": 6.395114311095905e-05, "loss": 0.046, "step": 43950 }, { "grad_norm": 0.36981508135795593, "learning_rate": 6.393526433471559e-05, "loss": 0.0412, "step": 43960 }, { "grad_norm": 0.40104684233665466, "learning_rate": 6.391938403453262e-05, "loss": 0.0361, "step": 43970 }, { "grad_norm": 0.2711935043334961, "learning_rate": 6.390350221214671e-05, "loss": 0.0345, "step": 43980 }, { "grad_norm": 0.3444893956184387, "learning_rate": 6.38876188692947e-05, "loss": 0.0377, "step": 43990 }, { "grad_norm": 0.7530564069747925, "learning_rate": 6.387173400771358e-05, "loss": 0.0456, "step": 44000 }, { "grad_norm": 0.35606732964515686, "learning_rate": 6.385584762914047e-05, "loss": 0.0392, "step": 44010 }, { "grad_norm": 0.34688225388526917, "learning_rate": 6.38399597353127e-05, "loss": 0.045, "step": 44020 }, { "grad_norm": 0.3741860091686249, "learning_rate": 6.382407032796775e-05, "loss": 0.0386, "step": 44030 }, { "grad_norm": 0.42221781611442566, "learning_rate": 6.380817940884325e-05, "loss": 0.039, "step": 44040 }, { "grad_norm": 0.39509138464927673, "learning_rate": 6.379228697967702e-05, "loss": 0.0392, "step": 44050 }, { "grad_norm": 0.2582338750362396, "learning_rate": 6.3776393042207e-05, "loss": 0.0388, "step": 44060 }, { "grad_norm": 0.38074737787246704, "learning_rate": 6.376049759817137e-05, "loss": 0.0382, "step": 44070 }, { "grad_norm": 0.3618744909763336, "learning_rate": 6.37446006493084e-05, "loss": 0.0325, "step": 44080 }, { "grad_norm": 1.8068276643753052, "learning_rate": 6.372870219735656e-05, "loss": 0.0324, "step": 44090 }, { "grad_norm": 0.5949942469596863, "learning_rate": 6.371280224405452e-05, "loss": 0.0357, "step": 44100 }, { "grad_norm": 0.44700688123703003, "learning_rate": 6.369690079114101e-05, "loss": 0.0381, "step": 44110 }, { "grad_norm": 0.32760313153266907, "learning_rate": 6.368099784035504e-05, "loss": 0.0418, "step": 44120 }, { "grad_norm": 0.3335956633090973, "learning_rate": 6.366509339343572e-05, "loss": 0.0411, "step": 44130 }, { "grad_norm": 0.34225404262542725, "learning_rate": 6.364918745212232e-05, "loss": 0.034, "step": 44140 }, { "grad_norm": 0.3104611039161682, "learning_rate": 6.363328001815431e-05, "loss": 0.0342, "step": 44150 }, { "grad_norm": 0.3800029754638672, "learning_rate": 6.361737109327128e-05, "loss": 0.0322, "step": 44160 }, { "grad_norm": 0.454205185174942, "learning_rate": 6.360146067921301e-05, "loss": 0.0386, "step": 44170 }, { "grad_norm": 0.5288757681846619, "learning_rate": 6.35855487777195e-05, "loss": 0.0378, "step": 44180 }, { "grad_norm": 0.34067195653915405, "learning_rate": 6.356963539053076e-05, "loss": 0.0363, "step": 44190 }, { "grad_norm": 0.39659813046455383, "learning_rate": 6.355372051938711e-05, "loss": 0.0359, "step": 44200 }, { "grad_norm": 0.4395281672477722, "learning_rate": 6.353780416602894e-05, "loss": 0.039, "step": 44210 }, { "grad_norm": 0.300841748714447, "learning_rate": 6.352188633219689e-05, "loss": 0.0351, "step": 44220 }, { "grad_norm": 0.5998445749282837, "learning_rate": 6.350596701963166e-05, "loss": 0.0349, "step": 44230 }, { "grad_norm": 0.33603742718696594, "learning_rate": 6.349004623007419e-05, "loss": 0.0351, "step": 44240 }, { "grad_norm": 0.5698198080062866, "learning_rate": 6.347412396526555e-05, "loss": 0.037, "step": 44250 }, { "grad_norm": 0.4125266373157501, "learning_rate": 6.345820022694696e-05, "loss": 0.0526, "step": 44260 }, { "grad_norm": 0.40106114745140076, "learning_rate": 6.344227501685984e-05, "loss": 0.0386, "step": 44270 }, { "grad_norm": 0.3447568118572235, "learning_rate": 6.342634833674572e-05, "loss": 0.0366, "step": 44280 }, { "grad_norm": 0.486378937959671, "learning_rate": 6.341042018834635e-05, "loss": 0.0462, "step": 44290 }, { "grad_norm": 0.3249393403530121, "learning_rate": 6.339449057340359e-05, "loss": 0.0344, "step": 44300 }, { "grad_norm": 0.33785197138786316, "learning_rate": 6.337855949365945e-05, "loss": 0.0389, "step": 44310 }, { "grad_norm": 0.32353031635284424, "learning_rate": 6.336262695085619e-05, "loss": 0.0417, "step": 44320 }, { "grad_norm": 0.37972864508628845, "learning_rate": 6.334669294673612e-05, "loss": 0.0376, "step": 44330 }, { "grad_norm": 0.2589975595474243, "learning_rate": 6.333075748304179e-05, "loss": 0.0385, "step": 44340 }, { "grad_norm": 0.3673405945301056, "learning_rate": 6.331482056151585e-05, "loss": 0.0323, "step": 44350 }, { "grad_norm": 0.630763590335846, "learning_rate": 6.329888218390117e-05, "loss": 0.0444, "step": 44360 }, { "grad_norm": 0.29748842120170593, "learning_rate": 6.328294235194072e-05, "loss": 0.0397, "step": 44370 }, { "grad_norm": 0.39412882924079895, "learning_rate": 6.326700106737765e-05, "loss": 0.0357, "step": 44380 }, { "grad_norm": 0.4038437306880951, "learning_rate": 6.325105833195531e-05, "loss": 0.0359, "step": 44390 }, { "grad_norm": 0.35994842648506165, "learning_rate": 6.323511414741715e-05, "loss": 0.0379, "step": 44400 }, { "grad_norm": 0.3293170928955078, "learning_rate": 6.321916851550678e-05, "loss": 0.035, "step": 44410 }, { "grad_norm": 0.4027264714241028, "learning_rate": 6.320322143796806e-05, "loss": 0.0357, "step": 44420 }, { "grad_norm": 0.3974480628967285, "learning_rate": 6.318727291654488e-05, "loss": 0.0361, "step": 44430 }, { "grad_norm": 0.45708543062210083, "learning_rate": 6.317132295298134e-05, "loss": 0.0386, "step": 44440 }, { "grad_norm": 0.3669370412826538, "learning_rate": 6.315537154902173e-05, "loss": 0.0368, "step": 44450 }, { "grad_norm": 0.2988233268260956, "learning_rate": 6.313941870641048e-05, "loss": 0.0376, "step": 44460 }, { "grad_norm": 0.33496570587158203, "learning_rate": 6.312346442689214e-05, "loss": 0.035, "step": 44470 }, { "grad_norm": 0.3893566429615021, "learning_rate": 6.310750871221147e-05, "loss": 0.0399, "step": 44480 }, { "grad_norm": 0.2542894780635834, "learning_rate": 6.309155156411335e-05, "loss": 0.0361, "step": 44490 }, { "grad_norm": 0.2817131578922272, "learning_rate": 6.307559298434284e-05, "loss": 0.0421, "step": 44500 }, { "grad_norm": 0.3548469543457031, "learning_rate": 6.305963297464512e-05, "loss": 0.0404, "step": 44510 }, { "grad_norm": 0.3762679398059845, "learning_rate": 6.304367153676561e-05, "loss": 0.0334, "step": 44520 }, { "grad_norm": 0.3405458629131317, "learning_rate": 6.302770867244978e-05, "loss": 0.0337, "step": 44530 }, { "grad_norm": 0.31824272871017456, "learning_rate": 6.301174438344328e-05, "loss": 0.039, "step": 44540 }, { "grad_norm": 0.2909514605998993, "learning_rate": 6.299577867149202e-05, "loss": 0.0404, "step": 44550 }, { "grad_norm": 0.4372059404850006, "learning_rate": 6.297981153834192e-05, "loss": 0.0378, "step": 44560 }, { "grad_norm": 0.4268491864204407, "learning_rate": 6.296384298573916e-05, "loss": 0.0388, "step": 44570 }, { "grad_norm": 0.4289407432079315, "learning_rate": 6.294787301543001e-05, "loss": 0.0403, "step": 44580 }, { "grad_norm": 0.33810585737228394, "learning_rate": 6.293190162916095e-05, "loss": 0.0318, "step": 44590 }, { "grad_norm": 0.38721898198127747, "learning_rate": 6.291592882867855e-05, "loss": 0.036, "step": 44600 }, { "grad_norm": 0.3967391848564148, "learning_rate": 6.28999546157296e-05, "loss": 0.0385, "step": 44610 }, { "grad_norm": 0.36013802886009216, "learning_rate": 6.288397899206102e-05, "loss": 0.0345, "step": 44620 }, { "grad_norm": 0.26888126134872437, "learning_rate": 6.286800195941984e-05, "loss": 0.0351, "step": 44630 }, { "grad_norm": 0.32980820536613464, "learning_rate": 6.285202351955334e-05, "loss": 0.0329, "step": 44640 }, { "grad_norm": 0.45511817932128906, "learning_rate": 6.283604367420887e-05, "loss": 0.0383, "step": 44650 }, { "grad_norm": 0.35145512223243713, "learning_rate": 6.282006242513394e-05, "loss": 0.0435, "step": 44660 }, { "grad_norm": 0.3712266683578491, "learning_rate": 6.280407977407628e-05, "loss": 0.0377, "step": 44670 }, { "grad_norm": 0.3424878418445587, "learning_rate": 6.27880957227837e-05, "loss": 0.0379, "step": 44680 }, { "grad_norm": 0.380272775888443, "learning_rate": 6.27721102730042e-05, "loss": 0.0337, "step": 44690 }, { "grad_norm": 0.3524571359157562, "learning_rate": 6.275612342648592e-05, "loss": 0.0373, "step": 44700 }, { "grad_norm": 0.3217064142227173, "learning_rate": 6.274013518497716e-05, "loss": 0.0363, "step": 44710 }, { "grad_norm": 0.5092492699623108, "learning_rate": 6.272414555022636e-05, "loss": 0.039, "step": 44720 }, { "grad_norm": 0.5423514246940613, "learning_rate": 6.270815452398215e-05, "loss": 0.0412, "step": 44730 }, { "grad_norm": 0.328995943069458, "learning_rate": 6.269216210799326e-05, "loss": 0.0332, "step": 44740 }, { "grad_norm": 0.6313387155532837, "learning_rate": 6.26761683040086e-05, "loss": 0.0371, "step": 44750 }, { "grad_norm": 0.40849655866622925, "learning_rate": 6.266017311377723e-05, "loss": 0.0395, "step": 44760 }, { "grad_norm": 0.38405919075012207, "learning_rate": 6.264417653904839e-05, "loss": 0.0396, "step": 44770 }, { "grad_norm": 0.41123104095458984, "learning_rate": 6.262817858157139e-05, "loss": 0.0343, "step": 44780 }, { "grad_norm": 0.38067081570625305, "learning_rate": 6.261217924309576e-05, "loss": 0.045, "step": 44790 }, { "grad_norm": 0.33081740140914917, "learning_rate": 6.259617852537118e-05, "loss": 0.0401, "step": 44800 }, { "grad_norm": 0.38253355026245117, "learning_rate": 6.258017643014747e-05, "loss": 0.0433, "step": 44810 }, { "grad_norm": 0.32156074047088623, "learning_rate": 6.256417295917456e-05, "loss": 0.0348, "step": 44820 }, { "grad_norm": 0.4432334005832672, "learning_rate": 6.254816811420258e-05, "loss": 0.0412, "step": 44830 }, { "grad_norm": 0.2676014006137848, "learning_rate": 6.253216189698183e-05, "loss": 0.0345, "step": 44840 }, { "grad_norm": 0.3142072558403015, "learning_rate": 6.251615430926267e-05, "loss": 0.0371, "step": 44850 }, { "grad_norm": 0.4655921459197998, "learning_rate": 6.25001453527957e-05, "loss": 0.0413, "step": 44860 }, { "grad_norm": 0.32018181681632996, "learning_rate": 6.248413502933164e-05, "loss": 0.0409, "step": 44870 }, { "grad_norm": 0.3258059620857239, "learning_rate": 6.246812334062133e-05, "loss": 0.0347, "step": 44880 }, { "grad_norm": 0.3683418929576874, "learning_rate": 6.245211028841579e-05, "loss": 0.0363, "step": 44890 }, { "grad_norm": 0.5987353920936584, "learning_rate": 6.24360958744662e-05, "loss": 0.0338, "step": 44900 }, { "grad_norm": 0.3837919235229492, "learning_rate": 6.242008010052387e-05, "loss": 0.0326, "step": 44910 }, { "grad_norm": 0.477327436208725, "learning_rate": 6.240406296834024e-05, "loss": 0.0389, "step": 44920 }, { "grad_norm": 0.4203064441680908, "learning_rate": 6.238804447966694e-05, "loss": 0.0418, "step": 44930 }, { "grad_norm": 0.428226113319397, "learning_rate": 6.237202463625573e-05, "loss": 0.0401, "step": 44940 }, { "grad_norm": 0.31725409626960754, "learning_rate": 6.235600343985848e-05, "loss": 0.0441, "step": 44950 }, { "grad_norm": 0.2852192223072052, "learning_rate": 6.233998089222729e-05, "loss": 0.0393, "step": 44960 }, { "grad_norm": 1.2202643156051636, "learning_rate": 6.232395699511433e-05, "loss": 0.0355, "step": 44970 }, { "grad_norm": 0.2671827971935272, "learning_rate": 6.230793175027199e-05, "loss": 0.0319, "step": 44980 }, { "grad_norm": 0.30502820014953613, "learning_rate": 6.22919051594527e-05, "loss": 0.0433, "step": 44990 }, { "grad_norm": 0.33133113384246826, "learning_rate": 6.227587722440917e-05, "loss": 0.0319, "step": 45000 }, { "grad_norm": 0.28157344460487366, "learning_rate": 6.225984794689414e-05, "loss": 0.0374, "step": 45010 }, { "grad_norm": 0.4527588486671448, "learning_rate": 6.22438173286606e-05, "loss": 0.0353, "step": 45020 }, { "grad_norm": 0.44093188643455505, "learning_rate": 6.222778537146159e-05, "loss": 0.0396, "step": 45030 }, { "grad_norm": 0.387158066034317, "learning_rate": 6.221175207705037e-05, "loss": 0.0328, "step": 45040 }, { "grad_norm": 0.3346082866191864, "learning_rate": 6.219571744718027e-05, "loss": 0.0403, "step": 45050 }, { "grad_norm": 3.319159507751465, "learning_rate": 6.217968148360487e-05, "loss": 0.0415, "step": 45060 }, { "grad_norm": 0.3685930073261261, "learning_rate": 6.216364418807782e-05, "loss": 0.0433, "step": 45070 }, { "grad_norm": 0.4576917290687561, "learning_rate": 6.214760556235292e-05, "loss": 0.0372, "step": 45080 }, { "grad_norm": 0.4519270360469818, "learning_rate": 6.213156560818414e-05, "loss": 0.0384, "step": 45090 }, { "grad_norm": 0.4227359890937805, "learning_rate": 6.211552432732559e-05, "loss": 0.0394, "step": 45100 }, { "grad_norm": 0.41782790422439575, "learning_rate": 6.20994817215315e-05, "loss": 0.0415, "step": 45110 }, { "grad_norm": 0.4005136489868164, "learning_rate": 6.208343779255628e-05, "loss": 0.0406, "step": 45120 }, { "grad_norm": 0.4187650978565216, "learning_rate": 6.206739254215449e-05, "loss": 0.0361, "step": 45130 }, { "grad_norm": 0.4514830410480499, "learning_rate": 6.205134597208077e-05, "loss": 0.0384, "step": 45140 }, { "grad_norm": 0.4201471209526062, "learning_rate": 6.203529808409e-05, "loss": 0.0344, "step": 45150 }, { "grad_norm": 0.23140494525432587, "learning_rate": 6.201924887993712e-05, "loss": 0.0362, "step": 45160 }, { "grad_norm": 0.4065161943435669, "learning_rate": 6.200319836137725e-05, "loss": 0.0367, "step": 45170 }, { "grad_norm": 0.3550943434238434, "learning_rate": 6.198714653016565e-05, "loss": 0.0389, "step": 45180 }, { "grad_norm": 0.4339110851287842, "learning_rate": 6.197109338805774e-05, "loss": 0.0428, "step": 45190 }, { "grad_norm": 0.31610333919525146, "learning_rate": 6.195503893680903e-05, "loss": 0.0396, "step": 45200 }, { "grad_norm": 0.6984142661094666, "learning_rate": 6.193898317817524e-05, "loss": 0.0351, "step": 45210 }, { "grad_norm": 0.31115368008613586, "learning_rate": 6.192292611391221e-05, "loss": 0.0388, "step": 45220 }, { "grad_norm": 0.328261137008667, "learning_rate": 6.190686774577591e-05, "loss": 0.0351, "step": 45230 }, { "grad_norm": 1.0500863790512085, "learning_rate": 6.189080807552245e-05, "loss": 0.035, "step": 45240 }, { "grad_norm": 0.2657225728034973, "learning_rate": 6.187474710490809e-05, "loss": 0.0403, "step": 45250 }, { "grad_norm": 0.38132408261299133, "learning_rate": 6.185868483568926e-05, "loss": 0.0354, "step": 45260 }, { "grad_norm": 0.5079218149185181, "learning_rate": 6.184262126962245e-05, "loss": 0.0365, "step": 45270 }, { "grad_norm": 0.342973530292511, "learning_rate": 6.182655640846442e-05, "loss": 0.0368, "step": 45280 }, { "grad_norm": 0.27398210763931274, "learning_rate": 6.181049025397196e-05, "loss": 0.0377, "step": 45290 }, { "grad_norm": 0.32092610001564026, "learning_rate": 6.179442280790202e-05, "loss": 0.0362, "step": 45300 }, { "grad_norm": 0.3199017643928528, "learning_rate": 6.177835407201174e-05, "loss": 0.0356, "step": 45310 }, { "grad_norm": 0.33401554822921753, "learning_rate": 6.176228404805839e-05, "loss": 0.0399, "step": 45320 }, { "grad_norm": 0.4002723693847656, "learning_rate": 6.174621273779932e-05, "loss": 0.043, "step": 45330 }, { "grad_norm": 0.4312285780906677, "learning_rate": 6.17301401429921e-05, "loss": 0.0442, "step": 45340 }, { "grad_norm": 0.3170653283596039, "learning_rate": 6.171406626539436e-05, "loss": 0.0322, "step": 45350 }, { "grad_norm": 0.671581506729126, "learning_rate": 6.169799110676398e-05, "loss": 0.0393, "step": 45360 }, { "grad_norm": 0.400577187538147, "learning_rate": 6.168191466885885e-05, "loss": 0.0438, "step": 45370 }, { "grad_norm": 0.3047610819339752, "learning_rate": 6.166583695343711e-05, "loss": 0.0393, "step": 45380 }, { "grad_norm": 0.35227593779563904, "learning_rate": 6.164975796225698e-05, "loss": 0.0385, "step": 45390 }, { "grad_norm": 0.3697374165058136, "learning_rate": 6.163367769707683e-05, "loss": 0.0334, "step": 45400 }, { "grad_norm": 0.41307955980300903, "learning_rate": 6.161759615965519e-05, "loss": 0.0396, "step": 45410 }, { "grad_norm": 0.4482290744781494, "learning_rate": 6.16015133517507e-05, "loss": 0.0394, "step": 45420 }, { "grad_norm": 0.3739679455757141, "learning_rate": 6.158542927512214e-05, "loss": 0.0409, "step": 45430 }, { "grad_norm": 0.30051541328430176, "learning_rate": 6.156934393152846e-05, "loss": 0.0389, "step": 45440 }, { "grad_norm": 0.3458985984325409, "learning_rate": 6.15532573227287e-05, "loss": 0.041, "step": 45450 }, { "grad_norm": 0.30650973320007324, "learning_rate": 6.153716945048212e-05, "loss": 0.0343, "step": 45460 }, { "grad_norm": 0.4194903075695038, "learning_rate": 6.152108031654802e-05, "loss": 0.0409, "step": 45470 }, { "grad_norm": 0.508976399898529, "learning_rate": 6.15049899226859e-05, "loss": 0.0389, "step": 45480 }, { "grad_norm": 0.3890330195426941, "learning_rate": 6.148889827065537e-05, "loss": 0.046, "step": 45490 }, { "grad_norm": 0.5615267157554626, "learning_rate": 6.147280536221622e-05, "loss": 0.0375, "step": 45500 }, { "grad_norm": 0.3933942914009094, "learning_rate": 6.145671119912832e-05, "loss": 0.0364, "step": 45510 }, { "grad_norm": 0.29981377720832825, "learning_rate": 6.144061578315169e-05, "loss": 0.0408, "step": 45520 }, { "grad_norm": 0.3345426321029663, "learning_rate": 6.142451911604654e-05, "loss": 0.0448, "step": 45530 }, { "grad_norm": 0.2950475513935089, "learning_rate": 6.140842119957315e-05, "loss": 0.0364, "step": 45540 }, { "grad_norm": 0.358613520860672, "learning_rate": 6.139232203549197e-05, "loss": 0.0388, "step": 45550 }, { "grad_norm": 0.29799774289131165, "learning_rate": 6.137622162556357e-05, "loss": 0.0358, "step": 45560 }, { "grad_norm": 0.4446188509464264, "learning_rate": 6.13601199715487e-05, "loss": 0.0343, "step": 45570 }, { "grad_norm": 0.28588882088661194, "learning_rate": 6.134401707520816e-05, "loss": 0.0394, "step": 45580 }, { "grad_norm": 0.4006318747997284, "learning_rate": 6.132791293830298e-05, "loss": 0.0337, "step": 45590 }, { "grad_norm": 0.5474902391433716, "learning_rate": 6.131180756259428e-05, "loss": 0.0435, "step": 45600 }, { "grad_norm": 0.3988974392414093, "learning_rate": 6.129570094984331e-05, "loss": 0.0382, "step": 45610 }, { "grad_norm": 0.39143696427345276, "learning_rate": 6.127959310181145e-05, "loss": 0.0412, "step": 45620 }, { "grad_norm": 0.30871257185935974, "learning_rate": 6.126348402026026e-05, "loss": 0.0361, "step": 45630 }, { "grad_norm": 0.44320541620254517, "learning_rate": 6.12473737069514e-05, "loss": 0.0302, "step": 45640 }, { "grad_norm": 0.3331235647201538, "learning_rate": 6.123126216364665e-05, "loss": 0.0364, "step": 45650 }, { "grad_norm": 0.3492863178253174, "learning_rate": 6.121514939210797e-05, "loss": 0.0386, "step": 45660 }, { "grad_norm": 0.3330366313457489, "learning_rate": 6.119903539409741e-05, "loss": 0.0379, "step": 45670 }, { "grad_norm": 0.3173966407775879, "learning_rate": 6.118292017137716e-05, "loss": 0.0307, "step": 45680 }, { "grad_norm": 0.28481167554855347, "learning_rate": 6.116680372570959e-05, "loss": 0.035, "step": 45690 }, { "grad_norm": 0.5185493230819702, "learning_rate": 6.115068605885713e-05, "loss": 0.0432, "step": 45700 }, { "grad_norm": 0.4269501566886902, "learning_rate": 6.113456717258243e-05, "loss": 0.0357, "step": 45710 }, { "grad_norm": 0.4274452030658722, "learning_rate": 6.11184470686482e-05, "loss": 0.0411, "step": 45720 }, { "grad_norm": 0.33642318844795227, "learning_rate": 6.110232574881731e-05, "loss": 0.0362, "step": 45730 }, { "grad_norm": 0.28750160336494446, "learning_rate": 6.108620321485277e-05, "loss": 0.0358, "step": 45740 }, { "grad_norm": 0.38013362884521484, "learning_rate": 6.107007946851773e-05, "loss": 0.0397, "step": 45750 }, { "grad_norm": 0.3213399052619934, "learning_rate": 6.105395451157542e-05, "loss": 0.0409, "step": 45760 }, { "grad_norm": 0.3243512511253357, "learning_rate": 6.103782834578928e-05, "loss": 0.0358, "step": 45770 }, { "grad_norm": 0.2952705919742584, "learning_rate": 6.102170097292281e-05, "loss": 0.0283, "step": 45780 }, { "grad_norm": 0.863513708114624, "learning_rate": 6.1005572394739716e-05, "loss": 0.0366, "step": 45790 }, { "grad_norm": 0.4530574083328247, "learning_rate": 6.0989442613003765e-05, "loss": 0.046, "step": 45800 }, { "grad_norm": 0.38784393668174744, "learning_rate": 6.097331162947888e-05, "loss": 0.0382, "step": 45810 }, { "grad_norm": 0.5411588549613953, "learning_rate": 6.095717944592914e-05, "loss": 0.0383, "step": 45820 }, { "grad_norm": 0.3714894950389862, "learning_rate": 6.094104606411873e-05, "loss": 0.0338, "step": 45830 }, { "grad_norm": 0.36254584789276123, "learning_rate": 6.0924911485811966e-05, "loss": 0.039, "step": 45840 }, { "grad_norm": 0.34209194779396057, "learning_rate": 6.090877571277331e-05, "loss": 0.0338, "step": 45850 }, { "grad_norm": 0.4213067293167114, "learning_rate": 6.089263874676736e-05, "loss": 0.0368, "step": 45860 }, { "grad_norm": 0.5220191478729248, "learning_rate": 6.0876500589558796e-05, "loss": 0.0394, "step": 45870 }, { "grad_norm": 0.4516684412956238, "learning_rate": 6.086036124291248e-05, "loss": 0.0372, "step": 45880 }, { "grad_norm": 0.2741973102092743, "learning_rate": 6.084422070859339e-05, "loss": 0.0378, "step": 45890 }, { "grad_norm": 0.4217641055583954, "learning_rate": 6.082807898836663e-05, "loss": 0.0381, "step": 45900 }, { "grad_norm": 0.2881215214729309, "learning_rate": 6.081193608399742e-05, "loss": 0.0413, "step": 45910 }, { "grad_norm": 0.4255719482898712, "learning_rate": 6.0795791997251164e-05, "loss": 0.0387, "step": 45920 }, { "grad_norm": 0.38083651661872864, "learning_rate": 6.0779646729893294e-05, "loss": 0.0333, "step": 45930 }, { "grad_norm": 0.44673892855644226, "learning_rate": 6.0763500283689476e-05, "loss": 0.0373, "step": 45940 }, { "grad_norm": 0.3814341127872467, "learning_rate": 6.0747352660405455e-05, "loss": 0.036, "step": 45950 }, { "grad_norm": 0.7427939772605896, "learning_rate": 6.073120386180709e-05, "loss": 0.0338, "step": 45960 }, { "grad_norm": 0.41919514536857605, "learning_rate": 6.0715053889660425e-05, "loss": 0.0337, "step": 45970 }, { "grad_norm": 0.3178916275501251, "learning_rate": 6.069890274573157e-05, "loss": 0.0328, "step": 45980 }, { "grad_norm": 0.48882246017456055, "learning_rate": 6.068275043178679e-05, "loss": 0.045, "step": 45990 }, { "grad_norm": 0.23792491853237152, "learning_rate": 6.066659694959248e-05, "loss": 0.0374, "step": 46000 }, { "grad_norm": 0.6026122570037842, "learning_rate": 6.0650442300915176e-05, "loss": 0.0387, "step": 46010 }, { "grad_norm": 0.43408671021461487, "learning_rate": 6.063428648752152e-05, "loss": 0.0423, "step": 46020 }, { "grad_norm": 0.3627638816833496, "learning_rate": 6.0618129511178266e-05, "loss": 0.0396, "step": 46030 }, { "grad_norm": 0.3138355016708374, "learning_rate": 6.060197137365234e-05, "loss": 0.0355, "step": 46040 }, { "grad_norm": 0.3211866319179535, "learning_rate": 6.058581207671077e-05, "loss": 0.0318, "step": 46050 }, { "grad_norm": 0.4623412787914276, "learning_rate": 6.056965162212072e-05, "loss": 0.0368, "step": 46060 }, { "grad_norm": 0.28941187262535095, "learning_rate": 6.055349001164943e-05, "loss": 0.0379, "step": 46070 }, { "grad_norm": 0.33670279383659363, "learning_rate": 6.0537327247064347e-05, "loss": 0.0356, "step": 46080 }, { "grad_norm": 0.3195646405220032, "learning_rate": 6.0521163330133014e-05, "loss": 0.0326, "step": 46090 }, { "grad_norm": 0.25000908970832825, "learning_rate": 6.050499826262306e-05, "loss": 0.0344, "step": 46100 }, { "grad_norm": 0.6085093021392822, "learning_rate": 6.0488832046302294e-05, "loss": 0.0364, "step": 46110 }, { "grad_norm": 0.38337814807891846, "learning_rate": 6.0472664682938626e-05, "loss": 0.0394, "step": 46120 }, { "grad_norm": 0.530471920967102, "learning_rate": 6.045649617430009e-05, "loss": 0.0403, "step": 46130 }, { "grad_norm": 0.44809490442276, "learning_rate": 6.0440326522154866e-05, "loss": 0.0405, "step": 46140 }, { "grad_norm": 0.32579439878463745, "learning_rate": 6.0424155728271224e-05, "loss": 0.0365, "step": 46150 }, { "grad_norm": 0.6552694439888, "learning_rate": 6.040798379441758e-05, "loss": 0.0439, "step": 46160 }, { "grad_norm": 0.31741154193878174, "learning_rate": 6.0391810722362485e-05, "loss": 0.0377, "step": 46170 }, { "grad_norm": 0.3921191990375519, "learning_rate": 6.037563651387458e-05, "loss": 0.036, "step": 46180 }, { "grad_norm": 0.4105887711048126, "learning_rate": 6.0359461170722666e-05, "loss": 0.0395, "step": 46190 }, { "grad_norm": 0.3403388261795044, "learning_rate": 6.034328469467566e-05, "loss": 0.0374, "step": 46200 }, { "grad_norm": 0.7105156779289246, "learning_rate": 6.0327107087502596e-05, "loss": 0.0402, "step": 46210 }, { "grad_norm": 0.5157450437545776, "learning_rate": 6.031092835097262e-05, "loss": 0.0467, "step": 46220 }, { "grad_norm": 0.5082080364227295, "learning_rate": 6.0294748486855024e-05, "loss": 0.0348, "step": 46230 }, { "grad_norm": 0.3668663203716278, "learning_rate": 6.0278567496919216e-05, "loss": 0.0396, "step": 46240 }, { "grad_norm": 0.47186917066574097, "learning_rate": 6.026238538293472e-05, "loss": 0.0388, "step": 46250 }, { "grad_norm": 0.29958489537239075, "learning_rate": 6.024620214667118e-05, "loss": 0.0305, "step": 46260 }, { "grad_norm": 0.4129674434661865, "learning_rate": 6.0230017789898384e-05, "loss": 0.0315, "step": 46270 }, { "grad_norm": 0.5059648156166077, "learning_rate": 6.021383231438622e-05, "loss": 0.034, "step": 46280 }, { "grad_norm": 0.5895408987998962, "learning_rate": 6.0197645721904704e-05, "loss": 0.0358, "step": 46290 }, { "grad_norm": 0.4850279688835144, "learning_rate": 6.0181458014224e-05, "loss": 0.0413, "step": 46300 }, { "grad_norm": 0.3019753098487854, "learning_rate": 6.016526919311435e-05, "loss": 0.038, "step": 46310 }, { "grad_norm": 0.35694360733032227, "learning_rate": 6.014907926034613e-05, "loss": 0.0368, "step": 46320 }, { "grad_norm": 0.3049367070198059, "learning_rate": 6.0132888217689875e-05, "loss": 0.0379, "step": 46330 }, { "grad_norm": 0.450819730758667, "learning_rate": 6.01166960669162e-05, "loss": 0.0328, "step": 46340 }, { "grad_norm": 0.32088157534599304, "learning_rate": 6.0100502809795845e-05, "loss": 0.036, "step": 46350 }, { "grad_norm": 0.3728615343570709, "learning_rate": 6.00843084480997e-05, "loss": 0.0345, "step": 46360 }, { "grad_norm": 0.3502810597419739, "learning_rate": 6.0068112983598736e-05, "loss": 0.0372, "step": 46370 }, { "grad_norm": 0.2897825241088867, "learning_rate": 6.005191641806407e-05, "loss": 0.0371, "step": 46380 }, { "grad_norm": 0.39842987060546875, "learning_rate": 6.003571875326694e-05, "loss": 0.0388, "step": 46390 }, { "grad_norm": 0.4279651939868927, "learning_rate": 6.00195199909787e-05, "loss": 0.0473, "step": 46400 }, { "grad_norm": 0.34787556529045105, "learning_rate": 6.000332013297082e-05, "loss": 0.042, "step": 46410 }, { "grad_norm": 0.8360465168952942, "learning_rate": 5.998711918101487e-05, "loss": 0.041, "step": 46420 }, { "grad_norm": 0.750335693359375, "learning_rate": 5.997091713688261e-05, "loss": 0.0391, "step": 46430 }, { "grad_norm": 1.4995087385177612, "learning_rate": 5.9954714002345836e-05, "loss": 0.0348, "step": 46440 }, { "grad_norm": 0.35826414823532104, "learning_rate": 5.993850977917649e-05, "loss": 0.041, "step": 46450 }, { "grad_norm": 0.5017663836479187, "learning_rate": 5.992230446914667e-05, "loss": 0.0348, "step": 46460 }, { "grad_norm": 0.2322854846715927, "learning_rate": 5.990609807402855e-05, "loss": 0.0337, "step": 46470 }, { "grad_norm": 0.3894027769565582, "learning_rate": 5.988989059559443e-05, "loss": 0.0379, "step": 46480 }, { "grad_norm": 0.314618855714798, "learning_rate": 5.987368203561675e-05, "loss": 0.0385, "step": 46490 }, { "grad_norm": 0.4425313472747803, "learning_rate": 5.9857472395868055e-05, "loss": 0.0323, "step": 46500 }, { "grad_norm": 0.33172616362571716, "learning_rate": 5.9841261678120983e-05, "loss": 0.0349, "step": 46510 }, { "grad_norm": 0.3018462359905243, "learning_rate": 5.982504988414834e-05, "loss": 0.0384, "step": 46520 }, { "grad_norm": 1.2352244853973389, "learning_rate": 5.9808837015723015e-05, "loss": 0.0356, "step": 46530 }, { "grad_norm": 0.5429360866546631, "learning_rate": 5.9792623074618016e-05, "loss": 0.038, "step": 46540 }, { "grad_norm": 0.3370540738105774, "learning_rate": 5.977640806260648e-05, "loss": 0.0404, "step": 46550 }, { "grad_norm": 0.5166813135147095, "learning_rate": 5.976019198146166e-05, "loss": 0.0446, "step": 46560 }, { "grad_norm": 0.3685235381126404, "learning_rate": 5.974397483295692e-05, "loss": 0.033, "step": 46570 }, { "grad_norm": 0.49561217427253723, "learning_rate": 5.972775661886575e-05, "loss": 0.0403, "step": 46580 }, { "grad_norm": 0.32553356885910034, "learning_rate": 5.971153734096173e-05, "loss": 0.0359, "step": 46590 }, { "grad_norm": 0.5846852660179138, "learning_rate": 5.969531700101859e-05, "loss": 0.042, "step": 46600 }, { "grad_norm": 0.5723956823348999, "learning_rate": 5.9679095600810155e-05, "loss": 0.0362, "step": 46610 }, { "grad_norm": 0.427386611700058, "learning_rate": 5.9662873142110384e-05, "loss": 0.0398, "step": 46620 }, { "grad_norm": 0.3484792411327362, "learning_rate": 5.964664962669333e-05, "loss": 0.034, "step": 46630 }, { "grad_norm": 0.524667501449585, "learning_rate": 5.9630425056333186e-05, "loss": 0.0404, "step": 46640 }, { "grad_norm": 0.3365730345249176, "learning_rate": 5.961419943280422e-05, "loss": 0.036, "step": 46650 }, { "grad_norm": 0.34894540905952454, "learning_rate": 5.959797275788087e-05, "loss": 0.0355, "step": 46660 }, { "grad_norm": 0.4633653163909912, "learning_rate": 5.958174503333765e-05, "loss": 0.0388, "step": 46670 }, { "grad_norm": 0.41293370723724365, "learning_rate": 5.9565516260949195e-05, "loss": 0.0336, "step": 46680 }, { "grad_norm": 0.3406127393245697, "learning_rate": 5.9549286442490273e-05, "loss": 0.035, "step": 46690 }, { "grad_norm": 0.3435140550136566, "learning_rate": 5.953305557973572e-05, "loss": 0.0343, "step": 46700 }, { "grad_norm": 0.3173603415489197, "learning_rate": 5.9516823674460564e-05, "loss": 0.0396, "step": 46710 }, { "grad_norm": 0.46602296829223633, "learning_rate": 5.9500590728439875e-05, "loss": 0.0379, "step": 46720 }, { "grad_norm": 0.37523460388183594, "learning_rate": 5.948435674344886e-05, "loss": 0.0323, "step": 46730 }, { "grad_norm": 0.40243837237358093, "learning_rate": 5.946812172126285e-05, "loss": 0.035, "step": 46740 }, { "grad_norm": 0.4867808222770691, "learning_rate": 5.945188566365729e-05, "loss": 0.0412, "step": 46750 }, { "grad_norm": 0.45809364318847656, "learning_rate": 5.943564857240773e-05, "loss": 0.0448, "step": 46760 }, { "grad_norm": 0.24592600762844086, "learning_rate": 5.941941044928981e-05, "loss": 0.0342, "step": 46770 }, { "grad_norm": 0.2823702394962311, "learning_rate": 5.940317129607935e-05, "loss": 0.0375, "step": 46780 }, { "grad_norm": 0.3741287887096405, "learning_rate": 5.9386931114552204e-05, "loss": 0.0307, "step": 46790 }, { "grad_norm": 0.2672373950481415, "learning_rate": 5.937068990648438e-05, "loss": 0.0354, "step": 46800 }, { "grad_norm": 0.336897075176239, "learning_rate": 5.935444767365199e-05, "loss": 0.0405, "step": 46810 }, { "grad_norm": 0.36492258310317993, "learning_rate": 5.933820441783129e-05, "loss": 0.03, "step": 46820 }, { "grad_norm": 0.3041551113128662, "learning_rate": 5.932196014079857e-05, "loss": 0.0403, "step": 46830 }, { "grad_norm": 0.43205538392066956, "learning_rate": 5.930571484433032e-05, "loss": 0.0434, "step": 46840 }, { "grad_norm": 0.49663612246513367, "learning_rate": 5.928946853020309e-05, "loss": 0.0404, "step": 46850 }, { "grad_norm": 0.34280917048454285, "learning_rate": 5.927322120019352e-05, "loss": 0.0353, "step": 46860 }, { "grad_norm": 0.3722216784954071, "learning_rate": 5.925697285607844e-05, "loss": 0.0307, "step": 46870 }, { "grad_norm": 0.29815229773521423, "learning_rate": 5.924072349963472e-05, "loss": 0.0354, "step": 46880 }, { "grad_norm": 0.34481990337371826, "learning_rate": 5.922447313263938e-05, "loss": 0.0403, "step": 46890 }, { "grad_norm": 0.5668580532073975, "learning_rate": 5.920822175686952e-05, "loss": 0.0339, "step": 46900 }, { "grad_norm": 0.3486260771751404, "learning_rate": 5.919196937410239e-05, "loss": 0.0377, "step": 46910 }, { "grad_norm": 0.5033727884292603, "learning_rate": 5.917571598611529e-05, "loss": 0.0415, "step": 46920 }, { "grad_norm": 0.35362887382507324, "learning_rate": 5.9159461594685686e-05, "loss": 0.0371, "step": 46930 }, { "grad_norm": 0.416374146938324, "learning_rate": 5.914320620159114e-05, "loss": 0.041, "step": 46940 }, { "grad_norm": 0.33144253492355347, "learning_rate": 5.912694980860931e-05, "loss": 0.0367, "step": 46950 }, { "grad_norm": 0.28553640842437744, "learning_rate": 5.9110692417517964e-05, "loss": 0.0302, "step": 46960 }, { "grad_norm": 0.38221168518066406, "learning_rate": 5.9094434030095e-05, "loss": 0.0326, "step": 46970 }, { "grad_norm": 0.4054659903049469, "learning_rate": 5.907817464811841e-05, "loss": 0.0288, "step": 46980 }, { "grad_norm": 0.4351066052913666, "learning_rate": 5.906191427336627e-05, "loss": 0.0365, "step": 46990 }, { "grad_norm": 0.5221766233444214, "learning_rate": 5.9045652907616835e-05, "loss": 0.042, "step": 47000 }, { "grad_norm": 0.3920450210571289, "learning_rate": 5.902939055264838e-05, "loss": 0.0395, "step": 47010 }, { "grad_norm": 0.3725632429122925, "learning_rate": 5.901312721023935e-05, "loss": 0.035, "step": 47020 }, { "grad_norm": 0.23148033022880554, "learning_rate": 5.8996862882168294e-05, "loss": 0.034, "step": 47030 }, { "grad_norm": 0.352850079536438, "learning_rate": 5.8980597570213826e-05, "loss": 0.0348, "step": 47040 }, { "grad_norm": 0.4044662415981293, "learning_rate": 5.896433127615471e-05, "loss": 0.0346, "step": 47050 }, { "grad_norm": 0.5548009276390076, "learning_rate": 5.894806400176981e-05, "loss": 0.0444, "step": 47060 }, { "grad_norm": 0.3082466423511505, "learning_rate": 5.893179574883808e-05, "loss": 0.0367, "step": 47070 }, { "grad_norm": 0.5684419870376587, "learning_rate": 5.8915526519138585e-05, "loss": 0.0342, "step": 47080 }, { "grad_norm": 0.35960251092910767, "learning_rate": 5.889925631445053e-05, "loss": 0.039, "step": 47090 }, { "grad_norm": 0.4125441014766693, "learning_rate": 5.888298513655318e-05, "loss": 0.0345, "step": 47100 }, { "grad_norm": 0.33765628933906555, "learning_rate": 5.886671298722592e-05, "loss": 0.0355, "step": 47110 }, { "grad_norm": 0.5475218892097473, "learning_rate": 5.885043986824828e-05, "loss": 0.0384, "step": 47120 }, { "grad_norm": 0.4182293117046356, "learning_rate": 5.883416578139982e-05, "loss": 0.0383, "step": 47130 }, { "grad_norm": 0.31334105134010315, "learning_rate": 5.881789072846029e-05, "loss": 0.0297, "step": 47140 }, { "grad_norm": 0.2933107316493988, "learning_rate": 5.8801614711209474e-05, "loss": 0.0419, "step": 47150 }, { "grad_norm": 0.4361640512943268, "learning_rate": 5.878533773142732e-05, "loss": 0.0487, "step": 47160 }, { "grad_norm": 0.3701564371585846, "learning_rate": 5.8769059790893834e-05, "loss": 0.0375, "step": 47170 }, { "grad_norm": 0.48821213841438293, "learning_rate": 5.8752780891389145e-05, "loss": 0.0351, "step": 47180 }, { "grad_norm": 0.47758162021636963, "learning_rate": 5.873650103469351e-05, "loss": 0.0419, "step": 47190 }, { "grad_norm": 0.47967734932899475, "learning_rate": 5.872022022258725e-05, "loss": 0.0352, "step": 47200 }, { "grad_norm": 0.35658836364746094, "learning_rate": 5.8703938456850814e-05, "loss": 0.037, "step": 47210 }, { "grad_norm": 0.36440080404281616, "learning_rate": 5.868765573926477e-05, "loss": 0.0349, "step": 47220 }, { "grad_norm": 0.34227651357650757, "learning_rate": 5.867137207160974e-05, "loss": 0.0307, "step": 47230 }, { "grad_norm": 0.4372599124908447, "learning_rate": 5.86550874556665e-05, "loss": 0.0354, "step": 47240 }, { "grad_norm": 0.5185545086860657, "learning_rate": 5.863880189321592e-05, "loss": 0.0356, "step": 47250 }, { "grad_norm": 0.4463267922401428, "learning_rate": 5.862251538603894e-05, "loss": 0.0343, "step": 47260 }, { "grad_norm": 0.3650536835193634, "learning_rate": 5.8606227935916656e-05, "loss": 0.0335, "step": 47270 }, { "grad_norm": 0.5517843961715698, "learning_rate": 5.858993954463021e-05, "loss": 0.0312, "step": 47280 }, { "grad_norm": 0.4315245747566223, "learning_rate": 5.8573650213960896e-05, "loss": 0.0354, "step": 47290 }, { "grad_norm": 0.28273090720176697, "learning_rate": 5.8557359945690084e-05, "loss": 0.0292, "step": 47300 }, { "grad_norm": 0.3331846296787262, "learning_rate": 5.8541068741599246e-05, "loss": 0.0331, "step": 47310 }, { "grad_norm": 0.39600852131843567, "learning_rate": 5.8524776603469985e-05, "loss": 0.0353, "step": 47320 }, { "grad_norm": 0.2848028838634491, "learning_rate": 5.850848353308397e-05, "loss": 0.0384, "step": 47330 }, { "grad_norm": 0.3724631071090698, "learning_rate": 5.849218953222297e-05, "loss": 0.0419, "step": 47340 }, { "grad_norm": 0.29810142517089844, "learning_rate": 5.847589460266891e-05, "loss": 0.0344, "step": 47350 }, { "grad_norm": 0.37099847197532654, "learning_rate": 5.845959874620375e-05, "loss": 0.029, "step": 47360 }, { "grad_norm": 0.42903685569763184, "learning_rate": 5.84433019646096e-05, "loss": 0.0352, "step": 47370 }, { "grad_norm": 0.34362852573394775, "learning_rate": 5.842700425966863e-05, "loss": 0.0349, "step": 47380 }, { "grad_norm": 0.44713738560676575, "learning_rate": 5.841070563316315e-05, "loss": 0.0345, "step": 47390 }, { "grad_norm": 0.41000109910964966, "learning_rate": 5.839440608687554e-05, "loss": 0.0373, "step": 47400 }, { "grad_norm": 0.5572345852851868, "learning_rate": 5.837810562258831e-05, "loss": 0.0397, "step": 47410 }, { "grad_norm": 0.45557066798210144, "learning_rate": 5.836180424208405e-05, "loss": 0.0362, "step": 47420 }, { "grad_norm": 0.40570759773254395, "learning_rate": 5.834550194714543e-05, "loss": 0.0372, "step": 47430 }, { "grad_norm": 0.33742207288742065, "learning_rate": 5.832919873955528e-05, "loss": 0.0393, "step": 47440 }, { "grad_norm": 0.44683757424354553, "learning_rate": 5.831289462109648e-05, "loss": 0.0332, "step": 47450 }, { "grad_norm": 0.21680286526679993, "learning_rate": 5.829658959355202e-05, "loss": 0.036, "step": 47460 }, { "grad_norm": 0.358693927526474, "learning_rate": 5.8280283658704995e-05, "loss": 0.0329, "step": 47470 }, { "grad_norm": 0.270440012216568, "learning_rate": 5.826397681833859e-05, "loss": 0.0416, "step": 47480 }, { "grad_norm": 0.3830929398536682, "learning_rate": 5.824766907423612e-05, "loss": 0.0339, "step": 47490 }, { "grad_norm": 0.3443821370601654, "learning_rate": 5.8231360428180945e-05, "loss": 0.0303, "step": 47500 }, { "grad_norm": 0.2895644009113312, "learning_rate": 5.821505088195658e-05, "loss": 0.0353, "step": 47510 }, { "grad_norm": 0.36291196942329407, "learning_rate": 5.819874043734661e-05, "loss": 0.0374, "step": 47520 }, { "grad_norm": 0.41242966055870056, "learning_rate": 5.8182429096134695e-05, "loss": 0.0349, "step": 47530 }, { "grad_norm": 0.4426429867744446, "learning_rate": 5.816611686010465e-05, "loss": 0.0405, "step": 47540 }, { "grad_norm": 0.5265633463859558, "learning_rate": 5.814980373104033e-05, "loss": 0.0364, "step": 47550 }, { "grad_norm": 0.28749942779541016, "learning_rate": 5.813348971072572e-05, "loss": 0.0388, "step": 47560 }, { "grad_norm": 0.37429627776145935, "learning_rate": 5.811717480094492e-05, "loss": 0.0395, "step": 47570 }, { "grad_norm": 0.34707576036453247, "learning_rate": 5.810085900348209e-05, "loss": 0.0471, "step": 47580 }, { "grad_norm": 0.22786575555801392, "learning_rate": 5.8084542320121483e-05, "loss": 0.0405, "step": 47590 }, { "grad_norm": 0.45042136311531067, "learning_rate": 5.8068224752647497e-05, "loss": 0.0395, "step": 47600 }, { "grad_norm": 0.7307100892066956, "learning_rate": 5.805190630284457e-05, "loss": 0.0436, "step": 47610 }, { "grad_norm": 0.3236772418022156, "learning_rate": 5.803558697249729e-05, "loss": 0.0452, "step": 47620 }, { "grad_norm": 0.4169257581233978, "learning_rate": 5.8019266763390276e-05, "loss": 0.035, "step": 47630 }, { "grad_norm": 0.3122507631778717, "learning_rate": 5.800294567730833e-05, "loss": 0.036, "step": 47640 }, { "grad_norm": 0.28661468625068665, "learning_rate": 5.798662371603628e-05, "loss": 0.0342, "step": 47650 }, { "grad_norm": 0.3865243196487427, "learning_rate": 5.797030088135904e-05, "loss": 0.0342, "step": 47660 }, { "grad_norm": 0.3594428598880768, "learning_rate": 5.7953977175061705e-05, "loss": 0.0314, "step": 47670 }, { "grad_norm": 0.348606139421463, "learning_rate": 5.793765259892938e-05, "loss": 0.0395, "step": 47680 }, { "grad_norm": 0.3881162106990814, "learning_rate": 5.792132715474729e-05, "loss": 0.0364, "step": 47690 }, { "grad_norm": 0.3967355489730835, "learning_rate": 5.790500084430078e-05, "loss": 0.0334, "step": 47700 }, { "grad_norm": 0.6253049373626709, "learning_rate": 5.7888673669375274e-05, "loss": 0.0408, "step": 47710 }, { "grad_norm": 0.3895057737827301, "learning_rate": 5.787234563175625e-05, "loss": 0.0366, "step": 47720 }, { "grad_norm": 0.4171830415725708, "learning_rate": 5.7856016733229355e-05, "loss": 0.0351, "step": 47730 }, { "grad_norm": 0.5597860217094421, "learning_rate": 5.7839686975580297e-05, "loss": 0.0411, "step": 47740 }, { "grad_norm": 0.42974746227264404, "learning_rate": 5.782335636059484e-05, "loss": 0.0346, "step": 47750 }, { "grad_norm": 0.34601250290870667, "learning_rate": 5.780702489005889e-05, "loss": 0.0391, "step": 47760 }, { "grad_norm": 0.2997110188007355, "learning_rate": 5.779069256575845e-05, "loss": 0.035, "step": 47770 }, { "grad_norm": 0.30372869968414307, "learning_rate": 5.7774359389479574e-05, "loss": 0.03, "step": 47780 }, { "grad_norm": 0.3427221179008484, "learning_rate": 5.775802536300845e-05, "loss": 0.0391, "step": 47790 }, { "grad_norm": 0.5221406817436218, "learning_rate": 5.774169048813134e-05, "loss": 0.0371, "step": 47800 }, { "grad_norm": 0.36181357502937317, "learning_rate": 5.77253547666346e-05, "loss": 0.0358, "step": 47810 }, { "grad_norm": 0.41126948595046997, "learning_rate": 5.770901820030465e-05, "loss": 0.0356, "step": 47820 }, { "grad_norm": 0.3241260349750519, "learning_rate": 5.769268079092809e-05, "loss": 0.0386, "step": 47830 }, { "grad_norm": 0.28411903977394104, "learning_rate": 5.767634254029151e-05, "loss": 0.0368, "step": 47840 }, { "grad_norm": 0.4548788070678711, "learning_rate": 5.7660003450181655e-05, "loss": 0.039, "step": 47850 }, { "grad_norm": 0.4200042188167572, "learning_rate": 5.764366352238534e-05, "loss": 0.0345, "step": 47860 }, { "grad_norm": 0.34404250979423523, "learning_rate": 5.7627322758689474e-05, "loss": 0.0372, "step": 47870 }, { "grad_norm": 0.6607881188392639, "learning_rate": 5.761098116088105e-05, "loss": 0.0334, "step": 47880 }, { "grad_norm": 0.5728980898857117, "learning_rate": 5.759463873074717e-05, "loss": 0.0451, "step": 47890 }, { "grad_norm": 0.45158952474594116, "learning_rate": 5.757829547007504e-05, "loss": 0.0397, "step": 47900 }, { "grad_norm": 0.4471808671951294, "learning_rate": 5.756195138065189e-05, "loss": 0.0379, "step": 47910 }, { "grad_norm": 0.5039457678794861, "learning_rate": 5.754560646426511e-05, "loss": 0.0348, "step": 47920 }, { "grad_norm": 0.3189406096935272, "learning_rate": 5.752926072270216e-05, "loss": 0.0322, "step": 47930 }, { "grad_norm": 0.3263135552406311, "learning_rate": 5.7512914157750563e-05, "loss": 0.0339, "step": 47940 }, { "grad_norm": 0.2447393834590912, "learning_rate": 5.749656677119798e-05, "loss": 0.0341, "step": 47950 }, { "grad_norm": 0.5671356916427612, "learning_rate": 5.7480218564832125e-05, "loss": 0.0378, "step": 47960 }, { "grad_norm": 0.3340681791305542, "learning_rate": 5.746386954044082e-05, "loss": 0.0308, "step": 47970 }, { "grad_norm": 0.30845993757247925, "learning_rate": 5.744751969981195e-05, "loss": 0.0353, "step": 47980 }, { "grad_norm": 0.33682945370674133, "learning_rate": 5.7431169044733526e-05, "loss": 0.0369, "step": 47990 }, { "grad_norm": 0.2852812111377716, "learning_rate": 5.741481757699364e-05, "loss": 0.0291, "step": 48000 }, { "grad_norm": 0.34973040223121643, "learning_rate": 5.7398465298380434e-05, "loss": 0.0363, "step": 48010 }, { "grad_norm": 0.2647886574268341, "learning_rate": 5.7382112210682193e-05, "loss": 0.0304, "step": 48020 }, { "grad_norm": 0.6757999658584595, "learning_rate": 5.7365758315687266e-05, "loss": 0.038, "step": 48030 }, { "grad_norm": 0.31074559688568115, "learning_rate": 5.734940361518407e-05, "loss": 0.0303, "step": 48040 }, { "grad_norm": 0.3496989905834198, "learning_rate": 5.733304811096116e-05, "loss": 0.0367, "step": 48050 }, { "grad_norm": 0.5023359656333923, "learning_rate": 5.731669180480713e-05, "loss": 0.0315, "step": 48060 }, { "grad_norm": 0.372952401638031, "learning_rate": 5.730033469851067e-05, "loss": 0.0353, "step": 48070 }, { "grad_norm": 0.4951840043067932, "learning_rate": 5.7283976793860607e-05, "loss": 0.0359, "step": 48080 }, { "grad_norm": 0.3334805965423584, "learning_rate": 5.726761809264577e-05, "loss": 0.0377, "step": 48090 }, { "grad_norm": 0.3144255578517914, "learning_rate": 5.7251258596655155e-05, "loss": 0.0324, "step": 48100 }, { "grad_norm": 0.8762198090553284, "learning_rate": 5.72348983076778e-05, "loss": 0.0366, "step": 48110 }, { "grad_norm": 0.8992159366607666, "learning_rate": 5.7218537227502854e-05, "loss": 0.0433, "step": 48120 }, { "grad_norm": 0.4019979238510132, "learning_rate": 5.720217535791951e-05, "loss": 0.0392, "step": 48130 }, { "grad_norm": 0.46603021025657654, "learning_rate": 5.718581270071711e-05, "loss": 0.0362, "step": 48140 }, { "grad_norm": 0.34558725357055664, "learning_rate": 5.716944925768505e-05, "loss": 0.0341, "step": 48150 }, { "grad_norm": 0.3902510106563568, "learning_rate": 5.7153085030612786e-05, "loss": 0.0376, "step": 48160 }, { "grad_norm": 0.44597700238227844, "learning_rate": 5.713672002128989e-05, "loss": 0.0383, "step": 48170 }, { "grad_norm": 0.5889028310775757, "learning_rate": 5.712035423150602e-05, "loss": 0.0404, "step": 48180 }, { "grad_norm": 0.45203453302383423, "learning_rate": 5.710398766305094e-05, "loss": 0.0347, "step": 48190 }, { "grad_norm": 0.4056242108345032, "learning_rate": 5.708762031771442e-05, "loss": 0.0336, "step": 48200 }, { "grad_norm": 0.3630967438220978, "learning_rate": 5.707125219728642e-05, "loss": 0.0336, "step": 48210 }, { "grad_norm": 0.256766676902771, "learning_rate": 5.7054883303556905e-05, "loss": 0.0319, "step": 48220 }, { "grad_norm": 0.3360600471496582, "learning_rate": 5.703851363831595e-05, "loss": 0.0311, "step": 48230 }, { "grad_norm": 0.37628400325775146, "learning_rate": 5.702214320335374e-05, "loss": 0.0337, "step": 48240 }, { "grad_norm": 0.5314980745315552, "learning_rate": 5.7005772000460514e-05, "loss": 0.0385, "step": 48250 }, { "grad_norm": 0.3499305248260498, "learning_rate": 5.698940003142656e-05, "loss": 0.0399, "step": 48260 }, { "grad_norm": 0.47516560554504395, "learning_rate": 5.697302729804236e-05, "loss": 0.0353, "step": 48270 }, { "grad_norm": 0.388257771730423, "learning_rate": 5.695665380209837e-05, "loss": 0.0351, "step": 48280 }, { "grad_norm": 0.38816267251968384, "learning_rate": 5.6940279545385165e-05, "loss": 0.0366, "step": 48290 }, { "grad_norm": 0.3472563922405243, "learning_rate": 5.692390452969344e-05, "loss": 0.0382, "step": 48300 }, { "grad_norm": 0.46151670813560486, "learning_rate": 5.690752875681392e-05, "loss": 0.0359, "step": 48310 }, { "grad_norm": 0.4940412640571594, "learning_rate": 5.6891152228537435e-05, "loss": 0.0415, "step": 48320 }, { "grad_norm": 0.4458955228328705, "learning_rate": 5.687477494665492e-05, "loss": 0.0336, "step": 48330 }, { "grad_norm": 0.3428913354873657, "learning_rate": 5.685839691295733e-05, "loss": 0.0313, "step": 48340 }, { "grad_norm": 0.5209921598434448, "learning_rate": 5.6842018129235786e-05, "loss": 0.0341, "step": 48350 }, { "grad_norm": 0.29096314311027527, "learning_rate": 5.6825638597281404e-05, "loss": 0.0416, "step": 48360 }, { "grad_norm": 0.4165002107620239, "learning_rate": 5.680925831888546e-05, "loss": 0.0401, "step": 48370 }, { "grad_norm": 0.4818289577960968, "learning_rate": 5.6792877295839274e-05, "loss": 0.0361, "step": 48380 }, { "grad_norm": 0.39972734451293945, "learning_rate": 5.6776495529934224e-05, "loss": 0.0362, "step": 48390 }, { "grad_norm": 0.3834535479545593, "learning_rate": 5.6760113022961824e-05, "loss": 0.0356, "step": 48400 }, { "grad_norm": 0.3759779930114746, "learning_rate": 5.6743729776713617e-05, "loss": 0.0367, "step": 48410 }, { "grad_norm": 0.44943249225616455, "learning_rate": 5.672734579298126e-05, "loss": 0.0332, "step": 48420 }, { "grad_norm": 0.4434659481048584, "learning_rate": 5.671096107355649e-05, "loss": 0.0344, "step": 48430 }, { "grad_norm": 0.5436574220657349, "learning_rate": 5.66945756202311e-05, "loss": 0.0429, "step": 48440 }, { "grad_norm": 0.42857807874679565, "learning_rate": 5.667818943479699e-05, "loss": 0.0365, "step": 48450 }, { "grad_norm": 0.35228458046913147, "learning_rate": 5.666180251904612e-05, "loss": 0.0346, "step": 48460 }, { "grad_norm": 0.6192077398300171, "learning_rate": 5.6645414874770555e-05, "loss": 0.0345, "step": 48470 }, { "grad_norm": 0.35588258504867554, "learning_rate": 5.66290265037624e-05, "loss": 0.0374, "step": 48480 }, { "grad_norm": 1.6506915092468262, "learning_rate": 5.661263740781386e-05, "loss": 0.0387, "step": 48490 }, { "grad_norm": 0.4973122477531433, "learning_rate": 5.6596247588717254e-05, "loss": 0.035, "step": 48500 }, { "grad_norm": 0.38949310779571533, "learning_rate": 5.6579857048264926e-05, "loss": 0.0436, "step": 48510 }, { "grad_norm": 0.3999602198600769, "learning_rate": 5.6563465788249314e-05, "loss": 0.0331, "step": 48520 }, { "grad_norm": 0.5626062750816345, "learning_rate": 5.6547073810462956e-05, "loss": 0.0332, "step": 48530 }, { "grad_norm": 0.23697331547737122, "learning_rate": 5.653068111669846e-05, "loss": 0.0343, "step": 48540 }, { "grad_norm": 0.4639265835285187, "learning_rate": 5.651428770874848e-05, "loss": 0.0353, "step": 48550 }, { "grad_norm": 0.27157723903656006, "learning_rate": 5.64978935884058e-05, "loss": 0.0291, "step": 48560 }, { "grad_norm": 0.363762229681015, "learning_rate": 5.6481498757463244e-05, "loss": 0.0377, "step": 48570 }, { "grad_norm": 0.35711005330085754, "learning_rate": 5.646510321771373e-05, "loss": 0.0411, "step": 48580 }, { "grad_norm": 0.340152382850647, "learning_rate": 5.644870697095024e-05, "loss": 0.0334, "step": 48590 }, { "grad_norm": 0.4415931701660156, "learning_rate": 5.643231001896586e-05, "loss": 0.0317, "step": 48600 }, { "grad_norm": 0.3057234287261963, "learning_rate": 5.6415912363553726e-05, "loss": 0.0355, "step": 48610 }, { "grad_norm": 0.35511666536331177, "learning_rate": 5.639951400650706e-05, "loss": 0.0406, "step": 48620 }, { "grad_norm": 0.2440229207277298, "learning_rate": 5.6383114949619165e-05, "loss": 0.0355, "step": 48630 }, { "grad_norm": 0.3209347724914551, "learning_rate": 5.636671519468342e-05, "loss": 0.0366, "step": 48640 }, { "grad_norm": 0.3939196467399597, "learning_rate": 5.635031474349327e-05, "loss": 0.031, "step": 48650 }, { "grad_norm": 0.506464958190918, "learning_rate": 5.6333913597842246e-05, "loss": 0.0353, "step": 48660 }, { "grad_norm": 0.3937678337097168, "learning_rate": 5.6317511759523955e-05, "loss": 0.0339, "step": 48670 }, { "grad_norm": 0.34449514746665955, "learning_rate": 5.630110923033207e-05, "loss": 0.0361, "step": 48680 }, { "grad_norm": 0.6478973627090454, "learning_rate": 5.628470601206036e-05, "loss": 0.0303, "step": 48690 }, { "grad_norm": 0.2566471993923187, "learning_rate": 5.626830210650266e-05, "loss": 0.0335, "step": 48700 }, { "grad_norm": 0.27693888545036316, "learning_rate": 5.625189751545285e-05, "loss": 0.0379, "step": 48710 }, { "grad_norm": 0.4292081296443939, "learning_rate": 5.6235492240704936e-05, "loss": 0.0403, "step": 48720 }, { "grad_norm": 0.363808810710907, "learning_rate": 5.621908628405296e-05, "loss": 0.0398, "step": 48730 }, { "grad_norm": 0.39766180515289307, "learning_rate": 5.620267964729106e-05, "loss": 0.0401, "step": 48740 }, { "grad_norm": 0.4084400534629822, "learning_rate": 5.618627233221344e-05, "loss": 0.0372, "step": 48750 }, { "grad_norm": 0.34089013934135437, "learning_rate": 5.616986434061438e-05, "loss": 0.0341, "step": 48760 }, { "grad_norm": 0.4857078492641449, "learning_rate": 5.615345567428822e-05, "loss": 0.0368, "step": 48770 }, { "grad_norm": 0.4203222393989563, "learning_rate": 5.613704633502941e-05, "loss": 0.0347, "step": 48780 }, { "grad_norm": 0.39580807089805603, "learning_rate": 5.612063632463245e-05, "loss": 0.0369, "step": 48790 }, { "grad_norm": 0.41738054156303406, "learning_rate": 5.610422564489188e-05, "loss": 0.0371, "step": 48800 }, { "grad_norm": 0.3999190628528595, "learning_rate": 5.608781429760239e-05, "loss": 0.0314, "step": 48810 }, { "grad_norm": 0.42426279187202454, "learning_rate": 5.607140228455866e-05, "loss": 0.0381, "step": 48820 }, { "grad_norm": 0.27517470717430115, "learning_rate": 5.605498960755553e-05, "loss": 0.039, "step": 48830 }, { "grad_norm": 0.2802095115184784, "learning_rate": 5.603857626838782e-05, "loss": 0.0315, "step": 48840 }, { "grad_norm": 0.3639681339263916, "learning_rate": 5.60221622688505e-05, "loss": 0.0328, "step": 48850 }, { "grad_norm": 0.42261409759521484, "learning_rate": 5.6005747610738565e-05, "loss": 0.0351, "step": 48860 }, { "grad_norm": 0.5426881313323975, "learning_rate": 5.59893322958471e-05, "loss": 0.0373, "step": 48870 }, { "grad_norm": 0.4370678663253784, "learning_rate": 5.5972916325971256e-05, "loss": 0.0376, "step": 48880 }, { "grad_norm": 0.2846969664096832, "learning_rate": 5.595649970290628e-05, "loss": 0.0341, "step": 48890 }, { "grad_norm": 0.4123368263244629, "learning_rate": 5.5940082428447426e-05, "loss": 0.0437, "step": 48900 }, { "grad_norm": 0.5340256094932556, "learning_rate": 5.592366450439012e-05, "loss": 0.0352, "step": 48910 }, { "grad_norm": 0.28559380769729614, "learning_rate": 5.590724593252975e-05, "loss": 0.0348, "step": 48920 }, { "grad_norm": 0.25458985567092896, "learning_rate": 5.589082671466184e-05, "loss": 0.0348, "step": 48930 }, { "grad_norm": 0.29115787148475647, "learning_rate": 5.587440685258199e-05, "loss": 0.0337, "step": 48940 }, { "grad_norm": 0.41416752338409424, "learning_rate": 5.585798634808583e-05, "loss": 0.04, "step": 48950 }, { "grad_norm": 0.33051598072052, "learning_rate": 5.584156520296909e-05, "loss": 0.0359, "step": 48960 }, { "grad_norm": 0.43382975459098816, "learning_rate": 5.582514341902757e-05, "loss": 0.0414, "step": 48970 }, { "grad_norm": 0.38069450855255127, "learning_rate": 5.580872099805713e-05, "loss": 0.0408, "step": 48980 }, { "grad_norm": 0.3565872013568878, "learning_rate": 5.5792297941853674e-05, "loss": 0.0406, "step": 48990 }, { "grad_norm": 0.3652602732181549, "learning_rate": 5.5775874252213247e-05, "loss": 0.0417, "step": 49000 }, { "grad_norm": 0.2604183256626129, "learning_rate": 5.575944993093189e-05, "loss": 0.0327, "step": 49010 }, { "grad_norm": 0.33032941818237305, "learning_rate": 5.574302497980574e-05, "loss": 0.0402, "step": 49020 }, { "grad_norm": 0.33477258682250977, "learning_rate": 5.5726599400631005e-05, "loss": 0.032, "step": 49030 }, { "grad_norm": 0.41512396931648254, "learning_rate": 5.571017319520399e-05, "loss": 0.0418, "step": 49040 }, { "grad_norm": 0.526858925819397, "learning_rate": 5.5693746365320984e-05, "loss": 0.0373, "step": 49050 }, { "grad_norm": 0.49138742685317993, "learning_rate": 5.567731891277848e-05, "loss": 0.0361, "step": 49060 }, { "grad_norm": 0.35183513164520264, "learning_rate": 5.566089083937288e-05, "loss": 0.0379, "step": 49070 }, { "grad_norm": 0.36900267004966736, "learning_rate": 5.564446214690079e-05, "loss": 0.0375, "step": 49080 }, { "grad_norm": 0.40271809697151184, "learning_rate": 5.56280328371588e-05, "loss": 0.0365, "step": 49090 }, { "grad_norm": 0.3102308511734009, "learning_rate": 5.5611602911943596e-05, "loss": 0.0366, "step": 49100 }, { "grad_norm": 0.468866765499115, "learning_rate": 5.559517237305194e-05, "loss": 0.0385, "step": 49110 }, { "grad_norm": 0.3840336203575134, "learning_rate": 5.557874122228064e-05, "loss": 0.0388, "step": 49120 }, { "grad_norm": 0.3135524094104767, "learning_rate": 5.55623094614266e-05, "loss": 0.0406, "step": 49130 }, { "grad_norm": 0.3903387784957886, "learning_rate": 5.5545877092286743e-05, "loss": 0.0385, "step": 49140 }, { "grad_norm": 0.2999500334262848, "learning_rate": 5.552944411665811e-05, "loss": 0.0345, "step": 49150 }, { "grad_norm": 0.40345433354377747, "learning_rate": 5.5513010536337795e-05, "loss": 0.0304, "step": 49160 }, { "grad_norm": 0.4882354140281677, "learning_rate": 5.5496576353122934e-05, "loss": 0.0355, "step": 49170 }, { "grad_norm": 0.32632124423980713, "learning_rate": 5.548014156881074e-05, "loss": 0.0339, "step": 49180 }, { "grad_norm": 0.23794794082641602, "learning_rate": 5.5463706185198494e-05, "loss": 0.0374, "step": 49190 }, { "grad_norm": 0.39881157875061035, "learning_rate": 5.544727020408358e-05, "loss": 0.0381, "step": 49200 }, { "grad_norm": 0.433749794960022, "learning_rate": 5.5430833627263366e-05, "loss": 0.0324, "step": 49210 }, { "grad_norm": 0.34923863410949707, "learning_rate": 5.5414396456535354e-05, "loss": 0.0346, "step": 49220 }, { "grad_norm": 0.4403200149536133, "learning_rate": 5.539795869369709e-05, "loss": 0.0402, "step": 49230 }, { "grad_norm": 0.3580209016799927, "learning_rate": 5.538152034054618e-05, "loss": 0.0308, "step": 49240 }, { "grad_norm": 0.3161790072917938, "learning_rate": 5.536508139888028e-05, "loss": 0.0354, "step": 49250 }, { "grad_norm": 0.41558918356895447, "learning_rate": 5.534864187049716e-05, "loss": 0.0368, "step": 49260 }, { "grad_norm": 0.356151819229126, "learning_rate": 5.5332201757194615e-05, "loss": 0.0326, "step": 49270 }, { "grad_norm": 0.3504892587661743, "learning_rate": 5.531576106077049e-05, "loss": 0.0368, "step": 49280 }, { "grad_norm": 0.4781608283519745, "learning_rate": 5.529931978302272e-05, "loss": 0.0334, "step": 49290 }, { "grad_norm": 0.3645978271961212, "learning_rate": 5.528287792574932e-05, "loss": 0.0379, "step": 49300 }, { "grad_norm": 1.0205693244934082, "learning_rate": 5.5266435490748324e-05, "loss": 0.0412, "step": 49310 }, { "grad_norm": 0.4107702374458313, "learning_rate": 5.524999247981787e-05, "loss": 0.0345, "step": 49320 }, { "grad_norm": 0.42493751645088196, "learning_rate": 5.523354889475613e-05, "loss": 0.0358, "step": 49330 }, { "grad_norm": 0.347211092710495, "learning_rate": 5.521710473736134e-05, "loss": 0.0333, "step": 49340 }, { "grad_norm": 0.988835871219635, "learning_rate": 5.5200660009431826e-05, "loss": 0.04, "step": 49350 }, { "grad_norm": 0.3632473051548004, "learning_rate": 5.518421471276596e-05, "loss": 0.0396, "step": 49360 }, { "grad_norm": 0.4468989372253418, "learning_rate": 5.5167768849162174e-05, "loss": 0.0411, "step": 49370 }, { "grad_norm": 0.35814520716667175, "learning_rate": 5.515132242041893e-05, "loss": 0.0341, "step": 49380 }, { "grad_norm": 0.531050980091095, "learning_rate": 5.513487542833483e-05, "loss": 0.0344, "step": 49390 }, { "grad_norm": 0.357476145029068, "learning_rate": 5.5118427874708466e-05, "loss": 0.0343, "step": 49400 }, { "grad_norm": 0.4147936999797821, "learning_rate": 5.510197976133853e-05, "loss": 0.0319, "step": 49410 }, { "grad_norm": 0.37349608540534973, "learning_rate": 5.508553109002376e-05, "loss": 0.0371, "step": 49420 }, { "grad_norm": 0.3673451542854309, "learning_rate": 5.5069081862562957e-05, "loss": 0.0326, "step": 49430 }, { "grad_norm": 0.4981631636619568, "learning_rate": 5.5052632080754965e-05, "loss": 0.0467, "step": 49440 }, { "grad_norm": 0.4100988507270813, "learning_rate": 5.503618174639874e-05, "loss": 0.0377, "step": 49450 }, { "grad_norm": 0.3471211791038513, "learning_rate": 5.501973086129325e-05, "loss": 0.0341, "step": 49460 }, { "grad_norm": 0.25696444511413574, "learning_rate": 5.500327942723753e-05, "loss": 0.0382, "step": 49470 }, { "grad_norm": 0.3685357868671417, "learning_rate": 5.498682744603071e-05, "loss": 0.0316, "step": 49480 }, { "grad_norm": 0.4563405513763428, "learning_rate": 5.497037491947194e-05, "loss": 0.033, "step": 49490 }, { "grad_norm": 0.34458255767822266, "learning_rate": 5.4953921849360424e-05, "loss": 0.0299, "step": 49500 }, { "grad_norm": 0.2796761393547058, "learning_rate": 5.493746823749547e-05, "loss": 0.0311, "step": 49510 }, { "grad_norm": 0.2700694501399994, "learning_rate": 5.49210140856764e-05, "loss": 0.0416, "step": 49520 }, { "grad_norm": 0.303688645362854, "learning_rate": 5.4904559395702647e-05, "loss": 0.0321, "step": 49530 }, { "grad_norm": 0.3578157126903534, "learning_rate": 5.488810416937364e-05, "loss": 0.0362, "step": 49540 }, { "grad_norm": 0.48919156193733215, "learning_rate": 5.4871648408488926e-05, "loss": 0.0349, "step": 49550 }, { "grad_norm": 0.3734986186027527, "learning_rate": 5.485519211484807e-05, "loss": 0.037, "step": 49560 }, { "grad_norm": 0.3512158989906311, "learning_rate": 5.4838735290250676e-05, "loss": 0.0428, "step": 49570 }, { "grad_norm": 0.42641761898994446, "learning_rate": 5.482227793649649e-05, "loss": 0.035, "step": 49580 }, { "grad_norm": 0.5759860277175903, "learning_rate": 5.480582005538524e-05, "loss": 0.0399, "step": 49590 }, { "grad_norm": 0.4344235956668854, "learning_rate": 5.478936164871671e-05, "loss": 0.0403, "step": 49600 }, { "grad_norm": 0.35947585105895996, "learning_rate": 5.477290271829081e-05, "loss": 0.0388, "step": 49610 }, { "grad_norm": 0.35807010531425476, "learning_rate": 5.475644326590744e-05, "loss": 0.0325, "step": 49620 }, { "grad_norm": 0.3139816224575043, "learning_rate": 5.473998329336658e-05, "loss": 0.0365, "step": 49630 }, { "grad_norm": 0.40826207399368286, "learning_rate": 5.4723522802468286e-05, "loss": 0.0349, "step": 49640 }, { "grad_norm": 0.49893057346343994, "learning_rate": 5.470706179501264e-05, "loss": 0.0351, "step": 49650 }, { "grad_norm": 0.45516151189804077, "learning_rate": 5.4690600272799776e-05, "loss": 0.0377, "step": 49660 }, { "grad_norm": 0.3934611678123474, "learning_rate": 5.467413823762993e-05, "loss": 0.0324, "step": 49670 }, { "grad_norm": 0.3950808048248291, "learning_rate": 5.465767569130335e-05, "loss": 0.0301, "step": 49680 }, { "grad_norm": 0.4221554398536682, "learning_rate": 5.464121263562036e-05, "loss": 0.0351, "step": 49690 }, { "grad_norm": 0.40978744626045227, "learning_rate": 5.4624749072381343e-05, "loss": 0.038, "step": 49700 }, { "grad_norm": 0.4050760269165039, "learning_rate": 5.460828500338672e-05, "loss": 0.0331, "step": 49710 }, { "grad_norm": 0.31773942708969116, "learning_rate": 5.459182043043698e-05, "loss": 0.0327, "step": 49720 }, { "grad_norm": 0.3284837603569031, "learning_rate": 5.457535535533265e-05, "loss": 0.0335, "step": 49730 }, { "grad_norm": 0.43046557903289795, "learning_rate": 5.4558889779874334e-05, "loss": 0.033, "step": 49740 }, { "grad_norm": 0.562469482421875, "learning_rate": 5.454242370586269e-05, "loss": 0.0383, "step": 49750 }, { "grad_norm": 0.27575841546058655, "learning_rate": 5.452595713509843e-05, "loss": 0.0301, "step": 49760 }, { "grad_norm": 0.3477112352848053, "learning_rate": 5.4509490069382274e-05, "loss": 0.0317, "step": 49770 }, { "grad_norm": 0.33810579776763916, "learning_rate": 5.449302251051509e-05, "loss": 0.0298, "step": 49780 }, { "grad_norm": 0.36876386404037476, "learning_rate": 5.44765544602977e-05, "loss": 0.0363, "step": 49790 }, { "grad_norm": 0.3091857433319092, "learning_rate": 5.446008592053107e-05, "loss": 0.035, "step": 49800 }, { "grad_norm": 0.28996458649635315, "learning_rate": 5.4443616893016135e-05, "loss": 0.036, "step": 49810 }, { "grad_norm": 0.2540101706981659, "learning_rate": 5.4427147379553934e-05, "loss": 0.0298, "step": 49820 }, { "grad_norm": 0.473198264837265, "learning_rate": 5.441067738194556e-05, "loss": 0.034, "step": 49830 }, { "grad_norm": 0.31202301383018494, "learning_rate": 5.439420690199214e-05, "loss": 0.0379, "step": 49840 }, { "grad_norm": 0.8623305559158325, "learning_rate": 5.437773594149484e-05, "loss": 0.0343, "step": 49850 }, { "grad_norm": 0.40612074732780457, "learning_rate": 5.436126450225495e-05, "loss": 0.038, "step": 49860 }, { "grad_norm": 0.797870934009552, "learning_rate": 5.434479258607371e-05, "loss": 0.0329, "step": 49870 }, { "grad_norm": 0.38882097601890564, "learning_rate": 5.43283201947525e-05, "loss": 0.0362, "step": 49880 }, { "grad_norm": 0.34818342328071594, "learning_rate": 5.431184733009268e-05, "loss": 0.0411, "step": 49890 }, { "grad_norm": 0.2787010967731476, "learning_rate": 5.4295373993895736e-05, "loss": 0.0358, "step": 49900 }, { "grad_norm": 0.34848687052726746, "learning_rate": 5.4278900187963157e-05, "loss": 0.0372, "step": 49910 }, { "grad_norm": 0.37188512086868286, "learning_rate": 5.426242591409646e-05, "loss": 0.0345, "step": 49920 }, { "grad_norm": 0.4333057105541229, "learning_rate": 5.42459511740973e-05, "loss": 0.0369, "step": 49930 }, { "grad_norm": 0.42875951528549194, "learning_rate": 5.422947596976729e-05, "loss": 0.0353, "step": 49940 }, { "grad_norm": 0.5237317681312561, "learning_rate": 5.4213000302908134e-05, "loss": 0.0354, "step": 49950 }, { "grad_norm": 0.479462593793869, "learning_rate": 5.419652417532162e-05, "loss": 0.0414, "step": 49960 }, { "grad_norm": 0.3843194842338562, "learning_rate": 5.4180047588809534e-05, "loss": 0.035, "step": 49970 }, { "grad_norm": 0.413459450006485, "learning_rate": 5.4163570545173704e-05, "loss": 0.0352, "step": 49980 }, { "grad_norm": 0.46961405873298645, "learning_rate": 5.414709304621608e-05, "loss": 0.0337, "step": 49990 }, { "grad_norm": 0.433423787355423, "learning_rate": 5.413061509373858e-05, "loss": 0.039, "step": 50000 }, { "grad_norm": 0.31446465849876404, "learning_rate": 5.411413668954324e-05, "loss": 0.0332, "step": 50010 }, { "grad_norm": 0.3275545537471771, "learning_rate": 5.4097657835432083e-05, "loss": 0.034, "step": 50020 }, { "grad_norm": 0.23945683240890503, "learning_rate": 5.408117853320723e-05, "loss": 0.032, "step": 50030 }, { "grad_norm": 0.36487558484077454, "learning_rate": 5.406469878467084e-05, "loss": 0.0354, "step": 50040 }, { "grad_norm": 0.37832459807395935, "learning_rate": 5.404821859162509e-05, "loss": 0.0326, "step": 50050 }, { "grad_norm": 0.2944389879703522, "learning_rate": 5.403173795587225e-05, "loss": 0.031, "step": 50060 }, { "grad_norm": 0.42534777522087097, "learning_rate": 5.4015256879214606e-05, "loss": 0.0294, "step": 50070 }, { "grad_norm": 0.5097178816795349, "learning_rate": 5.39987753634545e-05, "loss": 0.0355, "step": 50080 }, { "grad_norm": 0.48576903343200684, "learning_rate": 5.3982293410394335e-05, "loss": 0.0394, "step": 50090 }, { "grad_norm": 0.40225934982299805, "learning_rate": 5.396581102183655e-05, "loss": 0.0379, "step": 50100 }, { "grad_norm": 0.32356011867523193, "learning_rate": 5.3949328199583634e-05, "loss": 0.0316, "step": 50110 }, { "grad_norm": 0.3192272186279297, "learning_rate": 5.393284494543813e-05, "loss": 0.0366, "step": 50120 }, { "grad_norm": 0.44097939133644104, "learning_rate": 5.391636126120262e-05, "loss": 0.0297, "step": 50130 }, { "grad_norm": 0.3682015538215637, "learning_rate": 5.389987714867971e-05, "loss": 0.0394, "step": 50140 }, { "grad_norm": 0.3930530548095703, "learning_rate": 5.3883392609672114e-05, "loss": 0.0344, "step": 50150 }, { "grad_norm": 0.5471790432929993, "learning_rate": 5.3866907645982545e-05, "loss": 0.0331, "step": 50160 }, { "grad_norm": 0.3425286114215851, "learning_rate": 5.385042225941375e-05, "loss": 0.0339, "step": 50170 }, { "grad_norm": 0.37882307171821594, "learning_rate": 5.3833936451768585e-05, "loss": 0.0328, "step": 50180 }, { "grad_norm": 0.5404528379440308, "learning_rate": 5.3817450224849884e-05, "loss": 0.036, "step": 50190 }, { "grad_norm": 0.31428974866867065, "learning_rate": 5.380096358046056e-05, "loss": 0.0366, "step": 50200 }, { "grad_norm": 0.33357685804367065, "learning_rate": 5.378447652040359e-05, "loss": 0.0325, "step": 50210 }, { "grad_norm": 0.47122374176979065, "learning_rate": 5.376798904648195e-05, "loss": 0.0311, "step": 50220 }, { "grad_norm": 0.2772379517555237, "learning_rate": 5.375150116049869e-05, "loss": 0.0389, "step": 50230 }, { "grad_norm": 0.4306221306324005, "learning_rate": 5.373501286425691e-05, "loss": 0.0328, "step": 50240 }, { "grad_norm": 0.3226436674594879, "learning_rate": 5.3718524159559726e-05, "loss": 0.0323, "step": 50250 }, { "grad_norm": 0.4496372640132904, "learning_rate": 5.370203504821034e-05, "loss": 0.0344, "step": 50260 }, { "grad_norm": 0.32512909173965454, "learning_rate": 5.368554553201196e-05, "loss": 0.0351, "step": 50270 }, { "grad_norm": 0.42488545179367065, "learning_rate": 5.3669055612767874e-05, "loss": 0.0366, "step": 50280 }, { "grad_norm": 0.4170638620853424, "learning_rate": 5.3652565292281376e-05, "loss": 0.0417, "step": 50290 }, { "grad_norm": 0.4431516230106354, "learning_rate": 5.3636074572355824e-05, "loss": 0.0331, "step": 50300 }, { "grad_norm": 0.3696151077747345, "learning_rate": 5.361958345479463e-05, "loss": 0.0383, "step": 50310 }, { "grad_norm": 0.31573909521102905, "learning_rate": 5.360309194140123e-05, "loss": 0.0362, "step": 50320 }, { "grad_norm": 0.841157853603363, "learning_rate": 5.3586600033979106e-05, "loss": 0.0364, "step": 50330 }, { "grad_norm": 0.4435441493988037, "learning_rate": 5.357010773433181e-05, "loss": 0.0362, "step": 50340 }, { "grad_norm": 0.3385665714740753, "learning_rate": 5.3553615044262895e-05, "loss": 0.0336, "step": 50350 }, { "grad_norm": 0.6689051985740662, "learning_rate": 5.3537121965575964e-05, "loss": 0.0314, "step": 50360 }, { "grad_norm": 0.4437883794307709, "learning_rate": 5.352062850007471e-05, "loss": 0.0302, "step": 50370 }, { "grad_norm": 0.43432578444480896, "learning_rate": 5.350413464956282e-05, "loss": 0.0335, "step": 50380 }, { "grad_norm": 0.3357693552970886, "learning_rate": 5.348764041584403e-05, "loss": 0.0367, "step": 50390 }, { "grad_norm": 0.33145585656166077, "learning_rate": 5.347114580072212e-05, "loss": 0.0315, "step": 50400 }, { "grad_norm": 0.43526995182037354, "learning_rate": 5.345465080600093e-05, "loss": 0.0269, "step": 50410 }, { "grad_norm": 0.2971447706222534, "learning_rate": 5.3438155433484326e-05, "loss": 0.0319, "step": 50420 }, { "grad_norm": 0.5934613943099976, "learning_rate": 5.3421659684976197e-05, "loss": 0.0359, "step": 50430 }, { "grad_norm": 0.3840347230434418, "learning_rate": 5.340516356228052e-05, "loss": 0.0333, "step": 50440 }, { "grad_norm": 0.31598931550979614, "learning_rate": 5.338866706720128e-05, "loss": 0.0389, "step": 50450 }, { "grad_norm": 0.3743109405040741, "learning_rate": 5.337217020154249e-05, "loss": 0.0342, "step": 50460 }, { "grad_norm": 0.3107164800167084, "learning_rate": 5.335567296710825e-05, "loss": 0.0345, "step": 50470 }, { "grad_norm": 0.4832626283168793, "learning_rate": 5.333917536570265e-05, "loss": 0.0341, "step": 50480 }, { "grad_norm": 0.40168657898902893, "learning_rate": 5.332267739912986e-05, "loss": 0.0319, "step": 50490 }, { "grad_norm": 0.3226161003112793, "learning_rate": 5.330617906919405e-05, "loss": 0.0359, "step": 50500 }, { "grad_norm": 0.6650134921073914, "learning_rate": 5.328968037769949e-05, "loss": 0.0336, "step": 50510 }, { "grad_norm": 0.36215436458587646, "learning_rate": 5.327318132645042e-05, "loss": 0.0333, "step": 50520 }, { "grad_norm": 0.44525080919265747, "learning_rate": 5.325668191725116e-05, "loss": 0.0414, "step": 50530 }, { "grad_norm": 0.4561859369277954, "learning_rate": 5.324018215190606e-05, "loss": 0.0369, "step": 50540 }, { "grad_norm": 0.4241926670074463, "learning_rate": 5.3223682032219515e-05, "loss": 0.034, "step": 50550 }, { "grad_norm": 0.2819614112377167, "learning_rate": 5.320718155999595e-05, "loss": 0.0326, "step": 50560 }, { "grad_norm": 0.4891466200351715, "learning_rate": 5.3190680737039835e-05, "loss": 0.04, "step": 50570 }, { "grad_norm": 0.4988720417022705, "learning_rate": 5.317417956515567e-05, "loss": 0.0391, "step": 50580 }, { "grad_norm": 0.38653504848480225, "learning_rate": 5.315767804614798e-05, "loss": 0.0413, "step": 50590 }, { "grad_norm": 0.6161841154098511, "learning_rate": 5.3141176181821395e-05, "loss": 0.0339, "step": 50600 }, { "grad_norm": 0.44263896346092224, "learning_rate": 5.31246739739805e-05, "loss": 0.0385, "step": 50610 }, { "grad_norm": 0.46252501010894775, "learning_rate": 5.310817142442994e-05, "loss": 0.0374, "step": 50620 }, { "grad_norm": 0.42460817098617554, "learning_rate": 5.309166853497445e-05, "loss": 0.0386, "step": 50630 }, { "grad_norm": 0.3919922709465027, "learning_rate": 5.307516530741873e-05, "loss": 0.039, "step": 50640 }, { "grad_norm": 0.3649897575378418, "learning_rate": 5.305866174356754e-05, "loss": 0.0297, "step": 50650 }, { "grad_norm": 0.31975317001342773, "learning_rate": 5.304215784522571e-05, "loss": 0.0353, "step": 50660 }, { "grad_norm": 0.32443663477897644, "learning_rate": 5.302565361419808e-05, "loss": 0.0322, "step": 50670 }, { "grad_norm": 0.23891790211200714, "learning_rate": 5.3009149052289507e-05, "loss": 0.0319, "step": 50680 }, { "grad_norm": 0.34095990657806396, "learning_rate": 5.299264416130493e-05, "loss": 0.0297, "step": 50690 }, { "grad_norm": 1.2300013303756714, "learning_rate": 5.297613894304928e-05, "loss": 0.0306, "step": 50700 }, { "grad_norm": 0.7467108964920044, "learning_rate": 5.2959633399327534e-05, "loss": 0.0321, "step": 50710 }, { "grad_norm": 0.3850932717323303, "learning_rate": 5.294312753194476e-05, "loss": 0.0387, "step": 50720 }, { "grad_norm": 0.40812617540359497, "learning_rate": 5.292662134270596e-05, "loss": 0.0365, "step": 50730 }, { "grad_norm": 0.3180174231529236, "learning_rate": 5.291011483341626e-05, "loss": 0.0441, "step": 50740 }, { "grad_norm": 0.3360181450843811, "learning_rate": 5.2893608005880767e-05, "loss": 0.0408, "step": 50750 }, { "grad_norm": 0.48193326592445374, "learning_rate": 5.287710086190467e-05, "loss": 0.0361, "step": 50760 }, { "grad_norm": 0.4676819443702698, "learning_rate": 5.286059340329314e-05, "loss": 0.0397, "step": 50770 }, { "grad_norm": 0.2872442603111267, "learning_rate": 5.284408563185141e-05, "loss": 0.0361, "step": 50780 }, { "grad_norm": 0.3742412328720093, "learning_rate": 5.282757754938476e-05, "loss": 0.0333, "step": 50790 }, { "grad_norm": 0.4347550570964813, "learning_rate": 5.281106915769849e-05, "loss": 0.0339, "step": 50800 }, { "grad_norm": 0.3864270746707916, "learning_rate": 5.2794560458597897e-05, "loss": 0.0411, "step": 50810 }, { "grad_norm": 0.2779846787452698, "learning_rate": 5.27780514538884e-05, "loss": 0.038, "step": 50820 }, { "grad_norm": 0.37689557671546936, "learning_rate": 5.2761542145375365e-05, "loss": 0.039, "step": 50830 }, { "grad_norm": 0.4109010398387909, "learning_rate": 5.274503253486421e-05, "loss": 0.0359, "step": 50840 }, { "grad_norm": 0.5614544749259949, "learning_rate": 5.272852262416046e-05, "loss": 0.0356, "step": 50850 }, { "grad_norm": 0.36840885877609253, "learning_rate": 5.2712012415069555e-05, "loss": 0.0369, "step": 50860 }, { "grad_norm": 0.3528934121131897, "learning_rate": 5.269550190939705e-05, "loss": 0.0386, "step": 50870 }, { "grad_norm": 0.4161921441555023, "learning_rate": 5.267899110894852e-05, "loss": 0.0351, "step": 50880 }, { "grad_norm": 0.32757049798965454, "learning_rate": 5.266248001552955e-05, "loss": 0.0374, "step": 50890 }, { "grad_norm": 0.805081844329834, "learning_rate": 5.264596863094575e-05, "loss": 0.0317, "step": 50900 }, { "grad_norm": 0.5077993273735046, "learning_rate": 5.262945695700282e-05, "loss": 0.0415, "step": 50910 }, { "grad_norm": 0.5358324646949768, "learning_rate": 5.261294499550643e-05, "loss": 0.0366, "step": 50920 }, { "grad_norm": 0.4273630976676941, "learning_rate": 5.25964327482623e-05, "loss": 0.0384, "step": 50930 }, { "grad_norm": 0.4216172993183136, "learning_rate": 5.257992021707617e-05, "loss": 0.0368, "step": 50940 }, { "grad_norm": 0.4152250289916992, "learning_rate": 5.256340740375387e-05, "loss": 0.0364, "step": 50950 }, { "grad_norm": 0.32418641448020935, "learning_rate": 5.254689431010117e-05, "loss": 0.0373, "step": 50960 }, { "grad_norm": 0.38099202513694763, "learning_rate": 5.253038093792395e-05, "loss": 0.0417, "step": 50970 }, { "grad_norm": 0.4185802936553955, "learning_rate": 5.251386728902806e-05, "loss": 0.0363, "step": 50980 }, { "grad_norm": 0.28107950091362, "learning_rate": 5.2497353365219446e-05, "loss": 0.0336, "step": 50990 }, { "grad_norm": 0.4396098852157593, "learning_rate": 5.2480839168304e-05, "loss": 0.0333, "step": 51000 }, { "grad_norm": 0.43345025181770325, "learning_rate": 5.246432470008773e-05, "loss": 0.0318, "step": 51010 }, { "grad_norm": 0.3691292703151703, "learning_rate": 5.2447809962376616e-05, "loss": 0.0354, "step": 51020 }, { "grad_norm": 0.5044711828231812, "learning_rate": 5.243129495697666e-05, "loss": 0.0338, "step": 51030 }, { "grad_norm": 0.40478214621543884, "learning_rate": 5.241477968569397e-05, "loss": 0.0388, "step": 51040 }, { "grad_norm": 0.36786195635795593, "learning_rate": 5.23982641503346e-05, "loss": 0.0352, "step": 51050 }, { "grad_norm": 0.4215939939022064, "learning_rate": 5.238174835270465e-05, "loss": 0.033, "step": 51060 }, { "grad_norm": 0.5514323115348816, "learning_rate": 5.23652322946103e-05, "loss": 0.0358, "step": 51070 }, { "grad_norm": 0.2781006097793579, "learning_rate": 5.234871597785771e-05, "loss": 0.0404, "step": 51080 }, { "grad_norm": 0.4681585729122162, "learning_rate": 5.233219940425306e-05, "loss": 0.0314, "step": 51090 }, { "grad_norm": 0.49541959166526794, "learning_rate": 5.2315682575602576e-05, "loss": 0.0311, "step": 51100 }, { "grad_norm": 0.3157307505607605, "learning_rate": 5.229916549371254e-05, "loss": 0.0347, "step": 51110 }, { "grad_norm": 0.29732546210289, "learning_rate": 5.228264816038924e-05, "loss": 0.0307, "step": 51120 }, { "grad_norm": 0.36750826239585876, "learning_rate": 5.2266130577438935e-05, "loss": 0.0387, "step": 51130 }, { "grad_norm": 0.40605977177619934, "learning_rate": 5.224961274666802e-05, "loss": 0.0312, "step": 51140 }, { "grad_norm": 0.3068397343158722, "learning_rate": 5.223309466988283e-05, "loss": 0.0316, "step": 51150 }, { "grad_norm": 0.3032141327857971, "learning_rate": 5.221657634888976e-05, "loss": 0.0376, "step": 51160 }, { "grad_norm": 0.3610273599624634, "learning_rate": 5.2200057785495235e-05, "loss": 0.0343, "step": 51170 }, { "grad_norm": 0.22293983399868011, "learning_rate": 5.2183538981505706e-05, "loss": 0.0375, "step": 51180 }, { "grad_norm": 0.45059406757354736, "learning_rate": 5.216701993872762e-05, "loss": 0.0405, "step": 51190 }, { "grad_norm": 0.30296823382377625, "learning_rate": 5.215050065896752e-05, "loss": 0.0348, "step": 51200 }, { "grad_norm": 0.5157042741775513, "learning_rate": 5.2133981144031875e-05, "loss": 0.032, "step": 51210 }, { "grad_norm": 0.4022068381309509, "learning_rate": 5.211746139572727e-05, "loss": 0.0332, "step": 51220 }, { "grad_norm": 0.3595377802848816, "learning_rate": 5.210094141586026e-05, "loss": 0.0328, "step": 51230 }, { "grad_norm": 0.43298664689064026, "learning_rate": 5.2084421206237476e-05, "loss": 0.0326, "step": 51240 }, { "grad_norm": 0.5733505487442017, "learning_rate": 5.20679007686655e-05, "loss": 0.0346, "step": 51250 }, { "grad_norm": 0.2267715185880661, "learning_rate": 5.205138010495102e-05, "loss": 0.034, "step": 51260 }, { "grad_norm": 0.39533233642578125, "learning_rate": 5.20348592169007e-05, "loss": 0.0352, "step": 51270 }, { "grad_norm": 0.4101969301700592, "learning_rate": 5.201833810632123e-05, "loss": 0.0368, "step": 51280 }, { "grad_norm": 0.48632198572158813, "learning_rate": 5.200181677501933e-05, "loss": 0.0354, "step": 51290 }, { "grad_norm": 0.4229980707168579, "learning_rate": 5.1985295224801777e-05, "loss": 0.0347, "step": 51300 }, { "grad_norm": 0.39228200912475586, "learning_rate": 5.196877345747531e-05, "loss": 0.0345, "step": 51310 }, { "grad_norm": 0.4549016058444977, "learning_rate": 5.195225147484674e-05, "loss": 0.0358, "step": 51320 }, { "grad_norm": 0.6074178218841553, "learning_rate": 5.193572927872288e-05, "loss": 0.0402, "step": 51330 }, { "grad_norm": 0.30678319931030273, "learning_rate": 5.191920687091059e-05, "loss": 0.0377, "step": 51340 }, { "grad_norm": 0.28682178258895874, "learning_rate": 5.1902684253216695e-05, "loss": 0.0402, "step": 51350 }, { "grad_norm": 0.30752596259117126, "learning_rate": 5.1886161427448134e-05, "loss": 0.0305, "step": 51360 }, { "grad_norm": 0.4200826585292816, "learning_rate": 5.186963839541179e-05, "loss": 0.0282, "step": 51370 }, { "grad_norm": 0.4240366816520691, "learning_rate": 5.185311515891459e-05, "loss": 0.0323, "step": 51380 }, { "grad_norm": 0.3218856751918793, "learning_rate": 5.1836591719763516e-05, "loss": 0.0319, "step": 51390 }, { "grad_norm": 0.3522564172744751, "learning_rate": 5.182006807976554e-05, "loss": 0.033, "step": 51400 }, { "grad_norm": 0.3840024769306183, "learning_rate": 5.180354424072763e-05, "loss": 0.0364, "step": 51410 }, { "grad_norm": 0.457909494638443, "learning_rate": 5.178702020445686e-05, "loss": 0.0383, "step": 51420 }, { "grad_norm": 0.35377341508865356, "learning_rate": 5.177049597276024e-05, "loss": 0.0299, "step": 51430 }, { "grad_norm": 0.6822288036346436, "learning_rate": 5.175397154744482e-05, "loss": 0.035, "step": 51440 }, { "grad_norm": 0.4300072193145752, "learning_rate": 5.1737446930317736e-05, "loss": 0.0364, "step": 51450 }, { "grad_norm": 0.32224416732788086, "learning_rate": 5.172092212318604e-05, "loss": 0.0282, "step": 51460 }, { "grad_norm": 0.36124923825263977, "learning_rate": 5.170439712785692e-05, "loss": 0.0399, "step": 51470 }, { "grad_norm": 0.5429509878158569, "learning_rate": 5.1687871946137467e-05, "loss": 0.039, "step": 51480 }, { "grad_norm": 0.5714544057846069, "learning_rate": 5.16713465798349e-05, "loss": 0.0299, "step": 51490 }, { "grad_norm": 0.3782091438770294, "learning_rate": 5.165482103075637e-05, "loss": 0.0332, "step": 51500 }, { "grad_norm": 0.3940088152885437, "learning_rate": 5.163829530070909e-05, "loss": 0.0367, "step": 51510 }, { "grad_norm": 0.4325263798236847, "learning_rate": 5.162176939150032e-05, "loss": 0.0355, "step": 51520 }, { "grad_norm": 0.3222827911376953, "learning_rate": 5.160524330493728e-05, "loss": 0.0348, "step": 51530 }, { "grad_norm": 0.4180385172367096, "learning_rate": 5.158871704282724e-05, "loss": 0.0316, "step": 51540 }, { "grad_norm": 0.27041399478912354, "learning_rate": 5.15721906069775e-05, "loss": 0.0359, "step": 51550 }, { "grad_norm": 0.2954934239387512, "learning_rate": 5.155566399919536e-05, "loss": 0.0331, "step": 51560 }, { "grad_norm": 0.3587397336959839, "learning_rate": 5.153913722128813e-05, "loss": 0.0379, "step": 51570 }, { "grad_norm": 0.3754619359970093, "learning_rate": 5.1522610275063176e-05, "loss": 0.032, "step": 51580 }, { "grad_norm": 0.272939532995224, "learning_rate": 5.150608316232786e-05, "loss": 0.0318, "step": 51590 }, { "grad_norm": 0.4491005837917328, "learning_rate": 5.148955588488955e-05, "loss": 0.0288, "step": 51600 }, { "grad_norm": 0.42902064323425293, "learning_rate": 5.147302844455564e-05, "loss": 0.0366, "step": 51610 }, { "grad_norm": 0.5733738541603088, "learning_rate": 5.145650084313356e-05, "loss": 0.0314, "step": 51620 }, { "grad_norm": 0.3690718412399292, "learning_rate": 5.1439973082430734e-05, "loss": 0.034, "step": 51630 }, { "grad_norm": 0.4825464189052582, "learning_rate": 5.142344516425461e-05, "loss": 0.0349, "step": 51640 }, { "grad_norm": 0.3094305098056793, "learning_rate": 5.1406917090412673e-05, "loss": 0.0375, "step": 51650 }, { "grad_norm": 0.4276922643184662, "learning_rate": 5.1390388862712404e-05, "loss": 0.0292, "step": 51660 }, { "grad_norm": 0.4104102551937103, "learning_rate": 5.137386048296129e-05, "loss": 0.0371, "step": 51670 }, { "grad_norm": 0.2426466941833496, "learning_rate": 5.135733195296686e-05, "loss": 0.0292, "step": 51680 }, { "grad_norm": 0.34331607818603516, "learning_rate": 5.134080327453664e-05, "loss": 0.0372, "step": 51690 }, { "grad_norm": 0.329317182302475, "learning_rate": 5.1324274449478204e-05, "loss": 0.0357, "step": 51700 }, { "grad_norm": 0.5836275815963745, "learning_rate": 5.130774547959909e-05, "loss": 0.042, "step": 51710 }, { "grad_norm": 0.6366263628005981, "learning_rate": 5.129121636670692e-05, "loss": 0.0369, "step": 51720 }, { "grad_norm": 0.3740876615047455, "learning_rate": 5.1274687112609246e-05, "loss": 0.0305, "step": 51730 }, { "grad_norm": 0.26246708631515503, "learning_rate": 5.125815771911372e-05, "loss": 0.0294, "step": 51740 }, { "grad_norm": 2.076155662536621, "learning_rate": 5.1241628188027956e-05, "loss": 0.037, "step": 51750 }, { "grad_norm": 0.3249514102935791, "learning_rate": 5.1225098521159585e-05, "loss": 0.039, "step": 51760 }, { "grad_norm": 0.24094952642917633, "learning_rate": 5.12085687203163e-05, "loss": 0.0323, "step": 51770 }, { "grad_norm": 0.402392715215683, "learning_rate": 5.119203878730575e-05, "loss": 0.0333, "step": 51780 }, { "grad_norm": 0.3655901551246643, "learning_rate": 5.117550872393564e-05, "loss": 0.0343, "step": 51790 }, { "grad_norm": 0.44805026054382324, "learning_rate": 5.1158978532013645e-05, "loss": 0.0354, "step": 51800 }, { "grad_norm": 0.2889823913574219, "learning_rate": 5.114244821334751e-05, "loss": 0.0319, "step": 51810 }, { "grad_norm": 0.284750759601593, "learning_rate": 5.112591776974497e-05, "loss": 0.0405, "step": 51820 }, { "grad_norm": 0.31706541776657104, "learning_rate": 5.110938720301374e-05, "loss": 0.0376, "step": 51830 }, { "grad_norm": 0.34662407636642456, "learning_rate": 5.1092856514961595e-05, "loss": 0.0372, "step": 51840 }, { "grad_norm": 0.38661298155784607, "learning_rate": 5.107632570739631e-05, "loss": 0.0335, "step": 51850 }, { "grad_norm": 0.3626723885536194, "learning_rate": 5.1059794782125645e-05, "loss": 0.0326, "step": 51860 }, { "grad_norm": 0.45809656381607056, "learning_rate": 5.104326374095744e-05, "loss": 0.0326, "step": 51870 }, { "grad_norm": 0.28521567583084106, "learning_rate": 5.102673258569947e-05, "loss": 0.0323, "step": 51880 }, { "grad_norm": 0.340055376291275, "learning_rate": 5.101020131815956e-05, "loss": 0.0349, "step": 51890 }, { "grad_norm": 0.2878771126270294, "learning_rate": 5.099366994014556e-05, "loss": 0.038, "step": 51900 }, { "grad_norm": 0.3657682240009308, "learning_rate": 5.097713845346532e-05, "loss": 0.0256, "step": 51910 }, { "grad_norm": 0.28629007935523987, "learning_rate": 5.096060685992668e-05, "loss": 0.0385, "step": 51920 }, { "grad_norm": 0.2775910496711731, "learning_rate": 5.0944075161337514e-05, "loss": 0.0353, "step": 51930 }, { "grad_norm": 0.593878984451294, "learning_rate": 5.09275433595057e-05, "loss": 0.0351, "step": 51940 }, { "grad_norm": 0.3992031216621399, "learning_rate": 5.0911011456239157e-05, "loss": 0.0313, "step": 51950 }, { "grad_norm": 0.49269402027130127, "learning_rate": 5.089447945334574e-05, "loss": 0.0366, "step": 51960 }, { "grad_norm": 0.36610421538352966, "learning_rate": 5.087794735263343e-05, "loss": 0.0313, "step": 51970 }, { "grad_norm": 0.3568008244037628, "learning_rate": 5.086141515591011e-05, "loss": 0.0315, "step": 51980 }, { "grad_norm": 0.46196314692497253, "learning_rate": 5.084488286498371e-05, "loss": 0.0376, "step": 51990 }, { "grad_norm": 0.3361760377883911, "learning_rate": 5.08283504816622e-05, "loss": 0.0365, "step": 52000 }, { "grad_norm": 0.38349464535713196, "learning_rate": 5.081181800775353e-05, "loss": 0.0364, "step": 52010 }, { "grad_norm": 0.31485339999198914, "learning_rate": 5.079528544506565e-05, "loss": 0.0422, "step": 52020 }, { "grad_norm": 0.29891300201416016, "learning_rate": 5.077875279540657e-05, "loss": 0.0298, "step": 52030 }, { "grad_norm": 0.5750253796577454, "learning_rate": 5.076222006058424e-05, "loss": 0.0297, "step": 52040 }, { "grad_norm": 0.3607442378997803, "learning_rate": 5.074568724240667e-05, "loss": 0.0295, "step": 52050 }, { "grad_norm": 0.40815937519073486, "learning_rate": 5.072915434268187e-05, "loss": 0.0392, "step": 52060 }, { "grad_norm": 0.41805222630500793, "learning_rate": 5.071262136321785e-05, "loss": 0.0393, "step": 52070 }, { "grad_norm": 0.7973344922065735, "learning_rate": 5.069608830582262e-05, "loss": 0.0328, "step": 52080 }, { "grad_norm": 0.29257452487945557, "learning_rate": 5.067955517230423e-05, "loss": 0.0329, "step": 52090 }, { "grad_norm": 0.29899904131889343, "learning_rate": 5.06630219644707e-05, "loss": 0.03, "step": 52100 }, { "grad_norm": 0.2947874069213867, "learning_rate": 5.064648868413008e-05, "loss": 0.0267, "step": 52110 }, { "grad_norm": 0.28069236874580383, "learning_rate": 5.062995533309044e-05, "loss": 0.0377, "step": 52120 }, { "grad_norm": 0.31346791982650757, "learning_rate": 5.061342191315983e-05, "loss": 0.0381, "step": 52130 }, { "grad_norm": 0.339366614818573, "learning_rate": 5.0596888426146325e-05, "loss": 0.0361, "step": 52140 }, { "grad_norm": 0.3707111179828644, "learning_rate": 5.058035487385798e-05, "loss": 0.0335, "step": 52150 }, { "grad_norm": 0.5384906530380249, "learning_rate": 5.0563821258102915e-05, "loss": 0.0318, "step": 52160 }, { "grad_norm": 0.26399871706962585, "learning_rate": 5.05472875806892e-05, "loss": 0.0326, "step": 52170 }, { "grad_norm": 0.4517676532268524, "learning_rate": 5.053075384342494e-05, "loss": 0.0381, "step": 52180 }, { "grad_norm": 1.1123476028442383, "learning_rate": 5.0514220048118234e-05, "loss": 0.038, "step": 52190 }, { "grad_norm": 0.4375177025794983, "learning_rate": 5.049768619657721e-05, "loss": 0.0343, "step": 52200 }, { "grad_norm": 0.37047049403190613, "learning_rate": 5.048115229060996e-05, "loss": 0.0357, "step": 52210 }, { "grad_norm": 0.47586873173713684, "learning_rate": 5.046461833202462e-05, "loss": 0.0299, "step": 52220 }, { "grad_norm": 1.552503228187561, "learning_rate": 5.044808432262932e-05, "loss": 0.0279, "step": 52230 }, { "grad_norm": 0.27779433131217957, "learning_rate": 5.043155026423219e-05, "loss": 0.0308, "step": 52240 }, { "grad_norm": 1.4459911584854126, "learning_rate": 5.041501615864137e-05, "loss": 0.0346, "step": 52250 }, { "grad_norm": 0.23400986194610596, "learning_rate": 5.039848200766502e-05, "loss": 0.0322, "step": 52260 }, { "grad_norm": 0.38763758540153503, "learning_rate": 5.038194781311125e-05, "loss": 0.0418, "step": 52270 }, { "grad_norm": 0.265129953622818, "learning_rate": 5.0365413576788264e-05, "loss": 0.0267, "step": 52280 }, { "grad_norm": 0.6697215437889099, "learning_rate": 5.0348879300504194e-05, "loss": 0.035, "step": 52290 }, { "grad_norm": 0.24574759602546692, "learning_rate": 5.033234498606721e-05, "loss": 0.0316, "step": 52300 }, { "grad_norm": 0.5404148101806641, "learning_rate": 5.0315810635285476e-05, "loss": 0.0337, "step": 52310 }, { "grad_norm": 0.4763547480106354, "learning_rate": 5.0299276249967164e-05, "loss": 0.0299, "step": 52320 }, { "grad_norm": 0.5404433012008667, "learning_rate": 5.0282741831920454e-05, "loss": 0.0414, "step": 52330 }, { "grad_norm": 0.4941578805446625, "learning_rate": 5.026620738295351e-05, "loss": 0.0364, "step": 52340 }, { "grad_norm": 0.5634004473686218, "learning_rate": 5.024967290487455e-05, "loss": 0.0353, "step": 52350 }, { "grad_norm": 0.3869166374206543, "learning_rate": 5.023313839949172e-05, "loss": 0.0323, "step": 52360 }, { "grad_norm": 0.2963005304336548, "learning_rate": 5.0216603868613224e-05, "loss": 0.0336, "step": 52370 }, { "grad_norm": 0.3642178475856781, "learning_rate": 5.020006931404726e-05, "loss": 0.0307, "step": 52380 }, { "grad_norm": 0.3507370352745056, "learning_rate": 5.0183534737602014e-05, "loss": 0.0379, "step": 52390 }, { "grad_norm": 0.41619783639907837, "learning_rate": 5.0167000141085686e-05, "loss": 0.0355, "step": 52400 }, { "grad_norm": 0.3844202756881714, "learning_rate": 5.015046552630647e-05, "loss": 0.0364, "step": 52410 }, { "grad_norm": 0.26864054799079895, "learning_rate": 5.013393089507258e-05, "loss": 0.034, "step": 52420 }, { "grad_norm": 0.48526066541671753, "learning_rate": 5.011739624919219e-05, "loss": 0.0361, "step": 52430 }, { "grad_norm": 0.2923489809036255, "learning_rate": 5.0100861590473545e-05, "loss": 0.0319, "step": 52440 }, { "grad_norm": 0.3225059509277344, "learning_rate": 5.008432692072482e-05, "loss": 0.0432, "step": 52450 }, { "grad_norm": 0.3653843402862549, "learning_rate": 5.006779224175423e-05, "loss": 0.033, "step": 52460 }, { "grad_norm": 0.3799343407154083, "learning_rate": 5.005125755536998e-05, "loss": 0.0314, "step": 52470 }, { "grad_norm": 0.42524346709251404, "learning_rate": 5.003472286338029e-05, "loss": 0.0312, "step": 52480 }, { "grad_norm": 0.4880138039588928, "learning_rate": 5.001818816759335e-05, "loss": 0.0287, "step": 52490 }, { "grad_norm": 0.3507528603076935, "learning_rate": 5.000165346981738e-05, "loss": 0.0331, "step": 52500 }, { "grad_norm": 0.42275944352149963, "learning_rate": 4.9985118771860595e-05, "loss": 0.0328, "step": 52510 }, { "grad_norm": 0.4413544833660126, "learning_rate": 4.996858407553119e-05, "loss": 0.0299, "step": 52520 }, { "grad_norm": 0.24490821361541748, "learning_rate": 4.99520493826374e-05, "loss": 0.0314, "step": 52530 }, { "grad_norm": 0.3053531348705292, "learning_rate": 4.9935514694987384e-05, "loss": 0.0412, "step": 52540 }, { "grad_norm": 0.3689914047718048, "learning_rate": 4.991898001438941e-05, "loss": 0.0352, "step": 52550 }, { "grad_norm": 0.36331820487976074, "learning_rate": 4.990244534265166e-05, "loss": 0.0312, "step": 52560 }, { "grad_norm": 0.42190030217170715, "learning_rate": 4.988591068158234e-05, "loss": 0.0319, "step": 52570 }, { "grad_norm": 0.3061712086200714, "learning_rate": 4.986937603298965e-05, "loss": 0.0307, "step": 52580 }, { "grad_norm": 0.42739343643188477, "learning_rate": 4.985284139868181e-05, "loss": 0.0389, "step": 52590 }, { "grad_norm": 0.46007758378982544, "learning_rate": 4.9836306780467e-05, "loss": 0.0351, "step": 52600 }, { "grad_norm": 0.38165444135665894, "learning_rate": 4.981977218015343e-05, "loss": 0.0347, "step": 52610 }, { "grad_norm": 0.4387998878955841, "learning_rate": 4.980323759954932e-05, "loss": 0.0322, "step": 52620 }, { "grad_norm": 0.39635592699050903, "learning_rate": 4.978670304046284e-05, "loss": 0.0355, "step": 52630 }, { "grad_norm": 0.2528732120990753, "learning_rate": 4.9770168504702195e-05, "loss": 0.0344, "step": 52640 }, { "grad_norm": 0.3445977568626404, "learning_rate": 4.975363399407559e-05, "loss": 0.0373, "step": 52650 }, { "grad_norm": 0.45523175597190857, "learning_rate": 4.9737099510391194e-05, "loss": 0.0359, "step": 52660 }, { "grad_norm": 0.347787082195282, "learning_rate": 4.972056505545719e-05, "loss": 0.0371, "step": 52670 }, { "grad_norm": 0.35804417729377747, "learning_rate": 4.9704030631081775e-05, "loss": 0.0345, "step": 52680 }, { "grad_norm": 0.3597642183303833, "learning_rate": 4.968749623907313e-05, "loss": 0.0344, "step": 52690 }, { "grad_norm": 0.39137300848960876, "learning_rate": 4.967096188123943e-05, "loss": 0.0325, "step": 52700 }, { "grad_norm": 0.3821639120578766, "learning_rate": 4.965442755938884e-05, "loss": 0.0329, "step": 52710 }, { "grad_norm": 0.4461747705936432, "learning_rate": 4.963789327532954e-05, "loss": 0.0339, "step": 52720 }, { "grad_norm": 0.4020126461982727, "learning_rate": 4.962135903086968e-05, "loss": 0.0364, "step": 52730 }, { "grad_norm": 0.3534533679485321, "learning_rate": 4.960482482781741e-05, "loss": 0.033, "step": 52740 }, { "grad_norm": 0.6701934933662415, "learning_rate": 4.958829066798092e-05, "loss": 0.0341, "step": 52750 }, { "grad_norm": 0.6437177658081055, "learning_rate": 4.957175655316835e-05, "loss": 0.0385, "step": 52760 }, { "grad_norm": 0.38405612111091614, "learning_rate": 4.9555222485187834e-05, "loss": 0.03, "step": 52770 }, { "grad_norm": 0.2843431830406189, "learning_rate": 4.953868846584752e-05, "loss": 0.0342, "step": 52780 }, { "grad_norm": 0.43338117003440857, "learning_rate": 4.9522154496955544e-05, "loss": 0.0342, "step": 52790 }, { "grad_norm": 0.3477262556552887, "learning_rate": 4.950562058032001e-05, "loss": 0.0323, "step": 52800 }, { "grad_norm": 0.378068745136261, "learning_rate": 4.948908671774909e-05, "loss": 0.0328, "step": 52810 }, { "grad_norm": 0.29833483695983887, "learning_rate": 4.947255291105087e-05, "loss": 0.0314, "step": 52820 }, { "grad_norm": 0.2897838056087494, "learning_rate": 4.945601916203348e-05, "loss": 0.0342, "step": 52830 }, { "grad_norm": 0.3479025959968567, "learning_rate": 4.9439485472505016e-05, "loss": 0.032, "step": 52840 }, { "grad_norm": 0.34477102756500244, "learning_rate": 4.942295184427356e-05, "loss": 0.0331, "step": 52850 }, { "grad_norm": 0.4858569800853729, "learning_rate": 4.9406418279147234e-05, "loss": 0.0307, "step": 52860 }, { "grad_norm": 0.9309946298599243, "learning_rate": 4.9389884778934116e-05, "loss": 0.032, "step": 52870 }, { "grad_norm": 0.35216325521469116, "learning_rate": 4.937335134544228e-05, "loss": 0.0371, "step": 52880 }, { "grad_norm": 0.3937421441078186, "learning_rate": 4.935681798047979e-05, "loss": 0.0348, "step": 52890 }, { "grad_norm": 0.36717110872268677, "learning_rate": 4.934028468585473e-05, "loss": 0.0327, "step": 52900 }, { "grad_norm": 0.5679059028625488, "learning_rate": 4.932375146337514e-05, "loss": 0.0357, "step": 52910 }, { "grad_norm": 0.9450230002403259, "learning_rate": 4.930721831484908e-05, "loss": 0.0305, "step": 52920 }, { "grad_norm": 0.46491849422454834, "learning_rate": 4.929068524208456e-05, "loss": 0.0397, "step": 52930 }, { "grad_norm": 0.4168882966041565, "learning_rate": 4.927415224688965e-05, "loss": 0.0342, "step": 52940 }, { "grad_norm": 0.2866552770137787, "learning_rate": 4.9257619331072375e-05, "loss": 0.0315, "step": 52950 }, { "grad_norm": 0.44738927483558655, "learning_rate": 4.924108649644072e-05, "loss": 0.0344, "step": 52960 }, { "grad_norm": 0.4402548372745514, "learning_rate": 4.922455374480271e-05, "loss": 0.0337, "step": 52970 }, { "grad_norm": 0.3609345555305481, "learning_rate": 4.920802107796634e-05, "loss": 0.039, "step": 52980 }, { "grad_norm": 0.6464665532112122, "learning_rate": 4.9191488497739564e-05, "loss": 0.0291, "step": 52990 }, { "grad_norm": 0.5690604448318481, "learning_rate": 4.917495600593042e-05, "loss": 0.0313, "step": 53000 }, { "grad_norm": 0.631554901599884, "learning_rate": 4.915842360434684e-05, "loss": 0.0323, "step": 53010 }, { "grad_norm": 0.3612495958805084, "learning_rate": 4.914189129479679e-05, "loss": 0.032, "step": 53020 }, { "grad_norm": 0.3157528042793274, "learning_rate": 4.912535907908822e-05, "loss": 0.0358, "step": 53030 }, { "grad_norm": 0.43577495217323303, "learning_rate": 4.910882695902907e-05, "loss": 0.0372, "step": 53040 }, { "grad_norm": 0.467812716960907, "learning_rate": 4.9092294936427234e-05, "loss": 0.0335, "step": 53050 }, { "grad_norm": 0.3311432898044586, "learning_rate": 4.9075763013090685e-05, "loss": 0.0297, "step": 53060 }, { "grad_norm": 0.3271924555301666, "learning_rate": 4.9059231190827305e-05, "loss": 0.0281, "step": 53070 }, { "grad_norm": 0.308826208114624, "learning_rate": 4.904269947144499e-05, "loss": 0.0371, "step": 53080 }, { "grad_norm": 0.6792169213294983, "learning_rate": 4.902616785675161e-05, "loss": 0.0383, "step": 53090 }, { "grad_norm": 0.43805795907974243, "learning_rate": 4.900963634855505e-05, "loss": 0.0364, "step": 53100 }, { "grad_norm": 0.42681583762168884, "learning_rate": 4.8993104948663173e-05, "loss": 0.0357, "step": 53110 }, { "grad_norm": 0.3033006191253662, "learning_rate": 4.8976573658883823e-05, "loss": 0.034, "step": 53120 }, { "grad_norm": 0.5946709513664246, "learning_rate": 4.896004248102484e-05, "loss": 0.0361, "step": 53130 }, { "grad_norm": 0.4506807327270508, "learning_rate": 4.8943511416894045e-05, "loss": 0.0391, "step": 53140 }, { "grad_norm": 0.3699372708797455, "learning_rate": 4.892698046829926e-05, "loss": 0.0355, "step": 53150 }, { "grad_norm": 0.33046478033065796, "learning_rate": 4.8910449637048275e-05, "loss": 0.0359, "step": 53160 }, { "grad_norm": 0.30385449528694153, "learning_rate": 4.8893918924948884e-05, "loss": 0.0319, "step": 53170 }, { "grad_norm": 0.9431731104850769, "learning_rate": 4.8877388333808835e-05, "loss": 0.035, "step": 53180 }, { "grad_norm": 0.24147744476795197, "learning_rate": 4.8860857865435925e-05, "loss": 0.0331, "step": 53190 }, { "grad_norm": 0.28737547993659973, "learning_rate": 4.884432752163789e-05, "loss": 0.031, "step": 53200 }, { "grad_norm": 0.35219794511795044, "learning_rate": 4.8827797304222466e-05, "loss": 0.0379, "step": 53210 }, { "grad_norm": 0.43635496497154236, "learning_rate": 4.8811267214997366e-05, "loss": 0.0294, "step": 53220 }, { "grad_norm": 0.4168446362018585, "learning_rate": 4.879473725577029e-05, "loss": 0.0378, "step": 53230 }, { "grad_norm": 0.5446155667304993, "learning_rate": 4.877820742834891e-05, "loss": 0.0294, "step": 53240 }, { "grad_norm": 0.512025773525238, "learning_rate": 4.8761677734540956e-05, "loss": 0.0328, "step": 53250 }, { "grad_norm": 0.5237141847610474, "learning_rate": 4.874514817615407e-05, "loss": 0.0341, "step": 53260 }, { "grad_norm": 0.2894952893257141, "learning_rate": 4.872861875499588e-05, "loss": 0.0298, "step": 53270 }, { "grad_norm": 0.4845147132873535, "learning_rate": 4.871208947287404e-05, "loss": 0.0332, "step": 53280 }, { "grad_norm": 0.5491693019866943, "learning_rate": 4.869556033159615e-05, "loss": 0.0322, "step": 53290 }, { "grad_norm": 0.36091503500938416, "learning_rate": 4.8679031332969835e-05, "loss": 0.0334, "step": 53300 }, { "grad_norm": 0.610154390335083, "learning_rate": 4.8662502478802635e-05, "loss": 0.0343, "step": 53310 }, { "grad_norm": 0.37975284457206726, "learning_rate": 4.8645973770902176e-05, "loss": 0.0326, "step": 53320 }, { "grad_norm": 0.42267197370529175, "learning_rate": 4.862944521107599e-05, "loss": 0.0341, "step": 53330 }, { "grad_norm": 0.6378635168075562, "learning_rate": 4.861291680113162e-05, "loss": 0.0252, "step": 53340 }, { "grad_norm": 0.4336947202682495, "learning_rate": 4.8596388542876563e-05, "loss": 0.0337, "step": 53350 }, { "grad_norm": 0.40511783957481384, "learning_rate": 4.857986043811836e-05, "loss": 0.0337, "step": 53360 }, { "grad_norm": 0.6238943934440613, "learning_rate": 4.856333248866448e-05, "loss": 0.0341, "step": 53370 }, { "grad_norm": 0.41489556431770325, "learning_rate": 4.85468046963224e-05, "loss": 0.0282, "step": 53380 }, { "grad_norm": 0.3813101351261139, "learning_rate": 4.853027706289957e-05, "loss": 0.0278, "step": 53390 }, { "grad_norm": 0.33048829436302185, "learning_rate": 4.851374959020344e-05, "loss": 0.0336, "step": 53400 }, { "grad_norm": 0.29659706354141235, "learning_rate": 4.849722228004141e-05, "loss": 0.032, "step": 53410 }, { "grad_norm": 0.4482504725456238, "learning_rate": 4.848069513422091e-05, "loss": 0.0318, "step": 53420 }, { "grad_norm": 0.53264319896698, "learning_rate": 4.8464168154549294e-05, "loss": 0.0316, "step": 53430 }, { "grad_norm": 0.4292359948158264, "learning_rate": 4.844764134283392e-05, "loss": 0.0281, "step": 53440 }, { "grad_norm": 0.3746204078197479, "learning_rate": 4.843111470088218e-05, "loss": 0.035, "step": 53450 }, { "grad_norm": 0.2767391502857208, "learning_rate": 4.841458823050137e-05, "loss": 0.0282, "step": 53460 }, { "grad_norm": 0.3831104040145874, "learning_rate": 4.8398061933498816e-05, "loss": 0.0314, "step": 53470 }, { "grad_norm": 0.32107898592948914, "learning_rate": 4.83815358116818e-05, "loss": 0.0297, "step": 53480 }, { "grad_norm": 0.34138646721839905, "learning_rate": 4.8365009866857584e-05, "loss": 0.0321, "step": 53490 }, { "grad_norm": 0.3864513635635376, "learning_rate": 4.834848410083342e-05, "loss": 0.0317, "step": 53500 }, { "grad_norm": 0.36836057901382446, "learning_rate": 4.833195851541657e-05, "loss": 0.0334, "step": 53510 }, { "grad_norm": 1.6078364849090576, "learning_rate": 4.831543311241423e-05, "loss": 0.0292, "step": 53520 }, { "grad_norm": 0.33284780383110046, "learning_rate": 4.829890789363359e-05, "loss": 0.032, "step": 53530 }, { "grad_norm": 0.27621713280677795, "learning_rate": 4.828238286088182e-05, "loss": 0.0349, "step": 53540 }, { "grad_norm": 0.5076206922531128, "learning_rate": 4.826585801596609e-05, "loss": 0.0398, "step": 53550 }, { "grad_norm": 0.39492741227149963, "learning_rate": 4.8249333360693486e-05, "loss": 0.0325, "step": 53560 }, { "grad_norm": 0.3698011040687561, "learning_rate": 4.8232808896871185e-05, "loss": 0.0282, "step": 53570 }, { "grad_norm": 0.3348153829574585, "learning_rate": 4.821628462630624e-05, "loss": 0.0332, "step": 53580 }, { "grad_norm": 0.42320048809051514, "learning_rate": 4.8199760550805735e-05, "loss": 0.0318, "step": 53590 }, { "grad_norm": 0.3328370451927185, "learning_rate": 4.818323667217669e-05, "loss": 0.0312, "step": 53600 }, { "grad_norm": 0.406506210565567, "learning_rate": 4.8166712992226165e-05, "loss": 0.0302, "step": 53610 }, { "grad_norm": 0.38381025195121765, "learning_rate": 4.8150189512761154e-05, "loss": 0.031, "step": 53620 }, { "grad_norm": 0.3947625458240509, "learning_rate": 4.813366623558862e-05, "loss": 0.0335, "step": 53630 }, { "grad_norm": 0.39657700061798096, "learning_rate": 4.811714316251554e-05, "loss": 0.0306, "step": 53640 }, { "grad_norm": 0.3425171673297882, "learning_rate": 4.810062029534886e-05, "loss": 0.0295, "step": 53650 }, { "grad_norm": 0.468868225812912, "learning_rate": 4.8084097635895505e-05, "loss": 0.0369, "step": 53660 }, { "grad_norm": 0.2561746835708618, "learning_rate": 4.8067575185962335e-05, "loss": 0.0309, "step": 53670 }, { "grad_norm": 0.3726573586463928, "learning_rate": 4.805105294735625e-05, "loss": 0.0333, "step": 53680 }, { "grad_norm": 0.322480171918869, "learning_rate": 4.803453092188406e-05, "loss": 0.0301, "step": 53690 }, { "grad_norm": 0.2256719022989273, "learning_rate": 4.801800911135263e-05, "loss": 0.0293, "step": 53700 }, { "grad_norm": 0.3173922002315521, "learning_rate": 4.8001487517568744e-05, "loss": 0.0343, "step": 53710 }, { "grad_norm": 0.4379962980747223, "learning_rate": 4.798496614233918e-05, "loss": 0.0326, "step": 53720 }, { "grad_norm": 0.34826260805130005, "learning_rate": 4.796844498747069e-05, "loss": 0.0351, "step": 53730 }, { "grad_norm": 0.39709004759788513, "learning_rate": 4.795192405477e-05, "loss": 0.0329, "step": 53740 }, { "grad_norm": 0.42562246322631836, "learning_rate": 4.79354033460438e-05, "loss": 0.0353, "step": 53750 }, { "grad_norm": 0.38007012009620667, "learning_rate": 4.79188828630988e-05, "loss": 0.0305, "step": 53760 }, { "grad_norm": 0.5892594456672668, "learning_rate": 4.790236260774165e-05, "loss": 0.032, "step": 53770 }, { "grad_norm": 0.296315997838974, "learning_rate": 4.788584258177896e-05, "loss": 0.0373, "step": 53780 }, { "grad_norm": 0.36263468861579895, "learning_rate": 4.7869322787017356e-05, "loss": 0.0311, "step": 53790 }, { "grad_norm": 0.40713152289390564, "learning_rate": 4.78528032252634e-05, "loss": 0.0341, "step": 53800 }, { "grad_norm": 0.6850770711898804, "learning_rate": 4.783628389832366e-05, "loss": 0.0294, "step": 53810 }, { "grad_norm": 0.3390451967716217, "learning_rate": 4.7819764808004644e-05, "loss": 0.0311, "step": 53820 }, { "grad_norm": 0.4957430362701416, "learning_rate": 4.780324595611289e-05, "loss": 0.0363, "step": 53830 }, { "grad_norm": 0.33389484882354736, "learning_rate": 4.7786727344454854e-05, "loss": 0.0264, "step": 53840 }, { "grad_norm": 0.312360554933548, "learning_rate": 4.777020897483697e-05, "loss": 0.0292, "step": 53850 }, { "grad_norm": 0.4209454357624054, "learning_rate": 4.775369084906569e-05, "loss": 0.0341, "step": 53860 }, { "grad_norm": 0.3746173083782196, "learning_rate": 4.773717296894742e-05, "loss": 0.0332, "step": 53870 }, { "grad_norm": 0.3119533061981201, "learning_rate": 4.772065533628848e-05, "loss": 0.0384, "step": 53880 }, { "grad_norm": 0.47287946939468384, "learning_rate": 4.7704137952895265e-05, "loss": 0.0372, "step": 53890 }, { "grad_norm": 0.5677839517593384, "learning_rate": 4.768762082057407e-05, "loss": 0.0393, "step": 53900 }, { "grad_norm": 0.2822929322719574, "learning_rate": 4.7671103941131186e-05, "loss": 0.0366, "step": 53910 }, { "grad_norm": 0.34418585896492004, "learning_rate": 4.765458731637287e-05, "loss": 0.0407, "step": 53920 }, { "grad_norm": 0.4250960350036621, "learning_rate": 4.763807094810537e-05, "loss": 0.0323, "step": 53930 }, { "grad_norm": 0.36682775616645813, "learning_rate": 4.762155483813487e-05, "loss": 0.0341, "step": 53940 }, { "grad_norm": 0.24314430356025696, "learning_rate": 4.760503898826754e-05, "loss": 0.0314, "step": 53950 }, { "grad_norm": 0.2898198664188385, "learning_rate": 4.758852340030957e-05, "loss": 0.0299, "step": 53960 }, { "grad_norm": 0.4933239221572876, "learning_rate": 4.757200807606705e-05, "loss": 0.035, "step": 53970 }, { "grad_norm": 0.4146845042705536, "learning_rate": 4.755549301734607e-05, "loss": 0.0316, "step": 53980 }, { "grad_norm": 0.35776299238204956, "learning_rate": 4.753897822595271e-05, "loss": 0.0307, "step": 53990 }, { "grad_norm": 0.5052096843719482, "learning_rate": 4.7522463703692994e-05, "loss": 0.029, "step": 54000 }, { "grad_norm": 0.3729240298271179, "learning_rate": 4.75059494523729e-05, "loss": 0.0302, "step": 54010 }, { "grad_norm": 0.36425188183784485, "learning_rate": 4.748943547379843e-05, "loss": 0.0342, "step": 54020 }, { "grad_norm": 0.4050268232822418, "learning_rate": 4.7472921769775535e-05, "loss": 0.0385, "step": 54030 }, { "grad_norm": 0.2879685163497925, "learning_rate": 4.745640834211011e-05, "loss": 0.0301, "step": 54040 }, { "grad_norm": 0.7154462337493896, "learning_rate": 4.7439895192608054e-05, "loss": 0.038, "step": 54050 }, { "grad_norm": 0.22419531643390656, "learning_rate": 4.7423382323075186e-05, "loss": 0.0281, "step": 54060 }, { "grad_norm": 0.36753275990486145, "learning_rate": 4.740686973531737e-05, "loss": 0.0325, "step": 54070 }, { "grad_norm": 0.3522323966026306, "learning_rate": 4.7390357431140375e-05, "loss": 0.0326, "step": 54080 }, { "grad_norm": 0.3383682370185852, "learning_rate": 4.737384541234998e-05, "loss": 0.0306, "step": 54090 }, { "grad_norm": 0.35716986656188965, "learning_rate": 4.735733368075189e-05, "loss": 0.0302, "step": 54100 }, { "grad_norm": 0.4493143856525421, "learning_rate": 4.734082223815182e-05, "loss": 0.0342, "step": 54110 }, { "grad_norm": 0.3480767011642456, "learning_rate": 4.7324311086355436e-05, "loss": 0.0399, "step": 54120 }, { "grad_norm": 0.24944671988487244, "learning_rate": 4.730780022716836e-05, "loss": 0.0312, "step": 54130 }, { "grad_norm": 0.43959274888038635, "learning_rate": 4.729128966239619e-05, "loss": 0.0256, "step": 54140 }, { "grad_norm": 0.28766804933547974, "learning_rate": 4.7274779393844536e-05, "loss": 0.0305, "step": 54150 }, { "grad_norm": 0.30146145820617676, "learning_rate": 4.725826942331891e-05, "loss": 0.0347, "step": 54160 }, { "grad_norm": 0.6089681386947632, "learning_rate": 4.724175975262481e-05, "loss": 0.0339, "step": 54170 }, { "grad_norm": 0.33241143822669983, "learning_rate": 4.722525038356772e-05, "loss": 0.0316, "step": 54180 }, { "grad_norm": 0.47285833954811096, "learning_rate": 4.7208741317953075e-05, "loss": 0.0343, "step": 54190 }, { "grad_norm": 0.6402676701545715, "learning_rate": 4.7192232557586264e-05, "loss": 0.0335, "step": 54200 }, { "grad_norm": 0.4897245764732361, "learning_rate": 4.717572410427269e-05, "loss": 0.0332, "step": 54210 }, { "grad_norm": 0.5213755369186401, "learning_rate": 4.71592159598177e-05, "loss": 0.0305, "step": 54220 }, { "grad_norm": 0.32839158177375793, "learning_rate": 4.714270812602657e-05, "loss": 0.0296, "step": 54230 }, { "grad_norm": 0.3321395516395569, "learning_rate": 4.712620060470458e-05, "loss": 0.0292, "step": 54240 }, { "grad_norm": 0.5658741593360901, "learning_rate": 4.7109693397656986e-05, "loss": 0.0303, "step": 54250 }, { "grad_norm": 0.33650076389312744, "learning_rate": 4.709318650668894e-05, "loss": 0.035, "step": 54260 }, { "grad_norm": 0.8171434998512268, "learning_rate": 4.7076679933605676e-05, "loss": 0.032, "step": 54270 }, { "grad_norm": 0.36313778162002563, "learning_rate": 4.70601736802123e-05, "loss": 0.0352, "step": 54280 }, { "grad_norm": 0.34665822982788086, "learning_rate": 4.704366774831391e-05, "loss": 0.0325, "step": 54290 }, { "grad_norm": 0.9172725677490234, "learning_rate": 4.702716213971557e-05, "loss": 0.0266, "step": 54300 }, { "grad_norm": 0.30669674277305603, "learning_rate": 4.70106568562223e-05, "loss": 0.0347, "step": 54310 }, { "grad_norm": 0.6200782656669617, "learning_rate": 4.6994151899639105e-05, "loss": 0.0371, "step": 54320 }, { "grad_norm": 0.47046759724617004, "learning_rate": 4.697764727177093e-05, "loss": 0.0352, "step": 54330 }, { "grad_norm": 0.449520081281662, "learning_rate": 4.696114297442272e-05, "loss": 0.0348, "step": 54340 }, { "grad_norm": 0.37141767144203186, "learning_rate": 4.694463900939933e-05, "loss": 0.034, "step": 54350 }, { "grad_norm": 0.36171451210975647, "learning_rate": 4.692813537850564e-05, "loss": 0.0273, "step": 54360 }, { "grad_norm": 0.375985711812973, "learning_rate": 4.6911632083546445e-05, "loss": 0.0347, "step": 54370 }, { "grad_norm": 0.317237913608551, "learning_rate": 4.6895129126326525e-05, "loss": 0.0371, "step": 54380 }, { "grad_norm": 0.712943434715271, "learning_rate": 4.687862650865059e-05, "loss": 0.0322, "step": 54390 }, { "grad_norm": 0.4065004885196686, "learning_rate": 4.686212423232339e-05, "loss": 0.0311, "step": 54400 }, { "grad_norm": 0.40451329946517944, "learning_rate": 4.6845622299149565e-05, "loss": 0.0401, "step": 54410 }, { "grad_norm": 0.2812100648880005, "learning_rate": 4.682912071093374e-05, "loss": 0.0354, "step": 54420 }, { "grad_norm": 0.28885582089424133, "learning_rate": 4.681261946948052e-05, "loss": 0.0259, "step": 54430 }, { "grad_norm": 0.34840402007102966, "learning_rate": 4.679611857659443e-05, "loss": 0.0281, "step": 54440 }, { "grad_norm": 0.28136423230171204, "learning_rate": 4.6779618034079975e-05, "loss": 0.031, "step": 54450 }, { "grad_norm": 0.3993065655231476, "learning_rate": 4.6763117843741675e-05, "loss": 0.0337, "step": 54460 }, { "grad_norm": 0.3941417634487152, "learning_rate": 4.674661800738394e-05, "loss": 0.0304, "step": 54470 }, { "grad_norm": 0.4075668454170227, "learning_rate": 4.673011852681117e-05, "loss": 0.0289, "step": 54480 }, { "grad_norm": 0.48704150319099426, "learning_rate": 4.671361940382772e-05, "loss": 0.0328, "step": 54490 }, { "grad_norm": 0.29803305864334106, "learning_rate": 4.669712064023791e-05, "loss": 0.0331, "step": 54500 }, { "grad_norm": 0.35220420360565186, "learning_rate": 4.668062223784602e-05, "loss": 0.0287, "step": 54510 }, { "grad_norm": 0.3055218756198883, "learning_rate": 4.6664124198456274e-05, "loss": 0.0279, "step": 54520 }, { "grad_norm": 0.2852354645729065, "learning_rate": 4.6647626523872905e-05, "loss": 0.0304, "step": 54530 }, { "grad_norm": 0.33457067608833313, "learning_rate": 4.6631129215900066e-05, "loss": 0.0323, "step": 54540 }, { "grad_norm": 0.4082505404949188, "learning_rate": 4.661463227634187e-05, "loss": 0.0332, "step": 54550 }, { "grad_norm": 0.3878346085548401, "learning_rate": 4.659813570700238e-05, "loss": 0.0298, "step": 54560 }, { "grad_norm": 0.4767141342163086, "learning_rate": 4.658163950968568e-05, "loss": 0.0352, "step": 54570 }, { "grad_norm": 0.9948412179946899, "learning_rate": 4.6565143686195726e-05, "loss": 0.0324, "step": 54580 }, { "grad_norm": 0.27798992395401, "learning_rate": 4.654864823833649e-05, "loss": 0.0345, "step": 54590 }, { "grad_norm": 0.3167765140533447, "learning_rate": 4.65321531679119e-05, "loss": 0.0335, "step": 54600 }, { "grad_norm": 0.37056848406791687, "learning_rate": 4.6515658476725834e-05, "loss": 0.0326, "step": 54610 }, { "grad_norm": 0.30497974157333374, "learning_rate": 4.649916416658212e-05, "loss": 0.0319, "step": 54620 }, { "grad_norm": 0.25314590334892273, "learning_rate": 4.648267023928455e-05, "loss": 0.0289, "step": 54630 }, { "grad_norm": 0.46789297461509705, "learning_rate": 4.646617669663687e-05, "loss": 0.0395, "step": 54640 }, { "grad_norm": 0.291093111038208, "learning_rate": 4.644968354044277e-05, "loss": 0.0337, "step": 54650 }, { "grad_norm": 0.4124034643173218, "learning_rate": 4.643319077250596e-05, "loss": 0.0331, "step": 54660 }, { "grad_norm": 0.36132049560546875, "learning_rate": 4.6416698394630055e-05, "loss": 0.0322, "step": 54670 }, { "grad_norm": 0.3457026481628418, "learning_rate": 4.640020640861862e-05, "loss": 0.0314, "step": 54680 }, { "grad_norm": 0.28206366300582886, "learning_rate": 4.6383714816275195e-05, "loss": 0.0317, "step": 54690 }, { "grad_norm": 0.40688174962997437, "learning_rate": 4.636722361940329e-05, "loss": 0.0328, "step": 54700 }, { "grad_norm": 0.299690306186676, "learning_rate": 4.635073281980632e-05, "loss": 0.0295, "step": 54710 }, { "grad_norm": 0.3003305196762085, "learning_rate": 4.633424241928775e-05, "loss": 0.0286, "step": 54720 }, { "grad_norm": 0.306120365858078, "learning_rate": 4.6317752419650914e-05, "loss": 0.0315, "step": 54730 }, { "grad_norm": 0.24512311816215515, "learning_rate": 4.630126282269914e-05, "loss": 0.0316, "step": 54740 }, { "grad_norm": 0.33558812737464905, "learning_rate": 4.628477363023569e-05, "loss": 0.0345, "step": 54750 }, { "grad_norm": 0.4826323091983795, "learning_rate": 4.626828484406383e-05, "loss": 0.0395, "step": 54760 }, { "grad_norm": 0.3087748885154724, "learning_rate": 4.625179646598669e-05, "loss": 0.032, "step": 54770 }, { "grad_norm": 0.45307955145835876, "learning_rate": 4.6235308497807487e-05, "loss": 0.032, "step": 54780 }, { "grad_norm": 0.21815434098243713, "learning_rate": 4.6218820941329276e-05, "loss": 0.0313, "step": 54790 }, { "grad_norm": 0.2997281849384308, "learning_rate": 4.620233379835513e-05, "loss": 0.0277, "step": 54800 }, { "grad_norm": 0.376443088054657, "learning_rate": 4.618584707068804e-05, "loss": 0.0289, "step": 54810 }, { "grad_norm": 0.9380322694778442, "learning_rate": 4.616936076013099e-05, "loss": 0.0328, "step": 54820 }, { "grad_norm": 0.26824259757995605, "learning_rate": 4.615287486848688e-05, "loss": 0.0293, "step": 54830 }, { "grad_norm": 0.2778853178024292, "learning_rate": 4.6136389397558584e-05, "loss": 0.0288, "step": 54840 }, { "grad_norm": 0.3734741806983948, "learning_rate": 4.6119904349148935e-05, "loss": 0.029, "step": 54850 }, { "grad_norm": 0.38443419337272644, "learning_rate": 4.6103419725060726e-05, "loss": 0.038, "step": 54860 }, { "grad_norm": 0.3631535768508911, "learning_rate": 4.608693552709667e-05, "loss": 0.0344, "step": 54870 }, { "grad_norm": 0.41665172576904297, "learning_rate": 4.607045175705947e-05, "loss": 0.0318, "step": 54880 }, { "grad_norm": 0.37348857522010803, "learning_rate": 4.605396841675176e-05, "loss": 0.0345, "step": 54890 }, { "grad_norm": 0.4332737922668457, "learning_rate": 4.6037485507976094e-05, "loss": 0.0332, "step": 54900 }, { "grad_norm": 0.2877962589263916, "learning_rate": 4.602100303253509e-05, "loss": 0.0323, "step": 54910 }, { "grad_norm": 0.41535982489585876, "learning_rate": 4.600452099223122e-05, "loss": 0.031, "step": 54920 }, { "grad_norm": 0.4586166739463806, "learning_rate": 4.598803938886691e-05, "loss": 0.0343, "step": 54930 }, { "grad_norm": 0.47212815284729004, "learning_rate": 4.597155822424459e-05, "loss": 0.03, "step": 54940 }, { "grad_norm": 0.2914029061794281, "learning_rate": 4.59550775001666e-05, "loss": 0.0294, "step": 54950 }, { "grad_norm": 0.3863583207130432, "learning_rate": 4.5938597218435234e-05, "loss": 0.0322, "step": 54960 }, { "grad_norm": 0.3402671813964844, "learning_rate": 4.592211738085279e-05, "loss": 0.0389, "step": 54970 }, { "grad_norm": 0.40383774042129517, "learning_rate": 4.590563798922146e-05, "loss": 0.032, "step": 54980 }, { "grad_norm": 0.3333691954612732, "learning_rate": 4.5889159045343404e-05, "loss": 0.0334, "step": 54990 }, { "grad_norm": 0.32595956325531006, "learning_rate": 4.5872680551020734e-05, "loss": 0.0328, "step": 55000 }, { "grad_norm": 0.5256057977676392, "learning_rate": 4.585620250805552e-05, "loss": 0.0319, "step": 55010 }, { "grad_norm": 0.5206303000450134, "learning_rate": 4.583972491824975e-05, "loss": 0.0272, "step": 55020 }, { "grad_norm": 0.3635873794555664, "learning_rate": 4.582324778340541e-05, "loss": 0.0284, "step": 55030 }, { "grad_norm": 0.4271528720855713, "learning_rate": 4.5806771105324424e-05, "loss": 0.0296, "step": 55040 }, { "grad_norm": 0.45267540216445923, "learning_rate": 4.579029488580865e-05, "loss": 0.0261, "step": 55050 }, { "grad_norm": 0.5308568477630615, "learning_rate": 4.5773819126659876e-05, "loss": 0.0283, "step": 55060 }, { "grad_norm": 0.25627830624580383, "learning_rate": 4.575734382967991e-05, "loss": 0.0285, "step": 55070 }, { "grad_norm": 0.3744828999042511, "learning_rate": 4.574086899667044e-05, "loss": 0.0311, "step": 55080 }, { "grad_norm": 0.3460668921470642, "learning_rate": 4.5724394629433106e-05, "loss": 0.0338, "step": 55090 }, { "grad_norm": 0.5000987648963928, "learning_rate": 4.570792072976958e-05, "loss": 0.0386, "step": 55100 }, { "grad_norm": 0.31507977843284607, "learning_rate": 4.569144729948138e-05, "loss": 0.0265, "step": 55110 }, { "grad_norm": 0.23331502079963684, "learning_rate": 4.5674974340370033e-05, "loss": 0.0273, "step": 55120 }, { "grad_norm": 0.40359196066856384, "learning_rate": 4.565850185423699e-05, "loss": 0.0315, "step": 55130 }, { "grad_norm": 0.3495469093322754, "learning_rate": 4.564202984288365e-05, "loss": 0.0359, "step": 55140 }, { "grad_norm": 0.2916516661643982, "learning_rate": 4.5625558308111356e-05, "loss": 0.031, "step": 55150 }, { "grad_norm": 0.46811825037002563, "learning_rate": 4.560908725172145e-05, "loss": 0.035, "step": 55160 }, { "grad_norm": 0.301383912563324, "learning_rate": 4.559261667551516e-05, "loss": 0.037, "step": 55170 }, { "grad_norm": 0.3133619427680969, "learning_rate": 4.5576146581293685e-05, "loss": 0.039, "step": 55180 }, { "grad_norm": 0.4116348624229431, "learning_rate": 4.5559676970858164e-05, "loss": 0.0354, "step": 55190 }, { "grad_norm": 0.45025748014450073, "learning_rate": 4.5543207846009705e-05, "loss": 0.0289, "step": 55200 }, { "grad_norm": 0.38429856300354004, "learning_rate": 4.552673920854933e-05, "loss": 0.0331, "step": 55210 }, { "grad_norm": 0.31951388716697693, "learning_rate": 4.5510271060278006e-05, "loss": 0.028, "step": 55220 }, { "grad_norm": 0.37197044491767883, "learning_rate": 4.5493803402996704e-05, "loss": 0.0363, "step": 55230 }, { "grad_norm": 0.26155760884284973, "learning_rate": 4.5477336238506295e-05, "loss": 0.0335, "step": 55240 }, { "grad_norm": 0.27510347962379456, "learning_rate": 4.546086956860759e-05, "loss": 0.0314, "step": 55250 }, { "grad_norm": 0.5109288692474365, "learning_rate": 4.5444403395101374e-05, "loss": 0.0354, "step": 55260 }, { "grad_norm": 0.45744386315345764, "learning_rate": 4.542793771978834e-05, "loss": 0.0302, "step": 55270 }, { "grad_norm": 0.4293217360973358, "learning_rate": 4.541147254446916e-05, "loss": 0.0303, "step": 55280 }, { "grad_norm": 0.3082927465438843, "learning_rate": 4.539500787094445e-05, "loss": 0.0295, "step": 55290 }, { "grad_norm": 0.4398292601108551, "learning_rate": 4.537854370101475e-05, "loss": 0.0307, "step": 55300 }, { "grad_norm": 0.4023864269256592, "learning_rate": 4.5362080036480566e-05, "loss": 0.0277, "step": 55310 }, { "grad_norm": 0.33407923579216003, "learning_rate": 4.534561687914235e-05, "loss": 0.031, "step": 55320 }, { "grad_norm": 0.4293294847011566, "learning_rate": 4.5329154230800466e-05, "loss": 0.0327, "step": 55330 }, { "grad_norm": 0.35954034328460693, "learning_rate": 4.531269209325525e-05, "loss": 0.0277, "step": 55340 }, { "grad_norm": 0.34292200207710266, "learning_rate": 4.529623046830696e-05, "loss": 0.0307, "step": 55350 }, { "grad_norm": 0.49135372042655945, "learning_rate": 4.5279769357755866e-05, "loss": 0.03, "step": 55360 }, { "grad_norm": 0.372404545545578, "learning_rate": 4.5263308763402084e-05, "loss": 0.029, "step": 55370 }, { "grad_norm": 0.297133207321167, "learning_rate": 4.524684868704574e-05, "loss": 0.029, "step": 55380 }, { "grad_norm": 0.31682032346725464, "learning_rate": 4.523038913048688e-05, "loss": 0.03, "step": 55390 }, { "grad_norm": 0.3716500699520111, "learning_rate": 4.521393009552549e-05, "loss": 0.0313, "step": 55400 }, { "grad_norm": 0.3074764311313629, "learning_rate": 4.5197471583961484e-05, "loss": 0.0292, "step": 55410 }, { "grad_norm": 0.38053449988365173, "learning_rate": 4.51810135975948e-05, "loss": 0.0331, "step": 55420 }, { "grad_norm": 0.41490134596824646, "learning_rate": 4.516455613822521e-05, "loss": 0.0268, "step": 55430 }, { "grad_norm": 0.4190380275249481, "learning_rate": 4.514809920765249e-05, "loss": 0.0323, "step": 55440 }, { "grad_norm": 0.3801417648792267, "learning_rate": 4.5131642807676356e-05, "loss": 0.0325, "step": 55450 }, { "grad_norm": 0.4683704078197479, "learning_rate": 4.511518694009644e-05, "loss": 0.032, "step": 55460 }, { "grad_norm": 0.35330677032470703, "learning_rate": 4.509873160671231e-05, "loss": 0.0316, "step": 55470 }, { "grad_norm": 0.5815457701683044, "learning_rate": 4.508227680932356e-05, "loss": 0.0298, "step": 55480 }, { "grad_norm": 0.4150192439556122, "learning_rate": 4.5065822549729605e-05, "loss": 0.0318, "step": 55490 }, { "grad_norm": 0.3771189749240875, "learning_rate": 4.504936882972989e-05, "loss": 0.0371, "step": 55500 }, { "grad_norm": 0.25645190477371216, "learning_rate": 4.503291565112374e-05, "loss": 0.0305, "step": 55510 }, { "grad_norm": 0.3152173161506653, "learning_rate": 4.501646301571048e-05, "loss": 0.0288, "step": 55520 }, { "grad_norm": 0.429790198802948, "learning_rate": 4.500001092528933e-05, "loss": 0.0326, "step": 55530 }, { "grad_norm": 0.3813479542732239, "learning_rate": 4.4983559381659455e-05, "loss": 0.0364, "step": 55540 }, { "grad_norm": 0.4897684156894684, "learning_rate": 4.496710838661999e-05, "loss": 0.031, "step": 55550 }, { "grad_norm": 0.26120543479919434, "learning_rate": 4.495065794196999e-05, "loss": 0.0324, "step": 55560 }, { "grad_norm": 0.41783612966537476, "learning_rate": 4.4934208049508455e-05, "loss": 0.0399, "step": 55570 }, { "grad_norm": 0.34752604365348816, "learning_rate": 4.4917758711034303e-05, "loss": 0.0341, "step": 55580 }, { "grad_norm": 0.4438459575176239, "learning_rate": 4.490130992834642e-05, "loss": 0.0352, "step": 55590 }, { "grad_norm": 1.1206101179122925, "learning_rate": 4.488486170324359e-05, "loss": 0.031, "step": 55600 }, { "grad_norm": 0.23682032525539398, "learning_rate": 4.486841403752463e-05, "loss": 0.0294, "step": 55610 }, { "grad_norm": 0.22643575072288513, "learning_rate": 4.4851966932988174e-05, "loss": 0.0302, "step": 55620 }, { "grad_norm": 0.30132123827934265, "learning_rate": 4.4835520391432894e-05, "loss": 0.0316, "step": 55630 }, { "grad_norm": 0.27271148562431335, "learning_rate": 4.481907441465733e-05, "loss": 0.0368, "step": 55640 }, { "grad_norm": 0.3127739131450653, "learning_rate": 4.480262900446e-05, "loss": 0.0327, "step": 55650 }, { "grad_norm": 0.3817220628261566, "learning_rate": 4.478618416263933e-05, "loss": 0.0277, "step": 55660 }, { "grad_norm": 0.31368768215179443, "learning_rate": 4.476973989099373e-05, "loss": 0.034, "step": 55670 }, { "grad_norm": 0.3958408534526825, "learning_rate": 4.475329619132153e-05, "loss": 0.0329, "step": 55680 }, { "grad_norm": 0.34972846508026123, "learning_rate": 4.4736853065420956e-05, "loss": 0.0281, "step": 55690 }, { "grad_norm": 0.4826822280883789, "learning_rate": 4.472041051509023e-05, "loss": 0.029, "step": 55700 }, { "grad_norm": 0.5765683054924011, "learning_rate": 4.470396854212746e-05, "loss": 0.0329, "step": 55710 }, { "grad_norm": 0.414530873298645, "learning_rate": 4.468752714833072e-05, "loss": 0.0301, "step": 55720 }, { "grad_norm": 0.46923187375068665, "learning_rate": 4.467108633549803e-05, "loss": 0.0355, "step": 55730 }, { "grad_norm": 0.4034045338630676, "learning_rate": 4.4654646105427335e-05, "loss": 0.0309, "step": 55740 }, { "grad_norm": 0.7654080390930176, "learning_rate": 4.463820645991651e-05, "loss": 0.0342, "step": 55750 }, { "grad_norm": 0.45556730031967163, "learning_rate": 4.462176740076334e-05, "loss": 0.0357, "step": 55760 }, { "grad_norm": 1.6803795099258423, "learning_rate": 4.4605328929765624e-05, "loss": 0.0415, "step": 55770 }, { "grad_norm": 0.29985862970352173, "learning_rate": 4.458889104872102e-05, "loss": 0.0359, "step": 55780 }, { "grad_norm": 0.24970926344394684, "learning_rate": 4.457245375942715e-05, "loss": 0.0315, "step": 55790 }, { "grad_norm": 0.42827868461608887, "learning_rate": 4.455601706368157e-05, "loss": 0.0302, "step": 55800 }, { "grad_norm": 0.25966396927833557, "learning_rate": 4.453958096328179e-05, "loss": 0.0291, "step": 55810 }, { "grad_norm": 0.4201928973197937, "learning_rate": 4.452314546002523e-05, "loss": 0.0387, "step": 55820 }, { "grad_norm": 0.3413456380367279, "learning_rate": 4.4506710555709246e-05, "loss": 0.0289, "step": 55830 }, { "grad_norm": 0.3085794150829315, "learning_rate": 4.449027625213114e-05, "loss": 0.0302, "step": 55840 }, { "grad_norm": 0.40566328167915344, "learning_rate": 4.4473842551088135e-05, "loss": 0.0299, "step": 55850 }, { "grad_norm": 0.49833106994628906, "learning_rate": 4.445740945437738e-05, "loss": 0.0317, "step": 55860 }, { "grad_norm": 0.3882232904434204, "learning_rate": 4.444097696379602e-05, "loss": 0.0339, "step": 55870 }, { "grad_norm": 0.4156492054462433, "learning_rate": 4.442454508114106e-05, "loss": 0.0301, "step": 55880 }, { "grad_norm": 0.6177104711532593, "learning_rate": 4.4408113808209465e-05, "loss": 0.0355, "step": 55890 }, { "grad_norm": 0.32930999994277954, "learning_rate": 4.439168314679813e-05, "loss": 0.0347, "step": 55900 }, { "grad_norm": 0.37754860520362854, "learning_rate": 4.43752530987039e-05, "loss": 0.0289, "step": 55910 }, { "grad_norm": 0.34290581941604614, "learning_rate": 4.4358823665723515e-05, "loss": 0.0341, "step": 55920 }, { "grad_norm": 0.404246985912323, "learning_rate": 4.434239484965371e-05, "loss": 0.0314, "step": 55930 }, { "grad_norm": 0.23002977669239044, "learning_rate": 4.4325966652291103e-05, "loss": 0.0315, "step": 55940 }, { "grad_norm": 0.3343439996242523, "learning_rate": 4.430953907543225e-05, "loss": 0.0331, "step": 55950 }, { "grad_norm": 0.4426741302013397, "learning_rate": 4.4293112120873645e-05, "loss": 0.0389, "step": 55960 }, { "grad_norm": 0.3882279396057129, "learning_rate": 4.427668579041172e-05, "loss": 0.0322, "step": 55970 }, { "grad_norm": 0.33352360129356384, "learning_rate": 4.4260260085842816e-05, "loss": 0.0381, "step": 55980 }, { "grad_norm": 0.38074207305908203, "learning_rate": 4.4243835008963256e-05, "loss": 0.0295, "step": 55990 }, { "grad_norm": 0.5900468230247498, "learning_rate": 4.422741056156925e-05, "loss": 0.0288, "step": 56000 }, { "grad_norm": 0.6434919834136963, "learning_rate": 4.421098674545693e-05, "loss": 0.0356, "step": 56010 }, { "grad_norm": 0.34895455837249756, "learning_rate": 4.419456356242241e-05, "loss": 0.0343, "step": 56020 }, { "grad_norm": 0.39483383297920227, "learning_rate": 4.4178141014261685e-05, "loss": 0.0311, "step": 56030 }, { "grad_norm": 0.2734115719795227, "learning_rate": 4.416171910277071e-05, "loss": 0.0311, "step": 56040 }, { "grad_norm": 0.4118254780769348, "learning_rate": 4.414529782974533e-05, "loss": 0.0319, "step": 56050 }, { "grad_norm": 0.3316885232925415, "learning_rate": 4.41288771969814e-05, "loss": 0.0252, "step": 56060 }, { "grad_norm": 0.2583087384700775, "learning_rate": 4.411245720627464e-05, "loss": 0.0294, "step": 56070 }, { "grad_norm": 0.47577106952667236, "learning_rate": 4.40960378594207e-05, "loss": 0.0291, "step": 56080 }, { "grad_norm": 0.6194763779640198, "learning_rate": 4.407961915821519e-05, "loss": 0.04, "step": 56090 }, { "grad_norm": 0.5076204538345337, "learning_rate": 4.406320110445362e-05, "loss": 0.0329, "step": 56100 }, { "grad_norm": 0.4121984839439392, "learning_rate": 4.404678369993144e-05, "loss": 0.0388, "step": 56110 }, { "grad_norm": 0.46922484040260315, "learning_rate": 4.403036694644406e-05, "loss": 0.0323, "step": 56120 }, { "grad_norm": 0.33583277463912964, "learning_rate": 4.4013950845786764e-05, "loss": 0.0296, "step": 56130 }, { "grad_norm": 0.32341140508651733, "learning_rate": 4.399753539975482e-05, "loss": 0.0301, "step": 56140 }, { "grad_norm": 0.44996967911720276, "learning_rate": 4.398112061014337e-05, "loss": 0.0267, "step": 56150 }, { "grad_norm": 0.2776208221912384, "learning_rate": 4.396470647874753e-05, "loss": 0.0318, "step": 56160 }, { "grad_norm": 0.3375483751296997, "learning_rate": 4.394829300736229e-05, "loss": 0.0259, "step": 56170 }, { "grad_norm": 0.3142717480659485, "learning_rate": 4.393188019778265e-05, "loss": 0.0273, "step": 56180 }, { "grad_norm": 0.6935877203941345, "learning_rate": 4.391546805180347e-05, "loss": 0.0302, "step": 56190 }, { "grad_norm": 0.8520881533622742, "learning_rate": 4.389905657121955e-05, "loss": 0.0382, "step": 56200 }, { "grad_norm": 0.3730109632015228, "learning_rate": 4.3882645757825635e-05, "loss": 0.0298, "step": 56210 }, { "grad_norm": 0.41171032190322876, "learning_rate": 4.386623561341637e-05, "loss": 0.0323, "step": 56220 }, { "grad_norm": 0.6873319149017334, "learning_rate": 4.384982613978637e-05, "loss": 0.0335, "step": 56230 }, { "grad_norm": 0.531511127948761, "learning_rate": 4.383341733873012e-05, "loss": 0.0311, "step": 56240 }, { "grad_norm": 0.41778072714805603, "learning_rate": 4.381700921204209e-05, "loss": 0.0291, "step": 56250 }, { "grad_norm": 0.6078941226005554, "learning_rate": 4.380060176151663e-05, "loss": 0.0363, "step": 56260 }, { "grad_norm": 0.36501866579055786, "learning_rate": 4.378419498894805e-05, "loss": 0.0279, "step": 56270 }, { "grad_norm": 0.39009925723075867, "learning_rate": 4.376778889613056e-05, "loss": 0.0344, "step": 56280 }, { "grad_norm": 0.40490201115608215, "learning_rate": 4.375138348485831e-05, "loss": 0.0291, "step": 56290 }, { "grad_norm": 0.35127532482147217, "learning_rate": 4.3734978756925346e-05, "loss": 0.0257, "step": 56300 }, { "grad_norm": 0.3201589286327362, "learning_rate": 4.3718574714125706e-05, "loss": 0.0273, "step": 56310 }, { "grad_norm": 0.3579804599285126, "learning_rate": 4.370217135825329e-05, "loss": 0.0351, "step": 56320 }, { "grad_norm": 0.2740401327610016, "learning_rate": 4.368576869110194e-05, "loss": 0.036, "step": 56330 }, { "grad_norm": 0.5321789979934692, "learning_rate": 4.366936671446544e-05, "loss": 0.0303, "step": 56340 }, { "grad_norm": 0.5549731254577637, "learning_rate": 4.3652965430137474e-05, "loss": 0.0389, "step": 56350 }, { "grad_norm": 0.2199542224407196, "learning_rate": 4.3636564839911646e-05, "loss": 0.0276, "step": 56360 }, { "grad_norm": 0.4477132558822632, "learning_rate": 4.3620164945581545e-05, "loss": 0.0345, "step": 56370 }, { "grad_norm": 0.36312583088874817, "learning_rate": 4.360376574894061e-05, "loss": 0.0355, "step": 56380 }, { "grad_norm": 0.45698150992393494, "learning_rate": 4.358736725178224e-05, "loss": 0.0351, "step": 56390 }, { "grad_norm": 0.33800312876701355, "learning_rate": 4.357096945589974e-05, "loss": 0.0345, "step": 56400 }, { "grad_norm": 0.31267049908638, "learning_rate": 4.355457236308636e-05, "loss": 0.0323, "step": 56410 }, { "grad_norm": 0.32308855652809143, "learning_rate": 4.353817597513526e-05, "loss": 0.0295, "step": 56420 }, { "grad_norm": 0.37422850728034973, "learning_rate": 4.352178029383948e-05, "loss": 0.0338, "step": 56430 }, { "grad_norm": 0.35714906454086304, "learning_rate": 4.3505385320992105e-05, "loss": 0.0318, "step": 56440 }, { "grad_norm": 0.36730489134788513, "learning_rate": 4.348899105838602e-05, "loss": 0.0336, "step": 56450 }, { "grad_norm": 0.5659957528114319, "learning_rate": 4.3472597507814087e-05, "loss": 0.0278, "step": 56460 }, { "grad_norm": 0.4330874979496002, "learning_rate": 4.3456204671069066e-05, "loss": 0.0297, "step": 56470 }, { "grad_norm": 0.3287857174873352, "learning_rate": 4.343981254994367e-05, "loss": 0.0366, "step": 56480 }, { "grad_norm": 0.35304322838783264, "learning_rate": 4.34234211462305e-05, "loss": 0.0301, "step": 56490 }, { "grad_norm": 0.4191371202468872, "learning_rate": 4.340703046172213e-05, "loss": 0.0334, "step": 56500 }, { "grad_norm": 0.44252827763557434, "learning_rate": 4.339064049821097e-05, "loss": 0.0353, "step": 56510 }, { "grad_norm": 0.3201550245285034, "learning_rate": 4.3374251257489446e-05, "loss": 0.0332, "step": 56520 }, { "grad_norm": 0.3572198748588562, "learning_rate": 4.335786274134984e-05, "loss": 0.0287, "step": 56530 }, { "grad_norm": 0.3360825181007385, "learning_rate": 4.334147495158438e-05, "loss": 0.0314, "step": 56540 }, { "grad_norm": 0.5261409878730774, "learning_rate": 4.33250878899852e-05, "loss": 0.0304, "step": 56550 }, { "grad_norm": 0.4553544521331787, "learning_rate": 4.330870155834437e-05, "loss": 0.0317, "step": 56560 }, { "grad_norm": 0.47988080978393555, "learning_rate": 4.3292315958453884e-05, "loss": 0.0285, "step": 56570 }, { "grad_norm": 0.38393810391426086, "learning_rate": 4.3275931092105646e-05, "loss": 0.0298, "step": 56580 }, { "grad_norm": 0.3726871609687805, "learning_rate": 4.325954696109147e-05, "loss": 0.0369, "step": 56590 }, { "grad_norm": 0.3258435130119324, "learning_rate": 4.32431635672031e-05, "loss": 0.0307, "step": 56600 }, { "grad_norm": 0.3534228801727295, "learning_rate": 4.3226780912232214e-05, "loss": 0.0364, "step": 56610 }, { "grad_norm": 0.35091203451156616, "learning_rate": 4.321039899797036e-05, "loss": 0.0321, "step": 56620 }, { "grad_norm": 0.34580671787261963, "learning_rate": 4.319401782620908e-05, "loss": 0.0329, "step": 56630 }, { "grad_norm": 0.4607798457145691, "learning_rate": 4.317763739873978e-05, "loss": 0.0326, "step": 56640 }, { "grad_norm": 0.4114476442337036, "learning_rate": 4.31612577173538e-05, "loss": 0.0307, "step": 56650 }, { "grad_norm": 0.30490976572036743, "learning_rate": 4.314487878384239e-05, "loss": 0.0341, "step": 56660 }, { "grad_norm": 0.41484588384628296, "learning_rate": 4.3128500599996736e-05, "loss": 0.0288, "step": 56670 }, { "grad_norm": 0.3603796064853668, "learning_rate": 4.31121231676079e-05, "loss": 0.0291, "step": 56680 }, { "grad_norm": 0.5262439250946045, "learning_rate": 4.309574648846694e-05, "loss": 0.0308, "step": 56690 }, { "grad_norm": 0.3152260482311249, "learning_rate": 4.3079370564364755e-05, "loss": 0.0299, "step": 56700 }, { "grad_norm": 0.7637972235679626, "learning_rate": 4.306299539709221e-05, "loss": 0.0274, "step": 56710 }, { "grad_norm": 0.4530671238899231, "learning_rate": 4.3046620988440046e-05, "loss": 0.0336, "step": 56720 }, { "grad_norm": 0.8766067624092102, "learning_rate": 4.303024734019897e-05, "loss": 0.032, "step": 56730 }, { "grad_norm": 0.3422817587852478, "learning_rate": 4.3013874454159565e-05, "loss": 0.0277, "step": 56740 }, { "grad_norm": 0.25760617852211, "learning_rate": 4.299750233211233e-05, "loss": 0.0331, "step": 56750 }, { "grad_norm": 0.29229751229286194, "learning_rate": 4.2981130975847715e-05, "loss": 0.0269, "step": 56760 }, { "grad_norm": 0.4053787291049957, "learning_rate": 4.2964760387156076e-05, "loss": 0.0349, "step": 56770 }, { "grad_norm": 0.5267964005470276, "learning_rate": 4.294839056782767e-05, "loss": 0.0331, "step": 56780 }, { "grad_norm": 0.45648735761642456, "learning_rate": 4.293202151965266e-05, "loss": 0.0319, "step": 56790 }, { "grad_norm": 0.41324710845947266, "learning_rate": 4.291565324442115e-05, "loss": 0.0315, "step": 56800 }, { "grad_norm": 0.38451918959617615, "learning_rate": 4.2899285743923135e-05, "loss": 0.0399, "step": 56810 }, { "grad_norm": 0.41683992743492126, "learning_rate": 4.288291901994857e-05, "loss": 0.0297, "step": 56820 }, { "grad_norm": 0.48182690143585205, "learning_rate": 4.286655307428729e-05, "loss": 0.0343, "step": 56830 }, { "grad_norm": 0.4996727705001831, "learning_rate": 4.285018790872903e-05, "loss": 0.0333, "step": 56840 }, { "grad_norm": 0.4438294470310211, "learning_rate": 4.283382352506347e-05, "loss": 0.0307, "step": 56850 }, { "grad_norm": 1.0118972063064575, "learning_rate": 4.2817459925080205e-05, "loss": 0.0353, "step": 56860 }, { "grad_norm": 0.4217279553413391, "learning_rate": 4.28010971105687e-05, "loss": 0.03, "step": 56870 }, { "grad_norm": 0.3343019485473633, "learning_rate": 4.278473508331841e-05, "loss": 0.0275, "step": 56880 }, { "grad_norm": 0.43809518218040466, "learning_rate": 4.276837384511864e-05, "loss": 0.0339, "step": 56890 }, { "grad_norm": 0.37736278772354126, "learning_rate": 4.275201339775864e-05, "loss": 0.0295, "step": 56900 }, { "grad_norm": 0.317345529794693, "learning_rate": 4.273565374302756e-05, "loss": 0.0303, "step": 56910 }, { "grad_norm": 0.4197971224784851, "learning_rate": 4.2719294882714454e-05, "loss": 0.0309, "step": 56920 }, { "grad_norm": 0.43169039487838745, "learning_rate": 4.270293681860831e-05, "loss": 0.0321, "step": 56930 }, { "grad_norm": 0.37876859307289124, "learning_rate": 4.268657955249802e-05, "loss": 0.0336, "step": 56940 }, { "grad_norm": 0.8674798011779785, "learning_rate": 4.267022308617241e-05, "loss": 0.0297, "step": 56950 }, { "grad_norm": 0.292447030544281, "learning_rate": 4.265386742142018e-05, "loss": 0.0311, "step": 56960 }, { "grad_norm": 0.3990921676158905, "learning_rate": 4.263751256002996e-05, "loss": 0.029, "step": 56970 }, { "grad_norm": 0.33593952655792236, "learning_rate": 4.26211585037903e-05, "loss": 0.0314, "step": 56980 }, { "grad_norm": 0.3079148828983307, "learning_rate": 4.2604805254489656e-05, "loss": 0.0301, "step": 56990 }, { "grad_norm": 0.36301034688949585, "learning_rate": 4.258845281391638e-05, "loss": 0.0335, "step": 57000 }, { "grad_norm": 0.23810335993766785, "learning_rate": 4.257210118385876e-05, "loss": 0.0252, "step": 57010 }, { "grad_norm": 0.40333807468414307, "learning_rate": 4.2555750366104994e-05, "loss": 0.0279, "step": 57020 }, { "grad_norm": 0.49318090081214905, "learning_rate": 4.2539400362443184e-05, "loss": 0.0293, "step": 57030 }, { "grad_norm": 0.29835185408592224, "learning_rate": 4.252305117466133e-05, "loss": 0.0309, "step": 57040 }, { "grad_norm": 0.32049059867858887, "learning_rate": 4.2506702804547356e-05, "loss": 0.0267, "step": 57050 }, { "grad_norm": 0.43508949875831604, "learning_rate": 4.249035525388907e-05, "loss": 0.0362, "step": 57060 }, { "grad_norm": 0.4839094579219818, "learning_rate": 4.247400852447427e-05, "loss": 0.0331, "step": 57070 }, { "grad_norm": 0.45261117815971375, "learning_rate": 4.245766261809059e-05, "loss": 0.03, "step": 57080 }, { "grad_norm": 0.24275735020637512, "learning_rate": 4.244131753652558e-05, "loss": 0.0338, "step": 57090 }, { "grad_norm": 0.4299893081188202, "learning_rate": 4.242497328156672e-05, "loss": 0.0295, "step": 57100 }, { "grad_norm": 0.4000517427921295, "learning_rate": 4.240862985500139e-05, "loss": 0.0312, "step": 57110 }, { "grad_norm": 0.39483925700187683, "learning_rate": 4.2392287258616894e-05, "loss": 0.0311, "step": 57120 }, { "grad_norm": 0.3465203642845154, "learning_rate": 4.2375945494200396e-05, "loss": 0.0267, "step": 57130 }, { "grad_norm": 0.3589111864566803, "learning_rate": 4.2359604563539054e-05, "loss": 0.033, "step": 57140 }, { "grad_norm": 0.26694318652153015, "learning_rate": 4.2343264468419885e-05, "loss": 0.0326, "step": 57150 }, { "grad_norm": 0.39593955874443054, "learning_rate": 4.232692521062979e-05, "loss": 0.0293, "step": 57160 }, { "grad_norm": 0.4112774431705475, "learning_rate": 4.231058679195562e-05, "loss": 0.031, "step": 57170 }, { "grad_norm": 0.40555864572525024, "learning_rate": 4.229424921418411e-05, "loss": 0.0317, "step": 57180 }, { "grad_norm": 0.3686043620109558, "learning_rate": 4.2277912479101916e-05, "loss": 0.0261, "step": 57190 }, { "grad_norm": 0.4692055284976959, "learning_rate": 4.226157658849562e-05, "loss": 0.0362, "step": 57200 }, { "grad_norm": 0.32973697781562805, "learning_rate": 4.2245241544151664e-05, "loss": 0.0293, "step": 57210 }, { "grad_norm": 0.4522504508495331, "learning_rate": 4.2228907347856425e-05, "loss": 0.0304, "step": 57220 }, { "grad_norm": 0.4791714549064636, "learning_rate": 4.2212574001396214e-05, "loss": 0.0347, "step": 57230 }, { "grad_norm": 0.2547803521156311, "learning_rate": 4.21962415065572e-05, "loss": 0.0299, "step": 57240 }, { "grad_norm": 0.2691260278224945, "learning_rate": 4.217990986512548e-05, "loss": 0.0355, "step": 57250 }, { "grad_norm": 0.5806664228439331, "learning_rate": 4.216357907888704e-05, "loss": 0.0315, "step": 57260 }, { "grad_norm": 0.36477169394493103, "learning_rate": 4.2147249149627824e-05, "loss": 0.0247, "step": 57270 }, { "grad_norm": 0.43547675013542175, "learning_rate": 4.213092007913364e-05, "loss": 0.0325, "step": 57280 }, { "grad_norm": 0.3964310884475708, "learning_rate": 4.2114591869190205e-05, "loss": 0.0381, "step": 57290 }, { "grad_norm": 0.40013158321380615, "learning_rate": 4.209826452158315e-05, "loss": 0.0309, "step": 57300 }, { "grad_norm": 0.36048346757888794, "learning_rate": 4.2081938038098e-05, "loss": 0.0259, "step": 57310 }, { "grad_norm": 0.35759881138801575, "learning_rate": 4.2065612420520185e-05, "loss": 0.0284, "step": 57320 }, { "grad_norm": 0.7055469751358032, "learning_rate": 4.204928767063509e-05, "loss": 0.0301, "step": 57330 }, { "grad_norm": 0.2785334885120392, "learning_rate": 4.2032963790227944e-05, "loss": 0.0293, "step": 57340 }, { "grad_norm": 0.28710219264030457, "learning_rate": 4.2016640781083887e-05, "loss": 0.0345, "step": 57350 }, { "grad_norm": 0.4926055669784546, "learning_rate": 4.2000318644988004e-05, "loss": 0.0319, "step": 57360 }, { "grad_norm": 0.3207772672176361, "learning_rate": 4.198399738372524e-05, "loss": 0.0288, "step": 57370 }, { "grad_norm": 0.3537520468235016, "learning_rate": 4.196767699908045e-05, "loss": 0.0326, "step": 57380 }, { "grad_norm": 0.259068101644516, "learning_rate": 4.195135749283845e-05, "loss": 0.033, "step": 57390 }, { "grad_norm": 0.3158872127532959, "learning_rate": 4.1935038866783894e-05, "loss": 0.0301, "step": 57400 }, { "grad_norm": 0.6179509162902832, "learning_rate": 4.1918721122701355e-05, "loss": 0.0319, "step": 57410 }, { "grad_norm": 0.3133501410484314, "learning_rate": 4.190240426237533e-05, "loss": 0.0375, "step": 57420 }, { "grad_norm": 0.3550359606742859, "learning_rate": 4.188608828759019e-05, "loss": 0.0301, "step": 57430 }, { "grad_norm": 0.3886452317237854, "learning_rate": 4.1869773200130246e-05, "loss": 0.0317, "step": 57440 }, { "grad_norm": 0.6312093734741211, "learning_rate": 4.185345900177967e-05, "loss": 0.0327, "step": 57450 }, { "grad_norm": 1.745237112045288, "learning_rate": 4.183714569432258e-05, "loss": 0.0305, "step": 57460 }, { "grad_norm": 1.364004373550415, "learning_rate": 4.182083327954295e-05, "loss": 0.0313, "step": 57470 }, { "grad_norm": 0.2277461737394333, "learning_rate": 4.1804521759224714e-05, "loss": 0.0281, "step": 57480 }, { "grad_norm": 0.3229061961174011, "learning_rate": 4.178821113515165e-05, "loss": 0.0314, "step": 57490 }, { "grad_norm": 0.38729238510131836, "learning_rate": 4.1771901409107476e-05, "loss": 0.0295, "step": 57500 }, { "grad_norm": 0.5259706377983093, "learning_rate": 4.175559258287577e-05, "loss": 0.0285, "step": 57510 }, { "grad_norm": 0.347397118806839, "learning_rate": 4.173928465824008e-05, "loss": 0.0287, "step": 57520 }, { "grad_norm": 0.42474034428596497, "learning_rate": 4.17229776369838e-05, "loss": 0.0272, "step": 57530 }, { "grad_norm": 0.4059854745864868, "learning_rate": 4.170667152089025e-05, "loss": 0.0305, "step": 57540 }, { "grad_norm": 0.7550467252731323, "learning_rate": 4.1690366311742626e-05, "loss": 0.0307, "step": 57550 }, { "grad_norm": 0.435829758644104, "learning_rate": 4.1674062011324055e-05, "loss": 0.0379, "step": 57560 }, { "grad_norm": 0.2895828187465668, "learning_rate": 4.165775862141752e-05, "loss": 0.0309, "step": 57570 }, { "grad_norm": 0.30826833844184875, "learning_rate": 4.164145614380598e-05, "loss": 0.0321, "step": 57580 }, { "grad_norm": 0.3415578305721283, "learning_rate": 4.162515458027223e-05, "loss": 0.0271, "step": 57590 }, { "grad_norm": 0.38544100522994995, "learning_rate": 4.160885393259899e-05, "loss": 0.0318, "step": 57600 }, { "grad_norm": 0.27487751841545105, "learning_rate": 4.159255420256886e-05, "loss": 0.031, "step": 57610 }, { "grad_norm": 0.41830047965049744, "learning_rate": 4.157625539196436e-05, "loss": 0.0296, "step": 57620 }, { "grad_norm": 0.2739938497543335, "learning_rate": 4.155995750256791e-05, "loss": 0.0303, "step": 57630 }, { "grad_norm": 0.2703005373477936, "learning_rate": 4.15436605361618e-05, "loss": 0.0311, "step": 57640 }, { "grad_norm": 0.3007141649723053, "learning_rate": 4.152736449452827e-05, "loss": 0.0275, "step": 57650 }, { "grad_norm": 0.3163016140460968, "learning_rate": 4.151106937944943e-05, "loss": 0.0321, "step": 57660 }, { "grad_norm": 0.3094744086265564, "learning_rate": 4.149477519270727e-05, "loss": 0.0296, "step": 57670 }, { "grad_norm": 0.28370797634124756, "learning_rate": 4.147848193608371e-05, "loss": 0.0262, "step": 57680 }, { "grad_norm": 0.3809349238872528, "learning_rate": 4.146218961136056e-05, "loss": 0.0294, "step": 57690 }, { "grad_norm": 0.3507677912712097, "learning_rate": 4.14458982203195e-05, "loss": 0.0366, "step": 57700 }, { "grad_norm": 0.39480099081993103, "learning_rate": 4.1429607764742165e-05, "loss": 0.0312, "step": 57710 }, { "grad_norm": 0.3913031220436096, "learning_rate": 4.141331824641003e-05, "loss": 0.0334, "step": 57720 }, { "grad_norm": 0.325411856174469, "learning_rate": 4.139702966710452e-05, "loss": 0.0299, "step": 57730 }, { "grad_norm": 0.26418086886405945, "learning_rate": 4.13807420286069e-05, "loss": 0.0273, "step": 57740 }, { "grad_norm": 0.6624311208724976, "learning_rate": 4.136445533269837e-05, "loss": 0.0338, "step": 57750 }, { "grad_norm": 0.5680555701255798, "learning_rate": 4.134816958116003e-05, "loss": 0.0306, "step": 57760 }, { "grad_norm": 0.4792761206626892, "learning_rate": 4.133188477577283e-05, "loss": 0.0281, "step": 57770 }, { "grad_norm": 0.5559018850326538, "learning_rate": 4.131560091831771e-05, "loss": 0.0309, "step": 57780 }, { "grad_norm": 0.26191917061805725, "learning_rate": 4.1299318010575405e-05, "loss": 0.0338, "step": 57790 }, { "grad_norm": 0.2877444326877594, "learning_rate": 4.128303605432661e-05, "loss": 0.0273, "step": 57800 }, { "grad_norm": 0.2290182262659073, "learning_rate": 4.126675505135188e-05, "loss": 0.0306, "step": 57810 }, { "grad_norm": 0.4775741696357727, "learning_rate": 4.125047500343169e-05, "loss": 0.0315, "step": 57820 }, { "grad_norm": 0.4358971416950226, "learning_rate": 4.1234195912346376e-05, "loss": 0.0296, "step": 57830 }, { "grad_norm": 0.2602914273738861, "learning_rate": 4.1217917779876235e-05, "loss": 0.0285, "step": 57840 }, { "grad_norm": 0.42130985856056213, "learning_rate": 4.120164060780141e-05, "loss": 0.0344, "step": 57850 }, { "grad_norm": 0.3085819184780121, "learning_rate": 4.1185364397901935e-05, "loss": 0.0319, "step": 57860 }, { "grad_norm": 0.3418174088001251, "learning_rate": 4.116908915195776e-05, "loss": 0.034, "step": 57870 }, { "grad_norm": 0.27908727526664734, "learning_rate": 4.115281487174872e-05, "loss": 0.0285, "step": 57880 }, { "grad_norm": 0.46234989166259766, "learning_rate": 4.113654155905452e-05, "loss": 0.0289, "step": 57890 }, { "grad_norm": 0.2709851861000061, "learning_rate": 4.112026921565483e-05, "loss": 0.0323, "step": 57900 }, { "grad_norm": 0.8186090588569641, "learning_rate": 4.1103997843329147e-05, "loss": 0.0353, "step": 57910 }, { "grad_norm": 0.4623488187789917, "learning_rate": 4.108772744385689e-05, "loss": 0.0297, "step": 57920 }, { "grad_norm": 0.36491379141807556, "learning_rate": 4.1071458019017346e-05, "loss": 0.0302, "step": 57930 }, { "grad_norm": 0.24399207532405853, "learning_rate": 4.1055189570589734e-05, "loss": 0.0257, "step": 57940 }, { "grad_norm": 0.4586707055568695, "learning_rate": 4.103892210035315e-05, "loss": 0.0347, "step": 57950 }, { "grad_norm": 0.2368869185447693, "learning_rate": 4.102265561008656e-05, "loss": 0.0326, "step": 57960 }, { "grad_norm": 0.41559964418411255, "learning_rate": 4.100639010156885e-05, "loss": 0.0295, "step": 57970 }, { "grad_norm": 0.25581005215644836, "learning_rate": 4.099012557657882e-05, "loss": 0.0295, "step": 57980 }, { "grad_norm": 0.4257027804851532, "learning_rate": 4.09738620368951e-05, "loss": 0.0333, "step": 57990 }, { "grad_norm": 0.38261571526527405, "learning_rate": 4.095759948429626e-05, "loss": 0.034, "step": 58000 }, { "grad_norm": 0.46090152859687805, "learning_rate": 4.094133792056075e-05, "loss": 0.03, "step": 58010 }, { "grad_norm": 0.3806944489479065, "learning_rate": 4.092507734746688e-05, "loss": 0.0324, "step": 58020 }, { "grad_norm": 0.2294028401374817, "learning_rate": 4.090881776679293e-05, "loss": 0.0302, "step": 58030 }, { "grad_norm": 0.47681209444999695, "learning_rate": 4.0892559180316994e-05, "loss": 0.0306, "step": 58040 }, { "grad_norm": 0.6270396113395691, "learning_rate": 4.08763015898171e-05, "loss": 0.033, "step": 58050 }, { "grad_norm": 0.6785837411880493, "learning_rate": 4.0860044997071145e-05, "loss": 0.0333, "step": 58060 }, { "grad_norm": 0.3516695499420166, "learning_rate": 4.084378940385693e-05, "loss": 0.0295, "step": 58070 }, { "grad_norm": 0.4310651123523712, "learning_rate": 4.082753481195211e-05, "loss": 0.0322, "step": 58080 }, { "grad_norm": 0.2834429442882538, "learning_rate": 4.0811281223134324e-05, "loss": 0.0243, "step": 58090 }, { "grad_norm": 0.36726540327072144, "learning_rate": 4.0795028639181e-05, "loss": 0.0295, "step": 58100 }, { "grad_norm": 0.4325776696205139, "learning_rate": 4.077877706186951e-05, "loss": 0.0306, "step": 58110 }, { "grad_norm": 0.8383517265319824, "learning_rate": 4.076252649297709e-05, "loss": 0.0254, "step": 58120 }, { "grad_norm": 0.34871771931648254, "learning_rate": 4.07462769342809e-05, "loss": 0.0329, "step": 58130 }, { "grad_norm": 0.4026888906955719, "learning_rate": 4.073002838755793e-05, "loss": 0.0296, "step": 58140 }, { "grad_norm": 0.42750728130340576, "learning_rate": 4.0713780854585115e-05, "loss": 0.0317, "step": 58150 }, { "grad_norm": 0.4080483615398407, "learning_rate": 4.069753433713927e-05, "loss": 0.0315, "step": 58160 }, { "grad_norm": 0.4468386471271515, "learning_rate": 4.068128883699709e-05, "loss": 0.0262, "step": 58170 }, { "grad_norm": 0.5338056087493896, "learning_rate": 4.0665044355935155e-05, "loss": 0.0306, "step": 58180 }, { "grad_norm": 0.4390696585178375, "learning_rate": 4.064880089572993e-05, "loss": 0.0315, "step": 58190 }, { "grad_norm": 0.3368895947933197, "learning_rate": 4.063255845815779e-05, "loss": 0.0344, "step": 58200 }, { "grad_norm": 0.4317586421966553, "learning_rate": 4.0616317044994945e-05, "loss": 0.0284, "step": 58210 }, { "grad_norm": 0.5247766971588135, "learning_rate": 4.0600076658017585e-05, "loss": 0.0319, "step": 58220 }, { "grad_norm": 0.29930737614631653, "learning_rate": 4.0583837299001706e-05, "loss": 0.0284, "step": 58230 }, { "grad_norm": 0.355911523103714, "learning_rate": 4.056759896972322e-05, "loss": 0.0313, "step": 58240 }, { "grad_norm": 0.43919187784194946, "learning_rate": 4.055136167195793e-05, "loss": 0.036, "step": 58250 }, { "grad_norm": 0.3471791744232178, "learning_rate": 4.053512540748152e-05, "loss": 0.0331, "step": 58260 }, { "grad_norm": 0.5470978021621704, "learning_rate": 4.051889017806955e-05, "loss": 0.0349, "step": 58270 }, { "grad_norm": 0.2660946547985077, "learning_rate": 4.0502655985497516e-05, "loss": 0.0316, "step": 58280 }, { "grad_norm": 0.43874385952949524, "learning_rate": 4.048642283154074e-05, "loss": 0.0319, "step": 58290 }, { "grad_norm": 0.3685198128223419, "learning_rate": 4.0470190717974464e-05, "loss": 0.0319, "step": 58300 }, { "grad_norm": 0.6036027669906616, "learning_rate": 4.045395964657381e-05, "loss": 0.0317, "step": 58310 }, { "grad_norm": 0.3824273943901062, "learning_rate": 4.043772961911376e-05, "loss": 0.0321, "step": 58320 }, { "grad_norm": 0.44942474365234375, "learning_rate": 4.042150063736924e-05, "loss": 0.0318, "step": 58330 }, { "grad_norm": 0.31256771087646484, "learning_rate": 4.0405272703114976e-05, "loss": 0.0308, "step": 58340 }, { "grad_norm": 0.4334716498851776, "learning_rate": 4.038904581812569e-05, "loss": 0.0307, "step": 58350 }, { "grad_norm": 0.256618857383728, "learning_rate": 4.03728199841759e-05, "loss": 0.029, "step": 58360 }, { "grad_norm": 0.3367430567741394, "learning_rate": 4.0356595203040044e-05, "loss": 0.0296, "step": 58370 }, { "grad_norm": 0.2599734663963318, "learning_rate": 4.0340371476492426e-05, "loss": 0.0342, "step": 58380 }, { "grad_norm": 0.6094357967376709, "learning_rate": 4.032414880630727e-05, "loss": 0.032, "step": 58390 }, { "grad_norm": 0.33900439739227295, "learning_rate": 4.0307927194258646e-05, "loss": 0.0259, "step": 58400 }, { "grad_norm": 0.3270275294780731, "learning_rate": 4.029170664212054e-05, "loss": 0.0266, "step": 58410 }, { "grad_norm": 0.36714059114456177, "learning_rate": 4.0275487151666794e-05, "loss": 0.0273, "step": 58420 }, { "grad_norm": 0.4032808244228363, "learning_rate": 4.025926872467116e-05, "loss": 0.0342, "step": 58430 }, { "grad_norm": 0.2896570861339569, "learning_rate": 4.024305136290725e-05, "loss": 0.0246, "step": 58440 }, { "grad_norm": 0.35318803787231445, "learning_rate": 4.022683506814858e-05, "loss": 0.0291, "step": 58450 }, { "grad_norm": 0.34758439660072327, "learning_rate": 4.021061984216853e-05, "loss": 0.0348, "step": 58460 }, { "grad_norm": 0.3564889430999756, "learning_rate": 4.019440568674035e-05, "loss": 0.0329, "step": 58470 }, { "grad_norm": 0.32688596844673157, "learning_rate": 4.017819260363724e-05, "loss": 0.0357, "step": 58480 }, { "grad_norm": 0.45704224705696106, "learning_rate": 4.016198059463223e-05, "loss": 0.0355, "step": 58490 }, { "grad_norm": 0.4753507375717163, "learning_rate": 4.0145769661498233e-05, "loss": 0.028, "step": 58500 }, { "grad_norm": 0.34036797285079956, "learning_rate": 4.012955980600804e-05, "loss": 0.0286, "step": 58510 }, { "grad_norm": 0.3262937664985657, "learning_rate": 4.0113351029934355e-05, "loss": 0.0327, "step": 58520 }, { "grad_norm": 0.44943249225616455, "learning_rate": 4.0097143335049714e-05, "loss": 0.0381, "step": 58530 }, { "grad_norm": 0.29129427671432495, "learning_rate": 4.0080936723126606e-05, "loss": 0.0296, "step": 58540 }, { "grad_norm": 0.6226974725723267, "learning_rate": 4.006473119593735e-05, "loss": 0.0306, "step": 58550 }, { "grad_norm": 0.318461537361145, "learning_rate": 4.004852675525415e-05, "loss": 0.0314, "step": 58560 }, { "grad_norm": 0.44882482290267944, "learning_rate": 4.00323234028491e-05, "loss": 0.0301, "step": 58570 }, { "grad_norm": 0.5455401539802551, "learning_rate": 4.0016121140494186e-05, "loss": 0.0297, "step": 58580 }, { "grad_norm": 0.3913288712501526, "learning_rate": 3.999991996996122e-05, "loss": 0.0297, "step": 58590 }, { "grad_norm": 0.3396722972393036, "learning_rate": 3.9983719893022e-05, "loss": 0.0343, "step": 58600 }, { "grad_norm": 0.3662760257720947, "learning_rate": 3.996752091144812e-05, "loss": 0.036, "step": 58610 }, { "grad_norm": 0.44720712304115295, "learning_rate": 3.9951323027011055e-05, "loss": 0.0317, "step": 58620 }, { "grad_norm": 0.4160776436328888, "learning_rate": 3.993512624148219e-05, "loss": 0.0301, "step": 58630 }, { "grad_norm": 0.7588050961494446, "learning_rate": 3.99189305566328e-05, "loss": 0.0279, "step": 58640 }, { "grad_norm": 0.49021491408348083, "learning_rate": 3.990273597423401e-05, "loss": 0.0333, "step": 58650 }, { "grad_norm": 0.27816471457481384, "learning_rate": 3.988654249605681e-05, "loss": 0.0262, "step": 58660 }, { "grad_norm": 0.3765210807323456, "learning_rate": 3.987035012387213e-05, "loss": 0.0296, "step": 58670 }, { "grad_norm": 0.6632745265960693, "learning_rate": 3.9854158859450735e-05, "loss": 0.03, "step": 58680 }, { "grad_norm": 0.30981355905532837, "learning_rate": 3.983796870456328e-05, "loss": 0.0308, "step": 58690 }, { "grad_norm": 0.29300814867019653, "learning_rate": 3.982177966098028e-05, "loss": 0.0288, "step": 58700 }, { "grad_norm": 0.7165430188179016, "learning_rate": 3.980559173047216e-05, "loss": 0.0318, "step": 58710 }, { "grad_norm": 0.38775286078453064, "learning_rate": 3.9789404914809176e-05, "loss": 0.0312, "step": 58720 }, { "grad_norm": 0.2867262661457062, "learning_rate": 3.9773219215761544e-05, "loss": 0.0287, "step": 58730 }, { "grad_norm": 0.32229045033454895, "learning_rate": 3.975703463509928e-05, "loss": 0.0261, "step": 58740 }, { "grad_norm": 0.3262832462787628, "learning_rate": 3.974085117459231e-05, "loss": 0.0274, "step": 58750 }, { "grad_norm": 0.32865849137306213, "learning_rate": 3.972466883601043e-05, "loss": 0.0316, "step": 58760 }, { "grad_norm": 0.3781434893608093, "learning_rate": 3.9708487621123316e-05, "loss": 0.0335, "step": 58770 }, { "grad_norm": 0.3359447419643402, "learning_rate": 3.969230753170051e-05, "loss": 0.0285, "step": 58780 }, { "grad_norm": 0.38060876727104187, "learning_rate": 3.967612856951146e-05, "loss": 0.0287, "step": 58790 }, { "grad_norm": 0.3994459807872772, "learning_rate": 3.9659950736325476e-05, "loss": 0.029, "step": 58800 }, { "grad_norm": 0.31263554096221924, "learning_rate": 3.964377403391174e-05, "loss": 0.0317, "step": 58810 }, { "grad_norm": 0.5046395659446716, "learning_rate": 3.9627598464039303e-05, "loss": 0.0303, "step": 58820 }, { "grad_norm": 0.3629932403564453, "learning_rate": 3.9611424028477096e-05, "loss": 0.0352, "step": 58830 }, { "grad_norm": 0.3820132315158844, "learning_rate": 3.959525072899394e-05, "loss": 0.0305, "step": 58840 }, { "grad_norm": 0.3078172504901886, "learning_rate": 3.9579078567358506e-05, "loss": 0.0247, "step": 58850 }, { "grad_norm": 0.6423926949501038, "learning_rate": 3.956290754533939e-05, "loss": 0.0346, "step": 58860 }, { "grad_norm": 0.37233537435531616, "learning_rate": 3.954673766470501e-05, "loss": 0.0315, "step": 58870 }, { "grad_norm": 0.45460060238838196, "learning_rate": 3.9530568927223674e-05, "loss": 0.0311, "step": 58880 }, { "grad_norm": 0.37278997898101807, "learning_rate": 3.95144013346636e-05, "loss": 0.0304, "step": 58890 }, { "grad_norm": 0.2888754904270172, "learning_rate": 3.949823488879282e-05, "loss": 0.0276, "step": 58900 }, { "grad_norm": 0.2987070679664612, "learning_rate": 3.9482069591379276e-05, "loss": 0.0297, "step": 58910 }, { "grad_norm": 0.5347754955291748, "learning_rate": 3.9465905444190784e-05, "loss": 0.0293, "step": 58920 }, { "grad_norm": 0.3458676338195801, "learning_rate": 3.944974244899505e-05, "loss": 0.0275, "step": 58930 }, { "grad_norm": 0.492012619972229, "learning_rate": 3.9433580607559616e-05, "loss": 0.0308, "step": 58940 }, { "grad_norm": 0.3379075229167938, "learning_rate": 3.941741992165193e-05, "loss": 0.0307, "step": 58950 }, { "grad_norm": 0.5595303773880005, "learning_rate": 3.940126039303928e-05, "loss": 0.0324, "step": 58960 }, { "grad_norm": 0.4254045784473419, "learning_rate": 3.938510202348886e-05, "loss": 0.0274, "step": 58970 }, { "grad_norm": 0.2910240590572357, "learning_rate": 3.93689448147677e-05, "loss": 0.0245, "step": 58980 }, { "grad_norm": 0.3376959264278412, "learning_rate": 3.935278876864278e-05, "loss": 0.0295, "step": 58990 }, { "grad_norm": 0.3239898681640625, "learning_rate": 3.933663388688087e-05, "loss": 0.0301, "step": 59000 }, { "grad_norm": 0.3524060547351837, "learning_rate": 3.9320480171248645e-05, "loss": 0.0345, "step": 59010 }, { "grad_norm": 0.5345194935798645, "learning_rate": 3.930432762351265e-05, "loss": 0.0308, "step": 59020 }, { "grad_norm": 0.32214272022247314, "learning_rate": 3.928817624543931e-05, "loss": 0.027, "step": 59030 }, { "grad_norm": 0.32439011335372925, "learning_rate": 3.9272026038794885e-05, "loss": 0.0299, "step": 59040 }, { "grad_norm": 0.4736301898956299, "learning_rate": 3.9255877005345586e-05, "loss": 0.0311, "step": 59050 }, { "grad_norm": 0.4856892228126526, "learning_rate": 3.923972914685743e-05, "loss": 0.0287, "step": 59060 }, { "grad_norm": 0.2987615466117859, "learning_rate": 3.922358246509631e-05, "loss": 0.0332, "step": 59070 }, { "grad_norm": 0.39541932940483093, "learning_rate": 3.920743696182801e-05, "loss": 0.0333, "step": 59080 }, { "grad_norm": 0.3041103184223175, "learning_rate": 3.9191292638818164e-05, "loss": 0.0269, "step": 59090 }, { "grad_norm": 0.2907423973083496, "learning_rate": 3.9175149497832316e-05, "loss": 0.026, "step": 59100 }, { "grad_norm": 0.37776973843574524, "learning_rate": 3.915900754063584e-05, "loss": 0.0295, "step": 59110 }, { "grad_norm": 0.29906216263771057, "learning_rate": 3.9142866768994014e-05, "loss": 0.0301, "step": 59120 }, { "grad_norm": 0.34205764532089233, "learning_rate": 3.912672718467193e-05, "loss": 0.0276, "step": 59130 }, { "grad_norm": 0.5212987661361694, "learning_rate": 3.911058878943463e-05, "loss": 0.0305, "step": 59140 }, { "grad_norm": 0.32632750272750854, "learning_rate": 3.9094451585046974e-05, "loss": 0.0357, "step": 59150 }, { "grad_norm": 0.41126495599746704, "learning_rate": 3.90783155732737e-05, "loss": 0.0318, "step": 59160 }, { "grad_norm": 0.33450648188591003, "learning_rate": 3.9062180755879373e-05, "loss": 0.027, "step": 59170 }, { "grad_norm": 0.47439807653427124, "learning_rate": 3.904604713462855e-05, "loss": 0.0276, "step": 59180 }, { "grad_norm": 0.43944627046585083, "learning_rate": 3.902991471128554e-05, "loss": 0.0363, "step": 59190 }, { "grad_norm": 0.48195698857307434, "learning_rate": 3.901378348761456e-05, "loss": 0.0293, "step": 59200 }, { "grad_norm": 0.34301772713661194, "learning_rate": 3.899765346537971e-05, "loss": 0.0303, "step": 59210 }, { "grad_norm": 0.38463476300239563, "learning_rate": 3.898152464634493e-05, "loss": 0.0372, "step": 59220 }, { "grad_norm": 0.342342346906662, "learning_rate": 3.896539703227402e-05, "loss": 0.0379, "step": 59230 }, { "grad_norm": 0.37659743428230286, "learning_rate": 3.894927062493073e-05, "loss": 0.034, "step": 59240 }, { "grad_norm": 0.4030200242996216, "learning_rate": 3.893314542607859e-05, "loss": 0.0259, "step": 59250 }, { "grad_norm": 0.4588591158390045, "learning_rate": 3.891702143748103e-05, "loss": 0.0298, "step": 59260 }, { "grad_norm": 0.3974328935146332, "learning_rate": 3.890089866090133e-05, "loss": 0.0305, "step": 59270 }, { "grad_norm": 0.43610256910324097, "learning_rate": 3.888477709810267e-05, "loss": 0.0312, "step": 59280 }, { "grad_norm": 0.3140447437763214, "learning_rate": 3.886865675084806e-05, "loss": 0.0323, "step": 59290 }, { "grad_norm": 0.7334879040718079, "learning_rate": 3.885253762090043e-05, "loss": 0.0289, "step": 59300 }, { "grad_norm": 0.39130261540412903, "learning_rate": 3.8836419710022534e-05, "loss": 0.0263, "step": 59310 }, { "grad_norm": 0.5092320442199707, "learning_rate": 3.882030301997698e-05, "loss": 0.0333, "step": 59320 }, { "grad_norm": 0.3596719801425934, "learning_rate": 3.880418755252629e-05, "loss": 0.0276, "step": 59330 }, { "grad_norm": 0.3092108964920044, "learning_rate": 3.87880733094328e-05, "loss": 0.0262, "step": 59340 }, { "grad_norm": 0.39712509512901306, "learning_rate": 3.877196029245877e-05, "loss": 0.0251, "step": 59350 }, { "grad_norm": 0.3504175841808319, "learning_rate": 3.875584850336627e-05, "loss": 0.0259, "step": 59360 }, { "grad_norm": 0.4020783603191376, "learning_rate": 3.873973794391728e-05, "loss": 0.0281, "step": 59370 }, { "grad_norm": 0.25891217589378357, "learning_rate": 3.872362861587361e-05, "loss": 0.0261, "step": 59380 }, { "grad_norm": 0.3644947111606598, "learning_rate": 3.870752052099698e-05, "loss": 0.0328, "step": 59390 }, { "grad_norm": 0.6666892170906067, "learning_rate": 3.8691413661048916e-05, "loss": 0.0345, "step": 59400 }, { "grad_norm": 0.22902922332286835, "learning_rate": 3.867530803779085e-05, "loss": 0.0325, "step": 59410 }, { "grad_norm": 0.5215607285499573, "learning_rate": 3.8659203652984054e-05, "loss": 0.0282, "step": 59420 }, { "grad_norm": 0.7385847568511963, "learning_rate": 3.864310050838972e-05, "loss": 0.0256, "step": 59430 }, { "grad_norm": 0.4056553244590759, "learning_rate": 3.8626998605768835e-05, "loss": 0.03, "step": 59440 }, { "grad_norm": 0.2689123749732971, "learning_rate": 3.8610897946882294e-05, "loss": 0.0354, "step": 59450 }, { "grad_norm": 0.26296767592430115, "learning_rate": 3.859479853349082e-05, "loss": 0.024, "step": 59460 }, { "grad_norm": 0.34752485156059265, "learning_rate": 3.857870036735503e-05, "loss": 0.0286, "step": 59470 }, { "grad_norm": 0.2702682316303253, "learning_rate": 3.856260345023538e-05, "loss": 0.0272, "step": 59480 }, { "grad_norm": 0.26822710037231445, "learning_rate": 3.854650778389224e-05, "loss": 0.0313, "step": 59490 }, { "grad_norm": 0.3701217472553253, "learning_rate": 3.8530413370085786e-05, "loss": 0.0296, "step": 59500 }, { "grad_norm": 0.22275908291339874, "learning_rate": 3.851432021057608e-05, "loss": 0.0247, "step": 59510 }, { "grad_norm": 0.2390827238559723, "learning_rate": 3.849822830712306e-05, "loss": 0.0386, "step": 59520 }, { "grad_norm": 0.4470285475254059, "learning_rate": 3.848213766148649e-05, "loss": 0.034, "step": 59530 }, { "grad_norm": 0.3038654625415802, "learning_rate": 3.846604827542602e-05, "loss": 0.0271, "step": 59540 }, { "grad_norm": 0.37137481570243835, "learning_rate": 3.844996015070117e-05, "loss": 0.0292, "step": 59550 }, { "grad_norm": 0.33942660689353943, "learning_rate": 3.843387328907132e-05, "loss": 0.0308, "step": 59560 }, { "grad_norm": 0.26936179399490356, "learning_rate": 3.8417787692295696e-05, "loss": 0.0264, "step": 59570 }, { "grad_norm": 0.3195418119430542, "learning_rate": 3.84017033621334e-05, "loss": 0.0279, "step": 59580 }, { "grad_norm": 0.5574753284454346, "learning_rate": 3.838562030034338e-05, "loss": 0.0367, "step": 59590 }, { "grad_norm": 0.26428771018981934, "learning_rate": 3.836953850868447e-05, "loss": 0.0301, "step": 59600 }, { "grad_norm": 0.2936360239982605, "learning_rate": 3.835345798891533e-05, "loss": 0.026, "step": 59610 }, { "grad_norm": 0.3937585949897766, "learning_rate": 3.833737874279452e-05, "loss": 0.028, "step": 59620 }, { "grad_norm": 0.2540796399116516, "learning_rate": 3.8321300772080433e-05, "loss": 0.0319, "step": 59630 }, { "grad_norm": 0.3555344343185425, "learning_rate": 3.8305224078531345e-05, "loss": 0.0337, "step": 59640 }, { "grad_norm": 0.39844781160354614, "learning_rate": 3.828914866390537e-05, "loss": 0.0315, "step": 59650 }, { "grad_norm": 0.5146704316139221, "learning_rate": 3.827307452996048e-05, "loss": 0.0296, "step": 59660 }, { "grad_norm": 0.2939479947090149, "learning_rate": 3.825700167845454e-05, "loss": 0.0296, "step": 59670 }, { "grad_norm": 0.3073062300682068, "learning_rate": 3.8240930111145214e-05, "loss": 0.0305, "step": 59680 }, { "grad_norm": 0.43543484807014465, "learning_rate": 3.8224859829790116e-05, "loss": 0.0309, "step": 59690 }, { "grad_norm": 0.354497492313385, "learning_rate": 3.820879083614664e-05, "loss": 0.0346, "step": 59700 }, { "grad_norm": 0.28999340534210205, "learning_rate": 3.819272313197208e-05, "loss": 0.0246, "step": 59710 }, { "grad_norm": 0.32335177063941956, "learning_rate": 3.817665671902355e-05, "loss": 0.0316, "step": 59720 }, { "grad_norm": 0.28206339478492737, "learning_rate": 3.816059159905808e-05, "loss": 0.0277, "step": 59730 }, { "grad_norm": 0.3171691596508026, "learning_rate": 3.8144527773832476e-05, "loss": 0.0289, "step": 59740 }, { "grad_norm": 0.33725759387016296, "learning_rate": 3.812846524510352e-05, "loss": 0.0351, "step": 59750 }, { "grad_norm": 0.2645367681980133, "learning_rate": 3.811240401462775e-05, "loss": 0.0304, "step": 59760 }, { "grad_norm": 0.2919531464576721, "learning_rate": 3.809634408416162e-05, "loss": 0.0314, "step": 59770 }, { "grad_norm": 0.3415542542934418, "learning_rate": 3.8080285455461394e-05, "loss": 0.0264, "step": 59780 }, { "grad_norm": 0.29200947284698486, "learning_rate": 3.806422813028323e-05, "loss": 0.0289, "step": 59790 }, { "grad_norm": 0.3907841444015503, "learning_rate": 3.8048172110383114e-05, "loss": 0.0309, "step": 59800 }, { "grad_norm": 0.5871901512145996, "learning_rate": 3.8032117397516944e-05, "loss": 0.0342, "step": 59810 }, { "grad_norm": 0.5450229048728943, "learning_rate": 3.801606399344042e-05, "loss": 0.0332, "step": 59820 }, { "grad_norm": 0.37820306420326233, "learning_rate": 3.800001189990913e-05, "loss": 0.0297, "step": 59830 }, { "grad_norm": 0.42170625925064087, "learning_rate": 3.798396111867847e-05, "loss": 0.0332, "step": 59840 }, { "grad_norm": 0.5552884340286255, "learning_rate": 3.796791165150378e-05, "loss": 0.0358, "step": 59850 }, { "grad_norm": 0.382178395986557, "learning_rate": 3.795186350014018e-05, "loss": 0.0289, "step": 59860 }, { "grad_norm": 1.6063779592514038, "learning_rate": 3.793581666634266e-05, "loss": 0.028, "step": 59870 }, { "grad_norm": 0.36222103238105774, "learning_rate": 3.791977115186609e-05, "loss": 0.0314, "step": 59880 }, { "grad_norm": 0.32652705907821655, "learning_rate": 3.79037269584652e-05, "loss": 0.0292, "step": 59890 }, { "grad_norm": 0.19541847705841064, "learning_rate": 3.7887684087894545e-05, "loss": 0.0348, "step": 59900 }, { "grad_norm": 0.44875413179397583, "learning_rate": 3.787164254190854e-05, "loss": 0.0279, "step": 59910 }, { "grad_norm": 0.213990718126297, "learning_rate": 3.7855602322261473e-05, "loss": 0.026, "step": 59920 }, { "grad_norm": 0.3375113904476166, "learning_rate": 3.783956343070746e-05, "loss": 0.0335, "step": 59930 }, { "grad_norm": 0.3586488664150238, "learning_rate": 3.7823525869000526e-05, "loss": 0.0257, "step": 59940 }, { "grad_norm": 0.38614171743392944, "learning_rate": 3.78074896388945e-05, "loss": 0.0288, "step": 59950 }, { "grad_norm": 0.40816617012023926, "learning_rate": 3.7791454742143075e-05, "loss": 0.0242, "step": 59960 }, { "grad_norm": 0.5524848699569702, "learning_rate": 3.7775421180499795e-05, "loss": 0.0271, "step": 59970 }, { "grad_norm": 0.2715451419353485, "learning_rate": 3.7759388955718086e-05, "loss": 0.027, "step": 59980 }, { "grad_norm": 0.5440803170204163, "learning_rate": 3.774335806955117e-05, "loss": 0.0293, "step": 59990 }, { "grad_norm": 0.3233690857887268, "learning_rate": 3.7727328523752215e-05, "loss": 0.0305, "step": 60000 }, { "grad_norm": 0.34935563802719116, "learning_rate": 3.771130032007415e-05, "loss": 0.0281, "step": 60010 }, { "grad_norm": 0.7173222899436951, "learning_rate": 3.769527346026982e-05, "loss": 0.0277, "step": 60020 }, { "grad_norm": 0.4422612190246582, "learning_rate": 3.767924794609188e-05, "loss": 0.033, "step": 60030 }, { "grad_norm": 0.2640003561973572, "learning_rate": 3.7663223779292854e-05, "loss": 0.0317, "step": 60040 }, { "grad_norm": 0.4267788827419281, "learning_rate": 3.764720096162512e-05, "loss": 0.0314, "step": 60050 }, { "grad_norm": 0.4026969075202942, "learning_rate": 3.763117949484092e-05, "loss": 0.027, "step": 60060 }, { "grad_norm": 0.24621011316776276, "learning_rate": 3.7615159380692344e-05, "loss": 0.0249, "step": 60070 }, { "grad_norm": 0.40699365735054016, "learning_rate": 3.759914062093132e-05, "loss": 0.0315, "step": 60080 }, { "grad_norm": 0.5977046489715576, "learning_rate": 3.7583123217309615e-05, "loss": 0.0337, "step": 60090 }, { "grad_norm": 0.2996485233306885, "learning_rate": 3.7567107171578904e-05, "loss": 0.0276, "step": 60100 }, { "grad_norm": 0.33516767621040344, "learning_rate": 3.755109248549066e-05, "loss": 0.0321, "step": 60110 }, { "grad_norm": 0.2915796935558319, "learning_rate": 3.75350791607962e-05, "loss": 0.0325, "step": 60120 }, { "grad_norm": 0.32411128282546997, "learning_rate": 3.751906719924676e-05, "loss": 0.0292, "step": 60130 }, { "grad_norm": 0.27516162395477295, "learning_rate": 3.750305660259337e-05, "loss": 0.0289, "step": 60140 }, { "grad_norm": 0.3444010019302368, "learning_rate": 3.748704737258693e-05, "loss": 0.0303, "step": 60150 }, { "grad_norm": 0.5134755969047546, "learning_rate": 3.747103951097816e-05, "loss": 0.0272, "step": 60160 }, { "grad_norm": 0.2903999984264374, "learning_rate": 3.745503301951768e-05, "loss": 0.0328, "step": 60170 }, { "grad_norm": 0.2772400677204132, "learning_rate": 3.743902789995592e-05, "loss": 0.0314, "step": 60180 }, { "grad_norm": 0.36910557746887207, "learning_rate": 3.7423024154043183e-05, "loss": 0.0305, "step": 60190 }, { "grad_norm": 0.2611521780490875, "learning_rate": 3.740702178352963e-05, "loss": 0.0319, "step": 60200 }, { "grad_norm": 0.3445073068141937, "learning_rate": 3.739102079016523e-05, "loss": 0.0327, "step": 60210 }, { "grad_norm": 0.41955915093421936, "learning_rate": 3.7375021175699846e-05, "loss": 0.0316, "step": 60220 }, { "grad_norm": 0.2868266701698303, "learning_rate": 3.735902294188317e-05, "loss": 0.0286, "step": 60230 }, { "grad_norm": 0.4109423756599426, "learning_rate": 3.734302609046473e-05, "loss": 0.0325, "step": 60240 }, { "grad_norm": 0.3617579936981201, "learning_rate": 3.732703062319391e-05, "loss": 0.0304, "step": 60250 }, { "grad_norm": 0.43086251616477966, "learning_rate": 3.731103654181999e-05, "loss": 0.0313, "step": 60260 }, { "grad_norm": 0.38920295238494873, "learning_rate": 3.729504384809203e-05, "loss": 0.0293, "step": 60270 }, { "grad_norm": 0.28392061591148376, "learning_rate": 3.727905254375898e-05, "loss": 0.0343, "step": 60280 }, { "grad_norm": 0.36543235182762146, "learning_rate": 3.7263062630569614e-05, "loss": 0.0252, "step": 60290 }, { "grad_norm": 0.41218462586402893, "learning_rate": 3.724707411027256e-05, "loss": 0.0287, "step": 60300 }, { "grad_norm": 0.3432914614677429, "learning_rate": 3.723108698461631e-05, "loss": 0.0266, "step": 60310 }, { "grad_norm": 0.3703325390815735, "learning_rate": 3.72151012553492e-05, "loss": 0.0321, "step": 60320 }, { "grad_norm": 0.3993805944919586, "learning_rate": 3.719911692421939e-05, "loss": 0.0294, "step": 60330 }, { "grad_norm": 0.26783451437950134, "learning_rate": 3.71831339929749e-05, "loss": 0.0363, "step": 60340 }, { "grad_norm": 0.4404415488243103, "learning_rate": 3.716715246336362e-05, "loss": 0.0259, "step": 60350 }, { "grad_norm": 0.3608551025390625, "learning_rate": 3.715117233713324e-05, "loss": 0.0313, "step": 60360 }, { "grad_norm": 0.7764652967453003, "learning_rate": 3.713519361603135e-05, "loss": 0.0296, "step": 60370 }, { "grad_norm": 0.3586021363735199, "learning_rate": 3.711921630180532e-05, "loss": 0.0306, "step": 60380 }, { "grad_norm": 0.23885157704353333, "learning_rate": 3.710324039620245e-05, "loss": 0.0295, "step": 60390 }, { "grad_norm": 0.8642804622650146, "learning_rate": 3.708726590096982e-05, "loss": 0.0338, "step": 60400 }, { "grad_norm": 0.4851354956626892, "learning_rate": 3.707129281785437e-05, "loss": 0.0345, "step": 60410 }, { "grad_norm": 0.2553766667842865, "learning_rate": 3.705532114860291e-05, "loss": 0.0283, "step": 60420 }, { "grad_norm": 0.4265575408935547, "learning_rate": 3.703935089496207e-05, "loss": 0.0285, "step": 60430 }, { "grad_norm": 0.5570845603942871, "learning_rate": 3.702338205867829e-05, "loss": 0.0356, "step": 60440 }, { "grad_norm": 0.516327977180481, "learning_rate": 3.7007414641497976e-05, "loss": 0.0318, "step": 60450 }, { "grad_norm": 0.46844395995140076, "learning_rate": 3.699144864516726e-05, "loss": 0.026, "step": 60460 }, { "grad_norm": 0.48631441593170166, "learning_rate": 3.6975484071432153e-05, "loss": 0.0294, "step": 60470 }, { "grad_norm": 0.6080129146575928, "learning_rate": 3.695952092203854e-05, "loss": 0.0279, "step": 60480 }, { "grad_norm": 0.39720600843429565, "learning_rate": 3.69435591987321e-05, "loss": 0.028, "step": 60490 }, { "grad_norm": 0.40219196677207947, "learning_rate": 3.6927598903258374e-05, "loss": 0.0335, "step": 60500 }, { "grad_norm": 0.3592950999736786, "learning_rate": 3.691164003736281e-05, "loss": 0.0249, "step": 60510 }, { "grad_norm": 0.31224533915519714, "learning_rate": 3.689568260279061e-05, "loss": 0.0281, "step": 60520 }, { "grad_norm": 0.23402245342731476, "learning_rate": 3.687972660128686e-05, "loss": 0.03, "step": 60530 }, { "grad_norm": 0.31413236260414124, "learning_rate": 3.686377203459648e-05, "loss": 0.0334, "step": 60540 }, { "grad_norm": 0.3849943280220032, "learning_rate": 3.6847818904464226e-05, "loss": 0.0352, "step": 60550 }, { "grad_norm": 0.30655092000961304, "learning_rate": 3.683186721263474e-05, "loss": 0.0286, "step": 60560 }, { "grad_norm": 0.313399076461792, "learning_rate": 3.6815916960852435e-05, "loss": 0.0284, "step": 60570 }, { "grad_norm": 0.307523638010025, "learning_rate": 3.679996815086165e-05, "loss": 0.0269, "step": 60580 }, { "grad_norm": 0.33969274163246155, "learning_rate": 3.678402078440649e-05, "loss": 0.026, "step": 60590 }, { "grad_norm": 0.4826972782611847, "learning_rate": 3.676807486323096e-05, "loss": 0.0277, "step": 60600 }, { "grad_norm": 0.3351215720176697, "learning_rate": 3.6752130389078864e-05, "loss": 0.0342, "step": 60610 }, { "grad_norm": 0.4144783020019531, "learning_rate": 3.673618736369388e-05, "loss": 0.0313, "step": 60620 }, { "grad_norm": 0.4163048565387726, "learning_rate": 3.6720245788819474e-05, "loss": 0.034, "step": 60630 }, { "grad_norm": 0.45801323652267456, "learning_rate": 3.6704305666199044e-05, "loss": 0.0324, "step": 60640 }, { "grad_norm": 0.34451356530189514, "learning_rate": 3.668836699757576e-05, "loss": 0.0264, "step": 60650 }, { "grad_norm": 0.3621886670589447, "learning_rate": 3.667242978469265e-05, "loss": 0.0314, "step": 60660 }, { "grad_norm": 0.42712193727493286, "learning_rate": 3.6656494029292575e-05, "loss": 0.0282, "step": 60670 }, { "grad_norm": 0.40383094549179077, "learning_rate": 3.664055973311825e-05, "loss": 0.0273, "step": 60680 }, { "grad_norm": 0.3650606870651245, "learning_rate": 3.662462689791221e-05, "loss": 0.0271, "step": 60690 }, { "grad_norm": 0.31534433364868164, "learning_rate": 3.660869552541689e-05, "loss": 0.0315, "step": 60700 }, { "grad_norm": 0.28516170382499695, "learning_rate": 3.659276561737448e-05, "loss": 0.0317, "step": 60710 }, { "grad_norm": 0.5301467776298523, "learning_rate": 3.657683717552708e-05, "loss": 0.0293, "step": 60720 }, { "grad_norm": 0.40657439827919006, "learning_rate": 3.6560910201616574e-05, "loss": 0.0311, "step": 60730 }, { "grad_norm": 0.23690038919448853, "learning_rate": 3.654498469738472e-05, "loss": 0.0267, "step": 60740 }, { "grad_norm": 0.4067670702934265, "learning_rate": 3.6529060664573126e-05, "loss": 0.0294, "step": 60750 }, { "grad_norm": 0.3749687671661377, "learning_rate": 3.6513138104923176e-05, "loss": 0.0286, "step": 60760 }, { "grad_norm": 0.5299895405769348, "learning_rate": 3.6497217020176176e-05, "loss": 0.035, "step": 60770 }, { "grad_norm": 0.43555495142936707, "learning_rate": 3.648129741207323e-05, "loss": 0.0255, "step": 60780 }, { "grad_norm": 0.5782963037490845, "learning_rate": 3.646537928235527e-05, "loss": 0.0337, "step": 60790 }, { "grad_norm": 0.31710943579673767, "learning_rate": 3.6449462632763075e-05, "loss": 0.0255, "step": 60800 }, { "grad_norm": 0.44339144229888916, "learning_rate": 3.643354746503729e-05, "loss": 0.0282, "step": 60810 }, { "grad_norm": 0.3516450822353363, "learning_rate": 3.6417633780918345e-05, "loss": 0.0289, "step": 60820 }, { "grad_norm": 0.33068299293518066, "learning_rate": 3.640172158214655e-05, "loss": 0.0276, "step": 60830 }, { "grad_norm": 0.3376500606536865, "learning_rate": 3.6385810870462035e-05, "loss": 0.0279, "step": 60840 }, { "grad_norm": 0.3375605046749115, "learning_rate": 3.636990164760479e-05, "loss": 0.024, "step": 60850 }, { "grad_norm": 0.7978036403656006, "learning_rate": 3.6353993915314596e-05, "loss": 0.0267, "step": 60860 }, { "grad_norm": 0.4116632044315338, "learning_rate": 3.6338087675331124e-05, "loss": 0.0277, "step": 60870 }, { "grad_norm": 0.513988733291626, "learning_rate": 3.632218292939383e-05, "loss": 0.0272, "step": 60880 }, { "grad_norm": 0.3299410045146942, "learning_rate": 3.6306279679242033e-05, "loss": 0.0276, "step": 60890 }, { "grad_norm": 0.45133092999458313, "learning_rate": 3.629037792661492e-05, "loss": 0.0266, "step": 60900 }, { "grad_norm": 0.35569095611572266, "learning_rate": 3.627447767325147e-05, "loss": 0.0274, "step": 60910 }, { "grad_norm": 0.4659663736820221, "learning_rate": 3.62585789208905e-05, "loss": 0.034, "step": 60920 }, { "grad_norm": 0.2956309914588928, "learning_rate": 3.624268167127068e-05, "loss": 0.0257, "step": 60930 }, { "grad_norm": 0.46975570917129517, "learning_rate": 3.622678592613051e-05, "loss": 0.0305, "step": 60940 }, { "grad_norm": 0.4097931981086731, "learning_rate": 3.6210891687208304e-05, "loss": 0.0367, "step": 60950 }, { "grad_norm": 0.41457757353782654, "learning_rate": 3.619499895624228e-05, "loss": 0.028, "step": 60960 }, { "grad_norm": 0.520061194896698, "learning_rate": 3.61791077349704e-05, "loss": 0.0351, "step": 60970 }, { "grad_norm": 0.4378231465816498, "learning_rate": 3.616321802513053e-05, "loss": 0.0286, "step": 60980 }, { "grad_norm": 0.2962559461593628, "learning_rate": 3.6147329828460325e-05, "loss": 0.0306, "step": 60990 }, { "grad_norm": 0.43158087134361267, "learning_rate": 3.6131443146697306e-05, "loss": 0.0361, "step": 61000 }, { "grad_norm": 0.4441055357456207, "learning_rate": 3.611555798157879e-05, "loss": 0.0305, "step": 61010 }, { "grad_norm": 0.3124358355998993, "learning_rate": 3.6099674334842e-05, "loss": 0.0288, "step": 61020 }, { "grad_norm": 0.41559767723083496, "learning_rate": 3.6083792208223934e-05, "loss": 0.0288, "step": 61030 }, { "grad_norm": 0.3134961426258087, "learning_rate": 3.606791160346142e-05, "loss": 0.0309, "step": 61040 }, { "grad_norm": 1.3225362300872803, "learning_rate": 3.6052032522291134e-05, "loss": 0.0298, "step": 61050 }, { "grad_norm": 0.3179236948490143, "learning_rate": 3.603615496644962e-05, "loss": 0.0378, "step": 61060 }, { "grad_norm": 0.3228034973144531, "learning_rate": 3.60202789376732e-05, "loss": 0.0243, "step": 61070 }, { "grad_norm": 0.3059404492378235, "learning_rate": 3.6004404437698036e-05, "loss": 0.0297, "step": 61080 }, { "grad_norm": 0.33347731828689575, "learning_rate": 3.5988531468260184e-05, "loss": 0.0291, "step": 61090 }, { "grad_norm": 0.28470367193222046, "learning_rate": 3.597266003109546e-05, "loss": 0.0296, "step": 61100 }, { "grad_norm": 0.38379812240600586, "learning_rate": 3.5956790127939557e-05, "loss": 0.0317, "step": 61110 }, { "grad_norm": 0.35306864976882935, "learning_rate": 3.594092176052796e-05, "loss": 0.0364, "step": 61120 }, { "grad_norm": 0.3695659637451172, "learning_rate": 3.592505493059603e-05, "loss": 0.0288, "step": 61130 }, { "grad_norm": 0.4845888018608093, "learning_rate": 3.5909189639878905e-05, "loss": 0.0363, "step": 61140 }, { "grad_norm": 0.29323384165763855, "learning_rate": 3.5893325890111626e-05, "loss": 0.0309, "step": 61150 }, { "grad_norm": 0.480007529258728, "learning_rate": 3.5877463683029034e-05, "loss": 0.0349, "step": 61160 }, { "grad_norm": 0.36298230290412903, "learning_rate": 3.586160302036578e-05, "loss": 0.0279, "step": 61170 }, { "grad_norm": 0.4478786289691925, "learning_rate": 3.584574390385636e-05, "loss": 0.0282, "step": 61180 }, { "grad_norm": 0.5104972720146179, "learning_rate": 3.582988633523511e-05, "loss": 0.0272, "step": 61190 }, { "grad_norm": 0.3753458559513092, "learning_rate": 3.581403031623616e-05, "loss": 0.0265, "step": 61200 }, { "grad_norm": 0.46147096157073975, "learning_rate": 3.579817584859354e-05, "loss": 0.0285, "step": 61210 }, { "grad_norm": 0.29411354660987854, "learning_rate": 3.578232293404107e-05, "loss": 0.0299, "step": 61220 }, { "grad_norm": 0.2680078446865082, "learning_rate": 3.5766471574312385e-05, "loss": 0.0277, "step": 61230 }, { "grad_norm": 0.4342884421348572, "learning_rate": 3.5750621771140964e-05, "loss": 0.0244, "step": 61240 }, { "grad_norm": 0.4509493410587311, "learning_rate": 3.5734773526260115e-05, "loss": 0.0267, "step": 61250 }, { "grad_norm": 0.29896727204322815, "learning_rate": 3.5718926841402996e-05, "loss": 0.0287, "step": 61260 }, { "grad_norm": 0.3208310902118683, "learning_rate": 3.5703081718302555e-05, "loss": 0.0297, "step": 61270 }, { "grad_norm": 0.40194788575172424, "learning_rate": 3.568723815869162e-05, "loss": 0.0326, "step": 61280 }, { "grad_norm": 0.44553831219673157, "learning_rate": 3.5671396164302777e-05, "loss": 0.0302, "step": 61290 }, { "grad_norm": 0.4861428141593933, "learning_rate": 3.565555573686853e-05, "loss": 0.0276, "step": 61300 }, { "grad_norm": 0.3605996370315552, "learning_rate": 3.563971687812113e-05, "loss": 0.0373, "step": 61310 }, { "grad_norm": 0.5261538028717041, "learning_rate": 3.5623879589792706e-05, "loss": 0.0263, "step": 61320 }, { "grad_norm": 0.23042365908622742, "learning_rate": 3.5608043873615185e-05, "loss": 0.0229, "step": 61330 }, { "grad_norm": 0.38701730966567993, "learning_rate": 3.559220973132035e-05, "loss": 0.0296, "step": 61340 }, { "grad_norm": 0.36842772364616394, "learning_rate": 3.557637716463981e-05, "loss": 0.0268, "step": 61350 }, { "grad_norm": 0.3824273645877838, "learning_rate": 3.556054617530497e-05, "loss": 0.0312, "step": 61360 }, { "grad_norm": 0.3661669194698334, "learning_rate": 3.5544716765047106e-05, "loss": 0.033, "step": 61370 }, { "grad_norm": 0.2761199474334717, "learning_rate": 3.5528888935597274e-05, "loss": 0.0286, "step": 61380 }, { "grad_norm": 0.4732750952243805, "learning_rate": 3.551306268868636e-05, "loss": 0.0221, "step": 61390 }, { "grad_norm": 0.33546045422554016, "learning_rate": 3.5497238026045166e-05, "loss": 0.0294, "step": 61400 }, { "grad_norm": 0.6395872831344604, "learning_rate": 3.548141494940421e-05, "loss": 0.0324, "step": 61410 }, { "grad_norm": 0.3491553068161011, "learning_rate": 3.546559346049388e-05, "loss": 0.0254, "step": 61420 }, { "grad_norm": 0.4693342447280884, "learning_rate": 3.544977356104441e-05, "loss": 0.029, "step": 61430 }, { "grad_norm": 0.3617619276046753, "learning_rate": 3.543395525278582e-05, "loss": 0.0254, "step": 61440 }, { "grad_norm": 0.5273459553718567, "learning_rate": 3.5418138537447985e-05, "loss": 0.0316, "step": 61450 }, { "grad_norm": 0.27180877327919006, "learning_rate": 3.5402323416760574e-05, "loss": 0.0286, "step": 61460 }, { "grad_norm": 0.23824883997440338, "learning_rate": 3.538650989245316e-05, "loss": 0.0263, "step": 61470 }, { "grad_norm": 0.2343309372663498, "learning_rate": 3.537069796625504e-05, "loss": 0.029, "step": 61480 }, { "grad_norm": 0.2867155969142914, "learning_rate": 3.53548876398954e-05, "loss": 0.0232, "step": 61490 }, { "grad_norm": 0.2926926016807556, "learning_rate": 3.5339078915103214e-05, "loss": 0.0251, "step": 61500 }, { "grad_norm": 0.29792484641075134, "learning_rate": 3.532327179360733e-05, "loss": 0.0315, "step": 61510 }, { "grad_norm": 0.3593597114086151, "learning_rate": 3.530746627713636e-05, "loss": 0.0262, "step": 61520 }, { "grad_norm": 0.3459865152835846, "learning_rate": 3.5291662367418805e-05, "loss": 0.025, "step": 61530 }, { "grad_norm": 0.3145887553691864, "learning_rate": 3.527586006618293e-05, "loss": 0.0299, "step": 61540 }, { "grad_norm": 0.3547314703464508, "learning_rate": 3.526005937515687e-05, "loss": 0.0354, "step": 61550 }, { "grad_norm": 0.43758368492126465, "learning_rate": 3.524426029606856e-05, "loss": 0.0323, "step": 61560 }, { "grad_norm": 0.29486358165740967, "learning_rate": 3.5228462830645746e-05, "loss": 0.0232, "step": 61570 }, { "grad_norm": 0.4446758031845093, "learning_rate": 3.5212666980616036e-05, "loss": 0.0299, "step": 61580 }, { "grad_norm": 0.38049569725990295, "learning_rate": 3.519687274770681e-05, "loss": 0.0255, "step": 61590 }, { "grad_norm": 0.3242024779319763, "learning_rate": 3.518108013364535e-05, "loss": 0.0265, "step": 61600 }, { "grad_norm": 0.3146022856235504, "learning_rate": 3.516528914015868e-05, "loss": 0.0286, "step": 61610 }, { "grad_norm": 0.416096955537796, "learning_rate": 3.5149499768973694e-05, "loss": 0.0287, "step": 61620 }, { "grad_norm": 0.30024904012680054, "learning_rate": 3.513371202181708e-05, "loss": 0.0284, "step": 61630 }, { "grad_norm": 0.3544002175331116, "learning_rate": 3.511792590041537e-05, "loss": 0.033, "step": 61640 }, { "grad_norm": 0.5098653435707092, "learning_rate": 3.510214140649488e-05, "loss": 0.0263, "step": 61650 }, { "grad_norm": 0.46843770146369934, "learning_rate": 3.508635854178184e-05, "loss": 0.0297, "step": 61660 }, { "grad_norm": 0.26161977648735046, "learning_rate": 3.50705773080022e-05, "loss": 0.0274, "step": 61670 }, { "grad_norm": 0.34490513801574707, "learning_rate": 3.505479770688178e-05, "loss": 0.0275, "step": 61680 }, { "grad_norm": 0.33216336369514465, "learning_rate": 3.503901974014622e-05, "loss": 0.0273, "step": 61690 }, { "grad_norm": 0.5865685939788818, "learning_rate": 3.502324340952095e-05, "loss": 0.0292, "step": 61700 }, { "grad_norm": 0.5450891256332397, "learning_rate": 3.500746871673126e-05, "loss": 0.0283, "step": 61710 }, { "grad_norm": 0.2497413009405136, "learning_rate": 3.499169566350227e-05, "loss": 0.0311, "step": 61720 }, { "grad_norm": 0.2643522620201111, "learning_rate": 3.4975924251558876e-05, "loss": 0.0339, "step": 61730 }, { "grad_norm": 0.4559076428413391, "learning_rate": 3.496015448262582e-05, "loss": 0.0323, "step": 61740 }, { "grad_norm": 0.3096384108066559, "learning_rate": 3.494438635842764e-05, "loss": 0.0279, "step": 61750 }, { "grad_norm": 0.5146913528442383, "learning_rate": 3.492861988068874e-05, "loss": 0.032, "step": 61760 }, { "grad_norm": 0.4517003297805786, "learning_rate": 3.491285505113332e-05, "loss": 0.0263, "step": 61770 }, { "grad_norm": 0.2784380614757538, "learning_rate": 3.4897091871485376e-05, "loss": 0.0246, "step": 61780 }, { "grad_norm": 0.2777532935142517, "learning_rate": 3.488133034346875e-05, "loss": 0.0311, "step": 61790 }, { "grad_norm": 0.5839009284973145, "learning_rate": 3.486557046880713e-05, "loss": 0.031, "step": 61800 }, { "grad_norm": 0.2063588947057724, "learning_rate": 3.484981224922395e-05, "loss": 0.0263, "step": 61810 }, { "grad_norm": 0.6811996698379517, "learning_rate": 3.4834055686442544e-05, "loss": 0.0253, "step": 61820 }, { "grad_norm": 0.5204756259918213, "learning_rate": 3.4818300782186e-05, "loss": 0.0321, "step": 61830 }, { "grad_norm": 0.28500422835350037, "learning_rate": 3.4802547538177236e-05, "loss": 0.0278, "step": 61840 }, { "grad_norm": 0.2994634807109833, "learning_rate": 3.478679595613905e-05, "loss": 0.0287, "step": 61850 }, { "grad_norm": 0.2882651090621948, "learning_rate": 3.4771046037793975e-05, "loss": 0.0205, "step": 61860 }, { "grad_norm": 0.48363086581230164, "learning_rate": 3.4755297784864416e-05, "loss": 0.0248, "step": 61870 }, { "grad_norm": 0.3572525978088379, "learning_rate": 3.473955119907257e-05, "loss": 0.023, "step": 61880 }, { "grad_norm": 0.33367598056793213, "learning_rate": 3.472380628214046e-05, "loss": 0.0238, "step": 61890 }, { "grad_norm": 0.255771279335022, "learning_rate": 3.4708063035789906e-05, "loss": 0.0277, "step": 61900 }, { "grad_norm": 0.34554800391197205, "learning_rate": 3.4692321461742614e-05, "loss": 0.0289, "step": 61910 }, { "grad_norm": 0.2210932970046997, "learning_rate": 3.4676581561720023e-05, "loss": 0.0308, "step": 61920 }, { "grad_norm": 0.38086748123168945, "learning_rate": 3.466084333744344e-05, "loss": 0.0236, "step": 61930 }, { "grad_norm": 0.36675965785980225, "learning_rate": 3.4645106790633964e-05, "loss": 0.0283, "step": 61940 }, { "grad_norm": 0.5287811756134033, "learning_rate": 3.4629371923012526e-05, "loss": 0.0256, "step": 61950 }, { "grad_norm": 0.27161192893981934, "learning_rate": 3.4613638736299856e-05, "loss": 0.0329, "step": 61960 }, { "grad_norm": 0.4154455065727234, "learning_rate": 3.459790723221652e-05, "loss": 0.0237, "step": 61970 }, { "grad_norm": 1.017695426940918, "learning_rate": 3.45821774124829e-05, "loss": 0.0275, "step": 61980 }, { "grad_norm": 0.285540908575058, "learning_rate": 3.456644927881918e-05, "loss": 0.034, "step": 61990 }, { "grad_norm": 0.3763059973716736, "learning_rate": 3.4550722832945354e-05, "loss": 0.0302, "step": 62000 }, { "grad_norm": 0.4153389632701874, "learning_rate": 3.453499807658126e-05, "loss": 0.0306, "step": 62010 }, { "grad_norm": 0.35411813855171204, "learning_rate": 3.451927501144653e-05, "loss": 0.0293, "step": 62020 }, { "grad_norm": 0.358533650636673, "learning_rate": 3.450355363926059e-05, "loss": 0.0276, "step": 62030 }, { "grad_norm": 0.25557592511177063, "learning_rate": 3.448783396174273e-05, "loss": 0.0286, "step": 62040 }, { "grad_norm": 0.3283678889274597, "learning_rate": 3.4472115980612034e-05, "loss": 0.0241, "step": 62050 }, { "grad_norm": 0.36082693934440613, "learning_rate": 3.4456399697587396e-05, "loss": 0.0235, "step": 62060 }, { "grad_norm": 0.46070024371147156, "learning_rate": 3.4440685114387506e-05, "loss": 0.0369, "step": 62070 }, { "grad_norm": 0.3089667856693268, "learning_rate": 3.442497223273091e-05, "loss": 0.0311, "step": 62080 }, { "grad_norm": 0.22797733545303345, "learning_rate": 3.4409261054335926e-05, "loss": 0.0279, "step": 62090 }, { "grad_norm": 0.27237164974212646, "learning_rate": 3.439355158092069e-05, "loss": 0.029, "step": 62100 }, { "grad_norm": 0.3973372280597687, "learning_rate": 3.437784381420322e-05, "loss": 0.0297, "step": 62110 }, { "grad_norm": 0.570720911026001, "learning_rate": 3.436213775590126e-05, "loss": 0.0252, "step": 62120 }, { "grad_norm": 0.3527185618877411, "learning_rate": 3.4346433407732405e-05, "loss": 0.0287, "step": 62130 }, { "grad_norm": 0.36174678802490234, "learning_rate": 3.433073077141405e-05, "loss": 0.0278, "step": 62140 }, { "grad_norm": 0.2577648460865021, "learning_rate": 3.431502984866343e-05, "loss": 0.0314, "step": 62150 }, { "grad_norm": 0.26490992307662964, "learning_rate": 3.429933064119753e-05, "loss": 0.0237, "step": 62160 }, { "grad_norm": 0.2576507329940796, "learning_rate": 3.4283633150733255e-05, "loss": 0.0292, "step": 62170 }, { "grad_norm": 0.4077650308609009, "learning_rate": 3.4267937378987225e-05, "loss": 0.0259, "step": 62180 }, { "grad_norm": 0.32908323407173157, "learning_rate": 3.425224332767592e-05, "loss": 0.0316, "step": 62190 }, { "grad_norm": 0.296132892370224, "learning_rate": 3.4236550998515595e-05, "loss": 0.0234, "step": 62200 }, { "grad_norm": 0.3062323033809662, "learning_rate": 3.422086039322235e-05, "loss": 0.0234, "step": 62210 }, { "grad_norm": 0.29738229513168335, "learning_rate": 3.420517151351209e-05, "loss": 0.0256, "step": 62220 }, { "grad_norm": 0.33506232500076294, "learning_rate": 3.418948436110053e-05, "loss": 0.0292, "step": 62230 }, { "grad_norm": 0.2901447117328644, "learning_rate": 3.417379893770319e-05, "loss": 0.0266, "step": 62240 }, { "grad_norm": 0.4457056224346161, "learning_rate": 3.41581152450354e-05, "loss": 0.0276, "step": 62250 }, { "grad_norm": 0.31905514001846313, "learning_rate": 3.414243328481231e-05, "loss": 0.0268, "step": 62260 }, { "grad_norm": 0.370435506105423, "learning_rate": 3.412675305874888e-05, "loss": 0.0282, "step": 62270 }, { "grad_norm": 0.3432117700576782, "learning_rate": 3.4111074568559876e-05, "loss": 0.0308, "step": 62280 }, { "grad_norm": 0.41840800642967224, "learning_rate": 3.4095397815959846e-05, "loss": 0.0263, "step": 62290 }, { "grad_norm": 0.2930276095867157, "learning_rate": 3.407972280266322e-05, "loss": 0.0295, "step": 62300 }, { "grad_norm": 0.5202215909957886, "learning_rate": 3.4064049530384166e-05, "loss": 0.0267, "step": 62310 }, { "grad_norm": 0.3737333118915558, "learning_rate": 3.40483780008367e-05, "loss": 0.0281, "step": 62320 }, { "grad_norm": 0.985963761806488, "learning_rate": 3.4032708215734635e-05, "loss": 0.0298, "step": 62330 }, { "grad_norm": 0.5296128988265991, "learning_rate": 3.4017040176791584e-05, "loss": 0.0233, "step": 62340 }, { "grad_norm": 0.3693578541278839, "learning_rate": 3.4001373885720976e-05, "loss": 0.0288, "step": 62350 }, { "grad_norm": 0.38500070571899414, "learning_rate": 3.398570934423608e-05, "loss": 0.0333, "step": 62360 }, { "grad_norm": 0.23356425762176514, "learning_rate": 3.397004655404994e-05, "loss": 0.0289, "step": 62370 }, { "grad_norm": 0.24525925517082214, "learning_rate": 3.395438551687541e-05, "loss": 0.0268, "step": 62380 }, { "grad_norm": 0.3525067865848541, "learning_rate": 3.393872623442515e-05, "loss": 0.0255, "step": 62390 }, { "grad_norm": 0.24136784672737122, "learning_rate": 3.392306870841164e-05, "loss": 0.0269, "step": 62400 }, { "grad_norm": 0.4862362742424011, "learning_rate": 3.390741294054715e-05, "loss": 0.0323, "step": 62410 }, { "grad_norm": 0.3295219838619232, "learning_rate": 3.38917589325438e-05, "loss": 0.0264, "step": 62420 }, { "grad_norm": 0.3404320180416107, "learning_rate": 3.387610668611349e-05, "loss": 0.0264, "step": 62430 }, { "grad_norm": 0.3426511287689209, "learning_rate": 3.38604562029679e-05, "loss": 0.0344, "step": 62440 }, { "grad_norm": 0.3208913803100586, "learning_rate": 3.384480748481855e-05, "loss": 0.0316, "step": 62450 }, { "grad_norm": 0.2684735059738159, "learning_rate": 3.382916053337676e-05, "loss": 0.0284, "step": 62460 }, { "grad_norm": 0.4332285225391388, "learning_rate": 3.381351535035366e-05, "loss": 0.0309, "step": 62470 }, { "grad_norm": 0.4016023576259613, "learning_rate": 3.379787193746019e-05, "loss": 0.0322, "step": 62480 }, { "grad_norm": 0.4167860150337219, "learning_rate": 3.378223029640709e-05, "loss": 0.0248, "step": 62490 }, { "grad_norm": 0.3474137485027313, "learning_rate": 3.376659042890489e-05, "loss": 0.0303, "step": 62500 }, { "grad_norm": 0.2692294120788574, "learning_rate": 3.375095233666397e-05, "loss": 0.0282, "step": 62510 }, { "grad_norm": 1.1159861087799072, "learning_rate": 3.373531602139447e-05, "loss": 0.0248, "step": 62520 }, { "grad_norm": 0.45520371198654175, "learning_rate": 3.371968148480635e-05, "loss": 0.0264, "step": 62530 }, { "grad_norm": 1.0941869020462036, "learning_rate": 3.370404872860937e-05, "loss": 0.0298, "step": 62540 }, { "grad_norm": 0.21470682322978973, "learning_rate": 3.3688417754513145e-05, "loss": 0.0273, "step": 62550 }, { "grad_norm": 0.32875117659568787, "learning_rate": 3.367278856422703e-05, "loss": 0.0348, "step": 62560 }, { "grad_norm": 0.51165771484375, "learning_rate": 3.365716115946021e-05, "loss": 0.0294, "step": 62570 }, { "grad_norm": 0.633712887763977, "learning_rate": 3.3641535541921666e-05, "loss": 0.0288, "step": 62580 }, { "grad_norm": 0.33170199394226074, "learning_rate": 3.36259117133202e-05, "loss": 0.0347, "step": 62590 }, { "grad_norm": 0.41198796033859253, "learning_rate": 3.36102896753644e-05, "loss": 0.031, "step": 62600 }, { "grad_norm": 0.2996537685394287, "learning_rate": 3.35946694297627e-05, "loss": 0.0312, "step": 62610 }, { "grad_norm": 0.5855415463447571, "learning_rate": 3.357905097822328e-05, "loss": 0.0296, "step": 62620 }, { "grad_norm": 0.3097180426120758, "learning_rate": 3.356343432245416e-05, "loss": 0.0265, "step": 62630 }, { "grad_norm": 0.3966919183731079, "learning_rate": 3.354781946416315e-05, "loss": 0.0296, "step": 62640 }, { "grad_norm": 0.4898870289325714, "learning_rate": 3.3532206405057864e-05, "loss": 0.0319, "step": 62650 }, { "grad_norm": 0.31139370799064636, "learning_rate": 3.351659514684573e-05, "loss": 0.0271, "step": 62660 }, { "grad_norm": 0.5752801299095154, "learning_rate": 3.350098569123395e-05, "loss": 0.0292, "step": 62670 }, { "grad_norm": 0.5360574126243591, "learning_rate": 3.348537803992958e-05, "loss": 0.0312, "step": 62680 }, { "grad_norm": 0.33781012892723083, "learning_rate": 3.346977219463945e-05, "loss": 0.0349, "step": 62690 }, { "grad_norm": 0.2684086263179779, "learning_rate": 3.345416815707017e-05, "loss": 0.0317, "step": 62700 }, { "grad_norm": 0.29289716482162476, "learning_rate": 3.3438565928928185e-05, "loss": 0.0277, "step": 62710 }, { "grad_norm": 0.2844989597797394, "learning_rate": 3.3422965511919736e-05, "loss": 0.03, "step": 62720 }, { "grad_norm": 0.32603347301483154, "learning_rate": 3.3407366907750846e-05, "loss": 0.0241, "step": 62730 }, { "grad_norm": 0.2734818458557129, "learning_rate": 3.339177011812737e-05, "loss": 0.0296, "step": 62740 }, { "grad_norm": 0.4707992374897003, "learning_rate": 3.3376175144754935e-05, "loss": 0.0288, "step": 62750 }, { "grad_norm": 0.4055987000465393, "learning_rate": 3.3360581989339004e-05, "loss": 0.0291, "step": 62760 }, { "grad_norm": 0.5611565113067627, "learning_rate": 3.3344990653584804e-05, "loss": 0.0307, "step": 62770 }, { "grad_norm": 0.41357648372650146, "learning_rate": 3.332940113919739e-05, "loss": 0.0298, "step": 62780 }, { "grad_norm": 0.4351855516433716, "learning_rate": 3.331381344788161e-05, "loss": 0.0301, "step": 62790 }, { "grad_norm": 0.3348608911037445, "learning_rate": 3.3298227581342076e-05, "loss": 0.0274, "step": 62800 }, { "grad_norm": 0.37052157521247864, "learning_rate": 3.328264354128329e-05, "loss": 0.0266, "step": 62810 }, { "grad_norm": 0.27663516998291016, "learning_rate": 3.3267061329409465e-05, "loss": 0.0293, "step": 62820 }, { "grad_norm": 0.6675108671188354, "learning_rate": 3.325148094742466e-05, "loss": 0.0283, "step": 62830 }, { "grad_norm": 0.3136903643608093, "learning_rate": 3.323590239703271e-05, "loss": 0.0343, "step": 62840 }, { "grad_norm": 0.32168418169021606, "learning_rate": 3.322032567993728e-05, "loss": 0.0301, "step": 62850 }, { "grad_norm": 0.31554320454597473, "learning_rate": 3.3204750797841774e-05, "loss": 0.0259, "step": 62860 }, { "grad_norm": 0.35149824619293213, "learning_rate": 3.31891777524495e-05, "loss": 0.0254, "step": 62870 }, { "grad_norm": 0.2816377878189087, "learning_rate": 3.317360654546347e-05, "loss": 0.0316, "step": 62880 }, { "grad_norm": 0.2760234475135803, "learning_rate": 3.3158037178586534e-05, "loss": 0.028, "step": 62890 }, { "grad_norm": 0.25228166580200195, "learning_rate": 3.3142469653521316e-05, "loss": 0.025, "step": 62900 }, { "grad_norm": 0.4476873278617859, "learning_rate": 3.3126903971970284e-05, "loss": 0.0241, "step": 62910 }, { "grad_norm": 0.3786749541759491, "learning_rate": 3.311134013563564e-05, "loss": 0.0316, "step": 62920 }, { "grad_norm": 0.2563769221305847, "learning_rate": 3.309577814621947e-05, "loss": 0.0238, "step": 62930 }, { "grad_norm": 0.38937389850616455, "learning_rate": 3.308021800542359e-05, "loss": 0.0252, "step": 62940 }, { "grad_norm": 0.35096126794815063, "learning_rate": 3.306465971494963e-05, "loss": 0.0332, "step": 62950 }, { "grad_norm": 0.4906157851219177, "learning_rate": 3.304910327649901e-05, "loss": 0.0269, "step": 62960 }, { "grad_norm": 0.26786142587661743, "learning_rate": 3.303354869177297e-05, "loss": 0.0268, "step": 62970 }, { "grad_norm": 0.3028166890144348, "learning_rate": 3.301799596247255e-05, "loss": 0.0324, "step": 62980 }, { "grad_norm": 0.45896539092063904, "learning_rate": 3.300244509029854e-05, "loss": 0.0283, "step": 62990 }, { "grad_norm": 0.3532068133354187, "learning_rate": 3.2986896076951585e-05, "loss": 0.0289, "step": 63000 }, { "grad_norm": 0.4357883036136627, "learning_rate": 3.297134892413211e-05, "loss": 0.0276, "step": 63010 }, { "grad_norm": 0.34186092019081116, "learning_rate": 3.295580363354031e-05, "loss": 0.0323, "step": 63020 }, { "grad_norm": 0.3453133702278137, "learning_rate": 3.2940260206876194e-05, "loss": 0.0286, "step": 63030 }, { "grad_norm": 0.25972190499305725, "learning_rate": 3.292471864583958e-05, "loss": 0.0283, "step": 63040 }, { "grad_norm": 0.24455079436302185, "learning_rate": 3.290917895213004e-05, "loss": 0.027, "step": 63050 }, { "grad_norm": 0.3932330906391144, "learning_rate": 3.289364112744701e-05, "loss": 0.0312, "step": 63060 }, { "grad_norm": 0.39565935730934143, "learning_rate": 3.287810517348967e-05, "loss": 0.0324, "step": 63070 }, { "grad_norm": 1.401796579360962, "learning_rate": 3.2862571091957014e-05, "loss": 0.0345, "step": 63080 }, { "grad_norm": 0.46047061681747437, "learning_rate": 3.284703888454781e-05, "loss": 0.0315, "step": 63090 }, { "grad_norm": 0.32701677083969116, "learning_rate": 3.2831508552960646e-05, "loss": 0.0286, "step": 63100 }, { "grad_norm": 0.4140342175960541, "learning_rate": 3.281598009889387e-05, "loss": 0.0314, "step": 63110 }, { "grad_norm": 0.3783699572086334, "learning_rate": 3.28004535240457e-05, "loss": 0.0291, "step": 63120 }, { "grad_norm": 0.27881377935409546, "learning_rate": 3.278492883011407e-05, "loss": 0.024, "step": 63130 }, { "grad_norm": 0.4123094975948334, "learning_rate": 3.276940601879675e-05, "loss": 0.0274, "step": 63140 }, { "grad_norm": 0.3515207767486572, "learning_rate": 3.275388509179127e-05, "loss": 0.0311, "step": 63150 }, { "grad_norm": 0.2955622375011444, "learning_rate": 3.2738366050794985e-05, "loss": 0.0226, "step": 63160 }, { "grad_norm": 0.4029429256916046, "learning_rate": 3.2722848897505034e-05, "loss": 0.0295, "step": 63170 }, { "grad_norm": 0.2789786458015442, "learning_rate": 3.2707333633618356e-05, "loss": 0.0236, "step": 63180 }, { "grad_norm": 0.4328194260597229, "learning_rate": 3.269182026083167e-05, "loss": 0.0262, "step": 63190 }, { "grad_norm": 0.3329848647117615, "learning_rate": 3.2676308780841516e-05, "loss": 0.0242, "step": 63200 }, { "grad_norm": 0.45853564143180847, "learning_rate": 3.266079919534416e-05, "loss": 0.0268, "step": 63210 }, { "grad_norm": 0.3073749542236328, "learning_rate": 3.264529150603575e-05, "loss": 0.0266, "step": 63220 }, { "grad_norm": 0.34766414761543274, "learning_rate": 3.262978571461217e-05, "loss": 0.0304, "step": 63230 }, { "grad_norm": 0.5564694404602051, "learning_rate": 3.26142818227691e-05, "loss": 0.0325, "step": 63240 }, { "grad_norm": 0.2537081241607666, "learning_rate": 3.259877983220203e-05, "loss": 0.0248, "step": 63250 }, { "grad_norm": 0.35066699981689453, "learning_rate": 3.2583279744606246e-05, "loss": 0.0257, "step": 63260 }, { "grad_norm": 0.2581159472465515, "learning_rate": 3.25677815616768e-05, "loss": 0.0275, "step": 63270 }, { "grad_norm": 0.5195747017860413, "learning_rate": 3.255228528510856e-05, "loss": 0.0289, "step": 63280 }, { "grad_norm": 0.35147783160209656, "learning_rate": 3.253679091659617e-05, "loss": 0.0304, "step": 63290 }, { "grad_norm": 0.45048969984054565, "learning_rate": 3.252129845783404e-05, "loss": 0.0302, "step": 63300 }, { "grad_norm": 0.30541616678237915, "learning_rate": 3.2505807910516456e-05, "loss": 0.0312, "step": 63310 }, { "grad_norm": 0.5363004207611084, "learning_rate": 3.249031927633742e-05, "loss": 0.0302, "step": 63320 }, { "grad_norm": 0.27999016642570496, "learning_rate": 3.2474832556990737e-05, "loss": 0.0291, "step": 63330 }, { "grad_norm": 0.5148749351501465, "learning_rate": 3.245934775417002e-05, "loss": 0.0334, "step": 63340 }, { "grad_norm": 0.4123685657978058, "learning_rate": 3.244386486956866e-05, "loss": 0.034, "step": 63350 }, { "grad_norm": 0.3355613052845001, "learning_rate": 3.2428383904879846e-05, "loss": 0.0316, "step": 63360 }, { "grad_norm": 0.38784554600715637, "learning_rate": 3.241290486179652e-05, "loss": 0.0278, "step": 63370 }, { "grad_norm": 0.5500847101211548, "learning_rate": 3.239742774201151e-05, "loss": 0.0281, "step": 63380 }, { "grad_norm": 0.29830360412597656, "learning_rate": 3.238195254721733e-05, "loss": 0.029, "step": 63390 }, { "grad_norm": 0.47761961817741394, "learning_rate": 3.236647927910633e-05, "loss": 0.0314, "step": 63400 }, { "grad_norm": 0.3484663665294647, "learning_rate": 3.2351007939370646e-05, "loss": 0.031, "step": 63410 }, { "grad_norm": 0.4638134837150574, "learning_rate": 3.2335538529702184e-05, "loss": 0.0295, "step": 63420 }, { "grad_norm": 0.38283923268318176, "learning_rate": 3.232007105179268e-05, "loss": 0.0284, "step": 63430 }, { "grad_norm": 0.43362724781036377, "learning_rate": 3.230460550733363e-05, "loss": 0.0261, "step": 63440 }, { "grad_norm": 0.3830740749835968, "learning_rate": 3.2289141898016315e-05, "loss": 0.0275, "step": 63450 }, { "grad_norm": 0.4041467010974884, "learning_rate": 3.227368022553181e-05, "loss": 0.0286, "step": 63460 }, { "grad_norm": 0.542104184627533, "learning_rate": 3.225822049157099e-05, "loss": 0.0259, "step": 63470 }, { "grad_norm": 0.35112929344177246, "learning_rate": 3.2242762697824494e-05, "loss": 0.0267, "step": 63480 }, { "grad_norm": 0.22511044144630432, "learning_rate": 3.222730684598279e-05, "loss": 0.0279, "step": 63490 }, { "grad_norm": 0.3541194200515747, "learning_rate": 3.221185293773605e-05, "loss": 0.0318, "step": 63500 }, { "grad_norm": 0.473818838596344, "learning_rate": 3.2196400974774354e-05, "loss": 0.0271, "step": 63510 }, { "grad_norm": 0.38777869939804077, "learning_rate": 3.2180950958787485e-05, "loss": 0.0282, "step": 63520 }, { "grad_norm": 0.4152892231941223, "learning_rate": 3.2165502891465024e-05, "loss": 0.0267, "step": 63530 }, { "grad_norm": 0.44891098141670227, "learning_rate": 3.215005677449636e-05, "loss": 0.0291, "step": 63540 }, { "grad_norm": 0.28291046619415283, "learning_rate": 3.2134612609570644e-05, "loss": 0.0274, "step": 63550 }, { "grad_norm": 0.6330392360687256, "learning_rate": 3.211917039837682e-05, "loss": 0.0271, "step": 63560 }, { "grad_norm": 0.3327805995941162, "learning_rate": 3.210373014260365e-05, "loss": 0.0257, "step": 63570 }, { "grad_norm": 0.40178823471069336, "learning_rate": 3.208829184393965e-05, "loss": 0.0293, "step": 63580 }, { "grad_norm": 0.43156003952026367, "learning_rate": 3.207285550407312e-05, "loss": 0.0252, "step": 63590 }, { "grad_norm": 0.26027050614356995, "learning_rate": 3.205742112469217e-05, "loss": 0.0279, "step": 63600 }, { "grad_norm": 0.35640454292297363, "learning_rate": 3.204198870748465e-05, "loss": 0.0301, "step": 63610 }, { "grad_norm": 0.18392746150493622, "learning_rate": 3.202655825413824e-05, "loss": 0.0238, "step": 63620 }, { "grad_norm": 0.41848066449165344, "learning_rate": 3.2011129766340416e-05, "loss": 0.0302, "step": 63630 }, { "grad_norm": 0.3090815842151642, "learning_rate": 3.1995703245778384e-05, "loss": 0.0274, "step": 63640 }, { "grad_norm": 0.2638150751590729, "learning_rate": 3.1980278694139186e-05, "loss": 0.0322, "step": 63650 }, { "grad_norm": 0.3323535621166229, "learning_rate": 3.196485611310962e-05, "loss": 0.0223, "step": 63660 }, { "grad_norm": 0.8787282705307007, "learning_rate": 3.1949435504376257e-05, "loss": 0.0287, "step": 63670 }, { "grad_norm": 0.46782436966896057, "learning_rate": 3.1934016869625505e-05, "loss": 0.0309, "step": 63680 }, { "grad_norm": 0.24783384799957275, "learning_rate": 3.191860021054349e-05, "loss": 0.0274, "step": 63690 }, { "grad_norm": 0.4542105495929718, "learning_rate": 3.190318552881618e-05, "loss": 0.03, "step": 63700 }, { "grad_norm": 0.6155300736427307, "learning_rate": 3.1887772826129285e-05, "loss": 0.0285, "step": 63710 }, { "grad_norm": 0.2969760596752167, "learning_rate": 3.1872362104168335e-05, "loss": 0.0305, "step": 63720 }, { "grad_norm": 0.48172834515571594, "learning_rate": 3.185695336461861e-05, "loss": 0.0259, "step": 63730 }, { "grad_norm": 0.2473832368850708, "learning_rate": 3.184154660916519e-05, "loss": 0.0266, "step": 63740 }, { "grad_norm": 0.30713313817977905, "learning_rate": 3.1826141839492906e-05, "loss": 0.0255, "step": 63750 }, { "grad_norm": 0.31288155913352966, "learning_rate": 3.181073905728644e-05, "loss": 0.0325, "step": 63760 }, { "grad_norm": 0.24592292308807373, "learning_rate": 3.1795338264230214e-05, "loss": 0.0255, "step": 63770 }, { "grad_norm": 0.3238982856273651, "learning_rate": 3.1779939462008426e-05, "loss": 0.0267, "step": 63780 }, { "grad_norm": 0.38657695055007935, "learning_rate": 3.1764542652305066e-05, "loss": 0.031, "step": 63790 }, { "grad_norm": 0.6273025274276733, "learning_rate": 3.17491478368039e-05, "loss": 0.0284, "step": 63800 }, { "grad_norm": 0.3360389173030853, "learning_rate": 3.173375501718847e-05, "loss": 0.0271, "step": 63810 }, { "grad_norm": 0.2834327816963196, "learning_rate": 3.171836419514215e-05, "loss": 0.0312, "step": 63820 }, { "grad_norm": 0.2507680058479309, "learning_rate": 3.170297537234803e-05, "loss": 0.0245, "step": 63830 }, { "grad_norm": 0.3482278287410736, "learning_rate": 3.168758855048901e-05, "loss": 0.0243, "step": 63840 }, { "grad_norm": 0.4987611770629883, "learning_rate": 3.1672203731247785e-05, "loss": 0.0302, "step": 63850 }, { "grad_norm": 0.287418395280838, "learning_rate": 3.16568209163068e-05, "loss": 0.0299, "step": 63860 }, { "grad_norm": 0.32943156361579895, "learning_rate": 3.1641440107348295e-05, "loss": 0.0309, "step": 63870 }, { "grad_norm": 0.3644442856311798, "learning_rate": 3.162606130605429e-05, "loss": 0.0328, "step": 63880 }, { "grad_norm": 0.7817613482475281, "learning_rate": 3.16106845141066e-05, "loss": 0.0289, "step": 63890 }, { "grad_norm": 0.9012225270271301, "learning_rate": 3.159530973318682e-05, "loss": 0.0314, "step": 63900 }, { "grad_norm": 0.4051879346370697, "learning_rate": 3.1579936964976286e-05, "loss": 0.0357, "step": 63910 }, { "grad_norm": 0.3561503291130066, "learning_rate": 3.156456621115615e-05, "loss": 0.027, "step": 63920 }, { "grad_norm": 0.4293330907821655, "learning_rate": 3.154919747340734e-05, "loss": 0.026, "step": 63930 }, { "grad_norm": 0.37582242488861084, "learning_rate": 3.153383075341054e-05, "loss": 0.0255, "step": 63940 }, { "grad_norm": 0.5114501714706421, "learning_rate": 3.1518466052846264e-05, "loss": 0.0298, "step": 63950 }, { "grad_norm": 0.5831794738769531, "learning_rate": 3.150310337339474e-05, "loss": 0.0297, "step": 63960 }, { "grad_norm": 0.3328409194946289, "learning_rate": 3.148774271673603e-05, "loss": 0.0243, "step": 63970 }, { "grad_norm": 0.43030381202697754, "learning_rate": 3.147238408454995e-05, "loss": 0.0226, "step": 63980 }, { "grad_norm": 0.3608616590499878, "learning_rate": 3.1457027478516086e-05, "loss": 0.026, "step": 63990 }, { "grad_norm": 0.36382368206977844, "learning_rate": 3.144167290031382e-05, "loss": 0.0294, "step": 64000 }, { "grad_norm": 0.5437451004981995, "learning_rate": 3.142632035162227e-05, "loss": 0.0291, "step": 64010 }, { "grad_norm": 0.5993497371673584, "learning_rate": 3.141096983412044e-05, "loss": 0.0293, "step": 64020 }, { "grad_norm": 0.24265745282173157, "learning_rate": 3.139562134948699e-05, "loss": 0.0291, "step": 64030 }, { "grad_norm": 0.3534071743488312, "learning_rate": 3.1380274899400424e-05, "loss": 0.027, "step": 64040 }, { "grad_norm": 0.3228525221347809, "learning_rate": 3.1364930485539e-05, "loss": 0.0255, "step": 64050 }, { "grad_norm": 0.33871158957481384, "learning_rate": 3.134958810958076e-05, "loss": 0.0269, "step": 64060 }, { "grad_norm": 0.48637425899505615, "learning_rate": 3.13342477732035e-05, "loss": 0.0313, "step": 64070 }, { "grad_norm": 0.29544636607170105, "learning_rate": 3.131890947808487e-05, "loss": 0.0266, "step": 64080 }, { "grad_norm": 0.46663084626197815, "learning_rate": 3.13035732259022e-05, "loss": 0.0277, "step": 64090 }, { "grad_norm": 0.42029163241386414, "learning_rate": 3.1288239018332656e-05, "loss": 0.0203, "step": 64100 }, { "grad_norm": 0.23924539983272552, "learning_rate": 3.1272906857053164e-05, "loss": 0.0238, "step": 64110 }, { "grad_norm": 0.373434454202652, "learning_rate": 3.125757674374042e-05, "loss": 0.0283, "step": 64120 }, { "grad_norm": 0.3991251587867737, "learning_rate": 3.124224868007088e-05, "loss": 0.0249, "step": 64130 }, { "grad_norm": 0.5469040870666504, "learning_rate": 3.1226922667720854e-05, "loss": 0.0238, "step": 64140 }, { "grad_norm": 0.4143456220626831, "learning_rate": 3.121159870836633e-05, "loss": 0.0269, "step": 64150 }, { "grad_norm": 0.4607572555541992, "learning_rate": 3.119627680368312e-05, "loss": 0.0265, "step": 64160 }, { "grad_norm": 0.3356870114803314, "learning_rate": 3.1180956955346816e-05, "loss": 0.029, "step": 64170 }, { "grad_norm": 0.25199922919273376, "learning_rate": 3.116563916503276e-05, "loss": 0.0269, "step": 64180 }, { "grad_norm": 0.4799714684486389, "learning_rate": 3.1150323434416095e-05, "loss": 0.0286, "step": 64190 }, { "grad_norm": 0.7413994669914246, "learning_rate": 3.1135009765171696e-05, "loss": 0.0297, "step": 64200 }, { "grad_norm": 0.3488253951072693, "learning_rate": 3.111969815897429e-05, "loss": 0.0239, "step": 64210 }, { "grad_norm": 0.5937039852142334, "learning_rate": 3.1104388617498295e-05, "loss": 0.0244, "step": 64220 }, { "grad_norm": 0.33855199813842773, "learning_rate": 3.1089081142417966e-05, "loss": 0.0278, "step": 64230 }, { "grad_norm": 0.2668561041355133, "learning_rate": 3.1073775735407285e-05, "loss": 0.025, "step": 64240 }, { "grad_norm": 0.41792187094688416, "learning_rate": 3.1058472398140035e-05, "loss": 0.0336, "step": 64250 }, { "grad_norm": 0.29209238290786743, "learning_rate": 3.1043171132289736e-05, "loss": 0.026, "step": 64260 }, { "grad_norm": 0.34867802262306213, "learning_rate": 3.1027871939529774e-05, "loss": 0.0345, "step": 64270 }, { "grad_norm": 0.403315007686615, "learning_rate": 3.1012574821533214e-05, "loss": 0.0245, "step": 64280 }, { "grad_norm": 0.26681751012802124, "learning_rate": 3.099727977997291e-05, "loss": 0.0301, "step": 64290 }, { "grad_norm": 0.2780910134315491, "learning_rate": 3.098198681652154e-05, "loss": 0.0233, "step": 64300 }, { "grad_norm": 0.3074667453765869, "learning_rate": 3.0966695932851475e-05, "loss": 0.0296, "step": 64310 }, { "grad_norm": 0.3902347683906555, "learning_rate": 3.0951407130634926e-05, "loss": 0.0236, "step": 64320 }, { "grad_norm": 0.41149741411209106, "learning_rate": 3.0936120411543864e-05, "loss": 0.0242, "step": 64330 }, { "grad_norm": 0.45469391345977783, "learning_rate": 3.0920835777250015e-05, "loss": 0.0279, "step": 64340 }, { "grad_norm": 0.3107496500015259, "learning_rate": 3.090555322942488e-05, "loss": 0.0278, "step": 64350 }, { "grad_norm": 0.3902626931667328, "learning_rate": 3.089027276973973e-05, "loss": 0.0281, "step": 64360 }, { "grad_norm": 0.35719507932662964, "learning_rate": 3.0874994399865615e-05, "loss": 0.0224, "step": 64370 }, { "grad_norm": 0.23171062767505646, "learning_rate": 3.085971812147337e-05, "loss": 0.0244, "step": 64380 }, { "grad_norm": 0.24417150020599365, "learning_rate": 3.0844443936233557e-05, "loss": 0.0291, "step": 64390 }, { "grad_norm": 0.3919192850589752, "learning_rate": 3.082917184581657e-05, "loss": 0.0246, "step": 64400 }, { "grad_norm": 0.38391250371932983, "learning_rate": 3.0813901851892513e-05, "loss": 0.034, "step": 64410 }, { "grad_norm": 0.4474771022796631, "learning_rate": 3.0798633956131316e-05, "loss": 0.031, "step": 64420 }, { "grad_norm": 0.3681749403476715, "learning_rate": 3.078336816020264e-05, "loss": 0.027, "step": 64430 }, { "grad_norm": 0.30324116349220276, "learning_rate": 3.0768104465775935e-05, "loss": 0.0229, "step": 64440 }, { "grad_norm": 0.20695914328098297, "learning_rate": 3.0752842874520396e-05, "loss": 0.0266, "step": 64450 }, { "grad_norm": 0.29151275753974915, "learning_rate": 3.073758338810503e-05, "loss": 0.0297, "step": 64460 }, { "grad_norm": 0.2304506003856659, "learning_rate": 3.072232600819859e-05, "loss": 0.0238, "step": 64470 }, { "grad_norm": 0.276588499546051, "learning_rate": 3.070707073646959e-05, "loss": 0.0283, "step": 64480 }, { "grad_norm": 0.2949449121952057, "learning_rate": 3.069181757458633e-05, "loss": 0.0239, "step": 64490 }, { "grad_norm": 0.31990471482276917, "learning_rate": 3.067656652421687e-05, "loss": 0.0261, "step": 64500 }, { "grad_norm": 0.3308354318141937, "learning_rate": 3.0661317587029025e-05, "loss": 0.0344, "step": 64510 }, { "grad_norm": 0.25176119804382324, "learning_rate": 3.064607076469043e-05, "loss": 0.0267, "step": 64520 }, { "grad_norm": 0.30542489886283875, "learning_rate": 3.063082605886845e-05, "loss": 0.0257, "step": 64530 }, { "grad_norm": 0.23022307455539703, "learning_rate": 3.061558347123019e-05, "loss": 0.026, "step": 64540 }, { "grad_norm": 0.28159040212631226, "learning_rate": 3.060034300344259e-05, "loss": 0.0304, "step": 64550 }, { "grad_norm": 0.4616406559944153, "learning_rate": 3.058510465717231e-05, "loss": 0.0321, "step": 64560 }, { "grad_norm": 0.49544909596443176, "learning_rate": 3.0569868434085804e-05, "loss": 0.0286, "step": 64570 }, { "grad_norm": 0.33107322454452515, "learning_rate": 3.055463433584924e-05, "loss": 0.028, "step": 64580 }, { "grad_norm": 0.37802961468696594, "learning_rate": 3.0539402364128656e-05, "loss": 0.0329, "step": 64590 }, { "grad_norm": 0.5762015581130981, "learning_rate": 3.052417252058977e-05, "loss": 0.0303, "step": 64600 }, { "grad_norm": 0.3329995572566986, "learning_rate": 3.0508944806898092e-05, "loss": 0.0272, "step": 64610 }, { "grad_norm": 0.36493921279907227, "learning_rate": 3.04937192247189e-05, "loss": 0.0297, "step": 64620 }, { "grad_norm": 0.21502043306827545, "learning_rate": 3.047849577571725e-05, "loss": 0.0246, "step": 64630 }, { "grad_norm": 0.44731542468070984, "learning_rate": 3.046327446155795e-05, "loss": 0.0274, "step": 64640 }, { "grad_norm": 0.48150911927223206, "learning_rate": 3.0448055283905587e-05, "loss": 0.0251, "step": 64650 }, { "grad_norm": 0.3767548203468323, "learning_rate": 3.0432838244424494e-05, "loss": 0.0262, "step": 64660 }, { "grad_norm": 0.28504306077957153, "learning_rate": 3.04176233447788e-05, "loss": 0.0299, "step": 64670 }, { "grad_norm": 0.34447532892227173, "learning_rate": 3.040241058663238e-05, "loss": 0.0281, "step": 64680 }, { "grad_norm": 0.2880516052246094, "learning_rate": 3.038719997164887e-05, "loss": 0.0277, "step": 64690 }, { "grad_norm": 0.4443417489528656, "learning_rate": 3.0371991501491683e-05, "loss": 0.0268, "step": 64700 }, { "grad_norm": 0.2954839766025543, "learning_rate": 3.0356785177823975e-05, "loss": 0.0266, "step": 64710 }, { "grad_norm": 0.3330495059490204, "learning_rate": 3.0341581002308726e-05, "loss": 0.0264, "step": 64720 }, { "grad_norm": 0.2807554006576538, "learning_rate": 3.0326378976608627e-05, "loss": 0.0278, "step": 64730 }, { "grad_norm": 0.35466304421424866, "learning_rate": 3.0311179102386146e-05, "loss": 0.0292, "step": 64740 }, { "grad_norm": 0.28372102975845337, "learning_rate": 3.0295981381303513e-05, "loss": 0.0215, "step": 64750 }, { "grad_norm": 0.4131467938423157, "learning_rate": 3.0280785815022727e-05, "loss": 0.0239, "step": 64760 }, { "grad_norm": 0.4375881254673004, "learning_rate": 3.0265592405205544e-05, "loss": 0.0326, "step": 64770 }, { "grad_norm": 0.4146098494529724, "learning_rate": 3.0250401153513517e-05, "loss": 0.028, "step": 64780 }, { "grad_norm": 0.3376372456550598, "learning_rate": 3.0235212061607936e-05, "loss": 0.0293, "step": 64790 }, { "grad_norm": 0.5978971123695374, "learning_rate": 3.0220025131149842e-05, "loss": 0.0313, "step": 64800 }, { "grad_norm": 0.3156045973300934, "learning_rate": 3.020484036380007e-05, "loss": 0.0318, "step": 64810 }, { "grad_norm": 0.3229987323284149, "learning_rate": 3.0189657761219188e-05, "loss": 0.0289, "step": 64820 }, { "grad_norm": 0.37996581196784973, "learning_rate": 3.017447732506753e-05, "loss": 0.0297, "step": 64830 }, { "grad_norm": 0.43962809443473816, "learning_rate": 3.015929905700524e-05, "loss": 0.034, "step": 64840 }, { "grad_norm": 0.27465492486953735, "learning_rate": 3.014412295869218e-05, "loss": 0.0257, "step": 64850 }, { "grad_norm": 0.30715006589889526, "learning_rate": 3.0128949031787977e-05, "loss": 0.0286, "step": 64860 }, { "grad_norm": 0.4380781650543213, "learning_rate": 3.0113777277952022e-05, "loss": 0.0319, "step": 64870 }, { "grad_norm": 0.24816358089447021, "learning_rate": 3.0098607698843496e-05, "loss": 0.0258, "step": 64880 }, { "grad_norm": 0.2849748432636261, "learning_rate": 3.0083440296121312e-05, "loss": 0.0341, "step": 64890 }, { "grad_norm": 0.29823705554008484, "learning_rate": 3.0068275071444142e-05, "loss": 0.0222, "step": 64900 }, { "grad_norm": 0.49041837453842163, "learning_rate": 3.0053112026470442e-05, "loss": 0.0325, "step": 64910 }, { "grad_norm": 1.1774983406066895, "learning_rate": 3.0037951162858434e-05, "loss": 0.029, "step": 64920 }, { "grad_norm": 0.33860480785369873, "learning_rate": 3.002279248226607e-05, "loss": 0.0259, "step": 64930 }, { "grad_norm": 0.3584882915019989, "learning_rate": 3.000763598635109e-05, "loss": 0.0317, "step": 64940 }, { "grad_norm": 0.31091776490211487, "learning_rate": 2.999248167677098e-05, "loss": 0.03, "step": 64950 }, { "grad_norm": 0.2765946388244629, "learning_rate": 2.9977329555182976e-05, "loss": 0.0291, "step": 64960 }, { "grad_norm": 0.42580434679985046, "learning_rate": 2.9962179623244124e-05, "loss": 0.0241, "step": 64970 }, { "grad_norm": 0.4104701578617096, "learning_rate": 2.9947031882611175e-05, "loss": 0.0316, "step": 64980 }, { "grad_norm": 0.3697282373905182, "learning_rate": 2.9931886334940685e-05, "loss": 0.0287, "step": 64990 }, { "grad_norm": 0.3108045160770416, "learning_rate": 2.9916742981888923e-05, "loss": 0.0288, "step": 65000 }, { "grad_norm": 0.2696591913700104, "learning_rate": 2.990160182511196e-05, "loss": 0.0274, "step": 65010 }, { "grad_norm": 0.3034100830554962, "learning_rate": 2.9886462866265574e-05, "loss": 0.0262, "step": 65020 }, { "grad_norm": 0.3860147297382355, "learning_rate": 2.9871326107005392e-05, "loss": 0.0284, "step": 65030 }, { "grad_norm": 0.5307589769363403, "learning_rate": 2.9856191548986735e-05, "loss": 0.0295, "step": 65040 }, { "grad_norm": 0.2926582992076874, "learning_rate": 2.9841059193864672e-05, "loss": 0.0277, "step": 65050 }, { "grad_norm": 0.4576801359653473, "learning_rate": 2.982592904329407e-05, "loss": 0.0234, "step": 65060 }, { "grad_norm": 0.41720300912857056, "learning_rate": 2.981080109892954e-05, "loss": 0.0313, "step": 65070 }, { "grad_norm": 0.3880479633808136, "learning_rate": 2.9795675362425434e-05, "loss": 0.0294, "step": 65080 }, { "grad_norm": 0.33788591623306274, "learning_rate": 2.97805518354359e-05, "loss": 0.0279, "step": 65090 }, { "grad_norm": 0.38888561725616455, "learning_rate": 2.9765430519614825e-05, "loss": 0.0297, "step": 65100 }, { "grad_norm": 0.3321823179721832, "learning_rate": 2.975031141661584e-05, "loss": 0.0265, "step": 65110 }, { "grad_norm": 0.24962779879570007, "learning_rate": 2.9735194528092347e-05, "loss": 0.0248, "step": 65120 }, { "grad_norm": 0.252289742231369, "learning_rate": 2.9720079855697525e-05, "loss": 0.028, "step": 65130 }, { "grad_norm": 0.4371703267097473, "learning_rate": 2.970496740108427e-05, "loss": 0.0268, "step": 65140 }, { "grad_norm": 0.28597792983055115, "learning_rate": 2.968985716590527e-05, "loss": 0.0281, "step": 65150 }, { "grad_norm": 0.36514806747436523, "learning_rate": 2.9674749151812953e-05, "loss": 0.0302, "step": 65160 }, { "grad_norm": 0.26420873403549194, "learning_rate": 2.965964336045951e-05, "loss": 0.0263, "step": 65170 }, { "grad_norm": 0.33109331130981445, "learning_rate": 2.9644539793496894e-05, "loss": 0.0346, "step": 65180 }, { "grad_norm": 0.3569492697715759, "learning_rate": 2.9629438452576807e-05, "loss": 0.0296, "step": 65190 }, { "grad_norm": 0.3415340185165405, "learning_rate": 2.9614339339350698e-05, "loss": 0.0257, "step": 65200 }, { "grad_norm": 0.2498738318681717, "learning_rate": 2.959924245546979e-05, "loss": 0.0233, "step": 65210 }, { "grad_norm": 0.3828113377094269, "learning_rate": 2.9584147802585038e-05, "loss": 0.0265, "step": 65220 }, { "grad_norm": 0.23693180084228516, "learning_rate": 2.956905538234721e-05, "loss": 0.0258, "step": 65230 }, { "grad_norm": 0.38977786898612976, "learning_rate": 2.9553965196406764e-05, "loss": 0.0251, "step": 65240 }, { "grad_norm": 0.435730516910553, "learning_rate": 2.9538877246413943e-05, "loss": 0.0282, "step": 65250 }, { "grad_norm": 0.24789117276668549, "learning_rate": 2.952379153401874e-05, "loss": 0.0302, "step": 65260 }, { "grad_norm": 0.32749778032302856, "learning_rate": 2.9508708060870914e-05, "loss": 0.023, "step": 65270 }, { "grad_norm": 0.34531164169311523, "learning_rate": 2.9493626828619948e-05, "loss": 0.0294, "step": 65280 }, { "grad_norm": 0.3263886868953705, "learning_rate": 2.9478547838915137e-05, "loss": 0.021, "step": 65290 }, { "grad_norm": 0.3730127513408661, "learning_rate": 2.946347109340548e-05, "loss": 0.0279, "step": 65300 }, { "grad_norm": 0.9853928089141846, "learning_rate": 2.944839659373976e-05, "loss": 0.0255, "step": 65310 }, { "grad_norm": 0.3206762969493866, "learning_rate": 2.943332434156647e-05, "loss": 0.0286, "step": 65320 }, { "grad_norm": 0.46781685948371887, "learning_rate": 2.9418254338533912e-05, "loss": 0.0312, "step": 65330 }, { "grad_norm": 0.3223958909511566, "learning_rate": 2.940318658629011e-05, "loss": 0.028, "step": 65340 }, { "grad_norm": 0.47608762979507446, "learning_rate": 2.938812108648287e-05, "loss": 0.0313, "step": 65350 }, { "grad_norm": 0.39886894822120667, "learning_rate": 2.937305784075972e-05, "loss": 0.0273, "step": 65360 }, { "grad_norm": 0.45078927278518677, "learning_rate": 2.9357996850767937e-05, "loss": 0.0277, "step": 65370 }, { "grad_norm": 0.3412940204143524, "learning_rate": 2.9342938118154595e-05, "loss": 0.0234, "step": 65380 }, { "grad_norm": 0.2773967385292053, "learning_rate": 2.9327881644566485e-05, "loss": 0.0236, "step": 65390 }, { "grad_norm": 0.31200048327445984, "learning_rate": 2.9312827431650157e-05, "loss": 0.0296, "step": 65400 }, { "grad_norm": 0.22816266119480133, "learning_rate": 2.92977754810519e-05, "loss": 0.0212, "step": 65410 }, { "grad_norm": 0.3448956310749054, "learning_rate": 2.928272579441781e-05, "loss": 0.0262, "step": 65420 }, { "grad_norm": 0.19407156109809875, "learning_rate": 2.9267678373393676e-05, "loss": 0.0264, "step": 65430 }, { "grad_norm": 0.5597550868988037, "learning_rate": 2.925263321962507e-05, "loss": 0.0246, "step": 65440 }, { "grad_norm": 0.2726023495197296, "learning_rate": 2.9237590334757292e-05, "loss": 0.0259, "step": 65450 }, { "grad_norm": 0.3763737380504608, "learning_rate": 2.922254972043543e-05, "loss": 0.0288, "step": 65460 }, { "grad_norm": 0.2866254448890686, "learning_rate": 2.9207511378304265e-05, "loss": 0.0235, "step": 65470 }, { "grad_norm": 0.6305656433105469, "learning_rate": 2.9192475310008428e-05, "loss": 0.0222, "step": 65480 }, { "grad_norm": 0.41605931520462036, "learning_rate": 2.917744151719219e-05, "loss": 0.0314, "step": 65490 }, { "grad_norm": 0.26738405227661133, "learning_rate": 2.916241000149964e-05, "loss": 0.0235, "step": 65500 }, { "grad_norm": 0.3609100878238678, "learning_rate": 2.9147380764574622e-05, "loss": 0.0277, "step": 65510 }, { "grad_norm": 0.35667309165000916, "learning_rate": 2.9132353808060674e-05, "loss": 0.0251, "step": 65520 }, { "grad_norm": 0.28520211577415466, "learning_rate": 2.9117329133601134e-05, "loss": 0.0288, "step": 65530 }, { "grad_norm": 0.3263363540172577, "learning_rate": 2.9102306742839093e-05, "loss": 0.0237, "step": 65540 }, { "grad_norm": 1.348644495010376, "learning_rate": 2.9087286637417375e-05, "loss": 0.0296, "step": 65550 }, { "grad_norm": 0.3142498731613159, "learning_rate": 2.9072268818978537e-05, "loss": 0.0225, "step": 65560 }, { "grad_norm": 0.4111466705799103, "learning_rate": 2.905725328916493e-05, "loss": 0.0231, "step": 65570 }, { "grad_norm": 0.3690522313117981, "learning_rate": 2.9042240049618608e-05, "loss": 0.0252, "step": 65580 }, { "grad_norm": 0.6176183819770813, "learning_rate": 2.9027229101981413e-05, "loss": 0.0279, "step": 65590 }, { "grad_norm": 0.27984386682510376, "learning_rate": 2.901222044789488e-05, "loss": 0.0288, "step": 65600 }, { "grad_norm": 0.30025848746299744, "learning_rate": 2.89972140890004e-05, "loss": 0.0226, "step": 65610 }, { "grad_norm": 0.6043758392333984, "learning_rate": 2.8982210026938988e-05, "loss": 0.0324, "step": 65620 }, { "grad_norm": 0.22551438212394714, "learning_rate": 2.896720826335151e-05, "loss": 0.0227, "step": 65630 }, { "grad_norm": 0.24474693834781647, "learning_rate": 2.8952208799878493e-05, "loss": 0.0214, "step": 65640 }, { "grad_norm": 0.30292221903800964, "learning_rate": 2.89372116381603e-05, "loss": 0.0275, "step": 65650 }, { "grad_norm": 0.46440577507019043, "learning_rate": 2.8922216779836932e-05, "loss": 0.0289, "step": 65660 }, { "grad_norm": 0.36057814955711365, "learning_rate": 2.8907224226548285e-05, "loss": 0.0269, "step": 65670 }, { "grad_norm": 0.37483277916908264, "learning_rate": 2.8892233979933868e-05, "loss": 0.0235, "step": 65680 }, { "grad_norm": 0.4679096043109894, "learning_rate": 2.8877246041633026e-05, "loss": 0.0284, "step": 65690 }, { "grad_norm": 0.31965911388397217, "learning_rate": 2.886226041328478e-05, "loss": 0.0246, "step": 65700 }, { "grad_norm": 0.24888961017131805, "learning_rate": 2.8847277096527957e-05, "loss": 0.0288, "step": 65710 }, { "grad_norm": 0.3768364489078522, "learning_rate": 2.8832296093001105e-05, "loss": 0.0279, "step": 65720 }, { "grad_norm": 0.3300226926803589, "learning_rate": 2.8817317404342547e-05, "loss": 0.0314, "step": 65730 }, { "grad_norm": 0.7502053380012512, "learning_rate": 2.880234103219029e-05, "loss": 0.0287, "step": 65740 }, { "grad_norm": 0.5634337067604065, "learning_rate": 2.878736697818215e-05, "loss": 0.0321, "step": 65750 }, { "grad_norm": 0.2873249351978302, "learning_rate": 2.877239524395568e-05, "loss": 0.0267, "step": 65760 }, { "grad_norm": 0.2536914348602295, "learning_rate": 2.8757425831148134e-05, "loss": 0.0237, "step": 65770 }, { "grad_norm": 0.23281247913837433, "learning_rate": 2.874245874139658e-05, "loss": 0.0243, "step": 65780 }, { "grad_norm": 0.3681158423423767, "learning_rate": 2.8727493976337737e-05, "loss": 0.0268, "step": 65790 }, { "grad_norm": 0.31854960322380066, "learning_rate": 2.8712531537608218e-05, "loss": 0.0249, "step": 65800 }, { "grad_norm": 0.24983660876750946, "learning_rate": 2.8697571426844217e-05, "loss": 0.0241, "step": 65810 }, { "grad_norm": 0.4730391800403595, "learning_rate": 2.86826136456818e-05, "loss": 0.0253, "step": 65820 }, { "grad_norm": 0.562497079372406, "learning_rate": 2.8667658195756685e-05, "loss": 0.0243, "step": 65830 }, { "grad_norm": 0.30586183071136475, "learning_rate": 2.8652705078704412e-05, "loss": 0.0259, "step": 65840 }, { "grad_norm": 0.3277687132358551, "learning_rate": 2.863775429616018e-05, "loss": 0.0264, "step": 65850 }, { "grad_norm": 0.3319133520126343, "learning_rate": 2.862280584975906e-05, "loss": 0.0274, "step": 65860 }, { "grad_norm": 0.26267877221107483, "learning_rate": 2.8607859741135735e-05, "loss": 0.0274, "step": 65870 }, { "grad_norm": 0.4159221649169922, "learning_rate": 2.8592915971924723e-05, "loss": 0.0312, "step": 65880 }, { "grad_norm": 0.409054696559906, "learning_rate": 2.8577974543760212e-05, "loss": 0.0307, "step": 65890 }, { "grad_norm": 0.418886661529541, "learning_rate": 2.8563035458276222e-05, "loss": 0.0309, "step": 65900 }, { "grad_norm": 0.26092004776000977, "learning_rate": 2.8548098717106424e-05, "loss": 0.0251, "step": 65910 }, { "grad_norm": 0.47520941495895386, "learning_rate": 2.85331643218843e-05, "loss": 0.0273, "step": 65920 }, { "grad_norm": 0.3928074836730957, "learning_rate": 2.8518232274243046e-05, "loss": 0.0253, "step": 65930 }, { "grad_norm": 1.4564250707626343, "learning_rate": 2.8503302575815638e-05, "loss": 0.0307, "step": 65940 }, { "grad_norm": 0.3433646261692047, "learning_rate": 2.8488375228234716e-05, "loss": 0.0273, "step": 65950 }, { "grad_norm": 0.35443830490112305, "learning_rate": 2.8473450233132738e-05, "loss": 0.0237, "step": 65960 }, { "grad_norm": 0.5198174715042114, "learning_rate": 2.8458527592141898e-05, "loss": 0.0288, "step": 65970 }, { "grad_norm": 0.318482369184494, "learning_rate": 2.8443607306894073e-05, "loss": 0.0267, "step": 65980 }, { "grad_norm": 0.40526366233825684, "learning_rate": 2.8428689379020946e-05, "loss": 0.0276, "step": 65990 }, { "grad_norm": 0.3832627236843109, "learning_rate": 2.841377381015391e-05, "loss": 0.0269, "step": 66000 }, { "grad_norm": 0.3588263988494873, "learning_rate": 2.8398860601924144e-05, "loss": 0.0241, "step": 66010 }, { "grad_norm": 0.6268190145492554, "learning_rate": 2.838394975596248e-05, "loss": 0.0289, "step": 66020 }, { "grad_norm": 0.39638423919677734, "learning_rate": 2.83690412738996e-05, "loss": 0.0279, "step": 66030 }, { "grad_norm": 0.27951520681381226, "learning_rate": 2.835413515736579e-05, "loss": 0.0276, "step": 66040 }, { "grad_norm": 0.2189832180738449, "learning_rate": 2.8339231407991267e-05, "loss": 0.0314, "step": 66050 }, { "grad_norm": 0.26172736287117004, "learning_rate": 2.832433002740581e-05, "loss": 0.0222, "step": 66060 }, { "grad_norm": 0.3073519766330719, "learning_rate": 2.8309431017239047e-05, "loss": 0.0259, "step": 66070 }, { "grad_norm": 0.25394171476364136, "learning_rate": 2.829453437912028e-05, "loss": 0.0261, "step": 66080 }, { "grad_norm": 0.3588811457157135, "learning_rate": 2.8279640114678617e-05, "loss": 0.03, "step": 66090 }, { "grad_norm": 0.3274032473564148, "learning_rate": 2.8264748225542836e-05, "loss": 0.0253, "step": 66100 }, { "grad_norm": 0.6108270883560181, "learning_rate": 2.8249858713341516e-05, "loss": 0.0319, "step": 66110 }, { "grad_norm": 0.5560659766197205, "learning_rate": 2.823497157970294e-05, "loss": 0.0228, "step": 66120 }, { "grad_norm": 0.3758290708065033, "learning_rate": 2.8220086826255164e-05, "loss": 0.0326, "step": 66130 }, { "grad_norm": 0.3719692826271057, "learning_rate": 2.8205204454625934e-05, "loss": 0.0267, "step": 66140 }, { "grad_norm": 0.43544426560401917, "learning_rate": 2.8190324466442798e-05, "loss": 0.0243, "step": 66150 }, { "grad_norm": 0.2635924220085144, "learning_rate": 2.8175446863332965e-05, "loss": 0.0324, "step": 66160 }, { "grad_norm": 0.3461681604385376, "learning_rate": 2.816057164692345e-05, "loss": 0.0281, "step": 66170 }, { "grad_norm": 0.3171575963497162, "learning_rate": 2.8145698818840992e-05, "loss": 0.0277, "step": 66180 }, { "grad_norm": 0.32354795932769775, "learning_rate": 2.8130828380712065e-05, "loss": 0.0256, "step": 66190 }, { "grad_norm": 0.3561362326145172, "learning_rate": 2.811596033416285e-05, "loss": 0.0269, "step": 66200 }, { "grad_norm": 0.49091723561286926, "learning_rate": 2.8101094680819307e-05, "loss": 0.0297, "step": 66210 }, { "grad_norm": 0.32295170426368713, "learning_rate": 2.808623142230715e-05, "loss": 0.021, "step": 66220 }, { "grad_norm": 0.4478400647640228, "learning_rate": 2.8071370560251757e-05, "loss": 0.0272, "step": 66230 }, { "grad_norm": 0.3137821555137634, "learning_rate": 2.805651209627831e-05, "loss": 0.0248, "step": 66240 }, { "grad_norm": 0.41576674580574036, "learning_rate": 2.804165603201171e-05, "loss": 0.0271, "step": 66250 }, { "grad_norm": 0.267104834318161, "learning_rate": 2.802680236907661e-05, "loss": 0.0242, "step": 66260 }, { "grad_norm": 0.2889954149723053, "learning_rate": 2.8011951109097344e-05, "loss": 0.0257, "step": 66270 }, { "grad_norm": 0.3546932637691498, "learning_rate": 2.7997102253698072e-05, "loss": 0.034, "step": 66280 }, { "grad_norm": 0.22015540301799774, "learning_rate": 2.7982255804502588e-05, "loss": 0.0271, "step": 66290 }, { "grad_norm": 0.5189133286476135, "learning_rate": 2.7967411763134516e-05, "loss": 0.0293, "step": 66300 }, { "grad_norm": 0.4447178840637207, "learning_rate": 2.7952570131217166e-05, "loss": 0.0241, "step": 66310 }, { "grad_norm": 0.3773128092288971, "learning_rate": 2.7937730910373618e-05, "loss": 0.0255, "step": 66320 }, { "grad_norm": 0.34129074215888977, "learning_rate": 2.7922894102226628e-05, "loss": 0.028, "step": 66330 }, { "grad_norm": 0.6222065091133118, "learning_rate": 2.790805970839876e-05, "loss": 0.0287, "step": 66340 }, { "grad_norm": 0.36431950330734253, "learning_rate": 2.789322773051225e-05, "loss": 0.0292, "step": 66350 }, { "grad_norm": 0.30058345198631287, "learning_rate": 2.7878398170189124e-05, "loss": 0.0332, "step": 66360 }, { "grad_norm": 0.5313395261764526, "learning_rate": 2.7863571029051112e-05, "loss": 0.0218, "step": 66370 }, { "grad_norm": 0.5758703947067261, "learning_rate": 2.784874630871971e-05, "loss": 0.0263, "step": 66380 }, { "grad_norm": 1.1323935985565186, "learning_rate": 2.7833924010816086e-05, "loss": 0.0215, "step": 66390 }, { "grad_norm": 0.35196974873542786, "learning_rate": 2.7819104136961226e-05, "loss": 0.0278, "step": 66400 }, { "grad_norm": 0.4239758849143982, "learning_rate": 2.780428668877577e-05, "loss": 0.0309, "step": 66410 }, { "grad_norm": 0.35645192861557007, "learning_rate": 2.7789471667880152e-05, "loss": 0.0314, "step": 66420 }, { "grad_norm": 0.2942177951335907, "learning_rate": 2.7774659075894523e-05, "loss": 0.0301, "step": 66430 }, { "grad_norm": 0.4163394868373871, "learning_rate": 2.7759848914438778e-05, "loss": 0.0234, "step": 66440 }, { "grad_norm": 0.2980126440525055, "learning_rate": 2.7745041185132496e-05, "loss": 0.0283, "step": 66450 }, { "grad_norm": 0.35236597061157227, "learning_rate": 2.773023588959506e-05, "loss": 0.0258, "step": 66460 }, { "grad_norm": 0.31634774804115295, "learning_rate": 2.7715433029445555e-05, "loss": 0.0304, "step": 66470 }, { "grad_norm": 0.18988794088363647, "learning_rate": 2.7700632606302766e-05, "loss": 0.023, "step": 66480 }, { "grad_norm": 0.4175565838813782, "learning_rate": 2.768583462178528e-05, "loss": 0.0244, "step": 66490 }, { "grad_norm": 0.261810302734375, "learning_rate": 2.767103907751136e-05, "loss": 0.0299, "step": 66500 }, { "grad_norm": 0.34324899315834045, "learning_rate": 2.7656245975099065e-05, "loss": 0.0337, "step": 66510 }, { "grad_norm": 0.39895254373550415, "learning_rate": 2.7641455316166093e-05, "loss": 0.0293, "step": 66520 }, { "grad_norm": 0.2507306933403015, "learning_rate": 2.7626667102329973e-05, "loss": 0.0224, "step": 66530 }, { "grad_norm": 0.22338245809078217, "learning_rate": 2.761188133520788e-05, "loss": 0.03, "step": 66540 }, { "grad_norm": 0.3172606825828552, "learning_rate": 2.7597098016416784e-05, "loss": 0.0291, "step": 66550 }, { "grad_norm": 0.3009982109069824, "learning_rate": 2.7582317147573368e-05, "loss": 0.028, "step": 66560 }, { "grad_norm": 0.33565518260002136, "learning_rate": 2.7567538730294064e-05, "loss": 0.0242, "step": 66570 }, { "grad_norm": 0.40357717871665955, "learning_rate": 2.7552762766194973e-05, "loss": 0.028, "step": 66580 }, { "grad_norm": 0.3933795392513275, "learning_rate": 2.753798925689202e-05, "loss": 0.0242, "step": 66590 }, { "grad_norm": 0.31731730699539185, "learning_rate": 2.7523218204000774e-05, "loss": 0.0263, "step": 66600 }, { "grad_norm": 0.1998940408229828, "learning_rate": 2.7508449609136612e-05, "loss": 0.0281, "step": 66610 }, { "grad_norm": 0.7140131592750549, "learning_rate": 2.749368347391454e-05, "loss": 0.0258, "step": 66620 }, { "grad_norm": 0.3116242289543152, "learning_rate": 2.7478919799949443e-05, "loss": 0.0262, "step": 66630 }, { "grad_norm": 0.3336976170539856, "learning_rate": 2.7464158588855803e-05, "loss": 0.0225, "step": 66640 }, { "grad_norm": 0.40685969591140747, "learning_rate": 2.7449399842247918e-05, "loss": 0.0333, "step": 66650 }, { "grad_norm": 0.4642961323261261, "learning_rate": 2.7434643561739736e-05, "loss": 0.0308, "step": 66660 }, { "grad_norm": 0.3102168142795563, "learning_rate": 2.7419889748945016e-05, "loss": 0.0279, "step": 66670 }, { "grad_norm": 0.28019699454307556, "learning_rate": 2.7405138405477203e-05, "loss": 0.0289, "step": 66680 }, { "grad_norm": 0.3019068241119385, "learning_rate": 2.73903895329495e-05, "loss": 0.0204, "step": 66690 }, { "grad_norm": 0.5282983183860779, "learning_rate": 2.737564313297478e-05, "loss": 0.0248, "step": 66700 }, { "grad_norm": 0.37310391664505005, "learning_rate": 2.736089920716572e-05, "loss": 0.0259, "step": 66710 }, { "grad_norm": 0.5144487023353577, "learning_rate": 2.7346157757134706e-05, "loss": 0.0243, "step": 66720 }, { "grad_norm": 0.3156072795391083, "learning_rate": 2.733141878449379e-05, "loss": 0.0258, "step": 66730 }, { "grad_norm": 0.23187804222106934, "learning_rate": 2.7316682290854844e-05, "loss": 0.0257, "step": 66740 }, { "grad_norm": 0.3516106903553009, "learning_rate": 2.7301948277829414e-05, "loss": 0.024, "step": 66750 }, { "grad_norm": 0.3160593807697296, "learning_rate": 2.7287216747028814e-05, "loss": 0.0254, "step": 66760 }, { "grad_norm": 0.23813368380069733, "learning_rate": 2.7272487700064025e-05, "loss": 0.0291, "step": 66770 }, { "grad_norm": 0.3886781334877014, "learning_rate": 2.725776113854583e-05, "loss": 0.0224, "step": 66780 }, { "grad_norm": 0.37313327193260193, "learning_rate": 2.7243037064084664e-05, "loss": 0.0257, "step": 66790 }, { "grad_norm": 0.5626474022865295, "learning_rate": 2.7228315478290766e-05, "loss": 0.0265, "step": 66800 }, { "grad_norm": 0.2678588926792145, "learning_rate": 2.721359638277402e-05, "loss": 0.0284, "step": 66810 }, { "grad_norm": 0.4124676585197449, "learning_rate": 2.7198879779144148e-05, "loss": 0.0309, "step": 66820 }, { "grad_norm": 0.24452194571495056, "learning_rate": 2.7184165669010485e-05, "loss": 0.0244, "step": 66830 }, { "grad_norm": 1.3175287246704102, "learning_rate": 2.716945405398218e-05, "loss": 0.0253, "step": 66840 }, { "grad_norm": 0.3002380430698395, "learning_rate": 2.715474493566803e-05, "loss": 0.0299, "step": 66850 }, { "grad_norm": 0.28852298855781555, "learning_rate": 2.7140038315676652e-05, "loss": 0.0256, "step": 66860 }, { "grad_norm": 0.22342048585414886, "learning_rate": 2.7125334195616275e-05, "loss": 0.027, "step": 66870 }, { "grad_norm": 0.4294407367706299, "learning_rate": 2.7110632577094997e-05, "loss": 0.0293, "step": 66880 }, { "grad_norm": 0.29811835289001465, "learning_rate": 2.709593346172051e-05, "loss": 0.024, "step": 66890 }, { "grad_norm": 1.79319167137146, "learning_rate": 2.708123685110032e-05, "loss": 0.0267, "step": 66900 }, { "grad_norm": 0.43461355566978455, "learning_rate": 2.7066542746841588e-05, "loss": 0.0264, "step": 66910 }, { "grad_norm": 0.3505176305770874, "learning_rate": 2.7051851150551267e-05, "loss": 0.0247, "step": 66920 }, { "grad_norm": 0.46663251519203186, "learning_rate": 2.7037162063835996e-05, "loss": 0.027, "step": 66930 }, { "grad_norm": 0.5846151113510132, "learning_rate": 2.702247548830218e-05, "loss": 0.03, "step": 66940 }, { "grad_norm": 0.3364624083042145, "learning_rate": 2.7007791425555883e-05, "loss": 0.03, "step": 66950 }, { "grad_norm": 0.39732012152671814, "learning_rate": 2.6993109877202945e-05, "loss": 0.0221, "step": 66960 }, { "grad_norm": 0.45536455512046814, "learning_rate": 2.6978430844848935e-05, "loss": 0.0241, "step": 66970 }, { "grad_norm": 0.30078190565109253, "learning_rate": 2.69637543300991e-05, "loss": 0.0237, "step": 66980 }, { "grad_norm": 0.5564455986022949, "learning_rate": 2.694908033455848e-05, "loss": 0.0244, "step": 66990 }, { "grad_norm": 0.29815343022346497, "learning_rate": 2.6934408859831733e-05, "loss": 0.0257, "step": 67000 }, { "grad_norm": 0.3360791504383087, "learning_rate": 2.69197399075234e-05, "loss": 0.027, "step": 67010 }, { "grad_norm": 0.3159322738647461, "learning_rate": 2.6905073479237584e-05, "loss": 0.0288, "step": 67020 }, { "grad_norm": 0.37325334548950195, "learning_rate": 2.6890409576578235e-05, "loss": 0.0234, "step": 67030 }, { "grad_norm": 0.373939573764801, "learning_rate": 2.687574820114892e-05, "loss": 0.0278, "step": 67040 }, { "grad_norm": 1.0804243087768555, "learning_rate": 2.686108935455305e-05, "loss": 0.0292, "step": 67050 }, { "grad_norm": 0.3372390866279602, "learning_rate": 2.6846433038393603e-05, "loss": 0.0283, "step": 67060 }, { "grad_norm": 0.39813661575317383, "learning_rate": 2.6831779254273482e-05, "loss": 0.023, "step": 67070 }, { "grad_norm": 0.21706216037273407, "learning_rate": 2.6817128003795122e-05, "loss": 0.0252, "step": 67080 }, { "grad_norm": 0.30255553126335144, "learning_rate": 2.680247928856081e-05, "loss": 0.026, "step": 67090 }, { "grad_norm": 0.40475985407829285, "learning_rate": 2.6787833110172466e-05, "loss": 0.0263, "step": 67100 }, { "grad_norm": 0.2714594304561615, "learning_rate": 2.6773189470231807e-05, "loss": 0.0251, "step": 67110 }, { "grad_norm": 0.30138564109802246, "learning_rate": 2.675854837034022e-05, "loss": 0.0287, "step": 67120 }, { "grad_norm": 0.32755833864212036, "learning_rate": 2.6743909812098833e-05, "loss": 0.0247, "step": 67130 }, { "grad_norm": 0.37671372294425964, "learning_rate": 2.67292737971085e-05, "loss": 0.0281, "step": 67140 }, { "grad_norm": 0.2691194713115692, "learning_rate": 2.671464032696982e-05, "loss": 0.026, "step": 67150 }, { "grad_norm": 0.25470080971717834, "learning_rate": 2.6700009403283055e-05, "loss": 0.0268, "step": 67160 }, { "grad_norm": 0.3908607065677643, "learning_rate": 2.6685381027648214e-05, "loss": 0.0265, "step": 67170 }, { "grad_norm": 0.4338429570198059, "learning_rate": 2.6670755201665086e-05, "loss": 0.0248, "step": 67180 }, { "grad_norm": 0.3425050675868988, "learning_rate": 2.6656131926933066e-05, "loss": 0.0239, "step": 67190 }, { "grad_norm": 0.38047873973846436, "learning_rate": 2.664151120505136e-05, "loss": 0.0264, "step": 67200 }, { "grad_norm": 0.3112944960594177, "learning_rate": 2.6626893037618873e-05, "loss": 0.022, "step": 67210 }, { "grad_norm": 0.5055158734321594, "learning_rate": 2.6612277426234243e-05, "loss": 0.0296, "step": 67220 }, { "grad_norm": 0.19542604684829712, "learning_rate": 2.659766437249577e-05, "loss": 0.0272, "step": 67230 }, { "grad_norm": 0.7584558725357056, "learning_rate": 2.6583053878001557e-05, "loss": 0.0278, "step": 67240 }, { "grad_norm": 0.5032864212989807, "learning_rate": 2.6568445944349323e-05, "loss": 0.0275, "step": 67250 }, { "grad_norm": 0.3559758961200714, "learning_rate": 2.6553840573136657e-05, "loss": 0.0281, "step": 67260 }, { "grad_norm": 0.3581940829753876, "learning_rate": 2.6539237765960706e-05, "loss": 0.0258, "step": 67270 }, { "grad_norm": 0.21649131178855896, "learning_rate": 2.6524637524418465e-05, "loss": 0.0211, "step": 67280 }, { "grad_norm": 0.3620903193950653, "learning_rate": 2.651003985010655e-05, "loss": 0.0251, "step": 67290 }, { "grad_norm": 0.35902994871139526, "learning_rate": 2.649544474462138e-05, "loss": 0.0261, "step": 67300 }, { "grad_norm": 0.43773189187049866, "learning_rate": 2.6480852209559015e-05, "loss": 0.0272, "step": 67310 }, { "grad_norm": 0.3358488380908966, "learning_rate": 2.6466262246515296e-05, "loss": 0.0306, "step": 67320 }, { "grad_norm": 0.36866074800491333, "learning_rate": 2.6451674857085746e-05, "loss": 0.0279, "step": 67330 }, { "grad_norm": 0.3590224087238312, "learning_rate": 2.6437090042865655e-05, "loss": 0.0259, "step": 67340 }, { "grad_norm": 0.44857367873191833, "learning_rate": 2.6422507805449947e-05, "loss": 0.024, "step": 67350 }, { "grad_norm": 0.4917750358581543, "learning_rate": 2.6407928146433357e-05, "loss": 0.0241, "step": 67360 }, { "grad_norm": 0.5506998896598816, "learning_rate": 2.639335106741026e-05, "loss": 0.0302, "step": 67370 }, { "grad_norm": 0.3731849491596222, "learning_rate": 2.6378776569974796e-05, "loss": 0.0235, "step": 67380 }, { "grad_norm": 0.5965118408203125, "learning_rate": 2.6364204655720814e-05, "loss": 0.024, "step": 67390 }, { "grad_norm": 1.0794486999511719, "learning_rate": 2.6349635326241894e-05, "loss": 0.0279, "step": 67400 }, { "grad_norm": 0.5981380343437195, "learning_rate": 2.633506858313127e-05, "loss": 0.0289, "step": 67410 }, { "grad_norm": 0.3052843511104584, "learning_rate": 2.6320504427981975e-05, "loss": 0.0266, "step": 67420 }, { "grad_norm": 0.38043978810310364, "learning_rate": 2.6305942862386734e-05, "loss": 0.0242, "step": 67430 }, { "grad_norm": 0.5085129737854004, "learning_rate": 2.629138388793793e-05, "loss": 0.0279, "step": 67440 }, { "grad_norm": 0.5890266299247742, "learning_rate": 2.6276827506227743e-05, "loss": 0.0238, "step": 67450 }, { "grad_norm": 0.32545116543769836, "learning_rate": 2.6262273718848037e-05, "loss": 0.0263, "step": 67460 }, { "grad_norm": 0.4839571714401245, "learning_rate": 2.6247722527390407e-05, "loss": 0.0263, "step": 67470 }, { "grad_norm": 0.32405585050582886, "learning_rate": 2.6233173933446104e-05, "loss": 0.0264, "step": 67480 }, { "grad_norm": 0.2689788341522217, "learning_rate": 2.6218627938606188e-05, "loss": 0.0269, "step": 67490 }, { "grad_norm": 0.8171402812004089, "learning_rate": 2.620408454446135e-05, "loss": 0.0328, "step": 67500 }, { "grad_norm": 0.38817811012268066, "learning_rate": 2.6189543752602046e-05, "loss": 0.0247, "step": 67510 }, { "grad_norm": 0.4395659863948822, "learning_rate": 2.6175005564618444e-05, "loss": 0.028, "step": 67520 }, { "grad_norm": 0.37982967495918274, "learning_rate": 2.6160469982100428e-05, "loss": 0.0271, "step": 67530 }, { "grad_norm": 0.3548373579978943, "learning_rate": 2.6145937006637555e-05, "loss": 0.0253, "step": 67540 }, { "grad_norm": 0.8952263593673706, "learning_rate": 2.6131406639819166e-05, "loss": 0.0293, "step": 67550 }, { "grad_norm": 0.3297516703605652, "learning_rate": 2.611687888323424e-05, "loss": 0.024, "step": 67560 }, { "grad_norm": 0.3363897502422333, "learning_rate": 2.610235373847153e-05, "loss": 0.0219, "step": 67570 }, { "grad_norm": 0.45330068469047546, "learning_rate": 2.6087831207119494e-05, "loss": 0.0241, "step": 67580 }, { "grad_norm": 0.4913918375968933, "learning_rate": 2.60733112907663e-05, "loss": 0.0269, "step": 67590 }, { "grad_norm": 0.3657798171043396, "learning_rate": 2.6058793990999786e-05, "loss": 0.0215, "step": 67600 }, { "grad_norm": 0.24826323986053467, "learning_rate": 2.6044279309407593e-05, "loss": 0.0252, "step": 67610 }, { "grad_norm": 0.41126465797424316, "learning_rate": 2.6029767247576976e-05, "loss": 0.0252, "step": 67620 }, { "grad_norm": 0.36545464396476746, "learning_rate": 2.601525780709497e-05, "loss": 0.0262, "step": 67630 }, { "grad_norm": 0.27180859446525574, "learning_rate": 2.6000750989548318e-05, "loss": 0.0219, "step": 67640 }, { "grad_norm": 0.562074601650238, "learning_rate": 2.5986246796523462e-05, "loss": 0.0238, "step": 67650 }, { "grad_norm": 0.3843011260032654, "learning_rate": 2.5971745229606546e-05, "loss": 0.03, "step": 67660 }, { "grad_norm": 0.4802277088165283, "learning_rate": 2.5957246290383442e-05, "loss": 0.024, "step": 67670 }, { "grad_norm": 0.4654598534107208, "learning_rate": 2.594274998043975e-05, "loss": 0.0242, "step": 67680 }, { "grad_norm": 0.5398851633071899, "learning_rate": 2.5928256301360733e-05, "loss": 0.0244, "step": 67690 }, { "grad_norm": 0.29743266105651855, "learning_rate": 2.591376525473142e-05, "loss": 0.0272, "step": 67700 }, { "grad_norm": 0.30393558740615845, "learning_rate": 2.5899276842136523e-05, "loss": 0.0248, "step": 67710 }, { "grad_norm": 0.24742624163627625, "learning_rate": 2.5884791065160495e-05, "loss": 0.0216, "step": 67720 }, { "grad_norm": 0.3011032044887543, "learning_rate": 2.5870307925387436e-05, "loss": 0.0296, "step": 67730 }, { "grad_norm": 0.32640254497528076, "learning_rate": 2.585582742440125e-05, "loss": 0.0225, "step": 67740 }, { "grad_norm": 0.1524355411529541, "learning_rate": 2.5841349563785455e-05, "loss": 0.0283, "step": 67750 }, { "grad_norm": 0.49822041392326355, "learning_rate": 2.5826874345123355e-05, "loss": 0.0246, "step": 67760 }, { "grad_norm": 0.38106125593185425, "learning_rate": 2.581240176999792e-05, "loss": 0.0283, "step": 67770 }, { "grad_norm": 0.25724536180496216, "learning_rate": 2.579793183999189e-05, "loss": 0.0274, "step": 67780 }, { "grad_norm": 0.37537309527397156, "learning_rate": 2.5783464556687627e-05, "loss": 0.0252, "step": 67790 }, { "grad_norm": 0.3054778277873993, "learning_rate": 2.576899992166729e-05, "loss": 0.0309, "step": 67800 }, { "grad_norm": 0.37288644909858704, "learning_rate": 2.5754537936512668e-05, "loss": 0.0226, "step": 67810 }, { "grad_norm": 0.3459065556526184, "learning_rate": 2.574007860280535e-05, "loss": 0.0214, "step": 67820 }, { "grad_norm": 0.25771230459213257, "learning_rate": 2.5725621922126518e-05, "loss": 0.0263, "step": 67830 }, { "grad_norm": 0.299320787191391, "learning_rate": 2.571116789605722e-05, "loss": 0.0236, "step": 67840 }, { "grad_norm": 0.2730449438095093, "learning_rate": 2.5696716526178054e-05, "loss": 0.0224, "step": 67850 }, { "grad_norm": 0.35419246554374695, "learning_rate": 2.568226781406946e-05, "loss": 0.0274, "step": 67860 }, { "grad_norm": 0.4106481671333313, "learning_rate": 2.5667821761311474e-05, "loss": 0.0257, "step": 67870 }, { "grad_norm": 0.33019378781318665, "learning_rate": 2.5653378369483917e-05, "loss": 0.0273, "step": 67880 }, { "grad_norm": 0.3706456422805786, "learning_rate": 2.5638937640166292e-05, "loss": 0.0228, "step": 67890 }, { "grad_norm": 0.7654100060462952, "learning_rate": 2.5624499574937842e-05, "loss": 0.0254, "step": 67900 }, { "grad_norm": 0.32533910870552063, "learning_rate": 2.5610064175377456e-05, "loss": 0.0311, "step": 67910 }, { "grad_norm": 0.419575572013855, "learning_rate": 2.5595631443063782e-05, "loss": 0.0255, "step": 67920 }, { "grad_norm": 0.4127349853515625, "learning_rate": 2.5581201379575177e-05, "loss": 0.0228, "step": 67930 }, { "grad_norm": 0.5242903828620911, "learning_rate": 2.5566773986489655e-05, "loss": 0.027, "step": 67940 }, { "grad_norm": 0.33191248774528503, "learning_rate": 2.5552349265385e-05, "loss": 0.0267, "step": 67950 }, { "grad_norm": 0.33921974897384644, "learning_rate": 2.5537927217838675e-05, "loss": 0.0266, "step": 67960 }, { "grad_norm": 0.4193321764469147, "learning_rate": 2.5523507845427864e-05, "loss": 0.0274, "step": 67970 }, { "grad_norm": 0.3399694859981537, "learning_rate": 2.5509091149729413e-05, "loss": 0.0259, "step": 67980 }, { "grad_norm": 0.40527939796447754, "learning_rate": 2.5494677132319962e-05, "loss": 0.0245, "step": 67990 }, { "grad_norm": 0.2467389851808548, "learning_rate": 2.548026579477575e-05, "loss": 0.0239, "step": 68000 }, { "grad_norm": 0.3777180016040802, "learning_rate": 2.5465857138672834e-05, "loss": 0.0217, "step": 68010 }, { "grad_norm": 0.29185208678245544, "learning_rate": 2.5451451165586848e-05, "loss": 0.0257, "step": 68020 }, { "grad_norm": 0.32724347710609436, "learning_rate": 2.543704787709329e-05, "loss": 0.0223, "step": 68030 }, { "grad_norm": 0.2704867124557495, "learning_rate": 2.542264727476723e-05, "loss": 0.023, "step": 68040 }, { "grad_norm": 0.3408522605895996, "learning_rate": 2.5408249360183535e-05, "loss": 0.0304, "step": 68050 }, { "grad_norm": 0.28234192728996277, "learning_rate": 2.5393854134916694e-05, "loss": 0.0295, "step": 68060 }, { "grad_norm": 0.3191206455230713, "learning_rate": 2.5379461600540993e-05, "loss": 0.0249, "step": 68070 }, { "grad_norm": 0.32212021946907043, "learning_rate": 2.5365071758630317e-05, "loss": 0.0285, "step": 68080 }, { "grad_norm": 0.28697365522384644, "learning_rate": 2.535068461075839e-05, "loss": 0.0228, "step": 68090 }, { "grad_norm": 0.2491818070411682, "learning_rate": 2.5336300158498516e-05, "loss": 0.0217, "step": 68100 }, { "grad_norm": 0.38971754908561707, "learning_rate": 2.5321918403423793e-05, "loss": 0.032, "step": 68110 }, { "grad_norm": 0.4383903443813324, "learning_rate": 2.5307539347106957e-05, "loss": 0.028, "step": 68120 }, { "grad_norm": 0.2966785430908203, "learning_rate": 2.529316299112049e-05, "loss": 0.028, "step": 68130 }, { "grad_norm": 0.241258904337883, "learning_rate": 2.5278789337036567e-05, "loss": 0.0265, "step": 68140 }, { "grad_norm": 0.2243266999721527, "learning_rate": 2.52644183864271e-05, "loss": 0.0255, "step": 68150 }, { "grad_norm": 0.2607869505882263, "learning_rate": 2.525005014086363e-05, "loss": 0.0243, "step": 68160 }, { "grad_norm": 0.530403196811676, "learning_rate": 2.5235684601917465e-05, "loss": 0.0295, "step": 68170 }, { "grad_norm": 0.5024782419204712, "learning_rate": 2.5221321771159618e-05, "loss": 0.0294, "step": 68180 }, { "grad_norm": 0.4755762815475464, "learning_rate": 2.5206961650160743e-05, "loss": 0.0222, "step": 68190 }, { "grad_norm": 0.5105541348457336, "learning_rate": 2.519260424049129e-05, "loss": 0.027, "step": 68200 }, { "grad_norm": 0.33010008931159973, "learning_rate": 2.5178249543721304e-05, "loss": 0.0243, "step": 68210 }, { "grad_norm": 0.42039093375205994, "learning_rate": 2.5163897561420668e-05, "loss": 0.0287, "step": 68220 }, { "grad_norm": 0.6347780227661133, "learning_rate": 2.514954829515883e-05, "loss": 0.0258, "step": 68230 }, { "grad_norm": 0.4986073672771454, "learning_rate": 2.513520174650505e-05, "loss": 0.0279, "step": 68240 }, { "grad_norm": 0.3912642002105713, "learning_rate": 2.5120857917028207e-05, "loss": 0.0297, "step": 68250 }, { "grad_norm": 0.46675199270248413, "learning_rate": 2.5106516808296953e-05, "loss": 0.0262, "step": 68260 }, { "grad_norm": 0.4468066096305847, "learning_rate": 2.5092178421879552e-05, "loss": 0.0272, "step": 68270 }, { "grad_norm": 0.646661639213562, "learning_rate": 2.5077842759344116e-05, "loss": 0.0244, "step": 68280 }, { "grad_norm": 0.36491796374320984, "learning_rate": 2.5063509822258304e-05, "loss": 0.0244, "step": 68290 }, { "grad_norm": 0.2781144678592682, "learning_rate": 2.5049179612189576e-05, "loss": 0.0285, "step": 68300 }, { "grad_norm": 0.34412896633148193, "learning_rate": 2.5034852130705043e-05, "loss": 0.0255, "step": 68310 }, { "grad_norm": 0.5244229435920715, "learning_rate": 2.5020527379371572e-05, "loss": 0.0226, "step": 68320 }, { "grad_norm": 0.30058497190475464, "learning_rate": 2.500620535975564e-05, "loss": 0.0269, "step": 68330 }, { "grad_norm": 0.44365301728248596, "learning_rate": 2.4991886073423515e-05, "loss": 0.025, "step": 68340 }, { "grad_norm": 0.9089003205299377, "learning_rate": 2.4977569521941134e-05, "loss": 0.0273, "step": 68350 }, { "grad_norm": 0.2905559241771698, "learning_rate": 2.4963255706874143e-05, "loss": 0.0256, "step": 68360 }, { "grad_norm": 0.32609352469444275, "learning_rate": 2.4948944629787853e-05, "loss": 0.0232, "step": 68370 }, { "grad_norm": 0.3211199939250946, "learning_rate": 2.493463629224731e-05, "loss": 0.0287, "step": 68380 }, { "grad_norm": 0.5111425518989563, "learning_rate": 2.492033069581728e-05, "loss": 0.0293, "step": 68390 }, { "grad_norm": 0.44574373960494995, "learning_rate": 2.4906027842062137e-05, "loss": 0.0233, "step": 68400 }, { "grad_norm": 0.23942328989505768, "learning_rate": 2.48917277325461e-05, "loss": 0.0255, "step": 68410 }, { "grad_norm": 0.48732835054397583, "learning_rate": 2.487743036883295e-05, "loss": 0.0227, "step": 68420 }, { "grad_norm": 0.2690373361110687, "learning_rate": 2.486313575248626e-05, "loss": 0.0287, "step": 68430 }, { "grad_norm": 0.20180842280387878, "learning_rate": 2.4848843885069235e-05, "loss": 0.0209, "step": 68440 }, { "grad_norm": 0.8605381846427917, "learning_rate": 2.483455476814484e-05, "loss": 0.0284, "step": 68450 }, { "grad_norm": 0.43809372186660767, "learning_rate": 2.4820268403275672e-05, "loss": 0.0273, "step": 68460 }, { "grad_norm": 0.44075167179107666, "learning_rate": 2.480598479202413e-05, "loss": 0.0234, "step": 68470 }, { "grad_norm": 0.396799236536026, "learning_rate": 2.4791703935952193e-05, "loss": 0.0268, "step": 68480 }, { "grad_norm": 0.20911215245723724, "learning_rate": 2.477742583662163e-05, "loss": 0.0267, "step": 68490 }, { "grad_norm": 0.23937365412712097, "learning_rate": 2.476315049559384e-05, "loss": 0.0231, "step": 68500 }, { "grad_norm": 0.7621972560882568, "learning_rate": 2.4748877914429987e-05, "loss": 0.0261, "step": 68510 }, { "grad_norm": 0.4847479462623596, "learning_rate": 2.4734608094690857e-05, "loss": 0.028, "step": 68520 }, { "grad_norm": 0.48716387152671814, "learning_rate": 2.4720341037937006e-05, "loss": 0.0282, "step": 68530 }, { "grad_norm": 0.2962423264980316, "learning_rate": 2.4706076745728645e-05, "loss": 0.0257, "step": 68540 }, { "grad_norm": 0.8775045275688171, "learning_rate": 2.469181521962573e-05, "loss": 0.0237, "step": 68550 }, { "grad_norm": 0.38579171895980835, "learning_rate": 2.467755646118783e-05, "loss": 0.0265, "step": 68560 }, { "grad_norm": 0.4822833240032196, "learning_rate": 2.46633004719743e-05, "loss": 0.0252, "step": 68570 }, { "grad_norm": 0.34719154238700867, "learning_rate": 2.4649047253544123e-05, "loss": 0.0261, "step": 68580 }, { "grad_norm": 0.32891908288002014, "learning_rate": 2.4634796807456016e-05, "loss": 0.0272, "step": 68590 }, { "grad_norm": 0.3381110429763794, "learning_rate": 2.46205491352684e-05, "loss": 0.0236, "step": 68600 }, { "grad_norm": 0.43765872716903687, "learning_rate": 2.4606304238539375e-05, "loss": 0.0201, "step": 68610 }, { "grad_norm": 0.4258427917957306, "learning_rate": 2.4592062118826753e-05, "loss": 0.0191, "step": 68620 }, { "grad_norm": 0.41045263409614563, "learning_rate": 2.4577822777688004e-05, "loss": 0.0268, "step": 68630 }, { "grad_norm": 0.4471472203731537, "learning_rate": 2.4563586216680347e-05, "loss": 0.0318, "step": 68640 }, { "grad_norm": 0.3534301817417145, "learning_rate": 2.4549352437360622e-05, "loss": 0.0245, "step": 68650 }, { "grad_norm": 0.3404395282268524, "learning_rate": 2.4535121441285493e-05, "loss": 0.0235, "step": 68660 }, { "grad_norm": 0.34292304515838623, "learning_rate": 2.4520893230011172e-05, "loss": 0.0231, "step": 68670 }, { "grad_norm": 0.33400288224220276, "learning_rate": 2.450666780509368e-05, "loss": 0.0336, "step": 68680 }, { "grad_norm": 0.6152405142784119, "learning_rate": 2.4492445168088656e-05, "loss": 0.0274, "step": 68690 }, { "grad_norm": 1.285457968711853, "learning_rate": 2.447822532055149e-05, "loss": 0.027, "step": 68700 }, { "grad_norm": 0.2729494869709015, "learning_rate": 2.4464008264037212e-05, "loss": 0.022, "step": 68710 }, { "grad_norm": 1.0191775560379028, "learning_rate": 2.4449794000100605e-05, "loss": 0.0257, "step": 68720 }, { "grad_norm": 0.43332430720329285, "learning_rate": 2.4435582530296108e-05, "loss": 0.0261, "step": 68730 }, { "grad_norm": 0.39866480231285095, "learning_rate": 2.4421373856177887e-05, "loss": 0.0295, "step": 68740 }, { "grad_norm": 0.325869083404541, "learning_rate": 2.440716797929975e-05, "loss": 0.0307, "step": 68750 }, { "grad_norm": 0.48919013142585754, "learning_rate": 2.439296490121526e-05, "loss": 0.0297, "step": 68760 }, { "grad_norm": 0.7320546507835388, "learning_rate": 2.4378764623477608e-05, "loss": 0.0296, "step": 68770 }, { "grad_norm": 0.34908509254455566, "learning_rate": 2.436456714763974e-05, "loss": 0.028, "step": 68780 }, { "grad_norm": 1.006162166595459, "learning_rate": 2.4350372475254262e-05, "loss": 0.0291, "step": 68790 }, { "grad_norm": 0.3354743719100952, "learning_rate": 2.433618060787351e-05, "loss": 0.0243, "step": 68800 }, { "grad_norm": 0.32776445150375366, "learning_rate": 2.432199154704944e-05, "loss": 0.0281, "step": 68810 }, { "grad_norm": 0.4317849576473236, "learning_rate": 2.430780529433377e-05, "loss": 0.0202, "step": 68820 }, { "grad_norm": 0.22059756517410278, "learning_rate": 2.42936218512779e-05, "loss": 0.0219, "step": 68830 }, { "grad_norm": 0.34598681330680847, "learning_rate": 2.427944121943288e-05, "loss": 0.0249, "step": 68840 }, { "grad_norm": 0.37698373198509216, "learning_rate": 2.4265263400349497e-05, "loss": 0.027, "step": 68850 }, { "grad_norm": 0.5241605043411255, "learning_rate": 2.4251088395578213e-05, "loss": 0.0267, "step": 68860 }, { "grad_norm": 0.23494400084018707, "learning_rate": 2.4236916206669214e-05, "loss": 0.0244, "step": 68870 }, { "grad_norm": 0.4206249415874481, "learning_rate": 2.4222746835172304e-05, "loss": 0.0232, "step": 68880 }, { "grad_norm": 0.5434720516204834, "learning_rate": 2.4208580282637065e-05, "loss": 0.0262, "step": 68890 }, { "grad_norm": 0.6254711747169495, "learning_rate": 2.4194416550612687e-05, "loss": 0.0243, "step": 68900 }, { "grad_norm": 0.3522275984287262, "learning_rate": 2.418025564064812e-05, "loss": 0.0232, "step": 68910 }, { "grad_norm": 0.7786727547645569, "learning_rate": 2.4166097554291978e-05, "loss": 0.0287, "step": 68920 }, { "grad_norm": 0.6052305698394775, "learning_rate": 2.4151942293092583e-05, "loss": 0.0254, "step": 68930 }, { "grad_norm": 0.7401440739631653, "learning_rate": 2.41377898585979e-05, "loss": 0.0288, "step": 68940 }, { "grad_norm": 0.2505247890949249, "learning_rate": 2.4123640252355652e-05, "loss": 0.0244, "step": 68950 }, { "grad_norm": 0.37099191546440125, "learning_rate": 2.410949347591318e-05, "loss": 0.0276, "step": 68960 }, { "grad_norm": 0.3308010399341583, "learning_rate": 2.4095349530817578e-05, "loss": 0.0238, "step": 68970 }, { "grad_norm": 0.4096291959285736, "learning_rate": 2.4081208418615603e-05, "loss": 0.0288, "step": 68980 }, { "grad_norm": 0.30472928285598755, "learning_rate": 2.4067070140853725e-05, "loss": 0.0244, "step": 68990 }, { "grad_norm": 0.31346777081489563, "learning_rate": 2.405293469907805e-05, "loss": 0.0264, "step": 69000 }, { "grad_norm": 0.3153184950351715, "learning_rate": 2.4038802094834446e-05, "loss": 0.0251, "step": 69010 }, { "grad_norm": 0.3424266278743744, "learning_rate": 2.4024672329668395e-05, "loss": 0.0263, "step": 69020 }, { "grad_norm": 0.46000760793685913, "learning_rate": 2.4010545405125125e-05, "loss": 0.0299, "step": 69030 }, { "grad_norm": 0.2288215458393097, "learning_rate": 2.3996421322749528e-05, "loss": 0.0276, "step": 69040 }, { "grad_norm": 0.5553911328315735, "learning_rate": 2.3982300084086224e-05, "loss": 0.0284, "step": 69050 }, { "grad_norm": 0.31416624784469604, "learning_rate": 2.396818169067944e-05, "loss": 0.0284, "step": 69060 }, { "grad_norm": 0.4017788767814636, "learning_rate": 2.3954066144073178e-05, "loss": 0.0248, "step": 69070 }, { "grad_norm": 0.3490482568740845, "learning_rate": 2.39399534458111e-05, "loss": 0.0286, "step": 69080 }, { "grad_norm": 0.47913405299186707, "learning_rate": 2.392584359743651e-05, "loss": 0.0281, "step": 69090 }, { "grad_norm": 0.29367876052856445, "learning_rate": 2.3911736600492463e-05, "loss": 0.0237, "step": 69100 }, { "grad_norm": 0.3563425540924072, "learning_rate": 2.3897632456521672e-05, "loss": 0.0262, "step": 69110 }, { "grad_norm": 0.2901303172111511, "learning_rate": 2.388353116706658e-05, "loss": 0.0218, "step": 69120 }, { "grad_norm": 0.44691529870033264, "learning_rate": 2.3869432733669232e-05, "loss": 0.0292, "step": 69130 }, { "grad_norm": 0.36996471881866455, "learning_rate": 2.385533715787145e-05, "loss": 0.0244, "step": 69140 }, { "grad_norm": 0.292357474565506, "learning_rate": 2.3841244441214673e-05, "loss": 0.024, "step": 69150 }, { "grad_norm": 0.3113532066345215, "learning_rate": 2.3827154585240073e-05, "loss": 0.0277, "step": 69160 }, { "grad_norm": 0.32012665271759033, "learning_rate": 2.3813067591488498e-05, "loss": 0.0255, "step": 69170 }, { "grad_norm": 0.5273666381835938, "learning_rate": 2.3798983461500494e-05, "loss": 0.0298, "step": 69180 }, { "grad_norm": 0.4134533107280731, "learning_rate": 2.3784902196816256e-05, "loss": 0.0238, "step": 69190 }, { "grad_norm": 0.4333416521549225, "learning_rate": 2.3770823798975716e-05, "loss": 0.0249, "step": 69200 }, { "grad_norm": 0.339481920003891, "learning_rate": 2.3756748269518424e-05, "loss": 0.0308, "step": 69210 }, { "grad_norm": 0.33669042587280273, "learning_rate": 2.374267560998371e-05, "loss": 0.0224, "step": 69220 }, { "grad_norm": 0.43416690826416016, "learning_rate": 2.3728605821910477e-05, "loss": 0.0302, "step": 69230 }, { "grad_norm": 0.41755664348602295, "learning_rate": 2.3714538906837453e-05, "loss": 0.0278, "step": 69240 }, { "grad_norm": 0.5458061695098877, "learning_rate": 2.3700474866302913e-05, "loss": 0.0232, "step": 69250 }, { "grad_norm": 0.28722304105758667, "learning_rate": 2.3686413701844917e-05, "loss": 0.028, "step": 69260 }, { "grad_norm": 0.30489328503608704, "learning_rate": 2.367235541500114e-05, "loss": 0.0269, "step": 69270 }, { "grad_norm": 0.3312833309173584, "learning_rate": 2.3658300007308993e-05, "loss": 0.0261, "step": 69280 }, { "grad_norm": 0.33375707268714905, "learning_rate": 2.364424748030555e-05, "loss": 0.0365, "step": 69290 }, { "grad_norm": 0.3385920524597168, "learning_rate": 2.3630197835527595e-05, "loss": 0.0286, "step": 69300 }, { "grad_norm": 0.2797302305698395, "learning_rate": 2.3616151074511546e-05, "loss": 0.0251, "step": 69310 }, { "grad_norm": 0.24905911087989807, "learning_rate": 2.3602107198793545e-05, "loss": 0.0249, "step": 69320 }, { "grad_norm": 0.2899095416069031, "learning_rate": 2.3588066209909427e-05, "loss": 0.0222, "step": 69330 }, { "grad_norm": 1.115557074546814, "learning_rate": 2.3574028109394662e-05, "loss": 0.0256, "step": 69340 }, { "grad_norm": 0.31451666355133057, "learning_rate": 2.3559992898784454e-05, "loss": 0.0235, "step": 69350 }, { "grad_norm": 0.9322333931922913, "learning_rate": 2.354596057961367e-05, "loss": 0.0318, "step": 69360 }, { "grad_norm": 0.39155760407447815, "learning_rate": 2.353193115341688e-05, "loss": 0.0271, "step": 69370 }, { "grad_norm": 0.32923421263694763, "learning_rate": 2.3517904621728294e-05, "loss": 0.0229, "step": 69380 }, { "grad_norm": 0.7006142735481262, "learning_rate": 2.3503880986081855e-05, "loss": 0.0239, "step": 69390 }, { "grad_norm": 0.3110876679420471, "learning_rate": 2.3489860248011136e-05, "loss": 0.0288, "step": 69400 }, { "grad_norm": 0.7445245385169983, "learning_rate": 2.3475842409049464e-05, "loss": 0.0259, "step": 69410 }, { "grad_norm": 0.3473752737045288, "learning_rate": 2.346182747072975e-05, "loss": 0.021, "step": 69420 }, { "grad_norm": 0.2284974455833435, "learning_rate": 2.3447815434584718e-05, "loss": 0.0222, "step": 69430 }, { "grad_norm": 0.2833571135997772, "learning_rate": 2.3433806302146656e-05, "loss": 0.0259, "step": 69440 }, { "grad_norm": 0.7323948740959167, "learning_rate": 2.3419800074947612e-05, "loss": 0.0279, "step": 69450 }, { "grad_norm": 0.4195047914981842, "learning_rate": 2.3405796754519248e-05, "loss": 0.0261, "step": 69460 }, { "grad_norm": 0.3359646797180176, "learning_rate": 2.3391796342392992e-05, "loss": 0.0263, "step": 69470 }, { "grad_norm": 0.287180632352829, "learning_rate": 2.337779884009984e-05, "loss": 0.0244, "step": 69480 }, { "grad_norm": 0.5461944341659546, "learning_rate": 2.3363804249170624e-05, "loss": 0.0286, "step": 69490 }, { "grad_norm": 0.34164002537727356, "learning_rate": 2.3349812571135703e-05, "loss": 0.0263, "step": 69500 }, { "grad_norm": 0.40471914410591125, "learning_rate": 2.333582380752523e-05, "loss": 0.0277, "step": 69510 }, { "grad_norm": 0.28457245230674744, "learning_rate": 2.3321837959868963e-05, "loss": 0.027, "step": 69520 }, { "grad_norm": 0.5075979232788086, "learning_rate": 2.3307855029696383e-05, "loss": 0.022, "step": 69530 }, { "grad_norm": 0.2730395197868347, "learning_rate": 2.3293875018536648e-05, "loss": 0.0241, "step": 69540 }, { "grad_norm": 0.25251656770706177, "learning_rate": 2.327989792791861e-05, "loss": 0.0197, "step": 69550 }, { "grad_norm": 0.3522312343120575, "learning_rate": 2.326592375937074e-05, "loss": 0.0232, "step": 69560 }, { "grad_norm": 0.3317674994468689, "learning_rate": 2.3251952514421248e-05, "loss": 0.024, "step": 69570 }, { "grad_norm": 0.4357701241970062, "learning_rate": 2.323798419459804e-05, "loss": 0.0217, "step": 69580 }, { "grad_norm": 0.5727059841156006, "learning_rate": 2.3224018801428625e-05, "loss": 0.024, "step": 69590 }, { "grad_norm": 0.8742018342018127, "learning_rate": 2.3210056336440273e-05, "loss": 0.0227, "step": 69600 }, { "grad_norm": 0.47491151094436646, "learning_rate": 2.3196096801159844e-05, "loss": 0.0261, "step": 69610 }, { "grad_norm": 0.29581382870674133, "learning_rate": 2.3182140197114012e-05, "loss": 0.028, "step": 69620 }, { "grad_norm": 0.35080912709236145, "learning_rate": 2.3168186525828985e-05, "loss": 0.0253, "step": 69630 }, { "grad_norm": 0.30852994322776794, "learning_rate": 2.3154235788830757e-05, "loss": 0.0217, "step": 69640 }, { "grad_norm": 0.22503727674484253, "learning_rate": 2.3140287987644927e-05, "loss": 0.0264, "step": 69650 }, { "grad_norm": 0.4012376368045807, "learning_rate": 2.3126343123796833e-05, "loss": 0.0281, "step": 69660 }, { "grad_norm": 0.35012784600257874, "learning_rate": 2.3112401198811418e-05, "loss": 0.0232, "step": 69670 }, { "grad_norm": 0.43761274218559265, "learning_rate": 2.3098462214213418e-05, "loss": 0.0248, "step": 69680 }, { "grad_norm": 0.39656001329421997, "learning_rate": 2.308452617152712e-05, "loss": 0.0216, "step": 69690 }, { "grad_norm": 0.30030813813209534, "learning_rate": 2.3070593072276603e-05, "loss": 0.0256, "step": 69700 }, { "grad_norm": 0.25597044825553894, "learning_rate": 2.3056662917985518e-05, "loss": 0.0252, "step": 69710 }, { "grad_norm": 0.3125760555267334, "learning_rate": 2.3042735710177283e-05, "loss": 0.0263, "step": 69720 }, { "grad_norm": 0.38697177171707153, "learning_rate": 2.302881145037492e-05, "loss": 0.0298, "step": 69730 }, { "grad_norm": 0.3022669553756714, "learning_rate": 2.3014890140101192e-05, "loss": 0.0306, "step": 69740 }, { "grad_norm": 0.2623298168182373, "learning_rate": 2.30009717808785e-05, "loss": 0.0226, "step": 69750 }, { "grad_norm": 0.861669659614563, "learning_rate": 2.2987056374228967e-05, "loss": 0.0256, "step": 69760 }, { "grad_norm": 0.4245539605617523, "learning_rate": 2.297314392167431e-05, "loss": 0.0272, "step": 69770 }, { "grad_norm": 0.4499732255935669, "learning_rate": 2.295923442473601e-05, "loss": 0.0243, "step": 69780 }, { "grad_norm": 0.34615224599838257, "learning_rate": 2.294532788493519e-05, "loss": 0.0224, "step": 69790 }, { "grad_norm": 0.3072474002838135, "learning_rate": 2.293142430379262e-05, "loss": 0.03, "step": 69800 }, { "grad_norm": 0.2378838062286377, "learning_rate": 2.291752368282879e-05, "loss": 0.0234, "step": 69810 }, { "grad_norm": 0.32900315523147583, "learning_rate": 2.290362602356385e-05, "loss": 0.0255, "step": 69820 }, { "grad_norm": 0.3055538535118103, "learning_rate": 2.288973132751765e-05, "loss": 0.0212, "step": 69830 }, { "grad_norm": 0.33896604180336, "learning_rate": 2.287583959620965e-05, "loss": 0.0322, "step": 69840 }, { "grad_norm": 0.24727441370487213, "learning_rate": 2.2861950831159074e-05, "loss": 0.0233, "step": 69850 }, { "grad_norm": 0.44004619121551514, "learning_rate": 2.284806503388472e-05, "loss": 0.0263, "step": 69860 }, { "grad_norm": 0.31089168787002563, "learning_rate": 2.283418220590518e-05, "loss": 0.0281, "step": 69870 }, { "grad_norm": 0.40804171562194824, "learning_rate": 2.282030234873862e-05, "loss": 0.0264, "step": 69880 }, { "grad_norm": 0.352135568857193, "learning_rate": 2.2806425463902948e-05, "loss": 0.0237, "step": 69890 }, { "grad_norm": 0.2953062355518341, "learning_rate": 2.279255155291568e-05, "loss": 0.0255, "step": 69900 }, { "grad_norm": 0.6033146381378174, "learning_rate": 2.277868061729409e-05, "loss": 0.0236, "step": 69910 }, { "grad_norm": 0.2817077040672302, "learning_rate": 2.276481265855504e-05, "loss": 0.0256, "step": 69920 }, { "grad_norm": 0.6064243316650391, "learning_rate": 2.2750947678215128e-05, "loss": 0.0254, "step": 69930 }, { "grad_norm": 0.3984629213809967, "learning_rate": 2.2737085677790616e-05, "loss": 0.0237, "step": 69940 }, { "grad_norm": 0.33052390813827515, "learning_rate": 2.2723226658797443e-05, "loss": 0.0276, "step": 69950 }, { "grad_norm": 0.5590982437133789, "learning_rate": 2.270937062275117e-05, "loss": 0.0283, "step": 69960 }, { "grad_norm": 0.2591577470302582, "learning_rate": 2.2695517571167113e-05, "loss": 0.0264, "step": 69970 }, { "grad_norm": 0.40680718421936035, "learning_rate": 2.268166750556019e-05, "loss": 0.0246, "step": 69980 }, { "grad_norm": 0.5778293609619141, "learning_rate": 2.266782042744503e-05, "loss": 0.0234, "step": 69990 }, { "grad_norm": 0.39327049255371094, "learning_rate": 2.2653976338335935e-05, "loss": 0.022, "step": 70000 }, { "grad_norm": 0.6984977722167969, "learning_rate": 2.2640135239746895e-05, "loss": 0.0294, "step": 70010 }, { "grad_norm": 0.30082637071609497, "learning_rate": 2.2626297133191515e-05, "loss": 0.0296, "step": 70020 }, { "grad_norm": 0.2451525330543518, "learning_rate": 2.261246202018312e-05, "loss": 0.027, "step": 70030 }, { "grad_norm": 0.3615821301937103, "learning_rate": 2.2598629902234718e-05, "loss": 0.0216, "step": 70040 }, { "grad_norm": 0.4819420278072357, "learning_rate": 2.258480078085894e-05, "loss": 0.0281, "step": 70050 }, { "grad_norm": 0.42899227142333984, "learning_rate": 2.2570974657568126e-05, "loss": 0.0289, "step": 70060 }, { "grad_norm": 0.600972056388855, "learning_rate": 2.2557151533874286e-05, "loss": 0.023, "step": 70070 }, { "grad_norm": 0.28907424211502075, "learning_rate": 2.2543331411289116e-05, "loss": 0.0248, "step": 70080 }, { "grad_norm": 0.24816913902759552, "learning_rate": 2.252951429132392e-05, "loss": 0.0238, "step": 70090 }, { "grad_norm": 0.3596203625202179, "learning_rate": 2.2515700175489762e-05, "loss": 0.0249, "step": 70100 }, { "grad_norm": 0.40078219771385193, "learning_rate": 2.2501889065297288e-05, "loss": 0.0214, "step": 70110 }, { "grad_norm": 0.47703009843826294, "learning_rate": 2.2488080962256887e-05, "loss": 0.0216, "step": 70120 }, { "grad_norm": 0.4956733286380768, "learning_rate": 2.2474275867878585e-05, "loss": 0.0248, "step": 70130 }, { "grad_norm": 0.2307180017232895, "learning_rate": 2.246047378367211e-05, "loss": 0.0236, "step": 70140 }, { "grad_norm": 0.3865920305252075, "learning_rate": 2.2446674711146798e-05, "loss": 0.0272, "step": 70150 }, { "grad_norm": 0.2304202765226364, "learning_rate": 2.2432878651811734e-05, "loss": 0.0277, "step": 70160 }, { "grad_norm": 0.29546478390693665, "learning_rate": 2.2419085607175594e-05, "loss": 0.0278, "step": 70170 }, { "grad_norm": 0.4111652374267578, "learning_rate": 2.240529557874678e-05, "loss": 0.0252, "step": 70180 }, { "grad_norm": 0.3273935317993164, "learning_rate": 2.2391508568033358e-05, "loss": 0.0213, "step": 70190 }, { "grad_norm": 0.2667255997657776, "learning_rate": 2.237772457654307e-05, "loss": 0.0329, "step": 70200 }, { "grad_norm": 0.3603014051914215, "learning_rate": 2.236394360578327e-05, "loss": 0.0286, "step": 70210 }, { "grad_norm": 0.44398412108421326, "learning_rate": 2.2350165657261073e-05, "loss": 0.0324, "step": 70220 }, { "grad_norm": 0.3561549186706543, "learning_rate": 2.2336390732483164e-05, "loss": 0.0303, "step": 70230 }, { "grad_norm": 0.3473694324493408, "learning_rate": 2.2322618832955976e-05, "loss": 0.0266, "step": 70240 }, { "grad_norm": 0.30847951769828796, "learning_rate": 2.2308849960185578e-05, "loss": 0.0279, "step": 70250 }, { "grad_norm": 0.3746853768825531, "learning_rate": 2.229508411567774e-05, "loss": 0.0215, "step": 70260 }, { "grad_norm": 0.3820447623729706, "learning_rate": 2.2281321300937823e-05, "loss": 0.0215, "step": 70270 }, { "grad_norm": 0.578689694404602, "learning_rate": 2.2267561517470937e-05, "loss": 0.0211, "step": 70280 }, { "grad_norm": 0.3268459141254425, "learning_rate": 2.2253804766781845e-05, "loss": 0.0226, "step": 70290 }, { "grad_norm": 0.3641420006752014, "learning_rate": 2.2240051050374933e-05, "loss": 0.0283, "step": 70300 }, { "grad_norm": 0.2739660441875458, "learning_rate": 2.22263003697543e-05, "loss": 0.0209, "step": 70310 }, { "grad_norm": 0.40036365389823914, "learning_rate": 2.2212552726423702e-05, "loss": 0.0272, "step": 70320 }, { "grad_norm": 0.4005877375602722, "learning_rate": 2.2198808121886578e-05, "loss": 0.0218, "step": 70330 }, { "grad_norm": 0.31706032156944275, "learning_rate": 2.218506655764598e-05, "loss": 0.0259, "step": 70340 }, { "grad_norm": 0.37189534306526184, "learning_rate": 2.2171328035204704e-05, "loss": 0.0298, "step": 70350 }, { "grad_norm": 0.34871792793273926, "learning_rate": 2.2157592556065133e-05, "loss": 0.0236, "step": 70360 }, { "grad_norm": 0.3931249678134918, "learning_rate": 2.2143860121729382e-05, "loss": 0.0267, "step": 70370 }, { "grad_norm": 0.8389230966567993, "learning_rate": 2.2130130733699206e-05, "loss": 0.0305, "step": 70380 }, { "grad_norm": 0.47294989228248596, "learning_rate": 2.2116404393476052e-05, "loss": 0.0239, "step": 70390 }, { "grad_norm": 0.7703770399093628, "learning_rate": 2.2102681102560967e-05, "loss": 0.0326, "step": 70400 }, { "grad_norm": 0.3224412500858307, "learning_rate": 2.208896086245476e-05, "loss": 0.0284, "step": 70410 }, { "grad_norm": 0.5245215892791748, "learning_rate": 2.2075243674657808e-05, "loss": 0.0254, "step": 70420 }, { "grad_norm": 0.23429903388023376, "learning_rate": 2.206152954067024e-05, "loss": 0.0272, "step": 70430 }, { "grad_norm": 0.492382287979126, "learning_rate": 2.2047818461991758e-05, "loss": 0.0282, "step": 70440 }, { "grad_norm": 0.3265984356403351, "learning_rate": 2.2034110440121864e-05, "loss": 0.0246, "step": 70450 }, { "grad_norm": 0.9930707216262817, "learning_rate": 2.2020405476559585e-05, "loss": 0.0264, "step": 70460 }, { "grad_norm": 0.3228192627429962, "learning_rate": 2.2006703572803722e-05, "loss": 0.0262, "step": 70470 }, { "grad_norm": 0.4485194981098175, "learning_rate": 2.1993004730352646e-05, "loss": 0.0274, "step": 70480 }, { "grad_norm": 0.2624373137950897, "learning_rate": 2.1979308950704464e-05, "loss": 0.0215, "step": 70490 }, { "grad_norm": 0.2699717581272125, "learning_rate": 2.1965616235356923e-05, "loss": 0.0242, "step": 70500 }, { "grad_norm": 0.3810025453567505, "learning_rate": 2.1951926585807465e-05, "loss": 0.027, "step": 70510 }, { "grad_norm": 0.3126092851161957, "learning_rate": 2.193824000355313e-05, "loss": 0.0226, "step": 70520 }, { "grad_norm": 0.40069106221199036, "learning_rate": 2.192455649009067e-05, "loss": 0.0265, "step": 70530 }, { "grad_norm": 0.3418942987918854, "learning_rate": 2.1910876046916523e-05, "loss": 0.0245, "step": 70540 }, { "grad_norm": 0.27167290449142456, "learning_rate": 2.1897198675526725e-05, "loss": 0.0198, "step": 70550 }, { "grad_norm": 1.3807035684585571, "learning_rate": 2.1883524377417024e-05, "loss": 0.0215, "step": 70560 }, { "grad_norm": 0.5747936367988586, "learning_rate": 2.1869853154082825e-05, "loss": 0.0222, "step": 70570 }, { "grad_norm": 0.31633877754211426, "learning_rate": 2.1856185007019215e-05, "loss": 0.0317, "step": 70580 }, { "grad_norm": 0.3335433304309845, "learning_rate": 2.1842519937720874e-05, "loss": 0.0255, "step": 70590 }, { "grad_norm": 0.6048555374145508, "learning_rate": 2.182885794768224e-05, "loss": 0.0234, "step": 70600 }, { "grad_norm": 0.23062826693058014, "learning_rate": 2.1815199038397326e-05, "loss": 0.0308, "step": 70610 }, { "grad_norm": 0.2638061046600342, "learning_rate": 2.1801543211359893e-05, "loss": 0.0254, "step": 70620 }, { "grad_norm": 0.4304582476615906, "learning_rate": 2.178789046806326e-05, "loss": 0.0222, "step": 70630 }, { "grad_norm": 0.35051319003105164, "learning_rate": 2.1774240810000547e-05, "loss": 0.0218, "step": 70640 }, { "grad_norm": 0.21037942171096802, "learning_rate": 2.17605942386644e-05, "loss": 0.0226, "step": 70650 }, { "grad_norm": 0.30835723876953125, "learning_rate": 2.174695075554723e-05, "loss": 0.028, "step": 70660 }, { "grad_norm": 0.19782276451587677, "learning_rate": 2.173331036214103e-05, "loss": 0.0234, "step": 70670 }, { "grad_norm": 0.23179037868976593, "learning_rate": 2.171967305993753e-05, "loss": 0.0228, "step": 70680 }, { "grad_norm": 0.45444804430007935, "learning_rate": 2.1706038850428024e-05, "loss": 0.0279, "step": 70690 }, { "grad_norm": 0.2424953579902649, "learning_rate": 2.169240773510361e-05, "loss": 0.0236, "step": 70700 }, { "grad_norm": 0.24931631982326508, "learning_rate": 2.1678779715454906e-05, "loss": 0.0299, "step": 70710 }, { "grad_norm": 0.7831935882568359, "learning_rate": 2.166515479297229e-05, "loss": 0.0241, "step": 70720 }, { "grad_norm": 0.6885095238685608, "learning_rate": 2.1651532969145732e-05, "loss": 0.0338, "step": 70730 }, { "grad_norm": 0.35624930262565613, "learning_rate": 2.1637914245464903e-05, "loss": 0.0275, "step": 70740 }, { "grad_norm": 0.3260132670402527, "learning_rate": 2.1624298623419133e-05, "loss": 0.0254, "step": 70750 }, { "grad_norm": 0.40991222858428955, "learning_rate": 2.1610686104497413e-05, "loss": 0.0235, "step": 70760 }, { "grad_norm": 0.6533056497573853, "learning_rate": 2.1597076690188366e-05, "loss": 0.03, "step": 70770 }, { "grad_norm": 0.26464852690696716, "learning_rate": 2.1583470381980315e-05, "loss": 0.0189, "step": 70780 }, { "grad_norm": 0.43748217821121216, "learning_rate": 2.156986718136123e-05, "loss": 0.0244, "step": 70790 }, { "grad_norm": 0.43251341581344604, "learning_rate": 2.155626708981871e-05, "loss": 0.029, "step": 70800 }, { "grad_norm": 0.3359593152999878, "learning_rate": 2.1542670108840075e-05, "loss": 0.0235, "step": 70810 }, { "grad_norm": 0.32561999559402466, "learning_rate": 2.152907623991221e-05, "loss": 0.0196, "step": 70820 }, { "grad_norm": 0.368205189704895, "learning_rate": 2.1515485484521802e-05, "loss": 0.0262, "step": 70830 }, { "grad_norm": 0.2636246979236603, "learning_rate": 2.150189784415506e-05, "loss": 0.029, "step": 70840 }, { "grad_norm": 0.3752668797969818, "learning_rate": 2.148831332029794e-05, "loss": 0.0291, "step": 70850 }, { "grad_norm": 0.2607884705066681, "learning_rate": 2.147473191443599e-05, "loss": 0.0226, "step": 70860 }, { "grad_norm": 0.37199926376342773, "learning_rate": 2.1461153628054492e-05, "loss": 0.0288, "step": 70870 }, { "grad_norm": 0.4019221365451813, "learning_rate": 2.1447578462638284e-05, "loss": 0.0287, "step": 70880 }, { "grad_norm": 0.3637678027153015, "learning_rate": 2.143400641967201e-05, "loss": 0.0259, "step": 70890 }, { "grad_norm": 0.5411887168884277, "learning_rate": 2.1420437500639823e-05, "loss": 0.0236, "step": 70900 }, { "grad_norm": 0.2928728461265564, "learning_rate": 2.1406871707025646e-05, "loss": 0.02, "step": 70910 }, { "grad_norm": 0.30339300632476807, "learning_rate": 2.1393309040312963e-05, "loss": 0.0283, "step": 70920 }, { "grad_norm": 0.3052574694156647, "learning_rate": 2.137974950198502e-05, "loss": 0.0234, "step": 70930 }, { "grad_norm": 0.36904817819595337, "learning_rate": 2.1366193093524617e-05, "loss": 0.0241, "step": 70940 }, { "grad_norm": 0.23267410695552826, "learning_rate": 2.1352639816414288e-05, "loss": 0.0231, "step": 70950 }, { "grad_norm": 0.28751611709594727, "learning_rate": 2.1339089672136202e-05, "loss": 0.0234, "step": 70960 }, { "grad_norm": 0.469976007938385, "learning_rate": 2.1325542662172193e-05, "loss": 0.0205, "step": 70970 }, { "grad_norm": 0.5583021640777588, "learning_rate": 2.1311998788003712e-05, "loss": 0.0222, "step": 70980 }, { "grad_norm": 0.39696207642555237, "learning_rate": 2.1298458051111907e-05, "loss": 0.0315, "step": 70990 }, { "grad_norm": 1.4894648790359497, "learning_rate": 2.1284920452977602e-05, "loss": 0.024, "step": 71000 }, { "grad_norm": 0.2282608300447464, "learning_rate": 2.12713859950812e-05, "loss": 0.0308, "step": 71010 }, { "grad_norm": 0.3472738265991211, "learning_rate": 2.1257854678902832e-05, "loss": 0.0294, "step": 71020 }, { "grad_norm": 0.35344788432121277, "learning_rate": 2.124432650592227e-05, "loss": 0.031, "step": 71030 }, { "grad_norm": 0.4281068444252014, "learning_rate": 2.1230801477618937e-05, "loss": 0.0253, "step": 71040 }, { "grad_norm": 0.285459965467453, "learning_rate": 2.121727959547189e-05, "loss": 0.0218, "step": 71050 }, { "grad_norm": 0.34233027696609497, "learning_rate": 2.1203760860959886e-05, "loss": 0.0245, "step": 71060 }, { "grad_norm": 0.2519929111003876, "learning_rate": 2.119024527556127e-05, "loss": 0.0257, "step": 71070 }, { "grad_norm": 0.39849892258644104, "learning_rate": 2.117673284075415e-05, "loss": 0.027, "step": 71080 }, { "grad_norm": 0.4000822901725769, "learning_rate": 2.1163223558016165e-05, "loss": 0.0216, "step": 71090 }, { "grad_norm": 0.4116247892379761, "learning_rate": 2.1149717428824722e-05, "loss": 0.0252, "step": 71100 }, { "grad_norm": 0.6206857562065125, "learning_rate": 2.1136214454656784e-05, "loss": 0.0263, "step": 71110 }, { "grad_norm": 0.295240193605423, "learning_rate": 2.1122714636989056e-05, "loss": 0.0257, "step": 71120 }, { "grad_norm": 0.5158593654632568, "learning_rate": 2.1109217977297823e-05, "loss": 0.025, "step": 71130 }, { "grad_norm": 0.4956069588661194, "learning_rate": 2.1095724477059077e-05, "loss": 0.0249, "step": 71140 }, { "grad_norm": 0.3218659460544586, "learning_rate": 2.1082234137748434e-05, "loss": 0.0218, "step": 71150 }, { "grad_norm": 0.35288235545158386, "learning_rate": 2.106874696084122e-05, "loss": 0.0271, "step": 71160 }, { "grad_norm": 0.21561314165592194, "learning_rate": 2.1055262947812315e-05, "loss": 0.0277, "step": 71170 }, { "grad_norm": 0.3888889253139496, "learning_rate": 2.104178210013636e-05, "loss": 0.0288, "step": 71180 }, { "grad_norm": 0.286283403635025, "learning_rate": 2.1028304419287554e-05, "loss": 0.0221, "step": 71190 }, { "grad_norm": 0.4062581956386566, "learning_rate": 2.101482990673982e-05, "loss": 0.0224, "step": 71200 }, { "grad_norm": 0.327222615480423, "learning_rate": 2.1001358563966707e-05, "loss": 0.019, "step": 71210 }, { "grad_norm": 0.5754268169403076, "learning_rate": 2.098789039244144e-05, "loss": 0.0279, "step": 71220 }, { "grad_norm": 0.3283323645591736, "learning_rate": 2.097442539363685e-05, "loss": 0.0324, "step": 71230 }, { "grad_norm": 0.21820394694805145, "learning_rate": 2.0960963569025448e-05, "loss": 0.0266, "step": 71240 }, { "grad_norm": 0.4649803936481476, "learning_rate": 2.0947504920079435e-05, "loss": 0.0253, "step": 71250 }, { "grad_norm": 0.33072805404663086, "learning_rate": 2.093404944827059e-05, "loss": 0.0267, "step": 71260 }, { "grad_norm": 0.6837061047554016, "learning_rate": 2.0920597155070394e-05, "loss": 0.0234, "step": 71270 }, { "grad_norm": 0.2830149531364441, "learning_rate": 2.090714804194997e-05, "loss": 0.0259, "step": 71280 }, { "grad_norm": 0.30347055196762085, "learning_rate": 2.0893702110380114e-05, "loss": 0.0246, "step": 71290 }, { "grad_norm": 0.22333365678787231, "learning_rate": 2.0880259361831216e-05, "loss": 0.0226, "step": 71300 }, { "grad_norm": 0.2883814573287964, "learning_rate": 2.0866819797773397e-05, "loss": 0.0272, "step": 71310 }, { "grad_norm": 0.3629929721355438, "learning_rate": 2.085338341967634e-05, "loss": 0.0234, "step": 71320 }, { "grad_norm": 0.2695450484752655, "learning_rate": 2.083995022900946e-05, "loss": 0.0219, "step": 71330 }, { "grad_norm": 0.25731658935546875, "learning_rate": 2.0826520227241768e-05, "loss": 0.0249, "step": 71340 }, { "grad_norm": 0.5070192813873291, "learning_rate": 2.081309341584199e-05, "loss": 0.0245, "step": 71350 }, { "grad_norm": 0.6240041851997375, "learning_rate": 2.0799669796278413e-05, "loss": 0.026, "step": 71360 }, { "grad_norm": 0.22408097982406616, "learning_rate": 2.078624937001906e-05, "loss": 0.0257, "step": 71370 }, { "grad_norm": 0.3271527886390686, "learning_rate": 2.0772832138531544e-05, "loss": 0.0294, "step": 71380 }, { "grad_norm": 0.9027778506278992, "learning_rate": 2.0759418103283158e-05, "loss": 0.0231, "step": 71390 }, { "grad_norm": 0.27867671847343445, "learning_rate": 2.0746007265740846e-05, "loss": 0.0244, "step": 71400 }, { "grad_norm": 0.307288259267807, "learning_rate": 2.0732599627371215e-05, "loss": 0.022, "step": 71410 }, { "grad_norm": 0.41312721371650696, "learning_rate": 2.0719195189640468e-05, "loss": 0.0272, "step": 71420 }, { "grad_norm": 0.2921597957611084, "learning_rate": 2.0705793954014528e-05, "loss": 0.0319, "step": 71430 }, { "grad_norm": 0.38360145688056946, "learning_rate": 2.06923959219589e-05, "loss": 0.021, "step": 71440 }, { "grad_norm": 0.3531421720981598, "learning_rate": 2.0679001094938793e-05, "loss": 0.0263, "step": 71450 }, { "grad_norm": 0.3265538513660431, "learning_rate": 2.066560947441904e-05, "loss": 0.0246, "step": 71460 }, { "grad_norm": 0.2935521602630615, "learning_rate": 2.0652221061864152e-05, "loss": 0.0223, "step": 71470 }, { "grad_norm": 0.4755634069442749, "learning_rate": 2.0638835858738226e-05, "loss": 0.0283, "step": 71480 }, { "grad_norm": 0.3298548460006714, "learning_rate": 2.0625453866505068e-05, "loss": 0.0227, "step": 71490 }, { "grad_norm": 0.22263869643211365, "learning_rate": 2.0612075086628123e-05, "loss": 0.0223, "step": 71500 }, { "grad_norm": 0.43772050738334656, "learning_rate": 2.059869952057045e-05, "loss": 0.0231, "step": 71510 }, { "grad_norm": 0.2115958034992218, "learning_rate": 2.0585327169794793e-05, "loss": 0.0211, "step": 71520 }, { "grad_norm": 0.2709663212299347, "learning_rate": 2.0571958035763534e-05, "loss": 0.0259, "step": 71530 }, { "grad_norm": 0.28960222005844116, "learning_rate": 2.0558592119938715e-05, "loss": 0.0266, "step": 71540 }, { "grad_norm": 0.27975836396217346, "learning_rate": 2.054522942378198e-05, "loss": 0.0229, "step": 71550 }, { "grad_norm": 0.24016883969306946, "learning_rate": 2.053186994875469e-05, "loss": 0.0251, "step": 71560 }, { "grad_norm": 0.265331506729126, "learning_rate": 2.0518513696317775e-05, "loss": 0.0249, "step": 71570 }, { "grad_norm": 0.2146628499031067, "learning_rate": 2.050516066793188e-05, "loss": 0.0228, "step": 71580 }, { "grad_norm": 0.3065416216850281, "learning_rate": 2.0491810865057276e-05, "loss": 0.0279, "step": 71590 }, { "grad_norm": 0.2787626087665558, "learning_rate": 2.0478464289153882e-05, "loss": 0.0266, "step": 71600 }, { "grad_norm": 0.37050285935401917, "learning_rate": 2.0465120941681232e-05, "loss": 0.0224, "step": 71610 }, { "grad_norm": 0.23595920205116272, "learning_rate": 2.0451780824098566e-05, "loss": 0.0259, "step": 71620 }, { "grad_norm": 0.3620372414588928, "learning_rate": 2.043844393786471e-05, "loss": 0.0229, "step": 71630 }, { "grad_norm": 0.44732409715652466, "learning_rate": 2.0425110284438193e-05, "loss": 0.0244, "step": 71640 }, { "grad_norm": 0.584704577922821, "learning_rate": 2.041177986527712e-05, "loss": 0.0255, "step": 71650 }, { "grad_norm": 0.31847718358039856, "learning_rate": 2.0398452681839347e-05, "loss": 0.0275, "step": 71660 }, { "grad_norm": 0.5997727513313293, "learning_rate": 2.0385128735582272e-05, "loss": 0.0198, "step": 71670 }, { "grad_norm": 0.3344055116176605, "learning_rate": 2.0371808027963003e-05, "loss": 0.0281, "step": 71680 }, { "grad_norm": 0.259209007024765, "learning_rate": 2.035849056043825e-05, "loss": 0.028, "step": 71690 }, { "grad_norm": 0.3017269968986511, "learning_rate": 2.03451763344644e-05, "loss": 0.027, "step": 71700 }, { "grad_norm": 0.2683965861797333, "learning_rate": 2.0331865351497482e-05, "loss": 0.0259, "step": 71710 }, { "grad_norm": 0.2868134379386902, "learning_rate": 2.0318557612993185e-05, "loss": 0.0235, "step": 71720 }, { "grad_norm": 0.7373490333557129, "learning_rate": 2.0305253120406782e-05, "loss": 0.0258, "step": 71730 }, { "grad_norm": 0.4113517105579376, "learning_rate": 2.029195187519325e-05, "loss": 0.0262, "step": 71740 }, { "grad_norm": 0.3635060489177704, "learning_rate": 2.0278653878807217e-05, "loss": 0.0196, "step": 71750 }, { "grad_norm": 0.405901163816452, "learning_rate": 2.0265359132702892e-05, "loss": 0.0263, "step": 71760 }, { "grad_norm": 0.32821372151374817, "learning_rate": 2.0252067638334182e-05, "loss": 0.0242, "step": 71770 }, { "grad_norm": 0.2747848927974701, "learning_rate": 2.023877939715464e-05, "loss": 0.0256, "step": 71780 }, { "grad_norm": 0.35808905959129333, "learning_rate": 2.0225494410617448e-05, "loss": 0.0212, "step": 71790 }, { "grad_norm": 0.3078625500202179, "learning_rate": 2.02122126801754e-05, "loss": 0.0241, "step": 71800 }, { "grad_norm": 0.3468334674835205, "learning_rate": 2.0198934207281018e-05, "loss": 0.028, "step": 71810 }, { "grad_norm": 0.29160386323928833, "learning_rate": 2.0185658993386357e-05, "loss": 0.0245, "step": 71820 }, { "grad_norm": 0.19352976977825165, "learning_rate": 2.0172387039943226e-05, "loss": 0.0236, "step": 71830 }, { "grad_norm": 0.6447402834892273, "learning_rate": 2.0159118348402968e-05, "loss": 0.0221, "step": 71840 }, { "grad_norm": 0.3471059799194336, "learning_rate": 2.0145852920216697e-05, "loss": 0.0259, "step": 71850 }, { "grad_norm": 0.32904765009880066, "learning_rate": 2.0132590756835045e-05, "loss": 0.026, "step": 71860 }, { "grad_norm": 0.2891179919242859, "learning_rate": 2.0119331859708386e-05, "loss": 0.0239, "step": 71870 }, { "grad_norm": 0.3672860562801361, "learning_rate": 2.0106076230286654e-05, "loss": 0.0239, "step": 71880 }, { "grad_norm": 0.2587907314300537, "learning_rate": 2.0092823870019488e-05, "loss": 0.0234, "step": 71890 }, { "grad_norm": 0.2742649018764496, "learning_rate": 2.0079574780356115e-05, "loss": 0.0213, "step": 71900 }, { "grad_norm": 0.5374442934989929, "learning_rate": 2.0066328962745486e-05, "loss": 0.0296, "step": 71910 }, { "grad_norm": 0.27091601490974426, "learning_rate": 2.00530864186361e-05, "loss": 0.0225, "step": 71920 }, { "grad_norm": 0.2826148569583893, "learning_rate": 2.003984714947618e-05, "loss": 0.0245, "step": 71930 }, { "grad_norm": 0.366053968667984, "learning_rate": 2.0026611156713515e-05, "loss": 0.0244, "step": 71940 }, { "grad_norm": 0.4045426845550537, "learning_rate": 2.0013378441795584e-05, "loss": 0.0281, "step": 71950 }, { "grad_norm": 0.3635939061641693, "learning_rate": 2.0000149006169506e-05, "loss": 0.0273, "step": 71960 }, { "grad_norm": 0.2533029317855835, "learning_rate": 1.9986922851282048e-05, "loss": 0.0271, "step": 71970 }, { "grad_norm": 0.2656193673610687, "learning_rate": 1.997369997857956e-05, "loss": 0.026, "step": 71980 }, { "grad_norm": 0.326694518327713, "learning_rate": 1.9960480389508103e-05, "loss": 0.0206, "step": 71990 }, { "grad_norm": 0.25400155782699585, "learning_rate": 1.9947264085513363e-05, "loss": 0.0252, "step": 72000 }, { "grad_norm": 0.2681862413883209, "learning_rate": 1.993405106804062e-05, "loss": 0.0222, "step": 72010 }, { "grad_norm": 0.3594294488430023, "learning_rate": 1.9920841338534867e-05, "loss": 0.0238, "step": 72020 }, { "grad_norm": 0.3473467528820038, "learning_rate": 1.9907634898440646e-05, "loss": 0.022, "step": 72030 }, { "grad_norm": 0.32618361711502075, "learning_rate": 1.989443174920227e-05, "loss": 0.0253, "step": 72040 }, { "grad_norm": 0.23459038138389587, "learning_rate": 1.988123189226355e-05, "loss": 0.0271, "step": 72050 }, { "grad_norm": 0.5203304290771484, "learning_rate": 1.986803532906805e-05, "loss": 0.0278, "step": 72060 }, { "grad_norm": 0.39823514223098755, "learning_rate": 1.985484206105889e-05, "loss": 0.0225, "step": 72070 }, { "grad_norm": 0.42344462871551514, "learning_rate": 1.9841652089678887e-05, "loss": 0.0298, "step": 72080 }, { "grad_norm": 0.3047083020210266, "learning_rate": 1.9828465416370435e-05, "loss": 0.026, "step": 72090 }, { "grad_norm": 0.30577802658081055, "learning_rate": 1.9815282042575685e-05, "loss": 0.0229, "step": 72100 }, { "grad_norm": 0.5998049974441528, "learning_rate": 1.9802101969736285e-05, "loss": 0.0293, "step": 72110 }, { "grad_norm": 0.30897748470306396, "learning_rate": 1.9788925199293633e-05, "loss": 0.0254, "step": 72120 }, { "grad_norm": 0.31929266452789307, "learning_rate": 1.977575173268868e-05, "loss": 0.027, "step": 72130 }, { "grad_norm": 0.5071617364883423, "learning_rate": 1.9762581571362094e-05, "loss": 0.0251, "step": 72140 }, { "grad_norm": 0.23969490826129913, "learning_rate": 1.9749414716754112e-05, "loss": 0.0241, "step": 72150 }, { "grad_norm": 0.6864398121833801, "learning_rate": 1.973625117030465e-05, "loss": 0.0216, "step": 72160 }, { "grad_norm": 0.2625255286693573, "learning_rate": 1.972309093345326e-05, "loss": 0.0225, "step": 72170 }, { "grad_norm": 0.34163394570350647, "learning_rate": 1.970993400763915e-05, "loss": 0.0248, "step": 72180 }, { "grad_norm": 0.5630565881729126, "learning_rate": 1.969678039430109e-05, "loss": 0.0245, "step": 72190 }, { "grad_norm": 0.2429341971874237, "learning_rate": 1.9683630094877564e-05, "loss": 0.0296, "step": 72200 }, { "grad_norm": 0.3035077750682831, "learning_rate": 1.9670483110806696e-05, "loss": 0.0238, "step": 72210 }, { "grad_norm": 0.44066181778907776, "learning_rate": 1.9657339443526172e-05, "loss": 0.025, "step": 72220 }, { "grad_norm": 0.4321458637714386, "learning_rate": 1.964419909447339e-05, "loss": 0.022, "step": 72230 }, { "grad_norm": 0.6703920364379883, "learning_rate": 1.963106206508536e-05, "loss": 0.0294, "step": 72240 }, { "grad_norm": 0.43983450531959534, "learning_rate": 1.9617928356798738e-05, "loss": 0.0246, "step": 72250 }, { "grad_norm": 0.3932640850543976, "learning_rate": 1.960479797104977e-05, "loss": 0.0217, "step": 72260 }, { "grad_norm": 0.44018274545669556, "learning_rate": 1.9591670909274418e-05, "loss": 0.0219, "step": 72270 }, { "grad_norm": 0.3519209027290344, "learning_rate": 1.9578547172908184e-05, "loss": 0.0245, "step": 72280 }, { "grad_norm": 0.23577378690242767, "learning_rate": 1.9565426763386325e-05, "loss": 0.0284, "step": 72290 }, { "grad_norm": 0.3904346823692322, "learning_rate": 1.9552309682143622e-05, "loss": 0.0286, "step": 72300 }, { "grad_norm": 0.452671617269516, "learning_rate": 1.9539195930614574e-05, "loss": 0.0258, "step": 72310 }, { "grad_norm": 0.29575327038764954, "learning_rate": 1.952608551023324e-05, "loss": 0.0218, "step": 72320 }, { "grad_norm": 0.6412233710289001, "learning_rate": 1.9512978422433397e-05, "loss": 0.0303, "step": 72330 }, { "grad_norm": 0.7053651809692383, "learning_rate": 1.9499874668648376e-05, "loss": 0.0209, "step": 72340 }, { "grad_norm": 0.3019300103187561, "learning_rate": 1.9486774250311213e-05, "loss": 0.0209, "step": 72350 }, { "grad_norm": 0.45314064621925354, "learning_rate": 1.9473677168854537e-05, "loss": 0.026, "step": 72360 }, { "grad_norm": 0.6594672203063965, "learning_rate": 1.946058342571065e-05, "loss": 0.0237, "step": 72370 }, { "grad_norm": 0.27588051557540894, "learning_rate": 1.9447493022311424e-05, "loss": 0.0237, "step": 72380 }, { "grad_norm": 0.2231636345386505, "learning_rate": 1.9434405960088442e-05, "loss": 0.0221, "step": 72390 }, { "grad_norm": 0.40640169382095337, "learning_rate": 1.9421322240472857e-05, "loss": 0.0241, "step": 72400 }, { "grad_norm": 0.29471614956855774, "learning_rate": 1.9408241864895494e-05, "loss": 0.0229, "step": 72410 }, { "grad_norm": 0.3482278883457184, "learning_rate": 1.939516483478681e-05, "loss": 0.0227, "step": 72420 }, { "grad_norm": 0.3730645775794983, "learning_rate": 1.9382091151576902e-05, "loss": 0.0198, "step": 72430 }, { "grad_norm": 1.0036005973815918, "learning_rate": 1.9369020816695454e-05, "loss": 0.0231, "step": 72440 }, { "grad_norm": 0.5149733424186707, "learning_rate": 1.935595383157184e-05, "loss": 0.0248, "step": 72450 }, { "grad_norm": 0.28599998354911804, "learning_rate": 1.9342890197635057e-05, "loss": 0.0224, "step": 72460 }, { "grad_norm": 0.41844621300697327, "learning_rate": 1.9329829916313685e-05, "loss": 0.0271, "step": 72470 }, { "grad_norm": 0.6959719061851501, "learning_rate": 1.9316772989036013e-05, "loss": 0.026, "step": 72480 }, { "grad_norm": 0.26249778270721436, "learning_rate": 1.930371941722991e-05, "loss": 0.0264, "step": 72490 }, { "grad_norm": 0.551937997341156, "learning_rate": 1.929066920232292e-05, "loss": 0.0219, "step": 72500 }, { "grad_norm": 0.7610813975334167, "learning_rate": 1.9277622345742152e-05, "loss": 0.0236, "step": 72510 }, { "grad_norm": 0.5006235837936401, "learning_rate": 1.926457884891444e-05, "loss": 0.0257, "step": 72520 }, { "grad_norm": 0.34737420082092285, "learning_rate": 1.9251538713266147e-05, "loss": 0.0225, "step": 72530 }, { "grad_norm": 0.23789745569229126, "learning_rate": 1.9238501940223358e-05, "loss": 0.0291, "step": 72540 }, { "grad_norm": 0.45825299620628357, "learning_rate": 1.922546853121175e-05, "loss": 0.0244, "step": 72550 }, { "grad_norm": 0.33033743500709534, "learning_rate": 1.9212438487656647e-05, "loss": 0.0206, "step": 72560 }, { "grad_norm": 0.3510056436061859, "learning_rate": 1.9199411810982965e-05, "loss": 0.0234, "step": 72570 }, { "grad_norm": 0.3653404116630554, "learning_rate": 1.918638850261532e-05, "loss": 0.024, "step": 72580 }, { "grad_norm": 0.5506674647331238, "learning_rate": 1.9173368563977885e-05, "loss": 0.0247, "step": 72590 }, { "grad_norm": 0.45661646127700806, "learning_rate": 1.9160351996494513e-05, "loss": 0.0229, "step": 72600 }, { "grad_norm": 0.5872125625610352, "learning_rate": 1.9147338801588677e-05, "loss": 0.0209, "step": 72610 }, { "grad_norm": 0.439781129360199, "learning_rate": 1.913432898068351e-05, "loss": 0.0282, "step": 72620 }, { "grad_norm": 0.38268449902534485, "learning_rate": 1.912132253520169e-05, "loss": 0.0237, "step": 72630 }, { "grad_norm": 0.4387826919555664, "learning_rate": 1.910831946656564e-05, "loss": 0.02, "step": 72640 }, { "grad_norm": 0.4907900094985962, "learning_rate": 1.909531977619731e-05, "loss": 0.0232, "step": 72650 }, { "grad_norm": 0.4264086186885834, "learning_rate": 1.908232346551834e-05, "loss": 0.0258, "step": 72660 }, { "grad_norm": 0.9036845564842224, "learning_rate": 1.9069330535949993e-05, "loss": 0.0269, "step": 72670 }, { "grad_norm": 0.27785545587539673, "learning_rate": 1.9056340988913167e-05, "loss": 0.0273, "step": 72680 }, { "grad_norm": 0.19371411204338074, "learning_rate": 1.9043354825828357e-05, "loss": 0.0229, "step": 72690 }, { "grad_norm": 0.9041423797607422, "learning_rate": 1.9030372048115724e-05, "loss": 0.0277, "step": 72700 }, { "grad_norm": 0.2960739731788635, "learning_rate": 1.9017392657195056e-05, "loss": 0.0219, "step": 72710 }, { "grad_norm": 0.2775256335735321, "learning_rate": 1.9004416654485724e-05, "loss": 0.0273, "step": 72720 }, { "grad_norm": 0.2730454206466675, "learning_rate": 1.8991444041406785e-05, "loss": 0.0216, "step": 72730 }, { "grad_norm": 0.37703412771224976, "learning_rate": 1.8978474819376912e-05, "loss": 0.024, "step": 72740 }, { "grad_norm": 0.5269631147384644, "learning_rate": 1.8965508989814406e-05, "loss": 0.0268, "step": 72750 }, { "grad_norm": 0.39784878492355347, "learning_rate": 1.8952546554137158e-05, "loss": 0.0245, "step": 72760 }, { "grad_norm": 0.3988281190395355, "learning_rate": 1.8939587513762757e-05, "loss": 0.0249, "step": 72770 }, { "grad_norm": 0.3535812497138977, "learning_rate": 1.892663187010835e-05, "loss": 0.0327, "step": 72780 }, { "grad_norm": 0.5456036925315857, "learning_rate": 1.891367962459077e-05, "loss": 0.0227, "step": 72790 }, { "grad_norm": 0.4782053232192993, "learning_rate": 1.890073077862644e-05, "loss": 0.0205, "step": 72800 }, { "grad_norm": 0.29839739203453064, "learning_rate": 1.888778533363146e-05, "loss": 0.0235, "step": 72810 }, { "grad_norm": 0.37643998861312866, "learning_rate": 1.8874843291021478e-05, "loss": 0.0243, "step": 72820 }, { "grad_norm": 0.3885578513145447, "learning_rate": 1.8861904652211853e-05, "loss": 0.0242, "step": 72830 }, { "grad_norm": 2.5919723510742188, "learning_rate": 1.8848969418617507e-05, "loss": 0.0247, "step": 72840 }, { "grad_norm": 0.27307289838790894, "learning_rate": 1.8836037591653044e-05, "loss": 0.0244, "step": 72850 }, { "grad_norm": 0.31338924169540405, "learning_rate": 1.882310917273262e-05, "loss": 0.026, "step": 72860 }, { "grad_norm": 0.2322508990764618, "learning_rate": 1.881018416327014e-05, "loss": 0.0201, "step": 72870 }, { "grad_norm": 0.22807297110557556, "learning_rate": 1.8797262564679014e-05, "loss": 0.0264, "step": 72880 }, { "grad_norm": 0.18804873526096344, "learning_rate": 1.8784344378372355e-05, "loss": 0.0212, "step": 72890 }, { "grad_norm": 0.6486497521400452, "learning_rate": 1.8771429605762843e-05, "loss": 0.0305, "step": 72900 }, { "grad_norm": 0.8055946230888367, "learning_rate": 1.8758518248262835e-05, "loss": 0.0248, "step": 72910 }, { "grad_norm": 0.440856009721756, "learning_rate": 1.8745610307284305e-05, "loss": 0.0205, "step": 72920 }, { "grad_norm": 0.36931824684143066, "learning_rate": 1.8732705784238862e-05, "loss": 0.0228, "step": 72930 }, { "grad_norm": 0.3627467453479767, "learning_rate": 1.8719804680537682e-05, "loss": 0.0219, "step": 72940 }, { "grad_norm": 0.48882779479026794, "learning_rate": 1.870690699759164e-05, "loss": 0.0227, "step": 72950 }, { "grad_norm": 0.7504833340644836, "learning_rate": 1.8694012736811213e-05, "loss": 0.0298, "step": 72960 }, { "grad_norm": 0.3295375406742096, "learning_rate": 1.8681121899606463e-05, "loss": 0.0263, "step": 72970 }, { "grad_norm": 0.333927720785141, "learning_rate": 1.8668234487387144e-05, "loss": 0.0247, "step": 72980 }, { "grad_norm": 0.34508216381073, "learning_rate": 1.8655350501562592e-05, "loss": 0.0239, "step": 72990 }, { "grad_norm": 0.6692547798156738, "learning_rate": 1.8642469943541796e-05, "loss": 0.0249, "step": 73000 }, { "grad_norm": 0.34118252992630005, "learning_rate": 1.8629592814733328e-05, "loss": 0.0232, "step": 73010 }, { "grad_norm": 0.2679862678050995, "learning_rate": 1.8616719116545438e-05, "loss": 0.0214, "step": 73020 }, { "grad_norm": 0.30584585666656494, "learning_rate": 1.8603848850385952e-05, "loss": 0.0222, "step": 73030 }, { "grad_norm": 0.22757788002490997, "learning_rate": 1.859098201766236e-05, "loss": 0.0257, "step": 73040 }, { "grad_norm": 0.3768504559993744, "learning_rate": 1.8578118619781726e-05, "loss": 0.0246, "step": 73050 }, { "grad_norm": 0.38255369663238525, "learning_rate": 1.856525865815083e-05, "loss": 0.0244, "step": 73060 }, { "grad_norm": 0.30756044387817383, "learning_rate": 1.8552402134175966e-05, "loss": 0.0247, "step": 73070 }, { "grad_norm": 0.4611373841762543, "learning_rate": 1.8539549049263143e-05, "loss": 0.0225, "step": 73080 }, { "grad_norm": 0.3469174802303314, "learning_rate": 1.8526699404817917e-05, "loss": 0.0221, "step": 73090 }, { "grad_norm": 0.35761788487434387, "learning_rate": 1.8513853202245546e-05, "loss": 0.0214, "step": 73100 }, { "grad_norm": 0.30287033319473267, "learning_rate": 1.850101044295081e-05, "loss": 0.0223, "step": 73110 }, { "grad_norm": 0.2929368317127228, "learning_rate": 1.848817112833825e-05, "loss": 0.0218, "step": 73120 }, { "grad_norm": 0.5546918511390686, "learning_rate": 1.8475335259811905e-05, "loss": 0.0252, "step": 73130 }, { "grad_norm": 0.26179739832878113, "learning_rate": 1.8462502838775508e-05, "loss": 0.0223, "step": 73140 }, { "grad_norm": 0.3153218626976013, "learning_rate": 1.8449673866632378e-05, "loss": 0.0278, "step": 73150 }, { "grad_norm": 0.38161996006965637, "learning_rate": 1.8436848344785473e-05, "loss": 0.0287, "step": 73160 }, { "grad_norm": 0.3749748170375824, "learning_rate": 1.8424026274637378e-05, "loss": 0.0273, "step": 73170 }, { "grad_norm": 0.4571489989757538, "learning_rate": 1.8411207657590312e-05, "loss": 0.0248, "step": 73180 }, { "grad_norm": 0.3074566721916199, "learning_rate": 1.839839249504607e-05, "loss": 0.026, "step": 73190 }, { "grad_norm": 0.2964816391468048, "learning_rate": 1.8385580788406108e-05, "loss": 0.025, "step": 73200 }, { "grad_norm": 0.28373634815216064, "learning_rate": 1.837277253907152e-05, "loss": 0.0242, "step": 73210 }, { "grad_norm": 0.35629305243492126, "learning_rate": 1.8359967748442957e-05, "loss": 0.0269, "step": 73220 }, { "grad_norm": 0.3089102804660797, "learning_rate": 1.8347166417920774e-05, "loss": 0.0281, "step": 73230 }, { "grad_norm": 0.313165545463562, "learning_rate": 1.833436854890484e-05, "loss": 0.0212, "step": 73240 }, { "grad_norm": 0.6225000023841858, "learning_rate": 1.8321574142794796e-05, "loss": 0.02, "step": 73250 }, { "grad_norm": 0.36762186884880066, "learning_rate": 1.8308783200989748e-05, "loss": 0.0234, "step": 73260 }, { "grad_norm": 0.2830173969268799, "learning_rate": 1.829599572488855e-05, "loss": 0.0224, "step": 73270 }, { "grad_norm": 0.47087204456329346, "learning_rate": 1.828321171588957e-05, "loss": 0.0217, "step": 73280 }, { "grad_norm": 0.3474941849708557, "learning_rate": 1.827043117539089e-05, "loss": 0.0258, "step": 73290 }, { "grad_norm": 0.39545267820358276, "learning_rate": 1.8257654104790116e-05, "loss": 0.0248, "step": 73300 }, { "grad_norm": 0.43386921286582947, "learning_rate": 1.82448805054846e-05, "loss": 0.0194, "step": 73310 }, { "grad_norm": 0.3832152187824249, "learning_rate": 1.823211037887119e-05, "loss": 0.0222, "step": 73320 }, { "grad_norm": 0.3279017210006714, "learning_rate": 1.821934372634645e-05, "loss": 0.0287, "step": 73330 }, { "grad_norm": 0.3077211380004883, "learning_rate": 1.8206580549306472e-05, "loss": 0.0235, "step": 73340 }, { "grad_norm": 0.2725926637649536, "learning_rate": 1.819382084914707e-05, "loss": 0.0246, "step": 73350 }, { "grad_norm": 0.2650521397590637, "learning_rate": 1.8181064627263577e-05, "loss": 0.023, "step": 73360 }, { "grad_norm": 0.9182790517807007, "learning_rate": 1.8168311885051022e-05, "loss": 0.027, "step": 73370 }, { "grad_norm": 0.2622566819190979, "learning_rate": 1.815556262390402e-05, "loss": 0.0238, "step": 73380 }, { "grad_norm": 0.46543410420417786, "learning_rate": 1.8142816845216826e-05, "loss": 0.0253, "step": 73390 }, { "grad_norm": 0.4306625723838806, "learning_rate": 1.8130074550383276e-05, "loss": 0.0233, "step": 73400 }, { "grad_norm": 0.2478322982788086, "learning_rate": 1.8117335740796848e-05, "loss": 0.0271, "step": 73410 }, { "grad_norm": 0.31113138794898987, "learning_rate": 1.810460041785067e-05, "loss": 0.0262, "step": 73420 }, { "grad_norm": 0.34930548071861267, "learning_rate": 1.809186858293742e-05, "loss": 0.0226, "step": 73430 }, { "grad_norm": 0.3997471332550049, "learning_rate": 1.8079140237449448e-05, "loss": 0.0277, "step": 73440 }, { "grad_norm": 0.33617058396339417, "learning_rate": 1.8066415382778707e-05, "loss": 0.0224, "step": 73450 }, { "grad_norm": 0.3008999824523926, "learning_rate": 1.8053694020316786e-05, "loss": 0.0217, "step": 73460 }, { "grad_norm": 0.35812339186668396, "learning_rate": 1.8040976151454837e-05, "loss": 0.0219, "step": 73470 }, { "grad_norm": 0.4283340275287628, "learning_rate": 1.8028261777583704e-05, "loss": 0.0205, "step": 73480 }, { "grad_norm": 0.3486768305301666, "learning_rate": 1.8015550900093758e-05, "loss": 0.0275, "step": 73490 }, { "grad_norm": 0.28757575154304504, "learning_rate": 1.800284352037512e-05, "loss": 0.0245, "step": 73500 }, { "grad_norm": 0.262684166431427, "learning_rate": 1.799013963981738e-05, "loss": 0.0277, "step": 73510 }, { "grad_norm": 0.26035016775131226, "learning_rate": 1.797743925980987e-05, "loss": 0.0262, "step": 73520 }, { "grad_norm": 0.3610457181930542, "learning_rate": 1.7964742381741435e-05, "loss": 0.0218, "step": 73530 }, { "grad_norm": 0.308790922164917, "learning_rate": 1.7952049007000627e-05, "loss": 0.0263, "step": 73540 }, { "grad_norm": 0.2886921465396881, "learning_rate": 1.7939359136975535e-05, "loss": 0.027, "step": 73550 }, { "grad_norm": 0.43763816356658936, "learning_rate": 1.7926672773053932e-05, "loss": 0.0241, "step": 73560 }, { "grad_norm": 0.3983011841773987, "learning_rate": 1.791398991662317e-05, "loss": 0.0218, "step": 73570 }, { "grad_norm": 0.4634645879268646, "learning_rate": 1.7901310569070244e-05, "loss": 0.0232, "step": 73580 }, { "grad_norm": 0.23202575743198395, "learning_rate": 1.7888634731781724e-05, "loss": 0.0245, "step": 73590 }, { "grad_norm": 0.47487887740135193, "learning_rate": 1.7875962406143844e-05, "loss": 0.0298, "step": 73600 }, { "grad_norm": 0.3290727436542511, "learning_rate": 1.7863293593542403e-05, "loss": 0.0233, "step": 73610 }, { "grad_norm": 0.2360890805721283, "learning_rate": 1.7850628295362853e-05, "loss": 0.0194, "step": 73620 }, { "grad_norm": 0.2699038088321686, "learning_rate": 1.7837966512990263e-05, "loss": 0.0227, "step": 73630 }, { "grad_norm": 0.2897210419178009, "learning_rate": 1.782530824780931e-05, "loss": 0.0272, "step": 73640 }, { "grad_norm": 0.40719425678253174, "learning_rate": 1.7812653501204263e-05, "loss": 0.0203, "step": 73650 }, { "grad_norm": 0.31169724464416504, "learning_rate": 1.780000227455903e-05, "loss": 0.0256, "step": 73660 }, { "grad_norm": 0.2723815441131592, "learning_rate": 1.778735456925716e-05, "loss": 0.0236, "step": 73670 }, { "grad_norm": 0.26064079999923706, "learning_rate": 1.7774710386681737e-05, "loss": 0.0247, "step": 73680 }, { "grad_norm": 0.5185626745223999, "learning_rate": 1.7762069728215545e-05, "loss": 0.0275, "step": 73690 }, { "grad_norm": 0.2523815631866455, "learning_rate": 1.7749432595240932e-05, "loss": 0.0232, "step": 73700 }, { "grad_norm": 0.6269403696060181, "learning_rate": 1.773679898913991e-05, "loss": 0.0226, "step": 73710 }, { "grad_norm": 0.4226517081260681, "learning_rate": 1.7724168911294015e-05, "loss": 0.0273, "step": 73720 }, { "grad_norm": 0.27333539724349976, "learning_rate": 1.7711542363084504e-05, "loss": 0.0234, "step": 73730 }, { "grad_norm": 0.44761523604393005, "learning_rate": 1.7698919345892158e-05, "loss": 0.0196, "step": 73740 }, { "grad_norm": 0.4686349630355835, "learning_rate": 1.768629986109743e-05, "loss": 0.0238, "step": 73750 }, { "grad_norm": 0.25878000259399414, "learning_rate": 1.7673683910080357e-05, "loss": 0.0246, "step": 73760 }, { "grad_norm": 0.423471063375473, "learning_rate": 1.7661071494220638e-05, "loss": 0.0232, "step": 73770 }, { "grad_norm": 0.2574974000453949, "learning_rate": 1.7648462614897498e-05, "loss": 0.0223, "step": 73780 }, { "grad_norm": 0.9808139204978943, "learning_rate": 1.7635857273489863e-05, "loss": 0.027, "step": 73790 }, { "grad_norm": 0.31784912943840027, "learning_rate": 1.7623255471376198e-05, "loss": 0.0224, "step": 73800 }, { "grad_norm": 0.2629607021808624, "learning_rate": 1.7610657209934646e-05, "loss": 0.0229, "step": 73810 }, { "grad_norm": 0.3417867124080658, "learning_rate": 1.7598062490542915e-05, "loss": 0.0268, "step": 73820 }, { "grad_norm": 0.41852691769599915, "learning_rate": 1.7585471314578377e-05, "loss": 0.0239, "step": 73830 }, { "grad_norm": 0.27515941858291626, "learning_rate": 1.7572883683417945e-05, "loss": 0.0239, "step": 73840 }, { "grad_norm": 0.3690643310546875, "learning_rate": 1.7560299598438212e-05, "loss": 0.0275, "step": 73850 }, { "grad_norm": 0.40307867527008057, "learning_rate": 1.7547719061015334e-05, "loss": 0.0223, "step": 73860 }, { "grad_norm": 0.24492117762565613, "learning_rate": 1.75351420725251e-05, "loss": 0.0215, "step": 73870 }, { "grad_norm": 0.35417938232421875, "learning_rate": 1.7522568634342923e-05, "loss": 0.0275, "step": 73880 }, { "grad_norm": 0.3511885702610016, "learning_rate": 1.7509998747843825e-05, "loss": 0.027, "step": 73890 }, { "grad_norm": 0.40194371342658997, "learning_rate": 1.7497432414402403e-05, "loss": 0.0286, "step": 73900 }, { "grad_norm": 1.0898754596710205, "learning_rate": 1.7484869635392903e-05, "loss": 0.0289, "step": 73910 }, { "grad_norm": 0.29113301634788513, "learning_rate": 1.747231041218919e-05, "loss": 0.0296, "step": 73920 }, { "grad_norm": 0.42771708965301514, "learning_rate": 1.7459754746164692e-05, "loss": 0.0294, "step": 73930 }, { "grad_norm": 0.26391497254371643, "learning_rate": 1.744720263869249e-05, "loss": 0.024, "step": 73940 }, { "grad_norm": 0.3404898941516876, "learning_rate": 1.7434654091145263e-05, "loss": 0.0258, "step": 73950 }, { "grad_norm": 0.22392718493938446, "learning_rate": 1.742210910489532e-05, "loss": 0.0213, "step": 73960 }, { "grad_norm": 0.2644776999950409, "learning_rate": 1.7409567681314532e-05, "loss": 0.0193, "step": 73970 }, { "grad_norm": 0.37983694672584534, "learning_rate": 1.7397029821774434e-05, "loss": 0.0239, "step": 73980 }, { "grad_norm": 0.4085520803928375, "learning_rate": 1.7384495527646126e-05, "loss": 0.0261, "step": 73990 }, { "grad_norm": 0.3145223557949066, "learning_rate": 1.737196480030035e-05, "loss": 0.0217, "step": 74000 }, { "grad_norm": 0.217180997133255, "learning_rate": 1.7359437641107447e-05, "loss": 0.0246, "step": 74010 }, { "grad_norm": 0.3792514204978943, "learning_rate": 1.7346914051437387e-05, "loss": 0.0237, "step": 74020 }, { "grad_norm": 0.3069063723087311, "learning_rate": 1.733439403265969e-05, "loss": 0.0221, "step": 74030 }, { "grad_norm": 0.34976500272750854, "learning_rate": 1.732187758614357e-05, "loss": 0.0184, "step": 74040 }, { "grad_norm": 0.4120343029499054, "learning_rate": 1.730936471325777e-05, "loss": 0.0231, "step": 74050 }, { "grad_norm": 0.3573293387889862, "learning_rate": 1.7296855415370718e-05, "loss": 0.0214, "step": 74060 }, { "grad_norm": 0.3406585454940796, "learning_rate": 1.728434969385035e-05, "loss": 0.0317, "step": 74070 }, { "grad_norm": 0.5011216998100281, "learning_rate": 1.7271847550064346e-05, "loss": 0.0232, "step": 74080 }, { "grad_norm": 0.3162759244441986, "learning_rate": 1.725934898537987e-05, "loss": 0.0248, "step": 74090 }, { "grad_norm": 0.793066680431366, "learning_rate": 1.7246854001163783e-05, "loss": 0.021, "step": 74100 }, { "grad_norm": 0.3111281096935272, "learning_rate": 1.7234362598782476e-05, "loss": 0.0206, "step": 74110 }, { "grad_norm": 0.38161253929138184, "learning_rate": 1.722187477960202e-05, "loss": 0.0295, "step": 74120 }, { "grad_norm": 0.8740904331207275, "learning_rate": 1.7209390544988057e-05, "loss": 0.0283, "step": 74130 }, { "grad_norm": 0.302852064371109, "learning_rate": 1.7196909896305856e-05, "loss": 0.0255, "step": 74140 }, { "grad_norm": 0.2247660905122757, "learning_rate": 1.718443283492026e-05, "loss": 0.0212, "step": 74150 }, { "grad_norm": 0.43968334794044495, "learning_rate": 1.7171959362195746e-05, "loss": 0.023, "step": 74160 }, { "grad_norm": 0.39569345116615295, "learning_rate": 1.715948947949642e-05, "loss": 0.0257, "step": 74170 }, { "grad_norm": 0.30511775612831116, "learning_rate": 1.714702318818593e-05, "loss": 0.0305, "step": 74180 }, { "grad_norm": 0.41331085562705994, "learning_rate": 1.7134560489627592e-05, "loss": 0.023, "step": 74190 }, { "grad_norm": 0.28843045234680176, "learning_rate": 1.712210138518431e-05, "loss": 0.0262, "step": 74200 }, { "grad_norm": 0.4669901728630066, "learning_rate": 1.71096458762186e-05, "loss": 0.0235, "step": 74210 }, { "grad_norm": 0.3761461675167084, "learning_rate": 1.7097193964092552e-05, "loss": 0.0195, "step": 74220 }, { "grad_norm": 0.33792150020599365, "learning_rate": 1.708474565016792e-05, "loss": 0.0235, "step": 74230 }, { "grad_norm": 0.3816910982131958, "learning_rate": 1.7072300935806e-05, "loss": 0.0211, "step": 74240 }, { "grad_norm": 0.2811123728752136, "learning_rate": 1.705985982236776e-05, "loss": 0.0231, "step": 74250 }, { "grad_norm": 0.2196705937385559, "learning_rate": 1.7047422311213697e-05, "loss": 0.0233, "step": 74260 }, { "grad_norm": 0.35728558897972107, "learning_rate": 1.703498840370401e-05, "loss": 0.0242, "step": 74270 }, { "grad_norm": 0.5391125679016113, "learning_rate": 1.702255810119842e-05, "loss": 0.0188, "step": 74280 }, { "grad_norm": 0.1900080442428589, "learning_rate": 1.7010131405056305e-05, "loss": 0.0241, "step": 74290 }, { "grad_norm": 0.410948246717453, "learning_rate": 1.6997708316636613e-05, "loss": 0.0271, "step": 74300 }, { "grad_norm": 0.41309303045272827, "learning_rate": 1.6985288837297932e-05, "loss": 0.0218, "step": 74310 }, { "grad_norm": 0.6167073845863342, "learning_rate": 1.6972872968398394e-05, "loss": 0.0229, "step": 74320 }, { "grad_norm": 0.6785249710083008, "learning_rate": 1.6960460711295845e-05, "loss": 0.0256, "step": 74330 }, { "grad_norm": 0.47339433431625366, "learning_rate": 1.6948052067347623e-05, "loss": 0.0217, "step": 74340 }, { "grad_norm": 0.5794737339019775, "learning_rate": 1.693564703791075e-05, "loss": 0.0217, "step": 74350 }, { "grad_norm": 0.4434860348701477, "learning_rate": 1.6923245624341793e-05, "loss": 0.0239, "step": 74360 }, { "grad_norm": 0.37022650241851807, "learning_rate": 1.6910847827996962e-05, "loss": 0.0267, "step": 74370 }, { "grad_norm": 0.28809526562690735, "learning_rate": 1.689845365023206e-05, "loss": 0.0252, "step": 74380 }, { "grad_norm": 0.43500977754592896, "learning_rate": 1.6886063092402505e-05, "loss": 0.0243, "step": 74390 }, { "grad_norm": 0.3345162570476532, "learning_rate": 1.6873676155863317e-05, "loss": 0.0238, "step": 74400 }, { "grad_norm": 0.6222532987594604, "learning_rate": 1.6861292841969085e-05, "loss": 0.0247, "step": 74410 }, { "grad_norm": 0.29999610781669617, "learning_rate": 1.684891315207406e-05, "loss": 0.0236, "step": 74420 }, { "grad_norm": 0.6001031398773193, "learning_rate": 1.6836537087532034e-05, "loss": 0.0239, "step": 74430 }, { "grad_norm": 0.5800351500511169, "learning_rate": 1.6824164649696465e-05, "loss": 0.0236, "step": 74440 }, { "grad_norm": 0.3075854182243347, "learning_rate": 1.681179583992034e-05, "loss": 0.0298, "step": 74450 }, { "grad_norm": 0.2697696089744568, "learning_rate": 1.679943065955636e-05, "loss": 0.0217, "step": 74460 }, { "grad_norm": 0.23066382110118866, "learning_rate": 1.678706910995671e-05, "loss": 0.0235, "step": 74470 }, { "grad_norm": 0.22018185257911682, "learning_rate": 1.6774711192473262e-05, "loss": 0.027, "step": 74480 }, { "grad_norm": 0.3505617678165436, "learning_rate": 1.676235690845743e-05, "loss": 0.0207, "step": 74490 }, { "grad_norm": 0.2964279055595398, "learning_rate": 1.6750006259260285e-05, "loss": 0.0221, "step": 74500 }, { "grad_norm": 0.44599854946136475, "learning_rate": 1.673765924623244e-05, "loss": 0.0243, "step": 74510 }, { "grad_norm": 0.43674376606941223, "learning_rate": 1.67253158707242e-05, "loss": 0.0219, "step": 74520 }, { "grad_norm": 0.3376162648200989, "learning_rate": 1.671297613408537e-05, "loss": 0.0233, "step": 74530 }, { "grad_norm": 0.3841089606285095, "learning_rate": 1.6700640037665444e-05, "loss": 0.0223, "step": 74540 }, { "grad_norm": 0.491922527551651, "learning_rate": 1.668830758281344e-05, "loss": 0.0223, "step": 74550 }, { "grad_norm": 0.42782750725746155, "learning_rate": 1.6675978770878052e-05, "loss": 0.025, "step": 74560 }, { "grad_norm": 0.22767524421215057, "learning_rate": 1.6663653603207506e-05, "loss": 0.0231, "step": 74570 }, { "grad_norm": 0.33842167258262634, "learning_rate": 1.6651332081149685e-05, "loss": 0.0226, "step": 74580 }, { "grad_norm": 0.24737434089183807, "learning_rate": 1.6639014206052046e-05, "loss": 0.0238, "step": 74590 }, { "grad_norm": 0.5778904557228088, "learning_rate": 1.6626699979261657e-05, "loss": 0.0274, "step": 74600 }, { "grad_norm": 0.3735724985599518, "learning_rate": 1.6614389402125197e-05, "loss": 0.0201, "step": 74610 }, { "grad_norm": 0.24867737293243408, "learning_rate": 1.66020824759889e-05, "loss": 0.0257, "step": 74620 }, { "grad_norm": 0.601478099822998, "learning_rate": 1.6589779202198675e-05, "loss": 0.0245, "step": 74630 }, { "grad_norm": 0.6217794418334961, "learning_rate": 1.6577479582099924e-05, "loss": 0.0301, "step": 74640 }, { "grad_norm": 0.4582340717315674, "learning_rate": 1.6565183617037788e-05, "loss": 0.0303, "step": 74650 }, { "grad_norm": 0.4085732698440552, "learning_rate": 1.6552891308356894e-05, "loss": 0.0235, "step": 74660 }, { "grad_norm": 0.25570663809776306, "learning_rate": 1.6540602657401526e-05, "loss": 0.0214, "step": 74670 }, { "grad_norm": 0.3473706841468811, "learning_rate": 1.6528317665515537e-05, "loss": 0.02, "step": 74680 }, { "grad_norm": 0.677317202091217, "learning_rate": 1.6516036334042416e-05, "loss": 0.0242, "step": 74690 }, { "grad_norm": 0.4864552617073059, "learning_rate": 1.650375866432519e-05, "loss": 0.0221, "step": 74700 }, { "grad_norm": 0.37669074535369873, "learning_rate": 1.649148465770659e-05, "loss": 0.0224, "step": 74710 }, { "grad_norm": 0.4016841650009155, "learning_rate": 1.647921431552883e-05, "loss": 0.0294, "step": 74720 }, { "grad_norm": 0.2913592755794525, "learning_rate": 1.646694763913381e-05, "loss": 0.0228, "step": 74730 }, { "grad_norm": 0.35840481519699097, "learning_rate": 1.6454684629862966e-05, "loss": 0.0211, "step": 74740 }, { "grad_norm": 0.29713302850723267, "learning_rate": 1.6442425289057388e-05, "loss": 0.023, "step": 74750 }, { "grad_norm": 0.3159884214401245, "learning_rate": 1.6430169618057718e-05, "loss": 0.0254, "step": 74760 }, { "grad_norm": 0.4182262420654297, "learning_rate": 1.641791761820422e-05, "loss": 0.0253, "step": 74770 }, { "grad_norm": 0.4339064359664917, "learning_rate": 1.640566929083676e-05, "loss": 0.0249, "step": 74780 }, { "grad_norm": 0.4915742576122284, "learning_rate": 1.6393424637294823e-05, "loss": 0.0239, "step": 74790 }, { "grad_norm": 0.5379935503005981, "learning_rate": 1.638118365891742e-05, "loss": 0.0243, "step": 74800 }, { "grad_norm": 0.298493355512619, "learning_rate": 1.636894635704323e-05, "loss": 0.0234, "step": 74810 }, { "grad_norm": 0.2464107722043991, "learning_rate": 1.6356712733010517e-05, "loss": 0.0234, "step": 74820 }, { "grad_norm": 0.24873241782188416, "learning_rate": 1.6344482788157096e-05, "loss": 0.0194, "step": 74830 }, { "grad_norm": 0.3699325919151306, "learning_rate": 1.6332256523820444e-05, "loss": 0.021, "step": 74840 }, { "grad_norm": 0.3808042109012604, "learning_rate": 1.6320033941337597e-05, "loss": 0.0229, "step": 74850 }, { "grad_norm": 0.31653791666030884, "learning_rate": 1.6307815042045216e-05, "loss": 0.022, "step": 74860 }, { "grad_norm": 1.1275473833084106, "learning_rate": 1.6295599827279513e-05, "loss": 0.0241, "step": 74870 }, { "grad_norm": 0.5376415252685547, "learning_rate": 1.6283388298376356e-05, "loss": 0.0243, "step": 74880 }, { "grad_norm": 0.3054843246936798, "learning_rate": 1.6271180456671125e-05, "loss": 0.022, "step": 74890 }, { "grad_norm": 0.4767642319202423, "learning_rate": 1.625897630349893e-05, "loss": 0.0224, "step": 74900 }, { "grad_norm": 0.2584119141101837, "learning_rate": 1.6246775840194335e-05, "loss": 0.0289, "step": 74910 }, { "grad_norm": 0.30827832221984863, "learning_rate": 1.6234579068091604e-05, "loss": 0.0267, "step": 74920 }, { "grad_norm": 0.3379507064819336, "learning_rate": 1.6222385988524524e-05, "loss": 0.0234, "step": 74930 }, { "grad_norm": 0.5131978988647461, "learning_rate": 1.6210196602826544e-05, "loss": 0.0308, "step": 74940 }, { "grad_norm": 0.53190678358078, "learning_rate": 1.6198010912330646e-05, "loss": 0.0243, "step": 74950 }, { "grad_norm": 0.37144970893859863, "learning_rate": 1.6185828918369456e-05, "loss": 0.0243, "step": 74960 }, { "grad_norm": 0.24843758344650269, "learning_rate": 1.617365062227517e-05, "loss": 0.0249, "step": 74970 }, { "grad_norm": 0.24684438109397888, "learning_rate": 1.616147602537962e-05, "loss": 0.024, "step": 74980 }, { "grad_norm": 0.31505680084228516, "learning_rate": 1.614930512901415e-05, "loss": 0.0227, "step": 74990 }, { "grad_norm": 0.27823737263679504, "learning_rate": 1.6137137934509804e-05, "loss": 0.025, "step": 75000 }, { "grad_norm": 0.7007945775985718, "learning_rate": 1.6124974443197123e-05, "loss": 0.0278, "step": 75010 }, { "grad_norm": 0.3717256784439087, "learning_rate": 1.6112814656406306e-05, "loss": 0.0268, "step": 75020 }, { "grad_norm": 0.7336598038673401, "learning_rate": 1.6100658575467125e-05, "loss": 0.0311, "step": 75030 }, { "grad_norm": 0.29247650504112244, "learning_rate": 1.6088506201708974e-05, "loss": 0.022, "step": 75040 }, { "grad_norm": 0.28194448351860046, "learning_rate": 1.6076357536460785e-05, "loss": 0.0219, "step": 75050 }, { "grad_norm": 0.28142476081848145, "learning_rate": 1.606421258105113e-05, "loss": 0.0227, "step": 75060 }, { "grad_norm": 0.6224523186683655, "learning_rate": 1.6052071336808183e-05, "loss": 0.025, "step": 75070 }, { "grad_norm": 0.4228598475456238, "learning_rate": 1.6039933805059658e-05, "loss": 0.0188, "step": 75080 }, { "grad_norm": 0.3134070336818695, "learning_rate": 1.6027799987132915e-05, "loss": 0.0239, "step": 75090 }, { "grad_norm": 0.5486074686050415, "learning_rate": 1.6015669884354888e-05, "loss": 0.0217, "step": 75100 }, { "grad_norm": 0.39935317635536194, "learning_rate": 1.6003543498052125e-05, "loss": 0.0214, "step": 75110 }, { "grad_norm": 0.2865619957447052, "learning_rate": 1.5991420829550715e-05, "loss": 0.0222, "step": 75120 }, { "grad_norm": 0.314096063375473, "learning_rate": 1.5979301880176405e-05, "loss": 0.027, "step": 75130 }, { "grad_norm": 0.3864959180355072, "learning_rate": 1.5967186651254485e-05, "loss": 0.0248, "step": 75140 }, { "grad_norm": 0.30603352189064026, "learning_rate": 1.5955075144109853e-05, "loss": 0.0223, "step": 75150 }, { "grad_norm": 0.35222095251083374, "learning_rate": 1.594296736006703e-05, "loss": 0.026, "step": 75160 }, { "grad_norm": 0.3440001904964447, "learning_rate": 1.59308633004501e-05, "loss": 0.0227, "step": 75170 }, { "grad_norm": 0.5443105697631836, "learning_rate": 1.5918762966582724e-05, "loss": 0.0226, "step": 75180 }, { "grad_norm": 0.25490453839302063, "learning_rate": 1.5906666359788203e-05, "loss": 0.0184, "step": 75190 }, { "grad_norm": 0.4708441197872162, "learning_rate": 1.589457348138938e-05, "loss": 0.0243, "step": 75200 }, { "grad_norm": 0.3198142349720001, "learning_rate": 1.588248433270872e-05, "loss": 0.0204, "step": 75210 }, { "grad_norm": 0.1424100250005722, "learning_rate": 1.587039891506828e-05, "loss": 0.0231, "step": 75220 }, { "grad_norm": 0.274719774723053, "learning_rate": 1.5858317229789715e-05, "loss": 0.023, "step": 75230 }, { "grad_norm": 0.32673904299736023, "learning_rate": 1.584623927819423e-05, "loss": 0.0242, "step": 75240 }, { "grad_norm": 0.5063863396644592, "learning_rate": 1.583416506160269e-05, "loss": 0.0242, "step": 75250 }, { "grad_norm": 0.3656681180000305, "learning_rate": 1.5822094581335472e-05, "loss": 0.0199, "step": 75260 }, { "grad_norm": 0.20514915883541107, "learning_rate": 1.5810027838712604e-05, "loss": 0.023, "step": 75270 }, { "grad_norm": 0.32105353474617004, "learning_rate": 1.5797964835053687e-05, "loss": 0.0218, "step": 75280 }, { "grad_norm": 0.45519372820854187, "learning_rate": 1.5785905571677933e-05, "loss": 0.025, "step": 75290 }, { "grad_norm": 0.43202850222587585, "learning_rate": 1.5773850049904086e-05, "loss": 0.0248, "step": 75300 }, { "grad_norm": 0.38803941011428833, "learning_rate": 1.5761798271050543e-05, "loss": 0.0214, "step": 75310 }, { "grad_norm": 0.8479677438735962, "learning_rate": 1.5749750236435277e-05, "loss": 0.0279, "step": 75320 }, { "grad_norm": 0.3452187478542328, "learning_rate": 1.5737705947375815e-05, "loss": 0.0235, "step": 75330 }, { "grad_norm": 0.39210355281829834, "learning_rate": 1.5725665405189322e-05, "loss": 0.0248, "step": 75340 }, { "grad_norm": 0.36697739362716675, "learning_rate": 1.571362861119253e-05, "loss": 0.0271, "step": 75350 }, { "grad_norm": 0.4647171199321747, "learning_rate": 1.570159556670177e-05, "loss": 0.0293, "step": 75360 }, { "grad_norm": 0.3387574851512909, "learning_rate": 1.5689566273032946e-05, "loss": 0.0217, "step": 75370 }, { "grad_norm": 0.48415425419807434, "learning_rate": 1.5677540731501583e-05, "loss": 0.0275, "step": 75380 }, { "grad_norm": 0.31543388962745667, "learning_rate": 1.5665518943422746e-05, "loss": 0.0241, "step": 75390 }, { "grad_norm": 0.4499664008617401, "learning_rate": 1.5653500910111134e-05, "loss": 0.0227, "step": 75400 }, { "grad_norm": 0.2755969166755676, "learning_rate": 1.5641486632881023e-05, "loss": 0.0201, "step": 75410 }, { "grad_norm": 0.39366400241851807, "learning_rate": 1.56294761130463e-05, "loss": 0.0202, "step": 75420 }, { "grad_norm": 0.2700323760509491, "learning_rate": 1.5617469351920373e-05, "loss": 0.0233, "step": 75430 }, { "grad_norm": 0.21360470354557037, "learning_rate": 1.5605466350816315e-05, "loss": 0.0228, "step": 75440 }, { "grad_norm": 0.6164131164550781, "learning_rate": 1.559346711104674e-05, "loss": 0.022, "step": 75450 }, { "grad_norm": 0.2807055115699768, "learning_rate": 1.5581471633923883e-05, "loss": 0.0177, "step": 75460 }, { "grad_norm": 0.3886585235595703, "learning_rate": 1.5569479920759505e-05, "loss": 0.0195, "step": 75470 }, { "grad_norm": 0.3214776813983917, "learning_rate": 1.5557491972865072e-05, "loss": 0.0238, "step": 75480 }, { "grad_norm": 0.5783999562263489, "learning_rate": 1.554550779155152e-05, "loss": 0.0246, "step": 75490 }, { "grad_norm": 0.25729286670684814, "learning_rate": 1.5533527378129453e-05, "loss": 0.0193, "step": 75500 }, { "grad_norm": 0.5336719155311584, "learning_rate": 1.552155073390899e-05, "loss": 0.0262, "step": 75510 }, { "grad_norm": 0.27451762557029724, "learning_rate": 1.5509577860199903e-05, "loss": 0.0187, "step": 75520 }, { "grad_norm": 0.29633963108062744, "learning_rate": 1.5497608758311525e-05, "loss": 0.0219, "step": 75530 }, { "grad_norm": 0.34284669160842896, "learning_rate": 1.5485643429552797e-05, "loss": 0.0217, "step": 75540 }, { "grad_norm": 0.2481847107410431, "learning_rate": 1.5473681875232197e-05, "loss": 0.0234, "step": 75550 }, { "grad_norm": 0.41942232847213745, "learning_rate": 1.5461724096657837e-05, "loss": 0.0287, "step": 75560 }, { "grad_norm": 0.44516322016716003, "learning_rate": 1.5449770095137416e-05, "loss": 0.0235, "step": 75570 }, { "grad_norm": 0.33836016058921814, "learning_rate": 1.5437819871978178e-05, "loss": 0.0203, "step": 75580 }, { "grad_norm": 0.5194244980812073, "learning_rate": 1.5425873428486992e-05, "loss": 0.0228, "step": 75590 }, { "grad_norm": 0.5474206209182739, "learning_rate": 1.5413930765970312e-05, "loss": 0.0252, "step": 75600 }, { "grad_norm": 0.24843139946460724, "learning_rate": 1.5401991885734175e-05, "loss": 0.0222, "step": 75610 }, { "grad_norm": 0.3058306574821472, "learning_rate": 1.5390056789084174e-05, "loss": 0.025, "step": 75620 }, { "grad_norm": 0.33368414640426636, "learning_rate": 1.5378125477325545e-05, "loss": 0.0241, "step": 75630 }, { "grad_norm": 0.22810174524784088, "learning_rate": 1.5366197951763046e-05, "loss": 0.0262, "step": 75640 }, { "grad_norm": 0.3598068356513977, "learning_rate": 1.5354274213701075e-05, "loss": 0.0219, "step": 75650 }, { "grad_norm": 0.8512141108512878, "learning_rate": 1.5342354264443566e-05, "loss": 0.0242, "step": 75660 }, { "grad_norm": 0.38422009348869324, "learning_rate": 1.533043810529411e-05, "loss": 0.0209, "step": 75670 }, { "grad_norm": 0.33279362320899963, "learning_rate": 1.5318525737555807e-05, "loss": 0.0258, "step": 75680 }, { "grad_norm": 0.27458861470222473, "learning_rate": 1.53066171625314e-05, "loss": 0.0288, "step": 75690 }, { "grad_norm": 0.24463891983032227, "learning_rate": 1.529471238152317e-05, "loss": 0.0203, "step": 75700 }, { "grad_norm": 0.4091504216194153, "learning_rate": 1.5282811395833023e-05, "loss": 0.0218, "step": 75710 }, { "grad_norm": 0.35202521085739136, "learning_rate": 1.5270914206762397e-05, "loss": 0.0197, "step": 75720 }, { "grad_norm": 0.5296240448951721, "learning_rate": 1.525902081561241e-05, "loss": 0.0261, "step": 75730 }, { "grad_norm": 0.34970980882644653, "learning_rate": 1.5247131223683664e-05, "loss": 0.023, "step": 75740 }, { "grad_norm": 0.2639266550540924, "learning_rate": 1.5235245432276408e-05, "loss": 0.0257, "step": 75750 }, { "grad_norm": 0.21103732287883759, "learning_rate": 1.5223363442690424e-05, "loss": 0.0241, "step": 75760 }, { "grad_norm": 0.4882626533508301, "learning_rate": 1.5211485256225133e-05, "loss": 0.0199, "step": 75770 }, { "grad_norm": 0.25443777441978455, "learning_rate": 1.519961087417951e-05, "loss": 0.0222, "step": 75780 }, { "grad_norm": 0.33319634199142456, "learning_rate": 1.5187740297852137e-05, "loss": 0.0213, "step": 75790 }, { "grad_norm": 0.2732495963573456, "learning_rate": 1.517587352854113e-05, "loss": 0.0219, "step": 75800 }, { "grad_norm": 0.4441276788711548, "learning_rate": 1.516401056754423e-05, "loss": 0.0226, "step": 75810 }, { "grad_norm": 0.3910430669784546, "learning_rate": 1.5152151416158783e-05, "loss": 0.0212, "step": 75820 }, { "grad_norm": 0.17193345725536346, "learning_rate": 1.5140296075681642e-05, "loss": 0.0207, "step": 75830 }, { "grad_norm": 0.24959656596183777, "learning_rate": 1.5128444547409331e-05, "loss": 0.0232, "step": 75840 }, { "grad_norm": 0.3328976631164551, "learning_rate": 1.5116596832637864e-05, "loss": 0.024, "step": 75850 }, { "grad_norm": 0.39300069212913513, "learning_rate": 1.5104752932662952e-05, "loss": 0.0187, "step": 75860 }, { "grad_norm": 0.3890795111656189, "learning_rate": 1.5092912848779778e-05, "loss": 0.023, "step": 75870 }, { "grad_norm": 0.431124210357666, "learning_rate": 1.5081076582283194e-05, "loss": 0.02, "step": 75880 }, { "grad_norm": 0.22141431272029877, "learning_rate": 1.5069244134467553e-05, "loss": 0.0198, "step": 75890 }, { "grad_norm": 0.4481026828289032, "learning_rate": 1.5057415506626876e-05, "loss": 0.0231, "step": 75900 }, { "grad_norm": 0.28481215238571167, "learning_rate": 1.5045590700054668e-05, "loss": 0.0233, "step": 75910 }, { "grad_norm": 0.25689730048179626, "learning_rate": 1.5033769716044144e-05, "loss": 0.0246, "step": 75920 }, { "grad_norm": 0.4621858298778534, "learning_rate": 1.5021952555887975e-05, "loss": 0.0214, "step": 75930 }, { "grad_norm": 0.2719404101371765, "learning_rate": 1.50101392208785e-05, "loss": 0.0239, "step": 75940 }, { "grad_norm": 0.5274397730827332, "learning_rate": 1.4998329712307579e-05, "loss": 0.022, "step": 75950 }, { "grad_norm": 1.003617763519287, "learning_rate": 1.4986524031466708e-05, "loss": 0.0274, "step": 75960 }, { "grad_norm": 0.2887822091579437, "learning_rate": 1.4974722179646916e-05, "loss": 0.0217, "step": 75970 }, { "grad_norm": 0.43093550205230713, "learning_rate": 1.4962924158138847e-05, "loss": 0.0231, "step": 75980 }, { "grad_norm": 0.6022958159446716, "learning_rate": 1.4951129968232714e-05, "loss": 0.0257, "step": 75990 }, { "grad_norm": 0.5938349366188049, "learning_rate": 1.4939339611218329e-05, "loss": 0.0272, "step": 76000 }, { "grad_norm": 0.40601667761802673, "learning_rate": 1.4927553088385032e-05, "loss": 0.0302, "step": 76010 }, { "grad_norm": 0.6287176012992859, "learning_rate": 1.4915770401021801e-05, "loss": 0.0237, "step": 76020 }, { "grad_norm": 0.35429850220680237, "learning_rate": 1.4903991550417185e-05, "loss": 0.0193, "step": 76030 }, { "grad_norm": 0.32849398255348206, "learning_rate": 1.4892216537859277e-05, "loss": 0.024, "step": 76040 }, { "grad_norm": 0.4014377295970917, "learning_rate": 1.4880445364635775e-05, "loss": 0.0237, "step": 76050 }, { "grad_norm": 0.35982662439346313, "learning_rate": 1.4868678032033973e-05, "loss": 0.0172, "step": 76060 }, { "grad_norm": 0.2043747752904892, "learning_rate": 1.4856914541340739e-05, "loss": 0.0228, "step": 76070 }, { "grad_norm": 0.2776772379875183, "learning_rate": 1.4845154893842472e-05, "loss": 0.0198, "step": 76080 }, { "grad_norm": 0.1774396002292633, "learning_rate": 1.4833399090825234e-05, "loss": 0.0197, "step": 76090 }, { "grad_norm": 0.26935750246047974, "learning_rate": 1.4821647133574562e-05, "loss": 0.0248, "step": 76100 }, { "grad_norm": 0.2666120231151581, "learning_rate": 1.4809899023375706e-05, "loss": 0.0246, "step": 76110 }, { "grad_norm": 0.32654911279678345, "learning_rate": 1.4798154761513366e-05, "loss": 0.0209, "step": 76120 }, { "grad_norm": 0.35752394795417786, "learning_rate": 1.4786414349271916e-05, "loss": 0.0197, "step": 76130 }, { "grad_norm": 0.27320927381515503, "learning_rate": 1.4774677787935226e-05, "loss": 0.0222, "step": 76140 }, { "grad_norm": 0.27597740292549133, "learning_rate": 1.4762945078786838e-05, "loss": 0.0227, "step": 76150 }, { "grad_norm": 0.31095409393310547, "learning_rate": 1.4751216223109777e-05, "loss": 0.0228, "step": 76160 }, { "grad_norm": 0.93276047706604, "learning_rate": 1.4739491222186714e-05, "loss": 0.0312, "step": 76170 }, { "grad_norm": 0.32355666160583496, "learning_rate": 1.4727770077299879e-05, "loss": 0.0232, "step": 76180 }, { "grad_norm": 0.284315824508667, "learning_rate": 1.4716052789731094e-05, "loss": 0.025, "step": 76190 }, { "grad_norm": 0.2942422032356262, "learning_rate": 1.470433936076171e-05, "loss": 0.026, "step": 76200 }, { "grad_norm": 0.449855774641037, "learning_rate": 1.4692629791672719e-05, "loss": 0.0269, "step": 76210 }, { "grad_norm": 0.5248777270317078, "learning_rate": 1.468092408374464e-05, "loss": 0.0182, "step": 76220 }, { "grad_norm": 0.437453955411911, "learning_rate": 1.4669222238257602e-05, "loss": 0.0269, "step": 76230 }, { "grad_norm": 0.30475476384162903, "learning_rate": 1.4657524256491301e-05, "loss": 0.0271, "step": 76240 }, { "grad_norm": 0.47124677896499634, "learning_rate": 1.464583013972502e-05, "loss": 0.0238, "step": 76250 }, { "grad_norm": 0.3430596590042114, "learning_rate": 1.4634139889237591e-05, "loss": 0.0273, "step": 76260 }, { "grad_norm": 0.47061699628829956, "learning_rate": 1.4622453506307448e-05, "loss": 0.023, "step": 76270 }, { "grad_norm": 0.4239351153373718, "learning_rate": 1.461077099221262e-05, "loss": 0.0213, "step": 76280 }, { "grad_norm": 0.2074861079454422, "learning_rate": 1.4599092348230652e-05, "loss": 0.0188, "step": 76290 }, { "grad_norm": 0.2717408537864685, "learning_rate": 1.4587417575638724e-05, "loss": 0.0222, "step": 76300 }, { "grad_norm": 0.4652954936027527, "learning_rate": 1.4575746675713563e-05, "loss": 0.0266, "step": 76310 }, { "grad_norm": 0.36172154545783997, "learning_rate": 1.4564079649731504e-05, "loss": 0.0211, "step": 76320 }, { "grad_norm": 0.5888756513595581, "learning_rate": 1.4552416498968402e-05, "loss": 0.0225, "step": 76330 }, { "grad_norm": 0.21557816863059998, "learning_rate": 1.4540757224699747e-05, "loss": 0.0241, "step": 76340 }, { "grad_norm": 0.3183818757534027, "learning_rate": 1.452910182820056e-05, "loss": 0.0185, "step": 76350 }, { "grad_norm": 0.2384718358516693, "learning_rate": 1.4517450310745468e-05, "loss": 0.0196, "step": 76360 }, { "grad_norm": 0.6030998229980469, "learning_rate": 1.450580267360866e-05, "loss": 0.0276, "step": 76370 }, { "grad_norm": 0.2885321378707886, "learning_rate": 1.4494158918063922e-05, "loss": 0.0184, "step": 76380 }, { "grad_norm": 0.42477601766586304, "learning_rate": 1.4482519045384568e-05, "loss": 0.0224, "step": 76390 }, { "grad_norm": 0.4227369725704193, "learning_rate": 1.4470883056843548e-05, "loss": 0.021, "step": 76400 }, { "grad_norm": 0.6217485070228577, "learning_rate": 1.4459250953713322e-05, "loss": 0.0243, "step": 76410 }, { "grad_norm": 0.39545145630836487, "learning_rate": 1.444762273726598e-05, "loss": 0.0251, "step": 76420 }, { "grad_norm": 0.2391631305217743, "learning_rate": 1.443599840877316e-05, "loss": 0.0206, "step": 76430 }, { "grad_norm": 0.2040415108203888, "learning_rate": 1.4424377969506103e-05, "loss": 0.0215, "step": 76440 }, { "grad_norm": 0.4577184021472931, "learning_rate": 1.4412761420735565e-05, "loss": 0.0319, "step": 76450 }, { "grad_norm": 0.3582785129547119, "learning_rate": 1.4401148763731953e-05, "loss": 0.0234, "step": 76460 }, { "grad_norm": 0.36679768562316895, "learning_rate": 1.4389539999765173e-05, "loss": 0.0254, "step": 76470 }, { "grad_norm": 0.361182302236557, "learning_rate": 1.4377935130104758e-05, "loss": 0.0172, "step": 76480 }, { "grad_norm": 0.26133623719215393, "learning_rate": 1.4366334156019801e-05, "loss": 0.0237, "step": 76490 }, { "grad_norm": 0.35086700320243835, "learning_rate": 1.435473707877898e-05, "loss": 0.0227, "step": 76500 }, { "grad_norm": 0.38046956062316895, "learning_rate": 1.4343143899650508e-05, "loss": 0.0249, "step": 76510 }, { "grad_norm": 3.605677604675293, "learning_rate": 1.4331554619902204e-05, "loss": 0.0253, "step": 76520 }, { "grad_norm": 0.3392297923564911, "learning_rate": 1.4319969240801484e-05, "loss": 0.0274, "step": 76530 }, { "grad_norm": 0.3151942491531372, "learning_rate": 1.430838776361526e-05, "loss": 0.0277, "step": 76540 }, { "grad_norm": 0.25138038396835327, "learning_rate": 1.4296810189610093e-05, "loss": 0.021, "step": 76550 }, { "grad_norm": 0.3048262894153595, "learning_rate": 1.4285236520052087e-05, "loss": 0.0233, "step": 76560 }, { "grad_norm": 0.3113686442375183, "learning_rate": 1.4273666756206933e-05, "loss": 0.0204, "step": 76570 }, { "grad_norm": 0.30086737871170044, "learning_rate": 1.4262100899339848e-05, "loss": 0.0221, "step": 76580 }, { "grad_norm": 0.35978439450263977, "learning_rate": 1.4250538950715696e-05, "loss": 0.0212, "step": 76590 }, { "grad_norm": 0.36252689361572266, "learning_rate": 1.4238980911598842e-05, "loss": 0.0228, "step": 76600 }, { "grad_norm": 0.2939044237136841, "learning_rate": 1.4227426783253267e-05, "loss": 0.0222, "step": 76610 }, { "grad_norm": 0.39955100417137146, "learning_rate": 1.4215876566942521e-05, "loss": 0.0265, "step": 76620 }, { "grad_norm": 0.4228745698928833, "learning_rate": 1.4204330263929721e-05, "loss": 0.0232, "step": 76630 }, { "grad_norm": 0.35887613892555237, "learning_rate": 1.4192787875477537e-05, "loss": 0.02, "step": 76640 }, { "grad_norm": 0.21381425857543945, "learning_rate": 1.4181249402848246e-05, "loss": 0.0244, "step": 76650 }, { "grad_norm": 0.3765401542186737, "learning_rate": 1.4169714847303656e-05, "loss": 0.0266, "step": 76660 }, { "grad_norm": 0.32258838415145874, "learning_rate": 1.4158184210105197e-05, "loss": 0.0272, "step": 76670 }, { "grad_norm": 0.21965941786766052, "learning_rate": 1.4146657492513787e-05, "loss": 0.0214, "step": 76680 }, { "grad_norm": 0.43337714672088623, "learning_rate": 1.4135134695790047e-05, "loss": 0.0257, "step": 76690 }, { "grad_norm": 0.19903674721717834, "learning_rate": 1.4123615821194037e-05, "loss": 0.0217, "step": 76700 }, { "grad_norm": 0.5052392482757568, "learning_rate": 1.4112100869985473e-05, "loss": 0.0242, "step": 76710 }, { "grad_norm": 0.3684905171394348, "learning_rate": 1.4100589843423589e-05, "loss": 0.0232, "step": 76720 }, { "grad_norm": 0.3593766689300537, "learning_rate": 1.408908274276723e-05, "loss": 0.0243, "step": 76730 }, { "grad_norm": 0.339369535446167, "learning_rate": 1.4077579569274786e-05, "loss": 0.0307, "step": 76740 }, { "grad_norm": 0.34562891721725464, "learning_rate": 1.4066080324204245e-05, "loss": 0.0237, "step": 76750 }, { "grad_norm": 0.554166316986084, "learning_rate": 1.405458500881312e-05, "loss": 0.025, "step": 76760 }, { "grad_norm": 0.28164997696876526, "learning_rate": 1.4043093624358538e-05, "loss": 0.0213, "step": 76770 }, { "grad_norm": 0.33424681425094604, "learning_rate": 1.4031606172097189e-05, "loss": 0.0206, "step": 76780 }, { "grad_norm": 0.5804495215415955, "learning_rate": 1.40201226532853e-05, "loss": 0.0202, "step": 76790 }, { "grad_norm": 0.2776927053928375, "learning_rate": 1.4008643069178701e-05, "loss": 0.0237, "step": 76800 }, { "grad_norm": 0.45237523317337036, "learning_rate": 1.399716742103278e-05, "loss": 0.021, "step": 76810 }, { "grad_norm": 0.8581388592720032, "learning_rate": 1.398569571010252e-05, "loss": 0.0275, "step": 76820 }, { "grad_norm": 0.4745027720928192, "learning_rate": 1.3974227937642414e-05, "loss": 0.0249, "step": 76830 }, { "grad_norm": 0.25346460938453674, "learning_rate": 1.3962764104906595e-05, "loss": 0.0177, "step": 76840 }, { "grad_norm": 0.27112722396850586, "learning_rate": 1.3951304213148696e-05, "loss": 0.0186, "step": 76850 }, { "grad_norm": 0.45311686396598816, "learning_rate": 1.3939848263621979e-05, "loss": 0.0218, "step": 76860 }, { "grad_norm": 0.31969892978668213, "learning_rate": 1.3928396257579219e-05, "loss": 0.0197, "step": 76870 }, { "grad_norm": 0.4761044681072235, "learning_rate": 1.3916948196272833e-05, "loss": 0.0209, "step": 76880 }, { "grad_norm": 0.3424672484397888, "learning_rate": 1.3905504080954734e-05, "loss": 0.0197, "step": 76890 }, { "grad_norm": 0.35624879598617554, "learning_rate": 1.3894063912876454e-05, "loss": 0.0248, "step": 76900 }, { "grad_norm": 0.39484965801239014, "learning_rate": 1.3882627693289046e-05, "loss": 0.0254, "step": 76910 }, { "grad_norm": 0.31741371750831604, "learning_rate": 1.3871195423443183e-05, "loss": 0.0237, "step": 76920 }, { "grad_norm": 0.4496806561946869, "learning_rate": 1.3859767104589044e-05, "loss": 0.0214, "step": 76930 }, { "grad_norm": 0.44843563437461853, "learning_rate": 1.3848342737976467e-05, "loss": 0.026, "step": 76940 }, { "grad_norm": 0.6356617212295532, "learning_rate": 1.3836922324854757e-05, "loss": 0.0252, "step": 76950 }, { "grad_norm": 0.22909684479236603, "learning_rate": 1.382550586647287e-05, "loss": 0.0229, "step": 76960 }, { "grad_norm": 0.5823078751564026, "learning_rate": 1.3814093364079257e-05, "loss": 0.0254, "step": 76970 }, { "grad_norm": 0.5092142820358276, "learning_rate": 1.3802684818921991e-05, "loss": 0.0268, "step": 76980 }, { "grad_norm": 0.4383206367492676, "learning_rate": 1.3791280232248687e-05, "loss": 0.0281, "step": 76990 }, { "grad_norm": 0.19610212743282318, "learning_rate": 1.3779879605306555e-05, "loss": 0.0193, "step": 77000 }, { "grad_norm": 0.41753634810447693, "learning_rate": 1.376848293934232e-05, "loss": 0.0233, "step": 77010 }, { "grad_norm": 0.8360726833343506, "learning_rate": 1.3757090235602316e-05, "loss": 0.017, "step": 77020 }, { "grad_norm": 0.30232444405555725, "learning_rate": 1.3745701495332447e-05, "loss": 0.0235, "step": 77030 }, { "grad_norm": 0.3374538719654083, "learning_rate": 1.3734316719778145e-05, "loss": 0.0218, "step": 77040 }, { "grad_norm": 0.29477980732917786, "learning_rate": 1.3722935910184454e-05, "loss": 0.0226, "step": 77050 }, { "grad_norm": 0.37632250785827637, "learning_rate": 1.3711559067795921e-05, "loss": 0.0228, "step": 77060 }, { "grad_norm": 0.4589219391345978, "learning_rate": 1.3700186193856756e-05, "loss": 0.0287, "step": 77070 }, { "grad_norm": 0.3811824321746826, "learning_rate": 1.368881728961064e-05, "loss": 0.0225, "step": 77080 }, { "grad_norm": 0.40009990334510803, "learning_rate": 1.3677452356300885e-05, "loss": 0.0276, "step": 77090 }, { "grad_norm": 0.45738154649734497, "learning_rate": 1.3666091395170321e-05, "loss": 0.0208, "step": 77100 }, { "grad_norm": 0.34415966272354126, "learning_rate": 1.3654734407461384e-05, "loss": 0.0244, "step": 77110 }, { "grad_norm": 0.4371298551559448, "learning_rate": 1.3643381394416022e-05, "loss": 0.0233, "step": 77120 }, { "grad_norm": 0.4068349599838257, "learning_rate": 1.3632032357275838e-05, "loss": 0.0215, "step": 77130 }, { "grad_norm": 0.1303381621837616, "learning_rate": 1.3620687297281898e-05, "loss": 0.0223, "step": 77140 }, { "grad_norm": 0.245827779173851, "learning_rate": 1.3609346215674917e-05, "loss": 0.0297, "step": 77150 }, { "grad_norm": 0.29430729150772095, "learning_rate": 1.3598009113695109e-05, "loss": 0.0226, "step": 77160 }, { "grad_norm": 0.5039833784103394, "learning_rate": 1.358667599258231e-05, "loss": 0.0188, "step": 77170 }, { "grad_norm": 0.2537381649017334, "learning_rate": 1.357534685357586e-05, "loss": 0.0234, "step": 77180 }, { "grad_norm": 0.5142910480499268, "learning_rate": 1.3564021697914715e-05, "loss": 0.0224, "step": 77190 }, { "grad_norm": 1.3835192918777466, "learning_rate": 1.3552700526837381e-05, "loss": 0.0256, "step": 77200 }, { "grad_norm": 0.2862619161605835, "learning_rate": 1.3541383341581925e-05, "loss": 0.0258, "step": 77210 }, { "grad_norm": 0.45792385935783386, "learning_rate": 1.3530070143385965e-05, "loss": 0.0239, "step": 77220 }, { "grad_norm": 0.3902878165245056, "learning_rate": 1.3518760933486702e-05, "loss": 0.0235, "step": 77230 }, { "grad_norm": 0.3129573464393616, "learning_rate": 1.3507455713120908e-05, "loss": 0.0228, "step": 77240 }, { "grad_norm": 0.29878684878349304, "learning_rate": 1.3496154483524876e-05, "loss": 0.0241, "step": 77250 }, { "grad_norm": 0.2598719298839569, "learning_rate": 1.3484857245934506e-05, "loss": 0.0216, "step": 77260 }, { "grad_norm": 0.34168434143066406, "learning_rate": 1.3473564001585248e-05, "loss": 0.0247, "step": 77270 }, { "grad_norm": 0.30643394589424133, "learning_rate": 1.3462274751712129e-05, "loss": 0.022, "step": 77280 }, { "grad_norm": 0.7816944122314453, "learning_rate": 1.3450989497549698e-05, "loss": 0.0251, "step": 77290 }, { "grad_norm": 0.5630643963813782, "learning_rate": 1.3439708240332115e-05, "loss": 0.0246, "step": 77300 }, { "grad_norm": 0.3250194191932678, "learning_rate": 1.3428430981293043e-05, "loss": 0.0261, "step": 77310 }, { "grad_norm": 0.42769500613212585, "learning_rate": 1.3417157721665808e-05, "loss": 0.0223, "step": 77320 }, { "grad_norm": 0.3127586245536804, "learning_rate": 1.3405888462683186e-05, "loss": 0.0227, "step": 77330 }, { "grad_norm": 0.3362972140312195, "learning_rate": 1.3394623205577595e-05, "loss": 0.0249, "step": 77340 }, { "grad_norm": 0.3840181529521942, "learning_rate": 1.3383361951580963e-05, "loss": 0.0319, "step": 77350 }, { "grad_norm": 0.3935118317604065, "learning_rate": 1.337210470192483e-05, "loss": 0.02, "step": 77360 }, { "grad_norm": 0.2790268063545227, "learning_rate": 1.3360851457840245e-05, "loss": 0.022, "step": 77370 }, { "grad_norm": 0.3713001608848572, "learning_rate": 1.3349602220557856e-05, "loss": 0.0239, "step": 77380 }, { "grad_norm": 0.36859169602394104, "learning_rate": 1.3338356991307865e-05, "loss": 0.0203, "step": 77390 }, { "grad_norm": 0.49132582545280457, "learning_rate": 1.3327115771320048e-05, "loss": 0.0221, "step": 77400 }, { "grad_norm": 0.2719157338142395, "learning_rate": 1.3315878561823697e-05, "loss": 0.0247, "step": 77410 }, { "grad_norm": 0.5998214483261108, "learning_rate": 1.330464536404773e-05, "loss": 0.0214, "step": 77420 }, { "grad_norm": 0.4111245274543762, "learning_rate": 1.3293416179220563e-05, "loss": 0.0225, "step": 77430 }, { "grad_norm": 0.4127832353115082, "learning_rate": 1.328219100857021e-05, "loss": 0.0258, "step": 77440 }, { "grad_norm": 0.40510424971580505, "learning_rate": 1.327096985332425e-05, "loss": 0.0214, "step": 77450 }, { "grad_norm": 0.2795480787754059, "learning_rate": 1.3259752714709816e-05, "loss": 0.0229, "step": 77460 }, { "grad_norm": 0.22342237830162048, "learning_rate": 1.3248539593953569e-05, "loss": 0.0263, "step": 77470 }, { "grad_norm": 0.3507333993911743, "learning_rate": 1.323733049228178e-05, "loss": 0.0251, "step": 77480 }, { "grad_norm": 0.36512067914009094, "learning_rate": 1.3226125410920276e-05, "loss": 0.0234, "step": 77490 }, { "grad_norm": 0.25844481587409973, "learning_rate": 1.3214924351094388e-05, "loss": 0.0235, "step": 77500 }, { "grad_norm": 0.49951493740081787, "learning_rate": 1.3203727314029069e-05, "loss": 0.0246, "step": 77510 }, { "grad_norm": 0.3472023606300354, "learning_rate": 1.3192534300948811e-05, "loss": 0.0213, "step": 77520 }, { "grad_norm": 0.2180175483226776, "learning_rate": 1.3181345313077675e-05, "loss": 0.0177, "step": 77530 }, { "grad_norm": 0.45537546277046204, "learning_rate": 1.3170160351639243e-05, "loss": 0.0253, "step": 77540 }, { "grad_norm": 0.47954052686691284, "learning_rate": 1.3158979417856721e-05, "loss": 0.0197, "step": 77550 }, { "grad_norm": 0.24461255967617035, "learning_rate": 1.31478025129528e-05, "loss": 0.0182, "step": 77560 }, { "grad_norm": 0.29503095149993896, "learning_rate": 1.3136629638149794e-05, "loss": 0.0195, "step": 77570 }, { "grad_norm": 0.30742689967155457, "learning_rate": 1.3125460794669547e-05, "loss": 0.0224, "step": 77580 }, { "grad_norm": 0.318699449300766, "learning_rate": 1.3114295983733483e-05, "loss": 0.0201, "step": 77590 }, { "grad_norm": 0.3362986147403717, "learning_rate": 1.3103135206562534e-05, "loss": 0.0216, "step": 77600 }, { "grad_norm": 0.40714430809020996, "learning_rate": 1.3091978464377264e-05, "loss": 0.0228, "step": 77610 }, { "grad_norm": 0.45773622393608093, "learning_rate": 1.3080825758397724e-05, "loss": 0.0214, "step": 77620 }, { "grad_norm": 0.22417935729026794, "learning_rate": 1.3069677089843568e-05, "loss": 0.0216, "step": 77630 }, { "grad_norm": 0.3278961777687073, "learning_rate": 1.305853245993401e-05, "loss": 0.0245, "step": 77640 }, { "grad_norm": 0.44196948409080505, "learning_rate": 1.3047391869887809e-05, "loss": 0.0244, "step": 77650 }, { "grad_norm": 0.3764798939228058, "learning_rate": 1.3036255320923263e-05, "loss": 0.0237, "step": 77660 }, { "grad_norm": 1.1204074621200562, "learning_rate": 1.3025122814258273e-05, "loss": 0.0267, "step": 77670 }, { "grad_norm": 0.2875255346298218, "learning_rate": 1.3013994351110249e-05, "loss": 0.0172, "step": 77680 }, { "grad_norm": 0.35812339186668396, "learning_rate": 1.3002869932696194e-05, "loss": 0.0195, "step": 77690 }, { "grad_norm": 0.5339595675468445, "learning_rate": 1.2991749560232664e-05, "loss": 0.0226, "step": 77700 }, { "grad_norm": 0.17743642628192902, "learning_rate": 1.2980633234935768e-05, "loss": 0.02, "step": 77710 }, { "grad_norm": 0.49342915415763855, "learning_rate": 1.2969520958021153e-05, "loss": 0.0301, "step": 77720 }, { "grad_norm": 0.5644271373748779, "learning_rate": 1.295841273070405e-05, "loss": 0.0241, "step": 77730 }, { "grad_norm": 0.2964116036891937, "learning_rate": 1.2947308554199255e-05, "loss": 0.0235, "step": 77740 }, { "grad_norm": 0.6491528153419495, "learning_rate": 1.2936208429721075e-05, "loss": 0.0171, "step": 77750 }, { "grad_norm": 0.5867273211479187, "learning_rate": 1.2925112358483416e-05, "loss": 0.0218, "step": 77760 }, { "grad_norm": 0.2997020483016968, "learning_rate": 1.2914020341699728e-05, "loss": 0.0216, "step": 77770 }, { "grad_norm": 0.3009795844554901, "learning_rate": 1.2902932380583032e-05, "loss": 0.0211, "step": 77780 }, { "grad_norm": 0.16954608261585236, "learning_rate": 1.2891848476345863e-05, "loss": 0.0255, "step": 77790 }, { "grad_norm": 0.3397674560546875, "learning_rate": 1.2880768630200368e-05, "loss": 0.0249, "step": 77800 }, { "grad_norm": 0.36670583486557007, "learning_rate": 1.286969284335819e-05, "loss": 0.0264, "step": 77810 }, { "grad_norm": 0.4750382900238037, "learning_rate": 1.2858621117030578e-05, "loss": 0.0212, "step": 77820 }, { "grad_norm": 0.2626670300960541, "learning_rate": 1.2847553452428324e-05, "loss": 0.0236, "step": 77830 }, { "grad_norm": 0.6021392345428467, "learning_rate": 1.283648985076178e-05, "loss": 0.0257, "step": 77840 }, { "grad_norm": 0.32844749093055725, "learning_rate": 1.2825430313240817e-05, "loss": 0.022, "step": 77850 }, { "grad_norm": 0.22058285772800446, "learning_rate": 1.281437484107491e-05, "loss": 0.0223, "step": 77860 }, { "grad_norm": 0.25368720293045044, "learning_rate": 1.2803323435473058e-05, "loss": 0.0212, "step": 77870 }, { "grad_norm": 0.3292475938796997, "learning_rate": 1.2792276097643841e-05, "loss": 0.0199, "step": 77880 }, { "grad_norm": 0.30236467719078064, "learning_rate": 1.2781232828795342e-05, "loss": 0.0221, "step": 77890 }, { "grad_norm": 0.39475908875465393, "learning_rate": 1.2770193630135292e-05, "loss": 0.0247, "step": 77900 }, { "grad_norm": 0.21631695330142975, "learning_rate": 1.2759158502870877e-05, "loss": 0.0216, "step": 77910 }, { "grad_norm": 0.3173667788505554, "learning_rate": 1.2748127448208913e-05, "loss": 0.018, "step": 77920 }, { "grad_norm": 0.33531737327575684, "learning_rate": 1.2737100467355706e-05, "loss": 0.0205, "step": 77930 }, { "grad_norm": 0.46564432978630066, "learning_rate": 1.2726077561517169e-05, "loss": 0.0223, "step": 77940 }, { "grad_norm": 0.30840155482292175, "learning_rate": 1.2715058731898755e-05, "loss": 0.0257, "step": 77950 }, { "grad_norm": 0.2594335675239563, "learning_rate": 1.2704043979705471e-05, "loss": 0.0192, "step": 77960 }, { "grad_norm": 0.27092090249061584, "learning_rate": 1.269303330614185e-05, "loss": 0.0266, "step": 77970 }, { "grad_norm": 0.42810043692588806, "learning_rate": 1.2682026712412015e-05, "loss": 0.02, "step": 77980 }, { "grad_norm": 0.22700823843479156, "learning_rate": 1.2671024199719644e-05, "loss": 0.0199, "step": 77990 }, { "grad_norm": 0.40663373470306396, "learning_rate": 1.2660025769267936e-05, "loss": 0.0229, "step": 78000 }, { "grad_norm": 0.33685925602912903, "learning_rate": 1.2649031422259667e-05, "loss": 0.0236, "step": 78010 }, { "grad_norm": 0.4252106547355652, "learning_rate": 1.2638041159897163e-05, "loss": 0.0233, "step": 78020 }, { "grad_norm": 0.28496336936950684, "learning_rate": 1.262705498338232e-05, "loss": 0.026, "step": 78030 }, { "grad_norm": 0.2964639961719513, "learning_rate": 1.2616072893916542e-05, "loss": 0.0226, "step": 78040 }, { "grad_norm": 0.2877173125743866, "learning_rate": 1.2605094892700836e-05, "loss": 0.0206, "step": 78050 }, { "grad_norm": 0.3304012715816498, "learning_rate": 1.259412098093572e-05, "loss": 0.0229, "step": 78060 }, { "grad_norm": 0.3663547933101654, "learning_rate": 1.258315115982131e-05, "loss": 0.0241, "step": 78070 }, { "grad_norm": 0.42868006229400635, "learning_rate": 1.2572185430557203e-05, "loss": 0.0246, "step": 78080 }, { "grad_norm": 0.34727543592453003, "learning_rate": 1.2561223794342659e-05, "loss": 0.0263, "step": 78090 }, { "grad_norm": 0.37667927145957947, "learning_rate": 1.255026625237637e-05, "loss": 0.0194, "step": 78100 }, { "grad_norm": 1.1603771448135376, "learning_rate": 1.2539312805856678e-05, "loss": 0.0265, "step": 78110 }, { "grad_norm": 0.40649479627609253, "learning_rate": 1.25283634559814e-05, "loss": 0.0207, "step": 78120 }, { "grad_norm": 0.4947294294834137, "learning_rate": 1.2517418203947972e-05, "loss": 0.0236, "step": 78130 }, { "grad_norm": 0.6944331526756287, "learning_rate": 1.2506477050953308e-05, "loss": 0.0281, "step": 78140 }, { "grad_norm": 0.3310518264770508, "learning_rate": 1.2495539998193972e-05, "loss": 0.0192, "step": 78150 }, { "grad_norm": 0.4174948036670685, "learning_rate": 1.2484607046865975e-05, "loss": 0.0237, "step": 78160 }, { "grad_norm": 0.5178897380828857, "learning_rate": 1.2473678198164968e-05, "loss": 0.0266, "step": 78170 }, { "grad_norm": 0.3563709557056427, "learning_rate": 1.2462753453286075e-05, "loss": 0.0219, "step": 78180 }, { "grad_norm": 0.3396657109260559, "learning_rate": 1.2451832813424031e-05, "loss": 0.0259, "step": 78190 }, { "grad_norm": 0.24448837339878082, "learning_rate": 1.2440916279773095e-05, "loss": 0.0215, "step": 78200 }, { "grad_norm": 0.22613194584846497, "learning_rate": 1.2430003853527101e-05, "loss": 0.0214, "step": 78210 }, { "grad_norm": 0.4263124465942383, "learning_rate": 1.2419095535879383e-05, "loss": 0.0217, "step": 78220 }, { "grad_norm": 0.5758129954338074, "learning_rate": 1.2408191328022878e-05, "loss": 0.0261, "step": 78230 }, { "grad_norm": 0.3317745327949524, "learning_rate": 1.2397291231150066e-05, "loss": 0.0225, "step": 78240 }, { "grad_norm": 0.3342897295951843, "learning_rate": 1.2386395246452937e-05, "loss": 0.0277, "step": 78250 }, { "grad_norm": 0.300292044878006, "learning_rate": 1.2375503375123083e-05, "loss": 0.0192, "step": 78260 }, { "grad_norm": 0.32816800475120544, "learning_rate": 1.2364615618351583e-05, "loss": 0.0254, "step": 78270 }, { "grad_norm": 0.4736364185810089, "learning_rate": 1.2353731977329169e-05, "loss": 0.0217, "step": 78280 }, { "grad_norm": 0.3531486988067627, "learning_rate": 1.2342852453246007e-05, "loss": 0.0251, "step": 78290 }, { "grad_norm": 0.47965380549430847, "learning_rate": 1.2331977047291898e-05, "loss": 0.0241, "step": 78300 }, { "grad_norm": 0.5500918626785278, "learning_rate": 1.2321105760656137e-05, "loss": 0.0227, "step": 78310 }, { "grad_norm": 0.2483685165643692, "learning_rate": 1.2310238594527613e-05, "loss": 0.0217, "step": 78320 }, { "grad_norm": 0.3478772044181824, "learning_rate": 1.2299375550094693e-05, "loss": 0.0211, "step": 78330 }, { "grad_norm": 0.3131616711616516, "learning_rate": 1.2288516628545421e-05, "loss": 0.017, "step": 78340 }, { "grad_norm": 0.4283474385738373, "learning_rate": 1.2277661831067255e-05, "loss": 0.0209, "step": 78350 }, { "grad_norm": 0.5034064054489136, "learning_rate": 1.2266811158847286e-05, "loss": 0.0208, "step": 78360 }, { "grad_norm": 0.2580065429210663, "learning_rate": 1.2255964613072107e-05, "loss": 0.027, "step": 78370 }, { "grad_norm": 0.5506866574287415, "learning_rate": 1.2245122194927905e-05, "loss": 0.0192, "step": 78380 }, { "grad_norm": 0.27625831961631775, "learning_rate": 1.2234283905600364e-05, "loss": 0.0234, "step": 78390 }, { "grad_norm": 0.702903687953949, "learning_rate": 1.222344974627475e-05, "loss": 0.0222, "step": 78400 }, { "grad_norm": 0.41511213779449463, "learning_rate": 1.2212619718135875e-05, "loss": 0.0207, "step": 78410 }, { "grad_norm": 0.31140634417533875, "learning_rate": 1.2201793822368118e-05, "loss": 0.0213, "step": 78420 }, { "grad_norm": 0.5795956254005432, "learning_rate": 1.2190972060155337e-05, "loss": 0.0222, "step": 78430 }, { "grad_norm": 0.500886857509613, "learning_rate": 1.2180154432681013e-05, "loss": 0.0226, "step": 78440 }, { "grad_norm": 0.44152021408081055, "learning_rate": 1.2169340941128148e-05, "loss": 0.0182, "step": 78450 }, { "grad_norm": 0.7319707274436951, "learning_rate": 1.2158531586679267e-05, "loss": 0.0239, "step": 78460 }, { "grad_norm": 0.2088027447462082, "learning_rate": 1.2147726370516477e-05, "loss": 0.0278, "step": 78470 }, { "grad_norm": 0.31417715549468994, "learning_rate": 1.2136925293821422e-05, "loss": 0.0293, "step": 78480 }, { "grad_norm": 0.3400034010410309, "learning_rate": 1.2126128357775306e-05, "loss": 0.0223, "step": 78490 }, { "grad_norm": 0.2599449157714844, "learning_rate": 1.2115335563558838e-05, "loss": 0.0196, "step": 78500 }, { "grad_norm": 0.29648250341415405, "learning_rate": 1.2104546912352327e-05, "loss": 0.0183, "step": 78510 }, { "grad_norm": 0.3349163234233856, "learning_rate": 1.2093762405335557e-05, "loss": 0.0201, "step": 78520 }, { "grad_norm": 0.3102231025695801, "learning_rate": 1.2082982043687974e-05, "loss": 0.0184, "step": 78530 }, { "grad_norm": 0.4646563231945038, "learning_rate": 1.207220582858845e-05, "loss": 0.0211, "step": 78540 }, { "grad_norm": 0.4776574373245239, "learning_rate": 1.206143376121549e-05, "loss": 0.0218, "step": 78550 }, { "grad_norm": 0.2558300495147705, "learning_rate": 1.2050665842747078e-05, "loss": 0.0246, "step": 78560 }, { "grad_norm": 0.4207373559474945, "learning_rate": 1.2039902074360804e-05, "loss": 0.0212, "step": 78570 }, { "grad_norm": 0.48125502467155457, "learning_rate": 1.2029142457233755e-05, "loss": 0.0251, "step": 78580 }, { "grad_norm": 0.5225220322608948, "learning_rate": 1.2018386992542601e-05, "loss": 0.0237, "step": 78590 }, { "grad_norm": 0.48194369673728943, "learning_rate": 1.2007635681463542e-05, "loss": 0.0198, "step": 78600 }, { "grad_norm": 0.39359503984451294, "learning_rate": 1.1996888525172334e-05, "loss": 0.0215, "step": 78610 }, { "grad_norm": 0.31293532252311707, "learning_rate": 1.198614552484425e-05, "loss": 0.0243, "step": 78620 }, { "grad_norm": 0.26941901445388794, "learning_rate": 1.197540668165415e-05, "loss": 0.0217, "step": 78630 }, { "grad_norm": 0.32478272914886475, "learning_rate": 1.1964671996776395e-05, "loss": 0.0223, "step": 78640 }, { "grad_norm": 0.20690985023975372, "learning_rate": 1.1953941471384922e-05, "loss": 0.0248, "step": 78650 }, { "grad_norm": 0.17897337675094604, "learning_rate": 1.1943215106653205e-05, "loss": 0.0164, "step": 78660 }, { "grad_norm": 0.30846723914146423, "learning_rate": 1.193249290375429e-05, "loss": 0.0209, "step": 78670 }, { "grad_norm": 0.30721789598464966, "learning_rate": 1.1921774863860696e-05, "loss": 0.0219, "step": 78680 }, { "grad_norm": 0.34851476550102234, "learning_rate": 1.1911060988144556e-05, "loss": 0.0203, "step": 78690 }, { "grad_norm": 0.3156205415725708, "learning_rate": 1.1900351277777533e-05, "loss": 0.021, "step": 78700 }, { "grad_norm": 0.32395172119140625, "learning_rate": 1.1889645733930798e-05, "loss": 0.023, "step": 78710 }, { "grad_norm": 0.24779590964317322, "learning_rate": 1.1878944357775107e-05, "loss": 0.0202, "step": 78720 }, { "grad_norm": 0.2823375463485718, "learning_rate": 1.1868247150480749e-05, "loss": 0.0253, "step": 78730 }, { "grad_norm": 0.2354682832956314, "learning_rate": 1.1857554113217567e-05, "loss": 0.0193, "step": 78740 }, { "grad_norm": 0.4671825170516968, "learning_rate": 1.1846865247154903e-05, "loss": 0.0234, "step": 78750 }, { "grad_norm": 0.27688395977020264, "learning_rate": 1.1836180553461711e-05, "loss": 0.02, "step": 78760 }, { "grad_norm": 0.294763445854187, "learning_rate": 1.1825500033306419e-05, "loss": 0.0254, "step": 78770 }, { "grad_norm": 0.3466288447380066, "learning_rate": 1.1814823687857052e-05, "loss": 0.0232, "step": 78780 }, { "grad_norm": 0.6442304253578186, "learning_rate": 1.1804151518281153e-05, "loss": 0.0223, "step": 78790 }, { "grad_norm": 0.6157304048538208, "learning_rate": 1.179348352574583e-05, "loss": 0.0192, "step": 78800 }, { "grad_norm": 0.5933598875999451, "learning_rate": 1.1782819711417697e-05, "loss": 0.0266, "step": 78810 }, { "grad_norm": 0.510057806968689, "learning_rate": 1.1772160076462951e-05, "loss": 0.0276, "step": 78820 }, { "grad_norm": 0.25674355030059814, "learning_rate": 1.1761504622047293e-05, "loss": 0.0214, "step": 78830 }, { "grad_norm": 0.24115464091300964, "learning_rate": 1.1750853349335999e-05, "loss": 0.0239, "step": 78840 }, { "grad_norm": 0.4391096234321594, "learning_rate": 1.1740206259493874e-05, "loss": 0.0187, "step": 78850 }, { "grad_norm": 0.2849825322628021, "learning_rate": 1.1729563353685286e-05, "loss": 0.0238, "step": 78860 }, { "grad_norm": 0.2721230983734131, "learning_rate": 1.1718924633074091e-05, "loss": 0.0246, "step": 78870 }, { "grad_norm": 0.3211252987384796, "learning_rate": 1.1708290098823765e-05, "loss": 0.0189, "step": 78880 }, { "grad_norm": 0.9231786131858826, "learning_rate": 1.1697659752097245e-05, "loss": 0.0274, "step": 78890 }, { "grad_norm": 0.3818889260292053, "learning_rate": 1.1687033594057068e-05, "loss": 0.0205, "step": 78900 }, { "grad_norm": 0.3148019015789032, "learning_rate": 1.1676411625865297e-05, "loss": 0.0226, "step": 78910 }, { "grad_norm": 0.2918262183666229, "learning_rate": 1.1665793848683542e-05, "loss": 0.0205, "step": 78920 }, { "grad_norm": 0.25661805272102356, "learning_rate": 1.1655180263672927e-05, "loss": 0.0231, "step": 78930 }, { "grad_norm": 0.569284975528717, "learning_rate": 1.1644570871994142e-05, "loss": 0.022, "step": 78940 }, { "grad_norm": 0.2883044481277466, "learning_rate": 1.163396567480744e-05, "loss": 0.022, "step": 78950 }, { "grad_norm": 0.37141790986061096, "learning_rate": 1.1623364673272552e-05, "loss": 0.0243, "step": 78960 }, { "grad_norm": 0.3393898010253906, "learning_rate": 1.1612767868548807e-05, "loss": 0.0227, "step": 78970 }, { "grad_norm": 0.3101769685745239, "learning_rate": 1.1602175261795057e-05, "loss": 0.0265, "step": 78980 }, { "grad_norm": 0.5053300261497498, "learning_rate": 1.1591586854169696e-05, "loss": 0.0199, "step": 78990 }, { "grad_norm": 0.41102316975593567, "learning_rate": 1.158100264683064e-05, "loss": 0.0218, "step": 79000 }, { "grad_norm": 0.37414267659187317, "learning_rate": 1.1570422640935386e-05, "loss": 0.0282, "step": 79010 }, { "grad_norm": 0.4306983947753906, "learning_rate": 1.155984683764092e-05, "loss": 0.0264, "step": 79020 }, { "grad_norm": 0.5183374285697937, "learning_rate": 1.1549275238103812e-05, "loss": 0.0255, "step": 79030 }, { "grad_norm": 0.3893589675426483, "learning_rate": 1.1538707843480152e-05, "loss": 0.0245, "step": 79040 }, { "grad_norm": 0.31639039516448975, "learning_rate": 1.1528144654925593e-05, "loss": 0.0194, "step": 79050 }, { "grad_norm": 0.2754380404949188, "learning_rate": 1.1517585673595277e-05, "loss": 0.0175, "step": 79060 }, { "grad_norm": 0.2633601427078247, "learning_rate": 1.150703090064395e-05, "loss": 0.0243, "step": 79070 }, { "grad_norm": 0.6092178821563721, "learning_rate": 1.1496480337225835e-05, "loss": 0.0199, "step": 79080 }, { "grad_norm": 0.2927863895893097, "learning_rate": 1.1485933984494751e-05, "loss": 0.02, "step": 79090 }, { "grad_norm": 0.2213471680879593, "learning_rate": 1.1475391843604e-05, "loss": 0.019, "step": 79100 }, { "grad_norm": 0.8517100811004639, "learning_rate": 1.1464853915706503e-05, "loss": 0.0203, "step": 79110 }, { "grad_norm": 0.36136114597320557, "learning_rate": 1.1454320201954626e-05, "loss": 0.0234, "step": 79120 }, { "grad_norm": 0.23942026495933533, "learning_rate": 1.1443790703500356e-05, "loss": 0.018, "step": 79130 }, { "grad_norm": 0.5583993792533875, "learning_rate": 1.1433265421495154e-05, "loss": 0.0217, "step": 79140 }, { "grad_norm": 0.3299294412136078, "learning_rate": 1.1422744357090065e-05, "loss": 0.0197, "step": 79150 }, { "grad_norm": 1.3807854652404785, "learning_rate": 1.1412227511435647e-05, "loss": 0.0264, "step": 79160 }, { "grad_norm": 0.6911450624465942, "learning_rate": 1.1401714885682025e-05, "loss": 0.0179, "step": 79170 }, { "grad_norm": 0.367435485124588, "learning_rate": 1.1391206480978823e-05, "loss": 0.0218, "step": 79180 }, { "grad_norm": 0.39125609397888184, "learning_rate": 1.1380702298475227e-05, "loss": 0.0232, "step": 79190 }, { "grad_norm": 0.40190085768699646, "learning_rate": 1.1370202339319985e-05, "loss": 0.0165, "step": 79200 }, { "grad_norm": 0.26417097449302673, "learning_rate": 1.1359706604661319e-05, "loss": 0.023, "step": 79210 }, { "grad_norm": 0.19894637167453766, "learning_rate": 1.1349215095647042e-05, "loss": 0.0198, "step": 79220 }, { "grad_norm": 0.3640972971916199, "learning_rate": 1.1338727813424493e-05, "loss": 0.0213, "step": 79230 }, { "grad_norm": 0.9132712483406067, "learning_rate": 1.132824475914056e-05, "loss": 0.0209, "step": 79240 }, { "grad_norm": 0.40150004625320435, "learning_rate": 1.131776593394162e-05, "loss": 0.0251, "step": 79250 }, { "grad_norm": 0.3919628858566284, "learning_rate": 1.1307291338973657e-05, "loss": 0.0214, "step": 79260 }, { "grad_norm": 0.29887038469314575, "learning_rate": 1.1296820975382121e-05, "loss": 0.021, "step": 79270 }, { "grad_norm": 0.4211939573287964, "learning_rate": 1.1286354844312074e-05, "loss": 0.0213, "step": 79280 }, { "grad_norm": 0.23224471509456635, "learning_rate": 1.1275892946908023e-05, "loss": 0.0184, "step": 79290 }, { "grad_norm": 1.5064011812210083, "learning_rate": 1.126543528431413e-05, "loss": 0.0192, "step": 79300 }, { "grad_norm": 1.015058159828186, "learning_rate": 1.125498185767398e-05, "loss": 0.0234, "step": 79310 }, { "grad_norm": 0.2893197536468506, "learning_rate": 1.1244532668130781e-05, "loss": 0.0195, "step": 79320 }, { "grad_norm": 0.6156572699546814, "learning_rate": 1.1234087716827203e-05, "loss": 0.0248, "step": 79330 }, { "grad_norm": 0.3557305335998535, "learning_rate": 1.1223647004905524e-05, "loss": 0.0222, "step": 79340 }, { "grad_norm": 0.37064847350120544, "learning_rate": 1.1213210533507485e-05, "loss": 0.0306, "step": 79350 }, { "grad_norm": 0.2592249810695648, "learning_rate": 1.1202778303774447e-05, "loss": 0.0194, "step": 79360 }, { "grad_norm": 0.4751371443271637, "learning_rate": 1.1192350316847234e-05, "loss": 0.0246, "step": 79370 }, { "grad_norm": 0.2949487864971161, "learning_rate": 1.1181926573866258e-05, "loss": 0.0222, "step": 79380 }, { "grad_norm": 0.2620634436607361, "learning_rate": 1.1171507075971416e-05, "loss": 0.0218, "step": 79390 }, { "grad_norm": 0.4850708544254303, "learning_rate": 1.1161091824302177e-05, "loss": 0.0224, "step": 79400 }, { "grad_norm": 0.4970077574253082, "learning_rate": 1.115068081999755e-05, "loss": 0.0255, "step": 79410 }, { "grad_norm": 0.39355722069740295, "learning_rate": 1.1140274064196071e-05, "loss": 0.0211, "step": 79420 }, { "grad_norm": 0.40578436851501465, "learning_rate": 1.112987155803578e-05, "loss": 0.0209, "step": 79430 }, { "grad_norm": 0.29598474502563477, "learning_rate": 1.1119473302654298e-05, "loss": 0.0186, "step": 79440 }, { "grad_norm": 0.25373587012290955, "learning_rate": 1.1109079299188769e-05, "loss": 0.024, "step": 79450 }, { "grad_norm": 0.4037511348724365, "learning_rate": 1.1098689548775847e-05, "loss": 0.023, "step": 79460 }, { "grad_norm": 0.2831014096736908, "learning_rate": 1.1088304052551762e-05, "loss": 0.0263, "step": 79470 }, { "grad_norm": 0.26784148812294006, "learning_rate": 1.1077922811652215e-05, "loss": 0.0253, "step": 79480 }, { "grad_norm": 0.6305545568466187, "learning_rate": 1.1067545827212534e-05, "loss": 0.0227, "step": 79490 }, { "grad_norm": 0.6807557344436646, "learning_rate": 1.1057173100367496e-05, "loss": 0.023, "step": 79500 }, { "grad_norm": 0.582560658454895, "learning_rate": 1.1046804632251474e-05, "loss": 0.0172, "step": 79510 }, { "grad_norm": 0.3653571307659149, "learning_rate": 1.1036440423998318e-05, "loss": 0.0226, "step": 79520 }, { "grad_norm": 0.4530188739299774, "learning_rate": 1.1026080476741469e-05, "loss": 0.0241, "step": 79530 }, { "grad_norm": 0.33639055490493774, "learning_rate": 1.1015724791613836e-05, "loss": 0.0223, "step": 79540 }, { "grad_norm": 0.3183474838733673, "learning_rate": 1.1005373369747951e-05, "loss": 0.0203, "step": 79550 }, { "grad_norm": 0.29549121856689453, "learning_rate": 1.09950262122758e-05, "loss": 0.0214, "step": 79560 }, { "grad_norm": 0.4575745463371277, "learning_rate": 1.0984683320328948e-05, "loss": 0.0275, "step": 79570 }, { "grad_norm": 0.5668193101882935, "learning_rate": 1.0974344695038458e-05, "loss": 0.0246, "step": 79580 }, { "grad_norm": 0.3412233591079712, "learning_rate": 1.0964010337534968e-05, "loss": 0.0192, "step": 79590 }, { "grad_norm": 0.2596994638442993, "learning_rate": 1.0953680248948611e-05, "loss": 0.0236, "step": 79600 }, { "grad_norm": 0.39641863107681274, "learning_rate": 1.0943354430409069e-05, "loss": 0.0267, "step": 79610 }, { "grad_norm": 0.3468237519264221, "learning_rate": 1.0933032883045574e-05, "loss": 0.0191, "step": 79620 }, { "grad_norm": 0.21412834525108337, "learning_rate": 1.0922715607986872e-05, "loss": 0.0254, "step": 79630 }, { "grad_norm": 0.3472290635108948, "learning_rate": 1.0912402606361234e-05, "loss": 0.0218, "step": 79640 }, { "grad_norm": 0.5667555928230286, "learning_rate": 1.090209387929647e-05, "loss": 0.0178, "step": 79650 }, { "grad_norm": 0.4684138596057892, "learning_rate": 1.0891789427919957e-05, "loss": 0.0218, "step": 79660 }, { "grad_norm": 0.3114331364631653, "learning_rate": 1.088148925335854e-05, "loss": 0.0214, "step": 79670 }, { "grad_norm": 0.5210157036781311, "learning_rate": 1.087119335673864e-05, "loss": 0.021, "step": 79680 }, { "grad_norm": 0.45995450019836426, "learning_rate": 1.0860901739186208e-05, "loss": 0.0191, "step": 79690 }, { "grad_norm": 0.25264570116996765, "learning_rate": 1.085061440182673e-05, "loss": 0.0258, "step": 79700 }, { "grad_norm": 1.0171821117401123, "learning_rate": 1.0840331345785193e-05, "loss": 0.0243, "step": 79710 }, { "grad_norm": 0.33783817291259766, "learning_rate": 1.083005257218615e-05, "loss": 0.0232, "step": 79720 }, { "grad_norm": 1.1276228427886963, "learning_rate": 1.0819778082153642e-05, "loss": 0.0208, "step": 79730 }, { "grad_norm": 0.6393880248069763, "learning_rate": 1.0809507876811325e-05, "loss": 0.0233, "step": 79740 }, { "grad_norm": 0.2873404026031494, "learning_rate": 1.0799241957282292e-05, "loss": 0.0266, "step": 79750 }, { "grad_norm": 0.33295756578445435, "learning_rate": 1.078898032468923e-05, "loss": 0.0196, "step": 79760 }, { "grad_norm": 0.2579014301300049, "learning_rate": 1.0778722980154315e-05, "loss": 0.02, "step": 79770 }, { "grad_norm": 0.739835262298584, "learning_rate": 1.0768469924799302e-05, "loss": 0.0281, "step": 79780 }, { "grad_norm": 0.20549577474594116, "learning_rate": 1.0758221159745419e-05, "loss": 0.0211, "step": 79790 }, { "grad_norm": 0.362232506275177, "learning_rate": 1.074797668611347e-05, "loss": 0.0221, "step": 79800 }, { "grad_norm": 0.25920867919921875, "learning_rate": 1.0737736505023777e-05, "loss": 0.0217, "step": 79810 }, { "grad_norm": 0.21079717576503754, "learning_rate": 1.0727500617596203e-05, "loss": 0.021, "step": 79820 }, { "grad_norm": 0.3018738925457001, "learning_rate": 1.0717269024950104e-05, "loss": 0.0235, "step": 79830 }, { "grad_norm": 0.2588127553462982, "learning_rate": 1.070704172820442e-05, "loss": 0.0215, "step": 79840 }, { "grad_norm": 0.160988450050354, "learning_rate": 1.0696818728477559e-05, "loss": 0.0213, "step": 79850 }, { "grad_norm": 0.28277838230133057, "learning_rate": 1.0686600026887511e-05, "loss": 0.0239, "step": 79860 }, { "grad_norm": 0.2675643563270569, "learning_rate": 1.0676385624551777e-05, "loss": 0.0167, "step": 79870 }, { "grad_norm": 0.27366718649864197, "learning_rate": 1.0666175522587402e-05, "loss": 0.0215, "step": 79880 }, { "grad_norm": 0.7084622383117676, "learning_rate": 1.0655969722110925e-05, "loss": 0.019, "step": 79890 }, { "grad_norm": 0.4749646782875061, "learning_rate": 1.0645768224238439e-05, "loss": 0.0255, "step": 79900 }, { "grad_norm": 0.3788180351257324, "learning_rate": 1.063557103008559e-05, "loss": 0.028, "step": 79910 }, { "grad_norm": 0.2623283565044403, "learning_rate": 1.0625378140767494e-05, "loss": 0.0217, "step": 79920 }, { "grad_norm": 0.4283934235572815, "learning_rate": 1.0615189557398846e-05, "loss": 0.0204, "step": 79930 }, { "grad_norm": 0.44535207748413086, "learning_rate": 1.060500528109386e-05, "loss": 0.0192, "step": 79940 }, { "grad_norm": 0.24649661779403687, "learning_rate": 1.0594825312966272e-05, "loss": 0.0215, "step": 79950 }, { "grad_norm": 0.39229515194892883, "learning_rate": 1.0584649654129336e-05, "loss": 0.0221, "step": 79960 }, { "grad_norm": 0.30081748962402344, "learning_rate": 1.0574478305695863e-05, "loss": 0.0182, "step": 79970 }, { "grad_norm": 0.6511485576629639, "learning_rate": 1.0564311268778154e-05, "loss": 0.0224, "step": 79980 }, { "grad_norm": 0.5642109513282776, "learning_rate": 1.0554148544488073e-05, "loss": 0.0197, "step": 79990 }, { "grad_norm": 0.34399524331092834, "learning_rate": 1.0543990133937004e-05, "loss": 0.0197, "step": 80000 }, { "grad_norm": 0.5671305656433105, "learning_rate": 1.0533836038235868e-05, "loss": 0.0237, "step": 80010 }, { "grad_norm": 1.407564640045166, "learning_rate": 1.052368625849507e-05, "loss": 0.0235, "step": 80020 }, { "grad_norm": 0.5656149387359619, "learning_rate": 1.05135407958246e-05, "loss": 0.0219, "step": 80030 }, { "grad_norm": 0.23472754657268524, "learning_rate": 1.0503399651333934e-05, "loss": 0.0212, "step": 80040 }, { "grad_norm": 0.4387829601764679, "learning_rate": 1.04932628261321e-05, "loss": 0.0206, "step": 80050 }, { "grad_norm": 0.25593727827072144, "learning_rate": 1.0483130321327644e-05, "loss": 0.0207, "step": 80060 }, { "grad_norm": 0.3442217707633972, "learning_rate": 1.0473002138028654e-05, "loss": 0.0219, "step": 80070 }, { "grad_norm": 0.33880579471588135, "learning_rate": 1.0462878277342713e-05, "loss": 0.0187, "step": 80080 }, { "grad_norm": 0.3944133520126343, "learning_rate": 1.0452758740376972e-05, "loss": 0.0239, "step": 80090 }, { "grad_norm": 0.38882219791412354, "learning_rate": 1.0442643528238067e-05, "loss": 0.0172, "step": 80100 }, { "grad_norm": 0.2586967349052429, "learning_rate": 1.0432532642032188e-05, "loss": 0.0209, "step": 80110 }, { "grad_norm": 0.25568699836730957, "learning_rate": 1.042242608286505e-05, "loss": 0.0236, "step": 80120 }, { "grad_norm": 0.38863447308540344, "learning_rate": 1.041232385184191e-05, "loss": 0.0179, "step": 80130 }, { "grad_norm": 0.28856736421585083, "learning_rate": 1.0402225950067495e-05, "loss": 0.0199, "step": 80140 }, { "grad_norm": 0.30516067147254944, "learning_rate": 1.0392132378646119e-05, "loss": 0.0247, "step": 80150 }, { "grad_norm": 0.3299314081668854, "learning_rate": 1.0382043138681607e-05, "loss": 0.0224, "step": 80160 }, { "grad_norm": 0.27249574661254883, "learning_rate": 1.0371958231277284e-05, "loss": 0.022, "step": 80170 }, { "grad_norm": 0.35410261154174805, "learning_rate": 1.0361877657536024e-05, "loss": 0.0222, "step": 80180 }, { "grad_norm": 0.2346048206090927, "learning_rate": 1.035180141856023e-05, "loss": 0.0252, "step": 80190 }, { "grad_norm": 0.34570494294166565, "learning_rate": 1.0341729515451836e-05, "loss": 0.0226, "step": 80200 }, { "grad_norm": 0.3170139193534851, "learning_rate": 1.0331661949312265e-05, "loss": 0.0217, "step": 80210 }, { "grad_norm": 0.44992226362228394, "learning_rate": 1.0321598721242514e-05, "loss": 0.0201, "step": 80220 }, { "grad_norm": 0.4187375009059906, "learning_rate": 1.0311539832343064e-05, "loss": 0.0197, "step": 80230 }, { "grad_norm": 0.34222063422203064, "learning_rate": 1.0301485283713942e-05, "loss": 0.0227, "step": 80240 }, { "grad_norm": 0.3680195212364197, "learning_rate": 1.0291435076454703e-05, "loss": 0.0265, "step": 80250 }, { "grad_norm": 0.3148448169231415, "learning_rate": 1.028138921166444e-05, "loss": 0.0193, "step": 80260 }, { "grad_norm": 0.27513784170150757, "learning_rate": 1.0271347690441719e-05, "loss": 0.0258, "step": 80270 }, { "grad_norm": 0.76393061876297, "learning_rate": 1.0261310513884698e-05, "loss": 0.0227, "step": 80280 }, { "grad_norm": 1.2355917692184448, "learning_rate": 1.0251277683090998e-05, "loss": 0.0194, "step": 80290 }, { "grad_norm": 0.2570533752441406, "learning_rate": 1.0241249199157826e-05, "loss": 0.0179, "step": 80300 }, { "grad_norm": 0.20600877702236176, "learning_rate": 1.0231225063181837e-05, "loss": 0.0217, "step": 80310 }, { "grad_norm": 0.5006436705589294, "learning_rate": 1.0221205276259316e-05, "loss": 0.018, "step": 80320 }, { "grad_norm": 0.5358235836029053, "learning_rate": 1.0211189839485958e-05, "loss": 0.0205, "step": 80330 }, { "grad_norm": 0.6123287677764893, "learning_rate": 1.0201178753957075e-05, "loss": 0.0252, "step": 80340 }, { "grad_norm": 0.4053846001625061, "learning_rate": 1.0191172020767432e-05, "loss": 0.0212, "step": 80350 }, { "grad_norm": 0.3719913959503174, "learning_rate": 1.0181169641011362e-05, "loss": 0.026, "step": 80360 }, { "grad_norm": 0.2631169855594635, "learning_rate": 1.0171171615782721e-05, "loss": 0.0184, "step": 80370 }, { "grad_norm": 0.1992134153842926, "learning_rate": 1.0161177946174866e-05, "loss": 0.02, "step": 80380 }, { "grad_norm": 0.39780193567276, "learning_rate": 1.0151188633280706e-05, "loss": 0.0205, "step": 80390 }, { "grad_norm": 0.27678364515304565, "learning_rate": 1.0141203678192635e-05, "loss": 0.0271, "step": 80400 }, { "grad_norm": 0.4536985158920288, "learning_rate": 1.0131223082002612e-05, "loss": 0.0219, "step": 80410 }, { "grad_norm": 0.2607167065143585, "learning_rate": 1.012124684580208e-05, "loss": 0.0205, "step": 80420 }, { "grad_norm": 0.20300821959972382, "learning_rate": 1.0111274970682039e-05, "loss": 0.0198, "step": 80430 }, { "grad_norm": 0.26557356119155884, "learning_rate": 1.010130745773299e-05, "loss": 0.0286, "step": 80440 }, { "grad_norm": 0.5747720003128052, "learning_rate": 1.0091344308044986e-05, "loss": 0.0213, "step": 80450 }, { "grad_norm": 0.3084496259689331, "learning_rate": 1.0081385522707554e-05, "loss": 0.025, "step": 80460 }, { "grad_norm": 0.40248075127601624, "learning_rate": 1.0071431102809803e-05, "loss": 0.0204, "step": 80470 }, { "grad_norm": 0.3652232885360718, "learning_rate": 1.0061481049440297e-05, "loss": 0.0219, "step": 80480 }, { "grad_norm": 0.3174658417701721, "learning_rate": 1.0051535363687187e-05, "loss": 0.0194, "step": 80490 }, { "grad_norm": 0.31227174401283264, "learning_rate": 1.0041594046638087e-05, "loss": 0.0158, "step": 80500 }, { "grad_norm": 0.44873520731925964, "learning_rate": 1.0031657099380204e-05, "loss": 0.0215, "step": 80510 }, { "grad_norm": 0.3228916823863983, "learning_rate": 1.0021724523000204e-05, "loss": 0.0244, "step": 80520 }, { "grad_norm": 0.31501781940460205, "learning_rate": 1.001179631858431e-05, "loss": 0.0206, "step": 80530 }, { "grad_norm": 0.319828599691391, "learning_rate": 1.0001872487218245e-05, "loss": 0.0213, "step": 80540 }, { "grad_norm": 0.6314842700958252, "learning_rate": 9.991953029987272e-06, "loss": 0.0227, "step": 80550 }, { "grad_norm": 0.37804076075553894, "learning_rate": 9.982037947976147e-06, "loss": 0.0246, "step": 80560 }, { "grad_norm": 0.27370402216911316, "learning_rate": 9.972127242269208e-06, "loss": 0.021, "step": 80570 }, { "grad_norm": 0.4043944478034973, "learning_rate": 9.962220913950238e-06, "loss": 0.0223, "step": 80580 }, { "grad_norm": 0.34095096588134766, "learning_rate": 9.952318964102591e-06, "loss": 0.0234, "step": 80590 }, { "grad_norm": 0.4740200340747833, "learning_rate": 9.942421393809148e-06, "loss": 0.0187, "step": 80600 }, { "grad_norm": 0.3315405547618866, "learning_rate": 9.932528204152264e-06, "loss": 0.0197, "step": 80610 }, { "grad_norm": 0.5320997834205627, "learning_rate": 9.922639396213856e-06, "loss": 0.0239, "step": 80620 }, { "grad_norm": 0.3423282504081726, "learning_rate": 9.912754971075344e-06, "loss": 0.0279, "step": 80630 }, { "grad_norm": 0.25315573811531067, "learning_rate": 9.902874929817696e-06, "loss": 0.0206, "step": 80640 }, { "grad_norm": 0.4564969539642334, "learning_rate": 9.892999273521342e-06, "loss": 0.0286, "step": 80650 }, { "grad_norm": 0.6770726442337036, "learning_rate": 9.883128003266302e-06, "loss": 0.023, "step": 80660 }, { "grad_norm": 0.34796592593193054, "learning_rate": 9.87326112013206e-06, "loss": 0.0183, "step": 80670 }, { "grad_norm": 0.3141193687915802, "learning_rate": 9.863398625197662e-06, "loss": 0.0211, "step": 80680 }, { "grad_norm": 0.27550604939460754, "learning_rate": 9.853540519541615e-06, "loss": 0.0202, "step": 80690 }, { "grad_norm": 0.3194412589073181, "learning_rate": 9.843686804242053e-06, "loss": 0.0215, "step": 80700 }, { "grad_norm": 1.2168231010437012, "learning_rate": 9.833837480376506e-06, "loss": 0.0251, "step": 80710 }, { "grad_norm": 0.6380451917648315, "learning_rate": 9.82399254902212e-06, "loss": 0.0244, "step": 80720 }, { "grad_norm": 0.37542885541915894, "learning_rate": 9.814152011255494e-06, "loss": 0.0243, "step": 80730 }, { "grad_norm": 0.48618990182876587, "learning_rate": 9.804315868152797e-06, "loss": 0.0211, "step": 80740 }, { "grad_norm": 0.29428839683532715, "learning_rate": 9.79448412078966e-06, "loss": 0.0207, "step": 80750 }, { "grad_norm": 0.30562740564346313, "learning_rate": 9.784656770241312e-06, "loss": 0.0196, "step": 80760 }, { "grad_norm": 0.27010008692741394, "learning_rate": 9.77483381758243e-06, "loss": 0.0177, "step": 80770 }, { "grad_norm": 0.28052064776420593, "learning_rate": 9.76501526388725e-06, "loss": 0.0227, "step": 80780 }, { "grad_norm": 0.2757604122161865, "learning_rate": 9.7552011102295e-06, "loss": 0.0211, "step": 80790 }, { "grad_norm": 0.6902315616607666, "learning_rate": 9.745391357682449e-06, "loss": 0.0248, "step": 80800 }, { "grad_norm": 0.5636873245239258, "learning_rate": 9.735586007318875e-06, "loss": 0.0209, "step": 80810 }, { "grad_norm": 0.27817466855049133, "learning_rate": 9.725785060211096e-06, "loss": 0.0265, "step": 80820 }, { "grad_norm": 0.7392181754112244, "learning_rate": 9.715988517430896e-06, "loss": 0.0238, "step": 80830 }, { "grad_norm": 0.2910546362400055, "learning_rate": 9.706196380049625e-06, "loss": 0.0181, "step": 80840 }, { "grad_norm": 0.8565326929092407, "learning_rate": 9.69640864913815e-06, "loss": 0.0221, "step": 80850 }, { "grad_norm": 0.5448391437530518, "learning_rate": 9.686625325766818e-06, "loss": 0.025, "step": 80860 }, { "grad_norm": 0.8208802342414856, "learning_rate": 9.67684641100554e-06, "loss": 0.0219, "step": 80870 }, { "grad_norm": 0.6149570941925049, "learning_rate": 9.667071905923692e-06, "loss": 0.0258, "step": 80880 }, { "grad_norm": 0.6098200082778931, "learning_rate": 9.657301811590242e-06, "loss": 0.0204, "step": 80890 }, { "grad_norm": 0.5488348007202148, "learning_rate": 9.647536129073603e-06, "loss": 0.0201, "step": 80900 }, { "grad_norm": 0.2368641346693039, "learning_rate": 9.637774859441751e-06, "loss": 0.0261, "step": 80910 }, { "grad_norm": 0.502855658531189, "learning_rate": 9.628018003762152e-06, "loss": 0.0241, "step": 80920 }, { "grad_norm": 0.21552708745002747, "learning_rate": 9.618265563101813e-06, "loss": 0.0187, "step": 80930 }, { "grad_norm": 0.2786753177642822, "learning_rate": 9.608517538527211e-06, "loss": 0.0215, "step": 80940 }, { "grad_norm": 0.3012571930885315, "learning_rate": 9.598773931104431e-06, "loss": 0.0215, "step": 80950 }, { "grad_norm": 0.4511515498161316, "learning_rate": 9.589034741898983e-06, "loss": 0.0223, "step": 80960 }, { "grad_norm": 0.4198475182056427, "learning_rate": 9.579299971975947e-06, "loss": 0.017, "step": 80970 }, { "grad_norm": 0.6559696793556213, "learning_rate": 9.569569622399882e-06, "loss": 0.0218, "step": 80980 }, { "grad_norm": 0.3219086527824402, "learning_rate": 9.559843694234916e-06, "loss": 0.0186, "step": 80990 }, { "grad_norm": 0.3361577093601227, "learning_rate": 9.550122188544625e-06, "loss": 0.0207, "step": 81000 }, { "grad_norm": 0.4481527805328369, "learning_rate": 9.540405106392152e-06, "loss": 0.0206, "step": 81010 }, { "grad_norm": 0.3472151458263397, "learning_rate": 9.53069244884015e-06, "loss": 0.0251, "step": 81020 }, { "grad_norm": 0.16757415235042572, "learning_rate": 9.52098421695079e-06, "loss": 0.0214, "step": 81030 }, { "grad_norm": 0.22164514660835266, "learning_rate": 9.511280411785728e-06, "loss": 0.0192, "step": 81040 }, { "grad_norm": 0.4930211901664734, "learning_rate": 9.50158103440616e-06, "loss": 0.0218, "step": 81050 }, { "grad_norm": 0.2987940311431885, "learning_rate": 9.491886085872815e-06, "loss": 0.0218, "step": 81060 }, { "grad_norm": 0.35904306173324585, "learning_rate": 9.482195567245895e-06, "loss": 0.0329, "step": 81070 }, { "grad_norm": 0.3185996115207672, "learning_rate": 9.472509479585151e-06, "loss": 0.0238, "step": 81080 }, { "grad_norm": 0.3863837718963623, "learning_rate": 9.462827823949833e-06, "loss": 0.0269, "step": 81090 }, { "grad_norm": 0.6240180134773254, "learning_rate": 9.453150601398735e-06, "loss": 0.0225, "step": 81100 }, { "grad_norm": 0.48892828822135925, "learning_rate": 9.44347781299011e-06, "loss": 0.0241, "step": 81110 }, { "grad_norm": 0.3337486684322357, "learning_rate": 9.43380945978179e-06, "loss": 0.0264, "step": 81120 }, { "grad_norm": 1.0468955039978027, "learning_rate": 9.424145542831047e-06, "loss": 0.0215, "step": 81130 }, { "grad_norm": 1.8926385641098022, "learning_rate": 9.414486063194778e-06, "loss": 0.02, "step": 81140 }, { "grad_norm": 0.30138662457466125, "learning_rate": 9.404831021929273e-06, "loss": 0.0216, "step": 81150 }, { "grad_norm": 0.24662964046001434, "learning_rate": 9.395180420090427e-06, "loss": 0.018, "step": 81160 }, { "grad_norm": 0.3541744351387024, "learning_rate": 9.385534258733591e-06, "loss": 0.0182, "step": 81170 }, { "grad_norm": 0.2846076488494873, "learning_rate": 9.375892538913677e-06, "loss": 0.0206, "step": 81180 }, { "grad_norm": 0.2466617375612259, "learning_rate": 9.36625526168507e-06, "loss": 0.0198, "step": 81190 }, { "grad_norm": 0.43900373578071594, "learning_rate": 9.356622428101686e-06, "loss": 0.0203, "step": 81200 }, { "grad_norm": 0.5220575928688049, "learning_rate": 9.346994039216972e-06, "loss": 0.0191, "step": 81210 }, { "grad_norm": 0.4499857425689697, "learning_rate": 9.337370096083886e-06, "loss": 0.0222, "step": 81220 }, { "grad_norm": 0.7081239223480225, "learning_rate": 9.327750599754854e-06, "loss": 0.0169, "step": 81230 }, { "grad_norm": 0.3753662109375, "learning_rate": 9.318135551281875e-06, "loss": 0.0222, "step": 81240 }, { "grad_norm": 0.22965438663959503, "learning_rate": 9.308524951716424e-06, "loss": 0.0215, "step": 81250 }, { "grad_norm": 0.4023354947566986, "learning_rate": 9.298918802109497e-06, "loss": 0.0227, "step": 81260 }, { "grad_norm": 0.5969729423522949, "learning_rate": 9.289317103511625e-06, "loss": 0.0206, "step": 81270 }, { "grad_norm": 0.4722331762313843, "learning_rate": 9.279719856972835e-06, "loss": 0.0187, "step": 81280 }, { "grad_norm": 0.4259868562221527, "learning_rate": 9.270127063542645e-06, "loss": 0.023, "step": 81290 }, { "grad_norm": 0.3532835841178894, "learning_rate": 9.260538724270124e-06, "loss": 0.0244, "step": 81300 }, { "grad_norm": 0.4906117916107178, "learning_rate": 9.250954840203852e-06, "loss": 0.0257, "step": 81310 }, { "grad_norm": 0.25894543528556824, "learning_rate": 9.241375412391878e-06, "loss": 0.0195, "step": 81320 }, { "grad_norm": 0.3274286985397339, "learning_rate": 9.23180044188181e-06, "loss": 0.0181, "step": 81330 }, { "grad_norm": 0.39234718680381775, "learning_rate": 9.222229929720755e-06, "loss": 0.0156, "step": 81340 }, { "grad_norm": 0.22014698386192322, "learning_rate": 9.212663876955336e-06, "loss": 0.0162, "step": 81350 }, { "grad_norm": 0.3736426830291748, "learning_rate": 9.20310228463166e-06, "loss": 0.02, "step": 81360 }, { "grad_norm": 0.39812538027763367, "learning_rate": 9.19354515379539e-06, "loss": 0.0248, "step": 81370 }, { "grad_norm": 0.33086320757865906, "learning_rate": 9.183992485491666e-06, "loss": 0.0195, "step": 81380 }, { "grad_norm": 0.3491614758968353, "learning_rate": 9.174444280765154e-06, "loss": 0.0232, "step": 81390 }, { "grad_norm": 0.19621998071670532, "learning_rate": 9.164900540660032e-06, "loss": 0.0222, "step": 81400 }, { "grad_norm": 0.8517836332321167, "learning_rate": 9.155361266220008e-06, "loss": 0.0219, "step": 81410 }, { "grad_norm": 0.8176590204238892, "learning_rate": 9.145826458488254e-06, "loss": 0.0191, "step": 81420 }, { "grad_norm": 0.3282289206981659, "learning_rate": 9.136296118507509e-06, "loss": 0.0197, "step": 81430 }, { "grad_norm": 0.16390444338321686, "learning_rate": 9.126770247319966e-06, "loss": 0.0227, "step": 81440 }, { "grad_norm": 0.448881059885025, "learning_rate": 9.11724884596738e-06, "loss": 0.0226, "step": 81450 }, { "grad_norm": 0.13509739935398102, "learning_rate": 9.107731915490996e-06, "loss": 0.0185, "step": 81460 }, { "grad_norm": 0.2207895815372467, "learning_rate": 9.098219456931579e-06, "loss": 0.0243, "step": 81470 }, { "grad_norm": 0.43822091817855835, "learning_rate": 9.088711471329376e-06, "loss": 0.0196, "step": 81480 }, { "grad_norm": 0.3545336425304413, "learning_rate": 9.079207959724184e-06, "loss": 0.0228, "step": 81490 }, { "grad_norm": 0.26956498622894287, "learning_rate": 9.069708923155279e-06, "loss": 0.0216, "step": 81500 }, { "grad_norm": 0.35336893796920776, "learning_rate": 9.060214362661468e-06, "loss": 0.0178, "step": 81510 }, { "grad_norm": 0.30161958932876587, "learning_rate": 9.050724279281064e-06, "loss": 0.0229, "step": 81520 }, { "grad_norm": 0.2603616416454315, "learning_rate": 9.041238674051894e-06, "loss": 0.0178, "step": 81530 }, { "grad_norm": 0.2893913984298706, "learning_rate": 9.031757548011277e-06, "loss": 0.0235, "step": 81540 }, { "grad_norm": 0.2802931070327759, "learning_rate": 9.022280902196051e-06, "loss": 0.0227, "step": 81550 }, { "grad_norm": 0.7943295836448669, "learning_rate": 9.012808737642592e-06, "loss": 0.0196, "step": 81560 }, { "grad_norm": 0.35933300852775574, "learning_rate": 9.003341055386739e-06, "loss": 0.0224, "step": 81570 }, { "grad_norm": 0.2606038749217987, "learning_rate": 8.993877856463862e-06, "loss": 0.0193, "step": 81580 }, { "grad_norm": 0.2618131935596466, "learning_rate": 8.984419141908856e-06, "loss": 0.0188, "step": 81590 }, { "grad_norm": 0.47157400846481323, "learning_rate": 8.974964912756117e-06, "loss": 0.0199, "step": 81600 }, { "grad_norm": 0.8218373656272888, "learning_rate": 8.965515170039523e-06, "loss": 0.0233, "step": 81610 }, { "grad_norm": 0.45304396748542786, "learning_rate": 8.956069914792514e-06, "loss": 0.0206, "step": 81620 }, { "grad_norm": 0.26550614833831787, "learning_rate": 8.946629148047969e-06, "loss": 0.0217, "step": 81630 }, { "grad_norm": 0.32857444882392883, "learning_rate": 8.93719287083834e-06, "loss": 0.0194, "step": 81640 }, { "grad_norm": 0.4465011656284332, "learning_rate": 8.927761084195569e-06, "loss": 0.0205, "step": 81650 }, { "grad_norm": 0.3857864439487457, "learning_rate": 8.918333789151096e-06, "loss": 0.0199, "step": 81660 }, { "grad_norm": 0.5075552463531494, "learning_rate": 8.90891098673587e-06, "loss": 0.0222, "step": 81670 }, { "grad_norm": 0.31294819712638855, "learning_rate": 8.89949267798037e-06, "loss": 0.0192, "step": 81680 }, { "grad_norm": 0.4412868916988373, "learning_rate": 8.890078863914542e-06, "loss": 0.0196, "step": 81690 }, { "grad_norm": 0.4638003706932068, "learning_rate": 8.880669545567894e-06, "loss": 0.0194, "step": 81700 }, { "grad_norm": 0.3951517343521118, "learning_rate": 8.871264723969381e-06, "loss": 0.0204, "step": 81710 }, { "grad_norm": 0.437149316072464, "learning_rate": 8.86186440014754e-06, "loss": 0.0187, "step": 81720 }, { "grad_norm": 0.5101723074913025, "learning_rate": 8.852468575130351e-06, "loss": 0.0216, "step": 81730 }, { "grad_norm": 0.3513679802417755, "learning_rate": 8.843077249945347e-06, "loss": 0.0173, "step": 81740 }, { "grad_norm": 0.2657313346862793, "learning_rate": 8.833690425619523e-06, "loss": 0.0168, "step": 81750 }, { "grad_norm": 0.34632787108421326, "learning_rate": 8.82430810317942e-06, "loss": 0.0222, "step": 81760 }, { "grad_norm": 0.34386393427848816, "learning_rate": 8.814930283651074e-06, "loss": 0.0202, "step": 81770 }, { "grad_norm": 4.476842403411865, "learning_rate": 8.805556968060048e-06, "loss": 0.0194, "step": 81780 }, { "grad_norm": 0.5375921130180359, "learning_rate": 8.796188157431362e-06, "loss": 0.019, "step": 81790 }, { "grad_norm": 0.29668697714805603, "learning_rate": 8.786823852789595e-06, "loss": 0.0228, "step": 81800 }, { "grad_norm": 0.3502840995788574, "learning_rate": 8.777464055158813e-06, "loss": 0.0243, "step": 81810 }, { "grad_norm": 0.3442508578300476, "learning_rate": 8.768108765562578e-06, "loss": 0.0188, "step": 81820 }, { "grad_norm": 0.4760729968547821, "learning_rate": 8.758757985023975e-06, "loss": 0.0187, "step": 81830 }, { "grad_norm": 0.32757124304771423, "learning_rate": 8.749411714565598e-06, "loss": 0.0197, "step": 81840 }, { "grad_norm": 0.3120107352733612, "learning_rate": 8.74006995520954e-06, "loss": 0.0218, "step": 81850 }, { "grad_norm": 0.6787510514259338, "learning_rate": 8.730732707977397e-06, "loss": 0.0192, "step": 81860 }, { "grad_norm": 0.35996556282043457, "learning_rate": 8.721399973890281e-06, "loss": 0.0254, "step": 81870 }, { "grad_norm": 0.2858773171901703, "learning_rate": 8.712071753968792e-06, "loss": 0.0199, "step": 81880 }, { "grad_norm": 0.37439802289009094, "learning_rate": 8.702748049233072e-06, "loss": 0.024, "step": 81890 }, { "grad_norm": 0.513042688369751, "learning_rate": 8.693428860702718e-06, "loss": 0.019, "step": 81900 }, { "grad_norm": 0.22536274790763855, "learning_rate": 8.684114189396898e-06, "loss": 0.021, "step": 81910 }, { "grad_norm": 0.2925085425376892, "learning_rate": 8.674804036334222e-06, "loss": 0.0179, "step": 81920 }, { "grad_norm": 0.7135785818099976, "learning_rate": 8.665498402532857e-06, "loss": 0.022, "step": 81930 }, { "grad_norm": 0.36571305990219116, "learning_rate": 8.656197289010432e-06, "loss": 0.0245, "step": 81940 }, { "grad_norm": 0.15733793377876282, "learning_rate": 8.646900696784122e-06, "loss": 0.0192, "step": 81950 }, { "grad_norm": 0.9060038924217224, "learning_rate": 8.637608626870553e-06, "loss": 0.0203, "step": 81960 }, { "grad_norm": 0.5508208870887756, "learning_rate": 8.628321080285944e-06, "loss": 0.0191, "step": 81970 }, { "grad_norm": 0.2603261470794678, "learning_rate": 8.619038058045925e-06, "loss": 0.0226, "step": 81980 }, { "grad_norm": 0.3956504166126251, "learning_rate": 8.609759561165704e-06, "loss": 0.0214, "step": 81990 }, { "grad_norm": 0.3565793037414551, "learning_rate": 8.600485590659934e-06, "loss": 0.0214, "step": 82000 }, { "grad_norm": 0.43176114559173584, "learning_rate": 8.591216147542813e-06, "loss": 0.0239, "step": 82010 }, { "grad_norm": 0.258124977350235, "learning_rate": 8.581951232828045e-06, "loss": 0.024, "step": 82020 }, { "grad_norm": 0.5050873756408691, "learning_rate": 8.572690847528825e-06, "loss": 0.0203, "step": 82030 }, { "grad_norm": 0.820381760597229, "learning_rate": 8.563434992657838e-06, "loss": 0.0223, "step": 82040 }, { "grad_norm": 0.3155370354652405, "learning_rate": 8.554183669227307e-06, "loss": 0.0177, "step": 82050 }, { "grad_norm": 0.47260811924934387, "learning_rate": 8.544936878248937e-06, "loss": 0.0215, "step": 82060 }, { "grad_norm": 0.31140920519828796, "learning_rate": 8.535694620733942e-06, "loss": 0.0202, "step": 82070 }, { "grad_norm": 0.528369665145874, "learning_rate": 8.526456897693053e-06, "loss": 0.0216, "step": 82080 }, { "grad_norm": 0.404163122177124, "learning_rate": 8.517223710136456e-06, "loss": 0.0216, "step": 82090 }, { "grad_norm": 0.29904991388320923, "learning_rate": 8.507995059073931e-06, "loss": 0.0185, "step": 82100 }, { "grad_norm": 0.5078774690628052, "learning_rate": 8.498770945514673e-06, "loss": 0.0217, "step": 82110 }, { "grad_norm": 0.4540751576423645, "learning_rate": 8.489551370467436e-06, "loss": 0.0202, "step": 82120 }, { "grad_norm": 0.4021296501159668, "learning_rate": 8.480336334940442e-06, "loss": 0.0257, "step": 82130 }, { "grad_norm": 0.5026265382766724, "learning_rate": 8.471125839941458e-06, "loss": 0.0223, "step": 82140 }, { "grad_norm": 0.26780617237091064, "learning_rate": 8.461919886477682e-06, "loss": 0.0248, "step": 82150 }, { "grad_norm": 0.5161992311477661, "learning_rate": 8.452718475555927e-06, "loss": 0.0214, "step": 82160 }, { "grad_norm": 0.4424575865268707, "learning_rate": 8.443521608182398e-06, "loss": 0.0214, "step": 82170 }, { "grad_norm": 0.5586820840835571, "learning_rate": 8.43432928536288e-06, "loss": 0.021, "step": 82180 }, { "grad_norm": 0.2864498496055603, "learning_rate": 8.425141508102607e-06, "loss": 0.017, "step": 82190 }, { "grad_norm": 0.2517399191856384, "learning_rate": 8.415958277406365e-06, "loss": 0.0195, "step": 82200 }, { "grad_norm": 0.4920392334461212, "learning_rate": 8.406779594278392e-06, "loss": 0.0211, "step": 82210 }, { "grad_norm": 0.6766623258590698, "learning_rate": 8.397605459722468e-06, "loss": 0.0199, "step": 82220 }, { "grad_norm": 0.3992379605770111, "learning_rate": 8.388435874741863e-06, "loss": 0.0175, "step": 82230 }, { "grad_norm": 0.5242770910263062, "learning_rate": 8.37927084033936e-06, "loss": 0.0219, "step": 82240 }, { "grad_norm": 0.2786344587802887, "learning_rate": 8.370110357517208e-06, "loss": 0.0196, "step": 82250 }, { "grad_norm": 0.3780357241630554, "learning_rate": 8.360954427277201e-06, "loss": 0.021, "step": 82260 }, { "grad_norm": 0.5221051573753357, "learning_rate": 8.351803050620627e-06, "loss": 0.0221, "step": 82270 }, { "grad_norm": 0.38998523354530334, "learning_rate": 8.34265622854824e-06, "loss": 0.0174, "step": 82280 }, { "grad_norm": 0.3399237096309662, "learning_rate": 8.333513962060335e-06, "loss": 0.0208, "step": 82290 }, { "grad_norm": 0.1546163111925125, "learning_rate": 8.324376252156701e-06, "loss": 0.0182, "step": 82300 }, { "grad_norm": 0.31548458337783813, "learning_rate": 8.315243099836633e-06, "loss": 0.0179, "step": 82310 }, { "grad_norm": 0.2708713710308075, "learning_rate": 8.3061145060989e-06, "loss": 0.02, "step": 82320 }, { "grad_norm": 0.29041996598243713, "learning_rate": 8.296990471941806e-06, "loss": 0.0204, "step": 82330 }, { "grad_norm": 0.32754290103912354, "learning_rate": 8.287870998363112e-06, "loss": 0.0184, "step": 82340 }, { "grad_norm": 0.38818883895874023, "learning_rate": 8.278756086360156e-06, "loss": 0.016, "step": 82350 }, { "grad_norm": 0.43609917163848877, "learning_rate": 8.269645736929693e-06, "loss": 0.0248, "step": 82360 }, { "grad_norm": 0.4047186076641083, "learning_rate": 8.260539951068048e-06, "loss": 0.0253, "step": 82370 }, { "grad_norm": 0.3286842703819275, "learning_rate": 8.251438729770983e-06, "loss": 0.02, "step": 82380 }, { "grad_norm": 0.29817280173301697, "learning_rate": 8.242342074033826e-06, "loss": 0.0273, "step": 82390 }, { "grad_norm": 0.3179386258125305, "learning_rate": 8.233249984851344e-06, "loss": 0.0237, "step": 82400 }, { "grad_norm": 0.37226957082748413, "learning_rate": 8.224162463217849e-06, "loss": 0.0205, "step": 82410 }, { "grad_norm": 0.4350636601448059, "learning_rate": 8.215079510127139e-06, "loss": 0.0235, "step": 82420 }, { "grad_norm": 0.3980742394924164, "learning_rate": 8.206001126572521e-06, "loss": 0.0165, "step": 82430 }, { "grad_norm": 0.24309773743152618, "learning_rate": 8.196927313546776e-06, "loss": 0.0181, "step": 82440 }, { "grad_norm": 0.5482433438301086, "learning_rate": 8.18785807204222e-06, "loss": 0.0207, "step": 82450 }, { "grad_norm": 0.28834226727485657, "learning_rate": 8.178793403050627e-06, "loss": 0.0195, "step": 82460 }, { "grad_norm": 1.8324358463287354, "learning_rate": 8.16973330756331e-06, "loss": 0.0218, "step": 82470 }, { "grad_norm": 0.5066149830818176, "learning_rate": 8.16067778657107e-06, "loss": 0.0262, "step": 82480 }, { "grad_norm": 0.2613844573497772, "learning_rate": 8.151626841064214e-06, "loss": 0.0237, "step": 82490 }, { "grad_norm": 0.24896064400672913, "learning_rate": 8.142580472032518e-06, "loss": 0.022, "step": 82500 }, { "grad_norm": 0.3130342364311218, "learning_rate": 8.133538680465285e-06, "loss": 0.0162, "step": 82510 }, { "grad_norm": 0.17729653418064117, "learning_rate": 8.124501467351336e-06, "loss": 0.0218, "step": 82520 }, { "grad_norm": 0.41126593947410583, "learning_rate": 8.11546883367893e-06, "loss": 0.0208, "step": 82530 }, { "grad_norm": 0.2705073654651642, "learning_rate": 8.106440780435882e-06, "loss": 0.0255, "step": 82540 }, { "grad_norm": 0.3081508278846741, "learning_rate": 8.097417308609485e-06, "loss": 0.0261, "step": 82550 }, { "grad_norm": 0.7390328645706177, "learning_rate": 8.088398419186544e-06, "loss": 0.0219, "step": 82560 }, { "grad_norm": 0.3424462080001831, "learning_rate": 8.079384113153321e-06, "loss": 0.0207, "step": 82570 }, { "grad_norm": 0.642723798751831, "learning_rate": 8.070374391495644e-06, "loss": 0.0187, "step": 82580 }, { "grad_norm": 0.4626835286617279, "learning_rate": 8.061369255198764e-06, "loss": 0.0207, "step": 82590 }, { "grad_norm": 0.4123421013355255, "learning_rate": 8.052368705247493e-06, "loss": 0.0216, "step": 82600 }, { "grad_norm": 0.374637633562088, "learning_rate": 8.043372742626114e-06, "loss": 0.0212, "step": 82610 }, { "grad_norm": 0.5549618601799011, "learning_rate": 8.034381368318422e-06, "loss": 0.0191, "step": 82620 }, { "grad_norm": 0.17033950984477997, "learning_rate": 8.025394583307677e-06, "loss": 0.0217, "step": 82630 }, { "grad_norm": 0.4977470636367798, "learning_rate": 8.016412388576683e-06, "loss": 0.0207, "step": 82640 }, { "grad_norm": 0.3609525263309479, "learning_rate": 8.007434785107699e-06, "loss": 0.0203, "step": 82650 }, { "grad_norm": 0.25958332419395447, "learning_rate": 7.998461773882509e-06, "loss": 0.0194, "step": 82660 }, { "grad_norm": 0.43254441022872925, "learning_rate": 7.989493355882393e-06, "loss": 0.0201, "step": 82670 }, { "grad_norm": 0.27426713705062866, "learning_rate": 7.980529532088137e-06, "loss": 0.0213, "step": 82680 }, { "grad_norm": 0.4456322193145752, "learning_rate": 7.97157030347998e-06, "loss": 0.0215, "step": 82690 }, { "grad_norm": 0.25932297110557556, "learning_rate": 7.96261567103772e-06, "loss": 0.0238, "step": 82700 }, { "grad_norm": 0.2755168080329895, "learning_rate": 7.953665635740593e-06, "loss": 0.0204, "step": 82710 }, { "grad_norm": 0.36782342195510864, "learning_rate": 7.94472019856738e-06, "loss": 0.0197, "step": 82720 }, { "grad_norm": 0.20430192351341248, "learning_rate": 7.935779360496337e-06, "loss": 0.0176, "step": 82730 }, { "grad_norm": 0.5890440940856934, "learning_rate": 7.92684312250523e-06, "loss": 0.0195, "step": 82740 }, { "grad_norm": 0.46739301085472107, "learning_rate": 7.917911485571289e-06, "loss": 0.0219, "step": 82750 }, { "grad_norm": 0.483796089887619, "learning_rate": 7.90898445067128e-06, "loss": 0.0171, "step": 82760 }, { "grad_norm": 0.31122004985809326, "learning_rate": 7.900062018781462e-06, "loss": 0.0252, "step": 82770 }, { "grad_norm": 0.36065465211868286, "learning_rate": 7.891144190877553e-06, "loss": 0.0255, "step": 82780 }, { "grad_norm": 0.33715102076530457, "learning_rate": 7.882230967934806e-06, "loss": 0.0173, "step": 82790 }, { "grad_norm": 0.5148841142654419, "learning_rate": 7.873322350927959e-06, "loss": 0.0192, "step": 82800 }, { "grad_norm": 0.6405506134033203, "learning_rate": 7.864418340831253e-06, "loss": 0.0246, "step": 82810 }, { "grad_norm": 0.3024239242076874, "learning_rate": 7.855518938618395e-06, "loss": 0.0188, "step": 82820 }, { "grad_norm": 0.5262792110443115, "learning_rate": 7.846624145262633e-06, "loss": 0.0245, "step": 82830 }, { "grad_norm": 0.5236847996711731, "learning_rate": 7.837733961736672e-06, "loss": 0.0207, "step": 82840 }, { "grad_norm": 0.3991010785102844, "learning_rate": 7.828848389012727e-06, "loss": 0.0201, "step": 82850 }, { "grad_norm": 0.46208181977272034, "learning_rate": 7.819967428062524e-06, "loss": 0.0226, "step": 82860 }, { "grad_norm": 0.3259158432483673, "learning_rate": 7.811091079857274e-06, "loss": 0.0219, "step": 82870 }, { "grad_norm": 0.34826892614364624, "learning_rate": 7.802219345367667e-06, "loss": 0.0184, "step": 82880 }, { "grad_norm": 0.3109346032142639, "learning_rate": 7.793352225563916e-06, "loss": 0.0183, "step": 82890 }, { "grad_norm": 0.60244220495224, "learning_rate": 7.7844897214157e-06, "loss": 0.0228, "step": 82900 }, { "grad_norm": 0.3478146195411682, "learning_rate": 7.775631833892228e-06, "loss": 0.0228, "step": 82910 }, { "grad_norm": 0.5546814203262329, "learning_rate": 7.76677856396215e-06, "loss": 0.0207, "step": 82920 }, { "grad_norm": 0.3425653278827667, "learning_rate": 7.757929912593697e-06, "loss": 0.0276, "step": 82930 }, { "grad_norm": 0.46475374698638916, "learning_rate": 7.749085880754509e-06, "loss": 0.0192, "step": 82940 }, { "grad_norm": 0.5901358723640442, "learning_rate": 7.740246469411777e-06, "loss": 0.0239, "step": 82950 }, { "grad_norm": 0.4236070215702057, "learning_rate": 7.731411679532142e-06, "loss": 0.0184, "step": 82960 }, { "grad_norm": 0.19890323281288147, "learning_rate": 7.722581512081779e-06, "loss": 0.0231, "step": 82970 }, { "grad_norm": 0.6640766263008118, "learning_rate": 7.713755968026331e-06, "loss": 0.0216, "step": 82980 }, { "grad_norm": 0.3031066954135895, "learning_rate": 7.704935048330969e-06, "loss": 0.0219, "step": 82990 }, { "grad_norm": 0.5337163805961609, "learning_rate": 7.696118753960313e-06, "loss": 0.0206, "step": 83000 }, { "grad_norm": 0.256737619638443, "learning_rate": 7.687307085878503e-06, "loss": 0.0242, "step": 83010 }, { "grad_norm": 0.3631593585014343, "learning_rate": 7.678500045049186e-06, "loss": 0.0197, "step": 83020 }, { "grad_norm": 0.5260511636734009, "learning_rate": 7.669697632435463e-06, "loss": 0.0197, "step": 83030 }, { "grad_norm": 0.20122723281383514, "learning_rate": 7.660899848999963e-06, "loss": 0.0178, "step": 83040 }, { "grad_norm": 0.2685356140136719, "learning_rate": 7.652106695704803e-06, "loss": 0.0218, "step": 83050 }, { "grad_norm": 0.6152853965759277, "learning_rate": 7.643318173511598e-06, "loss": 0.0202, "step": 83060 }, { "grad_norm": 0.33893901109695435, "learning_rate": 7.634534283381423e-06, "loss": 0.0205, "step": 83070 }, { "grad_norm": 0.2549422085285187, "learning_rate": 7.625755026274889e-06, "loss": 0.0203, "step": 83080 }, { "grad_norm": 0.38987740874290466, "learning_rate": 7.616980403152069e-06, "loss": 0.0189, "step": 83090 }, { "grad_norm": 0.3653576970100403, "learning_rate": 7.608210414972561e-06, "loss": 0.0212, "step": 83100 }, { "grad_norm": 0.25051215291023254, "learning_rate": 7.599445062695404e-06, "loss": 0.0208, "step": 83110 }, { "grad_norm": 0.395630419254303, "learning_rate": 7.590684347279214e-06, "loss": 0.0227, "step": 83120 }, { "grad_norm": 0.39266499876976013, "learning_rate": 7.581928269682004e-06, "loss": 0.0219, "step": 83130 }, { "grad_norm": 0.3653024137020111, "learning_rate": 7.573176830861361e-06, "loss": 0.0218, "step": 83140 }, { "grad_norm": 0.2927153706550598, "learning_rate": 7.5644300317743e-06, "loss": 0.0196, "step": 83150 }, { "grad_norm": 0.4066086709499359, "learning_rate": 7.555687873377376e-06, "loss": 0.0204, "step": 83160 }, { "grad_norm": 0.4015410244464874, "learning_rate": 7.546950356626592e-06, "loss": 0.0285, "step": 83170 }, { "grad_norm": 0.34351110458374023, "learning_rate": 7.538217482477516e-06, "loss": 0.0193, "step": 83180 }, { "grad_norm": 0.4153967499732971, "learning_rate": 7.529489251885119e-06, "loss": 0.021, "step": 83190 }, { "grad_norm": 0.3241827189922333, "learning_rate": 7.520765665803941e-06, "loss": 0.0206, "step": 83200 }, { "grad_norm": 0.30415260791778564, "learning_rate": 7.512046725187949e-06, "loss": 0.0212, "step": 83210 }, { "grad_norm": 0.47779974341392517, "learning_rate": 7.503332430990645e-06, "loss": 0.0194, "step": 83220 }, { "grad_norm": 0.3484482169151306, "learning_rate": 7.494622784165017e-06, "loss": 0.018, "step": 83230 }, { "grad_norm": 0.2677086293697357, "learning_rate": 7.485917785663543e-06, "loss": 0.023, "step": 83240 }, { "grad_norm": 0.260513573884964, "learning_rate": 7.477217436438172e-06, "loss": 0.0196, "step": 83250 }, { "grad_norm": 0.32758814096450806, "learning_rate": 7.46852173744037e-06, "loss": 0.0183, "step": 83260 }, { "grad_norm": 0.41040948033332825, "learning_rate": 7.4598306896211004e-06, "loss": 0.0229, "step": 83270 }, { "grad_norm": 0.5050115585327148, "learning_rate": 7.451144293930773e-06, "loss": 0.0228, "step": 83280 }, { "grad_norm": 0.324465811252594, "learning_rate": 7.4424625513193455e-06, "loss": 0.0224, "step": 83290 }, { "grad_norm": 0.3423333764076233, "learning_rate": 7.433785462736209e-06, "loss": 0.0199, "step": 83300 }, { "grad_norm": 0.28112301230430603, "learning_rate": 7.425113029130315e-06, "loss": 0.0164, "step": 83310 }, { "grad_norm": 0.5747941732406616, "learning_rate": 7.416445251450044e-06, "loss": 0.0165, "step": 83320 }, { "grad_norm": 0.7654414772987366, "learning_rate": 7.4077821306433035e-06, "loss": 0.0288, "step": 83330 }, { "grad_norm": 0.3435736298561096, "learning_rate": 7.399123667657465e-06, "loss": 0.0192, "step": 83340 }, { "grad_norm": 0.5797106027603149, "learning_rate": 7.390469863439426e-06, "loss": 0.0224, "step": 83350 }, { "grad_norm": 1.5230085849761963, "learning_rate": 7.381820718935512e-06, "loss": 0.0195, "step": 83360 }, { "grad_norm": 0.8880825638771057, "learning_rate": 7.373176235091633e-06, "loss": 0.0229, "step": 83370 }, { "grad_norm": 0.3769659996032715, "learning_rate": 7.364536412853101e-06, "loss": 0.02, "step": 83380 }, { "grad_norm": 0.39558860659599304, "learning_rate": 7.355901253164776e-06, "loss": 0.0198, "step": 83390 }, { "grad_norm": 0.4271148145198822, "learning_rate": 7.34727075697097e-06, "loss": 0.019, "step": 83400 }, { "grad_norm": 0.20490361750125885, "learning_rate": 7.3386449252155095e-06, "loss": 0.0167, "step": 83410 }, { "grad_norm": 0.6895883679389954, "learning_rate": 7.330023758841692e-06, "loss": 0.0201, "step": 83420 }, { "grad_norm": 0.43326523900032043, "learning_rate": 7.321407258792323e-06, "loss": 0.0213, "step": 83430 }, { "grad_norm": 0.3745209574699402, "learning_rate": 7.312795426009694e-06, "loss": 0.0195, "step": 83440 }, { "grad_norm": 0.3459973931312561, "learning_rate": 7.304188261435585e-06, "loss": 0.0191, "step": 83450 }, { "grad_norm": 0.30812570452690125, "learning_rate": 7.295585766011242e-06, "loss": 0.0189, "step": 83460 }, { "grad_norm": 0.6172653436660767, "learning_rate": 7.286987940677436e-06, "loss": 0.0158, "step": 83470 }, { "grad_norm": 0.3489744961261749, "learning_rate": 7.278394786374426e-06, "loss": 0.0269, "step": 83480 }, { "grad_norm": 0.37651094794273376, "learning_rate": 7.269806304041915e-06, "loss": 0.0191, "step": 83490 }, { "grad_norm": 0.44767892360687256, "learning_rate": 7.261222494619147e-06, "loss": 0.0211, "step": 83500 }, { "grad_norm": 0.221099391579628, "learning_rate": 7.252643359044825e-06, "loss": 0.0196, "step": 83510 }, { "grad_norm": 0.41208842396736145, "learning_rate": 7.244068898257172e-06, "loss": 0.0196, "step": 83520 }, { "grad_norm": 0.6221688389778137, "learning_rate": 7.2354991131938495e-06, "loss": 0.0237, "step": 83530 }, { "grad_norm": 0.5820417404174805, "learning_rate": 7.2269340047920574e-06, "loss": 0.0266, "step": 83540 }, { "grad_norm": 0.6478205323219299, "learning_rate": 7.218373573988435e-06, "loss": 0.0211, "step": 83550 }, { "grad_norm": 0.6999391913414001, "learning_rate": 7.209817821719178e-06, "loss": 0.02, "step": 83560 }, { "grad_norm": 0.39495715498924255, "learning_rate": 7.201266748919894e-06, "loss": 0.0241, "step": 83570 }, { "grad_norm": 0.25760260224342346, "learning_rate": 7.19272035652575e-06, "loss": 0.0185, "step": 83580 }, { "grad_norm": 0.5462153553962708, "learning_rate": 7.184178645471334e-06, "loss": 0.0249, "step": 83590 }, { "grad_norm": 0.7082539796829224, "learning_rate": 7.175641616690775e-06, "loss": 0.0188, "step": 83600 }, { "grad_norm": 0.8476572632789612, "learning_rate": 7.167109271117661e-06, "loss": 0.0212, "step": 83610 }, { "grad_norm": 0.49340546131134033, "learning_rate": 7.158581609685072e-06, "loss": 0.0186, "step": 83620 }, { "grad_norm": 0.3622308075428009, "learning_rate": 7.150058633325591e-06, "loss": 0.0226, "step": 83630 }, { "grad_norm": 0.22533011436462402, "learning_rate": 7.141540342971287e-06, "loss": 0.0225, "step": 83640 }, { "grad_norm": 0.2619938552379608, "learning_rate": 7.133026739553677e-06, "loss": 0.0231, "step": 83650 }, { "grad_norm": 0.24499715864658356, "learning_rate": 7.124517824003829e-06, "loss": 0.0202, "step": 83660 }, { "grad_norm": 0.7047871947288513, "learning_rate": 7.116013597252236e-06, "loss": 0.0218, "step": 83670 }, { "grad_norm": 0.3917883336544037, "learning_rate": 7.107514060228921e-06, "loss": 0.0206, "step": 83680 }, { "grad_norm": 0.5829023718833923, "learning_rate": 7.099019213863378e-06, "loss": 0.0215, "step": 83690 }, { "grad_norm": 0.4279494285583496, "learning_rate": 7.090529059084611e-06, "loss": 0.02, "step": 83700 }, { "grad_norm": 0.33119019865989685, "learning_rate": 7.082043596821058e-06, "loss": 0.0172, "step": 83710 }, { "grad_norm": 0.42938339710235596, "learning_rate": 7.07356282800069e-06, "loss": 0.0218, "step": 83720 }, { "grad_norm": 0.39760106801986694, "learning_rate": 7.0650867535509645e-06, "loss": 0.0201, "step": 83730 }, { "grad_norm": 0.4751950204372406, "learning_rate": 7.0566153743987895e-06, "loss": 0.0192, "step": 83740 }, { "grad_norm": 0.4179997742176056, "learning_rate": 7.04814869147059e-06, "loss": 0.0208, "step": 83750 }, { "grad_norm": 0.6870294809341431, "learning_rate": 7.0396867056922725e-06, "loss": 0.0231, "step": 83760 }, { "grad_norm": 0.3889513313770294, "learning_rate": 7.031229417989244e-06, "loss": 0.0186, "step": 83770 }, { "grad_norm": 0.517954409122467, "learning_rate": 7.022776829286348e-06, "loss": 0.0254, "step": 83780 }, { "grad_norm": 0.38093674182891846, "learning_rate": 7.014328940507975e-06, "loss": 0.0231, "step": 83790 }, { "grad_norm": 0.2643943727016449, "learning_rate": 7.005885752577945e-06, "loss": 0.0206, "step": 83800 }, { "grad_norm": 0.5183923244476318, "learning_rate": 6.997447266419615e-06, "loss": 0.023, "step": 83810 }, { "grad_norm": 0.30925604701042175, "learning_rate": 6.989013482955797e-06, "loss": 0.0212, "step": 83820 }, { "grad_norm": 0.8673983216285706, "learning_rate": 6.980584403108808e-06, "loss": 0.0204, "step": 83830 }, { "grad_norm": 0.34171509742736816, "learning_rate": 6.972160027800417e-06, "loss": 0.0217, "step": 83840 }, { "grad_norm": 0.44960659742355347, "learning_rate": 6.963740357951931e-06, "loss": 0.0205, "step": 83850 }, { "grad_norm": 0.6257283091545105, "learning_rate": 6.95532539448408e-06, "loss": 0.0207, "step": 83860 }, { "grad_norm": 0.19423070549964905, "learning_rate": 6.946915138317128e-06, "loss": 0.0184, "step": 83870 }, { "grad_norm": 0.3916184902191162, "learning_rate": 6.9385095903708065e-06, "loss": 0.0176, "step": 83880 }, { "grad_norm": 0.23054540157318115, "learning_rate": 6.930108751564346e-06, "loss": 0.0166, "step": 83890 }, { "grad_norm": 0.9395455718040466, "learning_rate": 6.921712622816428e-06, "loss": 0.0176, "step": 83900 }, { "grad_norm": 0.41377660632133484, "learning_rate": 6.913321205045259e-06, "loss": 0.0232, "step": 83910 }, { "grad_norm": 0.29189157485961914, "learning_rate": 6.904934499168491e-06, "loss": 0.0175, "step": 83920 }, { "grad_norm": 0.24905894696712494, "learning_rate": 6.896552506103299e-06, "loss": 0.0208, "step": 83930 }, { "grad_norm": 0.28451207280158997, "learning_rate": 6.888175226766313e-06, "loss": 0.0205, "step": 83940 }, { "grad_norm": 0.4558318257331848, "learning_rate": 6.879802662073686e-06, "loss": 0.0176, "step": 83950 }, { "grad_norm": 0.3030690848827362, "learning_rate": 6.871434812940991e-06, "loss": 0.0207, "step": 83960 }, { "grad_norm": 0.3017388880252838, "learning_rate": 6.863071680283345e-06, "loss": 0.0201, "step": 83970 }, { "grad_norm": 0.6571839451789856, "learning_rate": 6.854713265015339e-06, "loss": 0.0206, "step": 83980 }, { "grad_norm": 0.4298149049282074, "learning_rate": 6.8463595680510025e-06, "loss": 0.0202, "step": 83990 }, { "grad_norm": 0.21501289308071136, "learning_rate": 6.838010590303911e-06, "loss": 0.0199, "step": 84000 }, { "grad_norm": 0.3129118084907532, "learning_rate": 6.829666332687085e-06, "loss": 0.0188, "step": 84010 }, { "grad_norm": 0.5489510893821716, "learning_rate": 6.821326796113059e-06, "loss": 0.023, "step": 84020 }, { "grad_norm": 0.40580588579177856, "learning_rate": 6.812991981493799e-06, "loss": 0.0189, "step": 84030 }, { "grad_norm": 0.5698217749595642, "learning_rate": 6.804661889740821e-06, "loss": 0.0217, "step": 84040 }, { "grad_norm": 0.31541121006011963, "learning_rate": 6.7963365217650625e-06, "loss": 0.0242, "step": 84050 }, { "grad_norm": 0.5539846420288086, "learning_rate": 6.788015878476983e-06, "loss": 0.0224, "step": 84060 }, { "grad_norm": 0.3593023121356964, "learning_rate": 6.779699960786523e-06, "loss": 0.0185, "step": 84070 }, { "grad_norm": 0.449197381734848, "learning_rate": 6.771388769603104e-06, "loss": 0.0192, "step": 84080 }, { "grad_norm": 0.5321406126022339, "learning_rate": 6.763082305835605e-06, "loss": 0.0215, "step": 84090 }, { "grad_norm": 0.9255896806716919, "learning_rate": 6.754780570392433e-06, "loss": 0.0197, "step": 84100 }, { "grad_norm": 0.3489057719707489, "learning_rate": 6.7464835641814215e-06, "loss": 0.0196, "step": 84110 }, { "grad_norm": 0.4206455647945404, "learning_rate": 6.7381912881099515e-06, "loss": 0.0222, "step": 84120 }, { "grad_norm": 0.2761669456958771, "learning_rate": 6.72990374308482e-06, "loss": 0.024, "step": 84130 }, { "grad_norm": 0.3415602147579193, "learning_rate": 6.721620930012379e-06, "loss": 0.0215, "step": 84140 }, { "grad_norm": 0.30928629636764526, "learning_rate": 6.713342849798393e-06, "loss": 0.0176, "step": 84150 }, { "grad_norm": 0.2603646516799927, "learning_rate": 6.705069503348166e-06, "loss": 0.0183, "step": 84160 }, { "grad_norm": 0.2710809111595154, "learning_rate": 6.696800891566435e-06, "loss": 0.0207, "step": 84170 }, { "grad_norm": 0.2725832164287567, "learning_rate": 6.688537015357449e-06, "loss": 0.0215, "step": 84180 }, { "grad_norm": 0.3271930515766144, "learning_rate": 6.680277875624941e-06, "loss": 0.0203, "step": 84190 }, { "grad_norm": 0.781843364238739, "learning_rate": 6.672023473272121e-06, "loss": 0.0215, "step": 84200 }, { "grad_norm": 0.4470410645008087, "learning_rate": 6.663773809201667e-06, "loss": 0.0222, "step": 84210 }, { "grad_norm": 0.2263040840625763, "learning_rate": 6.655528884315748e-06, "loss": 0.0196, "step": 84220 }, { "grad_norm": 0.888934850692749, "learning_rate": 6.647288699516036e-06, "loss": 0.0199, "step": 84230 }, { "grad_norm": 0.35297343134880066, "learning_rate": 6.639053255703642e-06, "loss": 0.0193, "step": 84240 }, { "grad_norm": 0.429576575756073, "learning_rate": 6.6308225537791925e-06, "loss": 0.0196, "step": 84250 }, { "grad_norm": 1.7675942182540894, "learning_rate": 6.622596594642777e-06, "loss": 0.0224, "step": 84260 }, { "grad_norm": 0.47656264901161194, "learning_rate": 6.614375379193999e-06, "loss": 0.0209, "step": 84270 }, { "grad_norm": 0.3484589755535126, "learning_rate": 6.606158908331889e-06, "loss": 0.0195, "step": 84280 }, { "grad_norm": 0.39351537823677063, "learning_rate": 6.597947182955005e-06, "loss": 0.0225, "step": 84290 }, { "grad_norm": 0.460025817155838, "learning_rate": 6.589740203961359e-06, "loss": 0.0197, "step": 84300 }, { "grad_norm": 0.3222447335720062, "learning_rate": 6.581537972248464e-06, "loss": 0.0222, "step": 84310 }, { "grad_norm": 0.22951331734657288, "learning_rate": 6.573340488713275e-06, "loss": 0.0217, "step": 84320 }, { "grad_norm": 0.24983371794223785, "learning_rate": 6.565147754252304e-06, "loss": 0.0228, "step": 84330 }, { "grad_norm": 0.3404529094696045, "learning_rate": 6.556959769761462e-06, "loss": 0.0214, "step": 84340 }, { "grad_norm": 0.4395776093006134, "learning_rate": 6.548776536136192e-06, "loss": 0.0206, "step": 84350 }, { "grad_norm": 0.5248048901557922, "learning_rate": 6.54059805427138e-06, "loss": 0.021, "step": 84360 }, { "grad_norm": 0.6312980651855469, "learning_rate": 6.532424325061437e-06, "loss": 0.0187, "step": 84370 }, { "grad_norm": 0.45154598355293274, "learning_rate": 6.5242553494001935e-06, "loss": 0.0189, "step": 84380 }, { "grad_norm": 0.33383709192276, "learning_rate": 6.516091128181046e-06, "loss": 0.0202, "step": 84390 }, { "grad_norm": 0.2757430374622345, "learning_rate": 6.507931662296785e-06, "loss": 0.023, "step": 84400 }, { "grad_norm": 0.29579320549964905, "learning_rate": 6.4997769526397354e-06, "loss": 0.0231, "step": 84410 }, { "grad_norm": 0.2155650556087494, "learning_rate": 6.491627000101663e-06, "loss": 0.0214, "step": 84420 }, { "grad_norm": 0.6838479042053223, "learning_rate": 6.4834818055738545e-06, "loss": 0.0201, "step": 84430 }, { "grad_norm": 0.31560391187667847, "learning_rate": 6.4753413699470465e-06, "loss": 0.0221, "step": 84440 }, { "grad_norm": 0.37429285049438477, "learning_rate": 6.467205694111478e-06, "loss": 0.0203, "step": 84450 }, { "grad_norm": 0.5904160141944885, "learning_rate": 6.459074778956836e-06, "loss": 0.0219, "step": 84460 }, { "grad_norm": 0.33177635073661804, "learning_rate": 6.450948625372311e-06, "loss": 0.0208, "step": 84470 }, { "grad_norm": 0.7347350120544434, "learning_rate": 6.442827234246585e-06, "loss": 0.0207, "step": 84480 }, { "grad_norm": 0.6190847158432007, "learning_rate": 6.43471060646777e-06, "loss": 0.0202, "step": 84490 }, { "grad_norm": 0.3413824737071991, "learning_rate": 6.426598742923512e-06, "loss": 0.0196, "step": 84500 }, { "grad_norm": 0.2420840561389923, "learning_rate": 6.418491644500885e-06, "loss": 0.0229, "step": 84510 }, { "grad_norm": 0.4118216037750244, "learning_rate": 6.410389312086512e-06, "loss": 0.0193, "step": 84520 }, { "grad_norm": 0.6601697206497192, "learning_rate": 6.4022917465664135e-06, "loss": 0.0205, "step": 84530 }, { "grad_norm": 0.8727052211761475, "learning_rate": 6.3941989488261575e-06, "loss": 0.0204, "step": 84540 }, { "grad_norm": 0.35089367628097534, "learning_rate": 6.386110919750726e-06, "loss": 0.0188, "step": 84550 }, { "grad_norm": 0.210068941116333, "learning_rate": 6.378027660224651e-06, "loss": 0.0187, "step": 84560 }, { "grad_norm": 0.4250825345516205, "learning_rate": 6.369949171131856e-06, "loss": 0.0173, "step": 84570 }, { "grad_norm": 0.4532061219215393, "learning_rate": 6.361875453355848e-06, "loss": 0.0216, "step": 84580 }, { "grad_norm": 0.20528092980384827, "learning_rate": 6.353806507779525e-06, "loss": 0.0177, "step": 84590 }, { "grad_norm": 0.3997621536254883, "learning_rate": 6.3457423352853044e-06, "loss": 0.0214, "step": 84600 }, { "grad_norm": 0.4082123339176178, "learning_rate": 6.337682936755062e-06, "loss": 0.0177, "step": 84610 }, { "grad_norm": 0.2677406966686249, "learning_rate": 6.329628313070179e-06, "loss": 0.0196, "step": 84620 }, { "grad_norm": 0.549332857131958, "learning_rate": 6.321578465111477e-06, "loss": 0.027, "step": 84630 }, { "grad_norm": 0.2560683786869049, "learning_rate": 6.31353339375928e-06, "loss": 0.019, "step": 84640 }, { "grad_norm": 0.3536185324192047, "learning_rate": 6.305493099893395e-06, "loss": 0.0209, "step": 84650 }, { "grad_norm": 0.35756590962409973, "learning_rate": 6.297457584393096e-06, "loss": 0.0221, "step": 84660 }, { "grad_norm": 0.34867793321609497, "learning_rate": 6.289426848137126e-06, "loss": 0.0235, "step": 84670 }, { "grad_norm": 0.7858589291572571, "learning_rate": 6.281400892003713e-06, "loss": 0.0158, "step": 84680 }, { "grad_norm": 0.5038840174674988, "learning_rate": 6.273379716870581e-06, "loss": 0.0235, "step": 84690 }, { "grad_norm": 0.3141466975212097, "learning_rate": 6.265363323614887e-06, "loss": 0.0214, "step": 84700 }, { "grad_norm": 0.7437950968742371, "learning_rate": 6.257351713113308e-06, "loss": 0.0203, "step": 84710 }, { "grad_norm": 0.3396817743778229, "learning_rate": 6.249344886241981e-06, "loss": 0.0189, "step": 84720 }, { "grad_norm": 0.4580746293067932, "learning_rate": 6.2413428438765295e-06, "loss": 0.0244, "step": 84730 }, { "grad_norm": 0.49709728360176086, "learning_rate": 6.233345586892026e-06, "loss": 0.0201, "step": 84740 }, { "grad_norm": 0.22405947744846344, "learning_rate": 6.225353116163052e-06, "loss": 0.0197, "step": 84750 }, { "grad_norm": 0.36703625321388245, "learning_rate": 6.217365432563627e-06, "loss": 0.0208, "step": 84760 }, { "grad_norm": 0.43546947836875916, "learning_rate": 6.209382536967318e-06, "loss": 0.0194, "step": 84770 }, { "grad_norm": 0.6141558885574341, "learning_rate": 6.201404430247082e-06, "loss": 0.0221, "step": 84780 }, { "grad_norm": 0.2403871864080429, "learning_rate": 6.1934311132754165e-06, "loss": 0.0167, "step": 84790 }, { "grad_norm": 0.31471946835517883, "learning_rate": 6.185462586924257e-06, "loss": 0.0165, "step": 84800 }, { "grad_norm": 0.71335369348526, "learning_rate": 6.177498852065039e-06, "loss": 0.0176, "step": 84810 }, { "grad_norm": 0.5703242421150208, "learning_rate": 6.169539909568655e-06, "loss": 0.0219, "step": 84820 }, { "grad_norm": 1.7408146858215332, "learning_rate": 6.161585760305488e-06, "loss": 0.0192, "step": 84830 }, { "grad_norm": 0.4793788194656372, "learning_rate": 6.153636405145391e-06, "loss": 0.0234, "step": 84840 }, { "grad_norm": 0.23588459193706512, "learning_rate": 6.145691844957702e-06, "loss": 0.0204, "step": 84850 }, { "grad_norm": 0.6652854681015015, "learning_rate": 6.1377520806112145e-06, "loss": 0.0251, "step": 84860 }, { "grad_norm": 0.3008788228034973, "learning_rate": 6.129817112974218e-06, "loss": 0.0168, "step": 84870 }, { "grad_norm": 0.34698009490966797, "learning_rate": 6.121886942914457e-06, "loss": 0.0197, "step": 84880 }, { "grad_norm": 0.8420652151107788, "learning_rate": 6.11396157129917e-06, "loss": 0.0231, "step": 84890 }, { "grad_norm": 0.2875663936138153, "learning_rate": 6.106040998995066e-06, "loss": 0.0174, "step": 84900 }, { "grad_norm": 0.34008049964904785, "learning_rate": 6.098125226868329e-06, "loss": 0.0233, "step": 84910 }, { "grad_norm": 0.6882621049880981, "learning_rate": 6.090214255784604e-06, "loss": 0.0175, "step": 84920 }, { "grad_norm": 0.26378941535949707, "learning_rate": 6.0823080866090346e-06, "loss": 0.0218, "step": 84930 }, { "grad_norm": 1.4087135791778564, "learning_rate": 6.074406720206233e-06, "loss": 0.0207, "step": 84940 }, { "grad_norm": 0.6035940051078796, "learning_rate": 6.066510157440259e-06, "loss": 0.0226, "step": 84950 }, { "grad_norm": 0.30250871181488037, "learning_rate": 6.058618399174676e-06, "loss": 0.0247, "step": 84960 }, { "grad_norm": 0.22164222598075867, "learning_rate": 6.050731446272528e-06, "loss": 0.0172, "step": 84970 }, { "grad_norm": 0.5501391291618347, "learning_rate": 6.042849299596321e-06, "loss": 0.0213, "step": 84980 }, { "grad_norm": 0.35159918665885925, "learning_rate": 6.034971960008013e-06, "loss": 0.0231, "step": 84990 }, { "grad_norm": 0.3418063819408417, "learning_rate": 6.027099428369082e-06, "loss": 0.0195, "step": 85000 }, { "grad_norm": 0.34549951553344727, "learning_rate": 6.019231705540435e-06, "loss": 0.0198, "step": 85010 }, { "grad_norm": 0.23559510707855225, "learning_rate": 6.011368792382488e-06, "loss": 0.0182, "step": 85020 }, { "grad_norm": 1.1326061487197876, "learning_rate": 6.00351068975511e-06, "loss": 0.0237, "step": 85030 }, { "grad_norm": 0.45687416195869446, "learning_rate": 5.9956573985176674e-06, "loss": 0.0228, "step": 85040 }, { "grad_norm": 0.2684800624847412, "learning_rate": 5.987808919528959e-06, "loss": 0.0224, "step": 85050 }, { "grad_norm": 0.4178098440170288, "learning_rate": 5.979965253647307e-06, "loss": 0.0206, "step": 85060 }, { "grad_norm": 0.3002922832965851, "learning_rate": 5.9721264017304655e-06, "loss": 0.0202, "step": 85070 }, { "grad_norm": 0.25476253032684326, "learning_rate": 5.964292364635682e-06, "loss": 0.0222, "step": 85080 }, { "grad_norm": 0.5267492532730103, "learning_rate": 5.956463143219682e-06, "loss": 0.0194, "step": 85090 }, { "grad_norm": 0.8400317430496216, "learning_rate": 5.948638738338663e-06, "loss": 0.0215, "step": 85100 }, { "grad_norm": 0.23274803161621094, "learning_rate": 5.940819150848271e-06, "loss": 0.0162, "step": 85110 }, { "grad_norm": 0.41674578189849854, "learning_rate": 5.933004381603663e-06, "loss": 0.0191, "step": 85120 }, { "grad_norm": 0.2504208981990814, "learning_rate": 5.92519443145943e-06, "loss": 0.0203, "step": 85130 }, { "grad_norm": 0.31837931275367737, "learning_rate": 5.917389301269671e-06, "loss": 0.0193, "step": 85140 }, { "grad_norm": 0.3738299310207367, "learning_rate": 5.9095889918879435e-06, "loss": 0.0201, "step": 85150 }, { "grad_norm": 0.5385423898696899, "learning_rate": 5.901793504167286e-06, "loss": 0.0214, "step": 85160 }, { "grad_norm": 0.4785352349281311, "learning_rate": 5.894002838960172e-06, "loss": 0.0212, "step": 85170 }, { "grad_norm": 0.23024925589561462, "learning_rate": 5.886216997118604e-06, "loss": 0.0236, "step": 85180 }, { "grad_norm": 0.5341675877571106, "learning_rate": 5.878435979494023e-06, "loss": 0.0325, "step": 85190 }, { "grad_norm": 0.3680935800075531, "learning_rate": 5.870659786937344e-06, "loss": 0.0183, "step": 85200 }, { "grad_norm": 0.4333093464374542, "learning_rate": 5.862888420298962e-06, "loss": 0.025, "step": 85210 }, { "grad_norm": 0.23168018460273743, "learning_rate": 5.855121880428743e-06, "loss": 0.0197, "step": 85220 }, { "grad_norm": 0.3106440007686615, "learning_rate": 5.84736016817603e-06, "loss": 0.02, "step": 85230 }, { "grad_norm": 0.5482251048088074, "learning_rate": 5.839603284389622e-06, "loss": 0.0232, "step": 85240 }, { "grad_norm": 0.8462015986442566, "learning_rate": 5.831851229917812e-06, "loss": 0.0218, "step": 85250 }, { "grad_norm": 0.22371919453144073, "learning_rate": 5.824104005608338e-06, "loss": 0.0193, "step": 85260 }, { "grad_norm": 0.3039928078651428, "learning_rate": 5.81636161230843e-06, "loss": 0.0204, "step": 85270 }, { "grad_norm": 0.8402809500694275, "learning_rate": 5.808624050864791e-06, "loss": 0.0252, "step": 85280 }, { "grad_norm": 0.1900862157344818, "learning_rate": 5.800891322123592e-06, "loss": 0.0162, "step": 85290 }, { "grad_norm": 0.8488490581512451, "learning_rate": 5.793163426930454e-06, "loss": 0.0217, "step": 85300 }, { "grad_norm": 0.6608096361160278, "learning_rate": 5.785440366130512e-06, "loss": 0.0167, "step": 85310 }, { "grad_norm": 0.25099095702171326, "learning_rate": 5.777722140568326e-06, "loss": 0.0182, "step": 85320 }, { "grad_norm": 0.1891019344329834, "learning_rate": 5.7700087510879685e-06, "loss": 0.0187, "step": 85330 }, { "grad_norm": 0.7212088704109192, "learning_rate": 5.762300198532933e-06, "loss": 0.023, "step": 85340 }, { "grad_norm": 0.31708472967147827, "learning_rate": 5.7545964837462606e-06, "loss": 0.0205, "step": 85350 }, { "grad_norm": 0.4874199628829956, "learning_rate": 5.746897607570384e-06, "loss": 0.0185, "step": 85360 }, { "grad_norm": 0.29100915789604187, "learning_rate": 5.7392035708472626e-06, "loss": 0.0163, "step": 85370 }, { "grad_norm": 0.3885997235774994, "learning_rate": 5.731514374418278e-06, "loss": 0.0228, "step": 85380 }, { "grad_norm": 0.5866086483001709, "learning_rate": 5.7238300191243325e-06, "loss": 0.0257, "step": 85390 }, { "grad_norm": 0.4076814651489258, "learning_rate": 5.716150505805762e-06, "loss": 0.0219, "step": 85400 }, { "grad_norm": 0.29929399490356445, "learning_rate": 5.708475835302407e-06, "loss": 0.0199, "step": 85410 }, { "grad_norm": 0.38170287013053894, "learning_rate": 5.70080600845353e-06, "loss": 0.021, "step": 85420 }, { "grad_norm": 0.4059300124645233, "learning_rate": 5.693141026097909e-06, "loss": 0.0185, "step": 85430 }, { "grad_norm": 0.3523057699203491, "learning_rate": 5.6854808890737745e-06, "loss": 0.0205, "step": 85440 }, { "grad_norm": 0.2540232837200165, "learning_rate": 5.6778255982188225e-06, "loss": 0.0201, "step": 85450 }, { "grad_norm": 0.6098946332931519, "learning_rate": 5.670175154370217e-06, "loss": 0.0217, "step": 85460 }, { "grad_norm": 0.6765543818473816, "learning_rate": 5.662529558364616e-06, "loss": 0.0219, "step": 85470 }, { "grad_norm": 0.5996211767196655, "learning_rate": 5.654888811038128e-06, "loss": 0.0205, "step": 85480 }, { "grad_norm": 0.38090819120407104, "learning_rate": 5.647252913226314e-06, "loss": 0.0199, "step": 85490 }, { "grad_norm": 0.3889746367931366, "learning_rate": 5.639621865764255e-06, "loss": 0.0213, "step": 85500 }, { "grad_norm": 0.6157989501953125, "learning_rate": 5.63199566948644e-06, "loss": 0.0242, "step": 85510 }, { "grad_norm": 0.2918924391269684, "learning_rate": 5.6243743252268835e-06, "loss": 0.0172, "step": 85520 }, { "grad_norm": 0.35400545597076416, "learning_rate": 5.616757833819009e-06, "loss": 0.0172, "step": 85530 }, { "grad_norm": 0.5597934722900391, "learning_rate": 5.609146196095794e-06, "loss": 0.0169, "step": 85540 }, { "grad_norm": 0.29932478070259094, "learning_rate": 5.6015394128896005e-06, "loss": 0.0245, "step": 85550 }, { "grad_norm": 0.5495296716690063, "learning_rate": 5.593937485032308e-06, "loss": 0.0208, "step": 85560 }, { "grad_norm": 0.3782452344894409, "learning_rate": 5.586340413355246e-06, "loss": 0.0193, "step": 85570 }, { "grad_norm": 0.8481273651123047, "learning_rate": 5.578748198689227e-06, "loss": 0.0261, "step": 85580 }, { "grad_norm": 0.27063557505607605, "learning_rate": 5.571160841864504e-06, "loss": 0.021, "step": 85590 }, { "grad_norm": 0.7835729122161865, "learning_rate": 5.563578343710846e-06, "loss": 0.0213, "step": 85600 }, { "grad_norm": 0.4835982918739319, "learning_rate": 5.556000705057446e-06, "loss": 0.0228, "step": 85610 }, { "grad_norm": 0.20852628350257874, "learning_rate": 5.548427926732996e-06, "loss": 0.0172, "step": 85620 }, { "grad_norm": 1.7881470918655396, "learning_rate": 5.540860009565629e-06, "loss": 0.029, "step": 85630 }, { "grad_norm": 0.7864094376564026, "learning_rate": 5.533296954382966e-06, "loss": 0.0226, "step": 85640 }, { "grad_norm": 0.24896712601184845, "learning_rate": 5.525738762012095e-06, "loss": 0.0197, "step": 85650 }, { "grad_norm": 0.4572392404079437, "learning_rate": 5.518185433279577e-06, "loss": 0.0228, "step": 85660 }, { "grad_norm": 0.2311018407344818, "learning_rate": 5.510636969011407e-06, "loss": 0.0237, "step": 85670 }, { "grad_norm": 0.8365724682807922, "learning_rate": 5.503093370033096e-06, "loss": 0.0172, "step": 85680 }, { "grad_norm": 0.37149035930633545, "learning_rate": 5.495554637169603e-06, "loss": 0.0202, "step": 85690 }, { "grad_norm": 0.3531511425971985, "learning_rate": 5.488020771245328e-06, "loss": 0.0223, "step": 85700 }, { "grad_norm": 0.3569405674934387, "learning_rate": 5.480491773084195e-06, "loss": 0.0192, "step": 85710 }, { "grad_norm": 0.6484642028808594, "learning_rate": 5.472967643509525e-06, "loss": 0.023, "step": 85720 }, { "grad_norm": 0.2374563217163086, "learning_rate": 5.46544838334419e-06, "loss": 0.0171, "step": 85730 }, { "grad_norm": 0.4004645347595215, "learning_rate": 5.4579339934104525e-06, "loss": 0.0239, "step": 85740 }, { "grad_norm": 0.23390914499759674, "learning_rate": 5.450424474530091e-06, "loss": 0.0186, "step": 85750 }, { "grad_norm": 0.30703261494636536, "learning_rate": 5.442919827524323e-06, "loss": 0.0264, "step": 85760 }, { "grad_norm": 0.18532420694828033, "learning_rate": 5.435420053213863e-06, "loss": 0.0182, "step": 85770 }, { "grad_norm": 0.4143240749835968, "learning_rate": 5.427925152418845e-06, "loss": 0.0215, "step": 85780 }, { "grad_norm": 0.2362920194864273, "learning_rate": 5.420435125958934e-06, "loss": 0.0202, "step": 85790 }, { "grad_norm": 0.3431181311607361, "learning_rate": 5.412949974653209e-06, "loss": 0.0195, "step": 85800 }, { "grad_norm": 0.3568473160266876, "learning_rate": 5.405469699320248e-06, "loss": 0.0227, "step": 85810 }, { "grad_norm": 0.36133041977882385, "learning_rate": 5.397994300778064e-06, "loss": 0.0167, "step": 85820 }, { "grad_norm": 0.3896494209766388, "learning_rate": 5.39052377984417e-06, "loss": 0.0197, "step": 85830 }, { "grad_norm": 0.3740622103214264, "learning_rate": 5.383058137335523e-06, "loss": 0.0176, "step": 85840 }, { "grad_norm": 0.35458558797836304, "learning_rate": 5.375597374068553e-06, "loss": 0.0174, "step": 85850 }, { "grad_norm": 0.7153537273406982, "learning_rate": 5.368141490859163e-06, "loss": 0.0244, "step": 85860 }, { "grad_norm": 0.268120676279068, "learning_rate": 5.360690488522729e-06, "loss": 0.0218, "step": 85870 }, { "grad_norm": 0.4872739017009735, "learning_rate": 5.353244367874055e-06, "loss": 0.0211, "step": 85880 }, { "grad_norm": 0.3635665774345398, "learning_rate": 5.345803129727456e-06, "loss": 0.018, "step": 85890 }, { "grad_norm": 0.614302396774292, "learning_rate": 5.338366774896697e-06, "loss": 0.0179, "step": 85900 }, { "grad_norm": 0.3985607326030731, "learning_rate": 5.330935304194995e-06, "loss": 0.0244, "step": 85910 }, { "grad_norm": 0.7759233117103577, "learning_rate": 5.323508718435044e-06, "loss": 0.0198, "step": 85920 }, { "grad_norm": 0.5086357593536377, "learning_rate": 5.316087018429011e-06, "loss": 0.0214, "step": 85930 }, { "grad_norm": 0.2730962336063385, "learning_rate": 5.30867020498853e-06, "loss": 0.0226, "step": 85940 }, { "grad_norm": 0.37661629915237427, "learning_rate": 5.30125827892467e-06, "loss": 0.0218, "step": 85950 }, { "grad_norm": 0.27265921235084534, "learning_rate": 5.293851241048015e-06, "loss": 0.0229, "step": 85960 }, { "grad_norm": 0.5468228459358215, "learning_rate": 5.286449092168555e-06, "loss": 0.0195, "step": 85970 }, { "grad_norm": 0.46222853660583496, "learning_rate": 5.2790518330958105e-06, "loss": 0.0181, "step": 85980 }, { "grad_norm": 0.2848261594772339, "learning_rate": 5.271659464638717e-06, "loss": 0.0204, "step": 85990 }, { "grad_norm": 0.24690106511116028, "learning_rate": 5.264271987605701e-06, "loss": 0.0207, "step": 86000 }, { "grad_norm": 0.3425275683403015, "learning_rate": 5.256889402804632e-06, "loss": 0.0203, "step": 86010 }, { "grad_norm": 0.48155248165130615, "learning_rate": 5.249511711042881e-06, "loss": 0.0177, "step": 86020 }, { "grad_norm": 0.20535150170326233, "learning_rate": 5.242138913127237e-06, "loss": 0.0182, "step": 86030 }, { "grad_norm": 0.604171097278595, "learning_rate": 5.234771009863987e-06, "loss": 0.0196, "step": 86040 }, { "grad_norm": 0.48067617416381836, "learning_rate": 5.227408002058881e-06, "loss": 0.0202, "step": 86050 }, { "grad_norm": 0.31511420011520386, "learning_rate": 5.220049890517126e-06, "loss": 0.0178, "step": 86060 }, { "grad_norm": 0.29612997174263, "learning_rate": 5.212696676043377e-06, "loss": 0.0167, "step": 86070 }, { "grad_norm": 0.2680895924568176, "learning_rate": 5.205348359441798e-06, "loss": 0.0189, "step": 86080 }, { "grad_norm": 0.3452719449996948, "learning_rate": 5.198004941515966e-06, "loss": 0.0218, "step": 86090 }, { "grad_norm": 0.26759639382362366, "learning_rate": 5.190666423068952e-06, "loss": 0.0225, "step": 86100 }, { "grad_norm": 0.4078972339630127, "learning_rate": 5.183332804903284e-06, "loss": 0.0187, "step": 86110 }, { "grad_norm": 0.4085598886013031, "learning_rate": 5.176004087820979e-06, "loss": 0.0191, "step": 86120 }, { "grad_norm": 0.37727639079093933, "learning_rate": 5.168680272623455e-06, "loss": 0.022, "step": 86130 }, { "grad_norm": 0.2855319082736969, "learning_rate": 5.161361360111661e-06, "loss": 0.0222, "step": 86140 }, { "grad_norm": 0.3094460964202881, "learning_rate": 5.154047351085984e-06, "loss": 0.0187, "step": 86150 }, { "grad_norm": 0.32180100679397583, "learning_rate": 5.1467382463462385e-06, "loss": 0.0173, "step": 86160 }, { "grad_norm": 0.28815892338752747, "learning_rate": 5.139434046691788e-06, "loss": 0.0224, "step": 86170 }, { "grad_norm": 0.4683496654033661, "learning_rate": 5.13213475292137e-06, "loss": 0.0165, "step": 86180 }, { "grad_norm": 0.3819679617881775, "learning_rate": 5.12484036583325e-06, "loss": 0.0175, "step": 86190 }, { "grad_norm": 0.5251056551933289, "learning_rate": 5.117550886225109e-06, "loss": 0.0185, "step": 86200 }, { "grad_norm": 0.31451115012168884, "learning_rate": 5.11026631489413e-06, "loss": 0.0196, "step": 86210 }, { "grad_norm": 0.9490247368812561, "learning_rate": 5.102986652636926e-06, "loss": 0.0182, "step": 86220 }, { "grad_norm": 0.39115869998931885, "learning_rate": 5.095711900249606e-06, "loss": 0.021, "step": 86230 }, { "grad_norm": 0.3001561164855957, "learning_rate": 5.088442058527715e-06, "loss": 0.0197, "step": 86240 }, { "grad_norm": 0.30711597204208374, "learning_rate": 5.08117712826629e-06, "loss": 0.0204, "step": 86250 }, { "grad_norm": 0.3748950958251953, "learning_rate": 5.073917110259785e-06, "loss": 0.0176, "step": 86260 }, { "grad_norm": 0.3718004822731018, "learning_rate": 5.066662005302175e-06, "loss": 0.0188, "step": 86270 }, { "grad_norm": 0.41015714406967163, "learning_rate": 5.059411814186838e-06, "loss": 0.0161, "step": 86280 }, { "grad_norm": 0.28475937247276306, "learning_rate": 5.05216653770666e-06, "loss": 0.0203, "step": 86290 }, { "grad_norm": 0.2718280553817749, "learning_rate": 5.044926176653969e-06, "loss": 0.0181, "step": 86300 }, { "grad_norm": 0.31257688999176025, "learning_rate": 5.037690731820571e-06, "loss": 0.0166, "step": 86310 }, { "grad_norm": 0.3540048599243164, "learning_rate": 5.030460203997711e-06, "loss": 0.0198, "step": 86320 }, { "grad_norm": 1.2247023582458496, "learning_rate": 5.0232345939761174e-06, "loss": 0.0226, "step": 86330 }, { "grad_norm": 0.5467809438705444, "learning_rate": 5.016013902545957e-06, "loss": 0.0247, "step": 86340 }, { "grad_norm": 0.4425293207168579, "learning_rate": 5.008798130496889e-06, "loss": 0.0216, "step": 86350 }, { "grad_norm": 0.289032518863678, "learning_rate": 5.001587278618008e-06, "loss": 0.0195, "step": 86360 }, { "grad_norm": 0.3101847469806671, "learning_rate": 4.994381347697891e-06, "loss": 0.019, "step": 86370 }, { "grad_norm": 0.49031010270118713, "learning_rate": 4.987180338524572e-06, "loss": 0.0216, "step": 86380 }, { "grad_norm": 0.31013211607933044, "learning_rate": 4.9799842518855275e-06, "loss": 0.0241, "step": 86390 }, { "grad_norm": 0.3279802203178406, "learning_rate": 4.9727930885677285e-06, "loss": 0.0175, "step": 86400 }, { "grad_norm": 0.48283737897872925, "learning_rate": 4.965606849357573e-06, "loss": 0.0183, "step": 86410 }, { "grad_norm": 0.3703644573688507, "learning_rate": 4.9584255350409445e-06, "loss": 0.0206, "step": 86420 }, { "grad_norm": 0.24299347400665283, "learning_rate": 4.951249146403175e-06, "loss": 0.0184, "step": 86430 }, { "grad_norm": 0.6621761322021484, "learning_rate": 4.9440776842290815e-06, "loss": 0.0319, "step": 86440 }, { "grad_norm": 0.2963361442089081, "learning_rate": 4.936911149302903e-06, "loss": 0.0203, "step": 86450 }, { "grad_norm": 0.8285318613052368, "learning_rate": 4.929749542408379e-06, "loss": 0.0276, "step": 86460 }, { "grad_norm": 0.48809710144996643, "learning_rate": 4.922592864328673e-06, "loss": 0.0185, "step": 86470 }, { "grad_norm": 0.23653627932071686, "learning_rate": 4.915441115846436e-06, "loss": 0.0236, "step": 86480 }, { "grad_norm": 0.5108718872070312, "learning_rate": 4.908294297743782e-06, "loss": 0.0202, "step": 86490 }, { "grad_norm": 0.52701336145401, "learning_rate": 4.901152410802273e-06, "loss": 0.0227, "step": 86500 }, { "grad_norm": 0.37164515256881714, "learning_rate": 4.894015455802919e-06, "loss": 0.0202, "step": 86510 }, { "grad_norm": 0.2253437489271164, "learning_rate": 4.8868834335262346e-06, "loss": 0.0219, "step": 86520 }, { "grad_norm": 0.2164200246334076, "learning_rate": 4.87975634475214e-06, "loss": 0.0208, "step": 86530 }, { "grad_norm": 0.644864022731781, "learning_rate": 4.8726341902600614e-06, "loss": 0.0219, "step": 86540 }, { "grad_norm": 0.3292497992515564, "learning_rate": 4.865516970828843e-06, "loss": 0.0187, "step": 86550 }, { "grad_norm": 0.3410935401916504, "learning_rate": 4.85840468723685e-06, "loss": 0.0187, "step": 86560 }, { "grad_norm": 0.2776276469230652, "learning_rate": 4.85129734026184e-06, "loss": 0.0224, "step": 86570 }, { "grad_norm": 0.4366376996040344, "learning_rate": 4.844194930681073e-06, "loss": 0.0193, "step": 86580 }, { "grad_norm": 0.4103776514530182, "learning_rate": 4.837097459271267e-06, "loss": 0.0201, "step": 86590 }, { "grad_norm": 0.39126867055892944, "learning_rate": 4.830004926808573e-06, "loss": 0.019, "step": 86600 }, { "grad_norm": 0.4594947099685669, "learning_rate": 4.822917334068627e-06, "loss": 0.0193, "step": 86610 }, { "grad_norm": 0.30851784348487854, "learning_rate": 4.81583468182652e-06, "loss": 0.0224, "step": 86620 }, { "grad_norm": 0.767821192741394, "learning_rate": 4.808756970856809e-06, "loss": 0.0163, "step": 86630 }, { "grad_norm": 1.1220101118087769, "learning_rate": 4.801684201933476e-06, "loss": 0.0212, "step": 86640 }, { "grad_norm": 0.5674508213996887, "learning_rate": 4.794616375830019e-06, "loss": 0.0265, "step": 86650 }, { "grad_norm": 0.4004090130329132, "learning_rate": 4.787553493319341e-06, "loss": 0.0218, "step": 86660 }, { "grad_norm": 0.2917740046977997, "learning_rate": 4.780495555173836e-06, "loss": 0.0181, "step": 86670 }, { "grad_norm": 0.2574410140514374, "learning_rate": 4.7734425621653475e-06, "loss": 0.0221, "step": 86680 }, { "grad_norm": 0.5562705397605896, "learning_rate": 4.766394515065198e-06, "loss": 0.0196, "step": 86690 }, { "grad_norm": 0.5353840589523315, "learning_rate": 4.75935141464412e-06, "loss": 0.0234, "step": 86700 }, { "grad_norm": 0.33486393094062805, "learning_rate": 4.75231326167237e-06, "loss": 0.0167, "step": 86710 }, { "grad_norm": 0.211909681558609, "learning_rate": 4.745280056919599e-06, "loss": 0.0182, "step": 86720 }, { "grad_norm": 0.32526373863220215, "learning_rate": 4.738251801154975e-06, "loss": 0.0249, "step": 86730 }, { "grad_norm": 0.3263991177082062, "learning_rate": 4.731228495147061e-06, "loss": 0.0187, "step": 86740 }, { "grad_norm": 0.34105122089385986, "learning_rate": 4.7242101396639595e-06, "loss": 0.02, "step": 86750 }, { "grad_norm": 0.22550012171268463, "learning_rate": 4.717196735473151e-06, "loss": 0.0181, "step": 86760 }, { "grad_norm": 0.6743561625480652, "learning_rate": 4.710188283341643e-06, "loss": 0.0223, "step": 86770 }, { "grad_norm": 2.3551018238067627, "learning_rate": 4.703184784035835e-06, "loss": 0.016, "step": 86780 }, { "grad_norm": 1.4008909463882446, "learning_rate": 4.696186238321642e-06, "loss": 0.022, "step": 86790 }, { "grad_norm": 0.3207634687423706, "learning_rate": 4.689192646964408e-06, "loss": 0.0205, "step": 86800 }, { "grad_norm": 0.5019890069961548, "learning_rate": 4.682204010728952e-06, "loss": 0.0205, "step": 86810 }, { "grad_norm": 0.3224261999130249, "learning_rate": 4.67522033037952e-06, "loss": 0.0223, "step": 86820 }, { "grad_norm": 0.23366956412792206, "learning_rate": 4.668241606679846e-06, "loss": 0.0212, "step": 86830 }, { "grad_norm": 0.38011088967323303, "learning_rate": 4.66126784039313e-06, "loss": 0.0241, "step": 86840 }, { "grad_norm": 0.2879718244075775, "learning_rate": 4.6542990322819836e-06, "loss": 0.0262, "step": 86850 }, { "grad_norm": 0.254200279712677, "learning_rate": 4.64733518310852e-06, "loss": 0.0185, "step": 86860 }, { "grad_norm": 0.32544419169425964, "learning_rate": 4.640376293634291e-06, "loss": 0.0194, "step": 86870 }, { "grad_norm": 0.6275415420532227, "learning_rate": 4.633422364620327e-06, "loss": 0.0238, "step": 86880 }, { "grad_norm": 0.4718250632286072, "learning_rate": 4.6264733968270745e-06, "loss": 0.0213, "step": 86890 }, { "grad_norm": 1.4857641458511353, "learning_rate": 4.619529391014482e-06, "loss": 0.0207, "step": 86900 }, { "grad_norm": 0.5595002174377441, "learning_rate": 4.61259034794192e-06, "loss": 0.0164, "step": 86910 }, { "grad_norm": 0.33003076910972595, "learning_rate": 4.605656268368247e-06, "loss": 0.0214, "step": 86920 }, { "grad_norm": 0.9057531356811523, "learning_rate": 4.598727153051735e-06, "loss": 0.0187, "step": 86930 }, { "grad_norm": 0.39532265067100525, "learning_rate": 4.591803002750178e-06, "loss": 0.0168, "step": 86940 }, { "grad_norm": 0.5307992100715637, "learning_rate": 4.58488381822077e-06, "loss": 0.0197, "step": 86950 }, { "grad_norm": 0.17300936579704285, "learning_rate": 4.5779696002201946e-06, "loss": 0.0203, "step": 86960 }, { "grad_norm": 0.41292738914489746, "learning_rate": 4.571060349504563e-06, "loss": 0.0211, "step": 86970 }, { "grad_norm": 0.3391699194908142, "learning_rate": 4.564156066829478e-06, "loss": 0.0203, "step": 86980 }, { "grad_norm": 0.6379764676094055, "learning_rate": 4.557256752949957e-06, "loss": 0.0196, "step": 86990 }, { "grad_norm": 0.5864599347114563, "learning_rate": 4.550362408620534e-06, "loss": 0.02, "step": 87000 }, { "grad_norm": 0.6987375020980835, "learning_rate": 4.543473034595136e-06, "loss": 0.0227, "step": 87010 }, { "grad_norm": 0.34481197595596313, "learning_rate": 4.536588631627192e-06, "loss": 0.0185, "step": 87020 }, { "grad_norm": 0.6753196120262146, "learning_rate": 4.5297092004695565e-06, "loss": 0.0212, "step": 87030 }, { "grad_norm": 0.21571993827819824, "learning_rate": 4.522834741874555e-06, "loss": 0.0203, "step": 87040 }, { "grad_norm": 0.43144020438194275, "learning_rate": 4.515965256593974e-06, "loss": 0.0232, "step": 87050 }, { "grad_norm": 0.40045976638793945, "learning_rate": 4.509100745379058e-06, "loss": 0.0238, "step": 87060 }, { "grad_norm": 0.36097511649131775, "learning_rate": 4.502241208980479e-06, "loss": 0.0209, "step": 87070 }, { "grad_norm": 0.5509045124053955, "learning_rate": 4.495386648148398e-06, "loss": 0.0219, "step": 87080 }, { "grad_norm": 0.3136180341243744, "learning_rate": 4.488537063632431e-06, "loss": 0.0209, "step": 87090 }, { "grad_norm": 0.7662746906280518, "learning_rate": 4.481692456181608e-06, "loss": 0.0176, "step": 87100 }, { "grad_norm": 0.8998585343360901, "learning_rate": 4.474852826544479e-06, "loss": 0.0185, "step": 87110 }, { "grad_norm": 0.5386587977409363, "learning_rate": 4.468018175468974e-06, "loss": 0.021, "step": 87120 }, { "grad_norm": 1.6069341897964478, "learning_rate": 4.461188503702568e-06, "loss": 0.0204, "step": 87130 }, { "grad_norm": 1.1143122911453247, "learning_rate": 4.454363811992107e-06, "loss": 0.0182, "step": 87140 }, { "grad_norm": 0.33164384961128235, "learning_rate": 4.447544101083956e-06, "loss": 0.0176, "step": 87150 }, { "grad_norm": 0.4473397731781006, "learning_rate": 4.440729371723884e-06, "loss": 0.0193, "step": 87160 }, { "grad_norm": 0.5583550333976746, "learning_rate": 4.433919624657162e-06, "loss": 0.0181, "step": 87170 }, { "grad_norm": 0.3477427363395691, "learning_rate": 4.427114860628462e-06, "loss": 0.0183, "step": 87180 }, { "grad_norm": 0.5817933082580566, "learning_rate": 4.420315080381982e-06, "loss": 0.0238, "step": 87190 }, { "grad_norm": 0.7050455212593079, "learning_rate": 4.4135202846613045e-06, "loss": 0.0251, "step": 87200 }, { "grad_norm": 0.32410135865211487, "learning_rate": 4.406730474209525e-06, "loss": 0.0228, "step": 87210 }, { "grad_norm": 0.32408440113067627, "learning_rate": 4.399945649769138e-06, "loss": 0.0178, "step": 87220 }, { "grad_norm": 0.40513625741004944, "learning_rate": 4.393165812082151e-06, "loss": 0.0183, "step": 87230 }, { "grad_norm": 0.3174936771392822, "learning_rate": 4.386390961889969e-06, "loss": 0.017, "step": 87240 }, { "grad_norm": 0.7754033207893372, "learning_rate": 4.3796210999334895e-06, "loss": 0.0182, "step": 87250 }, { "grad_norm": 0.4812167286872864, "learning_rate": 4.372856226953065e-06, "loss": 0.0154, "step": 87260 }, { "grad_norm": 0.44829103350639343, "learning_rate": 4.366096343688486e-06, "loss": 0.022, "step": 87270 }, { "grad_norm": 0.27306514978408813, "learning_rate": 4.359341450878995e-06, "loss": 0.0223, "step": 87280 }, { "grad_norm": 0.3762519955635071, "learning_rate": 4.352591549263302e-06, "loss": 0.0209, "step": 87290 }, { "grad_norm": 0.358561247587204, "learning_rate": 4.345846639579582e-06, "loss": 0.0198, "step": 87300 }, { "grad_norm": 0.32068559527397156, "learning_rate": 4.339106722565417e-06, "loss": 0.0202, "step": 87310 }, { "grad_norm": 0.32397720217704773, "learning_rate": 4.332371798957896e-06, "loss": 0.0232, "step": 87320 }, { "grad_norm": 0.374939501285553, "learning_rate": 4.3256418694935295e-06, "loss": 0.0198, "step": 87330 }, { "grad_norm": 0.38013094663619995, "learning_rate": 4.3189169349083125e-06, "loss": 0.022, "step": 87340 }, { "grad_norm": 0.3454555571079254, "learning_rate": 4.312196995937645e-06, "loss": 0.0211, "step": 87350 }, { "grad_norm": 0.4498080313205719, "learning_rate": 4.305482053316434e-06, "loss": 0.0222, "step": 87360 }, { "grad_norm": 0.24935805797576904, "learning_rate": 4.2987721077789925e-06, "loss": 0.0212, "step": 87370 }, { "grad_norm": 0.5462095141410828, "learning_rate": 4.2920671600591325e-06, "loss": 0.0213, "step": 87380 }, { "grad_norm": 1.2120875120162964, "learning_rate": 4.285367210890084e-06, "loss": 0.0194, "step": 87390 }, { "grad_norm": 0.7239019870758057, "learning_rate": 4.278672261004551e-06, "loss": 0.0232, "step": 87400 }, { "grad_norm": 0.22885026037693024, "learning_rate": 4.2719823111346704e-06, "loss": 0.0208, "step": 87410 }, { "grad_norm": 0.48538434505462646, "learning_rate": 4.265297362012061e-06, "loss": 0.0231, "step": 87420 }, { "grad_norm": 0.4728192090988159, "learning_rate": 4.258617414367766e-06, "loss": 0.0223, "step": 87430 }, { "grad_norm": 0.24494782090187073, "learning_rate": 4.251942468932296e-06, "loss": 0.0213, "step": 87440 }, { "grad_norm": 1.114209532737732, "learning_rate": 4.2452725264356175e-06, "loss": 0.0234, "step": 87450 }, { "grad_norm": 0.38058069348335266, "learning_rate": 4.2386075876071575e-06, "loss": 0.0202, "step": 87460 }, { "grad_norm": 0.7937028408050537, "learning_rate": 4.231947653175755e-06, "loss": 0.0185, "step": 87470 }, { "grad_norm": 0.5233355760574341, "learning_rate": 4.225292723869761e-06, "loss": 0.0212, "step": 87480 }, { "grad_norm": 0.2522035539150238, "learning_rate": 4.218642800416916e-06, "loss": 0.0193, "step": 87490 }, { "grad_norm": 0.27003777027130127, "learning_rate": 4.211997883544472e-06, "loss": 0.0187, "step": 87500 }, { "grad_norm": 1.026662826538086, "learning_rate": 4.205357973979096e-06, "loss": 0.0204, "step": 87510 }, { "grad_norm": 0.45681676268577576, "learning_rate": 4.198723072446925e-06, "loss": 0.0194, "step": 87520 }, { "grad_norm": 0.2629525065422058, "learning_rate": 4.192093179673534e-06, "loss": 0.0244, "step": 87530 }, { "grad_norm": 0.25965020060539246, "learning_rate": 4.185468296383959e-06, "loss": 0.0165, "step": 87540 }, { "grad_norm": 0.3332175612449646, "learning_rate": 4.1788484233027005e-06, "loss": 0.0209, "step": 87550 }, { "grad_norm": 0.20679661631584167, "learning_rate": 4.172233561153677e-06, "loss": 0.0201, "step": 87560 }, { "grad_norm": 0.46346792578697205, "learning_rate": 4.165623710660288e-06, "loss": 0.0212, "step": 87570 }, { "grad_norm": 0.2690589129924774, "learning_rate": 4.1590188725453835e-06, "loss": 0.0179, "step": 87580 }, { "grad_norm": 0.6284784078598022, "learning_rate": 4.152419047531259e-06, "loss": 0.0226, "step": 87590 }, { "grad_norm": 0.24331459403038025, "learning_rate": 4.145824236339651e-06, "loss": 0.0212, "step": 87600 }, { "grad_norm": 0.8834213018417358, "learning_rate": 4.1392344396917695e-06, "loss": 0.0196, "step": 87610 }, { "grad_norm": 0.6545521020889282, "learning_rate": 4.132649658308252e-06, "loss": 0.0182, "step": 87620 }, { "grad_norm": 0.3661327362060547, "learning_rate": 4.126069892909213e-06, "loss": 0.0202, "step": 87630 }, { "grad_norm": 0.39305543899536133, "learning_rate": 4.119495144214191e-06, "loss": 0.0216, "step": 87640 }, { "grad_norm": 0.6461391448974609, "learning_rate": 4.112925412942214e-06, "loss": 0.0199, "step": 87650 }, { "grad_norm": 0.7832736968994141, "learning_rate": 4.106360699811712e-06, "loss": 0.0241, "step": 87660 }, { "grad_norm": 0.29486680030822754, "learning_rate": 4.099801005540616e-06, "loss": 0.0192, "step": 87670 }, { "grad_norm": 0.49165284633636475, "learning_rate": 4.093246330846256e-06, "loss": 0.0206, "step": 87680 }, { "grad_norm": 0.3872055411338806, "learning_rate": 4.086696676445467e-06, "loss": 0.0214, "step": 87690 }, { "grad_norm": 0.21572358906269073, "learning_rate": 4.0801520430544914e-06, "loss": 0.0179, "step": 87700 }, { "grad_norm": 1.8779116868972778, "learning_rate": 4.073612431389062e-06, "loss": 0.0225, "step": 87710 }, { "grad_norm": 0.37275514006614685, "learning_rate": 4.067077842164319e-06, "loss": 0.0212, "step": 87720 }, { "grad_norm": 0.2936387360095978, "learning_rate": 4.0605482760948945e-06, "loss": 0.019, "step": 87730 }, { "grad_norm": 0.7868932485580444, "learning_rate": 4.054023733894829e-06, "loss": 0.0165, "step": 87740 }, { "grad_norm": 0.3773958683013916, "learning_rate": 4.047504216277653e-06, "loss": 0.0223, "step": 87750 }, { "grad_norm": 0.25618019700050354, "learning_rate": 4.040989723956323e-06, "loss": 0.0214, "step": 87760 }, { "grad_norm": 0.3278043270111084, "learning_rate": 4.03448025764327e-06, "loss": 0.0211, "step": 87770 }, { "grad_norm": 0.21118305623531342, "learning_rate": 4.027975818050339e-06, "loss": 0.0148, "step": 87780 }, { "grad_norm": 0.6037051677703857, "learning_rate": 4.021476405888858e-06, "loss": 0.0196, "step": 87790 }, { "grad_norm": 0.40262144804000854, "learning_rate": 4.01498202186959e-06, "loss": 0.0191, "step": 87800 }, { "grad_norm": 0.3298685550689697, "learning_rate": 4.008492666702751e-06, "loss": 0.0231, "step": 87810 }, { "grad_norm": 0.36734750866889954, "learning_rate": 4.002008341098001e-06, "loss": 0.0183, "step": 87820 }, { "grad_norm": 0.3057738244533539, "learning_rate": 3.995529045764462e-06, "loss": 0.0196, "step": 87830 }, { "grad_norm": 0.3820458650588989, "learning_rate": 3.989054781410712e-06, "loss": 0.0182, "step": 87840 }, { "grad_norm": 0.4612034857273102, "learning_rate": 3.982585548744744e-06, "loss": 0.018, "step": 87850 }, { "grad_norm": 0.542019248008728, "learning_rate": 3.9761213484740435e-06, "loss": 0.0164, "step": 87860 }, { "grad_norm": 0.510747492313385, "learning_rate": 3.9696621813055055e-06, "loss": 0.0227, "step": 87870 }, { "grad_norm": 0.35383445024490356, "learning_rate": 3.9632080479455044e-06, "loss": 0.0197, "step": 87880 }, { "grad_norm": 0.28906145691871643, "learning_rate": 3.956758949099859e-06, "loss": 0.0206, "step": 87890 }, { "grad_norm": 0.26307976245880127, "learning_rate": 3.9503148854738385e-06, "loss": 0.0234, "step": 87900 }, { "grad_norm": 0.4649456739425659, "learning_rate": 3.943875857772139e-06, "loss": 0.0222, "step": 87910 }, { "grad_norm": 0.3103873133659363, "learning_rate": 3.9374418666989375e-06, "loss": 0.0189, "step": 87920 }, { "grad_norm": 0.5277453660964966, "learning_rate": 3.931012912957838e-06, "loss": 0.0189, "step": 87930 }, { "grad_norm": 0.22756944596767426, "learning_rate": 3.9245889972519045e-06, "loss": 0.0169, "step": 87940 }, { "grad_norm": 0.21630650758743286, "learning_rate": 3.9181701202836265e-06, "loss": 0.0195, "step": 87950 }, { "grad_norm": 0.7783811688423157, "learning_rate": 3.911756282755003e-06, "loss": 0.0209, "step": 87960 }, { "grad_norm": 0.699855387210846, "learning_rate": 3.905347485367411e-06, "loss": 0.0208, "step": 87970 }, { "grad_norm": 0.21431250870227814, "learning_rate": 3.898943728821725e-06, "loss": 0.0211, "step": 87980 }, { "grad_norm": 0.6476022601127625, "learning_rate": 3.892545013818227e-06, "loss": 0.0168, "step": 87990 }, { "grad_norm": 0.5693400502204895, "learning_rate": 3.886151341056687e-06, "loss": 0.022, "step": 88000 }, { "grad_norm": 0.5565578937530518, "learning_rate": 3.8797627112363066e-06, "loss": 0.0199, "step": 88010 }, { "grad_norm": 0.5598085522651672, "learning_rate": 3.873379125055743e-06, "loss": 0.0203, "step": 88020 }, { "grad_norm": 0.6564863324165344, "learning_rate": 3.867000583213082e-06, "loss": 0.0214, "step": 88030 }, { "grad_norm": 0.5268188118934631, "learning_rate": 3.860627086405871e-06, "loss": 0.0215, "step": 88040 }, { "grad_norm": 0.25600093603134155, "learning_rate": 3.854258635331126e-06, "loss": 0.0165, "step": 88050 }, { "grad_norm": 0.33886659145355225, "learning_rate": 3.8478952306852665e-06, "loss": 0.0184, "step": 88060 }, { "grad_norm": 0.5306302905082703, "learning_rate": 3.841536873164198e-06, "loss": 0.0173, "step": 88070 }, { "grad_norm": 0.5080151557922363, "learning_rate": 3.8351835634632626e-06, "loss": 0.019, "step": 88080 }, { "grad_norm": 0.7748992443084717, "learning_rate": 3.828835302277245e-06, "loss": 0.0157, "step": 88090 }, { "grad_norm": 3.9794199466705322, "learning_rate": 3.822492090300378e-06, "loss": 0.0179, "step": 88100 }, { "grad_norm": 0.4126427173614502, "learning_rate": 3.8161539282263566e-06, "loss": 0.0217, "step": 88110 }, { "grad_norm": 0.2590842843055725, "learning_rate": 3.8098208167482995e-06, "loss": 0.02, "step": 88120 }, { "grad_norm": 0.6492237448692322, "learning_rate": 3.803492756558796e-06, "loss": 0.0185, "step": 88130 }, { "grad_norm": 0.4087149202823639, "learning_rate": 3.797169748349855e-06, "loss": 0.023, "step": 88140 }, { "grad_norm": 0.7049340009689331, "learning_rate": 3.790851792812977e-06, "loss": 0.0198, "step": 88150 }, { "grad_norm": 0.5637466907501221, "learning_rate": 3.784538890639072e-06, "loss": 0.0192, "step": 88160 }, { "grad_norm": 0.4946177303791046, "learning_rate": 3.7782310425185153e-06, "loss": 0.0182, "step": 88170 }, { "grad_norm": 0.258441686630249, "learning_rate": 3.77192824914111e-06, "loss": 0.0172, "step": 88180 }, { "grad_norm": 0.3023606240749359, "learning_rate": 3.7656305111961333e-06, "loss": 0.0196, "step": 88190 }, { "grad_norm": 1.4129962921142578, "learning_rate": 3.7593378293722782e-06, "loss": 0.0204, "step": 88200 }, { "grad_norm": 0.5337965488433838, "learning_rate": 3.7530502043577333e-06, "loss": 0.0209, "step": 88210 }, { "grad_norm": 2.926725149154663, "learning_rate": 3.746767636840076e-06, "loss": 0.0215, "step": 88220 }, { "grad_norm": 0.8630157113075256, "learning_rate": 3.74049012750638e-06, "loss": 0.0215, "step": 88230 }, { "grad_norm": 1.3549045324325562, "learning_rate": 3.734217677043128e-06, "loss": 0.0201, "step": 88240 }, { "grad_norm": 0.46499842405319214, "learning_rate": 3.727950286136267e-06, "loss": 0.0195, "step": 88250 }, { "grad_norm": 0.5142017602920532, "learning_rate": 3.7216879554711925e-06, "loss": 0.0188, "step": 88260 }, { "grad_norm": 0.21385294198989868, "learning_rate": 3.715430685732757e-06, "loss": 0.0176, "step": 88270 }, { "grad_norm": 0.2717320919036865, "learning_rate": 3.7091784776052243e-06, "loss": 0.0185, "step": 88280 }, { "grad_norm": 0.3209323287010193, "learning_rate": 3.7029313317723356e-06, "loss": 0.0183, "step": 88290 }, { "grad_norm": 0.35725510120391846, "learning_rate": 3.6966892489172787e-06, "loss": 0.0195, "step": 88300 }, { "grad_norm": 0.3680044114589691, "learning_rate": 3.6904522297226575e-06, "loss": 0.0162, "step": 88310 }, { "grad_norm": 2.28901743888855, "learning_rate": 3.6842202748705647e-06, "loss": 0.0185, "step": 88320 }, { "grad_norm": 0.30585071444511414, "learning_rate": 3.677993385042494e-06, "loss": 0.0177, "step": 88330 }, { "grad_norm": 0.3707505166530609, "learning_rate": 3.67177156091944e-06, "loss": 0.0204, "step": 88340 }, { "grad_norm": 0.3733045160770416, "learning_rate": 3.6655548031817866e-06, "loss": 0.0186, "step": 88350 }, { "grad_norm": 0.2647753655910492, "learning_rate": 3.6593431125094e-06, "loss": 0.0217, "step": 88360 }, { "grad_norm": 0.394795686006546, "learning_rate": 3.6531364895815766e-06, "loss": 0.0213, "step": 88370 }, { "grad_norm": 0.2508315443992615, "learning_rate": 3.6469349350770677e-06, "loss": 0.0168, "step": 88380 }, { "grad_norm": 0.20427650213241577, "learning_rate": 3.6407384496740525e-06, "loss": 0.021, "step": 88390 }, { "grad_norm": 0.5420339107513428, "learning_rate": 3.6345470340501954e-06, "loss": 0.0215, "step": 88400 }, { "grad_norm": 0.6839079856872559, "learning_rate": 3.6283606888825537e-06, "loss": 0.0171, "step": 88410 }, { "grad_norm": 0.31758901476860046, "learning_rate": 3.622179414847676e-06, "loss": 0.0167, "step": 88420 }, { "grad_norm": 0.425615519285202, "learning_rate": 3.616003212621527e-06, "loss": 0.022, "step": 88430 }, { "grad_norm": 0.6118277907371521, "learning_rate": 3.609832082879533e-06, "loss": 0.021, "step": 88440 }, { "grad_norm": 0.5579307079315186, "learning_rate": 3.6036660262965485e-06, "loss": 0.0232, "step": 88450 }, { "grad_norm": 0.36498793959617615, "learning_rate": 3.597505043546895e-06, "loss": 0.023, "step": 88460 }, { "grad_norm": 0.4500439167022705, "learning_rate": 3.5913491353043227e-06, "loss": 0.0203, "step": 88470 }, { "grad_norm": 0.32276833057403564, "learning_rate": 3.5851983022420375e-06, "loss": 0.0226, "step": 88480 }, { "grad_norm": 0.3964834213256836, "learning_rate": 3.5790525450326797e-06, "loss": 0.0191, "step": 88490 }, { "grad_norm": 0.3379053771495819, "learning_rate": 3.572911864348344e-06, "loss": 0.0195, "step": 88500 }, { "grad_norm": 0.3250057101249695, "learning_rate": 3.5667762608605717e-06, "loss": 0.0182, "step": 88510 }, { "grad_norm": 0.25993266701698303, "learning_rate": 3.5606457352403255e-06, "loss": 0.021, "step": 88520 }, { "grad_norm": 0.48654142022132874, "learning_rate": 3.554520288158042e-06, "loss": 0.0175, "step": 88530 }, { "grad_norm": 0.6661980152130127, "learning_rate": 3.5483999202835903e-06, "loss": 0.0232, "step": 88540 }, { "grad_norm": 0.44603845477104187, "learning_rate": 3.5422846322862913e-06, "loss": 0.0231, "step": 88550 }, { "grad_norm": 0.7220129370689392, "learning_rate": 3.5361744248348925e-06, "loss": 0.0194, "step": 88560 }, { "grad_norm": 0.5182387828826904, "learning_rate": 3.5300692985976047e-06, "loss": 0.0206, "step": 88570 }, { "grad_norm": 2.410292387008667, "learning_rate": 3.52396925424206e-06, "loss": 0.0192, "step": 88580 }, { "grad_norm": 0.22577986121177673, "learning_rate": 3.5178742924353747e-06, "loss": 0.0169, "step": 88590 }, { "grad_norm": 0.22468425333499908, "learning_rate": 3.511784413844066e-06, "loss": 0.0185, "step": 88600 }, { "grad_norm": 0.3814685344696045, "learning_rate": 3.5056996191341287e-06, "loss": 0.0221, "step": 88610 }, { "grad_norm": 0.3652781546115875, "learning_rate": 3.4996199089709692e-06, "loss": 0.0207, "step": 88620 }, { "grad_norm": 1.041744589805603, "learning_rate": 3.4935452840194725e-06, "loss": 0.0196, "step": 88630 }, { "grad_norm": 0.23175974190235138, "learning_rate": 3.48747574494393e-06, "loss": 0.0247, "step": 88640 }, { "grad_norm": 0.36830034852027893, "learning_rate": 3.4814112924081154e-06, "loss": 0.0216, "step": 88650 }, { "grad_norm": 0.8361423015594482, "learning_rate": 3.475351927075221e-06, "loss": 0.0205, "step": 88660 }, { "grad_norm": 0.3113590180873871, "learning_rate": 3.469297649607894e-06, "loss": 0.0246, "step": 88670 }, { "grad_norm": 0.2710326015949249, "learning_rate": 3.463248460668217e-06, "loss": 0.0185, "step": 88680 }, { "grad_norm": 0.31894662976264954, "learning_rate": 3.457204360917726e-06, "loss": 0.0177, "step": 88690 }, { "grad_norm": 0.28331896662712097, "learning_rate": 3.4511653510173827e-06, "loss": 0.0216, "step": 88700 }, { "grad_norm": 0.15246039628982544, "learning_rate": 3.4451314316276074e-06, "loss": 0.0192, "step": 88710 }, { "grad_norm": 0.3586559295654297, "learning_rate": 3.4391026034082674e-06, "loss": 0.0179, "step": 88720 }, { "grad_norm": 0.2203211933374405, "learning_rate": 3.433078867018674e-06, "loss": 0.0152, "step": 88730 }, { "grad_norm": 0.4373282790184021, "learning_rate": 3.427060223117551e-06, "loss": 0.0211, "step": 88740 }, { "grad_norm": 0.3427080810070038, "learning_rate": 3.4210466723631042e-06, "loss": 0.0194, "step": 88750 }, { "grad_norm": 0.43261072039604187, "learning_rate": 3.415038215412969e-06, "loss": 0.0217, "step": 88760 }, { "grad_norm": 0.2542893588542938, "learning_rate": 3.4090348529242088e-06, "loss": 0.0226, "step": 88770 }, { "grad_norm": 0.5519663691520691, "learning_rate": 3.4030365855533487e-06, "loss": 0.0211, "step": 88780 }, { "grad_norm": 0.5521636605262756, "learning_rate": 3.397043413956347e-06, "loss": 0.0188, "step": 88790 }, { "grad_norm": 0.2515847980976105, "learning_rate": 3.3910553387886237e-06, "loss": 0.0184, "step": 88800 }, { "grad_norm": 1.845127820968628, "learning_rate": 3.3850723607049996e-06, "loss": 0.0216, "step": 88810 }, { "grad_norm": 0.49734190106391907, "learning_rate": 3.3790944803597902e-06, "loss": 0.0188, "step": 88820 }, { "grad_norm": 0.36758267879486084, "learning_rate": 3.373121698406706e-06, "loss": 0.0204, "step": 88830 }, { "grad_norm": 0.5553813576698303, "learning_rate": 3.3671540154989302e-06, "loss": 0.0228, "step": 88840 }, { "grad_norm": 0.25347062945365906, "learning_rate": 3.361191432289079e-06, "loss": 0.0213, "step": 88850 }, { "grad_norm": 0.5491201281547546, "learning_rate": 3.3552339494292207e-06, "loss": 0.0202, "step": 88860 }, { "grad_norm": 0.3927823305130005, "learning_rate": 3.349281567570839e-06, "loss": 0.0191, "step": 88870 }, { "grad_norm": 0.37641626596450806, "learning_rate": 3.3433342873648965e-06, "loss": 0.0171, "step": 88880 }, { "grad_norm": 0.6907494068145752, "learning_rate": 3.337392109461762e-06, "loss": 0.0223, "step": 88890 }, { "grad_norm": 0.4393814504146576, "learning_rate": 3.331455034511266e-06, "loss": 0.0228, "step": 88900 }, { "grad_norm": 0.3380548357963562, "learning_rate": 3.325523063162689e-06, "loss": 0.0244, "step": 88910 }, { "grad_norm": 0.4439391791820526, "learning_rate": 3.3195961960647393e-06, "loss": 0.021, "step": 88920 }, { "grad_norm": 0.4127773940563202, "learning_rate": 3.31367443386556e-06, "loss": 0.023, "step": 88930 }, { "grad_norm": 0.7404200434684753, "learning_rate": 3.3077577772127556e-06, "loss": 0.021, "step": 88940 }, { "grad_norm": 0.31485122442245483, "learning_rate": 3.301846226753358e-06, "loss": 0.0213, "step": 88950 }, { "grad_norm": 0.38518112897872925, "learning_rate": 3.2959397831338444e-06, "loss": 0.0262, "step": 88960 }, { "grad_norm": 2.4588749408721924, "learning_rate": 3.2900384470001432e-06, "loss": 0.0184, "step": 88970 }, { "grad_norm": 0.7402082681655884, "learning_rate": 3.28414221899761e-06, "loss": 0.0209, "step": 88980 }, { "grad_norm": 0.46119338274002075, "learning_rate": 3.278251099771046e-06, "loss": 0.0186, "step": 88990 }, { "grad_norm": 0.3693411350250244, "learning_rate": 3.2723650899646906e-06, "loss": 0.02, "step": 89000 }, { "grad_norm": 0.2692776620388031, "learning_rate": 3.2664841902222466e-06, "loss": 0.0197, "step": 89010 }, { "grad_norm": 0.3056795299053192, "learning_rate": 3.2606084011868208e-06, "loss": 0.0187, "step": 89020 }, { "grad_norm": 0.33025628328323364, "learning_rate": 3.254737723500989e-06, "loss": 0.0196, "step": 89030 }, { "grad_norm": 1.1729360818862915, "learning_rate": 3.2488721578067595e-06, "loss": 0.0164, "step": 89040 }, { "grad_norm": 0.6115491390228271, "learning_rate": 3.243011704745591e-06, "loss": 0.0217, "step": 89050 }, { "grad_norm": 0.26099637150764465, "learning_rate": 3.237156364958355e-06, "loss": 0.018, "step": 89060 }, { "grad_norm": 0.2277243286371231, "learning_rate": 3.231306139085394e-06, "loss": 0.02, "step": 89070 }, { "grad_norm": 0.6793321967124939, "learning_rate": 3.2254610277664797e-06, "loss": 0.021, "step": 89080 }, { "grad_norm": 0.3235434591770172, "learning_rate": 3.2196210316408183e-06, "loss": 0.0207, "step": 89090 }, { "grad_norm": 0.463094562292099, "learning_rate": 3.2137861513470644e-06, "loss": 0.0195, "step": 89100 }, { "grad_norm": 0.3987541198730469, "learning_rate": 3.2079563875233252e-06, "loss": 0.0191, "step": 89110 }, { "grad_norm": 0.40859878063201904, "learning_rate": 3.2021317408071184e-06, "loss": 0.0205, "step": 89120 }, { "grad_norm": 0.3239486813545227, "learning_rate": 3.196312211835434e-06, "loss": 0.018, "step": 89130 }, { "grad_norm": 0.751221776008606, "learning_rate": 3.1904978012446686e-06, "loss": 0.0206, "step": 89140 }, { "grad_norm": 0.5227933526039124, "learning_rate": 3.184688509670691e-06, "loss": 0.0227, "step": 89150 }, { "grad_norm": 0.47352510690689087, "learning_rate": 3.178884337748783e-06, "loss": 0.0224, "step": 89160 }, { "grad_norm": 0.2630073130130768, "learning_rate": 3.173085286113697e-06, "loss": 0.022, "step": 89170 }, { "grad_norm": 0.5739642381668091, "learning_rate": 3.1672913553995986e-06, "loss": 0.0199, "step": 89180 }, { "grad_norm": 0.8029351830482483, "learning_rate": 3.1615025462401138e-06, "loss": 0.0193, "step": 89190 }, { "grad_norm": 0.307889461517334, "learning_rate": 3.1557188592682808e-06, "loss": 0.0191, "step": 89200 }, { "grad_norm": 0.2289372682571411, "learning_rate": 3.1499402951166045e-06, "loss": 0.0187, "step": 89210 }, { "grad_norm": 1.9166817665100098, "learning_rate": 3.144166854417019e-06, "loss": 0.0153, "step": 89220 }, { "grad_norm": 0.69721519947052, "learning_rate": 3.1383985378009073e-06, "loss": 0.0188, "step": 89230 }, { "grad_norm": 0.34316161274909973, "learning_rate": 3.1326353458990654e-06, "loss": 0.0213, "step": 89240 }, { "grad_norm": 0.5679020285606384, "learning_rate": 3.126877279341761e-06, "loss": 0.0207, "step": 89250 }, { "grad_norm": 0.5430834889411926, "learning_rate": 3.1211243387586964e-06, "loss": 0.0211, "step": 89260 }, { "grad_norm": 0.4174463450908661, "learning_rate": 3.115376524778979e-06, "loss": 0.0244, "step": 89270 }, { "grad_norm": 0.2581562399864197, "learning_rate": 3.109633838031195e-06, "loss": 0.016, "step": 89280 }, { "grad_norm": 0.5778873562812805, "learning_rate": 3.103896279143359e-06, "loss": 0.0194, "step": 89290 }, { "grad_norm": 0.37155070900917053, "learning_rate": 3.098163848742924e-06, "loss": 0.0186, "step": 89300 }, { "grad_norm": 0.29598507285118103, "learning_rate": 3.0924365474567675e-06, "loss": 0.0224, "step": 89310 }, { "grad_norm": 0.8752530217170715, "learning_rate": 3.086714375911237e-06, "loss": 0.0193, "step": 89320 }, { "grad_norm": 0.23843048512935638, "learning_rate": 3.0809973347320775e-06, "loss": 0.0191, "step": 89330 }, { "grad_norm": 0.3226708769798279, "learning_rate": 3.075285424544516e-06, "loss": 0.0205, "step": 89340 }, { "grad_norm": 0.3083750903606415, "learning_rate": 3.069578645973176e-06, "loss": 0.0227, "step": 89350 }, { "grad_norm": 0.7293597459793091, "learning_rate": 3.0638769996421735e-06, "loss": 0.0212, "step": 89360 }, { "grad_norm": 0.28267553448677063, "learning_rate": 3.058180486175005e-06, "loss": 0.0196, "step": 89370 }, { "grad_norm": 0.24212464690208435, "learning_rate": 3.052489106194645e-06, "loss": 0.0192, "step": 89380 }, { "grad_norm": 0.27788466215133667, "learning_rate": 3.0468028603234946e-06, "loss": 0.0201, "step": 89390 }, { "grad_norm": 0.3114328384399414, "learning_rate": 3.04112174918339e-06, "loss": 0.0202, "step": 89400 }, { "grad_norm": 0.6643602252006531, "learning_rate": 3.0354457733955954e-06, "loss": 0.0249, "step": 89410 }, { "grad_norm": 0.4121372103691101, "learning_rate": 3.029774933580859e-06, "loss": 0.0213, "step": 89420 }, { "grad_norm": 0.29748597741127014, "learning_rate": 3.0241092303593065e-06, "loss": 0.019, "step": 89430 }, { "grad_norm": 0.3555295169353485, "learning_rate": 3.018448664350554e-06, "loss": 0.021, "step": 89440 }, { "grad_norm": 0.2626303732395172, "learning_rate": 3.0127932361736055e-06, "loss": 0.017, "step": 89450 }, { "grad_norm": 1.1046875715255737, "learning_rate": 3.00714294644695e-06, "loss": 0.0188, "step": 89460 }, { "grad_norm": 0.6260037422180176, "learning_rate": 3.0014977957884828e-06, "loss": 0.0232, "step": 89470 }, { "grad_norm": 1.0310059785842896, "learning_rate": 2.99585778481557e-06, "loss": 0.0194, "step": 89480 }, { "grad_norm": 0.5369212031364441, "learning_rate": 2.990222914144969e-06, "loss": 0.0197, "step": 89490 }, { "grad_norm": 0.4397680461406708, "learning_rate": 2.984593184392914e-06, "loss": 0.0227, "step": 89500 }, { "grad_norm": 0.3576357662677765, "learning_rate": 2.978968596175069e-06, "loss": 0.0184, "step": 89510 }, { "grad_norm": 0.8725321292877197, "learning_rate": 2.9733491501065137e-06, "loss": 0.0182, "step": 89520 }, { "grad_norm": 0.25636351108551025, "learning_rate": 2.9677348468018005e-06, "loss": 0.0204, "step": 89530 }, { "grad_norm": 0.47141483426094055, "learning_rate": 2.9621256868748726e-06, "loss": 0.0181, "step": 89540 }, { "grad_norm": 0.2925419509410858, "learning_rate": 2.956521670939183e-06, "loss": 0.0238, "step": 89550 }, { "grad_norm": 0.9034027457237244, "learning_rate": 2.950922799607542e-06, "loss": 0.019, "step": 89560 }, { "grad_norm": 0.5269219279289246, "learning_rate": 2.9453290734922535e-06, "loss": 0.0185, "step": 89570 }, { "grad_norm": 0.42232367396354675, "learning_rate": 2.939740493205029e-06, "loss": 0.0182, "step": 89580 }, { "grad_norm": 0.5064329504966736, "learning_rate": 2.9341570593570345e-06, "loss": 0.0172, "step": 89590 }, { "grad_norm": 0.4003695845603943, "learning_rate": 2.928578772558843e-06, "loss": 0.0204, "step": 89600 }, { "grad_norm": 0.3671402335166931, "learning_rate": 2.923005633420528e-06, "loss": 0.0181, "step": 89610 }, { "grad_norm": 0.6761907339096069, "learning_rate": 2.9174376425515237e-06, "loss": 0.016, "step": 89620 }, { "grad_norm": 0.34003183245658875, "learning_rate": 2.91187480056076e-06, "loss": 0.021, "step": 89630 }, { "grad_norm": 0.30993372201919556, "learning_rate": 2.9063171080565664e-06, "loss": 0.0215, "step": 89640 }, { "grad_norm": 0.36905768513679504, "learning_rate": 2.900764565646741e-06, "loss": 0.0229, "step": 89650 }, { "grad_norm": 0.2956681251525879, "learning_rate": 2.895217173938475e-06, "loss": 0.0242, "step": 89660 }, { "grad_norm": 0.3396986722946167, "learning_rate": 2.8896749335384445e-06, "loss": 0.0222, "step": 89670 }, { "grad_norm": 1.5087230205535889, "learning_rate": 2.884137845052731e-06, "loss": 0.0208, "step": 89680 }, { "grad_norm": 0.35102590918540955, "learning_rate": 2.8786059090868734e-06, "loss": 0.0177, "step": 89690 }, { "grad_norm": 0.7740440964698792, "learning_rate": 2.87307912624582e-06, "loss": 0.0183, "step": 89700 }, { "grad_norm": 1.1280126571655273, "learning_rate": 2.867557497133977e-06, "loss": 0.0185, "step": 89710 }, { "grad_norm": 0.39508283138275146, "learning_rate": 2.862041022355194e-06, "loss": 0.0184, "step": 89720 }, { "grad_norm": 0.5497466921806335, "learning_rate": 2.8565297025127226e-06, "loss": 0.0187, "step": 89730 }, { "grad_norm": 0.44625017046928406, "learning_rate": 2.8510235382092853e-06, "loss": 0.0235, "step": 89740 }, { "grad_norm": 0.19898827373981476, "learning_rate": 2.845522530047029e-06, "loss": 0.0157, "step": 89750 }, { "grad_norm": 0.37221887707710266, "learning_rate": 2.8400266786275387e-06, "loss": 0.0213, "step": 89760 }, { "grad_norm": 0.2483285516500473, "learning_rate": 2.834535984551817e-06, "loss": 0.0158, "step": 89770 }, { "grad_norm": 0.18063414096832275, "learning_rate": 2.8290504484203397e-06, "loss": 0.0157, "step": 89780 }, { "grad_norm": 0.19498993456363678, "learning_rate": 2.8235700708329705e-06, "loss": 0.0178, "step": 89790 }, { "grad_norm": 0.49084439873695374, "learning_rate": 2.818094852389058e-06, "loss": 0.0194, "step": 89800 }, { "grad_norm": 0.41830748319625854, "learning_rate": 2.8126247936873516e-06, "loss": 0.0218, "step": 89810 }, { "grad_norm": 1.0058468580245972, "learning_rate": 2.8071598953260614e-06, "loss": 0.0225, "step": 89820 }, { "grad_norm": 0.2876463234424591, "learning_rate": 2.8017001579028035e-06, "loss": 0.017, "step": 89830 }, { "grad_norm": 0.27692970633506775, "learning_rate": 2.7962455820146617e-06, "loss": 0.0176, "step": 89840 }, { "grad_norm": 0.23126035928726196, "learning_rate": 2.7907961682581253e-06, "loss": 0.0272, "step": 89850 }, { "grad_norm": 0.3869359493255615, "learning_rate": 2.7853519172291453e-06, "loss": 0.0254, "step": 89860 }, { "grad_norm": 0.5429466962814331, "learning_rate": 2.7799128295230904e-06, "loss": 0.0214, "step": 89870 }, { "grad_norm": 0.31172823905944824, "learning_rate": 2.774478905734784e-06, "loss": 0.0176, "step": 89880 }, { "grad_norm": 0.4849848449230194, "learning_rate": 2.7690501464584516e-06, "loss": 0.0176, "step": 89890 }, { "grad_norm": 0.29507049918174744, "learning_rate": 2.763626552287796e-06, "loss": 0.0171, "step": 89900 }, { "grad_norm": 0.37615716457366943, "learning_rate": 2.7582081238159083e-06, "loss": 0.0144, "step": 89910 }, { "grad_norm": 0.3778369128704071, "learning_rate": 2.7527948616353548e-06, "loss": 0.0231, "step": 89920 }, { "grad_norm": 0.43942588567733765, "learning_rate": 2.7473867663381226e-06, "loss": 0.0171, "step": 89930 }, { "grad_norm": 0.22532983124256134, "learning_rate": 2.7419838385156383e-06, "loss": 0.0166, "step": 89940 }, { "grad_norm": 0.3542008399963379, "learning_rate": 2.7365860787587407e-06, "loss": 0.0217, "step": 89950 }, { "grad_norm": 0.2987840175628662, "learning_rate": 2.7311934876577295e-06, "loss": 0.0168, "step": 89960 }, { "grad_norm": 0.28538334369659424, "learning_rate": 2.725806065802339e-06, "loss": 0.0213, "step": 89970 }, { "grad_norm": 0.464894562959671, "learning_rate": 2.720423813781714e-06, "loss": 0.0197, "step": 89980 }, { "grad_norm": 0.3692210018634796, "learning_rate": 2.715046732184462e-06, "loss": 0.0181, "step": 89990 }, { "grad_norm": 0.40891334414482117, "learning_rate": 2.7096748215986013e-06, "loss": 0.0214, "step": 90000 } ], "logging_steps": 10, "max_steps": 100000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 10000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 24, "trial_name": null, "trial_params": null }