{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 100000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 0.7189611196517944, "learning_rate": 1.8e-07, "loss": 1.3842, "step": 10 }, { "grad_norm": 0.6879875063896179, "learning_rate": 3.8e-07, "loss": 1.3737, "step": 20 }, { "grad_norm": 0.6126565337181091, "learning_rate": 5.8e-07, "loss": 1.3811, "step": 30 }, { "grad_norm": 0.6504364609718323, "learning_rate": 7.8e-07, "loss": 1.3828, "step": 40 }, { "grad_norm": 0.5468832850456238, "learning_rate": 9.8e-07, "loss": 1.3818, "step": 50 }, { "grad_norm": 0.7315312623977661, "learning_rate": 1.18e-06, "loss": 1.346, "step": 60 }, { "grad_norm": 0.9557607173919678, "learning_rate": 1.3800000000000001e-06, "loss": 1.3308, "step": 70 }, { "grad_norm": 0.8067951798439026, "learning_rate": 1.5800000000000003e-06, "loss": 1.3094, "step": 80 }, { "grad_norm": 0.673760712146759, "learning_rate": 1.7800000000000001e-06, "loss": 1.3023, "step": 90 }, { "grad_norm": 1.1376993656158447, "learning_rate": 1.98e-06, "loss": 1.284, "step": 100 }, { "grad_norm": 3.015726327896118, "learning_rate": 2.1800000000000003e-06, "loss": 1.2712, "step": 110 }, { "grad_norm": 8.025469779968262, "learning_rate": 2.38e-06, "loss": 1.2373, "step": 120 }, { "grad_norm": 1.365903377532959, "learning_rate": 2.5800000000000003e-06, "loss": 1.2071, "step": 130 }, { "grad_norm": 0.8577656149864197, "learning_rate": 2.78e-06, "loss": 1.1965, "step": 140 }, { "grad_norm": 1.5047162771224976, "learning_rate": 2.9800000000000003e-06, "loss": 1.1902, "step": 150 }, { "grad_norm": 1.6267668008804321, "learning_rate": 3.1800000000000005e-06, "loss": 1.1636, "step": 160 }, { "grad_norm": 1.369300127029419, "learning_rate": 3.38e-06, "loss": 1.1552, "step": 170 }, { "grad_norm": 1.186592698097229, "learning_rate": 3.58e-06, "loss": 1.14, "step": 180 }, { "grad_norm": 1.2511482238769531, "learning_rate": 3.7800000000000002e-06, "loss": 1.145, "step": 190 }, { "grad_norm": 1.7414101362228394, "learning_rate": 3.98e-06, "loss": 1.1205, "step": 200 }, { "grad_norm": 1.342803716659546, "learning_rate": 4.18e-06, "loss": 1.0974, "step": 210 }, { "grad_norm": 1.5488823652267456, "learning_rate": 4.38e-06, "loss": 1.1055, "step": 220 }, { "grad_norm": 1.298343300819397, "learning_rate": 4.58e-06, "loss": 1.0898, "step": 230 }, { "grad_norm": 1.3222663402557373, "learning_rate": 4.780000000000001e-06, "loss": 1.089, "step": 240 }, { "grad_norm": 1.451127290725708, "learning_rate": 4.98e-06, "loss": 1.0947, "step": 250 }, { "grad_norm": 0.9744407534599304, "learning_rate": 5.18e-06, "loss": 1.0842, "step": 260 }, { "grad_norm": 1.1717950105667114, "learning_rate": 5.38e-06, "loss": 1.0733, "step": 270 }, { "grad_norm": 1.713626742362976, "learning_rate": 5.580000000000001e-06, "loss": 1.0824, "step": 280 }, { "grad_norm": 1.225218415260315, "learning_rate": 5.78e-06, "loss": 1.068, "step": 290 }, { "grad_norm": 1.0755562782287598, "learning_rate": 5.98e-06, "loss": 1.0584, "step": 300 }, { "grad_norm": 2.0199029445648193, "learning_rate": 6.18e-06, "loss": 1.0409, "step": 310 }, { "grad_norm": 1.5886393785476685, "learning_rate": 6.38e-06, "loss": 1.0189, "step": 320 }, { "grad_norm": 1.2317723035812378, "learning_rate": 6.58e-06, "loss": 0.988, "step": 330 }, { "grad_norm": 1.7935892343521118, "learning_rate": 6.78e-06, "loss": 0.9767, "step": 340 }, { "grad_norm": 1.568803071975708, "learning_rate": 6.98e-06, "loss": 0.9533, "step": 350 }, { "grad_norm": 1.7755630016326904, "learning_rate": 7.180000000000001e-06, "loss": 0.9367, "step": 360 }, { "grad_norm": 1.7311232089996338, "learning_rate": 7.3800000000000005e-06, "loss": 0.9089, "step": 370 }, { "grad_norm": 2.0171332359313965, "learning_rate": 7.580000000000001e-06, "loss": 0.8685, "step": 380 }, { "grad_norm": 1.9253841638565063, "learning_rate": 7.78e-06, "loss": 0.8377, "step": 390 }, { "grad_norm": 2.010132074356079, "learning_rate": 7.98e-06, "loss": 0.7991, "step": 400 }, { "grad_norm": 1.7832802534103394, "learning_rate": 8.18e-06, "loss": 0.7692, "step": 410 }, { "grad_norm": 1.5162575244903564, "learning_rate": 8.380000000000001e-06, "loss": 0.7336, "step": 420 }, { "grad_norm": 2.259995222091675, "learning_rate": 8.580000000000001e-06, "loss": 0.6914, "step": 430 }, { "grad_norm": 2.1575515270233154, "learning_rate": 8.78e-06, "loss": 0.6633, "step": 440 }, { "grad_norm": 2.043316125869751, "learning_rate": 8.98e-06, "loss": 0.6064, "step": 450 }, { "grad_norm": 2.5843749046325684, "learning_rate": 9.180000000000002e-06, "loss": 0.5853, "step": 460 }, { "grad_norm": 2.260324001312256, "learning_rate": 9.38e-06, "loss": 0.551, "step": 470 }, { "grad_norm": 2.336383581161499, "learning_rate": 9.58e-06, "loss": 0.5056, "step": 480 }, { "grad_norm": 2.123439311981201, "learning_rate": 9.78e-06, "loss": 0.4822, "step": 490 }, { "grad_norm": 2.1499857902526855, "learning_rate": 9.980000000000001e-06, "loss": 0.447, "step": 500 }, { "grad_norm": 2.327699661254883, "learning_rate": 1.018e-05, "loss": 0.4284, "step": 510 }, { "grad_norm": 2.4474050998687744, "learning_rate": 1.038e-05, "loss": 0.3978, "step": 520 }, { "grad_norm": 2.3307323455810547, "learning_rate": 1.058e-05, "loss": 0.3799, "step": 530 }, { "grad_norm": 2.236637592315674, "learning_rate": 1.0780000000000002e-05, "loss": 0.3561, "step": 540 }, { "grad_norm": 2.771460771560669, "learning_rate": 1.098e-05, "loss": 0.3716, "step": 550 }, { "grad_norm": 2.610137939453125, "learning_rate": 1.118e-05, "loss": 0.3468, "step": 560 }, { "grad_norm": 2.8033499717712402, "learning_rate": 1.1380000000000001e-05, "loss": 0.3404, "step": 570 }, { "grad_norm": 2.3770477771759033, "learning_rate": 1.1580000000000001e-05, "loss": 0.3208, "step": 580 }, { "grad_norm": 2.4397497177124023, "learning_rate": 1.178e-05, "loss": 0.2941, "step": 590 }, { "grad_norm": 2.228759527206421, "learning_rate": 1.198e-05, "loss": 0.2881, "step": 600 }, { "grad_norm": 2.286252498626709, "learning_rate": 1.2180000000000002e-05, "loss": 0.2729, "step": 610 }, { "grad_norm": 3.141627550125122, "learning_rate": 1.238e-05, "loss": 0.2864, "step": 620 }, { "grad_norm": 2.458711862564087, "learning_rate": 1.258e-05, "loss": 0.2269, "step": 630 }, { "grad_norm": 2.295487880706787, "learning_rate": 1.278e-05, "loss": 0.2442, "step": 640 }, { "grad_norm": 2.466015100479126, "learning_rate": 1.2980000000000001e-05, "loss": 0.2144, "step": 650 }, { "grad_norm": 2.092846393585205, "learning_rate": 1.3180000000000001e-05, "loss": 0.2326, "step": 660 }, { "grad_norm": 2.677034378051758, "learning_rate": 1.338e-05, "loss": 0.2223, "step": 670 }, { "grad_norm": 2.288968563079834, "learning_rate": 1.358e-05, "loss": 0.2034, "step": 680 }, { "grad_norm": 2.3460679054260254, "learning_rate": 1.3780000000000002e-05, "loss": 0.2035, "step": 690 }, { "grad_norm": 1.9059590101242065, "learning_rate": 1.3980000000000002e-05, "loss": 0.1989, "step": 700 }, { "grad_norm": 2.3070943355560303, "learning_rate": 1.4180000000000001e-05, "loss": 0.2044, "step": 710 }, { "grad_norm": 2.563772678375244, "learning_rate": 1.4380000000000001e-05, "loss": 0.1952, "step": 720 }, { "grad_norm": 2.42689847946167, "learning_rate": 1.4580000000000003e-05, "loss": 0.1984, "step": 730 }, { "grad_norm": 3.070462703704834, "learning_rate": 1.4779999999999999e-05, "loss": 0.1828, "step": 740 }, { "grad_norm": 2.0091800689697266, "learning_rate": 1.4979999999999999e-05, "loss": 0.1911, "step": 750 }, { "grad_norm": 2.2731034755706787, "learning_rate": 1.518e-05, "loss": 0.1748, "step": 760 }, { "grad_norm": 2.4065184593200684, "learning_rate": 1.538e-05, "loss": 0.1976, "step": 770 }, { "grad_norm": 2.1232128143310547, "learning_rate": 1.558e-05, "loss": 0.1796, "step": 780 }, { "grad_norm": 2.0898525714874268, "learning_rate": 1.578e-05, "loss": 0.1792, "step": 790 }, { "grad_norm": 2.4013845920562744, "learning_rate": 1.598e-05, "loss": 0.1867, "step": 800 }, { "grad_norm": 2.1258385181427, "learning_rate": 1.618e-05, "loss": 0.1922, "step": 810 }, { "grad_norm": 2.6432642936706543, "learning_rate": 1.6380000000000002e-05, "loss": 0.1709, "step": 820 }, { "grad_norm": 2.3923213481903076, "learning_rate": 1.658e-05, "loss": 0.1775, "step": 830 }, { "grad_norm": 2.2662131786346436, "learning_rate": 1.6780000000000002e-05, "loss": 0.1558, "step": 840 }, { "grad_norm": 2.107659339904785, "learning_rate": 1.698e-05, "loss": 0.1536, "step": 850 }, { "grad_norm": 1.9157633781433105, "learning_rate": 1.718e-05, "loss": 0.1689, "step": 860 }, { "grad_norm": 2.148111343383789, "learning_rate": 1.7380000000000003e-05, "loss": 0.1665, "step": 870 }, { "grad_norm": 1.98641037940979, "learning_rate": 1.758e-05, "loss": 0.1605, "step": 880 }, { "grad_norm": 1.9659124612808228, "learning_rate": 1.7780000000000003e-05, "loss": 0.1647, "step": 890 }, { "grad_norm": 2.02353572845459, "learning_rate": 1.798e-05, "loss": 0.1694, "step": 900 }, { "grad_norm": 2.0694003105163574, "learning_rate": 1.818e-05, "loss": 0.1683, "step": 910 }, { "grad_norm": 2.061793327331543, "learning_rate": 1.838e-05, "loss": 0.1436, "step": 920 }, { "grad_norm": 2.888801097869873, "learning_rate": 1.858e-05, "loss": 0.159, "step": 930 }, { "grad_norm": 2.250309705734253, "learning_rate": 1.878e-05, "loss": 0.1692, "step": 940 }, { "grad_norm": 1.4626997709274292, "learning_rate": 1.898e-05, "loss": 0.1535, "step": 950 }, { "grad_norm": 2.2808635234832764, "learning_rate": 1.918e-05, "loss": 0.145, "step": 960 }, { "grad_norm": 1.7916803359985352, "learning_rate": 1.938e-05, "loss": 0.1487, "step": 970 }, { "grad_norm": 1.6974761486053467, "learning_rate": 1.9580000000000002e-05, "loss": 0.1417, "step": 980 }, { "grad_norm": 2.119649648666382, "learning_rate": 1.978e-05, "loss": 0.1308, "step": 990 }, { "grad_norm": 1.5869832038879395, "learning_rate": 1.9980000000000002e-05, "loss": 0.1578, "step": 1000 }, { "grad_norm": 1.7303175926208496, "learning_rate": 2.0180000000000003e-05, "loss": 0.1323, "step": 1010 }, { "grad_norm": 1.6401172876358032, "learning_rate": 2.038e-05, "loss": 0.1439, "step": 1020 }, { "grad_norm": 2.6017205715179443, "learning_rate": 2.0580000000000003e-05, "loss": 0.1501, "step": 1030 }, { "grad_norm": 1.8912180662155151, "learning_rate": 2.078e-05, "loss": 0.1325, "step": 1040 }, { "grad_norm": 2.1562752723693848, "learning_rate": 2.098e-05, "loss": 0.13, "step": 1050 }, { "grad_norm": 2.1845991611480713, "learning_rate": 2.118e-05, "loss": 0.1412, "step": 1060 }, { "grad_norm": 1.7572767734527588, "learning_rate": 2.138e-05, "loss": 0.1337, "step": 1070 }, { "grad_norm": 1.197597861289978, "learning_rate": 2.158e-05, "loss": 0.1453, "step": 1080 }, { "grad_norm": 1.7756402492523193, "learning_rate": 2.178e-05, "loss": 0.137, "step": 1090 }, { "grad_norm": 1.4899978637695312, "learning_rate": 2.198e-05, "loss": 0.1258, "step": 1100 }, { "grad_norm": 2.1386632919311523, "learning_rate": 2.218e-05, "loss": 0.1466, "step": 1110 }, { "grad_norm": 1.6205018758773804, "learning_rate": 2.2380000000000003e-05, "loss": 0.1293, "step": 1120 }, { "grad_norm": 1.5366488695144653, "learning_rate": 2.258e-05, "loss": 0.1438, "step": 1130 }, { "grad_norm": 1.9599796533584595, "learning_rate": 2.2780000000000002e-05, "loss": 0.1299, "step": 1140 }, { "grad_norm": 1.6680326461791992, "learning_rate": 2.298e-05, "loss": 0.132, "step": 1150 }, { "grad_norm": 1.864543080329895, "learning_rate": 2.318e-05, "loss": 0.1291, "step": 1160 }, { "grad_norm": 1.5462638139724731, "learning_rate": 2.3380000000000003e-05, "loss": 0.1209, "step": 1170 }, { "grad_norm": 1.7018221616744995, "learning_rate": 2.358e-05, "loss": 0.1235, "step": 1180 }, { "grad_norm": 2.031799793243408, "learning_rate": 2.3780000000000003e-05, "loss": 0.1213, "step": 1190 }, { "grad_norm": 1.878738284111023, "learning_rate": 2.398e-05, "loss": 0.1368, "step": 1200 }, { "grad_norm": 1.6468981504440308, "learning_rate": 2.418e-05, "loss": 0.1135, "step": 1210 }, { "grad_norm": 1.3681260347366333, "learning_rate": 2.438e-05, "loss": 0.1268, "step": 1220 }, { "grad_norm": 1.4214974641799927, "learning_rate": 2.4580000000000002e-05, "loss": 0.1276, "step": 1230 }, { "grad_norm": 1.3952361345291138, "learning_rate": 2.478e-05, "loss": 0.1155, "step": 1240 }, { "grad_norm": 1.632112979888916, "learning_rate": 2.498e-05, "loss": 0.1211, "step": 1250 }, { "grad_norm": 2.013136863708496, "learning_rate": 2.5180000000000003e-05, "loss": 0.1265, "step": 1260 }, { "grad_norm": 1.5496902465820312, "learning_rate": 2.5380000000000004e-05, "loss": 0.1284, "step": 1270 }, { "grad_norm": 1.7339837551116943, "learning_rate": 2.5580000000000002e-05, "loss": 0.1187, "step": 1280 }, { "grad_norm": 1.310723066329956, "learning_rate": 2.5779999999999997e-05, "loss": 0.1199, "step": 1290 }, { "grad_norm": 1.516270399093628, "learning_rate": 2.598e-05, "loss": 0.115, "step": 1300 }, { "grad_norm": 1.450111985206604, "learning_rate": 2.618e-05, "loss": 0.1226, "step": 1310 }, { "grad_norm": 1.408196210861206, "learning_rate": 2.6379999999999998e-05, "loss": 0.1314, "step": 1320 }, { "grad_norm": 1.8964042663574219, "learning_rate": 2.658e-05, "loss": 0.1212, "step": 1330 }, { "grad_norm": 1.7944577932357788, "learning_rate": 2.678e-05, "loss": 0.1263, "step": 1340 }, { "grad_norm": 2.0621535778045654, "learning_rate": 2.698e-05, "loss": 0.1143, "step": 1350 }, { "grad_norm": 1.5734115839004517, "learning_rate": 2.718e-05, "loss": 0.1162, "step": 1360 }, { "grad_norm": 1.663405418395996, "learning_rate": 2.738e-05, "loss": 0.1161, "step": 1370 }, { "grad_norm": 1.220220923423767, "learning_rate": 2.758e-05, "loss": 0.1172, "step": 1380 }, { "grad_norm": 1.2429158687591553, "learning_rate": 2.778e-05, "loss": 0.1182, "step": 1390 }, { "grad_norm": 1.382276177406311, "learning_rate": 2.798e-05, "loss": 0.1147, "step": 1400 }, { "grad_norm": 1.5186676979064941, "learning_rate": 2.818e-05, "loss": 0.1237, "step": 1410 }, { "grad_norm": 1.2450891733169556, "learning_rate": 2.8380000000000003e-05, "loss": 0.1317, "step": 1420 }, { "grad_norm": 1.6646844148635864, "learning_rate": 2.858e-05, "loss": 0.1122, "step": 1430 }, { "grad_norm": 1.3237745761871338, "learning_rate": 2.8780000000000002e-05, "loss": 0.1226, "step": 1440 }, { "grad_norm": 1.4063031673431396, "learning_rate": 2.898e-05, "loss": 0.099, "step": 1450 }, { "grad_norm": 1.4078712463378906, "learning_rate": 2.9180000000000002e-05, "loss": 0.1117, "step": 1460 }, { "grad_norm": 1.270833969116211, "learning_rate": 2.9380000000000003e-05, "loss": 0.1122, "step": 1470 }, { "grad_norm": 1.326695442199707, "learning_rate": 2.958e-05, "loss": 0.1201, "step": 1480 }, { "grad_norm": 1.279596209526062, "learning_rate": 2.9780000000000003e-05, "loss": 0.1108, "step": 1490 }, { "grad_norm": 1.3284372091293335, "learning_rate": 2.998e-05, "loss": 0.1108, "step": 1500 }, { "grad_norm": 1.3078763484954834, "learning_rate": 3.0180000000000002e-05, "loss": 0.118, "step": 1510 }, { "grad_norm": 2.006213665008545, "learning_rate": 3.0380000000000004e-05, "loss": 0.1198, "step": 1520 }, { "grad_norm": 1.1841135025024414, "learning_rate": 3.058e-05, "loss": 0.1182, "step": 1530 }, { "grad_norm": 1.7675119638442993, "learning_rate": 3.078e-05, "loss": 0.11, "step": 1540 }, { "grad_norm": 1.1877492666244507, "learning_rate": 3.0980000000000005e-05, "loss": 0.1151, "step": 1550 }, { "grad_norm": 1.0714967250823975, "learning_rate": 3.118e-05, "loss": 0.1175, "step": 1560 }, { "grad_norm": 1.6107757091522217, "learning_rate": 3.138e-05, "loss": 0.1159, "step": 1570 }, { "grad_norm": 1.2059396505355835, "learning_rate": 3.1580000000000006e-05, "loss": 0.1183, "step": 1580 }, { "grad_norm": 1.2940316200256348, "learning_rate": 3.1780000000000004e-05, "loss": 0.1156, "step": 1590 }, { "grad_norm": 1.2308064699172974, "learning_rate": 3.198e-05, "loss": 0.1217, "step": 1600 }, { "grad_norm": 1.3787295818328857, "learning_rate": 3.218e-05, "loss": 0.1147, "step": 1610 }, { "grad_norm": 1.262906789779663, "learning_rate": 3.238e-05, "loss": 0.1115, "step": 1620 }, { "grad_norm": 1.2992266416549683, "learning_rate": 3.2579999999999996e-05, "loss": 0.1143, "step": 1630 }, { "grad_norm": 1.3992689847946167, "learning_rate": 3.278e-05, "loss": 0.1068, "step": 1640 }, { "grad_norm": 1.3863149881362915, "learning_rate": 3.298e-05, "loss": 0.1167, "step": 1650 }, { "grad_norm": 1.3933520317077637, "learning_rate": 3.318e-05, "loss": 0.1163, "step": 1660 }, { "grad_norm": 1.5092604160308838, "learning_rate": 3.338e-05, "loss": 0.1136, "step": 1670 }, { "grad_norm": 1.2302381992340088, "learning_rate": 3.358e-05, "loss": 0.1088, "step": 1680 }, { "grad_norm": 1.4704314470291138, "learning_rate": 3.378e-05, "loss": 0.1152, "step": 1690 }, { "grad_norm": 1.644866943359375, "learning_rate": 3.398e-05, "loss": 0.1082, "step": 1700 }, { "grad_norm": 0.919418215751648, "learning_rate": 3.418e-05, "loss": 0.1236, "step": 1710 }, { "grad_norm": 1.3294237852096558, "learning_rate": 3.438e-05, "loss": 0.1119, "step": 1720 }, { "grad_norm": 1.5318598747253418, "learning_rate": 3.4580000000000004e-05, "loss": 0.1142, "step": 1730 }, { "grad_norm": 1.2117044925689697, "learning_rate": 3.478e-05, "loss": 0.105, "step": 1740 }, { "grad_norm": 1.0003740787506104, "learning_rate": 3.498e-05, "loss": 0.1086, "step": 1750 }, { "grad_norm": 1.3572481870651245, "learning_rate": 3.518e-05, "loss": 0.1133, "step": 1760 }, { "grad_norm": 0.8142414689064026, "learning_rate": 3.5380000000000003e-05, "loss": 0.1093, "step": 1770 }, { "grad_norm": 1.303159475326538, "learning_rate": 3.558e-05, "loss": 0.1127, "step": 1780 }, { "grad_norm": 1.3594236373901367, "learning_rate": 3.578e-05, "loss": 0.109, "step": 1790 }, { "grad_norm": 1.144524097442627, "learning_rate": 3.5980000000000004e-05, "loss": 0.1122, "step": 1800 }, { "grad_norm": 1.4413033723831177, "learning_rate": 3.618e-05, "loss": 0.1091, "step": 1810 }, { "grad_norm": 0.9457780718803406, "learning_rate": 3.638e-05, "loss": 0.1048, "step": 1820 }, { "grad_norm": 1.0731059312820435, "learning_rate": 3.6580000000000006e-05, "loss": 0.0995, "step": 1830 }, { "grad_norm": 1.0861279964447021, "learning_rate": 3.6780000000000004e-05, "loss": 0.099, "step": 1840 }, { "grad_norm": 1.165177345275879, "learning_rate": 3.698e-05, "loss": 0.1, "step": 1850 }, { "grad_norm": 1.6080818176269531, "learning_rate": 3.7180000000000007e-05, "loss": 0.1071, "step": 1860 }, { "grad_norm": 1.2707765102386475, "learning_rate": 3.7380000000000005e-05, "loss": 0.104, "step": 1870 }, { "grad_norm": 1.087878704071045, "learning_rate": 3.758e-05, "loss": 0.0995, "step": 1880 }, { "grad_norm": 1.0919369459152222, "learning_rate": 3.778000000000001e-05, "loss": 0.1082, "step": 1890 }, { "grad_norm": 0.9082920551300049, "learning_rate": 3.7980000000000006e-05, "loss": 0.1001, "step": 1900 }, { "grad_norm": 1.2655599117279053, "learning_rate": 3.818e-05, "loss": 0.1028, "step": 1910 }, { "grad_norm": 1.1986820697784424, "learning_rate": 3.838e-05, "loss": 0.1077, "step": 1920 }, { "grad_norm": 1.1930919885635376, "learning_rate": 3.858e-05, "loss": 0.1058, "step": 1930 }, { "grad_norm": 1.2883825302124023, "learning_rate": 3.878e-05, "loss": 0.1099, "step": 1940 }, { "grad_norm": 0.9399268627166748, "learning_rate": 3.898e-05, "loss": 0.1035, "step": 1950 }, { "grad_norm": 1.0352765321731567, "learning_rate": 3.918e-05, "loss": 0.1077, "step": 1960 }, { "grad_norm": 0.8671998977661133, "learning_rate": 3.938e-05, "loss": 0.1049, "step": 1970 }, { "grad_norm": 1.0644190311431885, "learning_rate": 3.958e-05, "loss": 0.1032, "step": 1980 }, { "grad_norm": 1.1739046573638916, "learning_rate": 3.978e-05, "loss": 0.0985, "step": 1990 }, { "grad_norm": 1.2296907901763916, "learning_rate": 3.998e-05, "loss": 0.1205, "step": 2000 }, { "grad_norm": 1.113714575767517, "learning_rate": 4.018e-05, "loss": 0.0983, "step": 2010 }, { "grad_norm": 1.1107817888259888, "learning_rate": 4.038e-05, "loss": 0.1111, "step": 2020 }, { "grad_norm": 1.3059853315353394, "learning_rate": 4.058e-05, "loss": 0.1083, "step": 2030 }, { "grad_norm": 0.9844303727149963, "learning_rate": 4.078e-05, "loss": 0.1041, "step": 2040 }, { "grad_norm": 1.0621423721313477, "learning_rate": 4.0980000000000004e-05, "loss": 0.1067, "step": 2050 }, { "grad_norm": 1.323367714881897, "learning_rate": 4.118e-05, "loss": 0.117, "step": 2060 }, { "grad_norm": 1.0997182130813599, "learning_rate": 4.138e-05, "loss": 0.1109, "step": 2070 }, { "grad_norm": 1.421099066734314, "learning_rate": 4.1580000000000005e-05, "loss": 0.1035, "step": 2080 }, { "grad_norm": 1.098201036453247, "learning_rate": 4.178e-05, "loss": 0.099, "step": 2090 }, { "grad_norm": 0.7171524167060852, "learning_rate": 4.198e-05, "loss": 0.096, "step": 2100 }, { "grad_norm": 0.875731885433197, "learning_rate": 4.2180000000000006e-05, "loss": 0.0925, "step": 2110 }, { "grad_norm": 0.9568856954574585, "learning_rate": 4.2380000000000004e-05, "loss": 0.0976, "step": 2120 }, { "grad_norm": 1.1401640176773071, "learning_rate": 4.258e-05, "loss": 0.0955, "step": 2130 }, { "grad_norm": 0.9796532392501831, "learning_rate": 4.278e-05, "loss": 0.0978, "step": 2140 }, { "grad_norm": 0.9539077877998352, "learning_rate": 4.2980000000000005e-05, "loss": 0.095, "step": 2150 }, { "grad_norm": 0.8313602209091187, "learning_rate": 4.318e-05, "loss": 0.0953, "step": 2160 }, { "grad_norm": 0.848667323589325, "learning_rate": 4.338e-05, "loss": 0.0887, "step": 2170 }, { "grad_norm": 1.3781222105026245, "learning_rate": 4.3580000000000006e-05, "loss": 0.0956, "step": 2180 }, { "grad_norm": 1.0424203872680664, "learning_rate": 4.3780000000000004e-05, "loss": 0.1005, "step": 2190 }, { "grad_norm": 1.0165292024612427, "learning_rate": 4.398e-05, "loss": 0.1015, "step": 2200 }, { "grad_norm": 1.0463976860046387, "learning_rate": 4.418000000000001e-05, "loss": 0.0874, "step": 2210 }, { "grad_norm": 0.923625111579895, "learning_rate": 4.438e-05, "loss": 0.0881, "step": 2220 }, { "grad_norm": 0.9913981556892395, "learning_rate": 4.458e-05, "loss": 0.0928, "step": 2230 }, { "grad_norm": 0.8590376973152161, "learning_rate": 4.478e-05, "loss": 0.1037, "step": 2240 }, { "grad_norm": 0.9148945808410645, "learning_rate": 4.498e-05, "loss": 0.1027, "step": 2250 }, { "grad_norm": 1.0050692558288574, "learning_rate": 4.518e-05, "loss": 0.0906, "step": 2260 }, { "grad_norm": 0.9408490061759949, "learning_rate": 4.538e-05, "loss": 0.0869, "step": 2270 }, { "grad_norm": 1.0718791484832764, "learning_rate": 4.558e-05, "loss": 0.0991, "step": 2280 }, { "grad_norm": 1.0510295629501343, "learning_rate": 4.578e-05, "loss": 0.0946, "step": 2290 }, { "grad_norm": 0.961672306060791, "learning_rate": 4.5980000000000004e-05, "loss": 0.0997, "step": 2300 }, { "grad_norm": 1.186691403388977, "learning_rate": 4.618e-05, "loss": 0.0951, "step": 2310 }, { "grad_norm": 0.9263490438461304, "learning_rate": 4.638e-05, "loss": 0.0898, "step": 2320 }, { "grad_norm": 0.931897759437561, "learning_rate": 4.6580000000000005e-05, "loss": 0.1003, "step": 2330 }, { "grad_norm": 0.8543550968170166, "learning_rate": 4.678e-05, "loss": 0.1013, "step": 2340 }, { "grad_norm": 1.1699862480163574, "learning_rate": 4.698e-05, "loss": 0.102, "step": 2350 }, { "grad_norm": 0.8312776684761047, "learning_rate": 4.718e-05, "loss": 0.0956, "step": 2360 }, { "grad_norm": 0.8071889877319336, "learning_rate": 4.7380000000000004e-05, "loss": 0.0965, "step": 2370 }, { "grad_norm": 1.0148301124572754, "learning_rate": 4.758e-05, "loss": 0.0971, "step": 2380 }, { "grad_norm": 1.007031798362732, "learning_rate": 4.778e-05, "loss": 0.1089, "step": 2390 }, { "grad_norm": 1.1150444746017456, "learning_rate": 4.7980000000000005e-05, "loss": 0.1018, "step": 2400 }, { "grad_norm": 1.0318759679794312, "learning_rate": 4.818e-05, "loss": 0.1082, "step": 2410 }, { "grad_norm": 0.8206075429916382, "learning_rate": 4.838e-05, "loss": 0.1024, "step": 2420 }, { "grad_norm": 0.8261545300483704, "learning_rate": 4.8580000000000006e-05, "loss": 0.0921, "step": 2430 }, { "grad_norm": 1.106752634048462, "learning_rate": 4.8780000000000004e-05, "loss": 0.0977, "step": 2440 }, { "grad_norm": 0.8635814785957336, "learning_rate": 4.898e-05, "loss": 0.0935, "step": 2450 }, { "grad_norm": 0.8426365256309509, "learning_rate": 4.918000000000001e-05, "loss": 0.0883, "step": 2460 }, { "grad_norm": 0.8103430271148682, "learning_rate": 4.9380000000000005e-05, "loss": 0.0881, "step": 2470 }, { "grad_norm": 0.881854772567749, "learning_rate": 4.958e-05, "loss": 0.1038, "step": 2480 }, { "grad_norm": 0.9143291711807251, "learning_rate": 4.978e-05, "loss": 0.0949, "step": 2490 }, { "grad_norm": 0.8488584756851196, "learning_rate": 4.9980000000000006e-05, "loss": 0.0943, "step": 2500 }, { "grad_norm": 0.6950796246528625, "learning_rate": 5.0180000000000004e-05, "loss": 0.0828, "step": 2510 }, { "grad_norm": 0.6685064435005188, "learning_rate": 5.038e-05, "loss": 0.0862, "step": 2520 }, { "grad_norm": 0.9115505218505859, "learning_rate": 5.058000000000001e-05, "loss": 0.097, "step": 2530 }, { "grad_norm": 0.8384989500045776, "learning_rate": 5.0780000000000005e-05, "loss": 0.1018, "step": 2540 }, { "grad_norm": 0.6235048174858093, "learning_rate": 5.098e-05, "loss": 0.0894, "step": 2550 }, { "grad_norm": 0.8883695602416992, "learning_rate": 5.118000000000001e-05, "loss": 0.0919, "step": 2560 }, { "grad_norm": 0.9244877696037292, "learning_rate": 5.1380000000000006e-05, "loss": 0.0849, "step": 2570 }, { "grad_norm": 0.8085609674453735, "learning_rate": 5.1580000000000004e-05, "loss": 0.0945, "step": 2580 }, { "grad_norm": 1.0315029621124268, "learning_rate": 5.178000000000001e-05, "loss": 0.0927, "step": 2590 }, { "grad_norm": 0.8419883251190186, "learning_rate": 5.198000000000001e-05, "loss": 0.0964, "step": 2600 }, { "grad_norm": 0.9061105251312256, "learning_rate": 5.2180000000000005e-05, "loss": 0.0862, "step": 2610 }, { "grad_norm": 0.8982510566711426, "learning_rate": 5.238000000000001e-05, "loss": 0.0868, "step": 2620 }, { "grad_norm": 0.7257839441299438, "learning_rate": 5.258000000000001e-05, "loss": 0.0929, "step": 2630 }, { "grad_norm": 0.8872969150543213, "learning_rate": 5.2780000000000006e-05, "loss": 0.1023, "step": 2640 }, { "grad_norm": 1.0402151346206665, "learning_rate": 5.2980000000000004e-05, "loss": 0.0882, "step": 2650 }, { "grad_norm": 0.7583954334259033, "learning_rate": 5.318000000000001e-05, "loss": 0.0816, "step": 2660 }, { "grad_norm": 0.6525409817695618, "learning_rate": 5.338000000000001e-05, "loss": 0.0805, "step": 2670 }, { "grad_norm": 0.7373108863830566, "learning_rate": 5.3580000000000005e-05, "loss": 0.0804, "step": 2680 }, { "grad_norm": 0.6780685782432556, "learning_rate": 5.378e-05, "loss": 0.0843, "step": 2690 }, { "grad_norm": 0.5857313275337219, "learning_rate": 5.3979999999999995e-05, "loss": 0.0952, "step": 2700 }, { "grad_norm": 0.7565743327140808, "learning_rate": 5.418e-05, "loss": 0.086, "step": 2710 }, { "grad_norm": 1.053936243057251, "learning_rate": 5.438e-05, "loss": 0.0787, "step": 2720 }, { "grad_norm": 0.9540563821792603, "learning_rate": 5.4579999999999996e-05, "loss": 0.0869, "step": 2730 }, { "grad_norm": 0.7491376996040344, "learning_rate": 5.478e-05, "loss": 0.0777, "step": 2740 }, { "grad_norm": 0.5707305669784546, "learning_rate": 5.498e-05, "loss": 0.0749, "step": 2750 }, { "grad_norm": 0.7395896315574646, "learning_rate": 5.518e-05, "loss": 0.0799, "step": 2760 }, { "grad_norm": 0.8352436423301697, "learning_rate": 5.538e-05, "loss": 0.0762, "step": 2770 }, { "grad_norm": 0.615799069404602, "learning_rate": 5.558e-05, "loss": 0.0756, "step": 2780 }, { "grad_norm": 0.8265267610549927, "learning_rate": 5.578e-05, "loss": 0.0821, "step": 2790 }, { "grad_norm": 0.8592129349708557, "learning_rate": 5.5979999999999996e-05, "loss": 0.0793, "step": 2800 }, { "grad_norm": 0.6081315875053406, "learning_rate": 5.618e-05, "loss": 0.0811, "step": 2810 }, { "grad_norm": 0.8103339076042175, "learning_rate": 5.638e-05, "loss": 0.0846, "step": 2820 }, { "grad_norm": 1.1430416107177734, "learning_rate": 5.658e-05, "loss": 0.0774, "step": 2830 }, { "grad_norm": 0.9563974142074585, "learning_rate": 5.678e-05, "loss": 0.084, "step": 2840 }, { "grad_norm": 0.6825783252716064, "learning_rate": 5.698e-05, "loss": 0.086, "step": 2850 }, { "grad_norm": 0.9513275623321533, "learning_rate": 5.718e-05, "loss": 0.0977, "step": 2860 }, { "grad_norm": 0.676594078540802, "learning_rate": 5.738e-05, "loss": 0.0827, "step": 2870 }, { "grad_norm": 0.8023530840873718, "learning_rate": 5.758e-05, "loss": 0.0803, "step": 2880 }, { "grad_norm": 1.1440242528915405, "learning_rate": 5.778e-05, "loss": 0.0807, "step": 2890 }, { "grad_norm": 0.6456476449966431, "learning_rate": 5.7980000000000004e-05, "loss": 0.075, "step": 2900 }, { "grad_norm": 0.7342642545700073, "learning_rate": 5.818e-05, "loss": 0.087, "step": 2910 }, { "grad_norm": 0.7696200013160706, "learning_rate": 5.838e-05, "loss": 0.087, "step": 2920 }, { "grad_norm": 1.045058012008667, "learning_rate": 5.858e-05, "loss": 0.0855, "step": 2930 }, { "grad_norm": 0.6459250450134277, "learning_rate": 5.878e-05, "loss": 0.0787, "step": 2940 }, { "grad_norm": 1.0029280185699463, "learning_rate": 5.898e-05, "loss": 0.082, "step": 2950 }, { "grad_norm": 0.7057162523269653, "learning_rate": 5.918e-05, "loss": 0.0874, "step": 2960 }, { "grad_norm": 0.7591049671173096, "learning_rate": 5.9380000000000004e-05, "loss": 0.0876, "step": 2970 }, { "grad_norm": 0.690856397151947, "learning_rate": 5.958e-05, "loss": 0.0981, "step": 2980 }, { "grad_norm": 0.5918506383895874, "learning_rate": 5.978e-05, "loss": 0.0754, "step": 2990 }, { "grad_norm": 0.7166872024536133, "learning_rate": 5.9980000000000005e-05, "loss": 0.1003, "step": 3000 }, { "grad_norm": 0.6933836936950684, "learning_rate": 6.018e-05, "loss": 0.086, "step": 3010 }, { "grad_norm": 0.7374159693717957, "learning_rate": 6.038e-05, "loss": 0.0895, "step": 3020 }, { "grad_norm": 0.8100686073303223, "learning_rate": 6.0580000000000006e-05, "loss": 0.0881, "step": 3030 }, { "grad_norm": 0.8509901762008667, "learning_rate": 6.0780000000000004e-05, "loss": 0.0922, "step": 3040 }, { "grad_norm": 0.8186345100402832, "learning_rate": 6.098e-05, "loss": 0.0892, "step": 3050 }, { "grad_norm": 0.857949435710907, "learning_rate": 6.118000000000001e-05, "loss": 0.0807, "step": 3060 }, { "grad_norm": 0.6590606570243835, "learning_rate": 6.138e-05, "loss": 0.0882, "step": 3070 }, { "grad_norm": 0.6944153308868408, "learning_rate": 6.158e-05, "loss": 0.092, "step": 3080 }, { "grad_norm": 0.7946065664291382, "learning_rate": 6.178000000000001e-05, "loss": 0.084, "step": 3090 }, { "grad_norm": 1.2768237590789795, "learning_rate": 6.198e-05, "loss": 0.0903, "step": 3100 }, { "grad_norm": 0.7326115965843201, "learning_rate": 6.218e-05, "loss": 0.0898, "step": 3110 }, { "grad_norm": 0.5565741658210754, "learning_rate": 6.238000000000001e-05, "loss": 0.0799, "step": 3120 }, { "grad_norm": 0.7671065330505371, "learning_rate": 6.258e-05, "loss": 0.0862, "step": 3130 }, { "grad_norm": 0.9156190156936646, "learning_rate": 6.278e-05, "loss": 0.086, "step": 3140 }, { "grad_norm": 0.6571135520935059, "learning_rate": 6.298000000000001e-05, "loss": 0.0743, "step": 3150 }, { "grad_norm": 0.6871444582939148, "learning_rate": 6.318e-05, "loss": 0.0872, "step": 3160 }, { "grad_norm": 0.5972366333007812, "learning_rate": 6.338e-05, "loss": 0.0759, "step": 3170 }, { "grad_norm": 0.808833122253418, "learning_rate": 6.358000000000001e-05, "loss": 0.0874, "step": 3180 }, { "grad_norm": 0.7456704378128052, "learning_rate": 6.378e-05, "loss": 0.0949, "step": 3190 }, { "grad_norm": 0.7001237273216248, "learning_rate": 6.398000000000001e-05, "loss": 0.0817, "step": 3200 }, { "grad_norm": 0.8260010480880737, "learning_rate": 6.418000000000001e-05, "loss": 0.0857, "step": 3210 }, { "grad_norm": 0.64581698179245, "learning_rate": 6.438e-05, "loss": 0.0856, "step": 3220 }, { "grad_norm": 0.6626496911048889, "learning_rate": 6.458000000000001e-05, "loss": 0.0822, "step": 3230 }, { "grad_norm": 0.7366704344749451, "learning_rate": 6.478000000000001e-05, "loss": 0.0762, "step": 3240 }, { "grad_norm": 0.6865326166152954, "learning_rate": 6.498e-05, "loss": 0.0831, "step": 3250 }, { "grad_norm": 0.6987797021865845, "learning_rate": 6.518000000000001e-05, "loss": 0.0846, "step": 3260 }, { "grad_norm": 0.6453518867492676, "learning_rate": 6.538000000000001e-05, "loss": 0.0765, "step": 3270 }, { "grad_norm": 0.7633005976676941, "learning_rate": 6.558e-05, "loss": 0.0854, "step": 3280 }, { "grad_norm": 0.6999467015266418, "learning_rate": 6.578000000000001e-05, "loss": 0.0797, "step": 3290 }, { "grad_norm": 0.8009114861488342, "learning_rate": 6.598e-05, "loss": 0.0826, "step": 3300 }, { "grad_norm": 0.7196442484855652, "learning_rate": 6.618e-05, "loss": 0.0797, "step": 3310 }, { "grad_norm": 0.6590496301651001, "learning_rate": 6.638e-05, "loss": 0.0737, "step": 3320 }, { "grad_norm": 0.5808851718902588, "learning_rate": 6.658e-05, "loss": 0.084, "step": 3330 }, { "grad_norm": 0.7148438096046448, "learning_rate": 6.678e-05, "loss": 0.0882, "step": 3340 }, { "grad_norm": 0.6251967549324036, "learning_rate": 6.698e-05, "loss": 0.0817, "step": 3350 }, { "grad_norm": 0.552983820438385, "learning_rate": 6.718e-05, "loss": 0.0755, "step": 3360 }, { "grad_norm": 0.572636604309082, "learning_rate": 6.738e-05, "loss": 0.0754, "step": 3370 }, { "grad_norm": 0.5914811491966248, "learning_rate": 6.758e-05, "loss": 0.0782, "step": 3380 }, { "grad_norm": 0.8025500774383545, "learning_rate": 6.778e-05, "loss": 0.0865, "step": 3390 }, { "grad_norm": 0.7269631028175354, "learning_rate": 6.798e-05, "loss": 0.0967, "step": 3400 }, { "grad_norm": 0.606291651725769, "learning_rate": 6.818e-05, "loss": 0.0805, "step": 3410 }, { "grad_norm": 0.7778864502906799, "learning_rate": 6.838e-05, "loss": 0.079, "step": 3420 }, { "grad_norm": 0.6020950078964233, "learning_rate": 6.858e-05, "loss": 0.0852, "step": 3430 }, { "grad_norm": 0.6206157207489014, "learning_rate": 6.878e-05, "loss": 0.0772, "step": 3440 }, { "grad_norm": 0.8345757722854614, "learning_rate": 6.898e-05, "loss": 0.084, "step": 3450 }, { "grad_norm": 0.484793484210968, "learning_rate": 6.918e-05, "loss": 0.0714, "step": 3460 }, { "grad_norm": 0.6630068421363831, "learning_rate": 6.938e-05, "loss": 0.0803, "step": 3470 }, { "grad_norm": 0.7774220108985901, "learning_rate": 6.958e-05, "loss": 0.0743, "step": 3480 }, { "grad_norm": 0.6729553937911987, "learning_rate": 6.978e-05, "loss": 0.0812, "step": 3490 }, { "grad_norm": 0.8562501668930054, "learning_rate": 6.998e-05, "loss": 0.0748, "step": 3500 }, { "grad_norm": 0.7375395894050598, "learning_rate": 7.018e-05, "loss": 0.0883, "step": 3510 }, { "grad_norm": 0.6727158427238464, "learning_rate": 7.038e-05, "loss": 0.0762, "step": 3520 }, { "grad_norm": 0.6608932614326477, "learning_rate": 7.058e-05, "loss": 0.076, "step": 3530 }, { "grad_norm": 0.8893786072731018, "learning_rate": 7.078e-05, "loss": 0.0948, "step": 3540 }, { "grad_norm": 0.6602856516838074, "learning_rate": 7.098e-05, "loss": 0.0817, "step": 3550 }, { "grad_norm": 0.5520594120025635, "learning_rate": 7.118e-05, "loss": 0.0816, "step": 3560 }, { "grad_norm": 0.5546630024909973, "learning_rate": 7.138e-05, "loss": 0.0813, "step": 3570 }, { "grad_norm": 0.5587507486343384, "learning_rate": 7.158e-05, "loss": 0.0674, "step": 3580 }, { "grad_norm": 0.5648815631866455, "learning_rate": 7.178000000000001e-05, "loss": 0.0799, "step": 3590 }, { "grad_norm": 0.4752579629421234, "learning_rate": 7.198e-05, "loss": 0.0768, "step": 3600 }, { "grad_norm": 0.6604898571968079, "learning_rate": 7.218e-05, "loss": 0.0737, "step": 3610 }, { "grad_norm": 0.6930554509162903, "learning_rate": 7.238000000000001e-05, "loss": 0.0789, "step": 3620 }, { "grad_norm": 0.6759325861930847, "learning_rate": 7.258e-05, "loss": 0.0796, "step": 3630 }, { "grad_norm": 0.5716832280158997, "learning_rate": 7.278e-05, "loss": 0.0741, "step": 3640 }, { "grad_norm": 0.6348925232887268, "learning_rate": 7.298000000000001e-05, "loss": 0.0898, "step": 3650 }, { "grad_norm": 0.5529987215995789, "learning_rate": 7.318e-05, "loss": 0.0755, "step": 3660 }, { "grad_norm": 0.6864245533943176, "learning_rate": 7.338e-05, "loss": 0.0759, "step": 3670 }, { "grad_norm": 0.738946259021759, "learning_rate": 7.358000000000001e-05, "loss": 0.0828, "step": 3680 }, { "grad_norm": 0.63352370262146, "learning_rate": 7.378e-05, "loss": 0.0776, "step": 3690 }, { "grad_norm": 0.5932672619819641, "learning_rate": 7.398e-05, "loss": 0.0759, "step": 3700 }, { "grad_norm": 0.4575257897377014, "learning_rate": 7.418000000000001e-05, "loss": 0.0766, "step": 3710 }, { "grad_norm": 0.5841390490531921, "learning_rate": 7.438e-05, "loss": 0.0689, "step": 3720 }, { "grad_norm": 0.4735065996646881, "learning_rate": 7.458000000000001e-05, "loss": 0.0714, "step": 3730 }, { "grad_norm": 0.5657786130905151, "learning_rate": 7.478e-05, "loss": 0.0776, "step": 3740 }, { "grad_norm": 0.8493554592132568, "learning_rate": 7.498e-05, "loss": 0.08, "step": 3750 }, { "grad_norm": 0.6025906801223755, "learning_rate": 7.518000000000001e-05, "loss": 0.0744, "step": 3760 }, { "grad_norm": 0.6211271286010742, "learning_rate": 7.538e-05, "loss": 0.0826, "step": 3770 }, { "grad_norm": 0.6026765704154968, "learning_rate": 7.558e-05, "loss": 0.0737, "step": 3780 }, { "grad_norm": 0.6184554100036621, "learning_rate": 7.578000000000001e-05, "loss": 0.077, "step": 3790 }, { "grad_norm": 0.7196559309959412, "learning_rate": 7.598e-05, "loss": 0.0792, "step": 3800 }, { "grad_norm": 0.5573127865791321, "learning_rate": 7.618e-05, "loss": 0.0807, "step": 3810 }, { "grad_norm": 0.6073847413063049, "learning_rate": 7.638000000000001e-05, "loss": 0.0831, "step": 3820 }, { "grad_norm": 0.6834830045700073, "learning_rate": 7.658e-05, "loss": 0.0747, "step": 3830 }, { "grad_norm": 0.5097864270210266, "learning_rate": 7.678000000000001e-05, "loss": 0.0722, "step": 3840 }, { "grad_norm": 0.7749480605125427, "learning_rate": 7.698000000000001e-05, "loss": 0.078, "step": 3850 }, { "grad_norm": 1.1099562644958496, "learning_rate": 7.718e-05, "loss": 0.0761, "step": 3860 }, { "grad_norm": 0.7431067824363708, "learning_rate": 7.738000000000001e-05, "loss": 0.0849, "step": 3870 }, { "grad_norm": 0.6288900971412659, "learning_rate": 7.758000000000001e-05, "loss": 0.0966, "step": 3880 }, { "grad_norm": 0.6047532558441162, "learning_rate": 7.778e-05, "loss": 0.0726, "step": 3890 }, { "grad_norm": 0.6339138150215149, "learning_rate": 7.798000000000001e-05, "loss": 0.0837, "step": 3900 }, { "grad_norm": 0.7294380068778992, "learning_rate": 7.818000000000001e-05, "loss": 0.0764, "step": 3910 }, { "grad_norm": 0.7005537748336792, "learning_rate": 7.838e-05, "loss": 0.0801, "step": 3920 }, { "grad_norm": 0.6013553738594055, "learning_rate": 7.858000000000001e-05, "loss": 0.0801, "step": 3930 }, { "grad_norm": 0.4303536117076874, "learning_rate": 7.878e-05, "loss": 0.0832, "step": 3940 }, { "grad_norm": 0.593902587890625, "learning_rate": 7.897999999999999e-05, "loss": 0.0824, "step": 3950 }, { "grad_norm": 0.4634755551815033, "learning_rate": 7.918e-05, "loss": 0.0849, "step": 3960 }, { "grad_norm": 0.47628673911094666, "learning_rate": 7.938e-05, "loss": 0.0802, "step": 3970 }, { "grad_norm": 0.4862218201160431, "learning_rate": 7.958e-05, "loss": 0.077, "step": 3980 }, { "grad_norm": 0.7384225726127625, "learning_rate": 7.978e-05, "loss": 0.0758, "step": 3990 }, { "grad_norm": 0.5225085020065308, "learning_rate": 7.998e-05, "loss": 0.0825, "step": 4000 }, { "grad_norm": 0.6288861036300659, "learning_rate": 8.018e-05, "loss": 0.0751, "step": 4010 }, { "grad_norm": 0.5802978277206421, "learning_rate": 8.038e-05, "loss": 0.0783, "step": 4020 }, { "grad_norm": 0.4617580473423004, "learning_rate": 8.058e-05, "loss": 0.0785, "step": 4030 }, { "grad_norm": 0.554490327835083, "learning_rate": 8.078e-05, "loss": 0.069, "step": 4040 }, { "grad_norm": 0.561223030090332, "learning_rate": 8.098e-05, "loss": 0.0697, "step": 4050 }, { "grad_norm": 0.6351848244667053, "learning_rate": 8.118e-05, "loss": 0.0699, "step": 4060 }, { "grad_norm": 0.5925267338752747, "learning_rate": 8.138e-05, "loss": 0.0826, "step": 4070 }, { "grad_norm": 0.6076407432556152, "learning_rate": 8.158e-05, "loss": 0.0746, "step": 4080 }, { "grad_norm": 0.6873326301574707, "learning_rate": 8.178e-05, "loss": 0.0825, "step": 4090 }, { "grad_norm": 0.5837229490280151, "learning_rate": 8.198e-05, "loss": 0.073, "step": 4100 }, { "grad_norm": 0.47032517194747925, "learning_rate": 8.218e-05, "loss": 0.0787, "step": 4110 }, { "grad_norm": 0.5934879183769226, "learning_rate": 8.238000000000001e-05, "loss": 0.0764, "step": 4120 }, { "grad_norm": 0.5329772233963013, "learning_rate": 8.258e-05, "loss": 0.0741, "step": 4130 }, { "grad_norm": 0.530764639377594, "learning_rate": 8.278e-05, "loss": 0.0743, "step": 4140 }, { "grad_norm": 0.6409987807273865, "learning_rate": 8.298000000000001e-05, "loss": 0.0828, "step": 4150 }, { "grad_norm": 0.4797435998916626, "learning_rate": 8.318e-05, "loss": 0.0708, "step": 4160 }, { "grad_norm": 0.6627424955368042, "learning_rate": 8.338e-05, "loss": 0.0766, "step": 4170 }, { "grad_norm": 0.5560194849967957, "learning_rate": 8.358e-05, "loss": 0.0773, "step": 4180 }, { "grad_norm": 0.4889698624610901, "learning_rate": 8.378e-05, "loss": 0.0739, "step": 4190 }, { "grad_norm": 0.6080042719841003, "learning_rate": 8.398e-05, "loss": 0.0738, "step": 4200 }, { "grad_norm": 0.41038841009140015, "learning_rate": 8.418e-05, "loss": 0.0624, "step": 4210 }, { "grad_norm": 0.7000577449798584, "learning_rate": 8.438e-05, "loss": 0.0803, "step": 4220 }, { "grad_norm": 0.4373207986354828, "learning_rate": 8.458e-05, "loss": 0.0725, "step": 4230 }, { "grad_norm": 0.559677004814148, "learning_rate": 8.478e-05, "loss": 0.069, "step": 4240 }, { "grad_norm": 0.4664388597011566, "learning_rate": 8.498e-05, "loss": 0.0638, "step": 4250 }, { "grad_norm": 0.6189731359481812, "learning_rate": 8.518000000000001e-05, "loss": 0.0797, "step": 4260 }, { "grad_norm": 0.6943873167037964, "learning_rate": 8.538e-05, "loss": 0.073, "step": 4270 }, { "grad_norm": 0.5757395625114441, "learning_rate": 8.558e-05, "loss": 0.0719, "step": 4280 }, { "grad_norm": 0.6816046833992004, "learning_rate": 8.578000000000001e-05, "loss": 0.0698, "step": 4290 }, { "grad_norm": 0.6394875645637512, "learning_rate": 8.598e-05, "loss": 0.0766, "step": 4300 }, { "grad_norm": 0.47773632407188416, "learning_rate": 8.618e-05, "loss": 0.0676, "step": 4310 }, { "grad_norm": 0.5969908237457275, "learning_rate": 8.638000000000001e-05, "loss": 0.0671, "step": 4320 }, { "grad_norm": 0.6275129318237305, "learning_rate": 8.658e-05, "loss": 0.0671, "step": 4330 }, { "grad_norm": 0.47604578733444214, "learning_rate": 8.678e-05, "loss": 0.0817, "step": 4340 }, { "grad_norm": 0.5178322792053223, "learning_rate": 8.698000000000001e-05, "loss": 0.0736, "step": 4350 }, { "grad_norm": 0.5744819045066833, "learning_rate": 8.718e-05, "loss": 0.078, "step": 4360 }, { "grad_norm": 0.5818108916282654, "learning_rate": 8.738000000000001e-05, "loss": 0.0717, "step": 4370 }, { "grad_norm": 0.5232332348823547, "learning_rate": 8.758000000000001e-05, "loss": 0.0743, "step": 4380 }, { "grad_norm": 0.5728979110717773, "learning_rate": 8.778e-05, "loss": 0.069, "step": 4390 }, { "grad_norm": 0.5032474994659424, "learning_rate": 8.798000000000001e-05, "loss": 0.0742, "step": 4400 }, { "grad_norm": 0.47315868735313416, "learning_rate": 8.818000000000001e-05, "loss": 0.0639, "step": 4410 }, { "grad_norm": 0.5139799118041992, "learning_rate": 8.838e-05, "loss": 0.0706, "step": 4420 }, { "grad_norm": 0.5024729371070862, "learning_rate": 8.858000000000001e-05, "loss": 0.0685, "step": 4430 }, { "grad_norm": 0.620482325553894, "learning_rate": 8.878000000000001e-05, "loss": 0.0762, "step": 4440 }, { "grad_norm": 0.5298937559127808, "learning_rate": 8.898e-05, "loss": 0.0708, "step": 4450 }, { "grad_norm": 0.5584441423416138, "learning_rate": 8.918000000000001e-05, "loss": 0.0738, "step": 4460 }, { "grad_norm": 0.4228626787662506, "learning_rate": 8.938e-05, "loss": 0.0762, "step": 4470 }, { "grad_norm": 0.7000007033348083, "learning_rate": 8.958e-05, "loss": 0.0712, "step": 4480 }, { "grad_norm": 0.4099729359149933, "learning_rate": 8.978000000000001e-05, "loss": 0.0879, "step": 4490 }, { "grad_norm": 0.540330171585083, "learning_rate": 8.998e-05, "loss": 0.0784, "step": 4500 }, { "grad_norm": 0.48325517773628235, "learning_rate": 9.018000000000001e-05, "loss": 0.0835, "step": 4510 }, { "grad_norm": 0.5041128993034363, "learning_rate": 9.038000000000001e-05, "loss": 0.0786, "step": 4520 }, { "grad_norm": 0.5561689138412476, "learning_rate": 9.058e-05, "loss": 0.0713, "step": 4530 }, { "grad_norm": 0.598008930683136, "learning_rate": 9.078000000000001e-05, "loss": 0.0883, "step": 4540 }, { "grad_norm": 0.5965846180915833, "learning_rate": 9.098000000000001e-05, "loss": 0.0776, "step": 4550 }, { "grad_norm": 0.667057454586029, "learning_rate": 9.118e-05, "loss": 0.0734, "step": 4560 }, { "grad_norm": 0.4929996430873871, "learning_rate": 9.138e-05, "loss": 0.0727, "step": 4570 }, { "grad_norm": 0.8682854175567627, "learning_rate": 9.158e-05, "loss": 0.0694, "step": 4580 }, { "grad_norm": 0.45686888694763184, "learning_rate": 9.178e-05, "loss": 0.0781, "step": 4590 }, { "grad_norm": 0.6048151850700378, "learning_rate": 9.198e-05, "loss": 0.0683, "step": 4600 }, { "grad_norm": 0.715619683265686, "learning_rate": 9.218e-05, "loss": 0.0843, "step": 4610 }, { "grad_norm": 0.49714288115501404, "learning_rate": 9.238e-05, "loss": 0.0743, "step": 4620 }, { "grad_norm": 0.5853820443153381, "learning_rate": 9.258e-05, "loss": 0.0722, "step": 4630 }, { "grad_norm": 0.5438616275787354, "learning_rate": 9.278e-05, "loss": 0.0781, "step": 4640 }, { "grad_norm": 0.5904598236083984, "learning_rate": 9.298e-05, "loss": 0.0823, "step": 4650 }, { "grad_norm": 0.7669607400894165, "learning_rate": 9.318e-05, "loss": 0.0962, "step": 4660 }, { "grad_norm": 0.6656083464622498, "learning_rate": 9.338e-05, "loss": 0.0811, "step": 4670 }, { "grad_norm": 0.5444590449333191, "learning_rate": 9.358e-05, "loss": 0.0842, "step": 4680 }, { "grad_norm": 0.5830551981925964, "learning_rate": 9.378e-05, "loss": 0.0712, "step": 4690 }, { "grad_norm": 0.560207188129425, "learning_rate": 9.398e-05, "loss": 0.0744, "step": 4700 }, { "grad_norm": 0.49252140522003174, "learning_rate": 9.418e-05, "loss": 0.0767, "step": 4710 }, { "grad_norm": 0.4995827376842499, "learning_rate": 9.438e-05, "loss": 0.0836, "step": 4720 }, { "grad_norm": 0.5100601315498352, "learning_rate": 9.458e-05, "loss": 0.0702, "step": 4730 }, { "grad_norm": 0.537359893321991, "learning_rate": 9.478e-05, "loss": 0.0759, "step": 4740 }, { "grad_norm": 0.5443449020385742, "learning_rate": 9.498e-05, "loss": 0.0813, "step": 4750 }, { "grad_norm": 0.4861164689064026, "learning_rate": 9.518000000000001e-05, "loss": 0.0796, "step": 4760 }, { "grad_norm": 0.6012775897979736, "learning_rate": 9.538e-05, "loss": 0.0768, "step": 4770 }, { "grad_norm": 0.45128268003463745, "learning_rate": 9.558e-05, "loss": 0.0739, "step": 4780 }, { "grad_norm": 0.5162208080291748, "learning_rate": 9.578000000000001e-05, "loss": 0.0738, "step": 4790 }, { "grad_norm": 0.4127836525440216, "learning_rate": 9.598e-05, "loss": 0.0802, "step": 4800 }, { "grad_norm": 0.47622179985046387, "learning_rate": 9.618e-05, "loss": 0.0723, "step": 4810 }, { "grad_norm": 0.49518612027168274, "learning_rate": 9.638000000000001e-05, "loss": 0.0711, "step": 4820 }, { "grad_norm": 0.4334183931350708, "learning_rate": 9.658e-05, "loss": 0.0717, "step": 4830 }, { "grad_norm": 0.48778271675109863, "learning_rate": 9.678e-05, "loss": 0.0764, "step": 4840 }, { "grad_norm": 0.4829970598220825, "learning_rate": 9.698000000000001e-05, "loss": 0.0754, "step": 4850 }, { "grad_norm": 0.6221193671226501, "learning_rate": 9.718e-05, "loss": 0.0698, "step": 4860 }, { "grad_norm": 0.45151668787002563, "learning_rate": 9.738e-05, "loss": 0.0722, "step": 4870 }, { "grad_norm": 0.5216935276985168, "learning_rate": 9.758000000000001e-05, "loss": 0.0667, "step": 4880 }, { "grad_norm": 0.5287445783615112, "learning_rate": 9.778e-05, "loss": 0.0723, "step": 4890 }, { "grad_norm": 0.48619383573532104, "learning_rate": 9.798000000000001e-05, "loss": 0.0747, "step": 4900 }, { "grad_norm": 0.5160641670227051, "learning_rate": 9.818000000000001e-05, "loss": 0.0794, "step": 4910 }, { "grad_norm": 0.4473460912704468, "learning_rate": 9.838e-05, "loss": 0.0725, "step": 4920 }, { "grad_norm": 0.5076394081115723, "learning_rate": 9.858000000000001e-05, "loss": 0.0629, "step": 4930 }, { "grad_norm": 0.5090827941894531, "learning_rate": 9.878e-05, "loss": 0.0665, "step": 4940 }, { "grad_norm": 0.588241696357727, "learning_rate": 9.898e-05, "loss": 0.0793, "step": 4950 }, { "grad_norm": 0.4024425446987152, "learning_rate": 9.918000000000001e-05, "loss": 0.0682, "step": 4960 }, { "grad_norm": 0.6328859329223633, "learning_rate": 9.938e-05, "loss": 0.074, "step": 4970 }, { "grad_norm": 0.5322676301002502, "learning_rate": 9.958e-05, "loss": 0.0761, "step": 4980 }, { "grad_norm": 0.49438580870628357, "learning_rate": 9.978000000000001e-05, "loss": 0.0876, "step": 4990 }, { "grad_norm": 0.5291131734848022, "learning_rate": 9.998e-05, "loss": 0.0816, "step": 5000 }, { "grad_norm": 0.49052131175994873, "learning_rate": 9.999999778549045e-05, "loss": 0.069, "step": 5010 }, { "grad_norm": 0.5164645910263062, "learning_rate": 9.999999013039593e-05, "loss": 0.0811, "step": 5020 }, { "grad_norm": 0.4731025993824005, "learning_rate": 9.999997700737766e-05, "loss": 0.0801, "step": 5030 }, { "grad_norm": 0.38054606318473816, "learning_rate": 9.999995841643709e-05, "loss": 0.0696, "step": 5040 }, { "grad_norm": 0.49473893642425537, "learning_rate": 9.999993435757623e-05, "loss": 0.0625, "step": 5050 }, { "grad_norm": 0.51503586769104, "learning_rate": 9.999990483079773e-05, "loss": 0.0675, "step": 5060 }, { "grad_norm": 0.5018263459205627, "learning_rate": 9.999986983610481e-05, "loss": 0.0689, "step": 5070 }, { "grad_norm": 0.4313884973526001, "learning_rate": 9.99998293735013e-05, "loss": 0.0648, "step": 5080 }, { "grad_norm": 0.5139185786247253, "learning_rate": 9.999978344299161e-05, "loss": 0.0792, "step": 5090 }, { "grad_norm": 0.6292670369148254, "learning_rate": 9.99997320445808e-05, "loss": 0.088, "step": 5100 }, { "grad_norm": 0.5136967301368713, "learning_rate": 9.999967517827444e-05, "loss": 0.0774, "step": 5110 }, { "grad_norm": 0.5683696269989014, "learning_rate": 9.999961284407879e-05, "loss": 0.0784, "step": 5120 }, { "grad_norm": 0.632708728313446, "learning_rate": 9.999954504200067e-05, "loss": 0.0759, "step": 5130 }, { "grad_norm": 0.5204380750656128, "learning_rate": 9.999947177204744e-05, "loss": 0.0781, "step": 5140 }, { "grad_norm": 0.5112631320953369, "learning_rate": 9.999939303422718e-05, "loss": 0.0814, "step": 5150 }, { "grad_norm": 0.4644043743610382, "learning_rate": 9.999930882854847e-05, "loss": 0.0654, "step": 5160 }, { "grad_norm": 0.5312240123748779, "learning_rate": 9.999921915502051e-05, "loss": 0.0787, "step": 5170 }, { "grad_norm": 0.5331892371177673, "learning_rate": 9.99991240136531e-05, "loss": 0.071, "step": 5180 }, { "grad_norm": 0.5582438707351685, "learning_rate": 9.999902340445668e-05, "loss": 0.0821, "step": 5190 }, { "grad_norm": 0.5821518301963806, "learning_rate": 9.999891732744224e-05, "loss": 0.0817, "step": 5200 }, { "grad_norm": 0.49870991706848145, "learning_rate": 9.999880578262135e-05, "loss": 0.0822, "step": 5210 }, { "grad_norm": 0.587320864200592, "learning_rate": 9.999868877000624e-05, "loss": 0.0797, "step": 5220 }, { "grad_norm": 0.44917482137680054, "learning_rate": 9.99985662896097e-05, "loss": 0.0734, "step": 5230 }, { "grad_norm": 0.43927302956581116, "learning_rate": 9.999843834144513e-05, "loss": 0.0771, "step": 5240 }, { "grad_norm": 0.5609484314918518, "learning_rate": 9.99983049255265e-05, "loss": 0.0746, "step": 5250 }, { "grad_norm": 0.5167183876037598, "learning_rate": 9.999816604186843e-05, "loss": 0.0724, "step": 5260 }, { "grad_norm": 0.42886215448379517, "learning_rate": 9.999802169048609e-05, "loss": 0.073, "step": 5270 }, { "grad_norm": 0.6077436804771423, "learning_rate": 9.999787187139527e-05, "loss": 0.0622, "step": 5280 }, { "grad_norm": 0.9057908654212952, "learning_rate": 9.999771658461234e-05, "loss": 0.0841, "step": 5290 }, { "grad_norm": 0.49722081422805786, "learning_rate": 9.999755583015431e-05, "loss": 0.0733, "step": 5300 }, { "grad_norm": 0.5178150534629822, "learning_rate": 9.999738960803874e-05, "loss": 0.08, "step": 5310 }, { "grad_norm": 0.5769935846328735, "learning_rate": 9.99972179182838e-05, "loss": 0.0717, "step": 5320 }, { "grad_norm": 1.4073344469070435, "learning_rate": 9.99970407609083e-05, "loss": 0.0793, "step": 5330 }, { "grad_norm": 0.49528977274894714, "learning_rate": 9.999685813593159e-05, "loss": 0.0762, "step": 5340 }, { "grad_norm": 0.46771344542503357, "learning_rate": 9.999667004337362e-05, "loss": 0.0794, "step": 5350 }, { "grad_norm": 0.6613882780075073, "learning_rate": 9.9996476483255e-05, "loss": 0.0741, "step": 5360 }, { "grad_norm": 0.4034974277019501, "learning_rate": 9.999627745559688e-05, "loss": 0.0857, "step": 5370 }, { "grad_norm": 0.4304448664188385, "learning_rate": 9.999607296042101e-05, "loss": 0.0765, "step": 5380 }, { "grad_norm": 0.4914211332798004, "learning_rate": 9.99958629977498e-05, "loss": 0.0709, "step": 5390 }, { "grad_norm": 0.47826045751571655, "learning_rate": 9.999564756760615e-05, "loss": 0.0659, "step": 5400 }, { "grad_norm": 0.43430501222610474, "learning_rate": 9.999542667001366e-05, "loss": 0.076, "step": 5410 }, { "grad_norm": 0.4199322760105133, "learning_rate": 9.999520030499647e-05, "loss": 0.0691, "step": 5420 }, { "grad_norm": 0.5946688652038574, "learning_rate": 9.999496847257936e-05, "loss": 0.068, "step": 5430 }, { "grad_norm": 0.4285048842430115, "learning_rate": 9.999473117278764e-05, "loss": 0.0727, "step": 5440 }, { "grad_norm": 0.5303278565406799, "learning_rate": 9.999448840564731e-05, "loss": 0.0728, "step": 5450 }, { "grad_norm": 0.5558827519416809, "learning_rate": 9.999424017118488e-05, "loss": 0.0734, "step": 5460 }, { "grad_norm": 0.4261855483055115, "learning_rate": 9.999398646942751e-05, "loss": 0.0727, "step": 5470 }, { "grad_norm": 0.48034217953681946, "learning_rate": 9.999372730040296e-05, "loss": 0.0735, "step": 5480 }, { "grad_norm": 0.46086132526397705, "learning_rate": 9.999346266413953e-05, "loss": 0.0662, "step": 5490 }, { "grad_norm": 0.42237454652786255, "learning_rate": 9.99931925606662e-05, "loss": 0.0767, "step": 5500 }, { "grad_norm": 0.4877132177352905, "learning_rate": 9.99929169900125e-05, "loss": 0.0726, "step": 5510 }, { "grad_norm": 0.4656377136707306, "learning_rate": 9.999263595220855e-05, "loss": 0.0761, "step": 5520 }, { "grad_norm": 0.5220670700073242, "learning_rate": 9.99923494472851e-05, "loss": 0.0719, "step": 5530 }, { "grad_norm": 0.4505177438259125, "learning_rate": 9.999205747527348e-05, "loss": 0.0629, "step": 5540 }, { "grad_norm": 0.4189141094684601, "learning_rate": 9.999176003620561e-05, "loss": 0.0742, "step": 5550 }, { "grad_norm": 0.5175735950469971, "learning_rate": 9.999145713011405e-05, "loss": 0.0719, "step": 5560 }, { "grad_norm": 0.5476521253585815, "learning_rate": 9.999114875703186e-05, "loss": 0.0739, "step": 5570 }, { "grad_norm": 0.6168053150177002, "learning_rate": 9.999083491699281e-05, "loss": 0.0832, "step": 5580 }, { "grad_norm": 0.5285096764564514, "learning_rate": 9.999051561003123e-05, "loss": 0.0707, "step": 5590 }, { "grad_norm": 0.44296127557754517, "learning_rate": 9.999019083618202e-05, "loss": 0.071, "step": 5600 }, { "grad_norm": 0.45953911542892456, "learning_rate": 9.99898605954807e-05, "loss": 0.0637, "step": 5610 }, { "grad_norm": 0.4203157424926758, "learning_rate": 9.998952488796338e-05, "loss": 0.0631, "step": 5620 }, { "grad_norm": 0.41872289776802063, "learning_rate": 9.998918371366676e-05, "loss": 0.0627, "step": 5630 }, { "grad_norm": 0.43416154384613037, "learning_rate": 9.99888370726282e-05, "loss": 0.0646, "step": 5640 }, { "grad_norm": 0.5541588068008423, "learning_rate": 9.998848496488556e-05, "loss": 0.0692, "step": 5650 }, { "grad_norm": 0.4368991255760193, "learning_rate": 9.998812739047736e-05, "loss": 0.0648, "step": 5660 }, { "grad_norm": 0.5708600878715515, "learning_rate": 9.99877643494427e-05, "loss": 0.0752, "step": 5670 }, { "grad_norm": 0.44752970337867737, "learning_rate": 9.998739584182128e-05, "loss": 0.0638, "step": 5680 }, { "grad_norm": 0.4353484511375427, "learning_rate": 9.998702186765342e-05, "loss": 0.064, "step": 5690 }, { "grad_norm": 0.5218685865402222, "learning_rate": 9.998664242698e-05, "loss": 0.0627, "step": 5700 }, { "grad_norm": 0.38261738419532776, "learning_rate": 9.998625751984251e-05, "loss": 0.0644, "step": 5710 }, { "grad_norm": 0.3583544194698334, "learning_rate": 9.998586714628307e-05, "loss": 0.0602, "step": 5720 }, { "grad_norm": 0.36457309126853943, "learning_rate": 9.998547130634432e-05, "loss": 0.0662, "step": 5730 }, { "grad_norm": 0.3909115195274353, "learning_rate": 9.99850700000696e-05, "loss": 0.0643, "step": 5740 }, { "grad_norm": 0.4612206816673279, "learning_rate": 9.998466322750278e-05, "loss": 0.0658, "step": 5750 }, { "grad_norm": 0.37122273445129395, "learning_rate": 9.998425098868834e-05, "loss": 0.0577, "step": 5760 }, { "grad_norm": 0.36136189103126526, "learning_rate": 9.998383328367136e-05, "loss": 0.0705, "step": 5770 }, { "grad_norm": 0.5161547064781189, "learning_rate": 9.99834101124975e-05, "loss": 0.0586, "step": 5780 }, { "grad_norm": 0.372330904006958, "learning_rate": 9.998298147521309e-05, "loss": 0.0626, "step": 5790 }, { "grad_norm": 0.42282921075820923, "learning_rate": 9.998254737186496e-05, "loss": 0.0593, "step": 5800 }, { "grad_norm": 0.42929962277412415, "learning_rate": 9.99821078025006e-05, "loss": 0.06, "step": 5810 }, { "grad_norm": 0.5005477666854858, "learning_rate": 9.998166276716807e-05, "loss": 0.0633, "step": 5820 }, { "grad_norm": 0.49574220180511475, "learning_rate": 9.998121226591606e-05, "loss": 0.0622, "step": 5830 }, { "grad_norm": 0.4102029800415039, "learning_rate": 9.998075629879382e-05, "loss": 0.0585, "step": 5840 }, { "grad_norm": 0.45708122849464417, "learning_rate": 9.99802948658512e-05, "loss": 0.0663, "step": 5850 }, { "grad_norm": 0.40234240889549255, "learning_rate": 9.99798279671387e-05, "loss": 0.0695, "step": 5860 }, { "grad_norm": 0.35180872678756714, "learning_rate": 9.997935560270734e-05, "loss": 0.0764, "step": 5870 }, { "grad_norm": 0.4983593225479126, "learning_rate": 9.997887777260879e-05, "loss": 0.059, "step": 5880 }, { "grad_norm": 0.5685657262802124, "learning_rate": 9.997839447689532e-05, "loss": 0.0697, "step": 5890 }, { "grad_norm": 0.38514840602874756, "learning_rate": 9.997790571561978e-05, "loss": 0.0656, "step": 5900 }, { "grad_norm": 0.3824426531791687, "learning_rate": 9.99774114888356e-05, "loss": 0.0635, "step": 5910 }, { "grad_norm": 0.5508184432983398, "learning_rate": 9.997691179659684e-05, "loss": 0.0736, "step": 5920 }, { "grad_norm": 0.4454099237918854, "learning_rate": 9.997640663895815e-05, "loss": 0.0602, "step": 5930 }, { "grad_norm": 0.37570223212242126, "learning_rate": 9.997589601597477e-05, "loss": 0.0579, "step": 5940 }, { "grad_norm": 0.42120519280433655, "learning_rate": 9.997537992770252e-05, "loss": 0.0585, "step": 5950 }, { "grad_norm": 0.3838179409503937, "learning_rate": 9.997485837419788e-05, "loss": 0.0656, "step": 5960 }, { "grad_norm": 0.6030815839767456, "learning_rate": 9.997433135551786e-05, "loss": 0.0658, "step": 5970 }, { "grad_norm": 0.48595139384269714, "learning_rate": 9.997379887172009e-05, "loss": 0.0621, "step": 5980 }, { "grad_norm": 0.37119898200035095, "learning_rate": 9.997326092286281e-05, "loss": 0.067, "step": 5990 }, { "grad_norm": 0.4046497642993927, "learning_rate": 9.997271750900486e-05, "loss": 0.0645, "step": 6000 }, { "grad_norm": 0.5719247460365295, "learning_rate": 9.997216863020565e-05, "loss": 0.0798, "step": 6010 }, { "grad_norm": 0.38377225399017334, "learning_rate": 9.99716142865252e-05, "loss": 0.0636, "step": 6020 }, { "grad_norm": 0.4321150481700897, "learning_rate": 9.997105447802415e-05, "loss": 0.0606, "step": 6030 }, { "grad_norm": 0.40192341804504395, "learning_rate": 9.997048920476373e-05, "loss": 0.0675, "step": 6040 }, { "grad_norm": 0.40328335762023926, "learning_rate": 9.996991846680572e-05, "loss": 0.0641, "step": 6050 }, { "grad_norm": 0.4530247449874878, "learning_rate": 9.996934226421257e-05, "loss": 0.0603, "step": 6060 }, { "grad_norm": 0.4035988748073578, "learning_rate": 9.996876059704726e-05, "loss": 0.0636, "step": 6070 }, { "grad_norm": 0.43289703130722046, "learning_rate": 9.996817346537343e-05, "loss": 0.0619, "step": 6080 }, { "grad_norm": 0.5130596160888672, "learning_rate": 9.996758086925526e-05, "loss": 0.0699, "step": 6090 }, { "grad_norm": 0.38367196917533875, "learning_rate": 9.996698280875759e-05, "loss": 0.0588, "step": 6100 }, { "grad_norm": 0.41483399271965027, "learning_rate": 9.99663792839458e-05, "loss": 0.0642, "step": 6110 }, { "grad_norm": 0.33656588196754456, "learning_rate": 9.99657702948859e-05, "loss": 0.0646, "step": 6120 }, { "grad_norm": 0.3918236792087555, "learning_rate": 9.996515584164448e-05, "loss": 0.0628, "step": 6130 }, { "grad_norm": 0.3430735170841217, "learning_rate": 9.996453592428873e-05, "loss": 0.0737, "step": 6140 }, { "grad_norm": 0.36451664566993713, "learning_rate": 9.996391054288646e-05, "loss": 0.0606, "step": 6150 }, { "grad_norm": 0.453954815864563, "learning_rate": 9.996327969750605e-05, "loss": 0.068, "step": 6160 }, { "grad_norm": 0.43986669182777405, "learning_rate": 9.996264338821649e-05, "loss": 0.0719, "step": 6170 }, { "grad_norm": 0.46557462215423584, "learning_rate": 9.996200161508735e-05, "loss": 0.0647, "step": 6180 }, { "grad_norm": 0.4129607081413269, "learning_rate": 9.996135437818885e-05, "loss": 0.0711, "step": 6190 }, { "grad_norm": 0.3876471519470215, "learning_rate": 9.996070167759175e-05, "loss": 0.062, "step": 6200 }, { "grad_norm": 0.5564759373664856, "learning_rate": 9.996004351336743e-05, "loss": 0.0759, "step": 6210 }, { "grad_norm": 0.4940526783466339, "learning_rate": 9.995937988558785e-05, "loss": 0.0632, "step": 6220 }, { "grad_norm": 0.5607624650001526, "learning_rate": 9.995871079432561e-05, "loss": 0.0727, "step": 6230 }, { "grad_norm": 0.424428254365921, "learning_rate": 9.995803623965389e-05, "loss": 0.0622, "step": 6240 }, { "grad_norm": 0.46570292115211487, "learning_rate": 9.995735622164641e-05, "loss": 0.0692, "step": 6250 }, { "grad_norm": 0.43021780252456665, "learning_rate": 9.995667074037758e-05, "loss": 0.0712, "step": 6260 }, { "grad_norm": 0.4543602466583252, "learning_rate": 9.995597979592232e-05, "loss": 0.0679, "step": 6270 }, { "grad_norm": 0.327990859746933, "learning_rate": 9.995528338835625e-05, "loss": 0.067, "step": 6280 }, { "grad_norm": 0.37201106548309326, "learning_rate": 9.995458151775547e-05, "loss": 0.0672, "step": 6290 }, { "grad_norm": 0.4097701609134674, "learning_rate": 9.995387418419677e-05, "loss": 0.0644, "step": 6300 }, { "grad_norm": 0.43241143226623535, "learning_rate": 9.99531613877575e-05, "loss": 0.0619, "step": 6310 }, { "grad_norm": 0.4897738993167877, "learning_rate": 9.995244312851559e-05, "loss": 0.06, "step": 6320 }, { "grad_norm": 0.49865829944610596, "learning_rate": 9.995171940654961e-05, "loss": 0.0626, "step": 6330 }, { "grad_norm": 0.4315635859966278, "learning_rate": 9.995099022193871e-05, "loss": 0.071, "step": 6340 }, { "grad_norm": 0.4546506106853485, "learning_rate": 9.995025557476261e-05, "loss": 0.0629, "step": 6350 }, { "grad_norm": 0.38422366976737976, "learning_rate": 9.994951546510165e-05, "loss": 0.0662, "step": 6360 }, { "grad_norm": 0.4476645290851593, "learning_rate": 9.994876989303679e-05, "loss": 0.0781, "step": 6370 }, { "grad_norm": 0.43022191524505615, "learning_rate": 9.994801885864955e-05, "loss": 0.0708, "step": 6380 }, { "grad_norm": 0.5919501185417175, "learning_rate": 9.994726236202205e-05, "loss": 0.0601, "step": 6390 }, { "grad_norm": 0.37695592641830444, "learning_rate": 9.994650040323704e-05, "loss": 0.0687, "step": 6400 }, { "grad_norm": 0.3000831604003906, "learning_rate": 9.994573298237784e-05, "loss": 0.0593, "step": 6410 }, { "grad_norm": 0.4135439693927765, "learning_rate": 9.994496009952837e-05, "loss": 0.0612, "step": 6420 }, { "grad_norm": 0.4417782425880432, "learning_rate": 9.994418175477316e-05, "loss": 0.0575, "step": 6430 }, { "grad_norm": 0.38231393694877625, "learning_rate": 9.994339794819733e-05, "loss": 0.0638, "step": 6440 }, { "grad_norm": 0.44975873827934265, "learning_rate": 9.994260867988658e-05, "loss": 0.0824, "step": 6450 }, { "grad_norm": 0.3625616431236267, "learning_rate": 9.994181394992723e-05, "loss": 0.0658, "step": 6460 }, { "grad_norm": 0.37617194652557373, "learning_rate": 9.994101375840618e-05, "loss": 0.0617, "step": 6470 }, { "grad_norm": 0.37499406933784485, "learning_rate": 9.994020810541098e-05, "loss": 0.0691, "step": 6480 }, { "grad_norm": 0.6404722929000854, "learning_rate": 9.99393969910297e-05, "loss": 0.071, "step": 6490 }, { "grad_norm": 0.44253119826316833, "learning_rate": 9.993858041535104e-05, "loss": 0.0666, "step": 6500 }, { "grad_norm": 0.46142610907554626, "learning_rate": 9.99377583784643e-05, "loss": 0.0729, "step": 6510 }, { "grad_norm": 0.4148353636264801, "learning_rate": 9.993693088045939e-05, "loss": 0.0678, "step": 6520 }, { "grad_norm": 0.49056822061538696, "learning_rate": 9.99360979214268e-05, "loss": 0.0666, "step": 6530 }, { "grad_norm": 0.4692450761795044, "learning_rate": 9.99352595014576e-05, "loss": 0.0692, "step": 6540 }, { "grad_norm": 0.44601237773895264, "learning_rate": 9.993441562064354e-05, "loss": 0.0673, "step": 6550 }, { "grad_norm": 0.484372079372406, "learning_rate": 9.993356627907685e-05, "loss": 0.0644, "step": 6560 }, { "grad_norm": 0.3588715195655823, "learning_rate": 9.99327114768504e-05, "loss": 0.0746, "step": 6570 }, { "grad_norm": 0.4095497131347656, "learning_rate": 9.99318512140577e-05, "loss": 0.0625, "step": 6580 }, { "grad_norm": 0.4535367786884308, "learning_rate": 9.993098549079284e-05, "loss": 0.0647, "step": 6590 }, { "grad_norm": 0.46141767501831055, "learning_rate": 9.993011430715047e-05, "loss": 0.0675, "step": 6600 }, { "grad_norm": 0.34798339009284973, "learning_rate": 9.992923766322586e-05, "loss": 0.0578, "step": 6610 }, { "grad_norm": 0.5391647815704346, "learning_rate": 9.99283555591149e-05, "loss": 0.0684, "step": 6620 }, { "grad_norm": 0.45257288217544556, "learning_rate": 9.992746799491404e-05, "loss": 0.0722, "step": 6630 }, { "grad_norm": 0.36936095356941223, "learning_rate": 9.992657497072033e-05, "loss": 0.0711, "step": 6640 }, { "grad_norm": 0.41869086027145386, "learning_rate": 9.992567648663147e-05, "loss": 0.0612, "step": 6650 }, { "grad_norm": 0.4438968002796173, "learning_rate": 9.992477254274568e-05, "loss": 0.0671, "step": 6660 }, { "grad_norm": 0.4635665714740753, "learning_rate": 9.992386313916183e-05, "loss": 0.0645, "step": 6670 }, { "grad_norm": 0.41569432616233826, "learning_rate": 9.992294827597934e-05, "loss": 0.0686, "step": 6680 }, { "grad_norm": 0.46623069047927856, "learning_rate": 9.992202795329831e-05, "loss": 0.0693, "step": 6690 }, { "grad_norm": 0.39485839009284973, "learning_rate": 9.992110217121936e-05, "loss": 0.0683, "step": 6700 }, { "grad_norm": 0.4322608411312103, "learning_rate": 9.992017092984372e-05, "loss": 0.0638, "step": 6710 }, { "grad_norm": 0.3310049772262573, "learning_rate": 9.991923422927326e-05, "loss": 0.0695, "step": 6720 }, { "grad_norm": 0.4363126754760742, "learning_rate": 9.991829206961037e-05, "loss": 0.0655, "step": 6730 }, { "grad_norm": 0.3317543864250183, "learning_rate": 9.991734445095813e-05, "loss": 0.0578, "step": 6740 }, { "grad_norm": 0.3861159682273865, "learning_rate": 9.991639137342015e-05, "loss": 0.0651, "step": 6750 }, { "grad_norm": 0.3667494058609009, "learning_rate": 9.991543283710064e-05, "loss": 0.0574, "step": 6760 }, { "grad_norm": 0.4360527992248535, "learning_rate": 9.991446884210445e-05, "loss": 0.0655, "step": 6770 }, { "grad_norm": 0.4662325978279114, "learning_rate": 9.9913499388537e-05, "loss": 0.0647, "step": 6780 }, { "grad_norm": 0.40767809748649597, "learning_rate": 9.99125244765043e-05, "loss": 0.0665, "step": 6790 }, { "grad_norm": 0.37115198373794556, "learning_rate": 9.991154410611296e-05, "loss": 0.0693, "step": 6800 }, { "grad_norm": 0.35370486974716187, "learning_rate": 9.99105582774702e-05, "loss": 0.0574, "step": 6810 }, { "grad_norm": 0.36919891834259033, "learning_rate": 9.990956699068384e-05, "loss": 0.0761, "step": 6820 }, { "grad_norm": 0.4010396897792816, "learning_rate": 9.990857024586224e-05, "loss": 0.0582, "step": 6830 }, { "grad_norm": 0.5102018713951111, "learning_rate": 9.990756804311446e-05, "loss": 0.0575, "step": 6840 }, { "grad_norm": 0.45484885573387146, "learning_rate": 9.990656038255006e-05, "loss": 0.0617, "step": 6850 }, { "grad_norm": 0.38484227657318115, "learning_rate": 9.990554726427926e-05, "loss": 0.0511, "step": 6860 }, { "grad_norm": 0.42205438017845154, "learning_rate": 9.990452868841284e-05, "loss": 0.0678, "step": 6870 }, { "grad_norm": 0.4089653491973877, "learning_rate": 9.99035046550622e-05, "loss": 0.0569, "step": 6880 }, { "grad_norm": 0.4647107422351837, "learning_rate": 9.99024751643393e-05, "loss": 0.0492, "step": 6890 }, { "grad_norm": 0.40133845806121826, "learning_rate": 9.990144021635677e-05, "loss": 0.0612, "step": 6900 }, { "grad_norm": 0.4160388708114624, "learning_rate": 9.990039981122775e-05, "loss": 0.0651, "step": 6910 }, { "grad_norm": 0.37411656975746155, "learning_rate": 9.989935394906602e-05, "loss": 0.0588, "step": 6920 }, { "grad_norm": 0.4458743929862976, "learning_rate": 9.989830262998598e-05, "loss": 0.0641, "step": 6930 }, { "grad_norm": 0.4522882103919983, "learning_rate": 9.989724585410259e-05, "loss": 0.0624, "step": 6940 }, { "grad_norm": 0.4230647385120392, "learning_rate": 9.989618362153139e-05, "loss": 0.0657, "step": 6950 }, { "grad_norm": 0.37606051564216614, "learning_rate": 9.989511593238859e-05, "loss": 0.0628, "step": 6960 }, { "grad_norm": 0.5020167827606201, "learning_rate": 9.98940427867909e-05, "loss": 0.066, "step": 6970 }, { "grad_norm": 0.3995010554790497, "learning_rate": 9.989296418485573e-05, "loss": 0.0569, "step": 6980 }, { "grad_norm": 0.3896755278110504, "learning_rate": 9.989188012670101e-05, "loss": 0.0573, "step": 6990 }, { "grad_norm": 0.4247087240219116, "learning_rate": 9.989079061244528e-05, "loss": 0.0608, "step": 7000 }, { "grad_norm": 0.4415290057659149, "learning_rate": 9.988969564220769e-05, "loss": 0.0689, "step": 7010 }, { "grad_norm": 0.556317925453186, "learning_rate": 9.988859521610801e-05, "loss": 0.0626, "step": 7020 }, { "grad_norm": 0.41079631447792053, "learning_rate": 9.988748933426656e-05, "loss": 0.0569, "step": 7030 }, { "grad_norm": 0.35834163427352905, "learning_rate": 9.988637799680428e-05, "loss": 0.0614, "step": 7040 }, { "grad_norm": 0.4347209334373474, "learning_rate": 9.98852612038427e-05, "loss": 0.0604, "step": 7050 }, { "grad_norm": 0.3844144642353058, "learning_rate": 9.988413895550397e-05, "loss": 0.0567, "step": 7060 }, { "grad_norm": 0.3297899663448334, "learning_rate": 9.98830112519108e-05, "loss": 0.0579, "step": 7070 }, { "grad_norm": 0.341901570558548, "learning_rate": 9.98818780931865e-05, "loss": 0.0579, "step": 7080 }, { "grad_norm": 0.4134485125541687, "learning_rate": 9.988073947945502e-05, "loss": 0.0583, "step": 7090 }, { "grad_norm": 0.452449768781662, "learning_rate": 9.987959541084087e-05, "loss": 0.0545, "step": 7100 }, { "grad_norm": 0.45959293842315674, "learning_rate": 9.987844588746915e-05, "loss": 0.0601, "step": 7110 }, { "grad_norm": 0.5529808402061462, "learning_rate": 9.987729090946558e-05, "loss": 0.0639, "step": 7120 }, { "grad_norm": 0.5394242405891418, "learning_rate": 9.987613047695647e-05, "loss": 0.0647, "step": 7130 }, { "grad_norm": 0.36505991220474243, "learning_rate": 9.987496459006871e-05, "loss": 0.0723, "step": 7140 }, { "grad_norm": 0.4494795799255371, "learning_rate": 9.987379324892982e-05, "loss": 0.0708, "step": 7150 }, { "grad_norm": 0.41829925775527954, "learning_rate": 9.987261645366788e-05, "loss": 0.0642, "step": 7160 }, { "grad_norm": 0.6050693988800049, "learning_rate": 9.987143420441158e-05, "loss": 0.0619, "step": 7170 }, { "grad_norm": 0.329700231552124, "learning_rate": 9.987024650129022e-05, "loss": 0.0553, "step": 7180 }, { "grad_norm": 0.34796223044395447, "learning_rate": 9.986905334443368e-05, "loss": 0.0592, "step": 7190 }, { "grad_norm": 0.35675036907196045, "learning_rate": 9.986785473397245e-05, "loss": 0.0598, "step": 7200 }, { "grad_norm": 0.47443249821662903, "learning_rate": 9.98666506700376e-05, "loss": 0.0653, "step": 7210 }, { "grad_norm": 0.3935351073741913, "learning_rate": 9.986544115276081e-05, "loss": 0.0591, "step": 7220 }, { "grad_norm": 0.48930996656417847, "learning_rate": 9.986422618227433e-05, "loss": 0.0609, "step": 7230 }, { "grad_norm": 0.4651105999946594, "learning_rate": 9.986300575871106e-05, "loss": 0.0691, "step": 7240 }, { "grad_norm": 0.39909592270851135, "learning_rate": 9.986177988220444e-05, "loss": 0.0607, "step": 7250 }, { "grad_norm": 0.5259432196617126, "learning_rate": 9.986054855288856e-05, "loss": 0.0747, "step": 7260 }, { "grad_norm": 0.33880752325057983, "learning_rate": 9.985931177089802e-05, "loss": 0.0592, "step": 7270 }, { "grad_norm": 0.5283288359642029, "learning_rate": 9.985806953636814e-05, "loss": 0.0574, "step": 7280 }, { "grad_norm": 0.4170963168144226, "learning_rate": 9.985682184943471e-05, "loss": 0.0657, "step": 7290 }, { "grad_norm": 0.39979690313339233, "learning_rate": 9.98555687102342e-05, "loss": 0.0583, "step": 7300 }, { "grad_norm": 0.4185498356819153, "learning_rate": 9.985431011890367e-05, "loss": 0.0594, "step": 7310 }, { "grad_norm": 0.46838605403900146, "learning_rate": 9.985304607558075e-05, "loss": 0.0597, "step": 7320 }, { "grad_norm": 0.4357512891292572, "learning_rate": 9.985177658040364e-05, "loss": 0.0563, "step": 7330 }, { "grad_norm": 0.3326265811920166, "learning_rate": 9.985050163351119e-05, "loss": 0.06, "step": 7340 }, { "grad_norm": 0.5809574723243713, "learning_rate": 9.984922123504286e-05, "loss": 0.0664, "step": 7350 }, { "grad_norm": 0.31531110405921936, "learning_rate": 9.984793538513862e-05, "loss": 0.0545, "step": 7360 }, { "grad_norm": 0.3842657208442688, "learning_rate": 9.984664408393912e-05, "loss": 0.0663, "step": 7370 }, { "grad_norm": 0.39964815974235535, "learning_rate": 9.984534733158556e-05, "loss": 0.0656, "step": 7380 }, { "grad_norm": 0.3942583501338959, "learning_rate": 9.984404512821977e-05, "loss": 0.0553, "step": 7390 }, { "grad_norm": 0.4093293845653534, "learning_rate": 9.984273747398411e-05, "loss": 0.0581, "step": 7400 }, { "grad_norm": 0.43302619457244873, "learning_rate": 9.984142436902165e-05, "loss": 0.0564, "step": 7410 }, { "grad_norm": 0.4145091772079468, "learning_rate": 9.984010581347596e-05, "loss": 0.0625, "step": 7420 }, { "grad_norm": 0.4393395483493805, "learning_rate": 9.983878180749121e-05, "loss": 0.0566, "step": 7430 }, { "grad_norm": 0.36055755615234375, "learning_rate": 9.983745235121222e-05, "loss": 0.0559, "step": 7440 }, { "grad_norm": 0.6840464472770691, "learning_rate": 9.983611744478438e-05, "loss": 0.0625, "step": 7450 }, { "grad_norm": 0.38546204566955566, "learning_rate": 9.983477708835365e-05, "loss": 0.0644, "step": 7460 }, { "grad_norm": 0.3674996495246887, "learning_rate": 9.983343128206664e-05, "loss": 0.0642, "step": 7470 }, { "grad_norm": 0.4194955825805664, "learning_rate": 9.983208002607049e-05, "loss": 0.0602, "step": 7480 }, { "grad_norm": 0.4573762118816376, "learning_rate": 9.9830723320513e-05, "loss": 0.0656, "step": 7490 }, { "grad_norm": 0.3648015558719635, "learning_rate": 9.982936116554254e-05, "loss": 0.0617, "step": 7500 }, { "grad_norm": 0.37464022636413574, "learning_rate": 9.982799356130803e-05, "loss": 0.063, "step": 7510 }, { "grad_norm": 0.40271928906440735, "learning_rate": 9.982662050795908e-05, "loss": 0.0651, "step": 7520 }, { "grad_norm": 0.33954358100891113, "learning_rate": 9.982524200564583e-05, "loss": 0.0544, "step": 7530 }, { "grad_norm": 0.4201747179031372, "learning_rate": 9.982385805451901e-05, "loss": 0.0576, "step": 7540 }, { "grad_norm": 0.4577139914035797, "learning_rate": 9.982246865472998e-05, "loss": 0.0613, "step": 7550 }, { "grad_norm": 0.4329332709312439, "learning_rate": 9.982107380643069e-05, "loss": 0.0613, "step": 7560 }, { "grad_norm": 0.5131918787956238, "learning_rate": 9.981967350977368e-05, "loss": 0.0661, "step": 7570 }, { "grad_norm": 0.41799235343933105, "learning_rate": 9.981826776491208e-05, "loss": 0.0586, "step": 7580 }, { "grad_norm": 0.35743099451065063, "learning_rate": 9.98168565719996e-05, "loss": 0.0651, "step": 7590 }, { "grad_norm": 0.3659997582435608, "learning_rate": 9.98154399311906e-05, "loss": 0.0506, "step": 7600 }, { "grad_norm": 0.3403373658657074, "learning_rate": 9.981401784263997e-05, "loss": 0.0643, "step": 7610 }, { "grad_norm": 0.3607451319694519, "learning_rate": 9.981259030650326e-05, "loss": 0.066, "step": 7620 }, { "grad_norm": 0.3998834788799286, "learning_rate": 9.981115732293655e-05, "loss": 0.0637, "step": 7630 }, { "grad_norm": 0.3928876519203186, "learning_rate": 9.980971889209659e-05, "loss": 0.0653, "step": 7640 }, { "grad_norm": 0.573489785194397, "learning_rate": 9.980827501414064e-05, "loss": 0.0618, "step": 7650 }, { "grad_norm": 0.30805692076683044, "learning_rate": 9.980682568922663e-05, "loss": 0.0523, "step": 7660 }, { "grad_norm": 0.4455167353153229, "learning_rate": 9.980537091751304e-05, "loss": 0.058, "step": 7670 }, { "grad_norm": 0.2866877615451813, "learning_rate": 9.980391069915897e-05, "loss": 0.057, "step": 7680 }, { "grad_norm": 0.3836933970451355, "learning_rate": 9.98024450343241e-05, "loss": 0.0664, "step": 7690 }, { "grad_norm": 0.4176088273525238, "learning_rate": 9.980097392316872e-05, "loss": 0.057, "step": 7700 }, { "grad_norm": 0.3973361849784851, "learning_rate": 9.97994973658537e-05, "loss": 0.061, "step": 7710 }, { "grad_norm": 0.3778476417064667, "learning_rate": 9.979801536254054e-05, "loss": 0.0667, "step": 7720 }, { "grad_norm": 0.4614599645137787, "learning_rate": 9.979652791339127e-05, "loss": 0.0612, "step": 7730 }, { "grad_norm": 0.3748883903026581, "learning_rate": 9.97950350185686e-05, "loss": 0.057, "step": 7740 }, { "grad_norm": 0.30991485714912415, "learning_rate": 9.979353667823574e-05, "loss": 0.056, "step": 7750 }, { "grad_norm": 0.38634443283081055, "learning_rate": 9.979203289255658e-05, "loss": 0.0545, "step": 7760 }, { "grad_norm": 0.37213900685310364, "learning_rate": 9.979052366169557e-05, "loss": 0.0656, "step": 7770 }, { "grad_norm": 0.4185262620449066, "learning_rate": 9.978900898581775e-05, "loss": 0.0584, "step": 7780 }, { "grad_norm": 0.48884809017181396, "learning_rate": 9.978748886508875e-05, "loss": 0.0666, "step": 7790 }, { "grad_norm": 0.411932110786438, "learning_rate": 9.978596329967484e-05, "loss": 0.0652, "step": 7800 }, { "grad_norm": 0.37519556283950806, "learning_rate": 9.978443228974284e-05, "loss": 0.0606, "step": 7810 }, { "grad_norm": 0.3020993769168854, "learning_rate": 9.978289583546015e-05, "loss": 0.0585, "step": 7820 }, { "grad_norm": 0.3633405268192291, "learning_rate": 9.978135393699484e-05, "loss": 0.0549, "step": 7830 }, { "grad_norm": 0.3701154291629791, "learning_rate": 9.977980659451548e-05, "loss": 0.0672, "step": 7840 }, { "grad_norm": 0.4059952199459076, "learning_rate": 9.977825380819135e-05, "loss": 0.0557, "step": 7850 }, { "grad_norm": 0.359036386013031, "learning_rate": 9.97766955781922e-05, "loss": 0.0509, "step": 7860 }, { "grad_norm": 0.40672725439071655, "learning_rate": 9.977513190468848e-05, "loss": 0.0617, "step": 7870 }, { "grad_norm": 0.31665658950805664, "learning_rate": 9.977356278785116e-05, "loss": 0.0629, "step": 7880 }, { "grad_norm": 0.3061903417110443, "learning_rate": 9.977198822785184e-05, "loss": 0.0578, "step": 7890 }, { "grad_norm": 0.33737248182296753, "learning_rate": 9.977040822486273e-05, "loss": 0.0577, "step": 7900 }, { "grad_norm": 0.404171347618103, "learning_rate": 9.97688227790566e-05, "loss": 0.0627, "step": 7910 }, { "grad_norm": 0.30413004755973816, "learning_rate": 9.976723189060684e-05, "loss": 0.0582, "step": 7920 }, { "grad_norm": 0.42284664511680603, "learning_rate": 9.976563555968742e-05, "loss": 0.0572, "step": 7930 }, { "grad_norm": 0.8420502543449402, "learning_rate": 9.976403378647292e-05, "loss": 0.066, "step": 7940 }, { "grad_norm": 0.407503604888916, "learning_rate": 9.97624265711385e-05, "loss": 0.0596, "step": 7950 }, { "grad_norm": 0.2892839014530182, "learning_rate": 9.976081391385993e-05, "loss": 0.054, "step": 7960 }, { "grad_norm": 0.503128170967102, "learning_rate": 9.975919581481356e-05, "loss": 0.0641, "step": 7970 }, { "grad_norm": 0.39427804946899414, "learning_rate": 9.975757227417634e-05, "loss": 0.0568, "step": 7980 }, { "grad_norm": 1.1973217725753784, "learning_rate": 9.975594329212586e-05, "loss": 0.0588, "step": 7990 }, { "grad_norm": 0.3240153193473816, "learning_rate": 9.97543088688402e-05, "loss": 0.0621, "step": 8000 }, { "grad_norm": 0.5876399874687195, "learning_rate": 9.975266900449814e-05, "loss": 0.057, "step": 8010 }, { "grad_norm": 0.4510999321937561, "learning_rate": 9.975102369927898e-05, "loss": 0.0664, "step": 8020 }, { "grad_norm": 0.3822314143180847, "learning_rate": 9.974937295336269e-05, "loss": 0.0609, "step": 8030 }, { "grad_norm": 0.38998278975486755, "learning_rate": 9.974771676692975e-05, "loss": 0.0638, "step": 8040 }, { "grad_norm": 0.39862462878227234, "learning_rate": 9.974605514016131e-05, "loss": 0.0725, "step": 8050 }, { "grad_norm": 0.47256729006767273, "learning_rate": 9.974438807323907e-05, "loss": 0.066, "step": 8060 }, { "grad_norm": 0.4301372468471527, "learning_rate": 9.974271556634535e-05, "loss": 0.0702, "step": 8070 }, { "grad_norm": 0.4797617495059967, "learning_rate": 9.974103761966302e-05, "loss": 0.0709, "step": 8080 }, { "grad_norm": 0.35936352610588074, "learning_rate": 9.973935423337563e-05, "loss": 0.0625, "step": 8090 }, { "grad_norm": 0.3770734369754791, "learning_rate": 9.973766540766722e-05, "loss": 0.0654, "step": 8100 }, { "grad_norm": 0.4340161979198456, "learning_rate": 9.97359711427225e-05, "loss": 0.0576, "step": 8110 }, { "grad_norm": 0.3264179825782776, "learning_rate": 9.973427143872677e-05, "loss": 0.0593, "step": 8120 }, { "grad_norm": 0.36904123425483704, "learning_rate": 9.973256629586589e-05, "loss": 0.0542, "step": 8130 }, { "grad_norm": 0.36021459102630615, "learning_rate": 9.973085571432632e-05, "loss": 0.0545, "step": 8140 }, { "grad_norm": 0.34750205278396606, "learning_rate": 9.972913969429513e-05, "loss": 0.0554, "step": 8150 }, { "grad_norm": 0.40453222393989563, "learning_rate": 9.972741823596e-05, "loss": 0.0596, "step": 8160 }, { "grad_norm": 0.3307398557662964, "learning_rate": 9.972569133950917e-05, "loss": 0.0593, "step": 8170 }, { "grad_norm": 0.4588780105113983, "learning_rate": 9.972395900513151e-05, "loss": 0.0577, "step": 8180 }, { "grad_norm": 0.41947469115257263, "learning_rate": 9.972222123301645e-05, "loss": 0.056, "step": 8190 }, { "grad_norm": 0.39160290360450745, "learning_rate": 9.972047802335403e-05, "loss": 0.0565, "step": 8200 }, { "grad_norm": 0.3781588077545166, "learning_rate": 9.971872937633488e-05, "loss": 0.0607, "step": 8210 }, { "grad_norm": 0.3147294521331787, "learning_rate": 9.971697529215024e-05, "loss": 0.0562, "step": 8220 }, { "grad_norm": 0.3678285777568817, "learning_rate": 9.971521577099192e-05, "loss": 0.0614, "step": 8230 }, { "grad_norm": 0.347555011510849, "learning_rate": 9.971345081305236e-05, "loss": 0.055, "step": 8240 }, { "grad_norm": 0.3524329960346222, "learning_rate": 9.971168041852456e-05, "loss": 0.0558, "step": 8250 }, { "grad_norm": 0.3922632038593292, "learning_rate": 9.970990458760215e-05, "loss": 0.0612, "step": 8260 }, { "grad_norm": 0.46013104915618896, "learning_rate": 9.970812332047929e-05, "loss": 0.0523, "step": 8270 }, { "grad_norm": 0.3995570242404938, "learning_rate": 9.97063366173508e-05, "loss": 0.0554, "step": 8280 }, { "grad_norm": 0.3885401487350464, "learning_rate": 9.970454447841207e-05, "loss": 0.0586, "step": 8290 }, { "grad_norm": 0.3982822597026825, "learning_rate": 9.970274690385909e-05, "loss": 0.0589, "step": 8300 }, { "grad_norm": 0.37244823575019836, "learning_rate": 9.970094389388844e-05, "loss": 0.0545, "step": 8310 }, { "grad_norm": 0.44553142786026, "learning_rate": 9.969913544869728e-05, "loss": 0.0622, "step": 8320 }, { "grad_norm": 0.33606547117233276, "learning_rate": 9.96973215684834e-05, "loss": 0.0619, "step": 8330 }, { "grad_norm": 0.4077901542186737, "learning_rate": 9.969550225344513e-05, "loss": 0.0619, "step": 8340 }, { "grad_norm": 0.3286632299423218, "learning_rate": 9.969367750378147e-05, "loss": 0.0578, "step": 8350 }, { "grad_norm": 0.2780546545982361, "learning_rate": 9.969184731969194e-05, "loss": 0.0508, "step": 8360 }, { "grad_norm": 0.401652991771698, "learning_rate": 9.96900117013767e-05, "loss": 0.0534, "step": 8370 }, { "grad_norm": 0.320396363735199, "learning_rate": 9.96881706490365e-05, "loss": 0.0561, "step": 8380 }, { "grad_norm": 0.39059528708457947, "learning_rate": 9.968632416287265e-05, "loss": 0.0545, "step": 8390 }, { "grad_norm": 0.33297061920166016, "learning_rate": 9.96844722430871e-05, "loss": 0.061, "step": 8400 }, { "grad_norm": 0.38531172275543213, "learning_rate": 9.968261488988235e-05, "loss": 0.0621, "step": 8410 }, { "grad_norm": 0.4005593955516815, "learning_rate": 9.968075210346155e-05, "loss": 0.0521, "step": 8420 }, { "grad_norm": 0.4273936152458191, "learning_rate": 9.967888388402839e-05, "loss": 0.0594, "step": 8430 }, { "grad_norm": 0.2787460386753082, "learning_rate": 9.967701023178717e-05, "loss": 0.0531, "step": 8440 }, { "grad_norm": 0.3735837936401367, "learning_rate": 9.967513114694282e-05, "loss": 0.0587, "step": 8450 }, { "grad_norm": 0.31081199645996094, "learning_rate": 9.967324662970079e-05, "loss": 0.0448, "step": 8460 }, { "grad_norm": 0.4096432626247406, "learning_rate": 9.96713566802672e-05, "loss": 0.0534, "step": 8470 }, { "grad_norm": 0.4099937081336975, "learning_rate": 9.966946129884873e-05, "loss": 0.0544, "step": 8480 }, { "grad_norm": 0.3584698438644409, "learning_rate": 9.966756048565265e-05, "loss": 0.063, "step": 8490 }, { "grad_norm": 0.4987274706363678, "learning_rate": 9.966565424088681e-05, "loss": 0.0543, "step": 8500 }, { "grad_norm": 0.3494637906551361, "learning_rate": 9.96637425647597e-05, "loss": 0.0543, "step": 8510 }, { "grad_norm": 0.3509569764137268, "learning_rate": 9.966182545748038e-05, "loss": 0.0529, "step": 8520 }, { "grad_norm": 0.3306320011615753, "learning_rate": 9.96599029192585e-05, "loss": 0.0504, "step": 8530 }, { "grad_norm": 0.5311315059661865, "learning_rate": 9.965797495030428e-05, "loss": 0.0564, "step": 8540 }, { "grad_norm": 0.42939823865890503, "learning_rate": 9.96560415508286e-05, "loss": 0.062, "step": 8550 }, { "grad_norm": 0.3830183744430542, "learning_rate": 9.965410272104286e-05, "loss": 0.0609, "step": 8560 }, { "grad_norm": 0.3717516362667084, "learning_rate": 9.96521584611591e-05, "loss": 0.0579, "step": 8570 }, { "grad_norm": 0.42564257979393005, "learning_rate": 9.965020877138994e-05, "loss": 0.0641, "step": 8580 }, { "grad_norm": 0.4262835383415222, "learning_rate": 9.964825365194861e-05, "loss": 0.0538, "step": 8590 }, { "grad_norm": 0.42182686924934387, "learning_rate": 9.96462931030489e-05, "loss": 0.0613, "step": 8600 }, { "grad_norm": 0.37526124715805054, "learning_rate": 9.96443271249052e-05, "loss": 0.0581, "step": 8610 }, { "grad_norm": 0.4280022382736206, "learning_rate": 9.964235571773255e-05, "loss": 0.063, "step": 8620 }, { "grad_norm": 0.5214548110961914, "learning_rate": 9.96403788817465e-05, "loss": 0.0687, "step": 8630 }, { "grad_norm": 0.5032354593276978, "learning_rate": 9.963839661716325e-05, "loss": 0.0604, "step": 8640 }, { "grad_norm": 0.3600229322910309, "learning_rate": 9.963640892419958e-05, "loss": 0.0593, "step": 8650 }, { "grad_norm": 0.4443100094795227, "learning_rate": 9.963441580307286e-05, "loss": 0.0557, "step": 8660 }, { "grad_norm": 0.38702094554901123, "learning_rate": 9.963241725400104e-05, "loss": 0.0528, "step": 8670 }, { "grad_norm": 0.4599396288394928, "learning_rate": 9.963041327720271e-05, "loss": 0.0569, "step": 8680 }, { "grad_norm": 0.394950270652771, "learning_rate": 9.962840387289697e-05, "loss": 0.0561, "step": 8690 }, { "grad_norm": 0.46822357177734375, "learning_rate": 9.962638904130363e-05, "loss": 0.0682, "step": 8700 }, { "grad_norm": 0.3647553324699402, "learning_rate": 9.962436878264298e-05, "loss": 0.0501, "step": 8710 }, { "grad_norm": 0.3525070548057556, "learning_rate": 9.962234309713598e-05, "loss": 0.0558, "step": 8720 }, { "grad_norm": 0.31531935930252075, "learning_rate": 9.962031198500414e-05, "loss": 0.0556, "step": 8730 }, { "grad_norm": 0.28145235776901245, "learning_rate": 9.961827544646958e-05, "loss": 0.0594, "step": 8740 }, { "grad_norm": 0.393627405166626, "learning_rate": 9.961623348175501e-05, "loss": 0.0541, "step": 8750 }, { "grad_norm": 0.626246988773346, "learning_rate": 9.961418609108377e-05, "loss": 0.0667, "step": 8760 }, { "grad_norm": 0.35444197058677673, "learning_rate": 9.961213327467971e-05, "loss": 0.068, "step": 8770 }, { "grad_norm": 0.45091378688812256, "learning_rate": 9.961007503276736e-05, "loss": 0.0565, "step": 8780 }, { "grad_norm": 0.38891953229904175, "learning_rate": 9.960801136557179e-05, "loss": 0.058, "step": 8790 }, { "grad_norm": 0.2874760329723358, "learning_rate": 9.960594227331866e-05, "loss": 0.0557, "step": 8800 }, { "grad_norm": 0.3940553665161133, "learning_rate": 9.960386775623429e-05, "loss": 0.0531, "step": 8810 }, { "grad_norm": 0.48804140090942383, "learning_rate": 9.96017878145455e-05, "loss": 0.0577, "step": 8820 }, { "grad_norm": 0.3730660080909729, "learning_rate": 9.959970244847977e-05, "loss": 0.0542, "step": 8830 }, { "grad_norm": 0.38784006237983704, "learning_rate": 9.959761165826518e-05, "loss": 0.068, "step": 8840 }, { "grad_norm": 0.3722481429576874, "learning_rate": 9.959551544413033e-05, "loss": 0.0517, "step": 8850 }, { "grad_norm": 0.3467041850090027, "learning_rate": 9.959341380630448e-05, "loss": 0.0507, "step": 8860 }, { "grad_norm": 0.26796266436576843, "learning_rate": 9.959130674501746e-05, "loss": 0.0567, "step": 8870 }, { "grad_norm": 0.4019233286380768, "learning_rate": 9.958919426049968e-05, "loss": 0.0541, "step": 8880 }, { "grad_norm": 0.3319242298603058, "learning_rate": 9.958707635298219e-05, "loss": 0.052, "step": 8890 }, { "grad_norm": 0.3713124394416809, "learning_rate": 9.958495302269657e-05, "loss": 0.0578, "step": 8900 }, { "grad_norm": 0.38597357273101807, "learning_rate": 9.958282426987503e-05, "loss": 0.0492, "step": 8910 }, { "grad_norm": 0.41201478242874146, "learning_rate": 9.95806900947504e-05, "loss": 0.0609, "step": 8920 }, { "grad_norm": 0.4098012447357178, "learning_rate": 9.957855049755604e-05, "loss": 0.0554, "step": 8930 }, { "grad_norm": 0.37132036685943604, "learning_rate": 9.957640547852593e-05, "loss": 0.0463, "step": 8940 }, { "grad_norm": 0.37500903010368347, "learning_rate": 9.957425503789466e-05, "loss": 0.0555, "step": 8950 }, { "grad_norm": 0.4248208701610565, "learning_rate": 9.957209917589738e-05, "loss": 0.0506, "step": 8960 }, { "grad_norm": 0.2705501317977905, "learning_rate": 9.956993789276987e-05, "loss": 0.0522, "step": 8970 }, { "grad_norm": 0.45251357555389404, "learning_rate": 9.956777118874847e-05, "loss": 0.0671, "step": 8980 }, { "grad_norm": 0.3289561867713928, "learning_rate": 9.956559906407016e-05, "loss": 0.0572, "step": 8990 }, { "grad_norm": 0.343601793050766, "learning_rate": 9.956342151897245e-05, "loss": 0.0505, "step": 9000 }, { "grad_norm": 0.31240299344062805, "learning_rate": 9.956123855369346e-05, "loss": 0.0512, "step": 9010 }, { "grad_norm": 0.40541768074035645, "learning_rate": 9.955905016847196e-05, "loss": 0.0538, "step": 9020 }, { "grad_norm": 0.40280023217201233, "learning_rate": 9.955685636354723e-05, "loss": 0.0531, "step": 9030 }, { "grad_norm": 0.30498039722442627, "learning_rate": 9.95546571391592e-05, "loss": 0.0546, "step": 9040 }, { "grad_norm": 0.31044188141822815, "learning_rate": 9.955245249554837e-05, "loss": 0.0506, "step": 9050 }, { "grad_norm": 0.4219100773334503, "learning_rate": 9.955024243295582e-05, "loss": 0.0542, "step": 9060 }, { "grad_norm": 0.4302344024181366, "learning_rate": 9.954802695162328e-05, "loss": 0.0623, "step": 9070 }, { "grad_norm": 0.3289872407913208, "learning_rate": 9.954580605179302e-05, "loss": 0.0503, "step": 9080 }, { "grad_norm": 0.3419932425022125, "learning_rate": 9.954357973370788e-05, "loss": 0.0488, "step": 9090 }, { "grad_norm": 0.31915414333343506, "learning_rate": 9.954134799761135e-05, "loss": 0.0458, "step": 9100 }, { "grad_norm": 0.6428102850914001, "learning_rate": 9.953911084374748e-05, "loss": 0.0458, "step": 9110 }, { "grad_norm": 0.38081488013267517, "learning_rate": 9.953686827236093e-05, "loss": 0.0532, "step": 9120 }, { "grad_norm": 0.5098811388015747, "learning_rate": 9.953462028369695e-05, "loss": 0.0576, "step": 9130 }, { "grad_norm": 0.28850704431533813, "learning_rate": 9.953236687800136e-05, "loss": 0.0544, "step": 9140 }, { "grad_norm": 0.3849986493587494, "learning_rate": 9.95301080555206e-05, "loss": 0.0606, "step": 9150 }, { "grad_norm": 0.37708503007888794, "learning_rate": 9.952784381650171e-05, "loss": 0.0568, "step": 9160 }, { "grad_norm": 0.42161130905151367, "learning_rate": 9.952557416119226e-05, "loss": 0.0577, "step": 9170 }, { "grad_norm": 0.28928565979003906, "learning_rate": 9.95232990898405e-05, "loss": 0.0586, "step": 9180 }, { "grad_norm": 0.3125385642051697, "learning_rate": 9.95210186026952e-05, "loss": 0.0542, "step": 9190 }, { "grad_norm": 0.3480173349380493, "learning_rate": 9.951873270000576e-05, "loss": 0.0499, "step": 9200 }, { "grad_norm": 0.47091707587242126, "learning_rate": 9.951644138202216e-05, "loss": 0.0535, "step": 9210 }, { "grad_norm": 0.35888275504112244, "learning_rate": 9.951414464899498e-05, "loss": 0.0626, "step": 9220 }, { "grad_norm": 0.42521342635154724, "learning_rate": 9.951184250117538e-05, "loss": 0.0599, "step": 9230 }, { "grad_norm": 0.39390936493873596, "learning_rate": 9.950953493881513e-05, "loss": 0.0553, "step": 9240 }, { "grad_norm": 0.44807568192481995, "learning_rate": 9.950722196216658e-05, "loss": 0.0577, "step": 9250 }, { "grad_norm": 0.41828906536102295, "learning_rate": 9.950490357148265e-05, "loss": 0.0625, "step": 9260 }, { "grad_norm": 0.4676493704319, "learning_rate": 9.950257976701692e-05, "loss": 0.0622, "step": 9270 }, { "grad_norm": 0.4388819634914398, "learning_rate": 9.950025054902348e-05, "loss": 0.0613, "step": 9280 }, { "grad_norm": 0.7130244970321655, "learning_rate": 9.949791591775706e-05, "loss": 0.0539, "step": 9290 }, { "grad_norm": 0.36238986253738403, "learning_rate": 9.949557587347298e-05, "loss": 0.0669, "step": 9300 }, { "grad_norm": 0.3555923402309418, "learning_rate": 9.949323041642713e-05, "loss": 0.0568, "step": 9310 }, { "grad_norm": 0.4042954444885254, "learning_rate": 9.949087954687602e-05, "loss": 0.0688, "step": 9320 }, { "grad_norm": 0.4245396554470062, "learning_rate": 9.948852326507672e-05, "loss": 0.0631, "step": 9330 }, { "grad_norm": 0.423776239156723, "learning_rate": 9.948616157128694e-05, "loss": 0.0655, "step": 9340 }, { "grad_norm": 0.4731197655200958, "learning_rate": 9.948379446576493e-05, "loss": 0.0633, "step": 9350 }, { "grad_norm": 0.33370241522789, "learning_rate": 9.948142194876952e-05, "loss": 0.0625, "step": 9360 }, { "grad_norm": 0.3168065845966339, "learning_rate": 9.947904402056024e-05, "loss": 0.055, "step": 9370 }, { "grad_norm": 0.39040201902389526, "learning_rate": 9.947666068139708e-05, "loss": 0.0583, "step": 9380 }, { "grad_norm": 0.28952959179878235, "learning_rate": 9.947427193154071e-05, "loss": 0.0573, "step": 9390 }, { "grad_norm": 0.40290120244026184, "learning_rate": 9.947187777125233e-05, "loss": 0.0674, "step": 9400 }, { "grad_norm": 0.4254664480686188, "learning_rate": 9.946947820079377e-05, "loss": 0.0531, "step": 9410 }, { "grad_norm": 0.3378146290779114, "learning_rate": 9.946707322042747e-05, "loss": 0.0595, "step": 9420 }, { "grad_norm": 0.38270312547683716, "learning_rate": 9.94646628304164e-05, "loss": 0.0513, "step": 9430 }, { "grad_norm": 0.3665335178375244, "learning_rate": 9.946224703102418e-05, "loss": 0.0562, "step": 9440 }, { "grad_norm": 0.332445353269577, "learning_rate": 9.945982582251498e-05, "loss": 0.0566, "step": 9450 }, { "grad_norm": 0.3769785761833191, "learning_rate": 9.94573992051536e-05, "loss": 0.0505, "step": 9460 }, { "grad_norm": 0.3929274082183838, "learning_rate": 9.94549671792054e-05, "loss": 0.0576, "step": 9470 }, { "grad_norm": 0.48017147183418274, "learning_rate": 9.945252974493635e-05, "loss": 0.0565, "step": 9480 }, { "grad_norm": 0.4505764842033386, "learning_rate": 9.9450086902613e-05, "loss": 0.0549, "step": 9490 }, { "grad_norm": 0.4742209315299988, "learning_rate": 9.944763865250248e-05, "loss": 0.0607, "step": 9500 }, { "grad_norm": 0.3717540502548218, "learning_rate": 9.944518499487254e-05, "loss": 0.0597, "step": 9510 }, { "grad_norm": 0.3240413963794708, "learning_rate": 9.944272592999151e-05, "loss": 0.0585, "step": 9520 }, { "grad_norm": 0.42559942603111267, "learning_rate": 9.94402614581283e-05, "loss": 0.0592, "step": 9530 }, { "grad_norm": 0.3681601881980896, "learning_rate": 9.943779157955244e-05, "loss": 0.0541, "step": 9540 }, { "grad_norm": 0.35682258009910583, "learning_rate": 9.943531629453403e-05, "loss": 0.0537, "step": 9550 }, { "grad_norm": 0.2830066382884979, "learning_rate": 9.943283560334375e-05, "loss": 0.0535, "step": 9560 }, { "grad_norm": 0.3406347334384918, "learning_rate": 9.943034950625288e-05, "loss": 0.0533, "step": 9570 }, { "grad_norm": 0.530053436756134, "learning_rate": 9.942785800353332e-05, "loss": 0.0581, "step": 9580 }, { "grad_norm": 0.291338175535202, "learning_rate": 9.942536109545751e-05, "loss": 0.0566, "step": 9590 }, { "grad_norm": 0.2755907475948334, "learning_rate": 9.942285878229853e-05, "loss": 0.0572, "step": 9600 }, { "grad_norm": 0.39689505100250244, "learning_rate": 9.942035106433001e-05, "loss": 0.0492, "step": 9610 }, { "grad_norm": 0.34307679533958435, "learning_rate": 9.94178379418262e-05, "loss": 0.0492, "step": 9620 }, { "grad_norm": 0.4505964517593384, "learning_rate": 9.941531941506194e-05, "loss": 0.0577, "step": 9630 }, { "grad_norm": 0.25173842906951904, "learning_rate": 9.941279548431263e-05, "loss": 0.052, "step": 9640 }, { "grad_norm": 0.37087535858154297, "learning_rate": 9.941026614985431e-05, "loss": 0.0532, "step": 9650 }, { "grad_norm": 0.31947892904281616, "learning_rate": 9.940773141196357e-05, "loss": 0.0494, "step": 9660 }, { "grad_norm": 0.4363892376422882, "learning_rate": 9.94051912709176e-05, "loss": 0.0533, "step": 9670 }, { "grad_norm": 0.3562085032463074, "learning_rate": 9.940264572699421e-05, "loss": 0.0504, "step": 9680 }, { "grad_norm": 0.38832229375839233, "learning_rate": 9.940009478047174e-05, "loss": 0.0501, "step": 9690 }, { "grad_norm": 0.389514684677124, "learning_rate": 9.939753843162918e-05, "loss": 0.0525, "step": 9700 }, { "grad_norm": 0.35366111993789673, "learning_rate": 9.939497668074609e-05, "loss": 0.043, "step": 9710 }, { "grad_norm": 0.4405096173286438, "learning_rate": 9.93924095281026e-05, "loss": 0.0527, "step": 9720 }, { "grad_norm": 0.3771911859512329, "learning_rate": 9.938983697397948e-05, "loss": 0.0603, "step": 9730 }, { "grad_norm": 0.3360302746295929, "learning_rate": 9.938725901865805e-05, "loss": 0.0538, "step": 9740 }, { "grad_norm": 0.33220499753952026, "learning_rate": 9.93846756624202e-05, "loss": 0.0537, "step": 9750 }, { "grad_norm": 0.30960074067115784, "learning_rate": 9.938208690554849e-05, "loss": 0.0531, "step": 9760 }, { "grad_norm": 0.5758510231971741, "learning_rate": 9.9379492748326e-05, "loss": 0.0552, "step": 9770 }, { "grad_norm": 0.33598682284355164, "learning_rate": 9.937689319103641e-05, "loss": 0.0555, "step": 9780 }, { "grad_norm": 0.36248984932899475, "learning_rate": 9.937428823396404e-05, "loss": 0.0531, "step": 9790 }, { "grad_norm": 0.3541453182697296, "learning_rate": 9.937167787739372e-05, "loss": 0.057, "step": 9800 }, { "grad_norm": 0.34269315004348755, "learning_rate": 9.936906212161095e-05, "loss": 0.0524, "step": 9810 }, { "grad_norm": 0.3698427081108093, "learning_rate": 9.936644096690176e-05, "loss": 0.0523, "step": 9820 }, { "grad_norm": 0.4159679710865021, "learning_rate": 9.936381441355282e-05, "loss": 0.0551, "step": 9830 }, { "grad_norm": 0.3193890154361725, "learning_rate": 9.936118246185136e-05, "loss": 0.0516, "step": 9840 }, { "grad_norm": 0.36785024404525757, "learning_rate": 9.935854511208518e-05, "loss": 0.0634, "step": 9850 }, { "grad_norm": 0.3873262107372284, "learning_rate": 9.935590236454272e-05, "loss": 0.0559, "step": 9860 }, { "grad_norm": 0.4535679519176483, "learning_rate": 9.935325421951298e-05, "loss": 0.0495, "step": 9870 }, { "grad_norm": 0.34267234802246094, "learning_rate": 9.935060067728557e-05, "loss": 0.0564, "step": 9880 }, { "grad_norm": 0.36288219690322876, "learning_rate": 9.934794173815067e-05, "loss": 0.0582, "step": 9890 }, { "grad_norm": 0.3602055609226227, "learning_rate": 9.934527740239906e-05, "loss": 0.0524, "step": 9900 }, { "grad_norm": 0.366064190864563, "learning_rate": 9.934260767032209e-05, "loss": 0.0576, "step": 9910 }, { "grad_norm": 0.4162465035915375, "learning_rate": 9.933993254221172e-05, "loss": 0.0528, "step": 9920 }, { "grad_norm": 0.3116911053657532, "learning_rate": 9.933725201836053e-05, "loss": 0.0486, "step": 9930 }, { "grad_norm": 0.428658127784729, "learning_rate": 9.933456609906162e-05, "loss": 0.0455, "step": 9940 }, { "grad_norm": 0.36864781379699707, "learning_rate": 9.933187478460875e-05, "loss": 0.0573, "step": 9950 }, { "grad_norm": 0.3546521067619324, "learning_rate": 9.93291780752962e-05, "loss": 0.0557, "step": 9960 }, { "grad_norm": 0.3045267164707184, "learning_rate": 9.932647597141893e-05, "loss": 0.0532, "step": 9970 }, { "grad_norm": 0.3815228343009949, "learning_rate": 9.932376847327239e-05, "loss": 0.0587, "step": 9980 }, { "grad_norm": 0.2592238783836365, "learning_rate": 9.932105558115268e-05, "loss": 0.0493, "step": 9990 }, { "grad_norm": 0.2887296974658966, "learning_rate": 9.931833729535651e-05, "loss": 0.0598, "step": 10000 }, { "grad_norm": 0.438918799161911, "learning_rate": 9.931561361618111e-05, "loss": 0.0552, "step": 10010 }, { "grad_norm": 0.36791232228279114, "learning_rate": 9.931288454392435e-05, "loss": 0.0611, "step": 10020 }, { "grad_norm": 0.3735913336277008, "learning_rate": 9.931015007888467e-05, "loss": 0.0518, "step": 10030 }, { "grad_norm": 0.3512902855873108, "learning_rate": 9.930741022136112e-05, "loss": 0.0549, "step": 10040 }, { "grad_norm": 0.3634640872478485, "learning_rate": 9.930466497165333e-05, "loss": 0.048, "step": 10050 }, { "grad_norm": 0.35280606150627136, "learning_rate": 9.93019143300615e-05, "loss": 0.0479, "step": 10060 }, { "grad_norm": 0.28536519408226013, "learning_rate": 9.929915829688644e-05, "loss": 0.0473, "step": 10070 }, { "grad_norm": 0.33448588848114014, "learning_rate": 9.929639687242955e-05, "loss": 0.0508, "step": 10080 }, { "grad_norm": 0.4158271849155426, "learning_rate": 9.929363005699281e-05, "loss": 0.0473, "step": 10090 }, { "grad_norm": 0.28215673565864563, "learning_rate": 9.92908578508788e-05, "loss": 0.0535, "step": 10100 }, { "grad_norm": 0.3381006121635437, "learning_rate": 9.928808025439069e-05, "loss": 0.0535, "step": 10110 }, { "grad_norm": 0.29944413900375366, "learning_rate": 9.928529726783223e-05, "loss": 0.0549, "step": 10120 }, { "grad_norm": 0.3460116982460022, "learning_rate": 9.928250889150774e-05, "loss": 0.0521, "step": 10130 }, { "grad_norm": 0.43959927558898926, "learning_rate": 9.92797151257222e-05, "loss": 0.0484, "step": 10140 }, { "grad_norm": 0.349239319562912, "learning_rate": 9.927691597078108e-05, "loss": 0.0556, "step": 10150 }, { "grad_norm": 0.4691343605518341, "learning_rate": 9.927411142699053e-05, "loss": 0.0519, "step": 10160 }, { "grad_norm": 0.32956650853157043, "learning_rate": 9.927130149465725e-05, "loss": 0.0553, "step": 10170 }, { "grad_norm": 0.4293186068534851, "learning_rate": 9.92684861740885e-05, "loss": 0.0536, "step": 10180 }, { "grad_norm": 0.3584590554237366, "learning_rate": 9.926566546559217e-05, "loss": 0.0562, "step": 10190 }, { "grad_norm": 0.35539090633392334, "learning_rate": 9.926283936947673e-05, "loss": 0.0571, "step": 10200 }, { "grad_norm": 0.3400117754936218, "learning_rate": 9.926000788605126e-05, "loss": 0.0531, "step": 10210 }, { "grad_norm": 0.2730989158153534, "learning_rate": 9.92571710156254e-05, "loss": 0.0604, "step": 10220 }, { "grad_norm": 0.529777467250824, "learning_rate": 9.925432875850936e-05, "loss": 0.0555, "step": 10230 }, { "grad_norm": 0.40689384937286377, "learning_rate": 9.925148111501396e-05, "loss": 0.0564, "step": 10240 }, { "grad_norm": 0.3635563254356384, "learning_rate": 9.924862808545066e-05, "loss": 0.0527, "step": 10250 }, { "grad_norm": 0.26151445508003235, "learning_rate": 9.924576967013141e-05, "loss": 0.0515, "step": 10260 }, { "grad_norm": 0.5215846300125122, "learning_rate": 9.924290586936887e-05, "loss": 0.055, "step": 10270 }, { "grad_norm": 0.3275161385536194, "learning_rate": 9.924003668347614e-05, "loss": 0.0538, "step": 10280 }, { "grad_norm": 0.35093438625335693, "learning_rate": 9.923716211276704e-05, "loss": 0.0506, "step": 10290 }, { "grad_norm": 0.3303266167640686, "learning_rate": 9.923428215755594e-05, "loss": 0.0527, "step": 10300 }, { "grad_norm": 0.4024588465690613, "learning_rate": 9.923139681815775e-05, "loss": 0.0498, "step": 10310 }, { "grad_norm": 0.37018975615501404, "learning_rate": 9.922850609488801e-05, "loss": 0.0529, "step": 10320 }, { "grad_norm": 0.44000720977783203, "learning_rate": 9.922560998806287e-05, "loss": 0.0502, "step": 10330 }, { "grad_norm": 0.32153457403182983, "learning_rate": 9.922270849799905e-05, "loss": 0.0471, "step": 10340 }, { "grad_norm": 0.39553794264793396, "learning_rate": 9.92198016250138e-05, "loss": 0.0541, "step": 10350 }, { "grad_norm": 0.3773152828216553, "learning_rate": 9.921688936942506e-05, "loss": 0.0498, "step": 10360 }, { "grad_norm": 0.2856157124042511, "learning_rate": 9.921397173155129e-05, "loss": 0.0573, "step": 10370 }, { "grad_norm": 0.47694915533065796, "learning_rate": 9.921104871171157e-05, "loss": 0.0514, "step": 10380 }, { "grad_norm": 0.31777656078338623, "learning_rate": 9.920812031022554e-05, "loss": 0.0476, "step": 10390 }, { "grad_norm": 0.29740849137306213, "learning_rate": 9.920518652741348e-05, "loss": 0.0479, "step": 10400 }, { "grad_norm": 0.9420305490493774, "learning_rate": 9.920224736359618e-05, "loss": 0.0497, "step": 10410 }, { "grad_norm": 0.439217746257782, "learning_rate": 9.91993028190951e-05, "loss": 0.0659, "step": 10420 }, { "grad_norm": 0.368869423866272, "learning_rate": 9.919635289423222e-05, "loss": 0.0625, "step": 10430 }, { "grad_norm": 0.31846436858177185, "learning_rate": 9.919339758933015e-05, "loss": 0.0685, "step": 10440 }, { "grad_norm": 0.41544386744499207, "learning_rate": 9.919043690471209e-05, "loss": 0.0548, "step": 10450 }, { "grad_norm": 0.32491689920425415, "learning_rate": 9.91874708407018e-05, "loss": 0.0588, "step": 10460 }, { "grad_norm": 0.46440020203590393, "learning_rate": 9.918449939762367e-05, "loss": 0.0521, "step": 10470 }, { "grad_norm": 0.3891432583332062, "learning_rate": 9.91815225758026e-05, "loss": 0.0492, "step": 10480 }, { "grad_norm": 0.29858192801475525, "learning_rate": 9.917854037556419e-05, "loss": 0.0475, "step": 10490 }, { "grad_norm": 0.30942875146865845, "learning_rate": 9.917555279723454e-05, "loss": 0.0592, "step": 10500 }, { "grad_norm": 0.432425856590271, "learning_rate": 9.917255984114036e-05, "loss": 0.0539, "step": 10510 }, { "grad_norm": 0.3122553527355194, "learning_rate": 9.916956150760896e-05, "loss": 0.0578, "step": 10520 }, { "grad_norm": 0.39529818296432495, "learning_rate": 9.916655779696826e-05, "loss": 0.051, "step": 10530 }, { "grad_norm": 0.3502283990383148, "learning_rate": 9.916354870954671e-05, "loss": 0.0562, "step": 10540 }, { "grad_norm": 0.35735732316970825, "learning_rate": 9.91605342456734e-05, "loss": 0.054, "step": 10550 }, { "grad_norm": 0.3163277208805084, "learning_rate": 9.915751440567795e-05, "loss": 0.0523, "step": 10560 }, { "grad_norm": 0.3215142786502838, "learning_rate": 9.915448918989066e-05, "loss": 0.0492, "step": 10570 }, { "grad_norm": 0.3831300735473633, "learning_rate": 9.915145859864232e-05, "loss": 0.0545, "step": 10580 }, { "grad_norm": 0.32038983702659607, "learning_rate": 9.914842263226437e-05, "loss": 0.0551, "step": 10590 }, { "grad_norm": 0.3798322379589081, "learning_rate": 9.914538129108882e-05, "loss": 0.0613, "step": 10600 }, { "grad_norm": 0.41126230359077454, "learning_rate": 9.914233457544825e-05, "loss": 0.0557, "step": 10610 }, { "grad_norm": 0.40996962785720825, "learning_rate": 9.913928248567586e-05, "loss": 0.0567, "step": 10620 }, { "grad_norm": 0.2811562120914459, "learning_rate": 9.913622502210542e-05, "loss": 0.0583, "step": 10630 }, { "grad_norm": 0.36554479598999023, "learning_rate": 9.913316218507128e-05, "loss": 0.0606, "step": 10640 }, { "grad_norm": 0.3829156458377838, "learning_rate": 9.91300939749084e-05, "loss": 0.0491, "step": 10650 }, { "grad_norm": 0.37858113646507263, "learning_rate": 9.91270203919523e-05, "loss": 0.0516, "step": 10660 }, { "grad_norm": 0.35023266077041626, "learning_rate": 9.912394143653912e-05, "loss": 0.0526, "step": 10670 }, { "grad_norm": 0.31600284576416016, "learning_rate": 9.912085710900555e-05, "loss": 0.0538, "step": 10680 }, { "grad_norm": 0.3380908668041229, "learning_rate": 9.911776740968892e-05, "loss": 0.0582, "step": 10690 }, { "grad_norm": 0.4055658280849457, "learning_rate": 9.911467233892709e-05, "loss": 0.0498, "step": 10700 }, { "grad_norm": 0.37367677688598633, "learning_rate": 9.911157189705853e-05, "loss": 0.0465, "step": 10710 }, { "grad_norm": 0.3253156840801239, "learning_rate": 9.910846608442229e-05, "loss": 0.0459, "step": 10720 }, { "grad_norm": 0.3773154318332672, "learning_rate": 9.910535490135805e-05, "loss": 0.0522, "step": 10730 }, { "grad_norm": 0.3290378749370575, "learning_rate": 9.910223834820603e-05, "loss": 0.0494, "step": 10740 }, { "grad_norm": 0.36831995844841003, "learning_rate": 9.909911642530703e-05, "loss": 0.0574, "step": 10750 }, { "grad_norm": 0.29535746574401855, "learning_rate": 9.909598913300249e-05, "loss": 0.0517, "step": 10760 }, { "grad_norm": 0.35525086522102356, "learning_rate": 9.909285647163438e-05, "loss": 0.0537, "step": 10770 }, { "grad_norm": 0.43262842297554016, "learning_rate": 9.908971844154531e-05, "loss": 0.0422, "step": 10780 }, { "grad_norm": 0.3244515061378479, "learning_rate": 9.908657504307843e-05, "loss": 0.0441, "step": 10790 }, { "grad_norm": 0.35472723841667175, "learning_rate": 9.908342627657751e-05, "loss": 0.0466, "step": 10800 }, { "grad_norm": 0.4416818618774414, "learning_rate": 9.908027214238689e-05, "loss": 0.0473, "step": 10810 }, { "grad_norm": 0.31951549649238586, "learning_rate": 9.90771126408515e-05, "loss": 0.043, "step": 10820 }, { "grad_norm": 0.3610607981681824, "learning_rate": 9.907394777231685e-05, "loss": 0.051, "step": 10830 }, { "grad_norm": 0.28672900795936584, "learning_rate": 9.907077753712905e-05, "loss": 0.0444, "step": 10840 }, { "grad_norm": 0.28586992621421814, "learning_rate": 9.906760193563482e-05, "loss": 0.0522, "step": 10850 }, { "grad_norm": 0.3585466146469116, "learning_rate": 9.906442096818139e-05, "loss": 0.051, "step": 10860 }, { "grad_norm": 0.4295330047607422, "learning_rate": 9.906123463511665e-05, "loss": 0.0558, "step": 10870 }, { "grad_norm": 0.6416710019111633, "learning_rate": 9.905804293678907e-05, "loss": 0.0528, "step": 10880 }, { "grad_norm": 0.3750714659690857, "learning_rate": 9.905484587354766e-05, "loss": 0.0556, "step": 10890 }, { "grad_norm": 0.35604533553123474, "learning_rate": 9.905164344574205e-05, "loss": 0.0512, "step": 10900 }, { "grad_norm": 0.3586248457431793, "learning_rate": 9.904843565372248e-05, "loss": 0.0529, "step": 10910 }, { "grad_norm": 0.3217654526233673, "learning_rate": 9.904522249783972e-05, "loss": 0.0473, "step": 10920 }, { "grad_norm": 0.6042903661727905, "learning_rate": 9.904200397844517e-05, "loss": 0.0572, "step": 10930 }, { "grad_norm": 0.3435654938220978, "learning_rate": 9.903878009589078e-05, "loss": 0.0533, "step": 10940 }, { "grad_norm": 0.38009995222091675, "learning_rate": 9.903555085052915e-05, "loss": 0.0559, "step": 10950 }, { "grad_norm": 0.29119786620140076, "learning_rate": 9.903231624271338e-05, "loss": 0.0523, "step": 10960 }, { "grad_norm": 0.40129542350769043, "learning_rate": 9.902907627279724e-05, "loss": 0.0547, "step": 10970 }, { "grad_norm": 0.32613998651504517, "learning_rate": 9.902583094113504e-05, "loss": 0.0558, "step": 10980 }, { "grad_norm": 0.35042068362236023, "learning_rate": 9.902258024808168e-05, "loss": 0.0502, "step": 10990 }, { "grad_norm": 0.28072649240493774, "learning_rate": 9.901932419399264e-05, "loss": 0.0471, "step": 11000 }, { "grad_norm": 0.3410891890525818, "learning_rate": 9.9016062779224e-05, "loss": 0.0493, "step": 11010 }, { "grad_norm": 0.5682991147041321, "learning_rate": 9.901279600413242e-05, "loss": 0.0567, "step": 11020 }, { "grad_norm": 0.36237284541130066, "learning_rate": 9.900952386907518e-05, "loss": 0.0562, "step": 11030 }, { "grad_norm": 0.3899892568588257, "learning_rate": 9.90062463744101e-05, "loss": 0.0545, "step": 11040 }, { "grad_norm": 0.2778593599796295, "learning_rate": 9.900296352049558e-05, "loss": 0.0527, "step": 11050 }, { "grad_norm": 0.3590244650840759, "learning_rate": 9.899967530769065e-05, "loss": 0.0502, "step": 11060 }, { "grad_norm": 0.3212517499923706, "learning_rate": 9.899638173635489e-05, "loss": 0.045, "step": 11070 }, { "grad_norm": 0.43239402770996094, "learning_rate": 9.899308280684849e-05, "loss": 0.0576, "step": 11080 }, { "grad_norm": 0.44585174322128296, "learning_rate": 9.898977851953222e-05, "loss": 0.0565, "step": 11090 }, { "grad_norm": 0.32268962264060974, "learning_rate": 9.898646887476741e-05, "loss": 0.0506, "step": 11100 }, { "grad_norm": 0.35716402530670166, "learning_rate": 9.898315387291603e-05, "loss": 0.0493, "step": 11110 }, { "grad_norm": 0.4529339373111725, "learning_rate": 9.89798335143406e-05, "loss": 0.0468, "step": 11120 }, { "grad_norm": 0.3428956866264343, "learning_rate": 9.897650779940419e-05, "loss": 0.0576, "step": 11130 }, { "grad_norm": 0.4063076078891754, "learning_rate": 9.897317672847054e-05, "loss": 0.0518, "step": 11140 }, { "grad_norm": 0.4661111533641815, "learning_rate": 9.89698403019039e-05, "loss": 0.0542, "step": 11150 }, { "grad_norm": 0.41088107228279114, "learning_rate": 9.896649852006917e-05, "loss": 0.0509, "step": 11160 }, { "grad_norm": 0.2934872806072235, "learning_rate": 9.896315138333177e-05, "loss": 0.057, "step": 11170 }, { "grad_norm": 0.2822530269622803, "learning_rate": 9.895979889205774e-05, "loss": 0.0479, "step": 11180 }, { "grad_norm": 0.3607785701751709, "learning_rate": 9.895644104661372e-05, "loss": 0.0591, "step": 11190 }, { "grad_norm": 0.3522959053516388, "learning_rate": 9.895307784736691e-05, "loss": 0.0468, "step": 11200 }, { "grad_norm": 0.4942750930786133, "learning_rate": 9.894970929468512e-05, "loss": 0.0544, "step": 11210 }, { "grad_norm": 0.28227683901786804, "learning_rate": 9.89463353889367e-05, "loss": 0.0511, "step": 11220 }, { "grad_norm": 0.44951051473617554, "learning_rate": 9.894295613049065e-05, "loss": 0.0515, "step": 11230 }, { "grad_norm": 0.3476425111293793, "learning_rate": 9.893957151971649e-05, "loss": 0.0482, "step": 11240 }, { "grad_norm": 0.334195077419281, "learning_rate": 9.893618155698436e-05, "loss": 0.0474, "step": 11250 }, { "grad_norm": 0.3692297041416168, "learning_rate": 9.8932786242665e-05, "loss": 0.0503, "step": 11260 }, { "grad_norm": 0.357715904712677, "learning_rate": 9.89293855771297e-05, "loss": 0.0495, "step": 11270 }, { "grad_norm": 0.3478684723377228, "learning_rate": 9.892597956075036e-05, "loss": 0.0492, "step": 11280 }, { "grad_norm": 0.4265323579311371, "learning_rate": 9.892256819389947e-05, "loss": 0.0512, "step": 11290 }, { "grad_norm": 0.41699057817459106, "learning_rate": 9.891915147695006e-05, "loss": 0.0594, "step": 11300 }, { "grad_norm": 0.43288496136665344, "learning_rate": 9.891572941027577e-05, "loss": 0.0676, "step": 11310 }, { "grad_norm": 0.3918743133544922, "learning_rate": 9.89123019942509e-05, "loss": 0.0576, "step": 11320 }, { "grad_norm": 0.4193892478942871, "learning_rate": 9.89088692292502e-05, "loss": 0.0483, "step": 11330 }, { "grad_norm": 0.5959249138832092, "learning_rate": 9.89054311156491e-05, "loss": 0.0575, "step": 11340 }, { "grad_norm": 0.3582562804222107, "learning_rate": 9.890198765382357e-05, "loss": 0.053, "step": 11350 }, { "grad_norm": 0.4078196883201599, "learning_rate": 9.889853884415021e-05, "loss": 0.0512, "step": 11360 }, { "grad_norm": 0.36776527762413025, "learning_rate": 9.889508468700614e-05, "loss": 0.0524, "step": 11370 }, { "grad_norm": 0.5208590626716614, "learning_rate": 9.889162518276915e-05, "loss": 0.0553, "step": 11380 }, { "grad_norm": 0.3768851161003113, "learning_rate": 9.888816033181752e-05, "loss": 0.0564, "step": 11390 }, { "grad_norm": 0.30350568890571594, "learning_rate": 9.888469013453018e-05, "loss": 0.0487, "step": 11400 }, { "grad_norm": 0.3035303056240082, "learning_rate": 9.888121459128663e-05, "loss": 0.0518, "step": 11410 }, { "grad_norm": 0.4010630249977112, "learning_rate": 9.887773370246693e-05, "loss": 0.0519, "step": 11420 }, { "grad_norm": 0.3512534499168396, "learning_rate": 9.887424746845177e-05, "loss": 0.0522, "step": 11430 }, { "grad_norm": 0.3094578981399536, "learning_rate": 9.887075588962239e-05, "loss": 0.0533, "step": 11440 }, { "grad_norm": 0.38356781005859375, "learning_rate": 9.88672589663606e-05, "loss": 0.0549, "step": 11450 }, { "grad_norm": 0.404192179441452, "learning_rate": 9.886375669904886e-05, "loss": 0.0597, "step": 11460 }, { "grad_norm": 0.4021691381931305, "learning_rate": 9.886024908807014e-05, "loss": 0.0539, "step": 11470 }, { "grad_norm": 0.32714757323265076, "learning_rate": 9.885673613380806e-05, "loss": 0.0529, "step": 11480 }, { "grad_norm": 0.34075039625167847, "learning_rate": 9.885321783664676e-05, "loss": 0.0555, "step": 11490 }, { "grad_norm": 0.42014336585998535, "learning_rate": 9.884969419697101e-05, "loss": 0.0546, "step": 11500 }, { "grad_norm": 0.47422903776168823, "learning_rate": 9.884616521516614e-05, "loss": 0.0543, "step": 11510 }, { "grad_norm": 0.3569459319114685, "learning_rate": 9.88426308916181e-05, "loss": 0.0468, "step": 11520 }, { "grad_norm": 0.29631251096725464, "learning_rate": 9.883909122671335e-05, "loss": 0.0497, "step": 11530 }, { "grad_norm": 0.4942767918109894, "learning_rate": 9.883554622083904e-05, "loss": 0.0535, "step": 11540 }, { "grad_norm": 0.3519213795661926, "learning_rate": 9.88319958743828e-05, "loss": 0.0481, "step": 11550 }, { "grad_norm": 0.31522172689437866, "learning_rate": 9.882844018773291e-05, "loss": 0.0527, "step": 11560 }, { "grad_norm": 0.35459214448928833, "learning_rate": 9.882487916127823e-05, "loss": 0.0538, "step": 11570 }, { "grad_norm": 0.30800169706344604, "learning_rate": 9.882131279540815e-05, "loss": 0.0574, "step": 11580 }, { "grad_norm": 0.3718252182006836, "learning_rate": 9.881774109051271e-05, "loss": 0.051, "step": 11590 }, { "grad_norm": 0.30248892307281494, "learning_rate": 9.881416404698252e-05, "loss": 0.0526, "step": 11600 }, { "grad_norm": 0.36020776629447937, "learning_rate": 9.881058166520873e-05, "loss": 0.0491, "step": 11610 }, { "grad_norm": 0.479646772146225, "learning_rate": 9.880699394558311e-05, "loss": 0.0497, "step": 11620 }, { "grad_norm": 0.33582639694213867, "learning_rate": 9.880340088849801e-05, "loss": 0.0563, "step": 11630 }, { "grad_norm": 0.3523387312889099, "learning_rate": 9.879980249434637e-05, "loss": 0.046, "step": 11640 }, { "grad_norm": 0.2775977551937103, "learning_rate": 9.879619876352168e-05, "loss": 0.0469, "step": 11650 }, { "grad_norm": 0.2686833143234253, "learning_rate": 9.879258969641809e-05, "loss": 0.0536, "step": 11660 }, { "grad_norm": 0.3107149004936218, "learning_rate": 9.878897529343023e-05, "loss": 0.0474, "step": 11670 }, { "grad_norm": 0.37658634781837463, "learning_rate": 9.878535555495338e-05, "loss": 0.0479, "step": 11680 }, { "grad_norm": 0.27436333894729614, "learning_rate": 9.87817304813834e-05, "loss": 0.0449, "step": 11690 }, { "grad_norm": 0.3169119358062744, "learning_rate": 9.877810007311671e-05, "loss": 0.0503, "step": 11700 }, { "grad_norm": 0.25256577134132385, "learning_rate": 9.877446433055035e-05, "loss": 0.045, "step": 11710 }, { "grad_norm": 0.2813498377799988, "learning_rate": 9.877082325408191e-05, "loss": 0.0539, "step": 11720 }, { "grad_norm": 0.3282819986343384, "learning_rate": 9.876717684410954e-05, "loss": 0.0423, "step": 11730 }, { "grad_norm": 0.37405121326446533, "learning_rate": 9.876352510103204e-05, "loss": 0.054, "step": 11740 }, { "grad_norm": 0.33197954297065735, "learning_rate": 9.875986802524875e-05, "loss": 0.0467, "step": 11750 }, { "grad_norm": 0.38919854164123535, "learning_rate": 9.87562056171596e-05, "loss": 0.0523, "step": 11760 }, { "grad_norm": 0.461404025554657, "learning_rate": 9.875253787716511e-05, "loss": 0.0477, "step": 11770 }, { "grad_norm": 0.28578391671180725, "learning_rate": 9.874886480566637e-05, "loss": 0.0458, "step": 11780 }, { "grad_norm": 0.3819754719734192, "learning_rate": 9.874518640306507e-05, "loss": 0.0483, "step": 11790 }, { "grad_norm": 0.46453985571861267, "learning_rate": 9.874150266976347e-05, "loss": 0.0562, "step": 11800 }, { "grad_norm": 0.28347161412239075, "learning_rate": 9.873781360616443e-05, "loss": 0.0515, "step": 11810 }, { "grad_norm": 0.38722383975982666, "learning_rate": 9.873411921267137e-05, "loss": 0.0536, "step": 11820 }, { "grad_norm": 0.39838796854019165, "learning_rate": 9.873041948968829e-05, "loss": 0.0459, "step": 11830 }, { "grad_norm": 0.4479532539844513, "learning_rate": 9.872671443761981e-05, "loss": 0.0556, "step": 11840 }, { "grad_norm": 0.31409913301467896, "learning_rate": 9.872300405687109e-05, "loss": 0.048, "step": 11850 }, { "grad_norm": 0.28073400259017944, "learning_rate": 9.871928834784792e-05, "loss": 0.0463, "step": 11860 }, { "grad_norm": 0.2592845857143402, "learning_rate": 9.871556731095661e-05, "loss": 0.0557, "step": 11870 }, { "grad_norm": 0.30166661739349365, "learning_rate": 9.871184094660411e-05, "loss": 0.0517, "step": 11880 }, { "grad_norm": 0.38797295093536377, "learning_rate": 9.870810925519791e-05, "loss": 0.0555, "step": 11890 }, { "grad_norm": 0.32074812054634094, "learning_rate": 9.870437223714612e-05, "loss": 0.0497, "step": 11900 }, { "grad_norm": 0.35084912180900574, "learning_rate": 9.87006298928574e-05, "loss": 0.0442, "step": 11910 }, { "grad_norm": 0.3068421185016632, "learning_rate": 9.869688222274103e-05, "loss": 0.0434, "step": 11920 }, { "grad_norm": 0.3072677552700043, "learning_rate": 9.869312922720681e-05, "loss": 0.0513, "step": 11930 }, { "grad_norm": 0.33333495259284973, "learning_rate": 9.868937090666521e-05, "loss": 0.0516, "step": 11940 }, { "grad_norm": 0.4368855655193329, "learning_rate": 9.86856072615272e-05, "loss": 0.0521, "step": 11950 }, { "grad_norm": 0.35655006766319275, "learning_rate": 9.868183829220438e-05, "loss": 0.0507, "step": 11960 }, { "grad_norm": 0.43375736474990845, "learning_rate": 9.867806399910893e-05, "loss": 0.051, "step": 11970 }, { "grad_norm": 0.3733741044998169, "learning_rate": 9.867428438265356e-05, "loss": 0.0518, "step": 11980 }, { "grad_norm": 0.3427622318267822, "learning_rate": 9.867049944325165e-05, "loss": 0.0519, "step": 11990 }, { "grad_norm": 0.33429205417633057, "learning_rate": 9.86667091813171e-05, "loss": 0.0549, "step": 12000 }, { "grad_norm": 0.35122233629226685, "learning_rate": 9.866291359726438e-05, "loss": 0.0444, "step": 12010 }, { "grad_norm": 0.4178922772407532, "learning_rate": 9.865911269150861e-05, "loss": 0.0453, "step": 12020 }, { "grad_norm": 0.4660907983779907, "learning_rate": 9.865530646446544e-05, "loss": 0.0464, "step": 12030 }, { "grad_norm": 0.31217294931411743, "learning_rate": 9.86514949165511e-05, "loss": 0.0438, "step": 12040 }, { "grad_norm": 0.3381092846393585, "learning_rate": 9.864767804818243e-05, "loss": 0.0408, "step": 12050 }, { "grad_norm": 0.28189581632614136, "learning_rate": 9.86438558597768e-05, "loss": 0.0431, "step": 12060 }, { "grad_norm": 0.31328365206718445, "learning_rate": 9.864002835175225e-05, "loss": 0.0496, "step": 12070 }, { "grad_norm": 0.23391079902648926, "learning_rate": 9.863619552452734e-05, "loss": 0.047, "step": 12080 }, { "grad_norm": 0.4086768329143524, "learning_rate": 9.863235737852119e-05, "loss": 0.0432, "step": 12090 }, { "grad_norm": 0.3468124568462372, "learning_rate": 9.862851391415356e-05, "loss": 0.049, "step": 12100 }, { "grad_norm": 0.34902918338775635, "learning_rate": 9.862466513184477e-05, "loss": 0.0473, "step": 12110 }, { "grad_norm": 0.2791256904602051, "learning_rate": 9.86208110320157e-05, "loss": 0.0449, "step": 12120 }, { "grad_norm": 0.2988507151603699, "learning_rate": 9.861695161508784e-05, "loss": 0.0481, "step": 12130 }, { "grad_norm": 0.31242653727531433, "learning_rate": 9.861308688148324e-05, "loss": 0.0464, "step": 12140 }, { "grad_norm": 0.2784920334815979, "learning_rate": 9.860921683162455e-05, "loss": 0.0523, "step": 12150 }, { "grad_norm": 0.31956931948661804, "learning_rate": 9.860534146593499e-05, "loss": 0.0437, "step": 12160 }, { "grad_norm": 0.3577985465526581, "learning_rate": 9.860146078483836e-05, "loss": 0.0453, "step": 12170 }, { "grad_norm": 0.2809070944786072, "learning_rate": 9.859757478875905e-05, "loss": 0.0439, "step": 12180 }, { "grad_norm": 0.28153589367866516, "learning_rate": 9.859368347812204e-05, "loss": 0.0462, "step": 12190 }, { "grad_norm": 0.3104923963546753, "learning_rate": 9.858978685335285e-05, "loss": 0.0528, "step": 12200 }, { "grad_norm": 0.24283699691295624, "learning_rate": 9.858588491487763e-05, "loss": 0.0463, "step": 12210 }, { "grad_norm": 0.34565168619155884, "learning_rate": 9.858197766312308e-05, "loss": 0.0466, "step": 12220 }, { "grad_norm": 0.26242443919181824, "learning_rate": 9.857806509851649e-05, "loss": 0.0558, "step": 12230 }, { "grad_norm": 0.3222377598285675, "learning_rate": 9.857414722148574e-05, "loss": 0.0481, "step": 12240 }, { "grad_norm": 0.3465644121170044, "learning_rate": 9.857022403245928e-05, "loss": 0.0438, "step": 12250 }, { "grad_norm": 0.370950847864151, "learning_rate": 9.856629553186615e-05, "loss": 0.0513, "step": 12260 }, { "grad_norm": 0.3061610460281372, "learning_rate": 9.856236172013595e-05, "loss": 0.0514, "step": 12270 }, { "grad_norm": 0.2841220796108246, "learning_rate": 9.85584225976989e-05, "loss": 0.0482, "step": 12280 }, { "grad_norm": 0.3358941376209259, "learning_rate": 9.855447816498575e-05, "loss": 0.0521, "step": 12290 }, { "grad_norm": 0.32255589962005615, "learning_rate": 9.855052842242787e-05, "loss": 0.0555, "step": 12300 }, { "grad_norm": 0.3627440333366394, "learning_rate": 9.85465733704572e-05, "loss": 0.0495, "step": 12310 }, { "grad_norm": 0.3556196987628937, "learning_rate": 9.854261300950624e-05, "loss": 0.0474, "step": 12320 }, { "grad_norm": 0.32869553565979004, "learning_rate": 9.853864734000813e-05, "loss": 0.0514, "step": 12330 }, { "grad_norm": 0.28054550290107727, "learning_rate": 9.85346763623965e-05, "loss": 0.0457, "step": 12340 }, { "grad_norm": 0.3140217065811157, "learning_rate": 9.853070007710564e-05, "loss": 0.0448, "step": 12350 }, { "grad_norm": 0.31876713037490845, "learning_rate": 9.85267184845704e-05, "loss": 0.0471, "step": 12360 }, { "grad_norm": 0.31505364179611206, "learning_rate": 9.852273158522616e-05, "loss": 0.0489, "step": 12370 }, { "grad_norm": 0.2873947024345398, "learning_rate": 9.851873937950896e-05, "loss": 0.0459, "step": 12380 }, { "grad_norm": 0.32910069823265076, "learning_rate": 9.851474186785537e-05, "loss": 0.0452, "step": 12390 }, { "grad_norm": 0.3893514573574066, "learning_rate": 9.851073905070254e-05, "loss": 0.0588, "step": 12400 }, { "grad_norm": 0.29688990116119385, "learning_rate": 9.850673092848824e-05, "loss": 0.0495, "step": 12410 }, { "grad_norm": 0.3019835352897644, "learning_rate": 9.850271750165077e-05, "loss": 0.0498, "step": 12420 }, { "grad_norm": 0.34834399819374084, "learning_rate": 9.849869877062902e-05, "loss": 0.0545, "step": 12430 }, { "grad_norm": 0.38310572504997253, "learning_rate": 9.849467473586252e-05, "loss": 0.0511, "step": 12440 }, { "grad_norm": 0.3561399579048157, "learning_rate": 9.849064539779127e-05, "loss": 0.0532, "step": 12450 }, { "grad_norm": 0.5081175565719604, "learning_rate": 9.848661075685594e-05, "loss": 0.0477, "step": 12460 }, { "grad_norm": 0.2717787027359009, "learning_rate": 9.848257081349778e-05, "loss": 0.0448, "step": 12470 }, { "grad_norm": 0.3515077233314514, "learning_rate": 9.847852556815856e-05, "loss": 0.0482, "step": 12480 }, { "grad_norm": 0.4701513350009918, "learning_rate": 9.847447502128067e-05, "loss": 0.0529, "step": 12490 }, { "grad_norm": 0.3319222331047058, "learning_rate": 9.847041917330708e-05, "loss": 0.0449, "step": 12500 }, { "grad_norm": 0.34558913111686707, "learning_rate": 9.846635802468132e-05, "loss": 0.052, "step": 12510 }, { "grad_norm": 0.33461204171180725, "learning_rate": 9.84622915758475e-05, "loss": 0.0505, "step": 12520 }, { "grad_norm": 0.2494317591190338, "learning_rate": 9.845821982725034e-05, "loss": 0.0432, "step": 12530 }, { "grad_norm": 0.35770466923713684, "learning_rate": 9.845414277933514e-05, "loss": 0.0442, "step": 12540 }, { "grad_norm": 0.2728145122528076, "learning_rate": 9.845006043254771e-05, "loss": 0.0467, "step": 12550 }, { "grad_norm": 0.3417663872241974, "learning_rate": 9.844597278733451e-05, "loss": 0.0509, "step": 12560 }, { "grad_norm": 0.3160945773124695, "learning_rate": 9.844187984414259e-05, "loss": 0.0463, "step": 12570 }, { "grad_norm": 0.3567648231983185, "learning_rate": 9.84377816034195e-05, "loss": 0.047, "step": 12580 }, { "grad_norm": 0.4014863669872284, "learning_rate": 9.843367806561345e-05, "loss": 0.0582, "step": 12590 }, { "grad_norm": 0.3270188868045807, "learning_rate": 9.842956923117317e-05, "loss": 0.0472, "step": 12600 }, { "grad_norm": 0.39865562319755554, "learning_rate": 9.842545510054802e-05, "loss": 0.0517, "step": 12610 }, { "grad_norm": 0.39565232396125793, "learning_rate": 9.842133567418792e-05, "loss": 0.0433, "step": 12620 }, { "grad_norm": 0.32672831416130066, "learning_rate": 9.841721095254333e-05, "loss": 0.0518, "step": 12630 }, { "grad_norm": 0.42257797718048096, "learning_rate": 9.841308093606537e-05, "loss": 0.0462, "step": 12640 }, { "grad_norm": 0.43426916003227234, "learning_rate": 9.840894562520565e-05, "loss": 0.0542, "step": 12650 }, { "grad_norm": 0.36453768610954285, "learning_rate": 9.840480502041642e-05, "loss": 0.0477, "step": 12660 }, { "grad_norm": 0.3201111853122711, "learning_rate": 9.840065912215049e-05, "loss": 0.0546, "step": 12670 }, { "grad_norm": 0.3710078299045563, "learning_rate": 9.839650793086124e-05, "loss": 0.0463, "step": 12680 }, { "grad_norm": 0.3159351050853729, "learning_rate": 9.839235144700265e-05, "loss": 0.0543, "step": 12690 }, { "grad_norm": 0.30835655331611633, "learning_rate": 9.838818967102926e-05, "loss": 0.0494, "step": 12700 }, { "grad_norm": 0.2725861668586731, "learning_rate": 9.83840226033962e-05, "loss": 0.0499, "step": 12710 }, { "grad_norm": 0.2766207158565521, "learning_rate": 9.837985024455918e-05, "loss": 0.0476, "step": 12720 }, { "grad_norm": 0.35191142559051514, "learning_rate": 9.837567259497447e-05, "loss": 0.049, "step": 12730 }, { "grad_norm": 0.34474825859069824, "learning_rate": 9.837148965509894e-05, "loss": 0.0428, "step": 12740 }, { "grad_norm": 0.38295528292655945, "learning_rate": 9.836730142539001e-05, "loss": 0.0475, "step": 12750 }, { "grad_norm": 0.42371857166290283, "learning_rate": 9.836310790630574e-05, "loss": 0.0524, "step": 12760 }, { "grad_norm": 0.35715460777282715, "learning_rate": 9.83589090983047e-05, "loss": 0.048, "step": 12770 }, { "grad_norm": 0.3128608167171478, "learning_rate": 9.835470500184605e-05, "loss": 0.0477, "step": 12780 }, { "grad_norm": 0.3367336094379425, "learning_rate": 9.835049561738957e-05, "loss": 0.0407, "step": 12790 }, { "grad_norm": 0.44591110944747925, "learning_rate": 9.834628094539558e-05, "loss": 0.0531, "step": 12800 }, { "grad_norm": 0.35090985894203186, "learning_rate": 9.834206098632499e-05, "loss": 0.0508, "step": 12810 }, { "grad_norm": 0.2791661322116852, "learning_rate": 9.833783574063931e-05, "loss": 0.0504, "step": 12820 }, { "grad_norm": 0.2955075800418854, "learning_rate": 9.833360520880058e-05, "loss": 0.0552, "step": 12830 }, { "grad_norm": 0.32139360904693604, "learning_rate": 9.832936939127144e-05, "loss": 0.046, "step": 12840 }, { "grad_norm": 0.34117183089256287, "learning_rate": 9.832512828851515e-05, "loss": 0.0417, "step": 12850 }, { "grad_norm": 0.3619537651538849, "learning_rate": 9.832088190099546e-05, "loss": 0.0459, "step": 12860 }, { "grad_norm": 0.3558710217475891, "learning_rate": 9.831663022917679e-05, "loss": 0.0508, "step": 12870 }, { "grad_norm": 0.381354421377182, "learning_rate": 9.831237327352407e-05, "loss": 0.0511, "step": 12880 }, { "grad_norm": 0.28136852383613586, "learning_rate": 9.830811103450286e-05, "loss": 0.0453, "step": 12890 }, { "grad_norm": 0.3464166522026062, "learning_rate": 9.830384351257924e-05, "loss": 0.0491, "step": 12900 }, { "grad_norm": 0.3294628858566284, "learning_rate": 9.829957070821993e-05, "loss": 0.0483, "step": 12910 }, { "grad_norm": 0.27872762084007263, "learning_rate": 9.829529262189218e-05, "loss": 0.0433, "step": 12920 }, { "grad_norm": 0.3082878887653351, "learning_rate": 9.829100925406385e-05, "loss": 0.0494, "step": 12930 }, { "grad_norm": 0.47862401604652405, "learning_rate": 9.828672060520333e-05, "loss": 0.0516, "step": 12940 }, { "grad_norm": 0.33191153407096863, "learning_rate": 9.828242667577966e-05, "loss": 0.0524, "step": 12950 }, { "grad_norm": 0.3397899568080902, "learning_rate": 9.82781274662624e-05, "loss": 0.0448, "step": 12960 }, { "grad_norm": 0.40437695384025574, "learning_rate": 9.82738229771217e-05, "loss": 0.0514, "step": 12970 }, { "grad_norm": 0.36376750469207764, "learning_rate": 9.826951320882829e-05, "loss": 0.0522, "step": 12980 }, { "grad_norm": 0.38291722536087036, "learning_rate": 9.826519816185351e-05, "loss": 0.0455, "step": 12990 }, { "grad_norm": 0.2949780225753784, "learning_rate": 9.826087783666921e-05, "loss": 0.0449, "step": 13000 }, { "grad_norm": 0.3088294565677643, "learning_rate": 9.825655223374787e-05, "loss": 0.05, "step": 13010 }, { "grad_norm": 0.30415916442871094, "learning_rate": 9.825222135356253e-05, "loss": 0.0502, "step": 13020 }, { "grad_norm": 0.3430902361869812, "learning_rate": 9.82478851965868e-05, "loss": 0.0553, "step": 13030 }, { "grad_norm": 0.34271517395973206, "learning_rate": 9.82435437632949e-05, "loss": 0.0511, "step": 13040 }, { "grad_norm": 0.36007171869277954, "learning_rate": 9.823919705416158e-05, "loss": 0.0528, "step": 13050 }, { "grad_norm": 0.31173214316368103, "learning_rate": 9.82348450696622e-05, "loss": 0.0448, "step": 13060 }, { "grad_norm": 0.33253446221351624, "learning_rate": 9.823048781027268e-05, "loss": 0.0438, "step": 13070 }, { "grad_norm": 0.2812059819698334, "learning_rate": 9.822612527646953e-05, "loss": 0.0476, "step": 13080 }, { "grad_norm": 0.33429187536239624, "learning_rate": 9.822175746872984e-05, "loss": 0.0369, "step": 13090 }, { "grad_norm": 0.3608219623565674, "learning_rate": 9.821738438753123e-05, "loss": 0.0539, "step": 13100 }, { "grad_norm": 0.24893172085285187, "learning_rate": 9.821300603335196e-05, "loss": 0.0431, "step": 13110 }, { "grad_norm": 0.27129700779914856, "learning_rate": 9.820862240667085e-05, "loss": 0.0508, "step": 13120 }, { "grad_norm": 0.3024444580078125, "learning_rate": 9.820423350796726e-05, "loss": 0.0489, "step": 13130 }, { "grad_norm": 0.345572292804718, "learning_rate": 9.819983933772118e-05, "loss": 0.0584, "step": 13140 }, { "grad_norm": 0.579224169254303, "learning_rate": 9.819543989641314e-05, "loss": 0.0515, "step": 13150 }, { "grad_norm": 0.2903270423412323, "learning_rate": 9.819103518452423e-05, "loss": 0.0447, "step": 13160 }, { "grad_norm": 0.3540622889995575, "learning_rate": 9.818662520253618e-05, "loss": 0.0491, "step": 13170 }, { "grad_norm": 0.3248685598373413, "learning_rate": 9.818220995093126e-05, "loss": 0.0495, "step": 13180 }, { "grad_norm": 0.39567381143569946, "learning_rate": 9.817778943019228e-05, "loss": 0.0553, "step": 13190 }, { "grad_norm": 0.3602243959903717, "learning_rate": 9.81733636408027e-05, "loss": 0.052, "step": 13200 }, { "grad_norm": 0.31635838747024536, "learning_rate": 9.816893258324649e-05, "loss": 0.0476, "step": 13210 }, { "grad_norm": 0.25972092151641846, "learning_rate": 9.816449625800823e-05, "loss": 0.0573, "step": 13220 }, { "grad_norm": 0.25910940766334534, "learning_rate": 9.816005466557308e-05, "loss": 0.0432, "step": 13230 }, { "grad_norm": 0.2951022982597351, "learning_rate": 9.815560780642674e-05, "loss": 0.0418, "step": 13240 }, { "grad_norm": 0.26253291964530945, "learning_rate": 9.815115568105555e-05, "loss": 0.058, "step": 13250 }, { "grad_norm": 0.3538634181022644, "learning_rate": 9.814669828994638e-05, "loss": 0.048, "step": 13260 }, { "grad_norm": 0.39389288425445557, "learning_rate": 9.814223563358665e-05, "loss": 0.0534, "step": 13270 }, { "grad_norm": 0.36143577098846436, "learning_rate": 9.813776771246443e-05, "loss": 0.0505, "step": 13280 }, { "grad_norm": 0.33008328080177307, "learning_rate": 9.813329452706829e-05, "loss": 0.0444, "step": 13290 }, { "grad_norm": 0.37769615650177, "learning_rate": 9.812881607788744e-05, "loss": 0.0485, "step": 13300 }, { "grad_norm": 0.2696126699447632, "learning_rate": 9.812433236541163e-05, "loss": 0.0537, "step": 13310 }, { "grad_norm": 0.44052186608314514, "learning_rate": 9.811984339013116e-05, "loss": 0.0538, "step": 13320 }, { "grad_norm": 0.28388887643814087, "learning_rate": 9.811534915253698e-05, "loss": 0.0516, "step": 13330 }, { "grad_norm": 0.35663697123527527, "learning_rate": 9.811084965312056e-05, "loss": 0.0509, "step": 13340 }, { "grad_norm": 0.27785325050354004, "learning_rate": 9.810634489237396e-05, "loss": 0.0468, "step": 13350 }, { "grad_norm": 0.4582175612449646, "learning_rate": 9.81018348707898e-05, "loss": 0.0523, "step": 13360 }, { "grad_norm": 0.4146946370601654, "learning_rate": 9.809731958886131e-05, "loss": 0.0496, "step": 13370 }, { "grad_norm": 0.3466791808605194, "learning_rate": 9.809279904708224e-05, "loss": 0.0531, "step": 13380 }, { "grad_norm": 0.23986969888210297, "learning_rate": 9.808827324594699e-05, "loss": 0.051, "step": 13390 }, { "grad_norm": 0.4266051650047302, "learning_rate": 9.808374218595046e-05, "loss": 0.0488, "step": 13400 }, { "grad_norm": 0.2951606512069702, "learning_rate": 9.80792058675882e-05, "loss": 0.0476, "step": 13410 }, { "grad_norm": 0.3807160258293152, "learning_rate": 9.807466429135627e-05, "loss": 0.0449, "step": 13420 }, { "grad_norm": 0.32290560007095337, "learning_rate": 9.807011745775132e-05, "loss": 0.0483, "step": 13430 }, { "grad_norm": 0.4458112418651581, "learning_rate": 9.806556536727061e-05, "loss": 0.0486, "step": 13440 }, { "grad_norm": 0.3723248541355133, "learning_rate": 9.806100802041193e-05, "loss": 0.0493, "step": 13450 }, { "grad_norm": 0.2870277762413025, "learning_rate": 9.805644541767368e-05, "loss": 0.047, "step": 13460 }, { "grad_norm": 0.35142162442207336, "learning_rate": 9.805187755955478e-05, "loss": 0.0415, "step": 13470 }, { "grad_norm": 0.36712533235549927, "learning_rate": 9.804730444655483e-05, "loss": 0.047, "step": 13480 }, { "grad_norm": 0.31945744156837463, "learning_rate": 9.804272607917388e-05, "loss": 0.0434, "step": 13490 }, { "grad_norm": 0.5091832280158997, "learning_rate": 9.803814245791265e-05, "loss": 0.045, "step": 13500 }, { "grad_norm": 0.30633246898651123, "learning_rate": 9.803355358327239e-05, "loss": 0.0441, "step": 13510 }, { "grad_norm": 0.3266007900238037, "learning_rate": 9.802895945575492e-05, "loss": 0.0465, "step": 13520 }, { "grad_norm": 0.3311074376106262, "learning_rate": 9.802436007586266e-05, "loss": 0.0468, "step": 13530 }, { "grad_norm": 0.31500813364982605, "learning_rate": 9.801975544409858e-05, "loss": 0.0407, "step": 13540 }, { "grad_norm": 0.3598085045814514, "learning_rate": 9.801514556096625e-05, "loss": 0.0465, "step": 13550 }, { "grad_norm": 0.528684139251709, "learning_rate": 9.801053042696977e-05, "loss": 0.0477, "step": 13560 }, { "grad_norm": 0.3192172646522522, "learning_rate": 9.800591004261388e-05, "loss": 0.0476, "step": 13570 }, { "grad_norm": 0.3170643448829651, "learning_rate": 9.800128440840385e-05, "loss": 0.0518, "step": 13580 }, { "grad_norm": 0.30971503257751465, "learning_rate": 9.799665352484552e-05, "loss": 0.0457, "step": 13590 }, { "grad_norm": 0.3257831037044525, "learning_rate": 9.799201739244532e-05, "loss": 0.0485, "step": 13600 }, { "grad_norm": 0.3534996509552002, "learning_rate": 9.798737601171025e-05, "loss": 0.0511, "step": 13610 }, { "grad_norm": 0.36455240845680237, "learning_rate": 9.79827293831479e-05, "loss": 0.049, "step": 13620 }, { "grad_norm": 0.3256911039352417, "learning_rate": 9.797807750726638e-05, "loss": 0.0541, "step": 13630 }, { "grad_norm": 0.22839808464050293, "learning_rate": 9.797342038457446e-05, "loss": 0.0559, "step": 13640 }, { "grad_norm": 0.3770243227481842, "learning_rate": 9.796875801558141e-05, "loss": 0.0466, "step": 13650 }, { "grad_norm": 0.35938870906829834, "learning_rate": 9.79640904007971e-05, "loss": 0.0499, "step": 13660 }, { "grad_norm": 0.4581371247768402, "learning_rate": 9.795941754073199e-05, "loss": 0.0436, "step": 13670 }, { "grad_norm": 0.3566678762435913, "learning_rate": 9.795473943589705e-05, "loss": 0.0534, "step": 13680 }, { "grad_norm": 0.33917656540870667, "learning_rate": 9.795005608680394e-05, "loss": 0.0496, "step": 13690 }, { "grad_norm": 0.3531329333782196, "learning_rate": 9.794536749396477e-05, "loss": 0.0507, "step": 13700 }, { "grad_norm": 0.3541075587272644, "learning_rate": 9.79406736578923e-05, "loss": 0.0525, "step": 13710 }, { "grad_norm": 0.30711978673934937, "learning_rate": 9.793597457909984e-05, "loss": 0.0519, "step": 13720 }, { "grad_norm": 0.3025117814540863, "learning_rate": 9.793127025810127e-05, "loss": 0.0476, "step": 13730 }, { "grad_norm": 0.2891595959663391, "learning_rate": 9.792656069541104e-05, "loss": 0.0398, "step": 13740 }, { "grad_norm": 0.3576807379722595, "learning_rate": 9.79218458915442e-05, "loss": 0.0497, "step": 13750 }, { "grad_norm": 0.2561618983745575, "learning_rate": 9.791712584701634e-05, "loss": 0.0519, "step": 13760 }, { "grad_norm": 0.3007572591304779, "learning_rate": 9.791240056234364e-05, "loss": 0.0442, "step": 13770 }, { "grad_norm": 0.29023921489715576, "learning_rate": 9.790767003804283e-05, "loss": 0.0467, "step": 13780 }, { "grad_norm": 0.29901817440986633, "learning_rate": 9.790293427463126e-05, "loss": 0.0477, "step": 13790 }, { "grad_norm": 0.28597667813301086, "learning_rate": 9.789819327262684e-05, "loss": 0.0429, "step": 13800 }, { "grad_norm": 0.2907049059867859, "learning_rate": 9.7893447032548e-05, "loss": 0.0409, "step": 13810 }, { "grad_norm": 0.2907193601131439, "learning_rate": 9.78886955549138e-05, "loss": 0.0434, "step": 13820 }, { "grad_norm": 0.2814364433288574, "learning_rate": 9.788393884024387e-05, "loss": 0.0404, "step": 13830 }, { "grad_norm": 0.43211260437965393, "learning_rate": 9.787917688905836e-05, "loss": 0.0465, "step": 13840 }, { "grad_norm": 0.32958534359931946, "learning_rate": 9.787440970187807e-05, "loss": 0.0455, "step": 13850 }, { "grad_norm": 0.38426870107650757, "learning_rate": 9.786963727922429e-05, "loss": 0.0536, "step": 13860 }, { "grad_norm": 0.29726919531822205, "learning_rate": 9.786485962161897e-05, "loss": 0.0444, "step": 13870 }, { "grad_norm": 0.31046172976493835, "learning_rate": 9.786007672958455e-05, "loss": 0.0466, "step": 13880 }, { "grad_norm": 0.2803924083709717, "learning_rate": 9.78552886036441e-05, "loss": 0.0522, "step": 13890 }, { "grad_norm": 0.29004397988319397, "learning_rate": 9.785049524432124e-05, "loss": 0.0476, "step": 13900 }, { "grad_norm": 0.3796479403972626, "learning_rate": 9.784569665214016e-05, "loss": 0.049, "step": 13910 }, { "grad_norm": 0.31802427768707275, "learning_rate": 9.784089282762563e-05, "loss": 0.0436, "step": 13920 }, { "grad_norm": 0.2877451479434967, "learning_rate": 9.7836083771303e-05, "loss": 0.0476, "step": 13930 }, { "grad_norm": 0.38175246119499207, "learning_rate": 9.783126948369817e-05, "loss": 0.0404, "step": 13940 }, { "grad_norm": 0.35611873865127563, "learning_rate": 9.78264499653376e-05, "loss": 0.0452, "step": 13950 }, { "grad_norm": 0.3085852861404419, "learning_rate": 9.782162521674838e-05, "loss": 0.045, "step": 13960 }, { "grad_norm": 0.2897314429283142, "learning_rate": 9.781679523845812e-05, "loss": 0.0428, "step": 13970 }, { "grad_norm": 0.29077160358428955, "learning_rate": 9.781196003099502e-05, "loss": 0.039, "step": 13980 }, { "grad_norm": 0.3630986213684082, "learning_rate": 9.780711959488786e-05, "loss": 0.0449, "step": 13990 }, { "grad_norm": 0.30688291788101196, "learning_rate": 9.780227393066599e-05, "loss": 0.0477, "step": 14000 }, { "grad_norm": 0.3988196551799774, "learning_rate": 9.77974230388593e-05, "loss": 0.049, "step": 14010 }, { "grad_norm": 0.31189435720443726, "learning_rate": 9.779256691999829e-05, "loss": 0.0512, "step": 14020 }, { "grad_norm": 0.30710649490356445, "learning_rate": 9.778770557461403e-05, "loss": 0.0566, "step": 14030 }, { "grad_norm": 0.5363497138023376, "learning_rate": 9.778283900323812e-05, "loss": 0.0529, "step": 14040 }, { "grad_norm": 0.2928323745727539, "learning_rate": 9.777796720640277e-05, "loss": 0.0442, "step": 14050 }, { "grad_norm": 0.27832773327827454, "learning_rate": 9.777309018464078e-05, "loss": 0.0402, "step": 14060 }, { "grad_norm": 0.26643872261047363, "learning_rate": 9.776820793848547e-05, "loss": 0.0433, "step": 14070 }, { "grad_norm": 0.4282245635986328, "learning_rate": 9.776332046847075e-05, "loss": 0.0525, "step": 14080 }, { "grad_norm": 0.36192265152931213, "learning_rate": 9.775842777513111e-05, "loss": 0.0485, "step": 14090 }, { "grad_norm": 0.2922598719596863, "learning_rate": 9.775352985900163e-05, "loss": 0.0432, "step": 14100 }, { "grad_norm": 0.39382731914520264, "learning_rate": 9.774862672061791e-05, "loss": 0.0503, "step": 14110 }, { "grad_norm": 0.43917956948280334, "learning_rate": 9.774371836051616e-05, "loss": 0.057, "step": 14120 }, { "grad_norm": 0.3782644271850586, "learning_rate": 9.773880477923315e-05, "loss": 0.0496, "step": 14130 }, { "grad_norm": 0.3658880591392517, "learning_rate": 9.773388597730623e-05, "loss": 0.0459, "step": 14140 }, { "grad_norm": 0.3356606662273407, "learning_rate": 9.77289619552733e-05, "loss": 0.0492, "step": 14150 }, { "grad_norm": 0.2719391882419586, "learning_rate": 9.772403271367285e-05, "loss": 0.0469, "step": 14160 }, { "grad_norm": 0.3704564571380615, "learning_rate": 9.771909825304396e-05, "loss": 0.0484, "step": 14170 }, { "grad_norm": 0.311000794172287, "learning_rate": 9.771415857392619e-05, "loss": 0.0452, "step": 14180 }, { "grad_norm": 0.3326418697834015, "learning_rate": 9.770921367685978e-05, "loss": 0.0404, "step": 14190 }, { "grad_norm": 0.3863583207130432, "learning_rate": 9.770426356238551e-05, "loss": 0.0429, "step": 14200 }, { "grad_norm": 0.44443315267562866, "learning_rate": 9.769930823104469e-05, "loss": 0.0492, "step": 14210 }, { "grad_norm": 0.4790307581424713, "learning_rate": 9.769434768337926e-05, "loss": 0.0514, "step": 14220 }, { "grad_norm": 0.32658934593200684, "learning_rate": 9.768938191993164e-05, "loss": 0.0518, "step": 14230 }, { "grad_norm": 0.4303484559059143, "learning_rate": 9.768441094124494e-05, "loss": 0.0526, "step": 14240 }, { "grad_norm": 0.37214165925979614, "learning_rate": 9.767943474786275e-05, "loss": 0.0449, "step": 14250 }, { "grad_norm": 0.5291056036949158, "learning_rate": 9.767445334032923e-05, "loss": 0.0494, "step": 14260 }, { "grad_norm": 0.3258307874202728, "learning_rate": 9.766946671918919e-05, "loss": 0.0482, "step": 14270 }, { "grad_norm": 0.28485238552093506, "learning_rate": 9.766447488498796e-05, "loss": 0.0521, "step": 14280 }, { "grad_norm": 0.30247122049331665, "learning_rate": 9.765947783827139e-05, "loss": 0.0402, "step": 14290 }, { "grad_norm": 0.3442925214767456, "learning_rate": 9.765447557958599e-05, "loss": 0.0424, "step": 14300 }, { "grad_norm": 0.30874016880989075, "learning_rate": 9.764946810947879e-05, "loss": 0.0444, "step": 14310 }, { "grad_norm": 0.2651129364967346, "learning_rate": 9.764445542849738e-05, "loss": 0.0415, "step": 14320 }, { "grad_norm": 0.383593887090683, "learning_rate": 9.763943753718998e-05, "loss": 0.0453, "step": 14330 }, { "grad_norm": 0.29209664463996887, "learning_rate": 9.76344144361053e-05, "loss": 0.0439, "step": 14340 }, { "grad_norm": 0.32993340492248535, "learning_rate": 9.762938612579269e-05, "loss": 0.0487, "step": 14350 }, { "grad_norm": 0.25824856758117676, "learning_rate": 9.762435260680202e-05, "loss": 0.0468, "step": 14360 }, { "grad_norm": 0.373007208108902, "learning_rate": 9.761931387968373e-05, "loss": 0.0506, "step": 14370 }, { "grad_norm": 0.33159735798835754, "learning_rate": 9.76142699449889e-05, "loss": 0.0466, "step": 14380 }, { "grad_norm": 0.30563661456108093, "learning_rate": 9.760922080326908e-05, "loss": 0.0533, "step": 14390 }, { "grad_norm": 0.3828514516353607, "learning_rate": 9.760416645507644e-05, "loss": 0.0465, "step": 14400 }, { "grad_norm": 0.24309095740318298, "learning_rate": 9.759910690096375e-05, "loss": 0.0439, "step": 14410 }, { "grad_norm": 0.2818514406681061, "learning_rate": 9.759404214148429e-05, "loss": 0.0528, "step": 14420 }, { "grad_norm": 0.4253278374671936, "learning_rate": 9.758897217719191e-05, "loss": 0.0463, "step": 14430 }, { "grad_norm": 0.3355671167373657, "learning_rate": 9.758389700864113e-05, "loss": 0.0512, "step": 14440 }, { "grad_norm": 0.3139122426509857, "learning_rate": 9.757881663638688e-05, "loss": 0.0446, "step": 14450 }, { "grad_norm": 0.3057890236377716, "learning_rate": 9.757373106098478e-05, "loss": 0.0476, "step": 14460 }, { "grad_norm": 0.3034583032131195, "learning_rate": 9.756864028299097e-05, "loss": 0.0403, "step": 14470 }, { "grad_norm": 0.21134288609027863, "learning_rate": 9.75635443029622e-05, "loss": 0.0424, "step": 14480 }, { "grad_norm": 0.2746664583683014, "learning_rate": 9.755844312145572e-05, "loss": 0.0411, "step": 14490 }, { "grad_norm": 0.27472707629203796, "learning_rate": 9.755333673902941e-05, "loss": 0.0424, "step": 14500 }, { "grad_norm": 0.33981847763061523, "learning_rate": 9.75482251562417e-05, "loss": 0.0465, "step": 14510 }, { "grad_norm": 0.33284813165664673, "learning_rate": 9.754310837365155e-05, "loss": 0.0401, "step": 14520 }, { "grad_norm": 0.40666598081588745, "learning_rate": 9.753798639181856e-05, "loss": 0.0418, "step": 14530 }, { "grad_norm": 0.32391032576560974, "learning_rate": 9.753285921130286e-05, "loss": 0.0465, "step": 14540 }, { "grad_norm": 0.25537097454071045, "learning_rate": 9.752772683266512e-05, "loss": 0.0414, "step": 14550 }, { "grad_norm": 0.3249761760234833, "learning_rate": 9.752258925646665e-05, "loss": 0.0397, "step": 14560 }, { "grad_norm": 0.30733054876327515, "learning_rate": 9.751744648326926e-05, "loss": 0.0434, "step": 14570 }, { "grad_norm": 0.2570759952068329, "learning_rate": 9.751229851363536e-05, "loss": 0.0466, "step": 14580 }, { "grad_norm": 0.2620013356208801, "learning_rate": 9.750714534812793e-05, "loss": 0.0471, "step": 14590 }, { "grad_norm": 0.31507694721221924, "learning_rate": 9.750198698731053e-05, "loss": 0.0439, "step": 14600 }, { "grad_norm": 0.3552902936935425, "learning_rate": 9.749682343174722e-05, "loss": 0.0428, "step": 14610 }, { "grad_norm": 0.2843281924724579, "learning_rate": 9.749165468200272e-05, "loss": 0.0435, "step": 14620 }, { "grad_norm": 0.280730664730072, "learning_rate": 9.748648073864229e-05, "loss": 0.0434, "step": 14630 }, { "grad_norm": 0.3750157058238983, "learning_rate": 9.748130160223168e-05, "loss": 0.0473, "step": 14640 }, { "grad_norm": 0.2716808319091797, "learning_rate": 9.747611727333734e-05, "loss": 0.0454, "step": 14650 }, { "grad_norm": 0.3916192948818207, "learning_rate": 9.74709277525262e-05, "loss": 0.0495, "step": 14660 }, { "grad_norm": 0.23641341924667358, "learning_rate": 9.746573304036576e-05, "loss": 0.0463, "step": 14670 }, { "grad_norm": 0.2939438223838806, "learning_rate": 9.746053313742412e-05, "loss": 0.0401, "step": 14680 }, { "grad_norm": 0.27165162563323975, "learning_rate": 9.745532804426994e-05, "loss": 0.0384, "step": 14690 }, { "grad_norm": 0.3543950021266937, "learning_rate": 9.745011776147242e-05, "loss": 0.0539, "step": 14700 }, { "grad_norm": 0.4004017412662506, "learning_rate": 9.744490228960138e-05, "loss": 0.0407, "step": 14710 }, { "grad_norm": 0.3156217932701111, "learning_rate": 9.743968162922713e-05, "loss": 0.0428, "step": 14720 }, { "grad_norm": 0.26550018787384033, "learning_rate": 9.743445578092064e-05, "loss": 0.0454, "step": 14730 }, { "grad_norm": 0.34510982036590576, "learning_rate": 9.742922474525338e-05, "loss": 0.0472, "step": 14740 }, { "grad_norm": 0.34270599484443665, "learning_rate": 9.742398852279741e-05, "loss": 0.047, "step": 14750 }, { "grad_norm": 0.34916913509368896, "learning_rate": 9.741874711412535e-05, "loss": 0.0457, "step": 14760 }, { "grad_norm": 0.3056657016277313, "learning_rate": 9.741350051981042e-05, "loss": 0.0449, "step": 14770 }, { "grad_norm": 0.32145169377326965, "learning_rate": 9.740824874042633e-05, "loss": 0.0429, "step": 14780 }, { "grad_norm": 0.3499097526073456, "learning_rate": 9.740299177654746e-05, "loss": 0.0389, "step": 14790 }, { "grad_norm": 0.39772528409957886, "learning_rate": 9.739772962874867e-05, "loss": 0.0446, "step": 14800 }, { "grad_norm": 0.2801424264907837, "learning_rate": 9.739246229760541e-05, "loss": 0.0544, "step": 14810 }, { "grad_norm": 0.3374302387237549, "learning_rate": 9.738718978369376e-05, "loss": 0.0427, "step": 14820 }, { "grad_norm": 0.29023486375808716, "learning_rate": 9.738191208759025e-05, "loss": 0.0444, "step": 14830 }, { "grad_norm": 0.39192813634872437, "learning_rate": 9.73766292098721e-05, "loss": 0.0424, "step": 14840 }, { "grad_norm": 0.36325639486312866, "learning_rate": 9.737134115111699e-05, "loss": 0.0471, "step": 14850 }, { "grad_norm": 0.2781795263290405, "learning_rate": 9.736604791190323e-05, "loss": 0.0502, "step": 14860 }, { "grad_norm": 0.286232590675354, "learning_rate": 9.73607494928097e-05, "loss": 0.0404, "step": 14870 }, { "grad_norm": 0.29781267046928406, "learning_rate": 9.735544589441581e-05, "loss": 0.0494, "step": 14880 }, { "grad_norm": 0.3010658323764801, "learning_rate": 9.735013711730154e-05, "loss": 0.0435, "step": 14890 }, { "grad_norm": 0.2711239159107208, "learning_rate": 9.734482316204747e-05, "loss": 0.0441, "step": 14900 }, { "grad_norm": 0.4134131371974945, "learning_rate": 9.733950402923473e-05, "loss": 0.0402, "step": 14910 }, { "grad_norm": 0.298249214887619, "learning_rate": 9.7334179719445e-05, "loss": 0.0503, "step": 14920 }, { "grad_norm": 0.3386479914188385, "learning_rate": 9.732885023326053e-05, "loss": 0.0496, "step": 14930 }, { "grad_norm": 0.3511277139186859, "learning_rate": 9.732351557126418e-05, "loss": 0.0444, "step": 14940 }, { "grad_norm": 0.3976418077945709, "learning_rate": 9.731817573403929e-05, "loss": 0.0416, "step": 14950 }, { "grad_norm": 0.5082374811172485, "learning_rate": 9.731283072216985e-05, "loss": 0.0449, "step": 14960 }, { "grad_norm": 0.2665858566761017, "learning_rate": 9.730748053624039e-05, "loss": 0.0424, "step": 14970 }, { "grad_norm": 0.32603690028190613, "learning_rate": 9.730212517683598e-05, "loss": 0.0461, "step": 14980 }, { "grad_norm": 0.3141821026802063, "learning_rate": 9.729676464454228e-05, "loss": 0.0479, "step": 14990 }, { "grad_norm": 0.3060629367828369, "learning_rate": 9.72913989399455e-05, "loss": 0.0497, "step": 15000 }, { "grad_norm": 0.3278250992298126, "learning_rate": 9.728602806363242e-05, "loss": 0.0481, "step": 15010 }, { "grad_norm": 0.35824814438819885, "learning_rate": 9.728065201619043e-05, "loss": 0.0463, "step": 15020 }, { "grad_norm": 0.3603852391242981, "learning_rate": 9.727527079820742e-05, "loss": 0.0509, "step": 15030 }, { "grad_norm": 0.23687607049942017, "learning_rate": 9.726988441027186e-05, "loss": 0.0458, "step": 15040 }, { "grad_norm": 0.3637133240699768, "learning_rate": 9.726449285297281e-05, "loss": 0.0428, "step": 15050 }, { "grad_norm": 0.32233697175979614, "learning_rate": 9.72590961268999e-05, "loss": 0.0496, "step": 15060 }, { "grad_norm": 0.4141780138015747, "learning_rate": 9.725369423264328e-05, "loss": 0.0458, "step": 15070 }, { "grad_norm": 0.34259235858917236, "learning_rate": 9.72482871707937e-05, "loss": 0.0467, "step": 15080 }, { "grad_norm": 0.33549556136131287, "learning_rate": 9.724287494194247e-05, "loss": 0.0454, "step": 15090 }, { "grad_norm": 0.29212796688079834, "learning_rate": 9.723745754668147e-05, "loss": 0.0479, "step": 15100 }, { "grad_norm": 0.3430784046649933, "learning_rate": 9.723203498560313e-05, "loss": 0.043, "step": 15110 }, { "grad_norm": 0.3121868968009949, "learning_rate": 9.722660725930046e-05, "loss": 0.0471, "step": 15120 }, { "grad_norm": 0.2774896025657654, "learning_rate": 9.722117436836702e-05, "loss": 0.0405, "step": 15130 }, { "grad_norm": 0.31243255734443665, "learning_rate": 9.721573631339696e-05, "loss": 0.0419, "step": 15140 }, { "grad_norm": 0.2814342975616455, "learning_rate": 9.721029309498494e-05, "loss": 0.0414, "step": 15150 }, { "grad_norm": 0.40552979707717896, "learning_rate": 9.720484471372627e-05, "loss": 0.049, "step": 15160 }, { "grad_norm": 0.3590978682041168, "learning_rate": 9.719939117021673e-05, "loss": 0.0404, "step": 15170 }, { "grad_norm": 0.3017476797103882, "learning_rate": 9.719393246505275e-05, "loss": 0.0394, "step": 15180 }, { "grad_norm": 0.3304344713687897, "learning_rate": 9.718846859883128e-05, "loss": 0.043, "step": 15190 }, { "grad_norm": 0.29548412561416626, "learning_rate": 9.718299957214982e-05, "loss": 0.0402, "step": 15200 }, { "grad_norm": 0.27640682458877563, "learning_rate": 9.717752538560646e-05, "loss": 0.0443, "step": 15210 }, { "grad_norm": 0.30071207880973816, "learning_rate": 9.717204603979986e-05, "loss": 0.0424, "step": 15220 }, { "grad_norm": 0.30058467388153076, "learning_rate": 9.716656153532922e-05, "loss": 0.0471, "step": 15230 }, { "grad_norm": 0.27045702934265137, "learning_rate": 9.716107187279434e-05, "loss": 0.0438, "step": 15240 }, { "grad_norm": 0.28740930557250977, "learning_rate": 9.715557705279555e-05, "loss": 0.0423, "step": 15250 }, { "grad_norm": 0.3529955744743347, "learning_rate": 9.715007707593372e-05, "loss": 0.0418, "step": 15260 }, { "grad_norm": 0.28353944420814514, "learning_rate": 9.714457194281036e-05, "loss": 0.046, "step": 15270 }, { "grad_norm": 0.3509443998336792, "learning_rate": 9.713906165402751e-05, "loss": 0.0553, "step": 15280 }, { "grad_norm": 0.24739313125610352, "learning_rate": 9.713354621018774e-05, "loss": 0.0441, "step": 15290 }, { "grad_norm": 0.3049010634422302, "learning_rate": 9.712802561189422e-05, "loss": 0.0466, "step": 15300 }, { "grad_norm": 0.2630464434623718, "learning_rate": 9.712249985975069e-05, "loss": 0.0397, "step": 15310 }, { "grad_norm": 0.5321766138076782, "learning_rate": 9.71169689543614e-05, "loss": 0.0437, "step": 15320 }, { "grad_norm": 0.2773062288761139, "learning_rate": 9.711143289633123e-05, "loss": 0.0447, "step": 15330 }, { "grad_norm": 0.2891623377799988, "learning_rate": 9.710589168626561e-05, "loss": 0.0383, "step": 15340 }, { "grad_norm": 0.30707642436027527, "learning_rate": 9.710034532477048e-05, "loss": 0.0445, "step": 15350 }, { "grad_norm": 0.3045267164707184, "learning_rate": 9.709479381245239e-05, "loss": 0.0458, "step": 15360 }, { "grad_norm": 0.31282511353492737, "learning_rate": 9.708923714991847e-05, "loss": 0.0449, "step": 15370 }, { "grad_norm": 0.2947273254394531, "learning_rate": 9.708367533777638e-05, "loss": 0.0418, "step": 15380 }, { "grad_norm": 0.3597317337989807, "learning_rate": 9.707810837663431e-05, "loss": 0.0461, "step": 15390 }, { "grad_norm": 0.3169351816177368, "learning_rate": 9.707253626710113e-05, "loss": 0.0441, "step": 15400 }, { "grad_norm": 0.32806578278541565, "learning_rate": 9.706695900978613e-05, "loss": 0.0417, "step": 15410 }, { "grad_norm": 0.30076321959495544, "learning_rate": 9.706137660529926e-05, "loss": 0.0419, "step": 15420 }, { "grad_norm": 0.208360493183136, "learning_rate": 9.705578905425101e-05, "loss": 0.0366, "step": 15430 }, { "grad_norm": 0.35257333517074585, "learning_rate": 9.705019635725241e-05, "loss": 0.0428, "step": 15440 }, { "grad_norm": 0.35319995880126953, "learning_rate": 9.704459851491508e-05, "loss": 0.0453, "step": 15450 }, { "grad_norm": 0.26026463508605957, "learning_rate": 9.703899552785118e-05, "loss": 0.0394, "step": 15460 }, { "grad_norm": 0.3096758723258972, "learning_rate": 9.703338739667346e-05, "loss": 0.0432, "step": 15470 }, { "grad_norm": 0.26592299342155457, "learning_rate": 9.70277741219952e-05, "loss": 0.0364, "step": 15480 }, { "grad_norm": 0.3030959665775299, "learning_rate": 9.702215570443027e-05, "loss": 0.0402, "step": 15490 }, { "grad_norm": 0.2965589165687561, "learning_rate": 9.701653214459309e-05, "loss": 0.0436, "step": 15500 }, { "grad_norm": 0.28693076968193054, "learning_rate": 9.701090344309865e-05, "loss": 0.0464, "step": 15510 }, { "grad_norm": 0.2909204363822937, "learning_rate": 9.700526960056247e-05, "loss": 0.0396, "step": 15520 }, { "grad_norm": 0.2868652939796448, "learning_rate": 9.699963061760068e-05, "loss": 0.0443, "step": 15530 }, { "grad_norm": 0.4522731602191925, "learning_rate": 9.699398649482997e-05, "loss": 0.044, "step": 15540 }, { "grad_norm": 0.36470213532447815, "learning_rate": 9.698833723286753e-05, "loss": 0.0455, "step": 15550 }, { "grad_norm": 0.35817763209342957, "learning_rate": 9.698268283233118e-05, "loss": 0.0457, "step": 15560 }, { "grad_norm": 0.32192739844322205, "learning_rate": 9.697702329383929e-05, "loss": 0.042, "step": 15570 }, { "grad_norm": 0.2347530722618103, "learning_rate": 9.697135861801074e-05, "loss": 0.0415, "step": 15580 }, { "grad_norm": 0.26759251952171326, "learning_rate": 9.696568880546505e-05, "loss": 0.0465, "step": 15590 }, { "grad_norm": 0.3186815083026886, "learning_rate": 9.696001385682223e-05, "loss": 0.0438, "step": 15600 }, { "grad_norm": 0.2803758978843689, "learning_rate": 9.695433377270291e-05, "loss": 0.0501, "step": 15610 }, { "grad_norm": 0.30160969495773315, "learning_rate": 9.694864855372824e-05, "loss": 0.0404, "step": 15620 }, { "grad_norm": 0.42841392755508423, "learning_rate": 9.694295820051995e-05, "loss": 0.0427, "step": 15630 }, { "grad_norm": 0.32873082160949707, "learning_rate": 9.693726271370032e-05, "loss": 0.0462, "step": 15640 }, { "grad_norm": 0.5937984585762024, "learning_rate": 9.693156209389221e-05, "loss": 0.0488, "step": 15650 }, { "grad_norm": 0.36570897698402405, "learning_rate": 9.692585634171905e-05, "loss": 0.0432, "step": 15660 }, { "grad_norm": 0.34070083498954773, "learning_rate": 9.692014545780476e-05, "loss": 0.0429, "step": 15670 }, { "grad_norm": 0.31672191619873047, "learning_rate": 9.691442944277393e-05, "loss": 0.0462, "step": 15680 }, { "grad_norm": 0.35786256194114685, "learning_rate": 9.690870829725162e-05, "loss": 0.0369, "step": 15690 }, { "grad_norm": 0.31428301334381104, "learning_rate": 9.69029820218635e-05, "loss": 0.044, "step": 15700 }, { "grad_norm": 0.33645132184028625, "learning_rate": 9.689725061723579e-05, "loss": 0.0395, "step": 15710 }, { "grad_norm": 0.3795012831687927, "learning_rate": 9.689151408399527e-05, "loss": 0.0409, "step": 15720 }, { "grad_norm": 0.23152576386928558, "learning_rate": 9.688577242276924e-05, "loss": 0.0456, "step": 15730 }, { "grad_norm": 0.2632371783256531, "learning_rate": 9.688002563418566e-05, "loss": 0.0384, "step": 15740 }, { "grad_norm": 0.35081055760383606, "learning_rate": 9.687427371887293e-05, "loss": 0.0462, "step": 15750 }, { "grad_norm": 0.36402952671051025, "learning_rate": 9.686851667746012e-05, "loss": 0.0486, "step": 15760 }, { "grad_norm": 0.3953661024570465, "learning_rate": 9.686275451057677e-05, "loss": 0.0478, "step": 15770 }, { "grad_norm": 0.31463125348091125, "learning_rate": 9.685698721885308e-05, "loss": 0.0374, "step": 15780 }, { "grad_norm": 0.31162822246551514, "learning_rate": 9.68512148029197e-05, "loss": 0.0444, "step": 15790 }, { "grad_norm": 0.38760578632354736, "learning_rate": 9.684543726340791e-05, "loss": 0.0454, "step": 15800 }, { "grad_norm": 0.2892177402973175, "learning_rate": 9.683965460094952e-05, "loss": 0.0471, "step": 15810 }, { "grad_norm": 0.33988574147224426, "learning_rate": 9.683386681617694e-05, "loss": 0.0506, "step": 15820 }, { "grad_norm": 0.35007444024086, "learning_rate": 9.68280739097231e-05, "loss": 0.0443, "step": 15830 }, { "grad_norm": 0.3110329210758209, "learning_rate": 9.682227588222148e-05, "loss": 0.0426, "step": 15840 }, { "grad_norm": 0.29703548550605774, "learning_rate": 9.681647273430618e-05, "loss": 0.0421, "step": 15850 }, { "grad_norm": 0.3491549789905548, "learning_rate": 9.681066446661182e-05, "loss": 0.0473, "step": 15860 }, { "grad_norm": 0.32357001304626465, "learning_rate": 9.680485107977357e-05, "loss": 0.0442, "step": 15870 }, { "grad_norm": 0.4433218240737915, "learning_rate": 9.679903257442716e-05, "loss": 0.0507, "step": 15880 }, { "grad_norm": 0.2369937002658844, "learning_rate": 9.679320895120891e-05, "loss": 0.0398, "step": 15890 }, { "grad_norm": 0.3115740418434143, "learning_rate": 9.67873802107557e-05, "loss": 0.0426, "step": 15900 }, { "grad_norm": 0.3075025975704193, "learning_rate": 9.67815463537049e-05, "loss": 0.0483, "step": 15910 }, { "grad_norm": 0.42613211274147034, "learning_rate": 9.677570738069457e-05, "loss": 0.047, "step": 15920 }, { "grad_norm": 0.318473219871521, "learning_rate": 9.676986329236318e-05, "loss": 0.0444, "step": 15930 }, { "grad_norm": 0.30910971760749817, "learning_rate": 9.676401408934987e-05, "loss": 0.0423, "step": 15940 }, { "grad_norm": 0.304619163274765, "learning_rate": 9.675815977229428e-05, "loss": 0.0514, "step": 15950 }, { "grad_norm": 0.5484680533409119, "learning_rate": 9.675230034183664e-05, "loss": 0.0436, "step": 15960 }, { "grad_norm": 0.38982924818992615, "learning_rate": 9.674643579861773e-05, "loss": 0.0465, "step": 15970 }, { "grad_norm": 0.3782029151916504, "learning_rate": 9.674056614327886e-05, "loss": 0.0482, "step": 15980 }, { "grad_norm": 0.30078625679016113, "learning_rate": 9.673469137646198e-05, "loss": 0.0498, "step": 15990 }, { "grad_norm": 0.2952653467655182, "learning_rate": 9.67288114988095e-05, "loss": 0.0477, "step": 16000 }, { "grad_norm": 0.353102445602417, "learning_rate": 9.672292651096447e-05, "loss": 0.0534, "step": 16010 }, { "grad_norm": 0.3797408640384674, "learning_rate": 9.671703641357042e-05, "loss": 0.0505, "step": 16020 }, { "grad_norm": 0.37431803345680237, "learning_rate": 9.67111412072715e-05, "loss": 0.0512, "step": 16030 }, { "grad_norm": 0.2801273763179779, "learning_rate": 9.670524089271242e-05, "loss": 0.0455, "step": 16040 }, { "grad_norm": 0.28491130471229553, "learning_rate": 9.669933547053842e-05, "loss": 0.0439, "step": 16050 }, { "grad_norm": 0.36413219571113586, "learning_rate": 9.669342494139531e-05, "loss": 0.0481, "step": 16060 }, { "grad_norm": 0.2674226462841034, "learning_rate": 9.668750930592943e-05, "loss": 0.0459, "step": 16070 }, { "grad_norm": 0.3229173421859741, "learning_rate": 9.668158856478775e-05, "loss": 0.0496, "step": 16080 }, { "grad_norm": 0.4002511203289032, "learning_rate": 9.66756627186177e-05, "loss": 0.0406, "step": 16090 }, { "grad_norm": 0.3614268898963928, "learning_rate": 9.666973176806737e-05, "loss": 0.0428, "step": 16100 }, { "grad_norm": 0.35973188281059265, "learning_rate": 9.666379571378534e-05, "loss": 0.039, "step": 16110 }, { "grad_norm": 0.29047995805740356, "learning_rate": 9.665785455642076e-05, "loss": 0.0382, "step": 16120 }, { "grad_norm": 0.2899315357208252, "learning_rate": 9.665190829662337e-05, "loss": 0.0445, "step": 16130 }, { "grad_norm": 0.46114471554756165, "learning_rate": 9.664595693504342e-05, "loss": 0.0451, "step": 16140 }, { "grad_norm": 0.3389485478401184, "learning_rate": 9.664000047233175e-05, "loss": 0.0426, "step": 16150 }, { "grad_norm": 0.2658313810825348, "learning_rate": 9.663403890913976e-05, "loss": 0.0495, "step": 16160 }, { "grad_norm": 0.34302711486816406, "learning_rate": 9.662807224611938e-05, "loss": 0.0462, "step": 16170 }, { "grad_norm": 0.2546341121196747, "learning_rate": 9.662210048392311e-05, "loss": 0.0382, "step": 16180 }, { "grad_norm": 0.2959740161895752, "learning_rate": 9.661612362320405e-05, "loss": 0.0425, "step": 16190 }, { "grad_norm": 0.27989664673805237, "learning_rate": 9.661014166461579e-05, "loss": 0.0395, "step": 16200 }, { "grad_norm": 0.2846079170703888, "learning_rate": 9.66041546088125e-05, "loss": 0.0495, "step": 16210 }, { "grad_norm": 0.3815038502216339, "learning_rate": 9.659816245644895e-05, "loss": 0.0395, "step": 16220 }, { "grad_norm": 0.2876931428909302, "learning_rate": 9.65921652081804e-05, "loss": 0.0405, "step": 16230 }, { "grad_norm": 0.227733314037323, "learning_rate": 9.658616286466271e-05, "loss": 0.043, "step": 16240 }, { "grad_norm": 0.40751704573631287, "learning_rate": 9.65801554265523e-05, "loss": 0.0471, "step": 16250 }, { "grad_norm": 0.30162671208381653, "learning_rate": 9.657414289450612e-05, "loss": 0.0384, "step": 16260 }, { "grad_norm": 0.4308367669582367, "learning_rate": 9.656812526918171e-05, "loss": 0.0418, "step": 16270 }, { "grad_norm": 0.343974769115448, "learning_rate": 9.656210255123712e-05, "loss": 0.0474, "step": 16280 }, { "grad_norm": 0.2867889404296875, "learning_rate": 9.6556074741331e-05, "loss": 0.0427, "step": 16290 }, { "grad_norm": 0.22744160890579224, "learning_rate": 9.655004184012256e-05, "loss": 0.0448, "step": 16300 }, { "grad_norm": 0.36800894141197205, "learning_rate": 9.654400384827152e-05, "loss": 0.0456, "step": 16310 }, { "grad_norm": 0.2735383212566376, "learning_rate": 9.653796076643818e-05, "loss": 0.0408, "step": 16320 }, { "grad_norm": 0.31054553389549255, "learning_rate": 9.653191259528344e-05, "loss": 0.0419, "step": 16330 }, { "grad_norm": 0.31522849202156067, "learning_rate": 9.65258593354687e-05, "loss": 0.0501, "step": 16340 }, { "grad_norm": 0.28238558769226074, "learning_rate": 9.651980098765591e-05, "loss": 0.0464, "step": 16350 }, { "grad_norm": 0.28682172298431396, "learning_rate": 9.651373755250765e-05, "loss": 0.0477, "step": 16360 }, { "grad_norm": 0.26184138655662537, "learning_rate": 9.650766903068697e-05, "loss": 0.0458, "step": 16370 }, { "grad_norm": 0.37046992778778076, "learning_rate": 9.650159542285753e-05, "loss": 0.0444, "step": 16380 }, { "grad_norm": 0.33881181478500366, "learning_rate": 9.649551672968353e-05, "loss": 0.0429, "step": 16390 }, { "grad_norm": 0.2766590118408203, "learning_rate": 9.648943295182973e-05, "loss": 0.0365, "step": 16400 }, { "grad_norm": 0.32779476046562195, "learning_rate": 9.648334408996144e-05, "loss": 0.0451, "step": 16410 }, { "grad_norm": 0.3074224889278412, "learning_rate": 9.647725014474452e-05, "loss": 0.0421, "step": 16420 }, { "grad_norm": 0.2473163902759552, "learning_rate": 9.64711511168454e-05, "loss": 0.0434, "step": 16430 }, { "grad_norm": 0.3028188645839691, "learning_rate": 9.646504700693108e-05, "loss": 0.0371, "step": 16440 }, { "grad_norm": 0.23368509113788605, "learning_rate": 9.645893781566907e-05, "loss": 0.0395, "step": 16450 }, { "grad_norm": 0.4118861258029938, "learning_rate": 9.645282354372744e-05, "loss": 0.0408, "step": 16460 }, { "grad_norm": 0.3999190032482147, "learning_rate": 9.644670419177491e-05, "loss": 0.0425, "step": 16470 }, { "grad_norm": 0.3144058585166931, "learning_rate": 9.644057976048062e-05, "loss": 0.0392, "step": 16480 }, { "grad_norm": 0.3819354176521301, "learning_rate": 9.643445025051435e-05, "loss": 0.0439, "step": 16490 }, { "grad_norm": 0.5631871819496155, "learning_rate": 9.642831566254641e-05, "loss": 0.0449, "step": 16500 }, { "grad_norm": 0.30717000365257263, "learning_rate": 9.642217599724769e-05, "loss": 0.0453, "step": 16510 }, { "grad_norm": 0.2829464375972748, "learning_rate": 9.64160312552896e-05, "loss": 0.0379, "step": 16520 }, { "grad_norm": 0.4415261745452881, "learning_rate": 9.64098814373441e-05, "loss": 0.0465, "step": 16530 }, { "grad_norm": 0.3378543555736542, "learning_rate": 9.640372654408374e-05, "loss": 0.0534, "step": 16540 }, { "grad_norm": 0.2700168788433075, "learning_rate": 9.639756657618162e-05, "loss": 0.0504, "step": 16550 }, { "grad_norm": 0.3015792667865753, "learning_rate": 9.639140153431138e-05, "loss": 0.0512, "step": 16560 }, { "grad_norm": 0.4623999297618866, "learning_rate": 9.638523141914721e-05, "loss": 0.0497, "step": 16570 }, { "grad_norm": 0.22876089811325073, "learning_rate": 9.637905623136388e-05, "loss": 0.0453, "step": 16580 }, { "grad_norm": 0.23523873090744019, "learning_rate": 9.637287597163669e-05, "loss": 0.0397, "step": 16590 }, { "grad_norm": 0.3100273609161377, "learning_rate": 9.63666906406415e-05, "loss": 0.0401, "step": 16600 }, { "grad_norm": 0.29331526160240173, "learning_rate": 9.636050023905473e-05, "loss": 0.0424, "step": 16610 }, { "grad_norm": 0.26433151960372925, "learning_rate": 9.635430476755336e-05, "loss": 0.0476, "step": 16620 }, { "grad_norm": 0.26269543170928955, "learning_rate": 9.63481042268149e-05, "loss": 0.0447, "step": 16630 }, { "grad_norm": 0.23791241645812988, "learning_rate": 9.634189861751745e-05, "loss": 0.0428, "step": 16640 }, { "grad_norm": 0.28460848331451416, "learning_rate": 9.633568794033967e-05, "loss": 0.0419, "step": 16650 }, { "grad_norm": 0.3598819375038147, "learning_rate": 9.63294721959607e-05, "loss": 0.0386, "step": 16660 }, { "grad_norm": 0.2602587044239044, "learning_rate": 9.63232513850603e-05, "loss": 0.0429, "step": 16670 }, { "grad_norm": 0.3515322506427765, "learning_rate": 9.631702550831878e-05, "loss": 0.0376, "step": 16680 }, { "grad_norm": 0.3808808922767639, "learning_rate": 9.631079456641698e-05, "loss": 0.0389, "step": 16690 }, { "grad_norm": 0.32899579405784607, "learning_rate": 9.630455856003632e-05, "loss": 0.0382, "step": 16700 }, { "grad_norm": 0.2186407446861267, "learning_rate": 9.629831748985876e-05, "loss": 0.0377, "step": 16710 }, { "grad_norm": 0.3217272162437439, "learning_rate": 9.629207135656679e-05, "loss": 0.0382, "step": 16720 }, { "grad_norm": 0.3062385320663452, "learning_rate": 9.628582016084353e-05, "loss": 0.0399, "step": 16730 }, { "grad_norm": 0.28421035408973694, "learning_rate": 9.627956390337254e-05, "loss": 0.0415, "step": 16740 }, { "grad_norm": 0.32723698019981384, "learning_rate": 9.627330258483802e-05, "loss": 0.0372, "step": 16750 }, { "grad_norm": 0.3454698622226715, "learning_rate": 9.62670362059247e-05, "loss": 0.045, "step": 16760 }, { "grad_norm": 0.3432324230670929, "learning_rate": 9.626076476731786e-05, "loss": 0.0419, "step": 16770 }, { "grad_norm": 0.2522853910923004, "learning_rate": 9.625448826970336e-05, "loss": 0.0404, "step": 16780 }, { "grad_norm": 0.32553830742836, "learning_rate": 9.624820671376755e-05, "loss": 0.0375, "step": 16790 }, { "grad_norm": 0.25039181113243103, "learning_rate": 9.62419201001974e-05, "loss": 0.0443, "step": 16800 }, { "grad_norm": 0.2075188010931015, "learning_rate": 9.623562842968037e-05, "loss": 0.0392, "step": 16810 }, { "grad_norm": 0.24862805008888245, "learning_rate": 9.622933170290454e-05, "loss": 0.0436, "step": 16820 }, { "grad_norm": 0.2670648992061615, "learning_rate": 9.622302992055849e-05, "loss": 0.0397, "step": 16830 }, { "grad_norm": 0.3908593952655792, "learning_rate": 9.62167230833314e-05, "loss": 0.0456, "step": 16840 }, { "grad_norm": 0.3107130527496338, "learning_rate": 9.621041119191295e-05, "loss": 0.0392, "step": 16850 }, { "grad_norm": 0.2667558491230011, "learning_rate": 9.620409424699342e-05, "loss": 0.0385, "step": 16860 }, { "grad_norm": 0.35526132583618164, "learning_rate": 9.619777224926359e-05, "loss": 0.0383, "step": 16870 }, { "grad_norm": 0.2631467580795288, "learning_rate": 9.619144519941485e-05, "loss": 0.0419, "step": 16880 }, { "grad_norm": 0.28812524676322937, "learning_rate": 9.618511309813912e-05, "loss": 0.0384, "step": 16890 }, { "grad_norm": 0.28528398275375366, "learning_rate": 9.617877594612886e-05, "loss": 0.0418, "step": 16900 }, { "grad_norm": 0.28414663672447205, "learning_rate": 9.617243374407707e-05, "loss": 0.0363, "step": 16910 }, { "grad_norm": 0.39786651730537415, "learning_rate": 9.616608649267736e-05, "loss": 0.0495, "step": 16920 }, { "grad_norm": 0.3016234040260315, "learning_rate": 9.615973419262385e-05, "loss": 0.0432, "step": 16930 }, { "grad_norm": 0.3633803427219391, "learning_rate": 9.615337684461119e-05, "loss": 0.0399, "step": 16940 }, { "grad_norm": 0.26335078477859497, "learning_rate": 9.614701444933465e-05, "loss": 0.0399, "step": 16950 }, { "grad_norm": 0.2964004874229431, "learning_rate": 9.614064700748997e-05, "loss": 0.0418, "step": 16960 }, { "grad_norm": 0.2737901210784912, "learning_rate": 9.613427451977352e-05, "loss": 0.0376, "step": 16970 }, { "grad_norm": 0.42229634523391724, "learning_rate": 9.612789698688216e-05, "loss": 0.0471, "step": 16980 }, { "grad_norm": 0.29131442308425903, "learning_rate": 9.612151440951334e-05, "loss": 0.0433, "step": 16990 }, { "grad_norm": 0.5950684547424316, "learning_rate": 9.611512678836506e-05, "loss": 0.0402, "step": 17000 }, { "grad_norm": 0.3649860620498657, "learning_rate": 9.610873412413584e-05, "loss": 0.0428, "step": 17010 }, { "grad_norm": 0.33355778455734253, "learning_rate": 9.610233641752476e-05, "loss": 0.0494, "step": 17020 }, { "grad_norm": 0.31380870938301086, "learning_rate": 9.609593366923151e-05, "loss": 0.0387, "step": 17030 }, { "grad_norm": 0.2708700895309448, "learning_rate": 9.608952587995625e-05, "loss": 0.0475, "step": 17040 }, { "grad_norm": 0.2574889361858368, "learning_rate": 9.608311305039972e-05, "loss": 0.0442, "step": 17050 }, { "grad_norm": 0.2712864577770233, "learning_rate": 9.607669518126326e-05, "loss": 0.0479, "step": 17060 }, { "grad_norm": 0.3336707651615143, "learning_rate": 9.607027227324866e-05, "loss": 0.0474, "step": 17070 }, { "grad_norm": 0.3272700905799866, "learning_rate": 9.606384432705837e-05, "loss": 0.0404, "step": 17080 }, { "grad_norm": 0.28378626704216003, "learning_rate": 9.60574113433953e-05, "loss": 0.0449, "step": 17090 }, { "grad_norm": 0.32551833987236023, "learning_rate": 9.6050973322963e-05, "loss": 0.0402, "step": 17100 }, { "grad_norm": 0.42754319310188293, "learning_rate": 9.604453026646547e-05, "loss": 0.0458, "step": 17110 }, { "grad_norm": 0.5415604114532471, "learning_rate": 9.603808217460735e-05, "loss": 0.046, "step": 17120 }, { "grad_norm": 0.36627280712127686, "learning_rate": 9.603162904809377e-05, "loss": 0.0448, "step": 17130 }, { "grad_norm": 0.29427433013916016, "learning_rate": 9.602517088763045e-05, "loss": 0.0402, "step": 17140 }, { "grad_norm": 0.3481740355491638, "learning_rate": 9.601870769392365e-05, "loss": 0.0409, "step": 17150 }, { "grad_norm": 0.2768818140029907, "learning_rate": 9.601223946768017e-05, "loss": 0.0422, "step": 17160 }, { "grad_norm": 0.37003564834594727, "learning_rate": 9.600576620960734e-05, "loss": 0.0419, "step": 17170 }, { "grad_norm": 0.46014630794525146, "learning_rate": 9.599928792041308e-05, "loss": 0.0445, "step": 17180 }, { "grad_norm": 0.32520443201065063, "learning_rate": 9.599280460080587e-05, "loss": 0.0448, "step": 17190 }, { "grad_norm": 0.29360878467559814, "learning_rate": 9.59863162514947e-05, "loss": 0.0412, "step": 17200 }, { "grad_norm": 0.3467155396938324, "learning_rate": 9.597982287318911e-05, "loss": 0.0439, "step": 17210 }, { "grad_norm": 0.4283860921859741, "learning_rate": 9.597332446659923e-05, "loss": 0.0429, "step": 17220 }, { "grad_norm": 0.3016991913318634, "learning_rate": 9.59668210324357e-05, "loss": 0.0439, "step": 17230 }, { "grad_norm": 0.3802063763141632, "learning_rate": 9.596031257140974e-05, "loss": 0.0435, "step": 17240 }, { "grad_norm": 0.31672948598861694, "learning_rate": 9.59537990842331e-05, "loss": 0.0381, "step": 17250 }, { "grad_norm": 0.25721675157546997, "learning_rate": 9.594728057161806e-05, "loss": 0.0453, "step": 17260 }, { "grad_norm": 0.3420756757259369, "learning_rate": 9.594075703427752e-05, "loss": 0.0415, "step": 17270 }, { "grad_norm": 0.32831117510795593, "learning_rate": 9.593422847292486e-05, "loss": 0.0395, "step": 17280 }, { "grad_norm": 0.30626845359802246, "learning_rate": 9.592769488827402e-05, "loss": 0.0459, "step": 17290 }, { "grad_norm": 0.2634623050689697, "learning_rate": 9.592115628103952e-05, "loss": 0.0425, "step": 17300 }, { "grad_norm": 0.39185723662376404, "learning_rate": 9.591461265193643e-05, "loss": 0.0424, "step": 17310 }, { "grad_norm": 0.2986023724079132, "learning_rate": 9.590806400168032e-05, "loss": 0.041, "step": 17320 }, { "grad_norm": 0.3381703197956085, "learning_rate": 9.590151033098735e-05, "loss": 0.0425, "step": 17330 }, { "grad_norm": 0.3268478810787201, "learning_rate": 9.589495164057423e-05, "loss": 0.0472, "step": 17340 }, { "grad_norm": 0.31076952815055847, "learning_rate": 9.58883879311582e-05, "loss": 0.046, "step": 17350 }, { "grad_norm": 0.37347501516342163, "learning_rate": 9.588181920345705e-05, "loss": 0.0458, "step": 17360 }, { "grad_norm": 0.23729205131530762, "learning_rate": 9.587524545818913e-05, "loss": 0.0416, "step": 17370 }, { "grad_norm": 0.3170870542526245, "learning_rate": 9.586866669607335e-05, "loss": 0.043, "step": 17380 }, { "grad_norm": 0.3618462085723877, "learning_rate": 9.586208291782915e-05, "loss": 0.0375, "step": 17390 }, { "grad_norm": 0.384040892124176, "learning_rate": 9.58554941241765e-05, "loss": 0.0469, "step": 17400 }, { "grad_norm": 0.3245178461074829, "learning_rate": 9.584890031583596e-05, "loss": 0.0417, "step": 17410 }, { "grad_norm": 0.29453709721565247, "learning_rate": 9.584230149352861e-05, "loss": 0.0352, "step": 17420 }, { "grad_norm": 0.30352553725242615, "learning_rate": 9.58356976579761e-05, "loss": 0.0381, "step": 17430 }, { "grad_norm": 0.3452381491661072, "learning_rate": 9.58290888099006e-05, "loss": 0.0428, "step": 17440 }, { "grad_norm": 0.26678547263145447, "learning_rate": 9.582247495002486e-05, "loss": 0.0461, "step": 17450 }, { "grad_norm": 0.26927557587623596, "learning_rate": 9.581585607907214e-05, "loss": 0.0401, "step": 17460 }, { "grad_norm": 0.34516721963882446, "learning_rate": 9.580923219776628e-05, "loss": 0.0396, "step": 17470 }, { "grad_norm": 0.3053296208381653, "learning_rate": 9.580260330683167e-05, "loss": 0.0393, "step": 17480 }, { "grad_norm": 0.2403024137020111, "learning_rate": 9.579596940699322e-05, "loss": 0.0383, "step": 17490 }, { "grad_norm": 0.2878586947917938, "learning_rate": 9.578933049897643e-05, "loss": 0.042, "step": 17500 }, { "grad_norm": 0.26436087489128113, "learning_rate": 9.578268658350728e-05, "loss": 0.0389, "step": 17510 }, { "grad_norm": 0.23963811993598938, "learning_rate": 9.577603766131235e-05, "loss": 0.0362, "step": 17520 }, { "grad_norm": 0.2214564085006714, "learning_rate": 9.576938373311878e-05, "loss": 0.0336, "step": 17530 }, { "grad_norm": 0.3070088028907776, "learning_rate": 9.576272479965421e-05, "loss": 0.0365, "step": 17540 }, { "grad_norm": 0.3513237237930298, "learning_rate": 9.575606086164687e-05, "loss": 0.0377, "step": 17550 }, { "grad_norm": 0.30551856756210327, "learning_rate": 9.57493919198255e-05, "loss": 0.0414, "step": 17560 }, { "grad_norm": 0.27559894323349, "learning_rate": 9.57427179749194e-05, "loss": 0.0374, "step": 17570 }, { "grad_norm": 0.29534298181533813, "learning_rate": 9.573603902765846e-05, "loss": 0.0338, "step": 17580 }, { "grad_norm": 0.24197088181972504, "learning_rate": 9.572935507877304e-05, "loss": 0.0374, "step": 17590 }, { "grad_norm": 0.27601999044418335, "learning_rate": 9.57226661289941e-05, "loss": 0.0441, "step": 17600 }, { "grad_norm": 0.4929561913013458, "learning_rate": 9.571597217905315e-05, "loss": 0.0408, "step": 17610 }, { "grad_norm": 0.43010213971138, "learning_rate": 9.57092732296822e-05, "loss": 0.0437, "step": 17620 }, { "grad_norm": 0.3260008990764618, "learning_rate": 9.570256928161385e-05, "loss": 0.046, "step": 17630 }, { "grad_norm": 0.36272257566452026, "learning_rate": 9.569586033558126e-05, "loss": 0.0453, "step": 17640 }, { "grad_norm": 0.303329199552536, "learning_rate": 9.568914639231807e-05, "loss": 0.0491, "step": 17650 }, { "grad_norm": 0.288501501083374, "learning_rate": 9.568242745255852e-05, "loss": 0.04, "step": 17660 }, { "grad_norm": 0.23755981028079987, "learning_rate": 9.567570351703739e-05, "loss": 0.038, "step": 17670 }, { "grad_norm": 0.4388062655925751, "learning_rate": 9.566897458649001e-05, "loss": 0.0464, "step": 17680 }, { "grad_norm": 0.2962532341480255, "learning_rate": 9.566224066165221e-05, "loss": 0.0443, "step": 17690 }, { "grad_norm": 0.3704446852207184, "learning_rate": 9.565550174326043e-05, "loss": 0.0462, "step": 17700 }, { "grad_norm": 0.345060259103775, "learning_rate": 9.564875783205162e-05, "loss": 0.0436, "step": 17710 }, { "grad_norm": 0.3127157390117645, "learning_rate": 9.564200892876328e-05, "loss": 0.0452, "step": 17720 }, { "grad_norm": 0.3401845097541809, "learning_rate": 9.563525503413348e-05, "loss": 0.0453, "step": 17730 }, { "grad_norm": 0.2872157394886017, "learning_rate": 9.562849614890079e-05, "loss": 0.0384, "step": 17740 }, { "grad_norm": 0.2805516719818115, "learning_rate": 9.562173227380436e-05, "loss": 0.0378, "step": 17750 }, { "grad_norm": 0.31095197796821594, "learning_rate": 9.561496340958389e-05, "loss": 0.04, "step": 17760 }, { "grad_norm": 0.2944924235343933, "learning_rate": 9.560818955697959e-05, "loss": 0.0423, "step": 17770 }, { "grad_norm": 0.31539633870124817, "learning_rate": 9.560141071673228e-05, "loss": 0.0452, "step": 17780 }, { "grad_norm": 0.30558255314826965, "learning_rate": 9.559462688958323e-05, "loss": 0.0435, "step": 17790 }, { "grad_norm": 0.2798837721347809, "learning_rate": 9.558783807627434e-05, "loss": 0.0403, "step": 17800 }, { "grad_norm": 0.22173108160495758, "learning_rate": 9.558104427754801e-05, "loss": 0.0442, "step": 17810 }, { "grad_norm": 0.3517598509788513, "learning_rate": 9.557424549414722e-05, "loss": 0.0435, "step": 17820 }, { "grad_norm": 0.474528431892395, "learning_rate": 9.556744172681546e-05, "loss": 0.0418, "step": 17830 }, { "grad_norm": 0.30388978123664856, "learning_rate": 9.556063297629677e-05, "loss": 0.0459, "step": 17840 }, { "grad_norm": 0.32702749967575073, "learning_rate": 9.555381924333578e-05, "loss": 0.0452, "step": 17850 }, { "grad_norm": 0.2566956877708435, "learning_rate": 9.554700052867758e-05, "loss": 0.0368, "step": 17860 }, { "grad_norm": 0.25342050194740295, "learning_rate": 9.554017683306789e-05, "loss": 0.0378, "step": 17870 }, { "grad_norm": 0.25890272855758667, "learning_rate": 9.553334815725294e-05, "loss": 0.0403, "step": 17880 }, { "grad_norm": 0.37081974744796753, "learning_rate": 9.552651450197949e-05, "loss": 0.0454, "step": 17890 }, { "grad_norm": 0.3219183087348938, "learning_rate": 9.551967586799486e-05, "loss": 0.0466, "step": 17900 }, { "grad_norm": 0.40597593784332275, "learning_rate": 9.551283225604692e-05, "loss": 0.0522, "step": 17910 }, { "grad_norm": 0.3648420572280884, "learning_rate": 9.550598366688406e-05, "loss": 0.0417, "step": 17920 }, { "grad_norm": 0.2764841914176941, "learning_rate": 9.549913010125526e-05, "loss": 0.041, "step": 17930 }, { "grad_norm": 0.33121177554130554, "learning_rate": 9.549227155990999e-05, "loss": 0.0379, "step": 17940 }, { "grad_norm": 0.31917351484298706, "learning_rate": 9.548540804359828e-05, "loss": 0.0415, "step": 17950 }, { "grad_norm": 0.3117912709712982, "learning_rate": 9.547853955307077e-05, "loss": 0.046, "step": 17960 }, { "grad_norm": 0.3187728822231293, "learning_rate": 9.547166608907853e-05, "loss": 0.0365, "step": 17970 }, { "grad_norm": 0.2665868103504181, "learning_rate": 9.546478765237326e-05, "loss": 0.0467, "step": 17980 }, { "grad_norm": 0.37967005372047424, "learning_rate": 9.545790424370715e-05, "loss": 0.0507, "step": 17990 }, { "grad_norm": 0.2856180667877197, "learning_rate": 9.5451015863833e-05, "loss": 0.0492, "step": 18000 }, { "grad_norm": 0.3466992974281311, "learning_rate": 9.544412251350408e-05, "loss": 0.0507, "step": 18010 }, { "grad_norm": 0.32982879877090454, "learning_rate": 9.543722419347422e-05, "loss": 0.0409, "step": 18020 }, { "grad_norm": 0.20151163637638092, "learning_rate": 9.543032090449788e-05, "loss": 0.0366, "step": 18030 }, { "grad_norm": 0.2144405096769333, "learning_rate": 9.542341264732992e-05, "loss": 0.0401, "step": 18040 }, { "grad_norm": 0.3698650598526001, "learning_rate": 9.541649942272585e-05, "loss": 0.0441, "step": 18050 }, { "grad_norm": 0.2977275848388672, "learning_rate": 9.54095812314417e-05, "loss": 0.0447, "step": 18060 }, { "grad_norm": 0.4260321855545044, "learning_rate": 9.540265807423401e-05, "loss": 0.0426, "step": 18070 }, { "grad_norm": 0.27275630831718445, "learning_rate": 9.53957299518599e-05, "loss": 0.041, "step": 18080 }, { "grad_norm": 0.296827495098114, "learning_rate": 9.5388796865077e-05, "loss": 0.042, "step": 18090 }, { "grad_norm": 0.4142874777317047, "learning_rate": 9.538185881464353e-05, "loss": 0.0441, "step": 18100 }, { "grad_norm": 0.297258198261261, "learning_rate": 9.537491580131821e-05, "loss": 0.0436, "step": 18110 }, { "grad_norm": 0.36190488934516907, "learning_rate": 9.53679678258603e-05, "loss": 0.0359, "step": 18120 }, { "grad_norm": 0.2953573763370514, "learning_rate": 9.536101488902966e-05, "loss": 0.0403, "step": 18130 }, { "grad_norm": 0.26806509494781494, "learning_rate": 9.535405699158663e-05, "loss": 0.0424, "step": 18140 }, { "grad_norm": 0.26284825801849365, "learning_rate": 9.53470941342921e-05, "loss": 0.0446, "step": 18150 }, { "grad_norm": 0.31628745794296265, "learning_rate": 9.534012631790756e-05, "loss": 0.0477, "step": 18160 }, { "grad_norm": 0.3488289713859558, "learning_rate": 9.533315354319494e-05, "loss": 0.0436, "step": 18170 }, { "grad_norm": 0.2644602060317993, "learning_rate": 9.532617581091682e-05, "loss": 0.039, "step": 18180 }, { "grad_norm": 0.2959763705730438, "learning_rate": 9.531919312183629e-05, "loss": 0.0441, "step": 18190 }, { "grad_norm": 0.26641377806663513, "learning_rate": 9.531220547671688e-05, "loss": 0.0381, "step": 18200 }, { "grad_norm": 0.2817487120628357, "learning_rate": 9.530521287632285e-05, "loss": 0.049, "step": 18210 }, { "grad_norm": 0.2988262176513672, "learning_rate": 9.529821532141884e-05, "loss": 0.0388, "step": 18220 }, { "grad_norm": 0.2654905915260315, "learning_rate": 9.52912128127701e-05, "loss": 0.0417, "step": 18230 }, { "grad_norm": 0.25451841950416565, "learning_rate": 9.528420535114244e-05, "loss": 0.0441, "step": 18240 }, { "grad_norm": 0.4105472266674042, "learning_rate": 9.527719293730215e-05, "loss": 0.0408, "step": 18250 }, { "grad_norm": 0.35969141125679016, "learning_rate": 9.527017557201611e-05, "loss": 0.048, "step": 18260 }, { "grad_norm": 0.32978489995002747, "learning_rate": 9.526315325605176e-05, "loss": 0.0452, "step": 18270 }, { "grad_norm": 0.31864991784095764, "learning_rate": 9.525612599017699e-05, "loss": 0.0483, "step": 18280 }, { "grad_norm": 0.3059356212615967, "learning_rate": 9.524909377516033e-05, "loss": 0.037, "step": 18290 }, { "grad_norm": 0.32788538932800293, "learning_rate": 9.524205661177081e-05, "loss": 0.0414, "step": 18300 }, { "grad_norm": 0.27420923113822937, "learning_rate": 9.523501450077801e-05, "loss": 0.0408, "step": 18310 }, { "grad_norm": 0.29910531640052795, "learning_rate": 9.522796744295202e-05, "loss": 0.0379, "step": 18320 }, { "grad_norm": 0.3667720854282379, "learning_rate": 9.522091543906352e-05, "loss": 0.046, "step": 18330 }, { "grad_norm": 0.2761314809322357, "learning_rate": 9.521385848988369e-05, "loss": 0.0397, "step": 18340 }, { "grad_norm": 0.2869720160961151, "learning_rate": 9.520679659618428e-05, "loss": 0.0411, "step": 18350 }, { "grad_norm": 0.3426254689693451, "learning_rate": 9.519972975873754e-05, "loss": 0.0401, "step": 18360 }, { "grad_norm": 0.24342983961105347, "learning_rate": 9.519265797831633e-05, "loss": 0.0371, "step": 18370 }, { "grad_norm": 0.2963721454143524, "learning_rate": 9.518558125569399e-05, "loss": 0.0375, "step": 18380 }, { "grad_norm": 0.26263564825057983, "learning_rate": 9.517849959164442e-05, "loss": 0.0372, "step": 18390 }, { "grad_norm": 0.36299633979797363, "learning_rate": 9.517141298694205e-05, "loss": 0.044, "step": 18400 }, { "grad_norm": 0.3414997458457947, "learning_rate": 9.516432144236188e-05, "loss": 0.0449, "step": 18410 }, { "grad_norm": 0.32385268807411194, "learning_rate": 9.515722495867941e-05, "loss": 0.0385, "step": 18420 }, { "grad_norm": 0.4039923846721649, "learning_rate": 9.515012353667072e-05, "loss": 0.0432, "step": 18430 }, { "grad_norm": 0.33224987983703613, "learning_rate": 9.51430171771124e-05, "loss": 0.0401, "step": 18440 }, { "grad_norm": 0.4295211732387543, "learning_rate": 9.513590588078159e-05, "loss": 0.0443, "step": 18450 }, { "grad_norm": 0.3000915050506592, "learning_rate": 9.512878964845597e-05, "loss": 0.043, "step": 18460 }, { "grad_norm": 0.39333534240722656, "learning_rate": 9.512166848091377e-05, "loss": 0.0404, "step": 18470 }, { "grad_norm": 0.3371150493621826, "learning_rate": 9.511454237893376e-05, "loss": 0.0424, "step": 18480 }, { "grad_norm": 0.4125612676143646, "learning_rate": 9.51074113432952e-05, "loss": 0.0405, "step": 18490 }, { "grad_norm": 0.38185620307922363, "learning_rate": 9.510027537477797e-05, "loss": 0.0403, "step": 18500 }, { "grad_norm": 0.2869214415550232, "learning_rate": 9.509313447416242e-05, "loss": 0.0374, "step": 18510 }, { "grad_norm": 0.3284566402435303, "learning_rate": 9.508598864222949e-05, "loss": 0.0398, "step": 18520 }, { "grad_norm": 0.37063685059547424, "learning_rate": 9.507883787976062e-05, "loss": 0.0419, "step": 18530 }, { "grad_norm": 0.4892098307609558, "learning_rate": 9.507168218753781e-05, "loss": 0.0385, "step": 18540 }, { "grad_norm": 0.46637317538261414, "learning_rate": 9.506452156634362e-05, "loss": 0.0439, "step": 18550 }, { "grad_norm": 0.23532414436340332, "learning_rate": 9.505735601696109e-05, "loss": 0.0457, "step": 18560 }, { "grad_norm": 0.3435328006744385, "learning_rate": 9.505018554017385e-05, "loss": 0.0405, "step": 18570 }, { "grad_norm": 0.2816906273365021, "learning_rate": 9.504301013676604e-05, "loss": 0.0389, "step": 18580 }, { "grad_norm": 0.27247491478919983, "learning_rate": 9.503582980752238e-05, "loss": 0.0393, "step": 18590 }, { "grad_norm": 0.32804474234580994, "learning_rate": 9.502864455322809e-05, "loss": 0.0431, "step": 18600 }, { "grad_norm": 0.34697476029396057, "learning_rate": 9.502145437466891e-05, "loss": 0.0381, "step": 18610 }, { "grad_norm": 0.25897371768951416, "learning_rate": 9.501425927263116e-05, "loss": 0.0401, "step": 18620 }, { "grad_norm": 0.4268742501735687, "learning_rate": 9.500705924790172e-05, "loss": 0.0377, "step": 18630 }, { "grad_norm": 0.2765524387359619, "learning_rate": 9.499985430126794e-05, "loss": 0.0362, "step": 18640 }, { "grad_norm": 0.37186822295188904, "learning_rate": 9.499264443351775e-05, "loss": 0.0389, "step": 18650 }, { "grad_norm": 0.2972698509693146, "learning_rate": 9.498542964543961e-05, "loss": 0.0366, "step": 18660 }, { "grad_norm": 0.2748803496360779, "learning_rate": 9.497820993782252e-05, "loss": 0.0364, "step": 18670 }, { "grad_norm": 0.27259698510169983, "learning_rate": 9.497098531145601e-05, "loss": 0.0421, "step": 18680 }, { "grad_norm": 0.4327303469181061, "learning_rate": 9.496375576713017e-05, "loss": 0.0421, "step": 18690 }, { "grad_norm": 0.24711742997169495, "learning_rate": 9.49565213056356e-05, "loss": 0.0429, "step": 18700 }, { "grad_norm": 0.2800405025482178, "learning_rate": 9.494928192776342e-05, "loss": 0.036, "step": 18710 }, { "grad_norm": 0.29259636998176575, "learning_rate": 9.494203763430538e-05, "loss": 0.0398, "step": 18720 }, { "grad_norm": 0.4015572965145111, "learning_rate": 9.493478842605366e-05, "loss": 0.0347, "step": 18730 }, { "grad_norm": 0.2619989812374115, "learning_rate": 9.492753430380105e-05, "loss": 0.0361, "step": 18740 }, { "grad_norm": 0.2475527971982956, "learning_rate": 9.492027526834083e-05, "loss": 0.0388, "step": 18750 }, { "grad_norm": 0.33629748225212097, "learning_rate": 9.491301132046684e-05, "loss": 0.041, "step": 18760 }, { "grad_norm": 0.3086952567100525, "learning_rate": 9.490574246097345e-05, "loss": 0.0355, "step": 18770 }, { "grad_norm": 0.30268973112106323, "learning_rate": 9.48984686906556e-05, "loss": 0.036, "step": 18780 }, { "grad_norm": 0.2273128628730774, "learning_rate": 9.489119001030871e-05, "loss": 0.0354, "step": 18790 }, { "grad_norm": 0.32524845004081726, "learning_rate": 9.488390642072878e-05, "loss": 0.0403, "step": 18800 }, { "grad_norm": 0.31948766112327576, "learning_rate": 9.48766179227123e-05, "loss": 0.0342, "step": 18810 }, { "grad_norm": 0.3526419699192047, "learning_rate": 9.486932451705636e-05, "loss": 0.0356, "step": 18820 }, { "grad_norm": 0.2752958834171295, "learning_rate": 9.486202620455857e-05, "loss": 0.041, "step": 18830 }, { "grad_norm": 0.36466994881629944, "learning_rate": 9.485472298601704e-05, "loss": 0.0408, "step": 18840 }, { "grad_norm": 0.355692982673645, "learning_rate": 9.484741486223043e-05, "loss": 0.0447, "step": 18850 }, { "grad_norm": 0.27401936054229736, "learning_rate": 9.484010183399797e-05, "loss": 0.0378, "step": 18860 }, { "grad_norm": 0.2950071096420288, "learning_rate": 9.483278390211938e-05, "loss": 0.0385, "step": 18870 }, { "grad_norm": 0.2884671092033386, "learning_rate": 9.482546106739496e-05, "loss": 0.0358, "step": 18880 }, { "grad_norm": 0.29619598388671875, "learning_rate": 9.48181333306255e-05, "loss": 0.0397, "step": 18890 }, { "grad_norm": 0.2762194275856018, "learning_rate": 9.481080069261237e-05, "loss": 0.0376, "step": 18900 }, { "grad_norm": 0.28808602690696716, "learning_rate": 9.480346315415745e-05, "loss": 0.0369, "step": 18910 }, { "grad_norm": 0.35705530643463135, "learning_rate": 9.479612071606314e-05, "loss": 0.0416, "step": 18920 }, { "grad_norm": 0.37362873554229736, "learning_rate": 9.478877337913244e-05, "loss": 0.0429, "step": 18930 }, { "grad_norm": 0.41337454319000244, "learning_rate": 9.478142114416881e-05, "loss": 0.0474, "step": 18940 }, { "grad_norm": 0.30315956473350525, "learning_rate": 9.47740640119763e-05, "loss": 0.0432, "step": 18950 }, { "grad_norm": 0.3705739974975586, "learning_rate": 9.476670198335947e-05, "loss": 0.0425, "step": 18960 }, { "grad_norm": 0.2648918032646179, "learning_rate": 9.47593350591234e-05, "loss": 0.042, "step": 18970 }, { "grad_norm": 0.21682284772396088, "learning_rate": 9.475196324007376e-05, "loss": 0.0367, "step": 18980 }, { "grad_norm": 0.33236655592918396, "learning_rate": 9.474458652701669e-05, "loss": 0.0377, "step": 18990 }, { "grad_norm": 0.27268335223197937, "learning_rate": 9.473720492075892e-05, "loss": 0.0396, "step": 19000 }, { "grad_norm": 0.2748596966266632, "learning_rate": 9.472981842210768e-05, "loss": 0.0536, "step": 19010 }, { "grad_norm": 0.28468558192253113, "learning_rate": 9.472242703187074e-05, "loss": 0.0401, "step": 19020 }, { "grad_norm": 0.26357346773147583, "learning_rate": 9.471503075085643e-05, "loss": 0.0356, "step": 19030 }, { "grad_norm": 0.31876373291015625, "learning_rate": 9.470762957987359e-05, "loss": 0.0396, "step": 19040 }, { "grad_norm": 0.28555336594581604, "learning_rate": 9.470022351973158e-05, "loss": 0.0356, "step": 19050 }, { "grad_norm": 0.31873685121536255, "learning_rate": 9.469281257124034e-05, "loss": 0.0376, "step": 19060 }, { "grad_norm": 0.28364160656929016, "learning_rate": 9.46853967352103e-05, "loss": 0.039, "step": 19070 }, { "grad_norm": 0.24366901814937592, "learning_rate": 9.467797601245246e-05, "loss": 0.0423, "step": 19080 }, { "grad_norm": 0.2247345745563507, "learning_rate": 9.467055040377834e-05, "loss": 0.0491, "step": 19090 }, { "grad_norm": 0.33495229482650757, "learning_rate": 9.466311990999999e-05, "loss": 0.0454, "step": 19100 }, { "grad_norm": 0.28254076838493347, "learning_rate": 9.465568453193e-05, "loss": 0.039, "step": 19110 }, { "grad_norm": 0.297055184841156, "learning_rate": 9.464824427038148e-05, "loss": 0.0441, "step": 19120 }, { "grad_norm": 0.3393535912036896, "learning_rate": 9.46407991261681e-05, "loss": 0.0386, "step": 19130 }, { "grad_norm": 0.35237404704093933, "learning_rate": 9.463334910010404e-05, "loss": 0.0411, "step": 19140 }, { "grad_norm": 0.33266013860702515, "learning_rate": 9.462589419300403e-05, "loss": 0.045, "step": 19150 }, { "grad_norm": 0.34752580523490906, "learning_rate": 9.461843440568333e-05, "loss": 0.0397, "step": 19160 }, { "grad_norm": 0.27326545119285583, "learning_rate": 9.461096973895773e-05, "loss": 0.0405, "step": 19170 }, { "grad_norm": 0.30597418546676636, "learning_rate": 9.460350019364355e-05, "loss": 0.0404, "step": 19180 }, { "grad_norm": 0.2319726198911667, "learning_rate": 9.459602577055764e-05, "loss": 0.0325, "step": 19190 }, { "grad_norm": 0.2696608603000641, "learning_rate": 9.45885464705174e-05, "loss": 0.0402, "step": 19200 }, { "grad_norm": 0.3179985582828522, "learning_rate": 9.458106229434076e-05, "loss": 0.041, "step": 19210 }, { "grad_norm": 0.2813761234283447, "learning_rate": 9.457357324284617e-05, "loss": 0.0389, "step": 19220 }, { "grad_norm": 0.2283802479505539, "learning_rate": 9.456607931685262e-05, "loss": 0.032, "step": 19230 }, { "grad_norm": 0.2551376521587372, "learning_rate": 9.455858051717965e-05, "loss": 0.0329, "step": 19240 }, { "grad_norm": 0.24503223598003387, "learning_rate": 9.45510768446473e-05, "loss": 0.0363, "step": 19250 }, { "grad_norm": 0.3947776257991791, "learning_rate": 9.454356830007618e-05, "loss": 0.0375, "step": 19260 }, { "grad_norm": 0.3513856828212738, "learning_rate": 9.45360548842874e-05, "loss": 0.0459, "step": 19270 }, { "grad_norm": 0.2319161295890808, "learning_rate": 9.452853659810261e-05, "loss": 0.0377, "step": 19280 }, { "grad_norm": 0.46878182888031006, "learning_rate": 9.452101344234401e-05, "loss": 0.0456, "step": 19290 }, { "grad_norm": 0.2728790044784546, "learning_rate": 9.451348541783431e-05, "loss": 0.0413, "step": 19300 }, { "grad_norm": 0.3741236925125122, "learning_rate": 9.450595252539678e-05, "loss": 0.0442, "step": 19310 }, { "grad_norm": 0.3242413401603699, "learning_rate": 9.449841476585518e-05, "loss": 0.0382, "step": 19320 }, { "grad_norm": 0.38512957096099854, "learning_rate": 9.449087214003384e-05, "loss": 0.0398, "step": 19330 }, { "grad_norm": 0.31384551525115967, "learning_rate": 9.448332464875765e-05, "loss": 0.0419, "step": 19340 }, { "grad_norm": 0.36935558915138245, "learning_rate": 9.447577229285192e-05, "loss": 0.0371, "step": 19350 }, { "grad_norm": 0.2934320271015167, "learning_rate": 9.446821507314261e-05, "loss": 0.034, "step": 19360 }, { "grad_norm": 0.3375256061553955, "learning_rate": 9.446065299045617e-05, "loss": 0.0363, "step": 19370 }, { "grad_norm": 0.26348796486854553, "learning_rate": 9.445308604561955e-05, "loss": 0.043, "step": 19380 }, { "grad_norm": 0.3419700562953949, "learning_rate": 9.444551423946028e-05, "loss": 0.0368, "step": 19390 }, { "grad_norm": 0.3171050548553467, "learning_rate": 9.443793757280638e-05, "loss": 0.0416, "step": 19400 }, { "grad_norm": 0.27815553545951843, "learning_rate": 9.443035604648646e-05, "loss": 0.0404, "step": 19410 }, { "grad_norm": 0.3001198470592499, "learning_rate": 9.44227696613296e-05, "loss": 0.0417, "step": 19420 }, { "grad_norm": 0.3038581311702728, "learning_rate": 9.441517841816542e-05, "loss": 0.0442, "step": 19430 }, { "grad_norm": 0.34444087743759155, "learning_rate": 9.440758231782413e-05, "loss": 0.0366, "step": 19440 }, { "grad_norm": 0.3619679808616638, "learning_rate": 9.439998136113639e-05, "loss": 0.0416, "step": 19450 }, { "grad_norm": 0.2989976704120636, "learning_rate": 9.439237554893344e-05, "loss": 0.0388, "step": 19460 }, { "grad_norm": 0.32317864894866943, "learning_rate": 9.438476488204705e-05, "loss": 0.0384, "step": 19470 }, { "grad_norm": 0.28723663091659546, "learning_rate": 9.43771493613095e-05, "loss": 0.0425, "step": 19480 }, { "grad_norm": 0.44889310002326965, "learning_rate": 9.436952898755362e-05, "loss": 0.039, "step": 19490 }, { "grad_norm": 0.30609774589538574, "learning_rate": 9.436190376161276e-05, "loss": 0.0443, "step": 19500 }, { "grad_norm": 0.3026175796985626, "learning_rate": 9.43542736843208e-05, "loss": 0.0474, "step": 19510 }, { "grad_norm": 0.264139860868454, "learning_rate": 9.434663875651216e-05, "loss": 0.0372, "step": 19520 }, { "grad_norm": 0.3338431417942047, "learning_rate": 9.433899897902177e-05, "loss": 0.0401, "step": 19530 }, { "grad_norm": 0.2925910949707031, "learning_rate": 9.433135435268511e-05, "loss": 0.0406, "step": 19540 }, { "grad_norm": 0.2973836660385132, "learning_rate": 9.432370487833819e-05, "loss": 0.0378, "step": 19550 }, { "grad_norm": 0.19612878561019897, "learning_rate": 9.431605055681756e-05, "loss": 0.0352, "step": 19560 }, { "grad_norm": 0.3339715600013733, "learning_rate": 9.430839138896026e-05, "loss": 0.0464, "step": 19570 }, { "grad_norm": 0.3351176381111145, "learning_rate": 9.43007273756039e-05, "loss": 0.037, "step": 19580 }, { "grad_norm": 0.3222159743309021, "learning_rate": 9.429305851758658e-05, "loss": 0.0374, "step": 19590 }, { "grad_norm": 0.2647070288658142, "learning_rate": 9.428538481574699e-05, "loss": 0.0358, "step": 19600 }, { "grad_norm": 0.26754772663116455, "learning_rate": 9.42777062709243e-05, "loss": 0.0482, "step": 19610 }, { "grad_norm": 0.29331088066101074, "learning_rate": 9.427002288395821e-05, "loss": 0.0407, "step": 19620 }, { "grad_norm": 0.25046229362487793, "learning_rate": 9.426233465568898e-05, "loss": 0.0422, "step": 19630 }, { "grad_norm": 0.3466441333293915, "learning_rate": 9.42546415869574e-05, "loss": 0.0398, "step": 19640 }, { "grad_norm": 0.2598443627357483, "learning_rate": 9.424694367860473e-05, "loss": 0.0513, "step": 19650 }, { "grad_norm": 0.3783765137195587, "learning_rate": 9.423924093147284e-05, "loss": 0.0419, "step": 19660 }, { "grad_norm": 0.36904725432395935, "learning_rate": 9.423153334640407e-05, "loss": 0.0427, "step": 19670 }, { "grad_norm": 0.219173401594162, "learning_rate": 9.42238209242413e-05, "loss": 0.0432, "step": 19680 }, { "grad_norm": 0.3041026294231415, "learning_rate": 9.421610366582798e-05, "loss": 0.0443, "step": 19690 }, { "grad_norm": 0.3221820890903473, "learning_rate": 9.420838157200803e-05, "loss": 0.0472, "step": 19700 }, { "grad_norm": 0.261104553937912, "learning_rate": 9.420065464362594e-05, "loss": 0.0428, "step": 19710 }, { "grad_norm": 0.2669680118560791, "learning_rate": 9.419292288152673e-05, "loss": 0.0419, "step": 19720 }, { "grad_norm": 0.3711620271205902, "learning_rate": 9.418518628655588e-05, "loss": 0.0352, "step": 19730 }, { "grad_norm": 0.9988202452659607, "learning_rate": 9.417744485955951e-05, "loss": 0.0339, "step": 19740 }, { "grad_norm": 0.383710116147995, "learning_rate": 9.41696986013842e-05, "loss": 0.0481, "step": 19750 }, { "grad_norm": 0.3178221583366394, "learning_rate": 9.416194751287705e-05, "loss": 0.043, "step": 19760 }, { "grad_norm": 0.33633431792259216, "learning_rate": 9.415419159488572e-05, "loss": 0.0397, "step": 19770 }, { "grad_norm": 0.25578030943870544, "learning_rate": 9.414643084825837e-05, "loss": 0.0409, "step": 19780 }, { "grad_norm": 0.3015790283679962, "learning_rate": 9.413866527384372e-05, "loss": 0.0389, "step": 19790 }, { "grad_norm": 0.28914791345596313, "learning_rate": 9.4130894872491e-05, "loss": 0.0373, "step": 19800 }, { "grad_norm": 0.3288598954677582, "learning_rate": 9.412311964504998e-05, "loss": 0.0378, "step": 19810 }, { "grad_norm": 0.2822599411010742, "learning_rate": 9.411533959237091e-05, "loss": 0.0387, "step": 19820 }, { "grad_norm": 0.25842228531837463, "learning_rate": 9.410755471530464e-05, "loss": 0.0364, "step": 19830 }, { "grad_norm": 0.3129352629184723, "learning_rate": 9.40997650147025e-05, "loss": 0.0366, "step": 19840 }, { "grad_norm": 0.280333548784256, "learning_rate": 9.409197049141637e-05, "loss": 0.037, "step": 19850 }, { "grad_norm": 0.31350380182266235, "learning_rate": 9.408417114629863e-05, "loss": 0.0356, "step": 19860 }, { "grad_norm": 0.32454851269721985, "learning_rate": 9.40763669802022e-05, "loss": 0.0346, "step": 19870 }, { "grad_norm": 0.3015909194946289, "learning_rate": 9.406855799398056e-05, "loss": 0.0384, "step": 19880 }, { "grad_norm": 0.28331509232521057, "learning_rate": 9.406074418848767e-05, "loss": 0.0471, "step": 19890 }, { "grad_norm": 0.24210722744464874, "learning_rate": 9.405292556457805e-05, "loss": 0.0387, "step": 19900 }, { "grad_norm": 0.23665574193000793, "learning_rate": 9.404510212310671e-05, "loss": 0.035, "step": 19910 }, { "grad_norm": 0.2574698030948639, "learning_rate": 9.403727386492924e-05, "loss": 0.0404, "step": 19920 }, { "grad_norm": 0.2432027906179428, "learning_rate": 9.40294407909017e-05, "loss": 0.0374, "step": 19930 }, { "grad_norm": 0.26111099123954773, "learning_rate": 9.40216029018807e-05, "loss": 0.0358, "step": 19940 }, { "grad_norm": 0.28234079480171204, "learning_rate": 9.401376019872338e-05, "loss": 0.0344, "step": 19950 }, { "grad_norm": 0.30514419078826904, "learning_rate": 9.400591268228746e-05, "loss": 0.0375, "step": 19960 }, { "grad_norm": 0.27939754724502563, "learning_rate": 9.399806035343106e-05, "loss": 0.0352, "step": 19970 }, { "grad_norm": 0.34281548857688904, "learning_rate": 9.399020321301294e-05, "loss": 0.0438, "step": 19980 }, { "grad_norm": 0.3660121262073517, "learning_rate": 9.398234126189234e-05, "loss": 0.0432, "step": 19990 }, { "grad_norm": 0.32771170139312744, "learning_rate": 9.397447450092902e-05, "loss": 0.0386, "step": 20000 }, { "grad_norm": 0.24567770957946777, "learning_rate": 9.39666029309833e-05, "loss": 0.0371, "step": 20010 }, { "grad_norm": 0.300313800573349, "learning_rate": 9.395872655291596e-05, "loss": 0.0365, "step": 20020 }, { "grad_norm": 0.3661552965641022, "learning_rate": 9.395084536758838e-05, "loss": 0.0442, "step": 20030 }, { "grad_norm": 0.26093724370002747, "learning_rate": 9.394295937586243e-05, "loss": 0.0483, "step": 20040 }, { "grad_norm": 0.2664765417575836, "learning_rate": 9.393506857860052e-05, "loss": 0.0366, "step": 20050 }, { "grad_norm": 0.23903365433216095, "learning_rate": 9.392717297666555e-05, "loss": 0.0428, "step": 20060 }, { "grad_norm": 0.34040841460227966, "learning_rate": 9.391927257092101e-05, "loss": 0.0423, "step": 20070 }, { "grad_norm": 0.3938352167606354, "learning_rate": 9.391136736223085e-05, "loss": 0.0416, "step": 20080 }, { "grad_norm": 0.24641203880310059, "learning_rate": 9.390345735145956e-05, "loss": 0.0407, "step": 20090 }, { "grad_norm": 0.3832078278064728, "learning_rate": 9.389554253947219e-05, "loss": 0.0383, "step": 20100 }, { "grad_norm": 0.32283031940460205, "learning_rate": 9.388762292713428e-05, "loss": 0.0373, "step": 20110 }, { "grad_norm": 0.22005103528499603, "learning_rate": 9.38796985153119e-05, "loss": 0.0408, "step": 20120 }, { "grad_norm": 0.2666882276535034, "learning_rate": 9.387176930487169e-05, "loss": 0.0382, "step": 20130 }, { "grad_norm": 0.2208363264799118, "learning_rate": 9.386383529668072e-05, "loss": 0.039, "step": 20140 }, { "grad_norm": 0.23100338876247406, "learning_rate": 9.385589649160669e-05, "loss": 0.048, "step": 20150 }, { "grad_norm": 0.25242242217063904, "learning_rate": 9.384795289051775e-05, "loss": 0.0437, "step": 20160 }, { "grad_norm": 0.3444317877292633, "learning_rate": 9.384000449428261e-05, "loss": 0.0446, "step": 20170 }, { "grad_norm": 0.3252018690109253, "learning_rate": 9.383205130377048e-05, "loss": 0.0435, "step": 20180 }, { "grad_norm": 0.3101976811885834, "learning_rate": 9.382409331985114e-05, "loss": 0.0386, "step": 20190 }, { "grad_norm": 0.2915959656238556, "learning_rate": 9.381613054339482e-05, "loss": 0.0375, "step": 20200 }, { "grad_norm": 0.2540869414806366, "learning_rate": 9.380816297527235e-05, "loss": 0.0391, "step": 20210 }, { "grad_norm": 0.26418858766555786, "learning_rate": 9.380019061635506e-05, "loss": 0.0348, "step": 20220 }, { "grad_norm": 0.348240464925766, "learning_rate": 9.379221346751474e-05, "loss": 0.0404, "step": 20230 }, { "grad_norm": 0.32422322034835815, "learning_rate": 9.378423152962382e-05, "loss": 0.0398, "step": 20240 }, { "grad_norm": 0.3046052157878876, "learning_rate": 9.377624480355517e-05, "loss": 0.0353, "step": 20250 }, { "grad_norm": 0.3004095256328583, "learning_rate": 9.376825329018219e-05, "loss": 0.04, "step": 20260 }, { "grad_norm": 0.425052672624588, "learning_rate": 9.376025699037884e-05, "loss": 0.0391, "step": 20270 }, { "grad_norm": 0.28181880712509155, "learning_rate": 9.37522559050196e-05, "loss": 0.0385, "step": 20280 }, { "grad_norm": 0.3137870728969574, "learning_rate": 9.37442500349794e-05, "loss": 0.0423, "step": 20290 }, { "grad_norm": 0.26176440715789795, "learning_rate": 9.373623938113381e-05, "loss": 0.0378, "step": 20300 }, { "grad_norm": 0.21585726737976074, "learning_rate": 9.372822394435883e-05, "loss": 0.0389, "step": 20310 }, { "grad_norm": 0.24945320188999176, "learning_rate": 9.372020372553102e-05, "loss": 0.0358, "step": 20320 }, { "grad_norm": 0.2785467207431793, "learning_rate": 9.371217872552746e-05, "loss": 0.0425, "step": 20330 }, { "grad_norm": 0.2738749086856842, "learning_rate": 9.370414894522576e-05, "loss": 0.0332, "step": 20340 }, { "grad_norm": 0.6697885394096375, "learning_rate": 9.369611438550406e-05, "loss": 0.0354, "step": 20350 }, { "grad_norm": 0.2860734760761261, "learning_rate": 9.368807504724095e-05, "loss": 0.041, "step": 20360 }, { "grad_norm": 0.34622666239738464, "learning_rate": 9.368003093131565e-05, "loss": 0.0406, "step": 20370 }, { "grad_norm": 0.3110044598579407, "learning_rate": 9.367198203860785e-05, "loss": 0.0415, "step": 20380 }, { "grad_norm": 0.2707207500934601, "learning_rate": 9.366392836999774e-05, "loss": 0.0398, "step": 20390 }, { "grad_norm": 0.2930516302585602, "learning_rate": 9.365586992636607e-05, "loss": 0.0394, "step": 20400 }, { "grad_norm": 0.27662143111228943, "learning_rate": 9.364780670859412e-05, "loss": 0.0398, "step": 20410 }, { "grad_norm": 0.37233859300613403, "learning_rate": 9.363973871756364e-05, "loss": 0.0455, "step": 20420 }, { "grad_norm": 0.29750964045524597, "learning_rate": 9.363166595415696e-05, "loss": 0.0465, "step": 20430 }, { "grad_norm": 0.31972208619117737, "learning_rate": 9.362358841925686e-05, "loss": 0.0403, "step": 20440 }, { "grad_norm": 0.31160858273506165, "learning_rate": 9.361550611374674e-05, "loss": 0.0356, "step": 20450 }, { "grad_norm": 0.2551408112049103, "learning_rate": 9.360741903851043e-05, "loss": 0.0392, "step": 20460 }, { "grad_norm": 0.3013663589954376, "learning_rate": 9.359932719443236e-05, "loss": 0.0349, "step": 20470 }, { "grad_norm": 0.21176666021347046, "learning_rate": 9.35912305823974e-05, "loss": 0.0344, "step": 20480 }, { "grad_norm": 0.25681403279304504, "learning_rate": 9.358312920329101e-05, "loss": 0.0356, "step": 20490 }, { "grad_norm": 0.3121112287044525, "learning_rate": 9.357502305799914e-05, "loss": 0.0383, "step": 20500 }, { "grad_norm": 0.28359976410865784, "learning_rate": 9.356691214740824e-05, "loss": 0.0373, "step": 20510 }, { "grad_norm": 0.2630835175514221, "learning_rate": 9.355879647240535e-05, "loss": 0.0459, "step": 20520 }, { "grad_norm": 0.29107168316841125, "learning_rate": 9.355067603387798e-05, "loss": 0.0421, "step": 20530 }, { "grad_norm": 0.23559628427028656, "learning_rate": 9.354255083271412e-05, "loss": 0.0406, "step": 20540 }, { "grad_norm": 0.3191528022289276, "learning_rate": 9.353442086980239e-05, "loss": 0.0408, "step": 20550 }, { "grad_norm": 0.25973057746887207, "learning_rate": 9.352628614603185e-05, "loss": 0.0399, "step": 20560 }, { "grad_norm": 0.2462521344423294, "learning_rate": 9.351814666229209e-05, "loss": 0.0339, "step": 20570 }, { "grad_norm": 0.19920134544372559, "learning_rate": 9.351000241947324e-05, "loss": 0.0351, "step": 20580 }, { "grad_norm": 0.2279890477657318, "learning_rate": 9.350185341846594e-05, "loss": 0.0363, "step": 20590 }, { "grad_norm": 0.2641448378562927, "learning_rate": 9.349369966016134e-05, "loss": 0.0392, "step": 20600 }, { "grad_norm": 0.3675730228424072, "learning_rate": 9.348554114545117e-05, "loss": 0.0386, "step": 20610 }, { "grad_norm": 0.340164452791214, "learning_rate": 9.347737787522758e-05, "loss": 0.04, "step": 20620 }, { "grad_norm": 0.28064075112342834, "learning_rate": 9.346920985038332e-05, "loss": 0.0368, "step": 20630 }, { "grad_norm": 0.3078615665435791, "learning_rate": 9.346103707181162e-05, "loss": 0.0336, "step": 20640 }, { "grad_norm": 0.2874072194099426, "learning_rate": 9.345285954040626e-05, "loss": 0.0424, "step": 20650 }, { "grad_norm": 0.24762652814388275, "learning_rate": 9.34446772570615e-05, "loss": 0.0387, "step": 20660 }, { "grad_norm": 0.2735818922519684, "learning_rate": 9.343649022267214e-05, "loss": 0.0366, "step": 20670 }, { "grad_norm": 0.29747524857521057, "learning_rate": 9.342829843813353e-05, "loss": 0.0383, "step": 20680 }, { "grad_norm": 0.2686634063720703, "learning_rate": 9.342010190434149e-05, "loss": 0.0409, "step": 20690 }, { "grad_norm": 0.28798022866249084, "learning_rate": 9.34119006221924e-05, "loss": 0.0446, "step": 20700 }, { "grad_norm": 0.3446081280708313, "learning_rate": 9.340369459258313e-05, "loss": 0.0407, "step": 20710 }, { "grad_norm": 0.3400697410106659, "learning_rate": 9.339548381641106e-05, "loss": 0.0427, "step": 20720 }, { "grad_norm": 0.2783224284648895, "learning_rate": 9.338726829457413e-05, "loss": 0.0375, "step": 20730 }, { "grad_norm": 0.29492080211639404, "learning_rate": 9.337904802797078e-05, "loss": 0.0391, "step": 20740 }, { "grad_norm": 0.34762153029441833, "learning_rate": 9.337082301749993e-05, "loss": 0.0398, "step": 20750 }, { "grad_norm": 0.28806090354919434, "learning_rate": 9.336259326406109e-05, "loss": 0.0385, "step": 20760 }, { "grad_norm": 0.2355358600616455, "learning_rate": 9.335435876855427e-05, "loss": 0.037, "step": 20770 }, { "grad_norm": 0.394796758890152, "learning_rate": 9.334611953187994e-05, "loss": 0.0388, "step": 20780 }, { "grad_norm": 0.23275963962078094, "learning_rate": 9.333787555493914e-05, "loss": 0.0376, "step": 20790 }, { "grad_norm": 0.2640189528465271, "learning_rate": 9.332962683863345e-05, "loss": 0.0339, "step": 20800 }, { "grad_norm": 0.33121877908706665, "learning_rate": 9.332137338386489e-05, "loss": 0.041, "step": 20810 }, { "grad_norm": 0.2769424617290497, "learning_rate": 9.33131151915361e-05, "loss": 0.0422, "step": 20820 }, { "grad_norm": 0.3290959894657135, "learning_rate": 9.330485226255012e-05, "loss": 0.0365, "step": 20830 }, { "grad_norm": 0.37955597043037415, "learning_rate": 9.329658459781061e-05, "loss": 0.033, "step": 20840 }, { "grad_norm": 0.2225121706724167, "learning_rate": 9.328831219822172e-05, "loss": 0.0352, "step": 20850 }, { "grad_norm": 0.29501891136169434, "learning_rate": 9.328003506468808e-05, "loss": 0.0426, "step": 20860 }, { "grad_norm": 0.31270715594291687, "learning_rate": 9.327175319811488e-05, "loss": 0.0382, "step": 20870 }, { "grad_norm": 0.28572919964790344, "learning_rate": 9.326346659940781e-05, "loss": 0.0388, "step": 20880 }, { "grad_norm": 0.71739262342453, "learning_rate": 9.325517526947308e-05, "loss": 0.0375, "step": 20890 }, { "grad_norm": 0.2834027111530304, "learning_rate": 9.32468792092174e-05, "loss": 0.0418, "step": 20900 }, { "grad_norm": 0.3421855866909027, "learning_rate": 9.323857841954803e-05, "loss": 0.0369, "step": 20910 }, { "grad_norm": 0.3794124722480774, "learning_rate": 9.323027290137276e-05, "loss": 0.0402, "step": 20920 }, { "grad_norm": 0.2565392851829529, "learning_rate": 9.322196265559981e-05, "loss": 0.0349, "step": 20930 }, { "grad_norm": 0.23109018802642822, "learning_rate": 9.321364768313803e-05, "loss": 0.0425, "step": 20940 }, { "grad_norm": 0.3417356610298157, "learning_rate": 9.32053279848967e-05, "loss": 0.0339, "step": 20950 }, { "grad_norm": 0.22059808671474457, "learning_rate": 9.319700356178567e-05, "loss": 0.0382, "step": 20960 }, { "grad_norm": 0.29423972964286804, "learning_rate": 9.318867441471527e-05, "loss": 0.0314, "step": 20970 }, { "grad_norm": 0.24853020906448364, "learning_rate": 9.318034054459637e-05, "loss": 0.0353, "step": 20980 }, { "grad_norm": 0.27743223309516907, "learning_rate": 9.317200195234034e-05, "loss": 0.0369, "step": 20990 }, { "grad_norm": 0.3993227481842041, "learning_rate": 9.316365863885909e-05, "loss": 0.0352, "step": 21000 }, { "grad_norm": 0.31503450870513916, "learning_rate": 9.315531060506502e-05, "loss": 0.0388, "step": 21010 }, { "grad_norm": 0.31366729736328125, "learning_rate": 9.314695785187108e-05, "loss": 0.0376, "step": 21020 }, { "grad_norm": 0.34509575366973877, "learning_rate": 9.313860038019069e-05, "loss": 0.0409, "step": 21030 }, { "grad_norm": 0.3379817008972168, "learning_rate": 9.313023819093782e-05, "loss": 0.0426, "step": 21040 }, { "grad_norm": 0.3332487940788269, "learning_rate": 9.312187128502695e-05, "loss": 0.0364, "step": 21050 }, { "grad_norm": 0.38302600383758545, "learning_rate": 9.311349966337307e-05, "loss": 0.0421, "step": 21060 }, { "grad_norm": 0.30324140191078186, "learning_rate": 9.310512332689169e-05, "loss": 0.04, "step": 21070 }, { "grad_norm": 0.27808478474617004, "learning_rate": 9.309674227649883e-05, "loss": 0.0355, "step": 21080 }, { "grad_norm": 0.33350640535354614, "learning_rate": 9.308835651311103e-05, "loss": 0.0422, "step": 21090 }, { "grad_norm": 0.42341649532318115, "learning_rate": 9.307996603764533e-05, "loss": 0.0389, "step": 21100 }, { "grad_norm": 0.25130927562713623, "learning_rate": 9.307157085101932e-05, "loss": 0.0337, "step": 21110 }, { "grad_norm": 0.2553439140319824, "learning_rate": 9.306317095415109e-05, "loss": 0.0352, "step": 21120 }, { "grad_norm": 0.3064124286174774, "learning_rate": 9.305476634795922e-05, "loss": 0.0414, "step": 21130 }, { "grad_norm": 0.2543655037879944, "learning_rate": 9.304635703336284e-05, "loss": 0.034, "step": 21140 }, { "grad_norm": 0.2465922087430954, "learning_rate": 9.303794301128157e-05, "loss": 0.0335, "step": 21150 }, { "grad_norm": 0.2619902789592743, "learning_rate": 9.302952428263555e-05, "loss": 0.0312, "step": 21160 }, { "grad_norm": 0.21266622841358185, "learning_rate": 9.302110084834545e-05, "loss": 0.0391, "step": 21170 }, { "grad_norm": 0.2605288624763489, "learning_rate": 9.301267270933245e-05, "loss": 0.0366, "step": 21180 }, { "grad_norm": 0.31471362709999084, "learning_rate": 9.300423986651823e-05, "loss": 0.0372, "step": 21190 }, { "grad_norm": 0.22792866826057434, "learning_rate": 9.299580232082501e-05, "loss": 0.0337, "step": 21200 }, { "grad_norm": 0.31298667192459106, "learning_rate": 9.298736007317547e-05, "loss": 0.0399, "step": 21210 }, { "grad_norm": 0.3241824507713318, "learning_rate": 9.297891312449288e-05, "loss": 0.0348, "step": 21220 }, { "grad_norm": 0.3351902961730957, "learning_rate": 9.297046147570094e-05, "loss": 0.0392, "step": 21230 }, { "grad_norm": 0.30143797397613525, "learning_rate": 9.296200512772396e-05, "loss": 0.0393, "step": 21240 }, { "grad_norm": 0.2195899337530136, "learning_rate": 9.295354408148668e-05, "loss": 0.0306, "step": 21250 }, { "grad_norm": 0.24853673577308655, "learning_rate": 9.294507833791441e-05, "loss": 0.0342, "step": 21260 }, { "grad_norm": 0.23777472972869873, "learning_rate": 9.293660789793295e-05, "loss": 0.032, "step": 21270 }, { "grad_norm": 0.2293638288974762, "learning_rate": 9.292813276246858e-05, "loss": 0.0361, "step": 21280 }, { "grad_norm": 0.23646315932273865, "learning_rate": 9.291965293244816e-05, "loss": 0.0338, "step": 21290 }, { "grad_norm": 0.2827253043651581, "learning_rate": 9.291116840879904e-05, "loss": 0.0326, "step": 21300 }, { "grad_norm": 0.24106836318969727, "learning_rate": 9.290267919244904e-05, "loss": 0.0364, "step": 21310 }, { "grad_norm": 0.2549281418323517, "learning_rate": 9.289418528432655e-05, "loss": 0.033, "step": 21320 }, { "grad_norm": 0.32685422897338867, "learning_rate": 9.288568668536045e-05, "loss": 0.0382, "step": 21330 }, { "grad_norm": 0.2699177861213684, "learning_rate": 9.287718339648013e-05, "loss": 0.0358, "step": 21340 }, { "grad_norm": 0.26579269766807556, "learning_rate": 9.28686754186155e-05, "loss": 0.0333, "step": 21350 }, { "grad_norm": 0.30087190866470337, "learning_rate": 9.286016275269698e-05, "loss": 0.0333, "step": 21360 }, { "grad_norm": 0.25274330377578735, "learning_rate": 9.285164539965551e-05, "loss": 0.0362, "step": 21370 }, { "grad_norm": 0.4132579565048218, "learning_rate": 9.284312336042251e-05, "loss": 0.0405, "step": 21380 }, { "grad_norm": 0.2875617742538452, "learning_rate": 9.283459663592996e-05, "loss": 0.0368, "step": 21390 }, { "grad_norm": 0.23896822333335876, "learning_rate": 9.282606522711033e-05, "loss": 0.0373, "step": 21400 }, { "grad_norm": 0.34930479526519775, "learning_rate": 9.281752913489657e-05, "loss": 0.0358, "step": 21410 }, { "grad_norm": 0.2527092695236206, "learning_rate": 9.280898836022222e-05, "loss": 0.0343, "step": 21420 }, { "grad_norm": 0.289424866437912, "learning_rate": 9.280044290402126e-05, "loss": 0.0391, "step": 21430 }, { "grad_norm": 0.300649493932724, "learning_rate": 9.279189276722821e-05, "loss": 0.0385, "step": 21440 }, { "grad_norm": 0.20683668553829193, "learning_rate": 9.278333795077812e-05, "loss": 0.0349, "step": 21450 }, { "grad_norm": 0.3083503842353821, "learning_rate": 9.27747784556065e-05, "loss": 0.0379, "step": 21460 }, { "grad_norm": 0.2839822471141815, "learning_rate": 9.276621428264942e-05, "loss": 0.0329, "step": 21470 }, { "grad_norm": 0.2442197948694229, "learning_rate": 9.275764543284345e-05, "loss": 0.0383, "step": 21480 }, { "grad_norm": 0.27868127822875977, "learning_rate": 9.274907190712566e-05, "loss": 0.0373, "step": 21490 }, { "grad_norm": 0.38789913058280945, "learning_rate": 9.274049370643363e-05, "loss": 0.0369, "step": 21500 }, { "grad_norm": 0.3416953682899475, "learning_rate": 9.273191083170547e-05, "loss": 0.0368, "step": 21510 }, { "grad_norm": 0.2489299476146698, "learning_rate": 9.27233232838798e-05, "loss": 0.0368, "step": 21520 }, { "grad_norm": 0.5061928629875183, "learning_rate": 9.27147310638957e-05, "loss": 0.0372, "step": 21530 }, { "grad_norm": 0.2232120782136917, "learning_rate": 9.270613417269286e-05, "loss": 0.0391, "step": 21540 }, { "grad_norm": 0.31349679827690125, "learning_rate": 9.269753261121138e-05, "loss": 0.0387, "step": 21550 }, { "grad_norm": 0.3413074016571045, "learning_rate": 9.268892638039194e-05, "loss": 0.0372, "step": 21560 }, { "grad_norm": 0.2963986396789551, "learning_rate": 9.268031548117569e-05, "loss": 0.0429, "step": 21570 }, { "grad_norm": 0.30409374833106995, "learning_rate": 9.26716999145043e-05, "loss": 0.037, "step": 21580 }, { "grad_norm": 0.4351845681667328, "learning_rate": 9.266307968131998e-05, "loss": 0.038, "step": 21590 }, { "grad_norm": 0.24924889206886292, "learning_rate": 9.26544547825654e-05, "loss": 0.0443, "step": 21600 }, { "grad_norm": 0.22573573887348175, "learning_rate": 9.264582521918376e-05, "loss": 0.0338, "step": 21610 }, { "grad_norm": 0.2469622641801834, "learning_rate": 9.263719099211881e-05, "loss": 0.0318, "step": 21620 }, { "grad_norm": 0.24908412992954254, "learning_rate": 9.262855210231476e-05, "loss": 0.0387, "step": 21630 }, { "grad_norm": 0.6755164861679077, "learning_rate": 9.261990855071633e-05, "loss": 0.041, "step": 21640 }, { "grad_norm": 0.3097418546676636, "learning_rate": 9.261126033826878e-05, "loss": 0.0366, "step": 21650 }, { "grad_norm": 0.32765936851501465, "learning_rate": 9.260260746591786e-05, "loss": 0.0448, "step": 21660 }, { "grad_norm": 0.3664017915725708, "learning_rate": 9.259394993460985e-05, "loss": 0.0397, "step": 21670 }, { "grad_norm": 0.2276221662759781, "learning_rate": 9.258528774529151e-05, "loss": 0.0406, "step": 21680 }, { "grad_norm": 0.2741946578025818, "learning_rate": 9.257662089891013e-05, "loss": 0.0375, "step": 21690 }, { "grad_norm": 0.29118138551712036, "learning_rate": 9.25679493964135e-05, "loss": 0.0302, "step": 21700 }, { "grad_norm": 0.31701818108558655, "learning_rate": 9.255927323874994e-05, "loss": 0.0418, "step": 21710 }, { "grad_norm": 0.2740970551967621, "learning_rate": 9.255059242686822e-05, "loss": 0.0357, "step": 21720 }, { "grad_norm": 0.2721051871776581, "learning_rate": 9.254190696171769e-05, "loss": 0.0329, "step": 21730 }, { "grad_norm": 0.26048949360847473, "learning_rate": 9.25332168442482e-05, "loss": 0.0371, "step": 21740 }, { "grad_norm": 0.30258867144584656, "learning_rate": 9.252452207541004e-05, "loss": 0.0368, "step": 21750 }, { "grad_norm": 0.27723273634910583, "learning_rate": 9.251582265615409e-05, "loss": 0.0371, "step": 21760 }, { "grad_norm": 0.23352928459644318, "learning_rate": 9.250711858743169e-05, "loss": 0.0351, "step": 21770 }, { "grad_norm": 0.303081214427948, "learning_rate": 9.24984098701947e-05, "loss": 0.0401, "step": 21780 }, { "grad_norm": 0.3232305347919464, "learning_rate": 9.248969650539552e-05, "loss": 0.0332, "step": 21790 }, { "grad_norm": 0.33593228459358215, "learning_rate": 9.2480978493987e-05, "loss": 0.0364, "step": 21800 }, { "grad_norm": 0.2037256509065628, "learning_rate": 9.247225583692256e-05, "loss": 0.0351, "step": 21810 }, { "grad_norm": 0.29411378502845764, "learning_rate": 9.246352853515607e-05, "loss": 0.0353, "step": 21820 }, { "grad_norm": 0.30124613642692566, "learning_rate": 9.245479658964194e-05, "loss": 0.0327, "step": 21830 }, { "grad_norm": 0.21216446161270142, "learning_rate": 9.244606000133507e-05, "loss": 0.0386, "step": 21840 }, { "grad_norm": 0.20837976038455963, "learning_rate": 9.24373187711909e-05, "loss": 0.0354, "step": 21850 }, { "grad_norm": 0.3082909882068634, "learning_rate": 9.242857290016537e-05, "loss": 0.0381, "step": 21860 }, { "grad_norm": 0.30018165707588196, "learning_rate": 9.241982238921488e-05, "loss": 0.0352, "step": 21870 }, { "grad_norm": 0.2698197066783905, "learning_rate": 9.24110672392964e-05, "loss": 0.0415, "step": 21880 }, { "grad_norm": 0.2745720148086548, "learning_rate": 9.240230745136737e-05, "loss": 0.0336, "step": 21890 }, { "grad_norm": 0.2764868140220642, "learning_rate": 9.239354302638575e-05, "loss": 0.032, "step": 21900 }, { "grad_norm": 0.2721467614173889, "learning_rate": 9.238477396531e-05, "loss": 0.0336, "step": 21910 }, { "grad_norm": 0.27385640144348145, "learning_rate": 9.23760002690991e-05, "loss": 0.0332, "step": 21920 }, { "grad_norm": 0.27219387888908386, "learning_rate": 9.236722193871252e-05, "loss": 0.0333, "step": 21930 }, { "grad_norm": 0.3121076226234436, "learning_rate": 9.235843897511023e-05, "loss": 0.0322, "step": 21940 }, { "grad_norm": 0.43406760692596436, "learning_rate": 9.234965137925276e-05, "loss": 0.0392, "step": 21950 }, { "grad_norm": 0.24135075509548187, "learning_rate": 9.234085915210108e-05, "loss": 0.0344, "step": 21960 }, { "grad_norm": 0.3258422017097473, "learning_rate": 9.23320622946167e-05, "loss": 0.0382, "step": 21970 }, { "grad_norm": 0.2540751099586487, "learning_rate": 9.232326080776163e-05, "loss": 0.0353, "step": 21980 }, { "grad_norm": 0.2112571895122528, "learning_rate": 9.23144546924984e-05, "loss": 0.0397, "step": 21990 }, { "grad_norm": 0.27107492089271545, "learning_rate": 9.230564394979e-05, "loss": 0.0337, "step": 22000 }, { "grad_norm": 0.29364025592803955, "learning_rate": 9.22968285806e-05, "loss": 0.0332, "step": 22010 }, { "grad_norm": 0.3128189742565155, "learning_rate": 9.228800858589242e-05, "loss": 0.0352, "step": 22020 }, { "grad_norm": 0.19987963140010834, "learning_rate": 9.227918396663179e-05, "loss": 0.0355, "step": 22030 }, { "grad_norm": 0.3186514973640442, "learning_rate": 9.227035472378319e-05, "loss": 0.0409, "step": 22040 }, { "grad_norm": 0.2434864342212677, "learning_rate": 9.226152085831213e-05, "loss": 0.0371, "step": 22050 }, { "grad_norm": 0.2826649248600006, "learning_rate": 9.22526823711847e-05, "loss": 0.0387, "step": 22060 }, { "grad_norm": 0.2557315230369568, "learning_rate": 9.224383926336745e-05, "loss": 0.0353, "step": 22070 }, { "grad_norm": 0.2948925197124481, "learning_rate": 9.223499153582744e-05, "loss": 0.0371, "step": 22080 }, { "grad_norm": 0.2840483486652374, "learning_rate": 9.222613918953226e-05, "loss": 0.0374, "step": 22090 }, { "grad_norm": 0.3085497319698334, "learning_rate": 9.221728222544999e-05, "loss": 0.042, "step": 22100 }, { "grad_norm": 0.2749359905719757, "learning_rate": 9.22084206445492e-05, "loss": 0.0387, "step": 22110 }, { "grad_norm": 0.3190944790840149, "learning_rate": 9.2199554447799e-05, "loss": 0.0364, "step": 22120 }, { "grad_norm": 0.2908312678337097, "learning_rate": 9.219068363616897e-05, "loss": 0.0359, "step": 22130 }, { "grad_norm": 0.2354220300912857, "learning_rate": 9.218180821062919e-05, "loss": 0.0347, "step": 22140 }, { "grad_norm": 0.2671967148780823, "learning_rate": 9.21729281721503e-05, "loss": 0.0381, "step": 22150 }, { "grad_norm": 0.22749872505664825, "learning_rate": 9.216404352170339e-05, "loss": 0.0441, "step": 22160 }, { "grad_norm": 0.3058473467826843, "learning_rate": 9.215515426026007e-05, "loss": 0.0374, "step": 22170 }, { "grad_norm": 0.24591337144374847, "learning_rate": 9.214626038879246e-05, "loss": 0.0375, "step": 22180 }, { "grad_norm": 0.3506157398223877, "learning_rate": 9.21373619082732e-05, "loss": 0.0408, "step": 22190 }, { "grad_norm": 0.36473748087882996, "learning_rate": 9.212845881967535e-05, "loss": 0.0425, "step": 22200 }, { "grad_norm": 0.3234219551086426, "learning_rate": 9.211955112397262e-05, "loss": 0.0378, "step": 22210 }, { "grad_norm": 0.2867027521133423, "learning_rate": 9.211063882213909e-05, "loss": 0.0375, "step": 22220 }, { "grad_norm": 0.32291778922080994, "learning_rate": 9.210172191514942e-05, "loss": 0.0426, "step": 22230 }, { "grad_norm": 0.28804248571395874, "learning_rate": 9.209280040397874e-05, "loss": 0.0413, "step": 22240 }, { "grad_norm": 0.37124887108802795, "learning_rate": 9.208387428960268e-05, "loss": 0.0397, "step": 22250 }, { "grad_norm": 0.3652448356151581, "learning_rate": 9.20749435729974e-05, "loss": 0.0399, "step": 22260 }, { "grad_norm": 0.48961764574050903, "learning_rate": 9.206600825513957e-05, "loss": 0.0453, "step": 22270 }, { "grad_norm": 0.34079885482788086, "learning_rate": 9.20570683370063e-05, "loss": 0.0372, "step": 22280 }, { "grad_norm": 0.2965315878391266, "learning_rate": 9.204812381957528e-05, "loss": 0.0415, "step": 22290 }, { "grad_norm": 0.28549134731292725, "learning_rate": 9.203917470382465e-05, "loss": 0.0437, "step": 22300 }, { "grad_norm": 0.2655712366104126, "learning_rate": 9.203022099073309e-05, "loss": 0.0348, "step": 22310 }, { "grad_norm": 0.3712618947029114, "learning_rate": 9.202126268127976e-05, "loss": 0.0386, "step": 22320 }, { "grad_norm": 0.3451271653175354, "learning_rate": 9.20122997764443e-05, "loss": 0.0345, "step": 22330 }, { "grad_norm": 0.3271554410457611, "learning_rate": 9.200333227720692e-05, "loss": 0.037, "step": 22340 }, { "grad_norm": 0.3223828971385956, "learning_rate": 9.199436018454826e-05, "loss": 0.0427, "step": 22350 }, { "grad_norm": 0.358338862657547, "learning_rate": 9.198538349944952e-05, "loss": 0.0371, "step": 22360 }, { "grad_norm": 0.3274436295032501, "learning_rate": 9.197640222289234e-05, "loss": 0.0348, "step": 22370 }, { "grad_norm": 0.3911176025867462, "learning_rate": 9.196741635585895e-05, "loss": 0.0383, "step": 22380 }, { "grad_norm": 0.37504151463508606, "learning_rate": 9.195842589933199e-05, "loss": 0.033, "step": 22390 }, { "grad_norm": 0.40893805027008057, "learning_rate": 9.194943085429466e-05, "loss": 0.0355, "step": 22400 }, { "grad_norm": 0.2645050883293152, "learning_rate": 9.194043122173065e-05, "loss": 0.0403, "step": 22410 }, { "grad_norm": 0.3140178322792053, "learning_rate": 9.193142700262413e-05, "loss": 0.0366, "step": 22420 }, { "grad_norm": 0.2716907262802124, "learning_rate": 9.192241819795979e-05, "loss": 0.0338, "step": 22430 }, { "grad_norm": 0.307184100151062, "learning_rate": 9.191340480872284e-05, "loss": 0.0372, "step": 22440 }, { "grad_norm": 0.2453993558883667, "learning_rate": 9.190438683589895e-05, "loss": 0.0334, "step": 22450 }, { "grad_norm": 0.3429996371269226, "learning_rate": 9.189536428047432e-05, "loss": 0.0368, "step": 22460 }, { "grad_norm": 0.2695983946323395, "learning_rate": 9.188633714343564e-05, "loss": 0.0361, "step": 22470 }, { "grad_norm": 0.3499194085597992, "learning_rate": 9.18773054257701e-05, "loss": 0.0369, "step": 22480 }, { "grad_norm": 0.2463364601135254, "learning_rate": 9.18682691284654e-05, "loss": 0.0322, "step": 22490 }, { "grad_norm": 0.2021474838256836, "learning_rate": 9.185922825250974e-05, "loss": 0.0358, "step": 22500 }, { "grad_norm": 0.3247249722480774, "learning_rate": 9.185018279889181e-05, "loss": 0.0385, "step": 22510 }, { "grad_norm": 0.2220538854598999, "learning_rate": 9.184113276860082e-05, "loss": 0.0374, "step": 22520 }, { "grad_norm": 0.28703197836875916, "learning_rate": 9.183207816262645e-05, "loss": 0.0408, "step": 22530 }, { "grad_norm": 0.2958586812019348, "learning_rate": 9.182301898195891e-05, "loss": 0.0408, "step": 22540 }, { "grad_norm": 0.2631906270980835, "learning_rate": 9.181395522758889e-05, "loss": 0.0308, "step": 22550 }, { "grad_norm": 0.21469014883041382, "learning_rate": 9.180488690050759e-05, "loss": 0.0382, "step": 22560 }, { "grad_norm": 0.31366613507270813, "learning_rate": 9.179581400170671e-05, "loss": 0.0335, "step": 22570 }, { "grad_norm": 0.29102277755737305, "learning_rate": 9.178673653217845e-05, "loss": 0.0349, "step": 22580 }, { "grad_norm": 0.2730175852775574, "learning_rate": 9.177765449291551e-05, "loss": 0.0408, "step": 22590 }, { "grad_norm": 0.2584906220436096, "learning_rate": 9.176856788491109e-05, "loss": 0.0366, "step": 22600 }, { "grad_norm": 0.3593570590019226, "learning_rate": 9.175947670915887e-05, "loss": 0.0376, "step": 22610 }, { "grad_norm": 0.30204275250434875, "learning_rate": 9.175038096665309e-05, "loss": 0.0451, "step": 22620 }, { "grad_norm": 0.2550041377544403, "learning_rate": 9.17412806583884e-05, "loss": 0.0337, "step": 22630 }, { "grad_norm": 0.2997146248817444, "learning_rate": 9.173217578536002e-05, "loss": 0.0324, "step": 22640 }, { "grad_norm": 0.24509337544441223, "learning_rate": 9.172306634856362e-05, "loss": 0.043, "step": 22650 }, { "grad_norm": 0.38749873638153076, "learning_rate": 9.171395234899545e-05, "loss": 0.0347, "step": 22660 }, { "grad_norm": 0.27878400683403015, "learning_rate": 9.170483378765214e-05, "loss": 0.0387, "step": 22670 }, { "grad_norm": 0.29664352536201477, "learning_rate": 9.169571066553091e-05, "loss": 0.0384, "step": 22680 }, { "grad_norm": 0.32104602456092834, "learning_rate": 9.168658298362946e-05, "loss": 0.0361, "step": 22690 }, { "grad_norm": 0.2426285594701767, "learning_rate": 9.167745074294598e-05, "loss": 0.0392, "step": 22700 }, { "grad_norm": 0.32837992906570435, "learning_rate": 9.166831394447913e-05, "loss": 0.0336, "step": 22710 }, { "grad_norm": 0.28295522928237915, "learning_rate": 9.165917258922812e-05, "loss": 0.0348, "step": 22720 }, { "grad_norm": 0.3304803967475891, "learning_rate": 9.165002667819262e-05, "loss": 0.0348, "step": 22730 }, { "grad_norm": 0.24879670143127441, "learning_rate": 9.164087621237282e-05, "loss": 0.0342, "step": 22740 }, { "grad_norm": 0.2957207262516022, "learning_rate": 9.163172119276942e-05, "loss": 0.035, "step": 22750 }, { "grad_norm": 0.3386390209197998, "learning_rate": 9.162256162038358e-05, "loss": 0.0426, "step": 22760 }, { "grad_norm": 0.2948877811431885, "learning_rate": 9.161339749621698e-05, "loss": 0.039, "step": 22770 }, { "grad_norm": 0.26652008295059204, "learning_rate": 9.160422882127177e-05, "loss": 0.0385, "step": 22780 }, { "grad_norm": 0.2940284311771393, "learning_rate": 9.159505559655069e-05, "loss": 0.04, "step": 22790 }, { "grad_norm": 0.24131229519844055, "learning_rate": 9.158587782305684e-05, "loss": 0.0373, "step": 22800 }, { "grad_norm": 0.29435861110687256, "learning_rate": 9.157669550179391e-05, "loss": 0.0373, "step": 22810 }, { "grad_norm": 0.28061893582344055, "learning_rate": 9.156750863376609e-05, "loss": 0.0441, "step": 22820 }, { "grad_norm": 0.3214043974876404, "learning_rate": 9.155831721997801e-05, "loss": 0.0407, "step": 22830 }, { "grad_norm": 0.27465739846229553, "learning_rate": 9.154912126143484e-05, "loss": 0.0435, "step": 22840 }, { "grad_norm": 0.2578173577785492, "learning_rate": 9.153992075914224e-05, "loss": 0.0456, "step": 22850 }, { "grad_norm": 0.27958759665489197, "learning_rate": 9.153071571410635e-05, "loss": 0.0334, "step": 22860 }, { "grad_norm": 0.31444206833839417, "learning_rate": 9.152150612733384e-05, "loss": 0.0386, "step": 22870 }, { "grad_norm": 0.32007643580436707, "learning_rate": 9.151229199983184e-05, "loss": 0.0372, "step": 22880 }, { "grad_norm": 0.30069321393966675, "learning_rate": 9.150307333260802e-05, "loss": 0.0379, "step": 22890 }, { "grad_norm": 0.293183833360672, "learning_rate": 9.149385012667048e-05, "loss": 0.0363, "step": 22900 }, { "grad_norm": 0.24725690484046936, "learning_rate": 9.148462238302788e-05, "loss": 0.0362, "step": 22910 }, { "grad_norm": 0.34958845376968384, "learning_rate": 9.147539010268936e-05, "loss": 0.0368, "step": 22920 }, { "grad_norm": 0.26869964599609375, "learning_rate": 9.14661532866645e-05, "loss": 0.037, "step": 22930 }, { "grad_norm": 0.21783758699893951, "learning_rate": 9.145691193596348e-05, "loss": 0.0344, "step": 22940 }, { "grad_norm": 0.3126935064792633, "learning_rate": 9.144766605159691e-05, "loss": 0.0333, "step": 22950 }, { "grad_norm": 0.26317647099494934, "learning_rate": 9.14384156345759e-05, "loss": 0.0344, "step": 22960 }, { "grad_norm": 0.3378838002681732, "learning_rate": 9.142916068591204e-05, "loss": 0.0327, "step": 22970 }, { "grad_norm": 0.28905388712882996, "learning_rate": 9.141990120661746e-05, "loss": 0.0296, "step": 22980 }, { "grad_norm": 0.28713512420654297, "learning_rate": 9.141063719770475e-05, "loss": 0.0402, "step": 22990 }, { "grad_norm": 0.4686838686466217, "learning_rate": 9.140136866018704e-05, "loss": 0.0331, "step": 23000 }, { "grad_norm": 0.31182652711868286, "learning_rate": 9.139209559507788e-05, "loss": 0.033, "step": 23010 }, { "grad_norm": 0.27693551778793335, "learning_rate": 9.13828180033914e-05, "loss": 0.035, "step": 23020 }, { "grad_norm": 0.23281258344650269, "learning_rate": 9.137353588614212e-05, "loss": 0.0319, "step": 23030 }, { "grad_norm": 0.29330679774284363, "learning_rate": 9.136424924434519e-05, "loss": 0.0399, "step": 23040 }, { "grad_norm": 0.5869851112365723, "learning_rate": 9.135495807901615e-05, "loss": 0.0392, "step": 23050 }, { "grad_norm": 0.3879142105579376, "learning_rate": 9.134566239117108e-05, "loss": 0.0444, "step": 23060 }, { "grad_norm": 0.2808111011981964, "learning_rate": 9.13363621818265e-05, "loss": 0.0375, "step": 23070 }, { "grad_norm": 0.37866654992103577, "learning_rate": 9.132705745199953e-05, "loss": 0.0449, "step": 23080 }, { "grad_norm": 0.24745304882526398, "learning_rate": 9.131774820270768e-05, "loss": 0.0348, "step": 23090 }, { "grad_norm": 0.27307307720184326, "learning_rate": 9.130843443496901e-05, "loss": 0.0342, "step": 23100 }, { "grad_norm": 0.28194111585617065, "learning_rate": 9.129911614980206e-05, "loss": 0.04, "step": 23110 }, { "grad_norm": 0.32046055793762207, "learning_rate": 9.128979334822584e-05, "loss": 0.0369, "step": 23120 }, { "grad_norm": 0.32565826177597046, "learning_rate": 9.128046603125992e-05, "loss": 0.0446, "step": 23130 }, { "grad_norm": 0.3192822337150574, "learning_rate": 9.12711341999243e-05, "loss": 0.0406, "step": 23140 }, { "grad_norm": 0.2614830732345581, "learning_rate": 9.12617978552395e-05, "loss": 0.0367, "step": 23150 }, { "grad_norm": 0.28604909777641296, "learning_rate": 9.12524569982265e-05, "loss": 0.0317, "step": 23160 }, { "grad_norm": 0.23813988268375397, "learning_rate": 9.124311162990684e-05, "loss": 0.034, "step": 23170 }, { "grad_norm": 0.37138301134109497, "learning_rate": 9.12337617513025e-05, "loss": 0.0339, "step": 23180 }, { "grad_norm": 0.29792267084121704, "learning_rate": 9.122440736343596e-05, "loss": 0.0298, "step": 23190 }, { "grad_norm": 0.3265335261821747, "learning_rate": 9.12150484673302e-05, "loss": 0.0306, "step": 23200 }, { "grad_norm": 0.3291257917881012, "learning_rate": 9.120568506400873e-05, "loss": 0.0348, "step": 23210 }, { "grad_norm": 0.3803579807281494, "learning_rate": 9.119631715449548e-05, "loss": 0.033, "step": 23220 }, { "grad_norm": 0.27002066373825073, "learning_rate": 9.118694473981493e-05, "loss": 0.0349, "step": 23230 }, { "grad_norm": 0.2718620300292969, "learning_rate": 9.117756782099203e-05, "loss": 0.0394, "step": 23240 }, { "grad_norm": 0.23743169009685516, "learning_rate": 9.11681863990522e-05, "loss": 0.0308, "step": 23250 }, { "grad_norm": 0.24607691168785095, "learning_rate": 9.115880047502142e-05, "loss": 0.0336, "step": 23260 }, { "grad_norm": 0.2616867125034332, "learning_rate": 9.114941004992609e-05, "loss": 0.032, "step": 23270 }, { "grad_norm": 0.3469882905483246, "learning_rate": 9.114001512479317e-05, "loss": 0.0379, "step": 23280 }, { "grad_norm": 0.2784469723701477, "learning_rate": 9.113061570065003e-05, "loss": 0.037, "step": 23290 }, { "grad_norm": 0.2671855688095093, "learning_rate": 9.112121177852459e-05, "loss": 0.0356, "step": 23300 }, { "grad_norm": 0.2977921664714813, "learning_rate": 9.111180335944527e-05, "loss": 0.037, "step": 23310 }, { "grad_norm": 0.2970384955406189, "learning_rate": 9.110239044444093e-05, "loss": 0.0344, "step": 23320 }, { "grad_norm": 0.2543644309043884, "learning_rate": 9.109297303454099e-05, "loss": 0.0329, "step": 23330 }, { "grad_norm": 0.3919602334499359, "learning_rate": 9.108355113077526e-05, "loss": 0.0418, "step": 23340 }, { "grad_norm": 0.35632508993148804, "learning_rate": 9.107412473417419e-05, "loss": 0.0443, "step": 23350 }, { "grad_norm": 0.4256526231765747, "learning_rate": 9.106469384576858e-05, "loss": 0.0372, "step": 23360 }, { "grad_norm": 0.27858778834342957, "learning_rate": 9.105525846658978e-05, "loss": 0.0353, "step": 23370 }, { "grad_norm": 0.42289942502975464, "learning_rate": 9.104581859766965e-05, "loss": 0.0448, "step": 23380 }, { "grad_norm": 0.3609008193016052, "learning_rate": 9.10363742400405e-05, "loss": 0.0437, "step": 23390 }, { "grad_norm": 0.29967552423477173, "learning_rate": 9.102692539473518e-05, "loss": 0.0407, "step": 23400 }, { "grad_norm": 0.2290559709072113, "learning_rate": 9.101747206278697e-05, "loss": 0.038, "step": 23410 }, { "grad_norm": 0.2539924681186676, "learning_rate": 9.100801424522968e-05, "loss": 0.0338, "step": 23420 }, { "grad_norm": 0.2687823176383972, "learning_rate": 9.099855194309762e-05, "loss": 0.0399, "step": 23430 }, { "grad_norm": 0.36850738525390625, "learning_rate": 9.098908515742554e-05, "loss": 0.0387, "step": 23440 }, { "grad_norm": 0.2559226453304291, "learning_rate": 9.097961388924873e-05, "loss": 0.0352, "step": 23450 }, { "grad_norm": 0.36735254526138306, "learning_rate": 9.097013813960298e-05, "loss": 0.037, "step": 23460 }, { "grad_norm": 0.33498328924179077, "learning_rate": 9.09606579095245e-05, "loss": 0.0346, "step": 23470 }, { "grad_norm": 0.22375880181789398, "learning_rate": 9.095117320005008e-05, "loss": 0.0314, "step": 23480 }, { "grad_norm": 0.24807807803153992, "learning_rate": 9.094168401221691e-05, "loss": 0.0318, "step": 23490 }, { "grad_norm": 0.25098028779029846, "learning_rate": 9.093219034706273e-05, "loss": 0.0334, "step": 23500 }, { "grad_norm": 0.22796814143657684, "learning_rate": 9.092269220562577e-05, "loss": 0.0309, "step": 23510 }, { "grad_norm": 0.36283791065216064, "learning_rate": 9.09131895889447e-05, "loss": 0.0367, "step": 23520 }, { "grad_norm": 0.24293015897274017, "learning_rate": 9.090368249805873e-05, "loss": 0.0304, "step": 23530 }, { "grad_norm": 0.3531458377838135, "learning_rate": 9.089417093400754e-05, "loss": 0.0364, "step": 23540 }, { "grad_norm": 0.2737610340118408, "learning_rate": 9.088465489783131e-05, "loss": 0.0388, "step": 23550 }, { "grad_norm": 0.2544586658477783, "learning_rate": 9.087513439057068e-05, "loss": 0.039, "step": 23560 }, { "grad_norm": 0.308885782957077, "learning_rate": 9.08656094132668e-05, "loss": 0.0287, "step": 23570 }, { "grad_norm": 0.23705455660820007, "learning_rate": 9.085607996696134e-05, "loss": 0.0307, "step": 23580 }, { "grad_norm": 0.30507853627204895, "learning_rate": 9.084654605269639e-05, "loss": 0.035, "step": 23590 }, { "grad_norm": 0.2883889675140381, "learning_rate": 9.083700767151457e-05, "loss": 0.0386, "step": 23600 }, { "grad_norm": 0.2557024359703064, "learning_rate": 9.082746482445898e-05, "loss": 0.0331, "step": 23610 }, { "grad_norm": 0.3429860770702362, "learning_rate": 9.081791751257325e-05, "loss": 0.0304, "step": 23620 }, { "grad_norm": 0.27696630358695984, "learning_rate": 9.080836573690142e-05, "loss": 0.0357, "step": 23630 }, { "grad_norm": 0.2566133141517639, "learning_rate": 9.079880949848805e-05, "loss": 0.0347, "step": 23640 }, { "grad_norm": 0.21642857789993286, "learning_rate": 9.078924879837822e-05, "loss": 0.0335, "step": 23650 }, { "grad_norm": 0.30945512652397156, "learning_rate": 9.077968363761747e-05, "loss": 0.0394, "step": 23660 }, { "grad_norm": 0.43371865153312683, "learning_rate": 9.077011401725182e-05, "loss": 0.0336, "step": 23670 }, { "grad_norm": 0.27214089035987854, "learning_rate": 9.07605399383278e-05, "loss": 0.0385, "step": 23680 }, { "grad_norm": 0.23398229479789734, "learning_rate": 9.075096140189243e-05, "loss": 0.037, "step": 23690 }, { "grad_norm": 0.27576160430908203, "learning_rate": 9.074137840899318e-05, "loss": 0.032, "step": 23700 }, { "grad_norm": 0.8788318037986755, "learning_rate": 9.073179096067804e-05, "loss": 0.0386, "step": 23710 }, { "grad_norm": 0.440986305475235, "learning_rate": 9.072219905799549e-05, "loss": 0.0336, "step": 23720 }, { "grad_norm": 0.3195445239543915, "learning_rate": 9.071260270199447e-05, "loss": 0.0387, "step": 23730 }, { "grad_norm": 0.2726370394229889, "learning_rate": 9.070300189372441e-05, "loss": 0.0384, "step": 23740 }, { "grad_norm": 0.37982895970344543, "learning_rate": 9.069339663423528e-05, "loss": 0.0384, "step": 23750 }, { "grad_norm": 0.3492242097854614, "learning_rate": 9.068378692457747e-05, "loss": 0.0427, "step": 23760 }, { "grad_norm": 0.3805731534957886, "learning_rate": 9.067417276580189e-05, "loss": 0.0386, "step": 23770 }, { "grad_norm": 0.2459040880203247, "learning_rate": 9.066455415895993e-05, "loss": 0.0362, "step": 23780 }, { "grad_norm": 0.46186840534210205, "learning_rate": 9.065493110510346e-05, "loss": 0.0381, "step": 23790 }, { "grad_norm": 0.24166351556777954, "learning_rate": 9.064530360528484e-05, "loss": 0.0464, "step": 23800 }, { "grad_norm": 0.36312076449394226, "learning_rate": 9.063567166055695e-05, "loss": 0.0357, "step": 23810 }, { "grad_norm": 0.3190240263938904, "learning_rate": 9.062603527197308e-05, "loss": 0.038, "step": 23820 }, { "grad_norm": 0.29615476727485657, "learning_rate": 9.06163944405871e-05, "loss": 0.0356, "step": 23830 }, { "grad_norm": 0.2965685725212097, "learning_rate": 9.060674916745327e-05, "loss": 0.0408, "step": 23840 }, { "grad_norm": 0.2683122158050537, "learning_rate": 9.05970994536264e-05, "loss": 0.0399, "step": 23850 }, { "grad_norm": 0.2599143087863922, "learning_rate": 9.05874453001618e-05, "loss": 0.035, "step": 23860 }, { "grad_norm": 0.32164299488067627, "learning_rate": 9.057778670811517e-05, "loss": 0.034, "step": 23870 }, { "grad_norm": 0.30959388613700867, "learning_rate": 9.056812367854281e-05, "loss": 0.0367, "step": 23880 }, { "grad_norm": 0.3656349778175354, "learning_rate": 9.055845621250143e-05, "loss": 0.0422, "step": 23890 }, { "grad_norm": 0.2674049735069275, "learning_rate": 9.054878431104825e-05, "loss": 0.0377, "step": 23900 }, { "grad_norm": 0.389810711145401, "learning_rate": 9.0539107975241e-05, "loss": 0.0388, "step": 23910 }, { "grad_norm": 0.3439479470252991, "learning_rate": 9.052942720613784e-05, "loss": 0.0333, "step": 23920 }, { "grad_norm": 0.27509206533432007, "learning_rate": 9.051974200479745e-05, "loss": 0.0351, "step": 23930 }, { "grad_norm": 0.27946603298187256, "learning_rate": 9.051005237227901e-05, "loss": 0.0359, "step": 23940 }, { "grad_norm": 0.2666011452674866, "learning_rate": 9.050035830964215e-05, "loss": 0.0333, "step": 23950 }, { "grad_norm": 0.2871773838996887, "learning_rate": 9.049065981794698e-05, "loss": 0.036, "step": 23960 }, { "grad_norm": 0.20868735015392303, "learning_rate": 9.048095689825414e-05, "loss": 0.0341, "step": 23970 }, { "grad_norm": 0.2882884442806244, "learning_rate": 9.047124955162472e-05, "loss": 0.0329, "step": 23980 }, { "grad_norm": 0.22366183996200562, "learning_rate": 9.046153777912028e-05, "loss": 0.0351, "step": 23990 }, { "grad_norm": 0.30850788950920105, "learning_rate": 9.045182158180292e-05, "loss": 0.0354, "step": 24000 }, { "grad_norm": 0.2629830539226532, "learning_rate": 9.044210096073516e-05, "loss": 0.0328, "step": 24010 }, { "grad_norm": 0.4839721620082855, "learning_rate": 9.043237591698004e-05, "loss": 0.0421, "step": 24020 }, { "grad_norm": 0.3189566433429718, "learning_rate": 9.04226464516011e-05, "loss": 0.0314, "step": 24030 }, { "grad_norm": 0.2785499095916748, "learning_rate": 9.041291256566229e-05, "loss": 0.0375, "step": 24040 }, { "grad_norm": 0.3128184378147125, "learning_rate": 9.040317426022814e-05, "loss": 0.031, "step": 24050 }, { "grad_norm": 0.33281058073043823, "learning_rate": 9.03934315363636e-05, "loss": 0.0335, "step": 24060 }, { "grad_norm": 0.28156864643096924, "learning_rate": 9.038368439513409e-05, "loss": 0.0359, "step": 24070 }, { "grad_norm": 0.23099753260612488, "learning_rate": 9.03739328376056e-05, "loss": 0.0385, "step": 24080 }, { "grad_norm": 0.2530372440814972, "learning_rate": 9.036417686484451e-05, "loss": 0.0364, "step": 24090 }, { "grad_norm": 0.36455872654914856, "learning_rate": 9.035441647791773e-05, "loss": 0.0365, "step": 24100 }, { "grad_norm": 0.3461378216743469, "learning_rate": 9.034465167789263e-05, "loss": 0.0378, "step": 24110 }, { "grad_norm": 0.2613055408000946, "learning_rate": 9.033488246583706e-05, "loss": 0.0321, "step": 24120 }, { "grad_norm": 0.21068429946899414, "learning_rate": 9.032510884281941e-05, "loss": 0.0296, "step": 24130 }, { "grad_norm": 0.27776896953582764, "learning_rate": 9.031533080990848e-05, "loss": 0.0345, "step": 24140 }, { "grad_norm": 0.21990814805030823, "learning_rate": 9.030554836817358e-05, "loss": 0.0312, "step": 24150 }, { "grad_norm": 0.2915439307689667, "learning_rate": 9.029576151868451e-05, "loss": 0.0326, "step": 24160 }, { "grad_norm": 0.20194970071315765, "learning_rate": 9.028597026251155e-05, "loss": 0.0303, "step": 24170 }, { "grad_norm": 0.2517285645008087, "learning_rate": 9.027617460072547e-05, "loss": 0.0342, "step": 24180 }, { "grad_norm": 0.24468547105789185, "learning_rate": 9.026637453439745e-05, "loss": 0.0344, "step": 24190 }, { "grad_norm": 0.214238703250885, "learning_rate": 9.025657006459927e-05, "loss": 0.0284, "step": 24200 }, { "grad_norm": 0.27175983786582947, "learning_rate": 9.024676119240311e-05, "loss": 0.0354, "step": 24210 }, { "grad_norm": 0.23570401966571808, "learning_rate": 9.023694791888166e-05, "loss": 0.0301, "step": 24220 }, { "grad_norm": 0.292179673910141, "learning_rate": 9.022713024510808e-05, "loss": 0.0331, "step": 24230 }, { "grad_norm": 0.1878628134727478, "learning_rate": 9.021730817215601e-05, "loss": 0.0371, "step": 24240 }, { "grad_norm": 0.3441106677055359, "learning_rate": 9.02074817010996e-05, "loss": 0.0344, "step": 24250 }, { "grad_norm": 0.2792530059814453, "learning_rate": 9.019765083301342e-05, "loss": 0.0299, "step": 24260 }, { "grad_norm": 0.30244180560112, "learning_rate": 9.01878155689726e-05, "loss": 0.0314, "step": 24270 }, { "grad_norm": 0.2672591507434845, "learning_rate": 9.017797591005268e-05, "loss": 0.0305, "step": 24280 }, { "grad_norm": 0.31165850162506104, "learning_rate": 9.016813185732972e-05, "loss": 0.0346, "step": 24290 }, { "grad_norm": 0.27066299319267273, "learning_rate": 9.015828341188027e-05, "loss": 0.0314, "step": 24300 }, { "grad_norm": 0.3194079101085663, "learning_rate": 9.01484305747813e-05, "loss": 0.0363, "step": 24310 }, { "grad_norm": 0.24699652194976807, "learning_rate": 9.013857334711033e-05, "loss": 0.0335, "step": 24320 }, { "grad_norm": 0.3150935769081116, "learning_rate": 9.012871172994534e-05, "loss": 0.036, "step": 24330 }, { "grad_norm": 0.2816064655780792, "learning_rate": 9.011884572436476e-05, "loss": 0.0379, "step": 24340 }, { "grad_norm": 0.3180297315120697, "learning_rate": 9.010897533144754e-05, "loss": 0.033, "step": 24350 }, { "grad_norm": 0.25423723459243774, "learning_rate": 9.009910055227306e-05, "loss": 0.0333, "step": 24360 }, { "grad_norm": 0.17264889180660248, "learning_rate": 9.008922138792124e-05, "loss": 0.0293, "step": 24370 }, { "grad_norm": 0.2515614330768585, "learning_rate": 9.007933783947244e-05, "loss": 0.0308, "step": 24380 }, { "grad_norm": 0.32212644815444946, "learning_rate": 9.006944990800752e-05, "loss": 0.0393, "step": 24390 }, { "grad_norm": 0.3158515691757202, "learning_rate": 9.005955759460779e-05, "loss": 0.0366, "step": 24400 }, { "grad_norm": 0.31508868932724, "learning_rate": 9.004966090035508e-05, "loss": 0.0342, "step": 24410 }, { "grad_norm": 0.25463777780532837, "learning_rate": 9.003975982633166e-05, "loss": 0.0348, "step": 24420 }, { "grad_norm": 0.2596348822116852, "learning_rate": 9.00298543736203e-05, "loss": 0.0327, "step": 24430 }, { "grad_norm": 0.25071096420288086, "learning_rate": 9.001994454330427e-05, "loss": 0.0351, "step": 24440 }, { "grad_norm": 0.2962448000907898, "learning_rate": 9.001003033646727e-05, "loss": 0.0337, "step": 24450 }, { "grad_norm": 0.2666556239128113, "learning_rate": 9.00001117541935e-05, "loss": 0.0313, "step": 24460 }, { "grad_norm": 0.20467180013656616, "learning_rate": 8.999018879756764e-05, "loss": 0.0321, "step": 24470 }, { "grad_norm": 0.26045751571655273, "learning_rate": 8.998026146767487e-05, "loss": 0.038, "step": 24480 }, { "grad_norm": 0.34824338555336, "learning_rate": 8.99703297656008e-05, "loss": 0.0385, "step": 24490 }, { "grad_norm": 0.33936652541160583, "learning_rate": 8.996039369243156e-05, "loss": 0.0406, "step": 24500 }, { "grad_norm": 0.36294421553611755, "learning_rate": 8.995045324925378e-05, "loss": 0.0396, "step": 24510 }, { "grad_norm": 0.3003394901752472, "learning_rate": 8.994050843715448e-05, "loss": 0.0371, "step": 24520 }, { "grad_norm": 0.26562070846557617, "learning_rate": 8.993055925722121e-05, "loss": 0.04, "step": 24530 }, { "grad_norm": 0.2866482138633728, "learning_rate": 8.992060571054202e-05, "loss": 0.0357, "step": 24540 }, { "grad_norm": 0.24738819897174835, "learning_rate": 8.991064779820542e-05, "loss": 0.0338, "step": 24550 }, { "grad_norm": 0.3226715624332428, "learning_rate": 8.990068552130036e-05, "loss": 0.0357, "step": 24560 }, { "grad_norm": 0.24260132014751434, "learning_rate": 8.989071888091634e-05, "loss": 0.0328, "step": 24570 }, { "grad_norm": 0.2457539290189743, "learning_rate": 8.988074787814329e-05, "loss": 0.0314, "step": 24580 }, { "grad_norm": 0.2444826364517212, "learning_rate": 8.987077251407158e-05, "loss": 0.0326, "step": 24590 }, { "grad_norm": 0.28436505794525146, "learning_rate": 8.986079278979216e-05, "loss": 0.0329, "step": 24600 }, { "grad_norm": 0.28853631019592285, "learning_rate": 8.985080870639635e-05, "loss": 0.034, "step": 24610 }, { "grad_norm": 0.41382449865341187, "learning_rate": 8.984082026497603e-05, "loss": 0.0336, "step": 24620 }, { "grad_norm": 0.1957833617925644, "learning_rate": 8.98308274666235e-05, "loss": 0.0288, "step": 24630 }, { "grad_norm": 0.22864307463169098, "learning_rate": 8.982083031243155e-05, "loss": 0.0296, "step": 24640 }, { "grad_norm": 0.22967034578323364, "learning_rate": 8.98108288034935e-05, "loss": 0.032, "step": 24650 }, { "grad_norm": 0.27619513869285583, "learning_rate": 8.980082294090305e-05, "loss": 0.0348, "step": 24660 }, { "grad_norm": 0.30053380131721497, "learning_rate": 8.979081272575443e-05, "loss": 0.032, "step": 24670 }, { "grad_norm": 0.28631994128227234, "learning_rate": 8.978079815914236e-05, "loss": 0.0344, "step": 24680 }, { "grad_norm": 0.30803728103637695, "learning_rate": 8.977077924216202e-05, "loss": 0.0347, "step": 24690 }, { "grad_norm": 0.27912473678588867, "learning_rate": 8.976075597590905e-05, "loss": 0.0301, "step": 24700 }, { "grad_norm": 0.31120240688323975, "learning_rate": 8.975072836147958e-05, "loss": 0.0366, "step": 24710 }, { "grad_norm": 0.33341294527053833, "learning_rate": 8.974069639997025e-05, "loss": 0.0387, "step": 24720 }, { "grad_norm": 0.288896381855011, "learning_rate": 8.973066009247808e-05, "loss": 0.0332, "step": 24730 }, { "grad_norm": 0.27978751063346863, "learning_rate": 8.972061944010066e-05, "loss": 0.0349, "step": 24740 }, { "grad_norm": 0.2603835165500641, "learning_rate": 8.971057444393603e-05, "loss": 0.034, "step": 24750 }, { "grad_norm": 0.24462448060512543, "learning_rate": 8.970052510508268e-05, "loss": 0.034, "step": 24760 }, { "grad_norm": 0.3350871205329895, "learning_rate": 8.969047142463959e-05, "loss": 0.0348, "step": 24770 }, { "grad_norm": 0.26731300354003906, "learning_rate": 8.968041340370621e-05, "loss": 0.0331, "step": 24780 }, { "grad_norm": 0.27851709723472595, "learning_rate": 8.96703510433825e-05, "loss": 0.0384, "step": 24790 }, { "grad_norm": 0.290044903755188, "learning_rate": 8.966028434476883e-05, "loss": 0.0343, "step": 24800 }, { "grad_norm": 0.2787891924381256, "learning_rate": 8.96502133089661e-05, "loss": 0.0298, "step": 24810 }, { "grad_norm": 0.20513680577278137, "learning_rate": 8.964013793707564e-05, "loss": 0.0343, "step": 24820 }, { "grad_norm": 0.29379040002822876, "learning_rate": 8.963005823019932e-05, "loss": 0.0301, "step": 24830 }, { "grad_norm": 0.3470582664012909, "learning_rate": 8.961997418943939e-05, "loss": 0.0351, "step": 24840 }, { "grad_norm": 0.24132303893566132, "learning_rate": 8.960988581589865e-05, "loss": 0.0338, "step": 24850 }, { "grad_norm": 0.3913041949272156, "learning_rate": 8.959979311068037e-05, "loss": 0.0357, "step": 24860 }, { "grad_norm": 0.24355529248714447, "learning_rate": 8.958969607488823e-05, "loss": 0.0327, "step": 24870 }, { "grad_norm": 0.27408111095428467, "learning_rate": 8.957959470962647e-05, "loss": 0.0346, "step": 24880 }, { "grad_norm": 0.26165175437927246, "learning_rate": 8.956948901599971e-05, "loss": 0.0387, "step": 24890 }, { "grad_norm": 0.2538264989852905, "learning_rate": 8.955937899511315e-05, "loss": 0.0409, "step": 24900 }, { "grad_norm": 0.22913916409015656, "learning_rate": 8.954926464807238e-05, "loss": 0.0312, "step": 24910 }, { "grad_norm": 0.3763355314731598, "learning_rate": 8.953914597598347e-05, "loss": 0.0333, "step": 24920 }, { "grad_norm": 0.3143777549266815, "learning_rate": 8.952902297995303e-05, "loss": 0.0348, "step": 24930 }, { "grad_norm": 0.2704022228717804, "learning_rate": 8.951889566108804e-05, "loss": 0.0403, "step": 24940 }, { "grad_norm": 0.332329124212265, "learning_rate": 8.950876402049606e-05, "loss": 0.0329, "step": 24950 }, { "grad_norm": 0.21271538734436035, "learning_rate": 8.949862805928504e-05, "loss": 0.0301, "step": 24960 }, { "grad_norm": 0.27382856607437134, "learning_rate": 8.948848777856343e-05, "loss": 0.0304, "step": 24970 }, { "grad_norm": 0.28349611163139343, "learning_rate": 8.947834317944017e-05, "loss": 0.0342, "step": 24980 }, { "grad_norm": 0.20767803490161896, "learning_rate": 8.946819426302466e-05, "loss": 0.0348, "step": 24990 }, { "grad_norm": 0.24960441887378693, "learning_rate": 8.945804103042676e-05, "loss": 0.0309, "step": 25000 }, { "grad_norm": 0.2723786532878876, "learning_rate": 8.944788348275681e-05, "loss": 0.0356, "step": 25010 }, { "grad_norm": 0.2509235739707947, "learning_rate": 8.943772162112565e-05, "loss": 0.0365, "step": 25020 }, { "grad_norm": 0.24506182968616486, "learning_rate": 8.942755544664454e-05, "loss": 0.035, "step": 25030 }, { "grad_norm": 0.30748966336250305, "learning_rate": 8.941738496042525e-05, "loss": 0.0422, "step": 25040 }, { "grad_norm": 0.2409573644399643, "learning_rate": 8.940721016357999e-05, "loss": 0.034, "step": 25050 }, { "grad_norm": 0.21799726784229279, "learning_rate": 8.939703105722148e-05, "loss": 0.0382, "step": 25060 }, { "grad_norm": 0.2995268404483795, "learning_rate": 8.93868476424629e-05, "loss": 0.04, "step": 25070 }, { "grad_norm": 0.3174600601196289, "learning_rate": 8.937665992041786e-05, "loss": 0.0349, "step": 25080 }, { "grad_norm": 0.33979934453964233, "learning_rate": 8.93664678922005e-05, "loss": 0.0352, "step": 25090 }, { "grad_norm": 0.33747050166130066, "learning_rate": 8.93562715589254e-05, "loss": 0.0328, "step": 25100 }, { "grad_norm": 0.29243701696395874, "learning_rate": 8.934607092170762e-05, "loss": 0.033, "step": 25110 }, { "grad_norm": 0.2721221148967743, "learning_rate": 8.933586598166266e-05, "loss": 0.0328, "step": 25120 }, { "grad_norm": 0.26924580335617065, "learning_rate": 8.932565673990655e-05, "loss": 0.0344, "step": 25130 }, { "grad_norm": 0.35188794136047363, "learning_rate": 8.931544319755574e-05, "loss": 0.0321, "step": 25140 }, { "grad_norm": 0.30635231733322144, "learning_rate": 8.930522535572718e-05, "loss": 0.0328, "step": 25150 }, { "grad_norm": 0.3282792568206787, "learning_rate": 8.929500321553826e-05, "loss": 0.0327, "step": 25160 }, { "grad_norm": 0.27203601598739624, "learning_rate": 8.928477677810686e-05, "loss": 0.0351, "step": 25170 }, { "grad_norm": 0.242325559258461, "learning_rate": 8.927454604455137e-05, "loss": 0.0369, "step": 25180 }, { "grad_norm": 0.29182547330856323, "learning_rate": 8.926431101599053e-05, "loss": 0.0365, "step": 25190 }, { "grad_norm": 0.3303379416465759, "learning_rate": 8.925407169354369e-05, "loss": 0.0388, "step": 25200 }, { "grad_norm": 0.3385946750640869, "learning_rate": 8.92438280783306e-05, "loss": 0.034, "step": 25210 }, { "grad_norm": 0.33448150753974915, "learning_rate": 8.923358017147146e-05, "loss": 0.0403, "step": 25220 }, { "grad_norm": 0.2993879020214081, "learning_rate": 8.922332797408697e-05, "loss": 0.0427, "step": 25230 }, { "grad_norm": 0.2731856405735016, "learning_rate": 8.921307148729831e-05, "loss": 0.0376, "step": 25240 }, { "grad_norm": 0.28023043274879456, "learning_rate": 8.920281071222712e-05, "loss": 0.0393, "step": 25250 }, { "grad_norm": 0.24843107163906097, "learning_rate": 8.919254564999548e-05, "loss": 0.0323, "step": 25260 }, { "grad_norm": 0.40917712450027466, "learning_rate": 8.918227630172598e-05, "loss": 0.0339, "step": 25270 }, { "grad_norm": 0.3152557909488678, "learning_rate": 8.917200266854165e-05, "loss": 0.0408, "step": 25280 }, { "grad_norm": 0.2550700008869171, "learning_rate": 8.9161724751566e-05, "loss": 0.0362, "step": 25290 }, { "grad_norm": 0.2609485685825348, "learning_rate": 8.915144255192302e-05, "loss": 0.0336, "step": 25300 }, { "grad_norm": 0.2635272443294525, "learning_rate": 8.914115607073714e-05, "loss": 0.0293, "step": 25310 }, { "grad_norm": 0.21198676526546478, "learning_rate": 8.913086530913327e-05, "loss": 0.0302, "step": 25320 }, { "grad_norm": 0.27048397064208984, "learning_rate": 8.912057026823681e-05, "loss": 0.0313, "step": 25330 }, { "grad_norm": 0.26356878876686096, "learning_rate": 8.91102709491736e-05, "loss": 0.0319, "step": 25340 }, { "grad_norm": 0.2744753658771515, "learning_rate": 8.909996735306996e-05, "loss": 0.0342, "step": 25350 }, { "grad_norm": 0.2896430790424347, "learning_rate": 8.908965948105268e-05, "loss": 0.034, "step": 25360 }, { "grad_norm": 0.3472297191619873, "learning_rate": 8.907934733424901e-05, "loss": 0.0377, "step": 25370 }, { "grad_norm": 0.2555129826068878, "learning_rate": 8.906903091378666e-05, "loss": 0.0346, "step": 25380 }, { "grad_norm": 0.22637656331062317, "learning_rate": 8.905871022079384e-05, "loss": 0.0335, "step": 25390 }, { "grad_norm": 0.4143131375312805, "learning_rate": 8.90483852563992e-05, "loss": 0.0316, "step": 25400 }, { "grad_norm": 0.21438655257225037, "learning_rate": 8.903805602173185e-05, "loss": 0.0301, "step": 25410 }, { "grad_norm": 0.3000431954860687, "learning_rate": 8.902772251792137e-05, "loss": 0.0315, "step": 25420 }, { "grad_norm": 0.2434542328119278, "learning_rate": 8.901738474609786e-05, "loss": 0.0336, "step": 25430 }, { "grad_norm": 0.3057737648487091, "learning_rate": 8.900704270739179e-05, "loss": 0.0329, "step": 25440 }, { "grad_norm": 0.2089717537164688, "learning_rate": 8.89966964029342e-05, "loss": 0.0312, "step": 25450 }, { "grad_norm": 0.20960457623004913, "learning_rate": 8.898634583385652e-05, "loss": 0.032, "step": 25460 }, { "grad_norm": 0.36805030703544617, "learning_rate": 8.897599100129065e-05, "loss": 0.0314, "step": 25470 }, { "grad_norm": 0.3030906915664673, "learning_rate": 8.896563190636903e-05, "loss": 0.0341, "step": 25480 }, { "grad_norm": 0.23600825667381287, "learning_rate": 8.895526855022448e-05, "loss": 0.0298, "step": 25490 }, { "grad_norm": 0.2381887435913086, "learning_rate": 8.894490093399033e-05, "loss": 0.0306, "step": 25500 }, { "grad_norm": 0.26105573773384094, "learning_rate": 8.893452905880035e-05, "loss": 0.0343, "step": 25510 }, { "grad_norm": 0.21940678358078003, "learning_rate": 8.892415292578883e-05, "loss": 0.0287, "step": 25520 }, { "grad_norm": 0.3380388021469116, "learning_rate": 8.891377253609046e-05, "loss": 0.034, "step": 25530 }, { "grad_norm": 0.6765331029891968, "learning_rate": 8.890338789084043e-05, "loss": 0.0359, "step": 25540 }, { "grad_norm": 0.3544960021972656, "learning_rate": 8.88929989911744e-05, "loss": 0.0337, "step": 25550 }, { "grad_norm": 0.3077586591243744, "learning_rate": 8.888260583822847e-05, "loss": 0.0402, "step": 25560 }, { "grad_norm": 0.28329944610595703, "learning_rate": 8.887220843313921e-05, "loss": 0.0342, "step": 25570 }, { "grad_norm": 0.3828032612800598, "learning_rate": 8.88618067770437e-05, "loss": 0.0338, "step": 25580 }, { "grad_norm": 0.36352744698524475, "learning_rate": 8.885140087107942e-05, "loss": 0.0351, "step": 25590 }, { "grad_norm": 0.3071819245815277, "learning_rate": 8.884099071638436e-05, "loss": 0.0334, "step": 25600 }, { "grad_norm": 0.25106799602508545, "learning_rate": 8.883057631409695e-05, "loss": 0.0352, "step": 25610 }, { "grad_norm": 0.2883274555206299, "learning_rate": 8.882015766535608e-05, "loss": 0.0308, "step": 25620 }, { "grad_norm": 0.27348992228507996, "learning_rate": 8.880973477130115e-05, "loss": 0.0348, "step": 25630 }, { "grad_norm": 0.2801946699619293, "learning_rate": 8.879930763307197e-05, "loss": 0.0346, "step": 25640 }, { "grad_norm": 0.22316499054431915, "learning_rate": 8.878887625180884e-05, "loss": 0.0276, "step": 25650 }, { "grad_norm": 0.22920465469360352, "learning_rate": 8.877844062865253e-05, "loss": 0.0358, "step": 25660 }, { "grad_norm": 0.22808006405830383, "learning_rate": 8.876800076474424e-05, "loss": 0.0312, "step": 25670 }, { "grad_norm": 0.28337565064430237, "learning_rate": 8.875755666122568e-05, "loss": 0.0295, "step": 25680 }, { "grad_norm": 0.2042342722415924, "learning_rate": 8.8747108319239e-05, "loss": 0.0306, "step": 25690 }, { "grad_norm": 0.3702949583530426, "learning_rate": 8.87366557399268e-05, "loss": 0.0339, "step": 25700 }, { "grad_norm": 0.296762079000473, "learning_rate": 8.872619892443217e-05, "loss": 0.0355, "step": 25710 }, { "grad_norm": 0.2703660726547241, "learning_rate": 8.871573787389865e-05, "loss": 0.0331, "step": 25720 }, { "grad_norm": 0.20225636661052704, "learning_rate": 8.870527258947024e-05, "loss": 0.0307, "step": 25730 }, { "grad_norm": 0.27426043152809143, "learning_rate": 8.869480307229143e-05, "loss": 0.0341, "step": 25740 }, { "grad_norm": 0.23075971007347107, "learning_rate": 8.868432932350712e-05, "loss": 0.0345, "step": 25750 }, { "grad_norm": 0.4122255742549896, "learning_rate": 8.867385134426272e-05, "loss": 0.0344, "step": 25760 }, { "grad_norm": 0.29156574606895447, "learning_rate": 8.866336913570407e-05, "loss": 0.0332, "step": 25770 }, { "grad_norm": 0.40141382813453674, "learning_rate": 8.865288269897751e-05, "loss": 0.0398, "step": 25780 }, { "grad_norm": 0.27884629368782043, "learning_rate": 8.864239203522981e-05, "loss": 0.0365, "step": 25790 }, { "grad_norm": 0.2872637212276459, "learning_rate": 8.863189714560822e-05, "loss": 0.0385, "step": 25800 }, { "grad_norm": 0.26525720953941345, "learning_rate": 8.862139803126043e-05, "loss": 0.0322, "step": 25810 }, { "grad_norm": 0.3376012444496155, "learning_rate": 8.861089469333463e-05, "loss": 0.0314, "step": 25820 }, { "grad_norm": 0.2569099962711334, "learning_rate": 8.860038713297944e-05, "loss": 0.0347, "step": 25830 }, { "grad_norm": 0.2479589581489563, "learning_rate": 8.858987535134394e-05, "loss": 0.0327, "step": 25840 }, { "grad_norm": 0.3156946003437042, "learning_rate": 8.857935934957769e-05, "loss": 0.0377, "step": 25850 }, { "grad_norm": 0.27687010169029236, "learning_rate": 8.856883912883071e-05, "loss": 0.0332, "step": 25860 }, { "grad_norm": 0.2711528539657593, "learning_rate": 8.855831469025346e-05, "loss": 0.0325, "step": 25870 }, { "grad_norm": 0.2493167668581009, "learning_rate": 8.854778603499689e-05, "loss": 0.036, "step": 25880 }, { "grad_norm": 0.2067718803882599, "learning_rate": 8.85372531642124e-05, "loss": 0.0368, "step": 25890 }, { "grad_norm": 0.2450428456068039, "learning_rate": 8.852671607905185e-05, "loss": 0.0352, "step": 25900 }, { "grad_norm": 0.22911792993545532, "learning_rate": 8.851617478066754e-05, "loss": 0.034, "step": 25910 }, { "grad_norm": 0.28198450803756714, "learning_rate": 8.850562927021227e-05, "loss": 0.0315, "step": 25920 }, { "grad_norm": 0.3423287272453308, "learning_rate": 8.849507954883928e-05, "loss": 0.033, "step": 25930 }, { "grad_norm": 0.34440523386001587, "learning_rate": 8.848452561770226e-05, "loss": 0.0293, "step": 25940 }, { "grad_norm": 0.28538286685943604, "learning_rate": 8.847396747795538e-05, "loss": 0.0268, "step": 25950 }, { "grad_norm": 0.26677775382995605, "learning_rate": 8.846340513075327e-05, "loss": 0.0346, "step": 25960 }, { "grad_norm": 0.33066293597221375, "learning_rate": 8.845283857725099e-05, "loss": 0.0411, "step": 25970 }, { "grad_norm": 0.24476103484630585, "learning_rate": 8.844226781860409e-05, "loss": 0.0383, "step": 25980 }, { "grad_norm": 0.3208696246147156, "learning_rate": 8.84316928559686e-05, "loss": 0.0318, "step": 25990 }, { "grad_norm": 0.30376961827278137, "learning_rate": 8.842111369050094e-05, "loss": 0.0292, "step": 26000 }, { "grad_norm": 0.26125648617744446, "learning_rate": 8.841053032335808e-05, "loss": 0.0294, "step": 26010 }, { "grad_norm": 0.1534688025712967, "learning_rate": 8.839994275569735e-05, "loss": 0.0359, "step": 26020 }, { "grad_norm": 0.22237218916416168, "learning_rate": 8.838935098867662e-05, "loss": 0.0288, "step": 26030 }, { "grad_norm": 0.2129458487033844, "learning_rate": 8.837875502345418e-05, "loss": 0.0321, "step": 26040 }, { "grad_norm": 0.272148460149765, "learning_rate": 8.83681548611888e-05, "loss": 0.0321, "step": 26050 }, { "grad_norm": 0.2660568952560425, "learning_rate": 8.835755050303969e-05, "loss": 0.0314, "step": 26060 }, { "grad_norm": 0.19520412385463715, "learning_rate": 8.834694195016653e-05, "loss": 0.0348, "step": 26070 }, { "grad_norm": 0.33786606788635254, "learning_rate": 8.833632920372942e-05, "loss": 0.0376, "step": 26080 }, { "grad_norm": 0.5037880539894104, "learning_rate": 8.832571226488903e-05, "loss": 0.0405, "step": 26090 }, { "grad_norm": 0.22870968282222748, "learning_rate": 8.831509113480634e-05, "loss": 0.0342, "step": 26100 }, { "grad_norm": 0.34005483984947205, "learning_rate": 8.83044658146429e-05, "loss": 0.0369, "step": 26110 }, { "grad_norm": 0.23284587264060974, "learning_rate": 8.829383630556067e-05, "loss": 0.0293, "step": 26120 }, { "grad_norm": 0.2340555638074875, "learning_rate": 8.828320260872207e-05, "loss": 0.0308, "step": 26130 }, { "grad_norm": 0.3519551455974579, "learning_rate": 8.827256472529e-05, "loss": 0.0303, "step": 26140 }, { "grad_norm": 0.2667143642902374, "learning_rate": 8.826192265642778e-05, "loss": 0.0305, "step": 26150 }, { "grad_norm": 0.3138575851917267, "learning_rate": 8.825127640329923e-05, "loss": 0.0303, "step": 26160 }, { "grad_norm": 0.27352672815322876, "learning_rate": 8.824062596706861e-05, "loss": 0.0329, "step": 26170 }, { "grad_norm": 0.3044557571411133, "learning_rate": 8.822997134890062e-05, "loss": 0.0346, "step": 26180 }, { "grad_norm": 0.21880409121513367, "learning_rate": 8.821931254996044e-05, "loss": 0.0302, "step": 26190 }, { "grad_norm": 0.24450533092021942, "learning_rate": 8.82086495714137e-05, "loss": 0.0325, "step": 26200 }, { "grad_norm": 0.22993503510951996, "learning_rate": 8.81979824144265e-05, "loss": 0.0323, "step": 26210 }, { "grad_norm": 0.24282969534397125, "learning_rate": 8.818731108016536e-05, "loss": 0.0349, "step": 26220 }, { "grad_norm": 0.2792733609676361, "learning_rate": 8.81766355697973e-05, "loss": 0.0337, "step": 26230 }, { "grad_norm": 0.1771792769432068, "learning_rate": 8.816595588448977e-05, "loss": 0.0285, "step": 26240 }, { "grad_norm": 0.25932231545448303, "learning_rate": 8.81552720254107e-05, "loss": 0.0294, "step": 26250 }, { "grad_norm": 0.3010694980621338, "learning_rate": 8.814458399372842e-05, "loss": 0.0365, "step": 26260 }, { "grad_norm": 0.20127323269844055, "learning_rate": 8.813389179061181e-05, "loss": 0.03, "step": 26270 }, { "grad_norm": 0.190082848072052, "learning_rate": 8.812319541723012e-05, "loss": 0.0307, "step": 26280 }, { "grad_norm": 0.2720588445663452, "learning_rate": 8.811249487475309e-05, "loss": 0.0298, "step": 26290 }, { "grad_norm": 0.3599182963371277, "learning_rate": 8.810179016435092e-05, "loss": 0.0369, "step": 26300 }, { "grad_norm": 0.2912788391113281, "learning_rate": 8.809108128719428e-05, "loss": 0.0327, "step": 26310 }, { "grad_norm": 0.2938084900379181, "learning_rate": 8.808036824445424e-05, "loss": 0.0318, "step": 26320 }, { "grad_norm": 0.33660832047462463, "learning_rate": 8.806965103730238e-05, "loss": 0.0341, "step": 26330 }, { "grad_norm": 0.23660343885421753, "learning_rate": 8.805892966691074e-05, "loss": 0.034, "step": 26340 }, { "grad_norm": 0.22633139789104462, "learning_rate": 8.804820413445175e-05, "loss": 0.0298, "step": 26350 }, { "grad_norm": 0.237315371632576, "learning_rate": 8.803747444109837e-05, "loss": 0.0357, "step": 26360 }, { "grad_norm": 0.25295722484588623, "learning_rate": 8.802674058802399e-05, "loss": 0.0338, "step": 26370 }, { "grad_norm": 0.21325761079788208, "learning_rate": 8.801600257640241e-05, "loss": 0.0299, "step": 26380 }, { "grad_norm": 0.2959606647491455, "learning_rate": 8.800526040740795e-05, "loss": 0.0374, "step": 26390 }, { "grad_norm": 0.20990918576717377, "learning_rate": 8.799451408221535e-05, "loss": 0.0304, "step": 26400 }, { "grad_norm": 0.26238757371902466, "learning_rate": 8.798376360199982e-05, "loss": 0.0337, "step": 26410 }, { "grad_norm": 0.27511337399482727, "learning_rate": 8.797300896793701e-05, "loss": 0.031, "step": 26420 }, { "grad_norm": 0.3645128011703491, "learning_rate": 8.796225018120302e-05, "loss": 0.0351, "step": 26430 }, { "grad_norm": 0.24156445264816284, "learning_rate": 8.795148724297444e-05, "loss": 0.0304, "step": 26440 }, { "grad_norm": 0.27981463074684143, "learning_rate": 8.794072015442825e-05, "loss": 0.0363, "step": 26450 }, { "grad_norm": 0.31965482234954834, "learning_rate": 8.792994891674198e-05, "loss": 0.0326, "step": 26460 }, { "grad_norm": 0.26527145504951477, "learning_rate": 8.79191735310935e-05, "loss": 0.0368, "step": 26470 }, { "grad_norm": 0.3172682523727417, "learning_rate": 8.790839399866122e-05, "loss": 0.0338, "step": 26480 }, { "grad_norm": 0.30411428213119507, "learning_rate": 8.789761032062397e-05, "loss": 0.0323, "step": 26490 }, { "grad_norm": 0.25428855419158936, "learning_rate": 8.788682249816103e-05, "loss": 0.0312, "step": 26500 }, { "grad_norm": 0.25099650025367737, "learning_rate": 8.787603053245215e-05, "loss": 0.0336, "step": 26510 }, { "grad_norm": 0.3270021677017212, "learning_rate": 8.78652344246775e-05, "loss": 0.0289, "step": 26520 }, { "grad_norm": 0.29029330611228943, "learning_rate": 8.785443417601776e-05, "loss": 0.0362, "step": 26530 }, { "grad_norm": 0.2530509829521179, "learning_rate": 8.784362978765401e-05, "loss": 0.0286, "step": 26540 }, { "grad_norm": 0.29943132400512695, "learning_rate": 8.783282126076779e-05, "loss": 0.0315, "step": 26550 }, { "grad_norm": 0.2628575265407562, "learning_rate": 8.782200859654112e-05, "loss": 0.0333, "step": 26560 }, { "grad_norm": 0.2637436091899872, "learning_rate": 8.781119179615646e-05, "loss": 0.0289, "step": 26570 }, { "grad_norm": 0.24169592559337616, "learning_rate": 8.780037086079674e-05, "loss": 0.0302, "step": 26580 }, { "grad_norm": 0.24018822610378265, "learning_rate": 8.778954579164527e-05, "loss": 0.031, "step": 26590 }, { "grad_norm": 0.2765547037124634, "learning_rate": 8.777871658988588e-05, "loss": 0.0337, "step": 26600 }, { "grad_norm": 0.3164844810962677, "learning_rate": 8.776788325670285e-05, "loss": 0.0309, "step": 26610 }, { "grad_norm": 0.2192407250404358, "learning_rate": 8.775704579328089e-05, "loss": 0.0298, "step": 26620 }, { "grad_norm": 0.2655571401119232, "learning_rate": 8.774620420080517e-05, "loss": 0.031, "step": 26630 }, { "grad_norm": 0.23692099750041962, "learning_rate": 8.773535848046131e-05, "loss": 0.032, "step": 26640 }, { "grad_norm": 0.2713676393032074, "learning_rate": 8.772450863343538e-05, "loss": 0.0295, "step": 26650 }, { "grad_norm": 0.25977787375450134, "learning_rate": 8.77136546609139e-05, "loss": 0.0338, "step": 26660 }, { "grad_norm": 0.30202242732048035, "learning_rate": 8.770279656408385e-05, "loss": 0.0373, "step": 26670 }, { "grad_norm": 0.18729573488235474, "learning_rate": 8.769193434413265e-05, "loss": 0.029, "step": 26680 }, { "grad_norm": 0.3067074716091156, "learning_rate": 8.76810680022482e-05, "loss": 0.0293, "step": 26690 }, { "grad_norm": 0.23645780980587006, "learning_rate": 8.767019753961878e-05, "loss": 0.0337, "step": 26700 }, { "grad_norm": 0.563589870929718, "learning_rate": 8.765932295743321e-05, "loss": 0.0348, "step": 26710 }, { "grad_norm": 0.290563702583313, "learning_rate": 8.764844425688068e-05, "loss": 0.0366, "step": 26720 }, { "grad_norm": 0.31099405884742737, "learning_rate": 8.763756143915092e-05, "loss": 0.0314, "step": 26730 }, { "grad_norm": 0.27214542031288147, "learning_rate": 8.7626674505434e-05, "loss": 0.0349, "step": 26740 }, { "grad_norm": 0.2787144184112549, "learning_rate": 8.761578345692053e-05, "loss": 0.0328, "step": 26750 }, { "grad_norm": 0.2935802936553955, "learning_rate": 8.760488829480156e-05, "loss": 0.0307, "step": 26760 }, { "grad_norm": 0.23887182772159576, "learning_rate": 8.759398902026854e-05, "loss": 0.0299, "step": 26770 }, { "grad_norm": 0.2500286400318146, "learning_rate": 8.758308563451339e-05, "loss": 0.0324, "step": 26780 }, { "grad_norm": 0.35110464692115784, "learning_rate": 8.75721781387285e-05, "loss": 0.0331, "step": 26790 }, { "grad_norm": 0.39192137122154236, "learning_rate": 8.75612665341067e-05, "loss": 0.0338, "step": 26800 }, { "grad_norm": 0.3303094804286957, "learning_rate": 8.755035082184126e-05, "loss": 0.0319, "step": 26810 }, { "grad_norm": 0.2577213943004608, "learning_rate": 8.753943100312592e-05, "loss": 0.0279, "step": 26820 }, { "grad_norm": 0.22128497064113617, "learning_rate": 8.752850707915484e-05, "loss": 0.0301, "step": 26830 }, { "grad_norm": 0.2738178074359894, "learning_rate": 8.751757905112264e-05, "loss": 0.0337, "step": 26840 }, { "grad_norm": 0.3537549078464508, "learning_rate": 8.75066469202244e-05, "loss": 0.0364, "step": 26850 }, { "grad_norm": 0.3287091851234436, "learning_rate": 8.749571068765567e-05, "loss": 0.0411, "step": 26860 }, { "grad_norm": 0.29457247257232666, "learning_rate": 8.748477035461238e-05, "loss": 0.0353, "step": 26870 }, { "grad_norm": 0.2967138886451721, "learning_rate": 8.747382592229095e-05, "loss": 0.0444, "step": 26880 }, { "grad_norm": 0.27634841203689575, "learning_rate": 8.746287739188828e-05, "loss": 0.0309, "step": 26890 }, { "grad_norm": 0.29560625553131104, "learning_rate": 8.745192476460165e-05, "loss": 0.0381, "step": 26900 }, { "grad_norm": 0.24490240216255188, "learning_rate": 8.744096804162882e-05, "loss": 0.034, "step": 26910 }, { "grad_norm": 0.23576730489730835, "learning_rate": 8.743000722416804e-05, "loss": 0.0314, "step": 26920 }, { "grad_norm": 0.29939156770706177, "learning_rate": 8.741904231341793e-05, "loss": 0.0374, "step": 26930 }, { "grad_norm": 0.2889866828918457, "learning_rate": 8.740807331057762e-05, "loss": 0.0385, "step": 26940 }, { "grad_norm": 0.25362297892570496, "learning_rate": 8.739710021684667e-05, "loss": 0.0352, "step": 26950 }, { "grad_norm": 0.28916651010513306, "learning_rate": 8.738612303342503e-05, "loss": 0.0333, "step": 26960 }, { "grad_norm": 0.2743520140647888, "learning_rate": 8.73751417615132e-05, "loss": 0.0302, "step": 26970 }, { "grad_norm": 0.2564791440963745, "learning_rate": 8.736415640231208e-05, "loss": 0.03, "step": 26980 }, { "grad_norm": 0.398373544216156, "learning_rate": 8.735316695702297e-05, "loss": 0.0332, "step": 26990 }, { "grad_norm": 0.21289746463298798, "learning_rate": 8.734217342684769e-05, "loss": 0.0303, "step": 27000 }, { "grad_norm": 0.3003264367580414, "learning_rate": 8.733117581298847e-05, "loss": 0.0354, "step": 27010 }, { "grad_norm": 0.32369017601013184, "learning_rate": 8.732017411664796e-05, "loss": 0.0341, "step": 27020 }, { "grad_norm": 0.2816360294818878, "learning_rate": 8.730916833902936e-05, "loss": 0.0309, "step": 27030 }, { "grad_norm": 0.27020353078842163, "learning_rate": 8.729815848133618e-05, "loss": 0.0327, "step": 27040 }, { "grad_norm": 0.36035677790641785, "learning_rate": 8.728714454477247e-05, "loss": 0.0306, "step": 27050 }, { "grad_norm": 0.2920210063457489, "learning_rate": 8.727612653054269e-05, "loss": 0.0376, "step": 27060 }, { "grad_norm": 0.27329185605049133, "learning_rate": 8.726510443985176e-05, "loss": 0.0342, "step": 27070 }, { "grad_norm": 0.4202868342399597, "learning_rate": 8.725407827390503e-05, "loss": 0.0372, "step": 27080 }, { "grad_norm": 0.22304196655750275, "learning_rate": 8.724304803390833e-05, "loss": 0.0354, "step": 27090 }, { "grad_norm": 0.28214535117149353, "learning_rate": 8.723201372106788e-05, "loss": 0.0359, "step": 27100 }, { "grad_norm": 0.28076544404029846, "learning_rate": 8.722097533659038e-05, "loss": 0.0377, "step": 27110 }, { "grad_norm": 0.3287573754787445, "learning_rate": 8.720993288168299e-05, "loss": 0.0413, "step": 27120 }, { "grad_norm": 0.30489110946655273, "learning_rate": 8.719888635755327e-05, "loss": 0.0328, "step": 27130 }, { "grad_norm": 0.36688128113746643, "learning_rate": 8.718783576540928e-05, "loss": 0.0372, "step": 27140 }, { "grad_norm": 0.3689222037792206, "learning_rate": 8.717678110645948e-05, "loss": 0.0343, "step": 27150 }, { "grad_norm": 0.19993172585964203, "learning_rate": 8.716572238191279e-05, "loss": 0.0304, "step": 27160 }, { "grad_norm": 0.35040318965911865, "learning_rate": 8.715465959297857e-05, "loss": 0.029, "step": 27170 }, { "grad_norm": 0.3475591838359833, "learning_rate": 8.714359274086665e-05, "loss": 0.0396, "step": 27180 }, { "grad_norm": 0.24539870023727417, "learning_rate": 8.713252182678726e-05, "loss": 0.0324, "step": 27190 }, { "grad_norm": 0.2827298641204834, "learning_rate": 8.712144685195112e-05, "loss": 0.0319, "step": 27200 }, { "grad_norm": 0.24622073769569397, "learning_rate": 8.711036781756936e-05, "loss": 0.0325, "step": 27210 }, { "grad_norm": 0.2698618769645691, "learning_rate": 8.709928472485357e-05, "loss": 0.037, "step": 27220 }, { "grad_norm": 0.2571980655193329, "learning_rate": 8.708819757501579e-05, "loss": 0.0314, "step": 27230 }, { "grad_norm": 0.27415674924850464, "learning_rate": 8.707710636926846e-05, "loss": 0.037, "step": 27240 }, { "grad_norm": 0.32339930534362793, "learning_rate": 8.706601110882455e-05, "loss": 0.033, "step": 27250 }, { "grad_norm": 0.23196940124034882, "learning_rate": 8.705491179489738e-05, "loss": 0.0281, "step": 27260 }, { "grad_norm": 0.23479971289634705, "learning_rate": 8.704380842870077e-05, "loss": 0.0344, "step": 27270 }, { "grad_norm": 0.20345862209796906, "learning_rate": 8.703270101144895e-05, "loss": 0.0319, "step": 27280 }, { "grad_norm": 0.2279641330242157, "learning_rate": 8.702158954435664e-05, "loss": 0.0344, "step": 27290 }, { "grad_norm": 0.2584981322288513, "learning_rate": 8.701047402863896e-05, "loss": 0.0307, "step": 27300 }, { "grad_norm": 0.27374330163002014, "learning_rate": 8.699935446551148e-05, "loss": 0.0313, "step": 27310 }, { "grad_norm": 0.29649513959884644, "learning_rate": 8.698823085619022e-05, "loss": 0.0356, "step": 27320 }, { "grad_norm": 0.2688013017177582, "learning_rate": 8.697710320189166e-05, "loss": 0.0306, "step": 27330 }, { "grad_norm": 0.2417846918106079, "learning_rate": 8.696597150383268e-05, "loss": 0.0368, "step": 27340 }, { "grad_norm": 0.29664328694343567, "learning_rate": 8.695483576323063e-05, "loss": 0.0319, "step": 27350 }, { "grad_norm": 0.28315359354019165, "learning_rate": 8.69436959813033e-05, "loss": 0.0328, "step": 27360 }, { "grad_norm": 0.2780151665210724, "learning_rate": 8.693255215926892e-05, "loss": 0.0332, "step": 27370 }, { "grad_norm": 0.251024067401886, "learning_rate": 8.692140429834617e-05, "loss": 0.0326, "step": 27380 }, { "grad_norm": 0.2434665560722351, "learning_rate": 8.691025239975415e-05, "loss": 0.0314, "step": 27390 }, { "grad_norm": 0.24579881131649017, "learning_rate": 8.689909646471243e-05, "loss": 0.0326, "step": 27400 }, { "grad_norm": 0.3791252672672272, "learning_rate": 8.688793649444099e-05, "loss": 0.032, "step": 27410 }, { "grad_norm": 0.24115869402885437, "learning_rate": 8.687677249016029e-05, "loss": 0.0339, "step": 27420 }, { "grad_norm": 0.3351120352745056, "learning_rate": 8.686560445309118e-05, "loss": 0.0364, "step": 27430 }, { "grad_norm": 0.23234491050243378, "learning_rate": 8.685443238445499e-05, "loss": 0.0296, "step": 27440 }, { "grad_norm": 0.23154614865779877, "learning_rate": 8.68432562854735e-05, "loss": 0.0277, "step": 27450 }, { "grad_norm": 0.35474467277526855, "learning_rate": 8.683207615736887e-05, "loss": 0.0329, "step": 27460 }, { "grad_norm": 0.29634711146354675, "learning_rate": 8.682089200136379e-05, "loss": 0.0361, "step": 27470 }, { "grad_norm": 0.2886871099472046, "learning_rate": 8.680970381868132e-05, "loss": 0.0298, "step": 27480 }, { "grad_norm": 0.2890731394290924, "learning_rate": 8.679851161054498e-05, "loss": 0.0296, "step": 27490 }, { "grad_norm": 0.24075277149677277, "learning_rate": 8.678731537817873e-05, "loss": 0.0258, "step": 27500 }, { "grad_norm": 0.21589644253253937, "learning_rate": 8.677611512280697e-05, "loss": 0.0304, "step": 27510 }, { "grad_norm": 0.23150546848773956, "learning_rate": 8.676491084565457e-05, "loss": 0.0303, "step": 27520 }, { "grad_norm": 0.2606438100337982, "learning_rate": 8.675370254794678e-05, "loss": 0.0315, "step": 27530 }, { "grad_norm": 0.17605407536029816, "learning_rate": 8.674249023090935e-05, "loss": 0.034, "step": 27540 }, { "grad_norm": 0.28055259585380554, "learning_rate": 8.673127389576843e-05, "loss": 0.0332, "step": 27550 }, { "grad_norm": 0.3368113338947296, "learning_rate": 8.67200535437506e-05, "loss": 0.035, "step": 27560 }, { "grad_norm": 0.26342812180519104, "learning_rate": 8.670882917608296e-05, "loss": 0.0293, "step": 27570 }, { "grad_norm": 0.23792892694473267, "learning_rate": 8.669760079399292e-05, "loss": 0.0318, "step": 27580 }, { "grad_norm": 0.23984524607658386, "learning_rate": 8.668636839870845e-05, "loss": 0.0295, "step": 27590 }, { "grad_norm": 0.28380969166755676, "learning_rate": 8.667513199145789e-05, "loss": 0.0331, "step": 27600 }, { "grad_norm": 0.16702920198440552, "learning_rate": 8.666389157347002e-05, "loss": 0.0272, "step": 27610 }, { "grad_norm": 0.2721172571182251, "learning_rate": 8.66526471459741e-05, "loss": 0.0336, "step": 27620 }, { "grad_norm": 0.26987284421920776, "learning_rate": 8.66413987101998e-05, "loss": 0.0317, "step": 27630 }, { "grad_norm": 0.2405092567205429, "learning_rate": 8.663014626737723e-05, "loss": 0.0327, "step": 27640 }, { "grad_norm": 0.24385230243206024, "learning_rate": 8.661888981873691e-05, "loss": 0.0292, "step": 27650 }, { "grad_norm": 0.21872535347938538, "learning_rate": 8.660762936550988e-05, "loss": 0.0276, "step": 27660 }, { "grad_norm": 0.23425598442554474, "learning_rate": 8.659636490892753e-05, "loss": 0.0264, "step": 27670 }, { "grad_norm": 0.38015303015708923, "learning_rate": 8.658509645022174e-05, "loss": 0.03, "step": 27680 }, { "grad_norm": 0.2377278208732605, "learning_rate": 8.657382399062481e-05, "loss": 0.0305, "step": 27690 }, { "grad_norm": 0.2349749505519867, "learning_rate": 8.656254753136946e-05, "loss": 0.0325, "step": 27700 }, { "grad_norm": 0.28844380378723145, "learning_rate": 8.655126707368891e-05, "loss": 0.0327, "step": 27710 }, { "grad_norm": 0.30788519978523254, "learning_rate": 8.653998261881672e-05, "loss": 0.0325, "step": 27720 }, { "grad_norm": 0.2566644847393036, "learning_rate": 8.652869416798699e-05, "loss": 0.0381, "step": 27730 }, { "grad_norm": 0.2298620641231537, "learning_rate": 8.651740172243417e-05, "loss": 0.0267, "step": 27740 }, { "grad_norm": 0.26212024688720703, "learning_rate": 8.65061052833932e-05, "loss": 0.0331, "step": 27750 }, { "grad_norm": 0.22395552694797516, "learning_rate": 8.649480485209945e-05, "loss": 0.0297, "step": 27760 }, { "grad_norm": 0.2982400953769684, "learning_rate": 8.64835004297887e-05, "loss": 0.028, "step": 27770 }, { "grad_norm": 0.2192167490720749, "learning_rate": 8.64721920176972e-05, "loss": 0.0298, "step": 27780 }, { "grad_norm": 0.330990731716156, "learning_rate": 8.646087961706164e-05, "loss": 0.035, "step": 27790 }, { "grad_norm": 0.3052099943161011, "learning_rate": 8.644956322911908e-05, "loss": 0.0305, "step": 27800 }, { "grad_norm": 0.19829142093658447, "learning_rate": 8.643824285510709e-05, "loss": 0.03, "step": 27810 }, { "grad_norm": 0.2388705313205719, "learning_rate": 8.642691849626364e-05, "loss": 0.0246, "step": 27820 }, { "grad_norm": 0.2798060178756714, "learning_rate": 8.641559015382717e-05, "loss": 0.0304, "step": 27830 }, { "grad_norm": 0.22418923676013947, "learning_rate": 8.640425782903649e-05, "loss": 0.0342, "step": 27840 }, { "grad_norm": 0.3104627728462219, "learning_rate": 8.639292152313091e-05, "loss": 0.0294, "step": 27850 }, { "grad_norm": 0.27428585290908813, "learning_rate": 8.638158123735015e-05, "loss": 0.0332, "step": 27860 }, { "grad_norm": 0.22354300320148468, "learning_rate": 8.637023697293436e-05, "loss": 0.0251, "step": 27870 }, { "grad_norm": 0.196346715092659, "learning_rate": 8.635888873112414e-05, "loss": 0.0298, "step": 27880 }, { "grad_norm": 0.24707965552806854, "learning_rate": 8.634753651316052e-05, "loss": 0.0295, "step": 27890 }, { "grad_norm": 0.28519493341445923, "learning_rate": 8.633618032028496e-05, "loss": 0.0331, "step": 27900 }, { "grad_norm": 0.21237999200820923, "learning_rate": 8.632482015373934e-05, "loss": 0.0266, "step": 27910 }, { "grad_norm": 0.22879773378372192, "learning_rate": 8.6313456014766e-05, "loss": 0.0323, "step": 27920 }, { "grad_norm": 0.2968192398548126, "learning_rate": 8.630208790460771e-05, "loss": 0.0288, "step": 27930 }, { "grad_norm": 0.24819354712963104, "learning_rate": 8.629071582450768e-05, "loss": 0.0302, "step": 27940 }, { "grad_norm": 0.24036641418933868, "learning_rate": 8.62793397757095e-05, "loss": 0.0306, "step": 27950 }, { "grad_norm": 0.34631943702697754, "learning_rate": 8.626795975945729e-05, "loss": 0.0355, "step": 27960 }, { "grad_norm": 0.2656095623970032, "learning_rate": 8.625657577699551e-05, "loss": 0.0361, "step": 27970 }, { "grad_norm": 0.23019596934318542, "learning_rate": 8.624518782956914e-05, "loss": 0.0338, "step": 27980 }, { "grad_norm": 0.2561720609664917, "learning_rate": 8.62337959184235e-05, "loss": 0.031, "step": 27990 }, { "grad_norm": 0.29902487993240356, "learning_rate": 8.622240004480441e-05, "loss": 0.0386, "step": 28000 }, { "grad_norm": 0.2695724368095398, "learning_rate": 8.621100020995814e-05, "loss": 0.0321, "step": 28010 }, { "grad_norm": 0.24151833355426788, "learning_rate": 8.619959641513132e-05, "loss": 0.0317, "step": 28020 }, { "grad_norm": 0.6327409744262695, "learning_rate": 8.618818866157105e-05, "loss": 0.0381, "step": 28030 }, { "grad_norm": 0.29932209849357605, "learning_rate": 8.617677695052487e-05, "loss": 0.0365, "step": 28040 }, { "grad_norm": 0.25418663024902344, "learning_rate": 8.616536128324078e-05, "loss": 0.034, "step": 28050 }, { "grad_norm": 0.22755523025989532, "learning_rate": 8.615394166096712e-05, "loss": 0.0362, "step": 28060 }, { "grad_norm": 0.3060307204723358, "learning_rate": 8.614251808495279e-05, "loss": 0.0306, "step": 28070 }, { "grad_norm": 0.25917568802833557, "learning_rate": 8.6131090556447e-05, "loss": 0.035, "step": 28080 }, { "grad_norm": 0.29757216572761536, "learning_rate": 8.611965907669947e-05, "loss": 0.0317, "step": 28090 }, { "grad_norm": 0.3113250136375427, "learning_rate": 8.610822364696034e-05, "loss": 0.0402, "step": 28100 }, { "grad_norm": 0.2809675335884094, "learning_rate": 8.609678426848015e-05, "loss": 0.0364, "step": 28110 }, { "grad_norm": 0.26261553168296814, "learning_rate": 8.60853409425099e-05, "loss": 0.0295, "step": 28120 }, { "grad_norm": 0.3278332054615021, "learning_rate": 8.607389367030104e-05, "loss": 0.0352, "step": 28130 }, { "grad_norm": 0.2671268582344055, "learning_rate": 8.606244245310538e-05, "loss": 0.0343, "step": 28140 }, { "grad_norm": 0.2899438738822937, "learning_rate": 8.605098729217525e-05, "loss": 0.0323, "step": 28150 }, { "grad_norm": 0.26368704438209534, "learning_rate": 8.603952818876335e-05, "loss": 0.0312, "step": 28160 }, { "grad_norm": 0.3226470649242401, "learning_rate": 8.602806514412281e-05, "loss": 0.0344, "step": 28170 }, { "grad_norm": 0.3830994963645935, "learning_rate": 8.601659815950726e-05, "loss": 0.0354, "step": 28180 }, { "grad_norm": 0.30927717685699463, "learning_rate": 8.600512723617067e-05, "loss": 0.0298, "step": 28190 }, { "grad_norm": 0.2599824368953705, "learning_rate": 8.59936523753675e-05, "loss": 0.0347, "step": 28200 }, { "grad_norm": 0.2808815538883209, "learning_rate": 8.598217357835264e-05, "loss": 0.0313, "step": 28210 }, { "grad_norm": 0.2436937540769577, "learning_rate": 8.597069084638135e-05, "loss": 0.0301, "step": 28220 }, { "grad_norm": 0.2758329212665558, "learning_rate": 8.595920418070939e-05, "loss": 0.0281, "step": 28230 }, { "grad_norm": 0.26022377610206604, "learning_rate": 8.594771358259295e-05, "loss": 0.0324, "step": 28240 }, { "grad_norm": 0.17558906972408295, "learning_rate": 8.593621905328858e-05, "loss": 0.0265, "step": 28250 }, { "grad_norm": 0.23517276346683502, "learning_rate": 8.592472059405333e-05, "loss": 0.0326, "step": 28260 }, { "grad_norm": 0.21315091848373413, "learning_rate": 8.591321820614464e-05, "loss": 0.0278, "step": 28270 }, { "grad_norm": 0.25451627373695374, "learning_rate": 8.590171189082041e-05, "loss": 0.0322, "step": 28280 }, { "grad_norm": 0.2081630975008011, "learning_rate": 8.589020164933894e-05, "loss": 0.0274, "step": 28290 }, { "grad_norm": 0.22478072345256805, "learning_rate": 8.587868748295898e-05, "loss": 0.0324, "step": 28300 }, { "grad_norm": 0.22536098957061768, "learning_rate": 8.586716939293971e-05, "loss": 0.034, "step": 28310 }, { "grad_norm": 0.23970507085323334, "learning_rate": 8.58556473805407e-05, "loss": 0.0305, "step": 28320 }, { "grad_norm": 0.28999683260917664, "learning_rate": 8.584412144702202e-05, "loss": 0.0321, "step": 28330 }, { "grad_norm": 0.25272315740585327, "learning_rate": 8.58325915936441e-05, "loss": 0.0312, "step": 28340 }, { "grad_norm": 0.2884247303009033, "learning_rate": 8.582105782166783e-05, "loss": 0.0361, "step": 28350 }, { "grad_norm": 0.315287709236145, "learning_rate": 8.580952013235455e-05, "loss": 0.0354, "step": 28360 }, { "grad_norm": 0.29062339663505554, "learning_rate": 8.579797852696596e-05, "loss": 0.0339, "step": 28370 }, { "grad_norm": 0.22665604948997498, "learning_rate": 8.578643300676428e-05, "loss": 0.0297, "step": 28380 }, { "grad_norm": 0.2241717278957367, "learning_rate": 8.577488357301209e-05, "loss": 0.0331, "step": 28390 }, { "grad_norm": 0.2881399393081665, "learning_rate": 8.576333022697242e-05, "loss": 0.0344, "step": 28400 }, { "grad_norm": 0.4259205460548401, "learning_rate": 8.575177296990873e-05, "loss": 0.0294, "step": 28410 }, { "grad_norm": 0.3167412281036377, "learning_rate": 8.574021180308489e-05, "loss": 0.0332, "step": 28420 }, { "grad_norm": 0.27570754289627075, "learning_rate": 8.572864672776523e-05, "loss": 0.0324, "step": 28430 }, { "grad_norm": 0.28769996762275696, "learning_rate": 8.571707774521447e-05, "loss": 0.0308, "step": 28440 }, { "grad_norm": 0.2848484516143799, "learning_rate": 8.57055048566978e-05, "loss": 0.0305, "step": 28450 }, { "grad_norm": 0.21084997057914734, "learning_rate": 8.569392806348078e-05, "loss": 0.029, "step": 28460 }, { "grad_norm": 0.24623508751392365, "learning_rate": 8.568234736682947e-05, "loss": 0.0278, "step": 28470 }, { "grad_norm": 0.2830069661140442, "learning_rate": 8.567076276801029e-05, "loss": 0.0328, "step": 28480 }, { "grad_norm": 0.2555927038192749, "learning_rate": 8.565917426829013e-05, "loss": 0.0325, "step": 28490 }, { "grad_norm": 0.2510935664176941, "learning_rate": 8.564758186893628e-05, "loss": 0.0297, "step": 28500 }, { "grad_norm": 0.2886267900466919, "learning_rate": 8.563598557121649e-05, "loss": 0.0289, "step": 28510 }, { "grad_norm": 0.24755911529064178, "learning_rate": 8.562438537639888e-05, "loss": 0.0265, "step": 28520 }, { "grad_norm": 0.24299772083759308, "learning_rate": 8.561278128575206e-05, "loss": 0.0301, "step": 28530 }, { "grad_norm": 0.3760422170162201, "learning_rate": 8.5601173300545e-05, "loss": 0.0316, "step": 28540 }, { "grad_norm": 0.35643357038497925, "learning_rate": 8.558956142204717e-05, "loss": 0.0323, "step": 28550 }, { "grad_norm": 0.23188327252864838, "learning_rate": 8.55779456515284e-05, "loss": 0.0323, "step": 28560 }, { "grad_norm": 0.25830090045928955, "learning_rate": 8.556632599025898e-05, "loss": 0.0305, "step": 28570 }, { "grad_norm": 0.2409992665052414, "learning_rate": 8.555470243950964e-05, "loss": 0.0317, "step": 28580 }, { "grad_norm": 0.23119503259658813, "learning_rate": 8.554307500055148e-05, "loss": 0.0329, "step": 28590 }, { "grad_norm": 0.20135939121246338, "learning_rate": 8.553144367465609e-05, "loss": 0.039, "step": 28600 }, { "grad_norm": 0.26401442289352417, "learning_rate": 8.551980846309544e-05, "loss": 0.0284, "step": 28610 }, { "grad_norm": 0.28785741329193115, "learning_rate": 8.550816936714193e-05, "loss": 0.0314, "step": 28620 }, { "grad_norm": 0.3290264904499054, "learning_rate": 8.549652638806841e-05, "loss": 0.0281, "step": 28630 }, { "grad_norm": 0.24020783603191376, "learning_rate": 8.548487952714812e-05, "loss": 0.0281, "step": 28640 }, { "grad_norm": 0.24570216238498688, "learning_rate": 8.547322878565478e-05, "loss": 0.0289, "step": 28650 }, { "grad_norm": 0.3071651756763458, "learning_rate": 8.546157416486245e-05, "loss": 0.0335, "step": 28660 }, { "grad_norm": 0.1856987029314041, "learning_rate": 8.54499156660457e-05, "loss": 0.0273, "step": 28670 }, { "grad_norm": 0.2463897317647934, "learning_rate": 8.543825329047947e-05, "loss": 0.0271, "step": 28680 }, { "grad_norm": 0.22037866711616516, "learning_rate": 8.542658703943913e-05, "loss": 0.0335, "step": 28690 }, { "grad_norm": 0.22368097305297852, "learning_rate": 8.541491691420051e-05, "loss": 0.0265, "step": 28700 }, { "grad_norm": 0.2773926258087158, "learning_rate": 8.54032429160398e-05, "loss": 0.0289, "step": 28710 }, { "grad_norm": 0.2213306427001953, "learning_rate": 8.539156504623369e-05, "loss": 0.033, "step": 28720 }, { "grad_norm": 0.29788708686828613, "learning_rate": 8.537988330605923e-05, "loss": 0.0257, "step": 28730 }, { "grad_norm": 0.21488310396671295, "learning_rate": 8.536819769679393e-05, "loss": 0.0259, "step": 28740 }, { "grad_norm": 0.27301111817359924, "learning_rate": 8.53565082197157e-05, "loss": 0.0295, "step": 28750 }, { "grad_norm": 0.24335090816020966, "learning_rate": 8.534481487610289e-05, "loss": 0.0266, "step": 28760 }, { "grad_norm": 0.21632753312587738, "learning_rate": 8.533311766723428e-05, "loss": 0.0296, "step": 28770 }, { "grad_norm": 0.2352132648229599, "learning_rate": 8.532141659438901e-05, "loss": 0.0325, "step": 28780 }, { "grad_norm": 0.27173924446105957, "learning_rate": 8.530971165884675e-05, "loss": 0.0338, "step": 28790 }, { "grad_norm": 0.23760497570037842, "learning_rate": 8.529800286188752e-05, "loss": 0.0283, "step": 28800 }, { "grad_norm": 0.231704443693161, "learning_rate": 8.528629020479175e-05, "loss": 0.0328, "step": 28810 }, { "grad_norm": 0.30192771553993225, "learning_rate": 8.527457368884033e-05, "loss": 0.0319, "step": 28820 }, { "grad_norm": 0.24316851794719696, "learning_rate": 8.526285331531458e-05, "loss": 0.0308, "step": 28830 }, { "grad_norm": 0.32486429810523987, "learning_rate": 8.525112908549621e-05, "loss": 0.0258, "step": 28840 }, { "grad_norm": 0.2513510584831238, "learning_rate": 8.523940100066735e-05, "loss": 0.0282, "step": 28850 }, { "grad_norm": 0.22328299283981323, "learning_rate": 8.52276690621106e-05, "loss": 0.0284, "step": 28860 }, { "grad_norm": 0.22719238698482513, "learning_rate": 8.521593327110889e-05, "loss": 0.0275, "step": 28870 }, { "grad_norm": 0.29922112822532654, "learning_rate": 8.520419362894569e-05, "loss": 0.0294, "step": 28880 }, { "grad_norm": 0.29170283675193787, "learning_rate": 8.51924501369048e-05, "loss": 0.0281, "step": 28890 }, { "grad_norm": 0.2817547023296356, "learning_rate": 8.518070279627047e-05, "loss": 0.0355, "step": 28900 }, { "grad_norm": 0.36989110708236694, "learning_rate": 8.516895160832737e-05, "loss": 0.0303, "step": 28910 }, { "grad_norm": 0.23863516747951508, "learning_rate": 8.515719657436061e-05, "loss": 0.029, "step": 28920 }, { "grad_norm": 0.25893911719322205, "learning_rate": 8.514543769565568e-05, "loss": 0.0316, "step": 28930 }, { "grad_norm": 0.23471052944660187, "learning_rate": 8.513367497349853e-05, "loss": 0.0316, "step": 28940 }, { "grad_norm": 0.25073155760765076, "learning_rate": 8.51219084091755e-05, "loss": 0.0254, "step": 28950 }, { "grad_norm": 0.2305244654417038, "learning_rate": 8.511013800397338e-05, "loss": 0.0297, "step": 28960 }, { "grad_norm": 0.2385271042585373, "learning_rate": 8.509836375917937e-05, "loss": 0.0276, "step": 28970 }, { "grad_norm": 0.25901150703430176, "learning_rate": 8.508658567608104e-05, "loss": 0.0325, "step": 28980 }, { "grad_norm": 0.2705598771572113, "learning_rate": 8.507480375596647e-05, "loss": 0.0271, "step": 28990 }, { "grad_norm": 0.254306823015213, "learning_rate": 8.506301800012408e-05, "loss": 0.0337, "step": 29000 }, { "grad_norm": 0.2947525382041931, "learning_rate": 8.505122840984278e-05, "loss": 0.0332, "step": 29010 }, { "grad_norm": 0.2147730141878128, "learning_rate": 8.503943498641182e-05, "loss": 0.0308, "step": 29020 }, { "grad_norm": 0.1926053762435913, "learning_rate": 8.502763773112095e-05, "loss": 0.0251, "step": 29030 }, { "grad_norm": 0.3105071783065796, "learning_rate": 8.501583664526026e-05, "loss": 0.0298, "step": 29040 }, { "grad_norm": 0.25694358348846436, "learning_rate": 8.500403173012032e-05, "loss": 0.03, "step": 29050 }, { "grad_norm": 0.2605682611465454, "learning_rate": 8.499222298699211e-05, "loss": 0.027, "step": 29060 }, { "grad_norm": 0.2535853087902069, "learning_rate": 8.498041041716701e-05, "loss": 0.0286, "step": 29070 }, { "grad_norm": 0.17837078869342804, "learning_rate": 8.496859402193681e-05, "loss": 0.0342, "step": 29080 }, { "grad_norm": 0.19995248317718506, "learning_rate": 8.495677380259374e-05, "loss": 0.0298, "step": 29090 }, { "grad_norm": 0.2874472439289093, "learning_rate": 8.494494976043045e-05, "loss": 0.0313, "step": 29100 }, { "grad_norm": 0.26394540071487427, "learning_rate": 8.493312189673998e-05, "loss": 0.031, "step": 29110 }, { "grad_norm": 0.24826723337173462, "learning_rate": 8.492129021281584e-05, "loss": 0.03, "step": 29120 }, { "grad_norm": 0.29240164160728455, "learning_rate": 8.490945470995188e-05, "loss": 0.0322, "step": 29130 }, { "grad_norm": 0.280571848154068, "learning_rate": 8.489761538944247e-05, "loss": 0.0273, "step": 29140 }, { "grad_norm": 0.3440963327884674, "learning_rate": 8.48857722525823e-05, "loss": 0.034, "step": 29150 }, { "grad_norm": 0.3071203827857971, "learning_rate": 8.487392530066652e-05, "loss": 0.0337, "step": 29160 }, { "grad_norm": 0.5771895051002502, "learning_rate": 8.486207453499069e-05, "loss": 0.0338, "step": 29170 }, { "grad_norm": 0.2808007597923279, "learning_rate": 8.485021995685082e-05, "loss": 0.0303, "step": 29180 }, { "grad_norm": 0.22257100045681, "learning_rate": 8.483836156754328e-05, "loss": 0.0295, "step": 29190 }, { "grad_norm": 0.2373453825712204, "learning_rate": 8.482649936836491e-05, "loss": 0.0304, "step": 29200 }, { "grad_norm": 0.29034489393234253, "learning_rate": 8.481463336061293e-05, "loss": 0.031, "step": 29210 }, { "grad_norm": 0.2792394459247589, "learning_rate": 8.480276354558496e-05, "loss": 0.0284, "step": 29220 }, { "grad_norm": 0.20144598186016083, "learning_rate": 8.479088992457913e-05, "loss": 0.0279, "step": 29230 }, { "grad_norm": 0.32532799243927, "learning_rate": 8.477901249889387e-05, "loss": 0.0318, "step": 29240 }, { "grad_norm": 0.2559632658958435, "learning_rate": 8.47671312698281e-05, "loss": 0.0309, "step": 29250 }, { "grad_norm": 0.30340346693992615, "learning_rate": 8.475524623868112e-05, "loss": 0.0322, "step": 29260 }, { "grad_norm": 0.2345718890428543, "learning_rate": 8.474335740675266e-05, "loss": 0.0317, "step": 29270 }, { "grad_norm": 0.22408723831176758, "learning_rate": 8.473146477534289e-05, "loss": 0.031, "step": 29280 }, { "grad_norm": 0.3957621455192566, "learning_rate": 8.471956834575232e-05, "loss": 0.0293, "step": 29290 }, { "grad_norm": 0.25531789660453796, "learning_rate": 8.470766811928197e-05, "loss": 0.0335, "step": 29300 }, { "grad_norm": 0.2568913996219635, "learning_rate": 8.469576409723323e-05, "loss": 0.0399, "step": 29310 }, { "grad_norm": 0.2917158007621765, "learning_rate": 8.468385628090788e-05, "loss": 0.0322, "step": 29320 }, { "grad_norm": 0.23278485238552094, "learning_rate": 8.467194467160815e-05, "loss": 0.034, "step": 29330 }, { "grad_norm": 0.29782429337501526, "learning_rate": 8.466002927063667e-05, "loss": 0.0326, "step": 29340 }, { "grad_norm": 0.25096410512924194, "learning_rate": 8.464811007929651e-05, "loss": 0.0305, "step": 29350 }, { "grad_norm": 0.22958716750144958, "learning_rate": 8.463618709889114e-05, "loss": 0.0373, "step": 29360 }, { "grad_norm": 0.3167720139026642, "learning_rate": 8.462426033072442e-05, "loss": 0.0441, "step": 29370 }, { "grad_norm": 0.3341645896434784, "learning_rate": 8.461232977610061e-05, "loss": 0.0483, "step": 29380 }, { "grad_norm": 0.29672473669052124, "learning_rate": 8.46003954363245e-05, "loss": 0.0386, "step": 29390 }, { "grad_norm": 0.3109791576862335, "learning_rate": 8.458845731270115e-05, "loss": 0.0337, "step": 29400 }, { "grad_norm": 0.29147565364837646, "learning_rate": 8.45765154065361e-05, "loss": 0.0316, "step": 29410 }, { "grad_norm": 0.2752673923969269, "learning_rate": 8.456456971913532e-05, "loss": 0.0282, "step": 29420 }, { "grad_norm": 0.2614745497703552, "learning_rate": 8.455262025180517e-05, "loss": 0.0334, "step": 29430 }, { "grad_norm": 0.2522784173488617, "learning_rate": 8.454066700585242e-05, "loss": 0.0331, "step": 29440 }, { "grad_norm": 0.2798691987991333, "learning_rate": 8.452870998258423e-05, "loss": 0.0324, "step": 29450 }, { "grad_norm": 0.30934569239616394, "learning_rate": 8.451674918330825e-05, "loss": 0.0312, "step": 29460 }, { "grad_norm": 0.24597392976284027, "learning_rate": 8.450478460933246e-05, "loss": 0.0336, "step": 29470 }, { "grad_norm": 0.24407626688480377, "learning_rate": 8.449281626196532e-05, "loss": 0.0393, "step": 29480 }, { "grad_norm": 0.3084338307380676, "learning_rate": 8.448084414251564e-05, "loss": 0.0324, "step": 29490 }, { "grad_norm": 0.2767760157585144, "learning_rate": 8.446886825229271e-05, "loss": 0.0343, "step": 29500 }, { "grad_norm": 0.29121270775794983, "learning_rate": 8.445688859260615e-05, "loss": 0.0278, "step": 29510 }, { "grad_norm": 0.24899013340473175, "learning_rate": 8.444490516476606e-05, "loss": 0.0328, "step": 29520 }, { "grad_norm": 0.26124632358551025, "learning_rate": 8.443291797008293e-05, "loss": 0.0295, "step": 29530 }, { "grad_norm": 0.3220629394054413, "learning_rate": 8.442092700986765e-05, "loss": 0.0309, "step": 29540 }, { "grad_norm": 0.305101215839386, "learning_rate": 8.440893228543156e-05, "loss": 0.0291, "step": 29550 }, { "grad_norm": 0.34797292947769165, "learning_rate": 8.439693379808638e-05, "loss": 0.0302, "step": 29560 }, { "grad_norm": 0.2688521444797516, "learning_rate": 8.43849315491442e-05, "loss": 0.0351, "step": 29570 }, { "grad_norm": 0.2281866818666458, "learning_rate": 8.437292553991763e-05, "loss": 0.0346, "step": 29580 }, { "grad_norm": 0.2918848693370819, "learning_rate": 8.436091577171959e-05, "loss": 0.029, "step": 29590 }, { "grad_norm": 0.1985761523246765, "learning_rate": 8.434890224586347e-05, "loss": 0.0316, "step": 29600 }, { "grad_norm": 0.24711112678050995, "learning_rate": 8.433688496366303e-05, "loss": 0.0264, "step": 29610 }, { "grad_norm": 0.2605181634426117, "learning_rate": 8.432486392643248e-05, "loss": 0.0288, "step": 29620 }, { "grad_norm": 0.27932995557785034, "learning_rate": 8.431283913548643e-05, "loss": 0.0281, "step": 29630 }, { "grad_norm": 0.24358157813549042, "learning_rate": 8.430081059213985e-05, "loss": 0.0326, "step": 29640 }, { "grad_norm": 0.23564042150974274, "learning_rate": 8.428877829770823e-05, "loss": 0.0287, "step": 29650 }, { "grad_norm": 0.3386896550655365, "learning_rate": 8.427674225350735e-05, "loss": 0.0309, "step": 29660 }, { "grad_norm": 0.3140616714954376, "learning_rate": 8.426470246085347e-05, "loss": 0.0325, "step": 29670 }, { "grad_norm": 0.20466922223567963, "learning_rate": 8.425265892106324e-05, "loss": 0.0311, "step": 29680 }, { "grad_norm": 0.26119861006736755, "learning_rate": 8.424061163545374e-05, "loss": 0.0277, "step": 29690 }, { "grad_norm": 0.23006519675254822, "learning_rate": 8.422856060534243e-05, "loss": 0.031, "step": 29700 }, { "grad_norm": 0.33508992195129395, "learning_rate": 8.421650583204718e-05, "loss": 0.0273, "step": 29710 }, { "grad_norm": 0.23269279301166534, "learning_rate": 8.420444731688633e-05, "loss": 0.0297, "step": 29720 }, { "grad_norm": 0.20565547049045563, "learning_rate": 8.419238506117852e-05, "loss": 0.0253, "step": 29730 }, { "grad_norm": 0.8002240061759949, "learning_rate": 8.418031906624289e-05, "loss": 0.0299, "step": 29740 }, { "grad_norm": 0.3996932804584503, "learning_rate": 8.416824933339898e-05, "loss": 0.0379, "step": 29750 }, { "grad_norm": 0.3406488597393036, "learning_rate": 8.415617586396667e-05, "loss": 0.0386, "step": 29760 }, { "grad_norm": 0.23506681621074677, "learning_rate": 8.414409865926632e-05, "loss": 0.034, "step": 29770 }, { "grad_norm": 0.2861521542072296, "learning_rate": 8.413201772061867e-05, "loss": 0.0317, "step": 29780 }, { "grad_norm": 0.5150829553604126, "learning_rate": 8.411993304934488e-05, "loss": 0.0309, "step": 29790 }, { "grad_norm": 0.26086917519569397, "learning_rate": 8.410784464676654e-05, "loss": 0.0305, "step": 29800 }, { "grad_norm": 0.2889534831047058, "learning_rate": 8.409575251420556e-05, "loss": 0.0284, "step": 29810 }, { "grad_norm": 0.2549799084663391, "learning_rate": 8.408365665298435e-05, "loss": 0.0339, "step": 29820 }, { "grad_norm": 0.33775797486305237, "learning_rate": 8.40715570644257e-05, "loss": 0.0387, "step": 29830 }, { "grad_norm": 0.3417617678642273, "learning_rate": 8.40594537498528e-05, "loss": 0.0367, "step": 29840 }, { "grad_norm": 0.25382500886917114, "learning_rate": 8.404734671058924e-05, "loss": 0.0377, "step": 29850 }, { "grad_norm": 0.23568423092365265, "learning_rate": 8.403523594795902e-05, "loss": 0.0378, "step": 29860 }, { "grad_norm": 0.20593369007110596, "learning_rate": 8.402312146328659e-05, "loss": 0.0295, "step": 29870 }, { "grad_norm": 0.27962926030158997, "learning_rate": 8.401100325789675e-05, "loss": 0.0367, "step": 29880 }, { "grad_norm": 0.2663573920726776, "learning_rate": 8.399888133311472e-05, "loss": 0.0354, "step": 29890 }, { "grad_norm": 0.28044530749320984, "learning_rate": 8.398675569026613e-05, "loss": 0.0314, "step": 29900 }, { "grad_norm": 0.22873210906982422, "learning_rate": 8.397462633067705e-05, "loss": 0.0318, "step": 29910 }, { "grad_norm": 0.2620837688446045, "learning_rate": 8.396249325567392e-05, "loss": 0.0241, "step": 29920 }, { "grad_norm": 0.2760600745677948, "learning_rate": 8.395035646658357e-05, "loss": 0.036, "step": 29930 }, { "grad_norm": 0.25147441029548645, "learning_rate": 8.39382159647333e-05, "loss": 0.0311, "step": 29940 }, { "grad_norm": 0.24813206493854523, "learning_rate": 8.392607175145075e-05, "loss": 0.0301, "step": 29950 }, { "grad_norm": 0.25357356667518616, "learning_rate": 8.3913923828064e-05, "loss": 0.0283, "step": 29960 }, { "grad_norm": 0.2768002450466156, "learning_rate": 8.390177219590152e-05, "loss": 0.0315, "step": 29970 }, { "grad_norm": 0.27031564712524414, "learning_rate": 8.388961685629222e-05, "loss": 0.0262, "step": 29980 }, { "grad_norm": 0.2988479733467102, "learning_rate": 8.387745781056536e-05, "loss": 0.0357, "step": 29990 }, { "grad_norm": 0.292532742023468, "learning_rate": 8.386529506005065e-05, "loss": 0.0285, "step": 30000 }, { "grad_norm": 0.2397385686635971, "learning_rate": 8.38531286060782e-05, "loss": 0.0273, "step": 30010 }, { "grad_norm": 0.26640281081199646, "learning_rate": 8.384095844997849e-05, "loss": 0.0309, "step": 30020 }, { "grad_norm": 0.4628446102142334, "learning_rate": 8.382878459308245e-05, "loss": 0.0308, "step": 30030 }, { "grad_norm": 0.22988519072532654, "learning_rate": 8.381660703672138e-05, "loss": 0.0264, "step": 30040 }, { "grad_norm": 0.2704271674156189, "learning_rate": 8.380442578222702e-05, "loss": 0.0264, "step": 30050 }, { "grad_norm": 0.26556307077407837, "learning_rate": 8.379224083093146e-05, "loss": 0.0281, "step": 30060 }, { "grad_norm": 0.2968332767486572, "learning_rate": 8.378005218416727e-05, "loss": 0.0283, "step": 30070 }, { "grad_norm": 0.25420820713043213, "learning_rate": 8.376785984326735e-05, "loss": 0.0254, "step": 30080 }, { "grad_norm": 0.1966623216867447, "learning_rate": 8.375566380956506e-05, "loss": 0.0237, "step": 30090 }, { "grad_norm": 0.2482372373342514, "learning_rate": 8.374346408439411e-05, "loss": 0.0274, "step": 30100 }, { "grad_norm": 0.2316712886095047, "learning_rate": 8.373126066908868e-05, "loss": 0.0243, "step": 30110 }, { "grad_norm": 0.25250381231307983, "learning_rate": 8.371905356498326e-05, "loss": 0.0252, "step": 30120 }, { "grad_norm": 0.2165532410144806, "learning_rate": 8.370684277341288e-05, "loss": 0.0269, "step": 30130 }, { "grad_norm": 0.27470165491104126, "learning_rate": 8.369462829571282e-05, "loss": 0.0277, "step": 30140 }, { "grad_norm": 0.3414822518825531, "learning_rate": 8.36824101332189e-05, "loss": 0.0281, "step": 30150 }, { "grad_norm": 0.24318335950374603, "learning_rate": 8.367018828726721e-05, "loss": 0.0284, "step": 30160 }, { "grad_norm": 0.17227540910243988, "learning_rate": 8.365796275919438e-05, "loss": 0.0268, "step": 30170 }, { "grad_norm": 0.2381562739610672, "learning_rate": 8.364573355033734e-05, "loss": 0.0267, "step": 30180 }, { "grad_norm": 0.2577355206012726, "learning_rate": 8.363350066203346e-05, "loss": 0.0275, "step": 30190 }, { "grad_norm": 0.28701266646385193, "learning_rate": 8.362126409562053e-05, "loss": 0.0315, "step": 30200 }, { "grad_norm": 0.28054049611091614, "learning_rate": 8.360902385243667e-05, "loss": 0.0318, "step": 30210 }, { "grad_norm": 0.25667276978492737, "learning_rate": 8.359677993382052e-05, "loss": 0.0327, "step": 30220 }, { "grad_norm": 0.23690061271190643, "learning_rate": 8.358453234111103e-05, "loss": 0.0235, "step": 30230 }, { "grad_norm": 0.24511852860450745, "learning_rate": 8.357228107564756e-05, "loss": 0.0315, "step": 30240 }, { "grad_norm": 0.3031696677207947, "learning_rate": 8.356002613876993e-05, "loss": 0.0336, "step": 30250 }, { "grad_norm": 0.26829442381858826, "learning_rate": 8.35477675318183e-05, "loss": 0.031, "step": 30260 }, { "grad_norm": 0.26911890506744385, "learning_rate": 8.353550525613323e-05, "loss": 0.0315, "step": 30270 }, { "grad_norm": 0.3034963309764862, "learning_rate": 8.352323931305572e-05, "loss": 0.0299, "step": 30280 }, { "grad_norm": 0.24648083746433258, "learning_rate": 8.351096970392717e-05, "loss": 0.0326, "step": 30290 }, { "grad_norm": 0.33678245544433594, "learning_rate": 8.349869643008937e-05, "loss": 0.0341, "step": 30300 }, { "grad_norm": 0.18965213000774384, "learning_rate": 8.348641949288449e-05, "loss": 0.0308, "step": 30310 }, { "grad_norm": 0.22077424824237823, "learning_rate": 8.34741388936551e-05, "loss": 0.0308, "step": 30320 }, { "grad_norm": 0.2696404755115509, "learning_rate": 8.346185463374423e-05, "loss": 0.029, "step": 30330 }, { "grad_norm": 0.24653317034244537, "learning_rate": 8.344956671449524e-05, "loss": 0.0285, "step": 30340 }, { "grad_norm": 0.3020426034927368, "learning_rate": 8.343727513725192e-05, "loss": 0.0347, "step": 30350 }, { "grad_norm": 0.402793824672699, "learning_rate": 8.342497990335847e-05, "loss": 0.0318, "step": 30360 }, { "grad_norm": 0.22276197373867035, "learning_rate": 8.341268101415946e-05, "loss": 0.0284, "step": 30370 }, { "grad_norm": 0.279347687959671, "learning_rate": 8.34003784709999e-05, "loss": 0.036, "step": 30380 }, { "grad_norm": 0.22672797739505768, "learning_rate": 8.338807227522516e-05, "loss": 0.0251, "step": 30390 }, { "grad_norm": 0.3373226821422577, "learning_rate": 8.337576242818103e-05, "loss": 0.0278, "step": 30400 }, { "grad_norm": 0.3887947201728821, "learning_rate": 8.336344893121372e-05, "loss": 0.0327, "step": 30410 }, { "grad_norm": 0.2311752885580063, "learning_rate": 8.335113178566977e-05, "loss": 0.0291, "step": 30420 }, { "grad_norm": 0.24565349519252777, "learning_rate": 8.33388109928962e-05, "loss": 0.0295, "step": 30430 }, { "grad_norm": 0.32044854760169983, "learning_rate": 8.33264865542404e-05, "loss": 0.0287, "step": 30440 }, { "grad_norm": 0.3019433617591858, "learning_rate": 8.331415847105013e-05, "loss": 0.0278, "step": 30450 }, { "grad_norm": 0.23625534772872925, "learning_rate": 8.330182674467357e-05, "loss": 0.0284, "step": 30460 }, { "grad_norm": 0.16249173879623413, "learning_rate": 8.32894913764593e-05, "loss": 0.0249, "step": 30470 }, { "grad_norm": 0.21754388511180878, "learning_rate": 8.327715236775633e-05, "loss": 0.0323, "step": 30480 }, { "grad_norm": 0.2491663694381714, "learning_rate": 8.326480971991398e-05, "loss": 0.026, "step": 30490 }, { "grad_norm": 0.2615678012371063, "learning_rate": 8.325246343428206e-05, "loss": 0.0248, "step": 30500 }, { "grad_norm": 0.30412253737449646, "learning_rate": 8.324011351221072e-05, "loss": 0.0262, "step": 30510 }, { "grad_norm": 0.20667442679405212, "learning_rate": 8.322775995505057e-05, "loss": 0.0328, "step": 30520 }, { "grad_norm": 0.22762566804885864, "learning_rate": 8.321540276415254e-05, "loss": 0.0264, "step": 30530 }, { "grad_norm": 0.3049929738044739, "learning_rate": 8.320304194086798e-05, "loss": 0.0305, "step": 30540 }, { "grad_norm": 0.2491758167743683, "learning_rate": 8.31906774865487e-05, "loss": 0.0311, "step": 30550 }, { "grad_norm": 0.3144790828227997, "learning_rate": 8.317830940254682e-05, "loss": 0.0299, "step": 30560 }, { "grad_norm": 0.2710724174976349, "learning_rate": 8.316593769021491e-05, "loss": 0.0287, "step": 30570 }, { "grad_norm": 0.2155836522579193, "learning_rate": 8.315356235090592e-05, "loss": 0.027, "step": 30580 }, { "grad_norm": 0.21409867703914642, "learning_rate": 8.314118338597319e-05, "loss": 0.0285, "step": 30590 }, { "grad_norm": 0.28861820697784424, "learning_rate": 8.312880079677048e-05, "loss": 0.0307, "step": 30600 }, { "grad_norm": 0.22330160439014435, "learning_rate": 8.311641458465191e-05, "loss": 0.0299, "step": 30610 }, { "grad_norm": 0.3540857136249542, "learning_rate": 8.310402475097205e-05, "loss": 0.0316, "step": 30620 }, { "grad_norm": 0.283242404460907, "learning_rate": 8.309163129708581e-05, "loss": 0.029, "step": 30630 }, { "grad_norm": 0.33927568793296814, "learning_rate": 8.307923422434852e-05, "loss": 0.0376, "step": 30640 }, { "grad_norm": 0.3289174735546112, "learning_rate": 8.30668335341159e-05, "loss": 0.0319, "step": 30650 }, { "grad_norm": 0.38276243209838867, "learning_rate": 8.305442922774408e-05, "loss": 0.031, "step": 30660 }, { "grad_norm": 0.24894732236862183, "learning_rate": 8.304202130658959e-05, "loss": 0.0323, "step": 30670 }, { "grad_norm": 0.2733711302280426, "learning_rate": 8.302960977200931e-05, "loss": 0.0334, "step": 30680 }, { "grad_norm": 0.28340771794319153, "learning_rate": 8.301719462536058e-05, "loss": 0.0304, "step": 30690 }, { "grad_norm": 0.30186977982521057, "learning_rate": 8.300477586800108e-05, "loss": 0.0292, "step": 30700 }, { "grad_norm": 0.2580084204673767, "learning_rate": 8.299235350128892e-05, "loss": 0.0298, "step": 30710 }, { "grad_norm": 0.25602254271507263, "learning_rate": 8.297992752658258e-05, "loss": 0.0308, "step": 30720 }, { "grad_norm": 0.31766262650489807, "learning_rate": 8.296749794524094e-05, "loss": 0.0366, "step": 30730 }, { "grad_norm": 0.22999343276023865, "learning_rate": 8.295506475862332e-05, "loss": 0.0275, "step": 30740 }, { "grad_norm": 0.25314196944236755, "learning_rate": 8.294262796808933e-05, "loss": 0.0341, "step": 30750 }, { "grad_norm": 0.26321810483932495, "learning_rate": 8.293018757499909e-05, "loss": 0.0326, "step": 30760 }, { "grad_norm": 0.2778775095939636, "learning_rate": 8.291774358071305e-05, "loss": 0.0332, "step": 30770 }, { "grad_norm": 0.26455676555633545, "learning_rate": 8.290529598659205e-05, "loss": 0.0288, "step": 30780 }, { "grad_norm": 0.3245035409927368, "learning_rate": 8.289284479399738e-05, "loss": 0.0312, "step": 30790 }, { "grad_norm": 0.268134206533432, "learning_rate": 8.288039000429064e-05, "loss": 0.0329, "step": 30800 }, { "grad_norm": 0.2800338864326477, "learning_rate": 8.286793161883388e-05, "loss": 0.0338, "step": 30810 }, { "grad_norm": 0.2845776081085205, "learning_rate": 8.285546963898954e-05, "loss": 0.0361, "step": 30820 }, { "grad_norm": 0.269604355096817, "learning_rate": 8.284300406612044e-05, "loss": 0.0313, "step": 30830 }, { "grad_norm": 0.3010546565055847, "learning_rate": 8.283053490158978e-05, "loss": 0.0301, "step": 30840 }, { "grad_norm": 0.22335387766361237, "learning_rate": 8.28180621467612e-05, "loss": 0.0294, "step": 30850 }, { "grad_norm": 0.25493595004081726, "learning_rate": 8.280558580299867e-05, "loss": 0.029, "step": 30860 }, { "grad_norm": 0.26468056440353394, "learning_rate": 8.279310587166661e-05, "loss": 0.0324, "step": 30870 }, { "grad_norm": 0.21124625205993652, "learning_rate": 8.278062235412978e-05, "loss": 0.0295, "step": 30880 }, { "grad_norm": 0.23237495124340057, "learning_rate": 8.276813525175339e-05, "loss": 0.0303, "step": 30890 }, { "grad_norm": 0.22169408202171326, "learning_rate": 8.2755644565903e-05, "loss": 0.0296, "step": 30900 }, { "grad_norm": 0.2246840000152588, "learning_rate": 8.274315029794454e-05, "loss": 0.0349, "step": 30910 }, { "grad_norm": 0.2028099000453949, "learning_rate": 8.273065244924443e-05, "loss": 0.0239, "step": 30920 }, { "grad_norm": 0.21179655194282532, "learning_rate": 8.271815102116936e-05, "loss": 0.0321, "step": 30930 }, { "grad_norm": 0.30640366673469543, "learning_rate": 8.270564601508648e-05, "loss": 0.0331, "step": 30940 }, { "grad_norm": 0.27589812874794006, "learning_rate": 8.269313743236333e-05, "loss": 0.0325, "step": 30950 }, { "grad_norm": 0.26049545407295227, "learning_rate": 8.268062527436783e-05, "loss": 0.0317, "step": 30960 }, { "grad_norm": 0.2516367435455322, "learning_rate": 8.266810954246828e-05, "loss": 0.0362, "step": 30970 }, { "grad_norm": 0.30833011865615845, "learning_rate": 8.265559023803338e-05, "loss": 0.032, "step": 30980 }, { "grad_norm": 0.27719828486442566, "learning_rate": 8.264306736243223e-05, "loss": 0.0322, "step": 30990 }, { "grad_norm": 0.3660530745983124, "learning_rate": 8.263054091703432e-05, "loss": 0.0338, "step": 31000 }, { "grad_norm": 0.24588380753993988, "learning_rate": 8.26180109032095e-05, "loss": 0.0347, "step": 31010 }, { "grad_norm": 0.2281464785337448, "learning_rate": 8.260547732232807e-05, "loss": 0.0297, "step": 31020 }, { "grad_norm": 0.282691091299057, "learning_rate": 8.259294017576063e-05, "loss": 0.0284, "step": 31030 }, { "grad_norm": 0.35071060061454773, "learning_rate": 8.258039946487828e-05, "loss": 0.0337, "step": 31040 }, { "grad_norm": 0.2683883309364319, "learning_rate": 8.256785519105241e-05, "loss": 0.0296, "step": 31050 }, { "grad_norm": 0.2273825854063034, "learning_rate": 8.255530735565488e-05, "loss": 0.03, "step": 31060 }, { "grad_norm": 0.24877561628818512, "learning_rate": 8.254275596005787e-05, "loss": 0.0286, "step": 31070 }, { "grad_norm": 0.275162011384964, "learning_rate": 8.2530201005634e-05, "loss": 0.0305, "step": 31080 }, { "grad_norm": 0.5593597292900085, "learning_rate": 8.251764249375626e-05, "loss": 0.0339, "step": 31090 }, { "grad_norm": 0.31893035769462585, "learning_rate": 8.250508042579803e-05, "loss": 0.0312, "step": 31100 }, { "grad_norm": 0.2628602683544159, "learning_rate": 8.249251480313307e-05, "loss": 0.0273, "step": 31110 }, { "grad_norm": 0.2431182563304901, "learning_rate": 8.247994562713555e-05, "loss": 0.0293, "step": 31120 }, { "grad_norm": 0.3293149173259735, "learning_rate": 8.246737289917998e-05, "loss": 0.0302, "step": 31130 }, { "grad_norm": 0.19691017270088196, "learning_rate": 8.245479662064135e-05, "loss": 0.0252, "step": 31140 }, { "grad_norm": 0.23824414610862732, "learning_rate": 8.244221679289496e-05, "loss": 0.0256, "step": 31150 }, { "grad_norm": 0.27711206674575806, "learning_rate": 8.242963341731652e-05, "loss": 0.0279, "step": 31160 }, { "grad_norm": 0.22146917879581451, "learning_rate": 8.24170464952821e-05, "loss": 0.0283, "step": 31170 }, { "grad_norm": 0.24490076303482056, "learning_rate": 8.240445602816824e-05, "loss": 0.0293, "step": 31180 }, { "grad_norm": 0.1894853562116623, "learning_rate": 8.239186201735179e-05, "loss": 0.0326, "step": 31190 }, { "grad_norm": 0.31666526198387146, "learning_rate": 8.237926446420998e-05, "loss": 0.0285, "step": 31200 }, { "grad_norm": 0.29982006549835205, "learning_rate": 8.236666337012052e-05, "loss": 0.033, "step": 31210 }, { "grad_norm": 0.21712136268615723, "learning_rate": 8.23540587364614e-05, "loss": 0.0282, "step": 31220 }, { "grad_norm": 0.26897063851356506, "learning_rate": 8.234145056461107e-05, "loss": 0.0297, "step": 31230 }, { "grad_norm": 0.24762120842933655, "learning_rate": 8.232883885594832e-05, "loss": 0.0253, "step": 31240 }, { "grad_norm": 0.24716916680335999, "learning_rate": 8.231622361185236e-05, "loss": 0.027, "step": 31250 }, { "grad_norm": 0.27776190638542175, "learning_rate": 8.230360483370278e-05, "loss": 0.0271, "step": 31260 }, { "grad_norm": 0.20602291822433472, "learning_rate": 8.229098252287953e-05, "loss": 0.0276, "step": 31270 }, { "grad_norm": 0.2134481966495514, "learning_rate": 8.2278356680763e-05, "loss": 0.0292, "step": 31280 }, { "grad_norm": 0.2540072798728943, "learning_rate": 8.22657273087339e-05, "loss": 0.0301, "step": 31290 }, { "grad_norm": 0.2540978193283081, "learning_rate": 8.225309440817336e-05, "loss": 0.0336, "step": 31300 }, { "grad_norm": 0.274540513753891, "learning_rate": 8.224045798046293e-05, "loss": 0.0329, "step": 31310 }, { "grad_norm": 0.2803345322608948, "learning_rate": 8.22278180269845e-05, "loss": 0.0271, "step": 31320 }, { "grad_norm": 0.26069211959838867, "learning_rate": 8.221517454912031e-05, "loss": 0.0323, "step": 31330 }, { "grad_norm": 0.2632782757282257, "learning_rate": 8.220252754825308e-05, "loss": 0.0258, "step": 31340 }, { "grad_norm": 0.2634093463420868, "learning_rate": 8.218987702576586e-05, "loss": 0.0271, "step": 31350 }, { "grad_norm": 0.33343368768692017, "learning_rate": 8.217722298304207e-05, "loss": 0.0308, "step": 31360 }, { "grad_norm": 0.2778778374195099, "learning_rate": 8.216456542146557e-05, "loss": 0.0267, "step": 31370 }, { "grad_norm": 0.2598889172077179, "learning_rate": 8.215190434242055e-05, "loss": 0.0281, "step": 31380 }, { "grad_norm": 0.20400367677211761, "learning_rate": 8.213923974729161e-05, "loss": 0.024, "step": 31390 }, { "grad_norm": 0.22562366724014282, "learning_rate": 8.212657163746373e-05, "loss": 0.0256, "step": 31400 }, { "grad_norm": 0.18873906135559082, "learning_rate": 8.211390001432227e-05, "loss": 0.0271, "step": 31410 }, { "grad_norm": 0.23904848098754883, "learning_rate": 8.210122487925297e-05, "loss": 0.0279, "step": 31420 }, { "grad_norm": 0.16794008016586304, "learning_rate": 8.208854623364202e-05, "loss": 0.0296, "step": 31430 }, { "grad_norm": 0.23463480174541473, "learning_rate": 8.207586407887589e-05, "loss": 0.0278, "step": 31440 }, { "grad_norm": 0.25064489245414734, "learning_rate": 8.206317841634148e-05, "loss": 0.027, "step": 31450 }, { "grad_norm": 0.22632408142089844, "learning_rate": 8.205048924742609e-05, "loss": 0.0308, "step": 31460 }, { "grad_norm": 0.22814740240573883, "learning_rate": 8.203779657351738e-05, "loss": 0.0331, "step": 31470 }, { "grad_norm": 0.23470352590084076, "learning_rate": 8.20251003960034e-05, "loss": 0.0324, "step": 31480 }, { "grad_norm": 0.2396182119846344, "learning_rate": 8.201240071627258e-05, "loss": 0.0262, "step": 31490 }, { "grad_norm": 0.23728063702583313, "learning_rate": 8.199969753571377e-05, "loss": 0.0302, "step": 31500 }, { "grad_norm": 0.22087547183036804, "learning_rate": 8.198699085571615e-05, "loss": 0.0261, "step": 31510 }, { "grad_norm": 0.21729396283626556, "learning_rate": 8.197428067766928e-05, "loss": 0.0395, "step": 31520 }, { "grad_norm": 0.2672380208969116, "learning_rate": 8.196156700296316e-05, "loss": 0.0317, "step": 31530 }, { "grad_norm": 0.2789834141731262, "learning_rate": 8.194884983298814e-05, "loss": 0.0355, "step": 31540 }, { "grad_norm": 0.2514617443084717, "learning_rate": 8.193612916913491e-05, "loss": 0.0284, "step": 31550 }, { "grad_norm": 0.320053368806839, "learning_rate": 8.192340501279463e-05, "loss": 0.0304, "step": 31560 }, { "grad_norm": 0.280678927898407, "learning_rate": 8.191067736535876e-05, "loss": 0.0335, "step": 31570 }, { "grad_norm": 0.28430214524269104, "learning_rate": 8.18979462282192e-05, "loss": 0.0313, "step": 31580 }, { "grad_norm": 0.23518258333206177, "learning_rate": 8.188521160276819e-05, "loss": 0.032, "step": 31590 }, { "grad_norm": 0.21469458937644958, "learning_rate": 8.187247349039837e-05, "loss": 0.0277, "step": 31600 }, { "grad_norm": 0.2772291898727417, "learning_rate": 8.185973189250278e-05, "loss": 0.0321, "step": 31610 }, { "grad_norm": 0.2805357277393341, "learning_rate": 8.184698681047482e-05, "loss": 0.0295, "step": 31620 }, { "grad_norm": 0.28556814789772034, "learning_rate": 8.183423824570827e-05, "loss": 0.036, "step": 31630 }, { "grad_norm": 0.19960105419158936, "learning_rate": 8.182148619959725e-05, "loss": 0.0249, "step": 31640 }, { "grad_norm": 0.27223849296569824, "learning_rate": 8.180873067353636e-05, "loss": 0.0293, "step": 31650 }, { "grad_norm": 0.23218458890914917, "learning_rate": 8.179597166892052e-05, "loss": 0.0226, "step": 31660 }, { "grad_norm": 0.20012541115283966, "learning_rate": 8.178320918714501e-05, "loss": 0.0314, "step": 31670 }, { "grad_norm": 0.1979704350233078, "learning_rate": 8.177044322960554e-05, "loss": 0.025, "step": 31680 }, { "grad_norm": 0.26631513237953186, "learning_rate": 8.175767379769816e-05, "loss": 0.0273, "step": 31690 }, { "grad_norm": 0.312764972448349, "learning_rate": 8.174490089281932e-05, "loss": 0.0282, "step": 31700 }, { "grad_norm": 0.2674005329608917, "learning_rate": 8.173212451636584e-05, "loss": 0.0307, "step": 31710 }, { "grad_norm": 0.2246864289045334, "learning_rate": 8.171934466973493e-05, "loss": 0.0327, "step": 31720 }, { "grad_norm": 0.18548591434955597, "learning_rate": 8.170656135432418e-05, "loss": 0.0279, "step": 31730 }, { "grad_norm": 0.2543645203113556, "learning_rate": 8.169377457153155e-05, "loss": 0.0289, "step": 31740 }, { "grad_norm": 0.26260876655578613, "learning_rate": 8.168098432275539e-05, "loss": 0.031, "step": 31750 }, { "grad_norm": 0.2817261815071106, "learning_rate": 8.166819060939442e-05, "loss": 0.0308, "step": 31760 }, { "grad_norm": 0.19880683720111847, "learning_rate": 8.165539343284772e-05, "loss": 0.0309, "step": 31770 }, { "grad_norm": 0.20875705778598785, "learning_rate": 8.16425927945148e-05, "loss": 0.0315, "step": 31780 }, { "grad_norm": 0.32979628443717957, "learning_rate": 8.162978869579551e-05, "loss": 0.0357, "step": 31790 }, { "grad_norm": 0.2637382447719574, "learning_rate": 8.161698113809007e-05, "loss": 0.0348, "step": 31800 }, { "grad_norm": 0.27524885535240173, "learning_rate": 8.160417012279911e-05, "loss": 0.0306, "step": 31810 }, { "grad_norm": 0.24197043478488922, "learning_rate": 8.159135565132363e-05, "loss": 0.0305, "step": 31820 }, { "grad_norm": 0.3115691840648651, "learning_rate": 8.157853772506498e-05, "loss": 0.0352, "step": 31830 }, { "grad_norm": 0.23142606019973755, "learning_rate": 8.156571634542494e-05, "loss": 0.0321, "step": 31840 }, { "grad_norm": 0.27819114923477173, "learning_rate": 8.15528915138056e-05, "loss": 0.0272, "step": 31850 }, { "grad_norm": 0.29751577973365784, "learning_rate": 8.154006323160949e-05, "loss": 0.0298, "step": 31860 }, { "grad_norm": 0.2701883316040039, "learning_rate": 8.152723150023949e-05, "loss": 0.0301, "step": 31870 }, { "grad_norm": 0.23595723509788513, "learning_rate": 8.151439632109886e-05, "loss": 0.0281, "step": 31880 }, { "grad_norm": 0.2338618040084839, "learning_rate": 8.150155769559122e-05, "loss": 0.0266, "step": 31890 }, { "grad_norm": 0.2642364203929901, "learning_rate": 8.148871562512058e-05, "loss": 0.0298, "step": 31900 }, { "grad_norm": 0.27228647470474243, "learning_rate": 8.147587011109136e-05, "loss": 0.0269, "step": 31910 }, { "grad_norm": 0.2434777319431305, "learning_rate": 8.14630211549083e-05, "loss": 0.0314, "step": 31920 }, { "grad_norm": 0.23682966828346252, "learning_rate": 8.145016875797655e-05, "loss": 0.0317, "step": 31930 }, { "grad_norm": 0.2304771989583969, "learning_rate": 8.143731292170164e-05, "loss": 0.0282, "step": 31940 }, { "grad_norm": 0.207182377576828, "learning_rate": 8.142445364748944e-05, "loss": 0.0286, "step": 31950 }, { "grad_norm": 0.2568843364715576, "learning_rate": 8.141159093674624e-05, "loss": 0.0273, "step": 31960 }, { "grad_norm": 0.23843546211719513, "learning_rate": 8.139872479087869e-05, "loss": 0.0318, "step": 31970 }, { "grad_norm": 0.24618184566497803, "learning_rate": 8.13858552112938e-05, "loss": 0.0269, "step": 31980 }, { "grad_norm": 0.21440595388412476, "learning_rate": 8.137298219939895e-05, "loss": 0.0334, "step": 31990 }, { "grad_norm": 0.33892256021499634, "learning_rate": 8.136010575660196e-05, "loss": 0.0334, "step": 32000 }, { "grad_norm": 0.18804828822612762, "learning_rate": 8.134722588431095e-05, "loss": 0.0285, "step": 32010 }, { "grad_norm": 0.267068088054657, "learning_rate": 8.133434258393444e-05, "loss": 0.0337, "step": 32020 }, { "grad_norm": 0.2699849307537079, "learning_rate": 8.132145585688134e-05, "loss": 0.0307, "step": 32030 }, { "grad_norm": 0.23705841600894928, "learning_rate": 8.130856570456093e-05, "loss": 0.0298, "step": 32040 }, { "grad_norm": 0.23091284930706024, "learning_rate": 8.129567212838283e-05, "loss": 0.028, "step": 32050 }, { "grad_norm": 0.2974855303764343, "learning_rate": 8.128277512975708e-05, "loss": 0.0275, "step": 32060 }, { "grad_norm": 0.2535378336906433, "learning_rate": 8.126987471009408e-05, "loss": 0.0265, "step": 32070 }, { "grad_norm": 0.21153071522712708, "learning_rate": 8.125697087080459e-05, "loss": 0.0242, "step": 32080 }, { "grad_norm": 0.247707337141037, "learning_rate": 8.124406361329976e-05, "loss": 0.0267, "step": 32090 }, { "grad_norm": 0.21213741600513458, "learning_rate": 8.123115293899111e-05, "loss": 0.0271, "step": 32100 }, { "grad_norm": 0.21566538512706757, "learning_rate": 8.121823884929055e-05, "loss": 0.027, "step": 32110 }, { "grad_norm": 0.2810993194580078, "learning_rate": 8.12053213456103e-05, "loss": 0.0307, "step": 32120 }, { "grad_norm": 0.25995707511901855, "learning_rate": 8.119240042936303e-05, "loss": 0.0284, "step": 32130 }, { "grad_norm": 0.31627941131591797, "learning_rate": 8.117947610196175e-05, "loss": 0.0287, "step": 32140 }, { "grad_norm": 0.2579256594181061, "learning_rate": 8.116654836481982e-05, "loss": 0.0278, "step": 32150 }, { "grad_norm": 0.30195391178131104, "learning_rate": 8.115361721935106e-05, "loss": 0.0303, "step": 32160 }, { "grad_norm": 0.33971932530403137, "learning_rate": 8.114068266696953e-05, "loss": 0.0278, "step": 32170 }, { "grad_norm": 0.30077701807022095, "learning_rate": 8.112774470908978e-05, "loss": 0.0314, "step": 32180 }, { "grad_norm": 0.28169316053390503, "learning_rate": 8.111480334712665e-05, "loss": 0.0284, "step": 32190 }, { "grad_norm": 0.28661274909973145, "learning_rate": 8.110185858249542e-05, "loss": 0.0307, "step": 32200 }, { "grad_norm": 0.24180178344249725, "learning_rate": 8.108891041661168e-05, "loss": 0.0272, "step": 32210 }, { "grad_norm": 0.2843475937843323, "learning_rate": 8.107595885089146e-05, "loss": 0.0263, "step": 32220 }, { "grad_norm": 0.25380751490592957, "learning_rate": 8.106300388675108e-05, "loss": 0.0306, "step": 32230 }, { "grad_norm": 0.2323123961687088, "learning_rate": 8.105004552560731e-05, "loss": 0.0277, "step": 32240 }, { "grad_norm": 0.3600471317768097, "learning_rate": 8.103708376887724e-05, "loss": 0.0294, "step": 32250 }, { "grad_norm": 0.21124130487442017, "learning_rate": 8.102411861797836e-05, "loss": 0.0294, "step": 32260 }, { "grad_norm": 0.22584083676338196, "learning_rate": 8.101115007432851e-05, "loss": 0.0296, "step": 32270 }, { "grad_norm": 0.2199704945087433, "learning_rate": 8.09981781393459e-05, "loss": 0.029, "step": 32280 }, { "grad_norm": 0.2759917676448822, "learning_rate": 8.098520281444915e-05, "loss": 0.0318, "step": 32290 }, { "grad_norm": 0.3119020164012909, "learning_rate": 8.09722241010572e-05, "loss": 0.0385, "step": 32300 }, { "grad_norm": 0.2753755748271942, "learning_rate": 8.095924200058939e-05, "loss": 0.0277, "step": 32310 }, { "grad_norm": 0.32276761531829834, "learning_rate": 8.094625651446541e-05, "loss": 0.0319, "step": 32320 }, { "grad_norm": 0.25781548023223877, "learning_rate": 8.093326764410536e-05, "loss": 0.0273, "step": 32330 }, { "grad_norm": 0.29634982347488403, "learning_rate": 8.092027539092966e-05, "loss": 0.0302, "step": 32340 }, { "grad_norm": 0.2624419331550598, "learning_rate": 8.090727975635913e-05, "loss": 0.0295, "step": 32350 }, { "grad_norm": 0.25035834312438965, "learning_rate": 8.089428074181497e-05, "loss": 0.0256, "step": 32360 }, { "grad_norm": 0.30418336391448975, "learning_rate": 8.088127834871871e-05, "loss": 0.0261, "step": 32370 }, { "grad_norm": 0.23308275640010834, "learning_rate": 8.086827257849226e-05, "loss": 0.0271, "step": 32380 }, { "grad_norm": 0.2348172515630722, "learning_rate": 8.085526343255795e-05, "loss": 0.0254, "step": 32390 }, { "grad_norm": 0.27584195137023926, "learning_rate": 8.084225091233842e-05, "loss": 0.028, "step": 32400 }, { "grad_norm": 0.27148401737213135, "learning_rate": 8.082923501925668e-05, "loss": 0.0283, "step": 32410 }, { "grad_norm": 0.17922575771808624, "learning_rate": 8.081621575473617e-05, "loss": 0.0293, "step": 32420 }, { "grad_norm": 0.3011679947376251, "learning_rate": 8.080319312020064e-05, "loss": 0.0333, "step": 32430 }, { "grad_norm": 0.2240116000175476, "learning_rate": 8.079016711707421e-05, "loss": 0.03, "step": 32440 }, { "grad_norm": 0.31634417176246643, "learning_rate": 8.077713774678139e-05, "loss": 0.0271, "step": 32450 }, { "grad_norm": 0.2892787754535675, "learning_rate": 8.076410501074707e-05, "loss": 0.029, "step": 32460 }, { "grad_norm": 0.27506023645401, "learning_rate": 8.075106891039647e-05, "loss": 0.0326, "step": 32470 }, { "grad_norm": 0.20718412101268768, "learning_rate": 8.073802944715523e-05, "loss": 0.0292, "step": 32480 }, { "grad_norm": 0.17858357727527618, "learning_rate": 8.072498662244929e-05, "loss": 0.0247, "step": 32490 }, { "grad_norm": 0.22783301770687103, "learning_rate": 8.0711940437705e-05, "loss": 0.0246, "step": 32500 }, { "grad_norm": 0.2521284222602844, "learning_rate": 8.06988908943491e-05, "loss": 0.0287, "step": 32510 }, { "grad_norm": 0.2819381654262543, "learning_rate": 8.068583799380863e-05, "loss": 0.0334, "step": 32520 }, { "grad_norm": 0.26900529861450195, "learning_rate": 8.067278173751104e-05, "loss": 0.031, "step": 32530 }, { "grad_norm": 0.2995706796646118, "learning_rate": 8.065972212688417e-05, "loss": 0.0283, "step": 32540 }, { "grad_norm": 0.2854052484035492, "learning_rate": 8.064665916335618e-05, "loss": 0.0304, "step": 32550 }, { "grad_norm": 0.3530203402042389, "learning_rate": 8.063359284835564e-05, "loss": 0.0356, "step": 32560 }, { "grad_norm": 0.3157041072845459, "learning_rate": 8.062052318331142e-05, "loss": 0.0278, "step": 32570 }, { "grad_norm": 0.23961229622364044, "learning_rate": 8.060745016965283e-05, "loss": 0.0244, "step": 32580 }, { "grad_norm": 0.2546701431274414, "learning_rate": 8.059437380880952e-05, "loss": 0.0262, "step": 32590 }, { "grad_norm": 0.3138073980808258, "learning_rate": 8.058129410221146e-05, "loss": 0.0285, "step": 32600 }, { "grad_norm": 0.2116805464029312, "learning_rate": 8.056821105128908e-05, "loss": 0.0291, "step": 32610 }, { "grad_norm": 0.3462533950805664, "learning_rate": 8.05551246574731e-05, "loss": 0.034, "step": 32620 }, { "grad_norm": 0.3189466595649719, "learning_rate": 8.05420349221946e-05, "loss": 0.0332, "step": 32630 }, { "grad_norm": 0.22813203930854797, "learning_rate": 8.05289418468851e-05, "loss": 0.0283, "step": 32640 }, { "grad_norm": 0.2581312656402588, "learning_rate": 8.051584543297642e-05, "loss": 0.0292, "step": 32650 }, { "grad_norm": 0.22291210293769836, "learning_rate": 8.050274568190074e-05, "loss": 0.0283, "step": 32660 }, { "grad_norm": 0.2964385151863098, "learning_rate": 8.048964259509067e-05, "loss": 0.03, "step": 32670 }, { "grad_norm": 0.1686561405658722, "learning_rate": 8.047653617397914e-05, "loss": 0.0267, "step": 32680 }, { "grad_norm": 0.21721208095550537, "learning_rate": 8.046342641999941e-05, "loss": 0.0272, "step": 32690 }, { "grad_norm": 0.25402161478996277, "learning_rate": 8.045031333458517e-05, "loss": 0.029, "step": 32700 }, { "grad_norm": 0.27228838205337524, "learning_rate": 8.043719691917047e-05, "loss": 0.0334, "step": 32710 }, { "grad_norm": 0.2539026141166687, "learning_rate": 8.042407717518966e-05, "loss": 0.0311, "step": 32720 }, { "grad_norm": 0.25079840421676636, "learning_rate": 8.041095410407751e-05, "loss": 0.0259, "step": 32730 }, { "grad_norm": 0.21852533519268036, "learning_rate": 8.039782770726913e-05, "loss": 0.0258, "step": 32740 }, { "grad_norm": 0.24694423377513885, "learning_rate": 8.038469798620004e-05, "loss": 0.0239, "step": 32750 }, { "grad_norm": 0.41818827390670776, "learning_rate": 8.037156494230604e-05, "loss": 0.0283, "step": 32760 }, { "grad_norm": 0.23876510560512543, "learning_rate": 8.035842857702338e-05, "loss": 0.0264, "step": 32770 }, { "grad_norm": 0.2611294984817505, "learning_rate": 8.03452888917886e-05, "loss": 0.0303, "step": 32780 }, { "grad_norm": 0.2899409830570221, "learning_rate": 8.033214588803866e-05, "loss": 0.0332, "step": 32790 }, { "grad_norm": 0.2236192524433136, "learning_rate": 8.031899956721083e-05, "loss": 0.0301, "step": 32800 }, { "grad_norm": 0.30727818608283997, "learning_rate": 8.030584993074282e-05, "loss": 0.0342, "step": 32810 }, { "grad_norm": 0.3162018060684204, "learning_rate": 8.02926969800726e-05, "loss": 0.0275, "step": 32820 }, { "grad_norm": 0.2994557023048401, "learning_rate": 8.027954071663859e-05, "loss": 0.0354, "step": 32830 }, { "grad_norm": 0.320331871509552, "learning_rate": 8.026638114187954e-05, "loss": 0.0323, "step": 32840 }, { "grad_norm": 0.27741703391075134, "learning_rate": 8.025321825723456e-05, "loss": 0.026, "step": 32850 }, { "grad_norm": 0.225600928068161, "learning_rate": 8.02400520641431e-05, "loss": 0.0262, "step": 32860 }, { "grad_norm": 0.2263726443052292, "learning_rate": 8.022688256404501e-05, "loss": 0.0283, "step": 32870 }, { "grad_norm": 0.28348207473754883, "learning_rate": 8.02137097583805e-05, "loss": 0.0245, "step": 32880 }, { "grad_norm": 0.22083544731140137, "learning_rate": 8.02005336485901e-05, "loss": 0.0299, "step": 32890 }, { "grad_norm": 0.18899163603782654, "learning_rate": 8.018735423611476e-05, "loss": 0.0272, "step": 32900 }, { "grad_norm": 0.20126871764659882, "learning_rate": 8.017417152239574e-05, "loss": 0.0253, "step": 32910 }, { "grad_norm": 0.21570751070976257, "learning_rate": 8.01609855088747e-05, "loss": 0.0332, "step": 32920 }, { "grad_norm": 0.2070460468530655, "learning_rate": 8.014779619699362e-05, "loss": 0.0309, "step": 32930 }, { "grad_norm": 0.2785138785839081, "learning_rate": 8.013460358819489e-05, "loss": 0.0281, "step": 32940 }, { "grad_norm": 0.2200293242931366, "learning_rate": 8.01214076839212e-05, "loss": 0.0295, "step": 32950 }, { "grad_norm": 0.28789791464805603, "learning_rate": 8.010820848561565e-05, "loss": 0.0288, "step": 32960 }, { "grad_norm": 0.22220061719417572, "learning_rate": 8.009500599472171e-05, "loss": 0.03, "step": 32970 }, { "grad_norm": 0.22822298109531403, "learning_rate": 8.008180021268314e-05, "loss": 0.0241, "step": 32980 }, { "grad_norm": 0.23772649466991425, "learning_rate": 8.006859114094414e-05, "loss": 0.0251, "step": 32990 }, { "grad_norm": 0.21115735173225403, "learning_rate": 8.005537878094921e-05, "loss": 0.0297, "step": 33000 }, { "grad_norm": 0.25532373785972595, "learning_rate": 8.004216313414323e-05, "loss": 0.0255, "step": 33010 }, { "grad_norm": 0.2275335192680359, "learning_rate": 8.002894420197149e-05, "loss": 0.0264, "step": 33020 }, { "grad_norm": 0.2295459657907486, "learning_rate": 8.001572198587954e-05, "loss": 0.0273, "step": 33030 }, { "grad_norm": 0.32379794120788574, "learning_rate": 8.000249648731338e-05, "loss": 0.0287, "step": 33040 }, { "grad_norm": 0.2754579186439514, "learning_rate": 7.998926770771928e-05, "loss": 0.0288, "step": 33050 }, { "grad_norm": 0.21131612360477448, "learning_rate": 7.997603564854397e-05, "loss": 0.0268, "step": 33060 }, { "grad_norm": 0.2473432719707489, "learning_rate": 7.996280031123448e-05, "loss": 0.0241, "step": 33070 }, { "grad_norm": 0.22878707945346832, "learning_rate": 7.994956169723818e-05, "loss": 0.0284, "step": 33080 }, { "grad_norm": 0.2328636646270752, "learning_rate": 7.993631980800285e-05, "loss": 0.0233, "step": 33090 }, { "grad_norm": 0.2309371680021286, "learning_rate": 7.992307464497659e-05, "loss": 0.0246, "step": 33100 }, { "grad_norm": 0.2263847142457962, "learning_rate": 7.990982620960787e-05, "loss": 0.0301, "step": 33110 }, { "grad_norm": 0.20153114199638367, "learning_rate": 7.989657450334554e-05, "loss": 0.027, "step": 33120 }, { "grad_norm": 0.3236373960971832, "learning_rate": 7.988331952763877e-05, "loss": 0.0326, "step": 33130 }, { "grad_norm": 0.2800212800502777, "learning_rate": 7.987006128393709e-05, "loss": 0.0295, "step": 33140 }, { "grad_norm": 0.20177233219146729, "learning_rate": 7.985679977369043e-05, "loss": 0.0297, "step": 33150 }, { "grad_norm": 0.31112128496170044, "learning_rate": 7.984353499834902e-05, "loss": 0.026, "step": 33160 }, { "grad_norm": 0.20694665610790253, "learning_rate": 7.983026695936351e-05, "loss": 0.0254, "step": 33170 }, { "grad_norm": 0.2553165853023529, "learning_rate": 7.981699565818486e-05, "loss": 0.0267, "step": 33180 }, { "grad_norm": 0.23908618092536926, "learning_rate": 7.980372109626437e-05, "loss": 0.0261, "step": 33190 }, { "grad_norm": 0.271993488073349, "learning_rate": 7.979044327505375e-05, "loss": 0.025, "step": 33200 }, { "grad_norm": 0.36431559920310974, "learning_rate": 7.977716219600506e-05, "loss": 0.0285, "step": 33210 }, { "grad_norm": 0.27114859223365784, "learning_rate": 7.97638778605707e-05, "loss": 0.0258, "step": 33220 }, { "grad_norm": 0.27955782413482666, "learning_rate": 7.975059027020338e-05, "loss": 0.0297, "step": 33230 }, { "grad_norm": 0.20928582549095154, "learning_rate": 7.973729942635623e-05, "loss": 0.0255, "step": 33240 }, { "grad_norm": 0.2922416925430298, "learning_rate": 7.972400533048273e-05, "loss": 0.0312, "step": 33250 }, { "grad_norm": 0.21562404930591583, "learning_rate": 7.97107079840367e-05, "loss": 0.0291, "step": 33260 }, { "grad_norm": 0.2638216018676758, "learning_rate": 7.969740738847231e-05, "loss": 0.0293, "step": 33270 }, { "grad_norm": 0.34979575872421265, "learning_rate": 7.968410354524411e-05, "loss": 0.0254, "step": 33280 }, { "grad_norm": 0.3157016336917877, "learning_rate": 7.967079645580697e-05, "loss": 0.0318, "step": 33290 }, { "grad_norm": 0.29292070865631104, "learning_rate": 7.965748612161612e-05, "loss": 0.038, "step": 33300 }, { "grad_norm": 0.25842365622520447, "learning_rate": 7.96441725441272e-05, "loss": 0.0267, "step": 33310 }, { "grad_norm": 0.25666993856430054, "learning_rate": 7.963085572479614e-05, "loss": 0.0271, "step": 33320 }, { "grad_norm": 0.24066424369812012, "learning_rate": 7.961753566507924e-05, "loss": 0.0266, "step": 33330 }, { "grad_norm": 0.2373461127281189, "learning_rate": 7.960421236643316e-05, "loss": 0.0378, "step": 33340 }, { "grad_norm": 0.24380438029766083, "learning_rate": 7.959088583031496e-05, "loss": 0.0288, "step": 33350 }, { "grad_norm": 0.2857665419578552, "learning_rate": 7.957755605818194e-05, "loss": 0.0287, "step": 33360 }, { "grad_norm": 0.2308880239725113, "learning_rate": 7.956422305149185e-05, "loss": 0.0258, "step": 33370 }, { "grad_norm": 0.22794654965400696, "learning_rate": 7.95508868117028e-05, "loss": 0.0306, "step": 33380 }, { "grad_norm": 0.24368639290332794, "learning_rate": 7.953754734027318e-05, "loss": 0.0249, "step": 33390 }, { "grad_norm": 0.2527843415737152, "learning_rate": 7.952420463866182e-05, "loss": 0.0351, "step": 33400 }, { "grad_norm": 0.1807025521993637, "learning_rate": 7.951085870832782e-05, "loss": 0.0274, "step": 33410 }, { "grad_norm": 0.23329326510429382, "learning_rate": 7.949750955073067e-05, "loss": 0.0281, "step": 33420 }, { "grad_norm": 0.19256019592285156, "learning_rate": 7.948415716733022e-05, "loss": 0.0291, "step": 33430 }, { "grad_norm": 0.3843322694301605, "learning_rate": 7.947080155958669e-05, "loss": 0.0247, "step": 33440 }, { "grad_norm": 0.21318434178829193, "learning_rate": 7.94574427289606e-05, "loss": 0.0235, "step": 33450 }, { "grad_norm": 0.34074118733406067, "learning_rate": 7.944408067691284e-05, "loss": 0.0308, "step": 33460 }, { "grad_norm": 0.33339428901672363, "learning_rate": 7.943071540490473e-05, "loss": 0.0325, "step": 33470 }, { "grad_norm": 0.46061915159225464, "learning_rate": 7.94173469143978e-05, "loss": 0.0336, "step": 33480 }, { "grad_norm": 0.2591339945793152, "learning_rate": 7.940397520685406e-05, "loss": 0.033, "step": 33490 }, { "grad_norm": 0.25509190559387207, "learning_rate": 7.939060028373577e-05, "loss": 0.0303, "step": 33500 }, { "grad_norm": 0.2162807285785675, "learning_rate": 7.937722214650565e-05, "loss": 0.0291, "step": 33510 }, { "grad_norm": 0.28744420409202576, "learning_rate": 7.936384079662666e-05, "loss": 0.0295, "step": 33520 }, { "grad_norm": 0.27625522017478943, "learning_rate": 7.93504562355622e-05, "loss": 0.0297, "step": 33530 }, { "grad_norm": 0.3043854534626007, "learning_rate": 7.933706846477599e-05, "loss": 0.0339, "step": 33540 }, { "grad_norm": 0.2651287317276001, "learning_rate": 7.932367748573206e-05, "loss": 0.0313, "step": 33550 }, { "grad_norm": 0.24883030354976654, "learning_rate": 7.931028329989485e-05, "loss": 0.0285, "step": 33560 }, { "grad_norm": 0.2204706072807312, "learning_rate": 7.929688590872913e-05, "loss": 0.0282, "step": 33570 }, { "grad_norm": 0.21148613095283508, "learning_rate": 7.928348531370003e-05, "loss": 0.0266, "step": 33580 }, { "grad_norm": 0.23137904703617096, "learning_rate": 7.927008151627297e-05, "loss": 0.0252, "step": 33590 }, { "grad_norm": 0.2630447745323181, "learning_rate": 7.925667451791383e-05, "loss": 0.0282, "step": 33600 }, { "grad_norm": 0.21269811689853668, "learning_rate": 7.924326432008874e-05, "loss": 0.0252, "step": 33610 }, { "grad_norm": 0.16680899262428284, "learning_rate": 7.922985092426422e-05, "loss": 0.0229, "step": 33620 }, { "grad_norm": 0.2511916160583496, "learning_rate": 7.921643433190717e-05, "loss": 0.0271, "step": 33630 }, { "grad_norm": 0.2151658535003662, "learning_rate": 7.920301454448478e-05, "loss": 0.0255, "step": 33640 }, { "grad_norm": 0.1739290952682495, "learning_rate": 7.918959156346461e-05, "loss": 0.0238, "step": 33650 }, { "grad_norm": 0.25084781646728516, "learning_rate": 7.91761653903146e-05, "loss": 0.0291, "step": 33660 }, { "grad_norm": 0.6767206192016602, "learning_rate": 7.916273602650302e-05, "loss": 0.0279, "step": 33670 }, { "grad_norm": 0.23763932287693024, "learning_rate": 7.914930347349847e-05, "loss": 0.0258, "step": 33680 }, { "grad_norm": 0.25615039467811584, "learning_rate": 7.913586773276992e-05, "loss": 0.0308, "step": 33690 }, { "grad_norm": 0.2148589789867401, "learning_rate": 7.912242880578667e-05, "loss": 0.0279, "step": 33700 }, { "grad_norm": 0.2560988664627075, "learning_rate": 7.910898669401839e-05, "loss": 0.0286, "step": 33710 }, { "grad_norm": 0.2897895872592926, "learning_rate": 7.909554139893511e-05, "loss": 0.0268, "step": 33720 }, { "grad_norm": 0.25734037160873413, "learning_rate": 7.908209292200715e-05, "loss": 0.0286, "step": 33730 }, { "grad_norm": 0.38654032349586487, "learning_rate": 7.906864126470523e-05, "loss": 0.0289, "step": 33740 }, { "grad_norm": 0.24916306138038635, "learning_rate": 7.905518642850041e-05, "loss": 0.0274, "step": 33750 }, { "grad_norm": 0.22217629849910736, "learning_rate": 7.904172841486409e-05, "loss": 0.026, "step": 33760 }, { "grad_norm": 0.19826363027095795, "learning_rate": 7.902826722526801e-05, "loss": 0.0288, "step": 33770 }, { "grad_norm": 0.16834834218025208, "learning_rate": 7.901480286118427e-05, "loss": 0.0205, "step": 33780 }, { "grad_norm": 0.16515906155109406, "learning_rate": 7.900133532408531e-05, "loss": 0.0254, "step": 33790 }, { "grad_norm": 0.2304764688014984, "learning_rate": 7.898786461544395e-05, "loss": 0.0299, "step": 33800 }, { "grad_norm": 0.2007022351026535, "learning_rate": 7.897439073673325e-05, "loss": 0.0274, "step": 33810 }, { "grad_norm": 0.23696592450141907, "learning_rate": 7.896091368942677e-05, "loss": 0.0273, "step": 33820 }, { "grad_norm": 0.23675020039081573, "learning_rate": 7.894743347499832e-05, "loss": 0.0298, "step": 33830 }, { "grad_norm": 0.26252660155296326, "learning_rate": 7.893395009492203e-05, "loss": 0.0302, "step": 33840 }, { "grad_norm": 0.20742058753967285, "learning_rate": 7.892046355067248e-05, "loss": 0.0278, "step": 33850 }, { "grad_norm": 0.2686765193939209, "learning_rate": 7.890697384372451e-05, "loss": 0.0261, "step": 33860 }, { "grad_norm": 0.20218557119369507, "learning_rate": 7.889348097555336e-05, "loss": 0.024, "step": 33870 }, { "grad_norm": 0.2374296486377716, "learning_rate": 7.887998494763455e-05, "loss": 0.0253, "step": 33880 }, { "grad_norm": 0.21702411770820618, "learning_rate": 7.886648576144404e-05, "loss": 0.0288, "step": 33890 }, { "grad_norm": 0.17737337946891785, "learning_rate": 7.885298341845802e-05, "loss": 0.0278, "step": 33900 }, { "grad_norm": 0.19763332605361938, "learning_rate": 7.883947792015311e-05, "loss": 0.0249, "step": 33910 }, { "grad_norm": 0.2146986573934555, "learning_rate": 7.882596926800628e-05, "loss": 0.0308, "step": 33920 }, { "grad_norm": 0.26030945777893066, "learning_rate": 7.881245746349477e-05, "loss": 0.0286, "step": 33930 }, { "grad_norm": 0.2758169174194336, "learning_rate": 7.879894250809623e-05, "loss": 0.0292, "step": 33940 }, { "grad_norm": 0.28223395347595215, "learning_rate": 7.878542440328865e-05, "loss": 0.0266, "step": 33950 }, { "grad_norm": 0.207730233669281, "learning_rate": 7.877190315055031e-05, "loss": 0.026, "step": 33960 }, { "grad_norm": 0.2836690843105316, "learning_rate": 7.875837875135991e-05, "loss": 0.0295, "step": 33970 }, { "grad_norm": 0.2967776954174042, "learning_rate": 7.874485120719646e-05, "loss": 0.0279, "step": 33980 }, { "grad_norm": 0.2537866532802582, "learning_rate": 7.873132051953928e-05, "loss": 0.0266, "step": 33990 }, { "grad_norm": 0.23076680302619934, "learning_rate": 7.87177866898681e-05, "loss": 0.0262, "step": 34000 }, { "grad_norm": 0.2714769244194031, "learning_rate": 7.870424971966294e-05, "loss": 0.0316, "step": 34010 }, { "grad_norm": 0.24499087035655975, "learning_rate": 7.869070961040419e-05, "loss": 0.03, "step": 34020 }, { "grad_norm": 0.1976386159658432, "learning_rate": 7.867716636357257e-05, "loss": 0.0272, "step": 34030 }, { "grad_norm": 0.2080833464860916, "learning_rate": 7.866361998064915e-05, "loss": 0.027, "step": 34040 }, { "grad_norm": 0.25554993748664856, "learning_rate": 7.865007046311534e-05, "loss": 0.0294, "step": 34050 }, { "grad_norm": 0.2147407829761505, "learning_rate": 7.86365178124529e-05, "loss": 0.0268, "step": 34060 }, { "grad_norm": 0.37169817090034485, "learning_rate": 7.862296203014394e-05, "loss": 0.0294, "step": 34070 }, { "grad_norm": 0.27797731757164, "learning_rate": 7.860940311767088e-05, "loss": 0.03, "step": 34080 }, { "grad_norm": 0.22982358932495117, "learning_rate": 7.85958410765165e-05, "loss": 0.0295, "step": 34090 }, { "grad_norm": 0.2216164767742157, "learning_rate": 7.858227590816394e-05, "loss": 0.0269, "step": 34100 }, { "grad_norm": 0.16250483691692352, "learning_rate": 7.856870761409664e-05, "loss": 0.0275, "step": 34110 }, { "grad_norm": 0.2688240110874176, "learning_rate": 7.855513619579846e-05, "loss": 0.0318, "step": 34120 }, { "grad_norm": 0.194392591714859, "learning_rate": 7.85415616547535e-05, "loss": 0.0248, "step": 34130 }, { "grad_norm": 0.2273523509502411, "learning_rate": 7.852798399244627e-05, "loss": 0.0287, "step": 34140 }, { "grad_norm": 0.194852814078331, "learning_rate": 7.851440321036161e-05, "loss": 0.0227, "step": 34150 }, { "grad_norm": 0.2979293465614319, "learning_rate": 7.850081930998468e-05, "loss": 0.0258, "step": 34160 }, { "grad_norm": 0.25937801599502563, "learning_rate": 7.8487232292801e-05, "loss": 0.0281, "step": 34170 }, { "grad_norm": 0.26445919275283813, "learning_rate": 7.847364216029642e-05, "loss": 0.0272, "step": 34180 }, { "grad_norm": 0.2196597456932068, "learning_rate": 7.846004891395716e-05, "loss": 0.0217, "step": 34190 }, { "grad_norm": 0.27265676856040955, "learning_rate": 7.844645255526972e-05, "loss": 0.0262, "step": 34200 }, { "grad_norm": 0.2567417025566101, "learning_rate": 7.843285308572101e-05, "loss": 0.0248, "step": 34210 }, { "grad_norm": 0.3181752562522888, "learning_rate": 7.841925050679823e-05, "loss": 0.0257, "step": 34220 }, { "grad_norm": 0.21172140538692474, "learning_rate": 7.840564481998895e-05, "loss": 0.0303, "step": 34230 }, { "grad_norm": 0.22317905724048615, "learning_rate": 7.839203602678105e-05, "loss": 0.0274, "step": 34240 }, { "grad_norm": 0.19183599948883057, "learning_rate": 7.837842412866279e-05, "loss": 0.0261, "step": 34250 }, { "grad_norm": 0.21135196089744568, "learning_rate": 7.836480912712272e-05, "loss": 0.0247, "step": 34260 }, { "grad_norm": 0.24912074208259583, "learning_rate": 7.835119102364976e-05, "loss": 0.0294, "step": 34270 }, { "grad_norm": 0.25734081864356995, "learning_rate": 7.83375698197332e-05, "loss": 0.0268, "step": 34280 }, { "grad_norm": 0.25936540961265564, "learning_rate": 7.83239455168626e-05, "loss": 0.0266, "step": 34290 }, { "grad_norm": 0.23616763949394226, "learning_rate": 7.83103181165279e-05, "loss": 0.0322, "step": 34300 }, { "grad_norm": 0.19396190345287323, "learning_rate": 7.829668762021937e-05, "loss": 0.0311, "step": 34310 }, { "grad_norm": 0.21419599652290344, "learning_rate": 7.828305402942764e-05, "loss": 0.0253, "step": 34320 }, { "grad_norm": 0.20969024300575256, "learning_rate": 7.826941734564363e-05, "loss": 0.0238, "step": 34330 }, { "grad_norm": 0.28929638862609863, "learning_rate": 7.825577757035865e-05, "loss": 0.0273, "step": 34340 }, { "grad_norm": 0.2590159475803375, "learning_rate": 7.824213470506431e-05, "loss": 0.0264, "step": 34350 }, { "grad_norm": 0.2591547667980194, "learning_rate": 7.822848875125257e-05, "loss": 0.0262, "step": 34360 }, { "grad_norm": 0.32075539231300354, "learning_rate": 7.821483971041576e-05, "loss": 0.0275, "step": 34370 }, { "grad_norm": 0.21975553035736084, "learning_rate": 7.820118758404649e-05, "loss": 0.0292, "step": 34380 }, { "grad_norm": 0.21659481525421143, "learning_rate": 7.818753237363776e-05, "loss": 0.0291, "step": 34390 }, { "grad_norm": 0.19605298340320587, "learning_rate": 7.817387408068286e-05, "loss": 0.0235, "step": 34400 }, { "grad_norm": 0.21932603418827057, "learning_rate": 7.816021270667544e-05, "loss": 0.0247, "step": 34410 }, { "grad_norm": 0.22822913527488708, "learning_rate": 7.81465482531095e-05, "loss": 0.0232, "step": 34420 }, { "grad_norm": 0.23647785186767578, "learning_rate": 7.813288072147938e-05, "loss": 0.0247, "step": 34430 }, { "grad_norm": 0.27691471576690674, "learning_rate": 7.811921011327972e-05, "loss": 0.0251, "step": 34440 }, { "grad_norm": 0.2412179857492447, "learning_rate": 7.810553643000549e-05, "loss": 0.0261, "step": 34450 }, { "grad_norm": 0.263097882270813, "learning_rate": 7.809185967315206e-05, "loss": 0.03, "step": 34460 }, { "grad_norm": 0.2508170008659363, "learning_rate": 7.80781798442151e-05, "loss": 0.0316, "step": 34470 }, { "grad_norm": 0.31000131368637085, "learning_rate": 7.806449694469061e-05, "loss": 0.0356, "step": 34480 }, { "grad_norm": 0.20993147790431976, "learning_rate": 7.805081097607492e-05, "loss": 0.0329, "step": 34490 }, { "grad_norm": 0.2614315152168274, "learning_rate": 7.803712193986474e-05, "loss": 0.0318, "step": 34500 }, { "grad_norm": 0.1694452464580536, "learning_rate": 7.802342983755702e-05, "loss": 0.0289, "step": 34510 }, { "grad_norm": 0.2202405482530594, "learning_rate": 7.800973467064918e-05, "loss": 0.0279, "step": 34520 }, { "grad_norm": 0.330742746591568, "learning_rate": 7.799603644063884e-05, "loss": 0.0258, "step": 34530 }, { "grad_norm": 0.23683884739875793, "learning_rate": 7.798233514902405e-05, "loss": 0.0239, "step": 34540 }, { "grad_norm": 0.23521938920021057, "learning_rate": 7.796863079730318e-05, "loss": 0.0299, "step": 34550 }, { "grad_norm": 0.2122577726840973, "learning_rate": 7.795492338697488e-05, "loss": 0.0241, "step": 34560 }, { "grad_norm": 0.21650445461273193, "learning_rate": 7.794121291953819e-05, "loss": 0.0239, "step": 34570 }, { "grad_norm": 0.20517493784427643, "learning_rate": 7.792749939649246e-05, "loss": 0.0233, "step": 34580 }, { "grad_norm": 0.25443652272224426, "learning_rate": 7.79137828193374e-05, "loss": 0.0276, "step": 34590 }, { "grad_norm": 0.29107943177223206, "learning_rate": 7.790006318957301e-05, "loss": 0.0315, "step": 34600 }, { "grad_norm": 0.2545715570449829, "learning_rate": 7.788634050869965e-05, "loss": 0.0235, "step": 34610 }, { "grad_norm": 0.2755877375602722, "learning_rate": 7.787261477821803e-05, "loss": 0.0304, "step": 34620 }, { "grad_norm": 0.21713845431804657, "learning_rate": 7.785888599962916e-05, "loss": 0.0251, "step": 34630 }, { "grad_norm": 0.2598937153816223, "learning_rate": 7.784515417443439e-05, "loss": 0.0232, "step": 34640 }, { "grad_norm": 0.27760687470436096, "learning_rate": 7.783141930413545e-05, "loss": 0.0318, "step": 34650 }, { "grad_norm": 0.3174797594547272, "learning_rate": 7.78176813902343e-05, "loss": 0.0271, "step": 34660 }, { "grad_norm": 0.31958431005477905, "learning_rate": 7.780394043423336e-05, "loss": 0.0313, "step": 34670 }, { "grad_norm": 0.25841647386550903, "learning_rate": 7.77901964376353e-05, "loss": 0.0246, "step": 34680 }, { "grad_norm": 0.8574975728988647, "learning_rate": 7.777644940194316e-05, "loss": 0.028, "step": 34690 }, { "grad_norm": 0.26516857743263245, "learning_rate": 7.776269932866023e-05, "loss": 0.0357, "step": 34700 }, { "grad_norm": 0.257937490940094, "learning_rate": 7.774894621929026e-05, "loss": 0.0329, "step": 34710 }, { "grad_norm": 0.3762343227863312, "learning_rate": 7.773519007533725e-05, "loss": 0.0277, "step": 34720 }, { "grad_norm": 0.21324048936367035, "learning_rate": 7.772143089830556e-05, "loss": 0.0282, "step": 34730 }, { "grad_norm": 0.19459328055381775, "learning_rate": 7.770766868969985e-05, "loss": 0.0261, "step": 34740 }, { "grad_norm": 0.21242362260818481, "learning_rate": 7.769390345102518e-05, "loss": 0.0276, "step": 34750 }, { "grad_norm": 0.25832638144493103, "learning_rate": 7.768013518378683e-05, "loss": 0.0252, "step": 34760 }, { "grad_norm": 0.23773612082004547, "learning_rate": 7.766636388949053e-05, "loss": 0.0304, "step": 34770 }, { "grad_norm": 0.27023205161094666, "learning_rate": 7.765258956964229e-05, "loss": 0.0276, "step": 34780 }, { "grad_norm": 0.24887602031230927, "learning_rate": 7.76388122257484e-05, "loss": 0.0256, "step": 34790 }, { "grad_norm": 0.25272467732429504, "learning_rate": 7.762503185931558e-05, "loss": 0.03, "step": 34800 }, { "grad_norm": 0.2356470674276352, "learning_rate": 7.76112484718508e-05, "loss": 0.0238, "step": 34810 }, { "grad_norm": 0.23400278389453888, "learning_rate": 7.75974620648614e-05, "loss": 0.0305, "step": 34820 }, { "grad_norm": 0.21220959722995758, "learning_rate": 7.758367263985503e-05, "loss": 0.0288, "step": 34830 }, { "grad_norm": 0.2372892200946808, "learning_rate": 7.75698801983397e-05, "loss": 0.0279, "step": 34840 }, { "grad_norm": 0.31919196248054504, "learning_rate": 7.755608474182372e-05, "loss": 0.0224, "step": 34850 }, { "grad_norm": 0.3071989119052887, "learning_rate": 7.754228627181574e-05, "loss": 0.0306, "step": 34860 }, { "grad_norm": 0.319399893283844, "learning_rate": 7.752848478982476e-05, "loss": 0.0327, "step": 34870 }, { "grad_norm": 0.30460309982299805, "learning_rate": 7.751468029736006e-05, "loss": 0.028, "step": 34880 }, { "grad_norm": 0.3220232427120209, "learning_rate": 7.750087279593129e-05, "loss": 0.0245, "step": 34890 }, { "grad_norm": 0.28496041893959045, "learning_rate": 7.748706228704843e-05, "loss": 0.034, "step": 34900 }, { "grad_norm": 0.25097090005874634, "learning_rate": 7.747324877222176e-05, "loss": 0.0272, "step": 34910 }, { "grad_norm": 0.21781864762306213, "learning_rate": 7.745943225296188e-05, "loss": 0.0256, "step": 34920 }, { "grad_norm": 0.22649389505386353, "learning_rate": 7.744561273077981e-05, "loss": 0.029, "step": 34930 }, { "grad_norm": 0.19417449831962585, "learning_rate": 7.743179020718678e-05, "loss": 0.0274, "step": 34940 }, { "grad_norm": 0.3562029302120209, "learning_rate": 7.741796468369443e-05, "loss": 0.029, "step": 34950 }, { "grad_norm": 0.2386690229177475, "learning_rate": 7.740413616181466e-05, "loss": 0.0296, "step": 34960 }, { "grad_norm": 0.27515485882759094, "learning_rate": 7.739030464305978e-05, "loss": 0.028, "step": 34970 }, { "grad_norm": 0.30014193058013916, "learning_rate": 7.737647012894235e-05, "loss": 0.0256, "step": 34980 }, { "grad_norm": 0.23422430455684662, "learning_rate": 7.736263262097532e-05, "loss": 0.0256, "step": 34990 }, { "grad_norm": 0.38922277092933655, "learning_rate": 7.734879212067192e-05, "loss": 0.0343, "step": 35000 }, { "grad_norm": 0.32521504163742065, "learning_rate": 7.733494862954573e-05, "loss": 0.0323, "step": 35010 }, { "grad_norm": 0.26015084981918335, "learning_rate": 7.732110214911066e-05, "loss": 0.0292, "step": 35020 }, { "grad_norm": 0.2739339768886566, "learning_rate": 7.730725268088094e-05, "loss": 0.0272, "step": 35030 }, { "grad_norm": 0.21258416771888733, "learning_rate": 7.729340022637112e-05, "loss": 0.029, "step": 35040 }, { "grad_norm": 0.23557019233703613, "learning_rate": 7.727954478709607e-05, "loss": 0.0296, "step": 35050 }, { "grad_norm": 0.3498714566230774, "learning_rate": 7.726568636457103e-05, "loss": 0.0285, "step": 35060 }, { "grad_norm": 0.18723709881305695, "learning_rate": 7.725182496031153e-05, "loss": 0.0257, "step": 35070 }, { "grad_norm": 0.2192571461200714, "learning_rate": 7.723796057583341e-05, "loss": 0.0286, "step": 35080 }, { "grad_norm": 0.3032819926738739, "learning_rate": 7.722409321265287e-05, "loss": 0.0273, "step": 35090 }, { "grad_norm": 0.22044186294078827, "learning_rate": 7.721022287228645e-05, "loss": 0.0294, "step": 35100 }, { "grad_norm": 0.24491570889949799, "learning_rate": 7.719634955625094e-05, "loss": 0.0306, "step": 35110 }, { "grad_norm": 0.2903697192668915, "learning_rate": 7.718247326606355e-05, "loss": 0.0282, "step": 35120 }, { "grad_norm": 0.3482487201690674, "learning_rate": 7.716859400324175e-05, "loss": 0.0291, "step": 35130 }, { "grad_norm": 0.24389421939849854, "learning_rate": 7.715471176930335e-05, "loss": 0.0242, "step": 35140 }, { "grad_norm": 0.35200434923171997, "learning_rate": 7.714082656576651e-05, "loss": 0.0248, "step": 35150 }, { "grad_norm": 0.21407335996627808, "learning_rate": 7.712693839414968e-05, "loss": 0.0264, "step": 35160 }, { "grad_norm": 0.2870533764362335, "learning_rate": 7.711304725597164e-05, "loss": 0.029, "step": 35170 }, { "grad_norm": 0.2310180515050888, "learning_rate": 7.709915315275151e-05, "loss": 0.0287, "step": 35180 }, { "grad_norm": 0.22644881904125214, "learning_rate": 7.708525608600876e-05, "loss": 0.0248, "step": 35190 }, { "grad_norm": 0.2053668349981308, "learning_rate": 7.707135605726311e-05, "loss": 0.0263, "step": 35200 }, { "grad_norm": 0.26605451107025146, "learning_rate": 7.705745306803466e-05, "loss": 0.0282, "step": 35210 }, { "grad_norm": 0.24889366328716278, "learning_rate": 7.704354711984383e-05, "loss": 0.0291, "step": 35220 }, { "grad_norm": 0.2744556963443756, "learning_rate": 7.702963821421133e-05, "loss": 0.0234, "step": 35230 }, { "grad_norm": 0.23798055946826935, "learning_rate": 7.701572635265826e-05, "loss": 0.0281, "step": 35240 }, { "grad_norm": 0.27466288208961487, "learning_rate": 7.700181153670596e-05, "loss": 0.0262, "step": 35250 }, { "grad_norm": 0.18429747223854065, "learning_rate": 7.698789376787616e-05, "loss": 0.0281, "step": 35260 }, { "grad_norm": 0.3272179067134857, "learning_rate": 7.697397304769085e-05, "loss": 0.0297, "step": 35270 }, { "grad_norm": 0.26266399025917053, "learning_rate": 7.696004937767241e-05, "loss": 0.028, "step": 35280 }, { "grad_norm": 0.18996292352676392, "learning_rate": 7.694612275934352e-05, "loss": 0.0244, "step": 35290 }, { "grad_norm": 0.273884654045105, "learning_rate": 7.693219319422714e-05, "loss": 0.0244, "step": 35300 }, { "grad_norm": 0.23375172913074493, "learning_rate": 7.69182606838466e-05, "loss": 0.0232, "step": 35310 }, { "grad_norm": 0.25969403982162476, "learning_rate": 7.690432522972558e-05, "loss": 0.0246, "step": 35320 }, { "grad_norm": 0.3117648661136627, "learning_rate": 7.689038683338796e-05, "loss": 0.027, "step": 35330 }, { "grad_norm": 0.1786770522594452, "learning_rate": 7.687644549635808e-05, "loss": 0.0203, "step": 35340 }, { "grad_norm": 0.2810441851615906, "learning_rate": 7.686250122016053e-05, "loss": 0.0277, "step": 35350 }, { "grad_norm": 0.24965834617614746, "learning_rate": 7.684855400632022e-05, "loss": 0.0284, "step": 35360 }, { "grad_norm": 0.21464814245700836, "learning_rate": 7.683460385636243e-05, "loss": 0.0223, "step": 35370 }, { "grad_norm": 0.19792594015598297, "learning_rate": 7.68206507718127e-05, "loss": 0.0267, "step": 35380 }, { "grad_norm": 0.2454625815153122, "learning_rate": 7.680669475419693e-05, "loss": 0.023, "step": 35390 }, { "grad_norm": 0.2307521551847458, "learning_rate": 7.679273580504132e-05, "loss": 0.026, "step": 35400 }, { "grad_norm": 0.2855593264102936, "learning_rate": 7.67787739258724e-05, "loss": 0.0226, "step": 35410 }, { "grad_norm": 0.21317461133003235, "learning_rate": 7.676480911821705e-05, "loss": 0.0269, "step": 35420 }, { "grad_norm": 0.20401114225387573, "learning_rate": 7.675084138360238e-05, "loss": 0.0269, "step": 35430 }, { "grad_norm": 0.22887888550758362, "learning_rate": 7.673687072355592e-05, "loss": 0.0328, "step": 35440 }, { "grad_norm": 0.21815915405750275, "learning_rate": 7.67228971396055e-05, "loss": 0.0283, "step": 35450 }, { "grad_norm": 0.20975445210933685, "learning_rate": 7.670892063327922e-05, "loss": 0.0263, "step": 35460 }, { "grad_norm": 0.20288780331611633, "learning_rate": 7.669494120610552e-05, "loss": 0.0324, "step": 35470 }, { "grad_norm": 0.30964556336402893, "learning_rate": 7.66809588596132e-05, "loss": 0.0283, "step": 35480 }, { "grad_norm": 0.22762681543827057, "learning_rate": 7.666697359533132e-05, "loss": 0.0254, "step": 35490 }, { "grad_norm": 0.21032030880451202, "learning_rate": 7.665298541478932e-05, "loss": 0.0229, "step": 35500 }, { "grad_norm": 0.2561022937297821, "learning_rate": 7.66389943195169e-05, "loss": 0.028, "step": 35510 }, { "grad_norm": 0.24228346347808838, "learning_rate": 7.66250003110441e-05, "loss": 0.0341, "step": 35520 }, { "grad_norm": 0.24339349567890167, "learning_rate": 7.66110033909013e-05, "loss": 0.0313, "step": 35530 }, { "grad_norm": 0.24402368068695068, "learning_rate": 7.659700356061918e-05, "loss": 0.0279, "step": 35540 }, { "grad_norm": 0.2744808793067932, "learning_rate": 7.658300082172875e-05, "loss": 0.0261, "step": 35550 }, { "grad_norm": 0.3018224835395813, "learning_rate": 7.65689951757613e-05, "loss": 0.0277, "step": 35560 }, { "grad_norm": 0.27783334255218506, "learning_rate": 7.655498662424849e-05, "loss": 0.0259, "step": 35570 }, { "grad_norm": 0.21789948642253876, "learning_rate": 7.654097516872227e-05, "loss": 0.03, "step": 35580 }, { "grad_norm": 0.29147306084632874, "learning_rate": 7.65269608107149e-05, "loss": 0.0245, "step": 35590 }, { "grad_norm": 0.23043370246887207, "learning_rate": 7.6512943551759e-05, "loss": 0.0264, "step": 35600 }, { "grad_norm": 0.24784991145133972, "learning_rate": 7.649892339338743e-05, "loss": 0.0282, "step": 35610 }, { "grad_norm": 0.3050554096698761, "learning_rate": 7.648490033713344e-05, "loss": 0.029, "step": 35620 }, { "grad_norm": 0.27866330742836, "learning_rate": 7.647087438453058e-05, "loss": 0.0315, "step": 35630 }, { "grad_norm": 0.23944605886936188, "learning_rate": 7.64568455371127e-05, "loss": 0.0257, "step": 35640 }, { "grad_norm": 0.3024308383464813, "learning_rate": 7.644281379641396e-05, "loss": 0.0247, "step": 35650 }, { "grad_norm": 0.24503985047340393, "learning_rate": 7.642877916396887e-05, "loss": 0.0242, "step": 35660 }, { "grad_norm": 0.22634465992450714, "learning_rate": 7.641474164131221e-05, "loss": 0.029, "step": 35670 }, { "grad_norm": 0.2902089059352875, "learning_rate": 7.640070122997913e-05, "loss": 0.0271, "step": 35680 }, { "grad_norm": 0.34345415234565735, "learning_rate": 7.638665793150505e-05, "loss": 0.0259, "step": 35690 }, { "grad_norm": 0.2419169843196869, "learning_rate": 7.637261174742574e-05, "loss": 0.0261, "step": 35700 }, { "grad_norm": 0.3451967239379883, "learning_rate": 7.635856267927725e-05, "loss": 0.0296, "step": 35710 }, { "grad_norm": 0.21642735600471497, "learning_rate": 7.634451072859597e-05, "loss": 0.0226, "step": 35720 }, { "grad_norm": 0.29903337359428406, "learning_rate": 7.633045589691863e-05, "loss": 0.0288, "step": 35730 }, { "grad_norm": 0.2258489429950714, "learning_rate": 7.63163981857822e-05, "loss": 0.0297, "step": 35740 }, { "grad_norm": 0.25214916467666626, "learning_rate": 7.630233759672403e-05, "loss": 0.0232, "step": 35750 }, { "grad_norm": 0.2579795718193054, "learning_rate": 7.628827413128177e-05, "loss": 0.0284, "step": 35760 }, { "grad_norm": 0.22698242962360382, "learning_rate": 7.627420779099338e-05, "loss": 0.026, "step": 35770 }, { "grad_norm": 0.3046051561832428, "learning_rate": 7.626013857739711e-05, "loss": 0.0316, "step": 35780 }, { "grad_norm": 0.24834491312503815, "learning_rate": 7.624606649203158e-05, "loss": 0.0264, "step": 35790 }, { "grad_norm": 0.24566510319709778, "learning_rate": 7.623199153643569e-05, "loss": 0.0263, "step": 35800 }, { "grad_norm": 0.26743483543395996, "learning_rate": 7.621791371214863e-05, "loss": 0.0267, "step": 35810 }, { "grad_norm": 0.22889527678489685, "learning_rate": 7.620383302070995e-05, "loss": 0.0282, "step": 35820 }, { "grad_norm": 0.21895672380924225, "learning_rate": 7.61897494636595e-05, "loss": 0.0288, "step": 35830 }, { "grad_norm": 0.217783123254776, "learning_rate": 7.617566304253739e-05, "loss": 0.0278, "step": 35840 }, { "grad_norm": 0.3724122643470764, "learning_rate": 7.616157375888416e-05, "loss": 0.0236, "step": 35850 }, { "grad_norm": 0.2382335513830185, "learning_rate": 7.614748161424053e-05, "loss": 0.0237, "step": 35860 }, { "grad_norm": 0.22220122814178467, "learning_rate": 7.613338661014763e-05, "loss": 0.0288, "step": 35870 }, { "grad_norm": 0.17549727857112885, "learning_rate": 7.611928874814686e-05, "loss": 0.0255, "step": 35880 }, { "grad_norm": 0.20035420358181, "learning_rate": 7.610518802977996e-05, "loss": 0.0238, "step": 35890 }, { "grad_norm": 0.2384677231311798, "learning_rate": 7.609108445658893e-05, "loss": 0.022, "step": 35900 }, { "grad_norm": 0.31721293926239014, "learning_rate": 7.607697803011612e-05, "loss": 0.0228, "step": 35910 }, { "grad_norm": 0.20406094193458557, "learning_rate": 7.606286875190421e-05, "loss": 0.0291, "step": 35920 }, { "grad_norm": 0.23533590137958527, "learning_rate": 7.604875662349617e-05, "loss": 0.0287, "step": 35930 }, { "grad_norm": 0.291970431804657, "learning_rate": 7.603464164643525e-05, "loss": 0.0286, "step": 35940 }, { "grad_norm": 0.24140848219394684, "learning_rate": 7.602052382226507e-05, "loss": 0.0274, "step": 35950 }, { "grad_norm": 0.25192856788635254, "learning_rate": 7.600640315252954e-05, "loss": 0.0222, "step": 35960 }, { "grad_norm": 0.22786743938922882, "learning_rate": 7.599227963877284e-05, "loss": 0.0216, "step": 35970 }, { "grad_norm": 0.25754520297050476, "learning_rate": 7.597815328253954e-05, "loss": 0.0311, "step": 35980 }, { "grad_norm": 0.31130918860435486, "learning_rate": 7.596402408537443e-05, "loss": 0.0288, "step": 35990 }, { "grad_norm": 0.19589349627494812, "learning_rate": 7.59498920488227e-05, "loss": 0.0248, "step": 36000 }, { "grad_norm": 0.30918657779693604, "learning_rate": 7.593575717442979e-05, "loss": 0.0296, "step": 36010 }, { "grad_norm": 0.2252449244260788, "learning_rate": 7.592161946374147e-05, "loss": 0.0252, "step": 36020 }, { "grad_norm": 0.1889026015996933, "learning_rate": 7.590747891830381e-05, "loss": 0.0257, "step": 36030 }, { "grad_norm": 0.20537802577018738, "learning_rate": 7.58933355396632e-05, "loss": 0.0235, "step": 36040 }, { "grad_norm": 0.21341969072818756, "learning_rate": 7.587918932936636e-05, "loss": 0.0232, "step": 36050 }, { "grad_norm": 0.23403844237327576, "learning_rate": 7.586504028896028e-05, "loss": 0.0258, "step": 36060 }, { "grad_norm": 1.178760290145874, "learning_rate": 7.585088841999228e-05, "loss": 0.0291, "step": 36070 }, { "grad_norm": 0.27926766872406006, "learning_rate": 7.583673372400999e-05, "loss": 0.0258, "step": 36080 }, { "grad_norm": 0.24679185450077057, "learning_rate": 7.582257620256134e-05, "loss": 0.0314, "step": 36090 }, { "grad_norm": 0.24093833565711975, "learning_rate": 7.580841585719458e-05, "loss": 0.0262, "step": 36100 }, { "grad_norm": 0.22798150777816772, "learning_rate": 7.579425268945825e-05, "loss": 0.0265, "step": 36110 }, { "grad_norm": 0.2396136075258255, "learning_rate": 7.578008670090127e-05, "loss": 0.0312, "step": 36120 }, { "grad_norm": 0.22451364994049072, "learning_rate": 7.576591789307272e-05, "loss": 0.0293, "step": 36130 }, { "grad_norm": 0.18026936054229736, "learning_rate": 7.575174626752216e-05, "loss": 0.029, "step": 36140 }, { "grad_norm": 0.45350953936576843, "learning_rate": 7.573757182579934e-05, "loss": 0.0275, "step": 36150 }, { "grad_norm": 0.200513556599617, "learning_rate": 7.572339456945435e-05, "loss": 0.0306, "step": 36160 }, { "grad_norm": 0.24940472841262817, "learning_rate": 7.570921450003762e-05, "loss": 0.0263, "step": 36170 }, { "grad_norm": 0.2102155089378357, "learning_rate": 7.569503161909983e-05, "loss": 0.0271, "step": 36180 }, { "grad_norm": 0.23927348852157593, "learning_rate": 7.568084592819202e-05, "loss": 0.0246, "step": 36190 }, { "grad_norm": 0.3500610888004303, "learning_rate": 7.566665742886551e-05, "loss": 0.0279, "step": 36200 }, { "grad_norm": 0.3009044826030731, "learning_rate": 7.565246612267194e-05, "loss": 0.0299, "step": 36210 }, { "grad_norm": 0.1895747184753418, "learning_rate": 7.563827201116325e-05, "loss": 0.025, "step": 36220 }, { "grad_norm": 0.2953811287879944, "learning_rate": 7.562407509589164e-05, "loss": 0.0305, "step": 36230 }, { "grad_norm": 0.2450917363166809, "learning_rate": 7.560987537840973e-05, "loss": 0.0315, "step": 36240 }, { "grad_norm": 0.35576319694519043, "learning_rate": 7.559567286027036e-05, "loss": 0.0285, "step": 36250 }, { "grad_norm": 0.2611991763114929, "learning_rate": 7.558146754302668e-05, "loss": 0.0251, "step": 36260 }, { "grad_norm": 0.2182353287935257, "learning_rate": 7.556725942823217e-05, "loss": 0.0237, "step": 36270 }, { "grad_norm": 0.25372380018234253, "learning_rate": 7.555304851744061e-05, "loss": 0.0232, "step": 36280 }, { "grad_norm": 0.2190633863210678, "learning_rate": 7.553883481220608e-05, "loss": 0.0242, "step": 36290 }, { "grad_norm": 0.18383197486400604, "learning_rate": 7.552461831408298e-05, "loss": 0.0254, "step": 36300 }, { "grad_norm": 0.27671876549720764, "learning_rate": 7.551039902462599e-05, "loss": 0.0277, "step": 36310 }, { "grad_norm": 0.24828478693962097, "learning_rate": 7.549617694539011e-05, "loss": 0.0269, "step": 36320 }, { "grad_norm": 0.22912541031837463, "learning_rate": 7.548195207793066e-05, "loss": 0.0287, "step": 36330 }, { "grad_norm": 0.2332303524017334, "learning_rate": 7.546772442380323e-05, "loss": 0.0236, "step": 36340 }, { "grad_norm": 0.19307269155979156, "learning_rate": 7.545349398456376e-05, "loss": 0.0262, "step": 36350 }, { "grad_norm": 0.2126353234052658, "learning_rate": 7.543926076176845e-05, "loss": 0.0253, "step": 36360 }, { "grad_norm": 0.1864239126443863, "learning_rate": 7.542502475697385e-05, "loss": 0.0279, "step": 36370 }, { "grad_norm": 0.16253359615802765, "learning_rate": 7.541078597173675e-05, "loss": 0.0269, "step": 36380 }, { "grad_norm": 0.2471880316734314, "learning_rate": 7.539654440761431e-05, "loss": 0.0247, "step": 36390 }, { "grad_norm": 0.23364433646202087, "learning_rate": 7.538230006616395e-05, "loss": 0.0261, "step": 36400 }, { "grad_norm": 0.21784523129463196, "learning_rate": 7.536805294894342e-05, "loss": 0.0239, "step": 36410 }, { "grad_norm": 0.2554183304309845, "learning_rate": 7.535380305751076e-05, "loss": 0.026, "step": 36420 }, { "grad_norm": 0.3132549822330475, "learning_rate": 7.533955039342431e-05, "loss": 0.0286, "step": 36430 }, { "grad_norm": 0.2296724170446396, "learning_rate": 7.532529495824274e-05, "loss": 0.024, "step": 36440 }, { "grad_norm": 0.2303319126367569, "learning_rate": 7.531103675352497e-05, "loss": 0.0262, "step": 36450 }, { "grad_norm": 0.27360767126083374, "learning_rate": 7.529677578083029e-05, "loss": 0.028, "step": 36460 }, { "grad_norm": 0.25906902551651, "learning_rate": 7.528251204171823e-05, "loss": 0.0297, "step": 36470 }, { "grad_norm": 0.19342726469039917, "learning_rate": 7.526824553774867e-05, "loss": 0.0284, "step": 36480 }, { "grad_norm": 0.31305500864982605, "learning_rate": 7.525397627048178e-05, "loss": 0.0271, "step": 36490 }, { "grad_norm": 0.25737541913986206, "learning_rate": 7.523970424147802e-05, "loss": 0.024, "step": 36500 }, { "grad_norm": 0.3039933145046234, "learning_rate": 7.522542945229813e-05, "loss": 0.026, "step": 36510 }, { "grad_norm": 0.21435239911079407, "learning_rate": 7.521115190450321e-05, "loss": 0.0232, "step": 36520 }, { "grad_norm": 0.2050512135028839, "learning_rate": 7.519687159965462e-05, "loss": 0.033, "step": 36530 }, { "grad_norm": 0.24228724837303162, "learning_rate": 7.518258853931403e-05, "loss": 0.0244, "step": 36540 }, { "grad_norm": 0.2554648816585541, "learning_rate": 7.516830272504342e-05, "loss": 0.0224, "step": 36550 }, { "grad_norm": 0.25173458456993103, "learning_rate": 7.515401415840509e-05, "loss": 0.0266, "step": 36560 }, { "grad_norm": 0.21764957904815674, "learning_rate": 7.513972284096155e-05, "loss": 0.0256, "step": 36570 }, { "grad_norm": 0.2661465108394623, "learning_rate": 7.512542877427574e-05, "loss": 0.0248, "step": 36580 }, { "grad_norm": 0.25823837518692017, "learning_rate": 7.51111319599108e-05, "loss": 0.0222, "step": 36590 }, { "grad_norm": 0.22337111830711365, "learning_rate": 7.509683239943024e-05, "loss": 0.0211, "step": 36600 }, { "grad_norm": 0.33880695700645447, "learning_rate": 7.50825300943978e-05, "loss": 0.032, "step": 36610 }, { "grad_norm": 0.24666692316532135, "learning_rate": 7.50682250463776e-05, "loss": 0.0223, "step": 36620 }, { "grad_norm": 0.22134970128536224, "learning_rate": 7.5053917256934e-05, "loss": 0.0237, "step": 36630 }, { "grad_norm": 0.20531541109085083, "learning_rate": 7.503960672763166e-05, "loss": 0.025, "step": 36640 }, { "grad_norm": 0.25631487369537354, "learning_rate": 7.502529346003559e-05, "loss": 0.0272, "step": 36650 }, { "grad_norm": 0.2605782151222229, "learning_rate": 7.501097745571106e-05, "loss": 0.031, "step": 36660 }, { "grad_norm": 0.2007080316543579, "learning_rate": 7.499665871622361e-05, "loss": 0.0248, "step": 36670 }, { "grad_norm": 0.2540506422519684, "learning_rate": 7.498233724313917e-05, "loss": 0.0255, "step": 36680 }, { "grad_norm": 0.20513233542442322, "learning_rate": 7.496801303802387e-05, "loss": 0.027, "step": 36690 }, { "grad_norm": 0.20728614926338196, "learning_rate": 7.495368610244423e-05, "loss": 0.0213, "step": 36700 }, { "grad_norm": 0.19963352382183075, "learning_rate": 7.493935643796697e-05, "loss": 0.0225, "step": 36710 }, { "grad_norm": 0.21754205226898193, "learning_rate": 7.49250240461592e-05, "loss": 0.0249, "step": 36720 }, { "grad_norm": 0.20044635236263275, "learning_rate": 7.491068892858826e-05, "loss": 0.0215, "step": 36730 }, { "grad_norm": 0.2183687835931778, "learning_rate": 7.489635108682184e-05, "loss": 0.0255, "step": 36740 }, { "grad_norm": 0.28818196058273315, "learning_rate": 7.48820105224279e-05, "loss": 0.024, "step": 36750 }, { "grad_norm": 0.2069648802280426, "learning_rate": 7.48676672369747e-05, "loss": 0.0298, "step": 36760 }, { "grad_norm": 0.19007688760757446, "learning_rate": 7.485332123203079e-05, "loss": 0.0215, "step": 36770 }, { "grad_norm": 0.23999367654323578, "learning_rate": 7.483897250916504e-05, "loss": 0.0256, "step": 36780 }, { "grad_norm": 0.20924481749534607, "learning_rate": 7.48246210699466e-05, "loss": 0.0242, "step": 36790 }, { "grad_norm": 0.3374742269515991, "learning_rate": 7.481026691594492e-05, "loss": 0.0281, "step": 36800 }, { "grad_norm": 0.2162267416715622, "learning_rate": 7.479591004872975e-05, "loss": 0.0228, "step": 36810 }, { "grad_norm": 0.3069724440574646, "learning_rate": 7.478155046987114e-05, "loss": 0.0229, "step": 36820 }, { "grad_norm": 0.20270104706287384, "learning_rate": 7.476718818093944e-05, "loss": 0.0213, "step": 36830 }, { "grad_norm": 0.21251240372657776, "learning_rate": 7.475282318350527e-05, "loss": 0.024, "step": 36840 }, { "grad_norm": 0.23002606630325317, "learning_rate": 7.473845547913957e-05, "loss": 0.0216, "step": 36850 }, { "grad_norm": 0.23832380771636963, "learning_rate": 7.472408506941357e-05, "loss": 0.0277, "step": 36860 }, { "grad_norm": 0.2045930027961731, "learning_rate": 7.470971195589881e-05, "loss": 0.026, "step": 36870 }, { "grad_norm": 0.21554790437221527, "learning_rate": 7.46953361401671e-05, "loss": 0.0237, "step": 36880 }, { "grad_norm": 0.27910953760147095, "learning_rate": 7.468095762379055e-05, "loss": 0.0266, "step": 36890 }, { "grad_norm": 0.20272673666477203, "learning_rate": 7.466657640834158e-05, "loss": 0.0267, "step": 36900 }, { "grad_norm": 0.339559406042099, "learning_rate": 7.46521924953929e-05, "loss": 0.0228, "step": 36910 }, { "grad_norm": 0.2373640090227127, "learning_rate": 7.463780588651752e-05, "loss": 0.0246, "step": 36920 }, { "grad_norm": 0.21721279621124268, "learning_rate": 7.462341658328873e-05, "loss": 0.0261, "step": 36930 }, { "grad_norm": 0.25911015272140503, "learning_rate": 7.460902458728012e-05, "loss": 0.0256, "step": 36940 }, { "grad_norm": 0.26735007762908936, "learning_rate": 7.459462990006558e-05, "loss": 0.0254, "step": 36950 }, { "grad_norm": 0.1754174828529358, "learning_rate": 7.458023252321929e-05, "loss": 0.0244, "step": 36960 }, { "grad_norm": 0.2191867232322693, "learning_rate": 7.456583245831573e-05, "loss": 0.0262, "step": 36970 }, { "grad_norm": 0.2784283757209778, "learning_rate": 7.455142970692968e-05, "loss": 0.0298, "step": 36980 }, { "grad_norm": 0.32682180404663086, "learning_rate": 7.453702427063615e-05, "loss": 0.0245, "step": 36990 }, { "grad_norm": 0.2839975655078888, "learning_rate": 7.452261615101057e-05, "loss": 0.0287, "step": 37000 }, { "grad_norm": 0.19826067984104156, "learning_rate": 7.450820534962856e-05, "loss": 0.029, "step": 37010 }, { "grad_norm": 0.18943431973457336, "learning_rate": 7.449379186806604e-05, "loss": 0.0264, "step": 37020 }, { "grad_norm": 0.2661258280277252, "learning_rate": 7.44793757078993e-05, "loss": 0.0274, "step": 37030 }, { "grad_norm": 0.18837833404541016, "learning_rate": 7.44649568707048e-05, "loss": 0.0261, "step": 37040 }, { "grad_norm": 0.253152996301651, "learning_rate": 7.445053535805942e-05, "loss": 0.026, "step": 37050 }, { "grad_norm": 0.3042517602443695, "learning_rate": 7.443611117154024e-05, "loss": 0.0263, "step": 37060 }, { "grad_norm": 0.20763427019119263, "learning_rate": 7.442168431272469e-05, "loss": 0.0259, "step": 37070 }, { "grad_norm": 0.21372227370738983, "learning_rate": 7.440725478319044e-05, "loss": 0.0286, "step": 37080 }, { "grad_norm": 0.22405996918678284, "learning_rate": 7.439282258451553e-05, "loss": 0.0263, "step": 37090 }, { "grad_norm": 0.23710671067237854, "learning_rate": 7.43783877182782e-05, "loss": 0.0267, "step": 37100 }, { "grad_norm": 0.2182334065437317, "learning_rate": 7.436395018605707e-05, "loss": 0.0231, "step": 37110 }, { "grad_norm": 0.22261501848697662, "learning_rate": 7.434950998943094e-05, "loss": 0.0268, "step": 37120 }, { "grad_norm": 0.2197270691394806, "learning_rate": 7.433506712997902e-05, "loss": 0.0218, "step": 37130 }, { "grad_norm": 0.25053903460502625, "learning_rate": 7.432062160928077e-05, "loss": 0.0243, "step": 37140 }, { "grad_norm": 0.31256362795829773, "learning_rate": 7.430617342891588e-05, "loss": 0.0245, "step": 37150 }, { "grad_norm": 0.1893065720796585, "learning_rate": 7.429172259046442e-05, "loss": 0.0268, "step": 37160 }, { "grad_norm": 0.3294454514980316, "learning_rate": 7.42772690955067e-05, "loss": 0.0253, "step": 37170 }, { "grad_norm": 0.268375962972641, "learning_rate": 7.426281294562332e-05, "loss": 0.0293, "step": 37180 }, { "grad_norm": 0.21306076645851135, "learning_rate": 7.424835414239519e-05, "loss": 0.0246, "step": 37190 }, { "grad_norm": 0.2940666973590851, "learning_rate": 7.423389268740352e-05, "loss": 0.0266, "step": 37200 }, { "grad_norm": 0.28272324800491333, "learning_rate": 7.42194285822298e-05, "loss": 0.0299, "step": 37210 }, { "grad_norm": 0.29922667145729065, "learning_rate": 7.420496182845576e-05, "loss": 0.0277, "step": 37220 }, { "grad_norm": 0.24327288568019867, "learning_rate": 7.419049242766348e-05, "loss": 0.0305, "step": 37230 }, { "grad_norm": 0.28319990634918213, "learning_rate": 7.417602038143533e-05, "loss": 0.0266, "step": 37240 }, { "grad_norm": 0.2610841989517212, "learning_rate": 7.416154569135393e-05, "loss": 0.0303, "step": 37250 }, { "grad_norm": 0.18849590420722961, "learning_rate": 7.414706835900222e-05, "loss": 0.0271, "step": 37260 }, { "grad_norm": 0.21128545701503754, "learning_rate": 7.413258838596345e-05, "loss": 0.0264, "step": 37270 }, { "grad_norm": 0.19896627962589264, "learning_rate": 7.411810577382106e-05, "loss": 0.0271, "step": 37280 }, { "grad_norm": 0.20909163355827332, "learning_rate": 7.410362052415889e-05, "loss": 0.0259, "step": 37290 }, { "grad_norm": 0.26291346549987793, "learning_rate": 7.408913263856102e-05, "loss": 0.0249, "step": 37300 }, { "grad_norm": 0.3060694932937622, "learning_rate": 7.407464211861183e-05, "loss": 0.0276, "step": 37310 }, { "grad_norm": 0.20368492603302002, "learning_rate": 7.406014896589597e-05, "loss": 0.0245, "step": 37320 }, { "grad_norm": 0.30379942059516907, "learning_rate": 7.404565318199838e-05, "loss": 0.0233, "step": 37330 }, { "grad_norm": 0.29970479011535645, "learning_rate": 7.403115476850432e-05, "loss": 0.025, "step": 37340 }, { "grad_norm": 0.2769697308540344, "learning_rate": 7.401665372699932e-05, "loss": 0.0257, "step": 37350 }, { "grad_norm": 0.2527448534965515, "learning_rate": 7.400215005906916e-05, "loss": 0.0306, "step": 37360 }, { "grad_norm": 0.32024210691452026, "learning_rate": 7.398764376629996e-05, "loss": 0.0289, "step": 37370 }, { "grad_norm": 0.22666975855827332, "learning_rate": 7.39731348502781e-05, "loss": 0.0217, "step": 37380 }, { "grad_norm": 0.24124088883399963, "learning_rate": 7.395862331259028e-05, "loss": 0.0252, "step": 37390 }, { "grad_norm": 0.16029179096221924, "learning_rate": 7.394410915482341e-05, "loss": 0.0227, "step": 37400 }, { "grad_norm": 0.2186984121799469, "learning_rate": 7.392959237856476e-05, "loss": 0.0249, "step": 37410 }, { "grad_norm": 0.2573776841163635, "learning_rate": 7.391507298540187e-05, "loss": 0.0228, "step": 37420 }, { "grad_norm": 0.2510468363761902, "learning_rate": 7.390055097692257e-05, "loss": 0.0255, "step": 37430 }, { "grad_norm": 0.23507051169872284, "learning_rate": 7.388602635471494e-05, "loss": 0.0242, "step": 37440 }, { "grad_norm": 0.18208301067352295, "learning_rate": 7.387149912036738e-05, "loss": 0.0247, "step": 37450 }, { "grad_norm": 0.27953004837036133, "learning_rate": 7.385696927546854e-05, "loss": 0.0286, "step": 37460 }, { "grad_norm": 0.24649111926555634, "learning_rate": 7.384243682160743e-05, "loss": 0.0249, "step": 37470 }, { "grad_norm": 0.21475747227668762, "learning_rate": 7.382790176037329e-05, "loss": 0.0223, "step": 37480 }, { "grad_norm": 0.19992271065711975, "learning_rate": 7.381336409335562e-05, "loss": 0.0278, "step": 37490 }, { "grad_norm": 0.19537724554538727, "learning_rate": 7.379882382214426e-05, "loss": 0.025, "step": 37500 }, { "grad_norm": 0.21652965247631073, "learning_rate": 7.378428094832931e-05, "loss": 0.024, "step": 37510 }, { "grad_norm": 0.21403375267982483, "learning_rate": 7.376973547350114e-05, "loss": 0.0228, "step": 37520 }, { "grad_norm": 0.20713889598846436, "learning_rate": 7.375518739925046e-05, "loss": 0.0255, "step": 37530 }, { "grad_norm": 0.241330087184906, "learning_rate": 7.374063672716818e-05, "loss": 0.0227, "step": 37540 }, { "grad_norm": 0.24540024995803833, "learning_rate": 7.372608345884558e-05, "loss": 0.0258, "step": 37550 }, { "grad_norm": 0.20747292041778564, "learning_rate": 7.371152759587415e-05, "loss": 0.025, "step": 37560 }, { "grad_norm": 0.4114927053451538, "learning_rate": 7.369696913984572e-05, "loss": 0.0239, "step": 37570 }, { "grad_norm": 0.26829174160957336, "learning_rate": 7.368240809235237e-05, "loss": 0.0241, "step": 37580 }, { "grad_norm": 0.2178249955177307, "learning_rate": 7.366784445498649e-05, "loss": 0.0227, "step": 37590 }, { "grad_norm": 0.2637344002723694, "learning_rate": 7.36532782293407e-05, "loss": 0.0315, "step": 37600 }, { "grad_norm": 0.29473578929901123, "learning_rate": 7.363870941700797e-05, "loss": 0.024, "step": 37610 }, { "grad_norm": 0.2589428722858429, "learning_rate": 7.362413801958152e-05, "loss": 0.024, "step": 37620 }, { "grad_norm": 0.20185332000255585, "learning_rate": 7.360956403865486e-05, "loss": 0.0292, "step": 37630 }, { "grad_norm": 0.28390786051750183, "learning_rate": 7.359498747582177e-05, "loss": 0.0311, "step": 37640 }, { "grad_norm": 0.22296202182769775, "learning_rate": 7.358040833267634e-05, "loss": 0.0272, "step": 37650 }, { "grad_norm": 0.22053393721580505, "learning_rate": 7.356582661081288e-05, "loss": 0.028, "step": 37660 }, { "grad_norm": 0.22758635878562927, "learning_rate": 7.355124231182607e-05, "loss": 0.0238, "step": 37670 }, { "grad_norm": 0.22310833632946014, "learning_rate": 7.353665543731083e-05, "loss": 0.0247, "step": 37680 }, { "grad_norm": 0.24433039128780365, "learning_rate": 7.352206598886229e-05, "loss": 0.0217, "step": 37690 }, { "grad_norm": 0.21831803023815155, "learning_rate": 7.350747396807601e-05, "loss": 0.0244, "step": 37700 }, { "grad_norm": 0.30981361865997314, "learning_rate": 7.349287937654772e-05, "loss": 0.0235, "step": 37710 }, { "grad_norm": 0.244099959731102, "learning_rate": 7.347828221587345e-05, "loss": 0.0248, "step": 37720 }, { "grad_norm": 0.2625199556350708, "learning_rate": 7.346368248764955e-05, "loss": 0.0224, "step": 37730 }, { "grad_norm": 0.3286174535751343, "learning_rate": 7.34490801934726e-05, "loss": 0.024, "step": 37740 }, { "grad_norm": 0.2188960164785385, "learning_rate": 7.343447533493947e-05, "loss": 0.0265, "step": 37750 }, { "grad_norm": 0.22441159188747406, "learning_rate": 7.341986791364739e-05, "loss": 0.0244, "step": 37760 }, { "grad_norm": 0.2538504898548126, "learning_rate": 7.340525793119374e-05, "loss": 0.0288, "step": 37770 }, { "grad_norm": 0.2487100511789322, "learning_rate": 7.33906453891763e-05, "loss": 0.025, "step": 37780 }, { "grad_norm": 0.24260947108268738, "learning_rate": 7.337603028919303e-05, "loss": 0.0263, "step": 37790 }, { "grad_norm": 0.23166346549987793, "learning_rate": 7.336141263284225e-05, "loss": 0.0254, "step": 37800 }, { "grad_norm": 0.34863153100013733, "learning_rate": 7.334679242172249e-05, "loss": 0.0228, "step": 37810 }, { "grad_norm": 0.2520638108253479, "learning_rate": 7.33321696574326e-05, "loss": 0.0265, "step": 37820 }, { "grad_norm": 0.1746188849210739, "learning_rate": 7.331754434157174e-05, "loss": 0.025, "step": 37830 }, { "grad_norm": 0.30771031975746155, "learning_rate": 7.330291647573928e-05, "loss": 0.0287, "step": 37840 }, { "grad_norm": 0.21528205275535583, "learning_rate": 7.32882860615349e-05, "loss": 0.0264, "step": 37850 }, { "grad_norm": 0.21235822141170502, "learning_rate": 7.327365310055858e-05, "loss": 0.028, "step": 37860 }, { "grad_norm": 0.23764708638191223, "learning_rate": 7.325901759441055e-05, "loss": 0.0235, "step": 37870 }, { "grad_norm": 0.23409412801265717, "learning_rate": 7.324437954469132e-05, "loss": 0.0268, "step": 37880 }, { "grad_norm": 0.24682869017124176, "learning_rate": 7.32297389530017e-05, "loss": 0.0266, "step": 37890 }, { "grad_norm": 0.21075166761875153, "learning_rate": 7.321509582094274e-05, "loss": 0.0254, "step": 37900 }, { "grad_norm": 0.2332024723291397, "learning_rate": 7.320045015011582e-05, "loss": 0.0219, "step": 37910 }, { "grad_norm": 0.25820931792259216, "learning_rate": 7.318580194212253e-05, "loss": 0.0271, "step": 37920 }, { "grad_norm": 0.19127438962459564, "learning_rate": 7.317115119856483e-05, "loss": 0.0227, "step": 37930 }, { "grad_norm": 0.27038389444351196, "learning_rate": 7.315649792104485e-05, "loss": 0.0262, "step": 37940 }, { "grad_norm": 0.35153448581695557, "learning_rate": 7.314184211116507e-05, "loss": 0.0293, "step": 37950 }, { "grad_norm": 0.27900829911231995, "learning_rate": 7.312718377052823e-05, "loss": 0.0255, "step": 37960 }, { "grad_norm": 0.26419082283973694, "learning_rate": 7.311252290073737e-05, "loss": 0.0316, "step": 37970 }, { "grad_norm": 0.3437168598175049, "learning_rate": 7.309785950339572e-05, "loss": 0.0263, "step": 37980 }, { "grad_norm": 0.23496362566947937, "learning_rate": 7.308319358010692e-05, "loss": 0.0302, "step": 37990 }, { "grad_norm": 0.2992594242095947, "learning_rate": 7.306852513247474e-05, "loss": 0.0257, "step": 38000 }, { "grad_norm": 0.21420879662036896, "learning_rate": 7.305385416210336e-05, "loss": 0.0236, "step": 38010 }, { "grad_norm": 0.24357075989246368, "learning_rate": 7.303918067059711e-05, "loss": 0.025, "step": 38020 }, { "grad_norm": 0.26578351855278015, "learning_rate": 7.302450465956075e-05, "loss": 0.0281, "step": 38030 }, { "grad_norm": 0.29717373847961426, "learning_rate": 7.300982613059914e-05, "loss": 0.0264, "step": 38040 }, { "grad_norm": 0.24371589720249176, "learning_rate": 7.299514508531757e-05, "loss": 0.027, "step": 38050 }, { "grad_norm": 0.2732519209384918, "learning_rate": 7.298046152532148e-05, "loss": 0.0269, "step": 38060 }, { "grad_norm": 0.260797381401062, "learning_rate": 7.296577545221668e-05, "loss": 0.0263, "step": 38070 }, { "grad_norm": 0.20671889185905457, "learning_rate": 7.29510868676092e-05, "loss": 0.0211, "step": 38080 }, { "grad_norm": 0.19003993272781372, "learning_rate": 7.293639577310538e-05, "loss": 0.0229, "step": 38090 }, { "grad_norm": 0.2983749806880951, "learning_rate": 7.292170217031179e-05, "loss": 0.021, "step": 38100 }, { "grad_norm": 0.24433766305446625, "learning_rate": 7.290700606083532e-05, "loss": 0.0234, "step": 38110 }, { "grad_norm": 0.20350754261016846, "learning_rate": 7.28923074462831e-05, "loss": 0.0215, "step": 38120 }, { "grad_norm": 0.2820392847061157, "learning_rate": 7.287760632826255e-05, "loss": 0.0246, "step": 38130 }, { "grad_norm": 0.19286072254180908, "learning_rate": 7.286290270838139e-05, "loss": 0.0256, "step": 38140 }, { "grad_norm": 0.2727586627006531, "learning_rate": 7.284819658824756e-05, "loss": 0.0287, "step": 38150 }, { "grad_norm": 0.1760915368795395, "learning_rate": 7.283348796946929e-05, "loss": 0.029, "step": 38160 }, { "grad_norm": 0.25889474153518677, "learning_rate": 7.28187768536551e-05, "loss": 0.0268, "step": 38170 }, { "grad_norm": 0.2058410793542862, "learning_rate": 7.280406324241379e-05, "loss": 0.0205, "step": 38180 }, { "grad_norm": 0.22234806418418884, "learning_rate": 7.278934713735441e-05, "loss": 0.0263, "step": 38190 }, { "grad_norm": 0.26016417145729065, "learning_rate": 7.277462854008629e-05, "loss": 0.023, "step": 38200 }, { "grad_norm": 0.22863967716693878, "learning_rate": 7.275990745221904e-05, "loss": 0.0181, "step": 38210 }, { "grad_norm": 0.2482001930475235, "learning_rate": 7.274518387536252e-05, "loss": 0.0233, "step": 38220 }, { "grad_norm": 0.2376280426979065, "learning_rate": 7.273045781112689e-05, "loss": 0.0243, "step": 38230 }, { "grad_norm": 0.2574462890625, "learning_rate": 7.271572926112257e-05, "loss": 0.0251, "step": 38240 }, { "grad_norm": 0.1661529690027237, "learning_rate": 7.270099822696024e-05, "loss": 0.0241, "step": 38250 }, { "grad_norm": 0.1797553151845932, "learning_rate": 7.26862647102509e-05, "loss": 0.021, "step": 38260 }, { "grad_norm": 0.22379840910434723, "learning_rate": 7.267152871260573e-05, "loss": 0.0208, "step": 38270 }, { "grad_norm": 0.24895624816417694, "learning_rate": 7.265679023563629e-05, "loss": 0.0222, "step": 38280 }, { "grad_norm": 0.2383168339729309, "learning_rate": 7.264204928095433e-05, "loss": 0.0254, "step": 38290 }, { "grad_norm": 0.1984308958053589, "learning_rate": 7.262730585017188e-05, "loss": 0.0243, "step": 38300 }, { "grad_norm": 0.212418332695961, "learning_rate": 7.261255994490131e-05, "loss": 0.0244, "step": 38310 }, { "grad_norm": 0.28594115376472473, "learning_rate": 7.259781156675517e-05, "loss": 0.0232, "step": 38320 }, { "grad_norm": 0.17380593717098236, "learning_rate": 7.258306071734633e-05, "loss": 0.0233, "step": 38330 }, { "grad_norm": 0.2971343398094177, "learning_rate": 7.256830739828792e-05, "loss": 0.0236, "step": 38340 }, { "grad_norm": 0.3745144307613373, "learning_rate": 7.255355161119336e-05, "loss": 0.0274, "step": 38350 }, { "grad_norm": 0.2281854748725891, "learning_rate": 7.253879335767628e-05, "loss": 0.0247, "step": 38360 }, { "grad_norm": 0.2932676076889038, "learning_rate": 7.252403263935067e-05, "loss": 0.027, "step": 38370 }, { "grad_norm": 0.4138113856315613, "learning_rate": 7.25092694578307e-05, "loss": 0.0291, "step": 38380 }, { "grad_norm": 0.24252785742282867, "learning_rate": 7.249450381473087e-05, "loss": 0.0267, "step": 38390 }, { "grad_norm": 0.2038959413766861, "learning_rate": 7.247973571166593e-05, "loss": 0.0255, "step": 38400 }, { "grad_norm": 0.24031376838684082, "learning_rate": 7.246496515025089e-05, "loss": 0.0239, "step": 38410 }, { "grad_norm": 0.1701529622077942, "learning_rate": 7.245019213210101e-05, "loss": 0.023, "step": 38420 }, { "grad_norm": 0.3138541579246521, "learning_rate": 7.24354166588319e-05, "loss": 0.0257, "step": 38430 }, { "grad_norm": 0.2175561934709549, "learning_rate": 7.242063873205935e-05, "loss": 0.0202, "step": 38440 }, { "grad_norm": 0.16825860738754272, "learning_rate": 7.240585835339946e-05, "loss": 0.0243, "step": 38450 }, { "grad_norm": 0.25206753611564636, "learning_rate": 7.239107552446857e-05, "loss": 0.0244, "step": 38460 }, { "grad_norm": 0.21444177627563477, "learning_rate": 7.237629024688334e-05, "loss": 0.0211, "step": 38470 }, { "grad_norm": 0.20647110044956207, "learning_rate": 7.236150252226064e-05, "loss": 0.0239, "step": 38480 }, { "grad_norm": 0.2374476194381714, "learning_rate": 7.234671235221765e-05, "loss": 0.0256, "step": 38490 }, { "grad_norm": 0.2120506912469864, "learning_rate": 7.233191973837179e-05, "loss": 0.0261, "step": 38500 }, { "grad_norm": 0.2276383936405182, "learning_rate": 7.231712468234078e-05, "loss": 0.0218, "step": 38510 }, { "grad_norm": 0.18276414275169373, "learning_rate": 7.230232718574254e-05, "loss": 0.0261, "step": 38520 }, { "grad_norm": 0.2680802047252655, "learning_rate": 7.228752725019535e-05, "loss": 0.0249, "step": 38530 }, { "grad_norm": 0.21768860518932343, "learning_rate": 7.227272487731769e-05, "loss": 0.0196, "step": 38540 }, { "grad_norm": 0.19638866186141968, "learning_rate": 7.225792006872831e-05, "loss": 0.0236, "step": 38550 }, { "grad_norm": 0.29192981123924255, "learning_rate": 7.224311282604628e-05, "loss": 0.0259, "step": 38560 }, { "grad_norm": 0.22505763173103333, "learning_rate": 7.222830315089085e-05, "loss": 0.029, "step": 38570 }, { "grad_norm": 0.1564243733882904, "learning_rate": 7.22134910448816e-05, "loss": 0.023, "step": 38580 }, { "grad_norm": 0.31861844658851624, "learning_rate": 7.219867650963839e-05, "loss": 0.0203, "step": 38590 }, { "grad_norm": 0.4394116997718811, "learning_rate": 7.21838595467813e-05, "loss": 0.0241, "step": 38600 }, { "grad_norm": 0.2933160960674286, "learning_rate": 7.216904015793067e-05, "loss": 0.0231, "step": 38610 }, { "grad_norm": 0.2532348036766052, "learning_rate": 7.215421834470713e-05, "loss": 0.0254, "step": 38620 }, { "grad_norm": 0.2961815297603607, "learning_rate": 7.213939410873158e-05, "loss": 0.0274, "step": 38630 }, { "grad_norm": 0.16299791634082794, "learning_rate": 7.212456745162519e-05, "loss": 0.0236, "step": 38640 }, { "grad_norm": 0.2491084188222885, "learning_rate": 7.210973837500937e-05, "loss": 0.027, "step": 38650 }, { "grad_norm": 0.21957926452159882, "learning_rate": 7.209490688050578e-05, "loss": 0.0218, "step": 38660 }, { "grad_norm": 0.3636610209941864, "learning_rate": 7.208007296973641e-05, "loss": 0.0305, "step": 38670 }, { "grad_norm": 0.19306215643882751, "learning_rate": 7.206523664432345e-05, "loss": 0.024, "step": 38680 }, { "grad_norm": 0.2068105787038803, "learning_rate": 7.205039790588939e-05, "loss": 0.0233, "step": 38690 }, { "grad_norm": 0.18887318670749664, "learning_rate": 7.203555675605697e-05, "loss": 0.0213, "step": 38700 }, { "grad_norm": 0.19355955719947815, "learning_rate": 7.202071319644917e-05, "loss": 0.021, "step": 38710 }, { "grad_norm": 0.2105187028646469, "learning_rate": 7.200586722868932e-05, "loss": 0.0252, "step": 38720 }, { "grad_norm": 0.21708592772483826, "learning_rate": 7.199101885440088e-05, "loss": 0.0218, "step": 38730 }, { "grad_norm": 0.26911163330078125, "learning_rate": 7.197616807520771e-05, "loss": 0.0245, "step": 38740 }, { "grad_norm": 0.5099918246269226, "learning_rate": 7.196131489273381e-05, "loss": 0.0241, "step": 38750 }, { "grad_norm": 0.19745267927646637, "learning_rate": 7.194645930860355e-05, "loss": 0.0218, "step": 38760 }, { "grad_norm": 0.2832915186882019, "learning_rate": 7.19316013244415e-05, "loss": 0.0243, "step": 38770 }, { "grad_norm": 0.19100338220596313, "learning_rate": 7.191674094187248e-05, "loss": 0.0293, "step": 38780 }, { "grad_norm": 0.3236902058124542, "learning_rate": 7.190187816252165e-05, "loss": 0.0265, "step": 38790 }, { "grad_norm": 0.2178642898797989, "learning_rate": 7.188701298801435e-05, "loss": 0.0259, "step": 38800 }, { "grad_norm": 0.24971969425678253, "learning_rate": 7.18721454199762e-05, "loss": 0.0277, "step": 38810 }, { "grad_norm": 0.22317153215408325, "learning_rate": 7.185727546003312e-05, "loss": 0.025, "step": 38820 }, { "grad_norm": 0.350557804107666, "learning_rate": 7.184240310981126e-05, "loss": 0.0243, "step": 38830 }, { "grad_norm": 0.20367015898227692, "learning_rate": 7.182752837093704e-05, "loss": 0.021, "step": 38840 }, { "grad_norm": 0.19069692492485046, "learning_rate": 7.181265124503711e-05, "loss": 0.0222, "step": 38850 }, { "grad_norm": 0.20515558123588562, "learning_rate": 7.179777173373847e-05, "loss": 0.0242, "step": 38860 }, { "grad_norm": 0.1806616634130478, "learning_rate": 7.178288983866826e-05, "loss": 0.024, "step": 38870 }, { "grad_norm": 0.23366552591323853, "learning_rate": 7.176800556145397e-05, "loss": 0.0252, "step": 38880 }, { "grad_norm": 0.26724565029144287, "learning_rate": 7.175311890372334e-05, "loss": 0.0228, "step": 38890 }, { "grad_norm": 0.20164261758327484, "learning_rate": 7.17382298671043e-05, "loss": 0.025, "step": 38900 }, { "grad_norm": 0.21174131333827972, "learning_rate": 7.172333845322515e-05, "loss": 0.0219, "step": 38910 }, { "grad_norm": 0.17891661822795868, "learning_rate": 7.170844466371436e-05, "loss": 0.0228, "step": 38920 }, { "grad_norm": 0.18830639123916626, "learning_rate": 7.16935485002007e-05, "loss": 0.032, "step": 38930 }, { "grad_norm": 0.18592019379138947, "learning_rate": 7.167864996431319e-05, "loss": 0.0216, "step": 38940 }, { "grad_norm": 0.28295472264289856, "learning_rate": 7.166374905768111e-05, "loss": 0.023, "step": 38950 }, { "grad_norm": 0.43868181109428406, "learning_rate": 7.164884578193401e-05, "loss": 0.0247, "step": 38960 }, { "grad_norm": 0.23733006417751312, "learning_rate": 7.163394013870168e-05, "loss": 0.0249, "step": 38970 }, { "grad_norm": 0.21384134888648987, "learning_rate": 7.161903212961419e-05, "loss": 0.0235, "step": 38980 }, { "grad_norm": 0.204920694231987, "learning_rate": 7.160412175630183e-05, "loss": 0.0234, "step": 38990 }, { "grad_norm": 0.24186848104000092, "learning_rate": 7.158920902039521e-05, "loss": 0.0205, "step": 39000 }, { "grad_norm": 0.2124408781528473, "learning_rate": 7.157429392352514e-05, "loss": 0.0242, "step": 39010 }, { "grad_norm": 0.2343573123216629, "learning_rate": 7.155937646732275e-05, "loss": 0.03, "step": 39020 }, { "grad_norm": 0.20887865126132965, "learning_rate": 7.154445665341933e-05, "loss": 0.0253, "step": 39030 }, { "grad_norm": 0.22824636101722717, "learning_rate": 7.152953448344654e-05, "loss": 0.025, "step": 39040 }, { "grad_norm": 0.21072889864444733, "learning_rate": 7.151460995903624e-05, "loss": 0.0247, "step": 39050 }, { "grad_norm": 0.2541373372077942, "learning_rate": 7.149968308182052e-05, "loss": 0.0238, "step": 39060 }, { "grad_norm": 0.25812050700187683, "learning_rate": 7.14847538534318e-05, "loss": 0.0216, "step": 39070 }, { "grad_norm": 0.22290495038032532, "learning_rate": 7.14698222755027e-05, "loss": 0.0275, "step": 39080 }, { "grad_norm": 0.21019034087657928, "learning_rate": 7.14548883496661e-05, "loss": 0.0232, "step": 39090 }, { "grad_norm": 0.29005202651023865, "learning_rate": 7.143995207755517e-05, "loss": 0.029, "step": 39100 }, { "grad_norm": 0.1866532564163208, "learning_rate": 7.142501346080333e-05, "loss": 0.0218, "step": 39110 }, { "grad_norm": 0.26868343353271484, "learning_rate": 7.141007250104421e-05, "loss": 0.0232, "step": 39120 }, { "grad_norm": 0.2184489220380783, "learning_rate": 7.139512919991176e-05, "loss": 0.0225, "step": 39130 }, { "grad_norm": 0.20098300278186798, "learning_rate": 7.138018355904014e-05, "loss": 0.0222, "step": 39140 }, { "grad_norm": 0.1739760786294937, "learning_rate": 7.13652355800638e-05, "loss": 0.0207, "step": 39150 }, { "grad_norm": 0.20632849633693695, "learning_rate": 7.135028526461743e-05, "loss": 0.0208, "step": 39160 }, { "grad_norm": 0.23371262848377228, "learning_rate": 7.133533261433594e-05, "loss": 0.0232, "step": 39170 }, { "grad_norm": 0.4118814170360565, "learning_rate": 7.132037763085457e-05, "loss": 0.0264, "step": 39180 }, { "grad_norm": 0.22517257928848267, "learning_rate": 7.130542031580875e-05, "loss": 0.0246, "step": 39190 }, { "grad_norm": 0.15834401547908783, "learning_rate": 7.12904606708342e-05, "loss": 0.0229, "step": 39200 }, { "grad_norm": 0.2509422302246094, "learning_rate": 7.127549869756687e-05, "loss": 0.0215, "step": 39210 }, { "grad_norm": 0.2708394527435303, "learning_rate": 7.126053439764299e-05, "loss": 0.02, "step": 39220 }, { "grad_norm": 0.22040322422981262, "learning_rate": 7.124556777269904e-05, "loss": 0.0224, "step": 39230 }, { "grad_norm": 0.21665048599243164, "learning_rate": 7.123059882437174e-05, "loss": 0.024, "step": 39240 }, { "grad_norm": 0.23997502028942108, "learning_rate": 7.121562755429807e-05, "loss": 0.0243, "step": 39250 }, { "grad_norm": 0.22323879599571228, "learning_rate": 7.120065396411528e-05, "loss": 0.022, "step": 39260 }, { "grad_norm": 0.24866439402103424, "learning_rate": 7.118567805546084e-05, "loss": 0.0258, "step": 39270 }, { "grad_norm": 0.1895042359828949, "learning_rate": 7.117069982997248e-05, "loss": 0.023, "step": 39280 }, { "grad_norm": 0.19594059884548187, "learning_rate": 7.115571928928825e-05, "loss": 0.0246, "step": 39290 }, { "grad_norm": 0.14124436676502228, "learning_rate": 7.114073643504635e-05, "loss": 0.0201, "step": 39300 }, { "grad_norm": 0.21532537043094635, "learning_rate": 7.11257512688853e-05, "loss": 0.0239, "step": 39310 }, { "grad_norm": 0.23328107595443726, "learning_rate": 7.111076379244384e-05, "loss": 0.0219, "step": 39320 }, { "grad_norm": 0.2688004672527313, "learning_rate": 7.109577400736101e-05, "loss": 0.0197, "step": 39330 }, { "grad_norm": 0.23266395926475525, "learning_rate": 7.108078191527605e-05, "loss": 0.027, "step": 39340 }, { "grad_norm": 0.25991177558898926, "learning_rate": 7.106578751782847e-05, "loss": 0.027, "step": 39350 }, { "grad_norm": 0.24601536989212036, "learning_rate": 7.105079081665803e-05, "loss": 0.0236, "step": 39360 }, { "grad_norm": 0.145075261592865, "learning_rate": 7.103579181340476e-05, "loss": 0.0266, "step": 39370 }, { "grad_norm": 0.20929349958896637, "learning_rate": 7.102079050970893e-05, "loss": 0.0219, "step": 39380 }, { "grad_norm": 0.17829963564872742, "learning_rate": 7.100578690721105e-05, "loss": 0.0227, "step": 39390 }, { "grad_norm": 0.20910727977752686, "learning_rate": 7.09907810075519e-05, "loss": 0.0207, "step": 39400 }, { "grad_norm": 0.21363234519958496, "learning_rate": 7.097577281237249e-05, "loss": 0.0263, "step": 39410 }, { "grad_norm": 0.20362748205661774, "learning_rate": 7.09607623233141e-05, "loss": 0.0237, "step": 39420 }, { "grad_norm": 0.22347111999988556, "learning_rate": 7.094574954201827e-05, "loss": 0.0188, "step": 39430 }, { "grad_norm": 0.20116589963436127, "learning_rate": 7.093073447012675e-05, "loss": 0.0252, "step": 39440 }, { "grad_norm": 0.24535837769508362, "learning_rate": 7.09157171092816e-05, "loss": 0.0268, "step": 39450 }, { "grad_norm": 0.21667997539043427, "learning_rate": 7.090069746112504e-05, "loss": 0.023, "step": 39460 }, { "grad_norm": 0.22001324594020844, "learning_rate": 7.088567552729965e-05, "loss": 0.0248, "step": 39470 }, { "grad_norm": 0.23388943076133728, "learning_rate": 7.087065130944818e-05, "loss": 0.0255, "step": 39480 }, { "grad_norm": 0.20675355195999146, "learning_rate": 7.085562480921366e-05, "loss": 0.0233, "step": 39490 }, { "grad_norm": 0.2928963005542755, "learning_rate": 7.084059602823937e-05, "loss": 0.0236, "step": 39500 }, { "grad_norm": 0.21189893782138824, "learning_rate": 7.082556496816882e-05, "loss": 0.0194, "step": 39510 }, { "grad_norm": 0.1654176115989685, "learning_rate": 7.081053163064582e-05, "loss": 0.0243, "step": 39520 }, { "grad_norm": 0.1888871043920517, "learning_rate": 7.079549601731434e-05, "loss": 0.026, "step": 39530 }, { "grad_norm": 0.2595691382884979, "learning_rate": 7.07804581298187e-05, "loss": 0.0275, "step": 39540 }, { "grad_norm": 0.257061630487442, "learning_rate": 7.07654179698034e-05, "loss": 0.0237, "step": 39550 }, { "grad_norm": 0.23582008481025696, "learning_rate": 7.075037553891321e-05, "loss": 0.0222, "step": 39560 }, { "grad_norm": 0.24097585678100586, "learning_rate": 7.073533083879315e-05, "loss": 0.0226, "step": 39570 }, { "grad_norm": 0.3200519382953644, "learning_rate": 7.072028387108849e-05, "loss": 0.0231, "step": 39580 }, { "grad_norm": 0.2441716194152832, "learning_rate": 7.070523463744472e-05, "loss": 0.0237, "step": 39590 }, { "grad_norm": 0.30585741996765137, "learning_rate": 7.069018313950763e-05, "loss": 0.0229, "step": 39600 }, { "grad_norm": 0.2370944321155548, "learning_rate": 7.067512937892322e-05, "loss": 0.0233, "step": 39610 }, { "grad_norm": 0.23719577491283417, "learning_rate": 7.066007335733775e-05, "loss": 0.0233, "step": 39620 }, { "grad_norm": 0.3113500773906708, "learning_rate": 7.064501507639772e-05, "loss": 0.0273, "step": 39630 }, { "grad_norm": 0.17838238179683685, "learning_rate": 7.062995453774987e-05, "loss": 0.025, "step": 39640 }, { "grad_norm": 0.30703914165496826, "learning_rate": 7.061489174304121e-05, "loss": 0.0255, "step": 39650 }, { "grad_norm": 0.40220507979393005, "learning_rate": 7.0599826693919e-05, "loss": 0.0275, "step": 39660 }, { "grad_norm": 0.2202024906873703, "learning_rate": 7.05847593920307e-05, "loss": 0.0248, "step": 39670 }, { "grad_norm": 0.23211048543453217, "learning_rate": 7.056968983902406e-05, "loss": 0.0255, "step": 39680 }, { "grad_norm": 0.16443410515785217, "learning_rate": 7.055461803654706e-05, "loss": 0.0236, "step": 39690 }, { "grad_norm": 0.23208776116371155, "learning_rate": 7.053954398624794e-05, "loss": 0.0231, "step": 39700 }, { "grad_norm": 0.21522551774978638, "learning_rate": 7.052446768977518e-05, "loss": 0.0268, "step": 39710 }, { "grad_norm": 0.2901465892791748, "learning_rate": 7.050938914877748e-05, "loss": 0.0218, "step": 39720 }, { "grad_norm": 0.2253042459487915, "learning_rate": 7.049430836490382e-05, "loss": 0.0245, "step": 39730 }, { "grad_norm": 0.20685310661792755, "learning_rate": 7.047922533980342e-05, "loss": 0.0274, "step": 39740 }, { "grad_norm": 0.22451326251029968, "learning_rate": 7.046414007512571e-05, "loss": 0.0253, "step": 39750 }, { "grad_norm": 0.17338962852954865, "learning_rate": 7.044905257252042e-05, "loss": 0.0237, "step": 39760 }, { "grad_norm": 0.21422600746154785, "learning_rate": 7.04339628336375e-05, "loss": 0.0206, "step": 39770 }, { "grad_norm": 0.2880673408508301, "learning_rate": 7.041887086012711e-05, "loss": 0.0267, "step": 39780 }, { "grad_norm": 0.20727337896823883, "learning_rate": 7.040377665363969e-05, "loss": 0.0224, "step": 39790 }, { "grad_norm": 0.24013671278953552, "learning_rate": 7.038868021582594e-05, "loss": 0.0249, "step": 39800 }, { "grad_norm": 0.2757672965526581, "learning_rate": 7.037358154833679e-05, "loss": 0.0224, "step": 39810 }, { "grad_norm": 0.20622067153453827, "learning_rate": 7.035848065282339e-05, "loss": 0.0305, "step": 39820 }, { "grad_norm": 0.1750463992357254, "learning_rate": 7.034337753093714e-05, "loss": 0.0227, "step": 39830 }, { "grad_norm": 0.20560668408870697, "learning_rate": 7.032827218432972e-05, "loss": 0.0227, "step": 39840 }, { "grad_norm": 0.401431143283844, "learning_rate": 7.031316461465302e-05, "loss": 0.0212, "step": 39850 }, { "grad_norm": 0.22128844261169434, "learning_rate": 7.029805482355915e-05, "loss": 0.0238, "step": 39860 }, { "grad_norm": 0.21947841346263885, "learning_rate": 7.028294281270055e-05, "loss": 0.0238, "step": 39870 }, { "grad_norm": 0.22658951580524445, "learning_rate": 7.026782858372982e-05, "loss": 0.0249, "step": 39880 }, { "grad_norm": 0.2568283677101135, "learning_rate": 7.025271213829982e-05, "loss": 0.0226, "step": 39890 }, { "grad_norm": 0.2214960902929306, "learning_rate": 7.023759347806366e-05, "loss": 0.0246, "step": 39900 }, { "grad_norm": 0.2180868238210678, "learning_rate": 7.022247260467472e-05, "loss": 0.024, "step": 39910 }, { "grad_norm": 0.27304723858833313, "learning_rate": 7.020734951978658e-05, "loss": 0.0254, "step": 39920 }, { "grad_norm": 0.19184669852256775, "learning_rate": 7.019222422505307e-05, "loss": 0.0243, "step": 39930 }, { "grad_norm": 0.21789324283599854, "learning_rate": 7.017709672212827e-05, "loss": 0.0267, "step": 39940 }, { "grad_norm": 0.2132725864648819, "learning_rate": 7.016196701266652e-05, "loss": 0.0247, "step": 39950 }, { "grad_norm": 0.28160879015922546, "learning_rate": 7.014683509832238e-05, "loss": 0.0224, "step": 39960 }, { "grad_norm": 0.24010924994945526, "learning_rate": 7.013170098075063e-05, "loss": 0.0241, "step": 39970 }, { "grad_norm": 0.29206761717796326, "learning_rate": 7.011656466160632e-05, "loss": 0.0234, "step": 39980 }, { "grad_norm": 0.2717410922050476, "learning_rate": 7.010142614254475e-05, "loss": 0.0272, "step": 39990 }, { "grad_norm": 0.16757628321647644, "learning_rate": 7.008628542522147e-05, "loss": 0.0225, "step": 40000 }, { "grad_norm": 0.25330379605293274, "learning_rate": 7.007114251129218e-05, "loss": 0.035, "step": 40010 }, { "grad_norm": 0.2877977192401886, "learning_rate": 7.005599740241293e-05, "loss": 0.0232, "step": 40020 }, { "grad_norm": 0.2280188798904419, "learning_rate": 7.004085010023996e-05, "loss": 0.0234, "step": 40030 }, { "grad_norm": 0.17020411789417267, "learning_rate": 7.002570060642976e-05, "loss": 0.0249, "step": 40040 }, { "grad_norm": 0.2382514327764511, "learning_rate": 7.001054892263903e-05, "loss": 0.0267, "step": 40050 }, { "grad_norm": 0.3278485834598541, "learning_rate": 6.999539505052477e-05, "loss": 0.0285, "step": 40060 }, { "grad_norm": 0.24629049003124237, "learning_rate": 6.998023899174418e-05, "loss": 0.0269, "step": 40070 }, { "grad_norm": 0.24483232200145721, "learning_rate": 6.996508074795467e-05, "loss": 0.0238, "step": 40080 }, { "grad_norm": 0.22199364006519318, "learning_rate": 6.994992032081396e-05, "loss": 0.0245, "step": 40090 }, { "grad_norm": 0.18498528003692627, "learning_rate": 6.993475771197995e-05, "loss": 0.0245, "step": 40100 }, { "grad_norm": 0.23890919983386993, "learning_rate": 6.991959292311082e-05, "loss": 0.025, "step": 40110 }, { "grad_norm": 0.1447506546974182, "learning_rate": 6.990442595586495e-05, "loss": 0.0241, "step": 40120 }, { "grad_norm": 0.21113014221191406, "learning_rate": 6.988925681190098e-05, "loss": 0.0238, "step": 40130 }, { "grad_norm": 0.2711774706840515, "learning_rate": 6.987408549287778e-05, "loss": 0.0289, "step": 40140 }, { "grad_norm": 0.37318798899650574, "learning_rate": 6.985891200045449e-05, "loss": 0.0264, "step": 40150 }, { "grad_norm": 0.28217288851737976, "learning_rate": 6.984373633629045e-05, "loss": 0.0244, "step": 40160 }, { "grad_norm": 0.23556320369243622, "learning_rate": 6.98285585020452e-05, "loss": 0.0278, "step": 40170 }, { "grad_norm": 0.2045128494501114, "learning_rate": 6.981337849937864e-05, "loss": 0.0256, "step": 40180 }, { "grad_norm": 0.23297816514968872, "learning_rate": 6.979819632995078e-05, "loss": 0.0244, "step": 40190 }, { "grad_norm": 0.2659843862056732, "learning_rate": 6.978301199542193e-05, "loss": 0.0308, "step": 40200 }, { "grad_norm": 0.26498815417289734, "learning_rate": 6.976782549745263e-05, "loss": 0.0301, "step": 40210 }, { "grad_norm": 0.3310891389846802, "learning_rate": 6.975263683770367e-05, "loss": 0.0264, "step": 40220 }, { "grad_norm": 0.24541136622428894, "learning_rate": 6.973744601783604e-05, "loss": 0.0225, "step": 40230 }, { "grad_norm": 0.26360127329826355, "learning_rate": 6.972225303951097e-05, "loss": 0.0284, "step": 40240 }, { "grad_norm": 0.27830106019973755, "learning_rate": 6.970705790438998e-05, "loss": 0.0238, "step": 40250 }, { "grad_norm": 0.2737811505794525, "learning_rate": 6.969186061413477e-05, "loss": 0.0228, "step": 40260 }, { "grad_norm": 0.3810705542564392, "learning_rate": 6.967666117040727e-05, "loss": 0.0289, "step": 40270 }, { "grad_norm": 0.30961549282073975, "learning_rate": 6.966145957486972e-05, "loss": 0.0221, "step": 40280 }, { "grad_norm": 0.21297647058963776, "learning_rate": 6.964625582918449e-05, "loss": 0.0225, "step": 40290 }, { "grad_norm": 0.2774161994457245, "learning_rate": 6.963104993501425e-05, "loss": 0.0305, "step": 40300 }, { "grad_norm": 0.26059189438819885, "learning_rate": 6.961584189402192e-05, "loss": 0.0233, "step": 40310 }, { "grad_norm": 0.16720092296600342, "learning_rate": 6.96006317078706e-05, "loss": 0.0256, "step": 40320 }, { "grad_norm": 0.3135337829589844, "learning_rate": 6.958541937822367e-05, "loss": 0.0258, "step": 40330 }, { "grad_norm": 0.37201476097106934, "learning_rate": 6.957020490674473e-05, "loss": 0.0243, "step": 40340 }, { "grad_norm": 0.21835777163505554, "learning_rate": 6.95549882950976e-05, "loss": 0.022, "step": 40350 }, { "grad_norm": 0.2066330760717392, "learning_rate": 6.953976954494635e-05, "loss": 0.0253, "step": 40360 }, { "grad_norm": 0.21509486436843872, "learning_rate": 6.952454865795528e-05, "loss": 0.0273, "step": 40370 }, { "grad_norm": 0.24624374508857727, "learning_rate": 6.950932563578892e-05, "loss": 0.0256, "step": 40380 }, { "grad_norm": 0.34122228622436523, "learning_rate": 6.949410048011206e-05, "loss": 0.0252, "step": 40390 }, { "grad_norm": 0.17714326083660126, "learning_rate": 6.947887319258966e-05, "loss": 0.019, "step": 40400 }, { "grad_norm": 0.18607032299041748, "learning_rate": 6.946364377488696e-05, "loss": 0.0233, "step": 40410 }, { "grad_norm": 0.2386520802974701, "learning_rate": 6.944841222866947e-05, "loss": 0.0221, "step": 40420 }, { "grad_norm": 0.23072142899036407, "learning_rate": 6.943317855560284e-05, "loss": 0.0201, "step": 40430 }, { "grad_norm": 0.19840726256370544, "learning_rate": 6.941794275735302e-05, "loss": 0.0282, "step": 40440 }, { "grad_norm": 0.38147950172424316, "learning_rate": 6.94027048355862e-05, "loss": 0.0227, "step": 40450 }, { "grad_norm": 0.1414809226989746, "learning_rate": 6.938746479196871e-05, "loss": 0.0252, "step": 40460 }, { "grad_norm": 0.22276800870895386, "learning_rate": 6.937222262816724e-05, "loss": 0.0211, "step": 40470 }, { "grad_norm": 0.3471159040927887, "learning_rate": 6.935697834584865e-05, "loss": 0.0254, "step": 40480 }, { "grad_norm": 0.22968925535678864, "learning_rate": 6.934173194667998e-05, "loss": 0.0245, "step": 40490 }, { "grad_norm": 0.16692005097866058, "learning_rate": 6.93264834323286e-05, "loss": 0.022, "step": 40500 }, { "grad_norm": 0.2232583612203598, "learning_rate": 6.931123280446206e-05, "loss": 0.0261, "step": 40510 }, { "grad_norm": 0.20603592693805695, "learning_rate": 6.929598006474811e-05, "loss": 0.0225, "step": 40520 }, { "grad_norm": 0.2358548492193222, "learning_rate": 6.92807252148548e-05, "loss": 0.0229, "step": 40530 }, { "grad_norm": 0.23133635520935059, "learning_rate": 6.926546825645037e-05, "loss": 0.0186, "step": 40540 }, { "grad_norm": 0.23502157628536224, "learning_rate": 6.92502091912033e-05, "loss": 0.0218, "step": 40550 }, { "grad_norm": 0.2475523203611374, "learning_rate": 6.923494802078229e-05, "loss": 0.0226, "step": 40560 }, { "grad_norm": 0.2620142102241516, "learning_rate": 6.921968474685628e-05, "loss": 0.0237, "step": 40570 }, { "grad_norm": 0.27894264459609985, "learning_rate": 6.920441937109445e-05, "loss": 0.0262, "step": 40580 }, { "grad_norm": 0.28529447317123413, "learning_rate": 6.91891518951662e-05, "loss": 0.0214, "step": 40590 }, { "grad_norm": 0.2222684919834137, "learning_rate": 6.917388232074114e-05, "loss": 0.0227, "step": 40600 }, { "grad_norm": 0.20620138943195343, "learning_rate": 6.915861064948914e-05, "loss": 0.0281, "step": 40610 }, { "grad_norm": 0.2040877640247345, "learning_rate": 6.914333688308025e-05, "loss": 0.0283, "step": 40620 }, { "grad_norm": 0.24231621623039246, "learning_rate": 6.912806102318487e-05, "loss": 0.0237, "step": 40630 }, { "grad_norm": 0.3024965524673462, "learning_rate": 6.911278307147347e-05, "loss": 0.0266, "step": 40640 }, { "grad_norm": 0.1532583385705948, "learning_rate": 6.909750302961684e-05, "loss": 0.0232, "step": 40650 }, { "grad_norm": 0.2998828887939453, "learning_rate": 6.9082220899286e-05, "loss": 0.025, "step": 40660 }, { "grad_norm": 0.28045231103897095, "learning_rate": 6.906693668215219e-05, "loss": 0.0257, "step": 40670 }, { "grad_norm": 0.3249436914920807, "learning_rate": 6.905165037988683e-05, "loss": 0.023, "step": 40680 }, { "grad_norm": 0.17385299503803253, "learning_rate": 6.903636199416164e-05, "loss": 0.0227, "step": 40690 }, { "grad_norm": 0.25994881987571716, "learning_rate": 6.902107152664851e-05, "loss": 0.0309, "step": 40700 }, { "grad_norm": 0.23562559485435486, "learning_rate": 6.90057789790196e-05, "loss": 0.0281, "step": 40710 }, { "grad_norm": 0.23913425207138062, "learning_rate": 6.899048435294728e-05, "loss": 0.023, "step": 40720 }, { "grad_norm": 0.2540616989135742, "learning_rate": 6.897518765010415e-05, "loss": 0.0244, "step": 40730 }, { "grad_norm": 0.20154975354671478, "learning_rate": 6.895988887216302e-05, "loss": 0.0233, "step": 40740 }, { "grad_norm": 0.2387688159942627, "learning_rate": 6.894458802079694e-05, "loss": 0.0244, "step": 40750 }, { "grad_norm": 0.3824100196361542, "learning_rate": 6.892928509767922e-05, "loss": 0.0304, "step": 40760 }, { "grad_norm": 0.255036324262619, "learning_rate": 6.891398010448333e-05, "loss": 0.0303, "step": 40770 }, { "grad_norm": 0.26426398754119873, "learning_rate": 6.8898673042883e-05, "loss": 0.0264, "step": 40780 }, { "grad_norm": 0.20086142420768738, "learning_rate": 6.888336391455222e-05, "loss": 0.0211, "step": 40790 }, { "grad_norm": 0.27799054980278015, "learning_rate": 6.886805272116513e-05, "loss": 0.0286, "step": 40800 }, { "grad_norm": 0.22558999061584473, "learning_rate": 6.885273946439617e-05, "loss": 0.0255, "step": 40810 }, { "grad_norm": 0.21285498142242432, "learning_rate": 6.883742414591998e-05, "loss": 0.0225, "step": 40820 }, { "grad_norm": 0.2246098667383194, "learning_rate": 6.88221067674114e-05, "loss": 0.0194, "step": 40830 }, { "grad_norm": 0.36151647567749023, "learning_rate": 6.88067873305455e-05, "loss": 0.0203, "step": 40840 }, { "grad_norm": 0.25070086121559143, "learning_rate": 6.879146583699765e-05, "loss": 0.02, "step": 40850 }, { "grad_norm": 0.19242256879806519, "learning_rate": 6.877614228844334e-05, "loss": 0.0221, "step": 40860 }, { "grad_norm": 0.2371354103088379, "learning_rate": 6.876081668655832e-05, "loss": 0.0199, "step": 40870 }, { "grad_norm": 0.23662033677101135, "learning_rate": 6.874548903301861e-05, "loss": 0.0289, "step": 40880 }, { "grad_norm": 0.2068629264831543, "learning_rate": 6.873015932950039e-05, "loss": 0.0223, "step": 40890 }, { "grad_norm": 0.260667622089386, "learning_rate": 6.871482757768012e-05, "loss": 0.024, "step": 40900 }, { "grad_norm": 0.2605046033859253, "learning_rate": 6.869949377923442e-05, "loss": 0.0262, "step": 40910 }, { "grad_norm": 0.22153180837631226, "learning_rate": 6.868415793584022e-05, "loss": 0.0216, "step": 40920 }, { "grad_norm": 0.17354311048984528, "learning_rate": 6.86688200491746e-05, "loss": 0.0236, "step": 40930 }, { "grad_norm": 0.23098962008953094, "learning_rate": 6.865348012091488e-05, "loss": 0.0249, "step": 40940 }, { "grad_norm": 0.2352762669324875, "learning_rate": 6.86381381527386e-05, "loss": 0.0283, "step": 40950 }, { "grad_norm": 0.17659468948841095, "learning_rate": 6.862279414632359e-05, "loss": 0.023, "step": 40960 }, { "grad_norm": 0.25407591462135315, "learning_rate": 6.860744810334778e-05, "loss": 0.0311, "step": 40970 }, { "grad_norm": 0.23074756562709808, "learning_rate": 6.859210002548943e-05, "loss": 0.0235, "step": 40980 }, { "grad_norm": 0.20351724326610565, "learning_rate": 6.8576749914427e-05, "loss": 0.0237, "step": 40990 }, { "grad_norm": 0.2123223841190338, "learning_rate": 6.85613977718391e-05, "loss": 0.0236, "step": 41000 }, { "grad_norm": 0.1710900068283081, "learning_rate": 6.854604359940464e-05, "loss": 0.0196, "step": 41010 }, { "grad_norm": 0.2555790841579437, "learning_rate": 6.853068739880276e-05, "loss": 0.0224, "step": 41020 }, { "grad_norm": 0.1999550312757492, "learning_rate": 6.851532917171276e-05, "loss": 0.028, "step": 41030 }, { "grad_norm": 0.24336360394954681, "learning_rate": 6.849996891981418e-05, "loss": 0.0235, "step": 41040 }, { "grad_norm": 0.234670028090477, "learning_rate": 6.848460664478684e-05, "loss": 0.0284, "step": 41050 }, { "grad_norm": 0.22904373705387115, "learning_rate": 6.846924234831069e-05, "loss": 0.0257, "step": 41060 }, { "grad_norm": 0.2515070140361786, "learning_rate": 6.845387603206597e-05, "loss": 0.0228, "step": 41070 }, { "grad_norm": 0.20066499710083008, "learning_rate": 6.843850769773312e-05, "loss": 0.0189, "step": 41080 }, { "grad_norm": 0.2154521942138672, "learning_rate": 6.842313734699278e-05, "loss": 0.0184, "step": 41090 }, { "grad_norm": 0.18483635783195496, "learning_rate": 6.840776498152584e-05, "loss": 0.0255, "step": 41100 }, { "grad_norm": 0.2691235840320587, "learning_rate": 6.83923906030134e-05, "loss": 0.0248, "step": 41110 }, { "grad_norm": 0.2033386081457138, "learning_rate": 6.837701421313677e-05, "loss": 0.0261, "step": 41120 }, { "grad_norm": 0.18392740190029144, "learning_rate": 6.83616358135775e-05, "loss": 0.0234, "step": 41130 }, { "grad_norm": 0.19770771265029907, "learning_rate": 6.834625540601734e-05, "loss": 0.0218, "step": 41140 }, { "grad_norm": 0.243526428937912, "learning_rate": 6.833087299213829e-05, "loss": 0.0202, "step": 41150 }, { "grad_norm": 0.25011080503463745, "learning_rate": 6.83154885736225e-05, "loss": 0.0301, "step": 41160 }, { "grad_norm": 0.244230255484581, "learning_rate": 6.830010215215244e-05, "loss": 0.0216, "step": 41170 }, { "grad_norm": 0.14645260572433472, "learning_rate": 6.828471372941072e-05, "loss": 0.0201, "step": 41180 }, { "grad_norm": 0.21234887838363647, "learning_rate": 6.826932330708018e-05, "loss": 0.0221, "step": 41190 }, { "grad_norm": 0.23718176782131195, "learning_rate": 6.825393088684393e-05, "loss": 0.0204, "step": 41200 }, { "grad_norm": 0.23757831752300262, "learning_rate": 6.823853647038524e-05, "loss": 0.0241, "step": 41210 }, { "grad_norm": 0.23792649805545807, "learning_rate": 6.822314005938762e-05, "loss": 0.0222, "step": 41220 }, { "grad_norm": 0.2694324851036072, "learning_rate": 6.820774165553479e-05, "loss": 0.0187, "step": 41230 }, { "grad_norm": 0.2688738703727722, "learning_rate": 6.819234126051072e-05, "loss": 0.0204, "step": 41240 }, { "grad_norm": 0.27048566937446594, "learning_rate": 6.817693887599956e-05, "loss": 0.0236, "step": 41250 }, { "grad_norm": 0.23422439396381378, "learning_rate": 6.816153450368568e-05, "loss": 0.026, "step": 41260 }, { "grad_norm": 0.2154899686574936, "learning_rate": 6.814612814525372e-05, "loss": 0.0237, "step": 41270 }, { "grad_norm": 0.25473639369010925, "learning_rate": 6.813071980238846e-05, "loss": 0.0216, "step": 41280 }, { "grad_norm": 0.2251034677028656, "learning_rate": 6.811530947677492e-05, "loss": 0.0249, "step": 41290 }, { "grad_norm": 0.1841275542974472, "learning_rate": 6.809989717009839e-05, "loss": 0.0246, "step": 41300 }, { "grad_norm": 0.2515886723995209, "learning_rate": 6.808448288404431e-05, "loss": 0.0253, "step": 41310 }, { "grad_norm": 0.22124838829040527, "learning_rate": 6.806906662029838e-05, "loss": 0.0257, "step": 41320 }, { "grad_norm": 0.21079547703266144, "learning_rate": 6.805364838054647e-05, "loss": 0.024, "step": 41330 }, { "grad_norm": 0.3045833706855774, "learning_rate": 6.803822816647474e-05, "loss": 0.0237, "step": 41340 }, { "grad_norm": 0.19378623366355896, "learning_rate": 6.802280597976949e-05, "loss": 0.0203, "step": 41350 }, { "grad_norm": 0.21045921742916107, "learning_rate": 6.800738182211727e-05, "loss": 0.0285, "step": 41360 }, { "grad_norm": 0.33612751960754395, "learning_rate": 6.799195569520485e-05, "loss": 0.0277, "step": 41370 }, { "grad_norm": 0.27314555644989014, "learning_rate": 6.79765276007192e-05, "loss": 0.0249, "step": 41380 }, { "grad_norm": 0.22131949663162231, "learning_rate": 6.796109754034753e-05, "loss": 0.0273, "step": 41390 }, { "grad_norm": 0.20899616181850433, "learning_rate": 6.794566551577724e-05, "loss": 0.0223, "step": 41400 }, { "grad_norm": 0.1596423238515854, "learning_rate": 6.793023152869594e-05, "loss": 0.0194, "step": 41410 }, { "grad_norm": 0.23683063685894012, "learning_rate": 6.791479558079147e-05, "loss": 0.0267, "step": 41420 }, { "grad_norm": 0.24673503637313843, "learning_rate": 6.789935767375191e-05, "loss": 0.0249, "step": 41430 }, { "grad_norm": 0.1975037306547165, "learning_rate": 6.78839178092655e-05, "loss": 0.0226, "step": 41440 }, { "grad_norm": 0.22092044353485107, "learning_rate": 6.786847598902072e-05, "loss": 0.0251, "step": 41450 }, { "grad_norm": 0.25465330481529236, "learning_rate": 6.785303221470627e-05, "loss": 0.0269, "step": 41460 }, { "grad_norm": 0.24030713737010956, "learning_rate": 6.783758648801108e-05, "loss": 0.0261, "step": 41470 }, { "grad_norm": 0.24663794040679932, "learning_rate": 6.782213881062422e-05, "loss": 0.0282, "step": 41480 }, { "grad_norm": 0.1942249834537506, "learning_rate": 6.780668918423508e-05, "loss": 0.0268, "step": 41490 }, { "grad_norm": 0.18893767893314362, "learning_rate": 6.779123761053317e-05, "loss": 0.0224, "step": 41500 }, { "grad_norm": 0.24165084958076477, "learning_rate": 6.777578409120826e-05, "loss": 0.0264, "step": 41510 }, { "grad_norm": 0.2385156899690628, "learning_rate": 6.776032862795033e-05, "loss": 0.0269, "step": 41520 }, { "grad_norm": 0.19034980237483978, "learning_rate": 6.774487122244956e-05, "loss": 0.0296, "step": 41530 }, { "grad_norm": 0.21120379865169525, "learning_rate": 6.772941187639637e-05, "loss": 0.0243, "step": 41540 }, { "grad_norm": 0.28800168633461, "learning_rate": 6.771395059148134e-05, "loss": 0.0229, "step": 41550 }, { "grad_norm": 0.22651033103466034, "learning_rate": 6.76984873693953e-05, "loss": 0.0219, "step": 41560 }, { "grad_norm": 0.2143089473247528, "learning_rate": 6.768302221182931e-05, "loss": 0.0226, "step": 41570 }, { "grad_norm": 0.22544367611408234, "learning_rate": 6.766755512047457e-05, "loss": 0.0218, "step": 41580 }, { "grad_norm": 0.28717008233070374, "learning_rate": 6.765208609702259e-05, "loss": 0.0268, "step": 41590 }, { "grad_norm": 0.35839101672172546, "learning_rate": 6.763661514316499e-05, "loss": 0.0235, "step": 41600 }, { "grad_norm": 0.1819111853837967, "learning_rate": 6.76211422605937e-05, "loss": 0.0205, "step": 41610 }, { "grad_norm": 0.2702285349369049, "learning_rate": 6.760566745100076e-05, "loss": 0.0225, "step": 41620 }, { "grad_norm": 0.2855518162250519, "learning_rate": 6.75901907160785e-05, "loss": 0.0274, "step": 41630 }, { "grad_norm": 0.16690391302108765, "learning_rate": 6.757471205751943e-05, "loss": 0.0207, "step": 41640 }, { "grad_norm": 0.21775200963020325, "learning_rate": 6.75592314770163e-05, "loss": 0.0229, "step": 41650 }, { "grad_norm": 0.1950998157262802, "learning_rate": 6.7543748976262e-05, "loss": 0.0235, "step": 41660 }, { "grad_norm": 0.19027279317378998, "learning_rate": 6.752826455694968e-05, "loss": 0.0222, "step": 41670 }, { "grad_norm": 0.37793052196502686, "learning_rate": 6.751277822077271e-05, "loss": 0.0213, "step": 41680 }, { "grad_norm": 0.1716262847185135, "learning_rate": 6.749728996942465e-05, "loss": 0.0196, "step": 41690 }, { "grad_norm": 0.226430743932724, "learning_rate": 6.748179980459924e-05, "loss": 0.0232, "step": 41700 }, { "grad_norm": 0.27903318405151367, "learning_rate": 6.746630772799052e-05, "loss": 0.0256, "step": 41710 }, { "grad_norm": 0.2800166606903076, "learning_rate": 6.745081374129262e-05, "loss": 0.023, "step": 41720 }, { "grad_norm": 0.208798348903656, "learning_rate": 6.743531784619998e-05, "loss": 0.0206, "step": 41730 }, { "grad_norm": 0.1905028522014618, "learning_rate": 6.741982004440719e-05, "loss": 0.0213, "step": 41740 }, { "grad_norm": 0.21830224990844727, "learning_rate": 6.740432033760907e-05, "loss": 0.0215, "step": 41750 }, { "grad_norm": 0.2847350537776947, "learning_rate": 6.738881872750066e-05, "loss": 0.0284, "step": 41760 }, { "grad_norm": 0.26485955715179443, "learning_rate": 6.737331521577715e-05, "loss": 0.0233, "step": 41770 }, { "grad_norm": 0.16959211230278015, "learning_rate": 6.735780980413403e-05, "loss": 0.022, "step": 41780 }, { "grad_norm": 0.22533579170703888, "learning_rate": 6.734230249426692e-05, "loss": 0.0186, "step": 41790 }, { "grad_norm": 0.24770556390285492, "learning_rate": 6.732679328787168e-05, "loss": 0.023, "step": 41800 }, { "grad_norm": 0.2560035288333893, "learning_rate": 6.731128218664438e-05, "loss": 0.0284, "step": 41810 }, { "grad_norm": 0.23296642303466797, "learning_rate": 6.729576919228129e-05, "loss": 0.0196, "step": 41820 }, { "grad_norm": 0.2518368661403656, "learning_rate": 6.728025430647888e-05, "loss": 0.024, "step": 41830 }, { "grad_norm": 0.17781147360801697, "learning_rate": 6.726473753093383e-05, "loss": 0.0235, "step": 41840 }, { "grad_norm": 0.2196919173002243, "learning_rate": 6.724921886734305e-05, "loss": 0.0196, "step": 41850 }, { "grad_norm": 0.2500529885292053, "learning_rate": 6.723369831740363e-05, "loss": 0.0228, "step": 41860 }, { "grad_norm": 0.22110655903816223, "learning_rate": 6.721817588281288e-05, "loss": 0.026, "step": 41870 }, { "grad_norm": 0.20014019310474396, "learning_rate": 6.720265156526828e-05, "loss": 0.0207, "step": 41880 }, { "grad_norm": 0.2491474598646164, "learning_rate": 6.718712536646758e-05, "loss": 0.0203, "step": 41890 }, { "grad_norm": 0.24865487217903137, "learning_rate": 6.71715972881087e-05, "loss": 0.02, "step": 41900 }, { "grad_norm": 0.20911334455013275, "learning_rate": 6.715606733188974e-05, "loss": 0.0231, "step": 41910 }, { "grad_norm": 0.18784499168395996, "learning_rate": 6.714053549950909e-05, "loss": 0.0253, "step": 41920 }, { "grad_norm": 0.17575182020664215, "learning_rate": 6.71250017926652e-05, "loss": 0.0237, "step": 41930 }, { "grad_norm": 0.17281067371368408, "learning_rate": 6.71094662130569e-05, "loss": 0.0217, "step": 41940 }, { "grad_norm": 0.4389086961746216, "learning_rate": 6.709392876238307e-05, "loss": 0.0222, "step": 41950 }, { "grad_norm": 0.274431437253952, "learning_rate": 6.707838944234289e-05, "loss": 0.0236, "step": 41960 }, { "grad_norm": 0.22639304399490356, "learning_rate": 6.706284825463574e-05, "loss": 0.02, "step": 41970 }, { "grad_norm": 0.20259791612625122, "learning_rate": 6.704730520096114e-05, "loss": 0.023, "step": 41980 }, { "grad_norm": 0.17057383060455322, "learning_rate": 6.703176028301888e-05, "loss": 0.0288, "step": 41990 }, { "grad_norm": 0.16969169676303864, "learning_rate": 6.701621350250892e-05, "loss": 0.0272, "step": 42000 }, { "grad_norm": 0.18253584206104279, "learning_rate": 6.700066486113144e-05, "loss": 0.0219, "step": 42010 }, { "grad_norm": 0.17493265867233276, "learning_rate": 6.69851143605868e-05, "loss": 0.0205, "step": 42020 }, { "grad_norm": 0.2366630584001541, "learning_rate": 6.696956200257561e-05, "loss": 0.0233, "step": 42030 }, { "grad_norm": 0.21015623211860657, "learning_rate": 6.695400778879863e-05, "loss": 0.0232, "step": 42040 }, { "grad_norm": 0.25808200240135193, "learning_rate": 6.693845172095683e-05, "loss": 0.0229, "step": 42050 }, { "grad_norm": 0.16009218990802765, "learning_rate": 6.692289380075142e-05, "loss": 0.019, "step": 42060 }, { "grad_norm": 0.21493209898471832, "learning_rate": 6.690733402988379e-05, "loss": 0.0241, "step": 42070 }, { "grad_norm": 0.2535351514816284, "learning_rate": 6.689177241005553e-05, "loss": 0.0241, "step": 42080 }, { "grad_norm": 0.23296165466308594, "learning_rate": 6.687620894296844e-05, "loss": 0.0235, "step": 42090 }, { "grad_norm": 0.22751398384571075, "learning_rate": 6.686064363032451e-05, "loss": 0.0258, "step": 42100 }, { "grad_norm": 0.1629992574453354, "learning_rate": 6.684507647382596e-05, "loss": 0.0208, "step": 42110 }, { "grad_norm": 0.23857848346233368, "learning_rate": 6.682950747517513e-05, "loss": 0.0203, "step": 42120 }, { "grad_norm": 0.27440622448921204, "learning_rate": 6.68139366360747e-05, "loss": 0.0318, "step": 42130 }, { "grad_norm": 0.25747978687286377, "learning_rate": 6.679836395822744e-05, "loss": 0.0232, "step": 42140 }, { "grad_norm": 0.22254173457622528, "learning_rate": 6.678278944333633e-05, "loss": 0.0252, "step": 42150 }, { "grad_norm": 0.21694858372211456, "learning_rate": 6.676721309310462e-05, "loss": 0.0222, "step": 42160 }, { "grad_norm": 0.23312227427959442, "learning_rate": 6.67516349092357e-05, "loss": 0.0232, "step": 42170 }, { "grad_norm": 0.2265801727771759, "learning_rate": 6.673605489343315e-05, "loss": 0.0218, "step": 42180 }, { "grad_norm": 0.18985752761363983, "learning_rate": 6.672047304740081e-05, "loss": 0.0208, "step": 42190 }, { "grad_norm": 0.23496022820472717, "learning_rate": 6.670488937284268e-05, "loss": 0.0239, "step": 42200 }, { "grad_norm": 0.20303237438201904, "learning_rate": 6.668930387146294e-05, "loss": 0.0191, "step": 42210 }, { "grad_norm": 0.16978719830513, "learning_rate": 6.667371654496605e-05, "loss": 0.0251, "step": 42220 }, { "grad_norm": 0.20945750176906586, "learning_rate": 6.665812739505659e-05, "loss": 0.025, "step": 42230 }, { "grad_norm": 0.2990019917488098, "learning_rate": 6.664253642343935e-05, "loss": 0.0216, "step": 42240 }, { "grad_norm": 0.2578694224357605, "learning_rate": 6.662694363181935e-05, "loss": 0.0254, "step": 42250 }, { "grad_norm": 0.25360429286956787, "learning_rate": 6.66113490219018e-05, "loss": 0.0255, "step": 42260 }, { "grad_norm": 0.2733902931213379, "learning_rate": 6.659575259539207e-05, "loss": 0.0214, "step": 42270 }, { "grad_norm": 0.18986129760742188, "learning_rate": 6.65801543539958e-05, "loss": 0.0267, "step": 42280 }, { "grad_norm": 0.2160511165857315, "learning_rate": 6.65645542994188e-05, "loss": 0.026, "step": 42290 }, { "grad_norm": 0.18324920535087585, "learning_rate": 6.654895243336702e-05, "loss": 0.0229, "step": 42300 }, { "grad_norm": 0.2413322478532791, "learning_rate": 6.65333487575467e-05, "loss": 0.0254, "step": 42310 }, { "grad_norm": 0.23720872402191162, "learning_rate": 6.651774327366422e-05, "loss": 0.0187, "step": 42320 }, { "grad_norm": 0.20668388903141022, "learning_rate": 6.650213598342613e-05, "loss": 0.0225, "step": 42330 }, { "grad_norm": 0.16750629246234894, "learning_rate": 6.648652688853931e-05, "loss": 0.0198, "step": 42340 }, { "grad_norm": 0.23069418966770172, "learning_rate": 6.647091599071066e-05, "loss": 0.024, "step": 42350 }, { "grad_norm": 0.24398139119148254, "learning_rate": 6.645530329164743e-05, "loss": 0.0204, "step": 42360 }, { "grad_norm": 0.30939674377441406, "learning_rate": 6.643968879305694e-05, "loss": 0.0312, "step": 42370 }, { "grad_norm": 0.26669424772262573, "learning_rate": 6.642407249664683e-05, "loss": 0.0251, "step": 42380 }, { "grad_norm": 0.2795252799987793, "learning_rate": 6.640845440412483e-05, "loss": 0.0236, "step": 42390 }, { "grad_norm": 0.20941397547721863, "learning_rate": 6.639283451719893e-05, "loss": 0.0248, "step": 42400 }, { "grad_norm": 0.27439993619918823, "learning_rate": 6.637721283757729e-05, "loss": 0.0204, "step": 42410 }, { "grad_norm": 0.1770736128091812, "learning_rate": 6.636158936696828e-05, "loss": 0.0201, "step": 42420 }, { "grad_norm": 0.21953459084033966, "learning_rate": 6.634596410708047e-05, "loss": 0.0258, "step": 42430 }, { "grad_norm": 0.2687278687953949, "learning_rate": 6.633033705962257e-05, "loss": 0.0196, "step": 42440 }, { "grad_norm": 0.29735642671585083, "learning_rate": 6.631470822630359e-05, "loss": 0.0235, "step": 42450 }, { "grad_norm": 0.2098473608493805, "learning_rate": 6.629907760883263e-05, "loss": 0.0222, "step": 42460 }, { "grad_norm": 0.1891375631093979, "learning_rate": 6.628344520891905e-05, "loss": 0.0191, "step": 42470 }, { "grad_norm": 0.24814090132713318, "learning_rate": 6.62678110282724e-05, "loss": 0.0236, "step": 42480 }, { "grad_norm": 0.21210649609565735, "learning_rate": 6.625217506860237e-05, "loss": 0.0215, "step": 42490 }, { "grad_norm": 0.22964362800121307, "learning_rate": 6.623653733161892e-05, "loss": 0.0197, "step": 42500 }, { "grad_norm": 0.21261471509933472, "learning_rate": 6.622089781903217e-05, "loss": 0.0212, "step": 42510 }, { "grad_norm": 0.2472122758626938, "learning_rate": 6.620525653255241e-05, "loss": 0.0224, "step": 42520 }, { "grad_norm": 0.1891014277935028, "learning_rate": 6.618961347389015e-05, "loss": 0.0253, "step": 42530 }, { "grad_norm": 0.19924680888652802, "learning_rate": 6.617396864475613e-05, "loss": 0.0213, "step": 42540 }, { "grad_norm": 0.27584463357925415, "learning_rate": 6.61583220468612e-05, "loss": 0.0242, "step": 42550 }, { "grad_norm": 0.27076736092567444, "learning_rate": 6.614267368191645e-05, "loss": 0.0254, "step": 42560 }, { "grad_norm": 0.23157177865505219, "learning_rate": 6.61270235516332e-05, "loss": 0.0234, "step": 42570 }, { "grad_norm": 0.26383236050605774, "learning_rate": 6.611137165772287e-05, "loss": 0.0256, "step": 42580 }, { "grad_norm": 0.3178122937679291, "learning_rate": 6.60957180018972e-05, "loss": 0.0261, "step": 42590 }, { "grad_norm": 0.18875578045845032, "learning_rate": 6.608006258586797e-05, "loss": 0.0213, "step": 42600 }, { "grad_norm": 0.18653139472007751, "learning_rate": 6.606440541134731e-05, "loss": 0.0264, "step": 42610 }, { "grad_norm": 0.2672123312950134, "learning_rate": 6.60487464800474e-05, "loss": 0.0245, "step": 42620 }, { "grad_norm": 0.2223028689622879, "learning_rate": 6.603308579368071e-05, "loss": 0.0214, "step": 42630 }, { "grad_norm": 0.25140446424484253, "learning_rate": 6.601742335395987e-05, "loss": 0.0267, "step": 42640 }, { "grad_norm": 0.2874581515789032, "learning_rate": 6.600175916259769e-05, "loss": 0.0218, "step": 42650 }, { "grad_norm": 0.258969783782959, "learning_rate": 6.598609322130718e-05, "loss": 0.0213, "step": 42660 }, { "grad_norm": 0.2175803780555725, "learning_rate": 6.597042553180155e-05, "loss": 0.0294, "step": 42670 }, { "grad_norm": 0.20880252122879028, "learning_rate": 6.59547560957942e-05, "loss": 0.021, "step": 42680 }, { "grad_norm": 0.24533642828464508, "learning_rate": 6.59390849149987e-05, "loss": 0.0221, "step": 42690 }, { "grad_norm": 0.3038634955883026, "learning_rate": 6.592341199112886e-05, "loss": 0.0264, "step": 42700 }, { "grad_norm": 0.1935223489999771, "learning_rate": 6.590773732589861e-05, "loss": 0.0219, "step": 42710 }, { "grad_norm": 0.19752486050128937, "learning_rate": 6.589206092102213e-05, "loss": 0.0218, "step": 42720 }, { "grad_norm": 0.2943305969238281, "learning_rate": 6.587638277821375e-05, "loss": 0.0205, "step": 42730 }, { "grad_norm": 0.12372338771820068, "learning_rate": 6.586070289918803e-05, "loss": 0.0221, "step": 42740 }, { "grad_norm": 0.2163742333650589, "learning_rate": 6.584502128565968e-05, "loss": 0.0217, "step": 42750 }, { "grad_norm": 0.25335803627967834, "learning_rate": 6.582933793934363e-05, "loss": 0.0199, "step": 42760 }, { "grad_norm": 0.34215763211250305, "learning_rate": 6.581365286195498e-05, "loss": 0.0245, "step": 42770 }, { "grad_norm": 0.29785728454589844, "learning_rate": 6.579796605520903e-05, "loss": 0.0228, "step": 42780 }, { "grad_norm": 0.16196675598621368, "learning_rate": 6.578227752082127e-05, "loss": 0.0196, "step": 42790 }, { "grad_norm": 0.18879815936088562, "learning_rate": 6.576658726050735e-05, "loss": 0.0233, "step": 42800 }, { "grad_norm": 0.16788850724697113, "learning_rate": 6.575089527598316e-05, "loss": 0.0241, "step": 42810 }, { "grad_norm": 0.1450549066066742, "learning_rate": 6.573520156896476e-05, "loss": 0.0203, "step": 42820 }, { "grad_norm": 0.23911578953266144, "learning_rate": 6.571950614116835e-05, "loss": 0.0232, "step": 42830 }, { "grad_norm": 0.24837203323841095, "learning_rate": 6.570380899431039e-05, "loss": 0.0216, "step": 42840 }, { "grad_norm": 0.23079994320869446, "learning_rate": 6.568811013010749e-05, "loss": 0.0238, "step": 42850 }, { "grad_norm": 0.20665417611598969, "learning_rate": 6.567240955027645e-05, "loss": 0.0221, "step": 42860 }, { "grad_norm": 0.2929147779941559, "learning_rate": 6.565670725653427e-05, "loss": 0.0224, "step": 42870 }, { "grad_norm": 0.1557670682668686, "learning_rate": 6.564100325059812e-05, "loss": 0.0213, "step": 42880 }, { "grad_norm": 0.18076550960540771, "learning_rate": 6.562529753418538e-05, "loss": 0.0241, "step": 42890 }, { "grad_norm": 0.18796178698539734, "learning_rate": 6.56095901090136e-05, "loss": 0.0235, "step": 42900 }, { "grad_norm": 0.2004653662443161, "learning_rate": 6.55938809768005e-05, "loss": 0.0248, "step": 42910 }, { "grad_norm": 0.27086931467056274, "learning_rate": 6.557817013926402e-05, "loss": 0.0197, "step": 42920 }, { "grad_norm": 0.26422399282455444, "learning_rate": 6.556245759812227e-05, "loss": 0.0228, "step": 42930 }, { "grad_norm": 0.19808487594127655, "learning_rate": 6.554674335509357e-05, "loss": 0.0208, "step": 42940 }, { "grad_norm": 0.29528674483299255, "learning_rate": 6.553102741189638e-05, "loss": 0.0214, "step": 42950 }, { "grad_norm": 0.22740653157234192, "learning_rate": 6.55153097702494e-05, "loss": 0.0229, "step": 42960 }, { "grad_norm": 0.2021261304616928, "learning_rate": 6.549959043187146e-05, "loss": 0.0192, "step": 42970 }, { "grad_norm": 0.28022974729537964, "learning_rate": 6.54838693984816e-05, "loss": 0.0225, "step": 42980 }, { "grad_norm": 0.22207152843475342, "learning_rate": 6.546814667179909e-05, "loss": 0.0229, "step": 42990 }, { "grad_norm": 0.3296186327934265, "learning_rate": 6.545242225354328e-05, "loss": 0.0207, "step": 43000 }, { "grad_norm": 0.23157146573066711, "learning_rate": 6.543669614543383e-05, "loss": 0.0177, "step": 43010 }, { "grad_norm": 0.24092082679271698, "learning_rate": 6.542096834919049e-05, "loss": 0.0196, "step": 43020 }, { "grad_norm": 0.1881721168756485, "learning_rate": 6.540523886653324e-05, "loss": 0.022, "step": 43030 }, { "grad_norm": 0.25650709867477417, "learning_rate": 6.538950769918222e-05, "loss": 0.0243, "step": 43040 }, { "grad_norm": 0.1717960238456726, "learning_rate": 6.537377484885779e-05, "loss": 0.025, "step": 43050 }, { "grad_norm": 0.2145952582359314, "learning_rate": 6.535804031728045e-05, "loss": 0.0206, "step": 43060 }, { "grad_norm": 0.2292531132698059, "learning_rate": 6.534230410617092e-05, "loss": 0.0211, "step": 43070 }, { "grad_norm": 0.22858572006225586, "learning_rate": 6.532656621725007e-05, "loss": 0.021, "step": 43080 }, { "grad_norm": 0.25844404101371765, "learning_rate": 6.531082665223898e-05, "loss": 0.0206, "step": 43090 }, { "grad_norm": 0.24297191202640533, "learning_rate": 6.529508541285889e-05, "loss": 0.0206, "step": 43100 }, { "grad_norm": 0.23482078313827515, "learning_rate": 6.527934250083128e-05, "loss": 0.0231, "step": 43110 }, { "grad_norm": 0.2193371206521988, "learning_rate": 6.526359791787775e-05, "loss": 0.0236, "step": 43120 }, { "grad_norm": 0.2193305343389511, "learning_rate": 6.524785166572009e-05, "loss": 0.0182, "step": 43130 }, { "grad_norm": 0.1854533553123474, "learning_rate": 6.523210374608029e-05, "loss": 0.0204, "step": 43140 }, { "grad_norm": 0.2286466360092163, "learning_rate": 6.521635416068054e-05, "loss": 0.0248, "step": 43150 }, { "grad_norm": 0.21905449032783508, "learning_rate": 6.520060291124317e-05, "loss": 0.0219, "step": 43160 }, { "grad_norm": 0.2951601445674896, "learning_rate": 6.518484999949072e-05, "loss": 0.0228, "step": 43170 }, { "grad_norm": 0.28678077459335327, "learning_rate": 6.516909542714591e-05, "loss": 0.0245, "step": 43180 }, { "grad_norm": 0.203409343957901, "learning_rate": 6.515333919593161e-05, "loss": 0.0243, "step": 43190 }, { "grad_norm": 0.22344987094402313, "learning_rate": 6.513758130757094e-05, "loss": 0.0233, "step": 43200 }, { "grad_norm": 0.2110714167356491, "learning_rate": 6.512182176378713e-05, "loss": 0.0217, "step": 43210 }, { "grad_norm": 0.2287951409816742, "learning_rate": 6.510606056630362e-05, "loss": 0.0213, "step": 43220 }, { "grad_norm": 0.23075300455093384, "learning_rate": 6.509029771684405e-05, "loss": 0.0197, "step": 43230 }, { "grad_norm": 0.19299066066741943, "learning_rate": 6.50745332171322e-05, "loss": 0.0258, "step": 43240 }, { "grad_norm": 0.2646755576133728, "learning_rate": 6.505876706889207e-05, "loss": 0.0241, "step": 43250 }, { "grad_norm": 0.248456209897995, "learning_rate": 6.504299927384781e-05, "loss": 0.0238, "step": 43260 }, { "grad_norm": 0.19587700068950653, "learning_rate": 6.502722983372378e-05, "loss": 0.0248, "step": 43270 }, { "grad_norm": 0.34603750705718994, "learning_rate": 6.501145875024446e-05, "loss": 0.0224, "step": 43280 }, { "grad_norm": 0.1894567608833313, "learning_rate": 6.499568602513459e-05, "loss": 0.0215, "step": 43290 }, { "grad_norm": 0.20769868791103363, "learning_rate": 6.497991166011903e-05, "loss": 0.0195, "step": 43300 }, { "grad_norm": 0.48959454894065857, "learning_rate": 6.49641356569229e-05, "loss": 0.0201, "step": 43310 }, { "grad_norm": 0.2665981650352478, "learning_rate": 6.494835801727135e-05, "loss": 0.0255, "step": 43320 }, { "grad_norm": 0.15652057528495789, "learning_rate": 6.493257874288986e-05, "loss": 0.0244, "step": 43330 }, { "grad_norm": 0.20453865826129913, "learning_rate": 6.491679783550399e-05, "loss": 0.0234, "step": 43340 }, { "grad_norm": 0.23908264935016632, "learning_rate": 6.490101529683957e-05, "loss": 0.0248, "step": 43350 }, { "grad_norm": 0.17362099885940552, "learning_rate": 6.488523112862248e-05, "loss": 0.0192, "step": 43360 }, { "grad_norm": 0.21100634336471558, "learning_rate": 6.486944533257891e-05, "loss": 0.022, "step": 43370 }, { "grad_norm": 0.18890421092510223, "learning_rate": 6.485365791043517e-05, "loss": 0.0249, "step": 43380 }, { "grad_norm": 0.1402520388364792, "learning_rate": 6.483786886391772e-05, "loss": 0.023, "step": 43390 }, { "grad_norm": 0.26524117588996887, "learning_rate": 6.482207819475323e-05, "loss": 0.0222, "step": 43400 }, { "grad_norm": 0.27474841475486755, "learning_rate": 6.480628590466857e-05, "loss": 0.0266, "step": 43410 }, { "grad_norm": 0.25002387166023254, "learning_rate": 6.479049199539074e-05, "loss": 0.0221, "step": 43420 }, { "grad_norm": 0.22616980969905853, "learning_rate": 6.477469646864693e-05, "loss": 0.0271, "step": 43430 }, { "grad_norm": 0.21044015884399414, "learning_rate": 6.475889932616454e-05, "loss": 0.0288, "step": 43440 }, { "grad_norm": 0.25571033358573914, "learning_rate": 6.474310056967111e-05, "loss": 0.0243, "step": 43450 }, { "grad_norm": 0.2404278963804245, "learning_rate": 6.472730020089437e-05, "loss": 0.026, "step": 43460 }, { "grad_norm": 0.2012513130903244, "learning_rate": 6.471149822156222e-05, "loss": 0.0216, "step": 43470 }, { "grad_norm": 0.1958148032426834, "learning_rate": 6.469569463340275e-05, "loss": 0.0199, "step": 43480 }, { "grad_norm": 0.22521111369132996, "learning_rate": 6.467988943814418e-05, "loss": 0.0226, "step": 43490 }, { "grad_norm": 0.18919630348682404, "learning_rate": 6.4664082637515e-05, "loss": 0.0196, "step": 43500 }, { "grad_norm": 0.2413119226694107, "learning_rate": 6.464827423324377e-05, "loss": 0.0192, "step": 43510 }, { "grad_norm": 0.18381717801094055, "learning_rate": 6.463246422705931e-05, "loss": 0.0215, "step": 43520 }, { "grad_norm": 0.19241146743297577, "learning_rate": 6.461665262069055e-05, "loss": 0.0217, "step": 43530 }, { "grad_norm": 0.21307392418384552, "learning_rate": 6.460083941586665e-05, "loss": 0.0248, "step": 43540 }, { "grad_norm": 0.18729346990585327, "learning_rate": 6.45850246143169e-05, "loss": 0.0185, "step": 43550 }, { "grad_norm": 0.2015230357646942, "learning_rate": 6.456920821777077e-05, "loss": 0.0227, "step": 43560 }, { "grad_norm": 0.28256717324256897, "learning_rate": 6.455339022795796e-05, "loss": 0.0245, "step": 43570 }, { "grad_norm": 0.23653045296669006, "learning_rate": 6.453757064660825e-05, "loss": 0.0193, "step": 43580 }, { "grad_norm": 0.16841688752174377, "learning_rate": 6.452174947545169e-05, "loss": 0.0225, "step": 43590 }, { "grad_norm": 0.2319643348455429, "learning_rate": 6.450592671621842e-05, "loss": 0.0203, "step": 43600 }, { "grad_norm": 0.18208111822605133, "learning_rate": 6.449010237063882e-05, "loss": 0.0211, "step": 43610 }, { "grad_norm": 0.3260903060436249, "learning_rate": 6.447427644044343e-05, "loss": 0.0209, "step": 43620 }, { "grad_norm": 0.27042141556739807, "learning_rate": 6.44584489273629e-05, "loss": 0.0233, "step": 43630 }, { "grad_norm": 0.20257337391376495, "learning_rate": 6.444261983312817e-05, "loss": 0.022, "step": 43640 }, { "grad_norm": 0.21991217136383057, "learning_rate": 6.442678915947023e-05, "loss": 0.022, "step": 43650 }, { "grad_norm": 0.3189152479171753, "learning_rate": 6.441095690812032e-05, "loss": 0.0205, "step": 43660 }, { "grad_norm": 0.26213398575782776, "learning_rate": 6.439512308080982e-05, "loss": 0.0229, "step": 43670 }, { "grad_norm": 0.3224998414516449, "learning_rate": 6.43792876792703e-05, "loss": 0.0239, "step": 43680 }, { "grad_norm": 0.26647356152534485, "learning_rate": 6.43634507052335e-05, "loss": 0.0252, "step": 43690 }, { "grad_norm": 0.2062014490365982, "learning_rate": 6.434761216043135e-05, "loss": 0.0241, "step": 43700 }, { "grad_norm": 0.18340900540351868, "learning_rate": 6.433177204659589e-05, "loss": 0.0229, "step": 43710 }, { "grad_norm": 0.2894969582557678, "learning_rate": 6.431593036545938e-05, "loss": 0.02, "step": 43720 }, { "grad_norm": 0.21357740461826324, "learning_rate": 6.430008711875426e-05, "loss": 0.0211, "step": 43730 }, { "grad_norm": 0.2161114662885666, "learning_rate": 6.42842423082131e-05, "loss": 0.0221, "step": 43740 }, { "grad_norm": 0.17865093052387238, "learning_rate": 6.426839593556868e-05, "loss": 0.0202, "step": 43750 }, { "grad_norm": 0.1938420981168747, "learning_rate": 6.425254800255392e-05, "loss": 0.0189, "step": 43760 }, { "grad_norm": 0.17973484098911285, "learning_rate": 6.423669851090196e-05, "loss": 0.0206, "step": 43770 }, { "grad_norm": 0.22626210749149323, "learning_rate": 6.422084746234604e-05, "loss": 0.0226, "step": 43780 }, { "grad_norm": 0.25697746872901917, "learning_rate": 6.420499485861965e-05, "loss": 0.0229, "step": 43790 }, { "grad_norm": 0.19145193696022034, "learning_rate": 6.418914070145636e-05, "loss": 0.0225, "step": 43800 }, { "grad_norm": 0.20528525114059448, "learning_rate": 6.417328499258996e-05, "loss": 0.0199, "step": 43810 }, { "grad_norm": 0.22320310771465302, "learning_rate": 6.415742773375446e-05, "loss": 0.0207, "step": 43820 }, { "grad_norm": 0.2132444828748703, "learning_rate": 6.414156892668393e-05, "loss": 0.0175, "step": 43830 }, { "grad_norm": 0.23377321660518646, "learning_rate": 6.412570857311267e-05, "loss": 0.0221, "step": 43840 }, { "grad_norm": 0.25259795784950256, "learning_rate": 6.410984667477518e-05, "loss": 0.0229, "step": 43850 }, { "grad_norm": 0.21140941977500916, "learning_rate": 6.409398323340607e-05, "loss": 0.0215, "step": 43860 }, { "grad_norm": 0.30823037028312683, "learning_rate": 6.407811825074012e-05, "loss": 0.0227, "step": 43870 }, { "grad_norm": 0.23294594883918762, "learning_rate": 6.406225172851234e-05, "loss": 0.0221, "step": 43880 }, { "grad_norm": 0.17619727551937103, "learning_rate": 6.404638366845786e-05, "loss": 0.0196, "step": 43890 }, { "grad_norm": 0.22917050123214722, "learning_rate": 6.403051407231196e-05, "loss": 0.0216, "step": 43900 }, { "grad_norm": 0.19290472567081451, "learning_rate": 6.401464294181016e-05, "loss": 0.0216, "step": 43910 }, { "grad_norm": 0.18659870326519012, "learning_rate": 6.399877027868808e-05, "loss": 0.0222, "step": 43920 }, { "grad_norm": 0.2510974705219269, "learning_rate": 6.39828960846815e-05, "loss": 0.0262, "step": 43930 }, { "grad_norm": 0.18436631560325623, "learning_rate": 6.396702036152644e-05, "loss": 0.0232, "step": 43940 }, { "grad_norm": 0.17825709283351898, "learning_rate": 6.395114311095905e-05, "loss": 0.0264, "step": 43950 }, { "grad_norm": 0.24020174145698547, "learning_rate": 6.393526433471559e-05, "loss": 0.0241, "step": 43960 }, { "grad_norm": 0.21427159011363983, "learning_rate": 6.391938403453262e-05, "loss": 0.0198, "step": 43970 }, { "grad_norm": 0.173656165599823, "learning_rate": 6.390350221214671e-05, "loss": 0.0201, "step": 43980 }, { "grad_norm": 0.21018442511558533, "learning_rate": 6.38876188692947e-05, "loss": 0.0175, "step": 43990 }, { "grad_norm": 0.27160489559173584, "learning_rate": 6.387173400771358e-05, "loss": 0.0197, "step": 44000 }, { "grad_norm": 0.24937109649181366, "learning_rate": 6.385584762914047e-05, "loss": 0.0233, "step": 44010 }, { "grad_norm": 0.1872449517250061, "learning_rate": 6.38399597353127e-05, "loss": 0.018, "step": 44020 }, { "grad_norm": 0.1934531331062317, "learning_rate": 6.382407032796775e-05, "loss": 0.0215, "step": 44030 }, { "grad_norm": 0.21374410390853882, "learning_rate": 6.380817940884325e-05, "loss": 0.0227, "step": 44040 }, { "grad_norm": 0.1904640942811966, "learning_rate": 6.379228697967702e-05, "loss": 0.021, "step": 44050 }, { "grad_norm": 0.24132545292377472, "learning_rate": 6.3776393042207e-05, "loss": 0.02, "step": 44060 }, { "grad_norm": 0.28738781809806824, "learning_rate": 6.376049759817137e-05, "loss": 0.0233, "step": 44070 }, { "grad_norm": 0.20328299701213837, "learning_rate": 6.37446006493084e-05, "loss": 0.0209, "step": 44080 }, { "grad_norm": 0.16686375439167023, "learning_rate": 6.372870219735656e-05, "loss": 0.0193, "step": 44090 }, { "grad_norm": 0.21618866920471191, "learning_rate": 6.371280224405452e-05, "loss": 0.0215, "step": 44100 }, { "grad_norm": 0.3504765033721924, "learning_rate": 6.369690079114101e-05, "loss": 0.019, "step": 44110 }, { "grad_norm": 0.21547123789787292, "learning_rate": 6.368099784035504e-05, "loss": 0.0226, "step": 44120 }, { "grad_norm": 0.24279248714447021, "learning_rate": 6.366509339343572e-05, "loss": 0.0236, "step": 44130 }, { "grad_norm": 0.23297998309135437, "learning_rate": 6.364918745212232e-05, "loss": 0.0204, "step": 44140 }, { "grad_norm": 0.21874955296516418, "learning_rate": 6.363328001815431e-05, "loss": 0.0215, "step": 44150 }, { "grad_norm": 0.20613384246826172, "learning_rate": 6.361737109327128e-05, "loss": 0.0205, "step": 44160 }, { "grad_norm": 0.2756938934326172, "learning_rate": 6.360146067921301e-05, "loss": 0.0212, "step": 44170 }, { "grad_norm": 0.3152322769165039, "learning_rate": 6.35855487777195e-05, "loss": 0.0227, "step": 44180 }, { "grad_norm": 0.193433940410614, "learning_rate": 6.356963539053076e-05, "loss": 0.0248, "step": 44190 }, { "grad_norm": 0.2720535397529602, "learning_rate": 6.355372051938711e-05, "loss": 0.0242, "step": 44200 }, { "grad_norm": 0.1830073744058609, "learning_rate": 6.353780416602894e-05, "loss": 0.0186, "step": 44210 }, { "grad_norm": 0.22349965572357178, "learning_rate": 6.352188633219689e-05, "loss": 0.023, "step": 44220 }, { "grad_norm": 0.20449379086494446, "learning_rate": 6.350596701963166e-05, "loss": 0.0229, "step": 44230 }, { "grad_norm": 0.24200871586799622, "learning_rate": 6.349004623007419e-05, "loss": 0.0211, "step": 44240 }, { "grad_norm": 0.18009325861930847, "learning_rate": 6.347412396526555e-05, "loss": 0.0192, "step": 44250 }, { "grad_norm": 0.2146432250738144, "learning_rate": 6.345820022694696e-05, "loss": 0.0266, "step": 44260 }, { "grad_norm": 0.18105463683605194, "learning_rate": 6.344227501685984e-05, "loss": 0.0197, "step": 44270 }, { "grad_norm": 0.3480710983276367, "learning_rate": 6.342634833674572e-05, "loss": 0.022, "step": 44280 }, { "grad_norm": 0.29981645941734314, "learning_rate": 6.341042018834635e-05, "loss": 0.0205, "step": 44290 }, { "grad_norm": 0.2411905825138092, "learning_rate": 6.339449057340359e-05, "loss": 0.0179, "step": 44300 }, { "grad_norm": 0.19546754658222198, "learning_rate": 6.337855949365945e-05, "loss": 0.02, "step": 44310 }, { "grad_norm": 0.2257654219865799, "learning_rate": 6.336262695085619e-05, "loss": 0.0201, "step": 44320 }, { "grad_norm": 0.1653570681810379, "learning_rate": 6.334669294673612e-05, "loss": 0.02, "step": 44330 }, { "grad_norm": 0.20268607139587402, "learning_rate": 6.333075748304179e-05, "loss": 0.023, "step": 44340 }, { "grad_norm": 0.1306125372648239, "learning_rate": 6.331482056151585e-05, "loss": 0.0182, "step": 44350 }, { "grad_norm": 0.22269515693187714, "learning_rate": 6.329888218390117e-05, "loss": 0.0241, "step": 44360 }, { "grad_norm": 0.2290686070919037, "learning_rate": 6.328294235194072e-05, "loss": 0.0214, "step": 44370 }, { "grad_norm": 0.23321601748466492, "learning_rate": 6.326700106737765e-05, "loss": 0.0207, "step": 44380 }, { "grad_norm": 0.18830129504203796, "learning_rate": 6.325105833195531e-05, "loss": 0.0208, "step": 44390 }, { "grad_norm": 0.21265172958374023, "learning_rate": 6.323511414741715e-05, "loss": 0.0221, "step": 44400 }, { "grad_norm": 0.2390899956226349, "learning_rate": 6.321916851550678e-05, "loss": 0.0225, "step": 44410 }, { "grad_norm": 0.18616323173046112, "learning_rate": 6.320322143796806e-05, "loss": 0.0219, "step": 44420 }, { "grad_norm": 0.24267831444740295, "learning_rate": 6.318727291654488e-05, "loss": 0.0228, "step": 44430 }, { "grad_norm": 0.2740359902381897, "learning_rate": 6.317132295298134e-05, "loss": 0.0247, "step": 44440 }, { "grad_norm": 0.31140708923339844, "learning_rate": 6.315537154902173e-05, "loss": 0.0208, "step": 44450 }, { "grad_norm": 0.15871110558509827, "learning_rate": 6.313941870641048e-05, "loss": 0.0201, "step": 44460 }, { "grad_norm": 0.20215626060962677, "learning_rate": 6.312346442689214e-05, "loss": 0.0179, "step": 44470 }, { "grad_norm": 0.15793101489543915, "learning_rate": 6.310750871221147e-05, "loss": 0.0209, "step": 44480 }, { "grad_norm": 0.18490085005760193, "learning_rate": 6.309155156411335e-05, "loss": 0.0228, "step": 44490 }, { "grad_norm": 0.21492613852024078, "learning_rate": 6.307559298434284e-05, "loss": 0.0293, "step": 44500 }, { "grad_norm": 0.2322075366973877, "learning_rate": 6.305963297464512e-05, "loss": 0.0262, "step": 44510 }, { "grad_norm": 0.2294325828552246, "learning_rate": 6.304367153676561e-05, "loss": 0.0188, "step": 44520 }, { "grad_norm": 0.2230655997991562, "learning_rate": 6.302770867244978e-05, "loss": 0.0216, "step": 44530 }, { "grad_norm": 0.1575051248073578, "learning_rate": 6.301174438344328e-05, "loss": 0.0184, "step": 44540 }, { "grad_norm": 0.20040389895439148, "learning_rate": 6.299577867149202e-05, "loss": 0.0208, "step": 44550 }, { "grad_norm": 0.18437109887599945, "learning_rate": 6.297981153834192e-05, "loss": 0.0218, "step": 44560 }, { "grad_norm": 0.17232654988765717, "learning_rate": 6.296384298573916e-05, "loss": 0.0206, "step": 44570 }, { "grad_norm": 0.20693065226078033, "learning_rate": 6.294787301543001e-05, "loss": 0.0187, "step": 44580 }, { "grad_norm": 0.18954883515834808, "learning_rate": 6.293190162916095e-05, "loss": 0.0193, "step": 44590 }, { "grad_norm": 0.2587721645832062, "learning_rate": 6.291592882867855e-05, "loss": 0.0192, "step": 44600 }, { "grad_norm": 0.18756307661533356, "learning_rate": 6.28999546157296e-05, "loss": 0.023, "step": 44610 }, { "grad_norm": 0.2572883069515228, "learning_rate": 6.288397899206102e-05, "loss": 0.0221, "step": 44620 }, { "grad_norm": 0.23222804069519043, "learning_rate": 6.286800195941984e-05, "loss": 0.0249, "step": 44630 }, { "grad_norm": 0.19782213866710663, "learning_rate": 6.285202351955334e-05, "loss": 0.022, "step": 44640 }, { "grad_norm": 0.1893714815378189, "learning_rate": 6.283604367420887e-05, "loss": 0.023, "step": 44650 }, { "grad_norm": 0.3173452913761139, "learning_rate": 6.282006242513394e-05, "loss": 0.0259, "step": 44660 }, { "grad_norm": 0.24543441832065582, "learning_rate": 6.280407977407628e-05, "loss": 0.0233, "step": 44670 }, { "grad_norm": 0.23103375732898712, "learning_rate": 6.27880957227837e-05, "loss": 0.0225, "step": 44680 }, { "grad_norm": 0.2442137449979782, "learning_rate": 6.27721102730042e-05, "loss": 0.0188, "step": 44690 }, { "grad_norm": 0.16995856165885925, "learning_rate": 6.275612342648592e-05, "loss": 0.0196, "step": 44700 }, { "grad_norm": 0.21291254460811615, "learning_rate": 6.274013518497716e-05, "loss": 0.0197, "step": 44710 }, { "grad_norm": 0.23902080953121185, "learning_rate": 6.272414555022636e-05, "loss": 0.0223, "step": 44720 }, { "grad_norm": 0.19720648229122162, "learning_rate": 6.270815452398215e-05, "loss": 0.0211, "step": 44730 }, { "grad_norm": 0.22618655860424042, "learning_rate": 6.269216210799326e-05, "loss": 0.0249, "step": 44740 }, { "grad_norm": 0.22651462256908417, "learning_rate": 6.26761683040086e-05, "loss": 0.0236, "step": 44750 }, { "grad_norm": 0.20288647711277008, "learning_rate": 6.266017311377723e-05, "loss": 0.0233, "step": 44760 }, { "grad_norm": 0.23697532713413239, "learning_rate": 6.264417653904839e-05, "loss": 0.0257, "step": 44770 }, { "grad_norm": 0.2447235882282257, "learning_rate": 6.262817858157139e-05, "loss": 0.0237, "step": 44780 }, { "grad_norm": 0.26497504115104675, "learning_rate": 6.261217924309576e-05, "loss": 0.0227, "step": 44790 }, { "grad_norm": 0.21580012142658234, "learning_rate": 6.259617852537118e-05, "loss": 0.0271, "step": 44800 }, { "grad_norm": 0.1970006227493286, "learning_rate": 6.258017643014747e-05, "loss": 0.0221, "step": 44810 }, { "grad_norm": 0.18114754557609558, "learning_rate": 6.256417295917456e-05, "loss": 0.0199, "step": 44820 }, { "grad_norm": 0.2322908490896225, "learning_rate": 6.254816811420258e-05, "loss": 0.0271, "step": 44830 }, { "grad_norm": 0.27782997488975525, "learning_rate": 6.253216189698183e-05, "loss": 0.0237, "step": 44840 }, { "grad_norm": 0.20005260407924652, "learning_rate": 6.251615430926267e-05, "loss": 0.0212, "step": 44850 }, { "grad_norm": 0.21450836956501007, "learning_rate": 6.25001453527957e-05, "loss": 0.023, "step": 44860 }, { "grad_norm": 0.23909994959831238, "learning_rate": 6.248413502933164e-05, "loss": 0.0261, "step": 44870 }, { "grad_norm": 0.2428053766489029, "learning_rate": 6.246812334062133e-05, "loss": 0.0219, "step": 44880 }, { "grad_norm": 0.29903972148895264, "learning_rate": 6.245211028841579e-05, "loss": 0.0248, "step": 44890 }, { "grad_norm": 0.179621160030365, "learning_rate": 6.24360958744662e-05, "loss": 0.0182, "step": 44900 }, { "grad_norm": 0.2631463408470154, "learning_rate": 6.242008010052387e-05, "loss": 0.0203, "step": 44910 }, { "grad_norm": 0.24422946572303772, "learning_rate": 6.240406296834024e-05, "loss": 0.022, "step": 44920 }, { "grad_norm": 0.26738467812538147, "learning_rate": 6.238804447966694e-05, "loss": 0.0216, "step": 44930 }, { "grad_norm": 0.22716465592384338, "learning_rate": 6.237202463625573e-05, "loss": 0.0211, "step": 44940 }, { "grad_norm": 0.1841079145669937, "learning_rate": 6.235600343985848e-05, "loss": 0.0196, "step": 44950 }, { "grad_norm": 0.1996650993824005, "learning_rate": 6.233998089222729e-05, "loss": 0.0208, "step": 44960 }, { "grad_norm": 0.17763076722621918, "learning_rate": 6.232395699511433e-05, "loss": 0.0196, "step": 44970 }, { "grad_norm": 0.17771250009536743, "learning_rate": 6.230793175027199e-05, "loss": 0.0201, "step": 44980 }, { "grad_norm": 0.22537651658058167, "learning_rate": 6.22919051594527e-05, "loss": 0.0251, "step": 44990 }, { "grad_norm": 0.22825241088867188, "learning_rate": 6.227587722440917e-05, "loss": 0.0208, "step": 45000 }, { "grad_norm": 0.2383386790752411, "learning_rate": 6.225984794689414e-05, "loss": 0.0236, "step": 45010 }, { "grad_norm": 0.2053341418504715, "learning_rate": 6.22438173286606e-05, "loss": 0.0217, "step": 45020 }, { "grad_norm": 0.271251380443573, "learning_rate": 6.222778537146159e-05, "loss": 0.0243, "step": 45030 }, { "grad_norm": 0.182183638215065, "learning_rate": 6.221175207705037e-05, "loss": 0.018, "step": 45040 }, { "grad_norm": 0.15066690742969513, "learning_rate": 6.219571744718027e-05, "loss": 0.0195, "step": 45050 }, { "grad_norm": 0.14285537600517273, "learning_rate": 6.217968148360487e-05, "loss": 0.0252, "step": 45060 }, { "grad_norm": 0.20818154513835907, "learning_rate": 6.216364418807782e-05, "loss": 0.0239, "step": 45070 }, { "grad_norm": 0.23243863880634308, "learning_rate": 6.214760556235292e-05, "loss": 0.0204, "step": 45080 }, { "grad_norm": 0.21621853113174438, "learning_rate": 6.213156560818414e-05, "loss": 0.0248, "step": 45090 }, { "grad_norm": 0.229569673538208, "learning_rate": 6.211552432732559e-05, "loss": 0.0197, "step": 45100 }, { "grad_norm": 0.2164597511291504, "learning_rate": 6.20994817215315e-05, "loss": 0.0205, "step": 45110 }, { "grad_norm": 0.16420474648475647, "learning_rate": 6.208343779255628e-05, "loss": 0.0191, "step": 45120 }, { "grad_norm": 0.1831391453742981, "learning_rate": 6.206739254215449e-05, "loss": 0.0193, "step": 45130 }, { "grad_norm": 0.19732250273227692, "learning_rate": 6.205134597208077e-05, "loss": 0.0173, "step": 45140 }, { "grad_norm": 0.1827966421842575, "learning_rate": 6.203529808409e-05, "loss": 0.017, "step": 45150 }, { "grad_norm": 0.20979216694831848, "learning_rate": 6.201924887993712e-05, "loss": 0.0181, "step": 45160 }, { "grad_norm": 0.16292287409305573, "learning_rate": 6.200319836137725e-05, "loss": 0.0192, "step": 45170 }, { "grad_norm": 0.11144399642944336, "learning_rate": 6.198714653016565e-05, "loss": 0.0187, "step": 45180 }, { "grad_norm": 0.20942790806293488, "learning_rate": 6.197109338805774e-05, "loss": 0.0208, "step": 45190 }, { "grad_norm": 0.2981693744659424, "learning_rate": 6.195503893680903e-05, "loss": 0.0231, "step": 45200 }, { "grad_norm": 0.2668697237968445, "learning_rate": 6.193898317817524e-05, "loss": 0.0181, "step": 45210 }, { "grad_norm": 0.16158796846866608, "learning_rate": 6.192292611391221e-05, "loss": 0.0229, "step": 45220 }, { "grad_norm": 0.18745538592338562, "learning_rate": 6.190686774577591e-05, "loss": 0.0184, "step": 45230 }, { "grad_norm": 0.18975284695625305, "learning_rate": 6.189080807552245e-05, "loss": 0.0186, "step": 45240 }, { "grad_norm": 0.22964566946029663, "learning_rate": 6.187474710490809e-05, "loss": 0.0219, "step": 45250 }, { "grad_norm": 0.21888726949691772, "learning_rate": 6.185868483568926e-05, "loss": 0.022, "step": 45260 }, { "grad_norm": 0.20929747819900513, "learning_rate": 6.184262126962245e-05, "loss": 0.0222, "step": 45270 }, { "grad_norm": 0.2712996006011963, "learning_rate": 6.182655640846442e-05, "loss": 0.0224, "step": 45280 }, { "grad_norm": 0.23000116646289825, "learning_rate": 6.181049025397196e-05, "loss": 0.0253, "step": 45290 }, { "grad_norm": 0.2920461893081665, "learning_rate": 6.179442280790202e-05, "loss": 0.0221, "step": 45300 }, { "grad_norm": 0.24269340932369232, "learning_rate": 6.177835407201174e-05, "loss": 0.0226, "step": 45310 }, { "grad_norm": 0.20712825655937195, "learning_rate": 6.176228404805839e-05, "loss": 0.0197, "step": 45320 }, { "grad_norm": 0.2204979658126831, "learning_rate": 6.174621273779932e-05, "loss": 0.0224, "step": 45330 }, { "grad_norm": 0.19595986604690552, "learning_rate": 6.17301401429921e-05, "loss": 0.0226, "step": 45340 }, { "grad_norm": 0.23039913177490234, "learning_rate": 6.171406626539436e-05, "loss": 0.0187, "step": 45350 }, { "grad_norm": 0.21893399953842163, "learning_rate": 6.169799110676398e-05, "loss": 0.0224, "step": 45360 }, { "grad_norm": 0.2731306254863739, "learning_rate": 6.168191466885885e-05, "loss": 0.023, "step": 45370 }, { "grad_norm": 0.22619540989398956, "learning_rate": 6.166583695343711e-05, "loss": 0.0195, "step": 45380 }, { "grad_norm": 0.24351799488067627, "learning_rate": 6.164975796225698e-05, "loss": 0.0231, "step": 45390 }, { "grad_norm": 0.24579906463623047, "learning_rate": 6.163367769707683e-05, "loss": 0.0206, "step": 45400 }, { "grad_norm": 0.3246658146381378, "learning_rate": 6.161759615965519e-05, "loss": 0.0239, "step": 45410 }, { "grad_norm": 0.2486550509929657, "learning_rate": 6.16015133517507e-05, "loss": 0.0239, "step": 45420 }, { "grad_norm": 0.26512765884399414, "learning_rate": 6.158542927512214e-05, "loss": 0.0234, "step": 45430 }, { "grad_norm": 0.23296652734279633, "learning_rate": 6.156934393152846e-05, "loss": 0.0185, "step": 45440 }, { "grad_norm": 0.22486697137355804, "learning_rate": 6.15532573227287e-05, "loss": 0.0219, "step": 45450 }, { "grad_norm": 0.28936198353767395, "learning_rate": 6.153716945048212e-05, "loss": 0.024, "step": 45460 }, { "grad_norm": 0.1826363205909729, "learning_rate": 6.152108031654802e-05, "loss": 0.0231, "step": 45470 }, { "grad_norm": 0.20081186294555664, "learning_rate": 6.15049899226859e-05, "loss": 0.02, "step": 45480 }, { "grad_norm": 0.19468452036380768, "learning_rate": 6.148889827065537e-05, "loss": 0.0246, "step": 45490 }, { "grad_norm": 0.27889227867126465, "learning_rate": 6.147280536221622e-05, "loss": 0.0183, "step": 45500 }, { "grad_norm": 0.29775023460388184, "learning_rate": 6.145671119912832e-05, "loss": 0.0197, "step": 45510 }, { "grad_norm": 0.2487950176000595, "learning_rate": 6.144061578315169e-05, "loss": 0.0223, "step": 45520 }, { "grad_norm": 0.19414956867694855, "learning_rate": 6.142451911604654e-05, "loss": 0.0249, "step": 45530 }, { "grad_norm": 0.2113180011510849, "learning_rate": 6.140842119957315e-05, "loss": 0.0202, "step": 45540 }, { "grad_norm": 0.17700879275798798, "learning_rate": 6.139232203549197e-05, "loss": 0.0204, "step": 45550 }, { "grad_norm": 0.2185465693473816, "learning_rate": 6.137622162556357e-05, "loss": 0.0208, "step": 45560 }, { "grad_norm": 0.18558822572231293, "learning_rate": 6.13601199715487e-05, "loss": 0.0202, "step": 45570 }, { "grad_norm": 0.208497554063797, "learning_rate": 6.134401707520816e-05, "loss": 0.0228, "step": 45580 }, { "grad_norm": 0.2030297964811325, "learning_rate": 6.132791293830298e-05, "loss": 0.0201, "step": 45590 }, { "grad_norm": 0.242881640791893, "learning_rate": 6.131180756259428e-05, "loss": 0.0216, "step": 45600 }, { "grad_norm": 0.22470144927501678, "learning_rate": 6.129570094984331e-05, "loss": 0.023, "step": 45610 }, { "grad_norm": 0.1430220901966095, "learning_rate": 6.127959310181145e-05, "loss": 0.0255, "step": 45620 }, { "grad_norm": 0.26532605290412903, "learning_rate": 6.126348402026026e-05, "loss": 0.0188, "step": 45630 }, { "grad_norm": 0.19247542321681976, "learning_rate": 6.12473737069514e-05, "loss": 0.0196, "step": 45640 }, { "grad_norm": 0.21628771722316742, "learning_rate": 6.123126216364665e-05, "loss": 0.0212, "step": 45650 }, { "grad_norm": 0.24364621937274933, "learning_rate": 6.121514939210797e-05, "loss": 0.0233, "step": 45660 }, { "grad_norm": 0.1849154680967331, "learning_rate": 6.119903539409741e-05, "loss": 0.0215, "step": 45670 }, { "grad_norm": 0.1896241456270218, "learning_rate": 6.118292017137716e-05, "loss": 0.0178, "step": 45680 }, { "grad_norm": 0.2033589482307434, "learning_rate": 6.116680372570959e-05, "loss": 0.0242, "step": 45690 }, { "grad_norm": 0.21415860950946808, "learning_rate": 6.115068605885713e-05, "loss": 0.0231, "step": 45700 }, { "grad_norm": 0.2716216742992401, "learning_rate": 6.113456717258243e-05, "loss": 0.023, "step": 45710 }, { "grad_norm": 0.15157942473888397, "learning_rate": 6.11184470686482e-05, "loss": 0.0217, "step": 45720 }, { "grad_norm": 0.16620691120624542, "learning_rate": 6.110232574881731e-05, "loss": 0.0207, "step": 45730 }, { "grad_norm": 0.2717105448246002, "learning_rate": 6.108620321485277e-05, "loss": 0.0228, "step": 45740 }, { "grad_norm": 0.28997692465782166, "learning_rate": 6.107007946851773e-05, "loss": 0.0204, "step": 45750 }, { "grad_norm": 0.19578896462917328, "learning_rate": 6.105395451157542e-05, "loss": 0.02, "step": 45760 }, { "grad_norm": 0.23156508803367615, "learning_rate": 6.103782834578928e-05, "loss": 0.0209, "step": 45770 }, { "grad_norm": 0.2581489682197571, "learning_rate": 6.102170097292281e-05, "loss": 0.0204, "step": 45780 }, { "grad_norm": 0.23085401952266693, "learning_rate": 6.1005572394739716e-05, "loss": 0.0222, "step": 45790 }, { "grad_norm": 0.2300877720117569, "learning_rate": 6.0989442613003765e-05, "loss": 0.024, "step": 45800 }, { "grad_norm": 0.196931853890419, "learning_rate": 6.097331162947888e-05, "loss": 0.0243, "step": 45810 }, { "grad_norm": 0.18340089917182922, "learning_rate": 6.095717944592914e-05, "loss": 0.0225, "step": 45820 }, { "grad_norm": 0.20702287554740906, "learning_rate": 6.094104606411873e-05, "loss": 0.0191, "step": 45830 }, { "grad_norm": 0.19624735414981842, "learning_rate": 6.0924911485811966e-05, "loss": 0.0251, "step": 45840 }, { "grad_norm": 0.1673271507024765, "learning_rate": 6.090877571277331e-05, "loss": 0.02, "step": 45850 }, { "grad_norm": 0.16998374462127686, "learning_rate": 6.089263874676736e-05, "loss": 0.0205, "step": 45860 }, { "grad_norm": 0.17302913963794708, "learning_rate": 6.0876500589558796e-05, "loss": 0.021, "step": 45870 }, { "grad_norm": 0.2525324523448944, "learning_rate": 6.086036124291248e-05, "loss": 0.0214, "step": 45880 }, { "grad_norm": 0.19050738215446472, "learning_rate": 6.084422070859339e-05, "loss": 0.022, "step": 45890 }, { "grad_norm": 0.18883150815963745, "learning_rate": 6.082807898836663e-05, "loss": 0.023, "step": 45900 }, { "grad_norm": 0.3127208650112152, "learning_rate": 6.081193608399742e-05, "loss": 0.0221, "step": 45910 }, { "grad_norm": 0.20586040616035461, "learning_rate": 6.0795791997251164e-05, "loss": 0.0217, "step": 45920 }, { "grad_norm": 0.1703265905380249, "learning_rate": 6.0779646729893294e-05, "loss": 0.0206, "step": 45930 }, { "grad_norm": 0.25181642174720764, "learning_rate": 6.0763500283689476e-05, "loss": 0.0216, "step": 45940 }, { "grad_norm": 0.26187485456466675, "learning_rate": 6.0747352660405455e-05, "loss": 0.0212, "step": 45950 }, { "grad_norm": 0.19020499289035797, "learning_rate": 6.073120386180709e-05, "loss": 0.0189, "step": 45960 }, { "grad_norm": 0.19958709180355072, "learning_rate": 6.0715053889660425e-05, "loss": 0.0188, "step": 45970 }, { "grad_norm": 0.19130881130695343, "learning_rate": 6.069890274573157e-05, "loss": 0.0175, "step": 45980 }, { "grad_norm": 0.16061343252658844, "learning_rate": 6.068275043178679e-05, "loss": 0.0244, "step": 45990 }, { "grad_norm": 0.18466225266456604, "learning_rate": 6.066659694959248e-05, "loss": 0.0209, "step": 46000 }, { "grad_norm": 0.16529983282089233, "learning_rate": 6.0650442300915176e-05, "loss": 0.0195, "step": 46010 }, { "grad_norm": 0.2225680947303772, "learning_rate": 6.063428648752152e-05, "loss": 0.0193, "step": 46020 }, { "grad_norm": 0.18398188054561615, "learning_rate": 6.0618129511178266e-05, "loss": 0.0213, "step": 46030 }, { "grad_norm": 0.2470063418149948, "learning_rate": 6.060197137365234e-05, "loss": 0.0193, "step": 46040 }, { "grad_norm": 0.21772867441177368, "learning_rate": 6.058581207671077e-05, "loss": 0.0186, "step": 46050 }, { "grad_norm": 0.21226249635219574, "learning_rate": 6.056965162212072e-05, "loss": 0.0193, "step": 46060 }, { "grad_norm": 0.22945737838745117, "learning_rate": 6.055349001164943e-05, "loss": 0.0205, "step": 46070 }, { "grad_norm": 0.21533650159835815, "learning_rate": 6.0537327247064347e-05, "loss": 0.0195, "step": 46080 }, { "grad_norm": 0.2998642027378082, "learning_rate": 6.0521163330133014e-05, "loss": 0.019, "step": 46090 }, { "grad_norm": 0.19868098199367523, "learning_rate": 6.050499826262306e-05, "loss": 0.0199, "step": 46100 }, { "grad_norm": 0.23634999990463257, "learning_rate": 6.0488832046302294e-05, "loss": 0.016, "step": 46110 }, { "grad_norm": 0.23018400371074677, "learning_rate": 6.0472664682938626e-05, "loss": 0.0257, "step": 46120 }, { "grad_norm": 0.2086435854434967, "learning_rate": 6.045649617430009e-05, "loss": 0.0211, "step": 46130 }, { "grad_norm": 0.21482928097248077, "learning_rate": 6.0440326522154866e-05, "loss": 0.0206, "step": 46140 }, { "grad_norm": 0.19110184907913208, "learning_rate": 6.0424155728271224e-05, "loss": 0.0173, "step": 46150 }, { "grad_norm": 0.20241691172122955, "learning_rate": 6.040798379441758e-05, "loss": 0.0232, "step": 46160 }, { "grad_norm": 0.15486939251422882, "learning_rate": 6.0391810722362485e-05, "loss": 0.0211, "step": 46170 }, { "grad_norm": 0.2502240836620331, "learning_rate": 6.037563651387458e-05, "loss": 0.025, "step": 46180 }, { "grad_norm": 0.20903806388378143, "learning_rate": 6.0359461170722666e-05, "loss": 0.0193, "step": 46190 }, { "grad_norm": 0.24448198080062866, "learning_rate": 6.034328469467566e-05, "loss": 0.0207, "step": 46200 }, { "grad_norm": 0.2296837866306305, "learning_rate": 6.0327107087502596e-05, "loss": 0.0228, "step": 46210 }, { "grad_norm": 0.1601477414369583, "learning_rate": 6.031092835097262e-05, "loss": 0.0192, "step": 46220 }, { "grad_norm": 0.2003832459449768, "learning_rate": 6.0294748486855024e-05, "loss": 0.0191, "step": 46230 }, { "grad_norm": 0.24135194718837738, "learning_rate": 6.0278567496919216e-05, "loss": 0.02, "step": 46240 }, { "grad_norm": 0.16732726991176605, "learning_rate": 6.026238538293472e-05, "loss": 0.0194, "step": 46250 }, { "grad_norm": 0.15820442140102386, "learning_rate": 6.024620214667118e-05, "loss": 0.0188, "step": 46260 }, { "grad_norm": 0.20405541360378265, "learning_rate": 6.0230017789898384e-05, "loss": 0.0193, "step": 46270 }, { "grad_norm": 0.17587481439113617, "learning_rate": 6.021383231438622e-05, "loss": 0.0205, "step": 46280 }, { "grad_norm": 0.33035022020339966, "learning_rate": 6.0197645721904704e-05, "loss": 0.0176, "step": 46290 }, { "grad_norm": 0.25199711322784424, "learning_rate": 6.0181458014224e-05, "loss": 0.025, "step": 46300 }, { "grad_norm": 0.21901540458202362, "learning_rate": 6.016526919311435e-05, "loss": 0.019, "step": 46310 }, { "grad_norm": 0.2709788978099823, "learning_rate": 6.014907926034613e-05, "loss": 0.0204, "step": 46320 }, { "grad_norm": 0.16290925443172455, "learning_rate": 6.0132888217689875e-05, "loss": 0.0227, "step": 46330 }, { "grad_norm": 0.2234066277742386, "learning_rate": 6.01166960669162e-05, "loss": 0.0207, "step": 46340 }, { "grad_norm": 0.21975712478160858, "learning_rate": 6.0100502809795845e-05, "loss": 0.0233, "step": 46350 }, { "grad_norm": 0.18806369602680206, "learning_rate": 6.00843084480997e-05, "loss": 0.0203, "step": 46360 }, { "grad_norm": 0.24434274435043335, "learning_rate": 6.0068112983598736e-05, "loss": 0.0239, "step": 46370 }, { "grad_norm": 0.28739985823631287, "learning_rate": 6.005191641806407e-05, "loss": 0.0225, "step": 46380 }, { "grad_norm": 0.26098111271858215, "learning_rate": 6.003571875326694e-05, "loss": 0.0231, "step": 46390 }, { "grad_norm": 0.20411892235279083, "learning_rate": 6.00195199909787e-05, "loss": 0.0237, "step": 46400 }, { "grad_norm": 0.1797320395708084, "learning_rate": 6.000332013297082e-05, "loss": 0.02, "step": 46410 }, { "grad_norm": 0.18672452867031097, "learning_rate": 5.998711918101487e-05, "loss": 0.0193, "step": 46420 }, { "grad_norm": 0.1837330013513565, "learning_rate": 5.997091713688261e-05, "loss": 0.0183, "step": 46430 }, { "grad_norm": 0.17580735683441162, "learning_rate": 5.9954714002345836e-05, "loss": 0.0173, "step": 46440 }, { "grad_norm": 0.31975430250167847, "learning_rate": 5.993850977917649e-05, "loss": 0.0202, "step": 46450 }, { "grad_norm": 0.25217121839523315, "learning_rate": 5.992230446914667e-05, "loss": 0.0205, "step": 46460 }, { "grad_norm": 0.18035852909088135, "learning_rate": 5.990609807402855e-05, "loss": 0.0191, "step": 46470 }, { "grad_norm": 0.22879557311534882, "learning_rate": 5.988989059559443e-05, "loss": 0.0194, "step": 46480 }, { "grad_norm": 0.29199016094207764, "learning_rate": 5.987368203561675e-05, "loss": 0.02, "step": 46490 }, { "grad_norm": 0.22797343134880066, "learning_rate": 5.9857472395868055e-05, "loss": 0.0188, "step": 46500 }, { "grad_norm": 0.19310759007930756, "learning_rate": 5.9841261678120983e-05, "loss": 0.0185, "step": 46510 }, { "grad_norm": 0.17709879577159882, "learning_rate": 5.982504988414834e-05, "loss": 0.0171, "step": 46520 }, { "grad_norm": 0.21157307922840118, "learning_rate": 5.9808837015723015e-05, "loss": 0.0176, "step": 46530 }, { "grad_norm": 0.1304878294467926, "learning_rate": 5.9792623074618016e-05, "loss": 0.0194, "step": 46540 }, { "grad_norm": 0.19122909009456635, "learning_rate": 5.977640806260648e-05, "loss": 0.0198, "step": 46550 }, { "grad_norm": 0.26937010884284973, "learning_rate": 5.976019198146166e-05, "loss": 0.0218, "step": 46560 }, { "grad_norm": 0.17970672249794006, "learning_rate": 5.974397483295692e-05, "loss": 0.0219, "step": 46570 }, { "grad_norm": 0.18672426044940948, "learning_rate": 5.972775661886575e-05, "loss": 0.0209, "step": 46580 }, { "grad_norm": 0.18274401128292084, "learning_rate": 5.971153734096173e-05, "loss": 0.0192, "step": 46590 }, { "grad_norm": 0.20135906338691711, "learning_rate": 5.969531700101859e-05, "loss": 0.0204, "step": 46600 }, { "grad_norm": 0.2015281617641449, "learning_rate": 5.9679095600810155e-05, "loss": 0.021, "step": 46610 }, { "grad_norm": 0.19138960540294647, "learning_rate": 5.9662873142110384e-05, "loss": 0.0217, "step": 46620 }, { "grad_norm": 0.25125378370285034, "learning_rate": 5.964664962669333e-05, "loss": 0.022, "step": 46630 }, { "grad_norm": 0.20104023814201355, "learning_rate": 5.9630425056333186e-05, "loss": 0.0216, "step": 46640 }, { "grad_norm": 0.18326298892498016, "learning_rate": 5.961419943280422e-05, "loss": 0.0169, "step": 46650 }, { "grad_norm": 0.16873247921466827, "learning_rate": 5.959797275788087e-05, "loss": 0.02, "step": 46660 }, { "grad_norm": 0.19972416758537292, "learning_rate": 5.958174503333765e-05, "loss": 0.0181, "step": 46670 }, { "grad_norm": 0.2477429360151291, "learning_rate": 5.9565516260949195e-05, "loss": 0.017, "step": 46680 }, { "grad_norm": 0.22839891910552979, "learning_rate": 5.9549286442490273e-05, "loss": 0.0216, "step": 46690 }, { "grad_norm": 0.23791703581809998, "learning_rate": 5.953305557973572e-05, "loss": 0.0221, "step": 46700 }, { "grad_norm": 0.18948687613010406, "learning_rate": 5.9516823674460564e-05, "loss": 0.0204, "step": 46710 }, { "grad_norm": 0.20134209096431732, "learning_rate": 5.9500590728439875e-05, "loss": 0.0202, "step": 46720 }, { "grad_norm": 0.20227929949760437, "learning_rate": 5.948435674344886e-05, "loss": 0.0211, "step": 46730 }, { "grad_norm": 0.1738853007555008, "learning_rate": 5.946812172126285e-05, "loss": 0.0218, "step": 46740 }, { "grad_norm": 0.19912216067314148, "learning_rate": 5.945188566365729e-05, "loss": 0.0241, "step": 46750 }, { "grad_norm": 0.22952908277511597, "learning_rate": 5.943564857240773e-05, "loss": 0.0232, "step": 46760 }, { "grad_norm": 0.17979496717453003, "learning_rate": 5.941941044928981e-05, "loss": 0.0221, "step": 46770 }, { "grad_norm": 0.28360995650291443, "learning_rate": 5.940317129607935e-05, "loss": 0.0228, "step": 46780 }, { "grad_norm": 0.1896696835756302, "learning_rate": 5.9386931114552204e-05, "loss": 0.0204, "step": 46790 }, { "grad_norm": 0.26775917410850525, "learning_rate": 5.937068990648438e-05, "loss": 0.024, "step": 46800 }, { "grad_norm": 0.2894566059112549, "learning_rate": 5.935444767365199e-05, "loss": 0.0267, "step": 46810 }, { "grad_norm": 0.22820617258548737, "learning_rate": 5.933820441783129e-05, "loss": 0.0213, "step": 46820 }, { "grad_norm": 0.2307496964931488, "learning_rate": 5.932196014079857e-05, "loss": 0.0227, "step": 46830 }, { "grad_norm": 0.18072474002838135, "learning_rate": 5.930571484433032e-05, "loss": 0.0247, "step": 46840 }, { "grad_norm": 0.2485787272453308, "learning_rate": 5.928946853020309e-05, "loss": 0.0259, "step": 46850 }, { "grad_norm": 0.29091012477874756, "learning_rate": 5.927322120019352e-05, "loss": 0.0219, "step": 46860 }, { "grad_norm": 0.19874265789985657, "learning_rate": 5.925697285607844e-05, "loss": 0.0242, "step": 46870 }, { "grad_norm": 0.22063395380973816, "learning_rate": 5.924072349963472e-05, "loss": 0.0238, "step": 46880 }, { "grad_norm": 0.20165328681468964, "learning_rate": 5.922447313263938e-05, "loss": 0.0275, "step": 46890 }, { "grad_norm": 0.1741485744714737, "learning_rate": 5.920822175686952e-05, "loss": 0.0201, "step": 46900 }, { "grad_norm": 0.18460151553153992, "learning_rate": 5.919196937410239e-05, "loss": 0.0196, "step": 46910 }, { "grad_norm": 0.22052550315856934, "learning_rate": 5.917571598611529e-05, "loss": 0.0231, "step": 46920 }, { "grad_norm": 0.24571555852890015, "learning_rate": 5.9159461594685686e-05, "loss": 0.0256, "step": 46930 }, { "grad_norm": 0.25285032391548157, "learning_rate": 5.914320620159114e-05, "loss": 0.0245, "step": 46940 }, { "grad_norm": 0.2469419687986374, "learning_rate": 5.912694980860931e-05, "loss": 0.0211, "step": 46950 }, { "grad_norm": 0.2158176153898239, "learning_rate": 5.9110692417517964e-05, "loss": 0.0207, "step": 46960 }, { "grad_norm": 0.2385857105255127, "learning_rate": 5.9094434030095e-05, "loss": 0.0195, "step": 46970 }, { "grad_norm": 0.19118013978004456, "learning_rate": 5.907817464811841e-05, "loss": 0.0218, "step": 46980 }, { "grad_norm": 0.21811942756175995, "learning_rate": 5.906191427336627e-05, "loss": 0.0256, "step": 46990 }, { "grad_norm": 0.22404180467128754, "learning_rate": 5.9045652907616835e-05, "loss": 0.0246, "step": 47000 }, { "grad_norm": 0.23551930487155914, "learning_rate": 5.902939055264838e-05, "loss": 0.0222, "step": 47010 }, { "grad_norm": 0.19219091534614563, "learning_rate": 5.901312721023935e-05, "loss": 0.0261, "step": 47020 }, { "grad_norm": 0.1950749307870865, "learning_rate": 5.8996862882168294e-05, "loss": 0.0234, "step": 47030 }, { "grad_norm": 0.3130800724029541, "learning_rate": 5.8980597570213826e-05, "loss": 0.0252, "step": 47040 }, { "grad_norm": 0.2240944355726242, "learning_rate": 5.896433127615471e-05, "loss": 0.0219, "step": 47050 }, { "grad_norm": 0.26462337374687195, "learning_rate": 5.894806400176981e-05, "loss": 0.0258, "step": 47060 }, { "grad_norm": 0.1999545693397522, "learning_rate": 5.893179574883808e-05, "loss": 0.0237, "step": 47070 }, { "grad_norm": 0.24353419244289398, "learning_rate": 5.8915526519138585e-05, "loss": 0.0219, "step": 47080 }, { "grad_norm": 0.23667791485786438, "learning_rate": 5.889925631445053e-05, "loss": 0.0205, "step": 47090 }, { "grad_norm": 0.23412400484085083, "learning_rate": 5.888298513655318e-05, "loss": 0.0216, "step": 47100 }, { "grad_norm": 0.21925856173038483, "learning_rate": 5.886671298722592e-05, "loss": 0.0181, "step": 47110 }, { "grad_norm": 0.191481813788414, "learning_rate": 5.885043986824828e-05, "loss": 0.0183, "step": 47120 }, { "grad_norm": 0.1795223355293274, "learning_rate": 5.883416578139982e-05, "loss": 0.0222, "step": 47130 }, { "grad_norm": 0.18632113933563232, "learning_rate": 5.881789072846029e-05, "loss": 0.0218, "step": 47140 }, { "grad_norm": 0.1841869205236435, "learning_rate": 5.8801614711209474e-05, "loss": 0.0202, "step": 47150 }, { "grad_norm": 0.24290622770786285, "learning_rate": 5.878533773142732e-05, "loss": 0.026, "step": 47160 }, { "grad_norm": 0.19355495274066925, "learning_rate": 5.8769059790893834e-05, "loss": 0.0188, "step": 47170 }, { "grad_norm": 0.16958381235599518, "learning_rate": 5.8752780891389145e-05, "loss": 0.0188, "step": 47180 }, { "grad_norm": 0.27737221121788025, "learning_rate": 5.873650103469351e-05, "loss": 0.0226, "step": 47190 }, { "grad_norm": 0.234089657664299, "learning_rate": 5.872022022258725e-05, "loss": 0.0168, "step": 47200 }, { "grad_norm": 0.14992420375347137, "learning_rate": 5.8703938456850814e-05, "loss": 0.0172, "step": 47210 }, { "grad_norm": 0.17410171031951904, "learning_rate": 5.868765573926477e-05, "loss": 0.0164, "step": 47220 }, { "grad_norm": 0.1709209382534027, "learning_rate": 5.867137207160974e-05, "loss": 0.0152, "step": 47230 }, { "grad_norm": 0.20747599005699158, "learning_rate": 5.86550874556665e-05, "loss": 0.0153, "step": 47240 }, { "grad_norm": 0.19360247254371643, "learning_rate": 5.863880189321592e-05, "loss": 0.0207, "step": 47250 }, { "grad_norm": 0.18641456961631775, "learning_rate": 5.862251538603894e-05, "loss": 0.0234, "step": 47260 }, { "grad_norm": 0.257517009973526, "learning_rate": 5.8606227935916656e-05, "loss": 0.0211, "step": 47270 }, { "grad_norm": 0.2330872118473053, "learning_rate": 5.858993954463021e-05, "loss": 0.0229, "step": 47280 }, { "grad_norm": 0.23722802102565765, "learning_rate": 5.8573650213960896e-05, "loss": 0.0216, "step": 47290 }, { "grad_norm": 0.20025938749313354, "learning_rate": 5.8557359945690084e-05, "loss": 0.0178, "step": 47300 }, { "grad_norm": 0.2856602668762207, "learning_rate": 5.8541068741599246e-05, "loss": 0.0217, "step": 47310 }, { "grad_norm": 0.1765010803937912, "learning_rate": 5.8524776603469985e-05, "loss": 0.0231, "step": 47320 }, { "grad_norm": 0.17394115030765533, "learning_rate": 5.850848353308397e-05, "loss": 0.0199, "step": 47330 }, { "grad_norm": 0.27849602699279785, "learning_rate": 5.849218953222297e-05, "loss": 0.0186, "step": 47340 }, { "grad_norm": 0.16924238204956055, "learning_rate": 5.847589460266891e-05, "loss": 0.0172, "step": 47350 }, { "grad_norm": 0.17562627792358398, "learning_rate": 5.845959874620375e-05, "loss": 0.018, "step": 47360 }, { "grad_norm": 0.16956910490989685, "learning_rate": 5.84433019646096e-05, "loss": 0.0171, "step": 47370 }, { "grad_norm": 0.24131126701831818, "learning_rate": 5.842700425966863e-05, "loss": 0.0184, "step": 47380 }, { "grad_norm": 0.24944524466991425, "learning_rate": 5.841070563316315e-05, "loss": 0.0187, "step": 47390 }, { "grad_norm": 0.25588974356651306, "learning_rate": 5.839440608687554e-05, "loss": 0.0226, "step": 47400 }, { "grad_norm": 0.21812531352043152, "learning_rate": 5.837810562258831e-05, "loss": 0.0183, "step": 47410 }, { "grad_norm": 0.17124365270137787, "learning_rate": 5.836180424208405e-05, "loss": 0.0176, "step": 47420 }, { "grad_norm": 0.16379530727863312, "learning_rate": 5.834550194714543e-05, "loss": 0.0206, "step": 47430 }, { "grad_norm": 0.20741790533065796, "learning_rate": 5.832919873955528e-05, "loss": 0.0212, "step": 47440 }, { "grad_norm": 0.2632391154766083, "learning_rate": 5.831289462109648e-05, "loss": 0.0191, "step": 47450 }, { "grad_norm": 0.2057378739118576, "learning_rate": 5.829658959355202e-05, "loss": 0.0215, "step": 47460 }, { "grad_norm": 0.19373486936092377, "learning_rate": 5.8280283658704995e-05, "loss": 0.0168, "step": 47470 }, { "grad_norm": 0.2241217941045761, "learning_rate": 5.826397681833859e-05, "loss": 0.0237, "step": 47480 }, { "grad_norm": 0.2452155351638794, "learning_rate": 5.824766907423612e-05, "loss": 0.0202, "step": 47490 }, { "grad_norm": 0.2325916290283203, "learning_rate": 5.8231360428180945e-05, "loss": 0.0192, "step": 47500 }, { "grad_norm": 0.4044409692287445, "learning_rate": 5.821505088195658e-05, "loss": 0.0195, "step": 47510 }, { "grad_norm": 0.1935366839170456, "learning_rate": 5.819874043734661e-05, "loss": 0.0241, "step": 47520 }, { "grad_norm": 0.17962795495986938, "learning_rate": 5.8182429096134695e-05, "loss": 0.0218, "step": 47530 }, { "grad_norm": 0.1758139431476593, "learning_rate": 5.816611686010465e-05, "loss": 0.0211, "step": 47540 }, { "grad_norm": 0.18399131298065186, "learning_rate": 5.814980373104033e-05, "loss": 0.0184, "step": 47550 }, { "grad_norm": 0.2228081077337265, "learning_rate": 5.813348971072572e-05, "loss": 0.021, "step": 47560 }, { "grad_norm": 0.14732660353183746, "learning_rate": 5.811717480094492e-05, "loss": 0.02, "step": 47570 }, { "grad_norm": 0.1697869747877121, "learning_rate": 5.810085900348209e-05, "loss": 0.0198, "step": 47580 }, { "grad_norm": 0.2389049381017685, "learning_rate": 5.8084542320121483e-05, "loss": 0.0202, "step": 47590 }, { "grad_norm": 0.3180103003978729, "learning_rate": 5.8068224752647497e-05, "loss": 0.0211, "step": 47600 }, { "grad_norm": 0.19563762843608856, "learning_rate": 5.805190630284457e-05, "loss": 0.0219, "step": 47610 }, { "grad_norm": 0.2187429815530777, "learning_rate": 5.803558697249729e-05, "loss": 0.0195, "step": 47620 }, { "grad_norm": 0.24636700749397278, "learning_rate": 5.8019266763390276e-05, "loss": 0.0193, "step": 47630 }, { "grad_norm": 0.16393674910068512, "learning_rate": 5.800294567730833e-05, "loss": 0.0198, "step": 47640 }, { "grad_norm": 0.1708841770887375, "learning_rate": 5.798662371603628e-05, "loss": 0.0157, "step": 47650 }, { "grad_norm": 0.1457788646221161, "learning_rate": 5.797030088135904e-05, "loss": 0.015, "step": 47660 }, { "grad_norm": 0.15233586728572845, "learning_rate": 5.7953977175061705e-05, "loss": 0.018, "step": 47670 }, { "grad_norm": 0.2201276421546936, "learning_rate": 5.793765259892938e-05, "loss": 0.019, "step": 47680 }, { "grad_norm": 0.31449854373931885, "learning_rate": 5.792132715474729e-05, "loss": 0.021, "step": 47690 }, { "grad_norm": 0.17638860642910004, "learning_rate": 5.790500084430078e-05, "loss": 0.0173, "step": 47700 }, { "grad_norm": 0.21100099384784698, "learning_rate": 5.7888673669375274e-05, "loss": 0.0205, "step": 47710 }, { "grad_norm": 0.20442011952400208, "learning_rate": 5.787234563175625e-05, "loss": 0.0194, "step": 47720 }, { "grad_norm": 0.2180153727531433, "learning_rate": 5.7856016733229355e-05, "loss": 0.0181, "step": 47730 }, { "grad_norm": 0.2013155221939087, "learning_rate": 5.7839686975580297e-05, "loss": 0.0244, "step": 47740 }, { "grad_norm": 0.22704452276229858, "learning_rate": 5.782335636059484e-05, "loss": 0.0194, "step": 47750 }, { "grad_norm": 0.2782270312309265, "learning_rate": 5.780702489005889e-05, "loss": 0.023, "step": 47760 }, { "grad_norm": 0.24252177774906158, "learning_rate": 5.779069256575845e-05, "loss": 0.0199, "step": 47770 }, { "grad_norm": 0.15420158207416534, "learning_rate": 5.7774359389479574e-05, "loss": 0.0206, "step": 47780 }, { "grad_norm": 0.24336561560630798, "learning_rate": 5.775802536300845e-05, "loss": 0.02, "step": 47790 }, { "grad_norm": 0.15946216881275177, "learning_rate": 5.774169048813134e-05, "loss": 0.0225, "step": 47800 }, { "grad_norm": 0.17694774270057678, "learning_rate": 5.77253547666346e-05, "loss": 0.0213, "step": 47810 }, { "grad_norm": 0.19434280693531036, "learning_rate": 5.770901820030465e-05, "loss": 0.018, "step": 47820 }, { "grad_norm": 0.21160826086997986, "learning_rate": 5.769268079092809e-05, "loss": 0.0174, "step": 47830 }, { "grad_norm": 0.23539555072784424, "learning_rate": 5.767634254029151e-05, "loss": 0.0212, "step": 47840 }, { "grad_norm": 0.19787247478961945, "learning_rate": 5.7660003450181655e-05, "loss": 0.0188, "step": 47850 }, { "grad_norm": 0.14313791692256927, "learning_rate": 5.764366352238534e-05, "loss": 0.0189, "step": 47860 }, { "grad_norm": 0.22188323736190796, "learning_rate": 5.7627322758689474e-05, "loss": 0.0198, "step": 47870 }, { "grad_norm": 0.2294168770313263, "learning_rate": 5.761098116088105e-05, "loss": 0.0203, "step": 47880 }, { "grad_norm": 0.30501502752304077, "learning_rate": 5.759463873074717e-05, "loss": 0.02, "step": 47890 }, { "grad_norm": 0.2604392468929291, "learning_rate": 5.757829547007504e-05, "loss": 0.025, "step": 47900 }, { "grad_norm": 0.24921296536922455, "learning_rate": 5.756195138065189e-05, "loss": 0.0221, "step": 47910 }, { "grad_norm": 0.17900347709655762, "learning_rate": 5.754560646426511e-05, "loss": 0.0184, "step": 47920 }, { "grad_norm": 0.16724641621112823, "learning_rate": 5.752926072270216e-05, "loss": 0.0201, "step": 47930 }, { "grad_norm": 0.22116082906723022, "learning_rate": 5.7512914157750563e-05, "loss": 0.018, "step": 47940 }, { "grad_norm": 0.15409308671951294, "learning_rate": 5.749656677119798e-05, "loss": 0.0173, "step": 47950 }, { "grad_norm": 0.21286676824092865, "learning_rate": 5.7480218564832125e-05, "loss": 0.0185, "step": 47960 }, { "grad_norm": 0.12491817772388458, "learning_rate": 5.746386954044082e-05, "loss": 0.0213, "step": 47970 }, { "grad_norm": 0.19288024306297302, "learning_rate": 5.744751969981195e-05, "loss": 0.0214, "step": 47980 }, { "grad_norm": 0.15564125776290894, "learning_rate": 5.7431169044733526e-05, "loss": 0.0197, "step": 47990 }, { "grad_norm": 0.2674106955528259, "learning_rate": 5.741481757699364e-05, "loss": 0.0216, "step": 48000 }, { "grad_norm": 0.25864097476005554, "learning_rate": 5.7398465298380434e-05, "loss": 0.0222, "step": 48010 }, { "grad_norm": 0.27503135800361633, "learning_rate": 5.7382112210682193e-05, "loss": 0.0206, "step": 48020 }, { "grad_norm": 0.21266265213489532, "learning_rate": 5.7365758315687266e-05, "loss": 0.0234, "step": 48030 }, { "grad_norm": 0.22496679425239563, "learning_rate": 5.734940361518407e-05, "loss": 0.0193, "step": 48040 }, { "grad_norm": 0.24753724038600922, "learning_rate": 5.733304811096116e-05, "loss": 0.0205, "step": 48050 }, { "grad_norm": 0.23094190657138824, "learning_rate": 5.731669180480713e-05, "loss": 0.0194, "step": 48060 }, { "grad_norm": 0.18906590342521667, "learning_rate": 5.730033469851067e-05, "loss": 0.0187, "step": 48070 }, { "grad_norm": 0.19074906408786774, "learning_rate": 5.7283976793860607e-05, "loss": 0.0204, "step": 48080 }, { "grad_norm": 0.21181000769138336, "learning_rate": 5.726761809264577e-05, "loss": 0.0197, "step": 48090 }, { "grad_norm": 0.20320813357830048, "learning_rate": 5.7251258596655155e-05, "loss": 0.0221, "step": 48100 }, { "grad_norm": 0.2233193814754486, "learning_rate": 5.72348983076778e-05, "loss": 0.0228, "step": 48110 }, { "grad_norm": 0.2416813224554062, "learning_rate": 5.7218537227502854e-05, "loss": 0.0223, "step": 48120 }, { "grad_norm": 0.21206997334957123, "learning_rate": 5.720217535791951e-05, "loss": 0.0216, "step": 48130 }, { "grad_norm": 0.2470131814479828, "learning_rate": 5.718581270071711e-05, "loss": 0.0196, "step": 48140 }, { "grad_norm": 0.24304291605949402, "learning_rate": 5.716944925768505e-05, "loss": 0.0175, "step": 48150 }, { "grad_norm": 0.18197527527809143, "learning_rate": 5.7153085030612786e-05, "loss": 0.0206, "step": 48160 }, { "grad_norm": 0.23796230554580688, "learning_rate": 5.713672002128989e-05, "loss": 0.0213, "step": 48170 }, { "grad_norm": 0.26969724893569946, "learning_rate": 5.712035423150602e-05, "loss": 0.0208, "step": 48180 }, { "grad_norm": 0.29426515102386475, "learning_rate": 5.710398766305094e-05, "loss": 0.0185, "step": 48190 }, { "grad_norm": 0.15520209074020386, "learning_rate": 5.708762031771442e-05, "loss": 0.0222, "step": 48200 }, { "grad_norm": 0.2169274240732193, "learning_rate": 5.707125219728642e-05, "loss": 0.0186, "step": 48210 }, { "grad_norm": 0.22123725712299347, "learning_rate": 5.7054883303556905e-05, "loss": 0.0214, "step": 48220 }, { "grad_norm": 0.19516071677207947, "learning_rate": 5.703851363831595e-05, "loss": 0.0195, "step": 48230 }, { "grad_norm": 0.20742689073085785, "learning_rate": 5.702214320335374e-05, "loss": 0.0207, "step": 48240 }, { "grad_norm": 0.22079937160015106, "learning_rate": 5.7005772000460514e-05, "loss": 0.021, "step": 48250 }, { "grad_norm": 0.18490590155124664, "learning_rate": 5.698940003142656e-05, "loss": 0.0169, "step": 48260 }, { "grad_norm": 0.15531225502490997, "learning_rate": 5.697302729804236e-05, "loss": 0.0173, "step": 48270 }, { "grad_norm": 0.17716340720653534, "learning_rate": 5.695665380209837e-05, "loss": 0.0197, "step": 48280 }, { "grad_norm": 0.23355136811733246, "learning_rate": 5.6940279545385165e-05, "loss": 0.0217, "step": 48290 }, { "grad_norm": 0.15888306498527527, "learning_rate": 5.692390452969344e-05, "loss": 0.0187, "step": 48300 }, { "grad_norm": 0.1964024305343628, "learning_rate": 5.690752875681392e-05, "loss": 0.0206, "step": 48310 }, { "grad_norm": 0.1359967589378357, "learning_rate": 5.6891152228537435e-05, "loss": 0.0177, "step": 48320 }, { "grad_norm": 0.21849578619003296, "learning_rate": 5.687477494665492e-05, "loss": 0.0163, "step": 48330 }, { "grad_norm": 0.24528203904628754, "learning_rate": 5.685839691295733e-05, "loss": 0.0171, "step": 48340 }, { "grad_norm": 0.23352836072444916, "learning_rate": 5.6842018129235786e-05, "loss": 0.0205, "step": 48350 }, { "grad_norm": 0.17718347907066345, "learning_rate": 5.6825638597281404e-05, "loss": 0.0215, "step": 48360 }, { "grad_norm": 0.24715305864810944, "learning_rate": 5.680925831888546e-05, "loss": 0.0262, "step": 48370 }, { "grad_norm": 0.2956782579421997, "learning_rate": 5.6792877295839274e-05, "loss": 0.0207, "step": 48380 }, { "grad_norm": 0.18648062646389008, "learning_rate": 5.6776495529934224e-05, "loss": 0.0198, "step": 48390 }, { "grad_norm": 0.20785564184188843, "learning_rate": 5.6760113022961824e-05, "loss": 0.0169, "step": 48400 }, { "grad_norm": 0.21460887789726257, "learning_rate": 5.6743729776713617e-05, "loss": 0.0196, "step": 48410 }, { "grad_norm": 0.3513580858707428, "learning_rate": 5.672734579298126e-05, "loss": 0.019, "step": 48420 }, { "grad_norm": 0.189551442861557, "learning_rate": 5.671096107355649e-05, "loss": 0.0181, "step": 48430 }, { "grad_norm": 0.24988767504692078, "learning_rate": 5.66945756202311e-05, "loss": 0.0297, "step": 48440 }, { "grad_norm": 1.0653793811798096, "learning_rate": 5.667818943479699e-05, "loss": 0.0248, "step": 48450 }, { "grad_norm": 0.3824174106121063, "learning_rate": 5.666180251904612e-05, "loss": 0.0231, "step": 48460 }, { "grad_norm": 0.20551007986068726, "learning_rate": 5.6645414874770555e-05, "loss": 0.0247, "step": 48470 }, { "grad_norm": 0.18013931810855865, "learning_rate": 5.66290265037624e-05, "loss": 0.0222, "step": 48480 }, { "grad_norm": 0.30350086092948914, "learning_rate": 5.661263740781386e-05, "loss": 0.0269, "step": 48490 }, { "grad_norm": 0.1922898292541504, "learning_rate": 5.6596247588717254e-05, "loss": 0.0227, "step": 48500 }, { "grad_norm": 0.19243095815181732, "learning_rate": 5.6579857048264926e-05, "loss": 0.0221, "step": 48510 }, { "grad_norm": 0.22239430248737335, "learning_rate": 5.6563465788249314e-05, "loss": 0.0207, "step": 48520 }, { "grad_norm": 0.23254315555095673, "learning_rate": 5.6547073810462956e-05, "loss": 0.0216, "step": 48530 }, { "grad_norm": 0.20793363451957703, "learning_rate": 5.653068111669846e-05, "loss": 0.0202, "step": 48540 }, { "grad_norm": 0.23512153327465057, "learning_rate": 5.651428770874848e-05, "loss": 0.0186, "step": 48550 }, { "grad_norm": 0.1783216893672943, "learning_rate": 5.64978935884058e-05, "loss": 0.0177, "step": 48560 }, { "grad_norm": 0.2266884595155716, "learning_rate": 5.6481498757463244e-05, "loss": 0.0244, "step": 48570 }, { "grad_norm": 0.15095263719558716, "learning_rate": 5.646510321771373e-05, "loss": 0.0203, "step": 48580 }, { "grad_norm": 0.2239929586648941, "learning_rate": 5.644870697095024e-05, "loss": 0.0248, "step": 48590 }, { "grad_norm": 0.24633225798606873, "learning_rate": 5.643231001896586e-05, "loss": 0.0206, "step": 48600 }, { "grad_norm": 0.2054406851530075, "learning_rate": 5.6415912363553726e-05, "loss": 0.0197, "step": 48610 }, { "grad_norm": 0.2313813716173172, "learning_rate": 5.639951400650706e-05, "loss": 0.0196, "step": 48620 }, { "grad_norm": 0.2736619710922241, "learning_rate": 5.6383114949619165e-05, "loss": 0.0207, "step": 48630 }, { "grad_norm": 0.2125421017408371, "learning_rate": 5.636671519468342e-05, "loss": 0.0206, "step": 48640 }, { "grad_norm": 0.22305236756801605, "learning_rate": 5.635031474349327e-05, "loss": 0.0197, "step": 48650 }, { "grad_norm": 0.1609496772289276, "learning_rate": 5.6333913597842246e-05, "loss": 0.0206, "step": 48660 }, { "grad_norm": 0.16334772109985352, "learning_rate": 5.6317511759523955e-05, "loss": 0.0201, "step": 48670 }, { "grad_norm": 0.19594846665859222, "learning_rate": 5.630110923033207e-05, "loss": 0.0193, "step": 48680 }, { "grad_norm": 0.25328415632247925, "learning_rate": 5.628470601206036e-05, "loss": 0.0199, "step": 48690 }, { "grad_norm": 0.27132275700569153, "learning_rate": 5.626830210650266e-05, "loss": 0.0199, "step": 48700 }, { "grad_norm": 0.23181024193763733, "learning_rate": 5.625189751545285e-05, "loss": 0.0225, "step": 48710 }, { "grad_norm": 0.15053518116474152, "learning_rate": 5.6235492240704936e-05, "loss": 0.0184, "step": 48720 }, { "grad_norm": 0.18078546226024628, "learning_rate": 5.621908628405296e-05, "loss": 0.0176, "step": 48730 }, { "grad_norm": 0.20021341741085052, "learning_rate": 5.620267964729106e-05, "loss": 0.0215, "step": 48740 }, { "grad_norm": 0.21543405950069427, "learning_rate": 5.618627233221344e-05, "loss": 0.0179, "step": 48750 }, { "grad_norm": 0.1882188320159912, "learning_rate": 5.616986434061438e-05, "loss": 0.0159, "step": 48760 }, { "grad_norm": 0.2224922627210617, "learning_rate": 5.615345567428822e-05, "loss": 0.0186, "step": 48770 }, { "grad_norm": 0.1473071426153183, "learning_rate": 5.613704633502941e-05, "loss": 0.0206, "step": 48780 }, { "grad_norm": 0.1805226057767868, "learning_rate": 5.612063632463245e-05, "loss": 0.0193, "step": 48790 }, { "grad_norm": 0.28277426958084106, "learning_rate": 5.610422564489188e-05, "loss": 0.0186, "step": 48800 }, { "grad_norm": 0.2595684230327606, "learning_rate": 5.608781429760239e-05, "loss": 0.0188, "step": 48810 }, { "grad_norm": 0.23895788192749023, "learning_rate": 5.607140228455866e-05, "loss": 0.0233, "step": 48820 }, { "grad_norm": 0.2206163853406906, "learning_rate": 5.605498960755553e-05, "loss": 0.02, "step": 48830 }, { "grad_norm": 0.2083227038383484, "learning_rate": 5.603857626838782e-05, "loss": 0.0193, "step": 48840 }, { "grad_norm": 0.28693607449531555, "learning_rate": 5.60221622688505e-05, "loss": 0.0206, "step": 48850 }, { "grad_norm": 0.20075109601020813, "learning_rate": 5.6005747610738565e-05, "loss": 0.0187, "step": 48860 }, { "grad_norm": 0.24730505049228668, "learning_rate": 5.59893322958471e-05, "loss": 0.0204, "step": 48870 }, { "grad_norm": 0.2397819459438324, "learning_rate": 5.5972916325971256e-05, "loss": 0.0209, "step": 48880 }, { "grad_norm": 0.19284555315971375, "learning_rate": 5.595649970290628e-05, "loss": 0.0179, "step": 48890 }, { "grad_norm": 0.2204466462135315, "learning_rate": 5.5940082428447426e-05, "loss": 0.021, "step": 48900 }, { "grad_norm": 0.1726316213607788, "learning_rate": 5.592366450439012e-05, "loss": 0.0189, "step": 48910 }, { "grad_norm": 0.1905570775270462, "learning_rate": 5.590724593252975e-05, "loss": 0.0179, "step": 48920 }, { "grad_norm": 0.16880054771900177, "learning_rate": 5.589082671466184e-05, "loss": 0.0183, "step": 48930 }, { "grad_norm": 0.204849973320961, "learning_rate": 5.587440685258199e-05, "loss": 0.0165, "step": 48940 }, { "grad_norm": 0.2095097452402115, "learning_rate": 5.585798634808583e-05, "loss": 0.02, "step": 48950 }, { "grad_norm": 0.13488584756851196, "learning_rate": 5.584156520296909e-05, "loss": 0.0173, "step": 48960 }, { "grad_norm": 0.26904571056365967, "learning_rate": 5.582514341902757e-05, "loss": 0.0212, "step": 48970 }, { "grad_norm": 0.31696006655693054, "learning_rate": 5.580872099805713e-05, "loss": 0.0219, "step": 48980 }, { "grad_norm": 0.23979848623275757, "learning_rate": 5.5792297941853674e-05, "loss": 0.0205, "step": 48990 }, { "grad_norm": 0.2937689423561096, "learning_rate": 5.5775874252213247e-05, "loss": 0.0227, "step": 49000 }, { "grad_norm": 0.33583199977874756, "learning_rate": 5.575944993093189e-05, "loss": 0.0204, "step": 49010 }, { "grad_norm": 0.23017123341560364, "learning_rate": 5.574302497980574e-05, "loss": 0.0216, "step": 49020 }, { "grad_norm": 0.19637271761894226, "learning_rate": 5.5726599400631005e-05, "loss": 0.019, "step": 49030 }, { "grad_norm": 0.19597505033016205, "learning_rate": 5.571017319520399e-05, "loss": 0.0205, "step": 49040 }, { "grad_norm": 0.188230499625206, "learning_rate": 5.5693746365320984e-05, "loss": 0.0203, "step": 49050 }, { "grad_norm": 0.19899572432041168, "learning_rate": 5.567731891277848e-05, "loss": 0.0231, "step": 49060 }, { "grad_norm": 0.1976555436849594, "learning_rate": 5.566089083937288e-05, "loss": 0.0207, "step": 49070 }, { "grad_norm": 0.28066286444664, "learning_rate": 5.564446214690079e-05, "loss": 0.0256, "step": 49080 }, { "grad_norm": 0.23834459483623505, "learning_rate": 5.56280328371588e-05, "loss": 0.0208, "step": 49090 }, { "grad_norm": 0.16089065372943878, "learning_rate": 5.5611602911943596e-05, "loss": 0.0171, "step": 49100 }, { "grad_norm": 0.21909886598587036, "learning_rate": 5.559517237305194e-05, "loss": 0.0205, "step": 49110 }, { "grad_norm": 0.29151779413223267, "learning_rate": 5.557874122228064e-05, "loss": 0.0188, "step": 49120 }, { "grad_norm": 0.22446003556251526, "learning_rate": 5.55623094614266e-05, "loss": 0.023, "step": 49130 }, { "grad_norm": 0.20513074100017548, "learning_rate": 5.5545877092286743e-05, "loss": 0.0227, "step": 49140 }, { "grad_norm": 0.1759837418794632, "learning_rate": 5.552944411665811e-05, "loss": 0.0216, "step": 49150 }, { "grad_norm": 0.18655610084533691, "learning_rate": 5.5513010536337795e-05, "loss": 0.0162, "step": 49160 }, { "grad_norm": 0.19408844411373138, "learning_rate": 5.5496576353122934e-05, "loss": 0.021, "step": 49170 }, { "grad_norm": 0.18174822628498077, "learning_rate": 5.548014156881074e-05, "loss": 0.0174, "step": 49180 }, { "grad_norm": 0.20086561143398285, "learning_rate": 5.5463706185198494e-05, "loss": 0.022, "step": 49190 }, { "grad_norm": 0.19245551526546478, "learning_rate": 5.544727020408358e-05, "loss": 0.0179, "step": 49200 }, { "grad_norm": 0.21320119500160217, "learning_rate": 5.5430833627263366e-05, "loss": 0.0185, "step": 49210 }, { "grad_norm": 0.20647257566452026, "learning_rate": 5.5414396456535354e-05, "loss": 0.019, "step": 49220 }, { "grad_norm": 0.22381986677646637, "learning_rate": 5.539795869369709e-05, "loss": 0.0196, "step": 49230 }, { "grad_norm": 0.23224793374538422, "learning_rate": 5.538152034054618e-05, "loss": 0.02, "step": 49240 }, { "grad_norm": 0.14042188227176666, "learning_rate": 5.536508139888028e-05, "loss": 0.0146, "step": 49250 }, { "grad_norm": 0.20904742181301117, "learning_rate": 5.534864187049716e-05, "loss": 0.018, "step": 49260 }, { "grad_norm": 0.20896296203136444, "learning_rate": 5.5332201757194615e-05, "loss": 0.0175, "step": 49270 }, { "grad_norm": 0.1897696703672409, "learning_rate": 5.531576106077049e-05, "loss": 0.0172, "step": 49280 }, { "grad_norm": 0.20029783248901367, "learning_rate": 5.529931978302272e-05, "loss": 0.0163, "step": 49290 }, { "grad_norm": 0.15688571333885193, "learning_rate": 5.528287792574932e-05, "loss": 0.0201, "step": 49300 }, { "grad_norm": 0.21447183191776276, "learning_rate": 5.5266435490748324e-05, "loss": 0.0207, "step": 49310 }, { "grad_norm": 0.44476521015167236, "learning_rate": 5.524999247981787e-05, "loss": 0.0172, "step": 49320 }, { "grad_norm": 0.18797782063484192, "learning_rate": 5.523354889475613e-05, "loss": 0.0188, "step": 49330 }, { "grad_norm": 0.16218580305576324, "learning_rate": 5.521710473736134e-05, "loss": 0.0187, "step": 49340 }, { "grad_norm": 0.29758918285369873, "learning_rate": 5.5200660009431826e-05, "loss": 0.0184, "step": 49350 }, { "grad_norm": 0.16933412849903107, "learning_rate": 5.518421471276596e-05, "loss": 0.0194, "step": 49360 }, { "grad_norm": 0.26522359251976013, "learning_rate": 5.5167768849162174e-05, "loss": 0.0198, "step": 49370 }, { "grad_norm": 0.23042581975460052, "learning_rate": 5.515132242041893e-05, "loss": 0.0177, "step": 49380 }, { "grad_norm": 0.18627940118312836, "learning_rate": 5.513487542833483e-05, "loss": 0.0206, "step": 49390 }, { "grad_norm": 0.22298629581928253, "learning_rate": 5.5118427874708466e-05, "loss": 0.018, "step": 49400 }, { "grad_norm": 0.17695677280426025, "learning_rate": 5.510197976133853e-05, "loss": 0.0177, "step": 49410 }, { "grad_norm": 0.17272555828094482, "learning_rate": 5.508553109002376e-05, "loss": 0.0185, "step": 49420 }, { "grad_norm": 0.1820480227470398, "learning_rate": 5.5069081862562957e-05, "loss": 0.0209, "step": 49430 }, { "grad_norm": 0.20176184177398682, "learning_rate": 5.5052632080754965e-05, "loss": 0.0218, "step": 49440 }, { "grad_norm": 0.2614593505859375, "learning_rate": 5.503618174639874e-05, "loss": 0.0169, "step": 49450 }, { "grad_norm": 0.20321336388587952, "learning_rate": 5.501973086129325e-05, "loss": 0.0182, "step": 49460 }, { "grad_norm": 0.25964704155921936, "learning_rate": 5.500327942723753e-05, "loss": 0.0179, "step": 49470 }, { "grad_norm": 0.22112229466438293, "learning_rate": 5.498682744603071e-05, "loss": 0.0192, "step": 49480 }, { "grad_norm": 0.1554659605026245, "learning_rate": 5.497037491947194e-05, "loss": 0.0193, "step": 49490 }, { "grad_norm": 0.23216095566749573, "learning_rate": 5.4953921849360424e-05, "loss": 0.0189, "step": 49500 }, { "grad_norm": 0.25943735241889954, "learning_rate": 5.493746823749547e-05, "loss": 0.0186, "step": 49510 }, { "grad_norm": 0.18582358956336975, "learning_rate": 5.49210140856764e-05, "loss": 0.0208, "step": 49520 }, { "grad_norm": 0.19086775183677673, "learning_rate": 5.4904559395702647e-05, "loss": 0.0173, "step": 49530 }, { "grad_norm": 0.18674413859844208, "learning_rate": 5.488810416937364e-05, "loss": 0.0168, "step": 49540 }, { "grad_norm": 0.1730448603630066, "learning_rate": 5.4871648408488926e-05, "loss": 0.017, "step": 49550 }, { "grad_norm": 0.18243661522865295, "learning_rate": 5.485519211484807e-05, "loss": 0.0189, "step": 49560 }, { "grad_norm": 0.26934611797332764, "learning_rate": 5.4838735290250676e-05, "loss": 0.0253, "step": 49570 }, { "grad_norm": 0.18532198667526245, "learning_rate": 5.482227793649649e-05, "loss": 0.0214, "step": 49580 }, { "grad_norm": 0.14742900431156158, "learning_rate": 5.480582005538524e-05, "loss": 0.0192, "step": 49590 }, { "grad_norm": 0.24726775288581848, "learning_rate": 5.478936164871671e-05, "loss": 0.0197, "step": 49600 }, { "grad_norm": 0.2125403881072998, "learning_rate": 5.477290271829081e-05, "loss": 0.0226, "step": 49610 }, { "grad_norm": 0.22675971686840057, "learning_rate": 5.475644326590744e-05, "loss": 0.0194, "step": 49620 }, { "grad_norm": 0.17156782746315002, "learning_rate": 5.473998329336658e-05, "loss": 0.021, "step": 49630 }, { "grad_norm": 0.21380412578582764, "learning_rate": 5.4723522802468286e-05, "loss": 0.0161, "step": 49640 }, { "grad_norm": 0.22271902859210968, "learning_rate": 5.470706179501264e-05, "loss": 0.0182, "step": 49650 }, { "grad_norm": 0.20115618407726288, "learning_rate": 5.4690600272799776e-05, "loss": 0.019, "step": 49660 }, { "grad_norm": 0.16459009051322937, "learning_rate": 5.467413823762993e-05, "loss": 0.0178, "step": 49670 }, { "grad_norm": 0.16087451577186584, "learning_rate": 5.465767569130335e-05, "loss": 0.0186, "step": 49680 }, { "grad_norm": 0.170426607131958, "learning_rate": 5.464121263562036e-05, "loss": 0.0161, "step": 49690 }, { "grad_norm": 0.17921729385852814, "learning_rate": 5.4624749072381343e-05, "loss": 0.021, "step": 49700 }, { "grad_norm": 0.18093085289001465, "learning_rate": 5.460828500338672e-05, "loss": 0.0168, "step": 49710 }, { "grad_norm": 0.17809893190860748, "learning_rate": 5.459182043043698e-05, "loss": 0.0157, "step": 49720 }, { "grad_norm": 0.19561655819416046, "learning_rate": 5.457535535533265e-05, "loss": 0.0156, "step": 49730 }, { "grad_norm": 0.20473498106002808, "learning_rate": 5.4558889779874334e-05, "loss": 0.0189, "step": 49740 }, { "grad_norm": 0.18971401453018188, "learning_rate": 5.454242370586269e-05, "loss": 0.0205, "step": 49750 }, { "grad_norm": 0.16716912388801575, "learning_rate": 5.452595713509843e-05, "loss": 0.0205, "step": 49760 }, { "grad_norm": 0.231257826089859, "learning_rate": 5.4509490069382274e-05, "loss": 0.0161, "step": 49770 }, { "grad_norm": 0.19719262421131134, "learning_rate": 5.449302251051509e-05, "loss": 0.0172, "step": 49780 }, { "grad_norm": 0.19682534039020538, "learning_rate": 5.44765544602977e-05, "loss": 0.0176, "step": 49790 }, { "grad_norm": 0.1754101663827896, "learning_rate": 5.446008592053107e-05, "loss": 0.0171, "step": 49800 }, { "grad_norm": 0.15372855961322784, "learning_rate": 5.4443616893016135e-05, "loss": 0.02, "step": 49810 }, { "grad_norm": 0.18007436394691467, "learning_rate": 5.4427147379553934e-05, "loss": 0.017, "step": 49820 }, { "grad_norm": 0.14356356859207153, "learning_rate": 5.441067738194556e-05, "loss": 0.0169, "step": 49830 }, { "grad_norm": 0.15150998532772064, "learning_rate": 5.439420690199214e-05, "loss": 0.0223, "step": 49840 }, { "grad_norm": 0.19626174867153168, "learning_rate": 5.437773594149484e-05, "loss": 0.0181, "step": 49850 }, { "grad_norm": 0.14828260242938995, "learning_rate": 5.436126450225495e-05, "loss": 0.0173, "step": 49860 }, { "grad_norm": 0.17065531015396118, "learning_rate": 5.434479258607371e-05, "loss": 0.017, "step": 49870 }, { "grad_norm": 0.16150708496570587, "learning_rate": 5.43283201947525e-05, "loss": 0.0164, "step": 49880 }, { "grad_norm": 0.18934409320354462, "learning_rate": 5.431184733009268e-05, "loss": 0.0187, "step": 49890 }, { "grad_norm": 0.21182781457901, "learning_rate": 5.4295373993895736e-05, "loss": 0.016, "step": 49900 }, { "grad_norm": 0.19802868366241455, "learning_rate": 5.4278900187963157e-05, "loss": 0.0185, "step": 49910 }, { "grad_norm": 0.18039025366306305, "learning_rate": 5.426242591409646e-05, "loss": 0.0186, "step": 49920 }, { "grad_norm": 0.16887453198432922, "learning_rate": 5.42459511740973e-05, "loss": 0.0159, "step": 49930 }, { "grad_norm": 0.22536516189575195, "learning_rate": 5.422947596976729e-05, "loss": 0.0178, "step": 49940 }, { "grad_norm": 0.16209310293197632, "learning_rate": 5.4213000302908134e-05, "loss": 0.0166, "step": 49950 }, { "grad_norm": 0.20450706779956818, "learning_rate": 5.419652417532162e-05, "loss": 0.0258, "step": 49960 }, { "grad_norm": 0.2004106342792511, "learning_rate": 5.4180047588809534e-05, "loss": 0.0192, "step": 49970 }, { "grad_norm": 0.2053513079881668, "learning_rate": 5.4163570545173704e-05, "loss": 0.0187, "step": 49980 }, { "grad_norm": 0.16225990653038025, "learning_rate": 5.414709304621608e-05, "loss": 0.0194, "step": 49990 }, { "grad_norm": 0.2063005566596985, "learning_rate": 5.413061509373858e-05, "loss": 0.0183, "step": 50000 }, { "grad_norm": 0.19850720465183258, "learning_rate": 5.411413668954324e-05, "loss": 0.0213, "step": 50010 }, { "grad_norm": 0.1966789960861206, "learning_rate": 5.4097657835432083e-05, "loss": 0.0175, "step": 50020 }, { "grad_norm": 0.19768434762954712, "learning_rate": 5.408117853320723e-05, "loss": 0.0177, "step": 50030 }, { "grad_norm": 0.193183034658432, "learning_rate": 5.406469878467084e-05, "loss": 0.0187, "step": 50040 }, { "grad_norm": 0.22785694897174835, "learning_rate": 5.404821859162509e-05, "loss": 0.0191, "step": 50050 }, { "grad_norm": 0.19786563515663147, "learning_rate": 5.403173795587225e-05, "loss": 0.0196, "step": 50060 }, { "grad_norm": 0.2419421523809433, "learning_rate": 5.4015256879214606e-05, "loss": 0.0179, "step": 50070 }, { "grad_norm": 0.19788064062595367, "learning_rate": 5.39987753634545e-05, "loss": 0.0189, "step": 50080 }, { "grad_norm": 0.26816603541374207, "learning_rate": 5.3982293410394335e-05, "loss": 0.0203, "step": 50090 }, { "grad_norm": 0.26668867468833923, "learning_rate": 5.396581102183655e-05, "loss": 0.0182, "step": 50100 }, { "grad_norm": 0.22511352598667145, "learning_rate": 5.3949328199583634e-05, "loss": 0.0179, "step": 50110 }, { "grad_norm": 0.16295279562473297, "learning_rate": 5.393284494543813e-05, "loss": 0.0189, "step": 50120 }, { "grad_norm": 0.23282219469547272, "learning_rate": 5.391636126120262e-05, "loss": 0.0184, "step": 50130 }, { "grad_norm": 0.20565764605998993, "learning_rate": 5.389987714867971e-05, "loss": 0.0231, "step": 50140 }, { "grad_norm": 0.21058529615402222, "learning_rate": 5.3883392609672114e-05, "loss": 0.0181, "step": 50150 }, { "grad_norm": 0.1890244036912918, "learning_rate": 5.3866907645982545e-05, "loss": 0.0176, "step": 50160 }, { "grad_norm": 0.20630714297294617, "learning_rate": 5.385042225941375e-05, "loss": 0.0203, "step": 50170 }, { "grad_norm": 0.18858957290649414, "learning_rate": 5.3833936451768585e-05, "loss": 0.0166, "step": 50180 }, { "grad_norm": 0.21657806634902954, "learning_rate": 5.3817450224849884e-05, "loss": 0.0187, "step": 50190 }, { "grad_norm": 0.22177769243717194, "learning_rate": 5.380096358046056e-05, "loss": 0.0206, "step": 50200 }, { "grad_norm": 0.17689792811870575, "learning_rate": 5.378447652040359e-05, "loss": 0.0163, "step": 50210 }, { "grad_norm": 0.15691909193992615, "learning_rate": 5.376798904648195e-05, "loss": 0.0191, "step": 50220 }, { "grad_norm": 0.24059423804283142, "learning_rate": 5.375150116049869e-05, "loss": 0.0223, "step": 50230 }, { "grad_norm": 0.17085741460323334, "learning_rate": 5.373501286425691e-05, "loss": 0.0179, "step": 50240 }, { "grad_norm": 0.17208842933177948, "learning_rate": 5.3718524159559726e-05, "loss": 0.0187, "step": 50250 }, { "grad_norm": 0.25025054812431335, "learning_rate": 5.370203504821034e-05, "loss": 0.0189, "step": 50260 }, { "grad_norm": 0.18623869121074677, "learning_rate": 5.368554553201196e-05, "loss": 0.0208, "step": 50270 }, { "grad_norm": 0.21042084693908691, "learning_rate": 5.3669055612767874e-05, "loss": 0.0188, "step": 50280 }, { "grad_norm": 0.18138746917247772, "learning_rate": 5.3652565292281376e-05, "loss": 0.0209, "step": 50290 }, { "grad_norm": 0.2042708843946457, "learning_rate": 5.3636074572355824e-05, "loss": 0.0163, "step": 50300 }, { "grad_norm": 0.18572647869586945, "learning_rate": 5.361958345479463e-05, "loss": 0.0186, "step": 50310 }, { "grad_norm": 0.1758616417646408, "learning_rate": 5.360309194140123e-05, "loss": 0.0172, "step": 50320 }, { "grad_norm": 0.204679936170578, "learning_rate": 5.3586600033979106e-05, "loss": 0.0189, "step": 50330 }, { "grad_norm": 0.24433983862400055, "learning_rate": 5.357010773433181e-05, "loss": 0.0196, "step": 50340 }, { "grad_norm": 0.2564089596271515, "learning_rate": 5.3553615044262895e-05, "loss": 0.0221, "step": 50350 }, { "grad_norm": 0.23350952565670013, "learning_rate": 5.3537121965575964e-05, "loss": 0.0181, "step": 50360 }, { "grad_norm": 0.18323948979377747, "learning_rate": 5.352062850007471e-05, "loss": 0.0158, "step": 50370 }, { "grad_norm": 0.21646539866924286, "learning_rate": 5.350413464956282e-05, "loss": 0.0208, "step": 50380 }, { "grad_norm": 0.18072783946990967, "learning_rate": 5.348764041584403e-05, "loss": 0.018, "step": 50390 }, { "grad_norm": 0.19811058044433594, "learning_rate": 5.347114580072212e-05, "loss": 0.0181, "step": 50400 }, { "grad_norm": 0.15475080907344818, "learning_rate": 5.345465080600093e-05, "loss": 0.0193, "step": 50410 }, { "grad_norm": 0.17784874141216278, "learning_rate": 5.3438155433484326e-05, "loss": 0.0178, "step": 50420 }, { "grad_norm": 0.29061585664749146, "learning_rate": 5.3421659684976197e-05, "loss": 0.0246, "step": 50430 }, { "grad_norm": 0.1361166089773178, "learning_rate": 5.340516356228052e-05, "loss": 0.0197, "step": 50440 }, { "grad_norm": 0.19782385230064392, "learning_rate": 5.338866706720128e-05, "loss": 0.0191, "step": 50450 }, { "grad_norm": 0.2262035310268402, "learning_rate": 5.337217020154249e-05, "loss": 0.0186, "step": 50460 }, { "grad_norm": 0.20843060314655304, "learning_rate": 5.335567296710825e-05, "loss": 0.0179, "step": 50470 }, { "grad_norm": 0.23466990888118744, "learning_rate": 5.333917536570265e-05, "loss": 0.0183, "step": 50480 }, { "grad_norm": 0.13969941437244415, "learning_rate": 5.332267739912986e-05, "loss": 0.019, "step": 50490 }, { "grad_norm": 0.22745482623577118, "learning_rate": 5.330617906919405e-05, "loss": 0.0171, "step": 50500 }, { "grad_norm": 0.23716670274734497, "learning_rate": 5.328968037769949e-05, "loss": 0.018, "step": 50510 }, { "grad_norm": 0.20157361030578613, "learning_rate": 5.327318132645042e-05, "loss": 0.0177, "step": 50520 }, { "grad_norm": 0.22670264542102814, "learning_rate": 5.325668191725116e-05, "loss": 0.0225, "step": 50530 }, { "grad_norm": 0.3690224587917328, "learning_rate": 5.324018215190606e-05, "loss": 0.0176, "step": 50540 }, { "grad_norm": 0.20523123443126678, "learning_rate": 5.3223682032219515e-05, "loss": 0.0209, "step": 50550 }, { "grad_norm": 0.16015706956386566, "learning_rate": 5.320718155999595e-05, "loss": 0.0203, "step": 50560 }, { "grad_norm": 0.21075654029846191, "learning_rate": 5.3190680737039835e-05, "loss": 0.022, "step": 50570 }, { "grad_norm": 0.16838446259498596, "learning_rate": 5.317417956515567e-05, "loss": 0.0189, "step": 50580 }, { "grad_norm": 0.18496350944042206, "learning_rate": 5.315767804614798e-05, "loss": 0.0218, "step": 50590 }, { "grad_norm": 0.1779976487159729, "learning_rate": 5.3141176181821395e-05, "loss": 0.0172, "step": 50600 }, { "grad_norm": 0.27116596698760986, "learning_rate": 5.31246739739805e-05, "loss": 0.0225, "step": 50610 }, { "grad_norm": 0.17146581411361694, "learning_rate": 5.310817142442994e-05, "loss": 0.0179, "step": 50620 }, { "grad_norm": 0.18232490122318268, "learning_rate": 5.309166853497445e-05, "loss": 0.0171, "step": 50630 }, { "grad_norm": 0.16593511402606964, "learning_rate": 5.307516530741873e-05, "loss": 0.0186, "step": 50640 }, { "grad_norm": 0.22048704326152802, "learning_rate": 5.305866174356754e-05, "loss": 0.0156, "step": 50650 }, { "grad_norm": 0.20048904418945312, "learning_rate": 5.304215784522571e-05, "loss": 0.0217, "step": 50660 }, { "grad_norm": 0.266912043094635, "learning_rate": 5.302565361419808e-05, "loss": 0.017, "step": 50670 }, { "grad_norm": 0.1973683089017868, "learning_rate": 5.3009149052289507e-05, "loss": 0.0168, "step": 50680 }, { "grad_norm": 0.2989617586135864, "learning_rate": 5.299264416130493e-05, "loss": 0.0186, "step": 50690 }, { "grad_norm": 0.21740268170833588, "learning_rate": 5.297613894304928e-05, "loss": 0.0163, "step": 50700 }, { "grad_norm": 0.2295757532119751, "learning_rate": 5.2959633399327534e-05, "loss": 0.0196, "step": 50710 }, { "grad_norm": 0.20133253931999207, "learning_rate": 5.294312753194476e-05, "loss": 0.0185, "step": 50720 }, { "grad_norm": 0.2321038395166397, "learning_rate": 5.292662134270596e-05, "loss": 0.0166, "step": 50730 }, { "grad_norm": 0.188203826546669, "learning_rate": 5.291011483341626e-05, "loss": 0.0216, "step": 50740 }, { "grad_norm": 0.1958378702402115, "learning_rate": 5.2893608005880767e-05, "loss": 0.0181, "step": 50750 }, { "grad_norm": 0.27948322892189026, "learning_rate": 5.287710086190467e-05, "loss": 0.0233, "step": 50760 }, { "grad_norm": 0.19401578605175018, "learning_rate": 5.286059340329314e-05, "loss": 0.0216, "step": 50770 }, { "grad_norm": 0.18310992419719696, "learning_rate": 5.284408563185141e-05, "loss": 0.0191, "step": 50780 }, { "grad_norm": 0.1773185431957245, "learning_rate": 5.282757754938476e-05, "loss": 0.0219, "step": 50790 }, { "grad_norm": 0.2137138843536377, "learning_rate": 5.281106915769849e-05, "loss": 0.0183, "step": 50800 }, { "grad_norm": 0.20947708189487457, "learning_rate": 5.2794560458597897e-05, "loss": 0.0237, "step": 50810 }, { "grad_norm": 0.15307903289794922, "learning_rate": 5.27780514538884e-05, "loss": 0.0198, "step": 50820 }, { "grad_norm": 0.1600755900144577, "learning_rate": 5.2761542145375365e-05, "loss": 0.019, "step": 50830 }, { "grad_norm": 0.15935590863227844, "learning_rate": 5.274503253486421e-05, "loss": 0.021, "step": 50840 }, { "grad_norm": 0.2706761360168457, "learning_rate": 5.272852262416046e-05, "loss": 0.0176, "step": 50850 }, { "grad_norm": 0.19064119458198547, "learning_rate": 5.2712012415069555e-05, "loss": 0.0208, "step": 50860 }, { "grad_norm": 0.22589166462421417, "learning_rate": 5.269550190939705e-05, "loss": 0.0215, "step": 50870 }, { "grad_norm": 0.16932512819766998, "learning_rate": 5.267899110894852e-05, "loss": 0.0171, "step": 50880 }, { "grad_norm": 0.1631772816181183, "learning_rate": 5.266248001552955e-05, "loss": 0.0169, "step": 50890 }, { "grad_norm": 0.2375696897506714, "learning_rate": 5.264596863094575e-05, "loss": 0.0174, "step": 50900 }, { "grad_norm": 0.17803914844989777, "learning_rate": 5.262945695700282e-05, "loss": 0.0236, "step": 50910 }, { "grad_norm": 0.2110443264245987, "learning_rate": 5.261294499550643e-05, "loss": 0.0178, "step": 50920 }, { "grad_norm": 0.194864422082901, "learning_rate": 5.25964327482623e-05, "loss": 0.0226, "step": 50930 }, { "grad_norm": 0.77398282289505, "learning_rate": 5.257992021707617e-05, "loss": 0.0216, "step": 50940 }, { "grad_norm": 0.1970011591911316, "learning_rate": 5.256340740375387e-05, "loss": 0.0229, "step": 50950 }, { "grad_norm": 0.1712784320116043, "learning_rate": 5.254689431010117e-05, "loss": 0.0195, "step": 50960 }, { "grad_norm": 0.21295541524887085, "learning_rate": 5.253038093792395e-05, "loss": 0.0196, "step": 50970 }, { "grad_norm": 0.1795375496149063, "learning_rate": 5.251386728902806e-05, "loss": 0.0217, "step": 50980 }, { "grad_norm": 0.20858004689216614, "learning_rate": 5.2497353365219446e-05, "loss": 0.0176, "step": 50990 }, { "grad_norm": 0.19424693286418915, "learning_rate": 5.2480839168304e-05, "loss": 0.0187, "step": 51000 }, { "grad_norm": 0.21871161460876465, "learning_rate": 5.246432470008773e-05, "loss": 0.022, "step": 51010 }, { "grad_norm": 0.19328279793262482, "learning_rate": 5.2447809962376616e-05, "loss": 0.0183, "step": 51020 }, { "grad_norm": 0.2264350950717926, "learning_rate": 5.243129495697666e-05, "loss": 0.02, "step": 51030 }, { "grad_norm": 0.23097608983516693, "learning_rate": 5.241477968569397e-05, "loss": 0.0258, "step": 51040 }, { "grad_norm": 0.23051440715789795, "learning_rate": 5.23982641503346e-05, "loss": 0.0207, "step": 51050 }, { "grad_norm": 0.1960393190383911, "learning_rate": 5.238174835270465e-05, "loss": 0.021, "step": 51060 }, { "grad_norm": 0.22642676532268524, "learning_rate": 5.23652322946103e-05, "loss": 0.0177, "step": 51070 }, { "grad_norm": 0.20434583723545074, "learning_rate": 5.234871597785771e-05, "loss": 0.0208, "step": 51080 }, { "grad_norm": 0.19242815673351288, "learning_rate": 5.233219940425306e-05, "loss": 0.0158, "step": 51090 }, { "grad_norm": 0.1590614914894104, "learning_rate": 5.2315682575602576e-05, "loss": 0.017, "step": 51100 }, { "grad_norm": 0.21921586990356445, "learning_rate": 5.229916549371254e-05, "loss": 0.0194, "step": 51110 }, { "grad_norm": 0.17635725438594818, "learning_rate": 5.228264816038924e-05, "loss": 0.0156, "step": 51120 }, { "grad_norm": 0.1909191757440567, "learning_rate": 5.2266130577438935e-05, "loss": 0.0203, "step": 51130 }, { "grad_norm": 0.2454080432653427, "learning_rate": 5.224961274666802e-05, "loss": 0.0167, "step": 51140 }, { "grad_norm": 0.21499812602996826, "learning_rate": 5.223309466988283e-05, "loss": 0.0191, "step": 51150 }, { "grad_norm": 0.2374224215745926, "learning_rate": 5.221657634888976e-05, "loss": 0.0205, "step": 51160 }, { "grad_norm": 0.17095211148262024, "learning_rate": 5.2200057785495235e-05, "loss": 0.0188, "step": 51170 }, { "grad_norm": 0.25898799300193787, "learning_rate": 5.2183538981505706e-05, "loss": 0.0256, "step": 51180 }, { "grad_norm": 0.23168416321277618, "learning_rate": 5.216701993872762e-05, "loss": 0.0236, "step": 51190 }, { "grad_norm": 0.21424807608127594, "learning_rate": 5.215050065896752e-05, "loss": 0.0179, "step": 51200 }, { "grad_norm": 0.22781042754650116, "learning_rate": 5.2133981144031875e-05, "loss": 0.0169, "step": 51210 }, { "grad_norm": 0.18367476761341095, "learning_rate": 5.211746139572727e-05, "loss": 0.0174, "step": 51220 }, { "grad_norm": 0.1863703429698944, "learning_rate": 5.210094141586026e-05, "loss": 0.0176, "step": 51230 }, { "grad_norm": 0.1746901571750641, "learning_rate": 5.2084421206237476e-05, "loss": 0.0189, "step": 51240 }, { "grad_norm": 0.15870314836502075, "learning_rate": 5.20679007686655e-05, "loss": 0.0189, "step": 51250 }, { "grad_norm": 0.25690385699272156, "learning_rate": 5.205138010495102e-05, "loss": 0.0231, "step": 51260 }, { "grad_norm": 0.1544434130191803, "learning_rate": 5.20348592169007e-05, "loss": 0.0166, "step": 51270 }, { "grad_norm": 0.21586069464683533, "learning_rate": 5.201833810632123e-05, "loss": 0.0187, "step": 51280 }, { "grad_norm": 0.1722850799560547, "learning_rate": 5.200181677501933e-05, "loss": 0.0168, "step": 51290 }, { "grad_norm": 0.19289803504943848, "learning_rate": 5.1985295224801777e-05, "loss": 0.0216, "step": 51300 }, { "grad_norm": 0.13418345153331757, "learning_rate": 5.196877345747531e-05, "loss": 0.0199, "step": 51310 }, { "grad_norm": 0.2152058482170105, "learning_rate": 5.195225147484674e-05, "loss": 0.0161, "step": 51320 }, { "grad_norm": 0.219629168510437, "learning_rate": 5.193572927872288e-05, "loss": 0.0211, "step": 51330 }, { "grad_norm": 0.1881895363330841, "learning_rate": 5.191920687091059e-05, "loss": 0.0184, "step": 51340 }, { "grad_norm": 0.2203981876373291, "learning_rate": 5.1902684253216695e-05, "loss": 0.0194, "step": 51350 }, { "grad_norm": 0.16603907942771912, "learning_rate": 5.1886161427448134e-05, "loss": 0.0183, "step": 51360 }, { "grad_norm": 0.20816676318645477, "learning_rate": 5.186963839541179e-05, "loss": 0.0145, "step": 51370 }, { "grad_norm": 0.34079864621162415, "learning_rate": 5.185311515891459e-05, "loss": 0.0187, "step": 51380 }, { "grad_norm": 0.1836700290441513, "learning_rate": 5.1836591719763516e-05, "loss": 0.0211, "step": 51390 }, { "grad_norm": 0.16910162568092346, "learning_rate": 5.182006807976554e-05, "loss": 0.0176, "step": 51400 }, { "grad_norm": 0.18667413294315338, "learning_rate": 5.180354424072763e-05, "loss": 0.0188, "step": 51410 }, { "grad_norm": 0.16777268052101135, "learning_rate": 5.178702020445686e-05, "loss": 0.0214, "step": 51420 }, { "grad_norm": 0.21074025332927704, "learning_rate": 5.177049597276024e-05, "loss": 0.0159, "step": 51430 }, { "grad_norm": 0.195515438914299, "learning_rate": 5.175397154744482e-05, "loss": 0.0165, "step": 51440 }, { "grad_norm": 0.2402518093585968, "learning_rate": 5.1737446930317736e-05, "loss": 0.0191, "step": 51450 }, { "grad_norm": 0.21760636568069458, "learning_rate": 5.172092212318604e-05, "loss": 0.0183, "step": 51460 }, { "grad_norm": 0.20428207516670227, "learning_rate": 5.170439712785692e-05, "loss": 0.0175, "step": 51470 }, { "grad_norm": 0.1998508870601654, "learning_rate": 5.1687871946137467e-05, "loss": 0.0208, "step": 51480 }, { "grad_norm": 0.18192236125469208, "learning_rate": 5.16713465798349e-05, "loss": 0.0164, "step": 51490 }, { "grad_norm": 0.24895749986171722, "learning_rate": 5.165482103075637e-05, "loss": 0.0181, "step": 51500 }, { "grad_norm": 0.23799175024032593, "learning_rate": 5.163829530070909e-05, "loss": 0.0187, "step": 51510 }, { "grad_norm": 0.21982088685035706, "learning_rate": 5.162176939150032e-05, "loss": 0.0223, "step": 51520 }, { "grad_norm": 0.20620186626911163, "learning_rate": 5.160524330493728e-05, "loss": 0.0178, "step": 51530 }, { "grad_norm": 0.3986780643463135, "learning_rate": 5.158871704282724e-05, "loss": 0.0164, "step": 51540 }, { "grad_norm": 0.20866818726062775, "learning_rate": 5.15721906069775e-05, "loss": 0.0208, "step": 51550 }, { "grad_norm": 0.2972528040409088, "learning_rate": 5.155566399919536e-05, "loss": 0.0184, "step": 51560 }, { "grad_norm": 0.16149239242076874, "learning_rate": 5.153913722128813e-05, "loss": 0.0186, "step": 51570 }, { "grad_norm": 0.2038833647966385, "learning_rate": 5.1522610275063176e-05, "loss": 0.0206, "step": 51580 }, { "grad_norm": 0.15076056122779846, "learning_rate": 5.150608316232786e-05, "loss": 0.0177, "step": 51590 }, { "grad_norm": 0.1897481083869934, "learning_rate": 5.148955588488955e-05, "loss": 0.0191, "step": 51600 }, { "grad_norm": 0.22739070653915405, "learning_rate": 5.147302844455564e-05, "loss": 0.02, "step": 51610 }, { "grad_norm": 0.16443988680839539, "learning_rate": 5.145650084313356e-05, "loss": 0.0167, "step": 51620 }, { "grad_norm": 0.1819653958082199, "learning_rate": 5.1439973082430734e-05, "loss": 0.0168, "step": 51630 }, { "grad_norm": 0.2626769244670868, "learning_rate": 5.142344516425461e-05, "loss": 0.0198, "step": 51640 }, { "grad_norm": 0.17932340502738953, "learning_rate": 5.1406917090412673e-05, "loss": 0.0223, "step": 51650 }, { "grad_norm": 0.20307905972003937, "learning_rate": 5.1390388862712404e-05, "loss": 0.0177, "step": 51660 }, { "grad_norm": 0.222602978348732, "learning_rate": 5.137386048296129e-05, "loss": 0.0213, "step": 51670 }, { "grad_norm": 0.17413488030433655, "learning_rate": 5.135733195296686e-05, "loss": 0.0184, "step": 51680 }, { "grad_norm": 0.1679484248161316, "learning_rate": 5.134080327453664e-05, "loss": 0.018, "step": 51690 }, { "grad_norm": 0.13874070346355438, "learning_rate": 5.1324274449478204e-05, "loss": 0.0193, "step": 51700 }, { "grad_norm": 0.2433299720287323, "learning_rate": 5.130774547959909e-05, "loss": 0.0227, "step": 51710 }, { "grad_norm": 0.17981672286987305, "learning_rate": 5.129121636670692e-05, "loss": 0.0186, "step": 51720 }, { "grad_norm": 0.14494688808918, "learning_rate": 5.1274687112609246e-05, "loss": 0.0158, "step": 51730 }, { "grad_norm": 0.23413236439228058, "learning_rate": 5.125815771911372e-05, "loss": 0.0176, "step": 51740 }, { "grad_norm": 0.17915049195289612, "learning_rate": 5.1241628188027956e-05, "loss": 0.0204, "step": 51750 }, { "grad_norm": 0.1748361736536026, "learning_rate": 5.1225098521159585e-05, "loss": 0.0194, "step": 51760 }, { "grad_norm": 0.16609075665473938, "learning_rate": 5.12085687203163e-05, "loss": 0.0177, "step": 51770 }, { "grad_norm": 0.1974012553691864, "learning_rate": 5.119203878730575e-05, "loss": 0.0179, "step": 51780 }, { "grad_norm": 0.18431678414344788, "learning_rate": 5.117550872393564e-05, "loss": 0.0159, "step": 51790 }, { "grad_norm": 0.191999152302742, "learning_rate": 5.1158978532013645e-05, "loss": 0.0174, "step": 51800 }, { "grad_norm": 0.12403783947229385, "learning_rate": 5.114244821334751e-05, "loss": 0.0193, "step": 51810 }, { "grad_norm": 0.36802566051483154, "learning_rate": 5.112591776974497e-05, "loss": 0.0223, "step": 51820 }, { "grad_norm": 0.2250073254108429, "learning_rate": 5.110938720301374e-05, "loss": 0.0213, "step": 51830 }, { "grad_norm": 0.16545473039150238, "learning_rate": 5.1092856514961595e-05, "loss": 0.0187, "step": 51840 }, { "grad_norm": 0.1788981407880783, "learning_rate": 5.107632570739631e-05, "loss": 0.016, "step": 51850 }, { "grad_norm": 0.1750698983669281, "learning_rate": 5.1059794782125645e-05, "loss": 0.0163, "step": 51860 }, { "grad_norm": 0.23540757596492767, "learning_rate": 5.104326374095744e-05, "loss": 0.0159, "step": 51870 }, { "grad_norm": 0.15044237673282623, "learning_rate": 5.102673258569947e-05, "loss": 0.0174, "step": 51880 }, { "grad_norm": 0.25705403089523315, "learning_rate": 5.101020131815956e-05, "loss": 0.0221, "step": 51890 }, { "grad_norm": 0.3432445228099823, "learning_rate": 5.099366994014556e-05, "loss": 0.0205, "step": 51900 }, { "grad_norm": 0.30550864338874817, "learning_rate": 5.097713845346532e-05, "loss": 0.0192, "step": 51910 }, { "grad_norm": 0.17294278740882874, "learning_rate": 5.096060685992668e-05, "loss": 0.0209, "step": 51920 }, { "grad_norm": 0.18050841987133026, "learning_rate": 5.0944075161337514e-05, "loss": 0.018, "step": 51930 }, { "grad_norm": 0.17770342528820038, "learning_rate": 5.09275433595057e-05, "loss": 0.017, "step": 51940 }, { "grad_norm": 0.22200945019721985, "learning_rate": 5.0911011456239157e-05, "loss": 0.0198, "step": 51950 }, { "grad_norm": 0.21181966364383698, "learning_rate": 5.089447945334574e-05, "loss": 0.0162, "step": 51960 }, { "grad_norm": 0.21389274299144745, "learning_rate": 5.087794735263343e-05, "loss": 0.0206, "step": 51970 }, { "grad_norm": 0.2468903362751007, "learning_rate": 5.086141515591011e-05, "loss": 0.0184, "step": 51980 }, { "grad_norm": 0.20951515436172485, "learning_rate": 5.084488286498371e-05, "loss": 0.0179, "step": 51990 }, { "grad_norm": 0.1716359257698059, "learning_rate": 5.08283504816622e-05, "loss": 0.016, "step": 52000 }, { "grad_norm": 0.1613587737083435, "learning_rate": 5.081181800775353e-05, "loss": 0.0219, "step": 52010 }, { "grad_norm": 0.21184110641479492, "learning_rate": 5.079528544506565e-05, "loss": 0.0241, "step": 52020 }, { "grad_norm": 0.1880985051393509, "learning_rate": 5.077875279540657e-05, "loss": 0.0182, "step": 52030 }, { "grad_norm": 0.19491642713546753, "learning_rate": 5.076222006058424e-05, "loss": 0.0176, "step": 52040 }, { "grad_norm": 0.17871452867984772, "learning_rate": 5.074568724240667e-05, "loss": 0.0201, "step": 52050 }, { "grad_norm": 0.1786375492811203, "learning_rate": 5.072915434268187e-05, "loss": 0.0217, "step": 52060 }, { "grad_norm": 0.1771683394908905, "learning_rate": 5.071262136321785e-05, "loss": 0.0191, "step": 52070 }, { "grad_norm": 0.2217240035533905, "learning_rate": 5.069608830582262e-05, "loss": 0.0207, "step": 52080 }, { "grad_norm": 0.16154597699642181, "learning_rate": 5.067955517230423e-05, "loss": 0.0172, "step": 52090 }, { "grad_norm": 0.2094929963350296, "learning_rate": 5.06630219644707e-05, "loss": 0.0195, "step": 52100 }, { "grad_norm": 0.24546509981155396, "learning_rate": 5.064648868413008e-05, "loss": 0.0175, "step": 52110 }, { "grad_norm": 0.20925262570381165, "learning_rate": 5.062995533309044e-05, "loss": 0.0191, "step": 52120 }, { "grad_norm": 0.145415261387825, "learning_rate": 5.061342191315983e-05, "loss": 0.0171, "step": 52130 }, { "grad_norm": 0.18307630717754364, "learning_rate": 5.0596888426146325e-05, "loss": 0.0202, "step": 52140 }, { "grad_norm": 0.15160392224788666, "learning_rate": 5.058035487385798e-05, "loss": 0.0179, "step": 52150 }, { "grad_norm": 0.2020210325717926, "learning_rate": 5.0563821258102915e-05, "loss": 0.0168, "step": 52160 }, { "grad_norm": 0.16234219074249268, "learning_rate": 5.05472875806892e-05, "loss": 0.0157, "step": 52170 }, { "grad_norm": 0.2123556286096573, "learning_rate": 5.053075384342494e-05, "loss": 0.0206, "step": 52180 }, { "grad_norm": 0.21335123479366302, "learning_rate": 5.0514220048118234e-05, "loss": 0.0201, "step": 52190 }, { "grad_norm": 0.17536111176013947, "learning_rate": 5.049768619657721e-05, "loss": 0.0162, "step": 52200 }, { "grad_norm": 0.18093378841876984, "learning_rate": 5.048115229060996e-05, "loss": 0.0165, "step": 52210 }, { "grad_norm": 0.13487882912158966, "learning_rate": 5.046461833202462e-05, "loss": 0.0159, "step": 52220 }, { "grad_norm": 0.1861899197101593, "learning_rate": 5.044808432262932e-05, "loss": 0.0172, "step": 52230 }, { "grad_norm": 0.18265880644321442, "learning_rate": 5.043155026423219e-05, "loss": 0.0153, "step": 52240 }, { "grad_norm": 0.1589871346950531, "learning_rate": 5.041501615864137e-05, "loss": 0.0169, "step": 52250 }, { "grad_norm": 0.16381308436393738, "learning_rate": 5.039848200766502e-05, "loss": 0.0184, "step": 52260 }, { "grad_norm": 0.17179027199745178, "learning_rate": 5.038194781311125e-05, "loss": 0.0186, "step": 52270 }, { "grad_norm": 0.16935735940933228, "learning_rate": 5.0365413576788264e-05, "loss": 0.0153, "step": 52280 }, { "grad_norm": 0.17801444232463837, "learning_rate": 5.0348879300504194e-05, "loss": 0.0197, "step": 52290 }, { "grad_norm": 0.15386679768562317, "learning_rate": 5.033234498606721e-05, "loss": 0.0148, "step": 52300 }, { "grad_norm": 0.20017729699611664, "learning_rate": 5.0315810635285476e-05, "loss": 0.0165, "step": 52310 }, { "grad_norm": 0.16118590533733368, "learning_rate": 5.0299276249967164e-05, "loss": 0.017, "step": 52320 }, { "grad_norm": 0.2069607824087143, "learning_rate": 5.0282741831920454e-05, "loss": 0.021, "step": 52330 }, { "grad_norm": 0.18800237774848938, "learning_rate": 5.026620738295351e-05, "loss": 0.0189, "step": 52340 }, { "grad_norm": 0.18018080294132233, "learning_rate": 5.024967290487455e-05, "loss": 0.0168, "step": 52350 }, { "grad_norm": 0.1965566724538803, "learning_rate": 5.023313839949172e-05, "loss": 0.0157, "step": 52360 }, { "grad_norm": 0.18444222211837769, "learning_rate": 5.0216603868613224e-05, "loss": 0.0193, "step": 52370 }, { "grad_norm": 0.18883109092712402, "learning_rate": 5.020006931404726e-05, "loss": 0.0175, "step": 52380 }, { "grad_norm": 0.1671755313873291, "learning_rate": 5.0183534737602014e-05, "loss": 0.0209, "step": 52390 }, { "grad_norm": 0.1599888652563095, "learning_rate": 5.0167000141085686e-05, "loss": 0.0165, "step": 52400 }, { "grad_norm": 0.2156515121459961, "learning_rate": 5.015046552630647e-05, "loss": 0.0191, "step": 52410 }, { "grad_norm": 0.22959798574447632, "learning_rate": 5.013393089507258e-05, "loss": 0.0181, "step": 52420 }, { "grad_norm": 0.19275373220443726, "learning_rate": 5.011739624919219e-05, "loss": 0.0194, "step": 52430 }, { "grad_norm": 0.21415597200393677, "learning_rate": 5.0100861590473545e-05, "loss": 0.0181, "step": 52440 }, { "grad_norm": 0.23107126355171204, "learning_rate": 5.008432692072482e-05, "loss": 0.022, "step": 52450 }, { "grad_norm": 0.16566231846809387, "learning_rate": 5.006779224175423e-05, "loss": 0.0184, "step": 52460 }, { "grad_norm": 0.21178552508354187, "learning_rate": 5.005125755536998e-05, "loss": 0.0195, "step": 52470 }, { "grad_norm": 0.2409166395664215, "learning_rate": 5.003472286338029e-05, "loss": 0.0198, "step": 52480 }, { "grad_norm": 0.17558778822422028, "learning_rate": 5.001818816759335e-05, "loss": 0.0148, "step": 52490 }, { "grad_norm": 0.21885259449481964, "learning_rate": 5.000165346981738e-05, "loss": 0.0166, "step": 52500 }, { "grad_norm": 0.17877867817878723, "learning_rate": 4.9985118771860595e-05, "loss": 0.019, "step": 52510 }, { "grad_norm": 0.17970427870750427, "learning_rate": 4.996858407553119e-05, "loss": 0.0149, "step": 52520 }, { "grad_norm": 0.1779971718788147, "learning_rate": 4.99520493826374e-05, "loss": 0.0147, "step": 52530 }, { "grad_norm": 0.1966538280248642, "learning_rate": 4.9935514694987384e-05, "loss": 0.019, "step": 52540 }, { "grad_norm": 0.20137423276901245, "learning_rate": 4.991898001438941e-05, "loss": 0.017, "step": 52550 }, { "grad_norm": 0.14413568377494812, "learning_rate": 4.990244534265166e-05, "loss": 0.0152, "step": 52560 }, { "grad_norm": 0.2148168981075287, "learning_rate": 4.988591068158234e-05, "loss": 0.0165, "step": 52570 }, { "grad_norm": 0.21621008217334747, "learning_rate": 4.986937603298965e-05, "loss": 0.0175, "step": 52580 }, { "grad_norm": 0.24400551617145538, "learning_rate": 4.985284139868181e-05, "loss": 0.0181, "step": 52590 }, { "grad_norm": 0.22176232933998108, "learning_rate": 4.9836306780467e-05, "loss": 0.0185, "step": 52600 }, { "grad_norm": 0.21373122930526733, "learning_rate": 4.981977218015343e-05, "loss": 0.0165, "step": 52610 }, { "grad_norm": 0.1696770340204239, "learning_rate": 4.980323759954932e-05, "loss": 0.0138, "step": 52620 }, { "grad_norm": 0.28422102332115173, "learning_rate": 4.978670304046284e-05, "loss": 0.0177, "step": 52630 }, { "grad_norm": 0.17285794019699097, "learning_rate": 4.9770168504702195e-05, "loss": 0.0183, "step": 52640 }, { "grad_norm": 0.16518527269363403, "learning_rate": 4.975363399407559e-05, "loss": 0.0176, "step": 52650 }, { "grad_norm": 0.15008266270160675, "learning_rate": 4.9737099510391194e-05, "loss": 0.0172, "step": 52660 }, { "grad_norm": 0.23615609109401703, "learning_rate": 4.972056505545719e-05, "loss": 0.0172, "step": 52670 }, { "grad_norm": 0.16743458807468414, "learning_rate": 4.9704030631081775e-05, "loss": 0.0174, "step": 52680 }, { "grad_norm": 0.2514016330242157, "learning_rate": 4.968749623907313e-05, "loss": 0.023, "step": 52690 }, { "grad_norm": 0.21818822622299194, "learning_rate": 4.967096188123943e-05, "loss": 0.0179, "step": 52700 }, { "grad_norm": 0.1718541979789734, "learning_rate": 4.965442755938884e-05, "loss": 0.0168, "step": 52710 }, { "grad_norm": 0.2197979837656021, "learning_rate": 4.963789327532954e-05, "loss": 0.0189, "step": 52720 }, { "grad_norm": 0.199501171708107, "learning_rate": 4.962135903086968e-05, "loss": 0.0162, "step": 52730 }, { "grad_norm": 0.10407627373933792, "learning_rate": 4.960482482781741e-05, "loss": 0.0153, "step": 52740 }, { "grad_norm": 0.2373294234275818, "learning_rate": 4.958829066798092e-05, "loss": 0.0173, "step": 52750 }, { "grad_norm": 0.24470870196819305, "learning_rate": 4.957175655316835e-05, "loss": 0.0197, "step": 52760 }, { "grad_norm": 0.16729934513568878, "learning_rate": 4.9555222485187834e-05, "loss": 0.0185, "step": 52770 }, { "grad_norm": 0.18246106803417206, "learning_rate": 4.953868846584752e-05, "loss": 0.0217, "step": 52780 }, { "grad_norm": 0.1658976674079895, "learning_rate": 4.9522154496955544e-05, "loss": 0.0158, "step": 52790 }, { "grad_norm": 0.19990001618862152, "learning_rate": 4.950562058032001e-05, "loss": 0.0176, "step": 52800 }, { "grad_norm": 0.20498113334178925, "learning_rate": 4.948908671774909e-05, "loss": 0.0181, "step": 52810 }, { "grad_norm": 0.21610209345817566, "learning_rate": 4.947255291105087e-05, "loss": 0.0158, "step": 52820 }, { "grad_norm": 0.19312845170497894, "learning_rate": 4.945601916203348e-05, "loss": 0.0172, "step": 52830 }, { "grad_norm": 0.20135056972503662, "learning_rate": 4.9439485472505016e-05, "loss": 0.0177, "step": 52840 }, { "grad_norm": 0.2339525669813156, "learning_rate": 4.942295184427356e-05, "loss": 0.0189, "step": 52850 }, { "grad_norm": 0.16235478222370148, "learning_rate": 4.9406418279147234e-05, "loss": 0.0159, "step": 52860 }, { "grad_norm": 0.18791580200195312, "learning_rate": 4.9389884778934116e-05, "loss": 0.0194, "step": 52870 }, { "grad_norm": 0.2081955075263977, "learning_rate": 4.937335134544228e-05, "loss": 0.0198, "step": 52880 }, { "grad_norm": 0.19078899919986725, "learning_rate": 4.935681798047979e-05, "loss": 0.018, "step": 52890 }, { "grad_norm": 0.26427674293518066, "learning_rate": 4.934028468585473e-05, "loss": 0.0175, "step": 52900 }, { "grad_norm": 0.21231664717197418, "learning_rate": 4.932375146337514e-05, "loss": 0.0196, "step": 52910 }, { "grad_norm": 0.2536777853965759, "learning_rate": 4.930721831484908e-05, "loss": 0.0172, "step": 52920 }, { "grad_norm": 0.2603636682033539, "learning_rate": 4.929068524208456e-05, "loss": 0.021, "step": 52930 }, { "grad_norm": 0.22147977352142334, "learning_rate": 4.927415224688965e-05, "loss": 0.0192, "step": 52940 }, { "grad_norm": 0.1969612091779709, "learning_rate": 4.9257619331072375e-05, "loss": 0.019, "step": 52950 }, { "grad_norm": 0.17770755290985107, "learning_rate": 4.924108649644072e-05, "loss": 0.0187, "step": 52960 }, { "grad_norm": 0.15070481598377228, "learning_rate": 4.922455374480271e-05, "loss": 0.0196, "step": 52970 }, { "grad_norm": 0.13602215051651, "learning_rate": 4.920802107796634e-05, "loss": 0.0187, "step": 52980 }, { "grad_norm": 0.19046777486801147, "learning_rate": 4.9191488497739564e-05, "loss": 0.0174, "step": 52990 }, { "grad_norm": 0.19723449647426605, "learning_rate": 4.917495600593042e-05, "loss": 0.0156, "step": 53000 }, { "grad_norm": 0.2054336965084076, "learning_rate": 4.915842360434684e-05, "loss": 0.0165, "step": 53010 }, { "grad_norm": 0.19099515676498413, "learning_rate": 4.914189129479679e-05, "loss": 0.0177, "step": 53020 }, { "grad_norm": 0.21450111269950867, "learning_rate": 4.912535907908822e-05, "loss": 0.0179, "step": 53030 }, { "grad_norm": 0.18042561411857605, "learning_rate": 4.910882695902907e-05, "loss": 0.0158, "step": 53040 }, { "grad_norm": 0.29853305220603943, "learning_rate": 4.9092294936427234e-05, "loss": 0.0208, "step": 53050 }, { "grad_norm": 0.18398329615592957, "learning_rate": 4.9075763013090685e-05, "loss": 0.0171, "step": 53060 }, { "grad_norm": 0.15196596086025238, "learning_rate": 4.9059231190827305e-05, "loss": 0.0172, "step": 53070 }, { "grad_norm": 0.20948119461536407, "learning_rate": 4.904269947144499e-05, "loss": 0.0247, "step": 53080 }, { "grad_norm": 0.24223899841308594, "learning_rate": 4.902616785675161e-05, "loss": 0.0159, "step": 53090 }, { "grad_norm": 0.1819378286600113, "learning_rate": 4.900963634855505e-05, "loss": 0.0154, "step": 53100 }, { "grad_norm": 0.21979719400405884, "learning_rate": 4.8993104948663173e-05, "loss": 0.0175, "step": 53110 }, { "grad_norm": 0.17658206820487976, "learning_rate": 4.8976573658883823e-05, "loss": 0.0197, "step": 53120 }, { "grad_norm": 0.21798540651798248, "learning_rate": 4.896004248102484e-05, "loss": 0.0193, "step": 53130 }, { "grad_norm": 0.1796303242444992, "learning_rate": 4.8943511416894045e-05, "loss": 0.0213, "step": 53140 }, { "grad_norm": 0.1997530311346054, "learning_rate": 4.892698046829926e-05, "loss": 0.0208, "step": 53150 }, { "grad_norm": 0.19855763018131256, "learning_rate": 4.8910449637048275e-05, "loss": 0.0179, "step": 53160 }, { "grad_norm": 0.2147134691476822, "learning_rate": 4.8893918924948884e-05, "loss": 0.018, "step": 53170 }, { "grad_norm": 0.17783978581428528, "learning_rate": 4.8877388333808835e-05, "loss": 0.0176, "step": 53180 }, { "grad_norm": 0.19467869400978088, "learning_rate": 4.8860857865435925e-05, "loss": 0.0168, "step": 53190 }, { "grad_norm": 0.16041319072246552, "learning_rate": 4.884432752163789e-05, "loss": 0.0188, "step": 53200 }, { "grad_norm": 0.16521745920181274, "learning_rate": 4.8827797304222466e-05, "loss": 0.0244, "step": 53210 }, { "grad_norm": 0.2043655961751938, "learning_rate": 4.8811267214997366e-05, "loss": 0.0183, "step": 53220 }, { "grad_norm": 0.1924675703048706, "learning_rate": 4.879473725577029e-05, "loss": 0.0253, "step": 53230 }, { "grad_norm": 0.18620802462100983, "learning_rate": 4.877820742834891e-05, "loss": 0.0152, "step": 53240 }, { "grad_norm": 0.2123568058013916, "learning_rate": 4.8761677734540956e-05, "loss": 0.0194, "step": 53250 }, { "grad_norm": 0.21102839708328247, "learning_rate": 4.874514817615407e-05, "loss": 0.0194, "step": 53260 }, { "grad_norm": 0.1386861354112625, "learning_rate": 4.872861875499588e-05, "loss": 0.0148, "step": 53270 }, { "grad_norm": 0.2457357496023178, "learning_rate": 4.871208947287404e-05, "loss": 0.0192, "step": 53280 }, { "grad_norm": 0.25526466965675354, "learning_rate": 4.869556033159615e-05, "loss": 0.0168, "step": 53290 }, { "grad_norm": 0.20179340243339539, "learning_rate": 4.8679031332969835e-05, "loss": 0.0191, "step": 53300 }, { "grad_norm": 0.2087644636631012, "learning_rate": 4.8662502478802635e-05, "loss": 0.0175, "step": 53310 }, { "grad_norm": 0.2408612221479416, "learning_rate": 4.8645973770902176e-05, "loss": 0.0179, "step": 53320 }, { "grad_norm": 0.2751012444496155, "learning_rate": 4.862944521107599e-05, "loss": 0.0177, "step": 53330 }, { "grad_norm": 0.2265009582042694, "learning_rate": 4.861291680113162e-05, "loss": 0.0169, "step": 53340 }, { "grad_norm": 0.2878219783306122, "learning_rate": 4.8596388542876563e-05, "loss": 0.0186, "step": 53350 }, { "grad_norm": 0.18015360832214355, "learning_rate": 4.857986043811836e-05, "loss": 0.0167, "step": 53360 }, { "grad_norm": 0.18427421152591705, "learning_rate": 4.856333248866448e-05, "loss": 0.0196, "step": 53370 }, { "grad_norm": 0.20640522241592407, "learning_rate": 4.85468046963224e-05, "loss": 0.0156, "step": 53380 }, { "grad_norm": 0.21990151703357697, "learning_rate": 4.853027706289957e-05, "loss": 0.0183, "step": 53390 }, { "grad_norm": 0.20460593700408936, "learning_rate": 4.851374959020344e-05, "loss": 0.0168, "step": 53400 }, { "grad_norm": 0.1780027449131012, "learning_rate": 4.849722228004141e-05, "loss": 0.0168, "step": 53410 }, { "grad_norm": 0.19898028671741486, "learning_rate": 4.848069513422091e-05, "loss": 0.0177, "step": 53420 }, { "grad_norm": 0.18983685970306396, "learning_rate": 4.8464168154549294e-05, "loss": 0.0176, "step": 53430 }, { "grad_norm": 0.175596684217453, "learning_rate": 4.844764134283392e-05, "loss": 0.0152, "step": 53440 }, { "grad_norm": 0.18753448128700256, "learning_rate": 4.843111470088218e-05, "loss": 0.017, "step": 53450 }, { "grad_norm": 0.16745205223560333, "learning_rate": 4.841458823050137e-05, "loss": 0.0147, "step": 53460 }, { "grad_norm": 0.2652800977230072, "learning_rate": 4.8398061933498816e-05, "loss": 0.0157, "step": 53470 }, { "grad_norm": 0.12738727033138275, "learning_rate": 4.83815358116818e-05, "loss": 0.0144, "step": 53480 }, { "grad_norm": 0.17846368253231049, "learning_rate": 4.8365009866857584e-05, "loss": 0.0182, "step": 53490 }, { "grad_norm": 0.18995298445224762, "learning_rate": 4.834848410083342e-05, "loss": 0.018, "step": 53500 }, { "grad_norm": 0.16695819795131683, "learning_rate": 4.833195851541657e-05, "loss": 0.0145, "step": 53510 }, { "grad_norm": 0.15496401488780975, "learning_rate": 4.831543311241423e-05, "loss": 0.0184, "step": 53520 }, { "grad_norm": 0.1853494942188263, "learning_rate": 4.829890789363359e-05, "loss": 0.0139, "step": 53530 }, { "grad_norm": 0.2713161110877991, "learning_rate": 4.828238286088182e-05, "loss": 0.016, "step": 53540 }, { "grad_norm": 0.3153771162033081, "learning_rate": 4.826585801596609e-05, "loss": 0.0218, "step": 53550 }, { "grad_norm": 0.21918196976184845, "learning_rate": 4.8249333360693486e-05, "loss": 0.0172, "step": 53560 }, { "grad_norm": 0.11798092722892761, "learning_rate": 4.8232808896871185e-05, "loss": 0.014, "step": 53570 }, { "grad_norm": 0.1540452390909195, "learning_rate": 4.821628462630624e-05, "loss": 0.0148, "step": 53580 }, { "grad_norm": 0.26728305220603943, "learning_rate": 4.8199760550805735e-05, "loss": 0.0164, "step": 53590 }, { "grad_norm": 0.15761145949363708, "learning_rate": 4.818323667217669e-05, "loss": 0.0191, "step": 53600 }, { "grad_norm": 0.20211035013198853, "learning_rate": 4.8166712992226165e-05, "loss": 0.0168, "step": 53610 }, { "grad_norm": 0.1348038911819458, "learning_rate": 4.8150189512761154e-05, "loss": 0.0137, "step": 53620 }, { "grad_norm": 0.14055471122264862, "learning_rate": 4.813366623558862e-05, "loss": 0.0174, "step": 53630 }, { "grad_norm": 0.1400933712720871, "learning_rate": 4.811714316251554e-05, "loss": 0.0157, "step": 53640 }, { "grad_norm": 0.2764463424682617, "learning_rate": 4.810062029534886e-05, "loss": 0.0162, "step": 53650 }, { "grad_norm": 0.1907973438501358, "learning_rate": 4.8084097635895505e-05, "loss": 0.0174, "step": 53660 }, { "grad_norm": 0.17026115953922272, "learning_rate": 4.8067575185962335e-05, "loss": 0.0157, "step": 53670 }, { "grad_norm": 0.1597965657711029, "learning_rate": 4.805105294735625e-05, "loss": 0.0164, "step": 53680 }, { "grad_norm": 0.24431000649929047, "learning_rate": 4.803453092188406e-05, "loss": 0.0184, "step": 53690 }, { "grad_norm": 0.18367600440979004, "learning_rate": 4.801800911135263e-05, "loss": 0.0156, "step": 53700 }, { "grad_norm": 0.16134867072105408, "learning_rate": 4.8001487517568744e-05, "loss": 0.015, "step": 53710 }, { "grad_norm": 0.21979352831840515, "learning_rate": 4.798496614233918e-05, "loss": 0.0172, "step": 53720 }, { "grad_norm": 0.2566852271556854, "learning_rate": 4.796844498747069e-05, "loss": 0.022, "step": 53730 }, { "grad_norm": 0.1969691962003708, "learning_rate": 4.795192405477e-05, "loss": 0.0189, "step": 53740 }, { "grad_norm": 0.2601926922798157, "learning_rate": 4.79354033460438e-05, "loss": 0.0186, "step": 53750 }, { "grad_norm": 0.16770431399345398, "learning_rate": 4.79188828630988e-05, "loss": 0.0166, "step": 53760 }, { "grad_norm": 0.17408670485019684, "learning_rate": 4.790236260774165e-05, "loss": 0.0163, "step": 53770 }, { "grad_norm": 0.2401031255722046, "learning_rate": 4.788584258177896e-05, "loss": 0.0226, "step": 53780 }, { "grad_norm": 0.19851863384246826, "learning_rate": 4.7869322787017356e-05, "loss": 0.0185, "step": 53790 }, { "grad_norm": 0.18796640634536743, "learning_rate": 4.78528032252634e-05, "loss": 0.0205, "step": 53800 }, { "grad_norm": 0.1909628063440323, "learning_rate": 4.783628389832366e-05, "loss": 0.0149, "step": 53810 }, { "grad_norm": 0.15678347647190094, "learning_rate": 4.7819764808004644e-05, "loss": 0.0197, "step": 53820 }, { "grad_norm": 0.14899976551532745, "learning_rate": 4.780324595611289e-05, "loss": 0.0177, "step": 53830 }, { "grad_norm": 0.19537580013275146, "learning_rate": 4.7786727344454854e-05, "loss": 0.0159, "step": 53840 }, { "grad_norm": 0.2044019252061844, "learning_rate": 4.777020897483697e-05, "loss": 0.0167, "step": 53850 }, { "grad_norm": 0.19415411353111267, "learning_rate": 4.775369084906569e-05, "loss": 0.0167, "step": 53860 }, { "grad_norm": 0.254534512758255, "learning_rate": 4.773717296894742e-05, "loss": 0.0183, "step": 53870 }, { "grad_norm": 0.1697891801595688, "learning_rate": 4.772065533628848e-05, "loss": 0.0194, "step": 53880 }, { "grad_norm": 0.1731739640235901, "learning_rate": 4.7704137952895265e-05, "loss": 0.0207, "step": 53890 }, { "grad_norm": 0.22904613614082336, "learning_rate": 4.768762082057407e-05, "loss": 0.0213, "step": 53900 }, { "grad_norm": 0.1507786363363266, "learning_rate": 4.7671103941131186e-05, "loss": 0.0176, "step": 53910 }, { "grad_norm": 0.2569246292114258, "learning_rate": 4.765458731637287e-05, "loss": 0.0216, "step": 53920 }, { "grad_norm": 0.2856162190437317, "learning_rate": 4.763807094810537e-05, "loss": 0.0161, "step": 53930 }, { "grad_norm": 0.21007134020328522, "learning_rate": 4.762155483813487e-05, "loss": 0.0193, "step": 53940 }, { "grad_norm": 0.1628027856349945, "learning_rate": 4.760503898826754e-05, "loss": 0.0167, "step": 53950 }, { "grad_norm": 0.16406196355819702, "learning_rate": 4.758852340030957e-05, "loss": 0.0168, "step": 53960 }, { "grad_norm": 0.15324005484580994, "learning_rate": 4.757200807606705e-05, "loss": 0.0172, "step": 53970 }, { "grad_norm": 0.17887412011623383, "learning_rate": 4.755549301734607e-05, "loss": 0.0166, "step": 53980 }, { "grad_norm": 0.14389097690582275, "learning_rate": 4.753897822595271e-05, "loss": 0.0165, "step": 53990 }, { "grad_norm": 0.15024146437644958, "learning_rate": 4.7522463703692994e-05, "loss": 0.0171, "step": 54000 }, { "grad_norm": 0.1801372468471527, "learning_rate": 4.75059494523729e-05, "loss": 0.019, "step": 54010 }, { "grad_norm": 0.21379618346691132, "learning_rate": 4.748943547379843e-05, "loss": 0.0162, "step": 54020 }, { "grad_norm": 0.1981593370437622, "learning_rate": 4.7472921769775535e-05, "loss": 0.0164, "step": 54030 }, { "grad_norm": 0.17345909774303436, "learning_rate": 4.745640834211011e-05, "loss": 0.0175, "step": 54040 }, { "grad_norm": 0.19860048592090607, "learning_rate": 4.7439895192608054e-05, "loss": 0.0201, "step": 54050 }, { "grad_norm": 0.20043827593326569, "learning_rate": 4.7423382323075186e-05, "loss": 0.0151, "step": 54060 }, { "grad_norm": 0.2117568850517273, "learning_rate": 4.740686973531737e-05, "loss": 0.0142, "step": 54070 }, { "grad_norm": 0.2094627320766449, "learning_rate": 4.7390357431140375e-05, "loss": 0.0181, "step": 54080 }, { "grad_norm": 0.17938371002674103, "learning_rate": 4.737384541234998e-05, "loss": 0.0153, "step": 54090 }, { "grad_norm": 0.2754381597042084, "learning_rate": 4.735733368075189e-05, "loss": 0.016, "step": 54100 }, { "grad_norm": 0.25724345445632935, "learning_rate": 4.734082223815182e-05, "loss": 0.0162, "step": 54110 }, { "grad_norm": 0.174764484167099, "learning_rate": 4.7324311086355436e-05, "loss": 0.017, "step": 54120 }, { "grad_norm": 0.2668500244617462, "learning_rate": 4.730780022716836e-05, "loss": 0.0208, "step": 54130 }, { "grad_norm": 0.2097952663898468, "learning_rate": 4.729128966239619e-05, "loss": 0.0139, "step": 54140 }, { "grad_norm": 0.15335135161876678, "learning_rate": 4.7274779393844536e-05, "loss": 0.0171, "step": 54150 }, { "grad_norm": 0.2146928757429123, "learning_rate": 4.725826942331891e-05, "loss": 0.0181, "step": 54160 }, { "grad_norm": 0.17014601826667786, "learning_rate": 4.724175975262481e-05, "loss": 0.018, "step": 54170 }, { "grad_norm": 0.17732849717140198, "learning_rate": 4.722525038356772e-05, "loss": 0.0158, "step": 54180 }, { "grad_norm": 0.20378224551677704, "learning_rate": 4.7208741317953075e-05, "loss": 0.0179, "step": 54190 }, { "grad_norm": 0.18974116444587708, "learning_rate": 4.7192232557586264e-05, "loss": 0.0172, "step": 54200 }, { "grad_norm": 0.16525810956954956, "learning_rate": 4.717572410427269e-05, "loss": 0.0144, "step": 54210 }, { "grad_norm": 0.2123980075120926, "learning_rate": 4.71592159598177e-05, "loss": 0.0152, "step": 54220 }, { "grad_norm": 0.16948454082012177, "learning_rate": 4.714270812602657e-05, "loss": 0.0147, "step": 54230 }, { "grad_norm": 0.26383334398269653, "learning_rate": 4.712620060470458e-05, "loss": 0.0154, "step": 54240 }, { "grad_norm": 0.21710149943828583, "learning_rate": 4.7109693397656986e-05, "loss": 0.0164, "step": 54250 }, { "grad_norm": 0.17716918885707855, "learning_rate": 4.709318650668894e-05, "loss": 0.0178, "step": 54260 }, { "grad_norm": 0.2374572604894638, "learning_rate": 4.7076679933605676e-05, "loss": 0.0152, "step": 54270 }, { "grad_norm": 0.2262469232082367, "learning_rate": 4.70601736802123e-05, "loss": 0.0173, "step": 54280 }, { "grad_norm": 0.19382157921791077, "learning_rate": 4.704366774831391e-05, "loss": 0.0204, "step": 54290 }, { "grad_norm": 0.2029658704996109, "learning_rate": 4.702716213971557e-05, "loss": 0.0145, "step": 54300 }, { "grad_norm": 0.20784422755241394, "learning_rate": 4.70106568562223e-05, "loss": 0.0204, "step": 54310 }, { "grad_norm": 0.27380022406578064, "learning_rate": 4.6994151899639105e-05, "loss": 0.0178, "step": 54320 }, { "grad_norm": 0.18623684346675873, "learning_rate": 4.697764727177093e-05, "loss": 0.0197, "step": 54330 }, { "grad_norm": 0.25320419669151306, "learning_rate": 4.696114297442272e-05, "loss": 0.0206, "step": 54340 }, { "grad_norm": 0.16759394109249115, "learning_rate": 4.694463900939933e-05, "loss": 0.0166, "step": 54350 }, { "grad_norm": 0.25158700346946716, "learning_rate": 4.692813537850564e-05, "loss": 0.0148, "step": 54360 }, { "grad_norm": 0.1787882298231125, "learning_rate": 4.6911632083546445e-05, "loss": 0.018, "step": 54370 }, { "grad_norm": 0.1549343317747116, "learning_rate": 4.6895129126326525e-05, "loss": 0.0215, "step": 54380 }, { "grad_norm": 0.17165489494800568, "learning_rate": 4.687862650865059e-05, "loss": 0.0165, "step": 54390 }, { "grad_norm": 0.26767998933792114, "learning_rate": 4.686212423232339e-05, "loss": 0.0155, "step": 54400 }, { "grad_norm": 0.19500108063220978, "learning_rate": 4.6845622299149565e-05, "loss": 0.0178, "step": 54410 }, { "grad_norm": 0.16350232064723969, "learning_rate": 4.682912071093374e-05, "loss": 0.0152, "step": 54420 }, { "grad_norm": 0.13045965135097504, "learning_rate": 4.681261946948052e-05, "loss": 0.0144, "step": 54430 }, { "grad_norm": 0.24577012658119202, "learning_rate": 4.679611857659443e-05, "loss": 0.0192, "step": 54440 }, { "grad_norm": 0.17237335443496704, "learning_rate": 4.6779618034079975e-05, "loss": 0.0177, "step": 54450 }, { "grad_norm": 0.2255227118730545, "learning_rate": 4.6763117843741675e-05, "loss": 0.017, "step": 54460 }, { "grad_norm": 0.24083787202835083, "learning_rate": 4.674661800738394e-05, "loss": 0.0171, "step": 54470 }, { "grad_norm": 0.24311983585357666, "learning_rate": 4.673011852681117e-05, "loss": 0.0172, "step": 54480 }, { "grad_norm": 0.23613202571868896, "learning_rate": 4.671361940382772e-05, "loss": 0.0181, "step": 54490 }, { "grad_norm": 0.19194097816944122, "learning_rate": 4.669712064023791e-05, "loss": 0.019, "step": 54500 }, { "grad_norm": 0.20906893908977509, "learning_rate": 4.668062223784602e-05, "loss": 0.0168, "step": 54510 }, { "grad_norm": 0.2062123864889145, "learning_rate": 4.6664124198456274e-05, "loss": 0.0193, "step": 54520 }, { "grad_norm": 0.18896543979644775, "learning_rate": 4.6647626523872905e-05, "loss": 0.0174, "step": 54530 }, { "grad_norm": 0.15175166726112366, "learning_rate": 4.6631129215900066e-05, "loss": 0.013, "step": 54540 }, { "grad_norm": 0.20194120705127716, "learning_rate": 4.661463227634187e-05, "loss": 0.0169, "step": 54550 }, { "grad_norm": 0.17741453647613525, "learning_rate": 4.659813570700238e-05, "loss": 0.0152, "step": 54560 }, { "grad_norm": 0.19723202288150787, "learning_rate": 4.658163950968568e-05, "loss": 0.0187, "step": 54570 }, { "grad_norm": 0.11597965657711029, "learning_rate": 4.6565143686195726e-05, "loss": 0.0188, "step": 54580 }, { "grad_norm": 0.15188652276992798, "learning_rate": 4.654864823833649e-05, "loss": 0.0188, "step": 54590 }, { "grad_norm": 0.2516043484210968, "learning_rate": 4.65321531679119e-05, "loss": 0.0208, "step": 54600 }, { "grad_norm": 0.24495171010494232, "learning_rate": 4.6515658476725834e-05, "loss": 0.0175, "step": 54610 }, { "grad_norm": 0.14473946392536163, "learning_rate": 4.649916416658212e-05, "loss": 0.0171, "step": 54620 }, { "grad_norm": 0.13656897842884064, "learning_rate": 4.648267023928455e-05, "loss": 0.0177, "step": 54630 }, { "grad_norm": 0.23302559554576874, "learning_rate": 4.646617669663687e-05, "loss": 0.0198, "step": 54640 }, { "grad_norm": 0.19568239152431488, "learning_rate": 4.644968354044277e-05, "loss": 0.0148, "step": 54650 }, { "grad_norm": 0.17210909724235535, "learning_rate": 4.643319077250596e-05, "loss": 0.017, "step": 54660 }, { "grad_norm": 0.15490463376045227, "learning_rate": 4.6416698394630055e-05, "loss": 0.0166, "step": 54670 }, { "grad_norm": 0.19876883924007416, "learning_rate": 4.640020640861862e-05, "loss": 0.0167, "step": 54680 }, { "grad_norm": 0.15973815321922302, "learning_rate": 4.6383714816275195e-05, "loss": 0.0163, "step": 54690 }, { "grad_norm": 0.1520192176103592, "learning_rate": 4.636722361940329e-05, "loss": 0.0169, "step": 54700 }, { "grad_norm": 0.3880350589752197, "learning_rate": 4.635073281980632e-05, "loss": 0.0161, "step": 54710 }, { "grad_norm": 0.16094830632209778, "learning_rate": 4.633424241928775e-05, "loss": 0.0166, "step": 54720 }, { "grad_norm": 0.17375244200229645, "learning_rate": 4.6317752419650914e-05, "loss": 0.0183, "step": 54730 }, { "grad_norm": 0.1663019210100174, "learning_rate": 4.630126282269914e-05, "loss": 0.0196, "step": 54740 }, { "grad_norm": 0.19338643550872803, "learning_rate": 4.628477363023569e-05, "loss": 0.0199, "step": 54750 }, { "grad_norm": 0.25631415843963623, "learning_rate": 4.626828484406383e-05, "loss": 0.0206, "step": 54760 }, { "grad_norm": 0.34447792172431946, "learning_rate": 4.625179646598669e-05, "loss": 0.0212, "step": 54770 }, { "grad_norm": 0.23537515103816986, "learning_rate": 4.6235308497807487e-05, "loss": 0.0175, "step": 54780 }, { "grad_norm": 0.20735473930835724, "learning_rate": 4.6218820941329276e-05, "loss": 0.0169, "step": 54790 }, { "grad_norm": 0.20901688933372498, "learning_rate": 4.620233379835513e-05, "loss": 0.0155, "step": 54800 }, { "grad_norm": 0.18528564274311066, "learning_rate": 4.618584707068804e-05, "loss": 0.0158, "step": 54810 }, { "grad_norm": 0.12304843962192535, "learning_rate": 4.616936076013099e-05, "loss": 0.016, "step": 54820 }, { "grad_norm": 0.13650435209274292, "learning_rate": 4.615287486848688e-05, "loss": 0.0164, "step": 54830 }, { "grad_norm": 0.15609993040561676, "learning_rate": 4.6136389397558584e-05, "loss": 0.0177, "step": 54840 }, { "grad_norm": 0.19433647394180298, "learning_rate": 4.6119904349148935e-05, "loss": 0.0166, "step": 54850 }, { "grad_norm": 0.15596823394298553, "learning_rate": 4.6103419725060726e-05, "loss": 0.0226, "step": 54860 }, { "grad_norm": 0.20864105224609375, "learning_rate": 4.608693552709667e-05, "loss": 0.0188, "step": 54870 }, { "grad_norm": 0.24392375349998474, "learning_rate": 4.607045175705947e-05, "loss": 0.0167, "step": 54880 }, { "grad_norm": 0.1675659418106079, "learning_rate": 4.605396841675176e-05, "loss": 0.0184, "step": 54890 }, { "grad_norm": 0.2463560551404953, "learning_rate": 4.6037485507976094e-05, "loss": 0.0188, "step": 54900 }, { "grad_norm": 0.15723058581352234, "learning_rate": 4.602100303253509e-05, "loss": 0.0135, "step": 54910 }, { "grad_norm": 0.1767248958349228, "learning_rate": 4.600452099223122e-05, "loss": 0.0175, "step": 54920 }, { "grad_norm": 0.2013268619775772, "learning_rate": 4.598803938886691e-05, "loss": 0.0212, "step": 54930 }, { "grad_norm": 0.14853248000144958, "learning_rate": 4.597155822424459e-05, "loss": 0.0181, "step": 54940 }, { "grad_norm": 0.17493605613708496, "learning_rate": 4.59550775001666e-05, "loss": 0.0174, "step": 54950 }, { "grad_norm": 0.20420828461647034, "learning_rate": 4.5938597218435234e-05, "loss": 0.0158, "step": 54960 }, { "grad_norm": 0.16627489030361176, "learning_rate": 4.592211738085279e-05, "loss": 0.0171, "step": 54970 }, { "grad_norm": 0.17844940721988678, "learning_rate": 4.590563798922146e-05, "loss": 0.0155, "step": 54980 }, { "grad_norm": 0.2201845496892929, "learning_rate": 4.5889159045343404e-05, "loss": 0.0201, "step": 54990 }, { "grad_norm": 0.2478088140487671, "learning_rate": 4.5872680551020734e-05, "loss": 0.0164, "step": 55000 }, { "grad_norm": 0.20414228737354279, "learning_rate": 4.585620250805552e-05, "loss": 0.0181, "step": 55010 }, { "grad_norm": 0.14432214200496674, "learning_rate": 4.583972491824975e-05, "loss": 0.0141, "step": 55020 }, { "grad_norm": 0.1817087084054947, "learning_rate": 4.582324778340541e-05, "loss": 0.016, "step": 55030 }, { "grad_norm": 0.1837722510099411, "learning_rate": 4.5806771105324424e-05, "loss": 0.0152, "step": 55040 }, { "grad_norm": 0.14316731691360474, "learning_rate": 4.579029488580865e-05, "loss": 0.0155, "step": 55050 }, { "grad_norm": 0.19493705034255981, "learning_rate": 4.5773819126659876e-05, "loss": 0.0171, "step": 55060 }, { "grad_norm": 0.19490782916545868, "learning_rate": 4.575734382967991e-05, "loss": 0.0165, "step": 55070 }, { "grad_norm": 0.22265557944774628, "learning_rate": 4.574086899667044e-05, "loss": 0.0175, "step": 55080 }, { "grad_norm": 0.1595577597618103, "learning_rate": 4.5724394629433106e-05, "loss": 0.0177, "step": 55090 }, { "grad_norm": 0.1909007430076599, "learning_rate": 4.570792072976958e-05, "loss": 0.0166, "step": 55100 }, { "grad_norm": 0.13392497599124908, "learning_rate": 4.569144729948138e-05, "loss": 0.0145, "step": 55110 }, { "grad_norm": 0.11188097298145294, "learning_rate": 4.5674974340370033e-05, "loss": 0.0134, "step": 55120 }, { "grad_norm": 0.2505453824996948, "learning_rate": 4.565850185423699e-05, "loss": 0.0175, "step": 55130 }, { "grad_norm": 0.19147108495235443, "learning_rate": 4.564202984288365e-05, "loss": 0.0177, "step": 55140 }, { "grad_norm": 0.21049368381500244, "learning_rate": 4.5625558308111356e-05, "loss": 0.0182, "step": 55150 }, { "grad_norm": 0.14221204817295074, "learning_rate": 4.560908725172145e-05, "loss": 0.0172, "step": 55160 }, { "grad_norm": 0.23783043026924133, "learning_rate": 4.559261667551516e-05, "loss": 0.0163, "step": 55170 }, { "grad_norm": 0.16083788871765137, "learning_rate": 4.5576146581293685e-05, "loss": 0.0172, "step": 55180 }, { "grad_norm": 0.16702508926391602, "learning_rate": 4.5559676970858164e-05, "loss": 0.0157, "step": 55190 }, { "grad_norm": 0.2036718726158142, "learning_rate": 4.5543207846009705e-05, "loss": 0.0171, "step": 55200 }, { "grad_norm": 0.20687328279018402, "learning_rate": 4.552673920854933e-05, "loss": 0.0187, "step": 55210 }, { "grad_norm": 0.11969535052776337, "learning_rate": 4.5510271060278006e-05, "loss": 0.0168, "step": 55220 }, { "grad_norm": 0.15528561174869537, "learning_rate": 4.5493803402996704e-05, "loss": 0.019, "step": 55230 }, { "grad_norm": 0.17341452836990356, "learning_rate": 4.5477336238506295e-05, "loss": 0.0197, "step": 55240 }, { "grad_norm": 0.2105669379234314, "learning_rate": 4.546086956860759e-05, "loss": 0.0167, "step": 55250 }, { "grad_norm": 0.24300748109817505, "learning_rate": 4.5444403395101374e-05, "loss": 0.0209, "step": 55260 }, { "grad_norm": 0.2023223489522934, "learning_rate": 4.542793771978834e-05, "loss": 0.0194, "step": 55270 }, { "grad_norm": 0.22005383670330048, "learning_rate": 4.541147254446916e-05, "loss": 0.0159, "step": 55280 }, { "grad_norm": 0.14905931055545807, "learning_rate": 4.539500787094445e-05, "loss": 0.0166, "step": 55290 }, { "grad_norm": 0.17268501222133636, "learning_rate": 4.537854370101475e-05, "loss": 0.0169, "step": 55300 }, { "grad_norm": 0.23853424191474915, "learning_rate": 4.5362080036480566e-05, "loss": 0.016, "step": 55310 }, { "grad_norm": 0.17298924922943115, "learning_rate": 4.534561687914235e-05, "loss": 0.0178, "step": 55320 }, { "grad_norm": 0.17888493835926056, "learning_rate": 4.5329154230800466e-05, "loss": 0.0152, "step": 55330 }, { "grad_norm": 0.1640983372926712, "learning_rate": 4.531269209325525e-05, "loss": 0.0164, "step": 55340 }, { "grad_norm": 0.24608896672725677, "learning_rate": 4.529623046830696e-05, "loss": 0.0156, "step": 55350 }, { "grad_norm": 0.14393644034862518, "learning_rate": 4.5279769357755866e-05, "loss": 0.0155, "step": 55360 }, { "grad_norm": 0.19821074604988098, "learning_rate": 4.5263308763402084e-05, "loss": 0.0179, "step": 55370 }, { "grad_norm": 0.1623850017786026, "learning_rate": 4.524684868704574e-05, "loss": 0.0148, "step": 55380 }, { "grad_norm": 0.17750486731529236, "learning_rate": 4.523038913048688e-05, "loss": 0.0178, "step": 55390 }, { "grad_norm": 0.15604625642299652, "learning_rate": 4.521393009552549e-05, "loss": 0.0167, "step": 55400 }, { "grad_norm": 0.17960037291049957, "learning_rate": 4.5197471583961484e-05, "loss": 0.0162, "step": 55410 }, { "grad_norm": 0.184526264667511, "learning_rate": 4.51810135975948e-05, "loss": 0.0165, "step": 55420 }, { "grad_norm": 0.17132991552352905, "learning_rate": 4.516455613822521e-05, "loss": 0.0148, "step": 55430 }, { "grad_norm": 0.1920565664768219, "learning_rate": 4.514809920765249e-05, "loss": 0.0159, "step": 55440 }, { "grad_norm": 0.21003638207912445, "learning_rate": 4.5131642807676356e-05, "loss": 0.0167, "step": 55450 }, { "grad_norm": 0.28564003109931946, "learning_rate": 4.511518694009644e-05, "loss": 0.0168, "step": 55460 }, { "grad_norm": 0.154961496591568, "learning_rate": 4.509873160671231e-05, "loss": 0.0159, "step": 55470 }, { "grad_norm": 0.1995086967945099, "learning_rate": 4.508227680932356e-05, "loss": 0.017, "step": 55480 }, { "grad_norm": 0.2679462134838104, "learning_rate": 4.5065822549729605e-05, "loss": 0.017, "step": 55490 }, { "grad_norm": 0.17327845096588135, "learning_rate": 4.504936882972989e-05, "loss": 0.016, "step": 55500 }, { "grad_norm": 0.2110484540462494, "learning_rate": 4.503291565112374e-05, "loss": 0.0156, "step": 55510 }, { "grad_norm": 0.194601908326149, "learning_rate": 4.501646301571048e-05, "loss": 0.0182, "step": 55520 }, { "grad_norm": 0.1875770539045334, "learning_rate": 4.500001092528933e-05, "loss": 0.0184, "step": 55530 }, { "grad_norm": 0.24645066261291504, "learning_rate": 4.4983559381659455e-05, "loss": 0.0202, "step": 55540 }, { "grad_norm": 0.24508869647979736, "learning_rate": 4.496710838661999e-05, "loss": 0.0145, "step": 55550 }, { "grad_norm": 0.18845711648464203, "learning_rate": 4.495065794196999e-05, "loss": 0.0145, "step": 55560 }, { "grad_norm": 0.15134887397289276, "learning_rate": 4.4934208049508455e-05, "loss": 0.0145, "step": 55570 }, { "grad_norm": 0.1885426789522171, "learning_rate": 4.4917758711034303e-05, "loss": 0.0173, "step": 55580 }, { "grad_norm": 0.27432188391685486, "learning_rate": 4.490130992834642e-05, "loss": 0.0178, "step": 55590 }, { "grad_norm": 0.17586830258369446, "learning_rate": 4.488486170324359e-05, "loss": 0.0172, "step": 55600 }, { "grad_norm": 0.18269231915473938, "learning_rate": 4.486841403752463e-05, "loss": 0.0193, "step": 55610 }, { "grad_norm": 0.14355838298797607, "learning_rate": 4.4851966932988174e-05, "loss": 0.0171, "step": 55620 }, { "grad_norm": 0.20465441048145294, "learning_rate": 4.4835520391432894e-05, "loss": 0.0215, "step": 55630 }, { "grad_norm": 0.1850769966840744, "learning_rate": 4.481907441465733e-05, "loss": 0.0172, "step": 55640 }, { "grad_norm": 0.2732677459716797, "learning_rate": 4.480262900446e-05, "loss": 0.0175, "step": 55650 }, { "grad_norm": 0.23991082608699799, "learning_rate": 4.478618416263933e-05, "loss": 0.0135, "step": 55660 }, { "grad_norm": 0.18959146738052368, "learning_rate": 4.476973989099373e-05, "loss": 0.0163, "step": 55670 }, { "grad_norm": 0.17737755179405212, "learning_rate": 4.475329619132153e-05, "loss": 0.0173, "step": 55680 }, { "grad_norm": 0.17035117745399475, "learning_rate": 4.4736853065420956e-05, "loss": 0.018, "step": 55690 }, { "grad_norm": 0.21928498148918152, "learning_rate": 4.472041051509023e-05, "loss": 0.0149, "step": 55700 }, { "grad_norm": 0.21979361772537231, "learning_rate": 4.470396854212746e-05, "loss": 0.0173, "step": 55710 }, { "grad_norm": 0.18327566981315613, "learning_rate": 4.468752714833072e-05, "loss": 0.0166, "step": 55720 }, { "grad_norm": 0.17268109321594238, "learning_rate": 4.467108633549803e-05, "loss": 0.0161, "step": 55730 }, { "grad_norm": 0.17533040046691895, "learning_rate": 4.4654646105427335e-05, "loss": 0.0164, "step": 55740 }, { "grad_norm": 0.18896418809890747, "learning_rate": 4.463820645991651e-05, "loss": 0.0187, "step": 55750 }, { "grad_norm": 0.2280692160129547, "learning_rate": 4.462176740076334e-05, "loss": 0.0165, "step": 55760 }, { "grad_norm": 0.1774933785200119, "learning_rate": 4.4605328929765624e-05, "loss": 0.0175, "step": 55770 }, { "grad_norm": 0.18454809486865997, "learning_rate": 4.458889104872102e-05, "loss": 0.0193, "step": 55780 }, { "grad_norm": 0.21704097092151642, "learning_rate": 4.457245375942715e-05, "loss": 0.0166, "step": 55790 }, { "grad_norm": 0.20698893070220947, "learning_rate": 4.455601706368157e-05, "loss": 0.0174, "step": 55800 }, { "grad_norm": 0.17144958674907684, "learning_rate": 4.453958096328179e-05, "loss": 0.0171, "step": 55810 }, { "grad_norm": 0.16838376224040985, "learning_rate": 4.452314546002523e-05, "loss": 0.023, "step": 55820 }, { "grad_norm": 0.161187544465065, "learning_rate": 4.4506710555709246e-05, "loss": 0.0178, "step": 55830 }, { "grad_norm": 0.23250912129878998, "learning_rate": 4.449027625213114e-05, "loss": 0.0169, "step": 55840 }, { "grad_norm": 0.2135433405637741, "learning_rate": 4.4473842551088135e-05, "loss": 0.0176, "step": 55850 }, { "grad_norm": 0.18289534747600555, "learning_rate": 4.445740945437738e-05, "loss": 0.0188, "step": 55860 }, { "grad_norm": 0.2390894740819931, "learning_rate": 4.444097696379602e-05, "loss": 0.0153, "step": 55870 }, { "grad_norm": 0.2051011323928833, "learning_rate": 4.442454508114106e-05, "loss": 0.0158, "step": 55880 }, { "grad_norm": 0.17828932404518127, "learning_rate": 4.4408113808209465e-05, "loss": 0.0192, "step": 55890 }, { "grad_norm": 0.20350125432014465, "learning_rate": 4.439168314679813e-05, "loss": 0.0204, "step": 55900 }, { "grad_norm": 0.1364467293024063, "learning_rate": 4.43752530987039e-05, "loss": 0.0169, "step": 55910 }, { "grad_norm": 0.183481365442276, "learning_rate": 4.4358823665723515e-05, "loss": 0.017, "step": 55920 }, { "grad_norm": 0.2514084577560425, "learning_rate": 4.434239484965371e-05, "loss": 0.0195, "step": 55930 }, { "grad_norm": 0.15478165447711945, "learning_rate": 4.4325966652291103e-05, "loss": 0.0184, "step": 55940 }, { "grad_norm": 0.2107049822807312, "learning_rate": 4.430953907543225e-05, "loss": 0.0158, "step": 55950 }, { "grad_norm": 0.1911424845457077, "learning_rate": 4.4293112120873645e-05, "loss": 0.0165, "step": 55960 }, { "grad_norm": 0.13204145431518555, "learning_rate": 4.427668579041172e-05, "loss": 0.0166, "step": 55970 }, { "grad_norm": 0.25783783197402954, "learning_rate": 4.4260260085842816e-05, "loss": 0.0198, "step": 55980 }, { "grad_norm": 0.22027702629566193, "learning_rate": 4.4243835008963256e-05, "loss": 0.014, "step": 55990 }, { "grad_norm": 0.1478758305311203, "learning_rate": 4.422741056156925e-05, "loss": 0.0186, "step": 56000 }, { "grad_norm": 0.23505590856075287, "learning_rate": 4.421098674545693e-05, "loss": 0.0165, "step": 56010 }, { "grad_norm": 0.1871444135904312, "learning_rate": 4.419456356242241e-05, "loss": 0.0196, "step": 56020 }, { "grad_norm": 0.21127928793430328, "learning_rate": 4.4178141014261685e-05, "loss": 0.0166, "step": 56030 }, { "grad_norm": 0.13468404114246368, "learning_rate": 4.416171910277071e-05, "loss": 0.0162, "step": 56040 }, { "grad_norm": 0.21623221039772034, "learning_rate": 4.414529782974533e-05, "loss": 0.0169, "step": 56050 }, { "grad_norm": 0.18940520286560059, "learning_rate": 4.41288771969814e-05, "loss": 0.0137, "step": 56060 }, { "grad_norm": 0.1282171607017517, "learning_rate": 4.411245720627464e-05, "loss": 0.0142, "step": 56070 }, { "grad_norm": 0.1766720712184906, "learning_rate": 4.40960378594207e-05, "loss": 0.0146, "step": 56080 }, { "grad_norm": 0.1164395660161972, "learning_rate": 4.407961915821519e-05, "loss": 0.018, "step": 56090 }, { "grad_norm": 0.19126182794570923, "learning_rate": 4.406320110445362e-05, "loss": 0.0169, "step": 56100 }, { "grad_norm": 0.2013719528913498, "learning_rate": 4.404678369993144e-05, "loss": 0.0183, "step": 56110 }, { "grad_norm": 0.30089256167411804, "learning_rate": 4.403036694644406e-05, "loss": 0.0179, "step": 56120 }, { "grad_norm": 0.31183189153671265, "learning_rate": 4.4013950845786764e-05, "loss": 0.0159, "step": 56130 }, { "grad_norm": 0.17198523879051208, "learning_rate": 4.399753539975482e-05, "loss": 0.0165, "step": 56140 }, { "grad_norm": 0.13651567697525024, "learning_rate": 4.398112061014337e-05, "loss": 0.0127, "step": 56150 }, { "grad_norm": 0.2155802845954895, "learning_rate": 4.396470647874753e-05, "loss": 0.0174, "step": 56160 }, { "grad_norm": 0.20466957986354828, "learning_rate": 4.394829300736229e-05, "loss": 0.0141, "step": 56170 }, { "grad_norm": 0.2295379787683487, "learning_rate": 4.393188019778265e-05, "loss": 0.0137, "step": 56180 }, { "grad_norm": 0.16236600279808044, "learning_rate": 4.391546805180347e-05, "loss": 0.0157, "step": 56190 }, { "grad_norm": 0.2003297507762909, "learning_rate": 4.389905657121955e-05, "loss": 0.0198, "step": 56200 }, { "grad_norm": 0.12644733488559723, "learning_rate": 4.3882645757825635e-05, "loss": 0.0132, "step": 56210 }, { "grad_norm": 0.2470024973154068, "learning_rate": 4.386623561341637e-05, "loss": 0.0164, "step": 56220 }, { "grad_norm": 0.20162558555603027, "learning_rate": 4.384982613978637e-05, "loss": 0.0138, "step": 56230 }, { "grad_norm": 0.18436017632484436, "learning_rate": 4.383341733873012e-05, "loss": 0.0147, "step": 56240 }, { "grad_norm": 0.1621575951576233, "learning_rate": 4.381700921204209e-05, "loss": 0.0145, "step": 56250 }, { "grad_norm": 0.17961831390857697, "learning_rate": 4.380060176151663e-05, "loss": 0.0143, "step": 56260 }, { "grad_norm": 0.14522138237953186, "learning_rate": 4.378419498894805e-05, "loss": 0.0166, "step": 56270 }, { "grad_norm": 0.2968105971813202, "learning_rate": 4.376778889613056e-05, "loss": 0.0161, "step": 56280 }, { "grad_norm": 0.19409199059009552, "learning_rate": 4.375138348485831e-05, "loss": 0.0128, "step": 56290 }, { "grad_norm": 0.3228069841861725, "learning_rate": 4.3734978756925346e-05, "loss": 0.0138, "step": 56300 }, { "grad_norm": 0.1941477507352829, "learning_rate": 4.3718574714125706e-05, "loss": 0.0144, "step": 56310 }, { "grad_norm": 0.16958044469356537, "learning_rate": 4.370217135825329e-05, "loss": 0.0168, "step": 56320 }, { "grad_norm": 0.24575383961200714, "learning_rate": 4.368576869110194e-05, "loss": 0.0203, "step": 56330 }, { "grad_norm": 0.16711732745170593, "learning_rate": 4.366936671446544e-05, "loss": 0.0157, "step": 56340 }, { "grad_norm": 0.1614135205745697, "learning_rate": 4.3652965430137474e-05, "loss": 0.0169, "step": 56350 }, { "grad_norm": 0.12186034768819809, "learning_rate": 4.3636564839911646e-05, "loss": 0.0144, "step": 56360 }, { "grad_norm": 0.17536813020706177, "learning_rate": 4.3620164945581545e-05, "loss": 0.0164, "step": 56370 }, { "grad_norm": 0.18108434975147247, "learning_rate": 4.360376574894061e-05, "loss": 0.0184, "step": 56380 }, { "grad_norm": 0.14115005731582642, "learning_rate": 4.358736725178224e-05, "loss": 0.0137, "step": 56390 }, { "grad_norm": 0.2822391390800476, "learning_rate": 4.357096945589974e-05, "loss": 0.0183, "step": 56400 }, { "grad_norm": 0.18193216621875763, "learning_rate": 4.355457236308636e-05, "loss": 0.0169, "step": 56410 }, { "grad_norm": 0.1704966127872467, "learning_rate": 4.353817597513526e-05, "loss": 0.0167, "step": 56420 }, { "grad_norm": 0.21301373839378357, "learning_rate": 4.352178029383948e-05, "loss": 0.0171, "step": 56430 }, { "grad_norm": 0.18118034303188324, "learning_rate": 4.3505385320992105e-05, "loss": 0.0179, "step": 56440 }, { "grad_norm": 0.1639486849308014, "learning_rate": 4.348899105838602e-05, "loss": 0.0158, "step": 56450 }, { "grad_norm": 0.17893265187740326, "learning_rate": 4.3472597507814087e-05, "loss": 0.0163, "step": 56460 }, { "grad_norm": 0.1557965874671936, "learning_rate": 4.3456204671069066e-05, "loss": 0.0179, "step": 56470 }, { "grad_norm": 0.17195817828178406, "learning_rate": 4.343981254994367e-05, "loss": 0.0196, "step": 56480 }, { "grad_norm": 0.18079593777656555, "learning_rate": 4.34234211462305e-05, "loss": 0.0151, "step": 56490 }, { "grad_norm": 0.1635551154613495, "learning_rate": 4.340703046172213e-05, "loss": 0.0177, "step": 56500 }, { "grad_norm": 0.16592194139957428, "learning_rate": 4.339064049821097e-05, "loss": 0.0174, "step": 56510 }, { "grad_norm": 0.180813267827034, "learning_rate": 4.3374251257489446e-05, "loss": 0.0152, "step": 56520 }, { "grad_norm": 0.1928117573261261, "learning_rate": 4.335786274134984e-05, "loss": 0.0146, "step": 56530 }, { "grad_norm": 0.19186030328273773, "learning_rate": 4.334147495158438e-05, "loss": 0.0143, "step": 56540 }, { "grad_norm": 0.1552610695362091, "learning_rate": 4.33250878899852e-05, "loss": 0.0167, "step": 56550 }, { "grad_norm": 0.14910458028316498, "learning_rate": 4.330870155834437e-05, "loss": 0.0169, "step": 56560 }, { "grad_norm": 0.2221510112285614, "learning_rate": 4.3292315958453884e-05, "loss": 0.014, "step": 56570 }, { "grad_norm": 0.45411962270736694, "learning_rate": 4.3275931092105646e-05, "loss": 0.0178, "step": 56580 }, { "grad_norm": 0.1852976530790329, "learning_rate": 4.325954696109147e-05, "loss": 0.0194, "step": 56590 }, { "grad_norm": 0.20628497004508972, "learning_rate": 4.32431635672031e-05, "loss": 0.0155, "step": 56600 }, { "grad_norm": 0.1635064333677292, "learning_rate": 4.3226780912232214e-05, "loss": 0.0162, "step": 56610 }, { "grad_norm": 0.21373334527015686, "learning_rate": 4.321039899797036e-05, "loss": 0.0184, "step": 56620 }, { "grad_norm": 0.27025607228279114, "learning_rate": 4.319401782620908e-05, "loss": 0.0167, "step": 56630 }, { "grad_norm": 0.19766215980052948, "learning_rate": 4.317763739873978e-05, "loss": 0.0195, "step": 56640 }, { "grad_norm": 0.29506662487983704, "learning_rate": 4.31612577173538e-05, "loss": 0.0137, "step": 56650 }, { "grad_norm": 0.1941450834274292, "learning_rate": 4.314487878384239e-05, "loss": 0.0164, "step": 56660 }, { "grad_norm": 0.18226531147956848, "learning_rate": 4.3128500599996736e-05, "loss": 0.0158, "step": 56670 }, { "grad_norm": 0.18456731736660004, "learning_rate": 4.31121231676079e-05, "loss": 0.0153, "step": 56680 }, { "grad_norm": 0.18744826316833496, "learning_rate": 4.309574648846694e-05, "loss": 0.0137, "step": 56690 }, { "grad_norm": 0.22332029044628143, "learning_rate": 4.3079370564364755e-05, "loss": 0.0178, "step": 56700 }, { "grad_norm": 0.1592743843793869, "learning_rate": 4.306299539709221e-05, "loss": 0.0144, "step": 56710 }, { "grad_norm": 0.19344156980514526, "learning_rate": 4.3046620988440046e-05, "loss": 0.014, "step": 56720 }, { "grad_norm": 0.196241095662117, "learning_rate": 4.303024734019897e-05, "loss": 0.0154, "step": 56730 }, { "grad_norm": 0.18295417726039886, "learning_rate": 4.3013874454159565e-05, "loss": 0.0148, "step": 56740 }, { "grad_norm": 0.2280624359846115, "learning_rate": 4.299750233211233e-05, "loss": 0.0161, "step": 56750 }, { "grad_norm": 0.16314314305782318, "learning_rate": 4.2981130975847715e-05, "loss": 0.0168, "step": 56760 }, { "grad_norm": 0.1685449481010437, "learning_rate": 4.2964760387156076e-05, "loss": 0.0159, "step": 56770 }, { "grad_norm": 0.39652034640312195, "learning_rate": 4.294839056782767e-05, "loss": 0.0175, "step": 56780 }, { "grad_norm": 0.1921456754207611, "learning_rate": 4.293202151965266e-05, "loss": 0.0139, "step": 56790 }, { "grad_norm": 0.18624064326286316, "learning_rate": 4.291565324442115e-05, "loss": 0.0158, "step": 56800 }, { "grad_norm": 0.20567768812179565, "learning_rate": 4.2899285743923135e-05, "loss": 0.0207, "step": 56810 }, { "grad_norm": 0.13451558351516724, "learning_rate": 4.288291901994857e-05, "loss": 0.0145, "step": 56820 }, { "grad_norm": 0.23591893911361694, "learning_rate": 4.286655307428729e-05, "loss": 0.0176, "step": 56830 }, { "grad_norm": 0.13876913487911224, "learning_rate": 4.285018790872903e-05, "loss": 0.0142, "step": 56840 }, { "grad_norm": 0.20933431386947632, "learning_rate": 4.283382352506347e-05, "loss": 0.0181, "step": 56850 }, { "grad_norm": 0.18621721863746643, "learning_rate": 4.2817459925080205e-05, "loss": 0.0138, "step": 56860 }, { "grad_norm": 0.18960709869861603, "learning_rate": 4.28010971105687e-05, "loss": 0.0146, "step": 56870 }, { "grad_norm": 0.18067660927772522, "learning_rate": 4.278473508331841e-05, "loss": 0.0152, "step": 56880 }, { "grad_norm": 0.1525789052248001, "learning_rate": 4.276837384511864e-05, "loss": 0.0162, "step": 56890 }, { "grad_norm": 0.23588821291923523, "learning_rate": 4.275201339775864e-05, "loss": 0.0154, "step": 56900 }, { "grad_norm": 0.17072780430316925, "learning_rate": 4.273565374302756e-05, "loss": 0.0133, "step": 56910 }, { "grad_norm": 0.19302774965763092, "learning_rate": 4.2719294882714454e-05, "loss": 0.0178, "step": 56920 }, { "grad_norm": 0.20194697380065918, "learning_rate": 4.270293681860831e-05, "loss": 0.0176, "step": 56930 }, { "grad_norm": 0.15932632982730865, "learning_rate": 4.268657955249802e-05, "loss": 0.0154, "step": 56940 }, { "grad_norm": 0.19599542021751404, "learning_rate": 4.267022308617241e-05, "loss": 0.0171, "step": 56950 }, { "grad_norm": 0.19407175481319427, "learning_rate": 4.265386742142018e-05, "loss": 0.019, "step": 56960 }, { "grad_norm": 0.1378016620874405, "learning_rate": 4.263751256002996e-05, "loss": 0.0147, "step": 56970 }, { "grad_norm": 0.2333369106054306, "learning_rate": 4.26211585037903e-05, "loss": 0.0164, "step": 56980 }, { "grad_norm": 0.21948833763599396, "learning_rate": 4.2604805254489656e-05, "loss": 0.0162, "step": 56990 }, { "grad_norm": 0.1501385122537613, "learning_rate": 4.258845281391638e-05, "loss": 0.0141, "step": 57000 }, { "grad_norm": 0.19958513975143433, "learning_rate": 4.257210118385876e-05, "loss": 0.0144, "step": 57010 }, { "grad_norm": 0.17757046222686768, "learning_rate": 4.2555750366104994e-05, "loss": 0.0166, "step": 57020 }, { "grad_norm": 0.19276756048202515, "learning_rate": 4.2539400362443184e-05, "loss": 0.0156, "step": 57030 }, { "grad_norm": 0.22990307211875916, "learning_rate": 4.252305117466133e-05, "loss": 0.0153, "step": 57040 }, { "grad_norm": 0.1869966834783554, "learning_rate": 4.2506702804547356e-05, "loss": 0.0178, "step": 57050 }, { "grad_norm": 0.19360709190368652, "learning_rate": 4.249035525388907e-05, "loss": 0.0196, "step": 57060 }, { "grad_norm": 0.180223748087883, "learning_rate": 4.247400852447427e-05, "loss": 0.0147, "step": 57070 }, { "grad_norm": 0.433184951543808, "learning_rate": 4.245766261809059e-05, "loss": 0.0171, "step": 57080 }, { "grad_norm": 0.15796172618865967, "learning_rate": 4.244131753652558e-05, "loss": 0.0199, "step": 57090 }, { "grad_norm": 0.18807369470596313, "learning_rate": 4.242497328156672e-05, "loss": 0.016, "step": 57100 }, { "grad_norm": 0.21383066475391388, "learning_rate": 4.240862985500139e-05, "loss": 0.0172, "step": 57110 }, { "grad_norm": 0.212602898478508, "learning_rate": 4.2392287258616894e-05, "loss": 0.0161, "step": 57120 }, { "grad_norm": 0.17743998765945435, "learning_rate": 4.2375945494200396e-05, "loss": 0.0139, "step": 57130 }, { "grad_norm": 0.1492079198360443, "learning_rate": 4.2359604563539054e-05, "loss": 0.0163, "step": 57140 }, { "grad_norm": 0.171275332570076, "learning_rate": 4.2343264468419885e-05, "loss": 0.0149, "step": 57150 }, { "grad_norm": 0.1366473287343979, "learning_rate": 4.232692521062979e-05, "loss": 0.0145, "step": 57160 }, { "grad_norm": 0.21796809136867523, "learning_rate": 4.231058679195562e-05, "loss": 0.0182, "step": 57170 }, { "grad_norm": 0.14142222702503204, "learning_rate": 4.229424921418411e-05, "loss": 0.0147, "step": 57180 }, { "grad_norm": 0.17423872649669647, "learning_rate": 4.2277912479101916e-05, "loss": 0.0146, "step": 57190 }, { "grad_norm": 0.16065746545791626, "learning_rate": 4.226157658849562e-05, "loss": 0.0161, "step": 57200 }, { "grad_norm": 0.17676763236522675, "learning_rate": 4.2245241544151664e-05, "loss": 0.0138, "step": 57210 }, { "grad_norm": 0.18273206055164337, "learning_rate": 4.2228907347856425e-05, "loss": 0.0145, "step": 57220 }, { "grad_norm": 0.12375643849372864, "learning_rate": 4.2212574001396214e-05, "loss": 0.0187, "step": 57230 }, { "grad_norm": 0.14617489278316498, "learning_rate": 4.21962415065572e-05, "loss": 0.0126, "step": 57240 }, { "grad_norm": 0.17119942605495453, "learning_rate": 4.217990986512548e-05, "loss": 0.0202, "step": 57250 }, { "grad_norm": 0.18658889830112457, "learning_rate": 4.216357907888704e-05, "loss": 0.0142, "step": 57260 }, { "grad_norm": 0.18726947903633118, "learning_rate": 4.2147249149627824e-05, "loss": 0.015, "step": 57270 }, { "grad_norm": 0.20456476509571075, "learning_rate": 4.213092007913364e-05, "loss": 0.0132, "step": 57280 }, { "grad_norm": 0.14569726586341858, "learning_rate": 4.2114591869190205e-05, "loss": 0.0175, "step": 57290 }, { "grad_norm": 0.1508222073316574, "learning_rate": 4.209826452158315e-05, "loss": 0.016, "step": 57300 }, { "grad_norm": 0.19629083573818207, "learning_rate": 4.2081938038098e-05, "loss": 0.0141, "step": 57310 }, { "grad_norm": 0.20028476417064667, "learning_rate": 4.2065612420520185e-05, "loss": 0.0156, "step": 57320 }, { "grad_norm": 0.16419363021850586, "learning_rate": 4.204928767063509e-05, "loss": 0.0158, "step": 57330 }, { "grad_norm": 0.20859356224536896, "learning_rate": 4.2032963790227944e-05, "loss": 0.0123, "step": 57340 }, { "grad_norm": 0.2260105013847351, "learning_rate": 4.2016640781083887e-05, "loss": 0.0161, "step": 57350 }, { "grad_norm": 0.17954909801483154, "learning_rate": 4.2000318644988004e-05, "loss": 0.0167, "step": 57360 }, { "grad_norm": 0.2558594346046448, "learning_rate": 4.198399738372524e-05, "loss": 0.0199, "step": 57370 }, { "grad_norm": 0.22062768042087555, "learning_rate": 4.196767699908045e-05, "loss": 0.017, "step": 57380 }, { "grad_norm": 0.14507220685482025, "learning_rate": 4.195135749283845e-05, "loss": 0.0188, "step": 57390 }, { "grad_norm": 0.1509903222322464, "learning_rate": 4.1935038866783894e-05, "loss": 0.0151, "step": 57400 }, { "grad_norm": 0.1645057201385498, "learning_rate": 4.1918721122701355e-05, "loss": 0.0155, "step": 57410 }, { "grad_norm": 0.17511384189128876, "learning_rate": 4.190240426237533e-05, "loss": 0.0168, "step": 57420 }, { "grad_norm": 0.25353094935417175, "learning_rate": 4.188608828759019e-05, "loss": 0.0148, "step": 57430 }, { "grad_norm": 0.15386183559894562, "learning_rate": 4.1869773200130246e-05, "loss": 0.014, "step": 57440 }, { "grad_norm": 0.16690610349178314, "learning_rate": 4.185345900177967e-05, "loss": 0.0187, "step": 57450 }, { "grad_norm": 0.19724036753177643, "learning_rate": 4.183714569432258e-05, "loss": 0.0152, "step": 57460 }, { "grad_norm": 0.19335944950580597, "learning_rate": 4.182083327954295e-05, "loss": 0.0144, "step": 57470 }, { "grad_norm": 0.18038824200630188, "learning_rate": 4.1804521759224714e-05, "loss": 0.0143, "step": 57480 }, { "grad_norm": 0.2261608988046646, "learning_rate": 4.178821113515165e-05, "loss": 0.0189, "step": 57490 }, { "grad_norm": 0.1766573041677475, "learning_rate": 4.1771901409107476e-05, "loss": 0.0168, "step": 57500 }, { "grad_norm": 0.15626437962055206, "learning_rate": 4.175559258287577e-05, "loss": 0.0193, "step": 57510 }, { "grad_norm": 0.16423599421977997, "learning_rate": 4.173928465824008e-05, "loss": 0.0142, "step": 57520 }, { "grad_norm": 0.1838427633047104, "learning_rate": 4.17229776369838e-05, "loss": 0.021, "step": 57530 }, { "grad_norm": 0.17405179142951965, "learning_rate": 4.170667152089025e-05, "loss": 0.0129, "step": 57540 }, { "grad_norm": 0.10691853612661362, "learning_rate": 4.1690366311742626e-05, "loss": 0.0173, "step": 57550 }, { "grad_norm": 0.15434366464614868, "learning_rate": 4.1674062011324055e-05, "loss": 0.0154, "step": 57560 }, { "grad_norm": 0.16107887029647827, "learning_rate": 4.165775862141752e-05, "loss": 0.0146, "step": 57570 }, { "grad_norm": 0.17249929904937744, "learning_rate": 4.164145614380598e-05, "loss": 0.0156, "step": 57580 }, { "grad_norm": 0.16889898478984833, "learning_rate": 4.162515458027223e-05, "loss": 0.0144, "step": 57590 }, { "grad_norm": 0.15860480070114136, "learning_rate": 4.160885393259899e-05, "loss": 0.0146, "step": 57600 }, { "grad_norm": 0.2044202834367752, "learning_rate": 4.159255420256886e-05, "loss": 0.0152, "step": 57610 }, { "grad_norm": 0.21196582913398743, "learning_rate": 4.157625539196436e-05, "loss": 0.0171, "step": 57620 }, { "grad_norm": 0.19739042222499847, "learning_rate": 4.155995750256791e-05, "loss": 0.0189, "step": 57630 }, { "grad_norm": 0.16587671637535095, "learning_rate": 4.15436605361618e-05, "loss": 0.0165, "step": 57640 }, { "grad_norm": 0.18490739166736603, "learning_rate": 4.152736449452827e-05, "loss": 0.0141, "step": 57650 }, { "grad_norm": 0.1751832664012909, "learning_rate": 4.151106937944943e-05, "loss": 0.0156, "step": 57660 }, { "grad_norm": 0.16904926300048828, "learning_rate": 4.149477519270727e-05, "loss": 0.0189, "step": 57670 }, { "grad_norm": 0.1583123654127121, "learning_rate": 4.147848193608371e-05, "loss": 0.0143, "step": 57680 }, { "grad_norm": 0.1419401913881302, "learning_rate": 4.146218961136056e-05, "loss": 0.0151, "step": 57690 }, { "grad_norm": 0.17743565142154694, "learning_rate": 4.14458982203195e-05, "loss": 0.0162, "step": 57700 }, { "grad_norm": 0.15158791840076447, "learning_rate": 4.1429607764742165e-05, "loss": 0.0173, "step": 57710 }, { "grad_norm": 0.36171233654022217, "learning_rate": 4.141331824641003e-05, "loss": 0.0157, "step": 57720 }, { "grad_norm": 0.3277859389781952, "learning_rate": 4.139702966710452e-05, "loss": 0.0183, "step": 57730 }, { "grad_norm": 0.14942584931850433, "learning_rate": 4.13807420286069e-05, "loss": 0.0161, "step": 57740 }, { "grad_norm": 0.15536178648471832, "learning_rate": 4.136445533269837e-05, "loss": 0.015, "step": 57750 }, { "grad_norm": 0.23497143387794495, "learning_rate": 4.134816958116003e-05, "loss": 0.0153, "step": 57760 }, { "grad_norm": 0.19486989080905914, "learning_rate": 4.133188477577283e-05, "loss": 0.0149, "step": 57770 }, { "grad_norm": 0.16754679381847382, "learning_rate": 4.131560091831771e-05, "loss": 0.0172, "step": 57780 }, { "grad_norm": 0.12295573949813843, "learning_rate": 4.1299318010575405e-05, "loss": 0.0189, "step": 57790 }, { "grad_norm": 0.2193789780139923, "learning_rate": 4.128303605432661e-05, "loss": 0.0164, "step": 57800 }, { "grad_norm": 0.14469262957572937, "learning_rate": 4.126675505135188e-05, "loss": 0.0139, "step": 57810 }, { "grad_norm": 0.16404911875724792, "learning_rate": 4.125047500343169e-05, "loss": 0.0152, "step": 57820 }, { "grad_norm": 0.212103009223938, "learning_rate": 4.1234195912346376e-05, "loss": 0.014, "step": 57830 }, { "grad_norm": 0.17115791141986847, "learning_rate": 4.1217917779876235e-05, "loss": 0.0136, "step": 57840 }, { "grad_norm": 0.11873944848775864, "learning_rate": 4.120164060780141e-05, "loss": 0.0164, "step": 57850 }, { "grad_norm": 0.1596090942621231, "learning_rate": 4.1185364397901935e-05, "loss": 0.0167, "step": 57860 }, { "grad_norm": 0.16130903363227844, "learning_rate": 4.116908915195776e-05, "loss": 0.0146, "step": 57870 }, { "grad_norm": 0.18640190362930298, "learning_rate": 4.115281487174872e-05, "loss": 0.0122, "step": 57880 }, { "grad_norm": 0.17362070083618164, "learning_rate": 4.113654155905452e-05, "loss": 0.0148, "step": 57890 }, { "grad_norm": 0.17823439836502075, "learning_rate": 4.112026921565483e-05, "loss": 0.0156, "step": 57900 }, { "grad_norm": 0.19225098192691803, "learning_rate": 4.1103997843329147e-05, "loss": 0.0166, "step": 57910 }, { "grad_norm": 0.18079569935798645, "learning_rate": 4.108772744385689e-05, "loss": 0.0139, "step": 57920 }, { "grad_norm": 0.20143859088420868, "learning_rate": 4.1071458019017346e-05, "loss": 0.0175, "step": 57930 }, { "grad_norm": 0.1800692230463028, "learning_rate": 4.1055189570589734e-05, "loss": 0.0154, "step": 57940 }, { "grad_norm": 0.20647366344928741, "learning_rate": 4.103892210035315e-05, "loss": 0.0165, "step": 57950 }, { "grad_norm": 0.14757665991783142, "learning_rate": 4.102265561008656e-05, "loss": 0.016, "step": 57960 }, { "grad_norm": 0.1723366677761078, "learning_rate": 4.100639010156885e-05, "loss": 0.0152, "step": 57970 }, { "grad_norm": 0.19908849895000458, "learning_rate": 4.099012557657882e-05, "loss": 0.0166, "step": 57980 }, { "grad_norm": 0.18207600712776184, "learning_rate": 4.09738620368951e-05, "loss": 0.013, "step": 57990 }, { "grad_norm": 0.18233372271060944, "learning_rate": 4.095759948429626e-05, "loss": 0.0169, "step": 58000 }, { "grad_norm": 0.23316361010074615, "learning_rate": 4.094133792056075e-05, "loss": 0.0143, "step": 58010 }, { "grad_norm": 0.16751179099082947, "learning_rate": 4.092507734746688e-05, "loss": 0.0146, "step": 58020 }, { "grad_norm": 0.20086906850337982, "learning_rate": 4.090881776679293e-05, "loss": 0.0154, "step": 58030 }, { "grad_norm": 0.18745878338813782, "learning_rate": 4.0892559180316994e-05, "loss": 0.0135, "step": 58040 }, { "grad_norm": 0.2234174907207489, "learning_rate": 4.08763015898171e-05, "loss": 0.0138, "step": 58050 }, { "grad_norm": 0.17509539425373077, "learning_rate": 4.0860044997071145e-05, "loss": 0.0142, "step": 58060 }, { "grad_norm": 0.25313806533813477, "learning_rate": 4.084378940385693e-05, "loss": 0.018, "step": 58070 }, { "grad_norm": 0.25539180636405945, "learning_rate": 4.082753481195211e-05, "loss": 0.0177, "step": 58080 }, { "grad_norm": 0.20151075720787048, "learning_rate": 4.0811281223134324e-05, "loss": 0.0129, "step": 58090 }, { "grad_norm": 0.16771526634693146, "learning_rate": 4.0795028639181e-05, "loss": 0.0124, "step": 58100 }, { "grad_norm": 0.21444956958293915, "learning_rate": 4.077877706186951e-05, "loss": 0.0157, "step": 58110 }, { "grad_norm": 0.19251251220703125, "learning_rate": 4.076252649297709e-05, "loss": 0.0131, "step": 58120 }, { "grad_norm": 0.19403527677059174, "learning_rate": 4.07462769342809e-05, "loss": 0.0158, "step": 58130 }, { "grad_norm": 0.20456627011299133, "learning_rate": 4.073002838755793e-05, "loss": 0.0148, "step": 58140 }, { "grad_norm": 0.21088695526123047, "learning_rate": 4.0713780854585115e-05, "loss": 0.0156, "step": 58150 }, { "grad_norm": 0.15569284558296204, "learning_rate": 4.069753433713927e-05, "loss": 0.0161, "step": 58160 }, { "grad_norm": 0.23489290475845337, "learning_rate": 4.068128883699709e-05, "loss": 0.0156, "step": 58170 }, { "grad_norm": 0.16954386234283447, "learning_rate": 4.0665044355935155e-05, "loss": 0.0159, "step": 58180 }, { "grad_norm": 0.20574218034744263, "learning_rate": 4.064880089572993e-05, "loss": 0.014, "step": 58190 }, { "grad_norm": 0.23488108813762665, "learning_rate": 4.063255845815779e-05, "loss": 0.0163, "step": 58200 }, { "grad_norm": 0.18104133009910583, "learning_rate": 4.0616317044994945e-05, "loss": 0.0162, "step": 58210 }, { "grad_norm": 0.15991169214248657, "learning_rate": 4.0600076658017585e-05, "loss": 0.0123, "step": 58220 }, { "grad_norm": 0.33520328998565674, "learning_rate": 4.0583837299001706e-05, "loss": 0.0179, "step": 58230 }, { "grad_norm": 0.18398946523666382, "learning_rate": 4.056759896972322e-05, "loss": 0.0157, "step": 58240 }, { "grad_norm": 0.1819511353969574, "learning_rate": 4.055136167195793e-05, "loss": 0.0188, "step": 58250 }, { "grad_norm": 0.13757286965847015, "learning_rate": 4.053512540748152e-05, "loss": 0.0121, "step": 58260 }, { "grad_norm": 0.2576718032360077, "learning_rate": 4.051889017806955e-05, "loss": 0.0183, "step": 58270 }, { "grad_norm": 0.14625734090805054, "learning_rate": 4.0502655985497516e-05, "loss": 0.0138, "step": 58280 }, { "grad_norm": 0.3110758364200592, "learning_rate": 4.048642283154074e-05, "loss": 0.0183, "step": 58290 }, { "grad_norm": 0.1760512888431549, "learning_rate": 4.0470190717974464e-05, "loss": 0.0143, "step": 58300 }, { "grad_norm": 0.17738588154315948, "learning_rate": 4.045395964657381e-05, "loss": 0.0191, "step": 58310 }, { "grad_norm": 0.20435263216495514, "learning_rate": 4.043772961911376e-05, "loss": 0.015, "step": 58320 }, { "grad_norm": 0.1831130087375641, "learning_rate": 4.042150063736924e-05, "loss": 0.0159, "step": 58330 }, { "grad_norm": 0.1694738119840622, "learning_rate": 4.0405272703114976e-05, "loss": 0.0165, "step": 58340 }, { "grad_norm": 0.1712665557861328, "learning_rate": 4.038904581812569e-05, "loss": 0.0156, "step": 58350 }, { "grad_norm": 0.13925224542617798, "learning_rate": 4.03728199841759e-05, "loss": 0.0152, "step": 58360 }, { "grad_norm": 0.17606480419635773, "learning_rate": 4.0356595203040044e-05, "loss": 0.0136, "step": 58370 }, { "grad_norm": 0.18053756654262543, "learning_rate": 4.0340371476492426e-05, "loss": 0.0174, "step": 58380 }, { "grad_norm": 0.18693457543849945, "learning_rate": 4.032414880630727e-05, "loss": 0.014, "step": 58390 }, { "grad_norm": 0.1517089605331421, "learning_rate": 4.0307927194258646e-05, "loss": 0.0137, "step": 58400 }, { "grad_norm": 0.14565908908843994, "learning_rate": 4.029170664212054e-05, "loss": 0.0155, "step": 58410 }, { "grad_norm": 0.16632068157196045, "learning_rate": 4.0275487151666794e-05, "loss": 0.0136, "step": 58420 }, { "grad_norm": 0.14074039459228516, "learning_rate": 4.025926872467116e-05, "loss": 0.0182, "step": 58430 }, { "grad_norm": 0.15267735719680786, "learning_rate": 4.024305136290725e-05, "loss": 0.012, "step": 58440 }, { "grad_norm": 0.2175782471895218, "learning_rate": 4.022683506814858e-05, "loss": 0.0151, "step": 58450 }, { "grad_norm": 0.19192026555538177, "learning_rate": 4.021061984216853e-05, "loss": 0.0159, "step": 58460 }, { "grad_norm": 0.19262459874153137, "learning_rate": 4.019440568674035e-05, "loss": 0.0157, "step": 58470 }, { "grad_norm": 0.23719266057014465, "learning_rate": 4.017819260363724e-05, "loss": 0.0188, "step": 58480 }, { "grad_norm": 0.19733558595180511, "learning_rate": 4.016198059463223e-05, "loss": 0.0169, "step": 58490 }, { "grad_norm": 0.1353185474872589, "learning_rate": 4.0145769661498233e-05, "loss": 0.0141, "step": 58500 }, { "grad_norm": 0.4549795687198639, "learning_rate": 4.012955980600804e-05, "loss": 0.0164, "step": 58510 }, { "grad_norm": 0.18384437263011932, "learning_rate": 4.0113351029934355e-05, "loss": 0.0145, "step": 58520 }, { "grad_norm": 0.1642380654811859, "learning_rate": 4.0097143335049714e-05, "loss": 0.017, "step": 58530 }, { "grad_norm": 0.5434411764144897, "learning_rate": 4.0080936723126606e-05, "loss": 0.0142, "step": 58540 }, { "grad_norm": 0.23707076907157898, "learning_rate": 4.006473119593735e-05, "loss": 0.0156, "step": 58550 }, { "grad_norm": 0.14486251771450043, "learning_rate": 4.004852675525415e-05, "loss": 0.0175, "step": 58560 }, { "grad_norm": 0.16986529529094696, "learning_rate": 4.00323234028491e-05, "loss": 0.016, "step": 58570 }, { "grad_norm": 0.1588120013475418, "learning_rate": 4.0016121140494186e-05, "loss": 0.0182, "step": 58580 }, { "grad_norm": 0.2247074842453003, "learning_rate": 3.999991996996122e-05, "loss": 0.0166, "step": 58590 }, { "grad_norm": 0.18489289283752441, "learning_rate": 3.9983719893022e-05, "loss": 0.0176, "step": 58600 }, { "grad_norm": 0.16268986463546753, "learning_rate": 3.996752091144812e-05, "loss": 0.0146, "step": 58610 }, { "grad_norm": 0.16814306378364563, "learning_rate": 3.9951323027011055e-05, "loss": 0.0164, "step": 58620 }, { "grad_norm": 0.15995892882347107, "learning_rate": 3.993512624148219e-05, "loss": 0.0166, "step": 58630 }, { "grad_norm": 0.21253986656665802, "learning_rate": 3.99189305566328e-05, "loss": 0.0152, "step": 58640 }, { "grad_norm": 0.2758246958255768, "learning_rate": 3.990273597423401e-05, "loss": 0.0199, "step": 58650 }, { "grad_norm": 0.10529027879238129, "learning_rate": 3.988654249605681e-05, "loss": 0.0133, "step": 58660 }, { "grad_norm": 0.13128577172756195, "learning_rate": 3.987035012387213e-05, "loss": 0.0154, "step": 58670 }, { "grad_norm": 0.15469254553318024, "learning_rate": 3.9854158859450735e-05, "loss": 0.014, "step": 58680 }, { "grad_norm": 0.23149321973323822, "learning_rate": 3.983796870456328e-05, "loss": 0.0191, "step": 58690 }, { "grad_norm": 0.16576868295669556, "learning_rate": 3.982177966098028e-05, "loss": 0.0144, "step": 58700 }, { "grad_norm": 0.1814160943031311, "learning_rate": 3.980559173047216e-05, "loss": 0.0161, "step": 58710 }, { "grad_norm": 0.226541206240654, "learning_rate": 3.9789404914809176e-05, "loss": 0.022, "step": 58720 }, { "grad_norm": 0.16154053807258606, "learning_rate": 3.9773219215761544e-05, "loss": 0.0164, "step": 58730 }, { "grad_norm": 0.1676248461008072, "learning_rate": 3.975703463509928e-05, "loss": 0.0166, "step": 58740 }, { "grad_norm": 0.20218217372894287, "learning_rate": 3.974085117459231e-05, "loss": 0.0149, "step": 58750 }, { "grad_norm": 0.18293742835521698, "learning_rate": 3.972466883601043e-05, "loss": 0.0159, "step": 58760 }, { "grad_norm": 0.2657666504383087, "learning_rate": 3.9708487621123316e-05, "loss": 0.0145, "step": 58770 }, { "grad_norm": 0.21257631480693817, "learning_rate": 3.969230753170051e-05, "loss": 0.0171, "step": 58780 }, { "grad_norm": 0.2023691087961197, "learning_rate": 3.967612856951146e-05, "loss": 0.0156, "step": 58790 }, { "grad_norm": 0.19236747920513153, "learning_rate": 3.9659950736325476e-05, "loss": 0.0158, "step": 58800 }, { "grad_norm": 0.32242363691329956, "learning_rate": 3.964377403391174e-05, "loss": 0.0132, "step": 58810 }, { "grad_norm": 0.1722579151391983, "learning_rate": 3.9627598464039303e-05, "loss": 0.0136, "step": 58820 }, { "grad_norm": 0.15396513044834137, "learning_rate": 3.9611424028477096e-05, "loss": 0.0136, "step": 58830 }, { "grad_norm": 0.19834519922733307, "learning_rate": 3.959525072899394e-05, "loss": 0.0155, "step": 58840 }, { "grad_norm": 0.18206581473350525, "learning_rate": 3.9579078567358506e-05, "loss": 0.015, "step": 58850 }, { "grad_norm": 0.14501020312309265, "learning_rate": 3.956290754533939e-05, "loss": 0.0154, "step": 58860 }, { "grad_norm": 0.2041306346654892, "learning_rate": 3.954673766470501e-05, "loss": 0.015, "step": 58870 }, { "grad_norm": 0.17621241509914398, "learning_rate": 3.9530568927223674e-05, "loss": 0.0127, "step": 58880 }, { "grad_norm": 0.21793586015701294, "learning_rate": 3.95144013346636e-05, "loss": 0.0176, "step": 58890 }, { "grad_norm": 0.13959263265132904, "learning_rate": 3.949823488879282e-05, "loss": 0.0138, "step": 58900 }, { "grad_norm": 0.184395894408226, "learning_rate": 3.9482069591379276e-05, "loss": 0.0165, "step": 58910 }, { "grad_norm": 0.15062861144542694, "learning_rate": 3.9465905444190784e-05, "loss": 0.0147, "step": 58920 }, { "grad_norm": 0.1406903862953186, "learning_rate": 3.944974244899505e-05, "loss": 0.013, "step": 58930 }, { "grad_norm": 0.1668069213628769, "learning_rate": 3.9433580607559616e-05, "loss": 0.0158, "step": 58940 }, { "grad_norm": 0.17247815430164337, "learning_rate": 3.941741992165193e-05, "loss": 0.0173, "step": 58950 }, { "grad_norm": 0.1901829093694687, "learning_rate": 3.940126039303928e-05, "loss": 0.0144, "step": 58960 }, { "grad_norm": 0.22636552155017853, "learning_rate": 3.938510202348886e-05, "loss": 0.0133, "step": 58970 }, { "grad_norm": 0.21134008467197418, "learning_rate": 3.93689448147677e-05, "loss": 0.014, "step": 58980 }, { "grad_norm": 0.20370234549045563, "learning_rate": 3.935278876864278e-05, "loss": 0.0166, "step": 58990 }, { "grad_norm": 0.27128058671951294, "learning_rate": 3.933663388688087e-05, "loss": 0.0178, "step": 59000 }, { "grad_norm": 0.23028692603111267, "learning_rate": 3.9320480171248645e-05, "loss": 0.019, "step": 59010 }, { "grad_norm": 0.17059826850891113, "learning_rate": 3.930432762351265e-05, "loss": 0.0142, "step": 59020 }, { "grad_norm": 0.2041853815317154, "learning_rate": 3.928817624543931e-05, "loss": 0.0183, "step": 59030 }, { "grad_norm": 0.17504863440990448, "learning_rate": 3.9272026038794885e-05, "loss": 0.0145, "step": 59040 }, { "grad_norm": 0.14416605234146118, "learning_rate": 3.9255877005345586e-05, "loss": 0.0156, "step": 59050 }, { "grad_norm": 0.13233786821365356, "learning_rate": 3.923972914685743e-05, "loss": 0.013, "step": 59060 }, { "grad_norm": 0.16871200501918793, "learning_rate": 3.922358246509631e-05, "loss": 0.0161, "step": 59070 }, { "grad_norm": 0.21459639072418213, "learning_rate": 3.920743696182801e-05, "loss": 0.018, "step": 59080 }, { "grad_norm": 0.17630864679813385, "learning_rate": 3.9191292638818164e-05, "loss": 0.0138, "step": 59090 }, { "grad_norm": 0.17519651353359222, "learning_rate": 3.9175149497832316e-05, "loss": 0.014, "step": 59100 }, { "grad_norm": 0.13216015696525574, "learning_rate": 3.915900754063584e-05, "loss": 0.0148, "step": 59110 }, { "grad_norm": 0.16014818847179413, "learning_rate": 3.9142866768994014e-05, "loss": 0.0172, "step": 59120 }, { "grad_norm": 0.19655270874500275, "learning_rate": 3.912672718467193e-05, "loss": 0.0147, "step": 59130 }, { "grad_norm": 0.1594698429107666, "learning_rate": 3.911058878943463e-05, "loss": 0.0162, "step": 59140 }, { "grad_norm": 0.1585044115781784, "learning_rate": 3.9094451585046974e-05, "loss": 0.0147, "step": 59150 }, { "grad_norm": 0.16224054992198944, "learning_rate": 3.90783155732737e-05, "loss": 0.015, "step": 59160 }, { "grad_norm": 0.20567388832569122, "learning_rate": 3.9062180755879373e-05, "loss": 0.0158, "step": 59170 }, { "grad_norm": 0.17384083569049835, "learning_rate": 3.904604713462855e-05, "loss": 0.0146, "step": 59180 }, { "grad_norm": 0.16178062558174133, "learning_rate": 3.902991471128554e-05, "loss": 0.0153, "step": 59190 }, { "grad_norm": 0.2405622899532318, "learning_rate": 3.901378348761456e-05, "loss": 0.0157, "step": 59200 }, { "grad_norm": 0.2302583009004593, "learning_rate": 3.899765346537971e-05, "loss": 0.0152, "step": 59210 }, { "grad_norm": 0.20525430142879486, "learning_rate": 3.898152464634493e-05, "loss": 0.0193, "step": 59220 }, { "grad_norm": 0.13870786130428314, "learning_rate": 3.896539703227402e-05, "loss": 0.0169, "step": 59230 }, { "grad_norm": 0.16499270498752594, "learning_rate": 3.894927062493073e-05, "loss": 0.0139, "step": 59240 }, { "grad_norm": 0.17256005108356476, "learning_rate": 3.893314542607859e-05, "loss": 0.012, "step": 59250 }, { "grad_norm": 0.12540781497955322, "learning_rate": 3.891702143748103e-05, "loss": 0.0155, "step": 59260 }, { "grad_norm": 0.19101585447788239, "learning_rate": 3.890089866090133e-05, "loss": 0.0137, "step": 59270 }, { "grad_norm": 0.24667738378047943, "learning_rate": 3.888477709810267e-05, "loss": 0.0143, "step": 59280 }, { "grad_norm": 0.11837579309940338, "learning_rate": 3.886865675084806e-05, "loss": 0.014, "step": 59290 }, { "grad_norm": 0.20733387768268585, "learning_rate": 3.885253762090043e-05, "loss": 0.0138, "step": 59300 }, { "grad_norm": 0.21029053628444672, "learning_rate": 3.8836419710022534e-05, "loss": 0.0133, "step": 59310 }, { "grad_norm": 0.1671397089958191, "learning_rate": 3.882030301997698e-05, "loss": 0.0166, "step": 59320 }, { "grad_norm": 0.22961384057998657, "learning_rate": 3.880418755252629e-05, "loss": 0.0169, "step": 59330 }, { "grad_norm": 0.21383585035800934, "learning_rate": 3.87880733094328e-05, "loss": 0.015, "step": 59340 }, { "grad_norm": 0.17109446227550507, "learning_rate": 3.877196029245877e-05, "loss": 0.0134, "step": 59350 }, { "grad_norm": 0.17738263309001923, "learning_rate": 3.875584850336627e-05, "loss": 0.0114, "step": 59360 }, { "grad_norm": 0.18017913401126862, "learning_rate": 3.873973794391728e-05, "loss": 0.0151, "step": 59370 }, { "grad_norm": 0.17232881486415863, "learning_rate": 3.872362861587361e-05, "loss": 0.0153, "step": 59380 }, { "grad_norm": 0.16322512924671173, "learning_rate": 3.870752052099698e-05, "loss": 0.0154, "step": 59390 }, { "grad_norm": 0.2045181840658188, "learning_rate": 3.8691413661048916e-05, "loss": 0.0155, "step": 59400 }, { "grad_norm": 0.19733873009681702, "learning_rate": 3.867530803779085e-05, "loss": 0.018, "step": 59410 }, { "grad_norm": 0.17292460799217224, "learning_rate": 3.8659203652984054e-05, "loss": 0.0132, "step": 59420 }, { "grad_norm": 0.14729559421539307, "learning_rate": 3.864310050838972e-05, "loss": 0.014, "step": 59430 }, { "grad_norm": 0.15433576703071594, "learning_rate": 3.8626998605768835e-05, "loss": 0.015, "step": 59440 }, { "grad_norm": 0.22074483335018158, "learning_rate": 3.8610897946882294e-05, "loss": 0.0225, "step": 59450 }, { "grad_norm": 0.2166355550289154, "learning_rate": 3.859479853349082e-05, "loss": 0.0118, "step": 59460 }, { "grad_norm": 0.1643405705690384, "learning_rate": 3.857870036735503e-05, "loss": 0.0119, "step": 59470 }, { "grad_norm": 0.23618505895137787, "learning_rate": 3.856260345023538e-05, "loss": 0.0159, "step": 59480 }, { "grad_norm": 0.13329318165779114, "learning_rate": 3.854650778389224e-05, "loss": 0.0167, "step": 59490 }, { "grad_norm": 0.1318281888961792, "learning_rate": 3.8530413370085786e-05, "loss": 0.0151, "step": 59500 }, { "grad_norm": 0.1580425202846527, "learning_rate": 3.851432021057608e-05, "loss": 0.0118, "step": 59510 }, { "grad_norm": 0.10061865299940109, "learning_rate": 3.849822830712306e-05, "loss": 0.0176, "step": 59520 }, { "grad_norm": 0.2457302212715149, "learning_rate": 3.848213766148649e-05, "loss": 0.0183, "step": 59530 }, { "grad_norm": 0.15386995673179626, "learning_rate": 3.846604827542602e-05, "loss": 0.0179, "step": 59540 }, { "grad_norm": 0.13569758832454681, "learning_rate": 3.844996015070117e-05, "loss": 0.0156, "step": 59550 }, { "grad_norm": 0.16616258025169373, "learning_rate": 3.843387328907132e-05, "loss": 0.0141, "step": 59560 }, { "grad_norm": 0.1665937751531601, "learning_rate": 3.8417787692295696e-05, "loss": 0.015, "step": 59570 }, { "grad_norm": 0.18340151011943817, "learning_rate": 3.84017033621334e-05, "loss": 0.0147, "step": 59580 }, { "grad_norm": 0.1436077207326889, "learning_rate": 3.838562030034338e-05, "loss": 0.0147, "step": 59590 }, { "grad_norm": 0.1772737056016922, "learning_rate": 3.836953850868447e-05, "loss": 0.0172, "step": 59600 }, { "grad_norm": 0.21930555999279022, "learning_rate": 3.835345798891533e-05, "loss": 0.0133, "step": 59610 }, { "grad_norm": 0.20784424245357513, "learning_rate": 3.833737874279452e-05, "loss": 0.0151, "step": 59620 }, { "grad_norm": 0.2716967463493347, "learning_rate": 3.8321300772080433e-05, "loss": 0.0174, "step": 59630 }, { "grad_norm": 0.16903257369995117, "learning_rate": 3.8305224078531345e-05, "loss": 0.0139, "step": 59640 }, { "grad_norm": 0.22415047883987427, "learning_rate": 3.828914866390537e-05, "loss": 0.0185, "step": 59650 }, { "grad_norm": 0.22338177263736725, "learning_rate": 3.827307452996048e-05, "loss": 0.0152, "step": 59660 }, { "grad_norm": 0.1710379719734192, "learning_rate": 3.825700167845454e-05, "loss": 0.0149, "step": 59670 }, { "grad_norm": 0.16482047736644745, "learning_rate": 3.8240930111145214e-05, "loss": 0.0173, "step": 59680 }, { "grad_norm": 0.15730434656143188, "learning_rate": 3.8224859829790116e-05, "loss": 0.0125, "step": 59690 }, { "grad_norm": 0.21644394099712372, "learning_rate": 3.820879083614664e-05, "loss": 0.021, "step": 59700 }, { "grad_norm": 0.16690506041049957, "learning_rate": 3.819272313197208e-05, "loss": 0.0172, "step": 59710 }, { "grad_norm": 0.17965741455554962, "learning_rate": 3.817665671902355e-05, "loss": 0.0163, "step": 59720 }, { "grad_norm": 0.24177883565425873, "learning_rate": 3.816059159905808e-05, "loss": 0.0147, "step": 59730 }, { "grad_norm": 0.17487680912017822, "learning_rate": 3.8144527773832476e-05, "loss": 0.0145, "step": 59740 }, { "grad_norm": 0.23194290697574615, "learning_rate": 3.812846524510352e-05, "loss": 0.0175, "step": 59750 }, { "grad_norm": 0.19551946222782135, "learning_rate": 3.811240401462775e-05, "loss": 0.0175, "step": 59760 }, { "grad_norm": 0.17860497534275055, "learning_rate": 3.809634408416162e-05, "loss": 0.019, "step": 59770 }, { "grad_norm": 0.1756526529788971, "learning_rate": 3.8080285455461394e-05, "loss": 0.0161, "step": 59780 }, { "grad_norm": 0.1651402860879898, "learning_rate": 3.806422813028323e-05, "loss": 0.016, "step": 59790 }, { "grad_norm": 0.16948555409908295, "learning_rate": 3.8048172110383114e-05, "loss": 0.0161, "step": 59800 }, { "grad_norm": 0.22257308661937714, "learning_rate": 3.8032117397516944e-05, "loss": 0.0149, "step": 59810 }, { "grad_norm": 0.2145165055990219, "learning_rate": 3.801606399344042e-05, "loss": 0.0202, "step": 59820 }, { "grad_norm": 0.19167065620422363, "learning_rate": 3.800001189990913e-05, "loss": 0.0177, "step": 59830 }, { "grad_norm": 0.23640406131744385, "learning_rate": 3.798396111867847e-05, "loss": 0.0149, "step": 59840 }, { "grad_norm": 0.1783660352230072, "learning_rate": 3.796791165150378e-05, "loss": 0.0191, "step": 59850 }, { "grad_norm": 0.16497306525707245, "learning_rate": 3.795186350014018e-05, "loss": 0.0155, "step": 59860 }, { "grad_norm": 0.1500440090894699, "learning_rate": 3.793581666634266e-05, "loss": 0.0172, "step": 59870 }, { "grad_norm": 0.1374140828847885, "learning_rate": 3.791977115186609e-05, "loss": 0.0187, "step": 59880 }, { "grad_norm": 0.21006529033184052, "learning_rate": 3.79037269584652e-05, "loss": 0.0168, "step": 59890 }, { "grad_norm": 0.16075019538402557, "learning_rate": 3.7887684087894545e-05, "loss": 0.0161, "step": 59900 }, { "grad_norm": 0.1845230907201767, "learning_rate": 3.787164254190854e-05, "loss": 0.0124, "step": 59910 }, { "grad_norm": 0.1443103700876236, "learning_rate": 3.7855602322261473e-05, "loss": 0.0155, "step": 59920 }, { "grad_norm": 0.18104761838912964, "learning_rate": 3.783956343070746e-05, "loss": 0.0161, "step": 59930 }, { "grad_norm": 0.17245660722255707, "learning_rate": 3.7823525869000526e-05, "loss": 0.0139, "step": 59940 }, { "grad_norm": 0.19773131608963013, "learning_rate": 3.78074896388945e-05, "loss": 0.0144, "step": 59950 }, { "grad_norm": 0.23266567289829254, "learning_rate": 3.7791454742143075e-05, "loss": 0.0136, "step": 59960 }, { "grad_norm": 0.19393497705459595, "learning_rate": 3.7775421180499795e-05, "loss": 0.0112, "step": 59970 }, { "grad_norm": 0.17116424441337585, "learning_rate": 3.7759388955718086e-05, "loss": 0.0163, "step": 59980 }, { "grad_norm": 0.1942400485277176, "learning_rate": 3.774335806955117e-05, "loss": 0.015, "step": 59990 }, { "grad_norm": 0.18268521130084991, "learning_rate": 3.7727328523752215e-05, "loss": 0.0143, "step": 60000 }, { "grad_norm": 0.17526081204414368, "learning_rate": 3.771130032007415e-05, "loss": 0.0138, "step": 60010 }, { "grad_norm": 0.1784512847661972, "learning_rate": 3.769527346026982e-05, "loss": 0.0146, "step": 60020 }, { "grad_norm": 0.1542792022228241, "learning_rate": 3.767924794609188e-05, "loss": 0.0184, "step": 60030 }, { "grad_norm": 0.17526817321777344, "learning_rate": 3.7663223779292854e-05, "loss": 0.0164, "step": 60040 }, { "grad_norm": 0.1760818064212799, "learning_rate": 3.764720096162512e-05, "loss": 0.0142, "step": 60050 }, { "grad_norm": 0.17752714455127716, "learning_rate": 3.763117949484092e-05, "loss": 0.0149, "step": 60060 }, { "grad_norm": 0.30211901664733887, "learning_rate": 3.7615159380692344e-05, "loss": 0.0123, "step": 60070 }, { "grad_norm": 0.21667972207069397, "learning_rate": 3.759914062093132e-05, "loss": 0.0137, "step": 60080 }, { "grad_norm": 0.23229072988033295, "learning_rate": 3.7583123217309615e-05, "loss": 0.0146, "step": 60090 }, { "grad_norm": 0.15466992557048798, "learning_rate": 3.7567107171578904e-05, "loss": 0.0134, "step": 60100 }, { "grad_norm": 0.1680803745985031, "learning_rate": 3.755109248549066e-05, "loss": 0.0151, "step": 60110 }, { "grad_norm": 0.1771594136953354, "learning_rate": 3.75350791607962e-05, "loss": 0.0194, "step": 60120 }, { "grad_norm": 0.197223961353302, "learning_rate": 3.751906719924676e-05, "loss": 0.0149, "step": 60130 }, { "grad_norm": 0.14052271842956543, "learning_rate": 3.750305660259337e-05, "loss": 0.0133, "step": 60140 }, { "grad_norm": 0.24261125922203064, "learning_rate": 3.748704737258693e-05, "loss": 0.0142, "step": 60150 }, { "grad_norm": 0.14634084701538086, "learning_rate": 3.747103951097816e-05, "loss": 0.0154, "step": 60160 }, { "grad_norm": 0.19099286198616028, "learning_rate": 3.745503301951768e-05, "loss": 0.0166, "step": 60170 }, { "grad_norm": 0.16745935380458832, "learning_rate": 3.743902789995592e-05, "loss": 0.0134, "step": 60180 }, { "grad_norm": 0.1901850402355194, "learning_rate": 3.7423024154043183e-05, "loss": 0.0143, "step": 60190 }, { "grad_norm": 0.17381936311721802, "learning_rate": 3.740702178352963e-05, "loss": 0.0147, "step": 60200 }, { "grad_norm": 0.1728156954050064, "learning_rate": 3.739102079016523e-05, "loss": 0.0149, "step": 60210 }, { "grad_norm": 0.15539337694644928, "learning_rate": 3.7375021175699846e-05, "loss": 0.0136, "step": 60220 }, { "grad_norm": 0.19856461882591248, "learning_rate": 3.735902294188317e-05, "loss": 0.0162, "step": 60230 }, { "grad_norm": 0.1641201227903366, "learning_rate": 3.734302609046473e-05, "loss": 0.0146, "step": 60240 }, { "grad_norm": 0.17716707289218903, "learning_rate": 3.732703062319391e-05, "loss": 0.0139, "step": 60250 }, { "grad_norm": 0.1562071293592453, "learning_rate": 3.731103654181999e-05, "loss": 0.0152, "step": 60260 }, { "grad_norm": 0.2099848985671997, "learning_rate": 3.729504384809203e-05, "loss": 0.0141, "step": 60270 }, { "grad_norm": 0.16600772738456726, "learning_rate": 3.727905254375898e-05, "loss": 0.0187, "step": 60280 }, { "grad_norm": 0.2129598706960678, "learning_rate": 3.7263062630569614e-05, "loss": 0.017, "step": 60290 }, { "grad_norm": 0.19670486450195312, "learning_rate": 3.724707411027256e-05, "loss": 0.0163, "step": 60300 }, { "grad_norm": 0.17132562398910522, "learning_rate": 3.723108698461631e-05, "loss": 0.0155, "step": 60310 }, { "grad_norm": 0.1601582169532776, "learning_rate": 3.72151012553492e-05, "loss": 0.0173, "step": 60320 }, { "grad_norm": 0.1937924027442932, "learning_rate": 3.719911692421939e-05, "loss": 0.0141, "step": 60330 }, { "grad_norm": 0.15164817869663239, "learning_rate": 3.71831339929749e-05, "loss": 0.0195, "step": 60340 }, { "grad_norm": 0.18016497790813446, "learning_rate": 3.716715246336362e-05, "loss": 0.0125, "step": 60350 }, { "grad_norm": 0.15075789391994476, "learning_rate": 3.715117233713324e-05, "loss": 0.0152, "step": 60360 }, { "grad_norm": 0.267707884311676, "learning_rate": 3.713519361603135e-05, "loss": 0.0136, "step": 60370 }, { "grad_norm": 0.20494292676448822, "learning_rate": 3.711921630180532e-05, "loss": 0.0164, "step": 60380 }, { "grad_norm": 0.119402214884758, "learning_rate": 3.710324039620245e-05, "loss": 0.0164, "step": 60390 }, { "grad_norm": 0.15241654217243195, "learning_rate": 3.708726590096982e-05, "loss": 0.0151, "step": 60400 }, { "grad_norm": 0.27880799770355225, "learning_rate": 3.707129281785437e-05, "loss": 0.0172, "step": 60410 }, { "grad_norm": 0.3292875289916992, "learning_rate": 3.705532114860291e-05, "loss": 0.0155, "step": 60420 }, { "grad_norm": 0.22031709551811218, "learning_rate": 3.703935089496207e-05, "loss": 0.0159, "step": 60430 }, { "grad_norm": 0.1361749768257141, "learning_rate": 3.702338205867829e-05, "loss": 0.0218, "step": 60440 }, { "grad_norm": 0.188876211643219, "learning_rate": 3.7007414641497976e-05, "loss": 0.0145, "step": 60450 }, { "grad_norm": 0.1449728161096573, "learning_rate": 3.699144864516726e-05, "loss": 0.012, "step": 60460 }, { "grad_norm": 0.14730586111545563, "learning_rate": 3.6975484071432153e-05, "loss": 0.0143, "step": 60470 }, { "grad_norm": 0.17160935699939728, "learning_rate": 3.695952092203854e-05, "loss": 0.017, "step": 60480 }, { "grad_norm": 0.1752105951309204, "learning_rate": 3.69435591987321e-05, "loss": 0.015, "step": 60490 }, { "grad_norm": 0.16985009610652924, "learning_rate": 3.6927598903258374e-05, "loss": 0.0237, "step": 60500 }, { "grad_norm": 0.15687130391597748, "learning_rate": 3.691164003736281e-05, "loss": 0.0157, "step": 60510 }, { "grad_norm": 0.15263760089874268, "learning_rate": 3.689568260279061e-05, "loss": 0.0165, "step": 60520 }, { "grad_norm": 0.18299680948257446, "learning_rate": 3.687972660128686e-05, "loss": 0.0168, "step": 60530 }, { "grad_norm": 0.19803626835346222, "learning_rate": 3.686377203459648e-05, "loss": 0.0163, "step": 60540 }, { "grad_norm": 0.20421186089515686, "learning_rate": 3.6847818904464226e-05, "loss": 0.0191, "step": 60550 }, { "grad_norm": 0.21550294756889343, "learning_rate": 3.683186721263474e-05, "loss": 0.016, "step": 60560 }, { "grad_norm": 0.17219534516334534, "learning_rate": 3.6815916960852435e-05, "loss": 0.0166, "step": 60570 }, { "grad_norm": 0.1897435486316681, "learning_rate": 3.679996815086165e-05, "loss": 0.0137, "step": 60580 }, { "grad_norm": 0.16638046503067017, "learning_rate": 3.678402078440649e-05, "loss": 0.0141, "step": 60590 }, { "grad_norm": 0.2672623097896576, "learning_rate": 3.676807486323096e-05, "loss": 0.0167, "step": 60600 }, { "grad_norm": 0.18278126418590546, "learning_rate": 3.6752130389078864e-05, "loss": 0.0138, "step": 60610 }, { "grad_norm": 0.18698355555534363, "learning_rate": 3.673618736369388e-05, "loss": 0.0141, "step": 60620 }, { "grad_norm": 0.19009016454219818, "learning_rate": 3.6720245788819474e-05, "loss": 0.0162, "step": 60630 }, { "grad_norm": 0.20132353901863098, "learning_rate": 3.6704305666199044e-05, "loss": 0.0168, "step": 60640 }, { "grad_norm": 0.167368546128273, "learning_rate": 3.668836699757576e-05, "loss": 0.0147, "step": 60650 }, { "grad_norm": 0.14532333612442017, "learning_rate": 3.667242978469265e-05, "loss": 0.0166, "step": 60660 }, { "grad_norm": 0.1306346356868744, "learning_rate": 3.6656494029292575e-05, "loss": 0.016, "step": 60670 }, { "grad_norm": 0.15690846741199493, "learning_rate": 3.664055973311825e-05, "loss": 0.0165, "step": 60680 }, { "grad_norm": 0.2026004195213318, "learning_rate": 3.662462689791221e-05, "loss": 0.0143, "step": 60690 }, { "grad_norm": 0.230680912733078, "learning_rate": 3.660869552541689e-05, "loss": 0.0144, "step": 60700 }, { "grad_norm": 0.17795546352863312, "learning_rate": 3.659276561737448e-05, "loss": 0.015, "step": 60710 }, { "grad_norm": 0.1413554549217224, "learning_rate": 3.657683717552708e-05, "loss": 0.0173, "step": 60720 }, { "grad_norm": 0.16905532777309418, "learning_rate": 3.6560910201616574e-05, "loss": 0.0172, "step": 60730 }, { "grad_norm": 0.17520073056221008, "learning_rate": 3.654498469738472e-05, "loss": 0.0142, "step": 60740 }, { "grad_norm": 0.15031492710113525, "learning_rate": 3.6529060664573126e-05, "loss": 0.015, "step": 60750 }, { "grad_norm": 0.145172119140625, "learning_rate": 3.6513138104923176e-05, "loss": 0.0162, "step": 60760 }, { "grad_norm": 0.23526589572429657, "learning_rate": 3.6497217020176176e-05, "loss": 0.0159, "step": 60770 }, { "grad_norm": 0.13848331570625305, "learning_rate": 3.648129741207323e-05, "loss": 0.0146, "step": 60780 }, { "grad_norm": 0.18190442025661469, "learning_rate": 3.646537928235527e-05, "loss": 0.0136, "step": 60790 }, { "grad_norm": 0.19403046369552612, "learning_rate": 3.6449462632763075e-05, "loss": 0.0154, "step": 60800 }, { "grad_norm": 0.15026727318763733, "learning_rate": 3.643354746503729e-05, "loss": 0.0157, "step": 60810 }, { "grad_norm": 0.1399969756603241, "learning_rate": 3.6417633780918345e-05, "loss": 0.014, "step": 60820 }, { "grad_norm": 0.20655840635299683, "learning_rate": 3.640172158214655e-05, "loss": 0.0164, "step": 60830 }, { "grad_norm": 0.15690301358699799, "learning_rate": 3.6385810870462035e-05, "loss": 0.0138, "step": 60840 }, { "grad_norm": 0.1973041445016861, "learning_rate": 3.636990164760479e-05, "loss": 0.0118, "step": 60850 }, { "grad_norm": 0.17259061336517334, "learning_rate": 3.6353993915314596e-05, "loss": 0.0121, "step": 60860 }, { "grad_norm": 0.2030731439590454, "learning_rate": 3.6338087675331124e-05, "loss": 0.0139, "step": 60870 }, { "grad_norm": 0.12863776087760925, "learning_rate": 3.632218292939383e-05, "loss": 0.0137, "step": 60880 }, { "grad_norm": 0.17418552935123444, "learning_rate": 3.6306279679242033e-05, "loss": 0.0135, "step": 60890 }, { "grad_norm": 0.22634541988372803, "learning_rate": 3.629037792661492e-05, "loss": 0.0165, "step": 60900 }, { "grad_norm": 0.15190739929676056, "learning_rate": 3.627447767325147e-05, "loss": 0.0123, "step": 60910 }, { "grad_norm": 0.17121492326259613, "learning_rate": 3.62585789208905e-05, "loss": 0.0133, "step": 60920 }, { "grad_norm": 0.16097195446491241, "learning_rate": 3.624268167127068e-05, "loss": 0.014, "step": 60930 }, { "grad_norm": 0.18005701899528503, "learning_rate": 3.622678592613051e-05, "loss": 0.0155, "step": 60940 }, { "grad_norm": 0.20038633048534393, "learning_rate": 3.6210891687208304e-05, "loss": 0.0131, "step": 60950 }, { "grad_norm": 0.1403942108154297, "learning_rate": 3.619499895624228e-05, "loss": 0.0131, "step": 60960 }, { "grad_norm": 0.20998317003250122, "learning_rate": 3.61791077349704e-05, "loss": 0.0163, "step": 60970 }, { "grad_norm": 0.16015322506427765, "learning_rate": 3.616321802513053e-05, "loss": 0.0168, "step": 60980 }, { "grad_norm": 0.17497892677783966, "learning_rate": 3.6147329828460325e-05, "loss": 0.0129, "step": 60990 }, { "grad_norm": 0.19836848974227905, "learning_rate": 3.6131443146697306e-05, "loss": 0.0162, "step": 61000 }, { "grad_norm": 0.17028166353702545, "learning_rate": 3.611555798157879e-05, "loss": 0.015, "step": 61010 }, { "grad_norm": 0.14507776498794556, "learning_rate": 3.6099674334842e-05, "loss": 0.0124, "step": 61020 }, { "grad_norm": 0.1530158370733261, "learning_rate": 3.6083792208223934e-05, "loss": 0.0156, "step": 61030 }, { "grad_norm": 0.19113072752952576, "learning_rate": 3.606791160346142e-05, "loss": 0.0146, "step": 61040 }, { "grad_norm": 0.15912462770938873, "learning_rate": 3.6052032522291134e-05, "loss": 0.0153, "step": 61050 }, { "grad_norm": 0.19059449434280396, "learning_rate": 3.603615496644962e-05, "loss": 0.0158, "step": 61060 }, { "grad_norm": 0.15198323130607605, "learning_rate": 3.60202789376732e-05, "loss": 0.0122, "step": 61070 }, { "grad_norm": 0.16794629395008087, "learning_rate": 3.6004404437698036e-05, "loss": 0.0146, "step": 61080 }, { "grad_norm": 0.12632589042186737, "learning_rate": 3.5988531468260184e-05, "loss": 0.0155, "step": 61090 }, { "grad_norm": 0.2367367297410965, "learning_rate": 3.597266003109546e-05, "loss": 0.0143, "step": 61100 }, { "grad_norm": 0.14515037834644318, "learning_rate": 3.5956790127939557e-05, "loss": 0.015, "step": 61110 }, { "grad_norm": 0.1533246636390686, "learning_rate": 3.594092176052796e-05, "loss": 0.0177, "step": 61120 }, { "grad_norm": 0.2130097597837448, "learning_rate": 3.592505493059603e-05, "loss": 0.0143, "step": 61130 }, { "grad_norm": 0.18630683422088623, "learning_rate": 3.5909189639878905e-05, "loss": 0.0221, "step": 61140 }, { "grad_norm": 0.19289608299732208, "learning_rate": 3.5893325890111626e-05, "loss": 0.0146, "step": 61150 }, { "grad_norm": 0.14962638914585114, "learning_rate": 3.5877463683029034e-05, "loss": 0.0175, "step": 61160 }, { "grad_norm": 0.1526431292295456, "learning_rate": 3.586160302036578e-05, "loss": 0.0141, "step": 61170 }, { "grad_norm": 0.1359158605337143, "learning_rate": 3.584574390385636e-05, "loss": 0.014, "step": 61180 }, { "grad_norm": 0.17888569831848145, "learning_rate": 3.582988633523511e-05, "loss": 0.0136, "step": 61190 }, { "grad_norm": 0.14892645180225372, "learning_rate": 3.581403031623616e-05, "loss": 0.015, "step": 61200 }, { "grad_norm": 0.13196437060832977, "learning_rate": 3.579817584859354e-05, "loss": 0.014, "step": 61210 }, { "grad_norm": 0.2175494134426117, "learning_rate": 3.578232293404107e-05, "loss": 0.0152, "step": 61220 }, { "grad_norm": 0.19779030978679657, "learning_rate": 3.5766471574312385e-05, "loss": 0.0131, "step": 61230 }, { "grad_norm": 0.14120353758335114, "learning_rate": 3.5750621771140964e-05, "loss": 0.0121, "step": 61240 }, { "grad_norm": 0.1366971880197525, "learning_rate": 3.5734773526260115e-05, "loss": 0.0144, "step": 61250 }, { "grad_norm": 0.18308037519454956, "learning_rate": 3.5718926841402996e-05, "loss": 0.0174, "step": 61260 }, { "grad_norm": 0.1666542887687683, "learning_rate": 3.5703081718302555e-05, "loss": 0.0149, "step": 61270 }, { "grad_norm": 0.20067353546619415, "learning_rate": 3.568723815869162e-05, "loss": 0.0148, "step": 61280 }, { "grad_norm": 0.20540522038936615, "learning_rate": 3.5671396164302777e-05, "loss": 0.0159, "step": 61290 }, { "grad_norm": 0.14260946214199066, "learning_rate": 3.565555573686853e-05, "loss": 0.0126, "step": 61300 }, { "grad_norm": 0.20566800236701965, "learning_rate": 3.563971687812113e-05, "loss": 0.0167, "step": 61310 }, { "grad_norm": 0.16949564218521118, "learning_rate": 3.5623879589792706e-05, "loss": 0.0155, "step": 61320 }, { "grad_norm": 0.17449595034122467, "learning_rate": 3.5608043873615185e-05, "loss": 0.0138, "step": 61330 }, { "grad_norm": 0.1460803896188736, "learning_rate": 3.559220973132035e-05, "loss": 0.0138, "step": 61340 }, { "grad_norm": 0.30373895168304443, "learning_rate": 3.557637716463981e-05, "loss": 0.0129, "step": 61350 }, { "grad_norm": 0.202446848154068, "learning_rate": 3.556054617530497e-05, "loss": 0.0153, "step": 61360 }, { "grad_norm": 0.11587465554475784, "learning_rate": 3.5544716765047106e-05, "loss": 0.0164, "step": 61370 }, { "grad_norm": 0.22779272496700287, "learning_rate": 3.5528888935597274e-05, "loss": 0.014, "step": 61380 }, { "grad_norm": 0.18565696477890015, "learning_rate": 3.551306268868636e-05, "loss": 0.0156, "step": 61390 }, { "grad_norm": 0.17826871573925018, "learning_rate": 3.5497238026045166e-05, "loss": 0.0182, "step": 61400 }, { "grad_norm": 0.1773260086774826, "learning_rate": 3.548141494940421e-05, "loss": 0.0137, "step": 61410 }, { "grad_norm": 0.18596668541431427, "learning_rate": 3.546559346049388e-05, "loss": 0.0124, "step": 61420 }, { "grad_norm": 0.18330813944339752, "learning_rate": 3.544977356104441e-05, "loss": 0.0144, "step": 61430 }, { "grad_norm": 0.18430151045322418, "learning_rate": 3.543395525278582e-05, "loss": 0.0149, "step": 61440 }, { "grad_norm": 0.21742281317710876, "learning_rate": 3.5418138537447985e-05, "loss": 0.0163, "step": 61450 }, { "grad_norm": 0.13159047067165375, "learning_rate": 3.5402323416760574e-05, "loss": 0.0122, "step": 61460 }, { "grad_norm": 0.18965904414653778, "learning_rate": 3.538650989245316e-05, "loss": 0.0152, "step": 61470 }, { "grad_norm": 0.13831104338169098, "learning_rate": 3.537069796625504e-05, "loss": 0.0142, "step": 61480 }, { "grad_norm": 0.24978744983673096, "learning_rate": 3.53548876398954e-05, "loss": 0.0118, "step": 61490 }, { "grad_norm": 0.13606414198875427, "learning_rate": 3.5339078915103214e-05, "loss": 0.0151, "step": 61500 }, { "grad_norm": 0.14547066390514374, "learning_rate": 3.532327179360733e-05, "loss": 0.0147, "step": 61510 }, { "grad_norm": 0.19063574075698853, "learning_rate": 3.530746627713636e-05, "loss": 0.0128, "step": 61520 }, { "grad_norm": 0.23734913766384125, "learning_rate": 3.5291662367418805e-05, "loss": 0.012, "step": 61530 }, { "grad_norm": 0.14122354984283447, "learning_rate": 3.527586006618293e-05, "loss": 0.0168, "step": 61540 }, { "grad_norm": 0.26219022274017334, "learning_rate": 3.526005937515687e-05, "loss": 0.0192, "step": 61550 }, { "grad_norm": 0.19099971652030945, "learning_rate": 3.524426029606856e-05, "loss": 0.0164, "step": 61560 }, { "grad_norm": 0.20997513830661774, "learning_rate": 3.5228462830645746e-05, "loss": 0.0133, "step": 61570 }, { "grad_norm": 0.2212866246700287, "learning_rate": 3.5212666980616036e-05, "loss": 0.0145, "step": 61580 }, { "grad_norm": 0.20097662508487701, "learning_rate": 3.519687274770681e-05, "loss": 0.015, "step": 61590 }, { "grad_norm": 0.30714961886405945, "learning_rate": 3.518108013364535e-05, "loss": 0.0166, "step": 61600 }, { "grad_norm": 0.2012624740600586, "learning_rate": 3.516528914015868e-05, "loss": 0.0169, "step": 61610 }, { "grad_norm": 0.17391452193260193, "learning_rate": 3.5149499768973694e-05, "loss": 0.0154, "step": 61620 }, { "grad_norm": 0.1544317901134491, "learning_rate": 3.513371202181708e-05, "loss": 0.0137, "step": 61630 }, { "grad_norm": 0.1737385094165802, "learning_rate": 3.511792590041537e-05, "loss": 0.0157, "step": 61640 }, { "grad_norm": 0.16901695728302002, "learning_rate": 3.510214140649488e-05, "loss": 0.0158, "step": 61650 }, { "grad_norm": 0.1768130511045456, "learning_rate": 3.508635854178184e-05, "loss": 0.0167, "step": 61660 }, { "grad_norm": 0.20595216751098633, "learning_rate": 3.50705773080022e-05, "loss": 0.0135, "step": 61670 }, { "grad_norm": 0.1527765542268753, "learning_rate": 3.505479770688178e-05, "loss": 0.014, "step": 61680 }, { "grad_norm": 0.20364251732826233, "learning_rate": 3.503901974014622e-05, "loss": 0.0136, "step": 61690 }, { "grad_norm": 0.13760527968406677, "learning_rate": 3.502324340952095e-05, "loss": 0.0153, "step": 61700 }, { "grad_norm": 0.20597094297409058, "learning_rate": 3.500746871673126e-05, "loss": 0.0125, "step": 61710 }, { "grad_norm": 0.6248281002044678, "learning_rate": 3.499169566350227e-05, "loss": 0.0179, "step": 61720 }, { "grad_norm": 0.16467274725437164, "learning_rate": 3.4975924251558876e-05, "loss": 0.0169, "step": 61730 }, { "grad_norm": 0.19127421081066132, "learning_rate": 3.496015448262582e-05, "loss": 0.0204, "step": 61740 }, { "grad_norm": 0.23332428932189941, "learning_rate": 3.494438635842764e-05, "loss": 0.0171, "step": 61750 }, { "grad_norm": 0.16811804473400116, "learning_rate": 3.492861988068874e-05, "loss": 0.0148, "step": 61760 }, { "grad_norm": 0.19689470529556274, "learning_rate": 3.491285505113332e-05, "loss": 0.0147, "step": 61770 }, { "grad_norm": 0.1522635519504547, "learning_rate": 3.4897091871485376e-05, "loss": 0.0137, "step": 61780 }, { "grad_norm": 0.22562560439109802, "learning_rate": 3.488133034346875e-05, "loss": 0.0148, "step": 61790 }, { "grad_norm": 0.19983264803886414, "learning_rate": 3.486557046880713e-05, "loss": 0.0152, "step": 61800 }, { "grad_norm": 0.2105906754732132, "learning_rate": 3.484981224922395e-05, "loss": 0.0165, "step": 61810 }, { "grad_norm": 0.17788955569267273, "learning_rate": 3.4834055686442544e-05, "loss": 0.0141, "step": 61820 }, { "grad_norm": 0.19053658843040466, "learning_rate": 3.4818300782186e-05, "loss": 0.0139, "step": 61830 }, { "grad_norm": 0.20536795258522034, "learning_rate": 3.4802547538177236e-05, "loss": 0.0163, "step": 61840 }, { "grad_norm": 0.1897292584180832, "learning_rate": 3.478679595613905e-05, "loss": 0.0183, "step": 61850 }, { "grad_norm": 0.12491541355848312, "learning_rate": 3.4771046037793975e-05, "loss": 0.0116, "step": 61860 }, { "grad_norm": 0.16511528193950653, "learning_rate": 3.4755297784864416e-05, "loss": 0.0135, "step": 61870 }, { "grad_norm": 0.18010248243808746, "learning_rate": 3.473955119907257e-05, "loss": 0.0127, "step": 61880 }, { "grad_norm": 0.14920102059841156, "learning_rate": 3.472380628214046e-05, "loss": 0.0114, "step": 61890 }, { "grad_norm": 0.18919353187084198, "learning_rate": 3.4708063035789906e-05, "loss": 0.0149, "step": 61900 }, { "grad_norm": 0.1448410451412201, "learning_rate": 3.4692321461742614e-05, "loss": 0.0115, "step": 61910 }, { "grad_norm": 0.1297789365053177, "learning_rate": 3.4676581561720023e-05, "loss": 0.0143, "step": 61920 }, { "grad_norm": 0.13880909979343414, "learning_rate": 3.466084333744344e-05, "loss": 0.0119, "step": 61930 }, { "grad_norm": 0.1956467628479004, "learning_rate": 3.4645106790633964e-05, "loss": 0.0154, "step": 61940 }, { "grad_norm": 0.1436014473438263, "learning_rate": 3.4629371923012526e-05, "loss": 0.0131, "step": 61950 }, { "grad_norm": 0.13339461386203766, "learning_rate": 3.4613638736299856e-05, "loss": 0.0145, "step": 61960 }, { "grad_norm": 0.1655425727367401, "learning_rate": 3.459790723221652e-05, "loss": 0.0149, "step": 61970 }, { "grad_norm": 0.1319342404603958, "learning_rate": 3.45821774124829e-05, "loss": 0.0169, "step": 61980 }, { "grad_norm": 0.18085448443889618, "learning_rate": 3.456644927881918e-05, "loss": 0.0151, "step": 61990 }, { "grad_norm": 0.1988959163427353, "learning_rate": 3.4550722832945354e-05, "loss": 0.0128, "step": 62000 }, { "grad_norm": 0.14806799590587616, "learning_rate": 3.453499807658126e-05, "loss": 0.0143, "step": 62010 }, { "grad_norm": 0.12955206632614136, "learning_rate": 3.451927501144653e-05, "loss": 0.0156, "step": 62020 }, { "grad_norm": 0.15567350387573242, "learning_rate": 3.450355363926059e-05, "loss": 0.0155, "step": 62030 }, { "grad_norm": 0.16152486205101013, "learning_rate": 3.448783396174273e-05, "loss": 0.0147, "step": 62040 }, { "grad_norm": 0.18548418581485748, "learning_rate": 3.4472115980612034e-05, "loss": 0.014, "step": 62050 }, { "grad_norm": 0.13412509858608246, "learning_rate": 3.4456399697587396e-05, "loss": 0.0136, "step": 62060 }, { "grad_norm": 0.2073907107114792, "learning_rate": 3.4440685114387506e-05, "loss": 0.0174, "step": 62070 }, { "grad_norm": 0.17163792252540588, "learning_rate": 3.442497223273091e-05, "loss": 0.0159, "step": 62080 }, { "grad_norm": 0.15323831140995026, "learning_rate": 3.4409261054335926e-05, "loss": 0.015, "step": 62090 }, { "grad_norm": 0.11188719421625137, "learning_rate": 3.439355158092069e-05, "loss": 0.014, "step": 62100 }, { "grad_norm": 0.17975470423698425, "learning_rate": 3.437784381420322e-05, "loss": 0.0145, "step": 62110 }, { "grad_norm": 0.12475201487541199, "learning_rate": 3.436213775590126e-05, "loss": 0.0119, "step": 62120 }, { "grad_norm": 0.2005312591791153, "learning_rate": 3.4346433407732405e-05, "loss": 0.015, "step": 62130 }, { "grad_norm": 0.11195625364780426, "learning_rate": 3.433073077141405e-05, "loss": 0.0139, "step": 62140 }, { "grad_norm": 0.1790318787097931, "learning_rate": 3.431502984866343e-05, "loss": 0.0192, "step": 62150 }, { "grad_norm": 0.1316610872745514, "learning_rate": 3.429933064119753e-05, "loss": 0.0129, "step": 62160 }, { "grad_norm": 0.1960139125585556, "learning_rate": 3.4283633150733255e-05, "loss": 0.0146, "step": 62170 }, { "grad_norm": 0.15382084250450134, "learning_rate": 3.4267937378987225e-05, "loss": 0.0118, "step": 62180 }, { "grad_norm": 0.20857784152030945, "learning_rate": 3.425224332767592e-05, "loss": 0.0159, "step": 62190 }, { "grad_norm": 0.14467011392116547, "learning_rate": 3.4236550998515595e-05, "loss": 0.0138, "step": 62200 }, { "grad_norm": 0.2262643575668335, "learning_rate": 3.422086039322235e-05, "loss": 0.0125, "step": 62210 }, { "grad_norm": 0.18221546709537506, "learning_rate": 3.420517151351209e-05, "loss": 0.016, "step": 62220 }, { "grad_norm": 0.11881387233734131, "learning_rate": 3.418948436110053e-05, "loss": 0.0174, "step": 62230 }, { "grad_norm": 0.1569199413061142, "learning_rate": 3.417379893770319e-05, "loss": 0.0166, "step": 62240 }, { "grad_norm": 0.15884509682655334, "learning_rate": 3.41581152450354e-05, "loss": 0.0138, "step": 62250 }, { "grad_norm": 0.13776826858520508, "learning_rate": 3.414243328481231e-05, "loss": 0.0149, "step": 62260 }, { "grad_norm": 0.15231382846832275, "learning_rate": 3.412675305874888e-05, "loss": 0.0137, "step": 62270 }, { "grad_norm": 0.1921265721321106, "learning_rate": 3.4111074568559876e-05, "loss": 0.0143, "step": 62280 }, { "grad_norm": 0.16071076691150665, "learning_rate": 3.4095397815959846e-05, "loss": 0.0133, "step": 62290 }, { "grad_norm": 0.22900377213954926, "learning_rate": 3.407972280266322e-05, "loss": 0.0138, "step": 62300 }, { "grad_norm": 0.11354949325323105, "learning_rate": 3.4064049530384166e-05, "loss": 0.0148, "step": 62310 }, { "grad_norm": 0.18996448814868927, "learning_rate": 3.40483780008367e-05, "loss": 0.0143, "step": 62320 }, { "grad_norm": 0.15358726680278778, "learning_rate": 3.4032708215734635e-05, "loss": 0.0151, "step": 62330 }, { "grad_norm": 0.2565928101539612, "learning_rate": 3.4017040176791584e-05, "loss": 0.0127, "step": 62340 }, { "grad_norm": 0.14467528462409973, "learning_rate": 3.4001373885720976e-05, "loss": 0.0125, "step": 62350 }, { "grad_norm": 0.15406309068202972, "learning_rate": 3.398570934423608e-05, "loss": 0.0159, "step": 62360 }, { "grad_norm": 0.1426580399274826, "learning_rate": 3.397004655404994e-05, "loss": 0.0125, "step": 62370 }, { "grad_norm": 0.1486644148826599, "learning_rate": 3.395438551687541e-05, "loss": 0.0138, "step": 62380 }, { "grad_norm": 0.13461841642856598, "learning_rate": 3.393872623442515e-05, "loss": 0.0133, "step": 62390 }, { "grad_norm": 0.14441508054733276, "learning_rate": 3.392306870841164e-05, "loss": 0.0143, "step": 62400 }, { "grad_norm": 0.18444420397281647, "learning_rate": 3.390741294054715e-05, "loss": 0.0167, "step": 62410 }, { "grad_norm": 0.1315993070602417, "learning_rate": 3.38917589325438e-05, "loss": 0.0132, "step": 62420 }, { "grad_norm": 0.12230654805898666, "learning_rate": 3.387610668611349e-05, "loss": 0.0105, "step": 62430 }, { "grad_norm": 0.22340275347232819, "learning_rate": 3.38604562029679e-05, "loss": 0.0166, "step": 62440 }, { "grad_norm": 0.26498961448669434, "learning_rate": 3.384480748481855e-05, "loss": 0.0151, "step": 62450 }, { "grad_norm": 0.1922302097082138, "learning_rate": 3.382916053337676e-05, "loss": 0.0123, "step": 62460 }, { "grad_norm": 0.16025623679161072, "learning_rate": 3.381351535035366e-05, "loss": 0.0172, "step": 62470 }, { "grad_norm": 0.1358744204044342, "learning_rate": 3.379787193746019e-05, "loss": 0.0128, "step": 62480 }, { "grad_norm": 0.11009596288204193, "learning_rate": 3.378223029640709e-05, "loss": 0.0138, "step": 62490 }, { "grad_norm": 0.20442719757556915, "learning_rate": 3.376659042890489e-05, "loss": 0.0129, "step": 62500 }, { "grad_norm": 0.17926521599292755, "learning_rate": 3.375095233666397e-05, "loss": 0.0144, "step": 62510 }, { "grad_norm": 0.13100925087928772, "learning_rate": 3.373531602139447e-05, "loss": 0.0126, "step": 62520 }, { "grad_norm": 0.21797418594360352, "learning_rate": 3.371968148480635e-05, "loss": 0.0127, "step": 62530 }, { "grad_norm": 0.186226025223732, "learning_rate": 3.370404872860937e-05, "loss": 0.016, "step": 62540 }, { "grad_norm": 0.15633739531040192, "learning_rate": 3.3688417754513145e-05, "loss": 0.0135, "step": 62550 }, { "grad_norm": 0.16956588625907898, "learning_rate": 3.367278856422703e-05, "loss": 0.0196, "step": 62560 }, { "grad_norm": 0.15949507057666779, "learning_rate": 3.365716115946021e-05, "loss": 0.0175, "step": 62570 }, { "grad_norm": 0.22104567289352417, "learning_rate": 3.3641535541921666e-05, "loss": 0.0149, "step": 62580 }, { "grad_norm": 0.2172219455242157, "learning_rate": 3.36259117133202e-05, "loss": 0.0163, "step": 62590 }, { "grad_norm": 0.1597977876663208, "learning_rate": 3.36102896753644e-05, "loss": 0.0146, "step": 62600 }, { "grad_norm": 0.1579057276248932, "learning_rate": 3.35946694297627e-05, "loss": 0.0132, "step": 62610 }, { "grad_norm": 0.18898536264896393, "learning_rate": 3.357905097822328e-05, "loss": 0.011, "step": 62620 }, { "grad_norm": 0.28845325112342834, "learning_rate": 3.356343432245416e-05, "loss": 0.014, "step": 62630 }, { "grad_norm": 0.25262364745140076, "learning_rate": 3.354781946416315e-05, "loss": 0.0136, "step": 62640 }, { "grad_norm": 0.15121042728424072, "learning_rate": 3.3532206405057864e-05, "loss": 0.0164, "step": 62650 }, { "grad_norm": 0.22517825663089752, "learning_rate": 3.351659514684573e-05, "loss": 0.0167, "step": 62660 }, { "grad_norm": 0.1675271987915039, "learning_rate": 3.350098569123395e-05, "loss": 0.0179, "step": 62670 }, { "grad_norm": 0.1486285924911499, "learning_rate": 3.348537803992958e-05, "loss": 0.0118, "step": 62680 }, { "grad_norm": 0.20830459892749786, "learning_rate": 3.346977219463945e-05, "loss": 0.0163, "step": 62690 }, { "grad_norm": 0.3393903076648712, "learning_rate": 3.345416815707017e-05, "loss": 0.0137, "step": 62700 }, { "grad_norm": 0.18531456589698792, "learning_rate": 3.3438565928928185e-05, "loss": 0.0145, "step": 62710 }, { "grad_norm": 0.16081002354621887, "learning_rate": 3.3422965511919736e-05, "loss": 0.0166, "step": 62720 }, { "grad_norm": 0.20693011581897736, "learning_rate": 3.3407366907750846e-05, "loss": 0.0155, "step": 62730 }, { "grad_norm": 0.14324423670768738, "learning_rate": 3.339177011812737e-05, "loss": 0.0165, "step": 62740 }, { "grad_norm": 0.271578311920166, "learning_rate": 3.3376175144754935e-05, "loss": 0.0159, "step": 62750 }, { "grad_norm": 0.1287883073091507, "learning_rate": 3.3360581989339004e-05, "loss": 0.0163, "step": 62760 }, { "grad_norm": 0.18693745136260986, "learning_rate": 3.3344990653584804e-05, "loss": 0.0159, "step": 62770 }, { "grad_norm": 0.15624435245990753, "learning_rate": 3.332940113919739e-05, "loss": 0.0146, "step": 62780 }, { "grad_norm": 0.19645527005195618, "learning_rate": 3.331381344788161e-05, "loss": 0.02, "step": 62790 }, { "grad_norm": 0.11628705263137817, "learning_rate": 3.3298227581342076e-05, "loss": 0.0126, "step": 62800 }, { "grad_norm": 0.15514865517616272, "learning_rate": 3.328264354128329e-05, "loss": 0.0132, "step": 62810 }, { "grad_norm": 0.17929653823375702, "learning_rate": 3.3267061329409465e-05, "loss": 0.016, "step": 62820 }, { "grad_norm": 0.18193688988685608, "learning_rate": 3.325148094742466e-05, "loss": 0.0132, "step": 62830 }, { "grad_norm": 0.1582607626914978, "learning_rate": 3.323590239703271e-05, "loss": 0.0171, "step": 62840 }, { "grad_norm": 0.192892923951149, "learning_rate": 3.322032567993728e-05, "loss": 0.015, "step": 62850 }, { "grad_norm": 0.1194608137011528, "learning_rate": 3.3204750797841774e-05, "loss": 0.0123, "step": 62860 }, { "grad_norm": 0.1097608357667923, "learning_rate": 3.31891777524495e-05, "loss": 0.0127, "step": 62870 }, { "grad_norm": 0.22213464975357056, "learning_rate": 3.317360654546347e-05, "loss": 0.0162, "step": 62880 }, { "grad_norm": 0.1342921406030655, "learning_rate": 3.3158037178586534e-05, "loss": 0.0145, "step": 62890 }, { "grad_norm": 0.20194412767887115, "learning_rate": 3.3142469653521316e-05, "loss": 0.0132, "step": 62900 }, { "grad_norm": 0.15319140255451202, "learning_rate": 3.3126903971970284e-05, "loss": 0.0141, "step": 62910 }, { "grad_norm": 0.18356412649154663, "learning_rate": 3.311134013563564e-05, "loss": 0.0192, "step": 62920 }, { "grad_norm": 0.21837297081947327, "learning_rate": 3.309577814621947e-05, "loss": 0.0139, "step": 62930 }, { "grad_norm": 0.17677180469036102, "learning_rate": 3.308021800542359e-05, "loss": 0.0142, "step": 62940 }, { "grad_norm": 0.1730283796787262, "learning_rate": 3.306465971494963e-05, "loss": 0.0148, "step": 62950 }, { "grad_norm": 0.13268132507801056, "learning_rate": 3.304910327649901e-05, "loss": 0.0121, "step": 62960 }, { "grad_norm": 0.11685744673013687, "learning_rate": 3.303354869177297e-05, "loss": 0.0153, "step": 62970 }, { "grad_norm": 0.1996338665485382, "learning_rate": 3.301799596247255e-05, "loss": 0.0131, "step": 62980 }, { "grad_norm": 0.1434859037399292, "learning_rate": 3.300244509029854e-05, "loss": 0.013, "step": 62990 }, { "grad_norm": 0.14406220614910126, "learning_rate": 3.2986896076951585e-05, "loss": 0.0152, "step": 63000 }, { "grad_norm": 0.1147104948759079, "learning_rate": 3.297134892413211e-05, "loss": 0.0122, "step": 63010 }, { "grad_norm": 0.1846507489681244, "learning_rate": 3.295580363354031e-05, "loss": 0.0136, "step": 63020 }, { "grad_norm": 0.20150870084762573, "learning_rate": 3.2940260206876194e-05, "loss": 0.0127, "step": 63030 }, { "grad_norm": 0.13443616032600403, "learning_rate": 3.292471864583958e-05, "loss": 0.0138, "step": 63040 }, { "grad_norm": 0.1626628190279007, "learning_rate": 3.290917895213004e-05, "loss": 0.0126, "step": 63050 }, { "grad_norm": 0.17931602895259857, "learning_rate": 3.289364112744701e-05, "loss": 0.014, "step": 63060 }, { "grad_norm": 0.2115829735994339, "learning_rate": 3.287810517348967e-05, "loss": 0.0183, "step": 63070 }, { "grad_norm": 0.15133753418922424, "learning_rate": 3.2862571091957014e-05, "loss": 0.0153, "step": 63080 }, { "grad_norm": 0.37411826848983765, "learning_rate": 3.284703888454781e-05, "loss": 0.0184, "step": 63090 }, { "grad_norm": 0.18129459023475647, "learning_rate": 3.2831508552960646e-05, "loss": 0.0156, "step": 63100 }, { "grad_norm": 0.18923130631446838, "learning_rate": 3.281598009889387e-05, "loss": 0.0144, "step": 63110 }, { "grad_norm": 0.1908646672964096, "learning_rate": 3.28004535240457e-05, "loss": 0.0133, "step": 63120 }, { "grad_norm": 0.14732737839221954, "learning_rate": 3.278492883011407e-05, "loss": 0.0103, "step": 63130 }, { "grad_norm": 0.1798088103532791, "learning_rate": 3.276940601879675e-05, "loss": 0.0114, "step": 63140 }, { "grad_norm": 0.22606554627418518, "learning_rate": 3.275388509179127e-05, "loss": 0.0134, "step": 63150 }, { "grad_norm": 0.16865837574005127, "learning_rate": 3.2738366050794985e-05, "loss": 0.0114, "step": 63160 }, { "grad_norm": 0.2108023464679718, "learning_rate": 3.2722848897505034e-05, "loss": 0.0157, "step": 63170 }, { "grad_norm": 0.15925362706184387, "learning_rate": 3.2707333633618356e-05, "loss": 0.0138, "step": 63180 }, { "grad_norm": 0.17720292508602142, "learning_rate": 3.269182026083167e-05, "loss": 0.0138, "step": 63190 }, { "grad_norm": 0.13383598625659943, "learning_rate": 3.2676308780841516e-05, "loss": 0.013, "step": 63200 }, { "grad_norm": 0.14223140478134155, "learning_rate": 3.266079919534416e-05, "loss": 0.0138, "step": 63210 }, { "grad_norm": 0.16126398742198944, "learning_rate": 3.264529150603575e-05, "loss": 0.014, "step": 63220 }, { "grad_norm": 0.1489250659942627, "learning_rate": 3.262978571461217e-05, "loss": 0.0136, "step": 63230 }, { "grad_norm": 0.25073757767677307, "learning_rate": 3.26142818227691e-05, "loss": 0.0192, "step": 63240 }, { "grad_norm": 0.16959542036056519, "learning_rate": 3.259877983220203e-05, "loss": 0.0121, "step": 63250 }, { "grad_norm": 0.1712043583393097, "learning_rate": 3.2583279744606246e-05, "loss": 0.0101, "step": 63260 }, { "grad_norm": 0.1707921326160431, "learning_rate": 3.25677815616768e-05, "loss": 0.015, "step": 63270 }, { "grad_norm": 0.12868353724479675, "learning_rate": 3.255228528510856e-05, "loss": 0.0149, "step": 63280 }, { "grad_norm": 0.2025837004184723, "learning_rate": 3.253679091659617e-05, "loss": 0.0147, "step": 63290 }, { "grad_norm": 0.18113331496715546, "learning_rate": 3.252129845783404e-05, "loss": 0.0153, "step": 63300 }, { "grad_norm": 0.1751357465982437, "learning_rate": 3.2505807910516456e-05, "loss": 0.0136, "step": 63310 }, { "grad_norm": 0.171706885099411, "learning_rate": 3.249031927633742e-05, "loss": 0.0164, "step": 63320 }, { "grad_norm": 0.17497484385967255, "learning_rate": 3.2474832556990737e-05, "loss": 0.0137, "step": 63330 }, { "grad_norm": 0.1853996217250824, "learning_rate": 3.245934775417002e-05, "loss": 0.0167, "step": 63340 }, { "grad_norm": 0.15957039594650269, "learning_rate": 3.244386486956866e-05, "loss": 0.0134, "step": 63350 }, { "grad_norm": 0.20535360276699066, "learning_rate": 3.2428383904879846e-05, "loss": 0.0167, "step": 63360 }, { "grad_norm": 0.26195111870765686, "learning_rate": 3.241290486179652e-05, "loss": 0.012, "step": 63370 }, { "grad_norm": 0.12029056996107101, "learning_rate": 3.239742774201151e-05, "loss": 0.0136, "step": 63380 }, { "grad_norm": 0.15139947831630707, "learning_rate": 3.238195254721733e-05, "loss": 0.0156, "step": 63390 }, { "grad_norm": 0.14402025938034058, "learning_rate": 3.236647927910633e-05, "loss": 0.014, "step": 63400 }, { "grad_norm": 0.1255647838115692, "learning_rate": 3.2351007939370646e-05, "loss": 0.0135, "step": 63410 }, { "grad_norm": 0.14797230064868927, "learning_rate": 3.2335538529702184e-05, "loss": 0.0134, "step": 63420 }, { "grad_norm": 0.18912935256958008, "learning_rate": 3.232007105179268e-05, "loss": 0.0156, "step": 63430 }, { "grad_norm": 0.16185717284679413, "learning_rate": 3.230460550733363e-05, "loss": 0.0102, "step": 63440 }, { "grad_norm": 0.1608424335718155, "learning_rate": 3.2289141898016315e-05, "loss": 0.0131, "step": 63450 }, { "grad_norm": 0.1555277556180954, "learning_rate": 3.227368022553181e-05, "loss": 0.0133, "step": 63460 }, { "grad_norm": 0.1609567552804947, "learning_rate": 3.225822049157099e-05, "loss": 0.0159, "step": 63470 }, { "grad_norm": 0.13804452121257782, "learning_rate": 3.2242762697824494e-05, "loss": 0.0125, "step": 63480 }, { "grad_norm": 0.1587376594543457, "learning_rate": 3.222730684598279e-05, "loss": 0.0161, "step": 63490 }, { "grad_norm": 0.14306584000587463, "learning_rate": 3.221185293773605e-05, "loss": 0.0155, "step": 63500 }, { "grad_norm": 0.18322031199932098, "learning_rate": 3.2196400974774354e-05, "loss": 0.0132, "step": 63510 }, { "grad_norm": 0.13247619569301605, "learning_rate": 3.2180950958787485e-05, "loss": 0.0143, "step": 63520 }, { "grad_norm": 0.15870004892349243, "learning_rate": 3.2165502891465024e-05, "loss": 0.0162, "step": 63530 }, { "grad_norm": 0.1529301106929779, "learning_rate": 3.215005677449636e-05, "loss": 0.0132, "step": 63540 }, { "grad_norm": 0.16793590784072876, "learning_rate": 3.2134612609570644e-05, "loss": 0.0131, "step": 63550 }, { "grad_norm": 0.1483837515115738, "learning_rate": 3.211917039837682e-05, "loss": 0.0148, "step": 63560 }, { "grad_norm": 0.15511788427829742, "learning_rate": 3.210373014260365e-05, "loss": 0.0143, "step": 63570 }, { "grad_norm": 0.2134050875902176, "learning_rate": 3.208829184393965e-05, "loss": 0.0163, "step": 63580 }, { "grad_norm": 0.1672612577676773, "learning_rate": 3.207285550407312e-05, "loss": 0.0139, "step": 63590 }, { "grad_norm": 0.1423557549715042, "learning_rate": 3.205742112469217e-05, "loss": 0.0147, "step": 63600 }, { "grad_norm": 0.22046101093292236, "learning_rate": 3.204198870748465e-05, "loss": 0.0168, "step": 63610 }, { "grad_norm": 0.13411076366901398, "learning_rate": 3.202655825413824e-05, "loss": 0.0122, "step": 63620 }, { "grad_norm": 0.14608284831047058, "learning_rate": 3.2011129766340416e-05, "loss": 0.0132, "step": 63630 }, { "grad_norm": 0.13698510825634003, "learning_rate": 3.1995703245778384e-05, "loss": 0.0109, "step": 63640 }, { "grad_norm": 0.23049478232860565, "learning_rate": 3.1980278694139186e-05, "loss": 0.0151, "step": 63650 }, { "grad_norm": 0.11693986505270004, "learning_rate": 3.196485611310962e-05, "loss": 0.0105, "step": 63660 }, { "grad_norm": 0.1056467592716217, "learning_rate": 3.1949435504376257e-05, "loss": 0.0109, "step": 63670 }, { "grad_norm": 0.1460014134645462, "learning_rate": 3.1934016869625505e-05, "loss": 0.0119, "step": 63680 }, { "grad_norm": 0.11938739567995071, "learning_rate": 3.191860021054349e-05, "loss": 0.0141, "step": 63690 }, { "grad_norm": 0.1823347806930542, "learning_rate": 3.190318552881618e-05, "loss": 0.0131, "step": 63700 }, { "grad_norm": 0.16591785848140717, "learning_rate": 3.1887772826129285e-05, "loss": 0.013, "step": 63710 }, { "grad_norm": 0.1535477489233017, "learning_rate": 3.1872362104168335e-05, "loss": 0.0156, "step": 63720 }, { "grad_norm": 0.19160474836826324, "learning_rate": 3.185695336461861e-05, "loss": 0.0153, "step": 63730 }, { "grad_norm": 0.14619936048984528, "learning_rate": 3.184154660916519e-05, "loss": 0.0132, "step": 63740 }, { "grad_norm": 0.14138120412826538, "learning_rate": 3.1826141839492906e-05, "loss": 0.0129, "step": 63750 }, { "grad_norm": 0.14634625613689423, "learning_rate": 3.181073905728644e-05, "loss": 0.0129, "step": 63760 }, { "grad_norm": 0.15815268456935883, "learning_rate": 3.1795338264230214e-05, "loss": 0.0125, "step": 63770 }, { "grad_norm": 0.12609535455703735, "learning_rate": 3.1779939462008426e-05, "loss": 0.0107, "step": 63780 }, { "grad_norm": 0.15267331898212433, "learning_rate": 3.1764542652305066e-05, "loss": 0.0119, "step": 63790 }, { "grad_norm": 0.18289503455162048, "learning_rate": 3.17491478368039e-05, "loss": 0.0121, "step": 63800 }, { "grad_norm": 0.16082753241062164, "learning_rate": 3.173375501718847e-05, "loss": 0.0137, "step": 63810 }, { "grad_norm": 0.14652509987354279, "learning_rate": 3.171836419514215e-05, "loss": 0.0141, "step": 63820 }, { "grad_norm": 0.16734819114208221, "learning_rate": 3.170297537234803e-05, "loss": 0.0131, "step": 63830 }, { "grad_norm": 0.20891845226287842, "learning_rate": 3.168758855048901e-05, "loss": 0.0126, "step": 63840 }, { "grad_norm": 0.16091394424438477, "learning_rate": 3.1672203731247785e-05, "loss": 0.0125, "step": 63850 }, { "grad_norm": 0.1752079576253891, "learning_rate": 3.16568209163068e-05, "loss": 0.0118, "step": 63860 }, { "grad_norm": 0.11887259781360626, "learning_rate": 3.1641440107348295e-05, "loss": 0.0156, "step": 63870 }, { "grad_norm": 0.19961664080619812, "learning_rate": 3.162606130605429e-05, "loss": 0.0186, "step": 63880 }, { "grad_norm": 0.24031329154968262, "learning_rate": 3.16106845141066e-05, "loss": 0.0147, "step": 63890 }, { "grad_norm": 0.24664521217346191, "learning_rate": 3.159530973318682e-05, "loss": 0.0148, "step": 63900 }, { "grad_norm": 0.16041648387908936, "learning_rate": 3.1579936964976286e-05, "loss": 0.0189, "step": 63910 }, { "grad_norm": 0.1645389050245285, "learning_rate": 3.156456621115615e-05, "loss": 0.014, "step": 63920 }, { "grad_norm": 0.16136467456817627, "learning_rate": 3.154919747340734e-05, "loss": 0.0134, "step": 63930 }, { "grad_norm": 0.15346011519432068, "learning_rate": 3.153383075341054e-05, "loss": 0.0132, "step": 63940 }, { "grad_norm": 0.1874593198299408, "learning_rate": 3.1518466052846264e-05, "loss": 0.0151, "step": 63950 }, { "grad_norm": 0.1541786789894104, "learning_rate": 3.150310337339474e-05, "loss": 0.0143, "step": 63960 }, { "grad_norm": 0.11726334691047668, "learning_rate": 3.148774271673603e-05, "loss": 0.0123, "step": 63970 }, { "grad_norm": 0.1346646100282669, "learning_rate": 3.147238408454995e-05, "loss": 0.0132, "step": 63980 }, { "grad_norm": 0.14923502504825592, "learning_rate": 3.1457027478516086e-05, "loss": 0.0124, "step": 63990 }, { "grad_norm": 0.16517414152622223, "learning_rate": 3.144167290031382e-05, "loss": 0.0143, "step": 64000 }, { "grad_norm": 0.18695440888404846, "learning_rate": 3.142632035162227e-05, "loss": 0.0147, "step": 64010 }, { "grad_norm": 0.1712917536497116, "learning_rate": 3.141096983412044e-05, "loss": 0.0166, "step": 64020 }, { "grad_norm": 0.19109179079532623, "learning_rate": 3.139562134948699e-05, "loss": 0.0134, "step": 64030 }, { "grad_norm": 0.14799974858760834, "learning_rate": 3.1380274899400424e-05, "loss": 0.0133, "step": 64040 }, { "grad_norm": 0.16796205937862396, "learning_rate": 3.1364930485539e-05, "loss": 0.0135, "step": 64050 }, { "grad_norm": 0.20315080881118774, "learning_rate": 3.134958810958076e-05, "loss": 0.0137, "step": 64060 }, { "grad_norm": 0.20703063905239105, "learning_rate": 3.13342477732035e-05, "loss": 0.013, "step": 64070 }, { "grad_norm": 0.13398566842079163, "learning_rate": 3.131890947808487e-05, "loss": 0.0148, "step": 64080 }, { "grad_norm": 0.16204962134361267, "learning_rate": 3.13035732259022e-05, "loss": 0.0143, "step": 64090 }, { "grad_norm": 0.18462006747722626, "learning_rate": 3.1288239018332656e-05, "loss": 0.0121, "step": 64100 }, { "grad_norm": 0.212051123380661, "learning_rate": 3.1272906857053164e-05, "loss": 0.0125, "step": 64110 }, { "grad_norm": 0.18185772001743317, "learning_rate": 3.125757674374042e-05, "loss": 0.013, "step": 64120 }, { "grad_norm": 0.12749320268630981, "learning_rate": 3.124224868007088e-05, "loss": 0.012, "step": 64130 }, { "grad_norm": 0.17506569623947144, "learning_rate": 3.1226922667720854e-05, "loss": 0.0122, "step": 64140 }, { "grad_norm": 0.16647936403751373, "learning_rate": 3.121159870836633e-05, "loss": 0.0177, "step": 64150 }, { "grad_norm": 0.18031850457191467, "learning_rate": 3.119627680368312e-05, "loss": 0.014, "step": 64160 }, { "grad_norm": 0.13533715903759003, "learning_rate": 3.1180956955346816e-05, "loss": 0.0146, "step": 64170 }, { "grad_norm": 0.2880708575248718, "learning_rate": 3.116563916503276e-05, "loss": 0.0122, "step": 64180 }, { "grad_norm": 0.17880885303020477, "learning_rate": 3.1150323434416095e-05, "loss": 0.0136, "step": 64190 }, { "grad_norm": 0.17806483805179596, "learning_rate": 3.1135009765171696e-05, "loss": 0.0144, "step": 64200 }, { "grad_norm": 0.20919932425022125, "learning_rate": 3.111969815897429e-05, "loss": 0.0174, "step": 64210 }, { "grad_norm": 0.11815004050731659, "learning_rate": 3.1104388617498295e-05, "loss": 0.0133, "step": 64220 }, { "grad_norm": 0.21202415227890015, "learning_rate": 3.1089081142417966e-05, "loss": 0.0147, "step": 64230 }, { "grad_norm": 0.1386653631925583, "learning_rate": 3.1073775735407285e-05, "loss": 0.0136, "step": 64240 }, { "grad_norm": 0.15606558322906494, "learning_rate": 3.1058472398140035e-05, "loss": 0.0141, "step": 64250 }, { "grad_norm": 0.1567949801683426, "learning_rate": 3.1043171132289736e-05, "loss": 0.0128, "step": 64260 }, { "grad_norm": 0.1227099746465683, "learning_rate": 3.1027871939529774e-05, "loss": 0.0155, "step": 64270 }, { "grad_norm": 0.17906132340431213, "learning_rate": 3.1012574821533214e-05, "loss": 0.0116, "step": 64280 }, { "grad_norm": 0.17373764514923096, "learning_rate": 3.099727977997291e-05, "loss": 0.0167, "step": 64290 }, { "grad_norm": 0.25525131821632385, "learning_rate": 3.098198681652154e-05, "loss": 0.0113, "step": 64300 }, { "grad_norm": 0.17079181969165802, "learning_rate": 3.0966695932851475e-05, "loss": 0.0128, "step": 64310 }, { "grad_norm": 0.1351926177740097, "learning_rate": 3.0951407130634926e-05, "loss": 0.0133, "step": 64320 }, { "grad_norm": 0.22207164764404297, "learning_rate": 3.0936120411543864e-05, "loss": 0.0133, "step": 64330 }, { "grad_norm": 0.15030387043952942, "learning_rate": 3.0920835777250015e-05, "loss": 0.0119, "step": 64340 }, { "grad_norm": 0.15792185068130493, "learning_rate": 3.090555322942488e-05, "loss": 0.0143, "step": 64350 }, { "grad_norm": 0.14226756989955902, "learning_rate": 3.089027276973973e-05, "loss": 0.0121, "step": 64360 }, { "grad_norm": 0.14401152729988098, "learning_rate": 3.0874994399865615e-05, "loss": 0.0103, "step": 64370 }, { "grad_norm": 0.12080750614404678, "learning_rate": 3.085971812147337e-05, "loss": 0.0129, "step": 64380 }, { "grad_norm": 0.13505449891090393, "learning_rate": 3.0844443936233557e-05, "loss": 0.0106, "step": 64390 }, { "grad_norm": 0.1469704955816269, "learning_rate": 3.082917184581657e-05, "loss": 0.0115, "step": 64400 }, { "grad_norm": 0.1693921834230423, "learning_rate": 3.0813901851892513e-05, "loss": 0.0169, "step": 64410 }, { "grad_norm": 0.16482692956924438, "learning_rate": 3.0798633956131316e-05, "loss": 0.0154, "step": 64420 }, { "grad_norm": 0.2770095467567444, "learning_rate": 3.078336816020264e-05, "loss": 0.0128, "step": 64430 }, { "grad_norm": 0.22550570964813232, "learning_rate": 3.0768104465775935e-05, "loss": 0.0138, "step": 64440 }, { "grad_norm": 0.16752716898918152, "learning_rate": 3.0752842874520396e-05, "loss": 0.0149, "step": 64450 }, { "grad_norm": 0.3140951693058014, "learning_rate": 3.073758338810503e-05, "loss": 0.0146, "step": 64460 }, { "grad_norm": 0.16695760190486908, "learning_rate": 3.072232600819859e-05, "loss": 0.0129, "step": 64470 }, { "grad_norm": 0.1766061931848526, "learning_rate": 3.070707073646959e-05, "loss": 0.014, "step": 64480 }, { "grad_norm": 0.22868315875530243, "learning_rate": 3.069181757458633e-05, "loss": 0.0125, "step": 64490 }, { "grad_norm": 0.14891168475151062, "learning_rate": 3.067656652421687e-05, "loss": 0.0122, "step": 64500 }, { "grad_norm": 0.2158103883266449, "learning_rate": 3.0661317587029025e-05, "loss": 0.0149, "step": 64510 }, { "grad_norm": 0.175658717751503, "learning_rate": 3.064607076469043e-05, "loss": 0.0141, "step": 64520 }, { "grad_norm": 0.12097404897212982, "learning_rate": 3.063082605886845e-05, "loss": 0.0137, "step": 64530 }, { "grad_norm": 0.16936741769313812, "learning_rate": 3.061558347123019e-05, "loss": 0.0128, "step": 64540 }, { "grad_norm": 0.14387990534305573, "learning_rate": 3.060034300344259e-05, "loss": 0.013, "step": 64550 }, { "grad_norm": 0.17282356321811676, "learning_rate": 3.058510465717231e-05, "loss": 0.0146, "step": 64560 }, { "grad_norm": 0.21836169064044952, "learning_rate": 3.0569868434085804e-05, "loss": 0.0118, "step": 64570 }, { "grad_norm": 0.15916699171066284, "learning_rate": 3.055463433584924e-05, "loss": 0.0168, "step": 64580 }, { "grad_norm": 0.13130006194114685, "learning_rate": 3.0539402364128656e-05, "loss": 0.0116, "step": 64590 }, { "grad_norm": 0.15141606330871582, "learning_rate": 3.052417252058977e-05, "loss": 0.014, "step": 64600 }, { "grad_norm": 0.10135152190923691, "learning_rate": 3.0508944806898092e-05, "loss": 0.0115, "step": 64610 }, { "grad_norm": 0.16124603152275085, "learning_rate": 3.04937192247189e-05, "loss": 0.0128, "step": 64620 }, { "grad_norm": 0.13627232611179352, "learning_rate": 3.047849577571725e-05, "loss": 0.0102, "step": 64630 }, { "grad_norm": 0.34745320677757263, "learning_rate": 3.046327446155795e-05, "loss": 0.0123, "step": 64640 }, { "grad_norm": 0.21172387897968292, "learning_rate": 3.0448055283905587e-05, "loss": 0.0118, "step": 64650 }, { "grad_norm": 0.18460378050804138, "learning_rate": 3.0432838244424494e-05, "loss": 0.0133, "step": 64660 }, { "grad_norm": 0.19933868944644928, "learning_rate": 3.04176233447788e-05, "loss": 0.013, "step": 64670 }, { "grad_norm": 0.20466719567775726, "learning_rate": 3.040241058663238e-05, "loss": 0.0134, "step": 64680 }, { "grad_norm": 0.1735433042049408, "learning_rate": 3.038719997164887e-05, "loss": 0.0154, "step": 64690 }, { "grad_norm": 0.20982322096824646, "learning_rate": 3.0371991501491683e-05, "loss": 0.0127, "step": 64700 }, { "grad_norm": 0.14921154081821442, "learning_rate": 3.0356785177823975e-05, "loss": 0.014, "step": 64710 }, { "grad_norm": 0.18056131899356842, "learning_rate": 3.0341581002308726e-05, "loss": 0.0138, "step": 64720 }, { "grad_norm": 0.19800686836242676, "learning_rate": 3.0326378976608627e-05, "loss": 0.0141, "step": 64730 }, { "grad_norm": 0.18250638246536255, "learning_rate": 3.0311179102386146e-05, "loss": 0.0128, "step": 64740 }, { "grad_norm": 0.16848208010196686, "learning_rate": 3.0295981381303513e-05, "loss": 0.0141, "step": 64750 }, { "grad_norm": 0.20279517769813538, "learning_rate": 3.0280785815022727e-05, "loss": 0.0119, "step": 64760 }, { "grad_norm": 0.18492142856121063, "learning_rate": 3.0265592405205544e-05, "loss": 0.0148, "step": 64770 }, { "grad_norm": 0.22090411186218262, "learning_rate": 3.0250401153513517e-05, "loss": 0.0168, "step": 64780 }, { "grad_norm": 0.1966417133808136, "learning_rate": 3.0235212061607936e-05, "loss": 0.0136, "step": 64790 }, { "grad_norm": 0.10426873713731766, "learning_rate": 3.0220025131149842e-05, "loss": 0.0136, "step": 64800 }, { "grad_norm": 0.17965491116046906, "learning_rate": 3.020484036380007e-05, "loss": 0.0189, "step": 64810 }, { "grad_norm": 0.21163243055343628, "learning_rate": 3.0189657761219188e-05, "loss": 0.0136, "step": 64820 }, { "grad_norm": 0.1819012463092804, "learning_rate": 3.017447732506753e-05, "loss": 0.0128, "step": 64830 }, { "grad_norm": 0.15905436873435974, "learning_rate": 3.015929905700524e-05, "loss": 0.0141, "step": 64840 }, { "grad_norm": 0.11719723045825958, "learning_rate": 3.014412295869218e-05, "loss": 0.013, "step": 64850 }, { "grad_norm": 0.21024051308631897, "learning_rate": 3.0128949031787977e-05, "loss": 0.0126, "step": 64860 }, { "grad_norm": 0.21082644164562225, "learning_rate": 3.0113777277952022e-05, "loss": 0.0148, "step": 64870 }, { "grad_norm": 0.12075924128293991, "learning_rate": 3.0098607698843496e-05, "loss": 0.0111, "step": 64880 }, { "grad_norm": 0.13957664370536804, "learning_rate": 3.0083440296121312e-05, "loss": 0.0157, "step": 64890 }, { "grad_norm": 0.18209506571292877, "learning_rate": 3.0068275071444142e-05, "loss": 0.0134, "step": 64900 }, { "grad_norm": 0.17350554466247559, "learning_rate": 3.0053112026470442e-05, "loss": 0.0155, "step": 64910 }, { "grad_norm": 0.11684007942676544, "learning_rate": 3.0037951162858434e-05, "loss": 0.0125, "step": 64920 }, { "grad_norm": 0.14149366319179535, "learning_rate": 3.002279248226607e-05, "loss": 0.014, "step": 64930 }, { "grad_norm": 0.19498838484287262, "learning_rate": 3.000763598635109e-05, "loss": 0.0152, "step": 64940 }, { "grad_norm": 0.24296307563781738, "learning_rate": 2.999248167677098e-05, "loss": 0.0146, "step": 64950 }, { "grad_norm": 0.15279339253902435, "learning_rate": 2.9977329555182976e-05, "loss": 0.0135, "step": 64960 }, { "grad_norm": 0.19547539949417114, "learning_rate": 2.9962179623244124e-05, "loss": 0.0136, "step": 64970 }, { "grad_norm": 0.16222186386585236, "learning_rate": 2.9947031882611175e-05, "loss": 0.013, "step": 64980 }, { "grad_norm": 0.13241714239120483, "learning_rate": 2.9931886334940685e-05, "loss": 0.0139, "step": 64990 }, { "grad_norm": 0.14791148900985718, "learning_rate": 2.9916742981888923e-05, "loss": 0.0116, "step": 65000 }, { "grad_norm": 0.23771147429943085, "learning_rate": 2.990160182511196e-05, "loss": 0.014, "step": 65010 }, { "grad_norm": 0.12848132848739624, "learning_rate": 2.9886462866265574e-05, "loss": 0.0123, "step": 65020 }, { "grad_norm": 0.12080176919698715, "learning_rate": 2.9871326107005392e-05, "loss": 0.0122, "step": 65030 }, { "grad_norm": 0.19773951172828674, "learning_rate": 2.9856191548986735e-05, "loss": 0.014, "step": 65040 }, { "grad_norm": 0.10001755505800247, "learning_rate": 2.9841059193864672e-05, "loss": 0.0128, "step": 65050 }, { "grad_norm": 0.14085769653320312, "learning_rate": 2.982592904329407e-05, "loss": 0.0128, "step": 65060 }, { "grad_norm": 0.31160658597946167, "learning_rate": 2.981080109892954e-05, "loss": 0.0217, "step": 65070 }, { "grad_norm": 0.20518195629119873, "learning_rate": 2.9795675362425434e-05, "loss": 0.0154, "step": 65080 }, { "grad_norm": 0.19344982504844666, "learning_rate": 2.97805518354359e-05, "loss": 0.0137, "step": 65090 }, { "grad_norm": 0.23791176080703735, "learning_rate": 2.9765430519614825e-05, "loss": 0.0119, "step": 65100 }, { "grad_norm": 0.19580507278442383, "learning_rate": 2.975031141661584e-05, "loss": 0.0132, "step": 65110 }, { "grad_norm": 0.1473396271467209, "learning_rate": 2.9735194528092347e-05, "loss": 0.011, "step": 65120 }, { "grad_norm": 0.15290412306785583, "learning_rate": 2.9720079855697525e-05, "loss": 0.0121, "step": 65130 }, { "grad_norm": 0.17272613942623138, "learning_rate": 2.970496740108427e-05, "loss": 0.0132, "step": 65140 }, { "grad_norm": 0.17009122669696808, "learning_rate": 2.968985716590527e-05, "loss": 0.0137, "step": 65150 }, { "grad_norm": 0.23406146466732025, "learning_rate": 2.9674749151812953e-05, "loss": 0.0182, "step": 65160 }, { "grad_norm": 0.21657833456993103, "learning_rate": 2.965964336045951e-05, "loss": 0.0127, "step": 65170 }, { "grad_norm": 0.14759719371795654, "learning_rate": 2.9644539793496894e-05, "loss": 0.017, "step": 65180 }, { "grad_norm": 0.16103729605674744, "learning_rate": 2.9629438452576807e-05, "loss": 0.0156, "step": 65190 }, { "grad_norm": 0.15689034759998322, "learning_rate": 2.9614339339350698e-05, "loss": 0.0155, "step": 65200 }, { "grad_norm": 0.17027969658374786, "learning_rate": 2.959924245546979e-05, "loss": 0.0111, "step": 65210 }, { "grad_norm": 0.1488676369190216, "learning_rate": 2.9584147802585038e-05, "loss": 0.0125, "step": 65220 }, { "grad_norm": 0.14024126529693604, "learning_rate": 2.956905538234721e-05, "loss": 0.0115, "step": 65230 }, { "grad_norm": 0.17264842987060547, "learning_rate": 2.9553965196406764e-05, "loss": 0.0099, "step": 65240 }, { "grad_norm": 0.1479291170835495, "learning_rate": 2.9538877246413943e-05, "loss": 0.0132, "step": 65250 }, { "grad_norm": 0.12993930280208588, "learning_rate": 2.952379153401874e-05, "loss": 0.0143, "step": 65260 }, { "grad_norm": 0.22909167408943176, "learning_rate": 2.9508708060870914e-05, "loss": 0.013, "step": 65270 }, { "grad_norm": 0.12109692394733429, "learning_rate": 2.9493626828619948e-05, "loss": 0.0136, "step": 65280 }, { "grad_norm": 0.13924334943294525, "learning_rate": 2.9478547838915137e-05, "loss": 0.0115, "step": 65290 }, { "grad_norm": 0.24616269767284393, "learning_rate": 2.946347109340548e-05, "loss": 0.0125, "step": 65300 }, { "grad_norm": 0.13575218617916107, "learning_rate": 2.944839659373976e-05, "loss": 0.0115, "step": 65310 }, { "grad_norm": 0.17942677438259125, "learning_rate": 2.943332434156647e-05, "loss": 0.0144, "step": 65320 }, { "grad_norm": 0.19504167139530182, "learning_rate": 2.9418254338533912e-05, "loss": 0.0126, "step": 65330 }, { "grad_norm": 0.1246054545044899, "learning_rate": 2.940318658629011e-05, "loss": 0.0117, "step": 65340 }, { "grad_norm": 0.20855282247066498, "learning_rate": 2.938812108648287e-05, "loss": 0.0178, "step": 65350 }, { "grad_norm": 0.1463296264410019, "learning_rate": 2.937305784075972e-05, "loss": 0.0128, "step": 65360 }, { "grad_norm": 0.17007091641426086, "learning_rate": 2.9357996850767937e-05, "loss": 0.0159, "step": 65370 }, { "grad_norm": 0.2156725376844406, "learning_rate": 2.9342938118154595e-05, "loss": 0.0125, "step": 65380 }, { "grad_norm": 0.1399611085653305, "learning_rate": 2.9327881644566485e-05, "loss": 0.0127, "step": 65390 }, { "grad_norm": 0.2233811467885971, "learning_rate": 2.9312827431650157e-05, "loss": 0.0159, "step": 65400 }, { "grad_norm": 0.1266620010137558, "learning_rate": 2.92977754810519e-05, "loss": 0.0126, "step": 65410 }, { "grad_norm": 0.10648105293512344, "learning_rate": 2.928272579441781e-05, "loss": 0.0116, "step": 65420 }, { "grad_norm": 0.12737731635570526, "learning_rate": 2.9267678373393676e-05, "loss": 0.0123, "step": 65430 }, { "grad_norm": 0.1620582491159439, "learning_rate": 2.925263321962507e-05, "loss": 0.0129, "step": 65440 }, { "grad_norm": 0.20640306174755096, "learning_rate": 2.9237590334757292e-05, "loss": 0.0132, "step": 65450 }, { "grad_norm": 0.17836152017116547, "learning_rate": 2.922254972043543e-05, "loss": 0.0134, "step": 65460 }, { "grad_norm": 0.11981701850891113, "learning_rate": 2.9207511378304265e-05, "loss": 0.0113, "step": 65470 }, { "grad_norm": 0.1296423375606537, "learning_rate": 2.9192475310008428e-05, "loss": 0.0123, "step": 65480 }, { "grad_norm": 0.16811040043830872, "learning_rate": 2.917744151719219e-05, "loss": 0.0138, "step": 65490 }, { "grad_norm": 0.19558265805244446, "learning_rate": 2.916241000149964e-05, "loss": 0.0125, "step": 65500 }, { "grad_norm": 0.1891011744737625, "learning_rate": 2.9147380764574622e-05, "loss": 0.0123, "step": 65510 }, { "grad_norm": 0.12036855518817902, "learning_rate": 2.9132353808060674e-05, "loss": 0.014, "step": 65520 }, { "grad_norm": 0.1279619187116623, "learning_rate": 2.9117329133601134e-05, "loss": 0.0134, "step": 65530 }, { "grad_norm": 0.12658144533634186, "learning_rate": 2.9102306742839093e-05, "loss": 0.0126, "step": 65540 }, { "grad_norm": 0.1653670072555542, "learning_rate": 2.9087286637417375e-05, "loss": 0.0157, "step": 65550 }, { "grad_norm": 0.2124529629945755, "learning_rate": 2.9072268818978537e-05, "loss": 0.0133, "step": 65560 }, { "grad_norm": 0.20148979127407074, "learning_rate": 2.905725328916493e-05, "loss": 0.0152, "step": 65570 }, { "grad_norm": 0.11138249188661575, "learning_rate": 2.9042240049618608e-05, "loss": 0.0138, "step": 65580 }, { "grad_norm": 0.2866821885108948, "learning_rate": 2.9027229101981413e-05, "loss": 0.0146, "step": 65590 }, { "grad_norm": 0.17868249118328094, "learning_rate": 2.901222044789488e-05, "loss": 0.0131, "step": 65600 }, { "grad_norm": 0.1895090639591217, "learning_rate": 2.89972140890004e-05, "loss": 0.012, "step": 65610 }, { "grad_norm": 0.13859784603118896, "learning_rate": 2.8982210026938988e-05, "loss": 0.0124, "step": 65620 }, { "grad_norm": 0.12817388772964478, "learning_rate": 2.896720826335151e-05, "loss": 0.014, "step": 65630 }, { "grad_norm": 0.1788530945777893, "learning_rate": 2.8952208799878493e-05, "loss": 0.0122, "step": 65640 }, { "grad_norm": 0.1773160845041275, "learning_rate": 2.89372116381603e-05, "loss": 0.0141, "step": 65650 }, { "grad_norm": 0.17519043385982513, "learning_rate": 2.8922216779836932e-05, "loss": 0.014, "step": 65660 }, { "grad_norm": 0.17753642797470093, "learning_rate": 2.8907224226548285e-05, "loss": 0.0143, "step": 65670 }, { "grad_norm": 0.17257340252399445, "learning_rate": 2.8892233979933868e-05, "loss": 0.0108, "step": 65680 }, { "grad_norm": 0.23846203088760376, "learning_rate": 2.8877246041633026e-05, "loss": 0.0129, "step": 65690 }, { "grad_norm": 0.16867266595363617, "learning_rate": 2.886226041328478e-05, "loss": 0.0127, "step": 65700 }, { "grad_norm": 0.1641126126050949, "learning_rate": 2.8847277096527957e-05, "loss": 0.0125, "step": 65710 }, { "grad_norm": 0.20069003105163574, "learning_rate": 2.8832296093001105e-05, "loss": 0.0115, "step": 65720 }, { "grad_norm": 0.211360901594162, "learning_rate": 2.8817317404342547e-05, "loss": 0.0133, "step": 65730 }, { "grad_norm": 0.09617524594068527, "learning_rate": 2.880234103219029e-05, "loss": 0.012, "step": 65740 }, { "grad_norm": 0.1801178753376007, "learning_rate": 2.878736697818215e-05, "loss": 0.0155, "step": 65750 }, { "grad_norm": 0.16132186353206635, "learning_rate": 2.877239524395568e-05, "loss": 0.0148, "step": 65760 }, { "grad_norm": 0.15239183604717255, "learning_rate": 2.8757425831148134e-05, "loss": 0.0144, "step": 65770 }, { "grad_norm": 0.15357306599617004, "learning_rate": 2.874245874139658e-05, "loss": 0.0104, "step": 65780 }, { "grad_norm": 0.15634681284427643, "learning_rate": 2.8727493976337737e-05, "loss": 0.0112, "step": 65790 }, { "grad_norm": 0.1864989846944809, "learning_rate": 2.8712531537608218e-05, "loss": 0.0123, "step": 65800 }, { "grad_norm": 0.13199156522750854, "learning_rate": 2.8697571426844217e-05, "loss": 0.0095, "step": 65810 }, { "grad_norm": 0.13302570581436157, "learning_rate": 2.86826136456818e-05, "loss": 0.0132, "step": 65820 }, { "grad_norm": 0.13548217713832855, "learning_rate": 2.8667658195756685e-05, "loss": 0.0137, "step": 65830 }, { "grad_norm": 0.11632304638624191, "learning_rate": 2.8652705078704412e-05, "loss": 0.0133, "step": 65840 }, { "grad_norm": 0.17647835612297058, "learning_rate": 2.863775429616018e-05, "loss": 0.0129, "step": 65850 }, { "grad_norm": 0.1455584317445755, "learning_rate": 2.862280584975906e-05, "loss": 0.0135, "step": 65860 }, { "grad_norm": 0.20959866046905518, "learning_rate": 2.8607859741135735e-05, "loss": 0.0133, "step": 65870 }, { "grad_norm": 0.29033392667770386, "learning_rate": 2.8592915971924723e-05, "loss": 0.0167, "step": 65880 }, { "grad_norm": 0.15609675645828247, "learning_rate": 2.8577974543760212e-05, "loss": 0.0135, "step": 65890 }, { "grad_norm": 0.2252196967601776, "learning_rate": 2.8563035458276222e-05, "loss": 0.014, "step": 65900 }, { "grad_norm": 0.18415185809135437, "learning_rate": 2.8548098717106424e-05, "loss": 0.012, "step": 65910 }, { "grad_norm": 0.14018796384334564, "learning_rate": 2.85331643218843e-05, "loss": 0.0095, "step": 65920 }, { "grad_norm": 0.18779024481773376, "learning_rate": 2.8518232274243046e-05, "loss": 0.0105, "step": 65930 }, { "grad_norm": 0.15579552948474884, "learning_rate": 2.8503302575815638e-05, "loss": 0.013, "step": 65940 }, { "grad_norm": 0.16079862415790558, "learning_rate": 2.8488375228234716e-05, "loss": 0.016, "step": 65950 }, { "grad_norm": 0.27688324451446533, "learning_rate": 2.8473450233132738e-05, "loss": 0.0121, "step": 65960 }, { "grad_norm": 0.13832028210163116, "learning_rate": 2.8458527592141898e-05, "loss": 0.0165, "step": 65970 }, { "grad_norm": 0.1636645495891571, "learning_rate": 2.8443607306894073e-05, "loss": 0.0116, "step": 65980 }, { "grad_norm": 0.15312913060188293, "learning_rate": 2.8428689379020946e-05, "loss": 0.014, "step": 65990 }, { "grad_norm": 0.10607674717903137, "learning_rate": 2.841377381015391e-05, "loss": 0.0128, "step": 66000 }, { "grad_norm": 0.18971043825149536, "learning_rate": 2.8398860601924144e-05, "loss": 0.0121, "step": 66010 }, { "grad_norm": 0.15631359815597534, "learning_rate": 2.838394975596248e-05, "loss": 0.0126, "step": 66020 }, { "grad_norm": 0.18555061519145966, "learning_rate": 2.83690412738996e-05, "loss": 0.0115, "step": 66030 }, { "grad_norm": 0.2055780589580536, "learning_rate": 2.835413515736579e-05, "loss": 0.0133, "step": 66040 }, { "grad_norm": 0.20853959023952484, "learning_rate": 2.8339231407991267e-05, "loss": 0.0158, "step": 66050 }, { "grad_norm": 0.15320104360580444, "learning_rate": 2.832433002740581e-05, "loss": 0.0122, "step": 66060 }, { "grad_norm": 0.17242197692394257, "learning_rate": 2.8309431017239047e-05, "loss": 0.0146, "step": 66070 }, { "grad_norm": 0.13004957139492035, "learning_rate": 2.829453437912028e-05, "loss": 0.0143, "step": 66080 }, { "grad_norm": 0.19149187207221985, "learning_rate": 2.8279640114678617e-05, "loss": 0.0163, "step": 66090 }, { "grad_norm": 0.20988567173480988, "learning_rate": 2.8264748225542836e-05, "loss": 0.0128, "step": 66100 }, { "grad_norm": 0.16928906738758087, "learning_rate": 2.8249858713341516e-05, "loss": 0.0142, "step": 66110 }, { "grad_norm": 0.18103636801242828, "learning_rate": 2.823497157970294e-05, "loss": 0.0122, "step": 66120 }, { "grad_norm": 0.18778300285339355, "learning_rate": 2.8220086826255164e-05, "loss": 0.0152, "step": 66130 }, { "grad_norm": 0.14683319628238678, "learning_rate": 2.8205204454625934e-05, "loss": 0.0105, "step": 66140 }, { "grad_norm": 0.13866889476776123, "learning_rate": 2.8190324466442798e-05, "loss": 0.012, "step": 66150 }, { "grad_norm": 0.19225037097930908, "learning_rate": 2.8175446863332965e-05, "loss": 0.0156, "step": 66160 }, { "grad_norm": 0.13469642400741577, "learning_rate": 2.816057164692345e-05, "loss": 0.0152, "step": 66170 }, { "grad_norm": 0.25656041502952576, "learning_rate": 2.8145698818840992e-05, "loss": 0.0153, "step": 66180 }, { "grad_norm": 0.17214322090148926, "learning_rate": 2.8130828380712065e-05, "loss": 0.0155, "step": 66190 }, { "grad_norm": 0.1570059359073639, "learning_rate": 2.811596033416285e-05, "loss": 0.0142, "step": 66200 }, { "grad_norm": 0.18532633781433105, "learning_rate": 2.8101094680819307e-05, "loss": 0.0149, "step": 66210 }, { "grad_norm": 0.30134493112564087, "learning_rate": 2.808623142230715e-05, "loss": 0.0119, "step": 66220 }, { "grad_norm": 0.18534179031848907, "learning_rate": 2.8071370560251757e-05, "loss": 0.0111, "step": 66230 }, { "grad_norm": 0.15651637315750122, "learning_rate": 2.805651209627831e-05, "loss": 0.0122, "step": 66240 }, { "grad_norm": 0.15153610706329346, "learning_rate": 2.804165603201171e-05, "loss": 0.0128, "step": 66250 }, { "grad_norm": 0.12069468200206757, "learning_rate": 2.802680236907661e-05, "loss": 0.0103, "step": 66260 }, { "grad_norm": 0.18872737884521484, "learning_rate": 2.8011951109097344e-05, "loss": 0.0126, "step": 66270 }, { "grad_norm": 0.1773184984922409, "learning_rate": 2.7997102253698072e-05, "loss": 0.0185, "step": 66280 }, { "grad_norm": 0.13756613433361053, "learning_rate": 2.7982255804502588e-05, "loss": 0.0142, "step": 66290 }, { "grad_norm": 0.291047066450119, "learning_rate": 2.7967411763134516e-05, "loss": 0.0141, "step": 66300 }, { "grad_norm": 0.19877515733242035, "learning_rate": 2.7952570131217166e-05, "loss": 0.0128, "step": 66310 }, { "grad_norm": 0.20879337191581726, "learning_rate": 2.7937730910373618e-05, "loss": 0.015, "step": 66320 }, { "grad_norm": 0.16198153793811798, "learning_rate": 2.7922894102226628e-05, "loss": 0.013, "step": 66330 }, { "grad_norm": 0.14693321287631989, "learning_rate": 2.790805970839876e-05, "loss": 0.0158, "step": 66340 }, { "grad_norm": 0.1596190184354782, "learning_rate": 2.789322773051225e-05, "loss": 0.0131, "step": 66350 }, { "grad_norm": 0.13319237530231476, "learning_rate": 2.7878398170189124e-05, "loss": 0.0154, "step": 66360 }, { "grad_norm": 0.17252393066883087, "learning_rate": 2.7863571029051112e-05, "loss": 0.0117, "step": 66370 }, { "grad_norm": 0.1737164556980133, "learning_rate": 2.784874630871971e-05, "loss": 0.0119, "step": 66380 }, { "grad_norm": 0.16859668493270874, "learning_rate": 2.7833924010816086e-05, "loss": 0.0114, "step": 66390 }, { "grad_norm": 0.12013063579797745, "learning_rate": 2.7819104136961226e-05, "loss": 0.0109, "step": 66400 }, { "grad_norm": 0.1529666930437088, "learning_rate": 2.780428668877577e-05, "loss": 0.016, "step": 66410 }, { "grad_norm": 0.16366881132125854, "learning_rate": 2.7789471667880152e-05, "loss": 0.0143, "step": 66420 }, { "grad_norm": 0.23908080160617828, "learning_rate": 2.7774659075894523e-05, "loss": 0.0119, "step": 66430 }, { "grad_norm": 0.20539093017578125, "learning_rate": 2.7759848914438778e-05, "loss": 0.0133, "step": 66440 }, { "grad_norm": 0.16373926401138306, "learning_rate": 2.7745041185132496e-05, "loss": 0.0127, "step": 66450 }, { "grad_norm": 0.16769039630889893, "learning_rate": 2.773023588959506e-05, "loss": 0.0126, "step": 66460 }, { "grad_norm": 0.1621955782175064, "learning_rate": 2.7715433029445555e-05, "loss": 0.0133, "step": 66470 }, { "grad_norm": 0.13764040172100067, "learning_rate": 2.7700632606302766e-05, "loss": 0.0125, "step": 66480 }, { "grad_norm": 0.21109502017498016, "learning_rate": 2.768583462178528e-05, "loss": 0.0133, "step": 66490 }, { "grad_norm": 0.19084884226322174, "learning_rate": 2.767103907751136e-05, "loss": 0.0124, "step": 66500 }, { "grad_norm": 0.15209035575389862, "learning_rate": 2.7656245975099065e-05, "loss": 0.0182, "step": 66510 }, { "grad_norm": 0.1692861169576645, "learning_rate": 2.7641455316166093e-05, "loss": 0.0118, "step": 66520 }, { "grad_norm": 0.15277087688446045, "learning_rate": 2.7626667102329973e-05, "loss": 0.0117, "step": 66530 }, { "grad_norm": 0.1801060438156128, "learning_rate": 2.761188133520788e-05, "loss": 0.0144, "step": 66540 }, { "grad_norm": 0.19495469331741333, "learning_rate": 2.7597098016416784e-05, "loss": 0.0142, "step": 66550 }, { "grad_norm": 0.18777316808700562, "learning_rate": 2.7582317147573368e-05, "loss": 0.0164, "step": 66560 }, { "grad_norm": 0.1896069347858429, "learning_rate": 2.7567538730294064e-05, "loss": 0.0137, "step": 66570 }, { "grad_norm": 0.15470311045646667, "learning_rate": 2.7552762766194973e-05, "loss": 0.0121, "step": 66580 }, { "grad_norm": 0.13955806195735931, "learning_rate": 2.753798925689202e-05, "loss": 0.0125, "step": 66590 }, { "grad_norm": 0.15491902828216553, "learning_rate": 2.7523218204000774e-05, "loss": 0.0122, "step": 66600 }, { "grad_norm": 0.18135765194892883, "learning_rate": 2.7508449609136612e-05, "loss": 0.014, "step": 66610 }, { "grad_norm": 0.14932993054389954, "learning_rate": 2.749368347391454e-05, "loss": 0.013, "step": 66620 }, { "grad_norm": 0.17395727336406708, "learning_rate": 2.7478919799949443e-05, "loss": 0.0123, "step": 66630 }, { "grad_norm": 0.15198135375976562, "learning_rate": 2.7464158588855803e-05, "loss": 0.01, "step": 66640 }, { "grad_norm": 0.15393149852752686, "learning_rate": 2.7449399842247918e-05, "loss": 0.0162, "step": 66650 }, { "grad_norm": 0.2260366678237915, "learning_rate": 2.7434643561739736e-05, "loss": 0.013, "step": 66660 }, { "grad_norm": 0.15031176805496216, "learning_rate": 2.7419889748945016e-05, "loss": 0.0135, "step": 66670 }, { "grad_norm": 0.14778508245944977, "learning_rate": 2.7405138405477203e-05, "loss": 0.01, "step": 66680 }, { "grad_norm": 0.17289145290851593, "learning_rate": 2.73903895329495e-05, "loss": 0.0112, "step": 66690 }, { "grad_norm": 0.15074126422405243, "learning_rate": 2.737564313297478e-05, "loss": 0.0148, "step": 66700 }, { "grad_norm": 0.1987941563129425, "learning_rate": 2.736089920716572e-05, "loss": 0.0128, "step": 66710 }, { "grad_norm": 0.18723055720329285, "learning_rate": 2.7346157757134706e-05, "loss": 0.0134, "step": 66720 }, { "grad_norm": 0.13536065816879272, "learning_rate": 2.733141878449379e-05, "loss": 0.0131, "step": 66730 }, { "grad_norm": 0.18343301117420197, "learning_rate": 2.7316682290854844e-05, "loss": 0.0128, "step": 66740 }, { "grad_norm": 0.15605945885181427, "learning_rate": 2.7301948277829414e-05, "loss": 0.0143, "step": 66750 }, { "grad_norm": 0.1507323831319809, "learning_rate": 2.7287216747028814e-05, "loss": 0.0155, "step": 66760 }, { "grad_norm": 0.13046099245548248, "learning_rate": 2.7272487700064025e-05, "loss": 0.0156, "step": 66770 }, { "grad_norm": 0.12489333003759384, "learning_rate": 2.725776113854583e-05, "loss": 0.0125, "step": 66780 }, { "grad_norm": 0.1517932116985321, "learning_rate": 2.7243037064084664e-05, "loss": 0.0117, "step": 66790 }, { "grad_norm": 0.222652405500412, "learning_rate": 2.7228315478290766e-05, "loss": 0.0138, "step": 66800 }, { "grad_norm": 0.18615677952766418, "learning_rate": 2.721359638277402e-05, "loss": 0.0142, "step": 66810 }, { "grad_norm": 0.19061174988746643, "learning_rate": 2.7198879779144148e-05, "loss": 0.0124, "step": 66820 }, { "grad_norm": 0.1686784029006958, "learning_rate": 2.7184165669010485e-05, "loss": 0.0109, "step": 66830 }, { "grad_norm": 0.1982114464044571, "learning_rate": 2.716945405398218e-05, "loss": 0.0131, "step": 66840 }, { "grad_norm": 0.49918314814567566, "learning_rate": 2.715474493566803e-05, "loss": 0.0144, "step": 66850 }, { "grad_norm": 0.15207596123218536, "learning_rate": 2.7140038315676652e-05, "loss": 0.0136, "step": 66860 }, { "grad_norm": 0.10187486559152603, "learning_rate": 2.7125334195616275e-05, "loss": 0.0157, "step": 66870 }, { "grad_norm": 0.138717919588089, "learning_rate": 2.7110632577094997e-05, "loss": 0.0132, "step": 66880 }, { "grad_norm": 0.154129758477211, "learning_rate": 2.709593346172051e-05, "loss": 0.0122, "step": 66890 }, { "grad_norm": 0.14726044237613678, "learning_rate": 2.708123685110032e-05, "loss": 0.0137, "step": 66900 }, { "grad_norm": 0.14708073437213898, "learning_rate": 2.7066542746841588e-05, "loss": 0.0116, "step": 66910 }, { "grad_norm": 0.1418059915304184, "learning_rate": 2.7051851150551267e-05, "loss": 0.0118, "step": 66920 }, { "grad_norm": 0.16739670932292938, "learning_rate": 2.7037162063835996e-05, "loss": 0.015, "step": 66930 }, { "grad_norm": 0.2049936056137085, "learning_rate": 2.702247548830218e-05, "loss": 0.0131, "step": 66940 }, { "grad_norm": 0.11637501418590546, "learning_rate": 2.7007791425555883e-05, "loss": 0.017, "step": 66950 }, { "grad_norm": 0.16760356724262238, "learning_rate": 2.6993109877202945e-05, "loss": 0.0135, "step": 66960 }, { "grad_norm": 0.12935014069080353, "learning_rate": 2.6978430844848935e-05, "loss": 0.0113, "step": 66970 }, { "grad_norm": 0.16227881610393524, "learning_rate": 2.69637543300991e-05, "loss": 0.0124, "step": 66980 }, { "grad_norm": 0.19211435317993164, "learning_rate": 2.694908033455848e-05, "loss": 0.0116, "step": 66990 }, { "grad_norm": 0.15293057262897491, "learning_rate": 2.6934408859831733e-05, "loss": 0.0119, "step": 67000 }, { "grad_norm": 0.1360243558883667, "learning_rate": 2.69197399075234e-05, "loss": 0.015, "step": 67010 }, { "grad_norm": 0.15401935577392578, "learning_rate": 2.6905073479237584e-05, "loss": 0.0157, "step": 67020 }, { "grad_norm": 0.12108343094587326, "learning_rate": 2.6890409576578235e-05, "loss": 0.0121, "step": 67030 }, { "grad_norm": 0.16841472685337067, "learning_rate": 2.687574820114892e-05, "loss": 0.0117, "step": 67040 }, { "grad_norm": 0.16668280959129333, "learning_rate": 2.686108935455305e-05, "loss": 0.0123, "step": 67050 }, { "grad_norm": 0.17922723293304443, "learning_rate": 2.6846433038393603e-05, "loss": 0.0139, "step": 67060 }, { "grad_norm": 0.11295215785503387, "learning_rate": 2.6831779254273482e-05, "loss": 0.011, "step": 67070 }, { "grad_norm": 0.15238387882709503, "learning_rate": 2.6817128003795122e-05, "loss": 0.0132, "step": 67080 }, { "grad_norm": 0.13823826611042023, "learning_rate": 2.680247928856081e-05, "loss": 0.0138, "step": 67090 }, { "grad_norm": 0.20410238206386566, "learning_rate": 2.6787833110172466e-05, "loss": 0.0132, "step": 67100 }, { "grad_norm": 0.18677771091461182, "learning_rate": 2.6773189470231807e-05, "loss": 0.0112, "step": 67110 }, { "grad_norm": 0.15037696063518524, "learning_rate": 2.675854837034022e-05, "loss": 0.0116, "step": 67120 }, { "grad_norm": 0.1734808087348938, "learning_rate": 2.6743909812098833e-05, "loss": 0.0145, "step": 67130 }, { "grad_norm": 0.19986452162265778, "learning_rate": 2.67292737971085e-05, "loss": 0.0138, "step": 67140 }, { "grad_norm": 0.16138224303722382, "learning_rate": 2.671464032696982e-05, "loss": 0.0128, "step": 67150 }, { "grad_norm": 0.1684214472770691, "learning_rate": 2.6700009403283055e-05, "loss": 0.0131, "step": 67160 }, { "grad_norm": 0.1584351509809494, "learning_rate": 2.6685381027648214e-05, "loss": 0.0126, "step": 67170 }, { "grad_norm": 0.1548522412776947, "learning_rate": 2.6670755201665086e-05, "loss": 0.012, "step": 67180 }, { "grad_norm": 0.14839287102222443, "learning_rate": 2.6656131926933066e-05, "loss": 0.0137, "step": 67190 }, { "grad_norm": 0.16047403216362, "learning_rate": 2.664151120505136e-05, "loss": 0.0128, "step": 67200 }, { "grad_norm": 0.1136007308959961, "learning_rate": 2.6626893037618873e-05, "loss": 0.0117, "step": 67210 }, { "grad_norm": 0.2043667584657669, "learning_rate": 2.6612277426234243e-05, "loss": 0.0178, "step": 67220 }, { "grad_norm": 0.1308017075061798, "learning_rate": 2.659766437249577e-05, "loss": 0.0124, "step": 67230 }, { "grad_norm": 0.1346157342195511, "learning_rate": 2.6583053878001557e-05, "loss": 0.0106, "step": 67240 }, { "grad_norm": 0.13383206725120544, "learning_rate": 2.6568445944349323e-05, "loss": 0.0143, "step": 67250 }, { "grad_norm": 0.15831977128982544, "learning_rate": 2.6553840573136657e-05, "loss": 0.0116, "step": 67260 }, { "grad_norm": 0.15435966849327087, "learning_rate": 2.6539237765960706e-05, "loss": 0.0126, "step": 67270 }, { "grad_norm": 0.12041158229112625, "learning_rate": 2.6524637524418465e-05, "loss": 0.01, "step": 67280 }, { "grad_norm": 0.1435440480709076, "learning_rate": 2.651003985010655e-05, "loss": 0.0127, "step": 67290 }, { "grad_norm": 0.18438251316547394, "learning_rate": 2.649544474462138e-05, "loss": 0.0131, "step": 67300 }, { "grad_norm": 0.15956220030784607, "learning_rate": 2.6480852209559015e-05, "loss": 0.0094, "step": 67310 }, { "grad_norm": 0.11939752101898193, "learning_rate": 2.6466262246515296e-05, "loss": 0.014, "step": 67320 }, { "grad_norm": 0.14315542578697205, "learning_rate": 2.6451674857085746e-05, "loss": 0.011, "step": 67330 }, { "grad_norm": 0.16213905811309814, "learning_rate": 2.6437090042865655e-05, "loss": 0.0132, "step": 67340 }, { "grad_norm": 0.15144425630569458, "learning_rate": 2.6422507805449947e-05, "loss": 0.013, "step": 67350 }, { "grad_norm": 0.1452675610780716, "learning_rate": 2.6407928146433357e-05, "loss": 0.0118, "step": 67360 }, { "grad_norm": 0.13297176361083984, "learning_rate": 2.639335106741026e-05, "loss": 0.016, "step": 67370 }, { "grad_norm": 0.14164894819259644, "learning_rate": 2.6378776569974796e-05, "loss": 0.0123, "step": 67380 }, { "grad_norm": 0.1825004518032074, "learning_rate": 2.6364204655720814e-05, "loss": 0.0131, "step": 67390 }, { "grad_norm": 0.09922194480895996, "learning_rate": 2.6349635326241894e-05, "loss": 0.0097, "step": 67400 }, { "grad_norm": 0.17677785456180573, "learning_rate": 2.633506858313127e-05, "loss": 0.0129, "step": 67410 }, { "grad_norm": 0.15802888572216034, "learning_rate": 2.6320504427981975e-05, "loss": 0.0128, "step": 67420 }, { "grad_norm": 0.1744532585144043, "learning_rate": 2.6305942862386734e-05, "loss": 0.0098, "step": 67430 }, { "grad_norm": 0.16272187232971191, "learning_rate": 2.629138388793793e-05, "loss": 0.0146, "step": 67440 }, { "grad_norm": 0.1664898693561554, "learning_rate": 2.6276827506227743e-05, "loss": 0.011, "step": 67450 }, { "grad_norm": 0.15206937491893768, "learning_rate": 2.6262273718848037e-05, "loss": 0.0119, "step": 67460 }, { "grad_norm": 0.22331836819648743, "learning_rate": 2.6247722527390407e-05, "loss": 0.0103, "step": 67470 }, { "grad_norm": 0.15546691417694092, "learning_rate": 2.6233173933446104e-05, "loss": 0.0118, "step": 67480 }, { "grad_norm": 0.16710364818572998, "learning_rate": 2.6218627938606188e-05, "loss": 0.0137, "step": 67490 }, { "grad_norm": 0.16367889940738678, "learning_rate": 2.620408454446135e-05, "loss": 0.0139, "step": 67500 }, { "grad_norm": 0.3204125761985779, "learning_rate": 2.6189543752602046e-05, "loss": 0.0126, "step": 67510 }, { "grad_norm": 0.11020998656749725, "learning_rate": 2.6175005564618444e-05, "loss": 0.0115, "step": 67520 }, { "grad_norm": 0.135324165225029, "learning_rate": 2.6160469982100428e-05, "loss": 0.0116, "step": 67530 }, { "grad_norm": 0.12898306548595428, "learning_rate": 2.6145937006637555e-05, "loss": 0.0133, "step": 67540 }, { "grad_norm": 0.2058311402797699, "learning_rate": 2.6131406639819166e-05, "loss": 0.013, "step": 67550 }, { "grad_norm": 0.12395988404750824, "learning_rate": 2.611687888323424e-05, "loss": 0.0121, "step": 67560 }, { "grad_norm": 0.21704134345054626, "learning_rate": 2.610235373847153e-05, "loss": 0.0134, "step": 67570 }, { "grad_norm": 0.1787814050912857, "learning_rate": 2.6087831207119494e-05, "loss": 0.0129, "step": 67580 }, { "grad_norm": 0.15603621304035187, "learning_rate": 2.60733112907663e-05, "loss": 0.015, "step": 67590 }, { "grad_norm": 0.16024741530418396, "learning_rate": 2.6058793990999786e-05, "loss": 0.0111, "step": 67600 }, { "grad_norm": 0.1626802235841751, "learning_rate": 2.6044279309407593e-05, "loss": 0.0129, "step": 67610 }, { "grad_norm": 0.17153792083263397, "learning_rate": 2.6029767247576976e-05, "loss": 0.0128, "step": 67620 }, { "grad_norm": 0.17527347803115845, "learning_rate": 2.601525780709497e-05, "loss": 0.0114, "step": 67630 }, { "grad_norm": 0.1207023411989212, "learning_rate": 2.6000750989548318e-05, "loss": 0.0105, "step": 67640 }, { "grad_norm": 0.18866699934005737, "learning_rate": 2.5986246796523462e-05, "loss": 0.0128, "step": 67650 }, { "grad_norm": 0.13804146647453308, "learning_rate": 2.5971745229606546e-05, "loss": 0.015, "step": 67660 }, { "grad_norm": 0.11677761375904083, "learning_rate": 2.5957246290383442e-05, "loss": 0.0126, "step": 67670 }, { "grad_norm": 0.1976359486579895, "learning_rate": 2.594274998043975e-05, "loss": 0.0111, "step": 67680 }, { "grad_norm": 0.1310073733329773, "learning_rate": 2.5928256301360733e-05, "loss": 0.0093, "step": 67690 }, { "grad_norm": 0.17972293496131897, "learning_rate": 2.591376525473142e-05, "loss": 0.0142, "step": 67700 }, { "grad_norm": 0.14899985492229462, "learning_rate": 2.5899276842136523e-05, "loss": 0.0116, "step": 67710 }, { "grad_norm": 0.11950451880693436, "learning_rate": 2.5884791065160495e-05, "loss": 0.009, "step": 67720 }, { "grad_norm": 0.14635123312473297, "learning_rate": 2.5870307925387436e-05, "loss": 0.013, "step": 67730 }, { "grad_norm": 0.11922167986631393, "learning_rate": 2.585582742440125e-05, "loss": 0.0101, "step": 67740 }, { "grad_norm": 0.14093266427516937, "learning_rate": 2.5841349563785455e-05, "loss": 0.0152, "step": 67750 }, { "grad_norm": 0.1949823796749115, "learning_rate": 2.5826874345123355e-05, "loss": 0.0125, "step": 67760 }, { "grad_norm": 0.22836069762706757, "learning_rate": 2.581240176999792e-05, "loss": 0.0125, "step": 67770 }, { "grad_norm": 0.13605418801307678, "learning_rate": 2.579793183999189e-05, "loss": 0.0149, "step": 67780 }, { "grad_norm": 0.17647124826908112, "learning_rate": 2.5783464556687627e-05, "loss": 0.0133, "step": 67790 }, { "grad_norm": 0.16595932841300964, "learning_rate": 2.576899992166729e-05, "loss": 0.015, "step": 67800 }, { "grad_norm": 0.251074880361557, "learning_rate": 2.5754537936512668e-05, "loss": 0.0119, "step": 67810 }, { "grad_norm": 0.11055424809455872, "learning_rate": 2.574007860280535e-05, "loss": 0.0104, "step": 67820 }, { "grad_norm": 0.09432309120893478, "learning_rate": 2.5725621922126518e-05, "loss": 0.0108, "step": 67830 }, { "grad_norm": 0.2732483744621277, "learning_rate": 2.571116789605722e-05, "loss": 0.0116, "step": 67840 }, { "grad_norm": 0.2835944890975952, "learning_rate": 2.5696716526178054e-05, "loss": 0.0181, "step": 67850 }, { "grad_norm": 0.1508944034576416, "learning_rate": 2.568226781406946e-05, "loss": 0.0123, "step": 67860 }, { "grad_norm": 0.19740134477615356, "learning_rate": 2.5667821761311474e-05, "loss": 0.012, "step": 67870 }, { "grad_norm": 0.15080049633979797, "learning_rate": 2.5653378369483917e-05, "loss": 0.0122, "step": 67880 }, { "grad_norm": 0.15358471870422363, "learning_rate": 2.5638937640166292e-05, "loss": 0.0103, "step": 67890 }, { "grad_norm": 0.13155117630958557, "learning_rate": 2.5624499574937842e-05, "loss": 0.0114, "step": 67900 }, { "grad_norm": 0.1511889100074768, "learning_rate": 2.5610064175377456e-05, "loss": 0.0165, "step": 67910 }, { "grad_norm": 0.13856549561023712, "learning_rate": 2.5595631443063782e-05, "loss": 0.011, "step": 67920 }, { "grad_norm": 0.15909473598003387, "learning_rate": 2.5581201379575177e-05, "loss": 0.0132, "step": 67930 }, { "grad_norm": 0.15371018648147583, "learning_rate": 2.5566773986489655e-05, "loss": 0.0133, "step": 67940 }, { "grad_norm": 0.17519712448120117, "learning_rate": 2.5552349265385e-05, "loss": 0.0141, "step": 67950 }, { "grad_norm": 0.12704041600227356, "learning_rate": 2.5537927217838675e-05, "loss": 0.0121, "step": 67960 }, { "grad_norm": 0.19345946609973907, "learning_rate": 2.5523507845427864e-05, "loss": 0.0115, "step": 67970 }, { "grad_norm": 0.1304602026939392, "learning_rate": 2.5509091149729413e-05, "loss": 0.013, "step": 67980 }, { "grad_norm": 0.1826995313167572, "learning_rate": 2.5494677132319962e-05, "loss": 0.0133, "step": 67990 }, { "grad_norm": 0.242606520652771, "learning_rate": 2.548026579477575e-05, "loss": 0.0145, "step": 68000 }, { "grad_norm": 0.1325809359550476, "learning_rate": 2.5465857138672834e-05, "loss": 0.0103, "step": 68010 }, { "grad_norm": 0.17752403020858765, "learning_rate": 2.5451451165586848e-05, "loss": 0.0132, "step": 68020 }, { "grad_norm": 0.134785458445549, "learning_rate": 2.543704787709329e-05, "loss": 0.0116, "step": 68030 }, { "grad_norm": 0.24205943942070007, "learning_rate": 2.542264727476723e-05, "loss": 0.0146, "step": 68040 }, { "grad_norm": 0.1842353194952011, "learning_rate": 2.5408249360183535e-05, "loss": 0.0135, "step": 68050 }, { "grad_norm": 0.1462765336036682, "learning_rate": 2.5393854134916694e-05, "loss": 0.0168, "step": 68060 }, { "grad_norm": 0.16372209787368774, "learning_rate": 2.5379461600540993e-05, "loss": 0.0118, "step": 68070 }, { "grad_norm": 0.16010157763957977, "learning_rate": 2.5365071758630317e-05, "loss": 0.0137, "step": 68080 }, { "grad_norm": 0.16549955308437347, "learning_rate": 2.535068461075839e-05, "loss": 0.0126, "step": 68090 }, { "grad_norm": 0.1467660516500473, "learning_rate": 2.5336300158498516e-05, "loss": 0.01, "step": 68100 }, { "grad_norm": 0.18554845452308655, "learning_rate": 2.5321918403423793e-05, "loss": 0.0171, "step": 68110 }, { "grad_norm": 0.21851567924022675, "learning_rate": 2.5307539347106957e-05, "loss": 0.0124, "step": 68120 }, { "grad_norm": 0.15911947190761566, "learning_rate": 2.529316299112049e-05, "loss": 0.0119, "step": 68130 }, { "grad_norm": 0.13908979296684265, "learning_rate": 2.5278789337036567e-05, "loss": 0.0136, "step": 68140 }, { "grad_norm": 0.12772347033023834, "learning_rate": 2.52644183864271e-05, "loss": 0.0119, "step": 68150 }, { "grad_norm": 0.1840488463640213, "learning_rate": 2.525005014086363e-05, "loss": 0.0126, "step": 68160 }, { "grad_norm": 0.139706090092659, "learning_rate": 2.5235684601917465e-05, "loss": 0.0123, "step": 68170 }, { "grad_norm": 0.20433928072452545, "learning_rate": 2.5221321771159618e-05, "loss": 0.0153, "step": 68180 }, { "grad_norm": 0.1944252997636795, "learning_rate": 2.5206961650160743e-05, "loss": 0.0108, "step": 68190 }, { "grad_norm": 0.13388694822788239, "learning_rate": 2.519260424049129e-05, "loss": 0.0121, "step": 68200 }, { "grad_norm": 0.13498173654079437, "learning_rate": 2.5178249543721304e-05, "loss": 0.0114, "step": 68210 }, { "grad_norm": 0.1584925651550293, "learning_rate": 2.5163897561420668e-05, "loss": 0.0147, "step": 68220 }, { "grad_norm": 0.10798283666372299, "learning_rate": 2.514954829515883e-05, "loss": 0.0122, "step": 68230 }, { "grad_norm": 0.17333948612213135, "learning_rate": 2.513520174650505e-05, "loss": 0.014, "step": 68240 }, { "grad_norm": 0.14030428230762482, "learning_rate": 2.5120857917028207e-05, "loss": 0.0142, "step": 68250 }, { "grad_norm": 0.17195384204387665, "learning_rate": 2.5106516808296953e-05, "loss": 0.0121, "step": 68260 }, { "grad_norm": 0.1649731993675232, "learning_rate": 2.5092178421879552e-05, "loss": 0.0125, "step": 68270 }, { "grad_norm": 0.16620519757270813, "learning_rate": 2.5077842759344116e-05, "loss": 0.0114, "step": 68280 }, { "grad_norm": 0.13243964314460754, "learning_rate": 2.5063509822258304e-05, "loss": 0.0131, "step": 68290 }, { "grad_norm": 0.14410783350467682, "learning_rate": 2.5049179612189576e-05, "loss": 0.0127, "step": 68300 }, { "grad_norm": 0.138869047164917, "learning_rate": 2.5034852130705043e-05, "loss": 0.0117, "step": 68310 }, { "grad_norm": 0.1884818822145462, "learning_rate": 2.5020527379371572e-05, "loss": 0.0106, "step": 68320 }, { "grad_norm": 0.23360654711723328, "learning_rate": 2.500620535975564e-05, "loss": 0.0122, "step": 68330 }, { "grad_norm": 0.1317928433418274, "learning_rate": 2.4991886073423515e-05, "loss": 0.0109, "step": 68340 }, { "grad_norm": 0.12518642842769623, "learning_rate": 2.4977569521941134e-05, "loss": 0.0139, "step": 68350 }, { "grad_norm": 0.17198458313941956, "learning_rate": 2.4963255706874143e-05, "loss": 0.0129, "step": 68360 }, { "grad_norm": 0.2116805762052536, "learning_rate": 2.4948944629787853e-05, "loss": 0.0114, "step": 68370 }, { "grad_norm": 0.16431814432144165, "learning_rate": 2.493463629224731e-05, "loss": 0.0118, "step": 68380 }, { "grad_norm": 0.16475562751293182, "learning_rate": 2.492033069581728e-05, "loss": 0.0115, "step": 68390 }, { "grad_norm": 0.19824530184268951, "learning_rate": 2.4906027842062137e-05, "loss": 0.0111, "step": 68400 }, { "grad_norm": 0.1365250200033188, "learning_rate": 2.48917277325461e-05, "loss": 0.0126, "step": 68410 }, { "grad_norm": 0.21357588469982147, "learning_rate": 2.487743036883295e-05, "loss": 0.0102, "step": 68420 }, { "grad_norm": 0.17806105315685272, "learning_rate": 2.486313575248626e-05, "loss": 0.0131, "step": 68430 }, { "grad_norm": 0.14133092761039734, "learning_rate": 2.4848843885069235e-05, "loss": 0.0126, "step": 68440 }, { "grad_norm": 0.1419098973274231, "learning_rate": 2.483455476814484e-05, "loss": 0.0129, "step": 68450 }, { "grad_norm": 0.14572717249393463, "learning_rate": 2.4820268403275672e-05, "loss": 0.0144, "step": 68460 }, { "grad_norm": 0.18457166850566864, "learning_rate": 2.480598479202413e-05, "loss": 0.0126, "step": 68470 }, { "grad_norm": 0.19066159427165985, "learning_rate": 2.4791703935952193e-05, "loss": 0.0145, "step": 68480 }, { "grad_norm": 0.14710819721221924, "learning_rate": 2.477742583662163e-05, "loss": 0.0125, "step": 68490 }, { "grad_norm": 0.13419531285762787, "learning_rate": 2.476315049559384e-05, "loss": 0.0117, "step": 68500 }, { "grad_norm": 0.1405983418226242, "learning_rate": 2.4748877914429987e-05, "loss": 0.013, "step": 68510 }, { "grad_norm": 0.16357894241809845, "learning_rate": 2.4734608094690857e-05, "loss": 0.0153, "step": 68520 }, { "grad_norm": 0.18109732866287231, "learning_rate": 2.4720341037937006e-05, "loss": 0.0124, "step": 68530 }, { "grad_norm": 0.13651742041110992, "learning_rate": 2.4706076745728645e-05, "loss": 0.0142, "step": 68540 }, { "grad_norm": 0.1721772998571396, "learning_rate": 2.469181521962573e-05, "loss": 0.0128, "step": 68550 }, { "grad_norm": 0.15626701712608337, "learning_rate": 2.467755646118783e-05, "loss": 0.0136, "step": 68560 }, { "grad_norm": 0.130808487534523, "learning_rate": 2.46633004719743e-05, "loss": 0.0118, "step": 68570 }, { "grad_norm": 0.21473461389541626, "learning_rate": 2.4649047253544123e-05, "loss": 0.0141, "step": 68580 }, { "grad_norm": 0.2625880241394043, "learning_rate": 2.4634796807456016e-05, "loss": 0.0098, "step": 68590 }, { "grad_norm": 0.18604686856269836, "learning_rate": 2.46205491352684e-05, "loss": 0.0139, "step": 68600 }, { "grad_norm": 0.19160474836826324, "learning_rate": 2.4606304238539375e-05, "loss": 0.0118, "step": 68610 }, { "grad_norm": 0.2287384420633316, "learning_rate": 2.4592062118826753e-05, "loss": 0.0121, "step": 68620 }, { "grad_norm": 0.13242743909358978, "learning_rate": 2.4577822777688004e-05, "loss": 0.013, "step": 68630 }, { "grad_norm": 0.20927438139915466, "learning_rate": 2.4563586216680347e-05, "loss": 0.013, "step": 68640 }, { "grad_norm": 0.136850968003273, "learning_rate": 2.4549352437360622e-05, "loss": 0.0129, "step": 68650 }, { "grad_norm": 0.12169977277517319, "learning_rate": 2.4535121441285493e-05, "loss": 0.012, "step": 68660 }, { "grad_norm": 0.10879532992839813, "learning_rate": 2.4520893230011172e-05, "loss": 0.0106, "step": 68670 }, { "grad_norm": 0.11889323592185974, "learning_rate": 2.450666780509368e-05, "loss": 0.0115, "step": 68680 }, { "grad_norm": 0.16579489409923553, "learning_rate": 2.4492445168088656e-05, "loss": 0.0127, "step": 68690 }, { "grad_norm": 0.1580076962709427, "learning_rate": 2.447822532055149e-05, "loss": 0.0129, "step": 68700 }, { "grad_norm": 0.20291072130203247, "learning_rate": 2.4464008264037212e-05, "loss": 0.0118, "step": 68710 }, { "grad_norm": 0.1369922012090683, "learning_rate": 2.4449794000100605e-05, "loss": 0.0149, "step": 68720 }, { "grad_norm": 0.1292252391576767, "learning_rate": 2.4435582530296108e-05, "loss": 0.0108, "step": 68730 }, { "grad_norm": 0.13364450633525848, "learning_rate": 2.4421373856177887e-05, "loss": 0.0115, "step": 68740 }, { "grad_norm": 0.20158687233924866, "learning_rate": 2.440716797929975e-05, "loss": 0.0109, "step": 68750 }, { "grad_norm": 0.19609585404396057, "learning_rate": 2.439296490121526e-05, "loss": 0.016, "step": 68760 }, { "grad_norm": 0.15430647134780884, "learning_rate": 2.4378764623477608e-05, "loss": 0.0184, "step": 68770 }, { "grad_norm": 0.18434666097164154, "learning_rate": 2.436456714763974e-05, "loss": 0.0143, "step": 68780 }, { "grad_norm": 0.1748039573431015, "learning_rate": 2.4350372475254262e-05, "loss": 0.0114, "step": 68790 }, { "grad_norm": 0.17108774185180664, "learning_rate": 2.433618060787351e-05, "loss": 0.0121, "step": 68800 }, { "grad_norm": 0.1747712343931198, "learning_rate": 2.432199154704944e-05, "loss": 0.0125, "step": 68810 }, { "grad_norm": 0.12427444756031036, "learning_rate": 2.430780529433377e-05, "loss": 0.01, "step": 68820 }, { "grad_norm": 0.11836816370487213, "learning_rate": 2.42936218512779e-05, "loss": 0.0111, "step": 68830 }, { "grad_norm": 0.1668161004781723, "learning_rate": 2.427944121943288e-05, "loss": 0.0119, "step": 68840 }, { "grad_norm": 0.18245528638362885, "learning_rate": 2.4265263400349497e-05, "loss": 0.0122, "step": 68850 }, { "grad_norm": 0.17570331692695618, "learning_rate": 2.4251088395578213e-05, "loss": 0.0114, "step": 68860 }, { "grad_norm": 0.18086785078048706, "learning_rate": 2.4236916206669214e-05, "loss": 0.0109, "step": 68870 }, { "grad_norm": 0.1570325791835785, "learning_rate": 2.4222746835172304e-05, "loss": 0.0099, "step": 68880 }, { "grad_norm": 0.09496982395648956, "learning_rate": 2.4208580282637065e-05, "loss": 0.0137, "step": 68890 }, { "grad_norm": 0.1507023125886917, "learning_rate": 2.4194416550612687e-05, "loss": 0.0116, "step": 68900 }, { "grad_norm": 0.17356380820274353, "learning_rate": 2.418025564064812e-05, "loss": 0.0121, "step": 68910 }, { "grad_norm": 0.16403993964195251, "learning_rate": 2.4166097554291978e-05, "loss": 0.0126, "step": 68920 }, { "grad_norm": 0.1865995228290558, "learning_rate": 2.4151942293092583e-05, "loss": 0.012, "step": 68930 }, { "grad_norm": 0.15562352538108826, "learning_rate": 2.41377898585979e-05, "loss": 0.0097, "step": 68940 }, { "grad_norm": 0.15323442220687866, "learning_rate": 2.4123640252355652e-05, "loss": 0.0124, "step": 68950 }, { "grad_norm": 0.15384076535701752, "learning_rate": 2.410949347591318e-05, "loss": 0.0127, "step": 68960 }, { "grad_norm": 0.14572373032569885, "learning_rate": 2.4095349530817578e-05, "loss": 0.0115, "step": 68970 }, { "grad_norm": 0.11731673032045364, "learning_rate": 2.4081208418615603e-05, "loss": 0.0106, "step": 68980 }, { "grad_norm": 0.2088858187198639, "learning_rate": 2.4067070140853725e-05, "loss": 0.014, "step": 68990 }, { "grad_norm": 0.14925798773765564, "learning_rate": 2.405293469907805e-05, "loss": 0.0096, "step": 69000 }, { "grad_norm": 0.2900471091270447, "learning_rate": 2.4038802094834446e-05, "loss": 0.0132, "step": 69010 }, { "grad_norm": 0.11978679895401001, "learning_rate": 2.4024672329668395e-05, "loss": 0.0135, "step": 69020 }, { "grad_norm": 0.25109636783599854, "learning_rate": 2.4010545405125125e-05, "loss": 0.016, "step": 69030 }, { "grad_norm": 0.16249315440654755, "learning_rate": 2.3996421322749528e-05, "loss": 0.012, "step": 69040 }, { "grad_norm": 0.2008085548877716, "learning_rate": 2.3982300084086224e-05, "loss": 0.0143, "step": 69050 }, { "grad_norm": 0.1607101708650589, "learning_rate": 2.396818169067944e-05, "loss": 0.0133, "step": 69060 }, { "grad_norm": 0.13802017271518707, "learning_rate": 2.3954066144073178e-05, "loss": 0.0117, "step": 69070 }, { "grad_norm": 0.11839078366756439, "learning_rate": 2.39399534458111e-05, "loss": 0.0103, "step": 69080 }, { "grad_norm": 0.15792518854141235, "learning_rate": 2.392584359743651e-05, "loss": 0.0126, "step": 69090 }, { "grad_norm": 0.11395733058452606, "learning_rate": 2.3911736600492463e-05, "loss": 0.0112, "step": 69100 }, { "grad_norm": 0.12969893217086792, "learning_rate": 2.3897632456521672e-05, "loss": 0.0115, "step": 69110 }, { "grad_norm": 0.1534849852323532, "learning_rate": 2.388353116706658e-05, "loss": 0.0133, "step": 69120 }, { "grad_norm": 0.16271229088306427, "learning_rate": 2.3869432733669232e-05, "loss": 0.0126, "step": 69130 }, { "grad_norm": 0.173444464802742, "learning_rate": 2.385533715787145e-05, "loss": 0.0117, "step": 69140 }, { "grad_norm": 0.13358454406261444, "learning_rate": 2.3841244441214673e-05, "loss": 0.0106, "step": 69150 }, { "grad_norm": 0.1299900859594345, "learning_rate": 2.3827154585240073e-05, "loss": 0.0141, "step": 69160 }, { "grad_norm": 0.14543311297893524, "learning_rate": 2.3813067591488498e-05, "loss": 0.0096, "step": 69170 }, { "grad_norm": 0.1454901546239853, "learning_rate": 2.3798983461500494e-05, "loss": 0.0117, "step": 69180 }, { "grad_norm": 0.19257216155529022, "learning_rate": 2.3784902196816256e-05, "loss": 0.0096, "step": 69190 }, { "grad_norm": 0.16867804527282715, "learning_rate": 2.3770823798975716e-05, "loss": 0.01, "step": 69200 }, { "grad_norm": 0.1412588655948639, "learning_rate": 2.3756748269518424e-05, "loss": 0.0139, "step": 69210 }, { "grad_norm": 0.1382775604724884, "learning_rate": 2.374267560998371e-05, "loss": 0.0101, "step": 69220 }, { "grad_norm": 0.16933412849903107, "learning_rate": 2.3728605821910477e-05, "loss": 0.0114, "step": 69230 }, { "grad_norm": 0.18235348165035248, "learning_rate": 2.3714538906837453e-05, "loss": 0.0122, "step": 69240 }, { "grad_norm": 0.20029069483280182, "learning_rate": 2.3700474866302913e-05, "loss": 0.0138, "step": 69250 }, { "grad_norm": 0.10026683658361435, "learning_rate": 2.3686413701844917e-05, "loss": 0.0131, "step": 69260 }, { "grad_norm": 0.1631421446800232, "learning_rate": 2.367235541500114e-05, "loss": 0.0127, "step": 69270 }, { "grad_norm": 0.17534366250038147, "learning_rate": 2.3658300007308993e-05, "loss": 0.0149, "step": 69280 }, { "grad_norm": 0.1494683176279068, "learning_rate": 2.364424748030555e-05, "loss": 0.0137, "step": 69290 }, { "grad_norm": 0.22344562411308289, "learning_rate": 2.3630197835527595e-05, "loss": 0.0167, "step": 69300 }, { "grad_norm": 0.1294705718755722, "learning_rate": 2.3616151074511546e-05, "loss": 0.0104, "step": 69310 }, { "grad_norm": 0.1898704618215561, "learning_rate": 2.3602107198793545e-05, "loss": 0.0107, "step": 69320 }, { "grad_norm": 0.19417977333068848, "learning_rate": 2.3588066209909427e-05, "loss": 0.0109, "step": 69330 }, { "grad_norm": 0.14779239892959595, "learning_rate": 2.3574028109394662e-05, "loss": 0.0115, "step": 69340 }, { "grad_norm": 0.15084336698055267, "learning_rate": 2.3559992898784454e-05, "loss": 0.011, "step": 69350 }, { "grad_norm": 0.18636174499988556, "learning_rate": 2.354596057961367e-05, "loss": 0.014, "step": 69360 }, { "grad_norm": 0.13897258043289185, "learning_rate": 2.353193115341688e-05, "loss": 0.0114, "step": 69370 }, { "grad_norm": 0.1416764259338379, "learning_rate": 2.3517904621728294e-05, "loss": 0.0118, "step": 69380 }, { "grad_norm": 0.14348803460597992, "learning_rate": 2.3503880986081855e-05, "loss": 0.0107, "step": 69390 }, { "grad_norm": 0.1109633594751358, "learning_rate": 2.3489860248011136e-05, "loss": 0.0128, "step": 69400 }, { "grad_norm": 0.16655628383159637, "learning_rate": 2.3475842409049464e-05, "loss": 0.0114, "step": 69410 }, { "grad_norm": 0.17211180925369263, "learning_rate": 2.346182747072975e-05, "loss": 0.0116, "step": 69420 }, { "grad_norm": 0.12665563821792603, "learning_rate": 2.3447815434584718e-05, "loss": 0.0119, "step": 69430 }, { "grad_norm": 0.16429263353347778, "learning_rate": 2.3433806302146656e-05, "loss": 0.0108, "step": 69440 }, { "grad_norm": 0.16501331329345703, "learning_rate": 2.3419800074947612e-05, "loss": 0.0115, "step": 69450 }, { "grad_norm": 0.14633525907993317, "learning_rate": 2.3405796754519248e-05, "loss": 0.0108, "step": 69460 }, { "grad_norm": 0.1621638834476471, "learning_rate": 2.3391796342392992e-05, "loss": 0.0129, "step": 69470 }, { "grad_norm": 0.14080622792243958, "learning_rate": 2.337779884009984e-05, "loss": 0.0116, "step": 69480 }, { "grad_norm": 0.15276187658309937, "learning_rate": 2.3363804249170624e-05, "loss": 0.0125, "step": 69490 }, { "grad_norm": 0.16279572248458862, "learning_rate": 2.3349812571135703e-05, "loss": 0.0123, "step": 69500 }, { "grad_norm": 0.1412864774465561, "learning_rate": 2.333582380752523e-05, "loss": 0.0135, "step": 69510 }, { "grad_norm": 0.12389881163835526, "learning_rate": 2.3321837959868963e-05, "loss": 0.0108, "step": 69520 }, { "grad_norm": 0.1121707335114479, "learning_rate": 2.3307855029696383e-05, "loss": 0.011, "step": 69530 }, { "grad_norm": 0.1953432857990265, "learning_rate": 2.3293875018536648e-05, "loss": 0.014, "step": 69540 }, { "grad_norm": 0.15289904177188873, "learning_rate": 2.327989792791861e-05, "loss": 0.0103, "step": 69550 }, { "grad_norm": 0.17126286029815674, "learning_rate": 2.326592375937074e-05, "loss": 0.0119, "step": 69560 }, { "grad_norm": 0.1678466498851776, "learning_rate": 2.3251952514421248e-05, "loss": 0.0111, "step": 69570 }, { "grad_norm": 0.20638461410999298, "learning_rate": 2.323798419459804e-05, "loss": 0.0119, "step": 69580 }, { "grad_norm": 0.14263315498828888, "learning_rate": 2.3224018801428625e-05, "loss": 0.0124, "step": 69590 }, { "grad_norm": 0.12965159118175507, "learning_rate": 2.3210056336440273e-05, "loss": 0.0124, "step": 69600 }, { "grad_norm": 0.16031435132026672, "learning_rate": 2.3196096801159844e-05, "loss": 0.0134, "step": 69610 }, { "grad_norm": 0.19301649928092957, "learning_rate": 2.3182140197114012e-05, "loss": 0.0146, "step": 69620 }, { "grad_norm": 0.10625042021274567, "learning_rate": 2.3168186525828985e-05, "loss": 0.0113, "step": 69630 }, { "grad_norm": 0.14091823995113373, "learning_rate": 2.3154235788830757e-05, "loss": 0.0118, "step": 69640 }, { "grad_norm": 0.1980113983154297, "learning_rate": 2.3140287987644927e-05, "loss": 0.0143, "step": 69650 }, { "grad_norm": 0.16606482863426208, "learning_rate": 2.3126343123796833e-05, "loss": 0.0155, "step": 69660 }, { "grad_norm": 0.15017423033714294, "learning_rate": 2.3112401198811418e-05, "loss": 0.0131, "step": 69670 }, { "grad_norm": 0.18278071284294128, "learning_rate": 2.3098462214213418e-05, "loss": 0.0122, "step": 69680 }, { "grad_norm": 0.17386887967586517, "learning_rate": 2.308452617152712e-05, "loss": 0.0117, "step": 69690 }, { "grad_norm": 0.13369505107402802, "learning_rate": 2.3070593072276603e-05, "loss": 0.0114, "step": 69700 }, { "grad_norm": 0.10875898599624634, "learning_rate": 2.3056662917985518e-05, "loss": 0.011, "step": 69710 }, { "grad_norm": 0.16374316811561584, "learning_rate": 2.3042735710177283e-05, "loss": 0.0121, "step": 69720 }, { "grad_norm": 0.16320832073688507, "learning_rate": 2.302881145037492e-05, "loss": 0.0171, "step": 69730 }, { "grad_norm": 0.15985780954360962, "learning_rate": 2.3014890140101192e-05, "loss": 0.0125, "step": 69740 }, { "grad_norm": 0.14186589419841766, "learning_rate": 2.30009717808785e-05, "loss": 0.0131, "step": 69750 }, { "grad_norm": 0.16291537880897522, "learning_rate": 2.2987056374228967e-05, "loss": 0.0121, "step": 69760 }, { "grad_norm": 0.18167240917682648, "learning_rate": 2.297314392167431e-05, "loss": 0.0137, "step": 69770 }, { "grad_norm": 0.19988945126533508, "learning_rate": 2.295923442473601e-05, "loss": 0.0122, "step": 69780 }, { "grad_norm": 0.15277962386608124, "learning_rate": 2.294532788493519e-05, "loss": 0.0111, "step": 69790 }, { "grad_norm": 0.14348578453063965, "learning_rate": 2.293142430379262e-05, "loss": 0.0125, "step": 69800 }, { "grad_norm": 0.14880435168743134, "learning_rate": 2.291752368282879e-05, "loss": 0.0118, "step": 69810 }, { "grad_norm": 0.18759550154209137, "learning_rate": 2.290362602356385e-05, "loss": 0.0118, "step": 69820 }, { "grad_norm": 0.17524884641170502, "learning_rate": 2.288973132751765e-05, "loss": 0.0115, "step": 69830 }, { "grad_norm": 0.18195056915283203, "learning_rate": 2.287583959620965e-05, "loss": 0.0127, "step": 69840 }, { "grad_norm": 0.16707247495651245, "learning_rate": 2.2861950831159074e-05, "loss": 0.0137, "step": 69850 }, { "grad_norm": 0.15550637245178223, "learning_rate": 2.284806503388472e-05, "loss": 0.0109, "step": 69860 }, { "grad_norm": 0.1666823923587799, "learning_rate": 2.283418220590518e-05, "loss": 0.0144, "step": 69870 }, { "grad_norm": 0.14134210348129272, "learning_rate": 2.282030234873862e-05, "loss": 0.0111, "step": 69880 }, { "grad_norm": 0.1359635591506958, "learning_rate": 2.2806425463902948e-05, "loss": 0.0113, "step": 69890 }, { "grad_norm": 0.23844075202941895, "learning_rate": 2.279255155291568e-05, "loss": 0.0124, "step": 69900 }, { "grad_norm": 0.13846901059150696, "learning_rate": 2.277868061729409e-05, "loss": 0.0117, "step": 69910 }, { "grad_norm": 0.14044003188610077, "learning_rate": 2.276481265855504e-05, "loss": 0.0144, "step": 69920 }, { "grad_norm": 0.16007736325263977, "learning_rate": 2.2750947678215128e-05, "loss": 0.0123, "step": 69930 }, { "grad_norm": 0.11221791058778763, "learning_rate": 2.2737085677790616e-05, "loss": 0.01, "step": 69940 }, { "grad_norm": 0.15398862957954407, "learning_rate": 2.2723226658797443e-05, "loss": 0.0136, "step": 69950 }, { "grad_norm": 0.14825859665870667, "learning_rate": 2.270937062275117e-05, "loss": 0.0115, "step": 69960 }, { "grad_norm": 0.12529313564300537, "learning_rate": 2.2695517571167113e-05, "loss": 0.0115, "step": 69970 }, { "grad_norm": 0.15730826556682587, "learning_rate": 2.268166750556019e-05, "loss": 0.0136, "step": 69980 }, { "grad_norm": 0.18946252763271332, "learning_rate": 2.266782042744503e-05, "loss": 0.0132, "step": 69990 }, { "grad_norm": 0.1881275177001953, "learning_rate": 2.2653976338335935e-05, "loss": 0.0105, "step": 70000 }, { "grad_norm": 0.17058736085891724, "learning_rate": 2.2640135239746895e-05, "loss": 0.0137, "step": 70010 }, { "grad_norm": 0.1367187649011612, "learning_rate": 2.2626297133191515e-05, "loss": 0.0132, "step": 70020 }, { "grad_norm": 0.15123343467712402, "learning_rate": 2.261246202018312e-05, "loss": 0.0115, "step": 70030 }, { "grad_norm": 0.20772908627986908, "learning_rate": 2.2598629902234718e-05, "loss": 0.0139, "step": 70040 }, { "grad_norm": 0.18010513484477997, "learning_rate": 2.258480078085894e-05, "loss": 0.0138, "step": 70050 }, { "grad_norm": 0.18331952393054962, "learning_rate": 2.2570974657568126e-05, "loss": 0.0125, "step": 70060 }, { "grad_norm": 0.1847584843635559, "learning_rate": 2.2557151533874286e-05, "loss": 0.0126, "step": 70070 }, { "grad_norm": 0.12141510844230652, "learning_rate": 2.2543331411289116e-05, "loss": 0.0135, "step": 70080 }, { "grad_norm": 0.18327651917934418, "learning_rate": 2.252951429132392e-05, "loss": 0.0097, "step": 70090 }, { "grad_norm": 0.14431814849376678, "learning_rate": 2.2515700175489762e-05, "loss": 0.0114, "step": 70100 }, { "grad_norm": 0.12926729023456573, "learning_rate": 2.2501889065297288e-05, "loss": 0.0128, "step": 70110 }, { "grad_norm": 0.1564987748861313, "learning_rate": 2.2488080962256887e-05, "loss": 0.0096, "step": 70120 }, { "grad_norm": 0.14527387917041779, "learning_rate": 2.2474275867878585e-05, "loss": 0.0116, "step": 70130 }, { "grad_norm": 0.15327589213848114, "learning_rate": 2.246047378367211e-05, "loss": 0.0151, "step": 70140 }, { "grad_norm": 0.1463724970817566, "learning_rate": 2.2446674711146798e-05, "loss": 0.0125, "step": 70150 }, { "grad_norm": 0.15809588134288788, "learning_rate": 2.2432878651811734e-05, "loss": 0.0152, "step": 70160 }, { "grad_norm": 0.15332666039466858, "learning_rate": 2.2419085607175594e-05, "loss": 0.0125, "step": 70170 }, { "grad_norm": 0.16763994097709656, "learning_rate": 2.240529557874678e-05, "loss": 0.0116, "step": 70180 }, { "grad_norm": 0.15099066495895386, "learning_rate": 2.2391508568033358e-05, "loss": 0.0119, "step": 70190 }, { "grad_norm": 0.12627804279327393, "learning_rate": 2.237772457654307e-05, "loss": 0.0143, "step": 70200 }, { "grad_norm": 0.16770219802856445, "learning_rate": 2.236394360578327e-05, "loss": 0.0115, "step": 70210 }, { "grad_norm": 0.2156907469034195, "learning_rate": 2.2350165657261073e-05, "loss": 0.0154, "step": 70220 }, { "grad_norm": 0.13921964168548584, "learning_rate": 2.2336390732483164e-05, "loss": 0.013, "step": 70230 }, { "grad_norm": 0.19316647946834564, "learning_rate": 2.2322618832955976e-05, "loss": 0.0141, "step": 70240 }, { "grad_norm": 0.16968247294425964, "learning_rate": 2.2308849960185578e-05, "loss": 0.013, "step": 70250 }, { "grad_norm": 0.14308837056159973, "learning_rate": 2.229508411567774e-05, "loss": 0.011, "step": 70260 }, { "grad_norm": 0.17529156804084778, "learning_rate": 2.2281321300937823e-05, "loss": 0.0117, "step": 70270 }, { "grad_norm": 0.1832498162984848, "learning_rate": 2.2267561517470937e-05, "loss": 0.0113, "step": 70280 }, { "grad_norm": 0.14992474019527435, "learning_rate": 2.2253804766781845e-05, "loss": 0.0103, "step": 70290 }, { "grad_norm": 0.11429088562726974, "learning_rate": 2.2240051050374933e-05, "loss": 0.0105, "step": 70300 }, { "grad_norm": 0.2197277694940567, "learning_rate": 2.22263003697543e-05, "loss": 0.0107, "step": 70310 }, { "grad_norm": 0.16467945277690887, "learning_rate": 2.2212552726423702e-05, "loss": 0.0132, "step": 70320 }, { "grad_norm": 0.18851304054260254, "learning_rate": 2.2198808121886578e-05, "loss": 0.0124, "step": 70330 }, { "grad_norm": 0.13656306266784668, "learning_rate": 2.218506655764598e-05, "loss": 0.0126, "step": 70340 }, { "grad_norm": 0.24218793213367462, "learning_rate": 2.2171328035204704e-05, "loss": 0.0142, "step": 70350 }, { "grad_norm": 0.13956694304943085, "learning_rate": 2.2157592556065133e-05, "loss": 0.0113, "step": 70360 }, { "grad_norm": 0.15215463936328888, "learning_rate": 2.2143860121729382e-05, "loss": 0.0133, "step": 70370 }, { "grad_norm": 0.18787389993667603, "learning_rate": 2.2130130733699206e-05, "loss": 0.0147, "step": 70380 }, { "grad_norm": 0.17097951471805573, "learning_rate": 2.2116404393476052e-05, "loss": 0.0109, "step": 70390 }, { "grad_norm": 0.21110856533050537, "learning_rate": 2.2102681102560967e-05, "loss": 0.0179, "step": 70400 }, { "grad_norm": 0.13487286865711212, "learning_rate": 2.208896086245476e-05, "loss": 0.0127, "step": 70410 }, { "grad_norm": 0.14060407876968384, "learning_rate": 2.2075243674657808e-05, "loss": 0.01, "step": 70420 }, { "grad_norm": 0.1422145962715149, "learning_rate": 2.206152954067024e-05, "loss": 0.0115, "step": 70430 }, { "grad_norm": 0.15983238816261292, "learning_rate": 2.2047818461991758e-05, "loss": 0.011, "step": 70440 }, { "grad_norm": 0.15786458551883698, "learning_rate": 2.2034110440121864e-05, "loss": 0.01, "step": 70450 }, { "grad_norm": 0.11892019957304001, "learning_rate": 2.2020405476559585e-05, "loss": 0.013, "step": 70460 }, { "grad_norm": 0.1616799384355545, "learning_rate": 2.2006703572803722e-05, "loss": 0.0116, "step": 70470 }, { "grad_norm": 0.17294026911258698, "learning_rate": 2.1993004730352646e-05, "loss": 0.0142, "step": 70480 }, { "grad_norm": 0.13921821117401123, "learning_rate": 2.1979308950704464e-05, "loss": 0.0106, "step": 70490 }, { "grad_norm": 0.08338797837495804, "learning_rate": 2.1965616235356923e-05, "loss": 0.0117, "step": 70500 }, { "grad_norm": 0.15956127643585205, "learning_rate": 2.1951926585807465e-05, "loss": 0.0137, "step": 70510 }, { "grad_norm": 0.12345018982887268, "learning_rate": 2.193824000355313e-05, "loss": 0.012, "step": 70520 }, { "grad_norm": 0.1708412617444992, "learning_rate": 2.192455649009067e-05, "loss": 0.0104, "step": 70530 }, { "grad_norm": 0.17913652956485748, "learning_rate": 2.1910876046916523e-05, "loss": 0.0136, "step": 70540 }, { "grad_norm": 0.10581230372190475, "learning_rate": 2.1897198675526725e-05, "loss": 0.0106, "step": 70550 }, { "grad_norm": 0.09101744741201401, "learning_rate": 2.1883524377417024e-05, "loss": 0.0106, "step": 70560 }, { "grad_norm": 0.1349002867937088, "learning_rate": 2.1869853154082825e-05, "loss": 0.0099, "step": 70570 }, { "grad_norm": 0.1645408272743225, "learning_rate": 2.1856185007019215e-05, "loss": 0.0135, "step": 70580 }, { "grad_norm": 0.14040541648864746, "learning_rate": 2.1842519937720874e-05, "loss": 0.0138, "step": 70590 }, { "grad_norm": 0.09346053749322891, "learning_rate": 2.182885794768224e-05, "loss": 0.0122, "step": 70600 }, { "grad_norm": 0.12498918920755386, "learning_rate": 2.1815199038397326e-05, "loss": 0.0123, "step": 70610 }, { "grad_norm": 0.14095693826675415, "learning_rate": 2.1801543211359893e-05, "loss": 0.0122, "step": 70620 }, { "grad_norm": 0.1801639199256897, "learning_rate": 2.178789046806326e-05, "loss": 0.0119, "step": 70630 }, { "grad_norm": 0.12555038928985596, "learning_rate": 2.1774240810000547e-05, "loss": 0.0124, "step": 70640 }, { "grad_norm": 0.11739795655012131, "learning_rate": 2.17605942386644e-05, "loss": 0.011, "step": 70650 }, { "grad_norm": 0.16650384664535522, "learning_rate": 2.174695075554723e-05, "loss": 0.0143, "step": 70660 }, { "grad_norm": 0.15098993480205536, "learning_rate": 2.173331036214103e-05, "loss": 0.0116, "step": 70670 }, { "grad_norm": 0.1482757329940796, "learning_rate": 2.171967305993753e-05, "loss": 0.0123, "step": 70680 }, { "grad_norm": 0.17826923727989197, "learning_rate": 2.1706038850428024e-05, "loss": 0.0129, "step": 70690 }, { "grad_norm": 0.15548938512802124, "learning_rate": 2.169240773510361e-05, "loss": 0.0088, "step": 70700 }, { "grad_norm": 0.14223867654800415, "learning_rate": 2.1678779715454906e-05, "loss": 0.0104, "step": 70710 }, { "grad_norm": 0.1216566413640976, "learning_rate": 2.166515479297229e-05, "loss": 0.0094, "step": 70720 }, { "grad_norm": 0.1397736370563507, "learning_rate": 2.1651532969145732e-05, "loss": 0.0171, "step": 70730 }, { "grad_norm": 0.136525496840477, "learning_rate": 2.1637914245464903e-05, "loss": 0.0114, "step": 70740 }, { "grad_norm": 0.13836051523685455, "learning_rate": 2.1624298623419133e-05, "loss": 0.0101, "step": 70750 }, { "grad_norm": 0.18315303325653076, "learning_rate": 2.1610686104497413e-05, "loss": 0.0096, "step": 70760 }, { "grad_norm": 0.12946294248104095, "learning_rate": 2.1597076690188366e-05, "loss": 0.014, "step": 70770 }, { "grad_norm": 0.16249892115592957, "learning_rate": 2.1583470381980315e-05, "loss": 0.0089, "step": 70780 }, { "grad_norm": 0.19161655008792877, "learning_rate": 2.156986718136123e-05, "loss": 0.0113, "step": 70790 }, { "grad_norm": 0.17945340275764465, "learning_rate": 2.155626708981871e-05, "loss": 0.0135, "step": 70800 }, { "grad_norm": 0.1754807084798813, "learning_rate": 2.1542670108840075e-05, "loss": 0.0119, "step": 70810 }, { "grad_norm": 0.18301774561405182, "learning_rate": 2.152907623991221e-05, "loss": 0.009, "step": 70820 }, { "grad_norm": 0.14099057018756866, "learning_rate": 2.1515485484521802e-05, "loss": 0.012, "step": 70830 }, { "grad_norm": 0.22747346758842468, "learning_rate": 2.150189784415506e-05, "loss": 0.0143, "step": 70840 }, { "grad_norm": 0.17343658208847046, "learning_rate": 2.148831332029794e-05, "loss": 0.0139, "step": 70850 }, { "grad_norm": 0.2007414847612381, "learning_rate": 2.147473191443599e-05, "loss": 0.0109, "step": 70860 }, { "grad_norm": 0.2218938022851944, "learning_rate": 2.1461153628054492e-05, "loss": 0.0161, "step": 70870 }, { "grad_norm": 0.11682560294866562, "learning_rate": 2.1447578462638284e-05, "loss": 0.011, "step": 70880 }, { "grad_norm": 0.12028307467699051, "learning_rate": 2.143400641967201e-05, "loss": 0.0134, "step": 70890 }, { "grad_norm": 0.17219431698322296, "learning_rate": 2.1420437500639823e-05, "loss": 0.0115, "step": 70900 }, { "grad_norm": 0.15459275245666504, "learning_rate": 2.1406871707025646e-05, "loss": 0.0101, "step": 70910 }, { "grad_norm": 0.15359143912792206, "learning_rate": 2.1393309040312963e-05, "loss": 0.0153, "step": 70920 }, { "grad_norm": 0.21737779676914215, "learning_rate": 2.137974950198502e-05, "loss": 0.0136, "step": 70930 }, { "grad_norm": 0.15588396787643433, "learning_rate": 2.1366193093524617e-05, "loss": 0.0116, "step": 70940 }, { "grad_norm": 0.11913885176181793, "learning_rate": 2.1352639816414288e-05, "loss": 0.0129, "step": 70950 }, { "grad_norm": 0.18703396618366241, "learning_rate": 2.1339089672136202e-05, "loss": 0.0115, "step": 70960 }, { "grad_norm": 0.21880854666233063, "learning_rate": 2.1325542662172193e-05, "loss": 0.0109, "step": 70970 }, { "grad_norm": 0.14441530406475067, "learning_rate": 2.1311998788003712e-05, "loss": 0.0091, "step": 70980 }, { "grad_norm": 0.11927711963653564, "learning_rate": 2.1298458051111907e-05, "loss": 0.0163, "step": 70990 }, { "grad_norm": 0.16761863231658936, "learning_rate": 2.1284920452977602e-05, "loss": 0.0121, "step": 71000 }, { "grad_norm": 0.17977678775787354, "learning_rate": 2.12713859950812e-05, "loss": 0.0147, "step": 71010 }, { "grad_norm": 0.1904757171869278, "learning_rate": 2.1257854678902832e-05, "loss": 0.0143, "step": 71020 }, { "grad_norm": 0.26683324575424194, "learning_rate": 2.124432650592227e-05, "loss": 0.0149, "step": 71030 }, { "grad_norm": 0.15801921486854553, "learning_rate": 2.1230801477618937e-05, "loss": 0.011, "step": 71040 }, { "grad_norm": 0.1520441621541977, "learning_rate": 2.121727959547189e-05, "loss": 0.0125, "step": 71050 }, { "grad_norm": 0.16921475529670715, "learning_rate": 2.1203760860959886e-05, "loss": 0.0128, "step": 71060 }, { "grad_norm": 0.2409859150648117, "learning_rate": 2.119024527556127e-05, "loss": 0.0146, "step": 71070 }, { "grad_norm": 0.14608879387378693, "learning_rate": 2.117673284075415e-05, "loss": 0.0118, "step": 71080 }, { "grad_norm": 0.19102132320404053, "learning_rate": 2.1163223558016165e-05, "loss": 0.0132, "step": 71090 }, { "grad_norm": 0.13716763257980347, "learning_rate": 2.1149717428824722e-05, "loss": 0.0096, "step": 71100 }, { "grad_norm": 0.15147128701210022, "learning_rate": 2.1136214454656784e-05, "loss": 0.0129, "step": 71110 }, { "grad_norm": 0.09992507100105286, "learning_rate": 2.1122714636989056e-05, "loss": 0.0092, "step": 71120 }, { "grad_norm": 0.18426012992858887, "learning_rate": 2.1109217977297823e-05, "loss": 0.0127, "step": 71130 }, { "grad_norm": 0.1840216964483261, "learning_rate": 2.1095724477059077e-05, "loss": 0.0122, "step": 71140 }, { "grad_norm": 0.13934548199176788, "learning_rate": 2.1082234137748434e-05, "loss": 0.0082, "step": 71150 }, { "grad_norm": 0.1288091540336609, "learning_rate": 2.106874696084122e-05, "loss": 0.0114, "step": 71160 }, { "grad_norm": 0.16549555957317352, "learning_rate": 2.1055262947812315e-05, "loss": 0.0118, "step": 71170 }, { "grad_norm": 0.1395273506641388, "learning_rate": 2.104178210013636e-05, "loss": 0.0102, "step": 71180 }, { "grad_norm": 0.17107626795768738, "learning_rate": 2.1028304419287554e-05, "loss": 0.0117, "step": 71190 }, { "grad_norm": 0.13683755695819855, "learning_rate": 2.101482990673982e-05, "loss": 0.0105, "step": 71200 }, { "grad_norm": 0.12031126022338867, "learning_rate": 2.1001358563966707e-05, "loss": 0.0102, "step": 71210 }, { "grad_norm": 0.1772654950618744, "learning_rate": 2.098789039244144e-05, "loss": 0.0133, "step": 71220 }, { "grad_norm": 0.1453229784965515, "learning_rate": 2.097442539363685e-05, "loss": 0.0147, "step": 71230 }, { "grad_norm": 0.11824934184551239, "learning_rate": 2.0960963569025448e-05, "loss": 0.0123, "step": 71240 }, { "grad_norm": 0.20066595077514648, "learning_rate": 2.0947504920079435e-05, "loss": 0.012, "step": 71250 }, { "grad_norm": 0.1289968490600586, "learning_rate": 2.093404944827059e-05, "loss": 0.0135, "step": 71260 }, { "grad_norm": 0.13766054809093475, "learning_rate": 2.0920597155070394e-05, "loss": 0.0126, "step": 71270 }, { "grad_norm": 0.22787976264953613, "learning_rate": 2.090714804194997e-05, "loss": 0.0135, "step": 71280 }, { "grad_norm": 0.13283149898052216, "learning_rate": 2.0893702110380114e-05, "loss": 0.0114, "step": 71290 }, { "grad_norm": 0.1248726099729538, "learning_rate": 2.0880259361831216e-05, "loss": 0.0113, "step": 71300 }, { "grad_norm": 0.12950214743614197, "learning_rate": 2.0866819797773397e-05, "loss": 0.0101, "step": 71310 }, { "grad_norm": 0.11057894676923752, "learning_rate": 2.085338341967634e-05, "loss": 0.0114, "step": 71320 }, { "grad_norm": 0.11693546921014786, "learning_rate": 2.083995022900946e-05, "loss": 0.01, "step": 71330 }, { "grad_norm": 0.11992734670639038, "learning_rate": 2.0826520227241768e-05, "loss": 0.0118, "step": 71340 }, { "grad_norm": 0.14775942265987396, "learning_rate": 2.081309341584199e-05, "loss": 0.0116, "step": 71350 }, { "grad_norm": 0.14993472397327423, "learning_rate": 2.0799669796278413e-05, "loss": 0.0114, "step": 71360 }, { "grad_norm": 0.24986720085144043, "learning_rate": 2.078624937001906e-05, "loss": 0.0139, "step": 71370 }, { "grad_norm": 0.17445574700832367, "learning_rate": 2.0772832138531544e-05, "loss": 0.0126, "step": 71380 }, { "grad_norm": 0.21192437410354614, "learning_rate": 2.0759418103283158e-05, "loss": 0.0111, "step": 71390 }, { "grad_norm": 0.13947096467018127, "learning_rate": 2.0746007265740846e-05, "loss": 0.0126, "step": 71400 }, { "grad_norm": 0.1203923374414444, "learning_rate": 2.0732599627371215e-05, "loss": 0.0092, "step": 71410 }, { "grad_norm": 0.14323878288269043, "learning_rate": 2.0719195189640468e-05, "loss": 0.0125, "step": 71420 }, { "grad_norm": 0.15720677375793457, "learning_rate": 2.0705793954014528e-05, "loss": 0.0184, "step": 71430 }, { "grad_norm": 0.14749367535114288, "learning_rate": 2.06923959219589e-05, "loss": 0.0105, "step": 71440 }, { "grad_norm": 0.14432457089424133, "learning_rate": 2.0679001094938793e-05, "loss": 0.0112, "step": 71450 }, { "grad_norm": 0.19917826354503632, "learning_rate": 2.066560947441904e-05, "loss": 0.0151, "step": 71460 }, { "grad_norm": 0.08465708792209625, "learning_rate": 2.0652221061864152e-05, "loss": 0.0104, "step": 71470 }, { "grad_norm": 0.18386420607566833, "learning_rate": 2.0638835858738226e-05, "loss": 0.0104, "step": 71480 }, { "grad_norm": 0.13430370390415192, "learning_rate": 2.0625453866505068e-05, "loss": 0.011, "step": 71490 }, { "grad_norm": 0.1725887656211853, "learning_rate": 2.0612075086628123e-05, "loss": 0.0128, "step": 71500 }, { "grad_norm": 0.19307735562324524, "learning_rate": 2.059869952057045e-05, "loss": 0.01, "step": 71510 }, { "grad_norm": 0.1698838770389557, "learning_rate": 2.0585327169794793e-05, "loss": 0.0111, "step": 71520 }, { "grad_norm": 0.12389026582241058, "learning_rate": 2.0571958035763534e-05, "loss": 0.0108, "step": 71530 }, { "grad_norm": 0.1664363592863083, "learning_rate": 2.0558592119938715e-05, "loss": 0.0142, "step": 71540 }, { "grad_norm": 0.09118450433015823, "learning_rate": 2.054522942378198e-05, "loss": 0.0096, "step": 71550 }, { "grad_norm": 0.1522817462682724, "learning_rate": 2.053186994875469e-05, "loss": 0.0118, "step": 71560 }, { "grad_norm": 0.18237271904945374, "learning_rate": 2.0518513696317775e-05, "loss": 0.0116, "step": 71570 }, { "grad_norm": 0.1297866255044937, "learning_rate": 2.050516066793188e-05, "loss": 0.0097, "step": 71580 }, { "grad_norm": 0.15673767030239105, "learning_rate": 2.0491810865057276e-05, "loss": 0.0147, "step": 71590 }, { "grad_norm": 0.1382240355014801, "learning_rate": 2.0478464289153882e-05, "loss": 0.012, "step": 71600 }, { "grad_norm": 0.19722840189933777, "learning_rate": 2.0465120941681232e-05, "loss": 0.0133, "step": 71610 }, { "grad_norm": 0.1750887930393219, "learning_rate": 2.0451780824098566e-05, "loss": 0.014, "step": 71620 }, { "grad_norm": 0.1719031184911728, "learning_rate": 2.043844393786471e-05, "loss": 0.0124, "step": 71630 }, { "grad_norm": 0.1274697184562683, "learning_rate": 2.0425110284438193e-05, "loss": 0.0105, "step": 71640 }, { "grad_norm": 0.16144639253616333, "learning_rate": 2.041177986527712e-05, "loss": 0.011, "step": 71650 }, { "grad_norm": 0.1590239554643631, "learning_rate": 2.0398452681839347e-05, "loss": 0.0114, "step": 71660 }, { "grad_norm": 0.12768922746181488, "learning_rate": 2.0385128735582272e-05, "loss": 0.0097, "step": 71670 }, { "grad_norm": 0.12173112481832504, "learning_rate": 2.0371808027963003e-05, "loss": 0.0124, "step": 71680 }, { "grad_norm": 0.11883264034986496, "learning_rate": 2.035849056043825e-05, "loss": 0.0144, "step": 71690 }, { "grad_norm": 0.0884840190410614, "learning_rate": 2.03451763344644e-05, "loss": 0.0097, "step": 71700 }, { "grad_norm": 0.13046890497207642, "learning_rate": 2.0331865351497482e-05, "loss": 0.0114, "step": 71710 }, { "grad_norm": 0.12982277572155, "learning_rate": 2.0318557612993185e-05, "loss": 0.0101, "step": 71720 }, { "grad_norm": 0.18330980837345123, "learning_rate": 2.0305253120406782e-05, "loss": 0.0133, "step": 71730 }, { "grad_norm": 0.1340123414993286, "learning_rate": 2.029195187519325e-05, "loss": 0.0158, "step": 71740 }, { "grad_norm": 0.12498881667852402, "learning_rate": 2.0278653878807217e-05, "loss": 0.0106, "step": 71750 }, { "grad_norm": 0.1484375, "learning_rate": 2.0265359132702892e-05, "loss": 0.01, "step": 71760 }, { "grad_norm": 0.181749165058136, "learning_rate": 2.0252067638334182e-05, "loss": 0.0103, "step": 71770 }, { "grad_norm": 0.173179030418396, "learning_rate": 2.023877939715464e-05, "loss": 0.0107, "step": 71780 }, { "grad_norm": 0.19334855675697327, "learning_rate": 2.0225494410617448e-05, "loss": 0.01, "step": 71790 }, { "grad_norm": 0.14166446030139923, "learning_rate": 2.02122126801754e-05, "loss": 0.0109, "step": 71800 }, { "grad_norm": 0.13171064853668213, "learning_rate": 2.0198934207281018e-05, "loss": 0.0121, "step": 71810 }, { "grad_norm": 0.23078858852386475, "learning_rate": 2.0185658993386357e-05, "loss": 0.0117, "step": 71820 }, { "grad_norm": 0.1108289584517479, "learning_rate": 2.0172387039943226e-05, "loss": 0.0102, "step": 71830 }, { "grad_norm": 0.17276933789253235, "learning_rate": 2.0159118348402968e-05, "loss": 0.0119, "step": 71840 }, { "grad_norm": 0.14338582754135132, "learning_rate": 2.0145852920216697e-05, "loss": 0.0127, "step": 71850 }, { "grad_norm": 0.1516067534685135, "learning_rate": 2.0132590756835045e-05, "loss": 0.0105, "step": 71860 }, { "grad_norm": 0.13519226014614105, "learning_rate": 2.0119331859708386e-05, "loss": 0.0105, "step": 71870 }, { "grad_norm": 0.1331917941570282, "learning_rate": 2.0106076230286654e-05, "loss": 0.0099, "step": 71880 }, { "grad_norm": 0.17544326186180115, "learning_rate": 2.0092823870019488e-05, "loss": 0.0129, "step": 71890 }, { "grad_norm": 0.17038078606128693, "learning_rate": 2.0079574780356115e-05, "loss": 0.0103, "step": 71900 }, { "grad_norm": 0.14094431698322296, "learning_rate": 2.0066328962745486e-05, "loss": 0.01, "step": 71910 }, { "grad_norm": 0.1427423655986786, "learning_rate": 2.00530864186361e-05, "loss": 0.0096, "step": 71920 }, { "grad_norm": 0.15104839205741882, "learning_rate": 2.003984714947618e-05, "loss": 0.0106, "step": 71930 }, { "grad_norm": 0.20138998329639435, "learning_rate": 2.0026611156713515e-05, "loss": 0.0113, "step": 71940 }, { "grad_norm": 0.15794168412685394, "learning_rate": 2.0013378441795584e-05, "loss": 0.011, "step": 71950 }, { "grad_norm": 0.208625927567482, "learning_rate": 2.0000149006169506e-05, "loss": 0.0108, "step": 71960 }, { "grad_norm": 0.18582239747047424, "learning_rate": 1.9986922851282048e-05, "loss": 0.0132, "step": 71970 }, { "grad_norm": 0.1743973344564438, "learning_rate": 1.997369997857956e-05, "loss": 0.0131, "step": 71980 }, { "grad_norm": 0.18128733336925507, "learning_rate": 1.9960480389508103e-05, "loss": 0.0123, "step": 71990 }, { "grad_norm": 0.14560575783252716, "learning_rate": 1.9947264085513363e-05, "loss": 0.0141, "step": 72000 }, { "grad_norm": 0.1643078476190567, "learning_rate": 1.993405106804062e-05, "loss": 0.0107, "step": 72010 }, { "grad_norm": 0.14064544439315796, "learning_rate": 1.9920841338534867e-05, "loss": 0.0118, "step": 72020 }, { "grad_norm": 0.15197378396987915, "learning_rate": 1.9907634898440646e-05, "loss": 0.0096, "step": 72030 }, { "grad_norm": 0.1535525768995285, "learning_rate": 1.989443174920227e-05, "loss": 0.0124, "step": 72040 }, { "grad_norm": 0.14851057529449463, "learning_rate": 1.988123189226355e-05, "loss": 0.013, "step": 72050 }, { "grad_norm": 0.15601840615272522, "learning_rate": 1.986803532906805e-05, "loss": 0.013, "step": 72060 }, { "grad_norm": 0.16930142045021057, "learning_rate": 1.985484206105889e-05, "loss": 0.009, "step": 72070 }, { "grad_norm": 0.13016219437122345, "learning_rate": 1.9841652089678887e-05, "loss": 0.0136, "step": 72080 }, { "grad_norm": 0.16104525327682495, "learning_rate": 1.9828465416370435e-05, "loss": 0.012, "step": 72090 }, { "grad_norm": 0.18821127712726593, "learning_rate": 1.9815282042575685e-05, "loss": 0.0086, "step": 72100 }, { "grad_norm": 0.16463345289230347, "learning_rate": 1.9802101969736285e-05, "loss": 0.0131, "step": 72110 }, { "grad_norm": 0.12378736585378647, "learning_rate": 1.9788925199293633e-05, "loss": 0.0091, "step": 72120 }, { "grad_norm": 0.12789566814899445, "learning_rate": 1.977575173268868e-05, "loss": 0.0121, "step": 72130 }, { "grad_norm": 0.1583508402109146, "learning_rate": 1.9762581571362094e-05, "loss": 0.0118, "step": 72140 }, { "grad_norm": 0.13941147923469543, "learning_rate": 1.9749414716754112e-05, "loss": 0.0124, "step": 72150 }, { "grad_norm": 0.14807532727718353, "learning_rate": 1.973625117030465e-05, "loss": 0.0109, "step": 72160 }, { "grad_norm": 0.15067888796329498, "learning_rate": 1.972309093345326e-05, "loss": 0.0133, "step": 72170 }, { "grad_norm": 0.14687225222587585, "learning_rate": 1.970993400763915e-05, "loss": 0.0119, "step": 72180 }, { "grad_norm": 0.12823449075222015, "learning_rate": 1.969678039430109e-05, "loss": 0.0108, "step": 72190 }, { "grad_norm": 0.1610414832830429, "learning_rate": 1.9683630094877564e-05, "loss": 0.0149, "step": 72200 }, { "grad_norm": 0.20179274678230286, "learning_rate": 1.9670483110806696e-05, "loss": 0.0125, "step": 72210 }, { "grad_norm": 0.14736846089363098, "learning_rate": 1.9657339443526172e-05, "loss": 0.0101, "step": 72220 }, { "grad_norm": 0.08865713328123093, "learning_rate": 1.964419909447339e-05, "loss": 0.0094, "step": 72230 }, { "grad_norm": 0.18694496154785156, "learning_rate": 1.963106206508536e-05, "loss": 0.0127, "step": 72240 }, { "grad_norm": 0.14162549376487732, "learning_rate": 1.9617928356798738e-05, "loss": 0.0102, "step": 72250 }, { "grad_norm": 0.1783844232559204, "learning_rate": 1.960479797104977e-05, "loss": 0.0114, "step": 72260 }, { "grad_norm": 0.12536264955997467, "learning_rate": 1.9591670909274418e-05, "loss": 0.0102, "step": 72270 }, { "grad_norm": 0.13777093589305878, "learning_rate": 1.9578547172908184e-05, "loss": 0.0117, "step": 72280 }, { "grad_norm": 0.15107642114162445, "learning_rate": 1.9565426763386325e-05, "loss": 0.0108, "step": 72290 }, { "grad_norm": 0.13563105463981628, "learning_rate": 1.9552309682143622e-05, "loss": 0.0117, "step": 72300 }, { "grad_norm": 0.1619448959827423, "learning_rate": 1.9539195930614574e-05, "loss": 0.0114, "step": 72310 }, { "grad_norm": 0.16617374122142792, "learning_rate": 1.952608551023324e-05, "loss": 0.0074, "step": 72320 }, { "grad_norm": 0.1252593696117401, "learning_rate": 1.9512978422433397e-05, "loss": 0.0131, "step": 72330 }, { "grad_norm": 0.18562503159046173, "learning_rate": 1.9499874668648376e-05, "loss": 0.0103, "step": 72340 }, { "grad_norm": 0.19877570867538452, "learning_rate": 1.9486774250311213e-05, "loss": 0.0098, "step": 72350 }, { "grad_norm": 0.33676740527153015, "learning_rate": 1.9473677168854537e-05, "loss": 0.012, "step": 72360 }, { "grad_norm": 0.13898199796676636, "learning_rate": 1.946058342571065e-05, "loss": 0.0104, "step": 72370 }, { "grad_norm": 0.15598788857460022, "learning_rate": 1.9447493022311424e-05, "loss": 0.0105, "step": 72380 }, { "grad_norm": 0.14179620146751404, "learning_rate": 1.9434405960088442e-05, "loss": 0.0099, "step": 72390 }, { "grad_norm": 0.11271895468235016, "learning_rate": 1.9421322240472857e-05, "loss": 0.0105, "step": 72400 }, { "grad_norm": 0.10861845314502716, "learning_rate": 1.9408241864895494e-05, "loss": 0.0113, "step": 72410 }, { "grad_norm": 0.16138654947280884, "learning_rate": 1.939516483478681e-05, "loss": 0.0091, "step": 72420 }, { "grad_norm": 0.14781518280506134, "learning_rate": 1.9382091151576902e-05, "loss": 0.0102, "step": 72430 }, { "grad_norm": 0.13341405987739563, "learning_rate": 1.9369020816695454e-05, "loss": 0.0099, "step": 72440 }, { "grad_norm": 0.15140381455421448, "learning_rate": 1.935595383157184e-05, "loss": 0.0104, "step": 72450 }, { "grad_norm": 0.15203239023685455, "learning_rate": 1.9342890197635057e-05, "loss": 0.0152, "step": 72460 }, { "grad_norm": 0.15322594344615936, "learning_rate": 1.9329829916313685e-05, "loss": 0.0106, "step": 72470 }, { "grad_norm": 0.1665249764919281, "learning_rate": 1.9316772989036013e-05, "loss": 0.0118, "step": 72480 }, { "grad_norm": 0.11823597550392151, "learning_rate": 1.930371941722991e-05, "loss": 0.013, "step": 72490 }, { "grad_norm": 0.1164325699210167, "learning_rate": 1.929066920232292e-05, "loss": 0.0109, "step": 72500 }, { "grad_norm": 0.2193230241537094, "learning_rate": 1.9277622345742152e-05, "loss": 0.0108, "step": 72510 }, { "grad_norm": 0.15101636946201324, "learning_rate": 1.926457884891444e-05, "loss": 0.0144, "step": 72520 }, { "grad_norm": 0.13830600678920746, "learning_rate": 1.9251538713266147e-05, "loss": 0.0109, "step": 72530 }, { "grad_norm": 0.1719801276922226, "learning_rate": 1.9238501940223358e-05, "loss": 0.0119, "step": 72540 }, { "grad_norm": 0.14602583646774292, "learning_rate": 1.922546853121175e-05, "loss": 0.0122, "step": 72550 }, { "grad_norm": 0.16344968974590302, "learning_rate": 1.9212438487656647e-05, "loss": 0.0118, "step": 72560 }, { "grad_norm": 0.11302413046360016, "learning_rate": 1.9199411810982965e-05, "loss": 0.01, "step": 72570 }, { "grad_norm": 0.1884429156780243, "learning_rate": 1.918638850261532e-05, "loss": 0.0123, "step": 72580 }, { "grad_norm": 0.13482557237148285, "learning_rate": 1.9173368563977885e-05, "loss": 0.0115, "step": 72590 }, { "grad_norm": 0.18182846903800964, "learning_rate": 1.9160351996494513e-05, "loss": 0.0103, "step": 72600 }, { "grad_norm": 0.14588986337184906, "learning_rate": 1.9147338801588677e-05, "loss": 0.0115, "step": 72610 }, { "grad_norm": 0.20708079636096954, "learning_rate": 1.913432898068351e-05, "loss": 0.0124, "step": 72620 }, { "grad_norm": 0.11265754699707031, "learning_rate": 1.912132253520169e-05, "loss": 0.0129, "step": 72630 }, { "grad_norm": 0.1443132907152176, "learning_rate": 1.910831946656564e-05, "loss": 0.01, "step": 72640 }, { "grad_norm": 0.13361366093158722, "learning_rate": 1.909531977619731e-05, "loss": 0.0099, "step": 72650 }, { "grad_norm": 0.12073321640491486, "learning_rate": 1.908232346551834e-05, "loss": 0.0134, "step": 72660 }, { "grad_norm": 0.13847751915454865, "learning_rate": 1.9069330535949993e-05, "loss": 0.0118, "step": 72670 }, { "grad_norm": 0.13023293018341064, "learning_rate": 1.9056340988913167e-05, "loss": 0.0119, "step": 72680 }, { "grad_norm": 0.11446599662303925, "learning_rate": 1.9043354825828357e-05, "loss": 0.0091, "step": 72690 }, { "grad_norm": 0.18330411612987518, "learning_rate": 1.9030372048115724e-05, "loss": 0.0135, "step": 72700 }, { "grad_norm": 0.14126329123973846, "learning_rate": 1.9017392657195056e-05, "loss": 0.0107, "step": 72710 }, { "grad_norm": 0.15511572360992432, "learning_rate": 1.9004416654485724e-05, "loss": 0.0101, "step": 72720 }, { "grad_norm": 0.11234892159700394, "learning_rate": 1.8991444041406785e-05, "loss": 0.0118, "step": 72730 }, { "grad_norm": 0.12821047008037567, "learning_rate": 1.8978474819376912e-05, "loss": 0.0119, "step": 72740 }, { "grad_norm": 0.11849594116210938, "learning_rate": 1.8965508989814406e-05, "loss": 0.0109, "step": 72750 }, { "grad_norm": 0.17246316373348236, "learning_rate": 1.8952546554137158e-05, "loss": 0.0137, "step": 72760 }, { "grad_norm": 0.129095196723938, "learning_rate": 1.8939587513762757e-05, "loss": 0.0117, "step": 72770 }, { "grad_norm": 0.1634775549173355, "learning_rate": 1.892663187010835e-05, "loss": 0.0143, "step": 72780 }, { "grad_norm": 0.15925265848636627, "learning_rate": 1.891367962459077e-05, "loss": 0.0102, "step": 72790 }, { "grad_norm": 0.14254838228225708, "learning_rate": 1.890073077862644e-05, "loss": 0.0102, "step": 72800 }, { "grad_norm": 0.14852501451969147, "learning_rate": 1.888778533363146e-05, "loss": 0.0102, "step": 72810 }, { "grad_norm": 0.13641659915447235, "learning_rate": 1.8874843291021478e-05, "loss": 0.0123, "step": 72820 }, { "grad_norm": 0.1543779969215393, "learning_rate": 1.8861904652211853e-05, "loss": 0.0116, "step": 72830 }, { "grad_norm": 0.12347957491874695, "learning_rate": 1.8848969418617507e-05, "loss": 0.0111, "step": 72840 }, { "grad_norm": 0.12203625589609146, "learning_rate": 1.8836037591653044e-05, "loss": 0.0102, "step": 72850 }, { "grad_norm": 0.13941526412963867, "learning_rate": 1.882310917273262e-05, "loss": 0.0133, "step": 72860 }, { "grad_norm": 0.165838360786438, "learning_rate": 1.881018416327014e-05, "loss": 0.0115, "step": 72870 }, { "grad_norm": 0.15864746272563934, "learning_rate": 1.8797262564679014e-05, "loss": 0.0121, "step": 72880 }, { "grad_norm": 0.1554974913597107, "learning_rate": 1.8784344378372355e-05, "loss": 0.0081, "step": 72890 }, { "grad_norm": 0.1688760370016098, "learning_rate": 1.8771429605762843e-05, "loss": 0.0128, "step": 72900 }, { "grad_norm": 0.16521525382995605, "learning_rate": 1.8758518248262835e-05, "loss": 0.0105, "step": 72910 }, { "grad_norm": 0.16201674938201904, "learning_rate": 1.8745610307284305e-05, "loss": 0.009, "step": 72920 }, { "grad_norm": 0.16187457740306854, "learning_rate": 1.8732705784238862e-05, "loss": 0.0128, "step": 72930 }, { "grad_norm": 0.14087505638599396, "learning_rate": 1.8719804680537682e-05, "loss": 0.0119, "step": 72940 }, { "grad_norm": 0.14658448100090027, "learning_rate": 1.870690699759164e-05, "loss": 0.0091, "step": 72950 }, { "grad_norm": 0.10168340057134628, "learning_rate": 1.8694012736811213e-05, "loss": 0.0129, "step": 72960 }, { "grad_norm": 0.19202087819576263, "learning_rate": 1.8681121899606463e-05, "loss": 0.0105, "step": 72970 }, { "grad_norm": 0.2010156512260437, "learning_rate": 1.8668234487387144e-05, "loss": 0.0111, "step": 72980 }, { "grad_norm": 0.12150047719478607, "learning_rate": 1.8655350501562592e-05, "loss": 0.0091, "step": 72990 }, { "grad_norm": 0.20329181849956512, "learning_rate": 1.8642469943541796e-05, "loss": 0.0139, "step": 73000 }, { "grad_norm": 0.21299663186073303, "learning_rate": 1.8629592814733328e-05, "loss": 0.0106, "step": 73010 }, { "grad_norm": 0.12999561429023743, "learning_rate": 1.8616719116545438e-05, "loss": 0.0104, "step": 73020 }, { "grad_norm": 0.12038201093673706, "learning_rate": 1.8603848850385952e-05, "loss": 0.0108, "step": 73030 }, { "grad_norm": 0.1554797887802124, "learning_rate": 1.859098201766236e-05, "loss": 0.0096, "step": 73040 }, { "grad_norm": 0.15810462832450867, "learning_rate": 1.8578118619781726e-05, "loss": 0.0101, "step": 73050 }, { "grad_norm": 0.11703695356845856, "learning_rate": 1.856525865815083e-05, "loss": 0.0103, "step": 73060 }, { "grad_norm": 0.1838141232728958, "learning_rate": 1.8552402134175966e-05, "loss": 0.0135, "step": 73070 }, { "grad_norm": 0.1823185235261917, "learning_rate": 1.8539549049263143e-05, "loss": 0.0109, "step": 73080 }, { "grad_norm": 0.1387137621641159, "learning_rate": 1.8526699404817917e-05, "loss": 0.0106, "step": 73090 }, { "grad_norm": 0.15442398190498352, "learning_rate": 1.8513853202245546e-05, "loss": 0.0116, "step": 73100 }, { "grad_norm": 0.20041075348854065, "learning_rate": 1.850101044295081e-05, "loss": 0.0089, "step": 73110 }, { "grad_norm": 0.11031949520111084, "learning_rate": 1.848817112833825e-05, "loss": 0.009, "step": 73120 }, { "grad_norm": 0.1092509850859642, "learning_rate": 1.8475335259811905e-05, "loss": 0.0112, "step": 73130 }, { "grad_norm": 0.18928569555282593, "learning_rate": 1.8462502838775508e-05, "loss": 0.0098, "step": 73140 }, { "grad_norm": 0.1378774344921112, "learning_rate": 1.8449673866632378e-05, "loss": 0.0135, "step": 73150 }, { "grad_norm": 0.1534431427717209, "learning_rate": 1.8436848344785473e-05, "loss": 0.0124, "step": 73160 }, { "grad_norm": 0.20149265229701996, "learning_rate": 1.8424026274637378e-05, "loss": 0.0142, "step": 73170 }, { "grad_norm": 0.1597488671541214, "learning_rate": 1.8411207657590312e-05, "loss": 0.0142, "step": 73180 }, { "grad_norm": 0.15828363597393036, "learning_rate": 1.839839249504607e-05, "loss": 0.0134, "step": 73190 }, { "grad_norm": 0.08721748739480972, "learning_rate": 1.8385580788406108e-05, "loss": 0.0103, "step": 73200 }, { "grad_norm": 0.13269320130348206, "learning_rate": 1.837277253907152e-05, "loss": 0.0129, "step": 73210 }, { "grad_norm": 0.13217604160308838, "learning_rate": 1.8359967748442957e-05, "loss": 0.0128, "step": 73220 }, { "grad_norm": 0.13793499767780304, "learning_rate": 1.8347166417920774e-05, "loss": 0.0143, "step": 73230 }, { "grad_norm": 0.1043131873011589, "learning_rate": 1.833436854890484e-05, "loss": 0.0109, "step": 73240 }, { "grad_norm": 0.17670780420303345, "learning_rate": 1.8321574142794796e-05, "loss": 0.011, "step": 73250 }, { "grad_norm": 0.11628113687038422, "learning_rate": 1.8308783200989748e-05, "loss": 0.0098, "step": 73260 }, { "grad_norm": 0.19024913012981415, "learning_rate": 1.829599572488855e-05, "loss": 0.0113, "step": 73270 }, { "grad_norm": 0.14488396048545837, "learning_rate": 1.828321171588957e-05, "loss": 0.01, "step": 73280 }, { "grad_norm": 0.17740853130817413, "learning_rate": 1.827043117539089e-05, "loss": 0.0138, "step": 73290 }, { "grad_norm": 0.1253228634595871, "learning_rate": 1.8257654104790116e-05, "loss": 0.012, "step": 73300 }, { "grad_norm": 0.15796777606010437, "learning_rate": 1.82448805054846e-05, "loss": 0.0083, "step": 73310 }, { "grad_norm": 0.13644076883792877, "learning_rate": 1.823211037887119e-05, "loss": 0.0111, "step": 73320 }, { "grad_norm": 0.14943520724773407, "learning_rate": 1.821934372634645e-05, "loss": 0.0096, "step": 73330 }, { "grad_norm": 0.15893715620040894, "learning_rate": 1.8206580549306472e-05, "loss": 0.0121, "step": 73340 }, { "grad_norm": 0.1624613255262375, "learning_rate": 1.819382084914707e-05, "loss": 0.0133, "step": 73350 }, { "grad_norm": 0.18727561831474304, "learning_rate": 1.8181064627263577e-05, "loss": 0.0108, "step": 73360 }, { "grad_norm": 0.18637049198150635, "learning_rate": 1.8168311885051022e-05, "loss": 0.0125, "step": 73370 }, { "grad_norm": 0.13843092322349548, "learning_rate": 1.815556262390402e-05, "loss": 0.0136, "step": 73380 }, { "grad_norm": 0.16243910789489746, "learning_rate": 1.8142816845216826e-05, "loss": 0.0114, "step": 73390 }, { "grad_norm": 0.15322069823741913, "learning_rate": 1.8130074550383276e-05, "loss": 0.0095, "step": 73400 }, { "grad_norm": 0.19828976690769196, "learning_rate": 1.8117335740796848e-05, "loss": 0.0093, "step": 73410 }, { "grad_norm": 0.14382264018058777, "learning_rate": 1.810460041785067e-05, "loss": 0.0119, "step": 73420 }, { "grad_norm": 0.12280623614788055, "learning_rate": 1.809186858293742e-05, "loss": 0.0093, "step": 73430 }, { "grad_norm": 0.134505033493042, "learning_rate": 1.8079140237449448e-05, "loss": 0.0118, "step": 73440 }, { "grad_norm": 0.1581626832485199, "learning_rate": 1.8066415382778707e-05, "loss": 0.0122, "step": 73450 }, { "grad_norm": 0.16529247164726257, "learning_rate": 1.8053694020316786e-05, "loss": 0.0137, "step": 73460 }, { "grad_norm": 0.13423249125480652, "learning_rate": 1.8040976151454837e-05, "loss": 0.0095, "step": 73470 }, { "grad_norm": 0.2683040499687195, "learning_rate": 1.8028261777583704e-05, "loss": 0.0116, "step": 73480 }, { "grad_norm": 0.1595821976661682, "learning_rate": 1.8015550900093758e-05, "loss": 0.0123, "step": 73490 }, { "grad_norm": 0.13752567768096924, "learning_rate": 1.800284352037512e-05, "loss": 0.0123, "step": 73500 }, { "grad_norm": 0.11200538277626038, "learning_rate": 1.799013963981738e-05, "loss": 0.0117, "step": 73510 }, { "grad_norm": 0.18097764253616333, "learning_rate": 1.797743925980987e-05, "loss": 0.0134, "step": 73520 }, { "grad_norm": 0.12708766758441925, "learning_rate": 1.7964742381741435e-05, "loss": 0.0082, "step": 73530 }, { "grad_norm": 0.13465791940689087, "learning_rate": 1.7952049007000627e-05, "loss": 0.0102, "step": 73540 }, { "grad_norm": 0.13584783673286438, "learning_rate": 1.7939359136975535e-05, "loss": 0.0098, "step": 73550 }, { "grad_norm": 0.19787341356277466, "learning_rate": 1.7926672773053932e-05, "loss": 0.0111, "step": 73560 }, { "grad_norm": 0.1355096846818924, "learning_rate": 1.791398991662317e-05, "loss": 0.0098, "step": 73570 }, { "grad_norm": 0.08023227006196976, "learning_rate": 1.7901310569070244e-05, "loss": 0.0086, "step": 73580 }, { "grad_norm": 0.09377270936965942, "learning_rate": 1.7888634731781724e-05, "loss": 0.0119, "step": 73590 }, { "grad_norm": 0.246269091963768, "learning_rate": 1.7875962406143844e-05, "loss": 0.0144, "step": 73600 }, { "grad_norm": 0.21618284285068512, "learning_rate": 1.7863293593542403e-05, "loss": 0.0115, "step": 73610 }, { "grad_norm": 0.14631062746047974, "learning_rate": 1.7850628295362853e-05, "loss": 0.01, "step": 73620 }, { "grad_norm": 0.11059578508138657, "learning_rate": 1.7837966512990263e-05, "loss": 0.0114, "step": 73630 }, { "grad_norm": 0.1864904761314392, "learning_rate": 1.782530824780931e-05, "loss": 0.0114, "step": 73640 }, { "grad_norm": 0.15399591624736786, "learning_rate": 1.7812653501204263e-05, "loss": 0.0114, "step": 73650 }, { "grad_norm": 0.1746908277273178, "learning_rate": 1.780000227455903e-05, "loss": 0.0124, "step": 73660 }, { "grad_norm": 0.30899426341056824, "learning_rate": 1.778735456925716e-05, "loss": 0.0106, "step": 73670 }, { "grad_norm": 0.1726117581129074, "learning_rate": 1.7774710386681737e-05, "loss": 0.0125, "step": 73680 }, { "grad_norm": 0.17117945849895477, "learning_rate": 1.7762069728215545e-05, "loss": 0.0128, "step": 73690 }, { "grad_norm": 0.14250394701957703, "learning_rate": 1.7749432595240932e-05, "loss": 0.011, "step": 73700 }, { "grad_norm": 0.10386321693658829, "learning_rate": 1.773679898913991e-05, "loss": 0.012, "step": 73710 }, { "grad_norm": 0.13634900748729706, "learning_rate": 1.7724168911294015e-05, "loss": 0.0119, "step": 73720 }, { "grad_norm": 0.18725664913654327, "learning_rate": 1.7711542363084504e-05, "loss": 0.0112, "step": 73730 }, { "grad_norm": 0.10913372039794922, "learning_rate": 1.7698919345892158e-05, "loss": 0.0106, "step": 73740 }, { "grad_norm": 0.20209413766860962, "learning_rate": 1.768629986109743e-05, "loss": 0.0108, "step": 73750 }, { "grad_norm": 0.23268185555934906, "learning_rate": 1.7673683910080357e-05, "loss": 0.0134, "step": 73760 }, { "grad_norm": 0.21848861873149872, "learning_rate": 1.7661071494220638e-05, "loss": 0.0096, "step": 73770 }, { "grad_norm": 0.13508443534374237, "learning_rate": 1.7648462614897498e-05, "loss": 0.0115, "step": 73780 }, { "grad_norm": 0.20423102378845215, "learning_rate": 1.7635857273489863e-05, "loss": 0.0107, "step": 73790 }, { "grad_norm": 0.13899780809879303, "learning_rate": 1.7623255471376198e-05, "loss": 0.0109, "step": 73800 }, { "grad_norm": 0.17792096734046936, "learning_rate": 1.7610657209934646e-05, "loss": 0.014, "step": 73810 }, { "grad_norm": 0.203928142786026, "learning_rate": 1.7598062490542915e-05, "loss": 0.0119, "step": 73820 }, { "grad_norm": 0.1305261105298996, "learning_rate": 1.7585471314578377e-05, "loss": 0.01, "step": 73830 }, { "grad_norm": 0.14888733625411987, "learning_rate": 1.7572883683417945e-05, "loss": 0.0111, "step": 73840 }, { "grad_norm": 0.1655118316411972, "learning_rate": 1.7560299598438212e-05, "loss": 0.0117, "step": 73850 }, { "grad_norm": 0.11023671180009842, "learning_rate": 1.7547719061015334e-05, "loss": 0.0102, "step": 73860 }, { "grad_norm": 0.22648903727531433, "learning_rate": 1.75351420725251e-05, "loss": 0.0106, "step": 73870 }, { "grad_norm": 0.17904256284236908, "learning_rate": 1.7522568634342923e-05, "loss": 0.013, "step": 73880 }, { "grad_norm": 0.2489725947380066, "learning_rate": 1.7509998747843825e-05, "loss": 0.0133, "step": 73890 }, { "grad_norm": 0.13829538226127625, "learning_rate": 1.7497432414402403e-05, "loss": 0.0119, "step": 73900 }, { "grad_norm": 0.11803461611270905, "learning_rate": 1.7484869635392903e-05, "loss": 0.0133, "step": 73910 }, { "grad_norm": 0.14496925473213196, "learning_rate": 1.747231041218919e-05, "loss": 0.0152, "step": 73920 }, { "grad_norm": 0.16151194274425507, "learning_rate": 1.7459754746164692e-05, "loss": 0.0115, "step": 73930 }, { "grad_norm": 0.12420506030321121, "learning_rate": 1.744720263869249e-05, "loss": 0.013, "step": 73940 }, { "grad_norm": 0.1559583693742752, "learning_rate": 1.7434654091145263e-05, "loss": 0.0106, "step": 73950 }, { "grad_norm": 0.11320409923791885, "learning_rate": 1.742210910489532e-05, "loss": 0.0097, "step": 73960 }, { "grad_norm": 0.14827001094818115, "learning_rate": 1.7409567681314532e-05, "loss": 0.0103, "step": 73970 }, { "grad_norm": 0.14171601831912994, "learning_rate": 1.7397029821774434e-05, "loss": 0.0125, "step": 73980 }, { "grad_norm": 0.24197320640087128, "learning_rate": 1.7384495527646126e-05, "loss": 0.015, "step": 73990 }, { "grad_norm": 0.11799347400665283, "learning_rate": 1.737196480030035e-05, "loss": 0.0098, "step": 74000 }, { "grad_norm": 0.12714698910713196, "learning_rate": 1.7359437641107447e-05, "loss": 0.0115, "step": 74010 }, { "grad_norm": 0.13718213140964508, "learning_rate": 1.7346914051437387e-05, "loss": 0.0125, "step": 74020 }, { "grad_norm": 0.14681720733642578, "learning_rate": 1.733439403265969e-05, "loss": 0.0111, "step": 74030 }, { "grad_norm": 0.09961813688278198, "learning_rate": 1.732187758614357e-05, "loss": 0.0092, "step": 74040 }, { "grad_norm": 0.14449094235897064, "learning_rate": 1.730936471325777e-05, "loss": 0.01, "step": 74050 }, { "grad_norm": 0.11860525608062744, "learning_rate": 1.7296855415370718e-05, "loss": 0.0087, "step": 74060 }, { "grad_norm": 0.13673286139965057, "learning_rate": 1.728434969385035e-05, "loss": 0.0138, "step": 74070 }, { "grad_norm": 0.1833215206861496, "learning_rate": 1.7271847550064346e-05, "loss": 0.012, "step": 74080 }, { "grad_norm": 0.14658643305301666, "learning_rate": 1.725934898537987e-05, "loss": 0.0131, "step": 74090 }, { "grad_norm": 0.11364869028329849, "learning_rate": 1.7246854001163783e-05, "loss": 0.011, "step": 74100 }, { "grad_norm": 0.17445413768291473, "learning_rate": 1.7234362598782476e-05, "loss": 0.0118, "step": 74110 }, { "grad_norm": 0.18924544751644135, "learning_rate": 1.722187477960202e-05, "loss": 0.0131, "step": 74120 }, { "grad_norm": 0.11761307716369629, "learning_rate": 1.7209390544988057e-05, "loss": 0.0114, "step": 74130 }, { "grad_norm": 0.12977270781993866, "learning_rate": 1.7196909896305856e-05, "loss": 0.011, "step": 74140 }, { "grad_norm": 0.14639835059642792, "learning_rate": 1.718443283492026e-05, "loss": 0.0104, "step": 74150 }, { "grad_norm": 0.166448175907135, "learning_rate": 1.7171959362195746e-05, "loss": 0.0112, "step": 74160 }, { "grad_norm": 0.15599843859672546, "learning_rate": 1.715948947949642e-05, "loss": 0.01, "step": 74170 }, { "grad_norm": 0.17290008068084717, "learning_rate": 1.714702318818593e-05, "loss": 0.0138, "step": 74180 }, { "grad_norm": 0.11714658886194229, "learning_rate": 1.7134560489627592e-05, "loss": 0.0107, "step": 74190 }, { "grad_norm": 0.1724948287010193, "learning_rate": 1.712210138518431e-05, "loss": 0.0142, "step": 74200 }, { "grad_norm": 0.1807892620563507, "learning_rate": 1.71096458762186e-05, "loss": 0.0103, "step": 74210 }, { "grad_norm": 0.11430726945400238, "learning_rate": 1.7097193964092552e-05, "loss": 0.0086, "step": 74220 }, { "grad_norm": 0.1477169692516327, "learning_rate": 1.708474565016792e-05, "loss": 0.0094, "step": 74230 }, { "grad_norm": 0.19802793860435486, "learning_rate": 1.7072300935806e-05, "loss": 0.0102, "step": 74240 }, { "grad_norm": 0.13565264642238617, "learning_rate": 1.705985982236776e-05, "loss": 0.0107, "step": 74250 }, { "grad_norm": 0.12559665739536285, "learning_rate": 1.7047422311213697e-05, "loss": 0.0103, "step": 74260 }, { "grad_norm": 0.1251639872789383, "learning_rate": 1.703498840370401e-05, "loss": 0.0105, "step": 74270 }, { "grad_norm": 0.18577519059181213, "learning_rate": 1.702255810119842e-05, "loss": 0.0091, "step": 74280 }, { "grad_norm": 0.10356734693050385, "learning_rate": 1.7010131405056305e-05, "loss": 0.01, "step": 74290 }, { "grad_norm": 0.1560581475496292, "learning_rate": 1.6997708316636613e-05, "loss": 0.0106, "step": 74300 }, { "grad_norm": 0.16434165835380554, "learning_rate": 1.6985288837297932e-05, "loss": 0.0094, "step": 74310 }, { "grad_norm": 0.14080198109149933, "learning_rate": 1.6972872968398394e-05, "loss": 0.0109, "step": 74320 }, { "grad_norm": 0.18796922266483307, "learning_rate": 1.6960460711295845e-05, "loss": 0.0132, "step": 74330 }, { "grad_norm": 0.11277607828378677, "learning_rate": 1.6948052067347623e-05, "loss": 0.0114, "step": 74340 }, { "grad_norm": 0.14482998847961426, "learning_rate": 1.693564703791075e-05, "loss": 0.0141, "step": 74350 }, { "grad_norm": 0.10473153740167618, "learning_rate": 1.6923245624341793e-05, "loss": 0.0099, "step": 74360 }, { "grad_norm": 0.1495157778263092, "learning_rate": 1.6910847827996962e-05, "loss": 0.0144, "step": 74370 }, { "grad_norm": 0.15145432949066162, "learning_rate": 1.689845365023206e-05, "loss": 0.0143, "step": 74380 }, { "grad_norm": 0.1913229376077652, "learning_rate": 1.6886063092402505e-05, "loss": 0.0118, "step": 74390 }, { "grad_norm": 0.12370409071445465, "learning_rate": 1.6873676155863317e-05, "loss": 0.0107, "step": 74400 }, { "grad_norm": 0.13293671607971191, "learning_rate": 1.6861292841969085e-05, "loss": 0.0114, "step": 74410 }, { "grad_norm": 0.13291624188423157, "learning_rate": 1.684891315207406e-05, "loss": 0.0127, "step": 74420 }, { "grad_norm": 0.17029434442520142, "learning_rate": 1.6836537087532034e-05, "loss": 0.0107, "step": 74430 }, { "grad_norm": 0.18872162699699402, "learning_rate": 1.6824164649696465e-05, "loss": 0.0101, "step": 74440 }, { "grad_norm": 0.1579783409833908, "learning_rate": 1.681179583992034e-05, "loss": 0.0134, "step": 74450 }, { "grad_norm": 0.10505414009094238, "learning_rate": 1.679943065955636e-05, "loss": 0.0107, "step": 74460 }, { "grad_norm": 0.11462762951850891, "learning_rate": 1.678706910995671e-05, "loss": 0.0104, "step": 74470 }, { "grad_norm": 0.16703100502490997, "learning_rate": 1.6774711192473262e-05, "loss": 0.0107, "step": 74480 }, { "grad_norm": 0.16872113943099976, "learning_rate": 1.676235690845743e-05, "loss": 0.0101, "step": 74490 }, { "grad_norm": 0.08912501484155655, "learning_rate": 1.6750006259260285e-05, "loss": 0.0106, "step": 74500 }, { "grad_norm": 0.1293356716632843, "learning_rate": 1.673765924623244e-05, "loss": 0.0098, "step": 74510 }, { "grad_norm": 0.11585267633199692, "learning_rate": 1.67253158707242e-05, "loss": 0.0104, "step": 74520 }, { "grad_norm": 0.15792115032672882, "learning_rate": 1.671297613408537e-05, "loss": 0.0095, "step": 74530 }, { "grad_norm": 0.15222878754138947, "learning_rate": 1.6700640037665444e-05, "loss": 0.0102, "step": 74540 }, { "grad_norm": 0.15559101104736328, "learning_rate": 1.668830758281344e-05, "loss": 0.0108, "step": 74550 }, { "grad_norm": 0.17568275332450867, "learning_rate": 1.6675978770878052e-05, "loss": 0.0118, "step": 74560 }, { "grad_norm": 0.1495954394340515, "learning_rate": 1.6663653603207506e-05, "loss": 0.01, "step": 74570 }, { "grad_norm": 0.1534176468849182, "learning_rate": 1.6651332081149685e-05, "loss": 0.0099, "step": 74580 }, { "grad_norm": 0.11230984330177307, "learning_rate": 1.6639014206052046e-05, "loss": 0.0085, "step": 74590 }, { "grad_norm": 0.16415941715240479, "learning_rate": 1.6626699979261657e-05, "loss": 0.013, "step": 74600 }, { "grad_norm": 0.12111258506774902, "learning_rate": 1.6614389402125197e-05, "loss": 0.0087, "step": 74610 }, { "grad_norm": 0.11850562691688538, "learning_rate": 1.66020824759889e-05, "loss": 0.0106, "step": 74620 }, { "grad_norm": 0.1750117838382721, "learning_rate": 1.6589779202198675e-05, "loss": 0.0111, "step": 74630 }, { "grad_norm": 0.11750256270170212, "learning_rate": 1.6577479582099924e-05, "loss": 0.0162, "step": 74640 }, { "grad_norm": 0.2375081330537796, "learning_rate": 1.6565183617037788e-05, "loss": 0.013, "step": 74650 }, { "grad_norm": 0.17465244233608246, "learning_rate": 1.6552891308356894e-05, "loss": 0.0118, "step": 74660 }, { "grad_norm": 0.10883209854364395, "learning_rate": 1.6540602657401526e-05, "loss": 0.0115, "step": 74670 }, { "grad_norm": 0.13556605577468872, "learning_rate": 1.6528317665515537e-05, "loss": 0.0102, "step": 74680 }, { "grad_norm": 0.2272201031446457, "learning_rate": 1.6516036334042416e-05, "loss": 0.0119, "step": 74690 }, { "grad_norm": 0.1436888724565506, "learning_rate": 1.650375866432519e-05, "loss": 0.01, "step": 74700 }, { "grad_norm": 0.22228200733661652, "learning_rate": 1.649148465770659e-05, "loss": 0.0093, "step": 74710 }, { "grad_norm": 0.09215915203094482, "learning_rate": 1.647921431552883e-05, "loss": 0.0122, "step": 74720 }, { "grad_norm": 0.13100461661815643, "learning_rate": 1.646694763913381e-05, "loss": 0.0107, "step": 74730 }, { "grad_norm": 0.1506727635860443, "learning_rate": 1.6454684629862966e-05, "loss": 0.0094, "step": 74740 }, { "grad_norm": 0.1436164826154709, "learning_rate": 1.6442425289057388e-05, "loss": 0.0097, "step": 74750 }, { "grad_norm": 0.1384734958410263, "learning_rate": 1.6430169618057718e-05, "loss": 0.0108, "step": 74760 }, { "grad_norm": 0.14296218752861023, "learning_rate": 1.641791761820422e-05, "loss": 0.0114, "step": 74770 }, { "grad_norm": 0.10826430469751358, "learning_rate": 1.640566929083676e-05, "loss": 0.0109, "step": 74780 }, { "grad_norm": 0.11353977769613266, "learning_rate": 1.6393424637294823e-05, "loss": 0.0099, "step": 74790 }, { "grad_norm": 0.10439030826091766, "learning_rate": 1.638118365891742e-05, "loss": 0.0109, "step": 74800 }, { "grad_norm": 0.1275663524866104, "learning_rate": 1.636894635704323e-05, "loss": 0.0119, "step": 74810 }, { "grad_norm": 0.1311689019203186, "learning_rate": 1.6356712733010517e-05, "loss": 0.0101, "step": 74820 }, { "grad_norm": 0.09120731800794601, "learning_rate": 1.6344482788157096e-05, "loss": 0.0095, "step": 74830 }, { "grad_norm": 0.13722434639930725, "learning_rate": 1.6332256523820444e-05, "loss": 0.0109, "step": 74840 }, { "grad_norm": 0.09466767311096191, "learning_rate": 1.6320033941337597e-05, "loss": 0.0092, "step": 74850 }, { "grad_norm": 0.1407405585050583, "learning_rate": 1.6307815042045216e-05, "loss": 0.0087, "step": 74860 }, { "grad_norm": 0.20419317483901978, "learning_rate": 1.6295599827279513e-05, "loss": 0.0147, "step": 74870 }, { "grad_norm": 0.15731129050254822, "learning_rate": 1.6283388298376356e-05, "loss": 0.0106, "step": 74880 }, { "grad_norm": 0.1704588383436203, "learning_rate": 1.6271180456671125e-05, "loss": 0.0091, "step": 74890 }, { "grad_norm": 0.1705683320760727, "learning_rate": 1.625897630349893e-05, "loss": 0.0104, "step": 74900 }, { "grad_norm": 0.17642788589000702, "learning_rate": 1.6246775840194335e-05, "loss": 0.0142, "step": 74910 }, { "grad_norm": 0.16351859271526337, "learning_rate": 1.6234579068091604e-05, "loss": 0.0095, "step": 74920 }, { "grad_norm": 0.1403992921113968, "learning_rate": 1.6222385988524524e-05, "loss": 0.0092, "step": 74930 }, { "grad_norm": 0.15546739101409912, "learning_rate": 1.6210196602826544e-05, "loss": 0.0128, "step": 74940 }, { "grad_norm": 0.14269597828388214, "learning_rate": 1.6198010912330646e-05, "loss": 0.0129, "step": 74950 }, { "grad_norm": 0.13975556194782257, "learning_rate": 1.6185828918369456e-05, "loss": 0.0105, "step": 74960 }, { "grad_norm": 0.18602149188518524, "learning_rate": 1.617365062227517e-05, "loss": 0.0107, "step": 74970 }, { "grad_norm": 0.15580882132053375, "learning_rate": 1.616147602537962e-05, "loss": 0.0109, "step": 74980 }, { "grad_norm": 0.1338481307029724, "learning_rate": 1.614930512901415e-05, "loss": 0.0106, "step": 74990 }, { "grad_norm": 0.12552136182785034, "learning_rate": 1.6137137934509804e-05, "loss": 0.0128, "step": 75000 }, { "grad_norm": 0.15904617309570312, "learning_rate": 1.6124974443197123e-05, "loss": 0.0117, "step": 75010 }, { "grad_norm": 0.1235537901520729, "learning_rate": 1.6112814656406306e-05, "loss": 0.01, "step": 75020 }, { "grad_norm": 0.17006835341453552, "learning_rate": 1.6100658575467125e-05, "loss": 0.0156, "step": 75030 }, { "grad_norm": 0.10480855405330658, "learning_rate": 1.6088506201708974e-05, "loss": 0.0097, "step": 75040 }, { "grad_norm": 0.15823645889759064, "learning_rate": 1.6076357536460785e-05, "loss": 0.0091, "step": 75050 }, { "grad_norm": 0.16071346402168274, "learning_rate": 1.606421258105113e-05, "loss": 0.0112, "step": 75060 }, { "grad_norm": 0.10960482060909271, "learning_rate": 1.6052071336808183e-05, "loss": 0.011, "step": 75070 }, { "grad_norm": 0.17298485338687897, "learning_rate": 1.6039933805059658e-05, "loss": 0.0091, "step": 75080 }, { "grad_norm": 0.15051379799842834, "learning_rate": 1.6027799987132915e-05, "loss": 0.0114, "step": 75090 }, { "grad_norm": 0.18997058272361755, "learning_rate": 1.6015669884354888e-05, "loss": 0.0098, "step": 75100 }, { "grad_norm": 0.11192119866609573, "learning_rate": 1.6003543498052125e-05, "loss": 0.0101, "step": 75110 }, { "grad_norm": 0.15027788281440735, "learning_rate": 1.5991420829550715e-05, "loss": 0.0116, "step": 75120 }, { "grad_norm": 0.14365044236183167, "learning_rate": 1.5979301880176405e-05, "loss": 0.0137, "step": 75130 }, { "grad_norm": 0.13874530792236328, "learning_rate": 1.5967186651254485e-05, "loss": 0.0125, "step": 75140 }, { "grad_norm": 0.19711019098758698, "learning_rate": 1.5955075144109853e-05, "loss": 0.0082, "step": 75150 }, { "grad_norm": 0.12995672225952148, "learning_rate": 1.594296736006703e-05, "loss": 0.0116, "step": 75160 }, { "grad_norm": 0.09679406136274338, "learning_rate": 1.59308633004501e-05, "loss": 0.0087, "step": 75170 }, { "grad_norm": 0.15062299370765686, "learning_rate": 1.5918762966582724e-05, "loss": 0.0113, "step": 75180 }, { "grad_norm": 0.1528618335723877, "learning_rate": 1.5906666359788203e-05, "loss": 0.0085, "step": 75190 }, { "grad_norm": 0.16419212520122528, "learning_rate": 1.589457348138938e-05, "loss": 0.009, "step": 75200 }, { "grad_norm": 0.13747569918632507, "learning_rate": 1.588248433270872e-05, "loss": 0.0093, "step": 75210 }, { "grad_norm": 0.12923642992973328, "learning_rate": 1.587039891506828e-05, "loss": 0.0121, "step": 75220 }, { "grad_norm": 0.17078974843025208, "learning_rate": 1.5858317229789715e-05, "loss": 0.0095, "step": 75230 }, { "grad_norm": 0.13770286738872528, "learning_rate": 1.584623927819423e-05, "loss": 0.0093, "step": 75240 }, { "grad_norm": 0.15899641811847687, "learning_rate": 1.583416506160269e-05, "loss": 0.011, "step": 75250 }, { "grad_norm": 0.11765195429325104, "learning_rate": 1.5822094581335472e-05, "loss": 0.0088, "step": 75260 }, { "grad_norm": 0.15909962356090546, "learning_rate": 1.5810027838712604e-05, "loss": 0.0097, "step": 75270 }, { "grad_norm": 0.13198497891426086, "learning_rate": 1.5797964835053687e-05, "loss": 0.01, "step": 75280 }, { "grad_norm": 0.1479368507862091, "learning_rate": 1.5785905571677933e-05, "loss": 0.012, "step": 75290 }, { "grad_norm": 0.18381039798259735, "learning_rate": 1.5773850049904086e-05, "loss": 0.0126, "step": 75300 }, { "grad_norm": 0.1786906123161316, "learning_rate": 1.5761798271050543e-05, "loss": 0.0125, "step": 75310 }, { "grad_norm": 0.1277237981557846, "learning_rate": 1.5749750236435277e-05, "loss": 0.0105, "step": 75320 }, { "grad_norm": 0.21799463033676147, "learning_rate": 1.5737705947375815e-05, "loss": 0.012, "step": 75330 }, { "grad_norm": 0.13336770236492157, "learning_rate": 1.5725665405189322e-05, "loss": 0.0116, "step": 75340 }, { "grad_norm": 0.16029399633407593, "learning_rate": 1.571362861119253e-05, "loss": 0.0117, "step": 75350 }, { "grad_norm": 0.1585446149110794, "learning_rate": 1.570159556670177e-05, "loss": 0.014, "step": 75360 }, { "grad_norm": 0.1736534982919693, "learning_rate": 1.5689566273032946e-05, "loss": 0.0133, "step": 75370 }, { "grad_norm": 0.21735836565494537, "learning_rate": 1.5677540731501583e-05, "loss": 0.014, "step": 75380 }, { "grad_norm": 0.18454721570014954, "learning_rate": 1.5665518943422746e-05, "loss": 0.0117, "step": 75390 }, { "grad_norm": 0.13994397222995758, "learning_rate": 1.5653500910111134e-05, "loss": 0.0091, "step": 75400 }, { "grad_norm": 0.17702434957027435, "learning_rate": 1.5641486632881023e-05, "loss": 0.0105, "step": 75410 }, { "grad_norm": 0.33011600375175476, "learning_rate": 1.56294761130463e-05, "loss": 0.0114, "step": 75420 }, { "grad_norm": 0.09340998530387878, "learning_rate": 1.5617469351920373e-05, "loss": 0.0103, "step": 75430 }, { "grad_norm": 0.1611890345811844, "learning_rate": 1.5605466350816315e-05, "loss": 0.0102, "step": 75440 }, { "grad_norm": 0.2803367078304291, "learning_rate": 1.559346711104674e-05, "loss": 0.0115, "step": 75450 }, { "grad_norm": 0.10408613830804825, "learning_rate": 1.5581471633923883e-05, "loss": 0.0102, "step": 75460 }, { "grad_norm": 0.12144482135772705, "learning_rate": 1.5569479920759505e-05, "loss": 0.0093, "step": 75470 }, { "grad_norm": 0.12961742281913757, "learning_rate": 1.5557491972865072e-05, "loss": 0.0106, "step": 75480 }, { "grad_norm": 0.16677498817443848, "learning_rate": 1.554550779155152e-05, "loss": 0.0122, "step": 75490 }, { "grad_norm": 0.26487329602241516, "learning_rate": 1.5533527378129453e-05, "loss": 0.009, "step": 75500 }, { "grad_norm": 0.1592729389667511, "learning_rate": 1.552155073390899e-05, "loss": 0.0111, "step": 75510 }, { "grad_norm": 0.16034702956676483, "learning_rate": 1.5509577860199903e-05, "loss": 0.0104, "step": 75520 }, { "grad_norm": 0.14467774331569672, "learning_rate": 1.5497608758311525e-05, "loss": 0.0094, "step": 75530 }, { "grad_norm": 0.21206186711788177, "learning_rate": 1.5485643429552797e-05, "loss": 0.0102, "step": 75540 }, { "grad_norm": 0.13112026453018188, "learning_rate": 1.5473681875232197e-05, "loss": 0.0108, "step": 75550 }, { "grad_norm": 0.1466754823923111, "learning_rate": 1.5461724096657837e-05, "loss": 0.0118, "step": 75560 }, { "grad_norm": 0.14656832814216614, "learning_rate": 1.5449770095137416e-05, "loss": 0.0112, "step": 75570 }, { "grad_norm": 0.14864113926887512, "learning_rate": 1.5437819871978178e-05, "loss": 0.0091, "step": 75580 }, { "grad_norm": 0.07724589854478836, "learning_rate": 1.5425873428486992e-05, "loss": 0.0098, "step": 75590 }, { "grad_norm": 0.19123756885528564, "learning_rate": 1.5413930765970312e-05, "loss": 0.0094, "step": 75600 }, { "grad_norm": 0.19503387808799744, "learning_rate": 1.5401991885734175e-05, "loss": 0.011, "step": 75610 }, { "grad_norm": 0.16525906324386597, "learning_rate": 1.5390056789084174e-05, "loss": 0.0104, "step": 75620 }, { "grad_norm": 0.13807864487171173, "learning_rate": 1.5378125477325545e-05, "loss": 0.0106, "step": 75630 }, { "grad_norm": 0.18167047202587128, "learning_rate": 1.5366197951763046e-05, "loss": 0.0125, "step": 75640 }, { "grad_norm": 0.14539898931980133, "learning_rate": 1.5354274213701075e-05, "loss": 0.0083, "step": 75650 }, { "grad_norm": 0.13144418597221375, "learning_rate": 1.5342354264443566e-05, "loss": 0.0092, "step": 75660 }, { "grad_norm": 0.1395578682422638, "learning_rate": 1.533043810529411e-05, "loss": 0.0101, "step": 75670 }, { "grad_norm": 0.1955375373363495, "learning_rate": 1.5318525737555807e-05, "loss": 0.0116, "step": 75680 }, { "grad_norm": 0.1341925710439682, "learning_rate": 1.53066171625314e-05, "loss": 0.0115, "step": 75690 }, { "grad_norm": 0.15165138244628906, "learning_rate": 1.529471238152317e-05, "loss": 0.0107, "step": 75700 }, { "grad_norm": 0.14971452951431274, "learning_rate": 1.5282811395833023e-05, "loss": 0.0088, "step": 75710 }, { "grad_norm": 0.14882229268550873, "learning_rate": 1.5270914206762397e-05, "loss": 0.0099, "step": 75720 }, { "grad_norm": 0.13520722091197968, "learning_rate": 1.525902081561241e-05, "loss": 0.0095, "step": 75730 }, { "grad_norm": 0.1171175017952919, "learning_rate": 1.5247131223683664e-05, "loss": 0.0095, "step": 75740 }, { "grad_norm": 0.18637672066688538, "learning_rate": 1.5235245432276408e-05, "loss": 0.011, "step": 75750 }, { "grad_norm": 0.10202010720968246, "learning_rate": 1.5223363442690424e-05, "loss": 0.0101, "step": 75760 }, { "grad_norm": 0.09697317332029343, "learning_rate": 1.5211485256225133e-05, "loss": 0.0091, "step": 75770 }, { "grad_norm": 0.11289657652378082, "learning_rate": 1.519961087417951e-05, "loss": 0.0104, "step": 75780 }, { "grad_norm": 0.12230592221021652, "learning_rate": 1.5187740297852137e-05, "loss": 0.0085, "step": 75790 }, { "grad_norm": 0.17455962300300598, "learning_rate": 1.517587352854113e-05, "loss": 0.0107, "step": 75800 }, { "grad_norm": 0.16252605617046356, "learning_rate": 1.516401056754423e-05, "loss": 0.0095, "step": 75810 }, { "grad_norm": 0.1486048549413681, "learning_rate": 1.5152151416158783e-05, "loss": 0.009, "step": 75820 }, { "grad_norm": 0.16630196571350098, "learning_rate": 1.5140296075681642e-05, "loss": 0.0104, "step": 75830 }, { "grad_norm": 0.2141132354736328, "learning_rate": 1.5128444547409331e-05, "loss": 0.0106, "step": 75840 }, { "grad_norm": 0.10863767564296722, "learning_rate": 1.5116596832637864e-05, "loss": 0.0105, "step": 75850 }, { "grad_norm": 0.16355907917022705, "learning_rate": 1.5104752932662952e-05, "loss": 0.0091, "step": 75860 }, { "grad_norm": 0.12143105268478394, "learning_rate": 1.5092912848779778e-05, "loss": 0.0088, "step": 75870 }, { "grad_norm": 0.13483013212680817, "learning_rate": 1.5081076582283194e-05, "loss": 0.0114, "step": 75880 }, { "grad_norm": 0.16430158913135529, "learning_rate": 1.5069244134467553e-05, "loss": 0.0105, "step": 75890 }, { "grad_norm": 0.15925511717796326, "learning_rate": 1.5057415506626876e-05, "loss": 0.0118, "step": 75900 }, { "grad_norm": 0.13236568868160248, "learning_rate": 1.5045590700054668e-05, "loss": 0.0111, "step": 75910 }, { "grad_norm": 0.11750360578298569, "learning_rate": 1.5033769716044144e-05, "loss": 0.0103, "step": 75920 }, { "grad_norm": 0.13523484766483307, "learning_rate": 1.5021952555887975e-05, "loss": 0.0123, "step": 75930 }, { "grad_norm": 0.4640519618988037, "learning_rate": 1.50101392208785e-05, "loss": 0.0113, "step": 75940 }, { "grad_norm": 0.15223106741905212, "learning_rate": 1.4998329712307579e-05, "loss": 0.0109, "step": 75950 }, { "grad_norm": 0.16226352751255035, "learning_rate": 1.4986524031466708e-05, "loss": 0.0112, "step": 75960 }, { "grad_norm": 0.1578119993209839, "learning_rate": 1.4974722179646916e-05, "loss": 0.0084, "step": 75970 }, { "grad_norm": 0.27390673756599426, "learning_rate": 1.4962924158138847e-05, "loss": 0.0113, "step": 75980 }, { "grad_norm": 0.20679780840873718, "learning_rate": 1.4951129968232714e-05, "loss": 0.014, "step": 75990 }, { "grad_norm": 0.16370920836925507, "learning_rate": 1.4939339611218329e-05, "loss": 0.0135, "step": 76000 }, { "grad_norm": 0.15709547698497772, "learning_rate": 1.4927553088385032e-05, "loss": 0.0105, "step": 76010 }, { "grad_norm": 0.17227481305599213, "learning_rate": 1.4915770401021801e-05, "loss": 0.0113, "step": 76020 }, { "grad_norm": 0.14866571128368378, "learning_rate": 1.4903991550417185e-05, "loss": 0.0095, "step": 76030 }, { "grad_norm": 0.17482054233551025, "learning_rate": 1.4892216537859277e-05, "loss": 0.0123, "step": 76040 }, { "grad_norm": 0.15030768513679504, "learning_rate": 1.4880445364635775e-05, "loss": 0.01, "step": 76050 }, { "grad_norm": 0.2052537500858307, "learning_rate": 1.4868678032033973e-05, "loss": 0.0089, "step": 76060 }, { "grad_norm": 0.15362614393234253, "learning_rate": 1.4856914541340739e-05, "loss": 0.0108, "step": 76070 }, { "grad_norm": 0.10935839265584946, "learning_rate": 1.4845154893842472e-05, "loss": 0.0083, "step": 76080 }, { "grad_norm": 0.1418595314025879, "learning_rate": 1.4833399090825234e-05, "loss": 0.0108, "step": 76090 }, { "grad_norm": 0.21016913652420044, "learning_rate": 1.4821647133574562e-05, "loss": 0.0155, "step": 76100 }, { "grad_norm": 0.14095763862133026, "learning_rate": 1.4809899023375706e-05, "loss": 0.0111, "step": 76110 }, { "grad_norm": 0.12183304876089096, "learning_rate": 1.4798154761513366e-05, "loss": 0.0096, "step": 76120 }, { "grad_norm": 0.14647795259952545, "learning_rate": 1.4786414349271916e-05, "loss": 0.0093, "step": 76130 }, { "grad_norm": 0.09396984428167343, "learning_rate": 1.4774677787935226e-05, "loss": 0.0086, "step": 76140 }, { "grad_norm": 0.13996919989585876, "learning_rate": 1.4762945078786838e-05, "loss": 0.0101, "step": 76150 }, { "grad_norm": 0.15732859075069427, "learning_rate": 1.4751216223109777e-05, "loss": 0.0114, "step": 76160 }, { "grad_norm": 0.20304162800312042, "learning_rate": 1.4739491222186714e-05, "loss": 0.0129, "step": 76170 }, { "grad_norm": 0.14754366874694824, "learning_rate": 1.4727770077299879e-05, "loss": 0.0109, "step": 76180 }, { "grad_norm": 0.25041434168815613, "learning_rate": 1.4716052789731094e-05, "loss": 0.0134, "step": 76190 }, { "grad_norm": 0.22893200814723969, "learning_rate": 1.470433936076171e-05, "loss": 0.0115, "step": 76200 }, { "grad_norm": 0.19080132246017456, "learning_rate": 1.4692629791672719e-05, "loss": 0.0125, "step": 76210 }, { "grad_norm": 0.1181035041809082, "learning_rate": 1.468092408374464e-05, "loss": 0.0093, "step": 76220 }, { "grad_norm": 0.13145668804645538, "learning_rate": 1.4669222238257602e-05, "loss": 0.012, "step": 76230 }, { "grad_norm": 0.12070871144533157, "learning_rate": 1.4657524256491301e-05, "loss": 0.0113, "step": 76240 }, { "grad_norm": 0.14820632338523865, "learning_rate": 1.464583013972502e-05, "loss": 0.0112, "step": 76250 }, { "grad_norm": 0.14320029318332672, "learning_rate": 1.4634139889237591e-05, "loss": 0.0128, "step": 76260 }, { "grad_norm": 0.11097417771816254, "learning_rate": 1.4622453506307448e-05, "loss": 0.01, "step": 76270 }, { "grad_norm": 0.11712288856506348, "learning_rate": 1.461077099221262e-05, "loss": 0.0107, "step": 76280 }, { "grad_norm": 0.14975175261497498, "learning_rate": 1.4599092348230652e-05, "loss": 0.0119, "step": 76290 }, { "grad_norm": 0.1625574380159378, "learning_rate": 1.4587417575638724e-05, "loss": 0.009, "step": 76300 }, { "grad_norm": 0.20186467468738556, "learning_rate": 1.4575746675713563e-05, "loss": 0.0105, "step": 76310 }, { "grad_norm": 0.1001972109079361, "learning_rate": 1.4564079649731504e-05, "loss": 0.0099, "step": 76320 }, { "grad_norm": 0.141641303896904, "learning_rate": 1.4552416498968402e-05, "loss": 0.0105, "step": 76330 }, { "grad_norm": 0.13940836489200592, "learning_rate": 1.4540757224699747e-05, "loss": 0.0129, "step": 76340 }, { "grad_norm": 0.1330864131450653, "learning_rate": 1.452910182820056e-05, "loss": 0.0082, "step": 76350 }, { "grad_norm": 0.13162755966186523, "learning_rate": 1.4517450310745468e-05, "loss": 0.0102, "step": 76360 }, { "grad_norm": 0.12955766916275024, "learning_rate": 1.450580267360866e-05, "loss": 0.0104, "step": 76370 }, { "grad_norm": 0.19487421214580536, "learning_rate": 1.4494158918063922e-05, "loss": 0.0104, "step": 76380 }, { "grad_norm": 0.09178996831178665, "learning_rate": 1.4482519045384568e-05, "loss": 0.0099, "step": 76390 }, { "grad_norm": 0.13645395636558533, "learning_rate": 1.4470883056843548e-05, "loss": 0.0092, "step": 76400 }, { "grad_norm": 0.13379691541194916, "learning_rate": 1.4459250953713322e-05, "loss": 0.0084, "step": 76410 }, { "grad_norm": 0.13970330357551575, "learning_rate": 1.444762273726598e-05, "loss": 0.01, "step": 76420 }, { "grad_norm": 0.19233663380146027, "learning_rate": 1.443599840877316e-05, "loss": 0.008, "step": 76430 }, { "grad_norm": 0.20008721947669983, "learning_rate": 1.4424377969506103e-05, "loss": 0.0111, "step": 76440 }, { "grad_norm": 0.13936221599578857, "learning_rate": 1.4412761420735565e-05, "loss": 0.0155, "step": 76450 }, { "grad_norm": 0.08965540677309036, "learning_rate": 1.4401148763731953e-05, "loss": 0.01, "step": 76460 }, { "grad_norm": 0.1583249419927597, "learning_rate": 1.4389539999765173e-05, "loss": 0.0116, "step": 76470 }, { "grad_norm": 0.1425808072090149, "learning_rate": 1.4377935130104758e-05, "loss": 0.008, "step": 76480 }, { "grad_norm": 0.13754087686538696, "learning_rate": 1.4366334156019801e-05, "loss": 0.0136, "step": 76490 }, { "grad_norm": 0.17334719002246857, "learning_rate": 1.435473707877898e-05, "loss": 0.0098, "step": 76500 }, { "grad_norm": 0.11303837597370148, "learning_rate": 1.4343143899650508e-05, "loss": 0.012, "step": 76510 }, { "grad_norm": 0.1476972997188568, "learning_rate": 1.4331554619902204e-05, "loss": 0.0109, "step": 76520 }, { "grad_norm": 0.13360555469989777, "learning_rate": 1.4319969240801484e-05, "loss": 0.0105, "step": 76530 }, { "grad_norm": 0.11433190107345581, "learning_rate": 1.430838776361526e-05, "loss": 0.0101, "step": 76540 }, { "grad_norm": 0.17553187906742096, "learning_rate": 1.4296810189610093e-05, "loss": 0.0114, "step": 76550 }, { "grad_norm": 0.13102200627326965, "learning_rate": 1.4285236520052087e-05, "loss": 0.01, "step": 76560 }, { "grad_norm": 0.11971690505743027, "learning_rate": 1.4273666756206933e-05, "loss": 0.0102, "step": 76570 }, { "grad_norm": 0.16038818657398224, "learning_rate": 1.4262100899339848e-05, "loss": 0.0119, "step": 76580 }, { "grad_norm": 0.13092118501663208, "learning_rate": 1.4250538950715696e-05, "loss": 0.0099, "step": 76590 }, { "grad_norm": 0.09144463390111923, "learning_rate": 1.4238980911598842e-05, "loss": 0.0098, "step": 76600 }, { "grad_norm": 0.14553730189800262, "learning_rate": 1.4227426783253267e-05, "loss": 0.0097, "step": 76610 }, { "grad_norm": 0.13787931203842163, "learning_rate": 1.4215876566942521e-05, "loss": 0.0106, "step": 76620 }, { "grad_norm": 0.09395848959684372, "learning_rate": 1.4204330263929721e-05, "loss": 0.0097, "step": 76630 }, { "grad_norm": 0.10750409960746765, "learning_rate": 1.4192787875477537e-05, "loss": 0.0102, "step": 76640 }, { "grad_norm": 0.10799595713615417, "learning_rate": 1.4181249402848246e-05, "loss": 0.0115, "step": 76650 }, { "grad_norm": 0.11273815482854843, "learning_rate": 1.4169714847303656e-05, "loss": 0.0103, "step": 76660 }, { "grad_norm": 0.2146148830652237, "learning_rate": 1.4158184210105197e-05, "loss": 0.0187, "step": 76670 }, { "grad_norm": 0.12563692033290863, "learning_rate": 1.4146657492513787e-05, "loss": 0.0101, "step": 76680 }, { "grad_norm": 0.10762445628643036, "learning_rate": 1.4135134695790047e-05, "loss": 0.0102, "step": 76690 }, { "grad_norm": 0.20461487770080566, "learning_rate": 1.4123615821194037e-05, "loss": 0.0109, "step": 76700 }, { "grad_norm": 0.12025585770606995, "learning_rate": 1.4112100869985473e-05, "loss": 0.0093, "step": 76710 }, { "grad_norm": 0.15984083712100983, "learning_rate": 1.4100589843423589e-05, "loss": 0.0097, "step": 76720 }, { "grad_norm": 0.11336325109004974, "learning_rate": 1.408908274276723e-05, "loss": 0.0098, "step": 76730 }, { "grad_norm": 0.2669624388217926, "learning_rate": 1.4077579569274786e-05, "loss": 0.0131, "step": 76740 }, { "grad_norm": 0.14621640741825104, "learning_rate": 1.4066080324204245e-05, "loss": 0.0096, "step": 76750 }, { "grad_norm": 0.2117798775434494, "learning_rate": 1.405458500881312e-05, "loss": 0.0116, "step": 76760 }, { "grad_norm": 0.09297521412372589, "learning_rate": 1.4043093624358538e-05, "loss": 0.0091, "step": 76770 }, { "grad_norm": 0.15172496438026428, "learning_rate": 1.4031606172097189e-05, "loss": 0.009, "step": 76780 }, { "grad_norm": 0.10180813819169998, "learning_rate": 1.40201226532853e-05, "loss": 0.0089, "step": 76790 }, { "grad_norm": 0.12779489159584045, "learning_rate": 1.4008643069178701e-05, "loss": 0.0099, "step": 76800 }, { "grad_norm": 0.16236834228038788, "learning_rate": 1.399716742103278e-05, "loss": 0.0098, "step": 76810 }, { "grad_norm": 0.21605359017848969, "learning_rate": 1.398569571010252e-05, "loss": 0.0136, "step": 76820 }, { "grad_norm": 0.11096976697444916, "learning_rate": 1.3974227937642414e-05, "loss": 0.0116, "step": 76830 }, { "grad_norm": 0.15529359877109528, "learning_rate": 1.3962764104906595e-05, "loss": 0.009, "step": 76840 }, { "grad_norm": 0.18395517766475677, "learning_rate": 1.3951304213148696e-05, "loss": 0.0119, "step": 76850 }, { "grad_norm": 0.13220849633216858, "learning_rate": 1.3939848263621979e-05, "loss": 0.0105, "step": 76860 }, { "grad_norm": 0.15451043844223022, "learning_rate": 1.3928396257579219e-05, "loss": 0.0095, "step": 76870 }, { "grad_norm": 0.17550064623355865, "learning_rate": 1.3916948196272833e-05, "loss": 0.0085, "step": 76880 }, { "grad_norm": 0.15101948380470276, "learning_rate": 1.3905504080954734e-05, "loss": 0.0096, "step": 76890 }, { "grad_norm": 0.13666865229606628, "learning_rate": 1.3894063912876454e-05, "loss": 0.0113, "step": 76900 }, { "grad_norm": 0.1480717658996582, "learning_rate": 1.3882627693289046e-05, "loss": 0.0115, "step": 76910 }, { "grad_norm": 0.1419588178396225, "learning_rate": 1.3871195423443183e-05, "loss": 0.0107, "step": 76920 }, { "grad_norm": 0.13297535479068756, "learning_rate": 1.3859767104589044e-05, "loss": 0.0096, "step": 76930 }, { "grad_norm": 0.12011035531759262, "learning_rate": 1.3848342737976467e-05, "loss": 0.0092, "step": 76940 }, { "grad_norm": 0.18280111253261566, "learning_rate": 1.3836922324854757e-05, "loss": 0.0129, "step": 76950 }, { "grad_norm": 0.12157953530550003, "learning_rate": 1.382550586647287e-05, "loss": 0.0087, "step": 76960 }, { "grad_norm": 0.20142152905464172, "learning_rate": 1.3814093364079257e-05, "loss": 0.0122, "step": 76970 }, { "grad_norm": 0.11417043954133987, "learning_rate": 1.3802684818921991e-05, "loss": 0.0109, "step": 76980 }, { "grad_norm": 0.1739671528339386, "learning_rate": 1.3791280232248687e-05, "loss": 0.0164, "step": 76990 }, { "grad_norm": 0.14277106523513794, "learning_rate": 1.3779879605306555e-05, "loss": 0.011, "step": 77000 }, { "grad_norm": 0.15571118891239166, "learning_rate": 1.376848293934232e-05, "loss": 0.0127, "step": 77010 }, { "grad_norm": 0.12721839547157288, "learning_rate": 1.3757090235602316e-05, "loss": 0.0092, "step": 77020 }, { "grad_norm": 0.12068958580493927, "learning_rate": 1.3745701495332447e-05, "loss": 0.0088, "step": 77030 }, { "grad_norm": 0.12223976850509644, "learning_rate": 1.3734316719778145e-05, "loss": 0.0096, "step": 77040 }, { "grad_norm": 0.134312242269516, "learning_rate": 1.3722935910184454e-05, "loss": 0.0084, "step": 77050 }, { "grad_norm": 0.11463628709316254, "learning_rate": 1.3711559067795921e-05, "loss": 0.0111, "step": 77060 }, { "grad_norm": 0.16668324172496796, "learning_rate": 1.3700186193856756e-05, "loss": 0.012, "step": 77070 }, { "grad_norm": 0.10813150554895401, "learning_rate": 1.368881728961064e-05, "loss": 0.0097, "step": 77080 }, { "grad_norm": 0.17700617015361786, "learning_rate": 1.3677452356300885e-05, "loss": 0.0146, "step": 77090 }, { "grad_norm": 0.14936241507530212, "learning_rate": 1.3666091395170321e-05, "loss": 0.0097, "step": 77100 }, { "grad_norm": 0.1702439934015274, "learning_rate": 1.3654734407461384e-05, "loss": 0.0108, "step": 77110 }, { "grad_norm": 0.16612689197063446, "learning_rate": 1.3643381394416022e-05, "loss": 0.0096, "step": 77120 }, { "grad_norm": 0.21913088858127594, "learning_rate": 1.3632032357275838e-05, "loss": 0.0106, "step": 77130 }, { "grad_norm": 0.10894303023815155, "learning_rate": 1.3620687297281898e-05, "loss": 0.0106, "step": 77140 }, { "grad_norm": 0.14424513280391693, "learning_rate": 1.3609346215674917e-05, "loss": 0.0111, "step": 77150 }, { "grad_norm": 0.1656447947025299, "learning_rate": 1.3598009113695109e-05, "loss": 0.0096, "step": 77160 }, { "grad_norm": 0.18563134968280792, "learning_rate": 1.358667599258231e-05, "loss": 0.0091, "step": 77170 }, { "grad_norm": 0.18256625533103943, "learning_rate": 1.357534685357586e-05, "loss": 0.0137, "step": 77180 }, { "grad_norm": 0.11784067749977112, "learning_rate": 1.3564021697914715e-05, "loss": 0.0094, "step": 77190 }, { "grad_norm": 0.2071860134601593, "learning_rate": 1.3552700526837381e-05, "loss": 0.0127, "step": 77200 }, { "grad_norm": 0.16944266855716705, "learning_rate": 1.3541383341581925e-05, "loss": 0.0104, "step": 77210 }, { "grad_norm": 0.1347435712814331, "learning_rate": 1.3530070143385965e-05, "loss": 0.01, "step": 77220 }, { "grad_norm": 0.1372135579586029, "learning_rate": 1.3518760933486702e-05, "loss": 0.0118, "step": 77230 }, { "grad_norm": 0.1775752156972885, "learning_rate": 1.3507455713120908e-05, "loss": 0.0108, "step": 77240 }, { "grad_norm": 0.1314828097820282, "learning_rate": 1.3496154483524876e-05, "loss": 0.0088, "step": 77250 }, { "grad_norm": 0.12566325068473816, "learning_rate": 1.3484857245934506e-05, "loss": 0.0097, "step": 77260 }, { "grad_norm": 0.15951040387153625, "learning_rate": 1.3473564001585248e-05, "loss": 0.009, "step": 77270 }, { "grad_norm": 0.10965277999639511, "learning_rate": 1.3462274751712129e-05, "loss": 0.0116, "step": 77280 }, { "grad_norm": 0.18340738117694855, "learning_rate": 1.3450989497549698e-05, "loss": 0.0101, "step": 77290 }, { "grad_norm": 0.23441918194293976, "learning_rate": 1.3439708240332115e-05, "loss": 0.0097, "step": 77300 }, { "grad_norm": 0.0914975255727768, "learning_rate": 1.3428430981293043e-05, "loss": 0.0095, "step": 77310 }, { "grad_norm": 0.1446097195148468, "learning_rate": 1.3417157721665808e-05, "loss": 0.0115, "step": 77320 }, { "grad_norm": 0.17859384417533875, "learning_rate": 1.3405888462683186e-05, "loss": 0.0093, "step": 77330 }, { "grad_norm": 0.13434357941150665, "learning_rate": 1.3394623205577595e-05, "loss": 0.0093, "step": 77340 }, { "grad_norm": 0.12229438871145248, "learning_rate": 1.3383361951580963e-05, "loss": 0.012, "step": 77350 }, { "grad_norm": 0.11246166378259659, "learning_rate": 1.337210470192483e-05, "loss": 0.008, "step": 77360 }, { "grad_norm": 0.10852111876010895, "learning_rate": 1.3360851457840245e-05, "loss": 0.0095, "step": 77370 }, { "grad_norm": 0.16142262518405914, "learning_rate": 1.3349602220557856e-05, "loss": 0.0104, "step": 77380 }, { "grad_norm": 0.12672801315784454, "learning_rate": 1.3338356991307865e-05, "loss": 0.0093, "step": 77390 }, { "grad_norm": 0.10751841217279434, "learning_rate": 1.3327115771320048e-05, "loss": 0.009, "step": 77400 }, { "grad_norm": 0.13875341415405273, "learning_rate": 1.3315878561823697e-05, "loss": 0.0097, "step": 77410 }, { "grad_norm": 0.10018976032733917, "learning_rate": 1.330464536404773e-05, "loss": 0.0097, "step": 77420 }, { "grad_norm": 0.13572752475738525, "learning_rate": 1.3293416179220563e-05, "loss": 0.0102, "step": 77430 }, { "grad_norm": 0.12392910569906235, "learning_rate": 1.328219100857021e-05, "loss": 0.0123, "step": 77440 }, { "grad_norm": 0.17733940482139587, "learning_rate": 1.327096985332425e-05, "loss": 0.013, "step": 77450 }, { "grad_norm": 0.13216198980808258, "learning_rate": 1.3259752714709816e-05, "loss": 0.0131, "step": 77460 }, { "grad_norm": 0.15470948815345764, "learning_rate": 1.3248539593953569e-05, "loss": 0.0114, "step": 77470 }, { "grad_norm": 0.12331976741552353, "learning_rate": 1.323733049228178e-05, "loss": 0.0112, "step": 77480 }, { "grad_norm": 0.16179072856903076, "learning_rate": 1.3226125410920276e-05, "loss": 0.0111, "step": 77490 }, { "grad_norm": 0.10915383696556091, "learning_rate": 1.3214924351094388e-05, "loss": 0.0112, "step": 77500 }, { "grad_norm": 0.17413197457790375, "learning_rate": 1.3203727314029069e-05, "loss": 0.0098, "step": 77510 }, { "grad_norm": 0.1594391167163849, "learning_rate": 1.3192534300948811e-05, "loss": 0.0091, "step": 77520 }, { "grad_norm": 0.14084340631961823, "learning_rate": 1.3181345313077675e-05, "loss": 0.0088, "step": 77530 }, { "grad_norm": 0.22311846911907196, "learning_rate": 1.3170160351639243e-05, "loss": 0.0119, "step": 77540 }, { "grad_norm": 0.16196589171886444, "learning_rate": 1.3158979417856721e-05, "loss": 0.0106, "step": 77550 }, { "grad_norm": 0.1138884425163269, "learning_rate": 1.31478025129528e-05, "loss": 0.0076, "step": 77560 }, { "grad_norm": 0.1456078141927719, "learning_rate": 1.3136629638149794e-05, "loss": 0.0091, "step": 77570 }, { "grad_norm": 0.1666204184293747, "learning_rate": 1.3125460794669547e-05, "loss": 0.0116, "step": 77580 }, { "grad_norm": 0.10139644145965576, "learning_rate": 1.3114295983733483e-05, "loss": 0.0115, "step": 77590 }, { "grad_norm": 0.11369127035140991, "learning_rate": 1.3103135206562534e-05, "loss": 0.0084, "step": 77600 }, { "grad_norm": 0.1503887176513672, "learning_rate": 1.3091978464377264e-05, "loss": 0.0093, "step": 77610 }, { "grad_norm": 0.13982979953289032, "learning_rate": 1.3080825758397724e-05, "loss": 0.0104, "step": 77620 }, { "grad_norm": 0.09355859458446503, "learning_rate": 1.3069677089843568e-05, "loss": 0.0106, "step": 77630 }, { "grad_norm": 0.10588357597589493, "learning_rate": 1.305853245993401e-05, "loss": 0.012, "step": 77640 }, { "grad_norm": 0.1408269852399826, "learning_rate": 1.3047391869887809e-05, "loss": 0.0099, "step": 77650 }, { "grad_norm": 0.15851455926895142, "learning_rate": 1.3036255320923263e-05, "loss": 0.0122, "step": 77660 }, { "grad_norm": 0.1354474276304245, "learning_rate": 1.3025122814258273e-05, "loss": 0.01, "step": 77670 }, { "grad_norm": 0.12657710909843445, "learning_rate": 1.3013994351110249e-05, "loss": 0.0091, "step": 77680 }, { "grad_norm": 0.19373533129692078, "learning_rate": 1.3002869932696194e-05, "loss": 0.0109, "step": 77690 }, { "grad_norm": 0.1607394814491272, "learning_rate": 1.2991749560232664e-05, "loss": 0.0099, "step": 77700 }, { "grad_norm": 0.13066457211971283, "learning_rate": 1.2980633234935768e-05, "loss": 0.012, "step": 77710 }, { "grad_norm": 0.1799500584602356, "learning_rate": 1.2969520958021153e-05, "loss": 0.0128, "step": 77720 }, { "grad_norm": 0.15457013249397278, "learning_rate": 1.295841273070405e-05, "loss": 0.0096, "step": 77730 }, { "grad_norm": 0.1229863166809082, "learning_rate": 1.2947308554199255e-05, "loss": 0.0102, "step": 77740 }, { "grad_norm": 0.1580832153558731, "learning_rate": 1.2936208429721075e-05, "loss": 0.0102, "step": 77750 }, { "grad_norm": 0.1721266508102417, "learning_rate": 1.2925112358483416e-05, "loss": 0.0106, "step": 77760 }, { "grad_norm": 0.13422299921512604, "learning_rate": 1.2914020341699728e-05, "loss": 0.0105, "step": 77770 }, { "grad_norm": 0.12374819815158844, "learning_rate": 1.2902932380583032e-05, "loss": 0.0094, "step": 77780 }, { "grad_norm": 0.0865088403224945, "learning_rate": 1.2891848476345863e-05, "loss": 0.0111, "step": 77790 }, { "grad_norm": 0.1540168970823288, "learning_rate": 1.2880768630200368e-05, "loss": 0.0108, "step": 77800 }, { "grad_norm": 0.14653539657592773, "learning_rate": 1.286969284335819e-05, "loss": 0.0108, "step": 77810 }, { "grad_norm": 0.08584823459386826, "learning_rate": 1.2858621117030578e-05, "loss": 0.0101, "step": 77820 }, { "grad_norm": 0.17641332745552063, "learning_rate": 1.2847553452428324e-05, "loss": 0.0093, "step": 77830 }, { "grad_norm": 0.1239163875579834, "learning_rate": 1.283648985076178e-05, "loss": 0.0094, "step": 77840 }, { "grad_norm": 0.1310458481311798, "learning_rate": 1.2825430313240817e-05, "loss": 0.0093, "step": 77850 }, { "grad_norm": 0.10810291767120361, "learning_rate": 1.281437484107491e-05, "loss": 0.0092, "step": 77860 }, { "grad_norm": 0.21253162622451782, "learning_rate": 1.2803323435473058e-05, "loss": 0.0085, "step": 77870 }, { "grad_norm": 0.18040885031223297, "learning_rate": 1.2792276097643841e-05, "loss": 0.0101, "step": 77880 }, { "grad_norm": 0.1755422204732895, "learning_rate": 1.2781232828795342e-05, "loss": 0.0113, "step": 77890 }, { "grad_norm": 0.1546889990568161, "learning_rate": 1.2770193630135292e-05, "loss": 0.0114, "step": 77900 }, { "grad_norm": 0.2570444345474243, "learning_rate": 1.2759158502870877e-05, "loss": 0.0118, "step": 77910 }, { "grad_norm": 0.17881211638450623, "learning_rate": 1.2748127448208913e-05, "loss": 0.0103, "step": 77920 }, { "grad_norm": 0.11421958357095718, "learning_rate": 1.2737100467355706e-05, "loss": 0.0095, "step": 77930 }, { "grad_norm": 0.10048062354326248, "learning_rate": 1.2726077561517169e-05, "loss": 0.012, "step": 77940 }, { "grad_norm": 0.16459539532661438, "learning_rate": 1.2715058731898755e-05, "loss": 0.0145, "step": 77950 }, { "grad_norm": 0.12861232459545135, "learning_rate": 1.2704043979705471e-05, "loss": 0.0093, "step": 77960 }, { "grad_norm": 0.17256803810596466, "learning_rate": 1.269303330614185e-05, "loss": 0.0169, "step": 77970 }, { "grad_norm": 0.19192542135715485, "learning_rate": 1.2682026712412015e-05, "loss": 0.0094, "step": 77980 }, { "grad_norm": 0.10918533802032471, "learning_rate": 1.2671024199719644e-05, "loss": 0.0087, "step": 77990 }, { "grad_norm": 0.1115463525056839, "learning_rate": 1.2660025769267936e-05, "loss": 0.0093, "step": 78000 }, { "grad_norm": 0.16554702818393707, "learning_rate": 1.2649031422259667e-05, "loss": 0.0131, "step": 78010 }, { "grad_norm": 0.12946727871894836, "learning_rate": 1.2638041159897163e-05, "loss": 0.0105, "step": 78020 }, { "grad_norm": 0.0810372456908226, "learning_rate": 1.262705498338232e-05, "loss": 0.0105, "step": 78030 }, { "grad_norm": 0.11488162726163864, "learning_rate": 1.2616072893916542e-05, "loss": 0.0088, "step": 78040 }, { "grad_norm": 0.3132184147834778, "learning_rate": 1.2605094892700836e-05, "loss": 0.0103, "step": 78050 }, { "grad_norm": 0.11209399253129959, "learning_rate": 1.259412098093572e-05, "loss": 0.0109, "step": 78060 }, { "grad_norm": 0.13609695434570312, "learning_rate": 1.258315115982131e-05, "loss": 0.0101, "step": 78070 }, { "grad_norm": 0.27364012598991394, "learning_rate": 1.2572185430557203e-05, "loss": 0.0146, "step": 78080 }, { "grad_norm": 0.11200922727584839, "learning_rate": 1.2561223794342659e-05, "loss": 0.0099, "step": 78090 }, { "grad_norm": 0.18729722499847412, "learning_rate": 1.255026625237637e-05, "loss": 0.0099, "step": 78100 }, { "grad_norm": 0.17435383796691895, "learning_rate": 1.2539312805856678e-05, "loss": 0.0142, "step": 78110 }, { "grad_norm": 0.10602771490812302, "learning_rate": 1.25283634559814e-05, "loss": 0.0092, "step": 78120 }, { "grad_norm": 0.25240498781204224, "learning_rate": 1.2517418203947972e-05, "loss": 0.0083, "step": 78130 }, { "grad_norm": 0.16391506791114807, "learning_rate": 1.2506477050953308e-05, "loss": 0.0124, "step": 78140 }, { "grad_norm": 0.09469930827617645, "learning_rate": 1.2495539998193972e-05, "loss": 0.0097, "step": 78150 }, { "grad_norm": 0.20121529698371887, "learning_rate": 1.2484607046865975e-05, "loss": 0.0081, "step": 78160 }, { "grad_norm": 0.08783475309610367, "learning_rate": 1.2473678198164968e-05, "loss": 0.0105, "step": 78170 }, { "grad_norm": 0.17883582413196564, "learning_rate": 1.2462753453286075e-05, "loss": 0.0117, "step": 78180 }, { "grad_norm": 0.10843195766210556, "learning_rate": 1.2451832813424031e-05, "loss": 0.0116, "step": 78190 }, { "grad_norm": 0.15071627497673035, "learning_rate": 1.2440916279773095e-05, "loss": 0.0087, "step": 78200 }, { "grad_norm": 0.14459799230098724, "learning_rate": 1.2430003853527101e-05, "loss": 0.0092, "step": 78210 }, { "grad_norm": 0.14904560148715973, "learning_rate": 1.2419095535879383e-05, "loss": 0.0098, "step": 78220 }, { "grad_norm": 0.17881707847118378, "learning_rate": 1.2408191328022878e-05, "loss": 0.0126, "step": 78230 }, { "grad_norm": 0.13865622878074646, "learning_rate": 1.2397291231150066e-05, "loss": 0.0091, "step": 78240 }, { "grad_norm": 0.18987323343753815, "learning_rate": 1.2386395246452937e-05, "loss": 0.0134, "step": 78250 }, { "grad_norm": 0.12262032181024551, "learning_rate": 1.2375503375123083e-05, "loss": 0.0102, "step": 78260 }, { "grad_norm": 0.179258793592453, "learning_rate": 1.2364615618351583e-05, "loss": 0.0109, "step": 78270 }, { "grad_norm": 0.17988908290863037, "learning_rate": 1.2353731977329169e-05, "loss": 0.0097, "step": 78280 }, { "grad_norm": 0.16234324872493744, "learning_rate": 1.2342852453246007e-05, "loss": 0.009, "step": 78290 }, { "grad_norm": 0.09531520307064056, "learning_rate": 1.2331977047291898e-05, "loss": 0.0088, "step": 78300 }, { "grad_norm": 0.12327715009450912, "learning_rate": 1.2321105760656137e-05, "loss": 0.0088, "step": 78310 }, { "grad_norm": 0.16256798803806305, "learning_rate": 1.2310238594527613e-05, "loss": 0.0119, "step": 78320 }, { "grad_norm": 0.11649076640605927, "learning_rate": 1.2299375550094693e-05, "loss": 0.0081, "step": 78330 }, { "grad_norm": 0.09158605337142944, "learning_rate": 1.2288516628545421e-05, "loss": 0.0085, "step": 78340 }, { "grad_norm": 0.09035789966583252, "learning_rate": 1.2277661831067255e-05, "loss": 0.0087, "step": 78350 }, { "grad_norm": 0.11614082008600235, "learning_rate": 1.2266811158847286e-05, "loss": 0.0093, "step": 78360 }, { "grad_norm": 0.12756969034671783, "learning_rate": 1.2255964613072107e-05, "loss": 0.0141, "step": 78370 }, { "grad_norm": 0.15862207114696503, "learning_rate": 1.2245122194927905e-05, "loss": 0.0084, "step": 78380 }, { "grad_norm": 0.11392594873905182, "learning_rate": 1.2234283905600364e-05, "loss": 0.0093, "step": 78390 }, { "grad_norm": 0.14181961119174957, "learning_rate": 1.222344974627475e-05, "loss": 0.0092, "step": 78400 }, { "grad_norm": 0.15761008858680725, "learning_rate": 1.2212619718135875e-05, "loss": 0.0097, "step": 78410 }, { "grad_norm": 0.1460326761007309, "learning_rate": 1.2201793822368118e-05, "loss": 0.0118, "step": 78420 }, { "grad_norm": 0.13368140161037445, "learning_rate": 1.2190972060155337e-05, "loss": 0.0107, "step": 78430 }, { "grad_norm": 0.17100340127944946, "learning_rate": 1.2180154432681013e-05, "loss": 0.0091, "step": 78440 }, { "grad_norm": 0.1257672756910324, "learning_rate": 1.2169340941128148e-05, "loss": 0.0112, "step": 78450 }, { "grad_norm": 0.14112024009227753, "learning_rate": 1.2158531586679267e-05, "loss": 0.0096, "step": 78460 }, { "grad_norm": 0.1218312606215477, "learning_rate": 1.2147726370516477e-05, "loss": 0.0097, "step": 78470 }, { "grad_norm": 0.15667645633220673, "learning_rate": 1.2136925293821422e-05, "loss": 0.016, "step": 78480 }, { "grad_norm": 0.12318778038024902, "learning_rate": 1.2126128357775306e-05, "loss": 0.0093, "step": 78490 }, { "grad_norm": 0.11902480572462082, "learning_rate": 1.2115335563558838e-05, "loss": 0.0089, "step": 78500 }, { "grad_norm": 0.1794152855873108, "learning_rate": 1.2104546912352327e-05, "loss": 0.0081, "step": 78510 }, { "grad_norm": 0.14430128037929535, "learning_rate": 1.2093762405335557e-05, "loss": 0.0086, "step": 78520 }, { "grad_norm": 0.11060429364442825, "learning_rate": 1.2082982043687974e-05, "loss": 0.0079, "step": 78530 }, { "grad_norm": 0.19250379502773285, "learning_rate": 1.207220582858845e-05, "loss": 0.0092, "step": 78540 }, { "grad_norm": 0.13206219673156738, "learning_rate": 1.206143376121549e-05, "loss": 0.0096, "step": 78550 }, { "grad_norm": 0.2463952898979187, "learning_rate": 1.2050665842747078e-05, "loss": 0.0151, "step": 78560 }, { "grad_norm": 0.13084767758846283, "learning_rate": 1.2039902074360804e-05, "loss": 0.0106, "step": 78570 }, { "grad_norm": 0.1475047618150711, "learning_rate": 1.2029142457233755e-05, "loss": 0.0138, "step": 78580 }, { "grad_norm": 0.1944749653339386, "learning_rate": 1.2018386992542601e-05, "loss": 0.0091, "step": 78590 }, { "grad_norm": 0.14427685737609863, "learning_rate": 1.2007635681463542e-05, "loss": 0.0106, "step": 78600 }, { "grad_norm": 0.1636778712272644, "learning_rate": 1.1996888525172334e-05, "loss": 0.0108, "step": 78610 }, { "grad_norm": 0.1755026876926422, "learning_rate": 1.198614552484425e-05, "loss": 0.013, "step": 78620 }, { "grad_norm": 0.12808962166309357, "learning_rate": 1.197540668165415e-05, "loss": 0.0093, "step": 78630 }, { "grad_norm": 0.17014755308628082, "learning_rate": 1.1964671996776395e-05, "loss": 0.0112, "step": 78640 }, { "grad_norm": 0.11889371275901794, "learning_rate": 1.1953941471384922e-05, "loss": 0.0124, "step": 78650 }, { "grad_norm": 0.1709960550069809, "learning_rate": 1.1943215106653205e-05, "loss": 0.0094, "step": 78660 }, { "grad_norm": 0.11408030241727829, "learning_rate": 1.193249290375429e-05, "loss": 0.0095, "step": 78670 }, { "grad_norm": 0.16726356744766235, "learning_rate": 1.1921774863860696e-05, "loss": 0.009, "step": 78680 }, { "grad_norm": 0.18078656494617462, "learning_rate": 1.1911060988144556e-05, "loss": 0.0114, "step": 78690 }, { "grad_norm": 0.17639482021331787, "learning_rate": 1.1900351277777533e-05, "loss": 0.0113, "step": 78700 }, { "grad_norm": 0.1848185807466507, "learning_rate": 1.1889645733930798e-05, "loss": 0.0106, "step": 78710 }, { "grad_norm": 0.1923157274723053, "learning_rate": 1.1878944357775107e-05, "loss": 0.0115, "step": 78720 }, { "grad_norm": 0.14072194695472717, "learning_rate": 1.1868247150480749e-05, "loss": 0.0101, "step": 78730 }, { "grad_norm": 0.14726674556732178, "learning_rate": 1.1857554113217567e-05, "loss": 0.0102, "step": 78740 }, { "grad_norm": 0.17673377692699432, "learning_rate": 1.1846865247154903e-05, "loss": 0.0118, "step": 78750 }, { "grad_norm": 0.13498786091804504, "learning_rate": 1.1836180553461711e-05, "loss": 0.0092, "step": 78760 }, { "grad_norm": 0.15282809734344482, "learning_rate": 1.1825500033306419e-05, "loss": 0.011, "step": 78770 }, { "grad_norm": 0.16191880404949188, "learning_rate": 1.1814823687857052e-05, "loss": 0.0093, "step": 78780 }, { "grad_norm": 0.14335612952709198, "learning_rate": 1.1804151518281153e-05, "loss": 0.0071, "step": 78790 }, { "grad_norm": 0.0999172106385231, "learning_rate": 1.179348352574583e-05, "loss": 0.0101, "step": 78800 }, { "grad_norm": 0.08479676395654678, "learning_rate": 1.1782819711417697e-05, "loss": 0.0087, "step": 78810 }, { "grad_norm": 0.20803923904895782, "learning_rate": 1.1772160076462951e-05, "loss": 0.0104, "step": 78820 }, { "grad_norm": 0.16494810581207275, "learning_rate": 1.1761504622047293e-05, "loss": 0.0095, "step": 78830 }, { "grad_norm": 0.12074413150548935, "learning_rate": 1.1750853349335999e-05, "loss": 0.0104, "step": 78840 }, { "grad_norm": 0.1353066861629486, "learning_rate": 1.1740206259493874e-05, "loss": 0.009, "step": 78850 }, { "grad_norm": 0.1046728864312172, "learning_rate": 1.1729563353685286e-05, "loss": 0.012, "step": 78860 }, { "grad_norm": 0.11579805612564087, "learning_rate": 1.1718924633074091e-05, "loss": 0.0128, "step": 78870 }, { "grad_norm": 0.12488388270139694, "learning_rate": 1.1708290098823765e-05, "loss": 0.01, "step": 78880 }, { "grad_norm": 0.13004127144813538, "learning_rate": 1.1697659752097245e-05, "loss": 0.0113, "step": 78890 }, { "grad_norm": 0.09544640779495239, "learning_rate": 1.1687033594057068e-05, "loss": 0.0073, "step": 78900 }, { "grad_norm": 0.19885726273059845, "learning_rate": 1.1676411625865297e-05, "loss": 0.009, "step": 78910 }, { "grad_norm": 0.15878424048423767, "learning_rate": 1.1665793848683542e-05, "loss": 0.0105, "step": 78920 }, { "grad_norm": 0.11162292212247849, "learning_rate": 1.1655180263672927e-05, "loss": 0.0085, "step": 78930 }, { "grad_norm": 0.1455342173576355, "learning_rate": 1.1644570871994142e-05, "loss": 0.0092, "step": 78940 }, { "grad_norm": 0.09079642593860626, "learning_rate": 1.163396567480744e-05, "loss": 0.0078, "step": 78950 }, { "grad_norm": 0.12760746479034424, "learning_rate": 1.1623364673272552e-05, "loss": 0.0103, "step": 78960 }, { "grad_norm": 0.19698649644851685, "learning_rate": 1.1612767868548807e-05, "loss": 0.0081, "step": 78970 }, { "grad_norm": 0.11725558340549469, "learning_rate": 1.1602175261795057e-05, "loss": 0.01, "step": 78980 }, { "grad_norm": 0.1308021992444992, "learning_rate": 1.1591586854169696e-05, "loss": 0.0086, "step": 78990 }, { "grad_norm": 0.1386333703994751, "learning_rate": 1.158100264683064e-05, "loss": 0.0097, "step": 79000 }, { "grad_norm": 0.12751080095767975, "learning_rate": 1.1570422640935386e-05, "loss": 0.0144, "step": 79010 }, { "grad_norm": 0.13395985960960388, "learning_rate": 1.155984683764092e-05, "loss": 0.0124, "step": 79020 }, { "grad_norm": 0.14450329542160034, "learning_rate": 1.1549275238103812e-05, "loss": 0.0119, "step": 79030 }, { "grad_norm": 0.23594790697097778, "learning_rate": 1.1538707843480152e-05, "loss": 0.0111, "step": 79040 }, { "grad_norm": 0.1268092393875122, "learning_rate": 1.1528144654925593e-05, "loss": 0.0094, "step": 79050 }, { "grad_norm": 0.12406516075134277, "learning_rate": 1.1517585673595277e-05, "loss": 0.0104, "step": 79060 }, { "grad_norm": 0.151638925075531, "learning_rate": 1.150703090064395e-05, "loss": 0.0103, "step": 79070 }, { "grad_norm": 0.10211902856826782, "learning_rate": 1.1496480337225835e-05, "loss": 0.0096, "step": 79080 }, { "grad_norm": 0.11742343008518219, "learning_rate": 1.1485933984494751e-05, "loss": 0.0085, "step": 79090 }, { "grad_norm": 0.1254035085439682, "learning_rate": 1.1475391843604e-05, "loss": 0.01, "step": 79100 }, { "grad_norm": 0.15948668122291565, "learning_rate": 1.1464853915706503e-05, "loss": 0.0099, "step": 79110 }, { "grad_norm": 0.15207551419734955, "learning_rate": 1.1454320201954626e-05, "loss": 0.0098, "step": 79120 }, { "grad_norm": 0.12332788854837418, "learning_rate": 1.1443790703500356e-05, "loss": 0.0083, "step": 79130 }, { "grad_norm": 0.09829486161470413, "learning_rate": 1.1433265421495154e-05, "loss": 0.0102, "step": 79140 }, { "grad_norm": 0.17514728009700775, "learning_rate": 1.1422744357090065e-05, "loss": 0.0094, "step": 79150 }, { "grad_norm": 0.20518016815185547, "learning_rate": 1.1412227511435647e-05, "loss": 0.013, "step": 79160 }, { "grad_norm": 0.1367042064666748, "learning_rate": 1.1401714885682025e-05, "loss": 0.0093, "step": 79170 }, { "grad_norm": 0.12593841552734375, "learning_rate": 1.1391206480978823e-05, "loss": 0.013, "step": 79180 }, { "grad_norm": 0.16373398900032043, "learning_rate": 1.1380702298475227e-05, "loss": 0.0104, "step": 79190 }, { "grad_norm": 0.1264587640762329, "learning_rate": 1.1370202339319985e-05, "loss": 0.0085, "step": 79200 }, { "grad_norm": 0.16559748351573944, "learning_rate": 1.1359706604661319e-05, "loss": 0.0134, "step": 79210 }, { "grad_norm": 0.12537075579166412, "learning_rate": 1.1349215095647042e-05, "loss": 0.0083, "step": 79220 }, { "grad_norm": 0.1810542494058609, "learning_rate": 1.1338727813424493e-05, "loss": 0.0095, "step": 79230 }, { "grad_norm": 0.12206059694290161, "learning_rate": 1.132824475914056e-05, "loss": 0.0101, "step": 79240 }, { "grad_norm": 0.17811870574951172, "learning_rate": 1.131776593394162e-05, "loss": 0.0117, "step": 79250 }, { "grad_norm": 0.14845985174179077, "learning_rate": 1.1307291338973657e-05, "loss": 0.0097, "step": 79260 }, { "grad_norm": 0.1651524007320404, "learning_rate": 1.1296820975382121e-05, "loss": 0.0098, "step": 79270 }, { "grad_norm": 0.11184792220592499, "learning_rate": 1.1286354844312074e-05, "loss": 0.0081, "step": 79280 }, { "grad_norm": 0.1680568903684616, "learning_rate": 1.1275892946908023e-05, "loss": 0.0087, "step": 79290 }, { "grad_norm": 0.15621636807918549, "learning_rate": 1.126543528431413e-05, "loss": 0.0078, "step": 79300 }, { "grad_norm": 0.11671435832977295, "learning_rate": 1.125498185767398e-05, "loss": 0.01, "step": 79310 }, { "grad_norm": 0.17593564093112946, "learning_rate": 1.1244532668130781e-05, "loss": 0.0121, "step": 79320 }, { "grad_norm": 0.15934933722019196, "learning_rate": 1.1234087716827203e-05, "loss": 0.0098, "step": 79330 }, { "grad_norm": 0.1631799340248108, "learning_rate": 1.1223647004905524e-05, "loss": 0.0097, "step": 79340 }, { "grad_norm": 0.2025667428970337, "learning_rate": 1.1213210533507485e-05, "loss": 0.0121, "step": 79350 }, { "grad_norm": 0.09283778071403503, "learning_rate": 1.1202778303774447e-05, "loss": 0.0098, "step": 79360 }, { "grad_norm": 0.1556718498468399, "learning_rate": 1.1192350316847234e-05, "loss": 0.0107, "step": 79370 }, { "grad_norm": 0.10268218070268631, "learning_rate": 1.1181926573866258e-05, "loss": 0.0099, "step": 79380 }, { "grad_norm": 0.12236688286066055, "learning_rate": 1.1171507075971416e-05, "loss": 0.0117, "step": 79390 }, { "grad_norm": 0.20773574709892273, "learning_rate": 1.1161091824302177e-05, "loss": 0.0097, "step": 79400 }, { "grad_norm": 0.14173398911952972, "learning_rate": 1.115068081999755e-05, "loss": 0.0123, "step": 79410 }, { "grad_norm": 0.11390849202871323, "learning_rate": 1.1140274064196071e-05, "loss": 0.0094, "step": 79420 }, { "grad_norm": 0.11370441317558289, "learning_rate": 1.112987155803578e-05, "loss": 0.0101, "step": 79430 }, { "grad_norm": 0.22111202776432037, "learning_rate": 1.1119473302654298e-05, "loss": 0.0098, "step": 79440 }, { "grad_norm": 0.12662459909915924, "learning_rate": 1.1109079299188769e-05, "loss": 0.0109, "step": 79450 }, { "grad_norm": 0.18504877388477325, "learning_rate": 1.1098689548775847e-05, "loss": 0.0103, "step": 79460 }, { "grad_norm": 0.14664705097675323, "learning_rate": 1.1088304052551762e-05, "loss": 0.0112, "step": 79470 }, { "grad_norm": 0.15823043882846832, "learning_rate": 1.1077922811652215e-05, "loss": 0.0146, "step": 79480 }, { "grad_norm": 0.12859760224819183, "learning_rate": 1.1067545827212534e-05, "loss": 0.0086, "step": 79490 }, { "grad_norm": 0.15526634454727173, "learning_rate": 1.1057173100367496e-05, "loss": 0.0088, "step": 79500 }, { "grad_norm": 0.13669736683368683, "learning_rate": 1.1046804632251474e-05, "loss": 0.0081, "step": 79510 }, { "grad_norm": 0.2295994907617569, "learning_rate": 1.1036440423998318e-05, "loss": 0.0116, "step": 79520 }, { "grad_norm": 0.2206064760684967, "learning_rate": 1.1026080476741469e-05, "loss": 0.0132, "step": 79530 }, { "grad_norm": 0.14421306550502777, "learning_rate": 1.1015724791613836e-05, "loss": 0.0088, "step": 79540 }, { "grad_norm": 0.1550522893667221, "learning_rate": 1.1005373369747951e-05, "loss": 0.008, "step": 79550 }, { "grad_norm": 0.12304791808128357, "learning_rate": 1.09950262122758e-05, "loss": 0.009, "step": 79560 }, { "grad_norm": 0.10280271619558334, "learning_rate": 1.0984683320328948e-05, "loss": 0.0115, "step": 79570 }, { "grad_norm": 0.1331128180027008, "learning_rate": 1.0974344695038458e-05, "loss": 0.0099, "step": 79580 }, { "grad_norm": 0.1632416695356369, "learning_rate": 1.0964010337534968e-05, "loss": 0.01, "step": 79590 }, { "grad_norm": 0.11106047034263611, "learning_rate": 1.0953680248948611e-05, "loss": 0.0114, "step": 79600 }, { "grad_norm": 0.17964474856853485, "learning_rate": 1.0943354430409069e-05, "loss": 0.0113, "step": 79610 }, { "grad_norm": 0.12508633732795715, "learning_rate": 1.0933032883045574e-05, "loss": 0.009, "step": 79620 }, { "grad_norm": 0.12000350654125214, "learning_rate": 1.0922715607986872e-05, "loss": 0.0109, "step": 79630 }, { "grad_norm": 0.10744944214820862, "learning_rate": 1.0912402606361234e-05, "loss": 0.0119, "step": 79640 }, { "grad_norm": 0.11316825449466705, "learning_rate": 1.090209387929647e-05, "loss": 0.0096, "step": 79650 }, { "grad_norm": 0.14514942467212677, "learning_rate": 1.0891789427919957e-05, "loss": 0.0098, "step": 79660 }, { "grad_norm": 0.1584622710943222, "learning_rate": 1.088148925335854e-05, "loss": 0.0105, "step": 79670 }, { "grad_norm": 0.21768566966056824, "learning_rate": 1.087119335673864e-05, "loss": 0.0099, "step": 79680 }, { "grad_norm": 0.1064792275428772, "learning_rate": 1.0860901739186208e-05, "loss": 0.0104, "step": 79690 }, { "grad_norm": 0.14953815937042236, "learning_rate": 1.085061440182673e-05, "loss": 0.0119, "step": 79700 }, { "grad_norm": 0.12428490817546844, "learning_rate": 1.0840331345785193e-05, "loss": 0.0102, "step": 79710 }, { "grad_norm": 0.09188037365674973, "learning_rate": 1.083005257218615e-05, "loss": 0.0096, "step": 79720 }, { "grad_norm": 0.15971964597702026, "learning_rate": 1.0819778082153642e-05, "loss": 0.0101, "step": 79730 }, { "grad_norm": 0.13434308767318726, "learning_rate": 1.0809507876811325e-05, "loss": 0.009, "step": 79740 }, { "grad_norm": 0.13020528852939606, "learning_rate": 1.0799241957282292e-05, "loss": 0.0103, "step": 79750 }, { "grad_norm": 0.12774892151355743, "learning_rate": 1.078898032468923e-05, "loss": 0.0089, "step": 79760 }, { "grad_norm": 0.08672377467155457, "learning_rate": 1.0778722980154315e-05, "loss": 0.0116, "step": 79770 }, { "grad_norm": 0.17727109789848328, "learning_rate": 1.0768469924799302e-05, "loss": 0.0133, "step": 79780 }, { "grad_norm": 0.1758522391319275, "learning_rate": 1.0758221159745419e-05, "loss": 0.0095, "step": 79790 }, { "grad_norm": 0.16015976667404175, "learning_rate": 1.074797668611347e-05, "loss": 0.013, "step": 79800 }, { "grad_norm": 0.09344349056482315, "learning_rate": 1.0737736505023777e-05, "loss": 0.0099, "step": 79810 }, { "grad_norm": 0.21699965000152588, "learning_rate": 1.0727500617596203e-05, "loss": 0.0085, "step": 79820 }, { "grad_norm": 0.16029700636863708, "learning_rate": 1.0717269024950104e-05, "loss": 0.0119, "step": 79830 }, { "grad_norm": 0.11880432069301605, "learning_rate": 1.070704172820442e-05, "loss": 0.0096, "step": 79840 }, { "grad_norm": 0.13911092281341553, "learning_rate": 1.0696818728477559e-05, "loss": 0.0096, "step": 79850 }, { "grad_norm": 0.08925985544919968, "learning_rate": 1.0686600026887511e-05, "loss": 0.0125, "step": 79860 }, { "grad_norm": 0.236466646194458, "learning_rate": 1.0676385624551777e-05, "loss": 0.0101, "step": 79870 }, { "grad_norm": 0.16655533015727997, "learning_rate": 1.0666175522587402e-05, "loss": 0.0092, "step": 79880 }, { "grad_norm": 0.1395842581987381, "learning_rate": 1.0655969722110925e-05, "loss": 0.0086, "step": 79890 }, { "grad_norm": 0.15758872032165527, "learning_rate": 1.0645768224238439e-05, "loss": 0.0092, "step": 79900 }, { "grad_norm": 0.1868884563446045, "learning_rate": 1.063557103008559e-05, "loss": 0.012, "step": 79910 }, { "grad_norm": 0.15136609971523285, "learning_rate": 1.0625378140767494e-05, "loss": 0.0099, "step": 79920 }, { "grad_norm": 0.16298876702785492, "learning_rate": 1.0615189557398846e-05, "loss": 0.0093, "step": 79930 }, { "grad_norm": 0.22375839948654175, "learning_rate": 1.060500528109386e-05, "loss": 0.0106, "step": 79940 }, { "grad_norm": 0.1456127017736435, "learning_rate": 1.0594825312966272e-05, "loss": 0.0111, "step": 79950 }, { "grad_norm": 0.1686200350522995, "learning_rate": 1.0584649654129336e-05, "loss": 0.0088, "step": 79960 }, { "grad_norm": 0.11270494759082794, "learning_rate": 1.0574478305695863e-05, "loss": 0.0089, "step": 79970 }, { "grad_norm": 0.14985813200473785, "learning_rate": 1.0564311268778154e-05, "loss": 0.0108, "step": 79980 }, { "grad_norm": 0.12860487401485443, "learning_rate": 1.0554148544488073e-05, "loss": 0.0099, "step": 79990 }, { "grad_norm": 0.20634308457374573, "learning_rate": 1.0543990133937004e-05, "loss": 0.0097, "step": 80000 }, { "grad_norm": 0.12352990359067917, "learning_rate": 1.0533836038235868e-05, "loss": 0.008, "step": 80010 }, { "grad_norm": 0.17185981571674347, "learning_rate": 1.052368625849507e-05, "loss": 0.0121, "step": 80020 }, { "grad_norm": 0.30279025435447693, "learning_rate": 1.05135407958246e-05, "loss": 0.0131, "step": 80030 }, { "grad_norm": 0.13310347497463226, "learning_rate": 1.0503399651333934e-05, "loss": 0.0111, "step": 80040 }, { "grad_norm": 0.13664372265338898, "learning_rate": 1.04932628261321e-05, "loss": 0.0107, "step": 80050 }, { "grad_norm": 0.08249207586050034, "learning_rate": 1.0483130321327644e-05, "loss": 0.0097, "step": 80060 }, { "grad_norm": 0.12543001770973206, "learning_rate": 1.0473002138028654e-05, "loss": 0.0133, "step": 80070 }, { "grad_norm": 0.18584923446178436, "learning_rate": 1.0462878277342713e-05, "loss": 0.0094, "step": 80080 }, { "grad_norm": 0.09419246762990952, "learning_rate": 1.0452758740376972e-05, "loss": 0.0126, "step": 80090 }, { "grad_norm": 0.12974046170711517, "learning_rate": 1.0442643528238067e-05, "loss": 0.0095, "step": 80100 }, { "grad_norm": 0.16275496780872345, "learning_rate": 1.0432532642032188e-05, "loss": 0.0097, "step": 80110 }, { "grad_norm": 0.19051958620548248, "learning_rate": 1.042242608286505e-05, "loss": 0.0112, "step": 80120 }, { "grad_norm": 0.136492058634758, "learning_rate": 1.041232385184191e-05, "loss": 0.0089, "step": 80130 }, { "grad_norm": 0.17794425785541534, "learning_rate": 1.0402225950067495e-05, "loss": 0.009, "step": 80140 }, { "grad_norm": 0.18497531116008759, "learning_rate": 1.0392132378646119e-05, "loss": 0.0115, "step": 80150 }, { "grad_norm": 0.11281662434339523, "learning_rate": 1.0382043138681607e-05, "loss": 0.0099, "step": 80160 }, { "grad_norm": 0.24645011126995087, "learning_rate": 1.0371958231277284e-05, "loss": 0.0112, "step": 80170 }, { "grad_norm": 0.16758686304092407, "learning_rate": 1.0361877657536024e-05, "loss": 0.0113, "step": 80180 }, { "grad_norm": 0.1797276884317398, "learning_rate": 1.035180141856023e-05, "loss": 0.0115, "step": 80190 }, { "grad_norm": 0.09252060204744339, "learning_rate": 1.0341729515451836e-05, "loss": 0.0104, "step": 80200 }, { "grad_norm": 0.10251297056674957, "learning_rate": 1.0331661949312265e-05, "loss": 0.0102, "step": 80210 }, { "grad_norm": 0.18142542243003845, "learning_rate": 1.0321598721242514e-05, "loss": 0.0076, "step": 80220 }, { "grad_norm": 0.13770243525505066, "learning_rate": 1.0311539832343064e-05, "loss": 0.0084, "step": 80230 }, { "grad_norm": 0.13177336752414703, "learning_rate": 1.0301485283713942e-05, "loss": 0.0125, "step": 80240 }, { "grad_norm": 0.318622350692749, "learning_rate": 1.0291435076454703e-05, "loss": 0.0142, "step": 80250 }, { "grad_norm": 0.09922700375318527, "learning_rate": 1.028138921166444e-05, "loss": 0.0107, "step": 80260 }, { "grad_norm": 0.10159110277891159, "learning_rate": 1.0271347690441719e-05, "loss": 0.012, "step": 80270 }, { "grad_norm": 0.24546878039836884, "learning_rate": 1.0261310513884698e-05, "loss": 0.0103, "step": 80280 }, { "grad_norm": 0.1962255835533142, "learning_rate": 1.0251277683090998e-05, "loss": 0.0107, "step": 80290 }, { "grad_norm": 0.0809626579284668, "learning_rate": 1.0241249199157826e-05, "loss": 0.0086, "step": 80300 }, { "grad_norm": 0.1706731766462326, "learning_rate": 1.0231225063181837e-05, "loss": 0.0102, "step": 80310 }, { "grad_norm": 0.13966457545757294, "learning_rate": 1.0221205276259316e-05, "loss": 0.0115, "step": 80320 }, { "grad_norm": 0.1040007695555687, "learning_rate": 1.0211189839485958e-05, "loss": 0.0115, "step": 80330 }, { "grad_norm": 0.13201166689395905, "learning_rate": 1.0201178753957075e-05, "loss": 0.0115, "step": 80340 }, { "grad_norm": 0.13214586675167084, "learning_rate": 1.0191172020767432e-05, "loss": 0.0098, "step": 80350 }, { "grad_norm": 0.21413566172122955, "learning_rate": 1.0181169641011362e-05, "loss": 0.0117, "step": 80360 }, { "grad_norm": 0.11214300990104675, "learning_rate": 1.0171171615782721e-05, "loss": 0.0117, "step": 80370 }, { "grad_norm": 0.12097525596618652, "learning_rate": 1.0161177946174866e-05, "loss": 0.01, "step": 80380 }, { "grad_norm": 0.17839273810386658, "learning_rate": 1.0151188633280706e-05, "loss": 0.0084, "step": 80390 }, { "grad_norm": 0.1976800560951233, "learning_rate": 1.0141203678192635e-05, "loss": 0.013, "step": 80400 }, { "grad_norm": 0.16021527349948883, "learning_rate": 1.0131223082002612e-05, "loss": 0.0109, "step": 80410 }, { "grad_norm": 0.15966598689556122, "learning_rate": 1.012124684580208e-05, "loss": 0.0081, "step": 80420 }, { "grad_norm": 0.15684396028518677, "learning_rate": 1.0111274970682039e-05, "loss": 0.0074, "step": 80430 }, { "grad_norm": 0.12601710855960846, "learning_rate": 1.010130745773299e-05, "loss": 0.0113, "step": 80440 }, { "grad_norm": 0.10346533358097076, "learning_rate": 1.0091344308044986e-05, "loss": 0.01, "step": 80450 }, { "grad_norm": 0.10014688968658447, "learning_rate": 1.0081385522707554e-05, "loss": 0.0116, "step": 80460 }, { "grad_norm": 0.17597723007202148, "learning_rate": 1.0071431102809803e-05, "loss": 0.0095, "step": 80470 }, { "grad_norm": 0.14203622937202454, "learning_rate": 1.0061481049440297e-05, "loss": 0.0079, "step": 80480 }, { "grad_norm": 0.15812747180461884, "learning_rate": 1.0051535363687187e-05, "loss": 0.0086, "step": 80490 }, { "grad_norm": 0.16406463086605072, "learning_rate": 1.0041594046638087e-05, "loss": 0.008, "step": 80500 }, { "grad_norm": 0.1563471555709839, "learning_rate": 1.0031657099380204e-05, "loss": 0.0081, "step": 80510 }, { "grad_norm": 0.1308101862668991, "learning_rate": 1.0021724523000204e-05, "loss": 0.0116, "step": 80520 }, { "grad_norm": 0.155845046043396, "learning_rate": 1.001179631858431e-05, "loss": 0.0088, "step": 80530 }, { "grad_norm": 0.11053679138422012, "learning_rate": 1.0001872487218245e-05, "loss": 0.0091, "step": 80540 }, { "grad_norm": 0.15152528882026672, "learning_rate": 9.991953029987272e-06, "loss": 0.0099, "step": 80550 }, { "grad_norm": 0.2202903926372528, "learning_rate": 9.982037947976147e-06, "loss": 0.0121, "step": 80560 }, { "grad_norm": 0.16075600683689117, "learning_rate": 9.972127242269208e-06, "loss": 0.0092, "step": 80570 }, { "grad_norm": 0.22427059710025787, "learning_rate": 9.962220913950238e-06, "loss": 0.0103, "step": 80580 }, { "grad_norm": 0.1347680538892746, "learning_rate": 9.952318964102591e-06, "loss": 0.008, "step": 80590 }, { "grad_norm": 0.15003074705600739, "learning_rate": 9.942421393809148e-06, "loss": 0.008, "step": 80600 }, { "grad_norm": 0.1200588122010231, "learning_rate": 9.932528204152264e-06, "loss": 0.0087, "step": 80610 }, { "grad_norm": 0.19954395294189453, "learning_rate": 9.922639396213856e-06, "loss": 0.0117, "step": 80620 }, { "grad_norm": 0.12590499222278595, "learning_rate": 9.912754971075344e-06, "loss": 0.0115, "step": 80630 }, { "grad_norm": 0.1222679540514946, "learning_rate": 9.902874929817696e-06, "loss": 0.0084, "step": 80640 }, { "grad_norm": 0.12652407586574554, "learning_rate": 9.892999273521342e-06, "loss": 0.0129, "step": 80650 }, { "grad_norm": 0.1894959807395935, "learning_rate": 9.883128003266302e-06, "loss": 0.0111, "step": 80660 }, { "grad_norm": 0.09637544304132462, "learning_rate": 9.87326112013206e-06, "loss": 0.0079, "step": 80670 }, { "grad_norm": 0.197807177901268, "learning_rate": 9.863398625197662e-06, "loss": 0.0097, "step": 80680 }, { "grad_norm": 0.1176917627453804, "learning_rate": 9.853540519541615e-06, "loss": 0.0095, "step": 80690 }, { "grad_norm": 0.1258642077445984, "learning_rate": 9.843686804242053e-06, "loss": 0.0087, "step": 80700 }, { "grad_norm": 0.15195724368095398, "learning_rate": 9.833837480376506e-06, "loss": 0.0103, "step": 80710 }, { "grad_norm": 0.13542594015598297, "learning_rate": 9.82399254902212e-06, "loss": 0.0133, "step": 80720 }, { "grad_norm": 0.13951626420021057, "learning_rate": 9.814152011255494e-06, "loss": 0.0092, "step": 80730 }, { "grad_norm": 0.1231963038444519, "learning_rate": 9.804315868152797e-06, "loss": 0.0094, "step": 80740 }, { "grad_norm": 0.12319900095462799, "learning_rate": 9.79448412078966e-06, "loss": 0.0081, "step": 80750 }, { "grad_norm": 0.12207479774951935, "learning_rate": 9.784656770241312e-06, "loss": 0.0098, "step": 80760 }, { "grad_norm": 0.14746378362178802, "learning_rate": 9.77483381758243e-06, "loss": 0.0083, "step": 80770 }, { "grad_norm": 0.14290840923786163, "learning_rate": 9.76501526388725e-06, "loss": 0.0091, "step": 80780 }, { "grad_norm": 0.14938172698020935, "learning_rate": 9.7552011102295e-06, "loss": 0.0108, "step": 80790 }, { "grad_norm": 0.17619547247886658, "learning_rate": 9.745391357682449e-06, "loss": 0.0101, "step": 80800 }, { "grad_norm": 0.09288472682237625, "learning_rate": 9.735586007318875e-06, "loss": 0.0095, "step": 80810 }, { "grad_norm": 0.19848068058490753, "learning_rate": 9.725785060211096e-06, "loss": 0.0083, "step": 80820 }, { "grad_norm": 0.1978732794523239, "learning_rate": 9.715988517430896e-06, "loss": 0.0085, "step": 80830 }, { "grad_norm": 0.17398206889629364, "learning_rate": 9.706196380049625e-06, "loss": 0.0109, "step": 80840 }, { "grad_norm": 0.29718631505966187, "learning_rate": 9.69640864913815e-06, "loss": 0.0107, "step": 80850 }, { "grad_norm": 0.12839849293231964, "learning_rate": 9.686625325766818e-06, "loss": 0.0126, "step": 80860 }, { "grad_norm": 0.17293676733970642, "learning_rate": 9.67684641100554e-06, "loss": 0.0095, "step": 80870 }, { "grad_norm": 0.13240893185138702, "learning_rate": 9.667071905923692e-06, "loss": 0.0134, "step": 80880 }, { "grad_norm": 0.13064543902873993, "learning_rate": 9.657301811590242e-06, "loss": 0.0116, "step": 80890 }, { "grad_norm": 0.07139847427606583, "learning_rate": 9.647536129073603e-06, "loss": 0.0074, "step": 80900 }, { "grad_norm": 0.17811940610408783, "learning_rate": 9.637774859441751e-06, "loss": 0.014, "step": 80910 }, { "grad_norm": 0.12017826735973358, "learning_rate": 9.628018003762152e-06, "loss": 0.0103, "step": 80920 }, { "grad_norm": 0.1632942110300064, "learning_rate": 9.618265563101813e-06, "loss": 0.0093, "step": 80930 }, { "grad_norm": 0.10570802539587021, "learning_rate": 9.608517538527211e-06, "loss": 0.0086, "step": 80940 }, { "grad_norm": 0.17100508511066437, "learning_rate": 9.598773931104431e-06, "loss": 0.0107, "step": 80950 }, { "grad_norm": 0.1145862489938736, "learning_rate": 9.589034741898983e-06, "loss": 0.0087, "step": 80960 }, { "grad_norm": 0.14779900014400482, "learning_rate": 9.579299971975947e-06, "loss": 0.0077, "step": 80970 }, { "grad_norm": 0.1336865872144699, "learning_rate": 9.569569622399882e-06, "loss": 0.0087, "step": 80980 }, { "grad_norm": 0.2065466195344925, "learning_rate": 9.559843694234916e-06, "loss": 0.0101, "step": 80990 }, { "grad_norm": 0.12019114941358566, "learning_rate": 9.550122188544625e-06, "loss": 0.0087, "step": 81000 }, { "grad_norm": 0.14489254355430603, "learning_rate": 9.540405106392152e-06, "loss": 0.0097, "step": 81010 }, { "grad_norm": 0.1471143513917923, "learning_rate": 9.53069244884015e-06, "loss": 0.0098, "step": 81020 }, { "grad_norm": 0.16022586822509766, "learning_rate": 9.52098421695079e-06, "loss": 0.0098, "step": 81030 }, { "grad_norm": 0.12501610815525055, "learning_rate": 9.511280411785728e-06, "loss": 0.0083, "step": 81040 }, { "grad_norm": 0.09456289559602737, "learning_rate": 9.50158103440616e-06, "loss": 0.0093, "step": 81050 }, { "grad_norm": 0.1923629492521286, "learning_rate": 9.491886085872815e-06, "loss": 0.0117, "step": 81060 }, { "grad_norm": 0.1918875277042389, "learning_rate": 9.482195567245895e-06, "loss": 0.019, "step": 81070 }, { "grad_norm": 0.1505597084760666, "learning_rate": 9.472509479585151e-06, "loss": 0.009, "step": 81080 }, { "grad_norm": 0.1159808486700058, "learning_rate": 9.462827823949833e-06, "loss": 0.0129, "step": 81090 }, { "grad_norm": 0.12322083115577698, "learning_rate": 9.453150601398735e-06, "loss": 0.0092, "step": 81100 }, { "grad_norm": 0.07655302435159683, "learning_rate": 9.44347781299011e-06, "loss": 0.0115, "step": 81110 }, { "grad_norm": 0.158365860581398, "learning_rate": 9.43380945978179e-06, "loss": 0.0119, "step": 81120 }, { "grad_norm": 0.09066488593816757, "learning_rate": 9.424145542831047e-06, "loss": 0.009, "step": 81130 }, { "grad_norm": 0.15395838022232056, "learning_rate": 9.414486063194778e-06, "loss": 0.0092, "step": 81140 }, { "grad_norm": 0.20244979858398438, "learning_rate": 9.404831021929273e-06, "loss": 0.0102, "step": 81150 }, { "grad_norm": 0.17342175543308258, "learning_rate": 9.395180420090427e-06, "loss": 0.0085, "step": 81160 }, { "grad_norm": 0.09635642915964127, "learning_rate": 9.385534258733591e-06, "loss": 0.0079, "step": 81170 }, { "grad_norm": 0.14080016314983368, "learning_rate": 9.375892538913677e-06, "loss": 0.0079, "step": 81180 }, { "grad_norm": 0.15542860329151154, "learning_rate": 9.36625526168507e-06, "loss": 0.0086, "step": 81190 }, { "grad_norm": 0.12901689112186432, "learning_rate": 9.356622428101686e-06, "loss": 0.0102, "step": 81200 }, { "grad_norm": 0.11538612097501755, "learning_rate": 9.346994039216972e-06, "loss": 0.0093, "step": 81210 }, { "grad_norm": 0.15501905977725983, "learning_rate": 9.337370096083886e-06, "loss": 0.011, "step": 81220 }, { "grad_norm": 0.1473708301782608, "learning_rate": 9.327750599754854e-06, "loss": 0.008, "step": 81230 }, { "grad_norm": 0.17698806524276733, "learning_rate": 9.318135551281875e-06, "loss": 0.0088, "step": 81240 }, { "grad_norm": 0.12853148579597473, "learning_rate": 9.308524951716424e-06, "loss": 0.0082, "step": 81250 }, { "grad_norm": 0.16951246559619904, "learning_rate": 9.298918802109497e-06, "loss": 0.0093, "step": 81260 }, { "grad_norm": 0.17019546031951904, "learning_rate": 9.289317103511625e-06, "loss": 0.0083, "step": 81270 }, { "grad_norm": 0.27434492111206055, "learning_rate": 9.279719856972835e-06, "loss": 0.0087, "step": 81280 }, { "grad_norm": 0.1490953266620636, "learning_rate": 9.270127063542645e-06, "loss": 0.0111, "step": 81290 }, { "grad_norm": 0.14871005713939667, "learning_rate": 9.260538724270124e-06, "loss": 0.0072, "step": 81300 }, { "grad_norm": 0.17095807194709778, "learning_rate": 9.250954840203852e-06, "loss": 0.01, "step": 81310 }, { "grad_norm": 0.10703055560588837, "learning_rate": 9.241375412391878e-06, "loss": 0.0075, "step": 81320 }, { "grad_norm": 0.1826460361480713, "learning_rate": 9.23180044188181e-06, "loss": 0.0093, "step": 81330 }, { "grad_norm": 0.13398058712482452, "learning_rate": 9.222229929720755e-06, "loss": 0.0079, "step": 81340 }, { "grad_norm": 0.08165697008371353, "learning_rate": 9.212663876955336e-06, "loss": 0.0079, "step": 81350 }, { "grad_norm": 0.12187009304761887, "learning_rate": 9.20310228463166e-06, "loss": 0.0119, "step": 81360 }, { "grad_norm": 0.12657521665096283, "learning_rate": 9.19354515379539e-06, "loss": 0.0125, "step": 81370 }, { "grad_norm": 0.1194690391421318, "learning_rate": 9.183992485491666e-06, "loss": 0.0086, "step": 81380 }, { "grad_norm": 0.13712386786937714, "learning_rate": 9.174444280765154e-06, "loss": 0.0076, "step": 81390 }, { "grad_norm": 0.10200648754835129, "learning_rate": 9.164900540660032e-06, "loss": 0.0108, "step": 81400 }, { "grad_norm": 0.18847443163394928, "learning_rate": 9.155361266220008e-06, "loss": 0.0094, "step": 81410 }, { "grad_norm": 0.14160789549350739, "learning_rate": 9.145826458488254e-06, "loss": 0.0095, "step": 81420 }, { "grad_norm": 0.1333475410938263, "learning_rate": 9.136296118507509e-06, "loss": 0.0101, "step": 81430 }, { "grad_norm": 0.11337433755397797, "learning_rate": 9.126770247319966e-06, "loss": 0.0122, "step": 81440 }, { "grad_norm": 0.1790134757757187, "learning_rate": 9.11724884596738e-06, "loss": 0.0081, "step": 81450 }, { "grad_norm": 0.1257716417312622, "learning_rate": 9.107731915490996e-06, "loss": 0.0068, "step": 81460 }, { "grad_norm": 0.10834888368844986, "learning_rate": 9.098219456931579e-06, "loss": 0.0132, "step": 81470 }, { "grad_norm": 0.18723079562187195, "learning_rate": 9.088711471329376e-06, "loss": 0.0098, "step": 81480 }, { "grad_norm": 0.12229710817337036, "learning_rate": 9.079207959724184e-06, "loss": 0.0107, "step": 81490 }, { "grad_norm": 0.10709957033395767, "learning_rate": 9.069708923155279e-06, "loss": 0.0076, "step": 81500 }, { "grad_norm": 0.22302885353565216, "learning_rate": 9.060214362661468e-06, "loss": 0.009, "step": 81510 }, { "grad_norm": 0.15853722393512726, "learning_rate": 9.050724279281064e-06, "loss": 0.0106, "step": 81520 }, { "grad_norm": 0.09447019547224045, "learning_rate": 9.041238674051894e-06, "loss": 0.0072, "step": 81530 }, { "grad_norm": 0.12123772501945496, "learning_rate": 9.031757548011277e-06, "loss": 0.0067, "step": 81540 }, { "grad_norm": 0.09930909425020218, "learning_rate": 9.022280902196051e-06, "loss": 0.01, "step": 81550 }, { "grad_norm": 0.15578599274158478, "learning_rate": 9.012808737642592e-06, "loss": 0.0093, "step": 81560 }, { "grad_norm": 0.11986422538757324, "learning_rate": 9.003341055386739e-06, "loss": 0.0106, "step": 81570 }, { "grad_norm": 0.13820765912532806, "learning_rate": 8.993877856463862e-06, "loss": 0.0077, "step": 81580 }, { "grad_norm": 0.0856361910700798, "learning_rate": 8.984419141908856e-06, "loss": 0.0096, "step": 81590 }, { "grad_norm": 0.15328770875930786, "learning_rate": 8.974964912756117e-06, "loss": 0.0096, "step": 81600 }, { "grad_norm": 0.14588862657546997, "learning_rate": 8.965515170039523e-06, "loss": 0.0112, "step": 81610 }, { "grad_norm": 0.1329236775636673, "learning_rate": 8.956069914792514e-06, "loss": 0.01, "step": 81620 }, { "grad_norm": 0.11722850799560547, "learning_rate": 8.946629148047969e-06, "loss": 0.0083, "step": 81630 }, { "grad_norm": 0.1629645973443985, "learning_rate": 8.93719287083834e-06, "loss": 0.0093, "step": 81640 }, { "grad_norm": 0.14485697448253632, "learning_rate": 8.927761084195569e-06, "loss": 0.0091, "step": 81650 }, { "grad_norm": 0.1064460277557373, "learning_rate": 8.918333789151096e-06, "loss": 0.0074, "step": 81660 }, { "grad_norm": 0.17302483320236206, "learning_rate": 8.90891098673587e-06, "loss": 0.0105, "step": 81670 }, { "grad_norm": 0.1727055460214615, "learning_rate": 8.89949267798037e-06, "loss": 0.008, "step": 81680 }, { "grad_norm": 0.13311944901943207, "learning_rate": 8.890078863914542e-06, "loss": 0.0066, "step": 81690 }, { "grad_norm": 0.14902415871620178, "learning_rate": 8.880669545567894e-06, "loss": 0.0085, "step": 81700 }, { "grad_norm": 0.09321542829275131, "learning_rate": 8.871264723969381e-06, "loss": 0.0081, "step": 81710 }, { "grad_norm": 0.11459153145551682, "learning_rate": 8.86186440014754e-06, "loss": 0.0091, "step": 81720 }, { "grad_norm": 0.13891582190990448, "learning_rate": 8.852468575130351e-06, "loss": 0.0085, "step": 81730 }, { "grad_norm": 0.3118378818035126, "learning_rate": 8.843077249945347e-06, "loss": 0.0082, "step": 81740 }, { "grad_norm": 0.08265454322099686, "learning_rate": 8.833690425619523e-06, "loss": 0.0084, "step": 81750 }, { "grad_norm": 0.1632901281118393, "learning_rate": 8.82430810317942e-06, "loss": 0.0084, "step": 81760 }, { "grad_norm": 0.09583524614572525, "learning_rate": 8.814930283651074e-06, "loss": 0.0098, "step": 81770 }, { "grad_norm": 0.12958376109600067, "learning_rate": 8.805556968060048e-06, "loss": 0.0098, "step": 81780 }, { "grad_norm": 0.13622000813484192, "learning_rate": 8.796188157431362e-06, "loss": 0.0096, "step": 81790 }, { "grad_norm": 0.126097172498703, "learning_rate": 8.786823852789595e-06, "loss": 0.0097, "step": 81800 }, { "grad_norm": 0.15599201619625092, "learning_rate": 8.777464055158813e-06, "loss": 0.0105, "step": 81810 }, { "grad_norm": 0.12532661855220795, "learning_rate": 8.768108765562578e-06, "loss": 0.0093, "step": 81820 }, { "grad_norm": 0.09464496374130249, "learning_rate": 8.758757985023975e-06, "loss": 0.009, "step": 81830 }, { "grad_norm": 0.1036892905831337, "learning_rate": 8.749411714565598e-06, "loss": 0.0084, "step": 81840 }, { "grad_norm": 0.1668758988380432, "learning_rate": 8.74006995520954e-06, "loss": 0.0097, "step": 81850 }, { "grad_norm": 0.11795788258314133, "learning_rate": 8.730732707977397e-06, "loss": 0.008, "step": 81860 }, { "grad_norm": 0.13930298388004303, "learning_rate": 8.721399973890281e-06, "loss": 0.0115, "step": 81870 }, { "grad_norm": 0.11657516658306122, "learning_rate": 8.712071753968792e-06, "loss": 0.0076, "step": 81880 }, { "grad_norm": 0.17441211640834808, "learning_rate": 8.702748049233072e-06, "loss": 0.0126, "step": 81890 }, { "grad_norm": 0.283474326133728, "learning_rate": 8.693428860702718e-06, "loss": 0.0102, "step": 81900 }, { "grad_norm": 0.13040822744369507, "learning_rate": 8.684114189396898e-06, "loss": 0.0092, "step": 81910 }, { "grad_norm": 0.1291598379611969, "learning_rate": 8.674804036334222e-06, "loss": 0.0085, "step": 81920 }, { "grad_norm": 0.17346669733524323, "learning_rate": 8.665498402532857e-06, "loss": 0.0114, "step": 81930 }, { "grad_norm": 0.20333638787269592, "learning_rate": 8.656197289010432e-06, "loss": 0.012, "step": 81940 }, { "grad_norm": 0.11004897952079773, "learning_rate": 8.646900696784122e-06, "loss": 0.0077, "step": 81950 }, { "grad_norm": 0.10740943998098373, "learning_rate": 8.637608626870553e-06, "loss": 0.0072, "step": 81960 }, { "grad_norm": 0.10870422422885895, "learning_rate": 8.628321080285944e-06, "loss": 0.0107, "step": 81970 }, { "grad_norm": 0.11414076387882233, "learning_rate": 8.619038058045925e-06, "loss": 0.0098, "step": 81980 }, { "grad_norm": 0.10421491414308548, "learning_rate": 8.609759561165704e-06, "loss": 0.0085, "step": 81990 }, { "grad_norm": 0.1272905021905899, "learning_rate": 8.600485590659934e-06, "loss": 0.0086, "step": 82000 }, { "grad_norm": 0.11753223091363907, "learning_rate": 8.591216147542813e-06, "loss": 0.0075, "step": 82010 }, { "grad_norm": 0.344928115606308, "learning_rate": 8.581951232828045e-06, "loss": 0.0116, "step": 82020 }, { "grad_norm": 0.14367550611495972, "learning_rate": 8.572690847528825e-06, "loss": 0.0086, "step": 82030 }, { "grad_norm": 0.21950563788414001, "learning_rate": 8.563434992657838e-06, "loss": 0.0092, "step": 82040 }, { "grad_norm": 0.12728935480117798, "learning_rate": 8.554183669227307e-06, "loss": 0.0092, "step": 82050 }, { "grad_norm": 0.13437069952487946, "learning_rate": 8.544936878248937e-06, "loss": 0.0096, "step": 82060 }, { "grad_norm": 0.2010834664106369, "learning_rate": 8.535694620733942e-06, "loss": 0.0083, "step": 82070 }, { "grad_norm": 0.15642254054546356, "learning_rate": 8.526456897693053e-06, "loss": 0.0082, "step": 82080 }, { "grad_norm": 0.08315100520849228, "learning_rate": 8.517223710136456e-06, "loss": 0.01, "step": 82090 }, { "grad_norm": 0.09103872627019882, "learning_rate": 8.507995059073931e-06, "loss": 0.0087, "step": 82100 }, { "grad_norm": 0.11688771098852158, "learning_rate": 8.498770945514673e-06, "loss": 0.0073, "step": 82110 }, { "grad_norm": 0.1587596982717514, "learning_rate": 8.489551370467436e-06, "loss": 0.0072, "step": 82120 }, { "grad_norm": 0.1657092571258545, "learning_rate": 8.480336334940442e-06, "loss": 0.011, "step": 82130 }, { "grad_norm": 0.13574399054050446, "learning_rate": 8.471125839941458e-06, "loss": 0.0093, "step": 82140 }, { "grad_norm": 0.11435133963823318, "learning_rate": 8.461919886477682e-06, "loss": 0.0105, "step": 82150 }, { "grad_norm": 0.09481655061244965, "learning_rate": 8.452718475555927e-06, "loss": 0.0086, "step": 82160 }, { "grad_norm": 0.13570967316627502, "learning_rate": 8.443521608182398e-06, "loss": 0.0097, "step": 82170 }, { "grad_norm": 0.13611896336078644, "learning_rate": 8.43432928536288e-06, "loss": 0.0082, "step": 82180 }, { "grad_norm": 0.13902096450328827, "learning_rate": 8.425141508102607e-06, "loss": 0.0093, "step": 82190 }, { "grad_norm": 0.1124328076839447, "learning_rate": 8.415958277406365e-06, "loss": 0.0096, "step": 82200 }, { "grad_norm": 0.11360307782888412, "learning_rate": 8.406779594278392e-06, "loss": 0.0089, "step": 82210 }, { "grad_norm": 0.11768233776092529, "learning_rate": 8.397605459722468e-06, "loss": 0.0081, "step": 82220 }, { "grad_norm": 0.09644544869661331, "learning_rate": 8.388435874741863e-06, "loss": 0.0075, "step": 82230 }, { "grad_norm": 0.13299015164375305, "learning_rate": 8.37927084033936e-06, "loss": 0.0077, "step": 82240 }, { "grad_norm": 0.18307346105575562, "learning_rate": 8.370110357517208e-06, "loss": 0.0069, "step": 82250 }, { "grad_norm": 0.10937755554914474, "learning_rate": 8.360954427277201e-06, "loss": 0.0071, "step": 82260 }, { "grad_norm": 0.19435907900333405, "learning_rate": 8.351803050620627e-06, "loss": 0.0093, "step": 82270 }, { "grad_norm": 0.13474051654338837, "learning_rate": 8.34265622854824e-06, "loss": 0.0075, "step": 82280 }, { "grad_norm": 0.17462417483329773, "learning_rate": 8.333513962060335e-06, "loss": 0.0087, "step": 82290 }, { "grad_norm": 0.15336300432682037, "learning_rate": 8.324376252156701e-06, "loss": 0.0082, "step": 82300 }, { "grad_norm": 0.08998627960681915, "learning_rate": 8.315243099836633e-06, "loss": 0.0072, "step": 82310 }, { "grad_norm": 0.12524092197418213, "learning_rate": 8.3061145060989e-06, "loss": 0.0078, "step": 82320 }, { "grad_norm": 0.11864809691905975, "learning_rate": 8.296990471941806e-06, "loss": 0.0109, "step": 82330 }, { "grad_norm": 0.11462897062301636, "learning_rate": 8.287870998363112e-06, "loss": 0.0091, "step": 82340 }, { "grad_norm": 0.1515873521566391, "learning_rate": 8.278756086360156e-06, "loss": 0.0087, "step": 82350 }, { "grad_norm": 0.13335879147052765, "learning_rate": 8.269645736929693e-06, "loss": 0.0111, "step": 82360 }, { "grad_norm": 0.1599932163953781, "learning_rate": 8.260539951068048e-06, "loss": 0.0135, "step": 82370 }, { "grad_norm": 0.16410522162914276, "learning_rate": 8.251438729770983e-06, "loss": 0.0102, "step": 82380 }, { "grad_norm": 0.13747555017471313, "learning_rate": 8.242342074033826e-06, "loss": 0.0141, "step": 82390 }, { "grad_norm": 0.1812952160835266, "learning_rate": 8.233249984851344e-06, "loss": 0.0118, "step": 82400 }, { "grad_norm": 0.13276879489421844, "learning_rate": 8.224162463217849e-06, "loss": 0.008, "step": 82410 }, { "grad_norm": 0.1866898089647293, "learning_rate": 8.215079510127139e-06, "loss": 0.0095, "step": 82420 }, { "grad_norm": 0.20677503943443298, "learning_rate": 8.206001126572521e-06, "loss": 0.0092, "step": 82430 }, { "grad_norm": 0.17528004944324493, "learning_rate": 8.196927313546776e-06, "loss": 0.0092, "step": 82440 }, { "grad_norm": 0.16225633025169373, "learning_rate": 8.18785807204222e-06, "loss": 0.009, "step": 82450 }, { "grad_norm": 0.12760813534259796, "learning_rate": 8.178793403050627e-06, "loss": 0.0081, "step": 82460 }, { "grad_norm": 0.11273079365491867, "learning_rate": 8.16973330756331e-06, "loss": 0.0112, "step": 82470 }, { "grad_norm": 0.1439342349767685, "learning_rate": 8.16067778657107e-06, "loss": 0.008, "step": 82480 }, { "grad_norm": 0.12219836562871933, "learning_rate": 8.151626841064214e-06, "loss": 0.0141, "step": 82490 }, { "grad_norm": 0.19329001009464264, "learning_rate": 8.142580472032518e-06, "loss": 0.0094, "step": 82500 }, { "grad_norm": 0.1400279849767685, "learning_rate": 8.133538680465285e-06, "loss": 0.0088, "step": 82510 }, { "grad_norm": 0.260120689868927, "learning_rate": 8.124501467351336e-06, "loss": 0.0092, "step": 82520 }, { "grad_norm": 0.15252186357975006, "learning_rate": 8.11546883367893e-06, "loss": 0.0085, "step": 82530 }, { "grad_norm": 0.11518945544958115, "learning_rate": 8.106440780435882e-06, "loss": 0.0124, "step": 82540 }, { "grad_norm": 0.1758388876914978, "learning_rate": 8.097417308609485e-06, "loss": 0.0102, "step": 82550 }, { "grad_norm": 0.16686835885047913, "learning_rate": 8.088398419186544e-06, "loss": 0.0094, "step": 82560 }, { "grad_norm": 0.15136592090129852, "learning_rate": 8.079384113153321e-06, "loss": 0.0076, "step": 82570 }, { "grad_norm": 0.11953562498092651, "learning_rate": 8.070374391495644e-06, "loss": 0.0086, "step": 82580 }, { "grad_norm": 0.10862242430448532, "learning_rate": 8.061369255198764e-06, "loss": 0.0073, "step": 82590 }, { "grad_norm": 0.16940325498580933, "learning_rate": 8.052368705247493e-06, "loss": 0.0092, "step": 82600 }, { "grad_norm": 0.2812436521053314, "learning_rate": 8.043372742626114e-06, "loss": 0.0109, "step": 82610 }, { "grad_norm": 0.2142210602760315, "learning_rate": 8.034381368318422e-06, "loss": 0.0106, "step": 82620 }, { "grad_norm": 0.13673244416713715, "learning_rate": 8.025394583307677e-06, "loss": 0.0104, "step": 82630 }, { "grad_norm": 0.1445879340171814, "learning_rate": 8.016412388576683e-06, "loss": 0.0095, "step": 82640 }, { "grad_norm": 0.15533649921417236, "learning_rate": 8.007434785107699e-06, "loss": 0.0103, "step": 82650 }, { "grad_norm": 0.17615662515163422, "learning_rate": 7.998461773882509e-06, "loss": 0.0108, "step": 82660 }, { "grad_norm": 0.13173629343509674, "learning_rate": 7.989493355882393e-06, "loss": 0.0096, "step": 82670 }, { "grad_norm": 0.07440457493066788, "learning_rate": 7.980529532088137e-06, "loss": 0.0099, "step": 82680 }, { "grad_norm": 0.14164364337921143, "learning_rate": 7.97157030347998e-06, "loss": 0.0106, "step": 82690 }, { "grad_norm": 0.11680776625871658, "learning_rate": 7.96261567103772e-06, "loss": 0.0107, "step": 82700 }, { "grad_norm": 0.09930705279111862, "learning_rate": 7.953665635740593e-06, "loss": 0.0103, "step": 82710 }, { "grad_norm": 0.14425279200077057, "learning_rate": 7.94472019856738e-06, "loss": 0.0082, "step": 82720 }, { "grad_norm": 0.19650982320308685, "learning_rate": 7.935779360496337e-06, "loss": 0.0075, "step": 82730 }, { "grad_norm": 0.08396394550800323, "learning_rate": 7.92684312250523e-06, "loss": 0.0091, "step": 82740 }, { "grad_norm": 0.12971974909305573, "learning_rate": 7.917911485571289e-06, "loss": 0.0106, "step": 82750 }, { "grad_norm": 0.13140404224395752, "learning_rate": 7.90898445067128e-06, "loss": 0.0075, "step": 82760 }, { "grad_norm": 0.12655538320541382, "learning_rate": 7.900062018781462e-06, "loss": 0.0089, "step": 82770 }, { "grad_norm": 0.22119930386543274, "learning_rate": 7.891144190877553e-06, "loss": 0.016, "step": 82780 }, { "grad_norm": 0.10865842550992966, "learning_rate": 7.882230967934806e-06, "loss": 0.0084, "step": 82790 }, { "grad_norm": 0.08579409122467041, "learning_rate": 7.873322350927959e-06, "loss": 0.0077, "step": 82800 }, { "grad_norm": 0.1128859668970108, "learning_rate": 7.864418340831253e-06, "loss": 0.0123, "step": 82810 }, { "grad_norm": 0.14920006692409515, "learning_rate": 7.855518938618395e-06, "loss": 0.0079, "step": 82820 }, { "grad_norm": 0.13567419350147247, "learning_rate": 7.846624145262633e-06, "loss": 0.0125, "step": 82830 }, { "grad_norm": 0.09878791123628616, "learning_rate": 7.837733961736672e-06, "loss": 0.0085, "step": 82840 }, { "grad_norm": 0.18607965111732483, "learning_rate": 7.828848389012727e-06, "loss": 0.0078, "step": 82850 }, { "grad_norm": 0.11468029022216797, "learning_rate": 7.819967428062524e-06, "loss": 0.0108, "step": 82860 }, { "grad_norm": 0.15953023731708527, "learning_rate": 7.811091079857274e-06, "loss": 0.0116, "step": 82870 }, { "grad_norm": 0.1648171842098236, "learning_rate": 7.802219345367667e-06, "loss": 0.0086, "step": 82880 }, { "grad_norm": 0.13859260082244873, "learning_rate": 7.793352225563916e-06, "loss": 0.009, "step": 82890 }, { "grad_norm": 0.12809081375598907, "learning_rate": 7.7844897214157e-06, "loss": 0.0096, "step": 82900 }, { "grad_norm": 0.1339259296655655, "learning_rate": 7.775631833892228e-06, "loss": 0.0097, "step": 82910 }, { "grad_norm": 0.10323884338140488, "learning_rate": 7.76677856396215e-06, "loss": 0.0078, "step": 82920 }, { "grad_norm": 0.12907803058624268, "learning_rate": 7.757929912593697e-06, "loss": 0.0112, "step": 82930 }, { "grad_norm": 0.1257869005203247, "learning_rate": 7.749085880754509e-06, "loss": 0.0101, "step": 82940 }, { "grad_norm": 0.10178607702255249, "learning_rate": 7.740246469411777e-06, "loss": 0.0105, "step": 82950 }, { "grad_norm": 0.15831273794174194, "learning_rate": 7.731411679532142e-06, "loss": 0.0092, "step": 82960 }, { "grad_norm": 0.11880135536193848, "learning_rate": 7.722581512081779e-06, "loss": 0.0104, "step": 82970 }, { "grad_norm": 0.1651073843240738, "learning_rate": 7.713755968026331e-06, "loss": 0.0104, "step": 82980 }, { "grad_norm": 0.14240026473999023, "learning_rate": 7.704935048330969e-06, "loss": 0.0103, "step": 82990 }, { "grad_norm": 0.13035237789154053, "learning_rate": 7.696118753960313e-06, "loss": 0.0082, "step": 83000 }, { "grad_norm": 0.16130028665065765, "learning_rate": 7.687307085878503e-06, "loss": 0.0098, "step": 83010 }, { "grad_norm": 0.12320446223020554, "learning_rate": 7.678500045049186e-06, "loss": 0.0089, "step": 83020 }, { "grad_norm": 0.1428554803133011, "learning_rate": 7.669697632435463e-06, "loss": 0.0084, "step": 83030 }, { "grad_norm": 0.18197336792945862, "learning_rate": 7.660899848999963e-06, "loss": 0.0084, "step": 83040 }, { "grad_norm": 0.14861927926540375, "learning_rate": 7.652106695704803e-06, "loss": 0.0089, "step": 83050 }, { "grad_norm": 0.12388419359922409, "learning_rate": 7.643318173511598e-06, "loss": 0.0072, "step": 83060 }, { "grad_norm": 0.1162344366312027, "learning_rate": 7.634534283381423e-06, "loss": 0.0093, "step": 83070 }, { "grad_norm": 0.10992923378944397, "learning_rate": 7.625755026274889e-06, "loss": 0.0071, "step": 83080 }, { "grad_norm": 0.15160439908504486, "learning_rate": 7.616980403152069e-06, "loss": 0.0101, "step": 83090 }, { "grad_norm": 0.15857496857643127, "learning_rate": 7.608210414972561e-06, "loss": 0.0124, "step": 83100 }, { "grad_norm": 0.1610700488090515, "learning_rate": 7.599445062695404e-06, "loss": 0.0085, "step": 83110 }, { "grad_norm": 0.1346324384212494, "learning_rate": 7.590684347279214e-06, "loss": 0.0089, "step": 83120 }, { "grad_norm": 0.15803436934947968, "learning_rate": 7.581928269682004e-06, "loss": 0.0109, "step": 83130 }, { "grad_norm": 0.10973408818244934, "learning_rate": 7.573176830861361e-06, "loss": 0.0084, "step": 83140 }, { "grad_norm": 0.11796355992555618, "learning_rate": 7.5644300317743e-06, "loss": 0.0079, "step": 83150 }, { "grad_norm": 0.089264877140522, "learning_rate": 7.555687873377376e-06, "loss": 0.01, "step": 83160 }, { "grad_norm": 0.17926821112632751, "learning_rate": 7.546950356626592e-06, "loss": 0.013, "step": 83170 }, { "grad_norm": 0.17850080132484436, "learning_rate": 7.538217482477516e-06, "loss": 0.0094, "step": 83180 }, { "grad_norm": 0.14220330119132996, "learning_rate": 7.529489251885119e-06, "loss": 0.0093, "step": 83190 }, { "grad_norm": 0.11181406676769257, "learning_rate": 7.520765665803941e-06, "loss": 0.0083, "step": 83200 }, { "grad_norm": 0.08978311717510223, "learning_rate": 7.512046725187949e-06, "loss": 0.0098, "step": 83210 }, { "grad_norm": 0.1317121386528015, "learning_rate": 7.503332430990645e-06, "loss": 0.0085, "step": 83220 }, { "grad_norm": 0.1599707156419754, "learning_rate": 7.494622784165017e-06, "loss": 0.0084, "step": 83230 }, { "grad_norm": 0.20596837997436523, "learning_rate": 7.485917785663543e-06, "loss": 0.0135, "step": 83240 }, { "grad_norm": 0.10726824402809143, "learning_rate": 7.477217436438172e-06, "loss": 0.0074, "step": 83250 }, { "grad_norm": 0.19296251237392426, "learning_rate": 7.46852173744037e-06, "loss": 0.0085, "step": 83260 }, { "grad_norm": 0.12269983440637589, "learning_rate": 7.4598306896211004e-06, "loss": 0.0083, "step": 83270 }, { "grad_norm": 0.13545766472816467, "learning_rate": 7.451144293930773e-06, "loss": 0.0088, "step": 83280 }, { "grad_norm": 0.09231115132570267, "learning_rate": 7.4424625513193455e-06, "loss": 0.0087, "step": 83290 }, { "grad_norm": 0.15549270808696747, "learning_rate": 7.433785462736209e-06, "loss": 0.0095, "step": 83300 }, { "grad_norm": 0.10989125818014145, "learning_rate": 7.425113029130315e-06, "loss": 0.009, "step": 83310 }, { "grad_norm": 0.1481945514678955, "learning_rate": 7.416445251450044e-06, "loss": 0.0073, "step": 83320 }, { "grad_norm": 0.16675668954849243, "learning_rate": 7.4077821306433035e-06, "loss": 0.0109, "step": 83330 }, { "grad_norm": 0.14777490496635437, "learning_rate": 7.399123667657465e-06, "loss": 0.0118, "step": 83340 }, { "grad_norm": 0.12636858224868774, "learning_rate": 7.390469863439426e-06, "loss": 0.0089, "step": 83350 }, { "grad_norm": 0.15132088959217072, "learning_rate": 7.381820718935512e-06, "loss": 0.0078, "step": 83360 }, { "grad_norm": 0.1310817152261734, "learning_rate": 7.373176235091633e-06, "loss": 0.0094, "step": 83370 }, { "grad_norm": 0.15570920705795288, "learning_rate": 7.364536412853101e-06, "loss": 0.0103, "step": 83380 }, { "grad_norm": 0.1668984740972519, "learning_rate": 7.355901253164776e-06, "loss": 0.0093, "step": 83390 }, { "grad_norm": 0.20367500185966492, "learning_rate": 7.34727075697097e-06, "loss": 0.0084, "step": 83400 }, { "grad_norm": 0.11491595208644867, "learning_rate": 7.3386449252155095e-06, "loss": 0.0096, "step": 83410 }, { "grad_norm": 0.1489303708076477, "learning_rate": 7.330023758841692e-06, "loss": 0.01, "step": 83420 }, { "grad_norm": 0.09311394393444061, "learning_rate": 7.321407258792323e-06, "loss": 0.0099, "step": 83430 }, { "grad_norm": 0.15812507271766663, "learning_rate": 7.312795426009694e-06, "loss": 0.0082, "step": 83440 }, { "grad_norm": 0.17460674047470093, "learning_rate": 7.304188261435585e-06, "loss": 0.009, "step": 83450 }, { "grad_norm": 0.14846138656139374, "learning_rate": 7.295585766011242e-06, "loss": 0.0088, "step": 83460 }, { "grad_norm": 0.07773713767528534, "learning_rate": 7.286987940677436e-06, "loss": 0.0074, "step": 83470 }, { "grad_norm": 0.17482887208461761, "learning_rate": 7.278394786374426e-06, "loss": 0.0097, "step": 83480 }, { "grad_norm": 0.10731739550828934, "learning_rate": 7.269806304041915e-06, "loss": 0.0083, "step": 83490 }, { "grad_norm": 0.11399063467979431, "learning_rate": 7.261222494619147e-06, "loss": 0.0081, "step": 83500 }, { "grad_norm": 0.17395326495170593, "learning_rate": 7.252643359044825e-06, "loss": 0.0093, "step": 83510 }, { "grad_norm": 0.1283310502767563, "learning_rate": 7.244068898257172e-06, "loss": 0.0078, "step": 83520 }, { "grad_norm": 0.15521714091300964, "learning_rate": 7.2354991131938495e-06, "loss": 0.0116, "step": 83530 }, { "grad_norm": 0.1277291178703308, "learning_rate": 7.2269340047920574e-06, "loss": 0.0121, "step": 83540 }, { "grad_norm": 0.13157430291175842, "learning_rate": 7.218373573988435e-06, "loss": 0.0094, "step": 83550 }, { "grad_norm": 0.11071522533893585, "learning_rate": 7.209817821719178e-06, "loss": 0.0089, "step": 83560 }, { "grad_norm": 0.20666545629501343, "learning_rate": 7.201266748919894e-06, "loss": 0.0109, "step": 83570 }, { "grad_norm": 0.09561269730329514, "learning_rate": 7.19272035652575e-06, "loss": 0.0084, "step": 83580 }, { "grad_norm": 0.12964436411857605, "learning_rate": 7.184178645471334e-06, "loss": 0.01, "step": 83590 }, { "grad_norm": 0.14475496113300323, "learning_rate": 7.175641616690775e-06, "loss": 0.0076, "step": 83600 }, { "grad_norm": 0.12938980758190155, "learning_rate": 7.167109271117661e-06, "loss": 0.0092, "step": 83610 }, { "grad_norm": 0.16043519973754883, "learning_rate": 7.158581609685072e-06, "loss": 0.0091, "step": 83620 }, { "grad_norm": 0.20113986730575562, "learning_rate": 7.150058633325591e-06, "loss": 0.0104, "step": 83630 }, { "grad_norm": 0.12118496000766754, "learning_rate": 7.141540342971287e-06, "loss": 0.0091, "step": 83640 }, { "grad_norm": 0.24291744828224182, "learning_rate": 7.133026739553677e-06, "loss": 0.0116, "step": 83650 }, { "grad_norm": 0.11653067916631699, "learning_rate": 7.124517824003829e-06, "loss": 0.0092, "step": 83660 }, { "grad_norm": 0.18136544525623322, "learning_rate": 7.116013597252236e-06, "loss": 0.0086, "step": 83670 }, { "grad_norm": 0.13336946070194244, "learning_rate": 7.107514060228921e-06, "loss": 0.0078, "step": 83680 }, { "grad_norm": 0.1319541335105896, "learning_rate": 7.099019213863378e-06, "loss": 0.0107, "step": 83690 }, { "grad_norm": 0.13220268487930298, "learning_rate": 7.090529059084611e-06, "loss": 0.0104, "step": 83700 }, { "grad_norm": 0.14429424703121185, "learning_rate": 7.082043596821058e-06, "loss": 0.0107, "step": 83710 }, { "grad_norm": 0.20201599597930908, "learning_rate": 7.07356282800069e-06, "loss": 0.0079, "step": 83720 }, { "grad_norm": 0.12880352139472961, "learning_rate": 7.0650867535509645e-06, "loss": 0.0084, "step": 83730 }, { "grad_norm": 0.1144728884100914, "learning_rate": 7.0566153743987895e-06, "loss": 0.0082, "step": 83740 }, { "grad_norm": 0.11797294020652771, "learning_rate": 7.04814869147059e-06, "loss": 0.0124, "step": 83750 }, { "grad_norm": 0.19219817221164703, "learning_rate": 7.0396867056922725e-06, "loss": 0.0122, "step": 83760 }, { "grad_norm": 0.12748733162879944, "learning_rate": 7.031229417989244e-06, "loss": 0.0073, "step": 83770 }, { "grad_norm": 0.1685764640569687, "learning_rate": 7.022776829286348e-06, "loss": 0.0105, "step": 83780 }, { "grad_norm": 0.13064229488372803, "learning_rate": 7.014328940507975e-06, "loss": 0.0078, "step": 83790 }, { "grad_norm": 0.10796958208084106, "learning_rate": 7.005885752577945e-06, "loss": 0.0097, "step": 83800 }, { "grad_norm": 0.15168194472789764, "learning_rate": 6.997447266419615e-06, "loss": 0.0116, "step": 83810 }, { "grad_norm": 0.12891364097595215, "learning_rate": 6.989013482955797e-06, "loss": 0.0092, "step": 83820 }, { "grad_norm": 0.1657608449459076, "learning_rate": 6.980584403108808e-06, "loss": 0.0083, "step": 83830 }, { "grad_norm": 0.13171721994876862, "learning_rate": 6.972160027800417e-06, "loss": 0.0098, "step": 83840 }, { "grad_norm": 0.1910334974527359, "learning_rate": 6.963740357951931e-06, "loss": 0.0095, "step": 83850 }, { "grad_norm": 0.1878986358642578, "learning_rate": 6.95532539448408e-06, "loss": 0.0092, "step": 83860 }, { "grad_norm": 0.1367952674627304, "learning_rate": 6.946915138317128e-06, "loss": 0.0088, "step": 83870 }, { "grad_norm": 0.1565781682729721, "learning_rate": 6.9385095903708065e-06, "loss": 0.0094, "step": 83880 }, { "grad_norm": 0.11339256912469864, "learning_rate": 6.930108751564346e-06, "loss": 0.0081, "step": 83890 }, { "grad_norm": 0.15476049482822418, "learning_rate": 6.921712622816428e-06, "loss": 0.0091, "step": 83900 }, { "grad_norm": 0.1991644650697708, "learning_rate": 6.913321205045259e-06, "loss": 0.0109, "step": 83910 }, { "grad_norm": 0.3025772273540497, "learning_rate": 6.904934499168491e-06, "loss": 0.0083, "step": 83920 }, { "grad_norm": 0.10250880569219589, "learning_rate": 6.896552506103299e-06, "loss": 0.008, "step": 83930 }, { "grad_norm": 0.12054718285799026, "learning_rate": 6.888175226766313e-06, "loss": 0.0085, "step": 83940 }, { "grad_norm": 0.14832130074501038, "learning_rate": 6.879802662073686e-06, "loss": 0.0101, "step": 83950 }, { "grad_norm": 0.13832546770572662, "learning_rate": 6.871434812940991e-06, "loss": 0.009, "step": 83960 }, { "grad_norm": 0.13833144307136536, "learning_rate": 6.863071680283345e-06, "loss": 0.007, "step": 83970 }, { "grad_norm": 0.11561109870672226, "learning_rate": 6.854713265015339e-06, "loss": 0.0102, "step": 83980 }, { "grad_norm": 0.13750728964805603, "learning_rate": 6.8463595680510025e-06, "loss": 0.0077, "step": 83990 }, { "grad_norm": 0.15626588463783264, "learning_rate": 6.838010590303911e-06, "loss": 0.0084, "step": 84000 }, { "grad_norm": 0.47072336077690125, "learning_rate": 6.829666332687085e-06, "loss": 0.0106, "step": 84010 }, { "grad_norm": 0.1565159410238266, "learning_rate": 6.821326796113059e-06, "loss": 0.0087, "step": 84020 }, { "grad_norm": 0.11630279570817947, "learning_rate": 6.812991981493799e-06, "loss": 0.0078, "step": 84030 }, { "grad_norm": 0.24813085794448853, "learning_rate": 6.804661889740821e-06, "loss": 0.0086, "step": 84040 }, { "grad_norm": 0.11341580003499985, "learning_rate": 6.7963365217650625e-06, "loss": 0.0088, "step": 84050 }, { "grad_norm": 0.12500408291816711, "learning_rate": 6.788015878476983e-06, "loss": 0.0104, "step": 84060 }, { "grad_norm": 0.38888975977897644, "learning_rate": 6.779699960786523e-06, "loss": 0.01, "step": 84070 }, { "grad_norm": 0.15786515176296234, "learning_rate": 6.771388769603104e-06, "loss": 0.0079, "step": 84080 }, { "grad_norm": 0.1278238445520401, "learning_rate": 6.763082305835605e-06, "loss": 0.0069, "step": 84090 }, { "grad_norm": 0.1386808604001999, "learning_rate": 6.754780570392433e-06, "loss": 0.0084, "step": 84100 }, { "grad_norm": 0.1698085069656372, "learning_rate": 6.7464835641814215e-06, "loss": 0.0093, "step": 84110 }, { "grad_norm": 0.21396435797214508, "learning_rate": 6.7381912881099515e-06, "loss": 0.0108, "step": 84120 }, { "grad_norm": 0.10818452388048172, "learning_rate": 6.72990374308482e-06, "loss": 0.0085, "step": 84130 }, { "grad_norm": 0.13971716165542603, "learning_rate": 6.721620930012379e-06, "loss": 0.0095, "step": 84140 }, { "grad_norm": 0.10019689053297043, "learning_rate": 6.713342849798393e-06, "loss": 0.0094, "step": 84150 }, { "grad_norm": 0.16301701962947845, "learning_rate": 6.705069503348166e-06, "loss": 0.0102, "step": 84160 }, { "grad_norm": 0.11416362226009369, "learning_rate": 6.696800891566435e-06, "loss": 0.0092, "step": 84170 }, { "grad_norm": 0.23197707533836365, "learning_rate": 6.688537015357449e-06, "loss": 0.0107, "step": 84180 }, { "grad_norm": 0.11981258541345596, "learning_rate": 6.680277875624941e-06, "loss": 0.0093, "step": 84190 }, { "grad_norm": 0.1674700379371643, "learning_rate": 6.672023473272121e-06, "loss": 0.0085, "step": 84200 }, { "grad_norm": 0.10515854507684708, "learning_rate": 6.663773809201667e-06, "loss": 0.0074, "step": 84210 }, { "grad_norm": 0.162069633603096, "learning_rate": 6.655528884315748e-06, "loss": 0.0092, "step": 84220 }, { "grad_norm": 0.18176637589931488, "learning_rate": 6.647288699516036e-06, "loss": 0.0072, "step": 84230 }, { "grad_norm": 0.15362347662448883, "learning_rate": 6.639053255703642e-06, "loss": 0.0081, "step": 84240 }, { "grad_norm": 0.10246239602565765, "learning_rate": 6.6308225537791925e-06, "loss": 0.0086, "step": 84250 }, { "grad_norm": 0.09495416283607483, "learning_rate": 6.622596594642777e-06, "loss": 0.0087, "step": 84260 }, { "grad_norm": 0.18475717306137085, "learning_rate": 6.614375379193999e-06, "loss": 0.0064, "step": 84270 }, { "grad_norm": 0.11028683185577393, "learning_rate": 6.606158908331889e-06, "loss": 0.0085, "step": 84280 }, { "grad_norm": 0.1819302886724472, "learning_rate": 6.597947182955005e-06, "loss": 0.0103, "step": 84290 }, { "grad_norm": 0.1516631841659546, "learning_rate": 6.589740203961359e-06, "loss": 0.0089, "step": 84300 }, { "grad_norm": 0.12650543451309204, "learning_rate": 6.581537972248464e-06, "loss": 0.0086, "step": 84310 }, { "grad_norm": 0.13347093760967255, "learning_rate": 6.573340488713275e-06, "loss": 0.0096, "step": 84320 }, { "grad_norm": 0.1027422696352005, "learning_rate": 6.565147754252304e-06, "loss": 0.0093, "step": 84330 }, { "grad_norm": 0.16561947762966156, "learning_rate": 6.556959769761462e-06, "loss": 0.0094, "step": 84340 }, { "grad_norm": 0.11991522461175919, "learning_rate": 6.548776536136192e-06, "loss": 0.0107, "step": 84350 }, { "grad_norm": 0.17590965330600739, "learning_rate": 6.54059805427138e-06, "loss": 0.0088, "step": 84360 }, { "grad_norm": 0.11242212355136871, "learning_rate": 6.532424325061437e-06, "loss": 0.0079, "step": 84370 }, { "grad_norm": 0.15785016119480133, "learning_rate": 6.5242553494001935e-06, "loss": 0.0095, "step": 84380 }, { "grad_norm": 0.11289041489362717, "learning_rate": 6.516091128181046e-06, "loss": 0.0098, "step": 84390 }, { "grad_norm": 0.20684078335762024, "learning_rate": 6.507931662296785e-06, "loss": 0.0108, "step": 84400 }, { "grad_norm": 0.19578608870506287, "learning_rate": 6.4997769526397354e-06, "loss": 0.0098, "step": 84410 }, { "grad_norm": 0.18207715451717377, "learning_rate": 6.491627000101663e-06, "loss": 0.0094, "step": 84420 }, { "grad_norm": 0.13240376114845276, "learning_rate": 6.4834818055738545e-06, "loss": 0.0061, "step": 84430 }, { "grad_norm": 0.10447896271944046, "learning_rate": 6.4753413699470465e-06, "loss": 0.0091, "step": 84440 }, { "grad_norm": 0.14610034227371216, "learning_rate": 6.467205694111478e-06, "loss": 0.0092, "step": 84450 }, { "grad_norm": 0.15231384336948395, "learning_rate": 6.459074778956836e-06, "loss": 0.0097, "step": 84460 }, { "grad_norm": 0.11083280295133591, "learning_rate": 6.450948625372311e-06, "loss": 0.0092, "step": 84470 }, { "grad_norm": 0.2220611572265625, "learning_rate": 6.442827234246585e-06, "loss": 0.0095, "step": 84480 }, { "grad_norm": 0.11321971565485, "learning_rate": 6.43471060646777e-06, "loss": 0.0082, "step": 84490 }, { "grad_norm": 0.19301103055477142, "learning_rate": 6.426598742923512e-06, "loss": 0.008, "step": 84500 }, { "grad_norm": 0.19844457507133484, "learning_rate": 6.418491644500885e-06, "loss": 0.0092, "step": 84510 }, { "grad_norm": 0.13377517461776733, "learning_rate": 6.410389312086512e-06, "loss": 0.009, "step": 84520 }, { "grad_norm": 0.23520071804523468, "learning_rate": 6.4022917465664135e-06, "loss": 0.0111, "step": 84530 }, { "grad_norm": 0.20659779012203217, "learning_rate": 6.3941989488261575e-06, "loss": 0.0101, "step": 84540 }, { "grad_norm": 0.1638440489768982, "learning_rate": 6.386110919750726e-06, "loss": 0.0084, "step": 84550 }, { "grad_norm": 0.15567496418952942, "learning_rate": 6.378027660224651e-06, "loss": 0.009, "step": 84560 }, { "grad_norm": 0.09010389447212219, "learning_rate": 6.369949171131856e-06, "loss": 0.0071, "step": 84570 }, { "grad_norm": 0.08775009214878082, "learning_rate": 6.361875453355848e-06, "loss": 0.0087, "step": 84580 }, { "grad_norm": 0.17441776394844055, "learning_rate": 6.353806507779525e-06, "loss": 0.0107, "step": 84590 }, { "grad_norm": 0.11968646198511124, "learning_rate": 6.3457423352853044e-06, "loss": 0.0103, "step": 84600 }, { "grad_norm": 0.1830121874809265, "learning_rate": 6.337682936755062e-06, "loss": 0.0083, "step": 84610 }, { "grad_norm": 0.17974510788917542, "learning_rate": 6.329628313070179e-06, "loss": 0.01, "step": 84620 }, { "grad_norm": 0.23947550356388092, "learning_rate": 6.321578465111477e-06, "loss": 0.0123, "step": 84630 }, { "grad_norm": 0.14347292482852936, "learning_rate": 6.31353339375928e-06, "loss": 0.01, "step": 84640 }, { "grad_norm": 0.11958977580070496, "learning_rate": 6.305493099893395e-06, "loss": 0.0092, "step": 84650 }, { "grad_norm": 0.17425177991390228, "learning_rate": 6.297457584393096e-06, "loss": 0.0097, "step": 84660 }, { "grad_norm": 0.12558667361736298, "learning_rate": 6.289426848137126e-06, "loss": 0.0108, "step": 84670 }, { "grad_norm": 0.1888613998889923, "learning_rate": 6.281400892003713e-06, "loss": 0.0084, "step": 84680 }, { "grad_norm": 0.3206470012664795, "learning_rate": 6.273379716870581e-06, "loss": 0.0101, "step": 84690 }, { "grad_norm": 0.13549938797950745, "learning_rate": 6.265363323614887e-06, "loss": 0.0115, "step": 84700 }, { "grad_norm": 0.1210380345582962, "learning_rate": 6.257351713113308e-06, "loss": 0.007, "step": 84710 }, { "grad_norm": 0.19357912242412567, "learning_rate": 6.249344886241981e-06, "loss": 0.0091, "step": 84720 }, { "grad_norm": 0.15853963792324066, "learning_rate": 6.2413428438765295e-06, "loss": 0.0128, "step": 84730 }, { "grad_norm": 0.15657314658164978, "learning_rate": 6.233345586892026e-06, "loss": 0.0098, "step": 84740 }, { "grad_norm": 0.15284141898155212, "learning_rate": 6.225353116163052e-06, "loss": 0.0118, "step": 84750 }, { "grad_norm": 0.10350769013166428, "learning_rate": 6.217365432563627e-06, "loss": 0.0095, "step": 84760 }, { "grad_norm": 0.11042211949825287, "learning_rate": 6.209382536967318e-06, "loss": 0.0087, "step": 84770 }, { "grad_norm": 0.16746893525123596, "learning_rate": 6.201404430247082e-06, "loss": 0.0113, "step": 84780 }, { "grad_norm": 0.10656511783599854, "learning_rate": 6.1934311132754165e-06, "loss": 0.0077, "step": 84790 }, { "grad_norm": 0.08698882907629013, "learning_rate": 6.185462586924257e-06, "loss": 0.0076, "step": 84800 }, { "grad_norm": 0.10932394862174988, "learning_rate": 6.177498852065039e-06, "loss": 0.0076, "step": 84810 }, { "grad_norm": 0.17133782804012299, "learning_rate": 6.169539909568655e-06, "loss": 0.0117, "step": 84820 }, { "grad_norm": 0.15483587980270386, "learning_rate": 6.161585760305488e-06, "loss": 0.0086, "step": 84830 }, { "grad_norm": 0.15829041600227356, "learning_rate": 6.153636405145391e-06, "loss": 0.0091, "step": 84840 }, { "grad_norm": 0.17867320775985718, "learning_rate": 6.145691844957702e-06, "loss": 0.0087, "step": 84850 }, { "grad_norm": 0.12556874752044678, "learning_rate": 6.1377520806112145e-06, "loss": 0.0128, "step": 84860 }, { "grad_norm": 0.14372509717941284, "learning_rate": 6.129817112974218e-06, "loss": 0.0069, "step": 84870 }, { "grad_norm": 0.1507088541984558, "learning_rate": 6.121886942914457e-06, "loss": 0.0076, "step": 84880 }, { "grad_norm": 0.1308446228504181, "learning_rate": 6.11396157129917e-06, "loss": 0.0104, "step": 84890 }, { "grad_norm": 0.11130771040916443, "learning_rate": 6.106040998995066e-06, "loss": 0.0083, "step": 84900 }, { "grad_norm": 0.23681038618087769, "learning_rate": 6.098125226868329e-06, "loss": 0.0114, "step": 84910 }, { "grad_norm": 0.13658727705478668, "learning_rate": 6.090214255784604e-06, "loss": 0.0097, "step": 84920 }, { "grad_norm": 0.1540914624929428, "learning_rate": 6.0823080866090346e-06, "loss": 0.0105, "step": 84930 }, { "grad_norm": 0.14224129915237427, "learning_rate": 6.074406720206233e-06, "loss": 0.0101, "step": 84940 }, { "grad_norm": 0.18702855706214905, "learning_rate": 6.066510157440259e-06, "loss": 0.0083, "step": 84950 }, { "grad_norm": 0.178900346159935, "learning_rate": 6.058618399174676e-06, "loss": 0.0112, "step": 84960 }, { "grad_norm": 0.09083041548728943, "learning_rate": 6.050731446272528e-06, "loss": 0.008, "step": 84970 }, { "grad_norm": 0.09671337902545929, "learning_rate": 6.042849299596321e-06, "loss": 0.0099, "step": 84980 }, { "grad_norm": 0.11114373058080673, "learning_rate": 6.034971960008013e-06, "loss": 0.0108, "step": 84990 }, { "grad_norm": 0.17308583855628967, "learning_rate": 6.027099428369082e-06, "loss": 0.0109, "step": 85000 }, { "grad_norm": 0.1702135056257248, "learning_rate": 6.019231705540435e-06, "loss": 0.008, "step": 85010 }, { "grad_norm": 0.17259785532951355, "learning_rate": 6.011368792382488e-06, "loss": 0.0089, "step": 85020 }, { "grad_norm": 0.16400043666362762, "learning_rate": 6.00351068975511e-06, "loss": 0.0099, "step": 85030 }, { "grad_norm": 0.10435555130243301, "learning_rate": 5.9956573985176674e-06, "loss": 0.01, "step": 85040 }, { "grad_norm": 0.10991272330284119, "learning_rate": 5.987808919528959e-06, "loss": 0.0075, "step": 85050 }, { "grad_norm": 0.19023080170154572, "learning_rate": 5.979965253647307e-06, "loss": 0.0095, "step": 85060 }, { "grad_norm": 0.14266780018806458, "learning_rate": 5.9721264017304655e-06, "loss": 0.0072, "step": 85070 }, { "grad_norm": 0.11839074641466141, "learning_rate": 5.964292364635682e-06, "loss": 0.0086, "step": 85080 }, { "grad_norm": 0.09595412760972977, "learning_rate": 5.956463143219682e-06, "loss": 0.0081, "step": 85090 }, { "grad_norm": 0.13128377497196198, "learning_rate": 5.948638738338663e-06, "loss": 0.0072, "step": 85100 }, { "grad_norm": 0.1058039665222168, "learning_rate": 5.940819150848271e-06, "loss": 0.0076, "step": 85110 }, { "grad_norm": 0.11951804161071777, "learning_rate": 5.933004381603663e-06, "loss": 0.0079, "step": 85120 }, { "grad_norm": 0.23668183386325836, "learning_rate": 5.92519443145943e-06, "loss": 0.0094, "step": 85130 }, { "grad_norm": 0.18755723536014557, "learning_rate": 5.917389301269671e-06, "loss": 0.007, "step": 85140 }, { "grad_norm": 0.2581532895565033, "learning_rate": 5.9095889918879435e-06, "loss": 0.011, "step": 85150 }, { "grad_norm": 0.12067648768424988, "learning_rate": 5.901793504167286e-06, "loss": 0.0074, "step": 85160 }, { "grad_norm": 0.15810614824295044, "learning_rate": 5.894002838960172e-06, "loss": 0.0107, "step": 85170 }, { "grad_norm": 0.15686960518360138, "learning_rate": 5.886216997118604e-06, "loss": 0.0083, "step": 85180 }, { "grad_norm": 0.2897896468639374, "learning_rate": 5.878435979494023e-06, "loss": 0.0145, "step": 85190 }, { "grad_norm": 0.11352957040071487, "learning_rate": 5.870659786937344e-06, "loss": 0.0086, "step": 85200 }, { "grad_norm": 0.16188053786754608, "learning_rate": 5.862888420298962e-06, "loss": 0.0088, "step": 85210 }, { "grad_norm": 0.15827523171901703, "learning_rate": 5.855121880428743e-06, "loss": 0.008, "step": 85220 }, { "grad_norm": 0.11056942492723465, "learning_rate": 5.84736016817603e-06, "loss": 0.0082, "step": 85230 }, { "grad_norm": 0.18593798577785492, "learning_rate": 5.839603284389622e-06, "loss": 0.0099, "step": 85240 }, { "grad_norm": 0.10476137697696686, "learning_rate": 5.831851229917812e-06, "loss": 0.0114, "step": 85250 }, { "grad_norm": 0.18799535930156708, "learning_rate": 5.824104005608338e-06, "loss": 0.011, "step": 85260 }, { "grad_norm": 0.1522839069366455, "learning_rate": 5.81636161230843e-06, "loss": 0.0086, "step": 85270 }, { "grad_norm": 0.09847336262464523, "learning_rate": 5.808624050864791e-06, "loss": 0.0088, "step": 85280 }, { "grad_norm": 0.12188737094402313, "learning_rate": 5.800891322123592e-06, "loss": 0.0085, "step": 85290 }, { "grad_norm": 0.11242533475160599, "learning_rate": 5.793163426930454e-06, "loss": 0.0101, "step": 85300 }, { "grad_norm": 0.18702897429466248, "learning_rate": 5.785440366130512e-06, "loss": 0.0081, "step": 85310 }, { "grad_norm": 0.10072290152311325, "learning_rate": 5.777722140568326e-06, "loss": 0.0078, "step": 85320 }, { "grad_norm": 0.11550584435462952, "learning_rate": 5.7700087510879685e-06, "loss": 0.0083, "step": 85330 }, { "grad_norm": 0.2468995451927185, "learning_rate": 5.762300198532933e-06, "loss": 0.0082, "step": 85340 }, { "grad_norm": 0.08601603657007217, "learning_rate": 5.7545964837462606e-06, "loss": 0.0093, "step": 85350 }, { "grad_norm": 0.17613187432289124, "learning_rate": 5.746897607570384e-06, "loss": 0.0111, "step": 85360 }, { "grad_norm": 0.1720331609249115, "learning_rate": 5.7392035708472626e-06, "loss": 0.0085, "step": 85370 }, { "grad_norm": 0.09701427072286606, "learning_rate": 5.731514374418278e-06, "loss": 0.0102, "step": 85380 }, { "grad_norm": 0.12889209389686584, "learning_rate": 5.7238300191243325e-06, "loss": 0.0089, "step": 85390 }, { "grad_norm": 0.2845468819141388, "learning_rate": 5.716150505805762e-06, "loss": 0.0122, "step": 85400 }, { "grad_norm": 0.1265772581100464, "learning_rate": 5.708475835302407e-06, "loss": 0.0094, "step": 85410 }, { "grad_norm": 0.2266370952129364, "learning_rate": 5.70080600845353e-06, "loss": 0.011, "step": 85420 }, { "grad_norm": 0.14410929381847382, "learning_rate": 5.693141026097909e-06, "loss": 0.0094, "step": 85430 }, { "grad_norm": 0.1577559858560562, "learning_rate": 5.6854808890737745e-06, "loss": 0.0092, "step": 85440 }, { "grad_norm": 0.17090469598770142, "learning_rate": 5.6778255982188225e-06, "loss": 0.01, "step": 85450 }, { "grad_norm": 0.12265585362911224, "learning_rate": 5.670175154370217e-06, "loss": 0.0086, "step": 85460 }, { "grad_norm": 0.11161148548126221, "learning_rate": 5.662529558364616e-06, "loss": 0.0089, "step": 85470 }, { "grad_norm": 0.20696479082107544, "learning_rate": 5.654888811038128e-06, "loss": 0.0111, "step": 85480 }, { "grad_norm": 0.1320701241493225, "learning_rate": 5.647252913226314e-06, "loss": 0.0071, "step": 85490 }, { "grad_norm": 0.09786001592874527, "learning_rate": 5.639621865764255e-06, "loss": 0.0083, "step": 85500 }, { "grad_norm": 0.36355993151664734, "learning_rate": 5.63199566948644e-06, "loss": 0.0109, "step": 85510 }, { "grad_norm": 0.08741098642349243, "learning_rate": 5.6243743252268835e-06, "loss": 0.0067, "step": 85520 }, { "grad_norm": 0.11178857088088989, "learning_rate": 5.616757833819009e-06, "loss": 0.0073, "step": 85530 }, { "grad_norm": 0.15425211191177368, "learning_rate": 5.609146196095794e-06, "loss": 0.0075, "step": 85540 }, { "grad_norm": 0.12215278297662735, "learning_rate": 5.6015394128896005e-06, "loss": 0.0097, "step": 85550 }, { "grad_norm": 0.12027543038129807, "learning_rate": 5.593937485032308e-06, "loss": 0.009, "step": 85560 }, { "grad_norm": 0.1280726194381714, "learning_rate": 5.586340413355246e-06, "loss": 0.0095, "step": 85570 }, { "grad_norm": 0.1492101550102234, "learning_rate": 5.578748198689227e-06, "loss": 0.0093, "step": 85580 }, { "grad_norm": 0.1253730058670044, "learning_rate": 5.571160841864504e-06, "loss": 0.0097, "step": 85590 }, { "grad_norm": 0.13190267980098724, "learning_rate": 5.563578343710846e-06, "loss": 0.0113, "step": 85600 }, { "grad_norm": 0.1618739515542984, "learning_rate": 5.556000705057446e-06, "loss": 0.0103, "step": 85610 }, { "grad_norm": 0.15208543837070465, "learning_rate": 5.548427926732996e-06, "loss": 0.0073, "step": 85620 }, { "grad_norm": 0.1843421906232834, "learning_rate": 5.540860009565629e-06, "loss": 0.012, "step": 85630 }, { "grad_norm": 0.2778920531272888, "learning_rate": 5.533296954382966e-06, "loss": 0.0095, "step": 85640 }, { "grad_norm": 0.13818667829036713, "learning_rate": 5.525738762012095e-06, "loss": 0.0084, "step": 85650 }, { "grad_norm": 0.15077783167362213, "learning_rate": 5.518185433279577e-06, "loss": 0.0086, "step": 85660 }, { "grad_norm": 0.23587757349014282, "learning_rate": 5.510636969011407e-06, "loss": 0.0102, "step": 85670 }, { "grad_norm": 0.15420575439929962, "learning_rate": 5.503093370033096e-06, "loss": 0.0093, "step": 85680 }, { "grad_norm": 0.10135509073734283, "learning_rate": 5.495554637169603e-06, "loss": 0.0075, "step": 85690 }, { "grad_norm": 0.15358766913414001, "learning_rate": 5.488020771245328e-06, "loss": 0.0095, "step": 85700 }, { "grad_norm": 0.15416158735752106, "learning_rate": 5.480491773084195e-06, "loss": 0.0078, "step": 85710 }, { "grad_norm": 0.13646961748600006, "learning_rate": 5.472967643509525e-06, "loss": 0.0093, "step": 85720 }, { "grad_norm": 0.15989796817302704, "learning_rate": 5.46544838334419e-06, "loss": 0.0064, "step": 85730 }, { "grad_norm": 0.16578465700149536, "learning_rate": 5.4579339934104525e-06, "loss": 0.0083, "step": 85740 }, { "grad_norm": 0.12510322034358978, "learning_rate": 5.450424474530091e-06, "loss": 0.0065, "step": 85750 }, { "grad_norm": 0.1320279985666275, "learning_rate": 5.442919827524323e-06, "loss": 0.0111, "step": 85760 }, { "grad_norm": 0.08539526909589767, "learning_rate": 5.435420053213863e-06, "loss": 0.0071, "step": 85770 }, { "grad_norm": 0.1282196193933487, "learning_rate": 5.427925152418845e-06, "loss": 0.0091, "step": 85780 }, { "grad_norm": 0.09881187230348587, "learning_rate": 5.420435125958934e-06, "loss": 0.0098, "step": 85790 }, { "grad_norm": 0.10419464111328125, "learning_rate": 5.412949974653209e-06, "loss": 0.0098, "step": 85800 }, { "grad_norm": 0.06685648113489151, "learning_rate": 5.405469699320248e-06, "loss": 0.0085, "step": 85810 }, { "grad_norm": 0.18146568536758423, "learning_rate": 5.397994300778064e-06, "loss": 0.008, "step": 85820 }, { "grad_norm": 0.14762680232524872, "learning_rate": 5.39052377984417e-06, "loss": 0.0083, "step": 85830 }, { "grad_norm": 0.12421954423189163, "learning_rate": 5.383058137335523e-06, "loss": 0.0104, "step": 85840 }, { "grad_norm": 0.12571115791797638, "learning_rate": 5.375597374068553e-06, "loss": 0.0083, "step": 85850 }, { "grad_norm": 0.09490824490785599, "learning_rate": 5.368141490859163e-06, "loss": 0.0088, "step": 85860 }, { "grad_norm": 0.1666821986436844, "learning_rate": 5.360690488522729e-06, "loss": 0.0077, "step": 85870 }, { "grad_norm": 0.2194025069475174, "learning_rate": 5.353244367874055e-06, "loss": 0.0095, "step": 85880 }, { "grad_norm": 0.11982092261314392, "learning_rate": 5.345803129727456e-06, "loss": 0.0073, "step": 85890 }, { "grad_norm": 0.11648871749639511, "learning_rate": 5.338366774896697e-06, "loss": 0.0086, "step": 85900 }, { "grad_norm": 0.1946120709180832, "learning_rate": 5.330935304194995e-06, "loss": 0.0141, "step": 85910 }, { "grad_norm": 0.15664789080619812, "learning_rate": 5.323508718435044e-06, "loss": 0.0093, "step": 85920 }, { "grad_norm": 0.23441636562347412, "learning_rate": 5.316087018429011e-06, "loss": 0.0114, "step": 85930 }, { "grad_norm": 0.14177775382995605, "learning_rate": 5.30867020498853e-06, "loss": 0.0083, "step": 85940 }, { "grad_norm": 0.14531990885734558, "learning_rate": 5.30125827892467e-06, "loss": 0.012, "step": 85950 }, { "grad_norm": 0.09157887101173401, "learning_rate": 5.293851241048015e-06, "loss": 0.009, "step": 85960 }, { "grad_norm": 0.20101605355739594, "learning_rate": 5.286449092168555e-06, "loss": 0.008, "step": 85970 }, { "grad_norm": 0.10531558096408844, "learning_rate": 5.2790518330958105e-06, "loss": 0.008, "step": 85980 }, { "grad_norm": 0.12404566258192062, "learning_rate": 5.271659464638717e-06, "loss": 0.0094, "step": 85990 }, { "grad_norm": 0.16516892611980438, "learning_rate": 5.264271987605701e-06, "loss": 0.0104, "step": 86000 }, { "grad_norm": 0.15080992877483368, "learning_rate": 5.256889402804632e-06, "loss": 0.0108, "step": 86010 }, { "grad_norm": 0.11329002678394318, "learning_rate": 5.249511711042881e-06, "loss": 0.0071, "step": 86020 }, { "grad_norm": 0.1825983226299286, "learning_rate": 5.242138913127237e-06, "loss": 0.0075, "step": 86030 }, { "grad_norm": 0.1463770866394043, "learning_rate": 5.234771009863987e-06, "loss": 0.0085, "step": 86040 }, { "grad_norm": 0.09847207367420197, "learning_rate": 5.227408002058881e-06, "loss": 0.0078, "step": 86050 }, { "grad_norm": 0.16004201769828796, "learning_rate": 5.220049890517126e-06, "loss": 0.0079, "step": 86060 }, { "grad_norm": 0.10535592585802078, "learning_rate": 5.212696676043377e-06, "loss": 0.0093, "step": 86070 }, { "grad_norm": 0.15786252915859222, "learning_rate": 5.205348359441798e-06, "loss": 0.009, "step": 86080 }, { "grad_norm": 0.0931905135512352, "learning_rate": 5.198004941515966e-06, "loss": 0.0086, "step": 86090 }, { "grad_norm": 0.13013066351413727, "learning_rate": 5.190666423068952e-06, "loss": 0.0094, "step": 86100 }, { "grad_norm": 0.08228980004787445, "learning_rate": 5.183332804903284e-06, "loss": 0.0087, "step": 86110 }, { "grad_norm": 0.2097640037536621, "learning_rate": 5.176004087820979e-06, "loss": 0.0098, "step": 86120 }, { "grad_norm": 0.17269723117351532, "learning_rate": 5.168680272623455e-06, "loss": 0.0115, "step": 86130 }, { "grad_norm": 0.1640682965517044, "learning_rate": 5.161361360111661e-06, "loss": 0.0099, "step": 86140 }, { "grad_norm": 0.08059081435203552, "learning_rate": 5.154047351085984e-06, "loss": 0.0086, "step": 86150 }, { "grad_norm": 0.15831045806407928, "learning_rate": 5.1467382463462385e-06, "loss": 0.0082, "step": 86160 }, { "grad_norm": 0.14885760843753815, "learning_rate": 5.139434046691788e-06, "loss": 0.0115, "step": 86170 }, { "grad_norm": 0.13654044270515442, "learning_rate": 5.13213475292137e-06, "loss": 0.0064, "step": 86180 }, { "grad_norm": 0.09400402754545212, "learning_rate": 5.12484036583325e-06, "loss": 0.0079, "step": 86190 }, { "grad_norm": 0.10823558270931244, "learning_rate": 5.117550886225109e-06, "loss": 0.0087, "step": 86200 }, { "grad_norm": 0.18695372343063354, "learning_rate": 5.11026631489413e-06, "loss": 0.0092, "step": 86210 }, { "grad_norm": 0.09102953225374222, "learning_rate": 5.102986652636926e-06, "loss": 0.0082, "step": 86220 }, { "grad_norm": 0.16798369586467743, "learning_rate": 5.095711900249606e-06, "loss": 0.0109, "step": 86230 }, { "grad_norm": 0.1460486203432083, "learning_rate": 5.088442058527715e-06, "loss": 0.0081, "step": 86240 }, { "grad_norm": 0.0930425375699997, "learning_rate": 5.08117712826629e-06, "loss": 0.0102, "step": 86250 }, { "grad_norm": 0.1101599782705307, "learning_rate": 5.073917110259785e-06, "loss": 0.0074, "step": 86260 }, { "grad_norm": 0.08936379104852676, "learning_rate": 5.066662005302175e-06, "loss": 0.0067, "step": 86270 }, { "grad_norm": 0.12695316970348358, "learning_rate": 5.059411814186838e-06, "loss": 0.0074, "step": 86280 }, { "grad_norm": 0.15600687265396118, "learning_rate": 5.05216653770666e-06, "loss": 0.009, "step": 86290 }, { "grad_norm": 0.0895421952009201, "learning_rate": 5.044926176653969e-06, "loss": 0.0073, "step": 86300 }, { "grad_norm": 0.1426076591014862, "learning_rate": 5.037690731820571e-06, "loss": 0.0072, "step": 86310 }, { "grad_norm": 0.17404988408088684, "learning_rate": 5.030460203997711e-06, "loss": 0.0081, "step": 86320 }, { "grad_norm": 0.10810743272304535, "learning_rate": 5.0232345939761174e-06, "loss": 0.0092, "step": 86330 }, { "grad_norm": 0.2647157907485962, "learning_rate": 5.016013902545957e-06, "loss": 0.0091, "step": 86340 }, { "grad_norm": 0.1359746903181076, "learning_rate": 5.008798130496889e-06, "loss": 0.0113, "step": 86350 }, { "grad_norm": 0.23695409297943115, "learning_rate": 5.001587278618008e-06, "loss": 0.0082, "step": 86360 }, { "grad_norm": 0.153273805975914, "learning_rate": 4.994381347697891e-06, "loss": 0.009, "step": 86370 }, { "grad_norm": 0.13754282891750336, "learning_rate": 4.987180338524572e-06, "loss": 0.0088, "step": 86380 }, { "grad_norm": 0.19011548161506653, "learning_rate": 4.9799842518855275e-06, "loss": 0.008, "step": 86390 }, { "grad_norm": 0.1477421522140503, "learning_rate": 4.9727930885677285e-06, "loss": 0.0097, "step": 86400 }, { "grad_norm": 0.1892872005701065, "learning_rate": 4.965606849357573e-06, "loss": 0.0091, "step": 86410 }, { "grad_norm": 0.10217159241437912, "learning_rate": 4.9584255350409445e-06, "loss": 0.0091, "step": 86420 }, { "grad_norm": 0.10008283704519272, "learning_rate": 4.951249146403175e-06, "loss": 0.0069, "step": 86430 }, { "grad_norm": 0.09676684439182281, "learning_rate": 4.9440776842290815e-06, "loss": 0.0132, "step": 86440 }, { "grad_norm": 0.24091103672981262, "learning_rate": 4.936911149302903e-06, "loss": 0.0088, "step": 86450 }, { "grad_norm": 0.12372682243585587, "learning_rate": 4.929749542408379e-06, "loss": 0.0135, "step": 86460 }, { "grad_norm": 0.1440398246049881, "learning_rate": 4.922592864328673e-06, "loss": 0.0084, "step": 86470 }, { "grad_norm": 0.12195856869220734, "learning_rate": 4.915441115846436e-06, "loss": 0.0089, "step": 86480 }, { "grad_norm": 0.2358681708574295, "learning_rate": 4.908294297743782e-06, "loss": 0.0088, "step": 86490 }, { "grad_norm": 0.10079219937324524, "learning_rate": 4.901152410802273e-06, "loss": 0.0107, "step": 86500 }, { "grad_norm": 0.1475319117307663, "learning_rate": 4.894015455802919e-06, "loss": 0.009, "step": 86510 }, { "grad_norm": 0.12485519796609879, "learning_rate": 4.8868834335262346e-06, "loss": 0.0082, "step": 86520 }, { "grad_norm": 0.08082010596990585, "learning_rate": 4.87975634475214e-06, "loss": 0.0085, "step": 86530 }, { "grad_norm": 0.10013750195503235, "learning_rate": 4.8726341902600614e-06, "loss": 0.0073, "step": 86540 }, { "grad_norm": 0.11493335664272308, "learning_rate": 4.865516970828843e-06, "loss": 0.0074, "step": 86550 }, { "grad_norm": 0.12604597210884094, "learning_rate": 4.85840468723685e-06, "loss": 0.0092, "step": 86560 }, { "grad_norm": 0.13277490437030792, "learning_rate": 4.85129734026184e-06, "loss": 0.0085, "step": 86570 }, { "grad_norm": 0.1681186705827713, "learning_rate": 4.844194930681073e-06, "loss": 0.0092, "step": 86580 }, { "grad_norm": 0.13200326263904572, "learning_rate": 4.837097459271267e-06, "loss": 0.0089, "step": 86590 }, { "grad_norm": 0.12761805951595306, "learning_rate": 4.830004926808573e-06, "loss": 0.0092, "step": 86600 }, { "grad_norm": 0.15728358924388885, "learning_rate": 4.822917334068627e-06, "loss": 0.0078, "step": 86610 }, { "grad_norm": 0.18449991941452026, "learning_rate": 4.81583468182652e-06, "loss": 0.0096, "step": 86620 }, { "grad_norm": 0.15347567200660706, "learning_rate": 4.808756970856809e-06, "loss": 0.0078, "step": 86630 }, { "grad_norm": 0.0935526043176651, "learning_rate": 4.801684201933476e-06, "loss": 0.0088, "step": 86640 }, { "grad_norm": 0.23572149872779846, "learning_rate": 4.794616375830019e-06, "loss": 0.0125, "step": 86650 }, { "grad_norm": 0.17923134565353394, "learning_rate": 4.787553493319341e-06, "loss": 0.0118, "step": 86660 }, { "grad_norm": 0.22127117216587067, "learning_rate": 4.780495555173836e-06, "loss": 0.0077, "step": 86670 }, { "grad_norm": 0.11189927905797958, "learning_rate": 4.7734425621653475e-06, "loss": 0.0098, "step": 86680 }, { "grad_norm": 0.13312378525733948, "learning_rate": 4.766394515065198e-06, "loss": 0.008, "step": 86690 }, { "grad_norm": 0.10600277781486511, "learning_rate": 4.75935141464412e-06, "loss": 0.0103, "step": 86700 }, { "grad_norm": 0.12718833982944489, "learning_rate": 4.75231326167237e-06, "loss": 0.0082, "step": 86710 }, { "grad_norm": 0.11912732571363449, "learning_rate": 4.745280056919599e-06, "loss": 0.0103, "step": 86720 }, { "grad_norm": 0.13846902549266815, "learning_rate": 4.738251801154975e-06, "loss": 0.0098, "step": 86730 }, { "grad_norm": 0.13694511353969574, "learning_rate": 4.731228495147061e-06, "loss": 0.0075, "step": 86740 }, { "grad_norm": 0.08977523446083069, "learning_rate": 4.7242101396639595e-06, "loss": 0.0093, "step": 86750 }, { "grad_norm": 0.12585996091365814, "learning_rate": 4.717196735473151e-06, "loss": 0.0075, "step": 86760 }, { "grad_norm": 0.16350767016410828, "learning_rate": 4.710188283341643e-06, "loss": 0.0088, "step": 86770 }, { "grad_norm": 0.13632814586162567, "learning_rate": 4.703184784035835e-06, "loss": 0.0071, "step": 86780 }, { "grad_norm": 0.0879330262541771, "learning_rate": 4.696186238321642e-06, "loss": 0.0104, "step": 86790 }, { "grad_norm": 0.10984384268522263, "learning_rate": 4.689192646964408e-06, "loss": 0.0082, "step": 86800 }, { "grad_norm": 0.13096627593040466, "learning_rate": 4.682204010728952e-06, "loss": 0.0108, "step": 86810 }, { "grad_norm": 0.08743739128112793, "learning_rate": 4.67522033037952e-06, "loss": 0.0074, "step": 86820 }, { "grad_norm": 0.07161251455545425, "learning_rate": 4.668241606679846e-06, "loss": 0.011, "step": 86830 }, { "grad_norm": 0.089427150785923, "learning_rate": 4.66126784039313e-06, "loss": 0.0111, "step": 86840 }, { "grad_norm": 0.08510520309209824, "learning_rate": 4.6542990322819836e-06, "loss": 0.0108, "step": 86850 }, { "grad_norm": 0.09633902460336685, "learning_rate": 4.64733518310852e-06, "loss": 0.0083, "step": 86860 }, { "grad_norm": 0.13463576138019562, "learning_rate": 4.640376293634291e-06, "loss": 0.0107, "step": 86870 }, { "grad_norm": 0.18240852653980255, "learning_rate": 4.633422364620327e-06, "loss": 0.0103, "step": 86880 }, { "grad_norm": 0.09240223467350006, "learning_rate": 4.6264733968270745e-06, "loss": 0.0081, "step": 86890 }, { "grad_norm": 0.1783149391412735, "learning_rate": 4.619529391014482e-06, "loss": 0.0092, "step": 86900 }, { "grad_norm": 0.11712820827960968, "learning_rate": 4.61259034794192e-06, "loss": 0.006, "step": 86910 }, { "grad_norm": 0.2526995539665222, "learning_rate": 4.605656268368247e-06, "loss": 0.01, "step": 86920 }, { "grad_norm": 0.08041497319936752, "learning_rate": 4.598727153051735e-06, "loss": 0.0067, "step": 86930 }, { "grad_norm": 0.15431317687034607, "learning_rate": 4.591803002750178e-06, "loss": 0.0094, "step": 86940 }, { "grad_norm": 0.10530314594507217, "learning_rate": 4.58488381822077e-06, "loss": 0.0095, "step": 86950 }, { "grad_norm": 0.1111752837896347, "learning_rate": 4.5779696002201946e-06, "loss": 0.0089, "step": 86960 }, { "grad_norm": 0.06791073828935623, "learning_rate": 4.571060349504563e-06, "loss": 0.0076, "step": 86970 }, { "grad_norm": 0.1323917955160141, "learning_rate": 4.564156066829478e-06, "loss": 0.0086, "step": 86980 }, { "grad_norm": 0.14704005420207977, "learning_rate": 4.557256752949957e-06, "loss": 0.0089, "step": 86990 }, { "grad_norm": 0.18900473415851593, "learning_rate": 4.550362408620534e-06, "loss": 0.0082, "step": 87000 }, { "grad_norm": 0.09257563203573227, "learning_rate": 4.543473034595136e-06, "loss": 0.0084, "step": 87010 }, { "grad_norm": 0.06589686125516891, "learning_rate": 4.536588631627192e-06, "loss": 0.0063, "step": 87020 }, { "grad_norm": 0.09228958934545517, "learning_rate": 4.5297092004695565e-06, "loss": 0.007, "step": 87030 }, { "grad_norm": 0.16007712483406067, "learning_rate": 4.522834741874555e-06, "loss": 0.0094, "step": 87040 }, { "grad_norm": 0.10173039883375168, "learning_rate": 4.515965256593974e-06, "loss": 0.006, "step": 87050 }, { "grad_norm": 0.19446943700313568, "learning_rate": 4.509100745379058e-06, "loss": 0.0103, "step": 87060 }, { "grad_norm": 0.11451677232980728, "learning_rate": 4.502241208980479e-06, "loss": 0.0093, "step": 87070 }, { "grad_norm": 0.1813661754131317, "learning_rate": 4.495386648148398e-06, "loss": 0.0062, "step": 87080 }, { "grad_norm": 0.09853839874267578, "learning_rate": 4.488537063632431e-06, "loss": 0.0066, "step": 87090 }, { "grad_norm": 0.13185851275920868, "learning_rate": 4.481692456181608e-06, "loss": 0.0094, "step": 87100 }, { "grad_norm": 0.16825881600379944, "learning_rate": 4.474852826544479e-06, "loss": 0.0092, "step": 87110 }, { "grad_norm": 0.14318102598190308, "learning_rate": 4.468018175468974e-06, "loss": 0.008, "step": 87120 }, { "grad_norm": 0.11326303333044052, "learning_rate": 4.461188503702568e-06, "loss": 0.0083, "step": 87130 }, { "grad_norm": 0.1751669943332672, "learning_rate": 4.454363811992107e-06, "loss": 0.0092, "step": 87140 }, { "grad_norm": 0.15291276574134827, "learning_rate": 4.447544101083956e-06, "loss": 0.0093, "step": 87150 }, { "grad_norm": 0.16780881583690643, "learning_rate": 4.440729371723884e-06, "loss": 0.009, "step": 87160 }, { "grad_norm": 0.14067280292510986, "learning_rate": 4.433919624657162e-06, "loss": 0.0083, "step": 87170 }, { "grad_norm": 0.18162867426872253, "learning_rate": 4.427114860628462e-06, "loss": 0.0079, "step": 87180 }, { "grad_norm": 0.14133408665657043, "learning_rate": 4.420315080381982e-06, "loss": 0.0094, "step": 87190 }, { "grad_norm": 0.13638348877429962, "learning_rate": 4.4135202846613045e-06, "loss": 0.0095, "step": 87200 }, { "grad_norm": 0.10876817256212234, "learning_rate": 4.406730474209525e-06, "loss": 0.0088, "step": 87210 }, { "grad_norm": 0.2174426019191742, "learning_rate": 4.399945649769138e-06, "loss": 0.0075, "step": 87220 }, { "grad_norm": 0.1070890873670578, "learning_rate": 4.393165812082151e-06, "loss": 0.0085, "step": 87230 }, { "grad_norm": 0.1487174779176712, "learning_rate": 4.386390961889969e-06, "loss": 0.0076, "step": 87240 }, { "grad_norm": 0.07134673744440079, "learning_rate": 4.3796210999334895e-06, "loss": 0.0066, "step": 87250 }, { "grad_norm": 0.09006457775831223, "learning_rate": 4.372856226953065e-06, "loss": 0.0068, "step": 87260 }, { "grad_norm": 0.09027885645627975, "learning_rate": 4.366096343688486e-06, "loss": 0.0102, "step": 87270 }, { "grad_norm": 0.09508125483989716, "learning_rate": 4.359341450878995e-06, "loss": 0.01, "step": 87280 }, { "grad_norm": 0.14771373569965363, "learning_rate": 4.352591549263302e-06, "loss": 0.0092, "step": 87290 }, { "grad_norm": 0.09459380060434341, "learning_rate": 4.345846639579582e-06, "loss": 0.0078, "step": 87300 }, { "grad_norm": 0.11411340534687042, "learning_rate": 4.339106722565417e-06, "loss": 0.0067, "step": 87310 }, { "grad_norm": 0.1284664124250412, "learning_rate": 4.332371798957896e-06, "loss": 0.0108, "step": 87320 }, { "grad_norm": 0.16071870923042297, "learning_rate": 4.3256418694935295e-06, "loss": 0.0077, "step": 87330 }, { "grad_norm": 0.1421840786933899, "learning_rate": 4.3189169349083125e-06, "loss": 0.01, "step": 87340 }, { "grad_norm": 0.13126371800899506, "learning_rate": 4.312196995937645e-06, "loss": 0.0094, "step": 87350 }, { "grad_norm": 0.0953783467411995, "learning_rate": 4.305482053316434e-06, "loss": 0.0095, "step": 87360 }, { "grad_norm": 0.14419177174568176, "learning_rate": 4.2987721077789925e-06, "loss": 0.011, "step": 87370 }, { "grad_norm": 0.1516299843788147, "learning_rate": 4.2920671600591325e-06, "loss": 0.0084, "step": 87380 }, { "grad_norm": 0.16089026629924774, "learning_rate": 4.285367210890084e-06, "loss": 0.0094, "step": 87390 }, { "grad_norm": 0.11901256442070007, "learning_rate": 4.278672261004551e-06, "loss": 0.0111, "step": 87400 }, { "grad_norm": 0.11979120224714279, "learning_rate": 4.2719823111346704e-06, "loss": 0.0108, "step": 87410 }, { "grad_norm": 0.0874219536781311, "learning_rate": 4.265297362012061e-06, "loss": 0.009, "step": 87420 }, { "grad_norm": 0.1083802580833435, "learning_rate": 4.258617414367766e-06, "loss": 0.0108, "step": 87430 }, { "grad_norm": 0.1308739185333252, "learning_rate": 4.251942468932296e-06, "loss": 0.0089, "step": 87440 }, { "grad_norm": 0.12203550338745117, "learning_rate": 4.2452725264356175e-06, "loss": 0.0081, "step": 87450 }, { "grad_norm": 0.12617121636867523, "learning_rate": 4.2386075876071575e-06, "loss": 0.0072, "step": 87460 }, { "grad_norm": 0.13131655752658844, "learning_rate": 4.231947653175755e-06, "loss": 0.0077, "step": 87470 }, { "grad_norm": 0.14015385508537292, "learning_rate": 4.225292723869761e-06, "loss": 0.0094, "step": 87480 }, { "grad_norm": 0.10669653862714767, "learning_rate": 4.218642800416916e-06, "loss": 0.0086, "step": 87490 }, { "grad_norm": 0.2294549196958542, "learning_rate": 4.211997883544472e-06, "loss": 0.0088, "step": 87500 }, { "grad_norm": 0.12398552894592285, "learning_rate": 4.205357973979096e-06, "loss": 0.0087, "step": 87510 }, { "grad_norm": 0.12548847496509552, "learning_rate": 4.198723072446925e-06, "loss": 0.0079, "step": 87520 }, { "grad_norm": 0.1563924252986908, "learning_rate": 4.192093179673534e-06, "loss": 0.0078, "step": 87530 }, { "grad_norm": 0.09643077850341797, "learning_rate": 4.185468296383959e-06, "loss": 0.007, "step": 87540 }, { "grad_norm": 0.20335030555725098, "learning_rate": 4.1788484233027005e-06, "loss": 0.0087, "step": 87550 }, { "grad_norm": 0.12744081020355225, "learning_rate": 4.172233561153677e-06, "loss": 0.0097, "step": 87560 }, { "grad_norm": 0.17049379646778107, "learning_rate": 4.165623710660288e-06, "loss": 0.0098, "step": 87570 }, { "grad_norm": 0.08746754378080368, "learning_rate": 4.1590188725453835e-06, "loss": 0.0084, "step": 87580 }, { "grad_norm": 0.1084228903055191, "learning_rate": 4.152419047531259e-06, "loss": 0.0096, "step": 87590 }, { "grad_norm": 0.8842363357543945, "learning_rate": 4.145824236339651e-06, "loss": 0.0099, "step": 87600 }, { "grad_norm": 0.11855155974626541, "learning_rate": 4.1392344396917695e-06, "loss": 0.0081, "step": 87610 }, { "grad_norm": 0.2035861313343048, "learning_rate": 4.132649658308252e-06, "loss": 0.0088, "step": 87620 }, { "grad_norm": 0.11475899815559387, "learning_rate": 4.126069892909213e-06, "loss": 0.0098, "step": 87630 }, { "grad_norm": 0.135282963514328, "learning_rate": 4.119495144214191e-06, "loss": 0.0079, "step": 87640 }, { "grad_norm": 0.1391192376613617, "learning_rate": 4.112925412942214e-06, "loss": 0.0086, "step": 87650 }, { "grad_norm": 0.09104583412408829, "learning_rate": 4.106360699811712e-06, "loss": 0.0093, "step": 87660 }, { "grad_norm": 0.09457340091466904, "learning_rate": 4.099801005540616e-06, "loss": 0.0084, "step": 87670 }, { "grad_norm": 0.1946609616279602, "learning_rate": 4.093246330846256e-06, "loss": 0.0095, "step": 87680 }, { "grad_norm": 0.18125106394290924, "learning_rate": 4.086696676445467e-06, "loss": 0.0102, "step": 87690 }, { "grad_norm": 0.1641596555709839, "learning_rate": 4.0801520430544914e-06, "loss": 0.0089, "step": 87700 }, { "grad_norm": 0.12137256562709808, "learning_rate": 4.073612431389062e-06, "loss": 0.0114, "step": 87710 }, { "grad_norm": 0.1048203557729721, "learning_rate": 4.067077842164319e-06, "loss": 0.0078, "step": 87720 }, { "grad_norm": 0.15453225374221802, "learning_rate": 4.0605482760948945e-06, "loss": 0.0103, "step": 87730 }, { "grad_norm": 0.09013006836175919, "learning_rate": 4.054023733894829e-06, "loss": 0.0076, "step": 87740 }, { "grad_norm": 0.11335807293653488, "learning_rate": 4.047504216277653e-06, "loss": 0.0088, "step": 87750 }, { "grad_norm": 0.10644929856061935, "learning_rate": 4.040989723956323e-06, "loss": 0.0077, "step": 87760 }, { "grad_norm": 0.14219297468662262, "learning_rate": 4.03448025764327e-06, "loss": 0.0083, "step": 87770 }, { "grad_norm": 0.11777865141630173, "learning_rate": 4.027975818050339e-06, "loss": 0.0066, "step": 87780 }, { "grad_norm": 0.22303804755210876, "learning_rate": 4.021476405888858e-06, "loss": 0.0096, "step": 87790 }, { "grad_norm": 0.14229880273342133, "learning_rate": 4.01498202186959e-06, "loss": 0.0077, "step": 87800 }, { "grad_norm": 0.12706083059310913, "learning_rate": 4.008492666702751e-06, "loss": 0.0117, "step": 87810 }, { "grad_norm": 0.08983733505010605, "learning_rate": 4.002008341098001e-06, "loss": 0.006, "step": 87820 }, { "grad_norm": 0.13753920793533325, "learning_rate": 3.995529045764462e-06, "loss": 0.0078, "step": 87830 }, { "grad_norm": 0.13955806195735931, "learning_rate": 3.989054781410712e-06, "loss": 0.0079, "step": 87840 }, { "grad_norm": 0.1483040750026703, "learning_rate": 3.982585548744744e-06, "loss": 0.0082, "step": 87850 }, { "grad_norm": 0.1277296245098114, "learning_rate": 3.9761213484740435e-06, "loss": 0.008, "step": 87860 }, { "grad_norm": 0.10295803099870682, "learning_rate": 3.9696621813055055e-06, "loss": 0.0111, "step": 87870 }, { "grad_norm": 0.13625074923038483, "learning_rate": 3.9632080479455044e-06, "loss": 0.0065, "step": 87880 }, { "grad_norm": 0.16071626543998718, "learning_rate": 3.956758949099859e-06, "loss": 0.0117, "step": 87890 }, { "grad_norm": 0.14152321219444275, "learning_rate": 3.9503148854738385e-06, "loss": 0.0117, "step": 87900 }, { "grad_norm": 0.1064383015036583, "learning_rate": 3.943875857772139e-06, "loss": 0.0098, "step": 87910 }, { "grad_norm": 0.13330470025539398, "learning_rate": 3.9374418666989375e-06, "loss": 0.0068, "step": 87920 }, { "grad_norm": 0.09966053813695908, "learning_rate": 3.931012912957838e-06, "loss": 0.0074, "step": 87930 }, { "grad_norm": 0.18418408930301666, "learning_rate": 3.9245889972519045e-06, "loss": 0.0075, "step": 87940 }, { "grad_norm": 0.09812821447849274, "learning_rate": 3.9181701202836265e-06, "loss": 0.0103, "step": 87950 }, { "grad_norm": 0.1245913878083229, "learning_rate": 3.911756282755003e-06, "loss": 0.01, "step": 87960 }, { "grad_norm": 0.11028246581554413, "learning_rate": 3.905347485367411e-06, "loss": 0.011, "step": 87970 }, { "grad_norm": 0.09749440848827362, "learning_rate": 3.898943728821725e-06, "loss": 0.0096, "step": 87980 }, { "grad_norm": 0.10757897049188614, "learning_rate": 3.892545013818227e-06, "loss": 0.0075, "step": 87990 }, { "grad_norm": 0.15489301085472107, "learning_rate": 3.886151341056687e-06, "loss": 0.0099, "step": 88000 }, { "grad_norm": 0.1099318265914917, "learning_rate": 3.8797627112363066e-06, "loss": 0.007, "step": 88010 }, { "grad_norm": 0.10643802583217621, "learning_rate": 3.873379125055743e-06, "loss": 0.009, "step": 88020 }, { "grad_norm": 0.10822310298681259, "learning_rate": 3.867000583213082e-06, "loss": 0.0093, "step": 88030 }, { "grad_norm": 0.13677743077278137, "learning_rate": 3.860627086405871e-06, "loss": 0.0088, "step": 88040 }, { "grad_norm": 0.08847339451313019, "learning_rate": 3.854258635331126e-06, "loss": 0.007, "step": 88050 }, { "grad_norm": 0.17036360502243042, "learning_rate": 3.8478952306852665e-06, "loss": 0.01, "step": 88060 }, { "grad_norm": 0.1723642796278, "learning_rate": 3.841536873164198e-06, "loss": 0.0091, "step": 88070 }, { "grad_norm": 0.10229583829641342, "learning_rate": 3.8351835634632626e-06, "loss": 0.0077, "step": 88080 }, { "grad_norm": 0.16361723840236664, "learning_rate": 3.828835302277245e-06, "loss": 0.0077, "step": 88090 }, { "grad_norm": 0.14534787833690643, "learning_rate": 3.822492090300378e-06, "loss": 0.0081, "step": 88100 }, { "grad_norm": 0.10468509793281555, "learning_rate": 3.8161539282263566e-06, "loss": 0.0083, "step": 88110 }, { "grad_norm": 0.19597555696964264, "learning_rate": 3.8098208167482995e-06, "loss": 0.0074, "step": 88120 }, { "grad_norm": 0.13885381817817688, "learning_rate": 3.803492756558796e-06, "loss": 0.0082, "step": 88130 }, { "grad_norm": 0.14410144090652466, "learning_rate": 3.797169748349855e-06, "loss": 0.0092, "step": 88140 }, { "grad_norm": 0.11468508839607239, "learning_rate": 3.790851792812977e-06, "loss": 0.0086, "step": 88150 }, { "grad_norm": 0.13319098949432373, "learning_rate": 3.784538890639072e-06, "loss": 0.0077, "step": 88160 }, { "grad_norm": 0.11496206372976303, "learning_rate": 3.7782310425185153e-06, "loss": 0.0083, "step": 88170 }, { "grad_norm": 0.08092658221721649, "learning_rate": 3.77192824914111e-06, "loss": 0.0072, "step": 88180 }, { "grad_norm": 0.17227782309055328, "learning_rate": 3.7656305111961333e-06, "loss": 0.0097, "step": 88190 }, { "grad_norm": 0.15243111550807953, "learning_rate": 3.7593378293722782e-06, "loss": 0.0089, "step": 88200 }, { "grad_norm": 0.1459481567144394, "learning_rate": 3.7530502043577333e-06, "loss": 0.0095, "step": 88210 }, { "grad_norm": 0.19585929811000824, "learning_rate": 3.746767636840076e-06, "loss": 0.0086, "step": 88220 }, { "grad_norm": 0.11514916270971298, "learning_rate": 3.74049012750638e-06, "loss": 0.0096, "step": 88230 }, { "grad_norm": 0.21316446363925934, "learning_rate": 3.734217677043128e-06, "loss": 0.0096, "step": 88240 }, { "grad_norm": 0.1303325593471527, "learning_rate": 3.727950286136267e-06, "loss": 0.0094, "step": 88250 }, { "grad_norm": 0.2007220834493637, "learning_rate": 3.7216879554711925e-06, "loss": 0.0073, "step": 88260 }, { "grad_norm": 0.11975402384996414, "learning_rate": 3.715430685732757e-06, "loss": 0.0081, "step": 88270 }, { "grad_norm": 0.11242306232452393, "learning_rate": 3.7091784776052243e-06, "loss": 0.0085, "step": 88280 }, { "grad_norm": 0.12489300221204758, "learning_rate": 3.7029313317723356e-06, "loss": 0.0092, "step": 88290 }, { "grad_norm": 0.11088739335536957, "learning_rate": 3.6966892489172787e-06, "loss": 0.0061, "step": 88300 }, { "grad_norm": 0.09686927497386932, "learning_rate": 3.6904522297226575e-06, "loss": 0.0082, "step": 88310 }, { "grad_norm": 0.1532883495092392, "learning_rate": 3.6842202748705647e-06, "loss": 0.0092, "step": 88320 }, { "grad_norm": 0.17222259938716888, "learning_rate": 3.677993385042494e-06, "loss": 0.008, "step": 88330 }, { "grad_norm": 0.10991792380809784, "learning_rate": 3.67177156091944e-06, "loss": 0.0103, "step": 88340 }, { "grad_norm": 0.1332341581583023, "learning_rate": 3.6655548031817866e-06, "loss": 0.0075, "step": 88350 }, { "grad_norm": 0.0795707106590271, "learning_rate": 3.6593431125094e-06, "loss": 0.0078, "step": 88360 }, { "grad_norm": 0.14321544766426086, "learning_rate": 3.6531364895815766e-06, "loss": 0.0085, "step": 88370 }, { "grad_norm": 0.12073154002428055, "learning_rate": 3.6469349350770677e-06, "loss": 0.0075, "step": 88380 }, { "grad_norm": 0.18281246721744537, "learning_rate": 3.6407384496740525e-06, "loss": 0.0106, "step": 88390 }, { "grad_norm": 0.09570904821157455, "learning_rate": 3.6345470340501954e-06, "loss": 0.0084, "step": 88400 }, { "grad_norm": 0.126506969332695, "learning_rate": 3.6283606888825537e-06, "loss": 0.0072, "step": 88410 }, { "grad_norm": 0.14829102158546448, "learning_rate": 3.622179414847676e-06, "loss": 0.0069, "step": 88420 }, { "grad_norm": 0.11731606721878052, "learning_rate": 3.616003212621527e-06, "loss": 0.0084, "step": 88430 }, { "grad_norm": 0.14777013659477234, "learning_rate": 3.609832082879533e-06, "loss": 0.0076, "step": 88440 }, { "grad_norm": 0.10985169559717178, "learning_rate": 3.6036660262965485e-06, "loss": 0.0103, "step": 88450 }, { "grad_norm": 0.0892244428396225, "learning_rate": 3.597505043546895e-06, "loss": 0.0107, "step": 88460 }, { "grad_norm": 0.1389096975326538, "learning_rate": 3.5913491353043227e-06, "loss": 0.0085, "step": 88470 }, { "grad_norm": 0.12863942980766296, "learning_rate": 3.5851983022420375e-06, "loss": 0.0078, "step": 88480 }, { "grad_norm": 0.13347972929477692, "learning_rate": 3.5790525450326797e-06, "loss": 0.0096, "step": 88490 }, { "grad_norm": 0.136402428150177, "learning_rate": 3.572911864348344e-06, "loss": 0.0077, "step": 88500 }, { "grad_norm": 0.1494472771883011, "learning_rate": 3.5667762608605717e-06, "loss": 0.0072, "step": 88510 }, { "grad_norm": 0.12316851317882538, "learning_rate": 3.5606457352403255e-06, "loss": 0.01, "step": 88520 }, { "grad_norm": 0.17615407705307007, "learning_rate": 3.554520288158042e-06, "loss": 0.0085, "step": 88530 }, { "grad_norm": 0.13626804947853088, "learning_rate": 3.5483999202835903e-06, "loss": 0.0101, "step": 88540 }, { "grad_norm": 0.13361574709415436, "learning_rate": 3.5422846322862913e-06, "loss": 0.0108, "step": 88550 }, { "grad_norm": 0.17633679509162903, "learning_rate": 3.5361744248348925e-06, "loss": 0.0069, "step": 88560 }, { "grad_norm": 0.1410345733165741, "learning_rate": 3.5300692985976047e-06, "loss": 0.0096, "step": 88570 }, { "grad_norm": 0.15518483519554138, "learning_rate": 3.52396925424206e-06, "loss": 0.007, "step": 88580 }, { "grad_norm": 0.20251160860061646, "learning_rate": 3.5178742924353747e-06, "loss": 0.0077, "step": 88590 }, { "grad_norm": 0.09732384979724884, "learning_rate": 3.511784413844066e-06, "loss": 0.0096, "step": 88600 }, { "grad_norm": 0.21006318926811218, "learning_rate": 3.5056996191341287e-06, "loss": 0.0084, "step": 88610 }, { "grad_norm": 0.15595167875289917, "learning_rate": 3.4996199089709692e-06, "loss": 0.0077, "step": 88620 }, { "grad_norm": 0.1378406137228012, "learning_rate": 3.4935452840194725e-06, "loss": 0.0093, "step": 88630 }, { "grad_norm": 0.14905551075935364, "learning_rate": 3.48747574494393e-06, "loss": 0.0103, "step": 88640 }, { "grad_norm": 0.08828609436750412, "learning_rate": 3.4814112924081154e-06, "loss": 0.0087, "step": 88650 }, { "grad_norm": 0.14500519633293152, "learning_rate": 3.475351927075221e-06, "loss": 0.0093, "step": 88660 }, { "grad_norm": 0.23745520412921906, "learning_rate": 3.469297649607894e-06, "loss": 0.0111, "step": 88670 }, { "grad_norm": 0.14946788549423218, "learning_rate": 3.463248460668217e-06, "loss": 0.0077, "step": 88680 }, { "grad_norm": 0.15092478692531586, "learning_rate": 3.457204360917726e-06, "loss": 0.0086, "step": 88690 }, { "grad_norm": 0.3770895302295685, "learning_rate": 3.4511653510173827e-06, "loss": 0.0095, "step": 88700 }, { "grad_norm": 0.1355462372303009, "learning_rate": 3.4451314316276074e-06, "loss": 0.0084, "step": 88710 }, { "grad_norm": 0.11015810072422028, "learning_rate": 3.4391026034082674e-06, "loss": 0.0086, "step": 88720 }, { "grad_norm": 0.09774089604616165, "learning_rate": 3.433078867018674e-06, "loss": 0.0053, "step": 88730 }, { "grad_norm": 0.11195914447307587, "learning_rate": 3.427060223117551e-06, "loss": 0.007, "step": 88740 }, { "grad_norm": 0.11079714447259903, "learning_rate": 3.4210466723631042e-06, "loss": 0.0082, "step": 88750 }, { "grad_norm": 0.10207468271255493, "learning_rate": 3.415038215412969e-06, "loss": 0.0079, "step": 88760 }, { "grad_norm": 0.07604917138814926, "learning_rate": 3.4090348529242088e-06, "loss": 0.0092, "step": 88770 }, { "grad_norm": 0.12364093214273453, "learning_rate": 3.4030365855533487e-06, "loss": 0.0099, "step": 88780 }, { "grad_norm": 0.08609844744205475, "learning_rate": 3.397043413956347e-06, "loss": 0.007, "step": 88790 }, { "grad_norm": 0.14243508875370026, "learning_rate": 3.3910553387886237e-06, "loss": 0.0082, "step": 88800 }, { "grad_norm": 0.21447841823101044, "learning_rate": 3.3850723607049996e-06, "loss": 0.0103, "step": 88810 }, { "grad_norm": 0.08645287901163101, "learning_rate": 3.3790944803597902e-06, "loss": 0.0076, "step": 88820 }, { "grad_norm": 0.16083268821239471, "learning_rate": 3.373121698406706e-06, "loss": 0.0088, "step": 88830 }, { "grad_norm": 0.11722338199615479, "learning_rate": 3.3671540154989302e-06, "loss": 0.0096, "step": 88840 }, { "grad_norm": 0.12787245213985443, "learning_rate": 3.361191432289079e-06, "loss": 0.0093, "step": 88850 }, { "grad_norm": 0.10905550420284271, "learning_rate": 3.3552339494292207e-06, "loss": 0.0074, "step": 88860 }, { "grad_norm": 0.09039399772882462, "learning_rate": 3.349281567570839e-06, "loss": 0.0086, "step": 88870 }, { "grad_norm": 0.10982968658208847, "learning_rate": 3.3433342873648965e-06, "loss": 0.0078, "step": 88880 }, { "grad_norm": 0.09587335586547852, "learning_rate": 3.337392109461762e-06, "loss": 0.01, "step": 88890 }, { "grad_norm": 0.16163435578346252, "learning_rate": 3.331455034511266e-06, "loss": 0.0081, "step": 88900 }, { "grad_norm": 0.12425598502159119, "learning_rate": 3.325523063162689e-06, "loss": 0.0082, "step": 88910 }, { "grad_norm": 0.13017544150352478, "learning_rate": 3.3195961960647393e-06, "loss": 0.0104, "step": 88920 }, { "grad_norm": 0.0826335996389389, "learning_rate": 3.31367443386556e-06, "loss": 0.0099, "step": 88930 }, { "grad_norm": 0.105558380484581, "learning_rate": 3.3077577772127556e-06, "loss": 0.0083, "step": 88940 }, { "grad_norm": 0.21492432057857513, "learning_rate": 3.301846226753358e-06, "loss": 0.0095, "step": 88950 }, { "grad_norm": 0.10894905030727386, "learning_rate": 3.2959397831338444e-06, "loss": 0.0088, "step": 88960 }, { "grad_norm": 0.1278691440820694, "learning_rate": 3.2900384470001432e-06, "loss": 0.0072, "step": 88970 }, { "grad_norm": 0.11690934747457504, "learning_rate": 3.28414221899761e-06, "loss": 0.0073, "step": 88980 }, { "grad_norm": 0.19339235126972198, "learning_rate": 3.278251099771046e-06, "loss": 0.0099, "step": 88990 }, { "grad_norm": 0.11634469032287598, "learning_rate": 3.2723650899646906e-06, "loss": 0.0083, "step": 89000 }, { "grad_norm": 0.10600434988737106, "learning_rate": 3.2664841902222466e-06, "loss": 0.008, "step": 89010 }, { "grad_norm": 0.0822845846414566, "learning_rate": 3.2606084011868208e-06, "loss": 0.0077, "step": 89020 }, { "grad_norm": 0.14912039041519165, "learning_rate": 3.254737723500989e-06, "loss": 0.0081, "step": 89030 }, { "grad_norm": 0.11638233065605164, "learning_rate": 3.2488721578067595e-06, "loss": 0.0077, "step": 89040 }, { "grad_norm": 0.10220037400722504, "learning_rate": 3.243011704745591e-06, "loss": 0.0081, "step": 89050 }, { "grad_norm": 0.14622285962104797, "learning_rate": 3.237156364958355e-06, "loss": 0.0063, "step": 89060 }, { "grad_norm": 0.11408331245183945, "learning_rate": 3.231306139085394e-06, "loss": 0.0076, "step": 89070 }, { "grad_norm": 0.16337013244628906, "learning_rate": 3.2254610277664797e-06, "loss": 0.0108, "step": 89080 }, { "grad_norm": 0.1239423081278801, "learning_rate": 3.2196210316408183e-06, "loss": 0.0082, "step": 89090 }, { "grad_norm": 0.11872896552085876, "learning_rate": 3.2137861513470644e-06, "loss": 0.0071, "step": 89100 }, { "grad_norm": 0.17685717344284058, "learning_rate": 3.2079563875233252e-06, "loss": 0.0083, "step": 89110 }, { "grad_norm": 0.08635272085666656, "learning_rate": 3.2021317408071184e-06, "loss": 0.0095, "step": 89120 }, { "grad_norm": 0.18133071064949036, "learning_rate": 3.196312211835434e-06, "loss": 0.0086, "step": 89130 }, { "grad_norm": 0.10055724531412125, "learning_rate": 3.1904978012446686e-06, "loss": 0.009, "step": 89140 }, { "grad_norm": 0.19451771676540375, "learning_rate": 3.184688509670691e-06, "loss": 0.0101, "step": 89150 }, { "grad_norm": 0.1269349902868271, "learning_rate": 3.178884337748783e-06, "loss": 0.0097, "step": 89160 }, { "grad_norm": 0.14527711272239685, "learning_rate": 3.173085286113697e-06, "loss": 0.0087, "step": 89170 }, { "grad_norm": 0.13707126677036285, "learning_rate": 3.1672913553995986e-06, "loss": 0.0091, "step": 89180 }, { "grad_norm": 0.16001808643341064, "learning_rate": 3.1615025462401138e-06, "loss": 0.0081, "step": 89190 }, { "grad_norm": 0.09785759449005127, "learning_rate": 3.1557188592682808e-06, "loss": 0.0086, "step": 89200 }, { "grad_norm": 0.11953248083591461, "learning_rate": 3.1499402951166045e-06, "loss": 0.0091, "step": 89210 }, { "grad_norm": 0.09415721148252487, "learning_rate": 3.144166854417019e-06, "loss": 0.008, "step": 89220 }, { "grad_norm": 0.22519421577453613, "learning_rate": 3.1383985378009073e-06, "loss": 0.0099, "step": 89230 }, { "grad_norm": 0.10805361717939377, "learning_rate": 3.1326353458990654e-06, "loss": 0.0092, "step": 89240 }, { "grad_norm": 0.16230975091457367, "learning_rate": 3.126877279341761e-06, "loss": 0.0072, "step": 89250 }, { "grad_norm": 0.12958423793315887, "learning_rate": 3.1211243387586964e-06, "loss": 0.0097, "step": 89260 }, { "grad_norm": 0.2572723627090454, "learning_rate": 3.115376524778979e-06, "loss": 0.0108, "step": 89270 }, { "grad_norm": 0.11093845963478088, "learning_rate": 3.109633838031195e-06, "loss": 0.0092, "step": 89280 }, { "grad_norm": 0.1679050326347351, "learning_rate": 3.103896279143359e-06, "loss": 0.008, "step": 89290 }, { "grad_norm": 0.1127469465136528, "learning_rate": 3.098163848742924e-06, "loss": 0.008, "step": 89300 }, { "grad_norm": 0.12308559566736221, "learning_rate": 3.0924365474567675e-06, "loss": 0.0099, "step": 89310 }, { "grad_norm": 0.12136673182249069, "learning_rate": 3.086714375911237e-06, "loss": 0.0109, "step": 89320 }, { "grad_norm": 0.11489737033843994, "learning_rate": 3.0809973347320775e-06, "loss": 0.0096, "step": 89330 }, { "grad_norm": 0.10866977274417877, "learning_rate": 3.075285424544516e-06, "loss": 0.01, "step": 89340 }, { "grad_norm": 0.08672704547643661, "learning_rate": 3.069578645973176e-06, "loss": 0.0071, "step": 89350 }, { "grad_norm": 0.16308142244815826, "learning_rate": 3.0638769996421735e-06, "loss": 0.0098, "step": 89360 }, { "grad_norm": 0.13241782784461975, "learning_rate": 3.058180486175005e-06, "loss": 0.0076, "step": 89370 }, { "grad_norm": 0.15694265067577362, "learning_rate": 3.052489106194645e-06, "loss": 0.0092, "step": 89380 }, { "grad_norm": 0.11418693512678146, "learning_rate": 3.0468028603234946e-06, "loss": 0.0085, "step": 89390 }, { "grad_norm": 0.11118272691965103, "learning_rate": 3.04112174918339e-06, "loss": 0.0088, "step": 89400 }, { "grad_norm": 0.17374630272388458, "learning_rate": 3.0354457733955954e-06, "loss": 0.01, "step": 89410 }, { "grad_norm": 0.10487662255764008, "learning_rate": 3.029774933580859e-06, "loss": 0.0088, "step": 89420 }, { "grad_norm": 0.1405649334192276, "learning_rate": 3.0241092303593065e-06, "loss": 0.0083, "step": 89430 }, { "grad_norm": 0.07396029680967331, "learning_rate": 3.018448664350554e-06, "loss": 0.0089, "step": 89440 }, { "grad_norm": 0.10469410568475723, "learning_rate": 3.0127932361736055e-06, "loss": 0.0057, "step": 89450 }, { "grad_norm": 0.2034245878458023, "learning_rate": 3.00714294644695e-06, "loss": 0.0094, "step": 89460 }, { "grad_norm": 0.12496482580900192, "learning_rate": 3.0014977957884828e-06, "loss": 0.0097, "step": 89470 }, { "grad_norm": 0.15458281338214874, "learning_rate": 2.99585778481557e-06, "loss": 0.0082, "step": 89480 }, { "grad_norm": 0.13484305143356323, "learning_rate": 2.990222914144969e-06, "loss": 0.0066, "step": 89490 }, { "grad_norm": 0.15777361392974854, "learning_rate": 2.984593184392914e-06, "loss": 0.0112, "step": 89500 }, { "grad_norm": 0.14985710382461548, "learning_rate": 2.978968596175069e-06, "loss": 0.0071, "step": 89510 }, { "grad_norm": 0.09875991195440292, "learning_rate": 2.9733491501065137e-06, "loss": 0.0076, "step": 89520 }, { "grad_norm": 0.12081552296876907, "learning_rate": 2.9677348468018005e-06, "loss": 0.0083, "step": 89530 }, { "grad_norm": 0.1661553531885147, "learning_rate": 2.9621256868748726e-06, "loss": 0.0075, "step": 89540 }, { "grad_norm": 0.14160814881324768, "learning_rate": 2.956521670939183e-06, "loss": 0.0095, "step": 89550 }, { "grad_norm": 0.16698651015758514, "learning_rate": 2.950922799607542e-06, "loss": 0.0082, "step": 89560 }, { "grad_norm": 0.14909468591213226, "learning_rate": 2.9453290734922535e-06, "loss": 0.0077, "step": 89570 }, { "grad_norm": 0.12367936223745346, "learning_rate": 2.939740493205029e-06, "loss": 0.0081, "step": 89580 }, { "grad_norm": 0.1155535951256752, "learning_rate": 2.9341570593570345e-06, "loss": 0.0066, "step": 89590 }, { "grad_norm": 0.09665708243846893, "learning_rate": 2.928578772558843e-06, "loss": 0.0085, "step": 89600 }, { "grad_norm": 0.0886746421456337, "learning_rate": 2.923005633420528e-06, "loss": 0.0064, "step": 89610 }, { "grad_norm": 0.11142878234386444, "learning_rate": 2.9174376425515237e-06, "loss": 0.0081, "step": 89620 }, { "grad_norm": 0.19447121024131775, "learning_rate": 2.91187480056076e-06, "loss": 0.0108, "step": 89630 }, { "grad_norm": 0.11360254138708115, "learning_rate": 2.9063171080565664e-06, "loss": 0.0084, "step": 89640 }, { "grad_norm": 0.1074838861823082, "learning_rate": 2.900764565646741e-06, "loss": 0.0105, "step": 89650 }, { "grad_norm": 0.29077842831611633, "learning_rate": 2.895217173938475e-06, "loss": 0.0111, "step": 89660 }, { "grad_norm": 0.16419222950935364, "learning_rate": 2.8896749335384445e-06, "loss": 0.0092, "step": 89670 }, { "grad_norm": 0.1347251981496811, "learning_rate": 2.884137845052731e-06, "loss": 0.0102, "step": 89680 }, { "grad_norm": 0.10265078395605087, "learning_rate": 2.8786059090868734e-06, "loss": 0.007, "step": 89690 }, { "grad_norm": 0.1174563616514206, "learning_rate": 2.87307912624582e-06, "loss": 0.012, "step": 89700 }, { "grad_norm": 0.14357294142246246, "learning_rate": 2.867557497133977e-06, "loss": 0.009, "step": 89710 }, { "grad_norm": 0.1999499648809433, "learning_rate": 2.862041022355194e-06, "loss": 0.0091, "step": 89720 }, { "grad_norm": 0.10543551295995712, "learning_rate": 2.8565297025127226e-06, "loss": 0.0089, "step": 89730 }, { "grad_norm": 0.13249452412128448, "learning_rate": 2.8510235382092853e-06, "loss": 0.0083, "step": 89740 }, { "grad_norm": 0.07933539152145386, "learning_rate": 2.845522530047029e-06, "loss": 0.0092, "step": 89750 }, { "grad_norm": 0.13072772324085236, "learning_rate": 2.8400266786275387e-06, "loss": 0.0082, "step": 89760 }, { "grad_norm": 0.09070379287004471, "learning_rate": 2.834535984551817e-06, "loss": 0.0079, "step": 89770 }, { "grad_norm": 0.13483847677707672, "learning_rate": 2.8290504484203397e-06, "loss": 0.0076, "step": 89780 }, { "grad_norm": 0.08155910670757294, "learning_rate": 2.8235700708329705e-06, "loss": 0.0088, "step": 89790 }, { "grad_norm": 0.2998940348625183, "learning_rate": 2.818094852389058e-06, "loss": 0.0077, "step": 89800 }, { "grad_norm": 0.15602761507034302, "learning_rate": 2.8126247936873516e-06, "loss": 0.0074, "step": 89810 }, { "grad_norm": 0.15040630102157593, "learning_rate": 2.8071598953260614e-06, "loss": 0.0097, "step": 89820 }, { "grad_norm": 0.08959247171878815, "learning_rate": 2.8017001579028035e-06, "loss": 0.0074, "step": 89830 }, { "grad_norm": 0.0738276019692421, "learning_rate": 2.7962455820146617e-06, "loss": 0.0092, "step": 89840 }, { "grad_norm": 0.09792403876781464, "learning_rate": 2.7907961682581253e-06, "loss": 0.0106, "step": 89850 }, { "grad_norm": 0.20583847165107727, "learning_rate": 2.7853519172291453e-06, "loss": 0.0124, "step": 89860 }, { "grad_norm": 0.0998949334025383, "learning_rate": 2.7799128295230904e-06, "loss": 0.0086, "step": 89870 }, { "grad_norm": 0.1335011124610901, "learning_rate": 2.774478905734784e-06, "loss": 0.0082, "step": 89880 }, { "grad_norm": 0.11438063532114029, "learning_rate": 2.7690501464584516e-06, "loss": 0.0082, "step": 89890 }, { "grad_norm": 0.09650150686502457, "learning_rate": 2.763626552287796e-06, "loss": 0.0083, "step": 89900 }, { "grad_norm": 0.10997875779867172, "learning_rate": 2.7582081238159083e-06, "loss": 0.0086, "step": 89910 }, { "grad_norm": 0.0794844850897789, "learning_rate": 2.7527948616353548e-06, "loss": 0.0086, "step": 89920 }, { "grad_norm": 0.14592553675174713, "learning_rate": 2.7473867663381226e-06, "loss": 0.007, "step": 89930 }, { "grad_norm": 0.15252411365509033, "learning_rate": 2.7419838385156383e-06, "loss": 0.0089, "step": 89940 }, { "grad_norm": 0.14110691845417023, "learning_rate": 2.7365860787587407e-06, "loss": 0.0083, "step": 89950 }, { "grad_norm": 0.11185404658317566, "learning_rate": 2.7311934876577295e-06, "loss": 0.0078, "step": 89960 }, { "grad_norm": 0.11866705864667892, "learning_rate": 2.725806065802339e-06, "loss": 0.0105, "step": 89970 }, { "grad_norm": 0.0912938266992569, "learning_rate": 2.720423813781714e-06, "loss": 0.0102, "step": 89980 }, { "grad_norm": 0.12196758389472961, "learning_rate": 2.715046732184462e-06, "loss": 0.0086, "step": 89990 }, { "grad_norm": 0.12115855515003204, "learning_rate": 2.7096748215986013e-06, "loss": 0.0084, "step": 90000 }, { "grad_norm": 0.11556095629930496, "learning_rate": 2.704308082611612e-06, "loss": 0.0092, "step": 90010 }, { "grad_norm": 0.13953839242458344, "learning_rate": 2.698946515810374e-06, "loss": 0.0086, "step": 90020 }, { "grad_norm": 0.1135924682021141, "learning_rate": 2.6935901217812367e-06, "loss": 0.0095, "step": 90030 }, { "grad_norm": 0.15786108374595642, "learning_rate": 2.688238901109952e-06, "loss": 0.0078, "step": 90040 }, { "grad_norm": 0.11868735402822495, "learning_rate": 2.68289285438173e-06, "loss": 0.0078, "step": 90050 }, { "grad_norm": 0.1308082938194275, "learning_rate": 2.6775519821812033e-06, "loss": 0.0083, "step": 90060 }, { "grad_norm": 0.14392778277397156, "learning_rate": 2.672216285092449e-06, "loss": 0.0088, "step": 90070 }, { "grad_norm": 0.1468891054391861, "learning_rate": 2.666885763698962e-06, "loss": 0.0093, "step": 90080 }, { "grad_norm": 0.18772326409816742, "learning_rate": 2.661560418583692e-06, "loss": 0.0084, "step": 90090 }, { "grad_norm": 0.13770842552185059, "learning_rate": 2.656240250328995e-06, "loss": 0.0076, "step": 90100 }, { "grad_norm": 0.14905594289302826, "learning_rate": 2.650925259516679e-06, "loss": 0.0081, "step": 90110 }, { "grad_norm": 0.157478928565979, "learning_rate": 2.645615446727995e-06, "loss": 0.0114, "step": 90120 }, { "grad_norm": 0.2008921205997467, "learning_rate": 2.6403108125436115e-06, "loss": 0.0092, "step": 90130 }, { "grad_norm": 0.1310180425643921, "learning_rate": 2.6350113575436264e-06, "loss": 0.0074, "step": 90140 }, { "grad_norm": 0.10632389783859253, "learning_rate": 2.629717082307598e-06, "loss": 0.0086, "step": 90150 }, { "grad_norm": 0.19654129445552826, "learning_rate": 2.6244279874144793e-06, "loss": 0.0095, "step": 90160 }, { "grad_norm": 0.13957767188549042, "learning_rate": 2.6191440734426862e-06, "loss": 0.0079, "step": 90170 }, { "grad_norm": 0.12589699029922485, "learning_rate": 2.613865340970062e-06, "loss": 0.0062, "step": 90180 }, { "grad_norm": 0.0765112116932869, "learning_rate": 2.608591790573889e-06, "loss": 0.0072, "step": 90190 }, { "grad_norm": 0.10358778387308121, "learning_rate": 2.603323422830861e-06, "loss": 0.0081, "step": 90200 }, { "grad_norm": 0.20175406336784363, "learning_rate": 2.5980602383171183e-06, "loss": 0.0079, "step": 90210 }, { "grad_norm": 0.12040332704782486, "learning_rate": 2.5928022376082496e-06, "loss": 0.0077, "step": 90220 }, { "grad_norm": 0.1339404582977295, "learning_rate": 2.5875494212792505e-06, "loss": 0.0066, "step": 90230 }, { "grad_norm": 0.15684455633163452, "learning_rate": 2.5823017899045565e-06, "loss": 0.0085, "step": 90240 }, { "grad_norm": 0.13142642378807068, "learning_rate": 2.577059344058047e-06, "loss": 0.0074, "step": 90250 }, { "grad_norm": 0.1929343044757843, "learning_rate": 2.5718220843130415e-06, "loss": 0.0078, "step": 90260 }, { "grad_norm": 0.09609754383563995, "learning_rate": 2.5665900112422535e-06, "loss": 0.0066, "step": 90270 }, { "grad_norm": 0.1619696468114853, "learning_rate": 2.561363125417876e-06, "loss": 0.0105, "step": 90280 }, { "grad_norm": 0.0948452353477478, "learning_rate": 2.556141427411496e-06, "loss": 0.0084, "step": 90290 }, { "grad_norm": 0.2047027051448822, "learning_rate": 2.550924917794156e-06, "loss": 0.0102, "step": 90300 }, { "grad_norm": 0.09651698172092438, "learning_rate": 2.545713597136329e-06, "loss": 0.0077, "step": 90310 }, { "grad_norm": 0.14596670866012573, "learning_rate": 2.5405074660079298e-06, "loss": 0.0095, "step": 90320 }, { "grad_norm": 0.10644862800836563, "learning_rate": 2.5353065249782647e-06, "loss": 0.0073, "step": 90330 }, { "grad_norm": 0.16275352239608765, "learning_rate": 2.530110774616129e-06, "loss": 0.0086, "step": 90340 }, { "grad_norm": 0.13126270473003387, "learning_rate": 2.524920215489701e-06, "loss": 0.0107, "step": 90350 }, { "grad_norm": 0.20875480771064758, "learning_rate": 2.519734848166627e-06, "loss": 0.0094, "step": 90360 }, { "grad_norm": 0.13933616876602173, "learning_rate": 2.514554673213948e-06, "loss": 0.0083, "step": 90370 }, { "grad_norm": 0.14410333335399628, "learning_rate": 2.5093796911981936e-06, "loss": 0.0091, "step": 90380 }, { "grad_norm": 0.15120361745357513, "learning_rate": 2.5042099026852672e-06, "loss": 0.0087, "step": 90390 }, { "grad_norm": 0.18543514609336853, "learning_rate": 2.499045308240544e-06, "loss": 0.008, "step": 90400 }, { "grad_norm": 0.10472708195447922, "learning_rate": 2.4938859084288003e-06, "loss": 0.0088, "step": 90410 }, { "grad_norm": 0.10972079634666443, "learning_rate": 2.4887317038142734e-06, "loss": 0.0078, "step": 90420 }, { "grad_norm": 0.07735613733530045, "learning_rate": 2.4835826949606124e-06, "loss": 0.0061, "step": 90430 }, { "grad_norm": 0.17112144827842712, "learning_rate": 2.4784388824309167e-06, "loss": 0.0067, "step": 90440 }, { "grad_norm": 0.08280447870492935, "learning_rate": 2.473300266787687e-06, "loss": 0.0068, "step": 90450 }, { "grad_norm": 0.19262471795082092, "learning_rate": 2.46816684859289e-06, "loss": 0.0087, "step": 90460 }, { "grad_norm": 0.17616599798202515, "learning_rate": 2.4630386284079043e-06, "loss": 0.0091, "step": 90470 }, { "grad_norm": 0.13626879453659058, "learning_rate": 2.4579156067935427e-06, "loss": 0.0095, "step": 90480 }, { "grad_norm": 0.18933330476284027, "learning_rate": 2.452797784310046e-06, "loss": 0.0089, "step": 90490 }, { "grad_norm": 0.13963554799556732, "learning_rate": 2.447685161517105e-06, "loss": 0.0134, "step": 90500 }, { "grad_norm": 0.18537858128547668, "learning_rate": 2.4425777389738224e-06, "loss": 0.0079, "step": 90510 }, { "grad_norm": 0.11950767785310745, "learning_rate": 2.43747551723873e-06, "loss": 0.0079, "step": 90520 }, { "grad_norm": 0.08548154681921005, "learning_rate": 2.4323784968698136e-06, "loss": 0.0077, "step": 90530 }, { "grad_norm": 0.13549622893333435, "learning_rate": 2.4272866784244565e-06, "loss": 0.0092, "step": 90540 }, { "grad_norm": 0.10460367798805237, "learning_rate": 2.422200062459518e-06, "loss": 0.0072, "step": 90550 }, { "grad_norm": 0.10060634464025497, "learning_rate": 2.41711864953123e-06, "loss": 0.009, "step": 90560 }, { "grad_norm": 0.1443660855293274, "learning_rate": 2.412042440195322e-06, "loss": 0.0081, "step": 90570 }, { "grad_norm": 0.10012122243642807, "learning_rate": 2.4069714350068993e-06, "loss": 0.008, "step": 90580 }, { "grad_norm": 0.11099059879779816, "learning_rate": 2.4019056345205294e-06, "loss": 0.0081, "step": 90590 }, { "grad_norm": 0.09196445345878601, "learning_rate": 2.396845039290191e-06, "loss": 0.01, "step": 90600 }, { "grad_norm": 0.08856155723333359, "learning_rate": 2.3917896498693194e-06, "loss": 0.0081, "step": 90610 }, { "grad_norm": 0.20959234237670898, "learning_rate": 2.3867394668107335e-06, "loss": 0.0086, "step": 90620 }, { "grad_norm": 0.10582567006349564, "learning_rate": 2.3816944906667527e-06, "loss": 0.0072, "step": 90630 }, { "grad_norm": 0.15941986441612244, "learning_rate": 2.3766547219890634e-06, "loss": 0.0091, "step": 90640 }, { "grad_norm": 0.19752316176891327, "learning_rate": 2.371620161328819e-06, "loss": 0.0091, "step": 90650 }, { "grad_norm": 0.07726141065359116, "learning_rate": 2.366590809236574e-06, "loss": 0.0083, "step": 90660 }, { "grad_norm": 0.1520884484052658, "learning_rate": 2.3615666662623494e-06, "loss": 0.0083, "step": 90670 }, { "grad_norm": 0.1777196228504181, "learning_rate": 2.3565477329555664e-06, "loss": 0.0106, "step": 90680 }, { "grad_norm": 0.21189747750759125, "learning_rate": 2.3515340098650975e-06, "loss": 0.0087, "step": 90690 }, { "grad_norm": 0.11109867691993713, "learning_rate": 2.3465254975392258e-06, "loss": 0.0071, "step": 90700 }, { "grad_norm": 0.19562307000160217, "learning_rate": 2.3415221965256805e-06, "loss": 0.0079, "step": 90710 }, { "grad_norm": 0.0870208665728569, "learning_rate": 2.3365241073716237e-06, "loss": 0.0084, "step": 90720 }, { "grad_norm": 0.27743420004844666, "learning_rate": 2.3315312306236183e-06, "loss": 0.0104, "step": 90730 }, { "grad_norm": 0.20022694766521454, "learning_rate": 2.326543566827699e-06, "loss": 0.0096, "step": 90740 }, { "grad_norm": 0.14990660548210144, "learning_rate": 2.3215611165292862e-06, "loss": 0.0098, "step": 90750 }, { "grad_norm": 0.14426255226135254, "learning_rate": 2.3165838802732765e-06, "loss": 0.0081, "step": 90760 }, { "grad_norm": 0.11496232450008392, "learning_rate": 2.311611858603957e-06, "loss": 0.0085, "step": 90770 }, { "grad_norm": 0.10119140148162842, "learning_rate": 2.3066450520650763e-06, "loss": 0.0109, "step": 90780 }, { "grad_norm": 0.1095384880900383, "learning_rate": 2.3016834611997773e-06, "loss": 0.0103, "step": 90790 }, { "grad_norm": 0.15092524886131287, "learning_rate": 2.296727086550665e-06, "loss": 0.0089, "step": 90800 }, { "grad_norm": 0.11120998859405518, "learning_rate": 2.2917759286597507e-06, "loss": 0.0086, "step": 90810 }, { "grad_norm": 0.0904601588845253, "learning_rate": 2.2868299880685006e-06, "loss": 0.0077, "step": 90820 }, { "grad_norm": 0.0807482972741127, "learning_rate": 2.2818892653177827e-06, "loss": 0.0071, "step": 90830 }, { "grad_norm": 0.1101904958486557, "learning_rate": 2.2769537609479196e-06, "loss": 0.0067, "step": 90840 }, { "grad_norm": 0.09970057010650635, "learning_rate": 2.2720234754986357e-06, "loss": 0.0057, "step": 90850 }, { "grad_norm": 0.15987341105937958, "learning_rate": 2.2670984095091108e-06, "loss": 0.0072, "step": 90860 }, { "grad_norm": 0.0828036367893219, "learning_rate": 2.262178563517936e-06, "loss": 0.0071, "step": 90870 }, { "grad_norm": 0.18372812867164612, "learning_rate": 2.257263938063137e-06, "loss": 0.0092, "step": 90880 }, { "grad_norm": 0.1706087291240692, "learning_rate": 2.2523545336821715e-06, "loss": 0.0127, "step": 90890 }, { "grad_norm": 0.14898785948753357, "learning_rate": 2.247450350911939e-06, "loss": 0.0066, "step": 90900 }, { "grad_norm": 0.07093994319438934, "learning_rate": 2.2425513902887328e-06, "loss": 0.0073, "step": 90910 }, { "grad_norm": 0.12337397783994675, "learning_rate": 2.2376576523483015e-06, "loss": 0.008, "step": 90920 }, { "grad_norm": 0.16081838309764862, "learning_rate": 2.232769137625829e-06, "loss": 0.0091, "step": 90930 }, { "grad_norm": 0.07601836323738098, "learning_rate": 2.227885846655903e-06, "loss": 0.0104, "step": 90940 }, { "grad_norm": 0.1470976322889328, "learning_rate": 2.2230077799725523e-06, "loss": 0.0095, "step": 90950 }, { "grad_norm": 0.11315618455410004, "learning_rate": 2.2181349381092396e-06, "loss": 0.0062, "step": 90960 }, { "grad_norm": 0.08913811296224594, "learning_rate": 2.2132673215988552e-06, "loss": 0.0074, "step": 90970 }, { "grad_norm": 0.1345108598470688, "learning_rate": 2.2084049309737065e-06, "loss": 0.0082, "step": 90980 }, { "grad_norm": 0.20556363463401794, "learning_rate": 2.2035477667655513e-06, "loss": 0.0062, "step": 90990 }, { "grad_norm": 0.1167207807302475, "learning_rate": 2.1986958295055315e-06, "loss": 0.0076, "step": 91000 }, { "grad_norm": 0.10240515321493149, "learning_rate": 2.193849119724284e-06, "loss": 0.0092, "step": 91010 }, { "grad_norm": 0.14454123377799988, "learning_rate": 2.1890076379518175e-06, "loss": 0.0078, "step": 91020 }, { "grad_norm": 0.11318553239107132, "learning_rate": 2.1841713847175925e-06, "loss": 0.0089, "step": 91030 }, { "grad_norm": 0.2254355251789093, "learning_rate": 2.1793403605504968e-06, "loss": 0.0132, "step": 91040 }, { "grad_norm": 0.10616457462310791, "learning_rate": 2.1745145659788413e-06, "loss": 0.0075, "step": 91050 }, { "grad_norm": 0.09823562949895859, "learning_rate": 2.1696940015303645e-06, "loss": 0.0068, "step": 91060 }, { "grad_norm": 0.12313251197338104, "learning_rate": 2.1648786677322397e-06, "loss": 0.0076, "step": 91070 }, { "grad_norm": 0.13040603697299957, "learning_rate": 2.1600685651110673e-06, "loss": 0.0083, "step": 91080 }, { "grad_norm": 0.11676699668169022, "learning_rate": 2.1552636941928717e-06, "loss": 0.0072, "step": 91090 }, { "grad_norm": 0.11359050124883652, "learning_rate": 2.1504640555031043e-06, "loss": 0.0076, "step": 91100 }, { "grad_norm": 0.08623821288347244, "learning_rate": 2.1456696495666506e-06, "loss": 0.0088, "step": 91110 }, { "grad_norm": 0.12038818746805191, "learning_rate": 2.1408804769078084e-06, "loss": 0.0087, "step": 91120 }, { "grad_norm": 0.19268523156642914, "learning_rate": 2.136096538050325e-06, "loss": 0.0102, "step": 91130 }, { "grad_norm": 0.09101323038339615, "learning_rate": 2.13131783351736e-06, "loss": 0.0066, "step": 91140 }, { "grad_norm": 0.1451372504234314, "learning_rate": 2.126544363831512e-06, "loss": 0.0087, "step": 91150 }, { "grad_norm": 0.11351412534713745, "learning_rate": 2.121776129514791e-06, "loss": 0.0084, "step": 91160 }, { "grad_norm": 0.20425862073898315, "learning_rate": 2.1170131310886523e-06, "loss": 0.0067, "step": 91170 }, { "grad_norm": 0.10823748260736465, "learning_rate": 2.112255369073968e-06, "loss": 0.0107, "step": 91180 }, { "grad_norm": 0.15118002891540527, "learning_rate": 2.1075028439910328e-06, "loss": 0.0081, "step": 91190 }, { "grad_norm": 0.07800185680389404, "learning_rate": 2.1027555563595813e-06, "loss": 0.01, "step": 91200 }, { "grad_norm": 0.16058294475078583, "learning_rate": 2.0980135066987706e-06, "loss": 0.0092, "step": 91210 }, { "grad_norm": 0.09276223182678223, "learning_rate": 2.093276695527191e-06, "loss": 0.008, "step": 91220 }, { "grad_norm": 0.10282187163829803, "learning_rate": 2.0885451233628396e-06, "loss": 0.0068, "step": 91230 }, { "grad_norm": 0.18184566497802734, "learning_rate": 2.0838187907231633e-06, "loss": 0.0081, "step": 91240 }, { "grad_norm": 0.1340354084968567, "learning_rate": 2.0790976981250153e-06, "loss": 0.0083, "step": 91250 }, { "grad_norm": 0.14828072488307953, "learning_rate": 2.0743818460847e-06, "loss": 0.0073, "step": 91260 }, { "grad_norm": 0.07981576770544052, "learning_rate": 2.069671235117926e-06, "loss": 0.0078, "step": 91270 }, { "grad_norm": 0.10749437659978867, "learning_rate": 2.0649658657398484e-06, "loss": 0.009, "step": 91280 }, { "grad_norm": 0.09247507899999619, "learning_rate": 2.060265738465034e-06, "loss": 0.0083, "step": 91290 }, { "grad_norm": 0.13210690021514893, "learning_rate": 2.055570853807487e-06, "loss": 0.007, "step": 91300 }, { "grad_norm": 0.09073757380247116, "learning_rate": 2.0508812122806196e-06, "loss": 0.0064, "step": 91310 }, { "grad_norm": 0.10432624071836472, "learning_rate": 2.046196814397294e-06, "loss": 0.0086, "step": 91320 }, { "grad_norm": 0.12993159890174866, "learning_rate": 2.0415176606697896e-06, "loss": 0.0081, "step": 91330 }, { "grad_norm": 0.13725127279758453, "learning_rate": 2.0368437516098126e-06, "loss": 0.0086, "step": 91340 }, { "grad_norm": 0.13170881569385529, "learning_rate": 2.032175087728483e-06, "loss": 0.0079, "step": 91350 }, { "grad_norm": 0.12245038896799088, "learning_rate": 2.027511669536375e-06, "loss": 0.0086, "step": 91360 }, { "grad_norm": 0.08773387968540192, "learning_rate": 2.0228534975434643e-06, "loss": 0.0071, "step": 91370 }, { "grad_norm": 0.13602006435394287, "learning_rate": 2.0182005722591548e-06, "loss": 0.0087, "step": 91380 }, { "grad_norm": 0.1279049962759018, "learning_rate": 2.0135528941922944e-06, "loss": 0.0069, "step": 91390 }, { "grad_norm": 0.09647543728351593, "learning_rate": 2.008910463851149e-06, "loss": 0.0085, "step": 91400 }, { "grad_norm": 0.1758819967508316, "learning_rate": 2.0042732817433963e-06, "loss": 0.0111, "step": 91410 }, { "grad_norm": 0.11740801483392715, "learning_rate": 1.999641348376158e-06, "loss": 0.0078, "step": 91420 }, { "grad_norm": 0.08144783228635788, "learning_rate": 1.9950146642559787e-06, "loss": 0.0084, "step": 91430 }, { "grad_norm": 0.20617549121379852, "learning_rate": 1.990393229888815e-06, "loss": 0.0072, "step": 91440 }, { "grad_norm": 0.22562459111213684, "learning_rate": 1.9857770457800684e-06, "loss": 0.0102, "step": 91450 }, { "grad_norm": 0.09028200060129166, "learning_rate": 1.9811661124345517e-06, "loss": 0.0071, "step": 91460 }, { "grad_norm": 0.1647145003080368, "learning_rate": 1.976560430356522e-06, "loss": 0.0091, "step": 91470 }, { "grad_norm": 0.11997902393341064, "learning_rate": 1.9719600000496383e-06, "loss": 0.0094, "step": 91480 }, { "grad_norm": 0.20119085907936096, "learning_rate": 1.967364822017004e-06, "loss": 0.01, "step": 91490 }, { "grad_norm": 0.17142891883850098, "learning_rate": 1.962774896761127e-06, "loss": 0.0123, "step": 91500 }, { "grad_norm": 0.17482440173625946, "learning_rate": 1.958190224783968e-06, "loss": 0.0091, "step": 91510 }, { "grad_norm": 0.1287485808134079, "learning_rate": 1.9536108065868975e-06, "loss": 0.0093, "step": 91520 }, { "grad_norm": 0.12671397626399994, "learning_rate": 1.949036642670715e-06, "loss": 0.0106, "step": 91530 }, { "grad_norm": 0.13287991285324097, "learning_rate": 1.9444677335356377e-06, "loss": 0.0104, "step": 91540 }, { "grad_norm": 0.3058326542377472, "learning_rate": 1.939904079681326e-06, "loss": 0.0104, "step": 91550 }, { "grad_norm": 0.11957911401987076, "learning_rate": 1.935345681606837e-06, "loss": 0.0111, "step": 91560 }, { "grad_norm": 0.0879470705986023, "learning_rate": 1.930792539810694e-06, "loss": 0.0104, "step": 91570 }, { "grad_norm": 0.16281092166900635, "learning_rate": 1.9262446547907877e-06, "loss": 0.0082, "step": 91580 }, { "grad_norm": 0.19178706407546997, "learning_rate": 1.921702027044503e-06, "loss": 0.0079, "step": 91590 }, { "grad_norm": 0.15938077867031097, "learning_rate": 1.917164657068593e-06, "loss": 0.0128, "step": 91600 }, { "grad_norm": 0.12588784098625183, "learning_rate": 1.912632545359272e-06, "loss": 0.0099, "step": 91610 }, { "grad_norm": 0.2079719752073288, "learning_rate": 1.9081056924121488e-06, "loss": 0.0084, "step": 91620 }, { "grad_norm": 0.1432129591703415, "learning_rate": 1.9035840987222887e-06, "loss": 0.0101, "step": 91630 }, { "grad_norm": 0.15446563065052032, "learning_rate": 1.8990677647841515e-06, "loss": 0.008, "step": 91640 }, { "grad_norm": 0.1128038838505745, "learning_rate": 1.8945566910916534e-06, "loss": 0.0072, "step": 91650 }, { "grad_norm": 0.17332938313484192, "learning_rate": 1.8900508781381054e-06, "loss": 0.0085, "step": 91660 }, { "grad_norm": 0.08740909397602081, "learning_rate": 1.8855503264162633e-06, "loss": 0.0061, "step": 91670 }, { "grad_norm": 0.1790817230939865, "learning_rate": 1.8810550364182998e-06, "loss": 0.0124, "step": 91680 }, { "grad_norm": 0.14413276314735413, "learning_rate": 1.8765650086358112e-06, "loss": 0.0104, "step": 91690 }, { "grad_norm": 0.14778347313404083, "learning_rate": 1.8720802435598151e-06, "loss": 0.0097, "step": 91700 }, { "grad_norm": 0.11800851672887802, "learning_rate": 1.8676007416807694e-06, "loss": 0.0079, "step": 91710 }, { "grad_norm": 0.23322664201259613, "learning_rate": 1.8631265034885437e-06, "loss": 0.0108, "step": 91720 }, { "grad_norm": 0.12196581810712814, "learning_rate": 1.858657529472424e-06, "loss": 0.0097, "step": 91730 }, { "grad_norm": 0.09287804365158081, "learning_rate": 1.8541938201211473e-06, "loss": 0.0113, "step": 91740 }, { "grad_norm": 0.08911534398794174, "learning_rate": 1.8497353759228398e-06, "loss": 0.0074, "step": 91750 }, { "grad_norm": 0.11817532777786255, "learning_rate": 1.8452821973650835e-06, "loss": 0.0105, "step": 91760 }, { "grad_norm": 0.21908068656921387, "learning_rate": 1.840834284934856e-06, "loss": 0.0082, "step": 91770 }, { "grad_norm": 0.09791151434183121, "learning_rate": 1.8363916391185954e-06, "loss": 0.0078, "step": 91780 }, { "grad_norm": 0.10540073364973068, "learning_rate": 1.8319542604021245e-06, "loss": 0.0072, "step": 91790 }, { "grad_norm": 0.088847815990448, "learning_rate": 1.8275221492707273e-06, "loss": 0.0067, "step": 91800 }, { "grad_norm": 0.0992172434926033, "learning_rate": 1.8230953062090717e-06, "loss": 0.008, "step": 91810 }, { "grad_norm": 0.10878337174654007, "learning_rate": 1.8186737317012926e-06, "loss": 0.0076, "step": 91820 }, { "grad_norm": 0.1323704570531845, "learning_rate": 1.8142574262308977e-06, "loss": 0.0084, "step": 91830 }, { "grad_norm": 0.13142934441566467, "learning_rate": 1.8098463902808727e-06, "loss": 0.0097, "step": 91840 }, { "grad_norm": 0.12171225994825363, "learning_rate": 1.805440624333593e-06, "loss": 0.0058, "step": 91850 }, { "grad_norm": 0.09406642615795135, "learning_rate": 1.8010401288708734e-06, "loss": 0.0071, "step": 91860 }, { "grad_norm": 0.14084786176681519, "learning_rate": 1.7966449043739343e-06, "loss": 0.0091, "step": 91870 }, { "grad_norm": 0.11835573613643646, "learning_rate": 1.79225495132343e-06, "loss": 0.0079, "step": 91880 }, { "grad_norm": 0.11561807245016098, "learning_rate": 1.7878702701994487e-06, "loss": 0.0078, "step": 91890 }, { "grad_norm": 0.16211266815662384, "learning_rate": 1.783490861481496e-06, "loss": 0.0069, "step": 91900 }, { "grad_norm": 0.091048464179039, "learning_rate": 1.7791167256484832e-06, "loss": 0.0066, "step": 91910 }, { "grad_norm": 0.21172256767749786, "learning_rate": 1.774747863178766e-06, "loss": 0.0076, "step": 91920 }, { "grad_norm": 0.13161160051822662, "learning_rate": 1.7703842745501186e-06, "loss": 0.0099, "step": 91930 }, { "grad_norm": 0.13257628679275513, "learning_rate": 1.7660259602397366e-06, "loss": 0.0075, "step": 91940 }, { "grad_norm": 0.06294240057468414, "learning_rate": 1.7616729207242388e-06, "loss": 0.008, "step": 91950 }, { "grad_norm": 0.11747386306524277, "learning_rate": 1.7573251564796556e-06, "loss": 0.0068, "step": 91960 }, { "grad_norm": 0.14283238351345062, "learning_rate": 1.7529826679814676e-06, "loss": 0.009, "step": 91970 }, { "grad_norm": 0.0914459228515625, "learning_rate": 1.7486454557045562e-06, "loss": 0.0082, "step": 91980 }, { "grad_norm": 0.16540855169296265, "learning_rate": 1.7443135201232364e-06, "loss": 0.0107, "step": 91990 }, { "grad_norm": 0.14911507070064545, "learning_rate": 1.7399868617112347e-06, "loss": 0.0073, "step": 92000 }, { "grad_norm": 0.14535702764987946, "learning_rate": 1.7356654809417171e-06, "loss": 0.0081, "step": 92010 }, { "grad_norm": 0.13336950540542603, "learning_rate": 1.7313493782872448e-06, "loss": 0.0079, "step": 92020 }, { "grad_norm": 0.09835833311080933, "learning_rate": 1.7270385542198453e-06, "loss": 0.0079, "step": 92030 }, { "grad_norm": 0.12222566455602646, "learning_rate": 1.7227330092109307e-06, "loss": 0.0073, "step": 92040 }, { "grad_norm": 0.07538717985153198, "learning_rate": 1.718432743731352e-06, "loss": 0.0076, "step": 92050 }, { "grad_norm": 0.11491751670837402, "learning_rate": 1.7141377582513718e-06, "loss": 0.0074, "step": 92060 }, { "grad_norm": 0.10065244138240814, "learning_rate": 1.709848053240698e-06, "loss": 0.0089, "step": 92070 }, { "grad_norm": 0.12435133755207062, "learning_rate": 1.7055636291684328e-06, "loss": 0.0084, "step": 92080 }, { "grad_norm": 0.10030601918697357, "learning_rate": 1.7012844865031185e-06, "loss": 0.0109, "step": 92090 }, { "grad_norm": 0.09060591459274292, "learning_rate": 1.6970106257127138e-06, "loss": 0.0084, "step": 92100 }, { "grad_norm": 0.16224928200244904, "learning_rate": 1.6927420472646117e-06, "loss": 0.0102, "step": 92110 }, { "grad_norm": 0.11163011938333511, "learning_rate": 1.6884787516256106e-06, "loss": 0.008, "step": 92120 }, { "grad_norm": 0.1544775813817978, "learning_rate": 1.6842207392619325e-06, "loss": 0.0091, "step": 92130 }, { "grad_norm": 0.13766004145145416, "learning_rate": 1.679968010639238e-06, "loss": 0.0076, "step": 92140 }, { "grad_norm": 0.15253083407878876, "learning_rate": 1.6757205662225829e-06, "loss": 0.0101, "step": 92150 }, { "grad_norm": 0.17852404713630676, "learning_rate": 1.6714784064764844e-06, "loss": 0.01, "step": 92160 }, { "grad_norm": 0.14559349417686462, "learning_rate": 1.6672415318648383e-06, "loss": 0.0072, "step": 92170 }, { "grad_norm": 0.11975900083780289, "learning_rate": 1.6630099428510016e-06, "loss": 0.0089, "step": 92180 }, { "grad_norm": 0.11024992913007736, "learning_rate": 1.6587836398977153e-06, "loss": 0.0077, "step": 92190 }, { "grad_norm": 0.15617159008979797, "learning_rate": 1.6545626234671762e-06, "loss": 0.0087, "step": 92200 }, { "grad_norm": 0.1325240135192871, "learning_rate": 1.650346894020971e-06, "loss": 0.0077, "step": 92210 }, { "grad_norm": 0.08845226466655731, "learning_rate": 1.6461364520201472e-06, "loss": 0.0088, "step": 92220 }, { "grad_norm": 0.13116860389709473, "learning_rate": 1.6419312979251366e-06, "loss": 0.0101, "step": 92230 }, { "grad_norm": 0.1393711417913437, "learning_rate": 1.6377314321958214e-06, "loss": 0.0116, "step": 92240 }, { "grad_norm": 0.1375073790550232, "learning_rate": 1.6335368552914787e-06, "loss": 0.0085, "step": 92250 }, { "grad_norm": 0.15135683119297028, "learning_rate": 1.6293475676708358e-06, "loss": 0.0109, "step": 92260 }, { "grad_norm": 0.1326492875814438, "learning_rate": 1.6251635697920097e-06, "loss": 0.0101, "step": 92270 }, { "grad_norm": 0.07396429777145386, "learning_rate": 1.6209848621125624e-06, "loss": 0.0082, "step": 92280 }, { "grad_norm": 0.1437101811170578, "learning_rate": 1.6168114450894778e-06, "loss": 0.0085, "step": 92290 }, { "grad_norm": 0.10311688482761383, "learning_rate": 1.612643319179158e-06, "loss": 0.0072, "step": 92300 }, { "grad_norm": 0.09600823372602463, "learning_rate": 1.6084804848374046e-06, "loss": 0.0084, "step": 92310 }, { "grad_norm": 0.11177727580070496, "learning_rate": 1.6043229425194817e-06, "loss": 0.0068, "step": 92320 }, { "grad_norm": 0.13609278202056885, "learning_rate": 1.6001706926800308e-06, "loss": 0.0059, "step": 92330 }, { "grad_norm": 0.1093870997428894, "learning_rate": 1.5960237357731444e-06, "loss": 0.0082, "step": 92340 }, { "grad_norm": 0.18078789114952087, "learning_rate": 1.5918820722523264e-06, "loss": 0.0076, "step": 92350 }, { "grad_norm": 0.10616117715835571, "learning_rate": 1.587745702570509e-06, "loss": 0.007, "step": 92360 }, { "grad_norm": 0.16303224861621857, "learning_rate": 1.5836146271800356e-06, "loss": 0.0099, "step": 92370 }, { "grad_norm": 0.09411586076021194, "learning_rate": 1.5794888465326728e-06, "loss": 0.0088, "step": 92380 }, { "grad_norm": 0.08613062649965286, "learning_rate": 1.5753683610796155e-06, "loss": 0.0079, "step": 92390 }, { "grad_norm": 0.09900306910276413, "learning_rate": 1.5712531712714528e-06, "loss": 0.0089, "step": 92400 }, { "grad_norm": 0.14620396494865417, "learning_rate": 1.5671432775582474e-06, "loss": 0.0077, "step": 92410 }, { "grad_norm": 0.10088418424129486, "learning_rate": 1.563038680389428e-06, "loss": 0.007, "step": 92420 }, { "grad_norm": 0.1295395791530609, "learning_rate": 1.5589393802138808e-06, "loss": 0.0078, "step": 92430 }, { "grad_norm": 0.1391305774450302, "learning_rate": 1.554845377479891e-06, "loss": 0.0071, "step": 92440 }, { "grad_norm": 0.11954385042190552, "learning_rate": 1.550756672635173e-06, "loss": 0.0076, "step": 92450 }, { "grad_norm": 0.1725049614906311, "learning_rate": 1.5466732661268635e-06, "loss": 0.0083, "step": 92460 }, { "grad_norm": 0.1576758474111557, "learning_rate": 1.5425951584015163e-06, "loss": 0.0082, "step": 92470 }, { "grad_norm": 0.13604100048542023, "learning_rate": 1.5385223499051026e-06, "loss": 0.007, "step": 92480 }, { "grad_norm": 0.10906735062599182, "learning_rate": 1.5344548410830328e-06, "loss": 0.0056, "step": 92490 }, { "grad_norm": 0.124510757625103, "learning_rate": 1.5303926323801121e-06, "loss": 0.0085, "step": 92500 }, { "grad_norm": 0.19298799335956573, "learning_rate": 1.5263357242405795e-06, "loss": 0.0082, "step": 92510 }, { "grad_norm": 0.11878993362188339, "learning_rate": 1.5222841171080915e-06, "loss": 0.0079, "step": 92520 }, { "grad_norm": 0.11801428347826004, "learning_rate": 1.5182378114257268e-06, "loss": 0.0091, "step": 92530 }, { "grad_norm": 0.13437850773334503, "learning_rate": 1.5141968076359869e-06, "loss": 0.0082, "step": 92540 }, { "grad_norm": 0.1047186478972435, "learning_rate": 1.5101611061807908e-06, "loss": 0.0084, "step": 92550 }, { "grad_norm": 0.14395809173583984, "learning_rate": 1.506130707501463e-06, "loss": 0.0085, "step": 92560 }, { "grad_norm": 0.1520835906267166, "learning_rate": 1.5021056120387734e-06, "loss": 0.0096, "step": 92570 }, { "grad_norm": 0.0987882912158966, "learning_rate": 1.4980858202329084e-06, "loss": 0.0076, "step": 92580 }, { "grad_norm": 0.0954274833202362, "learning_rate": 1.4940713325234445e-06, "loss": 0.0069, "step": 92590 }, { "grad_norm": 0.10509327799081802, "learning_rate": 1.4900621493494137e-06, "loss": 0.0084, "step": 92600 }, { "grad_norm": 0.1527477502822876, "learning_rate": 1.4860582711492544e-06, "loss": 0.0089, "step": 92610 }, { "grad_norm": 0.09509608149528503, "learning_rate": 1.4820596983608325e-06, "loss": 0.0087, "step": 92620 }, { "grad_norm": 0.11606528609991074, "learning_rate": 1.4780664314214044e-06, "loss": 0.0102, "step": 92630 }, { "grad_norm": 0.20127053558826447, "learning_rate": 1.474078470767687e-06, "loss": 0.0092, "step": 92640 }, { "grad_norm": 0.11018767952919006, "learning_rate": 1.470095816835787e-06, "loss": 0.0101, "step": 92650 }, { "grad_norm": 0.12052349001169205, "learning_rate": 1.4661184700612395e-06, "loss": 0.0063, "step": 92660 }, { "grad_norm": 0.12793117761611938, "learning_rate": 1.4621464308790133e-06, "loss": 0.0108, "step": 92670 }, { "grad_norm": 0.13979782164096832, "learning_rate": 1.4581796997234775e-06, "loss": 0.0082, "step": 92680 }, { "grad_norm": 0.13074636459350586, "learning_rate": 1.4542182770284297e-06, "loss": 0.0119, "step": 92690 }, { "grad_norm": 0.11581946909427643, "learning_rate": 1.450262163227084e-06, "loss": 0.0071, "step": 92700 }, { "grad_norm": 0.13680480420589447, "learning_rate": 1.4463113587520671e-06, "loss": 0.0091, "step": 92710 }, { "grad_norm": 0.1269996464252472, "learning_rate": 1.4423658640354443e-06, "loss": 0.0087, "step": 92720 }, { "grad_norm": 0.14879293739795685, "learning_rate": 1.438425679508687e-06, "loss": 0.0098, "step": 92730 }, { "grad_norm": 0.10940929502248764, "learning_rate": 1.4344908056026895e-06, "loss": 0.0091, "step": 92740 }, { "grad_norm": 0.19576221704483032, "learning_rate": 1.4305612427477521e-06, "loss": 0.0087, "step": 92750 }, { "grad_norm": 0.13978908956050873, "learning_rate": 1.42663699137362e-06, "loss": 0.0093, "step": 92760 }, { "grad_norm": 0.18356184661388397, "learning_rate": 1.4227180519094386e-06, "loss": 0.0103, "step": 92770 }, { "grad_norm": 0.05446247383952141, "learning_rate": 1.4188044247837706e-06, "loss": 0.0068, "step": 92780 }, { "grad_norm": 0.11211320012807846, "learning_rate": 1.4148961104246128e-06, "loss": 0.0074, "step": 92790 }, { "grad_norm": 0.11106973141431808, "learning_rate": 1.4109931092593731e-06, "loss": 0.0073, "step": 92800 }, { "grad_norm": 0.07769685238599777, "learning_rate": 1.4070954217148658e-06, "loss": 0.0082, "step": 92810 }, { "grad_norm": 0.15693041682243347, "learning_rate": 1.4032030482173497e-06, "loss": 0.0075, "step": 92820 }, { "grad_norm": 0.26186177134513855, "learning_rate": 1.3993159891924901e-06, "loss": 0.009, "step": 92830 }, { "grad_norm": 0.1048855409026146, "learning_rate": 1.3954342450653524e-06, "loss": 0.0077, "step": 92840 }, { "grad_norm": 0.12871865928173065, "learning_rate": 1.3915578162604582e-06, "loss": 0.0075, "step": 92850 }, { "grad_norm": 0.15238098800182343, "learning_rate": 1.3876867032017126e-06, "loss": 0.0099, "step": 92860 }, { "grad_norm": 0.08977743238210678, "learning_rate": 1.383820906312472e-06, "loss": 0.008, "step": 92870 }, { "grad_norm": 0.07757830619812012, "learning_rate": 1.3799604260154752e-06, "loss": 0.0086, "step": 92880 }, { "grad_norm": 0.08446396887302399, "learning_rate": 1.376105262732913e-06, "loss": 0.0078, "step": 92890 }, { "grad_norm": 0.12291106581687927, "learning_rate": 1.37225541688637e-06, "loss": 0.0096, "step": 92900 }, { "grad_norm": 0.156741663813591, "learning_rate": 1.3684108888968594e-06, "loss": 0.0065, "step": 92910 }, { "grad_norm": 0.16205616295337677, "learning_rate": 1.364571679184823e-06, "loss": 0.0101, "step": 92920 }, { "grad_norm": 0.17123010754585266, "learning_rate": 1.3607377881701077e-06, "loss": 0.0075, "step": 92930 }, { "grad_norm": 0.15032175183296204, "learning_rate": 1.3569092162719787e-06, "loss": 0.0064, "step": 92940 }, { "grad_norm": 0.10012595355510712, "learning_rate": 1.353085963909123e-06, "loss": 0.0111, "step": 92950 }, { "grad_norm": 0.09669599682092667, "learning_rate": 1.3492680314996454e-06, "loss": 0.0091, "step": 92960 }, { "grad_norm": 0.09051172435283661, "learning_rate": 1.345455419461078e-06, "loss": 0.0073, "step": 92970 }, { "grad_norm": 0.1471940279006958, "learning_rate": 1.3416481282103431e-06, "loss": 0.0092, "step": 92980 }, { "grad_norm": 0.09498313069343567, "learning_rate": 1.337846158163819e-06, "loss": 0.0081, "step": 92990 }, { "grad_norm": 0.15783649682998657, "learning_rate": 1.334049509737273e-06, "loss": 0.0085, "step": 93000 }, { "grad_norm": 0.10763977468013763, "learning_rate": 1.3302581833459116e-06, "loss": 0.0112, "step": 93010 }, { "grad_norm": 0.21780386567115784, "learning_rate": 1.3264721794043366e-06, "loss": 0.0089, "step": 93020 }, { "grad_norm": 0.11079157888889313, "learning_rate": 1.3226914983265836e-06, "loss": 0.008, "step": 93030 }, { "grad_norm": 0.11648274213075638, "learning_rate": 1.3189161405260997e-06, "loss": 0.0098, "step": 93040 }, { "grad_norm": 0.12205924093723297, "learning_rate": 1.3151461064157656e-06, "loss": 0.0077, "step": 93050 }, { "grad_norm": 0.07456608861684799, "learning_rate": 1.3113813964078459e-06, "loss": 0.0092, "step": 93060 }, { "grad_norm": 0.07934287190437317, "learning_rate": 1.3076220109140614e-06, "loss": 0.0082, "step": 93070 }, { "grad_norm": 0.14637216925621033, "learning_rate": 1.3038679503455275e-06, "loss": 0.0069, "step": 93080 }, { "grad_norm": 0.08179270476102829, "learning_rate": 1.3001192151127773e-06, "loss": 0.0102, "step": 93090 }, { "grad_norm": 0.23213602602481842, "learning_rate": 1.2963758056257657e-06, "loss": 0.0088, "step": 93100 }, { "grad_norm": 0.17387229204177856, "learning_rate": 1.2926377222938768e-06, "loss": 0.0085, "step": 93110 }, { "grad_norm": 0.13687315583229065, "learning_rate": 1.2889049655259e-06, "loss": 0.009, "step": 93120 }, { "grad_norm": 0.31512200832366943, "learning_rate": 1.2851775357300312e-06, "loss": 0.0075, "step": 93130 }, { "grad_norm": 0.12074091285467148, "learning_rate": 1.281455433313916e-06, "loss": 0.0088, "step": 93140 }, { "grad_norm": 0.2149454653263092, "learning_rate": 1.2777386586845796e-06, "loss": 0.0089, "step": 93150 }, { "grad_norm": 0.18078960478305817, "learning_rate": 1.2740272122484964e-06, "loss": 0.0093, "step": 93160 }, { "grad_norm": 0.15247970819473267, "learning_rate": 1.2703210944115308e-06, "loss": 0.0088, "step": 93170 }, { "grad_norm": 0.13613362610340118, "learning_rate": 1.2666203055789916e-06, "loss": 0.009, "step": 93180 }, { "grad_norm": 0.19591261446475983, "learning_rate": 1.2629248461555832e-06, "loss": 0.0086, "step": 93190 }, { "grad_norm": 0.12473490834236145, "learning_rate": 1.2592347165454431e-06, "loss": 0.0084, "step": 93200 }, { "grad_norm": 0.14318187534809113, "learning_rate": 1.2555499171521102e-06, "loss": 0.0077, "step": 93210 }, { "grad_norm": 0.15723174810409546, "learning_rate": 1.251870448378556e-06, "loss": 0.0091, "step": 93220 }, { "grad_norm": 0.07511748373508453, "learning_rate": 1.2481963106271478e-06, "loss": 0.0072, "step": 93230 }, { "grad_norm": 0.18329958617687225, "learning_rate": 1.2445275042997084e-06, "loss": 0.0108, "step": 93240 }, { "grad_norm": 0.12156528979539871, "learning_rate": 1.2408640297974283e-06, "loss": 0.0091, "step": 93250 }, { "grad_norm": 0.12551134824752808, "learning_rate": 1.2372058875209535e-06, "loss": 0.0104, "step": 93260 }, { "grad_norm": 0.122447170317173, "learning_rate": 1.2335530778703252e-06, "loss": 0.0066, "step": 93270 }, { "grad_norm": 0.12271778285503387, "learning_rate": 1.2299056012450182e-06, "loss": 0.0086, "step": 93280 }, { "grad_norm": 0.09634154289960861, "learning_rate": 1.2262634580439026e-06, "loss": 0.0079, "step": 93290 }, { "grad_norm": 0.24935252964496613, "learning_rate": 1.2226266486652926e-06, "loss": 0.0091, "step": 93300 }, { "grad_norm": 0.12876473367214203, "learning_rate": 1.2189951735068927e-06, "loss": 0.0068, "step": 93310 }, { "grad_norm": 0.1120937168598175, "learning_rate": 1.2153690329658407e-06, "loss": 0.0111, "step": 93320 }, { "grad_norm": 0.14571698009967804, "learning_rate": 1.2117482274386859e-06, "loss": 0.0092, "step": 93330 }, { "grad_norm": 0.2637118101119995, "learning_rate": 1.2081327573213953e-06, "loss": 0.0088, "step": 93340 }, { "grad_norm": 0.10377084463834763, "learning_rate": 1.2045226230093465e-06, "loss": 0.0078, "step": 93350 }, { "grad_norm": 0.11379015445709229, "learning_rate": 1.2009178248973296e-06, "loss": 0.0075, "step": 93360 }, { "grad_norm": 0.10173329710960388, "learning_rate": 1.1973183633795849e-06, "loss": 0.0072, "step": 93370 }, { "grad_norm": 0.07564523816108704, "learning_rate": 1.193724238849725e-06, "loss": 0.0086, "step": 93380 }, { "grad_norm": 0.27240270376205444, "learning_rate": 1.190135451700808e-06, "loss": 0.0106, "step": 93390 }, { "grad_norm": 0.08533421903848648, "learning_rate": 1.186552002325292e-06, "loss": 0.0077, "step": 93400 }, { "grad_norm": 0.1877705603837967, "learning_rate": 1.1829738911150578e-06, "loss": 0.0092, "step": 93410 }, { "grad_norm": 0.08847260475158691, "learning_rate": 1.1794011184613985e-06, "loss": 0.0069, "step": 93420 }, { "grad_norm": 0.09649495780467987, "learning_rate": 1.17583368475504e-06, "loss": 0.0085, "step": 93430 }, { "grad_norm": 0.13260672986507416, "learning_rate": 1.1722715903860982e-06, "loss": 0.0065, "step": 93440 }, { "grad_norm": 0.08149269223213196, "learning_rate": 1.168714835744128e-06, "loss": 0.0099, "step": 93450 }, { "grad_norm": 0.09038787335157394, "learning_rate": 1.165163421218085e-06, "loss": 0.0085, "step": 93460 }, { "grad_norm": 0.1497722715139389, "learning_rate": 1.1616173471963477e-06, "loss": 0.0077, "step": 93470 }, { "grad_norm": 0.09244055300951004, "learning_rate": 1.1580766140667109e-06, "loss": 0.0075, "step": 93480 }, { "grad_norm": 0.09044281393289566, "learning_rate": 1.1545412222163764e-06, "loss": 0.0103, "step": 93490 }, { "grad_norm": 0.13839899003505707, "learning_rate": 1.1510111720319795e-06, "loss": 0.0085, "step": 93500 }, { "grad_norm": 0.09469525516033173, "learning_rate": 1.1474864638995663e-06, "loss": 0.0071, "step": 93510 }, { "grad_norm": 0.11537081003189087, "learning_rate": 1.1439670982045735e-06, "loss": 0.008, "step": 93520 }, { "grad_norm": 0.087933748960495, "learning_rate": 1.1404530753318877e-06, "loss": 0.0102, "step": 93530 }, { "grad_norm": 0.07531892508268356, "learning_rate": 1.136944395665801e-06, "loss": 0.008, "step": 93540 }, { "grad_norm": 0.15310925245285034, "learning_rate": 1.1334410595900014e-06, "loss": 0.0086, "step": 93550 }, { "grad_norm": 0.09601669758558273, "learning_rate": 1.129943067487621e-06, "loss": 0.0077, "step": 93560 }, { "grad_norm": 0.09058631211519241, "learning_rate": 1.126450419741193e-06, "loss": 0.0101, "step": 93570 }, { "grad_norm": 0.15394385159015656, "learning_rate": 1.1229631167326671e-06, "loss": 0.0072, "step": 93580 }, { "grad_norm": 0.11104775220155716, "learning_rate": 1.1194811588434052e-06, "loss": 0.0096, "step": 93590 }, { "grad_norm": 0.11963950842618942, "learning_rate": 1.1160045464542023e-06, "loss": 0.0093, "step": 93600 }, { "grad_norm": 0.1315855234861374, "learning_rate": 1.1125332799452325e-06, "loss": 0.0102, "step": 93610 }, { "grad_norm": 0.10528966784477234, "learning_rate": 1.109067359696131e-06, "loss": 0.0092, "step": 93620 }, { "grad_norm": 0.10397949069738388, "learning_rate": 1.1056067860859166e-06, "loss": 0.0079, "step": 93630 }, { "grad_norm": 0.08051587641239166, "learning_rate": 1.102151559493031e-06, "loss": 0.0094, "step": 93640 }, { "grad_norm": 0.1357600837945938, "learning_rate": 1.0987016802953332e-06, "loss": 0.0085, "step": 93650 }, { "grad_norm": 0.1295468956232071, "learning_rate": 1.0952571488701048e-06, "loss": 0.0069, "step": 93660 }, { "grad_norm": 0.12492217868566513, "learning_rate": 1.0918179655940163e-06, "loss": 0.0084, "step": 93670 }, { "grad_norm": 0.1438339203596115, "learning_rate": 1.088384130843184e-06, "loss": 0.0073, "step": 93680 }, { "grad_norm": 0.17979225516319275, "learning_rate": 1.0849556449931296e-06, "loss": 0.0078, "step": 93690 }, { "grad_norm": 0.299724817276001, "learning_rate": 1.0815325084187866e-06, "loss": 0.0077, "step": 93700 }, { "grad_norm": 0.09040693938732147, "learning_rate": 1.0781147214944943e-06, "loss": 0.0095, "step": 93710 }, { "grad_norm": 0.14945682883262634, "learning_rate": 1.0747022845940313e-06, "loss": 0.0082, "step": 93720 }, { "grad_norm": 0.11564341187477112, "learning_rate": 1.0712951980905605e-06, "loss": 0.0065, "step": 93730 }, { "grad_norm": 0.1937304437160492, "learning_rate": 1.0678934623566839e-06, "loss": 0.0078, "step": 93740 }, { "grad_norm": 0.12121636420488358, "learning_rate": 1.0644970777644091e-06, "loss": 0.0092, "step": 93750 }, { "grad_norm": 0.2016143798828125, "learning_rate": 1.0611060446851673e-06, "loss": 0.0083, "step": 93760 }, { "grad_norm": 0.08537254482507706, "learning_rate": 1.0577203634897836e-06, "loss": 0.0079, "step": 93770 }, { "grad_norm": 0.13818393647670746, "learning_rate": 1.0543400345485178e-06, "loss": 0.0067, "step": 93780 }, { "grad_norm": 0.08959449827671051, "learning_rate": 1.0509650582310404e-06, "loss": 0.0078, "step": 93790 }, { "grad_norm": 0.08038564026355743, "learning_rate": 1.0475954349064287e-06, "loss": 0.0091, "step": 93800 }, { "grad_norm": 0.15100069344043732, "learning_rate": 1.0442311649431825e-06, "loss": 0.0067, "step": 93810 }, { "grad_norm": 0.14225010573863983, "learning_rate": 1.0408722487092071e-06, "loss": 0.007, "step": 93820 }, { "grad_norm": 0.15338896214962006, "learning_rate": 1.0375186865718424e-06, "loss": 0.0068, "step": 93830 }, { "grad_norm": 0.12406016141176224, "learning_rate": 1.0341704788978113e-06, "loss": 0.0085, "step": 93840 }, { "grad_norm": 0.0859779492020607, "learning_rate": 1.0308276260532879e-06, "loss": 0.0077, "step": 93850 }, { "grad_norm": 0.15000620484352112, "learning_rate": 1.0274901284038241e-06, "loss": 0.0083, "step": 93860 }, { "grad_norm": 0.12976425886154175, "learning_rate": 1.0241579863144114e-06, "loss": 0.0102, "step": 93870 }, { "grad_norm": 0.14617609977722168, "learning_rate": 1.0208312001494413e-06, "loss": 0.0091, "step": 93880 }, { "grad_norm": 0.14685885608196259, "learning_rate": 1.0175097702727453e-06, "loss": 0.008, "step": 93890 }, { "grad_norm": 0.08955959230661392, "learning_rate": 1.0141936970475275e-06, "loss": 0.0087, "step": 93900 }, { "grad_norm": 0.08928173780441284, "learning_rate": 1.0108829808364417e-06, "loss": 0.01, "step": 93910 }, { "grad_norm": 0.11513475328683853, "learning_rate": 1.0075776220015376e-06, "loss": 0.0074, "step": 93920 }, { "grad_norm": 0.1302335113286972, "learning_rate": 1.0042776209042816e-06, "loss": 0.0076, "step": 93930 }, { "grad_norm": 0.12275123596191406, "learning_rate": 1.000982977905568e-06, "loss": 0.0076, "step": 93940 }, { "grad_norm": 0.12103215605020523, "learning_rate": 9.976936933656867e-07, "loss": 0.0089, "step": 93950 }, { "grad_norm": 0.1254590004682541, "learning_rate": 9.944097676443442e-07, "loss": 0.0117, "step": 93960 }, { "grad_norm": 0.12705586850643158, "learning_rate": 9.911312011006757e-07, "loss": 0.0078, "step": 93970 }, { "grad_norm": 0.12818849086761475, "learning_rate": 9.878579940932109e-07, "loss": 0.0073, "step": 93980 }, { "grad_norm": 0.10388746857643127, "learning_rate": 9.84590146979908e-07, "loss": 0.0062, "step": 93990 }, { "grad_norm": 0.09727974981069565, "learning_rate": 9.813276601181365e-07, "loss": 0.007, "step": 94000 }, { "grad_norm": 0.06510210782289505, "learning_rate": 9.780705338646724e-07, "loss": 0.0076, "step": 94010 }, { "grad_norm": 0.115467868745327, "learning_rate": 9.748187685757083e-07, "loss": 0.0087, "step": 94020 }, { "grad_norm": 0.10692823678255081, "learning_rate": 9.715723646068543e-07, "loss": 0.0082, "step": 94030 }, { "grad_norm": 0.10156796872615814, "learning_rate": 9.68331322313143e-07, "loss": 0.0074, "step": 94040 }, { "grad_norm": 0.1355384886264801, "learning_rate": 9.650956420489909e-07, "loss": 0.0084, "step": 94050 }, { "grad_norm": 0.17619463801383972, "learning_rate": 9.618653241682595e-07, "loss": 0.0078, "step": 94060 }, { "grad_norm": 0.21620111167430878, "learning_rate": 9.586403690242053e-07, "loss": 0.0114, "step": 94070 }, { "grad_norm": 0.14911462366580963, "learning_rate": 9.554207769695122e-07, "loss": 0.0085, "step": 94080 }, { "grad_norm": 0.15561683475971222, "learning_rate": 9.522065483562659e-07, "loss": 0.0088, "step": 94090 }, { "grad_norm": 0.12354275584220886, "learning_rate": 9.489976835359681e-07, "loss": 0.0083, "step": 94100 }, { "grad_norm": 0.24254797399044037, "learning_rate": 9.457941828595329e-07, "loss": 0.009, "step": 94110 }, { "grad_norm": 0.18085487186908722, "learning_rate": 9.425960466772965e-07, "loss": 0.0086, "step": 94120 }, { "grad_norm": 0.15871857106685638, "learning_rate": 9.394032753390014e-07, "loss": 0.0088, "step": 94130 }, { "grad_norm": 0.20992760360240936, "learning_rate": 9.362158691938017e-07, "loss": 0.0087, "step": 94140 }, { "grad_norm": 0.14143876731395721, "learning_rate": 9.330338285902685e-07, "loss": 0.0077, "step": 94150 }, { "grad_norm": 0.11878326535224915, "learning_rate": 9.298571538763845e-07, "loss": 0.0073, "step": 94160 }, { "grad_norm": 0.10691069811582565, "learning_rate": 9.266858453995441e-07, "loss": 0.0082, "step": 94170 }, { "grad_norm": 0.12571948766708374, "learning_rate": 9.235199035065645e-07, "loss": 0.0082, "step": 94180 }, { "grad_norm": 0.13057975471019745, "learning_rate": 9.203593285436574e-07, "loss": 0.0086, "step": 94190 }, { "grad_norm": 0.1434415876865387, "learning_rate": 9.172041208564686e-07, "loss": 0.0095, "step": 94200 }, { "grad_norm": 0.10733119398355484, "learning_rate": 9.140542807900332e-07, "loss": 0.0085, "step": 94210 }, { "grad_norm": 0.11597082763910294, "learning_rate": 9.109098086888313e-07, "loss": 0.0071, "step": 94220 }, { "grad_norm": 0.1485157310962677, "learning_rate": 9.077707048967265e-07, "loss": 0.0084, "step": 94230 }, { "grad_norm": 0.09352944046258926, "learning_rate": 9.046369697570057e-07, "loss": 0.0064, "step": 94240 }, { "grad_norm": 0.10528544336557388, "learning_rate": 9.015086036123721e-07, "loss": 0.0075, "step": 94250 }, { "grad_norm": 0.1138664111495018, "learning_rate": 8.983856068049413e-07, "loss": 0.0087, "step": 94260 }, { "grad_norm": 0.1393204927444458, "learning_rate": 8.952679796762398e-07, "loss": 0.0082, "step": 94270 }, { "grad_norm": 0.1679684966802597, "learning_rate": 8.921557225671951e-07, "loss": 0.0081, "step": 94280 }, { "grad_norm": 0.08165140450000763, "learning_rate": 8.890488358181792e-07, "loss": 0.0092, "step": 94290 }, { "grad_norm": 0.08289731293916702, "learning_rate": 8.859473197689372e-07, "loss": 0.0102, "step": 94300 }, { "grad_norm": 0.10799183696508408, "learning_rate": 8.828511747586588e-07, "loss": 0.0086, "step": 94310 }, { "grad_norm": 0.07396281510591507, "learning_rate": 8.797604011259287e-07, "loss": 0.0082, "step": 94320 }, { "grad_norm": 0.1191043108701706, "learning_rate": 8.766749992087487e-07, "loss": 0.0078, "step": 94330 }, { "grad_norm": 0.09105324000120163, "learning_rate": 8.735949693445378e-07, "loss": 0.0068, "step": 94340 }, { "grad_norm": 0.1291993260383606, "learning_rate": 8.705203118701211e-07, "loss": 0.0102, "step": 94350 }, { "grad_norm": 0.11583790928125381, "learning_rate": 8.674510271217295e-07, "loss": 0.0088, "step": 94360 }, { "grad_norm": 0.09892355650663376, "learning_rate": 8.643871154350336e-07, "loss": 0.0073, "step": 94370 }, { "grad_norm": 0.12428414821624756, "learning_rate": 8.613285771450818e-07, "loss": 0.0073, "step": 94380 }, { "grad_norm": 0.14434754848480225, "learning_rate": 8.582754125863623e-07, "loss": 0.0083, "step": 94390 }, { "grad_norm": 0.13550813496112823, "learning_rate": 8.55227622092758e-07, "loss": 0.0111, "step": 94400 }, { "grad_norm": 0.1706879884004593, "learning_rate": 8.521852059975743e-07, "loss": 0.0098, "step": 94410 }, { "grad_norm": 0.11649162322282791, "learning_rate": 8.491481646335231e-07, "loss": 0.0078, "step": 94420 }, { "grad_norm": 0.10549430549144745, "learning_rate": 8.46116498332733e-07, "loss": 0.0087, "step": 94430 }, { "grad_norm": 0.1126561313867569, "learning_rate": 8.430902074267333e-07, "loss": 0.008, "step": 94440 }, { "grad_norm": 0.1258831024169922, "learning_rate": 8.400692922464925e-07, "loss": 0.0075, "step": 94450 }, { "grad_norm": 0.3274737298488617, "learning_rate": 8.370537531223577e-07, "loss": 0.0099, "step": 94460 }, { "grad_norm": 0.1443902999162674, "learning_rate": 8.340435903841092e-07, "loss": 0.0085, "step": 94470 }, { "grad_norm": 0.1569238007068634, "learning_rate": 8.310388043609341e-07, "loss": 0.0083, "step": 94480 }, { "grad_norm": 0.138786181807518, "learning_rate": 8.280393953814303e-07, "loss": 0.0091, "step": 94490 }, { "grad_norm": 0.10914846509695053, "learning_rate": 8.250453637736077e-07, "loss": 0.0074, "step": 94500 }, { "grad_norm": 0.15587075054645538, "learning_rate": 8.22056709864899e-07, "loss": 0.0104, "step": 94510 }, { "grad_norm": 0.13927637040615082, "learning_rate": 8.190734339821204e-07, "loss": 0.0059, "step": 94520 }, { "grad_norm": 0.22233064472675323, "learning_rate": 8.160955364515333e-07, "loss": 0.0085, "step": 94530 }, { "grad_norm": 0.17977015674114227, "learning_rate": 8.131230175987936e-07, "loss": 0.0109, "step": 94540 }, { "grad_norm": 0.08372090756893158, "learning_rate": 8.101558777489693e-07, "loss": 0.0068, "step": 94550 }, { "grad_norm": 0.11391541361808777, "learning_rate": 8.071941172265396e-07, "loss": 0.0083, "step": 94560 }, { "grad_norm": 0.13322551548480988, "learning_rate": 8.042377363553955e-07, "loss": 0.0086, "step": 94570 }, { "grad_norm": 0.0718279555439949, "learning_rate": 8.01286735458856e-07, "loss": 0.0096, "step": 94580 }, { "grad_norm": 0.08137853443622589, "learning_rate": 7.983411148596298e-07, "loss": 0.0066, "step": 94590 }, { "grad_norm": 0.11946417391300201, "learning_rate": 7.954008748798425e-07, "loss": 0.0069, "step": 94600 }, { "grad_norm": 0.1458108276128769, "learning_rate": 7.924660158410369e-07, "loss": 0.0076, "step": 94610 }, { "grad_norm": 0.0833212211728096, "learning_rate": 7.895365380641728e-07, "loss": 0.007, "step": 94620 }, { "grad_norm": 0.15951639413833618, "learning_rate": 7.866124418695941e-07, "loss": 0.009, "step": 94630 }, { "grad_norm": 0.10927926748991013, "learning_rate": 7.83693727577095e-07, "loss": 0.0081, "step": 94640 }, { "grad_norm": 0.13715748488903046, "learning_rate": 7.807803955058535e-07, "loss": 0.0075, "step": 94650 }, { "grad_norm": 0.16074970364570618, "learning_rate": 7.778724459744702e-07, "loss": 0.0076, "step": 94660 }, { "grad_norm": 0.09746603667736053, "learning_rate": 7.749698793009464e-07, "loss": 0.0082, "step": 94670 }, { "grad_norm": 0.14324036240577698, "learning_rate": 7.720726958027169e-07, "loss": 0.0077, "step": 94680 }, { "grad_norm": 0.196054145693779, "learning_rate": 7.691808957965951e-07, "loss": 0.0064, "step": 94690 }, { "grad_norm": 0.16338874399662018, "learning_rate": 7.662944795988336e-07, "loss": 0.0141, "step": 94700 }, { "grad_norm": 0.12463036179542542, "learning_rate": 7.634134475250854e-07, "loss": 0.0085, "step": 94710 }, { "grad_norm": 0.13110527396202087, "learning_rate": 7.605377998904262e-07, "loss": 0.0114, "step": 94720 }, { "grad_norm": 0.1243276298046112, "learning_rate": 7.5766753700931e-07, "loss": 0.0096, "step": 94730 }, { "grad_norm": 0.12089856714010239, "learning_rate": 7.54802659195647e-07, "loss": 0.0077, "step": 94740 }, { "grad_norm": 0.07528236508369446, "learning_rate": 7.519431667627253e-07, "loss": 0.0088, "step": 94750 }, { "grad_norm": 0.1794126033782959, "learning_rate": 7.490890600232503e-07, "loss": 0.0063, "step": 94760 }, { "grad_norm": 0.11512263119220734, "learning_rate": 7.462403392893502e-07, "loss": 0.0063, "step": 94770 }, { "grad_norm": 0.19807764887809753, "learning_rate": 7.433970048725536e-07, "loss": 0.0109, "step": 94780 }, { "grad_norm": 0.0910792127251625, "learning_rate": 7.405590570838062e-07, "loss": 0.0073, "step": 94790 }, { "grad_norm": 0.12874478101730347, "learning_rate": 7.377264962334596e-07, "loss": 0.008, "step": 94800 }, { "grad_norm": 0.08664782345294952, "learning_rate": 7.348993226312773e-07, "loss": 0.0074, "step": 94810 }, { "grad_norm": 0.15123052895069122, "learning_rate": 7.32077536586434e-07, "loss": 0.0114, "step": 94820 }, { "grad_norm": 0.13246652483940125, "learning_rate": 7.292611384075221e-07, "loss": 0.0074, "step": 94830 }, { "grad_norm": 0.06510455161333084, "learning_rate": 7.264501284025338e-07, "loss": 0.0084, "step": 94840 }, { "grad_norm": 0.18327739834785461, "learning_rate": 7.236445068788789e-07, "loss": 0.008, "step": 94850 }, { "grad_norm": 0.07563702017068863, "learning_rate": 7.20844274143373e-07, "loss": 0.0091, "step": 94860 }, { "grad_norm": 0.1369256228208542, "learning_rate": 7.180494305022489e-07, "loss": 0.0066, "step": 94870 }, { "grad_norm": 0.1385524868965149, "learning_rate": 7.152599762611456e-07, "loss": 0.0096, "step": 94880 }, { "grad_norm": 0.08664664626121521, "learning_rate": 7.124759117251079e-07, "loss": 0.0078, "step": 94890 }, { "grad_norm": 0.06302866339683533, "learning_rate": 7.096972371986033e-07, "loss": 0.0085, "step": 94900 }, { "grad_norm": 0.14822658896446228, "learning_rate": 7.069239529855054e-07, "loss": 0.0091, "step": 94910 }, { "grad_norm": 0.17266234755516052, "learning_rate": 7.041560593890939e-07, "loss": 0.0105, "step": 94920 }, { "grad_norm": 0.09403964132070541, "learning_rate": 7.013935567120599e-07, "loss": 0.0069, "step": 94930 }, { "grad_norm": 0.12587732076644897, "learning_rate": 6.986364452565009e-07, "loss": 0.007, "step": 94940 }, { "grad_norm": 0.26534411311149597, "learning_rate": 6.95884725323942e-07, "loss": 0.0085, "step": 94950 }, { "grad_norm": 0.23025600612163544, "learning_rate": 6.931383972152982e-07, "loss": 0.0084, "step": 94960 }, { "grad_norm": 0.07977578788995743, "learning_rate": 6.903974612309127e-07, "loss": 0.0072, "step": 94970 }, { "grad_norm": 0.1145758330821991, "learning_rate": 6.876619176705179e-07, "loss": 0.007, "step": 94980 }, { "grad_norm": 0.13027210533618927, "learning_rate": 6.8493176683328e-07, "loss": 0.007, "step": 94990 }, { "grad_norm": 0.09168915450572968, "learning_rate": 6.822070090177657e-07, "loss": 0.0067, "step": 95000 }, { "grad_norm": 0.11178918182849884, "learning_rate": 6.794876445219367e-07, "loss": 0.0075, "step": 95010 }, { "grad_norm": 0.1493862122297287, "learning_rate": 6.767736736431885e-07, "loss": 0.0091, "step": 95020 }, { "grad_norm": 0.14364050328731537, "learning_rate": 6.740650966783113e-07, "loss": 0.007, "step": 95030 }, { "grad_norm": 0.10736026614904404, "learning_rate": 6.713619139235239e-07, "loss": 0.0069, "step": 95040 }, { "grad_norm": 0.11092667281627655, "learning_rate": 6.686641256744286e-07, "loss": 0.0067, "step": 95050 }, { "grad_norm": 0.13271547853946686, "learning_rate": 6.659717322260562e-07, "loss": 0.0079, "step": 95060 }, { "grad_norm": 0.2569698989391327, "learning_rate": 6.632847338728377e-07, "loss": 0.0103, "step": 95070 }, { "grad_norm": 0.1737470030784607, "learning_rate": 6.60603130908627e-07, "loss": 0.0076, "step": 95080 }, { "grad_norm": 0.11981336772441864, "learning_rate": 6.579269236266783e-07, "loss": 0.0088, "step": 95090 }, { "grad_norm": 0.12533994019031525, "learning_rate": 6.552561123196632e-07, "loss": 0.0067, "step": 95100 }, { "grad_norm": 0.1737206131219864, "learning_rate": 6.525906972796425e-07, "loss": 0.0087, "step": 95110 }, { "grad_norm": 0.14836300909519196, "learning_rate": 6.499306787981108e-07, "loss": 0.0067, "step": 95120 }, { "grad_norm": 0.13723067939281464, "learning_rate": 6.472760571659631e-07, "loss": 0.0101, "step": 95130 }, { "grad_norm": 0.08917161822319031, "learning_rate": 6.446268326735061e-07, "loss": 0.0084, "step": 95140 }, { "grad_norm": 0.10338863730430603, "learning_rate": 6.41983005610458e-07, "loss": 0.0078, "step": 95150 }, { "grad_norm": 0.13892708718776703, "learning_rate": 6.393445762659378e-07, "loss": 0.0102, "step": 95160 }, { "grad_norm": 0.1178417056798935, "learning_rate": 6.36711544928481e-07, "loss": 0.0077, "step": 95170 }, { "grad_norm": 0.15376907587051392, "learning_rate": 6.340839118860298e-07, "loss": 0.008, "step": 95180 }, { "grad_norm": 0.17796552181243896, "learning_rate": 6.314616774259485e-07, "loss": 0.0071, "step": 95190 }, { "grad_norm": 0.13909421861171722, "learning_rate": 6.288448418349858e-07, "loss": 0.0075, "step": 95200 }, { "grad_norm": 0.13872918486595154, "learning_rate": 6.262334053993235e-07, "loss": 0.013, "step": 95210 }, { "grad_norm": 0.14902208745479584, "learning_rate": 6.236273684045501e-07, "loss": 0.0091, "step": 95220 }, { "grad_norm": 0.1198665052652359, "learning_rate": 6.210267311356488e-07, "loss": 0.0082, "step": 95230 }, { "grad_norm": 0.1290035992860794, "learning_rate": 6.184314938770197e-07, "loss": 0.0076, "step": 95240 }, { "grad_norm": 0.09914962202310562, "learning_rate": 6.158416569124858e-07, "loss": 0.007, "step": 95250 }, { "grad_norm": 0.20408600568771362, "learning_rate": 6.132572205252541e-07, "loss": 0.0075, "step": 95260 }, { "grad_norm": 0.19758029282093048, "learning_rate": 6.106781849979648e-07, "loss": 0.0087, "step": 95270 }, { "grad_norm": 0.12192526459693909, "learning_rate": 6.081045506126482e-07, "loss": 0.007, "step": 95280 }, { "grad_norm": 0.12656407058238983, "learning_rate": 6.055363176507678e-07, "loss": 0.0074, "step": 95290 }, { "grad_norm": 0.1153714582324028, "learning_rate": 6.029734863931602e-07, "loss": 0.0087, "step": 95300 }, { "grad_norm": 0.09038281440734863, "learning_rate": 6.004160571201179e-07, "loss": 0.0069, "step": 95310 }, { "grad_norm": 0.1345767378807068, "learning_rate": 5.978640301112947e-07, "loss": 0.0064, "step": 95320 }, { "grad_norm": 0.1355891078710556, "learning_rate": 5.9531740564579e-07, "loss": 0.0074, "step": 95330 }, { "grad_norm": 0.09131451696157455, "learning_rate": 5.927761840020918e-07, "loss": 0.0065, "step": 95340 }, { "grad_norm": 0.16647636890411377, "learning_rate": 5.902403654581113e-07, "loss": 0.0085, "step": 95350 }, { "grad_norm": 0.20062355697155, "learning_rate": 5.877099502911542e-07, "loss": 0.0073, "step": 95360 }, { "grad_norm": 0.10651080310344696, "learning_rate": 5.851849387779496e-07, "loss": 0.0067, "step": 95370 }, { "grad_norm": 0.1065891906619072, "learning_rate": 5.826653311946262e-07, "loss": 0.0098, "step": 95380 }, { "grad_norm": 0.091328926384449, "learning_rate": 5.801511278167249e-07, "loss": 0.0076, "step": 95390 }, { "grad_norm": 0.11177445203065872, "learning_rate": 5.776423289191868e-07, "loss": 0.0087, "step": 95400 }, { "grad_norm": 0.13846516609191895, "learning_rate": 5.751389347763869e-07, "loss": 0.0078, "step": 95410 }, { "grad_norm": 0.11547403782606125, "learning_rate": 5.726409456620841e-07, "loss": 0.0075, "step": 95420 }, { "grad_norm": 0.16882731020450592, "learning_rate": 5.701483618494541e-07, "loss": 0.0083, "step": 95430 }, { "grad_norm": 0.1642693132162094, "learning_rate": 5.67661183611079e-07, "loss": 0.0066, "step": 95440 }, { "grad_norm": 0.10115291178226471, "learning_rate": 5.651794112189579e-07, "loss": 0.0067, "step": 95450 }, { "grad_norm": 0.10665599256753922, "learning_rate": 5.627030449444903e-07, "loss": 0.0075, "step": 95460 }, { "grad_norm": 0.1506529301404953, "learning_rate": 5.60232085058493e-07, "loss": 0.009, "step": 95470 }, { "grad_norm": 0.11518316715955734, "learning_rate": 5.577665318311776e-07, "loss": 0.0079, "step": 95480 }, { "grad_norm": 0.08313377946615219, "learning_rate": 5.553063855321838e-07, "loss": 0.0066, "step": 95490 }, { "grad_norm": 0.10109417885541916, "learning_rate": 5.528516464305467e-07, "loss": 0.0122, "step": 95500 }, { "grad_norm": 0.098300039768219, "learning_rate": 5.504023147947068e-07, "loss": 0.007, "step": 95510 }, { "grad_norm": 0.1266084760427475, "learning_rate": 5.479583908925279e-07, "loss": 0.0094, "step": 95520 }, { "grad_norm": 0.15001273155212402, "learning_rate": 5.455198749912626e-07, "loss": 0.0087, "step": 95530 }, { "grad_norm": 0.13802166283130646, "learning_rate": 5.430867673575979e-07, "loss": 0.0078, "step": 95540 }, { "grad_norm": 0.23598721623420715, "learning_rate": 5.406590682575985e-07, "loss": 0.0068, "step": 95550 }, { "grad_norm": 0.12547780573368073, "learning_rate": 5.382367779567687e-07, "loss": 0.0076, "step": 95560 }, { "grad_norm": 0.19344617426395416, "learning_rate": 5.358198967200023e-07, "loss": 0.0078, "step": 95570 }, { "grad_norm": 0.12332101911306381, "learning_rate": 5.334084248115989e-07, "loss": 0.0071, "step": 95580 }, { "grad_norm": 0.12411713600158691, "learning_rate": 5.310023624952754e-07, "loss": 0.0102, "step": 95590 }, { "grad_norm": 0.15286052227020264, "learning_rate": 5.286017100341656e-07, "loss": 0.0063, "step": 95600 }, { "grad_norm": 0.16455857455730438, "learning_rate": 5.262064676907874e-07, "loss": 0.0067, "step": 95610 }, { "grad_norm": 0.14479036629199982, "learning_rate": 5.238166357270924e-07, "loss": 0.0087, "step": 95620 }, { "grad_norm": 0.11567308753728867, "learning_rate": 5.214322144044215e-07, "loss": 0.0068, "step": 95630 }, { "grad_norm": 0.17320632934570312, "learning_rate": 5.190532039835327e-07, "loss": 0.0104, "step": 95640 }, { "grad_norm": 0.15348011255264282, "learning_rate": 5.166796047245903e-07, "loss": 0.0098, "step": 95650 }, { "grad_norm": 0.16825102269649506, "learning_rate": 5.143114168871699e-07, "loss": 0.0082, "step": 95660 }, { "grad_norm": 0.061990559101104736, "learning_rate": 5.119486407302532e-07, "loss": 0.0057, "step": 95670 }, { "grad_norm": 0.12404259294271469, "learning_rate": 5.095912765122335e-07, "loss": 0.007, "step": 95680 }, { "grad_norm": 0.18838420510292053, "learning_rate": 5.07239324490899e-07, "loss": 0.0076, "step": 95690 }, { "grad_norm": 0.12709824740886688, "learning_rate": 5.048927849234608e-07, "loss": 0.0074, "step": 95700 }, { "grad_norm": 0.12428510189056396, "learning_rate": 5.025516580665358e-07, "loss": 0.0086, "step": 95710 }, { "grad_norm": 0.14160966873168945, "learning_rate": 5.002159441761412e-07, "loss": 0.0077, "step": 95720 }, { "grad_norm": 0.13685071468353271, "learning_rate": 4.978856435077062e-07, "loss": 0.0092, "step": 95730 }, { "grad_norm": 0.16910427808761597, "learning_rate": 4.955607563160769e-07, "loss": 0.0082, "step": 95740 }, { "grad_norm": 0.14499349892139435, "learning_rate": 4.932412828554944e-07, "loss": 0.0075, "step": 95750 }, { "grad_norm": 0.17680305242538452, "learning_rate": 4.909272233796114e-07, "loss": 0.0105, "step": 95760 }, { "grad_norm": 0.12974153459072113, "learning_rate": 4.88618578141492e-07, "loss": 0.0088, "step": 95770 }, { "grad_norm": 0.1626695692539215, "learning_rate": 4.863153473936011e-07, "loss": 0.0076, "step": 95780 }, { "grad_norm": 0.14132480323314667, "learning_rate": 4.840175313878259e-07, "loss": 0.0079, "step": 95790 }, { "grad_norm": 0.12050449103116989, "learning_rate": 4.817251303754433e-07, "loss": 0.0068, "step": 95800 }, { "grad_norm": 0.08277446031570435, "learning_rate": 4.794381446071583e-07, "loss": 0.0082, "step": 95810 }, { "grad_norm": 0.08589024096727371, "learning_rate": 4.771565743330597e-07, "loss": 0.0069, "step": 95820 }, { "grad_norm": 0.1635604053735733, "learning_rate": 4.7488041980266463e-07, "loss": 0.0078, "step": 95830 }, { "grad_norm": 0.20791399478912354, "learning_rate": 4.726096812648795e-07, "loss": 0.0097, "step": 95840 }, { "grad_norm": 0.15101608633995056, "learning_rate": 4.703443589680445e-07, "loss": 0.0079, "step": 95850 }, { "grad_norm": 0.15423236787319183, "learning_rate": 4.6808445315987826e-07, "loss": 0.0106, "step": 95860 }, { "grad_norm": 0.12996110320091248, "learning_rate": 4.658299640875274e-07, "loss": 0.0083, "step": 95870 }, { "grad_norm": 0.07075797766447067, "learning_rate": 4.635808919975393e-07, "loss": 0.0078, "step": 95880 }, { "grad_norm": 0.11759471893310547, "learning_rate": 4.6133723713586706e-07, "loss": 0.0061, "step": 95890 }, { "grad_norm": 0.1257716715335846, "learning_rate": 4.590989997478756e-07, "loss": 0.0085, "step": 95900 }, { "grad_norm": 0.1645462065935135, "learning_rate": 4.5686618007833027e-07, "loss": 0.0089, "step": 95910 }, { "grad_norm": 0.1822221279144287, "learning_rate": 4.54638778371419e-07, "loss": 0.0081, "step": 95920 }, { "grad_norm": 0.26259133219718933, "learning_rate": 4.524167948707192e-07, "loss": 0.012, "step": 95930 }, { "grad_norm": 0.1446375697851181, "learning_rate": 4.5020022981921986e-07, "loss": 0.0087, "step": 95940 }, { "grad_norm": 0.09986034035682678, "learning_rate": 4.47989083459327e-07, "loss": 0.0099, "step": 95950 }, { "grad_norm": 0.08233893662691116, "learning_rate": 4.457833560328528e-07, "loss": 0.0079, "step": 95960 }, { "grad_norm": 0.15550100803375244, "learning_rate": 4.4358304778100434e-07, "loss": 0.0072, "step": 95970 }, { "grad_norm": 0.22064635157585144, "learning_rate": 4.413881589444058e-07, "loss": 0.0106, "step": 95980 }, { "grad_norm": 0.14187024533748627, "learning_rate": 4.391986897630873e-07, "loss": 0.0093, "step": 95990 }, { "grad_norm": 0.09904038906097412, "learning_rate": 4.370146404764908e-07, "loss": 0.0086, "step": 96000 }, { "grad_norm": 0.19797898828983307, "learning_rate": 4.3483601132345285e-07, "loss": 0.0081, "step": 96010 }, { "grad_norm": 0.14506538212299347, "learning_rate": 4.32662802542233e-07, "loss": 0.0059, "step": 96020 }, { "grad_norm": 0.10598649084568024, "learning_rate": 4.3049501437047444e-07, "loss": 0.0083, "step": 96030 }, { "grad_norm": 0.18077285587787628, "learning_rate": 4.283326470452653e-07, "loss": 0.008, "step": 96040 }, { "grad_norm": 0.18671707808971405, "learning_rate": 4.2617570080306647e-07, "loss": 0.0077, "step": 96050 }, { "grad_norm": 0.07440707832574844, "learning_rate": 4.2402417587975585e-07, "loss": 0.0077, "step": 96060 }, { "grad_norm": 0.1081613302230835, "learning_rate": 4.218780725106286e-07, "loss": 0.0073, "step": 96070 }, { "grad_norm": 0.13573876023292542, "learning_rate": 4.197373909303748e-07, "loss": 0.0094, "step": 96080 }, { "grad_norm": 0.17146234214305878, "learning_rate": 4.17602131373096e-07, "loss": 0.0088, "step": 96090 }, { "grad_norm": 0.12554100155830383, "learning_rate": 4.154722940723055e-07, "loss": 0.0076, "step": 96100 }, { "grad_norm": 0.09082918614149094, "learning_rate": 4.1334787926091136e-07, "loss": 0.0084, "step": 96110 }, { "grad_norm": 0.17306479811668396, "learning_rate": 4.1122888717124444e-07, "loss": 0.0099, "step": 96120 }, { "grad_norm": 0.09419456869363785, "learning_rate": 4.091153180350249e-07, "loss": 0.0072, "step": 96130 }, { "grad_norm": 0.25916147232055664, "learning_rate": 4.0700717208340125e-07, "loss": 0.0107, "step": 96140 }, { "grad_norm": 0.1291879266500473, "learning_rate": 4.049044495469112e-07, "loss": 0.0071, "step": 96150 }, { "grad_norm": 0.07508177310228348, "learning_rate": 4.0280715065549866e-07, "loss": 0.0063, "step": 96160 }, { "grad_norm": 0.12779513001441956, "learning_rate": 4.0071527563853575e-07, "loss": 0.0084, "step": 96170 }, { "grad_norm": 0.08322439342737198, "learning_rate": 3.9862882472477823e-07, "loss": 0.0073, "step": 96180 }, { "grad_norm": 0.13102398812770844, "learning_rate": 3.9654779814239375e-07, "loss": 0.0081, "step": 96190 }, { "grad_norm": 0.10560956597328186, "learning_rate": 3.944721961189668e-07, "loss": 0.011, "step": 96200 }, { "grad_norm": 0.08983977884054184, "learning_rate": 3.9240201888147697e-07, "loss": 0.0074, "step": 96210 }, { "grad_norm": 0.2091570645570755, "learning_rate": 3.9033726665632096e-07, "loss": 0.0105, "step": 96220 }, { "grad_norm": 0.09087228775024414, "learning_rate": 3.882779396692959e-07, "loss": 0.008, "step": 96230 }, { "grad_norm": 0.10774694383144379, "learning_rate": 3.8622403814559947e-07, "loss": 0.0072, "step": 96240 }, { "grad_norm": 0.21351566910743713, "learning_rate": 3.841755623098575e-07, "loss": 0.011, "step": 96250 }, { "grad_norm": 0.26338356733322144, "learning_rate": 3.821325123860742e-07, "loss": 0.0082, "step": 96260 }, { "grad_norm": 0.1711747646331787, "learning_rate": 3.800948885976818e-07, "loss": 0.0089, "step": 96270 }, { "grad_norm": 0.09892908483743668, "learning_rate": 3.780626911675078e-07, "loss": 0.0084, "step": 96280 }, { "grad_norm": 0.07335267961025238, "learning_rate": 3.76035920317791e-07, "loss": 0.0099, "step": 96290 }, { "grad_norm": 0.12505196034908295, "learning_rate": 3.740145762701819e-07, "loss": 0.0072, "step": 96300 }, { "grad_norm": 0.11705637723207474, "learning_rate": 3.7199865924572585e-07, "loss": 0.0068, "step": 96310 }, { "grad_norm": 0.09854865819215775, "learning_rate": 3.699881694648799e-07, "loss": 0.0071, "step": 96320 }, { "grad_norm": 0.15514777600765228, "learning_rate": 3.679831071475126e-07, "loss": 0.0082, "step": 96330 }, { "grad_norm": 0.11281032115221024, "learning_rate": 3.65983472512893e-07, "loss": 0.0073, "step": 96340 }, { "grad_norm": 0.1292618364095688, "learning_rate": 3.6398926577969616e-07, "loss": 0.0069, "step": 96350 }, { "grad_norm": 0.11717595905065536, "learning_rate": 3.620004871660032e-07, "loss": 0.0085, "step": 96360 }, { "grad_norm": 0.08032716810703278, "learning_rate": 3.600171368893124e-07, "loss": 0.0067, "step": 96370 }, { "grad_norm": 0.15528516471385956, "learning_rate": 3.580392151665113e-07, "loss": 0.0073, "step": 96380 }, { "grad_norm": 0.10983465611934662, "learning_rate": 3.5606672221391024e-07, "loss": 0.0087, "step": 96390 }, { "grad_norm": 0.19504553079605103, "learning_rate": 3.540996582472089e-07, "loss": 0.0073, "step": 96400 }, { "grad_norm": 0.15572021901607513, "learning_rate": 3.521380234815297e-07, "loss": 0.0082, "step": 96410 }, { "grad_norm": 0.12968139350414276, "learning_rate": 3.5018181813138985e-07, "loss": 0.0072, "step": 96420 }, { "grad_norm": 0.09853977710008621, "learning_rate": 3.482310424107238e-07, "loss": 0.0073, "step": 96430 }, { "grad_norm": 0.09721986204385757, "learning_rate": 3.4628569653285537e-07, "loss": 0.0075, "step": 96440 }, { "grad_norm": 0.13914023339748383, "learning_rate": 3.44345780710531e-07, "loss": 0.0099, "step": 96450 }, { "grad_norm": 0.15861447155475616, "learning_rate": 3.424112951558978e-07, "loss": 0.0088, "step": 96460 }, { "grad_norm": 0.12138117104768753, "learning_rate": 3.40482240080503e-07, "loss": 0.008, "step": 96470 }, { "grad_norm": 0.13265059888362885, "learning_rate": 3.3855861569531133e-07, "loss": 0.0098, "step": 96480 }, { "grad_norm": 0.16425395011901855, "learning_rate": 3.3664042221068226e-07, "loss": 0.0083, "step": 96490 }, { "grad_norm": 0.0910363420844078, "learning_rate": 3.347276598363924e-07, "loss": 0.0095, "step": 96500 }, { "grad_norm": 0.0813172310590744, "learning_rate": 3.328203287816134e-07, "loss": 0.0087, "step": 96510 }, { "grad_norm": 0.08447765558958054, "learning_rate": 3.3091842925493386e-07, "loss": 0.0068, "step": 96520 }, { "grad_norm": 0.16855202615261078, "learning_rate": 3.2902196146433196e-07, "loss": 0.0085, "step": 96530 }, { "grad_norm": 0.10505085438489914, "learning_rate": 3.2713092561720836e-07, "loss": 0.0084, "step": 96540 }, { "grad_norm": 0.1049347072839737, "learning_rate": 3.2524532192036996e-07, "loss": 0.0085, "step": 96550 }, { "grad_norm": 0.4836167097091675, "learning_rate": 3.2336515058001836e-07, "loss": 0.0079, "step": 96560 }, { "grad_norm": 0.18838515877723694, "learning_rate": 3.2149041180176143e-07, "loss": 0.0079, "step": 96570 }, { "grad_norm": 0.16514769196510315, "learning_rate": 3.1962110579062954e-07, "loss": 0.0088, "step": 96580 }, { "grad_norm": 0.11998908966779709, "learning_rate": 3.177572327510314e-07, "loss": 0.0088, "step": 96590 }, { "grad_norm": 0.16683702170848846, "learning_rate": 3.158987928868151e-07, "loss": 0.0076, "step": 96600 }, { "grad_norm": 0.09941935539245605, "learning_rate": 3.1404578640120144e-07, "loss": 0.0078, "step": 96610 }, { "grad_norm": 0.18839521706104279, "learning_rate": 3.1219821349683933e-07, "loss": 0.0103, "step": 96620 }, { "grad_norm": 0.16566425561904907, "learning_rate": 3.103560743757783e-07, "loss": 0.0083, "step": 96630 }, { "grad_norm": 0.10374637693166733, "learning_rate": 3.0851936923946834e-07, "loss": 0.0085, "step": 96640 }, { "grad_norm": 0.11789827793836594, "learning_rate": 3.06688098288771e-07, "loss": 0.0072, "step": 96650 }, { "grad_norm": 0.1343001276254654, "learning_rate": 3.0486226172394825e-07, "loss": 0.0062, "step": 96660 }, { "grad_norm": 0.11473222076892853, "learning_rate": 3.030418597446738e-07, "loss": 0.008, "step": 96670 }, { "grad_norm": 0.12241344153881073, "learning_rate": 3.012268925500272e-07, "loss": 0.0076, "step": 96680 }, { "grad_norm": 0.11906374245882034, "learning_rate": 2.994173603384831e-07, "loss": 0.0091, "step": 96690 }, { "grad_norm": 0.09870199114084244, "learning_rate": 2.9761326330793317e-07, "loss": 0.0067, "step": 96700 }, { "grad_norm": 0.1462913155555725, "learning_rate": 2.958146016556751e-07, "loss": 0.0096, "step": 96710 }, { "grad_norm": 0.09733901917934418, "learning_rate": 2.94021375578396e-07, "loss": 0.0095, "step": 96720 }, { "grad_norm": 0.08742779493331909, "learning_rate": 2.9223358527221135e-07, "loss": 0.0098, "step": 96730 }, { "grad_norm": 0.08999242633581161, "learning_rate": 2.904512309326257e-07, "loss": 0.0083, "step": 96740 }, { "grad_norm": 0.1319592297077179, "learning_rate": 2.8867431275455545e-07, "loss": 0.0082, "step": 96750 }, { "grad_norm": 0.1843651384115219, "learning_rate": 2.86902830932323e-07, "loss": 0.0099, "step": 96760 }, { "grad_norm": 0.17049048840999603, "learning_rate": 2.851367856596565e-07, "loss": 0.0085, "step": 96770 }, { "grad_norm": 0.12520122528076172, "learning_rate": 2.83376177129685e-07, "loss": 0.0065, "step": 96780 }, { "grad_norm": 0.10214308649301529, "learning_rate": 2.8162100553494884e-07, "loss": 0.0069, "step": 96790 }, { "grad_norm": 0.1217527911067009, "learning_rate": 2.798712710673834e-07, "loss": 0.0098, "step": 96800 }, { "grad_norm": 0.07130039483308792, "learning_rate": 2.7812697391834676e-07, "loss": 0.0091, "step": 96810 }, { "grad_norm": 0.08060868084430695, "learning_rate": 2.7638811427858627e-07, "loss": 0.0069, "step": 96820 }, { "grad_norm": 0.11160045862197876, "learning_rate": 2.746546923382665e-07, "loss": 0.0075, "step": 96830 }, { "grad_norm": 0.08810095489025116, "learning_rate": 2.729267082869469e-07, "loss": 0.0066, "step": 96840 }, { "grad_norm": 0.14900510013103485, "learning_rate": 2.712041623135986e-07, "loss": 0.0074, "step": 96850 }, { "grad_norm": 0.12789997458457947, "learning_rate": 2.694870546065931e-07, "loss": 0.008, "step": 96860 }, { "grad_norm": 0.20850145816802979, "learning_rate": 2.677753853537246e-07, "loss": 0.0086, "step": 96870 }, { "grad_norm": 0.07022598385810852, "learning_rate": 2.660691547421601e-07, "loss": 0.0078, "step": 96880 }, { "grad_norm": 0.16304630041122437, "learning_rate": 2.6436836295850033e-07, "loss": 0.0083, "step": 96890 }, { "grad_norm": 0.18126562237739563, "learning_rate": 2.6267301018874093e-07, "loss": 0.0072, "step": 96900 }, { "grad_norm": 0.17210479080677032, "learning_rate": 2.6098309661828355e-07, "loss": 0.0092, "step": 96910 }, { "grad_norm": 0.0839337632060051, "learning_rate": 2.592986224319305e-07, "loss": 0.0086, "step": 96920 }, { "grad_norm": 0.07873689383268356, "learning_rate": 2.576195878139009e-07, "loss": 0.0071, "step": 96930 }, { "grad_norm": 0.10792805254459381, "learning_rate": 2.5594599294780364e-07, "loss": 0.0072, "step": 96940 }, { "grad_norm": 0.1199432834982872, "learning_rate": 2.542778380166644e-07, "loss": 0.0081, "step": 96950 }, { "grad_norm": 0.09471429139375687, "learning_rate": 2.526151232029095e-07, "loss": 0.0068, "step": 96960 }, { "grad_norm": 0.09510502219200134, "learning_rate": 2.5095784868836567e-07, "loss": 0.0083, "step": 96970 }, { "grad_norm": 0.07053811848163605, "learning_rate": 2.493060146542825e-07, "loss": 0.0118, "step": 96980 }, { "grad_norm": 0.20442481338977814, "learning_rate": 2.476596212812876e-07, "loss": 0.0094, "step": 96990 }, { "grad_norm": 0.09726055711507797, "learning_rate": 2.4601866874943703e-07, "loss": 0.0076, "step": 97000 }, { "grad_norm": 0.12238042056560516, "learning_rate": 2.4438315723818163e-07, "loss": 0.009, "step": 97010 }, { "grad_norm": 0.18661031126976013, "learning_rate": 2.4275308692637833e-07, "loss": 0.0084, "step": 97020 }, { "grad_norm": 0.1589110642671585, "learning_rate": 2.411284579922901e-07, "loss": 0.0077, "step": 97030 }, { "grad_norm": 0.11822521686553955, "learning_rate": 2.395092706135804e-07, "loss": 0.0137, "step": 97040 }, { "grad_norm": 0.12307682633399963, "learning_rate": 2.3789552496731872e-07, "loss": 0.0075, "step": 97050 }, { "grad_norm": 0.10443856567144394, "learning_rate": 2.3628722122999158e-07, "loss": 0.0096, "step": 97060 }, { "grad_norm": 0.12059825658798218, "learning_rate": 2.3468435957747503e-07, "loss": 0.009, "step": 97070 }, { "grad_norm": 0.11821231991052628, "learning_rate": 2.330869401850566e-07, "loss": 0.0076, "step": 97080 }, { "grad_norm": 0.1635945737361908, "learning_rate": 2.3149496322742437e-07, "loss": 0.0088, "step": 97090 }, { "grad_norm": 0.09569164365530014, "learning_rate": 2.2990842887868346e-07, "loss": 0.0077, "step": 97100 }, { "grad_norm": 0.09615253657102585, "learning_rate": 2.2832733731232292e-07, "loss": 0.0081, "step": 97110 }, { "grad_norm": 0.25743845105171204, "learning_rate": 2.2675168870124886e-07, "loss": 0.0079, "step": 97120 }, { "grad_norm": 0.12783172726631165, "learning_rate": 2.2518148321778455e-07, "loss": 0.0064, "step": 97130 }, { "grad_norm": 0.1296057403087616, "learning_rate": 2.2361672103363706e-07, "loss": 0.0084, "step": 97140 }, { "grad_norm": 0.10928612947463989, "learning_rate": 2.220574023199251e-07, "loss": 0.0075, "step": 97150 }, { "grad_norm": 0.09506770223379135, "learning_rate": 2.2050352724717894e-07, "loss": 0.0092, "step": 97160 }, { "grad_norm": 0.12226957827806473, "learning_rate": 2.1895509598532372e-07, "loss": 0.0079, "step": 97170 }, { "grad_norm": 0.10558874905109406, "learning_rate": 2.1741210870369066e-07, "loss": 0.009, "step": 97180 }, { "grad_norm": 0.14289957284927368, "learning_rate": 2.1587456557102814e-07, "loss": 0.0077, "step": 97190 }, { "grad_norm": 0.09144040942192078, "learning_rate": 2.143424667554683e-07, "loss": 0.0088, "step": 97200 }, { "grad_norm": 0.12089458107948303, "learning_rate": 2.128158124245716e-07, "loss": 0.0094, "step": 97210 }, { "grad_norm": 0.16976548731327057, "learning_rate": 2.1129460274527668e-07, "loss": 0.0086, "step": 97220 }, { "grad_norm": 0.11222628504037857, "learning_rate": 2.0977883788395046e-07, "loss": 0.0089, "step": 97230 }, { "grad_norm": 0.12135408073663712, "learning_rate": 2.0826851800634928e-07, "loss": 0.0077, "step": 97240 }, { "grad_norm": 0.1267249584197998, "learning_rate": 2.0676364327764653e-07, "loss": 0.0074, "step": 97250 }, { "grad_norm": 0.16946685314178467, "learning_rate": 2.0526421386240502e-07, "loss": 0.0083, "step": 97260 }, { "grad_norm": 0.14183761179447174, "learning_rate": 2.0377022992461025e-07, "loss": 0.0087, "step": 97270 }, { "grad_norm": 0.17176881432533264, "learning_rate": 2.0228169162763156e-07, "loss": 0.0095, "step": 97280 }, { "grad_norm": 0.15519793331623077, "learning_rate": 2.0079859913425536e-07, "loss": 0.0086, "step": 97290 }, { "grad_norm": 0.25512367486953735, "learning_rate": 1.9932095260667417e-07, "loss": 0.0106, "step": 97300 }, { "grad_norm": 0.05755137652158737, "learning_rate": 1.9784875220648092e-07, "loss": 0.0067, "step": 97310 }, { "grad_norm": 0.17552214860916138, "learning_rate": 1.9638199809466907e-07, "loss": 0.0104, "step": 97320 }, { "grad_norm": 0.1180691346526146, "learning_rate": 1.9492069043164364e-07, "loss": 0.0085, "step": 97330 }, { "grad_norm": 0.13020852208137512, "learning_rate": 1.9346482937721567e-07, "loss": 0.0093, "step": 97340 }, { "grad_norm": 0.20312832295894623, "learning_rate": 1.920144150905856e-07, "loss": 0.0089, "step": 97350 }, { "grad_norm": 0.14696267247200012, "learning_rate": 1.9056944773037655e-07, "loss": 0.0086, "step": 97360 }, { "grad_norm": 0.14640028774738312, "learning_rate": 1.8912992745460102e-07, "loss": 0.0074, "step": 97370 }, { "grad_norm": 0.12078005075454712, "learning_rate": 1.876958544206886e-07, "loss": 0.0067, "step": 97380 }, { "grad_norm": 0.08348837494850159, "learning_rate": 1.8626722878546942e-07, "loss": 0.0104, "step": 97390 }, { "grad_norm": 0.13320936262607574, "learning_rate": 1.8484405070516851e-07, "loss": 0.0096, "step": 97400 }, { "grad_norm": 0.17771941423416138, "learning_rate": 1.8342632033542805e-07, "loss": 0.0105, "step": 97410 }, { "grad_norm": 0.21498234570026398, "learning_rate": 1.8201403783129066e-07, "loss": 0.0084, "step": 97420 }, { "grad_norm": 0.11244802176952362, "learning_rate": 1.806072033471884e-07, "loss": 0.0078, "step": 97430 }, { "grad_norm": 0.09204237908124924, "learning_rate": 1.79205817036987e-07, "loss": 0.0069, "step": 97440 }, { "grad_norm": 0.10701896250247955, "learning_rate": 1.7780987905393065e-07, "loss": 0.0085, "step": 97450 }, { "grad_norm": 0.2087179273366928, "learning_rate": 1.764193895506805e-07, "loss": 0.0096, "step": 97460 }, { "grad_norm": 0.2627965211868286, "learning_rate": 1.7503434867929824e-07, "loss": 0.0079, "step": 97470 }, { "grad_norm": 0.12178997695446014, "learning_rate": 1.7365475659125165e-07, "loss": 0.0074, "step": 97480 }, { "grad_norm": 0.14330078661441803, "learning_rate": 1.7228061343740332e-07, "loss": 0.0082, "step": 97490 }, { "grad_norm": 0.17196182906627655, "learning_rate": 1.7091191936803308e-07, "loss": 0.0087, "step": 97500 }, { "grad_norm": 0.1505761742591858, "learning_rate": 1.6954867453281563e-07, "loss": 0.0066, "step": 97510 }, { "grad_norm": 0.07644262909889221, "learning_rate": 1.6819087908084285e-07, "loss": 0.007, "step": 97520 }, { "grad_norm": 0.23665183782577515, "learning_rate": 1.6683853316058485e-07, "loss": 0.0101, "step": 97530 }, { "grad_norm": 0.10578387975692749, "learning_rate": 1.6549163691995106e-07, "loss": 0.0073, "step": 97540 }, { "grad_norm": 0.17586705088615417, "learning_rate": 1.641501905062237e-07, "loss": 0.0074, "step": 97550 }, { "grad_norm": 0.13501769304275513, "learning_rate": 1.628141940661021e-07, "loss": 0.0103, "step": 97560 }, { "grad_norm": 0.11983788013458252, "learning_rate": 1.6148364774569158e-07, "loss": 0.0079, "step": 97570 }, { "grad_norm": 0.1461806744337082, "learning_rate": 1.6015855169050355e-07, "loss": 0.0079, "step": 97580 }, { "grad_norm": 0.12241428345441818, "learning_rate": 1.5883890604543873e-07, "loss": 0.007, "step": 97590 }, { "grad_norm": 0.0976567342877388, "learning_rate": 1.5752471095482057e-07, "loss": 0.0065, "step": 97600 }, { "grad_norm": 0.11950714886188507, "learning_rate": 1.5621596656235638e-07, "loss": 0.0063, "step": 97610 }, { "grad_norm": 0.14303292334079742, "learning_rate": 1.5491267301117608e-07, "loss": 0.006, "step": 97620 }, { "grad_norm": 0.11197635531425476, "learning_rate": 1.536148304438101e-07, "loss": 0.007, "step": 97630 }, { "grad_norm": 0.11991382390260696, "learning_rate": 1.5232243900217823e-07, "loss": 0.0085, "step": 97640 }, { "grad_norm": 0.14059415459632874, "learning_rate": 1.51035498827623e-07, "loss": 0.0107, "step": 97650 }, { "grad_norm": 0.15866810083389282, "learning_rate": 1.497540100608763e-07, "loss": 0.0075, "step": 97660 }, { "grad_norm": 0.11968018859624863, "learning_rate": 1.4847797284208153e-07, "loss": 0.0114, "step": 97670 }, { "grad_norm": 0.11971797794103622, "learning_rate": 1.472073873107882e-07, "loss": 0.0079, "step": 97680 }, { "grad_norm": 0.19189698994159698, "learning_rate": 1.459422536059407e-07, "loss": 0.0114, "step": 97690 }, { "grad_norm": 0.10232027620077133, "learning_rate": 1.4468257186589508e-07, "loss": 0.0065, "step": 97700 }, { "grad_norm": 0.13677182793617249, "learning_rate": 1.4342834222840217e-07, "loss": 0.01, "step": 97710 }, { "grad_norm": 0.11346635222434998, "learning_rate": 1.4217956483063566e-07, "loss": 0.0075, "step": 97720 }, { "grad_norm": 0.08866690844297409, "learning_rate": 1.409362398091474e-07, "loss": 0.0058, "step": 97730 }, { "grad_norm": 0.0785636305809021, "learning_rate": 1.3969836729990638e-07, "loss": 0.0106, "step": 97740 }, { "grad_norm": 0.19871635735034943, "learning_rate": 1.384659474382932e-07, "loss": 0.0081, "step": 97750 }, { "grad_norm": 0.11091205477714539, "learning_rate": 1.372389803590779e-07, "loss": 0.0074, "step": 97760 }, { "grad_norm": 0.1300904005765915, "learning_rate": 1.360174661964364e-07, "loss": 0.0076, "step": 97770 }, { "grad_norm": 0.16416895389556885, "learning_rate": 1.3480140508396188e-07, "loss": 0.0069, "step": 97780 }, { "grad_norm": 0.21394364535808563, "learning_rate": 1.335907971546313e-07, "loss": 0.0105, "step": 97790 }, { "grad_norm": 0.07453441619873047, "learning_rate": 1.3238564254083875e-07, "loss": 0.0078, "step": 97800 }, { "grad_norm": 0.12685807049274445, "learning_rate": 1.3118594137437323e-07, "loss": 0.0066, "step": 97810 }, { "grad_norm": 0.10974451899528503, "learning_rate": 1.2999169378644094e-07, "loss": 0.0067, "step": 97820 }, { "grad_norm": 0.12628936767578125, "learning_rate": 1.2880289990763738e-07, "loss": 0.0078, "step": 97830 }, { "grad_norm": 0.12264183163642883, "learning_rate": 1.2761955986796968e-07, "loss": 0.0077, "step": 97840 }, { "grad_norm": 0.09887038171291351, "learning_rate": 1.264416737968399e-07, "loss": 0.0086, "step": 97850 }, { "grad_norm": 0.15627871453762054, "learning_rate": 1.2526924182307275e-07, "loss": 0.0065, "step": 97860 }, { "grad_norm": 0.11635681241750717, "learning_rate": 1.2410226407487124e-07, "loss": 0.0103, "step": 97870 }, { "grad_norm": 0.09680549055337906, "learning_rate": 1.229407406798555e-07, "loss": 0.0106, "step": 97880 }, { "grad_norm": 0.16845011711120605, "learning_rate": 1.2178467176505726e-07, "loss": 0.0091, "step": 97890 }, { "grad_norm": 0.11983168125152588, "learning_rate": 1.2063405745689205e-07, "loss": 0.0079, "step": 97900 }, { "grad_norm": 0.1278536021709442, "learning_rate": 1.1948889788119812e-07, "loss": 0.0068, "step": 97910 }, { "grad_norm": 0.09462681412696838, "learning_rate": 1.1834919316320303e-07, "loss": 0.0088, "step": 97920 }, { "grad_norm": 0.15761029720306396, "learning_rate": 1.1721494342754048e-07, "loss": 0.0094, "step": 97930 }, { "grad_norm": 0.11589884012937546, "learning_rate": 1.1608614879825563e-07, "loss": 0.0083, "step": 97940 }, { "grad_norm": 0.20634476840496063, "learning_rate": 1.1496280939879422e-07, "loss": 0.0087, "step": 97950 }, { "grad_norm": 0.09920895099639893, "learning_rate": 1.1384492535199687e-07, "loss": 0.0082, "step": 97960 }, { "grad_norm": 0.07423578202724457, "learning_rate": 1.1273249678011578e-07, "loss": 0.0073, "step": 97970 }, { "grad_norm": 0.07946125417947769, "learning_rate": 1.1162552380480362e-07, "loss": 0.0068, "step": 97980 }, { "grad_norm": 0.11059129983186722, "learning_rate": 1.1052400654711359e-07, "loss": 0.0075, "step": 97990 }, { "grad_norm": 0.18219804763793945, "learning_rate": 1.0942794512751597e-07, "loss": 0.0104, "step": 98000 }, { "grad_norm": 0.12127053737640381, "learning_rate": 1.0833733966587045e-07, "loss": 0.0074, "step": 98010 }, { "grad_norm": 0.09092891961336136, "learning_rate": 1.0725219028143718e-07, "loss": 0.007, "step": 98020 }, { "grad_norm": 0.13170774281024933, "learning_rate": 1.0617249709289345e-07, "loss": 0.0094, "step": 98030 }, { "grad_norm": 0.4082958400249481, "learning_rate": 1.0509826021831703e-07, "loss": 0.0078, "step": 98040 }, { "grad_norm": 0.18245252966880798, "learning_rate": 1.0402947977517508e-07, "loss": 0.0101, "step": 98050 }, { "grad_norm": 0.0795755684375763, "learning_rate": 1.0296615588035185e-07, "loss": 0.0071, "step": 98060 }, { "grad_norm": 0.1705581545829773, "learning_rate": 1.0190828865012659e-07, "loss": 0.0096, "step": 98070 }, { "grad_norm": 0.1305503100156784, "learning_rate": 1.008558782001956e-07, "loss": 0.0069, "step": 98080 }, { "grad_norm": 0.09337779134511948, "learning_rate": 9.980892464563906e-08, "loss": 0.0078, "step": 98090 }, { "grad_norm": 0.10397607833147049, "learning_rate": 9.876742810095985e-08, "loss": 0.0088, "step": 98100 }, { "grad_norm": 0.09919881075620651, "learning_rate": 9.77313886800446e-08, "loss": 0.0081, "step": 98110 }, { "grad_norm": 0.15319156646728516, "learning_rate": 9.670080649619717e-08, "loss": 0.0067, "step": 98120 }, { "grad_norm": 0.14262384176254272, "learning_rate": 9.567568166212737e-08, "loss": 0.0074, "step": 98130 }, { "grad_norm": 0.10159044712781906, "learning_rate": 9.465601428992888e-08, "loss": 0.0072, "step": 98140 }, { "grad_norm": 0.08657589554786682, "learning_rate": 9.364180449111803e-08, "loss": 0.0087, "step": 98150 }, { "grad_norm": 0.09951850771903992, "learning_rate": 9.263305237661169e-08, "loss": 0.0166, "step": 98160 }, { "grad_norm": 0.17124545574188232, "learning_rate": 9.162975805671603e-08, "loss": 0.0077, "step": 98170 }, { "grad_norm": 0.11395618319511414, "learning_rate": 9.063192164115442e-08, "loss": 0.0077, "step": 98180 }, { "grad_norm": 0.13229897618293762, "learning_rate": 8.963954323904511e-08, "loss": 0.008, "step": 98190 }, { "grad_norm": 0.14440488815307617, "learning_rate": 8.865262295891796e-08, "loss": 0.0064, "step": 98200 }, { "grad_norm": 0.10863113403320312, "learning_rate": 8.767116090870331e-08, "loss": 0.0081, "step": 98210 }, { "grad_norm": 0.25171250104904175, "learning_rate": 8.669515719572086e-08, "loss": 0.0128, "step": 98220 }, { "grad_norm": 0.08312540501356125, "learning_rate": 8.572461192671855e-08, "loss": 0.0066, "step": 98230 }, { "grad_norm": 0.12232013791799545, "learning_rate": 8.475952520782815e-08, "loss": 0.0077, "step": 98240 }, { "grad_norm": 0.10583049803972244, "learning_rate": 8.379989714458747e-08, "loss": 0.007, "step": 98250 }, { "grad_norm": 0.12465452402830124, "learning_rate": 8.284572784194034e-08, "loss": 0.0086, "step": 98260 }, { "grad_norm": 0.1474941372871399, "learning_rate": 8.189701740424216e-08, "loss": 0.0096, "step": 98270 }, { "grad_norm": 0.07079204171895981, "learning_rate": 8.095376593522663e-08, "loss": 0.0075, "step": 98280 }, { "grad_norm": 0.1439376175403595, "learning_rate": 8.001597353806123e-08, "loss": 0.0072, "step": 98290 }, { "grad_norm": 0.11327829957008362, "learning_rate": 7.90836403152917e-08, "loss": 0.01, "step": 98300 }, { "grad_norm": 0.09712520986795425, "learning_rate": 7.815676636888092e-08, "loss": 0.0079, "step": 98310 }, { "grad_norm": 0.16032636165618896, "learning_rate": 7.723535180019226e-08, "loss": 0.0102, "step": 98320 }, { "grad_norm": 0.11807839572429657, "learning_rate": 7.6319396709984e-08, "loss": 0.0063, "step": 98330 }, { "grad_norm": 0.13228008151054382, "learning_rate": 7.540890119843158e-08, "loss": 0.0084, "step": 98340 }, { "grad_norm": 0.12117180973291397, "learning_rate": 7.450386536509979e-08, "loss": 0.0075, "step": 98350 }, { "grad_norm": 0.07707128673791885, "learning_rate": 7.360428930895947e-08, "loss": 0.0062, "step": 98360 }, { "grad_norm": 0.10768428444862366, "learning_rate": 7.271017312839302e-08, "loss": 0.0077, "step": 98370 }, { "grad_norm": 0.13093632459640503, "learning_rate": 7.182151692118333e-08, "loss": 0.0086, "step": 98380 }, { "grad_norm": 0.1004772037267685, "learning_rate": 7.093832078449713e-08, "loss": 0.0065, "step": 98390 }, { "grad_norm": 0.09494045376777649, "learning_rate": 7.006058481493493e-08, "loss": 0.0112, "step": 98400 }, { "grad_norm": 0.17252567410469055, "learning_rate": 6.918830910847552e-08, "loss": 0.0086, "step": 98410 }, { "grad_norm": 0.11933565139770508, "learning_rate": 6.832149376052033e-08, "loss": 0.008, "step": 98420 }, { "grad_norm": 0.11547316610813141, "learning_rate": 6.746013886584912e-08, "loss": 0.0074, "step": 98430 }, { "grad_norm": 0.16446176171302795, "learning_rate": 6.660424451866431e-08, "loss": 0.0078, "step": 98440 }, { "grad_norm": 0.13159671425819397, "learning_rate": 6.575381081256327e-08, "loss": 0.0087, "step": 98450 }, { "grad_norm": 0.1440056413412094, "learning_rate": 6.49088378405549e-08, "loss": 0.0078, "step": 98460 }, { "grad_norm": 0.13201001286506653, "learning_rate": 6.406932569503755e-08, "loss": 0.0086, "step": 98470 }, { "grad_norm": 0.11926054954528809, "learning_rate": 6.323527446782662e-08, "loss": 0.0088, "step": 98480 }, { "grad_norm": 0.14300821721553802, "learning_rate": 6.240668425012142e-08, "loss": 0.0092, "step": 98490 }, { "grad_norm": 0.08110763877630234, "learning_rate": 6.158355513254388e-08, "loss": 0.0112, "step": 98500 }, { "grad_norm": 0.11603128165006638, "learning_rate": 6.076588720510534e-08, "loss": 0.0071, "step": 98510 }, { "grad_norm": 0.11783228069543839, "learning_rate": 5.995368055722316e-08, "loss": 0.0094, "step": 98520 }, { "grad_norm": 0.1833401322364807, "learning_rate": 5.914693527773185e-08, "loss": 0.0069, "step": 98530 }, { "grad_norm": 0.14751790463924408, "learning_rate": 5.834565145483861e-08, "loss": 0.0105, "step": 98540 }, { "grad_norm": 0.10846436768770218, "learning_rate": 5.754982917618446e-08, "loss": 0.0065, "step": 98550 }, { "grad_norm": 0.12171569466590881, "learning_rate": 5.675946852878866e-08, "loss": 0.0073, "step": 98560 }, { "grad_norm": 0.14338836073875427, "learning_rate": 5.59745695990932e-08, "loss": 0.0085, "step": 98570 }, { "grad_norm": 0.06808973103761673, "learning_rate": 5.51951324729294e-08, "loss": 0.0069, "step": 98580 }, { "grad_norm": 0.15915259718894958, "learning_rate": 5.4421157235529093e-08, "loss": 0.0098, "step": 98590 }, { "grad_norm": 0.11170068383216858, "learning_rate": 5.365264397154679e-08, "loss": 0.0091, "step": 98600 }, { "grad_norm": 0.6683058738708496, "learning_rate": 5.288959276501526e-08, "loss": 0.0083, "step": 98610 }, { "grad_norm": 0.16010980308055878, "learning_rate": 5.2132003699378876e-08, "loss": 0.0088, "step": 98620 }, { "grad_norm": 0.1609208583831787, "learning_rate": 5.137987685749912e-08, "loss": 0.0081, "step": 98630 }, { "grad_norm": 0.05538836494088173, "learning_rate": 5.0633212321610227e-08, "loss": 0.007, "step": 98640 }, { "grad_norm": 0.12636426091194153, "learning_rate": 4.989201017338019e-08, "loss": 0.0076, "step": 98650 }, { "grad_norm": 0.1910606324672699, "learning_rate": 4.9156270493860846e-08, "loss": 0.0093, "step": 98660 }, { "grad_norm": 0.1156497374176979, "learning_rate": 4.842599336351561e-08, "loss": 0.0089, "step": 98670 }, { "grad_norm": 0.08305522799491882, "learning_rate": 4.7701178862197274e-08, "loss": 0.0081, "step": 98680 }, { "grad_norm": 0.09650954604148865, "learning_rate": 4.698182706918131e-08, "loss": 0.0088, "step": 98690 }, { "grad_norm": 0.16958652436733246, "learning_rate": 4.6267938063121466e-08, "loss": 0.0068, "step": 98700 }, { "grad_norm": 0.13878469169139862, "learning_rate": 4.555951192209973e-08, "loss": 0.0079, "step": 98710 }, { "grad_norm": 0.10269437730312347, "learning_rate": 4.485654872358747e-08, "loss": 0.0072, "step": 98720 }, { "grad_norm": 0.12791341543197632, "learning_rate": 4.415904854445097e-08, "loss": 0.0083, "step": 98730 }, { "grad_norm": 0.07999741286039352, "learning_rate": 4.34670114609792e-08, "loss": 0.0082, "step": 98740 }, { "grad_norm": 0.07330676168203354, "learning_rate": 4.278043754884498e-08, "loss": 0.0078, "step": 98750 }, { "grad_norm": 0.16905687749385834, "learning_rate": 4.209932688312712e-08, "loss": 0.0122, "step": 98760 }, { "grad_norm": 0.16902458667755127, "learning_rate": 4.14236795383216e-08, "loss": 0.0125, "step": 98770 }, { "grad_norm": 0.13828317821025848, "learning_rate": 4.075349558830821e-08, "loss": 0.0079, "step": 98780 }, { "grad_norm": 0.12680210173130035, "learning_rate": 4.008877510638387e-08, "loss": 0.0064, "step": 98790 }, { "grad_norm": 0.12162118405103683, "learning_rate": 3.942951816523488e-08, "loss": 0.0076, "step": 98800 }, { "grad_norm": 0.14452625811100006, "learning_rate": 3.8775724836959125e-08, "loss": 0.0103, "step": 98810 }, { "grad_norm": 0.11163070052862167, "learning_rate": 3.812739519305497e-08, "loss": 0.0094, "step": 98820 }, { "grad_norm": 0.11862180382013321, "learning_rate": 3.748452930442126e-08, "loss": 0.012, "step": 98830 }, { "grad_norm": 0.11855506151914597, "learning_rate": 3.6847127241362855e-08, "loss": 0.011, "step": 98840 }, { "grad_norm": 0.059680137783288956, "learning_rate": 3.621518907358512e-08, "loss": 0.009, "step": 98850 }, { "grad_norm": 0.08249017596244812, "learning_rate": 3.558871487019388e-08, "loss": 0.0096, "step": 98860 }, { "grad_norm": 0.11174152046442032, "learning_rate": 3.4967704699701007e-08, "loss": 0.0072, "step": 98870 }, { "grad_norm": 0.0817015990614891, "learning_rate": 3.435215863001884e-08, "loss": 0.0073, "step": 98880 }, { "grad_norm": 0.18559041619300842, "learning_rate": 3.374207672846019e-08, "loss": 0.0089, "step": 98890 }, { "grad_norm": 0.17015498876571655, "learning_rate": 3.313745906174392e-08, "loss": 0.0087, "step": 98900 }, { "grad_norm": 0.10131026804447174, "learning_rate": 3.253830569599492e-08, "loss": 0.0096, "step": 98910 }, { "grad_norm": 0.10256610065698624, "learning_rate": 3.1944616696727436e-08, "loss": 0.0072, "step": 98920 }, { "grad_norm": 0.1642349660396576, "learning_rate": 3.135639212887287e-08, "loss": 0.0076, "step": 98930 }, { "grad_norm": 0.09086530655622482, "learning_rate": 3.077363205675754e-08, "loss": 0.007, "step": 98940 }, { "grad_norm": 0.16933615505695343, "learning_rate": 3.0196336544113804e-08, "loss": 0.0084, "step": 98950 }, { "grad_norm": 0.11278267949819565, "learning_rate": 2.9624505654063383e-08, "loss": 0.0067, "step": 98960 }, { "grad_norm": 0.12482579052448273, "learning_rate": 2.90581394491507e-08, "loss": 0.0084, "step": 98970 }, { "grad_norm": 0.14466744661331177, "learning_rate": 2.8497237991309545e-08, "loss": 0.0092, "step": 98980 }, { "grad_norm": 0.1272234469652176, "learning_rate": 2.7941801341879735e-08, "loss": 0.0089, "step": 98990 }, { "grad_norm": 0.06499214470386505, "learning_rate": 2.7391829561601578e-08, "loss": 0.0103, "step": 99000 }, { "grad_norm": 0.07518211007118225, "learning_rate": 2.6847322710621403e-08, "loss": 0.0084, "step": 99010 }, { "grad_norm": 0.11835933476686478, "learning_rate": 2.6308280848486022e-08, "loss": 0.0076, "step": 99020 }, { "grad_norm": 0.1389731913805008, "learning_rate": 2.5774704034137177e-08, "loss": 0.0082, "step": 99030 }, { "grad_norm": 0.11714170128107071, "learning_rate": 2.524659232593929e-08, "loss": 0.0064, "step": 99040 }, { "grad_norm": 0.13320264220237732, "learning_rate": 2.472394578163506e-08, "loss": 0.0088, "step": 99050 }, { "grad_norm": 0.13754001259803772, "learning_rate": 2.4206764458378772e-08, "loss": 0.0076, "step": 99060 }, { "grad_norm": 0.16920140385627747, "learning_rate": 2.3695048412736285e-08, "loss": 0.0075, "step": 99070 }, { "grad_norm": 0.16406206786632538, "learning_rate": 2.3188797700668395e-08, "loss": 0.0082, "step": 99080 }, { "grad_norm": 0.10477238893508911, "learning_rate": 2.268801237753082e-08, "loss": 0.0078, "step": 99090 }, { "grad_norm": 0.11171247810125351, "learning_rate": 2.2192692498090862e-08, "loss": 0.0075, "step": 99100 }, { "grad_norm": 0.13591602444648743, "learning_rate": 2.170283811652185e-08, "loss": 0.0075, "step": 99110 }, { "grad_norm": 0.17286691069602966, "learning_rate": 2.12184492863865e-08, "loss": 0.0075, "step": 99120 }, { "grad_norm": 0.13147592544555664, "learning_rate": 2.073952606066465e-08, "loss": 0.0074, "step": 99130 }, { "grad_norm": 0.2015761286020279, "learning_rate": 2.026606849171997e-08, "loss": 0.0081, "step": 99140 }, { "grad_norm": 0.1585599035024643, "learning_rate": 1.9798076631333262e-08, "loss": 0.0085, "step": 99150 }, { "grad_norm": 0.1118311733007431, "learning_rate": 1.933555053069136e-08, "loss": 0.0088, "step": 99160 }, { "grad_norm": 0.0973859652876854, "learning_rate": 1.887849024036492e-08, "loss": 0.0104, "step": 99170 }, { "grad_norm": 0.10122644901275635, "learning_rate": 1.8426895810341736e-08, "loss": 0.0101, "step": 99180 }, { "grad_norm": 0.1731409728527069, "learning_rate": 1.798076729000453e-08, "loss": 0.0072, "step": 99190 }, { "grad_norm": 0.10418032854795456, "learning_rate": 1.75401047281476e-08, "loss": 0.009, "step": 99200 }, { "grad_norm": 0.12711600959300995, "learning_rate": 1.7104908172954626e-08, "loss": 0.0086, "step": 99210 }, { "grad_norm": 0.1120440661907196, "learning_rate": 1.6675177672020868e-08, "loss": 0.0068, "step": 99220 }, { "grad_norm": 0.16854026913642883, "learning_rate": 1.6250913272342073e-08, "loss": 0.0079, "step": 99230 }, { "grad_norm": 0.18617039918899536, "learning_rate": 1.5832115020314453e-08, "loss": 0.0077, "step": 99240 }, { "grad_norm": 0.10588373243808746, "learning_rate": 1.5418782961734712e-08, "loss": 0.0085, "step": 99250 }, { "grad_norm": 0.16136877238750458, "learning_rate": 1.501091714181113e-08, "loss": 0.0086, "step": 99260 }, { "grad_norm": 0.11436522006988525, "learning_rate": 1.460851760513582e-08, "loss": 0.0077, "step": 99270 }, { "grad_norm": 0.12903712689876556, "learning_rate": 1.4211584395723566e-08, "loss": 0.0084, "step": 99280 }, { "grad_norm": 0.1407465934753418, "learning_rate": 1.382011755697854e-08, "loss": 0.0083, "step": 99290 }, { "grad_norm": 0.14403334259986877, "learning_rate": 1.3434117131716494e-08, "loss": 0.0085, "step": 99300 }, { "grad_norm": 0.07187765091657639, "learning_rate": 1.3053583162142557e-08, "loss": 0.0073, "step": 99310 }, { "grad_norm": 0.1466134935617447, "learning_rate": 1.2678515689873438e-08, "loss": 0.008, "step": 99320 }, { "grad_norm": 0.11411724984645844, "learning_rate": 1.2308914755931877e-08, "loss": 0.0071, "step": 99330 }, { "grad_norm": 0.1389012336730957, "learning_rate": 1.1944780400729993e-08, "loss": 0.0102, "step": 99340 }, { "grad_norm": 0.19470974802970886, "learning_rate": 1.1586112664085935e-08, "loss": 0.0094, "step": 99350 }, { "grad_norm": 0.24177680909633636, "learning_rate": 1.1232911585234984e-08, "loss": 0.0084, "step": 99360 }, { "grad_norm": 0.1355278044939041, "learning_rate": 1.0885177202790697e-08, "loss": 0.0081, "step": 99370 }, { "grad_norm": 0.11309395730495453, "learning_rate": 1.0542909554789315e-08, "loss": 0.0108, "step": 99380 }, { "grad_norm": 0.14180070161819458, "learning_rate": 1.0206108678656456e-08, "loss": 0.0094, "step": 99390 }, { "grad_norm": 0.13671089708805084, "learning_rate": 9.874774611223769e-09, "loss": 0.0083, "step": 99400 }, { "grad_norm": 0.12144114822149277, "learning_rate": 9.548907388728933e-09, "loss": 0.0077, "step": 99410 }, { "grad_norm": 0.17137491703033447, "learning_rate": 9.228507046804557e-09, "loss": 0.0078, "step": 99420 }, { "grad_norm": 0.1455167979001999, "learning_rate": 8.91357362048928e-09, "loss": 0.0076, "step": 99430 }, { "grad_norm": 0.13383294641971588, "learning_rate": 8.604107144227769e-09, "loss": 0.0112, "step": 99440 }, { "grad_norm": 0.1303377002477646, "learning_rate": 8.300107651859623e-09, "loss": 0.0071, "step": 99450 }, { "grad_norm": 0.11172647774219513, "learning_rate": 8.001575176630472e-09, "loss": 0.0102, "step": 99460 }, { "grad_norm": 0.15362803637981415, "learning_rate": 7.708509751186422e-09, "loss": 0.0092, "step": 99470 }, { "grad_norm": 0.26858165860176086, "learning_rate": 7.420911407579611e-09, "loss": 0.008, "step": 99480 }, { "grad_norm": 0.08533700555562973, "learning_rate": 7.13878017725711e-09, "loss": 0.0073, "step": 99490 }, { "grad_norm": 0.08712811022996902, "learning_rate": 6.8621160910720125e-09, "loss": 0.0083, "step": 99500 }, { "grad_norm": 0.07517609745264053, "learning_rate": 6.59091917928345e-09, "loss": 0.0082, "step": 99510 }, { "grad_norm": 0.17117002606391907, "learning_rate": 6.32518947155103e-09, "loss": 0.0087, "step": 99520 }, { "grad_norm": 0.1461884081363678, "learning_rate": 6.064926996929288e-09, "loss": 0.0069, "step": 99530 }, { "grad_norm": 0.12011713534593582, "learning_rate": 5.810131783884343e-09, "loss": 0.0084, "step": 99540 }, { "grad_norm": 0.17126691341400146, "learning_rate": 5.5608038602772415e-09, "loss": 0.008, "step": 99550 }, { "grad_norm": 0.08188438415527344, "learning_rate": 5.316943253375062e-09, "loss": 0.0087, "step": 99560 }, { "grad_norm": 0.17485031485557556, "learning_rate": 5.07854998984536e-09, "loss": 0.0101, "step": 99570 }, { "grad_norm": 0.12749885022640228, "learning_rate": 4.845624095756174e-09, "loss": 0.0078, "step": 99580 }, { "grad_norm": 0.09760216623544693, "learning_rate": 4.618165596587121e-09, "loss": 0.0076, "step": 99590 }, { "grad_norm": 0.1703924834728241, "learning_rate": 4.396174517207197e-09, "loss": 0.0077, "step": 99600 }, { "grad_norm": 0.11714360117912292, "learning_rate": 4.179650881896979e-09, "loss": 0.0079, "step": 99610 }, { "grad_norm": 0.12460566312074661, "learning_rate": 3.968594714331975e-09, "loss": 0.0089, "step": 99620 }, { "grad_norm": 0.12205252796411514, "learning_rate": 3.763006037593719e-09, "loss": 0.0085, "step": 99630 }, { "grad_norm": 0.11734869331121445, "learning_rate": 3.5628848741586786e-09, "loss": 0.0077, "step": 99640 }, { "grad_norm": 0.13052481412887573, "learning_rate": 3.368231245926001e-09, "loss": 0.0091, "step": 99650 }, { "grad_norm": 0.16188688576221466, "learning_rate": 3.179045174167561e-09, "loss": 0.0066, "step": 99660 }, { "grad_norm": 0.16264043748378754, "learning_rate": 2.9953266795834657e-09, "loss": 0.0072, "step": 99670 }, { "grad_norm": 0.12389810383319855, "learning_rate": 2.817075782263201e-09, "loss": 0.0075, "step": 99680 }, { "grad_norm": 0.064769446849823, "learning_rate": 2.6442925016967325e-09, "loss": 0.0063, "step": 99690 }, { "grad_norm": 0.1127057895064354, "learning_rate": 2.4769768567800555e-09, "loss": 0.0105, "step": 99700 }, { "grad_norm": 0.11660845577716827, "learning_rate": 2.315128865809646e-09, "loss": 0.0063, "step": 99710 }, { "grad_norm": 0.08308093994855881, "learning_rate": 2.1587485464880097e-09, "loss": 0.0074, "step": 99720 }, { "grad_norm": 0.10569459199905396, "learning_rate": 2.0078359159125813e-09, "loss": 0.0083, "step": 99730 }, { "grad_norm": 0.20232999324798584, "learning_rate": 1.8623909905923776e-09, "loss": 0.009, "step": 99740 }, { "grad_norm": 0.12741103768348694, "learning_rate": 1.7224137864257916e-09, "loss": 0.0078, "step": 99750 }, { "grad_norm": 0.0916212946176529, "learning_rate": 1.5879043187283505e-09, "loss": 0.0084, "step": 99760 }, { "grad_norm": 0.10040013492107391, "learning_rate": 1.458862602204958e-09, "loss": 0.0074, "step": 99770 }, { "grad_norm": 0.17953434586524963, "learning_rate": 1.3352886509720996e-09, "loss": 0.0083, "step": 99780 }, { "grad_norm": 0.17618601024150848, "learning_rate": 1.2171824785356389e-09, "loss": 0.0097, "step": 99790 }, { "grad_norm": 0.10886667668819427, "learning_rate": 1.1045440978185718e-09, "loss": 0.0063, "step": 99800 }, { "grad_norm": 0.09109670668840408, "learning_rate": 9.973735211332714e-10, "loss": 0.0079, "step": 99810 }, { "grad_norm": 0.1264573484659195, "learning_rate": 8.956707602036929e-10, "loss": 0.0064, "step": 99820 }, { "grad_norm": 0.13382630050182343, "learning_rate": 7.994358261542712e-10, "loss": 0.0094, "step": 99830 }, { "grad_norm": 0.1099303811788559, "learning_rate": 7.086687295043692e-10, "loss": 0.0092, "step": 99840 }, { "grad_norm": 0.11528951674699783, "learning_rate": 6.23369480179381e-10, "loss": 0.01, "step": 99850 }, { "grad_norm": 0.07420440018177032, "learning_rate": 5.43538087510731e-10, "loss": 0.0096, "step": 99860 }, { "grad_norm": 0.10156292468309402, "learning_rate": 4.691745602303233e-10, "loss": 0.0076, "step": 99870 }, { "grad_norm": 0.14790965616703033, "learning_rate": 4.0027890646499034e-10, "loss": 0.0085, "step": 99880 }, { "grad_norm": 0.08510530740022659, "learning_rate": 3.368511337531466e-10, "loss": 0.0083, "step": 99890 }, { "grad_norm": 0.18081919848918915, "learning_rate": 2.788912490281348e-10, "loss": 0.0081, "step": 99900 }, { "grad_norm": 0.17981082201004028, "learning_rate": 2.2639925863487954e-10, "loss": 0.0082, "step": 99910 }, { "grad_norm": 0.17824655771255493, "learning_rate": 1.7937516830213163e-10, "loss": 0.0087, "step": 99920 }, { "grad_norm": 0.11844316869974136, "learning_rate": 1.3781898318687702e-10, "loss": 0.0098, "step": 99930 }, { "grad_norm": 0.15747566521167755, "learning_rate": 1.0173070781882566e-10, "loss": 0.0059, "step": 99940 }, { "grad_norm": 0.164200559258461, "learning_rate": 7.111034615592261e-11, "loss": 0.0083, "step": 99950 }, { "grad_norm": 0.11843521147966385, "learning_rate": 4.595790153993918e-11, "loss": 0.0074, "step": 99960 }, { "grad_norm": 0.17240485548973083, "learning_rate": 2.6273376729779585e-11, "loss": 0.0098, "step": 99970 }, { "grad_norm": 0.10519058257341385, "learning_rate": 1.2056773868174276e-11, "loss": 0.0096, "step": 99980 }, { "grad_norm": 0.13520073890686035, "learning_rate": 3.3080945149865926e-12, "loss": 0.0115, "step": 99990 }, { "grad_norm": 0.14285042881965637, "learning_rate": 2.7339625008337977e-14, "loss": 0.008, "step": 100000 } ], "logging_steps": 10, "max_steps": 100000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 20000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 24, "trial_name": null, "trial_params": null }