{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 80000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 2.2298784255981445, "learning_rate": 2.25e-07, "loss": 1.7242, "step": 10 }, { "grad_norm": 1.456304907798767, "learning_rate": 4.75e-07, "loss": 1.7133, "step": 20 }, { "grad_norm": 1.9889506101608276, "learning_rate": 7.25e-07, "loss": 1.7199, "step": 30 }, { "grad_norm": 1.9621801376342773, "learning_rate": 9.75e-07, "loss": 1.668, "step": 40 }, { "grad_norm": 1.5660237073898315, "learning_rate": 1.2250000000000001e-06, "loss": 1.7184, "step": 50 }, { "grad_norm": 2.5951008796691895, "learning_rate": 1.475e-06, "loss": 1.5348, "step": 60 }, { "grad_norm": 1.7581535577774048, "learning_rate": 1.7250000000000002e-06, "loss": 1.4531, "step": 70 }, { "grad_norm": 3.0696451663970947, "learning_rate": 1.975e-06, "loss": 1.3715, "step": 80 }, { "grad_norm": 1.5590258836746216, "learning_rate": 2.225e-06, "loss": 1.3062, "step": 90 }, { "grad_norm": 1.7032015323638916, "learning_rate": 2.4750000000000004e-06, "loss": 1.4078, "step": 100 }, { "grad_norm": 11.179523468017578, "learning_rate": 2.725e-06, "loss": 1.3414, "step": 110 }, { "grad_norm": 1.5377836227416992, "learning_rate": 2.975e-06, "loss": 1.2879, "step": 120 }, { "grad_norm": 2.4886057376861572, "learning_rate": 3.225e-06, "loss": 1.3047, "step": 130 }, { "grad_norm": 1.5185812711715698, "learning_rate": 3.4750000000000006e-06, "loss": 1.2262, "step": 140 }, { "grad_norm": 1.5803375244140625, "learning_rate": 3.725e-06, "loss": 1.2367, "step": 150 }, { "grad_norm": 4.03292179107666, "learning_rate": 3.975e-06, "loss": 1.184, "step": 160 }, { "grad_norm": 2.7000844478607178, "learning_rate": 4.225e-06, "loss": 1.1816, "step": 170 }, { "grad_norm": 2.769853115081787, "learning_rate": 4.475e-06, "loss": 1.1531, "step": 180 }, { "grad_norm": 1.9363588094711304, "learning_rate": 4.7250000000000005e-06, "loss": 1.1449, "step": 190 }, { "grad_norm": 2.195021629333496, "learning_rate": 4.975000000000001e-06, "loss": 1.1277, "step": 200 }, { "grad_norm": 1.5432817935943604, "learning_rate": 5.225e-06, "loss": 1.1148, "step": 210 }, { "grad_norm": 2.0386433601379395, "learning_rate": 5.475e-06, "loss": 1.1211, "step": 220 }, { "grad_norm": 2.212369680404663, "learning_rate": 5.725e-06, "loss": 1.1172, "step": 230 }, { "grad_norm": 1.4101619720458984, "learning_rate": 5.975e-06, "loss": 1.0838, "step": 240 }, { "grad_norm": 2.043989658355713, "learning_rate": 6.2250000000000005e-06, "loss": 1.0746, "step": 250 }, { "grad_norm": 1.2678929567337036, "learning_rate": 6.475000000000001e-06, "loss": 1.0557, "step": 260 }, { "grad_norm": 2.4053618907928467, "learning_rate": 6.725000000000001e-06, "loss": 1.0711, "step": 270 }, { "grad_norm": 2.3797340393066406, "learning_rate": 6.975000000000001e-06, "loss": 1.0678, "step": 280 }, { "grad_norm": 1.897119164466858, "learning_rate": 7.2249999999999994e-06, "loss": 1.0754, "step": 290 }, { "grad_norm": 1.6700385808944702, "learning_rate": 7.4750000000000004e-06, "loss": 1.0688, "step": 300 }, { "grad_norm": 1.7525588274002075, "learning_rate": 7.725e-06, "loss": 1.0719, "step": 310 }, { "grad_norm": 1.5169082880020142, "learning_rate": 7.975e-06, "loss": 1.035, "step": 320 }, { "grad_norm": 1.661608338356018, "learning_rate": 8.225e-06, "loss": 1.0705, "step": 330 }, { "grad_norm": 1.5686413049697876, "learning_rate": 8.475000000000001e-06, "loss": 1.0473, "step": 340 }, { "grad_norm": 1.702304720878601, "learning_rate": 8.725e-06, "loss": 1.0324, "step": 350 }, { "grad_norm": 1.739457368850708, "learning_rate": 8.975e-06, "loss": 1.0314, "step": 360 }, { "grad_norm": 1.397358775138855, "learning_rate": 9.225e-06, "loss": 1.0354, "step": 370 }, { "grad_norm": 2.0839264392852783, "learning_rate": 9.475e-06, "loss": 1.0072, "step": 380 }, { "grad_norm": 2.198821783065796, "learning_rate": 9.725000000000001e-06, "loss": 0.9705, "step": 390 }, { "grad_norm": 2.223331928253174, "learning_rate": 9.975e-06, "loss": 0.9258, "step": 400 }, { "grad_norm": 2.632025718688965, "learning_rate": 1.0225e-05, "loss": 0.8426, "step": 410 }, { "grad_norm": 3.230532169342041, "learning_rate": 1.0475e-05, "loss": 0.7645, "step": 420 }, { "grad_norm": 3.2102677822113037, "learning_rate": 1.0725e-05, "loss": 0.7205, "step": 430 }, { "grad_norm": 2.610898494720459, "learning_rate": 1.0975e-05, "loss": 0.6543, "step": 440 }, { "grad_norm": 3.277453660964966, "learning_rate": 1.1225e-05, "loss": 0.5794, "step": 450 }, { "grad_norm": 3.4922893047332764, "learning_rate": 1.1475000000000001e-05, "loss": 0.5425, "step": 460 }, { "grad_norm": 2.9619343280792236, "learning_rate": 1.1725e-05, "loss": 0.5197, "step": 470 }, { "grad_norm": 2.89013671875, "learning_rate": 1.1975e-05, "loss": 0.4993, "step": 480 }, { "grad_norm": 1.981811761856079, "learning_rate": 1.2225e-05, "loss": 0.4326, "step": 490 }, { "grad_norm": 3.233821153640747, "learning_rate": 1.2475e-05, "loss": 0.382, "step": 500 }, { "grad_norm": 2.7934303283691406, "learning_rate": 1.2725000000000001e-05, "loss": 0.368, "step": 510 }, { "grad_norm": 2.664538860321045, "learning_rate": 1.2975e-05, "loss": 0.3189, "step": 520 }, { "grad_norm": 3.0831573009490967, "learning_rate": 1.3225000000000001e-05, "loss": 0.3111, "step": 530 }, { "grad_norm": 2.8626461029052734, "learning_rate": 1.3475000000000002e-05, "loss": 0.2445, "step": 540 }, { "grad_norm": 3.2424328327178955, "learning_rate": 1.3725000000000002e-05, "loss": 0.2852, "step": 550 }, { "grad_norm": 3.306307554244995, "learning_rate": 1.3975000000000003e-05, "loss": 0.2752, "step": 560 }, { "grad_norm": 2.990680456161499, "learning_rate": 1.4225e-05, "loss": 0.2291, "step": 570 }, { "grad_norm": 3.3184938430786133, "learning_rate": 1.4475e-05, "loss": 0.2245, "step": 580 }, { "grad_norm": 2.7056925296783447, "learning_rate": 1.4725e-05, "loss": 0.2307, "step": 590 }, { "grad_norm": 2.6089489459991455, "learning_rate": 1.4975e-05, "loss": 0.213, "step": 600 }, { "grad_norm": 2.7556309700012207, "learning_rate": 1.5225e-05, "loss": 0.2077, "step": 610 }, { "grad_norm": 3.2376444339752197, "learning_rate": 1.5475e-05, "loss": 0.2141, "step": 620 }, { "grad_norm": 2.5349225997924805, "learning_rate": 1.5725e-05, "loss": 0.2014, "step": 630 }, { "grad_norm": 2.5553903579711914, "learning_rate": 1.5975000000000002e-05, "loss": 0.2057, "step": 640 }, { "grad_norm": 2.42868971824646, "learning_rate": 1.6225e-05, "loss": 0.1889, "step": 650 }, { "grad_norm": 2.9107158184051514, "learning_rate": 1.6475e-05, "loss": 0.2233, "step": 660 }, { "grad_norm": 2.150240182876587, "learning_rate": 1.6725000000000003e-05, "loss": 0.1553, "step": 670 }, { "grad_norm": 3.200167179107666, "learning_rate": 1.6975000000000003e-05, "loss": 0.19, "step": 680 }, { "grad_norm": 2.3171489238739014, "learning_rate": 1.7225e-05, "loss": 0.1833, "step": 690 }, { "grad_norm": 2.79947829246521, "learning_rate": 1.7475e-05, "loss": 0.182, "step": 700 }, { "grad_norm": 2.2762374877929688, "learning_rate": 1.7725e-05, "loss": 0.171, "step": 710 }, { "grad_norm": 3.147722005844116, "learning_rate": 1.7975e-05, "loss": 0.1666, "step": 720 }, { "grad_norm": 2.4783878326416016, "learning_rate": 1.8225e-05, "loss": 0.1637, "step": 730 }, { "grad_norm": 2.1964550018310547, "learning_rate": 1.8475000000000002e-05, "loss": 0.1374, "step": 740 }, { "grad_norm": 2.3240833282470703, "learning_rate": 1.8725e-05, "loss": 0.1584, "step": 750 }, { "grad_norm": 2.096346855163574, "learning_rate": 1.8975e-05, "loss": 0.1606, "step": 760 }, { "grad_norm": 2.046778440475464, "learning_rate": 1.9225e-05, "loss": 0.1707, "step": 770 }, { "grad_norm": 2.544495105743408, "learning_rate": 1.9475000000000002e-05, "loss": 0.2181, "step": 780 }, { "grad_norm": 1.5817638635635376, "learning_rate": 1.9725000000000002e-05, "loss": 0.1568, "step": 790 }, { "grad_norm": 2.3917245864868164, "learning_rate": 1.9975e-05, "loss": 0.1439, "step": 800 }, { "grad_norm": 1.7769970893859863, "learning_rate": 2.0225000000000004e-05, "loss": 0.1509, "step": 810 }, { "grad_norm": 2.5251805782318115, "learning_rate": 2.0475e-05, "loss": 0.1656, "step": 820 }, { "grad_norm": 2.49336838722229, "learning_rate": 2.0725e-05, "loss": 0.1748, "step": 830 }, { "grad_norm": 2.5175020694732666, "learning_rate": 2.0975e-05, "loss": 0.1594, "step": 840 }, { "grad_norm": 2.9579739570617676, "learning_rate": 2.1225e-05, "loss": 0.1584, "step": 850 }, { "grad_norm": 2.496579885482788, "learning_rate": 2.1475e-05, "loss": 0.1515, "step": 860 }, { "grad_norm": 1.9303358793258667, "learning_rate": 2.1725e-05, "loss": 0.1637, "step": 870 }, { "grad_norm": 1.7562605142593384, "learning_rate": 2.1975000000000002e-05, "loss": 0.1402, "step": 880 }, { "grad_norm": 2.0495340824127197, "learning_rate": 2.2225e-05, "loss": 0.141, "step": 890 }, { "grad_norm": 2.1349453926086426, "learning_rate": 2.2475e-05, "loss": 0.1411, "step": 900 }, { "grad_norm": 2.4634203910827637, "learning_rate": 2.2725000000000003e-05, "loss": 0.1362, "step": 910 }, { "grad_norm": 1.7922248840332031, "learning_rate": 2.2975000000000003e-05, "loss": 0.1179, "step": 920 }, { "grad_norm": 1.7750483751296997, "learning_rate": 2.3225000000000002e-05, "loss": 0.1428, "step": 930 }, { "grad_norm": 2.1115896701812744, "learning_rate": 2.3475e-05, "loss": 0.1463, "step": 940 }, { "grad_norm": 2.2020530700683594, "learning_rate": 2.3725e-05, "loss": 0.1234, "step": 950 }, { "grad_norm": 1.7637088298797607, "learning_rate": 2.3975e-05, "loss": 0.124, "step": 960 }, { "grad_norm": 2.1056230068206787, "learning_rate": 2.4225e-05, "loss": 0.1609, "step": 970 }, { "grad_norm": 1.7339425086975098, "learning_rate": 2.4475000000000002e-05, "loss": 0.1148, "step": 980 }, { "grad_norm": 2.483042001724243, "learning_rate": 2.4725e-05, "loss": 0.1124, "step": 990 }, { "grad_norm": 1.9671446084976196, "learning_rate": 2.4975e-05, "loss": 0.124, "step": 1000 }, { "grad_norm": 1.3712846040725708, "learning_rate": 2.5225e-05, "loss": 0.109, "step": 1010 }, { "grad_norm": 1.582654356956482, "learning_rate": 2.5475e-05, "loss": 0.1476, "step": 1020 }, { "grad_norm": 1.864896297454834, "learning_rate": 2.5725e-05, "loss": 0.1062, "step": 1030 }, { "grad_norm": 1.6700177192687988, "learning_rate": 2.5974999999999998e-05, "loss": 0.1293, "step": 1040 }, { "grad_norm": 2.324470043182373, "learning_rate": 2.6225e-05, "loss": 0.1318, "step": 1050 }, { "grad_norm": 1.3987772464752197, "learning_rate": 2.6475e-05, "loss": 0.1278, "step": 1060 }, { "grad_norm": 2.327559232711792, "learning_rate": 2.6725e-05, "loss": 0.1437, "step": 1070 }, { "grad_norm": 1.8907489776611328, "learning_rate": 2.6975000000000002e-05, "loss": 0.1377, "step": 1080 }, { "grad_norm": 2.1409084796905518, "learning_rate": 2.7225e-05, "loss": 0.1363, "step": 1090 }, { "grad_norm": 2.085498332977295, "learning_rate": 2.7475e-05, "loss": 0.1064, "step": 1100 }, { "grad_norm": 1.7662912607192993, "learning_rate": 2.7725e-05, "loss": 0.1535, "step": 1110 }, { "grad_norm": 1.64963698387146, "learning_rate": 2.7975000000000002e-05, "loss": 0.1091, "step": 1120 }, { "grad_norm": 1.746871829032898, "learning_rate": 2.8225e-05, "loss": 0.1401, "step": 1130 }, { "grad_norm": 2.3623592853546143, "learning_rate": 2.8475e-05, "loss": 0.1647, "step": 1140 }, { "grad_norm": 1.686938762664795, "learning_rate": 2.8725e-05, "loss": 0.1348, "step": 1150 }, { "grad_norm": 1.885180115699768, "learning_rate": 2.8975000000000003e-05, "loss": 0.1322, "step": 1160 }, { "grad_norm": 1.6104788780212402, "learning_rate": 2.9225000000000002e-05, "loss": 0.135, "step": 1170 }, { "grad_norm": 1.3156250715255737, "learning_rate": 2.9475e-05, "loss": 0.1036, "step": 1180 }, { "grad_norm": 1.5839684009552002, "learning_rate": 2.9725000000000004e-05, "loss": 0.1301, "step": 1190 }, { "grad_norm": 2.344107151031494, "learning_rate": 2.9975000000000004e-05, "loss": 0.1488, "step": 1200 }, { "grad_norm": 3.4778389930725098, "learning_rate": 3.0225000000000003e-05, "loss": 0.1505, "step": 1210 }, { "grad_norm": 1.535250186920166, "learning_rate": 3.0475000000000002e-05, "loss": 0.1338, "step": 1220 }, { "grad_norm": 1.504900574684143, "learning_rate": 3.0725e-05, "loss": 0.1145, "step": 1230 }, { "grad_norm": 2.0577187538146973, "learning_rate": 3.0975e-05, "loss": 0.127, "step": 1240 }, { "grad_norm": 1.5481061935424805, "learning_rate": 3.122500000000001e-05, "loss": 0.1097, "step": 1250 }, { "grad_norm": 1.803620457649231, "learning_rate": 3.1475e-05, "loss": 0.1395, "step": 1260 }, { "grad_norm": 1.3189525604248047, "learning_rate": 3.1725e-05, "loss": 0.1005, "step": 1270 }, { "grad_norm": 1.1624397039413452, "learning_rate": 3.1975e-05, "loss": 0.131, "step": 1280 }, { "grad_norm": 1.6329725980758667, "learning_rate": 3.2225e-05, "loss": 0.1162, "step": 1290 }, { "grad_norm": 1.943293571472168, "learning_rate": 3.2474999999999997e-05, "loss": 0.1114, "step": 1300 }, { "grad_norm": 1.729225993156433, "learning_rate": 3.2725e-05, "loss": 0.1213, "step": 1310 }, { "grad_norm": 1.2150542736053467, "learning_rate": 3.2975e-05, "loss": 0.1184, "step": 1320 }, { "grad_norm": 1.5122488737106323, "learning_rate": 3.3225e-05, "loss": 0.0945, "step": 1330 }, { "grad_norm": 1.3040146827697754, "learning_rate": 3.3475e-05, "loss": 0.1063, "step": 1340 }, { "grad_norm": 1.6083887815475464, "learning_rate": 3.3725e-05, "loss": 0.1415, "step": 1350 }, { "grad_norm": 2.082900285720825, "learning_rate": 3.3975e-05, "loss": 0.1266, "step": 1360 }, { "grad_norm": 1.4831280708312988, "learning_rate": 3.4225e-05, "loss": 0.1129, "step": 1370 }, { "grad_norm": 1.4909614324569702, "learning_rate": 3.4475000000000005e-05, "loss": 0.1051, "step": 1380 }, { "grad_norm": 1.232156753540039, "learning_rate": 3.4725000000000004e-05, "loss": 0.1052, "step": 1390 }, { "grad_norm": 1.3300786018371582, "learning_rate": 3.4975e-05, "loss": 0.1026, "step": 1400 }, { "grad_norm": 1.2861900329589844, "learning_rate": 3.5225e-05, "loss": 0.1191, "step": 1410 }, { "grad_norm": 1.8448030948638916, "learning_rate": 3.5475e-05, "loss": 0.1129, "step": 1420 }, { "grad_norm": 1.4426863193511963, "learning_rate": 3.5725e-05, "loss": 0.1011, "step": 1430 }, { "grad_norm": 1.3929029703140259, "learning_rate": 3.5975e-05, "loss": 0.1247, "step": 1440 }, { "grad_norm": 1.3699837923049927, "learning_rate": 3.6225000000000006e-05, "loss": 0.1302, "step": 1450 }, { "grad_norm": 0.9256219863891602, "learning_rate": 3.6475000000000006e-05, "loss": 0.1075, "step": 1460 }, { "grad_norm": 1.1944804191589355, "learning_rate": 3.6725000000000005e-05, "loss": 0.1168, "step": 1470 }, { "grad_norm": 1.870071530342102, "learning_rate": 3.6975000000000004e-05, "loss": 0.104, "step": 1480 }, { "grad_norm": 1.7432854175567627, "learning_rate": 3.7225000000000004e-05, "loss": 0.1082, "step": 1490 }, { "grad_norm": 0.9093107581138611, "learning_rate": 3.7475e-05, "loss": 0.1125, "step": 1500 }, { "grad_norm": 1.0381470918655396, "learning_rate": 3.7725e-05, "loss": 0.0923, "step": 1510 }, { "grad_norm": 1.7855560779571533, "learning_rate": 3.7975e-05, "loss": 0.099, "step": 1520 }, { "grad_norm": 1.1601051092147827, "learning_rate": 3.8225e-05, "loss": 0.1077, "step": 1530 }, { "grad_norm": 0.8841872215270996, "learning_rate": 3.8475e-05, "loss": 0.1081, "step": 1540 }, { "grad_norm": 1.0158809423446655, "learning_rate": 3.8725e-05, "loss": 0.0969, "step": 1550 }, { "grad_norm": 1.1540812253952026, "learning_rate": 3.8975e-05, "loss": 0.1239, "step": 1560 }, { "grad_norm": 1.3000612258911133, "learning_rate": 3.9225e-05, "loss": 0.0978, "step": 1570 }, { "grad_norm": 0.8529472947120667, "learning_rate": 3.9475000000000004e-05, "loss": 0.1111, "step": 1580 }, { "grad_norm": 1.8877301216125488, "learning_rate": 3.9725e-05, "loss": 0.117, "step": 1590 }, { "grad_norm": 1.1250687837600708, "learning_rate": 3.9975e-05, "loss": 0.1056, "step": 1600 }, { "grad_norm": 1.1738380193710327, "learning_rate": 4.0225e-05, "loss": 0.0816, "step": 1610 }, { "grad_norm": 0.7664106488227844, "learning_rate": 4.0475e-05, "loss": 0.1021, "step": 1620 }, { "grad_norm": 0.9673675298690796, "learning_rate": 4.0725e-05, "loss": 0.1381, "step": 1630 }, { "grad_norm": 1.7347890138626099, "learning_rate": 4.0975e-05, "loss": 0.1039, "step": 1640 }, { "grad_norm": 0.9964717626571655, "learning_rate": 4.1225e-05, "loss": 0.1074, "step": 1650 }, { "grad_norm": 0.9489464163780212, "learning_rate": 4.1475000000000005e-05, "loss": 0.101, "step": 1660 }, { "grad_norm": 0.9798162579536438, "learning_rate": 4.1725000000000005e-05, "loss": 0.0812, "step": 1670 }, { "grad_norm": 1.1665356159210205, "learning_rate": 4.1975000000000004e-05, "loss": 0.0976, "step": 1680 }, { "grad_norm": 0.9684575200080872, "learning_rate": 4.2225e-05, "loss": 0.0936, "step": 1690 }, { "grad_norm": 0.8323948979377747, "learning_rate": 4.2475e-05, "loss": 0.111, "step": 1700 }, { "grad_norm": 1.1920881271362305, "learning_rate": 4.2725e-05, "loss": 0.1237, "step": 1710 }, { "grad_norm": 1.1023179292678833, "learning_rate": 4.2975e-05, "loss": 0.1095, "step": 1720 }, { "grad_norm": 2.212022066116333, "learning_rate": 4.322500000000001e-05, "loss": 0.1062, "step": 1730 }, { "grad_norm": 1.133650541305542, "learning_rate": 4.3475000000000006e-05, "loss": 0.0956, "step": 1740 }, { "grad_norm": 1.3654613494873047, "learning_rate": 4.3725000000000006e-05, "loss": 0.1006, "step": 1750 }, { "grad_norm": 1.4878596067428589, "learning_rate": 4.3975e-05, "loss": 0.1075, "step": 1760 }, { "grad_norm": 1.3989571332931519, "learning_rate": 4.4225e-05, "loss": 0.0967, "step": 1770 }, { "grad_norm": 1.0092346668243408, "learning_rate": 4.4475e-05, "loss": 0.113, "step": 1780 }, { "grad_norm": 1.5690189599990845, "learning_rate": 4.4725e-05, "loss": 0.0867, "step": 1790 }, { "grad_norm": 1.0975697040557861, "learning_rate": 4.4975e-05, "loss": 0.0866, "step": 1800 }, { "grad_norm": 1.0900931358337402, "learning_rate": 4.5225e-05, "loss": 0.0932, "step": 1810 }, { "grad_norm": 1.4092172384262085, "learning_rate": 4.5475e-05, "loss": 0.1038, "step": 1820 }, { "grad_norm": 1.659443736076355, "learning_rate": 4.5725e-05, "loss": 0.1124, "step": 1830 }, { "grad_norm": 1.2397264242172241, "learning_rate": 4.5975e-05, "loss": 0.1188, "step": 1840 }, { "grad_norm": 1.4303464889526367, "learning_rate": 4.6225e-05, "loss": 0.1332, "step": 1850 }, { "grad_norm": 0.8290777802467346, "learning_rate": 4.6475000000000005e-05, "loss": 0.1113, "step": 1860 }, { "grad_norm": 0.9864966869354248, "learning_rate": 4.6725000000000004e-05, "loss": 0.0967, "step": 1870 }, { "grad_norm": 1.0712636709213257, "learning_rate": 4.6975000000000003e-05, "loss": 0.0902, "step": 1880 }, { "grad_norm": 1.035468578338623, "learning_rate": 4.7225e-05, "loss": 0.0902, "step": 1890 }, { "grad_norm": 0.9210202693939209, "learning_rate": 4.7475e-05, "loss": 0.0923, "step": 1900 }, { "grad_norm": 1.3829723596572876, "learning_rate": 4.7725e-05, "loss": 0.1283, "step": 1910 }, { "grad_norm": 1.0755506753921509, "learning_rate": 4.7975e-05, "loss": 0.1141, "step": 1920 }, { "grad_norm": 1.0059531927108765, "learning_rate": 4.822500000000001e-05, "loss": 0.097, "step": 1930 }, { "grad_norm": 1.4523354768753052, "learning_rate": 4.8475000000000006e-05, "loss": 0.1145, "step": 1940 }, { "grad_norm": 0.9766315817832947, "learning_rate": 4.8725000000000005e-05, "loss": 0.0971, "step": 1950 }, { "grad_norm": 1.3899714946746826, "learning_rate": 4.8975000000000005e-05, "loss": 0.1155, "step": 1960 }, { "grad_norm": 0.9843569397926331, "learning_rate": 4.9225000000000004e-05, "loss": 0.0891, "step": 1970 }, { "grad_norm": 1.6093567609786987, "learning_rate": 4.9475e-05, "loss": 0.1234, "step": 1980 }, { "grad_norm": 1.2971434593200684, "learning_rate": 4.9725e-05, "loss": 0.0951, "step": 1990 }, { "grad_norm": 1.029455542564392, "learning_rate": 4.9975e-05, "loss": 0.1095, "step": 2000 }, { "grad_norm": 0.9245067834854126, "learning_rate": 5.0225e-05, "loss": 0.1388, "step": 2010 }, { "grad_norm": 1.4763301610946655, "learning_rate": 5.047500000000001e-05, "loss": 0.1104, "step": 2020 }, { "grad_norm": 1.0758012533187866, "learning_rate": 5.0725e-05, "loss": 0.0995, "step": 2030 }, { "grad_norm": 0.8081463575363159, "learning_rate": 5.0975000000000006e-05, "loss": 0.1168, "step": 2040 }, { "grad_norm": 1.1746879816055298, "learning_rate": 5.1225e-05, "loss": 0.1102, "step": 2050 }, { "grad_norm": 1.25478196144104, "learning_rate": 5.1475000000000004e-05, "loss": 0.0928, "step": 2060 }, { "grad_norm": 0.9355568289756775, "learning_rate": 5.1725000000000004e-05, "loss": 0.0929, "step": 2070 }, { "grad_norm": 1.3387975692749023, "learning_rate": 5.197500000000001e-05, "loss": 0.095, "step": 2080 }, { "grad_norm": 1.362473964691162, "learning_rate": 5.2225e-05, "loss": 0.0984, "step": 2090 }, { "grad_norm": 1.1937904357910156, "learning_rate": 5.247500000000001e-05, "loss": 0.1017, "step": 2100 }, { "grad_norm": 0.7731842994689941, "learning_rate": 5.2725e-05, "loss": 0.1113, "step": 2110 }, { "grad_norm": 1.00146484375, "learning_rate": 5.297500000000001e-05, "loss": 0.103, "step": 2120 }, { "grad_norm": 1.4968355894088745, "learning_rate": 5.3225e-05, "loss": 0.1066, "step": 2130 }, { "grad_norm": 1.0985361337661743, "learning_rate": 5.3475e-05, "loss": 0.0896, "step": 2140 }, { "grad_norm": 1.0702381134033203, "learning_rate": 5.3725000000000005e-05, "loss": 0.1147, "step": 2150 }, { "grad_norm": 0.8574181795120239, "learning_rate": 5.3975e-05, "loss": 0.1094, "step": 2160 }, { "grad_norm": 0.8629910349845886, "learning_rate": 5.4225000000000003e-05, "loss": 0.0897, "step": 2170 }, { "grad_norm": 0.8283425569534302, "learning_rate": 5.4474999999999996e-05, "loss": 0.1076, "step": 2180 }, { "grad_norm": 0.8567692637443542, "learning_rate": 5.4725e-05, "loss": 0.091, "step": 2190 }, { "grad_norm": 0.8125368356704712, "learning_rate": 5.4975e-05, "loss": 0.0876, "step": 2200 }, { "grad_norm": 0.7449269890785217, "learning_rate": 5.522500000000001e-05, "loss": 0.0854, "step": 2210 }, { "grad_norm": 0.7430480718612671, "learning_rate": 5.5475e-05, "loss": 0.0838, "step": 2220 }, { "grad_norm": 1.0232347249984741, "learning_rate": 5.5725000000000006e-05, "loss": 0.0935, "step": 2230 }, { "grad_norm": 0.6379863619804382, "learning_rate": 5.5975e-05, "loss": 0.0837, "step": 2240 }, { "grad_norm": 0.6971792578697205, "learning_rate": 5.6225000000000005e-05, "loss": 0.0852, "step": 2250 }, { "grad_norm": 0.7175782918930054, "learning_rate": 5.6475e-05, "loss": 0.0874, "step": 2260 }, { "grad_norm": 0.9117566347122192, "learning_rate": 5.6725e-05, "loss": 0.0775, "step": 2270 }, { "grad_norm": 0.9167803525924683, "learning_rate": 5.6975e-05, "loss": 0.0799, "step": 2280 }, { "grad_norm": 1.432693600654602, "learning_rate": 5.722500000000001e-05, "loss": 0.1271, "step": 2290 }, { "grad_norm": 0.7663713097572327, "learning_rate": 5.7475e-05, "loss": 0.0897, "step": 2300 }, { "grad_norm": 1.1497551202774048, "learning_rate": 5.772500000000001e-05, "loss": 0.0897, "step": 2310 }, { "grad_norm": 0.8169030547142029, "learning_rate": 5.7975e-05, "loss": 0.0744, "step": 2320 }, { "grad_norm": 0.767641544342041, "learning_rate": 5.8225000000000006e-05, "loss": 0.104, "step": 2330 }, { "grad_norm": 0.719834566116333, "learning_rate": 5.8475000000000005e-05, "loss": 0.0965, "step": 2340 }, { "grad_norm": 0.6201204061508179, "learning_rate": 5.8725000000000004e-05, "loss": 0.0884, "step": 2350 }, { "grad_norm": 0.9653298854827881, "learning_rate": 5.8975000000000004e-05, "loss": 0.0869, "step": 2360 }, { "grad_norm": 0.5957003831863403, "learning_rate": 5.922500000000001e-05, "loss": 0.088, "step": 2370 }, { "grad_norm": 0.7461864948272705, "learning_rate": 5.9475e-05, "loss": 0.0913, "step": 2380 }, { "grad_norm": 0.780010998249054, "learning_rate": 5.9724999999999995e-05, "loss": 0.1049, "step": 2390 }, { "grad_norm": 0.9924521446228027, "learning_rate": 5.9975e-05, "loss": 0.1039, "step": 2400 }, { "grad_norm": 0.7118524312973022, "learning_rate": 6.0225e-05, "loss": 0.1039, "step": 2410 }, { "grad_norm": 0.9616999626159668, "learning_rate": 6.0475000000000006e-05, "loss": 0.139, "step": 2420 }, { "grad_norm": 1.0906481742858887, "learning_rate": 6.0725e-05, "loss": 0.1053, "step": 2430 }, { "grad_norm": 0.8569415807723999, "learning_rate": 6.0975000000000005e-05, "loss": 0.1047, "step": 2440 }, { "grad_norm": 0.7998971939086914, "learning_rate": 6.1225e-05, "loss": 0.0866, "step": 2450 }, { "grad_norm": 0.6349855065345764, "learning_rate": 6.1475e-05, "loss": 0.0887, "step": 2460 }, { "grad_norm": 0.958854615688324, "learning_rate": 6.1725e-05, "loss": 0.0798, "step": 2470 }, { "grad_norm": 0.807072639465332, "learning_rate": 6.1975e-05, "loss": 0.1058, "step": 2480 }, { "grad_norm": 0.732824444770813, "learning_rate": 6.2225e-05, "loss": 0.1039, "step": 2490 }, { "grad_norm": 0.8124560713768005, "learning_rate": 6.2475e-05, "loss": 0.0872, "step": 2500 }, { "grad_norm": 0.7900496125221252, "learning_rate": 6.2725e-05, "loss": 0.0837, "step": 2510 }, { "grad_norm": 0.8583679795265198, "learning_rate": 6.297500000000001e-05, "loss": 0.0889, "step": 2520 }, { "grad_norm": 0.7322919964790344, "learning_rate": 6.3225e-05, "loss": 0.0821, "step": 2530 }, { "grad_norm": 0.7582690119743347, "learning_rate": 6.347500000000001e-05, "loss": 0.0875, "step": 2540 }, { "grad_norm": 0.7379795908927917, "learning_rate": 6.3725e-05, "loss": 0.0896, "step": 2550 }, { "grad_norm": 0.8569535613059998, "learning_rate": 6.397500000000001e-05, "loss": 0.0831, "step": 2560 }, { "grad_norm": 0.8859396576881409, "learning_rate": 6.4225e-05, "loss": 0.08, "step": 2570 }, { "grad_norm": 0.6922135949134827, "learning_rate": 6.447500000000001e-05, "loss": 0.0819, "step": 2580 }, { "grad_norm": 0.66802978515625, "learning_rate": 6.4725e-05, "loss": 0.0809, "step": 2590 }, { "grad_norm": 0.6784030795097351, "learning_rate": 6.497500000000001e-05, "loss": 0.0879, "step": 2600 }, { "grad_norm": 0.8478695750236511, "learning_rate": 6.5225e-05, "loss": 0.0955, "step": 2610 }, { "grad_norm": 0.6356168389320374, "learning_rate": 6.5475e-05, "loss": 0.0715, "step": 2620 }, { "grad_norm": 0.7908731698989868, "learning_rate": 6.5725e-05, "loss": 0.0787, "step": 2630 }, { "grad_norm": 0.8621340394020081, "learning_rate": 6.5975e-05, "loss": 0.0986, "step": 2640 }, { "grad_norm": 1.0177642107009888, "learning_rate": 6.6225e-05, "loss": 0.0953, "step": 2650 }, { "grad_norm": 0.9772836565971375, "learning_rate": 6.6475e-05, "loss": 0.0962, "step": 2660 }, { "grad_norm": 1.104851245880127, "learning_rate": 6.672500000000001e-05, "loss": 0.1017, "step": 2670 }, { "grad_norm": 0.985365629196167, "learning_rate": 6.6975e-05, "loss": 0.1052, "step": 2680 }, { "grad_norm": 0.617050051689148, "learning_rate": 6.722500000000001e-05, "loss": 0.0834, "step": 2690 }, { "grad_norm": 0.7830052375793457, "learning_rate": 6.7475e-05, "loss": 0.072, "step": 2700 }, { "grad_norm": 0.7070849537849426, "learning_rate": 6.7725e-05, "loss": 0.087, "step": 2710 }, { "grad_norm": 0.761220395565033, "learning_rate": 6.7975e-05, "loss": 0.1146, "step": 2720 }, { "grad_norm": 1.0989171266555786, "learning_rate": 6.8225e-05, "loss": 0.0938, "step": 2730 }, { "grad_norm": 0.8478238582611084, "learning_rate": 6.8475e-05, "loss": 0.0755, "step": 2740 }, { "grad_norm": 0.8601119518280029, "learning_rate": 6.8725e-05, "loss": 0.0921, "step": 2750 }, { "grad_norm": 0.7022044658660889, "learning_rate": 6.8975e-05, "loss": 0.0804, "step": 2760 }, { "grad_norm": 0.8332458138465881, "learning_rate": 6.9225e-05, "loss": 0.0806, "step": 2770 }, { "grad_norm": 0.8674906492233276, "learning_rate": 6.9475e-05, "loss": 0.0959, "step": 2780 }, { "grad_norm": 0.7992690801620483, "learning_rate": 6.9725e-05, "loss": 0.0857, "step": 2790 }, { "grad_norm": 0.728879451751709, "learning_rate": 6.997500000000001e-05, "loss": 0.0914, "step": 2800 }, { "grad_norm": 0.8403785824775696, "learning_rate": 7.022500000000001e-05, "loss": 0.107, "step": 2810 }, { "grad_norm": 0.8630279302597046, "learning_rate": 7.0475e-05, "loss": 0.0993, "step": 2820 }, { "grad_norm": 0.7947811484336853, "learning_rate": 7.072500000000001e-05, "loss": 0.0841, "step": 2830 }, { "grad_norm": 0.6601839065551758, "learning_rate": 7.0975e-05, "loss": 0.1122, "step": 2840 }, { "grad_norm": 0.5235240459442139, "learning_rate": 7.122500000000001e-05, "loss": 0.0982, "step": 2850 }, { "grad_norm": 0.7541103363037109, "learning_rate": 7.1475e-05, "loss": 0.0819, "step": 2860 }, { "grad_norm": 0.7208298444747925, "learning_rate": 7.172500000000001e-05, "loss": 0.0743, "step": 2870 }, { "grad_norm": 0.7513206005096436, "learning_rate": 7.1975e-05, "loss": 0.0724, "step": 2880 }, { "grad_norm": 0.8177627325057983, "learning_rate": 7.2225e-05, "loss": 0.0779, "step": 2890 }, { "grad_norm": 0.789935290813446, "learning_rate": 7.2475e-05, "loss": 0.0855, "step": 2900 }, { "grad_norm": 0.9498433470726013, "learning_rate": 7.272499999999999e-05, "loss": 0.0728, "step": 2910 }, { "grad_norm": 0.6414762139320374, "learning_rate": 7.2975e-05, "loss": 0.0868, "step": 2920 }, { "grad_norm": 0.7908788919448853, "learning_rate": 7.3225e-05, "loss": 0.0663, "step": 2930 }, { "grad_norm": 0.7761130928993225, "learning_rate": 7.347500000000001e-05, "loss": 0.0707, "step": 2940 }, { "grad_norm": 0.8489480018615723, "learning_rate": 7.3725e-05, "loss": 0.0685, "step": 2950 }, { "grad_norm": 0.6192476153373718, "learning_rate": 7.397500000000001e-05, "loss": 0.0942, "step": 2960 }, { "grad_norm": 0.6137591600418091, "learning_rate": 7.4225e-05, "loss": 0.0897, "step": 2970 }, { "grad_norm": 0.6944708824157715, "learning_rate": 7.447500000000001e-05, "loss": 0.0842, "step": 2980 }, { "grad_norm": 0.7840366363525391, "learning_rate": 7.4725e-05, "loss": 0.0945, "step": 2990 }, { "grad_norm": 0.732486367225647, "learning_rate": 7.4975e-05, "loss": 0.096, "step": 3000 }, { "grad_norm": 0.7138053178787231, "learning_rate": 7.5225e-05, "loss": 0.0745, "step": 3010 }, { "grad_norm": 0.6189539432525635, "learning_rate": 7.5475e-05, "loss": 0.0809, "step": 3020 }, { "grad_norm": 0.7038538455963135, "learning_rate": 7.5725e-05, "loss": 0.0759, "step": 3030 }, { "grad_norm": 0.7025724649429321, "learning_rate": 7.5975e-05, "loss": 0.0928, "step": 3040 }, { "grad_norm": 0.7089672088623047, "learning_rate": 7.6225e-05, "loss": 0.0702, "step": 3050 }, { "grad_norm": 0.6425966620445251, "learning_rate": 7.6475e-05, "loss": 0.0779, "step": 3060 }, { "grad_norm": 0.6169239282608032, "learning_rate": 7.672500000000001e-05, "loss": 0.0656, "step": 3070 }, { "grad_norm": 0.6915908455848694, "learning_rate": 7.697500000000001e-05, "loss": 0.0792, "step": 3080 }, { "grad_norm": 0.9682517051696777, "learning_rate": 7.722500000000001e-05, "loss": 0.0965, "step": 3090 }, { "grad_norm": 0.7947483658790588, "learning_rate": 7.747500000000001e-05, "loss": 0.0859, "step": 3100 }, { "grad_norm": 0.5167046785354614, "learning_rate": 7.7725e-05, "loss": 0.0756, "step": 3110 }, { "grad_norm": 0.57271409034729, "learning_rate": 7.797500000000001e-05, "loss": 0.0708, "step": 3120 }, { "grad_norm": 0.5921311974525452, "learning_rate": 7.8225e-05, "loss": 0.0726, "step": 3130 }, { "grad_norm": 0.7732922434806824, "learning_rate": 7.8475e-05, "loss": 0.0817, "step": 3140 }, { "grad_norm": 0.5585314035415649, "learning_rate": 7.8725e-05, "loss": 0.0713, "step": 3150 }, { "grad_norm": 0.7185795307159424, "learning_rate": 7.8975e-05, "loss": 0.0964, "step": 3160 }, { "grad_norm": 0.7507059574127197, "learning_rate": 7.9225e-05, "loss": 0.0756, "step": 3170 }, { "grad_norm": 0.6614155769348145, "learning_rate": 7.9475e-05, "loss": 0.0928, "step": 3180 }, { "grad_norm": 0.7900053262710571, "learning_rate": 7.9725e-05, "loss": 0.0916, "step": 3190 }, { "grad_norm": 0.5823906064033508, "learning_rate": 7.9975e-05, "loss": 0.0802, "step": 3200 }, { "grad_norm": 0.7230386137962341, "learning_rate": 8.022500000000001e-05, "loss": 0.0785, "step": 3210 }, { "grad_norm": 1.201810359954834, "learning_rate": 8.0475e-05, "loss": 0.1027, "step": 3220 }, { "grad_norm": 0.7101589441299438, "learning_rate": 8.072500000000001e-05, "loss": 0.0919, "step": 3230 }, { "grad_norm": 0.403658002614975, "learning_rate": 8.0975e-05, "loss": 0.0721, "step": 3240 }, { "grad_norm": 0.5359125733375549, "learning_rate": 8.122500000000001e-05, "loss": 0.0898, "step": 3250 }, { "grad_norm": 0.8795015215873718, "learning_rate": 8.1475e-05, "loss": 0.0818, "step": 3260 }, { "grad_norm": 0.7432019710540771, "learning_rate": 8.172500000000001e-05, "loss": 0.0803, "step": 3270 }, { "grad_norm": 0.7150573134422302, "learning_rate": 8.1975e-05, "loss": 0.0816, "step": 3280 }, { "grad_norm": 0.679432213306427, "learning_rate": 8.2225e-05, "loss": 0.0707, "step": 3290 }, { "grad_norm": 0.6008446216583252, "learning_rate": 8.2475e-05, "loss": 0.0699, "step": 3300 }, { "grad_norm": 0.8597428798675537, "learning_rate": 8.2725e-05, "loss": 0.0795, "step": 3310 }, { "grad_norm": 0.6012914180755615, "learning_rate": 8.2975e-05, "loss": 0.0944, "step": 3320 }, { "grad_norm": 0.8057188987731934, "learning_rate": 8.3225e-05, "loss": 0.0705, "step": 3330 }, { "grad_norm": 0.7304126620292664, "learning_rate": 8.347500000000001e-05, "loss": 0.0763, "step": 3340 }, { "grad_norm": 0.7468079924583435, "learning_rate": 8.3725e-05, "loss": 0.0755, "step": 3350 }, { "grad_norm": 0.9952413439750671, "learning_rate": 8.397500000000001e-05, "loss": 0.066, "step": 3360 }, { "grad_norm": 0.6693013906478882, "learning_rate": 8.422500000000001e-05, "loss": 0.075, "step": 3370 }, { "grad_norm": 0.5827971696853638, "learning_rate": 8.447500000000001e-05, "loss": 0.1159, "step": 3380 }, { "grad_norm": 0.7804464101791382, "learning_rate": 8.4725e-05, "loss": 0.0701, "step": 3390 }, { "grad_norm": 0.6248950958251953, "learning_rate": 8.4975e-05, "loss": 0.0846, "step": 3400 }, { "grad_norm": 0.632793128490448, "learning_rate": 8.5225e-05, "loss": 0.0719, "step": 3410 }, { "grad_norm": 0.7789002060890198, "learning_rate": 8.5475e-05, "loss": 0.0704, "step": 3420 }, { "grad_norm": 0.6635804772377014, "learning_rate": 8.5725e-05, "loss": 0.0739, "step": 3430 }, { "grad_norm": 0.5102415680885315, "learning_rate": 8.5975e-05, "loss": 0.0857, "step": 3440 }, { "grad_norm": 0.5680745840072632, "learning_rate": 8.6225e-05, "loss": 0.074, "step": 3450 }, { "grad_norm": 0.5044145584106445, "learning_rate": 8.6475e-05, "loss": 0.0592, "step": 3460 }, { "grad_norm": 0.5930052399635315, "learning_rate": 8.672500000000001e-05, "loss": 0.082, "step": 3470 }, { "grad_norm": 0.6222184300422668, "learning_rate": 8.6975e-05, "loss": 0.081, "step": 3480 }, { "grad_norm": 0.5692686438560486, "learning_rate": 8.7225e-05, "loss": 0.0792, "step": 3490 }, { "grad_norm": 0.6030033826828003, "learning_rate": 8.747500000000001e-05, "loss": 0.0598, "step": 3500 }, { "grad_norm": 0.6913685202598572, "learning_rate": 8.7725e-05, "loss": 0.0727, "step": 3510 }, { "grad_norm": 0.7820367813110352, "learning_rate": 8.797500000000001e-05, "loss": 0.0776, "step": 3520 }, { "grad_norm": 0.586893618106842, "learning_rate": 8.8225e-05, "loss": 0.0742, "step": 3530 }, { "grad_norm": 0.708417534828186, "learning_rate": 8.847500000000001e-05, "loss": 0.0707, "step": 3540 }, { "grad_norm": 0.9352200627326965, "learning_rate": 8.8725e-05, "loss": 0.0892, "step": 3550 }, { "grad_norm": 0.739625096321106, "learning_rate": 8.897500000000001e-05, "loss": 0.0678, "step": 3560 }, { "grad_norm": 0.7439042329788208, "learning_rate": 8.9225e-05, "loss": 0.0724, "step": 3570 }, { "grad_norm": 0.8078579902648926, "learning_rate": 8.9475e-05, "loss": 0.0964, "step": 3580 }, { "grad_norm": 0.6100075244903564, "learning_rate": 8.9725e-05, "loss": 0.0884, "step": 3590 }, { "grad_norm": 0.5476158857345581, "learning_rate": 8.9975e-05, "loss": 0.0776, "step": 3600 }, { "grad_norm": 0.6785266995429993, "learning_rate": 9.0225e-05, "loss": 0.0818, "step": 3610 }, { "grad_norm": 0.7139505743980408, "learning_rate": 9.0475e-05, "loss": 0.0717, "step": 3620 }, { "grad_norm": 0.6407026648521423, "learning_rate": 9.072500000000001e-05, "loss": 0.0654, "step": 3630 }, { "grad_norm": 0.869198739528656, "learning_rate": 9.0975e-05, "loss": 0.0722, "step": 3640 }, { "grad_norm": 0.6660171747207642, "learning_rate": 9.122500000000001e-05, "loss": 0.0819, "step": 3650 }, { "grad_norm": 0.467261403799057, "learning_rate": 9.1475e-05, "loss": 0.0838, "step": 3660 }, { "grad_norm": 0.7115690112113953, "learning_rate": 9.172500000000001e-05, "loss": 0.0787, "step": 3670 }, { "grad_norm": 0.6593378782272339, "learning_rate": 9.1975e-05, "loss": 0.0835, "step": 3680 }, { "grad_norm": 0.6332399845123291, "learning_rate": 9.2225e-05, "loss": 0.0741, "step": 3690 }, { "grad_norm": 0.6476573348045349, "learning_rate": 9.2475e-05, "loss": 0.073, "step": 3700 }, { "grad_norm": 0.6917386054992676, "learning_rate": 9.2725e-05, "loss": 0.0742, "step": 3710 }, { "grad_norm": 0.5005589723587036, "learning_rate": 9.2975e-05, "loss": 0.0744, "step": 3720 }, { "grad_norm": 0.4901551902294159, "learning_rate": 9.3225e-05, "loss": 0.0607, "step": 3730 }, { "grad_norm": 0.5880137085914612, "learning_rate": 9.3475e-05, "loss": 0.0576, "step": 3740 }, { "grad_norm": 0.49788233637809753, "learning_rate": 9.3725e-05, "loss": 0.0797, "step": 3750 }, { "grad_norm": 0.7581449151039124, "learning_rate": 9.397500000000001e-05, "loss": 0.0668, "step": 3760 }, { "grad_norm": 0.6817517280578613, "learning_rate": 9.422500000000001e-05, "loss": 0.0838, "step": 3770 }, { "grad_norm": 0.8310948610305786, "learning_rate": 9.4475e-05, "loss": 0.074, "step": 3780 }, { "grad_norm": 0.6465156078338623, "learning_rate": 9.472500000000001e-05, "loss": 0.0801, "step": 3790 }, { "grad_norm": 0.4869101047515869, "learning_rate": 9.4975e-05, "loss": 0.0944, "step": 3800 }, { "grad_norm": 0.7236448526382446, "learning_rate": 9.522500000000001e-05, "loss": 0.0671, "step": 3810 }, { "grad_norm": 0.7121617197990417, "learning_rate": 9.5475e-05, "loss": 0.0935, "step": 3820 }, { "grad_norm": 0.6515911221504211, "learning_rate": 9.572500000000001e-05, "loss": 0.0817, "step": 3830 }, { "grad_norm": 0.6217439770698547, "learning_rate": 9.5975e-05, "loss": 0.075, "step": 3840 }, { "grad_norm": 0.9135004281997681, "learning_rate": 9.622500000000001e-05, "loss": 0.1006, "step": 3850 }, { "grad_norm": 0.6868797540664673, "learning_rate": 9.6475e-05, "loss": 0.0777, "step": 3860 }, { "grad_norm": 0.6351804137229919, "learning_rate": 9.6725e-05, "loss": 0.0853, "step": 3870 }, { "grad_norm": 0.6315440535545349, "learning_rate": 9.6975e-05, "loss": 0.0648, "step": 3880 }, { "grad_norm": 0.5314168930053711, "learning_rate": 9.7225e-05, "loss": 0.0885, "step": 3890 }, { "grad_norm": 0.717968225479126, "learning_rate": 9.747500000000001e-05, "loss": 0.0858, "step": 3900 }, { "grad_norm": 3.9156835079193115, "learning_rate": 9.7725e-05, "loss": 0.1102, "step": 3910 }, { "grad_norm": 0.9179057478904724, "learning_rate": 9.797500000000001e-05, "loss": 0.1254, "step": 3920 }, { "grad_norm": 0.8874500393867493, "learning_rate": 9.8225e-05, "loss": 0.1125, "step": 3930 }, { "grad_norm": 0.742614209651947, "learning_rate": 9.847500000000001e-05, "loss": 0.1127, "step": 3940 }, { "grad_norm": 0.9695990681648254, "learning_rate": 9.8725e-05, "loss": 0.0908, "step": 3950 }, { "grad_norm": 0.7198958396911621, "learning_rate": 9.897500000000001e-05, "loss": 0.0957, "step": 3960 }, { "grad_norm": 1.0188413858413696, "learning_rate": 9.9225e-05, "loss": 0.0756, "step": 3970 }, { "grad_norm": 0.792401134967804, "learning_rate": 9.9475e-05, "loss": 0.0962, "step": 3980 }, { "grad_norm": 0.7378698587417603, "learning_rate": 9.9725e-05, "loss": 0.094, "step": 3990 }, { "grad_norm": 0.6006528735160828, "learning_rate": 9.9975e-05, "loss": 0.0834, "step": 4000 }, { "grad_norm": 1.0248891115188599, "learning_rate": 9.999999653982884e-05, "loss": 0.0882, "step": 4010 }, { "grad_norm": 0.5673749446868896, "learning_rate": 9.999998457874392e-05, "loss": 0.0869, "step": 4020 }, { "grad_norm": 0.7280023097991943, "learning_rate": 9.999996407402913e-05, "loss": 0.0815, "step": 4030 }, { "grad_norm": 0.6349278688430786, "learning_rate": 9.999993502568801e-05, "loss": 0.0671, "step": 4040 }, { "grad_norm": 0.49280697107315063, "learning_rate": 9.999989743372548e-05, "loss": 0.0742, "step": 4050 }, { "grad_norm": 0.8093444108963013, "learning_rate": 9.999985129814798e-05, "loss": 0.0748, "step": 4060 }, { "grad_norm": 0.7473916411399841, "learning_rate": 9.99997966189634e-05, "loss": 0.065, "step": 4070 }, { "grad_norm": 0.7794133424758911, "learning_rate": 9.999973339618107e-05, "loss": 0.0855, "step": 4080 }, { "grad_norm": 0.5915987491607666, "learning_rate": 9.999966162981179e-05, "loss": 0.0737, "step": 4090 }, { "grad_norm": 0.49094364047050476, "learning_rate": 9.999958131986784e-05, "loss": 0.0728, "step": 4100 }, { "grad_norm": 0.576235294342041, "learning_rate": 9.999949246636293e-05, "loss": 0.072, "step": 4110 }, { "grad_norm": 0.7043496966362, "learning_rate": 9.999939506931224e-05, "loss": 0.0811, "step": 4120 }, { "grad_norm": 0.5554373264312744, "learning_rate": 9.999928912873243e-05, "loss": 0.0651, "step": 4130 }, { "grad_norm": 0.6200869083404541, "learning_rate": 9.999917464464159e-05, "loss": 0.0661, "step": 4140 }, { "grad_norm": 0.7626453042030334, "learning_rate": 9.999905161705929e-05, "loss": 0.0832, "step": 4150 }, { "grad_norm": 0.5470541715621948, "learning_rate": 9.999892004600653e-05, "loss": 0.0715, "step": 4160 }, { "grad_norm": 0.5708557963371277, "learning_rate": 9.999877993150581e-05, "loss": 0.0671, "step": 4170 }, { "grad_norm": 0.5965170860290527, "learning_rate": 9.999863127358108e-05, "loss": 0.0657, "step": 4180 }, { "grad_norm": 0.7661723494529724, "learning_rate": 9.999847407225773e-05, "loss": 0.064, "step": 4190 }, { "grad_norm": 0.7276044487953186, "learning_rate": 9.999830832756262e-05, "loss": 0.0783, "step": 4200 }, { "grad_norm": 0.42511454224586487, "learning_rate": 9.999813403952407e-05, "loss": 0.069, "step": 4210 }, { "grad_norm": 0.39360252022743225, "learning_rate": 9.999795120817187e-05, "loss": 0.0755, "step": 4220 }, { "grad_norm": 0.6759563684463501, "learning_rate": 9.999775983353725e-05, "loss": 0.0876, "step": 4230 }, { "grad_norm": 0.4542551338672638, "learning_rate": 9.999755991565292e-05, "loss": 0.0779, "step": 4240 }, { "grad_norm": 0.5490238070487976, "learning_rate": 9.999735145455303e-05, "loss": 0.0791, "step": 4250 }, { "grad_norm": 0.5346692800521851, "learning_rate": 9.99971344502732e-05, "loss": 0.0844, "step": 4260 }, { "grad_norm": 0.5317337512969971, "learning_rate": 9.999690890285053e-05, "loss": 0.075, "step": 4270 }, { "grad_norm": 0.7594810128211975, "learning_rate": 9.999667481232356e-05, "loss": 0.0708, "step": 4280 }, { "grad_norm": 0.686241090297699, "learning_rate": 9.999643217873225e-05, "loss": 0.0673, "step": 4290 }, { "grad_norm": 0.5513443350791931, "learning_rate": 9.999618100211809e-05, "loss": 0.0646, "step": 4300 }, { "grad_norm": 0.4853488504886627, "learning_rate": 9.999592128252402e-05, "loss": 0.0635, "step": 4310 }, { "grad_norm": 0.7061753273010254, "learning_rate": 9.999565301999437e-05, "loss": 0.0985, "step": 4320 }, { "grad_norm": 0.8197558522224426, "learning_rate": 9.999537621457502e-05, "loss": 0.0667, "step": 4330 }, { "grad_norm": 0.573205292224884, "learning_rate": 9.999509086631323e-05, "loss": 0.0593, "step": 4340 }, { "grad_norm": 0.7979260683059692, "learning_rate": 9.99947969752578e-05, "loss": 0.062, "step": 4350 }, { "grad_norm": 0.6869593262672424, "learning_rate": 9.999449454145891e-05, "loss": 0.0647, "step": 4360 }, { "grad_norm": 0.5565863251686096, "learning_rate": 9.999418356496827e-05, "loss": 0.0854, "step": 4370 }, { "grad_norm": 0.5896948575973511, "learning_rate": 9.999386404583899e-05, "loss": 0.0682, "step": 4380 }, { "grad_norm": 0.5952700972557068, "learning_rate": 9.999353598412568e-05, "loss": 0.0601, "step": 4390 }, { "grad_norm": 0.3789099454879761, "learning_rate": 9.999319937988442e-05, "loss": 0.0818, "step": 4400 }, { "grad_norm": 0.6105528473854065, "learning_rate": 9.999285423317268e-05, "loss": 0.061, "step": 4410 }, { "grad_norm": 0.680875301361084, "learning_rate": 9.999250054404947e-05, "loss": 0.0923, "step": 4420 }, { "grad_norm": 0.3836138844490051, "learning_rate": 9.99921383125752e-05, "loss": 0.0704, "step": 4430 }, { "grad_norm": 0.6301745772361755, "learning_rate": 9.99917675388118e-05, "loss": 0.0768, "step": 4440 }, { "grad_norm": 0.6740859150886536, "learning_rate": 9.99913882228226e-05, "loss": 0.082, "step": 4450 }, { "grad_norm": 0.6359063982963562, "learning_rate": 9.999100036467242e-05, "loss": 0.074, "step": 4460 }, { "grad_norm": 0.6096593737602234, "learning_rate": 9.999060396442753e-05, "loss": 0.063, "step": 4470 }, { "grad_norm": 0.5031774044036865, "learning_rate": 9.999019902215566e-05, "loss": 0.0692, "step": 4480 }, { "grad_norm": 0.6508592963218689, "learning_rate": 9.998978553792602e-05, "loss": 0.0803, "step": 4490 }, { "grad_norm": 0.6366022825241089, "learning_rate": 9.998936351180926e-05, "loss": 0.0728, "step": 4500 }, { "grad_norm": 0.4594060778617859, "learning_rate": 9.998893294387747e-05, "loss": 0.0969, "step": 4510 }, { "grad_norm": 0.6833745241165161, "learning_rate": 9.998849383420426e-05, "loss": 0.0868, "step": 4520 }, { "grad_norm": 0.5334557890892029, "learning_rate": 9.998804618286465e-05, "loss": 0.0898, "step": 4530 }, { "grad_norm": 0.7780174612998962, "learning_rate": 9.99875899899351e-05, "loss": 0.0951, "step": 4540 }, { "grad_norm": 0.5320178270339966, "learning_rate": 9.99871252554936e-05, "loss": 0.0854, "step": 4550 }, { "grad_norm": 0.4552043378353119, "learning_rate": 9.998665197961955e-05, "loss": 0.0739, "step": 4560 }, { "grad_norm": 0.47099238634109497, "learning_rate": 9.998617016239379e-05, "loss": 0.0778, "step": 4570 }, { "grad_norm": 0.7682055830955505, "learning_rate": 9.998567980389869e-05, "loss": 0.0777, "step": 4580 }, { "grad_norm": 0.5629297494888306, "learning_rate": 9.998518090421802e-05, "loss": 0.0687, "step": 4590 }, { "grad_norm": 0.47523990273475647, "learning_rate": 9.998467346343703e-05, "loss": 0.0691, "step": 4600 }, { "grad_norm": 0.7396674752235413, "learning_rate": 9.998415748164243e-05, "loss": 0.0702, "step": 4610 }, { "grad_norm": 0.614080011844635, "learning_rate": 9.998363295892238e-05, "loss": 0.0554, "step": 4620 }, { "grad_norm": 0.9786016941070557, "learning_rate": 9.998309989536652e-05, "loss": 0.0739, "step": 4630 }, { "grad_norm": 0.5116230249404907, "learning_rate": 9.998255829106593e-05, "loss": 0.0783, "step": 4640 }, { "grad_norm": 0.5494259595870972, "learning_rate": 9.998200814611316e-05, "loss": 0.0637, "step": 4650 }, { "grad_norm": 0.5533872246742249, "learning_rate": 9.998144946060219e-05, "loss": 0.0659, "step": 4660 }, { "grad_norm": 0.8109402060508728, "learning_rate": 9.998088223462852e-05, "loss": 0.0611, "step": 4670 }, { "grad_norm": 0.40653127431869507, "learning_rate": 9.998030646828905e-05, "loss": 0.0746, "step": 4680 }, { "grad_norm": 0.7120426297187805, "learning_rate": 9.997972216168217e-05, "loss": 0.0857, "step": 4690 }, { "grad_norm": 0.7006242871284485, "learning_rate": 9.997912931490771e-05, "loss": 0.0671, "step": 4700 }, { "grad_norm": 0.5439690351486206, "learning_rate": 9.9978527928067e-05, "loss": 0.0672, "step": 4710 }, { "grad_norm": 0.5449845790863037, "learning_rate": 9.997791800126277e-05, "loss": 0.0688, "step": 4720 }, { "grad_norm": 0.3178025186061859, "learning_rate": 9.997729953459927e-05, "loss": 0.084, "step": 4730 }, { "grad_norm": 0.4415518343448639, "learning_rate": 9.997667252818214e-05, "loss": 0.0717, "step": 4740 }, { "grad_norm": 0.712713897228241, "learning_rate": 9.997603698211855e-05, "loss": 0.0711, "step": 4750 }, { "grad_norm": 0.6199305653572083, "learning_rate": 9.99753928965171e-05, "loss": 0.0962, "step": 4760 }, { "grad_norm": 0.6169033646583557, "learning_rate": 9.997474027148781e-05, "loss": 0.1015, "step": 4770 }, { "grad_norm": 0.6185609698295593, "learning_rate": 9.997407910714223e-05, "loss": 0.0884, "step": 4780 }, { "grad_norm": 0.7355145215988159, "learning_rate": 9.997340940359332e-05, "loss": 0.0856, "step": 4790 }, { "grad_norm": 0.6608867049217224, "learning_rate": 9.997273116095552e-05, "loss": 0.074, "step": 4800 }, { "grad_norm": 0.9267734885215759, "learning_rate": 9.997204437934473e-05, "loss": 0.0949, "step": 4810 }, { "grad_norm": 0.7607278227806091, "learning_rate": 9.997134905887829e-05, "loss": 0.0852, "step": 4820 }, { "grad_norm": 0.5659125447273254, "learning_rate": 9.997064519967501e-05, "loss": 0.0811, "step": 4830 }, { "grad_norm": 0.5706393718719482, "learning_rate": 9.996993280185517e-05, "loss": 0.0834, "step": 4840 }, { "grad_norm": 0.5261604189872742, "learning_rate": 9.99692118655405e-05, "loss": 0.0804, "step": 4850 }, { "grad_norm": 0.5525628328323364, "learning_rate": 9.996848239085417e-05, "loss": 0.0817, "step": 4860 }, { "grad_norm": 0.5206994414329529, "learning_rate": 9.996774437792085e-05, "loss": 0.0786, "step": 4870 }, { "grad_norm": 0.7958998084068298, "learning_rate": 9.996699782686664e-05, "loss": 0.072, "step": 4880 }, { "grad_norm": 0.6589877605438232, "learning_rate": 9.996624273781909e-05, "loss": 0.0905, "step": 4890 }, { "grad_norm": 0.7029542326927185, "learning_rate": 9.996547911090725e-05, "loss": 0.0623, "step": 4900 }, { "grad_norm": 0.7353554368019104, "learning_rate": 9.996470694626157e-05, "loss": 0.0698, "step": 4910 }, { "grad_norm": 0.5048136115074158, "learning_rate": 9.996392624401403e-05, "loss": 0.0746, "step": 4920 }, { "grad_norm": 0.556999921798706, "learning_rate": 9.996313700429801e-05, "loss": 0.0626, "step": 4930 }, { "grad_norm": 0.4767974019050598, "learning_rate": 9.996233922724836e-05, "loss": 0.0581, "step": 4940 }, { "grad_norm": 0.5628023147583008, "learning_rate": 9.996153291300141e-05, "loss": 0.0693, "step": 4950 }, { "grad_norm": 0.6871451139450073, "learning_rate": 9.996071806169494e-05, "loss": 0.0606, "step": 4960 }, { "grad_norm": 0.664116621017456, "learning_rate": 9.995989467346817e-05, "loss": 0.0876, "step": 4970 }, { "grad_norm": 0.5804128646850586, "learning_rate": 9.995906274846183e-05, "loss": 0.0678, "step": 4980 }, { "grad_norm": 0.870029628276825, "learning_rate": 9.995822228681803e-05, "loss": 0.0694, "step": 4990 }, { "grad_norm": 0.43633338809013367, "learning_rate": 9.99573732886804e-05, "loss": 0.0816, "step": 5000 }, { "grad_norm": 0.5549604296684265, "learning_rate": 9.995651575419402e-05, "loss": 0.071, "step": 5010 }, { "grad_norm": 0.601064920425415, "learning_rate": 9.995564968350541e-05, "loss": 0.0627, "step": 5020 }, { "grad_norm": 0.6065890789031982, "learning_rate": 9.995477507676256e-05, "loss": 0.0607, "step": 5030 }, { "grad_norm": 0.5710084438323975, "learning_rate": 9.995389193411493e-05, "loss": 0.0636, "step": 5040 }, { "grad_norm": 0.46493080258369446, "learning_rate": 9.995300025571339e-05, "loss": 0.0673, "step": 5050 }, { "grad_norm": 0.5083101391792297, "learning_rate": 9.995210004171034e-05, "loss": 0.0545, "step": 5060 }, { "grad_norm": 0.5628308057785034, "learning_rate": 9.995119129225956e-05, "loss": 0.0705, "step": 5070 }, { "grad_norm": 0.6967461109161377, "learning_rate": 9.995027400751637e-05, "loss": 0.0759, "step": 5080 }, { "grad_norm": 0.6950468420982361, "learning_rate": 9.994934818763751e-05, "loss": 0.0662, "step": 5090 }, { "grad_norm": 0.4630851745605469, "learning_rate": 9.994841383278115e-05, "loss": 0.0683, "step": 5100 }, { "grad_norm": 0.6506832242012024, "learning_rate": 9.994747094310695e-05, "loss": 0.0742, "step": 5110 }, { "grad_norm": 0.5173658132553101, "learning_rate": 9.994651951877604e-05, "loss": 0.0893, "step": 5120 }, { "grad_norm": 0.5590020418167114, "learning_rate": 9.994555955995099e-05, "loss": 0.0814, "step": 5130 }, { "grad_norm": 0.6046057939529419, "learning_rate": 9.994459106679581e-05, "loss": 0.0716, "step": 5140 }, { "grad_norm": 0.7607628703117371, "learning_rate": 9.994361403947603e-05, "loss": 0.0662, "step": 5150 }, { "grad_norm": 0.5445544719696045, "learning_rate": 9.994262847815854e-05, "loss": 0.0774, "step": 5160 }, { "grad_norm": 0.6890463829040527, "learning_rate": 9.99416343830118e-05, "loss": 0.0738, "step": 5170 }, { "grad_norm": 0.7416422963142395, "learning_rate": 9.994063175420565e-05, "loss": 0.0665, "step": 5180 }, { "grad_norm": 0.6012803316116333, "learning_rate": 9.99396205919114e-05, "loss": 0.0965, "step": 5190 }, { "grad_norm": 0.5797383785247803, "learning_rate": 9.993860089630185e-05, "loss": 0.0587, "step": 5200 }, { "grad_norm": 0.45979559421539307, "learning_rate": 9.993757266755123e-05, "loss": 0.068, "step": 5210 }, { "grad_norm": 0.4200184941291809, "learning_rate": 9.993653590583522e-05, "loss": 0.0683, "step": 5220 }, { "grad_norm": 0.4681931734085083, "learning_rate": 9.993549061133102e-05, "loss": 0.0564, "step": 5230 }, { "grad_norm": 0.6981790661811829, "learning_rate": 9.993443678421719e-05, "loss": 0.0697, "step": 5240 }, { "grad_norm": 0.6591899394989014, "learning_rate": 9.993337442467384e-05, "loss": 0.0752, "step": 5250 }, { "grad_norm": 0.6767861843109131, "learning_rate": 9.993230353288248e-05, "loss": 0.0597, "step": 5260 }, { "grad_norm": 0.568640947341919, "learning_rate": 9.993122410902608e-05, "loss": 0.0682, "step": 5270 }, { "grad_norm": 0.7028656005859375, "learning_rate": 9.993013615328912e-05, "loss": 0.0666, "step": 5280 }, { "grad_norm": 0.4125896394252777, "learning_rate": 9.992903966585747e-05, "loss": 0.0643, "step": 5290 }, { "grad_norm": 0.4728047549724579, "learning_rate": 9.992793464691852e-05, "loss": 0.0656, "step": 5300 }, { "grad_norm": 0.6500008702278137, "learning_rate": 9.992682109666105e-05, "loss": 0.0677, "step": 5310 }, { "grad_norm": 0.5343496203422546, "learning_rate": 9.992569901527538e-05, "loss": 0.0698, "step": 5320 }, { "grad_norm": 0.6417036056518555, "learning_rate": 9.99245684029532e-05, "loss": 0.1025, "step": 5330 }, { "grad_norm": 0.5442426204681396, "learning_rate": 9.992342925988774e-05, "loss": 0.0659, "step": 5340 }, { "grad_norm": 0.5677492022514343, "learning_rate": 9.992228158627361e-05, "loss": 0.0649, "step": 5350 }, { "grad_norm": 0.4624585509300232, "learning_rate": 9.992112538230693e-05, "loss": 0.0762, "step": 5360 }, { "grad_norm": 0.45087602734565735, "learning_rate": 9.991996064818527e-05, "loss": 0.0625, "step": 5370 }, { "grad_norm": 0.5169437527656555, "learning_rate": 9.991878738410768e-05, "loss": 0.0536, "step": 5380 }, { "grad_norm": 0.5138689279556274, "learning_rate": 9.991760559027457e-05, "loss": 0.0623, "step": 5390 }, { "grad_norm": 0.7790486216545105, "learning_rate": 9.991641526688793e-05, "loss": 0.0807, "step": 5400 }, { "grad_norm": 0.46144968271255493, "learning_rate": 9.991521641415113e-05, "loss": 0.0667, "step": 5410 }, { "grad_norm": 0.5062205791473389, "learning_rate": 9.991400903226904e-05, "loss": 0.0604, "step": 5420 }, { "grad_norm": 0.6013466715812683, "learning_rate": 9.991279312144794e-05, "loss": 0.0632, "step": 5430 }, { "grad_norm": 0.7092994451522827, "learning_rate": 9.991156868189564e-05, "loss": 0.0753, "step": 5440 }, { "grad_norm": 0.8258692622184753, "learning_rate": 9.991033571382131e-05, "loss": 0.0552, "step": 5450 }, { "grad_norm": 0.5638810992240906, "learning_rate": 9.990909421743569e-05, "loss": 0.0621, "step": 5460 }, { "grad_norm": 0.6555312871932983, "learning_rate": 9.990784419295085e-05, "loss": 0.0602, "step": 5470 }, { "grad_norm": 0.4908044636249542, "learning_rate": 9.990658564058044e-05, "loss": 0.0597, "step": 5480 }, { "grad_norm": 0.47312527894973755, "learning_rate": 9.990531856053948e-05, "loss": 0.0586, "step": 5490 }, { "grad_norm": 0.47652116417884827, "learning_rate": 9.99040429530445e-05, "loss": 0.0615, "step": 5500 }, { "grad_norm": 0.5951025485992432, "learning_rate": 9.990275881831346e-05, "loss": 0.0623, "step": 5510 }, { "grad_norm": 0.5214774012565613, "learning_rate": 9.990146615656577e-05, "loss": 0.0563, "step": 5520 }, { "grad_norm": 0.4498390555381775, "learning_rate": 9.990016496802233e-05, "loss": 0.0572, "step": 5530 }, { "grad_norm": 0.5241314172744751, "learning_rate": 9.989885525290548e-05, "loss": 0.0482, "step": 5540 }, { "grad_norm": 0.3714304268360138, "learning_rate": 9.989753701143897e-05, "loss": 0.0518, "step": 5550 }, { "grad_norm": 0.49587172269821167, "learning_rate": 9.989621024384812e-05, "loss": 0.0561, "step": 5560 }, { "grad_norm": 0.5828692317008972, "learning_rate": 9.989487495035959e-05, "loss": 0.0578, "step": 5570 }, { "grad_norm": 0.5852890014648438, "learning_rate": 9.989353113120156e-05, "loss": 0.067, "step": 5580 }, { "grad_norm": 0.5432547330856323, "learning_rate": 9.989217878660366e-05, "loss": 0.0584, "step": 5590 }, { "grad_norm": 0.41558754444122314, "learning_rate": 9.989081791679695e-05, "loss": 0.067, "step": 5600 }, { "grad_norm": 0.6911340355873108, "learning_rate": 9.988944852201397e-05, "loss": 0.0598, "step": 5610 }, { "grad_norm": 0.42689064145088196, "learning_rate": 9.988807060248873e-05, "loss": 0.0584, "step": 5620 }, { "grad_norm": 0.4556451141834259, "learning_rate": 9.988668415845665e-05, "loss": 0.0524, "step": 5630 }, { "grad_norm": 0.6454123258590698, "learning_rate": 9.988528919015466e-05, "loss": 0.0644, "step": 5640 }, { "grad_norm": 0.4596400558948517, "learning_rate": 9.988388569782112e-05, "loss": 0.0691, "step": 5650 }, { "grad_norm": 0.45600426197052, "learning_rate": 9.988247368169583e-05, "loss": 0.0606, "step": 5660 }, { "grad_norm": 0.6440150141716003, "learning_rate": 9.988105314202007e-05, "loss": 0.0644, "step": 5670 }, { "grad_norm": 0.5486634969711304, "learning_rate": 9.987962407903659e-05, "loss": 0.0734, "step": 5680 }, { "grad_norm": 0.9155829548835754, "learning_rate": 9.987818649298957e-05, "loss": 0.0657, "step": 5690 }, { "grad_norm": 0.5129969120025635, "learning_rate": 9.987674038412465e-05, "loss": 0.0628, "step": 5700 }, { "grad_norm": 0.5093467831611633, "learning_rate": 9.987528575268891e-05, "loss": 0.0695, "step": 5710 }, { "grad_norm": 0.587615430355072, "learning_rate": 9.987382259893095e-05, "loss": 0.0573, "step": 5720 }, { "grad_norm": 0.6208682060241699, "learning_rate": 9.987235092310074e-05, "loss": 0.0583, "step": 5730 }, { "grad_norm": 0.40711164474487305, "learning_rate": 9.987087072544978e-05, "loss": 0.0668, "step": 5740 }, { "grad_norm": 0.5072969794273376, "learning_rate": 9.9869382006231e-05, "loss": 0.0579, "step": 5750 }, { "grad_norm": 0.5422557592391968, "learning_rate": 9.986788476569875e-05, "loss": 0.0663, "step": 5760 }, { "grad_norm": 0.5218154191970825, "learning_rate": 9.986637900410887e-05, "loss": 0.0607, "step": 5770 }, { "grad_norm": 0.5217679142951965, "learning_rate": 9.986486472171869e-05, "loss": 0.0688, "step": 5780 }, { "grad_norm": 0.5795953273773193, "learning_rate": 9.986334191878692e-05, "loss": 0.0571, "step": 5790 }, { "grad_norm": 0.5540916919708252, "learning_rate": 9.986181059557378e-05, "loss": 0.0761, "step": 5800 }, { "grad_norm": 0.5238950848579407, "learning_rate": 9.986027075234094e-05, "loss": 0.0619, "step": 5810 }, { "grad_norm": 0.3660737872123718, "learning_rate": 9.985872238935152e-05, "loss": 0.0616, "step": 5820 }, { "grad_norm": 0.5416193008422852, "learning_rate": 9.985716550687008e-05, "loss": 0.0608, "step": 5830 }, { "grad_norm": 0.39272117614746094, "learning_rate": 9.985560010516264e-05, "loss": 0.0629, "step": 5840 }, { "grad_norm": 0.5179275870323181, "learning_rate": 9.985402618449668e-05, "loss": 0.0527, "step": 5850 }, { "grad_norm": 0.37788766622543335, "learning_rate": 9.985244374514118e-05, "loss": 0.0683, "step": 5860 }, { "grad_norm": 0.4140208661556244, "learning_rate": 9.985085278736651e-05, "loss": 0.0678, "step": 5870 }, { "grad_norm": 0.8976212739944458, "learning_rate": 9.984925331144452e-05, "loss": 0.0747, "step": 5880 }, { "grad_norm": 0.5638300180435181, "learning_rate": 9.984764531764851e-05, "loss": 0.0694, "step": 5890 }, { "grad_norm": 0.9004281163215637, "learning_rate": 9.984602880625326e-05, "loss": 0.0896, "step": 5900 }, { "grad_norm": 0.566127598285675, "learning_rate": 9.9844403777535e-05, "loss": 0.0844, "step": 5910 }, { "grad_norm": 0.800446093082428, "learning_rate": 9.984277023177135e-05, "loss": 0.0643, "step": 5920 }, { "grad_norm": 0.4000251293182373, "learning_rate": 9.984112816924148e-05, "loss": 0.0623, "step": 5930 }, { "grad_norm": 0.4919607639312744, "learning_rate": 9.983947759022596e-05, "loss": 0.0625, "step": 5940 }, { "grad_norm": 0.5099521279335022, "learning_rate": 9.983781849500682e-05, "loss": 0.0635, "step": 5950 }, { "grad_norm": 0.42212823033332825, "learning_rate": 9.98361508838676e-05, "loss": 0.0583, "step": 5960 }, { "grad_norm": 0.8947172164916992, "learning_rate": 9.98344747570932e-05, "loss": 0.0775, "step": 5970 }, { "grad_norm": 0.7097366452217102, "learning_rate": 9.983279011497004e-05, "loss": 0.0791, "step": 5980 }, { "grad_norm": 0.43449267745018005, "learning_rate": 9.983109695778596e-05, "loss": 0.0856, "step": 5990 }, { "grad_norm": 0.5329252481460571, "learning_rate": 9.982939528583032e-05, "loss": 0.0899, "step": 6000 }, { "grad_norm": 0.5892184376716614, "learning_rate": 9.982768509939385e-05, "loss": 0.0595, "step": 6010 }, { "grad_norm": 0.4441240727901459, "learning_rate": 9.982596639876879e-05, "loss": 0.0581, "step": 6020 }, { "grad_norm": 0.4799131155014038, "learning_rate": 9.982423918424881e-05, "loss": 0.0534, "step": 6030 }, { "grad_norm": 0.6126622557640076, "learning_rate": 9.982250345612908e-05, "loss": 0.0499, "step": 6040 }, { "grad_norm": 0.6552585363388062, "learning_rate": 9.982075921470611e-05, "loss": 0.0705, "step": 6050 }, { "grad_norm": 0.6091848015785217, "learning_rate": 9.981900646027802e-05, "loss": 0.0517, "step": 6060 }, { "grad_norm": 0.5308831930160522, "learning_rate": 9.981724519314425e-05, "loss": 0.057, "step": 6070 }, { "grad_norm": 0.5504332184791565, "learning_rate": 9.981547541360581e-05, "loss": 0.0605, "step": 6080 }, { "grad_norm": 0.4889134168624878, "learning_rate": 9.981369712196508e-05, "loss": 0.0693, "step": 6090 }, { "grad_norm": 0.5455299019813538, "learning_rate": 9.981191031852592e-05, "loss": 0.0699, "step": 6100 }, { "grad_norm": 0.5107730031013489, "learning_rate": 9.981011500359362e-05, "loss": 0.0527, "step": 6110 }, { "grad_norm": 0.5554317235946655, "learning_rate": 9.9808311177475e-05, "loss": 0.0685, "step": 6120 }, { "grad_norm": 0.5622275471687317, "learning_rate": 9.980649884047826e-05, "loss": 0.0572, "step": 6130 }, { "grad_norm": 0.5835056900978088, "learning_rate": 9.980467799291307e-05, "loss": 0.0671, "step": 6140 }, { "grad_norm": 0.5352317690849304, "learning_rate": 9.980284863509058e-05, "loss": 0.0859, "step": 6150 }, { "grad_norm": 0.5379736423492432, "learning_rate": 9.980101076732334e-05, "loss": 0.0831, "step": 6160 }, { "grad_norm": 0.6115630269050598, "learning_rate": 9.979916438992544e-05, "loss": 0.0815, "step": 6170 }, { "grad_norm": 0.3908666670322418, "learning_rate": 9.979730950321237e-05, "loss": 0.0559, "step": 6180 }, { "grad_norm": 0.503819465637207, "learning_rate": 9.979544610750104e-05, "loss": 0.0672, "step": 6190 }, { "grad_norm": 0.4916321039199829, "learning_rate": 9.97935742031099e-05, "loss": 0.0569, "step": 6200 }, { "grad_norm": 0.539369523525238, "learning_rate": 9.979169379035878e-05, "loss": 0.0623, "step": 6210 }, { "grad_norm": 0.5622324347496033, "learning_rate": 9.978980486956899e-05, "loss": 0.0562, "step": 6220 }, { "grad_norm": 0.45410144329071045, "learning_rate": 9.978790744106332e-05, "loss": 0.0543, "step": 6230 }, { "grad_norm": 0.6411210894584656, "learning_rate": 9.978600150516594e-05, "loss": 0.062, "step": 6240 }, { "grad_norm": 0.4257316589355469, "learning_rate": 9.978408706220259e-05, "loss": 0.0685, "step": 6250 }, { "grad_norm": 0.4536980390548706, "learning_rate": 9.978216411250032e-05, "loss": 0.0521, "step": 6260 }, { "grad_norm": 0.44755470752716064, "learning_rate": 9.978023265638778e-05, "loss": 0.0537, "step": 6270 }, { "grad_norm": 0.5334983468055725, "learning_rate": 9.977829269419495e-05, "loss": 0.0602, "step": 6280 }, { "grad_norm": 0.48625949025154114, "learning_rate": 9.977634422625335e-05, "loss": 0.0662, "step": 6290 }, { "grad_norm": 0.4576539695262909, "learning_rate": 9.97743872528959e-05, "loss": 0.0656, "step": 6300 }, { "grad_norm": 0.5209015011787415, "learning_rate": 9.9772421774457e-05, "loss": 0.055, "step": 6310 }, { "grad_norm": 0.4497063457965851, "learning_rate": 9.977044779127252e-05, "loss": 0.0656, "step": 6320 }, { "grad_norm": 0.46779748797416687, "learning_rate": 9.976846530367971e-05, "loss": 0.0645, "step": 6330 }, { "grad_norm": 0.5337200164794922, "learning_rate": 9.976647431201735e-05, "loss": 0.082, "step": 6340 }, { "grad_norm": 0.5477796196937561, "learning_rate": 9.976447481662568e-05, "loss": 0.0646, "step": 6350 }, { "grad_norm": 0.4559324085712433, "learning_rate": 9.976246681784629e-05, "loss": 0.0656, "step": 6360 }, { "grad_norm": 0.40945255756378174, "learning_rate": 9.976045031602234e-05, "loss": 0.0622, "step": 6370 }, { "grad_norm": 0.6610015630722046, "learning_rate": 9.975842531149837e-05, "loss": 0.0575, "step": 6380 }, { "grad_norm": 0.6147937774658203, "learning_rate": 9.975639180462043e-05, "loss": 0.0668, "step": 6390 }, { "grad_norm": 0.4294508397579193, "learning_rate": 9.975434979573596e-05, "loss": 0.0625, "step": 6400 }, { "grad_norm": 0.44475895166397095, "learning_rate": 9.97522992851939e-05, "loss": 0.0553, "step": 6410 }, { "grad_norm": 0.5937820076942444, "learning_rate": 9.975024027334461e-05, "loss": 0.0646, "step": 6420 }, { "grad_norm": 0.42107322812080383, "learning_rate": 9.974817276053993e-05, "loss": 0.0539, "step": 6430 }, { "grad_norm": 0.8869693875312805, "learning_rate": 9.974609674713315e-05, "loss": 0.0576, "step": 6440 }, { "grad_norm": 0.4928419589996338, "learning_rate": 9.9744012233479e-05, "loss": 0.0526, "step": 6450 }, { "grad_norm": 0.4558889865875244, "learning_rate": 9.974191921993366e-05, "loss": 0.0539, "step": 6460 }, { "grad_norm": 0.5263728499412537, "learning_rate": 9.973981770685474e-05, "loss": 0.0683, "step": 6470 }, { "grad_norm": 0.499134361743927, "learning_rate": 9.97377076946014e-05, "loss": 0.0545, "step": 6480 }, { "grad_norm": 4.804525375366211, "learning_rate": 9.973558918353412e-05, "loss": 0.082, "step": 6490 }, { "grad_norm": 0.38381677865982056, "learning_rate": 9.973346217401494e-05, "loss": 0.0509, "step": 6500 }, { "grad_norm": 0.5449846386909485, "learning_rate": 9.973132666640726e-05, "loss": 0.0462, "step": 6510 }, { "grad_norm": 0.5690767168998718, "learning_rate": 9.972918266107602e-05, "loss": 0.0559, "step": 6520 }, { "grad_norm": 0.5075006484985352, "learning_rate": 9.972703015838756e-05, "loss": 0.0593, "step": 6530 }, { "grad_norm": 0.6201375722885132, "learning_rate": 9.97248691587097e-05, "loss": 0.0689, "step": 6540 }, { "grad_norm": 0.7037590742111206, "learning_rate": 9.972269966241166e-05, "loss": 0.0533, "step": 6550 }, { "grad_norm": 0.597108006477356, "learning_rate": 9.972052166986417e-05, "loss": 0.0675, "step": 6560 }, { "grad_norm": 0.772982120513916, "learning_rate": 9.971833518143938e-05, "loss": 0.0776, "step": 6570 }, { "grad_norm": 0.5231939554214478, "learning_rate": 9.971614019751093e-05, "loss": 0.0579, "step": 6580 }, { "grad_norm": 0.5944338440895081, "learning_rate": 9.971393671845383e-05, "loss": 0.0733, "step": 6590 }, { "grad_norm": 0.3339437246322632, "learning_rate": 9.971172474464464e-05, "loss": 0.0729, "step": 6600 }, { "grad_norm": 0.4806179702281952, "learning_rate": 9.97095042764613e-05, "loss": 0.0589, "step": 6610 }, { "grad_norm": 0.7073850035667419, "learning_rate": 9.970727531428324e-05, "loss": 0.0595, "step": 6620 }, { "grad_norm": 0.4677821099758148, "learning_rate": 9.970503785849132e-05, "loss": 0.0625, "step": 6630 }, { "grad_norm": 0.7329978346824646, "learning_rate": 9.970279190946788e-05, "loss": 0.0603, "step": 6640 }, { "grad_norm": 0.5591877102851868, "learning_rate": 9.970053746759667e-05, "loss": 0.0765, "step": 6650 }, { "grad_norm": 0.5066378712654114, "learning_rate": 9.969827453326292e-05, "loss": 0.0631, "step": 6660 }, { "grad_norm": 0.40526512265205383, "learning_rate": 9.969600310685332e-05, "loss": 0.06, "step": 6670 }, { "grad_norm": 0.5235259532928467, "learning_rate": 9.969372318875596e-05, "loss": 0.0664, "step": 6680 }, { "grad_norm": 0.57053142786026, "learning_rate": 9.969143477936043e-05, "loss": 0.0772, "step": 6690 }, { "grad_norm": 0.5670711994171143, "learning_rate": 9.968913787905775e-05, "loss": 0.0723, "step": 6700 }, { "grad_norm": 0.634070634841919, "learning_rate": 9.968683248824045e-05, "loss": 0.0784, "step": 6710 }, { "grad_norm": 0.4783041775226593, "learning_rate": 9.968451860730238e-05, "loss": 0.0811, "step": 6720 }, { "grad_norm": 0.6046650409698486, "learning_rate": 9.968219623663896e-05, "loss": 0.0738, "step": 6730 }, { "grad_norm": 0.5567582249641418, "learning_rate": 9.967986537664702e-05, "loss": 0.0929, "step": 6740 }, { "grad_norm": 0.43306857347488403, "learning_rate": 9.967752602772483e-05, "loss": 0.0717, "step": 6750 }, { "grad_norm": 0.616478681564331, "learning_rate": 9.967517819027212e-05, "loss": 0.0628, "step": 6760 }, { "grad_norm": 0.4250568747520447, "learning_rate": 9.967282186469009e-05, "loss": 0.0569, "step": 6770 }, { "grad_norm": 0.5115340948104858, "learning_rate": 9.967045705138135e-05, "loss": 0.0745, "step": 6780 }, { "grad_norm": 0.42937999963760376, "learning_rate": 9.966808375074998e-05, "loss": 0.0592, "step": 6790 }, { "grad_norm": 0.5131934881210327, "learning_rate": 9.966570196320154e-05, "loss": 0.061, "step": 6800 }, { "grad_norm": 0.4625800848007202, "learning_rate": 9.966331168914299e-05, "loss": 0.046, "step": 6810 }, { "grad_norm": 0.5773588418960571, "learning_rate": 9.966091292898277e-05, "loss": 0.0707, "step": 6820 }, { "grad_norm": 0.6965749859809875, "learning_rate": 9.965850568313076e-05, "loss": 0.0636, "step": 6830 }, { "grad_norm": 0.5287426710128784, "learning_rate": 9.965608995199827e-05, "loss": 0.0678, "step": 6840 }, { "grad_norm": 0.4280235767364502, "learning_rate": 9.965366573599812e-05, "loss": 0.0629, "step": 6850 }, { "grad_norm": 0.5541777610778809, "learning_rate": 9.965123303554453e-05, "loss": 0.0555, "step": 6860 }, { "grad_norm": 0.5842531323432922, "learning_rate": 9.964879185105317e-05, "loss": 0.0738, "step": 6870 }, { "grad_norm": 0.5315285325050354, "learning_rate": 9.964634218294119e-05, "loss": 0.0624, "step": 6880 }, { "grad_norm": 0.4341384768486023, "learning_rate": 9.964388403162714e-05, "loss": 0.061, "step": 6890 }, { "grad_norm": 0.5671449899673462, "learning_rate": 9.96414173975311e-05, "loss": 0.0724, "step": 6900 }, { "grad_norm": 0.44417497515678406, "learning_rate": 9.963894228107451e-05, "loss": 0.0631, "step": 6910 }, { "grad_norm": 0.3944939970970154, "learning_rate": 9.963645868268032e-05, "loss": 0.0665, "step": 6920 }, { "grad_norm": 0.5463246703147888, "learning_rate": 9.963396660277289e-05, "loss": 0.0613, "step": 6930 }, { "grad_norm": 0.5255232453346252, "learning_rate": 9.963146604177807e-05, "loss": 0.0538, "step": 6940 }, { "grad_norm": 0.508531391620636, "learning_rate": 9.962895700012311e-05, "loss": 0.0546, "step": 6950 }, { "grad_norm": 0.49816614389419556, "learning_rate": 9.962643947823677e-05, "loss": 0.0652, "step": 6960 }, { "grad_norm": 0.4418409764766693, "learning_rate": 9.962391347654921e-05, "loss": 0.0597, "step": 6970 }, { "grad_norm": 0.5667820572853088, "learning_rate": 9.962137899549204e-05, "loss": 0.0641, "step": 6980 }, { "grad_norm": 0.450850248336792, "learning_rate": 9.961883603549835e-05, "loss": 0.0549, "step": 6990 }, { "grad_norm": 0.4701966345310211, "learning_rate": 9.961628459700267e-05, "loss": 0.0662, "step": 7000 }, { "grad_norm": 0.4725888669490814, "learning_rate": 9.961372468044095e-05, "loss": 0.0602, "step": 7010 }, { "grad_norm": 0.8726286888122559, "learning_rate": 9.961115628625062e-05, "loss": 0.0551, "step": 7020 }, { "grad_norm": 0.5752240419387817, "learning_rate": 9.960857941487056e-05, "loss": 0.0856, "step": 7030 }, { "grad_norm": 0.6870790123939514, "learning_rate": 9.960599406674106e-05, "loss": 0.0641, "step": 7040 }, { "grad_norm": 0.43455395102500916, "learning_rate": 9.960340024230393e-05, "loss": 0.0573, "step": 7050 }, { "grad_norm": 0.4445975124835968, "learning_rate": 9.960079794200232e-05, "loss": 0.06, "step": 7060 }, { "grad_norm": 0.47916147112846375, "learning_rate": 9.959818716628096e-05, "loss": 0.0596, "step": 7070 }, { "grad_norm": 0.7548960447311401, "learning_rate": 9.95955679155859e-05, "loss": 0.0823, "step": 7080 }, { "grad_norm": 0.6051356196403503, "learning_rate": 9.959294019036472e-05, "loss": 0.0634, "step": 7090 }, { "grad_norm": 0.7320365309715271, "learning_rate": 9.959030399106646e-05, "loss": 0.0527, "step": 7100 }, { "grad_norm": 0.46535709500312805, "learning_rate": 9.958765931814153e-05, "loss": 0.0744, "step": 7110 }, { "grad_norm": 0.42633599042892456, "learning_rate": 9.958500617204184e-05, "loss": 0.0553, "step": 7120 }, { "grad_norm": 0.48647361993789673, "learning_rate": 9.958234455322075e-05, "loss": 0.0619, "step": 7130 }, { "grad_norm": 0.4949270784854889, "learning_rate": 9.957967446213308e-05, "loss": 0.0727, "step": 7140 }, { "grad_norm": 0.4370289742946625, "learning_rate": 9.957699589923501e-05, "loss": 0.0676, "step": 7150 }, { "grad_norm": 0.7032856941223145, "learning_rate": 9.957430886498431e-05, "loss": 0.072, "step": 7160 }, { "grad_norm": 0.5248855948448181, "learning_rate": 9.957161335984008e-05, "loss": 0.0541, "step": 7170 }, { "grad_norm": 0.7103868126869202, "learning_rate": 9.956890938426291e-05, "loss": 0.054, "step": 7180 }, { "grad_norm": 0.49597811698913574, "learning_rate": 9.956619693871482e-05, "loss": 0.0663, "step": 7190 }, { "grad_norm": 0.6892248392105103, "learning_rate": 9.956347602365934e-05, "loss": 0.0626, "step": 7200 }, { "grad_norm": 0.39949071407318115, "learning_rate": 9.956074663956135e-05, "loss": 0.0605, "step": 7210 }, { "grad_norm": 0.46196067333221436, "learning_rate": 9.955800878688726e-05, "loss": 0.0597, "step": 7220 }, { "grad_norm": 0.6237478256225586, "learning_rate": 9.955526246610489e-05, "loss": 0.0506, "step": 7230 }, { "grad_norm": 0.5923722386360168, "learning_rate": 9.955250767768349e-05, "loss": 0.0578, "step": 7240 }, { "grad_norm": 0.46619901061058044, "learning_rate": 9.95497444220938e-05, "loss": 0.0558, "step": 7250 }, { "grad_norm": 0.5000545382499695, "learning_rate": 9.954697269980797e-05, "loss": 0.055, "step": 7260 }, { "grad_norm": 0.3236261010169983, "learning_rate": 9.954419251129962e-05, "loss": 0.0609, "step": 7270 }, { "grad_norm": 0.5210586190223694, "learning_rate": 9.95414038570438e-05, "loss": 0.0601, "step": 7280 }, { "grad_norm": 0.44348233938217163, "learning_rate": 9.953860673751703e-05, "loss": 0.063, "step": 7290 }, { "grad_norm": 0.4140046238899231, "learning_rate": 9.953580115319725e-05, "loss": 0.0563, "step": 7300 }, { "grad_norm": 0.40607669949531555, "learning_rate": 9.953298710456387e-05, "loss": 0.0673, "step": 7310 }, { "grad_norm": 0.44427838921546936, "learning_rate": 9.953016459209771e-05, "loss": 0.0502, "step": 7320 }, { "grad_norm": 0.4247197210788727, "learning_rate": 9.952733361628108e-05, "loss": 0.0714, "step": 7330 }, { "grad_norm": 0.6313010454177856, "learning_rate": 9.952449417759772e-05, "loss": 0.068, "step": 7340 }, { "grad_norm": 0.4233396649360657, "learning_rate": 9.952164627653279e-05, "loss": 0.0576, "step": 7350 }, { "grad_norm": 0.5229185819625854, "learning_rate": 9.951878991357292e-05, "loss": 0.0561, "step": 7360 }, { "grad_norm": 0.6376730799674988, "learning_rate": 9.951592508920622e-05, "loss": 0.0599, "step": 7370 }, { "grad_norm": 0.3716120719909668, "learning_rate": 9.951305180392219e-05, "loss": 0.0538, "step": 7380 }, { "grad_norm": 0.6650915741920471, "learning_rate": 9.951017005821178e-05, "loss": 0.062, "step": 7390 }, { "grad_norm": 0.7303879857063293, "learning_rate": 9.95072798525674e-05, "loss": 0.0583, "step": 7400 }, { "grad_norm": 0.4595858156681061, "learning_rate": 9.950438118748293e-05, "loss": 0.0607, "step": 7410 }, { "grad_norm": 0.5084384083747864, "learning_rate": 9.950147406345366e-05, "loss": 0.0638, "step": 7420 }, { "grad_norm": 0.5458325743675232, "learning_rate": 9.949855848097635e-05, "loss": 0.0655, "step": 7430 }, { "grad_norm": 0.5362468957901001, "learning_rate": 9.949563444054916e-05, "loss": 0.0644, "step": 7440 }, { "grad_norm": 0.445383757352829, "learning_rate": 9.949270194267178e-05, "loss": 0.0542, "step": 7450 }, { "grad_norm": 0.44463130831718445, "learning_rate": 9.948976098784526e-05, "loss": 0.0602, "step": 7460 }, { "grad_norm": 0.6282782554626465, "learning_rate": 9.948681157657213e-05, "loss": 0.0673, "step": 7470 }, { "grad_norm": 0.6791053414344788, "learning_rate": 9.948385370935638e-05, "loss": 0.0794, "step": 7480 }, { "grad_norm": 0.54116290807724, "learning_rate": 9.94808873867034e-05, "loss": 0.0594, "step": 7490 }, { "grad_norm": 0.4625099003314972, "learning_rate": 9.947791260912009e-05, "loss": 0.0482, "step": 7500 }, { "grad_norm": 0.643775224685669, "learning_rate": 9.947492937711474e-05, "loss": 0.0514, "step": 7510 }, { "grad_norm": 0.5468482971191406, "learning_rate": 9.947193769119707e-05, "loss": 0.0491, "step": 7520 }, { "grad_norm": 0.4253377616405487, "learning_rate": 9.946893755187834e-05, "loss": 0.0434, "step": 7530 }, { "grad_norm": 0.3893895149230957, "learning_rate": 9.946592895967115e-05, "loss": 0.052, "step": 7540 }, { "grad_norm": 0.4589829444885254, "learning_rate": 9.94629119150896e-05, "loss": 0.0454, "step": 7550 }, { "grad_norm": 0.4223136007785797, "learning_rate": 9.94598864186492e-05, "loss": 0.0523, "step": 7560 }, { "grad_norm": 0.4321393668651581, "learning_rate": 9.945685247086696e-05, "loss": 0.0613, "step": 7570 }, { "grad_norm": 0.4117303490638733, "learning_rate": 9.945381007226129e-05, "loss": 0.056, "step": 7580 }, { "grad_norm": 0.5238384008407593, "learning_rate": 9.945075922335203e-05, "loss": 0.0573, "step": 7590 }, { "grad_norm": 0.32063496112823486, "learning_rate": 9.944769992466049e-05, "loss": 0.0688, "step": 7600 }, { "grad_norm": 0.4850861728191376, "learning_rate": 9.944463217670945e-05, "loss": 0.0636, "step": 7610 }, { "grad_norm": 0.4469912052154541, "learning_rate": 9.944155598002307e-05, "loss": 0.0552, "step": 7620 }, { "grad_norm": 0.37371596693992615, "learning_rate": 9.943847133512701e-05, "loss": 0.0547, "step": 7630 }, { "grad_norm": 0.4201574921607971, "learning_rate": 9.943537824254834e-05, "loss": 0.0512, "step": 7640 }, { "grad_norm": 0.4943617582321167, "learning_rate": 9.943227670281559e-05, "loss": 0.0464, "step": 7650 }, { "grad_norm": 0.45728766918182373, "learning_rate": 9.942916671645873e-05, "loss": 0.0545, "step": 7660 }, { "grad_norm": 0.426567405462265, "learning_rate": 9.942604828400916e-05, "loss": 0.0591, "step": 7670 }, { "grad_norm": 0.5905613303184509, "learning_rate": 9.942292140599975e-05, "loss": 0.0491, "step": 7680 }, { "grad_norm": 0.41681453585624695, "learning_rate": 9.94197860829648e-05, "loss": 0.0536, "step": 7690 }, { "grad_norm": 0.38989466428756714, "learning_rate": 9.941664231544004e-05, "loss": 0.0446, "step": 7700 }, { "grad_norm": 0.5462397336959839, "learning_rate": 9.941349010396264e-05, "loss": 0.0651, "step": 7710 }, { "grad_norm": 0.5115859508514404, "learning_rate": 9.941032944907125e-05, "loss": 0.0623, "step": 7720 }, { "grad_norm": 0.5270349383354187, "learning_rate": 9.940716035130596e-05, "loss": 0.0518, "step": 7730 }, { "grad_norm": 0.4954008460044861, "learning_rate": 9.940398281120821e-05, "loss": 0.0511, "step": 7740 }, { "grad_norm": 0.46840953826904297, "learning_rate": 9.940079682932102e-05, "loss": 0.044, "step": 7750 }, { "grad_norm": 0.6215953230857849, "learning_rate": 9.939760240618877e-05, "loss": 0.0617, "step": 7760 }, { "grad_norm": 0.5668230056762695, "learning_rate": 9.939439954235729e-05, "loss": 0.0649, "step": 7770 }, { "grad_norm": 0.4628986120223999, "learning_rate": 9.939118823837387e-05, "loss": 0.0546, "step": 7780 }, { "grad_norm": 0.4418438971042633, "learning_rate": 9.938796849478725e-05, "loss": 0.0601, "step": 7790 }, { "grad_norm": 0.5633415579795837, "learning_rate": 9.938474031214755e-05, "loss": 0.0572, "step": 7800 }, { "grad_norm": 0.44084471464157104, "learning_rate": 9.938150369100643e-05, "loss": 0.0451, "step": 7810 }, { "grad_norm": 0.5254381895065308, "learning_rate": 9.93782586319169e-05, "loss": 0.0461, "step": 7820 }, { "grad_norm": 0.8242443203926086, "learning_rate": 9.937500513543348e-05, "loss": 0.0531, "step": 7830 }, { "grad_norm": 0.5697723627090454, "learning_rate": 9.937174320211207e-05, "loss": 0.0618, "step": 7840 }, { "grad_norm": 0.5375595688819885, "learning_rate": 9.936847283251009e-05, "loss": 0.054, "step": 7850 }, { "grad_norm": 0.4541417062282562, "learning_rate": 9.936519402718632e-05, "loss": 0.0649, "step": 7860 }, { "grad_norm": 0.4602662920951843, "learning_rate": 9.936190678670102e-05, "loss": 0.0659, "step": 7870 }, { "grad_norm": 0.46443599462509155, "learning_rate": 9.935861111161593e-05, "loss": 0.0654, "step": 7880 }, { "grad_norm": 0.474616676568985, "learning_rate": 9.935530700249416e-05, "loss": 0.0458, "step": 7890 }, { "grad_norm": 0.4418122470378876, "learning_rate": 9.935199445990028e-05, "loss": 0.0774, "step": 7900 }, { "grad_norm": 0.3534408509731293, "learning_rate": 9.934867348440033e-05, "loss": 0.0664, "step": 7910 }, { "grad_norm": 0.4663161635398865, "learning_rate": 9.934534407656176e-05, "loss": 0.0822, "step": 7920 }, { "grad_norm": 0.4953087568283081, "learning_rate": 9.93420062369535e-05, "loss": 0.0614, "step": 7930 }, { "grad_norm": 0.5676716566085815, "learning_rate": 9.933865996614589e-05, "loss": 0.0589, "step": 7940 }, { "grad_norm": 0.7215105295181274, "learning_rate": 9.933530526471068e-05, "loss": 0.0602, "step": 7950 }, { "grad_norm": 0.7616533041000366, "learning_rate": 9.933194213322114e-05, "loss": 0.0535, "step": 7960 }, { "grad_norm": 0.4010379910469055, "learning_rate": 9.932857057225192e-05, "loss": 0.0584, "step": 7970 }, { "grad_norm": 0.5541771650314331, "learning_rate": 9.932519058237912e-05, "loss": 0.0547, "step": 7980 }, { "grad_norm": 0.5742183327674866, "learning_rate": 9.932180216418032e-05, "loss": 0.0559, "step": 7990 }, { "grad_norm": 0.3940316140651703, "learning_rate": 9.931840531823446e-05, "loss": 0.0509, "step": 8000 }, { "grad_norm": 0.3295437693595886, "learning_rate": 9.9315000045122e-05, "loss": 0.0492, "step": 8010 }, { "grad_norm": 0.521837055683136, "learning_rate": 9.931158634542481e-05, "loss": 0.0521, "step": 8020 }, { "grad_norm": 0.4891173243522644, "learning_rate": 9.930816421972617e-05, "loss": 0.0554, "step": 8030 }, { "grad_norm": 0.46587976813316345, "learning_rate": 9.930473366861086e-05, "loss": 0.0479, "step": 8040 }, { "grad_norm": 0.3890436291694641, "learning_rate": 9.930129469266505e-05, "loss": 0.0634, "step": 8050 }, { "grad_norm": 0.3559127449989319, "learning_rate": 9.929784729247638e-05, "loss": 0.0537, "step": 8060 }, { "grad_norm": 0.486928790807724, "learning_rate": 9.929439146863389e-05, "loss": 0.0518, "step": 8070 }, { "grad_norm": 0.6164573431015015, "learning_rate": 9.92909272217281e-05, "loss": 0.0712, "step": 8080 }, { "grad_norm": 0.5879020094871521, "learning_rate": 9.928745455235097e-05, "loss": 0.056, "step": 8090 }, { "grad_norm": 0.3864176273345947, "learning_rate": 9.928397346109588e-05, "loss": 0.06, "step": 8100 }, { "grad_norm": 0.6531065702438354, "learning_rate": 9.928048394855762e-05, "loss": 0.0635, "step": 8110 }, { "grad_norm": 0.5104966163635254, "learning_rate": 9.92769860153325e-05, "loss": 0.0632, "step": 8120 }, { "grad_norm": 0.4837489426136017, "learning_rate": 9.927347966201819e-05, "loss": 0.0601, "step": 8130 }, { "grad_norm": 0.4989352226257324, "learning_rate": 9.926996488921383e-05, "loss": 0.0637, "step": 8140 }, { "grad_norm": 0.6647611856460571, "learning_rate": 9.926644169752001e-05, "loss": 0.0492, "step": 8150 }, { "grad_norm": 0.4467063546180725, "learning_rate": 9.926291008753875e-05, "loss": 0.0598, "step": 8160 }, { "grad_norm": 0.5170247554779053, "learning_rate": 9.92593700598735e-05, "loss": 0.0567, "step": 8170 }, { "grad_norm": 0.40450945496559143, "learning_rate": 9.925582161512915e-05, "loss": 0.0509, "step": 8180 }, { "grad_norm": 0.6117929220199585, "learning_rate": 9.925226475391205e-05, "loss": 0.0585, "step": 8190 }, { "grad_norm": 0.33192354440689087, "learning_rate": 9.924869947682993e-05, "loss": 0.0483, "step": 8200 }, { "grad_norm": 0.4700356423854828, "learning_rate": 9.924512578449204e-05, "loss": 0.0459, "step": 8210 }, { "grad_norm": 0.3673359155654907, "learning_rate": 9.924154367750901e-05, "loss": 0.0478, "step": 8220 }, { "grad_norm": 0.3556532859802246, "learning_rate": 9.923795315649293e-05, "loss": 0.0492, "step": 8230 }, { "grad_norm": 0.5261340737342834, "learning_rate": 9.92343542220573e-05, "loss": 0.0588, "step": 8240 }, { "grad_norm": 0.723939836025238, "learning_rate": 9.92307468748171e-05, "loss": 0.0517, "step": 8250 }, { "grad_norm": 0.604206919670105, "learning_rate": 9.922713111538873e-05, "loss": 0.0471, "step": 8260 }, { "grad_norm": 0.4402112364768982, "learning_rate": 9.922350694439003e-05, "loss": 0.0492, "step": 8270 }, { "grad_norm": 0.4342346489429474, "learning_rate": 9.921987436244024e-05, "loss": 0.0584, "step": 8280 }, { "grad_norm": 0.4347570240497589, "learning_rate": 9.921623337016008e-05, "loss": 0.0455, "step": 8290 }, { "grad_norm": 0.38862666487693787, "learning_rate": 9.921258396817172e-05, "loss": 0.0556, "step": 8300 }, { "grad_norm": 0.48430371284484863, "learning_rate": 9.920892615709874e-05, "loss": 0.051, "step": 8310 }, { "grad_norm": 0.5458853244781494, "learning_rate": 9.920525993756612e-05, "loss": 0.0497, "step": 8320 }, { "grad_norm": 0.4748211205005646, "learning_rate": 9.920158531020036e-05, "loss": 0.054, "step": 8330 }, { "grad_norm": 0.40361565351486206, "learning_rate": 9.919790227562933e-05, "loss": 0.0493, "step": 8340 }, { "grad_norm": 0.5323659777641296, "learning_rate": 9.919421083448237e-05, "loss": 0.0686, "step": 8350 }, { "grad_norm": 0.455133318901062, "learning_rate": 9.919051098739022e-05, "loss": 0.052, "step": 8360 }, { "grad_norm": 0.4703785479068756, "learning_rate": 9.918680273498514e-05, "loss": 0.0506, "step": 8370 }, { "grad_norm": 0.6506497263908386, "learning_rate": 9.918308607790072e-05, "loss": 0.0652, "step": 8380 }, { "grad_norm": 0.3054303526878357, "learning_rate": 9.917936101677205e-05, "loss": 0.0453, "step": 8390 }, { "grad_norm": 0.5566979646682739, "learning_rate": 9.917562755223564e-05, "loss": 0.0464, "step": 8400 }, { "grad_norm": 0.44076693058013916, "learning_rate": 9.917188568492944e-05, "loss": 0.0503, "step": 8410 }, { "grad_norm": 0.5155674815177917, "learning_rate": 9.916813541549283e-05, "loss": 0.0469, "step": 8420 }, { "grad_norm": 0.49975281953811646, "learning_rate": 9.916437674456663e-05, "loss": 0.0494, "step": 8430 }, { "grad_norm": 0.3596594035625458, "learning_rate": 9.916060967279308e-05, "loss": 0.0466, "step": 8440 }, { "grad_norm": 0.3671681582927704, "learning_rate": 9.91568342008159e-05, "loss": 0.0541, "step": 8450 }, { "grad_norm": 0.49531376361846924, "learning_rate": 9.915305032928019e-05, "loss": 0.0536, "step": 8460 }, { "grad_norm": 0.5450968146324158, "learning_rate": 9.914925805883253e-05, "loss": 0.0566, "step": 8470 }, { "grad_norm": 0.5265876650810242, "learning_rate": 9.914545739012088e-05, "loss": 0.0524, "step": 8480 }, { "grad_norm": 0.5135941505432129, "learning_rate": 9.91416483237947e-05, "loss": 0.0582, "step": 8490 }, { "grad_norm": 0.49577298760414124, "learning_rate": 9.913783086050485e-05, "loss": 0.0603, "step": 8500 }, { "grad_norm": 0.3401779234409332, "learning_rate": 9.913400500090364e-05, "loss": 0.0476, "step": 8510 }, { "grad_norm": 0.5127879977226257, "learning_rate": 9.913017074564479e-05, "loss": 0.0385, "step": 8520 }, { "grad_norm": 0.5080923438072205, "learning_rate": 9.912632809538348e-05, "loss": 0.0567, "step": 8530 }, { "grad_norm": 0.5797358155250549, "learning_rate": 9.912247705077629e-05, "loss": 0.0544, "step": 8540 }, { "grad_norm": 0.3798142671585083, "learning_rate": 9.911861761248127e-05, "loss": 0.0601, "step": 8550 }, { "grad_norm": 0.38797810673713684, "learning_rate": 9.91147497811579e-05, "loss": 0.0584, "step": 8560 }, { "grad_norm": 0.6057915091514587, "learning_rate": 9.911087355746709e-05, "loss": 0.0933, "step": 8570 }, { "grad_norm": 0.6792982816696167, "learning_rate": 9.910698894207117e-05, "loss": 0.0576, "step": 8580 }, { "grad_norm": 0.6116101145744324, "learning_rate": 9.910309593563392e-05, "loss": 0.0551, "step": 8590 }, { "grad_norm": 0.42589545249938965, "learning_rate": 9.909919453882057e-05, "loss": 0.0436, "step": 8600 }, { "grad_norm": 0.5994938015937805, "learning_rate": 9.90952847522977e-05, "loss": 0.0455, "step": 8610 }, { "grad_norm": 0.3811785876750946, "learning_rate": 9.909136657673346e-05, "loss": 0.0516, "step": 8620 }, { "grad_norm": 0.4819124937057495, "learning_rate": 9.908744001279731e-05, "loss": 0.0486, "step": 8630 }, { "grad_norm": 0.6693864464759827, "learning_rate": 9.90835050611602e-05, "loss": 0.0575, "step": 8640 }, { "grad_norm": 0.4307042360305786, "learning_rate": 9.90795617224945e-05, "loss": 0.0603, "step": 8650 }, { "grad_norm": 0.4761705994606018, "learning_rate": 9.907560999747405e-05, "loss": 0.0474, "step": 8660 }, { "grad_norm": 0.2950363755226135, "learning_rate": 9.907164988677408e-05, "loss": 0.0385, "step": 8670 }, { "grad_norm": 0.2888356149196625, "learning_rate": 9.906768139107124e-05, "loss": 0.033, "step": 8680 }, { "grad_norm": 0.3162114918231964, "learning_rate": 9.906370451104367e-05, "loss": 0.0432, "step": 8690 }, { "grad_norm": 0.4150416851043701, "learning_rate": 9.905971924737088e-05, "loss": 0.0445, "step": 8700 }, { "grad_norm": 0.5218140482902527, "learning_rate": 9.905572560073387e-05, "loss": 0.0595, "step": 8710 }, { "grad_norm": 0.6749846935272217, "learning_rate": 9.905172357181501e-05, "loss": 0.0623, "step": 8720 }, { "grad_norm": 0.34645378589630127, "learning_rate": 9.904771316129817e-05, "loss": 0.0459, "step": 8730 }, { "grad_norm": 0.4780953824520111, "learning_rate": 9.904369436986862e-05, "loss": 0.0471, "step": 8740 }, { "grad_norm": 0.39113354682922363, "learning_rate": 9.903966719821303e-05, "loss": 0.0575, "step": 8750 }, { "grad_norm": 0.4185963273048401, "learning_rate": 9.903563164701956e-05, "loss": 0.0511, "step": 8760 }, { "grad_norm": 0.4358232617378235, "learning_rate": 9.903158771697778e-05, "loss": 0.0655, "step": 8770 }, { "grad_norm": 0.5732045769691467, "learning_rate": 9.902753540877867e-05, "loss": 0.0601, "step": 8780 }, { "grad_norm": 0.7076114416122437, "learning_rate": 9.902347472311466e-05, "loss": 0.0524, "step": 8790 }, { "grad_norm": 0.6034021377563477, "learning_rate": 9.901940566067962e-05, "loss": 0.0517, "step": 8800 }, { "grad_norm": 0.4152577221393585, "learning_rate": 9.901532822216883e-05, "loss": 0.0487, "step": 8810 }, { "grad_norm": 0.4814049303531647, "learning_rate": 9.901124240827904e-05, "loss": 0.0531, "step": 8820 }, { "grad_norm": 0.6517809629440308, "learning_rate": 9.900714821970835e-05, "loss": 0.0476, "step": 8830 }, { "grad_norm": 0.40238073468208313, "learning_rate": 9.900304565715641e-05, "loss": 0.0497, "step": 8840 }, { "grad_norm": 0.3334257900714874, "learning_rate": 9.899893472132419e-05, "loss": 0.0502, "step": 8850 }, { "grad_norm": 0.488071084022522, "learning_rate": 9.899481541291415e-05, "loss": 0.0459, "step": 8860 }, { "grad_norm": 0.4518069922924042, "learning_rate": 9.899068773263016e-05, "loss": 0.0516, "step": 8870 }, { "grad_norm": 0.5454759001731873, "learning_rate": 9.898655168117754e-05, "loss": 0.0537, "step": 8880 }, { "grad_norm": 0.5435086488723755, "learning_rate": 9.898240725926302e-05, "loss": 0.0576, "step": 8890 }, { "grad_norm": 0.3980844020843506, "learning_rate": 9.897825446759478e-05, "loss": 0.0528, "step": 8900 }, { "grad_norm": 0.4826858937740326, "learning_rate": 9.897409330688241e-05, "loss": 0.0485, "step": 8910 }, { "grad_norm": 0.587395429611206, "learning_rate": 9.896992377783692e-05, "loss": 0.0617, "step": 8920 }, { "grad_norm": 0.4524063169956207, "learning_rate": 9.89657458811708e-05, "loss": 0.0529, "step": 8930 }, { "grad_norm": 0.40646740794181824, "learning_rate": 9.896155961759792e-05, "loss": 0.0576, "step": 8940 }, { "grad_norm": 0.4514388144016266, "learning_rate": 9.895736498783361e-05, "loss": 0.0506, "step": 8950 }, { "grad_norm": 0.4074675440788269, "learning_rate": 9.895316199259462e-05, "loss": 0.0518, "step": 8960 }, { "grad_norm": 0.42566031217575073, "learning_rate": 9.894895063259909e-05, "loss": 0.0463, "step": 8970 }, { "grad_norm": 0.4873006045818329, "learning_rate": 9.894473090856667e-05, "loss": 0.0537, "step": 8980 }, { "grad_norm": 0.5102710127830505, "learning_rate": 9.894050282121839e-05, "loss": 0.0732, "step": 8990 }, { "grad_norm": 0.3785865306854248, "learning_rate": 9.893626637127668e-05, "loss": 0.0516, "step": 9000 }, { "grad_norm": 0.3611728250980377, "learning_rate": 9.893202155946546e-05, "loss": 0.0456, "step": 9010 }, { "grad_norm": 0.45154869556427, "learning_rate": 9.892776838651006e-05, "loss": 0.0574, "step": 9020 }, { "grad_norm": 0.4537663161754608, "learning_rate": 9.892350685313722e-05, "loss": 0.0592, "step": 9030 }, { "grad_norm": 0.4865742325782776, "learning_rate": 9.891923696007513e-05, "loss": 0.0658, "step": 9040 }, { "grad_norm": 0.4052159786224365, "learning_rate": 9.891495870805336e-05, "loss": 0.0507, "step": 9050 }, { "grad_norm": 0.49616339802742004, "learning_rate": 9.891067209780298e-05, "loss": 0.0553, "step": 9060 }, { "grad_norm": 0.20538119971752167, "learning_rate": 9.890637713005646e-05, "loss": 0.0361, "step": 9070 }, { "grad_norm": 0.37192633748054504, "learning_rate": 9.890207380554767e-05, "loss": 0.0428, "step": 9080 }, { "grad_norm": 0.41170740127563477, "learning_rate": 9.889776212501196e-05, "loss": 0.0581, "step": 9090 }, { "grad_norm": 0.4398897588253021, "learning_rate": 9.889344208918605e-05, "loss": 0.0587, "step": 9100 }, { "grad_norm": 0.5232097506523132, "learning_rate": 9.888911369880812e-05, "loss": 0.0561, "step": 9110 }, { "grad_norm": 0.5697697997093201, "learning_rate": 9.888477695461777e-05, "loss": 0.0518, "step": 9120 }, { "grad_norm": 0.4766896069049835, "learning_rate": 9.888043185735607e-05, "loss": 0.0476, "step": 9130 }, { "grad_norm": 0.4020838737487793, "learning_rate": 9.887607840776542e-05, "loss": 0.0392, "step": 9140 }, { "grad_norm": 0.4361814856529236, "learning_rate": 9.887171660658975e-05, "loss": 0.0561, "step": 9150 }, { "grad_norm": 0.3899551033973694, "learning_rate": 9.886734645457435e-05, "loss": 0.0408, "step": 9160 }, { "grad_norm": 0.40749049186706543, "learning_rate": 9.886296795246597e-05, "loss": 0.0435, "step": 9170 }, { "grad_norm": 0.4682026505470276, "learning_rate": 9.885858110101276e-05, "loss": 0.0449, "step": 9180 }, { "grad_norm": 0.31077831983566284, "learning_rate": 9.885418590096434e-05, "loss": 0.0433, "step": 9190 }, { "grad_norm": 0.5677563548088074, "learning_rate": 9.88497823530717e-05, "loss": 0.0549, "step": 9200 }, { "grad_norm": 0.5019437670707703, "learning_rate": 9.884537045808732e-05, "loss": 0.0665, "step": 9210 }, { "grad_norm": 0.3827044367790222, "learning_rate": 9.884095021676502e-05, "loss": 0.0597, "step": 9220 }, { "grad_norm": 0.44691866636276245, "learning_rate": 9.883652162986017e-05, "loss": 0.0547, "step": 9230 }, { "grad_norm": 0.3885970413684845, "learning_rate": 9.883208469812943e-05, "loss": 0.0578, "step": 9240 }, { "grad_norm": 0.46878594160079956, "learning_rate": 9.882763942233098e-05, "loss": 0.0475, "step": 9250 }, { "grad_norm": 0.2630952000617981, "learning_rate": 9.882318580322441e-05, "loss": 0.0431, "step": 9260 }, { "grad_norm": 0.37337857484817505, "learning_rate": 9.881872384157067e-05, "loss": 0.0531, "step": 9270 }, { "grad_norm": 0.45215630531311035, "learning_rate": 9.881425353813225e-05, "loss": 0.0539, "step": 9280 }, { "grad_norm": 0.6579791903495789, "learning_rate": 9.880977489367296e-05, "loss": 0.058, "step": 9290 }, { "grad_norm": 0.5978196859359741, "learning_rate": 9.88052879089581e-05, "loss": 0.0679, "step": 9300 }, { "grad_norm": 0.6540464758872986, "learning_rate": 9.880079258475434e-05, "loss": 0.0592, "step": 9310 }, { "grad_norm": 0.6270044445991516, "learning_rate": 9.879628892182985e-05, "loss": 0.0759, "step": 9320 }, { "grad_norm": 0.5714443325996399, "learning_rate": 9.879177692095416e-05, "loss": 0.0599, "step": 9330 }, { "grad_norm": 0.6162557005882263, "learning_rate": 9.878725658289825e-05, "loss": 0.0652, "step": 9340 }, { "grad_norm": 0.42647144198417664, "learning_rate": 9.878272790843454e-05, "loss": 0.052, "step": 9350 }, { "grad_norm": 0.4445285201072693, "learning_rate": 9.877819089833682e-05, "loss": 0.0456, "step": 9360 }, { "grad_norm": 0.35428282618522644, "learning_rate": 9.877364555338038e-05, "loss": 0.0623, "step": 9370 }, { "grad_norm": 0.4485381245613098, "learning_rate": 9.876909187434186e-05, "loss": 0.0518, "step": 9380 }, { "grad_norm": 0.4555874466896057, "learning_rate": 9.876452986199939e-05, "loss": 0.0461, "step": 9390 }, { "grad_norm": 0.4785311222076416, "learning_rate": 9.875995951713248e-05, "loss": 0.0593, "step": 9400 }, { "grad_norm": 0.49658772349357605, "learning_rate": 9.875538084052207e-05, "loss": 0.0554, "step": 9410 }, { "grad_norm": 0.729568600654602, "learning_rate": 9.875079383295053e-05, "loss": 0.0613, "step": 9420 }, { "grad_norm": 0.5221003293991089, "learning_rate": 9.874619849520167e-05, "loss": 0.0576, "step": 9430 }, { "grad_norm": 0.4751858115196228, "learning_rate": 9.874159482806069e-05, "loss": 0.0539, "step": 9440 }, { "grad_norm": 0.3511454164981842, "learning_rate": 9.873698283231426e-05, "loss": 0.0477, "step": 9450 }, { "grad_norm": 0.37872424721717834, "learning_rate": 9.87323625087504e-05, "loss": 0.0482, "step": 9460 }, { "grad_norm": 0.42815181612968445, "learning_rate": 9.872773385815863e-05, "loss": 0.0644, "step": 9470 }, { "grad_norm": 0.42701011896133423, "learning_rate": 9.872309688132986e-05, "loss": 0.0484, "step": 9480 }, { "grad_norm": 0.5254909992218018, "learning_rate": 9.871845157905639e-05, "loss": 0.0524, "step": 9490 }, { "grad_norm": 0.422732412815094, "learning_rate": 9.871379795213201e-05, "loss": 0.0419, "step": 9500 }, { "grad_norm": 0.5282423496246338, "learning_rate": 9.87091360013519e-05, "loss": 0.0464, "step": 9510 }, { "grad_norm": 0.5883885622024536, "learning_rate": 9.870446572751262e-05, "loss": 0.0501, "step": 9520 }, { "grad_norm": 0.5642241835594177, "learning_rate": 9.869978713141224e-05, "loss": 0.05, "step": 9530 }, { "grad_norm": 0.5693297982215881, "learning_rate": 9.869510021385016e-05, "loss": 0.0495, "step": 9540 }, { "grad_norm": 0.4743920564651489, "learning_rate": 9.869040497562727e-05, "loss": 0.065, "step": 9550 }, { "grad_norm": 0.4592972993850708, "learning_rate": 9.868570141754587e-05, "loss": 0.0634, "step": 9560 }, { "grad_norm": 0.741771399974823, "learning_rate": 9.868098954040965e-05, "loss": 0.0737, "step": 9570 }, { "grad_norm": 0.552486777305603, "learning_rate": 9.867626934502374e-05, "loss": 0.0577, "step": 9580 }, { "grad_norm": 0.47544634342193604, "learning_rate": 9.86715408321947e-05, "loss": 0.0519, "step": 9590 }, { "grad_norm": 0.4578838646411896, "learning_rate": 9.86668040027305e-05, "loss": 0.0506, "step": 9600 }, { "grad_norm": 0.5171000361442566, "learning_rate": 9.866205885744053e-05, "loss": 0.0542, "step": 9610 }, { "grad_norm": 0.4440520405769348, "learning_rate": 9.865730539713563e-05, "loss": 0.0511, "step": 9620 }, { "grad_norm": 0.6186878681182861, "learning_rate": 9.8652543622628e-05, "loss": 0.057, "step": 9630 }, { "grad_norm": 0.36847904324531555, "learning_rate": 9.864777353473132e-05, "loss": 0.0464, "step": 9640 }, { "grad_norm": 0.4456900954246521, "learning_rate": 9.864299513426068e-05, "loss": 0.0514, "step": 9650 }, { "grad_norm": 0.46710503101348877, "learning_rate": 9.863820842203254e-05, "loss": 0.0492, "step": 9660 }, { "grad_norm": 0.4537682831287384, "learning_rate": 9.863341339886483e-05, "loss": 0.0543, "step": 9670 }, { "grad_norm": 0.4802383780479431, "learning_rate": 9.86286100655769e-05, "loss": 0.0428, "step": 9680 }, { "grad_norm": 0.3942939043045044, "learning_rate": 9.862379842298953e-05, "loss": 0.052, "step": 9690 }, { "grad_norm": 0.5292214155197144, "learning_rate": 9.861897847192485e-05, "loss": 0.0456, "step": 9700 }, { "grad_norm": 0.60828697681427, "learning_rate": 9.86141502132065e-05, "loss": 0.0509, "step": 9710 }, { "grad_norm": 0.47518301010131836, "learning_rate": 9.860931364765946e-05, "loss": 0.0531, "step": 9720 }, { "grad_norm": 0.4878525733947754, "learning_rate": 9.860446877611021e-05, "loss": 0.0622, "step": 9730 }, { "grad_norm": 0.43835732340812683, "learning_rate": 9.859961559938655e-05, "loss": 0.0515, "step": 9740 }, { "grad_norm": 0.5224351286888123, "learning_rate": 9.85947541183178e-05, "loss": 0.0583, "step": 9750 }, { "grad_norm": 0.4763549864292145, "learning_rate": 9.858988433373463e-05, "loss": 0.0501, "step": 9760 }, { "grad_norm": 0.4473944306373596, "learning_rate": 9.858500624646918e-05, "loss": 0.0497, "step": 9770 }, { "grad_norm": 0.4282688796520233, "learning_rate": 9.858011985735497e-05, "loss": 0.043, "step": 9780 }, { "grad_norm": 0.5473127365112305, "learning_rate": 9.857522516722693e-05, "loss": 0.0478, "step": 9790 }, { "grad_norm": 0.47285550832748413, "learning_rate": 9.857032217692145e-05, "loss": 0.0478, "step": 9800 }, { "grad_norm": 0.6193031668663025, "learning_rate": 9.856541088727631e-05, "loss": 0.0533, "step": 9810 }, { "grad_norm": 0.5568262934684753, "learning_rate": 9.856049129913072e-05, "loss": 0.0524, "step": 9820 }, { "grad_norm": 0.4820765554904938, "learning_rate": 9.85555634133253e-05, "loss": 0.051, "step": 9830 }, { "grad_norm": 0.30784884095191956, "learning_rate": 9.855062723070208e-05, "loss": 0.0503, "step": 9840 }, { "grad_norm": 0.3259543180465698, "learning_rate": 9.854568275210454e-05, "loss": 0.0476, "step": 9850 }, { "grad_norm": 0.5279306769371033, "learning_rate": 9.854072997837754e-05, "loss": 0.0383, "step": 9860 }, { "grad_norm": 0.3535550832748413, "learning_rate": 9.853576891036737e-05, "loss": 0.0465, "step": 9870 }, { "grad_norm": 0.4347693622112274, "learning_rate": 9.853079954892177e-05, "loss": 0.0518, "step": 9880 }, { "grad_norm": 0.4146589934825897, "learning_rate": 9.852582189488983e-05, "loss": 0.0433, "step": 9890 }, { "grad_norm": 0.45798778533935547, "learning_rate": 9.852083594912212e-05, "loss": 0.0621, "step": 9900 }, { "grad_norm": 0.3329858183860779, "learning_rate": 9.851584171247058e-05, "loss": 0.0539, "step": 9910 }, { "grad_norm": 0.41284456849098206, "learning_rate": 9.851083918578863e-05, "loss": 0.0602, "step": 9920 }, { "grad_norm": 0.41141393780708313, "learning_rate": 9.850582836993103e-05, "loss": 0.0447, "step": 9930 }, { "grad_norm": 0.4535030126571655, "learning_rate": 9.850080926575397e-05, "loss": 0.0478, "step": 9940 }, { "grad_norm": 0.4935802221298218, "learning_rate": 9.849578187411515e-05, "loss": 0.0452, "step": 9950 }, { "grad_norm": 0.3944477140903473, "learning_rate": 9.849074619587354e-05, "loss": 0.0508, "step": 9960 }, { "grad_norm": 0.3858025074005127, "learning_rate": 9.848570223188964e-05, "loss": 0.0572, "step": 9970 }, { "grad_norm": 0.4294820725917816, "learning_rate": 9.848064998302531e-05, "loss": 0.0512, "step": 9980 }, { "grad_norm": 0.5720513463020325, "learning_rate": 9.847558945014386e-05, "loss": 0.0562, "step": 9990 }, { "grad_norm": 0.7633821368217468, "learning_rate": 9.847052063410996e-05, "loss": 0.0553, "step": 10000 }, { "grad_norm": 0.4319377839565277, "learning_rate": 9.846544353578977e-05, "loss": 0.0501, "step": 10010 }, { "grad_norm": 0.4610116481781006, "learning_rate": 9.846035815605081e-05, "loss": 0.0595, "step": 10020 }, { "grad_norm": 0.46035197377204895, "learning_rate": 9.845526449576204e-05, "loss": 0.0496, "step": 10030 }, { "grad_norm": 0.5759092569351196, "learning_rate": 9.845016255579383e-05, "loss": 0.0612, "step": 10040 }, { "grad_norm": 0.5021308660507202, "learning_rate": 9.844505233701794e-05, "loss": 0.0575, "step": 10050 }, { "grad_norm": 0.42679911851882935, "learning_rate": 9.843993384030757e-05, "loss": 0.0552, "step": 10060 }, { "grad_norm": 0.30864623188972473, "learning_rate": 9.843480706653737e-05, "loss": 0.0356, "step": 10070 }, { "grad_norm": 0.31764814257621765, "learning_rate": 9.84296720165833e-05, "loss": 0.0493, "step": 10080 }, { "grad_norm": 0.627045214176178, "learning_rate": 9.842452869132286e-05, "loss": 0.0483, "step": 10090 }, { "grad_norm": 0.4155269265174866, "learning_rate": 9.841937709163489e-05, "loss": 0.0538, "step": 10100 }, { "grad_norm": 0.4845089614391327, "learning_rate": 9.841421721839962e-05, "loss": 0.0387, "step": 10110 }, { "grad_norm": 0.27702760696411133, "learning_rate": 9.840904907249879e-05, "loss": 0.0483, "step": 10120 }, { "grad_norm": 0.49400594830513, "learning_rate": 9.840387265481545e-05, "loss": 0.0608, "step": 10130 }, { "grad_norm": 0.6243187785148621, "learning_rate": 9.839868796623411e-05, "loss": 0.0458, "step": 10140 }, { "grad_norm": 0.3940747082233429, "learning_rate": 9.839349500764072e-05, "loss": 0.0578, "step": 10150 }, { "grad_norm": 0.6179128885269165, "learning_rate": 9.83882937799226e-05, "loss": 0.0575, "step": 10160 }, { "grad_norm": 0.48165756464004517, "learning_rate": 9.838308428396849e-05, "loss": 0.0511, "step": 10170 }, { "grad_norm": 0.44408148527145386, "learning_rate": 9.837786652066854e-05, "loss": 0.0672, "step": 10180 }, { "grad_norm": 0.5279368758201599, "learning_rate": 9.837264049091437e-05, "loss": 0.0637, "step": 10190 }, { "grad_norm": 0.45702141523361206, "learning_rate": 9.836740619559893e-05, "loss": 0.0555, "step": 10200 }, { "grad_norm": 0.6705712080001831, "learning_rate": 9.836216363561659e-05, "loss": 0.0568, "step": 10210 }, { "grad_norm": 0.5475910902023315, "learning_rate": 9.835691281186322e-05, "loss": 0.0447, "step": 10220 }, { "grad_norm": 0.40376588702201843, "learning_rate": 9.8351653725236e-05, "loss": 0.0488, "step": 10230 }, { "grad_norm": 0.3318358361721039, "learning_rate": 9.83463863766336e-05, "loss": 0.0529, "step": 10240 }, { "grad_norm": 0.3394371569156647, "learning_rate": 9.834111076695602e-05, "loss": 0.0457, "step": 10250 }, { "grad_norm": 0.4319875240325928, "learning_rate": 9.833582689710477e-05, "loss": 0.0478, "step": 10260 }, { "grad_norm": 0.5659420490264893, "learning_rate": 9.833053476798268e-05, "loss": 0.0555, "step": 10270 }, { "grad_norm": 0.5240433216094971, "learning_rate": 9.832523438049404e-05, "loss": 0.0543, "step": 10280 }, { "grad_norm": 0.4450834095478058, "learning_rate": 9.831992573554454e-05, "loss": 0.0626, "step": 10290 }, { "grad_norm": 0.42510509490966797, "learning_rate": 9.831460883404128e-05, "loss": 0.041, "step": 10300 }, { "grad_norm": 0.4553431570529938, "learning_rate": 9.830928367689278e-05, "loss": 0.041, "step": 10310 }, { "grad_norm": 0.37032991647720337, "learning_rate": 9.830395026500896e-05, "loss": 0.0468, "step": 10320 }, { "grad_norm": 0.3735310137271881, "learning_rate": 9.829860859930115e-05, "loss": 0.0538, "step": 10330 }, { "grad_norm": 0.6488659381866455, "learning_rate": 9.829325868068212e-05, "loss": 0.0563, "step": 10340 }, { "grad_norm": 0.46496742963790894, "learning_rate": 9.8287900510066e-05, "loss": 0.0637, "step": 10350 }, { "grad_norm": 0.3297616243362427, "learning_rate": 9.828253408836834e-05, "loss": 0.0437, "step": 10360 }, { "grad_norm": 0.39663517475128174, "learning_rate": 9.827715941650615e-05, "loss": 0.0475, "step": 10370 }, { "grad_norm": 0.4596486985683441, "learning_rate": 9.82717764953978e-05, "loss": 0.057, "step": 10380 }, { "grad_norm": 0.39309149980545044, "learning_rate": 9.826638532596308e-05, "loss": 0.0625, "step": 10390 }, { "grad_norm": 0.47154322266578674, "learning_rate": 9.82609859091232e-05, "loss": 0.0644, "step": 10400 }, { "grad_norm": 0.6219792366027832, "learning_rate": 9.825557824580076e-05, "loss": 0.0603, "step": 10410 }, { "grad_norm": 0.513633131980896, "learning_rate": 9.82501623369198e-05, "loss": 0.0596, "step": 10420 }, { "grad_norm": 0.43688932061195374, "learning_rate": 9.824473818340574e-05, "loss": 0.0478, "step": 10430 }, { "grad_norm": 0.4610980451107025, "learning_rate": 9.823930578618541e-05, "loss": 0.0516, "step": 10440 }, { "grad_norm": 0.4400478005409241, "learning_rate": 9.823386514618709e-05, "loss": 0.0703, "step": 10450 }, { "grad_norm": 0.45910319685935974, "learning_rate": 9.82284162643404e-05, "loss": 0.0483, "step": 10460 }, { "grad_norm": 0.36168453097343445, "learning_rate": 9.822295914157642e-05, "loss": 0.0477, "step": 10470 }, { "grad_norm": 0.527407169342041, "learning_rate": 9.821749377882763e-05, "loss": 0.0555, "step": 10480 }, { "grad_norm": 0.41103219985961914, "learning_rate": 9.821202017702791e-05, "loss": 0.0486, "step": 10490 }, { "grad_norm": 0.5951488614082336, "learning_rate": 9.820653833711253e-05, "loss": 0.0454, "step": 10500 }, { "grad_norm": 0.5369540452957153, "learning_rate": 9.820104826001822e-05, "loss": 0.0553, "step": 10510 }, { "grad_norm": 0.6055701971054077, "learning_rate": 9.819554994668305e-05, "loss": 0.0498, "step": 10520 }, { "grad_norm": 0.49371805787086487, "learning_rate": 9.819004339804654e-05, "loss": 0.0448, "step": 10530 }, { "grad_norm": 0.46248191595077515, "learning_rate": 9.818452861504961e-05, "loss": 0.0484, "step": 10540 }, { "grad_norm": 0.42754706740379333, "learning_rate": 9.81790055986346e-05, "loss": 0.0434, "step": 10550 }, { "grad_norm": 0.4226903021335602, "learning_rate": 9.817347434974523e-05, "loss": 0.0399, "step": 10560 }, { "grad_norm": 0.4252609312534332, "learning_rate": 9.816793486932664e-05, "loss": 0.0425, "step": 10570 }, { "grad_norm": 0.7045584917068481, "learning_rate": 9.816238715832538e-05, "loss": 0.0492, "step": 10580 }, { "grad_norm": 0.4034467339515686, "learning_rate": 9.815683121768939e-05, "loss": 0.0478, "step": 10590 }, { "grad_norm": 0.3963279128074646, "learning_rate": 9.815126704836804e-05, "loss": 0.0588, "step": 10600 }, { "grad_norm": 0.31646251678466797, "learning_rate": 9.81456946513121e-05, "loss": 0.0703, "step": 10610 }, { "grad_norm": 0.5946424603462219, "learning_rate": 9.814011402747373e-05, "loss": 0.0603, "step": 10620 }, { "grad_norm": 0.3262813985347748, "learning_rate": 9.813452517780651e-05, "loss": 0.0477, "step": 10630 }, { "grad_norm": 0.4570903480052948, "learning_rate": 9.81289281032654e-05, "loss": 0.0539, "step": 10640 }, { "grad_norm": 0.5368927121162415, "learning_rate": 9.812332280480683e-05, "loss": 0.057, "step": 10650 }, { "grad_norm": 0.373793363571167, "learning_rate": 9.811770928338854e-05, "loss": 0.0495, "step": 10660 }, { "grad_norm": 0.6610448360443115, "learning_rate": 9.811208753996979e-05, "loss": 0.0547, "step": 10670 }, { "grad_norm": 0.41290536522865295, "learning_rate": 9.810645757551113e-05, "loss": 0.0398, "step": 10680 }, { "grad_norm": 0.5110064148902893, "learning_rate": 9.810081939097459e-05, "loss": 0.0475, "step": 10690 }, { "grad_norm": 0.5423712134361267, "learning_rate": 9.809517298732356e-05, "loss": 0.074, "step": 10700 }, { "grad_norm": 0.5035714507102966, "learning_rate": 9.80895183655229e-05, "loss": 0.0558, "step": 10710 }, { "grad_norm": 0.5366937518119812, "learning_rate": 9.808385552653877e-05, "loss": 0.0479, "step": 10720 }, { "grad_norm": 0.48906081914901733, "learning_rate": 9.807818447133886e-05, "loss": 0.0433, "step": 10730 }, { "grad_norm": 0.3975168466567993, "learning_rate": 9.807250520089215e-05, "loss": 0.0393, "step": 10740 }, { "grad_norm": 0.4815812408924103, "learning_rate": 9.806681771616908e-05, "loss": 0.0519, "step": 10750 }, { "grad_norm": 0.432318776845932, "learning_rate": 9.80611220181415e-05, "loss": 0.057, "step": 10760 }, { "grad_norm": 0.6417034864425659, "learning_rate": 9.805541810778264e-05, "loss": 0.0493, "step": 10770 }, { "grad_norm": 0.4975537955760956, "learning_rate": 9.804970598606716e-05, "loss": 0.055, "step": 10780 }, { "grad_norm": 0.5257596373558044, "learning_rate": 9.804398565397106e-05, "loss": 0.0544, "step": 10790 }, { "grad_norm": 0.5142531991004944, "learning_rate": 9.803825711247183e-05, "loss": 0.0562, "step": 10800 }, { "grad_norm": 0.5232230424880981, "learning_rate": 9.803252036254831e-05, "loss": 0.0415, "step": 10810 }, { "grad_norm": 0.4324894845485687, "learning_rate": 9.802677540518076e-05, "loss": 0.0501, "step": 10820 }, { "grad_norm": 0.6407973170280457, "learning_rate": 9.802102224135081e-05, "loss": 0.0518, "step": 10830 }, { "grad_norm": 0.5269061326980591, "learning_rate": 9.801526087204155e-05, "loss": 0.0527, "step": 10840 }, { "grad_norm": 0.7198730111122131, "learning_rate": 9.800949129823743e-05, "loss": 0.0495, "step": 10850 }, { "grad_norm": 0.4258001446723938, "learning_rate": 9.80037135209243e-05, "loss": 0.0394, "step": 10860 }, { "grad_norm": 0.6036926507949829, "learning_rate": 9.799792754108946e-05, "loss": 0.0448, "step": 10870 }, { "grad_norm": 0.4689074158668518, "learning_rate": 9.799213335972152e-05, "loss": 0.0493, "step": 10880 }, { "grad_norm": 0.4446744918823242, "learning_rate": 9.798633097781058e-05, "loss": 0.0515, "step": 10890 }, { "grad_norm": 0.5057311058044434, "learning_rate": 9.79805203963481e-05, "loss": 0.0476, "step": 10900 }, { "grad_norm": 0.5434181690216064, "learning_rate": 9.797470161632697e-05, "loss": 0.0473, "step": 10910 }, { "grad_norm": 0.30044451355934143, "learning_rate": 9.796887463874145e-05, "loss": 0.0442, "step": 10920 }, { "grad_norm": 0.23470738530158997, "learning_rate": 9.796303946458718e-05, "loss": 0.0481, "step": 10930 }, { "grad_norm": 0.5993635654449463, "learning_rate": 9.795719609486127e-05, "loss": 0.0532, "step": 10940 }, { "grad_norm": 0.43836960196495056, "learning_rate": 9.795134453056219e-05, "loss": 0.0461, "step": 10950 }, { "grad_norm": 0.4660307466983795, "learning_rate": 9.794548477268979e-05, "loss": 0.0392, "step": 10960 }, { "grad_norm": 0.3843594491481781, "learning_rate": 9.793961682224537e-05, "loss": 0.0387, "step": 10970 }, { "grad_norm": 0.5778015851974487, "learning_rate": 9.793374068023156e-05, "loss": 0.0565, "step": 10980 }, { "grad_norm": 0.38893526792526245, "learning_rate": 9.792785634765247e-05, "loss": 0.047, "step": 10990 }, { "grad_norm": 0.3113986551761627, "learning_rate": 9.792196382551357e-05, "loss": 0.0431, "step": 11000 }, { "grad_norm": 0.36130639910697937, "learning_rate": 9.791606311482171e-05, "loss": 0.051, "step": 11010 }, { "grad_norm": 0.39506974816322327, "learning_rate": 9.791015421658518e-05, "loss": 0.0496, "step": 11020 }, { "grad_norm": 0.645836591720581, "learning_rate": 9.790423713181362e-05, "loss": 0.0519, "step": 11030 }, { "grad_norm": 0.49609577655792236, "learning_rate": 9.789831186151814e-05, "loss": 0.0482, "step": 11040 }, { "grad_norm": 0.5652279853820801, "learning_rate": 9.789237840671118e-05, "loss": 0.0475, "step": 11050 }, { "grad_norm": 0.3735578656196594, "learning_rate": 9.78864367684066e-05, "loss": 0.0497, "step": 11060 }, { "grad_norm": 0.3390207588672638, "learning_rate": 9.788048694761968e-05, "loss": 0.0396, "step": 11070 }, { "grad_norm": 0.3405294418334961, "learning_rate": 9.787452894536709e-05, "loss": 0.0511, "step": 11080 }, { "grad_norm": 0.45457372069358826, "learning_rate": 9.786856276266685e-05, "loss": 0.046, "step": 11090 }, { "grad_norm": 0.3287164568901062, "learning_rate": 9.786258840053845e-05, "loss": 0.0512, "step": 11100 }, { "grad_norm": 0.44201430678367615, "learning_rate": 9.785660586000273e-05, "loss": 0.0441, "step": 11110 }, { "grad_norm": 0.44964176416397095, "learning_rate": 9.785061514208196e-05, "loss": 0.0397, "step": 11120 }, { "grad_norm": 0.44172510504722595, "learning_rate": 9.784461624779977e-05, "loss": 0.0524, "step": 11130 }, { "grad_norm": 0.2675134837627411, "learning_rate": 9.783860917818123e-05, "loss": 0.0451, "step": 11140 }, { "grad_norm": 0.3250475823879242, "learning_rate": 9.783259393425277e-05, "loss": 0.0367, "step": 11150 }, { "grad_norm": 0.3961936831474304, "learning_rate": 9.782657051704221e-05, "loss": 0.0357, "step": 11160 }, { "grad_norm": 0.3749127984046936, "learning_rate": 9.782053892757883e-05, "loss": 0.041, "step": 11170 }, { "grad_norm": 0.5333904027938843, "learning_rate": 9.781449916689324e-05, "loss": 0.042, "step": 11180 }, { "grad_norm": 0.4246068298816681, "learning_rate": 9.780845123601746e-05, "loss": 0.0506, "step": 11190 }, { "grad_norm": 0.49661004543304443, "learning_rate": 9.780239513598492e-05, "loss": 0.0511, "step": 11200 }, { "grad_norm": 0.5034843683242798, "learning_rate": 9.779633086783047e-05, "loss": 0.049, "step": 11210 }, { "grad_norm": 0.3437368869781494, "learning_rate": 9.779025843259031e-05, "loss": 0.0586, "step": 11220 }, { "grad_norm": 0.6332435011863708, "learning_rate": 9.778417783130204e-05, "loss": 0.0659, "step": 11230 }, { "grad_norm": 0.5026386380195618, "learning_rate": 9.777808906500468e-05, "loss": 0.0455, "step": 11240 }, { "grad_norm": 0.6212897300720215, "learning_rate": 9.777199213473862e-05, "loss": 0.0467, "step": 11250 }, { "grad_norm": 0.4051917791366577, "learning_rate": 9.77658870415457e-05, "loss": 0.0463, "step": 11260 }, { "grad_norm": 0.48237425088882446, "learning_rate": 9.775977378646906e-05, "loss": 0.0559, "step": 11270 }, { "grad_norm": 0.4217418134212494, "learning_rate": 9.775365237055331e-05, "loss": 0.0693, "step": 11280 }, { "grad_norm": 0.28753232955932617, "learning_rate": 9.774752279484445e-05, "loss": 0.0589, "step": 11290 }, { "grad_norm": 0.6554321646690369, "learning_rate": 9.774138506038984e-05, "loss": 0.0518, "step": 11300 }, { "grad_norm": 0.40593358874320984, "learning_rate": 9.773523916823826e-05, "loss": 0.048, "step": 11310 }, { "grad_norm": 0.4244164228439331, "learning_rate": 9.772908511943986e-05, "loss": 0.0421, "step": 11320 }, { "grad_norm": 0.29686084389686584, "learning_rate": 9.77229229150462e-05, "loss": 0.0467, "step": 11330 }, { "grad_norm": 0.48995181918144226, "learning_rate": 9.771675255611024e-05, "loss": 0.041, "step": 11340 }, { "grad_norm": 0.4213000535964966, "learning_rate": 9.771057404368632e-05, "loss": 0.068, "step": 11350 }, { "grad_norm": 0.48507222533226013, "learning_rate": 9.770438737883018e-05, "loss": 0.0453, "step": 11360 }, { "grad_norm": 0.4159007668495178, "learning_rate": 9.769819256259898e-05, "loss": 0.0416, "step": 11370 }, { "grad_norm": 0.39792370796203613, "learning_rate": 9.769198959605119e-05, "loss": 0.0474, "step": 11380 }, { "grad_norm": 0.44544172286987305, "learning_rate": 9.768577848024678e-05, "loss": 0.0621, "step": 11390 }, { "grad_norm": 0.5486425757408142, "learning_rate": 9.767955921624702e-05, "loss": 0.0592, "step": 11400 }, { "grad_norm": 0.6596385836601257, "learning_rate": 9.767333180511465e-05, "loss": 0.0429, "step": 11410 }, { "grad_norm": 0.4052330255508423, "learning_rate": 9.766709624791373e-05, "loss": 0.0626, "step": 11420 }, { "grad_norm": 0.34324440360069275, "learning_rate": 9.766085254570975e-05, "loss": 0.0476, "step": 11430 }, { "grad_norm": 0.3858569860458374, "learning_rate": 9.76546006995696e-05, "loss": 0.0419, "step": 11440 }, { "grad_norm": 0.33357134461402893, "learning_rate": 9.764834071056155e-05, "loss": 0.049, "step": 11450 }, { "grad_norm": 0.6635075807571411, "learning_rate": 9.764207257975526e-05, "loss": 0.0634, "step": 11460 }, { "grad_norm": 0.46804848313331604, "learning_rate": 9.763579630822179e-05, "loss": 0.0527, "step": 11470 }, { "grad_norm": 0.5023922324180603, "learning_rate": 9.762951189703356e-05, "loss": 0.048, "step": 11480 }, { "grad_norm": 0.6008262634277344, "learning_rate": 9.762321934726442e-05, "loss": 0.0582, "step": 11490 }, { "grad_norm": 0.4047873020172119, "learning_rate": 9.761691865998959e-05, "loss": 0.0438, "step": 11500 }, { "grad_norm": 0.47083261609077454, "learning_rate": 9.76106098362857e-05, "loss": 0.0463, "step": 11510 }, { "grad_norm": 0.4266377389431, "learning_rate": 9.760429287723072e-05, "loss": 0.0481, "step": 11520 }, { "grad_norm": 0.48058372735977173, "learning_rate": 9.759796778390406e-05, "loss": 0.0471, "step": 11530 }, { "grad_norm": 0.4427419602870941, "learning_rate": 9.759163455738653e-05, "loss": 0.0455, "step": 11540 }, { "grad_norm": 0.5068089365959167, "learning_rate": 9.75852931987603e-05, "loss": 0.0421, "step": 11550 }, { "grad_norm": 0.5941742062568665, "learning_rate": 9.757894370910891e-05, "loss": 0.0465, "step": 11560 }, { "grad_norm": 0.3702211081981659, "learning_rate": 9.757258608951733e-05, "loss": 0.0361, "step": 11570 }, { "grad_norm": 0.4366120994091034, "learning_rate": 9.75662203410719e-05, "loss": 0.0487, "step": 11580 }, { "grad_norm": 0.49517887830734253, "learning_rate": 9.755984646486034e-05, "loss": 0.0484, "step": 11590 }, { "grad_norm": 0.4518824815750122, "learning_rate": 9.75534644619718e-05, "loss": 0.0447, "step": 11600 }, { "grad_norm": 0.3560795485973358, "learning_rate": 9.754707433349676e-05, "loss": 0.0591, "step": 11610 }, { "grad_norm": 0.5232295989990234, "learning_rate": 9.754067608052715e-05, "loss": 0.0548, "step": 11620 }, { "grad_norm": 0.669028639793396, "learning_rate": 9.753426970415622e-05, "loss": 0.0594, "step": 11630 }, { "grad_norm": 0.4726513624191284, "learning_rate": 9.752785520547868e-05, "loss": 0.0441, "step": 11640 }, { "grad_norm": 0.547912061214447, "learning_rate": 9.752143258559056e-05, "loss": 0.0481, "step": 11650 }, { "grad_norm": 0.32576367259025574, "learning_rate": 9.751500184558933e-05, "loss": 0.0509, "step": 11660 }, { "grad_norm": 0.4598951041698456, "learning_rate": 9.750856298657383e-05, "loss": 0.036, "step": 11670 }, { "grad_norm": 0.5794382691383362, "learning_rate": 9.750211600964428e-05, "loss": 0.0452, "step": 11680 }, { "grad_norm": 0.40385982394218445, "learning_rate": 9.749566091590226e-05, "loss": 0.0423, "step": 11690 }, { "grad_norm": 0.46921855211257935, "learning_rate": 9.748919770645083e-05, "loss": 0.0445, "step": 11700 }, { "grad_norm": 0.4037110507488251, "learning_rate": 9.748272638239432e-05, "loss": 0.0441, "step": 11710 }, { "grad_norm": 0.5926459431648254, "learning_rate": 9.747624694483855e-05, "loss": 0.047, "step": 11720 }, { "grad_norm": 0.2759370803833008, "learning_rate": 9.746975939489065e-05, "loss": 0.0408, "step": 11730 }, { "grad_norm": 0.37114787101745605, "learning_rate": 9.746326373365918e-05, "loss": 0.0504, "step": 11740 }, { "grad_norm": 0.43141084909439087, "learning_rate": 9.745675996225403e-05, "loss": 0.0506, "step": 11750 }, { "grad_norm": 0.4411022663116455, "learning_rate": 9.745024808178657e-05, "loss": 0.0451, "step": 11760 }, { "grad_norm": 0.39446988701820374, "learning_rate": 9.744372809336947e-05, "loss": 0.0536, "step": 11770 }, { "grad_norm": 0.4688822329044342, "learning_rate": 9.743719999811682e-05, "loss": 0.0403, "step": 11780 }, { "grad_norm": 0.4308195412158966, "learning_rate": 9.743066379714412e-05, "loss": 0.0408, "step": 11790 }, { "grad_norm": 0.4334006905555725, "learning_rate": 9.74241194915682e-05, "loss": 0.0588, "step": 11800 }, { "grad_norm": 0.349723756313324, "learning_rate": 9.741756708250731e-05, "loss": 0.0566, "step": 11810 }, { "grad_norm": 0.3696272075176239, "learning_rate": 9.741100657108109e-05, "loss": 0.0484, "step": 11820 }, { "grad_norm": 0.4345083236694336, "learning_rate": 9.740443795841054e-05, "loss": 0.0622, "step": 11830 }, { "grad_norm": 0.442534863948822, "learning_rate": 9.739786124561805e-05, "loss": 0.0464, "step": 11840 }, { "grad_norm": 0.32108765840530396, "learning_rate": 9.73912764338274e-05, "loss": 0.0478, "step": 11850 }, { "grad_norm": 0.3149438202381134, "learning_rate": 9.738468352416377e-05, "loss": 0.051, "step": 11860 }, { "grad_norm": 0.3802957832813263, "learning_rate": 9.737808251775369e-05, "loss": 0.0475, "step": 11870 }, { "grad_norm": 0.33676156401634216, "learning_rate": 9.737147341572512e-05, "loss": 0.058, "step": 11880 }, { "grad_norm": 0.5085352063179016, "learning_rate": 9.736485621920735e-05, "loss": 0.0376, "step": 11890 }, { "grad_norm": 0.41333237290382385, "learning_rate": 9.735823092933108e-05, "loss": 0.0431, "step": 11900 }, { "grad_norm": 0.43696531653404236, "learning_rate": 9.735159754722838e-05, "loss": 0.0439, "step": 11910 }, { "grad_norm": 0.5265536308288574, "learning_rate": 9.734495607403275e-05, "loss": 0.0365, "step": 11920 }, { "grad_norm": 0.3086473047733307, "learning_rate": 9.733830651087901e-05, "loss": 0.0444, "step": 11930 }, { "grad_norm": 0.4668741822242737, "learning_rate": 9.733164885890338e-05, "loss": 0.0504, "step": 11940 }, { "grad_norm": 0.43095120787620544, "learning_rate": 9.732498311924349e-05, "loss": 0.0375, "step": 11950 }, { "grad_norm": 0.5285136103630066, "learning_rate": 9.731830929303833e-05, "loss": 0.0459, "step": 11960 }, { "grad_norm": 0.4531131088733673, "learning_rate": 9.731162738142827e-05, "loss": 0.0401, "step": 11970 }, { "grad_norm": 0.40627521276474, "learning_rate": 9.730493738555506e-05, "loss": 0.0405, "step": 11980 }, { "grad_norm": 0.3795287311077118, "learning_rate": 9.729823930656186e-05, "loss": 0.0423, "step": 11990 }, { "grad_norm": 0.6239099502563477, "learning_rate": 9.729153314559316e-05, "loss": 0.0583, "step": 12000 }, { "grad_norm": 0.46024060249328613, "learning_rate": 9.728481890379486e-05, "loss": 0.0502, "step": 12010 }, { "grad_norm": 0.4351509213447571, "learning_rate": 9.727809658231428e-05, "loss": 0.0459, "step": 12020 }, { "grad_norm": 0.5162484645843506, "learning_rate": 9.727136618230003e-05, "loss": 0.0481, "step": 12030 }, { "grad_norm": 0.6180964708328247, "learning_rate": 9.726462770490219e-05, "loss": 0.0491, "step": 12040 }, { "grad_norm": 0.4320570230484009, "learning_rate": 9.725788115127214e-05, "loss": 0.052, "step": 12050 }, { "grad_norm": 0.3718518912792206, "learning_rate": 9.725112652256274e-05, "loss": 0.0439, "step": 12060 }, { "grad_norm": 0.37903255224227905, "learning_rate": 9.724436381992812e-05, "loss": 0.0593, "step": 12070 }, { "grad_norm": 0.5695672035217285, "learning_rate": 9.723759304452387e-05, "loss": 0.0465, "step": 12080 }, { "grad_norm": 0.40049412846565247, "learning_rate": 9.72308141975069e-05, "loss": 0.0391, "step": 12090 }, { "grad_norm": 0.4546939432621002, "learning_rate": 9.722402728003557e-05, "loss": 0.0484, "step": 12100 }, { "grad_norm": 0.4344910681247711, "learning_rate": 9.721723229326953e-05, "loss": 0.0355, "step": 12110 }, { "grad_norm": 0.402415007352829, "learning_rate": 9.721042923836992e-05, "loss": 0.0359, "step": 12120 }, { "grad_norm": 0.43607816100120544, "learning_rate": 9.720361811649914e-05, "loss": 0.0411, "step": 12130 }, { "grad_norm": 0.31958648562431335, "learning_rate": 9.719679892882106e-05, "loss": 0.035, "step": 12140 }, { "grad_norm": 0.35361799597740173, "learning_rate": 9.718997167650085e-05, "loss": 0.0407, "step": 12150 }, { "grad_norm": 0.3615454435348511, "learning_rate": 9.718313636070515e-05, "loss": 0.0367, "step": 12160 }, { "grad_norm": 0.35603126883506775, "learning_rate": 9.717629298260192e-05, "loss": 0.0442, "step": 12170 }, { "grad_norm": 0.5075443983078003, "learning_rate": 9.716944154336047e-05, "loss": 0.0547, "step": 12180 }, { "grad_norm": 0.5382558107376099, "learning_rate": 9.716258204415157e-05, "loss": 0.0411, "step": 12190 }, { "grad_norm": 0.4613151550292969, "learning_rate": 9.715571448614728e-05, "loss": 0.0598, "step": 12200 }, { "grad_norm": 0.48829102516174316, "learning_rate": 9.71488388705211e-05, "loss": 0.0612, "step": 12210 }, { "grad_norm": 0.4793728291988373, "learning_rate": 9.714195519844788e-05, "loss": 0.0461, "step": 12220 }, { "grad_norm": 0.42521366477012634, "learning_rate": 9.713506347110386e-05, "loss": 0.0383, "step": 12230 }, { "grad_norm": 0.4048771560192108, "learning_rate": 9.712816368966662e-05, "loss": 0.0407, "step": 12240 }, { "grad_norm": 0.3963131010532379, "learning_rate": 9.712125585531517e-05, "loss": 0.0616, "step": 12250 }, { "grad_norm": 0.6164613366127014, "learning_rate": 9.711433996922988e-05, "loss": 0.0601, "step": 12260 }, { "grad_norm": 0.32673078775405884, "learning_rate": 9.710741603259245e-05, "loss": 0.0534, "step": 12270 }, { "grad_norm": 0.4662962257862091, "learning_rate": 9.710048404658603e-05, "loss": 0.04, "step": 12280 }, { "grad_norm": 0.329112708568573, "learning_rate": 9.709354401239508e-05, "loss": 0.0518, "step": 12290 }, { "grad_norm": 0.48763564229011536, "learning_rate": 9.708659593120546e-05, "loss": 0.0433, "step": 12300 }, { "grad_norm": 0.4290774166584015, "learning_rate": 9.707963980420443e-05, "loss": 0.0528, "step": 12310 }, { "grad_norm": 0.47171545028686523, "learning_rate": 9.707267563258058e-05, "loss": 0.055, "step": 12320 }, { "grad_norm": 0.46055251359939575, "learning_rate": 9.70657034175239e-05, "loss": 0.0468, "step": 12330 }, { "grad_norm": 0.48109114170074463, "learning_rate": 9.705872316022577e-05, "loss": 0.0513, "step": 12340 }, { "grad_norm": 0.4497084617614746, "learning_rate": 9.705173486187891e-05, "loss": 0.043, "step": 12350 }, { "grad_norm": 0.35527318716049194, "learning_rate": 9.704473852367741e-05, "loss": 0.0516, "step": 12360 }, { "grad_norm": 0.42453351616859436, "learning_rate": 9.70377341468168e-05, "loss": 0.0395, "step": 12370 }, { "grad_norm": 0.44832736253738403, "learning_rate": 9.703072173249389e-05, "loss": 0.0461, "step": 12380 }, { "grad_norm": 1.8208565711975098, "learning_rate": 9.702370128190693e-05, "loss": 0.0534, "step": 12390 }, { "grad_norm": 0.4512990117073059, "learning_rate": 9.701667279625552e-05, "loss": 0.0566, "step": 12400 }, { "grad_norm": 0.7565293908119202, "learning_rate": 9.700963627674065e-05, "loss": 0.0591, "step": 12410 }, { "grad_norm": 0.5273473858833313, "learning_rate": 9.700259172456466e-05, "loss": 0.0476, "step": 12420 }, { "grad_norm": 0.5885107517242432, "learning_rate": 9.699553914093124e-05, "loss": 0.0652, "step": 12430 }, { "grad_norm": 0.42144250869750977, "learning_rate": 9.698847852704553e-05, "loss": 0.0556, "step": 12440 }, { "grad_norm": 0.42519626021385193, "learning_rate": 9.6981409884114e-05, "loss": 0.0424, "step": 12450 }, { "grad_norm": 0.30960312485694885, "learning_rate": 9.697433321334443e-05, "loss": 0.0527, "step": 12460 }, { "grad_norm": 0.37791669368743896, "learning_rate": 9.696724851594607e-05, "loss": 0.0448, "step": 12470 }, { "grad_norm": 0.3537672460079193, "learning_rate": 9.696015579312952e-05, "loss": 0.0347, "step": 12480 }, { "grad_norm": 0.3393002450466156, "learning_rate": 9.695305504610668e-05, "loss": 0.0416, "step": 12490 }, { "grad_norm": 0.42668187618255615, "learning_rate": 9.694594627609092e-05, "loss": 0.0473, "step": 12500 }, { "grad_norm": 0.5566374659538269, "learning_rate": 9.693882948429691e-05, "loss": 0.0588, "step": 12510 }, { "grad_norm": 0.5025665163993835, "learning_rate": 9.693170467194071e-05, "loss": 0.0495, "step": 12520 }, { "grad_norm": 0.4822729527950287, "learning_rate": 9.692457184023977e-05, "loss": 0.0439, "step": 12530 }, { "grad_norm": 0.5405736565589905, "learning_rate": 9.691743099041291e-05, "loss": 0.0546, "step": 12540 }, { "grad_norm": 0.44858691096305847, "learning_rate": 9.691028212368027e-05, "loss": 0.0459, "step": 12550 }, { "grad_norm": 0.6523935198783875, "learning_rate": 9.690312524126342e-05, "loss": 0.0479, "step": 12560 }, { "grad_norm": 0.3382337987422943, "learning_rate": 9.689596034438527e-05, "loss": 0.0471, "step": 12570 }, { "grad_norm": 0.40521419048309326, "learning_rate": 9.688878743427012e-05, "loss": 0.0419, "step": 12580 }, { "grad_norm": 0.3501238524913788, "learning_rate": 9.688160651214359e-05, "loss": 0.0503, "step": 12590 }, { "grad_norm": 0.5107623934745789, "learning_rate": 9.687441757923273e-05, "loss": 0.0459, "step": 12600 }, { "grad_norm": 0.4555400013923645, "learning_rate": 9.68672206367659e-05, "loss": 0.0407, "step": 12610 }, { "grad_norm": 0.36875706911087036, "learning_rate": 9.686001568597291e-05, "loss": 0.0533, "step": 12620 }, { "grad_norm": 0.43235665559768677, "learning_rate": 9.685280272808486e-05, "loss": 0.0402, "step": 12630 }, { "grad_norm": 0.5264152884483337, "learning_rate": 9.684558176433424e-05, "loss": 0.0538, "step": 12640 }, { "grad_norm": 0.31753963232040405, "learning_rate": 9.683835279595495e-05, "loss": 0.0409, "step": 12650 }, { "grad_norm": 0.3703649640083313, "learning_rate": 9.683111582418216e-05, "loss": 0.0451, "step": 12660 }, { "grad_norm": 0.44565415382385254, "learning_rate": 9.682387085025254e-05, "loss": 0.046, "step": 12670 }, { "grad_norm": 0.4067932367324829, "learning_rate": 9.681661787540401e-05, "loss": 0.0479, "step": 12680 }, { "grad_norm": 0.4316827952861786, "learning_rate": 9.680935690087593e-05, "loss": 0.0435, "step": 12690 }, { "grad_norm": 0.7038748860359192, "learning_rate": 9.680208792790901e-05, "loss": 0.048, "step": 12700 }, { "grad_norm": 0.3919718563556671, "learning_rate": 9.679481095774529e-05, "loss": 0.0426, "step": 12710 }, { "grad_norm": 0.5013135075569153, "learning_rate": 9.678752599162822e-05, "loss": 0.0486, "step": 12720 }, { "grad_norm": 0.5682360529899597, "learning_rate": 9.678023303080259e-05, "loss": 0.042, "step": 12730 }, { "grad_norm": 0.3814045190811157, "learning_rate": 9.677293207651459e-05, "loss": 0.0536, "step": 12740 }, { "grad_norm": 0.39721161127090454, "learning_rate": 9.676562313001173e-05, "loss": 0.045, "step": 12750 }, { "grad_norm": 0.662899374961853, "learning_rate": 9.675830619254293e-05, "loss": 0.062, "step": 12760 }, { "grad_norm": 0.42449718713760376, "learning_rate": 9.675098126535843e-05, "loss": 0.0666, "step": 12770 }, { "grad_norm": 0.5459147095680237, "learning_rate": 9.674364834970988e-05, "loss": 0.0616, "step": 12780 }, { "grad_norm": 0.46337103843688965, "learning_rate": 9.673630744685028e-05, "loss": 0.0537, "step": 12790 }, { "grad_norm": 0.4589895009994507, "learning_rate": 9.672895855803397e-05, "loss": 0.0505, "step": 12800 }, { "grad_norm": 0.5184335708618164, "learning_rate": 9.672160168451667e-05, "loss": 0.0471, "step": 12810 }, { "grad_norm": 0.43025219440460205, "learning_rate": 9.671423682755549e-05, "loss": 0.0592, "step": 12820 }, { "grad_norm": 0.6339226961135864, "learning_rate": 9.670686398840888e-05, "loss": 0.0518, "step": 12830 }, { "grad_norm": 0.34990233182907104, "learning_rate": 9.669948316833664e-05, "loss": 0.0534, "step": 12840 }, { "grad_norm": 0.5235701203346252, "learning_rate": 9.669209436859997e-05, "loss": 0.046, "step": 12850 }, { "grad_norm": 0.48227158188819885, "learning_rate": 9.66846975904614e-05, "loss": 0.0543, "step": 12860 }, { "grad_norm": 0.5444728136062622, "learning_rate": 9.667729283518483e-05, "loss": 0.0435, "step": 12870 }, { "grad_norm": 0.5311805009841919, "learning_rate": 9.666988010403557e-05, "loss": 0.0453, "step": 12880 }, { "grad_norm": 0.3869835436344147, "learning_rate": 9.66624593982802e-05, "loss": 0.0441, "step": 12890 }, { "grad_norm": 0.38813072443008423, "learning_rate": 9.665503071918675e-05, "loss": 0.0587, "step": 12900 }, { "grad_norm": 0.36572083830833435, "learning_rate": 9.664759406802456e-05, "loss": 0.0425, "step": 12910 }, { "grad_norm": 0.39501500129699707, "learning_rate": 9.664014944606437e-05, "loss": 0.0362, "step": 12920 }, { "grad_norm": 0.3740244507789612, "learning_rate": 9.663269685457822e-05, "loss": 0.0411, "step": 12930 }, { "grad_norm": 0.5108780264854431, "learning_rate": 9.662523629483962e-05, "loss": 0.0461, "step": 12940 }, { "grad_norm": 0.3474109172821045, "learning_rate": 9.661776776812333e-05, "loss": 0.0358, "step": 12950 }, { "grad_norm": 0.3762547969818115, "learning_rate": 9.661029127570553e-05, "loss": 0.0349, "step": 12960 }, { "grad_norm": 0.32232651114463806, "learning_rate": 9.660280681886373e-05, "loss": 0.0403, "step": 12970 }, { "grad_norm": 0.34926947951316833, "learning_rate": 9.659531439887685e-05, "loss": 0.0358, "step": 12980 }, { "grad_norm": 0.48566603660583496, "learning_rate": 9.658781401702511e-05, "loss": 0.0518, "step": 12990 }, { "grad_norm": 0.43664801120758057, "learning_rate": 9.658030567459015e-05, "loss": 0.0553, "step": 13000 }, { "grad_norm": 0.5171744227409363, "learning_rate": 9.65727893728549e-05, "loss": 0.0735, "step": 13010 }, { "grad_norm": 0.5081170201301575, "learning_rate": 9.656526511310375e-05, "loss": 0.058, "step": 13020 }, { "grad_norm": 0.39553913474082947, "learning_rate": 9.655773289662233e-05, "loss": 0.0523, "step": 13030 }, { "grad_norm": 0.5497028827667236, "learning_rate": 9.655019272469772e-05, "loss": 0.0418, "step": 13040 }, { "grad_norm": 0.5392035841941833, "learning_rate": 9.654264459861832e-05, "loss": 0.0412, "step": 13050 }, { "grad_norm": 0.46611344814300537, "learning_rate": 9.653508851967391e-05, "loss": 0.0438, "step": 13060 }, { "grad_norm": 0.5839679837226868, "learning_rate": 9.65275244891556e-05, "loss": 0.0644, "step": 13070 }, { "grad_norm": 0.7205440998077393, "learning_rate": 9.651995250835591e-05, "loss": 0.0501, "step": 13080 }, { "grad_norm": 0.39983442425727844, "learning_rate": 9.651237257856862e-05, "loss": 0.0547, "step": 13090 }, { "grad_norm": 0.35021132230758667, "learning_rate": 9.6504784701089e-05, "loss": 0.0513, "step": 13100 }, { "grad_norm": 0.3796819746494293, "learning_rate": 9.649718887721357e-05, "loss": 0.0403, "step": 13110 }, { "grad_norm": 0.6112896800041199, "learning_rate": 9.648958510824028e-05, "loss": 0.0448, "step": 13120 }, { "grad_norm": 0.536353349685669, "learning_rate": 9.648197339546837e-05, "loss": 0.0523, "step": 13130 }, { "grad_norm": 0.562343955039978, "learning_rate": 9.647435374019851e-05, "loss": 0.0528, "step": 13140 }, { "grad_norm": 0.6996495723724365, "learning_rate": 9.646672614373266e-05, "loss": 0.0535, "step": 13150 }, { "grad_norm": 0.42320647835731506, "learning_rate": 9.645909060737418e-05, "loss": 0.0361, "step": 13160 }, { "grad_norm": 0.46246299147605896, "learning_rate": 9.645144713242778e-05, "loss": 0.0483, "step": 13170 }, { "grad_norm": 0.35040614008903503, "learning_rate": 9.64437957201995e-05, "loss": 0.0469, "step": 13180 }, { "grad_norm": 0.41117405891418457, "learning_rate": 9.643613637199678e-05, "loss": 0.0423, "step": 13190 }, { "grad_norm": 0.5878121256828308, "learning_rate": 9.642846908912839e-05, "loss": 0.0499, "step": 13200 }, { "grad_norm": 0.41554880142211914, "learning_rate": 9.642079387290444e-05, "loss": 0.0329, "step": 13210 }, { "grad_norm": 0.4275512993335724, "learning_rate": 9.641311072463644e-05, "loss": 0.0421, "step": 13220 }, { "grad_norm": 0.4186107814311981, "learning_rate": 9.640541964563722e-05, "loss": 0.0461, "step": 13230 }, { "grad_norm": 0.5563111901283264, "learning_rate": 9.639772063722096e-05, "loss": 0.0501, "step": 13240 }, { "grad_norm": 0.34290236234664917, "learning_rate": 9.639001370070324e-05, "loss": 0.039, "step": 13250 }, { "grad_norm": 0.4549674093723297, "learning_rate": 9.638229883740095e-05, "loss": 0.0428, "step": 13260 }, { "grad_norm": 0.31732597947120667, "learning_rate": 9.637457604863233e-05, "loss": 0.0388, "step": 13270 }, { "grad_norm": 0.5181941986083984, "learning_rate": 9.636684533571703e-05, "loss": 0.0333, "step": 13280 }, { "grad_norm": 0.5082500576972961, "learning_rate": 9.635910669997599e-05, "loss": 0.0402, "step": 13290 }, { "grad_norm": 0.40272057056427, "learning_rate": 9.635136014273154e-05, "loss": 0.0311, "step": 13300 }, { "grad_norm": 0.36297473311424255, "learning_rate": 9.634360566530735e-05, "loss": 0.0472, "step": 13310 }, { "grad_norm": 0.4636584222316742, "learning_rate": 9.633584326902845e-05, "loss": 0.0569, "step": 13320 }, { "grad_norm": 0.43242236971855164, "learning_rate": 9.632807295522124e-05, "loss": 0.0451, "step": 13330 }, { "grad_norm": 0.3774069845676422, "learning_rate": 9.632029472521342e-05, "loss": 0.0468, "step": 13340 }, { "grad_norm": 0.3135913014411926, "learning_rate": 9.631250858033409e-05, "loss": 0.0402, "step": 13350 }, { "grad_norm": 0.5309874415397644, "learning_rate": 9.630471452191371e-05, "loss": 0.039, "step": 13360 }, { "grad_norm": 0.3733164966106415, "learning_rate": 9.629691255128405e-05, "loss": 0.0411, "step": 13370 }, { "grad_norm": 0.5089719891548157, "learning_rate": 9.628910266977825e-05, "loss": 0.0392, "step": 13380 }, { "grad_norm": 0.3368435204029083, "learning_rate": 9.628128487873083e-05, "loss": 0.0409, "step": 13390 }, { "grad_norm": 0.5148091912269592, "learning_rate": 9.627345917947761e-05, "loss": 0.0501, "step": 13400 }, { "grad_norm": 0.4441317021846771, "learning_rate": 9.626562557335579e-05, "loss": 0.0467, "step": 13410 }, { "grad_norm": 0.35911983251571655, "learning_rate": 9.625778406170393e-05, "loss": 0.0491, "step": 13420 }, { "grad_norm": 0.451761931180954, "learning_rate": 9.624993464586193e-05, "loss": 0.0338, "step": 13430 }, { "grad_norm": 0.4024706780910492, "learning_rate": 9.624207732717105e-05, "loss": 0.0421, "step": 13440 }, { "grad_norm": 0.4892871379852295, "learning_rate": 9.623421210697386e-05, "loss": 0.0534, "step": 13450 }, { "grad_norm": 0.3893268406391144, "learning_rate": 9.622633898661434e-05, "loss": 0.0469, "step": 13460 }, { "grad_norm": 0.39128053188323975, "learning_rate": 9.621845796743778e-05, "loss": 0.0657, "step": 13470 }, { "grad_norm": 0.4808219373226166, "learning_rate": 9.621056905079082e-05, "loss": 0.053, "step": 13480 }, { "grad_norm": 0.5568147897720337, "learning_rate": 9.620267223802149e-05, "loss": 0.0549, "step": 13490 }, { "grad_norm": 0.31196701526641846, "learning_rate": 9.619476753047911e-05, "loss": 0.0319, "step": 13500 }, { "grad_norm": 0.3702337145805359, "learning_rate": 9.618685492951438e-05, "loss": 0.0463, "step": 13510 }, { "grad_norm": 0.49256041646003723, "learning_rate": 9.617893443647938e-05, "loss": 0.0476, "step": 13520 }, { "grad_norm": 0.42328187823295593, "learning_rate": 9.617100605272746e-05, "loss": 0.0488, "step": 13530 }, { "grad_norm": 0.3723820745944977, "learning_rate": 9.616306977961338e-05, "loss": 0.0473, "step": 13540 }, { "grad_norm": 0.31337305903434753, "learning_rate": 9.615512561849326e-05, "loss": 0.0362, "step": 13550 }, { "grad_norm": 0.3384978771209717, "learning_rate": 9.61471735707245e-05, "loss": 0.0505, "step": 13560 }, { "grad_norm": 0.43554461002349854, "learning_rate": 9.613921363766592e-05, "loss": 0.041, "step": 13570 }, { "grad_norm": 0.33470243215560913, "learning_rate": 9.613124582067763e-05, "loss": 0.0407, "step": 13580 }, { "grad_norm": 0.4030573070049286, "learning_rate": 9.612327012112112e-05, "loss": 0.0559, "step": 13590 }, { "grad_norm": 0.43051695823669434, "learning_rate": 9.611528654035921e-05, "loss": 0.0624, "step": 13600 }, { "grad_norm": 0.5503274202346802, "learning_rate": 9.610729507975611e-05, "loss": 0.0435, "step": 13610 }, { "grad_norm": 0.4597894549369812, "learning_rate": 9.609929574067731e-05, "loss": 0.0556, "step": 13620 }, { "grad_norm": 0.30296558141708374, "learning_rate": 9.609128852448967e-05, "loss": 0.038, "step": 13630 }, { "grad_norm": 0.48160648345947266, "learning_rate": 9.608327343256143e-05, "loss": 0.0402, "step": 13640 }, { "grad_norm": 0.4389536678791046, "learning_rate": 9.607525046626216e-05, "loss": 0.0486, "step": 13650 }, { "grad_norm": 0.5151249170303345, "learning_rate": 9.606721962696272e-05, "loss": 0.044, "step": 13660 }, { "grad_norm": 0.4012778699398041, "learning_rate": 9.60591809160354e-05, "loss": 0.0453, "step": 13670 }, { "grad_norm": 0.4461338520050049, "learning_rate": 9.605113433485378e-05, "loss": 0.0451, "step": 13680 }, { "grad_norm": 0.45740100741386414, "learning_rate": 9.604307988479279e-05, "loss": 0.0423, "step": 13690 }, { "grad_norm": 0.402048796415329, "learning_rate": 9.603501756722876e-05, "loss": 0.038, "step": 13700 }, { "grad_norm": 0.6193347573280334, "learning_rate": 9.602694738353927e-05, "loss": 0.0423, "step": 13710 }, { "grad_norm": 0.3762461543083191, "learning_rate": 9.601886933510331e-05, "loss": 0.0364, "step": 13720 }, { "grad_norm": 0.5158107876777649, "learning_rate": 9.60107834233012e-05, "loss": 0.041, "step": 13730 }, { "grad_norm": 0.3009028136730194, "learning_rate": 9.60026896495146e-05, "loss": 0.0324, "step": 13740 }, { "grad_norm": 0.5076273083686829, "learning_rate": 9.599458801512652e-05, "loss": 0.0432, "step": 13750 }, { "grad_norm": 0.47227415442466736, "learning_rate": 9.598647852152129e-05, "loss": 0.0613, "step": 13760 }, { "grad_norm": 0.5235857367515564, "learning_rate": 9.597836117008462e-05, "loss": 0.0446, "step": 13770 }, { "grad_norm": 0.6270670294761658, "learning_rate": 9.597023596220356e-05, "loss": 0.0512, "step": 13780 }, { "grad_norm": 0.4470062553882599, "learning_rate": 9.596210289926643e-05, "loss": 0.0407, "step": 13790 }, { "grad_norm": 0.3561553955078125, "learning_rate": 9.5953961982663e-05, "loss": 0.0453, "step": 13800 }, { "grad_norm": 0.5723956227302551, "learning_rate": 9.594581321378431e-05, "loss": 0.0508, "step": 13810 }, { "grad_norm": 0.49058735370635986, "learning_rate": 9.593765659402276e-05, "loss": 0.0526, "step": 13820 }, { "grad_norm": 0.46443694829940796, "learning_rate": 9.59294921247721e-05, "loss": 0.0512, "step": 13830 }, { "grad_norm": 0.3412626087665558, "learning_rate": 9.59213198074274e-05, "loss": 0.0527, "step": 13840 }, { "grad_norm": 0.4165308475494385, "learning_rate": 9.59131396433851e-05, "loss": 0.0417, "step": 13850 }, { "grad_norm": 0.34208086133003235, "learning_rate": 9.590495163404297e-05, "loss": 0.0392, "step": 13860 }, { "grad_norm": 0.29211896657943726, "learning_rate": 9.589675578080009e-05, "loss": 0.0386, "step": 13870 }, { "grad_norm": 0.3182264268398285, "learning_rate": 9.588855208505694e-05, "loss": 0.0347, "step": 13880 }, { "grad_norm": 0.4530583322048187, "learning_rate": 9.588034054821529e-05, "loss": 0.0497, "step": 13890 }, { "grad_norm": 0.3035835027694702, "learning_rate": 9.587212117167826e-05, "loss": 0.0359, "step": 13900 }, { "grad_norm": 0.5818781852722168, "learning_rate": 9.586389395685033e-05, "loss": 0.0414, "step": 13910 }, { "grad_norm": 0.5828718543052673, "learning_rate": 9.585565890513733e-05, "loss": 0.0475, "step": 13920 }, { "grad_norm": 0.4774647653102875, "learning_rate": 9.584741601794636e-05, "loss": 0.0463, "step": 13930 }, { "grad_norm": 0.5094736218452454, "learning_rate": 9.58391652966859e-05, "loss": 0.044, "step": 13940 }, { "grad_norm": 0.43456289172172546, "learning_rate": 9.583090674276583e-05, "loss": 0.0395, "step": 13950 }, { "grad_norm": 0.3570030629634857, "learning_rate": 9.582264035759726e-05, "loss": 0.043, "step": 13960 }, { "grad_norm": 0.4917267858982086, "learning_rate": 9.58143661425927e-05, "loss": 0.0644, "step": 13970 }, { "grad_norm": 0.5678696036338806, "learning_rate": 9.580608409916601e-05, "loss": 0.0521, "step": 13980 }, { "grad_norm": 0.3232181668281555, "learning_rate": 9.579779422873233e-05, "loss": 0.0432, "step": 13990 }, { "grad_norm": 0.386332631111145, "learning_rate": 9.578949653270819e-05, "loss": 0.0471, "step": 14000 }, { "grad_norm": 0.3461098372936249, "learning_rate": 9.578119101251144e-05, "loss": 0.0387, "step": 14010 }, { "grad_norm": 0.4553775489330292, "learning_rate": 9.577287766956127e-05, "loss": 0.0537, "step": 14020 }, { "grad_norm": 0.3943447768688202, "learning_rate": 9.57645565052782e-05, "loss": 0.0448, "step": 14030 }, { "grad_norm": 0.3930102288722992, "learning_rate": 9.575622752108407e-05, "loss": 0.0391, "step": 14040 }, { "grad_norm": 0.3761933147907257, "learning_rate": 9.57478907184021e-05, "loss": 0.0403, "step": 14050 }, { "grad_norm": 0.4636801481246948, "learning_rate": 9.573954609865681e-05, "loss": 0.0481, "step": 14060 }, { "grad_norm": 0.42332184314727783, "learning_rate": 9.573119366327408e-05, "loss": 0.0445, "step": 14070 }, { "grad_norm": 0.4368940591812134, "learning_rate": 9.57228334136811e-05, "loss": 0.0403, "step": 14080 }, { "grad_norm": 0.36070144176483154, "learning_rate": 9.571446535130641e-05, "loss": 0.037, "step": 14090 }, { "grad_norm": 0.5519856810569763, "learning_rate": 9.570608947757988e-05, "loss": 0.0402, "step": 14100 }, { "grad_norm": 0.32528963685035706, "learning_rate": 9.569770579393274e-05, "loss": 0.0482, "step": 14110 }, { "grad_norm": 0.3489520251750946, "learning_rate": 9.56893143017975e-05, "loss": 0.0435, "step": 14120 }, { "grad_norm": 0.4120558798313141, "learning_rate": 9.568091500260806e-05, "loss": 0.0496, "step": 14130 }, { "grad_norm": 0.3383130729198456, "learning_rate": 9.567250789779961e-05, "loss": 0.0429, "step": 14140 }, { "grad_norm": 0.48926231265068054, "learning_rate": 9.566409298880872e-05, "loss": 0.0421, "step": 14150 }, { "grad_norm": 0.43343958258628845, "learning_rate": 9.565567027707326e-05, "loss": 0.0572, "step": 14160 }, { "grad_norm": 0.42535528540611267, "learning_rate": 9.56472397640324e-05, "loss": 0.0457, "step": 14170 }, { "grad_norm": 0.4117426872253418, "learning_rate": 9.563880145112675e-05, "loss": 0.0429, "step": 14180 }, { "grad_norm": 0.6394652128219604, "learning_rate": 9.563035533979814e-05, "loss": 0.0491, "step": 14190 }, { "grad_norm": 0.41771605610847473, "learning_rate": 9.562190143148981e-05, "loss": 0.0418, "step": 14200 }, { "grad_norm": 0.28595492243766785, "learning_rate": 9.561343972764627e-05, "loss": 0.0354, "step": 14210 }, { "grad_norm": 0.3469604253768921, "learning_rate": 9.560497022971343e-05, "loss": 0.0494, "step": 14220 }, { "grad_norm": 0.31010687351226807, "learning_rate": 9.559649293913847e-05, "loss": 0.0454, "step": 14230 }, { "grad_norm": 0.5529469847679138, "learning_rate": 9.558800785736993e-05, "loss": 0.0491, "step": 14240 }, { "grad_norm": 0.3446567952632904, "learning_rate": 9.557951498585767e-05, "loss": 0.0437, "step": 14250 }, { "grad_norm": 0.2921867370605469, "learning_rate": 9.557101432605293e-05, "loss": 0.0362, "step": 14260 }, { "grad_norm": 0.44929566979408264, "learning_rate": 9.556250587940818e-05, "loss": 0.0505, "step": 14270 }, { "grad_norm": 0.5673851370811462, "learning_rate": 9.555398964737734e-05, "loss": 0.0444, "step": 14280 }, { "grad_norm": 0.39419588446617126, "learning_rate": 9.554546563141555e-05, "loss": 0.0386, "step": 14290 }, { "grad_norm": 0.4689921438694, "learning_rate": 9.553693383297937e-05, "loss": 0.0467, "step": 14300 }, { "grad_norm": 0.45880335569381714, "learning_rate": 9.552839425352663e-05, "loss": 0.0346, "step": 14310 }, { "grad_norm": 0.45106980204582214, "learning_rate": 9.551984689451652e-05, "loss": 0.0482, "step": 14320 }, { "grad_norm": 0.4217597246170044, "learning_rate": 9.551129175740953e-05, "loss": 0.0388, "step": 14330 }, { "grad_norm": 0.39172449707984924, "learning_rate": 9.550272884366754e-05, "loss": 0.0455, "step": 14340 }, { "grad_norm": 0.60927814245224, "learning_rate": 9.549415815475369e-05, "loss": 0.0457, "step": 14350 }, { "grad_norm": 0.31920191645622253, "learning_rate": 9.548557969213247e-05, "loss": 0.0369, "step": 14360 }, { "grad_norm": 0.42385977506637573, "learning_rate": 9.547699345726972e-05, "loss": 0.0414, "step": 14370 }, { "grad_norm": 0.42035797238349915, "learning_rate": 9.546839945163257e-05, "loss": 0.0382, "step": 14380 }, { "grad_norm": 0.4927510619163513, "learning_rate": 9.545979767668953e-05, "loss": 0.0525, "step": 14390 }, { "grad_norm": 0.4471638798713684, "learning_rate": 9.54511881339104e-05, "loss": 0.0387, "step": 14400 }, { "grad_norm": 0.44679784774780273, "learning_rate": 9.54425708247663e-05, "loss": 0.0389, "step": 14410 }, { "grad_norm": 0.40675222873687744, "learning_rate": 9.543394575072972e-05, "loss": 0.0395, "step": 14420 }, { "grad_norm": 0.3762909471988678, "learning_rate": 9.542531291327441e-05, "loss": 0.0375, "step": 14430 }, { "grad_norm": 0.3416166603565216, "learning_rate": 9.541667231387552e-05, "loss": 0.0353, "step": 14440 }, { "grad_norm": 0.3304896950721741, "learning_rate": 9.540802395400949e-05, "loss": 0.0399, "step": 14450 }, { "grad_norm": 0.377010315656662, "learning_rate": 9.539936783515406e-05, "loss": 0.0353, "step": 14460 }, { "grad_norm": 0.5229063034057617, "learning_rate": 9.539070395878835e-05, "loss": 0.0383, "step": 14470 }, { "grad_norm": 0.4615074396133423, "learning_rate": 9.538203232639277e-05, "loss": 0.036, "step": 14480 }, { "grad_norm": 0.39483389258384705, "learning_rate": 9.537335293944907e-05, "loss": 0.0402, "step": 14490 }, { "grad_norm": 0.4033648371696472, "learning_rate": 9.536466579944032e-05, "loss": 0.0531, "step": 14500 }, { "grad_norm": 0.4973951280117035, "learning_rate": 9.535597090785091e-05, "loss": 0.0428, "step": 14510 }, { "grad_norm": 0.3609781563282013, "learning_rate": 9.534726826616656e-05, "loss": 0.0342, "step": 14520 }, { "grad_norm": 0.5729681849479675, "learning_rate": 9.53385578758743e-05, "loss": 0.0547, "step": 14530 }, { "grad_norm": 0.45105838775634766, "learning_rate": 9.532983973846252e-05, "loss": 0.0446, "step": 14540 }, { "grad_norm": 0.5440917015075684, "learning_rate": 9.53211138554209e-05, "loss": 0.0405, "step": 14550 }, { "grad_norm": 0.3054599463939667, "learning_rate": 9.531238022824047e-05, "loss": 0.0396, "step": 14560 }, { "grad_norm": 0.3666191101074219, "learning_rate": 9.530363885841355e-05, "loss": 0.0344, "step": 14570 }, { "grad_norm": 0.49615156650543213, "learning_rate": 9.52948897474338e-05, "loss": 0.0465, "step": 14580 }, { "grad_norm": 0.3136923015117645, "learning_rate": 9.528613289679622e-05, "loss": 0.0508, "step": 14590 }, { "grad_norm": 0.4403083324432373, "learning_rate": 9.52773683079971e-05, "loss": 0.0432, "step": 14600 }, { "grad_norm": 0.38535746932029724, "learning_rate": 9.526859598253407e-05, "loss": 0.0352, "step": 14610 }, { "grad_norm": 0.47571858763694763, "learning_rate": 9.525981592190609e-05, "loss": 0.0385, "step": 14620 }, { "grad_norm": 0.43746158480644226, "learning_rate": 9.525102812761342e-05, "loss": 0.0382, "step": 14630 }, { "grad_norm": 0.388202428817749, "learning_rate": 9.524223260115768e-05, "loss": 0.0335, "step": 14640 }, { "grad_norm": 0.7605817317962646, "learning_rate": 9.523342934404175e-05, "loss": 0.0436, "step": 14650 }, { "grad_norm": 0.4826405644416809, "learning_rate": 9.522461835776989e-05, "loss": 0.0455, "step": 14660 }, { "grad_norm": 0.526417076587677, "learning_rate": 9.521579964384764e-05, "loss": 0.0502, "step": 14670 }, { "grad_norm": 0.4885443449020386, "learning_rate": 9.52069732037819e-05, "loss": 0.0439, "step": 14680 }, { "grad_norm": 0.4091547727584839, "learning_rate": 9.519813903908083e-05, "loss": 0.0437, "step": 14690 }, { "grad_norm": 0.3892230987548828, "learning_rate": 9.5189297151254e-05, "loss": 0.0497, "step": 14700 }, { "grad_norm": 0.580015242099762, "learning_rate": 9.518044754181218e-05, "loss": 0.0472, "step": 14710 }, { "grad_norm": 0.40132519602775574, "learning_rate": 9.51715902122676e-05, "loss": 0.0454, "step": 14720 }, { "grad_norm": 0.4249028265476227, "learning_rate": 9.516272516413368e-05, "loss": 0.0471, "step": 14730 }, { "grad_norm": 0.4558207392692566, "learning_rate": 9.515385239892525e-05, "loss": 0.0501, "step": 14740 }, { "grad_norm": 0.2899569869041443, "learning_rate": 9.514497191815839e-05, "loss": 0.0509, "step": 14750 }, { "grad_norm": 0.3643931448459625, "learning_rate": 9.513608372335055e-05, "loss": 0.0424, "step": 14760 }, { "grad_norm": 0.46097928285598755, "learning_rate": 9.512718781602045e-05, "loss": 0.0437, "step": 14770 }, { "grad_norm": 0.5342243313789368, "learning_rate": 9.511828419768823e-05, "loss": 0.0423, "step": 14780 }, { "grad_norm": 0.4659900665283203, "learning_rate": 9.510937286987521e-05, "loss": 0.0394, "step": 14790 }, { "grad_norm": 0.5743874907493591, "learning_rate": 9.510045383410408e-05, "loss": 0.0487, "step": 14800 }, { "grad_norm": 0.354861319065094, "learning_rate": 9.509152709189892e-05, "loss": 0.0431, "step": 14810 }, { "grad_norm": 0.5026644468307495, "learning_rate": 9.508259264478504e-05, "loss": 0.0444, "step": 14820 }, { "grad_norm": 0.3901246190071106, "learning_rate": 9.507365049428909e-05, "loss": 0.0458, "step": 14830 }, { "grad_norm": 0.3265673816204071, "learning_rate": 9.506470064193902e-05, "loss": 0.0521, "step": 14840 }, { "grad_norm": 0.45392993092536926, "learning_rate": 9.505574308926414e-05, "loss": 0.0655, "step": 14850 }, { "grad_norm": 0.41684994101524353, "learning_rate": 9.504677783779505e-05, "loss": 0.0423, "step": 14860 }, { "grad_norm": 0.3861827254295349, "learning_rate": 9.503780488906365e-05, "loss": 0.0436, "step": 14870 }, { "grad_norm": 0.509314775466919, "learning_rate": 9.502882424460319e-05, "loss": 0.0428, "step": 14880 }, { "grad_norm": 0.48995569348335266, "learning_rate": 9.501983590594821e-05, "loss": 0.0386, "step": 14890 }, { "grad_norm": 0.3301602900028229, "learning_rate": 9.501083987463455e-05, "loss": 0.0339, "step": 14900 }, { "grad_norm": 0.5336301326751709, "learning_rate": 9.500183615219942e-05, "loss": 0.0422, "step": 14910 }, { "grad_norm": 0.546282172203064, "learning_rate": 9.49928247401813e-05, "loss": 0.04, "step": 14920 }, { "grad_norm": 0.7227290272712708, "learning_rate": 9.498380564011997e-05, "loss": 0.0409, "step": 14930 }, { "grad_norm": 0.6780446171760559, "learning_rate": 9.497477885355656e-05, "loss": 0.0505, "step": 14940 }, { "grad_norm": 0.5481519103050232, "learning_rate": 9.496574438203353e-05, "loss": 0.0492, "step": 14950 }, { "grad_norm": 0.47502657771110535, "learning_rate": 9.495670222709459e-05, "loss": 0.0474, "step": 14960 }, { "grad_norm": 0.39619705080986023, "learning_rate": 9.494765239028483e-05, "loss": 0.0415, "step": 14970 }, { "grad_norm": 0.42605769634246826, "learning_rate": 9.493859487315057e-05, "loss": 0.0401, "step": 14980 }, { "grad_norm": 0.554203450679779, "learning_rate": 9.492952967723953e-05, "loss": 0.0389, "step": 14990 }, { "grad_norm": 0.4647393822669983, "learning_rate": 9.492045680410068e-05, "loss": 0.0513, "step": 15000 }, { "grad_norm": 0.44815781712532043, "learning_rate": 9.491137625528436e-05, "loss": 0.037, "step": 15010 }, { "grad_norm": 0.6216405034065247, "learning_rate": 9.490228803234215e-05, "loss": 0.0345, "step": 15020 }, { "grad_norm": 0.42089736461639404, "learning_rate": 9.489319213682701e-05, "loss": 0.0393, "step": 15030 }, { "grad_norm": 0.43502292037010193, "learning_rate": 9.488408857029316e-05, "loss": 0.0525, "step": 15040 }, { "grad_norm": 0.5412378907203674, "learning_rate": 9.487497733429616e-05, "loss": 0.0525, "step": 15050 }, { "grad_norm": 0.558055579662323, "learning_rate": 9.486585843039286e-05, "loss": 0.0534, "step": 15060 }, { "grad_norm": 0.39138469099998474, "learning_rate": 9.485673186014143e-05, "loss": 0.044, "step": 15070 }, { "grad_norm": 0.7528339624404907, "learning_rate": 9.484759762510137e-05, "loss": 0.0566, "step": 15080 }, { "grad_norm": 0.40083813667297363, "learning_rate": 9.483845572683346e-05, "loss": 0.0424, "step": 15090 }, { "grad_norm": 0.3976590037345886, "learning_rate": 9.48293061668998e-05, "loss": 0.043, "step": 15100 }, { "grad_norm": 0.48255598545074463, "learning_rate": 9.48201489468638e-05, "loss": 0.0421, "step": 15110 }, { "grad_norm": 0.345337837934494, "learning_rate": 9.481098406829016e-05, "loss": 0.0355, "step": 15120 }, { "grad_norm": 0.29878729581832886, "learning_rate": 9.480181153274495e-05, "loss": 0.0319, "step": 15130 }, { "grad_norm": 0.3889608681201935, "learning_rate": 9.479263134179548e-05, "loss": 0.0274, "step": 15140 }, { "grad_norm": 0.5619472861289978, "learning_rate": 9.478344349701039e-05, "loss": 0.0415, "step": 15150 }, { "grad_norm": 0.4276297688484192, "learning_rate": 9.477424799995964e-05, "loss": 0.0334, "step": 15160 }, { "grad_norm": 0.4812365472316742, "learning_rate": 9.476504485221448e-05, "loss": 0.0479, "step": 15170 }, { "grad_norm": 0.3388201892375946, "learning_rate": 9.475583405534748e-05, "loss": 0.04, "step": 15180 }, { "grad_norm": 0.4354267120361328, "learning_rate": 9.474661561093251e-05, "loss": 0.0387, "step": 15190 }, { "grad_norm": 0.47917354106903076, "learning_rate": 9.473738952054478e-05, "loss": 0.0499, "step": 15200 }, { "grad_norm": 0.28291967511177063, "learning_rate": 9.472815578576073e-05, "loss": 0.0386, "step": 15210 }, { "grad_norm": 0.7630964517593384, "learning_rate": 9.471891440815817e-05, "loss": 0.0412, "step": 15220 }, { "grad_norm": 0.3938199281692505, "learning_rate": 9.470966538931621e-05, "loss": 0.0362, "step": 15230 }, { "grad_norm": 0.3248332142829895, "learning_rate": 9.470040873081525e-05, "loss": 0.0387, "step": 15240 }, { "grad_norm": 0.5181470513343811, "learning_rate": 9.469114443423698e-05, "loss": 0.0494, "step": 15250 }, { "grad_norm": 0.5856170058250427, "learning_rate": 9.468187250116445e-05, "loss": 0.0416, "step": 15260 }, { "grad_norm": 0.38068887591362, "learning_rate": 9.467259293318197e-05, "loss": 0.035, "step": 15270 }, { "grad_norm": 0.2834935486316681, "learning_rate": 9.466330573187514e-05, "loss": 0.0352, "step": 15280 }, { "grad_norm": 0.5095102190971375, "learning_rate": 9.46540108988309e-05, "loss": 0.047, "step": 15290 }, { "grad_norm": 0.5523943901062012, "learning_rate": 9.46447084356375e-05, "loss": 0.041, "step": 15300 }, { "grad_norm": 0.4478524923324585, "learning_rate": 9.463539834388447e-05, "loss": 0.0513, "step": 15310 }, { "grad_norm": 0.507524847984314, "learning_rate": 9.462608062516263e-05, "loss": 0.042, "step": 15320 }, { "grad_norm": 1.3791112899780273, "learning_rate": 9.461675528106413e-05, "loss": 0.041, "step": 15330 }, { "grad_norm": 0.4505903422832489, "learning_rate": 9.460742231318244e-05, "loss": 0.0497, "step": 15340 }, { "grad_norm": 0.4788481593132019, "learning_rate": 9.459808172311229e-05, "loss": 0.052, "step": 15350 }, { "grad_norm": 0.3733186423778534, "learning_rate": 9.458873351244972e-05, "loss": 0.0402, "step": 15360 }, { "grad_norm": 0.3302479088306427, "learning_rate": 9.457937768279211e-05, "loss": 0.0467, "step": 15370 }, { "grad_norm": 0.46956685185432434, "learning_rate": 9.45700142357381e-05, "loss": 0.0435, "step": 15380 }, { "grad_norm": 0.40928199887275696, "learning_rate": 9.456064317288765e-05, "loss": 0.0445, "step": 15390 }, { "grad_norm": 0.3372962772846222, "learning_rate": 9.455126449584201e-05, "loss": 0.0389, "step": 15400 }, { "grad_norm": 0.37586402893066406, "learning_rate": 9.454187820620375e-05, "loss": 0.0353, "step": 15410 }, { "grad_norm": 0.48038873076438904, "learning_rate": 9.453248430557673e-05, "loss": 0.0382, "step": 15420 }, { "grad_norm": 0.5241243839263916, "learning_rate": 9.452308279556611e-05, "loss": 0.0554, "step": 15430 }, { "grad_norm": 0.653222382068634, "learning_rate": 9.451367367777835e-05, "loss": 0.0405, "step": 15440 }, { "grad_norm": 0.3091312050819397, "learning_rate": 9.450425695382122e-05, "loss": 0.0504, "step": 15450 }, { "grad_norm": 0.39937451481819153, "learning_rate": 9.449483262530375e-05, "loss": 0.0391, "step": 15460 }, { "grad_norm": 0.39093875885009766, "learning_rate": 9.448540069383633e-05, "loss": 0.0342, "step": 15470 }, { "grad_norm": 0.5335071682929993, "learning_rate": 9.447596116103061e-05, "loss": 0.0434, "step": 15480 }, { "grad_norm": 0.5430645942687988, "learning_rate": 9.446651402849955e-05, "loss": 0.0477, "step": 15490 }, { "grad_norm": 0.3707900047302246, "learning_rate": 9.44570592978574e-05, "loss": 0.0399, "step": 15500 }, { "grad_norm": 0.25368914008140564, "learning_rate": 9.444759697071972e-05, "loss": 0.0397, "step": 15510 }, { "grad_norm": 0.4975147247314453, "learning_rate": 9.443812704870336e-05, "loss": 0.0342, "step": 15520 }, { "grad_norm": 0.37321802973747253, "learning_rate": 9.442864953342649e-05, "loss": 0.0341, "step": 15530 }, { "grad_norm": 0.3827781677246094, "learning_rate": 9.441916442650852e-05, "loss": 0.0398, "step": 15540 }, { "grad_norm": 0.32730525732040405, "learning_rate": 9.440967172957023e-05, "loss": 0.0359, "step": 15550 }, { "grad_norm": 0.46789902448654175, "learning_rate": 9.440017144423364e-05, "loss": 0.04, "step": 15560 }, { "grad_norm": 0.8488776087760925, "learning_rate": 9.439066357212209e-05, "loss": 0.0497, "step": 15570 }, { "grad_norm": 0.36652475595474243, "learning_rate": 9.438114811486022e-05, "loss": 0.0471, "step": 15580 }, { "grad_norm": 0.32471439242362976, "learning_rate": 9.4371625074074e-05, "loss": 0.0367, "step": 15590 }, { "grad_norm": 0.37918591499328613, "learning_rate": 9.436209445139059e-05, "loss": 0.033, "step": 15600 }, { "grad_norm": 0.6393328905105591, "learning_rate": 9.435255624843855e-05, "loss": 0.0533, "step": 15610 }, { "grad_norm": 0.41782501339912415, "learning_rate": 9.43430104668477e-05, "loss": 0.0358, "step": 15620 }, { "grad_norm": 0.4544743597507477, "learning_rate": 9.433345710824914e-05, "loss": 0.0453, "step": 15630 }, { "grad_norm": 0.38256219029426575, "learning_rate": 9.432389617427529e-05, "loss": 0.0414, "step": 15640 }, { "grad_norm": 0.35384052991867065, "learning_rate": 9.431432766655984e-05, "loss": 0.039, "step": 15650 }, { "grad_norm": 0.42665332555770874, "learning_rate": 9.430475158673778e-05, "loss": 0.0442, "step": 15660 }, { "grad_norm": 0.360872745513916, "learning_rate": 9.429516793644542e-05, "loss": 0.0412, "step": 15670 }, { "grad_norm": 0.35415247082710266, "learning_rate": 9.428557671732034e-05, "loss": 0.0312, "step": 15680 }, { "grad_norm": 0.569317102432251, "learning_rate": 9.42759779310014e-05, "loss": 0.0385, "step": 15690 }, { "grad_norm": 0.47838133573532104, "learning_rate": 9.426637157912879e-05, "loss": 0.045, "step": 15700 }, { "grad_norm": 0.35972681641578674, "learning_rate": 9.425675766334397e-05, "loss": 0.0451, "step": 15710 }, { "grad_norm": 0.4229332506656647, "learning_rate": 9.424713618528968e-05, "loss": 0.0349, "step": 15720 }, { "grad_norm": 0.4281511902809143, "learning_rate": 9.423750714661e-05, "loss": 0.0487, "step": 15730 }, { "grad_norm": 0.3514706492424011, "learning_rate": 9.422787054895022e-05, "loss": 0.0309, "step": 15740 }, { "grad_norm": 0.3951966464519501, "learning_rate": 9.4218226393957e-05, "loss": 0.0492, "step": 15750 }, { "grad_norm": 0.39932718873023987, "learning_rate": 9.420857468327828e-05, "loss": 0.0417, "step": 15760 }, { "grad_norm": 0.41900894045829773, "learning_rate": 9.419891541856323e-05, "loss": 0.045, "step": 15770 }, { "grad_norm": 0.3167802691459656, "learning_rate": 9.41892486014624e-05, "loss": 0.0463, "step": 15780 }, { "grad_norm": 0.5290799140930176, "learning_rate": 9.417957423362756e-05, "loss": 0.0551, "step": 15790 }, { "grad_norm": 0.32303759455680847, "learning_rate": 9.416989231671178e-05, "loss": 0.0393, "step": 15800 }, { "grad_norm": 0.29439520835876465, "learning_rate": 9.416020285236946e-05, "loss": 0.0396, "step": 15810 }, { "grad_norm": 0.5604180693626404, "learning_rate": 9.415050584225626e-05, "loss": 0.042, "step": 15820 }, { "grad_norm": 0.4361097514629364, "learning_rate": 9.414080128802914e-05, "loss": 0.0391, "step": 15830 }, { "grad_norm": 0.3527112901210785, "learning_rate": 9.413108919134632e-05, "loss": 0.0364, "step": 15840 }, { "grad_norm": 0.41902264952659607, "learning_rate": 9.412136955386734e-05, "loss": 0.0359, "step": 15850 }, { "grad_norm": 0.4680146872997284, "learning_rate": 9.411164237725303e-05, "loss": 0.0361, "step": 15860 }, { "grad_norm": 0.45774200558662415, "learning_rate": 9.41019076631655e-05, "loss": 0.0398, "step": 15870 }, { "grad_norm": 0.42389556765556335, "learning_rate": 9.409216541326815e-05, "loss": 0.0471, "step": 15880 }, { "grad_norm": 0.3198574483394623, "learning_rate": 9.408241562922564e-05, "loss": 0.0396, "step": 15890 }, { "grad_norm": 0.36391693353652954, "learning_rate": 9.407265831270395e-05, "loss": 0.0326, "step": 15900 }, { "grad_norm": 0.29496148228645325, "learning_rate": 9.406289346537035e-05, "loss": 0.0368, "step": 15910 }, { "grad_norm": 0.41647857427597046, "learning_rate": 9.405312108889339e-05, "loss": 0.0356, "step": 15920 }, { "grad_norm": 0.35751253366470337, "learning_rate": 9.404334118494288e-05, "loss": 0.0408, "step": 15930 }, { "grad_norm": 0.2642008364200592, "learning_rate": 9.403355375518995e-05, "loss": 0.0388, "step": 15940 }, { "grad_norm": 0.24573718011379242, "learning_rate": 9.4023758801307e-05, "loss": 0.0378, "step": 15950 }, { "grad_norm": 0.25862178206443787, "learning_rate": 9.401395632496774e-05, "loss": 0.0332, "step": 15960 }, { "grad_norm": 0.3608960509300232, "learning_rate": 9.400414632784711e-05, "loss": 0.043, "step": 15970 }, { "grad_norm": 0.5779455900192261, "learning_rate": 9.39943288116214e-05, "loss": 0.0353, "step": 15980 }, { "grad_norm": 0.3356480002403259, "learning_rate": 9.398450377796815e-05, "loss": 0.0432, "step": 15990 }, { "grad_norm": 0.399041086435318, "learning_rate": 9.397467122856616e-05, "loss": 0.0433, "step": 16000 }, { "grad_norm": 0.3346885144710541, "learning_rate": 9.396483116509558e-05, "loss": 0.0302, "step": 16010 }, { "grad_norm": 0.30859771370887756, "learning_rate": 9.39549835892378e-05, "loss": 0.0333, "step": 16020 }, { "grad_norm": 0.3806462287902832, "learning_rate": 9.39451285026755e-05, "loss": 0.0394, "step": 16030 }, { "grad_norm": 0.42127740383148193, "learning_rate": 9.393526590709262e-05, "loss": 0.0348, "step": 16040 }, { "grad_norm": 0.45202937722206116, "learning_rate": 9.392539580417444e-05, "loss": 0.0346, "step": 16050 }, { "grad_norm": 0.3964751362800598, "learning_rate": 9.391551819560747e-05, "loss": 0.0357, "step": 16060 }, { "grad_norm": 0.46202656626701355, "learning_rate": 9.390563308307955e-05, "loss": 0.0328, "step": 16070 }, { "grad_norm": 0.4547281265258789, "learning_rate": 9.389574046827974e-05, "loss": 0.0427, "step": 16080 }, { "grad_norm": 0.39437127113342285, "learning_rate": 9.388584035289845e-05, "loss": 0.0391, "step": 16090 }, { "grad_norm": 0.42075851559638977, "learning_rate": 9.387593273862732e-05, "loss": 0.0442, "step": 16100 }, { "grad_norm": 0.4231313169002533, "learning_rate": 9.386601762715929e-05, "loss": 0.0431, "step": 16110 }, { "grad_norm": 0.4335596263408661, "learning_rate": 9.38560950201886e-05, "loss": 0.0468, "step": 16120 }, { "grad_norm": 0.5805754065513611, "learning_rate": 9.384616491941071e-05, "loss": 0.0412, "step": 16130 }, { "grad_norm": 0.5458234548568726, "learning_rate": 9.383622732652245e-05, "loss": 0.0318, "step": 16140 }, { "grad_norm": 0.31523406505584717, "learning_rate": 9.382628224322187e-05, "loss": 0.0439, "step": 16150 }, { "grad_norm": 0.3187142312526703, "learning_rate": 9.381632967120829e-05, "loss": 0.0414, "step": 16160 }, { "grad_norm": 0.4417238235473633, "learning_rate": 9.380636961218235e-05, "loss": 0.0389, "step": 16170 }, { "grad_norm": 0.37233683466911316, "learning_rate": 9.379640206784597e-05, "loss": 0.0314, "step": 16180 }, { "grad_norm": 0.47277703881263733, "learning_rate": 9.378642703990229e-05, "loss": 0.042, "step": 16190 }, { "grad_norm": 0.3557370901107788, "learning_rate": 9.37764445300558e-05, "loss": 0.0422, "step": 16200 }, { "grad_norm": 0.36499476432800293, "learning_rate": 9.376645454001222e-05, "loss": 0.0484, "step": 16210 }, { "grad_norm": 0.42904120683670044, "learning_rate": 9.375645707147858e-05, "loss": 0.0335, "step": 16220 }, { "grad_norm": 0.4341629445552826, "learning_rate": 9.374645212616316e-05, "loss": 0.05, "step": 16230 }, { "grad_norm": 0.4928911030292511, "learning_rate": 9.373643970577555e-05, "loss": 0.0413, "step": 16240 }, { "grad_norm": 0.42841464281082153, "learning_rate": 9.372641981202659e-05, "loss": 0.0526, "step": 16250 }, { "grad_norm": 0.3489953279495239, "learning_rate": 9.37163924466284e-05, "loss": 0.0421, "step": 16260 }, { "grad_norm": 0.39739087224006653, "learning_rate": 9.370635761129438e-05, "loss": 0.0403, "step": 16270 }, { "grad_norm": 0.4595629572868347, "learning_rate": 9.36963153077392e-05, "loss": 0.0443, "step": 16280 }, { "grad_norm": 0.3625541627407074, "learning_rate": 9.368626553767888e-05, "loss": 0.0393, "step": 16290 }, { "grad_norm": 0.3417719602584839, "learning_rate": 9.367620830283057e-05, "loss": 0.0299, "step": 16300 }, { "grad_norm": 0.47665220499038696, "learning_rate": 9.366614360491281e-05, "loss": 0.0392, "step": 16310 }, { "grad_norm": 0.3939836025238037, "learning_rate": 9.365607144564539e-05, "loss": 0.0336, "step": 16320 }, { "grad_norm": 0.34886372089385986, "learning_rate": 9.364599182674934e-05, "loss": 0.0467, "step": 16330 }, { "grad_norm": 0.3379485607147217, "learning_rate": 9.3635904749947e-05, "loss": 0.0349, "step": 16340 }, { "grad_norm": 0.44499123096466064, "learning_rate": 9.362581021696202e-05, "loss": 0.0357, "step": 16350 }, { "grad_norm": 0.34880509972572327, "learning_rate": 9.361570822951921e-05, "loss": 0.0427, "step": 16360 }, { "grad_norm": 0.475378155708313, "learning_rate": 9.360559878934476e-05, "loss": 0.0417, "step": 16370 }, { "grad_norm": 0.35848894715309143, "learning_rate": 9.359548189816611e-05, "loss": 0.0365, "step": 16380 }, { "grad_norm": 0.45730099081993103, "learning_rate": 9.358535755771193e-05, "loss": 0.043, "step": 16390 }, { "grad_norm": 0.29652684926986694, "learning_rate": 9.357522576971221e-05, "loss": 0.038, "step": 16400 }, { "grad_norm": 0.38220512866973877, "learning_rate": 9.356508653589819e-05, "loss": 0.0437, "step": 16410 }, { "grad_norm": 0.29693400859832764, "learning_rate": 9.355493985800237e-05, "loss": 0.04, "step": 16420 }, { "grad_norm": 0.4267085790634155, "learning_rate": 9.354478573775857e-05, "loss": 0.0378, "step": 16430 }, { "grad_norm": 0.4513190686702728, "learning_rate": 9.353462417690186e-05, "loss": 0.0416, "step": 16440 }, { "grad_norm": 0.3721033036708832, "learning_rate": 9.352445517716853e-05, "loss": 0.0388, "step": 16450 }, { "grad_norm": 0.3655655086040497, "learning_rate": 9.351427874029621e-05, "loss": 0.0419, "step": 16460 }, { "grad_norm": 0.3204815089702606, "learning_rate": 9.350409486802379e-05, "loss": 0.037, "step": 16470 }, { "grad_norm": 0.3801737129688263, "learning_rate": 9.349390356209138e-05, "loss": 0.0488, "step": 16480 }, { "grad_norm": 0.34795650839805603, "learning_rate": 9.348370482424042e-05, "loss": 0.0472, "step": 16490 }, { "grad_norm": 0.6034976840019226, "learning_rate": 9.347349865621357e-05, "loss": 0.0382, "step": 16500 }, { "grad_norm": 0.37149637937545776, "learning_rate": 9.346328505975481e-05, "loss": 0.0473, "step": 16510 }, { "grad_norm": 0.299872487783432, "learning_rate": 9.345306403660936e-05, "loss": 0.0459, "step": 16520 }, { "grad_norm": 0.518706202507019, "learning_rate": 9.344283558852371e-05, "loss": 0.0416, "step": 16530 }, { "grad_norm": 0.5960737466812134, "learning_rate": 9.343259971724563e-05, "loss": 0.0485, "step": 16540 }, { "grad_norm": 0.43245550990104675, "learning_rate": 9.342235642452413e-05, "loss": 0.0398, "step": 16550 }, { "grad_norm": 0.4913434386253357, "learning_rate": 9.341210571210954e-05, "loss": 0.0417, "step": 16560 }, { "grad_norm": 0.4952826201915741, "learning_rate": 9.340184758175338e-05, "loss": 0.039, "step": 16570 }, { "grad_norm": 0.40309563279151917, "learning_rate": 9.339158203520854e-05, "loss": 0.0485, "step": 16580 }, { "grad_norm": 0.3787982761859894, "learning_rate": 9.338130907422908e-05, "loss": 0.0345, "step": 16590 }, { "grad_norm": 0.3648316562175751, "learning_rate": 9.337102870057037e-05, "loss": 0.032, "step": 16600 }, { "grad_norm": 0.26193976402282715, "learning_rate": 9.336074091598907e-05, "loss": 0.0318, "step": 16610 }, { "grad_norm": 0.4058169722557068, "learning_rate": 9.335044572224306e-05, "loss": 0.0267, "step": 16620 }, { "grad_norm": 0.42333129048347473, "learning_rate": 9.334014312109151e-05, "loss": 0.0309, "step": 16630 }, { "grad_norm": 0.5651357173919678, "learning_rate": 9.332983311429486e-05, "loss": 0.0372, "step": 16640 }, { "grad_norm": 0.338476300239563, "learning_rate": 9.33195157036148e-05, "loss": 0.0377, "step": 16650 }, { "grad_norm": 0.3730369806289673, "learning_rate": 9.330919089081432e-05, "loss": 0.0338, "step": 16660 }, { "grad_norm": 0.4678986668586731, "learning_rate": 9.32988586776576e-05, "loss": 0.0403, "step": 16670 }, { "grad_norm": 0.4615095555782318, "learning_rate": 9.328851906591016e-05, "loss": 0.0334, "step": 16680 }, { "grad_norm": 0.5816299915313721, "learning_rate": 9.327817205733875e-05, "loss": 0.036, "step": 16690 }, { "grad_norm": 0.27703040838241577, "learning_rate": 9.326781765371142e-05, "loss": 0.0329, "step": 16700 }, { "grad_norm": 0.42260780930519104, "learning_rate": 9.325745585679741e-05, "loss": 0.0419, "step": 16710 }, { "grad_norm": 0.4463431239128113, "learning_rate": 9.32470866683673e-05, "loss": 0.056, "step": 16720 }, { "grad_norm": 0.5255961418151855, "learning_rate": 9.323671009019288e-05, "loss": 0.0523, "step": 16730 }, { "grad_norm": 0.4487669765949249, "learning_rate": 9.322632612404725e-05, "loss": 0.0445, "step": 16740 }, { "grad_norm": 0.36040690541267395, "learning_rate": 9.321593477170471e-05, "loss": 0.0513, "step": 16750 }, { "grad_norm": 0.3604166507720947, "learning_rate": 9.320553603494088e-05, "loss": 0.0402, "step": 16760 }, { "grad_norm": 0.46051329374313354, "learning_rate": 9.319512991553261e-05, "loss": 0.0436, "step": 16770 }, { "grad_norm": 0.458624929189682, "learning_rate": 9.318471641525803e-05, "loss": 0.0372, "step": 16780 }, { "grad_norm": 0.3885023891925812, "learning_rate": 9.317429553589652e-05, "loss": 0.0365, "step": 16790 }, { "grad_norm": 0.5068872570991516, "learning_rate": 9.316386727922873e-05, "loss": 0.0351, "step": 16800 }, { "grad_norm": 0.34409236907958984, "learning_rate": 9.315343164703656e-05, "loss": 0.0363, "step": 16810 }, { "grad_norm": 0.5469350814819336, "learning_rate": 9.314298864110316e-05, "loss": 0.0403, "step": 16820 }, { "grad_norm": 0.3661065399646759, "learning_rate": 9.313253826321295e-05, "loss": 0.0407, "step": 16830 }, { "grad_norm": 0.42690813541412354, "learning_rate": 9.312208051515165e-05, "loss": 0.045, "step": 16840 }, { "grad_norm": 0.4654434025287628, "learning_rate": 9.311161539870618e-05, "loss": 0.039, "step": 16850 }, { "grad_norm": 0.3113771677017212, "learning_rate": 9.310114291566474e-05, "loss": 0.0444, "step": 16860 }, { "grad_norm": 0.4533078670501709, "learning_rate": 9.309066306781679e-05, "loss": 0.037, "step": 16870 }, { "grad_norm": 0.40369847416877747, "learning_rate": 9.308017585695306e-05, "loss": 0.0362, "step": 16880 }, { "grad_norm": 0.41005659103393555, "learning_rate": 9.306968128486552e-05, "loss": 0.0331, "step": 16890 }, { "grad_norm": 0.7504719495773315, "learning_rate": 9.30591793533474e-05, "loss": 0.0431, "step": 16900 }, { "grad_norm": 0.5667011141777039, "learning_rate": 9.304867006419321e-05, "loss": 0.0443, "step": 16910 }, { "grad_norm": 0.36432960629463196, "learning_rate": 9.303815341919868e-05, "loss": 0.0359, "step": 16920 }, { "grad_norm": 0.4719550907611847, "learning_rate": 9.302762942016084e-05, "loss": 0.0381, "step": 16930 }, { "grad_norm": 0.5564181208610535, "learning_rate": 9.301709806887792e-05, "loss": 0.039, "step": 16940 }, { "grad_norm": 0.4458763301372528, "learning_rate": 9.300655936714948e-05, "loss": 0.0494, "step": 16950 }, { "grad_norm": 0.4994100332260132, "learning_rate": 9.299601331677627e-05, "loss": 0.0424, "step": 16960 }, { "grad_norm": 0.534355878829956, "learning_rate": 9.298545991956033e-05, "loss": 0.0399, "step": 16970 }, { "grad_norm": 0.31584152579307556, "learning_rate": 9.297489917730493e-05, "loss": 0.0418, "step": 16980 }, { "grad_norm": 0.35093769431114197, "learning_rate": 9.296433109181464e-05, "loss": 0.0438, "step": 16990 }, { "grad_norm": 0.7859265208244324, "learning_rate": 9.295375566489523e-05, "loss": 0.0435, "step": 17000 }, { "grad_norm": 0.3578079640865326, "learning_rate": 9.294317289835379e-05, "loss": 0.0448, "step": 17010 }, { "grad_norm": 0.48617982864379883, "learning_rate": 9.293258279399859e-05, "loss": 0.0454, "step": 17020 }, { "grad_norm": 0.2693389654159546, "learning_rate": 9.292198535363919e-05, "loss": 0.0363, "step": 17030 }, { "grad_norm": 0.34742170572280884, "learning_rate": 9.291138057908641e-05, "loss": 0.0374, "step": 17040 }, { "grad_norm": 0.48218995332717896, "learning_rate": 9.290076847215234e-05, "loss": 0.0356, "step": 17050 }, { "grad_norm": 0.6213827729225159, "learning_rate": 9.289014903465025e-05, "loss": 0.0397, "step": 17060 }, { "grad_norm": 0.2872277498245239, "learning_rate": 9.287952226839475e-05, "loss": 0.0296, "step": 17070 }, { "grad_norm": 0.4311879873275757, "learning_rate": 9.286888817520164e-05, "loss": 0.0332, "step": 17080 }, { "grad_norm": 0.28503239154815674, "learning_rate": 9.285824675688803e-05, "loss": 0.0314, "step": 17090 }, { "grad_norm": 0.362549364566803, "learning_rate": 9.28475980152722e-05, "loss": 0.0321, "step": 17100 }, { "grad_norm": 0.4108794033527374, "learning_rate": 9.283694195217379e-05, "loss": 0.033, "step": 17110 }, { "grad_norm": 0.44986844062805176, "learning_rate": 9.282627856941356e-05, "loss": 0.0357, "step": 17120 }, { "grad_norm": 0.34016865491867065, "learning_rate": 9.281560786881363e-05, "loss": 0.0304, "step": 17130 }, { "grad_norm": 0.43729302287101746, "learning_rate": 9.280492985219733e-05, "loss": 0.0346, "step": 17140 }, { "grad_norm": 0.24792176485061646, "learning_rate": 9.279424452138924e-05, "loss": 0.039, "step": 17150 }, { "grad_norm": 0.3567817807197571, "learning_rate": 9.278355187821517e-05, "loss": 0.0444, "step": 17160 }, { "grad_norm": 0.40960830450057983, "learning_rate": 9.277285192450224e-05, "loss": 0.0437, "step": 17170 }, { "grad_norm": 0.4552701711654663, "learning_rate": 9.276214466207875e-05, "loss": 0.0422, "step": 17180 }, { "grad_norm": 0.370373010635376, "learning_rate": 9.275143009277427e-05, "loss": 0.0307, "step": 17190 }, { "grad_norm": 0.4913177788257599, "learning_rate": 9.274070821841964e-05, "loss": 0.034, "step": 17200 }, { "grad_norm": 0.3484823405742645, "learning_rate": 9.272997904084696e-05, "loss": 0.0334, "step": 17210 }, { "grad_norm": 0.368309885263443, "learning_rate": 9.271924256188951e-05, "loss": 0.0355, "step": 17220 }, { "grad_norm": 0.44529739022254944, "learning_rate": 9.270849878338189e-05, "loss": 0.0396, "step": 17230 }, { "grad_norm": 0.3752562701702118, "learning_rate": 9.269774770715991e-05, "loss": 0.0353, "step": 17240 }, { "grad_norm": 0.5634811520576477, "learning_rate": 9.268698933506061e-05, "loss": 0.0448, "step": 17250 }, { "grad_norm": 0.37780386209487915, "learning_rate": 9.267622366892235e-05, "loss": 0.0326, "step": 17260 }, { "grad_norm": 0.39533236622810364, "learning_rate": 9.266545071058465e-05, "loss": 0.0349, "step": 17270 }, { "grad_norm": 0.27895793318748474, "learning_rate": 9.265467046188833e-05, "loss": 0.0316, "step": 17280 }, { "grad_norm": 0.4512782394886017, "learning_rate": 9.264388292467543e-05, "loss": 0.0387, "step": 17290 }, { "grad_norm": 0.33240002393722534, "learning_rate": 9.263308810078926e-05, "loss": 0.0337, "step": 17300 }, { "grad_norm": 0.38489988446235657, "learning_rate": 9.262228599207434e-05, "loss": 0.0314, "step": 17310 }, { "grad_norm": 0.3452615737915039, "learning_rate": 9.261147660037647e-05, "loss": 0.0234, "step": 17320 }, { "grad_norm": 0.2569207549095154, "learning_rate": 9.26006599275427e-05, "loss": 0.0321, "step": 17330 }, { "grad_norm": 0.38189488649368286, "learning_rate": 9.258983597542124e-05, "loss": 0.0329, "step": 17340 }, { "grad_norm": 0.3460424244403839, "learning_rate": 9.257900474586167e-05, "loss": 0.0279, "step": 17350 }, { "grad_norm": 0.4562113583087921, "learning_rate": 9.256816624071471e-05, "loss": 0.0425, "step": 17360 }, { "grad_norm": 0.35997727513313293, "learning_rate": 9.25573204618324e-05, "loss": 0.0393, "step": 17370 }, { "grad_norm": 0.29465657472610474, "learning_rate": 9.254646741106796e-05, "loss": 0.0381, "step": 17380 }, { "grad_norm": 0.3057857155799866, "learning_rate": 9.253560709027589e-05, "loss": 0.0417, "step": 17390 }, { "grad_norm": 0.38794785737991333, "learning_rate": 9.252473950131192e-05, "loss": 0.0469, "step": 17400 }, { "grad_norm": 0.3778064548969269, "learning_rate": 9.251386464603302e-05, "loss": 0.03, "step": 17410 }, { "grad_norm": 0.3737868368625641, "learning_rate": 9.250298252629741e-05, "loss": 0.0383, "step": 17420 }, { "grad_norm": 0.3972663879394531, "learning_rate": 9.249209314396454e-05, "loss": 0.0316, "step": 17430 }, { "grad_norm": 0.3399431109428406, "learning_rate": 9.248119650089513e-05, "loss": 0.0358, "step": 17440 }, { "grad_norm": 0.2548033595085144, "learning_rate": 9.247029259895108e-05, "loss": 0.0398, "step": 17450 }, { "grad_norm": 0.4523741602897644, "learning_rate": 9.24593814399956e-05, "loss": 0.0351, "step": 17460 }, { "grad_norm": 0.2974990904331207, "learning_rate": 9.244846302589309e-05, "loss": 0.0296, "step": 17470 }, { "grad_norm": 0.32832714915275574, "learning_rate": 9.243753735850923e-05, "loss": 0.037, "step": 17480 }, { "grad_norm": 0.3763536214828491, "learning_rate": 9.24266044397109e-05, "loss": 0.0334, "step": 17490 }, { "grad_norm": 0.34385472536087036, "learning_rate": 9.241566427136624e-05, "loss": 0.0389, "step": 17500 }, { "grad_norm": 0.40182429552078247, "learning_rate": 9.240471685534463e-05, "loss": 0.0398, "step": 17510 }, { "grad_norm": 0.28075259923934937, "learning_rate": 9.239376219351667e-05, "loss": 0.041, "step": 17520 }, { "grad_norm": 0.5229320526123047, "learning_rate": 9.238280028775425e-05, "loss": 0.04, "step": 17530 }, { "grad_norm": 0.3725567162036896, "learning_rate": 9.237183113993041e-05, "loss": 0.0423, "step": 17540 }, { "grad_norm": 0.2674903869628906, "learning_rate": 9.236085475191952e-05, "loss": 0.04, "step": 17550 }, { "grad_norm": 0.3862789273262024, "learning_rate": 9.234987112559709e-05, "loss": 0.0514, "step": 17560 }, { "grad_norm": 0.4606916308403015, "learning_rate": 9.233888026283999e-05, "loss": 0.0463, "step": 17570 }, { "grad_norm": 0.3976791799068451, "learning_rate": 9.232788216552619e-05, "loss": 0.0419, "step": 17580 }, { "grad_norm": 0.28675395250320435, "learning_rate": 9.231687683553502e-05, "loss": 0.0474, "step": 17590 }, { "grad_norm": 0.4307745397090912, "learning_rate": 9.230586427474698e-05, "loss": 0.0511, "step": 17600 }, { "grad_norm": 0.4655638635158539, "learning_rate": 9.229484448504379e-05, "loss": 0.0431, "step": 17610 }, { "grad_norm": 0.3847913146018982, "learning_rate": 9.228381746830843e-05, "loss": 0.0378, "step": 17620 }, { "grad_norm": 0.4771474599838257, "learning_rate": 9.227278322642514e-05, "loss": 0.033, "step": 17630 }, { "grad_norm": 0.28638598322868347, "learning_rate": 9.226174176127937e-05, "loss": 0.0368, "step": 17640 }, { "grad_norm": 0.3649148941040039, "learning_rate": 9.22506930747578e-05, "loss": 0.0362, "step": 17650 }, { "grad_norm": 0.41134127974510193, "learning_rate": 9.223963716874831e-05, "loss": 0.044, "step": 17660 }, { "grad_norm": 0.4670550227165222, "learning_rate": 9.222857404514012e-05, "loss": 0.0399, "step": 17670 }, { "grad_norm": 0.5582402944564819, "learning_rate": 9.221750370582355e-05, "loss": 0.0416, "step": 17680 }, { "grad_norm": 0.3914591073989868, "learning_rate": 9.220642615269028e-05, "loss": 0.0352, "step": 17690 }, { "grad_norm": 0.36319270730018616, "learning_rate": 9.219534138763311e-05, "loss": 0.0374, "step": 17700 }, { "grad_norm": 0.4824574291706085, "learning_rate": 9.218424941254613e-05, "loss": 0.0385, "step": 17710 }, { "grad_norm": 0.31798845529556274, "learning_rate": 9.217315022932468e-05, "loss": 0.038, "step": 17720 }, { "grad_norm": 0.37390971183776855, "learning_rate": 9.216204383986528e-05, "loss": 0.0384, "step": 17730 }, { "grad_norm": 0.35454508662223816, "learning_rate": 9.215093024606574e-05, "loss": 0.0471, "step": 17740 }, { "grad_norm": 0.41080155968666077, "learning_rate": 9.213980944982506e-05, "loss": 0.0553, "step": 17750 }, { "grad_norm": 0.35368484258651733, "learning_rate": 9.212868145304346e-05, "loss": 0.0467, "step": 17760 }, { "grad_norm": 0.30274420976638794, "learning_rate": 9.211754625762241e-05, "loss": 0.0381, "step": 17770 }, { "grad_norm": 0.5764412879943848, "learning_rate": 9.210640386546463e-05, "loss": 0.0481, "step": 17780 }, { "grad_norm": 0.3496193587779999, "learning_rate": 9.209525427847405e-05, "loss": 0.0339, "step": 17790 }, { "grad_norm": 0.3484799265861511, "learning_rate": 9.208409749855583e-05, "loss": 0.0382, "step": 17800 }, { "grad_norm": 0.3852540850639343, "learning_rate": 9.207293352761633e-05, "loss": 0.039, "step": 17810 }, { "grad_norm": 0.5147120952606201, "learning_rate": 9.206176236756319e-05, "loss": 0.0434, "step": 17820 }, { "grad_norm": 0.49264588952064514, "learning_rate": 9.205058402030525e-05, "loss": 0.0377, "step": 17830 }, { "grad_norm": 0.5155923366546631, "learning_rate": 9.203939848775259e-05, "loss": 0.0373, "step": 17840 }, { "grad_norm": 0.44983842968940735, "learning_rate": 9.202820577181652e-05, "loss": 0.0418, "step": 17850 }, { "grad_norm": 0.28116947412490845, "learning_rate": 9.201700587440953e-05, "loss": 0.0383, "step": 17860 }, { "grad_norm": 0.3772192597389221, "learning_rate": 9.200579879744544e-05, "loss": 0.0354, "step": 17870 }, { "grad_norm": 0.45001187920570374, "learning_rate": 9.199458454283918e-05, "loss": 0.0332, "step": 17880 }, { "grad_norm": 0.2929866909980774, "learning_rate": 9.198336311250697e-05, "loss": 0.0367, "step": 17890 }, { "grad_norm": 0.30929452180862427, "learning_rate": 9.197213450836626e-05, "loss": 0.0285, "step": 17900 }, { "grad_norm": 0.5823236703872681, "learning_rate": 9.19608987323357e-05, "loss": 0.0341, "step": 17910 }, { "grad_norm": 0.31022801995277405, "learning_rate": 9.194965578633517e-05, "loss": 0.0323, "step": 17920 }, { "grad_norm": 0.31062379479408264, "learning_rate": 9.193840567228582e-05, "loss": 0.0355, "step": 17930 }, { "grad_norm": 0.34882277250289917, "learning_rate": 9.192714839210994e-05, "loss": 0.032, "step": 17940 }, { "grad_norm": 0.4040907323360443, "learning_rate": 9.19158839477311e-05, "loss": 0.0472, "step": 17950 }, { "grad_norm": 0.36768054962158203, "learning_rate": 9.190461234107411e-05, "loss": 0.038, "step": 17960 }, { "grad_norm": 0.2231403887271881, "learning_rate": 9.189333357406496e-05, "loss": 0.0321, "step": 17970 }, { "grad_norm": 0.41800394654273987, "learning_rate": 9.188204764863089e-05, "loss": 0.0426, "step": 17980 }, { "grad_norm": 0.4594769775867462, "learning_rate": 9.187075456670033e-05, "loss": 0.0299, "step": 17990 }, { "grad_norm": 0.37939977645874023, "learning_rate": 9.1859454330203e-05, "loss": 0.0475, "step": 18000 }, { "grad_norm": 0.42591777443885803, "learning_rate": 9.18481469410698e-05, "loss": 0.0474, "step": 18010 }, { "grad_norm": 0.4475117325782776, "learning_rate": 9.183683240123281e-05, "loss": 0.0384, "step": 18020 }, { "grad_norm": 0.4148818552494049, "learning_rate": 9.182551071262541e-05, "loss": 0.0452, "step": 18030 }, { "grad_norm": 0.4309631586074829, "learning_rate": 9.181418187718218e-05, "loss": 0.0404, "step": 18040 }, { "grad_norm": 0.41764354705810547, "learning_rate": 9.180284589683888e-05, "loss": 0.0416, "step": 18050 }, { "grad_norm": 0.49179181456565857, "learning_rate": 9.17915027735325e-05, "loss": 0.0334, "step": 18060 }, { "grad_norm": 0.3504154682159424, "learning_rate": 9.178015250920133e-05, "loss": 0.0382, "step": 18070 }, { "grad_norm": 0.3764403760433197, "learning_rate": 9.176879510578477e-05, "loss": 0.0353, "step": 18080 }, { "grad_norm": 0.3864254057407379, "learning_rate": 9.17574305652235e-05, "loss": 0.0402, "step": 18090 }, { "grad_norm": 0.3306662440299988, "learning_rate": 9.174605888945942e-05, "loss": 0.0493, "step": 18100 }, { "grad_norm": 0.4256729483604431, "learning_rate": 9.173468008043564e-05, "loss": 0.0294, "step": 18110 }, { "grad_norm": 0.35903865098953247, "learning_rate": 9.172329414009648e-05, "loss": 0.0379, "step": 18120 }, { "grad_norm": 0.3944171667098999, "learning_rate": 9.171190107038747e-05, "loss": 0.0362, "step": 18130 }, { "grad_norm": 0.2749289870262146, "learning_rate": 9.170050087325541e-05, "loss": 0.0289, "step": 18140 }, { "grad_norm": 0.43280458450317383, "learning_rate": 9.168909355064824e-05, "loss": 0.0361, "step": 18150 }, { "grad_norm": 0.3390047252178192, "learning_rate": 9.167767910451519e-05, "loss": 0.0342, "step": 18160 }, { "grad_norm": 0.3020128905773163, "learning_rate": 9.166625753680669e-05, "loss": 0.0362, "step": 18170 }, { "grad_norm": 0.4515606164932251, "learning_rate": 9.165482884947431e-05, "loss": 0.0315, "step": 18180 }, { "grad_norm": 0.4086684286594391, "learning_rate": 9.164339304447098e-05, "loss": 0.042, "step": 18190 }, { "grad_norm": 0.3192857801914215, "learning_rate": 9.163195012375072e-05, "loss": 0.0324, "step": 18200 }, { "grad_norm": 0.29211920499801636, "learning_rate": 9.16205000892688e-05, "loss": 0.034, "step": 18210 }, { "grad_norm": 0.47168800234794617, "learning_rate": 9.160904294298175e-05, "loss": 0.0281, "step": 18220 }, { "grad_norm": 0.4007086157798767, "learning_rate": 9.159757868684727e-05, "loss": 0.0303, "step": 18230 }, { "grad_norm": 0.5091307163238525, "learning_rate": 9.15861073228243e-05, "loss": 0.0324, "step": 18240 }, { "grad_norm": 0.39866212010383606, "learning_rate": 9.157462885287296e-05, "loss": 0.0497, "step": 18250 }, { "grad_norm": 0.40801867842674255, "learning_rate": 9.156314327895461e-05, "loss": 0.0416, "step": 18260 }, { "grad_norm": 0.5816314220428467, "learning_rate": 9.155165060303185e-05, "loss": 0.0366, "step": 18270 }, { "grad_norm": 0.3293011784553528, "learning_rate": 9.154015082706841e-05, "loss": 0.0341, "step": 18280 }, { "grad_norm": 0.4846779406070709, "learning_rate": 9.152864395302936e-05, "loss": 0.0419, "step": 18290 }, { "grad_norm": 0.4705699384212494, "learning_rate": 9.151712998288085e-05, "loss": 0.0358, "step": 18300 }, { "grad_norm": 0.44931527972221375, "learning_rate": 9.150560891859031e-05, "loss": 0.0401, "step": 18310 }, { "grad_norm": 0.33627644181251526, "learning_rate": 9.14940807621264e-05, "loss": 0.0348, "step": 18320 }, { "grad_norm": 0.34364256262779236, "learning_rate": 9.148254551545894e-05, "loss": 0.0399, "step": 18330 }, { "grad_norm": 0.46985241770744324, "learning_rate": 9.147100318055901e-05, "loss": 0.0383, "step": 18340 }, { "grad_norm": 0.31092146039009094, "learning_rate": 9.145945375939888e-05, "loss": 0.0358, "step": 18350 }, { "grad_norm": 0.506902813911438, "learning_rate": 9.144789725395203e-05, "loss": 0.0306, "step": 18360 }, { "grad_norm": 0.3405759930610657, "learning_rate": 9.14363336661931e-05, "loss": 0.0367, "step": 18370 }, { "grad_norm": 0.2712154686450958, "learning_rate": 9.142476299809806e-05, "loss": 0.033, "step": 18380 }, { "grad_norm": 0.20919425785541534, "learning_rate": 9.1413185251644e-05, "loss": 0.0281, "step": 18390 }, { "grad_norm": 0.3677268326282501, "learning_rate": 9.140160042880923e-05, "loss": 0.0345, "step": 18400 }, { "grad_norm": 0.4180575907230377, "learning_rate": 9.139000853157327e-05, "loss": 0.0325, "step": 18410 }, { "grad_norm": 0.4417911469936371, "learning_rate": 9.137840956191688e-05, "loss": 0.0356, "step": 18420 }, { "grad_norm": 0.3415553867816925, "learning_rate": 9.136680352182199e-05, "loss": 0.0286, "step": 18430 }, { "grad_norm": 0.4877130091190338, "learning_rate": 9.135519041327177e-05, "loss": 0.0268, "step": 18440 }, { "grad_norm": 0.3918581008911133, "learning_rate": 9.134357023825058e-05, "loss": 0.0396, "step": 18450 }, { "grad_norm": 0.3559073507785797, "learning_rate": 9.133194299874398e-05, "loss": 0.0312, "step": 18460 }, { "grad_norm": 0.4916282594203949, "learning_rate": 9.132030869673876e-05, "loss": 0.0358, "step": 18470 }, { "grad_norm": 0.43689608573913574, "learning_rate": 9.130866733422288e-05, "loss": 0.0288, "step": 18480 }, { "grad_norm": 0.5139758586883545, "learning_rate": 9.129701891318556e-05, "loss": 0.0369, "step": 18490 }, { "grad_norm": 0.3997645676136017, "learning_rate": 9.128536343561718e-05, "loss": 0.0455, "step": 18500 }, { "grad_norm": 0.25787124037742615, "learning_rate": 9.127370090350934e-05, "loss": 0.037, "step": 18510 }, { "grad_norm": 0.6054443120956421, "learning_rate": 9.126203131885487e-05, "loss": 0.0373, "step": 18520 }, { "grad_norm": 0.5543069243431091, "learning_rate": 9.125035468364775e-05, "loss": 0.0392, "step": 18530 }, { "grad_norm": 0.4545454978942871, "learning_rate": 9.123867099988322e-05, "loss": 0.0304, "step": 18540 }, { "grad_norm": 0.3896959125995636, "learning_rate": 9.122698026955769e-05, "loss": 0.0359, "step": 18550 }, { "grad_norm": 0.5248709321022034, "learning_rate": 9.12152824946688e-05, "loss": 0.0464, "step": 18560 }, { "grad_norm": 0.3800099492073059, "learning_rate": 9.120357767721538e-05, "loss": 0.0364, "step": 18570 }, { "grad_norm": 0.4662027657032013, "learning_rate": 9.119186581919745e-05, "loss": 0.0372, "step": 18580 }, { "grad_norm": 0.5371197462081909, "learning_rate": 9.118014692261624e-05, "loss": 0.0336, "step": 18590 }, { "grad_norm": 0.6174297332763672, "learning_rate": 9.116842098947422e-05, "loss": 0.0521, "step": 18600 }, { "grad_norm": 0.30853429436683655, "learning_rate": 9.115668802177499e-05, "loss": 0.0329, "step": 18610 }, { "grad_norm": 0.31458356976509094, "learning_rate": 9.114494802152342e-05, "loss": 0.0537, "step": 18620 }, { "grad_norm": 0.5977941751480103, "learning_rate": 9.113320099072555e-05, "loss": 0.0646, "step": 18630 }, { "grad_norm": 0.48152634501457214, "learning_rate": 9.112144693138864e-05, "loss": 0.0389, "step": 18640 }, { "grad_norm": 0.5731042623519897, "learning_rate": 9.110968584552111e-05, "loss": 0.0352, "step": 18650 }, { "grad_norm": 0.24920563399791718, "learning_rate": 9.109791773513264e-05, "loss": 0.0334, "step": 18660 }, { "grad_norm": 0.3133090138435364, "learning_rate": 9.108614260223403e-05, "loss": 0.0357, "step": 18670 }, { "grad_norm": 0.5300940275192261, "learning_rate": 9.107436044883738e-05, "loss": 0.0354, "step": 18680 }, { "grad_norm": 0.38372376561164856, "learning_rate": 9.10625712769559e-05, "loss": 0.0322, "step": 18690 }, { "grad_norm": 0.2960982918739319, "learning_rate": 9.105077508860406e-05, "loss": 0.0264, "step": 18700 }, { "grad_norm": 0.3330300450325012, "learning_rate": 9.103897188579751e-05, "loss": 0.0294, "step": 18710 }, { "grad_norm": 0.4493769407272339, "learning_rate": 9.102716167055308e-05, "loss": 0.0264, "step": 18720 }, { "grad_norm": 0.5176987051963806, "learning_rate": 9.10153444448888e-05, "loss": 0.0528, "step": 18730 }, { "grad_norm": 0.4434530436992645, "learning_rate": 9.100352021082393e-05, "loss": 0.0284, "step": 18740 }, { "grad_norm": 0.3010551631450653, "learning_rate": 9.099168897037891e-05, "loss": 0.029, "step": 18750 }, { "grad_norm": 0.4152607321739197, "learning_rate": 9.097985072557538e-05, "loss": 0.0319, "step": 18760 }, { "grad_norm": 0.2953483462333679, "learning_rate": 9.096800547843615e-05, "loss": 0.0489, "step": 18770 }, { "grad_norm": 0.7732467651367188, "learning_rate": 9.095615323098526e-05, "loss": 0.0506, "step": 18780 }, { "grad_norm": 0.563679039478302, "learning_rate": 9.094429398524795e-05, "loss": 0.0385, "step": 18790 }, { "grad_norm": 0.44332849979400635, "learning_rate": 9.093242774325061e-05, "loss": 0.0376, "step": 18800 }, { "grad_norm": 0.3040858507156372, "learning_rate": 9.092055450702088e-05, "loss": 0.0315, "step": 18810 }, { "grad_norm": 0.5142569541931152, "learning_rate": 9.090867427858756e-05, "loss": 0.0357, "step": 18820 }, { "grad_norm": 0.3836390972137451, "learning_rate": 9.089678705998066e-05, "loss": 0.0349, "step": 18830 }, { "grad_norm": 0.3291219174861908, "learning_rate": 9.088489285323139e-05, "loss": 0.0318, "step": 18840 }, { "grad_norm": 0.30687907338142395, "learning_rate": 9.087299166037212e-05, "loss": 0.0328, "step": 18850 }, { "grad_norm": 0.46690651774406433, "learning_rate": 9.086108348343647e-05, "loss": 0.0406, "step": 18860 }, { "grad_norm": 0.5696375966072083, "learning_rate": 9.08491683244592e-05, "loss": 0.04, "step": 18870 }, { "grad_norm": 0.31647542119026184, "learning_rate": 9.08372461854763e-05, "loss": 0.0354, "step": 18880 }, { "grad_norm": 0.3327438533306122, "learning_rate": 9.082531706852492e-05, "loss": 0.0289, "step": 18890 }, { "grad_norm": 0.2817155420780182, "learning_rate": 9.081338097564342e-05, "loss": 0.0421, "step": 18900 }, { "grad_norm": 0.281390905380249, "learning_rate": 9.080143790887137e-05, "loss": 0.0339, "step": 18910 }, { "grad_norm": 0.41964274644851685, "learning_rate": 9.07894878702495e-05, "loss": 0.0362, "step": 18920 }, { "grad_norm": 0.31665414571762085, "learning_rate": 9.077753086181974e-05, "loss": 0.0389, "step": 18930 }, { "grad_norm": 0.39004820585250854, "learning_rate": 9.076556688562524e-05, "loss": 0.0312, "step": 18940 }, { "grad_norm": 0.34970614314079285, "learning_rate": 9.075359594371029e-05, "loss": 0.028, "step": 18950 }, { "grad_norm": 0.3763287663459778, "learning_rate": 9.07416180381204e-05, "loss": 0.036, "step": 18960 }, { "grad_norm": 0.3054503798484802, "learning_rate": 9.072963317090228e-05, "loss": 0.0344, "step": 18970 }, { "grad_norm": 0.41767483949661255, "learning_rate": 9.071764134410382e-05, "loss": 0.0339, "step": 18980 }, { "grad_norm": 0.39803773164749146, "learning_rate": 9.070564255977407e-05, "loss": 0.0287, "step": 18990 }, { "grad_norm": 0.5632229447364807, "learning_rate": 9.06936368199633e-05, "loss": 0.0479, "step": 19000 }, { "grad_norm": 0.42512834072113037, "learning_rate": 9.0681624126723e-05, "loss": 0.0398, "step": 19010 }, { "grad_norm": 0.41683536767959595, "learning_rate": 9.066960448210576e-05, "loss": 0.0398, "step": 19020 }, { "grad_norm": 0.3936285078525543, "learning_rate": 9.065757788816543e-05, "loss": 0.0319, "step": 19030 }, { "grad_norm": 0.35685667395591736, "learning_rate": 9.064554434695705e-05, "loss": 0.0341, "step": 19040 }, { "grad_norm": 0.30630746483802795, "learning_rate": 9.063350386053677e-05, "loss": 0.0329, "step": 19050 }, { "grad_norm": 0.3663657605648041, "learning_rate": 9.062145643096202e-05, "loss": 0.0333, "step": 19060 }, { "grad_norm": 0.28488802909851074, "learning_rate": 9.060940206029136e-05, "loss": 0.0345, "step": 19070 }, { "grad_norm": 0.31536999344825745, "learning_rate": 9.059734075058457e-05, "loss": 0.0361, "step": 19080 }, { "grad_norm": 0.34442877769470215, "learning_rate": 9.058527250390257e-05, "loss": 0.0381, "step": 19090 }, { "grad_norm": 0.5297933220863342, "learning_rate": 9.057319732230752e-05, "loss": 0.0444, "step": 19100 }, { "grad_norm": 0.3689417839050293, "learning_rate": 9.056111520786273e-05, "loss": 0.0354, "step": 19110 }, { "grad_norm": 0.3053085505962372, "learning_rate": 9.054902616263268e-05, "loss": 0.0337, "step": 19120 }, { "grad_norm": 0.344207227230072, "learning_rate": 9.05369301886831e-05, "loss": 0.048, "step": 19130 }, { "grad_norm": 0.3079240620136261, "learning_rate": 9.052482728808083e-05, "loss": 0.0311, "step": 19140 }, { "grad_norm": 0.40268778800964355, "learning_rate": 9.051271746289391e-05, "loss": 0.0286, "step": 19150 }, { "grad_norm": 0.619869589805603, "learning_rate": 9.050060071519162e-05, "loss": 0.0385, "step": 19160 }, { "grad_norm": 0.38367223739624023, "learning_rate": 9.048847704704437e-05, "loss": 0.0329, "step": 19170 }, { "grad_norm": 0.37065252661705017, "learning_rate": 9.047634646052376e-05, "loss": 0.0354, "step": 19180 }, { "grad_norm": 0.4620060622692108, "learning_rate": 9.046420895770256e-05, "loss": 0.0497, "step": 19190 }, { "grad_norm": 0.3684999942779541, "learning_rate": 9.045206454065473e-05, "loss": 0.0344, "step": 19200 }, { "grad_norm": 0.4145887494087219, "learning_rate": 9.043991321145546e-05, "loss": 0.0336, "step": 19210 }, { "grad_norm": 0.5166178345680237, "learning_rate": 9.042775497218105e-05, "loss": 0.0341, "step": 19220 }, { "grad_norm": 0.4706641435623169, "learning_rate": 9.041558982490901e-05, "loss": 0.0467, "step": 19230 }, { "grad_norm": 0.29291287064552307, "learning_rate": 9.040341777171805e-05, "loss": 0.0373, "step": 19240 }, { "grad_norm": 0.5256592631340027, "learning_rate": 9.039123881468802e-05, "loss": 0.0397, "step": 19250 }, { "grad_norm": 0.4370717704296112, "learning_rate": 9.037905295589998e-05, "loss": 0.0347, "step": 19260 }, { "grad_norm": 0.4141928553581238, "learning_rate": 9.036686019743617e-05, "loss": 0.029, "step": 19270 }, { "grad_norm": 0.20351646840572357, "learning_rate": 9.035466054137997e-05, "loss": 0.0351, "step": 19280 }, { "grad_norm": 0.357643187046051, "learning_rate": 9.0342453989816e-05, "loss": 0.0395, "step": 19290 }, { "grad_norm": 0.3223193287849426, "learning_rate": 9.033024054483e-05, "loss": 0.0354, "step": 19300 }, { "grad_norm": 0.5753912329673767, "learning_rate": 9.031802020850894e-05, "loss": 0.0406, "step": 19310 }, { "grad_norm": 0.4590586721897125, "learning_rate": 9.030579298294092e-05, "loss": 0.0421, "step": 19320 }, { "grad_norm": 0.39201462268829346, "learning_rate": 9.029355887021524e-05, "loss": 0.0472, "step": 19330 }, { "grad_norm": 0.7150716781616211, "learning_rate": 9.028131787242238e-05, "loss": 0.0537, "step": 19340 }, { "grad_norm": 0.27760976552963257, "learning_rate": 9.026906999165399e-05, "loss": 0.0432, "step": 19350 }, { "grad_norm": 0.32481324672698975, "learning_rate": 9.025681523000291e-05, "loss": 0.0311, "step": 19360 }, { "grad_norm": 0.3101690709590912, "learning_rate": 9.024455358956315e-05, "loss": 0.0326, "step": 19370 }, { "grad_norm": 0.44484737515449524, "learning_rate": 9.023228507242984e-05, "loss": 0.0399, "step": 19380 }, { "grad_norm": 0.3129139244556427, "learning_rate": 9.022000968069937e-05, "loss": 0.0369, "step": 19390 }, { "grad_norm": 0.3746558725833893, "learning_rate": 9.020772741646928e-05, "loss": 0.0377, "step": 19400 }, { "grad_norm": 0.5838225483894348, "learning_rate": 9.019543828183826e-05, "loss": 0.0424, "step": 19410 }, { "grad_norm": 0.3720093369483948, "learning_rate": 9.018314227890616e-05, "loss": 0.0355, "step": 19420 }, { "grad_norm": 0.6010069847106934, "learning_rate": 9.017083940977408e-05, "loss": 0.0483, "step": 19430 }, { "grad_norm": 0.19848327338695526, "learning_rate": 9.015852967654422e-05, "loss": 0.0313, "step": 19440 }, { "grad_norm": 0.36479002237319946, "learning_rate": 9.014621308131996e-05, "loss": 0.0326, "step": 19450 }, { "grad_norm": 0.495022177696228, "learning_rate": 9.01338896262059e-05, "loss": 0.0397, "step": 19460 }, { "grad_norm": 0.3332238495349884, "learning_rate": 9.012155931330777e-05, "loss": 0.037, "step": 19470 }, { "grad_norm": 0.28620001673698425, "learning_rate": 9.010922214473246e-05, "loss": 0.0331, "step": 19480 }, { "grad_norm": 0.3401290774345398, "learning_rate": 9.009687812258808e-05, "loss": 0.0508, "step": 19490 }, { "grad_norm": 0.42446014285087585, "learning_rate": 9.00845272489839e-05, "loss": 0.0434, "step": 19500 }, { "grad_norm": 0.39021220803260803, "learning_rate": 9.007216952603031e-05, "loss": 0.0418, "step": 19510 }, { "grad_norm": 0.3474824130535126, "learning_rate": 9.005980495583894e-05, "loss": 0.03, "step": 19520 }, { "grad_norm": 0.3347165882587433, "learning_rate": 9.004743354052252e-05, "loss": 0.0315, "step": 19530 }, { "grad_norm": 0.5676416754722595, "learning_rate": 9.003505528219503e-05, "loss": 0.0365, "step": 19540 }, { "grad_norm": 0.5669789910316467, "learning_rate": 9.002267018297154e-05, "loss": 0.0352, "step": 19550 }, { "grad_norm": 0.5903368592262268, "learning_rate": 9.001027824496834e-05, "loss": 0.043, "step": 19560 }, { "grad_norm": 0.47973254323005676, "learning_rate": 8.999787947030287e-05, "loss": 0.04, "step": 19570 }, { "grad_norm": 0.28995245695114136, "learning_rate": 8.998547386109376e-05, "loss": 0.0418, "step": 19580 }, { "grad_norm": 0.3136598765850067, "learning_rate": 8.997306141946073e-05, "loss": 0.0354, "step": 19590 }, { "grad_norm": 0.5864888429641724, "learning_rate": 8.996064214752481e-05, "loss": 0.038, "step": 19600 }, { "grad_norm": 0.4611058533191681, "learning_rate": 8.994821604740806e-05, "loss": 0.0358, "step": 19610 }, { "grad_norm": 0.3926291763782501, "learning_rate": 8.993578312123377e-05, "loss": 0.0465, "step": 19620 }, { "grad_norm": 0.39541009068489075, "learning_rate": 8.992334337112639e-05, "loss": 0.0372, "step": 19630 }, { "grad_norm": 0.3489963114261627, "learning_rate": 8.991089679921154e-05, "loss": 0.0314, "step": 19640 }, { "grad_norm": 0.37576282024383545, "learning_rate": 8.989844340761599e-05, "loss": 0.0406, "step": 19650 }, { "grad_norm": 0.36199140548706055, "learning_rate": 8.988598319846768e-05, "loss": 0.0353, "step": 19660 }, { "grad_norm": 0.27743932604789734, "learning_rate": 8.987351617389574e-05, "loss": 0.0315, "step": 19670 }, { "grad_norm": 0.2813369333744049, "learning_rate": 8.98610423360304e-05, "loss": 0.0289, "step": 19680 }, { "grad_norm": 0.31966832280158997, "learning_rate": 8.984856168700317e-05, "loss": 0.037, "step": 19690 }, { "grad_norm": 0.30365487933158875, "learning_rate": 8.983607422894658e-05, "loss": 0.0338, "step": 19700 }, { "grad_norm": 0.2772423326969147, "learning_rate": 8.982357996399442e-05, "loss": 0.0361, "step": 19710 }, { "grad_norm": 0.40653613209724426, "learning_rate": 8.981107889428164e-05, "loss": 0.0342, "step": 19720 }, { "grad_norm": 0.39411771297454834, "learning_rate": 8.979857102194428e-05, "loss": 0.0332, "step": 19730 }, { "grad_norm": 0.2497989386320114, "learning_rate": 8.978605634911968e-05, "loss": 0.0323, "step": 19740 }, { "grad_norm": 0.33451470732688904, "learning_rate": 8.977353487794616e-05, "loss": 0.029, "step": 19750 }, { "grad_norm": 0.8523989915847778, "learning_rate": 8.976100661056334e-05, "loss": 0.0403, "step": 19760 }, { "grad_norm": 0.3498835265636444, "learning_rate": 8.974847154911197e-05, "loss": 0.0325, "step": 19770 }, { "grad_norm": 0.36118990182876587, "learning_rate": 8.973592969573393e-05, "loss": 0.0413, "step": 19780 }, { "grad_norm": 0.3502088487148285, "learning_rate": 8.972338105257228e-05, "loss": 0.04, "step": 19790 }, { "grad_norm": 0.3978632986545563, "learning_rate": 8.971082562177125e-05, "loss": 0.0316, "step": 19800 }, { "grad_norm": 0.623081386089325, "learning_rate": 8.96982634054762e-05, "loss": 0.0336, "step": 19810 }, { "grad_norm": 0.4447407126426697, "learning_rate": 8.96856944058337e-05, "loss": 0.046, "step": 19820 }, { "grad_norm": 0.3560118079185486, "learning_rate": 8.967311862499144e-05, "loss": 0.0261, "step": 19830 }, { "grad_norm": 0.31257981061935425, "learning_rate": 8.966053606509825e-05, "loss": 0.0263, "step": 19840 }, { "grad_norm": 0.5914917588233948, "learning_rate": 8.964794672830417e-05, "loss": 0.0404, "step": 19850 }, { "grad_norm": 0.42860981822013855, "learning_rate": 8.963535061676038e-05, "loss": 0.0449, "step": 19860 }, { "grad_norm": 0.6923375725746155, "learning_rate": 8.962274773261918e-05, "loss": 0.0396, "step": 19870 }, { "grad_norm": 0.6002978086471558, "learning_rate": 8.961013807803409e-05, "loss": 0.0485, "step": 19880 }, { "grad_norm": 0.2819831669330597, "learning_rate": 8.959752165515973e-05, "loss": 0.0449, "step": 19890 }, { "grad_norm": 0.5620126128196716, "learning_rate": 8.958489846615193e-05, "loss": 0.0548, "step": 19900 }, { "grad_norm": 0.5387278199195862, "learning_rate": 8.957226851316762e-05, "loss": 0.0422, "step": 19910 }, { "grad_norm": 0.3912273943424225, "learning_rate": 8.955963179836493e-05, "loss": 0.0341, "step": 19920 }, { "grad_norm": 0.6490488648414612, "learning_rate": 8.954698832390312e-05, "loss": 0.0449, "step": 19930 }, { "grad_norm": 0.37447792291641235, "learning_rate": 8.953433809194263e-05, "loss": 0.0335, "step": 19940 }, { "grad_norm": 0.5768982768058777, "learning_rate": 8.9521681104645e-05, "loss": 0.0355, "step": 19950 }, { "grad_norm": 0.45225200057029724, "learning_rate": 8.9509017364173e-05, "loss": 0.0302, "step": 19960 }, { "grad_norm": 0.4435863792896271, "learning_rate": 8.949634687269052e-05, "loss": 0.0398, "step": 19970 }, { "grad_norm": 0.6348145604133606, "learning_rate": 8.948366963236259e-05, "loss": 0.031, "step": 19980 }, { "grad_norm": 0.3101019859313965, "learning_rate": 8.947098564535538e-05, "loss": 0.035, "step": 19990 }, { "grad_norm": 0.4263094663619995, "learning_rate": 8.945829491383627e-05, "loss": 0.037, "step": 20000 }, { "grad_norm": 0.3460792303085327, "learning_rate": 8.944559743997374e-05, "loss": 0.035, "step": 20010 }, { "grad_norm": 0.38574904203414917, "learning_rate": 8.943289322593746e-05, "loss": 0.0342, "step": 20020 }, { "grad_norm": 0.35784927010536194, "learning_rate": 8.942018227389821e-05, "loss": 0.0354, "step": 20030 }, { "grad_norm": 0.33790117502212524, "learning_rate": 8.940746458602795e-05, "loss": 0.0549, "step": 20040 }, { "grad_norm": 0.4488186836242676, "learning_rate": 8.939474016449979e-05, "loss": 0.045, "step": 20050 }, { "grad_norm": 0.2566605508327484, "learning_rate": 8.938200901148799e-05, "loss": 0.031, "step": 20060 }, { "grad_norm": 0.5514219999313354, "learning_rate": 8.936927112916795e-05, "loss": 0.0446, "step": 20070 }, { "grad_norm": 0.4234762191772461, "learning_rate": 8.935652651971622e-05, "loss": 0.0306, "step": 20080 }, { "grad_norm": 0.3463698923587799, "learning_rate": 8.934377518531052e-05, "loss": 0.0404, "step": 20090 }, { "grad_norm": 0.35089200735092163, "learning_rate": 8.933101712812967e-05, "loss": 0.0338, "step": 20100 }, { "grad_norm": 0.3580324053764343, "learning_rate": 8.931825235035374e-05, "loss": 0.0302, "step": 20110 }, { "grad_norm": 0.34666740894317627, "learning_rate": 8.930548085416382e-05, "loss": 0.0303, "step": 20120 }, { "grad_norm": 0.4500749111175537, "learning_rate": 8.92927026417422e-05, "loss": 0.0398, "step": 20130 }, { "grad_norm": 0.3201676309108734, "learning_rate": 8.92799177152724e-05, "loss": 0.0319, "step": 20140 }, { "grad_norm": 0.5221537351608276, "learning_rate": 8.926712607693895e-05, "loss": 0.0315, "step": 20150 }, { "grad_norm": 0.5236226916313171, "learning_rate": 8.925432772892762e-05, "loss": 0.0358, "step": 20160 }, { "grad_norm": 0.3373253047466278, "learning_rate": 8.924152267342529e-05, "loss": 0.0409, "step": 20170 }, { "grad_norm": 0.42660239338874817, "learning_rate": 8.922871091261998e-05, "loss": 0.0397, "step": 20180 }, { "grad_norm": 0.31154969334602356, "learning_rate": 8.92158924487009e-05, "loss": 0.0255, "step": 20190 }, { "grad_norm": 0.4468379616737366, "learning_rate": 8.920306728385834e-05, "loss": 0.0275, "step": 20200 }, { "grad_norm": 0.45120906829833984, "learning_rate": 8.919023542028379e-05, "loss": 0.0282, "step": 20210 }, { "grad_norm": 0.3031345307826996, "learning_rate": 8.917739686016988e-05, "loss": 0.0281, "step": 20220 }, { "grad_norm": 0.38362306356430054, "learning_rate": 8.916455160571033e-05, "loss": 0.0393, "step": 20230 }, { "grad_norm": 0.25926607847213745, "learning_rate": 8.915169965910008e-05, "loss": 0.028, "step": 20240 }, { "grad_norm": 0.29281315207481384, "learning_rate": 8.913884102253514e-05, "loss": 0.0273, "step": 20250 }, { "grad_norm": 0.29503753781318665, "learning_rate": 8.912597569821273e-05, "loss": 0.0262, "step": 20260 }, { "grad_norm": 0.2778354585170746, "learning_rate": 8.911310368833118e-05, "loss": 0.0297, "step": 20270 }, { "grad_norm": 0.3758085072040558, "learning_rate": 8.910022499508994e-05, "loss": 0.0324, "step": 20280 }, { "grad_norm": 0.40501460433006287, "learning_rate": 8.908733962068965e-05, "loss": 0.0413, "step": 20290 }, { "grad_norm": 0.4531136453151703, "learning_rate": 8.907444756733207e-05, "loss": 0.0277, "step": 20300 }, { "grad_norm": 0.4465048313140869, "learning_rate": 8.906154883722006e-05, "loss": 0.0384, "step": 20310 }, { "grad_norm": 0.21329285204410553, "learning_rate": 8.904864343255773e-05, "loss": 0.0283, "step": 20320 }, { "grad_norm": 0.3561351001262665, "learning_rate": 8.90357313555502e-05, "loss": 0.0292, "step": 20330 }, { "grad_norm": 0.4635617733001709, "learning_rate": 8.90228126084038e-05, "loss": 0.035, "step": 20340 }, { "grad_norm": 0.4900287091732025, "learning_rate": 8.900988719332601e-05, "loss": 0.0343, "step": 20350 }, { "grad_norm": 0.3142586052417755, "learning_rate": 8.899695511252542e-05, "loss": 0.0373, "step": 20360 }, { "grad_norm": 0.3892967402935028, "learning_rate": 8.898401636821176e-05, "loss": 0.0388, "step": 20370 }, { "grad_norm": 0.44848594069480896, "learning_rate": 8.897107096259593e-05, "loss": 0.0447, "step": 20380 }, { "grad_norm": 0.36050090193748474, "learning_rate": 8.895811889788994e-05, "loss": 0.037, "step": 20390 }, { "grad_norm": 0.3436700105667114, "learning_rate": 8.894516017630692e-05, "loss": 0.0304, "step": 20400 }, { "grad_norm": 0.31890183687210083, "learning_rate": 8.893219480006118e-05, "loss": 0.0296, "step": 20410 }, { "grad_norm": 0.3936845660209656, "learning_rate": 8.891922277136817e-05, "loss": 0.0325, "step": 20420 }, { "grad_norm": 0.45599034428596497, "learning_rate": 8.890624409244441e-05, "loss": 0.0303, "step": 20430 }, { "grad_norm": 0.44714125990867615, "learning_rate": 8.889325876550763e-05, "loss": 0.0451, "step": 20440 }, { "grad_norm": 0.34216219186782837, "learning_rate": 8.888026679277666e-05, "loss": 0.0326, "step": 20450 }, { "grad_norm": 0.27725568413734436, "learning_rate": 8.886726817647147e-05, "loss": 0.0506, "step": 20460 }, { "grad_norm": 0.30971044301986694, "learning_rate": 8.885426291881319e-05, "loss": 0.0379, "step": 20470 }, { "grad_norm": 0.4315445125102997, "learning_rate": 8.884125102202401e-05, "loss": 0.0357, "step": 20480 }, { "grad_norm": 0.40526890754699707, "learning_rate": 8.882823248832736e-05, "loss": 0.0301, "step": 20490 }, { "grad_norm": 0.32696351408958435, "learning_rate": 8.881520731994772e-05, "loss": 0.0349, "step": 20500 }, { "grad_norm": 0.49986639618873596, "learning_rate": 8.880217551911077e-05, "loss": 0.0364, "step": 20510 }, { "grad_norm": 0.47391945123672485, "learning_rate": 8.878913708804323e-05, "loss": 0.038, "step": 20520 }, { "grad_norm": 0.2944822311401367, "learning_rate": 8.877609202897308e-05, "loss": 0.0421, "step": 20530 }, { "grad_norm": 0.5118546485900879, "learning_rate": 8.876304034412933e-05, "loss": 0.0356, "step": 20540 }, { "grad_norm": 0.32708224654197693, "learning_rate": 8.874998203574214e-05, "loss": 0.0359, "step": 20550 }, { "grad_norm": 0.35810017585754395, "learning_rate": 8.873691710604284e-05, "loss": 0.0314, "step": 20560 }, { "grad_norm": 0.485767126083374, "learning_rate": 8.872384555726387e-05, "loss": 0.0332, "step": 20570 }, { "grad_norm": 0.4267512261867523, "learning_rate": 8.871076739163878e-05, "loss": 0.0335, "step": 20580 }, { "grad_norm": 0.26466765999794006, "learning_rate": 8.86976826114023e-05, "loss": 0.0363, "step": 20590 }, { "grad_norm": 0.32073575258255005, "learning_rate": 8.868459121879023e-05, "loss": 0.0386, "step": 20600 }, { "grad_norm": 0.45564791560173035, "learning_rate": 8.867149321603956e-05, "loss": 0.04, "step": 20610 }, { "grad_norm": 0.4985177218914032, "learning_rate": 8.865838860538835e-05, "loss": 0.0349, "step": 20620 }, { "grad_norm": 0.4539857506752014, "learning_rate": 8.864527738907585e-05, "loss": 0.031, "step": 20630 }, { "grad_norm": 0.31574201583862305, "learning_rate": 8.863215956934239e-05, "loss": 0.0469, "step": 20640 }, { "grad_norm": 0.39470696449279785, "learning_rate": 8.861903514842947e-05, "loss": 0.0348, "step": 20650 }, { "grad_norm": 0.37467190623283386, "learning_rate": 8.860590412857966e-05, "loss": 0.0315, "step": 20660 }, { "grad_norm": 0.22000382840633392, "learning_rate": 8.85927665120367e-05, "loss": 0.025, "step": 20670 }, { "grad_norm": 0.3783794343471527, "learning_rate": 8.857962230104546e-05, "loss": 0.0318, "step": 20680 }, { "grad_norm": 0.49539393186569214, "learning_rate": 8.856647149785193e-05, "loss": 0.0304, "step": 20690 }, { "grad_norm": 0.3506929576396942, "learning_rate": 8.855331410470322e-05, "loss": 0.0384, "step": 20700 }, { "grad_norm": 0.38004010915756226, "learning_rate": 8.854015012384756e-05, "loss": 0.034, "step": 20710 }, { "grad_norm": 0.436461478471756, "learning_rate": 8.852697955753433e-05, "loss": 0.0366, "step": 20720 }, { "grad_norm": 0.38045981526374817, "learning_rate": 8.851380240801399e-05, "loss": 0.0406, "step": 20730 }, { "grad_norm": 0.3457326889038086, "learning_rate": 8.850061867753818e-05, "loss": 0.0292, "step": 20740 }, { "grad_norm": 0.3678864538669586, "learning_rate": 8.848742836835963e-05, "loss": 0.0366, "step": 20750 }, { "grad_norm": 0.608887791633606, "learning_rate": 8.847423148273221e-05, "loss": 0.0356, "step": 20760 }, { "grad_norm": 0.4554661512374878, "learning_rate": 8.846102802291092e-05, "loss": 0.04, "step": 20770 }, { "grad_norm": 0.4915081560611725, "learning_rate": 8.844781799115183e-05, "loss": 0.0379, "step": 20780 }, { "grad_norm": 0.3505799472332001, "learning_rate": 8.84346013897122e-05, "loss": 0.0219, "step": 20790 }, { "grad_norm": 0.26255980134010315, "learning_rate": 8.842137822085038e-05, "loss": 0.0364, "step": 20800 }, { "grad_norm": 0.26868292689323425, "learning_rate": 8.840814848682585e-05, "loss": 0.028, "step": 20810 }, { "grad_norm": 0.4176049530506134, "learning_rate": 8.83949121898992e-05, "loss": 0.0438, "step": 20820 }, { "grad_norm": 0.35391274094581604, "learning_rate": 8.838166933233217e-05, "loss": 0.0313, "step": 20830 }, { "grad_norm": 0.4032523036003113, "learning_rate": 8.83684199163876e-05, "loss": 0.0349, "step": 20840 }, { "grad_norm": 0.3495248556137085, "learning_rate": 8.835516394432943e-05, "loss": 0.0302, "step": 20850 }, { "grad_norm": 0.36273306608200073, "learning_rate": 8.834190141842276e-05, "loss": 0.0247, "step": 20860 }, { "grad_norm": 0.4469996392726898, "learning_rate": 8.83286323409338e-05, "loss": 0.0349, "step": 20870 }, { "grad_norm": 0.3663352131843567, "learning_rate": 8.831535671412986e-05, "loss": 0.0339, "step": 20880 }, { "grad_norm": 0.327343612909317, "learning_rate": 8.830207454027938e-05, "loss": 0.0364, "step": 20890 }, { "grad_norm": 0.38637691736221313, "learning_rate": 8.828878582165192e-05, "loss": 0.0274, "step": 20900 }, { "grad_norm": 0.33201420307159424, "learning_rate": 8.827549056051818e-05, "loss": 0.037, "step": 20910 }, { "grad_norm": 0.4773756265640259, "learning_rate": 8.826218875914993e-05, "loss": 0.026, "step": 20920 }, { "grad_norm": 0.34596017003059387, "learning_rate": 8.82488804198201e-05, "loss": 0.0296, "step": 20930 }, { "grad_norm": 0.22919730842113495, "learning_rate": 8.82355655448027e-05, "loss": 0.0339, "step": 20940 }, { "grad_norm": 0.5048438310623169, "learning_rate": 8.822224413637293e-05, "loss": 0.0359, "step": 20950 }, { "grad_norm": 0.36163532733917236, "learning_rate": 8.820891619680697e-05, "loss": 0.0311, "step": 20960 }, { "grad_norm": 0.2937081456184387, "learning_rate": 8.819558172838227e-05, "loss": 0.0288, "step": 20970 }, { "grad_norm": 0.44830042123794556, "learning_rate": 8.818224073337731e-05, "loss": 0.0299, "step": 20980 }, { "grad_norm": 0.3154657185077667, "learning_rate": 8.816889321407169e-05, "loss": 0.028, "step": 20990 }, { "grad_norm": 0.39362800121307373, "learning_rate": 8.815553917274615e-05, "loss": 0.0347, "step": 21000 }, { "grad_norm": 0.37922564148902893, "learning_rate": 8.81421786116825e-05, "loss": 0.0387, "step": 21010 }, { "grad_norm": 0.5413735508918762, "learning_rate": 8.812881153316373e-05, "loss": 0.04, "step": 21020 }, { "grad_norm": 0.3211091160774231, "learning_rate": 8.81154379394739e-05, "loss": 0.0443, "step": 21030 }, { "grad_norm": 0.3985815942287445, "learning_rate": 8.810205783289818e-05, "loss": 0.034, "step": 21040 }, { "grad_norm": 0.35577788949012756, "learning_rate": 8.808867121572286e-05, "loss": 0.0377, "step": 21050 }, { "grad_norm": 0.47649848461151123, "learning_rate": 8.807527809023537e-05, "loss": 0.0505, "step": 21060 }, { "grad_norm": 0.503478467464447, "learning_rate": 8.80618784587242e-05, "loss": 0.0337, "step": 21070 }, { "grad_norm": 0.40865591168403625, "learning_rate": 8.804847232347902e-05, "loss": 0.0332, "step": 21080 }, { "grad_norm": 0.31832996010780334, "learning_rate": 8.803505968679054e-05, "loss": 0.0356, "step": 21090 }, { "grad_norm": 0.3653215765953064, "learning_rate": 8.802164055095061e-05, "loss": 0.043, "step": 21100 }, { "grad_norm": 0.4948742091655731, "learning_rate": 8.80082149182522e-05, "loss": 0.0345, "step": 21110 }, { "grad_norm": 0.3555126190185547, "learning_rate": 8.79947827909894e-05, "loss": 0.0275, "step": 21120 }, { "grad_norm": 0.4026808440685272, "learning_rate": 8.798134417145738e-05, "loss": 0.0362, "step": 21130 }, { "grad_norm": 0.4920814633369446, "learning_rate": 8.796789906195243e-05, "loss": 0.0411, "step": 21140 }, { "grad_norm": 0.42332133650779724, "learning_rate": 8.795444746477195e-05, "loss": 0.0324, "step": 21150 }, { "grad_norm": 0.298618346452713, "learning_rate": 8.794098938221446e-05, "loss": 0.0388, "step": 21160 }, { "grad_norm": 0.32691437005996704, "learning_rate": 8.792752481657957e-05, "loss": 0.041, "step": 21170 }, { "grad_norm": 0.3523212671279907, "learning_rate": 8.791405377016802e-05, "loss": 0.0299, "step": 21180 }, { "grad_norm": 0.5520089864730835, "learning_rate": 8.790057624528163e-05, "loss": 0.04, "step": 21190 }, { "grad_norm": 0.5657264590263367, "learning_rate": 8.788709224422333e-05, "loss": 0.0389, "step": 21200 }, { "grad_norm": 0.4717770218849182, "learning_rate": 8.787360176929717e-05, "loss": 0.0333, "step": 21210 }, { "grad_norm": 0.38040676712989807, "learning_rate": 8.786010482280834e-05, "loss": 0.039, "step": 21220 }, { "grad_norm": 0.2744053900241852, "learning_rate": 8.784660140706306e-05, "loss": 0.0302, "step": 21230 }, { "grad_norm": 0.31193917989730835, "learning_rate": 8.783309152436872e-05, "loss": 0.0332, "step": 21240 }, { "grad_norm": 0.5638653635978699, "learning_rate": 8.781957517703375e-05, "loss": 0.0341, "step": 21250 }, { "grad_norm": 0.6174744367599487, "learning_rate": 8.780605236736776e-05, "loss": 0.0448, "step": 21260 }, { "grad_norm": 0.4371091425418854, "learning_rate": 8.779252309768142e-05, "loss": 0.0397, "step": 21270 }, { "grad_norm": 0.5511850118637085, "learning_rate": 8.777898737028652e-05, "loss": 0.0463, "step": 21280 }, { "grad_norm": 0.49161985516548157, "learning_rate": 8.776544518749591e-05, "loss": 0.0393, "step": 21290 }, { "grad_norm": 0.4897412359714508, "learning_rate": 8.775189655162364e-05, "loss": 0.0374, "step": 21300 }, { "grad_norm": 0.41912007331848145, "learning_rate": 8.773834146498474e-05, "loss": 0.0305, "step": 21310 }, { "grad_norm": 0.3207215368747711, "learning_rate": 8.772477992989545e-05, "loss": 0.0266, "step": 21320 }, { "grad_norm": 0.4687289595603943, "learning_rate": 8.771121194867304e-05, "loss": 0.0397, "step": 21330 }, { "grad_norm": 0.2734229862689972, "learning_rate": 8.769763752363589e-05, "loss": 0.0298, "step": 21340 }, { "grad_norm": 0.22316351532936096, "learning_rate": 8.768405665710352e-05, "loss": 0.0325, "step": 21350 }, { "grad_norm": 0.32610759139060974, "learning_rate": 8.767046935139655e-05, "loss": 0.0333, "step": 21360 }, { "grad_norm": 0.4914608597755432, "learning_rate": 8.765687560883666e-05, "loss": 0.0287, "step": 21370 }, { "grad_norm": 0.33896759152412415, "learning_rate": 8.764327543174664e-05, "loss": 0.0309, "step": 21380 }, { "grad_norm": 0.4655258357524872, "learning_rate": 8.762966882245038e-05, "loss": 0.0307, "step": 21390 }, { "grad_norm": 0.4957077205181122, "learning_rate": 8.761605578327291e-05, "loss": 0.0396, "step": 21400 }, { "grad_norm": 0.30519983172416687, "learning_rate": 8.76024363165403e-05, "loss": 0.0432, "step": 21410 }, { "grad_norm": 0.3053940236568451, "learning_rate": 8.758881042457976e-05, "loss": 0.0405, "step": 21420 }, { "grad_norm": 0.3130364418029785, "learning_rate": 8.757517810971957e-05, "loss": 0.03, "step": 21430 }, { "grad_norm": 0.24744512140750885, "learning_rate": 8.756153937428913e-05, "loss": 0.0315, "step": 21440 }, { "grad_norm": 0.3800404965877533, "learning_rate": 8.754789422061889e-05, "loss": 0.0432, "step": 21450 }, { "grad_norm": 0.40670767426490784, "learning_rate": 8.753424265104052e-05, "loss": 0.0293, "step": 21460 }, { "grad_norm": 0.47903972864151, "learning_rate": 8.752058466788659e-05, "loss": 0.0462, "step": 21470 }, { "grad_norm": 0.3210621178150177, "learning_rate": 8.750692027349097e-05, "loss": 0.0252, "step": 21480 }, { "grad_norm": 0.36464786529541016, "learning_rate": 8.749324947018847e-05, "loss": 0.0275, "step": 21490 }, { "grad_norm": 0.6548550128936768, "learning_rate": 8.747957226031507e-05, "loss": 0.0314, "step": 21500 }, { "grad_norm": 0.34789931774139404, "learning_rate": 8.746588864620787e-05, "loss": 0.0308, "step": 21510 }, { "grad_norm": 0.5326499938964844, "learning_rate": 8.745219863020498e-05, "loss": 0.0342, "step": 21520 }, { "grad_norm": 0.38436466455459595, "learning_rate": 8.743850221464564e-05, "loss": 0.0316, "step": 21530 }, { "grad_norm": 0.3481849431991577, "learning_rate": 8.742479940187026e-05, "loss": 0.0355, "step": 21540 }, { "grad_norm": 0.3922131061553955, "learning_rate": 8.74110901942202e-05, "loss": 0.035, "step": 21550 }, { "grad_norm": 0.43094009160995483, "learning_rate": 8.739737459403803e-05, "loss": 0.0365, "step": 21560 }, { "grad_norm": 0.3756909966468811, "learning_rate": 8.738365260366737e-05, "loss": 0.0306, "step": 21570 }, { "grad_norm": 0.400284081697464, "learning_rate": 8.736992422545292e-05, "loss": 0.0276, "step": 21580 }, { "grad_norm": 0.36710700392723083, "learning_rate": 8.73561894617405e-05, "loss": 0.0348, "step": 21590 }, { "grad_norm": 0.40725213289260864, "learning_rate": 8.734244831487697e-05, "loss": 0.0288, "step": 21600 }, { "grad_norm": 0.38125377893447876, "learning_rate": 8.732870078721035e-05, "loss": 0.0462, "step": 21610 }, { "grad_norm": 0.3995232880115509, "learning_rate": 8.731494688108972e-05, "loss": 0.0303, "step": 21620 }, { "grad_norm": 0.36132243275642395, "learning_rate": 8.730118659886523e-05, "loss": 0.0361, "step": 21630 }, { "grad_norm": 0.3669811487197876, "learning_rate": 8.728741994288814e-05, "loss": 0.0355, "step": 21640 }, { "grad_norm": 0.4130707383155823, "learning_rate": 8.727364691551079e-05, "loss": 0.0329, "step": 21650 }, { "grad_norm": 0.4461206793785095, "learning_rate": 8.725986751908661e-05, "loss": 0.0337, "step": 21660 }, { "grad_norm": 0.4247026741504669, "learning_rate": 8.724608175597016e-05, "loss": 0.0341, "step": 21670 }, { "grad_norm": 0.30274227261543274, "learning_rate": 8.723228962851699e-05, "loss": 0.0259, "step": 21680 }, { "grad_norm": 0.3713722825050354, "learning_rate": 8.721849113908385e-05, "loss": 0.0305, "step": 21690 }, { "grad_norm": 0.3060263693332672, "learning_rate": 8.720468629002848e-05, "loss": 0.0354, "step": 21700 }, { "grad_norm": 0.28209948539733887, "learning_rate": 8.719087508370978e-05, "loss": 0.0243, "step": 21710 }, { "grad_norm": 0.5352857112884521, "learning_rate": 8.717705752248772e-05, "loss": 0.0324, "step": 21720 }, { "grad_norm": 0.6935817003250122, "learning_rate": 8.71632336087233e-05, "loss": 0.025, "step": 21730 }, { "grad_norm": 0.3777986764907837, "learning_rate": 8.71494033447787e-05, "loss": 0.031, "step": 21740 }, { "grad_norm": 0.3143879473209381, "learning_rate": 8.713556673301708e-05, "loss": 0.0299, "step": 21750 }, { "grad_norm": 0.5443257689476013, "learning_rate": 8.712172377580278e-05, "loss": 0.0489, "step": 21760 }, { "grad_norm": 0.3624301850795746, "learning_rate": 8.710787447550114e-05, "loss": 0.0332, "step": 21770 }, { "grad_norm": 0.26231899857521057, "learning_rate": 8.70940188344787e-05, "loss": 0.0441, "step": 21780 }, { "grad_norm": 0.3337545394897461, "learning_rate": 8.708015685510293e-05, "loss": 0.0345, "step": 21790 }, { "grad_norm": 0.36174118518829346, "learning_rate": 8.706628853974252e-05, "loss": 0.0396, "step": 21800 }, { "grad_norm": 0.1762162446975708, "learning_rate": 8.705241389076715e-05, "loss": 0.0255, "step": 21810 }, { "grad_norm": 0.3347809314727783, "learning_rate": 8.703853291054764e-05, "loss": 0.0314, "step": 21820 }, { "grad_norm": 0.43665072321891785, "learning_rate": 8.702464560145587e-05, "loss": 0.0336, "step": 21830 }, { "grad_norm": 0.45728543400764465, "learning_rate": 8.701075196586476e-05, "loss": 0.0276, "step": 21840 }, { "grad_norm": 0.2899768650531769, "learning_rate": 8.699685200614842e-05, "loss": 0.0294, "step": 21850 }, { "grad_norm": 0.44915544986724854, "learning_rate": 8.698294572468193e-05, "loss": 0.0345, "step": 21860 }, { "grad_norm": 0.27112722396850586, "learning_rate": 8.696903312384148e-05, "loss": 0.0335, "step": 21870 }, { "grad_norm": 0.33484816551208496, "learning_rate": 8.695511420600439e-05, "loss": 0.0239, "step": 21880 }, { "grad_norm": 0.48934808373451233, "learning_rate": 8.694118897354901e-05, "loss": 0.0277, "step": 21890 }, { "grad_norm": 0.24045148491859436, "learning_rate": 8.692725742885478e-05, "loss": 0.0269, "step": 21900 }, { "grad_norm": 0.47010064125061035, "learning_rate": 8.691331957430221e-05, "loss": 0.033, "step": 21910 }, { "grad_norm": 0.343341201543808, "learning_rate": 8.68993754122729e-05, "loss": 0.0305, "step": 21920 }, { "grad_norm": 0.2839774787425995, "learning_rate": 8.688542494514955e-05, "loss": 0.028, "step": 21930 }, { "grad_norm": 0.47470957040786743, "learning_rate": 8.68714681753159e-05, "loss": 0.0294, "step": 21940 }, { "grad_norm": 0.3476281464099884, "learning_rate": 8.685750510515676e-05, "loss": 0.0307, "step": 21950 }, { "grad_norm": 0.397971510887146, "learning_rate": 8.684353573705805e-05, "loss": 0.0355, "step": 21960 }, { "grad_norm": 0.39571085572242737, "learning_rate": 8.682956007340677e-05, "loss": 0.0394, "step": 21970 }, { "grad_norm": 0.40816837549209595, "learning_rate": 8.681557811659095e-05, "loss": 0.0308, "step": 21980 }, { "grad_norm": 0.40046846866607666, "learning_rate": 8.680158986899974e-05, "loss": 0.0272, "step": 21990 }, { "grad_norm": 0.30762621760368347, "learning_rate": 8.678759533302335e-05, "loss": 0.0297, "step": 22000 }, { "grad_norm": 0.3206136226654053, "learning_rate": 8.677359451105308e-05, "loss": 0.0307, "step": 22010 }, { "grad_norm": 0.3928978741168976, "learning_rate": 8.675958740548123e-05, "loss": 0.0416, "step": 22020 }, { "grad_norm": 0.3315986692905426, "learning_rate": 8.674557401870129e-05, "loss": 0.0383, "step": 22030 }, { "grad_norm": 0.3297584056854248, "learning_rate": 8.673155435310775e-05, "loss": 0.0331, "step": 22040 }, { "grad_norm": 0.42710644006729126, "learning_rate": 8.671752841109617e-05, "loss": 0.0321, "step": 22050 }, { "grad_norm": 0.5019413232803345, "learning_rate": 8.670349619506321e-05, "loss": 0.0369, "step": 22060 }, { "grad_norm": 0.33001986145973206, "learning_rate": 8.66894577074066e-05, "loss": 0.0306, "step": 22070 }, { "grad_norm": 0.33655473589897156, "learning_rate": 8.667541295052513e-05, "loss": 0.0309, "step": 22080 }, { "grad_norm": 0.44857144355773926, "learning_rate": 8.666136192681865e-05, "loss": 0.0382, "step": 22090 }, { "grad_norm": 0.395040899515152, "learning_rate": 8.664730463868811e-05, "loss": 0.03, "step": 22100 }, { "grad_norm": 0.35425105690956116, "learning_rate": 8.663324108853552e-05, "loss": 0.04, "step": 22110 }, { "grad_norm": 0.29908832907676697, "learning_rate": 8.661917127876395e-05, "loss": 0.0426, "step": 22120 }, { "grad_norm": 0.4201194941997528, "learning_rate": 8.660509521177754e-05, "loss": 0.0412, "step": 22130 }, { "grad_norm": 0.3143192231655121, "learning_rate": 8.65910128899815e-05, "loss": 0.0336, "step": 22140 }, { "grad_norm": 0.31390321254730225, "learning_rate": 8.657692431578214e-05, "loss": 0.0267, "step": 22150 }, { "grad_norm": 0.27084362506866455, "learning_rate": 8.656282949158679e-05, "loss": 0.0237, "step": 22160 }, { "grad_norm": 0.3436318635940552, "learning_rate": 8.654872841980388e-05, "loss": 0.0257, "step": 22170 }, { "grad_norm": 0.42357832193374634, "learning_rate": 8.653462110284289e-05, "loss": 0.0269, "step": 22180 }, { "grad_norm": 0.3146668076515198, "learning_rate": 8.652050754311437e-05, "loss": 0.0336, "step": 22190 }, { "grad_norm": 0.3261256217956543, "learning_rate": 8.650638774302995e-05, "loss": 0.033, "step": 22200 }, { "grad_norm": 0.4396288990974426, "learning_rate": 8.649226170500233e-05, "loss": 0.0389, "step": 22210 }, { "grad_norm": 0.36381617188453674, "learning_rate": 8.647812943144524e-05, "loss": 0.0338, "step": 22220 }, { "grad_norm": 0.3318311870098114, "learning_rate": 8.646399092477351e-05, "loss": 0.0293, "step": 22230 }, { "grad_norm": 0.4383511543273926, "learning_rate": 8.644984618740301e-05, "loss": 0.0396, "step": 22240 }, { "grad_norm": 0.47127634286880493, "learning_rate": 8.643569522175073e-05, "loss": 0.0297, "step": 22250 }, { "grad_norm": 0.4549095928668976, "learning_rate": 8.642153803023463e-05, "loss": 0.0342, "step": 22260 }, { "grad_norm": 0.39338988065719604, "learning_rate": 8.640737461527383e-05, "loss": 0.0311, "step": 22270 }, { "grad_norm": 0.333760529756546, "learning_rate": 8.639320497928845e-05, "loss": 0.0358, "step": 22280 }, { "grad_norm": 0.5102898478507996, "learning_rate": 8.637902912469969e-05, "loss": 0.0309, "step": 22290 }, { "grad_norm": 0.45982691645622253, "learning_rate": 8.636484705392982e-05, "loss": 0.0364, "step": 22300 }, { "grad_norm": 0.26761558651924133, "learning_rate": 8.635065876940216e-05, "loss": 0.0289, "step": 22310 }, { "grad_norm": 0.3974916338920593, "learning_rate": 8.633646427354112e-05, "loss": 0.0368, "step": 22320 }, { "grad_norm": 0.31269505620002747, "learning_rate": 8.632226356877213e-05, "loss": 0.0233, "step": 22330 }, { "grad_norm": 0.23507681488990784, "learning_rate": 8.630805665752173e-05, "loss": 0.0286, "step": 22340 }, { "grad_norm": 0.2751312255859375, "learning_rate": 8.629384354221748e-05, "loss": 0.0279, "step": 22350 }, { "grad_norm": 0.2745290994644165, "learning_rate": 8.627962422528797e-05, "loss": 0.0269, "step": 22360 }, { "grad_norm": 0.43859466910362244, "learning_rate": 8.626539870916296e-05, "loss": 0.0389, "step": 22370 }, { "grad_norm": 0.36767271161079407, "learning_rate": 8.625116699627317e-05, "loss": 0.0346, "step": 22380 }, { "grad_norm": 0.31293365359306335, "learning_rate": 8.623692908905041e-05, "loss": 0.0274, "step": 22390 }, { "grad_norm": 0.4555736780166626, "learning_rate": 8.622268498992755e-05, "loss": 0.0353, "step": 22400 }, { "grad_norm": 0.34629419445991516, "learning_rate": 8.620843470133851e-05, "loss": 0.0428, "step": 22410 }, { "grad_norm": 0.42535164952278137, "learning_rate": 8.619417822571829e-05, "loss": 0.032, "step": 22420 }, { "grad_norm": 0.3930799961090088, "learning_rate": 8.617991556550292e-05, "loss": 0.033, "step": 22430 }, { "grad_norm": 0.2768489718437195, "learning_rate": 8.616564672312952e-05, "loss": 0.0277, "step": 22440 }, { "grad_norm": 0.3926774263381958, "learning_rate": 8.61513717010362e-05, "loss": 0.038, "step": 22450 }, { "grad_norm": 0.381124883890152, "learning_rate": 8.613709050166221e-05, "loss": 0.0391, "step": 22460 }, { "grad_norm": 0.29712966084480286, "learning_rate": 8.61228031274478e-05, "loss": 0.0363, "step": 22470 }, { "grad_norm": 0.5809943079948425, "learning_rate": 8.610850958083431e-05, "loss": 0.0362, "step": 22480 }, { "grad_norm": 0.3434527814388275, "learning_rate": 8.609420986426409e-05, "loss": 0.0282, "step": 22490 }, { "grad_norm": 0.6145493984222412, "learning_rate": 8.60799039801806e-05, "loss": 0.033, "step": 22500 }, { "grad_norm": 0.47761133313179016, "learning_rate": 8.606559193102828e-05, "loss": 0.0282, "step": 22510 }, { "grad_norm": 0.2870239019393921, "learning_rate": 8.605127371925273e-05, "loss": 0.0381, "step": 22520 }, { "grad_norm": 0.5446380972862244, "learning_rate": 8.603694934730047e-05, "loss": 0.0425, "step": 22530 }, { "grad_norm": 0.2740435302257538, "learning_rate": 8.602261881761919e-05, "loss": 0.0243, "step": 22540 }, { "grad_norm": 0.26307594776153564, "learning_rate": 8.600828213265759e-05, "loss": 0.0429, "step": 22550 }, { "grad_norm": 0.4659283459186554, "learning_rate": 8.599393929486539e-05, "loss": 0.0439, "step": 22560 }, { "grad_norm": 0.2605478763580322, "learning_rate": 8.59795903066934e-05, "loss": 0.0247, "step": 22570 }, { "grad_norm": 0.2671355903148651, "learning_rate": 8.596523517059347e-05, "loss": 0.0295, "step": 22580 }, { "grad_norm": 0.4642617106437683, "learning_rate": 8.59508738890185e-05, "loss": 0.0253, "step": 22590 }, { "grad_norm": 0.3257347047328949, "learning_rate": 8.593650646442246e-05, "loss": 0.024, "step": 22600 }, { "grad_norm": 0.5443268418312073, "learning_rate": 8.59221328992603e-05, "loss": 0.0287, "step": 22610 }, { "grad_norm": 0.4576787054538727, "learning_rate": 8.590775319598813e-05, "loss": 0.0289, "step": 22620 }, { "grad_norm": 0.45977067947387695, "learning_rate": 8.589336735706301e-05, "loss": 0.0323, "step": 22630 }, { "grad_norm": 0.3472486734390259, "learning_rate": 8.587897538494307e-05, "loss": 0.0495, "step": 22640 }, { "grad_norm": 0.4104423224925995, "learning_rate": 8.586457728208756e-05, "loss": 0.031, "step": 22650 }, { "grad_norm": 0.3822866380214691, "learning_rate": 8.585017305095667e-05, "loss": 0.0365, "step": 22660 }, { "grad_norm": 0.3687185049057007, "learning_rate": 8.583576269401173e-05, "loss": 0.0305, "step": 22670 }, { "grad_norm": 0.3719487190246582, "learning_rate": 8.582134621371504e-05, "loss": 0.0319, "step": 22680 }, { "grad_norm": 0.22702014446258545, "learning_rate": 8.580692361253e-05, "loss": 0.0281, "step": 22690 }, { "grad_norm": 0.41407230496406555, "learning_rate": 8.579249489292104e-05, "loss": 0.0279, "step": 22700 }, { "grad_norm": 0.38349494338035583, "learning_rate": 8.577806005735363e-05, "loss": 0.0403, "step": 22710 }, { "grad_norm": 0.5531105399131775, "learning_rate": 8.576361910829429e-05, "loss": 0.0327, "step": 22720 }, { "grad_norm": 0.285665363073349, "learning_rate": 8.574917204821057e-05, "loss": 0.0297, "step": 22730 }, { "grad_norm": 0.3047773241996765, "learning_rate": 8.57347188795711e-05, "loss": 0.0338, "step": 22740 }, { "grad_norm": 0.39725160598754883, "learning_rate": 8.572025960484551e-05, "loss": 0.0373, "step": 22750 }, { "grad_norm": 0.34836557507514954, "learning_rate": 8.57057942265045e-05, "loss": 0.0383, "step": 22760 }, { "grad_norm": 0.2512058913707733, "learning_rate": 8.569132274701984e-05, "loss": 0.0365, "step": 22770 }, { "grad_norm": 0.4460318982601166, "learning_rate": 8.567684516886427e-05, "loss": 0.0312, "step": 22780 }, { "grad_norm": 0.33001917600631714, "learning_rate": 8.56623614945116e-05, "loss": 0.0293, "step": 22790 }, { "grad_norm": 0.36450472474098206, "learning_rate": 8.564787172643675e-05, "loss": 0.0269, "step": 22800 }, { "grad_norm": 0.36648958921432495, "learning_rate": 8.563337586711559e-05, "loss": 0.0359, "step": 22810 }, { "grad_norm": 0.5018454194068909, "learning_rate": 8.561887391902506e-05, "loss": 0.0405, "step": 22820 }, { "grad_norm": 0.3015282154083252, "learning_rate": 8.560436588464316e-05, "loss": 0.0279, "step": 22830 }, { "grad_norm": 0.3292903006076813, "learning_rate": 8.55898517664489e-05, "loss": 0.0364, "step": 22840 }, { "grad_norm": 0.5130276083946228, "learning_rate": 8.557533156692236e-05, "loss": 0.0445, "step": 22850 }, { "grad_norm": 0.36936211585998535, "learning_rate": 8.556080528854467e-05, "loss": 0.0285, "step": 22860 }, { "grad_norm": 0.3346671462059021, "learning_rate": 8.554627293379791e-05, "loss": 0.0315, "step": 22870 }, { "grad_norm": 0.4919247627258301, "learning_rate": 8.553173450516531e-05, "loss": 0.041, "step": 22880 }, { "grad_norm": 0.2952791750431061, "learning_rate": 8.551719000513108e-05, "loss": 0.0366, "step": 22890 }, { "grad_norm": 0.42954379320144653, "learning_rate": 8.550263943618049e-05, "loss": 0.0516, "step": 22900 }, { "grad_norm": 0.40303340554237366, "learning_rate": 8.54880828007998e-05, "loss": 0.0374, "step": 22910 }, { "grad_norm": 0.5682956576347351, "learning_rate": 8.547352010147637e-05, "loss": 0.0394, "step": 22920 }, { "grad_norm": 0.6060301065444946, "learning_rate": 8.545895134069855e-05, "loss": 0.032, "step": 22930 }, { "grad_norm": 0.4229646325111389, "learning_rate": 8.544437652095576e-05, "loss": 0.0301, "step": 22940 }, { "grad_norm": 0.45306289196014404, "learning_rate": 8.542979564473843e-05, "loss": 0.0373, "step": 22950 }, { "grad_norm": 0.41990941762924194, "learning_rate": 8.541520871453802e-05, "loss": 0.031, "step": 22960 }, { "grad_norm": 0.27428823709487915, "learning_rate": 8.540061573284705e-05, "loss": 0.027, "step": 22970 }, { "grad_norm": 0.369745671749115, "learning_rate": 8.538601670215906e-05, "loss": 0.0285, "step": 22980 }, { "grad_norm": 0.6055394411087036, "learning_rate": 8.537141162496864e-05, "loss": 0.0362, "step": 22990 }, { "grad_norm": 0.5691915154457092, "learning_rate": 8.535680050377137e-05, "loss": 0.0323, "step": 23000 }, { "grad_norm": 0.4956684410572052, "learning_rate": 8.534218334106391e-05, "loss": 0.0276, "step": 23010 }, { "grad_norm": 0.3099422752857208, "learning_rate": 8.532756013934393e-05, "loss": 0.0283, "step": 23020 }, { "grad_norm": 0.5186004042625427, "learning_rate": 8.531293090111012e-05, "loss": 0.0325, "step": 23030 }, { "grad_norm": 0.3049929141998291, "learning_rate": 8.529829562886225e-05, "loss": 0.0312, "step": 23040 }, { "grad_norm": 0.5229509472846985, "learning_rate": 8.528365432510105e-05, "loss": 0.0362, "step": 23050 }, { "grad_norm": 0.35903701186180115, "learning_rate": 8.526900699232833e-05, "loss": 0.0302, "step": 23060 }, { "grad_norm": 0.38921740651130676, "learning_rate": 8.525435363304695e-05, "loss": 0.0308, "step": 23070 }, { "grad_norm": 0.32862845063209534, "learning_rate": 8.523969424976072e-05, "loss": 0.0258, "step": 23080 }, { "grad_norm": 0.40226438641548157, "learning_rate": 8.522502884497457e-05, "loss": 0.0303, "step": 23090 }, { "grad_norm": 0.4022340774536133, "learning_rate": 8.521035742119437e-05, "loss": 0.0383, "step": 23100 }, { "grad_norm": 0.3315553367137909, "learning_rate": 8.519567998092712e-05, "loss": 0.0367, "step": 23110 }, { "grad_norm": 0.49722543358802795, "learning_rate": 8.518099652668075e-05, "loss": 0.0374, "step": 23120 }, { "grad_norm": 0.26318684220314026, "learning_rate": 8.516630706096429e-05, "loss": 0.0308, "step": 23130 }, { "grad_norm": 0.3620198965072632, "learning_rate": 8.515161158628773e-05, "loss": 0.0305, "step": 23140 }, { "grad_norm": 0.3095591068267822, "learning_rate": 8.513691010516216e-05, "loss": 0.0508, "step": 23150 }, { "grad_norm": 0.44371792674064636, "learning_rate": 8.512220262009966e-05, "loss": 0.036, "step": 23160 }, { "grad_norm": 0.45718297362327576, "learning_rate": 8.510748913361332e-05, "loss": 0.0332, "step": 23170 }, { "grad_norm": 0.40918123722076416, "learning_rate": 8.509276964821726e-05, "loss": 0.0233, "step": 23180 }, { "grad_norm": 0.4399576485157013, "learning_rate": 8.507804416642669e-05, "loss": 0.0337, "step": 23190 }, { "grad_norm": 0.6793482303619385, "learning_rate": 8.506331269075774e-05, "loss": 0.0368, "step": 23200 }, { "grad_norm": 0.3073870837688446, "learning_rate": 8.504857522372765e-05, "loss": 0.0306, "step": 23210 }, { "grad_norm": 0.36887168884277344, "learning_rate": 8.503383176785461e-05, "loss": 0.0292, "step": 23220 }, { "grad_norm": 0.38930854201316833, "learning_rate": 8.501908232565792e-05, "loss": 0.0289, "step": 23230 }, { "grad_norm": 0.31724196672439575, "learning_rate": 8.50043268996578e-05, "loss": 0.0305, "step": 23240 }, { "grad_norm": 0.31323912739753723, "learning_rate": 8.498956549237562e-05, "loss": 0.0329, "step": 23250 }, { "grad_norm": 0.41451600193977356, "learning_rate": 8.497479810633366e-05, "loss": 0.0361, "step": 23260 }, { "grad_norm": 0.5021040439605713, "learning_rate": 8.496002474405525e-05, "loss": 0.0368, "step": 23270 }, { "grad_norm": 0.23669101297855377, "learning_rate": 8.494524540806478e-05, "loss": 0.0361, "step": 23280 }, { "grad_norm": 0.3837323486804962, "learning_rate": 8.493046010088761e-05, "loss": 0.0427, "step": 23290 }, { "grad_norm": 0.39323484897613525, "learning_rate": 8.491566882505018e-05, "loss": 0.0364, "step": 23300 }, { "grad_norm": 0.39683040976524353, "learning_rate": 8.490087158307988e-05, "loss": 0.0404, "step": 23310 }, { "grad_norm": 0.4744022786617279, "learning_rate": 8.488606837750518e-05, "loss": 0.0321, "step": 23320 }, { "grad_norm": 0.25883492827415466, "learning_rate": 8.487125921085552e-05, "loss": 0.0308, "step": 23330 }, { "grad_norm": 0.37342122197151184, "learning_rate": 8.485644408566141e-05, "loss": 0.0332, "step": 23340 }, { "grad_norm": 0.46355196833610535, "learning_rate": 8.484162300445431e-05, "loss": 0.0356, "step": 23350 }, { "grad_norm": 0.35439351201057434, "learning_rate": 8.482679596976676e-05, "loss": 0.031, "step": 23360 }, { "grad_norm": 0.36990782618522644, "learning_rate": 8.48119629841323e-05, "loss": 0.0383, "step": 23370 }, { "grad_norm": 0.476969838142395, "learning_rate": 8.479712405008547e-05, "loss": 0.0363, "step": 23380 }, { "grad_norm": 0.3649286925792694, "learning_rate": 8.478227917016184e-05, "loss": 0.0416, "step": 23390 }, { "grad_norm": 0.2679590582847595, "learning_rate": 8.476742834689801e-05, "loss": 0.0361, "step": 23400 }, { "grad_norm": 0.5296474695205688, "learning_rate": 8.475257158283157e-05, "loss": 0.0481, "step": 23410 }, { "grad_norm": 0.3553653657436371, "learning_rate": 8.473770888050112e-05, "loss": 0.0258, "step": 23420 }, { "grad_norm": 0.47147130966186523, "learning_rate": 8.47228402424463e-05, "loss": 0.032, "step": 23430 }, { "grad_norm": 0.474077045917511, "learning_rate": 8.470796567120775e-05, "loss": 0.0306, "step": 23440 }, { "grad_norm": 0.5334368944168091, "learning_rate": 8.469308516932714e-05, "loss": 0.0383, "step": 23450 }, { "grad_norm": 0.36221081018447876, "learning_rate": 8.467819873934714e-05, "loss": 0.0313, "step": 23460 }, { "grad_norm": 0.35309717059135437, "learning_rate": 8.466330638381143e-05, "loss": 0.031, "step": 23470 }, { "grad_norm": 0.3237763047218323, "learning_rate": 8.464840810526469e-05, "loss": 0.0273, "step": 23480 }, { "grad_norm": 0.3714803457260132, "learning_rate": 8.463350390625264e-05, "loss": 0.028, "step": 23490 }, { "grad_norm": 0.32796335220336914, "learning_rate": 8.4618593789322e-05, "loss": 0.0435, "step": 23500 }, { "grad_norm": 0.5261627435684204, "learning_rate": 8.46036777570205e-05, "loss": 0.0344, "step": 23510 }, { "grad_norm": 0.4333586096763611, "learning_rate": 8.458875581189688e-05, "loss": 0.0575, "step": 23520 }, { "grad_norm": 0.33665600419044495, "learning_rate": 8.457382795650092e-05, "loss": 0.0301, "step": 23530 }, { "grad_norm": 0.302875280380249, "learning_rate": 8.455889419338335e-05, "loss": 0.033, "step": 23540 }, { "grad_norm": 0.32762208580970764, "learning_rate": 8.454395452509593e-05, "loss": 0.0281, "step": 23550 }, { "grad_norm": 0.5531992316246033, "learning_rate": 8.452900895419146e-05, "loss": 0.0375, "step": 23560 }, { "grad_norm": 0.2483566850423813, "learning_rate": 8.451405748322376e-05, "loss": 0.0295, "step": 23570 }, { "grad_norm": 0.3870358467102051, "learning_rate": 8.449910011474759e-05, "loss": 0.031, "step": 23580 }, { "grad_norm": 0.421017050743103, "learning_rate": 8.448413685131876e-05, "loss": 0.0414, "step": 23590 }, { "grad_norm": 0.25083422660827637, "learning_rate": 8.446916769549407e-05, "loss": 0.0468, "step": 23600 }, { "grad_norm": 0.4203372895717621, "learning_rate": 8.445419264983136e-05, "loss": 0.0409, "step": 23610 }, { "grad_norm": 0.341564804315567, "learning_rate": 8.443921171688947e-05, "loss": 0.0308, "step": 23620 }, { "grad_norm": 0.39662009477615356, "learning_rate": 8.442422489922819e-05, "loss": 0.0319, "step": 23630 }, { "grad_norm": 0.3806110918521881, "learning_rate": 8.440923219940838e-05, "loss": 0.0393, "step": 23640 }, { "grad_norm": 0.4814326763153076, "learning_rate": 8.439423361999189e-05, "loss": 0.0363, "step": 23650 }, { "grad_norm": 0.40708744525909424, "learning_rate": 8.437922916354155e-05, "loss": 0.033, "step": 23660 }, { "grad_norm": 0.3640247583389282, "learning_rate": 8.436421883262123e-05, "loss": 0.0294, "step": 23670 }, { "grad_norm": 0.24984829127788544, "learning_rate": 8.434920262979577e-05, "loss": 0.0383, "step": 23680 }, { "grad_norm": 0.36835771799087524, "learning_rate": 8.433418055763104e-05, "loss": 0.024, "step": 23690 }, { "grad_norm": 0.5859907865524292, "learning_rate": 8.431915261869389e-05, "loss": 0.0312, "step": 23700 }, { "grad_norm": 0.24754363298416138, "learning_rate": 8.43041188155522e-05, "loss": 0.0274, "step": 23710 }, { "grad_norm": 0.3060457408428192, "learning_rate": 8.428907915077481e-05, "loss": 0.0311, "step": 23720 }, { "grad_norm": 0.5202796459197998, "learning_rate": 8.42740336269316e-05, "loss": 0.0455, "step": 23730 }, { "grad_norm": 0.45162826776504517, "learning_rate": 8.425898224659345e-05, "loss": 0.0299, "step": 23740 }, { "grad_norm": 0.40723246335983276, "learning_rate": 8.42439250123322e-05, "loss": 0.03, "step": 23750 }, { "grad_norm": 0.4137848913669586, "learning_rate": 8.422886192672076e-05, "loss": 0.033, "step": 23760 }, { "grad_norm": 0.33381199836730957, "learning_rate": 8.421379299233297e-05, "loss": 0.0277, "step": 23770 }, { "grad_norm": 0.36096519231796265, "learning_rate": 8.419871821174371e-05, "loss": 0.029, "step": 23780 }, { "grad_norm": 0.2906184494495392, "learning_rate": 8.418363758752884e-05, "loss": 0.029, "step": 23790 }, { "grad_norm": 0.24325592815876007, "learning_rate": 8.416855112226523e-05, "loss": 0.0305, "step": 23800 }, { "grad_norm": 0.34762099385261536, "learning_rate": 8.415345881853075e-05, "loss": 0.0291, "step": 23810 }, { "grad_norm": 0.33719414472579956, "learning_rate": 8.413836067890426e-05, "loss": 0.0262, "step": 23820 }, { "grad_norm": 0.42343831062316895, "learning_rate": 8.41232567059656e-05, "loss": 0.0309, "step": 23830 }, { "grad_norm": 0.44601279497146606, "learning_rate": 8.410814690229565e-05, "loss": 0.0328, "step": 23840 }, { "grad_norm": 0.2443782091140747, "learning_rate": 8.409303127047626e-05, "loss": 0.0237, "step": 23850 }, { "grad_norm": 0.286268025636673, "learning_rate": 8.407790981309028e-05, "loss": 0.0262, "step": 23860 }, { "grad_norm": 0.3970860242843628, "learning_rate": 8.406278253272153e-05, "loss": 0.0353, "step": 23870 }, { "grad_norm": 0.27390772104263306, "learning_rate": 8.404764943195487e-05, "loss": 0.0278, "step": 23880 }, { "grad_norm": 0.4160352945327759, "learning_rate": 8.403251051337613e-05, "loss": 0.0323, "step": 23890 }, { "grad_norm": 0.5105619430541992, "learning_rate": 8.401736577957214e-05, "loss": 0.0306, "step": 23900 }, { "grad_norm": 0.2791496515274048, "learning_rate": 8.40022152331307e-05, "loss": 0.027, "step": 23910 }, { "grad_norm": 0.31012091040611267, "learning_rate": 8.398705887664064e-05, "loss": 0.0347, "step": 23920 }, { "grad_norm": 0.4730610251426697, "learning_rate": 8.397189671269177e-05, "loss": 0.0341, "step": 23930 }, { "grad_norm": 0.37502968311309814, "learning_rate": 8.395672874387488e-05, "loss": 0.0296, "step": 23940 }, { "grad_norm": 0.3446783721446991, "learning_rate": 8.394155497278177e-05, "loss": 0.0303, "step": 23950 }, { "grad_norm": 0.4283570647239685, "learning_rate": 8.392637540200523e-05, "loss": 0.0457, "step": 23960 }, { "grad_norm": 0.44970646500587463, "learning_rate": 8.391119003413902e-05, "loss": 0.0337, "step": 23970 }, { "grad_norm": 0.38708147406578064, "learning_rate": 8.38959988717779e-05, "loss": 0.0314, "step": 23980 }, { "grad_norm": 0.34642088413238525, "learning_rate": 8.388080191751764e-05, "loss": 0.041, "step": 23990 }, { "grad_norm": 0.4447937309741974, "learning_rate": 8.386559917395496e-05, "loss": 0.0284, "step": 24000 }, { "grad_norm": 0.23183701932430267, "learning_rate": 8.385039064368761e-05, "loss": 0.0237, "step": 24010 }, { "grad_norm": 0.27855634689331055, "learning_rate": 8.383517632931431e-05, "loss": 0.0295, "step": 24020 }, { "grad_norm": 0.36981841921806335, "learning_rate": 8.381995623343477e-05, "loss": 0.0298, "step": 24030 }, { "grad_norm": 0.34939271211624146, "learning_rate": 8.380473035864968e-05, "loss": 0.0334, "step": 24040 }, { "grad_norm": 0.26145729422569275, "learning_rate": 8.378949870756076e-05, "loss": 0.028, "step": 24050 }, { "grad_norm": 0.21586422622203827, "learning_rate": 8.377426128277063e-05, "loss": 0.0359, "step": 24060 }, { "grad_norm": 0.3740568161010742, "learning_rate": 8.375901808688298e-05, "loss": 0.0269, "step": 24070 }, { "grad_norm": 0.37285324931144714, "learning_rate": 8.374376912250246e-05, "loss": 0.0334, "step": 24080 }, { "grad_norm": 0.3249727487564087, "learning_rate": 8.372851439223468e-05, "loss": 0.0253, "step": 24090 }, { "grad_norm": 0.42159220576286316, "learning_rate": 8.371325389868627e-05, "loss": 0.0329, "step": 24100 }, { "grad_norm": 0.390227735042572, "learning_rate": 8.369798764446482e-05, "loss": 0.0283, "step": 24110 }, { "grad_norm": 0.43533989787101746, "learning_rate": 8.368271563217893e-05, "loss": 0.0296, "step": 24120 }, { "grad_norm": 0.3387891948223114, "learning_rate": 8.366743786443817e-05, "loss": 0.0434, "step": 24130 }, { "grad_norm": 0.2844676375389099, "learning_rate": 8.365215434385309e-05, "loss": 0.041, "step": 24140 }, { "grad_norm": 0.43533089756965637, "learning_rate": 8.36368650730352e-05, "loss": 0.0364, "step": 24150 }, { "grad_norm": 0.22760836780071259, "learning_rate": 8.362157005459705e-05, "loss": 0.0283, "step": 24160 }, { "grad_norm": 0.4033108949661255, "learning_rate": 8.360626929115213e-05, "loss": 0.0314, "step": 24170 }, { "grad_norm": 0.4248490035533905, "learning_rate": 8.359096278531492e-05, "loss": 0.0301, "step": 24180 }, { "grad_norm": 0.34371834993362427, "learning_rate": 8.357565053970088e-05, "loss": 0.0277, "step": 24190 }, { "grad_norm": 0.21552076935768127, "learning_rate": 8.356033255692647e-05, "loss": 0.0252, "step": 24200 }, { "grad_norm": 0.364006370306015, "learning_rate": 8.354500883960911e-05, "loss": 0.0264, "step": 24210 }, { "grad_norm": 0.2370893359184265, "learning_rate": 8.352967939036717e-05, "loss": 0.0299, "step": 24220 }, { "grad_norm": 0.2555486261844635, "learning_rate": 8.35143442118201e-05, "loss": 0.0349, "step": 24230 }, { "grad_norm": 0.22620433568954468, "learning_rate": 8.349900330658819e-05, "loss": 0.0261, "step": 24240 }, { "grad_norm": 0.39537057280540466, "learning_rate": 8.348365667729284e-05, "loss": 0.0268, "step": 24250 }, { "grad_norm": 0.3292252719402313, "learning_rate": 8.346830432655633e-05, "loss": 0.0239, "step": 24260 }, { "grad_norm": 0.40557631850242615, "learning_rate": 8.345294625700195e-05, "loss": 0.0251, "step": 24270 }, { "grad_norm": 0.36457449197769165, "learning_rate": 8.343758247125402e-05, "loss": 0.0295, "step": 24280 }, { "grad_norm": 0.38099581003189087, "learning_rate": 8.342221297193776e-05, "loss": 0.0343, "step": 24290 }, { "grad_norm": 0.47265174984931946, "learning_rate": 8.34068377616794e-05, "loss": 0.0406, "step": 24300 }, { "grad_norm": 0.597566545009613, "learning_rate": 8.339145684310615e-05, "loss": 0.0328, "step": 24310 }, { "grad_norm": 0.23869073390960693, "learning_rate": 8.337607021884618e-05, "loss": 0.0263, "step": 24320 }, { "grad_norm": 0.3620463013648987, "learning_rate": 8.336067789152867e-05, "loss": 0.0254, "step": 24330 }, { "grad_norm": 0.43778273463249207, "learning_rate": 8.334527986378369e-05, "loss": 0.0359, "step": 24340 }, { "grad_norm": 0.29158103466033936, "learning_rate": 8.332987613824239e-05, "loss": 0.0341, "step": 24350 }, { "grad_norm": 0.24407780170440674, "learning_rate": 8.331446671753685e-05, "loss": 0.0283, "step": 24360 }, { "grad_norm": 0.5034101605415344, "learning_rate": 8.329905160430007e-05, "loss": 0.0307, "step": 24370 }, { "grad_norm": 0.3833259046077728, "learning_rate": 8.328363080116611e-05, "loss": 0.0272, "step": 24380 }, { "grad_norm": 0.2732089161872864, "learning_rate": 8.326820431076997e-05, "loss": 0.0278, "step": 24390 }, { "grad_norm": 0.372581422328949, "learning_rate": 8.325277213574759e-05, "loss": 0.0367, "step": 24400 }, { "grad_norm": 0.31024691462516785, "learning_rate": 8.32373342787359e-05, "loss": 0.0292, "step": 24410 }, { "grad_norm": 0.38843274116516113, "learning_rate": 8.322189074237285e-05, "loss": 0.0337, "step": 24420 }, { "grad_norm": 0.40024200081825256, "learning_rate": 8.32064415292973e-05, "loss": 0.0355, "step": 24430 }, { "grad_norm": 0.27076220512390137, "learning_rate": 8.319098664214907e-05, "loss": 0.0295, "step": 24440 }, { "grad_norm": 0.3674837648868561, "learning_rate": 8.3175526083569e-05, "loss": 0.0336, "step": 24450 }, { "grad_norm": 0.3743603229522705, "learning_rate": 8.316005985619889e-05, "loss": 0.0376, "step": 24460 }, { "grad_norm": 0.3602047562599182, "learning_rate": 8.314458796268147e-05, "loss": 0.0349, "step": 24470 }, { "grad_norm": 0.2441636621952057, "learning_rate": 8.312911040566047e-05, "loss": 0.0291, "step": 24480 }, { "grad_norm": 0.2433849722146988, "learning_rate": 8.31136271877806e-05, "loss": 0.0265, "step": 24490 }, { "grad_norm": 0.42810431122779846, "learning_rate": 8.309813831168748e-05, "loss": 0.0325, "step": 24500 }, { "grad_norm": 0.4912082254886627, "learning_rate": 8.308264378002777e-05, "loss": 0.0292, "step": 24510 }, { "grad_norm": 0.35803624987602234, "learning_rate": 8.306714359544906e-05, "loss": 0.0373, "step": 24520 }, { "grad_norm": 0.357374906539917, "learning_rate": 8.30516377605999e-05, "loss": 0.0276, "step": 24530 }, { "grad_norm": 0.34840822219848633, "learning_rate": 8.30361262781298e-05, "loss": 0.0253, "step": 24540 }, { "grad_norm": 0.5141642689704895, "learning_rate": 8.302060915068924e-05, "loss": 0.0285, "step": 24550 }, { "grad_norm": 0.24043332040309906, "learning_rate": 8.300508638092972e-05, "loss": 0.0252, "step": 24560 }, { "grad_norm": 0.2513459622859955, "learning_rate": 8.298955797150361e-05, "loss": 0.027, "step": 24570 }, { "grad_norm": 0.4236547350883484, "learning_rate": 8.297402392506433e-05, "loss": 0.0275, "step": 24580 }, { "grad_norm": 0.2917407155036926, "learning_rate": 8.295848424426617e-05, "loss": 0.0237, "step": 24590 }, { "grad_norm": 0.338185578584671, "learning_rate": 8.29429389317645e-05, "loss": 0.0225, "step": 24600 }, { "grad_norm": 0.2606019973754883, "learning_rate": 8.292738799021556e-05, "loss": 0.0368, "step": 24610 }, { "grad_norm": 0.3160708248615265, "learning_rate": 8.291183142227656e-05, "loss": 0.0205, "step": 24620 }, { "grad_norm": 0.3734188377857208, "learning_rate": 8.289626923060572e-05, "loss": 0.0417, "step": 24630 }, { "grad_norm": 0.30764803290367126, "learning_rate": 8.288070141786218e-05, "loss": 0.0303, "step": 24640 }, { "grad_norm": 0.33216947317123413, "learning_rate": 8.286512798670605e-05, "loss": 0.0296, "step": 24650 }, { "grad_norm": 0.34934648871421814, "learning_rate": 8.284954893979842e-05, "loss": 0.0321, "step": 24660 }, { "grad_norm": 0.3170143961906433, "learning_rate": 8.283396427980131e-05, "loss": 0.0388, "step": 24670 }, { "grad_norm": 0.32440251111984253, "learning_rate": 8.281837400937771e-05, "loss": 0.0316, "step": 24680 }, { "grad_norm": 0.3459572494029999, "learning_rate": 8.28027781311916e-05, "loss": 0.0278, "step": 24690 }, { "grad_norm": 0.4158172905445099, "learning_rate": 8.278717664790785e-05, "loss": 0.0364, "step": 24700 }, { "grad_norm": 0.4444074034690857, "learning_rate": 8.277156956219234e-05, "loss": 0.0312, "step": 24710 }, { "grad_norm": 0.4038625955581665, "learning_rate": 8.275595687671189e-05, "loss": 0.0354, "step": 24720 }, { "grad_norm": 0.38942110538482666, "learning_rate": 8.27403385941343e-05, "loss": 0.0285, "step": 24730 }, { "grad_norm": 0.48770663142204285, "learning_rate": 8.272471471712828e-05, "loss": 0.0334, "step": 24740 }, { "grad_norm": 0.3462439775466919, "learning_rate": 8.270908524836355e-05, "loss": 0.0323, "step": 24750 }, { "grad_norm": 0.41278669238090515, "learning_rate": 8.269345019051074e-05, "loss": 0.0401, "step": 24760 }, { "grad_norm": 0.3010275363922119, "learning_rate": 8.267780954624147e-05, "loss": 0.0362, "step": 24770 }, { "grad_norm": 0.48382365703582764, "learning_rate": 8.266216331822827e-05, "loss": 0.0346, "step": 24780 }, { "grad_norm": 0.2969155013561249, "learning_rate": 8.264651150914469e-05, "loss": 0.0293, "step": 24790 }, { "grad_norm": 0.3540419042110443, "learning_rate": 8.263085412166517e-05, "loss": 0.0232, "step": 24800 }, { "grad_norm": 0.3262769281864166, "learning_rate": 8.261519115846514e-05, "loss": 0.0285, "step": 24810 }, { "grad_norm": 0.2841525375843048, "learning_rate": 8.259952262222096e-05, "loss": 0.0281, "step": 24820 }, { "grad_norm": 0.38046714663505554, "learning_rate": 8.258384851560997e-05, "loss": 0.0282, "step": 24830 }, { "grad_norm": 0.47620704770088196, "learning_rate": 8.256816884131044e-05, "loss": 0.0265, "step": 24840 }, { "grad_norm": 0.4532994031906128, "learning_rate": 8.255248360200159e-05, "loss": 0.0283, "step": 24850 }, { "grad_norm": 0.24670155346393585, "learning_rate": 8.253679280036359e-05, "loss": 0.0236, "step": 24860 }, { "grad_norm": 0.31701043248176575, "learning_rate": 8.252109643907762e-05, "loss": 0.0318, "step": 24870 }, { "grad_norm": 0.23817101120948792, "learning_rate": 8.250539452082569e-05, "loss": 0.0231, "step": 24880 }, { "grad_norm": 0.46487361192703247, "learning_rate": 8.248968704829087e-05, "loss": 0.0306, "step": 24890 }, { "grad_norm": 0.2861013412475586, "learning_rate": 8.247397402415714e-05, "loss": 0.0309, "step": 24900 }, { "grad_norm": 0.3770347237586975, "learning_rate": 8.24582554511094e-05, "loss": 0.0258, "step": 24910 }, { "grad_norm": 0.3880903720855713, "learning_rate": 8.244253133183355e-05, "loss": 0.0247, "step": 24920 }, { "grad_norm": 0.3178769052028656, "learning_rate": 8.24268016690164e-05, "loss": 0.0288, "step": 24930 }, { "grad_norm": 0.2893485724925995, "learning_rate": 8.241106646534571e-05, "loss": 0.0283, "step": 24940 }, { "grad_norm": 0.3126901090145111, "learning_rate": 8.23953257235102e-05, "loss": 0.0292, "step": 24950 }, { "grad_norm": 0.42700961232185364, "learning_rate": 8.237957944619956e-05, "loss": 0.0317, "step": 24960 }, { "grad_norm": 0.2808645963668823, "learning_rate": 8.236382763610437e-05, "loss": 0.0288, "step": 24970 }, { "grad_norm": 0.3041004538536072, "learning_rate": 8.234807029591619e-05, "loss": 0.0268, "step": 24980 }, { "grad_norm": 0.4313656687736511, "learning_rate": 8.233230742832752e-05, "loss": 0.0373, "step": 24990 }, { "grad_norm": 0.39138954877853394, "learning_rate": 8.231653903603178e-05, "loss": 0.0279, "step": 25000 }, { "grad_norm": 0.3318635821342468, "learning_rate": 8.23007651217234e-05, "loss": 0.0318, "step": 25010 }, { "grad_norm": 0.3522745370864868, "learning_rate": 8.228498568809769e-05, "loss": 0.028, "step": 25020 }, { "grad_norm": 0.3618367314338684, "learning_rate": 8.22692007378509e-05, "loss": 0.033, "step": 25030 }, { "grad_norm": 0.3179386854171753, "learning_rate": 8.225341027368028e-05, "loss": 0.0291, "step": 25040 }, { "grad_norm": 0.357546329498291, "learning_rate": 8.223761429828399e-05, "loss": 0.0305, "step": 25050 }, { "grad_norm": 0.35577112436294556, "learning_rate": 8.22218128143611e-05, "loss": 0.0318, "step": 25060 }, { "grad_norm": 0.32388395071029663, "learning_rate": 8.220600582461166e-05, "loss": 0.0271, "step": 25070 }, { "grad_norm": 0.28469231724739075, "learning_rate": 8.219019333173668e-05, "loss": 0.0284, "step": 25080 }, { "grad_norm": 0.39927470684051514, "learning_rate": 8.217437533843805e-05, "loss": 0.0315, "step": 25090 }, { "grad_norm": 0.2235809564590454, "learning_rate": 8.215855184741867e-05, "loss": 0.0283, "step": 25100 }, { "grad_norm": 0.2880808711051941, "learning_rate": 8.21427228613823e-05, "loss": 0.0254, "step": 25110 }, { "grad_norm": 0.41901257634162903, "learning_rate": 8.21268883830337e-05, "loss": 0.026, "step": 25120 }, { "grad_norm": 0.33623483777046204, "learning_rate": 8.211104841507855e-05, "loss": 0.0361, "step": 25130 }, { "grad_norm": 0.21361158788204193, "learning_rate": 8.209520296022346e-05, "loss": 0.0253, "step": 25140 }, { "grad_norm": 0.5017843246459961, "learning_rate": 8.207935202117599e-05, "loss": 0.0353, "step": 25150 }, { "grad_norm": 0.25258558988571167, "learning_rate": 8.206349560064463e-05, "loss": 0.0255, "step": 25160 }, { "grad_norm": 0.6358445286750793, "learning_rate": 8.204763370133881e-05, "loss": 0.0262, "step": 25170 }, { "grad_norm": 0.3154981732368469, "learning_rate": 8.203176632596892e-05, "loss": 0.0289, "step": 25180 }, { "grad_norm": 0.3853837549686432, "learning_rate": 8.20158934772462e-05, "loss": 0.0331, "step": 25190 }, { "grad_norm": 0.41725561022758484, "learning_rate": 8.200001515788294e-05, "loss": 0.0262, "step": 25200 }, { "grad_norm": 0.3802865147590637, "learning_rate": 8.198413137059228e-05, "loss": 0.0347, "step": 25210 }, { "grad_norm": 0.4121628403663635, "learning_rate": 8.196824211808835e-05, "loss": 0.0274, "step": 25220 }, { "grad_norm": 0.41145944595336914, "learning_rate": 8.195234740308617e-05, "loss": 0.0332, "step": 25230 }, { "grad_norm": 0.3411482274532318, "learning_rate": 8.193644722830171e-05, "loss": 0.0307, "step": 25240 }, { "grad_norm": 0.46245077252388, "learning_rate": 8.19205415964519e-05, "loss": 0.0207, "step": 25250 }, { "grad_norm": 0.20692358911037445, "learning_rate": 8.190463051025456e-05, "loss": 0.0246, "step": 25260 }, { "grad_norm": 0.3041267991065979, "learning_rate": 8.188871397242843e-05, "loss": 0.027, "step": 25270 }, { "grad_norm": 0.2854808270931244, "learning_rate": 8.187279198569326e-05, "loss": 0.0256, "step": 25280 }, { "grad_norm": 0.2988159656524658, "learning_rate": 8.185686455276966e-05, "loss": 0.0283, "step": 25290 }, { "grad_norm": 0.2143142968416214, "learning_rate": 8.184093167637921e-05, "loss": 0.0239, "step": 25300 }, { "grad_norm": 0.3179726004600525, "learning_rate": 8.182499335924437e-05, "loss": 0.0275, "step": 25310 }, { "grad_norm": 0.29029107093811035, "learning_rate": 8.18090496040886e-05, "loss": 0.0255, "step": 25320 }, { "grad_norm": 0.44882696866989136, "learning_rate": 8.179310041363621e-05, "loss": 0.0422, "step": 25330 }, { "grad_norm": 0.2711249589920044, "learning_rate": 8.17771457906125e-05, "loss": 0.0269, "step": 25340 }, { "grad_norm": 0.3193086087703705, "learning_rate": 8.176118573774371e-05, "loss": 0.0305, "step": 25350 }, { "grad_norm": 0.33152103424072266, "learning_rate": 8.174522025775692e-05, "loss": 0.0303, "step": 25360 }, { "grad_norm": 0.4909977912902832, "learning_rate": 8.172924935338022e-05, "loss": 0.0372, "step": 25370 }, { "grad_norm": 0.3087199330329895, "learning_rate": 8.171327302734262e-05, "loss": 0.0318, "step": 25380 }, { "grad_norm": 0.3199569880962372, "learning_rate": 8.169729128237401e-05, "loss": 0.0322, "step": 25390 }, { "grad_norm": 0.4473224878311157, "learning_rate": 8.168130412120525e-05, "loss": 0.0311, "step": 25400 }, { "grad_norm": 0.3603808581829071, "learning_rate": 8.16653115465681e-05, "loss": 0.0294, "step": 25410 }, { "grad_norm": 0.3435085713863373, "learning_rate": 8.164931356119526e-05, "loss": 0.0251, "step": 25420 }, { "grad_norm": 0.6748446226119995, "learning_rate": 8.163331016782032e-05, "loss": 0.0417, "step": 25430 }, { "grad_norm": 0.3398350179195404, "learning_rate": 8.161730136917785e-05, "loss": 0.028, "step": 25440 }, { "grad_norm": 0.4118909537792206, "learning_rate": 8.160128716800333e-05, "loss": 0.0365, "step": 25450 }, { "grad_norm": 0.266767680644989, "learning_rate": 8.158526756703313e-05, "loss": 0.03, "step": 25460 }, { "grad_norm": 0.4871802031993866, "learning_rate": 8.156924256900455e-05, "loss": 0.0355, "step": 25470 }, { "grad_norm": 0.4448726177215576, "learning_rate": 8.155321217665584e-05, "loss": 0.0306, "step": 25480 }, { "grad_norm": 0.3064424991607666, "learning_rate": 8.153717639272614e-05, "loss": 0.0273, "step": 25490 }, { "grad_norm": 0.3133241832256317, "learning_rate": 8.152113521995555e-05, "loss": 0.0302, "step": 25500 }, { "grad_norm": 0.3953215777873993, "learning_rate": 8.150508866108505e-05, "loss": 0.0297, "step": 25510 }, { "grad_norm": 0.3931196928024292, "learning_rate": 8.148903671885657e-05, "loss": 0.0318, "step": 25520 }, { "grad_norm": 0.34547075629234314, "learning_rate": 8.147297939601292e-05, "loss": 0.0423, "step": 25530 }, { "grad_norm": 0.579860508441925, "learning_rate": 8.145691669529792e-05, "loss": 0.0392, "step": 25540 }, { "grad_norm": 0.49050408601760864, "learning_rate": 8.144084861945618e-05, "loss": 0.0323, "step": 25550 }, { "grad_norm": 0.42165419459342957, "learning_rate": 8.142477517123333e-05, "loss": 0.0335, "step": 25560 }, { "grad_norm": 0.42730680108070374, "learning_rate": 8.140869635337586e-05, "loss": 0.0283, "step": 25570 }, { "grad_norm": 0.3617230951786041, "learning_rate": 8.139261216863123e-05, "loss": 0.0274, "step": 25580 }, { "grad_norm": 0.34508216381073, "learning_rate": 8.137652261974776e-05, "loss": 0.0356, "step": 25590 }, { "grad_norm": 0.31546759605407715, "learning_rate": 8.136042770947472e-05, "loss": 0.033, "step": 25600 }, { "grad_norm": 0.36940696835517883, "learning_rate": 8.134432744056228e-05, "loss": 0.0367, "step": 25610 }, { "grad_norm": 0.4398965537548065, "learning_rate": 8.132822181576158e-05, "loss": 0.0301, "step": 25620 }, { "grad_norm": 0.3998645544052124, "learning_rate": 8.131211083782459e-05, "loss": 0.0302, "step": 25630 }, { "grad_norm": 0.44066452980041504, "learning_rate": 8.129599450950424e-05, "loss": 0.0221, "step": 25640 }, { "grad_norm": 0.35408544540405273, "learning_rate": 8.127987283355438e-05, "loss": 0.0324, "step": 25650 }, { "grad_norm": 0.23960481584072113, "learning_rate": 8.126374581272976e-05, "loss": 0.0263, "step": 25660 }, { "grad_norm": 0.3049464225769043, "learning_rate": 8.124761344978605e-05, "loss": 0.0398, "step": 25670 }, { "grad_norm": 0.25736066699028015, "learning_rate": 8.12314757474798e-05, "loss": 0.0201, "step": 25680 }, { "grad_norm": 0.3433734178543091, "learning_rate": 8.121533270856856e-05, "loss": 0.0239, "step": 25690 }, { "grad_norm": 0.32785454392433167, "learning_rate": 8.119918433581069e-05, "loss": 0.0246, "step": 25700 }, { "grad_norm": 0.34070444107055664, "learning_rate": 8.118303063196551e-05, "loss": 0.0241, "step": 25710 }, { "grad_norm": 0.3495153486728668, "learning_rate": 8.116687159979326e-05, "loss": 0.0232, "step": 25720 }, { "grad_norm": 0.3547370731830597, "learning_rate": 8.115070724205508e-05, "loss": 0.0284, "step": 25730 }, { "grad_norm": 0.3282674252986908, "learning_rate": 8.113453756151296e-05, "loss": 0.0253, "step": 25740 }, { "grad_norm": 0.4442296326160431, "learning_rate": 8.111836256092995e-05, "loss": 0.0229, "step": 25750 }, { "grad_norm": 0.3283950984477997, "learning_rate": 8.110218224306985e-05, "loss": 0.0279, "step": 25760 }, { "grad_norm": 0.38558658957481384, "learning_rate": 8.108599661069745e-05, "loss": 0.0267, "step": 25770 }, { "grad_norm": 0.31877821683883667, "learning_rate": 8.106980566657845e-05, "loss": 0.0236, "step": 25780 }, { "grad_norm": 0.3155786693096161, "learning_rate": 8.10536094134794e-05, "loss": 0.027, "step": 25790 }, { "grad_norm": 0.36644700169563293, "learning_rate": 8.103740785416783e-05, "loss": 0.03, "step": 25800 }, { "grad_norm": 0.2806146740913391, "learning_rate": 8.102120099141212e-05, "loss": 0.0284, "step": 25810 }, { "grad_norm": 0.5564125180244446, "learning_rate": 8.100498882798163e-05, "loss": 0.0272, "step": 25820 }, { "grad_norm": 0.36140015721321106, "learning_rate": 8.09887713666465e-05, "loss": 0.0327, "step": 25830 }, { "grad_norm": 0.25046926736831665, "learning_rate": 8.09725486101779e-05, "loss": 0.0346, "step": 25840 }, { "grad_norm": 0.5177770256996155, "learning_rate": 8.095632056134784e-05, "loss": 0.0285, "step": 25850 }, { "grad_norm": 0.25915801525115967, "learning_rate": 8.094008722292925e-05, "loss": 0.0258, "step": 25860 }, { "grad_norm": 0.5083009600639343, "learning_rate": 8.092384859769598e-05, "loss": 0.0315, "step": 25870 }, { "grad_norm": 0.37652117013931274, "learning_rate": 8.090760468842275e-05, "loss": 0.0408, "step": 25880 }, { "grad_norm": 0.30663853883743286, "learning_rate": 8.089135549788521e-05, "loss": 0.031, "step": 25890 }, { "grad_norm": 0.27934470772743225, "learning_rate": 8.087510102885987e-05, "loss": 0.0246, "step": 25900 }, { "grad_norm": 0.2583511173725128, "learning_rate": 8.085884128412422e-05, "loss": 0.0365, "step": 25910 }, { "grad_norm": 0.36812588572502136, "learning_rate": 8.084257626645659e-05, "loss": 0.029, "step": 25920 }, { "grad_norm": 0.22224381566047668, "learning_rate": 8.08263059786362e-05, "loss": 0.0258, "step": 25930 }, { "grad_norm": 0.36330243945121765, "learning_rate": 8.081003042344325e-05, "loss": 0.0249, "step": 25940 }, { "grad_norm": 0.3155248761177063, "learning_rate": 8.079374960365872e-05, "loss": 0.0316, "step": 25950 }, { "grad_norm": 0.4260352551937103, "learning_rate": 8.077746352206463e-05, "loss": 0.0282, "step": 25960 }, { "grad_norm": 0.26684677600860596, "learning_rate": 8.076117218144377e-05, "loss": 0.0279, "step": 25970 }, { "grad_norm": 0.41479724645614624, "learning_rate": 8.074487558457991e-05, "loss": 0.0255, "step": 25980 }, { "grad_norm": 0.24125929176807404, "learning_rate": 8.072857373425768e-05, "loss": 0.0275, "step": 25990 }, { "grad_norm": 0.27655142545700073, "learning_rate": 8.071226663326264e-05, "loss": 0.0268, "step": 26000 }, { "grad_norm": 0.42752403020858765, "learning_rate": 8.069595428438121e-05, "loss": 0.0331, "step": 26010 }, { "grad_norm": 0.4655517637729645, "learning_rate": 8.067963669040072e-05, "loss": 0.0239, "step": 26020 }, { "grad_norm": 0.5454714894294739, "learning_rate": 8.066331385410942e-05, "loss": 0.0267, "step": 26030 }, { "grad_norm": 0.4289962649345398, "learning_rate": 8.064698577829641e-05, "loss": 0.0288, "step": 26040 }, { "grad_norm": 0.3405488431453705, "learning_rate": 8.063065246575175e-05, "loss": 0.0297, "step": 26050 }, { "grad_norm": 0.4076952040195465, "learning_rate": 8.061431391926631e-05, "loss": 0.0224, "step": 26060 }, { "grad_norm": 0.3632146716117859, "learning_rate": 8.059797014163195e-05, "loss": 0.0253, "step": 26070 }, { "grad_norm": 0.4900069236755371, "learning_rate": 8.058162113564133e-05, "loss": 0.0372, "step": 26080 }, { "grad_norm": 0.308820903301239, "learning_rate": 8.056526690408806e-05, "loss": 0.0381, "step": 26090 }, { "grad_norm": 0.2548111379146576, "learning_rate": 8.054890744976666e-05, "loss": 0.0298, "step": 26100 }, { "grad_norm": 0.2559564411640167, "learning_rate": 8.053254277547248e-05, "loss": 0.0216, "step": 26110 }, { "grad_norm": 0.3612932562828064, "learning_rate": 8.051617288400182e-05, "loss": 0.0295, "step": 26120 }, { "grad_norm": 0.33799514174461365, "learning_rate": 8.049979777815182e-05, "loss": 0.0411, "step": 26130 }, { "grad_norm": 0.39225706458091736, "learning_rate": 8.048341746072054e-05, "loss": 0.0307, "step": 26140 }, { "grad_norm": 0.512887179851532, "learning_rate": 8.046703193450696e-05, "loss": 0.0432, "step": 26150 }, { "grad_norm": 0.23858171701431274, "learning_rate": 8.04506412023109e-05, "loss": 0.0241, "step": 26160 }, { "grad_norm": 0.37548673152923584, "learning_rate": 8.043424526693306e-05, "loss": 0.0282, "step": 26170 }, { "grad_norm": 0.291440486907959, "learning_rate": 8.04178441311751e-05, "loss": 0.0252, "step": 26180 }, { "grad_norm": 0.373290091753006, "learning_rate": 8.04014377978395e-05, "loss": 0.0299, "step": 26190 }, { "grad_norm": 0.2516454756259918, "learning_rate": 8.038502626972967e-05, "loss": 0.0311, "step": 26200 }, { "grad_norm": 0.4021906852722168, "learning_rate": 8.036860954964989e-05, "loss": 0.0328, "step": 26210 }, { "grad_norm": 0.41902273893356323, "learning_rate": 8.035218764040531e-05, "loss": 0.0289, "step": 26220 }, { "grad_norm": 0.35399922728538513, "learning_rate": 8.033576054480199e-05, "loss": 0.0328, "step": 26230 }, { "grad_norm": 0.43688255548477173, "learning_rate": 8.031932826564688e-05, "loss": 0.023, "step": 26240 }, { "grad_norm": 0.3345430791378021, "learning_rate": 8.030289080574782e-05, "loss": 0.0262, "step": 26250 }, { "grad_norm": 0.484216570854187, "learning_rate": 8.028644816791349e-05, "loss": 0.0333, "step": 26260 }, { "grad_norm": 0.2776540517807007, "learning_rate": 8.027000035495351e-05, "loss": 0.0241, "step": 26270 }, { "grad_norm": 0.3700949251651764, "learning_rate": 8.025354736967836e-05, "loss": 0.0393, "step": 26280 }, { "grad_norm": 0.324810653924942, "learning_rate": 8.023708921489941e-05, "loss": 0.0337, "step": 26290 }, { "grad_norm": 0.2315192073583603, "learning_rate": 8.022062589342887e-05, "loss": 0.0329, "step": 26300 }, { "grad_norm": 0.7071444392204285, "learning_rate": 8.020415740807993e-05, "loss": 0.0434, "step": 26310 }, { "grad_norm": 0.40827980637550354, "learning_rate": 8.018768376166656e-05, "loss": 0.0256, "step": 26320 }, { "grad_norm": 0.33826008439064026, "learning_rate": 8.017120495700368e-05, "loss": 0.0294, "step": 26330 }, { "grad_norm": 0.32942166924476624, "learning_rate": 8.015472099690704e-05, "loss": 0.0278, "step": 26340 }, { "grad_norm": 0.19413283467292786, "learning_rate": 8.013823188419332e-05, "loss": 0.0217, "step": 26350 }, { "grad_norm": 0.2636334002017975, "learning_rate": 8.012173762168006e-05, "loss": 0.0281, "step": 26360 }, { "grad_norm": 0.2860932946205139, "learning_rate": 8.010523821218567e-05, "loss": 0.0204, "step": 26370 }, { "grad_norm": 0.2687477767467499, "learning_rate": 8.008873365852945e-05, "loss": 0.0214, "step": 26380 }, { "grad_norm": 0.3602549433708191, "learning_rate": 8.007222396353157e-05, "loss": 0.0231, "step": 26390 }, { "grad_norm": 0.45295053720474243, "learning_rate": 8.00557091300131e-05, "loss": 0.027, "step": 26400 }, { "grad_norm": 0.224347323179245, "learning_rate": 8.003918916079597e-05, "loss": 0.0341, "step": 26410 }, { "grad_norm": 0.3784966468811035, "learning_rate": 8.002266405870298e-05, "loss": 0.0288, "step": 26420 }, { "grad_norm": 0.3255341947078705, "learning_rate": 8.000613382655782e-05, "loss": 0.037, "step": 26430 }, { "grad_norm": 0.42135152220726013, "learning_rate": 7.998959846718505e-05, "loss": 0.0287, "step": 26440 }, { "grad_norm": 0.4890979528427124, "learning_rate": 7.997305798341012e-05, "loss": 0.0333, "step": 26450 }, { "grad_norm": 0.39742785692214966, "learning_rate": 7.995651237805937e-05, "loss": 0.0345, "step": 26460 }, { "grad_norm": 0.39079609513282776, "learning_rate": 7.993996165395996e-05, "loss": 0.0365, "step": 26470 }, { "grad_norm": 0.3605290949344635, "learning_rate": 7.992340581393996e-05, "loss": 0.0319, "step": 26480 }, { "grad_norm": 0.4283379316329956, "learning_rate": 7.990684486082831e-05, "loss": 0.0329, "step": 26490 }, { "grad_norm": 0.387021541595459, "learning_rate": 7.989027879745482e-05, "loss": 0.031, "step": 26500 }, { "grad_norm": 0.3002363443374634, "learning_rate": 7.98737076266502e-05, "loss": 0.0397, "step": 26510 }, { "grad_norm": 0.2647313177585602, "learning_rate": 7.985713135124598e-05, "loss": 0.027, "step": 26520 }, { "grad_norm": 0.4722326695919037, "learning_rate": 7.98405499740746e-05, "loss": 0.0364, "step": 26530 }, { "grad_norm": 0.6228498816490173, "learning_rate": 7.98239634979694e-05, "loss": 0.0345, "step": 26540 }, { "grad_norm": 0.4510497748851776, "learning_rate": 7.98073719257645e-05, "loss": 0.0276, "step": 26550 }, { "grad_norm": 0.7364043593406677, "learning_rate": 7.979077526029499e-05, "loss": 0.0328, "step": 26560 }, { "grad_norm": 0.29990434646606445, "learning_rate": 7.977417350439675e-05, "loss": 0.0242, "step": 26570 }, { "grad_norm": 0.3163500726222992, "learning_rate": 7.97575666609066e-05, "loss": 0.0275, "step": 26580 }, { "grad_norm": 0.5012360215187073, "learning_rate": 7.974095473266216e-05, "loss": 0.0297, "step": 26590 }, { "grad_norm": 0.4341354966163635, "learning_rate": 7.972433772250198e-05, "loss": 0.0345, "step": 26600 }, { "grad_norm": 0.2960723042488098, "learning_rate": 7.970771563326544e-05, "loss": 0.0267, "step": 26610 }, { "grad_norm": 0.33538752794265747, "learning_rate": 7.96910884677928e-05, "loss": 0.0291, "step": 26620 }, { "grad_norm": 0.33690574765205383, "learning_rate": 7.967445622892523e-05, "loss": 0.0296, "step": 26630 }, { "grad_norm": 0.1968059539794922, "learning_rate": 7.965781891950465e-05, "loss": 0.0225, "step": 26640 }, { "grad_norm": 0.366715669631958, "learning_rate": 7.964117654237397e-05, "loss": 0.0192, "step": 26650 }, { "grad_norm": 0.39337456226348877, "learning_rate": 7.962452910037692e-05, "loss": 0.0311, "step": 26660 }, { "grad_norm": 0.311769038438797, "learning_rate": 7.96078765963581e-05, "loss": 0.0254, "step": 26670 }, { "grad_norm": 0.41456738114356995, "learning_rate": 7.95912190331629e-05, "loss": 0.0279, "step": 26680 }, { "grad_norm": 0.28869926929473877, "learning_rate": 7.957455641363772e-05, "loss": 0.0252, "step": 26690 }, { "grad_norm": 0.5064225196838379, "learning_rate": 7.955788874062968e-05, "loss": 0.0378, "step": 26700 }, { "grad_norm": 0.4635598361492157, "learning_rate": 7.95412160169869e-05, "loss": 0.0299, "step": 26710 }, { "grad_norm": 0.5336678624153137, "learning_rate": 7.952453824555824e-05, "loss": 0.0189, "step": 26720 }, { "grad_norm": 0.2961617410182953, "learning_rate": 7.95078554291935e-05, "loss": 0.0201, "step": 26730 }, { "grad_norm": 0.31086310744285583, "learning_rate": 7.94911675707433e-05, "loss": 0.0244, "step": 26740 }, { "grad_norm": 0.3886480927467346, "learning_rate": 7.947447467305915e-05, "loss": 0.0303, "step": 26750 }, { "grad_norm": 0.3534455895423889, "learning_rate": 7.94577767389934e-05, "loss": 0.0267, "step": 26760 }, { "grad_norm": 0.31779730319976807, "learning_rate": 7.944107377139928e-05, "loss": 0.0246, "step": 26770 }, { "grad_norm": 0.27522173523902893, "learning_rate": 7.942436577313088e-05, "loss": 0.0279, "step": 26780 }, { "grad_norm": 0.28217199444770813, "learning_rate": 7.940765274704312e-05, "loss": 0.0309, "step": 26790 }, { "grad_norm": 0.30394354462623596, "learning_rate": 7.939093469599181e-05, "loss": 0.0219, "step": 26800 }, { "grad_norm": 0.20352372527122498, "learning_rate": 7.93742116228336e-05, "loss": 0.0248, "step": 26810 }, { "grad_norm": 0.3559007942676544, "learning_rate": 7.935748353042602e-05, "loss": 0.028, "step": 26820 }, { "grad_norm": 0.35209885239601135, "learning_rate": 7.934075042162744e-05, "loss": 0.0309, "step": 26830 }, { "grad_norm": 0.6358603239059448, "learning_rate": 7.932401229929705e-05, "loss": 0.0348, "step": 26840 }, { "grad_norm": 0.4527646005153656, "learning_rate": 7.9307269166295e-05, "loss": 0.0324, "step": 26850 }, { "grad_norm": 0.32776299118995667, "learning_rate": 7.92905210254822e-05, "loss": 0.0244, "step": 26860 }, { "grad_norm": 0.5962305665016174, "learning_rate": 7.927376787972045e-05, "loss": 0.028, "step": 26870 }, { "grad_norm": 0.3795536160469055, "learning_rate": 7.92570097318724e-05, "loss": 0.0332, "step": 26880 }, { "grad_norm": 0.3168638348579407, "learning_rate": 7.924024658480158e-05, "loss": 0.0379, "step": 26890 }, { "grad_norm": 0.44542253017425537, "learning_rate": 7.922347844137233e-05, "loss": 0.0309, "step": 26900 }, { "grad_norm": 0.32770222425460815, "learning_rate": 7.92067053044499e-05, "loss": 0.0318, "step": 26910 }, { "grad_norm": 0.4431760013103485, "learning_rate": 7.918992717690031e-05, "loss": 0.0353, "step": 26920 }, { "grad_norm": 0.5382694602012634, "learning_rate": 7.917314406159053e-05, "loss": 0.0358, "step": 26930 }, { "grad_norm": 0.596555233001709, "learning_rate": 7.915635596138832e-05, "loss": 0.0346, "step": 26940 }, { "grad_norm": 0.3025146424770355, "learning_rate": 7.913956287916228e-05, "loss": 0.0287, "step": 26950 }, { "grad_norm": 0.3439398407936096, "learning_rate": 7.912276481778193e-05, "loss": 0.0259, "step": 26960 }, { "grad_norm": 0.19430428743362427, "learning_rate": 7.910596178011759e-05, "loss": 0.026, "step": 26970 }, { "grad_norm": 0.38939979672431946, "learning_rate": 7.908915376904043e-05, "loss": 0.0251, "step": 26980 }, { "grad_norm": 0.2408418208360672, "learning_rate": 7.907234078742247e-05, "loss": 0.028, "step": 26990 }, { "grad_norm": 0.32407841086387634, "learning_rate": 7.90555228381366e-05, "loss": 0.0239, "step": 27000 }, { "grad_norm": 0.25760218501091003, "learning_rate": 7.903869992405656e-05, "loss": 0.0228, "step": 27010 }, { "grad_norm": 0.37705063819885254, "learning_rate": 7.902187204805691e-05, "loss": 0.0274, "step": 27020 }, { "grad_norm": 0.29621371626853943, "learning_rate": 7.900503921301308e-05, "loss": 0.024, "step": 27030 }, { "grad_norm": 0.32896631956100464, "learning_rate": 7.898820142180133e-05, "loss": 0.0304, "step": 27040 }, { "grad_norm": 0.44605258107185364, "learning_rate": 7.897135867729879e-05, "loss": 0.0282, "step": 27050 }, { "grad_norm": 0.31257933378219604, "learning_rate": 7.89545109823834e-05, "loss": 0.0234, "step": 27060 }, { "grad_norm": 0.43141862750053406, "learning_rate": 7.8937658339934e-05, "loss": 0.03, "step": 27070 }, { "grad_norm": 0.3925040066242218, "learning_rate": 7.892080075283026e-05, "loss": 0.0296, "step": 27080 }, { "grad_norm": 0.29601040482521057, "learning_rate": 7.890393822395263e-05, "loss": 0.032, "step": 27090 }, { "grad_norm": 0.3227587640285492, "learning_rate": 7.88870707561825e-05, "loss": 0.028, "step": 27100 }, { "grad_norm": 0.6126355528831482, "learning_rate": 7.887019835240203e-05, "loss": 0.0383, "step": 27110 }, { "grad_norm": 0.4321099817752838, "learning_rate": 7.885332101549427e-05, "loss": 0.0309, "step": 27120 }, { "grad_norm": 0.37091225385665894, "learning_rate": 7.883643874834308e-05, "loss": 0.0275, "step": 27130 }, { "grad_norm": 0.4451334476470947, "learning_rate": 7.881955155383321e-05, "loss": 0.0405, "step": 27140 }, { "grad_norm": 0.34639447927474976, "learning_rate": 7.880265943485017e-05, "loss": 0.0267, "step": 27150 }, { "grad_norm": 0.3107856512069702, "learning_rate": 7.878576239428038e-05, "loss": 0.0261, "step": 27160 }, { "grad_norm": 0.4877810776233673, "learning_rate": 7.87688604350111e-05, "loss": 0.0306, "step": 27170 }, { "grad_norm": 0.5289612412452698, "learning_rate": 7.875195355993042e-05, "loss": 0.0267, "step": 27180 }, { "grad_norm": 0.4339142143726349, "learning_rate": 7.873504177192724e-05, "loss": 0.0298, "step": 27190 }, { "grad_norm": 0.35478585958480835, "learning_rate": 7.87181250738913e-05, "loss": 0.0348, "step": 27200 }, { "grad_norm": 0.28349554538726807, "learning_rate": 7.870120346871324e-05, "loss": 0.0278, "step": 27210 }, { "grad_norm": 0.26270031929016113, "learning_rate": 7.86842769592845e-05, "loss": 0.0253, "step": 27220 }, { "grad_norm": 0.37429553270339966, "learning_rate": 7.866734554849732e-05, "loss": 0.0258, "step": 27230 }, { "grad_norm": 0.3022003471851349, "learning_rate": 7.865040923924486e-05, "loss": 0.0245, "step": 27240 }, { "grad_norm": 0.7093153595924377, "learning_rate": 7.863346803442104e-05, "loss": 0.0368, "step": 27250 }, { "grad_norm": 0.287383496761322, "learning_rate": 7.861652193692067e-05, "loss": 0.0208, "step": 27260 }, { "grad_norm": 0.6060658097267151, "learning_rate": 7.859957094963937e-05, "loss": 0.0203, "step": 27270 }, { "grad_norm": 0.38942739367485046, "learning_rate": 7.858261507547357e-05, "loss": 0.0216, "step": 27280 }, { "grad_norm": 0.2503901422023773, "learning_rate": 7.856565431732061e-05, "loss": 0.0214, "step": 27290 }, { "grad_norm": 0.2261558175086975, "learning_rate": 7.854868867807859e-05, "loss": 0.0199, "step": 27300 }, { "grad_norm": 0.2996388077735901, "learning_rate": 7.85317181606465e-05, "loss": 0.0266, "step": 27310 }, { "grad_norm": 0.26895320415496826, "learning_rate": 7.85147427679241e-05, "loss": 0.0249, "step": 27320 }, { "grad_norm": 0.30856984853744507, "learning_rate": 7.849776250281205e-05, "loss": 0.0207, "step": 27330 }, { "grad_norm": 0.6007102727890015, "learning_rate": 7.84807773682118e-05, "loss": 0.0289, "step": 27340 }, { "grad_norm": 0.20520900189876556, "learning_rate": 7.846378736702565e-05, "loss": 0.0303, "step": 27350 }, { "grad_norm": 0.44306159019470215, "learning_rate": 7.844679250215671e-05, "loss": 0.0303, "step": 27360 }, { "grad_norm": 0.2703341543674469, "learning_rate": 7.842979277650898e-05, "loss": 0.0291, "step": 27370 }, { "grad_norm": 0.3563581705093384, "learning_rate": 7.84127881929872e-05, "loss": 0.0283, "step": 27380 }, { "grad_norm": 0.2931148111820221, "learning_rate": 7.839577875449704e-05, "loss": 0.0293, "step": 27390 }, { "grad_norm": 0.4470328092575073, "learning_rate": 7.837876446394489e-05, "loss": 0.0276, "step": 27400 }, { "grad_norm": 0.3510046899318695, "learning_rate": 7.836174532423805e-05, "loss": 0.0317, "step": 27410 }, { "grad_norm": 0.2015310674905777, "learning_rate": 7.834472133828466e-05, "loss": 0.0258, "step": 27420 }, { "grad_norm": 0.5158534049987793, "learning_rate": 7.832769250899359e-05, "loss": 0.0296, "step": 27430 }, { "grad_norm": 0.4588307738304138, "learning_rate": 7.831065883927464e-05, "loss": 0.0356, "step": 27440 }, { "grad_norm": 0.424601286649704, "learning_rate": 7.829362033203841e-05, "loss": 0.0307, "step": 27450 }, { "grad_norm": 0.37808817625045776, "learning_rate": 7.827657699019628e-05, "loss": 0.0255, "step": 27460 }, { "grad_norm": 0.32659366726875305, "learning_rate": 7.825952881666052e-05, "loss": 0.0404, "step": 27470 }, { "grad_norm": 0.2860170304775238, "learning_rate": 7.824247581434418e-05, "loss": 0.0244, "step": 27480 }, { "grad_norm": 0.4269990622997284, "learning_rate": 7.822541798616116e-05, "loss": 0.0292, "step": 27490 }, { "grad_norm": 0.3069738447666168, "learning_rate": 7.820835533502617e-05, "loss": 0.0341, "step": 27500 }, { "grad_norm": 0.6990815997123718, "learning_rate": 7.819128786385475e-05, "loss": 0.0361, "step": 27510 }, { "grad_norm": 0.3420751094818115, "learning_rate": 7.817421557556329e-05, "loss": 0.0307, "step": 27520 }, { "grad_norm": 0.3649970293045044, "learning_rate": 7.815713847306893e-05, "loss": 0.0332, "step": 27530 }, { "grad_norm": 0.6694777011871338, "learning_rate": 7.81400565592897e-05, "loss": 0.0295, "step": 27540 }, { "grad_norm": 0.4367477595806122, "learning_rate": 7.812296983714444e-05, "loss": 0.0372, "step": 27550 }, { "grad_norm": 0.24459964036941528, "learning_rate": 7.810587830955281e-05, "loss": 0.0244, "step": 27560 }, { "grad_norm": 0.3612726926803589, "learning_rate": 7.808878197943528e-05, "loss": 0.0291, "step": 27570 }, { "grad_norm": 0.4487927556037903, "learning_rate": 7.807168084971312e-05, "loss": 0.0265, "step": 27580 }, { "grad_norm": 0.2837933301925659, "learning_rate": 7.805457492330849e-05, "loss": 0.0221, "step": 27590 }, { "grad_norm": 0.4004633128643036, "learning_rate": 7.803746420314428e-05, "loss": 0.0256, "step": 27600 }, { "grad_norm": 0.2974553108215332, "learning_rate": 7.802034869214428e-05, "loss": 0.0279, "step": 27610 }, { "grad_norm": 0.22855977714061737, "learning_rate": 7.800322839323303e-05, "loss": 0.0208, "step": 27620 }, { "grad_norm": 0.352438747882843, "learning_rate": 7.798610330933593e-05, "loss": 0.0263, "step": 27630 }, { "grad_norm": 0.315422385931015, "learning_rate": 7.796897344337922e-05, "loss": 0.0352, "step": 27640 }, { "grad_norm": 0.3327898383140564, "learning_rate": 7.795183879828989e-05, "loss": 0.0373, "step": 27650 }, { "grad_norm": 0.49031445384025574, "learning_rate": 7.793469937699579e-05, "loss": 0.0397, "step": 27660 }, { "grad_norm": 0.3109623193740845, "learning_rate": 7.791755518242558e-05, "loss": 0.0324, "step": 27670 }, { "grad_norm": 0.4086689054965973, "learning_rate": 7.790040621750876e-05, "loss": 0.0284, "step": 27680 }, { "grad_norm": 0.32837969064712524, "learning_rate": 7.788325248517558e-05, "loss": 0.0219, "step": 27690 }, { "grad_norm": 0.4609223008155823, "learning_rate": 7.786609398835715e-05, "loss": 0.0389, "step": 27700 }, { "grad_norm": 0.32854634523391724, "learning_rate": 7.784893072998541e-05, "loss": 0.0261, "step": 27710 }, { "grad_norm": 0.28573837876319885, "learning_rate": 7.783176271299306e-05, "loss": 0.0232, "step": 27720 }, { "grad_norm": 0.3580412268638611, "learning_rate": 7.781458994031368e-05, "loss": 0.0257, "step": 27730 }, { "grad_norm": 0.3456045687198639, "learning_rate": 7.779741241488161e-05, "loss": 0.0328, "step": 27740 }, { "grad_norm": 0.5006719827651978, "learning_rate": 7.7780230139632e-05, "loss": 0.0242, "step": 27750 }, { "grad_norm": 0.30843719840049744, "learning_rate": 7.776304311750087e-05, "loss": 0.0217, "step": 27760 }, { "grad_norm": 1.0347728729248047, "learning_rate": 7.7745851351425e-05, "loss": 0.0351, "step": 27770 }, { "grad_norm": 0.4885454773902893, "learning_rate": 7.772865484434197e-05, "loss": 0.0309, "step": 27780 }, { "grad_norm": 0.5516655445098877, "learning_rate": 7.77114535991902e-05, "loss": 0.0356, "step": 27790 }, { "grad_norm": 0.4718550443649292, "learning_rate": 7.769424761890893e-05, "loss": 0.0334, "step": 27800 }, { "grad_norm": 0.6435683369636536, "learning_rate": 7.767703690643817e-05, "loss": 0.043, "step": 27810 }, { "grad_norm": 0.23092754185199738, "learning_rate": 7.76598214647188e-05, "loss": 0.0302, "step": 27820 }, { "grad_norm": 0.4770656228065491, "learning_rate": 7.764260129669241e-05, "loss": 0.0351, "step": 27830 }, { "grad_norm": 0.44740602374076843, "learning_rate": 7.76253764053015e-05, "loss": 0.0423, "step": 27840 }, { "grad_norm": 0.2310788929462433, "learning_rate": 7.760814679348932e-05, "loss": 0.0255, "step": 27850 }, { "grad_norm": 0.6565482020378113, "learning_rate": 7.759091246419992e-05, "loss": 0.0307, "step": 27860 }, { "grad_norm": 0.31924912333488464, "learning_rate": 7.75736734203782e-05, "loss": 0.0248, "step": 27870 }, { "grad_norm": 0.4534851312637329, "learning_rate": 7.755642966496985e-05, "loss": 0.0426, "step": 27880 }, { "grad_norm": 0.43643563985824585, "learning_rate": 7.753918120092132e-05, "loss": 0.0349, "step": 27890 }, { "grad_norm": 0.37985721230506897, "learning_rate": 7.752192803117993e-05, "loss": 0.0319, "step": 27900 }, { "grad_norm": 0.4411304295063019, "learning_rate": 7.750467015869377e-05, "loss": 0.0292, "step": 27910 }, { "grad_norm": 0.3264487087726593, "learning_rate": 7.748740758641174e-05, "loss": 0.0299, "step": 27920 }, { "grad_norm": 0.40045639872550964, "learning_rate": 7.74701403172835e-05, "loss": 0.0355, "step": 27930 }, { "grad_norm": 0.3016695976257324, "learning_rate": 7.745286835425962e-05, "loss": 0.029, "step": 27940 }, { "grad_norm": 0.4666965901851654, "learning_rate": 7.743559170029138e-05, "loss": 0.0309, "step": 27950 }, { "grad_norm": 0.3110596537590027, "learning_rate": 7.741831035833087e-05, "loss": 0.038, "step": 27960 }, { "grad_norm": 0.3888750672340393, "learning_rate": 7.740102433133102e-05, "loss": 0.0298, "step": 27970 }, { "grad_norm": 0.2708633840084076, "learning_rate": 7.738373362224553e-05, "loss": 0.0242, "step": 27980 }, { "grad_norm": 0.35760292410850525, "learning_rate": 7.73664382340289e-05, "loss": 0.0244, "step": 27990 }, { "grad_norm": 0.43563276529312134, "learning_rate": 7.734913816963647e-05, "loss": 0.0346, "step": 28000 }, { "grad_norm": 0.4058198630809784, "learning_rate": 7.73318334320243e-05, "loss": 0.0251, "step": 28010 }, { "grad_norm": 0.4396870732307434, "learning_rate": 7.731452402414934e-05, "loss": 0.0332, "step": 28020 }, { "grad_norm": 0.3496159613132477, "learning_rate": 7.729720994896928e-05, "loss": 0.0248, "step": 28030 }, { "grad_norm": 0.26288318634033203, "learning_rate": 7.727989120944262e-05, "loss": 0.026, "step": 28040 }, { "grad_norm": 0.2413351684808731, "learning_rate": 7.726256780852865e-05, "loss": 0.0259, "step": 28050 }, { "grad_norm": 0.24240462481975555, "learning_rate": 7.724523974918749e-05, "loss": 0.0233, "step": 28060 }, { "grad_norm": 0.2786235809326172, "learning_rate": 7.722790703438002e-05, "loss": 0.0246, "step": 28070 }, { "grad_norm": 0.24801470339298248, "learning_rate": 7.72105696670679e-05, "loss": 0.0199, "step": 28080 }, { "grad_norm": 0.3317612409591675, "learning_rate": 7.719322765021364e-05, "loss": 0.0244, "step": 28090 }, { "grad_norm": 0.20568546652793884, "learning_rate": 7.717588098678051e-05, "loss": 0.0241, "step": 28100 }, { "grad_norm": 0.3699486255645752, "learning_rate": 7.715852967973258e-05, "loss": 0.0254, "step": 28110 }, { "grad_norm": 0.38660845160484314, "learning_rate": 7.714117373203474e-05, "loss": 0.0238, "step": 28120 }, { "grad_norm": 0.25257736444473267, "learning_rate": 7.712381314665259e-05, "loss": 0.0309, "step": 28130 }, { "grad_norm": 0.31334391236305237, "learning_rate": 7.710644792655261e-05, "loss": 0.0226, "step": 28140 }, { "grad_norm": 0.4860398769378662, "learning_rate": 7.708907807470207e-05, "loss": 0.0398, "step": 28150 }, { "grad_norm": 0.2534085810184479, "learning_rate": 7.707170359406896e-05, "loss": 0.0323, "step": 28160 }, { "grad_norm": 0.40790578722953796, "learning_rate": 7.705432448762213e-05, "loss": 0.031, "step": 28170 }, { "grad_norm": 0.25266072154045105, "learning_rate": 7.703694075833117e-05, "loss": 0.0362, "step": 28180 }, { "grad_norm": 0.3264906406402588, "learning_rate": 7.70195524091665e-05, "loss": 0.0454, "step": 28190 }, { "grad_norm": 0.47832831740379333, "learning_rate": 7.70021594430993e-05, "loss": 0.0302, "step": 28200 }, { "grad_norm": 0.32823964953422546, "learning_rate": 7.698476186310157e-05, "loss": 0.0249, "step": 28210 }, { "grad_norm": 0.33892515301704407, "learning_rate": 7.696735967214608e-05, "loss": 0.0224, "step": 28220 }, { "grad_norm": 0.3727912902832031, "learning_rate": 7.694995287320636e-05, "loss": 0.0269, "step": 28230 }, { "grad_norm": 0.25428590178489685, "learning_rate": 7.693254146925679e-05, "loss": 0.0248, "step": 28240 }, { "grad_norm": 0.3765193521976471, "learning_rate": 7.691512546327251e-05, "loss": 0.0399, "step": 28250 }, { "grad_norm": 0.2838205397129059, "learning_rate": 7.689770485822939e-05, "loss": 0.0314, "step": 28260 }, { "grad_norm": 0.4727115333080292, "learning_rate": 7.688027965710416e-05, "loss": 0.0227, "step": 28270 }, { "grad_norm": 0.4176926910877228, "learning_rate": 7.686284986287433e-05, "loss": 0.0325, "step": 28280 }, { "grad_norm": 0.239309623837471, "learning_rate": 7.684541547851817e-05, "loss": 0.0257, "step": 28290 }, { "grad_norm": 0.3217770457267761, "learning_rate": 7.68279765070147e-05, "loss": 0.0255, "step": 28300 }, { "grad_norm": 0.42857444286346436, "learning_rate": 7.68105329513438e-05, "loss": 0.0322, "step": 28310 }, { "grad_norm": 0.40783146023750305, "learning_rate": 7.67930848144861e-05, "loss": 0.0314, "step": 28320 }, { "grad_norm": 0.2936444878578186, "learning_rate": 7.6775632099423e-05, "loss": 0.0328, "step": 28330 }, { "grad_norm": 0.3295149505138397, "learning_rate": 7.675817480913667e-05, "loss": 0.0289, "step": 28340 }, { "grad_norm": 0.33278483152389526, "learning_rate": 7.674071294661011e-05, "loss": 0.0238, "step": 28350 }, { "grad_norm": 0.2660770118236542, "learning_rate": 7.672324651482707e-05, "loss": 0.0226, "step": 28360 }, { "grad_norm": 0.25129321217536926, "learning_rate": 7.670577551677209e-05, "loss": 0.0308, "step": 28370 }, { "grad_norm": 0.24377159774303436, "learning_rate": 7.668829995543047e-05, "loss": 0.0228, "step": 28380 }, { "grad_norm": 0.4200083315372467, "learning_rate": 7.667081983378832e-05, "loss": 0.0265, "step": 28390 }, { "grad_norm": 0.278860867023468, "learning_rate": 7.66533351548325e-05, "loss": 0.0219, "step": 28400 }, { "grad_norm": 0.39009904861450195, "learning_rate": 7.663584592155069e-05, "loss": 0.0238, "step": 28410 }, { "grad_norm": 0.30948156118392944, "learning_rate": 7.661835213693129e-05, "loss": 0.0299, "step": 28420 }, { "grad_norm": 0.4827541410923004, "learning_rate": 7.660085380396353e-05, "loss": 0.0271, "step": 28430 }, { "grad_norm": 0.49029725790023804, "learning_rate": 7.658335092563738e-05, "loss": 0.0271, "step": 28440 }, { "grad_norm": 0.4087710380554199, "learning_rate": 7.656584350494362e-05, "loss": 0.0281, "step": 28450 }, { "grad_norm": 0.36232060194015503, "learning_rate": 7.654833154487378e-05, "loss": 0.0263, "step": 28460 }, { "grad_norm": 0.3362904191017151, "learning_rate": 7.653081504842017e-05, "loss": 0.0248, "step": 28470 }, { "grad_norm": 0.4512111246585846, "learning_rate": 7.65132940185759e-05, "loss": 0.0343, "step": 28480 }, { "grad_norm": 0.3128792941570282, "learning_rate": 7.649576845833481e-05, "loss": 0.0242, "step": 28490 }, { "grad_norm": 0.32242581248283386, "learning_rate": 7.647823837069156e-05, "loss": 0.0301, "step": 28500 }, { "grad_norm": 0.35729727149009705, "learning_rate": 7.646070375864156e-05, "loss": 0.0218, "step": 28510 }, { "grad_norm": 0.28144580125808716, "learning_rate": 7.644316462518099e-05, "loss": 0.0303, "step": 28520 }, { "grad_norm": 0.4504086375236511, "learning_rate": 7.642562097330679e-05, "loss": 0.0272, "step": 28530 }, { "grad_norm": 0.6226091384887695, "learning_rate": 7.640807280601671e-05, "loss": 0.0268, "step": 28540 }, { "grad_norm": 0.34146934747695923, "learning_rate": 7.639052012630927e-05, "loss": 0.0374, "step": 28550 }, { "grad_norm": 0.41559073328971863, "learning_rate": 7.63729629371837e-05, "loss": 0.0309, "step": 28560 }, { "grad_norm": 0.32601198554039, "learning_rate": 7.635540124164009e-05, "loss": 0.0334, "step": 28570 }, { "grad_norm": 0.5253439545631409, "learning_rate": 7.633783504267922e-05, "loss": 0.028, "step": 28580 }, { "grad_norm": 0.2538597881793976, "learning_rate": 7.632026434330269e-05, "loss": 0.0295, "step": 28590 }, { "grad_norm": 0.27091848850250244, "learning_rate": 7.630268914651282e-05, "loss": 0.0211, "step": 28600 }, { "grad_norm": 0.2235511839389801, "learning_rate": 7.628510945531278e-05, "loss": 0.0361, "step": 28610 }, { "grad_norm": 0.3397577702999115, "learning_rate": 7.626752527270641e-05, "loss": 0.0323, "step": 28620 }, { "grad_norm": 0.24162611365318298, "learning_rate": 7.62499366016984e-05, "loss": 0.0258, "step": 28630 }, { "grad_norm": 0.2855992019176483, "learning_rate": 7.623234344529416e-05, "loss": 0.028, "step": 28640 }, { "grad_norm": 0.29049256443977356, "learning_rate": 7.62147458064999e-05, "loss": 0.0262, "step": 28650 }, { "grad_norm": 0.4640318751335144, "learning_rate": 7.619714368832254e-05, "loss": 0.0282, "step": 28660 }, { "grad_norm": 0.34423601627349854, "learning_rate": 7.61795370937698e-05, "loss": 0.0252, "step": 28670 }, { "grad_norm": 0.2638148367404938, "learning_rate": 7.61619260258502e-05, "loss": 0.028, "step": 28680 }, { "grad_norm": 0.4034470319747925, "learning_rate": 7.614431048757298e-05, "loss": 0.0268, "step": 28690 }, { "grad_norm": 0.25927361845970154, "learning_rate": 7.612669048194814e-05, "loss": 0.0313, "step": 28700 }, { "grad_norm": 0.3505948483943939, "learning_rate": 7.610906601198646e-05, "loss": 0.0263, "step": 28710 }, { "grad_norm": 0.2812603712081909, "learning_rate": 7.60914370806995e-05, "loss": 0.0335, "step": 28720 }, { "grad_norm": 0.19233809411525726, "learning_rate": 7.607380369109953e-05, "loss": 0.0382, "step": 28730 }, { "grad_norm": 0.30219465494155884, "learning_rate": 7.605616584619961e-05, "loss": 0.0193, "step": 28740 }, { "grad_norm": 0.3244791328907013, "learning_rate": 7.603852354901362e-05, "loss": 0.0238, "step": 28750 }, { "grad_norm": 0.2659243643283844, "learning_rate": 7.602087680255609e-05, "loss": 0.0381, "step": 28760 }, { "grad_norm": 0.38908952474594116, "learning_rate": 7.600322560984238e-05, "loss": 0.0244, "step": 28770 }, { "grad_norm": 0.3045126795768738, "learning_rate": 7.598556997388863e-05, "loss": 0.031, "step": 28780 }, { "grad_norm": 0.3151797354221344, "learning_rate": 7.596790989771166e-05, "loss": 0.0267, "step": 28790 }, { "grad_norm": 0.30621278285980225, "learning_rate": 7.595024538432914e-05, "loss": 0.0275, "step": 28800 }, { "grad_norm": 0.29084059596061707, "learning_rate": 7.59325764367594e-05, "loss": 0.0272, "step": 28810 }, { "grad_norm": 0.24118083715438843, "learning_rate": 7.59149030580216e-05, "loss": 0.0263, "step": 28820 }, { "grad_norm": 0.43589991331100464, "learning_rate": 7.589722525113562e-05, "loss": 0.0249, "step": 28830 }, { "grad_norm": 0.2823154032230377, "learning_rate": 7.587954301912216e-05, "loss": 0.0214, "step": 28840 }, { "grad_norm": 0.578715980052948, "learning_rate": 7.586185636500263e-05, "loss": 0.0308, "step": 28850 }, { "grad_norm": 0.24162162840366364, "learning_rate": 7.584416529179914e-05, "loss": 0.0267, "step": 28860 }, { "grad_norm": 0.5380929708480835, "learning_rate": 7.582646980253465e-05, "loss": 0.0296, "step": 28870 }, { "grad_norm": 0.3341025114059448, "learning_rate": 7.580876990023282e-05, "loss": 0.036, "step": 28880 }, { "grad_norm": 0.345841646194458, "learning_rate": 7.579106558791809e-05, "loss": 0.0231, "step": 28890 }, { "grad_norm": 0.2678109407424927, "learning_rate": 7.577335686861565e-05, "loss": 0.0281, "step": 28900 }, { "grad_norm": 0.36713582277297974, "learning_rate": 7.575564374535141e-05, "loss": 0.0275, "step": 28910 }, { "grad_norm": 0.31488120555877686, "learning_rate": 7.573792622115207e-05, "loss": 0.0198, "step": 28920 }, { "grad_norm": 0.27426883578300476, "learning_rate": 7.572020429904507e-05, "loss": 0.0245, "step": 28930 }, { "grad_norm": 0.5308612585067749, "learning_rate": 7.570247798205861e-05, "loss": 0.0294, "step": 28940 }, { "grad_norm": 0.5246650576591492, "learning_rate": 7.568474727322164e-05, "loss": 0.0328, "step": 28950 }, { "grad_norm": 0.32201340794563293, "learning_rate": 7.566701217556384e-05, "loss": 0.0275, "step": 28960 }, { "grad_norm": 0.46820759773254395, "learning_rate": 7.564927269211564e-05, "loss": 0.029, "step": 28970 }, { "grad_norm": 0.23780500888824463, "learning_rate": 7.563152882590824e-05, "loss": 0.0204, "step": 28980 }, { "grad_norm": 0.2666490375995636, "learning_rate": 7.56137805799736e-05, "loss": 0.0244, "step": 28990 }, { "grad_norm": 0.361039936542511, "learning_rate": 7.559602795734439e-05, "loss": 0.0271, "step": 29000 }, { "grad_norm": 0.2920312285423279, "learning_rate": 7.557827096105408e-05, "loss": 0.0242, "step": 29010 }, { "grad_norm": 0.3527258038520813, "learning_rate": 7.55605095941368e-05, "loss": 0.0295, "step": 29020 }, { "grad_norm": 0.30911776423454285, "learning_rate": 7.554274385962752e-05, "loss": 0.02, "step": 29030 }, { "grad_norm": 0.39285019040107727, "learning_rate": 7.552497376056191e-05, "loss": 0.0284, "step": 29040 }, { "grad_norm": 0.345548540353775, "learning_rate": 7.550719929997639e-05, "loss": 0.0266, "step": 29050 }, { "grad_norm": 0.37805140018463135, "learning_rate": 7.548942048090813e-05, "loss": 0.0297, "step": 29060 }, { "grad_norm": 0.3004775047302246, "learning_rate": 7.547163730639506e-05, "loss": 0.0234, "step": 29070 }, { "grad_norm": 0.24541613459587097, "learning_rate": 7.545384977947583e-05, "loss": 0.0228, "step": 29080 }, { "grad_norm": 0.31669846177101135, "learning_rate": 7.543605790318981e-05, "loss": 0.0212, "step": 29090 }, { "grad_norm": 0.2681392729282379, "learning_rate": 7.54182616805772e-05, "loss": 0.0283, "step": 29100 }, { "grad_norm": 0.24364621937274933, "learning_rate": 7.540046111467885e-05, "loss": 0.0194, "step": 29110 }, { "grad_norm": 0.2893572449684143, "learning_rate": 7.53826562085364e-05, "loss": 0.0225, "step": 29120 }, { "grad_norm": 0.2771414518356323, "learning_rate": 7.536484696519221e-05, "loss": 0.0167, "step": 29130 }, { "grad_norm": 0.39308860898017883, "learning_rate": 7.534703338768942e-05, "loss": 0.024, "step": 29140 }, { "grad_norm": 0.2656644582748413, "learning_rate": 7.532921547907185e-05, "loss": 0.0253, "step": 29150 }, { "grad_norm": 0.29489636421203613, "learning_rate": 7.531139324238412e-05, "loss": 0.0296, "step": 29160 }, { "grad_norm": 0.30485203862190247, "learning_rate": 7.529356668067157e-05, "loss": 0.0306, "step": 29170 }, { "grad_norm": 0.3454568088054657, "learning_rate": 7.527573579698023e-05, "loss": 0.0233, "step": 29180 }, { "grad_norm": 0.3921933174133301, "learning_rate": 7.525790059435693e-05, "loss": 0.0233, "step": 29190 }, { "grad_norm": 0.25188761949539185, "learning_rate": 7.524006107584926e-05, "loss": 0.0263, "step": 29200 }, { "grad_norm": 0.40557652711868286, "learning_rate": 7.522221724450544e-05, "loss": 0.0232, "step": 29210 }, { "grad_norm": 0.197922945022583, "learning_rate": 7.520436910337451e-05, "loss": 0.0352, "step": 29220 }, { "grad_norm": 0.2756437361240387, "learning_rate": 7.518651665550627e-05, "loss": 0.0257, "step": 29230 }, { "grad_norm": 0.24410811066627502, "learning_rate": 7.516865990395117e-05, "loss": 0.0375, "step": 29240 }, { "grad_norm": 0.4135983884334564, "learning_rate": 7.515079885176047e-05, "loss": 0.0203, "step": 29250 }, { "grad_norm": 0.24537144601345062, "learning_rate": 7.513293350198612e-05, "loss": 0.0236, "step": 29260 }, { "grad_norm": 0.363786906003952, "learning_rate": 7.511506385768081e-05, "loss": 0.0252, "step": 29270 }, { "grad_norm": 0.3258274793624878, "learning_rate": 7.509718992189801e-05, "loss": 0.0283, "step": 29280 }, { "grad_norm": 0.5529504418373108, "learning_rate": 7.507931169769182e-05, "loss": 0.0333, "step": 29290 }, { "grad_norm": 0.4247831404209137, "learning_rate": 7.506142918811722e-05, "loss": 0.0305, "step": 29300 }, { "grad_norm": 0.2687915861606598, "learning_rate": 7.504354239622978e-05, "loss": 0.0378, "step": 29310 }, { "grad_norm": 0.3069445490837097, "learning_rate": 7.50256513250859e-05, "loss": 0.031, "step": 29320 }, { "grad_norm": 0.21177703142166138, "learning_rate": 7.500775597774265e-05, "loss": 0.0251, "step": 29330 }, { "grad_norm": 0.29791924357414246, "learning_rate": 7.498985635725788e-05, "loss": 0.0227, "step": 29340 }, { "grad_norm": 0.40826550126075745, "learning_rate": 7.497195246669012e-05, "loss": 0.0319, "step": 29350 }, { "grad_norm": 0.33337247371673584, "learning_rate": 7.495404430909868e-05, "loss": 0.0257, "step": 29360 }, { "grad_norm": 0.4294278621673584, "learning_rate": 7.493613188754356e-05, "loss": 0.029, "step": 29370 }, { "grad_norm": 0.3789740204811096, "learning_rate": 7.49182152050855e-05, "loss": 0.0274, "step": 29380 }, { "grad_norm": 0.2970108389854431, "learning_rate": 7.490029426478598e-05, "loss": 0.0273, "step": 29390 }, { "grad_norm": 0.25577542185783386, "learning_rate": 7.488236906970719e-05, "loss": 0.0256, "step": 29400 }, { "grad_norm": 0.4257349967956543, "learning_rate": 7.486443962291207e-05, "loss": 0.0261, "step": 29410 }, { "grad_norm": 0.4093496799468994, "learning_rate": 7.484650592746424e-05, "loss": 0.0234, "step": 29420 }, { "grad_norm": 0.31438252329826355, "learning_rate": 7.482856798642811e-05, "loss": 0.0221, "step": 29430 }, { "grad_norm": 0.36260485649108887, "learning_rate": 7.481062580286878e-05, "loss": 0.0281, "step": 29440 }, { "grad_norm": 0.37101006507873535, "learning_rate": 7.479267937985208e-05, "loss": 0.0246, "step": 29450 }, { "grad_norm": 0.33974340558052063, "learning_rate": 7.477472872044456e-05, "loss": 0.0245, "step": 29460 }, { "grad_norm": 0.29779741168022156, "learning_rate": 7.475677382771347e-05, "loss": 0.0272, "step": 29470 }, { "grad_norm": 0.3109644949436188, "learning_rate": 7.473881470472683e-05, "loss": 0.033, "step": 29480 }, { "grad_norm": 0.5696395635604858, "learning_rate": 7.472085135455337e-05, "loss": 0.0298, "step": 29490 }, { "grad_norm": 0.3569173812866211, "learning_rate": 7.470288378026256e-05, "loss": 0.0326, "step": 29500 }, { "grad_norm": 0.25062695145606995, "learning_rate": 7.468491198492451e-05, "loss": 0.0277, "step": 29510 }, { "grad_norm": 0.29460635781288147, "learning_rate": 7.466693597161013e-05, "loss": 0.0263, "step": 29520 }, { "grad_norm": 0.3099108338356018, "learning_rate": 7.464895574339104e-05, "loss": 0.0278, "step": 29530 }, { "grad_norm": 0.16476313769817352, "learning_rate": 7.463097130333958e-05, "loss": 0.041, "step": 29540 }, { "grad_norm": 0.3310697376728058, "learning_rate": 7.461298265452876e-05, "loss": 0.0282, "step": 29550 }, { "grad_norm": 0.16688846051692963, "learning_rate": 7.459498980003239e-05, "loss": 0.0184, "step": 29560 }, { "grad_norm": 0.260671466588974, "learning_rate": 7.457699274292493e-05, "loss": 0.0273, "step": 29570 }, { "grad_norm": 0.37632209062576294, "learning_rate": 7.455899148628159e-05, "loss": 0.031, "step": 29580 }, { "grad_norm": 0.32264167070388794, "learning_rate": 7.45409860331783e-05, "loss": 0.0235, "step": 29590 }, { "grad_norm": 0.4750022888183594, "learning_rate": 7.452297638669169e-05, "loss": 0.0237, "step": 29600 }, { "grad_norm": 0.42955610156059265, "learning_rate": 7.450496254989911e-05, "loss": 0.0168, "step": 29610 }, { "grad_norm": 0.3321113586425781, "learning_rate": 7.448694452587866e-05, "loss": 0.0232, "step": 29620 }, { "grad_norm": 0.2295374870300293, "learning_rate": 7.44689223177091e-05, "loss": 0.025, "step": 29630 }, { "grad_norm": 0.3949214518070221, "learning_rate": 7.445089592846994e-05, "loss": 0.0227, "step": 29640 }, { "grad_norm": 0.20243693888187408, "learning_rate": 7.443286536124141e-05, "loss": 0.024, "step": 29650 }, { "grad_norm": 0.3243148624897003, "learning_rate": 7.441483061910443e-05, "loss": 0.0241, "step": 29660 }, { "grad_norm": 0.3823709487915039, "learning_rate": 7.439679170514064e-05, "loss": 0.0297, "step": 29670 }, { "grad_norm": 0.24014315009117126, "learning_rate": 7.43787486224324e-05, "loss": 0.0248, "step": 29680 }, { "grad_norm": 0.3253510296344757, "learning_rate": 7.436070137406276e-05, "loss": 0.0299, "step": 29690 }, { "grad_norm": 0.4951140284538269, "learning_rate": 7.434264996311556e-05, "loss": 0.0178, "step": 29700 }, { "grad_norm": 0.3555349111557007, "learning_rate": 7.432459439267525e-05, "loss": 0.0228, "step": 29710 }, { "grad_norm": 0.23336590826511383, "learning_rate": 7.430653466582701e-05, "loss": 0.0215, "step": 29720 }, { "grad_norm": 0.2786698639392853, "learning_rate": 7.42884707856568e-05, "loss": 0.0309, "step": 29730 }, { "grad_norm": 0.27395451068878174, "learning_rate": 7.427040275525122e-05, "loss": 0.0215, "step": 29740 }, { "grad_norm": 0.26054897904396057, "learning_rate": 7.42523305776976e-05, "loss": 0.0281, "step": 29750 }, { "grad_norm": 0.21822232007980347, "learning_rate": 7.4234254256084e-05, "loss": 0.021, "step": 29760 }, { "grad_norm": 0.3573191463947296, "learning_rate": 7.421617379349915e-05, "loss": 0.0289, "step": 29770 }, { "grad_norm": 0.2733097970485687, "learning_rate": 7.41980891930325e-05, "loss": 0.0249, "step": 29780 }, { "grad_norm": 0.2600206434726715, "learning_rate": 7.418000045777425e-05, "loss": 0.0207, "step": 29790 }, { "grad_norm": 0.4391912817955017, "learning_rate": 7.416190759081523e-05, "loss": 0.0216, "step": 29800 }, { "grad_norm": 0.32124781608581543, "learning_rate": 7.414381059524704e-05, "loss": 0.0206, "step": 29810 }, { "grad_norm": 0.28676608204841614, "learning_rate": 7.412570947416195e-05, "loss": 0.0215, "step": 29820 }, { "grad_norm": 0.4819580316543579, "learning_rate": 7.410760423065295e-05, "loss": 0.0357, "step": 29830 }, { "grad_norm": 0.34574297070503235, "learning_rate": 7.408949486781372e-05, "loss": 0.0335, "step": 29840 }, { "grad_norm": 0.32982343435287476, "learning_rate": 7.407138138873868e-05, "loss": 0.0303, "step": 29850 }, { "grad_norm": 0.223562091588974, "learning_rate": 7.405326379652292e-05, "loss": 0.0215, "step": 29860 }, { "grad_norm": 0.3009975552558899, "learning_rate": 7.403514209426222e-05, "loss": 0.0187, "step": 29870 }, { "grad_norm": 0.3646843731403351, "learning_rate": 7.40170162850531e-05, "loss": 0.0285, "step": 29880 }, { "grad_norm": 0.35006263852119446, "learning_rate": 7.399888637199278e-05, "loss": 0.0243, "step": 29890 }, { "grad_norm": 0.41403982043266296, "learning_rate": 7.398075235817914e-05, "loss": 0.0367, "step": 29900 }, { "grad_norm": 0.3687722980976105, "learning_rate": 7.39626142467108e-05, "loss": 0.0245, "step": 29910 }, { "grad_norm": 0.40248867869377136, "learning_rate": 7.394447204068706e-05, "loss": 0.0292, "step": 29920 }, { "grad_norm": 0.4771430790424347, "learning_rate": 7.392632574320793e-05, "loss": 0.0349, "step": 29930 }, { "grad_norm": 0.30231162905693054, "learning_rate": 7.390817535737411e-05, "loss": 0.0218, "step": 29940 }, { "grad_norm": 0.2880362868309021, "learning_rate": 7.389002088628703e-05, "loss": 0.027, "step": 29950 }, { "grad_norm": 0.26903021335601807, "learning_rate": 7.387186233304877e-05, "loss": 0.0306, "step": 29960 }, { "grad_norm": 0.2647477090358734, "learning_rate": 7.385369970076212e-05, "loss": 0.0249, "step": 29970 }, { "grad_norm": 0.33786576986312866, "learning_rate": 7.38355329925306e-05, "loss": 0.0293, "step": 29980 }, { "grad_norm": 0.31935253739356995, "learning_rate": 7.381736221145838e-05, "loss": 0.0298, "step": 29990 }, { "grad_norm": 0.30598974227905273, "learning_rate": 7.37991873606504e-05, "loss": 0.0218, "step": 30000 }, { "grad_norm": 0.2548796534538269, "learning_rate": 7.378100844321218e-05, "loss": 0.0282, "step": 30010 }, { "grad_norm": 0.383817195892334, "learning_rate": 7.376282546225004e-05, "loss": 0.0229, "step": 30020 }, { "grad_norm": 0.2907787263393402, "learning_rate": 7.374463842087094e-05, "loss": 0.0225, "step": 30030 }, { "grad_norm": 0.3771417438983917, "learning_rate": 7.372644732218254e-05, "loss": 0.0237, "step": 30040 }, { "grad_norm": 0.43065598607063293, "learning_rate": 7.370825216929322e-05, "loss": 0.0322, "step": 30050 }, { "grad_norm": 0.3134291470050812, "learning_rate": 7.369005296531205e-05, "loss": 0.0283, "step": 30060 }, { "grad_norm": 0.3120073080062866, "learning_rate": 7.367184971334873e-05, "loss": 0.0254, "step": 30070 }, { "grad_norm": 0.28943565487861633, "learning_rate": 7.365364241651371e-05, "loss": 0.0333, "step": 30080 }, { "grad_norm": 0.3216521143913269, "learning_rate": 7.363543107791815e-05, "loss": 0.025, "step": 30090 }, { "grad_norm": 0.5298242568969727, "learning_rate": 7.361721570067384e-05, "loss": 0.0268, "step": 30100 }, { "grad_norm": 0.4520522356033325, "learning_rate": 7.359899628789331e-05, "loss": 0.0303, "step": 30110 }, { "grad_norm": 0.34405702352523804, "learning_rate": 7.358077284268974e-05, "loss": 0.0305, "step": 30120 }, { "grad_norm": 0.552922785282135, "learning_rate": 7.356254536817702e-05, "loss": 0.0289, "step": 30130 }, { "grad_norm": 0.36726078391075134, "learning_rate": 7.354431386746973e-05, "loss": 0.0288, "step": 30140 }, { "grad_norm": 0.6919853091239929, "learning_rate": 7.352607834368316e-05, "loss": 0.0274, "step": 30150 }, { "grad_norm": 0.24707818031311035, "learning_rate": 7.350783879993324e-05, "loss": 0.0272, "step": 30160 }, { "grad_norm": 0.3278075158596039, "learning_rate": 7.348959523933658e-05, "loss": 0.0244, "step": 30170 }, { "grad_norm": 0.2890552580356598, "learning_rate": 7.347134766501057e-05, "loss": 0.0223, "step": 30180 }, { "grad_norm": 0.30312439799308777, "learning_rate": 7.345309608007315e-05, "loss": 0.0243, "step": 30190 }, { "grad_norm": 0.22124481201171875, "learning_rate": 7.343484048764308e-05, "loss": 0.0256, "step": 30200 }, { "grad_norm": 0.3289332985877991, "learning_rate": 7.341658089083972e-05, "loss": 0.024, "step": 30210 }, { "grad_norm": 0.25635600090026855, "learning_rate": 7.339831729278313e-05, "loss": 0.0257, "step": 30220 }, { "grad_norm": 0.4044071435928345, "learning_rate": 7.338004969659404e-05, "loss": 0.0249, "step": 30230 }, { "grad_norm": 0.3759373724460602, "learning_rate": 7.336177810539391e-05, "loss": 0.027, "step": 30240 }, { "grad_norm": 0.41928693652153015, "learning_rate": 7.334350252230485e-05, "loss": 0.0223, "step": 30250 }, { "grad_norm": 0.3007175624370575, "learning_rate": 7.332522295044965e-05, "loss": 0.0171, "step": 30260 }, { "grad_norm": 0.4120174050331116, "learning_rate": 7.33069393929518e-05, "loss": 0.0236, "step": 30270 }, { "grad_norm": 0.2898833751678467, "learning_rate": 7.328865185293545e-05, "loss": 0.0369, "step": 30280 }, { "grad_norm": 0.4148879945278168, "learning_rate": 7.327036033352546e-05, "loss": 0.0284, "step": 30290 }, { "grad_norm": 0.2753335237503052, "learning_rate": 7.325206483784733e-05, "loss": 0.0173, "step": 30300 }, { "grad_norm": 0.3010120391845703, "learning_rate": 7.323376536902724e-05, "loss": 0.0373, "step": 30310 }, { "grad_norm": 0.3413354754447937, "learning_rate": 7.321546193019213e-05, "loss": 0.028, "step": 30320 }, { "grad_norm": 0.351171612739563, "learning_rate": 7.31971545244695e-05, "loss": 0.0302, "step": 30330 }, { "grad_norm": 0.30035167932510376, "learning_rate": 7.31788431549876e-05, "loss": 0.0243, "step": 30340 }, { "grad_norm": 0.2724401652812958, "learning_rate": 7.316052782487534e-05, "loss": 0.0259, "step": 30350 }, { "grad_norm": 0.6367337107658386, "learning_rate": 7.314220853726234e-05, "loss": 0.032, "step": 30360 }, { "grad_norm": 0.3978276252746582, "learning_rate": 7.312388529527884e-05, "loss": 0.0282, "step": 30370 }, { "grad_norm": 0.3999420702457428, "learning_rate": 7.310555810205577e-05, "loss": 0.0285, "step": 30380 }, { "grad_norm": 0.2755468785762787, "learning_rate": 7.308722696072476e-05, "loss": 0.0285, "step": 30390 }, { "grad_norm": 0.32368361949920654, "learning_rate": 7.306889187441811e-05, "loss": 0.0253, "step": 30400 }, { "grad_norm": 0.37782925367355347, "learning_rate": 7.305055284626876e-05, "loss": 0.0179, "step": 30410 }, { "grad_norm": 0.313466876745224, "learning_rate": 7.303220987941037e-05, "loss": 0.024, "step": 30420 }, { "grad_norm": 0.351240336894989, "learning_rate": 7.301386297697726e-05, "loss": 0.0294, "step": 30430 }, { "grad_norm": 0.3363878130912781, "learning_rate": 7.299551214210438e-05, "loss": 0.0246, "step": 30440 }, { "grad_norm": 0.23966234922409058, "learning_rate": 7.297715737792738e-05, "loss": 0.0314, "step": 30450 }, { "grad_norm": 0.37822553515434265, "learning_rate": 7.295879868758265e-05, "loss": 0.0262, "step": 30460 }, { "grad_norm": 0.3143341839313507, "learning_rate": 7.294043607420713e-05, "loss": 0.025, "step": 30470 }, { "grad_norm": 0.4808027744293213, "learning_rate": 7.292206954093852e-05, "loss": 0.0241, "step": 30480 }, { "grad_norm": 0.3036462366580963, "learning_rate": 7.290369909091515e-05, "loss": 0.0247, "step": 30490 }, { "grad_norm": 0.7739918828010559, "learning_rate": 7.2885324727276e-05, "loss": 0.0236, "step": 30500 }, { "grad_norm": 0.29296422004699707, "learning_rate": 7.286694645316076e-05, "loss": 0.0232, "step": 30510 }, { "grad_norm": 0.33301031589508057, "learning_rate": 7.284856427170982e-05, "loss": 0.0252, "step": 30520 }, { "grad_norm": 0.3318466544151306, "learning_rate": 7.283017818606414e-05, "loss": 0.0286, "step": 30530 }, { "grad_norm": 0.39393338561058044, "learning_rate": 7.28117881993654e-05, "loss": 0.0329, "step": 30540 }, { "grad_norm": 0.5406108498573303, "learning_rate": 7.279339431475598e-05, "loss": 0.0284, "step": 30550 }, { "grad_norm": 0.27809077501296997, "learning_rate": 7.277499653537887e-05, "loss": 0.0243, "step": 30560 }, { "grad_norm": 0.4341135025024414, "learning_rate": 7.275659486437776e-05, "loss": 0.0302, "step": 30570 }, { "grad_norm": 0.2724534869194031, "learning_rate": 7.273818930489695e-05, "loss": 0.0359, "step": 30580 }, { "grad_norm": 0.29637762904167175, "learning_rate": 7.271977986008151e-05, "loss": 0.0267, "step": 30590 }, { "grad_norm": 0.40464693307876587, "learning_rate": 7.270136653307705e-05, "loss": 0.0236, "step": 30600 }, { "grad_norm": 0.3424641191959381, "learning_rate": 7.268294932702994e-05, "loss": 0.0282, "step": 30610 }, { "grad_norm": 0.35460713505744934, "learning_rate": 7.266452824508719e-05, "loss": 0.0262, "step": 30620 }, { "grad_norm": 0.38738882541656494, "learning_rate": 7.264610329039643e-05, "loss": 0.0214, "step": 30630 }, { "grad_norm": 0.3195997476577759, "learning_rate": 7.262767446610599e-05, "loss": 0.0231, "step": 30640 }, { "grad_norm": 0.38145390152931213, "learning_rate": 7.260924177536485e-05, "loss": 0.0327, "step": 30650 }, { "grad_norm": 0.303005188703537, "learning_rate": 7.259080522132265e-05, "loss": 0.0203, "step": 30660 }, { "grad_norm": 0.2680114209651947, "learning_rate": 7.257236480712972e-05, "loss": 0.0236, "step": 30670 }, { "grad_norm": 0.33077430725097656, "learning_rate": 7.255392053593697e-05, "loss": 0.0233, "step": 30680 }, { "grad_norm": 0.37080079317092896, "learning_rate": 7.253547241089607e-05, "loss": 0.0179, "step": 30690 }, { "grad_norm": 0.2864413261413574, "learning_rate": 7.251702043515927e-05, "loss": 0.0281, "step": 30700 }, { "grad_norm": 0.23921269178390503, "learning_rate": 7.249856461187952e-05, "loss": 0.0252, "step": 30710 }, { "grad_norm": 0.33829912543296814, "learning_rate": 7.248010494421042e-05, "loss": 0.0244, "step": 30720 }, { "grad_norm": 0.36622312664985657, "learning_rate": 7.246164143530622e-05, "loss": 0.0347, "step": 30730 }, { "grad_norm": 0.29213979840278625, "learning_rate": 7.244317408832181e-05, "loss": 0.015, "step": 30740 }, { "grad_norm": 0.5099638104438782, "learning_rate": 7.242470290641279e-05, "loss": 0.0252, "step": 30750 }, { "grad_norm": 0.2890555262565613, "learning_rate": 7.240622789273536e-05, "loss": 0.0208, "step": 30760 }, { "grad_norm": 0.33059918880462646, "learning_rate": 7.238774905044638e-05, "loss": 0.0219, "step": 30770 }, { "grad_norm": 0.35070890188217163, "learning_rate": 7.236926638270341e-05, "loss": 0.0277, "step": 30780 }, { "grad_norm": 0.563237726688385, "learning_rate": 7.23507798926646e-05, "loss": 0.0288, "step": 30790 }, { "grad_norm": 0.29466673731803894, "learning_rate": 7.23322895834888e-05, "loss": 0.0234, "step": 30800 }, { "grad_norm": 0.3488032817840576, "learning_rate": 7.231379545833552e-05, "loss": 0.029, "step": 30810 }, { "grad_norm": 0.3241041600704193, "learning_rate": 7.229529752036487e-05, "loss": 0.0217, "step": 30820 }, { "grad_norm": 0.26098722219467163, "learning_rate": 7.227679577273765e-05, "loss": 0.0295, "step": 30830 }, { "grad_norm": 0.430755615234375, "learning_rate": 7.225829021861529e-05, "loss": 0.0242, "step": 30840 }, { "grad_norm": 0.23472923040390015, "learning_rate": 7.223978086115992e-05, "loss": 0.0232, "step": 30850 }, { "grad_norm": 0.2990337014198303, "learning_rate": 7.222126770353425e-05, "loss": 0.0303, "step": 30860 }, { "grad_norm": 0.3025970160961151, "learning_rate": 7.22027507489017e-05, "loss": 0.0249, "step": 30870 }, { "grad_norm": 0.39996710419654846, "learning_rate": 7.218423000042627e-05, "loss": 0.0275, "step": 30880 }, { "grad_norm": 0.3232648968696594, "learning_rate": 7.216570546127268e-05, "loss": 0.0313, "step": 30890 }, { "grad_norm": 0.3512076139450073, "learning_rate": 7.214717713460626e-05, "loss": 0.0274, "step": 30900 }, { "grad_norm": 0.47285738587379456, "learning_rate": 7.2128645023593e-05, "loss": 0.0294, "step": 30910 }, { "grad_norm": 0.27136439085006714, "learning_rate": 7.211010913139951e-05, "loss": 0.0332, "step": 30920 }, { "grad_norm": 0.23294827342033386, "learning_rate": 7.209156946119308e-05, "loss": 0.0217, "step": 30930 }, { "grad_norm": 0.5219475030899048, "learning_rate": 7.207302601614166e-05, "loss": 0.0299, "step": 30940 }, { "grad_norm": 0.2353660762310028, "learning_rate": 7.205447879941378e-05, "loss": 0.0252, "step": 30950 }, { "grad_norm": 0.34261801838874817, "learning_rate": 7.203592781417866e-05, "loss": 0.0218, "step": 30960 }, { "grad_norm": 0.3687830865383148, "learning_rate": 7.201737306360617e-05, "loss": 0.0262, "step": 30970 }, { "grad_norm": 0.32974007725715637, "learning_rate": 7.19988145508668e-05, "loss": 0.0208, "step": 30980 }, { "grad_norm": 0.3322402536869049, "learning_rate": 7.198025227913168e-05, "loss": 0.0212, "step": 30990 }, { "grad_norm": 0.4145219027996063, "learning_rate": 7.196168625157261e-05, "loss": 0.0248, "step": 31000 }, { "grad_norm": 0.2504025399684906, "learning_rate": 7.194311647136201e-05, "loss": 0.0225, "step": 31010 }, { "grad_norm": 0.29777801036834717, "learning_rate": 7.192454294167297e-05, "loss": 0.0183, "step": 31020 }, { "grad_norm": 0.3529236316680908, "learning_rate": 7.190596566567917e-05, "loss": 0.0235, "step": 31030 }, { "grad_norm": 0.2743002772331238, "learning_rate": 7.188738464655496e-05, "loss": 0.0222, "step": 31040 }, { "grad_norm": 0.20403428375720978, "learning_rate": 7.186879988747533e-05, "loss": 0.0194, "step": 31050 }, { "grad_norm": 0.29039859771728516, "learning_rate": 7.185021139161592e-05, "loss": 0.0227, "step": 31060 }, { "grad_norm": 0.403800368309021, "learning_rate": 7.1831619162153e-05, "loss": 0.028, "step": 31070 }, { "grad_norm": 0.3590596318244934, "learning_rate": 7.181302320226345e-05, "loss": 0.0219, "step": 31080 }, { "grad_norm": 0.35121944546699524, "learning_rate": 7.179442351512482e-05, "loss": 0.0293, "step": 31090 }, { "grad_norm": 0.2595391869544983, "learning_rate": 7.177582010391528e-05, "loss": 0.023, "step": 31100 }, { "grad_norm": 0.3885224163532257, "learning_rate": 7.175721297181366e-05, "loss": 0.0263, "step": 31110 }, { "grad_norm": 0.4096481204032898, "learning_rate": 7.173860212199942e-05, "loss": 0.0245, "step": 31120 }, { "grad_norm": 0.3472417891025543, "learning_rate": 7.171998755765263e-05, "loss": 0.0267, "step": 31130 }, { "grad_norm": 0.4092135429382324, "learning_rate": 7.170136928195398e-05, "loss": 0.0234, "step": 31140 }, { "grad_norm": 0.32887041568756104, "learning_rate": 7.168274729808489e-05, "loss": 0.0239, "step": 31150 }, { "grad_norm": 0.2402181327342987, "learning_rate": 7.166412160922728e-05, "loss": 0.0218, "step": 31160 }, { "grad_norm": 0.378696471452713, "learning_rate": 7.164549221856382e-05, "loss": 0.0214, "step": 31170 }, { "grad_norm": 0.26570406556129456, "learning_rate": 7.162685912927775e-05, "loss": 0.0283, "step": 31180 }, { "grad_norm": 0.3811337351799011, "learning_rate": 7.160822234455294e-05, "loss": 0.025, "step": 31190 }, { "grad_norm": 0.2179296314716339, "learning_rate": 7.158958186757391e-05, "loss": 0.0168, "step": 31200 }, { "grad_norm": 0.48555538058280945, "learning_rate": 7.157093770152582e-05, "loss": 0.0195, "step": 31210 }, { "grad_norm": 0.3197055757045746, "learning_rate": 7.155228984959446e-05, "loss": 0.0248, "step": 31220 }, { "grad_norm": 0.27671635150909424, "learning_rate": 7.153363831496621e-05, "loss": 0.0269, "step": 31230 }, { "grad_norm": 0.3876916468143463, "learning_rate": 7.151498310082811e-05, "loss": 0.0318, "step": 31240 }, { "grad_norm": 0.34080585837364197, "learning_rate": 7.149632421036784e-05, "loss": 0.031, "step": 31250 }, { "grad_norm": 0.39944618940353394, "learning_rate": 7.147766164677369e-05, "loss": 0.0408, "step": 31260 }, { "grad_norm": 0.28249993920326233, "learning_rate": 7.145899541323459e-05, "loss": 0.0186, "step": 31270 }, { "grad_norm": 0.4020414650440216, "learning_rate": 7.144032551294007e-05, "loss": 0.0334, "step": 31280 }, { "grad_norm": 0.48253417015075684, "learning_rate": 7.14216519490803e-05, "loss": 0.0295, "step": 31290 }, { "grad_norm": 0.583311915397644, "learning_rate": 7.140297472484609e-05, "loss": 0.0391, "step": 31300 }, { "grad_norm": 0.32244518399238586, "learning_rate": 7.138429384342891e-05, "loss": 0.042, "step": 31310 }, { "grad_norm": 0.4215910732746124, "learning_rate": 7.136560930802074e-05, "loss": 0.0294, "step": 31320 }, { "grad_norm": 0.5064582228660583, "learning_rate": 7.134692112181431e-05, "loss": 0.0316, "step": 31330 }, { "grad_norm": 0.32756707072257996, "learning_rate": 7.13282292880029e-05, "loss": 0.0333, "step": 31340 }, { "grad_norm": 0.3159671723842621, "learning_rate": 7.130953380978043e-05, "loss": 0.0273, "step": 31350 }, { "grad_norm": 0.43985578417778015, "learning_rate": 7.129083469034144e-05, "loss": 0.0363, "step": 31360 }, { "grad_norm": 0.40183448791503906, "learning_rate": 7.127213193288112e-05, "loss": 0.0283, "step": 31370 }, { "grad_norm": 0.26717013120651245, "learning_rate": 7.125342554059522e-05, "loss": 0.0226, "step": 31380 }, { "grad_norm": 0.48277780413627625, "learning_rate": 7.12347155166802e-05, "loss": 0.0313, "step": 31390 }, { "grad_norm": 0.268635094165802, "learning_rate": 7.121600186433306e-05, "loss": 0.0277, "step": 31400 }, { "grad_norm": 0.2884058952331543, "learning_rate": 7.119728458675148e-05, "loss": 0.0221, "step": 31410 }, { "grad_norm": 0.3570469915866852, "learning_rate": 7.117856368713369e-05, "loss": 0.0211, "step": 31420 }, { "grad_norm": 0.4031667709350586, "learning_rate": 7.115983916867861e-05, "loss": 0.0278, "step": 31430 }, { "grad_norm": 0.40227189660072327, "learning_rate": 7.114111103458574e-05, "loss": 0.023, "step": 31440 }, { "grad_norm": 0.34717991948127747, "learning_rate": 7.11223792880552e-05, "loss": 0.0177, "step": 31450 }, { "grad_norm": 0.41192564368247986, "learning_rate": 7.110364393228773e-05, "loss": 0.0238, "step": 31460 }, { "grad_norm": 0.22912442684173584, "learning_rate": 7.108490497048471e-05, "loss": 0.0181, "step": 31470 }, { "grad_norm": 0.467103511095047, "learning_rate": 7.10661624058481e-05, "loss": 0.0227, "step": 31480 }, { "grad_norm": 0.2461460381746292, "learning_rate": 7.10474162415805e-05, "loss": 0.0223, "step": 31490 }, { "grad_norm": 0.2103404551744461, "learning_rate": 7.102866648088511e-05, "loss": 0.0219, "step": 31500 }, { "grad_norm": 0.3088257610797882, "learning_rate": 7.100991312696576e-05, "loss": 0.023, "step": 31510 }, { "grad_norm": 0.5652216672897339, "learning_rate": 7.099115618302686e-05, "loss": 0.0244, "step": 31520 }, { "grad_norm": 0.3893948495388031, "learning_rate": 7.097239565227349e-05, "loss": 0.0282, "step": 31530 }, { "grad_norm": 0.3070649206638336, "learning_rate": 7.09536315379113e-05, "loss": 0.0248, "step": 31540 }, { "grad_norm": 0.2355397790670395, "learning_rate": 7.093486384314656e-05, "loss": 0.0201, "step": 31550 }, { "grad_norm": 0.40504375100135803, "learning_rate": 7.091609257118616e-05, "loss": 0.0237, "step": 31560 }, { "grad_norm": 0.38874006271362305, "learning_rate": 7.08973177252376e-05, "loss": 0.0331, "step": 31570 }, { "grad_norm": 0.3238861858844757, "learning_rate": 7.087853930850898e-05, "loss": 0.0208, "step": 31580 }, { "grad_norm": 0.5204746723175049, "learning_rate": 7.085975732420903e-05, "loss": 0.0242, "step": 31590 }, { "grad_norm": 0.19521567225456238, "learning_rate": 7.084097177554706e-05, "loss": 0.0195, "step": 31600 }, { "grad_norm": 0.44364720582962036, "learning_rate": 7.082218266573301e-05, "loss": 0.0334, "step": 31610 }, { "grad_norm": 0.46594753861427307, "learning_rate": 7.080338999797743e-05, "loss": 0.0359, "step": 31620 }, { "grad_norm": 0.2606869637966156, "learning_rate": 7.07845937754915e-05, "loss": 0.0254, "step": 31630 }, { "grad_norm": 0.2763122022151947, "learning_rate": 7.076579400148693e-05, "loss": 0.0288, "step": 31640 }, { "grad_norm": 0.35166049003601074, "learning_rate": 7.074699067917611e-05, "loss": 0.0299, "step": 31650 }, { "grad_norm": 0.28095847368240356, "learning_rate": 7.072818381177201e-05, "loss": 0.0303, "step": 31660 }, { "grad_norm": 0.35630181431770325, "learning_rate": 7.070937340248823e-05, "loss": 0.0254, "step": 31670 }, { "grad_norm": 0.3399973511695862, "learning_rate": 7.069055945453893e-05, "loss": 0.0313, "step": 31680 }, { "grad_norm": 0.3656184673309326, "learning_rate": 7.067174197113892e-05, "loss": 0.0266, "step": 31690 }, { "grad_norm": 0.35921749472618103, "learning_rate": 7.065292095550355e-05, "loss": 0.0271, "step": 31700 }, { "grad_norm": 0.22777003049850464, "learning_rate": 7.063409641084887e-05, "loss": 0.0264, "step": 31710 }, { "grad_norm": 0.4050334393978119, "learning_rate": 7.061526834039145e-05, "loss": 0.0349, "step": 31720 }, { "grad_norm": 0.27220332622528076, "learning_rate": 7.05964367473485e-05, "loss": 0.0224, "step": 31730 }, { "grad_norm": 0.3719223439693451, "learning_rate": 7.057760163493783e-05, "loss": 0.0315, "step": 31740 }, { "grad_norm": 0.23501645028591156, "learning_rate": 7.055876300637783e-05, "loss": 0.0274, "step": 31750 }, { "grad_norm": 0.37679100036621094, "learning_rate": 7.053992086488753e-05, "loss": 0.0325, "step": 31760 }, { "grad_norm": 0.36470216512680054, "learning_rate": 7.052107521368651e-05, "loss": 0.0275, "step": 31770 }, { "grad_norm": 0.4841748774051666, "learning_rate": 7.0502226055995e-05, "loss": 0.033, "step": 31780 }, { "grad_norm": 0.2666904330253601, "learning_rate": 7.048337339503379e-05, "loss": 0.038, "step": 31790 }, { "grad_norm": 0.4620302617549896, "learning_rate": 7.046451723402427e-05, "loss": 0.025, "step": 31800 }, { "grad_norm": 0.4186922609806061, "learning_rate": 7.044565757618848e-05, "loss": 0.0343, "step": 31810 }, { "grad_norm": 0.2394506335258484, "learning_rate": 7.042679442474899e-05, "loss": 0.0271, "step": 31820 }, { "grad_norm": 0.24169208109378815, "learning_rate": 7.040792778292902e-05, "loss": 0.0369, "step": 31830 }, { "grad_norm": 0.35328084230422974, "learning_rate": 7.038905765395234e-05, "loss": 0.0359, "step": 31840 }, { "grad_norm": 0.44940927624702454, "learning_rate": 7.037018404104334e-05, "loss": 0.0279, "step": 31850 }, { "grad_norm": 0.4689325988292694, "learning_rate": 7.035130694742702e-05, "loss": 0.0232, "step": 31860 }, { "grad_norm": 0.36145010590553284, "learning_rate": 7.033242637632897e-05, "loss": 0.0265, "step": 31870 }, { "grad_norm": 0.4313502907752991, "learning_rate": 7.031354233097534e-05, "loss": 0.0281, "step": 31880 }, { "grad_norm": 0.29352545738220215, "learning_rate": 7.029465481459289e-05, "loss": 0.0353, "step": 31890 }, { "grad_norm": 0.2691177427768707, "learning_rate": 7.027576383040898e-05, "loss": 0.0251, "step": 31900 }, { "grad_norm": 0.3068411946296692, "learning_rate": 7.025686938165159e-05, "loss": 0.037, "step": 31910 }, { "grad_norm": 0.5218294262886047, "learning_rate": 7.023797147154924e-05, "loss": 0.0356, "step": 31920 }, { "grad_norm": 0.2890037000179291, "learning_rate": 7.021907010333111e-05, "loss": 0.0215, "step": 31930 }, { "grad_norm": 0.422232449054718, "learning_rate": 7.020016528022685e-05, "loss": 0.0257, "step": 31940 }, { "grad_norm": 0.2550339996814728, "learning_rate": 7.018125700546683e-05, "loss": 0.0295, "step": 31950 }, { "grad_norm": 0.381835013628006, "learning_rate": 7.016234528228196e-05, "loss": 0.0285, "step": 31960 }, { "grad_norm": 0.34268131852149963, "learning_rate": 7.014343011390372e-05, "loss": 0.0216, "step": 31970 }, { "grad_norm": 0.37429293990135193, "learning_rate": 7.01245115035642e-05, "loss": 0.0247, "step": 31980 }, { "grad_norm": 0.2766076624393463, "learning_rate": 7.010558945449606e-05, "loss": 0.0333, "step": 31990 }, { "grad_norm": 0.3091230094432831, "learning_rate": 7.008666396993258e-05, "loss": 0.02, "step": 32000 }, { "grad_norm": 0.4059317111968994, "learning_rate": 7.006773505310759e-05, "loss": 0.0312, "step": 32010 }, { "grad_norm": 0.3352925181388855, "learning_rate": 7.004880270725553e-05, "loss": 0.0281, "step": 32020 }, { "grad_norm": 0.3032725751399994, "learning_rate": 7.002986693561144e-05, "loss": 0.0242, "step": 32030 }, { "grad_norm": 0.3755210041999817, "learning_rate": 7.001092774141089e-05, "loss": 0.0271, "step": 32040 }, { "grad_norm": 0.31716427206993103, "learning_rate": 6.999198512789009e-05, "loss": 0.0282, "step": 32050 }, { "grad_norm": 0.4275621771812439, "learning_rate": 6.997303909828584e-05, "loss": 0.0252, "step": 32060 }, { "grad_norm": 0.29948267340660095, "learning_rate": 6.995408965583544e-05, "loss": 0.0242, "step": 32070 }, { "grad_norm": 0.31549111008644104, "learning_rate": 6.993513680377688e-05, "loss": 0.0175, "step": 32080 }, { "grad_norm": 0.2639622688293457, "learning_rate": 6.991618054534868e-05, "loss": 0.0289, "step": 32090 }, { "grad_norm": 0.3945445418357849, "learning_rate": 6.989722088378991e-05, "loss": 0.0253, "step": 32100 }, { "grad_norm": 0.39293330907821655, "learning_rate": 6.987825782234027e-05, "loss": 0.0279, "step": 32110 }, { "grad_norm": 0.4049125015735626, "learning_rate": 6.985929136424006e-05, "loss": 0.0254, "step": 32120 }, { "grad_norm": 0.41352763772010803, "learning_rate": 6.984032151273012e-05, "loss": 0.0226, "step": 32130 }, { "grad_norm": 0.3833019733428955, "learning_rate": 6.982134827105186e-05, "loss": 0.026, "step": 32140 }, { "grad_norm": 0.4516223669052124, "learning_rate": 6.980237164244729e-05, "loss": 0.0212, "step": 32150 }, { "grad_norm": 0.3183099329471588, "learning_rate": 6.9783391630159e-05, "loss": 0.0212, "step": 32160 }, { "grad_norm": 0.39084213972091675, "learning_rate": 6.976440823743015e-05, "loss": 0.0239, "step": 32170 }, { "grad_norm": 0.3177827298641205, "learning_rate": 6.974542146750451e-05, "loss": 0.0223, "step": 32180 }, { "grad_norm": 0.19151948392391205, "learning_rate": 6.972643132362637e-05, "loss": 0.0226, "step": 32190 }, { "grad_norm": 0.3998110890388489, "learning_rate": 6.970743780904064e-05, "loss": 0.034, "step": 32200 }, { "grad_norm": 0.371907114982605, "learning_rate": 6.968844092699277e-05, "loss": 0.0306, "step": 32210 }, { "grad_norm": 0.36035045981407166, "learning_rate": 6.966944068072883e-05, "loss": 0.0278, "step": 32220 }, { "grad_norm": 0.2796170115470886, "learning_rate": 6.965043707349545e-05, "loss": 0.018, "step": 32230 }, { "grad_norm": 0.25586050748825073, "learning_rate": 6.963143010853982e-05, "loss": 0.027, "step": 32240 }, { "grad_norm": 0.49407339096069336, "learning_rate": 6.961241978910971e-05, "loss": 0.0308, "step": 32250 }, { "grad_norm": 0.2527947723865509, "learning_rate": 6.959340611845344e-05, "loss": 0.0228, "step": 32260 }, { "grad_norm": 0.2571331262588501, "learning_rate": 6.957438909981995e-05, "loss": 0.0203, "step": 32270 }, { "grad_norm": 0.274501770734787, "learning_rate": 6.955536873645872e-05, "loss": 0.0155, "step": 32280 }, { "grad_norm": 0.21567006409168243, "learning_rate": 6.953634503161982e-05, "loss": 0.0198, "step": 32290 }, { "grad_norm": 0.341556578874588, "learning_rate": 6.951731798855387e-05, "loss": 0.0221, "step": 32300 }, { "grad_norm": 0.28155240416526794, "learning_rate": 6.949828761051208e-05, "loss": 0.0286, "step": 32310 }, { "grad_norm": 0.22608166933059692, "learning_rate": 6.947925390074622e-05, "loss": 0.0166, "step": 32320 }, { "grad_norm": 0.30507099628448486, "learning_rate": 6.946021686250863e-05, "loss": 0.0209, "step": 32330 }, { "grad_norm": 0.22997109591960907, "learning_rate": 6.94411764990522e-05, "loss": 0.0232, "step": 32340 }, { "grad_norm": 0.3032248318195343, "learning_rate": 6.942213281363044e-05, "loss": 0.0191, "step": 32350 }, { "grad_norm": 0.295215904712677, "learning_rate": 6.940308580949737e-05, "loss": 0.0205, "step": 32360 }, { "grad_norm": 0.21223197877407074, "learning_rate": 6.93840354899076e-05, "loss": 0.0174, "step": 32370 }, { "grad_norm": 0.30773136019706726, "learning_rate": 6.936498185811633e-05, "loss": 0.0238, "step": 32380 }, { "grad_norm": 0.21654240787029266, "learning_rate": 6.93459249173793e-05, "loss": 0.018, "step": 32390 }, { "grad_norm": 0.41694724559783936, "learning_rate": 6.932686467095279e-05, "loss": 0.0278, "step": 32400 }, { "grad_norm": 0.4786083400249481, "learning_rate": 6.930780112209373e-05, "loss": 0.0285, "step": 32410 }, { "grad_norm": 0.41697224974632263, "learning_rate": 6.92887342740595e-05, "loss": 0.027, "step": 32420 }, { "grad_norm": 0.30630406737327576, "learning_rate": 6.926966413010816e-05, "loss": 0.0319, "step": 32430 }, { "grad_norm": 0.28654763102531433, "learning_rate": 6.925059069349824e-05, "loss": 0.0213, "step": 32440 }, { "grad_norm": 0.27415117621421814, "learning_rate": 6.923151396748886e-05, "loss": 0.0235, "step": 32450 }, { "grad_norm": 0.3179130554199219, "learning_rate": 6.921243395533974e-05, "loss": 0.0203, "step": 32460 }, { "grad_norm": 0.40150952339172363, "learning_rate": 6.919335066031109e-05, "loss": 0.0299, "step": 32470 }, { "grad_norm": 0.2839738726615906, "learning_rate": 6.917426408566379e-05, "loss": 0.0255, "step": 32480 }, { "grad_norm": 0.2661172151565552, "learning_rate": 6.915517423465916e-05, "loss": 0.0231, "step": 32490 }, { "grad_norm": 0.4828955829143524, "learning_rate": 6.913608111055914e-05, "loss": 0.0251, "step": 32500 }, { "grad_norm": 0.28433895111083984, "learning_rate": 6.911698471662623e-05, "loss": 0.0229, "step": 32510 }, { "grad_norm": 0.27483367919921875, "learning_rate": 6.90978850561235e-05, "loss": 0.0207, "step": 32520 }, { "grad_norm": 0.22345423698425293, "learning_rate": 6.907878213231454e-05, "loss": 0.0197, "step": 32530 }, { "grad_norm": 0.5531277060508728, "learning_rate": 6.90596759484635e-05, "loss": 0.0261, "step": 32540 }, { "grad_norm": 0.39956560730934143, "learning_rate": 6.904056650783514e-05, "loss": 0.0313, "step": 32550 }, { "grad_norm": 0.34791916608810425, "learning_rate": 6.902145381369471e-05, "loss": 0.0247, "step": 32560 }, { "grad_norm": 0.1686495691537857, "learning_rate": 6.900233786930808e-05, "loss": 0.0224, "step": 32570 }, { "grad_norm": 0.33924081921577454, "learning_rate": 6.898321867794161e-05, "loss": 0.0222, "step": 32580 }, { "grad_norm": 0.30967283248901367, "learning_rate": 6.896409624286226e-05, "loss": 0.0245, "step": 32590 }, { "grad_norm": 0.2789785861968994, "learning_rate": 6.894497056733754e-05, "loss": 0.0239, "step": 32600 }, { "grad_norm": 0.40063029527664185, "learning_rate": 6.89258416546355e-05, "loss": 0.0209, "step": 32610 }, { "grad_norm": 0.29691648483276367, "learning_rate": 6.890670950802474e-05, "loss": 0.0273, "step": 32620 }, { "grad_norm": 0.20506097376346588, "learning_rate": 6.88875741307744e-05, "loss": 0.0216, "step": 32630 }, { "grad_norm": 0.3425416052341461, "learning_rate": 6.886843552615425e-05, "loss": 0.019, "step": 32640 }, { "grad_norm": 0.44528332352638245, "learning_rate": 6.884929369743451e-05, "loss": 0.0227, "step": 32650 }, { "grad_norm": 0.39736026525497437, "learning_rate": 6.8830148647886e-05, "loss": 0.0234, "step": 32660 }, { "grad_norm": 0.25496211647987366, "learning_rate": 6.88110003807801e-05, "loss": 0.0331, "step": 32670 }, { "grad_norm": 0.3073572516441345, "learning_rate": 6.87918488993887e-05, "loss": 0.0277, "step": 32680 }, { "grad_norm": 0.4894869923591614, "learning_rate": 6.877269420698431e-05, "loss": 0.0309, "step": 32690 }, { "grad_norm": 0.38288581371307373, "learning_rate": 6.875353630683989e-05, "loss": 0.0293, "step": 32700 }, { "grad_norm": 0.8215011954307556, "learning_rate": 6.873437520222905e-05, "loss": 0.0326, "step": 32710 }, { "grad_norm": 0.34617331624031067, "learning_rate": 6.871521089642585e-05, "loss": 0.0269, "step": 32720 }, { "grad_norm": 0.24166230857372284, "learning_rate": 6.869604339270498e-05, "loss": 0.0284, "step": 32730 }, { "grad_norm": 0.4151597023010254, "learning_rate": 6.867687269434164e-05, "loss": 0.0272, "step": 32740 }, { "grad_norm": 0.272536963224411, "learning_rate": 6.865769880461156e-05, "loss": 0.0262, "step": 32750 }, { "grad_norm": 0.23291607201099396, "learning_rate": 6.863852172679104e-05, "loss": 0.0245, "step": 32760 }, { "grad_norm": 0.524907648563385, "learning_rate": 6.861934146415693e-05, "loss": 0.0286, "step": 32770 }, { "grad_norm": 0.3093537986278534, "learning_rate": 6.86001580199866e-05, "loss": 0.0314, "step": 32780 }, { "grad_norm": 0.4753526747226715, "learning_rate": 6.858097139755798e-05, "loss": 0.0227, "step": 32790 }, { "grad_norm": 0.23708392679691315, "learning_rate": 6.856178160014955e-05, "loss": 0.0256, "step": 32800 }, { "grad_norm": 0.26901885867118835, "learning_rate": 6.85425886310403e-05, "loss": 0.023, "step": 32810 }, { "grad_norm": 0.38244470953941345, "learning_rate": 6.852339249350979e-05, "loss": 0.0289, "step": 32820 }, { "grad_norm": 0.3178005814552307, "learning_rate": 6.850419319083812e-05, "loss": 0.033, "step": 32830 }, { "grad_norm": 0.269265741109848, "learning_rate": 6.848499072630592e-05, "loss": 0.022, "step": 32840 }, { "grad_norm": 0.5717308521270752, "learning_rate": 6.846578510319439e-05, "loss": 0.0263, "step": 32850 }, { "grad_norm": 0.3502770960330963, "learning_rate": 6.844657632478519e-05, "loss": 0.0287, "step": 32860 }, { "grad_norm": 0.3355065882205963, "learning_rate": 6.842736439436063e-05, "loss": 0.0199, "step": 32870 }, { "grad_norm": 0.40837597846984863, "learning_rate": 6.84081493152035e-05, "loss": 0.0221, "step": 32880 }, { "grad_norm": 0.24631698429584503, "learning_rate": 6.83889310905971e-05, "loss": 0.0197, "step": 32890 }, { "grad_norm": 0.261505126953125, "learning_rate": 6.836970972382533e-05, "loss": 0.0248, "step": 32900 }, { "grad_norm": 0.18276581168174744, "learning_rate": 6.835048521817257e-05, "loss": 0.0228, "step": 32910 }, { "grad_norm": 0.3778412342071533, "learning_rate": 6.833125757692379e-05, "loss": 0.0312, "step": 32920 }, { "grad_norm": 0.2920396327972412, "learning_rate": 6.831202680336441e-05, "loss": 0.0184, "step": 32930 }, { "grad_norm": 0.2459837645292282, "learning_rate": 6.829279290078052e-05, "loss": 0.0243, "step": 32940 }, { "grad_norm": 0.4058130085468292, "learning_rate": 6.827355587245863e-05, "loss": 0.0238, "step": 32950 }, { "grad_norm": 0.19577690958976746, "learning_rate": 6.82543157216858e-05, "loss": 0.0194, "step": 32960 }, { "grad_norm": 0.32066479325294495, "learning_rate": 6.823507245174969e-05, "loss": 0.0213, "step": 32970 }, { "grad_norm": 0.3051808476448059, "learning_rate": 6.821582606593841e-05, "loss": 0.0181, "step": 32980 }, { "grad_norm": 0.2503421902656555, "learning_rate": 6.81965765675407e-05, "loss": 0.0204, "step": 32990 }, { "grad_norm": 0.40585458278656006, "learning_rate": 6.81773239598457e-05, "loss": 0.0229, "step": 33000 }, { "grad_norm": 0.186812624335289, "learning_rate": 6.815806824614319e-05, "loss": 0.0193, "step": 33010 }, { "grad_norm": 0.4968833923339844, "learning_rate": 6.813880942972343e-05, "loss": 0.0181, "step": 33020 }, { "grad_norm": 0.36576947569847107, "learning_rate": 6.811954751387726e-05, "loss": 0.0263, "step": 33030 }, { "grad_norm": 0.5312173962593079, "learning_rate": 6.810028250189598e-05, "loss": 0.0218, "step": 33040 }, { "grad_norm": 0.4600154459476471, "learning_rate": 6.808101439707147e-05, "loss": 0.0326, "step": 33050 }, { "grad_norm": 0.2580755650997162, "learning_rate": 6.806174320269609e-05, "loss": 0.0313, "step": 33060 }, { "grad_norm": 0.31165438890457153, "learning_rate": 6.804246892206281e-05, "loss": 0.0242, "step": 33070 }, { "grad_norm": 0.2671806514263153, "learning_rate": 6.802319155846506e-05, "loss": 0.0306, "step": 33080 }, { "grad_norm": 0.3167373836040497, "learning_rate": 6.800391111519679e-05, "loss": 0.0245, "step": 33090 }, { "grad_norm": 0.3045538365840912, "learning_rate": 6.798462759555253e-05, "loss": 0.0267, "step": 33100 }, { "grad_norm": 0.268869549036026, "learning_rate": 6.79653410028273e-05, "loss": 0.02, "step": 33110 }, { "grad_norm": 0.19419044256210327, "learning_rate": 6.794605134031663e-05, "loss": 0.0266, "step": 33120 }, { "grad_norm": 0.34813347458839417, "learning_rate": 6.792675861131661e-05, "loss": 0.0268, "step": 33130 }, { "grad_norm": 0.3096827268600464, "learning_rate": 6.790746281912386e-05, "loss": 0.0293, "step": 33140 }, { "grad_norm": 0.2752556800842285, "learning_rate": 6.788816396703546e-05, "loss": 0.0205, "step": 33150 }, { "grad_norm": 0.31495049595832825, "learning_rate": 6.78688620583491e-05, "loss": 0.0228, "step": 33160 }, { "grad_norm": 0.5058503746986389, "learning_rate": 6.784955709636292e-05, "loss": 0.0213, "step": 33170 }, { "grad_norm": 0.443678081035614, "learning_rate": 6.783024908437564e-05, "loss": 0.026, "step": 33180 }, { "grad_norm": 0.26044395565986633, "learning_rate": 6.781093802568641e-05, "loss": 0.0201, "step": 33190 }, { "grad_norm": 0.33718419075012207, "learning_rate": 6.779162392359504e-05, "loss": 0.0185, "step": 33200 }, { "grad_norm": 0.43335962295532227, "learning_rate": 6.777230678140172e-05, "loss": 0.0217, "step": 33210 }, { "grad_norm": 0.35483673214912415, "learning_rate": 6.775298660240726e-05, "loss": 0.0252, "step": 33220 }, { "grad_norm": 0.25633639097213745, "learning_rate": 6.773366338991292e-05, "loss": 0.0337, "step": 33230 }, { "grad_norm": 0.32687458395957947, "learning_rate": 6.771433714722052e-05, "loss": 0.0247, "step": 33240 }, { "grad_norm": 0.28722161054611206, "learning_rate": 6.769500787763239e-05, "loss": 0.0289, "step": 33250 }, { "grad_norm": 0.28162387013435364, "learning_rate": 6.76756755844514e-05, "loss": 0.0253, "step": 33260 }, { "grad_norm": 0.23577745258808136, "learning_rate": 6.765634027098087e-05, "loss": 0.0282, "step": 33270 }, { "grad_norm": 0.30507245659828186, "learning_rate": 6.763700194052468e-05, "loss": 0.0262, "step": 33280 }, { "grad_norm": 0.41561564803123474, "learning_rate": 6.761766059638723e-05, "loss": 0.0251, "step": 33290 }, { "grad_norm": 0.4468967914581299, "learning_rate": 6.759831624187345e-05, "loss": 0.0218, "step": 33300 }, { "grad_norm": 0.39427557587623596, "learning_rate": 6.757896888028871e-05, "loss": 0.0279, "step": 33310 }, { "grad_norm": 0.253309041261673, "learning_rate": 6.7559618514939e-05, "loss": 0.0251, "step": 33320 }, { "grad_norm": 0.4267646074295044, "learning_rate": 6.754026514913073e-05, "loss": 0.0242, "step": 33330 }, { "grad_norm": 0.27503275871276855, "learning_rate": 6.752090878617087e-05, "loss": 0.027, "step": 33340 }, { "grad_norm": 0.3310227394104004, "learning_rate": 6.75015494293669e-05, "loss": 0.0179, "step": 33350 }, { "grad_norm": 0.32785487174987793, "learning_rate": 6.74821870820268e-05, "loss": 0.019, "step": 33360 }, { "grad_norm": 0.15862944722175598, "learning_rate": 6.746282174745907e-05, "loss": 0.0252, "step": 33370 }, { "grad_norm": 0.3428398072719574, "learning_rate": 6.744345342897271e-05, "loss": 0.0247, "step": 33380 }, { "grad_norm": 0.3415713310241699, "learning_rate": 6.742408212987724e-05, "loss": 0.0237, "step": 33390 }, { "grad_norm": 0.24611695110797882, "learning_rate": 6.740470785348269e-05, "loss": 0.0158, "step": 33400 }, { "grad_norm": 0.30684223771095276, "learning_rate": 6.738533060309958e-05, "loss": 0.0337, "step": 33410 }, { "grad_norm": 0.2886050045490265, "learning_rate": 6.736595038203894e-05, "loss": 0.0214, "step": 33420 }, { "grad_norm": 0.35855984687805176, "learning_rate": 6.734656719361236e-05, "loss": 0.0276, "step": 33430 }, { "grad_norm": 0.22129715979099274, "learning_rate": 6.732718104113189e-05, "loss": 0.0226, "step": 33440 }, { "grad_norm": 0.3175240457057953, "learning_rate": 6.730779192791006e-05, "loss": 0.0179, "step": 33450 }, { "grad_norm": 0.27222585678100586, "learning_rate": 6.728839985725997e-05, "loss": 0.0203, "step": 33460 }, { "grad_norm": 0.2989107072353363, "learning_rate": 6.726900483249517e-05, "loss": 0.0195, "step": 33470 }, { "grad_norm": 0.37325477600097656, "learning_rate": 6.724960685692976e-05, "loss": 0.0211, "step": 33480 }, { "grad_norm": 0.3886723816394806, "learning_rate": 6.723020593387833e-05, "loss": 0.0221, "step": 33490 }, { "grad_norm": 0.36547744274139404, "learning_rate": 6.721080206665593e-05, "loss": 0.0278, "step": 33500 }, { "grad_norm": 0.1988804191350937, "learning_rate": 6.719139525857819e-05, "loss": 0.0271, "step": 33510 }, { "grad_norm": 0.23894543945789337, "learning_rate": 6.717198551296117e-05, "loss": 0.0213, "step": 33520 }, { "grad_norm": 0.8694857954978943, "learning_rate": 6.715257283312148e-05, "loss": 0.0263, "step": 33530 }, { "grad_norm": 0.37859266996383667, "learning_rate": 6.713315722237623e-05, "loss": 0.0246, "step": 33540 }, { "grad_norm": 0.5399892330169678, "learning_rate": 6.7113738684043e-05, "loss": 0.0303, "step": 33550 }, { "grad_norm": 0.2933160364627838, "learning_rate": 6.709431722143989e-05, "loss": 0.0276, "step": 33560 }, { "grad_norm": 0.2729542553424835, "learning_rate": 6.70748928378855e-05, "loss": 0.0298, "step": 33570 }, { "grad_norm": 0.3106006383895874, "learning_rate": 6.705546553669891e-05, "loss": 0.0207, "step": 33580 }, { "grad_norm": 0.32109206914901733, "learning_rate": 6.703603532119974e-05, "loss": 0.0224, "step": 33590 }, { "grad_norm": 0.23131056129932404, "learning_rate": 6.701660219470808e-05, "loss": 0.0193, "step": 33600 }, { "grad_norm": 0.23290829360485077, "learning_rate": 6.69971661605445e-05, "loss": 0.0184, "step": 33610 }, { "grad_norm": 0.28151360154151917, "learning_rate": 6.697772722203008e-05, "loss": 0.0227, "step": 33620 }, { "grad_norm": 0.3088677227497101, "learning_rate": 6.695828538248643e-05, "loss": 0.0204, "step": 33630 }, { "grad_norm": 0.33530017733573914, "learning_rate": 6.693884064523563e-05, "loss": 0.0207, "step": 33640 }, { "grad_norm": 0.37714290618896484, "learning_rate": 6.691939301360023e-05, "loss": 0.0167, "step": 33650 }, { "grad_norm": 0.3017638623714447, "learning_rate": 6.689994249090333e-05, "loss": 0.0176, "step": 33660 }, { "grad_norm": 0.31178799271583557, "learning_rate": 6.688048908046845e-05, "loss": 0.0209, "step": 33670 }, { "grad_norm": 0.16218818724155426, "learning_rate": 6.686103278561969e-05, "loss": 0.0165, "step": 33680 }, { "grad_norm": 0.3861725330352783, "learning_rate": 6.684157360968156e-05, "loss": 0.0232, "step": 33690 }, { "grad_norm": 0.2973552644252777, "learning_rate": 6.682211155597911e-05, "loss": 0.0166, "step": 33700 }, { "grad_norm": 0.36878448724746704, "learning_rate": 6.680264662783789e-05, "loss": 0.0234, "step": 33710 }, { "grad_norm": 0.4262727200984955, "learning_rate": 6.678317882858391e-05, "loss": 0.0265, "step": 33720 }, { "grad_norm": 0.2554619312286377, "learning_rate": 6.67637081615437e-05, "loss": 0.0235, "step": 33730 }, { "grad_norm": 0.1893453747034073, "learning_rate": 6.674423463004427e-05, "loss": 0.017, "step": 33740 }, { "grad_norm": 0.3032743036746979, "learning_rate": 6.672475823741308e-05, "loss": 0.0211, "step": 33750 }, { "grad_norm": 0.269051194190979, "learning_rate": 6.670527898697811e-05, "loss": 0.0223, "step": 33760 }, { "grad_norm": 0.18323677778244019, "learning_rate": 6.668579688206788e-05, "loss": 0.0238, "step": 33770 }, { "grad_norm": 0.2689245939254761, "learning_rate": 6.666631192601131e-05, "loss": 0.0171, "step": 33780 }, { "grad_norm": 0.22343796491622925, "learning_rate": 6.664682412213785e-05, "loss": 0.0234, "step": 33790 }, { "grad_norm": 0.33953067660331726, "learning_rate": 6.662733347377745e-05, "loss": 0.023, "step": 33800 }, { "grad_norm": 0.3329608738422394, "learning_rate": 6.660783998426051e-05, "loss": 0.024, "step": 33810 }, { "grad_norm": 0.2930600047111511, "learning_rate": 6.658834365691794e-05, "loss": 0.0221, "step": 33820 }, { "grad_norm": 0.32075273990631104, "learning_rate": 6.656884449508115e-05, "loss": 0.0233, "step": 33830 }, { "grad_norm": 0.2927626371383667, "learning_rate": 6.654934250208198e-05, "loss": 0.0269, "step": 33840 }, { "grad_norm": 0.39279699325561523, "learning_rate": 6.65298376812528e-05, "loss": 0.0288, "step": 33850 }, { "grad_norm": 0.2957425117492676, "learning_rate": 6.651033003592646e-05, "loss": 0.0258, "step": 33860 }, { "grad_norm": 0.25449225306510925, "learning_rate": 6.649081956943626e-05, "loss": 0.0219, "step": 33870 }, { "grad_norm": 0.2377353310585022, "learning_rate": 6.647130628511604e-05, "loss": 0.0274, "step": 33880 }, { "grad_norm": 0.4454909563064575, "learning_rate": 6.645179018630005e-05, "loss": 0.0269, "step": 33890 }, { "grad_norm": 0.30921670794487, "learning_rate": 6.643227127632309e-05, "loss": 0.0251, "step": 33900 }, { "grad_norm": 0.3857629597187042, "learning_rate": 6.641274955852038e-05, "loss": 0.0303, "step": 33910 }, { "grad_norm": 0.5574055314064026, "learning_rate": 6.639322503622768e-05, "loss": 0.0545, "step": 33920 }, { "grad_norm": 0.336988627910614, "learning_rate": 6.637369771278116e-05, "loss": 0.026, "step": 33930 }, { "grad_norm": 0.336769700050354, "learning_rate": 6.635416759151751e-05, "loss": 0.0207, "step": 33940 }, { "grad_norm": 0.2628577649593353, "learning_rate": 6.633463467577394e-05, "loss": 0.0299, "step": 33950 }, { "grad_norm": 0.3378629684448242, "learning_rate": 6.631509896888803e-05, "loss": 0.0221, "step": 33960 }, { "grad_norm": 0.27862271666526794, "learning_rate": 6.629556047419794e-05, "loss": 0.0178, "step": 33970 }, { "grad_norm": 0.3003343343734741, "learning_rate": 6.627601919504223e-05, "loss": 0.0197, "step": 33980 }, { "grad_norm": 0.2911309599876404, "learning_rate": 6.625647513476001e-05, "loss": 0.0182, "step": 33990 }, { "grad_norm": 0.41073527932167053, "learning_rate": 6.62369282966908e-05, "loss": 0.025, "step": 34000 }, { "grad_norm": 0.35069313645362854, "learning_rate": 6.621737868417464e-05, "loss": 0.0234, "step": 34010 }, { "grad_norm": 0.22181235253810883, "learning_rate": 6.619782630055198e-05, "loss": 0.023, "step": 34020 }, { "grad_norm": 0.25969842076301575, "learning_rate": 6.617827114916382e-05, "loss": 0.0265, "step": 34030 }, { "grad_norm": 0.22480949759483337, "learning_rate": 6.615871323335161e-05, "loss": 0.0278, "step": 34040 }, { "grad_norm": 0.44525426626205444, "learning_rate": 6.613915255645725e-05, "loss": 0.0252, "step": 34050 }, { "grad_norm": 0.4095677137374878, "learning_rate": 6.611958912182312e-05, "loss": 0.0216, "step": 34060 }, { "grad_norm": 0.6069485545158386, "learning_rate": 6.610002293279207e-05, "loss": 0.024, "step": 34070 }, { "grad_norm": 0.2638224959373474, "learning_rate": 6.608045399270746e-05, "loss": 0.0309, "step": 34080 }, { "grad_norm": 0.3532731235027313, "learning_rate": 6.606088230491304e-05, "loss": 0.0292, "step": 34090 }, { "grad_norm": 0.4118747413158417, "learning_rate": 6.604130787275312e-05, "loss": 0.0281, "step": 34100 }, { "grad_norm": 0.4380190968513489, "learning_rate": 6.602173069957242e-05, "loss": 0.029, "step": 34110 }, { "grad_norm": 0.4041662812232971, "learning_rate": 6.600215078871612e-05, "loss": 0.0256, "step": 34120 }, { "grad_norm": 0.45088696479797363, "learning_rate": 6.598256814352992e-05, "loss": 0.0298, "step": 34130 }, { "grad_norm": 0.4478030800819397, "learning_rate": 6.596298276735995e-05, "loss": 0.0206, "step": 34140 }, { "grad_norm": 0.33590471744537354, "learning_rate": 6.594339466355282e-05, "loss": 0.0265, "step": 34150 }, { "grad_norm": 0.3113204836845398, "learning_rate": 6.592380383545558e-05, "loss": 0.0246, "step": 34160 }, { "grad_norm": 0.2963636517524719, "learning_rate": 6.590421028641577e-05, "loss": 0.0314, "step": 34170 }, { "grad_norm": 0.3710557222366333, "learning_rate": 6.588461401978143e-05, "loss": 0.0312, "step": 34180 }, { "grad_norm": 0.2689637243747711, "learning_rate": 6.586501503890099e-05, "loss": 0.0258, "step": 34190 }, { "grad_norm": 0.28657469153404236, "learning_rate": 6.584541334712338e-05, "loss": 0.0244, "step": 34200 }, { "grad_norm": 0.31008729338645935, "learning_rate": 6.582580894779802e-05, "loss": 0.0257, "step": 34210 }, { "grad_norm": 0.3466331958770752, "learning_rate": 6.580620184427473e-05, "loss": 0.0276, "step": 34220 }, { "grad_norm": 0.20957952737808228, "learning_rate": 6.578659203990385e-05, "loss": 0.0227, "step": 34230 }, { "grad_norm": 0.2951942980289459, "learning_rate": 6.576697953803615e-05, "loss": 0.0237, "step": 34240 }, { "grad_norm": 0.3530837893486023, "learning_rate": 6.57473643420229e-05, "loss": 0.0245, "step": 34250 }, { "grad_norm": 0.3722061216831207, "learning_rate": 6.572774645521574e-05, "loss": 0.026, "step": 34260 }, { "grad_norm": 0.41498449444770813, "learning_rate": 6.570812588096688e-05, "loss": 0.0304, "step": 34270 }, { "grad_norm": 0.338066428899765, "learning_rate": 6.568850262262893e-05, "loss": 0.022, "step": 34280 }, { "grad_norm": 0.3036212921142578, "learning_rate": 6.566887668355497e-05, "loss": 0.0211, "step": 34290 }, { "grad_norm": 0.24147534370422363, "learning_rate": 6.564924806709851e-05, "loss": 0.0203, "step": 34300 }, { "grad_norm": 0.3504810631275177, "learning_rate": 6.562961677661359e-05, "loss": 0.0285, "step": 34310 }, { "grad_norm": 0.4086492359638214, "learning_rate": 6.56099828154546e-05, "loss": 0.0221, "step": 34320 }, { "grad_norm": 0.33350756764411926, "learning_rate": 6.55903461869765e-05, "loss": 0.0216, "step": 34330 }, { "grad_norm": 0.30475860834121704, "learning_rate": 6.557070689453465e-05, "loss": 0.018, "step": 34340 }, { "grad_norm": 0.2590806484222412, "learning_rate": 6.555106494148482e-05, "loss": 0.0281, "step": 34350 }, { "grad_norm": 0.40839773416519165, "learning_rate": 6.553142033118333e-05, "loss": 0.0255, "step": 34360 }, { "grad_norm": 0.41393423080444336, "learning_rate": 6.551177306698688e-05, "loss": 0.029, "step": 34370 }, { "grad_norm": 0.2202126383781433, "learning_rate": 6.549212315225267e-05, "loss": 0.023, "step": 34380 }, { "grad_norm": 0.2736908197402954, "learning_rate": 6.547247059033833e-05, "loss": 0.0242, "step": 34390 }, { "grad_norm": 0.34885260462760925, "learning_rate": 6.545281538460193e-05, "loss": 0.0247, "step": 34400 }, { "grad_norm": 0.22970810532569885, "learning_rate": 6.543315753840202e-05, "loss": 0.0233, "step": 34410 }, { "grad_norm": 0.30771273374557495, "learning_rate": 6.541349705509758e-05, "loss": 0.02, "step": 34420 }, { "grad_norm": 0.2550351917743683, "learning_rate": 6.539383393804805e-05, "loss": 0.0254, "step": 34430 }, { "grad_norm": 0.3687790632247925, "learning_rate": 6.537416819061333e-05, "loss": 0.026, "step": 34440 }, { "grad_norm": 0.17640002071857452, "learning_rate": 6.535449981615375e-05, "loss": 0.0306, "step": 34450 }, { "grad_norm": 0.3173871636390686, "learning_rate": 6.53348288180301e-05, "loss": 0.0206, "step": 34460 }, { "grad_norm": 0.24393224716186523, "learning_rate": 6.531515519960361e-05, "loss": 0.0295, "step": 34470 }, { "grad_norm": 0.5453435778617859, "learning_rate": 6.529547896423597e-05, "loss": 0.0273, "step": 34480 }, { "grad_norm": 0.39220571517944336, "learning_rate": 6.52758001152893e-05, "loss": 0.0257, "step": 34490 }, { "grad_norm": 0.22596310079097748, "learning_rate": 6.525611865612618e-05, "loss": 0.0242, "step": 34500 }, { "grad_norm": 0.3234485685825348, "learning_rate": 6.523643459010966e-05, "loss": 0.0215, "step": 34510 }, { "grad_norm": 0.4883192777633667, "learning_rate": 6.521674792060317e-05, "loss": 0.0254, "step": 34520 }, { "grad_norm": 0.3959271013736725, "learning_rate": 6.519705865097063e-05, "loss": 0.0262, "step": 34530 }, { "grad_norm": 0.29845428466796875, "learning_rate": 6.517736678457641e-05, "loss": 0.0208, "step": 34540 }, { "grad_norm": 0.4933452606201172, "learning_rate": 6.515767232478534e-05, "loss": 0.0228, "step": 34550 }, { "grad_norm": 0.3365289866924286, "learning_rate": 6.51379752749626e-05, "loss": 0.0207, "step": 34560 }, { "grad_norm": 0.45010319352149963, "learning_rate": 6.511827563847393e-05, "loss": 0.0247, "step": 34570 }, { "grad_norm": 0.28288307785987854, "learning_rate": 6.509857341868542e-05, "loss": 0.033, "step": 34580 }, { "grad_norm": 0.2250380516052246, "learning_rate": 6.507886861896367e-05, "loss": 0.0263, "step": 34590 }, { "grad_norm": 0.3027667701244354, "learning_rate": 6.505916124267567e-05, "loss": 0.0352, "step": 34600 }, { "grad_norm": 0.6219964623451233, "learning_rate": 6.503945129318891e-05, "loss": 0.0307, "step": 34610 }, { "grad_norm": 0.2394835501909256, "learning_rate": 6.501973877387122e-05, "loss": 0.0198, "step": 34620 }, { "grad_norm": 0.4215441942214966, "learning_rate": 6.500002368809098e-05, "loss": 0.0248, "step": 34630 }, { "grad_norm": 0.4185514450073242, "learning_rate": 6.498030603921694e-05, "loss": 0.026, "step": 34640 }, { "grad_norm": 0.6202311515808105, "learning_rate": 6.496058583061832e-05, "loss": 0.0305, "step": 34650 }, { "grad_norm": 0.2237073928117752, "learning_rate": 6.494086306566475e-05, "loss": 0.0172, "step": 34660 }, { "grad_norm": 0.36220282316207886, "learning_rate": 6.492113774772632e-05, "loss": 0.0264, "step": 34670 }, { "grad_norm": 0.3987637758255005, "learning_rate": 6.490140988017354e-05, "loss": 0.0256, "step": 34680 }, { "grad_norm": 0.4041312038898468, "learning_rate": 6.488167946637736e-05, "loss": 0.0193, "step": 34690 }, { "grad_norm": 0.4027598798274994, "learning_rate": 6.486194650970915e-05, "loss": 0.03, "step": 34700 }, { "grad_norm": 0.3259913921356201, "learning_rate": 6.48422110135408e-05, "loss": 0.0258, "step": 34710 }, { "grad_norm": 0.3584539592266083, "learning_rate": 6.482247298124451e-05, "loss": 0.0327, "step": 34720 }, { "grad_norm": 0.27786630392074585, "learning_rate": 6.480273241619297e-05, "loss": 0.0235, "step": 34730 }, { "grad_norm": 0.28736549615859985, "learning_rate": 6.478298932175933e-05, "loss": 0.0221, "step": 34740 }, { "grad_norm": 0.165510356426239, "learning_rate": 6.476324370131712e-05, "loss": 0.0194, "step": 34750 }, { "grad_norm": 0.46861860156059265, "learning_rate": 6.474349555824036e-05, "loss": 0.0277, "step": 34760 }, { "grad_norm": 0.24928028881549835, "learning_rate": 6.472374489590342e-05, "loss": 0.0217, "step": 34770 }, { "grad_norm": 0.32287925481796265, "learning_rate": 6.470399171768118e-05, "loss": 0.0286, "step": 34780 }, { "grad_norm": 0.32003968954086304, "learning_rate": 6.468423602694891e-05, "loss": 0.0204, "step": 34790 }, { "grad_norm": 0.3585580289363861, "learning_rate": 6.466447782708232e-05, "loss": 0.0182, "step": 34800 }, { "grad_norm": 0.38867366313934326, "learning_rate": 6.464471712145754e-05, "loss": 0.0262, "step": 34810 }, { "grad_norm": 0.2692846357822418, "learning_rate": 6.462495391345114e-05, "loss": 0.03, "step": 34820 }, { "grad_norm": 0.207396999001503, "learning_rate": 6.46051882064401e-05, "loss": 0.0188, "step": 34830 }, { "grad_norm": 0.3312571048736572, "learning_rate": 6.458542000380186e-05, "loss": 0.0193, "step": 34840 }, { "grad_norm": 0.29692551493644714, "learning_rate": 6.456564930891424e-05, "loss": 0.0223, "step": 34850 }, { "grad_norm": 0.35675275325775146, "learning_rate": 6.454587612515555e-05, "loss": 0.0265, "step": 34860 }, { "grad_norm": 0.21709877252578735, "learning_rate": 6.452610045590444e-05, "loss": 0.0242, "step": 34870 }, { "grad_norm": 0.3636498749256134, "learning_rate": 6.450632230454005e-05, "loss": 0.0184, "step": 34880 }, { "grad_norm": 0.3253977596759796, "learning_rate": 6.448654167444195e-05, "loss": 0.0287, "step": 34890 }, { "grad_norm": 0.2832447290420532, "learning_rate": 6.446675856899005e-05, "loss": 0.0156, "step": 34900 }, { "grad_norm": 0.2546078860759735, "learning_rate": 6.444697299156481e-05, "loss": 0.0197, "step": 34910 }, { "grad_norm": 0.29535338282585144, "learning_rate": 6.442718494554701e-05, "loss": 0.0212, "step": 34920 }, { "grad_norm": 0.30354422330856323, "learning_rate": 6.440739443431787e-05, "loss": 0.0171, "step": 34930 }, { "grad_norm": 0.29275232553482056, "learning_rate": 6.438760146125906e-05, "loss": 0.0193, "step": 34940 }, { "grad_norm": 0.3587654232978821, "learning_rate": 6.436780602975267e-05, "loss": 0.0211, "step": 34950 }, { "grad_norm": 0.22938229143619537, "learning_rate": 6.43480081431812e-05, "loss": 0.0284, "step": 34960 }, { "grad_norm": 0.3338093161582947, "learning_rate": 6.432820780492756e-05, "loss": 0.027, "step": 34970 }, { "grad_norm": 0.35623645782470703, "learning_rate": 6.430840501837506e-05, "loss": 0.0279, "step": 34980 }, { "grad_norm": 0.22604261338710785, "learning_rate": 6.428859978690748e-05, "loss": 0.0183, "step": 34990 }, { "grad_norm": 0.2979201376438141, "learning_rate": 6.426879211390901e-05, "loss": 0.0191, "step": 35000 }, { "grad_norm": 0.2844535708427429, "learning_rate": 6.424898200276422e-05, "loss": 0.022, "step": 35010 }, { "grad_norm": 0.24159374833106995, "learning_rate": 6.42291694568581e-05, "loss": 0.0224, "step": 35020 }, { "grad_norm": 0.3181571960449219, "learning_rate": 6.42093544795761e-05, "loss": 0.0258, "step": 35030 }, { "grad_norm": 0.2863505780696869, "learning_rate": 6.418953707430403e-05, "loss": 0.0291, "step": 35040 }, { "grad_norm": 0.4441314935684204, "learning_rate": 6.416971724442819e-05, "loss": 0.0233, "step": 35050 }, { "grad_norm": 0.3126890957355499, "learning_rate": 6.414989499333519e-05, "loss": 0.0241, "step": 35060 }, { "grad_norm": 0.2881414592266083, "learning_rate": 6.413007032441214e-05, "loss": 0.0229, "step": 35070 }, { "grad_norm": 0.2981286644935608, "learning_rate": 6.411024324104653e-05, "loss": 0.0224, "step": 35080 }, { "grad_norm": 0.2690493166446686, "learning_rate": 6.409041374662628e-05, "loss": 0.0195, "step": 35090 }, { "grad_norm": 0.2833879292011261, "learning_rate": 6.407058184453967e-05, "loss": 0.0332, "step": 35100 }, { "grad_norm": 0.3322078585624695, "learning_rate": 6.405074753817548e-05, "loss": 0.0243, "step": 35110 }, { "grad_norm": 0.44589272141456604, "learning_rate": 6.40309108309228e-05, "loss": 0.0282, "step": 35120 }, { "grad_norm": 0.3798638582229614, "learning_rate": 6.401107172617122e-05, "loss": 0.0327, "step": 35130 }, { "grad_norm": 0.4221253991127014, "learning_rate": 6.399123022731068e-05, "loss": 0.0416, "step": 35140 }, { "grad_norm": 0.29868853092193604, "learning_rate": 6.397138633773157e-05, "loss": 0.0215, "step": 35150 }, { "grad_norm": 0.3734951913356781, "learning_rate": 6.395154006082463e-05, "loss": 0.0224, "step": 35160 }, { "grad_norm": 0.3920116126537323, "learning_rate": 6.393169139998109e-05, "loss": 0.0225, "step": 35170 }, { "grad_norm": 0.3770543932914734, "learning_rate": 6.39118403585925e-05, "loss": 0.023, "step": 35180 }, { "grad_norm": 0.36095544695854187, "learning_rate": 6.38919869400509e-05, "loss": 0.0255, "step": 35190 }, { "grad_norm": 0.31578150391578674, "learning_rate": 6.387213114774865e-05, "loss": 0.0258, "step": 35200 }, { "grad_norm": 0.45536792278289795, "learning_rate": 6.385227298507863e-05, "loss": 0.056, "step": 35210 }, { "grad_norm": 0.30960842967033386, "learning_rate": 6.3832412455434e-05, "loss": 0.0237, "step": 35220 }, { "grad_norm": 0.3013031482696533, "learning_rate": 6.381254956220841e-05, "loss": 0.0275, "step": 35230 }, { "grad_norm": 0.2838110625743866, "learning_rate": 6.379268430879586e-05, "loss": 0.023, "step": 35240 }, { "grad_norm": 0.5557355284690857, "learning_rate": 6.37728166985908e-05, "loss": 0.0294, "step": 35250 }, { "grad_norm": 0.8291860818862915, "learning_rate": 6.375294673498804e-05, "loss": 0.0248, "step": 35260 }, { "grad_norm": 0.8396528363227844, "learning_rate": 6.373307442138284e-05, "loss": 0.0274, "step": 35270 }, { "grad_norm": 0.34938353300094604, "learning_rate": 6.371319976117081e-05, "loss": 0.0367, "step": 35280 }, { "grad_norm": 0.44334331154823303, "learning_rate": 6.3693322757748e-05, "loss": 0.0283, "step": 35290 }, { "grad_norm": 0.30696770548820496, "learning_rate": 6.367344341451086e-05, "loss": 0.0268, "step": 35300 }, { "grad_norm": 0.2800368070602417, "learning_rate": 6.36535617348562e-05, "loss": 0.0269, "step": 35310 }, { "grad_norm": 0.2808513939380646, "learning_rate": 6.363367772218128e-05, "loss": 0.0235, "step": 35320 }, { "grad_norm": 0.32279306650161743, "learning_rate": 6.36137913798837e-05, "loss": 0.0232, "step": 35330 }, { "grad_norm": 0.2685723602771759, "learning_rate": 6.359390271136151e-05, "loss": 0.0192, "step": 35340 }, { "grad_norm": 0.3102090656757355, "learning_rate": 6.357401172001314e-05, "loss": 0.0241, "step": 35350 }, { "grad_norm": 0.2787506580352783, "learning_rate": 6.355411840923742e-05, "loss": 0.022, "step": 35360 }, { "grad_norm": 0.2770788073539734, "learning_rate": 6.353422278243358e-05, "loss": 0.0214, "step": 35370 }, { "grad_norm": 0.4236556887626648, "learning_rate": 6.351432484300121e-05, "loss": 0.028, "step": 35380 }, { "grad_norm": 0.3215552270412445, "learning_rate": 6.349442459434036e-05, "loss": 0.0225, "step": 35390 }, { "grad_norm": 0.24380221962928772, "learning_rate": 6.34745220398514e-05, "loss": 0.031, "step": 35400 }, { "grad_norm": 0.2565545439720154, "learning_rate": 6.345461718293518e-05, "loss": 0.022, "step": 35410 }, { "grad_norm": 0.3534467816352844, "learning_rate": 6.343471002699286e-05, "loss": 0.0304, "step": 35420 }, { "grad_norm": 0.28322866559028625, "learning_rate": 6.341480057542602e-05, "loss": 0.0251, "step": 35430 }, { "grad_norm": 0.318327933549881, "learning_rate": 6.339488883163667e-05, "loss": 0.0246, "step": 35440 }, { "grad_norm": 0.4968356490135193, "learning_rate": 6.337497479902716e-05, "loss": 0.0255, "step": 35450 }, { "grad_norm": 0.4017643332481384, "learning_rate": 6.335505848100027e-05, "loss": 0.0265, "step": 35460 }, { "grad_norm": 0.512223482131958, "learning_rate": 6.333513988095915e-05, "loss": 0.0182, "step": 35470 }, { "grad_norm": 0.3120320737361908, "learning_rate": 6.331521900230735e-05, "loss": 0.0222, "step": 35480 }, { "grad_norm": 0.3543839156627655, "learning_rate": 6.329529584844878e-05, "loss": 0.022, "step": 35490 }, { "grad_norm": 0.31159842014312744, "learning_rate": 6.327537042278777e-05, "loss": 0.0258, "step": 35500 }, { "grad_norm": 0.32200688123703003, "learning_rate": 6.325544272872905e-05, "loss": 0.0211, "step": 35510 }, { "grad_norm": 0.39796754717826843, "learning_rate": 6.323551276967771e-05, "loss": 0.0265, "step": 35520 }, { "grad_norm": 0.3419071137905121, "learning_rate": 6.321558054903922e-05, "loss": 0.0209, "step": 35530 }, { "grad_norm": 0.2598966658115387, "learning_rate": 6.319564607021947e-05, "loss": 0.0225, "step": 35540 }, { "grad_norm": 0.37796443700790405, "learning_rate": 6.31757093366247e-05, "loss": 0.0203, "step": 35550 }, { "grad_norm": 0.37246575951576233, "learning_rate": 6.315577035166154e-05, "loss": 0.0199, "step": 35560 }, { "grad_norm": 0.2656288743019104, "learning_rate": 6.313582911873708e-05, "loss": 0.0168, "step": 35570 }, { "grad_norm": 0.34658604860305786, "learning_rate": 6.311588564125865e-05, "loss": 0.0175, "step": 35580 }, { "grad_norm": 0.42954397201538086, "learning_rate": 6.30959399226341e-05, "loss": 0.0208, "step": 35590 }, { "grad_norm": 0.28225263953208923, "learning_rate": 6.30759919662716e-05, "loss": 0.0348, "step": 35600 }, { "grad_norm": 0.2648703157901764, "learning_rate": 6.30560417755797e-05, "loss": 0.0205, "step": 35610 }, { "grad_norm": 0.1780417412519455, "learning_rate": 6.303608935396735e-05, "loss": 0.0168, "step": 35620 }, { "grad_norm": 0.20808455348014832, "learning_rate": 6.301613470484386e-05, "loss": 0.0182, "step": 35630 }, { "grad_norm": 0.3168855905532837, "learning_rate": 6.299617783161893e-05, "loss": 0.0222, "step": 35640 }, { "grad_norm": 0.31365951895713806, "learning_rate": 6.297621873770266e-05, "loss": 0.0186, "step": 35650 }, { "grad_norm": 0.2848111391067505, "learning_rate": 6.29562574265055e-05, "loss": 0.026, "step": 35660 }, { "grad_norm": 0.3891395628452301, "learning_rate": 6.293629390143834e-05, "loss": 0.025, "step": 35670 }, { "grad_norm": 0.3267706334590912, "learning_rate": 6.291632816591232e-05, "loss": 0.0215, "step": 35680 }, { "grad_norm": 0.33913230895996094, "learning_rate": 6.28963602233391e-05, "loss": 0.0237, "step": 35690 }, { "grad_norm": 0.3568381369113922, "learning_rate": 6.287639007713062e-05, "loss": 0.0205, "step": 35700 }, { "grad_norm": 0.2335795760154724, "learning_rate": 6.285641773069926e-05, "loss": 0.0207, "step": 35710 }, { "grad_norm": 0.20112043619155884, "learning_rate": 6.283644318745773e-05, "loss": 0.0166, "step": 35720 }, { "grad_norm": 0.3641316890716553, "learning_rate": 6.281646645081912e-05, "loss": 0.0229, "step": 35730 }, { "grad_norm": 0.23146618902683258, "learning_rate": 6.279648752419693e-05, "loss": 0.0192, "step": 35740 }, { "grad_norm": 0.3443063199520111, "learning_rate": 6.2776506411005e-05, "loss": 0.0235, "step": 35750 }, { "grad_norm": 0.2968631088733673, "learning_rate": 6.275652311465758e-05, "loss": 0.0169, "step": 35760 }, { "grad_norm": 0.21836185455322266, "learning_rate": 6.273653763856926e-05, "loss": 0.0238, "step": 35770 }, { "grad_norm": 0.3024172782897949, "learning_rate": 6.271654998615501e-05, "loss": 0.0254, "step": 35780 }, { "grad_norm": 0.26302096247673035, "learning_rate": 6.269656016083013e-05, "loss": 0.0195, "step": 35790 }, { "grad_norm": 0.3277972638607025, "learning_rate": 6.267656816601038e-05, "loss": 0.0207, "step": 35800 }, { "grad_norm": 0.597752571105957, "learning_rate": 6.265657400511185e-05, "loss": 0.0284, "step": 35810 }, { "grad_norm": 0.25101765990257263, "learning_rate": 6.263657768155098e-05, "loss": 0.0237, "step": 35820 }, { "grad_norm": 0.48677706718444824, "learning_rate": 6.261657919874457e-05, "loss": 0.0236, "step": 35830 }, { "grad_norm": 0.3046591579914093, "learning_rate": 6.259657856010986e-05, "loss": 0.0313, "step": 35840 }, { "grad_norm": 0.3325697183609009, "learning_rate": 6.257657576906439e-05, "loss": 0.0284, "step": 35850 }, { "grad_norm": 0.35768020153045654, "learning_rate": 6.255657082902609e-05, "loss": 0.0192, "step": 35860 }, { "grad_norm": 0.3074094355106354, "learning_rate": 6.253656374341325e-05, "loss": 0.0167, "step": 35870 }, { "grad_norm": 0.27024292945861816, "learning_rate": 6.251655451564457e-05, "loss": 0.0228, "step": 35880 }, { "grad_norm": 0.2632875442504883, "learning_rate": 6.249654314913902e-05, "loss": 0.0223, "step": 35890 }, { "grad_norm": 0.30363497138023376, "learning_rate": 6.247652964731604e-05, "loss": 0.0184, "step": 35900 }, { "grad_norm": 0.19953669607639313, "learning_rate": 6.245651401359537e-05, "loss": 0.0284, "step": 35910 }, { "grad_norm": 0.25379234552383423, "learning_rate": 6.243649625139715e-05, "loss": 0.0242, "step": 35920 }, { "grad_norm": 0.3418880105018616, "learning_rate": 6.241647636414185e-05, "loss": 0.0183, "step": 35930 }, { "grad_norm": 0.4438258707523346, "learning_rate": 6.239645435525034e-05, "loss": 0.0253, "step": 35940 }, { "grad_norm": 0.2971133291721344, "learning_rate": 6.237643022814381e-05, "loss": 0.0243, "step": 35950 }, { "grad_norm": 0.36182641983032227, "learning_rate": 6.235640398624386e-05, "loss": 0.0242, "step": 35960 }, { "grad_norm": 0.2702111005783081, "learning_rate": 6.233637563297243e-05, "loss": 0.0251, "step": 35970 }, { "grad_norm": 0.33024001121520996, "learning_rate": 6.23163451717518e-05, "loss": 0.0236, "step": 35980 }, { "grad_norm": 0.3397286832332611, "learning_rate": 6.229631260600463e-05, "loss": 0.023, "step": 35990 }, { "grad_norm": 0.5238834023475647, "learning_rate": 6.227627793915392e-05, "loss": 0.0303, "step": 36000 }, { "grad_norm": 0.2900165915489197, "learning_rate": 6.225624117462309e-05, "loss": 0.0181, "step": 36010 }, { "grad_norm": 0.3375806510448456, "learning_rate": 6.223620231583586e-05, "loss": 0.0216, "step": 36020 }, { "grad_norm": 0.32130613923072815, "learning_rate": 6.221616136621629e-05, "loss": 0.0214, "step": 36030 }, { "grad_norm": 0.33351123332977295, "learning_rate": 6.219611832918887e-05, "loss": 0.0188, "step": 36040 }, { "grad_norm": 0.2824496328830719, "learning_rate": 6.217607320817838e-05, "loss": 0.0179, "step": 36050 }, { "grad_norm": 0.3127155900001526, "learning_rate": 6.215602600661001e-05, "loss": 0.0245, "step": 36060 }, { "grad_norm": 0.3767285645008087, "learning_rate": 6.213597672790925e-05, "loss": 0.0229, "step": 36070 }, { "grad_norm": 0.2609204351902008, "learning_rate": 6.2115925375502e-05, "loss": 0.0209, "step": 36080 }, { "grad_norm": 0.31072694063186646, "learning_rate": 6.209587195281447e-05, "loss": 0.0231, "step": 36090 }, { "grad_norm": 0.27298423647880554, "learning_rate": 6.207581646327324e-05, "loss": 0.0217, "step": 36100 }, { "grad_norm": 0.19546298682689667, "learning_rate": 6.205575891030526e-05, "loss": 0.0151, "step": 36110 }, { "grad_norm": 0.22638700902462006, "learning_rate": 6.203569929733781e-05, "loss": 0.0165, "step": 36120 }, { "grad_norm": 0.2952471971511841, "learning_rate": 6.201563762779852e-05, "loss": 0.0207, "step": 36130 }, { "grad_norm": 0.29537564516067505, "learning_rate": 6.199557390511538e-05, "loss": 0.0212, "step": 36140 }, { "grad_norm": 0.4836036264896393, "learning_rate": 6.197550813271675e-05, "loss": 0.0207, "step": 36150 }, { "grad_norm": 0.2730239927768707, "learning_rate": 6.195544031403131e-05, "loss": 0.0233, "step": 36160 }, { "grad_norm": 0.30656903982162476, "learning_rate": 6.19353704524881e-05, "loss": 0.0225, "step": 36170 }, { "grad_norm": 0.3383297920227051, "learning_rate": 6.191529855151652e-05, "loss": 0.0185, "step": 36180 }, { "grad_norm": 0.31547093391418457, "learning_rate": 6.189522461454629e-05, "loss": 0.0244, "step": 36190 }, { "grad_norm": 0.28414174914360046, "learning_rate": 6.187514864500752e-05, "loss": 0.0255, "step": 36200 }, { "grad_norm": 0.4903146028518677, "learning_rate": 6.185507064633062e-05, "loss": 0.0265, "step": 36210 }, { "grad_norm": 0.48934730887413025, "learning_rate": 6.18349906219464e-05, "loss": 0.0231, "step": 36220 }, { "grad_norm": 0.5080440640449524, "learning_rate": 6.181490857528596e-05, "loss": 0.0224, "step": 36230 }, { "grad_norm": 0.205324187874794, "learning_rate": 6.179482450978077e-05, "loss": 0.0159, "step": 36240 }, { "grad_norm": 0.3666606843471527, "learning_rate": 6.177473842886269e-05, "loss": 0.0228, "step": 36250 }, { "grad_norm": 0.2535434067249298, "learning_rate": 6.175465033596382e-05, "loss": 0.0224, "step": 36260 }, { "grad_norm": 0.19554445147514343, "learning_rate": 6.173456023451671e-05, "loss": 0.0213, "step": 36270 }, { "grad_norm": 0.31430837512016296, "learning_rate": 6.171446812795422e-05, "loss": 0.0222, "step": 36280 }, { "grad_norm": 0.25850096344947815, "learning_rate": 6.169437401970949e-05, "loss": 0.033, "step": 36290 }, { "grad_norm": 0.20970207452774048, "learning_rate": 6.16742779132161e-05, "loss": 0.0181, "step": 36300 }, { "grad_norm": 0.2920694649219513, "learning_rate": 6.165417981190789e-05, "loss": 0.0179, "step": 36310 }, { "grad_norm": 0.31229710578918457, "learning_rate": 6.16340797192191e-05, "loss": 0.0201, "step": 36320 }, { "grad_norm": 0.2439647614955902, "learning_rate": 6.161397763858427e-05, "loss": 0.0183, "step": 36330 }, { "grad_norm": 0.28965693712234497, "learning_rate": 6.159387357343834e-05, "loss": 0.0184, "step": 36340 }, { "grad_norm": 0.2895953357219696, "learning_rate": 6.157376752721648e-05, "loss": 0.0217, "step": 36350 }, { "grad_norm": 0.336684912443161, "learning_rate": 6.155365950335428e-05, "loss": 0.023, "step": 36360 }, { "grad_norm": 0.36071792244911194, "learning_rate": 6.153354950528768e-05, "loss": 0.0177, "step": 36370 }, { "grad_norm": 0.31727123260498047, "learning_rate": 6.151343753645293e-05, "loss": 0.0197, "step": 36380 }, { "grad_norm": 0.24269647896289825, "learning_rate": 6.149332360028657e-05, "loss": 0.0172, "step": 36390 }, { "grad_norm": 0.24083329737186432, "learning_rate": 6.147320770022555e-05, "loss": 0.0201, "step": 36400 }, { "grad_norm": 0.21403783559799194, "learning_rate": 6.145308983970715e-05, "loss": 0.0173, "step": 36410 }, { "grad_norm": 0.3848820924758911, "learning_rate": 6.143297002216892e-05, "loss": 0.0183, "step": 36420 }, { "grad_norm": 0.35667285323143005, "learning_rate": 6.141284825104882e-05, "loss": 0.0249, "step": 36430 }, { "grad_norm": 0.36606451869010925, "learning_rate": 6.13927245297851e-05, "loss": 0.0243, "step": 36440 }, { "grad_norm": 0.32982733845710754, "learning_rate": 6.137259886181633e-05, "loss": 0.024, "step": 36450 }, { "grad_norm": 0.24670782685279846, "learning_rate": 6.135247125058145e-05, "loss": 0.02, "step": 36460 }, { "grad_norm": 0.3490044176578522, "learning_rate": 6.133234169951974e-05, "loss": 0.0128, "step": 36470 }, { "grad_norm": 0.278586208820343, "learning_rate": 6.131221021207078e-05, "loss": 0.0217, "step": 36480 }, { "grad_norm": 0.28707897663116455, "learning_rate": 6.129207679167448e-05, "loss": 0.0183, "step": 36490 }, { "grad_norm": 0.2737930715084076, "learning_rate": 6.127194144177109e-05, "loss": 0.0206, "step": 36500 }, { "grad_norm": 0.3139075040817261, "learning_rate": 6.125180416580118e-05, "loss": 0.016, "step": 36510 }, { "grad_norm": 0.2937723994255066, "learning_rate": 6.123166496720571e-05, "loss": 0.0163, "step": 36520 }, { "grad_norm": 0.3580864667892456, "learning_rate": 6.121152384942588e-05, "loss": 0.0233, "step": 36530 }, { "grad_norm": 0.2841458320617676, "learning_rate": 6.119138081590324e-05, "loss": 0.0197, "step": 36540 }, { "grad_norm": 0.2920999526977539, "learning_rate": 6.117123587007971e-05, "loss": 0.0173, "step": 36550 }, { "grad_norm": 0.3852091133594513, "learning_rate": 6.11510890153975e-05, "loss": 0.0237, "step": 36560 }, { "grad_norm": 0.31379300355911255, "learning_rate": 6.113094025529916e-05, "loss": 0.0233, "step": 36570 }, { "grad_norm": 0.42604029178619385, "learning_rate": 6.111078959322757e-05, "loss": 0.0292, "step": 36580 }, { "grad_norm": 0.4220943748950958, "learning_rate": 6.109063703262592e-05, "loss": 0.0315, "step": 36590 }, { "grad_norm": 0.2605605125427246, "learning_rate": 6.107048257693772e-05, "loss": 0.0279, "step": 36600 }, { "grad_norm": 0.27759110927581787, "learning_rate": 6.105032622960683e-05, "loss": 0.0217, "step": 36610 }, { "grad_norm": 0.25994405150413513, "learning_rate": 6.103016799407743e-05, "loss": 0.0211, "step": 36620 }, { "grad_norm": 0.40303388237953186, "learning_rate": 6.1010007873793984e-05, "loss": 0.0251, "step": 36630 }, { "grad_norm": 0.2899770438671112, "learning_rate": 6.098984587220131e-05, "loss": 0.0214, "step": 36640 }, { "grad_norm": 0.271054744720459, "learning_rate": 6.096968199274456e-05, "loss": 0.0228, "step": 36650 }, { "grad_norm": 0.2793172597885132, "learning_rate": 6.0949516238869166e-05, "loss": 0.0227, "step": 36660 }, { "grad_norm": 0.35105639696121216, "learning_rate": 6.092934861402092e-05, "loss": 0.0234, "step": 36670 }, { "grad_norm": 0.337420254945755, "learning_rate": 6.0909179121645924e-05, "loss": 0.0265, "step": 36680 }, { "grad_norm": 0.3168783187866211, "learning_rate": 6.0889007765190576e-05, "loss": 0.0258, "step": 36690 }, { "grad_norm": 0.25392526388168335, "learning_rate": 6.086883454810162e-05, "loss": 0.0222, "step": 36700 }, { "grad_norm": 0.4887002408504486, "learning_rate": 6.0848659473826084e-05, "loss": 0.0257, "step": 36710 }, { "grad_norm": 0.3555664122104645, "learning_rate": 6.082848254581138e-05, "loss": 0.0232, "step": 36720 }, { "grad_norm": 0.36324837803840637, "learning_rate": 6.080830376750517e-05, "loss": 0.0256, "step": 36730 }, { "grad_norm": 0.17903640866279602, "learning_rate": 6.0788123142355445e-05, "loss": 0.0145, "step": 36740 }, { "grad_norm": 0.2591833770275116, "learning_rate": 6.076794067381052e-05, "loss": 0.023, "step": 36750 }, { "grad_norm": 0.24525919556617737, "learning_rate": 6.074775636531905e-05, "loss": 0.0232, "step": 36760 }, { "grad_norm": 0.42744988203048706, "learning_rate": 6.072757022032997e-05, "loss": 0.0272, "step": 36770 }, { "grad_norm": 0.3323257863521576, "learning_rate": 6.070738224229253e-05, "loss": 0.0324, "step": 36780 }, { "grad_norm": 0.2953006625175476, "learning_rate": 6.0687192434656314e-05, "loss": 0.0247, "step": 36790 }, { "grad_norm": 0.28503820300102234, "learning_rate": 6.066700080087121e-05, "loss": 0.019, "step": 36800 }, { "grad_norm": 0.3509652614593506, "learning_rate": 6.0646807344387424e-05, "loss": 0.0196, "step": 36810 }, { "grad_norm": 0.4045184254646301, "learning_rate": 6.062661206865543e-05, "loss": 0.0258, "step": 36820 }, { "grad_norm": 0.2528630793094635, "learning_rate": 6.06064149771261e-05, "loss": 0.0223, "step": 36830 }, { "grad_norm": 0.7950335144996643, "learning_rate": 6.058621607325051e-05, "loss": 0.0199, "step": 36840 }, { "grad_norm": 0.30251431465148926, "learning_rate": 6.056601536048014e-05, "loss": 0.0201, "step": 36850 }, { "grad_norm": 0.3428458869457245, "learning_rate": 6.0545812842266725e-05, "loss": 0.0295, "step": 36860 }, { "grad_norm": 0.4866817593574524, "learning_rate": 6.052560852206232e-05, "loss": 0.0207, "step": 36870 }, { "grad_norm": 0.3339347839355469, "learning_rate": 6.05054024033193e-05, "loss": 0.0247, "step": 36880 }, { "grad_norm": 0.3301037847995758, "learning_rate": 6.048519448949032e-05, "loss": 0.021, "step": 36890 }, { "grad_norm": 0.24881848692893982, "learning_rate": 6.046498478402839e-05, "loss": 0.0232, "step": 36900 }, { "grad_norm": 0.31135061383247375, "learning_rate": 6.044477329038677e-05, "loss": 0.0188, "step": 36910 }, { "grad_norm": 0.3849307894706726, "learning_rate": 6.042456001201906e-05, "loss": 0.0228, "step": 36920 }, { "grad_norm": 0.3151019215583801, "learning_rate": 6.040434495237917e-05, "loss": 0.0194, "step": 36930 }, { "grad_norm": 0.46809691190719604, "learning_rate": 6.0384128114921256e-05, "loss": 0.0223, "step": 36940 }, { "grad_norm": 0.3122335970401764, "learning_rate": 6.036390950309987e-05, "loss": 0.0166, "step": 36950 }, { "grad_norm": 0.2812850773334503, "learning_rate": 6.0343689120369805e-05, "loss": 0.0206, "step": 36960 }, { "grad_norm": 0.3210470378398895, "learning_rate": 6.032346697018616e-05, "loss": 0.0232, "step": 36970 }, { "grad_norm": 0.28869935870170593, "learning_rate": 6.0303243056004375e-05, "loss": 0.0331, "step": 36980 }, { "grad_norm": 0.25084003806114197, "learning_rate": 6.0283017381280136e-05, "loss": 0.0201, "step": 36990 }, { "grad_norm": 0.29113471508026123, "learning_rate": 6.026278994946948e-05, "loss": 0.0198, "step": 37000 }, { "grad_norm": 0.21933071315288544, "learning_rate": 6.02425607640287e-05, "loss": 0.023, "step": 37010 }, { "grad_norm": 0.3928924798965454, "learning_rate": 6.022232982841441e-05, "loss": 0.0231, "step": 37020 }, { "grad_norm": 0.24488502740859985, "learning_rate": 6.020209714608355e-05, "loss": 0.0203, "step": 37030 }, { "grad_norm": 0.34892603754997253, "learning_rate": 6.018186272049331e-05, "loss": 0.0227, "step": 37040 }, { "grad_norm": 0.20007576048374176, "learning_rate": 6.01616265551012e-05, "loss": 0.0319, "step": 37050 }, { "grad_norm": 0.2847561538219452, "learning_rate": 6.014138865336503e-05, "loss": 0.0207, "step": 37060 }, { "grad_norm": 0.26656800508499146, "learning_rate": 6.0121149018742905e-05, "loss": 0.0148, "step": 37070 }, { "grad_norm": 0.25111818313598633, "learning_rate": 6.010090765469325e-05, "loss": 0.0234, "step": 37080 }, { "grad_norm": 0.28755345940589905, "learning_rate": 6.008066456467473e-05, "loss": 0.0158, "step": 37090 }, { "grad_norm": 0.13918209075927734, "learning_rate": 6.0060419752146335e-05, "loss": 0.0183, "step": 37100 }, { "grad_norm": 0.4199187755584717, "learning_rate": 6.0040173220567353e-05, "loss": 0.0208, "step": 37110 }, { "grad_norm": 0.2276892513036728, "learning_rate": 6.001992497339737e-05, "loss": 0.0206, "step": 37120 }, { "grad_norm": 0.25403377413749695, "learning_rate": 5.999967501409626e-05, "loss": 0.0241, "step": 37130 }, { "grad_norm": 0.19315043091773987, "learning_rate": 5.997942334612418e-05, "loss": 0.0203, "step": 37140 }, { "grad_norm": 0.23553594946861267, "learning_rate": 5.995916997294158e-05, "loss": 0.0165, "step": 37150 }, { "grad_norm": 0.2738024592399597, "learning_rate": 5.9938914898009214e-05, "loss": 0.0234, "step": 37160 }, { "grad_norm": 0.40955445170402527, "learning_rate": 5.991865812478813e-05, "loss": 0.0179, "step": 37170 }, { "grad_norm": 0.3674248456954956, "learning_rate": 5.989839965673964e-05, "loss": 0.0327, "step": 37180 }, { "grad_norm": 0.38883885741233826, "learning_rate": 5.987813949732539e-05, "loss": 0.0283, "step": 37190 }, { "grad_norm": 0.35835781693458557, "learning_rate": 5.9857877650007255e-05, "loss": 0.0237, "step": 37200 }, { "grad_norm": 0.2580469250679016, "learning_rate": 5.983761411824744e-05, "loss": 0.019, "step": 37210 }, { "grad_norm": 0.28503161668777466, "learning_rate": 5.981734890550844e-05, "loss": 0.0234, "step": 37220 }, { "grad_norm": 0.45360228419303894, "learning_rate": 5.979708201525301e-05, "loss": 0.0274, "step": 37230 }, { "grad_norm": 0.24881204962730408, "learning_rate": 5.977681345094422e-05, "loss": 0.021, "step": 37240 }, { "grad_norm": 0.23251774907112122, "learning_rate": 5.97565432160454e-05, "loss": 0.0193, "step": 37250 }, { "grad_norm": 0.23550494015216827, "learning_rate": 5.9736271314020186e-05, "loss": 0.0158, "step": 37260 }, { "grad_norm": 0.2316167801618576, "learning_rate": 5.971599774833251e-05, "loss": 0.0203, "step": 37270 }, { "grad_norm": 0.24204875528812408, "learning_rate": 5.9695722522446525e-05, "loss": 0.0229, "step": 37280 }, { "grad_norm": 0.2954282760620117, "learning_rate": 5.9675445639826765e-05, "loss": 0.0207, "step": 37290 }, { "grad_norm": 0.2898028492927551, "learning_rate": 5.965516710393796e-05, "loss": 0.0171, "step": 37300 }, { "grad_norm": 0.2497636079788208, "learning_rate": 5.963488691824516e-05, "loss": 0.0199, "step": 37310 }, { "grad_norm": 0.3524884283542633, "learning_rate": 5.96146050862137e-05, "loss": 0.021, "step": 37320 }, { "grad_norm": 0.22824092209339142, "learning_rate": 5.959432161130919e-05, "loss": 0.0201, "step": 37330 }, { "grad_norm": 0.41760754585266113, "learning_rate": 5.9574036496997545e-05, "loss": 0.0209, "step": 37340 }, { "grad_norm": 0.2541232109069824, "learning_rate": 5.955374974674488e-05, "loss": 0.0244, "step": 37350 }, { "grad_norm": 0.19595466554164886, "learning_rate": 5.9533461364017696e-05, "loss": 0.0203, "step": 37360 }, { "grad_norm": 0.20172613859176636, "learning_rate": 5.9513171352282716e-05, "loss": 0.0158, "step": 37370 }, { "grad_norm": 0.22450365126132965, "learning_rate": 5.949287971500692e-05, "loss": 0.0187, "step": 37380 }, { "grad_norm": 0.2044125497341156, "learning_rate": 5.947258645565762e-05, "loss": 0.0148, "step": 37390 }, { "grad_norm": 0.13590025901794434, "learning_rate": 5.945229157770237e-05, "loss": 0.0161, "step": 37400 }, { "grad_norm": 0.20000742375850677, "learning_rate": 5.9431995084609006e-05, "loss": 0.0209, "step": 37410 }, { "grad_norm": 0.273252934217453, "learning_rate": 5.941169697984564e-05, "loss": 0.0169, "step": 37420 }, { "grad_norm": 0.40201064944267273, "learning_rate": 5.9391397266880675e-05, "loss": 0.0212, "step": 37430 }, { "grad_norm": 0.4173870384693146, "learning_rate": 5.937109594918279e-05, "loss": 0.0361, "step": 37440 }, { "grad_norm": 0.29273462295532227, "learning_rate": 5.9350793030220884e-05, "loss": 0.0298, "step": 37450 }, { "grad_norm": 0.19689223170280457, "learning_rate": 5.933048851346421e-05, "loss": 0.0192, "step": 37460 }, { "grad_norm": 0.42805173993110657, "learning_rate": 5.931018240238224e-05, "loss": 0.0402, "step": 37470 }, { "grad_norm": 0.27330824732780457, "learning_rate": 5.928987470044471e-05, "loss": 0.0206, "step": 37480 }, { "grad_norm": 0.36891141533851624, "learning_rate": 5.9269565411121695e-05, "loss": 0.024, "step": 37490 }, { "grad_norm": 0.30635499954223633, "learning_rate": 5.924925453788347e-05, "loss": 0.0225, "step": 37500 }, { "grad_norm": 0.24067901074886322, "learning_rate": 5.92289420842006e-05, "loss": 0.025, "step": 37510 }, { "grad_norm": 0.29332199692726135, "learning_rate": 5.9208628053543945e-05, "loss": 0.0309, "step": 37520 }, { "grad_norm": 0.3792671859264374, "learning_rate": 5.918831244938462e-05, "loss": 0.0223, "step": 37530 }, { "grad_norm": 0.5436853170394897, "learning_rate": 5.9167995275194e-05, "loss": 0.0249, "step": 37540 }, { "grad_norm": 0.34302854537963867, "learning_rate": 5.914767653444373e-05, "loss": 0.0228, "step": 37550 }, { "grad_norm": 0.45328840613365173, "learning_rate": 5.912735623060572e-05, "loss": 0.0277, "step": 37560 }, { "grad_norm": 0.32582640647888184, "learning_rate": 5.910703436715217e-05, "loss": 0.0174, "step": 37570 }, { "grad_norm": 0.29951369762420654, "learning_rate": 5.908671094755552e-05, "loss": 0.0213, "step": 37580 }, { "grad_norm": 0.21626843512058258, "learning_rate": 5.906638597528851e-05, "loss": 0.0224, "step": 37590 }, { "grad_norm": 0.2937086522579193, "learning_rate": 5.9046059453824076e-05, "loss": 0.0209, "step": 37600 }, { "grad_norm": 0.2793009579181671, "learning_rate": 5.9025731386635505e-05, "loss": 0.0185, "step": 37610 }, { "grad_norm": 0.3561263382434845, "learning_rate": 5.900540177719629e-05, "loss": 0.017, "step": 37620 }, { "grad_norm": 0.3506627082824707, "learning_rate": 5.898507062898021e-05, "loss": 0.0198, "step": 37630 }, { "grad_norm": 0.3570871651172638, "learning_rate": 5.8964737945461316e-05, "loss": 0.0205, "step": 37640 }, { "grad_norm": 0.14492595195770264, "learning_rate": 5.8944403730113885e-05, "loss": 0.0188, "step": 37650 }, { "grad_norm": 0.38884490728378296, "learning_rate": 5.892406798641248e-05, "loss": 0.0269, "step": 37660 }, { "grad_norm": 0.21060962975025177, "learning_rate": 5.890373071783193e-05, "loss": 0.0196, "step": 37670 }, { "grad_norm": 0.34786155819892883, "learning_rate": 5.888339192784732e-05, "loss": 0.022, "step": 37680 }, { "grad_norm": 0.3348920941352844, "learning_rate": 5.8863051619934003e-05, "loss": 0.0206, "step": 37690 }, { "grad_norm": 0.26025545597076416, "learning_rate": 5.8842709797567554e-05, "loss": 0.0191, "step": 37700 }, { "grad_norm": 0.3013414144515991, "learning_rate": 5.8822366464223855e-05, "loss": 0.0192, "step": 37710 }, { "grad_norm": 0.3156901001930237, "learning_rate": 5.880202162337901e-05, "loss": 0.0225, "step": 37720 }, { "grad_norm": 0.31092768907546997, "learning_rate": 5.8781675278509405e-05, "loss": 0.0163, "step": 37730 }, { "grad_norm": 0.22484198212623596, "learning_rate": 5.8761327433091696e-05, "loss": 0.0194, "step": 37740 }, { "grad_norm": 0.6868064999580383, "learning_rate": 5.874097809060275e-05, "loss": 0.0306, "step": 37750 }, { "grad_norm": 0.30809563398361206, "learning_rate": 5.87206272545197e-05, "loss": 0.0327, "step": 37760 }, { "grad_norm": 0.37465956807136536, "learning_rate": 5.8700274928319955e-05, "loss": 0.0276, "step": 37770 }, { "grad_norm": 1.2216256856918335, "learning_rate": 5.867992111548118e-05, "loss": 0.0287, "step": 37780 }, { "grad_norm": 0.4081265330314636, "learning_rate": 5.865956581948131e-05, "loss": 0.0238, "step": 37790 }, { "grad_norm": 0.27294686436653137, "learning_rate": 5.863920904379845e-05, "loss": 0.0223, "step": 37800 }, { "grad_norm": 0.3155117332935333, "learning_rate": 5.8618850791911064e-05, "loss": 0.0219, "step": 37810 }, { "grad_norm": 0.2579362988471985, "learning_rate": 5.859849106729779e-05, "loss": 0.0226, "step": 37820 }, { "grad_norm": 0.2877509295940399, "learning_rate": 5.857812987343758e-05, "loss": 0.0239, "step": 37830 }, { "grad_norm": 0.2886989712715149, "learning_rate": 5.855776721380957e-05, "loss": 0.0224, "step": 37840 }, { "grad_norm": 0.2231038361787796, "learning_rate": 5.8537403091893217e-05, "loss": 0.0204, "step": 37850 }, { "grad_norm": 0.23099790513515472, "learning_rate": 5.851703751116816e-05, "loss": 0.0192, "step": 37860 }, { "grad_norm": 0.17209596931934357, "learning_rate": 5.8496670475114336e-05, "loss": 0.0172, "step": 37870 }, { "grad_norm": 0.20563286542892456, "learning_rate": 5.84763019872119e-05, "loss": 0.0286, "step": 37880 }, { "grad_norm": 0.3300706446170807, "learning_rate": 5.845593205094131e-05, "loss": 0.0179, "step": 37890 }, { "grad_norm": 0.30793455243110657, "learning_rate": 5.843556066978318e-05, "loss": 0.0221, "step": 37900 }, { "grad_norm": 0.19126956164836884, "learning_rate": 5.8415187847218455e-05, "loss": 0.0229, "step": 37910 }, { "grad_norm": 0.21689951419830322, "learning_rate": 5.839481358672827e-05, "loss": 0.0263, "step": 37920 }, { "grad_norm": 0.261201411485672, "learning_rate": 5.837443789179407e-05, "loss": 0.0187, "step": 37930 }, { "grad_norm": 0.3820212483406067, "learning_rate": 5.8354060765897445e-05, "loss": 0.0261, "step": 37940 }, { "grad_norm": 0.25009435415267944, "learning_rate": 5.8333682212520334e-05, "loss": 0.0327, "step": 37950 }, { "grad_norm": 0.35701534152030945, "learning_rate": 5.831330223514486e-05, "loss": 0.0245, "step": 37960 }, { "grad_norm": 0.2648361623287201, "learning_rate": 5.8292920837253396e-05, "loss": 0.0241, "step": 37970 }, { "grad_norm": 0.46039891242980957, "learning_rate": 5.827253802232857e-05, "loss": 0.0183, "step": 37980 }, { "grad_norm": 0.33936336636543274, "learning_rate": 5.825215379385327e-05, "loss": 0.0209, "step": 37990 }, { "grad_norm": 0.41061481833457947, "learning_rate": 5.823176815531057e-05, "loss": 0.0336, "step": 38000 }, { "grad_norm": 0.24711522459983826, "learning_rate": 5.8211381110183826e-05, "loss": 0.0288, "step": 38010 }, { "grad_norm": 0.26940008997917175, "learning_rate": 5.8190992661956645e-05, "loss": 0.0178, "step": 38020 }, { "grad_norm": 0.23038558661937714, "learning_rate": 5.817060281411284e-05, "loss": 0.0228, "step": 38030 }, { "grad_norm": 0.38805583119392395, "learning_rate": 5.815021157013647e-05, "loss": 0.0193, "step": 38040 }, { "grad_norm": 0.39668411016464233, "learning_rate": 5.8129818933511856e-05, "loss": 0.03, "step": 38050 }, { "grad_norm": 0.3749624490737915, "learning_rate": 5.8109424907723544e-05, "loss": 0.0241, "step": 38060 }, { "grad_norm": 0.27138635516166687, "learning_rate": 5.80890294962563e-05, "loss": 0.0196, "step": 38070 }, { "grad_norm": 0.5572101473808289, "learning_rate": 5.806863270259515e-05, "loss": 0.0225, "step": 38080 }, { "grad_norm": 0.2944258749485016, "learning_rate": 5.804823453022536e-05, "loss": 0.0186, "step": 38090 }, { "grad_norm": 0.2985639274120331, "learning_rate": 5.80278349826324e-05, "loss": 0.0223, "step": 38100 }, { "grad_norm": 0.2248975932598114, "learning_rate": 5.8007434063302014e-05, "loss": 0.0207, "step": 38110 }, { "grad_norm": 0.27379414439201355, "learning_rate": 5.7987031775720136e-05, "loss": 0.0177, "step": 38120 }, { "grad_norm": 0.252628356218338, "learning_rate": 5.7966628123372976e-05, "loss": 0.0245, "step": 38130 }, { "grad_norm": 0.47547370195388794, "learning_rate": 5.7946223109746956e-05, "loss": 0.0276, "step": 38140 }, { "grad_norm": 0.42386436462402344, "learning_rate": 5.7925816738328754e-05, "loss": 0.0326, "step": 38150 }, { "grad_norm": 0.5938906669616699, "learning_rate": 5.790540901260521e-05, "loss": 0.0285, "step": 38160 }, { "grad_norm": 0.4401269555091858, "learning_rate": 5.788499993606351e-05, "loss": 0.0253, "step": 38170 }, { "grad_norm": 0.26790985465049744, "learning_rate": 5.786458951219096e-05, "loss": 0.025, "step": 38180 }, { "grad_norm": 0.4164589047431946, "learning_rate": 5.784417774447517e-05, "loss": 0.0301, "step": 38190 }, { "grad_norm": 0.3225284516811371, "learning_rate": 5.782376463640393e-05, "loss": 0.0253, "step": 38200 }, { "grad_norm": 0.40520811080932617, "learning_rate": 5.780335019146531e-05, "loss": 0.0201, "step": 38210 }, { "grad_norm": 0.539462149143219, "learning_rate": 5.778293441314755e-05, "loss": 0.024, "step": 38220 }, { "grad_norm": 0.4016970694065094, "learning_rate": 5.776251730493917e-05, "loss": 0.0276, "step": 38230 }, { "grad_norm": 0.1923786699771881, "learning_rate": 5.774209887032887e-05, "loss": 0.0262, "step": 38240 }, { "grad_norm": 0.3441394567489624, "learning_rate": 5.772167911280565e-05, "loss": 0.0202, "step": 38250 }, { "grad_norm": 0.2763516306877136, "learning_rate": 5.770125803585864e-05, "loss": 0.0177, "step": 38260 }, { "grad_norm": 0.3454493582248688, "learning_rate": 5.768083564297726e-05, "loss": 0.0272, "step": 38270 }, { "grad_norm": 0.3312123119831085, "learning_rate": 5.766041193765114e-05, "loss": 0.0273, "step": 38280 }, { "grad_norm": 0.30930688977241516, "learning_rate": 5.763998692337015e-05, "loss": 0.0233, "step": 38290 }, { "grad_norm": 0.35758692026138306, "learning_rate": 5.761956060362433e-05, "loss": 0.0257, "step": 38300 }, { "grad_norm": 0.28022879362106323, "learning_rate": 5.7599132981904e-05, "loss": 0.0238, "step": 38310 }, { "grad_norm": 0.2786475419998169, "learning_rate": 5.75787040616997e-05, "loss": 0.0212, "step": 38320 }, { "grad_norm": 0.2944768965244293, "learning_rate": 5.755827384650212e-05, "loss": 0.0204, "step": 38330 }, { "grad_norm": 0.3884202837944031, "learning_rate": 5.753784233980228e-05, "loss": 0.0246, "step": 38340 }, { "grad_norm": 0.1726434975862503, "learning_rate": 5.751740954509135e-05, "loss": 0.016, "step": 38350 }, { "grad_norm": 0.5410749912261963, "learning_rate": 5.7496975465860715e-05, "loss": 0.0152, "step": 38360 }, { "grad_norm": 0.22918100655078888, "learning_rate": 5.747654010560202e-05, "loss": 0.0192, "step": 38370 }, { "grad_norm": 0.24630475044250488, "learning_rate": 5.7456103467807097e-05, "loss": 0.0254, "step": 38380 }, { "grad_norm": 0.2374948263168335, "learning_rate": 5.7435665555968046e-05, "loss": 0.0233, "step": 38390 }, { "grad_norm": 0.21577592194080353, "learning_rate": 5.74152263735771e-05, "loss": 0.0224, "step": 38400 }, { "grad_norm": 0.2259528636932373, "learning_rate": 5.739478592412677e-05, "loss": 0.0231, "step": 38410 }, { "grad_norm": 0.4233595132827759, "learning_rate": 5.7374344211109766e-05, "loss": 0.0288, "step": 38420 }, { "grad_norm": 0.27331218123435974, "learning_rate": 5.735390123801904e-05, "loss": 0.0228, "step": 38430 }, { "grad_norm": 0.3408249318599701, "learning_rate": 5.7333457008347704e-05, "loss": 0.0314, "step": 38440 }, { "grad_norm": 0.3702116012573242, "learning_rate": 5.7313011525589156e-05, "loss": 0.0248, "step": 38450 }, { "grad_norm": 0.35212182998657227, "learning_rate": 5.729256479323694e-05, "loss": 0.0202, "step": 38460 }, { "grad_norm": 0.3397960662841797, "learning_rate": 5.727211681478485e-05, "loss": 0.0246, "step": 38470 }, { "grad_norm": 0.2457178682088852, "learning_rate": 5.7251667593726886e-05, "loss": 0.0271, "step": 38480 }, { "grad_norm": 0.2503441274166107, "learning_rate": 5.723121713355728e-05, "loss": 0.0316, "step": 38490 }, { "grad_norm": 0.24600882828235626, "learning_rate": 5.721076543777044e-05, "loss": 0.0219, "step": 38500 }, { "grad_norm": 0.382161945104599, "learning_rate": 5.7190312509860986e-05, "loss": 0.0207, "step": 38510 }, { "grad_norm": 0.4625328481197357, "learning_rate": 5.716985835332379e-05, "loss": 0.027, "step": 38520 }, { "grad_norm": 0.26299068331718445, "learning_rate": 5.714940297165389e-05, "loss": 0.0208, "step": 38530 }, { "grad_norm": 0.303670734167099, "learning_rate": 5.712894636834656e-05, "loss": 0.0222, "step": 38540 }, { "grad_norm": 0.218377485871315, "learning_rate": 5.7108488546897276e-05, "loss": 0.0196, "step": 38550 }, { "grad_norm": 0.16988055408000946, "learning_rate": 5.708802951080172e-05, "loss": 0.0183, "step": 38560 }, { "grad_norm": 0.23917867243289948, "learning_rate": 5.706756926355576e-05, "loss": 0.0222, "step": 38570 }, { "grad_norm": 0.28335636854171753, "learning_rate": 5.704710780865554e-05, "loss": 0.0224, "step": 38580 }, { "grad_norm": 0.20780609548091888, "learning_rate": 5.7026645149597325e-05, "loss": 0.0215, "step": 38590 }, { "grad_norm": 0.3669348657131195, "learning_rate": 5.700618128987764e-05, "loss": 0.0301, "step": 38600 }, { "grad_norm": 0.2100071758031845, "learning_rate": 5.698571623299317e-05, "loss": 0.02, "step": 38610 }, { "grad_norm": 0.14342956244945526, "learning_rate": 5.696524998244086e-05, "loss": 0.0161, "step": 38620 }, { "grad_norm": 0.3228890597820282, "learning_rate": 5.6944782541717836e-05, "loss": 0.0277, "step": 38630 }, { "grad_norm": 0.260946661233902, "learning_rate": 5.69243139143214e-05, "loss": 0.0181, "step": 38640 }, { "grad_norm": 0.18736632168293, "learning_rate": 5.6903844103749125e-05, "loss": 0.022, "step": 38650 }, { "grad_norm": 0.27268457412719727, "learning_rate": 5.688337311349869e-05, "loss": 0.0183, "step": 38660 }, { "grad_norm": 0.32519474625587463, "learning_rate": 5.6862900947068074e-05, "loss": 0.0204, "step": 38670 }, { "grad_norm": 0.2804412841796875, "learning_rate": 5.6842427607955374e-05, "loss": 0.0283, "step": 38680 }, { "grad_norm": 0.2030884325504303, "learning_rate": 5.682195309965893e-05, "loss": 0.0199, "step": 38690 }, { "grad_norm": 0.3223126232624054, "learning_rate": 5.6801477425677294e-05, "loss": 0.0223, "step": 38700 }, { "grad_norm": 0.3995254337787628, "learning_rate": 5.678100058950917e-05, "loss": 0.0319, "step": 38710 }, { "grad_norm": 0.5896979570388794, "learning_rate": 5.676052259465352e-05, "loss": 0.0208, "step": 38720 }, { "grad_norm": 0.32085373997688293, "learning_rate": 5.674004344460945e-05, "loss": 0.0216, "step": 38730 }, { "grad_norm": 0.33164116740226746, "learning_rate": 5.6719563142876295e-05, "loss": 0.0193, "step": 38740 }, { "grad_norm": 0.4026186466217041, "learning_rate": 5.669908169295359e-05, "loss": 0.0226, "step": 38750 }, { "grad_norm": 0.27943816781044006, "learning_rate": 5.667859909834105e-05, "loss": 0.0206, "step": 38760 }, { "grad_norm": 0.25592368841171265, "learning_rate": 5.6658115362538565e-05, "loss": 0.016, "step": 38770 }, { "grad_norm": 0.2923930287361145, "learning_rate": 5.663763048904628e-05, "loss": 0.0179, "step": 38780 }, { "grad_norm": 0.3899855613708496, "learning_rate": 5.661714448136447e-05, "loss": 0.0167, "step": 38790 }, { "grad_norm": 0.37466275691986084, "learning_rate": 5.659665734299366e-05, "loss": 0.016, "step": 38800 }, { "grad_norm": 0.32841333746910095, "learning_rate": 5.6576169077434516e-05, "loss": 0.0231, "step": 38810 }, { "grad_norm": 0.13770630955696106, "learning_rate": 5.6555679688187944e-05, "loss": 0.0216, "step": 38820 }, { "grad_norm": 0.3982064127922058, "learning_rate": 5.6535189178755e-05, "loss": 0.0193, "step": 38830 }, { "grad_norm": 0.31626614928245544, "learning_rate": 5.651469755263698e-05, "loss": 0.0191, "step": 38840 }, { "grad_norm": 0.1822403073310852, "learning_rate": 5.6494204813335316e-05, "loss": 0.0254, "step": 38850 }, { "grad_norm": 0.2965816259384155, "learning_rate": 5.647371096435168e-05, "loss": 0.0158, "step": 38860 }, { "grad_norm": 0.17396007478237152, "learning_rate": 5.645321600918788e-05, "loss": 0.0185, "step": 38870 }, { "grad_norm": 0.28731128573417664, "learning_rate": 5.643271995134597e-05, "loss": 0.0252, "step": 38880 }, { "grad_norm": 0.3192174434661865, "learning_rate": 5.641222279432814e-05, "loss": 0.0248, "step": 38890 }, { "grad_norm": 0.31229013204574585, "learning_rate": 5.6391724541636834e-05, "loss": 0.0186, "step": 38900 }, { "grad_norm": 0.5208616256713867, "learning_rate": 5.6371225196774605e-05, "loss": 0.0236, "step": 38910 }, { "grad_norm": 0.31571775674819946, "learning_rate": 5.635072476324423e-05, "loss": 0.024, "step": 38920 }, { "grad_norm": 0.34710440039634705, "learning_rate": 5.63302232445487e-05, "loss": 0.0188, "step": 38930 }, { "grad_norm": 0.21082106232643127, "learning_rate": 5.6309720644191144e-05, "loss": 0.0175, "step": 38940 }, { "grad_norm": 0.4344130754470825, "learning_rate": 5.628921696567491e-05, "loss": 0.0193, "step": 38950 }, { "grad_norm": 0.36401835083961487, "learning_rate": 5.62687122125035e-05, "loss": 0.0195, "step": 38960 }, { "grad_norm": 0.3969275951385498, "learning_rate": 5.624820638818062e-05, "loss": 0.0164, "step": 38970 }, { "grad_norm": 0.23333407938480377, "learning_rate": 5.6227699496210164e-05, "loss": 0.0223, "step": 38980 }, { "grad_norm": 0.22841499745845795, "learning_rate": 5.6207191540096195e-05, "loss": 0.0218, "step": 38990 }, { "grad_norm": 0.23751460015773773, "learning_rate": 5.618668252334296e-05, "loss": 0.0182, "step": 39000 }, { "grad_norm": 0.23441413044929504, "learning_rate": 5.616617244945488e-05, "loss": 0.0209, "step": 39010 }, { "grad_norm": 0.2767494320869446, "learning_rate": 5.614566132193656e-05, "loss": 0.0244, "step": 39020 }, { "grad_norm": 0.17533642053604126, "learning_rate": 5.612514914429282e-05, "loss": 0.0208, "step": 39030 }, { "grad_norm": 0.4266168177127838, "learning_rate": 5.610463592002863e-05, "loss": 0.018, "step": 39040 }, { "grad_norm": 0.2941049337387085, "learning_rate": 5.608412165264909e-05, "loss": 0.0193, "step": 39050 }, { "grad_norm": 0.23750922083854675, "learning_rate": 5.606360634565959e-05, "loss": 0.0164, "step": 39060 }, { "grad_norm": 0.2375033050775528, "learning_rate": 5.604309000256559e-05, "loss": 0.0207, "step": 39070 }, { "grad_norm": 0.2408188283443451, "learning_rate": 5.6022572626872785e-05, "loss": 0.0166, "step": 39080 }, { "grad_norm": 0.3110527992248535, "learning_rate": 5.600205422208704e-05, "loss": 0.0191, "step": 39090 }, { "grad_norm": 0.25859129428863525, "learning_rate": 5.59815347917144e-05, "loss": 0.0222, "step": 39100 }, { "grad_norm": 0.2960147559642792, "learning_rate": 5.596101433926103e-05, "loss": 0.019, "step": 39110 }, { "grad_norm": 0.25229567289352417, "learning_rate": 5.5940492868233364e-05, "loss": 0.0156, "step": 39120 }, { "grad_norm": 0.3176618218421936, "learning_rate": 5.591997038213793e-05, "loss": 0.0197, "step": 39130 }, { "grad_norm": 0.2468293309211731, "learning_rate": 5.5899446884481475e-05, "loss": 0.0166, "step": 39140 }, { "grad_norm": 0.38274869322776794, "learning_rate": 5.5878922378770906e-05, "loss": 0.0299, "step": 39150 }, { "grad_norm": 0.3032037019729614, "learning_rate": 5.5858396868513285e-05, "loss": 0.0171, "step": 39160 }, { "grad_norm": 0.32302555441856384, "learning_rate": 5.583787035721586e-05, "loss": 0.022, "step": 39170 }, { "grad_norm": 0.2340671271085739, "learning_rate": 5.581734284838606e-05, "loss": 0.0148, "step": 39180 }, { "grad_norm": 0.3900830149650574, "learning_rate": 5.579681434553147e-05, "loss": 0.018, "step": 39190 }, { "grad_norm": 0.21493452787399292, "learning_rate": 5.5776284852159854e-05, "loss": 0.0152, "step": 39200 }, { "grad_norm": 0.16077068448066711, "learning_rate": 5.575575437177913e-05, "loss": 0.021, "step": 39210 }, { "grad_norm": 0.20227409899234772, "learning_rate": 5.573522290789742e-05, "loss": 0.018, "step": 39220 }, { "grad_norm": 0.30448442697525024, "learning_rate": 5.571469046402298e-05, "loss": 0.0163, "step": 39230 }, { "grad_norm": 0.25324440002441406, "learning_rate": 5.5694157043664205e-05, "loss": 0.0215, "step": 39240 }, { "grad_norm": 0.2421221137046814, "learning_rate": 5.567362265032975e-05, "loss": 0.0164, "step": 39250 }, { "grad_norm": 0.3529015779495239, "learning_rate": 5.565308728752836e-05, "loss": 0.0196, "step": 39260 }, { "grad_norm": 0.27083802223205566, "learning_rate": 5.5632550958768956e-05, "loss": 0.0207, "step": 39270 }, { "grad_norm": 0.4135487377643585, "learning_rate": 5.5612013667560636e-05, "loss": 0.0262, "step": 39280 }, { "grad_norm": 0.3096684515476227, "learning_rate": 5.5591475417412676e-05, "loss": 0.0362, "step": 39290 }, { "grad_norm": 0.32497668266296387, "learning_rate": 5.557093621183451e-05, "loss": 0.0236, "step": 39300 }, { "grad_norm": 0.2918798625469208, "learning_rate": 5.55503960543357e-05, "loss": 0.0213, "step": 39310 }, { "grad_norm": 0.29888659715652466, "learning_rate": 5.552985494842601e-05, "loss": 0.0231, "step": 39320 }, { "grad_norm": 0.22258083522319794, "learning_rate": 5.550931289761534e-05, "loss": 0.0194, "step": 39330 }, { "grad_norm": 0.4149337410926819, "learning_rate": 5.548876990541378e-05, "loss": 0.0213, "step": 39340 }, { "grad_norm": 0.41329243779182434, "learning_rate": 5.5468225975331564e-05, "loss": 0.0321, "step": 39350 }, { "grad_norm": 0.29073095321655273, "learning_rate": 5.544768111087909e-05, "loss": 0.019, "step": 39360 }, { "grad_norm": 0.21945127844810486, "learning_rate": 5.54271353155669e-05, "loss": 0.0219, "step": 39370 }, { "grad_norm": 0.20895880460739136, "learning_rate": 5.5406588592905715e-05, "loss": 0.0207, "step": 39380 }, { "grad_norm": 0.30517157912254333, "learning_rate": 5.5386040946406416e-05, "loss": 0.0261, "step": 39390 }, { "grad_norm": 0.3302744925022125, "learning_rate": 5.536549237958004e-05, "loss": 0.0285, "step": 39400 }, { "grad_norm": 0.1943526566028595, "learning_rate": 5.5344942895937744e-05, "loss": 0.0284, "step": 39410 }, { "grad_norm": 0.24070000648498535, "learning_rate": 5.5324392498990904e-05, "loss": 0.0193, "step": 39420 }, { "grad_norm": 0.4267045259475708, "learning_rate": 5.5303841192251e-05, "loss": 0.0198, "step": 39430 }, { "grad_norm": 0.31027570366859436, "learning_rate": 5.52832889792297e-05, "loss": 0.0256, "step": 39440 }, { "grad_norm": 0.23539865016937256, "learning_rate": 5.526273586343881e-05, "loss": 0.0212, "step": 39450 }, { "grad_norm": 0.32909178733825684, "learning_rate": 5.5242181848390306e-05, "loss": 0.016, "step": 39460 }, { "grad_norm": 0.29447442293167114, "learning_rate": 5.5221626937596285e-05, "loss": 0.0179, "step": 39470 }, { "grad_norm": 0.2248312383890152, "learning_rate": 5.520107113456903e-05, "loss": 0.0222, "step": 39480 }, { "grad_norm": 0.43922585248947144, "learning_rate": 5.5180514442820974e-05, "loss": 0.0242, "step": 39490 }, { "grad_norm": 0.20977894961833954, "learning_rate": 5.515995686586469e-05, "loss": 0.0143, "step": 39500 }, { "grad_norm": 0.18133790791034698, "learning_rate": 5.5139398407212916e-05, "loss": 0.0133, "step": 39510 }, { "grad_norm": 0.2762456238269806, "learning_rate": 5.511883907037849e-05, "loss": 0.0198, "step": 39520 }, { "grad_norm": 0.27382922172546387, "learning_rate": 5.509827885887449e-05, "loss": 0.0197, "step": 39530 }, { "grad_norm": 0.24242207407951355, "learning_rate": 5.507771777621406e-05, "loss": 0.0149, "step": 39540 }, { "grad_norm": 0.2587696313858032, "learning_rate": 5.505715582591052e-05, "loss": 0.031, "step": 39550 }, { "grad_norm": 0.3385961949825287, "learning_rate": 5.50365930114774e-05, "loss": 0.0283, "step": 39560 }, { "grad_norm": 0.20439474284648895, "learning_rate": 5.5016029336428255e-05, "loss": 0.0175, "step": 39570 }, { "grad_norm": 0.4578014016151428, "learning_rate": 5.49954648042769e-05, "loss": 0.019, "step": 39580 }, { "grad_norm": 0.23416121304035187, "learning_rate": 5.4974899418537226e-05, "loss": 0.0244, "step": 39590 }, { "grad_norm": 0.265657901763916, "learning_rate": 5.4954333182723316e-05, "loss": 0.0217, "step": 39600 }, { "grad_norm": 0.3995353579521179, "learning_rate": 5.493376610034937e-05, "loss": 0.0183, "step": 39610 }, { "grad_norm": 0.28971317410469055, "learning_rate": 5.4913198174929735e-05, "loss": 0.0189, "step": 39620 }, { "grad_norm": 0.38024669885635376, "learning_rate": 5.489262940997889e-05, "loss": 0.0297, "step": 39630 }, { "grad_norm": 0.18814978003501892, "learning_rate": 5.487205980901151e-05, "loss": 0.028, "step": 39640 }, { "grad_norm": 0.3104638159275055, "learning_rate": 5.485148937554234e-05, "loss": 0.0394, "step": 39650 }, { "grad_norm": 0.27342700958251953, "learning_rate": 5.483091811308635e-05, "loss": 0.0246, "step": 39660 }, { "grad_norm": 0.42503127455711365, "learning_rate": 5.4810346025158564e-05, "loss": 0.0237, "step": 39670 }, { "grad_norm": 0.4007159173488617, "learning_rate": 5.478977311527421e-05, "loss": 0.0215, "step": 39680 }, { "grad_norm": 0.3063925802707672, "learning_rate": 5.476919938694863e-05, "loss": 0.0211, "step": 39690 }, { "grad_norm": 0.21718910336494446, "learning_rate": 5.474862484369733e-05, "loss": 0.014, "step": 39700 }, { "grad_norm": 0.24935530126094818, "learning_rate": 5.472804948903589e-05, "loss": 0.0246, "step": 39710 }, { "grad_norm": 0.23663175106048584, "learning_rate": 5.470747332648013e-05, "loss": 0.013, "step": 39720 }, { "grad_norm": 0.25352755188941956, "learning_rate": 5.468689635954591e-05, "loss": 0.0265, "step": 39730 }, { "grad_norm": 0.2667825222015381, "learning_rate": 5.46663185917493e-05, "loss": 0.0195, "step": 39740 }, { "grad_norm": 0.26083147525787354, "learning_rate": 5.464574002660645e-05, "loss": 0.0228, "step": 39750 }, { "grad_norm": 0.3238420784473419, "learning_rate": 5.4625160667633724e-05, "loss": 0.0259, "step": 39760 }, { "grad_norm": 0.20409204065799713, "learning_rate": 5.4604580518347505e-05, "loss": 0.0193, "step": 39770 }, { "grad_norm": 0.2586224675178528, "learning_rate": 5.458399958226442e-05, "loss": 0.0164, "step": 39780 }, { "grad_norm": 0.25579723715782166, "learning_rate": 5.456341786290119e-05, "loss": 0.0149, "step": 39790 }, { "grad_norm": 0.27071648836135864, "learning_rate": 5.454283536377465e-05, "loss": 0.0193, "step": 39800 }, { "grad_norm": 0.24633221328258514, "learning_rate": 5.452225208840179e-05, "loss": 0.0184, "step": 39810 }, { "grad_norm": 0.15795010328292847, "learning_rate": 5.450166804029975e-05, "loss": 0.0209, "step": 39820 }, { "grad_norm": 0.38523733615875244, "learning_rate": 5.448108322298574e-05, "loss": 0.0178, "step": 39830 }, { "grad_norm": 0.24347230792045593, "learning_rate": 5.446049763997717e-05, "loss": 0.0184, "step": 39840 }, { "grad_norm": 0.29337602853775024, "learning_rate": 5.4439911294791546e-05, "loss": 0.0184, "step": 39850 }, { "grad_norm": 0.32511240243911743, "learning_rate": 5.441932419094652e-05, "loss": 0.015, "step": 39860 }, { "grad_norm": 0.36101052165031433, "learning_rate": 5.439873633195985e-05, "loss": 0.0196, "step": 39870 }, { "grad_norm": 0.27988553047180176, "learning_rate": 5.437814772134947e-05, "loss": 0.0229, "step": 39880 }, { "grad_norm": 0.22683507204055786, "learning_rate": 5.4357558362633366e-05, "loss": 0.0164, "step": 39890 }, { "grad_norm": 0.23942908644676208, "learning_rate": 5.4336968259329726e-05, "loss": 0.0143, "step": 39900 }, { "grad_norm": 0.22192810475826263, "learning_rate": 5.431637741495681e-05, "loss": 0.0216, "step": 39910 }, { "grad_norm": 0.3459923267364502, "learning_rate": 5.429578583303307e-05, "loss": 0.0199, "step": 39920 }, { "grad_norm": 0.20634444057941437, "learning_rate": 5.427519351707701e-05, "loss": 0.019, "step": 39930 }, { "grad_norm": 0.3043566942214966, "learning_rate": 5.42546004706073e-05, "loss": 0.0218, "step": 39940 }, { "grad_norm": 0.35381802916526794, "learning_rate": 5.4234006697142735e-05, "loss": 0.0192, "step": 39950 }, { "grad_norm": 0.2799130380153656, "learning_rate": 5.421341220020224e-05, "loss": 0.0203, "step": 39960 }, { "grad_norm": 0.2646469175815582, "learning_rate": 5.419281698330482e-05, "loss": 0.0254, "step": 39970 }, { "grad_norm": 0.2469346672296524, "learning_rate": 5.4172221049969665e-05, "loss": 0.0196, "step": 39980 }, { "grad_norm": 0.27610960602760315, "learning_rate": 5.415162440371604e-05, "loss": 0.0161, "step": 39990 }, { "grad_norm": 0.3390046954154968, "learning_rate": 5.413102704806334e-05, "loss": 0.0209, "step": 40000 }, { "grad_norm": 0.22573402523994446, "learning_rate": 5.41104289865311e-05, "loss": 0.0176, "step": 40010 }, { "grad_norm": 0.3107450604438782, "learning_rate": 5.408983022263898e-05, "loss": 0.0179, "step": 40020 }, { "grad_norm": 0.31418994069099426, "learning_rate": 5.406923075990673e-05, "loss": 0.0212, "step": 40030 }, { "grad_norm": 0.31463032960891724, "learning_rate": 5.404863060185423e-05, "loss": 0.0181, "step": 40040 }, { "grad_norm": 0.31877076625823975, "learning_rate": 5.40280297520015e-05, "loss": 0.0199, "step": 40050 }, { "grad_norm": 0.2584896385669708, "learning_rate": 5.400742821386865e-05, "loss": 0.0165, "step": 40060 }, { "grad_norm": 0.31518611311912537, "learning_rate": 5.398682599097592e-05, "loss": 0.0209, "step": 40070 }, { "grad_norm": 0.3786553740501404, "learning_rate": 5.396622308684367e-05, "loss": 0.0246, "step": 40080 }, { "grad_norm": 0.3123810291290283, "learning_rate": 5.394561950499236e-05, "loss": 0.0222, "step": 40090 }, { "grad_norm": 0.23166093230247498, "learning_rate": 5.39250152489426e-05, "loss": 0.0172, "step": 40100 }, { "grad_norm": 0.16275514662265778, "learning_rate": 5.390441032221507e-05, "loss": 0.0151, "step": 40110 }, { "grad_norm": 0.24035140872001648, "learning_rate": 5.388380472833062e-05, "loss": 0.0185, "step": 40120 }, { "grad_norm": 0.2853509485721588, "learning_rate": 5.386319847081016e-05, "loss": 0.0217, "step": 40130 }, { "grad_norm": 0.472883939743042, "learning_rate": 5.384259155317473e-05, "loss": 0.0224, "step": 40140 }, { "grad_norm": 0.38074713945388794, "learning_rate": 5.38219839789455e-05, "loss": 0.0262, "step": 40150 }, { "grad_norm": 0.6537005305290222, "learning_rate": 5.380137575164376e-05, "loss": 0.0262, "step": 40160 }, { "grad_norm": 0.2751692235469818, "learning_rate": 5.378076687479085e-05, "loss": 0.0215, "step": 40170 }, { "grad_norm": 0.17513038218021393, "learning_rate": 5.3760157351908305e-05, "loss": 0.023, "step": 40180 }, { "grad_norm": 0.27877822518348694, "learning_rate": 5.373954718651768e-05, "loss": 0.0238, "step": 40190 }, { "grad_norm": 0.28778520226478577, "learning_rate": 5.371893638214074e-05, "loss": 0.0156, "step": 40200 }, { "grad_norm": 0.3185630440711975, "learning_rate": 5.369832494229927e-05, "loss": 0.0249, "step": 40210 }, { "grad_norm": 0.34217098355293274, "learning_rate": 5.367771287051524e-05, "loss": 0.0277, "step": 40220 }, { "grad_norm": 0.16998445987701416, "learning_rate": 5.365710017031065e-05, "loss": 0.0197, "step": 40230 }, { "grad_norm": 0.2779598534107208, "learning_rate": 5.3636486845207654e-05, "loss": 0.0237, "step": 40240 }, { "grad_norm": 0.42746207118034363, "learning_rate": 5.3615872898728526e-05, "loss": 0.018, "step": 40250 }, { "grad_norm": 0.47994816303253174, "learning_rate": 5.3595258334395614e-05, "loss": 0.0199, "step": 40260 }, { "grad_norm": 0.28922924399375916, "learning_rate": 5.3574643155731396e-05, "loss": 0.0297, "step": 40270 }, { "grad_norm": 0.20193549990653992, "learning_rate": 5.3554027366258405e-05, "loss": 0.0216, "step": 40280 }, { "grad_norm": 0.2512524724006653, "learning_rate": 5.3533410969499355e-05, "loss": 0.0236, "step": 40290 }, { "grad_norm": 0.23269209265708923, "learning_rate": 5.3512793968977e-05, "loss": 0.0241, "step": 40300 }, { "grad_norm": 0.3401079773902893, "learning_rate": 5.349217636821423e-05, "loss": 0.03, "step": 40310 }, { "grad_norm": 0.23896627128124237, "learning_rate": 5.347155817073405e-05, "loss": 0.0217, "step": 40320 }, { "grad_norm": 0.23166806995868683, "learning_rate": 5.34509393800595e-05, "loss": 0.0161, "step": 40330 }, { "grad_norm": 0.3008493483066559, "learning_rate": 5.343031999971381e-05, "loss": 0.021, "step": 40340 }, { "grad_norm": 0.20388934016227722, "learning_rate": 5.340970003322026e-05, "loss": 0.0196, "step": 40350 }, { "grad_norm": 0.3581079840660095, "learning_rate": 5.33890794841022e-05, "loss": 0.0173, "step": 40360 }, { "grad_norm": 0.2577836513519287, "learning_rate": 5.336845835588318e-05, "loss": 0.0227, "step": 40370 }, { "grad_norm": 0.2872214913368225, "learning_rate": 5.334783665208674e-05, "loss": 0.019, "step": 40380 }, { "grad_norm": 0.4253438413143158, "learning_rate": 5.332721437623657e-05, "loss": 0.021, "step": 40390 }, { "grad_norm": 0.2613525092601776, "learning_rate": 5.3306591531856464e-05, "loss": 0.0204, "step": 40400 }, { "grad_norm": 0.265353262424469, "learning_rate": 5.3285968122470295e-05, "loss": 0.0234, "step": 40410 }, { "grad_norm": 0.2041931450366974, "learning_rate": 5.326534415160207e-05, "loss": 0.0261, "step": 40420 }, { "grad_norm": 0.354458749294281, "learning_rate": 5.324471962277582e-05, "loss": 0.0234, "step": 40430 }, { "grad_norm": 0.5272030830383301, "learning_rate": 5.3224094539515746e-05, "loss": 0.0237, "step": 40440 }, { "grad_norm": 0.18769709765911102, "learning_rate": 5.3203468905346076e-05, "loss": 0.0229, "step": 40450 }, { "grad_norm": 0.14387276768684387, "learning_rate": 5.3182842723791195e-05, "loss": 0.0203, "step": 40460 }, { "grad_norm": 0.18262585997581482, "learning_rate": 5.316221599837554e-05, "loss": 0.0177, "step": 40470 }, { "grad_norm": 0.5510495901107788, "learning_rate": 5.314158873262366e-05, "loss": 0.0242, "step": 40480 }, { "grad_norm": 0.2504841685295105, "learning_rate": 5.312096093006018e-05, "loss": 0.016, "step": 40490 }, { "grad_norm": 0.31966632604599, "learning_rate": 5.3100332594209846e-05, "loss": 0.0185, "step": 40500 }, { "grad_norm": 0.2798236906528473, "learning_rate": 5.3079703728597454e-05, "loss": 0.0224, "step": 40510 }, { "grad_norm": 0.2707116901874542, "learning_rate": 5.305907433674794e-05, "loss": 0.0216, "step": 40520 }, { "grad_norm": 0.3446427881717682, "learning_rate": 5.3038444422186284e-05, "loss": 0.0178, "step": 40530 }, { "grad_norm": 0.40689241886138916, "learning_rate": 5.301781398843757e-05, "loss": 0.0177, "step": 40540 }, { "grad_norm": 0.2542571425437927, "learning_rate": 5.299718303902699e-05, "loss": 0.0237, "step": 40550 }, { "grad_norm": 0.3472405672073364, "learning_rate": 5.297655157747979e-05, "loss": 0.0236, "step": 40560 }, { "grad_norm": 0.31006836891174316, "learning_rate": 5.295591960732136e-05, "loss": 0.0163, "step": 40570 }, { "grad_norm": 0.19674034416675568, "learning_rate": 5.293528713207708e-05, "loss": 0.019, "step": 40580 }, { "grad_norm": 0.21847814321517944, "learning_rate": 5.291465415527253e-05, "loss": 0.0208, "step": 40590 }, { "grad_norm": 0.3593502938747406, "learning_rate": 5.2894020680433296e-05, "loss": 0.0155, "step": 40600 }, { "grad_norm": 0.29831188917160034, "learning_rate": 5.287338671108507e-05, "loss": 0.017, "step": 40610 }, { "grad_norm": 0.22242587804794312, "learning_rate": 5.285275225075367e-05, "loss": 0.0176, "step": 40620 }, { "grad_norm": 0.3363821506500244, "learning_rate": 5.283211730296492e-05, "loss": 0.0161, "step": 40630 }, { "grad_norm": 0.38132813572883606, "learning_rate": 5.281148187124477e-05, "loss": 0.0186, "step": 40640 }, { "grad_norm": 0.4005310833454132, "learning_rate": 5.279084595911927e-05, "loss": 0.0184, "step": 40650 }, { "grad_norm": 0.32506197690963745, "learning_rate": 5.277020957011453e-05, "loss": 0.0308, "step": 40660 }, { "grad_norm": 0.30584481358528137, "learning_rate": 5.274957270775673e-05, "loss": 0.0256, "step": 40670 }, { "grad_norm": 0.2744576632976532, "learning_rate": 5.2728935375572164e-05, "loss": 0.015, "step": 40680 }, { "grad_norm": 0.20670337975025177, "learning_rate": 5.2708297577087176e-05, "loss": 0.0168, "step": 40690 }, { "grad_norm": 0.31009989976882935, "learning_rate": 5.2687659315828196e-05, "loss": 0.0188, "step": 40700 }, { "grad_norm": 0.1603742241859436, "learning_rate": 5.266702059532175e-05, "loss": 0.0174, "step": 40710 }, { "grad_norm": 0.38937488198280334, "learning_rate": 5.264638141909444e-05, "loss": 0.0255, "step": 40720 }, { "grad_norm": 0.22846083343029022, "learning_rate": 5.2625741790672925e-05, "loss": 0.0201, "step": 40730 }, { "grad_norm": 0.2575835883617401, "learning_rate": 5.2605101713583935e-05, "loss": 0.0197, "step": 40740 }, { "grad_norm": 0.24414072930812836, "learning_rate": 5.2584461191354315e-05, "loss": 0.0233, "step": 40750 }, { "grad_norm": 0.2095046192407608, "learning_rate": 5.2563820227510964e-05, "loss": 0.0169, "step": 40760 }, { "grad_norm": 0.31569674611091614, "learning_rate": 5.2543178825580864e-05, "loss": 0.0189, "step": 40770 }, { "grad_norm": 0.4873710572719574, "learning_rate": 5.2522536989091054e-05, "loss": 0.0264, "step": 40780 }, { "grad_norm": 0.2897956371307373, "learning_rate": 5.2501894721568655e-05, "loss": 0.0189, "step": 40790 }, { "grad_norm": 0.3098782002925873, "learning_rate": 5.248125202654089e-05, "loss": 0.0173, "step": 40800 }, { "grad_norm": 0.3939148187637329, "learning_rate": 5.246060890753501e-05, "loss": 0.0218, "step": 40810 }, { "grad_norm": 0.1971825510263443, "learning_rate": 5.243996536807837e-05, "loss": 0.0192, "step": 40820 }, { "grad_norm": 0.8732051253318787, "learning_rate": 5.2419321411698384e-05, "loss": 0.0154, "step": 40830 }, { "grad_norm": 0.378427118062973, "learning_rate": 5.239867704192253e-05, "loss": 0.0209, "step": 40840 }, { "grad_norm": 0.28631508350372314, "learning_rate": 5.237803226227838e-05, "loss": 0.0159, "step": 40850 }, { "grad_norm": 0.2232508659362793, "learning_rate": 5.235738707629354e-05, "loss": 0.0154, "step": 40860 }, { "grad_norm": 0.2138753980398178, "learning_rate": 5.233674148749575e-05, "loss": 0.0209, "step": 40870 }, { "grad_norm": 0.27105438709259033, "learning_rate": 5.231609549941272e-05, "loss": 0.0213, "step": 40880 }, { "grad_norm": 0.3051964044570923, "learning_rate": 5.229544911557233e-05, "loss": 0.0203, "step": 40890 }, { "grad_norm": 0.21548928320407867, "learning_rate": 5.227480233950246e-05, "loss": 0.0176, "step": 40900 }, { "grad_norm": 0.37471526861190796, "learning_rate": 5.22541551747311e-05, "loss": 0.0212, "step": 40910 }, { "grad_norm": 0.1917712241411209, "learning_rate": 5.223350762478626e-05, "loss": 0.013, "step": 40920 }, { "grad_norm": 0.26454395055770874, "learning_rate": 5.221285969319606e-05, "loss": 0.0154, "step": 40930 }, { "grad_norm": 0.1672043353319168, "learning_rate": 5.219221138348865e-05, "loss": 0.0174, "step": 40940 }, { "grad_norm": 0.20210570096969604, "learning_rate": 5.217156269919228e-05, "loss": 0.0188, "step": 40950 }, { "grad_norm": 0.3139609396457672, "learning_rate": 5.215091364383523e-05, "loss": 0.0201, "step": 40960 }, { "grad_norm": 0.31687238812446594, "learning_rate": 5.213026422094588e-05, "loss": 0.0211, "step": 40970 }, { "grad_norm": 0.3903524875640869, "learning_rate": 5.210961443405262e-05, "loss": 0.0181, "step": 40980 }, { "grad_norm": 0.27156543731689453, "learning_rate": 5.208896428668396e-05, "loss": 0.0148, "step": 40990 }, { "grad_norm": 0.2593756914138794, "learning_rate": 5.206831378236845e-05, "loss": 0.0195, "step": 41000 }, { "grad_norm": 0.26857537031173706, "learning_rate": 5.2047662924634666e-05, "loss": 0.0187, "step": 41010 }, { "grad_norm": 0.4210374355316162, "learning_rate": 5.20270117170113e-05, "loss": 0.0171, "step": 41020 }, { "grad_norm": 0.5702404379844666, "learning_rate": 5.200636016302707e-05, "loss": 0.0251, "step": 41030 }, { "grad_norm": 0.3012155294418335, "learning_rate": 5.1985708266210754e-05, "loss": 0.0199, "step": 41040 }, { "grad_norm": 0.5020608901977539, "learning_rate": 5.1965056030091206e-05, "loss": 0.0249, "step": 41050 }, { "grad_norm": 0.24336500465869904, "learning_rate": 5.194440345819732e-05, "loss": 0.0171, "step": 41060 }, { "grad_norm": 0.31589236855506897, "learning_rate": 5.1923750554058084e-05, "loss": 0.0193, "step": 41070 }, { "grad_norm": 0.21131502091884613, "learning_rate": 5.1903097321202476e-05, "loss": 0.0162, "step": 41080 }, { "grad_norm": 0.3851841390132904, "learning_rate": 5.18824437631596e-05, "loss": 0.0301, "step": 41090 }, { "grad_norm": 0.2163364291191101, "learning_rate": 5.186178988345856e-05, "loss": 0.0218, "step": 41100 }, { "grad_norm": 0.30695822834968567, "learning_rate": 5.184113568562855e-05, "loss": 0.0242, "step": 41110 }, { "grad_norm": 0.3879435360431671, "learning_rate": 5.18204811731988e-05, "loss": 0.0233, "step": 41120 }, { "grad_norm": 0.3962245583534241, "learning_rate": 5.179982634969862e-05, "loss": 0.0233, "step": 41130 }, { "grad_norm": 0.38155269622802734, "learning_rate": 5.177917121865734e-05, "loss": 0.0265, "step": 41140 }, { "grad_norm": 0.3397330939769745, "learning_rate": 5.1758515783604346e-05, "loss": 0.0199, "step": 41150 }, { "grad_norm": 0.31621652841567993, "learning_rate": 5.173786004806911e-05, "loss": 0.0217, "step": 41160 }, { "grad_norm": 0.4733025133609772, "learning_rate": 5.1717204015581135e-05, "loss": 0.0241, "step": 41170 }, { "grad_norm": 0.3139048218727112, "learning_rate": 5.169654768966994e-05, "loss": 0.0188, "step": 41180 }, { "grad_norm": 0.34665629267692566, "learning_rate": 5.1675891073865156e-05, "loss": 0.0202, "step": 41190 }, { "grad_norm": 0.2506008744239807, "learning_rate": 5.1655234171696424e-05, "loss": 0.015, "step": 41200 }, { "grad_norm": 0.22770968079566956, "learning_rate": 5.163457698669343e-05, "loss": 0.0165, "step": 41210 }, { "grad_norm": 0.38482826948165894, "learning_rate": 5.1613919522385936e-05, "loss": 0.0208, "step": 41220 }, { "grad_norm": 0.30270659923553467, "learning_rate": 5.1593261782303746e-05, "loss": 0.0315, "step": 41230 }, { "grad_norm": 0.23927277326583862, "learning_rate": 5.157260376997669e-05, "loss": 0.0174, "step": 41240 }, { "grad_norm": 0.27949801087379456, "learning_rate": 5.155194548893466e-05, "loss": 0.0235, "step": 41250 }, { "grad_norm": 0.22624552249908447, "learning_rate": 5.153128694270759e-05, "loss": 0.0233, "step": 41260 }, { "grad_norm": 0.22792088985443115, "learning_rate": 5.151062813482548e-05, "loss": 0.0259, "step": 41270 }, { "grad_norm": 0.3687157928943634, "learning_rate": 5.148996906881832e-05, "loss": 0.0244, "step": 41280 }, { "grad_norm": 0.4391099810600281, "learning_rate": 5.1469309748216196e-05, "loss": 0.0166, "step": 41290 }, { "grad_norm": 0.25801369547843933, "learning_rate": 5.144865017654923e-05, "loss": 0.0121, "step": 41300 }, { "grad_norm": 0.3846214711666107, "learning_rate": 5.1427990357347556e-05, "loss": 0.0233, "step": 41310 }, { "grad_norm": 0.24511313438415527, "learning_rate": 5.140733029414139e-05, "loss": 0.0174, "step": 41320 }, { "grad_norm": 0.27848494052886963, "learning_rate": 5.138666999046098e-05, "loss": 0.0211, "step": 41330 }, { "grad_norm": 0.3271337151527405, "learning_rate": 5.136600944983658e-05, "loss": 0.016, "step": 41340 }, { "grad_norm": 0.3426038920879364, "learning_rate": 5.134534867579853e-05, "loss": 0.018, "step": 41350 }, { "grad_norm": 0.33869579434394836, "learning_rate": 5.1324687671877184e-05, "loss": 0.0181, "step": 41360 }, { "grad_norm": 0.3368269205093384, "learning_rate": 5.130402644160296e-05, "loss": 0.0243, "step": 41370 }, { "grad_norm": 0.38232654333114624, "learning_rate": 5.128336498850628e-05, "loss": 0.0158, "step": 41380 }, { "grad_norm": 0.3510694205760956, "learning_rate": 5.126270331611761e-05, "loss": 0.0174, "step": 41390 }, { "grad_norm": 0.27750179171562195, "learning_rate": 5.124204142796748e-05, "loss": 0.0148, "step": 41400 }, { "grad_norm": 0.1646127551794052, "learning_rate": 5.122137932758644e-05, "loss": 0.0178, "step": 41410 }, { "grad_norm": 0.27020570635795593, "learning_rate": 5.1200717018505074e-05, "loss": 0.0158, "step": 41420 }, { "grad_norm": 0.20169900357723236, "learning_rate": 5.118005450425403e-05, "loss": 0.0158, "step": 41430 }, { "grad_norm": 0.23206497728824615, "learning_rate": 5.115939178836392e-05, "loss": 0.0142, "step": 41440 }, { "grad_norm": 0.3381384611129761, "learning_rate": 5.113872887436547e-05, "loss": 0.0197, "step": 41450 }, { "grad_norm": 0.34261369705200195, "learning_rate": 5.11180657657894e-05, "loss": 0.0225, "step": 41460 }, { "grad_norm": 0.38817495107650757, "learning_rate": 5.109740246616649e-05, "loss": 0.017, "step": 41470 }, { "grad_norm": 0.5947389602661133, "learning_rate": 5.10767389790275e-05, "loss": 0.0221, "step": 41480 }, { "grad_norm": 0.7326323986053467, "learning_rate": 5.1056075307903265e-05, "loss": 0.0204, "step": 41490 }, { "grad_norm": 0.2581115663051605, "learning_rate": 5.1035411456324634e-05, "loss": 0.0175, "step": 41500 }, { "grad_norm": 0.4297967255115509, "learning_rate": 5.1014747427822504e-05, "loss": 0.019, "step": 41510 }, { "grad_norm": 0.4461771845817566, "learning_rate": 5.0994083225927804e-05, "loss": 0.0229, "step": 41520 }, { "grad_norm": 0.2703193426132202, "learning_rate": 5.0973418854171475e-05, "loss": 0.0233, "step": 41530 }, { "grad_norm": 0.334251344203949, "learning_rate": 5.095275431608447e-05, "loss": 0.0195, "step": 41540 }, { "grad_norm": 0.2629280984401703, "learning_rate": 5.093208961519782e-05, "loss": 0.0205, "step": 41550 }, { "grad_norm": 0.24455766379833221, "learning_rate": 5.091142475504255e-05, "loss": 0.0176, "step": 41560 }, { "grad_norm": 0.352079302072525, "learning_rate": 5.089075973914971e-05, "loss": 0.0264, "step": 41570 }, { "grad_norm": 0.23242178559303284, "learning_rate": 5.08700945710504e-05, "loss": 0.0285, "step": 41580 }, { "grad_norm": 0.34966883063316345, "learning_rate": 5.0849429254275714e-05, "loss": 0.0174, "step": 41590 }, { "grad_norm": 0.19757430255413055, "learning_rate": 5.0828763792356804e-05, "loss": 0.0162, "step": 41600 }, { "grad_norm": 0.4478340744972229, "learning_rate": 5.080809818882483e-05, "loss": 0.0203, "step": 41610 }, { "grad_norm": 0.3669344186782837, "learning_rate": 5.078743244721097e-05, "loss": 0.0213, "step": 41620 }, { "grad_norm": 0.30833202600479126, "learning_rate": 5.0766766571046455e-05, "loss": 0.0221, "step": 41630 }, { "grad_norm": 0.31305670738220215, "learning_rate": 5.07461005638625e-05, "loss": 0.0155, "step": 41640 }, { "grad_norm": 0.1659351885318756, "learning_rate": 5.072543442919037e-05, "loss": 0.0149, "step": 41650 }, { "grad_norm": 0.27259716391563416, "learning_rate": 5.070476817056132e-05, "loss": 0.0314, "step": 41660 }, { "grad_norm": 0.3914475440979004, "learning_rate": 5.068410179150668e-05, "loss": 0.02, "step": 41670 }, { "grad_norm": 0.4198289215564728, "learning_rate": 5.066343529555775e-05, "loss": 0.0187, "step": 41680 }, { "grad_norm": 0.30468207597732544, "learning_rate": 5.064276868624587e-05, "loss": 0.0155, "step": 41690 }, { "grad_norm": 0.20840801298618317, "learning_rate": 5.0622101967102396e-05, "loss": 0.0147, "step": 41700 }, { "grad_norm": 0.25165942311286926, "learning_rate": 5.0601435141658714e-05, "loss": 0.0169, "step": 41710 }, { "grad_norm": 0.3101138174533844, "learning_rate": 5.058076821344622e-05, "loss": 0.0193, "step": 41720 }, { "grad_norm": 0.32804402709007263, "learning_rate": 5.056010118599632e-05, "loss": 0.0285, "step": 41730 }, { "grad_norm": 0.3063139021396637, "learning_rate": 5.053943406284044e-05, "loss": 0.0197, "step": 41740 }, { "grad_norm": 0.32600679993629456, "learning_rate": 5.051876684751006e-05, "loss": 0.0261, "step": 41750 }, { "grad_norm": 0.28010302782058716, "learning_rate": 5.0498099543536584e-05, "loss": 0.0169, "step": 41760 }, { "grad_norm": 0.20784230530261993, "learning_rate": 5.047743215445152e-05, "loss": 0.0169, "step": 41770 }, { "grad_norm": 0.23614133894443512, "learning_rate": 5.045676468378637e-05, "loss": 0.0153, "step": 41780 }, { "grad_norm": 0.3182823359966278, "learning_rate": 5.0436097135072626e-05, "loss": 0.0175, "step": 41790 }, { "grad_norm": 0.2861773371696472, "learning_rate": 5.041542951184181e-05, "loss": 0.0183, "step": 41800 }, { "grad_norm": 0.2046651840209961, "learning_rate": 5.039476181762545e-05, "loss": 0.0138, "step": 41810 }, { "grad_norm": 0.1595001071691513, "learning_rate": 5.037409405595508e-05, "loss": 0.0131, "step": 41820 }, { "grad_norm": 0.2796494960784912, "learning_rate": 5.035342623036229e-05, "loss": 0.0178, "step": 41830 }, { "grad_norm": 0.2993241250514984, "learning_rate": 5.033275834437862e-05, "loss": 0.0214, "step": 41840 }, { "grad_norm": 0.32360929250717163, "learning_rate": 5.031209040153564e-05, "loss": 0.0176, "step": 41850 }, { "grad_norm": 0.2442491352558136, "learning_rate": 5.0291422405364955e-05, "loss": 0.0263, "step": 41860 }, { "grad_norm": 0.30989933013916016, "learning_rate": 5.0270754359398133e-05, "loss": 0.0187, "step": 41870 }, { "grad_norm": 0.38970306515693665, "learning_rate": 5.025008626716682e-05, "loss": 0.015, "step": 41880 }, { "grad_norm": 0.24891284108161926, "learning_rate": 5.0229418132202585e-05, "loss": 0.0171, "step": 41890 }, { "grad_norm": 0.3193639814853668, "learning_rate": 5.020874995803707e-05, "loss": 0.0232, "step": 41900 }, { "grad_norm": 0.3376235365867615, "learning_rate": 5.01880817482019e-05, "loss": 0.0172, "step": 41910 }, { "grad_norm": 0.2158665806055069, "learning_rate": 5.01674135062287e-05, "loss": 0.0113, "step": 41920 }, { "grad_norm": 0.2817296087741852, "learning_rate": 5.0146745235649115e-05, "loss": 0.0152, "step": 41930 }, { "grad_norm": 0.31739893555641174, "learning_rate": 5.0126076939994795e-05, "loss": 0.0183, "step": 41940 }, { "grad_norm": 0.2990038990974426, "learning_rate": 5.010540862279736e-05, "loss": 0.0191, "step": 41950 }, { "grad_norm": 0.315885454416275, "learning_rate": 5.008474028758846e-05, "loss": 0.0182, "step": 41960 }, { "grad_norm": 0.23511618375778198, "learning_rate": 5.0064071937899784e-05, "loss": 0.0169, "step": 41970 }, { "grad_norm": 0.38074061274528503, "learning_rate": 5.004340357726296e-05, "loss": 0.0189, "step": 41980 }, { "grad_norm": 0.2657032310962677, "learning_rate": 5.002273520920965e-05, "loss": 0.0181, "step": 41990 }, { "grad_norm": 0.3526308834552765, "learning_rate": 5.000206683727151e-05, "loss": 0.0169, "step": 42000 }, { "grad_norm": 0.2924680709838867, "learning_rate": 4.998139846498021e-05, "loss": 0.018, "step": 42010 }, { "grad_norm": 0.36823511123657227, "learning_rate": 4.9960730095867405e-05, "loss": 0.0176, "step": 42020 }, { "grad_norm": 0.2364233136177063, "learning_rate": 4.9940061733464755e-05, "loss": 0.0214, "step": 42030 }, { "grad_norm": 0.22575530409812927, "learning_rate": 4.991939338130392e-05, "loss": 0.0307, "step": 42040 }, { "grad_norm": 0.4524588882923126, "learning_rate": 4.989872504291653e-05, "loss": 0.0275, "step": 42050 }, { "grad_norm": 0.263632595539093, "learning_rate": 4.9878056721834273e-05, "loss": 0.0213, "step": 42060 }, { "grad_norm": 0.20489224791526794, "learning_rate": 4.98573884215888e-05, "loss": 0.0169, "step": 42070 }, { "grad_norm": 0.2726769745349884, "learning_rate": 4.9836720145711715e-05, "loss": 0.018, "step": 42080 }, { "grad_norm": 0.28095555305480957, "learning_rate": 4.9816051897734725e-05, "loss": 0.021, "step": 42090 }, { "grad_norm": 0.33732080459594727, "learning_rate": 4.979538368118941e-05, "loss": 0.0232, "step": 42100 }, { "grad_norm": 0.38686639070510864, "learning_rate": 4.9774715499607446e-05, "loss": 0.0186, "step": 42110 }, { "grad_norm": 0.2237677127122879, "learning_rate": 4.9754047356520416e-05, "loss": 0.0262, "step": 42120 }, { "grad_norm": 0.3401719629764557, "learning_rate": 4.973337925545999e-05, "loss": 0.0203, "step": 42130 }, { "grad_norm": 0.36376720666885376, "learning_rate": 4.9712711199957746e-05, "loss": 0.0197, "step": 42140 }, { "grad_norm": 0.2575587332248688, "learning_rate": 4.969204319354531e-05, "loss": 0.0219, "step": 42150 }, { "grad_norm": 0.30481085181236267, "learning_rate": 4.9671375239754267e-05, "loss": 0.021, "step": 42160 }, { "grad_norm": 0.32314333319664, "learning_rate": 4.9650707342116195e-05, "loss": 0.02, "step": 42170 }, { "grad_norm": 0.23354756832122803, "learning_rate": 4.96300395041627e-05, "loss": 0.0139, "step": 42180 }, { "grad_norm": 0.2509557902812958, "learning_rate": 4.960937172942532e-05, "loss": 0.0202, "step": 42190 }, { "grad_norm": 0.2156214714050293, "learning_rate": 4.958870402143565e-05, "loss": 0.0186, "step": 42200 }, { "grad_norm": 0.22085455060005188, "learning_rate": 4.9568036383725186e-05, "loss": 0.0197, "step": 42210 }, { "grad_norm": 0.25135543942451477, "learning_rate": 4.9547368819825516e-05, "loss": 0.0191, "step": 42220 }, { "grad_norm": 0.4521521329879761, "learning_rate": 4.952670133326812e-05, "loss": 0.021, "step": 42230 }, { "grad_norm": 0.32833102345466614, "learning_rate": 4.950603392758453e-05, "loss": 0.0209, "step": 42240 }, { "grad_norm": 0.3128757178783417, "learning_rate": 4.948536660630621e-05, "loss": 0.018, "step": 42250 }, { "grad_norm": 0.3056167960166931, "learning_rate": 4.9464699372964685e-05, "loss": 0.0218, "step": 42260 }, { "grad_norm": 0.19578863680362701, "learning_rate": 4.9444032231091395e-05, "loss": 0.0187, "step": 42270 }, { "grad_norm": 0.29068413376808167, "learning_rate": 4.9423365184217776e-05, "loss": 0.0194, "step": 42280 }, { "grad_norm": 0.3154106140136719, "learning_rate": 4.940269823587529e-05, "loss": 0.02, "step": 42290 }, { "grad_norm": 0.15210360288619995, "learning_rate": 4.938203138959533e-05, "loss": 0.0149, "step": 42300 }, { "grad_norm": 0.24546384811401367, "learning_rate": 4.9361364648909325e-05, "loss": 0.0188, "step": 42310 }, { "grad_norm": 0.3380652964115143, "learning_rate": 4.9340698017348605e-05, "loss": 0.0159, "step": 42320 }, { "grad_norm": 0.2022760957479477, "learning_rate": 4.932003149844458e-05, "loss": 0.0176, "step": 42330 }, { "grad_norm": 0.37423720955848694, "learning_rate": 4.929936509572857e-05, "loss": 0.0263, "step": 42340 }, { "grad_norm": 0.3044484257698059, "learning_rate": 4.927869881273191e-05, "loss": 0.0195, "step": 42350 }, { "grad_norm": 0.2983860671520233, "learning_rate": 4.9258032652985894e-05, "loss": 0.0159, "step": 42360 }, { "grad_norm": 0.2793136239051819, "learning_rate": 4.9237366620021786e-05, "loss": 0.0188, "step": 42370 }, { "grad_norm": 0.21886484324932098, "learning_rate": 4.921670071737089e-05, "loss": 0.0139, "step": 42380 }, { "grad_norm": 0.2986140251159668, "learning_rate": 4.91960349485644e-05, "loss": 0.0214, "step": 42390 }, { "grad_norm": 0.3139273226261139, "learning_rate": 4.9175369317133566e-05, "loss": 0.0201, "step": 42400 }, { "grad_norm": 0.14764586091041565, "learning_rate": 4.915470382660953e-05, "loss": 0.0133, "step": 42410 }, { "grad_norm": 0.24534426629543304, "learning_rate": 4.9134038480523524e-05, "loss": 0.0171, "step": 42420 }, { "grad_norm": 0.22464486956596375, "learning_rate": 4.911337328240664e-05, "loss": 0.0216, "step": 42430 }, { "grad_norm": 0.26779189705848694, "learning_rate": 4.909270823579003e-05, "loss": 0.0175, "step": 42440 }, { "grad_norm": 0.30974894762039185, "learning_rate": 4.907204334420476e-05, "loss": 0.0214, "step": 42450 }, { "grad_norm": 0.2257726937532425, "learning_rate": 4.9051378611181884e-05, "loss": 0.0203, "step": 42460 }, { "grad_norm": 0.2855139672756195, "learning_rate": 4.9030714040252486e-05, "loss": 0.0178, "step": 42470 }, { "grad_norm": 0.256801038980484, "learning_rate": 4.901004963494752e-05, "loss": 0.0125, "step": 42480 }, { "grad_norm": 0.29967451095581055, "learning_rate": 4.898938539879802e-05, "loss": 0.0236, "step": 42490 }, { "grad_norm": 0.21098624169826508, "learning_rate": 4.8968721335334904e-05, "loss": 0.0215, "step": 42500 }, { "grad_norm": 0.3713010847568512, "learning_rate": 4.894805744808912e-05, "loss": 0.0202, "step": 42510 }, { "grad_norm": 0.26197782158851624, "learning_rate": 4.8927393740591524e-05, "loss": 0.0205, "step": 42520 }, { "grad_norm": 0.4219026565551758, "learning_rate": 4.890673021637302e-05, "loss": 0.0232, "step": 42530 }, { "grad_norm": 0.26396599411964417, "learning_rate": 4.888606687896442e-05, "loss": 0.0152, "step": 42540 }, { "grad_norm": 0.26375576853752136, "learning_rate": 4.886540373189652e-05, "loss": 0.0173, "step": 42550 }, { "grad_norm": 0.33881354331970215, "learning_rate": 4.88447407787001e-05, "loss": 0.0198, "step": 42560 }, { "grad_norm": 0.30267462134361267, "learning_rate": 4.8824078022905864e-05, "loss": 0.0216, "step": 42570 }, { "grad_norm": 0.5105552673339844, "learning_rate": 4.880341546804455e-05, "loss": 0.0218, "step": 42580 }, { "grad_norm": 0.3515016734600067, "learning_rate": 4.878275311764679e-05, "loss": 0.0222, "step": 42590 }, { "grad_norm": 0.359363853931427, "learning_rate": 4.8762090975243254e-05, "loss": 0.0164, "step": 42600 }, { "grad_norm": 0.24269936978816986, "learning_rate": 4.874142904436448e-05, "loss": 0.0196, "step": 42610 }, { "grad_norm": 0.21062393486499786, "learning_rate": 4.8720767328541094e-05, "loss": 0.0304, "step": 42620 }, { "grad_norm": 0.3034559190273285, "learning_rate": 4.870010583130358e-05, "loss": 0.0176, "step": 42630 }, { "grad_norm": 0.2078367918729782, "learning_rate": 4.867944455618241e-05, "loss": 0.019, "step": 42640 }, { "grad_norm": 0.22921234369277954, "learning_rate": 4.865878350670807e-05, "loss": 0.0223, "step": 42650 }, { "grad_norm": 0.26991429924964905, "learning_rate": 4.8638122686410914e-05, "loss": 0.0195, "step": 42660 }, { "grad_norm": 0.1413780003786087, "learning_rate": 4.861746209882137e-05, "loss": 0.0157, "step": 42670 }, { "grad_norm": 0.22765344381332397, "learning_rate": 4.859680174746972e-05, "loss": 0.0179, "step": 42680 }, { "grad_norm": 0.21396705508232117, "learning_rate": 4.857614163588629e-05, "loss": 0.0208, "step": 42690 }, { "grad_norm": 0.20900744199752808, "learning_rate": 4.855548176760131e-05, "loss": 0.0161, "step": 42700 }, { "grad_norm": 0.17354881763458252, "learning_rate": 4.8534822146144986e-05, "loss": 0.019, "step": 42710 }, { "grad_norm": 0.2761242389678955, "learning_rate": 4.851416277504749e-05, "loss": 0.0255, "step": 42720 }, { "grad_norm": 0.34825655817985535, "learning_rate": 4.8493503657838923e-05, "loss": 0.0183, "step": 42730 }, { "grad_norm": 0.1812153458595276, "learning_rate": 4.84728447980494e-05, "loss": 0.0173, "step": 42740 }, { "grad_norm": 0.34347689151763916, "learning_rate": 4.8452186199208914e-05, "loss": 0.0181, "step": 42750 }, { "grad_norm": 0.2835133969783783, "learning_rate": 4.843152786484749e-05, "loss": 0.018, "step": 42760 }, { "grad_norm": 0.2283516526222229, "learning_rate": 4.8410869798495035e-05, "loss": 0.0174, "step": 42770 }, { "grad_norm": 0.21373789012432098, "learning_rate": 4.8390212003681486e-05, "loss": 0.018, "step": 42780 }, { "grad_norm": 0.25871971249580383, "learning_rate": 4.836955448393667e-05, "loss": 0.0181, "step": 42790 }, { "grad_norm": 0.2192372828722, "learning_rate": 4.8348897242790406e-05, "loss": 0.0183, "step": 42800 }, { "grad_norm": 0.2472476065158844, "learning_rate": 4.832824028377243e-05, "loss": 0.0148, "step": 42810 }, { "grad_norm": 0.2873596251010895, "learning_rate": 4.830758361041249e-05, "loss": 0.016, "step": 42820 }, { "grad_norm": 0.5915916562080383, "learning_rate": 4.828692722624022e-05, "loss": 0.0181, "step": 42830 }, { "grad_norm": 0.23312360048294067, "learning_rate": 4.826627113478522e-05, "loss": 0.0314, "step": 42840 }, { "grad_norm": 0.29367202520370483, "learning_rate": 4.824561533957708e-05, "loss": 0.0155, "step": 42850 }, { "grad_norm": 0.2535249888896942, "learning_rate": 4.822495984414527e-05, "loss": 0.0189, "step": 42860 }, { "grad_norm": 0.2679515480995178, "learning_rate": 4.8204304652019304e-05, "loss": 0.0181, "step": 42870 }, { "grad_norm": 0.27641600370407104, "learning_rate": 4.8183649766728537e-05, "loss": 0.0179, "step": 42880 }, { "grad_norm": 0.4784647524356842, "learning_rate": 4.816299519180237e-05, "loss": 0.0199, "step": 42890 }, { "grad_norm": 0.2213865965604782, "learning_rate": 4.8142340930770066e-05, "loss": 0.0157, "step": 42900 }, { "grad_norm": 0.3159862458705902, "learning_rate": 4.81216869871609e-05, "loss": 0.0138, "step": 42910 }, { "grad_norm": 0.2464994490146637, "learning_rate": 4.8101033364504064e-05, "loss": 0.0135, "step": 42920 }, { "grad_norm": 0.2542431056499481, "learning_rate": 4.808038006632866e-05, "loss": 0.0196, "step": 42930 }, { "grad_norm": 0.3724849820137024, "learning_rate": 4.805972709616383e-05, "loss": 0.0205, "step": 42940 }, { "grad_norm": 0.2528633177280426, "learning_rate": 4.8039074457538556e-05, "loss": 0.0137, "step": 42950 }, { "grad_norm": 0.21316564083099365, "learning_rate": 4.801842215398184e-05, "loss": 0.0213, "step": 42960 }, { "grad_norm": 0.23147211968898773, "learning_rate": 4.799777018902256e-05, "loss": 0.0179, "step": 42970 }, { "grad_norm": 0.2242017388343811, "learning_rate": 4.797711856618961e-05, "loss": 0.0129, "step": 42980 }, { "grad_norm": 0.2679622769355774, "learning_rate": 4.7956467289011765e-05, "loss": 0.0158, "step": 42990 }, { "grad_norm": 0.2513456344604492, "learning_rate": 4.793581636101778e-05, "loss": 0.0153, "step": 43000 }, { "grad_norm": 0.40281957387924194, "learning_rate": 4.7915165785736326e-05, "loss": 0.0202, "step": 43010 }, { "grad_norm": 0.27139610052108765, "learning_rate": 4.789451556669599e-05, "loss": 0.0174, "step": 43020 }, { "grad_norm": 0.29034385085105896, "learning_rate": 4.7873865707425384e-05, "loss": 0.0193, "step": 43030 }, { "grad_norm": 0.3168696463108063, "learning_rate": 4.7853216211452975e-05, "loss": 0.018, "step": 43040 }, { "grad_norm": 0.30408018827438354, "learning_rate": 4.78325670823072e-05, "loss": 0.0151, "step": 43050 }, { "grad_norm": 0.245699942111969, "learning_rate": 4.781191832351641e-05, "loss": 0.0157, "step": 43060 }, { "grad_norm": 0.24459829926490784, "learning_rate": 4.7791269938608965e-05, "loss": 0.0196, "step": 43070 }, { "grad_norm": 0.28765302896499634, "learning_rate": 4.777062193111305e-05, "loss": 0.0167, "step": 43080 }, { "grad_norm": 0.33161425590515137, "learning_rate": 4.77499743045569e-05, "loss": 0.023, "step": 43090 }, { "grad_norm": 0.4166213274002075, "learning_rate": 4.7729327062468596e-05, "loss": 0.022, "step": 43100 }, { "grad_norm": 0.39884093403816223, "learning_rate": 4.770868020837619e-05, "loss": 0.0207, "step": 43110 }, { "grad_norm": 0.20707589387893677, "learning_rate": 4.768803374580768e-05, "loss": 0.0148, "step": 43120 }, { "grad_norm": 0.1947014182806015, "learning_rate": 4.766738767829094e-05, "loss": 0.0144, "step": 43130 }, { "grad_norm": 0.3052047789096832, "learning_rate": 4.764674200935388e-05, "loss": 0.0213, "step": 43140 }, { "grad_norm": 0.329546719789505, "learning_rate": 4.762609674252424e-05, "loss": 0.021, "step": 43150 }, { "grad_norm": 0.30566951632499695, "learning_rate": 4.760545188132974e-05, "loss": 0.0348, "step": 43160 }, { "grad_norm": 0.3276679515838623, "learning_rate": 4.758480742929801e-05, "loss": 0.0194, "step": 43170 }, { "grad_norm": 0.2235584259033203, "learning_rate": 4.756416338995664e-05, "loss": 0.0215, "step": 43180 }, { "grad_norm": 0.2078646868467331, "learning_rate": 4.7543519766833145e-05, "loss": 0.0141, "step": 43190 }, { "grad_norm": 0.32484716176986694, "learning_rate": 4.752287656345492e-05, "loss": 0.0195, "step": 43200 }, { "grad_norm": 0.29051244258880615, "learning_rate": 4.7502233783349355e-05, "loss": 0.018, "step": 43210 }, { "grad_norm": 0.19885630905628204, "learning_rate": 4.7481591430043694e-05, "loss": 0.0138, "step": 43220 }, { "grad_norm": 0.3440603017807007, "learning_rate": 4.746094950706521e-05, "loss": 0.0211, "step": 43230 }, { "grad_norm": 0.33739393949508667, "learning_rate": 4.744030801794099e-05, "loss": 0.0213, "step": 43240 }, { "grad_norm": 0.2580011188983917, "learning_rate": 4.741966696619813e-05, "loss": 0.0249, "step": 43250 }, { "grad_norm": 0.19840456545352936, "learning_rate": 4.739902635536361e-05, "loss": 0.0259, "step": 43260 }, { "grad_norm": 0.32964888215065, "learning_rate": 4.737838618896436e-05, "loss": 0.0168, "step": 43270 }, { "grad_norm": 0.49799442291259766, "learning_rate": 4.7357746470527203e-05, "loss": 0.022, "step": 43280 }, { "grad_norm": 0.24416659772396088, "learning_rate": 4.73371072035789e-05, "loss": 0.0196, "step": 43290 }, { "grad_norm": 0.33468934893608093, "learning_rate": 4.731646839164616e-05, "loss": 0.0174, "step": 43300 }, { "grad_norm": 0.35452982783317566, "learning_rate": 4.729583003825556e-05, "loss": 0.0197, "step": 43310 }, { "grad_norm": 0.22485440969467163, "learning_rate": 4.727519214693367e-05, "loss": 0.0279, "step": 43320 }, { "grad_norm": 0.3478551208972931, "learning_rate": 4.725455472120689e-05, "loss": 0.0209, "step": 43330 }, { "grad_norm": 0.25628429651260376, "learning_rate": 4.723391776460164e-05, "loss": 0.0221, "step": 43340 }, { "grad_norm": 0.30726319551467896, "learning_rate": 4.7213281280644186e-05, "loss": 0.016, "step": 43350 }, { "grad_norm": 0.22553424537181854, "learning_rate": 4.719264527286075e-05, "loss": 0.0202, "step": 43360 }, { "grad_norm": 0.25764599442481995, "learning_rate": 4.717200974477744e-05, "loss": 0.0222, "step": 43370 }, { "grad_norm": 0.5151048898696899, "learning_rate": 4.715137469992034e-05, "loss": 0.0225, "step": 43380 }, { "grad_norm": 0.24125705659389496, "learning_rate": 4.713074014181539e-05, "loss": 0.0185, "step": 43390 }, { "grad_norm": 0.16485090553760529, "learning_rate": 4.711010607398846e-05, "loss": 0.0133, "step": 43400 }, { "grad_norm": 0.21167448163032532, "learning_rate": 4.708947249996538e-05, "loss": 0.0133, "step": 43410 }, { "grad_norm": 0.25104376673698425, "learning_rate": 4.706883942327183e-05, "loss": 0.0174, "step": 43420 }, { "grad_norm": 0.1330847293138504, "learning_rate": 4.704820684743347e-05, "loss": 0.015, "step": 43430 }, { "grad_norm": 0.2697289288043976, "learning_rate": 4.702757477597581e-05, "loss": 0.0172, "step": 43440 }, { "grad_norm": 0.2694634795188904, "learning_rate": 4.700694321242435e-05, "loss": 0.0175, "step": 43450 }, { "grad_norm": 0.3762044906616211, "learning_rate": 4.698631216030441e-05, "loss": 0.0192, "step": 43460 }, { "grad_norm": 0.20899862051010132, "learning_rate": 4.6965681623141314e-05, "loss": 0.0131, "step": 43470 }, { "grad_norm": 0.2358524203300476, "learning_rate": 4.694505160446024e-05, "loss": 0.0112, "step": 43480 }, { "grad_norm": 0.2286195456981659, "learning_rate": 4.692442210778628e-05, "loss": 0.015, "step": 43490 }, { "grad_norm": 0.3841545283794403, "learning_rate": 4.690379313664448e-05, "loss": 0.0172, "step": 43500 }, { "grad_norm": 0.3269505798816681, "learning_rate": 4.688316469455973e-05, "loss": 0.0222, "step": 43510 }, { "grad_norm": 0.25690996646881104, "learning_rate": 4.6862536785056914e-05, "loss": 0.02, "step": 43520 }, { "grad_norm": 0.35102638602256775, "learning_rate": 4.684190941166073e-05, "loss": 0.0198, "step": 43530 }, { "grad_norm": 0.28860101103782654, "learning_rate": 4.682128257789587e-05, "loss": 0.0197, "step": 43540 }, { "grad_norm": 0.17702777683734894, "learning_rate": 4.680065628728687e-05, "loss": 0.0194, "step": 43550 }, { "grad_norm": 0.36013662815093994, "learning_rate": 4.678003054335822e-05, "loss": 0.022, "step": 43560 }, { "grad_norm": 0.31064119935035706, "learning_rate": 4.675940534963428e-05, "loss": 0.0207, "step": 43570 }, { "grad_norm": 0.2668758034706116, "learning_rate": 4.673878070963931e-05, "loss": 0.0183, "step": 43580 }, { "grad_norm": 0.17554573714733124, "learning_rate": 4.671815662689756e-05, "loss": 0.0119, "step": 43590 }, { "grad_norm": 0.23182415962219238, "learning_rate": 4.669753310493306e-05, "loss": 0.015, "step": 43600 }, { "grad_norm": 0.333164244890213, "learning_rate": 4.6676910147269845e-05, "loss": 0.0163, "step": 43610 }, { "grad_norm": 0.2840135097503662, "learning_rate": 4.665628775743177e-05, "loss": 0.0176, "step": 43620 }, { "grad_norm": 0.26018816232681274, "learning_rate": 4.66356659389427e-05, "loss": 0.014, "step": 43630 }, { "grad_norm": 0.49686017632484436, "learning_rate": 4.661504469532629e-05, "loss": 0.0208, "step": 43640 }, { "grad_norm": 0.23111915588378906, "learning_rate": 4.659442403010617e-05, "loss": 0.0195, "step": 43650 }, { "grad_norm": 0.2433045357465744, "learning_rate": 4.6573803946805845e-05, "loss": 0.0172, "step": 43660 }, { "grad_norm": 0.31041786074638367, "learning_rate": 4.6553184448948695e-05, "loss": 0.0188, "step": 43670 }, { "grad_norm": 0.16072911024093628, "learning_rate": 4.653256554005807e-05, "loss": 0.0137, "step": 43680 }, { "grad_norm": 0.22170156240463257, "learning_rate": 4.6511947223657145e-05, "loss": 0.0222, "step": 43690 }, { "grad_norm": 0.34234511852264404, "learning_rate": 4.649132950326906e-05, "loss": 0.0211, "step": 43700 }, { "grad_norm": 0.42496368288993835, "learning_rate": 4.647071238241679e-05, "loss": 0.0281, "step": 43710 }, { "grad_norm": 0.19986765086650848, "learning_rate": 4.6450095864623264e-05, "loss": 0.0176, "step": 43720 }, { "grad_norm": 0.29905590415000916, "learning_rate": 4.642947995341123e-05, "loss": 0.0165, "step": 43730 }, { "grad_norm": 0.2019273191690445, "learning_rate": 4.640886465230345e-05, "loss": 0.0192, "step": 43740 }, { "grad_norm": 0.41211119294166565, "learning_rate": 4.6388249964822485e-05, "loss": 0.0145, "step": 43750 }, { "grad_norm": 0.3078845739364624, "learning_rate": 4.63676358944908e-05, "loss": 0.0133, "step": 43760 }, { "grad_norm": 0.4345298111438751, "learning_rate": 4.634702244483081e-05, "loss": 0.0193, "step": 43770 }, { "grad_norm": 0.12879541516304016, "learning_rate": 4.632640961936475e-05, "loss": 0.0154, "step": 43780 }, { "grad_norm": 0.2280968427658081, "learning_rate": 4.6305797421614835e-05, "loss": 0.0121, "step": 43790 }, { "grad_norm": 0.3544617295265198, "learning_rate": 4.6285185855103105e-05, "loss": 0.0148, "step": 43800 }, { "grad_norm": 0.4285740256309509, "learning_rate": 4.626457492335151e-05, "loss": 0.0141, "step": 43810 }, { "grad_norm": 0.31874802708625793, "learning_rate": 4.624396462988188e-05, "loss": 0.0129, "step": 43820 }, { "grad_norm": 0.19715529680252075, "learning_rate": 4.6223354978215985e-05, "loss": 0.0186, "step": 43830 }, { "grad_norm": 0.3711439371109009, "learning_rate": 4.620274597187544e-05, "loss": 0.0177, "step": 43840 }, { "grad_norm": 0.28479743003845215, "learning_rate": 4.6182137614381736e-05, "loss": 0.0131, "step": 43850 }, { "grad_norm": 0.30690300464630127, "learning_rate": 4.616152990925631e-05, "loss": 0.0131, "step": 43860 }, { "grad_norm": 0.1916169822216034, "learning_rate": 4.614092286002041e-05, "loss": 0.0123, "step": 43870 }, { "grad_norm": 0.31678318977355957, "learning_rate": 4.612031647019528e-05, "loss": 0.0145, "step": 43880 }, { "grad_norm": 0.23453998565673828, "learning_rate": 4.609971074330193e-05, "loss": 0.0144, "step": 43890 }, { "grad_norm": 0.36100244522094727, "learning_rate": 4.607910568286136e-05, "loss": 0.0187, "step": 43900 }, { "grad_norm": 0.3705906867980957, "learning_rate": 4.605850129239438e-05, "loss": 0.0184, "step": 43910 }, { "grad_norm": 0.23143893480300903, "learning_rate": 4.603789757542174e-05, "loss": 0.0149, "step": 43920 }, { "grad_norm": 0.3752548396587372, "learning_rate": 4.601729453546404e-05, "loss": 0.0181, "step": 43930 }, { "grad_norm": 0.263058066368103, "learning_rate": 4.599669217604177e-05, "loss": 0.0154, "step": 43940 }, { "grad_norm": 0.31535208225250244, "learning_rate": 4.597609050067532e-05, "loss": 0.0227, "step": 43950 }, { "grad_norm": 0.26035261154174805, "learning_rate": 4.595548951288495e-05, "loss": 0.0197, "step": 43960 }, { "grad_norm": 0.25568535923957825, "learning_rate": 4.593488921619081e-05, "loss": 0.0179, "step": 43970 }, { "grad_norm": 0.4339941143989563, "learning_rate": 4.591428961411289e-05, "loss": 0.019, "step": 43980 }, { "grad_norm": 0.22628799080848694, "learning_rate": 4.589369071017117e-05, "loss": 0.0235, "step": 43990 }, { "grad_norm": 0.2485271543264389, "learning_rate": 4.587309250788538e-05, "loss": 0.0173, "step": 44000 }, { "grad_norm": 0.4754122495651245, "learning_rate": 4.585249501077522e-05, "loss": 0.0171, "step": 44010 }, { "grad_norm": 0.16849388182163239, "learning_rate": 4.583189822236021e-05, "loss": 0.0197, "step": 44020 }, { "grad_norm": 0.19470366835594177, "learning_rate": 4.5811302146159816e-05, "loss": 0.0208, "step": 44030 }, { "grad_norm": 0.24635672569274902, "learning_rate": 4.579070678569332e-05, "loss": 0.0193, "step": 44040 }, { "grad_norm": 0.4280654489994049, "learning_rate": 4.5770112144479904e-05, "loss": 0.0202, "step": 44050 }, { "grad_norm": 0.2821556031703949, "learning_rate": 4.5749518226038645e-05, "loss": 0.0168, "step": 44060 }, { "grad_norm": 0.28422069549560547, "learning_rate": 4.572892503388845e-05, "loss": 0.0195, "step": 44070 }, { "grad_norm": 0.2030417025089264, "learning_rate": 4.570833257154817e-05, "loss": 0.0142, "step": 44080 }, { "grad_norm": 0.32936424016952515, "learning_rate": 4.568774084253646e-05, "loss": 0.0218, "step": 44090 }, { "grad_norm": 0.3508985638618469, "learning_rate": 4.566714985037191e-05, "loss": 0.0244, "step": 44100 }, { "grad_norm": 0.3140784800052643, "learning_rate": 4.564655959857295e-05, "loss": 0.0203, "step": 44110 }, { "grad_norm": 0.30910879373550415, "learning_rate": 4.5625970090657893e-05, "loss": 0.0225, "step": 44120 }, { "grad_norm": 0.2544829249382019, "learning_rate": 4.560538133014491e-05, "loss": 0.0217, "step": 44130 }, { "grad_norm": 0.2897930443286896, "learning_rate": 4.5584793320552055e-05, "loss": 0.0197, "step": 44140 }, { "grad_norm": 0.29433831572532654, "learning_rate": 4.556420606539728e-05, "loss": 0.0187, "step": 44150 }, { "grad_norm": 0.20784428715705872, "learning_rate": 4.554361956819836e-05, "loss": 0.0186, "step": 44160 }, { "grad_norm": 0.24021416902542114, "learning_rate": 4.552303383247299e-05, "loss": 0.0149, "step": 44170 }, { "grad_norm": 0.25068721175193787, "learning_rate": 4.5502448861738664e-05, "loss": 0.0138, "step": 44180 }, { "grad_norm": 0.2803000211715698, "learning_rate": 4.548186465951284e-05, "loss": 0.0242, "step": 44190 }, { "grad_norm": 0.16567634046077728, "learning_rate": 4.546128122931277e-05, "loss": 0.016, "step": 44200 }, { "grad_norm": 0.3240645229816437, "learning_rate": 4.544069857465561e-05, "loss": 0.0156, "step": 44210 }, { "grad_norm": 0.3064705431461334, "learning_rate": 4.542011669905837e-05, "loss": 0.013, "step": 44220 }, { "grad_norm": 0.2794076204299927, "learning_rate": 4.53995356060379e-05, "loss": 0.0137, "step": 44230 }, { "grad_norm": 0.16290771961212158, "learning_rate": 4.5378955299110994e-05, "loss": 0.014, "step": 44240 }, { "grad_norm": 0.314897745847702, "learning_rate": 4.5358375781794233e-05, "loss": 0.0214, "step": 44250 }, { "grad_norm": 0.27458709478378296, "learning_rate": 4.533779705760411e-05, "loss": 0.0229, "step": 44260 }, { "grad_norm": 0.22838132083415985, "learning_rate": 4.5317219130056934e-05, "loss": 0.0192, "step": 44270 }, { "grad_norm": 0.1906471848487854, "learning_rate": 4.5296642002668964e-05, "loss": 0.0222, "step": 44280 }, { "grad_norm": 0.4162900149822235, "learning_rate": 4.5276065678956216e-05, "loss": 0.021, "step": 44290 }, { "grad_norm": 0.2506970465183258, "learning_rate": 4.525549016243466e-05, "loss": 0.0182, "step": 44300 }, { "grad_norm": 0.18841814994812012, "learning_rate": 4.523491545662008e-05, "loss": 0.0153, "step": 44310 }, { "grad_norm": 0.338509738445282, "learning_rate": 4.5214341565028104e-05, "loss": 0.0212, "step": 44320 }, { "grad_norm": 0.3612939417362213, "learning_rate": 4.519376849117428e-05, "loss": 0.0192, "step": 44330 }, { "grad_norm": 0.23779869079589844, "learning_rate": 4.517319623857395e-05, "loss": 0.0231, "step": 44340 }, { "grad_norm": 0.45128434896469116, "learning_rate": 4.5152624810742394e-05, "loss": 0.0201, "step": 44350 }, { "grad_norm": 0.32411667704582214, "learning_rate": 4.5132054211194664e-05, "loss": 0.0178, "step": 44360 }, { "grad_norm": 0.2636305093765259, "learning_rate": 4.511148444344574e-05, "loss": 0.0179, "step": 44370 }, { "grad_norm": 0.4497871398925781, "learning_rate": 4.509091551101041e-05, "loss": 0.0277, "step": 44380 }, { "grad_norm": 0.24318614602088928, "learning_rate": 4.507034741740338e-05, "loss": 0.0164, "step": 44390 }, { "grad_norm": 0.15480607748031616, "learning_rate": 4.5049780166139145e-05, "loss": 0.0212, "step": 44400 }, { "grad_norm": 0.4604562520980835, "learning_rate": 4.5029213760732075e-05, "loss": 0.0276, "step": 44410 }, { "grad_norm": 0.32032275199890137, "learning_rate": 4.5008648204696434e-05, "loss": 0.0211, "step": 44420 }, { "grad_norm": 0.2250615507364273, "learning_rate": 4.4988083501546284e-05, "loss": 0.0143, "step": 44430 }, { "grad_norm": 0.19652701914310455, "learning_rate": 4.4967519654795606e-05, "loss": 0.0155, "step": 44440 }, { "grad_norm": 0.20473046600818634, "learning_rate": 4.494695666795816e-05, "loss": 0.0189, "step": 44450 }, { "grad_norm": 0.321523517370224, "learning_rate": 4.4926394544547644e-05, "loss": 0.0189, "step": 44460 }, { "grad_norm": 0.35713115334510803, "learning_rate": 4.490583328807752e-05, "loss": 0.0182, "step": 44470 }, { "grad_norm": 0.22221580147743225, "learning_rate": 4.488527290206117e-05, "loss": 0.0147, "step": 44480 }, { "grad_norm": 0.17210493981838226, "learning_rate": 4.48647133900118e-05, "loss": 0.013, "step": 44490 }, { "grad_norm": 0.2330578863620758, "learning_rate": 4.484415475544244e-05, "loss": 0.0145, "step": 44500 }, { "grad_norm": 0.19764605164527893, "learning_rate": 4.4823597001866035e-05, "loss": 0.0121, "step": 44510 }, { "grad_norm": 0.5295576453208923, "learning_rate": 4.480304013279532e-05, "loss": 0.0184, "step": 44520 }, { "grad_norm": 0.20029424130916595, "learning_rate": 4.478248415174292e-05, "loss": 0.0121, "step": 44530 }, { "grad_norm": 0.2145150750875473, "learning_rate": 4.476192906222126e-05, "loss": 0.0226, "step": 44540 }, { "grad_norm": 0.3650346100330353, "learning_rate": 4.474137486774268e-05, "loss": 0.0247, "step": 44550 }, { "grad_norm": 0.2190667688846588, "learning_rate": 4.4720821571819296e-05, "loss": 0.0166, "step": 44560 }, { "grad_norm": 0.2275903820991516, "learning_rate": 4.470026917796314e-05, "loss": 0.0243, "step": 44570 }, { "grad_norm": 0.32468292117118835, "learning_rate": 4.4679717689686005e-05, "loss": 0.0306, "step": 44580 }, { "grad_norm": 0.273446649312973, "learning_rate": 4.465916711049963e-05, "loss": 0.0131, "step": 44590 }, { "grad_norm": 0.3175617456436157, "learning_rate": 4.4638617443915524e-05, "loss": 0.0183, "step": 44600 }, { "grad_norm": 0.42429468035697937, "learning_rate": 4.4618068693445055e-05, "loss": 0.0191, "step": 44610 }, { "grad_norm": 0.3668299913406372, "learning_rate": 4.459752086259946e-05, "loss": 0.0163, "step": 44620 }, { "grad_norm": 0.5418717265129089, "learning_rate": 4.457697395488977e-05, "loss": 0.0198, "step": 44630 }, { "grad_norm": 0.3071218729019165, "learning_rate": 4.455642797382693e-05, "loss": 0.0246, "step": 44640 }, { "grad_norm": 0.17713965475559235, "learning_rate": 4.4535882922921643e-05, "loss": 0.0198, "step": 44650 }, { "grad_norm": 0.3537560701370239, "learning_rate": 4.451533880568455e-05, "loss": 0.0191, "step": 44660 }, { "grad_norm": 0.4000324606895447, "learning_rate": 4.449479562562603e-05, "loss": 0.0206, "step": 44670 }, { "grad_norm": 0.26971471309661865, "learning_rate": 4.447425338625639e-05, "loss": 0.0227, "step": 44680 }, { "grad_norm": 0.273456335067749, "learning_rate": 4.4453712091085705e-05, "loss": 0.0192, "step": 44690 }, { "grad_norm": 0.3047838807106018, "learning_rate": 4.443317174362392e-05, "loss": 0.0217, "step": 44700 }, { "grad_norm": 0.23637279868125916, "learning_rate": 4.441263234738085e-05, "loss": 0.0147, "step": 44710 }, { "grad_norm": 0.3325249254703522, "learning_rate": 4.4392093905866086e-05, "loss": 0.0136, "step": 44720 }, { "grad_norm": 0.2813401222229004, "learning_rate": 4.437155642258911e-05, "loss": 0.0195, "step": 44730 }, { "grad_norm": 0.3256838619709015, "learning_rate": 4.4351019901059177e-05, "loss": 0.0214, "step": 44740 }, { "grad_norm": 0.3025226891040802, "learning_rate": 4.433048434478547e-05, "loss": 0.0269, "step": 44750 }, { "grad_norm": 0.22434361279010773, "learning_rate": 4.4309949757276906e-05, "loss": 0.0192, "step": 44760 }, { "grad_norm": 0.30371084809303284, "learning_rate": 4.428941614204233e-05, "loss": 0.0309, "step": 44770 }, { "grad_norm": 0.18764010071754456, "learning_rate": 4.426888350259034e-05, "loss": 0.0178, "step": 44780 }, { "grad_norm": 0.22709275782108307, "learning_rate": 4.424835184242939e-05, "loss": 0.025, "step": 44790 }, { "grad_norm": 0.19596083462238312, "learning_rate": 4.422782116506784e-05, "loss": 0.0126, "step": 44800 }, { "grad_norm": 0.608156144618988, "learning_rate": 4.4207291474013754e-05, "loss": 0.0199, "step": 44810 }, { "grad_norm": 0.277135968208313, "learning_rate": 4.418676277277514e-05, "loss": 0.0231, "step": 44820 }, { "grad_norm": 0.16163136065006256, "learning_rate": 4.416623506485976e-05, "loss": 0.0249, "step": 44830 }, { "grad_norm": 0.2797115743160248, "learning_rate": 4.414570835377527e-05, "loss": 0.0156, "step": 44840 }, { "grad_norm": 0.25106438994407654, "learning_rate": 4.412518264302909e-05, "loss": 0.0175, "step": 44850 }, { "grad_norm": 0.2896294593811035, "learning_rate": 4.4104657936128535e-05, "loss": 0.0135, "step": 44860 }, { "grad_norm": 0.3500630855560303, "learning_rate": 4.408413423658071e-05, "loss": 0.0194, "step": 44870 }, { "grad_norm": 0.28793224692344666, "learning_rate": 4.406361154789253e-05, "loss": 0.0183, "step": 44880 }, { "grad_norm": 0.39837756752967834, "learning_rate": 4.4043089873570776e-05, "loss": 0.0186, "step": 44890 }, { "grad_norm": 0.3265704810619354, "learning_rate": 4.4022569217122025e-05, "loss": 0.0168, "step": 44900 }, { "grad_norm": 0.2651529014110565, "learning_rate": 4.400204958205274e-05, "loss": 0.0124, "step": 44910 }, { "grad_norm": 0.3565864861011505, "learning_rate": 4.3981530971869125e-05, "loss": 0.023, "step": 44920 }, { "grad_norm": 0.22892558574676514, "learning_rate": 4.396101339007727e-05, "loss": 0.0179, "step": 44930 }, { "grad_norm": 0.11902187764644623, "learning_rate": 4.394049684018304e-05, "loss": 0.0206, "step": 44940 }, { "grad_norm": 0.27994364500045776, "learning_rate": 4.3919981325692186e-05, "loss": 0.0152, "step": 44950 }, { "grad_norm": 0.1660320907831192, "learning_rate": 4.389946685011024e-05, "loss": 0.0143, "step": 44960 }, { "grad_norm": 0.3455781638622284, "learning_rate": 4.387895341694255e-05, "loss": 0.0261, "step": 44970 }, { "grad_norm": 0.23487205803394318, "learning_rate": 4.3858441029694324e-05, "loss": 0.0215, "step": 44980 }, { "grad_norm": 0.22065329551696777, "learning_rate": 4.3837929691870527e-05, "loss": 0.0125, "step": 44990 }, { "grad_norm": 0.33756035566329956, "learning_rate": 4.381741940697604e-05, "loss": 0.0241, "step": 45000 }, { "grad_norm": 0.1263425052165985, "learning_rate": 4.379691017851547e-05, "loss": 0.0145, "step": 45010 }, { "grad_norm": 0.2477971464395523, "learning_rate": 4.3776402009993304e-05, "loss": 0.0223, "step": 45020 }, { "grad_norm": 0.17643749713897705, "learning_rate": 4.3755894904913794e-05, "loss": 0.0146, "step": 45030 }, { "grad_norm": 0.2340431660413742, "learning_rate": 4.373538886678109e-05, "loss": 0.0157, "step": 45040 }, { "grad_norm": 0.25727608799934387, "learning_rate": 4.371488389909909e-05, "loss": 0.015, "step": 45050 }, { "grad_norm": 0.2824594974517822, "learning_rate": 4.3694380005371515e-05, "loss": 0.0169, "step": 45060 }, { "grad_norm": 0.23900195956230164, "learning_rate": 4.367387718910196e-05, "loss": 0.0164, "step": 45070 }, { "grad_norm": 0.2155352085828781, "learning_rate": 4.3653375453793764e-05, "loss": 0.0163, "step": 45080 }, { "grad_norm": 0.22088158130645752, "learning_rate": 4.3632874802950136e-05, "loss": 0.0169, "step": 45090 }, { "grad_norm": 0.19026967883110046, "learning_rate": 4.3612375240074034e-05, "loss": 0.0154, "step": 45100 }, { "grad_norm": 0.2352093756198883, "learning_rate": 4.3591876768668325e-05, "loss": 0.018, "step": 45110 }, { "grad_norm": 0.3282748758792877, "learning_rate": 4.3571379392235605e-05, "loss": 0.0154, "step": 45120 }, { "grad_norm": 0.3551333546638489, "learning_rate": 4.3550883114278335e-05, "loss": 0.0155, "step": 45130 }, { "grad_norm": 0.24344798922538757, "learning_rate": 4.353038793829876e-05, "loss": 0.0172, "step": 45140 }, { "grad_norm": 0.3295263350009918, "learning_rate": 4.350989386779891e-05, "loss": 0.0128, "step": 45150 }, { "grad_norm": 0.3850118815898895, "learning_rate": 4.3489400906280724e-05, "loss": 0.0227, "step": 45160 }, { "grad_norm": 0.26351988315582275, "learning_rate": 4.3468909057245845e-05, "loss": 0.0165, "step": 45170 }, { "grad_norm": 0.18094445765018463, "learning_rate": 4.3448418324195794e-05, "loss": 0.019, "step": 45180 }, { "grad_norm": 0.3650659918785095, "learning_rate": 4.342792871063184e-05, "loss": 0.018, "step": 45190 }, { "grad_norm": 0.23070411384105682, "learning_rate": 4.3407440220055145e-05, "loss": 0.014, "step": 45200 }, { "grad_norm": 0.19452597200870514, "learning_rate": 4.33869528559666e-05, "loss": 0.01, "step": 45210 }, { "grad_norm": 0.44360795617103577, "learning_rate": 4.336646662186696e-05, "loss": 0.0102, "step": 45220 }, { "grad_norm": 0.27981993556022644, "learning_rate": 4.334598152125672e-05, "loss": 0.0153, "step": 45230 }, { "grad_norm": 0.22799047827720642, "learning_rate": 4.3325497557636276e-05, "loss": 0.0208, "step": 45240 }, { "grad_norm": 0.13990801572799683, "learning_rate": 4.330501473450574e-05, "loss": 0.0163, "step": 45250 }, { "grad_norm": 0.5192397236824036, "learning_rate": 4.328453305536507e-05, "loss": 0.0173, "step": 45260 }, { "grad_norm": 0.3477577865123749, "learning_rate": 4.326405252371404e-05, "loss": 0.0175, "step": 45270 }, { "grad_norm": 0.4849502742290497, "learning_rate": 4.324357314305221e-05, "loss": 0.0162, "step": 45280 }, { "grad_norm": 0.2781824767589569, "learning_rate": 4.3223094916878945e-05, "loss": 0.0173, "step": 45290 }, { "grad_norm": 0.25706812739372253, "learning_rate": 4.320261784869338e-05, "loss": 0.0228, "step": 45300 }, { "grad_norm": 0.26254329085350037, "learning_rate": 4.318214194199455e-05, "loss": 0.0193, "step": 45310 }, { "grad_norm": 0.3163752853870392, "learning_rate": 4.316166720028118e-05, "loss": 0.0138, "step": 45320 }, { "grad_norm": 0.3326606750488281, "learning_rate": 4.3141193627051864e-05, "loss": 0.015, "step": 45330 }, { "grad_norm": 0.37479129433631897, "learning_rate": 4.312072122580496e-05, "loss": 0.0213, "step": 45340 }, { "grad_norm": 0.27478280663490295, "learning_rate": 4.3100250000038646e-05, "loss": 0.0147, "step": 45350 }, { "grad_norm": 0.5101400017738342, "learning_rate": 4.307977995325091e-05, "loss": 0.0223, "step": 45360 }, { "grad_norm": 0.33616921305656433, "learning_rate": 4.30593110889395e-05, "loss": 0.0231, "step": 45370 }, { "grad_norm": 0.2407042682170868, "learning_rate": 4.3038843410602016e-05, "loss": 0.0132, "step": 45380 }, { "grad_norm": 0.32275891304016113, "learning_rate": 4.3018376921735774e-05, "loss": 0.0196, "step": 45390 }, { "grad_norm": 0.1614820808172226, "learning_rate": 4.299791162583799e-05, "loss": 0.0132, "step": 45400 }, { "grad_norm": 0.2889511287212372, "learning_rate": 4.29774475264056e-05, "loss": 0.0177, "step": 45410 }, { "grad_norm": 0.2931652367115021, "learning_rate": 4.2956984626935365e-05, "loss": 0.0136, "step": 45420 }, { "grad_norm": 0.33179447054862976, "learning_rate": 4.293652293092383e-05, "loss": 0.0166, "step": 45430 }, { "grad_norm": 0.21981485188007355, "learning_rate": 4.2916062441867324e-05, "loss": 0.0151, "step": 45440 }, { "grad_norm": 0.26673734188079834, "learning_rate": 4.289560316326201e-05, "loss": 0.0164, "step": 45450 }, { "grad_norm": 0.37436044216156006, "learning_rate": 4.28751450986038e-05, "loss": 0.0229, "step": 45460 }, { "grad_norm": 0.3882718086242676, "learning_rate": 4.2854688251388444e-05, "loss": 0.0194, "step": 45470 }, { "grad_norm": 0.3390754461288452, "learning_rate": 4.2834232625111425e-05, "loss": 0.0193, "step": 45480 }, { "grad_norm": 0.46069425344467163, "learning_rate": 4.2813778223268086e-05, "loss": 0.0178, "step": 45490 }, { "grad_norm": 0.27744999527931213, "learning_rate": 4.2793325049353477e-05, "loss": 0.0144, "step": 45500 }, { "grad_norm": 0.2930518686771393, "learning_rate": 4.2772873106862535e-05, "loss": 0.0184, "step": 45510 }, { "grad_norm": 0.2176855504512787, "learning_rate": 4.275242239928993e-05, "loss": 0.0158, "step": 45520 }, { "grad_norm": 0.17955242097377777, "learning_rate": 4.273197293013009e-05, "loss": 0.013, "step": 45530 }, { "grad_norm": 0.17468546330928802, "learning_rate": 4.271152470287731e-05, "loss": 0.015, "step": 45540 }, { "grad_norm": 0.2855221927165985, "learning_rate": 4.2691077721025594e-05, "loss": 0.0168, "step": 45550 }, { "grad_norm": 0.2944665551185608, "learning_rate": 4.267063198806883e-05, "loss": 0.017, "step": 45560 }, { "grad_norm": 0.23018909990787506, "learning_rate": 4.2650187507500574e-05, "loss": 0.0145, "step": 45570 }, { "grad_norm": 0.30694055557250977, "learning_rate": 4.2629744282814275e-05, "loss": 0.0181, "step": 45580 }, { "grad_norm": 0.2198864370584488, "learning_rate": 4.2609302317503074e-05, "loss": 0.0125, "step": 45590 }, { "grad_norm": 0.19812652468681335, "learning_rate": 4.258886161505999e-05, "loss": 0.0183, "step": 45600 }, { "grad_norm": 0.274983674287796, "learning_rate": 4.2568422178977775e-05, "loss": 0.0155, "step": 45610 }, { "grad_norm": 0.3643975555896759, "learning_rate": 4.254798401274894e-05, "loss": 0.0225, "step": 45620 }, { "grad_norm": 0.24288822710514069, "learning_rate": 4.252754711986583e-05, "loss": 0.0221, "step": 45630 }, { "grad_norm": 0.273899644613266, "learning_rate": 4.250711150382052e-05, "loss": 0.0194, "step": 45640 }, { "grad_norm": 0.27424654364585876, "learning_rate": 4.248667716810495e-05, "loss": 0.0215, "step": 45650 }, { "grad_norm": 0.24850016832351685, "learning_rate": 4.246624411621074e-05, "loss": 0.0168, "step": 45660 }, { "grad_norm": 0.46139389276504517, "learning_rate": 4.244581235162938e-05, "loss": 0.0204, "step": 45670 }, { "grad_norm": 0.3820473849773407, "learning_rate": 4.2425381877852075e-05, "loss": 0.0213, "step": 45680 }, { "grad_norm": 0.44145438075065613, "learning_rate": 4.2404952698369856e-05, "loss": 0.0206, "step": 45690 }, { "grad_norm": 0.21571826934814453, "learning_rate": 4.238452481667349e-05, "loss": 0.019, "step": 45700 }, { "grad_norm": 0.15883883833885193, "learning_rate": 4.2364098236253526e-05, "loss": 0.0202, "step": 45710 }, { "grad_norm": 0.305660218000412, "learning_rate": 4.2343672960600356e-05, "loss": 0.0166, "step": 45720 }, { "grad_norm": 0.2863485515117645, "learning_rate": 4.232324899320406e-05, "loss": 0.0152, "step": 45730 }, { "grad_norm": 0.20821718871593475, "learning_rate": 4.230282633755457e-05, "loss": 0.0134, "step": 45740 }, { "grad_norm": 0.19161026179790497, "learning_rate": 4.2282404997141515e-05, "loss": 0.0126, "step": 45750 }, { "grad_norm": 0.23714694380760193, "learning_rate": 4.2261984975454397e-05, "loss": 0.0146, "step": 45760 }, { "grad_norm": 0.3200992941856384, "learning_rate": 4.224156627598239e-05, "loss": 0.0142, "step": 45770 }, { "grad_norm": 0.21731741726398468, "learning_rate": 4.222114890221453e-05, "loss": 0.0151, "step": 45780 }, { "grad_norm": 0.345989853143692, "learning_rate": 4.2200732857639546e-05, "loss": 0.0193, "step": 45790 }, { "grad_norm": 0.240544855594635, "learning_rate": 4.2180318145746035e-05, "loss": 0.0178, "step": 45800 }, { "grad_norm": 0.3360804617404938, "learning_rate": 4.215990477002227e-05, "loss": 0.0209, "step": 45810 }, { "grad_norm": 0.2783425450325012, "learning_rate": 4.2139492733956356e-05, "loss": 0.0146, "step": 45820 }, { "grad_norm": 0.18049724400043488, "learning_rate": 4.211908204103615e-05, "loss": 0.0242, "step": 45830 }, { "grad_norm": 0.23602694272994995, "learning_rate": 4.2098672694749265e-05, "loss": 0.0295, "step": 45840 }, { "grad_norm": 0.32412245869636536, "learning_rate": 4.2078264698583133e-05, "loss": 0.0251, "step": 45850 }, { "grad_norm": 0.32287976145744324, "learning_rate": 4.205785805602488e-05, "loss": 0.0193, "step": 45860 }, { "grad_norm": 0.27840378880500793, "learning_rate": 4.203745277056149e-05, "loss": 0.0183, "step": 45870 }, { "grad_norm": 0.20404137670993805, "learning_rate": 4.201704884567964e-05, "loss": 0.017, "step": 45880 }, { "grad_norm": 0.2104872167110443, "learning_rate": 4.1996646284865816e-05, "loss": 0.0222, "step": 45890 }, { "grad_norm": 0.19928693771362305, "learning_rate": 4.197624509160625e-05, "loss": 0.0191, "step": 45900 }, { "grad_norm": 0.41450199484825134, "learning_rate": 4.195584526938692e-05, "loss": 0.0165, "step": 45910 }, { "grad_norm": 0.195278599858284, "learning_rate": 4.193544682169365e-05, "loss": 0.0197, "step": 45920 }, { "grad_norm": 0.22426731884479523, "learning_rate": 4.1915049752011946e-05, "loss": 0.0157, "step": 45930 }, { "grad_norm": 0.281009316444397, "learning_rate": 4.189465406382712e-05, "loss": 0.0137, "step": 45940 }, { "grad_norm": 0.2047388106584549, "learning_rate": 4.187425976062422e-05, "loss": 0.0167, "step": 45950 }, { "grad_norm": 0.14219960570335388, "learning_rate": 4.18538668458881e-05, "loss": 0.0124, "step": 45960 }, { "grad_norm": 0.3187878727912903, "learning_rate": 4.183347532310333e-05, "loss": 0.0141, "step": 45970 }, { "grad_norm": 0.16283968091011047, "learning_rate": 4.181308519575429e-05, "loss": 0.017, "step": 45980 }, { "grad_norm": 0.50107741355896, "learning_rate": 4.179269646732507e-05, "loss": 0.0291, "step": 45990 }, { "grad_norm": 0.2132793366909027, "learning_rate": 4.177230914129954e-05, "loss": 0.0177, "step": 46000 }, { "grad_norm": 0.22777774930000305, "learning_rate": 4.175192322116136e-05, "loss": 0.0156, "step": 46010 }, { "grad_norm": 0.30336472392082214, "learning_rate": 4.173153871039391e-05, "loss": 0.018, "step": 46020 }, { "grad_norm": 0.3698276877403259, "learning_rate": 4.171115561248036e-05, "loss": 0.0176, "step": 46030 }, { "grad_norm": 0.3091701865196228, "learning_rate": 4.16907739309036e-05, "loss": 0.0158, "step": 46040 }, { "grad_norm": 0.31765857338905334, "learning_rate": 4.167039366914633e-05, "loss": 0.0142, "step": 46050 }, { "grad_norm": 0.19718943536281586, "learning_rate": 4.165001483069096e-05, "loss": 0.0151, "step": 46060 }, { "grad_norm": 0.22412174940109253, "learning_rate": 4.16296374190197e-05, "loss": 0.0276, "step": 46070 }, { "grad_norm": 0.2095923274755478, "learning_rate": 4.1609261437614464e-05, "loss": 0.0181, "step": 46080 }, { "grad_norm": 0.530002772808075, "learning_rate": 4.158888688995696e-05, "loss": 0.0224, "step": 46090 }, { "grad_norm": 0.2902057468891144, "learning_rate": 4.1568513779528645e-05, "loss": 0.0154, "step": 46100 }, { "grad_norm": 0.2519642412662506, "learning_rate": 4.1548142109810704e-05, "loss": 0.0205, "step": 46110 }, { "grad_norm": 0.29828691482543945, "learning_rate": 4.152777188428414e-05, "loss": 0.0139, "step": 46120 }, { "grad_norm": 0.351974755525589, "learning_rate": 4.1507403106429646e-05, "loss": 0.0221, "step": 46130 }, { "grad_norm": 0.32698678970336914, "learning_rate": 4.148703577972768e-05, "loss": 0.0174, "step": 46140 }, { "grad_norm": 0.19509051740169525, "learning_rate": 4.146666990765846e-05, "loss": 0.0123, "step": 46150 }, { "grad_norm": 0.2212103009223938, "learning_rate": 4.1446305493701986e-05, "loss": 0.0171, "step": 46160 }, { "grad_norm": 0.3732857406139374, "learning_rate": 4.142594254133796e-05, "loss": 0.0183, "step": 46170 }, { "grad_norm": 0.2930203378200531, "learning_rate": 4.1405581054045835e-05, "loss": 0.0177, "step": 46180 }, { "grad_norm": 0.24452434480190277, "learning_rate": 4.1385221035304864e-05, "loss": 0.0115, "step": 46190 }, { "grad_norm": 0.2511179745197296, "learning_rate": 4.136486248859398e-05, "loss": 0.0188, "step": 46200 }, { "grad_norm": 0.18798477947711945, "learning_rate": 4.1344505417391955e-05, "loss": 0.018, "step": 46210 }, { "grad_norm": 0.21557143330574036, "learning_rate": 4.132414982517721e-05, "loss": 0.0125, "step": 46220 }, { "grad_norm": 0.24964219331741333, "learning_rate": 4.130379571542798e-05, "loss": 0.0179, "step": 46230 }, { "grad_norm": 0.201002299785614, "learning_rate": 4.12834430916222e-05, "loss": 0.0163, "step": 46240 }, { "grad_norm": 0.2780340015888214, "learning_rate": 4.126309195723763e-05, "loss": 0.0182, "step": 46250 }, { "grad_norm": 0.25840362906455994, "learning_rate": 4.124274231575168e-05, "loss": 0.0204, "step": 46260 }, { "grad_norm": 0.21408294141292572, "learning_rate": 4.122239417064154e-05, "loss": 0.0161, "step": 46270 }, { "grad_norm": 0.16084842383861542, "learning_rate": 4.1202047525384184e-05, "loss": 0.0167, "step": 46280 }, { "grad_norm": 0.2749235928058624, "learning_rate": 4.118170238345627e-05, "loss": 0.0161, "step": 46290 }, { "grad_norm": 0.21265654265880585, "learning_rate": 4.1161358748334256e-05, "loss": 0.0143, "step": 46300 }, { "grad_norm": 0.27436158061027527, "learning_rate": 4.1141016623494266e-05, "loss": 0.0173, "step": 46310 }, { "grad_norm": 0.2889970541000366, "learning_rate": 4.112067601241227e-05, "loss": 0.015, "step": 46320 }, { "grad_norm": 0.27228620648384094, "learning_rate": 4.110033691856387e-05, "loss": 0.0155, "step": 46330 }, { "grad_norm": 0.5159779787063599, "learning_rate": 4.107999934542451e-05, "loss": 0.0244, "step": 46340 }, { "grad_norm": 0.2350279986858368, "learning_rate": 4.105966329646928e-05, "loss": 0.0178, "step": 46350 }, { "grad_norm": 0.1788136065006256, "learning_rate": 4.103932877517308e-05, "loss": 0.013, "step": 46360 }, { "grad_norm": 0.23513934016227722, "learning_rate": 4.101899578501052e-05, "loss": 0.0128, "step": 46370 }, { "grad_norm": 0.280497670173645, "learning_rate": 4.099866432945595e-05, "loss": 0.0135, "step": 46380 }, { "grad_norm": 0.29465681314468384, "learning_rate": 4.097833441198346e-05, "loss": 0.0197, "step": 46390 }, { "grad_norm": 0.2554539144039154, "learning_rate": 4.095800603606685e-05, "loss": 0.0143, "step": 46400 }, { "grad_norm": 0.5575202107429504, "learning_rate": 4.093767920517975e-05, "loss": 0.0174, "step": 46410 }, { "grad_norm": 0.2977094352245331, "learning_rate": 4.091735392279539e-05, "loss": 0.0185, "step": 46420 }, { "grad_norm": 0.24493037164211273, "learning_rate": 4.089703019238685e-05, "loss": 0.0155, "step": 46430 }, { "grad_norm": 0.24630026519298553, "learning_rate": 4.0876708017426866e-05, "loss": 0.0148, "step": 46440 }, { "grad_norm": 0.2030494660139084, "learning_rate": 4.085638740138798e-05, "loss": 0.015, "step": 46450 }, { "grad_norm": 0.36018311977386475, "learning_rate": 4.08360683477424e-05, "loss": 0.0203, "step": 46460 }, { "grad_norm": 0.229298397898674, "learning_rate": 4.0815750859962085e-05, "loss": 0.0161, "step": 46470 }, { "grad_norm": 0.23410272598266602, "learning_rate": 4.079543494151879e-05, "loss": 0.0182, "step": 46480 }, { "grad_norm": 0.3168083429336548, "learning_rate": 4.07751205958839e-05, "loss": 0.016, "step": 46490 }, { "grad_norm": 0.33588534593582153, "learning_rate": 4.0754807826528615e-05, "loss": 0.0164, "step": 46500 }, { "grad_norm": 0.25980594754219055, "learning_rate": 4.073449663692379e-05, "loss": 0.0239, "step": 46510 }, { "grad_norm": 0.22321787476539612, "learning_rate": 4.07141870305401e-05, "loss": 0.0172, "step": 46520 }, { "grad_norm": 0.41784560680389404, "learning_rate": 4.0693879010847866e-05, "loss": 0.0318, "step": 46530 }, { "grad_norm": 0.22416174411773682, "learning_rate": 4.06735725813172e-05, "loss": 0.0174, "step": 46540 }, { "grad_norm": 0.32805442810058594, "learning_rate": 4.0653267745417903e-05, "loss": 0.0169, "step": 46550 }, { "grad_norm": 0.24576997756958008, "learning_rate": 4.063296450661949e-05, "loss": 0.0135, "step": 46560 }, { "grad_norm": 0.43668150901794434, "learning_rate": 4.061266286839128e-05, "loss": 0.0199, "step": 46570 }, { "grad_norm": 0.3258543908596039, "learning_rate": 4.0592362834202225e-05, "loss": 0.0175, "step": 46580 }, { "grad_norm": 0.3550901710987091, "learning_rate": 4.057206440752107e-05, "loss": 0.0164, "step": 46590 }, { "grad_norm": 0.1916816234588623, "learning_rate": 4.0551767591816245e-05, "loss": 0.0176, "step": 46600 }, { "grad_norm": 0.3321686387062073, "learning_rate": 4.0531472390555935e-05, "loss": 0.0213, "step": 46610 }, { "grad_norm": 0.2273704707622528, "learning_rate": 4.051117880720802e-05, "loss": 0.0138, "step": 46620 }, { "grad_norm": 0.14745372533798218, "learning_rate": 4.049088684524015e-05, "loss": 0.0186, "step": 46630 }, { "grad_norm": 0.22236600518226624, "learning_rate": 4.0470596508119636e-05, "loss": 0.0191, "step": 46640 }, { "grad_norm": 0.2562578022480011, "learning_rate": 4.0450307799313524e-05, "loss": 0.0236, "step": 46650 }, { "grad_norm": 0.21857066452503204, "learning_rate": 4.0430020722288656e-05, "loss": 0.0216, "step": 46660 }, { "grad_norm": 0.2147097885608673, "learning_rate": 4.040973528051148e-05, "loss": 0.0196, "step": 46670 }, { "grad_norm": 0.2982056438922882, "learning_rate": 4.038945147744827e-05, "loss": 0.0226, "step": 46680 }, { "grad_norm": 0.2028394341468811, "learning_rate": 4.0369169316564945e-05, "loss": 0.0169, "step": 46690 }, { "grad_norm": 0.2947790324687958, "learning_rate": 4.034888880132718e-05, "loss": 0.0151, "step": 46700 }, { "grad_norm": 0.41518551111221313, "learning_rate": 4.032860993520035e-05, "loss": 0.0182, "step": 46710 }, { "grad_norm": 0.334981232881546, "learning_rate": 4.030833272164959e-05, "loss": 0.0168, "step": 46720 }, { "grad_norm": 0.1763465702533722, "learning_rate": 4.0288057164139705e-05, "loss": 0.0164, "step": 46730 }, { "grad_norm": 0.15633250772953033, "learning_rate": 4.0267783266135205e-05, "loss": 0.0131, "step": 46740 }, { "grad_norm": 0.1882369965314865, "learning_rate": 4.024751103110039e-05, "loss": 0.0168, "step": 46750 }, { "grad_norm": 0.146014004945755, "learning_rate": 4.0227240462499176e-05, "loss": 0.019, "step": 46760 }, { "grad_norm": 0.24074965715408325, "learning_rate": 4.020697156379531e-05, "loss": 0.0174, "step": 46770 }, { "grad_norm": 0.275656521320343, "learning_rate": 4.018670433845215e-05, "loss": 0.0223, "step": 46780 }, { "grad_norm": 0.3052312135696411, "learning_rate": 4.016643878993284e-05, "loss": 0.0161, "step": 46790 }, { "grad_norm": 0.24603483080863953, "learning_rate": 4.014617492170017e-05, "loss": 0.0195, "step": 46800 }, { "grad_norm": 0.3131713271141052, "learning_rate": 4.0125912737216726e-05, "loss": 0.027, "step": 46810 }, { "grad_norm": 0.38375234603881836, "learning_rate": 4.0105652239944735e-05, "loss": 0.0257, "step": 46820 }, { "grad_norm": 0.32691648602485657, "learning_rate": 4.0085393433346144e-05, "loss": 0.0173, "step": 46830 }, { "grad_norm": 0.1888752281665802, "learning_rate": 4.006513632088268e-05, "loss": 0.0169, "step": 46840 }, { "grad_norm": 0.2749437689781189, "learning_rate": 4.004488090601567e-05, "loss": 0.0175, "step": 46850 }, { "grad_norm": 0.31173595786094666, "learning_rate": 4.002462719220626e-05, "loss": 0.0156, "step": 46860 }, { "grad_norm": 0.21442566812038422, "learning_rate": 4.000437518291522e-05, "loss": 0.0153, "step": 46870 }, { "grad_norm": 0.15899275243282318, "learning_rate": 3.9984124881603094e-05, "loss": 0.0162, "step": 46880 }, { "grad_norm": 0.1281324326992035, "learning_rate": 3.9963876291730086e-05, "loss": 0.0143, "step": 46890 }, { "grad_norm": 0.21952378749847412, "learning_rate": 3.994362941675614e-05, "loss": 0.018, "step": 46900 }, { "grad_norm": 0.290472149848938, "learning_rate": 3.992338426014088e-05, "loss": 0.0185, "step": 46910 }, { "grad_norm": 0.19411340355873108, "learning_rate": 3.9903140825343636e-05, "loss": 0.0166, "step": 46920 }, { "grad_norm": 0.23331399261951447, "learning_rate": 3.98828991158235e-05, "loss": 0.0111, "step": 46930 }, { "grad_norm": 0.2503957152366638, "learning_rate": 3.9862659135039185e-05, "loss": 0.0195, "step": 46940 }, { "grad_norm": 0.3392942547798157, "learning_rate": 3.984242088644918e-05, "loss": 0.0179, "step": 46950 }, { "grad_norm": 0.23026074469089508, "learning_rate": 3.9822184373511615e-05, "loss": 0.0143, "step": 46960 }, { "grad_norm": 0.25497332215309143, "learning_rate": 3.98019495996844e-05, "loss": 0.0186, "step": 46970 }, { "grad_norm": 0.31626614928245544, "learning_rate": 3.978171656842507e-05, "loss": 0.0165, "step": 46980 }, { "grad_norm": 0.2311558872461319, "learning_rate": 3.976148528319091e-05, "loss": 0.013, "step": 46990 }, { "grad_norm": 0.4203788638114929, "learning_rate": 3.974125574743888e-05, "loss": 0.0166, "step": 47000 }, { "grad_norm": 0.3147893249988556, "learning_rate": 3.9721027964625686e-05, "loss": 0.019, "step": 47010 }, { "grad_norm": 0.27848556637763977, "learning_rate": 3.9700801938207676e-05, "loss": 0.0188, "step": 47020 }, { "grad_norm": 0.30651044845581055, "learning_rate": 3.9680577671640916e-05, "loss": 0.0151, "step": 47030 }, { "grad_norm": 0.32194483280181885, "learning_rate": 3.966035516838121e-05, "loss": 0.0217, "step": 47040 }, { "grad_norm": 0.2171851247549057, "learning_rate": 3.9640134431884014e-05, "loss": 0.0123, "step": 47050 }, { "grad_norm": 0.2335238754749298, "learning_rate": 3.961991546560451e-05, "loss": 0.0137, "step": 47060 }, { "grad_norm": 0.2116672396659851, "learning_rate": 3.959969827299753e-05, "loss": 0.0199, "step": 47070 }, { "grad_norm": 0.2796604633331299, "learning_rate": 3.9579482857517684e-05, "loss": 0.0152, "step": 47080 }, { "grad_norm": 0.2123987078666687, "learning_rate": 3.955926922261921e-05, "loss": 0.0114, "step": 47090 }, { "grad_norm": 0.21251851320266724, "learning_rate": 3.9539057371756084e-05, "loss": 0.0155, "step": 47100 }, { "grad_norm": 0.27478575706481934, "learning_rate": 3.951884730838195e-05, "loss": 0.0137, "step": 47110 }, { "grad_norm": 0.18622691929340363, "learning_rate": 3.949863903595012e-05, "loss": 0.011, "step": 47120 }, { "grad_norm": 0.3036573827266693, "learning_rate": 3.947843255791369e-05, "loss": 0.011, "step": 47130 }, { "grad_norm": 0.11985266208648682, "learning_rate": 3.9458227877725364e-05, "loss": 0.0121, "step": 47140 }, { "grad_norm": 0.24912086129188538, "learning_rate": 3.943802499883758e-05, "loss": 0.0107, "step": 47150 }, { "grad_norm": 0.20656679570674896, "learning_rate": 3.9417823924702437e-05, "loss": 0.014, "step": 47160 }, { "grad_norm": 0.28041157126426697, "learning_rate": 3.939762465877178e-05, "loss": 0.0139, "step": 47170 }, { "grad_norm": 0.14828985929489136, "learning_rate": 3.937742720449708e-05, "loss": 0.0157, "step": 47180 }, { "grad_norm": 0.36911043524742126, "learning_rate": 3.9357231565329563e-05, "loss": 0.0135, "step": 47190 }, { "grad_norm": 0.14019185304641724, "learning_rate": 3.933703774472008e-05, "loss": 0.0136, "step": 47200 }, { "grad_norm": 0.17375408113002777, "learning_rate": 3.93168457461192e-05, "loss": 0.0135, "step": 47210 }, { "grad_norm": 0.212344691157341, "learning_rate": 3.9296655572977216e-05, "loss": 0.0158, "step": 47220 }, { "grad_norm": 0.3352063298225403, "learning_rate": 3.927646722874404e-05, "loss": 0.0175, "step": 47230 }, { "grad_norm": 0.17676430940628052, "learning_rate": 3.925628071686934e-05, "loss": 0.0135, "step": 47240 }, { "grad_norm": 0.20793631672859192, "learning_rate": 3.9236096040802415e-05, "loss": 0.02, "step": 47250 }, { "grad_norm": 0.3759891390800476, "learning_rate": 3.9215913203992294e-05, "loss": 0.0248, "step": 47260 }, { "grad_norm": 0.43471506237983704, "learning_rate": 3.9195732209887645e-05, "loss": 0.0148, "step": 47270 }, { "grad_norm": 0.21830090880393982, "learning_rate": 3.9175553061936875e-05, "loss": 0.015, "step": 47280 }, { "grad_norm": 0.2011135369539261, "learning_rate": 3.9155375763588045e-05, "loss": 0.0239, "step": 47290 }, { "grad_norm": 0.17650893330574036, "learning_rate": 3.913520031828889e-05, "loss": 0.0156, "step": 47300 }, { "grad_norm": 0.2546602487564087, "learning_rate": 3.911502672948685e-05, "loss": 0.0141, "step": 47310 }, { "grad_norm": 0.3132300078868866, "learning_rate": 3.9094855000629014e-05, "loss": 0.014, "step": 47320 }, { "grad_norm": 0.18118298053741455, "learning_rate": 3.907468513516223e-05, "loss": 0.0118, "step": 47330 }, { "grad_norm": 0.2300807535648346, "learning_rate": 3.905451713653294e-05, "loss": 0.0178, "step": 47340 }, { "grad_norm": 0.2888008952140808, "learning_rate": 3.903435100818731e-05, "loss": 0.0196, "step": 47350 }, { "grad_norm": 0.25029173493385315, "learning_rate": 3.901418675357117e-05, "loss": 0.018, "step": 47360 }, { "grad_norm": 0.25921517610549927, "learning_rate": 3.8994024376130075e-05, "loss": 0.0227, "step": 47370 }, { "grad_norm": 0.3505181074142456, "learning_rate": 3.8973863879309194e-05, "loss": 0.0182, "step": 47380 }, { "grad_norm": 0.3175927996635437, "learning_rate": 3.8953705266553394e-05, "loss": 0.011, "step": 47390 }, { "grad_norm": 0.20588617026805878, "learning_rate": 3.893354854130727e-05, "loss": 0.0167, "step": 47400 }, { "grad_norm": 0.33149200677871704, "learning_rate": 3.8913393707015006e-05, "loss": 0.0177, "step": 47410 }, { "grad_norm": 0.3603432774543762, "learning_rate": 3.889324076712056e-05, "loss": 0.0222, "step": 47420 }, { "grad_norm": 0.2771678566932678, "learning_rate": 3.8873089725067476e-05, "loss": 0.014, "step": 47430 }, { "grad_norm": 0.25882628560066223, "learning_rate": 3.885294058429905e-05, "loss": 0.0255, "step": 47440 }, { "grad_norm": 0.21416917443275452, "learning_rate": 3.8832793348258206e-05, "loss": 0.0183, "step": 47450 }, { "grad_norm": 0.28307846188545227, "learning_rate": 3.881264802038756e-05, "loss": 0.0272, "step": 47460 }, { "grad_norm": 0.26158809661865234, "learning_rate": 3.879250460412939e-05, "loss": 0.0216, "step": 47470 }, { "grad_norm": 0.22107334434986115, "learning_rate": 3.8772363102925644e-05, "loss": 0.0142, "step": 47480 }, { "grad_norm": 0.2636443078517914, "learning_rate": 3.875222352021798e-05, "loss": 0.0175, "step": 47490 }, { "grad_norm": 0.23368148505687714, "learning_rate": 3.8732085859447686e-05, "loss": 0.0166, "step": 47500 }, { "grad_norm": 0.29376766085624695, "learning_rate": 3.871195012405575e-05, "loss": 0.0156, "step": 47510 }, { "grad_norm": 0.2282097041606903, "learning_rate": 3.869181631748278e-05, "loss": 0.018, "step": 47520 }, { "grad_norm": 0.29442575573921204, "learning_rate": 3.867168444316915e-05, "loss": 0.0163, "step": 47530 }, { "grad_norm": 0.20518936216831207, "learning_rate": 3.865155450455481e-05, "loss": 0.0147, "step": 47540 }, { "grad_norm": 0.17073775827884674, "learning_rate": 3.8631426505079426e-05, "loss": 0.0178, "step": 47550 }, { "grad_norm": 0.3392283022403717, "learning_rate": 3.8611300448182304e-05, "loss": 0.0276, "step": 47560 }, { "grad_norm": 0.2686064541339874, "learning_rate": 3.859117633730248e-05, "loss": 0.0145, "step": 47570 }, { "grad_norm": 0.29455411434173584, "learning_rate": 3.857105417587858e-05, "loss": 0.0199, "step": 47580 }, { "grad_norm": 0.2903238832950592, "learning_rate": 3.855093396734894e-05, "loss": 0.0161, "step": 47590 }, { "grad_norm": 0.23372267186641693, "learning_rate": 3.8530815715151545e-05, "loss": 0.0281, "step": 47600 }, { "grad_norm": 0.3104911744594574, "learning_rate": 3.851069942272405e-05, "loss": 0.016, "step": 47610 }, { "grad_norm": 0.2766530513763428, "learning_rate": 3.849058509350382e-05, "loss": 0.0238, "step": 47620 }, { "grad_norm": 0.23412734270095825, "learning_rate": 3.8470472730927783e-05, "loss": 0.0235, "step": 47630 }, { "grad_norm": 0.22563274204730988, "learning_rate": 3.845036233843264e-05, "loss": 0.0131, "step": 47640 }, { "grad_norm": 0.292000949382782, "learning_rate": 3.843025391945469e-05, "loss": 0.0165, "step": 47650 }, { "grad_norm": 0.18995501101016998, "learning_rate": 3.841014747742992e-05, "loss": 0.0181, "step": 47660 }, { "grad_norm": 0.20631395280361176, "learning_rate": 3.839004301579397e-05, "loss": 0.0145, "step": 47670 }, { "grad_norm": 0.3072059750556946, "learning_rate": 3.8369940537982097e-05, "loss": 0.0221, "step": 47680 }, { "grad_norm": 0.20011252164840698, "learning_rate": 3.834984004742933e-05, "loss": 0.0169, "step": 47690 }, { "grad_norm": 0.22241611778736115, "learning_rate": 3.832974154757026e-05, "loss": 0.0145, "step": 47700 }, { "grad_norm": 0.31708481907844543, "learning_rate": 3.830964504183919e-05, "loss": 0.0177, "step": 47710 }, { "grad_norm": 0.25089889764785767, "learning_rate": 3.828955053367003e-05, "loss": 0.021, "step": 47720 }, { "grad_norm": 0.24986016750335693, "learning_rate": 3.826945802649642e-05, "loss": 0.0167, "step": 47730 }, { "grad_norm": 0.16709934175014496, "learning_rate": 3.824936752375159e-05, "loss": 0.0223, "step": 47740 }, { "grad_norm": 0.2691437602043152, "learning_rate": 3.822927902886848e-05, "loss": 0.0159, "step": 47750 }, { "grad_norm": 0.3038611114025116, "learning_rate": 3.8209192545279653e-05, "loss": 0.0153, "step": 47760 }, { "grad_norm": 0.1862792819738388, "learning_rate": 3.8189108076417326e-05, "loss": 0.0152, "step": 47770 }, { "grad_norm": 0.23206564784049988, "learning_rate": 3.816902562571342e-05, "loss": 0.0213, "step": 47780 }, { "grad_norm": 0.24457068741321564, "learning_rate": 3.814894519659944e-05, "loss": 0.015, "step": 47790 }, { "grad_norm": 0.22037553787231445, "learning_rate": 3.812886679250661e-05, "loss": 0.0156, "step": 47800 }, { "grad_norm": 0.2758572995662689, "learning_rate": 3.810879041686575e-05, "loss": 0.0137, "step": 47810 }, { "grad_norm": 0.46403399109840393, "learning_rate": 3.808871607310741e-05, "loss": 0.0239, "step": 47820 }, { "grad_norm": 0.36251601576805115, "learning_rate": 3.806864376466169e-05, "loss": 0.0153, "step": 47830 }, { "grad_norm": 0.17302192747592926, "learning_rate": 3.804857349495845e-05, "loss": 0.0122, "step": 47840 }, { "grad_norm": 0.384574830532074, "learning_rate": 3.802850526742713e-05, "loss": 0.0381, "step": 47850 }, { "grad_norm": 0.27064231038093567, "learning_rate": 3.800843908549683e-05, "loss": 0.0156, "step": 47860 }, { "grad_norm": 0.26480376720428467, "learning_rate": 3.7988374952596325e-05, "loss": 0.0152, "step": 47870 }, { "grad_norm": 0.23174767196178436, "learning_rate": 3.7968312872154e-05, "loss": 0.0164, "step": 47880 }, { "grad_norm": 0.29595401883125305, "learning_rate": 3.7948252847597965e-05, "loss": 0.0133, "step": 47890 }, { "grad_norm": 0.197785884141922, "learning_rate": 3.7928194882355885e-05, "loss": 0.0133, "step": 47900 }, { "grad_norm": 0.23094545304775238, "learning_rate": 3.790813897985515e-05, "loss": 0.0147, "step": 47910 }, { "grad_norm": 0.3707341253757477, "learning_rate": 3.7888085143522726e-05, "loss": 0.0174, "step": 47920 }, { "grad_norm": 0.3474085330963135, "learning_rate": 3.7868033376785314e-05, "loss": 0.0126, "step": 47930 }, { "grad_norm": 0.29023653268814087, "learning_rate": 3.784798368306919e-05, "loss": 0.0135, "step": 47940 }, { "grad_norm": 0.17428256571292877, "learning_rate": 3.782793606580029e-05, "loss": 0.012, "step": 47950 }, { "grad_norm": 0.2751086354255676, "learning_rate": 3.7807890528404205e-05, "loss": 0.0204, "step": 47960 }, { "grad_norm": 0.19179315865039825, "learning_rate": 3.778784707430616e-05, "loss": 0.0127, "step": 47970 }, { "grad_norm": 0.2597096860408783, "learning_rate": 3.776780570693107e-05, "loss": 0.0145, "step": 47980 }, { "grad_norm": 0.27120640873908997, "learning_rate": 3.774776642970342e-05, "loss": 0.0194, "step": 47990 }, { "grad_norm": 0.29228201508522034, "learning_rate": 3.77277292460474e-05, "loss": 0.0128, "step": 48000 }, { "grad_norm": 0.2285386472940445, "learning_rate": 3.770769415938678e-05, "loss": 0.0126, "step": 48010 }, { "grad_norm": 0.2304440289735794, "learning_rate": 3.768766117314506e-05, "loss": 0.0188, "step": 48020 }, { "grad_norm": 0.34626519680023193, "learning_rate": 3.76676302907453e-05, "loss": 0.0204, "step": 48030 }, { "grad_norm": 0.2850887179374695, "learning_rate": 3.764760151561021e-05, "loss": 0.017, "step": 48040 }, { "grad_norm": 0.24715082347393036, "learning_rate": 3.76275748511622e-05, "loss": 0.0144, "step": 48050 }, { "grad_norm": 0.2434479147195816, "learning_rate": 3.7607550300823255e-05, "loss": 0.0224, "step": 48060 }, { "grad_norm": 0.317842960357666, "learning_rate": 3.7587527868015044e-05, "loss": 0.0268, "step": 48070 }, { "grad_norm": 0.35163092613220215, "learning_rate": 3.756750755615881e-05, "loss": 0.0168, "step": 48080 }, { "grad_norm": 0.342352956533432, "learning_rate": 3.754748936867553e-05, "loss": 0.0207, "step": 48090 }, { "grad_norm": 0.29360783100128174, "learning_rate": 3.752747330898573e-05, "loss": 0.0132, "step": 48100 }, { "grad_norm": 0.2561569809913635, "learning_rate": 3.750745938050962e-05, "loss": 0.0168, "step": 48110 }, { "grad_norm": 0.27050355076789856, "learning_rate": 3.7487447586667025e-05, "loss": 0.0164, "step": 48120 }, { "grad_norm": 0.25752466917037964, "learning_rate": 3.7467437930877426e-05, "loss": 0.0117, "step": 48130 }, { "grad_norm": 0.14162860810756683, "learning_rate": 3.744743041655992e-05, "loss": 0.0149, "step": 48140 }, { "grad_norm": 0.1395616978406906, "learning_rate": 3.742742504713324e-05, "loss": 0.0162, "step": 48150 }, { "grad_norm": 0.13246721029281616, "learning_rate": 3.740742182601576e-05, "loss": 0.0145, "step": 48160 }, { "grad_norm": 0.13959874212741852, "learning_rate": 3.7387420756625464e-05, "loss": 0.0191, "step": 48170 }, { "grad_norm": 0.22279590368270874, "learning_rate": 3.736742184238002e-05, "loss": 0.0188, "step": 48180 }, { "grad_norm": 0.15846984088420868, "learning_rate": 3.7347425086696684e-05, "loss": 0.0141, "step": 48190 }, { "grad_norm": 0.22426870465278625, "learning_rate": 3.732743049299235e-05, "loss": 0.0159, "step": 48200 }, { "grad_norm": 0.21402029693126678, "learning_rate": 3.730743806468354e-05, "loss": 0.0133, "step": 48210 }, { "grad_norm": 0.30207589268684387, "learning_rate": 3.7287447805186436e-05, "loss": 0.0131, "step": 48220 }, { "grad_norm": 0.2285110205411911, "learning_rate": 3.726745971791682e-05, "loss": 0.0182, "step": 48230 }, { "grad_norm": 0.3514857590198517, "learning_rate": 3.724747380629008e-05, "loss": 0.0239, "step": 48240 }, { "grad_norm": 0.34483373165130615, "learning_rate": 3.72274900737213e-05, "loss": 0.0211, "step": 48250 }, { "grad_norm": 0.17621217668056488, "learning_rate": 3.7207508523625123e-05, "loss": 0.0161, "step": 48260 }, { "grad_norm": 0.4252382218837738, "learning_rate": 3.718752915941588e-05, "loss": 0.0196, "step": 48270 }, { "grad_norm": 0.2732878625392914, "learning_rate": 3.7167551984507464e-05, "loss": 0.0126, "step": 48280 }, { "grad_norm": 0.25367507338523865, "learning_rate": 3.714757700231346e-05, "loss": 0.0166, "step": 48290 }, { "grad_norm": 0.1940729171037674, "learning_rate": 3.712760421624703e-05, "loss": 0.013, "step": 48300 }, { "grad_norm": 0.24822621047496796, "learning_rate": 3.710763362972099e-05, "loss": 0.0139, "step": 48310 }, { "grad_norm": 0.14422833919525146, "learning_rate": 3.708766524614774e-05, "loss": 0.0179, "step": 48320 }, { "grad_norm": 0.1595412790775299, "learning_rate": 3.7067699068939335e-05, "loss": 0.0109, "step": 48330 }, { "grad_norm": 0.33751723170280457, "learning_rate": 3.704773510150748e-05, "loss": 0.0163, "step": 48340 }, { "grad_norm": 0.3230511546134949, "learning_rate": 3.702777334726344e-05, "loss": 0.0166, "step": 48350 }, { "grad_norm": 0.20979809761047363, "learning_rate": 3.7007813809618164e-05, "loss": 0.0142, "step": 48360 }, { "grad_norm": 0.296894371509552, "learning_rate": 3.6987856491982145e-05, "loss": 0.0233, "step": 48370 }, { "grad_norm": 0.22393928468227386, "learning_rate": 3.696790139776558e-05, "loss": 0.0209, "step": 48380 }, { "grad_norm": 0.2265842705965042, "learning_rate": 3.6947948530378235e-05, "loss": 0.0191, "step": 48390 }, { "grad_norm": 0.23231863975524902, "learning_rate": 3.6927997893229516e-05, "loss": 0.0133, "step": 48400 }, { "grad_norm": 0.2545551657676697, "learning_rate": 3.690804948972845e-05, "loss": 0.0135, "step": 48410 }, { "grad_norm": 0.15391407907009125, "learning_rate": 3.688810332328363e-05, "loss": 0.0127, "step": 48420 }, { "grad_norm": 0.2555018365383148, "learning_rate": 3.686815939730336e-05, "loss": 0.0163, "step": 48430 }, { "grad_norm": 0.265703022480011, "learning_rate": 3.684821771519548e-05, "loss": 0.0188, "step": 48440 }, { "grad_norm": 0.2865939736366272, "learning_rate": 3.68282782803675e-05, "loss": 0.0128, "step": 48450 }, { "grad_norm": 0.20523585379123688, "learning_rate": 3.6808341096226504e-05, "loss": 0.0239, "step": 48460 }, { "grad_norm": 0.3631777763366699, "learning_rate": 3.678840616617924e-05, "loss": 0.0192, "step": 48470 }, { "grad_norm": 0.3773652911186218, "learning_rate": 3.6768473493632e-05, "loss": 0.0177, "step": 48480 }, { "grad_norm": 0.21116338670253754, "learning_rate": 3.6748543081990783e-05, "loss": 0.0139, "step": 48490 }, { "grad_norm": 0.31360018253326416, "learning_rate": 3.672861493466112e-05, "loss": 0.0163, "step": 48500 }, { "grad_norm": 0.21532395482063293, "learning_rate": 3.670868905504818e-05, "loss": 0.0136, "step": 48510 }, { "grad_norm": 0.33220210671424866, "learning_rate": 3.6688765446556784e-05, "loss": 0.0183, "step": 48520 }, { "grad_norm": 0.27368953824043274, "learning_rate": 3.6668844112591276e-05, "loss": 0.0186, "step": 48530 }, { "grad_norm": 0.24052152037620544, "learning_rate": 3.664892505655573e-05, "loss": 0.02, "step": 48540 }, { "grad_norm": 0.28687217831611633, "learning_rate": 3.662900828185373e-05, "loss": 0.021, "step": 48550 }, { "grad_norm": 0.17630882561206818, "learning_rate": 3.6609093791888516e-05, "loss": 0.0154, "step": 48560 }, { "grad_norm": 0.21722640097141266, "learning_rate": 3.658918159006292e-05, "loss": 0.0166, "step": 48570 }, { "grad_norm": 0.1738342046737671, "learning_rate": 3.656927167977942e-05, "loss": 0.0177, "step": 48580 }, { "grad_norm": 0.15938693284988403, "learning_rate": 3.654936406444006e-05, "loss": 0.0168, "step": 48590 }, { "grad_norm": 0.1917891651391983, "learning_rate": 3.65294587474465e-05, "loss": 0.0148, "step": 48600 }, { "grad_norm": 0.23097747564315796, "learning_rate": 3.650955573220002e-05, "loss": 0.0438, "step": 48610 }, { "grad_norm": 0.27311018109321594, "learning_rate": 3.648965502210149e-05, "loss": 0.015, "step": 48620 }, { "grad_norm": 0.22888539731502533, "learning_rate": 3.646975662055142e-05, "loss": 0.0137, "step": 48630 }, { "grad_norm": 0.3345423638820648, "learning_rate": 3.644986053094987e-05, "loss": 0.0161, "step": 48640 }, { "grad_norm": 0.30579185485839844, "learning_rate": 3.642996675669659e-05, "loss": 0.0154, "step": 48650 }, { "grad_norm": 0.2322729378938675, "learning_rate": 3.641007530119083e-05, "loss": 0.0169, "step": 48660 }, { "grad_norm": 0.32217106223106384, "learning_rate": 3.639018616783153e-05, "loss": 0.019, "step": 48670 }, { "grad_norm": 0.2091611921787262, "learning_rate": 3.637029936001719e-05, "loss": 0.0143, "step": 48680 }, { "grad_norm": 0.3387294411659241, "learning_rate": 3.6350414881145886e-05, "loss": 0.0188, "step": 48690 }, { "grad_norm": 0.19918040931224823, "learning_rate": 3.6330532734615386e-05, "loss": 0.0144, "step": 48700 }, { "grad_norm": 0.24431374669075012, "learning_rate": 3.6310652923822975e-05, "loss": 0.0207, "step": 48710 }, { "grad_norm": 0.22636613249778748, "learning_rate": 3.629077545216558e-05, "loss": 0.0175, "step": 48720 }, { "grad_norm": 0.19651339948177338, "learning_rate": 3.627090032303969e-05, "loss": 0.0233, "step": 48730 }, { "grad_norm": 0.2765634059906006, "learning_rate": 3.625102753984146e-05, "loss": 0.0212, "step": 48740 }, { "grad_norm": 0.4089859426021576, "learning_rate": 3.623115710596659e-05, "loss": 0.0155, "step": 48750 }, { "grad_norm": 0.450339674949646, "learning_rate": 3.6211289024810395e-05, "loss": 0.0177, "step": 48760 }, { "grad_norm": 0.3231160342693329, "learning_rate": 3.619142329976777e-05, "loss": 0.0137, "step": 48770 }, { "grad_norm": 0.1766941100358963, "learning_rate": 3.6171559934233247e-05, "loss": 0.0129, "step": 48780 }, { "grad_norm": 0.13748577237129211, "learning_rate": 3.615169893160093e-05, "loss": 0.0151, "step": 48790 }, { "grad_norm": 0.1051752120256424, "learning_rate": 3.61318402952645e-05, "loss": 0.0131, "step": 48800 }, { "grad_norm": 0.4110204875469208, "learning_rate": 3.6111984028617285e-05, "loss": 0.019, "step": 48810 }, { "grad_norm": 0.2468067705631256, "learning_rate": 3.6092130135052134e-05, "loss": 0.0183, "step": 48820 }, { "grad_norm": 0.25539901852607727, "learning_rate": 3.6072278617961584e-05, "loss": 0.0157, "step": 48830 }, { "grad_norm": 0.34202349185943604, "learning_rate": 3.605242948073767e-05, "loss": 0.0165, "step": 48840 }, { "grad_norm": 0.2690732181072235, "learning_rate": 3.603258272677212e-05, "loss": 0.0175, "step": 48850 }, { "grad_norm": 0.31209737062454224, "learning_rate": 3.601273835945616e-05, "loss": 0.0171, "step": 48860 }, { "grad_norm": 0.1555311232805252, "learning_rate": 3.5992896382180664e-05, "loss": 0.016, "step": 48870 }, { "grad_norm": 0.3631940186023712, "learning_rate": 3.597305679833609e-05, "loss": 0.022, "step": 48880 }, { "grad_norm": 0.260930597782135, "learning_rate": 3.595321961131245e-05, "loss": 0.0157, "step": 48890 }, { "grad_norm": 0.24890922009944916, "learning_rate": 3.593338482449942e-05, "loss": 0.0244, "step": 48900 }, { "grad_norm": 0.3148301839828491, "learning_rate": 3.591355244128618e-05, "loss": 0.0278, "step": 48910 }, { "grad_norm": 0.2943148612976074, "learning_rate": 3.589372246506158e-05, "loss": 0.0198, "step": 48920 }, { "grad_norm": 0.16671158373355865, "learning_rate": 3.5873894899213984e-05, "loss": 0.0158, "step": 48930 }, { "grad_norm": 0.21210159361362457, "learning_rate": 3.5854069747131416e-05, "loss": 0.0166, "step": 48940 }, { "grad_norm": 0.30588534474372864, "learning_rate": 3.583424701220143e-05, "loss": 0.0197, "step": 48950 }, { "grad_norm": 0.24866269528865814, "learning_rate": 3.581442669781121e-05, "loss": 0.0168, "step": 48960 }, { "grad_norm": 0.2651105225086212, "learning_rate": 3.579460880734749e-05, "loss": 0.0178, "step": 48970 }, { "grad_norm": 0.23731274902820587, "learning_rate": 3.577479334419657e-05, "loss": 0.0156, "step": 48980 }, { "grad_norm": 0.29532748460769653, "learning_rate": 3.575498031174444e-05, "loss": 0.0195, "step": 48990 }, { "grad_norm": 0.2075706273317337, "learning_rate": 3.573516971337657e-05, "loss": 0.015, "step": 49000 }, { "grad_norm": 0.2331879436969757, "learning_rate": 3.5715361552478046e-05, "loss": 0.0197, "step": 49010 }, { "grad_norm": 0.329338401556015, "learning_rate": 3.5695555832433536e-05, "loss": 0.0132, "step": 49020 }, { "grad_norm": 0.3842356503009796, "learning_rate": 3.5675752556627325e-05, "loss": 0.0137, "step": 49030 }, { "grad_norm": 0.2581861615180969, "learning_rate": 3.565595172844322e-05, "loss": 0.0173, "step": 49040 }, { "grad_norm": 0.2178088128566742, "learning_rate": 3.5636153351264666e-05, "loss": 0.0169, "step": 49050 }, { "grad_norm": 0.3429992198944092, "learning_rate": 3.5616357428474655e-05, "loss": 0.02, "step": 49060 }, { "grad_norm": 0.25666722655296326, "learning_rate": 3.559656396345575e-05, "loss": 0.0198, "step": 49070 }, { "grad_norm": 0.23324823379516602, "learning_rate": 3.5576772959590146e-05, "loss": 0.0185, "step": 49080 }, { "grad_norm": 0.8581755757331848, "learning_rate": 3.5556984420259545e-05, "loss": 0.0191, "step": 49090 }, { "grad_norm": 0.31145524978637695, "learning_rate": 3.5537198348845305e-05, "loss": 0.0202, "step": 49100 }, { "grad_norm": 0.19824209809303284, "learning_rate": 3.551741474872831e-05, "loss": 0.016, "step": 49110 }, { "grad_norm": 0.2686786651611328, "learning_rate": 3.549763362328903e-05, "loss": 0.0203, "step": 49120 }, { "grad_norm": 0.23350656032562256, "learning_rate": 3.5477854975907515e-05, "loss": 0.0146, "step": 49130 }, { "grad_norm": 0.16646456718444824, "learning_rate": 3.5458078809963416e-05, "loss": 0.0199, "step": 49140 }, { "grad_norm": 0.2948823571205139, "learning_rate": 3.543830512883594e-05, "loss": 0.0165, "step": 49150 }, { "grad_norm": 0.21818526089191437, "learning_rate": 3.5418533935903824e-05, "loss": 0.0174, "step": 49160 }, { "grad_norm": 0.37298113107681274, "learning_rate": 3.539876523454547e-05, "loss": 0.0187, "step": 49170 }, { "grad_norm": 0.33826249837875366, "learning_rate": 3.537899902813878e-05, "loss": 0.0151, "step": 49180 }, { "grad_norm": 0.2487943470478058, "learning_rate": 3.5359235320061293e-05, "loss": 0.0182, "step": 49190 }, { "grad_norm": 0.21464593708515167, "learning_rate": 3.533947411369003e-05, "loss": 0.0131, "step": 49200 }, { "grad_norm": 0.1514582484960556, "learning_rate": 3.53197154124017e-05, "loss": 0.0164, "step": 49210 }, { "grad_norm": 0.2728247046470642, "learning_rate": 3.52999592195725e-05, "loss": 0.017, "step": 49220 }, { "grad_norm": 0.24923072755336761, "learning_rate": 3.5280205538578224e-05, "loss": 0.015, "step": 49230 }, { "grad_norm": 0.31642481684684753, "learning_rate": 3.5260454372794236e-05, "loss": 0.0197, "step": 49240 }, { "grad_norm": 0.1969045102596283, "learning_rate": 3.524070572559545e-05, "loss": 0.017, "step": 49250 }, { "grad_norm": 0.3170184791088104, "learning_rate": 3.5220959600356395e-05, "loss": 0.0173, "step": 49260 }, { "grad_norm": 0.17035865783691406, "learning_rate": 3.5201216000451145e-05, "loss": 0.0128, "step": 49270 }, { "grad_norm": 0.24047450721263885, "learning_rate": 3.5181474929253335e-05, "loss": 0.0143, "step": 49280 }, { "grad_norm": 0.2770829498767853, "learning_rate": 3.516173639013615e-05, "loss": 0.0198, "step": 49290 }, { "grad_norm": 0.3316069543361664, "learning_rate": 3.5142000386472406e-05, "loss": 0.0163, "step": 49300 }, { "grad_norm": 0.25186413526535034, "learning_rate": 3.5122266921634427e-05, "loss": 0.0158, "step": 49310 }, { "grad_norm": 0.2929663360118866, "learning_rate": 3.510253599899413e-05, "loss": 0.0159, "step": 49320 }, { "grad_norm": 0.3031275272369385, "learning_rate": 3.5082807621922965e-05, "loss": 0.0142, "step": 49330 }, { "grad_norm": 0.4149836301803589, "learning_rate": 3.506308179379201e-05, "loss": 0.0211, "step": 49340 }, { "grad_norm": 0.1844155639410019, "learning_rate": 3.5043358517971844e-05, "loss": 0.0141, "step": 49350 }, { "grad_norm": 0.17891106009483337, "learning_rate": 3.502363779783264e-05, "loss": 0.0179, "step": 49360 }, { "grad_norm": 0.2007402628660202, "learning_rate": 3.500391963674415e-05, "loss": 0.0142, "step": 49370 }, { "grad_norm": 0.31819233298301697, "learning_rate": 3.4984204038075615e-05, "loss": 0.0191, "step": 49380 }, { "grad_norm": 0.19822604954242706, "learning_rate": 3.496449100519595e-05, "loss": 0.0193, "step": 49390 }, { "grad_norm": 0.25039729475975037, "learning_rate": 3.494478054147354e-05, "loss": 0.016, "step": 49400 }, { "grad_norm": 0.29273733496665955, "learning_rate": 3.4925072650276395e-05, "loss": 0.0157, "step": 49410 }, { "grad_norm": 0.2424522191286087, "learning_rate": 3.4905367334972016e-05, "loss": 0.0152, "step": 49420 }, { "grad_norm": 0.25280818343162537, "learning_rate": 3.488566459892752e-05, "loss": 0.0183, "step": 49430 }, { "grad_norm": 0.25043854117393494, "learning_rate": 3.4865964445509585e-05, "loss": 0.0174, "step": 49440 }, { "grad_norm": 0.2198641002178192, "learning_rate": 3.484626687808438e-05, "loss": 0.0175, "step": 49450 }, { "grad_norm": 0.24689145386219025, "learning_rate": 3.4826571900017735e-05, "loss": 0.0125, "step": 49460 }, { "grad_norm": 0.2245585024356842, "learning_rate": 3.480687951467495e-05, "loss": 0.013, "step": 49470 }, { "grad_norm": 0.21183127164840698, "learning_rate": 3.4787189725420925e-05, "loss": 0.0133, "step": 49480 }, { "grad_norm": 0.21706844866275787, "learning_rate": 3.4767502535620086e-05, "loss": 0.0149, "step": 49490 }, { "grad_norm": 0.2741633653640747, "learning_rate": 3.474781794863648e-05, "loss": 0.0253, "step": 49500 }, { "grad_norm": 0.37087351083755493, "learning_rate": 3.472813596783363e-05, "loss": 0.0241, "step": 49510 }, { "grad_norm": 0.2967221736907959, "learning_rate": 3.470845659657466e-05, "loss": 0.0188, "step": 49520 }, { "grad_norm": 0.3490864038467407, "learning_rate": 3.468877983822223e-05, "loss": 0.0165, "step": 49530 }, { "grad_norm": 0.273812860250473, "learning_rate": 3.466910569613855e-05, "loss": 0.0125, "step": 49540 }, { "grad_norm": 0.30951714515686035, "learning_rate": 3.464943417368542e-05, "loss": 0.0156, "step": 49550 }, { "grad_norm": 0.24242570996284485, "learning_rate": 3.462976527422415e-05, "loss": 0.0121, "step": 49560 }, { "grad_norm": 0.3104992210865021, "learning_rate": 3.461009900111562e-05, "loss": 0.0176, "step": 49570 }, { "grad_norm": 0.28567764163017273, "learning_rate": 3.459043535772023e-05, "loss": 0.015, "step": 49580 }, { "grad_norm": 0.21894033253192902, "learning_rate": 3.4570774347398014e-05, "loss": 0.0151, "step": 49590 }, { "grad_norm": 0.21269278228282928, "learning_rate": 3.4551115973508454e-05, "loss": 0.0134, "step": 49600 }, { "grad_norm": 0.29417237639427185, "learning_rate": 3.453146023941066e-05, "loss": 0.0184, "step": 49610 }, { "grad_norm": 0.26449403166770935, "learning_rate": 3.451180714846325e-05, "loss": 0.0165, "step": 49620 }, { "grad_norm": 0.20918413996696472, "learning_rate": 3.449215670402438e-05, "loss": 0.0151, "step": 49630 }, { "grad_norm": 0.3286256194114685, "learning_rate": 3.447250890945181e-05, "loss": 0.0137, "step": 49640 }, { "grad_norm": 0.19322124123573303, "learning_rate": 3.4452863768102754e-05, "loss": 0.0163, "step": 49650 }, { "grad_norm": 0.28274309635162354, "learning_rate": 3.443322128333409e-05, "loss": 0.0177, "step": 49660 }, { "grad_norm": 0.510622501373291, "learning_rate": 3.441358145850215e-05, "loss": 0.0207, "step": 49670 }, { "grad_norm": 0.23270530998706818, "learning_rate": 3.439394429696286e-05, "loss": 0.0191, "step": 49680 }, { "grad_norm": 0.21914200484752655, "learning_rate": 3.4374309802071644e-05, "loss": 0.0143, "step": 49690 }, { "grad_norm": 0.2621558606624603, "learning_rate": 3.435467797718353e-05, "loss": 0.0134, "step": 49700 }, { "grad_norm": 0.15555953979492188, "learning_rate": 3.433504882565306e-05, "loss": 0.0123, "step": 49710 }, { "grad_norm": 0.17249834537506104, "learning_rate": 3.43154223508343e-05, "loss": 0.0162, "step": 49720 }, { "grad_norm": 0.3250235617160797, "learning_rate": 3.429579855608089e-05, "loss": 0.0139, "step": 49730 }, { "grad_norm": 0.2698376476764679, "learning_rate": 3.427617744474597e-05, "loss": 0.0126, "step": 49740 }, { "grad_norm": 0.409878671169281, "learning_rate": 3.425655902018231e-05, "loss": 0.0167, "step": 49750 }, { "grad_norm": 0.19980661571025848, "learning_rate": 3.423694328574211e-05, "loss": 0.0118, "step": 49760 }, { "grad_norm": 0.48334747552871704, "learning_rate": 3.42173302447772e-05, "loss": 0.0155, "step": 49770 }, { "grad_norm": 0.5289807915687561, "learning_rate": 3.419771990063886e-05, "loss": 0.015, "step": 49780 }, { "grad_norm": 0.29095587134361267, "learning_rate": 3.417811225667803e-05, "loss": 0.0151, "step": 49790 }, { "grad_norm": 0.2504155933856964, "learning_rate": 3.415850731624508e-05, "loss": 0.0152, "step": 49800 }, { "grad_norm": 0.3122783601284027, "learning_rate": 3.4138905082689945e-05, "loss": 0.02, "step": 49810 }, { "grad_norm": 0.11111918836832047, "learning_rate": 3.4119305559362145e-05, "loss": 0.0158, "step": 49820 }, { "grad_norm": 0.20014536380767822, "learning_rate": 3.4099708749610684e-05, "loss": 0.0175, "step": 49830 }, { "grad_norm": 0.33981239795684814, "learning_rate": 3.408011465678413e-05, "loss": 0.0145, "step": 49840 }, { "grad_norm": 0.33154961466789246, "learning_rate": 3.406052328423055e-05, "loss": 0.0132, "step": 49850 }, { "grad_norm": 0.2176208347082138, "learning_rate": 3.4040934635297615e-05, "loss": 0.0171, "step": 49860 }, { "grad_norm": 0.1861436367034912, "learning_rate": 3.4021348713332466e-05, "loss": 0.012, "step": 49870 }, { "grad_norm": 0.17692741751670837, "learning_rate": 3.4001765521681807e-05, "loss": 0.0124, "step": 49880 }, { "grad_norm": 0.19603800773620605, "learning_rate": 3.398218506369188e-05, "loss": 0.0223, "step": 49890 }, { "grad_norm": 0.1483350545167923, "learning_rate": 3.3962607342708404e-05, "loss": 0.0167, "step": 49900 }, { "grad_norm": 0.3504995107650757, "learning_rate": 3.394303236207673e-05, "loss": 0.0113, "step": 49910 }, { "grad_norm": 0.23145657777786255, "learning_rate": 3.392346012514166e-05, "loss": 0.0146, "step": 49920 }, { "grad_norm": 0.27880460023880005, "learning_rate": 3.390389063524757e-05, "loss": 0.011, "step": 49930 }, { "grad_norm": 0.1871163696050644, "learning_rate": 3.3884323895738324e-05, "loss": 0.015, "step": 49940 }, { "grad_norm": 0.3716733753681183, "learning_rate": 3.386475990995738e-05, "loss": 0.0211, "step": 49950 }, { "grad_norm": 0.195831760764122, "learning_rate": 3.384519868124765e-05, "loss": 0.012, "step": 49960 }, { "grad_norm": 0.12666752934455872, "learning_rate": 3.3825640212951645e-05, "loss": 0.0116, "step": 49970 }, { "grad_norm": 0.33948907256126404, "learning_rate": 3.380608450841134e-05, "loss": 0.0152, "step": 49980 }, { "grad_norm": 0.3022748827934265, "learning_rate": 3.3786531570968305e-05, "loss": 0.0115, "step": 49990 }, { "grad_norm": 0.18867740035057068, "learning_rate": 3.3766981403963584e-05, "loss": 0.0157, "step": 50000 }, { "grad_norm": 0.30206799507141113, "learning_rate": 3.374743401073775e-05, "loss": 0.0114, "step": 50010 }, { "grad_norm": 0.23813050985336304, "learning_rate": 3.372788939463095e-05, "loss": 0.0118, "step": 50020 }, { "grad_norm": 0.3872247338294983, "learning_rate": 3.370834755898281e-05, "loss": 0.0168, "step": 50030 }, { "grad_norm": 0.31215792894363403, "learning_rate": 3.3688808507132493e-05, "loss": 0.0198, "step": 50040 }, { "grad_norm": 0.4031508266925812, "learning_rate": 3.3669272242418685e-05, "loss": 0.0166, "step": 50050 }, { "grad_norm": 0.24003538489341736, "learning_rate": 3.364973876817961e-05, "loss": 0.0149, "step": 50060 }, { "grad_norm": 0.3708912432193756, "learning_rate": 3.363020808775299e-05, "loss": 0.0125, "step": 50070 }, { "grad_norm": 0.207148939371109, "learning_rate": 3.361068020447611e-05, "loss": 0.0133, "step": 50080 }, { "grad_norm": 0.4186270833015442, "learning_rate": 3.3591155121685724e-05, "loss": 0.015, "step": 50090 }, { "grad_norm": 0.26735761761665344, "learning_rate": 3.357163284271814e-05, "loss": 0.0158, "step": 50100 }, { "grad_norm": 0.1550597995519638, "learning_rate": 3.355211337090919e-05, "loss": 0.0158, "step": 50110 }, { "grad_norm": 0.17900793254375458, "learning_rate": 3.353259670959421e-05, "loss": 0.0158, "step": 50120 }, { "grad_norm": 0.13080373406410217, "learning_rate": 3.351308286210808e-05, "loss": 0.0145, "step": 50130 }, { "grad_norm": 0.24206002056598663, "learning_rate": 3.3493571831785156e-05, "loss": 0.0111, "step": 50140 }, { "grad_norm": 0.13708698749542236, "learning_rate": 3.347406362195936e-05, "loss": 0.0162, "step": 50150 }, { "grad_norm": 0.1838447004556656, "learning_rate": 3.345455823596411e-05, "loss": 0.0154, "step": 50160 }, { "grad_norm": 0.18411888182163239, "learning_rate": 3.3435055677132346e-05, "loss": 0.0128, "step": 50170 }, { "grad_norm": 0.32719647884368896, "learning_rate": 3.3415555948796505e-05, "loss": 0.0145, "step": 50180 }, { "grad_norm": 0.2797486484050751, "learning_rate": 3.3396059054288556e-05, "loss": 0.0147, "step": 50190 }, { "grad_norm": 0.2255845069885254, "learning_rate": 3.3376564996940015e-05, "loss": 0.0215, "step": 50200 }, { "grad_norm": 0.15952421724796295, "learning_rate": 3.3357073780081836e-05, "loss": 0.0143, "step": 50210 }, { "grad_norm": 0.24505864083766937, "learning_rate": 3.333758540704459e-05, "loss": 0.0129, "step": 50220 }, { "grad_norm": 0.18509447574615479, "learning_rate": 3.331809988115827e-05, "loss": 0.0125, "step": 50230 }, { "grad_norm": 0.23411962389945984, "learning_rate": 3.329861720575244e-05, "loss": 0.0127, "step": 50240 }, { "grad_norm": 0.18478044867515564, "learning_rate": 3.3279137384156126e-05, "loss": 0.0152, "step": 50250 }, { "grad_norm": 0.2355683594942093, "learning_rate": 3.3259660419697934e-05, "loss": 0.0168, "step": 50260 }, { "grad_norm": 0.2021569460630417, "learning_rate": 3.3240186315705926e-05, "loss": 0.0186, "step": 50270 }, { "grad_norm": 0.22974540293216705, "learning_rate": 3.322071507550769e-05, "loss": 0.0257, "step": 50280 }, { "grad_norm": 0.2652594745159149, "learning_rate": 3.320124670243033e-05, "loss": 0.0166, "step": 50290 }, { "grad_norm": 0.25884902477264404, "learning_rate": 3.318178119980045e-05, "loss": 0.0142, "step": 50300 }, { "grad_norm": 0.5741244554519653, "learning_rate": 3.316231857094421e-05, "loss": 0.0148, "step": 50310 }, { "grad_norm": 0.23264852166175842, "learning_rate": 3.31428588191872e-05, "loss": 0.0186, "step": 50320 }, { "grad_norm": 0.27000853419303894, "learning_rate": 3.312340194785458e-05, "loss": 0.0129, "step": 50330 }, { "grad_norm": 0.37165966629981995, "learning_rate": 3.3103947960270975e-05, "loss": 0.0142, "step": 50340 }, { "grad_norm": 0.22650901973247528, "learning_rate": 3.308449685976058e-05, "loss": 0.0117, "step": 50350 }, { "grad_norm": 0.3618364930152893, "learning_rate": 3.3065048649647024e-05, "loss": 0.0135, "step": 50360 }, { "grad_norm": 0.3915880024433136, "learning_rate": 3.304560333325348e-05, "loss": 0.013, "step": 50370 }, { "grad_norm": 0.2520025968551636, "learning_rate": 3.3026160913902624e-05, "loss": 0.0158, "step": 50380 }, { "grad_norm": 0.39428573846817017, "learning_rate": 3.300672139491662e-05, "loss": 0.0201, "step": 50390 }, { "grad_norm": 0.199285089969635, "learning_rate": 3.298728477961717e-05, "loss": 0.0145, "step": 50400 }, { "grad_norm": 0.16465748846530914, "learning_rate": 3.296785107132544e-05, "loss": 0.0165, "step": 50410 }, { "grad_norm": 0.2636229991912842, "learning_rate": 3.2948420273362156e-05, "loss": 0.0144, "step": 50420 }, { "grad_norm": 0.28603532910346985, "learning_rate": 3.292899238904747e-05, "loss": 0.0142, "step": 50430 }, { "grad_norm": 0.19977693259716034, "learning_rate": 3.29095674217011e-05, "loss": 0.011, "step": 50440 }, { "grad_norm": 0.16401901841163635, "learning_rate": 3.289014537464224e-05, "loss": 0.0177, "step": 50450 }, { "grad_norm": 0.16455796360969543, "learning_rate": 3.287072625118955e-05, "loss": 0.0179, "step": 50460 }, { "grad_norm": 0.18514850735664368, "learning_rate": 3.285131005466129e-05, "loss": 0.0131, "step": 50470 }, { "grad_norm": 0.2153513878583908, "learning_rate": 3.2831896788375105e-05, "loss": 0.0108, "step": 50480 }, { "grad_norm": 0.24750231206417084, "learning_rate": 3.281248645564822e-05, "loss": 0.0156, "step": 50490 }, { "grad_norm": 0.229548841714859, "learning_rate": 3.2793079059797306e-05, "loss": 0.0215, "step": 50500 }, { "grad_norm": 0.24660411477088928, "learning_rate": 3.277367460413859e-05, "loss": 0.0185, "step": 50510 }, { "grad_norm": 0.17683185636997223, "learning_rate": 3.275427309198773e-05, "loss": 0.0163, "step": 50520 }, { "grad_norm": 0.21928778290748596, "learning_rate": 3.273487452665993e-05, "loss": 0.0146, "step": 50530 }, { "grad_norm": 0.23449960350990295, "learning_rate": 3.271547891146986e-05, "loss": 0.0193, "step": 50540 }, { "grad_norm": 0.2723940908908844, "learning_rate": 3.269608624973173e-05, "loss": 0.0122, "step": 50550 }, { "grad_norm": 0.1655828058719635, "learning_rate": 3.26766965447592e-05, "loss": 0.0137, "step": 50560 }, { "grad_norm": 0.2032012939453125, "learning_rate": 3.2657309799865424e-05, "loss": 0.0199, "step": 50570 }, { "grad_norm": 0.1820051372051239, "learning_rate": 3.2637926018363084e-05, "loss": 0.0124, "step": 50580 }, { "grad_norm": 0.3804347515106201, "learning_rate": 3.261854520356432e-05, "loss": 0.0147, "step": 50590 }, { "grad_norm": 0.16458287835121155, "learning_rate": 3.2599167358780816e-05, "loss": 0.0135, "step": 50600 }, { "grad_norm": 0.3435417115688324, "learning_rate": 3.2579792487323676e-05, "loss": 0.0121, "step": 50610 }, { "grad_norm": 0.224808469414711, "learning_rate": 3.256042059250358e-05, "loss": 0.0179, "step": 50620 }, { "grad_norm": 0.4814926087856293, "learning_rate": 3.254105167763062e-05, "loss": 0.0208, "step": 50630 }, { "grad_norm": 0.14533258974552155, "learning_rate": 3.252168574601443e-05, "loss": 0.0159, "step": 50640 }, { "grad_norm": 0.2447379231452942, "learning_rate": 3.2502322800964115e-05, "loss": 0.0125, "step": 50650 }, { "grad_norm": 0.29138773679733276, "learning_rate": 3.248296284578825e-05, "loss": 0.0192, "step": 50660 }, { "grad_norm": 0.23350490629673004, "learning_rate": 3.246360588379497e-05, "loss": 0.0188, "step": 50670 }, { "grad_norm": 0.3928588330745697, "learning_rate": 3.24442519182918e-05, "loss": 0.0192, "step": 50680 }, { "grad_norm": 0.3431989550590515, "learning_rate": 3.2424900952585845e-05, "loss": 0.0166, "step": 50690 }, { "grad_norm": 0.29901283979415894, "learning_rate": 3.240555298998362e-05, "loss": 0.0195, "step": 50700 }, { "grad_norm": 0.21915923058986664, "learning_rate": 3.2386208033791194e-05, "loss": 0.0104, "step": 50710 }, { "grad_norm": 0.38595810532569885, "learning_rate": 3.2366866087314064e-05, "loss": 0.014, "step": 50720 }, { "grad_norm": 0.2924899756908417, "learning_rate": 3.234752715385727e-05, "loss": 0.0147, "step": 50730 }, { "grad_norm": 0.35366955399513245, "learning_rate": 3.232819123672529e-05, "loss": 0.0236, "step": 50740 }, { "grad_norm": 0.45381394028663635, "learning_rate": 3.2308858339222076e-05, "loss": 0.0201, "step": 50750 }, { "grad_norm": 0.2529886066913605, "learning_rate": 3.2289528464651144e-05, "loss": 0.0175, "step": 50760 }, { "grad_norm": 0.236948162317276, "learning_rate": 3.22702016163154e-05, "loss": 0.0166, "step": 50770 }, { "grad_norm": 0.24263916909694672, "learning_rate": 3.2250877797517295e-05, "loss": 0.0158, "step": 50780 }, { "grad_norm": 0.15226799249649048, "learning_rate": 3.223155701155872e-05, "loss": 0.0214, "step": 50790 }, { "grad_norm": 0.27020159363746643, "learning_rate": 3.2212239261741105e-05, "loss": 0.0187, "step": 50800 }, { "grad_norm": 0.21665291488170624, "learning_rate": 3.219292455136528e-05, "loss": 0.0137, "step": 50810 }, { "grad_norm": 0.30966657400131226, "learning_rate": 3.217361288373165e-05, "loss": 0.014, "step": 50820 }, { "grad_norm": 0.22763286530971527, "learning_rate": 3.215430426214002e-05, "loss": 0.0139, "step": 50830 }, { "grad_norm": 0.17713235318660736, "learning_rate": 3.2134998689889705e-05, "loss": 0.0093, "step": 50840 }, { "grad_norm": 0.26324015855789185, "learning_rate": 3.211569617027952e-05, "loss": 0.0106, "step": 50850 }, { "grad_norm": 0.11880394071340561, "learning_rate": 3.209639670660769e-05, "loss": 0.0148, "step": 50860 }, { "grad_norm": 0.5016542077064514, "learning_rate": 3.207710030217202e-05, "loss": 0.0181, "step": 50870 }, { "grad_norm": 0.24185729026794434, "learning_rate": 3.205780696026972e-05, "loss": 0.0138, "step": 50880 }, { "grad_norm": 0.20255698263645172, "learning_rate": 3.203851668419749e-05, "loss": 0.0123, "step": 50890 }, { "grad_norm": 0.2829890549182892, "learning_rate": 3.201922947725149e-05, "loss": 0.0139, "step": 50900 }, { "grad_norm": 0.235092431306839, "learning_rate": 3.199994534272742e-05, "loss": 0.0157, "step": 50910 }, { "grad_norm": 0.22960051894187927, "learning_rate": 3.1980664283920394e-05, "loss": 0.0167, "step": 50920 }, { "grad_norm": 0.20928193628787994, "learning_rate": 3.196138630412499e-05, "loss": 0.0141, "step": 50930 }, { "grad_norm": 0.21178746223449707, "learning_rate": 3.194211140663532e-05, "loss": 0.0118, "step": 50940 }, { "grad_norm": 0.2222081422805786, "learning_rate": 3.1922839594744914e-05, "loss": 0.0161, "step": 50950 }, { "grad_norm": 0.3314826488494873, "learning_rate": 3.190357087174683e-05, "loss": 0.0185, "step": 50960 }, { "grad_norm": 0.16446825861930847, "learning_rate": 3.188430524093353e-05, "loss": 0.0145, "step": 50970 }, { "grad_norm": 0.2647886574268341, "learning_rate": 3.1865042705597014e-05, "loss": 0.0203, "step": 50980 }, { "grad_norm": 0.41642245650291443, "learning_rate": 3.184578326902868e-05, "loss": 0.0174, "step": 50990 }, { "grad_norm": 0.2999310791492462, "learning_rate": 3.182652693451949e-05, "loss": 0.0178, "step": 51000 }, { "grad_norm": 0.6011447906494141, "learning_rate": 3.180727370535981e-05, "loss": 0.0166, "step": 51010 }, { "grad_norm": 0.30946028232574463, "learning_rate": 3.178802358483945e-05, "loss": 0.0168, "step": 51020 }, { "grad_norm": 0.2469760626554489, "learning_rate": 3.17687765762478e-05, "loss": 0.0177, "step": 51030 }, { "grad_norm": 0.2048453837633133, "learning_rate": 3.174953268287357e-05, "loss": 0.0201, "step": 51040 }, { "grad_norm": 0.3346022665500641, "learning_rate": 3.173029190800508e-05, "loss": 0.013, "step": 51050 }, { "grad_norm": 0.3568671941757202, "learning_rate": 3.171105425493e-05, "loss": 0.0197, "step": 51060 }, { "grad_norm": 0.19585150480270386, "learning_rate": 3.169181972693557e-05, "loss": 0.0129, "step": 51070 }, { "grad_norm": 0.14560675621032715, "learning_rate": 3.1672588327308404e-05, "loss": 0.0131, "step": 51080 }, { "grad_norm": 0.3495214283466339, "learning_rate": 3.1653360059334646e-05, "loss": 0.0177, "step": 51090 }, { "grad_norm": 0.6737074851989746, "learning_rate": 3.163413492629985e-05, "loss": 0.0266, "step": 51100 }, { "grad_norm": 0.23608671128749847, "learning_rate": 3.161491293148912e-05, "loss": 0.0125, "step": 51110 }, { "grad_norm": 0.22421813011169434, "learning_rate": 3.1595694078186925e-05, "loss": 0.0129, "step": 51120 }, { "grad_norm": 0.2843988835811615, "learning_rate": 3.157647836967724e-05, "loss": 0.0134, "step": 51130 }, { "grad_norm": 0.19180701673030853, "learning_rate": 3.155726580924354e-05, "loss": 0.0115, "step": 51140 }, { "grad_norm": 0.3613205552101135, "learning_rate": 3.1538056400168676e-05, "loss": 0.0204, "step": 51150 }, { "grad_norm": 0.2488248646259308, "learning_rate": 3.151885014573506e-05, "loss": 0.0209, "step": 51160 }, { "grad_norm": 0.34150272607803345, "learning_rate": 3.149964704922448e-05, "loss": 0.02, "step": 51170 }, { "grad_norm": 0.28802788257598877, "learning_rate": 3.148044711391825e-05, "loss": 0.0121, "step": 51180 }, { "grad_norm": 0.2022721767425537, "learning_rate": 3.146125034309707e-05, "loss": 0.0141, "step": 51190 }, { "grad_norm": 0.23754167556762695, "learning_rate": 3.1442056740041195e-05, "loss": 0.0184, "step": 51200 }, { "grad_norm": 0.23646634817123413, "learning_rate": 3.142286630803026e-05, "loss": 0.0222, "step": 51210 }, { "grad_norm": 0.179347425699234, "learning_rate": 3.140367905034337e-05, "loss": 0.0156, "step": 51220 }, { "grad_norm": 0.26189595460891724, "learning_rate": 3.138449497025914e-05, "loss": 0.0144, "step": 51230 }, { "grad_norm": 0.25127658247947693, "learning_rate": 3.136531407105557e-05, "loss": 0.016, "step": 51240 }, { "grad_norm": 0.26477304100990295, "learning_rate": 3.1346136356010184e-05, "loss": 0.0172, "step": 51250 }, { "grad_norm": 0.33116549253463745, "learning_rate": 3.132696182839988e-05, "loss": 0.0156, "step": 51260 }, { "grad_norm": 0.25636550784111023, "learning_rate": 3.130779049150111e-05, "loss": 0.021, "step": 51270 }, { "grad_norm": 0.255449503660202, "learning_rate": 3.128862234858971e-05, "loss": 0.0157, "step": 51280 }, { "grad_norm": 0.17245110869407654, "learning_rate": 3.1269457402941005e-05, "loss": 0.0146, "step": 51290 }, { "grad_norm": 0.1599278599023819, "learning_rate": 3.125029565782974e-05, "loss": 0.016, "step": 51300 }, { "grad_norm": 0.4528312683105469, "learning_rate": 3.123113711653013e-05, "loss": 0.017, "step": 51310 }, { "grad_norm": 0.273688405752182, "learning_rate": 3.121198178231587e-05, "loss": 0.0188, "step": 51320 }, { "grad_norm": 0.317405104637146, "learning_rate": 3.119282965846006e-05, "loss": 0.0199, "step": 51330 }, { "grad_norm": 0.3952895700931549, "learning_rate": 3.1173680748235303e-05, "loss": 0.0173, "step": 51340 }, { "grad_norm": 0.24972060322761536, "learning_rate": 3.115453505491358e-05, "loss": 0.0137, "step": 51350 }, { "grad_norm": 0.18252959847450256, "learning_rate": 3.1135392581766404e-05, "loss": 0.0152, "step": 51360 }, { "grad_norm": 0.26458725333213806, "learning_rate": 3.111625333206467e-05, "loss": 0.0138, "step": 51370 }, { "grad_norm": 0.1690535992383957, "learning_rate": 3.109711730907881e-05, "loss": 0.0127, "step": 51380 }, { "grad_norm": 0.23007605969905853, "learning_rate": 3.107798451607859e-05, "loss": 0.0142, "step": 51390 }, { "grad_norm": 0.20318809151649475, "learning_rate": 3.105885495633328e-05, "loss": 0.0129, "step": 51400 }, { "grad_norm": 0.12122871726751328, "learning_rate": 3.103972863311163e-05, "loss": 0.0115, "step": 51410 }, { "grad_norm": 0.2646850049495697, "learning_rate": 3.102060554968178e-05, "loss": 0.0166, "step": 51420 }, { "grad_norm": 0.21340616047382355, "learning_rate": 3.100148570931137e-05, "loss": 0.0133, "step": 51430 }, { "grad_norm": 0.25949975848197937, "learning_rate": 3.098236911526744e-05, "loss": 0.016, "step": 51440 }, { "grad_norm": 0.16594655811786652, "learning_rate": 3.0963255770816495e-05, "loss": 0.01, "step": 51450 }, { "grad_norm": 0.15914350748062134, "learning_rate": 3.094414567922447e-05, "loss": 0.0135, "step": 51460 }, { "grad_norm": 0.1927049309015274, "learning_rate": 3.092503884375678e-05, "loss": 0.0153, "step": 51470 }, { "grad_norm": 0.3486212193965912, "learning_rate": 3.0905935267678255e-05, "loss": 0.0186, "step": 51480 }, { "grad_norm": 0.24476979672908783, "learning_rate": 3.0886834954253166e-05, "loss": 0.0134, "step": 51490 }, { "grad_norm": 0.30424320697784424, "learning_rate": 3.086773790674524e-05, "loss": 0.0144, "step": 51500 }, { "grad_norm": 0.22116915881633759, "learning_rate": 3.0848644128417617e-05, "loss": 0.0139, "step": 51510 }, { "grad_norm": 0.30355218052864075, "learning_rate": 3.082955362253294e-05, "loss": 0.0154, "step": 51520 }, { "grad_norm": 0.21372465789318085, "learning_rate": 3.081046639235323e-05, "loss": 0.0162, "step": 51530 }, { "grad_norm": 0.2511298656463623, "learning_rate": 3.0791382441139986e-05, "loss": 0.011, "step": 51540 }, { "grad_norm": 0.1791069060564041, "learning_rate": 3.077230177215411e-05, "loss": 0.0201, "step": 51550 }, { "grad_norm": 0.4746868312358856, "learning_rate": 3.0753224388656e-05, "loss": 0.0166, "step": 51560 }, { "grad_norm": 0.2678053677082062, "learning_rate": 3.073415029390544e-05, "loss": 0.0253, "step": 51570 }, { "grad_norm": 0.14124439656734467, "learning_rate": 3.071507949116166e-05, "loss": 0.011, "step": 51580 }, { "grad_norm": 0.22980494797229767, "learning_rate": 3.069601198368337e-05, "loss": 0.0118, "step": 51590 }, { "grad_norm": 0.18645641207695007, "learning_rate": 3.067694777472864e-05, "loss": 0.016, "step": 51600 }, { "grad_norm": 0.2827545404434204, "learning_rate": 3.065788686755508e-05, "loss": 0.0116, "step": 51610 }, { "grad_norm": 0.21795593202114105, "learning_rate": 3.063882926541961e-05, "loss": 0.0141, "step": 51620 }, { "grad_norm": 0.2587359547615051, "learning_rate": 3.061977497157872e-05, "loss": 0.0128, "step": 51630 }, { "grad_norm": 0.3094545006752014, "learning_rate": 3.0600723989288224e-05, "loss": 0.0144, "step": 51640 }, { "grad_norm": 0.3264698386192322, "learning_rate": 3.058167632180343e-05, "loss": 0.0179, "step": 51650 }, { "grad_norm": 0.19771380722522736, "learning_rate": 3.0562631972379074e-05, "loss": 0.0161, "step": 51660 }, { "grad_norm": 0.23634149134159088, "learning_rate": 3.0543590944269276e-05, "loss": 0.011, "step": 51670 }, { "grad_norm": 0.3493232727050781, "learning_rate": 3.052455324072766e-05, "loss": 0.0155, "step": 51680 }, { "grad_norm": 0.12560993432998657, "learning_rate": 3.0505518865007244e-05, "loss": 0.0128, "step": 51690 }, { "grad_norm": 0.23539979755878448, "learning_rate": 3.048648782036048e-05, "loss": 0.0149, "step": 51700 }, { "grad_norm": 0.257127583026886, "learning_rate": 3.0467460110039236e-05, "loss": 0.013, "step": 51710 }, { "grad_norm": 0.23532971739768982, "learning_rate": 3.0448435737294866e-05, "loss": 0.016, "step": 51720 }, { "grad_norm": 0.18451660871505737, "learning_rate": 3.042941470537808e-05, "loss": 0.0137, "step": 51730 }, { "grad_norm": 0.15285243093967438, "learning_rate": 3.0410397017539072e-05, "loss": 0.013, "step": 51740 }, { "grad_norm": 0.2038314938545227, "learning_rate": 3.039138267702742e-05, "loss": 0.0105, "step": 51750 }, { "grad_norm": 0.16430194675922394, "learning_rate": 3.0372371687092193e-05, "loss": 0.0131, "step": 51760 }, { "grad_norm": 0.24228535592556, "learning_rate": 3.0353364050981823e-05, "loss": 0.0143, "step": 51770 }, { "grad_norm": 0.19225876033306122, "learning_rate": 3.033435977194418e-05, "loss": 0.0188, "step": 51780 }, { "grad_norm": 0.18386854231357574, "learning_rate": 3.0315358853226618e-05, "loss": 0.0144, "step": 51790 }, { "grad_norm": 0.1882208287715912, "learning_rate": 3.0296361298075844e-05, "loss": 0.0131, "step": 51800 }, { "grad_norm": 0.18086779117584229, "learning_rate": 3.027736710973803e-05, "loss": 0.0144, "step": 51810 }, { "grad_norm": 0.14079514145851135, "learning_rate": 3.025837629145875e-05, "loss": 0.0117, "step": 51820 }, { "grad_norm": 0.13994041085243225, "learning_rate": 3.0239388846483048e-05, "loss": 0.017, "step": 51830 }, { "grad_norm": 0.1819819211959839, "learning_rate": 3.022040477805532e-05, "loss": 0.015, "step": 51840 }, { "grad_norm": 0.3035750389099121, "learning_rate": 3.020142408941946e-05, "loss": 0.0152, "step": 51850 }, { "grad_norm": 0.14561718702316284, "learning_rate": 3.018244678381873e-05, "loss": 0.0155, "step": 51860 }, { "grad_norm": 0.3571781814098358, "learning_rate": 3.0163472864495812e-05, "loss": 0.0208, "step": 51870 }, { "grad_norm": 0.18770840764045715, "learning_rate": 3.014450233469287e-05, "loss": 0.0124, "step": 51880 }, { "grad_norm": 0.21947266161441803, "learning_rate": 3.0125535197651422e-05, "loss": 0.0181, "step": 51890 }, { "grad_norm": 0.29408571124076843, "learning_rate": 3.0106571456612448e-05, "loss": 0.0137, "step": 51900 }, { "grad_norm": 0.36085352301597595, "learning_rate": 3.0087611114816305e-05, "loss": 0.0221, "step": 51910 }, { "grad_norm": 0.32705554366111755, "learning_rate": 3.006865417550284e-05, "loss": 0.0132, "step": 51920 }, { "grad_norm": 0.37257668375968933, "learning_rate": 3.0049700641911242e-05, "loss": 0.0124, "step": 51930 }, { "grad_norm": 0.41546985507011414, "learning_rate": 3.0030750517280175e-05, "loss": 0.014, "step": 51940 }, { "grad_norm": 0.21428345143795013, "learning_rate": 3.0011803804847682e-05, "loss": 0.0188, "step": 51950 }, { "grad_norm": 0.350026398897171, "learning_rate": 2.9992860507851224e-05, "loss": 0.0133, "step": 51960 }, { "grad_norm": 0.4234236478805542, "learning_rate": 2.9973920629527723e-05, "loss": 0.0149, "step": 51970 }, { "grad_norm": 0.2708020508289337, "learning_rate": 2.9954984173113453e-05, "loss": 0.0129, "step": 51980 }, { "grad_norm": 0.28453364968299866, "learning_rate": 2.993605114184418e-05, "loss": 0.0183, "step": 51990 }, { "grad_norm": 0.32042643427848816, "learning_rate": 2.9917121538955005e-05, "loss": 0.0208, "step": 52000 }, { "grad_norm": 0.28507041931152344, "learning_rate": 2.98981953676805e-05, "loss": 0.0241, "step": 52010 }, { "grad_norm": 0.35006797313690186, "learning_rate": 2.9879272631254594e-05, "loss": 0.0158, "step": 52020 }, { "grad_norm": 0.2193412184715271, "learning_rate": 2.986035333291073e-05, "loss": 0.0133, "step": 52030 }, { "grad_norm": 0.3314300775527954, "learning_rate": 2.9841437475881646e-05, "loss": 0.0241, "step": 52040 }, { "grad_norm": 0.2879173457622528, "learning_rate": 2.9822525063399555e-05, "loss": 0.0136, "step": 52050 }, { "grad_norm": 0.24721011519432068, "learning_rate": 2.9803616098696087e-05, "loss": 0.0202, "step": 52060 }, { "grad_norm": 0.20246398448944092, "learning_rate": 2.9784710585002228e-05, "loss": 0.02, "step": 52070 }, { "grad_norm": 0.22538331151008606, "learning_rate": 2.9765808525548467e-05, "loss": 0.0156, "step": 52080 }, { "grad_norm": 0.19193941354751587, "learning_rate": 2.974690992356461e-05, "loss": 0.0157, "step": 52090 }, { "grad_norm": 0.19540180265903473, "learning_rate": 2.9728014782279934e-05, "loss": 0.0167, "step": 52100 }, { "grad_norm": 0.19571520388126373, "learning_rate": 2.970912310492307e-05, "loss": 0.009, "step": 52110 }, { "grad_norm": 0.17700344324111938, "learning_rate": 2.9690234894722134e-05, "loss": 0.0114, "step": 52120 }, { "grad_norm": 0.10963302850723267, "learning_rate": 2.9671350154904577e-05, "loss": 0.0163, "step": 52130 }, { "grad_norm": 0.26844027638435364, "learning_rate": 2.9652468888697282e-05, "loss": 0.0142, "step": 52140 }, { "grad_norm": 0.14733758568763733, "learning_rate": 2.9633591099326562e-05, "loss": 0.0103, "step": 52150 }, { "grad_norm": 0.1619146466255188, "learning_rate": 2.9614716790018072e-05, "loss": 0.0118, "step": 52160 }, { "grad_norm": 0.21410655975341797, "learning_rate": 2.959584596399697e-05, "loss": 0.0115, "step": 52170 }, { "grad_norm": 0.2932843565940857, "learning_rate": 2.9576978624487717e-05, "loss": 0.0142, "step": 52180 }, { "grad_norm": 0.14111390709877014, "learning_rate": 2.9558114774714263e-05, "loss": 0.0132, "step": 52190 }, { "grad_norm": 0.2648874521255493, "learning_rate": 2.9539254417899897e-05, "loss": 0.0141, "step": 52200 }, { "grad_norm": 0.11207978427410126, "learning_rate": 2.9520397557267365e-05, "loss": 0.0113, "step": 52210 }, { "grad_norm": 0.16542209684848785, "learning_rate": 2.9501544196038765e-05, "loss": 0.0122, "step": 52220 }, { "grad_norm": 0.14836344122886658, "learning_rate": 2.9482694337435614e-05, "loss": 0.0166, "step": 52230 }, { "grad_norm": 0.20861269533634186, "learning_rate": 2.946384798467887e-05, "loss": 0.0095, "step": 52240 }, { "grad_norm": 0.1112215667963028, "learning_rate": 2.9445005140988824e-05, "loss": 0.0096, "step": 52250 }, { "grad_norm": 0.14551183581352234, "learning_rate": 2.942616580958524e-05, "loss": 0.0123, "step": 52260 }, { "grad_norm": 0.14458991587162018, "learning_rate": 2.9407329993687193e-05, "loss": 0.0108, "step": 52270 }, { "grad_norm": 0.15213342010974884, "learning_rate": 2.938849769651326e-05, "loss": 0.0104, "step": 52280 }, { "grad_norm": 0.22268524765968323, "learning_rate": 2.9369668921281345e-05, "loss": 0.0114, "step": 52290 }, { "grad_norm": 0.17989715933799744, "learning_rate": 2.9350843671208773e-05, "loss": 0.0152, "step": 52300 }, { "grad_norm": 0.22602742910385132, "learning_rate": 2.9332021949512245e-05, "loss": 0.0151, "step": 52310 }, { "grad_norm": 0.33451420068740845, "learning_rate": 2.9313203759407908e-05, "loss": 0.0131, "step": 52320 }, { "grad_norm": 0.2691921591758728, "learning_rate": 2.929438910411127e-05, "loss": 0.0221, "step": 52330 }, { "grad_norm": 0.33160680532455444, "learning_rate": 2.927557798683722e-05, "loss": 0.0153, "step": 52340 }, { "grad_norm": 0.20115822553634644, "learning_rate": 2.9256770410800095e-05, "loss": 0.0116, "step": 52350 }, { "grad_norm": 0.13605301082134247, "learning_rate": 2.9237966379213554e-05, "loss": 0.0145, "step": 52360 }, { "grad_norm": 0.23321224749088287, "learning_rate": 2.9219165895290736e-05, "loss": 0.0162, "step": 52370 }, { "grad_norm": 0.22571726143360138, "learning_rate": 2.9200368962244083e-05, "loss": 0.0157, "step": 52380 }, { "grad_norm": 0.1532447189092636, "learning_rate": 2.9181575583285536e-05, "loss": 0.0118, "step": 52390 }, { "grad_norm": 0.355135053396225, "learning_rate": 2.916278576162632e-05, "loss": 0.0149, "step": 52400 }, { "grad_norm": 0.16774554550647736, "learning_rate": 2.9143999500477116e-05, "loss": 0.0141, "step": 52410 }, { "grad_norm": 0.17637038230895996, "learning_rate": 2.9125216803048004e-05, "loss": 0.0157, "step": 52420 }, { "grad_norm": 0.22040219604969025, "learning_rate": 2.9106437672548403e-05, "loss": 0.0106, "step": 52430 }, { "grad_norm": 0.20146916806697845, "learning_rate": 2.9087662112187154e-05, "loss": 0.0127, "step": 52440 }, { "grad_norm": 0.23007845878601074, "learning_rate": 2.9068890125172498e-05, "loss": 0.0149, "step": 52450 }, { "grad_norm": 0.15592241287231445, "learning_rate": 2.9050121714712058e-05, "loss": 0.0177, "step": 52460 }, { "grad_norm": 0.23985743522644043, "learning_rate": 2.9031356884012835e-05, "loss": 0.0115, "step": 52470 }, { "grad_norm": 0.22298620641231537, "learning_rate": 2.9012595636281247e-05, "loss": 0.0162, "step": 52480 }, { "grad_norm": 0.153700053691864, "learning_rate": 2.8993837974723016e-05, "loss": 0.0265, "step": 52490 }, { "grad_norm": 0.23337160050868988, "learning_rate": 2.8975083902543394e-05, "loss": 0.0146, "step": 52500 }, { "grad_norm": 0.2444794923067093, "learning_rate": 2.895633342294688e-05, "loss": 0.0163, "step": 52510 }, { "grad_norm": 0.15652258694171906, "learning_rate": 2.893758653913744e-05, "loss": 0.0128, "step": 52520 }, { "grad_norm": 0.16021662950515747, "learning_rate": 2.8918843254318423e-05, "loss": 0.0122, "step": 52530 }, { "grad_norm": 0.28688716888427734, "learning_rate": 2.8900103571692483e-05, "loss": 0.0129, "step": 52540 }, { "grad_norm": 0.3549479842185974, "learning_rate": 2.8881367494461796e-05, "loss": 0.014, "step": 52550 }, { "grad_norm": 0.16763612627983093, "learning_rate": 2.886263502582779e-05, "loss": 0.0163, "step": 52560 }, { "grad_norm": 0.2583214342594147, "learning_rate": 2.8843906168991353e-05, "loss": 0.0133, "step": 52570 }, { "grad_norm": 0.1876537799835205, "learning_rate": 2.8825180927152728e-05, "loss": 0.0126, "step": 52580 }, { "grad_norm": 0.3021981418132782, "learning_rate": 2.8806459303511547e-05, "loss": 0.0131, "step": 52590 }, { "grad_norm": 0.2887984812259674, "learning_rate": 2.8787741301266852e-05, "loss": 0.014, "step": 52600 }, { "grad_norm": 0.2620052099227905, "learning_rate": 2.876902692361697e-05, "loss": 0.0119, "step": 52610 }, { "grad_norm": 0.16103069484233856, "learning_rate": 2.875031617375976e-05, "loss": 0.0105, "step": 52620 }, { "grad_norm": 0.17718574404716492, "learning_rate": 2.8731609054892318e-05, "loss": 0.0139, "step": 52630 }, { "grad_norm": 0.24450437724590302, "learning_rate": 2.871290557021119e-05, "loss": 0.0177, "step": 52640 }, { "grad_norm": 0.28480657935142517, "learning_rate": 2.8694205722912298e-05, "loss": 0.0163, "step": 52650 }, { "grad_norm": 0.4588285982608795, "learning_rate": 2.8675509516190936e-05, "loss": 0.014, "step": 52660 }, { "grad_norm": 0.21687521040439606, "learning_rate": 2.8656816953241772e-05, "loss": 0.0147, "step": 52670 }, { "grad_norm": 0.32537493109703064, "learning_rate": 2.863812803725887e-05, "loss": 0.0171, "step": 52680 }, { "grad_norm": 0.3962402641773224, "learning_rate": 2.8619442771435623e-05, "loss": 0.0124, "step": 52690 }, { "grad_norm": 0.16346625983715057, "learning_rate": 2.860076115896484e-05, "loss": 0.0189, "step": 52700 }, { "grad_norm": 0.23924648761749268, "learning_rate": 2.8582083203038712e-05, "loss": 0.0126, "step": 52710 }, { "grad_norm": 0.26829513907432556, "learning_rate": 2.8563408906848778e-05, "loss": 0.0158, "step": 52720 }, { "grad_norm": 0.25255104899406433, "learning_rate": 2.8544738273585993e-05, "loss": 0.0122, "step": 52730 }, { "grad_norm": 0.3466159999370575, "learning_rate": 2.8526071306440595e-05, "loss": 0.0129, "step": 52740 }, { "grad_norm": 0.25177088379859924, "learning_rate": 2.850740800860233e-05, "loss": 0.0174, "step": 52750 }, { "grad_norm": 0.18069912493228912, "learning_rate": 2.8488748383260178e-05, "loss": 0.014, "step": 52760 }, { "grad_norm": 0.24872437119483948, "learning_rate": 2.8470092433602635e-05, "loss": 0.0188, "step": 52770 }, { "grad_norm": 0.3176388144493103, "learning_rate": 2.845144016281745e-05, "loss": 0.0136, "step": 52780 }, { "grad_norm": 0.2565770447254181, "learning_rate": 2.8432791574091744e-05, "loss": 0.0229, "step": 52790 }, { "grad_norm": 0.21752743422985077, "learning_rate": 2.8414146670612134e-05, "loss": 0.0169, "step": 52800 }, { "grad_norm": 0.1807270497083664, "learning_rate": 2.8395505455564446e-05, "loss": 0.0176, "step": 52810 }, { "grad_norm": 0.25870683789253235, "learning_rate": 2.837686793213403e-05, "loss": 0.0111, "step": 52820 }, { "grad_norm": 0.21023103594779968, "learning_rate": 2.8358234103505477e-05, "loss": 0.0145, "step": 52830 }, { "grad_norm": 0.19476918876171112, "learning_rate": 2.8339603972862806e-05, "loss": 0.0125, "step": 52840 }, { "grad_norm": 0.1440933495759964, "learning_rate": 2.83209775433894e-05, "loss": 0.0134, "step": 52850 }, { "grad_norm": 0.1463867872953415, "learning_rate": 2.8302354818268013e-05, "loss": 0.0127, "step": 52860 }, { "grad_norm": 0.2254704236984253, "learning_rate": 2.8283735800680754e-05, "loss": 0.0135, "step": 52870 }, { "grad_norm": 0.3020128011703491, "learning_rate": 2.826512049380913e-05, "loss": 0.0139, "step": 52880 }, { "grad_norm": 0.16483379900455475, "learning_rate": 2.8246508900833935e-05, "loss": 0.0203, "step": 52890 }, { "grad_norm": 0.2541565001010895, "learning_rate": 2.822790102493541e-05, "loss": 0.0168, "step": 52900 }, { "grad_norm": 0.2739507853984833, "learning_rate": 2.820929686929314e-05, "loss": 0.0138, "step": 52910 }, { "grad_norm": 0.17511330544948578, "learning_rate": 2.819069643708605e-05, "loss": 0.0161, "step": 52920 }, { "grad_norm": 0.3022541403770447, "learning_rate": 2.8172099731492484e-05, "loss": 0.0182, "step": 52930 }, { "grad_norm": 0.3163541555404663, "learning_rate": 2.815350675569003e-05, "loss": 0.0257, "step": 52940 }, { "grad_norm": 0.2094830572605133, "learning_rate": 2.8134917512855807e-05, "loss": 0.0174, "step": 52950 }, { "grad_norm": 0.2138122171163559, "learning_rate": 2.8116332006166134e-05, "loss": 0.0117, "step": 52960 }, { "grad_norm": 0.3164273202419281, "learning_rate": 2.809775023879685e-05, "loss": 0.0236, "step": 52970 }, { "grad_norm": 0.23418772220611572, "learning_rate": 2.807917221392299e-05, "loss": 0.0167, "step": 52980 }, { "grad_norm": 0.19162407517433167, "learning_rate": 2.8060597934719067e-05, "loss": 0.0136, "step": 52990 }, { "grad_norm": 0.23384544253349304, "learning_rate": 2.8042027404358927e-05, "loss": 0.0118, "step": 53000 }, { "grad_norm": 0.23721081018447876, "learning_rate": 2.802346062601572e-05, "loss": 0.0155, "step": 53010 }, { "grad_norm": 0.39398136734962463, "learning_rate": 2.800489760286207e-05, "loss": 0.0181, "step": 53020 }, { "grad_norm": 0.25888460874557495, "learning_rate": 2.7986338338069834e-05, "loss": 0.0114, "step": 53030 }, { "grad_norm": 0.35916173458099365, "learning_rate": 2.7967782834810297e-05, "loss": 0.0201, "step": 53040 }, { "grad_norm": 0.16416876018047333, "learning_rate": 2.794923109625409e-05, "loss": 0.0098, "step": 53050 }, { "grad_norm": 0.25554749369621277, "learning_rate": 2.7930683125571205e-05, "loss": 0.0122, "step": 53060 }, { "grad_norm": 0.3286993205547333, "learning_rate": 2.7912138925930997e-05, "loss": 0.0186, "step": 53070 }, { "grad_norm": 0.22701305150985718, "learning_rate": 2.7893598500502117e-05, "loss": 0.021, "step": 53080 }, { "grad_norm": 0.3720249831676483, "learning_rate": 2.7875061852452644e-05, "loss": 0.0169, "step": 53090 }, { "grad_norm": 0.3250078856945038, "learning_rate": 2.7856528984949982e-05, "loss": 0.0124, "step": 53100 }, { "grad_norm": 0.29201799631118774, "learning_rate": 2.7837999901160888e-05, "loss": 0.0118, "step": 53110 }, { "grad_norm": 0.17945732176303864, "learning_rate": 2.7819474604251484e-05, "loss": 0.0121, "step": 53120 }, { "grad_norm": 0.1619863510131836, "learning_rate": 2.780095309738725e-05, "loss": 0.0113, "step": 53130 }, { "grad_norm": 0.1913098394870758, "learning_rate": 2.778243538373294e-05, "loss": 0.012, "step": 53140 }, { "grad_norm": 0.3537815511226654, "learning_rate": 2.7763921466452826e-05, "loss": 0.0188, "step": 53150 }, { "grad_norm": 0.20698201656341553, "learning_rate": 2.7745411348710336e-05, "loss": 0.0207, "step": 53160 }, { "grad_norm": 0.37519529461860657, "learning_rate": 2.7726905033668395e-05, "loss": 0.0149, "step": 53170 }, { "grad_norm": 0.2782491445541382, "learning_rate": 2.7708402524489214e-05, "loss": 0.015, "step": 53180 }, { "grad_norm": 0.2856753468513489, "learning_rate": 2.7689903824334364e-05, "loss": 0.0156, "step": 53190 }, { "grad_norm": 0.40033766627311707, "learning_rate": 2.7671408936364785e-05, "loss": 0.0167, "step": 53200 }, { "grad_norm": 0.18498651683330536, "learning_rate": 2.765291786374069e-05, "loss": 0.0135, "step": 53210 }, { "grad_norm": 0.17236588895320892, "learning_rate": 2.7634430609621775e-05, "loss": 0.0112, "step": 53220 }, { "grad_norm": 0.31599289178848267, "learning_rate": 2.7615947177166956e-05, "loss": 0.0141, "step": 53230 }, { "grad_norm": 0.17655791342258453, "learning_rate": 2.7597467569534553e-05, "loss": 0.0106, "step": 53240 }, { "grad_norm": 0.17562523484230042, "learning_rate": 2.757899178988226e-05, "loss": 0.0124, "step": 53250 }, { "grad_norm": 0.1635541319847107, "learning_rate": 2.7560519841367005e-05, "loss": 0.0129, "step": 53260 }, { "grad_norm": 0.15784606337547302, "learning_rate": 2.7542051727145237e-05, "loss": 0.0147, "step": 53270 }, { "grad_norm": 0.2749975919723511, "learning_rate": 2.7523587450372578e-05, "loss": 0.0187, "step": 53280 }, { "grad_norm": 0.4977172315120697, "learning_rate": 2.750512701420409e-05, "loss": 0.0189, "step": 53290 }, { "grad_norm": 0.20513276755809784, "learning_rate": 2.7486670421794158e-05, "loss": 0.0117, "step": 53300 }, { "grad_norm": 0.18293438851833344, "learning_rate": 2.7468217676296515e-05, "loss": 0.0216, "step": 53310 }, { "grad_norm": 0.3705763816833496, "learning_rate": 2.7449768780864226e-05, "loss": 0.0156, "step": 53320 }, { "grad_norm": 0.41588351130485535, "learning_rate": 2.7431323738649724e-05, "loss": 0.0118, "step": 53330 }, { "grad_norm": 0.6677208542823792, "learning_rate": 2.7412882552804713e-05, "loss": 0.018, "step": 53340 }, { "grad_norm": 0.1844368577003479, "learning_rate": 2.739444522648032e-05, "loss": 0.0119, "step": 53350 }, { "grad_norm": 0.14734023809432983, "learning_rate": 2.737601176282697e-05, "loss": 0.0126, "step": 53360 }, { "grad_norm": 0.20457814633846283, "learning_rate": 2.735758216499445e-05, "loss": 0.0144, "step": 53370 }, { "grad_norm": 0.20360331237316132, "learning_rate": 2.7339156436131864e-05, "loss": 0.0154, "step": 53380 }, { "grad_norm": 0.24475793540477753, "learning_rate": 2.7320734579387663e-05, "loss": 0.0116, "step": 53390 }, { "grad_norm": 0.4579218029975891, "learning_rate": 2.730231659790966e-05, "loss": 0.0144, "step": 53400 }, { "grad_norm": 0.23173613846302032, "learning_rate": 2.728390249484494e-05, "loss": 0.0182, "step": 53410 }, { "grad_norm": 0.2195066213607788, "learning_rate": 2.7265492273340032e-05, "loss": 0.0119, "step": 53420 }, { "grad_norm": 0.40216490626335144, "learning_rate": 2.7247085936540678e-05, "loss": 0.013, "step": 53430 }, { "grad_norm": 0.16822585463523865, "learning_rate": 2.722868348759205e-05, "loss": 0.0148, "step": 53440 }, { "grad_norm": 0.21697306632995605, "learning_rate": 2.7210284929638635e-05, "loss": 0.0148, "step": 53450 }, { "grad_norm": 0.19200444221496582, "learning_rate": 2.7191890265824183e-05, "loss": 0.0116, "step": 53460 }, { "grad_norm": 0.17238226532936096, "learning_rate": 2.7173499499291926e-05, "loss": 0.0148, "step": 53470 }, { "grad_norm": 0.3511884808540344, "learning_rate": 2.7155112633184277e-05, "loss": 0.0135, "step": 53480 }, { "grad_norm": 0.3686644732952118, "learning_rate": 2.713672967064307e-05, "loss": 0.014, "step": 53490 }, { "grad_norm": 0.1644047349691391, "learning_rate": 2.711835061480945e-05, "loss": 0.0113, "step": 53500 }, { "grad_norm": 0.14484603703022003, "learning_rate": 2.7099975468823896e-05, "loss": 0.0144, "step": 53510 }, { "grad_norm": 0.3835228681564331, "learning_rate": 2.708160423582622e-05, "loss": 0.0159, "step": 53520 }, { "grad_norm": 0.623020589351654, "learning_rate": 2.706323691895557e-05, "loss": 0.0218, "step": 53530 }, { "grad_norm": 0.2776505649089813, "learning_rate": 2.70448735213504e-05, "loss": 0.0128, "step": 53540 }, { "grad_norm": 0.17485420405864716, "learning_rate": 2.702651404614852e-05, "loss": 0.0161, "step": 53550 }, { "grad_norm": 0.14380218088626862, "learning_rate": 2.7008158496487056e-05, "loss": 0.0155, "step": 53560 }, { "grad_norm": 0.2093087136745453, "learning_rate": 2.6989806875502487e-05, "loss": 0.0111, "step": 53570 }, { "grad_norm": 0.6082698106765747, "learning_rate": 2.6971459186330584e-05, "loss": 0.0141, "step": 53580 }, { "grad_norm": 0.24585328996181488, "learning_rate": 2.695311543210648e-05, "loss": 0.0132, "step": 53590 }, { "grad_norm": 0.24180907011032104, "learning_rate": 2.693477561596463e-05, "loss": 0.0149, "step": 53600 }, { "grad_norm": 0.23284226655960083, "learning_rate": 2.6916439741038756e-05, "loss": 0.0149, "step": 53610 }, { "grad_norm": 0.207721546292305, "learning_rate": 2.689810781046203e-05, "loss": 0.015, "step": 53620 }, { "grad_norm": 0.30711767077445984, "learning_rate": 2.6879779827366823e-05, "loss": 0.015, "step": 53630 }, { "grad_norm": 0.16161862015724182, "learning_rate": 2.6861455794884904e-05, "loss": 0.012, "step": 53640 }, { "grad_norm": 0.21659350395202637, "learning_rate": 2.6843135716147373e-05, "loss": 0.0108, "step": 53650 }, { "grad_norm": 0.19246821105480194, "learning_rate": 2.6824819594284556e-05, "loss": 0.0181, "step": 53660 }, { "grad_norm": 0.13513974845409393, "learning_rate": 2.6806507432426275e-05, "loss": 0.0093, "step": 53670 }, { "grad_norm": 0.18572083115577698, "learning_rate": 2.6788199233701512e-05, "loss": 0.0142, "step": 53680 }, { "grad_norm": 0.13848330080509186, "learning_rate": 2.6769895001238652e-05, "loss": 0.0152, "step": 53690 }, { "grad_norm": 0.22155936062335968, "learning_rate": 2.67515947381654e-05, "loss": 0.0153, "step": 53700 }, { "grad_norm": 0.20962661504745483, "learning_rate": 2.6733298447608745e-05, "loss": 0.0114, "step": 53710 }, { "grad_norm": 0.44587746262550354, "learning_rate": 2.6715006132695074e-05, "loss": 0.0167, "step": 53720 }, { "grad_norm": 0.24986200034618378, "learning_rate": 2.6696717796549985e-05, "loss": 0.0181, "step": 53730 }, { "grad_norm": 0.33539140224456787, "learning_rate": 2.667843344229848e-05, "loss": 0.018, "step": 53740 }, { "grad_norm": 0.36035969853401184, "learning_rate": 2.6660153073064846e-05, "loss": 0.0157, "step": 53750 }, { "grad_norm": 0.24743877351284027, "learning_rate": 2.664187669197271e-05, "loss": 0.0201, "step": 53760 }, { "grad_norm": 0.25093168020248413, "learning_rate": 2.6623604302145005e-05, "loss": 0.0108, "step": 53770 }, { "grad_norm": 0.3470214009284973, "learning_rate": 2.6605335906703975e-05, "loss": 0.0144, "step": 53780 }, { "grad_norm": 0.37720316648483276, "learning_rate": 2.65870715087712e-05, "loss": 0.0178, "step": 53790 }, { "grad_norm": 0.2688244879245758, "learning_rate": 2.6568811111467573e-05, "loss": 0.012, "step": 53800 }, { "grad_norm": 0.20819704234600067, "learning_rate": 2.6550554717913258e-05, "loss": 0.0137, "step": 53810 }, { "grad_norm": 0.28349757194519043, "learning_rate": 2.6532302331227805e-05, "loss": 0.0124, "step": 53820 }, { "grad_norm": 0.23125199973583221, "learning_rate": 2.651405395453004e-05, "loss": 0.0123, "step": 53830 }, { "grad_norm": 0.14307664334774017, "learning_rate": 2.6495809590938115e-05, "loss": 0.0182, "step": 53840 }, { "grad_norm": 0.21937702596187592, "learning_rate": 2.64775692435695e-05, "loss": 0.022, "step": 53850 }, { "grad_norm": 0.1834816038608551, "learning_rate": 2.6459332915540928e-05, "loss": 0.0118, "step": 53860 }, { "grad_norm": 0.18801234662532806, "learning_rate": 2.644110060996856e-05, "loss": 0.0173, "step": 53870 }, { "grad_norm": 0.14149154722690582, "learning_rate": 2.642287232996774e-05, "loss": 0.0145, "step": 53880 }, { "grad_norm": 0.23350206017494202, "learning_rate": 2.6404648078653205e-05, "loss": 0.0092, "step": 53890 }, { "grad_norm": 0.3296785056591034, "learning_rate": 2.6386427859139002e-05, "loss": 0.0134, "step": 53900 }, { "grad_norm": 0.3851519525051117, "learning_rate": 2.63682116745384e-05, "loss": 0.0164, "step": 53910 }, { "grad_norm": 0.15722522139549255, "learning_rate": 2.6349999527964138e-05, "loss": 0.0126, "step": 53920 }, { "grad_norm": 0.21705658733844757, "learning_rate": 2.63317914225281e-05, "loss": 0.0178, "step": 53930 }, { "grad_norm": 0.21705663204193115, "learning_rate": 2.6313587361341585e-05, "loss": 0.0149, "step": 53940 }, { "grad_norm": 0.4104657471179962, "learning_rate": 2.6295387347515165e-05, "loss": 0.0151, "step": 53950 }, { "grad_norm": 0.1887306421995163, "learning_rate": 2.6277191384158727e-05, "loss": 0.0153, "step": 53960 }, { "grad_norm": 0.17397461831569672, "learning_rate": 2.6258999474381457e-05, "loss": 0.0162, "step": 53970 }, { "grad_norm": 0.18422222137451172, "learning_rate": 2.624081162129186e-05, "loss": 0.0129, "step": 53980 }, { "grad_norm": 0.14869362115859985, "learning_rate": 2.6222627827997765e-05, "loss": 0.019, "step": 53990 }, { "grad_norm": 0.2541014850139618, "learning_rate": 2.6204448097606236e-05, "loss": 0.0108, "step": 54000 }, { "grad_norm": 0.21379558742046356, "learning_rate": 2.6186272433223726e-05, "loss": 0.013, "step": 54010 }, { "grad_norm": 0.1758526861667633, "learning_rate": 2.6168100837955943e-05, "loss": 0.0154, "step": 54020 }, { "grad_norm": 0.2745022475719452, "learning_rate": 2.6149933314907926e-05, "loss": 0.0235, "step": 54030 }, { "grad_norm": 0.3575540781021118, "learning_rate": 2.6131769867184e-05, "loss": 0.0213, "step": 54040 }, { "grad_norm": 0.18402984738349915, "learning_rate": 2.611361049788783e-05, "loss": 0.0129, "step": 54050 }, { "grad_norm": 0.423898845911026, "learning_rate": 2.6095455210122292e-05, "loss": 0.0136, "step": 54060 }, { "grad_norm": 0.19876718521118164, "learning_rate": 2.6077304006989712e-05, "loss": 0.0148, "step": 54070 }, { "grad_norm": 0.30723291635513306, "learning_rate": 2.6059156891591562e-05, "loss": 0.0142, "step": 54080 }, { "grad_norm": 0.17332683503627777, "learning_rate": 2.6041013867028718e-05, "loss": 0.0171, "step": 54090 }, { "grad_norm": 0.23614023625850677, "learning_rate": 2.6022874936401347e-05, "loss": 0.0103, "step": 54100 }, { "grad_norm": 0.1453668177127838, "learning_rate": 2.6004740102808832e-05, "loss": 0.0158, "step": 54110 }, { "grad_norm": 0.19346441328525543, "learning_rate": 2.598660936935e-05, "loss": 0.013, "step": 54120 }, { "grad_norm": 0.19969840347766876, "learning_rate": 2.5968482739122845e-05, "loss": 0.0172, "step": 54130 }, { "grad_norm": 0.4041232466697693, "learning_rate": 2.595036021522472e-05, "loss": 0.013, "step": 54140 }, { "grad_norm": 0.1952529400587082, "learning_rate": 2.5932241800752278e-05, "loss": 0.0131, "step": 54150 }, { "grad_norm": 0.29444923996925354, "learning_rate": 2.5914127498801453e-05, "loss": 0.0265, "step": 54160 }, { "grad_norm": 0.20856250822544098, "learning_rate": 2.5896017312467497e-05, "loss": 0.0161, "step": 54170 }, { "grad_norm": 0.3055075407028198, "learning_rate": 2.587791124484493e-05, "loss": 0.0165, "step": 54180 }, { "grad_norm": 0.16449399292469025, "learning_rate": 2.5859809299027615e-05, "loss": 0.0181, "step": 54190 }, { "grad_norm": 0.3025781512260437, "learning_rate": 2.5841711478108632e-05, "loss": 0.0182, "step": 54200 }, { "grad_norm": 0.27352145314216614, "learning_rate": 2.582361778518043e-05, "loss": 0.0152, "step": 54210 }, { "grad_norm": 0.17782633006572723, "learning_rate": 2.580552822333472e-05, "loss": 0.0147, "step": 54220 }, { "grad_norm": 0.2610708177089691, "learning_rate": 2.578744279566252e-05, "loss": 0.0135, "step": 54230 }, { "grad_norm": 0.47323381900787354, "learning_rate": 2.576936150525413e-05, "loss": 0.0256, "step": 54240 }, { "grad_norm": 0.2964257299900055, "learning_rate": 2.5751284355199168e-05, "loss": 0.0094, "step": 54250 }, { "grad_norm": 0.22009481489658356, "learning_rate": 2.573321134858646e-05, "loss": 0.0101, "step": 54260 }, { "grad_norm": 0.13144011795520782, "learning_rate": 2.5715142488504286e-05, "loss": 0.0127, "step": 54270 }, { "grad_norm": 0.22971339523792267, "learning_rate": 2.5697077778040042e-05, "loss": 0.0129, "step": 54280 }, { "grad_norm": 0.22430655360221863, "learning_rate": 2.5679017220280522e-05, "loss": 0.0296, "step": 54290 }, { "grad_norm": 0.14821648597717285, "learning_rate": 2.5660960818311796e-05, "loss": 0.0134, "step": 54300 }, { "grad_norm": 0.2840811014175415, "learning_rate": 2.564290857521915e-05, "loss": 0.0161, "step": 54310 }, { "grad_norm": 0.16265875101089478, "learning_rate": 2.5624860494087298e-05, "loss": 0.0131, "step": 54320 }, { "grad_norm": 0.17048296332359314, "learning_rate": 2.5606816578000115e-05, "loss": 0.0158, "step": 54330 }, { "grad_norm": 0.28926852345466614, "learning_rate": 2.558877683004082e-05, "loss": 0.0128, "step": 54340 }, { "grad_norm": 0.1417793333530426, "learning_rate": 2.557074125329192e-05, "loss": 0.011, "step": 54350 }, { "grad_norm": 0.18371979892253876, "learning_rate": 2.5552709850835195e-05, "loss": 0.0116, "step": 54360 }, { "grad_norm": 0.20557187497615814, "learning_rate": 2.5534682625751738e-05, "loss": 0.0104, "step": 54370 }, { "grad_norm": 0.24159222841262817, "learning_rate": 2.551665958112186e-05, "loss": 0.0245, "step": 54380 }, { "grad_norm": 0.23086591064929962, "learning_rate": 2.549864072002527e-05, "loss": 0.0132, "step": 54390 }, { "grad_norm": 0.14652062952518463, "learning_rate": 2.5480626045540858e-05, "loss": 0.0164, "step": 54400 }, { "grad_norm": 0.21133941411972046, "learning_rate": 2.546261556074684e-05, "loss": 0.0105, "step": 54410 }, { "grad_norm": 0.31070050597190857, "learning_rate": 2.5444609268720726e-05, "loss": 0.0127, "step": 54420 }, { "grad_norm": 0.2788427174091339, "learning_rate": 2.5426607172539297e-05, "loss": 0.014, "step": 54430 }, { "grad_norm": 0.1720380336046219, "learning_rate": 2.5408609275278617e-05, "loss": 0.0171, "step": 54440 }, { "grad_norm": 0.22238661348819733, "learning_rate": 2.5390615580014055e-05, "loss": 0.0134, "step": 54450 }, { "grad_norm": 0.26909610629081726, "learning_rate": 2.5372626089820207e-05, "loss": 0.0186, "step": 54460 }, { "grad_norm": 0.19705618917942047, "learning_rate": 2.5354640807770997e-05, "loss": 0.0088, "step": 54470 }, { "grad_norm": 0.2734924256801605, "learning_rate": 2.5336659736939622e-05, "loss": 0.0134, "step": 54480 }, { "grad_norm": 0.16122879087924957, "learning_rate": 2.5318682880398554e-05, "loss": 0.0177, "step": 54490 }, { "grad_norm": 0.2925751209259033, "learning_rate": 2.530071024121956e-05, "loss": 0.0169, "step": 54500 }, { "grad_norm": 0.28094416856765747, "learning_rate": 2.5282741822473627e-05, "loss": 0.0126, "step": 54510 }, { "grad_norm": 0.25861310958862305, "learning_rate": 2.526477762723114e-05, "loss": 0.0135, "step": 54520 }, { "grad_norm": 0.18123753368854523, "learning_rate": 2.5246817658561618e-05, "loss": 0.0124, "step": 54530 }, { "grad_norm": 0.37513965368270874, "learning_rate": 2.5228861919533965e-05, "loss": 0.0166, "step": 54540 }, { "grad_norm": 0.2711814343929291, "learning_rate": 2.5210910413216326e-05, "loss": 0.0213, "step": 54550 }, { "grad_norm": 0.4603184163570404, "learning_rate": 2.5192963142676086e-05, "loss": 0.0144, "step": 54560 }, { "grad_norm": 0.1800643503665924, "learning_rate": 2.517502011098001e-05, "loss": 0.0119, "step": 54570 }, { "grad_norm": 0.23444922268390656, "learning_rate": 2.5157081321193987e-05, "loss": 0.0147, "step": 54580 }, { "grad_norm": 0.21613875031471252, "learning_rate": 2.5139146776383356e-05, "loss": 0.0107, "step": 54590 }, { "grad_norm": 0.23422478139400482, "learning_rate": 2.5121216479612575e-05, "loss": 0.0111, "step": 54600 }, { "grad_norm": 0.22321730852127075, "learning_rate": 2.510329043394546e-05, "loss": 0.02, "step": 54610 }, { "grad_norm": 0.2249254584312439, "learning_rate": 2.508536864244508e-05, "loss": 0.0205, "step": 54620 }, { "grad_norm": 0.15910229086875916, "learning_rate": 2.5067451108173778e-05, "loss": 0.0136, "step": 54630 }, { "grad_norm": 0.19722440838813782, "learning_rate": 2.5049537834193204e-05, "loss": 0.011, "step": 54640 }, { "grad_norm": 0.21767692267894745, "learning_rate": 2.5031628823564194e-05, "loss": 0.0149, "step": 54650 }, { "grad_norm": 0.17648597061634064, "learning_rate": 2.5013724079346933e-05, "loss": 0.0118, "step": 54660 }, { "grad_norm": 0.22788214683532715, "learning_rate": 2.4995823604600854e-05, "loss": 0.0087, "step": 54670 }, { "grad_norm": 0.2105918675661087, "learning_rate": 2.497792740238465e-05, "loss": 0.0119, "step": 54680 }, { "grad_norm": 0.2601671516895294, "learning_rate": 2.49600354757563e-05, "loss": 0.0146, "step": 54690 }, { "grad_norm": 0.1913718730211258, "learning_rate": 2.494214782777306e-05, "loss": 0.016, "step": 54700 }, { "grad_norm": 0.3234413266181946, "learning_rate": 2.4924264461491386e-05, "loss": 0.0142, "step": 54710 }, { "grad_norm": 0.2941282093524933, "learning_rate": 2.4906385379967133e-05, "loss": 0.0199, "step": 54720 }, { "grad_norm": 0.3402807414531708, "learning_rate": 2.4888510586255285e-05, "loss": 0.0141, "step": 54730 }, { "grad_norm": 0.14889854192733765, "learning_rate": 2.487064008341018e-05, "loss": 0.0124, "step": 54740 }, { "grad_norm": 0.3715405762195587, "learning_rate": 2.4852773874485407e-05, "loss": 0.0148, "step": 54750 }, { "grad_norm": 0.2033620923757553, "learning_rate": 2.483491196253377e-05, "loss": 0.0152, "step": 54760 }, { "grad_norm": 0.22455404698848724, "learning_rate": 2.4817054350607443e-05, "loss": 0.016, "step": 54770 }, { "grad_norm": 0.26877936720848083, "learning_rate": 2.4799201041757742e-05, "loss": 0.0113, "step": 54780 }, { "grad_norm": 0.28705453872680664, "learning_rate": 2.4781352039035373e-05, "loss": 0.0127, "step": 54790 }, { "grad_norm": 0.13346204161643982, "learning_rate": 2.4763507345490194e-05, "loss": 0.0098, "step": 54800 }, { "grad_norm": 0.34051212668418884, "learning_rate": 2.4745666964171386e-05, "loss": 0.017, "step": 54810 }, { "grad_norm": 0.3758677840232849, "learning_rate": 2.4727830898127387e-05, "loss": 0.011, "step": 54820 }, { "grad_norm": 0.2070213109254837, "learning_rate": 2.4709999150405895e-05, "loss": 0.0148, "step": 54830 }, { "grad_norm": 0.41956692934036255, "learning_rate": 2.469217172405388e-05, "loss": 0.0117, "step": 54840 }, { "grad_norm": 0.15347091853618622, "learning_rate": 2.4674348622117527e-05, "loss": 0.0088, "step": 54850 }, { "grad_norm": 0.20414575934410095, "learning_rate": 2.465652984764234e-05, "loss": 0.0098, "step": 54860 }, { "grad_norm": 0.10940289497375488, "learning_rate": 2.4638715403673056e-05, "loss": 0.0088, "step": 54870 }, { "grad_norm": 0.3852083683013916, "learning_rate": 2.4620905293253676e-05, "loss": 0.0157, "step": 54880 }, { "grad_norm": 0.36524808406829834, "learning_rate": 2.4603099519427463e-05, "loss": 0.0135, "step": 54890 }, { "grad_norm": 0.25042441487312317, "learning_rate": 2.458529808523695e-05, "loss": 0.01, "step": 54900 }, { "grad_norm": 0.19963815808296204, "learning_rate": 2.456750099372387e-05, "loss": 0.0105, "step": 54910 }, { "grad_norm": 0.1654656082391739, "learning_rate": 2.4549708247929327e-05, "loss": 0.0148, "step": 54920 }, { "grad_norm": 0.24329160153865814, "learning_rate": 2.4531919850893554e-05, "loss": 0.0133, "step": 54930 }, { "grad_norm": 0.1630551517009735, "learning_rate": 2.4514135805656125e-05, "loss": 0.0108, "step": 54940 }, { "grad_norm": 0.3959355652332306, "learning_rate": 2.449635611525587e-05, "loss": 0.0163, "step": 54950 }, { "grad_norm": 0.1977481096982956, "learning_rate": 2.447858078273079e-05, "loss": 0.0163, "step": 54960 }, { "grad_norm": 0.3594561815261841, "learning_rate": 2.4460809811118275e-05, "loss": 0.0179, "step": 54970 }, { "grad_norm": 0.2994316518306732, "learning_rate": 2.444304320345483e-05, "loss": 0.0139, "step": 54980 }, { "grad_norm": 0.22371681034564972, "learning_rate": 2.4425280962776354e-05, "loss": 0.0155, "step": 54990 }, { "grad_norm": 0.2537473142147064, "learning_rate": 2.4407523092117875e-05, "loss": 0.0152, "step": 55000 }, { "grad_norm": 0.4255599081516266, "learning_rate": 2.438976959451374e-05, "loss": 0.0156, "step": 55010 }, { "grad_norm": 0.26445913314819336, "learning_rate": 2.4372020472997565e-05, "loss": 0.0137, "step": 55020 }, { "grad_norm": 0.4172627031803131, "learning_rate": 2.4354275730602122e-05, "loss": 0.0156, "step": 55030 }, { "grad_norm": 0.2250155210494995, "learning_rate": 2.433653537035958e-05, "loss": 0.015, "step": 55040 }, { "grad_norm": 0.3990301489830017, "learning_rate": 2.431879939530123e-05, "loss": 0.0147, "step": 55050 }, { "grad_norm": 0.18910792469978333, "learning_rate": 2.4301067808457684e-05, "loss": 0.0143, "step": 55060 }, { "grad_norm": 0.25123751163482666, "learning_rate": 2.428334061285878e-05, "loss": 0.0136, "step": 55070 }, { "grad_norm": 0.27549394965171814, "learning_rate": 2.426561781153361e-05, "loss": 0.0124, "step": 55080 }, { "grad_norm": 0.20776595175266266, "learning_rate": 2.424789940751052e-05, "loss": 0.0126, "step": 55090 }, { "grad_norm": 0.24029216170310974, "learning_rate": 2.423018540381712e-05, "loss": 0.0125, "step": 55100 }, { "grad_norm": 0.24462135136127472, "learning_rate": 2.421247580348021e-05, "loss": 0.015, "step": 55110 }, { "grad_norm": 0.20229770243167877, "learning_rate": 2.4194770609525897e-05, "loss": 0.0093, "step": 55120 }, { "grad_norm": 0.2003377228975296, "learning_rate": 2.417706982497951e-05, "loss": 0.0178, "step": 55130 }, { "grad_norm": 0.1766456514596939, "learning_rate": 2.415937345286563e-05, "loss": 0.0175, "step": 55140 }, { "grad_norm": 0.34683361649513245, "learning_rate": 2.4141681496208087e-05, "loss": 0.0115, "step": 55150 }, { "grad_norm": 0.2802715599536896, "learning_rate": 2.4123993958029946e-05, "loss": 0.0135, "step": 55160 }, { "grad_norm": 0.26336538791656494, "learning_rate": 2.4106310841353548e-05, "loss": 0.0133, "step": 55170 }, { "grad_norm": 0.14990635216236115, "learning_rate": 2.4088632149200398e-05, "loss": 0.0144, "step": 55180 }, { "grad_norm": 0.26110708713531494, "learning_rate": 2.4070957884591367e-05, "loss": 0.0127, "step": 55190 }, { "grad_norm": 0.540342390537262, "learning_rate": 2.4053288050546464e-05, "loss": 0.0176, "step": 55200 }, { "grad_norm": 0.29156023263931274, "learning_rate": 2.403562265008498e-05, "loss": 0.0148, "step": 55210 }, { "grad_norm": 0.4206583499908447, "learning_rate": 2.4017961686225483e-05, "loss": 0.0163, "step": 55220 }, { "grad_norm": 0.25694674253463745, "learning_rate": 2.400030516198568e-05, "loss": 0.0211, "step": 55230 }, { "grad_norm": 0.18542884290218353, "learning_rate": 2.3982653080382673e-05, "loss": 0.0114, "step": 55240 }, { "grad_norm": 0.24845023453235626, "learning_rate": 2.396500544443266e-05, "loss": 0.0107, "step": 55250 }, { "grad_norm": 0.4719460904598236, "learning_rate": 2.3947362257151156e-05, "loss": 0.01, "step": 55260 }, { "grad_norm": 0.45659273862838745, "learning_rate": 2.39297235215529e-05, "loss": 0.0188, "step": 55270 }, { "grad_norm": 0.23783068358898163, "learning_rate": 2.3912089240651873e-05, "loss": 0.0154, "step": 55280 }, { "grad_norm": 0.14146634936332703, "learning_rate": 2.389445941746129e-05, "loss": 0.0214, "step": 55290 }, { "grad_norm": 0.2508659064769745, "learning_rate": 2.3876834054993625e-05, "loss": 0.0135, "step": 55300 }, { "grad_norm": 0.23287473618984222, "learning_rate": 2.3859213156260522e-05, "loss": 0.0096, "step": 55310 }, { "grad_norm": 0.21400564908981323, "learning_rate": 2.3841596724272948e-05, "loss": 0.0152, "step": 55320 }, { "grad_norm": 0.21220357716083527, "learning_rate": 2.382398476204106e-05, "loss": 0.0132, "step": 55330 }, { "grad_norm": 0.19527696073055267, "learning_rate": 2.3806377272574254e-05, "loss": 0.0168, "step": 55340 }, { "grad_norm": 0.18433962762355804, "learning_rate": 2.3788774258881174e-05, "loss": 0.0118, "step": 55350 }, { "grad_norm": 0.3142447769641876, "learning_rate": 2.37711757239697e-05, "loss": 0.0124, "step": 55360 }, { "grad_norm": 0.417325496673584, "learning_rate": 2.3753581670846954e-05, "loss": 0.0141, "step": 55370 }, { "grad_norm": 0.14312313497066498, "learning_rate": 2.3735992102519216e-05, "loss": 0.0124, "step": 55380 }, { "grad_norm": 0.2758902311325073, "learning_rate": 2.371840702199215e-05, "loss": 0.0111, "step": 55390 }, { "grad_norm": 0.10497996211051941, "learning_rate": 2.3700826432270494e-05, "loss": 0.011, "step": 55400 }, { "grad_norm": 0.38902294635772705, "learning_rate": 2.3683250336358326e-05, "loss": 0.0179, "step": 55410 }, { "grad_norm": 0.13042490184307098, "learning_rate": 2.3665678737258923e-05, "loss": 0.0197, "step": 55420 }, { "grad_norm": 0.11739277839660645, "learning_rate": 2.3648111637974745e-05, "loss": 0.0127, "step": 55430 }, { "grad_norm": 0.2767407298088074, "learning_rate": 2.36305490415076e-05, "loss": 0.019, "step": 55440 }, { "grad_norm": 0.23709088563919067, "learning_rate": 2.3612990950858392e-05, "loss": 0.0092, "step": 55450 }, { "grad_norm": 0.20316018164157867, "learning_rate": 2.359543736902735e-05, "loss": 0.0121, "step": 55460 }, { "grad_norm": 0.20181791484355927, "learning_rate": 2.3577888299013896e-05, "loss": 0.0152, "step": 55470 }, { "grad_norm": 0.2870878577232361, "learning_rate": 2.3560343743816683e-05, "loss": 0.0141, "step": 55480 }, { "grad_norm": 0.24020136892795563, "learning_rate": 2.354280370643362e-05, "loss": 0.0182, "step": 55490 }, { "grad_norm": 0.17960606515407562, "learning_rate": 2.3525268189861777e-05, "loss": 0.0128, "step": 55500 }, { "grad_norm": 0.24021723866462708, "learning_rate": 2.3507737197097513e-05, "loss": 0.0091, "step": 55510 }, { "grad_norm": 0.24479208886623383, "learning_rate": 2.34902107311364e-05, "loss": 0.0104, "step": 55520 }, { "grad_norm": 0.20919956266880035, "learning_rate": 2.3472688794973225e-05, "loss": 0.0106, "step": 55530 }, { "grad_norm": 0.48459088802337646, "learning_rate": 2.3455171391602016e-05, "loss": 0.0145, "step": 55540 }, { "grad_norm": 0.34271541237831116, "learning_rate": 2.343765852401601e-05, "loss": 0.0099, "step": 55550 }, { "grad_norm": 0.10129711031913757, "learning_rate": 2.342015019520768e-05, "loss": 0.0132, "step": 55560 }, { "grad_norm": 0.23968692123889923, "learning_rate": 2.3402646408168742e-05, "loss": 0.0122, "step": 55570 }, { "grad_norm": 0.4723508954048157, "learning_rate": 2.3385147165890074e-05, "loss": 0.0197, "step": 55580 }, { "grad_norm": 0.1627665013074875, "learning_rate": 2.336765247136184e-05, "loss": 0.0102, "step": 55590 }, { "grad_norm": 0.3608945608139038, "learning_rate": 2.33501623275734e-05, "loss": 0.0191, "step": 55600 }, { "grad_norm": 0.15739335119724274, "learning_rate": 2.333267673751334e-05, "loss": 0.0148, "step": 55610 }, { "grad_norm": 0.12516403198242188, "learning_rate": 2.33151957041695e-05, "loss": 0.0151, "step": 55620 }, { "grad_norm": 0.21228915452957153, "learning_rate": 2.329771923052884e-05, "loss": 0.016, "step": 55630 }, { "grad_norm": 0.18529298901557922, "learning_rate": 2.3280247319577697e-05, "loss": 0.0148, "step": 55640 }, { "grad_norm": 0.26352718472480774, "learning_rate": 2.3262779974301473e-05, "loss": 0.0118, "step": 55650 }, { "grad_norm": 0.20442330837249756, "learning_rate": 2.3245317197684895e-05, "loss": 0.0107, "step": 55660 }, { "grad_norm": 0.20646589994430542, "learning_rate": 2.322785899271188e-05, "loss": 0.0122, "step": 55670 }, { "grad_norm": 0.1861828714609146, "learning_rate": 2.3210405362365507e-05, "loss": 0.0156, "step": 55680 }, { "grad_norm": 0.16570743918418884, "learning_rate": 2.31929563096282e-05, "loss": 0.0122, "step": 55690 }, { "grad_norm": 0.15928277373313904, "learning_rate": 2.317551183748146e-05, "loss": 0.0107, "step": 55700 }, { "grad_norm": 0.15664537250995636, "learning_rate": 2.3158071948906103e-05, "loss": 0.0113, "step": 55710 }, { "grad_norm": 0.08109377324581146, "learning_rate": 2.314063664688212e-05, "loss": 0.0083, "step": 55720 }, { "grad_norm": 0.3698611259460449, "learning_rate": 2.3123205934388725e-05, "loss": 0.0112, "step": 55730 }, { "grad_norm": 0.20325468480587006, "learning_rate": 2.310577981440436e-05, "loss": 0.0146, "step": 55740 }, { "grad_norm": 0.1650150716304779, "learning_rate": 2.3088358289906665e-05, "loss": 0.0123, "step": 55750 }, { "grad_norm": 0.2665189504623413, "learning_rate": 2.307094136387252e-05, "loss": 0.0139, "step": 55760 }, { "grad_norm": 0.23111768066883087, "learning_rate": 2.305352903927796e-05, "loss": 0.016, "step": 55770 }, { "grad_norm": 0.39978736639022827, "learning_rate": 2.303612131909831e-05, "loss": 0.0115, "step": 55780 }, { "grad_norm": 0.21253997087478638, "learning_rate": 2.3018718206308054e-05, "loss": 0.017, "step": 55790 }, { "grad_norm": 0.20404742658138275, "learning_rate": 2.3001319703880925e-05, "loss": 0.0143, "step": 55800 }, { "grad_norm": 0.2191387563943863, "learning_rate": 2.2983925814789835e-05, "loss": 0.0118, "step": 55810 }, { "grad_norm": 0.24583745002746582, "learning_rate": 2.2966536542006957e-05, "loss": 0.0128, "step": 55820 }, { "grad_norm": 0.20684508979320526, "learning_rate": 2.294915188850358e-05, "loss": 0.0125, "step": 55830 }, { "grad_norm": 0.2272505909204483, "learning_rate": 2.2931771857250333e-05, "loss": 0.0133, "step": 55840 }, { "grad_norm": 0.2591904401779175, "learning_rate": 2.2914396451216946e-05, "loss": 0.0117, "step": 55850 }, { "grad_norm": 0.265001118183136, "learning_rate": 2.2897025673372412e-05, "loss": 0.0126, "step": 55860 }, { "grad_norm": 0.24353833496570587, "learning_rate": 2.287965952668494e-05, "loss": 0.0136, "step": 55870 }, { "grad_norm": 0.3858806788921356, "learning_rate": 2.2862298014121873e-05, "loss": 0.015, "step": 55880 }, { "grad_norm": 0.26346680521965027, "learning_rate": 2.28449411386499e-05, "loss": 0.0191, "step": 55890 }, { "grad_norm": 0.3047417402267456, "learning_rate": 2.282758890323477e-05, "loss": 0.0133, "step": 55900 }, { "grad_norm": 0.17679312825202942, "learning_rate": 2.2810241310841528e-05, "loss": 0.0154, "step": 55910 }, { "grad_norm": 0.16464763879776, "learning_rate": 2.2792898364434412e-05, "loss": 0.0118, "step": 55920 }, { "grad_norm": 0.24828708171844482, "learning_rate": 2.2775560066976843e-05, "loss": 0.0154, "step": 55930 }, { "grad_norm": 0.30764254927635193, "learning_rate": 2.275822642143147e-05, "loss": 0.0184, "step": 55940 }, { "grad_norm": 0.1915614753961563, "learning_rate": 2.274089743076014e-05, "loss": 0.0135, "step": 55950 }, { "grad_norm": 0.19586040079593658, "learning_rate": 2.2723573097923917e-05, "loss": 0.0113, "step": 55960 }, { "grad_norm": 0.47331851720809937, "learning_rate": 2.2706253425883024e-05, "loss": 0.0124, "step": 55970 }, { "grad_norm": 0.28445741534233093, "learning_rate": 2.2688938417596933e-05, "loss": 0.0127, "step": 55980 }, { "grad_norm": 0.32884472608566284, "learning_rate": 2.2671628076024305e-05, "loss": 0.0147, "step": 55990 }, { "grad_norm": 0.2330811768770218, "learning_rate": 2.2654322404123008e-05, "loss": 0.0116, "step": 56000 }, { "grad_norm": 0.21847744286060333, "learning_rate": 2.2637021404850105e-05, "loss": 0.0118, "step": 56010 }, { "grad_norm": 0.2792873680591583, "learning_rate": 2.2619725081161885e-05, "loss": 0.0154, "step": 56020 }, { "grad_norm": 0.1251150220632553, "learning_rate": 2.260243343601376e-05, "loss": 0.0112, "step": 56030 }, { "grad_norm": 0.1326042115688324, "learning_rate": 2.2585146472360473e-05, "loss": 0.0118, "step": 56040 }, { "grad_norm": 0.3047316372394562, "learning_rate": 2.2567864193155834e-05, "loss": 0.0169, "step": 56050 }, { "grad_norm": 0.26868757605552673, "learning_rate": 2.255058660135294e-05, "loss": 0.0146, "step": 56060 }, { "grad_norm": 0.4351869523525238, "learning_rate": 2.253331369990407e-05, "loss": 0.0166, "step": 56070 }, { "grad_norm": 0.24198651313781738, "learning_rate": 2.251604549176063e-05, "loss": 0.0148, "step": 56080 }, { "grad_norm": 0.15942628681659698, "learning_rate": 2.249878197987337e-05, "loss": 0.0117, "step": 56090 }, { "grad_norm": 0.20888768136501312, "learning_rate": 2.2481523167192087e-05, "loss": 0.0113, "step": 56100 }, { "grad_norm": 0.4764288663864136, "learning_rate": 2.2464269056665867e-05, "loss": 0.0103, "step": 56110 }, { "grad_norm": 0.13728991150856018, "learning_rate": 2.2447019651242958e-05, "loss": 0.0141, "step": 56120 }, { "grad_norm": 0.17980796098709106, "learning_rate": 2.2429774953870818e-05, "loss": 0.0103, "step": 56130 }, { "grad_norm": 0.41054704785346985, "learning_rate": 2.241253496749611e-05, "loss": 0.0142, "step": 56140 }, { "grad_norm": 0.3630964159965515, "learning_rate": 2.2395299695064614e-05, "loss": 0.0126, "step": 56150 }, { "grad_norm": 0.2443217784166336, "learning_rate": 2.237806913952145e-05, "loss": 0.0121, "step": 56160 }, { "grad_norm": 0.16408982872962952, "learning_rate": 2.2360843303810798e-05, "loss": 0.0129, "step": 56170 }, { "grad_norm": 0.2671733498573303, "learning_rate": 2.2343622190876084e-05, "loss": 0.0111, "step": 56180 }, { "grad_norm": 0.155756413936615, "learning_rate": 2.2326405803659935e-05, "loss": 0.0091, "step": 56190 }, { "grad_norm": 0.2977259159088135, "learning_rate": 2.230919414510416e-05, "loss": 0.0113, "step": 56200 }, { "grad_norm": 0.34362438321113586, "learning_rate": 2.229198721814976e-05, "loss": 0.0108, "step": 56210 }, { "grad_norm": 0.2098422646522522, "learning_rate": 2.2274785025736948e-05, "loss": 0.0086, "step": 56220 }, { "grad_norm": 0.24057026207447052, "learning_rate": 2.225758757080507e-05, "loss": 0.015, "step": 56230 }, { "grad_norm": 0.239036425948143, "learning_rate": 2.2240394856292723e-05, "loss": 0.0116, "step": 56240 }, { "grad_norm": 0.23102936148643494, "learning_rate": 2.2223206885137664e-05, "loss": 0.0098, "step": 56250 }, { "grad_norm": 0.2758346498012543, "learning_rate": 2.2206023660276853e-05, "loss": 0.0136, "step": 56260 }, { "grad_norm": 0.40803682804107666, "learning_rate": 2.218884518464645e-05, "loss": 0.0142, "step": 56270 }, { "grad_norm": 0.40634283423423767, "learning_rate": 2.2171671461181732e-05, "loss": 0.0128, "step": 56280 }, { "grad_norm": 0.21895673871040344, "learning_rate": 2.2154502492817292e-05, "loss": 0.0137, "step": 56290 }, { "grad_norm": 0.1538853645324707, "learning_rate": 2.2137338282486782e-05, "loss": 0.0095, "step": 56300 }, { "grad_norm": 0.17031966149806976, "learning_rate": 2.2120178833123113e-05, "loss": 0.0108, "step": 56310 }, { "grad_norm": 0.144434854388237, "learning_rate": 2.210302414765838e-05, "loss": 0.0109, "step": 56320 }, { "grad_norm": 0.12987719476222992, "learning_rate": 2.2085874229023808e-05, "loss": 0.0095, "step": 56330 }, { "grad_norm": 0.22489267587661743, "learning_rate": 2.2068729080149907e-05, "loss": 0.0104, "step": 56340 }, { "grad_norm": 0.41594281792640686, "learning_rate": 2.205158870396625e-05, "loss": 0.0179, "step": 56350 }, { "grad_norm": 0.15989436209201813, "learning_rate": 2.2034453103401732e-05, "loss": 0.0097, "step": 56360 }, { "grad_norm": 0.17428156733512878, "learning_rate": 2.20173222813843e-05, "loss": 0.0088, "step": 56370 }, { "grad_norm": 0.1865798979997635, "learning_rate": 2.200019624084116e-05, "loss": 0.0138, "step": 56380 }, { "grad_norm": 0.20035456120967865, "learning_rate": 2.1983074984698687e-05, "loss": 0.0152, "step": 56390 }, { "grad_norm": 0.37268590927124023, "learning_rate": 2.1965958515882433e-05, "loss": 0.017, "step": 56400 }, { "grad_norm": 0.2025044709444046, "learning_rate": 2.1948846837317162e-05, "loss": 0.0122, "step": 56410 }, { "grad_norm": 0.23594903945922852, "learning_rate": 2.1931739951926738e-05, "loss": 0.0121, "step": 56420 }, { "grad_norm": 0.21276797354221344, "learning_rate": 2.191463786263429e-05, "loss": 0.0127, "step": 56430 }, { "grad_norm": 0.15079906582832336, "learning_rate": 2.1897540572362095e-05, "loss": 0.0125, "step": 56440 }, { "grad_norm": 0.23148545622825623, "learning_rate": 2.1880448084031614e-05, "loss": 0.0093, "step": 56450 }, { "grad_norm": 0.115655817091465, "learning_rate": 2.1863360400563482e-05, "loss": 0.013, "step": 56460 }, { "grad_norm": 0.22587010264396667, "learning_rate": 2.184627752487754e-05, "loss": 0.0124, "step": 56470 }, { "grad_norm": 0.15105217695236206, "learning_rate": 2.1829199459892725e-05, "loss": 0.0093, "step": 56480 }, { "grad_norm": 0.2462979406118393, "learning_rate": 2.1812126208527282e-05, "loss": 0.0116, "step": 56490 }, { "grad_norm": 0.14551545679569244, "learning_rate": 2.179505777369852e-05, "loss": 0.0142, "step": 56500 }, { "grad_norm": 0.11267606168985367, "learning_rate": 2.1777994158322974e-05, "loss": 0.0114, "step": 56510 }, { "grad_norm": 0.2085413783788681, "learning_rate": 2.176093536531637e-05, "loss": 0.015, "step": 56520 }, { "grad_norm": 0.25555869936943054, "learning_rate": 2.1743881397593537e-05, "loss": 0.0115, "step": 56530 }, { "grad_norm": 0.40645280480384827, "learning_rate": 2.1726832258068595e-05, "loss": 0.0238, "step": 56540 }, { "grad_norm": 0.40480396151542664, "learning_rate": 2.170978794965472e-05, "loss": 0.0125, "step": 56550 }, { "grad_norm": 0.2748659551143646, "learning_rate": 2.169274847526438e-05, "loss": 0.0104, "step": 56560 }, { "grad_norm": 0.28003039956092834, "learning_rate": 2.1675713837809103e-05, "loss": 0.0135, "step": 56570 }, { "grad_norm": 0.13390007615089417, "learning_rate": 2.1658684040199655e-05, "loss": 0.0107, "step": 56580 }, { "grad_norm": 0.19059590995311737, "learning_rate": 2.1641659085345974e-05, "loss": 0.0108, "step": 56590 }, { "grad_norm": 0.13662436604499817, "learning_rate": 2.1624638976157154e-05, "loss": 0.0135, "step": 56600 }, { "grad_norm": 0.22325581312179565, "learning_rate": 2.1607623715541476e-05, "loss": 0.0184, "step": 56610 }, { "grad_norm": 0.1922096312046051, "learning_rate": 2.159061330640636e-05, "loss": 0.0091, "step": 56620 }, { "grad_norm": 0.1746058613061905, "learning_rate": 2.1573607751658425e-05, "loss": 0.0129, "step": 56630 }, { "grad_norm": 0.1313355714082718, "learning_rate": 2.155660705420347e-05, "loss": 0.0129, "step": 56640 }, { "grad_norm": 0.17069339752197266, "learning_rate": 2.153961121694644e-05, "loss": 0.0151, "step": 56650 }, { "grad_norm": 0.15282022953033447, "learning_rate": 2.152262024279145e-05, "loss": 0.0161, "step": 56660 }, { "grad_norm": 0.46745190024375916, "learning_rate": 2.150563413464183e-05, "loss": 0.0152, "step": 56670 }, { "grad_norm": 0.12055493891239166, "learning_rate": 2.1488652895399974e-05, "loss": 0.0081, "step": 56680 }, { "grad_norm": 0.3054198920726776, "learning_rate": 2.1471676527967587e-05, "loss": 0.0114, "step": 56690 }, { "grad_norm": 0.20710936188697815, "learning_rate": 2.1454705035245414e-05, "loss": 0.0157, "step": 56700 }, { "grad_norm": 0.22420504689216614, "learning_rate": 2.143773842013343e-05, "loss": 0.0085, "step": 56710 }, { "grad_norm": 0.21852713823318481, "learning_rate": 2.1420776685530796e-05, "loss": 0.0104, "step": 56720 }, { "grad_norm": 0.23622193932533264, "learning_rate": 2.1403819834335742e-05, "loss": 0.0112, "step": 56730 }, { "grad_norm": 0.1963675618171692, "learning_rate": 2.1386867869445808e-05, "loss": 0.0104, "step": 56740 }, { "grad_norm": 0.10323062539100647, "learning_rate": 2.1369920793757548e-05, "loss": 0.0119, "step": 56750 }, { "grad_norm": 0.13609929382801056, "learning_rate": 2.1352978610166828e-05, "loss": 0.009, "step": 56760 }, { "grad_norm": 0.35750526189804077, "learning_rate": 2.1336041321568546e-05, "loss": 0.0116, "step": 56770 }, { "grad_norm": 0.18806836009025574, "learning_rate": 2.131910893085684e-05, "loss": 0.0102, "step": 56780 }, { "grad_norm": 0.14482958614826202, "learning_rate": 2.1302181440925012e-05, "loss": 0.0101, "step": 56790 }, { "grad_norm": 0.14744582772254944, "learning_rate": 2.128525885466546e-05, "loss": 0.0124, "step": 56800 }, { "grad_norm": 0.24171455204486847, "learning_rate": 2.1268341174969847e-05, "loss": 0.0152, "step": 56810 }, { "grad_norm": 0.2696917653083801, "learning_rate": 2.1251428404728894e-05, "loss": 0.0118, "step": 56820 }, { "grad_norm": 0.17661407589912415, "learning_rate": 2.123452054683256e-05, "loss": 0.0079, "step": 56830 }, { "grad_norm": 0.25985753536224365, "learning_rate": 2.1217617604169927e-05, "loss": 0.0154, "step": 56840 }, { "grad_norm": 0.2674267590045929, "learning_rate": 2.120071957962924e-05, "loss": 0.0126, "step": 56850 }, { "grad_norm": 0.18807627260684967, "learning_rate": 2.1183826476097917e-05, "loss": 0.0141, "step": 56860 }, { "grad_norm": 0.33591485023498535, "learning_rate": 2.116693829646254e-05, "loss": 0.0129, "step": 56870 }, { "grad_norm": 0.4341972768306732, "learning_rate": 2.1150055043608806e-05, "loss": 0.01, "step": 56880 }, { "grad_norm": 0.3433430790901184, "learning_rate": 2.113317672042162e-05, "loss": 0.0145, "step": 56890 }, { "grad_norm": 0.22596630454063416, "learning_rate": 2.1116303329785025e-05, "loss": 0.0103, "step": 56900 }, { "grad_norm": 0.3133322298526764, "learning_rate": 2.109943487458222e-05, "loss": 0.0124, "step": 56910 }, { "grad_norm": 0.1634858250617981, "learning_rate": 2.1082571357695574e-05, "loss": 0.0121, "step": 56920 }, { "grad_norm": 0.23890367150306702, "learning_rate": 2.1065712782006557e-05, "loss": 0.0124, "step": 56930 }, { "grad_norm": 0.25484758615493774, "learning_rate": 2.104885915039591e-05, "loss": 0.0112, "step": 56940 }, { "grad_norm": 0.21215704083442688, "learning_rate": 2.103201046574338e-05, "loss": 0.0092, "step": 56950 }, { "grad_norm": 0.2452547252178192, "learning_rate": 2.1015166730928022e-05, "loss": 0.0131, "step": 56960 }, { "grad_norm": 0.192989319562912, "learning_rate": 2.0998327948827912e-05, "loss": 0.0141, "step": 56970 }, { "grad_norm": 0.15653811395168304, "learning_rate": 2.0981494122320363e-05, "loss": 0.0131, "step": 56980 }, { "grad_norm": 0.27714240550994873, "learning_rate": 2.0964665254281822e-05, "loss": 0.0135, "step": 56990 }, { "grad_norm": 0.21150711178779602, "learning_rate": 2.094784134758784e-05, "loss": 0.0147, "step": 57000 }, { "grad_norm": 0.2732692062854767, "learning_rate": 2.0931022405113226e-05, "loss": 0.0209, "step": 57010 }, { "grad_norm": 0.33927786350250244, "learning_rate": 2.091420842973183e-05, "loss": 0.0125, "step": 57020 }, { "grad_norm": 0.22374220192432404, "learning_rate": 2.0897399424316715e-05, "loss": 0.0107, "step": 57030 }, { "grad_norm": 0.24528014659881592, "learning_rate": 2.0880595391740078e-05, "loss": 0.0109, "step": 57040 }, { "grad_norm": 0.18444909155368805, "learning_rate": 2.0863796334873277e-05, "loss": 0.0176, "step": 57050 }, { "grad_norm": 0.19685424864292145, "learning_rate": 2.08470022565868e-05, "loss": 0.0098, "step": 57060 }, { "grad_norm": 0.15786221623420715, "learning_rate": 2.0830213159750317e-05, "loss": 0.0097, "step": 57070 }, { "grad_norm": 0.23322655260562897, "learning_rate": 2.0813429047232596e-05, "loss": 0.0122, "step": 57080 }, { "grad_norm": 0.21492533385753632, "learning_rate": 2.0796649921901594e-05, "loss": 0.0132, "step": 57090 }, { "grad_norm": 0.22373345494270325, "learning_rate": 2.077987578662441e-05, "loss": 0.01, "step": 57100 }, { "grad_norm": 0.2318125218153, "learning_rate": 2.0763106644267277e-05, "loss": 0.0157, "step": 57110 }, { "grad_norm": 0.24336889386177063, "learning_rate": 2.0746342497695607e-05, "loss": 0.0092, "step": 57120 }, { "grad_norm": 0.25427570939064026, "learning_rate": 2.0729583349773886e-05, "loss": 0.0141, "step": 57130 }, { "grad_norm": 0.16653278470039368, "learning_rate": 2.0712829203365853e-05, "loss": 0.0121, "step": 57140 }, { "grad_norm": 0.1750716120004654, "learning_rate": 2.0696080061334267e-05, "loss": 0.0121, "step": 57150 }, { "grad_norm": 0.17003023624420166, "learning_rate": 2.067933592654117e-05, "loss": 0.0085, "step": 57160 }, { "grad_norm": 0.20314735174179077, "learning_rate": 2.066259680184763e-05, "loss": 0.0127, "step": 57170 }, { "grad_norm": 0.28172561526298523, "learning_rate": 2.0645862690113908e-05, "loss": 0.0195, "step": 57180 }, { "grad_norm": 0.3658161461353302, "learning_rate": 2.0629133594199436e-05, "loss": 0.0152, "step": 57190 }, { "grad_norm": 0.2111460268497467, "learning_rate": 2.0612409516962704e-05, "loss": 0.0134, "step": 57200 }, { "grad_norm": 0.3243809640407562, "learning_rate": 2.0595690461261467e-05, "loss": 0.0117, "step": 57210 }, { "grad_norm": 0.20923900604248047, "learning_rate": 2.0578976429952503e-05, "loss": 0.0097, "step": 57220 }, { "grad_norm": 0.22328698635101318, "learning_rate": 2.0562267425891802e-05, "loss": 0.0112, "step": 57230 }, { "grad_norm": 0.2668592631816864, "learning_rate": 2.0545563451934467e-05, "loss": 0.0127, "step": 57240 }, { "grad_norm": 0.14996367692947388, "learning_rate": 2.0528864510934764e-05, "loss": 0.0121, "step": 57250 }, { "grad_norm": 0.15008312463760376, "learning_rate": 2.0512170605746096e-05, "loss": 0.0143, "step": 57260 }, { "grad_norm": 0.1466095745563507, "learning_rate": 2.0495481739220963e-05, "loss": 0.0146, "step": 57270 }, { "grad_norm": 0.27776262164115906, "learning_rate": 2.0478797914211045e-05, "loss": 0.0129, "step": 57280 }, { "grad_norm": 0.1204892173409462, "learning_rate": 2.046211913356716e-05, "loss": 0.0154, "step": 57290 }, { "grad_norm": 0.17547020316123962, "learning_rate": 2.0445445400139247e-05, "loss": 0.0156, "step": 57300 }, { "grad_norm": 0.42586857080459595, "learning_rate": 2.0428776716776405e-05, "loss": 0.0115, "step": 57310 }, { "grad_norm": 0.15914291143417358, "learning_rate": 2.0412113086326856e-05, "loss": 0.0121, "step": 57320 }, { "grad_norm": 0.1828925609588623, "learning_rate": 2.0395454511637918e-05, "loss": 0.0096, "step": 57330 }, { "grad_norm": 0.20280204713344574, "learning_rate": 2.037880099555616e-05, "loss": 0.0123, "step": 57340 }, { "grad_norm": 0.1953071504831314, "learning_rate": 2.0362152540927144e-05, "loss": 0.0094, "step": 57350 }, { "grad_norm": 0.19004103541374207, "learning_rate": 2.0345509150595666e-05, "loss": 0.025, "step": 57360 }, { "grad_norm": 0.1701952964067459, "learning_rate": 2.0328870827405617e-05, "loss": 0.0156, "step": 57370 }, { "grad_norm": 0.19017507135868073, "learning_rate": 2.0312237574200043e-05, "loss": 0.0081, "step": 57380 }, { "grad_norm": 0.30479696393013, "learning_rate": 2.029560939382112e-05, "loss": 0.0111, "step": 57390 }, { "grad_norm": 0.18301835656166077, "learning_rate": 2.0278986289110097e-05, "loss": 0.0157, "step": 57400 }, { "grad_norm": 0.1470971703529358, "learning_rate": 2.0262368262907484e-05, "loss": 0.0088, "step": 57410 }, { "grad_norm": 0.2979868948459625, "learning_rate": 2.024575531805279e-05, "loss": 0.0127, "step": 57420 }, { "grad_norm": 0.14970694482326508, "learning_rate": 2.0229147457384735e-05, "loss": 0.0085, "step": 57430 }, { "grad_norm": 0.1633320450782776, "learning_rate": 2.0212544683741157e-05, "loss": 0.0159, "step": 57440 }, { "grad_norm": 0.36823970079421997, "learning_rate": 2.0195946999958976e-05, "loss": 0.0164, "step": 57450 }, { "grad_norm": 0.19159989058971405, "learning_rate": 2.017935440887434e-05, "loss": 0.0141, "step": 57460 }, { "grad_norm": 0.16897660493850708, "learning_rate": 2.0162766913322423e-05, "loss": 0.0111, "step": 57470 }, { "grad_norm": 0.35696911811828613, "learning_rate": 2.0146184516137588e-05, "loss": 0.0148, "step": 57480 }, { "grad_norm": 0.15111449360847473, "learning_rate": 2.012960722015332e-05, "loss": 0.0119, "step": 57490 }, { "grad_norm": 0.29632288217544556, "learning_rate": 2.0113035028202214e-05, "loss": 0.0129, "step": 57500 }, { "grad_norm": 0.26418113708496094, "learning_rate": 2.009646794311602e-05, "loss": 0.0122, "step": 57510 }, { "grad_norm": 0.16653430461883545, "learning_rate": 2.007990596772559e-05, "loss": 0.0151, "step": 57520 }, { "grad_norm": 0.1965745985507965, "learning_rate": 2.0063349104860923e-05, "loss": 0.0183, "step": 57530 }, { "grad_norm": 0.11537066847085953, "learning_rate": 2.0046797357351116e-05, "loss": 0.0183, "step": 57540 }, { "grad_norm": 0.23525908589363098, "learning_rate": 2.0030250728024412e-05, "loss": 0.0118, "step": 57550 }, { "grad_norm": 0.14238712191581726, "learning_rate": 2.001370921970819e-05, "loss": 0.008, "step": 57560 }, { "grad_norm": 0.15788598358631134, "learning_rate": 1.9997172835228932e-05, "loss": 0.0098, "step": 57570 }, { "grad_norm": 0.25432902574539185, "learning_rate": 1.9980641577412262e-05, "loss": 0.0123, "step": 57580 }, { "grad_norm": 0.3314925730228424, "learning_rate": 1.9964115449082925e-05, "loss": 0.0137, "step": 57590 }, { "grad_norm": 0.2456306368112564, "learning_rate": 1.9947594453064742e-05, "loss": 0.0157, "step": 57600 }, { "grad_norm": 0.3082168400287628, "learning_rate": 1.9931078592180774e-05, "loss": 0.0142, "step": 57610 }, { "grad_norm": 0.2652003765106201, "learning_rate": 1.9914567869253065e-05, "loss": 0.009, "step": 57620 }, { "grad_norm": 0.5789417624473572, "learning_rate": 1.989806228710287e-05, "loss": 0.0171, "step": 57630 }, { "grad_norm": 0.19832248985767365, "learning_rate": 1.9881561848550555e-05, "loss": 0.0112, "step": 57640 }, { "grad_norm": 0.2047560214996338, "learning_rate": 1.9865066556415544e-05, "loss": 0.0146, "step": 57650 }, { "grad_norm": 0.15667785704135895, "learning_rate": 1.98485764135165e-05, "loss": 0.011, "step": 57660 }, { "grad_norm": 0.111216239631176, "learning_rate": 1.983209142267109e-05, "loss": 0.0169, "step": 57670 }, { "grad_norm": 0.1649240404367447, "learning_rate": 1.9815611586696165e-05, "loss": 0.0099, "step": 57680 }, { "grad_norm": 0.16754141449928284, "learning_rate": 1.9799136908407667e-05, "loss": 0.0095, "step": 57690 }, { "grad_norm": 0.16763995587825775, "learning_rate": 1.9782667390620678e-05, "loss": 0.011, "step": 57700 }, { "grad_norm": 0.12913377583026886, "learning_rate": 1.976620303614939e-05, "loss": 0.0114, "step": 57710 }, { "grad_norm": 0.22939182817935944, "learning_rate": 1.9749743847807108e-05, "loss": 0.0109, "step": 57720 }, { "grad_norm": 0.2136606127023697, "learning_rate": 1.9733289828406272e-05, "loss": 0.0101, "step": 57730 }, { "grad_norm": 0.23966042697429657, "learning_rate": 1.9716840980758382e-05, "loss": 0.0119, "step": 57740 }, { "grad_norm": 0.17672386765480042, "learning_rate": 1.9700397307674134e-05, "loss": 0.0105, "step": 57750 }, { "grad_norm": 0.18205200135707855, "learning_rate": 1.968395881196328e-05, "loss": 0.0126, "step": 57760 }, { "grad_norm": 0.3932587802410126, "learning_rate": 1.966752549643473e-05, "loss": 0.0125, "step": 57770 }, { "grad_norm": 0.21246647834777832, "learning_rate": 1.965109736389647e-05, "loss": 0.0115, "step": 57780 }, { "grad_norm": 0.15552666783332825, "learning_rate": 1.9634674417155645e-05, "loss": 0.0088, "step": 57790 }, { "grad_norm": 0.2951981723308563, "learning_rate": 1.9618256659018434e-05, "loss": 0.0277, "step": 57800 }, { "grad_norm": 0.3591947853565216, "learning_rate": 1.9601844092290257e-05, "loss": 0.011, "step": 57810 }, { "grad_norm": 0.2869398295879364, "learning_rate": 1.9585436719775512e-05, "loss": 0.0088, "step": 57820 }, { "grad_norm": 0.13903240859508514, "learning_rate": 1.9569034544277793e-05, "loss": 0.0131, "step": 57830 }, { "grad_norm": 0.23397907614707947, "learning_rate": 1.9552637568599798e-05, "loss": 0.011, "step": 57840 }, { "grad_norm": 0.2624533772468567, "learning_rate": 1.953624579554327e-05, "loss": 0.0192, "step": 57850 }, { "grad_norm": 0.37909528613090515, "learning_rate": 1.951985922790918e-05, "loss": 0.0177, "step": 57860 }, { "grad_norm": 0.2585977017879486, "learning_rate": 1.9503477868497505e-05, "loss": 0.0124, "step": 57870 }, { "grad_norm": 0.3802746832370758, "learning_rate": 1.9487101720107375e-05, "loss": 0.0163, "step": 57880 }, { "grad_norm": 0.1960490196943283, "learning_rate": 1.9470730785537032e-05, "loss": 0.0122, "step": 57890 }, { "grad_norm": 0.3272843360900879, "learning_rate": 1.9454365067583823e-05, "loss": 0.0142, "step": 57900 }, { "grad_norm": 0.21107855439186096, "learning_rate": 1.9438004569044215e-05, "loss": 0.0114, "step": 57910 }, { "grad_norm": 0.18104298412799835, "learning_rate": 1.9421649292713724e-05, "loss": 0.0272, "step": 57920 }, { "grad_norm": 0.2079853117465973, "learning_rate": 1.9405299241387076e-05, "loss": 0.015, "step": 57930 }, { "grad_norm": 0.5091512203216553, "learning_rate": 1.9388954417858007e-05, "loss": 0.0138, "step": 57940 }, { "grad_norm": 0.3591064512729645, "learning_rate": 1.9372614824919417e-05, "loss": 0.014, "step": 57950 }, { "grad_norm": 0.24777336418628693, "learning_rate": 1.9356280465363284e-05, "loss": 0.0142, "step": 57960 }, { "grad_norm": 0.16829147934913635, "learning_rate": 1.9339951341980723e-05, "loss": 0.0105, "step": 57970 }, { "grad_norm": 0.5655989050865173, "learning_rate": 1.9323627457561916e-05, "loss": 0.0113, "step": 57980 }, { "grad_norm": 0.2909514307975769, "learning_rate": 1.9307308814896198e-05, "loss": 0.0108, "step": 57990 }, { "grad_norm": 0.1747969686985016, "learning_rate": 1.9290995416771935e-05, "loss": 0.0116, "step": 58000 }, { "grad_norm": 0.2592681050300598, "learning_rate": 1.9274687265976665e-05, "loss": 0.0138, "step": 58010 }, { "grad_norm": 0.17456236481666565, "learning_rate": 1.9258384365297e-05, "loss": 0.0117, "step": 58020 }, { "grad_norm": 0.19068722426891327, "learning_rate": 1.924208671751866e-05, "loss": 0.0126, "step": 58030 }, { "grad_norm": 0.3124861717224121, "learning_rate": 1.9225794325426492e-05, "loss": 0.0117, "step": 58040 }, { "grad_norm": 0.14288625121116638, "learning_rate": 1.920950719180436e-05, "loss": 0.0111, "step": 58050 }, { "grad_norm": 0.14336737990379333, "learning_rate": 1.919322531943536e-05, "loss": 0.0096, "step": 58060 }, { "grad_norm": 0.12372030317783356, "learning_rate": 1.917694871110157e-05, "loss": 0.011, "step": 58070 }, { "grad_norm": 0.2339693158864975, "learning_rate": 1.9160677369584234e-05, "loss": 0.0124, "step": 58080 }, { "grad_norm": 0.24324867129325867, "learning_rate": 1.9144411297663694e-05, "loss": 0.0141, "step": 58090 }, { "grad_norm": 0.22032563388347626, "learning_rate": 1.9128150498119328e-05, "loss": 0.0145, "step": 58100 }, { "grad_norm": 0.26876354217529297, "learning_rate": 1.9111894973729726e-05, "loss": 0.0113, "step": 58110 }, { "grad_norm": 0.45590144395828247, "learning_rate": 1.9095644727272454e-05, "loss": 0.0129, "step": 58120 }, { "grad_norm": 0.26863229274749756, "learning_rate": 1.907939976152429e-05, "loss": 0.0136, "step": 58130 }, { "grad_norm": 0.2562280297279358, "learning_rate": 1.906316007926101e-05, "loss": 0.0151, "step": 58140 }, { "grad_norm": 0.18852809071540833, "learning_rate": 1.904692568325755e-05, "loss": 0.0107, "step": 58150 }, { "grad_norm": 0.3088378310203552, "learning_rate": 1.9030696576287925e-05, "loss": 0.0137, "step": 58160 }, { "grad_norm": 0.2210724800825119, "learning_rate": 1.9014472761125242e-05, "loss": 0.0123, "step": 58170 }, { "grad_norm": 0.15099754929542542, "learning_rate": 1.899825424054172e-05, "loss": 0.0109, "step": 58180 }, { "grad_norm": 0.16562409698963165, "learning_rate": 1.898204101730863e-05, "loss": 0.0129, "step": 58190 }, { "grad_norm": 0.18619300425052643, "learning_rate": 1.8965833094196394e-05, "loss": 0.0187, "step": 58200 }, { "grad_norm": 0.22153633832931519, "learning_rate": 1.8949630473974495e-05, "loss": 0.0116, "step": 58210 }, { "grad_norm": 0.24302594363689423, "learning_rate": 1.8933433159411517e-05, "loss": 0.0126, "step": 58220 }, { "grad_norm": 0.2388579398393631, "learning_rate": 1.891724115327514e-05, "loss": 0.0121, "step": 58230 }, { "grad_norm": 0.26985713839530945, "learning_rate": 1.8901054458332152e-05, "loss": 0.0156, "step": 58240 }, { "grad_norm": 0.21393923461437225, "learning_rate": 1.8884873077348364e-05, "loss": 0.0131, "step": 58250 }, { "grad_norm": 0.19772256910800934, "learning_rate": 1.88686970130888e-05, "loss": 0.0124, "step": 58260 }, { "grad_norm": 0.23927351832389832, "learning_rate": 1.8852526268317456e-05, "loss": 0.0121, "step": 58270 }, { "grad_norm": 0.24251513183116913, "learning_rate": 1.883636084579749e-05, "loss": 0.0094, "step": 58280 }, { "grad_norm": 0.3091583847999573, "learning_rate": 1.8820200748291145e-05, "loss": 0.0136, "step": 58290 }, { "grad_norm": 0.21396297216415405, "learning_rate": 1.8804045978559686e-05, "loss": 0.0116, "step": 58300 }, { "grad_norm": 0.19223517179489136, "learning_rate": 1.8787896539363594e-05, "loss": 0.0099, "step": 58310 }, { "grad_norm": 0.12169840931892395, "learning_rate": 1.8771752433462298e-05, "loss": 0.0111, "step": 58320 }, { "grad_norm": 0.14427192509174347, "learning_rate": 1.8755613663614447e-05, "loss": 0.0133, "step": 58330 }, { "grad_norm": 0.33265233039855957, "learning_rate": 1.873948023257767e-05, "loss": 0.0121, "step": 58340 }, { "grad_norm": 0.20556005835533142, "learning_rate": 1.8723352143108747e-05, "loss": 0.0086, "step": 58350 }, { "grad_norm": 0.4101094603538513, "learning_rate": 1.870722939796352e-05, "loss": 0.0145, "step": 58360 }, { "grad_norm": 0.34123098850250244, "learning_rate": 1.869111199989693e-05, "loss": 0.0132, "step": 58370 }, { "grad_norm": 0.3868182599544525, "learning_rate": 1.8674999951663018e-05, "loss": 0.0112, "step": 58380 }, { "grad_norm": 0.3059656620025635, "learning_rate": 1.8658893256014857e-05, "loss": 0.0132, "step": 58390 }, { "grad_norm": 0.1593269556760788, "learning_rate": 1.8642791915704655e-05, "loss": 0.0107, "step": 58400 }, { "grad_norm": 0.23046629130840302, "learning_rate": 1.8626695933483697e-05, "loss": 0.0117, "step": 58410 }, { "grad_norm": 0.18330147862434387, "learning_rate": 1.8610605312102335e-05, "loss": 0.0105, "step": 58420 }, { "grad_norm": 0.2720387279987335, "learning_rate": 1.8594520054310032e-05, "loss": 0.0186, "step": 58430 }, { "grad_norm": 0.15941178798675537, "learning_rate": 1.8578440162855326e-05, "loss": 0.0117, "step": 58440 }, { "grad_norm": 0.25982969999313354, "learning_rate": 1.8562365640485774e-05, "loss": 0.0142, "step": 58450 }, { "grad_norm": 0.13824954628944397, "learning_rate": 1.854629648994815e-05, "loss": 0.0083, "step": 58460 }, { "grad_norm": 0.29373371601104736, "learning_rate": 1.8530232713988183e-05, "loss": 0.012, "step": 58470 }, { "grad_norm": 0.36876434087753296, "learning_rate": 1.851417431535074e-05, "loss": 0.0134, "step": 58480 }, { "grad_norm": 0.15127378702163696, "learning_rate": 1.8498121296779785e-05, "loss": 0.0174, "step": 58490 }, { "grad_norm": 0.2071746438741684, "learning_rate": 1.848207366101829e-05, "loss": 0.0143, "step": 58500 }, { "grad_norm": 0.34029218554496765, "learning_rate": 1.8466031410808422e-05, "loss": 0.0146, "step": 58510 }, { "grad_norm": 0.3089078664779663, "learning_rate": 1.844999454889129e-05, "loss": 0.0166, "step": 58520 }, { "grad_norm": 0.5170050859451294, "learning_rate": 1.8433963078007237e-05, "loss": 0.0145, "step": 58530 }, { "grad_norm": 0.0997048020362854, "learning_rate": 1.841793700089554e-05, "loss": 0.0086, "step": 58540 }, { "grad_norm": 0.28719717264175415, "learning_rate": 1.8401916320294644e-05, "loss": 0.0102, "step": 58550 }, { "grad_norm": 0.30971765518188477, "learning_rate": 1.838590103894205e-05, "loss": 0.0139, "step": 58560 }, { "grad_norm": 0.23528876900672913, "learning_rate": 1.8369891159574292e-05, "loss": 0.0101, "step": 58570 }, { "grad_norm": 0.16939705610275269, "learning_rate": 1.835388668492708e-05, "loss": 0.0139, "step": 58580 }, { "grad_norm": 0.24785444140434265, "learning_rate": 1.8337887617735095e-05, "loss": 0.0156, "step": 58590 }, { "grad_norm": 0.17333631217479706, "learning_rate": 1.8321893960732157e-05, "loss": 0.0114, "step": 58600 }, { "grad_norm": 0.29107582569122314, "learning_rate": 1.8305905716651138e-05, "loss": 0.0119, "step": 58610 }, { "grad_norm": 0.14042770862579346, "learning_rate": 1.8289922888224e-05, "loss": 0.0092, "step": 58620 }, { "grad_norm": 0.2792722284793854, "learning_rate": 1.8273945478181765e-05, "loss": 0.0117, "step": 58630 }, { "grad_norm": 0.25681498646736145, "learning_rate": 1.8257973489254558e-05, "loss": 0.0128, "step": 58640 }, { "grad_norm": 0.3242012858390808, "learning_rate": 1.824200692417152e-05, "loss": 0.0146, "step": 58650 }, { "grad_norm": 0.2374386340379715, "learning_rate": 1.8226045785660912e-05, "loss": 0.0104, "step": 58660 }, { "grad_norm": 0.2766285240650177, "learning_rate": 1.821009007645006e-05, "loss": 0.013, "step": 58670 }, { "grad_norm": 0.17927230894565582, "learning_rate": 1.8194139799265357e-05, "loss": 0.0107, "step": 58680 }, { "grad_norm": 0.15715067088603973, "learning_rate": 1.8178194956832295e-05, "loss": 0.0134, "step": 58690 }, { "grad_norm": 0.10226219892501831, "learning_rate": 1.8162255551875346e-05, "loss": 0.0073, "step": 58700 }, { "grad_norm": 0.18923139572143555, "learning_rate": 1.8146321587118194e-05, "loss": 0.013, "step": 58710 }, { "grad_norm": 0.17092861235141754, "learning_rate": 1.8130393065283448e-05, "loss": 0.0088, "step": 58720 }, { "grad_norm": 0.20877772569656372, "learning_rate": 1.8114469989092925e-05, "loss": 0.0085, "step": 58730 }, { "grad_norm": 0.34070727229118347, "learning_rate": 1.8098552361267397e-05, "loss": 0.0137, "step": 58740 }, { "grad_norm": 0.2448006123304367, "learning_rate": 1.8082640184526763e-05, "loss": 0.0136, "step": 58750 }, { "grad_norm": 0.17014215886592865, "learning_rate": 1.806673346158999e-05, "loss": 0.0146, "step": 58760 }, { "grad_norm": 0.18829794228076935, "learning_rate": 1.8050832195175067e-05, "loss": 0.016, "step": 58770 }, { "grad_norm": 0.2261200100183487, "learning_rate": 1.8034936387999136e-05, "loss": 0.0133, "step": 58780 }, { "grad_norm": 0.2827647030353546, "learning_rate": 1.8019046042778315e-05, "loss": 0.0127, "step": 58790 }, { "grad_norm": 0.36810779571533203, "learning_rate": 1.800316116222785e-05, "loss": 0.0164, "step": 58800 }, { "grad_norm": 0.19937250018119812, "learning_rate": 1.7987281749062018e-05, "loss": 0.024, "step": 58810 }, { "grad_norm": 0.13309621810913086, "learning_rate": 1.7971407805994195e-05, "loss": 0.0177, "step": 58820 }, { "grad_norm": 0.1529984027147293, "learning_rate": 1.7955539335736787e-05, "loss": 0.0114, "step": 58830 }, { "grad_norm": 0.3018408715724945, "learning_rate": 1.7939676341001304e-05, "loss": 0.0196, "step": 58840 }, { "grad_norm": 0.27200087904930115, "learning_rate": 1.7923818824498275e-05, "loss": 0.0119, "step": 58850 }, { "grad_norm": 0.24186132848262787, "learning_rate": 1.7907966788937315e-05, "loss": 0.0179, "step": 58860 }, { "grad_norm": 0.2097419947385788, "learning_rate": 1.7892120237027116e-05, "loss": 0.0122, "step": 58870 }, { "grad_norm": 0.1112714409828186, "learning_rate": 1.7876279171475413e-05, "loss": 0.0137, "step": 58880 }, { "grad_norm": 0.289659321308136, "learning_rate": 1.7860443594989028e-05, "loss": 0.0121, "step": 58890 }, { "grad_norm": 0.32050710916519165, "learning_rate": 1.784461351027379e-05, "loss": 0.0143, "step": 58900 }, { "grad_norm": 0.24217909574508667, "learning_rate": 1.7828788920034677e-05, "loss": 0.0145, "step": 58910 }, { "grad_norm": 0.2919289469718933, "learning_rate": 1.7812969826975623e-05, "loss": 0.0134, "step": 58920 }, { "grad_norm": 0.10621625185012817, "learning_rate": 1.7797156233799738e-05, "loss": 0.0131, "step": 58930 }, { "grad_norm": 0.28436073660850525, "learning_rate": 1.7781348143209094e-05, "loss": 0.0142, "step": 58940 }, { "grad_norm": 0.21159976720809937, "learning_rate": 1.7765545557904862e-05, "loss": 0.0182, "step": 58950 }, { "grad_norm": 0.31353598833084106, "learning_rate": 1.7749748480587302e-05, "loss": 0.0143, "step": 58960 }, { "grad_norm": 0.20598022639751434, "learning_rate": 1.773395691395564e-05, "loss": 0.0119, "step": 58970 }, { "grad_norm": 0.15806879103183746, "learning_rate": 1.7718170860708305e-05, "loss": 0.0128, "step": 58980 }, { "grad_norm": 0.2884369492530823, "learning_rate": 1.770239032354264e-05, "loss": 0.014, "step": 58990 }, { "grad_norm": 0.2697107791900635, "learning_rate": 1.7686615305155124e-05, "loss": 0.0207, "step": 59000 }, { "grad_norm": 0.27119797468185425, "learning_rate": 1.767084580824128e-05, "loss": 0.0132, "step": 59010 }, { "grad_norm": 0.6258769035339355, "learning_rate": 1.765508183549569e-05, "loss": 0.0161, "step": 59020 }, { "grad_norm": 0.26971113681793213, "learning_rate": 1.763932338961199e-05, "loss": 0.016, "step": 59030 }, { "grad_norm": 0.15312156081199646, "learning_rate": 1.762357047328284e-05, "loss": 0.0092, "step": 59040 }, { "grad_norm": 0.43326643109321594, "learning_rate": 1.7607823089199997e-05, "loss": 0.0173, "step": 59050 }, { "grad_norm": 0.27319881319999695, "learning_rate": 1.7592081240054265e-05, "loss": 0.0137, "step": 59060 }, { "grad_norm": 0.17101933062076569, "learning_rate": 1.7576344928535488e-05, "loss": 0.0141, "step": 59070 }, { "grad_norm": 0.24703602492809296, "learning_rate": 1.7560614157332573e-05, "loss": 0.0108, "step": 59080 }, { "grad_norm": 0.15329544246196747, "learning_rate": 1.7544888929133495e-05, "loss": 0.0127, "step": 59090 }, { "grad_norm": 0.14338108897209167, "learning_rate": 1.752916924662522e-05, "loss": 0.0097, "step": 59100 }, { "grad_norm": 0.21823656558990479, "learning_rate": 1.751345511249387e-05, "loss": 0.0114, "step": 59110 }, { "grad_norm": 0.17998212575912476, "learning_rate": 1.749774652942452e-05, "loss": 0.0081, "step": 59120 }, { "grad_norm": 0.19853748381137848, "learning_rate": 1.748204350010135e-05, "loss": 0.0112, "step": 59130 }, { "grad_norm": 0.28485092520713806, "learning_rate": 1.7466346027207574e-05, "loss": 0.0098, "step": 59140 }, { "grad_norm": 0.1776471734046936, "learning_rate": 1.745065411342547e-05, "loss": 0.0083, "step": 59150 }, { "grad_norm": 0.138381227850914, "learning_rate": 1.7434967761436366e-05, "loss": 0.0128, "step": 59160 }, { "grad_norm": 0.25611793994903564, "learning_rate": 1.741928697392058e-05, "loss": 0.0111, "step": 59170 }, { "grad_norm": 0.3760770857334137, "learning_rate": 1.7403611753557597e-05, "loss": 0.0102, "step": 59180 }, { "grad_norm": 0.1729700267314911, "learning_rate": 1.738794210302584e-05, "loss": 0.011, "step": 59190 }, { "grad_norm": 0.2851274907588959, "learning_rate": 1.7372278025002837e-05, "loss": 0.0097, "step": 59200 }, { "grad_norm": 0.4797000288963318, "learning_rate": 1.7356619522165164e-05, "loss": 0.0123, "step": 59210 }, { "grad_norm": 0.3915528655052185, "learning_rate": 1.7340966597188378e-05, "loss": 0.0117, "step": 59220 }, { "grad_norm": 0.15228009223937988, "learning_rate": 1.732531925274722e-05, "loss": 0.0128, "step": 59230 }, { "grad_norm": 0.13822795450687408, "learning_rate": 1.7309677491515318e-05, "loss": 0.0079, "step": 59240 }, { "grad_norm": 0.20349258184432983, "learning_rate": 1.7294041316165455e-05, "loss": 0.0095, "step": 59250 }, { "grad_norm": 0.23058725893497467, "learning_rate": 1.727841072936942e-05, "loss": 0.0142, "step": 59260 }, { "grad_norm": 0.30692654848098755, "learning_rate": 1.7262785733798058e-05, "loss": 0.0119, "step": 59270 }, { "grad_norm": 0.37476232647895813, "learning_rate": 1.7247166332121246e-05, "loss": 0.0102, "step": 59280 }, { "grad_norm": 0.23788747191429138, "learning_rate": 1.7231552527007933e-05, "loss": 0.0102, "step": 59290 }, { "grad_norm": 0.12505240738391876, "learning_rate": 1.721594432112606e-05, "loss": 0.0094, "step": 59300 }, { "grad_norm": 0.1540588140487671, "learning_rate": 1.7200341717142655e-05, "loss": 0.0132, "step": 59310 }, { "grad_norm": 0.3101484179496765, "learning_rate": 1.718474471772377e-05, "loss": 0.0113, "step": 59320 }, { "grad_norm": 0.15722137689590454, "learning_rate": 1.7169153325534528e-05, "loss": 0.011, "step": 59330 }, { "grad_norm": 0.12066444754600525, "learning_rate": 1.7153567543239045e-05, "loss": 0.011, "step": 59340 }, { "grad_norm": 0.1880682110786438, "learning_rate": 1.7137987373500525e-05, "loss": 0.01, "step": 59350 }, { "grad_norm": 0.21294470131397247, "learning_rate": 1.7122412818981198e-05, "loss": 0.0116, "step": 59360 }, { "grad_norm": 0.17174983024597168, "learning_rate": 1.7106843882342283e-05, "loss": 0.014, "step": 59370 }, { "grad_norm": 0.1250055879354477, "learning_rate": 1.709128056624415e-05, "loss": 0.0118, "step": 59380 }, { "grad_norm": 0.20717397332191467, "learning_rate": 1.70757228733461e-05, "loss": 0.009, "step": 59390 }, { "grad_norm": 0.16196870803833008, "learning_rate": 1.706017080630653e-05, "loss": 0.0113, "step": 59400 }, { "grad_norm": 0.22233077883720398, "learning_rate": 1.7044624367782873e-05, "loss": 0.011, "step": 59410 }, { "grad_norm": 0.2080588936805725, "learning_rate": 1.7029083560431553e-05, "loss": 0.0114, "step": 59420 }, { "grad_norm": 0.3928042948246002, "learning_rate": 1.701354838690813e-05, "loss": 0.0178, "step": 59430 }, { "grad_norm": 0.1263154298067093, "learning_rate": 1.6998018849867087e-05, "loss": 0.0097, "step": 59440 }, { "grad_norm": 0.30018922686576843, "learning_rate": 1.698249495196202e-05, "loss": 0.0138, "step": 59450 }, { "grad_norm": 0.16093915700912476, "learning_rate": 1.6966976695845528e-05, "loss": 0.0118, "step": 59460 }, { "grad_norm": 0.1864604949951172, "learning_rate": 1.6951464084169268e-05, "loss": 0.0104, "step": 59470 }, { "grad_norm": 0.17955650389194489, "learning_rate": 1.6935957119583916e-05, "loss": 0.0106, "step": 59480 }, { "grad_norm": 0.20760658383369446, "learning_rate": 1.6920455804739205e-05, "loss": 0.012, "step": 59490 }, { "grad_norm": 0.22861771285533905, "learning_rate": 1.6904960142283855e-05, "loss": 0.0112, "step": 59500 }, { "grad_norm": 0.08930157870054245, "learning_rate": 1.6889470134865666e-05, "loss": 0.0084, "step": 59510 }, { "grad_norm": 0.11655410379171371, "learning_rate": 1.687398578513145e-05, "loss": 0.0129, "step": 59520 }, { "grad_norm": 0.19253133237361908, "learning_rate": 1.6858507095727066e-05, "loss": 0.0086, "step": 59530 }, { "grad_norm": 0.2119349241256714, "learning_rate": 1.6843034069297403e-05, "loss": 0.0109, "step": 59540 }, { "grad_norm": 0.31774336099624634, "learning_rate": 1.682756670848637e-05, "loss": 0.0202, "step": 59550 }, { "grad_norm": 0.13737282156944275, "learning_rate": 1.6812105015936937e-05, "loss": 0.0091, "step": 59560 }, { "grad_norm": 0.24317730963230133, "learning_rate": 1.6796648994291027e-05, "loss": 0.0187, "step": 59570 }, { "grad_norm": 0.19022758305072784, "learning_rate": 1.678119864618973e-05, "loss": 0.0099, "step": 59580 }, { "grad_norm": 0.14692890644073486, "learning_rate": 1.6765753974273023e-05, "loss": 0.0082, "step": 59590 }, { "grad_norm": 0.059448160231113434, "learning_rate": 1.675031498118001e-05, "loss": 0.0088, "step": 59600 }, { "grad_norm": 0.1801087111234665, "learning_rate": 1.6734881669548804e-05, "loss": 0.01, "step": 59610 }, { "grad_norm": 0.24630926549434662, "learning_rate": 1.671945404201648e-05, "loss": 0.015, "step": 59620 }, { "grad_norm": 0.14728744328022003, "learning_rate": 1.670403210121927e-05, "loss": 0.0085, "step": 59630 }, { "grad_norm": 0.23419161140918732, "learning_rate": 1.6688615849792312e-05, "loss": 0.0082, "step": 59640 }, { "grad_norm": 0.3707854449748993, "learning_rate": 1.6673205290369832e-05, "loss": 0.0146, "step": 59650 }, { "grad_norm": 0.21937404572963715, "learning_rate": 1.6657800425585086e-05, "loss": 0.0141, "step": 59660 }, { "grad_norm": 0.19181562960147858, "learning_rate": 1.6642401258070327e-05, "loss": 0.0185, "step": 59670 }, { "grad_norm": 0.22802405059337616, "learning_rate": 1.6627007790456884e-05, "loss": 0.0178, "step": 59680 }, { "grad_norm": 0.175308495759964, "learning_rate": 1.6611620025375037e-05, "loss": 0.0125, "step": 59690 }, { "grad_norm": 0.17750154435634613, "learning_rate": 1.6596237965454154e-05, "loss": 0.0193, "step": 59700 }, { "grad_norm": 0.3139500617980957, "learning_rate": 1.6580861613322607e-05, "loss": 0.0149, "step": 59710 }, { "grad_norm": 0.22256821393966675, "learning_rate": 1.6565490971607796e-05, "loss": 0.0078, "step": 59720 }, { "grad_norm": 0.3151369094848633, "learning_rate": 1.655012604293614e-05, "loss": 0.0128, "step": 59730 }, { "grad_norm": 0.14381392300128937, "learning_rate": 1.653476682993309e-05, "loss": 0.0119, "step": 59740 }, { "grad_norm": 0.18566304445266724, "learning_rate": 1.651941333522311e-05, "loss": 0.0113, "step": 59750 }, { "grad_norm": 0.3122856914997101, "learning_rate": 1.6504065561429715e-05, "loss": 0.0122, "step": 59760 }, { "grad_norm": 0.11517348885536194, "learning_rate": 1.6488723511175385e-05, "loss": 0.0156, "step": 59770 }, { "grad_norm": 0.24073584377765656, "learning_rate": 1.6473387187081668e-05, "loss": 0.0108, "step": 59780 }, { "grad_norm": 0.15385201573371887, "learning_rate": 1.6458056591769123e-05, "loss": 0.0145, "step": 59790 }, { "grad_norm": 0.18020182847976685, "learning_rate": 1.6442731727857335e-05, "loss": 0.0117, "step": 59800 }, { "grad_norm": 0.20601822435855865, "learning_rate": 1.642741259796492e-05, "loss": 0.0081, "step": 59810 }, { "grad_norm": 0.26462334394454956, "learning_rate": 1.641209920470944e-05, "loss": 0.0126, "step": 59820 }, { "grad_norm": 0.19061316549777985, "learning_rate": 1.639679155070762e-05, "loss": 0.0092, "step": 59830 }, { "grad_norm": 0.18707123398780823, "learning_rate": 1.6381489638575048e-05, "loss": 0.0088, "step": 59840 }, { "grad_norm": 0.36880794167518616, "learning_rate": 1.636619347092643e-05, "loss": 0.012, "step": 59850 }, { "grad_norm": 0.3038088381290436, "learning_rate": 1.6350903050375476e-05, "loss": 0.0114, "step": 59860 }, { "grad_norm": 0.273990660905838, "learning_rate": 1.6335618379534856e-05, "loss": 0.013, "step": 59870 }, { "grad_norm": 0.2753737270832062, "learning_rate": 1.6320339461016364e-05, "loss": 0.0099, "step": 59880 }, { "grad_norm": 0.11974037438631058, "learning_rate": 1.6305066297430687e-05, "loss": 0.0106, "step": 59890 }, { "grad_norm": 0.145733043551445, "learning_rate": 1.6289798891387654e-05, "loss": 0.0108, "step": 59900 }, { "grad_norm": 0.21280264854431152, "learning_rate": 1.6274537245495995e-05, "loss": 0.0096, "step": 59910 }, { "grad_norm": 0.2397209107875824, "learning_rate": 1.6259281362363527e-05, "loss": 0.0099, "step": 59920 }, { "grad_norm": 0.20264080166816711, "learning_rate": 1.6244031244597068e-05, "loss": 0.0151, "step": 59930 }, { "grad_norm": 0.21879445016384125, "learning_rate": 1.622878689480244e-05, "loss": 0.0089, "step": 59940 }, { "grad_norm": 0.27528321743011475, "learning_rate": 1.6213548315584498e-05, "loss": 0.0112, "step": 59950 }, { "grad_norm": 0.1710326075553894, "learning_rate": 1.6198315509547074e-05, "loss": 0.0134, "step": 59960 }, { "grad_norm": 0.1999705284833908, "learning_rate": 1.6183088479293056e-05, "loss": 0.0088, "step": 59970 }, { "grad_norm": 0.16431353986263275, "learning_rate": 1.6167867227424316e-05, "loss": 0.0115, "step": 59980 }, { "grad_norm": 0.14772827923297882, "learning_rate": 1.615265175654176e-05, "loss": 0.0131, "step": 59990 }, { "grad_norm": 0.22914862632751465, "learning_rate": 1.613744206924529e-05, "loss": 0.0103, "step": 60000 }, { "grad_norm": 0.15730997920036316, "learning_rate": 1.6122238168133845e-05, "loss": 0.0091, "step": 60010 }, { "grad_norm": 0.15458446741104126, "learning_rate": 1.6107040055805305e-05, "loss": 0.0139, "step": 60020 }, { "grad_norm": 0.24445699155330658, "learning_rate": 1.6091847734856673e-05, "loss": 0.0116, "step": 60030 }, { "grad_norm": 0.26319801807403564, "learning_rate": 1.6076661207883865e-05, "loss": 0.0103, "step": 60040 }, { "grad_norm": 0.1577160805463791, "learning_rate": 1.6061480477481848e-05, "loss": 0.0101, "step": 60050 }, { "grad_norm": 0.19959260523319244, "learning_rate": 1.604630554624461e-05, "loss": 0.0098, "step": 60060 }, { "grad_norm": 0.3284403085708618, "learning_rate": 1.603113641676509e-05, "loss": 0.0126, "step": 60070 }, { "grad_norm": 0.14509811997413635, "learning_rate": 1.6015973091635338e-05, "loss": 0.0117, "step": 60080 }, { "grad_norm": 0.12376511842012405, "learning_rate": 1.6000815573446288e-05, "loss": 0.0082, "step": 60090 }, { "grad_norm": 0.3278922140598297, "learning_rate": 1.5985663864788002e-05, "loss": 0.0139, "step": 60100 }, { "grad_norm": 0.14498847723007202, "learning_rate": 1.597051796824946e-05, "loss": 0.0176, "step": 60110 }, { "grad_norm": 0.19653597474098206, "learning_rate": 1.5955377886418682e-05, "loss": 0.0108, "step": 60120 }, { "grad_norm": 0.1609855443239212, "learning_rate": 1.5940243621882704e-05, "loss": 0.0123, "step": 60130 }, { "grad_norm": 0.2795945405960083, "learning_rate": 1.5925115177227555e-05, "loss": 0.0171, "step": 60140 }, { "grad_norm": 0.2552225887775421, "learning_rate": 1.5909992555038288e-05, "loss": 0.0144, "step": 60150 }, { "grad_norm": 0.2184543013572693, "learning_rate": 1.5894875757898912e-05, "loss": 0.0143, "step": 60160 }, { "grad_norm": 0.21308454871177673, "learning_rate": 1.58797647883925e-05, "loss": 0.0136, "step": 60170 }, { "grad_norm": 0.25134554505348206, "learning_rate": 1.586465964910109e-05, "loss": 0.0162, "step": 60180 }, { "grad_norm": 0.14886225759983063, "learning_rate": 1.5849560342605734e-05, "loss": 0.0125, "step": 60190 }, { "grad_norm": 0.14271104335784912, "learning_rate": 1.583446687148651e-05, "loss": 0.0106, "step": 60200 }, { "grad_norm": 0.2644302546977997, "learning_rate": 1.581937923832248e-05, "loss": 0.0088, "step": 60210 }, { "grad_norm": 0.19455531239509583, "learning_rate": 1.5804297445691663e-05, "loss": 0.014, "step": 60220 }, { "grad_norm": 0.24867583811283112, "learning_rate": 1.578922149617119e-05, "loss": 0.0158, "step": 60230 }, { "grad_norm": 0.19786860048770905, "learning_rate": 1.5774151392337084e-05, "loss": 0.009, "step": 60240 }, { "grad_norm": 0.21205754578113556, "learning_rate": 1.575908713676442e-05, "loss": 0.0103, "step": 60250 }, { "grad_norm": 0.2951320707798004, "learning_rate": 1.5744028732027293e-05, "loss": 0.0158, "step": 60260 }, { "grad_norm": 0.2024976909160614, "learning_rate": 1.5728976180698723e-05, "loss": 0.0129, "step": 60270 }, { "grad_norm": 0.2666838467121124, "learning_rate": 1.5713929485350836e-05, "loss": 0.0123, "step": 60280 }, { "grad_norm": 0.1823519915342331, "learning_rate": 1.569888864855465e-05, "loss": 0.0148, "step": 60290 }, { "grad_norm": 0.1850183755159378, "learning_rate": 1.568385367288028e-05, "loss": 0.0126, "step": 60300 }, { "grad_norm": 0.16372616589069366, "learning_rate": 1.5668824560896755e-05, "loss": 0.0105, "step": 60310 }, { "grad_norm": 0.3873409628868103, "learning_rate": 1.5653801315172156e-05, "loss": 0.0144, "step": 60320 }, { "grad_norm": 0.33408138155937195, "learning_rate": 1.5638783938273554e-05, "loss": 0.013, "step": 60330 }, { "grad_norm": 0.3077911138534546, "learning_rate": 1.5623772432766966e-05, "loss": 0.0163, "step": 60340 }, { "grad_norm": 0.17402690649032593, "learning_rate": 1.5608766801217507e-05, "loss": 0.0134, "step": 60350 }, { "grad_norm": 0.20556621253490448, "learning_rate": 1.5593767046189184e-05, "loss": 0.0102, "step": 60360 }, { "grad_norm": 0.15663394331932068, "learning_rate": 1.5578773170245064e-05, "loss": 0.0096, "step": 60370 }, { "grad_norm": 0.31188568472862244, "learning_rate": 1.5563785175947182e-05, "loss": 0.0167, "step": 60380 }, { "grad_norm": 0.18988578021526337, "learning_rate": 1.5548803065856582e-05, "loss": 0.0115, "step": 60390 }, { "grad_norm": 0.20324012637138367, "learning_rate": 1.55338268425333e-05, "loss": 0.0092, "step": 60400 }, { "grad_norm": 0.13884906470775604, "learning_rate": 1.5518856508536373e-05, "loss": 0.0113, "step": 60410 }, { "grad_norm": 0.3796705901622772, "learning_rate": 1.5503892066423786e-05, "loss": 0.0139, "step": 60420 }, { "grad_norm": 0.3505062162876129, "learning_rate": 1.548893351875258e-05, "loss": 0.0187, "step": 60430 }, { "grad_norm": 0.3305474519729614, "learning_rate": 1.547398086807875e-05, "loss": 0.011, "step": 60440 }, { "grad_norm": 0.4273153841495514, "learning_rate": 1.5459034116957305e-05, "loss": 0.0174, "step": 60450 }, { "grad_norm": 0.14461830258369446, "learning_rate": 1.544409326794225e-05, "loss": 0.0203, "step": 60460 }, { "grad_norm": 0.23943158984184265, "learning_rate": 1.542915832358651e-05, "loss": 0.0153, "step": 60470 }, { "grad_norm": 0.192362979054451, "learning_rate": 1.541422928644214e-05, "loss": 0.0107, "step": 60480 }, { "grad_norm": 0.2451752871274948, "learning_rate": 1.5399306159060024e-05, "loss": 0.0154, "step": 60490 }, { "grad_norm": 0.15789055824279785, "learning_rate": 1.5384388943990185e-05, "loss": 0.0124, "step": 60500 }, { "grad_norm": 0.2178267389535904, "learning_rate": 1.5369477643781526e-05, "loss": 0.0096, "step": 60510 }, { "grad_norm": 0.1746068149805069, "learning_rate": 1.5354572260981985e-05, "loss": 0.0143, "step": 60520 }, { "grad_norm": 0.2768861949443817, "learning_rate": 1.533967279813851e-05, "loss": 0.0098, "step": 60530 }, { "grad_norm": 0.28611981868743896, "learning_rate": 1.5324779257796956e-05, "loss": 0.0096, "step": 60540 }, { "grad_norm": 0.15994612872600555, "learning_rate": 1.5309891642502293e-05, "loss": 0.0127, "step": 60550 }, { "grad_norm": 0.22043593227863312, "learning_rate": 1.5295009954798357e-05, "loss": 0.008, "step": 60560 }, { "grad_norm": 0.2384442239999771, "learning_rate": 1.5280134197228036e-05, "loss": 0.0149, "step": 60570 }, { "grad_norm": 0.17710231244564056, "learning_rate": 1.526526437233319e-05, "loss": 0.0091, "step": 60580 }, { "grad_norm": 0.2126409262418747, "learning_rate": 1.525040048265466e-05, "loss": 0.0113, "step": 60590 }, { "grad_norm": 0.2793186604976654, "learning_rate": 1.5235542530732288e-05, "loss": 0.0104, "step": 60600 }, { "grad_norm": 0.165439173579216, "learning_rate": 1.5220690519104901e-05, "loss": 0.0124, "step": 60610 }, { "grad_norm": 0.22083494067192078, "learning_rate": 1.5205844450310275e-05, "loss": 0.0116, "step": 60620 }, { "grad_norm": 0.2846008539199829, "learning_rate": 1.519100432688521e-05, "loss": 0.0116, "step": 60630 }, { "grad_norm": 0.18389412760734558, "learning_rate": 1.5176170151365476e-05, "loss": 0.0117, "step": 60640 }, { "grad_norm": 0.24266265332698822, "learning_rate": 1.5161341926285833e-05, "loss": 0.0113, "step": 60650 }, { "grad_norm": 0.13911060988903046, "learning_rate": 1.5146519654180025e-05, "loss": 0.0076, "step": 60660 }, { "grad_norm": 0.27079612016677856, "learning_rate": 1.5131703337580739e-05, "loss": 0.0153, "step": 60670 }, { "grad_norm": 0.2520648241043091, "learning_rate": 1.5116892979019731e-05, "loss": 0.0188, "step": 60680 }, { "grad_norm": 0.3026324510574341, "learning_rate": 1.5102088581027623e-05, "loss": 0.0076, "step": 60690 }, { "grad_norm": 0.22306594252586365, "learning_rate": 1.5087290146134154e-05, "loss": 0.01, "step": 60700 }, { "grad_norm": 0.18170391023159027, "learning_rate": 1.5072497676867915e-05, "loss": 0.0139, "step": 60710 }, { "grad_norm": 0.1611400544643402, "learning_rate": 1.505771117575655e-05, "loss": 0.0109, "step": 60720 }, { "grad_norm": 0.1708160787820816, "learning_rate": 1.5042930645326691e-05, "loss": 0.0093, "step": 60730 }, { "grad_norm": 0.18761976063251495, "learning_rate": 1.5028156088103879e-05, "loss": 0.0145, "step": 60740 }, { "grad_norm": 0.5018525123596191, "learning_rate": 1.5013387506612735e-05, "loss": 0.0145, "step": 60750 }, { "grad_norm": 0.20215588808059692, "learning_rate": 1.4998624903376767e-05, "loss": 0.0087, "step": 60760 }, { "grad_norm": 0.2490084022283554, "learning_rate": 1.4983868280918517e-05, "loss": 0.0108, "step": 60770 }, { "grad_norm": 0.2573862671852112, "learning_rate": 1.4969117641759478e-05, "loss": 0.0154, "step": 60780 }, { "grad_norm": 0.21796400845050812, "learning_rate": 1.495437298842014e-05, "loss": 0.0111, "step": 60790 }, { "grad_norm": 0.2064579576253891, "learning_rate": 1.4939634323419976e-05, "loss": 0.0098, "step": 60800 }, { "grad_norm": 0.29613739252090454, "learning_rate": 1.4924901649277384e-05, "loss": 0.014, "step": 60810 }, { "grad_norm": 0.1926073282957077, "learning_rate": 1.4910174968509793e-05, "loss": 0.0092, "step": 60820 }, { "grad_norm": 0.11962935328483582, "learning_rate": 1.4895454283633598e-05, "loss": 0.0081, "step": 60830 }, { "grad_norm": 0.17665693163871765, "learning_rate": 1.488073959716415e-05, "loss": 0.0105, "step": 60840 }, { "grad_norm": 0.2365332692861557, "learning_rate": 1.4866030911615791e-05, "loss": 0.0141, "step": 60850 }, { "grad_norm": 0.15303848683834076, "learning_rate": 1.4851328229501849e-05, "loss": 0.0107, "step": 60860 }, { "grad_norm": 0.15841908752918243, "learning_rate": 1.4836631553334562e-05, "loss": 0.0089, "step": 60870 }, { "grad_norm": 0.22630733251571655, "learning_rate": 1.4821940885625251e-05, "loss": 0.0107, "step": 60880 }, { "grad_norm": 0.24631379544734955, "learning_rate": 1.4807256228884109e-05, "loss": 0.0095, "step": 60890 }, { "grad_norm": 0.13465726375579834, "learning_rate": 1.4792577585620353e-05, "loss": 0.011, "step": 60900 }, { "grad_norm": 0.22093653678894043, "learning_rate": 1.4777904958342164e-05, "loss": 0.0109, "step": 60910 }, { "grad_norm": 0.09984778612852097, "learning_rate": 1.4763238349556691e-05, "loss": 0.009, "step": 60920 }, { "grad_norm": 0.175957590341568, "learning_rate": 1.4748577761770072e-05, "loss": 0.0166, "step": 60930 }, { "grad_norm": 0.21229003369808197, "learning_rate": 1.4733923197487354e-05, "loss": 0.0132, "step": 60940 }, { "grad_norm": 0.16544167697429657, "learning_rate": 1.4719274659212662e-05, "loss": 0.0123, "step": 60950 }, { "grad_norm": 0.25377893447875977, "learning_rate": 1.4704632149448983e-05, "loss": 0.012, "step": 60960 }, { "grad_norm": 0.38492801785469055, "learning_rate": 1.4689995670698343e-05, "loss": 0.0195, "step": 60970 }, { "grad_norm": 0.1832265704870224, "learning_rate": 1.4675365225461728e-05, "loss": 0.0121, "step": 60980 }, { "grad_norm": 0.2932940721511841, "learning_rate": 1.4660740816239032e-05, "loss": 0.0123, "step": 60990 }, { "grad_norm": 0.2107544094324112, "learning_rate": 1.464612244552923e-05, "loss": 0.0111, "step": 61000 }, { "grad_norm": 0.22119973599910736, "learning_rate": 1.4631510115830161e-05, "loss": 0.011, "step": 61010 }, { "grad_norm": 0.39649447798728943, "learning_rate": 1.4616903829638679e-05, "loss": 0.0126, "step": 61020 }, { "grad_norm": 0.16953112185001373, "learning_rate": 1.46023035894506e-05, "loss": 0.012, "step": 61030 }, { "grad_norm": 0.35307595133781433, "learning_rate": 1.4587709397760713e-05, "loss": 0.0136, "step": 61040 }, { "grad_norm": 0.2551000416278839, "learning_rate": 1.4573121257062755e-05, "loss": 0.0123, "step": 61050 }, { "grad_norm": 0.47432243824005127, "learning_rate": 1.4558539169849472e-05, "loss": 0.0166, "step": 61060 }, { "grad_norm": 0.24276413023471832, "learning_rate": 1.4543963138612499e-05, "loss": 0.0121, "step": 61070 }, { "grad_norm": 0.323091059923172, "learning_rate": 1.4529393165842498e-05, "loss": 0.0135, "step": 61080 }, { "grad_norm": 0.38646140694618225, "learning_rate": 1.4514829254029084e-05, "loss": 0.0131, "step": 61090 }, { "grad_norm": 0.17703086137771606, "learning_rate": 1.450027140566083e-05, "loss": 0.0118, "step": 61100 }, { "grad_norm": 0.20195771753787994, "learning_rate": 1.4485719623225268e-05, "loss": 0.0102, "step": 61110 }, { "grad_norm": 0.24671632051467896, "learning_rate": 1.4471173909208912e-05, "loss": 0.0141, "step": 61120 }, { "grad_norm": 0.11217174679040909, "learning_rate": 1.4456634266097236e-05, "loss": 0.0115, "step": 61130 }, { "grad_norm": 0.26270338892936707, "learning_rate": 1.444210069637461e-05, "loss": 0.0177, "step": 61140 }, { "grad_norm": 0.1665663719177246, "learning_rate": 1.4427573202524502e-05, "loss": 0.0068, "step": 61150 }, { "grad_norm": 0.18688255548477173, "learning_rate": 1.4413051787029208e-05, "loss": 0.0161, "step": 61160 }, { "grad_norm": 0.1453220397233963, "learning_rate": 1.439853645237006e-05, "loss": 0.0124, "step": 61170 }, { "grad_norm": 0.11291235685348511, "learning_rate": 1.4384027201027344e-05, "loss": 0.0076, "step": 61180 }, { "grad_norm": 0.18928825855255127, "learning_rate": 1.4369524035480253e-05, "loss": 0.0118, "step": 61190 }, { "grad_norm": 0.1585158258676529, "learning_rate": 1.4355026958207035e-05, "loss": 0.0102, "step": 61200 }, { "grad_norm": 0.10870198905467987, "learning_rate": 1.434053597168481e-05, "loss": 0.012, "step": 61210 }, { "grad_norm": 0.2666948139667511, "learning_rate": 1.432605107838969e-05, "loss": 0.0097, "step": 61220 }, { "grad_norm": 0.2754850387573242, "learning_rate": 1.4311572280796759e-05, "loss": 0.0107, "step": 61230 }, { "grad_norm": 0.13995617628097534, "learning_rate": 1.4297099581380047e-05, "loss": 0.0159, "step": 61240 }, { "grad_norm": 0.24904100596904755, "learning_rate": 1.4282632982612538e-05, "loss": 0.0115, "step": 61250 }, { "grad_norm": 0.11176622658967972, "learning_rate": 1.4268172486966197e-05, "loss": 0.013, "step": 61260 }, { "grad_norm": 0.32702404260635376, "learning_rate": 1.4253718096911894e-05, "loss": 0.0126, "step": 61270 }, { "grad_norm": 0.300273060798645, "learning_rate": 1.4239269814919504e-05, "loss": 0.0113, "step": 61280 }, { "grad_norm": 0.21364006400108337, "learning_rate": 1.4224827643457843e-05, "loss": 0.0159, "step": 61290 }, { "grad_norm": 0.13845132291316986, "learning_rate": 1.4210391584994686e-05, "loss": 0.0089, "step": 61300 }, { "grad_norm": 0.1687048077583313, "learning_rate": 1.4195961641996763e-05, "loss": 0.0127, "step": 61310 }, { "grad_norm": 0.17642436921596527, "learning_rate": 1.4181537816929751e-05, "loss": 0.0078, "step": 61320 }, { "grad_norm": 0.16353212296962738, "learning_rate": 1.4167120112258303e-05, "loss": 0.0143, "step": 61330 }, { "grad_norm": 0.26888811588287354, "learning_rate": 1.4152708530445963e-05, "loss": 0.0097, "step": 61340 }, { "grad_norm": 0.10149787366390228, "learning_rate": 1.4138303073955344e-05, "loss": 0.0081, "step": 61350 }, { "grad_norm": 0.2704191505908966, "learning_rate": 1.4123903745247897e-05, "loss": 0.0121, "step": 61360 }, { "grad_norm": 0.10061738640069962, "learning_rate": 1.4109510546784082e-05, "loss": 0.0065, "step": 61370 }, { "grad_norm": 0.16859564185142517, "learning_rate": 1.4095123481023325e-05, "loss": 0.0118, "step": 61380 }, { "grad_norm": 0.2503122389316559, "learning_rate": 1.4080742550423936e-05, "loss": 0.013, "step": 61390 }, { "grad_norm": 0.1409573256969452, "learning_rate": 1.4066367757443278e-05, "loss": 0.0127, "step": 61400 }, { "grad_norm": 0.29438140988349915, "learning_rate": 1.4051999104537572e-05, "loss": 0.0083, "step": 61410 }, { "grad_norm": 0.2786779999732971, "learning_rate": 1.4037636594162034e-05, "loss": 0.0127, "step": 61420 }, { "grad_norm": 0.28774645924568176, "learning_rate": 1.402328022877083e-05, "loss": 0.0094, "step": 61430 }, { "grad_norm": 0.1500544250011444, "learning_rate": 1.4008930010817073e-05, "loss": 0.0121, "step": 61440 }, { "grad_norm": 0.20794008672237396, "learning_rate": 1.3994585942752831e-05, "loss": 0.0082, "step": 61450 }, { "grad_norm": 0.2220044732093811, "learning_rate": 1.398024802702909e-05, "loss": 0.0076, "step": 61460 }, { "grad_norm": 0.2284765988588333, "learning_rate": 1.3965916266095819e-05, "loss": 0.0112, "step": 61470 }, { "grad_norm": 0.3538054823875427, "learning_rate": 1.395159066240192e-05, "loss": 0.0103, "step": 61480 }, { "grad_norm": 0.10610921680927277, "learning_rate": 1.3937271218395248e-05, "loss": 0.0091, "step": 61490 }, { "grad_norm": 0.26079222559928894, "learning_rate": 1.3922957936522613e-05, "loss": 0.0115, "step": 61500 }, { "grad_norm": 0.3234610855579376, "learning_rate": 1.3908650819229758e-05, "loss": 0.0122, "step": 61510 }, { "grad_norm": 0.2514275014400482, "learning_rate": 1.389434986896137e-05, "loss": 0.0112, "step": 61520 }, { "grad_norm": 0.15518632531166077, "learning_rate": 1.388005508816112e-05, "loss": 0.0085, "step": 61530 }, { "grad_norm": 0.21551547944545746, "learning_rate": 1.3865766479271557e-05, "loss": 0.0105, "step": 61540 }, { "grad_norm": 0.14140042662620544, "learning_rate": 1.3851484044734225e-05, "loss": 0.0141, "step": 61550 }, { "grad_norm": 0.21438150107860565, "learning_rate": 1.3837207786989608e-05, "loss": 0.0131, "step": 61560 }, { "grad_norm": 0.29966968297958374, "learning_rate": 1.382293770847713e-05, "loss": 0.0126, "step": 61570 }, { "grad_norm": 0.3375105857849121, "learning_rate": 1.3808673811635159e-05, "loss": 0.0143, "step": 61580 }, { "grad_norm": 0.21122421324253082, "learning_rate": 1.3794416098900975e-05, "loss": 0.0135, "step": 61590 }, { "grad_norm": 0.18426065146923065, "learning_rate": 1.3780164572710886e-05, "loss": 0.0158, "step": 61600 }, { "grad_norm": 0.10168901085853577, "learning_rate": 1.3765919235500035e-05, "loss": 0.0153, "step": 61610 }, { "grad_norm": 0.18512681126594543, "learning_rate": 1.375168008970259e-05, "loss": 0.0107, "step": 61620 }, { "grad_norm": 0.261405885219574, "learning_rate": 1.3737447137751635e-05, "loss": 0.0102, "step": 61630 }, { "grad_norm": 0.4864036738872528, "learning_rate": 1.372322038207915e-05, "loss": 0.013, "step": 61640 }, { "grad_norm": 0.35427194833755493, "learning_rate": 1.3708999825116159e-05, "loss": 0.0164, "step": 61650 }, { "grad_norm": 0.2364688515663147, "learning_rate": 1.3694785469292526e-05, "loss": 0.0171, "step": 61660 }, { "grad_norm": 0.2042403370141983, "learning_rate": 1.3680577317037101e-05, "loss": 0.0106, "step": 61670 }, { "grad_norm": 0.18826797604560852, "learning_rate": 1.3666375370777678e-05, "loss": 0.0106, "step": 61680 }, { "grad_norm": 0.16232411563396454, "learning_rate": 1.3652179632940981e-05, "loss": 0.01, "step": 61690 }, { "grad_norm": 0.1789727658033371, "learning_rate": 1.3637990105952664e-05, "loss": 0.0084, "step": 61700 }, { "grad_norm": 0.21854659914970398, "learning_rate": 1.3623806792237337e-05, "loss": 0.0103, "step": 61710 }, { "grad_norm": 0.15236787497997284, "learning_rate": 1.3609629694218551e-05, "loss": 0.0121, "step": 61720 }, { "grad_norm": 0.21049876511096954, "learning_rate": 1.3595458814318762e-05, "loss": 0.0106, "step": 61730 }, { "grad_norm": 0.2165946662425995, "learning_rate": 1.3581294154959389e-05, "loss": 0.0102, "step": 61740 }, { "grad_norm": 0.2939510643482208, "learning_rate": 1.3567135718560792e-05, "loss": 0.0104, "step": 61750 }, { "grad_norm": 0.3191847503185272, "learning_rate": 1.3552983507542261e-05, "loss": 0.0103, "step": 61760 }, { "grad_norm": 0.22660967707633972, "learning_rate": 1.3538837524322023e-05, "loss": 0.0119, "step": 61770 }, { "grad_norm": 0.09392619132995605, "learning_rate": 1.3524697771317251e-05, "loss": 0.0161, "step": 61780 }, { "grad_norm": 0.18636181950569153, "learning_rate": 1.3510564250943997e-05, "loss": 0.0137, "step": 61790 }, { "grad_norm": 0.30040547251701355, "learning_rate": 1.3496436965617353e-05, "loss": 0.0108, "step": 61800 }, { "grad_norm": 0.3123336136341095, "learning_rate": 1.3482315917751243e-05, "loss": 0.0133, "step": 61810 }, { "grad_norm": 0.18178005516529083, "learning_rate": 1.3468201109758582e-05, "loss": 0.0082, "step": 61820 }, { "grad_norm": 0.09464763849973679, "learning_rate": 1.3454092544051222e-05, "loss": 0.0118, "step": 61830 }, { "grad_norm": 0.35755303502082825, "learning_rate": 1.3439990223039878e-05, "loss": 0.0121, "step": 61840 }, { "grad_norm": 0.1390826553106308, "learning_rate": 1.3425894149134321e-05, "loss": 0.0122, "step": 61850 }, { "grad_norm": 0.19269561767578125, "learning_rate": 1.3411804324743132e-05, "loss": 0.0118, "step": 61860 }, { "grad_norm": 0.25527223944664, "learning_rate": 1.3397720752273896e-05, "loss": 0.016, "step": 61870 }, { "grad_norm": 0.17242532968521118, "learning_rate": 1.3383643434133108e-05, "loss": 0.0091, "step": 61880 }, { "grad_norm": 0.28706610202789307, "learning_rate": 1.3369572372726196e-05, "loss": 0.012, "step": 61890 }, { "grad_norm": 0.09551587700843811, "learning_rate": 1.3355507570457525e-05, "loss": 0.01, "step": 61900 }, { "grad_norm": 0.17131906747817993, "learning_rate": 1.334144902973038e-05, "loss": 0.0103, "step": 61910 }, { "grad_norm": 0.2025088369846344, "learning_rate": 1.3327396752946997e-05, "loss": 0.0117, "step": 61920 }, { "grad_norm": 0.0946778953075409, "learning_rate": 1.3313350742508496e-05, "loss": 0.0102, "step": 61930 }, { "grad_norm": 0.11961008608341217, "learning_rate": 1.3299311000814973e-05, "loss": 0.0092, "step": 61940 }, { "grad_norm": 0.18316969275474548, "learning_rate": 1.3285277530265432e-05, "loss": 0.0168, "step": 61950 }, { "grad_norm": 0.2777400314807892, "learning_rate": 1.3271250333257813e-05, "loss": 0.0107, "step": 61960 }, { "grad_norm": 0.13171102106571198, "learning_rate": 1.3257229412188977e-05, "loss": 0.0098, "step": 61970 }, { "grad_norm": 0.3422311246395111, "learning_rate": 1.3243214769454726e-05, "loss": 0.011, "step": 61980 }, { "grad_norm": 0.20000231266021729, "learning_rate": 1.3229206407449751e-05, "loss": 0.0107, "step": 61990 }, { "grad_norm": 0.1281467080116272, "learning_rate": 1.3215204328567742e-05, "loss": 0.0118, "step": 62000 }, { "grad_norm": 0.1255037784576416, "learning_rate": 1.3201208535201232e-05, "loss": 0.0115, "step": 62010 }, { "grad_norm": 0.16368815302848816, "learning_rate": 1.3187219029741737e-05, "loss": 0.0084, "step": 62020 }, { "grad_norm": 0.23437471687793732, "learning_rate": 1.3173235814579693e-05, "loss": 0.0096, "step": 62030 }, { "grad_norm": 0.2888222932815552, "learning_rate": 1.3159258892104398e-05, "loss": 0.0115, "step": 62040 }, { "grad_norm": 0.13698291778564453, "learning_rate": 1.3145288264704198e-05, "loss": 0.0146, "step": 62050 }, { "grad_norm": 0.11102647334337234, "learning_rate": 1.3131323934766237e-05, "loss": 0.0103, "step": 62060 }, { "grad_norm": 0.14399969577789307, "learning_rate": 1.311736590467666e-05, "loss": 0.013, "step": 62070 }, { "grad_norm": 0.3121497929096222, "learning_rate": 1.3103414176820505e-05, "loss": 0.0119, "step": 62080 }, { "grad_norm": 0.10841257125139236, "learning_rate": 1.308946875358174e-05, "loss": 0.0165, "step": 62090 }, { "grad_norm": 0.29336705803871155, "learning_rate": 1.3075529637343276e-05, "loss": 0.0211, "step": 62100 }, { "grad_norm": 0.3003551661968231, "learning_rate": 1.3061596830486883e-05, "loss": 0.0118, "step": 62110 }, { "grad_norm": 0.18538203835487366, "learning_rate": 1.3047670335393353e-05, "loss": 0.0079, "step": 62120 }, { "grad_norm": 0.1785411685705185, "learning_rate": 1.3033750154442298e-05, "loss": 0.011, "step": 62130 }, { "grad_norm": 0.23137298226356506, "learning_rate": 1.3019836290012316e-05, "loss": 0.0108, "step": 62140 }, { "grad_norm": 0.15314137935638428, "learning_rate": 1.30059287444809e-05, "loss": 0.013, "step": 62150 }, { "grad_norm": 0.2380487471818924, "learning_rate": 1.299202752022447e-05, "loss": 0.0199, "step": 62160 }, { "grad_norm": 0.26148349046707153, "learning_rate": 1.2978132619618371e-05, "loss": 0.0095, "step": 62170 }, { "grad_norm": 0.19568775594234467, "learning_rate": 1.2964244045036866e-05, "loss": 0.0174, "step": 62180 }, { "grad_norm": 0.305729478597641, "learning_rate": 1.295036179885311e-05, "loss": 0.0088, "step": 62190 }, { "grad_norm": 0.13832774758338928, "learning_rate": 1.293648588343922e-05, "loss": 0.0146, "step": 62200 }, { "grad_norm": 0.199406698346138, "learning_rate": 1.2922616301166196e-05, "loss": 0.0115, "step": 62210 }, { "grad_norm": 0.33371424674987793, "learning_rate": 1.2908753054403976e-05, "loss": 0.0122, "step": 62220 }, { "grad_norm": 0.13750436902046204, "learning_rate": 1.2894896145521429e-05, "loss": 0.0071, "step": 62230 }, { "grad_norm": 0.19293999671936035, "learning_rate": 1.2881045576886275e-05, "loss": 0.0127, "step": 62240 }, { "grad_norm": 0.22320672869682312, "learning_rate": 1.2867201350865254e-05, "loss": 0.0111, "step": 62250 }, { "grad_norm": 0.16454672813415527, "learning_rate": 1.2853363469823914e-05, "loss": 0.0122, "step": 62260 }, { "grad_norm": 0.22051839530467987, "learning_rate": 1.283953193612682e-05, "loss": 0.0092, "step": 62270 }, { "grad_norm": 0.1686476767063141, "learning_rate": 1.2825706752137372e-05, "loss": 0.0107, "step": 62280 }, { "grad_norm": 0.3396904468536377, "learning_rate": 1.2811887920217896e-05, "loss": 0.0134, "step": 62290 }, { "grad_norm": 0.19333821535110474, "learning_rate": 1.2798075442729707e-05, "loss": 0.0108, "step": 62300 }, { "grad_norm": 0.13298900425434113, "learning_rate": 1.2784269322032922e-05, "loss": 0.0112, "step": 62310 }, { "grad_norm": 0.1042407900094986, "learning_rate": 1.2770469560486686e-05, "loss": 0.0117, "step": 62320 }, { "grad_norm": 0.27441638708114624, "learning_rate": 1.2756676160448956e-05, "loss": 0.0087, "step": 62330 }, { "grad_norm": 0.16847047209739685, "learning_rate": 1.2742889124276663e-05, "loss": 0.0102, "step": 62340 }, { "grad_norm": 0.1778094619512558, "learning_rate": 1.2729108454325639e-05, "loss": 0.0123, "step": 62350 }, { "grad_norm": 0.30898502469062805, "learning_rate": 1.2715334152950614e-05, "loss": 0.0119, "step": 62360 }, { "grad_norm": 0.11881471425294876, "learning_rate": 1.2701566222505246e-05, "loss": 0.0089, "step": 62370 }, { "grad_norm": 0.3285519778728485, "learning_rate": 1.2687804665342107e-05, "loss": 0.0126, "step": 62380 }, { "grad_norm": 0.14311783015727997, "learning_rate": 1.267404948381265e-05, "loss": 0.0103, "step": 62390 }, { "grad_norm": 0.21302857995033264, "learning_rate": 1.2660300680267267e-05, "loss": 0.0088, "step": 62400 }, { "grad_norm": 0.23702308535575867, "learning_rate": 1.2646558257055257e-05, "loss": 0.0096, "step": 62410 }, { "grad_norm": 0.24892157316207886, "learning_rate": 1.263282221652482e-05, "loss": 0.0085, "step": 62420 }, { "grad_norm": 0.14401549100875854, "learning_rate": 1.2619092561023088e-05, "loss": 0.0102, "step": 62430 }, { "grad_norm": 0.2572649121284485, "learning_rate": 1.2605369292896036e-05, "loss": 0.0127, "step": 62440 }, { "grad_norm": 0.16288743913173676, "learning_rate": 1.2591652414488658e-05, "loss": 0.0113, "step": 62450 }, { "grad_norm": 0.38776257634162903, "learning_rate": 1.2577941928144732e-05, "loss": 0.0095, "step": 62460 }, { "grad_norm": 0.1331440657377243, "learning_rate": 1.256423783620706e-05, "loss": 0.0114, "step": 62470 }, { "grad_norm": 0.23741725087165833, "learning_rate": 1.2550540141017264e-05, "loss": 0.0063, "step": 62480 }, { "grad_norm": 0.1150272935628891, "learning_rate": 1.253684884491591e-05, "loss": 0.0081, "step": 62490 }, { "grad_norm": 0.2311524748802185, "learning_rate": 1.2523163950242483e-05, "loss": 0.0072, "step": 62500 }, { "grad_norm": 0.10058841109275818, "learning_rate": 1.2509485459335313e-05, "loss": 0.0102, "step": 62510 }, { "grad_norm": 0.11468750238418579, "learning_rate": 1.2495813374531739e-05, "loss": 0.0098, "step": 62520 }, { "grad_norm": 0.2532922923564911, "learning_rate": 1.2482147698167907e-05, "loss": 0.0072, "step": 62530 }, { "grad_norm": 0.11626560986042023, "learning_rate": 1.2468488432578911e-05, "loss": 0.009, "step": 62540 }, { "grad_norm": 0.1750262975692749, "learning_rate": 1.2454835580098761e-05, "loss": 0.0107, "step": 62550 }, { "grad_norm": 0.10068370401859283, "learning_rate": 1.2441189143060338e-05, "loss": 0.0087, "step": 62560 }, { "grad_norm": 0.21162699162960052, "learning_rate": 1.242754912379548e-05, "loss": 0.012, "step": 62570 }, { "grad_norm": 0.1560768038034439, "learning_rate": 1.2413915524634844e-05, "loss": 0.021, "step": 62580 }, { "grad_norm": 0.18242831528186798, "learning_rate": 1.2400288347908073e-05, "loss": 0.01, "step": 62590 }, { "grad_norm": 0.12926657497882843, "learning_rate": 1.2386667595943663e-05, "loss": 0.0078, "step": 62600 }, { "grad_norm": 0.4271498918533325, "learning_rate": 1.2373053271069035e-05, "loss": 0.0087, "step": 62610 }, { "grad_norm": 0.2317553460597992, "learning_rate": 1.2359445375610501e-05, "loss": 0.0127, "step": 62620 }, { "grad_norm": 0.25673148036003113, "learning_rate": 1.2345843911893301e-05, "loss": 0.0068, "step": 62630 }, { "grad_norm": 0.1537807583808899, "learning_rate": 1.2332248882241498e-05, "loss": 0.0209, "step": 62640 }, { "grad_norm": 0.3155604898929596, "learning_rate": 1.2318660288978178e-05, "loss": 0.01, "step": 62650 }, { "grad_norm": 0.18489114940166473, "learning_rate": 1.2305078134425213e-05, "loss": 0.0116, "step": 62660 }, { "grad_norm": 0.13357143104076385, "learning_rate": 1.2291502420903434e-05, "loss": 0.0122, "step": 62670 }, { "grad_norm": 0.13989651203155518, "learning_rate": 1.2277933150732567e-05, "loss": 0.0114, "step": 62680 }, { "grad_norm": 0.163691446185112, "learning_rate": 1.2264370326231216e-05, "loss": 0.0149, "step": 62690 }, { "grad_norm": 0.2077246755361557, "learning_rate": 1.2250813949716927e-05, "loss": 0.0116, "step": 62700 }, { "grad_norm": 0.49481040239334106, "learning_rate": 1.2237264023506063e-05, "loss": 0.009, "step": 62710 }, { "grad_norm": 0.2724255919456482, "learning_rate": 1.2223720549913987e-05, "loss": 0.0111, "step": 62720 }, { "grad_norm": 0.24133159220218658, "learning_rate": 1.221018353125487e-05, "loss": 0.0099, "step": 62730 }, { "grad_norm": 0.1670197695493698, "learning_rate": 1.2196652969841837e-05, "loss": 0.0121, "step": 62740 }, { "grad_norm": 0.34424057602882385, "learning_rate": 1.2183128867986904e-05, "loss": 0.0143, "step": 62750 }, { "grad_norm": 0.13729514181613922, "learning_rate": 1.2169611228000927e-05, "loss": 0.0141, "step": 62760 }, { "grad_norm": 0.209077388048172, "learning_rate": 1.2156100052193752e-05, "loss": 0.0086, "step": 62770 }, { "grad_norm": 0.23242990672588348, "learning_rate": 1.214259534287403e-05, "loss": 0.0093, "step": 62780 }, { "grad_norm": 0.13940100371837616, "learning_rate": 1.2129097102349363e-05, "loss": 0.0125, "step": 62790 }, { "grad_norm": 0.12660279870033264, "learning_rate": 1.2115605332926227e-05, "loss": 0.0101, "step": 62800 }, { "grad_norm": 0.12256287783384323, "learning_rate": 1.2102120036910003e-05, "loss": 0.0146, "step": 62810 }, { "grad_norm": 0.18865032494068146, "learning_rate": 1.2088641216604956e-05, "loss": 0.0171, "step": 62820 }, { "grad_norm": 0.1995278298854828, "learning_rate": 1.2075168874314264e-05, "loss": 0.011, "step": 62830 }, { "grad_norm": 0.14660149812698364, "learning_rate": 1.2061703012339942e-05, "loss": 0.0096, "step": 62840 }, { "grad_norm": 0.2266806811094284, "learning_rate": 1.204824363298297e-05, "loss": 0.0104, "step": 62850 }, { "grad_norm": 0.1973528265953064, "learning_rate": 1.2034790738543173e-05, "loss": 0.0169, "step": 62860 }, { "grad_norm": 0.15961790084838867, "learning_rate": 1.2021344331319284e-05, "loss": 0.0081, "step": 62870 }, { "grad_norm": 0.20502857863903046, "learning_rate": 1.200790441360894e-05, "loss": 0.0092, "step": 62880 }, { "grad_norm": 0.24352610111236572, "learning_rate": 1.199447098770864e-05, "loss": 0.0107, "step": 62890 }, { "grad_norm": 0.15977314114570618, "learning_rate": 1.198104405591381e-05, "loss": 0.012, "step": 62900 }, { "grad_norm": 0.16834141314029694, "learning_rate": 1.1967623620518697e-05, "loss": 0.0093, "step": 62910 }, { "grad_norm": 0.4853670001029968, "learning_rate": 1.1954209683816554e-05, "loss": 0.0161, "step": 62920 }, { "grad_norm": 0.1490097939968109, "learning_rate": 1.1940802248099402e-05, "loss": 0.0118, "step": 62930 }, { "grad_norm": 0.18910393118858337, "learning_rate": 1.1927401315658232e-05, "loss": 0.0128, "step": 62940 }, { "grad_norm": 0.24175642430782318, "learning_rate": 1.1914006888782898e-05, "loss": 0.0113, "step": 62950 }, { "grad_norm": 0.3612872064113617, "learning_rate": 1.1900618969762106e-05, "loss": 0.0102, "step": 62960 }, { "grad_norm": 0.1752750426530838, "learning_rate": 1.188723756088354e-05, "loss": 0.0107, "step": 62970 }, { "grad_norm": 0.19198019802570343, "learning_rate": 1.1873862664433672e-05, "loss": 0.0145, "step": 62980 }, { "grad_norm": 0.3711588978767395, "learning_rate": 1.1860494282697927e-05, "loss": 0.0122, "step": 62990 }, { "grad_norm": 0.2625628709793091, "learning_rate": 1.1847132417960588e-05, "loss": 0.0115, "step": 63000 }, { "grad_norm": 0.1980242133140564, "learning_rate": 1.1833777072504832e-05, "loss": 0.0135, "step": 63010 }, { "grad_norm": 0.1972891092300415, "learning_rate": 1.1820428248612731e-05, "loss": 0.0123, "step": 63020 }, { "grad_norm": 0.3769187927246094, "learning_rate": 1.1807085948565245e-05, "loss": 0.0111, "step": 63030 }, { "grad_norm": 0.4206341505050659, "learning_rate": 1.1793750174642172e-05, "loss": 0.0129, "step": 63040 }, { "grad_norm": 0.20315073430538177, "learning_rate": 1.1780420929122254e-05, "loss": 0.0137, "step": 63050 }, { "grad_norm": 0.2078545093536377, "learning_rate": 1.176709821428309e-05, "loss": 0.0122, "step": 63060 }, { "grad_norm": 0.1489381641149521, "learning_rate": 1.1753782032401173e-05, "loss": 0.0144, "step": 63070 }, { "grad_norm": 0.24437856674194336, "learning_rate": 1.1740472385751866e-05, "loss": 0.0091, "step": 63080 }, { "grad_norm": 0.15697453916072845, "learning_rate": 1.172716927660943e-05, "loss": 0.0103, "step": 63090 }, { "grad_norm": 0.24157774448394775, "learning_rate": 1.1713872707247015e-05, "loss": 0.0122, "step": 63100 }, { "grad_norm": 0.11561949551105499, "learning_rate": 1.1700582679936595e-05, "loss": 0.008, "step": 63110 }, { "grad_norm": 0.2549037039279938, "learning_rate": 1.1687299196949136e-05, "loss": 0.0124, "step": 63120 }, { "grad_norm": 0.14768445491790771, "learning_rate": 1.1674022260554374e-05, "loss": 0.0074, "step": 63130 }, { "grad_norm": 0.222062885761261, "learning_rate": 1.1660751873020987e-05, "loss": 0.0104, "step": 63140 }, { "grad_norm": 0.17296946048736572, "learning_rate": 1.1647488036616538e-05, "loss": 0.0121, "step": 63150 }, { "grad_norm": 0.14386865496635437, "learning_rate": 1.1634230753607417e-05, "loss": 0.014, "step": 63160 }, { "grad_norm": 0.22550933063030243, "learning_rate": 1.1620980026258982e-05, "loss": 0.011, "step": 63170 }, { "grad_norm": 0.12591606378555298, "learning_rate": 1.1607735856835373e-05, "loss": 0.0072, "step": 63180 }, { "grad_norm": 0.2938004434108734, "learning_rate": 1.1594498247599677e-05, "loss": 0.0124, "step": 63190 }, { "grad_norm": 0.14512695372104645, "learning_rate": 1.158126720081384e-05, "loss": 0.0157, "step": 63200 }, { "grad_norm": 0.20931093394756317, "learning_rate": 1.156804271873868e-05, "loss": 0.0102, "step": 63210 }, { "grad_norm": 0.1988527774810791, "learning_rate": 1.1554824803633924e-05, "loss": 0.0103, "step": 63220 }, { "grad_norm": 0.24158965051174164, "learning_rate": 1.154161345775811e-05, "loss": 0.0077, "step": 63230 }, { "grad_norm": 0.2859187126159668, "learning_rate": 1.1528408683368724e-05, "loss": 0.0115, "step": 63240 }, { "grad_norm": 0.4267948269844055, "learning_rate": 1.1515210482722088e-05, "loss": 0.0147, "step": 63250 }, { "grad_norm": 0.16928955912590027, "learning_rate": 1.150201885807342e-05, "loss": 0.0102, "step": 63260 }, { "grad_norm": 0.23328885436058044, "learning_rate": 1.1488833811676807e-05, "loss": 0.0083, "step": 63270 }, { "grad_norm": 0.1570722460746765, "learning_rate": 1.1475655345785213e-05, "loss": 0.0074, "step": 63280 }, { "grad_norm": 0.10797346383333206, "learning_rate": 1.1462483462650481e-05, "loss": 0.0105, "step": 63290 }, { "grad_norm": 0.20573295652866364, "learning_rate": 1.1449318164523331e-05, "loss": 0.0097, "step": 63300 }, { "grad_norm": 0.13783885538578033, "learning_rate": 1.1436159453653334e-05, "loss": 0.0076, "step": 63310 }, { "grad_norm": 0.3401317298412323, "learning_rate": 1.1423007332288955e-05, "loss": 0.0159, "step": 63320 }, { "grad_norm": 0.2831420600414276, "learning_rate": 1.1409861802677546e-05, "loss": 0.0118, "step": 63330 }, { "grad_norm": 0.1495094746351242, "learning_rate": 1.1396722867065313e-05, "loss": 0.0071, "step": 63340 }, { "grad_norm": 0.4263421595096588, "learning_rate": 1.1383590527697352e-05, "loss": 0.0123, "step": 63350 }, { "grad_norm": 0.12031285464763641, "learning_rate": 1.137046478681758e-05, "loss": 0.0098, "step": 63360 }, { "grad_norm": 0.2242913842201233, "learning_rate": 1.1357345646668888e-05, "loss": 0.0282, "step": 63370 }, { "grad_norm": 0.25824055075645447, "learning_rate": 1.1344233109492924e-05, "loss": 0.0094, "step": 63380 }, { "grad_norm": 0.13819019496440887, "learning_rate": 1.1331127177530292e-05, "loss": 0.0081, "step": 63390 }, { "grad_norm": 0.09673139452934265, "learning_rate": 1.1318027853020441e-05, "loss": 0.0103, "step": 63400 }, { "grad_norm": 0.07991810142993927, "learning_rate": 1.1304935138201645e-05, "loss": 0.0092, "step": 63410 }, { "grad_norm": 0.1899612993001938, "learning_rate": 1.1291849035311153e-05, "loss": 0.0127, "step": 63420 }, { "grad_norm": 0.13545016944408417, "learning_rate": 1.1278769546584972e-05, "loss": 0.0079, "step": 63430 }, { "grad_norm": 0.09004411101341248, "learning_rate": 1.1265696674258052e-05, "loss": 0.008, "step": 63440 }, { "grad_norm": 0.11921890079975128, "learning_rate": 1.1252630420564186e-05, "loss": 0.0075, "step": 63450 }, { "grad_norm": 0.2627914547920227, "learning_rate": 1.1239570787736036e-05, "loss": 0.0117, "step": 63460 }, { "grad_norm": 0.14124609529972076, "learning_rate": 1.1226517778005135e-05, "loss": 0.0102, "step": 63470 }, { "grad_norm": 0.2090311348438263, "learning_rate": 1.1213471393601893e-05, "loss": 0.0083, "step": 63480 }, { "grad_norm": 0.24122484028339386, "learning_rate": 1.1200431636755587e-05, "loss": 0.0137, "step": 63490 }, { "grad_norm": 0.39215150475502014, "learning_rate": 1.1187398509694336e-05, "loss": 0.0119, "step": 63500 }, { "grad_norm": 0.30974793434143066, "learning_rate": 1.1174372014645146e-05, "loss": 0.008, "step": 63510 }, { "grad_norm": 0.6870608925819397, "learning_rate": 1.1161352153833899e-05, "loss": 0.0122, "step": 63520 }, { "grad_norm": 0.1759670227766037, "learning_rate": 1.1148338929485325e-05, "loss": 0.0102, "step": 63530 }, { "grad_norm": 0.1508573442697525, "learning_rate": 1.113533234382304e-05, "loss": 0.0092, "step": 63540 }, { "grad_norm": 0.16846680641174316, "learning_rate": 1.1122332399069513e-05, "loss": 0.0124, "step": 63550 }, { "grad_norm": 0.22154898941516876, "learning_rate": 1.1109339097446047e-05, "loss": 0.0144, "step": 63560 }, { "grad_norm": 0.3109751343727112, "learning_rate": 1.1096352441172897e-05, "loss": 0.016, "step": 63570 }, { "grad_norm": 0.5457671880722046, "learning_rate": 1.1083372432469086e-05, "loss": 0.0074, "step": 63580 }, { "grad_norm": 0.1537274569272995, "learning_rate": 1.107039907355255e-05, "loss": 0.0104, "step": 63590 }, { "grad_norm": 0.18836885690689087, "learning_rate": 1.1057432366640103e-05, "loss": 0.0124, "step": 63600 }, { "grad_norm": 0.3387240171432495, "learning_rate": 1.1044472313947352e-05, "loss": 0.0143, "step": 63610 }, { "grad_norm": 0.24758820235729218, "learning_rate": 1.1031518917688877e-05, "loss": 0.0113, "step": 63620 }, { "grad_norm": 0.16558772325515747, "learning_rate": 1.1018572180078007e-05, "loss": 0.0138, "step": 63630 }, { "grad_norm": 0.22359329462051392, "learning_rate": 1.1005632103327018e-05, "loss": 0.0104, "step": 63640 }, { "grad_norm": 0.3826705515384674, "learning_rate": 1.0992698689646996e-05, "loss": 0.0122, "step": 63650 }, { "grad_norm": 0.1849268078804016, "learning_rate": 1.0979771941247919e-05, "loss": 0.0109, "step": 63660 }, { "grad_norm": 0.23389746248722076, "learning_rate": 1.0966851860338611e-05, "loss": 0.0111, "step": 63670 }, { "grad_norm": 0.11222468316555023, "learning_rate": 1.0953938449126766e-05, "loss": 0.0089, "step": 63680 }, { "grad_norm": 0.42023253440856934, "learning_rate": 1.0941031709818933e-05, "loss": 0.0138, "step": 63690 }, { "grad_norm": 0.21169191598892212, "learning_rate": 1.0928131644620509e-05, "loss": 0.0136, "step": 63700 }, { "grad_norm": 0.24879950284957886, "learning_rate": 1.0915238255735766e-05, "loss": 0.0122, "step": 63710 }, { "grad_norm": 0.09995061159133911, "learning_rate": 1.0902351545367833e-05, "loss": 0.0095, "step": 63720 }, { "grad_norm": 0.1927245706319809, "learning_rate": 1.0889471515718702e-05, "loss": 0.0073, "step": 63730 }, { "grad_norm": 0.3457117974758148, "learning_rate": 1.087659816898921e-05, "loss": 0.009, "step": 63740 }, { "grad_norm": 0.309980183839798, "learning_rate": 1.0863731507379082e-05, "loss": 0.0102, "step": 63750 }, { "grad_norm": 0.27397850155830383, "learning_rate": 1.0850871533086827e-05, "loss": 0.0126, "step": 63760 }, { "grad_norm": 0.40710020065307617, "learning_rate": 1.0838018248309927e-05, "loss": 0.0082, "step": 63770 }, { "grad_norm": 0.43857088685035706, "learning_rate": 1.0825171655244615e-05, "loss": 0.0118, "step": 63780 }, { "grad_norm": 0.14354175329208374, "learning_rate": 1.0812331756086025e-05, "loss": 0.009, "step": 63790 }, { "grad_norm": 0.2670125365257263, "learning_rate": 1.079949855302817e-05, "loss": 0.0093, "step": 63800 }, { "grad_norm": 0.23847652971744537, "learning_rate": 1.0786672048263852e-05, "loss": 0.0132, "step": 63810 }, { "grad_norm": 0.25214093923568726, "learning_rate": 1.0773852243984817e-05, "loss": 0.0147, "step": 63820 }, { "grad_norm": 0.23098458349704742, "learning_rate": 1.0761039142381586e-05, "loss": 0.0147, "step": 63830 }, { "grad_norm": 0.2099635899066925, "learning_rate": 1.0748232745643577e-05, "loss": 0.0121, "step": 63840 }, { "grad_norm": 0.1938861757516861, "learning_rate": 1.0735433055959055e-05, "loss": 0.0094, "step": 63850 }, { "grad_norm": 0.2340002804994583, "learning_rate": 1.0722640075515133e-05, "loss": 0.0085, "step": 63860 }, { "grad_norm": 0.1597401350736618, "learning_rate": 1.0709853806497795e-05, "loss": 0.012, "step": 63870 }, { "grad_norm": 0.13020840287208557, "learning_rate": 1.0697074251091831e-05, "loss": 0.0127, "step": 63880 }, { "grad_norm": 0.11004524677991867, "learning_rate": 1.0684301411480962e-05, "loss": 0.0125, "step": 63890 }, { "grad_norm": 0.17649491131305695, "learning_rate": 1.067153528984769e-05, "loss": 0.009, "step": 63900 }, { "grad_norm": 0.2231227159500122, "learning_rate": 1.0658775888373395e-05, "loss": 0.0119, "step": 63910 }, { "grad_norm": 0.15699081122875214, "learning_rate": 1.0646023209238314e-05, "loss": 0.015, "step": 63920 }, { "grad_norm": 0.09932376444339752, "learning_rate": 1.0633277254621537e-05, "loss": 0.0094, "step": 63930 }, { "grad_norm": 0.1402830332517624, "learning_rate": 1.0620538026700994e-05, "loss": 0.0091, "step": 63940 }, { "grad_norm": 0.1503482162952423, "learning_rate": 1.0607805527653486e-05, "loss": 0.0105, "step": 63950 }, { "grad_norm": 0.12594284117221832, "learning_rate": 1.059507975965462e-05, "loss": 0.0114, "step": 63960 }, { "grad_norm": 0.2086375206708908, "learning_rate": 1.0582360724878898e-05, "loss": 0.0077, "step": 63970 }, { "grad_norm": 0.502201497554779, "learning_rate": 1.0569648425499651e-05, "loss": 0.0086, "step": 63980 }, { "grad_norm": 0.34751763939857483, "learning_rate": 1.0556942863689063e-05, "loss": 0.0089, "step": 63990 }, { "grad_norm": 0.20418718457221985, "learning_rate": 1.054424404161819e-05, "loss": 0.0106, "step": 64000 }, { "grad_norm": 0.16158214211463928, "learning_rate": 1.053155196145686e-05, "loss": 0.0112, "step": 64010 }, { "grad_norm": 0.25722989439964294, "learning_rate": 1.0518866625373863e-05, "loss": 0.0094, "step": 64020 }, { "grad_norm": 0.1597861498594284, "learning_rate": 1.0506188035536735e-05, "loss": 0.0087, "step": 64030 }, { "grad_norm": 0.16467121243476868, "learning_rate": 1.0493516194111919e-05, "loss": 0.0086, "step": 64040 }, { "grad_norm": 0.15758520364761353, "learning_rate": 1.0480851103264688e-05, "loss": 0.0081, "step": 64050 }, { "grad_norm": 0.13664206862449646, "learning_rate": 1.046819276515913e-05, "loss": 0.007, "step": 64060 }, { "grad_norm": 0.35310229659080505, "learning_rate": 1.0455541181958256e-05, "loss": 0.0101, "step": 64070 }, { "grad_norm": 0.12816397845745087, "learning_rate": 1.0442896355823822e-05, "loss": 0.0066, "step": 64080 }, { "grad_norm": 0.17127026617527008, "learning_rate": 1.0430258288916539e-05, "loss": 0.0102, "step": 64090 }, { "grad_norm": 0.1427111029624939, "learning_rate": 1.0417626983395868e-05, "loss": 0.0082, "step": 64100 }, { "grad_norm": 0.14628472924232483, "learning_rate": 1.0405002441420165e-05, "loss": 0.0084, "step": 64110 }, { "grad_norm": 0.23245562613010406, "learning_rate": 1.039238466514662e-05, "loss": 0.0098, "step": 64120 }, { "grad_norm": 0.14339619874954224, "learning_rate": 1.0379773656731262e-05, "loss": 0.014, "step": 64130 }, { "grad_norm": 0.2107582539319992, "learning_rate": 1.0367169418328986e-05, "loss": 0.0081, "step": 64140 }, { "grad_norm": 0.5213342308998108, "learning_rate": 1.0354571952093484e-05, "loss": 0.0121, "step": 64150 }, { "grad_norm": 0.208945170044899, "learning_rate": 1.034198126017733e-05, "loss": 0.0108, "step": 64160 }, { "grad_norm": 0.22536158561706543, "learning_rate": 1.0329397344731928e-05, "loss": 0.0116, "step": 64170 }, { "grad_norm": 0.3712112605571747, "learning_rate": 1.0316820207907524e-05, "loss": 0.0129, "step": 64180 }, { "grad_norm": 0.13092166185379028, "learning_rate": 1.0304249851853215e-05, "loss": 0.012, "step": 64190 }, { "grad_norm": 0.23530475795269012, "learning_rate": 1.0291686278716933e-05, "loss": 0.0105, "step": 64200 }, { "grad_norm": 0.2610273063182831, "learning_rate": 1.0279129490645418e-05, "loss": 0.0108, "step": 64210 }, { "grad_norm": 0.11056039482355118, "learning_rate": 1.0266579489784328e-05, "loss": 0.0098, "step": 64220 }, { "grad_norm": 0.21040791273117065, "learning_rate": 1.0254036278278084e-05, "loss": 0.013, "step": 64230 }, { "grad_norm": 0.21417780220508575, "learning_rate": 1.0241499858269982e-05, "loss": 0.0086, "step": 64240 }, { "grad_norm": 0.2210068553686142, "learning_rate": 1.0228970231902169e-05, "loss": 0.0099, "step": 64250 }, { "grad_norm": 0.15483996272087097, "learning_rate": 1.0216447401315582e-05, "loss": 0.0097, "step": 64260 }, { "grad_norm": 0.2072867453098297, "learning_rate": 1.020393136865007e-05, "loss": 0.0084, "step": 64270 }, { "grad_norm": 0.29523253440856934, "learning_rate": 1.0191422136044242e-05, "loss": 0.0155, "step": 64280 }, { "grad_norm": 0.141886368393898, "learning_rate": 1.017891970563563e-05, "loss": 0.0097, "step": 64290 }, { "grad_norm": 0.17513394355773926, "learning_rate": 1.0166424079560516e-05, "loss": 0.0131, "step": 64300 }, { "grad_norm": 0.17570585012435913, "learning_rate": 1.0153935259954078e-05, "loss": 0.0109, "step": 64310 }, { "grad_norm": 0.5170493721961975, "learning_rate": 1.0141453248950311e-05, "loss": 0.0113, "step": 64320 }, { "grad_norm": 0.5671750903129578, "learning_rate": 1.0128978048682054e-05, "loss": 0.0111, "step": 64330 }, { "grad_norm": 0.34579238295555115, "learning_rate": 1.0116509661280982e-05, "loss": 0.0135, "step": 64340 }, { "grad_norm": 0.13595975935459137, "learning_rate": 1.0104048088877576e-05, "loss": 0.0091, "step": 64350 }, { "grad_norm": 0.1749090701341629, "learning_rate": 1.0091593333601201e-05, "loss": 0.0102, "step": 64360 }, { "grad_norm": 0.19008533656597137, "learning_rate": 1.0079145397580031e-05, "loss": 0.0102, "step": 64370 }, { "grad_norm": 0.35254740715026855, "learning_rate": 1.006670428294107e-05, "loss": 0.0129, "step": 64380 }, { "grad_norm": 0.0779842957854271, "learning_rate": 1.0054269991810166e-05, "loss": 0.0095, "step": 64390 }, { "grad_norm": 0.12973038852214813, "learning_rate": 1.0041842526312024e-05, "loss": 0.0117, "step": 64400 }, { "grad_norm": 0.10497037321329117, "learning_rate": 1.0029421888570101e-05, "loss": 0.0094, "step": 64410 }, { "grad_norm": 0.13430911302566528, "learning_rate": 1.0017008080706813e-05, "loss": 0.0221, "step": 64420 }, { "grad_norm": 0.1592235118150711, "learning_rate": 1.0004601104843287e-05, "loss": 0.0096, "step": 64430 }, { "grad_norm": 0.283153235912323, "learning_rate": 9.992200963099562e-06, "loss": 0.0146, "step": 64440 }, { "grad_norm": 0.13982057571411133, "learning_rate": 9.979807657594486e-06, "loss": 0.009, "step": 64450 }, { "grad_norm": 0.22004057466983795, "learning_rate": 9.967421190445703e-06, "loss": 0.0092, "step": 64460 }, { "grad_norm": 0.1820422261953354, "learning_rate": 9.955041563769769e-06, "loss": 0.0132, "step": 64470 }, { "grad_norm": 0.17716772854328156, "learning_rate": 9.942668779681974e-06, "loss": 0.013, "step": 64480 }, { "grad_norm": 0.29751336574554443, "learning_rate": 9.930302840296541e-06, "loss": 0.0098, "step": 64490 }, { "grad_norm": 0.19100341200828552, "learning_rate": 9.917943747726426e-06, "loss": 0.0092, "step": 64500 }, { "grad_norm": 0.31505969166755676, "learning_rate": 9.905591504083484e-06, "loss": 0.0115, "step": 64510 }, { "grad_norm": 0.18408726155757904, "learning_rate": 9.893246111478382e-06, "loss": 0.0073, "step": 64520 }, { "grad_norm": 0.3570125102996826, "learning_rate": 9.88090757202057e-06, "loss": 0.0091, "step": 64530 }, { "grad_norm": 0.18701227009296417, "learning_rate": 9.868575887818421e-06, "loss": 0.0117, "step": 64540 }, { "grad_norm": 0.17957329750061035, "learning_rate": 9.856251060979044e-06, "loss": 0.0112, "step": 64550 }, { "grad_norm": 0.14611981809139252, "learning_rate": 9.843933093608426e-06, "loss": 0.0126, "step": 64560 }, { "grad_norm": 0.08558999001979828, "learning_rate": 9.831621987811368e-06, "loss": 0.0103, "step": 64570 }, { "grad_norm": 0.24047741293907166, "learning_rate": 9.819317745691509e-06, "loss": 0.0107, "step": 64580 }, { "grad_norm": 0.21746262907981873, "learning_rate": 9.8070203693513e-06, "loss": 0.0084, "step": 64590 }, { "grad_norm": 0.27070069313049316, "learning_rate": 9.794729860892048e-06, "loss": 0.0089, "step": 64600 }, { "grad_norm": 0.23819787800312042, "learning_rate": 9.782446222413827e-06, "loss": 0.018, "step": 64610 }, { "grad_norm": 0.12591615319252014, "learning_rate": 9.770169456015598e-06, "loss": 0.0134, "step": 64620 }, { "grad_norm": 0.12070401012897491, "learning_rate": 9.75789956379512e-06, "loss": 0.0116, "step": 64630 }, { "grad_norm": 0.2032136619091034, "learning_rate": 9.74563654784898e-06, "loss": 0.0108, "step": 64640 }, { "grad_norm": 0.2048526406288147, "learning_rate": 9.733380410272596e-06, "loss": 0.0107, "step": 64650 }, { "grad_norm": 0.28415530920028687, "learning_rate": 9.721131153160207e-06, "loss": 0.0104, "step": 64660 }, { "grad_norm": 0.12745286524295807, "learning_rate": 9.708888778604886e-06, "loss": 0.0092, "step": 64670 }, { "grad_norm": 0.1635553538799286, "learning_rate": 9.696653288698488e-06, "loss": 0.0097, "step": 64680 }, { "grad_norm": 0.1565561443567276, "learning_rate": 9.684424685531763e-06, "loss": 0.0198, "step": 64690 }, { "grad_norm": 0.27921533584594727, "learning_rate": 9.672202971194216e-06, "loss": 0.0106, "step": 64700 }, { "grad_norm": 0.2310006320476532, "learning_rate": 9.659988147774213e-06, "loss": 0.0117, "step": 64710 }, { "grad_norm": 0.16588278114795685, "learning_rate": 9.647780217358942e-06, "loss": 0.0102, "step": 64720 }, { "grad_norm": 0.0944657102227211, "learning_rate": 9.635579182034376e-06, "loss": 0.0097, "step": 64730 }, { "grad_norm": 0.12483594566583633, "learning_rate": 9.623385043885386e-06, "loss": 0.0144, "step": 64740 }, { "grad_norm": 0.167876198887825, "learning_rate": 9.61119780499557e-06, "loss": 0.0071, "step": 64750 }, { "grad_norm": 0.20984289050102234, "learning_rate": 9.599017467447418e-06, "loss": 0.0101, "step": 64760 }, { "grad_norm": 0.1431494802236557, "learning_rate": 9.586844033322206e-06, "loss": 0.0091, "step": 64770 }, { "grad_norm": 0.3704933822154999, "learning_rate": 9.574677504700052e-06, "loss": 0.0149, "step": 64780 }, { "grad_norm": 0.2578992247581482, "learning_rate": 9.562517883659877e-06, "loss": 0.0204, "step": 64790 }, { "grad_norm": 0.16623297333717346, "learning_rate": 9.55036517227944e-06, "loss": 0.0108, "step": 64800 }, { "grad_norm": 0.1518038958311081, "learning_rate": 9.538219372635282e-06, "loss": 0.0097, "step": 64810 }, { "grad_norm": 0.19514884054660797, "learning_rate": 9.526080486802802e-06, "loss": 0.0081, "step": 64820 }, { "grad_norm": 0.13729766011238098, "learning_rate": 9.513948516856203e-06, "loss": 0.01, "step": 64830 }, { "grad_norm": 0.16066071391105652, "learning_rate": 9.501823464868504e-06, "loss": 0.0085, "step": 64840 }, { "grad_norm": 0.16637979447841644, "learning_rate": 9.489705332911547e-06, "loss": 0.0115, "step": 64850 }, { "grad_norm": 0.20219561457633972, "learning_rate": 9.477594123055994e-06, "loss": 0.0095, "step": 64860 }, { "grad_norm": 0.4020097553730011, "learning_rate": 9.465489837371321e-06, "loss": 0.0155, "step": 64870 }, { "grad_norm": 0.14378398656845093, "learning_rate": 9.453392477925794e-06, "loss": 0.0085, "step": 64880 }, { "grad_norm": 0.19319573044776917, "learning_rate": 9.441302046786566e-06, "loss": 0.0099, "step": 64890 }, { "grad_norm": 0.10309503972530365, "learning_rate": 9.429218546019519e-06, "loss": 0.0066, "step": 64900 }, { "grad_norm": 0.26214122772216797, "learning_rate": 9.41714197768941e-06, "loss": 0.0156, "step": 64910 }, { "grad_norm": 0.27625522017478943, "learning_rate": 9.405072343859805e-06, "loss": 0.0071, "step": 64920 }, { "grad_norm": 0.19278353452682495, "learning_rate": 9.393009646593043e-06, "loss": 0.0098, "step": 64930 }, { "grad_norm": 0.11602912843227386, "learning_rate": 9.38095388795035e-06, "loss": 0.0081, "step": 64940 }, { "grad_norm": 0.3378556966781616, "learning_rate": 9.36890506999169e-06, "loss": 0.0091, "step": 64950 }, { "grad_norm": 0.21417589485645294, "learning_rate": 9.356863194775894e-06, "loss": 0.0103, "step": 64960 }, { "grad_norm": 0.1458791047334671, "learning_rate": 9.344828264360583e-06, "loss": 0.0134, "step": 64970 }, { "grad_norm": 0.2469518631696701, "learning_rate": 9.332800280802201e-06, "loss": 0.0157, "step": 64980 }, { "grad_norm": 0.20288139581680298, "learning_rate": 9.32077924615602e-06, "loss": 0.0077, "step": 64990 }, { "grad_norm": 0.16091614961624146, "learning_rate": 9.308765162476063e-06, "loss": 0.0091, "step": 65000 }, { "grad_norm": 0.09356803447008133, "learning_rate": 9.296758031815239e-06, "loss": 0.011, "step": 65010 }, { "grad_norm": 0.3109152317047119, "learning_rate": 9.284757856225229e-06, "loss": 0.0091, "step": 65020 }, { "grad_norm": 0.20282557606697083, "learning_rate": 9.272764637756538e-06, "loss": 0.0133, "step": 65030 }, { "grad_norm": 0.1813976764678955, "learning_rate": 9.260778378458479e-06, "loss": 0.0132, "step": 65040 }, { "grad_norm": 0.22825802862644196, "learning_rate": 9.248799080379172e-06, "loss": 0.0099, "step": 65050 }, { "grad_norm": 0.18905974924564362, "learning_rate": 9.236826745565558e-06, "loss": 0.0087, "step": 65060 }, { "grad_norm": 0.188119575381279, "learning_rate": 9.224861376063388e-06, "loss": 0.0111, "step": 65070 }, { "grad_norm": 0.10951641947031021, "learning_rate": 9.212902973917192e-06, "loss": 0.0085, "step": 65080 }, { "grad_norm": 0.18506504595279694, "learning_rate": 9.20095154117035e-06, "loss": 0.008, "step": 65090 }, { "grad_norm": 0.19266356527805328, "learning_rate": 9.189007079865036e-06, "loss": 0.0079, "step": 65100 }, { "grad_norm": 0.15661858022212982, "learning_rate": 9.177069592042226e-06, "loss": 0.0072, "step": 65110 }, { "grad_norm": 0.2157014161348343, "learning_rate": 9.165139079741724e-06, "loss": 0.012, "step": 65120 }, { "grad_norm": 0.14790765941143036, "learning_rate": 9.153215545002098e-06, "loss": 0.016, "step": 65130 }, { "grad_norm": 0.17003858089447021, "learning_rate": 9.141298989860798e-06, "loss": 0.0141, "step": 65140 }, { "grad_norm": 0.4030402600765228, "learning_rate": 9.129389416353994e-06, "loss": 0.0076, "step": 65150 }, { "grad_norm": 0.14473824203014374, "learning_rate": 9.11748682651673e-06, "loss": 0.0076, "step": 65160 }, { "grad_norm": 0.2315686047077179, "learning_rate": 9.105591222382837e-06, "loss": 0.0079, "step": 65170 }, { "grad_norm": 0.15306277573108673, "learning_rate": 9.093702605984915e-06, "loss": 0.0094, "step": 65180 }, { "grad_norm": 0.2487597018480301, "learning_rate": 9.081820979354455e-06, "loss": 0.0114, "step": 65190 }, { "grad_norm": 0.252005934715271, "learning_rate": 9.069946344521663e-06, "loss": 0.009, "step": 65200 }, { "grad_norm": 0.13777436316013336, "learning_rate": 9.058078703515598e-06, "loss": 0.0095, "step": 65210 }, { "grad_norm": 0.25995543599128723, "learning_rate": 9.046218058364125e-06, "loss": 0.0191, "step": 65220 }, { "grad_norm": 0.16885091364383698, "learning_rate": 9.034364411093893e-06, "loss": 0.0105, "step": 65230 }, { "grad_norm": 0.1495903581380844, "learning_rate": 9.022517763730371e-06, "loss": 0.0115, "step": 65240 }, { "grad_norm": 0.15769653022289276, "learning_rate": 9.010678118297827e-06, "loss": 0.0102, "step": 65250 }, { "grad_norm": 0.3175254762172699, "learning_rate": 8.998845476819345e-06, "loss": 0.0108, "step": 65260 }, { "grad_norm": 0.2419050931930542, "learning_rate": 8.987019841316773e-06, "loss": 0.0098, "step": 65270 }, { "grad_norm": 0.7044961452484131, "learning_rate": 8.975201213810802e-06, "loss": 0.0097, "step": 65280 }, { "grad_norm": 0.13266998529434204, "learning_rate": 8.963389596320915e-06, "loss": 0.0084, "step": 65290 }, { "grad_norm": 0.2663775086402893, "learning_rate": 8.951584990865391e-06, "loss": 0.0118, "step": 65300 }, { "grad_norm": 0.33853501081466675, "learning_rate": 8.939787399461319e-06, "loss": 0.0102, "step": 65310 }, { "grad_norm": 0.6165851354598999, "learning_rate": 8.927996824124591e-06, "loss": 0.014, "step": 65320 }, { "grad_norm": 0.49104043841362, "learning_rate": 8.916213266869854e-06, "loss": 0.0137, "step": 65330 }, { "grad_norm": 0.11667746305465698, "learning_rate": 8.904436729710658e-06, "loss": 0.0087, "step": 65340 }, { "grad_norm": 0.1777881234884262, "learning_rate": 8.892667214659245e-06, "loss": 0.0069, "step": 65350 }, { "grad_norm": 0.23171886801719666, "learning_rate": 8.880904723726713e-06, "loss": 0.0095, "step": 65360 }, { "grad_norm": 0.12236964702606201, "learning_rate": 8.869149258922971e-06, "loss": 0.0051, "step": 65370 }, { "grad_norm": 0.15857161581516266, "learning_rate": 8.857400822256662e-06, "loss": 0.008, "step": 65380 }, { "grad_norm": 0.27879276871681213, "learning_rate": 8.845659415735324e-06, "loss": 0.0136, "step": 65390 }, { "grad_norm": 0.20310384035110474, "learning_rate": 8.83392504136521e-06, "loss": 0.0097, "step": 65400 }, { "grad_norm": 0.08267819136381149, "learning_rate": 8.822197701151407e-06, "loss": 0.0104, "step": 65410 }, { "grad_norm": 0.12312749028205872, "learning_rate": 8.810477397097804e-06, "loss": 0.0142, "step": 65420 }, { "grad_norm": 0.24808774888515472, "learning_rate": 8.798764131207077e-06, "loss": 0.0082, "step": 65430 }, { "grad_norm": 0.309712678194046, "learning_rate": 8.787057905480706e-06, "loss": 0.0107, "step": 65440 }, { "grad_norm": 0.2533702850341797, "learning_rate": 8.775358721918958e-06, "loss": 0.0094, "step": 65450 }, { "grad_norm": 0.19377471506595612, "learning_rate": 8.763666582520923e-06, "loss": 0.009, "step": 65460 }, { "grad_norm": 0.1472829431295395, "learning_rate": 8.751981489284445e-06, "loss": 0.0078, "step": 65470 }, { "grad_norm": 0.234577476978302, "learning_rate": 8.740303444206188e-06, "loss": 0.0105, "step": 65480 }, { "grad_norm": 0.1316896378993988, "learning_rate": 8.72863244928162e-06, "loss": 0.0086, "step": 65490 }, { "grad_norm": 0.1926126927137375, "learning_rate": 8.716968506504991e-06, "loss": 0.016, "step": 65500 }, { "grad_norm": 0.24645796418190002, "learning_rate": 8.70531161786935e-06, "loss": 0.011, "step": 65510 }, { "grad_norm": 0.14098794758319855, "learning_rate": 8.693661785366558e-06, "loss": 0.0128, "step": 65520 }, { "grad_norm": 0.191803440451622, "learning_rate": 8.682019010987208e-06, "loss": 0.0124, "step": 65530 }, { "grad_norm": 0.1251230686903, "learning_rate": 8.670383296720786e-06, "loss": 0.0094, "step": 65540 }, { "grad_norm": 0.1414685696363449, "learning_rate": 8.658754644555478e-06, "loss": 0.0084, "step": 65550 }, { "grad_norm": 0.16212499141693115, "learning_rate": 8.647133056478313e-06, "loss": 0.01, "step": 65560 }, { "grad_norm": 0.1689724624156952, "learning_rate": 8.635518534475123e-06, "loss": 0.0082, "step": 65570 }, { "grad_norm": 0.08429986238479614, "learning_rate": 8.623911080530467e-06, "loss": 0.0101, "step": 65580 }, { "grad_norm": 0.14860057830810547, "learning_rate": 8.61231069662779e-06, "loss": 0.0091, "step": 65590 }, { "grad_norm": 0.09835105389356613, "learning_rate": 8.600717384749252e-06, "loss": 0.0093, "step": 65600 }, { "grad_norm": 0.19089363515377045, "learning_rate": 8.58913114687584e-06, "loss": 0.0091, "step": 65610 }, { "grad_norm": 0.34620288014411926, "learning_rate": 8.57755198498732e-06, "loss": 0.0105, "step": 65620 }, { "grad_norm": 0.22102366387844086, "learning_rate": 8.565979901062265e-06, "loss": 0.0139, "step": 65630 }, { "grad_norm": 0.10776107758283615, "learning_rate": 8.554414897078033e-06, "loss": 0.0089, "step": 65640 }, { "grad_norm": 0.20212596654891968, "learning_rate": 8.542856975010727e-06, "loss": 0.0084, "step": 65650 }, { "grad_norm": 0.10644689202308655, "learning_rate": 8.531306136835337e-06, "loss": 0.0084, "step": 65660 }, { "grad_norm": 0.20057080686092377, "learning_rate": 8.519762384525548e-06, "loss": 0.0113, "step": 65670 }, { "grad_norm": 0.19238369166851044, "learning_rate": 8.508225720053875e-06, "loss": 0.01, "step": 65680 }, { "grad_norm": 0.14249476790428162, "learning_rate": 8.496696145391625e-06, "loss": 0.0093, "step": 65690 }, { "grad_norm": 0.11497490108013153, "learning_rate": 8.485173662508889e-06, "loss": 0.0124, "step": 65700 }, { "grad_norm": 0.1538129299879074, "learning_rate": 8.473658273374536e-06, "loss": 0.0087, "step": 65710 }, { "grad_norm": 0.24315069615840912, "learning_rate": 8.462149979956253e-06, "loss": 0.0079, "step": 65720 }, { "grad_norm": 0.09034747630357742, "learning_rate": 8.450648784220461e-06, "loss": 0.0129, "step": 65730 }, { "grad_norm": 0.27833661437034607, "learning_rate": 8.439154688132417e-06, "loss": 0.0105, "step": 65740 }, { "grad_norm": 0.20267564058303833, "learning_rate": 8.427667693656143e-06, "loss": 0.01, "step": 65750 }, { "grad_norm": 0.17417530715465546, "learning_rate": 8.416187802754454e-06, "loss": 0.0114, "step": 65760 }, { "grad_norm": 0.1390613317489624, "learning_rate": 8.404715017388965e-06, "loss": 0.0085, "step": 65770 }, { "grad_norm": 0.4826556146144867, "learning_rate": 8.393249339520015e-06, "loss": 0.0103, "step": 65780 }, { "grad_norm": 0.2788810431957245, "learning_rate": 8.381790771106834e-06, "loss": 0.0112, "step": 65790 }, { "grad_norm": 0.31129443645477295, "learning_rate": 8.370339314107339e-06, "loss": 0.009, "step": 65800 }, { "grad_norm": 0.2530806362628937, "learning_rate": 8.358894970478281e-06, "loss": 0.009, "step": 65810 }, { "grad_norm": 0.1949893981218338, "learning_rate": 8.347457742175196e-06, "loss": 0.0113, "step": 65820 }, { "grad_norm": 0.19289280474185944, "learning_rate": 8.33602763115236e-06, "loss": 0.0081, "step": 65830 }, { "grad_norm": 0.12866389751434326, "learning_rate": 8.324604639362916e-06, "loss": 0.0095, "step": 65840 }, { "grad_norm": 0.22758392989635468, "learning_rate": 8.31318876875869e-06, "loss": 0.009, "step": 65850 }, { "grad_norm": 0.1944243311882019, "learning_rate": 8.30178002129039e-06, "loss": 0.0079, "step": 65860 }, { "grad_norm": 0.2144709974527359, "learning_rate": 8.290378398907423e-06, "loss": 0.0095, "step": 65870 }, { "grad_norm": 0.18531273305416107, "learning_rate": 8.278983903558029e-06, "loss": 0.0099, "step": 65880 }, { "grad_norm": 0.22361619770526886, "learning_rate": 8.26759653718921e-06, "loss": 0.0083, "step": 65890 }, { "grad_norm": 0.12024220079183578, "learning_rate": 8.25621630174676e-06, "loss": 0.0103, "step": 65900 }, { "grad_norm": 0.17747199535369873, "learning_rate": 8.244843199175261e-06, "loss": 0.0132, "step": 65910 }, { "grad_norm": 0.15494532883167267, "learning_rate": 8.233477231418046e-06, "loss": 0.0071, "step": 65920 }, { "grad_norm": 0.17272140085697174, "learning_rate": 8.22211840041725e-06, "loss": 0.0145, "step": 65930 }, { "grad_norm": 0.27778518199920654, "learning_rate": 8.210766708113792e-06, "loss": 0.0156, "step": 65940 }, { "grad_norm": 0.23459811508655548, "learning_rate": 8.199422156447367e-06, "loss": 0.0086, "step": 65950 }, { "grad_norm": 0.20878487825393677, "learning_rate": 8.188084747356451e-06, "loss": 0.0149, "step": 65960 }, { "grad_norm": 0.21821224689483643, "learning_rate": 8.176754482778299e-06, "loss": 0.01, "step": 65970 }, { "grad_norm": 0.2349870502948761, "learning_rate": 8.165431364648918e-06, "loss": 0.0128, "step": 65980 }, { "grad_norm": 0.11874067038297653, "learning_rate": 8.154115394903162e-06, "loss": 0.009, "step": 65990 }, { "grad_norm": 0.22625508904457092, "learning_rate": 8.142806575474582e-06, "loss": 0.0096, "step": 66000 }, { "grad_norm": 0.1640658676624298, "learning_rate": 8.131504908295562e-06, "loss": 0.0083, "step": 66010 }, { "grad_norm": 0.23901870846748352, "learning_rate": 8.120210395297262e-06, "loss": 0.0076, "step": 66020 }, { "grad_norm": 0.29663875699043274, "learning_rate": 8.108923038409565e-06, "loss": 0.0098, "step": 66030 }, { "grad_norm": 0.19822782278060913, "learning_rate": 8.097642839561226e-06, "loss": 0.0095, "step": 66040 }, { "grad_norm": 0.14865955710411072, "learning_rate": 8.08636980067966e-06, "loss": 0.0087, "step": 66050 }, { "grad_norm": 0.1739751100540161, "learning_rate": 8.075103923691186e-06, "loss": 0.0077, "step": 66060 }, { "grad_norm": 0.13410115242004395, "learning_rate": 8.063845210520793e-06, "loss": 0.0089, "step": 66070 }, { "grad_norm": 0.1473112255334854, "learning_rate": 8.052593663092295e-06, "loss": 0.0101, "step": 66080 }, { "grad_norm": 0.2420368492603302, "learning_rate": 8.04134928332827e-06, "loss": 0.013, "step": 66090 }, { "grad_norm": 0.1757057160139084, "learning_rate": 8.030112073150086e-06, "loss": 0.0149, "step": 66100 }, { "grad_norm": 0.13579347729682922, "learning_rate": 8.018882034477881e-06, "loss": 0.0087, "step": 66110 }, { "grad_norm": 0.1405356228351593, "learning_rate": 8.007659169230541e-06, "loss": 0.0084, "step": 66120 }, { "grad_norm": 0.15850988030433655, "learning_rate": 7.996443479325755e-06, "loss": 0.0082, "step": 66130 }, { "grad_norm": 0.1984945833683014, "learning_rate": 7.985234966679977e-06, "loss": 0.0125, "step": 66140 }, { "grad_norm": 0.14524152874946594, "learning_rate": 7.974033633208438e-06, "loss": 0.0072, "step": 66150 }, { "grad_norm": 0.4016936123371124, "learning_rate": 7.962839480825135e-06, "loss": 0.0104, "step": 66160 }, { "grad_norm": 0.3153694272041321, "learning_rate": 7.951652511442858e-06, "loss": 0.0095, "step": 66170 }, { "grad_norm": 0.14815868437290192, "learning_rate": 7.940472726973125e-06, "loss": 0.0134, "step": 66180 }, { "grad_norm": 0.19497057795524597, "learning_rate": 7.929300129326289e-06, "loss": 0.0083, "step": 66190 }, { "grad_norm": 0.21398867666721344, "learning_rate": 7.91813472041142e-06, "loss": 0.0093, "step": 66200 }, { "grad_norm": 0.10756704211235046, "learning_rate": 7.906976502136376e-06, "loss": 0.0071, "step": 66210 }, { "grad_norm": 0.19955432415008545, "learning_rate": 7.89582547640782e-06, "loss": 0.0132, "step": 66220 }, { "grad_norm": 0.11512919515371323, "learning_rate": 7.884681645131115e-06, "loss": 0.0085, "step": 66230 }, { "grad_norm": 0.3142392635345459, "learning_rate": 7.87354501021048e-06, "loss": 0.009, "step": 66240 }, { "grad_norm": 0.12262119352817535, "learning_rate": 7.86241557354882e-06, "loss": 0.0135, "step": 66250 }, { "grad_norm": 0.39132967591285706, "learning_rate": 7.8512933370479e-06, "loss": 0.0105, "step": 66260 }, { "grad_norm": 0.1987254023551941, "learning_rate": 7.840178302608158e-06, "loss": 0.007, "step": 66270 }, { "grad_norm": 0.0897529199719429, "learning_rate": 7.82907047212888e-06, "loss": 0.0141, "step": 66280 }, { "grad_norm": 0.2229047566652298, "learning_rate": 7.81796984750809e-06, "loss": 0.0126, "step": 66290 }, { "grad_norm": 0.1546904295682907, "learning_rate": 7.806876430642546e-06, "loss": 0.0111, "step": 66300 }, { "grad_norm": 0.2032136768102646, "learning_rate": 7.795790223427862e-06, "loss": 0.0113, "step": 66310 }, { "grad_norm": 0.24261131882667542, "learning_rate": 7.784711227758324e-06, "loss": 0.0109, "step": 66320 }, { "grad_norm": 0.19831611216068268, "learning_rate": 7.773639445527053e-06, "loss": 0.0091, "step": 66330 }, { "grad_norm": 0.23267507553100586, "learning_rate": 7.762574878625905e-06, "loss": 0.0095, "step": 66340 }, { "grad_norm": 0.09878482669591904, "learning_rate": 7.751517528945513e-06, "loss": 0.0092, "step": 66350 }, { "grad_norm": 0.1555667668581009, "learning_rate": 7.740467398375278e-06, "loss": 0.0096, "step": 66360 }, { "grad_norm": 0.25034672021865845, "learning_rate": 7.729424488803378e-06, "loss": 0.0094, "step": 66370 }, { "grad_norm": 0.22973603010177612, "learning_rate": 7.71838880211671e-06, "loss": 0.012, "step": 66380 }, { "grad_norm": 0.16000379621982574, "learning_rate": 7.707360340200997e-06, "loss": 0.0083, "step": 66390 }, { "grad_norm": 0.1820489913225174, "learning_rate": 7.696339104940697e-06, "loss": 0.0108, "step": 66400 }, { "grad_norm": 0.1647074669599533, "learning_rate": 7.685325098219038e-06, "loss": 0.0103, "step": 66410 }, { "grad_norm": 0.15160568058490753, "learning_rate": 7.674318321918017e-06, "loss": 0.0134, "step": 66420 }, { "grad_norm": 0.2387358844280243, "learning_rate": 7.663318777918366e-06, "loss": 0.014, "step": 66430 }, { "grad_norm": 0.3591529428958893, "learning_rate": 7.652326468099647e-06, "loss": 0.0181, "step": 66440 }, { "grad_norm": 0.15998521447181702, "learning_rate": 7.641341394340096e-06, "loss": 0.0067, "step": 66450 }, { "grad_norm": 0.4125085175037384, "learning_rate": 7.630363558516818e-06, "loss": 0.0096, "step": 66460 }, { "grad_norm": 0.2472854107618332, "learning_rate": 7.619392962505578e-06, "loss": 0.0157, "step": 66470 }, { "grad_norm": 0.2801186442375183, "learning_rate": 7.6084296081809725e-06, "loss": 0.0105, "step": 66480 }, { "grad_norm": 0.2594738006591797, "learning_rate": 7.597473497416347e-06, "loss": 0.013, "step": 66490 }, { "grad_norm": 0.10132680088281631, "learning_rate": 7.586524632083764e-06, "loss": 0.0088, "step": 66500 }, { "grad_norm": 0.22584660351276398, "learning_rate": 7.5755830140541326e-06, "loss": 0.0122, "step": 66510 }, { "grad_norm": 0.22587618231773376, "learning_rate": 7.5646486451970475e-06, "loss": 0.0118, "step": 66520 }, { "grad_norm": 0.12201140820980072, "learning_rate": 7.553721527380897e-06, "loss": 0.0144, "step": 66530 }, { "grad_norm": 0.25284481048583984, "learning_rate": 7.542801662472826e-06, "loss": 0.0095, "step": 66540 }, { "grad_norm": 0.3032720386981964, "learning_rate": 7.5318890523387475e-06, "loss": 0.0086, "step": 66550 }, { "grad_norm": 0.16536051034927368, "learning_rate": 7.52098369884332e-06, "loss": 0.0092, "step": 66560 }, { "grad_norm": 0.16203181445598602, "learning_rate": 7.510085603849992e-06, "loss": 0.0108, "step": 66570 }, { "grad_norm": 0.19233913719654083, "learning_rate": 7.499194769220918e-06, "loss": 0.009, "step": 66580 }, { "grad_norm": 0.12089512497186661, "learning_rate": 7.48831119681706e-06, "loss": 0.0148, "step": 66590 }, { "grad_norm": 0.18029320240020752, "learning_rate": 7.477434888498119e-06, "loss": 0.0072, "step": 66600 }, { "grad_norm": 0.21234606206417084, "learning_rate": 7.466565846122564e-06, "loss": 0.0159, "step": 66610 }, { "grad_norm": 0.25485989451408386, "learning_rate": 7.455704071547626e-06, "loss": 0.0125, "step": 66620 }, { "grad_norm": 0.2755134701728821, "learning_rate": 7.444849566629247e-06, "loss": 0.0107, "step": 66630 }, { "grad_norm": 0.25895434617996216, "learning_rate": 7.434002333222212e-06, "loss": 0.0104, "step": 66640 }, { "grad_norm": 0.6393948197364807, "learning_rate": 7.423162373179976e-06, "loss": 0.0096, "step": 66650 }, { "grad_norm": 0.22229596972465515, "learning_rate": 7.412329688354835e-06, "loss": 0.0101, "step": 66660 }, { "grad_norm": 0.1732180267572403, "learning_rate": 7.40150428059776e-06, "loss": 0.0082, "step": 66670 }, { "grad_norm": 0.2696074843406677, "learning_rate": 7.3906861517585354e-06, "loss": 0.0103, "step": 66680 }, { "grad_norm": 0.33568093180656433, "learning_rate": 7.37987530368569e-06, "loss": 0.0123, "step": 66690 }, { "grad_norm": 0.16713844239711761, "learning_rate": 7.369071738226474e-06, "loss": 0.0104, "step": 66700 }, { "grad_norm": 0.13478858768939972, "learning_rate": 7.358275457226954e-06, "loss": 0.0061, "step": 66710 }, { "grad_norm": 0.1554066240787506, "learning_rate": 7.347486462531899e-06, "loss": 0.007, "step": 66720 }, { "grad_norm": 0.19611991941928864, "learning_rate": 7.336704755984858e-06, "loss": 0.011, "step": 66730 }, { "grad_norm": 0.3600022792816162, "learning_rate": 7.325930339428133e-06, "loss": 0.0084, "step": 66740 }, { "grad_norm": 0.24231095612049103, "learning_rate": 7.315163214702769e-06, "loss": 0.0129, "step": 66750 }, { "grad_norm": 0.4117286503314972, "learning_rate": 7.304403383648595e-06, "loss": 0.0147, "step": 66760 }, { "grad_norm": 0.2601562440395355, "learning_rate": 7.293650848104139e-06, "loss": 0.0089, "step": 66770 }, { "grad_norm": 0.23254799842834473, "learning_rate": 7.2829056099067374e-06, "loss": 0.0105, "step": 66780 }, { "grad_norm": 0.27772584557533264, "learning_rate": 7.2721676708924494e-06, "loss": 0.0111, "step": 66790 }, { "grad_norm": 0.5849252939224243, "learning_rate": 7.261437032896096e-06, "loss": 0.0222, "step": 66800 }, { "grad_norm": 0.22287499904632568, "learning_rate": 7.250713697751255e-06, "loss": 0.0085, "step": 66810 }, { "grad_norm": 0.1983107626438141, "learning_rate": 7.239997667290255e-06, "loss": 0.0101, "step": 66820 }, { "grad_norm": 0.12488582730293274, "learning_rate": 7.2292889433441425e-06, "loss": 0.0068, "step": 66830 }, { "grad_norm": 0.20563410222530365, "learning_rate": 7.218587527742793e-06, "loss": 0.0071, "step": 66840 }, { "grad_norm": 0.18550486862659454, "learning_rate": 7.207893422314749e-06, "loss": 0.0065, "step": 66850 }, { "grad_norm": 0.11847184598445892, "learning_rate": 7.1972066288873545e-06, "loss": 0.0124, "step": 66860 }, { "grad_norm": 0.17796985805034637, "learning_rate": 7.186527149286687e-06, "loss": 0.0239, "step": 66870 }, { "grad_norm": 0.17186382412910461, "learning_rate": 7.175854985337576e-06, "loss": 0.0098, "step": 66880 }, { "grad_norm": 0.09498010575771332, "learning_rate": 7.16519013886362e-06, "loss": 0.0115, "step": 66890 }, { "grad_norm": 0.231160506606102, "learning_rate": 7.154532611687109e-06, "loss": 0.0069, "step": 66900 }, { "grad_norm": 0.19995959103107452, "learning_rate": 7.143882405629177e-06, "loss": 0.0077, "step": 66910 }, { "grad_norm": 0.19922742247581482, "learning_rate": 7.133239522509605e-06, "loss": 0.0099, "step": 66920 }, { "grad_norm": 0.18789739906787872, "learning_rate": 7.1226039641469956e-06, "loss": 0.0088, "step": 66930 }, { "grad_norm": 0.7588435411453247, "learning_rate": 7.111975732358678e-06, "loss": 0.013, "step": 66940 }, { "grad_norm": 0.12669575214385986, "learning_rate": 7.101354828960693e-06, "loss": 0.0113, "step": 66950 }, { "grad_norm": 0.09395334869623184, "learning_rate": 7.090741255767918e-06, "loss": 0.0077, "step": 66960 }, { "grad_norm": 0.19742010533809662, "learning_rate": 7.080135014593875e-06, "loss": 0.0097, "step": 66970 }, { "grad_norm": 0.16846829652786255, "learning_rate": 7.069536107250896e-06, "loss": 0.0066, "step": 66980 }, { "grad_norm": 0.27392664551734924, "learning_rate": 7.058944535550049e-06, "loss": 0.0083, "step": 66990 }, { "grad_norm": 0.15570509433746338, "learning_rate": 7.048360301301138e-06, "loss": 0.0107, "step": 67000 }, { "grad_norm": 0.18145017325878143, "learning_rate": 7.03778340631272e-06, "loss": 0.0103, "step": 67010 }, { "grad_norm": 0.27032551169395447, "learning_rate": 7.0272138523921e-06, "loss": 0.0094, "step": 67020 }, { "grad_norm": 0.2559141218662262, "learning_rate": 7.016651641345334e-06, "loss": 0.0124, "step": 67030 }, { "grad_norm": 0.3618966341018677, "learning_rate": 7.0060967749772e-06, "loss": 0.0088, "step": 67040 }, { "grad_norm": 0.4283548593521118, "learning_rate": 6.995549255091238e-06, "loss": 0.0176, "step": 67050 }, { "grad_norm": 0.20424865186214447, "learning_rate": 6.98500908348973e-06, "loss": 0.0127, "step": 67060 }, { "grad_norm": 0.4614871144294739, "learning_rate": 6.974476261973711e-06, "loss": 0.011, "step": 67070 }, { "grad_norm": 0.19647298753261566, "learning_rate": 6.963950792342949e-06, "loss": 0.0063, "step": 67080 }, { "grad_norm": 0.12986062467098236, "learning_rate": 6.9534326763959715e-06, "loss": 0.0059, "step": 67090 }, { "grad_norm": 0.15114785730838776, "learning_rate": 6.942921915929995e-06, "loss": 0.0072, "step": 67100 }, { "grad_norm": 0.12100526690483093, "learning_rate": 6.9324185127410735e-06, "loss": 0.012, "step": 67110 }, { "grad_norm": 0.18014584481716156, "learning_rate": 6.921922468623921e-06, "loss": 0.0146, "step": 67120 }, { "grad_norm": 0.14676909148693085, "learning_rate": 6.911433785372023e-06, "loss": 0.0077, "step": 67130 }, { "grad_norm": 0.22149085998535156, "learning_rate": 6.900952464777632e-06, "loss": 0.0092, "step": 67140 }, { "grad_norm": 0.1927221119403839, "learning_rate": 6.8904785086316815e-06, "loss": 0.0106, "step": 67150 }, { "grad_norm": 0.27709856629371643, "learning_rate": 6.880011918723927e-06, "loss": 0.0101, "step": 67160 }, { "grad_norm": 0.18480689823627472, "learning_rate": 6.86955269684279e-06, "loss": 0.0078, "step": 67170 }, { "grad_norm": 0.26133856177330017, "learning_rate": 6.859100844775473e-06, "loss": 0.0117, "step": 67180 }, { "grad_norm": 0.1890401989221573, "learning_rate": 6.84865636430792e-06, "loss": 0.0099, "step": 67190 }, { "grad_norm": 0.0795571506023407, "learning_rate": 6.838219257224804e-06, "loss": 0.0069, "step": 67200 }, { "grad_norm": 0.11518628895282745, "learning_rate": 6.827789525309536e-06, "loss": 0.012, "step": 67210 }, { "grad_norm": 0.25446414947509766, "learning_rate": 6.81736717034428e-06, "loss": 0.0096, "step": 67220 }, { "grad_norm": 0.12081179767847061, "learning_rate": 6.806952194109933e-06, "loss": 0.0153, "step": 67230 }, { "grad_norm": 0.2235032171010971, "learning_rate": 6.796544598386112e-06, "loss": 0.009, "step": 67240 }, { "grad_norm": 0.2729392945766449, "learning_rate": 6.786144384951204e-06, "loss": 0.0104, "step": 67250 }, { "grad_norm": 0.2891649901866913, "learning_rate": 6.775751555582322e-06, "loss": 0.0089, "step": 67260 }, { "grad_norm": 0.13124440610408783, "learning_rate": 6.76536611205531e-06, "loss": 0.0103, "step": 67270 }, { "grad_norm": 0.07727354764938354, "learning_rate": 6.754988056144762e-06, "loss": 0.0119, "step": 67280 }, { "grad_norm": 0.6816655397415161, "learning_rate": 6.744617389624014e-06, "loss": 0.0121, "step": 67290 }, { "grad_norm": 0.33059072494506836, "learning_rate": 6.734254114265087e-06, "loss": 0.0098, "step": 67300 }, { "grad_norm": 0.2920779585838318, "learning_rate": 6.723898231838843e-06, "loss": 0.0111, "step": 67310 }, { "grad_norm": 0.14769992232322693, "learning_rate": 6.713549744114766e-06, "loss": 0.01, "step": 67320 }, { "grad_norm": 0.1719365268945694, "learning_rate": 6.703208652861159e-06, "loss": 0.0109, "step": 67330 }, { "grad_norm": 0.7520020604133606, "learning_rate": 6.69287495984503e-06, "loss": 0.0127, "step": 67340 }, { "grad_norm": 0.17397183179855347, "learning_rate": 6.6825486668321e-06, "loss": 0.0109, "step": 67350 }, { "grad_norm": 0.12870056927204132, "learning_rate": 6.672229775586886e-06, "loss": 0.0128, "step": 67360 }, { "grad_norm": 0.1504283994436264, "learning_rate": 6.661918287872576e-06, "loss": 0.0087, "step": 67370 }, { "grad_norm": 0.3204147517681122, "learning_rate": 6.6516142054511346e-06, "loss": 0.015, "step": 67380 }, { "grad_norm": 0.25412923097610474, "learning_rate": 6.641317530083241e-06, "loss": 0.0084, "step": 67390 }, { "grad_norm": 0.2932657301425934, "learning_rate": 6.631028263528322e-06, "loss": 0.0119, "step": 67400 }, { "grad_norm": 0.17652451992034912, "learning_rate": 6.620746407544537e-06, "loss": 0.0091, "step": 67410 }, { "grad_norm": 0.1495525985956192, "learning_rate": 6.610471963888742e-06, "loss": 0.0112, "step": 67420 }, { "grad_norm": 0.11714901030063629, "learning_rate": 6.600204934316606e-06, "loss": 0.01, "step": 67430 }, { "grad_norm": 0.10408054292201996, "learning_rate": 6.589945320582452e-06, "loss": 0.0106, "step": 67440 }, { "grad_norm": 0.16857308149337769, "learning_rate": 6.579693124439374e-06, "loss": 0.0093, "step": 67450 }, { "grad_norm": 0.2338963896036148, "learning_rate": 6.56944834763919e-06, "loss": 0.0114, "step": 67460 }, { "grad_norm": 0.232957124710083, "learning_rate": 6.5592109919324575e-06, "loss": 0.0074, "step": 67470 }, { "grad_norm": 0.16458258032798767, "learning_rate": 6.54898105906846e-06, "loss": 0.0099, "step": 67480 }, { "grad_norm": 0.3713107407093048, "learning_rate": 6.5387585507952155e-06, "loss": 0.0144, "step": 67490 }, { "grad_norm": 0.14624391496181488, "learning_rate": 6.528543468859461e-06, "loss": 0.0075, "step": 67500 }, { "grad_norm": 0.2165520191192627, "learning_rate": 6.518335815006682e-06, "loss": 0.0093, "step": 67510 }, { "grad_norm": 0.23702681064605713, "learning_rate": 6.5081355909810845e-06, "loss": 0.0072, "step": 67520 }, { "grad_norm": 0.18899697065353394, "learning_rate": 6.497942798525608e-06, "loss": 0.0113, "step": 67530 }, { "grad_norm": 0.58770352602005, "learning_rate": 6.487757439381936e-06, "loss": 0.0074, "step": 67540 }, { "grad_norm": 0.13462576270103455, "learning_rate": 6.477579515290433e-06, "loss": 0.0103, "step": 67550 }, { "grad_norm": 0.10534612089395523, "learning_rate": 6.467409027990273e-06, "loss": 0.0086, "step": 67560 }, { "grad_norm": 0.227759450674057, "learning_rate": 6.457245979219279e-06, "loss": 0.0117, "step": 67570 }, { "grad_norm": 0.3008137345314026, "learning_rate": 6.447090370714054e-06, "loss": 0.0125, "step": 67580 }, { "grad_norm": 0.19184610247612, "learning_rate": 6.436942204209917e-06, "loss": 0.0134, "step": 67590 }, { "grad_norm": 0.16829437017440796, "learning_rate": 6.4268014814408804e-06, "loss": 0.017, "step": 67600 }, { "grad_norm": 0.22318360209465027, "learning_rate": 6.41666820413977e-06, "loss": 0.01, "step": 67610 }, { "grad_norm": 0.28720152378082275, "learning_rate": 6.406542374038033e-06, "loss": 0.0127, "step": 67620 }, { "grad_norm": 0.32034140825271606, "learning_rate": 6.396423992865935e-06, "loss": 0.0091, "step": 67630 }, { "grad_norm": 0.11366260796785355, "learning_rate": 6.386313062352412e-06, "loss": 0.0072, "step": 67640 }, { "grad_norm": 0.2552899420261383, "learning_rate": 6.376209584225152e-06, "loss": 0.0069, "step": 67650 }, { "grad_norm": 0.13615287840366364, "learning_rate": 6.366113560210557e-06, "loss": 0.0096, "step": 67660 }, { "grad_norm": 0.4909149408340454, "learning_rate": 6.356024992033766e-06, "loss": 0.0115, "step": 67670 }, { "grad_norm": 0.19568666815757751, "learning_rate": 6.345943881418648e-06, "loss": 0.01, "step": 67680 }, { "grad_norm": 0.08428771048784256, "learning_rate": 6.33587023008777e-06, "loss": 0.008, "step": 67690 }, { "grad_norm": 0.1991758644580841, "learning_rate": 6.32580403976245e-06, "loss": 0.0088, "step": 67700 }, { "grad_norm": 0.27086400985717773, "learning_rate": 6.31574531216273e-06, "loss": 0.0086, "step": 67710 }, { "grad_norm": 0.21240702271461487, "learning_rate": 6.305694049007371e-06, "loss": 0.0114, "step": 67720 }, { "grad_norm": 0.3451787829399109, "learning_rate": 6.2956502520138575e-06, "loss": 0.0103, "step": 67730 }, { "grad_norm": 0.19061408936977386, "learning_rate": 6.285613922898409e-06, "loss": 0.0131, "step": 67740 }, { "grad_norm": 0.13523158431053162, "learning_rate": 6.275585063375927e-06, "loss": 0.0111, "step": 67750 }, { "grad_norm": 0.15411032736301422, "learning_rate": 6.265563675160113e-06, "loss": 0.0084, "step": 67760 }, { "grad_norm": 0.9985978007316589, "learning_rate": 6.25554975996332e-06, "loss": 0.0065, "step": 67770 }, { "grad_norm": 0.3026605248451233, "learning_rate": 6.245543319496661e-06, "loss": 0.0097, "step": 67780 }, { "grad_norm": 0.131099134683609, "learning_rate": 6.2355443554699685e-06, "loss": 0.0108, "step": 67790 }, { "grad_norm": 0.21603308618068695, "learning_rate": 6.225552869591766e-06, "loss": 0.01, "step": 67800 }, { "grad_norm": 0.16071538627147675, "learning_rate": 6.215568863569365e-06, "loss": 0.012, "step": 67810 }, { "grad_norm": 0.35925427079200745, "learning_rate": 6.205592339108712e-06, "loss": 0.0111, "step": 67820 }, { "grad_norm": 0.12873350083827972, "learning_rate": 6.195623297914577e-06, "loss": 0.0068, "step": 67830 }, { "grad_norm": 0.1790817528963089, "learning_rate": 6.185661741690357e-06, "loss": 0.0115, "step": 67840 }, { "grad_norm": 0.22438643872737885, "learning_rate": 6.1757076721382145e-06, "loss": 0.0117, "step": 67850 }, { "grad_norm": 0.19101965427398682, "learning_rate": 6.165761090959038e-06, "loss": 0.0121, "step": 67860 }, { "grad_norm": 0.21616917848587036, "learning_rate": 6.155821999852424e-06, "loss": 0.0085, "step": 67870 }, { "grad_norm": 0.16005674004554749, "learning_rate": 6.145890400516696e-06, "loss": 0.0087, "step": 67880 }, { "grad_norm": 0.11080975830554962, "learning_rate": 6.1359662946488816e-06, "loss": 0.0076, "step": 67890 }, { "grad_norm": 0.25927597284317017, "learning_rate": 6.12604968394474e-06, "loss": 0.0101, "step": 67900 }, { "grad_norm": 0.1268823742866516, "learning_rate": 6.11614057009875e-06, "loss": 0.0095, "step": 67910 }, { "grad_norm": 0.2248566746711731, "learning_rate": 6.106238954804111e-06, "loss": 0.0077, "step": 67920 }, { "grad_norm": 0.20687997341156006, "learning_rate": 6.096344839752738e-06, "loss": 0.0117, "step": 67930 }, { "grad_norm": 0.08118723332881927, "learning_rate": 6.086458226635278e-06, "loss": 0.0102, "step": 67940 }, { "grad_norm": 0.2673821747303009, "learning_rate": 6.076579117141046e-06, "loss": 0.0158, "step": 67950 }, { "grad_norm": 0.24500904977321625, "learning_rate": 6.066707512958153e-06, "loss": 0.0084, "step": 67960 }, { "grad_norm": 0.11454007029533386, "learning_rate": 6.056843415773361e-06, "loss": 0.008, "step": 67970 }, { "grad_norm": 0.20033426582813263, "learning_rate": 6.0469868272721776e-06, "loss": 0.0087, "step": 67980 }, { "grad_norm": 0.19018179178237915, "learning_rate": 6.037137749138844e-06, "loss": 0.0105, "step": 67990 }, { "grad_norm": 0.26785314083099365, "learning_rate": 6.027296183056252e-06, "loss": 0.014, "step": 68000 }, { "grad_norm": 0.18239948153495789, "learning_rate": 6.017462130706114e-06, "loss": 0.0074, "step": 68010 }, { "grad_norm": 0.17539580166339874, "learning_rate": 6.0076355937687516e-06, "loss": 0.0087, "step": 68020 }, { "grad_norm": 0.19173453748226166, "learning_rate": 5.9978165739232925e-06, "loss": 0.01, "step": 68030 }, { "grad_norm": 0.11967433989048004, "learning_rate": 5.988005072847508e-06, "loss": 0.0083, "step": 68040 }, { "grad_norm": 0.1870584487915039, "learning_rate": 5.978201092217928e-06, "loss": 0.0122, "step": 68050 }, { "grad_norm": 0.1338779479265213, "learning_rate": 5.9684046337097895e-06, "loss": 0.006, "step": 68060 }, { "grad_norm": 0.12978357076644897, "learning_rate": 5.958615698997016e-06, "loss": 0.0107, "step": 68070 }, { "grad_norm": 0.17269094288349152, "learning_rate": 5.948834289752303e-06, "loss": 0.0074, "step": 68080 }, { "grad_norm": 0.19021956622600555, "learning_rate": 5.939060407646996e-06, "loss": 0.0107, "step": 68090 }, { "grad_norm": 0.150994211435318, "learning_rate": 5.929294054351198e-06, "loss": 0.0103, "step": 68100 }, { "grad_norm": 0.12729276716709137, "learning_rate": 5.919535231533707e-06, "loss": 0.0053, "step": 68110 }, { "grad_norm": 0.24041128158569336, "learning_rate": 5.909783940862046e-06, "loss": 0.0063, "step": 68120 }, { "grad_norm": 0.16814754903316498, "learning_rate": 5.900040184002436e-06, "loss": 0.0102, "step": 68130 }, { "grad_norm": 0.26128408312797546, "learning_rate": 5.890303962619831e-06, "loss": 0.0062, "step": 68140 }, { "grad_norm": 0.24001960456371307, "learning_rate": 5.88057527837787e-06, "loss": 0.0094, "step": 68150 }, { "grad_norm": 0.5611312389373779, "learning_rate": 5.870854132938924e-06, "loss": 0.0086, "step": 68160 }, { "grad_norm": 0.23628973960876465, "learning_rate": 5.861140527964071e-06, "loss": 0.0094, "step": 68170 }, { "grad_norm": 0.22474385797977448, "learning_rate": 5.851434465113098e-06, "loss": 0.0057, "step": 68180 }, { "grad_norm": 0.1919277012348175, "learning_rate": 5.841735946044524e-06, "loss": 0.0082, "step": 68190 }, { "grad_norm": 0.15585210919380188, "learning_rate": 5.832044972415523e-06, "loss": 0.0099, "step": 68200 }, { "grad_norm": 0.13144439458847046, "learning_rate": 5.8223615458820655e-06, "loss": 0.011, "step": 68210 }, { "grad_norm": 0.1556662768125534, "learning_rate": 5.812685668098733e-06, "loss": 0.0126, "step": 68220 }, { "grad_norm": 0.09542100876569748, "learning_rate": 5.803017340718913e-06, "loss": 0.0101, "step": 68230 }, { "grad_norm": 0.321705162525177, "learning_rate": 5.793356565394636e-06, "loss": 0.0115, "step": 68240 }, { "grad_norm": 0.16725555062294006, "learning_rate": 5.783703343776658e-06, "loss": 0.0102, "step": 68250 }, { "grad_norm": 0.1471373587846756, "learning_rate": 5.774057677514477e-06, "loss": 0.0071, "step": 68260 }, { "grad_norm": 0.17799118161201477, "learning_rate": 5.764419568256235e-06, "loss": 0.0097, "step": 68270 }, { "grad_norm": 0.5577620267868042, "learning_rate": 5.754789017648865e-06, "loss": 0.0132, "step": 68280 }, { "grad_norm": 0.19019107520580292, "learning_rate": 5.745166027337934e-06, "loss": 0.0087, "step": 68290 }, { "grad_norm": 0.132050022482872, "learning_rate": 5.735550598967754e-06, "loss": 0.0082, "step": 68300 }, { "grad_norm": 0.15095306932926178, "learning_rate": 5.725942734181339e-06, "loss": 0.011, "step": 68310 }, { "grad_norm": 0.18138213455677032, "learning_rate": 5.7163424346204146e-06, "loss": 0.0081, "step": 68320 }, { "grad_norm": 0.16426542401313782, "learning_rate": 5.706749701925407e-06, "loss": 0.0073, "step": 68330 }, { "grad_norm": 0.21473455429077148, "learning_rate": 5.697164537735461e-06, "loss": 0.0086, "step": 68340 }, { "grad_norm": 0.4185413420200348, "learning_rate": 5.687586943688406e-06, "loss": 0.0095, "step": 68350 }, { "grad_norm": 0.4778117537498474, "learning_rate": 5.678016921420792e-06, "loss": 0.0128, "step": 68360 }, { "grad_norm": 0.33873268961906433, "learning_rate": 5.668454472567875e-06, "loss": 0.0124, "step": 68370 }, { "grad_norm": 0.391928106546402, "learning_rate": 5.658899598763617e-06, "loss": 0.012, "step": 68380 }, { "grad_norm": 0.2639760971069336, "learning_rate": 5.6493523016407035e-06, "loss": 0.008, "step": 68390 }, { "grad_norm": 0.6290656924247742, "learning_rate": 5.6398125828304615e-06, "loss": 0.0101, "step": 68400 }, { "grad_norm": 0.17368988692760468, "learning_rate": 5.630280443963015e-06, "loss": 0.0094, "step": 68410 }, { "grad_norm": 0.1589137315750122, "learning_rate": 5.620755886667112e-06, "loss": 0.01, "step": 68420 }, { "grad_norm": 0.18996863067150116, "learning_rate": 5.611238912570271e-06, "loss": 0.0083, "step": 68430 }, { "grad_norm": 0.3978600800037384, "learning_rate": 5.601729523298649e-06, "loss": 0.0148, "step": 68440 }, { "grad_norm": 0.13900130987167358, "learning_rate": 5.592227720477161e-06, "loss": 0.0096, "step": 68450 }, { "grad_norm": 0.12016206234693527, "learning_rate": 5.582733505729415e-06, "loss": 0.0144, "step": 68460 }, { "grad_norm": 0.27483659982681274, "learning_rate": 5.573246880677668e-06, "loss": 0.0082, "step": 68470 }, { "grad_norm": 0.20260120928287506, "learning_rate": 5.56376784694298e-06, "loss": 0.0077, "step": 68480 }, { "grad_norm": 0.3205951750278473, "learning_rate": 5.554296406145027e-06, "loss": 0.0089, "step": 68490 }, { "grad_norm": 0.13844747841358185, "learning_rate": 5.544832559902219e-06, "loss": 0.0092, "step": 68500 }, { "grad_norm": 0.12327372282743454, "learning_rate": 5.53537630983168e-06, "loss": 0.0077, "step": 68510 }, { "grad_norm": 0.12000565975904465, "learning_rate": 5.5259276575492125e-06, "loss": 0.0118, "step": 68520 }, { "grad_norm": 0.23778453469276428, "learning_rate": 5.516486604669357e-06, "loss": 0.0133, "step": 68530 }, { "grad_norm": 0.3491027057170868, "learning_rate": 5.507053152805303e-06, "loss": 0.0149, "step": 68540 }, { "grad_norm": 0.1805919110774994, "learning_rate": 5.497627303568975e-06, "loss": 0.0168, "step": 68550 }, { "grad_norm": 0.21439334750175476, "learning_rate": 5.488209058571003e-06, "loss": 0.0078, "step": 68560 }, { "grad_norm": 0.27759262919425964, "learning_rate": 5.4787984194207054e-06, "loss": 0.0102, "step": 68570 }, { "grad_norm": 0.19939783215522766, "learning_rate": 5.469395387726095e-06, "loss": 0.0067, "step": 68580 }, { "grad_norm": 0.20022353529930115, "learning_rate": 5.45999996509391e-06, "loss": 0.0102, "step": 68590 }, { "grad_norm": 0.16267013549804688, "learning_rate": 5.450612153129536e-06, "loss": 0.0104, "step": 68600 }, { "grad_norm": 0.2746240794658661, "learning_rate": 5.4412319534371426e-06, "loss": 0.011, "step": 68610 }, { "grad_norm": 0.1807379424571991, "learning_rate": 5.431859367619513e-06, "loss": 0.0079, "step": 68620 }, { "grad_norm": 0.24607178568840027, "learning_rate": 5.422494397278172e-06, "loss": 0.009, "step": 68630 }, { "grad_norm": 0.3809070587158203, "learning_rate": 5.413137044013344e-06, "loss": 0.0098, "step": 68640 }, { "grad_norm": 0.18168242275714874, "learning_rate": 5.403787309423941e-06, "loss": 0.0093, "step": 68650 }, { "grad_norm": 0.8529438972473145, "learning_rate": 5.3944451951075835e-06, "loss": 0.0084, "step": 68660 }, { "grad_norm": 0.2527451515197754, "learning_rate": 5.385110702660562e-06, "loss": 0.0082, "step": 68670 }, { "grad_norm": 0.10464777797460556, "learning_rate": 5.375783833677922e-06, "loss": 0.0083, "step": 68680 }, { "grad_norm": 0.20882028341293335, "learning_rate": 5.36646458975334e-06, "loss": 0.0109, "step": 68690 }, { "grad_norm": 0.14666518568992615, "learning_rate": 5.3571529724792294e-06, "loss": 0.0105, "step": 68700 }, { "grad_norm": 0.28735190629959106, "learning_rate": 5.347848983446702e-06, "loss": 0.0096, "step": 68710 }, { "grad_norm": 0.232815220952034, "learning_rate": 5.3385526242455185e-06, "loss": 0.0112, "step": 68720 }, { "grad_norm": 0.18194323778152466, "learning_rate": 5.329263896464226e-06, "loss": 0.009, "step": 68730 }, { "grad_norm": 0.10359876602888107, "learning_rate": 5.3199828016899715e-06, "loss": 0.0091, "step": 68740 }, { "grad_norm": 0.16781598329544067, "learning_rate": 5.310709341508657e-06, "loss": 0.0108, "step": 68750 }, { "grad_norm": 0.18476252257823944, "learning_rate": 5.301443517504861e-06, "loss": 0.0099, "step": 68760 }, { "grad_norm": 0.20983676612377167, "learning_rate": 5.292185331261862e-06, "loss": 0.006, "step": 68770 }, { "grad_norm": 0.09280593693256378, "learning_rate": 5.28293478436162e-06, "loss": 0.0067, "step": 68780 }, { "grad_norm": 0.1233120933175087, "learning_rate": 5.273691878384829e-06, "loss": 0.0114, "step": 68790 }, { "grad_norm": 0.2983722686767578, "learning_rate": 5.264456614910812e-06, "loss": 0.009, "step": 68800 }, { "grad_norm": 0.2520180940628052, "learning_rate": 5.255228995517647e-06, "loss": 0.0114, "step": 68810 }, { "grad_norm": 0.174288809299469, "learning_rate": 5.246009021782067e-06, "loss": 0.011, "step": 68820 }, { "grad_norm": 0.16829410195350647, "learning_rate": 5.2367966952795225e-06, "loss": 0.016, "step": 68830 }, { "grad_norm": 0.15886977314949036, "learning_rate": 5.2275920175841485e-06, "loss": 0.0083, "step": 68840 }, { "grad_norm": 0.08719799667596817, "learning_rate": 5.218394990268771e-06, "loss": 0.0108, "step": 68850 }, { "grad_norm": 0.3742814064025879, "learning_rate": 5.209205614904916e-06, "loss": 0.0098, "step": 68860 }, { "grad_norm": 0.24250264465808868, "learning_rate": 5.200023893062777e-06, "loss": 0.0112, "step": 68870 }, { "grad_norm": 0.2223927527666092, "learning_rate": 5.190849826311289e-06, "loss": 0.0083, "step": 68880 }, { "grad_norm": 0.36052805185317993, "learning_rate": 5.181683416218025e-06, "loss": 0.0154, "step": 68890 }, { "grad_norm": 0.24568909406661987, "learning_rate": 5.172524664349276e-06, "loss": 0.0063, "step": 68900 }, { "grad_norm": 0.08230961859226227, "learning_rate": 5.1633735722700416e-06, "loss": 0.0073, "step": 68910 }, { "grad_norm": 0.12540222704410553, "learning_rate": 5.154230141543958e-06, "loss": 0.0079, "step": 68920 }, { "grad_norm": 0.3192415237426758, "learning_rate": 5.145094373733433e-06, "loss": 0.0114, "step": 68930 }, { "grad_norm": 0.1808042973279953, "learning_rate": 5.135966270399478e-06, "loss": 0.009, "step": 68940 }, { "grad_norm": 0.43247687816619873, "learning_rate": 5.126845833101857e-06, "loss": 0.0139, "step": 68950 }, { "grad_norm": 0.11415145546197891, "learning_rate": 5.1177330633989994e-06, "loss": 0.0089, "step": 68960 }, { "grad_norm": 0.41329142451286316, "learning_rate": 5.108627962848033e-06, "loss": 0.0203, "step": 68970 }, { "grad_norm": 0.16019310057163239, "learning_rate": 5.099530533004759e-06, "loss": 0.0093, "step": 68980 }, { "grad_norm": 0.17961904406547546, "learning_rate": 5.090440775423699e-06, "loss": 0.0099, "step": 68990 }, { "grad_norm": 0.1565854400396347, "learning_rate": 5.081358691658022e-06, "loss": 0.0092, "step": 69000 }, { "grad_norm": 0.22901475429534912, "learning_rate": 5.072284283259621e-06, "loss": 0.0113, "step": 69010 }, { "grad_norm": 0.20788723230361938, "learning_rate": 5.063217551779053e-06, "loss": 0.0107, "step": 69020 }, { "grad_norm": 0.47011598944664, "learning_rate": 5.054158498765588e-06, "loss": 0.0088, "step": 69030 }, { "grad_norm": 0.4066326916217804, "learning_rate": 5.0451071257671625e-06, "loss": 0.0104, "step": 69040 }, { "grad_norm": 0.2497805505990982, "learning_rate": 5.036063434330407e-06, "loss": 0.0106, "step": 69050 }, { "grad_norm": 0.11813046783208847, "learning_rate": 5.027027426000652e-06, "loss": 0.0061, "step": 69060 }, { "grad_norm": 0.10560401529073715, "learning_rate": 5.017999102321886e-06, "loss": 0.0066, "step": 69070 }, { "grad_norm": 0.35226863622665405, "learning_rate": 5.0089784648368224e-06, "loss": 0.0114, "step": 69080 }, { "grad_norm": 0.177008256316185, "learning_rate": 4.999965515086829e-06, "loss": 0.0089, "step": 69090 }, { "grad_norm": 0.16927188634872437, "learning_rate": 4.990960254611976e-06, "loss": 0.0088, "step": 69100 }, { "grad_norm": 0.32940369844436646, "learning_rate": 4.9819626849510194e-06, "loss": 0.0129, "step": 69110 }, { "grad_norm": 0.11506079882383347, "learning_rate": 4.97297280764138e-06, "loss": 0.0148, "step": 69120 }, { "grad_norm": 0.2721644937992096, "learning_rate": 4.963990624219211e-06, "loss": 0.01, "step": 69130 }, { "grad_norm": 0.10908344388008118, "learning_rate": 4.955016136219298e-06, "loss": 0.0067, "step": 69140 }, { "grad_norm": 0.10970178991556168, "learning_rate": 4.946049345175146e-06, "loss": 0.0068, "step": 69150 }, { "grad_norm": 0.29202800989151, "learning_rate": 4.937090252618937e-06, "loss": 0.0097, "step": 69160 }, { "grad_norm": 0.1597602218389511, "learning_rate": 4.928138860081521e-06, "loss": 0.0104, "step": 69170 }, { "grad_norm": 0.15248346328735352, "learning_rate": 4.919195169092472e-06, "loss": 0.0097, "step": 69180 }, { "grad_norm": 0.13105690479278564, "learning_rate": 4.910259181179994e-06, "loss": 0.009, "step": 69190 }, { "grad_norm": 0.16129682958126068, "learning_rate": 4.901330897871015e-06, "loss": 0.0082, "step": 69200 }, { "grad_norm": 0.12917490303516388, "learning_rate": 4.8924103206911375e-06, "loss": 0.0056, "step": 69210 }, { "grad_norm": 0.1741025596857071, "learning_rate": 4.883497451164637e-06, "loss": 0.0129, "step": 69220 }, { "grad_norm": 0.2016916573047638, "learning_rate": 4.874592290814484e-06, "loss": 0.0081, "step": 69230 }, { "grad_norm": 0.43109819293022156, "learning_rate": 4.865694841162327e-06, "loss": 0.0111, "step": 69240 }, { "grad_norm": 0.16520899534225464, "learning_rate": 4.856805103728496e-06, "loss": 0.0082, "step": 69250 }, { "grad_norm": 0.21166692674160004, "learning_rate": 4.847923080032007e-06, "loss": 0.0068, "step": 69260 }, { "grad_norm": 0.15247222781181335, "learning_rate": 4.8390487715905486e-06, "loss": 0.0096, "step": 69270 }, { "grad_norm": 0.16744796931743622, "learning_rate": 4.8301821799205e-06, "loss": 0.0095, "step": 69280 }, { "grad_norm": 0.1300305724143982, "learning_rate": 4.821323306536918e-06, "loss": 0.0051, "step": 69290 }, { "grad_norm": 0.10257407277822495, "learning_rate": 4.8124721529535455e-06, "loss": 0.0093, "step": 69300 }, { "grad_norm": 0.1499180644750595, "learning_rate": 4.803628720682807e-06, "loss": 0.0137, "step": 69310 }, { "grad_norm": 0.16317526996135712, "learning_rate": 4.794793011235776e-06, "loss": 0.0094, "step": 69320 }, { "grad_norm": 0.32023265957832336, "learning_rate": 4.78596502612228e-06, "loss": 0.0121, "step": 69330 }, { "grad_norm": 0.19669389724731445, "learning_rate": 4.777144766850738e-06, "loss": 0.0111, "step": 69340 }, { "grad_norm": 0.2793281078338623, "learning_rate": 4.768332234928313e-06, "loss": 0.008, "step": 69350 }, { "grad_norm": 0.06895571202039719, "learning_rate": 4.759527431860828e-06, "loss": 0.0085, "step": 69360 }, { "grad_norm": 0.24648581445217133, "learning_rate": 4.750730359152755e-06, "loss": 0.0106, "step": 69370 }, { "grad_norm": 0.09733567386865616, "learning_rate": 4.741941018307311e-06, "loss": 0.0104, "step": 69380 }, { "grad_norm": 0.08783333003520966, "learning_rate": 4.733159410826321e-06, "loss": 0.0095, "step": 69390 }, { "grad_norm": 0.12891097366809845, "learning_rate": 4.724385538210357e-06, "loss": 0.0098, "step": 69400 }, { "grad_norm": 0.15182551741600037, "learning_rate": 4.715619401958599e-06, "loss": 0.0095, "step": 69410 }, { "grad_norm": 0.2061886042356491, "learning_rate": 4.706861003568958e-06, "loss": 0.0108, "step": 69420 }, { "grad_norm": 0.1426648199558258, "learning_rate": 4.698110344538003e-06, "loss": 0.0103, "step": 69430 }, { "grad_norm": 0.3101186156272888, "learning_rate": 4.689367426360975e-06, "loss": 0.0084, "step": 69440 }, { "grad_norm": 0.319726824760437, "learning_rate": 4.680632250531819e-06, "loss": 0.0094, "step": 69450 }, { "grad_norm": 0.2595795691013336, "learning_rate": 4.671904818543115e-06, "loss": 0.0106, "step": 69460 }, { "grad_norm": 0.17975500226020813, "learning_rate": 4.66318513188615e-06, "loss": 0.0086, "step": 69470 }, { "grad_norm": 0.17003995180130005, "learning_rate": 4.654473192050884e-06, "loss": 0.0094, "step": 69480 }, { "grad_norm": 0.1278173327445984, "learning_rate": 4.64576900052594e-06, "loss": 0.0082, "step": 69490 }, { "grad_norm": 0.1616838276386261, "learning_rate": 4.637072558798638e-06, "loss": 0.008, "step": 69500 }, { "grad_norm": 0.1589943766593933, "learning_rate": 4.628383868354969e-06, "loss": 0.0103, "step": 69510 }, { "grad_norm": 0.15665557980537415, "learning_rate": 4.6197029306795595e-06, "loss": 0.0082, "step": 69520 }, { "grad_norm": 0.11300437152385712, "learning_rate": 4.611029747255779e-06, "loss": 0.0055, "step": 69530 }, { "grad_norm": 0.4059314727783203, "learning_rate": 4.6023643195656164e-06, "loss": 0.016, "step": 69540 }, { "grad_norm": 0.06245826557278633, "learning_rate": 4.593706649089768e-06, "loss": 0.0075, "step": 69550 }, { "grad_norm": 0.31254565715789795, "learning_rate": 4.585056737307597e-06, "loss": 0.0094, "step": 69560 }, { "grad_norm": 0.09826566278934479, "learning_rate": 4.576414585697103e-06, "loss": 0.0056, "step": 69570 }, { "grad_norm": 0.2524510324001312, "learning_rate": 4.567780195735044e-06, "loss": 0.0087, "step": 69580 }, { "grad_norm": 0.18738022446632385, "learning_rate": 4.559153568896757e-06, "loss": 0.007, "step": 69590 }, { "grad_norm": 0.2954549193382263, "learning_rate": 4.550534706656329e-06, "loss": 0.0085, "step": 69600 }, { "grad_norm": 0.41053465008735657, "learning_rate": 4.541923610486465e-06, "loss": 0.0106, "step": 69610 }, { "grad_norm": 0.16303496062755585, "learning_rate": 4.533320281858578e-06, "loss": 0.0085, "step": 69620 }, { "grad_norm": 0.22293825447559357, "learning_rate": 4.52472472224274e-06, "loss": 0.0102, "step": 69630 }, { "grad_norm": 0.21241997182369232, "learning_rate": 4.51613693310769e-06, "loss": 0.0152, "step": 69640 }, { "grad_norm": 0.19241222739219666, "learning_rate": 4.507556915920868e-06, "loss": 0.0092, "step": 69650 }, { "grad_norm": 0.19660770893096924, "learning_rate": 4.498984672148332e-06, "loss": 0.0115, "step": 69660 }, { "grad_norm": 0.22917236387729645, "learning_rate": 4.490420203254864e-06, "loss": 0.0087, "step": 69670 }, { "grad_norm": 0.24909356236457825, "learning_rate": 4.481863510703893e-06, "loss": 0.0112, "step": 69680 }, { "grad_norm": 0.1530800461769104, "learning_rate": 4.473314595957523e-06, "loss": 0.0109, "step": 69690 }, { "grad_norm": 0.2814444601535797, "learning_rate": 4.464773460476535e-06, "loss": 0.0113, "step": 69700 }, { "grad_norm": 0.290105402469635, "learning_rate": 4.456240105720372e-06, "loss": 0.0073, "step": 69710 }, { "grad_norm": 0.13296402990818024, "learning_rate": 4.4477145331471405e-06, "loss": 0.022, "step": 69720 }, { "grad_norm": 0.05312354117631912, "learning_rate": 4.439196744213653e-06, "loss": 0.0074, "step": 69730 }, { "grad_norm": 0.1671125739812851, "learning_rate": 4.430686740375339e-06, "loss": 0.0059, "step": 69740 }, { "grad_norm": 0.19631415605545044, "learning_rate": 4.422184523086342e-06, "loss": 0.0073, "step": 69750 }, { "grad_norm": 0.21251092851161957, "learning_rate": 4.41369009379946e-06, "loss": 0.0085, "step": 69760 }, { "grad_norm": 0.257275253534317, "learning_rate": 4.405203453966139e-06, "loss": 0.0065, "step": 69770 }, { "grad_norm": 0.19903205335140228, "learning_rate": 4.396724605036539e-06, "loss": 0.0082, "step": 69780 }, { "grad_norm": 0.19264037907123566, "learning_rate": 4.388253548459437e-06, "loss": 0.0087, "step": 69790 }, { "grad_norm": 0.1316075176000595, "learning_rate": 4.3797902856823395e-06, "loss": 0.0102, "step": 69800 }, { "grad_norm": 0.1884012669324875, "learning_rate": 4.371334818151357e-06, "loss": 0.0078, "step": 69810 }, { "grad_norm": 0.2251373827457428, "learning_rate": 4.362887147311306e-06, "loss": 0.0087, "step": 69820 }, { "grad_norm": 0.22883175313472748, "learning_rate": 4.354447274605672e-06, "loss": 0.0091, "step": 69830 }, { "grad_norm": 0.23438550531864166, "learning_rate": 4.346015201476572e-06, "loss": 0.0142, "step": 69840 }, { "grad_norm": 0.12139485776424408, "learning_rate": 4.337590929364855e-06, "loss": 0.009, "step": 69850 }, { "grad_norm": 0.34878915548324585, "learning_rate": 4.329174459709973e-06, "loss": 0.0078, "step": 69860 }, { "grad_norm": 0.1866760402917862, "learning_rate": 4.320765793950071e-06, "loss": 0.0086, "step": 69870 }, { "grad_norm": 0.390865683555603, "learning_rate": 4.312364933521962e-06, "loss": 0.0108, "step": 69880 }, { "grad_norm": 0.08583314716815948, "learning_rate": 4.303971879861129e-06, "loss": 0.0082, "step": 69890 }, { "grad_norm": 0.19118106365203857, "learning_rate": 4.295586634401716e-06, "loss": 0.0088, "step": 69900 }, { "grad_norm": 0.16190148890018463, "learning_rate": 4.287209198576536e-06, "loss": 0.007, "step": 69910 }, { "grad_norm": 0.17348213493824005, "learning_rate": 4.278839573817045e-06, "loss": 0.0099, "step": 69920 }, { "grad_norm": 0.2579919099807739, "learning_rate": 4.270477761553399e-06, "loss": 0.0104, "step": 69930 }, { "grad_norm": 0.09518659859895706, "learning_rate": 4.262123763214393e-06, "loss": 0.0059, "step": 69940 }, { "grad_norm": 0.19989904761314392, "learning_rate": 4.253777580227508e-06, "loss": 0.0064, "step": 69950 }, { "grad_norm": 0.12574565410614014, "learning_rate": 4.2454392140188755e-06, "loss": 0.009, "step": 69960 }, { "grad_norm": 0.14561216533184052, "learning_rate": 4.2371086660132785e-06, "loss": 0.0116, "step": 69970 }, { "grad_norm": 0.343831866979599, "learning_rate": 4.228785937634205e-06, "loss": 0.007, "step": 69980 }, { "grad_norm": 0.20875032246112823, "learning_rate": 4.220471030303758e-06, "loss": 0.0104, "step": 69990 }, { "grad_norm": 0.1843239963054657, "learning_rate": 4.212163945442754e-06, "loss": 0.0083, "step": 70000 }, { "grad_norm": 0.28431257605552673, "learning_rate": 4.203864684470621e-06, "loss": 0.0079, "step": 70010 }, { "grad_norm": 0.2680034041404724, "learning_rate": 4.19557324880549e-06, "loss": 0.0082, "step": 70020 }, { "grad_norm": 0.22171823680400848, "learning_rate": 4.187289639864145e-06, "loss": 0.0094, "step": 70030 }, { "grad_norm": 0.11057832092046738, "learning_rate": 4.1790138590619974e-06, "loss": 0.008, "step": 70040 }, { "grad_norm": 0.6188632249832153, "learning_rate": 4.170745907813195e-06, "loss": 0.011, "step": 70050 }, { "grad_norm": 0.14958465099334717, "learning_rate": 4.162485787530479e-06, "loss": 0.0105, "step": 70060 }, { "grad_norm": 0.23372629284858704, "learning_rate": 4.154233499625282e-06, "loss": 0.0062, "step": 70070 }, { "grad_norm": 0.18073059618473053, "learning_rate": 4.145989045507692e-06, "loss": 0.0093, "step": 70080 }, { "grad_norm": 0.10395349562168121, "learning_rate": 4.1377524265864666e-06, "loss": 0.0109, "step": 70090 }, { "grad_norm": 0.15272605419158936, "learning_rate": 4.1295236442690175e-06, "loss": 0.0072, "step": 70100 }, { "grad_norm": 0.11659500747919083, "learning_rate": 4.121302699961421e-06, "loss": 0.0078, "step": 70110 }, { "grad_norm": 0.30046355724334717, "learning_rate": 4.113089595068403e-06, "loss": 0.0136, "step": 70120 }, { "grad_norm": 0.421667218208313, "learning_rate": 4.104884330993364e-06, "loss": 0.0103, "step": 70130 }, { "grad_norm": 0.11006463319063187, "learning_rate": 4.0966869091383585e-06, "loss": 0.0145, "step": 70140 }, { "grad_norm": 0.17629188299179077, "learning_rate": 4.088497330904101e-06, "loss": 0.0082, "step": 70150 }, { "grad_norm": 0.08316981792449951, "learning_rate": 4.080315597689976e-06, "loss": 0.0097, "step": 70160 }, { "grad_norm": 0.2235880345106125, "learning_rate": 4.072141710893995e-06, "loss": 0.0138, "step": 70170 }, { "grad_norm": 0.1592431217432022, "learning_rate": 4.063975671912879e-06, "loss": 0.0102, "step": 70180 }, { "grad_norm": 0.15449896454811096, "learning_rate": 4.055817482141949e-06, "loss": 0.0074, "step": 70190 }, { "grad_norm": 0.20343472063541412, "learning_rate": 4.047667142975259e-06, "loss": 0.0089, "step": 70200 }, { "grad_norm": 0.08244462311267853, "learning_rate": 4.039524655805443e-06, "loss": 0.0118, "step": 70210 }, { "grad_norm": 0.2235753983259201, "learning_rate": 4.03139002202384e-06, "loss": 0.0104, "step": 70220 }, { "grad_norm": 0.3618670105934143, "learning_rate": 4.023263243020447e-06, "loss": 0.0125, "step": 70230 }, { "grad_norm": 0.23021088540554047, "learning_rate": 4.015144320183884e-06, "loss": 0.0122, "step": 70240 }, { "grad_norm": 0.20752961933612823, "learning_rate": 4.007033254901482e-06, "loss": 0.009, "step": 70250 }, { "grad_norm": 0.20079907774925232, "learning_rate": 3.9989300485591795e-06, "loss": 0.0093, "step": 70260 }, { "grad_norm": 0.18945570290088654, "learning_rate": 3.990834702541601e-06, "loss": 0.0182, "step": 70270 }, { "grad_norm": 0.2617653012275696, "learning_rate": 3.982747218232019e-06, "loss": 0.0096, "step": 70280 }, { "grad_norm": 0.16674494743347168, "learning_rate": 3.97466759701236e-06, "loss": 0.0124, "step": 70290 }, { "grad_norm": 0.34346267580986023, "learning_rate": 3.966595840263226e-06, "loss": 0.0112, "step": 70300 }, { "grad_norm": 0.19417229294776917, "learning_rate": 3.958531949363836e-06, "loss": 0.0081, "step": 70310 }, { "grad_norm": 0.12249507755041122, "learning_rate": 3.950475925692098e-06, "loss": 0.0101, "step": 70320 }, { "grad_norm": 0.216648668050766, "learning_rate": 3.9424277706245685e-06, "loss": 0.0074, "step": 70330 }, { "grad_norm": 0.10780621320009232, "learning_rate": 3.934387485536451e-06, "loss": 0.0117, "step": 70340 }, { "grad_norm": 0.14539366960525513, "learning_rate": 3.926355071801619e-06, "loss": 0.0094, "step": 70350 }, { "grad_norm": 0.1348692625761032, "learning_rate": 3.9183305307925965e-06, "loss": 0.0067, "step": 70360 }, { "grad_norm": 0.1936129927635193, "learning_rate": 3.910313863880533e-06, "loss": 0.0098, "step": 70370 }, { "grad_norm": 0.1880379617214203, "learning_rate": 3.902305072435292e-06, "loss": 0.0074, "step": 70380 }, { "grad_norm": 0.23589129745960236, "learning_rate": 3.894304157825329e-06, "loss": 0.0089, "step": 70390 }, { "grad_norm": 0.1046484187245369, "learning_rate": 3.886311121417791e-06, "loss": 0.0076, "step": 70400 }, { "grad_norm": 0.21901196241378784, "learning_rate": 3.878325964578472e-06, "loss": 0.0097, "step": 70410 }, { "grad_norm": 0.24542823433876038, "learning_rate": 3.870348688671815e-06, "loss": 0.0079, "step": 70420 }, { "grad_norm": 0.2022087126970291, "learning_rate": 3.862379295060931e-06, "loss": 0.0068, "step": 70430 }, { "grad_norm": 0.15156923234462738, "learning_rate": 3.854417785107534e-06, "loss": 0.0168, "step": 70440 }, { "grad_norm": 0.18064938485622406, "learning_rate": 3.8464641601720755e-06, "loss": 0.0082, "step": 70450 }, { "grad_norm": 0.195133239030838, "learning_rate": 3.838518421613579e-06, "loss": 0.0114, "step": 70460 }, { "grad_norm": 0.3109825551509857, "learning_rate": 3.830580570789766e-06, "loss": 0.0103, "step": 70470 }, { "grad_norm": 0.1366478055715561, "learning_rate": 3.822650609057005e-06, "loss": 0.0079, "step": 70480 }, { "grad_norm": 0.12522846460342407, "learning_rate": 3.814728537770285e-06, "loss": 0.0089, "step": 70490 }, { "grad_norm": 0.36768633127212524, "learning_rate": 3.806814358283306e-06, "loss": 0.0109, "step": 70500 }, { "grad_norm": 0.3635143041610718, "learning_rate": 3.7989080719483596e-06, "loss": 0.0085, "step": 70510 }, { "grad_norm": 0.18677377700805664, "learning_rate": 3.7910096801164143e-06, "loss": 0.0103, "step": 70520 }, { "grad_norm": 0.1437114179134369, "learning_rate": 3.7831191841371016e-06, "loss": 0.0069, "step": 70530 }, { "grad_norm": 0.2858976125717163, "learning_rate": 3.775236585358688e-06, "loss": 0.01, "step": 70540 }, { "grad_norm": 0.12199685722589493, "learning_rate": 3.7673618851280843e-06, "loss": 0.0108, "step": 70550 }, { "grad_norm": 0.22131922841072083, "learning_rate": 3.759495084790887e-06, "loss": 0.0096, "step": 70560 }, { "grad_norm": 0.2694813013076782, "learning_rate": 3.751636185691282e-06, "loss": 0.0061, "step": 70570 }, { "grad_norm": 0.08398398756980896, "learning_rate": 3.7437851891721607e-06, "loss": 0.0077, "step": 70580 }, { "grad_norm": 0.12178926169872284, "learning_rate": 3.7359420965750404e-06, "loss": 0.0087, "step": 70590 }, { "grad_norm": 0.11763245612382889, "learning_rate": 3.7281069092400922e-06, "loss": 0.0129, "step": 70600 }, { "grad_norm": 0.1734975278377533, "learning_rate": 3.7202796285061348e-06, "loss": 0.0069, "step": 70610 }, { "grad_norm": 0.11819315701723099, "learning_rate": 3.712460255710637e-06, "loss": 0.0106, "step": 70620 }, { "grad_norm": 0.20976898074150085, "learning_rate": 3.704648792189719e-06, "loss": 0.0112, "step": 70630 }, { "grad_norm": 0.25788530707359314, "learning_rate": 3.696845239278124e-06, "loss": 0.0127, "step": 70640 }, { "grad_norm": 0.21961568295955658, "learning_rate": 3.689049598309302e-06, "loss": 0.0133, "step": 70650 }, { "grad_norm": 0.21476009488105774, "learning_rate": 3.681261870615288e-06, "loss": 0.0165, "step": 70660 }, { "grad_norm": 0.14951495826244354, "learning_rate": 3.6734820575268004e-06, "loss": 0.0144, "step": 70670 }, { "grad_norm": 0.26823490858078003, "learning_rate": 3.665710160373209e-06, "loss": 0.0081, "step": 70680 }, { "grad_norm": 0.33211749792099, "learning_rate": 3.65794618048248e-06, "loss": 0.0079, "step": 70690 }, { "grad_norm": 0.1658501923084259, "learning_rate": 3.6501901191813125e-06, "loss": 0.0069, "step": 70700 }, { "grad_norm": 0.10935774445533752, "learning_rate": 3.642441977794975e-06, "loss": 0.0093, "step": 70710 }, { "grad_norm": 0.1662626564502716, "learning_rate": 3.634701757647424e-06, "loss": 0.0099, "step": 70720 }, { "grad_norm": 0.21805444359779358, "learning_rate": 3.6269694600612468e-06, "loss": 0.0064, "step": 70730 }, { "grad_norm": 0.2512490153312683, "learning_rate": 3.619245086357681e-06, "loss": 0.0101, "step": 70740 }, { "grad_norm": 0.58170086145401, "learning_rate": 3.611528637856615e-06, "loss": 0.0127, "step": 70750 }, { "grad_norm": 0.6724708080291748, "learning_rate": 3.6038201158765884e-06, "loss": 0.0085, "step": 70760 }, { "grad_norm": 0.5371375679969788, "learning_rate": 3.5961195217347534e-06, "loss": 0.0157, "step": 70770 }, { "grad_norm": 0.15025679767131805, "learning_rate": 3.588426856746946e-06, "loss": 0.0088, "step": 70780 }, { "grad_norm": 0.27726858854293823, "learning_rate": 3.5807421222276316e-06, "loss": 0.0128, "step": 70790 }, { "grad_norm": 0.11600928008556366, "learning_rate": 3.5730653194899156e-06, "loss": 0.0108, "step": 70800 }, { "grad_norm": 0.1592976301908493, "learning_rate": 3.565396449845554e-06, "loss": 0.0077, "step": 70810 }, { "grad_norm": 0.44927939772605896, "learning_rate": 3.557735514604954e-06, "loss": 0.0093, "step": 70820 }, { "grad_norm": 0.1232399269938469, "learning_rate": 3.5500825150771633e-06, "loss": 0.0075, "step": 70830 }, { "grad_norm": 0.10646726191043854, "learning_rate": 3.5424374525698466e-06, "loss": 0.0079, "step": 70840 }, { "grad_norm": 0.2796337306499481, "learning_rate": 3.5348003283893704e-06, "loss": 0.0085, "step": 70850 }, { "grad_norm": 0.1497238427400589, "learning_rate": 3.527171143840685e-06, "loss": 0.0093, "step": 70860 }, { "grad_norm": 0.19812645018100739, "learning_rate": 3.519549900227409e-06, "loss": 0.0105, "step": 70870 }, { "grad_norm": 0.18272551894187927, "learning_rate": 3.511936598851828e-06, "loss": 0.0098, "step": 70880 }, { "grad_norm": 0.26847943663597107, "learning_rate": 3.504331241014813e-06, "loss": 0.0081, "step": 70890 }, { "grad_norm": 0.30091339349746704, "learning_rate": 3.4967338280159414e-06, "loss": 0.0075, "step": 70900 }, { "grad_norm": 0.2114991843700409, "learning_rate": 3.4891443611533846e-06, "loss": 0.0128, "step": 70910 }, { "grad_norm": 0.1332789957523346, "learning_rate": 3.481562841723984e-06, "loss": 0.011, "step": 70920 }, { "grad_norm": 0.22261404991149902, "learning_rate": 3.473989271023215e-06, "loss": 0.0124, "step": 70930 }, { "grad_norm": 0.21730393171310425, "learning_rate": 3.4664236503451854e-06, "loss": 0.009, "step": 70940 }, { "grad_norm": 0.06459248065948486, "learning_rate": 3.4588659809826674e-06, "loss": 0.0094, "step": 70950 }, { "grad_norm": 0.2527131140232086, "learning_rate": 3.451316264227039e-06, "loss": 0.0134, "step": 70960 }, { "grad_norm": 0.151185542345047, "learning_rate": 3.4437745013683574e-06, "loss": 0.0075, "step": 70970 }, { "grad_norm": 0.13603314757347107, "learning_rate": 3.4362406936952916e-06, "loss": 0.0073, "step": 70980 }, { "grad_norm": 0.18004296720027924, "learning_rate": 3.4287148424951786e-06, "loss": 0.0085, "step": 70990 }, { "grad_norm": 0.07387933880090714, "learning_rate": 3.4211969490539617e-06, "loss": 0.0078, "step": 71000 }, { "grad_norm": 0.1617453396320343, "learning_rate": 3.4136870146562584e-06, "loss": 0.0074, "step": 71010 }, { "grad_norm": 0.13794708251953125, "learning_rate": 3.406185040585308e-06, "loss": 0.0076, "step": 71020 }, { "grad_norm": 0.20764683187007904, "learning_rate": 3.3986910281229966e-06, "loss": 0.0095, "step": 71030 }, { "grad_norm": 0.20550689101219177, "learning_rate": 3.391204978549828e-06, "loss": 0.0113, "step": 71040 }, { "grad_norm": 0.2642500400543213, "learning_rate": 3.3837268931449785e-06, "loss": 0.0175, "step": 71050 }, { "grad_norm": 0.1187295913696289, "learning_rate": 3.376256773186248e-06, "loss": 0.0109, "step": 71060 }, { "grad_norm": 0.16755211353302002, "learning_rate": 3.3687946199500664e-06, "loss": 0.012, "step": 71070 }, { "grad_norm": 0.18557381629943848, "learning_rate": 3.3613404347115295e-06, "loss": 0.0085, "step": 71080 }, { "grad_norm": 0.07461759448051453, "learning_rate": 3.35389421874433e-06, "loss": 0.0062, "step": 71090 }, { "grad_norm": 0.15015915036201477, "learning_rate": 3.3464559733208446e-06, "loss": 0.0074, "step": 71100 }, { "grad_norm": 0.369852215051651, "learning_rate": 3.3390256997120505e-06, "loss": 0.0151, "step": 71110 }, { "grad_norm": 0.2934114933013916, "learning_rate": 3.331603399187583e-06, "loss": 0.0151, "step": 71120 }, { "grad_norm": 0.18821442127227783, "learning_rate": 3.324189073015721e-06, "loss": 0.0096, "step": 71130 }, { "grad_norm": 0.16611731052398682, "learning_rate": 3.3167827224633465e-06, "loss": 0.0116, "step": 71140 }, { "grad_norm": 0.15041448175907135, "learning_rate": 3.309384348796024e-06, "loss": 0.0111, "step": 71150 }, { "grad_norm": 0.2555333971977234, "learning_rate": 3.301993953277921e-06, "loss": 0.0104, "step": 71160 }, { "grad_norm": 0.18727026879787445, "learning_rate": 3.29461153717186e-06, "loss": 0.0078, "step": 71170 }, { "grad_norm": 0.12484908849000931, "learning_rate": 3.287237101739293e-06, "loss": 0.0126, "step": 71180 }, { "grad_norm": 0.4492403566837311, "learning_rate": 3.2798706482403075e-06, "loss": 0.0123, "step": 71190 }, { "grad_norm": 0.26605042815208435, "learning_rate": 3.2725121779336285e-06, "loss": 0.0108, "step": 71200 }, { "grad_norm": 0.25040560960769653, "learning_rate": 3.265161692076618e-06, "loss": 0.0098, "step": 71210 }, { "grad_norm": 0.2160131186246872, "learning_rate": 3.257819191925282e-06, "loss": 0.0075, "step": 71220 }, { "grad_norm": 0.13357581198215485, "learning_rate": 3.2504846787342392e-06, "loss": 0.0112, "step": 71230 }, { "grad_norm": 0.12943251430988312, "learning_rate": 3.243158153756759e-06, "loss": 0.0102, "step": 71240 }, { "grad_norm": 0.0859222486615181, "learning_rate": 3.2358396182447514e-06, "loss": 0.0084, "step": 71250 }, { "grad_norm": 0.11782974004745483, "learning_rate": 3.2285290734487496e-06, "loss": 0.0108, "step": 71260 }, { "grad_norm": 0.12273546308279037, "learning_rate": 3.221226520617926e-06, "loss": 0.0097, "step": 71270 }, { "grad_norm": 0.19863630831241608, "learning_rate": 3.213931961000094e-06, "loss": 0.0126, "step": 71280 }, { "grad_norm": 0.2342141717672348, "learning_rate": 3.2066453958416733e-06, "loss": 0.0072, "step": 71290 }, { "grad_norm": 0.2664853632450104, "learning_rate": 3.199366826387773e-06, "loss": 0.009, "step": 71300 }, { "grad_norm": 0.13229040801525116, "learning_rate": 3.192096253882071e-06, "loss": 0.01, "step": 71310 }, { "grad_norm": 0.6711251735687256, "learning_rate": 3.1848336795669177e-06, "loss": 0.0114, "step": 71320 }, { "grad_norm": 0.1649841070175171, "learning_rate": 3.1775791046833035e-06, "loss": 0.0114, "step": 71330 }, { "grad_norm": 0.14988471567630768, "learning_rate": 3.170332530470804e-06, "loss": 0.0099, "step": 71340 }, { "grad_norm": 0.3301732838153839, "learning_rate": 3.1630939581677012e-06, "loss": 0.0114, "step": 71350 }, { "grad_norm": 0.18086205422878265, "learning_rate": 3.155863389010838e-06, "loss": 0.0113, "step": 71360 }, { "grad_norm": 0.12244300544261932, "learning_rate": 3.1486408242357323e-06, "loss": 0.0102, "step": 71370 }, { "grad_norm": 0.4473274350166321, "learning_rate": 3.1414262650765248e-06, "loss": 0.017, "step": 71380 }, { "grad_norm": 0.27243244647979736, "learning_rate": 3.134219712765979e-06, "loss": 0.009, "step": 71390 }, { "grad_norm": 0.16680093109607697, "learning_rate": 3.1270211685355044e-06, "loss": 0.0107, "step": 71400 }, { "grad_norm": 0.10173242539167404, "learning_rate": 3.1198306336151338e-06, "loss": 0.0097, "step": 71410 }, { "grad_norm": 0.19677354395389557, "learning_rate": 3.11264810923354e-06, "loss": 0.0105, "step": 71420 }, { "grad_norm": 0.14328616857528687, "learning_rate": 3.105473596618008e-06, "loss": 0.009, "step": 71430 }, { "grad_norm": 0.30131930112838745, "learning_rate": 3.0983070969944683e-06, "loss": 0.0098, "step": 71440 }, { "grad_norm": 0.1010834202170372, "learning_rate": 3.09114861158748e-06, "loss": 0.0076, "step": 71450 }, { "grad_norm": 0.18418538570404053, "learning_rate": 3.0839981416202314e-06, "loss": 0.0109, "step": 71460 }, { "grad_norm": 0.21292144060134888, "learning_rate": 3.0768556883145507e-06, "loss": 0.0061, "step": 71470 }, { "grad_norm": 0.13052234053611755, "learning_rate": 3.0697212528908834e-06, "loss": 0.0067, "step": 71480 }, { "grad_norm": 0.11687584221363068, "learning_rate": 3.0625948365683e-06, "loss": 0.009, "step": 71490 }, { "grad_norm": 0.09819872677326202, "learning_rate": 3.055476440564531e-06, "loss": 0.0112, "step": 71500 }, { "grad_norm": 0.15572558343410492, "learning_rate": 3.0483660660958924e-06, "loss": 0.0078, "step": 71510 }, { "grad_norm": 0.3891461491584778, "learning_rate": 3.0412637143773624e-06, "loss": 0.0117, "step": 71520 }, { "grad_norm": 0.1529858559370041, "learning_rate": 3.034169386622554e-06, "loss": 0.0125, "step": 71530 }, { "grad_norm": 0.1329629272222519, "learning_rate": 3.027083084043658e-06, "loss": 0.0111, "step": 71540 }, { "grad_norm": 0.1404361128807068, "learning_rate": 3.0200048078515676e-06, "loss": 0.0089, "step": 71550 }, { "grad_norm": 0.29989758133888245, "learning_rate": 3.012934559255742e-06, "loss": 0.0085, "step": 71560 }, { "grad_norm": 0.4310736060142517, "learning_rate": 3.0058723394642994e-06, "loss": 0.0108, "step": 71570 }, { "grad_norm": 0.10158492624759674, "learning_rate": 2.998818149683985e-06, "loss": 0.0168, "step": 71580 }, { "grad_norm": 0.2821391224861145, "learning_rate": 2.9917719911201627e-06, "loss": 0.011, "step": 71590 }, { "grad_norm": 0.09999343007802963, "learning_rate": 2.9847338649768352e-06, "loss": 0.0113, "step": 71600 }, { "grad_norm": 0.25299105048179626, "learning_rate": 2.977703772456608e-06, "loss": 0.0095, "step": 71610 }, { "grad_norm": 0.19566017389297485, "learning_rate": 2.9706817147607535e-06, "loss": 0.009, "step": 71620 }, { "grad_norm": 0.10399307310581207, "learning_rate": 2.9636676930891393e-06, "loss": 0.0106, "step": 71630 }, { "grad_norm": 0.16420680284500122, "learning_rate": 2.9566617086402625e-06, "loss": 0.0078, "step": 71640 }, { "grad_norm": 0.13939890265464783, "learning_rate": 2.9496637626112655e-06, "loss": 0.0103, "step": 71650 }, { "grad_norm": 0.1523434966802597, "learning_rate": 2.9426738561979027e-06, "loss": 0.0113, "step": 71660 }, { "grad_norm": 0.4439730942249298, "learning_rate": 2.9356919905945524e-06, "loss": 0.0269, "step": 71670 }, { "grad_norm": 0.226570263504982, "learning_rate": 2.928718166994243e-06, "loss": 0.0118, "step": 71680 }, { "grad_norm": 0.2358975112438202, "learning_rate": 2.9217523865885833e-06, "loss": 0.0112, "step": 71690 }, { "grad_norm": 0.09798508882522583, "learning_rate": 2.9147946505678537e-06, "loss": 0.0084, "step": 71700 }, { "grad_norm": 0.2771819829940796, "learning_rate": 2.9078449601209313e-06, "loss": 0.0078, "step": 71710 }, { "grad_norm": 0.4335983395576477, "learning_rate": 2.900903316435333e-06, "loss": 0.014, "step": 71720 }, { "grad_norm": 0.1594998985528946, "learning_rate": 2.8939697206971983e-06, "loss": 0.0116, "step": 71730 }, { "grad_norm": 0.10402462631464005, "learning_rate": 2.8870441740912746e-06, "loss": 0.0067, "step": 71740 }, { "grad_norm": 0.2138816863298416, "learning_rate": 2.8801266778009762e-06, "loss": 0.0107, "step": 71750 }, { "grad_norm": 0.17570476233959198, "learning_rate": 2.8732172330082798e-06, "loss": 0.0157, "step": 71760 }, { "grad_norm": 0.184209942817688, "learning_rate": 2.8663158408938517e-06, "loss": 0.0092, "step": 71770 }, { "grad_norm": 0.16955707967281342, "learning_rate": 2.859422502636938e-06, "loss": 0.0067, "step": 71780 }, { "grad_norm": 0.14441610872745514, "learning_rate": 2.8525372194154076e-06, "loss": 0.0067, "step": 71790 }, { "grad_norm": 0.19655677676200867, "learning_rate": 2.8456599924057913e-06, "loss": 0.009, "step": 71800 }, { "grad_norm": 0.22280439734458923, "learning_rate": 2.8387908227831995e-06, "loss": 0.0136, "step": 71810 }, { "grad_norm": 0.13974452018737793, "learning_rate": 2.831929711721404e-06, "loss": 0.0077, "step": 71820 }, { "grad_norm": 0.21889741718769073, "learning_rate": 2.825076660392767e-06, "loss": 0.0057, "step": 71830 }, { "grad_norm": 0.14142951369285583, "learning_rate": 2.8182316699682908e-06, "loss": 0.0093, "step": 71840 }, { "grad_norm": 0.17818595468997955, "learning_rate": 2.811394741617601e-06, "loss": 0.0084, "step": 71850 }, { "grad_norm": 0.5080118179321289, "learning_rate": 2.8045658765089356e-06, "loss": 0.0173, "step": 71860 }, { "grad_norm": 0.4866548478603363, "learning_rate": 2.7977450758091605e-06, "loss": 0.015, "step": 71870 }, { "grad_norm": 0.2238396555185318, "learning_rate": 2.790932340683783e-06, "loss": 0.0129, "step": 71880 }, { "grad_norm": 0.20564472675323486, "learning_rate": 2.7841276722968823e-06, "loss": 0.0082, "step": 71890 }, { "grad_norm": 0.09791535139083862, "learning_rate": 2.7773310718112067e-06, "loss": 0.0065, "step": 71900 }, { "grad_norm": 0.16460248827934265, "learning_rate": 2.7705425403881102e-06, "loss": 0.0076, "step": 71910 }, { "grad_norm": 0.1373036950826645, "learning_rate": 2.763762079187565e-06, "loss": 0.007, "step": 71920 }, { "grad_norm": 0.10897839069366455, "learning_rate": 2.756989689368178e-06, "loss": 0.0126, "step": 71930 }, { "grad_norm": 0.18740344047546387, "learning_rate": 2.750225372087134e-06, "loss": 0.0086, "step": 71940 }, { "grad_norm": 0.31563177704811096, "learning_rate": 2.7434691285003033e-06, "loss": 0.0111, "step": 71950 }, { "grad_norm": 0.34821102023124695, "learning_rate": 2.7367209597621178e-06, "loss": 0.0104, "step": 71960 }, { "grad_norm": 0.131551131606102, "learning_rate": 2.729980867025683e-06, "loss": 0.0102, "step": 71970 }, { "grad_norm": 0.26943764090538025, "learning_rate": 2.7232488514426724e-06, "loss": 0.0086, "step": 71980 }, { "grad_norm": 0.22868046164512634, "learning_rate": 2.7165249141634097e-06, "loss": 0.0077, "step": 71990 }, { "grad_norm": 0.17964473366737366, "learning_rate": 2.709809056336837e-06, "loss": 0.012, "step": 72000 }, { "grad_norm": 0.1928817629814148, "learning_rate": 2.703101279110498e-06, "loss": 0.011, "step": 72010 }, { "grad_norm": 0.22738075256347656, "learning_rate": 2.696401583630587e-06, "loss": 0.0155, "step": 72020 }, { "grad_norm": 0.16406850516796112, "learning_rate": 2.689709971041887e-06, "loss": 0.0103, "step": 72030 }, { "grad_norm": 0.16995450854301453, "learning_rate": 2.6830264424878126e-06, "loss": 0.0094, "step": 72040 }, { "grad_norm": 0.21126961708068848, "learning_rate": 2.676350999110394e-06, "loss": 0.0152, "step": 72050 }, { "grad_norm": 0.24944619834423065, "learning_rate": 2.6696836420502856e-06, "loss": 0.012, "step": 72060 }, { "grad_norm": 0.37646204233169556, "learning_rate": 2.663024372446765e-06, "loss": 0.0079, "step": 72070 }, { "grad_norm": 0.16074049472808838, "learning_rate": 2.6563731914377055e-06, "loss": 0.0106, "step": 72080 }, { "grad_norm": 0.14210836589336395, "learning_rate": 2.6497301001596087e-06, "loss": 0.0125, "step": 72090 }, { "grad_norm": 0.3525482714176178, "learning_rate": 2.643095099747611e-06, "loss": 0.0112, "step": 72100 }, { "grad_norm": 0.1352308690547943, "learning_rate": 2.636468191335445e-06, "loss": 0.0092, "step": 72110 }, { "grad_norm": 0.18178075551986694, "learning_rate": 2.629849376055471e-06, "loss": 0.01, "step": 72120 }, { "grad_norm": 0.3284956216812134, "learning_rate": 2.6232386550386678e-06, "loss": 0.0106, "step": 72130 }, { "grad_norm": 0.1630285233259201, "learning_rate": 2.6166360294146097e-06, "loss": 0.0093, "step": 72140 }, { "grad_norm": 0.27965787053108215, "learning_rate": 2.6100415003115275e-06, "loss": 0.0156, "step": 72150 }, { "grad_norm": 0.21619977056980133, "learning_rate": 2.603455068856225e-06, "loss": 0.0109, "step": 72160 }, { "grad_norm": 0.10048236697912216, "learning_rate": 2.5968767361741584e-06, "loss": 0.0094, "step": 72170 }, { "grad_norm": 0.18591895699501038, "learning_rate": 2.590306503389378e-06, "loss": 0.0102, "step": 72180 }, { "grad_norm": 0.11543980240821838, "learning_rate": 2.583744371624558e-06, "loss": 0.0134, "step": 72190 }, { "grad_norm": 0.24267318844795227, "learning_rate": 2.57719034200099e-06, "loss": 0.0084, "step": 72200 }, { "grad_norm": 0.18430060148239136, "learning_rate": 2.570644415638568e-06, "loss": 0.0109, "step": 72210 }, { "grad_norm": 0.10473116487264633, "learning_rate": 2.5641065936558295e-06, "loss": 0.0116, "step": 72220 }, { "grad_norm": 0.13934196531772614, "learning_rate": 2.557576877169898e-06, "loss": 0.01, "step": 72230 }, { "grad_norm": 0.14492686092853546, "learning_rate": 2.5510552672965205e-06, "loss": 0.0084, "step": 72240 }, { "grad_norm": 0.14654113352298737, "learning_rate": 2.544541765150077e-06, "loss": 0.0104, "step": 72250 }, { "grad_norm": 0.13339431583881378, "learning_rate": 2.5380363718435165e-06, "loss": 0.0093, "step": 72260 }, { "grad_norm": 0.15391190350055695, "learning_rate": 2.5315390884884714e-06, "loss": 0.007, "step": 72270 }, { "grad_norm": 0.1364917904138565, "learning_rate": 2.5250499161951214e-06, "loss": 0.0116, "step": 72280 }, { "grad_norm": 0.18884558975696564, "learning_rate": 2.518568856072301e-06, "loss": 0.0081, "step": 72290 }, { "grad_norm": 0.1459314525127411, "learning_rate": 2.512095909227441e-06, "loss": 0.0093, "step": 72300 }, { "grad_norm": 0.19136863946914673, "learning_rate": 2.50563107676659e-06, "loss": 0.0106, "step": 72310 }, { "grad_norm": 0.15620754659175873, "learning_rate": 2.4991743597944152e-06, "loss": 0.0097, "step": 72320 }, { "grad_norm": 0.1266719251871109, "learning_rate": 2.492725759414205e-06, "loss": 0.0095, "step": 72330 }, { "grad_norm": 0.5257719159126282, "learning_rate": 2.4862852767278234e-06, "loss": 0.0125, "step": 72340 }, { "grad_norm": 0.1258823275566101, "learning_rate": 2.479852912835784e-06, "loss": 0.0075, "step": 72350 }, { "grad_norm": 0.45362409949302673, "learning_rate": 2.4734286688372075e-06, "loss": 0.0127, "step": 72360 }, { "grad_norm": 0.5171413421630859, "learning_rate": 2.467012545829811e-06, "loss": 0.0149, "step": 72370 }, { "grad_norm": 0.13503867387771606, "learning_rate": 2.460604544909945e-06, "loss": 0.0093, "step": 72380 }, { "grad_norm": 0.18243741989135742, "learning_rate": 2.454204667172555e-06, "loss": 0.0071, "step": 72390 }, { "grad_norm": 0.11593993753194809, "learning_rate": 2.447812913711217e-06, "loss": 0.0101, "step": 72400 }, { "grad_norm": 0.14555861055850983, "learning_rate": 2.44142928561808e-06, "loss": 0.0131, "step": 72410 }, { "grad_norm": 0.19493722915649414, "learning_rate": 2.435053783983965e-06, "loss": 0.0089, "step": 72420 }, { "grad_norm": 0.15631701052188873, "learning_rate": 2.4286864098982453e-06, "loss": 0.0086, "step": 72430 }, { "grad_norm": 0.11364493519067764, "learning_rate": 2.4223271644489397e-06, "loss": 0.0111, "step": 72440 }, { "grad_norm": 0.21812890470027924, "learning_rate": 2.415976048722679e-06, "loss": 0.0083, "step": 72450 }, { "grad_norm": 0.10966145992279053, "learning_rate": 2.4096330638046673e-06, "loss": 0.0081, "step": 72460 }, { "grad_norm": 0.17494606971740723, "learning_rate": 2.4032982107787816e-06, "loss": 0.0126, "step": 72470 }, { "grad_norm": 0.1372814178466797, "learning_rate": 2.396971490727451e-06, "loss": 0.0096, "step": 72480 }, { "grad_norm": 0.09304957091808319, "learning_rate": 2.3906529047317493e-06, "loss": 0.0066, "step": 72490 }, { "grad_norm": 0.16224245727062225, "learning_rate": 2.3843424538713465e-06, "loss": 0.0082, "step": 72500 }, { "grad_norm": 0.28563061356544495, "learning_rate": 2.3780401392245298e-06, "loss": 0.0109, "step": 72510 }, { "grad_norm": 0.10514955222606659, "learning_rate": 2.3717459618681883e-06, "loss": 0.0081, "step": 72520 }, { "grad_norm": 0.20021654665470123, "learning_rate": 2.3654599228778274e-06, "loss": 0.0087, "step": 72530 }, { "grad_norm": 0.26245906949043274, "learning_rate": 2.3591820233275607e-06, "loss": 0.0093, "step": 72540 }, { "grad_norm": 0.1516042947769165, "learning_rate": 2.3529122642901024e-06, "loss": 0.0083, "step": 72550 }, { "grad_norm": 0.11972258985042572, "learning_rate": 2.3466506468367845e-06, "loss": 0.0127, "step": 72560 }, { "grad_norm": 0.10862264782190323, "learning_rate": 2.3403971720375506e-06, "loss": 0.0078, "step": 72570 }, { "grad_norm": 0.16569837927818298, "learning_rate": 2.334151840960952e-06, "loss": 0.0095, "step": 72580 }, { "grad_norm": 0.17929400503635406, "learning_rate": 2.3279146546741347e-06, "loss": 0.0075, "step": 72590 }, { "grad_norm": 0.12317843735218048, "learning_rate": 2.32168561424288e-06, "loss": 0.0068, "step": 72600 }, { "grad_norm": 1.1744557619094849, "learning_rate": 2.3154647207315307e-06, "loss": 0.0127, "step": 72610 }, { "grad_norm": 0.2166675329208374, "learning_rate": 2.309251975203103e-06, "loss": 0.0106, "step": 72620 }, { "grad_norm": 0.07616392523050308, "learning_rate": 2.303047378719159e-06, "loss": 0.0126, "step": 72630 }, { "grad_norm": 0.2598508894443512, "learning_rate": 2.2968509323399e-06, "loss": 0.0109, "step": 72640 }, { "grad_norm": 0.16947820782661438, "learning_rate": 2.290662637124147e-06, "loss": 0.0077, "step": 72650 }, { "grad_norm": 0.07025105506181717, "learning_rate": 2.2844824941292807e-06, "loss": 0.0076, "step": 72660 }, { "grad_norm": 0.3914799094200134, "learning_rate": 2.2783105044113406e-06, "loss": 0.0109, "step": 72670 }, { "grad_norm": 0.2725759744644165, "learning_rate": 2.272146669024944e-06, "loss": 0.0104, "step": 72680 }, { "grad_norm": 0.16783952713012695, "learning_rate": 2.26599098902332e-06, "loss": 0.0088, "step": 72690 }, { "grad_norm": 0.33471837639808655, "learning_rate": 2.2598434654583113e-06, "loss": 0.0116, "step": 72700 }, { "grad_norm": 0.3696013391017914, "learning_rate": 2.253704099380355e-06, "loss": 0.0146, "step": 72710 }, { "grad_norm": 0.18069994449615479, "learning_rate": 2.247572891838512e-06, "loss": 0.0106, "step": 72720 }, { "grad_norm": 0.602327287197113, "learning_rate": 2.241449843880422e-06, "loss": 0.0111, "step": 72730 }, { "grad_norm": 0.17767372727394104, "learning_rate": 2.235334956552354e-06, "loss": 0.0128, "step": 72740 }, { "grad_norm": 0.10526440292596817, "learning_rate": 2.2292282308991775e-06, "loss": 0.0069, "step": 72750 }, { "grad_norm": 0.2845573425292969, "learning_rate": 2.223129667964363e-06, "loss": 0.0098, "step": 72760 }, { "grad_norm": 0.25274303555488586, "learning_rate": 2.2170392687899834e-06, "loss": 0.0067, "step": 72770 }, { "grad_norm": 0.12024152278900146, "learning_rate": 2.210957034416733e-06, "loss": 0.0062, "step": 72780 }, { "grad_norm": 0.1660454273223877, "learning_rate": 2.2048829658838867e-06, "loss": 0.0132, "step": 72790 }, { "grad_norm": 0.0799790471792221, "learning_rate": 2.1988170642293525e-06, "loss": 0.0072, "step": 72800 }, { "grad_norm": 0.16293483972549438, "learning_rate": 2.1927593304896075e-06, "loss": 0.008, "step": 72810 }, { "grad_norm": 0.11558908224105835, "learning_rate": 2.1867097656997626e-06, "loss": 0.0055, "step": 72820 }, { "grad_norm": 0.06553023308515549, "learning_rate": 2.180668370893524e-06, "loss": 0.0064, "step": 72830 }, { "grad_norm": 0.19305653870105743, "learning_rate": 2.174635147103199e-06, "loss": 0.0064, "step": 72840 }, { "grad_norm": 0.27769410610198975, "learning_rate": 2.1686100953597075e-06, "loss": 0.009, "step": 72850 }, { "grad_norm": 0.1750466525554657, "learning_rate": 2.1625932166925433e-06, "loss": 0.0084, "step": 72860 }, { "grad_norm": 0.10712462663650513, "learning_rate": 2.1565845121298556e-06, "loss": 0.0069, "step": 72870 }, { "grad_norm": 0.1112259030342102, "learning_rate": 2.150583982698351e-06, "loss": 0.0101, "step": 72880 }, { "grad_norm": 0.08308541774749756, "learning_rate": 2.144591629423359e-06, "loss": 0.0057, "step": 72890 }, { "grad_norm": 0.16870296001434326, "learning_rate": 2.138607453328817e-06, "loss": 0.0098, "step": 72900 }, { "grad_norm": 0.26837173104286194, "learning_rate": 2.132631455437234e-06, "loss": 0.0096, "step": 72910 }, { "grad_norm": 0.22726255655288696, "learning_rate": 2.1266636367697714e-06, "loss": 0.0136, "step": 72920 }, { "grad_norm": 0.24233292043209076, "learning_rate": 2.120703998346152e-06, "loss": 0.0106, "step": 72930 }, { "grad_norm": 0.19692379236221313, "learning_rate": 2.1147525411847114e-06, "loss": 0.0101, "step": 72940 }, { "grad_norm": 0.20193138718605042, "learning_rate": 2.108809266302403e-06, "loss": 0.0075, "step": 72950 }, { "grad_norm": 0.29697906970977783, "learning_rate": 2.102874174714764e-06, "loss": 0.0101, "step": 72960 }, { "grad_norm": 0.2616409659385681, "learning_rate": 2.0969472674359393e-06, "loss": 0.0094, "step": 72970 }, { "grad_norm": 0.07745181024074554, "learning_rate": 2.091028545478668e-06, "loss": 0.0129, "step": 72980 }, { "grad_norm": 0.1295359581708908, "learning_rate": 2.08511800985432e-06, "loss": 0.009, "step": 72990 }, { "grad_norm": 0.07438524812459946, "learning_rate": 2.07921566157282e-06, "loss": 0.012, "step": 73000 }, { "grad_norm": 0.11905281245708466, "learning_rate": 2.073321501642728e-06, "loss": 0.0076, "step": 73010 }, { "grad_norm": 0.11527429521083832, "learning_rate": 2.0674355310711937e-06, "loss": 0.007, "step": 73020 }, { "grad_norm": 0.09974952042102814, "learning_rate": 2.0615577508639682e-06, "loss": 0.0088, "step": 73030 }, { "grad_norm": 0.12452156841754913, "learning_rate": 2.055688162025399e-06, "loss": 0.0071, "step": 73040 }, { "grad_norm": 0.13640309870243073, "learning_rate": 2.0498267655584546e-06, "loss": 0.0091, "step": 73050 }, { "grad_norm": 0.22463546693325043, "learning_rate": 2.0439735624646626e-06, "loss": 0.0099, "step": 73060 }, { "grad_norm": 0.11979062110185623, "learning_rate": 2.0381285537442008e-06, "loss": 0.0095, "step": 73070 }, { "grad_norm": 0.18134532868862152, "learning_rate": 2.032291740395803e-06, "loss": 0.0069, "step": 73080 }, { "grad_norm": 0.1596367359161377, "learning_rate": 2.0264631234168276e-06, "loss": 0.0118, "step": 73090 }, { "grad_norm": 0.3050866425037384, "learning_rate": 2.0206427038032326e-06, "loss": 0.0125, "step": 73100 }, { "grad_norm": 0.25643637776374817, "learning_rate": 2.0148304825495457e-06, "loss": 0.0088, "step": 73110 }, { "grad_norm": 0.10005790740251541, "learning_rate": 2.0090264606489496e-06, "loss": 0.0088, "step": 73120 }, { "grad_norm": 0.2053322196006775, "learning_rate": 2.003230639093162e-06, "loss": 0.0084, "step": 73130 }, { "grad_norm": 0.2316114902496338, "learning_rate": 1.9974430188725525e-06, "loss": 0.0136, "step": 73140 }, { "grad_norm": 0.2068309336900711, "learning_rate": 1.9916636009760513e-06, "loss": 0.0108, "step": 73150 }, { "grad_norm": 0.28221216797828674, "learning_rate": 1.985892386391208e-06, "loss": 0.0105, "step": 73160 }, { "grad_norm": 0.06790229678153992, "learning_rate": 1.980129376104173e-06, "loss": 0.0109, "step": 73170 }, { "grad_norm": 0.1256982535123825, "learning_rate": 1.9743745710996796e-06, "loss": 0.0076, "step": 73180 }, { "grad_norm": 0.1331498920917511, "learning_rate": 1.9686279723610757e-06, "loss": 0.0073, "step": 73190 }, { "grad_norm": 0.19917389750480652, "learning_rate": 1.962889580870281e-06, "loss": 0.0125, "step": 73200 }, { "grad_norm": 0.10261938720941544, "learning_rate": 1.9571593976078385e-06, "loss": 0.0086, "step": 73210 }, { "grad_norm": 0.36126431822776794, "learning_rate": 1.9514374235528876e-06, "loss": 0.0102, "step": 73220 }, { "grad_norm": 0.2379264086484909, "learning_rate": 1.9457236596831408e-06, "loss": 0.0103, "step": 73230 }, { "grad_norm": 0.11266063153743744, "learning_rate": 1.940018106974939e-06, "loss": 0.0112, "step": 73240 }, { "grad_norm": 0.10924694687128067, "learning_rate": 1.9343207664032025e-06, "loss": 0.007, "step": 73250 }, { "grad_norm": 0.18569433689117432, "learning_rate": 1.928631638941436e-06, "loss": 0.0074, "step": 73260 }, { "grad_norm": 0.38645413517951965, "learning_rate": 1.922950725561784e-06, "loss": 0.013, "step": 73270 }, { "grad_norm": 0.09584292769432068, "learning_rate": 1.9172780272349312e-06, "loss": 0.0068, "step": 73280 }, { "grad_norm": 0.1571073681116104, "learning_rate": 1.911613544930202e-06, "loss": 0.01, "step": 73290 }, { "grad_norm": 0.16763101518154144, "learning_rate": 1.9059572796155e-06, "loss": 0.0106, "step": 73300 }, { "grad_norm": 0.27719423174858093, "learning_rate": 1.9003092322573192e-06, "loss": 0.0089, "step": 73310 }, { "grad_norm": 0.07761435210704803, "learning_rate": 1.8946694038207647e-06, "loss": 0.0101, "step": 73320 }, { "grad_norm": 0.2294331192970276, "learning_rate": 1.8890377952695215e-06, "loss": 0.0095, "step": 73330 }, { "grad_norm": 0.18103431165218353, "learning_rate": 1.8834144075658865e-06, "loss": 0.0082, "step": 73340 }, { "grad_norm": 0.18090933561325073, "learning_rate": 1.8777992416707357e-06, "loss": 0.0125, "step": 73350 }, { "grad_norm": 0.23254543542861938, "learning_rate": 1.8721922985435458e-06, "loss": 0.008, "step": 73360 }, { "grad_norm": 0.16905567049980164, "learning_rate": 1.8665935791424006e-06, "loss": 0.01, "step": 73370 }, { "grad_norm": 0.28307536244392395, "learning_rate": 1.8610030844239512e-06, "loss": 0.0059, "step": 73380 }, { "grad_norm": 0.19676212966442108, "learning_rate": 1.8554208153434838e-06, "loss": 0.0086, "step": 73390 }, { "grad_norm": 0.18172475695610046, "learning_rate": 1.8498467728548296e-06, "loss": 0.0095, "step": 73400 }, { "grad_norm": 0.1646672785282135, "learning_rate": 1.8442809579104547e-06, "loss": 0.0068, "step": 73410 }, { "grad_norm": 0.15624162554740906, "learning_rate": 1.8387233714614039e-06, "loss": 0.0082, "step": 73420 }, { "grad_norm": 0.26805204153060913, "learning_rate": 1.8331740144573174e-06, "loss": 0.0113, "step": 73430 }, { "grad_norm": 0.3238465487957001, "learning_rate": 1.8276328878464255e-06, "loss": 0.0081, "step": 73440 }, { "grad_norm": 0.17786255478858948, "learning_rate": 1.82209999257556e-06, "loss": 0.0086, "step": 73450 }, { "grad_norm": 0.18193727731704712, "learning_rate": 1.8165753295901311e-06, "loss": 0.0096, "step": 73460 }, { "grad_norm": 0.10628315806388855, "learning_rate": 1.8110588998341616e-06, "loss": 0.0081, "step": 73470 }, { "grad_norm": 0.1940620392560959, "learning_rate": 1.805550704250253e-06, "loss": 0.0169, "step": 73480 }, { "grad_norm": 0.27377042174339294, "learning_rate": 1.8000507437796077e-06, "loss": 0.0084, "step": 73490 }, { "grad_norm": 0.24788731336593628, "learning_rate": 1.7945590193620242e-06, "loss": 0.0094, "step": 73500 }, { "grad_norm": 0.19934436678886414, "learning_rate": 1.7890755319358742e-06, "loss": 0.0078, "step": 73510 }, { "grad_norm": 0.2820119857788086, "learning_rate": 1.7836002824381525e-06, "loss": 0.0097, "step": 73520 }, { "grad_norm": 0.2144618183374405, "learning_rate": 1.7781332718044165e-06, "loss": 0.008, "step": 73530 }, { "grad_norm": 0.247928187251091, "learning_rate": 1.772674500968835e-06, "loss": 0.0113, "step": 73540 }, { "grad_norm": 0.30487552285194397, "learning_rate": 1.7672239708641624e-06, "loss": 0.0106, "step": 73550 }, { "grad_norm": 0.15836159884929657, "learning_rate": 1.7617816824217371e-06, "loss": 0.0077, "step": 73560 }, { "grad_norm": 0.436926007270813, "learning_rate": 1.7563476365715148e-06, "loss": 0.0126, "step": 73570 }, { "grad_norm": 0.22353209555149078, "learning_rate": 1.750921834242003e-06, "loss": 0.0077, "step": 73580 }, { "grad_norm": 0.17196668684482574, "learning_rate": 1.7455042763603436e-06, "loss": 0.0085, "step": 73590 }, { "grad_norm": 0.16802281141281128, "learning_rate": 1.7400949638522345e-06, "loss": 0.0055, "step": 73600 }, { "grad_norm": 0.08870744705200195, "learning_rate": 1.734693897641987e-06, "loss": 0.0219, "step": 73610 }, { "grad_norm": 0.33399608731269836, "learning_rate": 1.7293010786524955e-06, "loss": 0.0092, "step": 73620 }, { "grad_norm": 0.30363649129867554, "learning_rate": 1.72391650780524e-06, "loss": 0.008, "step": 73630 }, { "grad_norm": 0.2646234631538391, "learning_rate": 1.718540186020301e-06, "loss": 0.009, "step": 73640 }, { "grad_norm": 0.19062408804893494, "learning_rate": 1.7131721142163437e-06, "loss": 0.0097, "step": 73650 }, { "grad_norm": 0.380537211894989, "learning_rate": 1.7078122933106233e-06, "loss": 0.0129, "step": 73660 }, { "grad_norm": 0.14336027204990387, "learning_rate": 1.7024607242189905e-06, "loss": 0.0082, "step": 73670 }, { "grad_norm": 0.31588318943977356, "learning_rate": 1.6971174078558749e-06, "loss": 0.0075, "step": 73680 }, { "grad_norm": 0.09285067021846771, "learning_rate": 1.6917823451343073e-06, "loss": 0.0066, "step": 73690 }, { "grad_norm": 0.22358085215091705, "learning_rate": 1.6864555369659142e-06, "loss": 0.0091, "step": 73700 }, { "grad_norm": 0.1948227435350418, "learning_rate": 1.681136984260878e-06, "loss": 0.0094, "step": 73710 }, { "grad_norm": 0.1800937056541443, "learning_rate": 1.6758266879280172e-06, "loss": 0.008, "step": 73720 }, { "grad_norm": 0.13371898233890533, "learning_rate": 1.6705246488746996e-06, "loss": 0.0108, "step": 73730 }, { "grad_norm": 0.17021287977695465, "learning_rate": 1.6652308680069062e-06, "loss": 0.0092, "step": 73740 }, { "grad_norm": 0.14621326327323914, "learning_rate": 1.6599453462292081e-06, "loss": 0.0079, "step": 73750 }, { "grad_norm": 0.517746090888977, "learning_rate": 1.6546680844447326e-06, "loss": 0.0089, "step": 73760 }, { "grad_norm": 0.12999780476093292, "learning_rate": 1.6493990835552475e-06, "loss": 0.0073, "step": 73770 }, { "grad_norm": 0.17467115819454193, "learning_rate": 1.6441383444610547e-06, "loss": 0.008, "step": 73780 }, { "grad_norm": 0.12490756809711456, "learning_rate": 1.6388858680610908e-06, "loss": 0.0084, "step": 73790 }, { "grad_norm": 0.10902697592973709, "learning_rate": 1.6336416552528543e-06, "loss": 0.0078, "step": 73800 }, { "grad_norm": 0.04854034632444382, "learning_rate": 1.6284057069324288e-06, "loss": 0.0142, "step": 73810 }, { "grad_norm": 0.210420623421669, "learning_rate": 1.623178023994504e-06, "loss": 0.0123, "step": 73820 }, { "grad_norm": 0.14899896085262299, "learning_rate": 1.6179586073323483e-06, "loss": 0.0086, "step": 73830 }, { "grad_norm": 0.3266611099243164, "learning_rate": 1.612747457837821e-06, "loss": 0.0067, "step": 73840 }, { "grad_norm": 0.15969762206077576, "learning_rate": 1.607544576401354e-06, "loss": 0.0098, "step": 73850 }, { "grad_norm": 0.2522030174732208, "learning_rate": 1.6023499639119754e-06, "loss": 0.0107, "step": 73860 }, { "grad_norm": 0.11789938062429428, "learning_rate": 1.5971636212573138e-06, "loss": 0.0055, "step": 73870 }, { "grad_norm": 0.1619093418121338, "learning_rate": 1.5919855493235714e-06, "loss": 0.0076, "step": 73880 }, { "grad_norm": 0.14327549934387207, "learning_rate": 1.586815748995535e-06, "loss": 0.0069, "step": 73890 }, { "grad_norm": 0.23717820644378662, "learning_rate": 1.5816542211565866e-06, "loss": 0.0083, "step": 73900 }, { "grad_norm": 0.16129019856452942, "learning_rate": 1.5765009666886765e-06, "loss": 0.0103, "step": 73910 }, { "grad_norm": 0.43208563327789307, "learning_rate": 1.5713559864723781e-06, "loss": 0.01, "step": 73920 }, { "grad_norm": 0.31528615951538086, "learning_rate": 1.5662192813868104e-06, "loss": 0.0089, "step": 73930 }, { "grad_norm": 0.10793508589267731, "learning_rate": 1.561090852309699e-06, "loss": 0.0118, "step": 73940 }, { "grad_norm": 0.18090030550956726, "learning_rate": 1.5559707001173651e-06, "loss": 0.0104, "step": 73950 }, { "grad_norm": 0.22714652121067047, "learning_rate": 1.5508588256846757e-06, "loss": 0.0077, "step": 73960 }, { "grad_norm": 0.18709337711334229, "learning_rate": 1.5457552298851319e-06, "loss": 0.0085, "step": 73970 }, { "grad_norm": 0.2074408382177353, "learning_rate": 1.5406599135907918e-06, "loss": 0.0087, "step": 73980 }, { "grad_norm": 0.09126140922307968, "learning_rate": 1.5355728776723088e-06, "loss": 0.0108, "step": 73990 }, { "grad_norm": 0.07651066035032272, "learning_rate": 1.5304941229989155e-06, "loss": 0.0069, "step": 74000 }, { "grad_norm": 0.1087174192070961, "learning_rate": 1.5254236504384345e-06, "loss": 0.0073, "step": 74010 }, { "grad_norm": 0.18799465894699097, "learning_rate": 1.5203614608572726e-06, "loss": 0.0131, "step": 74020 }, { "grad_norm": 0.12826892733573914, "learning_rate": 1.5153075551204044e-06, "loss": 0.0085, "step": 74030 }, { "grad_norm": 0.21396496891975403, "learning_rate": 1.5102619340914225e-06, "loss": 0.0202, "step": 74040 }, { "grad_norm": 0.14514204859733582, "learning_rate": 1.505224598632482e-06, "loss": 0.0069, "step": 74050 }, { "grad_norm": 0.22394350171089172, "learning_rate": 1.5001955496043164e-06, "loss": 0.0102, "step": 74060 }, { "grad_norm": 0.39629191160202026, "learning_rate": 1.4951747878662604e-06, "loss": 0.0082, "step": 74070 }, { "grad_norm": 0.3342936336994171, "learning_rate": 1.4901623142762221e-06, "loss": 0.0079, "step": 74080 }, { "grad_norm": 0.12054219841957092, "learning_rate": 1.4851581296907001e-06, "loss": 0.0104, "step": 74090 }, { "grad_norm": 0.17365871369838715, "learning_rate": 1.4801622349647714e-06, "loss": 0.0075, "step": 74100 }, { "grad_norm": 0.2669146656990051, "learning_rate": 1.4751746309520976e-06, "loss": 0.009, "step": 74110 }, { "grad_norm": 0.22106583416461945, "learning_rate": 1.4701953185049132e-06, "loss": 0.012, "step": 74120 }, { "grad_norm": 0.0862741470336914, "learning_rate": 1.4652242984740661e-06, "loss": 0.0099, "step": 74130 }, { "grad_norm": 0.28824132680892944, "learning_rate": 1.4602615717089484e-06, "loss": 0.0124, "step": 74140 }, { "grad_norm": 0.15493330359458923, "learning_rate": 1.4553071390575657e-06, "loss": 0.0067, "step": 74150 }, { "grad_norm": 0.18282300233840942, "learning_rate": 1.450361001366496e-06, "loss": 0.0134, "step": 74160 }, { "grad_norm": 0.12454210221767426, "learning_rate": 1.4454231594809021e-06, "loss": 0.0086, "step": 74170 }, { "grad_norm": 0.10936511307954788, "learning_rate": 1.4404936142445036e-06, "loss": 0.0099, "step": 74180 }, { "grad_norm": 0.13886231184005737, "learning_rate": 1.4355723664996546e-06, "loss": 0.0072, "step": 74190 }, { "grad_norm": 0.2833266854286194, "learning_rate": 1.4306594170872433e-06, "loss": 0.0058, "step": 74200 }, { "grad_norm": 0.2191293090581894, "learning_rate": 1.4257547668467598e-06, "loss": 0.0094, "step": 74210 }, { "grad_norm": 0.1968504637479782, "learning_rate": 1.4208584166162886e-06, "loss": 0.0105, "step": 74220 }, { "grad_norm": 0.19399721920490265, "learning_rate": 1.4159703672324553e-06, "loss": 0.0111, "step": 74230 }, { "grad_norm": 0.1970411092042923, "learning_rate": 1.4110906195305252e-06, "loss": 0.0087, "step": 74240 }, { "grad_norm": 0.5449459552764893, "learning_rate": 1.406219174344292e-06, "loss": 0.0079, "step": 74250 }, { "grad_norm": 0.14068590104579926, "learning_rate": 1.401356032506157e-06, "loss": 0.0089, "step": 74260 }, { "grad_norm": 0.25079160928726196, "learning_rate": 1.3965011948471051e-06, "loss": 0.0111, "step": 74270 }, { "grad_norm": 0.16644079983234406, "learning_rate": 1.3916546621966842e-06, "loss": 0.0087, "step": 74280 }, { "grad_norm": 0.14345434308052063, "learning_rate": 1.3868164353830482e-06, "loss": 0.006, "step": 74290 }, { "grad_norm": 0.1635107696056366, "learning_rate": 1.3819865152329082e-06, "loss": 0.0086, "step": 74300 }, { "grad_norm": 0.21094562113285065, "learning_rate": 1.377164902571565e-06, "loss": 0.0119, "step": 74310 }, { "grad_norm": 0.20439325273036957, "learning_rate": 1.372351598222904e-06, "loss": 0.0087, "step": 74320 }, { "grad_norm": 0.17098355293273926, "learning_rate": 1.3675466030093843e-06, "loss": 0.008, "step": 74330 }, { "grad_norm": 0.19544415175914764, "learning_rate": 1.3627499177520486e-06, "loss": 0.0147, "step": 74340 }, { "grad_norm": 0.14991264045238495, "learning_rate": 1.3579615432705196e-06, "loss": 0.0127, "step": 74350 }, { "grad_norm": 0.24375109374523163, "learning_rate": 1.3531814803829978e-06, "loss": 0.0137, "step": 74360 }, { "grad_norm": 0.16767555475234985, "learning_rate": 1.3484097299062803e-06, "loss": 0.0097, "step": 74370 }, { "grad_norm": 0.26357772946357727, "learning_rate": 1.343646292655698e-06, "loss": 0.0091, "step": 74380 }, { "grad_norm": 0.21641191840171814, "learning_rate": 1.3388911694452277e-06, "loss": 0.0101, "step": 74390 }, { "grad_norm": 0.40863174200057983, "learning_rate": 1.3341443610873638e-06, "loss": 0.0165, "step": 74400 }, { "grad_norm": 0.20753422379493713, "learning_rate": 1.329405868393213e-06, "loss": 0.0104, "step": 74410 }, { "grad_norm": 0.1603255271911621, "learning_rate": 1.3246756921724613e-06, "loss": 0.0153, "step": 74420 }, { "grad_norm": 0.20788244903087616, "learning_rate": 1.3199538332333506e-06, "loss": 0.0069, "step": 74430 }, { "grad_norm": 0.15584321320056915, "learning_rate": 1.3152402923827411e-06, "loss": 0.0126, "step": 74440 }, { "grad_norm": 0.24713614583015442, "learning_rate": 1.3105350704260277e-06, "loss": 0.0092, "step": 74450 }, { "grad_norm": 0.24583734571933746, "learning_rate": 1.305838168167206e-06, "loss": 0.0094, "step": 74460 }, { "grad_norm": 0.16309356689453125, "learning_rate": 1.301149586408862e-06, "loss": 0.0067, "step": 74470 }, { "grad_norm": 0.12957130372524261, "learning_rate": 1.2964693259521322e-06, "loss": 0.0104, "step": 74480 }, { "grad_norm": 0.2342265099287033, "learning_rate": 1.2917973875967548e-06, "loss": 0.0073, "step": 74490 }, { "grad_norm": 0.505027174949646, "learning_rate": 1.2871337721410249e-06, "loss": 0.0098, "step": 74500 }, { "grad_norm": 0.16249391436576843, "learning_rate": 1.2824784803818379e-06, "loss": 0.0103, "step": 74510 }, { "grad_norm": 0.11855921149253845, "learning_rate": 1.2778315131146467e-06, "loss": 0.0076, "step": 74520 }, { "grad_norm": 0.32055965065956116, "learning_rate": 1.2731928711334994e-06, "loss": 0.0087, "step": 74530 }, { "grad_norm": 0.2882468104362488, "learning_rate": 1.2685625552310122e-06, "loss": 0.01, "step": 74540 }, { "grad_norm": 0.10666143894195557, "learning_rate": 1.263940566198374e-06, "loss": 0.0075, "step": 74550 }, { "grad_norm": 0.11817145347595215, "learning_rate": 1.259326904825353e-06, "loss": 0.0065, "step": 74560 }, { "grad_norm": 0.25359028577804565, "learning_rate": 1.2547215719003137e-06, "loss": 0.0085, "step": 74570 }, { "grad_norm": 0.27904126048088074, "learning_rate": 1.2501245682101703e-06, "loss": 0.0081, "step": 74580 }, { "grad_norm": 0.13769520819187164, "learning_rate": 1.2455358945404171e-06, "loss": 0.0102, "step": 74590 }, { "grad_norm": 0.1968740075826645, "learning_rate": 1.2409555516751493e-06, "loss": 0.0082, "step": 74600 }, { "grad_norm": 0.19791825115680695, "learning_rate": 1.2363835403970125e-06, "loss": 0.0073, "step": 74610 }, { "grad_norm": 0.22435617446899414, "learning_rate": 1.2318198614872489e-06, "loss": 0.0081, "step": 74620 }, { "grad_norm": 0.23003409802913666, "learning_rate": 1.2272645157256457e-06, "loss": 0.0072, "step": 74630 }, { "grad_norm": 0.16574880480766296, "learning_rate": 1.222717503890608e-06, "loss": 0.0125, "step": 74640 }, { "grad_norm": 0.3810429275035858, "learning_rate": 1.218178826759081e-06, "loss": 0.0109, "step": 74650 }, { "grad_norm": 0.3678455054759979, "learning_rate": 1.213648485106611e-06, "loss": 0.0077, "step": 74660 }, { "grad_norm": 0.11155669391155243, "learning_rate": 1.2091264797073066e-06, "loss": 0.0113, "step": 74670 }, { "grad_norm": 0.15581198036670685, "learning_rate": 1.2046128113338494e-06, "loss": 0.0213, "step": 74680 }, { "grad_norm": 0.2279512882232666, "learning_rate": 1.200107480757512e-06, "loss": 0.0117, "step": 74690 }, { "grad_norm": 0.20198477804660797, "learning_rate": 1.1956104887481168e-06, "loss": 0.0133, "step": 74700 }, { "grad_norm": 0.1707574874162674, "learning_rate": 1.191121836074094e-06, "loss": 0.0111, "step": 74710 }, { "grad_norm": 0.26742374897003174, "learning_rate": 1.1866415235024186e-06, "loss": 0.0091, "step": 74720 }, { "grad_norm": 0.2759973108768463, "learning_rate": 1.182169551798662e-06, "loss": 0.011, "step": 74730 }, { "grad_norm": 0.18117278814315796, "learning_rate": 1.177705921726957e-06, "loss": 0.0085, "step": 74740 }, { "grad_norm": 0.14783959090709686, "learning_rate": 1.173250634050016e-06, "loss": 0.0098, "step": 74750 }, { "grad_norm": 0.11213632673025131, "learning_rate": 1.1688036895291354e-06, "loss": 0.007, "step": 74760 }, { "grad_norm": 0.18834149837493896, "learning_rate": 1.1643650889241574e-06, "loss": 0.0072, "step": 74770 }, { "grad_norm": 0.15880277752876282, "learning_rate": 1.1599348329935311e-06, "loss": 0.0106, "step": 74780 }, { "grad_norm": 0.25344589352607727, "learning_rate": 1.1555129224942673e-06, "loss": 0.0124, "step": 74790 }, { "grad_norm": 0.03496390953660011, "learning_rate": 1.1510993581819396e-06, "loss": 0.0112, "step": 74800 }, { "grad_norm": 0.11094725131988525, "learning_rate": 1.1466941408107112e-06, "loss": 0.0111, "step": 74810 }, { "grad_norm": 0.21553544700145721, "learning_rate": 1.1422972711333247e-06, "loss": 0.0082, "step": 74820 }, { "grad_norm": 0.23471155762672424, "learning_rate": 1.1379087499010565e-06, "loss": 0.0077, "step": 74830 }, { "grad_norm": 0.1406334936618805, "learning_rate": 1.1335285778638127e-06, "loss": 0.0055, "step": 74840 }, { "grad_norm": 0.1645030379295349, "learning_rate": 1.1291567557700333e-06, "loss": 0.0089, "step": 74850 }, { "grad_norm": 0.2762071490287781, "learning_rate": 1.124793284366743e-06, "loss": 0.0095, "step": 74860 }, { "grad_norm": 0.10250234603881836, "learning_rate": 1.1204381643995455e-06, "loss": 0.0066, "step": 74870 }, { "grad_norm": 0.16047672927379608, "learning_rate": 1.116091396612595e-06, "loss": 0.0082, "step": 74880 }, { "grad_norm": 0.12474384903907776, "learning_rate": 1.1117529817486538e-06, "loss": 0.0088, "step": 74890 }, { "grad_norm": 0.16951295733451843, "learning_rate": 1.107422920549034e-06, "loss": 0.0068, "step": 74900 }, { "grad_norm": 0.24673840403556824, "learning_rate": 1.1031012137536157e-06, "loss": 0.0082, "step": 74910 }, { "grad_norm": 0.24203646183013916, "learning_rate": 1.0987878621008695e-06, "loss": 0.0092, "step": 74920 }, { "grad_norm": 0.22729413211345673, "learning_rate": 1.0944828663278284e-06, "loss": 0.0124, "step": 74930 }, { "grad_norm": 0.11335545033216476, "learning_rate": 1.0901862271700925e-06, "loss": 0.0117, "step": 74940 }, { "grad_norm": 0.09868872165679932, "learning_rate": 1.0858979453618467e-06, "loss": 0.0124, "step": 74950 }, { "grad_norm": 0.18113018572330475, "learning_rate": 1.0816180216358441e-06, "loss": 0.0099, "step": 74960 }, { "grad_norm": 0.19841083884239197, "learning_rate": 1.0773464567233937e-06, "loss": 0.0068, "step": 74970 }, { "grad_norm": 0.1671452671289444, "learning_rate": 1.0730832513543953e-06, "loss": 0.0082, "step": 74980 }, { "grad_norm": 0.29889634251594543, "learning_rate": 1.0688284062573218e-06, "loss": 0.01, "step": 74990 }, { "grad_norm": 0.41557785868644714, "learning_rate": 1.0645819221591969e-06, "loss": 0.0101, "step": 75000 }, { "grad_norm": 0.22718031704425812, "learning_rate": 1.0603437997856347e-06, "loss": 0.0091, "step": 75010 }, { "grad_norm": 0.2148815244436264, "learning_rate": 1.0561140398608228e-06, "loss": 0.0073, "step": 75020 }, { "grad_norm": 0.1123715341091156, "learning_rate": 1.0518926431074937e-06, "loss": 0.0079, "step": 75030 }, { "grad_norm": 0.10831102728843689, "learning_rate": 1.0476796102469877e-06, "loss": 0.0103, "step": 75040 }, { "grad_norm": 0.22328025102615356, "learning_rate": 1.0434749419991786e-06, "loss": 0.0098, "step": 75050 }, { "grad_norm": 0.10801015794277191, "learning_rate": 1.0392786390825415e-06, "loss": 0.0081, "step": 75060 }, { "grad_norm": 0.2890065312385559, "learning_rate": 1.0350907022141087e-06, "loss": 0.0093, "step": 75070 }, { "grad_norm": 0.21693886816501617, "learning_rate": 1.0309111321094744e-06, "loss": 0.007, "step": 75080 }, { "grad_norm": 0.16867336630821228, "learning_rate": 1.026739929482834e-06, "loss": 0.0066, "step": 75090 }, { "grad_norm": 0.16857950389385223, "learning_rate": 1.022577095046906e-06, "loss": 0.0082, "step": 75100 }, { "grad_norm": 0.0912846252322197, "learning_rate": 1.0184226295130217e-06, "loss": 0.0079, "step": 75110 }, { "grad_norm": 0.3430863916873932, "learning_rate": 1.0142765335910575e-06, "loss": 0.0107, "step": 75120 }, { "grad_norm": 0.28578418493270874, "learning_rate": 1.0101388079894746e-06, "loss": 0.0075, "step": 75130 }, { "grad_norm": 0.10652752220630646, "learning_rate": 1.0060094534152908e-06, "loss": 0.0107, "step": 75140 }, { "grad_norm": 0.125647634267807, "learning_rate": 1.0018884705741028e-06, "loss": 0.0079, "step": 75150 }, { "grad_norm": 0.31013765931129456, "learning_rate": 9.977758601700805e-07, "loss": 0.0116, "step": 75160 }, { "grad_norm": 0.07737602293491364, "learning_rate": 9.93671622905945e-07, "loss": 0.0137, "step": 75170 }, { "grad_norm": 0.1460110992193222, "learning_rate": 9.895757594830024e-07, "loss": 0.0117, "step": 75180 }, { "grad_norm": 0.16249901056289673, "learning_rate": 9.854882706011204e-07, "loss": 0.0077, "step": 75190 }, { "grad_norm": 0.21622128784656525, "learning_rate": 9.814091569587513e-07, "loss": 0.0084, "step": 75200 }, { "grad_norm": 0.2367006242275238, "learning_rate": 9.77338419252888e-07, "loss": 0.0135, "step": 75210 }, { "grad_norm": 0.11773397028446198, "learning_rate": 9.732760581791289e-07, "loss": 0.009, "step": 75220 }, { "grad_norm": 0.11830273270606995, "learning_rate": 9.692220744315972e-07, "loss": 0.0072, "step": 75230 }, { "grad_norm": 0.28486159443855286, "learning_rate": 9.651764687030162e-07, "loss": 0.0077, "step": 75240 }, { "grad_norm": 0.2054695338010788, "learning_rate": 9.611392416846776e-07, "loss": 0.0067, "step": 75250 }, { "grad_norm": 0.17280501127243042, "learning_rate": 9.571103940664238e-07, "loss": 0.0084, "step": 75260 }, { "grad_norm": 0.1051211878657341, "learning_rate": 9.530899265366822e-07, "loss": 0.0082, "step": 75270 }, { "grad_norm": 0.08934290707111359, "learning_rate": 9.490778397824307e-07, "loss": 0.0102, "step": 75280 }, { "grad_norm": 0.2386966049671173, "learning_rate": 9.450741344892322e-07, "loss": 0.0092, "step": 75290 }, { "grad_norm": 0.35715749859809875, "learning_rate": 9.410788113412116e-07, "loss": 0.0111, "step": 75300 }, { "grad_norm": 0.15974539518356323, "learning_rate": 9.370918710210563e-07, "loss": 0.0067, "step": 75310 }, { "grad_norm": 0.17154613137245178, "learning_rate": 9.331133142100268e-07, "loss": 0.0107, "step": 75320 }, { "grad_norm": 0.42735347151756287, "learning_rate": 9.291431415879459e-07, "loss": 0.0073, "step": 75330 }, { "grad_norm": 0.29683271050453186, "learning_rate": 9.251813538332155e-07, "loss": 0.0094, "step": 75340 }, { "grad_norm": 0.1065090075135231, "learning_rate": 9.212279516227884e-07, "loss": 0.0084, "step": 75350 }, { "grad_norm": 0.17715126276016235, "learning_rate": 9.172829356322021e-07, "loss": 0.0057, "step": 75360 }, { "grad_norm": 0.17545437812805176, "learning_rate": 9.133463065355452e-07, "loss": 0.0097, "step": 75370 }, { "grad_norm": 0.4298872947692871, "learning_rate": 9.094180650054796e-07, "loss": 0.0124, "step": 75380 }, { "grad_norm": 0.20020367205142975, "learning_rate": 9.054982117132404e-07, "loss": 0.0082, "step": 75390 }, { "grad_norm": 0.2844927906990051, "learning_rate": 9.015867473286143e-07, "loss": 0.0097, "step": 75400 }, { "grad_norm": 0.18442785739898682, "learning_rate": 8.976836725199778e-07, "loss": 0.0062, "step": 75410 }, { "grad_norm": 0.22148773074150085, "learning_rate": 8.937889879542416e-07, "loss": 0.0066, "step": 75420 }, { "grad_norm": 0.11626888811588287, "learning_rate": 8.899026942969068e-07, "loss": 0.0098, "step": 75430 }, { "grad_norm": 0.16424109041690826, "learning_rate": 8.860247922120424e-07, "loss": 0.0084, "step": 75440 }, { "grad_norm": 0.362107515335083, "learning_rate": 8.821552823622737e-07, "loss": 0.0082, "step": 75450 }, { "grad_norm": 0.14915098249912262, "learning_rate": 8.782941654087884e-07, "loss": 0.0078, "step": 75460 }, { "grad_norm": 0.31771308183670044, "learning_rate": 8.744414420113478e-07, "loss": 0.013, "step": 75470 }, { "grad_norm": 0.0966235026717186, "learning_rate": 8.705971128282753e-07, "loss": 0.0078, "step": 75480 }, { "grad_norm": 0.1717190146446228, "learning_rate": 8.66761178516473e-07, "loss": 0.0068, "step": 75490 }, { "grad_norm": 0.23838229477405548, "learning_rate": 8.629336397313781e-07, "loss": 0.0112, "step": 75500 }, { "grad_norm": 0.12169236689805984, "learning_rate": 8.591144971270282e-07, "loss": 0.0117, "step": 75510 }, { "grad_norm": 0.1383160799741745, "learning_rate": 8.553037513560069e-07, "loss": 0.0118, "step": 75520 }, { "grad_norm": 0.20621152222156525, "learning_rate": 8.515014030694546e-07, "loss": 0.0135, "step": 75530 }, { "grad_norm": 0.20126934349536896, "learning_rate": 8.47707452917107e-07, "loss": 0.0086, "step": 75540 }, { "grad_norm": 0.1682918518781662, "learning_rate": 8.439219015472343e-07, "loss": 0.0093, "step": 75550 }, { "grad_norm": 0.25882619619369507, "learning_rate": 8.401447496066861e-07, "loss": 0.0082, "step": 75560 }, { "grad_norm": 0.2549740672111511, "learning_rate": 8.363759977408792e-07, "loss": 0.0101, "step": 75570 }, { "grad_norm": 0.1959257423877716, "learning_rate": 8.326156465937817e-07, "loss": 0.0083, "step": 75580 }, { "grad_norm": 0.48144233226776123, "learning_rate": 8.28863696807941e-07, "loss": 0.0157, "step": 75590 }, { "grad_norm": 0.2580348551273346, "learning_rate": 8.251201490244609e-07, "loss": 0.0084, "step": 75600 }, { "grad_norm": 0.16699041426181793, "learning_rate": 8.213850038830129e-07, "loss": 0.0076, "step": 75610 }, { "grad_norm": 0.3107648193836212, "learning_rate": 8.176582620218309e-07, "loss": 0.0077, "step": 75620 }, { "grad_norm": 0.042816393077373505, "learning_rate": 8.139399240777057e-07, "loss": 0.0074, "step": 75630 }, { "grad_norm": 0.19419418275356293, "learning_rate": 8.102299906860122e-07, "loss": 0.0088, "step": 75640 }, { "grad_norm": 0.5083727836608887, "learning_rate": 8.065284624806657e-07, "loss": 0.0119, "step": 75650 }, { "grad_norm": 0.23081402480602264, "learning_rate": 8.028353400941601e-07, "loss": 0.0092, "step": 75660 }, { "grad_norm": 0.560276448726654, "learning_rate": 7.991506241575575e-07, "loss": 0.01, "step": 75670 }, { "grad_norm": 0.34614747762680054, "learning_rate": 7.954743153004541e-07, "loss": 0.0086, "step": 75680 }, { "grad_norm": 0.1694803386926651, "learning_rate": 7.918064141510528e-07, "loss": 0.0091, "step": 75690 }, { "grad_norm": 0.15261714160442352, "learning_rate": 7.881469213360859e-07, "loss": 0.0097, "step": 75700 }, { "grad_norm": 0.2738291621208191, "learning_rate": 7.844958374808642e-07, "loss": 0.0069, "step": 75710 }, { "grad_norm": 0.10840390622615814, "learning_rate": 7.808531632092608e-07, "loss": 0.0118, "step": 75720 }, { "grad_norm": 0.11255265027284622, "learning_rate": 7.772188991436946e-07, "loss": 0.0085, "step": 75730 }, { "grad_norm": 0.2505747377872467, "learning_rate": 7.735930459051799e-07, "loss": 0.0129, "step": 75740 }, { "grad_norm": 0.12296963483095169, "learning_rate": 7.699756041132655e-07, "loss": 0.0086, "step": 75750 }, { "grad_norm": 0.1456863284111023, "learning_rate": 7.663665743860793e-07, "loss": 0.0057, "step": 75760 }, { "grad_norm": 0.11249268054962158, "learning_rate": 7.627659573403001e-07, "loss": 0.0102, "step": 75770 }, { "grad_norm": 0.12422353774309158, "learning_rate": 7.591737535911802e-07, "loss": 0.0066, "step": 75780 }, { "grad_norm": 0.28563952445983887, "learning_rate": 7.555899637525288e-07, "loss": 0.0118, "step": 75790 }, { "grad_norm": 0.18082472681999207, "learning_rate": 7.520145884367058e-07, "loss": 0.0083, "step": 75800 }, { "grad_norm": 0.2143254429101944, "learning_rate": 7.484476282546615e-07, "loss": 0.0064, "step": 75810 }, { "grad_norm": 0.25355401635169983, "learning_rate": 7.448890838158806e-07, "loss": 0.0111, "step": 75820 }, { "grad_norm": 0.10346648842096329, "learning_rate": 7.413389557284267e-07, "loss": 0.0078, "step": 75830 }, { "grad_norm": 0.09459710121154785, "learning_rate": 7.377972445989201e-07, "loss": 0.0084, "step": 75840 }, { "grad_norm": 0.23735250532627106, "learning_rate": 7.342639510325377e-07, "loss": 0.0111, "step": 75850 }, { "grad_norm": 0.22023731470108032, "learning_rate": 7.307390756330246e-07, "loss": 0.0079, "step": 75860 }, { "grad_norm": 0.7841372489929199, "learning_rate": 7.272226190026876e-07, "loss": 0.0086, "step": 75870 }, { "grad_norm": 0.1486542820930481, "learning_rate": 7.237145817423907e-07, "loss": 0.0074, "step": 75880 }, { "grad_norm": 0.13404083251953125, "learning_rate": 7.202149644515654e-07, "loss": 0.0072, "step": 75890 }, { "grad_norm": 0.19231484830379486, "learning_rate": 7.167237677281946e-07, "loss": 0.0103, "step": 75900 }, { "grad_norm": 0.10244473814964294, "learning_rate": 7.132409921688288e-07, "loss": 0.0099, "step": 75910 }, { "grad_norm": 0.3961615264415741, "learning_rate": 7.09766638368592e-07, "loss": 0.0088, "step": 75920 }, { "grad_norm": 0.15834484994411469, "learning_rate": 7.063007069211313e-07, "loss": 0.0059, "step": 75930 }, { "grad_norm": 0.15323592722415924, "learning_rate": 7.028431984187067e-07, "loss": 0.007, "step": 75940 }, { "grad_norm": 0.16880905628204346, "learning_rate": 6.993941134520898e-07, "loss": 0.0068, "step": 75950 }, { "grad_norm": 0.13088871538639069, "learning_rate": 6.959534526106537e-07, "loss": 0.0104, "step": 75960 }, { "grad_norm": 0.16081641614437103, "learning_rate": 6.925212164822948e-07, "loss": 0.007, "step": 75970 }, { "grad_norm": 0.23865623772144318, "learning_rate": 6.890974056535049e-07, "loss": 0.0088, "step": 75980 }, { "grad_norm": 0.09565559774637222, "learning_rate": 6.856820207093106e-07, "loss": 0.009, "step": 75990 }, { "grad_norm": 0.1242758110165596, "learning_rate": 6.822750622333063e-07, "loss": 0.0055, "step": 76000 }, { "grad_norm": 0.1849711835384369, "learning_rate": 6.788765308076539e-07, "loss": 0.0088, "step": 76010 }, { "grad_norm": 0.19172832369804382, "learning_rate": 6.754864270130668e-07, "loss": 0.0082, "step": 76020 }, { "grad_norm": 0.09478209912776947, "learning_rate": 6.721047514288203e-07, "loss": 0.0082, "step": 76030 }, { "grad_norm": 0.3697066307067871, "learning_rate": 6.687315046327469e-07, "loss": 0.0086, "step": 76040 }, { "grad_norm": 0.24872072041034698, "learning_rate": 6.65366687201252e-07, "loss": 0.0096, "step": 76050 }, { "grad_norm": 0.4176446199417114, "learning_rate": 6.620102997092814e-07, "loss": 0.0115, "step": 76060 }, { "grad_norm": 0.2931411564350128, "learning_rate": 6.586623427303596e-07, "loss": 0.0122, "step": 76070 }, { "grad_norm": 0.13431596755981445, "learning_rate": 6.553228168365455e-07, "loss": 0.0088, "step": 76080 }, { "grad_norm": 0.3041187524795532, "learning_rate": 6.519917225984884e-07, "loss": 0.0133, "step": 76090 }, { "grad_norm": 0.2721696197986603, "learning_rate": 6.486690605853718e-07, "loss": 0.0065, "step": 76100 }, { "grad_norm": 0.23791727423667908, "learning_rate": 6.453548313649527e-07, "loss": 0.0096, "step": 76110 }, { "grad_norm": 0.26641610264778137, "learning_rate": 6.420490355035446e-07, "loss": 0.0082, "step": 76120 }, { "grad_norm": 0.15774066746234894, "learning_rate": 6.387516735660071e-07, "loss": 0.0086, "step": 76130 }, { "grad_norm": 0.16018548607826233, "learning_rate": 6.35462746115778e-07, "loss": 0.0088, "step": 76140 }, { "grad_norm": 0.3363765776157379, "learning_rate": 6.321822537148358e-07, "loss": 0.0069, "step": 76150 }, { "grad_norm": 0.06911013275384903, "learning_rate": 6.289101969237432e-07, "loss": 0.0064, "step": 76160 }, { "grad_norm": 0.22499707341194153, "learning_rate": 6.256465763015918e-07, "loss": 0.011, "step": 76170 }, { "grad_norm": 0.19795750081539154, "learning_rate": 6.22391392406052e-07, "loss": 0.0094, "step": 76180 }, { "grad_norm": 0.24312926828861237, "learning_rate": 6.191446457933403e-07, "loss": 0.0096, "step": 76190 }, { "grad_norm": 0.16807231307029724, "learning_rate": 6.159063370182406e-07, "loss": 0.0077, "step": 76200 }, { "grad_norm": 0.18478652834892273, "learning_rate": 6.126764666340879e-07, "loss": 0.0072, "step": 76210 }, { "grad_norm": 0.12964032590389252, "learning_rate": 6.094550351927852e-07, "loss": 0.0113, "step": 76220 }, { "grad_norm": 0.23660457134246826, "learning_rate": 6.062420432447757e-07, "loss": 0.0087, "step": 76230 }, { "grad_norm": 0.13546879589557648, "learning_rate": 6.030374913390813e-07, "loss": 0.008, "step": 76240 }, { "grad_norm": 0.2680642008781433, "learning_rate": 5.998413800232694e-07, "loss": 0.0083, "step": 76250 }, { "grad_norm": 0.07199003547430038, "learning_rate": 5.966537098434755e-07, "loss": 0.0086, "step": 76260 }, { "grad_norm": 0.2799842059612274, "learning_rate": 5.934744813443694e-07, "loss": 0.0086, "step": 76270 }, { "grad_norm": 0.12479337304830551, "learning_rate": 5.903036950692054e-07, "loss": 0.0115, "step": 76280 }, { "grad_norm": 0.11687369644641876, "learning_rate": 5.871413515597835e-07, "loss": 0.0084, "step": 76290 }, { "grad_norm": 0.2264859676361084, "learning_rate": 5.839874513564547e-07, "loss": 0.0075, "step": 76300 }, { "grad_norm": 0.14074037969112396, "learning_rate": 5.808419949981436e-07, "loss": 0.0098, "step": 76310 }, { "grad_norm": 0.07570463418960571, "learning_rate": 5.777049830223257e-07, "loss": 0.0072, "step": 76320 }, { "grad_norm": 0.16219530999660492, "learning_rate": 5.745764159650114e-07, "loss": 0.0054, "step": 76330 }, { "grad_norm": 0.13262632489204407, "learning_rate": 5.71456294360806e-07, "loss": 0.0073, "step": 76340 }, { "grad_norm": 0.2807037830352783, "learning_rate": 5.683446187428443e-07, "loss": 0.0117, "step": 76350 }, { "grad_norm": 0.1667332798242569, "learning_rate": 5.652413896428288e-07, "loss": 0.013, "step": 76360 }, { "grad_norm": 0.4623037278652191, "learning_rate": 5.621466075910131e-07, "loss": 0.0102, "step": 76370 }, { "grad_norm": 0.16256269812583923, "learning_rate": 5.590602731162187e-07, "loss": 0.0093, "step": 76380 }, { "grad_norm": 0.2043118178844452, "learning_rate": 5.559823867458125e-07, "loss": 0.0077, "step": 76390 }, { "grad_norm": 0.08252324163913727, "learning_rate": 5.529129490057128e-07, "loss": 0.0123, "step": 76400 }, { "grad_norm": 0.11558609455823898, "learning_rate": 5.498519604204167e-07, "loss": 0.0078, "step": 76410 }, { "grad_norm": 0.13434168696403503, "learning_rate": 5.467994215129501e-07, "loss": 0.0083, "step": 76420 }, { "grad_norm": 0.1036800891160965, "learning_rate": 5.437553328049183e-07, "loss": 0.0085, "step": 76430 }, { "grad_norm": 0.1970432549715042, "learning_rate": 5.407196948164661e-07, "loss": 0.0068, "step": 76440 }, { "grad_norm": 0.12079230695962906, "learning_rate": 5.376925080663009e-07, "loss": 0.0095, "step": 76450 }, { "grad_norm": 0.14629697799682617, "learning_rate": 5.346737730716977e-07, "loss": 0.0089, "step": 76460 }, { "grad_norm": 0.1316697895526886, "learning_rate": 5.316634903484607e-07, "loss": 0.007, "step": 76470 }, { "grad_norm": 0.25576719641685486, "learning_rate": 5.286616604109728e-07, "loss": 0.0095, "step": 76480 }, { "grad_norm": 0.04199203848838806, "learning_rate": 5.256682837721627e-07, "loss": 0.0113, "step": 76490 }, { "grad_norm": 0.09193917363882065, "learning_rate": 5.226833609435156e-07, "loss": 0.0063, "step": 76500 }, { "grad_norm": 0.19585011899471283, "learning_rate": 5.19706892435079e-07, "loss": 0.0094, "step": 76510 }, { "grad_norm": 0.09636647254228592, "learning_rate": 5.167388787554406e-07, "loss": 0.0103, "step": 76520 }, { "grad_norm": 0.24592836201190948, "learning_rate": 5.137793204117614e-07, "loss": 0.0121, "step": 76530 }, { "grad_norm": 0.31706157326698303, "learning_rate": 5.108282179097479e-07, "loss": 0.0095, "step": 76540 }, { "grad_norm": 0.15219977498054504, "learning_rate": 5.078855717536523e-07, "loss": 0.0104, "step": 76550 }, { "grad_norm": 0.10867314785718918, "learning_rate": 5.049513824463059e-07, "loss": 0.0067, "step": 76560 }, { "grad_norm": 0.15020273625850677, "learning_rate": 5.020256504890742e-07, "loss": 0.0072, "step": 76570 }, { "grad_norm": 0.11311034858226776, "learning_rate": 4.991083763818849e-07, "loss": 0.0075, "step": 76580 }, { "grad_norm": 0.2395237237215042, "learning_rate": 4.961995606232228e-07, "loss": 0.0111, "step": 76590 }, { "grad_norm": 0.19969668984413147, "learning_rate": 4.932992037101236e-07, "loss": 0.0099, "step": 76600 }, { "grad_norm": 0.30029425024986267, "learning_rate": 4.904073061381798e-07, "loss": 0.0115, "step": 76610 }, { "grad_norm": 0.19510379433631897, "learning_rate": 4.875238684015349e-07, "loss": 0.0071, "step": 76620 }, { "grad_norm": 0.3056548237800598, "learning_rate": 4.846488909928948e-07, "loss": 0.0087, "step": 76630 }, { "grad_norm": 0.10266422480344772, "learning_rate": 4.81782374403511e-07, "loss": 0.0096, "step": 76640 }, { "grad_norm": 0.18560095131397247, "learning_rate": 4.789243191231918e-07, "loss": 0.0068, "step": 76650 }, { "grad_norm": 0.20371849834918976, "learning_rate": 4.76074725640302e-07, "loss": 0.0128, "step": 76660 }, { "grad_norm": 0.11830482631921768, "learning_rate": 4.732335944417632e-07, "loss": 0.0142, "step": 76670 }, { "grad_norm": 0.2019728571176529, "learning_rate": 4.704009260130371e-07, "loss": 0.0117, "step": 76680 }, { "grad_norm": 0.5605457425117493, "learning_rate": 4.675767208381587e-07, "loss": 0.0119, "step": 76690 }, { "grad_norm": 0.643012523651123, "learning_rate": 4.647609793997032e-07, "loss": 0.0106, "step": 76700 }, { "grad_norm": 0.21944214403629303, "learning_rate": 4.619537021788023e-07, "loss": 0.0108, "step": 76710 }, { "grad_norm": 0.2490500658750534, "learning_rate": 4.5915488965515005e-07, "loss": 0.0085, "step": 76720 }, { "grad_norm": 0.2103683203458786, "learning_rate": 4.563645423069807e-07, "loss": 0.0104, "step": 76730 }, { "grad_norm": 0.27482369542121887, "learning_rate": 4.535826606110849e-07, "loss": 0.0077, "step": 76740 }, { "grad_norm": 0.2601459324359894, "learning_rate": 4.508092450428214e-07, "loss": 0.0081, "step": 76750 }, { "grad_norm": 0.39192602038383484, "learning_rate": 4.480442960760778e-07, "loss": 0.0122, "step": 76760 }, { "grad_norm": 0.16216367483139038, "learning_rate": 4.4528781418332053e-07, "loss": 0.0082, "step": 76770 }, { "grad_norm": 0.46608996391296387, "learning_rate": 4.4253979983555073e-07, "loss": 0.0139, "step": 76780 }, { "grad_norm": 0.30382630228996277, "learning_rate": 4.3980025350233154e-07, "loss": 0.0082, "step": 76790 }, { "grad_norm": 0.20981623232364655, "learning_rate": 4.370691756517664e-07, "loss": 0.0118, "step": 76800 }, { "grad_norm": 0.2971145808696747, "learning_rate": 4.3434656675053753e-07, "loss": 0.0118, "step": 76810 }, { "grad_norm": 0.22201450169086456, "learning_rate": 4.3163242726385613e-07, "loss": 0.0088, "step": 76820 }, { "grad_norm": 0.16286620497703552, "learning_rate": 4.289267576554956e-07, "loss": 0.0071, "step": 76830 }, { "grad_norm": 0.24885506927967072, "learning_rate": 4.262295583877807e-07, "loss": 0.0063, "step": 76840 }, { "grad_norm": 0.702744722366333, "learning_rate": 4.235408299215815e-07, "loss": 0.0106, "step": 76850 }, { "grad_norm": 0.1844603568315506, "learning_rate": 4.2086057271634173e-07, "loss": 0.0078, "step": 76860 }, { "grad_norm": 0.24305526912212372, "learning_rate": 4.181887872300394e-07, "loss": 0.009, "step": 76870 }, { "grad_norm": 0.16596969962120056, "learning_rate": 4.155254739192038e-07, "loss": 0.01, "step": 76880 }, { "grad_norm": 0.34539031982421875, "learning_rate": 4.128706332389265e-07, "loss": 0.0111, "step": 76890 }, { "grad_norm": 0.32405799627304077, "learning_rate": 4.1022426564284453e-07, "loss": 0.007, "step": 76900 }, { "grad_norm": 0.13772235810756683, "learning_rate": 4.075863715831574e-07, "loss": 0.0096, "step": 76910 }, { "grad_norm": 0.17618818581104279, "learning_rate": 4.0495695151059887e-07, "loss": 0.0093, "step": 76920 }, { "grad_norm": 0.14956222474575043, "learning_rate": 4.023360058744763e-07, "loss": 0.0075, "step": 76930 }, { "grad_norm": 0.15039893984794617, "learning_rate": 3.997235351226314e-07, "loss": 0.0078, "step": 76940 }, { "grad_norm": 0.15220415592193604, "learning_rate": 3.971195397014571e-07, "loss": 0.0089, "step": 76950 }, { "grad_norm": 0.17922784388065338, "learning_rate": 3.9452402005591417e-07, "loss": 0.011, "step": 76960 }, { "grad_norm": 0.12254934012889862, "learning_rate": 3.9193697662950334e-07, "loss": 0.0065, "step": 76970 }, { "grad_norm": 0.4867098331451416, "learning_rate": 3.893584098642822e-07, "loss": 0.0135, "step": 76980 }, { "grad_norm": 0.10891687124967575, "learning_rate": 3.867883202008538e-07, "loss": 0.0084, "step": 76990 }, { "grad_norm": 0.15653949975967407, "learning_rate": 3.842267080783779e-07, "loss": 0.0094, "step": 77000 }, { "grad_norm": 0.18491747975349426, "learning_rate": 3.816735739345656e-07, "loss": 0.0112, "step": 77010 }, { "grad_norm": 0.1910315603017807, "learning_rate": 3.791289182056679e-07, "loss": 0.0098, "step": 77020 }, { "grad_norm": 0.21474790573120117, "learning_rate": 3.7659274132650913e-07, "loss": 0.0099, "step": 77030 }, { "grad_norm": 0.059691060334444046, "learning_rate": 3.7406504373044826e-07, "loss": 0.0071, "step": 77040 }, { "grad_norm": 0.16667065024375916, "learning_rate": 3.7154582584939533e-07, "loss": 0.0084, "step": 77050 }, { "grad_norm": 0.1792726218700409, "learning_rate": 3.690350881138227e-07, "loss": 0.0073, "step": 77060 }, { "grad_norm": 0.12458913028240204, "learning_rate": 3.665328309527427e-07, "loss": 0.0066, "step": 77070 }, { "grad_norm": 0.16533152759075165, "learning_rate": 3.640390547937245e-07, "loss": 0.0086, "step": 77080 }, { "grad_norm": 0.3931215703487396, "learning_rate": 3.615537600628882e-07, "loss": 0.0088, "step": 77090 }, { "grad_norm": 0.14232856035232544, "learning_rate": 3.5907694718489427e-07, "loss": 0.0106, "step": 77100 }, { "grad_norm": 0.1539764106273651, "learning_rate": 3.566086165829707e-07, "loss": 0.0131, "step": 77110 }, { "grad_norm": 0.7320528626441956, "learning_rate": 3.5414876867888024e-07, "loss": 0.0089, "step": 77120 }, { "grad_norm": 0.3044643700122833, "learning_rate": 3.5169740389295326e-07, "loss": 0.0118, "step": 77130 }, { "grad_norm": 0.15105420351028442, "learning_rate": 3.4925452264405466e-07, "loss": 0.0072, "step": 77140 }, { "grad_norm": 0.2725377678871155, "learning_rate": 3.468201253496062e-07, "loss": 0.0071, "step": 77150 }, { "grad_norm": 0.27044451236724854, "learning_rate": 3.443942124255861e-07, "loss": 0.0087, "step": 77160 }, { "grad_norm": 0.1888153851032257, "learning_rate": 3.4197678428650183e-07, "loss": 0.0094, "step": 77170 }, { "grad_norm": 0.20990364253520966, "learning_rate": 3.3956784134544504e-07, "loss": 0.0082, "step": 77180 }, { "grad_norm": 0.0978960171341896, "learning_rate": 3.3716738401402547e-07, "loss": 0.008, "step": 77190 }, { "grad_norm": 0.18279998004436493, "learning_rate": 3.3477541270241495e-07, "loss": 0.0086, "step": 77200 }, { "grad_norm": 0.23258163034915924, "learning_rate": 3.3239192781934215e-07, "loss": 0.0084, "step": 77210 }, { "grad_norm": 0.18442298471927643, "learning_rate": 3.3001692977207563e-07, "loss": 0.0103, "step": 77220 }, { "grad_norm": 0.12128189951181412, "learning_rate": 3.276504189664409e-07, "loss": 0.0089, "step": 77230 }, { "grad_norm": 0.1506451815366745, "learning_rate": 3.252923958068088e-07, "loss": 0.0106, "step": 77240 }, { "grad_norm": 0.10437377542257309, "learning_rate": 3.2294286069609046e-07, "loss": 0.0083, "step": 77250 }, { "grad_norm": 0.06824491918087006, "learning_rate": 3.2060181403577583e-07, "loss": 0.0071, "step": 77260 }, { "grad_norm": 0.12786684930324554, "learning_rate": 3.1826925622587823e-07, "loss": 0.01, "step": 77270 }, { "grad_norm": 0.07860307395458221, "learning_rate": 3.159451876649622e-07, "loss": 0.0106, "step": 77280 }, { "grad_norm": 0.23757889866828918, "learning_rate": 3.1362960875015444e-07, "loss": 0.0093, "step": 77290 }, { "grad_norm": 0.5151330232620239, "learning_rate": 3.1132251987711637e-07, "loss": 0.0128, "step": 77300 }, { "grad_norm": 0.4242444336414337, "learning_rate": 3.0902392144007695e-07, "loss": 0.0101, "step": 77310 }, { "grad_norm": 0.17792676389217377, "learning_rate": 3.0673381383179436e-07, "loss": 0.0084, "step": 77320 }, { "grad_norm": 0.20090554654598236, "learning_rate": 3.0445219744358875e-07, "loss": 0.0069, "step": 77330 }, { "grad_norm": 0.20324820280075073, "learning_rate": 3.021790726653262e-07, "loss": 0.0123, "step": 77340 }, { "grad_norm": 0.4458092153072357, "learning_rate": 2.9991443988542366e-07, "loss": 0.0096, "step": 77350 }, { "grad_norm": 0.11701258271932602, "learning_rate": 2.976582994908439e-07, "loss": 0.0071, "step": 77360 }, { "grad_norm": 0.10131530463695526, "learning_rate": 2.954106518671007e-07, "loss": 0.0098, "step": 77370 }, { "grad_norm": 0.20173068344593048, "learning_rate": 2.9317149739825356e-07, "loss": 0.0061, "step": 77380 }, { "grad_norm": 0.36882850527763367, "learning_rate": 2.909408364669075e-07, "loss": 0.0084, "step": 77390 }, { "grad_norm": 0.5089243054389954, "learning_rate": 2.887186694542299e-07, "loss": 0.0129, "step": 77400 }, { "grad_norm": 0.09704890102148056, "learning_rate": 2.865049967399225e-07, "loss": 0.0079, "step": 77410 }, { "grad_norm": 0.35836413502693176, "learning_rate": 2.842998187022439e-07, "loss": 0.0079, "step": 77420 }, { "grad_norm": 0.2095605731010437, "learning_rate": 2.8210313571800374e-07, "loss": 0.0075, "step": 77430 }, { "grad_norm": 0.2947752773761749, "learning_rate": 2.7991494816255184e-07, "loss": 0.0067, "step": 77440 }, { "grad_norm": 0.24426592886447906, "learning_rate": 2.777352564097779e-07, "loss": 0.0066, "step": 77450 }, { "grad_norm": 0.3780815601348877, "learning_rate": 2.755640608321508e-07, "loss": 0.0094, "step": 77460 }, { "grad_norm": 0.3911254107952118, "learning_rate": 2.73401361800657e-07, "loss": 0.011, "step": 77470 }, { "grad_norm": 0.1886865794658661, "learning_rate": 2.7124715968484537e-07, "loss": 0.009, "step": 77480 }, { "grad_norm": 0.15794625878334045, "learning_rate": 2.691014548528104e-07, "loss": 0.0087, "step": 77490 }, { "grad_norm": 0.07470650225877762, "learning_rate": 2.669642476711864e-07, "loss": 0.0091, "step": 77500 }, { "grad_norm": 0.15909188985824585, "learning_rate": 2.648355385051815e-07, "loss": 0.0083, "step": 77510 }, { "grad_norm": 0.20187383890151978, "learning_rate": 2.627153277185157e-07, "loss": 0.0085, "step": 77520 }, { "grad_norm": 0.09874219447374344, "learning_rate": 2.606036156734881e-07, "loss": 0.0062, "step": 77530 }, { "grad_norm": 0.15418969094753265, "learning_rate": 2.5850040273092127e-07, "loss": 0.0082, "step": 77540 }, { "grad_norm": 0.15714982151985168, "learning_rate": 2.5640568925020536e-07, "loss": 0.0111, "step": 77550 }, { "grad_norm": 0.13664619624614716, "learning_rate": 2.5431947558927083e-07, "loss": 0.0125, "step": 77560 }, { "grad_norm": 0.12330403923988342, "learning_rate": 2.522417621045825e-07, "loss": 0.0078, "step": 77570 }, { "grad_norm": 0.1309766322374344, "learning_rate": 2.501725491511786e-07, "loss": 0.0101, "step": 77580 }, { "grad_norm": 0.14062131941318512, "learning_rate": 2.481118370826263e-07, "loss": 0.01, "step": 77590 }, { "grad_norm": 0.14025159180164337, "learning_rate": 2.4605962625104393e-07, "loss": 0.0117, "step": 77600 }, { "grad_norm": 0.22860808670520782, "learning_rate": 2.440159170070955e-07, "loss": 0.0112, "step": 77610 }, { "grad_norm": 0.19769912958145142, "learning_rate": 2.41980709699996e-07, "loss": 0.0074, "step": 77620 }, { "grad_norm": 0.3658769130706787, "learning_rate": 2.3995400467751174e-07, "loss": 0.0077, "step": 77630 }, { "grad_norm": 0.26795780658721924, "learning_rate": 2.3793580228594902e-07, "loss": 0.0149, "step": 77640 }, { "grad_norm": 0.1409837305545807, "learning_rate": 2.3592610287015982e-07, "loss": 0.0072, "step": 77650 }, { "grad_norm": 0.17890624701976776, "learning_rate": 2.3392490677354718e-07, "loss": 0.011, "step": 77660 }, { "grad_norm": 0.23655852675437927, "learning_rate": 2.3193221433806533e-07, "loss": 0.0059, "step": 77670 }, { "grad_norm": 0.25057971477508545, "learning_rate": 2.299480259042086e-07, "loss": 0.0094, "step": 77680 }, { "grad_norm": 0.13382652401924133, "learning_rate": 2.2797234181102244e-07, "loss": 0.0115, "step": 77690 }, { "grad_norm": 0.06249704957008362, "learning_rate": 2.2600516239609238e-07, "loss": 0.0059, "step": 77700 }, { "grad_norm": 0.11121995002031326, "learning_rate": 2.240464879955606e-07, "loss": 0.0082, "step": 77710 }, { "grad_norm": 0.24528877437114716, "learning_rate": 2.2209631894410387e-07, "loss": 0.0074, "step": 77720 }, { "grad_norm": 0.09338148683309555, "learning_rate": 2.2015465557496117e-07, "loss": 0.0093, "step": 77730 }, { "grad_norm": 0.2686498165130615, "learning_rate": 2.1822149821990602e-07, "loss": 0.0076, "step": 77740 }, { "grad_norm": 0.5041686296463013, "learning_rate": 2.1629684720926303e-07, "loss": 0.0145, "step": 77750 }, { "grad_norm": 0.20367643237113953, "learning_rate": 2.1438070287189693e-07, "loss": 0.0073, "step": 77760 }, { "grad_norm": 0.14235056936740875, "learning_rate": 2.1247306553523472e-07, "loss": 0.0094, "step": 77770 }, { "grad_norm": 0.12795454263687134, "learning_rate": 2.105739355252323e-07, "loss": 0.0111, "step": 77780 }, { "grad_norm": 0.1036822721362114, "learning_rate": 2.0868331316639678e-07, "loss": 0.0067, "step": 77790 }, { "grad_norm": 0.14565187692642212, "learning_rate": 2.0680119878179193e-07, "loss": 0.0071, "step": 77800 }, { "grad_norm": 0.22137993574142456, "learning_rate": 2.0492759269301054e-07, "loss": 0.0082, "step": 77810 }, { "grad_norm": 0.1109437420964241, "learning_rate": 2.0306249522021315e-07, "loss": 0.0077, "step": 77820 }, { "grad_norm": 0.15412893891334534, "learning_rate": 2.0120590668207816e-07, "loss": 0.0103, "step": 77830 }, { "grad_norm": 0.2223961353302002, "learning_rate": 1.993578273958574e-07, "loss": 0.0069, "step": 77840 }, { "grad_norm": 0.14895012974739075, "learning_rate": 1.975182576773371e-07, "loss": 0.0082, "step": 77850 }, { "grad_norm": 0.442948579788208, "learning_rate": 1.9568719784083812e-07, "loss": 0.0108, "step": 77860 }, { "grad_norm": 0.20262539386749268, "learning_rate": 1.9386464819924898e-07, "loss": 0.0114, "step": 77870 }, { "grad_norm": 0.10478635877370834, "learning_rate": 1.9205060906399285e-07, "loss": 0.0071, "step": 77880 }, { "grad_norm": 0.18073463439941406, "learning_rate": 1.9024508074503845e-07, "loss": 0.0078, "step": 77890 }, { "grad_norm": 0.1823413074016571, "learning_rate": 1.8844806355089452e-07, "loss": 0.0119, "step": 77900 }, { "grad_norm": 0.13494956493377686, "learning_rate": 1.8665955778863208e-07, "loss": 0.0137, "step": 77910 }, { "grad_norm": 0.08581213653087616, "learning_rate": 1.848795637638512e-07, "loss": 0.0067, "step": 77920 }, { "grad_norm": 0.22490592300891876, "learning_rate": 1.831080817807085e-07, "loss": 0.0087, "step": 77930 }, { "grad_norm": 0.22802209854125977, "learning_rate": 1.813451121418952e-07, "loss": 0.0108, "step": 77940 }, { "grad_norm": 0.12931226193904877, "learning_rate": 1.795906551486648e-07, "loss": 0.0067, "step": 77950 }, { "grad_norm": 0.20652173459529877, "learning_rate": 1.7784471110079414e-07, "loss": 0.0079, "step": 77960 }, { "grad_norm": 0.21581998467445374, "learning_rate": 1.7610728029662793e-07, "loss": 0.011, "step": 77970 }, { "grad_norm": 0.1319160759449005, "learning_rate": 1.7437836303303979e-07, "loss": 0.0084, "step": 77980 }, { "grad_norm": 0.14333076775074005, "learning_rate": 1.726579596054545e-07, "loss": 0.0079, "step": 77990 }, { "grad_norm": 0.3204701542854309, "learning_rate": 1.7094607030784803e-07, "loss": 0.0093, "step": 78000 }, { "grad_norm": 0.13726483285427094, "learning_rate": 1.6924269543272532e-07, "loss": 0.0085, "step": 78010 }, { "grad_norm": 0.08404093235731125, "learning_rate": 1.6754783527115348e-07, "loss": 0.0074, "step": 78020 }, { "grad_norm": 0.5483077168464661, "learning_rate": 1.6586149011273422e-07, "loss": 0.0096, "step": 78030 }, { "grad_norm": 0.09962275624275208, "learning_rate": 1.6418366024562038e-07, "loss": 0.0123, "step": 78040 }, { "grad_norm": 0.12867987155914307, "learning_rate": 1.6251434595651037e-07, "loss": 0.0074, "step": 78050 }, { "grad_norm": 0.15848660469055176, "learning_rate": 1.608535475306372e-07, "loss": 0.0062, "step": 78060 }, { "grad_norm": 0.2748711407184601, "learning_rate": 1.5920126525179048e-07, "loss": 0.0074, "step": 78070 }, { "grad_norm": 0.22855007648468018, "learning_rate": 1.5755749940229435e-07, "loss": 0.0097, "step": 78080 }, { "grad_norm": 0.2002369463443756, "learning_rate": 1.559222502630353e-07, "loss": 0.0065, "step": 78090 }, { "grad_norm": 0.15827713906764984, "learning_rate": 1.5429551811341757e-07, "loss": 0.0121, "step": 78100 }, { "grad_norm": 0.2436273992061615, "learning_rate": 1.5267730323141882e-07, "loss": 0.012, "step": 78110 }, { "grad_norm": 0.20879407227039337, "learning_rate": 1.5106760589353454e-07, "loss": 0.0085, "step": 78120 }, { "grad_norm": 0.14785034954547882, "learning_rate": 1.4946642637483355e-07, "loss": 0.0102, "step": 78130 }, { "grad_norm": 0.24160154163837433, "learning_rate": 1.4787376494889703e-07, "loss": 0.0082, "step": 78140 }, { "grad_norm": 0.08096646517515182, "learning_rate": 1.4628962188787955e-07, "loss": 0.0078, "step": 78150 }, { "grad_norm": 0.23746484518051147, "learning_rate": 1.4471399746247006e-07, "loss": 0.0124, "step": 78160 }, { "grad_norm": 0.20934277772903442, "learning_rate": 1.4314689194188103e-07, "loss": 0.0071, "step": 78170 }, { "grad_norm": 0.43103620409965515, "learning_rate": 1.4158830559390933e-07, "loss": 0.0107, "step": 78180 }, { "grad_norm": 0.17438878118991852, "learning_rate": 1.4003823868486422e-07, "loss": 0.0219, "step": 78190 }, { "grad_norm": 0.1779635101556778, "learning_rate": 1.3849669147960598e-07, "loss": 0.0077, "step": 78200 }, { "grad_norm": 0.09934873133897781, "learning_rate": 1.3696366424155726e-07, "loss": 0.0074, "step": 78210 }, { "grad_norm": 0.426053911447525, "learning_rate": 1.35439157232653e-07, "loss": 0.0138, "step": 78220 }, { "grad_norm": 0.17636533081531525, "learning_rate": 1.3392317071340144e-07, "loss": 0.0097, "step": 78230 }, { "grad_norm": 0.2639143168926239, "learning_rate": 1.3241570494283984e-07, "loss": 0.0104, "step": 78240 }, { "grad_norm": 0.23073814809322357, "learning_rate": 1.3091676017855103e-07, "loss": 0.0093, "step": 78250 }, { "grad_norm": 0.4942803382873535, "learning_rate": 1.2942633667666904e-07, "loss": 0.0099, "step": 78260 }, { "grad_norm": 0.11628612875938416, "learning_rate": 1.2794443469185679e-07, "loss": 0.0097, "step": 78270 }, { "grad_norm": 0.18178361654281616, "learning_rate": 1.2647105447734508e-07, "loss": 0.0078, "step": 78280 }, { "grad_norm": 0.23962998390197754, "learning_rate": 1.2500619628488254e-07, "loss": 0.007, "step": 78290 }, { "grad_norm": 0.2904914915561676, "learning_rate": 1.2354986036477446e-07, "loss": 0.0141, "step": 78300 }, { "grad_norm": 0.15366089344024658, "learning_rate": 1.221020469658718e-07, "loss": 0.0061, "step": 78310 }, { "grad_norm": 0.1750708818435669, "learning_rate": 1.2066275633556556e-07, "loss": 0.0076, "step": 78320 }, { "grad_norm": 0.1418900340795517, "learning_rate": 1.192319887197979e-07, "loss": 0.0066, "step": 78330 }, { "grad_norm": 0.21233324706554413, "learning_rate": 1.1780974436303438e-07, "loss": 0.0066, "step": 78340 }, { "grad_norm": 0.18862758576869965, "learning_rate": 1.1639602350830836e-07, "loss": 0.0076, "step": 78350 }, { "grad_norm": 0.2887609302997589, "learning_rate": 1.1499082639718217e-07, "loss": 0.0079, "step": 78360 }, { "grad_norm": 0.14431753754615784, "learning_rate": 1.1359415326976374e-07, "loss": 0.008, "step": 78370 }, { "grad_norm": 0.1690782606601715, "learning_rate": 1.1220600436470663e-07, "loss": 0.0082, "step": 78380 }, { "grad_norm": 0.1771915853023529, "learning_rate": 1.1082637991920996e-07, "loss": 0.0061, "step": 78390 }, { "grad_norm": 0.17941829562187195, "learning_rate": 1.0945528016901851e-07, "loss": 0.0095, "step": 78400 }, { "grad_norm": 0.12118633091449738, "learning_rate": 1.0809270534840599e-07, "loss": 0.0068, "step": 78410 }, { "grad_norm": 0.2850576341152191, "learning_rate": 1.067386556902028e-07, "loss": 0.006, "step": 78420 }, { "grad_norm": 0.08828128129243851, "learning_rate": 1.05393131425785e-07, "loss": 0.007, "step": 78430 }, { "grad_norm": 0.512039303779602, "learning_rate": 1.0405613278505199e-07, "loss": 0.0049, "step": 78440 }, { "grad_norm": 0.16198158264160156, "learning_rate": 1.027276599964766e-07, "loss": 0.0082, "step": 78450 }, { "grad_norm": 0.2494453340768814, "learning_rate": 1.0140771328704391e-07, "loss": 0.0052, "step": 78460 }, { "grad_norm": 0.05923125892877579, "learning_rate": 1.0009629288231237e-07, "loss": 0.0077, "step": 78470 }, { "grad_norm": 0.09276635944843292, "learning_rate": 9.879339900635276e-08, "loss": 0.0082, "step": 78480 }, { "grad_norm": 0.3163822889328003, "learning_rate": 9.74990318817981e-08, "loss": 0.0099, "step": 78490 }, { "grad_norm": 0.24734550714492798, "learning_rate": 9.621319172982701e-08, "loss": 0.0074, "step": 78500 }, { "grad_norm": 0.10821401327848434, "learning_rate": 9.493587877015264e-08, "loss": 0.0071, "step": 78510 }, { "grad_norm": 0.8937591910362244, "learning_rate": 9.366709322102263e-08, "loss": 0.0136, "step": 78520 }, { "grad_norm": 0.4032594561576843, "learning_rate": 9.240683529924688e-08, "loss": 0.0084, "step": 78530 }, { "grad_norm": 0.1361476629972458, "learning_rate": 9.115510522016979e-08, "loss": 0.006, "step": 78540 }, { "grad_norm": 0.13643647730350494, "learning_rate": 8.991190319767583e-08, "loss": 0.0076, "step": 78550 }, { "grad_norm": 0.07235460728406906, "learning_rate": 8.867722944419509e-08, "loss": 0.0096, "step": 78560 }, { "grad_norm": 0.27662134170532227, "learning_rate": 8.745108417069214e-08, "loss": 0.0084, "step": 78570 }, { "grad_norm": 0.207101508975029, "learning_rate": 8.623346758669381e-08, "loss": 0.0057, "step": 78580 }, { "grad_norm": 0.20546819269657135, "learning_rate": 8.502437990025037e-08, "loss": 0.008, "step": 78590 }, { "grad_norm": 0.3960334062576294, "learning_rate": 8.382382131795764e-08, "loss": 0.0103, "step": 78600 }, { "grad_norm": 0.20786140859127045, "learning_rate": 8.263179204496818e-08, "loss": 0.0107, "step": 78610 }, { "grad_norm": 0.2568912208080292, "learning_rate": 8.144829228496354e-08, "loss": 0.0079, "step": 78620 }, { "grad_norm": 0.26972219347953796, "learning_rate": 8.027332224016526e-08, "loss": 0.0087, "step": 78630 }, { "grad_norm": 0.26141157746315, "learning_rate": 7.910688211135164e-08, "loss": 0.0083, "step": 78640 }, { "grad_norm": 0.12799103558063507, "learning_rate": 7.794897209783546e-08, "loss": 0.0106, "step": 78650 }, { "grad_norm": 0.05032108351588249, "learning_rate": 7.679959239746403e-08, "loss": 0.0085, "step": 78660 }, { "grad_norm": 0.15976715087890625, "learning_rate": 7.565874320664135e-08, "loss": 0.0091, "step": 78670 }, { "grad_norm": 0.2320525050163269, "learning_rate": 7.45264247203059e-08, "loss": 0.0068, "step": 78680 }, { "grad_norm": 0.25701606273651123, "learning_rate": 7.340263713194184e-08, "loss": 0.0092, "step": 78690 }, { "grad_norm": 0.22490952908992767, "learning_rate": 7.228738063356777e-08, "loss": 0.0093, "step": 78700 }, { "grad_norm": 0.3541709780693054, "learning_rate": 7.118065541575903e-08, "loss": 0.0105, "step": 78710 }, { "grad_norm": 0.16447409987449646, "learning_rate": 7.008246166761435e-08, "loss": 0.0068, "step": 78720 }, { "grad_norm": 0.20306642353534698, "learning_rate": 6.899279957679472e-08, "loss": 0.0077, "step": 78730 }, { "grad_norm": 0.17050060629844666, "learning_rate": 6.791166932949011e-08, "loss": 0.0077, "step": 78740 }, { "grad_norm": 0.213981032371521, "learning_rate": 6.683907111043608e-08, "loss": 0.0086, "step": 78750 }, { "grad_norm": 0.08528155088424683, "learning_rate": 6.577500510290824e-08, "loss": 0.0064, "step": 78760 }, { "grad_norm": 0.0626634880900383, "learning_rate": 6.47194714887278e-08, "loss": 0.0232, "step": 78770 }, { "grad_norm": 0.3799107074737549, "learning_rate": 6.367247044825608e-08, "loss": 0.0089, "step": 78780 }, { "grad_norm": 0.15107055008411407, "learning_rate": 6.263400216039994e-08, "loss": 0.0066, "step": 78790 }, { "grad_norm": 0.27672532200813293, "learning_rate": 6.160406680260078e-08, "loss": 0.0236, "step": 78800 }, { "grad_norm": 0.23039250075817108, "learning_rate": 6.058266455084561e-08, "loss": 0.0076, "step": 78810 }, { "grad_norm": 0.14869818091392517, "learning_rate": 5.956979557967257e-08, "loss": 0.0076, "step": 78820 }, { "grad_norm": 0.14810286462306976, "learning_rate": 5.856546006214325e-08, "loss": 0.0088, "step": 78830 }, { "grad_norm": 0.12049911171197891, "learning_rate": 5.7569658169881466e-08, "loss": 0.0069, "step": 78840 }, { "grad_norm": 0.2679094076156616, "learning_rate": 5.658239007303445e-08, "loss": 0.0109, "step": 78850 }, { "grad_norm": 0.11066487431526184, "learning_rate": 5.560365594030059e-08, "loss": 0.0153, "step": 78860 }, { "grad_norm": 0.18224933743476868, "learning_rate": 5.463345593891833e-08, "loss": 0.0072, "step": 78870 }, { "grad_norm": 0.09794972091913223, "learning_rate": 5.367179023467173e-08, "loss": 0.0103, "step": 78880 }, { "grad_norm": 0.24365296959877014, "learning_rate": 5.271865899187378e-08, "loss": 0.0079, "step": 78890 }, { "grad_norm": 0.4932871460914612, "learning_rate": 5.177406237340532e-08, "loss": 0.0149, "step": 78900 }, { "grad_norm": 0.21470558643341064, "learning_rate": 5.083800054065946e-08, "loss": 0.0063, "step": 78910 }, { "grad_norm": 0.20988048613071442, "learning_rate": 4.991047365358603e-08, "loss": 0.009, "step": 78920 }, { "grad_norm": 0.4274558126926422, "learning_rate": 4.899148187067493e-08, "loss": 0.0204, "step": 78930 }, { "grad_norm": 0.08444773405790329, "learning_rate": 4.808102534895609e-08, "loss": 0.0082, "step": 78940 }, { "grad_norm": 0.09728754311800003, "learning_rate": 4.717910424400507e-08, "loss": 0.0068, "step": 78950 }, { "grad_norm": 0.3084228038787842, "learning_rate": 4.628571870993193e-08, "loss": 0.0103, "step": 78960 }, { "grad_norm": 0.20646139979362488, "learning_rate": 4.5400868899392324e-08, "loss": 0.0098, "step": 78970 }, { "grad_norm": 0.36889341473579407, "learning_rate": 4.4524554963576436e-08, "loss": 0.0149, "step": 78980 }, { "grad_norm": 0.34651926159858704, "learning_rate": 4.36567770522367e-08, "loss": 0.0066, "step": 78990 }, { "grad_norm": 0.2594287395477295, "learning_rate": 4.279753531364339e-08, "loss": 0.0085, "step": 79000 }, { "grad_norm": 0.35447531938552856, "learning_rate": 4.1946829894617955e-08, "loss": 0.0109, "step": 79010 }, { "grad_norm": 0.10345876961946487, "learning_rate": 4.11046609405219e-08, "loss": 0.0134, "step": 79020 }, { "grad_norm": 0.04896106570959091, "learning_rate": 4.027102859526233e-08, "loss": 0.0069, "step": 79030 }, { "grad_norm": 0.15977777540683746, "learning_rate": 3.9445933001280856e-08, "loss": 0.0089, "step": 79040 }, { "grad_norm": 0.07141826301813126, "learning_rate": 3.8629374299564704e-08, "loss": 0.0105, "step": 79050 }, { "grad_norm": 0.2293577939271927, "learning_rate": 3.78213526296467e-08, "loss": 0.0088, "step": 79060 }, { "grad_norm": 0.2887113094329834, "learning_rate": 3.702186812958308e-08, "loss": 0.0084, "step": 79070 }, { "grad_norm": 0.16343437135219574, "learning_rate": 3.623092093599789e-08, "loss": 0.0116, "step": 79080 }, { "grad_norm": 0.4238739311695099, "learning_rate": 3.544851118403303e-08, "loss": 0.0112, "step": 79090 }, { "grad_norm": 0.15568912029266357, "learning_rate": 3.46746390073871e-08, "loss": 0.0098, "step": 79100 }, { "grad_norm": 0.2503274977207184, "learning_rate": 3.390930453829322e-08, "loss": 0.0159, "step": 79110 }, { "grad_norm": 0.140639066696167, "learning_rate": 3.3152507907519005e-08, "loss": 0.0111, "step": 79120 }, { "grad_norm": 0.1177576407790184, "learning_rate": 3.240424924438323e-08, "loss": 0.0063, "step": 79130 }, { "grad_norm": 0.18087607622146606, "learning_rate": 3.1664528676750295e-08, "loss": 0.0082, "step": 79140 }, { "grad_norm": 0.1098385825753212, "learning_rate": 3.093334633100797e-08, "loss": 0.0195, "step": 79150 }, { "grad_norm": 0.23291164636611938, "learning_rate": 3.021070233210077e-08, "loss": 0.0084, "step": 79160 }, { "grad_norm": 0.1166108101606369, "learning_rate": 2.9496596803507692e-08, "loss": 0.0077, "step": 79170 }, { "grad_norm": 0.21756993234157562, "learning_rate": 2.879102986725335e-08, "loss": 0.0087, "step": 79180 }, { "grad_norm": 0.28552061319351196, "learning_rate": 2.809400164389131e-08, "loss": 0.01, "step": 79190 }, { "grad_norm": 0.12148722261190414, "learning_rate": 2.740551225253185e-08, "loss": 0.0079, "step": 79200 }, { "grad_norm": 0.17226240038871765, "learning_rate": 2.6725561810819754e-08, "loss": 0.0113, "step": 79210 }, { "grad_norm": 0.22697250545024872, "learning_rate": 2.605415043493431e-08, "loss": 0.0072, "step": 79220 }, { "grad_norm": 0.11682655662298203, "learning_rate": 2.5391278239605965e-08, "loss": 0.0062, "step": 79230 }, { "grad_norm": 0.16092108190059662, "learning_rate": 2.4736945338094118e-08, "loss": 0.0071, "step": 79240 }, { "grad_norm": 0.10956991463899612, "learning_rate": 2.4091151842214887e-08, "loss": 0.0063, "step": 79250 }, { "grad_norm": 0.1435144990682602, "learning_rate": 2.3453897862318884e-08, "loss": 0.0103, "step": 79260 }, { "grad_norm": 0.37677067518234253, "learning_rate": 2.2825183507285686e-08, "loss": 0.009, "step": 79270 }, { "grad_norm": 0.1824091523885727, "learning_rate": 2.220500888455157e-08, "loss": 0.0116, "step": 79280 }, { "grad_norm": 0.14450061321258545, "learning_rate": 2.1593374100081777e-08, "loss": 0.0087, "step": 79290 }, { "grad_norm": 0.1556342989206314, "learning_rate": 2.09902792583927e-08, "loss": 0.0114, "step": 79300 }, { "grad_norm": 0.2117323875427246, "learning_rate": 2.0395724462540788e-08, "loss": 0.0084, "step": 79310 }, { "grad_norm": 0.0912964940071106, "learning_rate": 1.9809709814111453e-08, "loss": 0.0067, "step": 79320 }, { "grad_norm": 0.1371891349554062, "learning_rate": 1.923223541324126e-08, "loss": 0.0131, "step": 79330 }, { "grad_norm": 0.210099458694458, "learning_rate": 1.8663301358606833e-08, "loss": 0.0103, "step": 79340 }, { "grad_norm": 0.28131648898124695, "learning_rate": 1.8102907747419295e-08, "loss": 0.0129, "step": 79350 }, { "grad_norm": 0.1253829151391983, "learning_rate": 1.7551054675435385e-08, "loss": 0.0064, "step": 79360 }, { "grad_norm": 0.15931499004364014, "learning_rate": 1.7007742236957447e-08, "loss": 0.0111, "step": 79370 }, { "grad_norm": 0.2209671586751938, "learning_rate": 1.647297052481678e-08, "loss": 0.0092, "step": 79380 }, { "grad_norm": 0.26622211933135986, "learning_rate": 1.5946739630390285e-08, "loss": 0.0117, "step": 79390 }, { "grad_norm": 0.15076054632663727, "learning_rate": 1.5429049643606032e-08, "loss": 0.0153, "step": 79400 }, { "grad_norm": 0.32346588373184204, "learning_rate": 1.4919900652909935e-08, "loss": 0.0064, "step": 79410 }, { "grad_norm": 0.23598520457744598, "learning_rate": 1.4419292745310175e-08, "loss": 0.0081, "step": 79420 }, { "grad_norm": 0.15536685287952423, "learning_rate": 1.3927226006343885e-08, "loss": 0.0086, "step": 79430 }, { "grad_norm": 0.2947721779346466, "learning_rate": 1.3443700520093805e-08, "loss": 0.0083, "step": 79440 }, { "grad_norm": 0.10751581192016602, "learning_rate": 1.296871636917718e-08, "loss": 0.0107, "step": 79450 }, { "grad_norm": 0.3117837905883789, "learning_rate": 1.2502273634762419e-08, "loss": 0.0102, "step": 79460 }, { "grad_norm": 0.21860986948013306, "learning_rate": 1.2044372396546876e-08, "loss": 0.0084, "step": 79470 }, { "grad_norm": 0.21553905308246613, "learning_rate": 1.1595012732773524e-08, "loss": 0.009, "step": 79480 }, { "grad_norm": 0.10820763558149338, "learning_rate": 1.1154194720225386e-08, "loss": 0.0083, "step": 79490 }, { "grad_norm": 0.30179235339164734, "learning_rate": 1.0721918434231093e-08, "loss": 0.0077, "step": 79500 }, { "grad_norm": 0.3909887671470642, "learning_rate": 1.0298183948648232e-08, "loss": 0.0083, "step": 79510 }, { "grad_norm": 0.37301626801490784, "learning_rate": 9.88299133588555e-09, "loss": 0.0085, "step": 79520 }, { "grad_norm": 0.2221021056175232, "learning_rate": 9.476340666891847e-09, "loss": 0.0124, "step": 79530 }, { "grad_norm": 0.11950311809778214, "learning_rate": 9.078232011139332e-09, "loss": 0.0069, "step": 79540 }, { "grad_norm": 0.14997027814388275, "learning_rate": 8.68866543666802e-09, "loss": 0.0089, "step": 79550 }, { "grad_norm": 0.24554608762264252, "learning_rate": 8.307641010035782e-09, "loss": 0.0105, "step": 79560 }, { "grad_norm": 0.1369452327489853, "learning_rate": 7.93515879635165e-09, "loss": 0.0083, "step": 79570 }, { "grad_norm": 0.10459040105342865, "learning_rate": 7.571218859264706e-09, "loss": 0.0085, "step": 79580 }, { "grad_norm": 0.18415023386478424, "learning_rate": 7.215821260958544e-09, "loss": 0.0113, "step": 79590 }, { "grad_norm": 0.28199970722198486, "learning_rate": 6.868966062162363e-09, "loss": 0.0066, "step": 79600 }, { "grad_norm": 0.19385091960430145, "learning_rate": 6.530653322145419e-09, "loss": 0.0085, "step": 79610 }, { "grad_norm": 0.27235278487205505, "learning_rate": 6.200883098717025e-09, "loss": 0.0085, "step": 79620 }, { "grad_norm": 0.26715564727783203, "learning_rate": 5.879655448226551e-09, "loss": 0.0136, "step": 79630 }, { "grad_norm": 0.15180204808712006, "learning_rate": 5.566970425557872e-09, "loss": 0.0077, "step": 79640 }, { "grad_norm": 0.25649774074554443, "learning_rate": 5.26282808414047e-09, "loss": 0.0088, "step": 79650 }, { "grad_norm": 0.19494378566741943, "learning_rate": 4.967228475949437e-09, "loss": 0.0105, "step": 79660 }, { "grad_norm": 0.21393068134784698, "learning_rate": 4.680171651494369e-09, "loss": 0.0087, "step": 79670 }, { "grad_norm": 0.36545032262802124, "learning_rate": 4.40165765981937e-09, "loss": 0.0086, "step": 79680 }, { "grad_norm": 0.10424897819757462, "learning_rate": 4.131686548519698e-09, "loss": 0.0067, "step": 79690 }, { "grad_norm": 0.16830025613307953, "learning_rate": 3.8702583637251214e-09, "loss": 0.0062, "step": 79700 }, { "grad_norm": 0.23582200706005096, "learning_rate": 3.617373150105463e-09, "loss": 0.0068, "step": 79710 }, { "grad_norm": 0.16211077570915222, "learning_rate": 3.3730309508706036e-09, "loss": 0.0072, "step": 79720 }, { "grad_norm": 0.1225225180387497, "learning_rate": 3.137231807781582e-09, "loss": 0.0055, "step": 79730 }, { "grad_norm": 0.12553738057613373, "learning_rate": 2.9099757611172894e-09, "loss": 0.0088, "step": 79740 }, { "grad_norm": 0.19021297991275787, "learning_rate": 2.6912628497133275e-09, "loss": 0.0075, "step": 79750 }, { "grad_norm": 0.1721561700105667, "learning_rate": 2.481093110945354e-09, "loss": 0.0117, "step": 79760 }, { "grad_norm": 0.0945877879858017, "learning_rate": 2.2794665807235327e-09, "loss": 0.0087, "step": 79770 }, { "grad_norm": 0.19593845307826996, "learning_rate": 2.0863832934980843e-09, "loss": 0.0055, "step": 79780 }, { "grad_norm": 0.13456015288829803, "learning_rate": 1.901843282264837e-09, "loss": 0.0077, "step": 79790 }, { "grad_norm": 0.20442892611026764, "learning_rate": 1.7258465785541245e-09, "loss": 0.0096, "step": 79800 }, { "grad_norm": 0.07798979431390762, "learning_rate": 1.5583932124418887e-09, "loss": 0.0073, "step": 79810 }, { "grad_norm": 0.34963154792785645, "learning_rate": 1.3994832125441282e-09, "loss": 0.0096, "step": 79820 }, { "grad_norm": 0.24187572300434113, "learning_rate": 1.2491166060057958e-09, "loss": 0.0105, "step": 79830 }, { "grad_norm": 0.18471114337444305, "learning_rate": 1.1072934185230032e-09, "loss": 0.0089, "step": 79840 }, { "grad_norm": 0.24562570452690125, "learning_rate": 9.740136743319195e-10, "loss": 0.0087, "step": 79850 }, { "grad_norm": 0.19662465155124664, "learning_rate": 8.492773962087696e-10, "loss": 0.0076, "step": 79860 }, { "grad_norm": 0.4764423072338104, "learning_rate": 7.330846054587337e-10, "loss": 0.0147, "step": 79870 }, { "grad_norm": 0.20272937417030334, "learning_rate": 6.254353219492526e-10, "loss": 0.0081, "step": 79880 }, { "grad_norm": 0.23056942224502563, "learning_rate": 5.263295640600685e-10, "loss": 0.0094, "step": 79890 }, { "grad_norm": 0.07286084443330765, "learning_rate": 4.357673487387359e-10, "loss": 0.0064, "step": 79900 }, { "grad_norm": 0.3544404208660126, "learning_rate": 3.537486914506616e-10, "loss": 0.0077, "step": 79910 }, { "grad_norm": 0.10800694674253464, "learning_rate": 2.8027360621241117e-10, "loss": 0.0065, "step": 79920 }, { "grad_norm": 0.16107812523841858, "learning_rate": 2.153421055806071e-10, "loss": 0.0104, "step": 79930 }, { "grad_norm": 0.0779181569814682, "learning_rate": 1.5895420064637733e-10, "loss": 0.0102, "step": 79940 }, { "grad_norm": 0.12315474450588226, "learning_rate": 1.1110990105200891e-10, "loss": 0.008, "step": 79950 }, { "grad_norm": 0.22940297424793243, "learning_rate": 7.180921496874326e-11, "loss": 0.0086, "step": 79960 }, { "grad_norm": 0.2701982855796814, "learning_rate": 4.10521491134297e-11, "loss": 0.0076, "step": 79970 }, { "grad_norm": 0.19711308181285858, "learning_rate": 1.883870874297422e-11, "loss": 0.0063, "step": 79980 }, { "grad_norm": 0.20381878316402435, "learning_rate": 5.1688976432373584e-12, "loss": 0.0097, "step": 79990 }, { "grad_norm": 0.20041076838970184, "learning_rate": 4.271816234080461e-14, "loss": 0.0077, "step": 80000 } ], "logging_steps": 10, "max_steps": 80000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 40000, "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": 8, "trial_name": null, "trial_params": null }