{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.6666666666666666, "eval_steps": 500, "global_step": 20000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 1.3314505815505981, "learning_rate": 6.000000000000001e-07, "loss": 1.2179, "step": 10 }, { "grad_norm": 1.5507653951644897, "learning_rate": 1.2666666666666667e-06, "loss": 1.2173, "step": 20 }, { "grad_norm": 1.178682565689087, "learning_rate": 1.9333333333333336e-06, "loss": 1.212, "step": 30 }, { "grad_norm": 0.68263179063797, "learning_rate": 2.6e-06, "loss": 1.1905, "step": 40 }, { "grad_norm": 0.5855809450149536, "learning_rate": 3.2666666666666666e-06, "loss": 1.1888, "step": 50 }, { "grad_norm": 0.3312073051929474, "learning_rate": 3.9333333333333335e-06, "loss": 1.1345, "step": 60 }, { "grad_norm": 0.3775692582130432, "learning_rate": 4.6e-06, "loss": 1.1437, "step": 70 }, { "grad_norm": 0.35685569047927856, "learning_rate": 5.266666666666667e-06, "loss": 1.1419, "step": 80 }, { "grad_norm": 0.4233785569667816, "learning_rate": 5.933333333333334e-06, "loss": 1.1079, "step": 90 }, { "grad_norm": 0.45181751251220703, "learning_rate": 6.6e-06, "loss": 1.1288, "step": 100 }, { "grad_norm": 0.4978986084461212, "learning_rate": 7.266666666666668e-06, "loss": 1.1431, "step": 110 }, { "grad_norm": 0.4479588270187378, "learning_rate": 7.933333333333334e-06, "loss": 1.1471, "step": 120 }, { "grad_norm": 0.5029391050338745, "learning_rate": 8.599999999999999e-06, "loss": 1.117, "step": 130 }, { "grad_norm": 0.8853879570960999, "learning_rate": 9.266666666666667e-06, "loss": 1.1064, "step": 140 }, { "grad_norm": 0.5811502933502197, "learning_rate": 9.933333333333334e-06, "loss": 1.1036, "step": 150 }, { "grad_norm": 0.7939465045928955, "learning_rate": 1.06e-05, "loss": 1.0916, "step": 160 }, { "grad_norm": 0.6251223087310791, "learning_rate": 1.1266666666666667e-05, "loss": 1.0927, "step": 170 }, { "grad_norm": 0.7339620590209961, "learning_rate": 1.1933333333333333e-05, "loss": 1.0813, "step": 180 }, { "grad_norm": 0.6614446640014648, "learning_rate": 1.2600000000000001e-05, "loss": 1.0855, "step": 190 }, { "grad_norm": 0.501140832901001, "learning_rate": 1.3266666666666666e-05, "loss": 1.0701, "step": 200 }, { "grad_norm": 0.7906708717346191, "learning_rate": 1.3933333333333334e-05, "loss": 1.0887, "step": 210 }, { "grad_norm": 0.6906638145446777, "learning_rate": 1.4599999999999999e-05, "loss": 1.0612, "step": 220 }, { "grad_norm": 0.7390074133872986, "learning_rate": 1.5266666666666667e-05, "loss": 1.0774, "step": 230 }, { "grad_norm": 0.5626471638679504, "learning_rate": 1.5933333333333332e-05, "loss": 1.0671, "step": 240 }, { "grad_norm": 0.7262792587280273, "learning_rate": 1.66e-05, "loss": 1.0679, "step": 250 }, { "grad_norm": 0.5234604477882385, "learning_rate": 1.726666666666667e-05, "loss": 1.0684, "step": 260 }, { "grad_norm": 0.5092823505401611, "learning_rate": 1.7933333333333337e-05, "loss": 1.0641, "step": 270 }, { "grad_norm": 0.6354775428771973, "learning_rate": 1.86e-05, "loss": 1.0725, "step": 280 }, { "grad_norm": 0.5659180283546448, "learning_rate": 1.926666666666667e-05, "loss": 1.0633, "step": 290 }, { "grad_norm": 0.8407434225082397, "learning_rate": 1.9933333333333334e-05, "loss": 1.0568, "step": 300 }, { "grad_norm": 0.7417938113212585, "learning_rate": 2.06e-05, "loss": 1.0474, "step": 310 }, { "grad_norm": 0.7800794243812561, "learning_rate": 2.1266666666666667e-05, "loss": 1.041, "step": 320 }, { "grad_norm": 0.7304247617721558, "learning_rate": 2.1933333333333332e-05, "loss": 1.052, "step": 330 }, { "grad_norm": 0.9249001741409302, "learning_rate": 2.26e-05, "loss": 1.027, "step": 340 }, { "grad_norm": 0.7380666732788086, "learning_rate": 2.326666666666667e-05, "loss": 1.0311, "step": 350 }, { "grad_norm": 0.7331956624984741, "learning_rate": 2.3933333333333337e-05, "loss": 1.0275, "step": 360 }, { "grad_norm": 0.9304872155189514, "learning_rate": 2.46e-05, "loss": 1.0111, "step": 370 }, { "grad_norm": 0.8601149916648865, "learning_rate": 2.5266666666666666e-05, "loss": 0.997, "step": 380 }, { "grad_norm": 0.9442141652107239, "learning_rate": 2.5933333333333338e-05, "loss": 0.9944, "step": 390 }, { "grad_norm": 0.8515417575836182, "learning_rate": 2.6600000000000003e-05, "loss": 0.9864, "step": 400 }, { "grad_norm": 0.9509739875793457, "learning_rate": 2.7266666666666668e-05, "loss": 0.955, "step": 410 }, { "grad_norm": 1.084962248802185, "learning_rate": 2.7933333333333332e-05, "loss": 0.9272, "step": 420 }, { "grad_norm": 0.9800599813461304, "learning_rate": 2.86e-05, "loss": 0.9069, "step": 430 }, { "grad_norm": 1.0645278692245483, "learning_rate": 2.926666666666667e-05, "loss": 0.8653, "step": 440 }, { "grad_norm": 1.049734354019165, "learning_rate": 2.9933333333333337e-05, "loss": 0.8439, "step": 450 }, { "grad_norm": 1.1224632263183594, "learning_rate": 3.06e-05, "loss": 0.8167, "step": 460 }, { "grad_norm": 1.0559662580490112, "learning_rate": 3.126666666666666e-05, "loss": 0.7786, "step": 470 }, { "grad_norm": 1.1540216207504272, "learning_rate": 3.1933333333333335e-05, "loss": 0.7453, "step": 480 }, { "grad_norm": 1.0124099254608154, "learning_rate": 3.26e-05, "loss": 0.7176, "step": 490 }, { "grad_norm": 1.3634672164916992, "learning_rate": 3.326666666666667e-05, "loss": 0.6796, "step": 500 }, { "grad_norm": 1.2447317838668823, "learning_rate": 3.3933333333333336e-05, "loss": 0.6478, "step": 510 }, { "grad_norm": 1.3635326623916626, "learning_rate": 3.46e-05, "loss": 0.6181, "step": 520 }, { "grad_norm": 1.189446210861206, "learning_rate": 3.526666666666667e-05, "loss": 0.5902, "step": 530 }, { "grad_norm": 0.9477949738502502, "learning_rate": 3.593333333333334e-05, "loss": 0.5247, "step": 540 }, { "grad_norm": 1.3037503957748413, "learning_rate": 3.66e-05, "loss": 0.5328, "step": 550 }, { "grad_norm": 1.5641331672668457, "learning_rate": 3.726666666666667e-05, "loss": 0.5011, "step": 560 }, { "grad_norm": 1.127152442932129, "learning_rate": 3.793333333333334e-05, "loss": 0.4387, "step": 570 }, { "grad_norm": 1.2061371803283691, "learning_rate": 3.86e-05, "loss": 0.407, "step": 580 }, { "grad_norm": 1.273613452911377, "learning_rate": 3.926666666666667e-05, "loss": 0.3972, "step": 590 }, { "grad_norm": 1.1315085887908936, "learning_rate": 3.993333333333333e-05, "loss": 0.3814, "step": 600 }, { "grad_norm": 1.5654405355453491, "learning_rate": 4.0600000000000004e-05, "loss": 0.3573, "step": 610 }, { "grad_norm": 1.298802375793457, "learning_rate": 4.126666666666667e-05, "loss": 0.335, "step": 620 }, { "grad_norm": 1.101986050605774, "learning_rate": 4.1933333333333334e-05, "loss": 0.324, "step": 630 }, { "grad_norm": 1.480060338973999, "learning_rate": 4.26e-05, "loss": 0.3052, "step": 640 }, { "grad_norm": 1.656790018081665, "learning_rate": 4.3266666666666664e-05, "loss": 0.2794, "step": 650 }, { "grad_norm": 1.57819664478302, "learning_rate": 4.3933333333333335e-05, "loss": 0.2932, "step": 660 }, { "grad_norm": 1.248097538948059, "learning_rate": 4.46e-05, "loss": 0.2431, "step": 670 }, { "grad_norm": 1.4449474811553955, "learning_rate": 4.526666666666667e-05, "loss": 0.2639, "step": 680 }, { "grad_norm": 1.2288621664047241, "learning_rate": 4.5933333333333336e-05, "loss": 0.2493, "step": 690 }, { "grad_norm": 1.2860236167907715, "learning_rate": 4.660000000000001e-05, "loss": 0.2454, "step": 700 }, { "grad_norm": 1.2656257152557373, "learning_rate": 4.726666666666667e-05, "loss": 0.2165, "step": 710 }, { "grad_norm": 1.2987090349197388, "learning_rate": 4.793333333333334e-05, "loss": 0.2253, "step": 720 }, { "grad_norm": 1.233296513557434, "learning_rate": 4.86e-05, "loss": 0.1972, "step": 730 }, { "grad_norm": 1.1606056690216064, "learning_rate": 4.926666666666667e-05, "loss": 0.1532, "step": 740 }, { "grad_norm": 0.9332922697067261, "learning_rate": 4.993333333333334e-05, "loss": 0.1803, "step": 750 }, { "grad_norm": 1.067075490951538, "learning_rate": 5.0600000000000003e-05, "loss": 0.1656, "step": 760 }, { "grad_norm": 1.1522564888000488, "learning_rate": 5.1266666666666675e-05, "loss": 0.1762, "step": 770 }, { "grad_norm": 0.9858033061027527, "learning_rate": 5.193333333333333e-05, "loss": 0.2118, "step": 780 }, { "grad_norm": 0.9792870879173279, "learning_rate": 5.2600000000000005e-05, "loss": 0.1497, "step": 790 }, { "grad_norm": 1.120789647102356, "learning_rate": 5.326666666666666e-05, "loss": 0.1549, "step": 800 }, { "grad_norm": 0.9506876468658447, "learning_rate": 5.3933333333333334e-05, "loss": 0.165, "step": 810 }, { "grad_norm": 0.9334282279014587, "learning_rate": 5.4600000000000006e-05, "loss": 0.1629, "step": 820 }, { "grad_norm": 1.1798714399337769, "learning_rate": 5.5266666666666664e-05, "loss": 0.1924, "step": 830 }, { "grad_norm": 1.1720198392868042, "learning_rate": 5.5933333333333335e-05, "loss": 0.1665, "step": 840 }, { "grad_norm": 0.849250316619873, "learning_rate": 5.66e-05, "loss": 0.1652, "step": 850 }, { "grad_norm": 0.9155775904655457, "learning_rate": 5.726666666666667e-05, "loss": 0.1528, "step": 860 }, { "grad_norm": 1.171196699142456, "learning_rate": 5.7933333333333337e-05, "loss": 0.1722, "step": 870 }, { "grad_norm": 0.9858519434928894, "learning_rate": 5.86e-05, "loss": 0.1412, "step": 880 }, { "grad_norm": 1.081385850906372, "learning_rate": 5.926666666666667e-05, "loss": 0.1561, "step": 890 }, { "grad_norm": 0.8932467699050903, "learning_rate": 5.9933333333333345e-05, "loss": 0.1579, "step": 900 }, { "grad_norm": 0.8548662662506104, "learning_rate": 6.06e-05, "loss": 0.1505, "step": 910 }, { "grad_norm": 0.7572005391120911, "learning_rate": 6.126666666666667e-05, "loss": 0.1391, "step": 920 }, { "grad_norm": 0.9849821925163269, "learning_rate": 6.193333333333333e-05, "loss": 0.1444, "step": 930 }, { "grad_norm": 0.9562931656837463, "learning_rate": 6.26e-05, "loss": 0.1566, "step": 940 }, { "grad_norm": 1.2492797374725342, "learning_rate": 6.326666666666667e-05, "loss": 0.1419, "step": 950 }, { "grad_norm": 0.9293448328971863, "learning_rate": 6.393333333333333e-05, "loss": 0.144, "step": 960 }, { "grad_norm": 0.858593761920929, "learning_rate": 6.460000000000001e-05, "loss": 0.1649, "step": 970 }, { "grad_norm": 0.8152620196342468, "learning_rate": 6.526666666666666e-05, "loss": 0.1397, "step": 980 }, { "grad_norm": 0.824725866317749, "learning_rate": 6.593333333333334e-05, "loss": 0.1409, "step": 990 }, { "grad_norm": 0.834763765335083, "learning_rate": 6.66e-05, "loss": 0.1463, "step": 1000 }, { "grad_norm": 0.7459161877632141, "learning_rate": 6.726666666666667e-05, "loss": 0.1465, "step": 1010 }, { "grad_norm": 0.9462399482727051, "learning_rate": 6.793333333333334e-05, "loss": 0.1646, "step": 1020 }, { "grad_norm": 0.9641590714454651, "learning_rate": 6.860000000000001e-05, "loss": 0.138, "step": 1030 }, { "grad_norm": 0.8934517502784729, "learning_rate": 6.926666666666667e-05, "loss": 0.144, "step": 1040 }, { "grad_norm": 0.8419318199157715, "learning_rate": 6.993333333333334e-05, "loss": 0.1457, "step": 1050 }, { "grad_norm": 0.7819538116455078, "learning_rate": 7.06e-05, "loss": 0.1535, "step": 1060 }, { "grad_norm": 0.8466137051582336, "learning_rate": 7.126666666666667e-05, "loss": 0.151, "step": 1070 }, { "grad_norm": 1.0024020671844482, "learning_rate": 7.193333333333334e-05, "loss": 0.1749, "step": 1080 }, { "grad_norm": 0.8885055780410767, "learning_rate": 7.26e-05, "loss": 0.1457, "step": 1090 }, { "grad_norm": 0.8589239716529846, "learning_rate": 7.326666666666667e-05, "loss": 0.1313, "step": 1100 }, { "grad_norm": 0.8235303163528442, "learning_rate": 7.393333333333333e-05, "loss": 0.1657, "step": 1110 }, { "grad_norm": 0.8818825483322144, "learning_rate": 7.46e-05, "loss": 0.1293, "step": 1120 }, { "grad_norm": 0.7876054644584656, "learning_rate": 7.526666666666668e-05, "loss": 0.1478, "step": 1130 }, { "grad_norm": 0.7659526467323303, "learning_rate": 7.593333333333334e-05, "loss": 0.1765, "step": 1140 }, { "grad_norm": 0.7604926228523254, "learning_rate": 7.66e-05, "loss": 0.1656, "step": 1150 }, { "grad_norm": 0.8470979332923889, "learning_rate": 7.726666666666667e-05, "loss": 0.1483, "step": 1160 }, { "grad_norm": 0.6591774225234985, "learning_rate": 7.793333333333333e-05, "loss": 0.1422, "step": 1170 }, { "grad_norm": 0.7418103218078613, "learning_rate": 7.860000000000001e-05, "loss": 0.1253, "step": 1180 }, { "grad_norm": 0.8148030042648315, "learning_rate": 7.926666666666666e-05, "loss": 0.1569, "step": 1190 }, { "grad_norm": 1.147688865661621, "learning_rate": 7.993333333333334e-05, "loss": 0.1548, "step": 1200 }, { "grad_norm": 1.0288184881210327, "learning_rate": 8.060000000000001e-05, "loss": 0.1759, "step": 1210 }, { "grad_norm": 0.7353565096855164, "learning_rate": 8.126666666666667e-05, "loss": 0.1465, "step": 1220 }, { "grad_norm": 0.7904092669487, "learning_rate": 8.193333333333334e-05, "loss": 0.1369, "step": 1230 }, { "grad_norm": 0.6588637232780457, "learning_rate": 8.26e-05, "loss": 0.1438, "step": 1240 }, { "grad_norm": 0.9510173201560974, "learning_rate": 8.326666666666667e-05, "loss": 0.1328, "step": 1250 }, { "grad_norm": 0.7137603759765625, "learning_rate": 8.393333333333335e-05, "loss": 0.1355, "step": 1260 }, { "grad_norm": 0.5818139314651489, "learning_rate": 8.46e-05, "loss": 0.1161, "step": 1270 }, { "grad_norm": 0.9060401320457458, "learning_rate": 8.526666666666667e-05, "loss": 0.1396, "step": 1280 }, { "grad_norm": 0.7194551825523376, "learning_rate": 8.593333333333333e-05, "loss": 0.1404, "step": 1290 }, { "grad_norm": 0.5931652784347534, "learning_rate": 8.66e-05, "loss": 0.1346, "step": 1300 }, { "grad_norm": 0.7356128692626953, "learning_rate": 8.726666666666667e-05, "loss": 0.1328, "step": 1310 }, { "grad_norm": 0.6496874094009399, "learning_rate": 8.793333333333333e-05, "loss": 0.1323, "step": 1320 }, { "grad_norm": 0.7630678415298462, "learning_rate": 8.86e-05, "loss": 0.121, "step": 1330 }, { "grad_norm": 0.6775751113891602, "learning_rate": 8.926666666666668e-05, "loss": 0.1141, "step": 1340 }, { "grad_norm": 0.6647273302078247, "learning_rate": 8.993333333333334e-05, "loss": 0.1472, "step": 1350 }, { "grad_norm": 0.7508830428123474, "learning_rate": 9.06e-05, "loss": 0.1329, "step": 1360 }, { "grad_norm": 0.6608383655548096, "learning_rate": 9.126666666666667e-05, "loss": 0.122, "step": 1370 }, { "grad_norm": 0.6588468551635742, "learning_rate": 9.193333333333334e-05, "loss": 0.1148, "step": 1380 }, { "grad_norm": 0.5878859162330627, "learning_rate": 9.260000000000001e-05, "loss": 0.1255, "step": 1390 }, { "grad_norm": 0.5673328042030334, "learning_rate": 9.326666666666667e-05, "loss": 0.1165, "step": 1400 }, { "grad_norm": 0.7046443223953247, "learning_rate": 9.393333333333334e-05, "loss": 0.1376, "step": 1410 }, { "grad_norm": 0.6791790127754211, "learning_rate": 9.46e-05, "loss": 0.1376, "step": 1420 }, { "grad_norm": 0.5826202630996704, "learning_rate": 9.526666666666667e-05, "loss": 0.1105, "step": 1430 }, { "grad_norm": 0.6816211938858032, "learning_rate": 9.593333333333334e-05, "loss": 0.12, "step": 1440 }, { "grad_norm": 0.7346307635307312, "learning_rate": 9.66e-05, "loss": 0.1457, "step": 1450 }, { "grad_norm": 0.7110230922698975, "learning_rate": 9.726666666666667e-05, "loss": 0.1356, "step": 1460 }, { "grad_norm": 0.6250454187393188, "learning_rate": 9.793333333333333e-05, "loss": 0.1378, "step": 1470 }, { "grad_norm": 0.7070907354354858, "learning_rate": 9.86e-05, "loss": 0.1106, "step": 1480 }, { "grad_norm": 0.541174590587616, "learning_rate": 9.926666666666668e-05, "loss": 0.1152, "step": 1490 }, { "grad_norm": 0.46901485323905945, "learning_rate": 9.993333333333334e-05, "loss": 0.125, "step": 1500 }, { "grad_norm": 0.43468305468559265, "learning_rate": 9.999997539434007e-05, "loss": 0.1087, "step": 1510 }, { "grad_norm": 0.7124031782150269, "learning_rate": 9.999989033776898e-05, "loss": 0.1236, "step": 1520 }, { "grad_norm": 0.5573038458824158, "learning_rate": 9.999974452661641e-05, "loss": 0.132, "step": 1530 }, { "grad_norm": 0.5823051929473877, "learning_rate": 9.999953796105959e-05, "loss": 0.1186, "step": 1540 }, { "grad_norm": 0.6794359683990479, "learning_rate": 9.999927064134949e-05, "loss": 0.1214, "step": 1550 }, { "grad_norm": 0.725846529006958, "learning_rate": 9.999894256781095e-05, "loss": 0.1408, "step": 1560 }, { "grad_norm": 0.598950982093811, "learning_rate": 9.99985537408426e-05, "loss": 0.1115, "step": 1570 }, { "grad_norm": 0.7087956666946411, "learning_rate": 9.999810416091688e-05, "loss": 0.1331, "step": 1580 }, { "grad_norm": 0.6899130344390869, "learning_rate": 9.99975938285801e-05, "loss": 0.1427, "step": 1590 }, { "grad_norm": 0.599158525466919, "learning_rate": 9.999702274445236e-05, "loss": 0.1306, "step": 1600 }, { "grad_norm": 0.7044597268104553, "learning_rate": 9.999639090922756e-05, "loss": 0.1057, "step": 1610 }, { "grad_norm": 0.6268491744995117, "learning_rate": 9.999569832367346e-05, "loss": 0.1255, "step": 1620 }, { "grad_norm": 0.5614187717437744, "learning_rate": 9.999494498863162e-05, "loss": 0.1545, "step": 1630 }, { "grad_norm": 0.5501335263252258, "learning_rate": 9.99941309050174e-05, "loss": 0.1217, "step": 1640 }, { "grad_norm": 0.5499849319458008, "learning_rate": 9.999325607381999e-05, "loss": 0.1267, "step": 1650 }, { "grad_norm": 0.44044461846351624, "learning_rate": 9.999232049610238e-05, "loss": 0.126, "step": 1660 }, { "grad_norm": 0.49347344040870667, "learning_rate": 9.999132417300142e-05, "loss": 0.1058, "step": 1670 }, { "grad_norm": 0.6333240866661072, "learning_rate": 9.99902671057277e-05, "loss": 0.1163, "step": 1680 }, { "grad_norm": 0.5511918663978577, "learning_rate": 9.998914929556569e-05, "loss": 0.1168, "step": 1690 }, { "grad_norm": 0.574537992477417, "learning_rate": 9.998797074387361e-05, "loss": 0.1187, "step": 1700 }, { "grad_norm": 0.6028810143470764, "learning_rate": 9.99867314520835e-05, "loss": 0.1316, "step": 1710 }, { "grad_norm": 0.4513114392757416, "learning_rate": 9.998543142170126e-05, "loss": 0.1211, "step": 1720 }, { "grad_norm": 0.4847932457923889, "learning_rate": 9.99840706543065e-05, "loss": 0.1147, "step": 1730 }, { "grad_norm": 0.4815286695957184, "learning_rate": 9.998264915155274e-05, "loss": 0.11, "step": 1740 }, { "grad_norm": 0.6546589136123657, "learning_rate": 9.998116691516718e-05, "loss": 0.1305, "step": 1750 }, { "grad_norm": 0.6325458884239197, "learning_rate": 9.997962394695091e-05, "loss": 0.1249, "step": 1760 }, { "grad_norm": 0.5429052114486694, "learning_rate": 9.997802024877875e-05, "loss": 0.1129, "step": 1770 }, { "grad_norm": 0.5844834446907043, "learning_rate": 9.99763558225994e-05, "loss": 0.1456, "step": 1780 }, { "grad_norm": 0.650963306427002, "learning_rate": 9.997463067043526e-05, "loss": 0.1163, "step": 1790 }, { "grad_norm": 0.5388178825378418, "learning_rate": 9.997284479438253e-05, "loss": 0.1189, "step": 1800 }, { "grad_norm": 0.43301138281822205, "learning_rate": 9.997099819661127e-05, "loss": 0.1067, "step": 1810 }, { "grad_norm": 0.5625163912773132, "learning_rate": 9.996909087936524e-05, "loss": 0.1218, "step": 1820 }, { "grad_norm": 0.6838710904121399, "learning_rate": 9.996712284496201e-05, "loss": 0.1181, "step": 1830 }, { "grad_norm": 0.7385373711585999, "learning_rate": 9.996509409579293e-05, "loss": 0.1291, "step": 1840 }, { "grad_norm": 0.6341670155525208, "learning_rate": 9.996300463432312e-05, "loss": 0.1467, "step": 1850 }, { "grad_norm": 0.6742323637008667, "learning_rate": 9.996085446309148e-05, "loss": 0.1402, "step": 1860 }, { "grad_norm": 0.3937931954860687, "learning_rate": 9.995864358471066e-05, "loss": 0.1098, "step": 1870 }, { "grad_norm": 0.5168664455413818, "learning_rate": 9.99563720018671e-05, "loss": 0.0913, "step": 1880 }, { "grad_norm": 0.6387127041816711, "learning_rate": 9.995403971732098e-05, "loss": 0.1066, "step": 1890 }, { "grad_norm": 0.5926902890205383, "learning_rate": 9.995164673390625e-05, "loss": 0.1004, "step": 1900 }, { "grad_norm": 0.568058431148529, "learning_rate": 9.994919305453059e-05, "loss": 0.1259, "step": 1910 }, { "grad_norm": 0.6348229050636292, "learning_rate": 9.994667868217548e-05, "loss": 0.1107, "step": 1920 }, { "grad_norm": 0.4053514301776886, "learning_rate": 9.99441036198961e-05, "loss": 0.1035, "step": 1930 }, { "grad_norm": 0.5806289911270142, "learning_rate": 9.99414678708214e-05, "loss": 0.111, "step": 1940 }, { "grad_norm": 0.521151065826416, "learning_rate": 9.993877143815407e-05, "loss": 0.0986, "step": 1950 }, { "grad_norm": 0.6184947490692139, "learning_rate": 9.993601432517053e-05, "loss": 0.1161, "step": 1960 }, { "grad_norm": 0.5990853905677795, "learning_rate": 9.993319653522091e-05, "loss": 0.0941, "step": 1970 }, { "grad_norm": 0.5944622159004211, "learning_rate": 9.993031807172911e-05, "loss": 0.1273, "step": 1980 }, { "grad_norm": 0.5887544751167297, "learning_rate": 9.992737893819273e-05, "loss": 0.1019, "step": 1990 }, { "grad_norm": 0.49060699343681335, "learning_rate": 9.992437913818312e-05, "loss": 0.126, "step": 2000 }, { "grad_norm": 0.38806021213531494, "learning_rate": 9.992131867534526e-05, "loss": 0.1398, "step": 2010 }, { "grad_norm": 0.6121862530708313, "learning_rate": 9.991819755339796e-05, "loss": 0.1115, "step": 2020 }, { "grad_norm": 0.45789533853530884, "learning_rate": 9.991501577613365e-05, "loss": 0.0937, "step": 2030 }, { "grad_norm": 0.4618423283100128, "learning_rate": 9.99117733474185e-05, "loss": 0.1158, "step": 2040 }, { "grad_norm": 0.5058461427688599, "learning_rate": 9.990847027119234e-05, "loss": 0.1087, "step": 2050 }, { "grad_norm": 0.5403944253921509, "learning_rate": 9.990510655146877e-05, "loss": 0.1089, "step": 2060 }, { "grad_norm": 0.6129620671272278, "learning_rate": 9.990168219233496e-05, "loss": 0.101, "step": 2070 }, { "grad_norm": 0.49735042452812195, "learning_rate": 9.989819719795188e-05, "loss": 0.0979, "step": 2080 }, { "grad_norm": 0.5445982217788696, "learning_rate": 9.989465157255412e-05, "loss": 0.1028, "step": 2090 }, { "grad_norm": 0.44239258766174316, "learning_rate": 9.989104532044994e-05, "loss": 0.1015, "step": 2100 }, { "grad_norm": 0.403062641620636, "learning_rate": 9.988737844602128e-05, "loss": 0.1296, "step": 2110 }, { "grad_norm": 0.4222145080566406, "learning_rate": 9.988365095372372e-05, "loss": 0.0996, "step": 2120 }, { "grad_norm": 0.4066031575202942, "learning_rate": 9.987986284808654e-05, "loss": 0.1031, "step": 2130 }, { "grad_norm": 0.5209201574325562, "learning_rate": 9.987601413371264e-05, "loss": 0.0835, "step": 2140 }, { "grad_norm": 0.37327784299850464, "learning_rate": 9.987210481527855e-05, "loss": 0.1194, "step": 2150 }, { "grad_norm": 0.36498987674713135, "learning_rate": 9.98681348975345e-05, "loss": 0.0953, "step": 2160 }, { "grad_norm": 0.4885857105255127, "learning_rate": 9.986410438530427e-05, "loss": 0.0983, "step": 2170 }, { "grad_norm": 0.3909528851509094, "learning_rate": 9.986001328348534e-05, "loss": 0.1147, "step": 2180 }, { "grad_norm": 0.4662936329841614, "learning_rate": 9.985586159704878e-05, "loss": 0.0994, "step": 2190 }, { "grad_norm": 0.5220146775245667, "learning_rate": 9.985164933103929e-05, "loss": 0.0929, "step": 2200 }, { "grad_norm": 0.43521198630332947, "learning_rate": 9.984737649057513e-05, "loss": 0.1034, "step": 2210 }, { "grad_norm": 0.49455365538597107, "learning_rate": 9.984304308084827e-05, "loss": 0.0955, "step": 2220 }, { "grad_norm": 0.3833504915237427, "learning_rate": 9.983864910712416e-05, "loss": 0.0991, "step": 2230 }, { "grad_norm": 0.5373802781105042, "learning_rate": 9.98341945747419e-05, "loss": 0.1002, "step": 2240 }, { "grad_norm": 0.5095442533493042, "learning_rate": 9.98296794891142e-05, "loss": 0.1142, "step": 2250 }, { "grad_norm": 0.3834283947944641, "learning_rate": 9.982510385572725e-05, "loss": 0.0883, "step": 2260 }, { "grad_norm": 0.44791552424430847, "learning_rate": 9.982046768014094e-05, "loss": 0.0911, "step": 2270 }, { "grad_norm": 0.47189849615097046, "learning_rate": 9.981577096798863e-05, "loss": 0.0856, "step": 2280 }, { "grad_norm": 0.5494545102119446, "learning_rate": 9.981101372497727e-05, "loss": 0.1238, "step": 2290 }, { "grad_norm": 0.4184468686580658, "learning_rate": 9.980619595688737e-05, "loss": 0.0866, "step": 2300 }, { "grad_norm": 0.5321471691131592, "learning_rate": 9.980131766957295e-05, "loss": 0.1032, "step": 2310 }, { "grad_norm": 0.358763724565506, "learning_rate": 9.979637886896163e-05, "loss": 0.0858, "step": 2320 }, { "grad_norm": 0.4319607615470886, "learning_rate": 9.979137956105447e-05, "loss": 0.1104, "step": 2330 }, { "grad_norm": 0.41809988021850586, "learning_rate": 9.978631975192613e-05, "loss": 0.0977, "step": 2340 }, { "grad_norm": 0.4701225459575653, "learning_rate": 9.978119944772475e-05, "loss": 0.0986, "step": 2350 }, { "grad_norm": 0.3694179356098175, "learning_rate": 9.977601865467197e-05, "loss": 0.1003, "step": 2360 }, { "grad_norm": 0.4124026298522949, "learning_rate": 9.977077737906297e-05, "loss": 0.1002, "step": 2370 }, { "grad_norm": 0.4824465215206146, "learning_rate": 9.976547562726636e-05, "loss": 0.1003, "step": 2380 }, { "grad_norm": 0.4390423595905304, "learning_rate": 9.976011340572429e-05, "loss": 0.117, "step": 2390 }, { "grad_norm": 0.42644989490509033, "learning_rate": 9.975469072095237e-05, "loss": 0.0992, "step": 2400 }, { "grad_norm": 0.4766586720943451, "learning_rate": 9.974920757953965e-05, "loss": 0.0902, "step": 2410 }, { "grad_norm": 0.49637940526008606, "learning_rate": 9.97436639881487e-05, "loss": 0.1299, "step": 2420 }, { "grad_norm": 0.39479148387908936, "learning_rate": 9.973805995351545e-05, "loss": 0.1039, "step": 2430 }, { "grad_norm": 0.41260361671447754, "learning_rate": 9.973239548244939e-05, "loss": 0.1082, "step": 2440 }, { "grad_norm": 0.50091952085495, "learning_rate": 9.972667058183333e-05, "loss": 0.0938, "step": 2450 }, { "grad_norm": 0.3772203028202057, "learning_rate": 9.972088525862362e-05, "loss": 0.0894, "step": 2460 }, { "grad_norm": 0.41473454236984253, "learning_rate": 9.971503951984995e-05, "loss": 0.0972, "step": 2470 }, { "grad_norm": 0.48513537645339966, "learning_rate": 9.970913337261543e-05, "loss": 0.1148, "step": 2480 }, { "grad_norm": 0.42165645956993103, "learning_rate": 9.97031668240966e-05, "loss": 0.0958, "step": 2490 }, { "grad_norm": 0.49460962414741516, "learning_rate": 9.969713988154339e-05, "loss": 0.0924, "step": 2500 }, { "grad_norm": 0.41141465306282043, "learning_rate": 9.969105255227906e-05, "loss": 0.0846, "step": 2510 }, { "grad_norm": 0.3961196541786194, "learning_rate": 9.968490484370035e-05, "loss": 0.0951, "step": 2520 }, { "grad_norm": 0.4131496250629425, "learning_rate": 9.967869676327726e-05, "loss": 0.078, "step": 2530 }, { "grad_norm": 0.5037717223167419, "learning_rate": 9.96724283185532e-05, "loss": 0.0893, "step": 2540 }, { "grad_norm": 0.3475835621356964, "learning_rate": 9.966609951714494e-05, "loss": 0.1065, "step": 2550 }, { "grad_norm": 0.37139299511909485, "learning_rate": 9.965971036674255e-05, "loss": 0.1043, "step": 2560 }, { "grad_norm": 0.4540638029575348, "learning_rate": 9.965326087510947e-05, "loss": 0.0989, "step": 2570 }, { "grad_norm": 0.5618516802787781, "learning_rate": 9.964675105008244e-05, "loss": 0.1035, "step": 2580 }, { "grad_norm": 0.49820294976234436, "learning_rate": 9.964018089957147e-05, "loss": 0.0875, "step": 2590 }, { "grad_norm": 0.5172275900840759, "learning_rate": 9.963355043155996e-05, "loss": 0.107, "step": 2600 }, { "grad_norm": 0.5463851094245911, "learning_rate": 9.962685965410455e-05, "loss": 0.0989, "step": 2610 }, { "grad_norm": 0.44594430923461914, "learning_rate": 9.962010857533514e-05, "loss": 0.0856, "step": 2620 }, { "grad_norm": 0.5318399667739868, "learning_rate": 9.961329720345493e-05, "loss": 0.0916, "step": 2630 }, { "grad_norm": 0.40706491470336914, "learning_rate": 9.96064255467404e-05, "loss": 0.1134, "step": 2640 }, { "grad_norm": 0.4159218668937683, "learning_rate": 9.959949361354126e-05, "loss": 0.1084, "step": 2650 }, { "grad_norm": 0.5552453398704529, "learning_rate": 9.959250141228045e-05, "loss": 0.1167, "step": 2660 }, { "grad_norm": 0.5203363299369812, "learning_rate": 9.958544895145414e-05, "loss": 0.1168, "step": 2670 }, { "grad_norm": 0.47239911556243896, "learning_rate": 9.957833623963177e-05, "loss": 0.1268, "step": 2680 }, { "grad_norm": 0.33993610739707947, "learning_rate": 9.957116328545593e-05, "loss": 0.0983, "step": 2690 }, { "grad_norm": 0.4728415608406067, "learning_rate": 9.956393009764244e-05, "loss": 0.0821, "step": 2700 }, { "grad_norm": 0.37891140580177307, "learning_rate": 9.955663668498032e-05, "loss": 0.098, "step": 2710 }, { "grad_norm": 0.4417562484741211, "learning_rate": 9.954928305633173e-05, "loss": 0.1134, "step": 2720 }, { "grad_norm": 0.3930230736732483, "learning_rate": 9.954186922063204e-05, "loss": 0.0938, "step": 2730 }, { "grad_norm": 0.36004993319511414, "learning_rate": 9.953439518688974e-05, "loss": 0.0893, "step": 2740 }, { "grad_norm": 0.3530312478542328, "learning_rate": 9.952686096418652e-05, "loss": 0.0972, "step": 2750 }, { "grad_norm": 0.5520556569099426, "learning_rate": 9.951926656167714e-05, "loss": 0.0903, "step": 2760 }, { "grad_norm": 0.4239886403083801, "learning_rate": 9.951161198858953e-05, "loss": 0.0956, "step": 2770 }, { "grad_norm": 0.35769134759902954, "learning_rate": 9.950389725422471e-05, "loss": 0.0962, "step": 2780 }, { "grad_norm": 0.45160043239593506, "learning_rate": 9.949612236795682e-05, "loss": 0.0994, "step": 2790 }, { "grad_norm": 0.4155345559120178, "learning_rate": 9.948828733923305e-05, "loss": 0.0932, "step": 2800 }, { "grad_norm": 0.46791842579841614, "learning_rate": 9.948039217757374e-05, "loss": 0.1243, "step": 2810 }, { "grad_norm": 0.4612118899822235, "learning_rate": 9.947243689257225e-05, "loss": 0.0978, "step": 2820 }, { "grad_norm": 0.398109495639801, "learning_rate": 9.946442149389497e-05, "loss": 0.0863, "step": 2830 }, { "grad_norm": 0.5814754962921143, "learning_rate": 9.945634599128139e-05, "loss": 0.1136, "step": 2840 }, { "grad_norm": 0.46510931849479675, "learning_rate": 9.944821039454402e-05, "loss": 0.1119, "step": 2850 }, { "grad_norm": 0.4345492124557495, "learning_rate": 9.944001471356835e-05, "loss": 0.0982, "step": 2860 }, { "grad_norm": 0.391938716173172, "learning_rate": 9.94317589583129e-05, "loss": 0.0901, "step": 2870 }, { "grad_norm": 0.43730536103248596, "learning_rate": 9.942344313880922e-05, "loss": 0.089, "step": 2880 }, { "grad_norm": 0.4900282919406891, "learning_rate": 9.941506726516179e-05, "loss": 0.0984, "step": 2890 }, { "grad_norm": 0.45530518889427185, "learning_rate": 9.94066313475481e-05, "loss": 0.1065, "step": 2900 }, { "grad_norm": 0.4501086473464966, "learning_rate": 9.939813539621857e-05, "loss": 0.0916, "step": 2910 }, { "grad_norm": 0.3313567340373993, "learning_rate": 9.93895794214966e-05, "loss": 0.0912, "step": 2920 }, { "grad_norm": 0.32529062032699585, "learning_rate": 9.938096343377852e-05, "loss": 0.0761, "step": 2930 }, { "grad_norm": 0.3651159703731537, "learning_rate": 9.937228744353353e-05, "loss": 0.0731, "step": 2940 }, { "grad_norm": 0.40185627341270447, "learning_rate": 9.936355146130379e-05, "loss": 0.0799, "step": 2950 }, { "grad_norm": 0.38098153471946716, "learning_rate": 9.935475549770435e-05, "loss": 0.0846, "step": 2960 }, { "grad_norm": 0.4270680248737335, "learning_rate": 9.934589956342315e-05, "loss": 0.0842, "step": 2970 }, { "grad_norm": 0.5225918889045715, "learning_rate": 9.933698366922093e-05, "loss": 0.0891, "step": 2980 }, { "grad_norm": 0.3626577854156494, "learning_rate": 9.93280078259314e-05, "loss": 0.0968, "step": 2990 }, { "grad_norm": 0.4718676507472992, "learning_rate": 9.931897204446104e-05, "loss": 0.0976, "step": 3000 }, { "grad_norm": 0.34026843309402466, "learning_rate": 9.930987633578915e-05, "loss": 0.0789, "step": 3010 }, { "grad_norm": 0.39102548360824585, "learning_rate": 9.93007207109679e-05, "loss": 0.0832, "step": 3020 }, { "grad_norm": 0.3950934410095215, "learning_rate": 9.929150518112224e-05, "loss": 0.0884, "step": 3030 }, { "grad_norm": 0.40357714891433716, "learning_rate": 9.928222975744991e-05, "loss": 0.0872, "step": 3040 }, { "grad_norm": 0.4253603518009186, "learning_rate": 9.92728944512214e-05, "loss": 0.0904, "step": 3050 }, { "grad_norm": 0.37781596183776855, "learning_rate": 9.926349927378001e-05, "loss": 0.0906, "step": 3060 }, { "grad_norm": 0.42643070220947266, "learning_rate": 9.925404423654174e-05, "loss": 0.0718, "step": 3070 }, { "grad_norm": 0.47125905752182007, "learning_rate": 9.924452935099537e-05, "loss": 0.0918, "step": 3080 }, { "grad_norm": 0.5146117210388184, "learning_rate": 9.92349546287024e-05, "loss": 0.1131, "step": 3090 }, { "grad_norm": 0.41173461079597473, "learning_rate": 9.9225320081297e-05, "loss": 0.0903, "step": 3100 }, { "grad_norm": 0.36617806553840637, "learning_rate": 9.921562572048606e-05, "loss": 0.0822, "step": 3110 }, { "grad_norm": 0.4469146430492401, "learning_rate": 9.920587155804913e-05, "loss": 0.0779, "step": 3120 }, { "grad_norm": 0.4347069263458252, "learning_rate": 9.919605760583845e-05, "loss": 0.0837, "step": 3130 }, { "grad_norm": 0.5212476849555969, "learning_rate": 9.91861838757789e-05, "loss": 0.086, "step": 3140 }, { "grad_norm": 0.37679511308670044, "learning_rate": 9.917625037986798e-05, "loss": 0.0768, "step": 3150 }, { "grad_norm": 0.5342316627502441, "learning_rate": 9.916625713017583e-05, "loss": 0.0892, "step": 3160 }, { "grad_norm": 0.4032861590385437, "learning_rate": 9.915620413884519e-05, "loss": 0.0767, "step": 3170 }, { "grad_norm": 0.4224550127983093, "learning_rate": 9.914609141809139e-05, "loss": 0.0998, "step": 3180 }, { "grad_norm": 0.39286476373672485, "learning_rate": 9.913591898020235e-05, "loss": 0.0966, "step": 3190 }, { "grad_norm": 0.40394288301467896, "learning_rate": 9.912568683753853e-05, "loss": 0.0881, "step": 3200 }, { "grad_norm": 0.3444503843784332, "learning_rate": 9.911539500253295e-05, "loss": 0.0775, "step": 3210 }, { "grad_norm": 0.41828569769859314, "learning_rate": 9.910504348769118e-05, "loss": 0.1134, "step": 3220 }, { "grad_norm": 0.405741810798645, "learning_rate": 9.909463230559127e-05, "loss": 0.0991, "step": 3230 }, { "grad_norm": 0.37217283248901367, "learning_rate": 9.908416146888376e-05, "loss": 0.0795, "step": 3240 }, { "grad_norm": 0.5055076479911804, "learning_rate": 9.907363099029175e-05, "loss": 0.0948, "step": 3250 }, { "grad_norm": 0.4393453896045685, "learning_rate": 9.906304088261073e-05, "loss": 0.0898, "step": 3260 }, { "grad_norm": 0.3707945942878723, "learning_rate": 9.905239115870872e-05, "loss": 0.091, "step": 3270 }, { "grad_norm": 0.4471019208431244, "learning_rate": 9.90416818315261e-05, "loss": 0.0814, "step": 3280 }, { "grad_norm": 0.4025392234325409, "learning_rate": 9.903091291407573e-05, "loss": 0.079, "step": 3290 }, { "grad_norm": 0.2657795548439026, "learning_rate": 9.902008441944286e-05, "loss": 0.0782, "step": 3300 }, { "grad_norm": 0.40615376830101013, "learning_rate": 9.900919636078512e-05, "loss": 0.0953, "step": 3310 }, { "grad_norm": 0.35301724076271057, "learning_rate": 9.899824875133255e-05, "loss": 0.1031, "step": 3320 }, { "grad_norm": 0.42564162611961365, "learning_rate": 9.898724160438749e-05, "loss": 0.0911, "step": 3330 }, { "grad_norm": 0.46213772892951965, "learning_rate": 9.89761749333247e-05, "loss": 0.0886, "step": 3340 }, { "grad_norm": 0.39841514825820923, "learning_rate": 9.896504875159122e-05, "loss": 0.0885, "step": 3350 }, { "grad_norm": 0.4602282643318176, "learning_rate": 9.89538630727064e-05, "loss": 0.0878, "step": 3360 }, { "grad_norm": 0.4212428629398346, "learning_rate": 9.894261791026189e-05, "loss": 0.0887, "step": 3370 }, { "grad_norm": 0.3691689670085907, "learning_rate": 9.893131327792165e-05, "loss": 0.1267, "step": 3380 }, { "grad_norm": 0.3867242932319641, "learning_rate": 9.891994918942182e-05, "loss": 0.0811, "step": 3390 }, { "grad_norm": 0.3816985487937927, "learning_rate": 9.890852565857092e-05, "loss": 0.0897, "step": 3400 }, { "grad_norm": 0.3411717712879181, "learning_rate": 9.889704269924954e-05, "loss": 0.0887, "step": 3410 }, { "grad_norm": 0.47457781434059143, "learning_rate": 9.888550032541059e-05, "loss": 0.0859, "step": 3420 }, { "grad_norm": 0.3714228868484497, "learning_rate": 9.887389855107916e-05, "loss": 0.0904, "step": 3430 }, { "grad_norm": 0.40049436688423157, "learning_rate": 9.886223739035248e-05, "loss": 0.1005, "step": 3440 }, { "grad_norm": 0.4535018503665924, "learning_rate": 9.885051685739997e-05, "loss": 0.0846, "step": 3450 }, { "grad_norm": 0.44700679183006287, "learning_rate": 9.883873696646316e-05, "loss": 0.0779, "step": 3460 }, { "grad_norm": 0.4428093135356903, "learning_rate": 9.882689773185575e-05, "loss": 0.0926, "step": 3470 }, { "grad_norm": 0.5374898910522461, "learning_rate": 9.881499916796353e-05, "loss": 0.098, "step": 3480 }, { "grad_norm": 0.38435623049736023, "learning_rate": 9.880304128924434e-05, "loss": 0.0839, "step": 3490 }, { "grad_norm": 0.40651312470436096, "learning_rate": 9.879102411022817e-05, "loss": 0.0767, "step": 3500 }, { "grad_norm": 0.35287466645240784, "learning_rate": 9.877894764551703e-05, "loss": 0.0835, "step": 3510 }, { "grad_norm": 0.40496355295181274, "learning_rate": 9.876681190978494e-05, "loss": 0.0835, "step": 3520 }, { "grad_norm": 0.40663811564445496, "learning_rate": 9.875461691777797e-05, "loss": 0.0835, "step": 3530 }, { "grad_norm": 0.37282705307006836, "learning_rate": 9.874236268431417e-05, "loss": 0.0844, "step": 3540 }, { "grad_norm": 0.36780479550361633, "learning_rate": 9.873004922428361e-05, "loss": 0.0843, "step": 3550 }, { "grad_norm": 0.4676786959171295, "learning_rate": 9.871767655264829e-05, "loss": 0.0711, "step": 3560 }, { "grad_norm": 0.3080819249153137, "learning_rate": 9.87052446844422e-05, "loss": 0.0878, "step": 3570 }, { "grad_norm": 0.4025127589702606, "learning_rate": 9.869275363477122e-05, "loss": 0.0974, "step": 3580 }, { "grad_norm": 0.4986024796962738, "learning_rate": 9.868020341881312e-05, "loss": 0.1015, "step": 3590 }, { "grad_norm": 0.4008992612361908, "learning_rate": 9.866759405181765e-05, "loss": 0.091, "step": 3600 }, { "grad_norm": 0.3436967432498932, "learning_rate": 9.865492554910633e-05, "loss": 0.0812, "step": 3610 }, { "grad_norm": 0.33228084444999695, "learning_rate": 9.864219792607262e-05, "loss": 0.0755, "step": 3620 }, { "grad_norm": 0.38205280900001526, "learning_rate": 9.862941119818177e-05, "loss": 0.0686, "step": 3630 }, { "grad_norm": 0.463189959526062, "learning_rate": 9.861656538097086e-05, "loss": 0.083, "step": 3640 }, { "grad_norm": 0.4077150523662567, "learning_rate": 9.860366049004877e-05, "loss": 0.0854, "step": 3650 }, { "grad_norm": 0.44698309898376465, "learning_rate": 9.859069654109615e-05, "loss": 0.0795, "step": 3660 }, { "grad_norm": 0.4371110200881958, "learning_rate": 9.857767354986545e-05, "loss": 0.0868, "step": 3670 }, { "grad_norm": 0.4124850630760193, "learning_rate": 9.856459153218078e-05, "loss": 0.0805, "step": 3680 }, { "grad_norm": 0.34927934408187866, "learning_rate": 9.855145050393808e-05, "loss": 0.0749, "step": 3690 }, { "grad_norm": 0.320297509431839, "learning_rate": 9.85382504811049e-05, "loss": 0.07, "step": 3700 }, { "grad_norm": 0.38276198506355286, "learning_rate": 9.852499147972054e-05, "loss": 0.0782, "step": 3710 }, { "grad_norm": 0.41985082626342773, "learning_rate": 9.851167351589592e-05, "loss": 0.0834, "step": 3720 }, { "grad_norm": 0.39600902795791626, "learning_rate": 9.849829660581363e-05, "loss": 0.0765, "step": 3730 }, { "grad_norm": 0.4088449478149414, "learning_rate": 9.848486076572787e-05, "loss": 0.0785, "step": 3740 }, { "grad_norm": 0.38219979405403137, "learning_rate": 9.847136601196446e-05, "loss": 0.0883, "step": 3750 }, { "grad_norm": 0.3892247974872589, "learning_rate": 9.845781236092078e-05, "loss": 0.0683, "step": 3760 }, { "grad_norm": 0.39979979395866394, "learning_rate": 9.844419982906583e-05, "loss": 0.0891, "step": 3770 }, { "grad_norm": 0.36324378848075867, "learning_rate": 9.843052843294008e-05, "loss": 0.0816, "step": 3780 }, { "grad_norm": 0.37790122628211975, "learning_rate": 9.841679818915559e-05, "loss": 0.0744, "step": 3790 }, { "grad_norm": 0.4262799620628357, "learning_rate": 9.840300911439591e-05, "loss": 0.1055, "step": 3800 }, { "grad_norm": 0.320901095867157, "learning_rate": 9.838916122541603e-05, "loss": 0.078, "step": 3810 }, { "grad_norm": 0.3061879575252533, "learning_rate": 9.837525453904246e-05, "loss": 0.0856, "step": 3820 }, { "grad_norm": 0.31549859046936035, "learning_rate": 9.836128907217314e-05, "loss": 0.0692, "step": 3830 }, { "grad_norm": 0.35183101892471313, "learning_rate": 9.834726484177743e-05, "loss": 0.0726, "step": 3840 }, { "grad_norm": 0.3567470908164978, "learning_rate": 9.833318186489609e-05, "loss": 0.0921, "step": 3850 }, { "grad_norm": 0.42611056566238403, "learning_rate": 9.831904015864126e-05, "loss": 0.0771, "step": 3860 }, { "grad_norm": 0.41698014736175537, "learning_rate": 9.830483974019645e-05, "loss": 0.0745, "step": 3870 }, { "grad_norm": 0.3562908172607422, "learning_rate": 9.82905806268165e-05, "loss": 0.0713, "step": 3880 }, { "grad_norm": 0.32529765367507935, "learning_rate": 9.82762628358276e-05, "loss": 0.0848, "step": 3890 }, { "grad_norm": 0.3613531291484833, "learning_rate": 9.826188638462718e-05, "loss": 0.0833, "step": 3900 }, { "grad_norm": 0.4655057191848755, "learning_rate": 9.824745129068402e-05, "loss": 0.0775, "step": 3910 }, { "grad_norm": 0.3808934986591339, "learning_rate": 9.82329575715381e-05, "loss": 0.0875, "step": 3920 }, { "grad_norm": 0.3568260669708252, "learning_rate": 9.821840524480066e-05, "loss": 0.0768, "step": 3930 }, { "grad_norm": 0.31418269872665405, "learning_rate": 9.820379432815414e-05, "loss": 0.0878, "step": 3940 }, { "grad_norm": 0.443363755941391, "learning_rate": 9.81891248393522e-05, "loss": 0.0882, "step": 3950 }, { "grad_norm": 0.44039997458457947, "learning_rate": 9.817439679621963e-05, "loss": 0.0911, "step": 3960 }, { "grad_norm": 0.4505123496055603, "learning_rate": 9.815961021665243e-05, "loss": 0.0708, "step": 3970 }, { "grad_norm": 0.3179085850715637, "learning_rate": 9.814476511861763e-05, "loss": 0.0883, "step": 3980 }, { "grad_norm": 0.40261784195899963, "learning_rate": 9.812986152015348e-05, "loss": 0.0902, "step": 3990 }, { "grad_norm": 0.4452002942562103, "learning_rate": 9.811489943936922e-05, "loss": 0.0801, "step": 4000 }, { "grad_norm": 0.3613608777523041, "learning_rate": 9.809987889444522e-05, "loss": 0.076, "step": 4010 }, { "grad_norm": 0.3730508089065552, "learning_rate": 9.808479990363282e-05, "loss": 0.0843, "step": 4020 }, { "grad_norm": 0.3806791305541992, "learning_rate": 9.806966248525445e-05, "loss": 0.0895, "step": 4030 }, { "grad_norm": 0.3366662263870239, "learning_rate": 9.805446665770348e-05, "loss": 0.0712, "step": 4040 }, { "grad_norm": 0.4037311375141144, "learning_rate": 9.803921243944429e-05, "loss": 0.0846, "step": 4050 }, { "grad_norm": 0.2969924211502075, "learning_rate": 9.802389984901218e-05, "loss": 0.0752, "step": 4060 }, { "grad_norm": 0.32146286964416504, "learning_rate": 9.80085289050134e-05, "loss": 0.0772, "step": 4070 }, { "grad_norm": 0.26851075887680054, "learning_rate": 9.799309962612508e-05, "loss": 0.088, "step": 4080 }, { "grad_norm": 0.3639536499977112, "learning_rate": 9.797761203109527e-05, "loss": 0.0779, "step": 4090 }, { "grad_norm": 0.3490807116031647, "learning_rate": 9.796206613874283e-05, "loss": 0.0701, "step": 4100 }, { "grad_norm": 0.373683899641037, "learning_rate": 9.794646196795754e-05, "loss": 0.0788, "step": 4110 }, { "grad_norm": 0.6021084785461426, "learning_rate": 9.793079953769987e-05, "loss": 0.0861, "step": 4120 }, { "grad_norm": 0.3044678568840027, "learning_rate": 9.79150788670012e-05, "loss": 0.0855, "step": 4130 }, { "grad_norm": 0.38936150074005127, "learning_rate": 9.78992999749636e-05, "loss": 0.0803, "step": 4140 }, { "grad_norm": 0.38170117139816284, "learning_rate": 9.788346288075994e-05, "loss": 0.0902, "step": 4150 }, { "grad_norm": 0.3732798993587494, "learning_rate": 9.786756760363373e-05, "loss": 0.0782, "step": 4160 }, { "grad_norm": 0.3583327829837799, "learning_rate": 9.78516141628993e-05, "loss": 0.078, "step": 4170 }, { "grad_norm": 0.35289281606674194, "learning_rate": 9.783560257794154e-05, "loss": 0.0737, "step": 4180 }, { "grad_norm": 0.35022667050361633, "learning_rate": 9.781953286821603e-05, "loss": 0.0795, "step": 4190 }, { "grad_norm": 0.36590373516082764, "learning_rate": 9.780340505324901e-05, "loss": 0.076, "step": 4200 }, { "grad_norm": 0.33465808629989624, "learning_rate": 9.778721915263727e-05, "loss": 0.0773, "step": 4210 }, { "grad_norm": 0.31979843974113464, "learning_rate": 9.777097518604824e-05, "loss": 0.0934, "step": 4220 }, { "grad_norm": 0.350436806678772, "learning_rate": 9.775467317321984e-05, "loss": 0.0928, "step": 4230 }, { "grad_norm": 0.32108908891677856, "learning_rate": 9.773831313396055e-05, "loss": 0.0897, "step": 4240 }, { "grad_norm": 0.37086448073387146, "learning_rate": 9.77218950881494e-05, "loss": 0.0846, "step": 4250 }, { "grad_norm": 0.3528153896331787, "learning_rate": 9.770541905573583e-05, "loss": 0.0838, "step": 4260 }, { "grad_norm": 0.3567950129508972, "learning_rate": 9.768888505673976e-05, "loss": 0.0786, "step": 4270 }, { "grad_norm": 0.3640773594379425, "learning_rate": 9.767229311125162e-05, "loss": 0.0694, "step": 4280 }, { "grad_norm": 0.4295765161514282, "learning_rate": 9.765564323943211e-05, "loss": 0.0766, "step": 4290 }, { "grad_norm": 0.36064642667770386, "learning_rate": 9.763893546151244e-05, "loss": 0.0654, "step": 4300 }, { "grad_norm": 0.37981274724006653, "learning_rate": 9.762216979779412e-05, "loss": 0.0803, "step": 4310 }, { "grad_norm": 0.4492497146129608, "learning_rate": 9.760534626864902e-05, "loss": 0.0998, "step": 4320 }, { "grad_norm": 0.43422701954841614, "learning_rate": 9.758846489451931e-05, "loss": 0.075, "step": 4330 }, { "grad_norm": 0.36232230067253113, "learning_rate": 9.757152569591748e-05, "loss": 0.0675, "step": 4340 }, { "grad_norm": 0.2780834436416626, "learning_rate": 9.75545286934262e-05, "loss": 0.0607, "step": 4350 }, { "grad_norm": 0.346218079328537, "learning_rate": 9.753747390769847e-05, "loss": 0.0713, "step": 4360 }, { "grad_norm": 0.39921265840530396, "learning_rate": 9.752036135945744e-05, "loss": 0.0779, "step": 4370 }, { "grad_norm": 0.3122653663158417, "learning_rate": 9.750319106949649e-05, "loss": 0.0804, "step": 4380 }, { "grad_norm": 0.39125514030456543, "learning_rate": 9.748596305867913e-05, "loss": 0.0685, "step": 4390 }, { "grad_norm": 0.39845985174179077, "learning_rate": 9.746867734793903e-05, "loss": 0.0887, "step": 4400 }, { "grad_norm": 0.39573806524276733, "learning_rate": 9.745133395827993e-05, "loss": 0.0816, "step": 4410 }, { "grad_norm": 0.43776023387908936, "learning_rate": 9.743393291077572e-05, "loss": 0.0947, "step": 4420 }, { "grad_norm": 0.4129426181316376, "learning_rate": 9.741647422657028e-05, "loss": 0.0846, "step": 4430 }, { "grad_norm": 0.29558053612709045, "learning_rate": 9.739895792687758e-05, "loss": 0.0814, "step": 4440 }, { "grad_norm": 0.3221525251865387, "learning_rate": 9.738138403298157e-05, "loss": 0.0884, "step": 4450 }, { "grad_norm": 0.3943980634212494, "learning_rate": 9.736375256623619e-05, "loss": 0.0757, "step": 4460 }, { "grad_norm": 0.34983253479003906, "learning_rate": 9.734606354806533e-05, "loss": 0.081, "step": 4470 }, { "grad_norm": 0.3389854431152344, "learning_rate": 9.73283169999628e-05, "loss": 0.0873, "step": 4480 }, { "grad_norm": 0.36907777190208435, "learning_rate": 9.731051294349238e-05, "loss": 0.0872, "step": 4490 }, { "grad_norm": 0.3549933135509491, "learning_rate": 9.729265140028762e-05, "loss": 0.0822, "step": 4500 }, { "grad_norm": 0.39877042174339294, "learning_rate": 9.727473239205201e-05, "loss": 0.105, "step": 4510 }, { "grad_norm": 0.3073529303073883, "learning_rate": 9.725675594055883e-05, "loss": 0.0772, "step": 4520 }, { "grad_norm": 0.33667460083961487, "learning_rate": 9.723872206765116e-05, "loss": 0.0788, "step": 4530 }, { "grad_norm": 0.4809662401676178, "learning_rate": 9.722063079524185e-05, "loss": 0.0945, "step": 4540 }, { "grad_norm": 0.42365628480911255, "learning_rate": 9.720248214531351e-05, "loss": 0.0917, "step": 4550 }, { "grad_norm": 0.3855436146259308, "learning_rate": 9.718427613991848e-05, "loss": 0.0739, "step": 4560 }, { "grad_norm": 0.38019341230392456, "learning_rate": 9.716601280117873e-05, "loss": 0.0911, "step": 4570 }, { "grad_norm": 0.39614251255989075, "learning_rate": 9.714769215128596e-05, "loss": 0.0846, "step": 4580 }, { "grad_norm": 0.3134308457374573, "learning_rate": 9.712931421250152e-05, "loss": 0.0771, "step": 4590 }, { "grad_norm": 0.3715668320655823, "learning_rate": 9.711087900715627e-05, "loss": 0.0793, "step": 4600 }, { "grad_norm": 0.34114009141921997, "learning_rate": 9.709238655765078e-05, "loss": 0.0796, "step": 4610 }, { "grad_norm": 0.34276294708251953, "learning_rate": 9.707383688645511e-05, "loss": 0.0688, "step": 4620 }, { "grad_norm": 0.4090898633003235, "learning_rate": 9.705523001610883e-05, "loss": 0.0817, "step": 4630 }, { "grad_norm": 0.3592730760574341, "learning_rate": 9.703656596922107e-05, "loss": 0.0761, "step": 4640 }, { "grad_norm": 0.3821835219860077, "learning_rate": 9.70178447684704e-05, "loss": 0.0688, "step": 4650 }, { "grad_norm": 0.4046807885169983, "learning_rate": 9.699906643660483e-05, "loss": 0.0818, "step": 4660 }, { "grad_norm": 0.4721336364746094, "learning_rate": 9.698023099644185e-05, "loss": 0.0747, "step": 4670 }, { "grad_norm": 0.4179655909538269, "learning_rate": 9.696133847086823e-05, "loss": 0.0794, "step": 4680 }, { "grad_norm": 0.3553186357021332, "learning_rate": 9.694238888284022e-05, "loss": 0.0991, "step": 4690 }, { "grad_norm": 0.3423902988433838, "learning_rate": 9.692338225538333e-05, "loss": 0.0799, "step": 4700 }, { "grad_norm": 0.41327622532844543, "learning_rate": 9.690431861159241e-05, "loss": 0.0872, "step": 4710 }, { "grad_norm": 0.40878844261169434, "learning_rate": 9.688519797463161e-05, "loss": 0.0738, "step": 4720 }, { "grad_norm": 0.34178587794303894, "learning_rate": 9.686602036773426e-05, "loss": 0.0934, "step": 4730 }, { "grad_norm": 0.3249698579311371, "learning_rate": 9.684678581420302e-05, "loss": 0.0818, "step": 4740 }, { "grad_norm": 0.30668920278549194, "learning_rate": 9.682749433740962e-05, "loss": 0.0709, "step": 4750 }, { "grad_norm": 0.3032350540161133, "learning_rate": 9.680814596079507e-05, "loss": 0.0781, "step": 4760 }, { "grad_norm": 0.3527970314025879, "learning_rate": 9.678874070786945e-05, "loss": 0.085, "step": 4770 }, { "grad_norm": 0.4545914828777313, "learning_rate": 9.676927860221199e-05, "loss": 0.0889, "step": 4780 }, { "grad_norm": 0.32243612408638, "learning_rate": 9.674975966747097e-05, "loss": 0.0734, "step": 4790 }, { "grad_norm": 0.3319143056869507, "learning_rate": 9.673018392736374e-05, "loss": 0.075, "step": 4800 }, { "grad_norm": 0.30873703956604004, "learning_rate": 9.671055140567667e-05, "loss": 0.0811, "step": 4810 }, { "grad_norm": 0.3133257031440735, "learning_rate": 9.669086212626511e-05, "loss": 0.0848, "step": 4820 }, { "grad_norm": 0.38964468240737915, "learning_rate": 9.667111611305341e-05, "loss": 0.0831, "step": 4830 }, { "grad_norm": 0.3691699802875519, "learning_rate": 9.665131339003486e-05, "loss": 0.0802, "step": 4840 }, { "grad_norm": 0.3550565540790558, "learning_rate": 9.663145398127158e-05, "loss": 0.0811, "step": 4850 }, { "grad_norm": 0.34546104073524475, "learning_rate": 9.661153791089467e-05, "loss": 0.0657, "step": 4860 }, { "grad_norm": 0.38052675127983093, "learning_rate": 9.659156520310402e-05, "loss": 0.0751, "step": 4870 }, { "grad_norm": 0.31614619493484497, "learning_rate": 9.657153588216835e-05, "loss": 0.0701, "step": 4880 }, { "grad_norm": 0.3901197910308838, "learning_rate": 9.655144997242516e-05, "loss": 0.0884, "step": 4890 }, { "grad_norm": 0.2958650588989258, "learning_rate": 9.653130749828075e-05, "loss": 0.0676, "step": 4900 }, { "grad_norm": 0.5070604681968689, "learning_rate": 9.65111084842101e-05, "loss": 0.079, "step": 4910 }, { "grad_norm": 0.3388635516166687, "learning_rate": 9.649085295475695e-05, "loss": 0.0811, "step": 4920 }, { "grad_norm": 0.373856782913208, "learning_rate": 9.647054093453365e-05, "loss": 0.0796, "step": 4930 }, { "grad_norm": 0.3570548892021179, "learning_rate": 9.645017244822123e-05, "loss": 0.0673, "step": 4940 }, { "grad_norm": 0.3147738575935364, "learning_rate": 9.642974752056931e-05, "loss": 0.0811, "step": 4950 }, { "grad_norm": 0.3751422166824341, "learning_rate": 9.640926617639613e-05, "loss": 0.0767, "step": 4960 }, { "grad_norm": 0.35979461669921875, "learning_rate": 9.638872844058843e-05, "loss": 0.0851, "step": 4970 }, { "grad_norm": 0.3602491319179535, "learning_rate": 9.63681343381015e-05, "loss": 0.0772, "step": 4980 }, { "grad_norm": 0.21682311594486237, "learning_rate": 9.634748389395914e-05, "loss": 0.0681, "step": 4990 }, { "grad_norm": 0.36068013310432434, "learning_rate": 9.632677713325353e-05, "loss": 0.0817, "step": 5000 }, { "grad_norm": 0.4300607144832611, "learning_rate": 9.63060140811454e-05, "loss": 0.0802, "step": 5010 }, { "grad_norm": 0.39744478464126587, "learning_rate": 9.628519476286379e-05, "loss": 0.0717, "step": 5020 }, { "grad_norm": 0.26179471611976624, "learning_rate": 9.626431920370612e-05, "loss": 0.0609, "step": 5030 }, { "grad_norm": 0.39819440245628357, "learning_rate": 9.624338742903819e-05, "loss": 0.071, "step": 5040 }, { "grad_norm": 0.32530081272125244, "learning_rate": 9.622239946429406e-05, "loss": 0.0837, "step": 5050 }, { "grad_norm": 0.2880935072898865, "learning_rate": 9.620135533497609e-05, "loss": 0.0616, "step": 5060 }, { "grad_norm": 0.29914334416389465, "learning_rate": 9.61802550666549e-05, "loss": 0.0982, "step": 5070 }, { "grad_norm": 0.3077535927295685, "learning_rate": 9.615909868496928e-05, "loss": 0.0862, "step": 5080 }, { "grad_norm": 0.3008662760257721, "learning_rate": 9.613788621562622e-05, "loss": 0.0677, "step": 5090 }, { "grad_norm": 0.37613266706466675, "learning_rate": 9.611661768440093e-05, "loss": 0.0706, "step": 5100 }, { "grad_norm": 0.37448614835739136, "learning_rate": 9.609529311713661e-05, "loss": 0.0834, "step": 5110 }, { "grad_norm": 0.30840733647346497, "learning_rate": 9.607391253974466e-05, "loss": 0.0906, "step": 5120 }, { "grad_norm": 0.32020097970962524, "learning_rate": 9.605247597820448e-05, "loss": 0.0806, "step": 5130 }, { "grad_norm": 0.3742756247520447, "learning_rate": 9.603098345856354e-05, "loss": 0.0823, "step": 5140 }, { "grad_norm": 0.34984713792800903, "learning_rate": 9.600943500693725e-05, "loss": 0.0685, "step": 5150 }, { "grad_norm": 0.31098803877830505, "learning_rate": 9.598783064950902e-05, "loss": 0.0925, "step": 5160 }, { "grad_norm": 0.3337903916835785, "learning_rate": 9.596617041253018e-05, "loss": 0.0747, "step": 5170 }, { "grad_norm": 0.2817566394805908, "learning_rate": 9.594445432231996e-05, "loss": 0.0679, "step": 5180 }, { "grad_norm": 0.31515994668006897, "learning_rate": 9.592268240526547e-05, "loss": 0.0961, "step": 5190 }, { "grad_norm": 0.32990148663520813, "learning_rate": 9.590085468782162e-05, "loss": 0.0628, "step": 5200 }, { "grad_norm": 0.3763536810874939, "learning_rate": 9.587897119651116e-05, "loss": 0.0746, "step": 5210 }, { "grad_norm": 0.3864741623401642, "learning_rate": 9.585703195792459e-05, "loss": 0.0715, "step": 5220 }, { "grad_norm": 0.40956375002861023, "learning_rate": 9.583503699872016e-05, "loss": 0.0705, "step": 5230 }, { "grad_norm": 0.3632172644138336, "learning_rate": 9.581298634562381e-05, "loss": 0.0698, "step": 5240 }, { "grad_norm": 0.37006452679634094, "learning_rate": 9.579088002542917e-05, "loss": 0.0897, "step": 5250 }, { "grad_norm": 0.3521057069301605, "learning_rate": 9.57687180649975e-05, "loss": 0.069, "step": 5260 }, { "grad_norm": 0.3896416425704956, "learning_rate": 9.574650049125768e-05, "loss": 0.0713, "step": 5270 }, { "grad_norm": 0.34124037623405457, "learning_rate": 9.572422733120614e-05, "loss": 0.0737, "step": 5280 }, { "grad_norm": 0.2962722182273865, "learning_rate": 9.570189861190689e-05, "loss": 0.0742, "step": 5290 }, { "grad_norm": 0.3290007710456848, "learning_rate": 9.56795143604914e-05, "loss": 0.0651, "step": 5300 }, { "grad_norm": 0.2914034426212311, "learning_rate": 9.565707460415869e-05, "loss": 0.0578, "step": 5310 }, { "grad_norm": 0.27696359157562256, "learning_rate": 9.563457937017515e-05, "loss": 0.0739, "step": 5320 }, { "grad_norm": 0.4333876073360443, "learning_rate": 9.56120286858746e-05, "loss": 0.0865, "step": 5330 }, { "grad_norm": 0.37735089659690857, "learning_rate": 9.558942257865829e-05, "loss": 0.0712, "step": 5340 }, { "grad_norm": 0.2352084219455719, "learning_rate": 9.556676107599472e-05, "loss": 0.0631, "step": 5350 }, { "grad_norm": 0.435627281665802, "learning_rate": 9.554404420541978e-05, "loss": 0.0716, "step": 5360 }, { "grad_norm": 0.3229515254497528, "learning_rate": 9.55212719945366e-05, "loss": 0.0584, "step": 5370 }, { "grad_norm": 0.3254145085811615, "learning_rate": 9.549844447101559e-05, "loss": 0.0613, "step": 5380 }, { "grad_norm": 0.2995050847530365, "learning_rate": 9.547556166259433e-05, "loss": 0.0666, "step": 5390 }, { "grad_norm": 0.35769715905189514, "learning_rate": 9.545262359707756e-05, "loss": 0.0756, "step": 5400 }, { "grad_norm": 0.2691922187805176, "learning_rate": 9.542963030233724e-05, "loss": 0.0667, "step": 5410 }, { "grad_norm": 0.37167802453041077, "learning_rate": 9.540658180631237e-05, "loss": 0.0684, "step": 5420 }, { "grad_norm": 0.3279150128364563, "learning_rate": 9.538347813700904e-05, "loss": 0.0662, "step": 5430 }, { "grad_norm": 0.30230098962783813, "learning_rate": 9.536031932250036e-05, "loss": 0.083, "step": 5440 }, { "grad_norm": 0.3736446499824524, "learning_rate": 9.533710539092653e-05, "loss": 0.0546, "step": 5450 }, { "grad_norm": 0.2874501049518585, "learning_rate": 9.531383637049464e-05, "loss": 0.0614, "step": 5460 }, { "grad_norm": 0.33796513080596924, "learning_rate": 9.529051228947875e-05, "loss": 0.0636, "step": 5470 }, { "grad_norm": 0.30049335956573486, "learning_rate": 9.52671331762198e-05, "loss": 0.0684, "step": 5480 }, { "grad_norm": 0.2800152599811554, "learning_rate": 9.524369905912565e-05, "loss": 0.0676, "step": 5490 }, { "grad_norm": 0.34990885853767395, "learning_rate": 9.522020996667092e-05, "loss": 0.0741, "step": 5500 }, { "grad_norm": 0.2770771384239197, "learning_rate": 9.519666592739709e-05, "loss": 0.0693, "step": 5510 }, { "grad_norm": 0.2811853289604187, "learning_rate": 9.517306696991241e-05, "loss": 0.0619, "step": 5520 }, { "grad_norm": 0.30597004294395447, "learning_rate": 9.51494131228918e-05, "loss": 0.0685, "step": 5530 }, { "grad_norm": 0.3272843360900879, "learning_rate": 9.512570441507695e-05, "loss": 0.0652, "step": 5540 }, { "grad_norm": 0.36188942193984985, "learning_rate": 9.510194087527615e-05, "loss": 0.0653, "step": 5550 }, { "grad_norm": 0.2715345025062561, "learning_rate": 9.507812253236435e-05, "loss": 0.0633, "step": 5560 }, { "grad_norm": 0.36622563004493713, "learning_rate": 9.505424941528309e-05, "loss": 0.0638, "step": 5570 }, { "grad_norm": 0.343437135219574, "learning_rate": 9.503032155304046e-05, "loss": 0.089, "step": 5580 }, { "grad_norm": 0.3444344103336334, "learning_rate": 9.500633897471106e-05, "loss": 0.062, "step": 5590 }, { "grad_norm": 0.3655736744403839, "learning_rate": 9.498230170943596e-05, "loss": 0.0756, "step": 5600 }, { "grad_norm": 0.32791855931282043, "learning_rate": 9.495820978642275e-05, "loss": 0.0676, "step": 5610 }, { "grad_norm": 0.35262367129325867, "learning_rate": 9.493406323494535e-05, "loss": 0.0634, "step": 5620 }, { "grad_norm": 0.2358086109161377, "learning_rate": 9.490986208434413e-05, "loss": 0.0553, "step": 5630 }, { "grad_norm": 0.4025478661060333, "learning_rate": 9.488560636402577e-05, "loss": 0.0614, "step": 5640 }, { "grad_norm": 0.4370748698711395, "learning_rate": 9.486129610346321e-05, "loss": 0.0818, "step": 5650 }, { "grad_norm": 0.34372904896736145, "learning_rate": 9.483693133219576e-05, "loss": 0.0664, "step": 5660 }, { "grad_norm": 0.3874727487564087, "learning_rate": 9.481251207982888e-05, "loss": 0.0674, "step": 5670 }, { "grad_norm": 0.3849292993545532, "learning_rate": 9.47880383760343e-05, "loss": 0.0712, "step": 5680 }, { "grad_norm": 0.37215951085090637, "learning_rate": 9.476351025054983e-05, "loss": 0.0728, "step": 5690 }, { "grad_norm": 0.34720635414123535, "learning_rate": 9.473892773317952e-05, "loss": 0.0631, "step": 5700 }, { "grad_norm": 0.29277053475379944, "learning_rate": 9.471429085379338e-05, "loss": 0.0778, "step": 5710 }, { "grad_norm": 0.34614256024360657, "learning_rate": 9.468959964232757e-05, "loss": 0.0635, "step": 5720 }, { "grad_norm": 0.32548248767852783, "learning_rate": 9.466485412878425e-05, "loss": 0.0678, "step": 5730 }, { "grad_norm": 0.30909767746925354, "learning_rate": 9.464005434323154e-05, "loss": 0.0765, "step": 5740 }, { "grad_norm": 0.319947212934494, "learning_rate": 9.461520031580352e-05, "loss": 0.0647, "step": 5750 }, { "grad_norm": 0.32429876923561096, "learning_rate": 9.459029207670019e-05, "loss": 0.0685, "step": 5760 }, { "grad_norm": 0.23448695242404938, "learning_rate": 9.456532965618737e-05, "loss": 0.0641, "step": 5770 }, { "grad_norm": 0.3355346620082855, "learning_rate": 9.454031308459681e-05, "loss": 0.0825, "step": 5780 }, { "grad_norm": 0.25151774287223816, "learning_rate": 9.451524239232595e-05, "loss": 0.0603, "step": 5790 }, { "grad_norm": 0.3149898052215576, "learning_rate": 9.449011760983809e-05, "loss": 0.0812, "step": 5800 }, { "grad_norm": 0.3684616982936859, "learning_rate": 9.446493876766218e-05, "loss": 0.0709, "step": 5810 }, { "grad_norm": 0.2988879680633545, "learning_rate": 9.44397058963929e-05, "loss": 0.0665, "step": 5820 }, { "grad_norm": 0.35528990626335144, "learning_rate": 9.441441902669056e-05, "loss": 0.0709, "step": 5830 }, { "grad_norm": 0.3166639804840088, "learning_rate": 9.43890781892811e-05, "loss": 0.0672, "step": 5840 }, { "grad_norm": 0.3343195617198944, "learning_rate": 9.436368341495603e-05, "loss": 0.0595, "step": 5850 }, { "grad_norm": 0.23363587260246277, "learning_rate": 9.43382347345724e-05, "loss": 0.0716, "step": 5860 }, { "grad_norm": 0.3526093065738678, "learning_rate": 9.431273217905273e-05, "loss": 0.0752, "step": 5870 }, { "grad_norm": 0.3657768964767456, "learning_rate": 9.428717577938504e-05, "loss": 0.08, "step": 5880 }, { "grad_norm": 0.3492029905319214, "learning_rate": 9.426156556662276e-05, "loss": 0.0671, "step": 5890 }, { "grad_norm": 0.33571958541870117, "learning_rate": 9.423590157188474e-05, "loss": 0.0846, "step": 5900 }, { "grad_norm": 0.22962675988674164, "learning_rate": 9.421018382635513e-05, "loss": 0.0699, "step": 5910 }, { "grad_norm": 0.34968316555023193, "learning_rate": 9.418441236128343e-05, "loss": 0.0637, "step": 5920 }, { "grad_norm": 0.2607312798500061, "learning_rate": 9.41585872079844e-05, "loss": 0.0605, "step": 5930 }, { "grad_norm": 0.2509673833847046, "learning_rate": 9.413270839783802e-05, "loss": 0.0659, "step": 5940 }, { "grad_norm": 0.27132976055145264, "learning_rate": 9.41067759622895e-05, "loss": 0.0575, "step": 5950 }, { "grad_norm": 0.2866787910461426, "learning_rate": 9.408078993284917e-05, "loss": 0.0595, "step": 5960 }, { "grad_norm": 0.3039873540401459, "learning_rate": 9.405475034109254e-05, "loss": 0.0645, "step": 5970 }, { "grad_norm": 0.41472241282463074, "learning_rate": 9.402865721866015e-05, "loss": 0.0668, "step": 5980 }, { "grad_norm": 0.2999403178691864, "learning_rate": 9.400251059725762e-05, "loss": 0.0791, "step": 5990 }, { "grad_norm": 0.37193819880485535, "learning_rate": 9.397631050865554e-05, "loss": 0.092, "step": 6000 }, { "grad_norm": 0.31223922967910767, "learning_rate": 9.395005698468949e-05, "loss": 0.0642, "step": 6010 }, { "grad_norm": 0.3581950068473816, "learning_rate": 9.392375005726e-05, "loss": 0.0635, "step": 6020 }, { "grad_norm": 0.24546028673648834, "learning_rate": 9.389738975833242e-05, "loss": 0.0675, "step": 6030 }, { "grad_norm": 0.3020484149456024, "learning_rate": 9.387097611993707e-05, "loss": 0.0619, "step": 6040 }, { "grad_norm": 0.39207983016967773, "learning_rate": 9.384450917416894e-05, "loss": 0.0714, "step": 6050 }, { "grad_norm": 0.35458531975746155, "learning_rate": 9.381798895318792e-05, "loss": 0.0608, "step": 6060 }, { "grad_norm": 0.35320064425468445, "learning_rate": 9.379141548921853e-05, "loss": 0.0679, "step": 6070 }, { "grad_norm": 0.3492647409439087, "learning_rate": 9.376478881455009e-05, "loss": 0.0737, "step": 6080 }, { "grad_norm": 0.26932772994041443, "learning_rate": 9.373810896153647e-05, "loss": 0.068, "step": 6090 }, { "grad_norm": 0.4199330806732178, "learning_rate": 9.371137596259623e-05, "loss": 0.0775, "step": 6100 }, { "grad_norm": 0.24818776547908783, "learning_rate": 9.368458985021248e-05, "loss": 0.0587, "step": 6110 }, { "grad_norm": 0.30468839406967163, "learning_rate": 9.365775065693287e-05, "loss": 0.0673, "step": 6120 }, { "grad_norm": 0.2986454367637634, "learning_rate": 9.363085841536957e-05, "loss": 0.0573, "step": 6130 }, { "grad_norm": 0.31049644947052, "learning_rate": 9.360391315819917e-05, "loss": 0.0683, "step": 6140 }, { "grad_norm": 0.3427204191684723, "learning_rate": 9.357691491816269e-05, "loss": 0.091, "step": 6150 }, { "grad_norm": 0.3594108521938324, "learning_rate": 9.354986372806557e-05, "loss": 0.0875, "step": 6160 }, { "grad_norm": 0.27852752804756165, "learning_rate": 9.352275962077752e-05, "loss": 0.0776, "step": 6170 }, { "grad_norm": 0.36947229504585266, "learning_rate": 9.349560262923262e-05, "loss": 0.0568, "step": 6180 }, { "grad_norm": 0.3593618869781494, "learning_rate": 9.346839278642913e-05, "loss": 0.0768, "step": 6190 }, { "grad_norm": 0.29034745693206787, "learning_rate": 9.344113012542963e-05, "loss": 0.062, "step": 6200 }, { "grad_norm": 0.29143935441970825, "learning_rate": 9.341381467936079e-05, "loss": 0.0711, "step": 6210 }, { "grad_norm": 0.39614546298980713, "learning_rate": 9.338644648141346e-05, "loss": 0.0572, "step": 6220 }, { "grad_norm": 0.297868549823761, "learning_rate": 9.335902556484257e-05, "loss": 0.0689, "step": 6230 }, { "grad_norm": 0.3448638617992401, "learning_rate": 9.333155196296712e-05, "loss": 0.0695, "step": 6240 }, { "grad_norm": 0.28475135564804077, "learning_rate": 9.330402570917016e-05, "loss": 0.0679, "step": 6250 }, { "grad_norm": 0.25220218300819397, "learning_rate": 9.327644683689865e-05, "loss": 0.0562, "step": 6260 }, { "grad_norm": 0.28449541330337524, "learning_rate": 9.324881537966354e-05, "loss": 0.0592, "step": 6270 }, { "grad_norm": 0.4012209177017212, "learning_rate": 9.322113137103964e-05, "loss": 0.0594, "step": 6280 }, { "grad_norm": 0.36757007241249084, "learning_rate": 9.319339484466564e-05, "loss": 0.0666, "step": 6290 }, { "grad_norm": 0.22058990597724915, "learning_rate": 9.316560583424403e-05, "loss": 0.0592, "step": 6300 }, { "grad_norm": 0.3151439428329468, "learning_rate": 9.313776437354108e-05, "loss": 0.0555, "step": 6310 }, { "grad_norm": 0.3250259459018707, "learning_rate": 9.31098704963868e-05, "loss": 0.0588, "step": 6320 }, { "grad_norm": 0.26238471269607544, "learning_rate": 9.308192423667485e-05, "loss": 0.0589, "step": 6330 }, { "grad_norm": 0.3021993339061737, "learning_rate": 9.305392562836262e-05, "loss": 0.0718, "step": 6340 }, { "grad_norm": 0.3117139935493469, "learning_rate": 9.3025874705471e-05, "loss": 0.0638, "step": 6350 }, { "grad_norm": 0.349623441696167, "learning_rate": 9.299777150208457e-05, "loss": 0.0661, "step": 6360 }, { "grad_norm": 0.3632104992866516, "learning_rate": 9.296961605235133e-05, "loss": 0.0581, "step": 6370 }, { "grad_norm": 0.3572457432746887, "learning_rate": 9.29414083904828e-05, "loss": 0.0593, "step": 6380 }, { "grad_norm": 0.3871201276779175, "learning_rate": 9.291314855075397e-05, "loss": 0.0696, "step": 6390 }, { "grad_norm": 0.27238932251930237, "learning_rate": 9.288483656750322e-05, "loss": 0.0662, "step": 6400 }, { "grad_norm": 0.2907124161720276, "learning_rate": 9.285647247513225e-05, "loss": 0.0535, "step": 6410 }, { "grad_norm": 0.28582894802093506, "learning_rate": 9.282805630810614e-05, "loss": 0.0683, "step": 6420 }, { "grad_norm": 0.3173106908798218, "learning_rate": 9.279958810095317e-05, "loss": 0.0633, "step": 6430 }, { "grad_norm": 0.31924968957901, "learning_rate": 9.277106788826494e-05, "loss": 0.0677, "step": 6440 }, { "grad_norm": 0.3618471622467041, "learning_rate": 9.274249570469618e-05, "loss": 0.0585, "step": 6450 }, { "grad_norm": 0.3017713129520416, "learning_rate": 9.271387158496476e-05, "loss": 0.0611, "step": 6460 }, { "grad_norm": 0.27444908022880554, "learning_rate": 9.268519556385173e-05, "loss": 0.0649, "step": 6470 }, { "grad_norm": 0.2426140308380127, "learning_rate": 9.265646767620113e-05, "loss": 0.0539, "step": 6480 }, { "grad_norm": 0.3995105028152466, "learning_rate": 9.262768795692006e-05, "loss": 0.0729, "step": 6490 }, { "grad_norm": 0.33021706342697144, "learning_rate": 9.259885644097861e-05, "loss": 0.0519, "step": 6500 }, { "grad_norm": 0.38711899518966675, "learning_rate": 9.256997316340976e-05, "loss": 0.0534, "step": 6510 }, { "grad_norm": 0.3509060740470886, "learning_rate": 9.254103815930943e-05, "loss": 0.061, "step": 6520 }, { "grad_norm": 0.2589935064315796, "learning_rate": 9.251205146383636e-05, "loss": 0.0583, "step": 6530 }, { "grad_norm": 0.2807754576206207, "learning_rate": 9.248301311221216e-05, "loss": 0.0724, "step": 6540 }, { "grad_norm": 0.3476012647151947, "learning_rate": 9.245392313972115e-05, "loss": 0.0537, "step": 6550 }, { "grad_norm": 0.2692490816116333, "learning_rate": 9.242478158171037e-05, "loss": 0.0651, "step": 6560 }, { "grad_norm": 0.4631505608558655, "learning_rate": 9.239558847358957e-05, "loss": 0.0736, "step": 6570 }, { "grad_norm": 0.2911108434200287, "learning_rate": 9.236634385083114e-05, "loss": 0.0576, "step": 6580 }, { "grad_norm": 0.4063659608364105, "learning_rate": 9.233704774897005e-05, "loss": 0.0807, "step": 6590 }, { "grad_norm": 0.3191191256046295, "learning_rate": 9.230770020360382e-05, "loss": 0.0775, "step": 6600 }, { "grad_norm": 0.2895166575908661, "learning_rate": 9.227830125039247e-05, "loss": 0.0553, "step": 6610 }, { "grad_norm": 0.23409207165241241, "learning_rate": 9.224885092505853e-05, "loss": 0.052, "step": 6620 }, { "grad_norm": 0.30995839834213257, "learning_rate": 9.221934926338691e-05, "loss": 0.06, "step": 6630 }, { "grad_norm": 0.2553870975971222, "learning_rate": 9.21897963012249e-05, "loss": 0.0483, "step": 6640 }, { "grad_norm": 0.2702620029449463, "learning_rate": 9.216019207448217e-05, "loss": 0.074, "step": 6650 }, { "grad_norm": 0.29379966855049133, "learning_rate": 9.21305366191306e-05, "loss": 0.0529, "step": 6660 }, { "grad_norm": 0.32158833742141724, "learning_rate": 9.21008299712044e-05, "loss": 0.0521, "step": 6670 }, { "grad_norm": 0.2277429848909378, "learning_rate": 9.207107216679993e-05, "loss": 0.0636, "step": 6680 }, { "grad_norm": 0.2905857563018799, "learning_rate": 9.204126324207574e-05, "loss": 0.0621, "step": 6690 }, { "grad_norm": 0.3196415305137634, "learning_rate": 9.201140323325248e-05, "loss": 0.0617, "step": 6700 }, { "grad_norm": 0.2943153977394104, "learning_rate": 9.198149217661287e-05, "loss": 0.0625, "step": 6710 }, { "grad_norm": 0.2521283030509949, "learning_rate": 9.195153010850167e-05, "loss": 0.0688, "step": 6720 }, { "grad_norm": 0.28602156043052673, "learning_rate": 9.192151706532562e-05, "loss": 0.0618, "step": 6730 }, { "grad_norm": 0.2941585183143616, "learning_rate": 9.189145308355337e-05, "loss": 0.0855, "step": 6740 }, { "grad_norm": 0.28725162148475647, "learning_rate": 9.186133819971555e-05, "loss": 0.0534, "step": 6750 }, { "grad_norm": 0.2897113263607025, "learning_rate": 9.183117245040455e-05, "loss": 0.0576, "step": 6760 }, { "grad_norm": 0.32031962275505066, "learning_rate": 9.180095587227458e-05, "loss": 0.0506, "step": 6770 }, { "grad_norm": 0.3022593557834625, "learning_rate": 9.177068850204167e-05, "loss": 0.065, "step": 6780 }, { "grad_norm": 0.3544381260871887, "learning_rate": 9.174037037648351e-05, "loss": 0.0564, "step": 6790 }, { "grad_norm": 0.29177233576774597, "learning_rate": 9.171000153243947e-05, "loss": 0.0706, "step": 6800 }, { "grad_norm": 0.29167279601097107, "learning_rate": 9.167958200681057e-05, "loss": 0.049, "step": 6810 }, { "grad_norm": 0.3325619399547577, "learning_rate": 9.164911183655943e-05, "loss": 0.0682, "step": 6820 }, { "grad_norm": 0.28959158062934875, "learning_rate": 9.161859105871013e-05, "loss": 0.0648, "step": 6830 }, { "grad_norm": 0.36841556429862976, "learning_rate": 9.158801971034832e-05, "loss": 0.0677, "step": 6840 }, { "grad_norm": 0.2530005872249603, "learning_rate": 9.155739782862107e-05, "loss": 0.0625, "step": 6850 }, { "grad_norm": 0.31044507026672363, "learning_rate": 9.152672545073687e-05, "loss": 0.0535, "step": 6860 }, { "grad_norm": 0.30744463205337524, "learning_rate": 9.149600261396551e-05, "loss": 0.0863, "step": 6870 }, { "grad_norm": 0.2991034984588623, "learning_rate": 9.146522935563817e-05, "loss": 0.0728, "step": 6880 }, { "grad_norm": 0.30861696600914, "learning_rate": 9.143440571314723e-05, "loss": 0.0593, "step": 6890 }, { "grad_norm": 0.3867635130882263, "learning_rate": 9.140353172394636e-05, "loss": 0.0697, "step": 6900 }, { "grad_norm": 0.3190160095691681, "learning_rate": 9.137260742555032e-05, "loss": 0.0562, "step": 6910 }, { "grad_norm": 0.3696102201938629, "learning_rate": 9.13416328555351e-05, "loss": 0.0707, "step": 6920 }, { "grad_norm": 0.3417053520679474, "learning_rate": 9.131060805153769e-05, "loss": 0.0716, "step": 6930 }, { "grad_norm": 0.3282773494720459, "learning_rate": 9.127953305125618e-05, "loss": 0.0553, "step": 6940 }, { "grad_norm": 0.37075749039649963, "learning_rate": 9.124840789244958e-05, "loss": 0.0564, "step": 6950 }, { "grad_norm": 0.236242413520813, "learning_rate": 9.121723261293794e-05, "loss": 0.0622, "step": 6960 }, { "grad_norm": 0.3080188035964966, "learning_rate": 9.118600725060214e-05, "loss": 0.0634, "step": 6970 }, { "grad_norm": 0.28138449788093567, "learning_rate": 9.115473184338392e-05, "loss": 0.0693, "step": 6980 }, { "grad_norm": 0.21389149129390717, "learning_rate": 9.11234064292859e-05, "loss": 0.0526, "step": 6990 }, { "grad_norm": 0.35704877972602844, "learning_rate": 9.109203104637137e-05, "loss": 0.0671, "step": 7000 }, { "grad_norm": 0.3446478545665741, "learning_rate": 9.10606057327644e-05, "loss": 0.0685, "step": 7010 }, { "grad_norm": 0.3182772397994995, "learning_rate": 9.10291305266497e-05, "loss": 0.0585, "step": 7020 }, { "grad_norm": 0.3335762619972229, "learning_rate": 9.099760546627261e-05, "loss": 0.0833, "step": 7030 }, { "grad_norm": 0.2443947046995163, "learning_rate": 9.096603058993907e-05, "loss": 0.0555, "step": 7040 }, { "grad_norm": 0.3097827434539795, "learning_rate": 9.093440593601552e-05, "loss": 0.053, "step": 7050 }, { "grad_norm": 0.28366947174072266, "learning_rate": 9.090273154292889e-05, "loss": 0.0532, "step": 7060 }, { "grad_norm": 0.27883103489875793, "learning_rate": 9.087100744916656e-05, "loss": 0.0628, "step": 7070 }, { "grad_norm": 0.33641988039016724, "learning_rate": 9.08392336932763e-05, "loss": 0.0773, "step": 7080 }, { "grad_norm": 0.3267173171043396, "learning_rate": 9.080741031386619e-05, "loss": 0.068, "step": 7090 }, { "grad_norm": 0.250838041305542, "learning_rate": 9.077553734960468e-05, "loss": 0.0512, "step": 7100 }, { "grad_norm": 0.29570168256759644, "learning_rate": 9.074361483922041e-05, "loss": 0.0666, "step": 7110 }, { "grad_norm": 0.33114975690841675, "learning_rate": 9.071164282150224e-05, "loss": 0.0518, "step": 7120 }, { "grad_norm": 0.26472166180610657, "learning_rate": 9.067962133529919e-05, "loss": 0.0541, "step": 7130 }, { "grad_norm": 0.26712125539779663, "learning_rate": 9.064755041952036e-05, "loss": 0.0736, "step": 7140 }, { "grad_norm": 0.34713247418403625, "learning_rate": 9.061543011313497e-05, "loss": 0.0754, "step": 7150 }, { "grad_norm": 0.3616321086883545, "learning_rate": 9.05832604551722e-05, "loss": 0.0699, "step": 7160 }, { "grad_norm": 0.32782942056655884, "learning_rate": 9.055104148472124e-05, "loss": 0.0681, "step": 7170 }, { "grad_norm": 0.3218350112438202, "learning_rate": 9.051877324093114e-05, "loss": 0.056, "step": 7180 }, { "grad_norm": 0.26620590686798096, "learning_rate": 9.04864557630109e-05, "loss": 0.0652, "step": 7190 }, { "grad_norm": 0.34265294671058655, "learning_rate": 9.045408909022928e-05, "loss": 0.058, "step": 7200 }, { "grad_norm": 0.2833801209926605, "learning_rate": 9.042167326191484e-05, "loss": 0.0665, "step": 7210 }, { "grad_norm": 0.28618761897087097, "learning_rate": 9.038920831745587e-05, "loss": 0.063, "step": 7220 }, { "grad_norm": 0.2561211884021759, "learning_rate": 9.035669429630036e-05, "loss": 0.0561, "step": 7230 }, { "grad_norm": 0.24540375173091888, "learning_rate": 9.032413123795588e-05, "loss": 0.0628, "step": 7240 }, { "grad_norm": 0.3402605652809143, "learning_rate": 9.029151918198962e-05, "loss": 0.068, "step": 7250 }, { "grad_norm": 0.26309555768966675, "learning_rate": 9.025885816802833e-05, "loss": 0.0539, "step": 7260 }, { "grad_norm": 0.3143633008003235, "learning_rate": 9.02261482357582e-05, "loss": 0.0495, "step": 7270 }, { "grad_norm": 0.30007120966911316, "learning_rate": 9.019338942492485e-05, "loss": 0.0509, "step": 7280 }, { "grad_norm": 0.33782634139060974, "learning_rate": 9.01605817753334e-05, "loss": 0.0583, "step": 7290 }, { "grad_norm": 0.4020177125930786, "learning_rate": 9.012772532684818e-05, "loss": 0.056, "step": 7300 }, { "grad_norm": 0.357867032289505, "learning_rate": 9.00948201193929e-05, "loss": 0.0595, "step": 7310 }, { "grad_norm": 0.3049239218235016, "learning_rate": 9.006186619295047e-05, "loss": 0.0478, "step": 7320 }, { "grad_norm": 0.2606947422027588, "learning_rate": 9.002886358756305e-05, "loss": 0.0611, "step": 7330 }, { "grad_norm": 0.30634990334510803, "learning_rate": 8.999581234333188e-05, "loss": 0.0599, "step": 7340 }, { "grad_norm": 0.29688361287117004, "learning_rate": 8.996271250041734e-05, "loss": 0.0505, "step": 7350 }, { "grad_norm": 0.31640371680259705, "learning_rate": 8.99295640990389e-05, "loss": 0.0548, "step": 7360 }, { "grad_norm": 0.21952670812606812, "learning_rate": 8.989636717947494e-05, "loss": 0.0614, "step": 7370 }, { "grad_norm": 0.21746519207954407, "learning_rate": 8.986312178206291e-05, "loss": 0.0473, "step": 7380 }, { "grad_norm": 0.23434697091579437, "learning_rate": 8.982982794719904e-05, "loss": 0.0602, "step": 7390 }, { "grad_norm": 0.25415071845054626, "learning_rate": 8.979648571533852e-05, "loss": 0.0587, "step": 7400 }, { "grad_norm": 0.3684408366680145, "learning_rate": 8.976309512699528e-05, "loss": 0.061, "step": 7410 }, { "grad_norm": 0.253826767206192, "learning_rate": 8.972965622274205e-05, "loss": 0.0685, "step": 7420 }, { "grad_norm": 0.29526665806770325, "learning_rate": 8.969616904321026e-05, "loss": 0.0674, "step": 7430 }, { "grad_norm": 0.35492202639579773, "learning_rate": 8.966263362908998e-05, "loss": 0.0612, "step": 7440 }, { "grad_norm": 0.34254756569862366, "learning_rate": 8.962905002112989e-05, "loss": 0.0528, "step": 7450 }, { "grad_norm": 0.23975923657417297, "learning_rate": 8.959541826013724e-05, "loss": 0.0637, "step": 7460 }, { "grad_norm": 0.2604082524776459, "learning_rate": 8.956173838697781e-05, "loss": 0.0635, "step": 7470 }, { "grad_norm": 0.31560009717941284, "learning_rate": 8.952801044257581e-05, "loss": 0.0698, "step": 7480 }, { "grad_norm": 0.3092935383319855, "learning_rate": 8.949423446791387e-05, "loss": 0.0574, "step": 7490 }, { "grad_norm": 0.2915799021720886, "learning_rate": 8.9460410504033e-05, "loss": 0.0525, "step": 7500 }, { "grad_norm": 0.29869532585144043, "learning_rate": 8.942653859203248e-05, "loss": 0.0482, "step": 7510 }, { "grad_norm": 0.369875431060791, "learning_rate": 8.93926187730699e-05, "loss": 0.047, "step": 7520 }, { "grad_norm": 0.24332013726234436, "learning_rate": 8.935865108836102e-05, "loss": 0.0496, "step": 7530 }, { "grad_norm": 0.26538342237472534, "learning_rate": 8.932463557917981e-05, "loss": 0.0546, "step": 7540 }, { "grad_norm": 0.29162678122520447, "learning_rate": 8.929057228685829e-05, "loss": 0.0541, "step": 7550 }, { "grad_norm": 0.28678497672080994, "learning_rate": 8.925646125278658e-05, "loss": 0.0516, "step": 7560 }, { "grad_norm": 0.35382333397865295, "learning_rate": 8.92223025184128e-05, "loss": 0.063, "step": 7570 }, { "grad_norm": 0.2611851692199707, "learning_rate": 8.918809612524303e-05, "loss": 0.0614, "step": 7580 }, { "grad_norm": 0.248627707362175, "learning_rate": 8.91538421148413e-05, "loss": 0.0621, "step": 7590 }, { "grad_norm": 0.3188674747943878, "learning_rate": 8.91195405288294e-05, "loss": 0.0727, "step": 7600 }, { "grad_norm": 0.2575676441192627, "learning_rate": 8.908519140888705e-05, "loss": 0.0581, "step": 7610 }, { "grad_norm": 0.35355809330940247, "learning_rate": 8.905079479675164e-05, "loss": 0.0564, "step": 7620 }, { "grad_norm": 0.28264912962913513, "learning_rate": 8.901635073421831e-05, "loss": 0.0605, "step": 7630 }, { "grad_norm": 0.27571675181388855, "learning_rate": 8.898185926313983e-05, "loss": 0.0618, "step": 7640 }, { "grad_norm": 0.3539575934410095, "learning_rate": 8.894732042542659e-05, "loss": 0.053, "step": 7650 }, { "grad_norm": 0.32171306014060974, "learning_rate": 8.891273426304654e-05, "loss": 0.0523, "step": 7660 }, { "grad_norm": 0.2886122763156891, "learning_rate": 8.887810081802515e-05, "loss": 0.0633, "step": 7670 }, { "grad_norm": 0.3257535994052887, "learning_rate": 8.884342013244529e-05, "loss": 0.0629, "step": 7680 }, { "grad_norm": 0.31563088297843933, "learning_rate": 8.880869224844725e-05, "loss": 0.0561, "step": 7690 }, { "grad_norm": 0.3266989588737488, "learning_rate": 8.877391720822875e-05, "loss": 0.0503, "step": 7700 }, { "grad_norm": 0.333683580160141, "learning_rate": 8.873909505404468e-05, "loss": 0.0697, "step": 7710 }, { "grad_norm": 0.3123229146003723, "learning_rate": 8.870422582820726e-05, "loss": 0.0612, "step": 7720 }, { "grad_norm": 0.29147085547447205, "learning_rate": 8.866930957308588e-05, "loss": 0.058, "step": 7730 }, { "grad_norm": 0.2849343717098236, "learning_rate": 8.86343463311071e-05, "loss": 0.0599, "step": 7740 }, { "grad_norm": 0.26260074973106384, "learning_rate": 8.859933614475452e-05, "loss": 0.048, "step": 7750 }, { "grad_norm": 0.27800849080085754, "learning_rate": 8.85642790565689e-05, "loss": 0.0596, "step": 7760 }, { "grad_norm": 0.2939211428165436, "learning_rate": 8.852917510914782e-05, "loss": 0.061, "step": 7770 }, { "grad_norm": 0.22196416556835175, "learning_rate": 8.84940243451459e-05, "loss": 0.0618, "step": 7780 }, { "grad_norm": 0.28265464305877686, "learning_rate": 8.845882680727469e-05, "loss": 0.0546, "step": 7790 }, { "grad_norm": 0.24199794232845306, "learning_rate": 8.842358253830246e-05, "loss": 0.051, "step": 7800 }, { "grad_norm": 0.2887401580810547, "learning_rate": 8.838829158105434e-05, "loss": 0.0557, "step": 7810 }, { "grad_norm": 0.22729307413101196, "learning_rate": 8.835295397841217e-05, "loss": 0.0444, "step": 7820 }, { "grad_norm": 0.2723986506462097, "learning_rate": 8.831756977331447e-05, "loss": 0.0497, "step": 7830 }, { "grad_norm": 0.29254522919654846, "learning_rate": 8.828213900875638e-05, "loss": 0.044, "step": 7840 }, { "grad_norm": 0.232318714261055, "learning_rate": 8.824666172778964e-05, "loss": 0.0487, "step": 7850 }, { "grad_norm": 0.3090560734272003, "learning_rate": 8.821113797352245e-05, "loss": 0.0645, "step": 7860 }, { "grad_norm": 0.27316391468048096, "learning_rate": 8.817556778911956e-05, "loss": 0.0659, "step": 7870 }, { "grad_norm": 0.33472269773483276, "learning_rate": 8.81399512178021e-05, "loss": 0.062, "step": 7880 }, { "grad_norm": 0.27600476145744324, "learning_rate": 8.810428830284752e-05, "loss": 0.0518, "step": 7890 }, { "grad_norm": 0.292817622423172, "learning_rate": 8.806857908758967e-05, "loss": 0.0693, "step": 7900 }, { "grad_norm": 0.2777009904384613, "learning_rate": 8.80328236154186e-05, "loss": 0.0642, "step": 7910 }, { "grad_norm": 0.37165528535842896, "learning_rate": 8.799702192978056e-05, "loss": 0.0854, "step": 7920 }, { "grad_norm": 0.2624307870864868, "learning_rate": 8.7961174074178e-05, "loss": 0.0556, "step": 7930 }, { "grad_norm": 0.30395254492759705, "learning_rate": 8.792528009216942e-05, "loss": 0.063, "step": 7940 }, { "grad_norm": 0.39645254611968994, "learning_rate": 8.788934002736943e-05, "loss": 0.0542, "step": 7950 }, { "grad_norm": 0.27897000312805176, "learning_rate": 8.785335392344857e-05, "loss": 0.0532, "step": 7960 }, { "grad_norm": 0.299923300743103, "learning_rate": 8.781732182413336e-05, "loss": 0.0517, "step": 7970 }, { "grad_norm": 0.30059191584587097, "learning_rate": 8.778124377320618e-05, "loss": 0.0532, "step": 7980 }, { "grad_norm": 0.32196223735809326, "learning_rate": 8.774511981450529e-05, "loss": 0.0472, "step": 7990 }, { "grad_norm": 0.3021567165851593, "learning_rate": 8.770894999192467e-05, "loss": 0.0473, "step": 8000 }, { "grad_norm": 0.28544217348098755, "learning_rate": 8.767273434941412e-05, "loss": 0.0494, "step": 8010 }, { "grad_norm": 0.3216398358345032, "learning_rate": 8.7636472930979e-05, "loss": 0.0563, "step": 8020 }, { "grad_norm": 0.2400844395160675, "learning_rate": 8.76001657806804e-05, "loss": 0.0494, "step": 8030 }, { "grad_norm": 0.28355246782302856, "learning_rate": 8.756381294263491e-05, "loss": 0.0483, "step": 8040 }, { "grad_norm": 0.28866615891456604, "learning_rate": 8.752741446101464e-05, "loss": 0.0564, "step": 8050 }, { "grad_norm": 0.24818548560142517, "learning_rate": 8.749097038004722e-05, "loss": 0.0444, "step": 8060 }, { "grad_norm": 0.27400869131088257, "learning_rate": 8.745448074401561e-05, "loss": 0.0488, "step": 8070 }, { "grad_norm": 0.27506697177886963, "learning_rate": 8.741794559725818e-05, "loss": 0.0683, "step": 8080 }, { "grad_norm": 0.2994530200958252, "learning_rate": 8.738136498416856e-05, "loss": 0.0567, "step": 8090 }, { "grad_norm": 0.27273520827293396, "learning_rate": 8.734473894919564e-05, "loss": 0.0627, "step": 8100 }, { "grad_norm": 0.35509568452835083, "learning_rate": 8.730806753684353e-05, "loss": 0.0608, "step": 8110 }, { "grad_norm": 0.21389254927635193, "learning_rate": 8.727135079167144e-05, "loss": 0.0546, "step": 8120 }, { "grad_norm": 0.28771549463272095, "learning_rate": 8.723458875829368e-05, "loss": 0.0663, "step": 8130 }, { "grad_norm": 0.2096043974161148, "learning_rate": 8.719778148137958e-05, "loss": 0.0602, "step": 8140 }, { "grad_norm": 0.264454185962677, "learning_rate": 8.716092900565346e-05, "loss": 0.0495, "step": 8150 }, { "grad_norm": 0.3526304066181183, "learning_rate": 8.712403137589456e-05, "loss": 0.0676, "step": 8160 }, { "grad_norm": 0.2823895812034607, "learning_rate": 8.708708863693697e-05, "loss": 0.0569, "step": 8170 }, { "grad_norm": 0.24215340614318848, "learning_rate": 8.70501008336696e-05, "loss": 0.0516, "step": 8180 }, { "grad_norm": 0.30370357632637024, "learning_rate": 8.701306801103611e-05, "loss": 0.0547, "step": 8190 }, { "grad_norm": 0.2672850787639618, "learning_rate": 8.697599021403489e-05, "loss": 0.05, "step": 8200 }, { "grad_norm": 0.24326245486736298, "learning_rate": 8.693886748771896e-05, "loss": 0.0415, "step": 8210 }, { "grad_norm": 0.21991650760173798, "learning_rate": 8.690169987719593e-05, "loss": 0.0454, "step": 8220 }, { "grad_norm": 0.28604015707969666, "learning_rate": 8.686448742762791e-05, "loss": 0.0494, "step": 8230 }, { "grad_norm": 0.24010229110717773, "learning_rate": 8.682723018423161e-05, "loss": 0.0535, "step": 8240 }, { "grad_norm": 0.3204074800014496, "learning_rate": 8.678992819227804e-05, "loss": 0.0486, "step": 8250 }, { "grad_norm": 0.26610270142555237, "learning_rate": 8.675258149709265e-05, "loss": 0.0409, "step": 8260 }, { "grad_norm": 0.2511579692363739, "learning_rate": 8.67151901440552e-05, "loss": 0.0431, "step": 8270 }, { "grad_norm": 0.3090936839580536, "learning_rate": 8.667775417859971e-05, "loss": 0.0599, "step": 8280 }, { "grad_norm": 0.310330331325531, "learning_rate": 8.664027364621441e-05, "loss": 0.0439, "step": 8290 }, { "grad_norm": 0.3341936469078064, "learning_rate": 8.660274859244168e-05, "loss": 0.0596, "step": 8300 }, { "grad_norm": 0.3190235197544098, "learning_rate": 8.656517906287799e-05, "loss": 0.0485, "step": 8310 }, { "grad_norm": 0.25204184651374817, "learning_rate": 8.652756510317387e-05, "loss": 0.0416, "step": 8320 }, { "grad_norm": 0.30357614159584045, "learning_rate": 8.648990675903382e-05, "loss": 0.0529, "step": 8330 }, { "grad_norm": 0.29918432235717773, "learning_rate": 8.64522040762163e-05, "loss": 0.0429, "step": 8340 }, { "grad_norm": 0.2640821933746338, "learning_rate": 8.64144571005336e-05, "loss": 0.0614, "step": 8350 }, { "grad_norm": 0.280055969953537, "learning_rate": 8.637666587785184e-05, "loss": 0.0516, "step": 8360 }, { "grad_norm": 0.290903240442276, "learning_rate": 8.633883045409096e-05, "loss": 0.0452, "step": 8370 }, { "grad_norm": 0.2710651755332947, "learning_rate": 8.630095087522457e-05, "loss": 0.0535, "step": 8380 }, { "grad_norm": 0.2425202876329422, "learning_rate": 8.62630271872799e-05, "loss": 0.0469, "step": 8390 }, { "grad_norm": 0.24323712289333344, "learning_rate": 8.62250594363378e-05, "loss": 0.0484, "step": 8400 }, { "grad_norm": 0.29924124479293823, "learning_rate": 8.61870476685327e-05, "loss": 0.054, "step": 8410 }, { "grad_norm": 0.22607800364494324, "learning_rate": 8.614899193005248e-05, "loss": 0.0446, "step": 8420 }, { "grad_norm": 0.3191208839416504, "learning_rate": 8.611089226713842e-05, "loss": 0.0509, "step": 8430 }, { "grad_norm": 0.32413366436958313, "learning_rate": 8.607274872608521e-05, "loss": 0.0502, "step": 8440 }, { "grad_norm": 0.28586545586586, "learning_rate": 8.603456135324088e-05, "loss": 0.0464, "step": 8450 }, { "grad_norm": 0.23136994242668152, "learning_rate": 8.599633019500665e-05, "loss": 0.0468, "step": 8460 }, { "grad_norm": 0.25228628516197205, "learning_rate": 8.595805529783702e-05, "loss": 0.044, "step": 8470 }, { "grad_norm": 0.2524213492870331, "learning_rate": 8.591973670823958e-05, "loss": 0.0459, "step": 8480 }, { "grad_norm": 0.29968252778053284, "learning_rate": 8.588137447277503e-05, "loss": 0.0531, "step": 8490 }, { "grad_norm": 0.3116902709007263, "learning_rate": 8.584296863805709e-05, "loss": 0.0563, "step": 8500 }, { "grad_norm": 0.3455532491207123, "learning_rate": 8.580451925075248e-05, "loss": 0.0496, "step": 8510 }, { "grad_norm": 0.2749980390071869, "learning_rate": 8.576602635758085e-05, "loss": 0.0364, "step": 8520 }, { "grad_norm": 0.22466090321540833, "learning_rate": 8.572749000531469e-05, "loss": 0.0461, "step": 8530 }, { "grad_norm": 0.23747560381889343, "learning_rate": 8.568891024077926e-05, "loss": 0.0495, "step": 8540 }, { "grad_norm": 0.31052207946777344, "learning_rate": 8.565028711085265e-05, "loss": 0.0525, "step": 8550 }, { "grad_norm": 0.27240148186683655, "learning_rate": 8.561162066246561e-05, "loss": 0.0512, "step": 8560 }, { "grad_norm": 0.2537265121936798, "learning_rate": 8.557291094260152e-05, "loss": 0.0775, "step": 8570 }, { "grad_norm": 0.3093377947807312, "learning_rate": 8.553415799829629e-05, "loss": 0.0597, "step": 8580 }, { "grad_norm": 0.26136863231658936, "learning_rate": 8.549536187663847e-05, "loss": 0.0532, "step": 8590 }, { "grad_norm": 0.22629521787166595, "learning_rate": 8.545652262476897e-05, "loss": 0.0402, "step": 8600 }, { "grad_norm": 0.3351418972015381, "learning_rate": 8.541764028988115e-05, "loss": 0.0461, "step": 8610 }, { "grad_norm": 0.29206499457359314, "learning_rate": 8.537871491922071e-05, "loss": 0.0562, "step": 8620 }, { "grad_norm": 0.26203903555870056, "learning_rate": 8.533974656008566e-05, "loss": 0.0454, "step": 8630 }, { "grad_norm": 0.3431650698184967, "learning_rate": 8.530073525982621e-05, "loss": 0.0516, "step": 8640 }, { "grad_norm": 0.2787887156009674, "learning_rate": 8.526168106584476e-05, "loss": 0.0528, "step": 8650 }, { "grad_norm": 0.23733307421207428, "learning_rate": 8.522258402559586e-05, "loss": 0.0478, "step": 8660 }, { "grad_norm": 0.30951014161109924, "learning_rate": 8.518344418658613e-05, "loss": 0.0468, "step": 8670 }, { "grad_norm": 0.25972819328308105, "learning_rate": 8.51442615963741e-05, "loss": 0.0407, "step": 8680 }, { "grad_norm": 0.27463382482528687, "learning_rate": 8.510503630257034e-05, "loss": 0.0454, "step": 8690 }, { "grad_norm": 0.31636977195739746, "learning_rate": 8.50657683528373e-05, "loss": 0.0461, "step": 8700 }, { "grad_norm": 0.2419833391904831, "learning_rate": 8.502645779488922e-05, "loss": 0.0546, "step": 8710 }, { "grad_norm": 0.3163353204727173, "learning_rate": 8.498710467649213e-05, "loss": 0.0693, "step": 8720 }, { "grad_norm": 0.232834130525589, "learning_rate": 8.494770904546381e-05, "loss": 0.0475, "step": 8730 }, { "grad_norm": 0.24939203262329102, "learning_rate": 8.490827094967363e-05, "loss": 0.0459, "step": 8740 }, { "grad_norm": 0.25879377126693726, "learning_rate": 8.486879043704263e-05, "loss": 0.0556, "step": 8750 }, { "grad_norm": 0.2087334543466568, "learning_rate": 8.482926755554332e-05, "loss": 0.0442, "step": 8760 }, { "grad_norm": 0.3335881233215332, "learning_rate": 8.478970235319975e-05, "loss": 0.0517, "step": 8770 }, { "grad_norm": 0.26189175248146057, "learning_rate": 8.475009487808737e-05, "loss": 0.0522, "step": 8780 }, { "grad_norm": 0.2205239236354828, "learning_rate": 8.471044517833298e-05, "loss": 0.0415, "step": 8790 }, { "grad_norm": 0.272456556558609, "learning_rate": 8.467075330211474e-05, "loss": 0.0386, "step": 8800 }, { "grad_norm": 0.2636466920375824, "learning_rate": 8.463101929766198e-05, "loss": 0.0505, "step": 8810 }, { "grad_norm": 0.2145378589630127, "learning_rate": 8.459124321325528e-05, "loss": 0.0447, "step": 8820 }, { "grad_norm": 0.22128890454769135, "learning_rate": 8.455142509722634e-05, "loss": 0.0465, "step": 8830 }, { "grad_norm": 0.2453867346048355, "learning_rate": 8.451156499795792e-05, "loss": 0.0453, "step": 8840 }, { "grad_norm": 0.2712414860725403, "learning_rate": 8.447166296388381e-05, "loss": 0.0397, "step": 8850 }, { "grad_norm": 0.26136285066604614, "learning_rate": 8.443171904348872e-05, "loss": 0.0414, "step": 8860 }, { "grad_norm": 0.29648956656455994, "learning_rate": 8.43917332853083e-05, "loss": 0.0467, "step": 8870 }, { "grad_norm": 0.2118912935256958, "learning_rate": 8.435170573792901e-05, "loss": 0.0476, "step": 8880 }, { "grad_norm": 0.30057474970817566, "learning_rate": 8.431163644998808e-05, "loss": 0.0508, "step": 8890 }, { "grad_norm": 0.2780839502811432, "learning_rate": 8.42715254701735e-05, "loss": 0.0476, "step": 8900 }, { "grad_norm": 0.28566572070121765, "learning_rate": 8.42313728472239e-05, "loss": 0.0434, "step": 8910 }, { "grad_norm": 0.36405450105667114, "learning_rate": 8.419117862992844e-05, "loss": 0.0573, "step": 8920 }, { "grad_norm": 0.2788868546485901, "learning_rate": 8.415094286712694e-05, "loss": 0.0409, "step": 8930 }, { "grad_norm": 0.3000909090042114, "learning_rate": 8.411066560770965e-05, "loss": 0.0494, "step": 8940 }, { "grad_norm": 0.2516811788082123, "learning_rate": 8.407034690061721e-05, "loss": 0.045, "step": 8950 }, { "grad_norm": 0.24953649938106537, "learning_rate": 8.402998679484067e-05, "loss": 0.0397, "step": 8960 }, { "grad_norm": 0.24034424126148224, "learning_rate": 8.398958533942135e-05, "loss": 0.0376, "step": 8970 }, { "grad_norm": 0.30077075958251953, "learning_rate": 8.394914258345083e-05, "loss": 0.0495, "step": 8980 }, { "grad_norm": 0.3470993638038635, "learning_rate": 8.390865857607089e-05, "loss": 0.069, "step": 8990 }, { "grad_norm": 0.2332250475883484, "learning_rate": 8.386813336647339e-05, "loss": 0.0456, "step": 9000 }, { "grad_norm": 0.2492009848356247, "learning_rate": 8.38275670039003e-05, "loss": 0.0376, "step": 9010 }, { "grad_norm": 0.2934356927871704, "learning_rate": 8.378695953764356e-05, "loss": 0.0456, "step": 9020 }, { "grad_norm": 0.2467339038848877, "learning_rate": 8.374631101704509e-05, "loss": 0.0608, "step": 9030 }, { "grad_norm": 0.23218746483325958, "learning_rate": 8.370562149149665e-05, "loss": 0.0553, "step": 9040 }, { "grad_norm": 0.2821784019470215, "learning_rate": 8.366489101043989e-05, "loss": 0.0452, "step": 9050 }, { "grad_norm": 0.22763745486736298, "learning_rate": 8.362411962336614e-05, "loss": 0.0532, "step": 9060 }, { "grad_norm": 0.23225706815719604, "learning_rate": 8.358330737981651e-05, "loss": 0.0371, "step": 9070 }, { "grad_norm": 0.21868430078029633, "learning_rate": 8.35424543293817e-05, "loss": 0.0442, "step": 9080 }, { "grad_norm": 0.21281558275222778, "learning_rate": 8.350156052170206e-05, "loss": 0.0552, "step": 9090 }, { "grad_norm": 0.3598162531852722, "learning_rate": 8.346062600646738e-05, "loss": 0.0579, "step": 9100 }, { "grad_norm": 0.2622242271900177, "learning_rate": 8.341965083341697e-05, "loss": 0.0475, "step": 9110 }, { "grad_norm": 0.2924858629703522, "learning_rate": 8.337863505233953e-05, "loss": 0.0507, "step": 9120 }, { "grad_norm": 0.2943890690803528, "learning_rate": 8.33375787130731e-05, "loss": 0.0538, "step": 9130 }, { "grad_norm": 0.30198705196380615, "learning_rate": 8.3296481865505e-05, "loss": 0.0393, "step": 9140 }, { "grad_norm": 0.3284444212913513, "learning_rate": 8.325534455957181e-05, "loss": 0.0578, "step": 9150 }, { "grad_norm": 0.3026992678642273, "learning_rate": 8.321416684525917e-05, "loss": 0.0368, "step": 9160 }, { "grad_norm": 0.24541352689266205, "learning_rate": 8.317294877260193e-05, "loss": 0.042, "step": 9170 }, { "grad_norm": 0.34712836146354675, "learning_rate": 8.313169039168394e-05, "loss": 0.0454, "step": 9180 }, { "grad_norm": 0.22341030836105347, "learning_rate": 8.3090391752638e-05, "loss": 0.036, "step": 9190 }, { "grad_norm": 0.16605167090892792, "learning_rate": 8.304905290564586e-05, "loss": 0.0481, "step": 9200 }, { "grad_norm": 0.24110513925552368, "learning_rate": 8.300767390093813e-05, "loss": 0.061, "step": 9210 }, { "grad_norm": 0.248176708817482, "learning_rate": 8.296625478879417e-05, "loss": 0.0574, "step": 9220 }, { "grad_norm": 0.2800736725330353, "learning_rate": 8.292479561954214e-05, "loss": 0.0446, "step": 9230 }, { "grad_norm": 0.27925363183021545, "learning_rate": 8.288329644355883e-05, "loss": 0.0524, "step": 9240 }, { "grad_norm": 0.1991814225912094, "learning_rate": 8.284175731126965e-05, "loss": 0.0381, "step": 9250 }, { "grad_norm": 0.26170802116394043, "learning_rate": 8.280017827314854e-05, "loss": 0.0432, "step": 9260 }, { "grad_norm": 0.32718122005462646, "learning_rate": 8.2758559379718e-05, "loss": 0.0466, "step": 9270 }, { "grad_norm": 0.2426115870475769, "learning_rate": 8.271690068154887e-05, "loss": 0.0453, "step": 9280 }, { "grad_norm": 0.2564951777458191, "learning_rate": 8.26752022292604e-05, "loss": 0.0585, "step": 9290 }, { "grad_norm": 0.2689344584941864, "learning_rate": 8.263346407352018e-05, "loss": 0.0505, "step": 9300 }, { "grad_norm": 0.30103445053100586, "learning_rate": 8.259168626504395e-05, "loss": 0.0454, "step": 9310 }, { "grad_norm": 0.2145436853170395, "learning_rate": 8.254986885459572e-05, "loss": 0.0628, "step": 9320 }, { "grad_norm": 0.2973059415817261, "learning_rate": 8.250801189298757e-05, "loss": 0.045, "step": 9330 }, { "grad_norm": 0.2533753216266632, "learning_rate": 8.246611543107967e-05, "loss": 0.0463, "step": 9340 }, { "grad_norm": 0.18445678055286407, "learning_rate": 8.242417951978018e-05, "loss": 0.042, "step": 9350 }, { "grad_norm": 0.246917724609375, "learning_rate": 8.238220421004517e-05, "loss": 0.0404, "step": 9360 }, { "grad_norm": 0.19765189290046692, "learning_rate": 8.234018955287859e-05, "loss": 0.0551, "step": 9370 }, { "grad_norm": 0.3127460777759552, "learning_rate": 8.229813559933224e-05, "loss": 0.0489, "step": 9380 }, { "grad_norm": 0.28748321533203125, "learning_rate": 8.22560424005056e-05, "loss": 0.0485, "step": 9390 }, { "grad_norm": 0.3034312427043915, "learning_rate": 8.22139100075459e-05, "loss": 0.0563, "step": 9400 }, { "grad_norm": 0.22079165279865265, "learning_rate": 8.217173847164799e-05, "loss": 0.0555, "step": 9410 }, { "grad_norm": 0.32796576619148254, "learning_rate": 8.212952784405422e-05, "loss": 0.0479, "step": 9420 }, { "grad_norm": 0.3491758704185486, "learning_rate": 8.208727817605451e-05, "loss": 0.0455, "step": 9430 }, { "grad_norm": 0.258953720331192, "learning_rate": 8.204498951898618e-05, "loss": 0.0497, "step": 9440 }, { "grad_norm": 0.2378096729516983, "learning_rate": 8.200266192423395e-05, "loss": 0.0437, "step": 9450 }, { "grad_norm": 0.2638334035873413, "learning_rate": 8.196029544322981e-05, "loss": 0.0402, "step": 9460 }, { "grad_norm": 0.25130248069763184, "learning_rate": 8.191789012745305e-05, "loss": 0.063, "step": 9470 }, { "grad_norm": 0.2027653306722641, "learning_rate": 8.187544602843014e-05, "loss": 0.0422, "step": 9480 }, { "grad_norm": 0.2631757855415344, "learning_rate": 8.183296319773466e-05, "loss": 0.0452, "step": 9490 }, { "grad_norm": 0.21387606859207153, "learning_rate": 8.179044168698721e-05, "loss": 0.0376, "step": 9500 }, { "grad_norm": 0.22583898901939392, "learning_rate": 8.174788154785548e-05, "loss": 0.0397, "step": 9510 }, { "grad_norm": 0.2272488921880722, "learning_rate": 8.170528283205403e-05, "loss": 0.0492, "step": 9520 }, { "grad_norm": 0.21835656464099884, "learning_rate": 8.166264559134435e-05, "loss": 0.0486, "step": 9530 }, { "grad_norm": 0.3125370144844055, "learning_rate": 8.161996987753467e-05, "loss": 0.0608, "step": 9540 }, { "grad_norm": 0.2734662890434265, "learning_rate": 8.157725574248e-05, "loss": 0.0501, "step": 9550 }, { "grad_norm": 0.27803707122802734, "learning_rate": 8.153450323808209e-05, "loss": 0.0539, "step": 9560 }, { "grad_norm": 0.375032514333725, "learning_rate": 8.149171241628923e-05, "loss": 0.0643, "step": 9570 }, { "grad_norm": 0.2398683726787567, "learning_rate": 8.144888332909631e-05, "loss": 0.0498, "step": 9580 }, { "grad_norm": 0.25993067026138306, "learning_rate": 8.14060160285447e-05, "loss": 0.0502, "step": 9590 }, { "grad_norm": 0.32848912477493286, "learning_rate": 8.136311056672223e-05, "loss": 0.0408, "step": 9600 }, { "grad_norm": 0.26502951979637146, "learning_rate": 8.132016699576308e-05, "loss": 0.0454, "step": 9610 }, { "grad_norm": 0.21933794021606445, "learning_rate": 8.127718536784771e-05, "loss": 0.0485, "step": 9620 }, { "grad_norm": 0.3090316951274872, "learning_rate": 8.12341657352029e-05, "loss": 0.0442, "step": 9630 }, { "grad_norm": 0.2781892418861389, "learning_rate": 8.119110815010151e-05, "loss": 0.0447, "step": 9640 }, { "grad_norm": 0.30298709869384766, "learning_rate": 8.114801266486261e-05, "loss": 0.0499, "step": 9650 }, { "grad_norm": 0.22665993869304657, "learning_rate": 8.110487933185122e-05, "loss": 0.05, "step": 9660 }, { "grad_norm": 0.2654498517513275, "learning_rate": 8.106170820347847e-05, "loss": 0.0536, "step": 9670 }, { "grad_norm": 0.2829221487045288, "learning_rate": 8.101849933220132e-05, "loss": 0.0375, "step": 9680 }, { "grad_norm": 0.2343592345714569, "learning_rate": 8.097525277052264e-05, "loss": 0.0442, "step": 9690 }, { "grad_norm": 0.1920456886291504, "learning_rate": 8.093196857099106e-05, "loss": 0.0389, "step": 9700 }, { "grad_norm": 0.3087690472602844, "learning_rate": 8.088864678620096e-05, "loss": 0.0426, "step": 9710 }, { "grad_norm": 0.2967287003993988, "learning_rate": 8.084528746879243e-05, "loss": 0.0476, "step": 9720 }, { "grad_norm": 0.2517305910587311, "learning_rate": 8.080189067145108e-05, "loss": 0.0563, "step": 9730 }, { "grad_norm": 0.2537696957588196, "learning_rate": 8.075845644690813e-05, "loss": 0.0429, "step": 9740 }, { "grad_norm": 0.266324907541275, "learning_rate": 8.07149848479403e-05, "loss": 0.0435, "step": 9750 }, { "grad_norm": 0.24241960048675537, "learning_rate": 8.067147592736962e-05, "loss": 0.0482, "step": 9760 }, { "grad_norm": 0.24011065065860748, "learning_rate": 8.062792973806357e-05, "loss": 0.0543, "step": 9770 }, { "grad_norm": 0.2036685347557068, "learning_rate": 8.058434633293485e-05, "loss": 0.0472, "step": 9780 }, { "grad_norm": 0.2262846827507019, "learning_rate": 8.054072576494143e-05, "loss": 0.047, "step": 9790 }, { "grad_norm": 0.2544848918914795, "learning_rate": 8.04970680870864e-05, "loss": 0.0404, "step": 9800 }, { "grad_norm": 0.29485446214675903, "learning_rate": 8.045337335241793e-05, "loss": 0.0475, "step": 9810 }, { "grad_norm": 0.29331570863723755, "learning_rate": 8.040964161402932e-05, "loss": 0.0529, "step": 9820 }, { "grad_norm": 0.25500836968421936, "learning_rate": 8.036587292505869e-05, "loss": 0.0468, "step": 9830 }, { "grad_norm": 0.18651726841926575, "learning_rate": 8.032206733868912e-05, "loss": 0.0558, "step": 9840 }, { "grad_norm": 0.233231320977211, "learning_rate": 8.027822490814858e-05, "loss": 0.055, "step": 9850 }, { "grad_norm": 0.19117946922779083, "learning_rate": 8.023434568670972e-05, "loss": 0.039, "step": 9860 }, { "grad_norm": 0.24533501267433167, "learning_rate": 8.019042972768992e-05, "loss": 0.0388, "step": 9870 }, { "grad_norm": 0.2981744408607483, "learning_rate": 8.014647708445124e-05, "loss": 0.0508, "step": 9880 }, { "grad_norm": 0.24683181941509247, "learning_rate": 8.010248781040027e-05, "loss": 0.0383, "step": 9890 }, { "grad_norm": 0.2710823118686676, "learning_rate": 8.005846195898814e-05, "loss": 0.0493, "step": 9900 }, { "grad_norm": 0.2593339681625366, "learning_rate": 8.00143995837104e-05, "loss": 0.0466, "step": 9910 }, { "grad_norm": 0.23342756927013397, "learning_rate": 7.997030073810698e-05, "loss": 0.0539, "step": 9920 }, { "grad_norm": 0.2977638840675354, "learning_rate": 7.992616547576217e-05, "loss": 0.0417, "step": 9930 }, { "grad_norm": 0.2937262952327728, "learning_rate": 7.988199385030445e-05, "loss": 0.0412, "step": 9940 }, { "grad_norm": 0.26820459961891174, "learning_rate": 7.98377859154065e-05, "loss": 0.035, "step": 9950 }, { "grad_norm": 0.2686287760734558, "learning_rate": 7.979354172478516e-05, "loss": 0.0374, "step": 9960 }, { "grad_norm": 0.2354012131690979, "learning_rate": 7.974926133220127e-05, "loss": 0.0475, "step": 9970 }, { "grad_norm": 0.2203461080789566, "learning_rate": 7.970494479145967e-05, "loss": 0.0436, "step": 9980 }, { "grad_norm": 0.290770024061203, "learning_rate": 7.966059215640919e-05, "loss": 0.0463, "step": 9990 }, { "grad_norm": 0.3050641417503357, "learning_rate": 7.961620348094241e-05, "loss": 0.0541, "step": 10000 }, { "grad_norm": 0.2578940987586975, "learning_rate": 7.957177881899577e-05, "loss": 0.0477, "step": 10010 }, { "grad_norm": 0.21045561134815216, "learning_rate": 7.952731822454944e-05, "loss": 0.0484, "step": 10020 }, { "grad_norm": 0.24383485317230225, "learning_rate": 7.948282175162722e-05, "loss": 0.0391, "step": 10030 }, { "grad_norm": 0.2611801326274872, "learning_rate": 7.943828945429652e-05, "loss": 0.0527, "step": 10040 }, { "grad_norm": 0.2806190252304077, "learning_rate": 7.939372138666827e-05, "loss": 0.0495, "step": 10050 }, { "grad_norm": 0.26162877678871155, "learning_rate": 7.934911760289692e-05, "loss": 0.0498, "step": 10060 }, { "grad_norm": 0.2220834195613861, "learning_rate": 7.930447815718022e-05, "loss": 0.0339, "step": 10070 }, { "grad_norm": 0.21768777072429657, "learning_rate": 7.925980310375934e-05, "loss": 0.0442, "step": 10080 }, { "grad_norm": 0.25125011801719666, "learning_rate": 7.921509249691865e-05, "loss": 0.0443, "step": 10090 }, { "grad_norm": 0.21868425607681274, "learning_rate": 7.917034639098579e-05, "loss": 0.0485, "step": 10100 }, { "grad_norm": 0.18830092251300812, "learning_rate": 7.912556484033146e-05, "loss": 0.0325, "step": 10110 }, { "grad_norm": 0.24465546011924744, "learning_rate": 7.908074789936952e-05, "loss": 0.0379, "step": 10120 }, { "grad_norm": 0.2569633722305298, "learning_rate": 7.903589562255673e-05, "loss": 0.0485, "step": 10130 }, { "grad_norm": 0.26118218898773193, "learning_rate": 7.899100806439285e-05, "loss": 0.0392, "step": 10140 }, { "grad_norm": 0.24857108294963837, "learning_rate": 7.894608527942049e-05, "loss": 0.0497, "step": 10150 }, { "grad_norm": 0.25448358058929443, "learning_rate": 7.89011273222251e-05, "loss": 0.0453, "step": 10160 }, { "grad_norm": 0.18138442933559418, "learning_rate": 7.885613424743482e-05, "loss": 0.0392, "step": 10170 }, { "grad_norm": 0.30224618315696716, "learning_rate": 7.881110610972044e-05, "loss": 0.0572, "step": 10180 }, { "grad_norm": 0.2780274450778961, "learning_rate": 7.876604296379544e-05, "loss": 0.0579, "step": 10190 }, { "grad_norm": 0.2592679560184479, "learning_rate": 7.872094486441579e-05, "loss": 0.0485, "step": 10200 }, { "grad_norm": 0.177821084856987, "learning_rate": 7.867581186637991e-05, "loss": 0.0491, "step": 10210 }, { "grad_norm": 0.252866268157959, "learning_rate": 7.863064402452866e-05, "loss": 0.0404, "step": 10220 }, { "grad_norm": 0.2094162553548813, "learning_rate": 7.858544139374524e-05, "loss": 0.0477, "step": 10230 }, { "grad_norm": 0.27572962641716003, "learning_rate": 7.854020402895508e-05, "loss": 0.045, "step": 10240 }, { "grad_norm": 0.23872201144695282, "learning_rate": 7.849493198512587e-05, "loss": 0.0414, "step": 10250 }, { "grad_norm": 0.2604624927043915, "learning_rate": 7.84496253172674e-05, "loss": 0.0371, "step": 10260 }, { "grad_norm": 0.2773470878601074, "learning_rate": 7.840428408043155e-05, "loss": 0.0509, "step": 10270 }, { "grad_norm": 0.3006497919559479, "learning_rate": 7.835890832971218e-05, "loss": 0.0476, "step": 10280 }, { "grad_norm": 0.27423354983329773, "learning_rate": 7.831349812024512e-05, "loss": 0.0536, "step": 10290 }, { "grad_norm": 0.3522675335407257, "learning_rate": 7.826805350720807e-05, "loss": 0.0446, "step": 10300 }, { "grad_norm": 0.2906223237514496, "learning_rate": 7.822257454582049e-05, "loss": 0.0395, "step": 10310 }, { "grad_norm": 0.2504901587963104, "learning_rate": 7.817706129134363e-05, "loss": 0.0504, "step": 10320 }, { "grad_norm": 0.30543819069862366, "learning_rate": 7.813151379908036e-05, "loss": 0.0496, "step": 10330 }, { "grad_norm": 0.30536848306655884, "learning_rate": 7.808593212437523e-05, "loss": 0.0491, "step": 10340 }, { "grad_norm": 0.225987508893013, "learning_rate": 7.804031632261421e-05, "loss": 0.0552, "step": 10350 }, { "grad_norm": 0.248750701546669, "learning_rate": 7.799466644922484e-05, "loss": 0.0399, "step": 10360 }, { "grad_norm": 0.2396802008152008, "learning_rate": 7.794898255967602e-05, "loss": 0.0406, "step": 10370 }, { "grad_norm": 0.20537549257278442, "learning_rate": 7.790326470947795e-05, "loss": 0.0423, "step": 10380 }, { "grad_norm": 0.3009016811847687, "learning_rate": 7.785751295418217e-05, "loss": 0.0443, "step": 10390 }, { "grad_norm": 0.3260253369808197, "learning_rate": 7.781172734938136e-05, "loss": 0.0477, "step": 10400 }, { "grad_norm": 0.27009686827659607, "learning_rate": 7.776590795070933e-05, "loss": 0.0448, "step": 10410 }, { "grad_norm": 0.19967836141586304, "learning_rate": 7.772005481384099e-05, "loss": 0.0516, "step": 10420 }, { "grad_norm": 0.23688632249832153, "learning_rate": 7.767416799449223e-05, "loss": 0.0363, "step": 10430 }, { "grad_norm": 0.26364654302597046, "learning_rate": 7.762824754841984e-05, "loss": 0.0397, "step": 10440 }, { "grad_norm": 0.21969135105609894, "learning_rate": 7.758229353142152e-05, "loss": 0.0511, "step": 10450 }, { "grad_norm": 0.2860223650932312, "learning_rate": 7.753630599933572e-05, "loss": 0.0409, "step": 10460 }, { "grad_norm": 0.22365671396255493, "learning_rate": 7.749028500804161e-05, "loss": 0.0416, "step": 10470 }, { "grad_norm": 0.22721393406391144, "learning_rate": 7.744423061345906e-05, "loss": 0.0382, "step": 10480 }, { "grad_norm": 0.1891147643327713, "learning_rate": 7.73981428715485e-05, "loss": 0.0396, "step": 10490 }, { "grad_norm": 0.27045831084251404, "learning_rate": 7.735202183831085e-05, "loss": 0.0379, "step": 10500 }, { "grad_norm": 0.2657162845134735, "learning_rate": 7.730586756978757e-05, "loss": 0.0521, "step": 10510 }, { "grad_norm": 0.1984698623418808, "learning_rate": 7.72596801220604e-05, "loss": 0.0348, "step": 10520 }, { "grad_norm": 0.27833282947540283, "learning_rate": 7.721345955125147e-05, "loss": 0.0499, "step": 10530 }, { "grad_norm": 0.25403696298599243, "learning_rate": 7.71672059135231e-05, "loss": 0.0415, "step": 10540 }, { "grad_norm": 0.2516438364982605, "learning_rate": 7.712091926507787e-05, "loss": 0.0367, "step": 10550 }, { "grad_norm": 0.20605601370334625, "learning_rate": 7.70745996621584e-05, "loss": 0.0338, "step": 10560 }, { "grad_norm": 0.25127294659614563, "learning_rate": 7.702824716104735e-05, "loss": 0.0357, "step": 10570 }, { "grad_norm": 0.27902910113334656, "learning_rate": 7.698186181806743e-05, "loss": 0.0429, "step": 10580 }, { "grad_norm": 0.24001774191856384, "learning_rate": 7.693544368958116e-05, "loss": 0.0441, "step": 10590 }, { "grad_norm": 0.16792209446430206, "learning_rate": 7.688899283199096e-05, "loss": 0.0551, "step": 10600 }, { "grad_norm": 0.22789961099624634, "learning_rate": 7.684250930173901e-05, "loss": 0.0691, "step": 10610 }, { "grad_norm": 0.28557243943214417, "learning_rate": 7.679599315530718e-05, "loss": 0.0586, "step": 10620 }, { "grad_norm": 0.24823269248008728, "learning_rate": 7.674944444921695e-05, "loss": 0.0399, "step": 10630 }, { "grad_norm": 0.24957014620304108, "learning_rate": 7.670286324002944e-05, "loss": 0.0459, "step": 10640 }, { "grad_norm": 0.2336244434118271, "learning_rate": 7.665624958434514e-05, "loss": 0.0492, "step": 10650 }, { "grad_norm": 0.24939082562923431, "learning_rate": 7.66096035388041e-05, "loss": 0.05, "step": 10660 }, { "grad_norm": 0.2778778374195099, "learning_rate": 7.656292516008563e-05, "loss": 0.0486, "step": 10670 }, { "grad_norm": 0.1928761601448059, "learning_rate": 7.651621450490837e-05, "loss": 0.034, "step": 10680 }, { "grad_norm": 0.21475808322429657, "learning_rate": 7.646947163003017e-05, "loss": 0.0351, "step": 10690 }, { "grad_norm": 0.2870038151741028, "learning_rate": 7.642269659224803e-05, "loss": 0.0663, "step": 10700 }, { "grad_norm": 0.2897229790687561, "learning_rate": 7.637588944839803e-05, "loss": 0.0431, "step": 10710 }, { "grad_norm": 0.16810865700244904, "learning_rate": 7.632905025535528e-05, "loss": 0.0401, "step": 10720 }, { "grad_norm": 0.23614375293254852, "learning_rate": 7.628217907003378e-05, "loss": 0.0375, "step": 10730 }, { "grad_norm": 0.19215163588523865, "learning_rate": 7.623527594938649e-05, "loss": 0.0332, "step": 10740 }, { "grad_norm": 0.2940616011619568, "learning_rate": 7.618834095040509e-05, "loss": 0.0446, "step": 10750 }, { "grad_norm": 0.2541223168373108, "learning_rate": 7.614137413012e-05, "loss": 0.0485, "step": 10760 }, { "grad_norm": 0.28154581785202026, "learning_rate": 7.609437554560042e-05, "loss": 0.0483, "step": 10770 }, { "grad_norm": 0.265682578086853, "learning_rate": 7.604734525395398e-05, "loss": 0.0491, "step": 10780 }, { "grad_norm": 0.2652435004711151, "learning_rate": 7.600028331232699e-05, "loss": 0.0448, "step": 10790 }, { "grad_norm": 0.317158579826355, "learning_rate": 7.595318977790408e-05, "loss": 0.066, "step": 10800 }, { "grad_norm": 0.2686416804790497, "learning_rate": 7.590606470790836e-05, "loss": 0.0427, "step": 10810 }, { "grad_norm": 0.2857540547847748, "learning_rate": 7.585890815960125e-05, "loss": 0.0515, "step": 10820 }, { "grad_norm": 0.3129342198371887, "learning_rate": 7.581172019028238e-05, "loss": 0.0514, "step": 10830 }, { "grad_norm": 0.2064976990222931, "learning_rate": 7.576450085728959e-05, "loss": 0.0578, "step": 10840 }, { "grad_norm": 0.2766593396663666, "learning_rate": 7.571725021799885e-05, "loss": 0.0469, "step": 10850 }, { "grad_norm": 0.26029807329177856, "learning_rate": 7.566996832982409e-05, "loss": 0.0362, "step": 10860 }, { "grad_norm": 0.2625162899494171, "learning_rate": 7.56226552502173e-05, "loss": 0.0456, "step": 10870 }, { "grad_norm": 0.22911669313907623, "learning_rate": 7.557531103666833e-05, "loss": 0.0511, "step": 10880 }, { "grad_norm": 0.24267111718654633, "learning_rate": 7.552793574670485e-05, "loss": 0.044, "step": 10890 }, { "grad_norm": 0.2638726532459259, "learning_rate": 7.54805294378923e-05, "loss": 0.0456, "step": 10900 }, { "grad_norm": 0.2676328420639038, "learning_rate": 7.543309216783384e-05, "loss": 0.0386, "step": 10910 }, { "grad_norm": 0.28208452463150024, "learning_rate": 7.53856239941702e-05, "loss": 0.0402, "step": 10920 }, { "grad_norm": 0.21513724327087402, "learning_rate": 7.533812497457972e-05, "loss": 0.0531, "step": 10930 }, { "grad_norm": 0.2702227830886841, "learning_rate": 7.529059516677814e-05, "loss": 0.0508, "step": 10940 }, { "grad_norm": 0.2964681088924408, "learning_rate": 7.524303462851872e-05, "loss": 0.0419, "step": 10950 }, { "grad_norm": 0.20003049075603485, "learning_rate": 7.519544341759192e-05, "loss": 0.0349, "step": 10960 }, { "grad_norm": 0.2265021950006485, "learning_rate": 7.514782159182562e-05, "loss": 0.0386, "step": 10970 }, { "grad_norm": 0.17672517895698547, "learning_rate": 7.510016920908481e-05, "loss": 0.0388, "step": 10980 }, { "grad_norm": 0.24378830194473267, "learning_rate": 7.505248632727158e-05, "loss": 0.0385, "step": 10990 }, { "grad_norm": 0.20270298421382904, "learning_rate": 7.50047730043252e-05, "loss": 0.0337, "step": 11000 }, { "grad_norm": 0.21337686479091644, "learning_rate": 7.495702929822183e-05, "loss": 0.04, "step": 11010 }, { "grad_norm": 0.22308598458766937, "learning_rate": 7.490925526697455e-05, "loss": 0.0463, "step": 11020 }, { "grad_norm": 0.3268086612224579, "learning_rate": 7.486145096863334e-05, "loss": 0.0443, "step": 11030 }, { "grad_norm": 0.22342976927757263, "learning_rate": 7.481361646128491e-05, "loss": 0.0353, "step": 11040 }, { "grad_norm": 0.20908647775650024, "learning_rate": 7.476575180305271e-05, "loss": 0.0366, "step": 11050 }, { "grad_norm": 0.2579111158847809, "learning_rate": 7.471785705209682e-05, "loss": 0.047, "step": 11060 }, { "grad_norm": 0.2698732316493988, "learning_rate": 7.466993226661387e-05, "loss": 0.0351, "step": 11070 }, { "grad_norm": 0.18085768818855286, "learning_rate": 7.462197750483699e-05, "loss": 0.0386, "step": 11080 }, { "grad_norm": 0.23768162727355957, "learning_rate": 7.457399282503574e-05, "loss": 0.039, "step": 11090 }, { "grad_norm": 0.2124752253293991, "learning_rate": 7.452597828551604e-05, "loss": 0.0446, "step": 11100 }, { "grad_norm": 0.24273571372032166, "learning_rate": 7.447793394462006e-05, "loss": 0.0451, "step": 11110 }, { "grad_norm": 0.22022798657417297, "learning_rate": 7.442985986072624e-05, "loss": 0.042, "step": 11120 }, { "grad_norm": 0.22581815719604492, "learning_rate": 7.438175609224908e-05, "loss": 0.0515, "step": 11130 }, { "grad_norm": 0.23451009392738342, "learning_rate": 7.433362269763924e-05, "loss": 0.0379, "step": 11140 }, { "grad_norm": 0.21864807605743408, "learning_rate": 7.428545973538329e-05, "loss": 0.0311, "step": 11150 }, { "grad_norm": 0.20803111791610718, "learning_rate": 7.42372672640038e-05, "loss": 0.0414, "step": 11160 }, { "grad_norm": 0.24474173784255981, "learning_rate": 7.418904534205917e-05, "loss": 0.0342, "step": 11170 }, { "grad_norm": 0.28393539786338806, "learning_rate": 7.414079402814356e-05, "loss": 0.0404, "step": 11180 }, { "grad_norm": 0.22944341599941254, "learning_rate": 7.409251338088687e-05, "loss": 0.0482, "step": 11190 }, { "grad_norm": 0.2568800151348114, "learning_rate": 7.404420345895467e-05, "loss": 0.0423, "step": 11200 }, { "grad_norm": 0.23763087391853333, "learning_rate": 7.399586432104804e-05, "loss": 0.0401, "step": 11210 }, { "grad_norm": 0.21309614181518555, "learning_rate": 7.394749602590358e-05, "loss": 0.0399, "step": 11220 }, { "grad_norm": 0.2498435378074646, "learning_rate": 7.389909863229336e-05, "loss": 0.0483, "step": 11230 }, { "grad_norm": 0.20342831313610077, "learning_rate": 7.385067219902477e-05, "loss": 0.0363, "step": 11240 }, { "grad_norm": 0.24537654221057892, "learning_rate": 7.380221678494049e-05, "loss": 0.0354, "step": 11250 }, { "grad_norm": 0.2454153448343277, "learning_rate": 7.37537324489184e-05, "loss": 0.03, "step": 11260 }, { "grad_norm": 0.22627699375152588, "learning_rate": 7.370521924987155e-05, "loss": 0.0402, "step": 11270 }, { "grad_norm": 0.25349244475364685, "learning_rate": 7.365667724674806e-05, "loss": 0.0526, "step": 11280 }, { "grad_norm": 0.30053776502609253, "learning_rate": 7.360810649853105e-05, "loss": 0.0436, "step": 11290 }, { "grad_norm": 0.21177314221858978, "learning_rate": 7.355950706423853e-05, "loss": 0.0395, "step": 11300 }, { "grad_norm": 0.27116310596466064, "learning_rate": 7.351087900292342e-05, "loss": 0.0377, "step": 11310 }, { "grad_norm": 0.18429580330848694, "learning_rate": 7.346222237367339e-05, "loss": 0.04, "step": 11320 }, { "grad_norm": 0.21463339030742645, "learning_rate": 7.341353723561081e-05, "loss": 0.0392, "step": 11330 }, { "grad_norm": 0.21680587530136108, "learning_rate": 7.336482364789277e-05, "loss": 0.0315, "step": 11340 }, { "grad_norm": 0.18615351617336273, "learning_rate": 7.331608166971082e-05, "loss": 0.0556, "step": 11350 }, { "grad_norm": 0.24947886168956757, "learning_rate": 7.326731136029108e-05, "loss": 0.0369, "step": 11360 }, { "grad_norm": 0.18601466715335846, "learning_rate": 7.321851277889408e-05, "loss": 0.0298, "step": 11370 }, { "grad_norm": 0.2520013451576233, "learning_rate": 7.316968598481469e-05, "loss": 0.0395, "step": 11380 }, { "grad_norm": 0.23677198588848114, "learning_rate": 7.312083103738206e-05, "loss": 0.0449, "step": 11390 }, { "grad_norm": 0.2398114800453186, "learning_rate": 7.307194799595958e-05, "loss": 0.0485, "step": 11400 }, { "grad_norm": 0.29467156529426575, "learning_rate": 7.302303691994473e-05, "loss": 0.0357, "step": 11410 }, { "grad_norm": 0.16948124766349792, "learning_rate": 7.29740978687691e-05, "loss": 0.0531, "step": 11420 }, { "grad_norm": 0.19902008771896362, "learning_rate": 7.292513090189825e-05, "loss": 0.0354, "step": 11430 }, { "grad_norm": 0.1696200966835022, "learning_rate": 7.287613607883163e-05, "loss": 0.0343, "step": 11440 }, { "grad_norm": 0.21216900646686554, "learning_rate": 7.282711345910263e-05, "loss": 0.0379, "step": 11450 }, { "grad_norm": 0.2653067409992218, "learning_rate": 7.27780631022783e-05, "loss": 0.0493, "step": 11460 }, { "grad_norm": 0.23933789134025574, "learning_rate": 7.272898506795948e-05, "loss": 0.0301, "step": 11470 }, { "grad_norm": 0.24619582295417786, "learning_rate": 7.267987941578058e-05, "loss": 0.0354, "step": 11480 }, { "grad_norm": 0.2665177881717682, "learning_rate": 7.263074620540962e-05, "loss": 0.0454, "step": 11490 }, { "grad_norm": 0.24458494782447815, "learning_rate": 7.25815854965481e-05, "loss": 0.0405, "step": 11500 }, { "grad_norm": 0.2463536113500595, "learning_rate": 7.253239734893089e-05, "loss": 0.0316, "step": 11510 }, { "grad_norm": 0.19586481153964996, "learning_rate": 7.248318182232623e-05, "loss": 0.0338, "step": 11520 }, { "grad_norm": 0.19392995536327362, "learning_rate": 7.243393897653565e-05, "loss": 0.0329, "step": 11530 }, { "grad_norm": 0.2108003944158554, "learning_rate": 7.23846688713938e-05, "loss": 0.04, "step": 11540 }, { "grad_norm": 0.23535338044166565, "learning_rate": 7.233537156676854e-05, "loss": 0.0468, "step": 11550 }, { "grad_norm": 0.1808994710445404, "learning_rate": 7.228604712256076e-05, "loss": 0.035, "step": 11560 }, { "grad_norm": 0.23540198802947998, "learning_rate": 7.223669559870426e-05, "loss": 0.0347, "step": 11570 }, { "grad_norm": 0.24007359147071838, "learning_rate": 7.218731705516585e-05, "loss": 0.0414, "step": 11580 }, { "grad_norm": 0.23431654274463654, "learning_rate": 7.21379115519451e-05, "loss": 0.0456, "step": 11590 }, { "grad_norm": 0.1845175176858902, "learning_rate": 7.20884791490743e-05, "loss": 0.0345, "step": 11600 }, { "grad_norm": 0.2715092599391937, "learning_rate": 7.203901990661857e-05, "loss": 0.0563, "step": 11610 }, { "grad_norm": 0.4943791329860687, "learning_rate": 7.198953388467549e-05, "loss": 0.0559, "step": 11620 }, { "grad_norm": 0.24055415391921997, "learning_rate": 7.194002114337526e-05, "loss": 0.0615, "step": 11630 }, { "grad_norm": 0.2225314825773239, "learning_rate": 7.189048174288054e-05, "loss": 0.0357, "step": 11640 }, { "grad_norm": 0.22788000106811523, "learning_rate": 7.184091574338636e-05, "loss": 0.0404, "step": 11650 }, { "grad_norm": 0.21855661273002625, "learning_rate": 7.179132320512009e-05, "loss": 0.0405, "step": 11660 }, { "grad_norm": 0.1644456386566162, "learning_rate": 7.174170418834134e-05, "loss": 0.0312, "step": 11670 }, { "grad_norm": 0.24407050013542175, "learning_rate": 7.169205875334189e-05, "loss": 0.0398, "step": 11680 }, { "grad_norm": 0.20870651304721832, "learning_rate": 7.164238696044562e-05, "loss": 0.0363, "step": 11690 }, { "grad_norm": 0.24472561478614807, "learning_rate": 7.159268887000846e-05, "loss": 0.0374, "step": 11700 }, { "grad_norm": 0.21102306246757507, "learning_rate": 7.154296454241827e-05, "loss": 0.038, "step": 11710 }, { "grad_norm": 0.2249128818511963, "learning_rate": 7.149321403809479e-05, "loss": 0.0406, "step": 11720 }, { "grad_norm": 0.2045842856168747, "learning_rate": 7.144343741748956e-05, "loss": 0.0318, "step": 11730 }, { "grad_norm": 0.22144712507724762, "learning_rate": 7.139363474108589e-05, "loss": 0.0382, "step": 11740 }, { "grad_norm": 0.22706323862075806, "learning_rate": 7.134380606939874e-05, "loss": 0.0429, "step": 11750 }, { "grad_norm": 0.23560352623462677, "learning_rate": 7.12939514629746e-05, "loss": 0.0381, "step": 11760 }, { "grad_norm": 0.28658241033554077, "learning_rate": 7.124407098239155e-05, "loss": 0.0408, "step": 11770 }, { "grad_norm": 0.2505757510662079, "learning_rate": 7.119416468825908e-05, "loss": 0.0335, "step": 11780 }, { "grad_norm": 0.22469937801361084, "learning_rate": 7.114423264121804e-05, "loss": 0.0339, "step": 11790 }, { "grad_norm": 0.20082518458366394, "learning_rate": 7.109427490194056e-05, "loss": 0.0383, "step": 11800 }, { "grad_norm": 0.24056535959243774, "learning_rate": 7.104429153113001e-05, "loss": 0.05, "step": 11810 }, { "grad_norm": 0.2954394519329071, "learning_rate": 7.099428258952092e-05, "loss": 0.0403, "step": 11820 }, { "grad_norm": 0.2527979612350464, "learning_rate": 7.094424813787883e-05, "loss": 0.0527, "step": 11830 }, { "grad_norm": 0.26914769411087036, "learning_rate": 7.089418823700034e-05, "loss": 0.0385, "step": 11840 }, { "grad_norm": 0.24587085843086243, "learning_rate": 7.084410294771298e-05, "loss": 0.0391, "step": 11850 }, { "grad_norm": 0.19684584438800812, "learning_rate": 7.079399233087504e-05, "loss": 0.0428, "step": 11860 }, { "grad_norm": 0.2711717486381531, "learning_rate": 7.074385644737568e-05, "loss": 0.0358, "step": 11870 }, { "grad_norm": 0.2524182200431824, "learning_rate": 7.069369535813473e-05, "loss": 0.0508, "step": 11880 }, { "grad_norm": 0.23026691377162933, "learning_rate": 7.06435091241026e-05, "loss": 0.032, "step": 11890 }, { "grad_norm": 0.254262775182724, "learning_rate": 7.059329780626034e-05, "loss": 0.0394, "step": 11900 }, { "grad_norm": 0.24266617000102997, "learning_rate": 7.054306146561944e-05, "loss": 0.0403, "step": 11910 }, { "grad_norm": 0.17615970969200134, "learning_rate": 7.049280016322178e-05, "loss": 0.0308, "step": 11920 }, { "grad_norm": 0.25071993470191956, "learning_rate": 7.044251396013957e-05, "loss": 0.0404, "step": 11930 }, { "grad_norm": 0.21976354718208313, "learning_rate": 7.039220291747528e-05, "loss": 0.0347, "step": 11940 }, { "grad_norm": 0.22326399385929108, "learning_rate": 7.034186709636159e-05, "loss": 0.031, "step": 11950 }, { "grad_norm": 0.2497735470533371, "learning_rate": 7.029150655796129e-05, "loss": 0.0341, "step": 11960 }, { "grad_norm": 0.23633377254009247, "learning_rate": 7.024112136346712e-05, "loss": 0.0385, "step": 11970 }, { "grad_norm": 0.2755773067474365, "learning_rate": 7.01907115741019e-05, "loss": 0.039, "step": 11980 }, { "grad_norm": 0.22025923430919647, "learning_rate": 7.014027725111825e-05, "loss": 0.0415, "step": 11990 }, { "grad_norm": 0.30035513639450073, "learning_rate": 7.008981845579864e-05, "loss": 0.0519, "step": 12000 }, { "grad_norm": 0.2716161608695984, "learning_rate": 7.003933524945528e-05, "loss": 0.0411, "step": 12010 }, { "grad_norm": 0.26304319500923157, "learning_rate": 6.998882769342998e-05, "loss": 0.0359, "step": 12020 }, { "grad_norm": 0.2722817659378052, "learning_rate": 6.993829584909423e-05, "loss": 0.0339, "step": 12030 }, { "grad_norm": 0.2567330598831177, "learning_rate": 6.988773977784895e-05, "loss": 0.0436, "step": 12040 }, { "grad_norm": 0.25035011768341064, "learning_rate": 6.983715954112454e-05, "loss": 0.0403, "step": 12050 }, { "grad_norm": 0.22804924845695496, "learning_rate": 6.978655520038079e-05, "loss": 0.0391, "step": 12060 }, { "grad_norm": 0.22187314927577972, "learning_rate": 6.97359268171067e-05, "loss": 0.0587, "step": 12070 }, { "grad_norm": 0.29983407258987427, "learning_rate": 6.968527445282056e-05, "loss": 0.0368, "step": 12080 }, { "grad_norm": 0.2270602583885193, "learning_rate": 6.963459816906977e-05, "loss": 0.0301, "step": 12090 }, { "grad_norm": 0.2328372448682785, "learning_rate": 6.958389802743077e-05, "loss": 0.0458, "step": 12100 }, { "grad_norm": 0.2344803661108017, "learning_rate": 6.953317408950903e-05, "loss": 0.0364, "step": 12110 }, { "grad_norm": 0.2034386694431305, "learning_rate": 6.94824264169389e-05, "loss": 0.0296, "step": 12120 }, { "grad_norm": 0.2114945650100708, "learning_rate": 6.94316550713836e-05, "loss": 0.0427, "step": 12130 }, { "grad_norm": 0.24942737817764282, "learning_rate": 6.938086011453513e-05, "loss": 0.0311, "step": 12140 }, { "grad_norm": 0.20851460099220276, "learning_rate": 6.93300416081141e-05, "loss": 0.0361, "step": 12150 }, { "grad_norm": 0.1744174063205719, "learning_rate": 6.927919961386984e-05, "loss": 0.0302, "step": 12160 }, { "grad_norm": 0.15538497269153595, "learning_rate": 6.922833419358013e-05, "loss": 0.0329, "step": 12170 }, { "grad_norm": 0.2193404585123062, "learning_rate": 6.917744540905124e-05, "loss": 0.045, "step": 12180 }, { "grad_norm": 0.24471823871135712, "learning_rate": 6.912653332211787e-05, "loss": 0.0331, "step": 12190 }, { "grad_norm": 0.20749132335186005, "learning_rate": 6.9075597994643e-05, "loss": 0.0456, "step": 12200 }, { "grad_norm": 0.21621958911418915, "learning_rate": 6.902463948851787e-05, "loss": 0.0565, "step": 12210 }, { "grad_norm": 0.2211291640996933, "learning_rate": 6.897365786566184e-05, "loss": 0.0358, "step": 12220 }, { "grad_norm": 0.24112826585769653, "learning_rate": 6.892265318802242e-05, "loss": 0.0389, "step": 12230 }, { "grad_norm": 0.17419292032718658, "learning_rate": 6.887162551757507e-05, "loss": 0.0328, "step": 12240 }, { "grad_norm": 0.241267129778862, "learning_rate": 6.882057491632326e-05, "loss": 0.0497, "step": 12250 }, { "grad_norm": 0.31025364995002747, "learning_rate": 6.876950144629824e-05, "loss": 0.0501, "step": 12260 }, { "grad_norm": 0.18689601123332977, "learning_rate": 6.87184051695591e-05, "loss": 0.0431, "step": 12270 }, { "grad_norm": 0.1674424558877945, "learning_rate": 6.866728614819268e-05, "loss": 0.035, "step": 12280 }, { "grad_norm": 0.21214507520198822, "learning_rate": 6.861614444431337e-05, "loss": 0.0441, "step": 12290 }, { "grad_norm": 0.29427072405815125, "learning_rate": 6.856498012006318e-05, "loss": 0.0288, "step": 12300 }, { "grad_norm": 0.23155993223190308, "learning_rate": 6.851379323761157e-05, "loss": 0.038, "step": 12310 }, { "grad_norm": 0.24062618613243103, "learning_rate": 6.846258385915545e-05, "loss": 0.0462, "step": 12320 }, { "grad_norm": 0.25328630208969116, "learning_rate": 6.841135204691902e-05, "loss": 0.0325, "step": 12330 }, { "grad_norm": 0.24309873580932617, "learning_rate": 6.836009786315377e-05, "loss": 0.0421, "step": 12340 }, { "grad_norm": 0.17222735285758972, "learning_rate": 6.830882137013839e-05, "loss": 0.0353, "step": 12350 }, { "grad_norm": 0.22162985801696777, "learning_rate": 6.825752263017863e-05, "loss": 0.0449, "step": 12360 }, { "grad_norm": 0.2479204535484314, "learning_rate": 6.820620170560731e-05, "loss": 0.0314, "step": 12370 }, { "grad_norm": 0.19916625320911407, "learning_rate": 6.815485865878418e-05, "loss": 0.0404, "step": 12380 }, { "grad_norm": 0.21626994013786316, "learning_rate": 6.810349355209588e-05, "loss": 0.0404, "step": 12390 }, { "grad_norm": 0.2153671234846115, "learning_rate": 6.805210644795588e-05, "loss": 0.0446, "step": 12400 }, { "grad_norm": 0.2895190119743347, "learning_rate": 6.800069740880433e-05, "loss": 0.0422, "step": 12410 }, { "grad_norm": 0.25635990500450134, "learning_rate": 6.794926649710807e-05, "loss": 0.0296, "step": 12420 }, { "grad_norm": 0.2938802242279053, "learning_rate": 6.789781377536052e-05, "loss": 0.0462, "step": 12430 }, { "grad_norm": 0.27418097853660583, "learning_rate": 6.784633930608158e-05, "loss": 0.0479, "step": 12440 }, { "grad_norm": 0.26554492115974426, "learning_rate": 6.779484315181759e-05, "loss": 0.0372, "step": 12450 }, { "grad_norm": 0.23115956783294678, "learning_rate": 6.774332537514122e-05, "loss": 0.0427, "step": 12460 }, { "grad_norm": 0.19176499545574188, "learning_rate": 6.769178603865143e-05, "loss": 0.035, "step": 12470 }, { "grad_norm": 0.20709866285324097, "learning_rate": 6.764022520497337e-05, "loss": 0.0284, "step": 12480 }, { "grad_norm": 0.22247517108917236, "learning_rate": 6.758864293675833e-05, "loss": 0.0317, "step": 12490 }, { "grad_norm": 0.19083453714847565, "learning_rate": 6.753703929668362e-05, "loss": 0.0395, "step": 12500 }, { "grad_norm": 0.17470797896385193, "learning_rate": 6.748541434745253e-05, "loss": 0.0504, "step": 12510 }, { "grad_norm": 0.24933850765228271, "learning_rate": 6.743376815179424e-05, "loss": 0.0362, "step": 12520 }, { "grad_norm": 0.21933867037296295, "learning_rate": 6.738210077246376e-05, "loss": 0.0342, "step": 12530 }, { "grad_norm": 0.24607732892036438, "learning_rate": 6.733041227224181e-05, "loss": 0.0453, "step": 12540 }, { "grad_norm": 0.27121445536613464, "learning_rate": 6.72787027139348e-05, "loss": 0.0446, "step": 12550 }, { "grad_norm": 0.22388683259487152, "learning_rate": 6.72269721603747e-05, "loss": 0.0317, "step": 12560 }, { "grad_norm": 0.26443353295326233, "learning_rate": 6.717522067441904e-05, "loss": 0.0395, "step": 12570 }, { "grad_norm": 0.2077624648809433, "learning_rate": 6.712344831895074e-05, "loss": 0.0355, "step": 12580 }, { "grad_norm": 0.3406327962875366, "learning_rate": 6.707165515687811e-05, "loss": 0.046, "step": 12590 }, { "grad_norm": 0.23670122027397156, "learning_rate": 6.70198412511347e-05, "loss": 0.0392, "step": 12600 }, { "grad_norm": 0.3246597349643707, "learning_rate": 6.69680066646793e-05, "loss": 0.0325, "step": 12610 }, { "grad_norm": 0.17881350219249725, "learning_rate": 6.691615146049583e-05, "loss": 0.0355, "step": 12620 }, { "grad_norm": 0.2141602486371994, "learning_rate": 6.686427570159324e-05, "loss": 0.0264, "step": 12630 }, { "grad_norm": 0.21175578236579895, "learning_rate": 6.681237945100548e-05, "loss": 0.0485, "step": 12640 }, { "grad_norm": 0.2573133707046509, "learning_rate": 6.676046277179139e-05, "loss": 0.0367, "step": 12650 }, { "grad_norm": 0.21999990940093994, "learning_rate": 6.670852572703462e-05, "loss": 0.0466, "step": 12660 }, { "grad_norm": 0.20427608489990234, "learning_rate": 6.665656837984359e-05, "loss": 0.0362, "step": 12670 }, { "grad_norm": 0.1901365965604782, "learning_rate": 6.660459079335135e-05, "loss": 0.043, "step": 12680 }, { "grad_norm": 0.19009681046009064, "learning_rate": 6.655259303071558e-05, "loss": 0.0377, "step": 12690 }, { "grad_norm": 0.216633602976799, "learning_rate": 6.650057515511848e-05, "loss": 0.0331, "step": 12700 }, { "grad_norm": 0.3249838948249817, "learning_rate": 6.644853722976667e-05, "loss": 0.0327, "step": 12710 }, { "grad_norm": 0.2593340277671814, "learning_rate": 6.639647931789113e-05, "loss": 0.0365, "step": 12720 }, { "grad_norm": 0.2197743058204651, "learning_rate": 6.634440148274713e-05, "loss": 0.0368, "step": 12730 }, { "grad_norm": 0.22154894471168518, "learning_rate": 6.629230378761415e-05, "loss": 0.0392, "step": 12740 }, { "grad_norm": 0.19753877818584442, "learning_rate": 6.624018629579581e-05, "loss": 0.0268, "step": 12750 }, { "grad_norm": 0.2964731752872467, "learning_rate": 6.618804907061976e-05, "loss": 0.0504, "step": 12760 }, { "grad_norm": 0.19998621940612793, "learning_rate": 6.613589217543766e-05, "loss": 0.0482, "step": 12770 }, { "grad_norm": 0.22149795293807983, "learning_rate": 6.608371567362504e-05, "loss": 0.0392, "step": 12780 }, { "grad_norm": 0.23694419860839844, "learning_rate": 6.60315196285813e-05, "loss": 0.0388, "step": 12790 }, { "grad_norm": 0.18529345095157623, "learning_rate": 6.597930410372954e-05, "loss": 0.0378, "step": 12800 }, { "grad_norm": 0.24495136737823486, "learning_rate": 6.592706916251653e-05, "loss": 0.0375, "step": 12810 }, { "grad_norm": 0.29078438878059387, "learning_rate": 6.587481486841267e-05, "loss": 0.0444, "step": 12820 }, { "grad_norm": 0.2644461393356323, "learning_rate": 6.582254128491184e-05, "loss": 0.0409, "step": 12830 }, { "grad_norm": 0.18726222217082977, "learning_rate": 6.577024847553139e-05, "loss": 0.032, "step": 12840 }, { "grad_norm": 0.20037280023097992, "learning_rate": 6.571793650381201e-05, "loss": 0.0343, "step": 12850 }, { "grad_norm": 0.23505312204360962, "learning_rate": 6.566560543331766e-05, "loss": 0.0443, "step": 12860 }, { "grad_norm": 0.19942626357078552, "learning_rate": 6.561325532763554e-05, "loss": 0.0307, "step": 12870 }, { "grad_norm": 0.2573022246360779, "learning_rate": 6.556088625037597e-05, "loss": 0.0284, "step": 12880 }, { "grad_norm": 0.2541186809539795, "learning_rate": 6.55084982651723e-05, "loss": 0.0374, "step": 12890 }, { "grad_norm": 0.19334091246128082, "learning_rate": 6.54560914356809e-05, "loss": 0.0504, "step": 12900 }, { "grad_norm": 0.18890081346035004, "learning_rate": 6.540366582558097e-05, "loss": 0.0329, "step": 12910 }, { "grad_norm": 0.28007879853248596, "learning_rate": 6.53512214985746e-05, "loss": 0.0272, "step": 12920 }, { "grad_norm": 0.20009079575538635, "learning_rate": 6.529875851838658e-05, "loss": 0.0332, "step": 12930 }, { "grad_norm": 0.2381957471370697, "learning_rate": 6.52462769487644e-05, "loss": 0.0349, "step": 12940 }, { "grad_norm": 0.2105613499879837, "learning_rate": 6.519377685347807e-05, "loss": 0.0268, "step": 12950 }, { "grad_norm": 0.1945790797472, "learning_rate": 6.514125829632021e-05, "loss": 0.0283, "step": 12960 }, { "grad_norm": 0.22145922482013702, "learning_rate": 6.508872134110577e-05, "loss": 0.0307, "step": 12970 }, { "grad_norm": 0.194236621260643, "learning_rate": 6.503616605167213e-05, "loss": 0.0275, "step": 12980 }, { "grad_norm": 0.20272472500801086, "learning_rate": 6.498359249187893e-05, "loss": 0.0331, "step": 12990 }, { "grad_norm": 0.19808664917945862, "learning_rate": 6.493100072560799e-05, "loss": 0.0329, "step": 13000 }, { "grad_norm": 0.17202453315258026, "learning_rate": 6.487839081676327e-05, "loss": 0.0541, "step": 13010 }, { "grad_norm": 0.25600817799568176, "learning_rate": 6.482576282927076e-05, "loss": 0.0477, "step": 13020 }, { "grad_norm": 0.2314600646495819, "learning_rate": 6.477311682707845e-05, "loss": 0.0484, "step": 13030 }, { "grad_norm": 0.24856223165988922, "learning_rate": 6.472045287415616e-05, "loss": 0.0318, "step": 13040 }, { "grad_norm": 0.19380344450473785, "learning_rate": 6.466777103449558e-05, "loss": 0.029, "step": 13050 }, { "grad_norm": 0.20660191774368286, "learning_rate": 6.461507137211012e-05, "loss": 0.0379, "step": 13060 }, { "grad_norm": 0.26954641938209534, "learning_rate": 6.456235395103483e-05, "loss": 0.0519, "step": 13070 }, { "grad_norm": 0.22808924317359924, "learning_rate": 6.450961883532634e-05, "loss": 0.0354, "step": 13080 }, { "grad_norm": 0.21892490983009338, "learning_rate": 6.445686608906283e-05, "loss": 0.0365, "step": 13090 }, { "grad_norm": 0.24737608432769775, "learning_rate": 6.44040957763438e-05, "loss": 0.0416, "step": 13100 }, { "grad_norm": 0.2534080445766449, "learning_rate": 6.435130796129018e-05, "loss": 0.0302, "step": 13110 }, { "grad_norm": 0.26672855019569397, "learning_rate": 6.429850270804416e-05, "loss": 0.0316, "step": 13120 }, { "grad_norm": 0.24525567889213562, "learning_rate": 6.424568008076909e-05, "loss": 0.045, "step": 13130 }, { "grad_norm": 0.24042318761348724, "learning_rate": 6.419284014364943e-05, "loss": 0.0377, "step": 13140 }, { "grad_norm": 0.274423748254776, "learning_rate": 6.41399829608907e-05, "loss": 0.0375, "step": 13150 }, { "grad_norm": 0.23492196202278137, "learning_rate": 6.408710859671938e-05, "loss": 0.025, "step": 13160 }, { "grad_norm": 0.20374901592731476, "learning_rate": 6.403421711538278e-05, "loss": 0.0321, "step": 13170 }, { "grad_norm": 0.16325557231903076, "learning_rate": 6.398130858114903e-05, "loss": 0.0318, "step": 13180 }, { "grad_norm": 0.15487812459468842, "learning_rate": 6.392838305830701e-05, "loss": 0.0275, "step": 13190 }, { "grad_norm": 0.28143882751464844, "learning_rate": 6.387544061116621e-05, "loss": 0.0372, "step": 13200 }, { "grad_norm": 0.23125436902046204, "learning_rate": 6.382248130405671e-05, "loss": 0.0254, "step": 13210 }, { "grad_norm": 0.20818158984184265, "learning_rate": 6.376950520132905e-05, "loss": 0.0316, "step": 13220 }, { "grad_norm": 0.24970830976963043, "learning_rate": 6.371651236735418e-05, "loss": 0.0368, "step": 13230 }, { "grad_norm": 0.25157371163368225, "learning_rate": 6.366350286652341e-05, "loss": 0.047, "step": 13240 }, { "grad_norm": 0.2017275094985962, "learning_rate": 6.361047676324827e-05, "loss": 0.0329, "step": 13250 }, { "grad_norm": 0.28877535462379456, "learning_rate": 6.355743412196047e-05, "loss": 0.0339, "step": 13260 }, { "grad_norm": 0.21218827366828918, "learning_rate": 6.350437500711184e-05, "loss": 0.0328, "step": 13270 }, { "grad_norm": 0.2324109822511673, "learning_rate": 6.345129948317418e-05, "loss": 0.0332, "step": 13280 }, { "grad_norm": 0.22021619975566864, "learning_rate": 6.33982076146393e-05, "loss": 0.0351, "step": 13290 }, { "grad_norm": 0.15297672152519226, "learning_rate": 6.334509946601879e-05, "loss": 0.0273, "step": 13300 }, { "grad_norm": 0.19468866288661957, "learning_rate": 6.329197510184406e-05, "loss": 0.0417, "step": 13310 }, { "grad_norm": 0.2844182550907135, "learning_rate": 6.323883458666624e-05, "loss": 0.0434, "step": 13320 }, { "grad_norm": 0.24092581868171692, "learning_rate": 6.318567798505605e-05, "loss": 0.0346, "step": 13330 }, { "grad_norm": 0.18064098060131073, "learning_rate": 6.313250536160378e-05, "loss": 0.0351, "step": 13340 }, { "grad_norm": 0.26283884048461914, "learning_rate": 6.307931678091917e-05, "loss": 0.0326, "step": 13350 }, { "grad_norm": 0.18806573748588562, "learning_rate": 6.302611230763138e-05, "loss": 0.0385, "step": 13360 }, { "grad_norm": 0.2127745896577835, "learning_rate": 6.297289200638887e-05, "loss": 0.0432, "step": 13370 }, { "grad_norm": 0.25090956687927246, "learning_rate": 6.29196559418593e-05, "loss": 0.0367, "step": 13380 }, { "grad_norm": 0.2764403223991394, "learning_rate": 6.286640417872951e-05, "loss": 0.038, "step": 13390 }, { "grad_norm": 0.1490848958492279, "learning_rate": 6.281313678170542e-05, "loss": 0.0427, "step": 13400 }, { "grad_norm": 0.21525076031684875, "learning_rate": 6.275985381551194e-05, "loss": 0.0334, "step": 13410 }, { "grad_norm": 0.24265480041503906, "learning_rate": 6.270655534489292e-05, "loss": 0.0407, "step": 13420 }, { "grad_norm": 0.19496959447860718, "learning_rate": 6.265324143461098e-05, "loss": 0.0283, "step": 13430 }, { "grad_norm": 0.21141229569911957, "learning_rate": 6.259991214944758e-05, "loss": 0.034, "step": 13440 }, { "grad_norm": 0.24766430258750916, "learning_rate": 6.254656755420283e-05, "loss": 0.05, "step": 13450 }, { "grad_norm": 0.21125049889087677, "learning_rate": 6.249320771369545e-05, "loss": 0.0444, "step": 13460 }, { "grad_norm": 0.2124139815568924, "learning_rate": 6.243983269276263e-05, "loss": 0.0557, "step": 13470 }, { "grad_norm": 0.21224653720855713, "learning_rate": 6.238644255626012e-05, "loss": 0.0357, "step": 13480 }, { "grad_norm": 0.20648019015789032, "learning_rate": 6.233303736906193e-05, "loss": 0.0386, "step": 13490 }, { "grad_norm": 0.2699463367462158, "learning_rate": 6.227961719606039e-05, "loss": 0.0265, "step": 13500 }, { "grad_norm": 0.21248631179332733, "learning_rate": 6.222618210216608e-05, "loss": 0.0436, "step": 13510 }, { "grad_norm": 0.21012914180755615, "learning_rate": 6.217273215230767e-05, "loss": 0.0467, "step": 13520 }, { "grad_norm": 0.2842153012752533, "learning_rate": 6.211926741143188e-05, "loss": 0.0421, "step": 13530 }, { "grad_norm": 0.20694586634635925, "learning_rate": 6.20657879445034e-05, "loss": 0.0389, "step": 13540 }, { "grad_norm": 0.17912805080413818, "learning_rate": 6.201229381650485e-05, "loss": 0.028, "step": 13550 }, { "grad_norm": 0.21808524429798126, "learning_rate": 6.195878509243661e-05, "loss": 0.0434, "step": 13560 }, { "grad_norm": 0.21829964220523834, "learning_rate": 6.190526183731685e-05, "loss": 0.0326, "step": 13570 }, { "grad_norm": 0.21521972119808197, "learning_rate": 6.185172411618138e-05, "loss": 0.0336, "step": 13580 }, { "grad_norm": 0.18848109245300293, "learning_rate": 6.179817199408355e-05, "loss": 0.0435, "step": 13590 }, { "grad_norm": 0.21676728129386902, "learning_rate": 6.174460553609426e-05, "loss": 0.0563, "step": 13600 }, { "grad_norm": 0.16626447439193726, "learning_rate": 6.16910248073018e-05, "loss": 0.0308, "step": 13610 }, { "grad_norm": 0.24551330506801605, "learning_rate": 6.16374298728118e-05, "loss": 0.0444, "step": 13620 }, { "grad_norm": 0.25852853059768677, "learning_rate": 6.158382079774716e-05, "loss": 0.036, "step": 13630 }, { "grad_norm": 0.24059216678142548, "learning_rate": 6.153019764724799e-05, "loss": 0.0321, "step": 13640 }, { "grad_norm": 0.2241460233926773, "learning_rate": 6.147656048647144e-05, "loss": 0.0419, "step": 13650 }, { "grad_norm": 0.28821223974227905, "learning_rate": 6.142290938059173e-05, "loss": 0.0433, "step": 13660 }, { "grad_norm": 0.20681250095367432, "learning_rate": 6.136924439480001e-05, "loss": 0.0337, "step": 13670 }, { "grad_norm": 0.20991379022598267, "learning_rate": 6.13155655943043e-05, "loss": 0.0316, "step": 13680 }, { "grad_norm": 0.28373655676841736, "learning_rate": 6.126187304432941e-05, "loss": 0.0319, "step": 13690 }, { "grad_norm": 0.18512597680091858, "learning_rate": 6.120816681011684e-05, "loss": 0.0358, "step": 13700 }, { "grad_norm": 0.2205517739057541, "learning_rate": 6.115444695692474e-05, "loss": 0.0349, "step": 13710 }, { "grad_norm": 0.24638262391090393, "learning_rate": 6.110071355002779e-05, "loss": 0.0298, "step": 13720 }, { "grad_norm": 0.21366435289382935, "learning_rate": 6.104696665471714e-05, "loss": 0.0282, "step": 13730 }, { "grad_norm": 0.1688917875289917, "learning_rate": 6.099320633630036e-05, "loss": 0.0263, "step": 13740 }, { "grad_norm": 0.22755487263202667, "learning_rate": 6.093943266010128e-05, "loss": 0.033, "step": 13750 }, { "grad_norm": 0.2662801146507263, "learning_rate": 6.088564569146e-05, "loss": 0.0561, "step": 13760 }, { "grad_norm": 0.185906320810318, "learning_rate": 6.083184549573274e-05, "loss": 0.0306, "step": 13770 }, { "grad_norm": 0.17642588913440704, "learning_rate": 6.077803213829184e-05, "loss": 0.0311, "step": 13780 }, { "grad_norm": 0.20144300162792206, "learning_rate": 6.0724205684525594e-05, "loss": 0.0342, "step": 13790 }, { "grad_norm": 0.24424582719802856, "learning_rate": 6.0670366199838215e-05, "loss": 0.0297, "step": 13800 }, { "grad_norm": 0.15441343188285828, "learning_rate": 6.061651374964974e-05, "loss": 0.0352, "step": 13810 }, { "grad_norm": 0.22088730335235596, "learning_rate": 6.0562648399396004e-05, "loss": 0.0348, "step": 13820 }, { "grad_norm": 0.20948739349842072, "learning_rate": 6.050877021452845e-05, "loss": 0.0388, "step": 13830 }, { "grad_norm": 0.221237912774086, "learning_rate": 6.04548792605142e-05, "loss": 0.0377, "step": 13840 }, { "grad_norm": 0.21071971952915192, "learning_rate": 6.04009756028358e-05, "loss": 0.0296, "step": 13850 }, { "grad_norm": 0.20621587336063385, "learning_rate": 6.0347059306991294e-05, "loss": 0.0331, "step": 13860 }, { "grad_norm": 0.19958564639091492, "learning_rate": 6.029313043849407e-05, "loss": 0.0377, "step": 13870 }, { "grad_norm": 0.2112370878458023, "learning_rate": 6.023918906287279e-05, "loss": 0.0257, "step": 13880 }, { "grad_norm": 0.2724873125553131, "learning_rate": 6.01852352456713e-05, "loss": 0.0466, "step": 13890 }, { "grad_norm": 0.18394580483436584, "learning_rate": 6.013126905244858e-05, "loss": 0.026, "step": 13900 }, { "grad_norm": 0.18135543167591095, "learning_rate": 6.007729054877864e-05, "loss": 0.0341, "step": 13910 }, { "grad_norm": 0.279083251953125, "learning_rate": 6.0023299800250466e-05, "loss": 0.0417, "step": 13920 }, { "grad_norm": 0.2419212907552719, "learning_rate": 5.9969296872467894e-05, "loss": 0.0349, "step": 13930 }, { "grad_norm": 0.20985545217990875, "learning_rate": 5.9915281831049585e-05, "loss": 0.034, "step": 13940 }, { "grad_norm": 0.18044120073318481, "learning_rate": 5.98612547416289e-05, "loss": 0.0303, "step": 13950 }, { "grad_norm": 0.35703420639038086, "learning_rate": 5.9807215669853855e-05, "loss": 0.0356, "step": 13960 }, { "grad_norm": 0.2560860812664032, "learning_rate": 5.9753164681387e-05, "loss": 0.0556, "step": 13970 }, { "grad_norm": 0.20197473466396332, "learning_rate": 5.969910184190539e-05, "loss": 0.0407, "step": 13980 }, { "grad_norm": 0.18795058131217957, "learning_rate": 5.9645027217100477e-05, "loss": 0.0365, "step": 13990 }, { "grad_norm": 0.20213642716407776, "learning_rate": 5.959094087267804e-05, "loss": 0.0347, "step": 14000 }, { "grad_norm": 0.21438384056091309, "learning_rate": 5.9536842874358065e-05, "loss": 0.0325, "step": 14010 }, { "grad_norm": 0.2526093125343323, "learning_rate": 5.9482733287874734e-05, "loss": 0.0429, "step": 14020 }, { "grad_norm": 0.2645587921142578, "learning_rate": 5.942861217897631e-05, "loss": 0.0419, "step": 14030 }, { "grad_norm": 0.20881277322769165, "learning_rate": 5.9374479613425e-05, "loss": 0.0304, "step": 14040 }, { "grad_norm": 0.2747591435909271, "learning_rate": 5.932033565699704e-05, "loss": 0.0279, "step": 14050 }, { "grad_norm": 0.22347021102905273, "learning_rate": 5.926618037548237e-05, "loss": 0.0356, "step": 14060 }, { "grad_norm": 0.20071546733379364, "learning_rate": 5.9212013834684824e-05, "loss": 0.0356, "step": 14070 }, { "grad_norm": 0.19067145884037018, "learning_rate": 5.9157836100421824e-05, "loss": 0.0311, "step": 14080 }, { "grad_norm": 0.23311163485050201, "learning_rate": 5.910364723852444e-05, "loss": 0.0265, "step": 14090 }, { "grad_norm": 0.17106449604034424, "learning_rate": 5.904944731483723e-05, "loss": 0.0395, "step": 14100 }, { "grad_norm": 0.19753076136112213, "learning_rate": 5.899523639521825e-05, "loss": 0.0373, "step": 14110 }, { "grad_norm": 0.17817974090576172, "learning_rate": 5.894101454553883e-05, "loss": 0.0289, "step": 14120 }, { "grad_norm": 0.24653135240077972, "learning_rate": 5.888678183168368e-05, "loss": 0.0405, "step": 14130 }, { "grad_norm": 0.20660801231861115, "learning_rate": 5.883253831955061e-05, "loss": 0.0352, "step": 14140 }, { "grad_norm": 0.2055606245994568, "learning_rate": 5.8778284075050625e-05, "loss": 0.0306, "step": 14150 }, { "grad_norm": 0.2755970358848572, "learning_rate": 5.872401916410777e-05, "loss": 0.0533, "step": 14160 }, { "grad_norm": 0.2650911509990692, "learning_rate": 5.8669743652659014e-05, "loss": 0.0376, "step": 14170 }, { "grad_norm": 0.16536276042461395, "learning_rate": 5.86154576066542e-05, "loss": 0.0378, "step": 14180 }, { "grad_norm": 0.24964435398578644, "learning_rate": 5.856116109205602e-05, "loss": 0.0402, "step": 14190 }, { "grad_norm": 0.17182773351669312, "learning_rate": 5.850685417483983e-05, "loss": 0.0313, "step": 14200 }, { "grad_norm": 0.17793631553649902, "learning_rate": 5.8452536920993684e-05, "loss": 0.0339, "step": 14210 }, { "grad_norm": 0.23347553610801697, "learning_rate": 5.839820939651817e-05, "loss": 0.0414, "step": 14220 }, { "grad_norm": 0.24528080224990845, "learning_rate": 5.834387166742632e-05, "loss": 0.0343, "step": 14230 }, { "grad_norm": 0.25292474031448364, "learning_rate": 5.828952379974364e-05, "loss": 0.0325, "step": 14240 }, { "grad_norm": 0.23097531497478485, "learning_rate": 5.8235165859507864e-05, "loss": 0.0434, "step": 14250 }, { "grad_norm": 0.20588140189647675, "learning_rate": 5.818079791276907e-05, "loss": 0.0322, "step": 14260 }, { "grad_norm": 0.2397916465997696, "learning_rate": 5.8126420025589415e-05, "loss": 0.0427, "step": 14270 }, { "grad_norm": 0.22882963716983795, "learning_rate": 5.807203226404313e-05, "loss": 0.0334, "step": 14280 }, { "grad_norm": 0.22552430629730225, "learning_rate": 5.801763469421652e-05, "loss": 0.0271, "step": 14290 }, { "grad_norm": 0.17810235917568207, "learning_rate": 5.796322738220774e-05, "loss": 0.0354, "step": 14300 }, { "grad_norm": 0.1888292133808136, "learning_rate": 5.79088103941268e-05, "loss": 0.0284, "step": 14310 }, { "grad_norm": 0.20722892880439758, "learning_rate": 5.785438379609549e-05, "loss": 0.0449, "step": 14320 }, { "grad_norm": 0.21571381390094757, "learning_rate": 5.779994765424724e-05, "loss": 0.0296, "step": 14330 }, { "grad_norm": 0.226677805185318, "learning_rate": 5.7745502034727114e-05, "loss": 0.0341, "step": 14340 }, { "grad_norm": 0.2376023679971695, "learning_rate": 5.769104700369164e-05, "loss": 0.0341, "step": 14350 }, { "grad_norm": 0.26159730553627014, "learning_rate": 5.7636582627308854e-05, "loss": 0.0295, "step": 14360 }, { "grad_norm": 0.18536368012428284, "learning_rate": 5.7582108971758095e-05, "loss": 0.0323, "step": 14370 }, { "grad_norm": 0.17811369895935059, "learning_rate": 5.752762610322999e-05, "loss": 0.0327, "step": 14380 }, { "grad_norm": 0.23001214861869812, "learning_rate": 5.747313408792636e-05, "loss": 0.0452, "step": 14390 }, { "grad_norm": 0.17020268738269806, "learning_rate": 5.741863299206014e-05, "loss": 0.0351, "step": 14400 }, { "grad_norm": 0.19543471932411194, "learning_rate": 5.73641228818553e-05, "loss": 0.0272, "step": 14410 }, { "grad_norm": 0.20464570820331573, "learning_rate": 5.730960382354676e-05, "loss": 0.0274, "step": 14420 }, { "grad_norm": 0.22223763167858124, "learning_rate": 5.725507588338035e-05, "loss": 0.0348, "step": 14430 }, { "grad_norm": 0.22092466056346893, "learning_rate": 5.7200539127612604e-05, "loss": 0.0284, "step": 14440 }, { "grad_norm": 0.1885436475276947, "learning_rate": 5.714599362251087e-05, "loss": 0.0348, "step": 14450 }, { "grad_norm": 0.21525132656097412, "learning_rate": 5.709143943435307e-05, "loss": 0.0358, "step": 14460 }, { "grad_norm": 0.20676089823246002, "learning_rate": 5.7036876629427646e-05, "loss": 0.0364, "step": 14470 }, { "grad_norm": 0.2533383369445801, "learning_rate": 5.698230527403361e-05, "loss": 0.0313, "step": 14480 }, { "grad_norm": 0.215689018368721, "learning_rate": 5.6927725434480274e-05, "loss": 0.0324, "step": 14490 }, { "grad_norm": 0.16442927718162537, "learning_rate": 5.687313717708728e-05, "loss": 0.0443, "step": 14500 }, { "grad_norm": 0.2006472647190094, "learning_rate": 5.681854056818453e-05, "loss": 0.0391, "step": 14510 }, { "grad_norm": 0.19963093101978302, "learning_rate": 5.676393567411205e-05, "loss": 0.0257, "step": 14520 }, { "grad_norm": 0.210092231631279, "learning_rate": 5.670932256121991e-05, "loss": 0.049, "step": 14530 }, { "grad_norm": 0.21249045431613922, "learning_rate": 5.6654701295868216e-05, "loss": 0.028, "step": 14540 }, { "grad_norm": 0.23003003001213074, "learning_rate": 5.660007194442697e-05, "loss": 0.0278, "step": 14550 }, { "grad_norm": 0.22197377681732178, "learning_rate": 5.6545434573275946e-05, "loss": 0.0317, "step": 14560 }, { "grad_norm": 0.20769688487052917, "learning_rate": 5.649078924880472e-05, "loss": 0.0228, "step": 14570 }, { "grad_norm": 0.2729897201061249, "learning_rate": 5.643613603741252e-05, "loss": 0.04, "step": 14580 }, { "grad_norm": 0.17789435386657715, "learning_rate": 5.6381475005508154e-05, "loss": 0.0436, "step": 14590 }, { "grad_norm": 0.16866141557693481, "learning_rate": 5.632680621950992e-05, "loss": 0.0371, "step": 14600 }, { "grad_norm": 0.17298470437526703, "learning_rate": 5.627212974584555e-05, "loss": 0.0267, "step": 14610 }, { "grad_norm": 0.17135876417160034, "learning_rate": 5.62174456509521e-05, "loss": 0.026, "step": 14620 }, { "grad_norm": 0.18694578111171722, "learning_rate": 5.616275400127594e-05, "loss": 0.03, "step": 14630 }, { "grad_norm": 0.17214657366275787, "learning_rate": 5.610805486327254e-05, "loss": 0.0262, "step": 14640 }, { "grad_norm": 0.1767214834690094, "learning_rate": 5.6053348303406536e-05, "loss": 0.0286, "step": 14650 }, { "grad_norm": 0.1809508502483368, "learning_rate": 5.599863438815156e-05, "loss": 0.0328, "step": 14660 }, { "grad_norm": 0.22690822184085846, "learning_rate": 5.594391318399017e-05, "loss": 0.0408, "step": 14670 }, { "grad_norm": 0.2647361755371094, "learning_rate": 5.588918475741377e-05, "loss": 0.0291, "step": 14680 }, { "grad_norm": 0.19611769914627075, "learning_rate": 5.5834449174922586e-05, "loss": 0.0341, "step": 14690 }, { "grad_norm": 0.26690688729286194, "learning_rate": 5.57797065030255e-05, "loss": 0.0343, "step": 14700 }, { "grad_norm": 0.21813450753688812, "learning_rate": 5.5724956808240016e-05, "loss": 0.0397, "step": 14710 }, { "grad_norm": 0.2533606290817261, "learning_rate": 5.567020015709219e-05, "loss": 0.0344, "step": 14720 }, { "grad_norm": 0.18032726645469666, "learning_rate": 5.561543661611649e-05, "loss": 0.0322, "step": 14730 }, { "grad_norm": 0.17682529985904694, "learning_rate": 5.556066625185583e-05, "loss": 0.0368, "step": 14740 }, { "grad_norm": 0.21177470684051514, "learning_rate": 5.550588913086131e-05, "loss": 0.0348, "step": 14750 }, { "grad_norm": 0.184139683842659, "learning_rate": 5.545110531969234e-05, "loss": 0.0348, "step": 14760 }, { "grad_norm": 0.1982123851776123, "learning_rate": 5.539631488491641e-05, "loss": 0.0314, "step": 14770 }, { "grad_norm": 0.20279338955879211, "learning_rate": 5.534151789310904e-05, "loss": 0.0326, "step": 14780 }, { "grad_norm": 0.18822696805000305, "learning_rate": 5.528671441085376e-05, "loss": 0.027, "step": 14790 }, { "grad_norm": 0.2218574583530426, "learning_rate": 5.523190450474197e-05, "loss": 0.0349, "step": 14800 }, { "grad_norm": 0.1621718853712082, "learning_rate": 5.5177088241372865e-05, "loss": 0.0269, "step": 14810 }, { "grad_norm": 0.23336485028266907, "learning_rate": 5.512226568735338e-05, "loss": 0.0345, "step": 14820 }, { "grad_norm": 0.22718682885169983, "learning_rate": 5.506743690929809e-05, "loss": 0.0421, "step": 14830 }, { "grad_norm": 0.20915882289409637, "learning_rate": 5.501260197382912e-05, "loss": 0.0482, "step": 14840 }, { "grad_norm": 0.2537815272808075, "learning_rate": 5.49577609475761e-05, "loss": 0.0508, "step": 14850 }, { "grad_norm": 0.2017880231142044, "learning_rate": 5.490291389717603e-05, "loss": 0.0298, "step": 14860 }, { "grad_norm": 0.20284682512283325, "learning_rate": 5.484806088927329e-05, "loss": 0.0307, "step": 14870 }, { "grad_norm": 0.250785768032074, "learning_rate": 5.479320199051942e-05, "loss": 0.0356, "step": 14880 }, { "grad_norm": 0.22989673912525177, "learning_rate": 5.473833726757314e-05, "loss": 0.0291, "step": 14890 }, { "grad_norm": 0.2581615746021271, "learning_rate": 5.4683466787100304e-05, "loss": 0.0297, "step": 14900 }, { "grad_norm": 0.26279112696647644, "learning_rate": 5.462859061577369e-05, "loss": 0.0346, "step": 14910 }, { "grad_norm": 0.192124143242836, "learning_rate": 5.4573708820273026e-05, "loss": 0.027, "step": 14920 }, { "grad_norm": 0.29621562361717224, "learning_rate": 5.451882146728488e-05, "loss": 0.0283, "step": 14930 }, { "grad_norm": 0.15645359456539154, "learning_rate": 5.446392862350255e-05, "loss": 0.0406, "step": 14940 }, { "grad_norm": 0.20572848618030548, "learning_rate": 5.440903035562603e-05, "loss": 0.0451, "step": 14950 }, { "grad_norm": 0.14745163917541504, "learning_rate": 5.435412673036188e-05, "loss": 0.0357, "step": 14960 }, { "grad_norm": 0.14853869378566742, "learning_rate": 5.429921781442318e-05, "loss": 0.0326, "step": 14970 }, { "grad_norm": 0.2311686873435974, "learning_rate": 5.424430367452945e-05, "loss": 0.0342, "step": 14980 }, { "grad_norm": 0.19216400384902954, "learning_rate": 5.418938437740655e-05, "loss": 0.032, "step": 14990 }, { "grad_norm": 0.27252471446990967, "learning_rate": 5.413445998978658e-05, "loss": 0.0395, "step": 15000 }, { "grad_norm": 0.18471798300743103, "learning_rate": 5.407953057840789e-05, "loss": 0.026, "step": 15010 }, { "grad_norm": 0.21567478775978088, "learning_rate": 5.4024596210014853e-05, "loss": 0.0268, "step": 15020 }, { "grad_norm": 0.21496179699897766, "learning_rate": 5.3969656951357935e-05, "loss": 0.0292, "step": 15030 }, { "grad_norm": 0.13888072967529297, "learning_rate": 5.3914712869193504e-05, "loss": 0.0419, "step": 15040 }, { "grad_norm": 0.20161934196949005, "learning_rate": 5.385976403028381e-05, "loss": 0.0444, "step": 15050 }, { "grad_norm": 0.18831965327262878, "learning_rate": 5.3804810501396875e-05, "loss": 0.0449, "step": 15060 }, { "grad_norm": 0.1883031278848648, "learning_rate": 5.374985234930639e-05, "loss": 0.037, "step": 15070 }, { "grad_norm": 0.2121371179819107, "learning_rate": 5.3694889640791724e-05, "loss": 0.0341, "step": 15080 }, { "grad_norm": 0.22552281618118286, "learning_rate": 5.3639922442637736e-05, "loss": 0.0406, "step": 15090 }, { "grad_norm": 0.16408714652061462, "learning_rate": 5.358495082163476e-05, "loss": 0.0316, "step": 15100 }, { "grad_norm": 0.2552129030227661, "learning_rate": 5.35299748445785e-05, "loss": 0.0322, "step": 15110 }, { "grad_norm": 0.20163169503211975, "learning_rate": 5.3474994578269944e-05, "loss": 0.0261, "step": 15120 }, { "grad_norm": 0.2346493899822235, "learning_rate": 5.34200100895153e-05, "loss": 0.0291, "step": 15130 }, { "grad_norm": 0.19860681891441345, "learning_rate": 5.3365021445125916e-05, "loss": 0.0243, "step": 15140 }, { "grad_norm": 0.1700667440891266, "learning_rate": 5.331002871191817e-05, "loss": 0.0344, "step": 15150 }, { "grad_norm": 0.2913209795951843, "learning_rate": 5.3255031956713443e-05, "loss": 0.0336, "step": 15160 }, { "grad_norm": 0.21377846598625183, "learning_rate": 5.320003124633795e-05, "loss": 0.0349, "step": 15170 }, { "grad_norm": 0.22017671167850494, "learning_rate": 5.314502664762274e-05, "loss": 0.03, "step": 15180 }, { "grad_norm": 0.26838091015815735, "learning_rate": 5.3090018227403604e-05, "loss": 0.0296, "step": 15190 }, { "grad_norm": 0.1960185468196869, "learning_rate": 5.3035006052520955e-05, "loss": 0.0401, "step": 15200 }, { "grad_norm": 0.16199010610580444, "learning_rate": 5.297999018981976e-05, "loss": 0.031, "step": 15210 }, { "grad_norm": 0.22078879177570343, "learning_rate": 5.2924970706149505e-05, "loss": 0.0308, "step": 15220 }, { "grad_norm": 0.15208177268505096, "learning_rate": 5.2869947668364014e-05, "loss": 0.023, "step": 15230 }, { "grad_norm": 0.14824433624744415, "learning_rate": 5.2814921143321504e-05, "loss": 0.0254, "step": 15240 }, { "grad_norm": 0.2143213450908661, "learning_rate": 5.275989119788436e-05, "loss": 0.0469, "step": 15250 }, { "grad_norm": 0.23614417016506195, "learning_rate": 5.270485789891919e-05, "loss": 0.0368, "step": 15260 }, { "grad_norm": 0.1763586550951004, "learning_rate": 5.264982131329661e-05, "loss": 0.0286, "step": 15270 }, { "grad_norm": 0.132499560713768, "learning_rate": 5.259478150789128e-05, "loss": 0.0266, "step": 15280 }, { "grad_norm": 0.19481080770492554, "learning_rate": 5.253973854958173e-05, "loss": 0.0421, "step": 15290 }, { "grad_norm": 0.16290707886219025, "learning_rate": 5.2484692505250376e-05, "loss": 0.0312, "step": 15300 }, { "grad_norm": 0.1605348140001297, "learning_rate": 5.242964344178333e-05, "loss": 0.0363, "step": 15310 }, { "grad_norm": 0.17221303284168243, "learning_rate": 5.237459142607041e-05, "loss": 0.0251, "step": 15320 }, { "grad_norm": 0.17506961524486542, "learning_rate": 5.231953652500498e-05, "loss": 0.0286, "step": 15330 }, { "grad_norm": 0.18882247805595398, "learning_rate": 5.2264478805483976e-05, "loss": 0.0334, "step": 15340 }, { "grad_norm": 0.15320280194282532, "learning_rate": 5.220941833440768e-05, "loss": 0.0345, "step": 15350 }, { "grad_norm": 0.18199583888053894, "learning_rate": 5.215435517867977e-05, "loss": 0.0274, "step": 15360 }, { "grad_norm": 0.1420661211013794, "learning_rate": 5.209928940520719e-05, "loss": 0.0341, "step": 15370 }, { "grad_norm": 0.1718255579471588, "learning_rate": 5.204422108090004e-05, "loss": 0.029, "step": 15380 }, { "grad_norm": 0.21667012572288513, "learning_rate": 5.19891502726715e-05, "loss": 0.0275, "step": 15390 }, { "grad_norm": 0.25363489985466003, "learning_rate": 5.193407704743782e-05, "loss": 0.0284, "step": 15400 }, { "grad_norm": 0.21951861679553986, "learning_rate": 5.187900147211815e-05, "loss": 0.0247, "step": 15410 }, { "grad_norm": 0.17643359303474426, "learning_rate": 5.182392361363453e-05, "loss": 0.0347, "step": 15420 }, { "grad_norm": 0.17100131511688232, "learning_rate": 5.176884353891172e-05, "loss": 0.0506, "step": 15430 }, { "grad_norm": 0.22440698742866516, "learning_rate": 5.171376131487722e-05, "loss": 0.0293, "step": 15440 }, { "grad_norm": 0.20129908621311188, "learning_rate": 5.165867700846113e-05, "loss": 0.0421, "step": 15450 }, { "grad_norm": 0.23704560101032257, "learning_rate": 5.160359068659606e-05, "loss": 0.0289, "step": 15460 }, { "grad_norm": 0.1999187469482422, "learning_rate": 5.154850241621712e-05, "loss": 0.0262, "step": 15470 }, { "grad_norm": 0.2796396017074585, "learning_rate": 5.149341226426172e-05, "loss": 0.0369, "step": 15480 }, { "grad_norm": 0.2607261836528778, "learning_rate": 5.14383202976696e-05, "loss": 0.0338, "step": 15490 }, { "grad_norm": 0.18188908696174622, "learning_rate": 5.1383226583382684e-05, "loss": 0.0308, "step": 15500 }, { "grad_norm": 0.19661398231983185, "learning_rate": 5.132813118834504e-05, "loss": 0.0376, "step": 15510 }, { "grad_norm": 0.20603814721107483, "learning_rate": 5.1273034179502777e-05, "loss": 0.0245, "step": 15520 }, { "grad_norm": 0.23150624334812164, "learning_rate": 5.1217935623803945e-05, "loss": 0.0247, "step": 15530 }, { "grad_norm": 0.2575047016143799, "learning_rate": 5.1162835588198476e-05, "loss": 0.0302, "step": 15540 }, { "grad_norm": 0.20175369083881378, "learning_rate": 5.110773413963813e-05, "loss": 0.0259, "step": 15550 }, { "grad_norm": 0.17817428708076477, "learning_rate": 5.105263134507636e-05, "loss": 0.0352, "step": 15560 }, { "grad_norm": 0.22121955454349518, "learning_rate": 5.099752727146824e-05, "loss": 0.0382, "step": 15570 }, { "grad_norm": 0.17684024572372437, "learning_rate": 5.094242198577042e-05, "loss": 0.0411, "step": 15580 }, { "grad_norm": 0.13726060092449188, "learning_rate": 5.088731555494102e-05, "loss": 0.0255, "step": 15590 }, { "grad_norm": 0.16934603452682495, "learning_rate": 5.0832208045939556e-05, "loss": 0.0273, "step": 15600 }, { "grad_norm": 0.17474150657653809, "learning_rate": 5.0777099525726844e-05, "loss": 0.0349, "step": 15610 }, { "grad_norm": 0.19553348422050476, "learning_rate": 5.072199006126494e-05, "loss": 0.0241, "step": 15620 }, { "grad_norm": 0.23863022029399872, "learning_rate": 5.0666879719517026e-05, "loss": 0.0363, "step": 15630 }, { "grad_norm": 0.21098420023918152, "learning_rate": 5.061176856744737e-05, "loss": 0.0349, "step": 15640 }, { "grad_norm": 0.21710583567619324, "learning_rate": 5.0556656672021205e-05, "loss": 0.0313, "step": 15650 }, { "grad_norm": 0.2581000328063965, "learning_rate": 5.050154410020472e-05, "loss": 0.038, "step": 15660 }, { "grad_norm": 0.20587125420570374, "learning_rate": 5.044643091896485e-05, "loss": 0.0302, "step": 15670 }, { "grad_norm": 0.1747017502784729, "learning_rate": 5.039131719526932e-05, "loss": 0.0238, "step": 15680 }, { "grad_norm": 0.20988409221172333, "learning_rate": 5.03362029960865e-05, "loss": 0.0298, "step": 15690 }, { "grad_norm": 0.18822050094604492, "learning_rate": 5.028108838838533e-05, "loss": 0.0324, "step": 15700 }, { "grad_norm": 0.21265903115272522, "learning_rate": 5.0225973439135276e-05, "loss": 0.0322, "step": 15710 }, { "grad_norm": 0.20345637202262878, "learning_rate": 5.017085821530617e-05, "loss": 0.0235, "step": 15720 }, { "grad_norm": 0.2761247158050537, "learning_rate": 5.011574278386822e-05, "loss": 0.0418, "step": 15730 }, { "grad_norm": 0.17604151368141174, "learning_rate": 5.006062721179189e-05, "loss": 0.0309, "step": 15740 }, { "grad_norm": 0.24079827964305878, "learning_rate": 5.000551156604777e-05, "loss": 0.0368, "step": 15750 }, { "grad_norm": 0.2525259852409363, "learning_rate": 4.9950395913606594e-05, "loss": 0.0327, "step": 15760 }, { "grad_norm": 0.2506251037120819, "learning_rate": 4.989528032143903e-05, "loss": 0.0319, "step": 15770 }, { "grad_norm": 0.14502988755702972, "learning_rate": 4.984016485651578e-05, "loss": 0.0285, "step": 15780 }, { "grad_norm": 0.2011520266532898, "learning_rate": 4.9785049585807274e-05, "loss": 0.0392, "step": 15790 }, { "grad_norm": 0.1632399559020996, "learning_rate": 4.972993457628381e-05, "loss": 0.0344, "step": 15800 }, { "grad_norm": 0.16983166337013245, "learning_rate": 4.9674819894915306e-05, "loss": 0.0293, "step": 15810 }, { "grad_norm": 0.16955873370170593, "learning_rate": 4.9619705608671265e-05, "loss": 0.0295, "step": 15820 }, { "grad_norm": 0.1897546947002411, "learning_rate": 4.956459178452079e-05, "loss": 0.0287, "step": 15830 }, { "grad_norm": 0.18416433036327362, "learning_rate": 4.950947848943235e-05, "loss": 0.0297, "step": 15840 }, { "grad_norm": 0.2066381573677063, "learning_rate": 4.9454365790373805e-05, "loss": 0.0289, "step": 15850 }, { "grad_norm": 0.1745268851518631, "learning_rate": 4.939925375431226e-05, "loss": 0.0304, "step": 15860 }, { "grad_norm": 0.15955421328544617, "learning_rate": 4.9344142448214046e-05, "loss": 0.0371, "step": 15870 }, { "grad_norm": 0.193961963057518, "learning_rate": 4.928903193904461e-05, "loss": 0.039, "step": 15880 }, { "grad_norm": 0.21447107195854187, "learning_rate": 4.923392229376842e-05, "loss": 0.0351, "step": 15890 }, { "grad_norm": 0.1806727796792984, "learning_rate": 4.9178813579348865e-05, "loss": 0.0259, "step": 15900 }, { "grad_norm": 0.25397220253944397, "learning_rate": 4.9123705862748253e-05, "loss": 0.0288, "step": 15910 }, { "grad_norm": 0.2024795264005661, "learning_rate": 4.9068599210927627e-05, "loss": 0.0343, "step": 15920 }, { "grad_norm": 0.17103463411331177, "learning_rate": 4.901349369084681e-05, "loss": 0.0339, "step": 15930 }, { "grad_norm": 0.15353049337863922, "learning_rate": 4.8958389369464156e-05, "loss": 0.0284, "step": 15940 }, { "grad_norm": 0.21755750477313995, "learning_rate": 4.890328631373666e-05, "loss": 0.0245, "step": 15950 }, { "grad_norm": 0.27723759412765503, "learning_rate": 4.88481845906197e-05, "loss": 0.0294, "step": 15960 }, { "grad_norm": 0.2088996171951294, "learning_rate": 4.879308426706707e-05, "loss": 0.0338, "step": 15970 }, { "grad_norm": 0.16813503205776215, "learning_rate": 4.873798541003084e-05, "loss": 0.0246, "step": 15980 }, { "grad_norm": 0.1947682648897171, "learning_rate": 4.868288808646136e-05, "loss": 0.0428, "step": 15990 }, { "grad_norm": 0.19788439571857452, "learning_rate": 4.862779236330705e-05, "loss": 0.0299, "step": 16000 }, { "grad_norm": 0.17415949702262878, "learning_rate": 4.8572698307514395e-05, "loss": 0.034, "step": 16010 }, { "grad_norm": 0.16078214347362518, "learning_rate": 4.85176059860279e-05, "loss": 0.029, "step": 16020 }, { "grad_norm": 0.16113144159317017, "learning_rate": 4.846251546578989e-05, "loss": 0.0337, "step": 16030 }, { "grad_norm": 0.13889248669147491, "learning_rate": 4.840742681374058e-05, "loss": 0.0298, "step": 16040 }, { "grad_norm": 0.19051149487495422, "learning_rate": 4.835234009681787e-05, "loss": 0.0278, "step": 16050 }, { "grad_norm": 0.15506196022033691, "learning_rate": 4.8297255381957296e-05, "loss": 0.0258, "step": 16060 }, { "grad_norm": 0.21094945073127747, "learning_rate": 4.824217273609198e-05, "loss": 0.0275, "step": 16070 }, { "grad_norm": 0.19023582339286804, "learning_rate": 4.8187092226152555e-05, "loss": 0.0301, "step": 16080 }, { "grad_norm": 0.29125678539276123, "learning_rate": 4.8132013919067016e-05, "loss": 0.0337, "step": 16090 }, { "grad_norm": 0.1849486082792282, "learning_rate": 4.807693788176071e-05, "loss": 0.0332, "step": 16100 }, { "grad_norm": 0.18120302259922028, "learning_rate": 4.8021864181156214e-05, "loss": 0.0302, "step": 16110 }, { "grad_norm": 0.14806979894638062, "learning_rate": 4.796679288417326e-05, "loss": 0.039, "step": 16120 }, { "grad_norm": 0.2937803268432617, "learning_rate": 4.791172405772866e-05, "loss": 0.0357, "step": 16130 }, { "grad_norm": 0.26411184668540955, "learning_rate": 4.785665776873625e-05, "loss": 0.0234, "step": 16140 }, { "grad_norm": 0.17579706013202667, "learning_rate": 4.7801594084106763e-05, "loss": 0.037, "step": 16150 }, { "grad_norm": 0.15579213201999664, "learning_rate": 4.774653307074775e-05, "loss": 0.0327, "step": 16160 }, { "grad_norm": 0.14350923895835876, "learning_rate": 4.769147479556355e-05, "loss": 0.0273, "step": 16170 }, { "grad_norm": 0.1724187433719635, "learning_rate": 4.763641932545515e-05, "loss": 0.022, "step": 16180 }, { "grad_norm": 0.19554978609085083, "learning_rate": 4.758136672732013e-05, "loss": 0.0284, "step": 16190 }, { "grad_norm": 0.16540776193141937, "learning_rate": 4.752631706805261e-05, "loss": 0.0292, "step": 16200 }, { "grad_norm": 0.16038380563259125, "learning_rate": 4.7471270414543114e-05, "loss": 0.0366, "step": 16210 }, { "grad_norm": 0.15587887167930603, "learning_rate": 4.741622683367849e-05, "loss": 0.0235, "step": 16220 }, { "grad_norm": 0.1741756796836853, "learning_rate": 4.736118639234191e-05, "loss": 0.0419, "step": 16230 }, { "grad_norm": 0.2180599719285965, "learning_rate": 4.730614915741267e-05, "loss": 0.0352, "step": 16240 }, { "grad_norm": 0.1936764419078827, "learning_rate": 4.7251115195766234e-05, "loss": 0.0293, "step": 16250 }, { "grad_norm": 0.2160690277814865, "learning_rate": 4.719608457427404e-05, "loss": 0.033, "step": 16260 }, { "grad_norm": 0.131814643740654, "learning_rate": 4.714105735980347e-05, "loss": 0.0281, "step": 16270 }, { "grad_norm": 0.1713540405035019, "learning_rate": 4.70860336192178e-05, "loss": 0.0327, "step": 16280 }, { "grad_norm": 0.15607048571109772, "learning_rate": 4.7031013419376034e-05, "loss": 0.0279, "step": 16290 }, { "grad_norm": 0.21170714497566223, "learning_rate": 4.6975996827132926e-05, "loss": 0.0221, "step": 16300 }, { "grad_norm": 0.20372989773750305, "learning_rate": 4.692098390933883e-05, "loss": 0.0339, "step": 16310 }, { "grad_norm": 0.21972832083702087, "learning_rate": 4.6865974732839615e-05, "loss": 0.0265, "step": 16320 }, { "grad_norm": 0.19425582885742188, "learning_rate": 4.6810969364476615e-05, "loss": 0.0369, "step": 16330 }, { "grad_norm": 0.18226289749145508, "learning_rate": 4.675596787108653e-05, "loss": 0.0262, "step": 16340 }, { "grad_norm": 0.1735088974237442, "learning_rate": 4.670097031950139e-05, "loss": 0.0224, "step": 16350 }, { "grad_norm": 0.13549451529979706, "learning_rate": 4.664597677654838e-05, "loss": 0.0348, "step": 16360 }, { "grad_norm": 0.19674170017242432, "learning_rate": 4.659098730904986e-05, "loss": 0.037, "step": 16370 }, { "grad_norm": 0.178396537899971, "learning_rate": 4.6536001983823204e-05, "loss": 0.0296, "step": 16380 }, { "grad_norm": 0.1782890111207962, "learning_rate": 4.648102086768078e-05, "loss": 0.0394, "step": 16390 }, { "grad_norm": 0.17553257942199707, "learning_rate": 4.6426044027429786e-05, "loss": 0.0327, "step": 16400 }, { "grad_norm": 0.19642296433448792, "learning_rate": 4.637107152987234e-05, "loss": 0.0322, "step": 16410 }, { "grad_norm": 0.161524698138237, "learning_rate": 4.631610344180515e-05, "loss": 0.0307, "step": 16420 }, { "grad_norm": 0.1767970323562622, "learning_rate": 4.626113983001965e-05, "loss": 0.0304, "step": 16430 }, { "grad_norm": 0.16338589787483215, "learning_rate": 4.620618076130182e-05, "loss": 0.0269, "step": 16440 }, { "grad_norm": 0.1478855311870575, "learning_rate": 4.6151226302432074e-05, "loss": 0.0264, "step": 16450 }, { "grad_norm": 0.18238675594329834, "learning_rate": 4.6096276520185304e-05, "loss": 0.0307, "step": 16460 }, { "grad_norm": 0.19141612946987152, "learning_rate": 4.604133148133066e-05, "loss": 0.027, "step": 16470 }, { "grad_norm": 0.18557555973529816, "learning_rate": 4.598639125263155e-05, "loss": 0.0376, "step": 16480 }, { "grad_norm": 0.20224137604236603, "learning_rate": 4.593145590084553e-05, "loss": 0.0366, "step": 16490 }, { "grad_norm": 0.2365896850824356, "learning_rate": 4.58765254927242e-05, "loss": 0.0267, "step": 16500 }, { "grad_norm": 0.20219387114048004, "learning_rate": 4.582160009501323e-05, "loss": 0.0307, "step": 16510 }, { "grad_norm": 0.13314999639987946, "learning_rate": 4.5766679774452143e-05, "loss": 0.033, "step": 16520 }, { "grad_norm": 0.21838825941085815, "learning_rate": 4.571176459777431e-05, "loss": 0.0307, "step": 16530 }, { "grad_norm": 0.19648206233978271, "learning_rate": 4.5656854631706846e-05, "loss": 0.0291, "step": 16540 }, { "grad_norm": 0.2696146070957184, "learning_rate": 4.560194994297054e-05, "loss": 0.033, "step": 16550 }, { "grad_norm": 0.19981855154037476, "learning_rate": 4.554705059827974e-05, "loss": 0.0381, "step": 16560 }, { "grad_norm": 0.1518850177526474, "learning_rate": 4.5492156664342365e-05, "loss": 0.0303, "step": 16570 }, { "grad_norm": 0.20303091406822205, "learning_rate": 4.5437268207859695e-05, "loss": 0.0368, "step": 16580 }, { "grad_norm": 0.1486111581325531, "learning_rate": 4.538238529552641e-05, "loss": 0.0235, "step": 16590 }, { "grad_norm": 0.1364857405424118, "learning_rate": 4.5327507994030394e-05, "loss": 0.0213, "step": 16600 }, { "grad_norm": 0.19706322252750397, "learning_rate": 4.527263637005273e-05, "loss": 0.0232, "step": 16610 }, { "grad_norm": 0.1735982745885849, "learning_rate": 4.521777049026766e-05, "loss": 0.0207, "step": 16620 }, { "grad_norm": 0.161606103181839, "learning_rate": 4.5162910421342386e-05, "loss": 0.0226, "step": 16630 }, { "grad_norm": 0.1964697539806366, "learning_rate": 4.510805622993706e-05, "loss": 0.0353, "step": 16640 }, { "grad_norm": 0.17394199967384338, "learning_rate": 4.505320798270467e-05, "loss": 0.0319, "step": 16650 }, { "grad_norm": 0.1281191110610962, "learning_rate": 4.499836574629105e-05, "loss": 0.0262, "step": 16660 }, { "grad_norm": 0.15290354192256927, "learning_rate": 4.494352958733465e-05, "loss": 0.0413, "step": 16670 }, { "grad_norm": 0.21224620938301086, "learning_rate": 4.488869957246663e-05, "loss": 0.0281, "step": 16680 }, { "grad_norm": 0.2281198799610138, "learning_rate": 4.4833875768310575e-05, "loss": 0.0318, "step": 16690 }, { "grad_norm": 0.19968409836292267, "learning_rate": 4.47790582414826e-05, "loss": 0.0247, "step": 16700 }, { "grad_norm": 0.18725042045116425, "learning_rate": 4.472424705859114e-05, "loss": 0.0314, "step": 16710 }, { "grad_norm": 0.22686535120010376, "learning_rate": 4.466944228623701e-05, "loss": 0.0442, "step": 16720 }, { "grad_norm": 0.23461544513702393, "learning_rate": 4.461464399101312e-05, "loss": 0.0375, "step": 16730 }, { "grad_norm": 0.15982210636138916, "learning_rate": 4.45598522395046e-05, "loss": 0.0301, "step": 16740 }, { "grad_norm": 0.19638901948928833, "learning_rate": 4.4505067098288577e-05, "loss": 0.0384, "step": 16750 }, { "grad_norm": 0.20113924145698547, "learning_rate": 4.445028863393417e-05, "loss": 0.0235, "step": 16760 }, { "grad_norm": 0.18027903139591217, "learning_rate": 4.439551691300236e-05, "loss": 0.0257, "step": 16770 }, { "grad_norm": 0.17856533825397491, "learning_rate": 4.434075200204599e-05, "loss": 0.0275, "step": 16780 }, { "grad_norm": 0.1546831727027893, "learning_rate": 4.428599396760956e-05, "loss": 0.0248, "step": 16790 }, { "grad_norm": 0.20696532726287842, "learning_rate": 4.423124287622926e-05, "loss": 0.0269, "step": 16800 }, { "grad_norm": 0.1875312477350235, "learning_rate": 4.417649879443282e-05, "loss": 0.0236, "step": 16810 }, { "grad_norm": 0.18301160633563995, "learning_rate": 4.4121761788739444e-05, "loss": 0.0287, "step": 16820 }, { "grad_norm": 0.14404737949371338, "learning_rate": 4.4067031925659805e-05, "loss": 0.0258, "step": 16830 }, { "grad_norm": 0.2055557370185852, "learning_rate": 4.4012309271695816e-05, "loss": 0.024, "step": 16840 }, { "grad_norm": 0.2510603666305542, "learning_rate": 4.395759389334067e-05, "loss": 0.0282, "step": 16850 }, { "grad_norm": 0.18159820139408112, "learning_rate": 4.3902885857078685e-05, "loss": 0.0329, "step": 16860 }, { "grad_norm": 0.23767076432704926, "learning_rate": 4.384818522938531e-05, "loss": 0.0226, "step": 16870 }, { "grad_norm": 0.13381364941596985, "learning_rate": 4.379349207672695e-05, "loss": 0.022, "step": 16880 }, { "grad_norm": 0.24738550186157227, "learning_rate": 4.373880646556098e-05, "loss": 0.0306, "step": 16890 }, { "grad_norm": 0.2408086657524109, "learning_rate": 4.3684128462335533e-05, "loss": 0.0242, "step": 16900 }, { "grad_norm": 0.17812231183052063, "learning_rate": 4.362945813348955e-05, "loss": 0.0224, "step": 16910 }, { "grad_norm": 0.16423828899860382, "learning_rate": 4.357479554545262e-05, "loss": 0.0255, "step": 16920 }, { "grad_norm": 0.2010805308818817, "learning_rate": 4.352014076464499e-05, "loss": 0.0312, "step": 16930 }, { "grad_norm": 0.17796847224235535, "learning_rate": 4.346549385747734e-05, "loss": 0.0277, "step": 16940 }, { "grad_norm": 0.15069518983364105, "learning_rate": 4.34108548903508e-05, "loss": 0.0398, "step": 16950 }, { "grad_norm": 0.23427198827266693, "learning_rate": 4.335622392965689e-05, "loss": 0.0262, "step": 16960 }, { "grad_norm": 0.19162601232528687, "learning_rate": 4.330160104177738e-05, "loss": 0.0358, "step": 16970 }, { "grad_norm": 0.1743127703666687, "learning_rate": 4.32469862930842e-05, "loss": 0.0316, "step": 16980 }, { "grad_norm": 0.18205121159553528, "learning_rate": 4.319237974993946e-05, "loss": 0.0355, "step": 16990 }, { "grad_norm": 0.2791115939617157, "learning_rate": 4.3137781478695236e-05, "loss": 0.0358, "step": 17000 }, { "grad_norm": 0.1633518487215042, "learning_rate": 4.308319154569358e-05, "loss": 0.0388, "step": 17010 }, { "grad_norm": 0.22917166352272034, "learning_rate": 4.302861001726641e-05, "loss": 0.0405, "step": 17020 }, { "grad_norm": 0.16011689603328705, "learning_rate": 4.2974036959735415e-05, "loss": 0.0244, "step": 17030 }, { "grad_norm": 0.17509062588214874, "learning_rate": 4.2919472439412034e-05, "loss": 0.0236, "step": 17040 }, { "grad_norm": 0.1841406375169754, "learning_rate": 4.286491652259729e-05, "loss": 0.023, "step": 17050 }, { "grad_norm": 0.2576296329498291, "learning_rate": 4.281036927558178e-05, "loss": 0.0393, "step": 17060 }, { "grad_norm": 0.17871570587158203, "learning_rate": 4.275583076464552e-05, "loss": 0.0224, "step": 17070 }, { "grad_norm": 0.2514389157295227, "learning_rate": 4.270130105605794e-05, "loss": 0.0249, "step": 17080 }, { "grad_norm": 0.20198608934879303, "learning_rate": 4.264678021607782e-05, "loss": 0.0281, "step": 17090 }, { "grad_norm": 0.13999813795089722, "learning_rate": 4.25922683109531e-05, "loss": 0.0239, "step": 17100 }, { "grad_norm": 0.20158660411834717, "learning_rate": 4.25377654069209e-05, "loss": 0.0291, "step": 17110 }, { "grad_norm": 0.21980403363704681, "learning_rate": 4.248327157020737e-05, "loss": 0.0244, "step": 17120 }, { "grad_norm": 0.20235925912857056, "learning_rate": 4.242878686702763e-05, "loss": 0.0249, "step": 17130 }, { "grad_norm": 0.18056462705135345, "learning_rate": 4.23743113635858e-05, "loss": 0.0197, "step": 17140 }, { "grad_norm": 0.19994911551475525, "learning_rate": 4.2319845126074705e-05, "loss": 0.0331, "step": 17150 }, { "grad_norm": 0.20517182350158691, "learning_rate": 4.226538822067598e-05, "loss": 0.0314, "step": 17160 }, { "grad_norm": 0.1546240597963333, "learning_rate": 4.2210940713559896e-05, "loss": 0.0321, "step": 17170 }, { "grad_norm": 0.14777404069900513, "learning_rate": 4.21565026708853e-05, "loss": 0.0247, "step": 17180 }, { "grad_norm": 0.22177310287952423, "learning_rate": 4.2102074158799535e-05, "loss": 0.0224, "step": 17190 }, { "grad_norm": 0.17324088513851166, "learning_rate": 4.204765524343841e-05, "loss": 0.0308, "step": 17200 }, { "grad_norm": 0.19211719930171967, "learning_rate": 4.1993245990926036e-05, "loss": 0.0263, "step": 17210 }, { "grad_norm": 0.177433580160141, "learning_rate": 4.193884646737474e-05, "loss": 0.0278, "step": 17220 }, { "grad_norm": 0.13847830891609192, "learning_rate": 4.1884456738885124e-05, "loss": 0.0244, "step": 17230 }, { "grad_norm": 0.21790458261966705, "learning_rate": 4.1830076871545795e-05, "loss": 0.0248, "step": 17240 }, { "grad_norm": 0.15523585677146912, "learning_rate": 4.177570693143346e-05, "loss": 0.0398, "step": 17250 }, { "grad_norm": 0.15436400473117828, "learning_rate": 4.172134698461271e-05, "loss": 0.0263, "step": 17260 }, { "grad_norm": 0.15093588829040527, "learning_rate": 4.166699709713599e-05, "loss": 0.0231, "step": 17270 }, { "grad_norm": 0.1275588870048523, "learning_rate": 4.1612657335043555e-05, "loss": 0.0233, "step": 17280 }, { "grad_norm": 0.19398851692676544, "learning_rate": 4.155832776436331e-05, "loss": 0.0342, "step": 17290 }, { "grad_norm": 0.24114355444908142, "learning_rate": 4.1504008451110845e-05, "loss": 0.0261, "step": 17300 }, { "grad_norm": 0.17283889651298523, "learning_rate": 4.1449699461289225e-05, "loss": 0.0284, "step": 17310 }, { "grad_norm": 0.1788194328546524, "learning_rate": 4.139540086088901e-05, "loss": 0.0202, "step": 17320 }, { "grad_norm": 0.15865406394004822, "learning_rate": 4.1341112715888106e-05, "loss": 0.0226, "step": 17330 }, { "grad_norm": 0.1783648133277893, "learning_rate": 4.128683509225172e-05, "loss": 0.0299, "step": 17340 }, { "grad_norm": 0.16359177231788635, "learning_rate": 4.123256805593231e-05, "loss": 0.0215, "step": 17350 }, { "grad_norm": 0.20935894548892975, "learning_rate": 4.117831167286943e-05, "loss": 0.0351, "step": 17360 }, { "grad_norm": 0.23411209881305695, "learning_rate": 4.1124066008989685e-05, "loss": 0.0302, "step": 17370 }, { "grad_norm": 0.16020114719867706, "learning_rate": 4.106983113020669e-05, "loss": 0.0251, "step": 17380 }, { "grad_norm": 0.13372816145420074, "learning_rate": 4.101560710242094e-05, "loss": 0.0252, "step": 17390 }, { "grad_norm": 0.2202974408864975, "learning_rate": 4.096139399151971e-05, "loss": 0.0321, "step": 17400 }, { "grad_norm": 0.16391226649284363, "learning_rate": 4.090719186337709e-05, "loss": 0.0204, "step": 17410 }, { "grad_norm": 0.1368219405412674, "learning_rate": 4.0853000783853746e-05, "loss": 0.0268, "step": 17420 }, { "grad_norm": 0.17324770987033844, "learning_rate": 4.0798820818796956e-05, "loss": 0.0232, "step": 17430 }, { "grad_norm": 0.25627389550209045, "learning_rate": 4.074465203404048e-05, "loss": 0.0311, "step": 17440 }, { "grad_norm": 0.13377155363559723, "learning_rate": 4.0690494495404495e-05, "loss": 0.0286, "step": 17450 }, { "grad_norm": 0.20062614977359772, "learning_rate": 4.0636348268695535e-05, "loss": 0.0256, "step": 17460 }, { "grad_norm": 0.16404034197330475, "learning_rate": 4.058221341970637e-05, "loss": 0.0241, "step": 17470 }, { "grad_norm": 0.16925594210624695, "learning_rate": 4.0528090014215945e-05, "loss": 0.0311, "step": 17480 }, { "grad_norm": 0.24375289678573608, "learning_rate": 4.047397811798929e-05, "loss": 0.0251, "step": 17490 }, { "grad_norm": 0.14308850467205048, "learning_rate": 4.041987779677745e-05, "loss": 0.02, "step": 17500 }, { "grad_norm": 0.2551865577697754, "learning_rate": 4.036578911631745e-05, "loss": 0.0281, "step": 17510 }, { "grad_norm": 0.16355840861797333, "learning_rate": 4.031171214233211e-05, "loss": 0.0274, "step": 17520 }, { "grad_norm": 0.2348547726869583, "learning_rate": 4.025764694053008e-05, "loss": 0.026, "step": 17530 }, { "grad_norm": 0.1249302327632904, "learning_rate": 4.020359357660566e-05, "loss": 0.0317, "step": 17540 }, { "grad_norm": 0.1952463984489441, "learning_rate": 4.014955211623875e-05, "loss": 0.0226, "step": 17550 }, { "grad_norm": 0.15928368270397186, "learning_rate": 4.0095522625094895e-05, "loss": 0.0416, "step": 17560 }, { "grad_norm": 0.15431587398052216, "learning_rate": 4.004150516882497e-05, "loss": 0.0341, "step": 17570 }, { "grad_norm": 0.1980421096086502, "learning_rate": 3.99874998130653e-05, "loss": 0.0295, "step": 17580 }, { "grad_norm": 0.14738374948501587, "learning_rate": 3.9933506623437455e-05, "loss": 0.0312, "step": 17590 }, { "grad_norm": 0.15701749920845032, "learning_rate": 3.9879525665548276e-05, "loss": 0.0458, "step": 17600 }, { "grad_norm": 0.1988798826932907, "learning_rate": 3.9825557004989704e-05, "loss": 0.0361, "step": 17610 }, { "grad_norm": 0.17472513020038605, "learning_rate": 3.977160070733877e-05, "loss": 0.0281, "step": 17620 }, { "grad_norm": 0.20336899161338806, "learning_rate": 3.971765683815746e-05, "loss": 0.0212, "step": 17630 }, { "grad_norm": 0.17503057420253754, "learning_rate": 3.966372546299265e-05, "loss": 0.0303, "step": 17640 }, { "grad_norm": 0.15202848613262177, "learning_rate": 3.960980664737604e-05, "loss": 0.023, "step": 17650 }, { "grad_norm": 0.23117199540138245, "learning_rate": 3.955590045682408e-05, "loss": 0.039, "step": 17660 }, { "grad_norm": 0.1704898476600647, "learning_rate": 3.950200695683788e-05, "loss": 0.0298, "step": 17670 }, { "grad_norm": 0.27092838287353516, "learning_rate": 3.944812621290314e-05, "loss": 0.0328, "step": 17680 }, { "grad_norm": 0.18377923965454102, "learning_rate": 3.939425829049002e-05, "loss": 0.0263, "step": 17690 }, { "grad_norm": 0.11225568503141403, "learning_rate": 3.9340403255053135e-05, "loss": 0.0314, "step": 17700 }, { "grad_norm": 0.17668499052524567, "learning_rate": 3.9286561172031407e-05, "loss": 0.0307, "step": 17710 }, { "grad_norm": 0.1555749922990799, "learning_rate": 3.923273210684809e-05, "loss": 0.0283, "step": 17720 }, { "grad_norm": 0.1380285769701004, "learning_rate": 3.9178916124910555e-05, "loss": 0.0229, "step": 17730 }, { "grad_norm": 0.1583070307970047, "learning_rate": 3.9125113291610274e-05, "loss": 0.0346, "step": 17740 }, { "grad_norm": 0.21410991251468658, "learning_rate": 3.907132367232278e-05, "loss": 0.043, "step": 17750 }, { "grad_norm": 0.17118176817893982, "learning_rate": 3.901754733240752e-05, "loss": 0.0339, "step": 17760 }, { "grad_norm": 0.1818895936012268, "learning_rate": 3.896378433720785e-05, "loss": 0.0223, "step": 17770 }, { "grad_norm": 0.17435546219348907, "learning_rate": 3.891003475205086e-05, "loss": 0.0383, "step": 17780 }, { "grad_norm": 0.17768867313861847, "learning_rate": 3.885629864224736e-05, "loss": 0.0174, "step": 17790 }, { "grad_norm": 0.17956306040287018, "learning_rate": 3.880257607309178e-05, "loss": 0.0202, "step": 17800 }, { "grad_norm": 0.2094353884458542, "learning_rate": 3.874886710986213e-05, "loss": 0.0327, "step": 17810 }, { "grad_norm": 0.1935320645570755, "learning_rate": 3.8695171817819834e-05, "loss": 0.0277, "step": 17820 }, { "grad_norm": 0.1698695868253708, "learning_rate": 3.8641490262209774e-05, "loss": 0.0268, "step": 17830 }, { "grad_norm": 0.18696099519729614, "learning_rate": 3.8587822508260084e-05, "loss": 0.025, "step": 17840 }, { "grad_norm": 0.19433924555778503, "learning_rate": 3.853416862118214e-05, "loss": 0.0242, "step": 17850 }, { "grad_norm": 0.167063906788826, "learning_rate": 3.848052866617049e-05, "loss": 0.0273, "step": 17860 }, { "grad_norm": 0.1941961646080017, "learning_rate": 3.84269027084027e-05, "loss": 0.0271, "step": 17870 }, { "grad_norm": 0.17196966707706451, "learning_rate": 3.83732908130394e-05, "loss": 0.0278, "step": 17880 }, { "grad_norm": 0.1909564584493637, "learning_rate": 3.831969304522409e-05, "loss": 0.028, "step": 17890 }, { "grad_norm": 0.1785920113325119, "learning_rate": 3.826610947008312e-05, "loss": 0.0215, "step": 17900 }, { "grad_norm": 0.21671876311302185, "learning_rate": 3.821254015272559e-05, "loss": 0.0306, "step": 17910 }, { "grad_norm": 0.1587906926870346, "learning_rate": 3.8158985158243214e-05, "loss": 0.0232, "step": 17920 }, { "grad_norm": 0.1281021237373352, "learning_rate": 3.810544455171044e-05, "loss": 0.0275, "step": 17930 }, { "grad_norm": 0.22444884479045868, "learning_rate": 3.8051918398184114e-05, "loss": 0.0249, "step": 17940 }, { "grad_norm": 0.16777116060256958, "learning_rate": 3.799840676270356e-05, "loss": 0.0344, "step": 17950 }, { "grad_norm": 0.2180856317281723, "learning_rate": 3.794490971029048e-05, "loss": 0.032, "step": 17960 }, { "grad_norm": 0.14472422003746033, "learning_rate": 3.789142730594881e-05, "loss": 0.0203, "step": 17970 }, { "grad_norm": 0.22646042704582214, "learning_rate": 3.783795961466471e-05, "loss": 0.0364, "step": 17980 }, { "grad_norm": 0.18454337120056152, "learning_rate": 3.778450670140651e-05, "loss": 0.021, "step": 17990 }, { "grad_norm": 0.16541297733783722, "learning_rate": 3.773106863112451e-05, "loss": 0.0313, "step": 18000 }, { "grad_norm": 0.19565095007419586, "learning_rate": 3.7677645468751e-05, "loss": 0.0418, "step": 18010 }, { "grad_norm": 0.17675285041332245, "learning_rate": 3.762423727920018e-05, "loss": 0.0257, "step": 18020 }, { "grad_norm": 0.17111623287200928, "learning_rate": 3.7570844127367995e-05, "loss": 0.038, "step": 18030 }, { "grad_norm": 0.15214599668979645, "learning_rate": 3.751746607813221e-05, "loss": 0.03, "step": 18040 }, { "grad_norm": 0.19278427958488464, "learning_rate": 3.746410319635217e-05, "loss": 0.0294, "step": 18050 }, { "grad_norm": 0.24276070296764374, "learning_rate": 3.7410755546868806e-05, "loss": 0.0269, "step": 18060 }, { "grad_norm": 0.18017792701721191, "learning_rate": 3.735742319450454e-05, "loss": 0.0259, "step": 18070 }, { "grad_norm": 0.15712124109268188, "learning_rate": 3.7304106204063184e-05, "loss": 0.0232, "step": 18080 }, { "grad_norm": 0.17800459265708923, "learning_rate": 3.725080464032996e-05, "loss": 0.0287, "step": 18090 }, { "grad_norm": 0.22226795554161072, "learning_rate": 3.719751856807125e-05, "loss": 0.0341, "step": 18100 }, { "grad_norm": 0.18403956294059753, "learning_rate": 3.71442480520347e-05, "loss": 0.0229, "step": 18110 }, { "grad_norm": 0.1695186346769333, "learning_rate": 3.709099315694897e-05, "loss": 0.026, "step": 18120 }, { "grad_norm": 0.23708264529705048, "learning_rate": 3.703775394752378e-05, "loss": 0.0282, "step": 18130 }, { "grad_norm": 0.1198526918888092, "learning_rate": 3.698453048844983e-05, "loss": 0.0211, "step": 18140 }, { "grad_norm": 0.18099425733089447, "learning_rate": 3.693132284439861e-05, "loss": 0.0249, "step": 18150 }, { "grad_norm": 0.1238897293806076, "learning_rate": 3.6878131080022414e-05, "loss": 0.027, "step": 18160 }, { "grad_norm": 0.17655684053897858, "learning_rate": 3.682495525995429e-05, "loss": 0.0259, "step": 18170 }, { "grad_norm": 0.1988743543624878, "learning_rate": 3.6771795448807846e-05, "loss": 0.0234, "step": 18180 }, { "grad_norm": 0.14156265556812286, "learning_rate": 3.671865171117724e-05, "loss": 0.028, "step": 18190 }, { "grad_norm": 0.12035544216632843, "learning_rate": 3.666552411163718e-05, "loss": 0.018, "step": 18200 }, { "grad_norm": 0.18832409381866455, "learning_rate": 3.661241271474269e-05, "loss": 0.0261, "step": 18210 }, { "grad_norm": 0.1847100853919983, "learning_rate": 3.6559317585029116e-05, "loss": 0.0179, "step": 18220 }, { "grad_norm": 0.16597479581832886, "learning_rate": 3.650623878701204e-05, "loss": 0.0261, "step": 18230 }, { "grad_norm": 0.16980786621570587, "learning_rate": 3.645317638518721e-05, "loss": 0.0226, "step": 18240 }, { "grad_norm": 0.21781304478645325, "learning_rate": 3.640013044403046e-05, "loss": 0.0309, "step": 18250 }, { "grad_norm": 0.1537376344203949, "learning_rate": 3.634710102799761e-05, "loss": 0.0239, "step": 18260 }, { "grad_norm": 0.18778207898139954, "learning_rate": 3.6294088201524394e-05, "loss": 0.0323, "step": 18270 }, { "grad_norm": 0.15444649755954742, "learning_rate": 3.62410920290264e-05, "loss": 0.018, "step": 18280 }, { "grad_norm": 0.17421935498714447, "learning_rate": 3.6188112574898954e-05, "loss": 0.0274, "step": 18290 }, { "grad_norm": 0.19357150793075562, "learning_rate": 3.6135149903517115e-05, "loss": 0.0223, "step": 18300 }, { "grad_norm": 0.09341525286436081, "learning_rate": 3.608220407923552e-05, "loss": 0.0262, "step": 18310 }, { "grad_norm": 0.22790499031543732, "learning_rate": 3.602927516638833e-05, "loss": 0.0299, "step": 18320 }, { "grad_norm": 0.1444433331489563, "learning_rate": 3.5976363229289165e-05, "loss": 0.0239, "step": 18330 }, { "grad_norm": 0.16749098896980286, "learning_rate": 3.5923468332231e-05, "loss": 0.0271, "step": 18340 }, { "grad_norm": 0.19176580011844635, "learning_rate": 3.587059053948616e-05, "loss": 0.0269, "step": 18350 }, { "grad_norm": 0.21202567219734192, "learning_rate": 3.5817729915306134e-05, "loss": 0.0235, "step": 18360 }, { "grad_norm": 0.18161249160766602, "learning_rate": 3.5764886523921566e-05, "loss": 0.0223, "step": 18370 }, { "grad_norm": 0.2187669575214386, "learning_rate": 3.571206042954214e-05, "loss": 0.0238, "step": 18380 }, { "grad_norm": 0.1260470151901245, "learning_rate": 3.5659251696356566e-05, "loss": 0.02, "step": 18390 }, { "grad_norm": 0.18398205935955048, "learning_rate": 3.56064603885324e-05, "loss": 0.0373, "step": 18400 }, { "grad_norm": 0.1526593416929245, "learning_rate": 3.5553686570216115e-05, "loss": 0.0299, "step": 18410 }, { "grad_norm": 0.10986694693565369, "learning_rate": 3.550093030553284e-05, "loss": 0.0299, "step": 18420 }, { "grad_norm": 0.181725412607193, "learning_rate": 3.544819165858642e-05, "loss": 0.0227, "step": 18430 }, { "grad_norm": 0.2112978994846344, "learning_rate": 3.539547069345926e-05, "loss": 0.0202, "step": 18440 }, { "grad_norm": 0.17450885474681854, "learning_rate": 3.5342767474212344e-05, "loss": 0.0301, "step": 18450 }, { "grad_norm": 0.2050676792860031, "learning_rate": 3.529008206488502e-05, "loss": 0.0264, "step": 18460 }, { "grad_norm": 0.20739780366420746, "learning_rate": 3.5237414529495055e-05, "loss": 0.0336, "step": 18470 }, { "grad_norm": 0.1394074410200119, "learning_rate": 3.5184764932038454e-05, "loss": 0.0249, "step": 18480 }, { "grad_norm": 0.16489040851593018, "learning_rate": 3.513213333648945e-05, "loss": 0.0196, "step": 18490 }, { "grad_norm": 0.1349034309387207, "learning_rate": 3.5079519806800374e-05, "loss": 0.0357, "step": 18500 }, { "grad_norm": 0.2241867631673813, "learning_rate": 3.502692440690165e-05, "loss": 0.0264, "step": 18510 }, { "grad_norm": 0.1894799768924713, "learning_rate": 3.497434720070165e-05, "loss": 0.0231, "step": 18520 }, { "grad_norm": 0.13524530827999115, "learning_rate": 3.4921788252086616e-05, "loss": 0.0306, "step": 18530 }, { "grad_norm": 0.13438217341899872, "learning_rate": 3.486924762492065e-05, "loss": 0.0206, "step": 18540 }, { "grad_norm": 0.16378630697727203, "learning_rate": 3.4816725383045534e-05, "loss": 0.0243, "step": 18550 }, { "grad_norm": 0.20007386803627014, "learning_rate": 3.476422159028079e-05, "loss": 0.0405, "step": 18560 }, { "grad_norm": 0.16197073459625244, "learning_rate": 3.471173631042345e-05, "loss": 0.0265, "step": 18570 }, { "grad_norm": 0.19861827790737152, "learning_rate": 3.465926960724808e-05, "loss": 0.0225, "step": 18580 }, { "grad_norm": 0.22620494663715363, "learning_rate": 3.460682154450666e-05, "loss": 0.0268, "step": 18590 }, { "grad_norm": 0.2333693951368332, "learning_rate": 3.4554392185928564e-05, "loss": 0.0408, "step": 18600 }, { "grad_norm": 0.1531403511762619, "learning_rate": 3.450198159522037e-05, "loss": 0.0223, "step": 18610 }, { "grad_norm": 0.12277983129024506, "learning_rate": 3.444958983606592e-05, "loss": 0.0382, "step": 18620 }, { "grad_norm": 0.28725090622901917, "learning_rate": 3.439721697212612e-05, "loss": 0.0552, "step": 18630 }, { "grad_norm": 0.17010430991649628, "learning_rate": 3.4344863067038954e-05, "loss": 0.0263, "step": 18640 }, { "grad_norm": 0.17477719485759735, "learning_rate": 3.429252818441935e-05, "loss": 0.0273, "step": 18650 }, { "grad_norm": 0.20451471209526062, "learning_rate": 3.4240212387859094e-05, "loss": 0.0294, "step": 18660 }, { "grad_norm": 0.1548232138156891, "learning_rate": 3.418791574092686e-05, "loss": 0.0296, "step": 18670 }, { "grad_norm": 0.14248524606227875, "learning_rate": 3.413563830716796e-05, "loss": 0.0249, "step": 18680 }, { "grad_norm": 0.17519253492355347, "learning_rate": 3.408338015010445e-05, "loss": 0.0218, "step": 18690 }, { "grad_norm": 0.15680065751075745, "learning_rate": 3.40311413332349e-05, "loss": 0.0202, "step": 18700 }, { "grad_norm": 0.12265601754188538, "learning_rate": 3.3978921920034365e-05, "loss": 0.0247, "step": 18710 }, { "grad_norm": 0.1399289220571518, "learning_rate": 3.392672197395441e-05, "loss": 0.0195, "step": 18720 }, { "grad_norm": 0.19597192108631134, "learning_rate": 3.387454155842287e-05, "loss": 0.0546, "step": 18730 }, { "grad_norm": 0.1330702006816864, "learning_rate": 3.382238073684386e-05, "loss": 0.018, "step": 18740 }, { "grad_norm": 0.19046880304813385, "learning_rate": 3.377023957259771e-05, "loss": 0.0181, "step": 18750 }, { "grad_norm": 0.19902783632278442, "learning_rate": 3.3718118129040835e-05, "loss": 0.0227, "step": 18760 }, { "grad_norm": 0.16072876751422882, "learning_rate": 3.3666016469505724e-05, "loss": 0.0403, "step": 18770 }, { "grad_norm": 0.24210183322429657, "learning_rate": 3.361393465730079e-05, "loss": 0.0372, "step": 18780 }, { "grad_norm": 0.16308537125587463, "learning_rate": 3.3561872755710366e-05, "loss": 0.0268, "step": 18790 }, { "grad_norm": 0.19627268612384796, "learning_rate": 3.350983082799456e-05, "loss": 0.0213, "step": 18800 }, { "grad_norm": 0.13842621445655823, "learning_rate": 3.34578089373892e-05, "loss": 0.0257, "step": 18810 }, { "grad_norm": 0.155190110206604, "learning_rate": 3.340580714710581e-05, "loss": 0.0268, "step": 18820 }, { "grad_norm": 0.11541121453046799, "learning_rate": 3.3353825520331466e-05, "loss": 0.0229, "step": 18830 }, { "grad_norm": 0.137100487947464, "learning_rate": 3.330186412022876e-05, "loss": 0.0242, "step": 18840 }, { "grad_norm": 0.14785906672477722, "learning_rate": 3.324992300993568e-05, "loss": 0.0249, "step": 18850 }, { "grad_norm": 0.2191990613937378, "learning_rate": 3.319800225256556e-05, "loss": 0.0305, "step": 18860 }, { "grad_norm": 0.1687200665473938, "learning_rate": 3.314610191120702e-05, "loss": 0.0298, "step": 18870 }, { "grad_norm": 0.17324794828891754, "learning_rate": 3.30942220489239e-05, "loss": 0.0262, "step": 18880 }, { "grad_norm": 0.18554995954036713, "learning_rate": 3.3042362728755086e-05, "loss": 0.0232, "step": 18890 }, { "grad_norm": 0.12349756807088852, "learning_rate": 3.299052401371456e-05, "loss": 0.0381, "step": 18900 }, { "grad_norm": 0.19428104162216187, "learning_rate": 3.293870596679125e-05, "loss": 0.0196, "step": 18910 }, { "grad_norm": 0.15741226077079773, "learning_rate": 3.288690865094895e-05, "loss": 0.0234, "step": 18920 }, { "grad_norm": 0.17115776240825653, "learning_rate": 3.283513212912632e-05, "loss": 0.0291, "step": 18930 }, { "grad_norm": 0.18184944987297058, "learning_rate": 3.2783376464236684e-05, "loss": 0.0178, "step": 18940 }, { "grad_norm": 0.13031050562858582, "learning_rate": 3.273164171916806e-05, "loss": 0.0178, "step": 18950 }, { "grad_norm": 0.1831790655851364, "learning_rate": 3.267992795678306e-05, "loss": 0.0264, "step": 18960 }, { "grad_norm": 0.16591621935367584, "learning_rate": 3.2628235239918744e-05, "loss": 0.0244, "step": 18970 }, { "grad_norm": 0.16486702859401703, "learning_rate": 3.2576563631386695e-05, "loss": 0.0276, "step": 18980 }, { "grad_norm": 0.15615999698638916, "learning_rate": 3.252491319397275e-05, "loss": 0.0219, "step": 18990 }, { "grad_norm": 0.19280681014060974, "learning_rate": 3.247328399043706e-05, "loss": 0.033, "step": 19000 }, { "grad_norm": 0.18981625139713287, "learning_rate": 3.242167608351399e-05, "loss": 0.0304, "step": 19010 }, { "grad_norm": 0.23294803500175476, "learning_rate": 3.2370089535911986e-05, "loss": 0.0229, "step": 19020 }, { "grad_norm": 0.19178524613380432, "learning_rate": 3.23185244103136e-05, "loss": 0.019, "step": 19030 }, { "grad_norm": 0.12617479264736176, "learning_rate": 3.22669807693753e-05, "loss": 0.0338, "step": 19040 }, { "grad_norm": 0.13756798207759857, "learning_rate": 3.2215458675727495e-05, "loss": 0.0258, "step": 19050 }, { "grad_norm": 0.16802600026130676, "learning_rate": 3.216395819197438e-05, "loss": 0.0255, "step": 19060 }, { "grad_norm": 0.16625455021858215, "learning_rate": 3.211247938069387e-05, "loss": 0.0318, "step": 19070 }, { "grad_norm": 0.14249365031719208, "learning_rate": 3.206102230443759e-05, "loss": 0.0209, "step": 19080 }, { "grad_norm": 0.13370518386363983, "learning_rate": 3.2009587025730764e-05, "loss": 0.033, "step": 19090 }, { "grad_norm": 0.17647968232631683, "learning_rate": 3.195817360707207e-05, "loss": 0.0311, "step": 19100 }, { "grad_norm": 0.15116044878959656, "learning_rate": 3.19067821109337e-05, "loss": 0.025, "step": 19110 }, { "grad_norm": 0.1175459697842598, "learning_rate": 3.185541259976114e-05, "loss": 0.0195, "step": 19120 }, { "grad_norm": 0.12096242606639862, "learning_rate": 3.180406513597316e-05, "loss": 0.039, "step": 19130 }, { "grad_norm": 0.19856831431388855, "learning_rate": 3.1752739781961835e-05, "loss": 0.0202, "step": 19140 }, { "grad_norm": 0.15564022958278656, "learning_rate": 3.170143660009228e-05, "loss": 0.0138, "step": 19150 }, { "grad_norm": 0.18099430203437805, "learning_rate": 3.16501556527027e-05, "loss": 0.0261, "step": 19160 }, { "grad_norm": 0.13409096002578735, "learning_rate": 3.1598897002104265e-05, "loss": 0.0186, "step": 19170 }, { "grad_norm": 0.1453680545091629, "learning_rate": 3.154766071058108e-05, "loss": 0.024, "step": 19180 }, { "grad_norm": 0.18109342455863953, "learning_rate": 3.149644684039008e-05, "loss": 0.0394, "step": 19190 }, { "grad_norm": 0.16009053587913513, "learning_rate": 3.144525545376095e-05, "loss": 0.0238, "step": 19200 }, { "grad_norm": 0.1380680650472641, "learning_rate": 3.139408661289603e-05, "loss": 0.0208, "step": 19210 }, { "grad_norm": 0.16467247903347015, "learning_rate": 3.134294037997032e-05, "loss": 0.0211, "step": 19220 }, { "grad_norm": 0.1572534292936325, "learning_rate": 3.129181681713127e-05, "loss": 0.0259, "step": 19230 }, { "grad_norm": 0.12720642983913422, "learning_rate": 3.1240715986498855e-05, "loss": 0.0264, "step": 19240 }, { "grad_norm": 0.1801363080739975, "learning_rate": 3.1189637950165394e-05, "loss": 0.0251, "step": 19250 }, { "grad_norm": 0.15153776109218597, "learning_rate": 3.1138582770195544e-05, "loss": 0.0187, "step": 19260 }, { "grad_norm": 0.13876523077487946, "learning_rate": 3.108755050862615e-05, "loss": 0.0158, "step": 19270 }, { "grad_norm": 0.14863558113574982, "learning_rate": 3.1036541227466204e-05, "loss": 0.0227, "step": 19280 }, { "grad_norm": 0.23464426398277283, "learning_rate": 3.0985554988696786e-05, "loss": 0.0339, "step": 19290 }, { "grad_norm": 0.13351686298847198, "learning_rate": 3.093459185427102e-05, "loss": 0.0241, "step": 19300 }, { "grad_norm": 0.15337908267974854, "learning_rate": 3.0883651886113905e-05, "loss": 0.0223, "step": 19310 }, { "grad_norm": 0.10946591943502426, "learning_rate": 3.0832735146122294e-05, "loss": 0.0344, "step": 19320 }, { "grad_norm": 0.19487658143043518, "learning_rate": 3.078184169616485e-05, "loss": 0.0402, "step": 19330 }, { "grad_norm": 0.22088512778282166, "learning_rate": 3.073097159808187e-05, "loss": 0.0444, "step": 19340 }, { "grad_norm": 0.11876487731933594, "learning_rate": 3.068012491368537e-05, "loss": 0.0347, "step": 19350 }, { "grad_norm": 0.1271064430475235, "learning_rate": 3.062930170475885e-05, "loss": 0.015, "step": 19360 }, { "grad_norm": 0.17901334166526794, "learning_rate": 3.057850203305729e-05, "loss": 0.0234, "step": 19370 }, { "grad_norm": 0.1852279156446457, "learning_rate": 3.052772596030708e-05, "loss": 0.0305, "step": 19380 }, { "grad_norm": 0.21314957737922668, "learning_rate": 3.0476973548205943e-05, "loss": 0.0301, "step": 19390 }, { "grad_norm": 0.13682803511619568, "learning_rate": 3.0426244858422847e-05, "loss": 0.0221, "step": 19400 }, { "grad_norm": 0.16261282563209534, "learning_rate": 3.0375539952597943e-05, "loss": 0.0205, "step": 19410 }, { "grad_norm": 0.1439482569694519, "learning_rate": 3.0324858892342468e-05, "loss": 0.0198, "step": 19420 }, { "grad_norm": 0.21430012583732605, "learning_rate": 3.0274201739238672e-05, "loss": 0.0413, "step": 19430 }, { "grad_norm": 0.15873530507087708, "learning_rate": 3.0223568554839786e-05, "loss": 0.0209, "step": 19440 }, { "grad_norm": 0.18716482818126678, "learning_rate": 3.0172959400669886e-05, "loss": 0.0218, "step": 19450 }, { "grad_norm": 0.13811276853084564, "learning_rate": 3.0122374338223902e-05, "loss": 0.0234, "step": 19460 }, { "grad_norm": 0.14398421347141266, "learning_rate": 3.0071813428967427e-05, "loss": 0.0327, "step": 19470 }, { "grad_norm": 0.15316765010356903, "learning_rate": 3.0021276734336746e-05, "loss": 0.0172, "step": 19480 }, { "grad_norm": 0.2094481885433197, "learning_rate": 2.997076431573871e-05, "loss": 0.0376, "step": 19490 }, { "grad_norm": 0.15081949532032013, "learning_rate": 2.9920276234550636e-05, "loss": 0.0209, "step": 19500 }, { "grad_norm": 0.1410532146692276, "learning_rate": 2.986981255212035e-05, "loss": 0.0346, "step": 19510 }, { "grad_norm": 0.12697730958461761, "learning_rate": 2.9819373329765977e-05, "loss": 0.0197, "step": 19520 }, { "grad_norm": 0.16508854925632477, "learning_rate": 2.9768958628775902e-05, "loss": 0.0187, "step": 19530 }, { "grad_norm": 0.11260800808668137, "learning_rate": 2.9718568510408762e-05, "loss": 0.0197, "step": 19540 }, { "grad_norm": 0.1703542321920395, "learning_rate": 2.9668203035893272e-05, "loss": 0.0263, "step": 19550 }, { "grad_norm": 0.20560993254184723, "learning_rate": 2.9617862266428287e-05, "loss": 0.0364, "step": 19560 }, { "grad_norm": 0.16784992814064026, "learning_rate": 2.9567546263182556e-05, "loss": 0.0268, "step": 19570 }, { "grad_norm": 0.17267780005931854, "learning_rate": 2.951725508729476e-05, "loss": 0.025, "step": 19580 }, { "grad_norm": 0.143809512257576, "learning_rate": 2.946698879987344e-05, "loss": 0.0224, "step": 19590 }, { "grad_norm": 0.19188521802425385, "learning_rate": 2.9416747461996853e-05, "loss": 0.0288, "step": 19600 }, { "grad_norm": 0.21879516541957855, "learning_rate": 2.9366531134712973e-05, "loss": 0.0225, "step": 19610 }, { "grad_norm": 0.1874500960111618, "learning_rate": 2.9316339879039367e-05, "loss": 0.0266, "step": 19620 }, { "grad_norm": 0.1121116504073143, "learning_rate": 2.9266173755963167e-05, "loss": 0.0267, "step": 19630 }, { "grad_norm": 0.19667023420333862, "learning_rate": 2.9216032826440926e-05, "loss": 0.0207, "step": 19640 }, { "grad_norm": 0.1904008984565735, "learning_rate": 2.9165917151398592e-05, "loss": 0.024, "step": 19650 }, { "grad_norm": 0.17708151042461395, "learning_rate": 2.9115826791731428e-05, "loss": 0.0206, "step": 19660 }, { "grad_norm": 0.1747199445962906, "learning_rate": 2.906576180830398e-05, "loss": 0.0211, "step": 19670 }, { "grad_norm": 0.14770294725894928, "learning_rate": 2.9015722261949917e-05, "loss": 0.0209, "step": 19680 }, { "grad_norm": 0.14373403787612915, "learning_rate": 2.8965708213471986e-05, "loss": 0.0279, "step": 19690 }, { "grad_norm": 0.14920255541801453, "learning_rate": 2.8915719723641975e-05, "loss": 0.022, "step": 19700 }, { "grad_norm": 0.11723116040229797, "learning_rate": 2.8865756853200604e-05, "loss": 0.0289, "step": 19710 }, { "grad_norm": 0.15885697305202484, "learning_rate": 2.8815819662857503e-05, "loss": 0.0287, "step": 19720 }, { "grad_norm": 0.16492117941379547, "learning_rate": 2.876590821329105e-05, "loss": 0.0217, "step": 19730 }, { "grad_norm": 0.15017856657505035, "learning_rate": 2.871602256514836e-05, "loss": 0.0239, "step": 19740 }, { "grad_norm": 0.1373099386692047, "learning_rate": 2.8666162779045203e-05, "loss": 0.0176, "step": 19750 }, { "grad_norm": 0.11794531345367432, "learning_rate": 2.8616328915565904e-05, "loss": 0.0325, "step": 19760 }, { "grad_norm": 0.2832516133785248, "learning_rate": 2.856652103526334e-05, "loss": 0.0195, "step": 19770 }, { "grad_norm": 0.1918783038854599, "learning_rate": 2.8516739198658755e-05, "loss": 0.0288, "step": 19780 }, { "grad_norm": 0.15655063092708588, "learning_rate": 2.846698346624177e-05, "loss": 0.0318, "step": 19790 }, { "grad_norm": 0.09872264415025711, "learning_rate": 2.841725389847032e-05, "loss": 0.021, "step": 19800 }, { "grad_norm": 0.23096135258674622, "learning_rate": 2.8367550555770506e-05, "loss": 0.0255, "step": 19810 }, { "grad_norm": 0.12892113626003265, "learning_rate": 2.831787349853655e-05, "loss": 0.0233, "step": 19820 }, { "grad_norm": 0.12921549379825592, "learning_rate": 2.8268222787130806e-05, "loss": 0.0165, "step": 19830 }, { "grad_norm": 0.13559190928936005, "learning_rate": 2.821859848188355e-05, "loss": 0.0205, "step": 19840 }, { "grad_norm": 0.21315738558769226, "learning_rate": 2.816900064309299e-05, "loss": 0.0395, "step": 19850 }, { "grad_norm": 0.20192131400108337, "learning_rate": 2.811942933102517e-05, "loss": 0.0318, "step": 19860 }, { "grad_norm": 0.18601413071155548, "learning_rate": 2.806988460591391e-05, "loss": 0.019, "step": 19870 }, { "grad_norm": 0.179751917719841, "learning_rate": 2.802036652796074e-05, "loss": 0.0297, "step": 19880 }, { "grad_norm": 0.12817206978797913, "learning_rate": 2.797087515733478e-05, "loss": 0.0272, "step": 19890 }, { "grad_norm": 0.21460416913032532, "learning_rate": 2.7921410554172723e-05, "loss": 0.0412, "step": 19900 }, { "grad_norm": 0.1487034410238266, "learning_rate": 2.787197277857871e-05, "loss": 0.0251, "step": 19910 }, { "grad_norm": 0.14532862603664398, "learning_rate": 2.782256189062429e-05, "loss": 0.0217, "step": 19920 }, { "grad_norm": 0.1354219615459442, "learning_rate": 2.777317795034839e-05, "loss": 0.0244, "step": 19930 }, { "grad_norm": 0.1775583028793335, "learning_rate": 2.7723821017757112e-05, "loss": 0.0251, "step": 19940 }, { "grad_norm": 0.12660355865955353, "learning_rate": 2.7674491152823822e-05, "loss": 0.0196, "step": 19950 }, { "grad_norm": 0.1475277543067932, "learning_rate": 2.7625188415488944e-05, "loss": 0.0247, "step": 19960 }, { "grad_norm": 0.19415554404258728, "learning_rate": 2.7575912865659924e-05, "loss": 0.0258, "step": 19970 }, { "grad_norm": 0.2152940183877945, "learning_rate": 2.7526664563211245e-05, "loss": 0.0269, "step": 19980 }, { "grad_norm": 0.16885441541671753, "learning_rate": 2.7477443567984224e-05, "loss": 0.0234, "step": 19990 }, { "grad_norm": 0.1688922494649887, "learning_rate": 2.7428249939787e-05, "loss": 0.0259, "step": 20000 } ], "logging_steps": 10, "max_steps": 30000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 10000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 8, "trial_name": null, "trial_params": null }