{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 80000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 1.0328662395477295, "learning_rate": 2.25e-07, "loss": 1.5136, "step": 10 }, { "grad_norm": 1.1689609289169312, "learning_rate": 4.75e-07, "loss": 1.5524, "step": 20 }, { "grad_norm": 1.2042536735534668, "learning_rate": 7.25e-07, "loss": 1.5357, "step": 30 }, { "grad_norm": 1.0271039009094238, "learning_rate": 9.75e-07, "loss": 1.5195, "step": 40 }, { "grad_norm": 1.208017110824585, "learning_rate": 1.2250000000000001e-06, "loss": 1.5034, "step": 50 }, { "grad_norm": 2.511094331741333, "learning_rate": 1.475e-06, "loss": 1.3896, "step": 60 }, { "grad_norm": 2.243584394454956, "learning_rate": 1.7250000000000002e-06, "loss": 1.2733, "step": 70 }, { "grad_norm": 2.1840550899505615, "learning_rate": 1.975e-06, "loss": 1.228, "step": 80 }, { "grad_norm": 1.2740987539291382, "learning_rate": 2.225e-06, "loss": 1.2112, "step": 90 }, { "grad_norm": 1.3027572631835938, "learning_rate": 2.4750000000000004e-06, "loss": 1.1961, "step": 100 }, { "grad_norm": 1.5236798524856567, "learning_rate": 2.725e-06, "loss": 1.1636, "step": 110 }, { "grad_norm": 1.586569905281067, "learning_rate": 2.975e-06, "loss": 1.1516, "step": 120 }, { "grad_norm": 1.2753092050552368, "learning_rate": 3.225e-06, "loss": 1.1591, "step": 130 }, { "grad_norm": 1.567997694015503, "learning_rate": 3.4750000000000006e-06, "loss": 1.1335, "step": 140 }, { "grad_norm": 1.5441985130310059, "learning_rate": 3.725e-06, "loss": 1.118, "step": 150 }, { "grad_norm": 3.1526119709014893, "learning_rate": 3.975e-06, "loss": 1.0893, "step": 160 }, { "grad_norm": 2.035835027694702, "learning_rate": 4.225e-06, "loss": 1.1407, "step": 170 }, { "grad_norm": 2.761136770248413, "learning_rate": 4.475e-06, "loss": 1.1353, "step": 180 }, { "grad_norm": 1.5223788022994995, "learning_rate": 4.7250000000000005e-06, "loss": 1.1368, "step": 190 }, { "grad_norm": 1.5674704313278198, "learning_rate": 4.975000000000001e-06, "loss": 1.1216, "step": 200 }, { "grad_norm": 1.5519776344299316, "learning_rate": 5.225e-06, "loss": 1.1126, "step": 210 }, { "grad_norm": 1.2724462747573853, "learning_rate": 5.475e-06, "loss": 1.1123, "step": 220 }, { "grad_norm": 1.5307323932647705, "learning_rate": 5.725e-06, "loss": 1.1219, "step": 230 }, { "grad_norm": 1.1449670791625977, "learning_rate": 5.975e-06, "loss": 1.0977, "step": 240 }, { "grad_norm": 1.287976622581482, "learning_rate": 6.2250000000000005e-06, "loss": 1.0636, "step": 250 }, { "grad_norm": 1.2962383031845093, "learning_rate": 6.475000000000001e-06, "loss": 1.0721, "step": 260 }, { "grad_norm": 1.8601900339126587, "learning_rate": 6.725000000000001e-06, "loss": 1.0961, "step": 270 }, { "grad_norm": 1.3811486959457397, "learning_rate": 6.975000000000001e-06, "loss": 1.0805, "step": 280 }, { "grad_norm": 1.97438645362854, "learning_rate": 7.2249999999999994e-06, "loss": 1.0801, "step": 290 }, { "grad_norm": 1.4476057291030884, "learning_rate": 7.4750000000000004e-06, "loss": 1.0729, "step": 300 }, { "grad_norm": 1.297060489654541, "learning_rate": 7.725e-06, "loss": 1.0642, "step": 310 }, { "grad_norm": 1.5158865451812744, "learning_rate": 7.975e-06, "loss": 1.0816, "step": 320 }, { "grad_norm": 1.7402772903442383, "learning_rate": 8.225e-06, "loss": 1.0721, "step": 330 }, { "grad_norm": 1.647832989692688, "learning_rate": 8.475000000000001e-06, "loss": 1.0608, "step": 340 }, { "grad_norm": 1.3170913457870483, "learning_rate": 8.725e-06, "loss": 1.0695, "step": 350 }, { "grad_norm": 1.152583360671997, "learning_rate": 8.975e-06, "loss": 1.0861, "step": 360 }, { "grad_norm": 1.4112064838409424, "learning_rate": 9.225e-06, "loss": 1.0654, "step": 370 }, { "grad_norm": 1.5936031341552734, "learning_rate": 9.475e-06, "loss": 1.0771, "step": 380 }, { "grad_norm": 1.25896155834198, "learning_rate": 9.725000000000001e-06, "loss": 1.0401, "step": 390 }, { "grad_norm": 1.0166233777999878, "learning_rate": 9.975e-06, "loss": 1.0596, "step": 400 }, { "grad_norm": 1.2879701852798462, "learning_rate": 1.0225e-05, "loss": 1.0545, "step": 410 }, { "grad_norm": 1.1139379739761353, "learning_rate": 1.0475e-05, "loss": 1.0529, "step": 420 }, { "grad_norm": 1.6513571739196777, "learning_rate": 1.0725e-05, "loss": 1.0364, "step": 430 }, { "grad_norm": 1.3485082387924194, "learning_rate": 1.0975e-05, "loss": 1.0402, "step": 440 }, { "grad_norm": 1.7847754955291748, "learning_rate": 1.1225e-05, "loss": 1.0039, "step": 450 }, { "grad_norm": 2.2866289615631104, "learning_rate": 1.1475000000000001e-05, "loss": 0.9934, "step": 460 }, { "grad_norm": 1.9982353448867798, "learning_rate": 1.1725e-05, "loss": 0.9749, "step": 470 }, { "grad_norm": 2.1074252128601074, "learning_rate": 1.1975e-05, "loss": 0.9542, "step": 480 }, { "grad_norm": 2.7386529445648193, "learning_rate": 1.2225e-05, "loss": 0.9124, "step": 490 }, { "grad_norm": 2.279681921005249, "learning_rate": 1.2475e-05, "loss": 0.8876, "step": 500 }, { "grad_norm": 2.0968544483184814, "learning_rate": 1.2725000000000001e-05, "loss": 0.8647, "step": 510 }, { "grad_norm": 2.0159950256347656, "learning_rate": 1.2975e-05, "loss": 0.8242, "step": 520 }, { "grad_norm": 2.3953628540039062, "learning_rate": 1.3225000000000001e-05, "loss": 0.808, "step": 530 }, { "grad_norm": 2.0120930671691895, "learning_rate": 1.3475000000000002e-05, "loss": 0.7253, "step": 540 }, { "grad_norm": 2.1331374645233154, "learning_rate": 1.3725000000000002e-05, "loss": 0.7254, "step": 550 }, { "grad_norm": 2.344621419906616, "learning_rate": 1.3975000000000003e-05, "loss": 0.7146, "step": 560 }, { "grad_norm": 2.1520438194274902, "learning_rate": 1.4225e-05, "loss": 0.6652, "step": 570 }, { "grad_norm": 2.5088860988616943, "learning_rate": 1.4475e-05, "loss": 0.6239, "step": 580 }, { "grad_norm": 2.295358896255493, "learning_rate": 1.4725e-05, "loss": 0.611, "step": 590 }, { "grad_norm": 2.3692541122436523, "learning_rate": 1.4975e-05, "loss": 0.5756, "step": 600 }, { "grad_norm": 2.920039653778076, "learning_rate": 1.5225e-05, "loss": 0.5584, "step": 610 }, { "grad_norm": 4.357720375061035, "learning_rate": 1.5475e-05, "loss": 0.5453, "step": 620 }, { "grad_norm": 2.8271546363830566, "learning_rate": 1.5725e-05, "loss": 0.5129, "step": 630 }, { "grad_norm": 2.8659210205078125, "learning_rate": 1.5975000000000002e-05, "loss": 0.4895, "step": 640 }, { "grad_norm": 2.5481314659118652, "learning_rate": 1.6225e-05, "loss": 0.4526, "step": 650 }, { "grad_norm": 2.933938980102539, "learning_rate": 1.6475e-05, "loss": 0.4537, "step": 660 }, { "grad_norm": 2.4897520542144775, "learning_rate": 1.6725000000000003e-05, "loss": 0.3941, "step": 670 }, { "grad_norm": 3.0426886081695557, "learning_rate": 1.6975000000000003e-05, "loss": 0.4049, "step": 680 }, { "grad_norm": 2.6086580753326416, "learning_rate": 1.7225e-05, "loss": 0.3878, "step": 690 }, { "grad_norm": 2.9501137733459473, "learning_rate": 1.7475e-05, "loss": 0.3906, "step": 700 }, { "grad_norm": 2.6428489685058594, "learning_rate": 1.7725e-05, "loss": 0.3273, "step": 710 }, { "grad_norm": 2.541508436203003, "learning_rate": 1.7975e-05, "loss": 0.3639, "step": 720 }, { "grad_norm": 2.2458267211914062, "learning_rate": 1.8225e-05, "loss": 0.3043, "step": 730 }, { "grad_norm": 2.2130587100982666, "learning_rate": 1.8475000000000002e-05, "loss": 0.2783, "step": 740 }, { "grad_norm": 2.5213332176208496, "learning_rate": 1.8725e-05, "loss": 0.2798, "step": 750 }, { "grad_norm": 2.1702418327331543, "learning_rate": 1.8975e-05, "loss": 0.2765, "step": 760 }, { "grad_norm": 2.2712292671203613, "learning_rate": 1.9225e-05, "loss": 0.2671, "step": 770 }, { "grad_norm": 2.9880294799804688, "learning_rate": 1.9475000000000002e-05, "loss": 0.2871, "step": 780 }, { "grad_norm": 2.2898311614990234, "learning_rate": 1.9725000000000002e-05, "loss": 0.2302, "step": 790 }, { "grad_norm": 2.7754762172698975, "learning_rate": 1.9975e-05, "loss": 0.2141, "step": 800 }, { "grad_norm": 2.878180742263794, "learning_rate": 2.0225000000000004e-05, "loss": 0.2244, "step": 810 }, { "grad_norm": 2.2064568996429443, "learning_rate": 2.0475e-05, "loss": 0.2137, "step": 820 }, { "grad_norm": 2.6666882038116455, "learning_rate": 2.0725e-05, "loss": 0.2422, "step": 830 }, { "grad_norm": 2.136551856994629, "learning_rate": 2.0975e-05, "loss": 0.2329, "step": 840 }, { "grad_norm": 1.8120746612548828, "learning_rate": 2.1225e-05, "loss": 0.221, "step": 850 }, { "grad_norm": 2.9197940826416016, "learning_rate": 2.1475e-05, "loss": 0.194, "step": 860 }, { "grad_norm": 2.022698163986206, "learning_rate": 2.1725e-05, "loss": 0.2251, "step": 870 }, { "grad_norm": 2.8860764503479004, "learning_rate": 2.1975000000000002e-05, "loss": 0.1905, "step": 880 }, { "grad_norm": 2.150132894515991, "learning_rate": 2.2225e-05, "loss": 0.1967, "step": 890 }, { "grad_norm": 2.1377644538879395, "learning_rate": 2.2475e-05, "loss": 0.2115, "step": 900 }, { "grad_norm": 2.836089611053467, "learning_rate": 2.2725000000000003e-05, "loss": 0.2006, "step": 910 }, { "grad_norm": 1.7531088590621948, "learning_rate": 2.2975000000000003e-05, "loss": 0.1768, "step": 920 }, { "grad_norm": 2.4497644901275635, "learning_rate": 2.3225000000000002e-05, "loss": 0.1913, "step": 930 }, { "grad_norm": 2.581801414489746, "learning_rate": 2.3475e-05, "loss": 0.1802, "step": 940 }, { "grad_norm": 1.9976794719696045, "learning_rate": 2.3725e-05, "loss": 0.1645, "step": 950 }, { "grad_norm": 2.4764323234558105, "learning_rate": 2.3975e-05, "loss": 0.1785, "step": 960 }, { "grad_norm": 2.0191893577575684, "learning_rate": 2.4225e-05, "loss": 0.2068, "step": 970 }, { "grad_norm": 1.8471037149429321, "learning_rate": 2.4475000000000002e-05, "loss": 0.1559, "step": 980 }, { "grad_norm": 2.3538401126861572, "learning_rate": 2.4725e-05, "loss": 0.145, "step": 990 }, { "grad_norm": 2.3089640140533447, "learning_rate": 2.4975e-05, "loss": 0.1572, "step": 1000 }, { "grad_norm": 1.7633494138717651, "learning_rate": 2.5225e-05, "loss": 0.1585, "step": 1010 }, { "grad_norm": 2.235785484313965, "learning_rate": 2.5475e-05, "loss": 0.1874, "step": 1020 }, { "grad_norm": 2.031984806060791, "learning_rate": 2.5725e-05, "loss": 0.1513, "step": 1030 }, { "grad_norm": 2.3918137550354004, "learning_rate": 2.5974999999999998e-05, "loss": 0.1598, "step": 1040 }, { "grad_norm": 2.6211469173431396, "learning_rate": 2.6225e-05, "loss": 0.1658, "step": 1050 }, { "grad_norm": 2.0548644065856934, "learning_rate": 2.6475e-05, "loss": 0.1753, "step": 1060 }, { "grad_norm": 1.9490976333618164, "learning_rate": 2.6725e-05, "loss": 0.1862, "step": 1070 }, { "grad_norm": 1.9773657321929932, "learning_rate": 2.6975000000000002e-05, "loss": 0.1943, "step": 1080 }, { "grad_norm": 1.6614948511123657, "learning_rate": 2.7225e-05, "loss": 0.1683, "step": 1090 }, { "grad_norm": 1.6311432123184204, "learning_rate": 2.7475e-05, "loss": 0.1261, "step": 1100 }, { "grad_norm": 2.296689033508301, "learning_rate": 2.7725e-05, "loss": 0.2031, "step": 1110 }, { "grad_norm": 1.4224168062210083, "learning_rate": 2.7975000000000002e-05, "loss": 0.135, "step": 1120 }, { "grad_norm": 1.7640620470046997, "learning_rate": 2.8225e-05, "loss": 0.1612, "step": 1130 }, { "grad_norm": 2.564225196838379, "learning_rate": 2.8475e-05, "loss": 0.204, "step": 1140 }, { "grad_norm": 1.9082331657409668, "learning_rate": 2.8725e-05, "loss": 0.1933, "step": 1150 }, { "grad_norm": 1.8484327793121338, "learning_rate": 2.8975000000000003e-05, "loss": 0.1614, "step": 1160 }, { "grad_norm": 1.615736722946167, "learning_rate": 2.9225000000000002e-05, "loss": 0.1666, "step": 1170 }, { "grad_norm": 1.7061376571655273, "learning_rate": 2.9475e-05, "loss": 0.1434, "step": 1180 }, { "grad_norm": 1.9195568561553955, "learning_rate": 2.9725000000000004e-05, "loss": 0.1569, "step": 1190 }, { "grad_norm": 3.1572186946868896, "learning_rate": 2.9975000000000004e-05, "loss": 0.1616, "step": 1200 }, { "grad_norm": 3.378594398498535, "learning_rate": 3.0225000000000003e-05, "loss": 0.1847, "step": 1210 }, { "grad_norm": 2.4771265983581543, "learning_rate": 3.0475000000000002e-05, "loss": 0.1643, "step": 1220 }, { "grad_norm": 1.82952880859375, "learning_rate": 3.0725e-05, "loss": 0.1604, "step": 1230 }, { "grad_norm": 1.7856000661849976, "learning_rate": 3.0975e-05, "loss": 0.1661, "step": 1240 }, { "grad_norm": 2.3240132331848145, "learning_rate": 3.122500000000001e-05, "loss": 0.1426, "step": 1250 }, { "grad_norm": 2.1474781036376953, "learning_rate": 3.1475e-05, "loss": 0.145, "step": 1260 }, { "grad_norm": 1.501360297203064, "learning_rate": 3.1725e-05, "loss": 0.1352, "step": 1270 }, { "grad_norm": 1.0712971687316895, "learning_rate": 3.1975e-05, "loss": 0.1471, "step": 1280 }, { "grad_norm": 1.6432610750198364, "learning_rate": 3.2225e-05, "loss": 0.1585, "step": 1290 }, { "grad_norm": 1.5312330722808838, "learning_rate": 3.2474999999999997e-05, "loss": 0.1402, "step": 1300 }, { "grad_norm": 1.959442377090454, "learning_rate": 3.2725e-05, "loss": 0.1381, "step": 1310 }, { "grad_norm": 1.79099440574646, "learning_rate": 3.2975e-05, "loss": 0.1489, "step": 1320 }, { "grad_norm": 1.605150818824768, "learning_rate": 3.3225e-05, "loss": 0.1221, "step": 1330 }, { "grad_norm": 1.6844086647033691, "learning_rate": 3.3475e-05, "loss": 0.1247, "step": 1340 }, { "grad_norm": 2.4498705863952637, "learning_rate": 3.3725e-05, "loss": 0.164, "step": 1350 }, { "grad_norm": 2.285212993621826, "learning_rate": 3.3975e-05, "loss": 0.1427, "step": 1360 }, { "grad_norm": 2.247990369796753, "learning_rate": 3.4225e-05, "loss": 0.1266, "step": 1370 }, { "grad_norm": 1.730440616607666, "learning_rate": 3.4475000000000005e-05, "loss": 0.1225, "step": 1380 }, { "grad_norm": 2.016171455383301, "learning_rate": 3.4725000000000004e-05, "loss": 0.1381, "step": 1390 }, { "grad_norm": 1.808403730392456, "learning_rate": 3.4975e-05, "loss": 0.1265, "step": 1400 }, { "grad_norm": 1.469344139099121, "learning_rate": 3.5225e-05, "loss": 0.1399, "step": 1410 }, { "grad_norm": 1.9100643396377563, "learning_rate": 3.5475e-05, "loss": 0.1353, "step": 1420 }, { "grad_norm": 1.4375065565109253, "learning_rate": 3.5725e-05, "loss": 0.1184, "step": 1430 }, { "grad_norm": 1.6596410274505615, "learning_rate": 3.5975e-05, "loss": 0.1499, "step": 1440 }, { "grad_norm": 1.3814241886138916, "learning_rate": 3.6225000000000006e-05, "loss": 0.1579, "step": 1450 }, { "grad_norm": 1.322798728942871, "learning_rate": 3.6475000000000006e-05, "loss": 0.1608, "step": 1460 }, { "grad_norm": 1.6310749053955078, "learning_rate": 3.6725000000000005e-05, "loss": 0.1425, "step": 1470 }, { "grad_norm": 1.3396354913711548, "learning_rate": 3.6975000000000004e-05, "loss": 0.131, "step": 1480 }, { "grad_norm": 1.8354558944702148, "learning_rate": 3.7225000000000004e-05, "loss": 0.1341, "step": 1490 }, { "grad_norm": 1.150276780128479, "learning_rate": 3.7475e-05, "loss": 0.1363, "step": 1500 }, { "grad_norm": 1.6193608045578003, "learning_rate": 3.7725e-05, "loss": 0.1136, "step": 1510 }, { "grad_norm": 1.657314419746399, "learning_rate": 3.7975e-05, "loss": 0.1346, "step": 1520 }, { "grad_norm": 1.3767895698547363, "learning_rate": 3.8225e-05, "loss": 0.139, "step": 1530 }, { "grad_norm": 1.325490117073059, "learning_rate": 3.8475e-05, "loss": 0.123, "step": 1540 }, { "grad_norm": 1.2461845874786377, "learning_rate": 3.8725e-05, "loss": 0.126, "step": 1550 }, { "grad_norm": 1.0879772901535034, "learning_rate": 3.8975e-05, "loss": 0.1466, "step": 1560 }, { "grad_norm": 1.988452672958374, "learning_rate": 3.9225e-05, "loss": 0.1153, "step": 1570 }, { "grad_norm": 1.239048957824707, "learning_rate": 3.9475000000000004e-05, "loss": 0.1262, "step": 1580 }, { "grad_norm": 1.487064242362976, "learning_rate": 3.9725e-05, "loss": 0.1353, "step": 1590 }, { "grad_norm": 1.421630620956421, "learning_rate": 3.9975e-05, "loss": 0.1156, "step": 1600 }, { "grad_norm": 1.2456861734390259, "learning_rate": 4.0225e-05, "loss": 0.1184, "step": 1610 }, { "grad_norm": 1.7707293033599854, "learning_rate": 4.0475e-05, "loss": 0.1366, "step": 1620 }, { "grad_norm": 1.5295865535736084, "learning_rate": 4.0725e-05, "loss": 0.1723, "step": 1630 }, { "grad_norm": 1.1927872896194458, "learning_rate": 4.0975e-05, "loss": 0.1286, "step": 1640 }, { "grad_norm": 1.4466147422790527, "learning_rate": 4.1225e-05, "loss": 0.15, "step": 1650 }, { "grad_norm": 1.4459306001663208, "learning_rate": 4.1475000000000005e-05, "loss": 0.138, "step": 1660 }, { "grad_norm": 1.148963212966919, "learning_rate": 4.1725000000000005e-05, "loss": 0.1098, "step": 1670 }, { "grad_norm": 1.6836074590682983, "learning_rate": 4.1975000000000004e-05, "loss": 0.1269, "step": 1680 }, { "grad_norm": 1.297226071357727, "learning_rate": 4.2225e-05, "loss": 0.1249, "step": 1690 }, { "grad_norm": 1.4558109045028687, "learning_rate": 4.2475e-05, "loss": 0.1363, "step": 1700 }, { "grad_norm": 1.8140822649002075, "learning_rate": 4.2725e-05, "loss": 0.158, "step": 1710 }, { "grad_norm": 1.4298676252365112, "learning_rate": 4.2975e-05, "loss": 0.1406, "step": 1720 }, { "grad_norm": 1.0250638723373413, "learning_rate": 4.322500000000001e-05, "loss": 0.133, "step": 1730 }, { "grad_norm": 1.4889627695083618, "learning_rate": 4.3475000000000006e-05, "loss": 0.1218, "step": 1740 }, { "grad_norm": 1.4854521751403809, "learning_rate": 4.3725000000000006e-05, "loss": 0.1333, "step": 1750 }, { "grad_norm": 1.5497492551803589, "learning_rate": 4.3975e-05, "loss": 0.1261, "step": 1760 }, { "grad_norm": 1.2332457304000854, "learning_rate": 4.4225e-05, "loss": 0.1143, "step": 1770 }, { "grad_norm": 1.2578283548355103, "learning_rate": 4.4475e-05, "loss": 0.1345, "step": 1780 }, { "grad_norm": 1.4776544570922852, "learning_rate": 4.4725e-05, "loss": 0.1194, "step": 1790 }, { "grad_norm": 1.428714394569397, "learning_rate": 4.4975e-05, "loss": 0.131, "step": 1800 }, { "grad_norm": 1.0270500183105469, "learning_rate": 4.5225e-05, "loss": 0.1245, "step": 1810 }, { "grad_norm": 1.6218750476837158, "learning_rate": 4.5475e-05, "loss": 0.132, "step": 1820 }, { "grad_norm": 1.9815082550048828, "learning_rate": 4.5725e-05, "loss": 0.1274, "step": 1830 }, { "grad_norm": 1.3202300071716309, "learning_rate": 4.5975e-05, "loss": 0.14, "step": 1840 }, { "grad_norm": 1.150829553604126, "learning_rate": 4.6225e-05, "loss": 0.1552, "step": 1850 }, { "grad_norm": 1.4913369417190552, "learning_rate": 4.6475000000000005e-05, "loss": 0.1598, "step": 1860 }, { "grad_norm": 1.2847685813903809, "learning_rate": 4.6725000000000004e-05, "loss": 0.1264, "step": 1870 }, { "grad_norm": 1.2177637815475464, "learning_rate": 4.6975000000000003e-05, "loss": 0.1104, "step": 1880 }, { "grad_norm": 1.4630566835403442, "learning_rate": 4.7225e-05, "loss": 0.1257, "step": 1890 }, { "grad_norm": 1.0128777027130127, "learning_rate": 4.7475e-05, "loss": 0.1148, "step": 1900 }, { "grad_norm": 1.7373851537704468, "learning_rate": 4.7725e-05, "loss": 0.1523, "step": 1910 }, { "grad_norm": 1.22897469997406, "learning_rate": 4.7975e-05, "loss": 0.1241, "step": 1920 }, { "grad_norm": 1.1518582105636597, "learning_rate": 4.822500000000001e-05, "loss": 0.11, "step": 1930 }, { "grad_norm": 1.355059027671814, "learning_rate": 4.8475000000000006e-05, "loss": 0.1265, "step": 1940 }, { "grad_norm": 1.133690357208252, "learning_rate": 4.8725000000000005e-05, "loss": 0.1027, "step": 1950 }, { "grad_norm": 1.4226763248443604, "learning_rate": 4.8975000000000005e-05, "loss": 0.1444, "step": 1960 }, { "grad_norm": 1.2100170850753784, "learning_rate": 4.9225000000000004e-05, "loss": 0.095, "step": 1970 }, { "grad_norm": 1.1051616668701172, "learning_rate": 4.9475e-05, "loss": 0.1438, "step": 1980 }, { "grad_norm": 1.4791134595870972, "learning_rate": 4.9725e-05, "loss": 0.1059, "step": 1990 }, { "grad_norm": 1.1801148653030396, "learning_rate": 4.9975e-05, "loss": 0.1322, "step": 2000 }, { "grad_norm": 1.0856540203094482, "learning_rate": 5.0225e-05, "loss": 0.1632, "step": 2010 }, { "grad_norm": 1.3778549432754517, "learning_rate": 5.047500000000001e-05, "loss": 0.1369, "step": 2020 }, { "grad_norm": 1.2730082273483276, "learning_rate": 5.0725e-05, "loss": 0.1133, "step": 2030 }, { "grad_norm": 1.445490837097168, "learning_rate": 5.0975000000000006e-05, "loss": 0.1295, "step": 2040 }, { "grad_norm": 1.1453441381454468, "learning_rate": 5.1225e-05, "loss": 0.1276, "step": 2050 }, { "grad_norm": 1.0250128507614136, "learning_rate": 5.1475000000000004e-05, "loss": 0.1215, "step": 2060 }, { "grad_norm": 0.9667590260505676, "learning_rate": 5.1725000000000004e-05, "loss": 0.1093, "step": 2070 }, { "grad_norm": 1.4803335666656494, "learning_rate": 5.197500000000001e-05, "loss": 0.1234, "step": 2080 }, { "grad_norm": 1.200923204421997, "learning_rate": 5.2225e-05, "loss": 0.1218, "step": 2090 }, { "grad_norm": 1.3695945739746094, "learning_rate": 5.247500000000001e-05, "loss": 0.12, "step": 2100 }, { "grad_norm": 1.0159200429916382, "learning_rate": 5.2725e-05, "loss": 0.1414, "step": 2110 }, { "grad_norm": 1.0376113653182983, "learning_rate": 5.297500000000001e-05, "loss": 0.1103, "step": 2120 }, { "grad_norm": 0.9988017082214355, "learning_rate": 5.3225e-05, "loss": 0.1116, "step": 2130 }, { "grad_norm": 1.0154062509536743, "learning_rate": 5.3475e-05, "loss": 0.1072, "step": 2140 }, { "grad_norm": 1.0814332962036133, "learning_rate": 5.3725000000000005e-05, "loss": 0.1377, "step": 2150 }, { "grad_norm": 1.1285018920898438, "learning_rate": 5.3975e-05, "loss": 0.1182, "step": 2160 }, { "grad_norm": 1.0869550704956055, "learning_rate": 5.4225000000000003e-05, "loss": 0.1102, "step": 2170 }, { "grad_norm": 1.1678965091705322, "learning_rate": 5.4474999999999996e-05, "loss": 0.1304, "step": 2180 }, { "grad_norm": 0.9459791779518127, "learning_rate": 5.4725e-05, "loss": 0.1151, "step": 2190 }, { "grad_norm": 1.05971360206604, "learning_rate": 5.4975e-05, "loss": 0.1146, "step": 2200 }, { "grad_norm": 0.9368377327919006, "learning_rate": 5.522500000000001e-05, "loss": 0.1192, "step": 2210 }, { "grad_norm": 1.2645275592803955, "learning_rate": 5.5475e-05, "loss": 0.1152, "step": 2220 }, { "grad_norm": 1.169435739517212, "learning_rate": 5.5725000000000006e-05, "loss": 0.1037, "step": 2230 }, { "grad_norm": 0.9205635786056519, "learning_rate": 5.5975e-05, "loss": 0.1009, "step": 2240 }, { "grad_norm": 1.0256762504577637, "learning_rate": 5.6225000000000005e-05, "loss": 0.1244, "step": 2250 }, { "grad_norm": 1.2559503316879272, "learning_rate": 5.6475e-05, "loss": 0.111, "step": 2260 }, { "grad_norm": 1.026831030845642, "learning_rate": 5.6725e-05, "loss": 0.105, "step": 2270 }, { "grad_norm": 1.0858796834945679, "learning_rate": 5.6975e-05, "loss": 0.1118, "step": 2280 }, { "grad_norm": 1.4110602140426636, "learning_rate": 5.722500000000001e-05, "loss": 0.142, "step": 2290 }, { "grad_norm": 1.145301103591919, "learning_rate": 5.7475e-05, "loss": 0.1028, "step": 2300 }, { "grad_norm": 1.004923939704895, "learning_rate": 5.772500000000001e-05, "loss": 0.1103, "step": 2310 }, { "grad_norm": 0.9659808278083801, "learning_rate": 5.7975e-05, "loss": 0.0924, "step": 2320 }, { "grad_norm": 1.018654465675354, "learning_rate": 5.8225000000000006e-05, "loss": 0.1294, "step": 2330 }, { "grad_norm": 0.9022621512413025, "learning_rate": 5.8475000000000005e-05, "loss": 0.1077, "step": 2340 }, { "grad_norm": 1.1307390928268433, "learning_rate": 5.8725000000000004e-05, "loss": 0.1204, "step": 2350 }, { "grad_norm": 0.9168297648429871, "learning_rate": 5.8975000000000004e-05, "loss": 0.1239, "step": 2360 }, { "grad_norm": 0.823207437992096, "learning_rate": 5.922500000000001e-05, "loss": 0.113, "step": 2370 }, { "grad_norm": 1.2945489883422852, "learning_rate": 5.9475e-05, "loss": 0.1205, "step": 2380 }, { "grad_norm": 1.0081720352172852, "learning_rate": 5.9724999999999995e-05, "loss": 0.1256, "step": 2390 }, { "grad_norm": 0.697842001914978, "learning_rate": 5.9975e-05, "loss": 0.1264, "step": 2400 }, { "grad_norm": 1.019767165184021, "learning_rate": 6.0225e-05, "loss": 0.1121, "step": 2410 }, { "grad_norm": 0.9723102450370789, "learning_rate": 6.0475000000000006e-05, "loss": 0.1485, "step": 2420 }, { "grad_norm": 1.250232219696045, "learning_rate": 6.0725e-05, "loss": 0.1163, "step": 2430 }, { "grad_norm": 1.2310651540756226, "learning_rate": 6.0975000000000005e-05, "loss": 0.1301, "step": 2440 }, { "grad_norm": 1.2303051948547363, "learning_rate": 6.1225e-05, "loss": 0.0985, "step": 2450 }, { "grad_norm": 0.9028247594833374, "learning_rate": 6.1475e-05, "loss": 0.1072, "step": 2460 }, { "grad_norm": 0.9706183075904846, "learning_rate": 6.1725e-05, "loss": 0.1115, "step": 2470 }, { "grad_norm": 1.0450119972229004, "learning_rate": 6.1975e-05, "loss": 0.1355, "step": 2480 }, { "grad_norm": 1.0372077226638794, "learning_rate": 6.2225e-05, "loss": 0.108, "step": 2490 }, { "grad_norm": 0.8526507019996643, "learning_rate": 6.2475e-05, "loss": 0.1028, "step": 2500 }, { "grad_norm": 0.904940664768219, "learning_rate": 6.2725e-05, "loss": 0.1027, "step": 2510 }, { "grad_norm": 0.8903220295906067, "learning_rate": 6.297500000000001e-05, "loss": 0.1154, "step": 2520 }, { "grad_norm": 1.2149169445037842, "learning_rate": 6.3225e-05, "loss": 0.0962, "step": 2530 }, { "grad_norm": 0.9712855815887451, "learning_rate": 6.347500000000001e-05, "loss": 0.109, "step": 2540 }, { "grad_norm": 0.9613971710205078, "learning_rate": 6.3725e-05, "loss": 0.1083, "step": 2550 }, { "grad_norm": 0.9605720043182373, "learning_rate": 6.397500000000001e-05, "loss": 0.1221, "step": 2560 }, { "grad_norm": 0.9375208616256714, "learning_rate": 6.4225e-05, "loss": 0.1088, "step": 2570 }, { "grad_norm": 0.9362852573394775, "learning_rate": 6.447500000000001e-05, "loss": 0.1202, "step": 2580 }, { "grad_norm": 0.7061117887496948, "learning_rate": 6.4725e-05, "loss": 0.1054, "step": 2590 }, { "grad_norm": 0.6797558069229126, "learning_rate": 6.497500000000001e-05, "loss": 0.1188, "step": 2600 }, { "grad_norm": 1.1706287860870361, "learning_rate": 6.5225e-05, "loss": 0.1207, "step": 2610 }, { "grad_norm": 0.8788914084434509, "learning_rate": 6.5475e-05, "loss": 0.0994, "step": 2620 }, { "grad_norm": 0.9512560367584229, "learning_rate": 6.5725e-05, "loss": 0.0983, "step": 2630 }, { "grad_norm": 0.9879671335220337, "learning_rate": 6.5975e-05, "loss": 0.1227, "step": 2640 }, { "grad_norm": 0.8501400947570801, "learning_rate": 6.6225e-05, "loss": 0.1174, "step": 2650 }, { "grad_norm": 0.8345410823822021, "learning_rate": 6.6475e-05, "loss": 0.1213, "step": 2660 }, { "grad_norm": 0.945735514163971, "learning_rate": 6.672500000000001e-05, "loss": 0.1332, "step": 2670 }, { "grad_norm": 0.9071312546730042, "learning_rate": 6.6975e-05, "loss": 0.1463, "step": 2680 }, { "grad_norm": 0.9229214787483215, "learning_rate": 6.722500000000001e-05, "loss": 0.1115, "step": 2690 }, { "grad_norm": 1.1614080667495728, "learning_rate": 6.7475e-05, "loss": 0.0977, "step": 2700 }, { "grad_norm": 0.7669945955276489, "learning_rate": 6.7725e-05, "loss": 0.1055, "step": 2710 }, { "grad_norm": 0.749561607837677, "learning_rate": 6.7975e-05, "loss": 0.1284, "step": 2720 }, { "grad_norm": 1.3414952754974365, "learning_rate": 6.8225e-05, "loss": 0.1107, "step": 2730 }, { "grad_norm": 0.6145594120025635, "learning_rate": 6.8475e-05, "loss": 0.1026, "step": 2740 }, { "grad_norm": 0.9295600056648254, "learning_rate": 6.8725e-05, "loss": 0.112, "step": 2750 }, { "grad_norm": 0.8891124725341797, "learning_rate": 6.8975e-05, "loss": 0.0967, "step": 2760 }, { "grad_norm": 0.8418741822242737, "learning_rate": 6.9225e-05, "loss": 0.108, "step": 2770 }, { "grad_norm": 0.8237596154212952, "learning_rate": 6.9475e-05, "loss": 0.1157, "step": 2780 }, { "grad_norm": 0.8060526251792908, "learning_rate": 6.9725e-05, "loss": 0.1135, "step": 2790 }, { "grad_norm": 0.8001449704170227, "learning_rate": 6.997500000000001e-05, "loss": 0.0956, "step": 2800 }, { "grad_norm": 1.0515576601028442, "learning_rate": 7.022500000000001e-05, "loss": 0.1402, "step": 2810 }, { "grad_norm": 1.2523713111877441, "learning_rate": 7.0475e-05, "loss": 0.1104, "step": 2820 }, { "grad_norm": 0.879880428314209, "learning_rate": 7.072500000000001e-05, "loss": 0.0993, "step": 2830 }, { "grad_norm": 0.8748754262924194, "learning_rate": 7.0975e-05, "loss": 0.1282, "step": 2840 }, { "grad_norm": 1.0113729238510132, "learning_rate": 7.122500000000001e-05, "loss": 0.1158, "step": 2850 }, { "grad_norm": 0.8133852481842041, "learning_rate": 7.1475e-05, "loss": 0.1029, "step": 2860 }, { "grad_norm": 0.7426754236221313, "learning_rate": 7.172500000000001e-05, "loss": 0.0863, "step": 2870 }, { "grad_norm": 0.8741538524627686, "learning_rate": 7.1975e-05, "loss": 0.0933, "step": 2880 }, { "grad_norm": 0.816001296043396, "learning_rate": 7.2225e-05, "loss": 0.0984, "step": 2890 }, { "grad_norm": 0.8128759264945984, "learning_rate": 7.2475e-05, "loss": 0.1094, "step": 2900 }, { "grad_norm": 0.8905858993530273, "learning_rate": 7.272499999999999e-05, "loss": 0.1031, "step": 2910 }, { "grad_norm": 0.7582471966743469, "learning_rate": 7.2975e-05, "loss": 0.1105, "step": 2920 }, { "grad_norm": 0.8837406039237976, "learning_rate": 7.3225e-05, "loss": 0.0848, "step": 2930 }, { "grad_norm": 0.7486787438392639, "learning_rate": 7.347500000000001e-05, "loss": 0.0979, "step": 2940 }, { "grad_norm": 0.9835020899772644, "learning_rate": 7.3725e-05, "loss": 0.096, "step": 2950 }, { "grad_norm": 0.6222724914550781, "learning_rate": 7.397500000000001e-05, "loss": 0.0978, "step": 2960 }, { "grad_norm": 0.6727334856987, "learning_rate": 7.4225e-05, "loss": 0.0868, "step": 2970 }, { "grad_norm": 0.8734923601150513, "learning_rate": 7.447500000000001e-05, "loss": 0.1025, "step": 2980 }, { "grad_norm": 0.8451271057128906, "learning_rate": 7.4725e-05, "loss": 0.1108, "step": 2990 }, { "grad_norm": 0.6489584445953369, "learning_rate": 7.4975e-05, "loss": 0.1005, "step": 3000 }, { "grad_norm": 0.7340289950370789, "learning_rate": 7.5225e-05, "loss": 0.0824, "step": 3010 }, { "grad_norm": 0.7004130482673645, "learning_rate": 7.5475e-05, "loss": 0.089, "step": 3020 }, { "grad_norm": 1.0088145732879639, "learning_rate": 7.5725e-05, "loss": 0.0952, "step": 3030 }, { "grad_norm": 0.9218837022781372, "learning_rate": 7.5975e-05, "loss": 0.1013, "step": 3040 }, { "grad_norm": 1.0240991115570068, "learning_rate": 7.6225e-05, "loss": 0.1102, "step": 3050 }, { "grad_norm": 0.6424623727798462, "learning_rate": 7.6475e-05, "loss": 0.1035, "step": 3060 }, { "grad_norm": 0.7994450926780701, "learning_rate": 7.672500000000001e-05, "loss": 0.0891, "step": 3070 }, { "grad_norm": 0.7884606122970581, "learning_rate": 7.697500000000001e-05, "loss": 0.1071, "step": 3080 }, { "grad_norm": 1.051068663597107, "learning_rate": 7.722500000000001e-05, "loss": 0.1227, "step": 3090 }, { "grad_norm": 0.9708969593048096, "learning_rate": 7.747500000000001e-05, "loss": 0.0962, "step": 3100 }, { "grad_norm": 0.7540818452835083, "learning_rate": 7.7725e-05, "loss": 0.0859, "step": 3110 }, { "grad_norm": 0.8590658903121948, "learning_rate": 7.797500000000001e-05, "loss": 0.0893, "step": 3120 }, { "grad_norm": 0.8453493118286133, "learning_rate": 7.8225e-05, "loss": 0.0926, "step": 3130 }, { "grad_norm": 0.7823694348335266, "learning_rate": 7.8475e-05, "loss": 0.111, "step": 3140 }, { "grad_norm": 0.8259350657463074, "learning_rate": 7.8725e-05, "loss": 0.0979, "step": 3150 }, { "grad_norm": 0.7200552225112915, "learning_rate": 7.8975e-05, "loss": 0.1003, "step": 3160 }, { "grad_norm": 0.7671678066253662, "learning_rate": 7.9225e-05, "loss": 0.0802, "step": 3170 }, { "grad_norm": 0.7029131650924683, "learning_rate": 7.9475e-05, "loss": 0.1098, "step": 3180 }, { "grad_norm": 0.7869446277618408, "learning_rate": 7.9725e-05, "loss": 0.1082, "step": 3190 }, { "grad_norm": 0.7521267533302307, "learning_rate": 7.9975e-05, "loss": 0.1045, "step": 3200 }, { "grad_norm": 0.8685215711593628, "learning_rate": 8.022500000000001e-05, "loss": 0.0856, "step": 3210 }, { "grad_norm": 0.9910880923271179, "learning_rate": 8.0475e-05, "loss": 0.1188, "step": 3220 }, { "grad_norm": 0.6529672741889954, "learning_rate": 8.072500000000001e-05, "loss": 0.103, "step": 3230 }, { "grad_norm": 0.4807416796684265, "learning_rate": 8.0975e-05, "loss": 0.0828, "step": 3240 }, { "grad_norm": 0.6473065614700317, "learning_rate": 8.122500000000001e-05, "loss": 0.107, "step": 3250 }, { "grad_norm": 0.8310666084289551, "learning_rate": 8.1475e-05, "loss": 0.1084, "step": 3260 }, { "grad_norm": 0.8657576441764832, "learning_rate": 8.172500000000001e-05, "loss": 0.1039, "step": 3270 }, { "grad_norm": 0.7877914309501648, "learning_rate": 8.1975e-05, "loss": 0.0849, "step": 3280 }, { "grad_norm": 0.7929897308349609, "learning_rate": 8.2225e-05, "loss": 0.0931, "step": 3290 }, { "grad_norm": 0.7079852223396301, "learning_rate": 8.2475e-05, "loss": 0.0899, "step": 3300 }, { "grad_norm": 1.1360104084014893, "learning_rate": 8.2725e-05, "loss": 0.1131, "step": 3310 }, { "grad_norm": 0.6342985033988953, "learning_rate": 8.2975e-05, "loss": 0.1182, "step": 3320 }, { "grad_norm": 0.6094357371330261, "learning_rate": 8.3225e-05, "loss": 0.1045, "step": 3330 }, { "grad_norm": 1.1308168172836304, "learning_rate": 8.347500000000001e-05, "loss": 0.1031, "step": 3340 }, { "grad_norm": 0.6325505971908569, "learning_rate": 8.3725e-05, "loss": 0.0965, "step": 3350 }, { "grad_norm": 0.8371400833129883, "learning_rate": 8.397500000000001e-05, "loss": 0.0967, "step": 3360 }, { "grad_norm": 0.769624650478363, "learning_rate": 8.422500000000001e-05, "loss": 0.103, "step": 3370 }, { "grad_norm": 0.7694975733757019, "learning_rate": 8.447500000000001e-05, "loss": 0.1511, "step": 3380 }, { "grad_norm": 0.8087249994277954, "learning_rate": 8.4725e-05, "loss": 0.0959, "step": 3390 }, { "grad_norm": 0.9223907589912415, "learning_rate": 8.4975e-05, "loss": 0.1008, "step": 3400 }, { "grad_norm": 1.2528499364852905, "learning_rate": 8.5225e-05, "loss": 0.1115, "step": 3410 }, { "grad_norm": 0.771386981010437, "learning_rate": 8.5475e-05, "loss": 0.0906, "step": 3420 }, { "grad_norm": 0.7208974957466125, "learning_rate": 8.5725e-05, "loss": 0.1046, "step": 3430 }, { "grad_norm": 0.6969635486602783, "learning_rate": 8.5975e-05, "loss": 0.114, "step": 3440 }, { "grad_norm": 0.9125112891197205, "learning_rate": 8.6225e-05, "loss": 0.1007, "step": 3450 }, { "grad_norm": 0.7293601632118225, "learning_rate": 8.6475e-05, "loss": 0.0896, "step": 3460 }, { "grad_norm": 0.6918874979019165, "learning_rate": 8.672500000000001e-05, "loss": 0.117, "step": 3470 }, { "grad_norm": 0.7723965048789978, "learning_rate": 8.6975e-05, "loss": 0.1164, "step": 3480 }, { "grad_norm": 0.7655568718910217, "learning_rate": 8.7225e-05, "loss": 0.1147, "step": 3490 }, { "grad_norm": 0.8602724671363831, "learning_rate": 8.747500000000001e-05, "loss": 0.0869, "step": 3500 }, { "grad_norm": 0.8766627907752991, "learning_rate": 8.7725e-05, "loss": 0.0992, "step": 3510 }, { "grad_norm": 0.6663209199905396, "learning_rate": 8.797500000000001e-05, "loss": 0.0955, "step": 3520 }, { "grad_norm": 0.9129557013511658, "learning_rate": 8.8225e-05, "loss": 0.1007, "step": 3530 }, { "grad_norm": 0.7264055609703064, "learning_rate": 8.847500000000001e-05, "loss": 0.104, "step": 3540 }, { "grad_norm": 0.7610189318656921, "learning_rate": 8.8725e-05, "loss": 0.1008, "step": 3550 }, { "grad_norm": 0.6456170082092285, "learning_rate": 8.897500000000001e-05, "loss": 0.0818, "step": 3560 }, { "grad_norm": 0.7188547849655151, "learning_rate": 8.9225e-05, "loss": 0.1059, "step": 3570 }, { "grad_norm": 0.7315117716789246, "learning_rate": 8.9475e-05, "loss": 0.1161, "step": 3580 }, { "grad_norm": 0.6749659776687622, "learning_rate": 8.9725e-05, "loss": 0.1047, "step": 3590 }, { "grad_norm": 0.7005340456962585, "learning_rate": 8.9975e-05, "loss": 0.0962, "step": 3600 }, { "grad_norm": 0.7710024118423462, "learning_rate": 9.0225e-05, "loss": 0.092, "step": 3610 }, { "grad_norm": 0.7919918894767761, "learning_rate": 9.0475e-05, "loss": 0.0939, "step": 3620 }, { "grad_norm": 0.6196247935295105, "learning_rate": 9.072500000000001e-05, "loss": 0.0799, "step": 3630 }, { "grad_norm": 0.8039494752883911, "learning_rate": 9.0975e-05, "loss": 0.0894, "step": 3640 }, { "grad_norm": 0.6730365753173828, "learning_rate": 9.122500000000001e-05, "loss": 0.0921, "step": 3650 }, { "grad_norm": 0.7997238636016846, "learning_rate": 9.1475e-05, "loss": 0.099, "step": 3660 }, { "grad_norm": 0.8394008278846741, "learning_rate": 9.172500000000001e-05, "loss": 0.1182, "step": 3670 }, { "grad_norm": 0.6162999868392944, "learning_rate": 9.1975e-05, "loss": 0.091, "step": 3680 }, { "grad_norm": 0.8285170197486877, "learning_rate": 9.2225e-05, "loss": 0.0987, "step": 3690 }, { "grad_norm": 0.7257769107818604, "learning_rate": 9.2475e-05, "loss": 0.0915, "step": 3700 }, { "grad_norm": 0.8368633389472961, "learning_rate": 9.2725e-05, "loss": 0.0974, "step": 3710 }, { "grad_norm": 0.8175399899482727, "learning_rate": 9.2975e-05, "loss": 0.1031, "step": 3720 }, { "grad_norm": 0.8685070276260376, "learning_rate": 9.3225e-05, "loss": 0.0931, "step": 3730 }, { "grad_norm": 0.5912356376647949, "learning_rate": 9.3475e-05, "loss": 0.1005, "step": 3740 }, { "grad_norm": 0.6671695113182068, "learning_rate": 9.3725e-05, "loss": 0.1195, "step": 3750 }, { "grad_norm": 0.6472877264022827, "learning_rate": 9.397500000000001e-05, "loss": 0.0941, "step": 3760 }, { "grad_norm": 0.6223649382591248, "learning_rate": 9.422500000000001e-05, "loss": 0.1024, "step": 3770 }, { "grad_norm": 0.7702328562736511, "learning_rate": 9.4475e-05, "loss": 0.1013, "step": 3780 }, { "grad_norm": 0.7327327728271484, "learning_rate": 9.472500000000001e-05, "loss": 0.1016, "step": 3790 }, { "grad_norm": 0.7549030780792236, "learning_rate": 9.4975e-05, "loss": 0.131, "step": 3800 }, { "grad_norm": 0.6883710026741028, "learning_rate": 9.522500000000001e-05, "loss": 0.0857, "step": 3810 }, { "grad_norm": 0.9502581357955933, "learning_rate": 9.5475e-05, "loss": 0.1161, "step": 3820 }, { "grad_norm": 0.7633139491081238, "learning_rate": 9.572500000000001e-05, "loss": 0.092, "step": 3830 }, { "grad_norm": 0.6718946695327759, "learning_rate": 9.5975e-05, "loss": 0.0893, "step": 3840 }, { "grad_norm": 0.6769065856933594, "learning_rate": 9.622500000000001e-05, "loss": 0.1145, "step": 3850 }, { "grad_norm": 1.025349497795105, "learning_rate": 9.6475e-05, "loss": 0.0898, "step": 3860 }, { "grad_norm": 1.0190850496292114, "learning_rate": 9.6725e-05, "loss": 0.0904, "step": 3870 }, { "grad_norm": 0.6726580262184143, "learning_rate": 9.6975e-05, "loss": 0.0916, "step": 3880 }, { "grad_norm": 0.7465143799781799, "learning_rate": 9.7225e-05, "loss": 0.0986, "step": 3890 }, { "grad_norm": 0.958745002746582, "learning_rate": 9.747500000000001e-05, "loss": 0.1027, "step": 3900 }, { "grad_norm": 0.8315201997756958, "learning_rate": 9.7725e-05, "loss": 0.1028, "step": 3910 }, { "grad_norm": 0.7692592144012451, "learning_rate": 9.797500000000001e-05, "loss": 0.1083, "step": 3920 }, { "grad_norm": 0.7633219957351685, "learning_rate": 9.8225e-05, "loss": 0.084, "step": 3930 }, { "grad_norm": 0.7450398206710815, "learning_rate": 9.847500000000001e-05, "loss": 0.1054, "step": 3940 }, { "grad_norm": 0.6582277417182922, "learning_rate": 9.8725e-05, "loss": 0.0964, "step": 3950 }, { "grad_norm": 0.8869377374649048, "learning_rate": 9.897500000000001e-05, "loss": 0.1161, "step": 3960 }, { "grad_norm": 0.7684257626533508, "learning_rate": 9.9225e-05, "loss": 0.0995, "step": 3970 }, { "grad_norm": 0.7523536682128906, "learning_rate": 9.9475e-05, "loss": 0.1112, "step": 3980 }, { "grad_norm": 0.9562953114509583, "learning_rate": 9.9725e-05, "loss": 0.1052, "step": 3990 }, { "grad_norm": 0.89043128490448, "learning_rate": 9.9975e-05, "loss": 0.1081, "step": 4000 }, { "grad_norm": 0.7551044821739197, "learning_rate": 9.999999653982884e-05, "loss": 0.1031, "step": 4010 }, { "grad_norm": 0.5894122123718262, "learning_rate": 9.999998457874392e-05, "loss": 0.0951, "step": 4020 }, { "grad_norm": 0.9385680556297302, "learning_rate": 9.999996407402913e-05, "loss": 0.102, "step": 4030 }, { "grad_norm": 0.635469377040863, "learning_rate": 9.999993502568801e-05, "loss": 0.0709, "step": 4040 }, { "grad_norm": 0.6419922113418579, "learning_rate": 9.999989743372548e-05, "loss": 0.1008, "step": 4050 }, { "grad_norm": 0.7395955324172974, "learning_rate": 9.999985129814798e-05, "loss": 0.0937, "step": 4060 }, { "grad_norm": 0.6271235942840576, "learning_rate": 9.99997966189634e-05, "loss": 0.0841, "step": 4070 }, { "grad_norm": 0.672227680683136, "learning_rate": 9.999973339618107e-05, "loss": 0.1083, "step": 4080 }, { "grad_norm": 0.7781266570091248, "learning_rate": 9.999966162981179e-05, "loss": 0.0977, "step": 4090 }, { "grad_norm": 0.9488053917884827, "learning_rate": 9.999958131986784e-05, "loss": 0.0929, "step": 4100 }, { "grad_norm": 0.6974180340766907, "learning_rate": 9.999949246636293e-05, "loss": 0.1044, "step": 4110 }, { "grad_norm": 0.6133376955986023, "learning_rate": 9.999939506931224e-05, "loss": 0.1083, "step": 4120 }, { "grad_norm": 0.669640839099884, "learning_rate": 9.999928912873243e-05, "loss": 0.0925, "step": 4130 }, { "grad_norm": 0.730303168296814, "learning_rate": 9.999917464464159e-05, "loss": 0.108, "step": 4140 }, { "grad_norm": 0.5989996790885925, "learning_rate": 9.999905161705929e-05, "loss": 0.1142, "step": 4150 }, { "grad_norm": 0.7011979818344116, "learning_rate": 9.999892004600653e-05, "loss": 0.0974, "step": 4160 }, { "grad_norm": 0.8539503812789917, "learning_rate": 9.999877993150581e-05, "loss": 0.0897, "step": 4170 }, { "grad_norm": 0.5839387774467468, "learning_rate": 9.999863127358108e-05, "loss": 0.0835, "step": 4180 }, { "grad_norm": 0.7205541729927063, "learning_rate": 9.999847407225773e-05, "loss": 0.0844, "step": 4190 }, { "grad_norm": 0.6833751201629639, "learning_rate": 9.999830832756262e-05, "loss": 0.0848, "step": 4200 }, { "grad_norm": 0.869945228099823, "learning_rate": 9.999813403952407e-05, "loss": 0.0874, "step": 4210 }, { "grad_norm": 0.7706557512283325, "learning_rate": 9.999795120817187e-05, "loss": 0.1029, "step": 4220 }, { "grad_norm": 0.8312711119651794, "learning_rate": 9.999775983353725e-05, "loss": 0.1206, "step": 4230 }, { "grad_norm": 0.6948394179344177, "learning_rate": 9.999755991565292e-05, "loss": 0.1092, "step": 4240 }, { "grad_norm": 0.6581438779830933, "learning_rate": 9.999735145455303e-05, "loss": 0.1076, "step": 4250 }, { "grad_norm": 0.6573034524917603, "learning_rate": 9.99971344502732e-05, "loss": 0.1157, "step": 4260 }, { "grad_norm": 0.5816931128501892, "learning_rate": 9.999690890285053e-05, "loss": 0.0992, "step": 4270 }, { "grad_norm": 0.5880184769630432, "learning_rate": 9.999667481232356e-05, "loss": 0.0761, "step": 4280 }, { "grad_norm": 0.5846673846244812, "learning_rate": 9.999643217873225e-05, "loss": 0.0865, "step": 4290 }, { "grad_norm": 0.6218336224555969, "learning_rate": 9.999618100211809e-05, "loss": 0.0829, "step": 4300 }, { "grad_norm": 0.5355157256126404, "learning_rate": 9.999592128252402e-05, "loss": 0.0932, "step": 4310 }, { "grad_norm": 0.9917404651641846, "learning_rate": 9.999565301999437e-05, "loss": 0.1265, "step": 4320 }, { "grad_norm": 0.6724204421043396, "learning_rate": 9.999537621457502e-05, "loss": 0.0879, "step": 4330 }, { "grad_norm": 0.6264986395835876, "learning_rate": 9.999509086631323e-05, "loss": 0.0726, "step": 4340 }, { "grad_norm": 0.5383528470993042, "learning_rate": 9.99947969752578e-05, "loss": 0.071, "step": 4350 }, { "grad_norm": 0.683336079120636, "learning_rate": 9.999449454145891e-05, "loss": 0.0767, "step": 4360 }, { "grad_norm": 0.7883095145225525, "learning_rate": 9.999418356496827e-05, "loss": 0.0841, "step": 4370 }, { "grad_norm": 0.6819047927856445, "learning_rate": 9.999386404583899e-05, "loss": 0.0924, "step": 4380 }, { "grad_norm": 0.8027373552322388, "learning_rate": 9.999353598412568e-05, "loss": 0.08, "step": 4390 }, { "grad_norm": 0.5058581829071045, "learning_rate": 9.999319937988442e-05, "loss": 0.0966, "step": 4400 }, { "grad_norm": 0.5957497358322144, "learning_rate": 9.999285423317268e-05, "loss": 0.0968, "step": 4410 }, { "grad_norm": 0.5793278813362122, "learning_rate": 9.999250054404947e-05, "loss": 0.1156, "step": 4420 }, { "grad_norm": 0.6404590606689453, "learning_rate": 9.99921383125752e-05, "loss": 0.0973, "step": 4430 }, { "grad_norm": 0.5830166339874268, "learning_rate": 9.99917675388118e-05, "loss": 0.102, "step": 4440 }, { "grad_norm": 0.6710487604141235, "learning_rate": 9.99913882228226e-05, "loss": 0.1109, "step": 4450 }, { "grad_norm": 0.7171286940574646, "learning_rate": 9.999100036467242e-05, "loss": 0.0951, "step": 4460 }, { "grad_norm": 0.5154182314872742, "learning_rate": 9.999060396442753e-05, "loss": 0.0888, "step": 4470 }, { "grad_norm": 0.71336430311203, "learning_rate": 9.999019902215566e-05, "loss": 0.101, "step": 4480 }, { "grad_norm": 0.600866436958313, "learning_rate": 9.998978553792602e-05, "loss": 0.0899, "step": 4490 }, { "grad_norm": 0.6614140272140503, "learning_rate": 9.998936351180926e-05, "loss": 0.0885, "step": 4500 }, { "grad_norm": 0.7481375932693481, "learning_rate": 9.998893294387747e-05, "loss": 0.1164, "step": 4510 }, { "grad_norm": 0.8211782574653625, "learning_rate": 9.998849383420426e-05, "loss": 0.0934, "step": 4520 }, { "grad_norm": 0.88144850730896, "learning_rate": 9.998804618286465e-05, "loss": 0.097, "step": 4530 }, { "grad_norm": 1.0518277883529663, "learning_rate": 9.99875899899351e-05, "loss": 0.1072, "step": 4540 }, { "grad_norm": 0.7541211843490601, "learning_rate": 9.99871252554936e-05, "loss": 0.1033, "step": 4550 }, { "grad_norm": 0.7077826261520386, "learning_rate": 9.998665197961955e-05, "loss": 0.0815, "step": 4560 }, { "grad_norm": 0.6648160815238953, "learning_rate": 9.998617016239379e-05, "loss": 0.1123, "step": 4570 }, { "grad_norm": 1.2402167320251465, "learning_rate": 9.998567980389869e-05, "loss": 0.1067, "step": 4580 }, { "grad_norm": 0.6102001070976257, "learning_rate": 9.998518090421802e-05, "loss": 0.0938, "step": 4590 }, { "grad_norm": 0.7516087293624878, "learning_rate": 9.998467346343703e-05, "loss": 0.0878, "step": 4600 }, { "grad_norm": 0.7254472374916077, "learning_rate": 9.998415748164243e-05, "loss": 0.0876, "step": 4610 }, { "grad_norm": 0.7435702085494995, "learning_rate": 9.998363295892238e-05, "loss": 0.087, "step": 4620 }, { "grad_norm": 0.746528685092926, "learning_rate": 9.998309989536652e-05, "loss": 0.0845, "step": 4630 }, { "grad_norm": 0.5647684931755066, "learning_rate": 9.998255829106593e-05, "loss": 0.1015, "step": 4640 }, { "grad_norm": 0.6734121441841125, "learning_rate": 9.998200814611316e-05, "loss": 0.0838, "step": 4650 }, { "grad_norm": 0.7601632475852966, "learning_rate": 9.998144946060219e-05, "loss": 0.0992, "step": 4660 }, { "grad_norm": 0.8371598720550537, "learning_rate": 9.998088223462852e-05, "loss": 0.0892, "step": 4670 }, { "grad_norm": 0.9239290356636047, "learning_rate": 9.998030646828905e-05, "loss": 0.0897, "step": 4680 }, { "grad_norm": 0.6446864604949951, "learning_rate": 9.997972216168217e-05, "loss": 0.0976, "step": 4690 }, { "grad_norm": 0.7450882196426392, "learning_rate": 9.997912931490771e-05, "loss": 0.0945, "step": 4700 }, { "grad_norm": 0.7272372841835022, "learning_rate": 9.9978527928067e-05, "loss": 0.0924, "step": 4710 }, { "grad_norm": 0.5405566096305847, "learning_rate": 9.997791800126277e-05, "loss": 0.08, "step": 4720 }, { "grad_norm": 0.5730356574058533, "learning_rate": 9.997729953459927e-05, "loss": 0.1017, "step": 4730 }, { "grad_norm": 0.5183407068252563, "learning_rate": 9.997667252818214e-05, "loss": 0.097, "step": 4740 }, { "grad_norm": 0.7369403839111328, "learning_rate": 9.997603698211855e-05, "loss": 0.0836, "step": 4750 }, { "grad_norm": 0.7781545519828796, "learning_rate": 9.99753928965171e-05, "loss": 0.0927, "step": 4760 }, { "grad_norm": 0.7871254682540894, "learning_rate": 9.997474027148781e-05, "loss": 0.0907, "step": 4770 }, { "grad_norm": 0.7987176775932312, "learning_rate": 9.997407910714223e-05, "loss": 0.1051, "step": 4780 }, { "grad_norm": 0.7373534440994263, "learning_rate": 9.997340940359332e-05, "loss": 0.0794, "step": 4790 }, { "grad_norm": 0.814274787902832, "learning_rate": 9.997273116095552e-05, "loss": 0.0865, "step": 4800 }, { "grad_norm": 0.6890215873718262, "learning_rate": 9.997204437934473e-05, "loss": 0.0863, "step": 4810 }, { "grad_norm": 0.765272855758667, "learning_rate": 9.997134905887829e-05, "loss": 0.0887, "step": 4820 }, { "grad_norm": 0.5566312074661255, "learning_rate": 9.997064519967501e-05, "loss": 0.0912, "step": 4830 }, { "grad_norm": 0.5481608510017395, "learning_rate": 9.996993280185517e-05, "loss": 0.0865, "step": 4840 }, { "grad_norm": 0.8913837671279907, "learning_rate": 9.99692118655405e-05, "loss": 0.0909, "step": 4850 }, { "grad_norm": 0.5989277958869934, "learning_rate": 9.996848239085417e-05, "loss": 0.0798, "step": 4860 }, { "grad_norm": 0.524890124797821, "learning_rate": 9.996774437792085e-05, "loss": 0.0813, "step": 4870 }, { "grad_norm": 0.5820927023887634, "learning_rate": 9.996699782686664e-05, "loss": 0.0763, "step": 4880 }, { "grad_norm": 0.6223388910293579, "learning_rate": 9.996624273781909e-05, "loss": 0.1016, "step": 4890 }, { "grad_norm": 0.6464083194732666, "learning_rate": 9.996547911090725e-05, "loss": 0.0768, "step": 4900 }, { "grad_norm": 0.7235556244850159, "learning_rate": 9.996470694626157e-05, "loss": 0.0855, "step": 4910 }, { "grad_norm": 0.5702294707298279, "learning_rate": 9.996392624401403e-05, "loss": 0.0833, "step": 4920 }, { "grad_norm": 0.693223774433136, "learning_rate": 9.996313700429801e-05, "loss": 0.0831, "step": 4930 }, { "grad_norm": 0.6161274909973145, "learning_rate": 9.996233922724836e-05, "loss": 0.0744, "step": 4940 }, { "grad_norm": 0.5246443748474121, "learning_rate": 9.996153291300141e-05, "loss": 0.1048, "step": 4950 }, { "grad_norm": 0.47407785058021545, "learning_rate": 9.996071806169494e-05, "loss": 0.0794, "step": 4960 }, { "grad_norm": 0.565161943435669, "learning_rate": 9.995989467346817e-05, "loss": 0.0926, "step": 4970 }, { "grad_norm": 0.5337037444114685, "learning_rate": 9.995906274846183e-05, "loss": 0.0877, "step": 4980 }, { "grad_norm": 0.7034614682197571, "learning_rate": 9.995822228681803e-05, "loss": 0.0925, "step": 4990 }, { "grad_norm": 0.5592183470726013, "learning_rate": 9.99573732886804e-05, "loss": 0.0904, "step": 5000 }, { "grad_norm": 0.6840940117835999, "learning_rate": 9.995651575419402e-05, "loss": 0.0891, "step": 5010 }, { "grad_norm": 0.664452314376831, "learning_rate": 9.995564968350541e-05, "loss": 0.0855, "step": 5020 }, { "grad_norm": 0.5547385811805725, "learning_rate": 9.995477507676256e-05, "loss": 0.0749, "step": 5030 }, { "grad_norm": 0.49265795946121216, "learning_rate": 9.995389193411493e-05, "loss": 0.0749, "step": 5040 }, { "grad_norm": 0.49629512429237366, "learning_rate": 9.995300025571339e-05, "loss": 0.0903, "step": 5050 }, { "grad_norm": 0.5139895081520081, "learning_rate": 9.995210004171034e-05, "loss": 0.0704, "step": 5060 }, { "grad_norm": 0.6944475173950195, "learning_rate": 9.995119129225956e-05, "loss": 0.1023, "step": 5070 }, { "grad_norm": 0.6022356152534485, "learning_rate": 9.995027400751637e-05, "loss": 0.0958, "step": 5080 }, { "grad_norm": 0.6919957399368286, "learning_rate": 9.994934818763751e-05, "loss": 0.0809, "step": 5090 }, { "grad_norm": 0.5696915984153748, "learning_rate": 9.994841383278115e-05, "loss": 0.0833, "step": 5100 }, { "grad_norm": 0.7108675241470337, "learning_rate": 9.994747094310695e-05, "loss": 0.0906, "step": 5110 }, { "grad_norm": 0.8169341683387756, "learning_rate": 9.994651951877604e-05, "loss": 0.1038, "step": 5120 }, { "grad_norm": 0.6593674421310425, "learning_rate": 9.994555955995099e-05, "loss": 0.0969, "step": 5130 }, { "grad_norm": 0.5479996800422668, "learning_rate": 9.994459106679581e-05, "loss": 0.0855, "step": 5140 }, { "grad_norm": 0.47798481583595276, "learning_rate": 9.994361403947603e-05, "loss": 0.0712, "step": 5150 }, { "grad_norm": 0.6549602150917053, "learning_rate": 9.994262847815854e-05, "loss": 0.1077, "step": 5160 }, { "grad_norm": 0.7830278277397156, "learning_rate": 9.99416343830118e-05, "loss": 0.0966, "step": 5170 }, { "grad_norm": 0.588175892829895, "learning_rate": 9.994063175420565e-05, "loss": 0.078, "step": 5180 }, { "grad_norm": 0.7305207252502441, "learning_rate": 9.99396205919114e-05, "loss": 0.1128, "step": 5190 }, { "grad_norm": 0.5265524983406067, "learning_rate": 9.993860089630185e-05, "loss": 0.0741, "step": 5200 }, { "grad_norm": 0.532950758934021, "learning_rate": 9.993757266755123e-05, "loss": 0.0844, "step": 5210 }, { "grad_norm": 0.6769217848777771, "learning_rate": 9.993653590583522e-05, "loss": 0.0812, "step": 5220 }, { "grad_norm": 0.4805484116077423, "learning_rate": 9.993549061133102e-05, "loss": 0.0785, "step": 5230 }, { "grad_norm": 0.5033955574035645, "learning_rate": 9.993443678421719e-05, "loss": 0.0817, "step": 5240 }, { "grad_norm": 0.676726758480072, "learning_rate": 9.993337442467384e-05, "loss": 0.0906, "step": 5250 }, { "grad_norm": 0.7177996635437012, "learning_rate": 9.993230353288248e-05, "loss": 0.0737, "step": 5260 }, { "grad_norm": 0.5749724507331848, "learning_rate": 9.993122410902608e-05, "loss": 0.0824, "step": 5270 }, { "grad_norm": 0.5085024237632751, "learning_rate": 9.993013615328912e-05, "loss": 0.0861, "step": 5280 }, { "grad_norm": 0.5508118867874146, "learning_rate": 9.992903966585747e-05, "loss": 0.091, "step": 5290 }, { "grad_norm": 0.523264467716217, "learning_rate": 9.992793464691852e-05, "loss": 0.0717, "step": 5300 }, { "grad_norm": 0.5478363037109375, "learning_rate": 9.992682109666105e-05, "loss": 0.0855, "step": 5310 }, { "grad_norm": 0.6109188199043274, "learning_rate": 9.992569901527538e-05, "loss": 0.0846, "step": 5320 }, { "grad_norm": 0.7151046395301819, "learning_rate": 9.99245684029532e-05, "loss": 0.1, "step": 5330 }, { "grad_norm": 0.7156488299369812, "learning_rate": 9.992342925988774e-05, "loss": 0.0822, "step": 5340 }, { "grad_norm": 0.5281471610069275, "learning_rate": 9.992228158627361e-05, "loss": 0.0754, "step": 5350 }, { "grad_norm": 0.577237069606781, "learning_rate": 9.992112538230693e-05, "loss": 0.08, "step": 5360 }, { "grad_norm": 0.5344470143318176, "learning_rate": 9.991996064818527e-05, "loss": 0.0799, "step": 5370 }, { "grad_norm": 0.6639402508735657, "learning_rate": 9.991878738410768e-05, "loss": 0.0723, "step": 5380 }, { "grad_norm": 0.6173306107521057, "learning_rate": 9.991760559027457e-05, "loss": 0.0769, "step": 5390 }, { "grad_norm": 0.46440523862838745, "learning_rate": 9.991641526688793e-05, "loss": 0.1077, "step": 5400 }, { "grad_norm": 0.504680335521698, "learning_rate": 9.991521641415113e-05, "loss": 0.0814, "step": 5410 }, { "grad_norm": 0.4069281816482544, "learning_rate": 9.991400903226904e-05, "loss": 0.0733, "step": 5420 }, { "grad_norm": 0.8015432953834534, "learning_rate": 9.991279312144794e-05, "loss": 0.0816, "step": 5430 }, { "grad_norm": 0.7303397059440613, "learning_rate": 9.991156868189564e-05, "loss": 0.0925, "step": 5440 }, { "grad_norm": 0.5952489376068115, "learning_rate": 9.991033571382131e-05, "loss": 0.0599, "step": 5450 }, { "grad_norm": 0.521420419216156, "learning_rate": 9.990909421743569e-05, "loss": 0.066, "step": 5460 }, { "grad_norm": 0.7935965061187744, "learning_rate": 9.990784419295085e-05, "loss": 0.0734, "step": 5470 }, { "grad_norm": 0.6624455451965332, "learning_rate": 9.990658564058044e-05, "loss": 0.0754, "step": 5480 }, { "grad_norm": 0.751872718334198, "learning_rate": 9.990531856053948e-05, "loss": 0.0849, "step": 5490 }, { "grad_norm": 0.5713666677474976, "learning_rate": 9.99040429530445e-05, "loss": 0.0913, "step": 5500 }, { "grad_norm": 0.6881601214408875, "learning_rate": 9.990275881831346e-05, "loss": 0.0885, "step": 5510 }, { "grad_norm": 0.493095338344574, "learning_rate": 9.990146615656577e-05, "loss": 0.0803, "step": 5520 }, { "grad_norm": 0.4791032373905182, "learning_rate": 9.990016496802233e-05, "loss": 0.0751, "step": 5530 }, { "grad_norm": 0.5048568248748779, "learning_rate": 9.989885525290548e-05, "loss": 0.0654, "step": 5540 }, { "grad_norm": 0.8875209093093872, "learning_rate": 9.989753701143897e-05, "loss": 0.0764, "step": 5550 }, { "grad_norm": 0.6379298567771912, "learning_rate": 9.989621024384812e-05, "loss": 0.0775, "step": 5560 }, { "grad_norm": 0.7552372217178345, "learning_rate": 9.989487495035959e-05, "loss": 0.0854, "step": 5570 }, { "grad_norm": 0.7418407797813416, "learning_rate": 9.989353113120156e-05, "loss": 0.0984, "step": 5580 }, { "grad_norm": 0.6307381987571716, "learning_rate": 9.989217878660366e-05, "loss": 0.0792, "step": 5590 }, { "grad_norm": 0.7856007814407349, "learning_rate": 9.989081791679695e-05, "loss": 0.0866, "step": 5600 }, { "grad_norm": 0.6324573755264282, "learning_rate": 9.988944852201397e-05, "loss": 0.0822, "step": 5610 }, { "grad_norm": 0.561950147151947, "learning_rate": 9.988807060248873e-05, "loss": 0.0737, "step": 5620 }, { "grad_norm": 0.5550152659416199, "learning_rate": 9.988668415845665e-05, "loss": 0.0698, "step": 5630 }, { "grad_norm": 0.724449098110199, "learning_rate": 9.988528919015466e-05, "loss": 0.0799, "step": 5640 }, { "grad_norm": 0.5707767009735107, "learning_rate": 9.988388569782112e-05, "loss": 0.0984, "step": 5650 }, { "grad_norm": 0.5718764662742615, "learning_rate": 9.988247368169583e-05, "loss": 0.0753, "step": 5660 }, { "grad_norm": 0.6652843356132507, "learning_rate": 9.988105314202007e-05, "loss": 0.0762, "step": 5670 }, { "grad_norm": 0.6974888443946838, "learning_rate": 9.987962407903659e-05, "loss": 0.0868, "step": 5680 }, { "grad_norm": 0.9026489853858948, "learning_rate": 9.987818649298957e-05, "loss": 0.0812, "step": 5690 }, { "grad_norm": 0.48835015296936035, "learning_rate": 9.987674038412465e-05, "loss": 0.0842, "step": 5700 }, { "grad_norm": 0.6081012487411499, "learning_rate": 9.987528575268891e-05, "loss": 0.082, "step": 5710 }, { "grad_norm": 0.569346010684967, "learning_rate": 9.987382259893095e-05, "loss": 0.0764, "step": 5720 }, { "grad_norm": 0.6468530297279358, "learning_rate": 9.987235092310074e-05, "loss": 0.0789, "step": 5730 }, { "grad_norm": 0.5714325308799744, "learning_rate": 9.987087072544978e-05, "loss": 0.0919, "step": 5740 }, { "grad_norm": 0.56179279088974, "learning_rate": 9.9869382006231e-05, "loss": 0.0747, "step": 5750 }, { "grad_norm": 0.572730302810669, "learning_rate": 9.986788476569875e-05, "loss": 0.0849, "step": 5760 }, { "grad_norm": 0.6408863663673401, "learning_rate": 9.986637900410887e-05, "loss": 0.0788, "step": 5770 }, { "grad_norm": 0.43984562158584595, "learning_rate": 9.986486472171869e-05, "loss": 0.0905, "step": 5780 }, { "grad_norm": 0.8146726489067078, "learning_rate": 9.986334191878692e-05, "loss": 0.0779, "step": 5790 }, { "grad_norm": 0.6222293972969055, "learning_rate": 9.986181059557378e-05, "loss": 0.0965, "step": 5800 }, { "grad_norm": 0.527115523815155, "learning_rate": 9.986027075234094e-05, "loss": 0.0781, "step": 5810 }, { "grad_norm": 0.47211703658103943, "learning_rate": 9.985872238935152e-05, "loss": 0.0833, "step": 5820 }, { "grad_norm": 0.613163948059082, "learning_rate": 9.985716550687008e-05, "loss": 0.0841, "step": 5830 }, { "grad_norm": 0.582830011844635, "learning_rate": 9.985560010516264e-05, "loss": 0.0745, "step": 5840 }, { "grad_norm": 0.5763563513755798, "learning_rate": 9.985402618449668e-05, "loss": 0.0732, "step": 5850 }, { "grad_norm": 0.7137253284454346, "learning_rate": 9.985244374514118e-05, "loss": 0.0999, "step": 5860 }, { "grad_norm": 0.4592060148715973, "learning_rate": 9.985085278736651e-05, "loss": 0.0838, "step": 5870 }, { "grad_norm": 0.6258699893951416, "learning_rate": 9.984925331144452e-05, "loss": 0.0829, "step": 5880 }, { "grad_norm": 0.5976858139038086, "learning_rate": 9.984764531764851e-05, "loss": 0.0866, "step": 5890 }, { "grad_norm": 0.6048508882522583, "learning_rate": 9.984602880625326e-05, "loss": 0.0965, "step": 5900 }, { "grad_norm": 0.5320999026298523, "learning_rate": 9.9844403777535e-05, "loss": 0.0817, "step": 5910 }, { "grad_norm": 0.7633347511291504, "learning_rate": 9.984277023177135e-05, "loss": 0.0765, "step": 5920 }, { "grad_norm": 0.5442296266555786, "learning_rate": 9.984112816924148e-05, "loss": 0.0734, "step": 5930 }, { "grad_norm": 0.5323933362960815, "learning_rate": 9.983947759022596e-05, "loss": 0.0782, "step": 5940 }, { "grad_norm": 0.5523156523704529, "learning_rate": 9.983781849500682e-05, "loss": 0.0728, "step": 5950 }, { "grad_norm": 0.7110001444816589, "learning_rate": 9.98361508838676e-05, "loss": 0.0751, "step": 5960 }, { "grad_norm": 0.6702684164047241, "learning_rate": 9.98344747570932e-05, "loss": 0.0783, "step": 5970 }, { "grad_norm": 0.8102836608886719, "learning_rate": 9.983279011497004e-05, "loss": 0.0847, "step": 5980 }, { "grad_norm": 0.7589284777641296, "learning_rate": 9.983109695778596e-05, "loss": 0.0973, "step": 5990 }, { "grad_norm": 0.5040148496627808, "learning_rate": 9.982939528583032e-05, "loss": 0.1086, "step": 6000 }, { "grad_norm": 0.5364176630973816, "learning_rate": 9.982768509939385e-05, "loss": 0.0838, "step": 6010 }, { "grad_norm": 0.6546822190284729, "learning_rate": 9.982596639876879e-05, "loss": 0.0805, "step": 6020 }, { "grad_norm": 0.6603705286979675, "learning_rate": 9.982423918424881e-05, "loss": 0.0865, "step": 6030 }, { "grad_norm": 0.6073014140129089, "learning_rate": 9.982250345612908e-05, "loss": 0.0792, "step": 6040 }, { "grad_norm": 0.5945658087730408, "learning_rate": 9.982075921470611e-05, "loss": 0.0878, "step": 6050 }, { "grad_norm": 0.5832676291465759, "learning_rate": 9.981900646027802e-05, "loss": 0.0785, "step": 6060 }, { "grad_norm": 0.515859842300415, "learning_rate": 9.981724519314425e-05, "loss": 0.0833, "step": 6070 }, { "grad_norm": 0.405252069234848, "learning_rate": 9.981547541360581e-05, "loss": 0.0879, "step": 6080 }, { "grad_norm": 0.6802921295166016, "learning_rate": 9.981369712196508e-05, "loss": 0.0871, "step": 6090 }, { "grad_norm": 0.6555290818214417, "learning_rate": 9.981191031852592e-05, "loss": 0.0985, "step": 6100 }, { "grad_norm": 0.47739967703819275, "learning_rate": 9.981011500359362e-05, "loss": 0.0732, "step": 6110 }, { "grad_norm": 0.5539408922195435, "learning_rate": 9.9808311177475e-05, "loss": 0.0793, "step": 6120 }, { "grad_norm": 0.5023041367530823, "learning_rate": 9.980649884047826e-05, "loss": 0.0632, "step": 6130 }, { "grad_norm": 0.6691679358482361, "learning_rate": 9.980467799291307e-05, "loss": 0.071, "step": 6140 }, { "grad_norm": 0.5908961296081543, "learning_rate": 9.980284863509058e-05, "loss": 0.1038, "step": 6150 }, { "grad_norm": 0.6370383501052856, "learning_rate": 9.980101076732334e-05, "loss": 0.1001, "step": 6160 }, { "grad_norm": 0.72963547706604, "learning_rate": 9.979916438992544e-05, "loss": 0.0868, "step": 6170 }, { "grad_norm": 0.5857909917831421, "learning_rate": 9.979730950321237e-05, "loss": 0.076, "step": 6180 }, { "grad_norm": 0.658112108707428, "learning_rate": 9.979544610750104e-05, "loss": 0.0901, "step": 6190 }, { "grad_norm": 0.6012122631072998, "learning_rate": 9.97935742031099e-05, "loss": 0.0771, "step": 6200 }, { "grad_norm": 0.7158591747283936, "learning_rate": 9.979169379035878e-05, "loss": 0.0835, "step": 6210 }, { "grad_norm": 0.5729275941848755, "learning_rate": 9.978980486956899e-05, "loss": 0.0771, "step": 6220 }, { "grad_norm": 0.444853276014328, "learning_rate": 9.978790744106332e-05, "loss": 0.0776, "step": 6230 }, { "grad_norm": 0.6297940611839294, "learning_rate": 9.978600150516594e-05, "loss": 0.0915, "step": 6240 }, { "grad_norm": 0.5209591388702393, "learning_rate": 9.978408706220259e-05, "loss": 0.0831, "step": 6250 }, { "grad_norm": 0.47320371866226196, "learning_rate": 9.978216411250032e-05, "loss": 0.0672, "step": 6260 }, { "grad_norm": 0.6665074229240417, "learning_rate": 9.978023265638778e-05, "loss": 0.084, "step": 6270 }, { "grad_norm": 0.5489752292633057, "learning_rate": 9.977829269419495e-05, "loss": 0.0878, "step": 6280 }, { "grad_norm": 0.6035710573196411, "learning_rate": 9.977634422625335e-05, "loss": 0.0813, "step": 6290 }, { "grad_norm": 0.5650031566619873, "learning_rate": 9.97743872528959e-05, "loss": 0.079, "step": 6300 }, { "grad_norm": 0.6504883766174316, "learning_rate": 9.9772421774457e-05, "loss": 0.0759, "step": 6310 }, { "grad_norm": 0.5861506462097168, "learning_rate": 9.977044779127252e-05, "loss": 0.0714, "step": 6320 }, { "grad_norm": 0.575430154800415, "learning_rate": 9.976846530367971e-05, "loss": 0.0736, "step": 6330 }, { "grad_norm": 0.7476906776428223, "learning_rate": 9.976647431201735e-05, "loss": 0.0931, "step": 6340 }, { "grad_norm": 0.4662312865257263, "learning_rate": 9.976447481662568e-05, "loss": 0.0724, "step": 6350 }, { "grad_norm": 0.4934978187084198, "learning_rate": 9.976246681784629e-05, "loss": 0.0777, "step": 6360 }, { "grad_norm": 0.5032116770744324, "learning_rate": 9.976045031602234e-05, "loss": 0.0665, "step": 6370 }, { "grad_norm": 0.5302112102508545, "learning_rate": 9.975842531149837e-05, "loss": 0.081, "step": 6380 }, { "grad_norm": 0.8142202496528625, "learning_rate": 9.975639180462043e-05, "loss": 0.0869, "step": 6390 }, { "grad_norm": 0.5986716151237488, "learning_rate": 9.975434979573596e-05, "loss": 0.0783, "step": 6400 }, { "grad_norm": 0.4230092465877533, "learning_rate": 9.97522992851939e-05, "loss": 0.0615, "step": 6410 }, { "grad_norm": 0.46313267946243286, "learning_rate": 9.975024027334461e-05, "loss": 0.0716, "step": 6420 }, { "grad_norm": 0.42409586906433105, "learning_rate": 9.974817276053993e-05, "loss": 0.0696, "step": 6430 }, { "grad_norm": 0.5587354898452759, "learning_rate": 9.974609674713315e-05, "loss": 0.0755, "step": 6440 }, { "grad_norm": 0.5103870034217834, "learning_rate": 9.9744012233479e-05, "loss": 0.069, "step": 6450 }, { "grad_norm": 0.5517359972000122, "learning_rate": 9.974191921993366e-05, "loss": 0.0683, "step": 6460 }, { "grad_norm": 0.5089661478996277, "learning_rate": 9.973981770685474e-05, "loss": 0.0696, "step": 6470 }, { "grad_norm": 0.5783112645149231, "learning_rate": 9.97377076946014e-05, "loss": 0.0628, "step": 6480 }, { "grad_norm": 0.6201136112213135, "learning_rate": 9.973558918353412e-05, "loss": 0.0789, "step": 6490 }, { "grad_norm": 0.4784587323665619, "learning_rate": 9.973346217401494e-05, "loss": 0.0659, "step": 6500 }, { "grad_norm": 0.578912079334259, "learning_rate": 9.973132666640726e-05, "loss": 0.0616, "step": 6510 }, { "grad_norm": 0.5413712859153748, "learning_rate": 9.972918266107602e-05, "loss": 0.0666, "step": 6520 }, { "grad_norm": 0.6305744051933289, "learning_rate": 9.972703015838756e-05, "loss": 0.0744, "step": 6530 }, { "grad_norm": 0.5716162919998169, "learning_rate": 9.97248691587097e-05, "loss": 0.0781, "step": 6540 }, { "grad_norm": 0.6169860363006592, "learning_rate": 9.972269966241166e-05, "loss": 0.0694, "step": 6550 }, { "grad_norm": 0.6230577230453491, "learning_rate": 9.972052166986417e-05, "loss": 0.0772, "step": 6560 }, { "grad_norm": 0.5851648449897766, "learning_rate": 9.971833518143938e-05, "loss": 0.0793, "step": 6570 }, { "grad_norm": 0.582610547542572, "learning_rate": 9.971614019751093e-05, "loss": 0.0679, "step": 6580 }, { "grad_norm": 0.6703491806983948, "learning_rate": 9.971393671845383e-05, "loss": 0.0793, "step": 6590 }, { "grad_norm": 0.47344425320625305, "learning_rate": 9.971172474464464e-05, "loss": 0.0915, "step": 6600 }, { "grad_norm": 0.5619080662727356, "learning_rate": 9.97095042764613e-05, "loss": 0.065, "step": 6610 }, { "grad_norm": 0.5164514780044556, "learning_rate": 9.970727531428324e-05, "loss": 0.0666, "step": 6620 }, { "grad_norm": 0.6527890563011169, "learning_rate": 9.970503785849132e-05, "loss": 0.0653, "step": 6630 }, { "grad_norm": 0.483933687210083, "learning_rate": 9.970279190946788e-05, "loss": 0.0595, "step": 6640 }, { "grad_norm": 0.4860641360282898, "learning_rate": 9.970053746759667e-05, "loss": 0.0733, "step": 6650 }, { "grad_norm": 0.44133874773979187, "learning_rate": 9.969827453326292e-05, "loss": 0.0655, "step": 6660 }, { "grad_norm": 0.39009228348731995, "learning_rate": 9.969600310685332e-05, "loss": 0.0609, "step": 6670 }, { "grad_norm": 0.47266843914985657, "learning_rate": 9.969372318875596e-05, "loss": 0.0758, "step": 6680 }, { "grad_norm": 0.5789667367935181, "learning_rate": 9.969143477936043e-05, "loss": 0.0821, "step": 6690 }, { "grad_norm": 0.5485224723815918, "learning_rate": 9.968913787905775e-05, "loss": 0.0801, "step": 6700 }, { "grad_norm": 0.686758816242218, "learning_rate": 9.968683248824045e-05, "loss": 0.0834, "step": 6710 }, { "grad_norm": 0.6226227283477783, "learning_rate": 9.968451860730238e-05, "loss": 0.095, "step": 6720 }, { "grad_norm": 0.6004628539085388, "learning_rate": 9.968219623663896e-05, "loss": 0.0778, "step": 6730 }, { "grad_norm": 0.5532531142234802, "learning_rate": 9.967986537664702e-05, "loss": 0.1058, "step": 6740 }, { "grad_norm": 0.49874356389045715, "learning_rate": 9.967752602772483e-05, "loss": 0.0766, "step": 6750 }, { "grad_norm": 0.5880352854728699, "learning_rate": 9.967517819027212e-05, "loss": 0.0788, "step": 6760 }, { "grad_norm": 0.7535375952720642, "learning_rate": 9.967282186469009e-05, "loss": 0.0607, "step": 6770 }, { "grad_norm": 0.7204724550247192, "learning_rate": 9.967045705138135e-05, "loss": 0.0798, "step": 6780 }, { "grad_norm": 0.5664523839950562, "learning_rate": 9.966808375074998e-05, "loss": 0.0728, "step": 6790 }, { "grad_norm": 0.5500394701957703, "learning_rate": 9.966570196320154e-05, "loss": 0.0765, "step": 6800 }, { "grad_norm": 0.6512762904167175, "learning_rate": 9.966331168914299e-05, "loss": 0.078, "step": 6810 }, { "grad_norm": 0.63872230052948, "learning_rate": 9.966091292898277e-05, "loss": 0.0886, "step": 6820 }, { "grad_norm": 0.6561751365661621, "learning_rate": 9.965850568313076e-05, "loss": 0.0807, "step": 6830 }, { "grad_norm": 0.6177130937576294, "learning_rate": 9.965608995199827e-05, "loss": 0.087, "step": 6840 }, { "grad_norm": 0.5641895532608032, "learning_rate": 9.965366573599812e-05, "loss": 0.0812, "step": 6850 }, { "grad_norm": 0.4698207378387451, "learning_rate": 9.965123303554453e-05, "loss": 0.0739, "step": 6860 }, { "grad_norm": 0.663740336894989, "learning_rate": 9.964879185105317e-05, "loss": 0.0836, "step": 6870 }, { "grad_norm": 0.472135454416275, "learning_rate": 9.964634218294119e-05, "loss": 0.0804, "step": 6880 }, { "grad_norm": 0.6114245653152466, "learning_rate": 9.964388403162714e-05, "loss": 0.0782, "step": 6890 }, { "grad_norm": 0.7680885791778564, "learning_rate": 9.96414173975311e-05, "loss": 0.0774, "step": 6900 }, { "grad_norm": 0.5138612985610962, "learning_rate": 9.963894228107451e-05, "loss": 0.0623, "step": 6910 }, { "grad_norm": 0.49595847725868225, "learning_rate": 9.963645868268032e-05, "loss": 0.074, "step": 6920 }, { "grad_norm": 0.5462167859077454, "learning_rate": 9.963396660277289e-05, "loss": 0.0795, "step": 6930 }, { "grad_norm": 0.6151776909828186, "learning_rate": 9.963146604177807e-05, "loss": 0.0679, "step": 6940 }, { "grad_norm": 0.5809751749038696, "learning_rate": 9.962895700012311e-05, "loss": 0.0701, "step": 6950 }, { "grad_norm": 0.5596855878829956, "learning_rate": 9.962643947823677e-05, "loss": 0.0773, "step": 6960 }, { "grad_norm": 0.5380963087081909, "learning_rate": 9.962391347654921e-05, "loss": 0.0828, "step": 6970 }, { "grad_norm": 0.5230687260627747, "learning_rate": 9.962137899549204e-05, "loss": 0.0829, "step": 6980 }, { "grad_norm": 0.5768856406211853, "learning_rate": 9.961883603549835e-05, "loss": 0.0674, "step": 6990 }, { "grad_norm": 0.5047861933708191, "learning_rate": 9.961628459700267e-05, "loss": 0.0727, "step": 7000 }, { "grad_norm": 0.5992554426193237, "learning_rate": 9.961372468044095e-05, "loss": 0.0811, "step": 7010 }, { "grad_norm": 0.5710558295249939, "learning_rate": 9.961115628625062e-05, "loss": 0.0674, "step": 7020 }, { "grad_norm": 0.6417977809906006, "learning_rate": 9.960857941487056e-05, "loss": 0.0905, "step": 7030 }, { "grad_norm": 0.5402456521987915, "learning_rate": 9.960599406674106e-05, "loss": 0.0795, "step": 7040 }, { "grad_norm": 0.6471374034881592, "learning_rate": 9.960340024230393e-05, "loss": 0.069, "step": 7050 }, { "grad_norm": 0.4585295617580414, "learning_rate": 9.960079794200232e-05, "loss": 0.0711, "step": 7060 }, { "grad_norm": 0.6774080991744995, "learning_rate": 9.959818716628096e-05, "loss": 0.0773, "step": 7070 }, { "grad_norm": 0.6237754821777344, "learning_rate": 9.95955679155859e-05, "loss": 0.0912, "step": 7080 }, { "grad_norm": 0.668380856513977, "learning_rate": 9.959294019036472e-05, "loss": 0.0894, "step": 7090 }, { "grad_norm": 0.7742587327957153, "learning_rate": 9.959030399106646e-05, "loss": 0.0759, "step": 7100 }, { "grad_norm": 0.6739948987960815, "learning_rate": 9.958765931814153e-05, "loss": 0.088, "step": 7110 }, { "grad_norm": 0.5484842658042908, "learning_rate": 9.958500617204184e-05, "loss": 0.0657, "step": 7120 }, { "grad_norm": 0.4767090380191803, "learning_rate": 9.958234455322075e-05, "loss": 0.0678, "step": 7130 }, { "grad_norm": 0.5482659935951233, "learning_rate": 9.957967446213308e-05, "loss": 0.0909, "step": 7140 }, { "grad_norm": 0.5634098649024963, "learning_rate": 9.957699589923501e-05, "loss": 0.086, "step": 7150 }, { "grad_norm": 0.9236034154891968, "learning_rate": 9.957430886498431e-05, "loss": 0.0795, "step": 7160 }, { "grad_norm": 0.9238659143447876, "learning_rate": 9.957161335984008e-05, "loss": 0.0784, "step": 7170 }, { "grad_norm": 0.5289351940155029, "learning_rate": 9.956890938426291e-05, "loss": 0.0657, "step": 7180 }, { "grad_norm": 0.4250810444355011, "learning_rate": 9.956619693871482e-05, "loss": 0.0787, "step": 7190 }, { "grad_norm": 0.6990803480148315, "learning_rate": 9.956347602365934e-05, "loss": 0.0805, "step": 7200 }, { "grad_norm": 0.4241727590560913, "learning_rate": 9.956074663956135e-05, "loss": 0.0783, "step": 7210 }, { "grad_norm": 0.6938480734825134, "learning_rate": 9.955800878688726e-05, "loss": 0.0691, "step": 7220 }, { "grad_norm": 0.6294370293617249, "learning_rate": 9.955526246610489e-05, "loss": 0.0777, "step": 7230 }, { "grad_norm": 0.48819082975387573, "learning_rate": 9.955250767768349e-05, "loss": 0.0761, "step": 7240 }, { "grad_norm": 0.44526466727256775, "learning_rate": 9.95497444220938e-05, "loss": 0.0692, "step": 7250 }, { "grad_norm": 0.599856972694397, "learning_rate": 9.954697269980797e-05, "loss": 0.0746, "step": 7260 }, { "grad_norm": 0.5500591397285461, "learning_rate": 9.954419251129962e-05, "loss": 0.0651, "step": 7270 }, { "grad_norm": 0.5073754191398621, "learning_rate": 9.95414038570438e-05, "loss": 0.0654, "step": 7280 }, { "grad_norm": 0.4546101689338684, "learning_rate": 9.953860673751703e-05, "loss": 0.0728, "step": 7290 }, { "grad_norm": 0.6256600618362427, "learning_rate": 9.953580115319725e-05, "loss": 0.0694, "step": 7300 }, { "grad_norm": 0.45563793182373047, "learning_rate": 9.953298710456387e-05, "loss": 0.077, "step": 7310 }, { "grad_norm": 0.5291128158569336, "learning_rate": 9.953016459209771e-05, "loss": 0.0753, "step": 7320 }, { "grad_norm": 0.5305390357971191, "learning_rate": 9.952733361628108e-05, "loss": 0.0652, "step": 7330 }, { "grad_norm": 0.435506671667099, "learning_rate": 9.952449417759772e-05, "loss": 0.0772, "step": 7340 }, { "grad_norm": 0.55458003282547, "learning_rate": 9.952164627653279e-05, "loss": 0.0703, "step": 7350 }, { "grad_norm": 0.5092149972915649, "learning_rate": 9.951878991357292e-05, "loss": 0.0689, "step": 7360 }, { "grad_norm": 0.5053775310516357, "learning_rate": 9.951592508920622e-05, "loss": 0.081, "step": 7370 }, { "grad_norm": 0.5104374885559082, "learning_rate": 9.951305180392219e-05, "loss": 0.0734, "step": 7380 }, { "grad_norm": 0.4405478239059448, "learning_rate": 9.951017005821178e-05, "loss": 0.0715, "step": 7390 }, { "grad_norm": 0.6354638338088989, "learning_rate": 9.95072798525674e-05, "loss": 0.0692, "step": 7400 }, { "grad_norm": 0.5548388361930847, "learning_rate": 9.950438118748293e-05, "loss": 0.0712, "step": 7410 }, { "grad_norm": 0.6351592540740967, "learning_rate": 9.950147406345366e-05, "loss": 0.0842, "step": 7420 }, { "grad_norm": 0.5588976144790649, "learning_rate": 9.949855848097635e-05, "loss": 0.0793, "step": 7430 }, { "grad_norm": 0.535103976726532, "learning_rate": 9.949563444054916e-05, "loss": 0.0836, "step": 7440 }, { "grad_norm": 0.5408717393875122, "learning_rate": 9.949270194267178e-05, "loss": 0.0673, "step": 7450 }, { "grad_norm": 0.6555287837982178, "learning_rate": 9.948976098784526e-05, "loss": 0.0721, "step": 7460 }, { "grad_norm": 0.4179551899433136, "learning_rate": 9.948681157657213e-05, "loss": 0.0781, "step": 7470 }, { "grad_norm": 0.5624316334724426, "learning_rate": 9.948385370935638e-05, "loss": 0.0832, "step": 7480 }, { "grad_norm": 0.5859940648078918, "learning_rate": 9.94808873867034e-05, "loss": 0.0786, "step": 7490 }, { "grad_norm": 0.5966731309890747, "learning_rate": 9.947791260912009e-05, "loss": 0.0737, "step": 7500 }, { "grad_norm": 0.5306152105331421, "learning_rate": 9.947492937711474e-05, "loss": 0.0661, "step": 7510 }, { "grad_norm": 0.506169855594635, "learning_rate": 9.947193769119707e-05, "loss": 0.0759, "step": 7520 }, { "grad_norm": 0.4951288402080536, "learning_rate": 9.946893755187834e-05, "loss": 0.0684, "step": 7530 }, { "grad_norm": 0.4919278919696808, "learning_rate": 9.946592895967115e-05, "loss": 0.073, "step": 7540 }, { "grad_norm": 0.5519096255302429, "learning_rate": 9.94629119150896e-05, "loss": 0.0616, "step": 7550 }, { "grad_norm": 0.48661917448043823, "learning_rate": 9.94598864186492e-05, "loss": 0.0676, "step": 7560 }, { "grad_norm": 0.5061178207397461, "learning_rate": 9.945685247086696e-05, "loss": 0.0795, "step": 7570 }, { "grad_norm": 0.7325006723403931, "learning_rate": 9.945381007226129e-05, "loss": 0.0844, "step": 7580 }, { "grad_norm": 0.45613348484039307, "learning_rate": 9.945075922335203e-05, "loss": 0.0744, "step": 7590 }, { "grad_norm": 0.5130472779273987, "learning_rate": 9.944769992466049e-05, "loss": 0.0915, "step": 7600 }, { "grad_norm": 0.4299246370792389, "learning_rate": 9.944463217670945e-05, "loss": 0.078, "step": 7610 }, { "grad_norm": 0.5011075735092163, "learning_rate": 9.944155598002307e-05, "loss": 0.0714, "step": 7620 }, { "grad_norm": 0.5537618398666382, "learning_rate": 9.943847133512701e-05, "loss": 0.0693, "step": 7630 }, { "grad_norm": 0.7432940006256104, "learning_rate": 9.943537824254834e-05, "loss": 0.0652, "step": 7640 }, { "grad_norm": 0.5480780601501465, "learning_rate": 9.943227670281559e-05, "loss": 0.0648, "step": 7650 }, { "grad_norm": 0.5158580541610718, "learning_rate": 9.942916671645873e-05, "loss": 0.0601, "step": 7660 }, { "grad_norm": 0.6222877502441406, "learning_rate": 9.942604828400916e-05, "loss": 0.0713, "step": 7670 }, { "grad_norm": 0.506312370300293, "learning_rate": 9.942292140599975e-05, "loss": 0.0639, "step": 7680 }, { "grad_norm": 0.6387598514556885, "learning_rate": 9.94197860829648e-05, "loss": 0.0714, "step": 7690 }, { "grad_norm": 0.5257500410079956, "learning_rate": 9.941664231544004e-05, "loss": 0.0592, "step": 7700 }, { "grad_norm": 0.5981201529502869, "learning_rate": 9.941349010396264e-05, "loss": 0.0833, "step": 7710 }, { "grad_norm": 0.5602328777313232, "learning_rate": 9.941032944907125e-05, "loss": 0.0766, "step": 7720 }, { "grad_norm": 0.49900922179222107, "learning_rate": 9.940716035130596e-05, "loss": 0.0719, "step": 7730 }, { "grad_norm": 0.6061708331108093, "learning_rate": 9.940398281120821e-05, "loss": 0.0712, "step": 7740 }, { "grad_norm": 0.46904313564300537, "learning_rate": 9.940079682932102e-05, "loss": 0.066, "step": 7750 }, { "grad_norm": 0.5721049308776855, "learning_rate": 9.939760240618877e-05, "loss": 0.0728, "step": 7760 }, { "grad_norm": 0.47355911135673523, "learning_rate": 9.939439954235729e-05, "loss": 0.0719, "step": 7770 }, { "grad_norm": 0.6511998176574707, "learning_rate": 9.939118823837387e-05, "loss": 0.0757, "step": 7780 }, { "grad_norm": 0.4592154026031494, "learning_rate": 9.938796849478725e-05, "loss": 0.0751, "step": 7790 }, { "grad_norm": 0.6189271807670593, "learning_rate": 9.938474031214755e-05, "loss": 0.0786, "step": 7800 }, { "grad_norm": 0.5747806429862976, "learning_rate": 9.938150369100643e-05, "loss": 0.0737, "step": 7810 }, { "grad_norm": 0.4456845819950104, "learning_rate": 9.93782586319169e-05, "loss": 0.0594, "step": 7820 }, { "grad_norm": 0.5833998322486877, "learning_rate": 9.937500513543348e-05, "loss": 0.0743, "step": 7830 }, { "grad_norm": 0.5224260687828064, "learning_rate": 9.937174320211207e-05, "loss": 0.0674, "step": 7840 }, { "grad_norm": 0.598086416721344, "learning_rate": 9.936847283251009e-05, "loss": 0.0793, "step": 7850 }, { "grad_norm": 0.5767605900764465, "learning_rate": 9.936519402718632e-05, "loss": 0.0733, "step": 7860 }, { "grad_norm": 0.5606869459152222, "learning_rate": 9.936190678670102e-05, "loss": 0.0905, "step": 7870 }, { "grad_norm": 0.5510154366493225, "learning_rate": 9.935861111161593e-05, "loss": 0.0808, "step": 7880 }, { "grad_norm": 0.5116999745368958, "learning_rate": 9.935530700249416e-05, "loss": 0.0764, "step": 7890 }, { "grad_norm": 0.407765656709671, "learning_rate": 9.935199445990028e-05, "loss": 0.0844, "step": 7900 }, { "grad_norm": 0.5020503997802734, "learning_rate": 9.934867348440033e-05, "loss": 0.0861, "step": 7910 }, { "grad_norm": 0.5364698767662048, "learning_rate": 9.934534407656176e-05, "loss": 0.1006, "step": 7920 }, { "grad_norm": 0.5505528450012207, "learning_rate": 9.93420062369535e-05, "loss": 0.073, "step": 7930 }, { "grad_norm": 0.550015389919281, "learning_rate": 9.933865996614589e-05, "loss": 0.0754, "step": 7940 }, { "grad_norm": 0.6690931916236877, "learning_rate": 9.933530526471068e-05, "loss": 0.0641, "step": 7950 }, { "grad_norm": 0.5254227519035339, "learning_rate": 9.933194213322114e-05, "loss": 0.0675, "step": 7960 }, { "grad_norm": 0.5162867307662964, "learning_rate": 9.932857057225192e-05, "loss": 0.0707, "step": 7970 }, { "grad_norm": 0.45454972982406616, "learning_rate": 9.932519058237912e-05, "loss": 0.0636, "step": 7980 }, { "grad_norm": 0.472073495388031, "learning_rate": 9.932180216418032e-05, "loss": 0.0649, "step": 7990 }, { "grad_norm": 0.5097000598907471, "learning_rate": 9.931840531823446e-05, "loss": 0.0585, "step": 8000 }, { "grad_norm": 0.5158175230026245, "learning_rate": 9.9315000045122e-05, "loss": 0.0576, "step": 8010 }, { "grad_norm": 0.5296661257743835, "learning_rate": 9.931158634542481e-05, "loss": 0.0588, "step": 8020 }, { "grad_norm": 0.4362868070602417, "learning_rate": 9.930816421972617e-05, "loss": 0.0578, "step": 8030 }, { "grad_norm": 0.6357753276824951, "learning_rate": 9.930473366861086e-05, "loss": 0.0631, "step": 8040 }, { "grad_norm": 0.5121113061904907, "learning_rate": 9.930129469266505e-05, "loss": 0.0697, "step": 8050 }, { "grad_norm": 0.5755566954612732, "learning_rate": 9.929784729247638e-05, "loss": 0.0635, "step": 8060 }, { "grad_norm": 0.4389675259590149, "learning_rate": 9.929439146863389e-05, "loss": 0.0625, "step": 8070 }, { "grad_norm": 0.6588121056556702, "learning_rate": 9.92909272217281e-05, "loss": 0.0835, "step": 8080 }, { "grad_norm": 0.6547896862030029, "learning_rate": 9.928745455235097e-05, "loss": 0.0822, "step": 8090 }, { "grad_norm": 0.5882904529571533, "learning_rate": 9.928397346109588e-05, "loss": 0.0887, "step": 8100 }, { "grad_norm": 0.6996004581451416, "learning_rate": 9.928048394855762e-05, "loss": 0.0811, "step": 8110 }, { "grad_norm": 0.4715519845485687, "learning_rate": 9.92769860153325e-05, "loss": 0.0737, "step": 8120 }, { "grad_norm": 0.6319652199745178, "learning_rate": 9.927347966201819e-05, "loss": 0.0766, "step": 8130 }, { "grad_norm": 0.5699169635772705, "learning_rate": 9.926996488921383e-05, "loss": 0.0803, "step": 8140 }, { "grad_norm": 0.5293652415275574, "learning_rate": 9.926644169752001e-05, "loss": 0.0693, "step": 8150 }, { "grad_norm": 0.45524469017982483, "learning_rate": 9.926291008753875e-05, "loss": 0.0778, "step": 8160 }, { "grad_norm": 0.5192133784294128, "learning_rate": 9.92593700598735e-05, "loss": 0.0724, "step": 8170 }, { "grad_norm": 0.6422827839851379, "learning_rate": 9.925582161512915e-05, "loss": 0.0762, "step": 8180 }, { "grad_norm": 0.48336178064346313, "learning_rate": 9.925226475391205e-05, "loss": 0.0736, "step": 8190 }, { "grad_norm": 0.5145729780197144, "learning_rate": 9.924869947682993e-05, "loss": 0.0676, "step": 8200 }, { "grad_norm": 0.3619310259819031, "learning_rate": 9.924512578449204e-05, "loss": 0.0538, "step": 8210 }, { "grad_norm": 0.4435158371925354, "learning_rate": 9.924154367750901e-05, "loss": 0.0668, "step": 8220 }, { "grad_norm": 0.5305666327476501, "learning_rate": 9.923795315649293e-05, "loss": 0.0818, "step": 8230 }, { "grad_norm": 0.4674217402935028, "learning_rate": 9.92343542220573e-05, "loss": 0.073, "step": 8240 }, { "grad_norm": 0.6476868987083435, "learning_rate": 9.92307468748171e-05, "loss": 0.0684, "step": 8250 }, { "grad_norm": 0.5272192358970642, "learning_rate": 9.922713111538873e-05, "loss": 0.062, "step": 8260 }, { "grad_norm": 0.43514564633369446, "learning_rate": 9.922350694439003e-05, "loss": 0.0631, "step": 8270 }, { "grad_norm": 0.6361083388328552, "learning_rate": 9.921987436244024e-05, "loss": 0.082, "step": 8280 }, { "grad_norm": 0.4965636730194092, "learning_rate": 9.921623337016008e-05, "loss": 0.0612, "step": 8290 }, { "grad_norm": 0.532420814037323, "learning_rate": 9.921258396817172e-05, "loss": 0.071, "step": 8300 }, { "grad_norm": 0.5163946151733398, "learning_rate": 9.920892615709874e-05, "loss": 0.0674, "step": 8310 }, { "grad_norm": 0.4259324371814728, "learning_rate": 9.920525993756612e-05, "loss": 0.059, "step": 8320 }, { "grad_norm": 0.669985830783844, "learning_rate": 9.920158531020036e-05, "loss": 0.0656, "step": 8330 }, { "grad_norm": 0.5277070999145508, "learning_rate": 9.919790227562933e-05, "loss": 0.0582, "step": 8340 }, { "grad_norm": 0.5451658964157104, "learning_rate": 9.919421083448237e-05, "loss": 0.0789, "step": 8350 }, { "grad_norm": 0.531855046749115, "learning_rate": 9.919051098739022e-05, "loss": 0.0605, "step": 8360 }, { "grad_norm": 0.45366519689559937, "learning_rate": 9.918680273498514e-05, "loss": 0.0596, "step": 8370 }, { "grad_norm": 0.49204006791114807, "learning_rate": 9.918308607790072e-05, "loss": 0.0685, "step": 8380 }, { "grad_norm": 0.5689148306846619, "learning_rate": 9.917936101677205e-05, "loss": 0.0556, "step": 8390 }, { "grad_norm": 0.6042755842208862, "learning_rate": 9.917562755223564e-05, "loss": 0.0635, "step": 8400 }, { "grad_norm": 0.6075617074966431, "learning_rate": 9.917188568492944e-05, "loss": 0.1083, "step": 8410 }, { "grad_norm": 0.4475400149822235, "learning_rate": 9.916813541549283e-05, "loss": 0.0609, "step": 8420 }, { "grad_norm": 0.6060808300971985, "learning_rate": 9.916437674456663e-05, "loss": 0.0634, "step": 8430 }, { "grad_norm": 0.5422413945198059, "learning_rate": 9.916060967279308e-05, "loss": 0.0645, "step": 8440 }, { "grad_norm": 0.6826819181442261, "learning_rate": 9.91568342008159e-05, "loss": 0.0635, "step": 8450 }, { "grad_norm": 0.45829281210899353, "learning_rate": 9.915305032928019e-05, "loss": 0.0633, "step": 8460 }, { "grad_norm": 0.4433273673057556, "learning_rate": 9.914925805883253e-05, "loss": 0.0631, "step": 8470 }, { "grad_norm": 0.467151403427124, "learning_rate": 9.914545739012088e-05, "loss": 0.0572, "step": 8480 }, { "grad_norm": 0.6252233982086182, "learning_rate": 9.91416483237947e-05, "loss": 0.0594, "step": 8490 }, { "grad_norm": 0.6408552527427673, "learning_rate": 9.913783086050485e-05, "loss": 0.0666, "step": 8500 }, { "grad_norm": 0.47359633445739746, "learning_rate": 9.913400500090364e-05, "loss": 0.062, "step": 8510 }, { "grad_norm": 0.5021724104881287, "learning_rate": 9.913017074564479e-05, "loss": 0.0523, "step": 8520 }, { "grad_norm": 0.48141855001449585, "learning_rate": 9.912632809538348e-05, "loss": 0.0689, "step": 8530 }, { "grad_norm": 0.4852023720741272, "learning_rate": 9.912247705077629e-05, "loss": 0.0693, "step": 8540 }, { "grad_norm": 0.7192977666854858, "learning_rate": 9.911861761248127e-05, "loss": 0.068, "step": 8550 }, { "grad_norm": 0.6276741027832031, "learning_rate": 9.91147497811579e-05, "loss": 0.0687, "step": 8560 }, { "grad_norm": 0.6135303974151611, "learning_rate": 9.911087355746709e-05, "loss": 0.098, "step": 8570 }, { "grad_norm": 0.4900945723056793, "learning_rate": 9.910698894207117e-05, "loss": 0.0826, "step": 8580 }, { "grad_norm": 0.5828286409378052, "learning_rate": 9.910309593563392e-05, "loss": 0.0755, "step": 8590 }, { "grad_norm": 0.32525554299354553, "learning_rate": 9.909919453882057e-05, "loss": 0.0551, "step": 8600 }, { "grad_norm": 0.6967249512672424, "learning_rate": 9.90952847522977e-05, "loss": 0.061, "step": 8610 }, { "grad_norm": 0.5707406401634216, "learning_rate": 9.909136657673346e-05, "loss": 0.0824, "step": 8620 }, { "grad_norm": 0.5742839574813843, "learning_rate": 9.908744001279731e-05, "loss": 0.0656, "step": 8630 }, { "grad_norm": 0.6645046472549438, "learning_rate": 9.90835050611602e-05, "loss": 0.0825, "step": 8640 }, { "grad_norm": 0.5503866076469421, "learning_rate": 9.90795617224945e-05, "loss": 0.0784, "step": 8650 }, { "grad_norm": 0.649146318435669, "learning_rate": 9.907560999747405e-05, "loss": 0.0677, "step": 8660 }, { "grad_norm": 0.4679626524448395, "learning_rate": 9.907164988677408e-05, "loss": 0.0587, "step": 8670 }, { "grad_norm": 0.5627082586288452, "learning_rate": 9.906768139107124e-05, "loss": 0.0608, "step": 8680 }, { "grad_norm": 0.49064362049102783, "learning_rate": 9.906370451104367e-05, "loss": 0.0608, "step": 8690 }, { "grad_norm": 0.43124842643737793, "learning_rate": 9.905971924737088e-05, "loss": 0.0646, "step": 8700 }, { "grad_norm": 0.4598067104816437, "learning_rate": 9.905572560073387e-05, "loss": 0.0731, "step": 8710 }, { "grad_norm": 0.8145584464073181, "learning_rate": 9.905172357181501e-05, "loss": 0.0898, "step": 8720 }, { "grad_norm": 0.5740137696266174, "learning_rate": 9.904771316129817e-05, "loss": 0.0684, "step": 8730 }, { "grad_norm": 0.55716872215271, "learning_rate": 9.904369436986862e-05, "loss": 0.0665, "step": 8740 }, { "grad_norm": 0.5659870505332947, "learning_rate": 9.903966719821303e-05, "loss": 0.0739, "step": 8750 }, { "grad_norm": 0.564585268497467, "learning_rate": 9.903563164701956e-05, "loss": 0.0673, "step": 8760 }, { "grad_norm": 0.4506534934043884, "learning_rate": 9.903158771697778e-05, "loss": 0.0681, "step": 8770 }, { "grad_norm": 0.5015162825584412, "learning_rate": 9.902753540877867e-05, "loss": 0.0647, "step": 8780 }, { "grad_norm": 0.5021986365318298, "learning_rate": 9.902347472311466e-05, "loss": 0.0557, "step": 8790 }, { "grad_norm": 0.5234400033950806, "learning_rate": 9.901940566067962e-05, "loss": 0.0499, "step": 8800 }, { "grad_norm": 0.3827537000179291, "learning_rate": 9.901532822216883e-05, "loss": 0.057, "step": 8810 }, { "grad_norm": 0.549988865852356, "learning_rate": 9.901124240827904e-05, "loss": 0.0647, "step": 8820 }, { "grad_norm": 0.5905101299285889, "learning_rate": 9.900714821970835e-05, "loss": 0.0568, "step": 8830 }, { "grad_norm": 0.7033087611198425, "learning_rate": 9.900304565715641e-05, "loss": 0.0622, "step": 8840 }, { "grad_norm": 0.4158196747303009, "learning_rate": 9.899893472132419e-05, "loss": 0.0614, "step": 8850 }, { "grad_norm": 0.5366539359092712, "learning_rate": 9.899481541291415e-05, "loss": 0.0607, "step": 8860 }, { "grad_norm": 0.7353037595748901, "learning_rate": 9.899068773263016e-05, "loss": 0.0666, "step": 8870 }, { "grad_norm": 0.46172550320625305, "learning_rate": 9.898655168117754e-05, "loss": 0.0691, "step": 8880 }, { "grad_norm": 0.4560208022594452, "learning_rate": 9.898240725926302e-05, "loss": 0.0651, "step": 8890 }, { "grad_norm": 0.4526781737804413, "learning_rate": 9.897825446759478e-05, "loss": 0.0609, "step": 8900 }, { "grad_norm": 0.480383962392807, "learning_rate": 9.897409330688241e-05, "loss": 0.0621, "step": 8910 }, { "grad_norm": 0.6904319524765015, "learning_rate": 9.896992377783692e-05, "loss": 0.0819, "step": 8920 }, { "grad_norm": 0.5892252922058105, "learning_rate": 9.89657458811708e-05, "loss": 0.0568, "step": 8930 }, { "grad_norm": 0.43852612376213074, "learning_rate": 9.896155961759792e-05, "loss": 0.0645, "step": 8940 }, { "grad_norm": 0.42871302366256714, "learning_rate": 9.895736498783361e-05, "loss": 0.0651, "step": 8950 }, { "grad_norm": 0.5306944847106934, "learning_rate": 9.895316199259462e-05, "loss": 0.053, "step": 8960 }, { "grad_norm": 0.5118286609649658, "learning_rate": 9.894895063259909e-05, "loss": 0.0492, "step": 8970 }, { "grad_norm": 0.4627131521701813, "learning_rate": 9.894473090856667e-05, "loss": 0.0582, "step": 8980 }, { "grad_norm": 0.5666720271110535, "learning_rate": 9.894050282121839e-05, "loss": 0.0874, "step": 8990 }, { "grad_norm": 0.5638395547866821, "learning_rate": 9.893626637127668e-05, "loss": 0.0619, "step": 9000 }, { "grad_norm": 0.4603935480117798, "learning_rate": 9.893202155946546e-05, "loss": 0.0511, "step": 9010 }, { "grad_norm": 0.6210193634033203, "learning_rate": 9.892776838651006e-05, "loss": 0.0579, "step": 9020 }, { "grad_norm": 0.5235388875007629, "learning_rate": 9.892350685313722e-05, "loss": 0.0736, "step": 9030 }, { "grad_norm": 0.7283201217651367, "learning_rate": 9.891923696007513e-05, "loss": 0.0733, "step": 9040 }, { "grad_norm": 0.6449484825134277, "learning_rate": 9.891495870805336e-05, "loss": 0.0635, "step": 9050 }, { "grad_norm": 0.44755467772483826, "learning_rate": 9.891067209780298e-05, "loss": 0.0748, "step": 9060 }, { "grad_norm": 0.5156524181365967, "learning_rate": 9.890637713005646e-05, "loss": 0.0566, "step": 9070 }, { "grad_norm": 0.5168870687484741, "learning_rate": 9.890207380554767e-05, "loss": 0.0665, "step": 9080 }, { "grad_norm": 0.5812755823135376, "learning_rate": 9.889776212501196e-05, "loss": 0.0742, "step": 9090 }, { "grad_norm": 0.5427368879318237, "learning_rate": 9.889344208918605e-05, "loss": 0.0826, "step": 9100 }, { "grad_norm": 0.6685421466827393, "learning_rate": 9.888911369880812e-05, "loss": 0.0786, "step": 9110 }, { "grad_norm": 0.4284568130970001, "learning_rate": 9.888477695461777e-05, "loss": 0.0682, "step": 9120 }, { "grad_norm": 0.5356078147888184, "learning_rate": 9.888043185735607e-05, "loss": 0.0756, "step": 9130 }, { "grad_norm": 0.4701652228832245, "learning_rate": 9.887607840776542e-05, "loss": 0.0607, "step": 9140 }, { "grad_norm": 0.5195785164833069, "learning_rate": 9.887171660658975e-05, "loss": 0.0719, "step": 9150 }, { "grad_norm": 0.39833515882492065, "learning_rate": 9.886734645457435e-05, "loss": 0.0545, "step": 9160 }, { "grad_norm": 0.5868363976478577, "learning_rate": 9.886296795246597e-05, "loss": 0.0597, "step": 9170 }, { "grad_norm": 0.43165886402130127, "learning_rate": 9.885858110101276e-05, "loss": 0.061, "step": 9180 }, { "grad_norm": 0.3600703477859497, "learning_rate": 9.885418590096434e-05, "loss": 0.0508, "step": 9190 }, { "grad_norm": 0.455751895904541, "learning_rate": 9.88497823530717e-05, "loss": 0.0634, "step": 9200 }, { "grad_norm": 0.5404902100563049, "learning_rate": 9.884537045808732e-05, "loss": 0.0824, "step": 9210 }, { "grad_norm": 0.5968615412712097, "learning_rate": 9.884095021676502e-05, "loss": 0.0804, "step": 9220 }, { "grad_norm": 0.6244338154792786, "learning_rate": 9.883652162986017e-05, "loss": 0.0682, "step": 9230 }, { "grad_norm": 0.3728165924549103, "learning_rate": 9.883208469812943e-05, "loss": 0.0748, "step": 9240 }, { "grad_norm": 0.4387248158454895, "learning_rate": 9.882763942233098e-05, "loss": 0.0576, "step": 9250 }, { "grad_norm": 0.5232730507850647, "learning_rate": 9.882318580322441e-05, "loss": 0.0623, "step": 9260 }, { "grad_norm": 0.47157734632492065, "learning_rate": 9.881872384157067e-05, "loss": 0.0642, "step": 9270 }, { "grad_norm": 0.5017635822296143, "learning_rate": 9.881425353813225e-05, "loss": 0.064, "step": 9280 }, { "grad_norm": 0.6261676549911499, "learning_rate": 9.880977489367296e-05, "loss": 0.0771, "step": 9290 }, { "grad_norm": 0.44527384638786316, "learning_rate": 9.88052879089581e-05, "loss": 0.0741, "step": 9300 }, { "grad_norm": 0.4212121069431305, "learning_rate": 9.880079258475434e-05, "loss": 0.0496, "step": 9310 }, { "grad_norm": 0.478679895401001, "learning_rate": 9.879628892182985e-05, "loss": 0.0772, "step": 9320 }, { "grad_norm": 0.4488307237625122, "learning_rate": 9.879177692095416e-05, "loss": 0.0618, "step": 9330 }, { "grad_norm": 0.5632327198982239, "learning_rate": 9.878725658289825e-05, "loss": 0.0797, "step": 9340 }, { "grad_norm": 0.45970892906188965, "learning_rate": 9.878272790843454e-05, "loss": 0.0588, "step": 9350 }, { "grad_norm": 0.4956205189228058, "learning_rate": 9.877819089833682e-05, "loss": 0.0554, "step": 9360 }, { "grad_norm": 0.5158644318580627, "learning_rate": 9.877364555338038e-05, "loss": 0.0749, "step": 9370 }, { "grad_norm": 0.40306100249290466, "learning_rate": 9.876909187434186e-05, "loss": 0.0714, "step": 9380 }, { "grad_norm": 0.5846615433692932, "learning_rate": 9.876452986199939e-05, "loss": 0.0668, "step": 9390 }, { "grad_norm": 0.5431602597236633, "learning_rate": 9.875995951713248e-05, "loss": 0.0757, "step": 9400 }, { "grad_norm": 0.602813184261322, "learning_rate": 9.875538084052207e-05, "loss": 0.0722, "step": 9410 }, { "grad_norm": 0.6492631435394287, "learning_rate": 9.875079383295053e-05, "loss": 0.0697, "step": 9420 }, { "grad_norm": 0.43143174052238464, "learning_rate": 9.874619849520167e-05, "loss": 0.0549, "step": 9430 }, { "grad_norm": 0.5224357843399048, "learning_rate": 9.874159482806069e-05, "loss": 0.0787, "step": 9440 }, { "grad_norm": 0.5983032584190369, "learning_rate": 9.873698283231426e-05, "loss": 0.0697, "step": 9450 }, { "grad_norm": 0.6210148334503174, "learning_rate": 9.87323625087504e-05, "loss": 0.0596, "step": 9460 }, { "grad_norm": 0.5956394672393799, "learning_rate": 9.872773385815863e-05, "loss": 0.0806, "step": 9470 }, { "grad_norm": 0.5593950152397156, "learning_rate": 9.872309688132986e-05, "loss": 0.0625, "step": 9480 }, { "grad_norm": 0.47497713565826416, "learning_rate": 9.871845157905639e-05, "loss": 0.0711, "step": 9490 }, { "grad_norm": 0.46989524364471436, "learning_rate": 9.871379795213201e-05, "loss": 0.0611, "step": 9500 }, { "grad_norm": 0.485297828912735, "learning_rate": 9.87091360013519e-05, "loss": 0.0754, "step": 9510 }, { "grad_norm": 0.4318355619907379, "learning_rate": 9.870446572751262e-05, "loss": 0.0668, "step": 9520 }, { "grad_norm": 0.41083136200904846, "learning_rate": 9.869978713141224e-05, "loss": 0.0632, "step": 9530 }, { "grad_norm": 0.6720036864280701, "learning_rate": 9.869510021385016e-05, "loss": 0.0744, "step": 9540 }, { "grad_norm": 0.4488694965839386, "learning_rate": 9.869040497562727e-05, "loss": 0.0635, "step": 9550 }, { "grad_norm": 0.6547280550003052, "learning_rate": 9.868570141754587e-05, "loss": 0.0785, "step": 9560 }, { "grad_norm": 0.6457273960113525, "learning_rate": 9.868098954040965e-05, "loss": 0.0857, "step": 9570 }, { "grad_norm": 0.546299934387207, "learning_rate": 9.867626934502374e-05, "loss": 0.0719, "step": 9580 }, { "grad_norm": 1.6365348100662231, "learning_rate": 9.86715408321947e-05, "loss": 0.0757, "step": 9590 }, { "grad_norm": 0.5569512844085693, "learning_rate": 9.86668040027305e-05, "loss": 0.0673, "step": 9600 }, { "grad_norm": 0.5079053640365601, "learning_rate": 9.866205885744053e-05, "loss": 0.07, "step": 9610 }, { "grad_norm": 0.5834388732910156, "learning_rate": 9.865730539713563e-05, "loss": 0.0815, "step": 9620 }, { "grad_norm": 0.4688837230205536, "learning_rate": 9.8652543622628e-05, "loss": 0.0683, "step": 9630 }, { "grad_norm": 0.4150487780570984, "learning_rate": 9.864777353473132e-05, "loss": 0.0629, "step": 9640 }, { "grad_norm": 0.3883180320262909, "learning_rate": 9.864299513426068e-05, "loss": 0.0574, "step": 9650 }, { "grad_norm": 0.5143182277679443, "learning_rate": 9.863820842203254e-05, "loss": 0.0627, "step": 9660 }, { "grad_norm": 0.5366613268852234, "learning_rate": 9.863341339886483e-05, "loss": 0.0682, "step": 9670 }, { "grad_norm": 0.4339425563812256, "learning_rate": 9.86286100655769e-05, "loss": 0.0623, "step": 9680 }, { "grad_norm": 0.6532087922096252, "learning_rate": 9.862379842298953e-05, "loss": 0.0682, "step": 9690 }, { "grad_norm": 0.5368595719337463, "learning_rate": 9.861897847192485e-05, "loss": 0.055, "step": 9700 }, { "grad_norm": 0.48057931661605835, "learning_rate": 9.86141502132065e-05, "loss": 0.0767, "step": 9710 }, { "grad_norm": 0.4927929937839508, "learning_rate": 9.860931364765946e-05, "loss": 0.0598, "step": 9720 }, { "grad_norm": 0.6723587512969971, "learning_rate": 9.860446877611021e-05, "loss": 0.0711, "step": 9730 }, { "grad_norm": 0.4992118179798126, "learning_rate": 9.859961559938655e-05, "loss": 0.0726, "step": 9740 }, { "grad_norm": 0.659618616104126, "learning_rate": 9.85947541183178e-05, "loss": 0.0679, "step": 9750 }, { "grad_norm": 0.6489319801330566, "learning_rate": 9.858988433373463e-05, "loss": 0.0722, "step": 9760 }, { "grad_norm": 0.41698601841926575, "learning_rate": 9.858500624646918e-05, "loss": 0.065, "step": 9770 }, { "grad_norm": 0.5095018148422241, "learning_rate": 9.858011985735497e-05, "loss": 0.0666, "step": 9780 }, { "grad_norm": 0.6848360300064087, "learning_rate": 9.857522516722693e-05, "loss": 0.0659, "step": 9790 }, { "grad_norm": 0.4398139417171478, "learning_rate": 9.857032217692145e-05, "loss": 0.0544, "step": 9800 }, { "grad_norm": 0.8461122512817383, "learning_rate": 9.856541088727631e-05, "loss": 0.0696, "step": 9810 }, { "grad_norm": 0.3973804712295532, "learning_rate": 9.856049129913072e-05, "loss": 0.0747, "step": 9820 }, { "grad_norm": 0.538549542427063, "learning_rate": 9.85555634133253e-05, "loss": 0.0716, "step": 9830 }, { "grad_norm": 0.42247968912124634, "learning_rate": 9.855062723070208e-05, "loss": 0.0729, "step": 9840 }, { "grad_norm": 0.40338048338890076, "learning_rate": 9.854568275210454e-05, "loss": 0.0763, "step": 9850 }, { "grad_norm": 0.4112570881843567, "learning_rate": 9.854072997837754e-05, "loss": 0.0562, "step": 9860 }, { "grad_norm": 0.4816242754459381, "learning_rate": 9.853576891036737e-05, "loss": 0.0649, "step": 9870 }, { "grad_norm": 0.7952947616577148, "learning_rate": 9.853079954892177e-05, "loss": 0.0755, "step": 9880 }, { "grad_norm": 0.4997425973415375, "learning_rate": 9.852582189488983e-05, "loss": 0.0615, "step": 9890 }, { "grad_norm": 0.512944221496582, "learning_rate": 9.852083594912212e-05, "loss": 0.0763, "step": 9900 }, { "grad_norm": 0.5388298034667969, "learning_rate": 9.851584171247058e-05, "loss": 0.0691, "step": 9910 }, { "grad_norm": 0.5173930525779724, "learning_rate": 9.851083918578863e-05, "loss": 0.0748, "step": 9920 }, { "grad_norm": 0.6089332699775696, "learning_rate": 9.850582836993103e-05, "loss": 0.0655, "step": 9930 }, { "grad_norm": 0.5722701549530029, "learning_rate": 9.850080926575397e-05, "loss": 0.0687, "step": 9940 }, { "grad_norm": 0.41971322894096375, "learning_rate": 9.849578187411515e-05, "loss": 0.0607, "step": 9950 }, { "grad_norm": 0.5092242956161499, "learning_rate": 9.849074619587354e-05, "loss": 0.0568, "step": 9960 }, { "grad_norm": 0.5285491347312927, "learning_rate": 9.848570223188964e-05, "loss": 0.0651, "step": 9970 }, { "grad_norm": 0.5079439282417297, "learning_rate": 9.848064998302531e-05, "loss": 0.0643, "step": 9980 }, { "grad_norm": 0.6742474436759949, "learning_rate": 9.847558945014386e-05, "loss": 0.0663, "step": 9990 }, { "grad_norm": 0.6656249165534973, "learning_rate": 9.847052063410996e-05, "loss": 0.0709, "step": 10000 }, { "grad_norm": 0.4007723331451416, "learning_rate": 9.846544353578977e-05, "loss": 0.0682, "step": 10010 }, { "grad_norm": 0.39746415615081787, "learning_rate": 9.846035815605081e-05, "loss": 0.0715, "step": 10020 }, { "grad_norm": 0.48376503586769104, "learning_rate": 9.845526449576204e-05, "loss": 0.0605, "step": 10030 }, { "grad_norm": 0.5282701849937439, "learning_rate": 9.845016255579383e-05, "loss": 0.0728, "step": 10040 }, { "grad_norm": 0.6024710536003113, "learning_rate": 9.844505233701794e-05, "loss": 0.0714, "step": 10050 }, { "grad_norm": 0.6086927056312561, "learning_rate": 9.843993384030757e-05, "loss": 0.0665, "step": 10060 }, { "grad_norm": 0.36891239881515503, "learning_rate": 9.843480706653737e-05, "loss": 0.05, "step": 10070 }, { "grad_norm": 0.43691617250442505, "learning_rate": 9.84296720165833e-05, "loss": 0.0673, "step": 10080 }, { "grad_norm": 0.604579508304596, "learning_rate": 9.842452869132286e-05, "loss": 0.0561, "step": 10090 }, { "grad_norm": 0.5216884016990662, "learning_rate": 9.841937709163489e-05, "loss": 0.0611, "step": 10100 }, { "grad_norm": 0.47738000750541687, "learning_rate": 9.841421721839962e-05, "loss": 0.0534, "step": 10110 }, { "grad_norm": 0.5832960605621338, "learning_rate": 9.840904907249879e-05, "loss": 0.0494, "step": 10120 }, { "grad_norm": 0.5370098948478699, "learning_rate": 9.840387265481545e-05, "loss": 0.0701, "step": 10130 }, { "grad_norm": 0.5291993618011475, "learning_rate": 9.839868796623411e-05, "loss": 0.0553, "step": 10140 }, { "grad_norm": 0.5148180723190308, "learning_rate": 9.839349500764072e-05, "loss": 0.0762, "step": 10150 }, { "grad_norm": 0.5747126936912537, "learning_rate": 9.83882937799226e-05, "loss": 0.0692, "step": 10160 }, { "grad_norm": 0.6048736572265625, "learning_rate": 9.838308428396849e-05, "loss": 0.0655, "step": 10170 }, { "grad_norm": 0.36611121892929077, "learning_rate": 9.837786652066854e-05, "loss": 0.0782, "step": 10180 }, { "grad_norm": 0.5866525769233704, "learning_rate": 9.837264049091437e-05, "loss": 0.0863, "step": 10190 }, { "grad_norm": 0.4636238217353821, "learning_rate": 9.836740619559893e-05, "loss": 0.067, "step": 10200 }, { "grad_norm": 0.48046064376831055, "learning_rate": 9.836216363561659e-05, "loss": 0.0593, "step": 10210 }, { "grad_norm": 0.5313364267349243, "learning_rate": 9.835691281186322e-05, "loss": 0.0708, "step": 10220 }, { "grad_norm": 0.6084886193275452, "learning_rate": 9.8351653725236e-05, "loss": 0.0684, "step": 10230 }, { "grad_norm": 0.5905547738075256, "learning_rate": 9.83463863766336e-05, "loss": 0.0636, "step": 10240 }, { "grad_norm": 0.694668173789978, "learning_rate": 9.834111076695602e-05, "loss": 0.0572, "step": 10250 }, { "grad_norm": 0.5030235648155212, "learning_rate": 9.833582689710477e-05, "loss": 0.0511, "step": 10260 }, { "grad_norm": 0.4572940170764923, "learning_rate": 9.833053476798268e-05, "loss": 0.0608, "step": 10270 }, { "grad_norm": 0.48276621103286743, "learning_rate": 9.832523438049404e-05, "loss": 0.0686, "step": 10280 }, { "grad_norm": 0.4988575279712677, "learning_rate": 9.831992573554454e-05, "loss": 0.0683, "step": 10290 }, { "grad_norm": 0.5633752346038818, "learning_rate": 9.831460883404128e-05, "loss": 0.0687, "step": 10300 }, { "grad_norm": 0.5413036346435547, "learning_rate": 9.830928367689278e-05, "loss": 0.0579, "step": 10310 }, { "grad_norm": 0.4716736078262329, "learning_rate": 9.830395026500896e-05, "loss": 0.0618, "step": 10320 }, { "grad_norm": 0.511012852191925, "learning_rate": 9.829860859930115e-05, "loss": 0.0596, "step": 10330 }, { "grad_norm": 0.5742474794387817, "learning_rate": 9.829325868068212e-05, "loss": 0.0647, "step": 10340 }, { "grad_norm": 0.4759950041770935, "learning_rate": 9.8287900510066e-05, "loss": 0.0692, "step": 10350 }, { "grad_norm": 0.35963648557662964, "learning_rate": 9.828253408836834e-05, "loss": 0.0528, "step": 10360 }, { "grad_norm": 0.5144962668418884, "learning_rate": 9.827715941650615e-05, "loss": 0.0547, "step": 10370 }, { "grad_norm": 0.5111762881278992, "learning_rate": 9.82717764953978e-05, "loss": 0.0558, "step": 10380 }, { "grad_norm": 0.4878023862838745, "learning_rate": 9.826638532596308e-05, "loss": 0.0563, "step": 10390 }, { "grad_norm": 0.37002918124198914, "learning_rate": 9.82609859091232e-05, "loss": 0.0606, "step": 10400 }, { "grad_norm": 0.44112157821655273, "learning_rate": 9.825557824580076e-05, "loss": 0.0561, "step": 10410 }, { "grad_norm": 0.5810962915420532, "learning_rate": 9.82501623369198e-05, "loss": 0.0586, "step": 10420 }, { "grad_norm": 0.43213796615600586, "learning_rate": 9.824473818340574e-05, "loss": 0.0561, "step": 10430 }, { "grad_norm": 0.42079344391822815, "learning_rate": 9.823930578618541e-05, "loss": 0.0474, "step": 10440 }, { "grad_norm": 0.625201404094696, "learning_rate": 9.823386514618709e-05, "loss": 0.0648, "step": 10450 }, { "grad_norm": 0.4418233335018158, "learning_rate": 9.82284162643404e-05, "loss": 0.0594, "step": 10460 }, { "grad_norm": 0.3886982500553131, "learning_rate": 9.822295914157642e-05, "loss": 0.0566, "step": 10470 }, { "grad_norm": 0.5384728908538818, "learning_rate": 9.821749377882763e-05, "loss": 0.0523, "step": 10480 }, { "grad_norm": 0.3660270571708679, "learning_rate": 9.821202017702791e-05, "loss": 0.052, "step": 10490 }, { "grad_norm": 0.4466174840927124, "learning_rate": 9.820653833711253e-05, "loss": 0.0526, "step": 10500 }, { "grad_norm": 0.49214887619018555, "learning_rate": 9.820104826001822e-05, "loss": 0.064, "step": 10510 }, { "grad_norm": 0.43746131658554077, "learning_rate": 9.819554994668305e-05, "loss": 0.0479, "step": 10520 }, { "grad_norm": 0.4511483907699585, "learning_rate": 9.819004339804654e-05, "loss": 0.0673, "step": 10530 }, { "grad_norm": 0.5050042867660522, "learning_rate": 9.818452861504961e-05, "loss": 0.0531, "step": 10540 }, { "grad_norm": 0.3938261568546295, "learning_rate": 9.81790055986346e-05, "loss": 0.0475, "step": 10550 }, { "grad_norm": 0.4987563192844391, "learning_rate": 9.817347434974523e-05, "loss": 0.0473, "step": 10560 }, { "grad_norm": 0.46224677562713623, "learning_rate": 9.816793486932664e-05, "loss": 0.0455, "step": 10570 }, { "grad_norm": 0.41080355644226074, "learning_rate": 9.816238715832538e-05, "loss": 0.0545, "step": 10580 }, { "grad_norm": 0.49587270617485046, "learning_rate": 9.815683121768939e-05, "loss": 0.0649, "step": 10590 }, { "grad_norm": 0.7100891470909119, "learning_rate": 9.815126704836804e-05, "loss": 0.0681, "step": 10600 }, { "grad_norm": 0.6847078204154968, "learning_rate": 9.81456946513121e-05, "loss": 0.0903, "step": 10610 }, { "grad_norm": 0.6598333716392517, "learning_rate": 9.814011402747373e-05, "loss": 0.0821, "step": 10620 }, { "grad_norm": 0.49999743700027466, "learning_rate": 9.813452517780651e-05, "loss": 0.0549, "step": 10630 }, { "grad_norm": 0.4349750280380249, "learning_rate": 9.81289281032654e-05, "loss": 0.0637, "step": 10640 }, { "grad_norm": 0.3975478708744049, "learning_rate": 9.812332280480683e-05, "loss": 0.0648, "step": 10650 }, { "grad_norm": 0.44644632935523987, "learning_rate": 9.811770928338854e-05, "loss": 0.0627, "step": 10660 }, { "grad_norm": 0.6306639909744263, "learning_rate": 9.811208753996979e-05, "loss": 0.0639, "step": 10670 }, { "grad_norm": 0.5384917259216309, "learning_rate": 9.810645757551113e-05, "loss": 0.0487, "step": 10680 }, { "grad_norm": 0.3813328742980957, "learning_rate": 9.810081939097459e-05, "loss": 0.0537, "step": 10690 }, { "grad_norm": 0.40530920028686523, "learning_rate": 9.809517298732356e-05, "loss": 0.074, "step": 10700 }, { "grad_norm": 0.5995267033576965, "learning_rate": 9.80895183655229e-05, "loss": 0.0623, "step": 10710 }, { "grad_norm": 0.5230422616004944, "learning_rate": 9.808385552653877e-05, "loss": 0.0546, "step": 10720 }, { "grad_norm": 0.42116984724998474, "learning_rate": 9.807818447133886e-05, "loss": 0.0496, "step": 10730 }, { "grad_norm": 0.5290555953979492, "learning_rate": 9.807250520089215e-05, "loss": 0.0538, "step": 10740 }, { "grad_norm": 0.3902396857738495, "learning_rate": 9.806681771616908e-05, "loss": 0.0655, "step": 10750 }, { "grad_norm": 0.4734194874763489, "learning_rate": 9.80611220181415e-05, "loss": 0.0643, "step": 10760 }, { "grad_norm": 0.6981947422027588, "learning_rate": 9.805541810778264e-05, "loss": 0.0692, "step": 10770 }, { "grad_norm": 0.5870326161384583, "learning_rate": 9.804970598606716e-05, "loss": 0.0665, "step": 10780 }, { "grad_norm": 0.44626057147979736, "learning_rate": 9.804398565397106e-05, "loss": 0.062, "step": 10790 }, { "grad_norm": 0.5635755062103271, "learning_rate": 9.803825711247183e-05, "loss": 0.0668, "step": 10800 }, { "grad_norm": 0.5859290361404419, "learning_rate": 9.803252036254831e-05, "loss": 0.0573, "step": 10810 }, { "grad_norm": 0.4864128828048706, "learning_rate": 9.802677540518076e-05, "loss": 0.0626, "step": 10820 }, { "grad_norm": 0.3841102123260498, "learning_rate": 9.802102224135081e-05, "loss": 0.056, "step": 10830 }, { "grad_norm": 0.4333098530769348, "learning_rate": 9.801526087204155e-05, "loss": 0.0716, "step": 10840 }, { "grad_norm": 0.5647684335708618, "learning_rate": 9.800949129823743e-05, "loss": 0.0576, "step": 10850 }, { "grad_norm": 0.5409406423568726, "learning_rate": 9.80037135209243e-05, "loss": 0.0483, "step": 10860 }, { "grad_norm": 0.6334624290466309, "learning_rate": 9.799792754108946e-05, "loss": 0.0587, "step": 10870 }, { "grad_norm": 0.5479588508605957, "learning_rate": 9.799213335972152e-05, "loss": 0.0651, "step": 10880 }, { "grad_norm": 0.46667200326919556, "learning_rate": 9.798633097781058e-05, "loss": 0.06, "step": 10890 }, { "grad_norm": 0.42816004157066345, "learning_rate": 9.79805203963481e-05, "loss": 0.0632, "step": 10900 }, { "grad_norm": 0.4086267650127411, "learning_rate": 9.797470161632697e-05, "loss": 0.0585, "step": 10910 }, { "grad_norm": 0.646603524684906, "learning_rate": 9.796887463874145e-05, "loss": 0.06, "step": 10920 }, { "grad_norm": 0.3319067656993866, "learning_rate": 9.796303946458718e-05, "loss": 0.0767, "step": 10930 }, { "grad_norm": 0.47041353583335876, "learning_rate": 9.795719609486127e-05, "loss": 0.069, "step": 10940 }, { "grad_norm": 0.5637128949165344, "learning_rate": 9.795134453056219e-05, "loss": 0.0779, "step": 10950 }, { "grad_norm": 0.41281992197036743, "learning_rate": 9.794548477268979e-05, "loss": 0.0526, "step": 10960 }, { "grad_norm": 0.4399227201938629, "learning_rate": 9.793961682224537e-05, "loss": 0.0604, "step": 10970 }, { "grad_norm": 0.5612607598304749, "learning_rate": 9.793374068023156e-05, "loss": 0.0624, "step": 10980 }, { "grad_norm": 0.4864497780799866, "learning_rate": 9.792785634765247e-05, "loss": 0.0577, "step": 10990 }, { "grad_norm": 0.45553144812583923, "learning_rate": 9.792196382551357e-05, "loss": 0.0566, "step": 11000 }, { "grad_norm": 0.5894051194190979, "learning_rate": 9.791606311482171e-05, "loss": 0.0623, "step": 11010 }, { "grad_norm": 0.4050065875053406, "learning_rate": 9.791015421658518e-05, "loss": 0.0553, "step": 11020 }, { "grad_norm": 0.6117287278175354, "learning_rate": 9.790423713181362e-05, "loss": 0.0661, "step": 11030 }, { "grad_norm": 0.5131316781044006, "learning_rate": 9.789831186151814e-05, "loss": 0.0596, "step": 11040 }, { "grad_norm": 0.7258656620979309, "learning_rate": 9.789237840671118e-05, "loss": 0.0583, "step": 11050 }, { "grad_norm": 0.5091056823730469, "learning_rate": 9.78864367684066e-05, "loss": 0.0671, "step": 11060 }, { "grad_norm": 0.5296409130096436, "learning_rate": 9.788048694761968e-05, "loss": 0.0535, "step": 11070 }, { "grad_norm": 0.7164278626441956, "learning_rate": 9.787452894536709e-05, "loss": 0.0604, "step": 11080 }, { "grad_norm": 0.4372442662715912, "learning_rate": 9.786856276266685e-05, "loss": 0.0577, "step": 11090 }, { "grad_norm": 0.3447108566761017, "learning_rate": 9.786258840053845e-05, "loss": 0.065, "step": 11100 }, { "grad_norm": 0.6004694104194641, "learning_rate": 9.785660586000273e-05, "loss": 0.0623, "step": 11110 }, { "grad_norm": 0.42642202973365784, "learning_rate": 9.785061514208196e-05, "loss": 0.0598, "step": 11120 }, { "grad_norm": 0.41115108132362366, "learning_rate": 9.784461624779977e-05, "loss": 0.0589, "step": 11130 }, { "grad_norm": 0.4486196041107178, "learning_rate": 9.783860917818123e-05, "loss": 0.0544, "step": 11140 }, { "grad_norm": 0.4117847979068756, "learning_rate": 9.783259393425277e-05, "loss": 0.045, "step": 11150 }, { "grad_norm": 0.6133565306663513, "learning_rate": 9.782657051704221e-05, "loss": 0.054, "step": 11160 }, { "grad_norm": 0.4594074785709381, "learning_rate": 9.782053892757883e-05, "loss": 0.0561, "step": 11170 }, { "grad_norm": 0.5255475640296936, "learning_rate": 9.781449916689324e-05, "loss": 0.0544, "step": 11180 }, { "grad_norm": 0.5214667320251465, "learning_rate": 9.780845123601746e-05, "loss": 0.0571, "step": 11190 }, { "grad_norm": 0.49846774339675903, "learning_rate": 9.780239513598492e-05, "loss": 0.0588, "step": 11200 }, { "grad_norm": 0.39375802874565125, "learning_rate": 9.779633086783047e-05, "loss": 0.0559, "step": 11210 }, { "grad_norm": 0.45706015825271606, "learning_rate": 9.779025843259031e-05, "loss": 0.0532, "step": 11220 }, { "grad_norm": 0.5706455111503601, "learning_rate": 9.778417783130204e-05, "loss": 0.0668, "step": 11230 }, { "grad_norm": 0.5305402278900146, "learning_rate": 9.777808906500468e-05, "loss": 0.0573, "step": 11240 }, { "grad_norm": 0.4933985769748688, "learning_rate": 9.777199213473862e-05, "loss": 0.043, "step": 11250 }, { "grad_norm": 0.49892476201057434, "learning_rate": 9.77658870415457e-05, "loss": 0.0449, "step": 11260 }, { "grad_norm": 0.4436250925064087, "learning_rate": 9.775977378646906e-05, "loss": 0.05, "step": 11270 }, { "grad_norm": 0.5442119240760803, "learning_rate": 9.775365237055331e-05, "loss": 0.0662, "step": 11280 }, { "grad_norm": 0.44805675745010376, "learning_rate": 9.774752279484445e-05, "loss": 0.0581, "step": 11290 }, { "grad_norm": 0.4007732570171356, "learning_rate": 9.774138506038984e-05, "loss": 0.055, "step": 11300 }, { "grad_norm": 0.5815008282661438, "learning_rate": 9.773523916823826e-05, "loss": 0.0591, "step": 11310 }, { "grad_norm": 0.42259106040000916, "learning_rate": 9.772908511943986e-05, "loss": 0.0586, "step": 11320 }, { "grad_norm": 0.39687833189964294, "learning_rate": 9.77229229150462e-05, "loss": 0.0524, "step": 11330 }, { "grad_norm": 0.5158635973930359, "learning_rate": 9.771675255611024e-05, "loss": 0.0509, "step": 11340 }, { "grad_norm": 0.5352692008018494, "learning_rate": 9.771057404368632e-05, "loss": 0.0831, "step": 11350 }, { "grad_norm": 0.420667827129364, "learning_rate": 9.770438737883018e-05, "loss": 0.0597, "step": 11360 }, { "grad_norm": 0.5376151204109192, "learning_rate": 9.769819256259898e-05, "loss": 0.0502, "step": 11370 }, { "grad_norm": 0.39619266986846924, "learning_rate": 9.769198959605119e-05, "loss": 0.0575, "step": 11380 }, { "grad_norm": 0.43073800206184387, "learning_rate": 9.768577848024678e-05, "loss": 0.0648, "step": 11390 }, { "grad_norm": 0.44861170649528503, "learning_rate": 9.767955921624702e-05, "loss": 0.0682, "step": 11400 }, { "grad_norm": 0.6038893461227417, "learning_rate": 9.767333180511465e-05, "loss": 0.0505, "step": 11410 }, { "grad_norm": 0.34445539116859436, "learning_rate": 9.766709624791373e-05, "loss": 0.0674, "step": 11420 }, { "grad_norm": 0.39816683530807495, "learning_rate": 9.766085254570975e-05, "loss": 0.0577, "step": 11430 }, { "grad_norm": 0.49254554510116577, "learning_rate": 9.76546006995696e-05, "loss": 0.0472, "step": 11440 }, { "grad_norm": 0.5533846020698547, "learning_rate": 9.764834071056155e-05, "loss": 0.0545, "step": 11450 }, { "grad_norm": 0.43040886521339417, "learning_rate": 9.764207257975526e-05, "loss": 0.0601, "step": 11460 }, { "grad_norm": 0.7517008781433105, "learning_rate": 9.763579630822179e-05, "loss": 0.0509, "step": 11470 }, { "grad_norm": 0.6216162443161011, "learning_rate": 9.762951189703356e-05, "loss": 0.0672, "step": 11480 }, { "grad_norm": 0.4522243142127991, "learning_rate": 9.762321934726442e-05, "loss": 0.0728, "step": 11490 }, { "grad_norm": 0.39372876286506653, "learning_rate": 9.761691865998959e-05, "loss": 0.0576, "step": 11500 }, { "grad_norm": 0.4790498912334442, "learning_rate": 9.76106098362857e-05, "loss": 0.06, "step": 11510 }, { "grad_norm": 0.49909940361976624, "learning_rate": 9.760429287723072e-05, "loss": 0.0607, "step": 11520 }, { "grad_norm": 0.5013665556907654, "learning_rate": 9.759796778390406e-05, "loss": 0.0573, "step": 11530 }, { "grad_norm": 0.3814966678619385, "learning_rate": 9.759163455738653e-05, "loss": 0.056, "step": 11540 }, { "grad_norm": 0.5530051589012146, "learning_rate": 9.75852931987603e-05, "loss": 0.0573, "step": 11550 }, { "grad_norm": 0.32553163170814514, "learning_rate": 9.757894370910891e-05, "loss": 0.0517, "step": 11560 }, { "grad_norm": 0.38376927375793457, "learning_rate": 9.757258608951733e-05, "loss": 0.0539, "step": 11570 }, { "grad_norm": 0.6145081520080566, "learning_rate": 9.75662203410719e-05, "loss": 0.063, "step": 11580 }, { "grad_norm": 0.4706471860408783, "learning_rate": 9.755984646486034e-05, "loss": 0.054, "step": 11590 }, { "grad_norm": 0.46679678559303284, "learning_rate": 9.75534644619718e-05, "loss": 0.052, "step": 11600 }, { "grad_norm": 0.4300582706928253, "learning_rate": 9.754707433349676e-05, "loss": 0.0637, "step": 11610 }, { "grad_norm": 0.5521615147590637, "learning_rate": 9.754067608052715e-05, "loss": 0.0647, "step": 11620 }, { "grad_norm": 0.5298306345939636, "learning_rate": 9.753426970415622e-05, "loss": 0.0708, "step": 11630 }, { "grad_norm": 0.6096660494804382, "learning_rate": 9.752785520547868e-05, "loss": 0.0521, "step": 11640 }, { "grad_norm": 0.49628931283950806, "learning_rate": 9.752143258559056e-05, "loss": 0.059, "step": 11650 }, { "grad_norm": 0.6865842342376709, "learning_rate": 9.751500184558933e-05, "loss": 0.0585, "step": 11660 }, { "grad_norm": 0.41409286856651306, "learning_rate": 9.750856298657383e-05, "loss": 0.0495, "step": 11670 }, { "grad_norm": 0.6536752581596375, "learning_rate": 9.750211600964428e-05, "loss": 0.0625, "step": 11680 }, { "grad_norm": 0.5298307538032532, "learning_rate": 9.749566091590226e-05, "loss": 0.0625, "step": 11690 }, { "grad_norm": 0.48154905438423157, "learning_rate": 9.748919770645083e-05, "loss": 0.0604, "step": 11700 }, { "grad_norm": 0.36767834424972534, "learning_rate": 9.748272638239432e-05, "loss": 0.054, "step": 11710 }, { "grad_norm": 0.517586350440979, "learning_rate": 9.747624694483855e-05, "loss": 0.0545, "step": 11720 }, { "grad_norm": 0.4086769223213196, "learning_rate": 9.746975939489065e-05, "loss": 0.0514, "step": 11730 }, { "grad_norm": 0.46032723784446716, "learning_rate": 9.746326373365918e-05, "loss": 0.0541, "step": 11740 }, { "grad_norm": 0.45891934633255005, "learning_rate": 9.745675996225403e-05, "loss": 0.0644, "step": 11750 }, { "grad_norm": 0.4396396577358246, "learning_rate": 9.745024808178657e-05, "loss": 0.0549, "step": 11760 }, { "grad_norm": 0.43892547488212585, "learning_rate": 9.744372809336947e-05, "loss": 0.0553, "step": 11770 }, { "grad_norm": 0.5791813731193542, "learning_rate": 9.743719999811682e-05, "loss": 0.0515, "step": 11780 }, { "grad_norm": 0.45333772897720337, "learning_rate": 9.743066379714412e-05, "loss": 0.0494, "step": 11790 }, { "grad_norm": 0.3274403512477875, "learning_rate": 9.74241194915682e-05, "loss": 0.0505, "step": 11800 }, { "grad_norm": 0.42046311497688293, "learning_rate": 9.741756708250731e-05, "loss": 0.0655, "step": 11810 }, { "grad_norm": 0.3470471203327179, "learning_rate": 9.741100657108109e-05, "loss": 0.0531, "step": 11820 }, { "grad_norm": 0.43529194593429565, "learning_rate": 9.740443795841054e-05, "loss": 0.0693, "step": 11830 }, { "grad_norm": 0.45743528008461, "learning_rate": 9.739786124561805e-05, "loss": 0.0496, "step": 11840 }, { "grad_norm": 0.3347649574279785, "learning_rate": 9.73912764338274e-05, "loss": 0.051, "step": 11850 }, { "grad_norm": 0.34440651535987854, "learning_rate": 9.738468352416377e-05, "loss": 0.0571, "step": 11860 }, { "grad_norm": 0.4347904920578003, "learning_rate": 9.737808251775369e-05, "loss": 0.055, "step": 11870 }, { "grad_norm": 0.4691619575023651, "learning_rate": 9.737147341572512e-05, "loss": 0.0661, "step": 11880 }, { "grad_norm": 0.5879796147346497, "learning_rate": 9.736485621920735e-05, "loss": 0.0502, "step": 11890 }, { "grad_norm": 0.5477827191352844, "learning_rate": 9.735823092933108e-05, "loss": 0.0617, "step": 11900 }, { "grad_norm": 0.48528754711151123, "learning_rate": 9.735159754722838e-05, "loss": 0.0632, "step": 11910 }, { "grad_norm": 0.4765531122684479, "learning_rate": 9.734495607403275e-05, "loss": 0.0496, "step": 11920 }, { "grad_norm": 0.486745148897171, "learning_rate": 9.733830651087901e-05, "loss": 0.0541, "step": 11930 }, { "grad_norm": 0.4001593291759491, "learning_rate": 9.733164885890338e-05, "loss": 0.0464, "step": 11940 }, { "grad_norm": 0.37598931789398193, "learning_rate": 9.732498311924349e-05, "loss": 0.0389, "step": 11950 }, { "grad_norm": 0.41210609674453735, "learning_rate": 9.731830929303833e-05, "loss": 0.0483, "step": 11960 }, { "grad_norm": 0.5776318907737732, "learning_rate": 9.731162738142827e-05, "loss": 0.0651, "step": 11970 }, { "grad_norm": 0.5080376267433167, "learning_rate": 9.730493738555506e-05, "loss": 0.0543, "step": 11980 }, { "grad_norm": 0.41969093680381775, "learning_rate": 9.729823930656186e-05, "loss": 0.0555, "step": 11990 }, { "grad_norm": 0.42516812682151794, "learning_rate": 9.729153314559316e-05, "loss": 0.0629, "step": 12000 }, { "grad_norm": 0.5110496878623962, "learning_rate": 9.728481890379486e-05, "loss": 0.0565, "step": 12010 }, { "grad_norm": 0.4503949284553528, "learning_rate": 9.727809658231428e-05, "loss": 0.0499, "step": 12020 }, { "grad_norm": 0.369256854057312, "learning_rate": 9.727136618230003e-05, "loss": 0.047, "step": 12030 }, { "grad_norm": 0.4903275668621063, "learning_rate": 9.726462770490219e-05, "loss": 0.0579, "step": 12040 }, { "grad_norm": 0.44384831190109253, "learning_rate": 9.725788115127214e-05, "loss": 0.0578, "step": 12050 }, { "grad_norm": 0.5298236608505249, "learning_rate": 9.725112652256274e-05, "loss": 0.0528, "step": 12060 }, { "grad_norm": 0.4735782742500305, "learning_rate": 9.724436381992812e-05, "loss": 0.0738, "step": 12070 }, { "grad_norm": 0.5773546099662781, "learning_rate": 9.723759304452387e-05, "loss": 0.0551, "step": 12080 }, { "grad_norm": 0.5169771909713745, "learning_rate": 9.72308141975069e-05, "loss": 0.0533, "step": 12090 }, { "grad_norm": 0.5524449348449707, "learning_rate": 9.722402728003557e-05, "loss": 0.0714, "step": 12100 }, { "grad_norm": 0.5255969762802124, "learning_rate": 9.721723229326953e-05, "loss": 0.0559, "step": 12110 }, { "grad_norm": 0.5472707748413086, "learning_rate": 9.721042923836992e-05, "loss": 0.0453, "step": 12120 }, { "grad_norm": 0.3395029306411743, "learning_rate": 9.720361811649914e-05, "loss": 0.0487, "step": 12130 }, { "grad_norm": 0.5194413065910339, "learning_rate": 9.719679892882106e-05, "loss": 0.0474, "step": 12140 }, { "grad_norm": 0.5079523921012878, "learning_rate": 9.718997167650085e-05, "loss": 0.0486, "step": 12150 }, { "grad_norm": 0.41054049134254456, "learning_rate": 9.718313636070515e-05, "loss": 0.0418, "step": 12160 }, { "grad_norm": 0.4685823321342468, "learning_rate": 9.717629298260192e-05, "loss": 0.0509, "step": 12170 }, { "grad_norm": 0.7236262559890747, "learning_rate": 9.716944154336047e-05, "loss": 0.0624, "step": 12180 }, { "grad_norm": 0.40511101484298706, "learning_rate": 9.716258204415157e-05, "loss": 0.0533, "step": 12190 }, { "grad_norm": 0.35355862975120544, "learning_rate": 9.715571448614728e-05, "loss": 0.0616, "step": 12200 }, { "grad_norm": 0.4287075698375702, "learning_rate": 9.71488388705211e-05, "loss": 0.0664, "step": 12210 }, { "grad_norm": 0.6142800450325012, "learning_rate": 9.714195519844788e-05, "loss": 0.0455, "step": 12220 }, { "grad_norm": 0.5355135798454285, "learning_rate": 9.713506347110386e-05, "loss": 0.0584, "step": 12230 }, { "grad_norm": 0.29435086250305176, "learning_rate": 9.712816368966662e-05, "loss": 0.0433, "step": 12240 }, { "grad_norm": 0.40911024808883667, "learning_rate": 9.712125585531517e-05, "loss": 0.0604, "step": 12250 }, { "grad_norm": 0.554077684879303, "learning_rate": 9.711433996922988e-05, "loss": 0.0718, "step": 12260 }, { "grad_norm": 0.47368738055229187, "learning_rate": 9.710741603259245e-05, "loss": 0.0598, "step": 12270 }, { "grad_norm": 0.40519851446151733, "learning_rate": 9.710048404658603e-05, "loss": 0.049, "step": 12280 }, { "grad_norm": 0.38451632857322693, "learning_rate": 9.709354401239508e-05, "loss": 0.054, "step": 12290 }, { "grad_norm": 0.48110467195510864, "learning_rate": 9.708659593120546e-05, "loss": 0.0468, "step": 12300 }, { "grad_norm": 0.5421892404556274, "learning_rate": 9.707963980420443e-05, "loss": 0.0475, "step": 12310 }, { "grad_norm": 0.5833951234817505, "learning_rate": 9.707267563258058e-05, "loss": 0.0631, "step": 12320 }, { "grad_norm": 0.3763349652290344, "learning_rate": 9.70657034175239e-05, "loss": 0.0535, "step": 12330 }, { "grad_norm": 0.5003994107246399, "learning_rate": 9.705872316022577e-05, "loss": 0.0604, "step": 12340 }, { "grad_norm": 0.6811301112174988, "learning_rate": 9.705173486187891e-05, "loss": 0.0554, "step": 12350 }, { "grad_norm": 0.49156033992767334, "learning_rate": 9.704473852367741e-05, "loss": 0.0636, "step": 12360 }, { "grad_norm": 0.4776475131511688, "learning_rate": 9.70377341468168e-05, "loss": 0.0457, "step": 12370 }, { "grad_norm": 0.354098379611969, "learning_rate": 9.703072173249389e-05, "loss": 0.0487, "step": 12380 }, { "grad_norm": 0.44805291295051575, "learning_rate": 9.702370128190693e-05, "loss": 0.0487, "step": 12390 }, { "grad_norm": 0.5002039670944214, "learning_rate": 9.701667279625552e-05, "loss": 0.0593, "step": 12400 }, { "grad_norm": 0.5378288626670837, "learning_rate": 9.700963627674065e-05, "loss": 0.0609, "step": 12410 }, { "grad_norm": 0.5778486132621765, "learning_rate": 9.700259172456466e-05, "loss": 0.0531, "step": 12420 }, { "grad_norm": 0.5893649458885193, "learning_rate": 9.699553914093124e-05, "loss": 0.0623, "step": 12430 }, { "grad_norm": 0.5741661190986633, "learning_rate": 9.698847852704553e-05, "loss": 0.0661, "step": 12440 }, { "grad_norm": 0.5081756114959717, "learning_rate": 9.6981409884114e-05, "loss": 0.0523, "step": 12450 }, { "grad_norm": 0.46647462248802185, "learning_rate": 9.697433321334443e-05, "loss": 0.0566, "step": 12460 }, { "grad_norm": 0.32966548204421997, "learning_rate": 9.696724851594607e-05, "loss": 0.0528, "step": 12470 }, { "grad_norm": 0.44350796937942505, "learning_rate": 9.696015579312952e-05, "loss": 0.0459, "step": 12480 }, { "grad_norm": 0.3978257179260254, "learning_rate": 9.695305504610668e-05, "loss": 0.0515, "step": 12490 }, { "grad_norm": 0.4652993381023407, "learning_rate": 9.694594627609092e-05, "loss": 0.0543, "step": 12500 }, { "grad_norm": 0.6129014492034912, "learning_rate": 9.693882948429691e-05, "loss": 0.0684, "step": 12510 }, { "grad_norm": 0.3872971832752228, "learning_rate": 9.693170467194071e-05, "loss": 0.0457, "step": 12520 }, { "grad_norm": 0.3239416182041168, "learning_rate": 9.692457184023977e-05, "loss": 0.0502, "step": 12530 }, { "grad_norm": 0.33713364601135254, "learning_rate": 9.691743099041291e-05, "loss": 0.0525, "step": 12540 }, { "grad_norm": 0.5117162466049194, "learning_rate": 9.691028212368027e-05, "loss": 0.0473, "step": 12550 }, { "grad_norm": 0.5128705501556396, "learning_rate": 9.690312524126342e-05, "loss": 0.0423, "step": 12560 }, { "grad_norm": 0.443406343460083, "learning_rate": 9.689596034438527e-05, "loss": 0.0493, "step": 12570 }, { "grad_norm": 0.28008216619491577, "learning_rate": 9.688878743427012e-05, "loss": 0.0446, "step": 12580 }, { "grad_norm": 0.35233446955680847, "learning_rate": 9.688160651214359e-05, "loss": 0.055, "step": 12590 }, { "grad_norm": 0.45977699756622314, "learning_rate": 9.687441757923273e-05, "loss": 0.0497, "step": 12600 }, { "grad_norm": 0.3841215968132019, "learning_rate": 9.68672206367659e-05, "loss": 0.0456, "step": 12610 }, { "grad_norm": 0.41375818848609924, "learning_rate": 9.686001568597291e-05, "loss": 0.0444, "step": 12620 }, { "grad_norm": 0.38841426372528076, "learning_rate": 9.685280272808486e-05, "loss": 0.0425, "step": 12630 }, { "grad_norm": 0.44310903549194336, "learning_rate": 9.684558176433424e-05, "loss": 0.0595, "step": 12640 }, { "grad_norm": 0.5031426548957825, "learning_rate": 9.683835279595495e-05, "loss": 0.0468, "step": 12650 }, { "grad_norm": 0.4648231267929077, "learning_rate": 9.683111582418216e-05, "loss": 0.06, "step": 12660 }, { "grad_norm": 0.3778114914894104, "learning_rate": 9.682387085025254e-05, "loss": 0.044, "step": 12670 }, { "grad_norm": 0.5070315599441528, "learning_rate": 9.681661787540401e-05, "loss": 0.0614, "step": 12680 }, { "grad_norm": 0.5273488163948059, "learning_rate": 9.680935690087593e-05, "loss": 0.0523, "step": 12690 }, { "grad_norm": 0.45934879779815674, "learning_rate": 9.680208792790901e-05, "loss": 0.057, "step": 12700 }, { "grad_norm": 0.44893497228622437, "learning_rate": 9.679481095774529e-05, "loss": 0.048, "step": 12710 }, { "grad_norm": 0.43224087357521057, "learning_rate": 9.678752599162822e-05, "loss": 0.0519, "step": 12720 }, { "grad_norm": 0.5331626534461975, "learning_rate": 9.678023303080259e-05, "loss": 0.049, "step": 12730 }, { "grad_norm": 0.5849446058273315, "learning_rate": 9.677293207651459e-05, "loss": 0.0493, "step": 12740 }, { "grad_norm": 0.3764193654060364, "learning_rate": 9.676562313001173e-05, "loss": 0.0484, "step": 12750 }, { "grad_norm": 0.5859647989273071, "learning_rate": 9.675830619254293e-05, "loss": 0.0674, "step": 12760 }, { "grad_norm": 0.5574691295623779, "learning_rate": 9.675098126535843e-05, "loss": 0.0672, "step": 12770 }, { "grad_norm": 0.5643901824951172, "learning_rate": 9.674364834970988e-05, "loss": 0.0637, "step": 12780 }, { "grad_norm": 0.4792410135269165, "learning_rate": 9.673630744685028e-05, "loss": 0.0609, "step": 12790 }, { "grad_norm": 0.4050619304180145, "learning_rate": 9.672895855803397e-05, "loss": 0.0557, "step": 12800 }, { "grad_norm": 0.6106849908828735, "learning_rate": 9.672160168451667e-05, "loss": 0.0524, "step": 12810 }, { "grad_norm": 0.5384910106658936, "learning_rate": 9.671423682755549e-05, "loss": 0.0674, "step": 12820 }, { "grad_norm": 0.622308611869812, "learning_rate": 9.670686398840888e-05, "loss": 0.0696, "step": 12830 }, { "grad_norm": 0.5618612170219421, "learning_rate": 9.669948316833664e-05, "loss": 0.048, "step": 12840 }, { "grad_norm": 0.37126895785331726, "learning_rate": 9.669209436859997e-05, "loss": 0.0479, "step": 12850 }, { "grad_norm": 0.3920834958553314, "learning_rate": 9.66846975904614e-05, "loss": 0.0486, "step": 12860 }, { "grad_norm": 0.48166385293006897, "learning_rate": 9.667729283518483e-05, "loss": 0.0447, "step": 12870 }, { "grad_norm": 0.5520930290222168, "learning_rate": 9.666988010403557e-05, "loss": 0.0457, "step": 12880 }, { "grad_norm": 0.33021751046180725, "learning_rate": 9.66624593982802e-05, "loss": 0.0464, "step": 12890 }, { "grad_norm": 0.4095058739185333, "learning_rate": 9.665503071918675e-05, "loss": 0.0642, "step": 12900 }, { "grad_norm": 0.49954867362976074, "learning_rate": 9.664759406802456e-05, "loss": 0.0453, "step": 12910 }, { "grad_norm": 0.4599502980709076, "learning_rate": 9.664014944606437e-05, "loss": 0.0375, "step": 12920 }, { "grad_norm": 0.3247120678424835, "learning_rate": 9.663269685457822e-05, "loss": 0.0506, "step": 12930 }, { "grad_norm": 0.414707750082016, "learning_rate": 9.662523629483962e-05, "loss": 0.0472, "step": 12940 }, { "grad_norm": 0.3710594177246094, "learning_rate": 9.661776776812333e-05, "loss": 0.0403, "step": 12950 }, { "grad_norm": 0.45042985677719116, "learning_rate": 9.661029127570553e-05, "loss": 0.046, "step": 12960 }, { "grad_norm": 0.48707062005996704, "learning_rate": 9.660280681886373e-05, "loss": 0.0443, "step": 12970 }, { "grad_norm": 0.30556946992874146, "learning_rate": 9.659531439887685e-05, "loss": 0.0384, "step": 12980 }, { "grad_norm": 0.454629123210907, "learning_rate": 9.658781401702511e-05, "loss": 0.0557, "step": 12990 }, { "grad_norm": 0.531115710735321, "learning_rate": 9.658030567459015e-05, "loss": 0.0455, "step": 13000 }, { "grad_norm": 0.3886556029319763, "learning_rate": 9.65727893728549e-05, "loss": 0.063, "step": 13010 }, { "grad_norm": 0.5053225159645081, "learning_rate": 9.656526511310375e-05, "loss": 0.0616, "step": 13020 }, { "grad_norm": 0.47039151191711426, "learning_rate": 9.655773289662233e-05, "loss": 0.0597, "step": 13030 }, { "grad_norm": 0.5106619000434875, "learning_rate": 9.655019272469772e-05, "loss": 0.0537, "step": 13040 }, { "grad_norm": 0.4104464054107666, "learning_rate": 9.654264459861832e-05, "loss": 0.0395, "step": 13050 }, { "grad_norm": 0.42878320813179016, "learning_rate": 9.653508851967391e-05, "loss": 0.0471, "step": 13060 }, { "grad_norm": 0.5868843793869019, "learning_rate": 9.65275244891556e-05, "loss": 0.0581, "step": 13070 }, { "grad_norm": 0.4486425220966339, "learning_rate": 9.651995250835591e-05, "loss": 0.0484, "step": 13080 }, { "grad_norm": 0.4431191086769104, "learning_rate": 9.651237257856862e-05, "loss": 0.0561, "step": 13090 }, { "grad_norm": 0.6279934048652649, "learning_rate": 9.6504784701089e-05, "loss": 0.0508, "step": 13100 }, { "grad_norm": 0.36333754658699036, "learning_rate": 9.649718887721357e-05, "loss": 0.0485, "step": 13110 }, { "grad_norm": 0.616618275642395, "learning_rate": 9.648958510824028e-05, "loss": 0.0493, "step": 13120 }, { "grad_norm": 0.5640337467193604, "learning_rate": 9.648197339546837e-05, "loss": 0.06, "step": 13130 }, { "grad_norm": 0.3802449107170105, "learning_rate": 9.647435374019851e-05, "loss": 0.0553, "step": 13140 }, { "grad_norm": 0.5292030572891235, "learning_rate": 9.646672614373266e-05, "loss": 0.0546, "step": 13150 }, { "grad_norm": 0.6624547243118286, "learning_rate": 9.645909060737418e-05, "loss": 0.0497, "step": 13160 }, { "grad_norm": 0.38488054275512695, "learning_rate": 9.645144713242778e-05, "loss": 0.0447, "step": 13170 }, { "grad_norm": 0.37788593769073486, "learning_rate": 9.64437957201995e-05, "loss": 0.0542, "step": 13180 }, { "grad_norm": 0.34683796763420105, "learning_rate": 9.643613637199678e-05, "loss": 0.0452, "step": 13190 }, { "grad_norm": 0.5273218154907227, "learning_rate": 9.642846908912839e-05, "loss": 0.0605, "step": 13200 }, { "grad_norm": 0.4402712881565094, "learning_rate": 9.642079387290444e-05, "loss": 0.0496, "step": 13210 }, { "grad_norm": 0.4012521207332611, "learning_rate": 9.641311072463644e-05, "loss": 0.0492, "step": 13220 }, { "grad_norm": 0.4089004099369049, "learning_rate": 9.640541964563722e-05, "loss": 0.0574, "step": 13230 }, { "grad_norm": 0.49065154790878296, "learning_rate": 9.639772063722096e-05, "loss": 0.0635, "step": 13240 }, { "grad_norm": 0.3508399724960327, "learning_rate": 9.639001370070324e-05, "loss": 0.0455, "step": 13250 }, { "grad_norm": 0.48096081614494324, "learning_rate": 9.638229883740095e-05, "loss": 0.0468, "step": 13260 }, { "grad_norm": 0.3791355788707733, "learning_rate": 9.637457604863233e-05, "loss": 0.0431, "step": 13270 }, { "grad_norm": 0.3492566645145416, "learning_rate": 9.636684533571703e-05, "loss": 0.0459, "step": 13280 }, { "grad_norm": 0.3172915279865265, "learning_rate": 9.635910669997599e-05, "loss": 0.0453, "step": 13290 }, { "grad_norm": 1.0009628534317017, "learning_rate": 9.635136014273154e-05, "loss": 0.0498, "step": 13300 }, { "grad_norm": 0.5209001898765564, "learning_rate": 9.634360566530735e-05, "loss": 0.0568, "step": 13310 }, { "grad_norm": 0.4980361759662628, "learning_rate": 9.633584326902845e-05, "loss": 0.0681, "step": 13320 }, { "grad_norm": 0.505439281463623, "learning_rate": 9.632807295522124e-05, "loss": 0.0506, "step": 13330 }, { "grad_norm": 0.4594617187976837, "learning_rate": 9.632029472521342e-05, "loss": 0.0478, "step": 13340 }, { "grad_norm": 0.4278815686702728, "learning_rate": 9.631250858033409e-05, "loss": 0.0485, "step": 13350 }, { "grad_norm": 0.46127697825431824, "learning_rate": 9.630471452191371e-05, "loss": 0.0541, "step": 13360 }, { "grad_norm": 0.29355350136756897, "learning_rate": 9.629691255128405e-05, "loss": 0.0619, "step": 13370 }, { "grad_norm": 0.5808064341545105, "learning_rate": 9.628910266977825e-05, "loss": 0.0595, "step": 13380 }, { "grad_norm": 0.43618836998939514, "learning_rate": 9.628128487873083e-05, "loss": 0.0576, "step": 13390 }, { "grad_norm": 0.4695419371128082, "learning_rate": 9.627345917947761e-05, "loss": 0.0597, "step": 13400 }, { "grad_norm": 0.5262807011604309, "learning_rate": 9.626562557335579e-05, "loss": 0.0502, "step": 13410 }, { "grad_norm": 0.7205910682678223, "learning_rate": 9.625778406170393e-05, "loss": 0.0529, "step": 13420 }, { "grad_norm": 0.6317233443260193, "learning_rate": 9.624993464586193e-05, "loss": 0.051, "step": 13430 }, { "grad_norm": 0.48445212841033936, "learning_rate": 9.624207732717105e-05, "loss": 0.0555, "step": 13440 }, { "grad_norm": 0.5178702473640442, "learning_rate": 9.623421210697386e-05, "loss": 0.0609, "step": 13450 }, { "grad_norm": 0.5910674333572388, "learning_rate": 9.622633898661434e-05, "loss": 0.0629, "step": 13460 }, { "grad_norm": 0.4830735921859741, "learning_rate": 9.621845796743778e-05, "loss": 0.0681, "step": 13470 }, { "grad_norm": 0.35277852416038513, "learning_rate": 9.621056905079082e-05, "loss": 0.0579, "step": 13480 }, { "grad_norm": 0.40155765414237976, "learning_rate": 9.620267223802149e-05, "loss": 0.0611, "step": 13490 }, { "grad_norm": 0.565239667892456, "learning_rate": 9.619476753047911e-05, "loss": 0.0537, "step": 13500 }, { "grad_norm": 0.5647477507591248, "learning_rate": 9.618685492951438e-05, "loss": 0.0609, "step": 13510 }, { "grad_norm": 0.4155292510986328, "learning_rate": 9.617893443647938e-05, "loss": 0.0675, "step": 13520 }, { "grad_norm": 0.48429355025291443, "learning_rate": 9.617100605272746e-05, "loss": 0.0658, "step": 13530 }, { "grad_norm": 0.44200751185417175, "learning_rate": 9.616306977961338e-05, "loss": 0.0561, "step": 13540 }, { "grad_norm": 0.4599977433681488, "learning_rate": 9.615512561849326e-05, "loss": 0.0481, "step": 13550 }, { "grad_norm": 0.5481541752815247, "learning_rate": 9.61471735707245e-05, "loss": 0.0561, "step": 13560 }, { "grad_norm": 0.38243719935417175, "learning_rate": 9.613921363766592e-05, "loss": 0.0456, "step": 13570 }, { "grad_norm": 0.4860456585884094, "learning_rate": 9.613124582067763e-05, "loss": 0.053, "step": 13580 }, { "grad_norm": 0.5474532246589661, "learning_rate": 9.612327012112112e-05, "loss": 0.0624, "step": 13590 }, { "grad_norm": 0.5052236318588257, "learning_rate": 9.611528654035921e-05, "loss": 0.0758, "step": 13600 }, { "grad_norm": 0.4002145230770111, "learning_rate": 9.610729507975611e-05, "loss": 0.0452, "step": 13610 }, { "grad_norm": 0.4125348925590515, "learning_rate": 9.609929574067731e-05, "loss": 0.0642, "step": 13620 }, { "grad_norm": 0.3728492856025696, "learning_rate": 9.609128852448967e-05, "loss": 0.042, "step": 13630 }, { "grad_norm": 0.40594857931137085, "learning_rate": 9.608327343256143e-05, "loss": 0.0488, "step": 13640 }, { "grad_norm": 0.44392696022987366, "learning_rate": 9.607525046626216e-05, "loss": 0.0516, "step": 13650 }, { "grad_norm": 0.4999943673610687, "learning_rate": 9.606721962696272e-05, "loss": 0.0561, "step": 13660 }, { "grad_norm": 0.4576769769191742, "learning_rate": 9.60591809160354e-05, "loss": 0.0533, "step": 13670 }, { "grad_norm": 0.3948780596256256, "learning_rate": 9.605113433485378e-05, "loss": 0.0504, "step": 13680 }, { "grad_norm": 0.6685864925384521, "learning_rate": 9.604307988479279e-05, "loss": 0.0484, "step": 13690 }, { "grad_norm": 0.3898003399372101, "learning_rate": 9.603501756722876e-05, "loss": 0.0524, "step": 13700 }, { "grad_norm": 0.41876980662345886, "learning_rate": 9.602694738353927e-05, "loss": 0.043, "step": 13710 }, { "grad_norm": 0.3511662483215332, "learning_rate": 9.601886933510331e-05, "loss": 0.0411, "step": 13720 }, { "grad_norm": 0.47972938418388367, "learning_rate": 9.60107834233012e-05, "loss": 0.0442, "step": 13730 }, { "grad_norm": 0.38915392756462097, "learning_rate": 9.60026896495146e-05, "loss": 0.0438, "step": 13740 }, { "grad_norm": 0.43271249532699585, "learning_rate": 9.599458801512652e-05, "loss": 0.0464, "step": 13750 }, { "grad_norm": 0.41417524218559265, "learning_rate": 9.598647852152129e-05, "loss": 0.068, "step": 13760 }, { "grad_norm": 0.3939221203327179, "learning_rate": 9.597836117008462e-05, "loss": 0.0572, "step": 13770 }, { "grad_norm": 0.43710511922836304, "learning_rate": 9.597023596220356e-05, "loss": 0.0449, "step": 13780 }, { "grad_norm": 0.5570892095565796, "learning_rate": 9.596210289926643e-05, "loss": 0.0431, "step": 13790 }, { "grad_norm": 0.4948311150074005, "learning_rate": 9.5953961982663e-05, "loss": 0.0391, "step": 13800 }, { "grad_norm": 0.45554086565971375, "learning_rate": 9.594581321378431e-05, "loss": 0.0486, "step": 13810 }, { "grad_norm": 0.3901311755180359, "learning_rate": 9.593765659402276e-05, "loss": 0.049, "step": 13820 }, { "grad_norm": 0.5237751603126526, "learning_rate": 9.59294921247721e-05, "loss": 0.0564, "step": 13830 }, { "grad_norm": 0.4998389780521393, "learning_rate": 9.59213198074274e-05, "loss": 0.0602, "step": 13840 }, { "grad_norm": 0.5254873633384705, "learning_rate": 9.59131396433851e-05, "loss": 0.046, "step": 13850 }, { "grad_norm": 0.5199874043464661, "learning_rate": 9.590495163404297e-05, "loss": 0.0619, "step": 13860 }, { "grad_norm": 0.34007108211517334, "learning_rate": 9.589675578080009e-05, "loss": 0.054, "step": 13870 }, { "grad_norm": 0.4613988697528839, "learning_rate": 9.588855208505694e-05, "loss": 0.0447, "step": 13880 }, { "grad_norm": 0.6762868165969849, "learning_rate": 9.588034054821529e-05, "loss": 0.0746, "step": 13890 }, { "grad_norm": 0.5612926483154297, "learning_rate": 9.587212117167826e-05, "loss": 0.0452, "step": 13900 }, { "grad_norm": 0.4577140510082245, "learning_rate": 9.586389395685033e-05, "loss": 0.0471, "step": 13910 }, { "grad_norm": 0.5878139138221741, "learning_rate": 9.585565890513733e-05, "loss": 0.0568, "step": 13920 }, { "grad_norm": 0.46715033054351807, "learning_rate": 9.584741601794636e-05, "loss": 0.049, "step": 13930 }, { "grad_norm": 0.5450780987739563, "learning_rate": 9.58391652966859e-05, "loss": 0.0545, "step": 13940 }, { "grad_norm": 0.5376446843147278, "learning_rate": 9.583090674276583e-05, "loss": 0.0501, "step": 13950 }, { "grad_norm": 0.37083256244659424, "learning_rate": 9.582264035759726e-05, "loss": 0.0474, "step": 13960 }, { "grad_norm": 0.5936755537986755, "learning_rate": 9.58143661425927e-05, "loss": 0.0882, "step": 13970 }, { "grad_norm": 0.560678243637085, "learning_rate": 9.580608409916601e-05, "loss": 0.0596, "step": 13980 }, { "grad_norm": 0.38605812191963196, "learning_rate": 9.579779422873233e-05, "loss": 0.0558, "step": 13990 }, { "grad_norm": 0.3805879056453705, "learning_rate": 9.578949653270819e-05, "loss": 0.0515, "step": 14000 }, { "grad_norm": 0.5333142876625061, "learning_rate": 9.578119101251144e-05, "loss": 0.0478, "step": 14010 }, { "grad_norm": 0.3812043070793152, "learning_rate": 9.577287766956127e-05, "loss": 0.0575, "step": 14020 }, { "grad_norm": 0.35816490650177, "learning_rate": 9.57645565052782e-05, "loss": 0.055, "step": 14030 }, { "grad_norm": 0.5386207103729248, "learning_rate": 9.575622752108407e-05, "loss": 0.0461, "step": 14040 }, { "grad_norm": 0.46950623393058777, "learning_rate": 9.57478907184021e-05, "loss": 0.0455, "step": 14050 }, { "grad_norm": 0.540614128112793, "learning_rate": 9.573954609865681e-05, "loss": 0.0532, "step": 14060 }, { "grad_norm": 0.4639800190925598, "learning_rate": 9.573119366327408e-05, "loss": 0.0496, "step": 14070 }, { "grad_norm": 0.5146223902702332, "learning_rate": 9.57228334136811e-05, "loss": 0.045, "step": 14080 }, { "grad_norm": 0.3315344452857971, "learning_rate": 9.571446535130641e-05, "loss": 0.0492, "step": 14090 }, { "grad_norm": 0.4204854667186737, "learning_rate": 9.570608947757988e-05, "loss": 0.0531, "step": 14100 }, { "grad_norm": 0.429479718208313, "learning_rate": 9.569770579393274e-05, "loss": 0.0528, "step": 14110 }, { "grad_norm": 0.37578102946281433, "learning_rate": 9.56893143017975e-05, "loss": 0.0506, "step": 14120 }, { "grad_norm": 0.3796025216579437, "learning_rate": 9.568091500260806e-05, "loss": 0.0641, "step": 14130 }, { "grad_norm": 0.42704781889915466, "learning_rate": 9.567250789779961e-05, "loss": 0.0567, "step": 14140 }, { "grad_norm": 0.5054807662963867, "learning_rate": 9.566409298880872e-05, "loss": 0.0453, "step": 14150 }, { "grad_norm": 0.5623239278793335, "learning_rate": 9.565567027707326e-05, "loss": 0.0686, "step": 14160 }, { "grad_norm": 0.5108526945114136, "learning_rate": 9.56472397640324e-05, "loss": 0.0542, "step": 14170 }, { "grad_norm": 0.40532082319259644, "learning_rate": 9.563880145112675e-05, "loss": 0.0473, "step": 14180 }, { "grad_norm": 0.7596218585968018, "learning_rate": 9.563035533979814e-05, "loss": 0.0574, "step": 14190 }, { "grad_norm": 0.4484008252620697, "learning_rate": 9.562190143148981e-05, "loss": 0.048, "step": 14200 }, { "grad_norm": 0.4326188564300537, "learning_rate": 9.561343972764627e-05, "loss": 0.0472, "step": 14210 }, { "grad_norm": 0.44746044278144836, "learning_rate": 9.560497022971343e-05, "loss": 0.0564, "step": 14220 }, { "grad_norm": 0.4395422637462616, "learning_rate": 9.559649293913847e-05, "loss": 0.0443, "step": 14230 }, { "grad_norm": 0.5401150584220886, "learning_rate": 9.558800785736993e-05, "loss": 0.0473, "step": 14240 }, { "grad_norm": 0.3487298786640167, "learning_rate": 9.557951498585767e-05, "loss": 0.0543, "step": 14250 }, { "grad_norm": 0.3297312259674072, "learning_rate": 9.557101432605293e-05, "loss": 0.0446, "step": 14260 }, { "grad_norm": 0.40787631273269653, "learning_rate": 9.556250587940818e-05, "loss": 0.0526, "step": 14270 }, { "grad_norm": 0.5429772138595581, "learning_rate": 9.555398964737734e-05, "loss": 0.0453, "step": 14280 }, { "grad_norm": 0.4578990340232849, "learning_rate": 9.554546563141555e-05, "loss": 0.0459, "step": 14290 }, { "grad_norm": 0.5284930467605591, "learning_rate": 9.553693383297937e-05, "loss": 0.0604, "step": 14300 }, { "grad_norm": 0.45727404952049255, "learning_rate": 9.552839425352663e-05, "loss": 0.0481, "step": 14310 }, { "grad_norm": 0.34804701805114746, "learning_rate": 9.551984689451652e-05, "loss": 0.0577, "step": 14320 }, { "grad_norm": 0.41687265038490295, "learning_rate": 9.551129175740953e-05, "loss": 0.0474, "step": 14330 }, { "grad_norm": 0.3671542704105377, "learning_rate": 9.550272884366754e-05, "loss": 0.0509, "step": 14340 }, { "grad_norm": 0.5388680696487427, "learning_rate": 9.549415815475369e-05, "loss": 0.0534, "step": 14350 }, { "grad_norm": 0.32774367928504944, "learning_rate": 9.548557969213247e-05, "loss": 0.046, "step": 14360 }, { "grad_norm": 0.9608216285705566, "learning_rate": 9.547699345726972e-05, "loss": 0.0551, "step": 14370 }, { "grad_norm": 0.4264232814311981, "learning_rate": 9.546839945163257e-05, "loss": 0.0489, "step": 14380 }, { "grad_norm": 0.3482356667518616, "learning_rate": 9.545979767668953e-05, "loss": 0.0619, "step": 14390 }, { "grad_norm": 0.4542010426521301, "learning_rate": 9.54511881339104e-05, "loss": 0.0505, "step": 14400 }, { "grad_norm": 0.44111156463623047, "learning_rate": 9.54425708247663e-05, "loss": 0.0396, "step": 14410 }, { "grad_norm": 0.5502352118492126, "learning_rate": 9.543394575072972e-05, "loss": 0.0472, "step": 14420 }, { "grad_norm": 0.4175001382827759, "learning_rate": 9.542531291327441e-05, "loss": 0.0457, "step": 14430 }, { "grad_norm": 0.4309792220592499, "learning_rate": 9.541667231387552e-05, "loss": 0.0425, "step": 14440 }, { "grad_norm": 0.4680728018283844, "learning_rate": 9.540802395400949e-05, "loss": 0.0485, "step": 14450 }, { "grad_norm": 0.4733924865722656, "learning_rate": 9.539936783515406e-05, "loss": 0.0468, "step": 14460 }, { "grad_norm": 0.4276488721370697, "learning_rate": 9.539070395878835e-05, "loss": 0.0543, "step": 14470 }, { "grad_norm": 0.4394306540489197, "learning_rate": 9.538203232639277e-05, "loss": 0.0461, "step": 14480 }, { "grad_norm": 0.36825573444366455, "learning_rate": 9.537335293944907e-05, "loss": 0.0523, "step": 14490 }, { "grad_norm": 0.447885662317276, "learning_rate": 9.536466579944032e-05, "loss": 0.0511, "step": 14500 }, { "grad_norm": 0.33665886521339417, "learning_rate": 9.535597090785091e-05, "loss": 0.0508, "step": 14510 }, { "grad_norm": 0.38642314076423645, "learning_rate": 9.534726826616656e-05, "loss": 0.0434, "step": 14520 }, { "grad_norm": 0.37506619095802307, "learning_rate": 9.53385578758743e-05, "loss": 0.0661, "step": 14530 }, { "grad_norm": 0.4463900327682495, "learning_rate": 9.532983973846252e-05, "loss": 0.0484, "step": 14540 }, { "grad_norm": 0.39269521832466125, "learning_rate": 9.53211138554209e-05, "loss": 0.0525, "step": 14550 }, { "grad_norm": 0.6417310237884521, "learning_rate": 9.531238022824047e-05, "loss": 0.0496, "step": 14560 }, { "grad_norm": 0.44874653220176697, "learning_rate": 9.530363885841355e-05, "loss": 0.0519, "step": 14570 }, { "grad_norm": 0.5659007430076599, "learning_rate": 9.52948897474338e-05, "loss": 0.0601, "step": 14580 }, { "grad_norm": 0.4313638508319855, "learning_rate": 9.528613289679622e-05, "loss": 0.0549, "step": 14590 }, { "grad_norm": 0.2981405258178711, "learning_rate": 9.52773683079971e-05, "loss": 0.0435, "step": 14600 }, { "grad_norm": 0.6486802101135254, "learning_rate": 9.526859598253407e-05, "loss": 0.0509, "step": 14610 }, { "grad_norm": 0.38626164197921753, "learning_rate": 9.525981592190609e-05, "loss": 0.046, "step": 14620 }, { "grad_norm": 0.5254563689231873, "learning_rate": 9.525102812761342e-05, "loss": 0.0436, "step": 14630 }, { "grad_norm": 0.4609138071537018, "learning_rate": 9.524223260115768e-05, "loss": 0.0458, "step": 14640 }, { "grad_norm": 0.4075065851211548, "learning_rate": 9.523342934404175e-05, "loss": 0.0474, "step": 14650 }, { "grad_norm": 0.32262563705444336, "learning_rate": 9.522461835776989e-05, "loss": 0.0413, "step": 14660 }, { "grad_norm": 0.5113041400909424, "learning_rate": 9.521579964384764e-05, "loss": 0.0588, "step": 14670 }, { "grad_norm": 0.3374783396720886, "learning_rate": 9.52069732037819e-05, "loss": 0.0406, "step": 14680 }, { "grad_norm": 0.4707943797111511, "learning_rate": 9.519813903908083e-05, "loss": 0.0446, "step": 14690 }, { "grad_norm": 0.5024896264076233, "learning_rate": 9.5189297151254e-05, "loss": 0.057, "step": 14700 }, { "grad_norm": 0.6219943165779114, "learning_rate": 9.518044754181218e-05, "loss": 0.0587, "step": 14710 }, { "grad_norm": 0.8526636958122253, "learning_rate": 9.51715902122676e-05, "loss": 0.0528, "step": 14720 }, { "grad_norm": 0.36108240485191345, "learning_rate": 9.516272516413368e-05, "loss": 0.0473, "step": 14730 }, { "grad_norm": 0.2906782925128937, "learning_rate": 9.515385239892525e-05, "loss": 0.0571, "step": 14740 }, { "grad_norm": 0.4819292724132538, "learning_rate": 9.514497191815839e-05, "loss": 0.056, "step": 14750 }, { "grad_norm": 0.5740299820899963, "learning_rate": 9.513608372335055e-05, "loss": 0.0561, "step": 14760 }, { "grad_norm": 0.6037437915802002, "learning_rate": 9.512718781602045e-05, "loss": 0.0531, "step": 14770 }, { "grad_norm": 0.4594002366065979, "learning_rate": 9.511828419768823e-05, "loss": 0.0515, "step": 14780 }, { "grad_norm": 0.3918030858039856, "learning_rate": 9.510937286987521e-05, "loss": 0.0383, "step": 14790 }, { "grad_norm": 0.6427013874053955, "learning_rate": 9.510045383410408e-05, "loss": 0.0529, "step": 14800 }, { "grad_norm": 0.38242390751838684, "learning_rate": 9.509152709189892e-05, "loss": 0.0446, "step": 14810 }, { "grad_norm": 0.2849431335926056, "learning_rate": 9.508259264478504e-05, "loss": 0.0538, "step": 14820 }, { "grad_norm": 0.49141883850097656, "learning_rate": 9.507365049428909e-05, "loss": 0.0618, "step": 14830 }, { "grad_norm": 0.708641767501831, "learning_rate": 9.506470064193902e-05, "loss": 0.0643, "step": 14840 }, { "grad_norm": 0.4484403729438782, "learning_rate": 9.505574308926414e-05, "loss": 0.0617, "step": 14850 }, { "grad_norm": 0.42658916115760803, "learning_rate": 9.504677783779505e-05, "loss": 0.0445, "step": 14860 }, { "grad_norm": 0.7566975951194763, "learning_rate": 9.503780488906365e-05, "loss": 0.0448, "step": 14870 }, { "grad_norm": 0.4233250617980957, "learning_rate": 9.502882424460319e-05, "loss": 0.0504, "step": 14880 }, { "grad_norm": 0.3779735565185547, "learning_rate": 9.501983590594821e-05, "loss": 0.0434, "step": 14890 }, { "grad_norm": 0.45819810032844543, "learning_rate": 9.501083987463455e-05, "loss": 0.0413, "step": 14900 }, { "grad_norm": 0.40303680300712585, "learning_rate": 9.500183615219942e-05, "loss": 0.046, "step": 14910 }, { "grad_norm": 0.46622711420059204, "learning_rate": 9.49928247401813e-05, "loss": 0.0438, "step": 14920 }, { "grad_norm": 0.6089573502540588, "learning_rate": 9.498380564011997e-05, "loss": 0.0431, "step": 14930 }, { "grad_norm": 0.3752000033855438, "learning_rate": 9.497477885355656e-05, "loss": 0.0516, "step": 14940 }, { "grad_norm": 0.4575439393520355, "learning_rate": 9.496574438203353e-05, "loss": 0.0635, "step": 14950 }, { "grad_norm": 0.39275792241096497, "learning_rate": 9.495670222709459e-05, "loss": 0.0515, "step": 14960 }, { "grad_norm": 0.44830209016799927, "learning_rate": 9.494765239028483e-05, "loss": 0.0488, "step": 14970 }, { "grad_norm": 0.3504635989665985, "learning_rate": 9.493859487315057e-05, "loss": 0.0419, "step": 14980 }, { "grad_norm": 0.4215793311595917, "learning_rate": 9.492952967723953e-05, "loss": 0.0437, "step": 14990 }, { "grad_norm": 0.511399507522583, "learning_rate": 9.492045680410068e-05, "loss": 0.0502, "step": 15000 }, { "grad_norm": 0.38199031352996826, "learning_rate": 9.491137625528436e-05, "loss": 0.0439, "step": 15010 }, { "grad_norm": 0.535870373249054, "learning_rate": 9.490228803234215e-05, "loss": 0.0446, "step": 15020 }, { "grad_norm": 0.44968974590301514, "learning_rate": 9.489319213682701e-05, "loss": 0.044, "step": 15030 }, { "grad_norm": 0.2962111532688141, "learning_rate": 9.488408857029316e-05, "loss": 0.0544, "step": 15040 }, { "grad_norm": 0.33721601963043213, "learning_rate": 9.487497733429616e-05, "loss": 0.0546, "step": 15050 }, { "grad_norm": 0.3886985778808594, "learning_rate": 9.486585843039286e-05, "loss": 0.0583, "step": 15060 }, { "grad_norm": 0.49429237842559814, "learning_rate": 9.485673186014143e-05, "loss": 0.0523, "step": 15070 }, { "grad_norm": 0.5129354000091553, "learning_rate": 9.484759762510137e-05, "loss": 0.0609, "step": 15080 }, { "grad_norm": 0.44734570384025574, "learning_rate": 9.483845572683346e-05, "loss": 0.0514, "step": 15090 }, { "grad_norm": 0.4252418875694275, "learning_rate": 9.48293061668998e-05, "loss": 0.0478, "step": 15100 }, { "grad_norm": 0.3664085268974304, "learning_rate": 9.48201489468638e-05, "loss": 0.0507, "step": 15110 }, { "grad_norm": 0.41223129630088806, "learning_rate": 9.481098406829016e-05, "loss": 0.0464, "step": 15120 }, { "grad_norm": 0.6134036183357239, "learning_rate": 9.480181153274495e-05, "loss": 0.0451, "step": 15130 }, { "grad_norm": 0.547081708908081, "learning_rate": 9.479263134179548e-05, "loss": 0.0395, "step": 15140 }, { "grad_norm": 0.5162649154663086, "learning_rate": 9.478344349701039e-05, "loss": 0.053, "step": 15150 }, { "grad_norm": 0.42325836420059204, "learning_rate": 9.477424799995964e-05, "loss": 0.0526, "step": 15160 }, { "grad_norm": 0.3607238233089447, "learning_rate": 9.476504485221448e-05, "loss": 0.0495, "step": 15170 }, { "grad_norm": 0.44021075963974, "learning_rate": 9.475583405534748e-05, "loss": 0.049, "step": 15180 }, { "grad_norm": 0.4325045049190521, "learning_rate": 9.474661561093251e-05, "loss": 0.0497, "step": 15190 }, { "grad_norm": 0.43776437640190125, "learning_rate": 9.473738952054478e-05, "loss": 0.056, "step": 15200 }, { "grad_norm": 0.2953924238681793, "learning_rate": 9.472815578576073e-05, "loss": 0.0453, "step": 15210 }, { "grad_norm": 0.5967522263526917, "learning_rate": 9.471891440815817e-05, "loss": 0.0511, "step": 15220 }, { "grad_norm": 0.4561527669429779, "learning_rate": 9.470966538931621e-05, "loss": 0.04, "step": 15230 }, { "grad_norm": 0.2938949465751648, "learning_rate": 9.470040873081525e-05, "loss": 0.0402, "step": 15240 }, { "grad_norm": 0.4484460651874542, "learning_rate": 9.469114443423698e-05, "loss": 0.0562, "step": 15250 }, { "grad_norm": 0.4168435335159302, "learning_rate": 9.468187250116445e-05, "loss": 0.0472, "step": 15260 }, { "grad_norm": 0.5859142541885376, "learning_rate": 9.467259293318197e-05, "loss": 0.0456, "step": 15270 }, { "grad_norm": 0.27922770380973816, "learning_rate": 9.466330573187514e-05, "loss": 0.0397, "step": 15280 }, { "grad_norm": 0.3605108857154846, "learning_rate": 9.46540108988309e-05, "loss": 0.0607, "step": 15290 }, { "grad_norm": 0.4694368541240692, "learning_rate": 9.46447084356375e-05, "loss": 0.0541, "step": 15300 }, { "grad_norm": 0.2608813941478729, "learning_rate": 9.463539834388447e-05, "loss": 0.0497, "step": 15310 }, { "grad_norm": 0.4163062870502472, "learning_rate": 9.462608062516263e-05, "loss": 0.0357, "step": 15320 }, { "grad_norm": 0.45465216040611267, "learning_rate": 9.461675528106413e-05, "loss": 0.0389, "step": 15330 }, { "grad_norm": 0.3661433160305023, "learning_rate": 9.460742231318244e-05, "loss": 0.0462, "step": 15340 }, { "grad_norm": 0.7345144152641296, "learning_rate": 9.459808172311229e-05, "loss": 0.0607, "step": 15350 }, { "grad_norm": 0.35217568278312683, "learning_rate": 9.458873351244972e-05, "loss": 0.0409, "step": 15360 }, { "grad_norm": 0.43929523229599, "learning_rate": 9.457937768279211e-05, "loss": 0.0456, "step": 15370 }, { "grad_norm": 0.4293931722640991, "learning_rate": 9.45700142357381e-05, "loss": 0.0435, "step": 15380 }, { "grad_norm": 0.3977937400341034, "learning_rate": 9.456064317288765e-05, "loss": 0.0405, "step": 15390 }, { "grad_norm": 0.4466230571269989, "learning_rate": 9.455126449584201e-05, "loss": 0.0455, "step": 15400 }, { "grad_norm": 0.6813721060752869, "learning_rate": 9.454187820620375e-05, "loss": 0.0357, "step": 15410 }, { "grad_norm": 0.5236808061599731, "learning_rate": 9.453248430557673e-05, "loss": 0.0541, "step": 15420 }, { "grad_norm": 0.5990610718727112, "learning_rate": 9.452308279556611e-05, "loss": 0.0653, "step": 15430 }, { "grad_norm": 0.41922420263290405, "learning_rate": 9.451367367777835e-05, "loss": 0.0421, "step": 15440 }, { "grad_norm": 0.37595370411872864, "learning_rate": 9.450425695382122e-05, "loss": 0.0515, "step": 15450 }, { "grad_norm": 0.43797215819358826, "learning_rate": 9.449483262530375e-05, "loss": 0.0497, "step": 15460 }, { "grad_norm": 0.4604327380657196, "learning_rate": 9.448540069383633e-05, "loss": 0.0431, "step": 15470 }, { "grad_norm": 0.39495185017585754, "learning_rate": 9.447596116103061e-05, "loss": 0.0457, "step": 15480 }, { "grad_norm": 0.38353675603866577, "learning_rate": 9.446651402849955e-05, "loss": 0.046, "step": 15490 }, { "grad_norm": 0.3438076078891754, "learning_rate": 9.44570592978574e-05, "loss": 0.0449, "step": 15500 }, { "grad_norm": 0.3794802725315094, "learning_rate": 9.444759697071972e-05, "loss": 0.0478, "step": 15510 }, { "grad_norm": 0.4572095572948456, "learning_rate": 9.443812704870336e-05, "loss": 0.0413, "step": 15520 }, { "grad_norm": 0.3932870626449585, "learning_rate": 9.442864953342649e-05, "loss": 0.0435, "step": 15530 }, { "grad_norm": 0.33670246601104736, "learning_rate": 9.441916442650852e-05, "loss": 0.0475, "step": 15540 }, { "grad_norm": 0.316232830286026, "learning_rate": 9.440967172957023e-05, "loss": 0.0419, "step": 15550 }, { "grad_norm": 0.4897022247314453, "learning_rate": 9.440017144423364e-05, "loss": 0.0508, "step": 15560 }, { "grad_norm": 0.4668068587779999, "learning_rate": 9.439066357212209e-05, "loss": 0.0491, "step": 15570 }, { "grad_norm": 0.38890960812568665, "learning_rate": 9.438114811486022e-05, "loss": 0.0544, "step": 15580 }, { "grad_norm": 0.2989889085292816, "learning_rate": 9.4371625074074e-05, "loss": 0.0387, "step": 15590 }, { "grad_norm": 0.578540563583374, "learning_rate": 9.436209445139059e-05, "loss": 0.0422, "step": 15600 }, { "grad_norm": 0.356240838766098, "learning_rate": 9.435255624843855e-05, "loss": 0.0473, "step": 15610 }, { "grad_norm": 0.320298969745636, "learning_rate": 9.43430104668477e-05, "loss": 0.0374, "step": 15620 }, { "grad_norm": 0.4710850417613983, "learning_rate": 9.433345710824914e-05, "loss": 0.0459, "step": 15630 }, { "grad_norm": 0.3611133396625519, "learning_rate": 9.432389617427529e-05, "loss": 0.0426, "step": 15640 }, { "grad_norm": 0.4175347685813904, "learning_rate": 9.431432766655984e-05, "loss": 0.0416, "step": 15650 }, { "grad_norm": 0.5860666632652283, "learning_rate": 9.430475158673778e-05, "loss": 0.0492, "step": 15660 }, { "grad_norm": 0.4531394839286804, "learning_rate": 9.429516793644542e-05, "loss": 0.0393, "step": 15670 }, { "grad_norm": 0.36013537645339966, "learning_rate": 9.428557671732034e-05, "loss": 0.0333, "step": 15680 }, { "grad_norm": 0.3083541989326477, "learning_rate": 9.42759779310014e-05, "loss": 0.041, "step": 15690 }, { "grad_norm": 0.6993846893310547, "learning_rate": 9.426637157912879e-05, "loss": 0.0485, "step": 15700 }, { "grad_norm": 0.4884318709373474, "learning_rate": 9.425675766334397e-05, "loss": 0.0446, "step": 15710 }, { "grad_norm": 0.4280475974082947, "learning_rate": 9.424713618528968e-05, "loss": 0.0397, "step": 15720 }, { "grad_norm": 0.602022111415863, "learning_rate": 9.423750714661e-05, "loss": 0.0567, "step": 15730 }, { "grad_norm": 0.40581247210502625, "learning_rate": 9.422787054895022e-05, "loss": 0.0458, "step": 15740 }, { "grad_norm": 0.376817524433136, "learning_rate": 9.4218226393957e-05, "loss": 0.0593, "step": 15750 }, { "grad_norm": 0.4063802659511566, "learning_rate": 9.420857468327828e-05, "loss": 0.0521, "step": 15760 }, { "grad_norm": 0.4773770570755005, "learning_rate": 9.419891541856323e-05, "loss": 0.0487, "step": 15770 }, { "grad_norm": 0.4825810492038727, "learning_rate": 9.41892486014624e-05, "loss": 0.0468, "step": 15780 }, { "grad_norm": 0.49994009733200073, "learning_rate": 9.417957423362756e-05, "loss": 0.0514, "step": 15790 }, { "grad_norm": 0.42212510108947754, "learning_rate": 9.416989231671178e-05, "loss": 0.0477, "step": 15800 }, { "grad_norm": 0.4929404854774475, "learning_rate": 9.416020285236946e-05, "loss": 0.0481, "step": 15810 }, { "grad_norm": 0.35948166251182556, "learning_rate": 9.415050584225626e-05, "loss": 0.0444, "step": 15820 }, { "grad_norm": 0.4937751591205597, "learning_rate": 9.414080128802914e-05, "loss": 0.0499, "step": 15830 }, { "grad_norm": 0.40695273876190186, "learning_rate": 9.413108919134632e-05, "loss": 0.0534, "step": 15840 }, { "grad_norm": 0.5528419017791748, "learning_rate": 9.412136955386734e-05, "loss": 0.0463, "step": 15850 }, { "grad_norm": 0.37761566042900085, "learning_rate": 9.411164237725303e-05, "loss": 0.0441, "step": 15860 }, { "grad_norm": 0.5408079624176025, "learning_rate": 9.41019076631655e-05, "loss": 0.0499, "step": 15870 }, { "grad_norm": 0.33010992407798767, "learning_rate": 9.409216541326815e-05, "loss": 0.0475, "step": 15880 }, { "grad_norm": 0.3406860828399658, "learning_rate": 9.408241562922564e-05, "loss": 0.0488, "step": 15890 }, { "grad_norm": 0.4023871123790741, "learning_rate": 9.407265831270395e-05, "loss": 0.0436, "step": 15900 }, { "grad_norm": 0.5719010829925537, "learning_rate": 9.406289346537035e-05, "loss": 0.0412, "step": 15910 }, { "grad_norm": 0.4783810079097748, "learning_rate": 9.405312108889339e-05, "loss": 0.044, "step": 15920 }, { "grad_norm": 0.5384085178375244, "learning_rate": 9.404334118494288e-05, "loss": 0.047, "step": 15930 }, { "grad_norm": 0.33166348934173584, "learning_rate": 9.403355375518995e-05, "loss": 0.0398, "step": 15940 }, { "grad_norm": 0.4312153160572052, "learning_rate": 9.4023758801307e-05, "loss": 0.0437, "step": 15950 }, { "grad_norm": 0.40556323528289795, "learning_rate": 9.401395632496774e-05, "loss": 0.041, "step": 15960 }, { "grad_norm": 0.46570712327957153, "learning_rate": 9.400414632784711e-05, "loss": 0.0511, "step": 15970 }, { "grad_norm": 0.38383597135543823, "learning_rate": 9.39943288116214e-05, "loss": 0.0451, "step": 15980 }, { "grad_norm": 0.47791972756385803, "learning_rate": 9.398450377796815e-05, "loss": 0.0486, "step": 15990 }, { "grad_norm": 0.36660581827163696, "learning_rate": 9.397467122856616e-05, "loss": 0.0435, "step": 16000 }, { "grad_norm": 0.4147270619869232, "learning_rate": 9.396483116509558e-05, "loss": 0.0463, "step": 16010 }, { "grad_norm": 0.37275806069374084, "learning_rate": 9.39549835892378e-05, "loss": 0.0469, "step": 16020 }, { "grad_norm": 0.3332223892211914, "learning_rate": 9.39451285026755e-05, "loss": 0.0474, "step": 16030 }, { "grad_norm": 0.5225622653961182, "learning_rate": 9.393526590709262e-05, "loss": 0.0453, "step": 16040 }, { "grad_norm": 0.4741164743900299, "learning_rate": 9.392539580417444e-05, "loss": 0.0414, "step": 16050 }, { "grad_norm": 0.333119660615921, "learning_rate": 9.391551819560747e-05, "loss": 0.044, "step": 16060 }, { "grad_norm": 0.4495294690132141, "learning_rate": 9.390563308307955e-05, "loss": 0.045, "step": 16070 }, { "grad_norm": 0.5093159675598145, "learning_rate": 9.389574046827974e-05, "loss": 0.0441, "step": 16080 }, { "grad_norm": 0.4108915328979492, "learning_rate": 9.388584035289845e-05, "loss": 0.0508, "step": 16090 }, { "grad_norm": 0.3937346041202545, "learning_rate": 9.387593273862732e-05, "loss": 0.0509, "step": 16100 }, { "grad_norm": 0.37609559297561646, "learning_rate": 9.386601762715929e-05, "loss": 0.0546, "step": 16110 }, { "grad_norm": 0.5017209649085999, "learning_rate": 9.38560950201886e-05, "loss": 0.0556, "step": 16120 }, { "grad_norm": 0.6117917895317078, "learning_rate": 9.384616491941071e-05, "loss": 0.0566, "step": 16130 }, { "grad_norm": 0.6171096563339233, "learning_rate": 9.383622732652245e-05, "loss": 0.0501, "step": 16140 }, { "grad_norm": 0.3556797206401825, "learning_rate": 9.382628224322187e-05, "loss": 0.056, "step": 16150 }, { "grad_norm": 0.2979637086391449, "learning_rate": 9.381632967120829e-05, "loss": 0.0517, "step": 16160 }, { "grad_norm": 0.5331001281738281, "learning_rate": 9.380636961218235e-05, "loss": 0.0414, "step": 16170 }, { "grad_norm": 0.40995270013809204, "learning_rate": 9.379640206784597e-05, "loss": 0.0407, "step": 16180 }, { "grad_norm": 0.4276105463504791, "learning_rate": 9.378642703990229e-05, "loss": 0.0371, "step": 16190 }, { "grad_norm": 0.38809964060783386, "learning_rate": 9.37764445300558e-05, "loss": 0.0466, "step": 16200 }, { "grad_norm": 0.22549153864383698, "learning_rate": 9.376645454001222e-05, "loss": 0.0488, "step": 16210 }, { "grad_norm": 0.3580251634120941, "learning_rate": 9.375645707147858e-05, "loss": 0.0387, "step": 16220 }, { "grad_norm": 0.37482956051826477, "learning_rate": 9.374645212616316e-05, "loss": 0.05, "step": 16230 }, { "grad_norm": 0.5508458018302917, "learning_rate": 9.373643970577555e-05, "loss": 0.0442, "step": 16240 }, { "grad_norm": 0.5792708396911621, "learning_rate": 9.372641981202659e-05, "loss": 0.0429, "step": 16250 }, { "grad_norm": 0.5211492776870728, "learning_rate": 9.37163924466284e-05, "loss": 0.0552, "step": 16260 }, { "grad_norm": 0.48471251130104065, "learning_rate": 9.370635761129438e-05, "loss": 0.0455, "step": 16270 }, { "grad_norm": 0.5093134045600891, "learning_rate": 9.36963153077392e-05, "loss": 0.0471, "step": 16280 }, { "grad_norm": 0.38388243317604065, "learning_rate": 9.368626553767888e-05, "loss": 0.0424, "step": 16290 }, { "grad_norm": 0.41685351729393005, "learning_rate": 9.367620830283057e-05, "loss": 0.0389, "step": 16300 }, { "grad_norm": 0.40004438161849976, "learning_rate": 9.366614360491281e-05, "loss": 0.0448, "step": 16310 }, { "grad_norm": 0.4242977201938629, "learning_rate": 9.365607144564539e-05, "loss": 0.0437, "step": 16320 }, { "grad_norm": 0.46907100081443787, "learning_rate": 9.364599182674934e-05, "loss": 0.0499, "step": 16330 }, { "grad_norm": 0.4003777503967285, "learning_rate": 9.3635904749947e-05, "loss": 0.0401, "step": 16340 }, { "grad_norm": 0.3946529030799866, "learning_rate": 9.362581021696202e-05, "loss": 0.0435, "step": 16350 }, { "grad_norm": 0.386516273021698, "learning_rate": 9.361570822951921e-05, "loss": 0.0484, "step": 16360 }, { "grad_norm": 0.3232499361038208, "learning_rate": 9.360559878934476e-05, "loss": 0.0486, "step": 16370 }, { "grad_norm": 0.34943628311157227, "learning_rate": 9.359548189816611e-05, "loss": 0.0391, "step": 16380 }, { "grad_norm": 0.40323251485824585, "learning_rate": 9.358535755771193e-05, "loss": 0.0502, "step": 16390 }, { "grad_norm": 0.5477654337882996, "learning_rate": 9.357522576971221e-05, "loss": 0.0491, "step": 16400 }, { "grad_norm": 0.4684268534183502, "learning_rate": 9.356508653589819e-05, "loss": 0.0475, "step": 16410 }, { "grad_norm": 0.43331092596054077, "learning_rate": 9.355493985800237e-05, "loss": 0.0466, "step": 16420 }, { "grad_norm": 0.4726651906967163, "learning_rate": 9.354478573775857e-05, "loss": 0.0411, "step": 16430 }, { "grad_norm": 0.32171446084976196, "learning_rate": 9.353462417690186e-05, "loss": 0.0415, "step": 16440 }, { "grad_norm": 0.4531373381614685, "learning_rate": 9.352445517716853e-05, "loss": 0.0422, "step": 16450 }, { "grad_norm": 0.37373846769332886, "learning_rate": 9.351427874029621e-05, "loss": 0.0445, "step": 16460 }, { "grad_norm": 0.5218920111656189, "learning_rate": 9.350409486802379e-05, "loss": 0.0445, "step": 16470 }, { "grad_norm": 0.4568745493888855, "learning_rate": 9.349390356209138e-05, "loss": 0.0462, "step": 16480 }, { "grad_norm": 0.39631232619285583, "learning_rate": 9.348370482424042e-05, "loss": 0.0405, "step": 16490 }, { "grad_norm": 0.5810245275497437, "learning_rate": 9.347349865621357e-05, "loss": 0.0393, "step": 16500 }, { "grad_norm": 0.3912624716758728, "learning_rate": 9.346328505975481e-05, "loss": 0.0428, "step": 16510 }, { "grad_norm": 0.42256423830986023, "learning_rate": 9.345306403660936e-05, "loss": 0.0496, "step": 16520 }, { "grad_norm": 0.5906659364700317, "learning_rate": 9.344283558852371e-05, "loss": 0.0508, "step": 16530 }, { "grad_norm": 0.4335789084434509, "learning_rate": 9.343259971724563e-05, "loss": 0.0466, "step": 16540 }, { "grad_norm": 0.4801372289657593, "learning_rate": 9.342235642452413e-05, "loss": 0.0521, "step": 16550 }, { "grad_norm": 0.463417649269104, "learning_rate": 9.341210571210954e-05, "loss": 0.0463, "step": 16560 }, { "grad_norm": 0.368213415145874, "learning_rate": 9.340184758175338e-05, "loss": 0.0428, "step": 16570 }, { "grad_norm": 0.3471206724643707, "learning_rate": 9.339158203520854e-05, "loss": 0.0534, "step": 16580 }, { "grad_norm": 0.5623409748077393, "learning_rate": 9.338130907422908e-05, "loss": 0.0394, "step": 16590 }, { "grad_norm": 0.4684542715549469, "learning_rate": 9.337102870057037e-05, "loss": 0.0362, "step": 16600 }, { "grad_norm": 0.3163854777812958, "learning_rate": 9.336074091598907e-05, "loss": 0.0369, "step": 16610 }, { "grad_norm": 0.2960915267467499, "learning_rate": 9.335044572224306e-05, "loss": 0.0355, "step": 16620 }, { "grad_norm": 0.5417312979698181, "learning_rate": 9.334014312109151e-05, "loss": 0.0363, "step": 16630 }, { "grad_norm": 0.5392457246780396, "learning_rate": 9.332983311429486e-05, "loss": 0.0503, "step": 16640 }, { "grad_norm": 0.4053120017051697, "learning_rate": 9.33195157036148e-05, "loss": 0.0485, "step": 16650 }, { "grad_norm": 0.3438643217086792, "learning_rate": 9.330919089081432e-05, "loss": 0.0404, "step": 16660 }, { "grad_norm": 0.35402563214302063, "learning_rate": 9.32988586776576e-05, "loss": 0.0552, "step": 16670 }, { "grad_norm": 0.3576315939426422, "learning_rate": 9.328851906591016e-05, "loss": 0.0439, "step": 16680 }, { "grad_norm": 0.41900140047073364, "learning_rate": 9.327817205733875e-05, "loss": 0.0505, "step": 16690 }, { "grad_norm": 0.5303101539611816, "learning_rate": 9.326781765371142e-05, "loss": 0.0422, "step": 16700 }, { "grad_norm": 0.525421679019928, "learning_rate": 9.325745585679741e-05, "loss": 0.0429, "step": 16710 }, { "grad_norm": 0.41432878375053406, "learning_rate": 9.32470866683673e-05, "loss": 0.0593, "step": 16720 }, { "grad_norm": 0.5313615798950195, "learning_rate": 9.323671009019288e-05, "loss": 0.0532, "step": 16730 }, { "grad_norm": 0.3480933904647827, "learning_rate": 9.322632612404725e-05, "loss": 0.0397, "step": 16740 }, { "grad_norm": 0.37634149193763733, "learning_rate": 9.321593477170471e-05, "loss": 0.0536, "step": 16750 }, { "grad_norm": 0.29678988456726074, "learning_rate": 9.320553603494088e-05, "loss": 0.0357, "step": 16760 }, { "grad_norm": 0.4601637125015259, "learning_rate": 9.319512991553261e-05, "loss": 0.0408, "step": 16770 }, { "grad_norm": 0.46597597002983093, "learning_rate": 9.318471641525803e-05, "loss": 0.0449, "step": 16780 }, { "grad_norm": 0.3716297447681427, "learning_rate": 9.317429553589652e-05, "loss": 0.0429, "step": 16790 }, { "grad_norm": 0.49249547719955444, "learning_rate": 9.316386727922873e-05, "loss": 0.0501, "step": 16800 }, { "grad_norm": 0.4212192893028259, "learning_rate": 9.315343164703656e-05, "loss": 0.0456, "step": 16810 }, { "grad_norm": 0.51121985912323, "learning_rate": 9.314298864110316e-05, "loss": 0.0401, "step": 16820 }, { "grad_norm": 0.3165232241153717, "learning_rate": 9.313253826321295e-05, "loss": 0.0391, "step": 16830 }, { "grad_norm": 0.32726815342903137, "learning_rate": 9.312208051515165e-05, "loss": 0.036, "step": 16840 }, { "grad_norm": 0.348410040140152, "learning_rate": 9.311161539870618e-05, "loss": 0.0369, "step": 16850 }, { "grad_norm": 0.3860686719417572, "learning_rate": 9.310114291566474e-05, "loss": 0.0471, "step": 16860 }, { "grad_norm": 0.3503755033016205, "learning_rate": 9.309066306781679e-05, "loss": 0.0401, "step": 16870 }, { "grad_norm": 0.39629119634628296, "learning_rate": 9.308017585695306e-05, "loss": 0.0525, "step": 16880 }, { "grad_norm": 0.37001973390579224, "learning_rate": 9.306968128486552e-05, "loss": 0.0468, "step": 16890 }, { "grad_norm": 0.3773685395717621, "learning_rate": 9.30591793533474e-05, "loss": 0.0393, "step": 16900 }, { "grad_norm": 0.4694444537162781, "learning_rate": 9.304867006419321e-05, "loss": 0.0413, "step": 16910 }, { "grad_norm": 0.4691469669342041, "learning_rate": 9.303815341919868e-05, "loss": 0.0512, "step": 16920 }, { "grad_norm": 0.46549665927886963, "learning_rate": 9.302762942016084e-05, "loss": 0.0506, "step": 16930 }, { "grad_norm": 0.3937391936779022, "learning_rate": 9.301709806887792e-05, "loss": 0.0417, "step": 16940 }, { "grad_norm": 0.44971203804016113, "learning_rate": 9.300655936714948e-05, "loss": 0.0672, "step": 16950 }, { "grad_norm": 0.5782698392868042, "learning_rate": 9.299601331677627e-05, "loss": 0.0485, "step": 16960 }, { "grad_norm": 0.42550361156463623, "learning_rate": 9.298545991956033e-05, "loss": 0.0495, "step": 16970 }, { "grad_norm": 0.31684330105781555, "learning_rate": 9.297489917730493e-05, "loss": 0.0455, "step": 16980 }, { "grad_norm": 0.4957515001296997, "learning_rate": 9.296433109181464e-05, "loss": 0.0539, "step": 16990 }, { "grad_norm": 0.5215907096862793, "learning_rate": 9.295375566489523e-05, "loss": 0.0443, "step": 17000 }, { "grad_norm": 0.4419822692871094, "learning_rate": 9.294317289835379e-05, "loss": 0.0462, "step": 17010 }, { "grad_norm": 0.48849332332611084, "learning_rate": 9.293258279399859e-05, "loss": 0.0518, "step": 17020 }, { "grad_norm": 0.2944369912147522, "learning_rate": 9.292198535363919e-05, "loss": 0.0363, "step": 17030 }, { "grad_norm": 0.3432421386241913, "learning_rate": 9.291138057908641e-05, "loss": 0.0369, "step": 17040 }, { "grad_norm": 0.4709196388721466, "learning_rate": 9.290076847215234e-05, "loss": 0.0386, "step": 17050 }, { "grad_norm": 0.5345579981803894, "learning_rate": 9.289014903465025e-05, "loss": 0.0538, "step": 17060 }, { "grad_norm": 0.4095999002456665, "learning_rate": 9.287952226839475e-05, "loss": 0.0376, "step": 17070 }, { "grad_norm": 0.4417373239994049, "learning_rate": 9.286888817520164e-05, "loss": 0.0349, "step": 17080 }, { "grad_norm": 0.42557549476623535, "learning_rate": 9.285824675688803e-05, "loss": 0.0398, "step": 17090 }, { "grad_norm": 0.44706282019615173, "learning_rate": 9.28475980152722e-05, "loss": 0.0482, "step": 17100 }, { "grad_norm": 0.44073671102523804, "learning_rate": 9.283694195217379e-05, "loss": 0.0521, "step": 17110 }, { "grad_norm": 0.5158327221870422, "learning_rate": 9.282627856941356e-05, "loss": 0.0456, "step": 17120 }, { "grad_norm": 0.405611127614975, "learning_rate": 9.281560786881363e-05, "loss": 0.0418, "step": 17130 }, { "grad_norm": 0.48653116822242737, "learning_rate": 9.280492985219733e-05, "loss": 0.0338, "step": 17140 }, { "grad_norm": 0.4924352169036865, "learning_rate": 9.279424452138924e-05, "loss": 0.0441, "step": 17150 }, { "grad_norm": 0.8840945959091187, "learning_rate": 9.278355187821517e-05, "loss": 0.0436, "step": 17160 }, { "grad_norm": 0.38716796040534973, "learning_rate": 9.277285192450224e-05, "loss": 0.0446, "step": 17170 }, { "grad_norm": 0.47929322719573975, "learning_rate": 9.276214466207875e-05, "loss": 0.0472, "step": 17180 }, { "grad_norm": 0.502765417098999, "learning_rate": 9.275143009277427e-05, "loss": 0.0466, "step": 17190 }, { "grad_norm": 0.40588119626045227, "learning_rate": 9.274070821841964e-05, "loss": 0.0441, "step": 17200 }, { "grad_norm": 0.36497098207473755, "learning_rate": 9.272997904084696e-05, "loss": 0.0429, "step": 17210 }, { "grad_norm": 0.4086656868457794, "learning_rate": 9.271924256188951e-05, "loss": 0.0533, "step": 17220 }, { "grad_norm": 0.46439194679260254, "learning_rate": 9.270849878338189e-05, "loss": 0.0481, "step": 17230 }, { "grad_norm": 0.5449007153511047, "learning_rate": 9.269774770715991e-05, "loss": 0.047, "step": 17240 }, { "grad_norm": 0.48318198323249817, "learning_rate": 9.268698933506061e-05, "loss": 0.0576, "step": 17250 }, { "grad_norm": 0.39047402143478394, "learning_rate": 9.267622366892235e-05, "loss": 0.05, "step": 17260 }, { "grad_norm": 0.2680776119232178, "learning_rate": 9.266545071058465e-05, "loss": 0.0374, "step": 17270 }, { "grad_norm": 0.32050105929374695, "learning_rate": 9.265467046188833e-05, "loss": 0.0383, "step": 17280 }, { "grad_norm": 0.31036069989204407, "learning_rate": 9.264388292467543e-05, "loss": 0.0445, "step": 17290 }, { "grad_norm": 0.4287029504776001, "learning_rate": 9.263308810078926e-05, "loss": 0.043, "step": 17300 }, { "grad_norm": 0.41581034660339355, "learning_rate": 9.262228599207434e-05, "loss": 0.0402, "step": 17310 }, { "grad_norm": 0.3302435576915741, "learning_rate": 9.261147660037647e-05, "loss": 0.0398, "step": 17320 }, { "grad_norm": 0.4548039138317108, "learning_rate": 9.26006599275427e-05, "loss": 0.0396, "step": 17330 }, { "grad_norm": 0.3051582872867584, "learning_rate": 9.258983597542124e-05, "loss": 0.0374, "step": 17340 }, { "grad_norm": 0.41067996621131897, "learning_rate": 9.257900474586167e-05, "loss": 0.0369, "step": 17350 }, { "grad_norm": 0.4629344642162323, "learning_rate": 9.256816624071471e-05, "loss": 0.0484, "step": 17360 }, { "grad_norm": 0.4121609926223755, "learning_rate": 9.25573204618324e-05, "loss": 0.045, "step": 17370 }, { "grad_norm": 0.45741578936576843, "learning_rate": 9.254646741106796e-05, "loss": 0.044, "step": 17380 }, { "grad_norm": 0.41777968406677246, "learning_rate": 9.253560709027589e-05, "loss": 0.0401, "step": 17390 }, { "grad_norm": 0.4700753390789032, "learning_rate": 9.252473950131192e-05, "loss": 0.0446, "step": 17400 }, { "grad_norm": 0.38150545954704285, "learning_rate": 9.251386464603302e-05, "loss": 0.0353, "step": 17410 }, { "grad_norm": 0.43364208936691284, "learning_rate": 9.250298252629741e-05, "loss": 0.0434, "step": 17420 }, { "grad_norm": 0.4611862897872925, "learning_rate": 9.249209314396454e-05, "loss": 0.0368, "step": 17430 }, { "grad_norm": 0.47503960132598877, "learning_rate": 9.248119650089513e-05, "loss": 0.0445, "step": 17440 }, { "grad_norm": 0.41352418065071106, "learning_rate": 9.247029259895108e-05, "loss": 0.0436, "step": 17450 }, { "grad_norm": 0.3341122269630432, "learning_rate": 9.24593814399956e-05, "loss": 0.0435, "step": 17460 }, { "grad_norm": 0.2946796119213104, "learning_rate": 9.244846302589309e-05, "loss": 0.0368, "step": 17470 }, { "grad_norm": 0.37971365451812744, "learning_rate": 9.243753735850923e-05, "loss": 0.0442, "step": 17480 }, { "grad_norm": 0.2993837296962738, "learning_rate": 9.24266044397109e-05, "loss": 0.0341, "step": 17490 }, { "grad_norm": 0.30283239483833313, "learning_rate": 9.241566427136624e-05, "loss": 0.035, "step": 17500 }, { "grad_norm": 0.4584568440914154, "learning_rate": 9.240471685534463e-05, "loss": 0.0421, "step": 17510 }, { "grad_norm": 0.6723180413246155, "learning_rate": 9.239376219351667e-05, "loss": 0.0477, "step": 17520 }, { "grad_norm": 0.34304967522621155, "learning_rate": 9.238280028775425e-05, "loss": 0.0454, "step": 17530 }, { "grad_norm": 0.47120124101638794, "learning_rate": 9.237183113993041e-05, "loss": 0.0458, "step": 17540 }, { "grad_norm": 0.4429546892642975, "learning_rate": 9.236085475191952e-05, "loss": 0.0389, "step": 17550 }, { "grad_norm": 0.48164355754852295, "learning_rate": 9.234987112559709e-05, "loss": 0.0689, "step": 17560 }, { "grad_norm": 0.5477330088615417, "learning_rate": 9.233888026283999e-05, "loss": 0.0511, "step": 17570 }, { "grad_norm": 0.3966647982597351, "learning_rate": 9.232788216552619e-05, "loss": 0.0493, "step": 17580 }, { "grad_norm": 0.5895764827728271, "learning_rate": 9.231687683553502e-05, "loss": 0.0518, "step": 17590 }, { "grad_norm": 0.366568922996521, "learning_rate": 9.230586427474698e-05, "loss": 0.0548, "step": 17600 }, { "grad_norm": 0.47932618856430054, "learning_rate": 9.229484448504379e-05, "loss": 0.0538, "step": 17610 }, { "grad_norm": 0.355192631483078, "learning_rate": 9.228381746830843e-05, "loss": 0.0447, "step": 17620 }, { "grad_norm": 0.37120288610458374, "learning_rate": 9.227278322642514e-05, "loss": 0.0367, "step": 17630 }, { "grad_norm": 0.36649543046951294, "learning_rate": 9.226174176127937e-05, "loss": 0.0379, "step": 17640 }, { "grad_norm": 0.4040161371231079, "learning_rate": 9.22506930747578e-05, "loss": 0.0384, "step": 17650 }, { "grad_norm": 0.6122265458106995, "learning_rate": 9.223963716874831e-05, "loss": 0.0504, "step": 17660 }, { "grad_norm": 0.44746339321136475, "learning_rate": 9.222857404514012e-05, "loss": 0.0471, "step": 17670 }, { "grad_norm": 0.47945791482925415, "learning_rate": 9.221750370582355e-05, "loss": 0.0453, "step": 17680 }, { "grad_norm": 0.3783722221851349, "learning_rate": 9.220642615269028e-05, "loss": 0.0737, "step": 17690 }, { "grad_norm": 0.36329954862594604, "learning_rate": 9.219534138763311e-05, "loss": 0.0528, "step": 17700 }, { "grad_norm": 0.513633668422699, "learning_rate": 9.218424941254613e-05, "loss": 0.0529, "step": 17710 }, { "grad_norm": 0.46655595302581787, "learning_rate": 9.217315022932468e-05, "loss": 0.0475, "step": 17720 }, { "grad_norm": 0.5464370846748352, "learning_rate": 9.216204383986528e-05, "loss": 0.0441, "step": 17730 }, { "grad_norm": 0.3927620053291321, "learning_rate": 9.215093024606574e-05, "loss": 0.0511, "step": 17740 }, { "grad_norm": 0.4038712978363037, "learning_rate": 9.213980944982506e-05, "loss": 0.0644, "step": 17750 }, { "grad_norm": 0.473486989736557, "learning_rate": 9.212868145304346e-05, "loss": 0.0469, "step": 17760 }, { "grad_norm": 0.33199450373649597, "learning_rate": 9.211754625762241e-05, "loss": 0.0411, "step": 17770 }, { "grad_norm": 0.3388751745223999, "learning_rate": 9.210640386546463e-05, "loss": 0.0515, "step": 17780 }, { "grad_norm": 0.3312314748764038, "learning_rate": 9.209525427847405e-05, "loss": 0.0357, "step": 17790 }, { "grad_norm": 0.3805021047592163, "learning_rate": 9.208409749855583e-05, "loss": 0.035, "step": 17800 }, { "grad_norm": 0.4814170002937317, "learning_rate": 9.207293352761633e-05, "loss": 0.0467, "step": 17810 }, { "grad_norm": 0.48299068212509155, "learning_rate": 9.206176236756319e-05, "loss": 0.0444, "step": 17820 }, { "grad_norm": 0.3579309582710266, "learning_rate": 9.205058402030525e-05, "loss": 0.0389, "step": 17830 }, { "grad_norm": 0.36727023124694824, "learning_rate": 9.203939848775259e-05, "loss": 0.0383, "step": 17840 }, { "grad_norm": 0.28299206495285034, "learning_rate": 9.202820577181652e-05, "loss": 0.0359, "step": 17850 }, { "grad_norm": 0.45464828610420227, "learning_rate": 9.201700587440953e-05, "loss": 0.0374, "step": 17860 }, { "grad_norm": 0.47840046882629395, "learning_rate": 9.200579879744544e-05, "loss": 0.0378, "step": 17870 }, { "grad_norm": 0.3828243017196655, "learning_rate": 9.199458454283918e-05, "loss": 0.0365, "step": 17880 }, { "grad_norm": 0.7289993166923523, "learning_rate": 9.198336311250697e-05, "loss": 0.0397, "step": 17890 }, { "grad_norm": 0.4068128764629364, "learning_rate": 9.197213450836626e-05, "loss": 0.0382, "step": 17900 }, { "grad_norm": 0.5329107046127319, "learning_rate": 9.19608987323357e-05, "loss": 0.0486, "step": 17910 }, { "grad_norm": 0.39935368299484253, "learning_rate": 9.194965578633517e-05, "loss": 0.0396, "step": 17920 }, { "grad_norm": 0.4251667261123657, "learning_rate": 9.193840567228582e-05, "loss": 0.0412, "step": 17930 }, { "grad_norm": 0.38306495547294617, "learning_rate": 9.192714839210994e-05, "loss": 0.0353, "step": 17940 }, { "grad_norm": 0.3663475513458252, "learning_rate": 9.19158839477311e-05, "loss": 0.0497, "step": 17950 }, { "grad_norm": 0.43275946378707886, "learning_rate": 9.190461234107411e-05, "loss": 0.0407, "step": 17960 }, { "grad_norm": 0.45405375957489014, "learning_rate": 9.189333357406496e-05, "loss": 0.0402, "step": 17970 }, { "grad_norm": 0.3899093270301819, "learning_rate": 9.188204764863089e-05, "loss": 0.0568, "step": 17980 }, { "grad_norm": 0.48575299978256226, "learning_rate": 9.187075456670033e-05, "loss": 0.0391, "step": 17990 }, { "grad_norm": 0.419404536485672, "learning_rate": 9.1859454330203e-05, "loss": 0.0411, "step": 18000 }, { "grad_norm": 0.4716870188713074, "learning_rate": 9.18481469410698e-05, "loss": 0.055, "step": 18010 }, { "grad_norm": 0.3842945098876953, "learning_rate": 9.183683240123281e-05, "loss": 0.0464, "step": 18020 }, { "grad_norm": 0.35966023802757263, "learning_rate": 9.182551071262541e-05, "loss": 0.0445, "step": 18030 }, { "grad_norm": 0.34824955463409424, "learning_rate": 9.181418187718218e-05, "loss": 0.0488, "step": 18040 }, { "grad_norm": 0.5301852226257324, "learning_rate": 9.180284589683888e-05, "loss": 0.0489, "step": 18050 }, { "grad_norm": 0.5582890510559082, "learning_rate": 9.17915027735325e-05, "loss": 0.0458, "step": 18060 }, { "grad_norm": 0.4295784831047058, "learning_rate": 9.178015250920133e-05, "loss": 0.0412, "step": 18070 }, { "grad_norm": 0.3440503776073456, "learning_rate": 9.176879510578477e-05, "loss": 0.0419, "step": 18080 }, { "grad_norm": 0.3629089891910553, "learning_rate": 9.17574305652235e-05, "loss": 0.0386, "step": 18090 }, { "grad_norm": 0.48678773641586304, "learning_rate": 9.174605888945942e-05, "loss": 0.0457, "step": 18100 }, { "grad_norm": 0.4318850040435791, "learning_rate": 9.173468008043564e-05, "loss": 0.0452, "step": 18110 }, { "grad_norm": 0.4112852215766907, "learning_rate": 9.172329414009648e-05, "loss": 0.0428, "step": 18120 }, { "grad_norm": 0.5019362568855286, "learning_rate": 9.171190107038747e-05, "loss": 0.0471, "step": 18130 }, { "grad_norm": 0.3141312003135681, "learning_rate": 9.170050087325541e-05, "loss": 0.0361, "step": 18140 }, { "grad_norm": 0.4315183162689209, "learning_rate": 9.168909355064824e-05, "loss": 0.0429, "step": 18150 }, { "grad_norm": 0.3830341696739197, "learning_rate": 9.167767910451519e-05, "loss": 0.0413, "step": 18160 }, { "grad_norm": 0.4099998474121094, "learning_rate": 9.166625753680669e-05, "loss": 0.0418, "step": 18170 }, { "grad_norm": 0.3640528619289398, "learning_rate": 9.165482884947431e-05, "loss": 0.0328, "step": 18180 }, { "grad_norm": 0.34412822127342224, "learning_rate": 9.164339304447098e-05, "loss": 0.0432, "step": 18190 }, { "grad_norm": 0.36313992738723755, "learning_rate": 9.163195012375072e-05, "loss": 0.0303, "step": 18200 }, { "grad_norm": 0.4210096001625061, "learning_rate": 9.16205000892688e-05, "loss": 0.0464, "step": 18210 }, { "grad_norm": 0.5611625909805298, "learning_rate": 9.160904294298175e-05, "loss": 0.0399, "step": 18220 }, { "grad_norm": 0.2980835437774658, "learning_rate": 9.159757868684727e-05, "loss": 0.0467, "step": 18230 }, { "grad_norm": 0.8466941714286804, "learning_rate": 9.15861073228243e-05, "loss": 0.0458, "step": 18240 }, { "grad_norm": 0.5579171776771545, "learning_rate": 9.157462885287296e-05, "loss": 0.0495, "step": 18250 }, { "grad_norm": 0.3812580406665802, "learning_rate": 9.156314327895461e-05, "loss": 0.0371, "step": 18260 }, { "grad_norm": 0.35773342847824097, "learning_rate": 9.155165060303185e-05, "loss": 0.0434, "step": 18270 }, { "grad_norm": 0.3809218406677246, "learning_rate": 9.154015082706841e-05, "loss": 0.0352, "step": 18280 }, { "grad_norm": 0.3990982472896576, "learning_rate": 9.152864395302936e-05, "loss": 0.0464, "step": 18290 }, { "grad_norm": 0.31902801990509033, "learning_rate": 9.151712998288085e-05, "loss": 0.0326, "step": 18300 }, { "grad_norm": 0.3216908872127533, "learning_rate": 9.150560891859031e-05, "loss": 0.0392, "step": 18310 }, { "grad_norm": 0.4325176477432251, "learning_rate": 9.14940807621264e-05, "loss": 0.0396, "step": 18320 }, { "grad_norm": 0.30212539434432983, "learning_rate": 9.148254551545894e-05, "loss": 0.0428, "step": 18330 }, { "grad_norm": 0.5330511927604675, "learning_rate": 9.147100318055901e-05, "loss": 0.0407, "step": 18340 }, { "grad_norm": 0.556839644908905, "learning_rate": 9.145945375939888e-05, "loss": 0.045, "step": 18350 }, { "grad_norm": 0.3598996698856354, "learning_rate": 9.144789725395203e-05, "loss": 0.0435, "step": 18360 }, { "grad_norm": 0.37975379824638367, "learning_rate": 9.14363336661931e-05, "loss": 0.0405, "step": 18370 }, { "grad_norm": 0.4911458492279053, "learning_rate": 9.142476299809806e-05, "loss": 0.0442, "step": 18380 }, { "grad_norm": 0.4046519696712494, "learning_rate": 9.1413185251644e-05, "loss": 0.0412, "step": 18390 }, { "grad_norm": 0.5573268532752991, "learning_rate": 9.140160042880923e-05, "loss": 0.0494, "step": 18400 }, { "grad_norm": 0.2879200279712677, "learning_rate": 9.139000853157327e-05, "loss": 0.0456, "step": 18410 }, { "grad_norm": 0.49945369362831116, "learning_rate": 9.137840956191688e-05, "loss": 0.0404, "step": 18420 }, { "grad_norm": 0.28073954582214355, "learning_rate": 9.136680352182199e-05, "loss": 0.0378, "step": 18430 }, { "grad_norm": 0.3186326324939728, "learning_rate": 9.135519041327177e-05, "loss": 0.038, "step": 18440 }, { "grad_norm": 0.43882232904434204, "learning_rate": 9.134357023825058e-05, "loss": 0.0503, "step": 18450 }, { "grad_norm": 0.44378116726875305, "learning_rate": 9.133194299874398e-05, "loss": 0.0513, "step": 18460 }, { "grad_norm": 0.5836512446403503, "learning_rate": 9.132030869673876e-05, "loss": 0.0472, "step": 18470 }, { "grad_norm": 0.32940080761909485, "learning_rate": 9.130866733422288e-05, "loss": 0.0447, "step": 18480 }, { "grad_norm": 0.4792351722717285, "learning_rate": 9.129701891318556e-05, "loss": 0.0445, "step": 18490 }, { "grad_norm": 0.36714091897010803, "learning_rate": 9.128536343561718e-05, "loss": 0.0512, "step": 18500 }, { "grad_norm": 0.457996666431427, "learning_rate": 9.127370090350934e-05, "loss": 0.0469, "step": 18510 }, { "grad_norm": 0.51473069190979, "learning_rate": 9.126203131885487e-05, "loss": 0.0503, "step": 18520 }, { "grad_norm": 0.3549615442752838, "learning_rate": 9.125035468364775e-05, "loss": 0.0457, "step": 18530 }, { "grad_norm": 0.4845008850097656, "learning_rate": 9.123867099988322e-05, "loss": 0.0346, "step": 18540 }, { "grad_norm": 0.3523663878440857, "learning_rate": 9.122698026955769e-05, "loss": 0.0448, "step": 18550 }, { "grad_norm": 0.21860361099243164, "learning_rate": 9.12152824946688e-05, "loss": 0.0546, "step": 18560 }, { "grad_norm": 0.443512499332428, "learning_rate": 9.120357767721538e-05, "loss": 0.0424, "step": 18570 }, { "grad_norm": 0.45990702509880066, "learning_rate": 9.119186581919745e-05, "loss": 0.0425, "step": 18580 }, { "grad_norm": 0.41456812620162964, "learning_rate": 9.118014692261624e-05, "loss": 0.0399, "step": 18590 }, { "grad_norm": 0.514858603477478, "learning_rate": 9.116842098947422e-05, "loss": 0.0563, "step": 18600 }, { "grad_norm": 0.37062904238700867, "learning_rate": 9.115668802177499e-05, "loss": 0.0383, "step": 18610 }, { "grad_norm": 0.2765119969844818, "learning_rate": 9.114494802152342e-05, "loss": 0.0447, "step": 18620 }, { "grad_norm": 0.47749966382980347, "learning_rate": 9.113320099072555e-05, "loss": 0.0657, "step": 18630 }, { "grad_norm": 0.31991031765937805, "learning_rate": 9.112144693138864e-05, "loss": 0.0449, "step": 18640 }, { "grad_norm": 0.48221614956855774, "learning_rate": 9.110968584552111e-05, "loss": 0.0434, "step": 18650 }, { "grad_norm": 0.5675318241119385, "learning_rate": 9.109791773513264e-05, "loss": 0.0438, "step": 18660 }, { "grad_norm": 0.3397976756095886, "learning_rate": 9.108614260223403e-05, "loss": 0.0422, "step": 18670 }, { "grad_norm": 0.3516388535499573, "learning_rate": 9.107436044883738e-05, "loss": 0.0395, "step": 18680 }, { "grad_norm": 0.39873388409614563, "learning_rate": 9.10625712769559e-05, "loss": 0.0407, "step": 18690 }, { "grad_norm": 0.3402842879295349, "learning_rate": 9.105077508860406e-05, "loss": 0.0381, "step": 18700 }, { "grad_norm": 0.2833511233329773, "learning_rate": 9.103897188579751e-05, "loss": 0.037, "step": 18710 }, { "grad_norm": 0.24729196727275848, "learning_rate": 9.102716167055308e-05, "loss": 0.0317, "step": 18720 }, { "grad_norm": 0.3595465421676636, "learning_rate": 9.10153444448888e-05, "loss": 0.0747, "step": 18730 }, { "grad_norm": 0.3755250573158264, "learning_rate": 9.100352021082393e-05, "loss": 0.0326, "step": 18740 }, { "grad_norm": 0.4094937741756439, "learning_rate": 9.099168897037891e-05, "loss": 0.0375, "step": 18750 }, { "grad_norm": 0.41682496666908264, "learning_rate": 9.097985072557538e-05, "loss": 0.0364, "step": 18760 }, { "grad_norm": 0.40483537316322327, "learning_rate": 9.096800547843615e-05, "loss": 0.0622, "step": 18770 }, { "grad_norm": 0.46298664808273315, "learning_rate": 9.095615323098526e-05, "loss": 0.0525, "step": 18780 }, { "grad_norm": 0.5178123116493225, "learning_rate": 9.094429398524795e-05, "loss": 0.0412, "step": 18790 }, { "grad_norm": 0.4007304310798645, "learning_rate": 9.093242774325061e-05, "loss": 0.0408, "step": 18800 }, { "grad_norm": 0.4400074779987335, "learning_rate": 9.092055450702088e-05, "loss": 0.0428, "step": 18810 }, { "grad_norm": 0.44320517778396606, "learning_rate": 9.090867427858756e-05, "loss": 0.0426, "step": 18820 }, { "grad_norm": 0.432971715927124, "learning_rate": 9.089678705998066e-05, "loss": 0.0454, "step": 18830 }, { "grad_norm": 0.3834591805934906, "learning_rate": 9.088489285323139e-05, "loss": 0.0441, "step": 18840 }, { "grad_norm": 0.5148734450340271, "learning_rate": 9.087299166037212e-05, "loss": 0.0463, "step": 18850 }, { "grad_norm": 0.3679112195968628, "learning_rate": 9.086108348343647e-05, "loss": 0.0475, "step": 18860 }, { "grad_norm": 0.6082296967506409, "learning_rate": 9.08491683244592e-05, "loss": 0.0532, "step": 18870 }, { "grad_norm": 0.5405243039131165, "learning_rate": 9.08372461854763e-05, "loss": 0.0421, "step": 18880 }, { "grad_norm": 0.41671448945999146, "learning_rate": 9.082531706852492e-05, "loss": 0.0446, "step": 18890 }, { "grad_norm": 0.339776873588562, "learning_rate": 9.081338097564342e-05, "loss": 0.0591, "step": 18900 }, { "grad_norm": 0.23345695436000824, "learning_rate": 9.080143790887137e-05, "loss": 0.0383, "step": 18910 }, { "grad_norm": 0.7254964113235474, "learning_rate": 9.07894878702495e-05, "loss": 0.0439, "step": 18920 }, { "grad_norm": 0.558225154876709, "learning_rate": 9.077753086181974e-05, "loss": 0.0547, "step": 18930 }, { "grad_norm": 0.3003431260585785, "learning_rate": 9.076556688562524e-05, "loss": 0.0391, "step": 18940 }, { "grad_norm": 0.33998793363571167, "learning_rate": 9.075359594371029e-05, "loss": 0.0395, "step": 18950 }, { "grad_norm": 0.3237994909286499, "learning_rate": 9.07416180381204e-05, "loss": 0.0396, "step": 18960 }, { "grad_norm": 0.4473417401313782, "learning_rate": 9.072963317090228e-05, "loss": 0.0515, "step": 18970 }, { "grad_norm": 0.40626856684684753, "learning_rate": 9.071764134410382e-05, "loss": 0.0374, "step": 18980 }, { "grad_norm": 0.45148593187332153, "learning_rate": 9.070564255977407e-05, "loss": 0.0365, "step": 18990 }, { "grad_norm": 0.3720068633556366, "learning_rate": 9.06936368199633e-05, "loss": 0.0406, "step": 19000 }, { "grad_norm": 0.5690690875053406, "learning_rate": 9.0681624126723e-05, "loss": 0.0426, "step": 19010 }, { "grad_norm": 0.5037522912025452, "learning_rate": 9.066960448210576e-05, "loss": 0.0448, "step": 19020 }, { "grad_norm": 0.33842000365257263, "learning_rate": 9.065757788816543e-05, "loss": 0.0337, "step": 19030 }, { "grad_norm": 0.46390119194984436, "learning_rate": 9.064554434695705e-05, "loss": 0.0466, "step": 19040 }, { "grad_norm": 0.3055019974708557, "learning_rate": 9.063350386053677e-05, "loss": 0.0396, "step": 19050 }, { "grad_norm": 0.38627201318740845, "learning_rate": 9.062145643096202e-05, "loss": 0.0381, "step": 19060 }, { "grad_norm": 0.3598109185695648, "learning_rate": 9.060940206029136e-05, "loss": 0.0502, "step": 19070 }, { "grad_norm": 0.4854452311992645, "learning_rate": 9.059734075058457e-05, "loss": 0.0385, "step": 19080 }, { "grad_norm": 0.34673693776130676, "learning_rate": 9.058527250390257e-05, "loss": 0.0512, "step": 19090 }, { "grad_norm": 0.32052892446517944, "learning_rate": 9.057319732230752e-05, "loss": 0.05, "step": 19100 }, { "grad_norm": 0.32521405816078186, "learning_rate": 9.056111520786273e-05, "loss": 0.0354, "step": 19110 }, { "grad_norm": 0.41221264004707336, "learning_rate": 9.054902616263268e-05, "loss": 0.0393, "step": 19120 }, { "grad_norm": 0.24185383319854736, "learning_rate": 9.05369301886831e-05, "loss": 0.0522, "step": 19130 }, { "grad_norm": 0.32477664947509766, "learning_rate": 9.052482728808083e-05, "loss": 0.0356, "step": 19140 }, { "grad_norm": 0.28687241673469543, "learning_rate": 9.051271746289391e-05, "loss": 0.0302, "step": 19150 }, { "grad_norm": 0.4332177937030792, "learning_rate": 9.050060071519162e-05, "loss": 0.0436, "step": 19160 }, { "grad_norm": 0.35298705101013184, "learning_rate": 9.048847704704437e-05, "loss": 0.0434, "step": 19170 }, { "grad_norm": 0.45323559641838074, "learning_rate": 9.047634646052376e-05, "loss": 0.0465, "step": 19180 }, { "grad_norm": 0.525418221950531, "learning_rate": 9.046420895770256e-05, "loss": 0.0531, "step": 19190 }, { "grad_norm": 0.4335762858390808, "learning_rate": 9.045206454065473e-05, "loss": 0.0459, "step": 19200 }, { "grad_norm": 0.3874301314353943, "learning_rate": 9.043991321145546e-05, "loss": 0.0386, "step": 19210 }, { "grad_norm": 0.43202248215675354, "learning_rate": 9.042775497218105e-05, "loss": 0.0333, "step": 19220 }, { "grad_norm": 0.3507378101348877, "learning_rate": 9.041558982490901e-05, "loss": 0.0414, "step": 19230 }, { "grad_norm": 0.3228272795677185, "learning_rate": 9.040341777171805e-05, "loss": 0.0376, "step": 19240 }, { "grad_norm": 0.3568195104598999, "learning_rate": 9.039123881468802e-05, "loss": 0.0378, "step": 19250 }, { "grad_norm": 0.25250571966171265, "learning_rate": 9.037905295589998e-05, "loss": 0.0321, "step": 19260 }, { "grad_norm": 0.6099206805229187, "learning_rate": 9.036686019743617e-05, "loss": 0.0349, "step": 19270 }, { "grad_norm": 0.46500787138938904, "learning_rate": 9.035466054137997e-05, "loss": 0.0432, "step": 19280 }, { "grad_norm": 0.4638304114341736, "learning_rate": 9.0342453989816e-05, "loss": 0.0584, "step": 19290 }, { "grad_norm": 0.3873019516468048, "learning_rate": 9.033024054483e-05, "loss": 0.0422, "step": 19300 }, { "grad_norm": 0.44971972703933716, "learning_rate": 9.031802020850894e-05, "loss": 0.0377, "step": 19310 }, { "grad_norm": 0.38021278381347656, "learning_rate": 9.030579298294092e-05, "loss": 0.0527, "step": 19320 }, { "grad_norm": 0.39783433079719543, "learning_rate": 9.029355887021524e-05, "loss": 0.0552, "step": 19330 }, { "grad_norm": 0.6072502732276917, "learning_rate": 9.028131787242238e-05, "loss": 0.0638, "step": 19340 }, { "grad_norm": 0.3566831350326538, "learning_rate": 9.026906999165399e-05, "loss": 0.0545, "step": 19350 }, { "grad_norm": 0.39775553345680237, "learning_rate": 9.025681523000291e-05, "loss": 0.0407, "step": 19360 }, { "grad_norm": 0.4111008942127228, "learning_rate": 9.024455358956315e-05, "loss": 0.0421, "step": 19370 }, { "grad_norm": 0.4235496520996094, "learning_rate": 9.023228507242984e-05, "loss": 0.045, "step": 19380 }, { "grad_norm": 0.4794710874557495, "learning_rate": 9.022000968069937e-05, "loss": 0.046, "step": 19390 }, { "grad_norm": 0.3170353174209595, "learning_rate": 9.020772741646928e-05, "loss": 0.0391, "step": 19400 }, { "grad_norm": 0.4781688153743744, "learning_rate": 9.019543828183826e-05, "loss": 0.042, "step": 19410 }, { "grad_norm": 0.6918141841888428, "learning_rate": 9.018314227890616e-05, "loss": 0.0364, "step": 19420 }, { "grad_norm": 0.4989679455757141, "learning_rate": 9.017083940977408e-05, "loss": 0.0601, "step": 19430 }, { "grad_norm": 0.39123114943504333, "learning_rate": 9.015852967654422e-05, "loss": 0.0388, "step": 19440 }, { "grad_norm": 0.3817597031593323, "learning_rate": 9.014621308131996e-05, "loss": 0.0428, "step": 19450 }, { "grad_norm": 0.38111430406570435, "learning_rate": 9.01338896262059e-05, "loss": 0.0392, "step": 19460 }, { "grad_norm": 0.3221248686313629, "learning_rate": 9.012155931330777e-05, "loss": 0.0463, "step": 19470 }, { "grad_norm": 0.3069191873073578, "learning_rate": 9.010922214473246e-05, "loss": 0.0364, "step": 19480 }, { "grad_norm": 0.3286932110786438, "learning_rate": 9.009687812258808e-05, "loss": 0.0477, "step": 19490 }, { "grad_norm": 0.41639286279678345, "learning_rate": 9.00845272489839e-05, "loss": 0.045, "step": 19500 }, { "grad_norm": 0.48480305075645447, "learning_rate": 9.007216952603031e-05, "loss": 0.0513, "step": 19510 }, { "grad_norm": 0.4683266878128052, "learning_rate": 9.005980495583894e-05, "loss": 0.0375, "step": 19520 }, { "grad_norm": 0.4022987484931946, "learning_rate": 9.004743354052252e-05, "loss": 0.0429, "step": 19530 }, { "grad_norm": 0.3442816734313965, "learning_rate": 9.003505528219503e-05, "loss": 0.0397, "step": 19540 }, { "grad_norm": 0.37235668301582336, "learning_rate": 9.002267018297154e-05, "loss": 0.0453, "step": 19550 }, { "grad_norm": 0.6409799456596375, "learning_rate": 9.001027824496834e-05, "loss": 0.0507, "step": 19560 }, { "grad_norm": 0.32483673095703125, "learning_rate": 8.999787947030287e-05, "loss": 0.0397, "step": 19570 }, { "grad_norm": 0.3012678623199463, "learning_rate": 8.998547386109376e-05, "loss": 0.0487, "step": 19580 }, { "grad_norm": 0.4974212646484375, "learning_rate": 8.997306141946073e-05, "loss": 0.0381, "step": 19590 }, { "grad_norm": 0.5887474417686462, "learning_rate": 8.996064214752481e-05, "loss": 0.0465, "step": 19600 }, { "grad_norm": 0.3840291202068329, "learning_rate": 8.994821604740806e-05, "loss": 0.0361, "step": 19610 }, { "grad_norm": 0.3849497437477112, "learning_rate": 8.993578312123377e-05, "loss": 0.042, "step": 19620 }, { "grad_norm": 0.344363808631897, "learning_rate": 8.992334337112639e-05, "loss": 0.0417, "step": 19630 }, { "grad_norm": 0.3963110148906708, "learning_rate": 8.991089679921154e-05, "loss": 0.0325, "step": 19640 }, { "grad_norm": 0.3853728175163269, "learning_rate": 8.989844340761599e-05, "loss": 0.0423, "step": 19650 }, { "grad_norm": 0.42036378383636475, "learning_rate": 8.988598319846768e-05, "loss": 0.0364, "step": 19660 }, { "grad_norm": 0.35276705026626587, "learning_rate": 8.987351617389574e-05, "loss": 0.0414, "step": 19670 }, { "grad_norm": 0.3334489166736603, "learning_rate": 8.98610423360304e-05, "loss": 0.0359, "step": 19680 }, { "grad_norm": 0.33359137177467346, "learning_rate": 8.984856168700317e-05, "loss": 0.0407, "step": 19690 }, { "grad_norm": 0.38171809911727905, "learning_rate": 8.983607422894658e-05, "loss": 0.0412, "step": 19700 }, { "grad_norm": 0.31430038809776306, "learning_rate": 8.982357996399442e-05, "loss": 0.0454, "step": 19710 }, { "grad_norm": 0.46111100912094116, "learning_rate": 8.981107889428164e-05, "loss": 0.0392, "step": 19720 }, { "grad_norm": 0.39179977774620056, "learning_rate": 8.979857102194428e-05, "loss": 0.0359, "step": 19730 }, { "grad_norm": 0.43635115027427673, "learning_rate": 8.978605634911968e-05, "loss": 0.0408, "step": 19740 }, { "grad_norm": 0.5822969079017639, "learning_rate": 8.977353487794616e-05, "loss": 0.0369, "step": 19750 }, { "grad_norm": 0.37051913142204285, "learning_rate": 8.976100661056334e-05, "loss": 0.0487, "step": 19760 }, { "grad_norm": 0.4457007050514221, "learning_rate": 8.974847154911197e-05, "loss": 0.0399, "step": 19770 }, { "grad_norm": 0.3530732989311218, "learning_rate": 8.973592969573393e-05, "loss": 0.0489, "step": 19780 }, { "grad_norm": 0.42880916595458984, "learning_rate": 8.972338105257228e-05, "loss": 0.0528, "step": 19790 }, { "grad_norm": 0.3025886118412018, "learning_rate": 8.971082562177125e-05, "loss": 0.0351, "step": 19800 }, { "grad_norm": 0.34141016006469727, "learning_rate": 8.96982634054762e-05, "loss": 0.0391, "step": 19810 }, { "grad_norm": 0.3432392179965973, "learning_rate": 8.96856944058337e-05, "loss": 0.0433, "step": 19820 }, { "grad_norm": 0.3473479449748993, "learning_rate": 8.967311862499144e-05, "loss": 0.0335, "step": 19830 }, { "grad_norm": 0.3756226599216461, "learning_rate": 8.966053606509825e-05, "loss": 0.0327, "step": 19840 }, { "grad_norm": 0.46503686904907227, "learning_rate": 8.964794672830417e-05, "loss": 0.05, "step": 19850 }, { "grad_norm": 0.4088941216468811, "learning_rate": 8.963535061676038e-05, "loss": 0.0462, "step": 19860 }, { "grad_norm": 0.4948892295360565, "learning_rate": 8.962274773261918e-05, "loss": 0.0378, "step": 19870 }, { "grad_norm": 0.3806103467941284, "learning_rate": 8.961013807803409e-05, "loss": 0.0475, "step": 19880 }, { "grad_norm": 0.4270681142807007, "learning_rate": 8.959752165515973e-05, "loss": 0.0454, "step": 19890 }, { "grad_norm": 0.5237202644348145, "learning_rate": 8.958489846615193e-05, "loss": 0.0645, "step": 19900 }, { "grad_norm": 0.5495606064796448, "learning_rate": 8.957226851316762e-05, "loss": 0.0479, "step": 19910 }, { "grad_norm": 0.37260881066322327, "learning_rate": 8.955963179836493e-05, "loss": 0.0396, "step": 19920 }, { "grad_norm": 0.42991364002227783, "learning_rate": 8.954698832390312e-05, "loss": 0.0442, "step": 19930 }, { "grad_norm": 0.33924955129623413, "learning_rate": 8.953433809194263e-05, "loss": 0.0393, "step": 19940 }, { "grad_norm": 0.32537150382995605, "learning_rate": 8.9521681104645e-05, "loss": 0.0391, "step": 19950 }, { "grad_norm": 0.49121737480163574, "learning_rate": 8.9509017364173e-05, "loss": 0.0454, "step": 19960 }, { "grad_norm": 0.36876723170280457, "learning_rate": 8.949634687269052e-05, "loss": 0.036, "step": 19970 }, { "grad_norm": 0.6263166666030884, "learning_rate": 8.948366963236259e-05, "loss": 0.0398, "step": 19980 }, { "grad_norm": 0.5745136737823486, "learning_rate": 8.947098564535538e-05, "loss": 0.0441, "step": 19990 }, { "grad_norm": 0.465251088142395, "learning_rate": 8.945829491383627e-05, "loss": 0.0437, "step": 20000 }, { "grad_norm": 0.3583953082561493, "learning_rate": 8.944559743997374e-05, "loss": 0.0401, "step": 20010 }, { "grad_norm": 0.5218141674995422, "learning_rate": 8.943289322593746e-05, "loss": 0.0394, "step": 20020 }, { "grad_norm": 0.5800313949584961, "learning_rate": 8.942018227389821e-05, "loss": 0.0352, "step": 20030 }, { "grad_norm": 0.5320491790771484, "learning_rate": 8.940746458602795e-05, "loss": 0.0511, "step": 20040 }, { "grad_norm": 0.45846763253211975, "learning_rate": 8.939474016449979e-05, "loss": 0.041, "step": 20050 }, { "grad_norm": 0.4573795199394226, "learning_rate": 8.938200901148799e-05, "loss": 0.0334, "step": 20060 }, { "grad_norm": 0.5433993935585022, "learning_rate": 8.936927112916795e-05, "loss": 0.0435, "step": 20070 }, { "grad_norm": 0.3879251778125763, "learning_rate": 8.935652651971622e-05, "loss": 0.0375, "step": 20080 }, { "grad_norm": 0.3762998580932617, "learning_rate": 8.934377518531052e-05, "loss": 0.0466, "step": 20090 }, { "grad_norm": 0.3466789722442627, "learning_rate": 8.933101712812967e-05, "loss": 0.0447, "step": 20100 }, { "grad_norm": 0.2925940454006195, "learning_rate": 8.931825235035374e-05, "loss": 0.0376, "step": 20110 }, { "grad_norm": 0.4181855320930481, "learning_rate": 8.930548085416382e-05, "loss": 0.0392, "step": 20120 }, { "grad_norm": 0.4254727363586426, "learning_rate": 8.92927026417422e-05, "loss": 0.0389, "step": 20130 }, { "grad_norm": 0.4249049425125122, "learning_rate": 8.92799177152724e-05, "loss": 0.0351, "step": 20140 }, { "grad_norm": 0.30808866024017334, "learning_rate": 8.926712607693895e-05, "loss": 0.0346, "step": 20150 }, { "grad_norm": 0.23511089384555817, "learning_rate": 8.925432772892762e-05, "loss": 0.0438, "step": 20160 }, { "grad_norm": 0.4728407859802246, "learning_rate": 8.924152267342529e-05, "loss": 0.0509, "step": 20170 }, { "grad_norm": 0.37727540731430054, "learning_rate": 8.922871091261998e-05, "loss": 0.0449, "step": 20180 }, { "grad_norm": 0.35594967007637024, "learning_rate": 8.92158924487009e-05, "loss": 0.0353, "step": 20190 }, { "grad_norm": 0.3599487245082855, "learning_rate": 8.920306728385834e-05, "loss": 0.0379, "step": 20200 }, { "grad_norm": 0.28280702233314514, "learning_rate": 8.919023542028379e-05, "loss": 0.0374, "step": 20210 }, { "grad_norm": 0.34982186555862427, "learning_rate": 8.917739686016988e-05, "loss": 0.0382, "step": 20220 }, { "grad_norm": 0.5623500943183899, "learning_rate": 8.916455160571033e-05, "loss": 0.0492, "step": 20230 }, { "grad_norm": 0.37548449635505676, "learning_rate": 8.915169965910008e-05, "loss": 0.0365, "step": 20240 }, { "grad_norm": 0.35807305574417114, "learning_rate": 8.913884102253514e-05, "loss": 0.0394, "step": 20250 }, { "grad_norm": 0.3772886395454407, "learning_rate": 8.912597569821273e-05, "loss": 0.0304, "step": 20260 }, { "grad_norm": 0.40086376667022705, "learning_rate": 8.911310368833118e-05, "loss": 0.0343, "step": 20270 }, { "grad_norm": 0.392812043428421, "learning_rate": 8.910022499508994e-05, "loss": 0.0328, "step": 20280 }, { "grad_norm": 0.32291629910469055, "learning_rate": 8.908733962068965e-05, "loss": 0.0455, "step": 20290 }, { "grad_norm": 0.3247435390949249, "learning_rate": 8.907444756733207e-05, "loss": 0.0374, "step": 20300 }, { "grad_norm": 0.4940406084060669, "learning_rate": 8.906154883722006e-05, "loss": 0.045, "step": 20310 }, { "grad_norm": 0.4877251088619232, "learning_rate": 8.904864343255773e-05, "loss": 0.0384, "step": 20320 }, { "grad_norm": 0.3788592219352722, "learning_rate": 8.90357313555502e-05, "loss": 0.0385, "step": 20330 }, { "grad_norm": 0.35362687706947327, "learning_rate": 8.90228126084038e-05, "loss": 0.0424, "step": 20340 }, { "grad_norm": 0.4087488353252411, "learning_rate": 8.900988719332601e-05, "loss": 0.0485, "step": 20350 }, { "grad_norm": 0.34729012846946716, "learning_rate": 8.899695511252542e-05, "loss": 0.0376, "step": 20360 }, { "grad_norm": 0.40569162368774414, "learning_rate": 8.898401636821176e-05, "loss": 0.0406, "step": 20370 }, { "grad_norm": 0.3506811559200287, "learning_rate": 8.897107096259593e-05, "loss": 0.0497, "step": 20380 }, { "grad_norm": 0.37993910908699036, "learning_rate": 8.895811889788994e-05, "loss": 0.0405, "step": 20390 }, { "grad_norm": 0.26173821091651917, "learning_rate": 8.894516017630692e-05, "loss": 0.035, "step": 20400 }, { "grad_norm": 0.403800368309021, "learning_rate": 8.893219480006118e-05, "loss": 0.0363, "step": 20410 }, { "grad_norm": 0.45609530806541443, "learning_rate": 8.891922277136817e-05, "loss": 0.0375, "step": 20420 }, { "grad_norm": 0.36136120557785034, "learning_rate": 8.890624409244441e-05, "loss": 0.0332, "step": 20430 }, { "grad_norm": 0.4214688539505005, "learning_rate": 8.889325876550763e-05, "loss": 0.0411, "step": 20440 }, { "grad_norm": 0.3423345983028412, "learning_rate": 8.888026679277666e-05, "loss": 0.04, "step": 20450 }, { "grad_norm": 0.3148459196090698, "learning_rate": 8.886726817647147e-05, "loss": 0.0591, "step": 20460 }, { "grad_norm": 0.3474551737308502, "learning_rate": 8.885426291881319e-05, "loss": 0.0515, "step": 20470 }, { "grad_norm": 0.3410370647907257, "learning_rate": 8.884125102202401e-05, "loss": 0.0378, "step": 20480 }, { "grad_norm": 0.40472519397735596, "learning_rate": 8.882823248832736e-05, "loss": 0.0335, "step": 20490 }, { "grad_norm": 0.2967834770679474, "learning_rate": 8.881520731994772e-05, "loss": 0.0368, "step": 20500 }, { "grad_norm": 0.4534974992275238, "learning_rate": 8.880217551911077e-05, "loss": 0.0414, "step": 20510 }, { "grad_norm": 0.3177756369113922, "learning_rate": 8.878913708804323e-05, "loss": 0.0389, "step": 20520 }, { "grad_norm": 0.35602593421936035, "learning_rate": 8.877609202897308e-05, "loss": 0.0392, "step": 20530 }, { "grad_norm": 0.414837121963501, "learning_rate": 8.876304034412933e-05, "loss": 0.0369, "step": 20540 }, { "grad_norm": 0.4215880334377289, "learning_rate": 8.874998203574214e-05, "loss": 0.0412, "step": 20550 }, { "grad_norm": 0.4456618130207062, "learning_rate": 8.873691710604284e-05, "loss": 0.036, "step": 20560 }, { "grad_norm": 0.4458702802658081, "learning_rate": 8.872384555726387e-05, "loss": 0.0336, "step": 20570 }, { "grad_norm": 0.408117413520813, "learning_rate": 8.871076739163878e-05, "loss": 0.0377, "step": 20580 }, { "grad_norm": 0.34394407272338867, "learning_rate": 8.86976826114023e-05, "loss": 0.039, "step": 20590 }, { "grad_norm": 0.4165799021720886, "learning_rate": 8.868459121879023e-05, "loss": 0.0407, "step": 20600 }, { "grad_norm": 0.4048711359500885, "learning_rate": 8.867149321603956e-05, "loss": 0.045, "step": 20610 }, { "grad_norm": 0.342088520526886, "learning_rate": 8.865838860538835e-05, "loss": 0.0364, "step": 20620 }, { "grad_norm": 0.4395991563796997, "learning_rate": 8.864527738907585e-05, "loss": 0.0439, "step": 20630 }, { "grad_norm": 0.4532506465911865, "learning_rate": 8.863215956934239e-05, "loss": 0.0403, "step": 20640 }, { "grad_norm": 0.30837225914001465, "learning_rate": 8.861903514842947e-05, "loss": 0.0301, "step": 20650 }, { "grad_norm": 0.4818606674671173, "learning_rate": 8.860590412857966e-05, "loss": 0.0349, "step": 20660 }, { "grad_norm": 0.26370498538017273, "learning_rate": 8.85927665120367e-05, "loss": 0.031, "step": 20670 }, { "grad_norm": 0.3875586688518524, "learning_rate": 8.857962230104546e-05, "loss": 0.0339, "step": 20680 }, { "grad_norm": 0.3645019233226776, "learning_rate": 8.856647149785193e-05, "loss": 0.0363, "step": 20690 }, { "grad_norm": 0.3186663091182709, "learning_rate": 8.855331410470322e-05, "loss": 0.0374, "step": 20700 }, { "grad_norm": 0.39752107858657837, "learning_rate": 8.854015012384756e-05, "loss": 0.0334, "step": 20710 }, { "grad_norm": 0.3557101786136627, "learning_rate": 8.852697955753433e-05, "loss": 0.0392, "step": 20720 }, { "grad_norm": 0.32240721583366394, "learning_rate": 8.851380240801399e-05, "loss": 0.0551, "step": 20730 }, { "grad_norm": 0.337643563747406, "learning_rate": 8.850061867753818e-05, "loss": 0.0335, "step": 20740 }, { "grad_norm": 0.5540109872817993, "learning_rate": 8.848742836835963e-05, "loss": 0.0477, "step": 20750 }, { "grad_norm": 0.31959933042526245, "learning_rate": 8.847423148273221e-05, "loss": 0.0509, "step": 20760 }, { "grad_norm": 0.41701337695121765, "learning_rate": 8.846102802291092e-05, "loss": 0.0582, "step": 20770 }, { "grad_norm": 0.4535450041294098, "learning_rate": 8.844781799115183e-05, "loss": 0.0534, "step": 20780 }, { "grad_norm": 0.36400899291038513, "learning_rate": 8.84346013897122e-05, "loss": 0.0387, "step": 20790 }, { "grad_norm": 0.4134143888950348, "learning_rate": 8.842137822085038e-05, "loss": 0.0503, "step": 20800 }, { "grad_norm": 0.37764549255371094, "learning_rate": 8.840814848682585e-05, "loss": 0.0375, "step": 20810 }, { "grad_norm": 0.28028881549835205, "learning_rate": 8.83949121898992e-05, "loss": 0.0522, "step": 20820 }, { "grad_norm": 0.34217140078544617, "learning_rate": 8.838166933233217e-05, "loss": 0.0322, "step": 20830 }, { "grad_norm": 0.2493356615304947, "learning_rate": 8.83684199163876e-05, "loss": 0.0384, "step": 20840 }, { "grad_norm": 0.4120466709136963, "learning_rate": 8.835516394432943e-05, "loss": 0.034, "step": 20850 }, { "grad_norm": 0.2787671685218811, "learning_rate": 8.834190141842276e-05, "loss": 0.0445, "step": 20860 }, { "grad_norm": 0.6983239054679871, "learning_rate": 8.83286323409338e-05, "loss": 0.0405, "step": 20870 }, { "grad_norm": 0.44794338941574097, "learning_rate": 8.831535671412986e-05, "loss": 0.0414, "step": 20880 }, { "grad_norm": 0.289734810590744, "learning_rate": 8.830207454027938e-05, "loss": 0.0428, "step": 20890 }, { "grad_norm": 0.5737387537956238, "learning_rate": 8.828878582165192e-05, "loss": 0.0374, "step": 20900 }, { "grad_norm": 0.3897026777267456, "learning_rate": 8.827549056051818e-05, "loss": 0.0358, "step": 20910 }, { "grad_norm": 0.2598131597042084, "learning_rate": 8.826218875914993e-05, "loss": 0.0339, "step": 20920 }, { "grad_norm": 0.2350131720304489, "learning_rate": 8.82488804198201e-05, "loss": 0.0329, "step": 20930 }, { "grad_norm": 0.27528977394104004, "learning_rate": 8.82355655448027e-05, "loss": 0.034, "step": 20940 }, { "grad_norm": 0.34923285245895386, "learning_rate": 8.822224413637293e-05, "loss": 0.0373, "step": 20950 }, { "grad_norm": 0.38436266779899597, "learning_rate": 8.820891619680697e-05, "loss": 0.0375, "step": 20960 }, { "grad_norm": 0.4057960510253906, "learning_rate": 8.819558172838227e-05, "loss": 0.0362, "step": 20970 }, { "grad_norm": 0.43136146664619446, "learning_rate": 8.818224073337731e-05, "loss": 0.0301, "step": 20980 }, { "grad_norm": 0.40619900822639465, "learning_rate": 8.816889321407169e-05, "loss": 0.0361, "step": 20990 }, { "grad_norm": 0.2928810715675354, "learning_rate": 8.815553917274615e-05, "loss": 0.0405, "step": 21000 }, { "grad_norm": 0.39814144372940063, "learning_rate": 8.81421786116825e-05, "loss": 0.0447, "step": 21010 }, { "grad_norm": 0.34628862142562866, "learning_rate": 8.812881153316373e-05, "loss": 0.0337, "step": 21020 }, { "grad_norm": 0.47795936465263367, "learning_rate": 8.81154379394739e-05, "loss": 0.0417, "step": 21030 }, { "grad_norm": 0.47528141736984253, "learning_rate": 8.810205783289818e-05, "loss": 0.0344, "step": 21040 }, { "grad_norm": 0.7692768573760986, "learning_rate": 8.808867121572286e-05, "loss": 0.0376, "step": 21050 }, { "grad_norm": 0.38830047845840454, "learning_rate": 8.807527809023537e-05, "loss": 0.045, "step": 21060 }, { "grad_norm": 0.29230642318725586, "learning_rate": 8.80618784587242e-05, "loss": 0.0421, "step": 21070 }, { "grad_norm": 0.49571147561073303, "learning_rate": 8.804847232347902e-05, "loss": 0.0427, "step": 21080 }, { "grad_norm": 0.3154810965061188, "learning_rate": 8.803505968679054e-05, "loss": 0.044, "step": 21090 }, { "grad_norm": 0.4361841082572937, "learning_rate": 8.802164055095061e-05, "loss": 0.0479, "step": 21100 }, { "grad_norm": 0.35536158084869385, "learning_rate": 8.80082149182522e-05, "loss": 0.0367, "step": 21110 }, { "grad_norm": 0.3320657014846802, "learning_rate": 8.79947827909894e-05, "loss": 0.0371, "step": 21120 }, { "grad_norm": 0.35245293378829956, "learning_rate": 8.798134417145738e-05, "loss": 0.0374, "step": 21130 }, { "grad_norm": 0.40797629952430725, "learning_rate": 8.796789906195243e-05, "loss": 0.042, "step": 21140 }, { "grad_norm": 0.41407424211502075, "learning_rate": 8.795444746477195e-05, "loss": 0.0346, "step": 21150 }, { "grad_norm": 0.3565029799938202, "learning_rate": 8.794098938221446e-05, "loss": 0.0435, "step": 21160 }, { "grad_norm": 0.3706984221935272, "learning_rate": 8.792752481657957e-05, "loss": 0.0457, "step": 21170 }, { "grad_norm": 0.40371519327163696, "learning_rate": 8.791405377016802e-05, "loss": 0.0467, "step": 21180 }, { "grad_norm": 0.4608851969242096, "learning_rate": 8.790057624528163e-05, "loss": 0.0466, "step": 21190 }, { "grad_norm": 0.4156792163848877, "learning_rate": 8.788709224422333e-05, "loss": 0.0448, "step": 21200 }, { "grad_norm": 0.4889819324016571, "learning_rate": 8.787360176929717e-05, "loss": 0.0429, "step": 21210 }, { "grad_norm": 0.28606975078582764, "learning_rate": 8.786010482280834e-05, "loss": 0.0396, "step": 21220 }, { "grad_norm": 0.42785221338272095, "learning_rate": 8.784660140706306e-05, "loss": 0.0396, "step": 21230 }, { "grad_norm": 0.4259611964225769, "learning_rate": 8.783309152436872e-05, "loss": 0.0355, "step": 21240 }, { "grad_norm": 0.3568499982357025, "learning_rate": 8.781957517703375e-05, "loss": 0.0341, "step": 21250 }, { "grad_norm": 0.30789029598236084, "learning_rate": 8.780605236736776e-05, "loss": 0.0411, "step": 21260 }, { "grad_norm": 0.22222179174423218, "learning_rate": 8.779252309768142e-05, "loss": 0.0359, "step": 21270 }, { "grad_norm": 0.3180886209011078, "learning_rate": 8.777898737028652e-05, "loss": 0.0384, "step": 21280 }, { "grad_norm": 0.3465861678123474, "learning_rate": 8.776544518749591e-05, "loss": 0.036, "step": 21290 }, { "grad_norm": 0.34628424048423767, "learning_rate": 8.775189655162364e-05, "loss": 0.034, "step": 21300 }, { "grad_norm": 0.4315797984600067, "learning_rate": 8.773834146498474e-05, "loss": 0.0367, "step": 21310 }, { "grad_norm": 0.3292275071144104, "learning_rate": 8.772477992989545e-05, "loss": 0.041, "step": 21320 }, { "grad_norm": 0.42438629269599915, "learning_rate": 8.771121194867304e-05, "loss": 0.0516, "step": 21330 }, { "grad_norm": 0.3761439025402069, "learning_rate": 8.769763752363589e-05, "loss": 0.0392, "step": 21340 }, { "grad_norm": 0.32874900102615356, "learning_rate": 8.768405665710352e-05, "loss": 0.0418, "step": 21350 }, { "grad_norm": 0.6546608209609985, "learning_rate": 8.767046935139655e-05, "loss": 0.0429, "step": 21360 }, { "grad_norm": 0.3673970401287079, "learning_rate": 8.765687560883666e-05, "loss": 0.0388, "step": 21370 }, { "grad_norm": 0.3944304287433624, "learning_rate": 8.764327543174664e-05, "loss": 0.0389, "step": 21380 }, { "grad_norm": 0.37293189764022827, "learning_rate": 8.762966882245038e-05, "loss": 0.0428, "step": 21390 }, { "grad_norm": 0.9253076314926147, "learning_rate": 8.761605578327291e-05, "loss": 0.0461, "step": 21400 }, { "grad_norm": 0.3847813010215759, "learning_rate": 8.76024363165403e-05, "loss": 0.0454, "step": 21410 }, { "grad_norm": 0.33607861399650574, "learning_rate": 8.758881042457976e-05, "loss": 0.0446, "step": 21420 }, { "grad_norm": 0.4558216333389282, "learning_rate": 8.757517810971957e-05, "loss": 0.0403, "step": 21430 }, { "grad_norm": 0.3145231306552887, "learning_rate": 8.756153937428913e-05, "loss": 0.0381, "step": 21440 }, { "grad_norm": 0.42234987020492554, "learning_rate": 8.754789422061889e-05, "loss": 0.0485, "step": 21450 }, { "grad_norm": 0.3476654887199402, "learning_rate": 8.753424265104052e-05, "loss": 0.0405, "step": 21460 }, { "grad_norm": 0.5208801627159119, "learning_rate": 8.752058466788659e-05, "loss": 0.051, "step": 21470 }, { "grad_norm": 0.3859148323535919, "learning_rate": 8.750692027349097e-05, "loss": 0.037, "step": 21480 }, { "grad_norm": 0.44432127475738525, "learning_rate": 8.749324947018847e-05, "loss": 0.0334, "step": 21490 }, { "grad_norm": 0.3530706465244293, "learning_rate": 8.747957226031507e-05, "loss": 0.0382, "step": 21500 }, { "grad_norm": 0.339663565158844, "learning_rate": 8.746588864620787e-05, "loss": 0.0376, "step": 21510 }, { "grad_norm": 0.6140255928039551, "learning_rate": 8.745219863020498e-05, "loss": 0.0327, "step": 21520 }, { "grad_norm": 0.32237112522125244, "learning_rate": 8.743850221464564e-05, "loss": 0.0329, "step": 21530 }, { "grad_norm": 0.3401470482349396, "learning_rate": 8.742479940187026e-05, "loss": 0.0329, "step": 21540 }, { "grad_norm": 0.5146180391311646, "learning_rate": 8.74110901942202e-05, "loss": 0.0346, "step": 21550 }, { "grad_norm": 0.39077284932136536, "learning_rate": 8.739737459403803e-05, "loss": 0.0463, "step": 21560 }, { "grad_norm": 0.4071934223175049, "learning_rate": 8.738365260366737e-05, "loss": 0.0402, "step": 21570 }, { "grad_norm": 0.409741073846817, "learning_rate": 8.736992422545292e-05, "loss": 0.0431, "step": 21580 }, { "grad_norm": 0.5466285347938538, "learning_rate": 8.73561894617405e-05, "loss": 0.0502, "step": 21590 }, { "grad_norm": 0.31343409419059753, "learning_rate": 8.734244831487697e-05, "loss": 0.0339, "step": 21600 }, { "grad_norm": 0.2925306558609009, "learning_rate": 8.732870078721035e-05, "loss": 0.0441, "step": 21610 }, { "grad_norm": 0.40294235944747925, "learning_rate": 8.731494688108972e-05, "loss": 0.0379, "step": 21620 }, { "grad_norm": 0.30740442872047424, "learning_rate": 8.730118659886523e-05, "loss": 0.0411, "step": 21630 }, { "grad_norm": 0.7439016699790955, "learning_rate": 8.728741994288814e-05, "loss": 0.0371, "step": 21640 }, { "grad_norm": 0.43868130445480347, "learning_rate": 8.727364691551079e-05, "loss": 0.0421, "step": 21650 }, { "grad_norm": 0.5941896438598633, "learning_rate": 8.725986751908661e-05, "loss": 0.0449, "step": 21660 }, { "grad_norm": 0.4377099871635437, "learning_rate": 8.724608175597016e-05, "loss": 0.0423, "step": 21670 }, { "grad_norm": 0.37017351388931274, "learning_rate": 8.723228962851699e-05, "loss": 0.0345, "step": 21680 }, { "grad_norm": 0.42577365040779114, "learning_rate": 8.721849113908385e-05, "loss": 0.0425, "step": 21690 }, { "grad_norm": 0.4393180012702942, "learning_rate": 8.720468629002848e-05, "loss": 0.0409, "step": 21700 }, { "grad_norm": 0.3974250257015228, "learning_rate": 8.719087508370978e-05, "loss": 0.0374, "step": 21710 }, { "grad_norm": 0.5794216394424438, "learning_rate": 8.717705752248772e-05, "loss": 0.0481, "step": 21720 }, { "grad_norm": 0.6370225548744202, "learning_rate": 8.71632336087233e-05, "loss": 0.0372, "step": 21730 }, { "grad_norm": 0.46931105852127075, "learning_rate": 8.71494033447787e-05, "loss": 0.0373, "step": 21740 }, { "grad_norm": 0.6328515410423279, "learning_rate": 8.713556673301708e-05, "loss": 0.0339, "step": 21750 }, { "grad_norm": 0.4271431863307953, "learning_rate": 8.712172377580278e-05, "loss": 0.0515, "step": 21760 }, { "grad_norm": 0.4332505464553833, "learning_rate": 8.710787447550114e-05, "loss": 0.0403, "step": 21770 }, { "grad_norm": 0.40760430693626404, "learning_rate": 8.70940188344787e-05, "loss": 0.0504, "step": 21780 }, { "grad_norm": 0.4214969575405121, "learning_rate": 8.708015685510293e-05, "loss": 0.0398, "step": 21790 }, { "grad_norm": 0.5461729764938354, "learning_rate": 8.706628853974252e-05, "loss": 0.0412, "step": 21800 }, { "grad_norm": 0.3687708377838135, "learning_rate": 8.705241389076715e-05, "loss": 0.0349, "step": 21810 }, { "grad_norm": 0.3457154631614685, "learning_rate": 8.703853291054764e-05, "loss": 0.04, "step": 21820 }, { "grad_norm": 0.4601987898349762, "learning_rate": 8.702464560145587e-05, "loss": 0.0441, "step": 21830 }, { "grad_norm": 0.3421706259250641, "learning_rate": 8.701075196586476e-05, "loss": 0.0422, "step": 21840 }, { "grad_norm": 0.34663528203964233, "learning_rate": 8.699685200614842e-05, "loss": 0.0341, "step": 21850 }, { "grad_norm": 0.40280652046203613, "learning_rate": 8.698294572468193e-05, "loss": 0.043, "step": 21860 }, { "grad_norm": 0.30579686164855957, "learning_rate": 8.696903312384148e-05, "loss": 0.0439, "step": 21870 }, { "grad_norm": 0.37862643599510193, "learning_rate": 8.695511420600439e-05, "loss": 0.0355, "step": 21880 }, { "grad_norm": 0.3530524671077728, "learning_rate": 8.694118897354901e-05, "loss": 0.0302, "step": 21890 }, { "grad_norm": 0.3815290927886963, "learning_rate": 8.692725742885478e-05, "loss": 0.0283, "step": 21900 }, { "grad_norm": 0.48341044783592224, "learning_rate": 8.691331957430221e-05, "loss": 0.0367, "step": 21910 }, { "grad_norm": 0.4804910719394684, "learning_rate": 8.68993754122729e-05, "loss": 0.0404, "step": 21920 }, { "grad_norm": 0.4240535497665405, "learning_rate": 8.688542494514955e-05, "loss": 0.0333, "step": 21930 }, { "grad_norm": 0.42851507663726807, "learning_rate": 8.68714681753159e-05, "loss": 0.0374, "step": 21940 }, { "grad_norm": 0.39095747470855713, "learning_rate": 8.685750510515676e-05, "loss": 0.0366, "step": 21950 }, { "grad_norm": 0.44144678115844727, "learning_rate": 8.684353573705805e-05, "loss": 0.0336, "step": 21960 }, { "grad_norm": 0.3744187653064728, "learning_rate": 8.682956007340677e-05, "loss": 0.0297, "step": 21970 }, { "grad_norm": 0.32548266649246216, "learning_rate": 8.681557811659095e-05, "loss": 0.0333, "step": 21980 }, { "grad_norm": 0.34357523918151855, "learning_rate": 8.680158986899974e-05, "loss": 0.0319, "step": 21990 }, { "grad_norm": 0.3200289309024811, "learning_rate": 8.678759533302335e-05, "loss": 0.0337, "step": 22000 }, { "grad_norm": 0.46686306595802307, "learning_rate": 8.677359451105308e-05, "loss": 0.0344, "step": 22010 }, { "grad_norm": 0.3867539167404175, "learning_rate": 8.675958740548123e-05, "loss": 0.0412, "step": 22020 }, { "grad_norm": 0.31311267614364624, "learning_rate": 8.674557401870129e-05, "loss": 0.0409, "step": 22030 }, { "grad_norm": 0.5515240430831909, "learning_rate": 8.673155435310775e-05, "loss": 0.036, "step": 22040 }, { "grad_norm": 0.29734131693840027, "learning_rate": 8.671752841109617e-05, "loss": 0.0317, "step": 22050 }, { "grad_norm": 0.46239426732063293, "learning_rate": 8.670349619506321e-05, "loss": 0.0354, "step": 22060 }, { "grad_norm": 0.3811352252960205, "learning_rate": 8.66894577074066e-05, "loss": 0.0304, "step": 22070 }, { "grad_norm": 0.26945140957832336, "learning_rate": 8.667541295052513e-05, "loss": 0.0314, "step": 22080 }, { "grad_norm": 0.5261849761009216, "learning_rate": 8.666136192681865e-05, "loss": 0.0436, "step": 22090 }, { "grad_norm": 0.500211238861084, "learning_rate": 8.664730463868811e-05, "loss": 0.04, "step": 22100 }, { "grad_norm": 0.46764788031578064, "learning_rate": 8.663324108853552e-05, "loss": 0.0437, "step": 22110 }, { "grad_norm": 0.6987830400466919, "learning_rate": 8.661917127876395e-05, "loss": 0.0542, "step": 22120 }, { "grad_norm": 0.3425256609916687, "learning_rate": 8.660509521177754e-05, "loss": 0.0404, "step": 22130 }, { "grad_norm": 0.2749221920967102, "learning_rate": 8.65910128899815e-05, "loss": 0.0379, "step": 22140 }, { "grad_norm": 0.37383607029914856, "learning_rate": 8.657692431578214e-05, "loss": 0.0422, "step": 22150 }, { "grad_norm": 0.39820435643196106, "learning_rate": 8.656282949158679e-05, "loss": 0.0386, "step": 22160 }, { "grad_norm": 0.2962934672832489, "learning_rate": 8.654872841980388e-05, "loss": 0.0321, "step": 22170 }, { "grad_norm": 0.3436565101146698, "learning_rate": 8.653462110284289e-05, "loss": 0.0273, "step": 22180 }, { "grad_norm": 0.26202869415283203, "learning_rate": 8.652050754311437e-05, "loss": 0.0333, "step": 22190 }, { "grad_norm": 0.3808562457561493, "learning_rate": 8.650638774302995e-05, "loss": 0.0344, "step": 22200 }, { "grad_norm": 0.5322299599647522, "learning_rate": 8.649226170500233e-05, "loss": 0.0382, "step": 22210 }, { "grad_norm": 0.4419575035572052, "learning_rate": 8.647812943144524e-05, "loss": 0.0359, "step": 22220 }, { "grad_norm": 0.32736334204673767, "learning_rate": 8.646399092477351e-05, "loss": 0.031, "step": 22230 }, { "grad_norm": 0.48441481590270996, "learning_rate": 8.644984618740301e-05, "loss": 0.0416, "step": 22240 }, { "grad_norm": 0.309280127286911, "learning_rate": 8.643569522175073e-05, "loss": 0.0285, "step": 22250 }, { "grad_norm": 0.3208202123641968, "learning_rate": 8.642153803023463e-05, "loss": 0.0378, "step": 22260 }, { "grad_norm": 0.30189648270606995, "learning_rate": 8.640737461527383e-05, "loss": 0.0355, "step": 22270 }, { "grad_norm": 0.41439807415008545, "learning_rate": 8.639320497928845e-05, "loss": 0.038, "step": 22280 }, { "grad_norm": 0.35174131393432617, "learning_rate": 8.637902912469969e-05, "loss": 0.031, "step": 22290 }, { "grad_norm": 0.38692915439605713, "learning_rate": 8.636484705392982e-05, "loss": 0.0328, "step": 22300 }, { "grad_norm": 0.3206389546394348, "learning_rate": 8.635065876940216e-05, "loss": 0.0357, "step": 22310 }, { "grad_norm": 0.4669317305088043, "learning_rate": 8.633646427354112e-05, "loss": 0.0392, "step": 22320 }, { "grad_norm": 0.42620962858200073, "learning_rate": 8.632226356877213e-05, "loss": 0.0353, "step": 22330 }, { "grad_norm": 0.2982279658317566, "learning_rate": 8.630805665752173e-05, "loss": 0.0438, "step": 22340 }, { "grad_norm": 0.47769710421562195, "learning_rate": 8.629384354221748e-05, "loss": 0.0412, "step": 22350 }, { "grad_norm": 0.22808051109313965, "learning_rate": 8.627962422528797e-05, "loss": 0.0353, "step": 22360 }, { "grad_norm": 0.5912209749221802, "learning_rate": 8.626539870916296e-05, "loss": 0.0502, "step": 22370 }, { "grad_norm": 0.3547714352607727, "learning_rate": 8.625116699627317e-05, "loss": 0.0472, "step": 22380 }, { "grad_norm": 0.5867677330970764, "learning_rate": 8.623692908905041e-05, "loss": 0.0376, "step": 22390 }, { "grad_norm": 0.8213933706283569, "learning_rate": 8.622268498992755e-05, "loss": 0.0429, "step": 22400 }, { "grad_norm": 0.29858270287513733, "learning_rate": 8.620843470133851e-05, "loss": 0.05, "step": 22410 }, { "grad_norm": 0.3140195906162262, "learning_rate": 8.619417822571829e-05, "loss": 0.0399, "step": 22420 }, { "grad_norm": 0.35588356852531433, "learning_rate": 8.617991556550292e-05, "loss": 0.033, "step": 22430 }, { "grad_norm": 0.3327556550502777, "learning_rate": 8.616564672312952e-05, "loss": 0.0379, "step": 22440 }, { "grad_norm": 0.3805851638317108, "learning_rate": 8.61513717010362e-05, "loss": 0.0375, "step": 22450 }, { "grad_norm": 0.365444540977478, "learning_rate": 8.613709050166221e-05, "loss": 0.0393, "step": 22460 }, { "grad_norm": 0.34511974453926086, "learning_rate": 8.61228031274478e-05, "loss": 0.0465, "step": 22470 }, { "grad_norm": 0.559401273727417, "learning_rate": 8.610850958083431e-05, "loss": 0.0423, "step": 22480 }, { "grad_norm": 0.4532923996448517, "learning_rate": 8.609420986426409e-05, "loss": 0.0414, "step": 22490 }, { "grad_norm": 0.4541196823120117, "learning_rate": 8.60799039801806e-05, "loss": 0.0384, "step": 22500 }, { "grad_norm": 0.450872004032135, "learning_rate": 8.606559193102828e-05, "loss": 0.0325, "step": 22510 }, { "grad_norm": 0.5087043642997742, "learning_rate": 8.605127371925273e-05, "loss": 0.0518, "step": 22520 }, { "grad_norm": 0.37358155846595764, "learning_rate": 8.603694934730047e-05, "loss": 0.0367, "step": 22530 }, { "grad_norm": 0.31811732053756714, "learning_rate": 8.602261881761919e-05, "loss": 0.0269, "step": 22540 }, { "grad_norm": 0.21131834387779236, "learning_rate": 8.600828213265759e-05, "loss": 0.0445, "step": 22550 }, { "grad_norm": 0.3577582836151123, "learning_rate": 8.599393929486539e-05, "loss": 0.0474, "step": 22560 }, { "grad_norm": 0.324005663394928, "learning_rate": 8.59795903066934e-05, "loss": 0.0354, "step": 22570 }, { "grad_norm": 0.32240137457847595, "learning_rate": 8.596523517059347e-05, "loss": 0.035, "step": 22580 }, { "grad_norm": 0.41732123494148254, "learning_rate": 8.59508738890185e-05, "loss": 0.0313, "step": 22590 }, { "grad_norm": 0.26225847005844116, "learning_rate": 8.593650646442246e-05, "loss": 0.0348, "step": 22600 }, { "grad_norm": 0.3892320394515991, "learning_rate": 8.59221328992603e-05, "loss": 0.0312, "step": 22610 }, { "grad_norm": 0.3327404260635376, "learning_rate": 8.590775319598813e-05, "loss": 0.0338, "step": 22620 }, { "grad_norm": 0.36299511790275574, "learning_rate": 8.589336735706301e-05, "loss": 0.0354, "step": 22630 }, { "grad_norm": 0.3757041096687317, "learning_rate": 8.587897538494307e-05, "loss": 0.0512, "step": 22640 }, { "grad_norm": 0.32736361026763916, "learning_rate": 8.586457728208756e-05, "loss": 0.0386, "step": 22650 }, { "grad_norm": 0.41297006607055664, "learning_rate": 8.585017305095667e-05, "loss": 0.0368, "step": 22660 }, { "grad_norm": 0.3784513473510742, "learning_rate": 8.583576269401173e-05, "loss": 0.0355, "step": 22670 }, { "grad_norm": 0.37333643436431885, "learning_rate": 8.582134621371504e-05, "loss": 0.0352, "step": 22680 }, { "grad_norm": 0.5967885851860046, "learning_rate": 8.580692361253e-05, "loss": 0.0346, "step": 22690 }, { "grad_norm": 0.338823527097702, "learning_rate": 8.579249489292104e-05, "loss": 0.0329, "step": 22700 }, { "grad_norm": 0.589467465877533, "learning_rate": 8.577806005735363e-05, "loss": 0.043, "step": 22710 }, { "grad_norm": 0.3742583692073822, "learning_rate": 8.576361910829429e-05, "loss": 0.0312, "step": 22720 }, { "grad_norm": 0.40348297357559204, "learning_rate": 8.574917204821057e-05, "loss": 0.0422, "step": 22730 }, { "grad_norm": 0.4291239380836487, "learning_rate": 8.57347188795711e-05, "loss": 0.038, "step": 22740 }, { "grad_norm": 0.3859820067882538, "learning_rate": 8.572025960484551e-05, "loss": 0.0384, "step": 22750 }, { "grad_norm": 0.37094148993492126, "learning_rate": 8.57057942265045e-05, "loss": 0.0458, "step": 22760 }, { "grad_norm": 0.4744853675365448, "learning_rate": 8.569132274701984e-05, "loss": 0.035, "step": 22770 }, { "grad_norm": 0.36742478609085083, "learning_rate": 8.567684516886427e-05, "loss": 0.0353, "step": 22780 }, { "grad_norm": 0.4153997600078583, "learning_rate": 8.56623614945116e-05, "loss": 0.0345, "step": 22790 }, { "grad_norm": 0.4414551854133606, "learning_rate": 8.564787172643675e-05, "loss": 0.0391, "step": 22800 }, { "grad_norm": 0.2732792794704437, "learning_rate": 8.563337586711559e-05, "loss": 0.0397, "step": 22810 }, { "grad_norm": 0.3765203058719635, "learning_rate": 8.561887391902506e-05, "loss": 0.0431, "step": 22820 }, { "grad_norm": 0.48653340339660645, "learning_rate": 8.560436588464316e-05, "loss": 0.0322, "step": 22830 }, { "grad_norm": 0.3647955358028412, "learning_rate": 8.55898517664489e-05, "loss": 0.0414, "step": 22840 }, { "grad_norm": 0.571749746799469, "learning_rate": 8.557533156692236e-05, "loss": 0.0513, "step": 22850 }, { "grad_norm": 0.3625350296497345, "learning_rate": 8.556080528854467e-05, "loss": 0.0292, "step": 22860 }, { "grad_norm": 0.373230904340744, "learning_rate": 8.554627293379791e-05, "loss": 0.0332, "step": 22870 }, { "grad_norm": 0.38020142912864685, "learning_rate": 8.553173450516531e-05, "loss": 0.0452, "step": 22880 }, { "grad_norm": 0.37049049139022827, "learning_rate": 8.551719000513108e-05, "loss": 0.0335, "step": 22890 }, { "grad_norm": 0.42454978823661804, "learning_rate": 8.550263943618049e-05, "loss": 0.0515, "step": 22900 }, { "grad_norm": 0.3277033865451813, "learning_rate": 8.54880828007998e-05, "loss": 0.0441, "step": 22910 }, { "grad_norm": 0.335036963224411, "learning_rate": 8.547352010147637e-05, "loss": 0.0409, "step": 22920 }, { "grad_norm": 0.34275612235069275, "learning_rate": 8.545895134069855e-05, "loss": 0.0319, "step": 22930 }, { "grad_norm": 0.3854862153530121, "learning_rate": 8.544437652095576e-05, "loss": 0.0329, "step": 22940 }, { "grad_norm": 0.6430736184120178, "learning_rate": 8.542979564473843e-05, "loss": 0.0456, "step": 22950 }, { "grad_norm": 0.32970330119132996, "learning_rate": 8.541520871453802e-05, "loss": 0.0321, "step": 22960 }, { "grad_norm": 0.41088947653770447, "learning_rate": 8.540061573284705e-05, "loss": 0.0278, "step": 22970 }, { "grad_norm": 0.35101887583732605, "learning_rate": 8.538601670215906e-05, "loss": 0.0319, "step": 22980 }, { "grad_norm": 0.3453889489173889, "learning_rate": 8.537141162496864e-05, "loss": 0.036, "step": 22990 }, { "grad_norm": 0.39382317662239075, "learning_rate": 8.535680050377137e-05, "loss": 0.0339, "step": 23000 }, { "grad_norm": 0.25481870770454407, "learning_rate": 8.534218334106391e-05, "loss": 0.0352, "step": 23010 }, { "grad_norm": 0.3666202425956726, "learning_rate": 8.532756013934393e-05, "loss": 0.0381, "step": 23020 }, { "grad_norm": 0.36304596066474915, "learning_rate": 8.531293090111012e-05, "loss": 0.0462, "step": 23030 }, { "grad_norm": 0.37155839800834656, "learning_rate": 8.529829562886225e-05, "loss": 0.0353, "step": 23040 }, { "grad_norm": 0.4161291718482971, "learning_rate": 8.528365432510105e-05, "loss": 0.0403, "step": 23050 }, { "grad_norm": 0.39452239871025085, "learning_rate": 8.526900699232833e-05, "loss": 0.0352, "step": 23060 }, { "grad_norm": 0.4362095594406128, "learning_rate": 8.525435363304695e-05, "loss": 0.0375, "step": 23070 }, { "grad_norm": 0.43777531385421753, "learning_rate": 8.523969424976072e-05, "loss": 0.0321, "step": 23080 }, { "grad_norm": 0.27540817856788635, "learning_rate": 8.522502884497457e-05, "loss": 0.0306, "step": 23090 }, { "grad_norm": 0.32398396730422974, "learning_rate": 8.521035742119437e-05, "loss": 0.0355, "step": 23100 }, { "grad_norm": 0.23921501636505127, "learning_rate": 8.519567998092712e-05, "loss": 0.047, "step": 23110 }, { "grad_norm": 0.3138614892959595, "learning_rate": 8.518099652668075e-05, "loss": 0.032, "step": 23120 }, { "grad_norm": 0.2793996334075928, "learning_rate": 8.516630706096429e-05, "loss": 0.0287, "step": 23130 }, { "grad_norm": 0.3560847043991089, "learning_rate": 8.515161158628773e-05, "loss": 0.0277, "step": 23140 }, { "grad_norm": 0.4104834794998169, "learning_rate": 8.513691010516216e-05, "loss": 0.0497, "step": 23150 }, { "grad_norm": 0.3239120543003082, "learning_rate": 8.512220262009966e-05, "loss": 0.0403, "step": 23160 }, { "grad_norm": 0.6201300621032715, "learning_rate": 8.510748913361332e-05, "loss": 0.0392, "step": 23170 }, { "grad_norm": 0.40128761529922485, "learning_rate": 8.509276964821726e-05, "loss": 0.0318, "step": 23180 }, { "grad_norm": 0.36462241411209106, "learning_rate": 8.507804416642669e-05, "loss": 0.0371, "step": 23190 }, { "grad_norm": 0.40799179673194885, "learning_rate": 8.506331269075774e-05, "loss": 0.0336, "step": 23200 }, { "grad_norm": 0.39343857765197754, "learning_rate": 8.504857522372765e-05, "loss": 0.0311, "step": 23210 }, { "grad_norm": 0.37223538756370544, "learning_rate": 8.503383176785461e-05, "loss": 0.0302, "step": 23220 }, { "grad_norm": 0.32973727583885193, "learning_rate": 8.501908232565792e-05, "loss": 0.0334, "step": 23230 }, { "grad_norm": 0.42907625436782837, "learning_rate": 8.50043268996578e-05, "loss": 0.0356, "step": 23240 }, { "grad_norm": 0.29899007081985474, "learning_rate": 8.498956549237562e-05, "loss": 0.0387, "step": 23250 }, { "grad_norm": 0.2838830351829529, "learning_rate": 8.497479810633366e-05, "loss": 0.0336, "step": 23260 }, { "grad_norm": 0.5141830444335938, "learning_rate": 8.496002474405525e-05, "loss": 0.0421, "step": 23270 }, { "grad_norm": 0.4322335720062256, "learning_rate": 8.494524540806478e-05, "loss": 0.0392, "step": 23280 }, { "grad_norm": 0.5642908215522766, "learning_rate": 8.493046010088761e-05, "loss": 0.0521, "step": 23290 }, { "grad_norm": 0.7264232039451599, "learning_rate": 8.491566882505018e-05, "loss": 0.0475, "step": 23300 }, { "grad_norm": 0.49027031660079956, "learning_rate": 8.490087158307988e-05, "loss": 0.0416, "step": 23310 }, { "grad_norm": 0.5451589822769165, "learning_rate": 8.488606837750518e-05, "loss": 0.0353, "step": 23320 }, { "grad_norm": 0.30956751108169556, "learning_rate": 8.487125921085552e-05, "loss": 0.0339, "step": 23330 }, { "grad_norm": 0.4668065905570984, "learning_rate": 8.485644408566141e-05, "loss": 0.0371, "step": 23340 }, { "grad_norm": 0.3635803461074829, "learning_rate": 8.484162300445431e-05, "loss": 0.036, "step": 23350 }, { "grad_norm": 0.4146682322025299, "learning_rate": 8.482679596976676e-05, "loss": 0.0316, "step": 23360 }, { "grad_norm": 0.32216718792915344, "learning_rate": 8.48119629841323e-05, "loss": 0.0419, "step": 23370 }, { "grad_norm": 0.4735075533390045, "learning_rate": 8.479712405008547e-05, "loss": 0.0465, "step": 23380 }, { "grad_norm": 0.4683929979801178, "learning_rate": 8.478227917016184e-05, "loss": 0.0485, "step": 23390 }, { "grad_norm": 0.2824096977710724, "learning_rate": 8.476742834689801e-05, "loss": 0.0397, "step": 23400 }, { "grad_norm": 0.36352863907814026, "learning_rate": 8.475257158283157e-05, "loss": 0.0438, "step": 23410 }, { "grad_norm": 0.4781700074672699, "learning_rate": 8.473770888050112e-05, "loss": 0.0308, "step": 23420 }, { "grad_norm": 0.5855337381362915, "learning_rate": 8.47228402424463e-05, "loss": 0.0396, "step": 23430 }, { "grad_norm": 0.4715762734413147, "learning_rate": 8.470796567120775e-05, "loss": 0.0336, "step": 23440 }, { "grad_norm": 0.4367011785507202, "learning_rate": 8.469308516932714e-05, "loss": 0.0343, "step": 23450 }, { "grad_norm": 0.5458894968032837, "learning_rate": 8.467819873934714e-05, "loss": 0.031, "step": 23460 }, { "grad_norm": 0.32714855670928955, "learning_rate": 8.466330638381143e-05, "loss": 0.0348, "step": 23470 }, { "grad_norm": 0.3258351981639862, "learning_rate": 8.464840810526469e-05, "loss": 0.0311, "step": 23480 }, { "grad_norm": 0.3832985460758209, "learning_rate": 8.463350390625264e-05, "loss": 0.0454, "step": 23490 }, { "grad_norm": 0.372287780046463, "learning_rate": 8.4618593789322e-05, "loss": 0.0641, "step": 23500 }, { "grad_norm": 0.4219245910644531, "learning_rate": 8.46036777570205e-05, "loss": 0.0443, "step": 23510 }, { "grad_norm": 0.40532153844833374, "learning_rate": 8.458875581189688e-05, "loss": 0.0594, "step": 23520 }, { "grad_norm": 0.3442855477333069, "learning_rate": 8.457382795650092e-05, "loss": 0.0374, "step": 23530 }, { "grad_norm": 0.3230403959751129, "learning_rate": 8.455889419338335e-05, "loss": 0.0401, "step": 23540 }, { "grad_norm": 0.31315138936042786, "learning_rate": 8.454395452509593e-05, "loss": 0.0341, "step": 23550 }, { "grad_norm": 0.2864880859851837, "learning_rate": 8.452900895419146e-05, "loss": 0.0362, "step": 23560 }, { "grad_norm": 0.2844725251197815, "learning_rate": 8.451405748322376e-05, "loss": 0.0386, "step": 23570 }, { "grad_norm": 0.3415977656841278, "learning_rate": 8.449910011474759e-05, "loss": 0.029, "step": 23580 }, { "grad_norm": 0.3417700529098511, "learning_rate": 8.448413685131876e-05, "loss": 0.0446, "step": 23590 }, { "grad_norm": 0.3909222185611725, "learning_rate": 8.446916769549407e-05, "loss": 0.0419, "step": 23600 }, { "grad_norm": 0.5145734548568726, "learning_rate": 8.445419264983136e-05, "loss": 0.0447, "step": 23610 }, { "grad_norm": 0.2967034876346588, "learning_rate": 8.443921171688947e-05, "loss": 0.0389, "step": 23620 }, { "grad_norm": 0.4325139820575714, "learning_rate": 8.442422489922819e-05, "loss": 0.0326, "step": 23630 }, { "grad_norm": 0.4521223306655884, "learning_rate": 8.440923219940838e-05, "loss": 0.0444, "step": 23640 }, { "grad_norm": 0.7590405941009521, "learning_rate": 8.439423361999189e-05, "loss": 0.0369, "step": 23650 }, { "grad_norm": 0.34646469354629517, "learning_rate": 8.437922916354155e-05, "loss": 0.0413, "step": 23660 }, { "grad_norm": 0.22165703773498535, "learning_rate": 8.436421883262123e-05, "loss": 0.034, "step": 23670 }, { "grad_norm": 0.3154849708080292, "learning_rate": 8.434920262979577e-05, "loss": 0.0457, "step": 23680 }, { "grad_norm": 0.33443015813827515, "learning_rate": 8.433418055763104e-05, "loss": 0.0329, "step": 23690 }, { "grad_norm": 0.37781593203544617, "learning_rate": 8.431915261869389e-05, "loss": 0.0476, "step": 23700 }, { "grad_norm": 0.28648093342781067, "learning_rate": 8.43041188155522e-05, "loss": 0.0352, "step": 23710 }, { "grad_norm": 0.34438619017601013, "learning_rate": 8.428907915077481e-05, "loss": 0.0327, "step": 23720 }, { "grad_norm": 0.3955234885215759, "learning_rate": 8.42740336269316e-05, "loss": 0.0386, "step": 23730 }, { "grad_norm": 0.7608140110969543, "learning_rate": 8.425898224659345e-05, "loss": 0.0347, "step": 23740 }, { "grad_norm": 0.344527006149292, "learning_rate": 8.42439250123322e-05, "loss": 0.0342, "step": 23750 }, { "grad_norm": 0.3834726810455322, "learning_rate": 8.422886192672076e-05, "loss": 0.0388, "step": 23760 }, { "grad_norm": 0.30270880460739136, "learning_rate": 8.421379299233297e-05, "loss": 0.0374, "step": 23770 }, { "grad_norm": 0.7403445839881897, "learning_rate": 8.419871821174371e-05, "loss": 0.0375, "step": 23780 }, { "grad_norm": 0.3811553716659546, "learning_rate": 8.418363758752884e-05, "loss": 0.0387, "step": 23790 }, { "grad_norm": 0.3257497549057007, "learning_rate": 8.416855112226523e-05, "loss": 0.0357, "step": 23800 }, { "grad_norm": 0.33047258853912354, "learning_rate": 8.415345881853075e-05, "loss": 0.034, "step": 23810 }, { "grad_norm": 0.37501686811447144, "learning_rate": 8.413836067890426e-05, "loss": 0.0319, "step": 23820 }, { "grad_norm": 1.3052563667297363, "learning_rate": 8.41232567059656e-05, "loss": 0.0376, "step": 23830 }, { "grad_norm": 0.3909584581851959, "learning_rate": 8.410814690229565e-05, "loss": 0.0435, "step": 23840 }, { "grad_norm": 0.4665144085884094, "learning_rate": 8.409303127047626e-05, "loss": 0.0389, "step": 23850 }, { "grad_norm": 0.44122377038002014, "learning_rate": 8.407790981309028e-05, "loss": 0.0385, "step": 23860 }, { "grad_norm": 0.41161152720451355, "learning_rate": 8.406278253272153e-05, "loss": 0.0462, "step": 23870 }, { "grad_norm": 0.41591256856918335, "learning_rate": 8.404764943195487e-05, "loss": 0.0347, "step": 23880 }, { "grad_norm": 0.24823400378227234, "learning_rate": 8.403251051337613e-05, "loss": 0.044, "step": 23890 }, { "grad_norm": 0.316142201423645, "learning_rate": 8.401736577957214e-05, "loss": 0.0338, "step": 23900 }, { "grad_norm": 0.36774295568466187, "learning_rate": 8.40022152331307e-05, "loss": 0.037, "step": 23910 }, { "grad_norm": 0.39602652192115784, "learning_rate": 8.398705887664064e-05, "loss": 0.0404, "step": 23920 }, { "grad_norm": 0.2688405215740204, "learning_rate": 8.397189671269177e-05, "loss": 0.0295, "step": 23930 }, { "grad_norm": 0.4082188904285431, "learning_rate": 8.395672874387488e-05, "loss": 0.0352, "step": 23940 }, { "grad_norm": 0.3428686857223511, "learning_rate": 8.394155497278177e-05, "loss": 0.0295, "step": 23950 }, { "grad_norm": 0.6832461357116699, "learning_rate": 8.392637540200523e-05, "loss": 0.0427, "step": 23960 }, { "grad_norm": 0.25173047184944153, "learning_rate": 8.391119003413902e-05, "loss": 0.0333, "step": 23970 }, { "grad_norm": 0.30446410179138184, "learning_rate": 8.38959988717779e-05, "loss": 0.0344, "step": 23980 }, { "grad_norm": 0.3221684396266937, "learning_rate": 8.388080191751764e-05, "loss": 0.0383, "step": 23990 }, { "grad_norm": 0.3188439905643463, "learning_rate": 8.386559917395496e-05, "loss": 0.0333, "step": 24000 }, { "grad_norm": 0.2781255543231964, "learning_rate": 8.385039064368761e-05, "loss": 0.0263, "step": 24010 }, { "grad_norm": 0.27797263860702515, "learning_rate": 8.383517632931431e-05, "loss": 0.0284, "step": 24020 }, { "grad_norm": 0.30799350142478943, "learning_rate": 8.381995623343477e-05, "loss": 0.0307, "step": 24030 }, { "grad_norm": 0.4939205050468445, "learning_rate": 8.380473035864968e-05, "loss": 0.039, "step": 24040 }, { "grad_norm": 0.37379124760627747, "learning_rate": 8.378949870756076e-05, "loss": 0.0392, "step": 24050 }, { "grad_norm": 0.2350614070892334, "learning_rate": 8.377426128277063e-05, "loss": 0.0344, "step": 24060 }, { "grad_norm": 0.29163047671318054, "learning_rate": 8.375901808688298e-05, "loss": 0.027, "step": 24070 }, { "grad_norm": 0.48131808638572693, "learning_rate": 8.374376912250246e-05, "loss": 0.0414, "step": 24080 }, { "grad_norm": 0.31003662943840027, "learning_rate": 8.372851439223468e-05, "loss": 0.0324, "step": 24090 }, { "grad_norm": 0.3472422659397125, "learning_rate": 8.371325389868627e-05, "loss": 0.0417, "step": 24100 }, { "grad_norm": 0.35672780871391296, "learning_rate": 8.369798764446482e-05, "loss": 0.0336, "step": 24110 }, { "grad_norm": 0.4617539346218109, "learning_rate": 8.368271563217893e-05, "loss": 0.0321, "step": 24120 }, { "grad_norm": 0.3513234257698059, "learning_rate": 8.366743786443817e-05, "loss": 0.0479, "step": 24130 }, { "grad_norm": 0.44865378737449646, "learning_rate": 8.365215434385309e-05, "loss": 0.041, "step": 24140 }, { "grad_norm": 0.2971499264240265, "learning_rate": 8.36368650730352e-05, "loss": 0.0324, "step": 24150 }, { "grad_norm": 0.38706761598587036, "learning_rate": 8.362157005459705e-05, "loss": 0.039, "step": 24160 }, { "grad_norm": 0.2903778851032257, "learning_rate": 8.360626929115213e-05, "loss": 0.0307, "step": 24170 }, { "grad_norm": 0.3287951350212097, "learning_rate": 8.359096278531492e-05, "loss": 0.0368, "step": 24180 }, { "grad_norm": 0.2929770052433014, "learning_rate": 8.357565053970088e-05, "loss": 0.0292, "step": 24190 }, { "grad_norm": 0.3400157690048218, "learning_rate": 8.356033255692647e-05, "loss": 0.0333, "step": 24200 }, { "grad_norm": 0.3288123607635498, "learning_rate": 8.354500883960911e-05, "loss": 0.0284, "step": 24210 }, { "grad_norm": 0.33333146572113037, "learning_rate": 8.352967939036717e-05, "loss": 0.0383, "step": 24220 }, { "grad_norm": 0.3813519775867462, "learning_rate": 8.35143442118201e-05, "loss": 0.0364, "step": 24230 }, { "grad_norm": 0.2199416607618332, "learning_rate": 8.349900330658819e-05, "loss": 0.0334, "step": 24240 }, { "grad_norm": 0.3823058307170868, "learning_rate": 8.348365667729284e-05, "loss": 0.0336, "step": 24250 }, { "grad_norm": 0.3585173189640045, "learning_rate": 8.346830432655633e-05, "loss": 0.0252, "step": 24260 }, { "grad_norm": 0.4745064973831177, "learning_rate": 8.345294625700195e-05, "loss": 0.036, "step": 24270 }, { "grad_norm": 0.3815319240093231, "learning_rate": 8.343758247125402e-05, "loss": 0.0299, "step": 24280 }, { "grad_norm": 0.46224960684776306, "learning_rate": 8.342221297193776e-05, "loss": 0.0438, "step": 24290 }, { "grad_norm": 0.35279154777526855, "learning_rate": 8.34068377616794e-05, "loss": 0.0369, "step": 24300 }, { "grad_norm": 0.5841526389122009, "learning_rate": 8.339145684310615e-05, "loss": 0.0359, "step": 24310 }, { "grad_norm": 0.7785378098487854, "learning_rate": 8.337607021884618e-05, "loss": 0.0363, "step": 24320 }, { "grad_norm": 0.34956103563308716, "learning_rate": 8.336067789152867e-05, "loss": 0.0292, "step": 24330 }, { "grad_norm": 0.4703511595726013, "learning_rate": 8.334527986378369e-05, "loss": 0.0421, "step": 24340 }, { "grad_norm": 0.36453118920326233, "learning_rate": 8.332987613824239e-05, "loss": 0.0427, "step": 24350 }, { "grad_norm": 0.3630222976207733, "learning_rate": 8.331446671753685e-05, "loss": 0.0442, "step": 24360 }, { "grad_norm": 0.3023649752140045, "learning_rate": 8.329905160430007e-05, "loss": 0.0323, "step": 24370 }, { "grad_norm": 0.3862791061401367, "learning_rate": 8.328363080116611e-05, "loss": 0.0347, "step": 24380 }, { "grad_norm": 0.3037950098514557, "learning_rate": 8.326820431076997e-05, "loss": 0.0301, "step": 24390 }, { "grad_norm": 0.41693079471588135, "learning_rate": 8.325277213574759e-05, "loss": 0.036, "step": 24400 }, { "grad_norm": 0.5030235052108765, "learning_rate": 8.32373342787359e-05, "loss": 0.0357, "step": 24410 }, { "grad_norm": 0.8779670596122742, "learning_rate": 8.322189074237285e-05, "loss": 0.0357, "step": 24420 }, { "grad_norm": 0.2783895432949066, "learning_rate": 8.32064415292973e-05, "loss": 0.0326, "step": 24430 }, { "grad_norm": 0.3653310239315033, "learning_rate": 8.319098664214907e-05, "loss": 0.0292, "step": 24440 }, { "grad_norm": 0.31100448966026306, "learning_rate": 8.3175526083569e-05, "loss": 0.039, "step": 24450 }, { "grad_norm": 0.40006139874458313, "learning_rate": 8.316005985619889e-05, "loss": 0.0371, "step": 24460 }, { "grad_norm": 0.32423484325408936, "learning_rate": 8.314458796268147e-05, "loss": 0.0362, "step": 24470 }, { "grad_norm": 0.2794496715068817, "learning_rate": 8.312911040566047e-05, "loss": 0.0296, "step": 24480 }, { "grad_norm": 0.3847601115703583, "learning_rate": 8.31136271877806e-05, "loss": 0.0347, "step": 24490 }, { "grad_norm": 0.4395477771759033, "learning_rate": 8.309813831168748e-05, "loss": 0.0375, "step": 24500 }, { "grad_norm": 0.3733198046684265, "learning_rate": 8.308264378002777e-05, "loss": 0.0264, "step": 24510 }, { "grad_norm": 0.40919268131256104, "learning_rate": 8.306714359544906e-05, "loss": 0.0396, "step": 24520 }, { "grad_norm": 0.4525337219238281, "learning_rate": 8.30516377605999e-05, "loss": 0.029, "step": 24530 }, { "grad_norm": 0.30689430236816406, "learning_rate": 8.30361262781298e-05, "loss": 0.0291, "step": 24540 }, { "grad_norm": 0.4305822551250458, "learning_rate": 8.302060915068924e-05, "loss": 0.0405, "step": 24550 }, { "grad_norm": 0.39419615268707275, "learning_rate": 8.300508638092972e-05, "loss": 0.0279, "step": 24560 }, { "grad_norm": 0.7873024344444275, "learning_rate": 8.298955797150361e-05, "loss": 0.0331, "step": 24570 }, { "grad_norm": 0.5440157651901245, "learning_rate": 8.297402392506433e-05, "loss": 0.0355, "step": 24580 }, { "grad_norm": 0.3555432856082916, "learning_rate": 8.295848424426617e-05, "loss": 0.0325, "step": 24590 }, { "grad_norm": 0.29331493377685547, "learning_rate": 8.29429389317645e-05, "loss": 0.0293, "step": 24600 }, { "grad_norm": 0.313083678483963, "learning_rate": 8.292738799021556e-05, "loss": 0.0386, "step": 24610 }, { "grad_norm": 0.36193880438804626, "learning_rate": 8.291183142227656e-05, "loss": 0.0302, "step": 24620 }, { "grad_norm": 0.28286558389663696, "learning_rate": 8.289626923060572e-05, "loss": 0.0438, "step": 24630 }, { "grad_norm": 0.4239165484905243, "learning_rate": 8.288070141786218e-05, "loss": 0.0364, "step": 24640 }, { "grad_norm": 0.29508495330810547, "learning_rate": 8.286512798670605e-05, "loss": 0.0359, "step": 24650 }, { "grad_norm": 0.4233156740665436, "learning_rate": 8.284954893979842e-05, "loss": 0.0394, "step": 24660 }, { "grad_norm": 0.2510438859462738, "learning_rate": 8.283396427980131e-05, "loss": 0.038, "step": 24670 }, { "grad_norm": 0.3623015284538269, "learning_rate": 8.281837400937771e-05, "loss": 0.0383, "step": 24680 }, { "grad_norm": 0.4119403660297394, "learning_rate": 8.28027781311916e-05, "loss": 0.0319, "step": 24690 }, { "grad_norm": 0.3292015492916107, "learning_rate": 8.278717664790785e-05, "loss": 0.0326, "step": 24700 }, { "grad_norm": 0.4924238920211792, "learning_rate": 8.277156956219234e-05, "loss": 0.0351, "step": 24710 }, { "grad_norm": 0.499586820602417, "learning_rate": 8.275595687671189e-05, "loss": 0.0299, "step": 24720 }, { "grad_norm": 0.5844864845275879, "learning_rate": 8.27403385941343e-05, "loss": 0.0352, "step": 24730 }, { "grad_norm": 0.6847178936004639, "learning_rate": 8.272471471712828e-05, "loss": 0.0403, "step": 24740 }, { "grad_norm": 0.2375834733247757, "learning_rate": 8.270908524836355e-05, "loss": 0.0368, "step": 24750 }, { "grad_norm": 0.36646702885627747, "learning_rate": 8.269345019051074e-05, "loss": 0.045, "step": 24760 }, { "grad_norm": 0.37710678577423096, "learning_rate": 8.267780954624147e-05, "loss": 0.0383, "step": 24770 }, { "grad_norm": 0.3237892687320709, "learning_rate": 8.266216331822827e-05, "loss": 0.0337, "step": 24780 }, { "grad_norm": 0.4002845585346222, "learning_rate": 8.264651150914469e-05, "loss": 0.0353, "step": 24790 }, { "grad_norm": 0.2783460021018982, "learning_rate": 8.263085412166517e-05, "loss": 0.0296, "step": 24800 }, { "grad_norm": 0.32585620880126953, "learning_rate": 8.261519115846514e-05, "loss": 0.0301, "step": 24810 }, { "grad_norm": 0.32856687903404236, "learning_rate": 8.259952262222096e-05, "loss": 0.0296, "step": 24820 }, { "grad_norm": 0.3375509977340698, "learning_rate": 8.258384851560997e-05, "loss": 0.0334, "step": 24830 }, { "grad_norm": 0.35823988914489746, "learning_rate": 8.256816884131044e-05, "loss": 0.0402, "step": 24840 }, { "grad_norm": 0.3565436005592346, "learning_rate": 8.255248360200159e-05, "loss": 0.0314, "step": 24850 }, { "grad_norm": 0.34052425622940063, "learning_rate": 8.253679280036359e-05, "loss": 0.0285, "step": 24860 }, { "grad_norm": 0.5294675827026367, "learning_rate": 8.252109643907762e-05, "loss": 0.051, "step": 24870 }, { "grad_norm": 0.34172821044921875, "learning_rate": 8.250539452082569e-05, "loss": 0.0311, "step": 24880 }, { "grad_norm": 0.39280185103416443, "learning_rate": 8.248968704829087e-05, "loss": 0.0318, "step": 24890 }, { "grad_norm": 0.3898251950740814, "learning_rate": 8.247397402415714e-05, "loss": 0.0373, "step": 24900 }, { "grad_norm": 0.34394946694374084, "learning_rate": 8.24582554511094e-05, "loss": 0.0389, "step": 24910 }, { "grad_norm": 0.32048821449279785, "learning_rate": 8.244253133183355e-05, "loss": 0.0309, "step": 24920 }, { "grad_norm": 0.2870766520500183, "learning_rate": 8.24268016690164e-05, "loss": 0.0365, "step": 24930 }, { "grad_norm": 0.4661267101764679, "learning_rate": 8.241106646534571e-05, "loss": 0.0369, "step": 24940 }, { "grad_norm": 0.3222093880176544, "learning_rate": 8.23953257235102e-05, "loss": 0.033, "step": 24950 }, { "grad_norm": 0.5076224207878113, "learning_rate": 8.237957944619956e-05, "loss": 0.0385, "step": 24960 }, { "grad_norm": 0.44340890645980835, "learning_rate": 8.236382763610437e-05, "loss": 0.0336, "step": 24970 }, { "grad_norm": 0.4388119578361511, "learning_rate": 8.234807029591619e-05, "loss": 0.0372, "step": 24980 }, { "grad_norm": 0.35912179946899414, "learning_rate": 8.233230742832752e-05, "loss": 0.0468, "step": 24990 }, { "grad_norm": 0.44867244362831116, "learning_rate": 8.231653903603178e-05, "loss": 0.0398, "step": 25000 }, { "grad_norm": 0.4095630645751953, "learning_rate": 8.23007651217234e-05, "loss": 0.0329, "step": 25010 }, { "grad_norm": 0.3589971959590912, "learning_rate": 8.228498568809769e-05, "loss": 0.0337, "step": 25020 }, { "grad_norm": 0.4331638813018799, "learning_rate": 8.22692007378509e-05, "loss": 0.0421, "step": 25030 }, { "grad_norm": 0.3234155774116516, "learning_rate": 8.225341027368028e-05, "loss": 0.0319, "step": 25040 }, { "grad_norm": 0.31930291652679443, "learning_rate": 8.223761429828399e-05, "loss": 0.0321, "step": 25050 }, { "grad_norm": 0.33364182710647583, "learning_rate": 8.22218128143611e-05, "loss": 0.0376, "step": 25060 }, { "grad_norm": 0.3685622811317444, "learning_rate": 8.220600582461166e-05, "loss": 0.0304, "step": 25070 }, { "grad_norm": 0.28735852241516113, "learning_rate": 8.219019333173668e-05, "loss": 0.0366, "step": 25080 }, { "grad_norm": 0.4844459295272827, "learning_rate": 8.217437533843805e-05, "loss": 0.0418, "step": 25090 }, { "grad_norm": 0.2527235746383667, "learning_rate": 8.215855184741867e-05, "loss": 0.0321, "step": 25100 }, { "grad_norm": 0.2980116307735443, "learning_rate": 8.21427228613823e-05, "loss": 0.04, "step": 25110 }, { "grad_norm": 0.3709414601325989, "learning_rate": 8.21268883830337e-05, "loss": 0.0385, "step": 25120 }, { "grad_norm": 0.43692904710769653, "learning_rate": 8.211104841507855e-05, "loss": 0.0418, "step": 25130 }, { "grad_norm": 0.30557578802108765, "learning_rate": 8.209520296022346e-05, "loss": 0.0361, "step": 25140 }, { "grad_norm": 0.3559313714504242, "learning_rate": 8.207935202117599e-05, "loss": 0.0323, "step": 25150 }, { "grad_norm": 0.27159857749938965, "learning_rate": 8.206349560064463e-05, "loss": 0.0319, "step": 25160 }, { "grad_norm": 0.3567597270011902, "learning_rate": 8.204763370133881e-05, "loss": 0.0285, "step": 25170 }, { "grad_norm": 0.3170469403266907, "learning_rate": 8.203176632596892e-05, "loss": 0.0307, "step": 25180 }, { "grad_norm": 0.5984862446784973, "learning_rate": 8.20158934772462e-05, "loss": 0.0371, "step": 25190 }, { "grad_norm": 0.29854053258895874, "learning_rate": 8.200001515788294e-05, "loss": 0.0313, "step": 25200 }, { "grad_norm": 0.3647870123386383, "learning_rate": 8.198413137059228e-05, "loss": 0.0365, "step": 25210 }, { "grad_norm": 0.2591680586338043, "learning_rate": 8.196824211808835e-05, "loss": 0.0315, "step": 25220 }, { "grad_norm": 0.3908758759498596, "learning_rate": 8.195234740308617e-05, "loss": 0.0395, "step": 25230 }, { "grad_norm": 0.3345467150211334, "learning_rate": 8.193644722830171e-05, "loss": 0.0273, "step": 25240 }, { "grad_norm": 0.6677744388580322, "learning_rate": 8.19205415964519e-05, "loss": 0.0297, "step": 25250 }, { "grad_norm": 0.30437031388282776, "learning_rate": 8.190463051025456e-05, "loss": 0.0376, "step": 25260 }, { "grad_norm": 0.644271194934845, "learning_rate": 8.188871397242843e-05, "loss": 0.0408, "step": 25270 }, { "grad_norm": 0.2941782474517822, "learning_rate": 8.187279198569326e-05, "loss": 0.0318, "step": 25280 }, { "grad_norm": 0.360895574092865, "learning_rate": 8.185686455276966e-05, "loss": 0.0383, "step": 25290 }, { "grad_norm": 0.3023626208305359, "learning_rate": 8.184093167637921e-05, "loss": 0.0398, "step": 25300 }, { "grad_norm": 0.3028942346572876, "learning_rate": 8.182499335924437e-05, "loss": 0.0376, "step": 25310 }, { "grad_norm": 0.5410653948783875, "learning_rate": 8.18090496040886e-05, "loss": 0.0316, "step": 25320 }, { "grad_norm": 0.2933052182197571, "learning_rate": 8.179310041363621e-05, "loss": 0.0367, "step": 25330 }, { "grad_norm": 0.47636643052101135, "learning_rate": 8.17771457906125e-05, "loss": 0.0283, "step": 25340 }, { "grad_norm": 0.4600982666015625, "learning_rate": 8.176118573774371e-05, "loss": 0.0336, "step": 25350 }, { "grad_norm": 0.4442882835865021, "learning_rate": 8.174522025775692e-05, "loss": 0.0363, "step": 25360 }, { "grad_norm": 0.4357227385044098, "learning_rate": 8.172924935338022e-05, "loss": 0.0349, "step": 25370 }, { "grad_norm": 0.5986688137054443, "learning_rate": 8.171327302734262e-05, "loss": 0.0381, "step": 25380 }, { "grad_norm": 0.3262079060077667, "learning_rate": 8.169729128237401e-05, "loss": 0.0276, "step": 25390 }, { "grad_norm": 0.3378647267818451, "learning_rate": 8.168130412120525e-05, "loss": 0.0312, "step": 25400 }, { "grad_norm": 0.3586094081401825, "learning_rate": 8.16653115465681e-05, "loss": 0.0384, "step": 25410 }, { "grad_norm": 0.3525805175304413, "learning_rate": 8.164931356119526e-05, "loss": 0.0297, "step": 25420 }, { "grad_norm": 0.4815495014190674, "learning_rate": 8.163331016782032e-05, "loss": 0.0417, "step": 25430 }, { "grad_norm": 0.37505632638931274, "learning_rate": 8.161730136917785e-05, "loss": 0.025, "step": 25440 }, { "grad_norm": 0.3258843421936035, "learning_rate": 8.160128716800333e-05, "loss": 0.0381, "step": 25450 }, { "grad_norm": 0.4431929290294647, "learning_rate": 8.158526756703313e-05, "loss": 0.03, "step": 25460 }, { "grad_norm": 0.38523054122924805, "learning_rate": 8.156924256900455e-05, "loss": 0.038, "step": 25470 }, { "grad_norm": 0.4087700843811035, "learning_rate": 8.155321217665584e-05, "loss": 0.0313, "step": 25480 }, { "grad_norm": 0.4382791817188263, "learning_rate": 8.153717639272614e-05, "loss": 0.0293, "step": 25490 }, { "grad_norm": 0.4541074335575104, "learning_rate": 8.152113521995555e-05, "loss": 0.0358, "step": 25500 }, { "grad_norm": 0.41395100951194763, "learning_rate": 8.150508866108505e-05, "loss": 0.0366, "step": 25510 }, { "grad_norm": 0.4635264575481415, "learning_rate": 8.148903671885657e-05, "loss": 0.031, "step": 25520 }, { "grad_norm": 0.37769216299057007, "learning_rate": 8.147297939601292e-05, "loss": 0.0349, "step": 25530 }, { "grad_norm": 0.46615099906921387, "learning_rate": 8.145691669529792e-05, "loss": 0.0344, "step": 25540 }, { "grad_norm": 0.3901209831237793, "learning_rate": 8.144084861945618e-05, "loss": 0.0307, "step": 25550 }, { "grad_norm": 0.35780391097068787, "learning_rate": 8.142477517123333e-05, "loss": 0.0322, "step": 25560 }, { "grad_norm": 0.3512577712535858, "learning_rate": 8.140869635337586e-05, "loss": 0.0363, "step": 25570 }, { "grad_norm": 0.32157355546951294, "learning_rate": 8.139261216863123e-05, "loss": 0.0365, "step": 25580 }, { "grad_norm": 0.25233834981918335, "learning_rate": 8.137652261974776e-05, "loss": 0.0306, "step": 25590 }, { "grad_norm": 0.38490888476371765, "learning_rate": 8.136042770947472e-05, "loss": 0.0326, "step": 25600 }, { "grad_norm": 0.34444060921669006, "learning_rate": 8.134432744056228e-05, "loss": 0.0368, "step": 25610 }, { "grad_norm": 0.38560640811920166, "learning_rate": 8.132822181576158e-05, "loss": 0.0365, "step": 25620 }, { "grad_norm": 0.4004804790019989, "learning_rate": 8.131211083782459e-05, "loss": 0.0355, "step": 25630 }, { "grad_norm": 0.3750070631504059, "learning_rate": 8.129599450950424e-05, "loss": 0.0362, "step": 25640 }, { "grad_norm": 0.37162622809410095, "learning_rate": 8.127987283355438e-05, "loss": 0.041, "step": 25650 }, { "grad_norm": 0.2475987672805786, "learning_rate": 8.126374581272976e-05, "loss": 0.0234, "step": 25660 }, { "grad_norm": 0.3285790681838989, "learning_rate": 8.124761344978605e-05, "loss": 0.0459, "step": 25670 }, { "grad_norm": 0.37891969084739685, "learning_rate": 8.12314757474798e-05, "loss": 0.0295, "step": 25680 }, { "grad_norm": 0.29584604501724243, "learning_rate": 8.121533270856856e-05, "loss": 0.0396, "step": 25690 }, { "grad_norm": 0.524185836315155, "learning_rate": 8.119918433581069e-05, "loss": 0.032, "step": 25700 }, { "grad_norm": 0.46507176756858826, "learning_rate": 8.118303063196551e-05, "loss": 0.0344, "step": 25710 }, { "grad_norm": 0.33354857563972473, "learning_rate": 8.116687159979326e-05, "loss": 0.0327, "step": 25720 }, { "grad_norm": 0.3510400354862213, "learning_rate": 8.115070724205508e-05, "loss": 0.0385, "step": 25730 }, { "grad_norm": 0.4972337782382965, "learning_rate": 8.113453756151296e-05, "loss": 0.0385, "step": 25740 }, { "grad_norm": 0.6852967739105225, "learning_rate": 8.111836256092995e-05, "loss": 0.0454, "step": 25750 }, { "grad_norm": 0.3429863750934601, "learning_rate": 8.110218224306985e-05, "loss": 0.0341, "step": 25760 }, { "grad_norm": 0.5574705004692078, "learning_rate": 8.108599661069745e-05, "loss": 0.034, "step": 25770 }, { "grad_norm": 0.454738974571228, "learning_rate": 8.106980566657845e-05, "loss": 0.0276, "step": 25780 }, { "grad_norm": 0.26731792092323303, "learning_rate": 8.10536094134794e-05, "loss": 0.0299, "step": 25790 }, { "grad_norm": 0.254260390996933, "learning_rate": 8.103740785416783e-05, "loss": 0.0313, "step": 25800 }, { "grad_norm": 0.43704065680503845, "learning_rate": 8.102120099141212e-05, "loss": 0.0293, "step": 25810 }, { "grad_norm": 0.3646673560142517, "learning_rate": 8.100498882798163e-05, "loss": 0.0315, "step": 25820 }, { "grad_norm": 0.5753058195114136, "learning_rate": 8.09887713666465e-05, "loss": 0.0347, "step": 25830 }, { "grad_norm": 0.34152185916900635, "learning_rate": 8.09725486101779e-05, "loss": 0.0359, "step": 25840 }, { "grad_norm": 0.3301008939743042, "learning_rate": 8.095632056134784e-05, "loss": 0.0356, "step": 25850 }, { "grad_norm": 0.37893030047416687, "learning_rate": 8.094008722292925e-05, "loss": 0.0328, "step": 25860 }, { "grad_norm": 0.3014223277568817, "learning_rate": 8.092384859769598e-05, "loss": 0.0327, "step": 25870 }, { "grad_norm": 0.4197273552417755, "learning_rate": 8.090760468842275e-05, "loss": 0.044, "step": 25880 }, { "grad_norm": 0.27131375670433044, "learning_rate": 8.089135549788521e-05, "loss": 0.0351, "step": 25890 }, { "grad_norm": 0.3442223370075226, "learning_rate": 8.087510102885987e-05, "loss": 0.0319, "step": 25900 }, { "grad_norm": 0.6033420562744141, "learning_rate": 8.085884128412422e-05, "loss": 0.0364, "step": 25910 }, { "grad_norm": 0.47067922353744507, "learning_rate": 8.084257626645659e-05, "loss": 0.0326, "step": 25920 }, { "grad_norm": 0.5201683640480042, "learning_rate": 8.08263059786362e-05, "loss": 0.04, "step": 25930 }, { "grad_norm": 0.48401159048080444, "learning_rate": 8.081003042344325e-05, "loss": 0.0325, "step": 25940 }, { "grad_norm": 0.3288896083831787, "learning_rate": 8.079374960365872e-05, "loss": 0.0339, "step": 25950 }, { "grad_norm": 0.4034730792045593, "learning_rate": 8.077746352206463e-05, "loss": 0.0308, "step": 25960 }, { "grad_norm": 0.43723833560943604, "learning_rate": 8.076117218144377e-05, "loss": 0.0322, "step": 25970 }, { "grad_norm": 0.5145173072814941, "learning_rate": 8.074487558457991e-05, "loss": 0.0308, "step": 25980 }, { "grad_norm": 0.34376004338264465, "learning_rate": 8.072857373425768e-05, "loss": 0.0321, "step": 25990 }, { "grad_norm": 0.49729257822036743, "learning_rate": 8.071226663326264e-05, "loss": 0.0303, "step": 26000 }, { "grad_norm": 0.37969422340393066, "learning_rate": 8.069595428438121e-05, "loss": 0.0384, "step": 26010 }, { "grad_norm": 0.3363208770751953, "learning_rate": 8.067963669040072e-05, "loss": 0.0379, "step": 26020 }, { "grad_norm": 0.5609307885169983, "learning_rate": 8.066331385410942e-05, "loss": 0.035, "step": 26030 }, { "grad_norm": 0.3601725101470947, "learning_rate": 8.064698577829641e-05, "loss": 0.0359, "step": 26040 }, { "grad_norm": 0.3292330801486969, "learning_rate": 8.063065246575175e-05, "loss": 0.035, "step": 26050 }, { "grad_norm": 0.3928469717502594, "learning_rate": 8.061431391926631e-05, "loss": 0.0321, "step": 26060 }, { "grad_norm": 0.20684756338596344, "learning_rate": 8.059797014163195e-05, "loss": 0.0334, "step": 26070 }, { "grad_norm": 0.40634796023368835, "learning_rate": 8.058162113564133e-05, "loss": 0.0449, "step": 26080 }, { "grad_norm": 0.41459178924560547, "learning_rate": 8.056526690408806e-05, "loss": 0.0379, "step": 26090 }, { "grad_norm": 0.5515298843383789, "learning_rate": 8.054890744976666e-05, "loss": 0.0332, "step": 26100 }, { "grad_norm": 0.2979048490524292, "learning_rate": 8.053254277547248e-05, "loss": 0.0253, "step": 26110 }, { "grad_norm": 0.3832165002822876, "learning_rate": 8.051617288400182e-05, "loss": 0.0439, "step": 26120 }, { "grad_norm": 0.2852690815925598, "learning_rate": 8.049979777815182e-05, "loss": 0.0335, "step": 26130 }, { "grad_norm": 0.34875789284706116, "learning_rate": 8.048341746072054e-05, "loss": 0.0323, "step": 26140 }, { "grad_norm": 0.3596162497997284, "learning_rate": 8.046703193450696e-05, "loss": 0.04, "step": 26150 }, { "grad_norm": 0.6126401424407959, "learning_rate": 8.04506412023109e-05, "loss": 0.0333, "step": 26160 }, { "grad_norm": 0.30113479495048523, "learning_rate": 8.043424526693306e-05, "loss": 0.0318, "step": 26170 }, { "grad_norm": 0.6364549398422241, "learning_rate": 8.04178441311751e-05, "loss": 0.0324, "step": 26180 }, { "grad_norm": 0.41675877571105957, "learning_rate": 8.04014377978395e-05, "loss": 0.0301, "step": 26190 }, { "grad_norm": 0.5165632367134094, "learning_rate": 8.038502626972967e-05, "loss": 0.0497, "step": 26200 }, { "grad_norm": 0.36396756768226624, "learning_rate": 8.036860954964989e-05, "loss": 0.0381, "step": 26210 }, { "grad_norm": 0.4060097634792328, "learning_rate": 8.035218764040531e-05, "loss": 0.0348, "step": 26220 }, { "grad_norm": 0.4334065020084381, "learning_rate": 8.033576054480199e-05, "loss": 0.0408, "step": 26230 }, { "grad_norm": 0.33453649282455444, "learning_rate": 8.031932826564688e-05, "loss": 0.0324, "step": 26240 }, { "grad_norm": 0.3332372307777405, "learning_rate": 8.030289080574782e-05, "loss": 0.0371, "step": 26250 }, { "grad_norm": 0.36303701996803284, "learning_rate": 8.028644816791349e-05, "loss": 0.0398, "step": 26260 }, { "grad_norm": 0.4381384551525116, "learning_rate": 8.027000035495351e-05, "loss": 0.0289, "step": 26270 }, { "grad_norm": 0.581272542476654, "learning_rate": 8.025354736967836e-05, "loss": 0.0437, "step": 26280 }, { "grad_norm": 0.27529552578926086, "learning_rate": 8.023708921489941e-05, "loss": 0.0433, "step": 26290 }, { "grad_norm": 0.5051521062850952, "learning_rate": 8.022062589342887e-05, "loss": 0.0431, "step": 26300 }, { "grad_norm": 0.26875314116477966, "learning_rate": 8.020415740807993e-05, "loss": 0.0434, "step": 26310 }, { "grad_norm": 0.3216550052165985, "learning_rate": 8.018768376166656e-05, "loss": 0.0351, "step": 26320 }, { "grad_norm": 0.4134422242641449, "learning_rate": 8.017120495700368e-05, "loss": 0.0364, "step": 26330 }, { "grad_norm": 0.5285114645957947, "learning_rate": 8.015472099690704e-05, "loss": 0.0332, "step": 26340 }, { "grad_norm": 0.21064426004886627, "learning_rate": 8.013823188419332e-05, "loss": 0.0284, "step": 26350 }, { "grad_norm": 0.32244783639907837, "learning_rate": 8.012173762168006e-05, "loss": 0.034, "step": 26360 }, { "grad_norm": 0.3433443307876587, "learning_rate": 8.010523821218567e-05, "loss": 0.029, "step": 26370 }, { "grad_norm": 0.3551679849624634, "learning_rate": 8.008873365852945e-05, "loss": 0.0318, "step": 26380 }, { "grad_norm": 0.26909106969833374, "learning_rate": 8.007222396353157e-05, "loss": 0.0342, "step": 26390 }, { "grad_norm": 0.40777263045310974, "learning_rate": 8.00557091300131e-05, "loss": 0.0324, "step": 26400 }, { "grad_norm": 0.3883471190929413, "learning_rate": 8.003918916079597e-05, "loss": 0.0313, "step": 26410 }, { "grad_norm": 0.30645906925201416, "learning_rate": 8.002266405870298e-05, "loss": 0.0283, "step": 26420 }, { "grad_norm": 0.2827458083629608, "learning_rate": 8.000613382655782e-05, "loss": 0.0397, "step": 26430 }, { "grad_norm": 0.37868815660476685, "learning_rate": 7.998959846718505e-05, "loss": 0.0378, "step": 26440 }, { "grad_norm": 0.34003084897994995, "learning_rate": 7.997305798341012e-05, "loss": 0.0394, "step": 26450 }, { "grad_norm": 0.40252014994621277, "learning_rate": 7.995651237805937e-05, "loss": 0.037, "step": 26460 }, { "grad_norm": 0.37489446997642517, "learning_rate": 7.993996165395996e-05, "loss": 0.0276, "step": 26470 }, { "grad_norm": 0.39521002769470215, "learning_rate": 7.992340581393996e-05, "loss": 0.0358, "step": 26480 }, { "grad_norm": 0.302516371011734, "learning_rate": 7.990684486082831e-05, "loss": 0.0334, "step": 26490 }, { "grad_norm": 0.46536335349082947, "learning_rate": 7.989027879745482e-05, "loss": 0.0368, "step": 26500 }, { "grad_norm": 0.289036363363266, "learning_rate": 7.98737076266502e-05, "loss": 0.0385, "step": 26510 }, { "grad_norm": 0.25536441802978516, "learning_rate": 7.985713135124598e-05, "loss": 0.0333, "step": 26520 }, { "grad_norm": 0.3645716607570648, "learning_rate": 7.98405499740746e-05, "loss": 0.0327, "step": 26530 }, { "grad_norm": 0.31978148221969604, "learning_rate": 7.98239634979694e-05, "loss": 0.0332, "step": 26540 }, { "grad_norm": 0.40436360239982605, "learning_rate": 7.98073719257645e-05, "loss": 0.0309, "step": 26550 }, { "grad_norm": 0.3343738317489624, "learning_rate": 7.979077526029499e-05, "loss": 0.0358, "step": 26560 }, { "grad_norm": 0.22246359288692474, "learning_rate": 7.977417350439675e-05, "loss": 0.0283, "step": 26570 }, { "grad_norm": 0.38780471682548523, "learning_rate": 7.97575666609066e-05, "loss": 0.0328, "step": 26580 }, { "grad_norm": 0.4101959466934204, "learning_rate": 7.974095473266216e-05, "loss": 0.031, "step": 26590 }, { "grad_norm": 0.413987934589386, "learning_rate": 7.972433772250198e-05, "loss": 0.0365, "step": 26600 }, { "grad_norm": 0.47963133454322815, "learning_rate": 7.970771563326544e-05, "loss": 0.0351, "step": 26610 }, { "grad_norm": 0.5414088368415833, "learning_rate": 7.96910884677928e-05, "loss": 0.0373, "step": 26620 }, { "grad_norm": 0.35535722970962524, "learning_rate": 7.967445622892523e-05, "loss": 0.036, "step": 26630 }, { "grad_norm": 0.25123241543769836, "learning_rate": 7.965781891950465e-05, "loss": 0.0317, "step": 26640 }, { "grad_norm": 0.31111225485801697, "learning_rate": 7.964117654237397e-05, "loss": 0.0271, "step": 26650 }, { "grad_norm": 0.3937869071960449, "learning_rate": 7.962452910037692e-05, "loss": 0.0355, "step": 26660 }, { "grad_norm": 0.4136907160282135, "learning_rate": 7.96078765963581e-05, "loss": 0.0282, "step": 26670 }, { "grad_norm": 0.4465350806713104, "learning_rate": 7.95912190331629e-05, "loss": 0.0375, "step": 26680 }, { "grad_norm": 0.42456644773483276, "learning_rate": 7.957455641363772e-05, "loss": 0.0325, "step": 26690 }, { "grad_norm": 0.4490965008735657, "learning_rate": 7.955788874062968e-05, "loss": 0.0442, "step": 26700 }, { "grad_norm": 0.3306862711906433, "learning_rate": 7.95412160169869e-05, "loss": 0.0366, "step": 26710 }, { "grad_norm": 0.3004748821258545, "learning_rate": 7.952453824555824e-05, "loss": 0.0299, "step": 26720 }, { "grad_norm": 0.30160537362098694, "learning_rate": 7.95078554291935e-05, "loss": 0.0224, "step": 26730 }, { "grad_norm": 0.34963497519493103, "learning_rate": 7.94911675707433e-05, "loss": 0.0248, "step": 26740 }, { "grad_norm": 0.3802882432937622, "learning_rate": 7.947447467305915e-05, "loss": 0.0359, "step": 26750 }, { "grad_norm": 0.25313258171081543, "learning_rate": 7.94577767389934e-05, "loss": 0.0286, "step": 26760 }, { "grad_norm": 0.528428852558136, "learning_rate": 7.944107377139928e-05, "loss": 0.0339, "step": 26770 }, { "grad_norm": 0.4140019416809082, "learning_rate": 7.942436577313088e-05, "loss": 0.0392, "step": 26780 }, { "grad_norm": 0.3176327645778656, "learning_rate": 7.940765274704312e-05, "loss": 0.0349, "step": 26790 }, { "grad_norm": 0.345107764005661, "learning_rate": 7.939093469599181e-05, "loss": 0.0324, "step": 26800 }, { "grad_norm": 0.34939566254615784, "learning_rate": 7.93742116228336e-05, "loss": 0.0316, "step": 26810 }, { "grad_norm": 0.32950180768966675, "learning_rate": 7.935748353042602e-05, "loss": 0.0366, "step": 26820 }, { "grad_norm": 0.3423955738544464, "learning_rate": 7.934075042162744e-05, "loss": 0.0311, "step": 26830 }, { "grad_norm": 0.5052993893623352, "learning_rate": 7.932401229929705e-05, "loss": 0.0356, "step": 26840 }, { "grad_norm": 0.3911014497280121, "learning_rate": 7.9307269166295e-05, "loss": 0.0336, "step": 26850 }, { "grad_norm": 0.2837841808795929, "learning_rate": 7.92905210254822e-05, "loss": 0.0355, "step": 26860 }, { "grad_norm": 0.34189626574516296, "learning_rate": 7.927376787972045e-05, "loss": 0.0349, "step": 26870 }, { "grad_norm": 0.3651842474937439, "learning_rate": 7.92570097318724e-05, "loss": 0.0372, "step": 26880 }, { "grad_norm": 0.2785234749317169, "learning_rate": 7.924024658480158e-05, "loss": 0.0281, "step": 26890 }, { "grad_norm": 0.34349745512008667, "learning_rate": 7.922347844137233e-05, "loss": 0.0377, "step": 26900 }, { "grad_norm": 0.18151119351387024, "learning_rate": 7.92067053044499e-05, "loss": 0.0314, "step": 26910 }, { "grad_norm": 0.46185460686683655, "learning_rate": 7.918992717690031e-05, "loss": 0.0369, "step": 26920 }, { "grad_norm": 0.39167845249176025, "learning_rate": 7.917314406159053e-05, "loss": 0.0345, "step": 26930 }, { "grad_norm": 0.5358321070671082, "learning_rate": 7.915635596138832e-05, "loss": 0.0449, "step": 26940 }, { "grad_norm": 0.282846063375473, "learning_rate": 7.913956287916228e-05, "loss": 0.0354, "step": 26950 }, { "grad_norm": 0.32985883951187134, "learning_rate": 7.912276481778193e-05, "loss": 0.0281, "step": 26960 }, { "grad_norm": 0.2637387812137604, "learning_rate": 7.910596178011759e-05, "loss": 0.0314, "step": 26970 }, { "grad_norm": 0.24391868710517883, "learning_rate": 7.908915376904043e-05, "loss": 0.0262, "step": 26980 }, { "grad_norm": 0.30222904682159424, "learning_rate": 7.907234078742247e-05, "loss": 0.0304, "step": 26990 }, { "grad_norm": 0.3796596825122833, "learning_rate": 7.90555228381366e-05, "loss": 0.0259, "step": 27000 }, { "grad_norm": 0.3071141839027405, "learning_rate": 7.903869992405656e-05, "loss": 0.0231, "step": 27010 }, { "grad_norm": 0.41063982248306274, "learning_rate": 7.902187204805691e-05, "loss": 0.0296, "step": 27020 }, { "grad_norm": 0.33366191387176514, "learning_rate": 7.900503921301308e-05, "loss": 0.0268, "step": 27030 }, { "grad_norm": 0.45626986026763916, "learning_rate": 7.898820142180133e-05, "loss": 0.0401, "step": 27040 }, { "grad_norm": 0.34734392166137695, "learning_rate": 7.897135867729879e-05, "loss": 0.0284, "step": 27050 }, { "grad_norm": 0.37273693084716797, "learning_rate": 7.89545109823834e-05, "loss": 0.0223, "step": 27060 }, { "grad_norm": 0.28084608912467957, "learning_rate": 7.8937658339934e-05, "loss": 0.0239, "step": 27070 }, { "grad_norm": 0.29784780740737915, "learning_rate": 7.892080075283026e-05, "loss": 0.0248, "step": 27080 }, { "grad_norm": 0.26875028014183044, "learning_rate": 7.890393822395263e-05, "loss": 0.0307, "step": 27090 }, { "grad_norm": 0.33814582228660583, "learning_rate": 7.88870707561825e-05, "loss": 0.0315, "step": 27100 }, { "grad_norm": 0.5015461444854736, "learning_rate": 7.887019835240203e-05, "loss": 0.0337, "step": 27110 }, { "grad_norm": 0.4003744125366211, "learning_rate": 7.885332101549427e-05, "loss": 0.0332, "step": 27120 }, { "grad_norm": 0.33766332268714905, "learning_rate": 7.883643874834308e-05, "loss": 0.0303, "step": 27130 }, { "grad_norm": 0.3837261497974396, "learning_rate": 7.881955155383321e-05, "loss": 0.0416, "step": 27140 }, { "grad_norm": 0.3968624770641327, "learning_rate": 7.880265943485017e-05, "loss": 0.0355, "step": 27150 }, { "grad_norm": 0.2887701094150543, "learning_rate": 7.878576239428038e-05, "loss": 0.032, "step": 27160 }, { "grad_norm": 0.408217191696167, "learning_rate": 7.87688604350111e-05, "loss": 0.0464, "step": 27170 }, { "grad_norm": 0.4845799207687378, "learning_rate": 7.875195355993042e-05, "loss": 0.0345, "step": 27180 }, { "grad_norm": 0.4538952708244324, "learning_rate": 7.873504177192724e-05, "loss": 0.0383, "step": 27190 }, { "grad_norm": 0.3090376555919647, "learning_rate": 7.87181250738913e-05, "loss": 0.0373, "step": 27200 }, { "grad_norm": 0.30423232913017273, "learning_rate": 7.870120346871324e-05, "loss": 0.0325, "step": 27210 }, { "grad_norm": 0.3915318548679352, "learning_rate": 7.86842769592845e-05, "loss": 0.0307, "step": 27220 }, { "grad_norm": 0.4252452850341797, "learning_rate": 7.866734554849732e-05, "loss": 0.028, "step": 27230 }, { "grad_norm": 0.350115567445755, "learning_rate": 7.865040923924486e-05, "loss": 0.0313, "step": 27240 }, { "grad_norm": 0.3830365538597107, "learning_rate": 7.863346803442104e-05, "loss": 0.0344, "step": 27250 }, { "grad_norm": 0.4073932468891144, "learning_rate": 7.861652193692067e-05, "loss": 0.0296, "step": 27260 }, { "grad_norm": 0.2815788984298706, "learning_rate": 7.859957094963937e-05, "loss": 0.0234, "step": 27270 }, { "grad_norm": 0.2696211040019989, "learning_rate": 7.858261507547357e-05, "loss": 0.0267, "step": 27280 }, { "grad_norm": 0.2804945409297943, "learning_rate": 7.856565431732061e-05, "loss": 0.0247, "step": 27290 }, { "grad_norm": 0.31824541091918945, "learning_rate": 7.854868867807859e-05, "loss": 0.0297, "step": 27300 }, { "grad_norm": 0.370016872882843, "learning_rate": 7.85317181606465e-05, "loss": 0.0369, "step": 27310 }, { "grad_norm": 0.2831013798713684, "learning_rate": 7.85147427679241e-05, "loss": 0.0287, "step": 27320 }, { "grad_norm": 0.2686633765697479, "learning_rate": 7.849776250281205e-05, "loss": 0.0231, "step": 27330 }, { "grad_norm": 0.3612128794193268, "learning_rate": 7.84807773682118e-05, "loss": 0.0323, "step": 27340 }, { "grad_norm": 0.236830934882164, "learning_rate": 7.846378736702565e-05, "loss": 0.0299, "step": 27350 }, { "grad_norm": 0.8111609816551208, "learning_rate": 7.844679250215671e-05, "loss": 0.0442, "step": 27360 }, { "grad_norm": 0.23321807384490967, "learning_rate": 7.842979277650898e-05, "loss": 0.0326, "step": 27370 }, { "grad_norm": 0.400936484336853, "learning_rate": 7.84127881929872e-05, "loss": 0.034, "step": 27380 }, { "grad_norm": 0.2928875982761383, "learning_rate": 7.839577875449704e-05, "loss": 0.0411, "step": 27390 }, { "grad_norm": 0.35135358572006226, "learning_rate": 7.837876446394489e-05, "loss": 0.031, "step": 27400 }, { "grad_norm": 0.29889509081840515, "learning_rate": 7.836174532423805e-05, "loss": 0.0428, "step": 27410 }, { "grad_norm": 0.3782186806201935, "learning_rate": 7.834472133828466e-05, "loss": 0.031, "step": 27420 }, { "grad_norm": 0.42157718539237976, "learning_rate": 7.832769250899359e-05, "loss": 0.0358, "step": 27430 }, { "grad_norm": 0.42476850748062134, "learning_rate": 7.831065883927464e-05, "loss": 0.0365, "step": 27440 }, { "grad_norm": 0.32505422830581665, "learning_rate": 7.829362033203841e-05, "loss": 0.0327, "step": 27450 }, { "grad_norm": 0.34775903820991516, "learning_rate": 7.827657699019628e-05, "loss": 0.0367, "step": 27460 }, { "grad_norm": 0.383926123380661, "learning_rate": 7.825952881666052e-05, "loss": 0.0452, "step": 27470 }, { "grad_norm": 0.4623265862464905, "learning_rate": 7.824247581434418e-05, "loss": 0.0344, "step": 27480 }, { "grad_norm": 0.3596716821193695, "learning_rate": 7.822541798616116e-05, "loss": 0.0332, "step": 27490 }, { "grad_norm": 0.5065045952796936, "learning_rate": 7.820835533502617e-05, "loss": 0.0411, "step": 27500 }, { "grad_norm": 0.398893803358078, "learning_rate": 7.819128786385475e-05, "loss": 0.0312, "step": 27510 }, { "grad_norm": 0.3205671012401581, "learning_rate": 7.817421557556329e-05, "loss": 0.0414, "step": 27520 }, { "grad_norm": 0.4089905023574829, "learning_rate": 7.815713847306893e-05, "loss": 0.0412, "step": 27530 }, { "grad_norm": 0.33511483669281006, "learning_rate": 7.81400565592897e-05, "loss": 0.0313, "step": 27540 }, { "grad_norm": 0.39097315073013306, "learning_rate": 7.812296983714444e-05, "loss": 0.0374, "step": 27550 }, { "grad_norm": 0.21002410352230072, "learning_rate": 7.810587830955281e-05, "loss": 0.0316, "step": 27560 }, { "grad_norm": 0.3648391366004944, "learning_rate": 7.808878197943528e-05, "loss": 0.036, "step": 27570 }, { "grad_norm": 0.4033432900905609, "learning_rate": 7.807168084971312e-05, "loss": 0.035, "step": 27580 }, { "grad_norm": 0.4263889193534851, "learning_rate": 7.805457492330849e-05, "loss": 0.0296, "step": 27590 }, { "grad_norm": 0.36330726742744446, "learning_rate": 7.803746420314428e-05, "loss": 0.0349, "step": 27600 }, { "grad_norm": 0.47790470719337463, "learning_rate": 7.802034869214428e-05, "loss": 0.0409, "step": 27610 }, { "grad_norm": 0.4168335497379303, "learning_rate": 7.800322839323303e-05, "loss": 0.0309, "step": 27620 }, { "grad_norm": 0.28107091784477234, "learning_rate": 7.798610330933593e-05, "loss": 0.0301, "step": 27630 }, { "grad_norm": 0.20494945347309113, "learning_rate": 7.796897344337922e-05, "loss": 0.0325, "step": 27640 }, { "grad_norm": 0.6426880955696106, "learning_rate": 7.795183879828989e-05, "loss": 0.035, "step": 27650 }, { "grad_norm": 0.36131352186203003, "learning_rate": 7.793469937699579e-05, "loss": 0.0435, "step": 27660 }, { "grad_norm": 0.3077658414840698, "learning_rate": 7.791755518242558e-05, "loss": 0.0384, "step": 27670 }, { "grad_norm": 0.5655857920646667, "learning_rate": 7.790040621750876e-05, "loss": 0.0363, "step": 27680 }, { "grad_norm": 0.3404262065887451, "learning_rate": 7.788325248517558e-05, "loss": 0.0292, "step": 27690 }, { "grad_norm": 0.32050377130508423, "learning_rate": 7.786609398835715e-05, "loss": 0.0334, "step": 27700 }, { "grad_norm": 0.3469405770301819, "learning_rate": 7.784893072998541e-05, "loss": 0.0322, "step": 27710 }, { "grad_norm": 0.274352103471756, "learning_rate": 7.783176271299306e-05, "loss": 0.03, "step": 27720 }, { "grad_norm": 0.31236228346824646, "learning_rate": 7.781458994031368e-05, "loss": 0.0252, "step": 27730 }, { "grad_norm": 0.3660326600074768, "learning_rate": 7.779741241488161e-05, "loss": 0.0391, "step": 27740 }, { "grad_norm": 0.4746340811252594, "learning_rate": 7.7780230139632e-05, "loss": 0.0331, "step": 27750 }, { "grad_norm": 0.2963196933269501, "learning_rate": 7.776304311750087e-05, "loss": 0.0257, "step": 27760 }, { "grad_norm": 0.4703509211540222, "learning_rate": 7.7745851351425e-05, "loss": 0.0408, "step": 27770 }, { "grad_norm": 0.2877594828605652, "learning_rate": 7.772865484434197e-05, "loss": 0.0283, "step": 27780 }, { "grad_norm": 0.4408940076828003, "learning_rate": 7.77114535991902e-05, "loss": 0.0412, "step": 27790 }, { "grad_norm": 0.3612724542617798, "learning_rate": 7.769424761890893e-05, "loss": 0.032, "step": 27800 }, { "grad_norm": 0.4071797728538513, "learning_rate": 7.767703690643817e-05, "loss": 0.0346, "step": 27810 }, { "grad_norm": 0.24843494594097137, "learning_rate": 7.76598214647188e-05, "loss": 0.0307, "step": 27820 }, { "grad_norm": 0.32234618067741394, "learning_rate": 7.764260129669241e-05, "loss": 0.0327, "step": 27830 }, { "grad_norm": 0.3187185227870941, "learning_rate": 7.76253764053015e-05, "loss": 0.0398, "step": 27840 }, { "grad_norm": 0.300799161195755, "learning_rate": 7.760814679348932e-05, "loss": 0.0245, "step": 27850 }, { "grad_norm": 0.29909446835517883, "learning_rate": 7.759091246419992e-05, "loss": 0.0265, "step": 27860 }, { "grad_norm": 0.6962044835090637, "learning_rate": 7.75736734203782e-05, "loss": 0.0317, "step": 27870 }, { "grad_norm": 0.42577388882637024, "learning_rate": 7.755642966496985e-05, "loss": 0.0429, "step": 27880 }, { "grad_norm": 0.3705647885799408, "learning_rate": 7.753918120092132e-05, "loss": 0.0367, "step": 27890 }, { "grad_norm": 0.24035397171974182, "learning_rate": 7.752192803117993e-05, "loss": 0.0283, "step": 27900 }, { "grad_norm": 0.556316614151001, "learning_rate": 7.750467015869377e-05, "loss": 0.034, "step": 27910 }, { "grad_norm": 0.3312474191188812, "learning_rate": 7.748740758641174e-05, "loss": 0.0364, "step": 27920 }, { "grad_norm": 0.30906733870506287, "learning_rate": 7.74701403172835e-05, "loss": 0.0363, "step": 27930 }, { "grad_norm": 0.3241662383079529, "learning_rate": 7.745286835425962e-05, "loss": 0.0348, "step": 27940 }, { "grad_norm": 0.45754945278167725, "learning_rate": 7.743559170029138e-05, "loss": 0.0339, "step": 27950 }, { "grad_norm": 0.5681965947151184, "learning_rate": 7.741831035833087e-05, "loss": 0.0418, "step": 27960 }, { "grad_norm": 0.3976037800312042, "learning_rate": 7.740102433133102e-05, "loss": 0.0279, "step": 27970 }, { "grad_norm": 0.23475053906440735, "learning_rate": 7.738373362224553e-05, "loss": 0.0247, "step": 27980 }, { "grad_norm": 0.42635610699653625, "learning_rate": 7.73664382340289e-05, "loss": 0.0278, "step": 27990 }, { "grad_norm": 0.37018898129463196, "learning_rate": 7.734913816963647e-05, "loss": 0.0347, "step": 28000 }, { "grad_norm": 0.49705538153648376, "learning_rate": 7.73318334320243e-05, "loss": 0.0281, "step": 28010 }, { "grad_norm": 0.45830538868904114, "learning_rate": 7.731452402414934e-05, "loss": 0.0377, "step": 28020 }, { "grad_norm": 0.37830662727355957, "learning_rate": 7.729720994896928e-05, "loss": 0.0378, "step": 28030 }, { "grad_norm": 0.3233081102371216, "learning_rate": 7.727989120944262e-05, "loss": 0.0339, "step": 28040 }, { "grad_norm": 0.45953404903411865, "learning_rate": 7.726256780852865e-05, "loss": 0.0274, "step": 28050 }, { "grad_norm": 0.34955498576164246, "learning_rate": 7.724523974918749e-05, "loss": 0.0291, "step": 28060 }, { "grad_norm": 0.281274676322937, "learning_rate": 7.722790703438002e-05, "loss": 0.0228, "step": 28070 }, { "grad_norm": 0.4180060923099518, "learning_rate": 7.72105696670679e-05, "loss": 0.021, "step": 28080 }, { "grad_norm": 0.4284021258354187, "learning_rate": 7.719322765021364e-05, "loss": 0.0317, "step": 28090 }, { "grad_norm": 0.24314741790294647, "learning_rate": 7.717588098678051e-05, "loss": 0.0332, "step": 28100 }, { "grad_norm": 0.2902485132217407, "learning_rate": 7.715852967973258e-05, "loss": 0.0282, "step": 28110 }, { "grad_norm": 0.3517349660396576, "learning_rate": 7.714117373203474e-05, "loss": 0.0301, "step": 28120 }, { "grad_norm": 0.2970719635486603, "learning_rate": 7.712381314665259e-05, "loss": 0.0287, "step": 28130 }, { "grad_norm": 0.39464980363845825, "learning_rate": 7.710644792655261e-05, "loss": 0.0292, "step": 28140 }, { "grad_norm": 0.5661234259605408, "learning_rate": 7.708907807470207e-05, "loss": 0.0341, "step": 28150 }, { "grad_norm": 0.3305959403514862, "learning_rate": 7.707170359406896e-05, "loss": 0.0324, "step": 28160 }, { "grad_norm": 0.3911058008670807, "learning_rate": 7.705432448762213e-05, "loss": 0.03, "step": 28170 }, { "grad_norm": 0.4445459842681885, "learning_rate": 7.703694075833117e-05, "loss": 0.0407, "step": 28180 }, { "grad_norm": 0.5800712704658508, "learning_rate": 7.70195524091665e-05, "loss": 0.0546, "step": 28190 }, { "grad_norm": 0.4542689919471741, "learning_rate": 7.70021594430993e-05, "loss": 0.0404, "step": 28200 }, { "grad_norm": 0.3875873386859894, "learning_rate": 7.698476186310157e-05, "loss": 0.0333, "step": 28210 }, { "grad_norm": 0.3248653709888458, "learning_rate": 7.696735967214608e-05, "loss": 0.0283, "step": 28220 }, { "grad_norm": 0.34727758169174194, "learning_rate": 7.694995287320636e-05, "loss": 0.0299, "step": 28230 }, { "grad_norm": 0.30009546875953674, "learning_rate": 7.693254146925679e-05, "loss": 0.0308, "step": 28240 }, { "grad_norm": 0.45736995339393616, "learning_rate": 7.691512546327251e-05, "loss": 0.0386, "step": 28250 }, { "grad_norm": 0.42469218373298645, "learning_rate": 7.689770485822939e-05, "loss": 0.0336, "step": 28260 }, { "grad_norm": 0.3064996600151062, "learning_rate": 7.688027965710416e-05, "loss": 0.024, "step": 28270 }, { "grad_norm": 0.43813279271125793, "learning_rate": 7.686284986287433e-05, "loss": 0.0319, "step": 28280 }, { "grad_norm": 0.4106810688972473, "learning_rate": 7.684541547851817e-05, "loss": 0.027, "step": 28290 }, { "grad_norm": 0.3937070071697235, "learning_rate": 7.68279765070147e-05, "loss": 0.0319, "step": 28300 }, { "grad_norm": 0.2954331934452057, "learning_rate": 7.68105329513438e-05, "loss": 0.0275, "step": 28310 }, { "grad_norm": 0.29626381397247314, "learning_rate": 7.67930848144861e-05, "loss": 0.0272, "step": 28320 }, { "grad_norm": 0.46308907866477966, "learning_rate": 7.6775632099423e-05, "loss": 0.0361, "step": 28330 }, { "grad_norm": 0.2924267053604126, "learning_rate": 7.675817480913667e-05, "loss": 0.0314, "step": 28340 }, { "grad_norm": 0.31216397881507874, "learning_rate": 7.674071294661011e-05, "loss": 0.0304, "step": 28350 }, { "grad_norm": 0.3716059923171997, "learning_rate": 7.672324651482707e-05, "loss": 0.0261, "step": 28360 }, { "grad_norm": 0.4964858889579773, "learning_rate": 7.670577551677209e-05, "loss": 0.0326, "step": 28370 }, { "grad_norm": 0.3412816822528839, "learning_rate": 7.668829995543047e-05, "loss": 0.0316, "step": 28380 }, { "grad_norm": 0.30661073327064514, "learning_rate": 7.667081983378832e-05, "loss": 0.0327, "step": 28390 }, { "grad_norm": 0.42020532488822937, "learning_rate": 7.66533351548325e-05, "loss": 0.03, "step": 28400 }, { "grad_norm": 0.4615213871002197, "learning_rate": 7.663584592155069e-05, "loss": 0.0323, "step": 28410 }, { "grad_norm": 0.3854929804801941, "learning_rate": 7.661835213693129e-05, "loss": 0.0354, "step": 28420 }, { "grad_norm": 0.5433143377304077, "learning_rate": 7.660085380396353e-05, "loss": 0.0362, "step": 28430 }, { "grad_norm": 0.569021463394165, "learning_rate": 7.658335092563738e-05, "loss": 0.0368, "step": 28440 }, { "grad_norm": 0.3336722254753113, "learning_rate": 7.656584350494362e-05, "loss": 0.0306, "step": 28450 }, { "grad_norm": 0.3621520698070526, "learning_rate": 7.654833154487378e-05, "loss": 0.0329, "step": 28460 }, { "grad_norm": 0.36077725887298584, "learning_rate": 7.653081504842017e-05, "loss": 0.0289, "step": 28470 }, { "grad_norm": 0.3792189955711365, "learning_rate": 7.65132940185759e-05, "loss": 0.0344, "step": 28480 }, { "grad_norm": 0.38962626457214355, "learning_rate": 7.649576845833481e-05, "loss": 0.0254, "step": 28490 }, { "grad_norm": 0.2729405462741852, "learning_rate": 7.647823837069156e-05, "loss": 0.0354, "step": 28500 }, { "grad_norm": 0.2693425416946411, "learning_rate": 7.646070375864156e-05, "loss": 0.0292, "step": 28510 }, { "grad_norm": 0.4434497058391571, "learning_rate": 7.644316462518099e-05, "loss": 0.0338, "step": 28520 }, { "grad_norm": 0.43576353788375854, "learning_rate": 7.642562097330679e-05, "loss": 0.037, "step": 28530 }, { "grad_norm": 0.24358676373958588, "learning_rate": 7.640807280601671e-05, "loss": 0.0309, "step": 28540 }, { "grad_norm": 0.41343146562576294, "learning_rate": 7.639052012630927e-05, "loss": 0.038, "step": 28550 }, { "grad_norm": 0.35388287901878357, "learning_rate": 7.63729629371837e-05, "loss": 0.0414, "step": 28560 }, { "grad_norm": 0.39584794640541077, "learning_rate": 7.635540124164009e-05, "loss": 0.044, "step": 28570 }, { "grad_norm": 0.25304439663887024, "learning_rate": 7.633783504267922e-05, "loss": 0.0337, "step": 28580 }, { "grad_norm": 0.28003671765327454, "learning_rate": 7.632026434330269e-05, "loss": 0.0302, "step": 28590 }, { "grad_norm": 0.4901497960090637, "learning_rate": 7.630268914651282e-05, "loss": 0.0306, "step": 28600 }, { "grad_norm": 0.42248257994651794, "learning_rate": 7.628510945531278e-05, "loss": 0.0392, "step": 28610 }, { "grad_norm": 0.2518310844898224, "learning_rate": 7.626752527270641e-05, "loss": 0.0376, "step": 28620 }, { "grad_norm": 0.27321159839630127, "learning_rate": 7.62499366016984e-05, "loss": 0.0359, "step": 28630 }, { "grad_norm": 0.3003395199775696, "learning_rate": 7.623234344529416e-05, "loss": 0.0332, "step": 28640 }, { "grad_norm": 0.3930182158946991, "learning_rate": 7.62147458064999e-05, "loss": 0.0335, "step": 28650 }, { "grad_norm": 0.44866421818733215, "learning_rate": 7.619714368832254e-05, "loss": 0.0362, "step": 28660 }, { "grad_norm": 0.441673219203949, "learning_rate": 7.61795370937698e-05, "loss": 0.0337, "step": 28670 }, { "grad_norm": 0.39331087470054626, "learning_rate": 7.61619260258502e-05, "loss": 0.0342, "step": 28680 }, { "grad_norm": 0.35701972246170044, "learning_rate": 7.614431048757298e-05, "loss": 0.0408, "step": 28690 }, { "grad_norm": 0.2909775674343109, "learning_rate": 7.612669048194814e-05, "loss": 0.0343, "step": 28700 }, { "grad_norm": 0.3910573422908783, "learning_rate": 7.610906601198646e-05, "loss": 0.0262, "step": 28710 }, { "grad_norm": 0.36527177691459656, "learning_rate": 7.60914370806995e-05, "loss": 0.0379, "step": 28720 }, { "grad_norm": 0.5579994320869446, "learning_rate": 7.607380369109953e-05, "loss": 0.052, "step": 28730 }, { "grad_norm": 0.25708845257759094, "learning_rate": 7.605616584619961e-05, "loss": 0.027, "step": 28740 }, { "grad_norm": 0.31618019938468933, "learning_rate": 7.603852354901362e-05, "loss": 0.0316, "step": 28750 }, { "grad_norm": 0.3058621883392334, "learning_rate": 7.602087680255609e-05, "loss": 0.0455, "step": 28760 }, { "grad_norm": 0.4594099819660187, "learning_rate": 7.600322560984238e-05, "loss": 0.0346, "step": 28770 }, { "grad_norm": 0.42621558904647827, "learning_rate": 7.598556997388863e-05, "loss": 0.0325, "step": 28780 }, { "grad_norm": 0.31767719984054565, "learning_rate": 7.596790989771166e-05, "loss": 0.0306, "step": 28790 }, { "grad_norm": 0.5239026546478271, "learning_rate": 7.595024538432914e-05, "loss": 0.0308, "step": 28800 }, { "grad_norm": 0.5035474896430969, "learning_rate": 7.59325764367594e-05, "loss": 0.0316, "step": 28810 }, { "grad_norm": 0.3008468747138977, "learning_rate": 7.59149030580216e-05, "loss": 0.0369, "step": 28820 }, { "grad_norm": 0.2665499150753021, "learning_rate": 7.589722525113562e-05, "loss": 0.0301, "step": 28830 }, { "grad_norm": 0.47493913769721985, "learning_rate": 7.587954301912216e-05, "loss": 0.0307, "step": 28840 }, { "grad_norm": 0.47212907671928406, "learning_rate": 7.586185636500263e-05, "loss": 0.0346, "step": 28850 }, { "grad_norm": 0.3620334267616272, "learning_rate": 7.584416529179914e-05, "loss": 0.0328, "step": 28860 }, { "grad_norm": 0.41814088821411133, "learning_rate": 7.582646980253465e-05, "loss": 0.0314, "step": 28870 }, { "grad_norm": 0.3443792760372162, "learning_rate": 7.580876990023282e-05, "loss": 0.0456, "step": 28880 }, { "grad_norm": 0.34373030066490173, "learning_rate": 7.579106558791809e-05, "loss": 0.0318, "step": 28890 }, { "grad_norm": 0.45078781247138977, "learning_rate": 7.577335686861565e-05, "loss": 0.0346, "step": 28900 }, { "grad_norm": 0.2561981678009033, "learning_rate": 7.575564374535141e-05, "loss": 0.0358, "step": 28910 }, { "grad_norm": 0.2580525577068329, "learning_rate": 7.573792622115207e-05, "loss": 0.0278, "step": 28920 }, { "grad_norm": 0.21587547659873962, "learning_rate": 7.572020429904507e-05, "loss": 0.026, "step": 28930 }, { "grad_norm": 0.3596758544445038, "learning_rate": 7.570247798205861e-05, "loss": 0.0353, "step": 28940 }, { "grad_norm": 0.20097795128822327, "learning_rate": 7.568474727322164e-05, "loss": 0.0345, "step": 28950 }, { "grad_norm": 0.22612929344177246, "learning_rate": 7.566701217556384e-05, "loss": 0.0274, "step": 28960 }, { "grad_norm": 0.3395518958568573, "learning_rate": 7.564927269211564e-05, "loss": 0.0343, "step": 28970 }, { "grad_norm": 0.311057448387146, "learning_rate": 7.563152882590824e-05, "loss": 0.0294, "step": 28980 }, { "grad_norm": 0.40183714032173157, "learning_rate": 7.56137805799736e-05, "loss": 0.0344, "step": 28990 }, { "grad_norm": 0.5317103266716003, "learning_rate": 7.559602795734439e-05, "loss": 0.033, "step": 29000 }, { "grad_norm": 0.22533108294010162, "learning_rate": 7.557827096105408e-05, "loss": 0.0287, "step": 29010 }, { "grad_norm": 0.3457191586494446, "learning_rate": 7.55605095941368e-05, "loss": 0.0315, "step": 29020 }, { "grad_norm": 0.32229557633399963, "learning_rate": 7.554274385962752e-05, "loss": 0.0369, "step": 29030 }, { "grad_norm": 0.404554158449173, "learning_rate": 7.552497376056191e-05, "loss": 0.0385, "step": 29040 }, { "grad_norm": 1.073930263519287, "learning_rate": 7.550719929997639e-05, "loss": 0.037, "step": 29050 }, { "grad_norm": 0.5527806282043457, "learning_rate": 7.548942048090813e-05, "loss": 0.031, "step": 29060 }, { "grad_norm": 0.3374090790748596, "learning_rate": 7.547163730639506e-05, "loss": 0.0337, "step": 29070 }, { "grad_norm": 0.33648979663848877, "learning_rate": 7.545384977947583e-05, "loss": 0.0324, "step": 29080 }, { "grad_norm": 0.29756632447242737, "learning_rate": 7.543605790318981e-05, "loss": 0.0393, "step": 29090 }, { "grad_norm": 0.7551282048225403, "learning_rate": 7.54182616805772e-05, "loss": 0.0433, "step": 29100 }, { "grad_norm": 0.26129135489463806, "learning_rate": 7.540046111467885e-05, "loss": 0.0262, "step": 29110 }, { "grad_norm": 0.39315229654312134, "learning_rate": 7.53826562085364e-05, "loss": 0.0307, "step": 29120 }, { "grad_norm": 0.40414246916770935, "learning_rate": 7.536484696519221e-05, "loss": 0.027, "step": 29130 }, { "grad_norm": 0.3046392500400543, "learning_rate": 7.534703338768942e-05, "loss": 0.0274, "step": 29140 }, { "grad_norm": 0.37553736567497253, "learning_rate": 7.532921547907185e-05, "loss": 0.0332, "step": 29150 }, { "grad_norm": 0.4585625231266022, "learning_rate": 7.531139324238412e-05, "loss": 0.0343, "step": 29160 }, { "grad_norm": 0.37299805879592896, "learning_rate": 7.529356668067157e-05, "loss": 0.0426, "step": 29170 }, { "grad_norm": 0.29838261008262634, "learning_rate": 7.527573579698023e-05, "loss": 0.0356, "step": 29180 }, { "grad_norm": 0.2967846393585205, "learning_rate": 7.525790059435693e-05, "loss": 0.0333, "step": 29190 }, { "grad_norm": 0.40133512020111084, "learning_rate": 7.524006107584926e-05, "loss": 0.0346, "step": 29200 }, { "grad_norm": 0.31209713220596313, "learning_rate": 7.522221724450544e-05, "loss": 0.0323, "step": 29210 }, { "grad_norm": 0.31243425607681274, "learning_rate": 7.520436910337451e-05, "loss": 0.0432, "step": 29220 }, { "grad_norm": 0.46618783473968506, "learning_rate": 7.518651665550627e-05, "loss": 0.0308, "step": 29230 }, { "grad_norm": 0.530129611492157, "learning_rate": 7.516865990395117e-05, "loss": 0.0388, "step": 29240 }, { "grad_norm": 0.37473589181900024, "learning_rate": 7.515079885176047e-05, "loss": 0.0301, "step": 29250 }, { "grad_norm": 0.29906126856803894, "learning_rate": 7.513293350198612e-05, "loss": 0.0346, "step": 29260 }, { "grad_norm": 0.3650013208389282, "learning_rate": 7.511506385768081e-05, "loss": 0.0329, "step": 29270 }, { "grad_norm": 0.2959655523300171, "learning_rate": 7.509718992189801e-05, "loss": 0.0337, "step": 29280 }, { "grad_norm": 0.4543204605579376, "learning_rate": 7.507931169769182e-05, "loss": 0.0347, "step": 29290 }, { "grad_norm": 0.36183255910873413, "learning_rate": 7.506142918811722e-05, "loss": 0.0364, "step": 29300 }, { "grad_norm": 0.2049059420824051, "learning_rate": 7.504354239622978e-05, "loss": 0.0408, "step": 29310 }, { "grad_norm": 0.3730924427509308, "learning_rate": 7.50256513250859e-05, "loss": 0.0337, "step": 29320 }, { "grad_norm": 0.35159212350845337, "learning_rate": 7.500775597774265e-05, "loss": 0.0362, "step": 29330 }, { "grad_norm": 0.4551222324371338, "learning_rate": 7.498985635725788e-05, "loss": 0.029, "step": 29340 }, { "grad_norm": 0.3396826684474945, "learning_rate": 7.497195246669012e-05, "loss": 0.032, "step": 29350 }, { "grad_norm": 0.28460264205932617, "learning_rate": 7.495404430909868e-05, "loss": 0.0313, "step": 29360 }, { "grad_norm": 0.3822285234928131, "learning_rate": 7.493613188754356e-05, "loss": 0.032, "step": 29370 }, { "grad_norm": 0.2779771685600281, "learning_rate": 7.49182152050855e-05, "loss": 0.0323, "step": 29380 }, { "grad_norm": 0.4418887197971344, "learning_rate": 7.490029426478598e-05, "loss": 0.0286, "step": 29390 }, { "grad_norm": 0.26261386275291443, "learning_rate": 7.488236906970719e-05, "loss": 0.0349, "step": 29400 }, { "grad_norm": 0.3620922565460205, "learning_rate": 7.486443962291207e-05, "loss": 0.032, "step": 29410 }, { "grad_norm": 0.28467127680778503, "learning_rate": 7.484650592746424e-05, "loss": 0.0309, "step": 29420 }, { "grad_norm": 0.784659206867218, "learning_rate": 7.482856798642811e-05, "loss": 0.0292, "step": 29430 }, { "grad_norm": 0.2645655572414398, "learning_rate": 7.481062580286878e-05, "loss": 0.0317, "step": 29440 }, { "grad_norm": 0.1935313493013382, "learning_rate": 7.479267937985208e-05, "loss": 0.0272, "step": 29450 }, { "grad_norm": 0.44782647490501404, "learning_rate": 7.477472872044456e-05, "loss": 0.0246, "step": 29460 }, { "grad_norm": 0.37906211614608765, "learning_rate": 7.475677382771347e-05, "loss": 0.0344, "step": 29470 }, { "grad_norm": 0.24915191531181335, "learning_rate": 7.473881470472683e-05, "loss": 0.0296, "step": 29480 }, { "grad_norm": 0.34896618127822876, "learning_rate": 7.472085135455337e-05, "loss": 0.0305, "step": 29490 }, { "grad_norm": 0.30083411931991577, "learning_rate": 7.470288378026256e-05, "loss": 0.035, "step": 29500 }, { "grad_norm": 0.3400328457355499, "learning_rate": 7.468491198492451e-05, "loss": 0.0278, "step": 29510 }, { "grad_norm": 0.27864184975624084, "learning_rate": 7.466693597161013e-05, "loss": 0.031, "step": 29520 }, { "grad_norm": 0.242918461561203, "learning_rate": 7.464895574339104e-05, "loss": 0.0305, "step": 29530 }, { "grad_norm": 0.22879186272621155, "learning_rate": 7.463097130333958e-05, "loss": 0.0371, "step": 29540 }, { "grad_norm": 0.4186605215072632, "learning_rate": 7.461298265452876e-05, "loss": 0.032, "step": 29550 }, { "grad_norm": 0.5339353084564209, "learning_rate": 7.459498980003239e-05, "loss": 0.0252, "step": 29560 }, { "grad_norm": 0.32391637563705444, "learning_rate": 7.457699274292493e-05, "loss": 0.0393, "step": 29570 }, { "grad_norm": 0.4689306914806366, "learning_rate": 7.455899148628159e-05, "loss": 0.0367, "step": 29580 }, { "grad_norm": 0.39516547322273254, "learning_rate": 7.45409860331783e-05, "loss": 0.0304, "step": 29590 }, { "grad_norm": 0.3028169274330139, "learning_rate": 7.452297638669169e-05, "loss": 0.0288, "step": 29600 }, { "grad_norm": 0.3500988781452179, "learning_rate": 7.450496254989911e-05, "loss": 0.0215, "step": 29610 }, { "grad_norm": 0.27375441789627075, "learning_rate": 7.448694452587866e-05, "loss": 0.0252, "step": 29620 }, { "grad_norm": 0.3548722267150879, "learning_rate": 7.44689223177091e-05, "loss": 0.0282, "step": 29630 }, { "grad_norm": 0.3272821605205536, "learning_rate": 7.445089592846994e-05, "loss": 0.0259, "step": 29640 }, { "grad_norm": 0.3120756447315216, "learning_rate": 7.443286536124141e-05, "loss": 0.037, "step": 29650 }, { "grad_norm": 0.2770218551158905, "learning_rate": 7.441483061910443e-05, "loss": 0.0397, "step": 29660 }, { "grad_norm": 0.42964330315589905, "learning_rate": 7.439679170514064e-05, "loss": 0.0343, "step": 29670 }, { "grad_norm": 0.3973914086818695, "learning_rate": 7.43787486224324e-05, "loss": 0.0289, "step": 29680 }, { "grad_norm": 0.38628464937210083, "learning_rate": 7.436070137406276e-05, "loss": 0.0298, "step": 29690 }, { "grad_norm": 0.3590812087059021, "learning_rate": 7.434264996311556e-05, "loss": 0.0354, "step": 29700 }, { "grad_norm": 0.5611732602119446, "learning_rate": 7.432459439267525e-05, "loss": 0.0408, "step": 29710 }, { "grad_norm": 0.3177172839641571, "learning_rate": 7.430653466582701e-05, "loss": 0.0295, "step": 29720 }, { "grad_norm": 0.294998437166214, "learning_rate": 7.42884707856568e-05, "loss": 0.0404, "step": 29730 }, { "grad_norm": 0.4064638018608093, "learning_rate": 7.427040275525122e-05, "loss": 0.03, "step": 29740 }, { "grad_norm": 0.26492130756378174, "learning_rate": 7.42523305776976e-05, "loss": 0.0277, "step": 29750 }, { "grad_norm": 0.39510682225227356, "learning_rate": 7.4234254256084e-05, "loss": 0.0243, "step": 29760 }, { "grad_norm": 0.20860402286052704, "learning_rate": 7.421617379349915e-05, "loss": 0.0309, "step": 29770 }, { "grad_norm": 0.2812904119491577, "learning_rate": 7.41980891930325e-05, "loss": 0.0274, "step": 29780 }, { "grad_norm": 0.2570006251335144, "learning_rate": 7.418000045777425e-05, "loss": 0.029, "step": 29790 }, { "grad_norm": 0.24917079508304596, "learning_rate": 7.416190759081523e-05, "loss": 0.0279, "step": 29800 }, { "grad_norm": 0.30195069313049316, "learning_rate": 7.414381059524704e-05, "loss": 0.0253, "step": 29810 }, { "grad_norm": 0.2715396285057068, "learning_rate": 7.412570947416195e-05, "loss": 0.0247, "step": 29820 }, { "grad_norm": 0.5662555694580078, "learning_rate": 7.410760423065295e-05, "loss": 0.047, "step": 29830 }, { "grad_norm": 0.4169611632823944, "learning_rate": 7.408949486781372e-05, "loss": 0.0389, "step": 29840 }, { "grad_norm": 0.4232618808746338, "learning_rate": 7.407138138873868e-05, "loss": 0.0344, "step": 29850 }, { "grad_norm": 0.4375253915786743, "learning_rate": 7.405326379652292e-05, "loss": 0.0352, "step": 29860 }, { "grad_norm": 0.28697091341018677, "learning_rate": 7.403514209426222e-05, "loss": 0.0254, "step": 29870 }, { "grad_norm": 0.3821946084499359, "learning_rate": 7.40170162850531e-05, "loss": 0.0351, "step": 29880 }, { "grad_norm": 0.535137951374054, "learning_rate": 7.399888637199278e-05, "loss": 0.0297, "step": 29890 }, { "grad_norm": 0.4185650050640106, "learning_rate": 7.398075235817914e-05, "loss": 0.0432, "step": 29900 }, { "grad_norm": 0.49333885312080383, "learning_rate": 7.39626142467108e-05, "loss": 0.0365, "step": 29910 }, { "grad_norm": 0.29803842306137085, "learning_rate": 7.394447204068706e-05, "loss": 0.0251, "step": 29920 }, { "grad_norm": 0.37979939579963684, "learning_rate": 7.392632574320793e-05, "loss": 0.0384, "step": 29930 }, { "grad_norm": 0.30064332485198975, "learning_rate": 7.390817535737411e-05, "loss": 0.0315, "step": 29940 }, { "grad_norm": 0.38093748688697815, "learning_rate": 7.389002088628703e-05, "loss": 0.0351, "step": 29950 }, { "grad_norm": 0.3601524531841278, "learning_rate": 7.387186233304877e-05, "loss": 0.0363, "step": 29960 }, { "grad_norm": 0.5801966786384583, "learning_rate": 7.385369970076212e-05, "loss": 0.047, "step": 29970 }, { "grad_norm": 0.4714210331439972, "learning_rate": 7.38355329925306e-05, "loss": 0.0392, "step": 29980 }, { "grad_norm": 0.26935911178588867, "learning_rate": 7.381736221145838e-05, "loss": 0.0339, "step": 29990 }, { "grad_norm": 0.41926056146621704, "learning_rate": 7.37991873606504e-05, "loss": 0.0312, "step": 30000 }, { "grad_norm": 0.3124074339866638, "learning_rate": 7.378100844321218e-05, "loss": 0.0304, "step": 30010 }, { "grad_norm": 0.27725332975387573, "learning_rate": 7.376282546225004e-05, "loss": 0.0267, "step": 30020 }, { "grad_norm": 0.2843749523162842, "learning_rate": 7.374463842087094e-05, "loss": 0.0239, "step": 30030 }, { "grad_norm": 0.3541340231895447, "learning_rate": 7.372644732218254e-05, "loss": 0.0276, "step": 30040 }, { "grad_norm": 0.5419456958770752, "learning_rate": 7.370825216929322e-05, "loss": 0.0424, "step": 30050 }, { "grad_norm": 0.5043339729309082, "learning_rate": 7.369005296531205e-05, "loss": 0.0346, "step": 30060 }, { "grad_norm": 0.3418753147125244, "learning_rate": 7.367184971334873e-05, "loss": 0.0438, "step": 30070 }, { "grad_norm": 0.2686588168144226, "learning_rate": 7.365364241651371e-05, "loss": 0.0323, "step": 30080 }, { "grad_norm": 0.33758851885795593, "learning_rate": 7.363543107791815e-05, "loss": 0.0283, "step": 30090 }, { "grad_norm": 0.3059486746788025, "learning_rate": 7.361721570067384e-05, "loss": 0.0262, "step": 30100 }, { "grad_norm": 0.3980819284915924, "learning_rate": 7.359899628789331e-05, "loss": 0.0311, "step": 30110 }, { "grad_norm": 0.4408142566680908, "learning_rate": 7.358077284268974e-05, "loss": 0.0384, "step": 30120 }, { "grad_norm": 0.5131651163101196, "learning_rate": 7.356254536817702e-05, "loss": 0.0316, "step": 30130 }, { "grad_norm": 0.44993847608566284, "learning_rate": 7.354431386746973e-05, "loss": 0.036, "step": 30140 }, { "grad_norm": 0.4593646228313446, "learning_rate": 7.352607834368316e-05, "loss": 0.0311, "step": 30150 }, { "grad_norm": 0.2886638045310974, "learning_rate": 7.350783879993324e-05, "loss": 0.0253, "step": 30160 }, { "grad_norm": 0.353447824716568, "learning_rate": 7.348959523933658e-05, "loss": 0.0332, "step": 30170 }, { "grad_norm": 0.2234516143798828, "learning_rate": 7.347134766501057e-05, "loss": 0.0236, "step": 30180 }, { "grad_norm": 0.33269354701042175, "learning_rate": 7.345309608007315e-05, "loss": 0.0263, "step": 30190 }, { "grad_norm": 0.7914199829101562, "learning_rate": 7.343484048764308e-05, "loss": 0.031, "step": 30200 }, { "grad_norm": 0.530529797077179, "learning_rate": 7.341658089083972e-05, "loss": 0.0345, "step": 30210 }, { "grad_norm": 0.29073113203048706, "learning_rate": 7.339831729278313e-05, "loss": 0.0304, "step": 30220 }, { "grad_norm": 0.3658835291862488, "learning_rate": 7.338004969659404e-05, "loss": 0.0315, "step": 30230 }, { "grad_norm": 0.4055863320827484, "learning_rate": 7.336177810539391e-05, "loss": 0.0393, "step": 30240 }, { "grad_norm": 0.2720356285572052, "learning_rate": 7.334350252230485e-05, "loss": 0.0226, "step": 30250 }, { "grad_norm": 0.2969365119934082, "learning_rate": 7.332522295044965e-05, "loss": 0.0311, "step": 30260 }, { "grad_norm": 0.4101611375808716, "learning_rate": 7.33069393929518e-05, "loss": 0.0326, "step": 30270 }, { "grad_norm": 0.3642785847187042, "learning_rate": 7.328865185293545e-05, "loss": 0.0418, "step": 30280 }, { "grad_norm": 0.3047906458377838, "learning_rate": 7.327036033352546e-05, "loss": 0.0295, "step": 30290 }, { "grad_norm": 0.4137886166572571, "learning_rate": 7.325206483784733e-05, "loss": 0.0275, "step": 30300 }, { "grad_norm": 0.46502235531806946, "learning_rate": 7.323376536902724e-05, "loss": 0.0434, "step": 30310 }, { "grad_norm": 0.3028445839881897, "learning_rate": 7.321546193019213e-05, "loss": 0.0321, "step": 30320 }, { "grad_norm": 0.43038779497146606, "learning_rate": 7.31971545244695e-05, "loss": 0.0335, "step": 30330 }, { "grad_norm": 0.29907211661338806, "learning_rate": 7.31788431549876e-05, "loss": 0.0265, "step": 30340 }, { "grad_norm": 0.2100236415863037, "learning_rate": 7.316052782487534e-05, "loss": 0.0239, "step": 30350 }, { "grad_norm": 0.32782548666000366, "learning_rate": 7.314220853726234e-05, "loss": 0.0326, "step": 30360 }, { "grad_norm": 0.3206528127193451, "learning_rate": 7.312388529527884e-05, "loss": 0.0345, "step": 30370 }, { "grad_norm": 0.29174214601516724, "learning_rate": 7.310555810205577e-05, "loss": 0.0273, "step": 30380 }, { "grad_norm": 0.39071375131607056, "learning_rate": 7.308722696072476e-05, "loss": 0.0362, "step": 30390 }, { "grad_norm": 0.42671284079551697, "learning_rate": 7.306889187441811e-05, "loss": 0.0329, "step": 30400 }, { "grad_norm": 0.28248029947280884, "learning_rate": 7.305055284626876e-05, "loss": 0.0294, "step": 30410 }, { "grad_norm": 0.3356655538082123, "learning_rate": 7.303220987941037e-05, "loss": 0.0285, "step": 30420 }, { "grad_norm": 0.24812084436416626, "learning_rate": 7.301386297697726e-05, "loss": 0.0328, "step": 30430 }, { "grad_norm": 0.4875231683254242, "learning_rate": 7.299551214210438e-05, "loss": 0.0301, "step": 30440 }, { "grad_norm": 0.3100786507129669, "learning_rate": 7.297715737792738e-05, "loss": 0.0302, "step": 30450 }, { "grad_norm": 0.34378454089164734, "learning_rate": 7.295879868758265e-05, "loss": 0.0358, "step": 30460 }, { "grad_norm": 0.33771812915802, "learning_rate": 7.294043607420713e-05, "loss": 0.0267, "step": 30470 }, { "grad_norm": 0.36164212226867676, "learning_rate": 7.292206954093852e-05, "loss": 0.0326, "step": 30480 }, { "grad_norm": 0.3019392788410187, "learning_rate": 7.290369909091515e-05, "loss": 0.027, "step": 30490 }, { "grad_norm": 0.307302325963974, "learning_rate": 7.2885324727276e-05, "loss": 0.0305, "step": 30500 }, { "grad_norm": 0.37027066946029663, "learning_rate": 7.286694645316076e-05, "loss": 0.0268, "step": 30510 }, { "grad_norm": 0.30053025484085083, "learning_rate": 7.284856427170982e-05, "loss": 0.0279, "step": 30520 }, { "grad_norm": 0.6514121890068054, "learning_rate": 7.283017818606414e-05, "loss": 0.0324, "step": 30530 }, { "grad_norm": 0.5415616631507874, "learning_rate": 7.28117881993654e-05, "loss": 0.0346, "step": 30540 }, { "grad_norm": 0.2840338945388794, "learning_rate": 7.279339431475598e-05, "loss": 0.0322, "step": 30550 }, { "grad_norm": 0.28773602843284607, "learning_rate": 7.277499653537887e-05, "loss": 0.0289, "step": 30560 }, { "grad_norm": 0.4040592312812805, "learning_rate": 7.275659486437776e-05, "loss": 0.0394, "step": 30570 }, { "grad_norm": 0.35393989086151123, "learning_rate": 7.273818930489695e-05, "loss": 0.0386, "step": 30580 }, { "grad_norm": 0.3775220811367035, "learning_rate": 7.271977986008151e-05, "loss": 0.028, "step": 30590 }, { "grad_norm": 0.34036263823509216, "learning_rate": 7.270136653307705e-05, "loss": 0.0326, "step": 30600 }, { "grad_norm": 0.4577457308769226, "learning_rate": 7.268294932702994e-05, "loss": 0.0297, "step": 30610 }, { "grad_norm": 0.2864637076854706, "learning_rate": 7.266452824508719e-05, "loss": 0.0365, "step": 30620 }, { "grad_norm": 0.33616673946380615, "learning_rate": 7.264610329039643e-05, "loss": 0.0265, "step": 30630 }, { "grad_norm": 0.3089682459831238, "learning_rate": 7.262767446610599e-05, "loss": 0.0269, "step": 30640 }, { "grad_norm": 0.49753931164741516, "learning_rate": 7.260924177536485e-05, "loss": 0.0349, "step": 30650 }, { "grad_norm": 0.2171599119901657, "learning_rate": 7.259080522132265e-05, "loss": 0.0305, "step": 30660 }, { "grad_norm": 0.2665381133556366, "learning_rate": 7.257236480712972e-05, "loss": 0.0285, "step": 30670 }, { "grad_norm": 0.3161475956439972, "learning_rate": 7.255392053593697e-05, "loss": 0.0331, "step": 30680 }, { "grad_norm": 0.29667550325393677, "learning_rate": 7.253547241089607e-05, "loss": 0.0275, "step": 30690 }, { "grad_norm": 0.3226753771305084, "learning_rate": 7.251702043515927e-05, "loss": 0.0355, "step": 30700 }, { "grad_norm": 0.46620073914527893, "learning_rate": 7.249856461187952e-05, "loss": 0.0296, "step": 30710 }, { "grad_norm": 0.39768901467323303, "learning_rate": 7.248010494421042e-05, "loss": 0.0305, "step": 30720 }, { "grad_norm": 0.2349640130996704, "learning_rate": 7.246164143530622e-05, "loss": 0.0408, "step": 30730 }, { "grad_norm": 0.23316960036754608, "learning_rate": 7.244317408832181e-05, "loss": 0.0302, "step": 30740 }, { "grad_norm": 0.3321290910243988, "learning_rate": 7.242470290641279e-05, "loss": 0.0351, "step": 30750 }, { "grad_norm": 0.38039225339889526, "learning_rate": 7.240622789273536e-05, "loss": 0.0293, "step": 30760 }, { "grad_norm": 0.23688046634197235, "learning_rate": 7.238774905044638e-05, "loss": 0.0227, "step": 30770 }, { "grad_norm": 0.4194701313972473, "learning_rate": 7.236926638270341e-05, "loss": 0.0289, "step": 30780 }, { "grad_norm": 0.3634781837463379, "learning_rate": 7.23507798926646e-05, "loss": 0.0307, "step": 30790 }, { "grad_norm": 0.31104153394699097, "learning_rate": 7.23322895834888e-05, "loss": 0.03, "step": 30800 }, { "grad_norm": 0.267777681350708, "learning_rate": 7.231379545833552e-05, "loss": 0.0258, "step": 30810 }, { "grad_norm": 0.4377979338169098, "learning_rate": 7.229529752036487e-05, "loss": 0.0284, "step": 30820 }, { "grad_norm": 0.29493871331214905, "learning_rate": 7.227679577273765e-05, "loss": 0.0284, "step": 30830 }, { "grad_norm": 0.27536118030548096, "learning_rate": 7.225829021861529e-05, "loss": 0.0273, "step": 30840 }, { "grad_norm": 0.21306942403316498, "learning_rate": 7.223978086115992e-05, "loss": 0.0204, "step": 30850 }, { "grad_norm": 0.31717532873153687, "learning_rate": 7.222126770353425e-05, "loss": 0.0331, "step": 30860 }, { "grad_norm": 0.2899538576602936, "learning_rate": 7.22027507489017e-05, "loss": 0.0254, "step": 30870 }, { "grad_norm": 0.31067344546318054, "learning_rate": 7.218423000042627e-05, "loss": 0.0239, "step": 30880 }, { "grad_norm": 0.4661141037940979, "learning_rate": 7.216570546127268e-05, "loss": 0.0305, "step": 30890 }, { "grad_norm": 0.4093337655067444, "learning_rate": 7.214717713460626e-05, "loss": 0.037, "step": 30900 }, { "grad_norm": 0.37098193168640137, "learning_rate": 7.2128645023593e-05, "loss": 0.0336, "step": 30910 }, { "grad_norm": 0.42073261737823486, "learning_rate": 7.211010913139951e-05, "loss": 0.033, "step": 30920 }, { "grad_norm": 0.4170282483100891, "learning_rate": 7.209156946119308e-05, "loss": 0.0275, "step": 30930 }, { "grad_norm": 0.3236752450466156, "learning_rate": 7.207302601614166e-05, "loss": 0.0263, "step": 30940 }, { "grad_norm": 0.47560444474220276, "learning_rate": 7.205447879941378e-05, "loss": 0.0349, "step": 30950 }, { "grad_norm": 0.30535653233528137, "learning_rate": 7.203592781417866e-05, "loss": 0.0261, "step": 30960 }, { "grad_norm": 0.2875419855117798, "learning_rate": 7.201737306360617e-05, "loss": 0.0294, "step": 30970 }, { "grad_norm": 0.4077332019805908, "learning_rate": 7.19988145508668e-05, "loss": 0.032, "step": 30980 }, { "grad_norm": 0.38125213980674744, "learning_rate": 7.198025227913168e-05, "loss": 0.0342, "step": 30990 }, { "grad_norm": 0.3902094066143036, "learning_rate": 7.196168625157261e-05, "loss": 0.0321, "step": 31000 }, { "grad_norm": 0.41184890270233154, "learning_rate": 7.194311647136201e-05, "loss": 0.0329, "step": 31010 }, { "grad_norm": 0.45537617802619934, "learning_rate": 7.192454294167297e-05, "loss": 0.0251, "step": 31020 }, { "grad_norm": 0.8260423541069031, "learning_rate": 7.190596566567917e-05, "loss": 0.0312, "step": 31030 }, { "grad_norm": 0.417161762714386, "learning_rate": 7.188738464655496e-05, "loss": 0.0339, "step": 31040 }, { "grad_norm": 0.3444209694862366, "learning_rate": 7.186879988747533e-05, "loss": 0.0339, "step": 31050 }, { "grad_norm": 0.23185274004936218, "learning_rate": 7.185021139161592e-05, "loss": 0.0285, "step": 31060 }, { "grad_norm": 0.2684885561466217, "learning_rate": 7.1831619162153e-05, "loss": 0.0268, "step": 31070 }, { "grad_norm": 0.3653075695037842, "learning_rate": 7.181302320226345e-05, "loss": 0.0254, "step": 31080 }, { "grad_norm": 0.23349010944366455, "learning_rate": 7.179442351512482e-05, "loss": 0.0312, "step": 31090 }, { "grad_norm": 0.27748483419418335, "learning_rate": 7.177582010391528e-05, "loss": 0.0283, "step": 31100 }, { "grad_norm": 0.4196608364582062, "learning_rate": 7.175721297181366e-05, "loss": 0.0246, "step": 31110 }, { "grad_norm": 0.40169280767440796, "learning_rate": 7.173860212199942e-05, "loss": 0.0282, "step": 31120 }, { "grad_norm": 0.32217758893966675, "learning_rate": 7.171998755765263e-05, "loss": 0.0279, "step": 31130 }, { "grad_norm": 0.2734277546405792, "learning_rate": 7.170136928195398e-05, "loss": 0.0265, "step": 31140 }, { "grad_norm": 0.32032424211502075, "learning_rate": 7.168274729808489e-05, "loss": 0.028, "step": 31150 }, { "grad_norm": 0.35187554359436035, "learning_rate": 7.166412160922728e-05, "loss": 0.031, "step": 31160 }, { "grad_norm": 0.3221050798892975, "learning_rate": 7.164549221856382e-05, "loss": 0.0288, "step": 31170 }, { "grad_norm": 0.3109513819217682, "learning_rate": 7.162685912927775e-05, "loss": 0.0416, "step": 31180 }, { "grad_norm": 0.3307175636291504, "learning_rate": 7.160822234455294e-05, "loss": 0.0294, "step": 31190 }, { "grad_norm": 0.414400190114975, "learning_rate": 7.158958186757391e-05, "loss": 0.0294, "step": 31200 }, { "grad_norm": 0.37316182255744934, "learning_rate": 7.157093770152582e-05, "loss": 0.0317, "step": 31210 }, { "grad_norm": 0.2620463967323303, "learning_rate": 7.155228984959446e-05, "loss": 0.029, "step": 31220 }, { "grad_norm": 0.35151582956314087, "learning_rate": 7.153363831496621e-05, "loss": 0.0336, "step": 31230 }, { "grad_norm": 0.3343490660190582, "learning_rate": 7.151498310082811e-05, "loss": 0.0343, "step": 31240 }, { "grad_norm": 0.3393506407737732, "learning_rate": 7.149632421036784e-05, "loss": 0.0238, "step": 31250 }, { "grad_norm": 0.45684319734573364, "learning_rate": 7.147766164677369e-05, "loss": 0.0524, "step": 31260 }, { "grad_norm": 0.2235148549079895, "learning_rate": 7.145899541323459e-05, "loss": 0.0205, "step": 31270 }, { "grad_norm": 0.4099755883216858, "learning_rate": 7.144032551294007e-05, "loss": 0.037, "step": 31280 }, { "grad_norm": 0.305378794670105, "learning_rate": 7.14216519490803e-05, "loss": 0.0303, "step": 31290 }, { "grad_norm": 0.550653338432312, "learning_rate": 7.140297472484609e-05, "loss": 0.0402, "step": 31300 }, { "grad_norm": 0.21362584829330444, "learning_rate": 7.138429384342891e-05, "loss": 0.0443, "step": 31310 }, { "grad_norm": 0.4131215810775757, "learning_rate": 7.136560930802074e-05, "loss": 0.0261, "step": 31320 }, { "grad_norm": 0.3967585563659668, "learning_rate": 7.134692112181431e-05, "loss": 0.0286, "step": 31330 }, { "grad_norm": 0.4386838376522064, "learning_rate": 7.13282292880029e-05, "loss": 0.0278, "step": 31340 }, { "grad_norm": 0.35658136010169983, "learning_rate": 7.130953380978043e-05, "loss": 0.0312, "step": 31350 }, { "grad_norm": 0.2593252658843994, "learning_rate": 7.129083469034144e-05, "loss": 0.0388, "step": 31360 }, { "grad_norm": 0.4586103856563568, "learning_rate": 7.127213193288112e-05, "loss": 0.0363, "step": 31370 }, { "grad_norm": 0.3230883777141571, "learning_rate": 7.125342554059522e-05, "loss": 0.0248, "step": 31380 }, { "grad_norm": 0.39658981561660767, "learning_rate": 7.12347155166802e-05, "loss": 0.0269, "step": 31390 }, { "grad_norm": 0.32617664337158203, "learning_rate": 7.121600186433306e-05, "loss": 0.0248, "step": 31400 }, { "grad_norm": 0.3624962270259857, "learning_rate": 7.119728458675148e-05, "loss": 0.0237, "step": 31410 }, { "grad_norm": 0.5665030479431152, "learning_rate": 7.117856368713369e-05, "loss": 0.0284, "step": 31420 }, { "grad_norm": 0.45495161414146423, "learning_rate": 7.115983916867861e-05, "loss": 0.0313, "step": 31430 }, { "grad_norm": 0.29889243841171265, "learning_rate": 7.114111103458574e-05, "loss": 0.0249, "step": 31440 }, { "grad_norm": 0.2573884427547455, "learning_rate": 7.11223792880552e-05, "loss": 0.0208, "step": 31450 }, { "grad_norm": 0.3622000217437744, "learning_rate": 7.110364393228773e-05, "loss": 0.0286, "step": 31460 }, { "grad_norm": 0.4626953899860382, "learning_rate": 7.108490497048471e-05, "loss": 0.0248, "step": 31470 }, { "grad_norm": 0.38382670283317566, "learning_rate": 7.10661624058481e-05, "loss": 0.0282, "step": 31480 }, { "grad_norm": 0.30543264746665955, "learning_rate": 7.10474162415805e-05, "loss": 0.0296, "step": 31490 }, { "grad_norm": 0.2442931979894638, "learning_rate": 7.102866648088511e-05, "loss": 0.029, "step": 31500 }, { "grad_norm": 0.35845881700515747, "learning_rate": 7.100991312696576e-05, "loss": 0.03, "step": 31510 }, { "grad_norm": 0.2477555125951767, "learning_rate": 7.099115618302686e-05, "loss": 0.0254, "step": 31520 }, { "grad_norm": 0.3711530864238739, "learning_rate": 7.097239565227349e-05, "loss": 0.0326, "step": 31530 }, { "grad_norm": 1.2722924947738647, "learning_rate": 7.09536315379113e-05, "loss": 0.0264, "step": 31540 }, { "grad_norm": 0.4319120943546295, "learning_rate": 7.093486384314656e-05, "loss": 0.0246, "step": 31550 }, { "grad_norm": 0.28913047909736633, "learning_rate": 7.091609257118616e-05, "loss": 0.0254, "step": 31560 }, { "grad_norm": 0.4643736481666565, "learning_rate": 7.08973177252376e-05, "loss": 0.0308, "step": 31570 }, { "grad_norm": 0.3864825963973999, "learning_rate": 7.087853930850898e-05, "loss": 0.0253, "step": 31580 }, { "grad_norm": 0.39622220396995544, "learning_rate": 7.085975732420903e-05, "loss": 0.0253, "step": 31590 }, { "grad_norm": 1.0341168642044067, "learning_rate": 7.084097177554706e-05, "loss": 0.023, "step": 31600 }, { "grad_norm": 0.35141435265541077, "learning_rate": 7.082218266573301e-05, "loss": 0.032, "step": 31610 }, { "grad_norm": 0.3600148856639862, "learning_rate": 7.080338999797743e-05, "loss": 0.0312, "step": 31620 }, { "grad_norm": 0.5652239918708801, "learning_rate": 7.07845937754915e-05, "loss": 0.0324, "step": 31630 }, { "grad_norm": 0.2627582848072052, "learning_rate": 7.076579400148693e-05, "loss": 0.0369, "step": 31640 }, { "grad_norm": 0.2654043138027191, "learning_rate": 7.074699067917611e-05, "loss": 0.0304, "step": 31650 }, { "grad_norm": 0.40770402550697327, "learning_rate": 7.072818381177201e-05, "loss": 0.0322, "step": 31660 }, { "grad_norm": 0.39664939045906067, "learning_rate": 7.070937340248823e-05, "loss": 0.0295, "step": 31670 }, { "grad_norm": 0.29318201541900635, "learning_rate": 7.069055945453893e-05, "loss": 0.0297, "step": 31680 }, { "grad_norm": 0.3124685287475586, "learning_rate": 7.067174197113892e-05, "loss": 0.0238, "step": 31690 }, { "grad_norm": 0.28727903962135315, "learning_rate": 7.065292095550355e-05, "loss": 0.03, "step": 31700 }, { "grad_norm": 0.33045992255210876, "learning_rate": 7.063409641084887e-05, "loss": 0.0275, "step": 31710 }, { "grad_norm": 0.393798291683197, "learning_rate": 7.061526834039145e-05, "loss": 0.0291, "step": 31720 }, { "grad_norm": 0.3187315762042999, "learning_rate": 7.05964367473485e-05, "loss": 0.0245, "step": 31730 }, { "grad_norm": 0.4114491939544678, "learning_rate": 7.057760163493783e-05, "loss": 0.0334, "step": 31740 }, { "grad_norm": 0.3740575313568115, "learning_rate": 7.055876300637783e-05, "loss": 0.0294, "step": 31750 }, { "grad_norm": 0.32449933886528015, "learning_rate": 7.053992086488753e-05, "loss": 0.0413, "step": 31760 }, { "grad_norm": 0.29714465141296387, "learning_rate": 7.052107521368651e-05, "loss": 0.0315, "step": 31770 }, { "grad_norm": 0.34219905734062195, "learning_rate": 7.0502226055995e-05, "loss": 0.0338, "step": 31780 }, { "grad_norm": 0.305071085691452, "learning_rate": 7.048337339503379e-05, "loss": 0.0364, "step": 31790 }, { "grad_norm": 0.40216585993766785, "learning_rate": 7.046451723402427e-05, "loss": 0.0291, "step": 31800 }, { "grad_norm": 0.274350106716156, "learning_rate": 7.044565757618848e-05, "loss": 0.0433, "step": 31810 }, { "grad_norm": 0.2760309875011444, "learning_rate": 7.042679442474899e-05, "loss": 0.0281, "step": 31820 }, { "grad_norm": 0.2884061634540558, "learning_rate": 7.040792778292902e-05, "loss": 0.0374, "step": 31830 }, { "grad_norm": 0.4018021821975708, "learning_rate": 7.038905765395234e-05, "loss": 0.0389, "step": 31840 }, { "grad_norm": 0.38572239875793457, "learning_rate": 7.037018404104334e-05, "loss": 0.0298, "step": 31850 }, { "grad_norm": 0.2662917673587799, "learning_rate": 7.035130694742702e-05, "loss": 0.0227, "step": 31860 }, { "grad_norm": 0.5550113320350647, "learning_rate": 7.033242637632897e-05, "loss": 0.0325, "step": 31870 }, { "grad_norm": 0.432381808757782, "learning_rate": 7.031354233097534e-05, "loss": 0.0295, "step": 31880 }, { "grad_norm": 0.39733558893203735, "learning_rate": 7.029465481459289e-05, "loss": 0.0407, "step": 31890 }, { "grad_norm": 0.3330020308494568, "learning_rate": 7.027576383040898e-05, "loss": 0.0245, "step": 31900 }, { "grad_norm": 0.33576804399490356, "learning_rate": 7.025686938165159e-05, "loss": 0.0307, "step": 31910 }, { "grad_norm": 0.3707152307033539, "learning_rate": 7.023797147154924e-05, "loss": 0.0443, "step": 31920 }, { "grad_norm": 0.19379834830760956, "learning_rate": 7.021907010333111e-05, "loss": 0.022, "step": 31930 }, { "grad_norm": 0.21712374687194824, "learning_rate": 7.020016528022685e-05, "loss": 0.0225, "step": 31940 }, { "grad_norm": 0.33300745487213135, "learning_rate": 7.018125700546683e-05, "loss": 0.0346, "step": 31950 }, { "grad_norm": 0.3606962561607361, "learning_rate": 7.016234528228196e-05, "loss": 0.0295, "step": 31960 }, { "grad_norm": 0.2777491509914398, "learning_rate": 7.014343011390372e-05, "loss": 0.0343, "step": 31970 }, { "grad_norm": 0.28447064757347107, "learning_rate": 7.01245115035642e-05, "loss": 0.0227, "step": 31980 }, { "grad_norm": 0.30443263053894043, "learning_rate": 7.010558945449606e-05, "loss": 0.0379, "step": 31990 }, { "grad_norm": 0.25477316975593567, "learning_rate": 7.008666396993258e-05, "loss": 0.0221, "step": 32000 }, { "grad_norm": 0.31545254588127136, "learning_rate": 7.006773505310759e-05, "loss": 0.0368, "step": 32010 }, { "grad_norm": 0.5439341068267822, "learning_rate": 7.004880270725553e-05, "loss": 0.0295, "step": 32020 }, { "grad_norm": 0.3596075475215912, "learning_rate": 7.002986693561144e-05, "loss": 0.0375, "step": 32030 }, { "grad_norm": 0.38410377502441406, "learning_rate": 7.001092774141089e-05, "loss": 0.0375, "step": 32040 }, { "grad_norm": 0.46106812357902527, "learning_rate": 6.999198512789009e-05, "loss": 0.0364, "step": 32050 }, { "grad_norm": 0.5069971084594727, "learning_rate": 6.997303909828584e-05, "loss": 0.0338, "step": 32060 }, { "grad_norm": 0.3066674768924713, "learning_rate": 6.995408965583544e-05, "loss": 0.0321, "step": 32070 }, { "grad_norm": 0.44367003440856934, "learning_rate": 6.993513680377688e-05, "loss": 0.0199, "step": 32080 }, { "grad_norm": 0.39705681800842285, "learning_rate": 6.991618054534868e-05, "loss": 0.0397, "step": 32090 }, { "grad_norm": 0.34409067034721375, "learning_rate": 6.989722088378991e-05, "loss": 0.0316, "step": 32100 }, { "grad_norm": 0.5806422829627991, "learning_rate": 6.987825782234027e-05, "loss": 0.0343, "step": 32110 }, { "grad_norm": 0.28979867696762085, "learning_rate": 6.985929136424006e-05, "loss": 0.0272, "step": 32120 }, { "grad_norm": 0.5938044786453247, "learning_rate": 6.984032151273012e-05, "loss": 0.0348, "step": 32130 }, { "grad_norm": 0.3305199444293976, "learning_rate": 6.982134827105186e-05, "loss": 0.0336, "step": 32140 }, { "grad_norm": 0.3945416808128357, "learning_rate": 6.980237164244729e-05, "loss": 0.0319, "step": 32150 }, { "grad_norm": 0.40309590101242065, "learning_rate": 6.9783391630159e-05, "loss": 0.0245, "step": 32160 }, { "grad_norm": 0.2915726602077484, "learning_rate": 6.976440823743015e-05, "loss": 0.0257, "step": 32170 }, { "grad_norm": 0.31734052300453186, "learning_rate": 6.974542146750451e-05, "loss": 0.0243, "step": 32180 }, { "grad_norm": 0.3561248183250427, "learning_rate": 6.972643132362637e-05, "loss": 0.0244, "step": 32190 }, { "grad_norm": 0.3434482216835022, "learning_rate": 6.970743780904064e-05, "loss": 0.0277, "step": 32200 }, { "grad_norm": 0.3494040369987488, "learning_rate": 6.968844092699277e-05, "loss": 0.0262, "step": 32210 }, { "grad_norm": 0.34297117590904236, "learning_rate": 6.966944068072883e-05, "loss": 0.0275, "step": 32220 }, { "grad_norm": 0.3428126275539398, "learning_rate": 6.965043707349545e-05, "loss": 0.0264, "step": 32230 }, { "grad_norm": 0.3226320743560791, "learning_rate": 6.963143010853982e-05, "loss": 0.032, "step": 32240 }, { "grad_norm": 0.4196108281612396, "learning_rate": 6.961241978910971e-05, "loss": 0.0356, "step": 32250 }, { "grad_norm": 0.4186302721500397, "learning_rate": 6.959340611845344e-05, "loss": 0.0314, "step": 32260 }, { "grad_norm": 0.44278717041015625, "learning_rate": 6.957438909981995e-05, "loss": 0.0284, "step": 32270 }, { "grad_norm": 0.29907652735710144, "learning_rate": 6.955536873645872e-05, "loss": 0.0229, "step": 32280 }, { "grad_norm": 0.36283496022224426, "learning_rate": 6.953634503161982e-05, "loss": 0.0287, "step": 32290 }, { "grad_norm": 0.3419273793697357, "learning_rate": 6.951731798855387e-05, "loss": 0.0273, "step": 32300 }, { "grad_norm": 0.3561791777610779, "learning_rate": 6.949828761051208e-05, "loss": 0.0334, "step": 32310 }, { "grad_norm": 0.3131959140300751, "learning_rate": 6.947925390074622e-05, "loss": 0.0245, "step": 32320 }, { "grad_norm": 0.3349446952342987, "learning_rate": 6.946021686250863e-05, "loss": 0.0208, "step": 32330 }, { "grad_norm": 0.4143305718898773, "learning_rate": 6.94411764990522e-05, "loss": 0.0258, "step": 32340 }, { "grad_norm": 0.29680460691452026, "learning_rate": 6.942213281363044e-05, "loss": 0.0245, "step": 32350 }, { "grad_norm": 0.30557981133461, "learning_rate": 6.940308580949737e-05, "loss": 0.0318, "step": 32360 }, { "grad_norm": 0.2701411843299866, "learning_rate": 6.93840354899076e-05, "loss": 0.0243, "step": 32370 }, { "grad_norm": 1.022577166557312, "learning_rate": 6.936498185811633e-05, "loss": 0.0276, "step": 32380 }, { "grad_norm": 0.29990866780281067, "learning_rate": 6.93459249173793e-05, "loss": 0.0242, "step": 32390 }, { "grad_norm": 0.40157032012939453, "learning_rate": 6.932686467095279e-05, "loss": 0.0293, "step": 32400 }, { "grad_norm": 0.3507286608219147, "learning_rate": 6.930780112209373e-05, "loss": 0.0343, "step": 32410 }, { "grad_norm": 0.44633960723876953, "learning_rate": 6.92887342740595e-05, "loss": 0.0298, "step": 32420 }, { "grad_norm": 0.3104740083217621, "learning_rate": 6.926966413010816e-05, "loss": 0.0362, "step": 32430 }, { "grad_norm": 0.24789154529571533, "learning_rate": 6.925059069349824e-05, "loss": 0.0273, "step": 32440 }, { "grad_norm": 0.29332995414733887, "learning_rate": 6.923151396748886e-05, "loss": 0.0286, "step": 32450 }, { "grad_norm": 0.24903370440006256, "learning_rate": 6.921243395533974e-05, "loss": 0.0286, "step": 32460 }, { "grad_norm": 0.3329099416732788, "learning_rate": 6.919335066031109e-05, "loss": 0.0306, "step": 32470 }, { "grad_norm": 0.3010532259941101, "learning_rate": 6.917426408566379e-05, "loss": 0.0319, "step": 32480 }, { "grad_norm": 0.44342026114463806, "learning_rate": 6.915517423465916e-05, "loss": 0.0348, "step": 32490 }, { "grad_norm": 0.3434251546859741, "learning_rate": 6.913608111055914e-05, "loss": 0.029, "step": 32500 }, { "grad_norm": 0.4661216735839844, "learning_rate": 6.911698471662623e-05, "loss": 0.0265, "step": 32510 }, { "grad_norm": 0.30157142877578735, "learning_rate": 6.90978850561235e-05, "loss": 0.0258, "step": 32520 }, { "grad_norm": 0.46947455406188965, "learning_rate": 6.907878213231454e-05, "loss": 0.0266, "step": 32530 }, { "grad_norm": 0.42109018564224243, "learning_rate": 6.90596759484635e-05, "loss": 0.0268, "step": 32540 }, { "grad_norm": 0.4261512756347656, "learning_rate": 6.904056650783514e-05, "loss": 0.0285, "step": 32550 }, { "grad_norm": 0.3046377897262573, "learning_rate": 6.902145381369471e-05, "loss": 0.0327, "step": 32560 }, { "grad_norm": 0.3592851459980011, "learning_rate": 6.900233786930808e-05, "loss": 0.0298, "step": 32570 }, { "grad_norm": 0.33829909563064575, "learning_rate": 6.898321867794161e-05, "loss": 0.0304, "step": 32580 }, { "grad_norm": 0.419609010219574, "learning_rate": 6.896409624286226e-05, "loss": 0.0329, "step": 32590 }, { "grad_norm": 0.35429292917251587, "learning_rate": 6.894497056733754e-05, "loss": 0.0292, "step": 32600 }, { "grad_norm": 0.2946276366710663, "learning_rate": 6.89258416546355e-05, "loss": 0.0315, "step": 32610 }, { "grad_norm": 0.34490031003952026, "learning_rate": 6.890670950802474e-05, "loss": 0.032, "step": 32620 }, { "grad_norm": 0.4497402608394623, "learning_rate": 6.88875741307744e-05, "loss": 0.0294, "step": 32630 }, { "grad_norm": 0.20772039890289307, "learning_rate": 6.886843552615425e-05, "loss": 0.0324, "step": 32640 }, { "grad_norm": 0.25485414266586304, "learning_rate": 6.884929369743451e-05, "loss": 0.0205, "step": 32650 }, { "grad_norm": 0.3177052140235901, "learning_rate": 6.8830148647886e-05, "loss": 0.0257, "step": 32660 }, { "grad_norm": 0.3011915981769562, "learning_rate": 6.88110003807801e-05, "loss": 0.0364, "step": 32670 }, { "grad_norm": 0.3675462007522583, "learning_rate": 6.87918488993887e-05, "loss": 0.0248, "step": 32680 }, { "grad_norm": 0.44077277183532715, "learning_rate": 6.877269420698431e-05, "loss": 0.0337, "step": 32690 }, { "grad_norm": 0.3243289589881897, "learning_rate": 6.875353630683989e-05, "loss": 0.0299, "step": 32700 }, { "grad_norm": 0.4751603305339813, "learning_rate": 6.873437520222905e-05, "loss": 0.0333, "step": 32710 }, { "grad_norm": 0.22423166036605835, "learning_rate": 6.871521089642585e-05, "loss": 0.0295, "step": 32720 }, { "grad_norm": 0.23377445340156555, "learning_rate": 6.869604339270498e-05, "loss": 0.0291, "step": 32730 }, { "grad_norm": 0.3997906744480133, "learning_rate": 6.867687269434164e-05, "loss": 0.0368, "step": 32740 }, { "grad_norm": 0.311392605304718, "learning_rate": 6.865769880461156e-05, "loss": 0.0268, "step": 32750 }, { "grad_norm": 0.2851020097732544, "learning_rate": 6.863852172679104e-05, "loss": 0.0264, "step": 32760 }, { "grad_norm": 0.33868467807769775, "learning_rate": 6.861934146415693e-05, "loss": 0.0285, "step": 32770 }, { "grad_norm": 0.43493252992630005, "learning_rate": 6.86001580199866e-05, "loss": 0.0336, "step": 32780 }, { "grad_norm": 0.29182371497154236, "learning_rate": 6.858097139755798e-05, "loss": 0.0236, "step": 32790 }, { "grad_norm": 0.38974249362945557, "learning_rate": 6.856178160014955e-05, "loss": 0.025, "step": 32800 }, { "grad_norm": 0.3636086881160736, "learning_rate": 6.85425886310403e-05, "loss": 0.0242, "step": 32810 }, { "grad_norm": 0.3125673532485962, "learning_rate": 6.852339249350979e-05, "loss": 0.0223, "step": 32820 }, { "grad_norm": 0.2142173796892166, "learning_rate": 6.850419319083812e-05, "loss": 0.0326, "step": 32830 }, { "grad_norm": 0.3760188817977905, "learning_rate": 6.848499072630592e-05, "loss": 0.0239, "step": 32840 }, { "grad_norm": 0.7593902349472046, "learning_rate": 6.846578510319439e-05, "loss": 0.0319, "step": 32850 }, { "grad_norm": 0.41824567317962646, "learning_rate": 6.844657632478519e-05, "loss": 0.0335, "step": 32860 }, { "grad_norm": 0.3104363977909088, "learning_rate": 6.842736439436063e-05, "loss": 0.0244, "step": 32870 }, { "grad_norm": 0.27004170417785645, "learning_rate": 6.84081493152035e-05, "loss": 0.027, "step": 32880 }, { "grad_norm": 0.2723597288131714, "learning_rate": 6.83889310905971e-05, "loss": 0.0232, "step": 32890 }, { "grad_norm": 0.30758169293403625, "learning_rate": 6.836970972382533e-05, "loss": 0.0392, "step": 32900 }, { "grad_norm": 0.2598913311958313, "learning_rate": 6.835048521817257e-05, "loss": 0.0277, "step": 32910 }, { "grad_norm": 0.2612287998199463, "learning_rate": 6.833125757692379e-05, "loss": 0.0277, "step": 32920 }, { "grad_norm": 0.2645219564437866, "learning_rate": 6.831202680336441e-05, "loss": 0.0243, "step": 32930 }, { "grad_norm": 0.27467525005340576, "learning_rate": 6.829279290078052e-05, "loss": 0.0228, "step": 32940 }, { "grad_norm": 0.3561386168003082, "learning_rate": 6.827355587245863e-05, "loss": 0.0339, "step": 32950 }, { "grad_norm": 0.24526306986808777, "learning_rate": 6.82543157216858e-05, "loss": 0.0273, "step": 32960 }, { "grad_norm": 0.26211127638816833, "learning_rate": 6.823507245174969e-05, "loss": 0.0283, "step": 32970 }, { "grad_norm": 0.40641334652900696, "learning_rate": 6.821582606593841e-05, "loss": 0.0235, "step": 32980 }, { "grad_norm": 0.30796632170677185, "learning_rate": 6.81965765675407e-05, "loss": 0.0226, "step": 32990 }, { "grad_norm": 0.3666396737098694, "learning_rate": 6.81773239598457e-05, "loss": 0.0296, "step": 33000 }, { "grad_norm": 0.31675463914871216, "learning_rate": 6.815806824614319e-05, "loss": 0.0298, "step": 33010 }, { "grad_norm": 0.3855495750904083, "learning_rate": 6.813880942972343e-05, "loss": 0.0277, "step": 33020 }, { "grad_norm": 0.4051041901111603, "learning_rate": 6.811954751387726e-05, "loss": 0.0293, "step": 33030 }, { "grad_norm": 0.2961912751197815, "learning_rate": 6.810028250189598e-05, "loss": 0.0259, "step": 33040 }, { "grad_norm": 0.4397744834423065, "learning_rate": 6.808101439707147e-05, "loss": 0.0357, "step": 33050 }, { "grad_norm": 0.43690207600593567, "learning_rate": 6.806174320269609e-05, "loss": 0.0297, "step": 33060 }, { "grad_norm": 0.37547197937965393, "learning_rate": 6.804246892206281e-05, "loss": 0.0317, "step": 33070 }, { "grad_norm": 0.519282877445221, "learning_rate": 6.802319155846506e-05, "loss": 0.0439, "step": 33080 }, { "grad_norm": 0.42792826890945435, "learning_rate": 6.800391111519679e-05, "loss": 0.0292, "step": 33090 }, { "grad_norm": 0.33715495467185974, "learning_rate": 6.798462759555253e-05, "loss": 0.0329, "step": 33100 }, { "grad_norm": 0.2136482149362564, "learning_rate": 6.79653410028273e-05, "loss": 0.0258, "step": 33110 }, { "grad_norm": 0.2369619607925415, "learning_rate": 6.794605134031663e-05, "loss": 0.028, "step": 33120 }, { "grad_norm": 0.635006844997406, "learning_rate": 6.792675861131661e-05, "loss": 0.0322, "step": 33130 }, { "grad_norm": 0.37776508927345276, "learning_rate": 6.790746281912386e-05, "loss": 0.0341, "step": 33140 }, { "grad_norm": 0.3132440447807312, "learning_rate": 6.788816396703546e-05, "loss": 0.0273, "step": 33150 }, { "grad_norm": 0.42155876755714417, "learning_rate": 6.78688620583491e-05, "loss": 0.0277, "step": 33160 }, { "grad_norm": 0.3637807071208954, "learning_rate": 6.784955709636292e-05, "loss": 0.0357, "step": 33170 }, { "grad_norm": 0.4489888548851013, "learning_rate": 6.783024908437564e-05, "loss": 0.0267, "step": 33180 }, { "grad_norm": 0.5103796124458313, "learning_rate": 6.781093802568641e-05, "loss": 0.0324, "step": 33190 }, { "grad_norm": 0.3059512972831726, "learning_rate": 6.779162392359504e-05, "loss": 0.0236, "step": 33200 }, { "grad_norm": 0.4256565272808075, "learning_rate": 6.777230678140172e-05, "loss": 0.0274, "step": 33210 }, { "grad_norm": 0.2803061604499817, "learning_rate": 6.775298660240726e-05, "loss": 0.0296, "step": 33220 }, { "grad_norm": 0.2585664391517639, "learning_rate": 6.773366338991292e-05, "loss": 0.0368, "step": 33230 }, { "grad_norm": 0.2750272750854492, "learning_rate": 6.771433714722052e-05, "loss": 0.0257, "step": 33240 }, { "grad_norm": 0.4798917770385742, "learning_rate": 6.769500787763239e-05, "loss": 0.0325, "step": 33250 }, { "grad_norm": 0.256832480430603, "learning_rate": 6.76756755844514e-05, "loss": 0.033, "step": 33260 }, { "grad_norm": 0.3702237606048584, "learning_rate": 6.765634027098087e-05, "loss": 0.0275, "step": 33270 }, { "grad_norm": 0.35422900319099426, "learning_rate": 6.763700194052468e-05, "loss": 0.0293, "step": 33280 }, { "grad_norm": 0.270569384098053, "learning_rate": 6.761766059638723e-05, "loss": 0.0247, "step": 33290 }, { "grad_norm": 0.34800779819488525, "learning_rate": 6.759831624187345e-05, "loss": 0.0301, "step": 33300 }, { "grad_norm": 0.3934338688850403, "learning_rate": 6.757896888028871e-05, "loss": 0.0273, "step": 33310 }, { "grad_norm": 0.3101211190223694, "learning_rate": 6.7559618514939e-05, "loss": 0.0287, "step": 33320 }, { "grad_norm": 0.3177088797092438, "learning_rate": 6.754026514913073e-05, "loss": 0.0261, "step": 33330 }, { "grad_norm": 0.26713404059410095, "learning_rate": 6.752090878617087e-05, "loss": 0.0361, "step": 33340 }, { "grad_norm": 0.3929915130138397, "learning_rate": 6.75015494293669e-05, "loss": 0.0269, "step": 33350 }, { "grad_norm": 0.31874752044677734, "learning_rate": 6.74821870820268e-05, "loss": 0.0214, "step": 33360 }, { "grad_norm": 0.2684786021709442, "learning_rate": 6.746282174745907e-05, "loss": 0.034, "step": 33370 }, { "grad_norm": 0.3920521140098572, "learning_rate": 6.744345342897271e-05, "loss": 0.0342, "step": 33380 }, { "grad_norm": 0.258738249540329, "learning_rate": 6.742408212987724e-05, "loss": 0.0352, "step": 33390 }, { "grad_norm": 0.29937678575515747, "learning_rate": 6.740470785348269e-05, "loss": 0.0223, "step": 33400 }, { "grad_norm": 0.3911701738834381, "learning_rate": 6.738533060309958e-05, "loss": 0.0305, "step": 33410 }, { "grad_norm": 0.27359071373939514, "learning_rate": 6.736595038203894e-05, "loss": 0.0272, "step": 33420 }, { "grad_norm": 0.29278939962387085, "learning_rate": 6.734656719361236e-05, "loss": 0.0328, "step": 33430 }, { "grad_norm": 0.2708149254322052, "learning_rate": 6.732718104113189e-05, "loss": 0.0285, "step": 33440 }, { "grad_norm": 0.4411870837211609, "learning_rate": 6.730779192791006e-05, "loss": 0.0264, "step": 33450 }, { "grad_norm": 0.2675973176956177, "learning_rate": 6.728839985725997e-05, "loss": 0.0278, "step": 33460 }, { "grad_norm": 0.3104143738746643, "learning_rate": 6.726900483249517e-05, "loss": 0.0232, "step": 33470 }, { "grad_norm": 0.24452617764472961, "learning_rate": 6.724960685692976e-05, "loss": 0.0258, "step": 33480 }, { "grad_norm": 0.2610297203063965, "learning_rate": 6.723020593387833e-05, "loss": 0.0219, "step": 33490 }, { "grad_norm": 0.37103262543678284, "learning_rate": 6.721080206665593e-05, "loss": 0.0267, "step": 33500 }, { "grad_norm": 0.446016401052475, "learning_rate": 6.719139525857819e-05, "loss": 0.0361, "step": 33510 }, { "grad_norm": 0.22309580445289612, "learning_rate": 6.717198551296117e-05, "loss": 0.0229, "step": 33520 }, { "grad_norm": 0.26730677485466003, "learning_rate": 6.715257283312148e-05, "loss": 0.0296, "step": 33530 }, { "grad_norm": 0.2276345044374466, "learning_rate": 6.713315722237623e-05, "loss": 0.023, "step": 33540 }, { "grad_norm": 0.3133530020713806, "learning_rate": 6.7113738684043e-05, "loss": 0.0336, "step": 33550 }, { "grad_norm": 0.5313854813575745, "learning_rate": 6.709431722143989e-05, "loss": 0.0268, "step": 33560 }, { "grad_norm": 0.36962980031967163, "learning_rate": 6.70748928378855e-05, "loss": 0.0342, "step": 33570 }, { "grad_norm": 0.42355838418006897, "learning_rate": 6.705546553669891e-05, "loss": 0.0242, "step": 33580 }, { "grad_norm": 0.19132965803146362, "learning_rate": 6.703603532119974e-05, "loss": 0.0219, "step": 33590 }, { "grad_norm": 0.36457815766334534, "learning_rate": 6.701660219470808e-05, "loss": 0.0279, "step": 33600 }, { "grad_norm": 0.45001718401908875, "learning_rate": 6.69971661605445e-05, "loss": 0.0272, "step": 33610 }, { "grad_norm": 0.3584843575954437, "learning_rate": 6.697772722203008e-05, "loss": 0.0257, "step": 33620 }, { "grad_norm": 0.4131649434566498, "learning_rate": 6.695828538248643e-05, "loss": 0.0278, "step": 33630 }, { "grad_norm": 0.6714658141136169, "learning_rate": 6.693884064523563e-05, "loss": 0.0353, "step": 33640 }, { "grad_norm": 0.2417031079530716, "learning_rate": 6.691939301360023e-05, "loss": 0.0225, "step": 33650 }, { "grad_norm": 0.4553913474082947, "learning_rate": 6.689994249090333e-05, "loss": 0.0283, "step": 33660 }, { "grad_norm": 0.5731977820396423, "learning_rate": 6.688048908046845e-05, "loss": 0.0357, "step": 33670 }, { "grad_norm": 0.25212597846984863, "learning_rate": 6.686103278561969e-05, "loss": 0.0238, "step": 33680 }, { "grad_norm": 0.2882511019706726, "learning_rate": 6.684157360968156e-05, "loss": 0.0326, "step": 33690 }, { "grad_norm": 0.2979156970977783, "learning_rate": 6.682211155597911e-05, "loss": 0.0233, "step": 33700 }, { "grad_norm": 0.3165440559387207, "learning_rate": 6.680264662783789e-05, "loss": 0.0265, "step": 33710 }, { "grad_norm": 0.31335657835006714, "learning_rate": 6.678317882858391e-05, "loss": 0.0288, "step": 33720 }, { "grad_norm": 0.3761846423149109, "learning_rate": 6.67637081615437e-05, "loss": 0.0251, "step": 33730 }, { "grad_norm": 0.2971174716949463, "learning_rate": 6.674423463004427e-05, "loss": 0.0235, "step": 33740 }, { "grad_norm": 0.3607199192047119, "learning_rate": 6.672475823741308e-05, "loss": 0.0271, "step": 33750 }, { "grad_norm": 0.47057512402534485, "learning_rate": 6.670527898697811e-05, "loss": 0.0333, "step": 33760 }, { "grad_norm": 0.32392382621765137, "learning_rate": 6.668579688206788e-05, "loss": 0.0335, "step": 33770 }, { "grad_norm": 0.33626529574394226, "learning_rate": 6.666631192601131e-05, "loss": 0.0272, "step": 33780 }, { "grad_norm": 0.3675614893436432, "learning_rate": 6.664682412213785e-05, "loss": 0.0283, "step": 33790 }, { "grad_norm": 0.38683682680130005, "learning_rate": 6.662733347377745e-05, "loss": 0.0253, "step": 33800 }, { "grad_norm": 0.32946500182151794, "learning_rate": 6.660783998426051e-05, "loss": 0.028, "step": 33810 }, { "grad_norm": 0.43053457140922546, "learning_rate": 6.658834365691794e-05, "loss": 0.0273, "step": 33820 }, { "grad_norm": 0.26407796144485474, "learning_rate": 6.656884449508115e-05, "loss": 0.0236, "step": 33830 }, { "grad_norm": 0.3656637668609619, "learning_rate": 6.654934250208198e-05, "loss": 0.0221, "step": 33840 }, { "grad_norm": 0.26811355352401733, "learning_rate": 6.65298376812528e-05, "loss": 0.0267, "step": 33850 }, { "grad_norm": 0.23162828385829926, "learning_rate": 6.651033003592646e-05, "loss": 0.0241, "step": 33860 }, { "grad_norm": 0.29720500111579895, "learning_rate": 6.649081956943626e-05, "loss": 0.0222, "step": 33870 }, { "grad_norm": 0.4111599326133728, "learning_rate": 6.647130628511604e-05, "loss": 0.027, "step": 33880 }, { "grad_norm": 0.3143327534198761, "learning_rate": 6.645179018630005e-05, "loss": 0.0239, "step": 33890 }, { "grad_norm": 0.4188513159751892, "learning_rate": 6.643227127632309e-05, "loss": 0.028, "step": 33900 }, { "grad_norm": 0.397585391998291, "learning_rate": 6.641274955852038e-05, "loss": 0.0322, "step": 33910 }, { "grad_norm": 0.4547061026096344, "learning_rate": 6.639322503622768e-05, "loss": 0.0451, "step": 33920 }, { "grad_norm": 0.3846021592617035, "learning_rate": 6.637369771278116e-05, "loss": 0.0342, "step": 33930 }, { "grad_norm": 0.2676609754562378, "learning_rate": 6.635416759151751e-05, "loss": 0.0328, "step": 33940 }, { "grad_norm": 0.3369656801223755, "learning_rate": 6.633463467577394e-05, "loss": 0.0444, "step": 33950 }, { "grad_norm": 0.6633026003837585, "learning_rate": 6.631509896888803e-05, "loss": 0.0325, "step": 33960 }, { "grad_norm": 0.2715129256248474, "learning_rate": 6.629556047419794e-05, "loss": 0.0236, "step": 33970 }, { "grad_norm": 0.37760457396507263, "learning_rate": 6.627601919504223e-05, "loss": 0.0249, "step": 33980 }, { "grad_norm": 0.35272789001464844, "learning_rate": 6.625647513476001e-05, "loss": 0.0255, "step": 33990 }, { "grad_norm": 0.5561496019363403, "learning_rate": 6.62369282966908e-05, "loss": 0.0397, "step": 34000 }, { "grad_norm": 0.3385658264160156, "learning_rate": 6.621737868417464e-05, "loss": 0.0316, "step": 34010 }, { "grad_norm": 0.23777399957180023, "learning_rate": 6.619782630055198e-05, "loss": 0.0296, "step": 34020 }, { "grad_norm": 0.3149891495704651, "learning_rate": 6.617827114916382e-05, "loss": 0.028, "step": 34030 }, { "grad_norm": 0.22047531604766846, "learning_rate": 6.615871323335161e-05, "loss": 0.0237, "step": 34040 }, { "grad_norm": 0.18482868373394012, "learning_rate": 6.613915255645725e-05, "loss": 0.0238, "step": 34050 }, { "grad_norm": 0.3739556670188904, "learning_rate": 6.611958912182312e-05, "loss": 0.0222, "step": 34060 }, { "grad_norm": 0.351627916097641, "learning_rate": 6.610002293279207e-05, "loss": 0.0283, "step": 34070 }, { "grad_norm": 0.28332483768463135, "learning_rate": 6.608045399270746e-05, "loss": 0.025, "step": 34080 }, { "grad_norm": 0.24853837490081787, "learning_rate": 6.606088230491304e-05, "loss": 0.0257, "step": 34090 }, { "grad_norm": 0.3327062726020813, "learning_rate": 6.604130787275312e-05, "loss": 0.024, "step": 34100 }, { "grad_norm": 0.23683366179466248, "learning_rate": 6.602173069957242e-05, "loss": 0.0238, "step": 34110 }, { "grad_norm": 0.31162211298942566, "learning_rate": 6.600215078871612e-05, "loss": 0.0252, "step": 34120 }, { "grad_norm": 0.35637012124061584, "learning_rate": 6.598256814352992e-05, "loss": 0.0237, "step": 34130 }, { "grad_norm": 0.4008612632751465, "learning_rate": 6.596298276735995e-05, "loss": 0.0251, "step": 34140 }, { "grad_norm": 0.5435084104537964, "learning_rate": 6.594339466355282e-05, "loss": 0.0277, "step": 34150 }, { "grad_norm": 0.29366275668144226, "learning_rate": 6.592380383545558e-05, "loss": 0.0331, "step": 34160 }, { "grad_norm": 0.36081892251968384, "learning_rate": 6.590421028641577e-05, "loss": 0.0314, "step": 34170 }, { "grad_norm": 0.35794341564178467, "learning_rate": 6.588461401978143e-05, "loss": 0.0253, "step": 34180 }, { "grad_norm": 0.47800782322883606, "learning_rate": 6.586501503890099e-05, "loss": 0.026, "step": 34190 }, { "grad_norm": 0.31727319955825806, "learning_rate": 6.584541334712338e-05, "loss": 0.0254, "step": 34200 }, { "grad_norm": 0.3277142643928528, "learning_rate": 6.582580894779802e-05, "loss": 0.0266, "step": 34210 }, { "grad_norm": 0.3434234857559204, "learning_rate": 6.580620184427473e-05, "loss": 0.0222, "step": 34220 }, { "grad_norm": 0.2690163254737854, "learning_rate": 6.578659203990385e-05, "loss": 0.0219, "step": 34230 }, { "grad_norm": 0.31501296162605286, "learning_rate": 6.576697953803615e-05, "loss": 0.0255, "step": 34240 }, { "grad_norm": 0.21407602727413177, "learning_rate": 6.57473643420229e-05, "loss": 0.0278, "step": 34250 }, { "grad_norm": 0.3505345582962036, "learning_rate": 6.572774645521574e-05, "loss": 0.0309, "step": 34260 }, { "grad_norm": 0.5347409844398499, "learning_rate": 6.570812588096688e-05, "loss": 0.0338, "step": 34270 }, { "grad_norm": 0.31036311388015747, "learning_rate": 6.568850262262893e-05, "loss": 0.0232, "step": 34280 }, { "grad_norm": 0.3139667510986328, "learning_rate": 6.566887668355497e-05, "loss": 0.0259, "step": 34290 }, { "grad_norm": 0.541851818561554, "learning_rate": 6.564924806709851e-05, "loss": 0.0246, "step": 34300 }, { "grad_norm": 0.2849547863006592, "learning_rate": 6.562961677661359e-05, "loss": 0.0353, "step": 34310 }, { "grad_norm": 0.44496068358421326, "learning_rate": 6.56099828154546e-05, "loss": 0.0331, "step": 34320 }, { "grad_norm": 0.36438116431236267, "learning_rate": 6.55903461869765e-05, "loss": 0.0256, "step": 34330 }, { "grad_norm": 0.21181215345859528, "learning_rate": 6.557070689453465e-05, "loss": 0.0285, "step": 34340 }, { "grad_norm": 0.24964699149131775, "learning_rate": 6.555106494148482e-05, "loss": 0.0364, "step": 34350 }, { "grad_norm": 0.4075709581375122, "learning_rate": 6.553142033118333e-05, "loss": 0.0268, "step": 34360 }, { "grad_norm": 0.3536464273929596, "learning_rate": 6.551177306698688e-05, "loss": 0.0349, "step": 34370 }, { "grad_norm": 0.23864220082759857, "learning_rate": 6.549212315225267e-05, "loss": 0.0238, "step": 34380 }, { "grad_norm": 0.22949878871440887, "learning_rate": 6.547247059033833e-05, "loss": 0.0259, "step": 34390 }, { "grad_norm": 0.43283650279045105, "learning_rate": 6.545281538460193e-05, "loss": 0.0276, "step": 34400 }, { "grad_norm": 0.24440698325634003, "learning_rate": 6.543315753840202e-05, "loss": 0.0215, "step": 34410 }, { "grad_norm": 0.2220776081085205, "learning_rate": 6.541349705509758e-05, "loss": 0.0234, "step": 34420 }, { "grad_norm": 0.3028528094291687, "learning_rate": 6.539383393804805e-05, "loss": 0.0245, "step": 34430 }, { "grad_norm": 0.29651591181755066, "learning_rate": 6.537416819061333e-05, "loss": 0.0321, "step": 34440 }, { "grad_norm": 0.27985879778862, "learning_rate": 6.535449981615375e-05, "loss": 0.0318, "step": 34450 }, { "grad_norm": 0.28399232029914856, "learning_rate": 6.53348288180301e-05, "loss": 0.0228, "step": 34460 }, { "grad_norm": 0.4143885672092438, "learning_rate": 6.531515519960361e-05, "loss": 0.0307, "step": 34470 }, { "grad_norm": 0.28989389538764954, "learning_rate": 6.529547896423597e-05, "loss": 0.0239, "step": 34480 }, { "grad_norm": 0.5107706189155579, "learning_rate": 6.52758001152893e-05, "loss": 0.0343, "step": 34490 }, { "grad_norm": 0.2813883125782013, "learning_rate": 6.525611865612618e-05, "loss": 0.0221, "step": 34500 }, { "grad_norm": 0.2874565124511719, "learning_rate": 6.523643459010966e-05, "loss": 0.0288, "step": 34510 }, { "grad_norm": 0.2977565824985504, "learning_rate": 6.521674792060317e-05, "loss": 0.0317, "step": 34520 }, { "grad_norm": 0.4340466260910034, "learning_rate": 6.519705865097063e-05, "loss": 0.0266, "step": 34530 }, { "grad_norm": 0.34309637546539307, "learning_rate": 6.517736678457641e-05, "loss": 0.0263, "step": 34540 }, { "grad_norm": 0.2855578362941742, "learning_rate": 6.515767232478534e-05, "loss": 0.0295, "step": 34550 }, { "grad_norm": 0.19316190481185913, "learning_rate": 6.51379752749626e-05, "loss": 0.0208, "step": 34560 }, { "grad_norm": 0.37892383337020874, "learning_rate": 6.511827563847393e-05, "loss": 0.0281, "step": 34570 }, { "grad_norm": 0.3043605089187622, "learning_rate": 6.509857341868542e-05, "loss": 0.0275, "step": 34580 }, { "grad_norm": 0.32386377453804016, "learning_rate": 6.507886861896367e-05, "loss": 0.028, "step": 34590 }, { "grad_norm": 0.3466118574142456, "learning_rate": 6.505916124267567e-05, "loss": 0.0335, "step": 34600 }, { "grad_norm": 0.35813409090042114, "learning_rate": 6.503945129318891e-05, "loss": 0.0295, "step": 34610 }, { "grad_norm": 0.49797946214675903, "learning_rate": 6.501973877387122e-05, "loss": 0.0231, "step": 34620 }, { "grad_norm": 0.5871413350105286, "learning_rate": 6.500002368809098e-05, "loss": 0.0318, "step": 34630 }, { "grad_norm": 0.41195520758628845, "learning_rate": 6.498030603921694e-05, "loss": 0.0365, "step": 34640 }, { "grad_norm": 0.3479364216327667, "learning_rate": 6.496058583061832e-05, "loss": 0.0316, "step": 34650 }, { "grad_norm": 0.3674974739551544, "learning_rate": 6.494086306566475e-05, "loss": 0.0229, "step": 34660 }, { "grad_norm": 0.3702174127101898, "learning_rate": 6.492113774772632e-05, "loss": 0.0284, "step": 34670 }, { "grad_norm": 0.5774827599525452, "learning_rate": 6.490140988017354e-05, "loss": 0.0309, "step": 34680 }, { "grad_norm": 0.5998229384422302, "learning_rate": 6.488167946637736e-05, "loss": 0.0311, "step": 34690 }, { "grad_norm": 0.41226837038993835, "learning_rate": 6.486194650970915e-05, "loss": 0.0304, "step": 34700 }, { "grad_norm": 0.39414864778518677, "learning_rate": 6.48422110135408e-05, "loss": 0.0335, "step": 34710 }, { "grad_norm": 0.26271700859069824, "learning_rate": 6.482247298124451e-05, "loss": 0.031, "step": 34720 }, { "grad_norm": 0.22960828244686127, "learning_rate": 6.480273241619297e-05, "loss": 0.0252, "step": 34730 }, { "grad_norm": 0.3142012357711792, "learning_rate": 6.478298932175933e-05, "loss": 0.0232, "step": 34740 }, { "grad_norm": 0.36541104316711426, "learning_rate": 6.476324370131712e-05, "loss": 0.0309, "step": 34750 }, { "grad_norm": 0.528383731842041, "learning_rate": 6.474349555824036e-05, "loss": 0.04, "step": 34760 }, { "grad_norm": 0.36069226264953613, "learning_rate": 6.472374489590342e-05, "loss": 0.0253, "step": 34770 }, { "grad_norm": 0.3476284146308899, "learning_rate": 6.470399171768118e-05, "loss": 0.0443, "step": 34780 }, { "grad_norm": 0.30653196573257446, "learning_rate": 6.468423602694891e-05, "loss": 0.0251, "step": 34790 }, { "grad_norm": 0.25191840529441833, "learning_rate": 6.466447782708232e-05, "loss": 0.0257, "step": 34800 }, { "grad_norm": 0.23197470605373383, "learning_rate": 6.464471712145754e-05, "loss": 0.0224, "step": 34810 }, { "grad_norm": 0.29350385069847107, "learning_rate": 6.462495391345114e-05, "loss": 0.0241, "step": 34820 }, { "grad_norm": 0.47703447937965393, "learning_rate": 6.46051882064401e-05, "loss": 0.0241, "step": 34830 }, { "grad_norm": 0.32546907663345337, "learning_rate": 6.458542000380186e-05, "loss": 0.0241, "step": 34840 }, { "grad_norm": 0.3614700436592102, "learning_rate": 6.456564930891424e-05, "loss": 0.0232, "step": 34850 }, { "grad_norm": 0.40389782190322876, "learning_rate": 6.454587612515555e-05, "loss": 0.0348, "step": 34860 }, { "grad_norm": 0.2012755274772644, "learning_rate": 6.452610045590444e-05, "loss": 0.0242, "step": 34870 }, { "grad_norm": 0.4108094573020935, "learning_rate": 6.450632230454005e-05, "loss": 0.0292, "step": 34880 }, { "grad_norm": 0.2972099781036377, "learning_rate": 6.448654167444195e-05, "loss": 0.0313, "step": 34890 }, { "grad_norm": 0.5599693059921265, "learning_rate": 6.446675856899005e-05, "loss": 0.0239, "step": 34900 }, { "grad_norm": 0.17945979535579681, "learning_rate": 6.444697299156481e-05, "loss": 0.0249, "step": 34910 }, { "grad_norm": 0.40281903743743896, "learning_rate": 6.442718494554701e-05, "loss": 0.0313, "step": 34920 }, { "grad_norm": 0.5399630069732666, "learning_rate": 6.440739443431787e-05, "loss": 0.0198, "step": 34930 }, { "grad_norm": 0.20124174654483795, "learning_rate": 6.438760146125906e-05, "loss": 0.0173, "step": 34940 }, { "grad_norm": 0.43102067708969116, "learning_rate": 6.436780602975267e-05, "loss": 0.0275, "step": 34950 }, { "grad_norm": 0.35012075304985046, "learning_rate": 6.43480081431812e-05, "loss": 0.03, "step": 34960 }, { "grad_norm": 0.3014392554759979, "learning_rate": 6.432820780492756e-05, "loss": 0.0276, "step": 34970 }, { "grad_norm": 0.3004119396209717, "learning_rate": 6.430840501837506e-05, "loss": 0.0298, "step": 34980 }, { "grad_norm": 0.27437618374824524, "learning_rate": 6.428859978690748e-05, "loss": 0.0284, "step": 34990 }, { "grad_norm": 0.4311124384403229, "learning_rate": 6.426879211390901e-05, "loss": 0.0241, "step": 35000 }, { "grad_norm": 0.7109077572822571, "learning_rate": 6.424898200276422e-05, "loss": 0.03, "step": 35010 }, { "grad_norm": 0.33038800954818726, "learning_rate": 6.42291694568581e-05, "loss": 0.0242, "step": 35020 }, { "grad_norm": 0.39729052782058716, "learning_rate": 6.42093544795761e-05, "loss": 0.0347, "step": 35030 }, { "grad_norm": 0.3630200922489166, "learning_rate": 6.418953707430403e-05, "loss": 0.0265, "step": 35040 }, { "grad_norm": 0.34551846981048584, "learning_rate": 6.416971724442819e-05, "loss": 0.0219, "step": 35050 }, { "grad_norm": 0.22340650856494904, "learning_rate": 6.414989499333519e-05, "loss": 0.019, "step": 35060 }, { "grad_norm": 0.25811800360679626, "learning_rate": 6.413007032441214e-05, "loss": 0.0222, "step": 35070 }, { "grad_norm": 0.40618354082107544, "learning_rate": 6.411024324104653e-05, "loss": 0.0281, "step": 35080 }, { "grad_norm": 0.42333823442459106, "learning_rate": 6.409041374662628e-05, "loss": 0.0215, "step": 35090 }, { "grad_norm": 0.2667546570301056, "learning_rate": 6.407058184453967e-05, "loss": 0.0358, "step": 35100 }, { "grad_norm": 0.3484819829463959, "learning_rate": 6.405074753817548e-05, "loss": 0.0253, "step": 35110 }, { "grad_norm": 0.37472084164619446, "learning_rate": 6.40309108309228e-05, "loss": 0.0315, "step": 35120 }, { "grad_norm": 0.2916752099990845, "learning_rate": 6.401107172617122e-05, "loss": 0.0249, "step": 35130 }, { "grad_norm": 0.36491483449935913, "learning_rate": 6.399123022731068e-05, "loss": 0.0422, "step": 35140 }, { "grad_norm": 0.42585811018943787, "learning_rate": 6.397138633773157e-05, "loss": 0.0246, "step": 35150 }, { "grad_norm": 0.46853578090667725, "learning_rate": 6.395154006082463e-05, "loss": 0.0206, "step": 35160 }, { "grad_norm": 0.28059613704681396, "learning_rate": 6.393169139998109e-05, "loss": 0.0254, "step": 35170 }, { "grad_norm": 0.32820361852645874, "learning_rate": 6.39118403585925e-05, "loss": 0.0242, "step": 35180 }, { "grad_norm": 0.29865458607673645, "learning_rate": 6.38919869400509e-05, "loss": 0.0247, "step": 35190 }, { "grad_norm": 0.3920862078666687, "learning_rate": 6.387213114774865e-05, "loss": 0.0283, "step": 35200 }, { "grad_norm": 0.2907763123512268, "learning_rate": 6.385227298507863e-05, "loss": 0.0477, "step": 35210 }, { "grad_norm": 0.40691161155700684, "learning_rate": 6.3832412455434e-05, "loss": 0.0233, "step": 35220 }, { "grad_norm": 0.29133397340774536, "learning_rate": 6.381254956220841e-05, "loss": 0.0267, "step": 35230 }, { "grad_norm": 0.26960647106170654, "learning_rate": 6.379268430879586e-05, "loss": 0.0289, "step": 35240 }, { "grad_norm": 0.3288393020629883, "learning_rate": 6.37728166985908e-05, "loss": 0.0222, "step": 35250 }, { "grad_norm": 0.3855660557746887, "learning_rate": 6.375294673498804e-05, "loss": 0.0211, "step": 35260 }, { "grad_norm": 0.2503397762775421, "learning_rate": 6.373307442138284e-05, "loss": 0.025, "step": 35270 }, { "grad_norm": 0.38787639141082764, "learning_rate": 6.371319976117081e-05, "loss": 0.036, "step": 35280 }, { "grad_norm": 0.24274523556232452, "learning_rate": 6.3693322757748e-05, "loss": 0.026, "step": 35290 }, { "grad_norm": 0.42827558517456055, "learning_rate": 6.367344341451086e-05, "loss": 0.028, "step": 35300 }, { "grad_norm": 0.412676602602005, "learning_rate": 6.36535617348562e-05, "loss": 0.0241, "step": 35310 }, { "grad_norm": 0.3476279079914093, "learning_rate": 6.363367772218128e-05, "loss": 0.0272, "step": 35320 }, { "grad_norm": 0.41531068086624146, "learning_rate": 6.36137913798837e-05, "loss": 0.0277, "step": 35330 }, { "grad_norm": 0.25264158844947815, "learning_rate": 6.359390271136151e-05, "loss": 0.0229, "step": 35340 }, { "grad_norm": 0.3152482509613037, "learning_rate": 6.357401172001314e-05, "loss": 0.0239, "step": 35350 }, { "grad_norm": 0.2747821509838104, "learning_rate": 6.355411840923742e-05, "loss": 0.0274, "step": 35360 }, { "grad_norm": 0.33435702323913574, "learning_rate": 6.353422278243358e-05, "loss": 0.0218, "step": 35370 }, { "grad_norm": 0.4383895993232727, "learning_rate": 6.351432484300121e-05, "loss": 0.0292, "step": 35380 }, { "grad_norm": 0.36104145646095276, "learning_rate": 6.349442459434036e-05, "loss": 0.0308, "step": 35390 }, { "grad_norm": 0.6232671141624451, "learning_rate": 6.34745220398514e-05, "loss": 0.0325, "step": 35400 }, { "grad_norm": 0.3327662944793701, "learning_rate": 6.345461718293518e-05, "loss": 0.0242, "step": 35410 }, { "grad_norm": 0.33406510949134827, "learning_rate": 6.343471002699286e-05, "loss": 0.0337, "step": 35420 }, { "grad_norm": 0.32301273941993713, "learning_rate": 6.341480057542602e-05, "loss": 0.0302, "step": 35430 }, { "grad_norm": 0.36815890669822693, "learning_rate": 6.339488883163667e-05, "loss": 0.026, "step": 35440 }, { "grad_norm": 0.2527957856655121, "learning_rate": 6.337497479902716e-05, "loss": 0.0233, "step": 35450 }, { "grad_norm": 0.29171791672706604, "learning_rate": 6.335505848100027e-05, "loss": 0.0244, "step": 35460 }, { "grad_norm": 0.4746594727039337, "learning_rate": 6.333513988095915e-05, "loss": 0.0257, "step": 35470 }, { "grad_norm": 0.44581538438796997, "learning_rate": 6.331521900230735e-05, "loss": 0.0241, "step": 35480 }, { "grad_norm": 0.6469883918762207, "learning_rate": 6.329529584844878e-05, "loss": 0.0264, "step": 35490 }, { "grad_norm": 0.37743398547172546, "learning_rate": 6.327537042278777e-05, "loss": 0.0384, "step": 35500 }, { "grad_norm": 0.3779277801513672, "learning_rate": 6.325544272872905e-05, "loss": 0.0253, "step": 35510 }, { "grad_norm": 0.26732122898101807, "learning_rate": 6.323551276967771e-05, "loss": 0.03, "step": 35520 }, { "grad_norm": 0.4659290909767151, "learning_rate": 6.321558054903922e-05, "loss": 0.0235, "step": 35530 }, { "grad_norm": 0.1702430248260498, "learning_rate": 6.319564607021947e-05, "loss": 0.0199, "step": 35540 }, { "grad_norm": 0.5175513625144958, "learning_rate": 6.31757093366247e-05, "loss": 0.0261, "step": 35550 }, { "grad_norm": 0.4118870496749878, "learning_rate": 6.315577035166154e-05, "loss": 0.0255, "step": 35560 }, { "grad_norm": 0.26017725467681885, "learning_rate": 6.313582911873708e-05, "loss": 0.0215, "step": 35570 }, { "grad_norm": 0.3026370108127594, "learning_rate": 6.311588564125865e-05, "loss": 0.0203, "step": 35580 }, { "grad_norm": 0.4294586479663849, "learning_rate": 6.30959399226341e-05, "loss": 0.0276, "step": 35590 }, { "grad_norm": 0.27899861335754395, "learning_rate": 6.30759919662716e-05, "loss": 0.0334, "step": 35600 }, { "grad_norm": 0.2877977192401886, "learning_rate": 6.30560417755797e-05, "loss": 0.0262, "step": 35610 }, { "grad_norm": 0.26223909854888916, "learning_rate": 6.303608935396735e-05, "loss": 0.025, "step": 35620 }, { "grad_norm": 0.2359170764684677, "learning_rate": 6.301613470484386e-05, "loss": 0.022, "step": 35630 }, { "grad_norm": 0.36155667901039124, "learning_rate": 6.299617783161893e-05, "loss": 0.0302, "step": 35640 }, { "grad_norm": 0.22196397185325623, "learning_rate": 6.297621873770266e-05, "loss": 0.026, "step": 35650 }, { "grad_norm": 0.35080137848854065, "learning_rate": 6.29562574265055e-05, "loss": 0.0284, "step": 35660 }, { "grad_norm": 0.4246968924999237, "learning_rate": 6.293629390143834e-05, "loss": 0.0274, "step": 35670 }, { "grad_norm": 0.22813209891319275, "learning_rate": 6.291632816591232e-05, "loss": 0.0193, "step": 35680 }, { "grad_norm": 0.40134045481681824, "learning_rate": 6.28963602233391e-05, "loss": 0.028, "step": 35690 }, { "grad_norm": 0.3556760251522064, "learning_rate": 6.287639007713062e-05, "loss": 0.0318, "step": 35700 }, { "grad_norm": 0.5022999048233032, "learning_rate": 6.285641773069926e-05, "loss": 0.0236, "step": 35710 }, { "grad_norm": 0.2937120795249939, "learning_rate": 6.283644318745773e-05, "loss": 0.0227, "step": 35720 }, { "grad_norm": 0.2662305235862732, "learning_rate": 6.281646645081912e-05, "loss": 0.0351, "step": 35730 }, { "grad_norm": 0.38376152515411377, "learning_rate": 6.279648752419693e-05, "loss": 0.0234, "step": 35740 }, { "grad_norm": 0.27487432956695557, "learning_rate": 6.2776506411005e-05, "loss": 0.0237, "step": 35750 }, { "grad_norm": 0.40781933069229126, "learning_rate": 6.275652311465758e-05, "loss": 0.0274, "step": 35760 }, { "grad_norm": 0.2875462472438812, "learning_rate": 6.273653763856926e-05, "loss": 0.0272, "step": 35770 }, { "grad_norm": 0.3151354491710663, "learning_rate": 6.271654998615501e-05, "loss": 0.0252, "step": 35780 }, { "grad_norm": 0.2812210023403168, "learning_rate": 6.269656016083013e-05, "loss": 0.0321, "step": 35790 }, { "grad_norm": 0.2755124270915985, "learning_rate": 6.267656816601038e-05, "loss": 0.0346, "step": 35800 }, { "grad_norm": 0.3129064738750458, "learning_rate": 6.265657400511185e-05, "loss": 0.0303, "step": 35810 }, { "grad_norm": 0.35375910997390747, "learning_rate": 6.263657768155098e-05, "loss": 0.0307, "step": 35820 }, { "grad_norm": 0.2674326002597809, "learning_rate": 6.261657919874457e-05, "loss": 0.0236, "step": 35830 }, { "grad_norm": 0.3240218758583069, "learning_rate": 6.259657856010986e-05, "loss": 0.0323, "step": 35840 }, { "grad_norm": 0.44451799988746643, "learning_rate": 6.257657576906439e-05, "loss": 0.0282, "step": 35850 }, { "grad_norm": 0.3781532049179077, "learning_rate": 6.255657082902609e-05, "loss": 0.0233, "step": 35860 }, { "grad_norm": 0.23772232234477997, "learning_rate": 6.253656374341325e-05, "loss": 0.0226, "step": 35870 }, { "grad_norm": 0.3225371241569519, "learning_rate": 6.251655451564457e-05, "loss": 0.025, "step": 35880 }, { "grad_norm": 0.29614028334617615, "learning_rate": 6.249654314913902e-05, "loss": 0.02, "step": 35890 }, { "grad_norm": 0.22676274180412292, "learning_rate": 6.247652964731604e-05, "loss": 0.0249, "step": 35900 }, { "grad_norm": 0.3568111062049866, "learning_rate": 6.245651401359537e-05, "loss": 0.0389, "step": 35910 }, { "grad_norm": 0.2952530086040497, "learning_rate": 6.243649625139715e-05, "loss": 0.0271, "step": 35920 }, { "grad_norm": 0.3556642234325409, "learning_rate": 6.241647636414185e-05, "loss": 0.0293, "step": 35930 }, { "grad_norm": 0.32967472076416016, "learning_rate": 6.239645435525034e-05, "loss": 0.0345, "step": 35940 }, { "grad_norm": 0.27085810899734497, "learning_rate": 6.237643022814381e-05, "loss": 0.0265, "step": 35950 }, { "grad_norm": 0.35168132185935974, "learning_rate": 6.235640398624386e-05, "loss": 0.0253, "step": 35960 }, { "grad_norm": 1.0557340383529663, "learning_rate": 6.233637563297243e-05, "loss": 0.0303, "step": 35970 }, { "grad_norm": 0.6658965349197388, "learning_rate": 6.23163451717518e-05, "loss": 0.0327, "step": 35980 }, { "grad_norm": 0.40491461753845215, "learning_rate": 6.229631260600463e-05, "loss": 0.0256, "step": 35990 }, { "grad_norm": 0.27782052755355835, "learning_rate": 6.227627793915392e-05, "loss": 0.0339, "step": 36000 }, { "grad_norm": 0.324034720659256, "learning_rate": 6.225624117462309e-05, "loss": 0.0284, "step": 36010 }, { "grad_norm": 0.31046581268310547, "learning_rate": 6.223620231583586e-05, "loss": 0.0311, "step": 36020 }, { "grad_norm": 0.6168403625488281, "learning_rate": 6.221616136621629e-05, "loss": 0.0253, "step": 36030 }, { "grad_norm": 0.2763499319553375, "learning_rate": 6.219611832918887e-05, "loss": 0.0293, "step": 36040 }, { "grad_norm": 0.3284352719783783, "learning_rate": 6.217607320817838e-05, "loss": 0.0266, "step": 36050 }, { "grad_norm": 0.42665114998817444, "learning_rate": 6.215602600661001e-05, "loss": 0.024, "step": 36060 }, { "grad_norm": 0.2984261214733124, "learning_rate": 6.213597672790925e-05, "loss": 0.0228, "step": 36070 }, { "grad_norm": 0.3413061499595642, "learning_rate": 6.2115925375502e-05, "loss": 0.0228, "step": 36080 }, { "grad_norm": 0.34932050108909607, "learning_rate": 6.209587195281447e-05, "loss": 0.0254, "step": 36090 }, { "grad_norm": 0.3046252429485321, "learning_rate": 6.207581646327324e-05, "loss": 0.0301, "step": 36100 }, { "grad_norm": 0.2749488353729248, "learning_rate": 6.205575891030526e-05, "loss": 0.0193, "step": 36110 }, { "grad_norm": 0.2606362998485565, "learning_rate": 6.203569929733781e-05, "loss": 0.022, "step": 36120 }, { "grad_norm": 0.3441266715526581, "learning_rate": 6.201563762779852e-05, "loss": 0.0249, "step": 36130 }, { "grad_norm": 0.8153908848762512, "learning_rate": 6.199557390511538e-05, "loss": 0.0225, "step": 36140 }, { "grad_norm": 0.31515392661094666, "learning_rate": 6.197550813271675e-05, "loss": 0.024, "step": 36150 }, { "grad_norm": 0.32050812244415283, "learning_rate": 6.195544031403131e-05, "loss": 0.0231, "step": 36160 }, { "grad_norm": 0.42259010672569275, "learning_rate": 6.19353704524881e-05, "loss": 0.0255, "step": 36170 }, { "grad_norm": 0.4423321783542633, "learning_rate": 6.191529855151652e-05, "loss": 0.0233, "step": 36180 }, { "grad_norm": 0.26174724102020264, "learning_rate": 6.189522461454629e-05, "loss": 0.025, "step": 36190 }, { "grad_norm": 0.28980034589767456, "learning_rate": 6.187514864500752e-05, "loss": 0.0293, "step": 36200 }, { "grad_norm": 0.4337232708930969, "learning_rate": 6.185507064633062e-05, "loss": 0.0262, "step": 36210 }, { "grad_norm": 0.3272239565849304, "learning_rate": 6.18349906219464e-05, "loss": 0.0278, "step": 36220 }, { "grad_norm": 0.319508820772171, "learning_rate": 6.181490857528596e-05, "loss": 0.0206, "step": 36230 }, { "grad_norm": 0.4753415584564209, "learning_rate": 6.179482450978077e-05, "loss": 0.0165, "step": 36240 }, { "grad_norm": 0.27555301785469055, "learning_rate": 6.177473842886269e-05, "loss": 0.0209, "step": 36250 }, { "grad_norm": 0.3644145727157593, "learning_rate": 6.175465033596382e-05, "loss": 0.028, "step": 36260 }, { "grad_norm": 0.23792040348052979, "learning_rate": 6.173456023451671e-05, "loss": 0.0237, "step": 36270 }, { "grad_norm": 0.3000107705593109, "learning_rate": 6.171446812795422e-05, "loss": 0.0262, "step": 36280 }, { "grad_norm": 0.277654767036438, "learning_rate": 6.169437401970949e-05, "loss": 0.0372, "step": 36290 }, { "grad_norm": 0.36544957756996155, "learning_rate": 6.16742779132161e-05, "loss": 0.0219, "step": 36300 }, { "grad_norm": 0.2922668755054474, "learning_rate": 6.165417981190789e-05, "loss": 0.0312, "step": 36310 }, { "grad_norm": 0.5714682340621948, "learning_rate": 6.16340797192191e-05, "loss": 0.0221, "step": 36320 }, { "grad_norm": 0.281732976436615, "learning_rate": 6.161397763858427e-05, "loss": 0.0299, "step": 36330 }, { "grad_norm": 0.4038659632205963, "learning_rate": 6.159387357343834e-05, "loss": 0.0203, "step": 36340 }, { "grad_norm": 0.22363623976707458, "learning_rate": 6.157376752721648e-05, "loss": 0.0293, "step": 36350 }, { "grad_norm": 0.5456743836402893, "learning_rate": 6.155365950335428e-05, "loss": 0.0291, "step": 36360 }, { "grad_norm": 0.6949955224990845, "learning_rate": 6.153354950528768e-05, "loss": 0.0303, "step": 36370 }, { "grad_norm": 0.27305322885513306, "learning_rate": 6.151343753645293e-05, "loss": 0.0295, "step": 36380 }, { "grad_norm": 0.29411056637763977, "learning_rate": 6.149332360028657e-05, "loss": 0.0255, "step": 36390 }, { "grad_norm": 0.29433631896972656, "learning_rate": 6.147320770022555e-05, "loss": 0.0273, "step": 36400 }, { "grad_norm": 0.23291553556919098, "learning_rate": 6.145308983970715e-05, "loss": 0.0257, "step": 36410 }, { "grad_norm": 0.49734777212142944, "learning_rate": 6.143297002216892e-05, "loss": 0.0258, "step": 36420 }, { "grad_norm": 0.386898010969162, "learning_rate": 6.141284825104882e-05, "loss": 0.0285, "step": 36430 }, { "grad_norm": 0.2057504951953888, "learning_rate": 6.13927245297851e-05, "loss": 0.0273, "step": 36440 }, { "grad_norm": 0.29010623693466187, "learning_rate": 6.137259886181633e-05, "loss": 0.0291, "step": 36450 }, { "grad_norm": 0.2873702943325043, "learning_rate": 6.135247125058145e-05, "loss": 0.0269, "step": 36460 }, { "grad_norm": 0.21861249208450317, "learning_rate": 6.133234169951974e-05, "loss": 0.0189, "step": 36470 }, { "grad_norm": 0.29154571890830994, "learning_rate": 6.131221021207078e-05, "loss": 0.0265, "step": 36480 }, { "grad_norm": 0.34198570251464844, "learning_rate": 6.129207679167448e-05, "loss": 0.0248, "step": 36490 }, { "grad_norm": 0.24822133779525757, "learning_rate": 6.127194144177109e-05, "loss": 0.0203, "step": 36500 }, { "grad_norm": 0.2972579300403595, "learning_rate": 6.125180416580118e-05, "loss": 0.023, "step": 36510 }, { "grad_norm": 0.2842023968696594, "learning_rate": 6.123166496720571e-05, "loss": 0.0253, "step": 36520 }, { "grad_norm": 1.0711121559143066, "learning_rate": 6.121152384942588e-05, "loss": 0.0204, "step": 36530 }, { "grad_norm": 0.3499973714351654, "learning_rate": 6.119138081590324e-05, "loss": 0.0267, "step": 36540 }, { "grad_norm": 0.2936144173145294, "learning_rate": 6.117123587007971e-05, "loss": 0.0207, "step": 36550 }, { "grad_norm": 0.33639898896217346, "learning_rate": 6.11510890153975e-05, "loss": 0.0279, "step": 36560 }, { "grad_norm": 0.5157155394554138, "learning_rate": 6.113094025529916e-05, "loss": 0.0258, "step": 36570 }, { "grad_norm": 0.22856555879116058, "learning_rate": 6.111078959322757e-05, "loss": 0.0339, "step": 36580 }, { "grad_norm": 0.3439182937145233, "learning_rate": 6.109063703262592e-05, "loss": 0.0352, "step": 36590 }, { "grad_norm": 0.2718384861946106, "learning_rate": 6.107048257693772e-05, "loss": 0.0248, "step": 36600 }, { "grad_norm": 0.3416726291179657, "learning_rate": 6.105032622960683e-05, "loss": 0.0269, "step": 36610 }, { "grad_norm": 0.38402026891708374, "learning_rate": 6.103016799407743e-05, "loss": 0.0266, "step": 36620 }, { "grad_norm": 0.33391010761260986, "learning_rate": 6.1010007873793984e-05, "loss": 0.0355, "step": 36630 }, { "grad_norm": 0.30307164788246155, "learning_rate": 6.098984587220131e-05, "loss": 0.0267, "step": 36640 }, { "grad_norm": 0.21980121731758118, "learning_rate": 6.096968199274456e-05, "loss": 0.0269, "step": 36650 }, { "grad_norm": 0.3484470844268799, "learning_rate": 6.0949516238869166e-05, "loss": 0.0263, "step": 36660 }, { "grad_norm": 0.2887048125267029, "learning_rate": 6.092934861402092e-05, "loss": 0.032, "step": 36670 }, { "grad_norm": 0.4606006145477295, "learning_rate": 6.0909179121645924e-05, "loss": 0.0356, "step": 36680 }, { "grad_norm": 0.32803434133529663, "learning_rate": 6.0889007765190576e-05, "loss": 0.0318, "step": 36690 }, { "grad_norm": 0.2113298624753952, "learning_rate": 6.086883454810162e-05, "loss": 0.0248, "step": 36700 }, { "grad_norm": 0.3060060143470764, "learning_rate": 6.0848659473826084e-05, "loss": 0.0224, "step": 36710 }, { "grad_norm": 0.38111230731010437, "learning_rate": 6.082848254581138e-05, "loss": 0.0263, "step": 36720 }, { "grad_norm": 0.28828126192092896, "learning_rate": 6.080830376750517e-05, "loss": 0.0227, "step": 36730 }, { "grad_norm": 0.329355925321579, "learning_rate": 6.0788123142355445e-05, "loss": 0.0232, "step": 36740 }, { "grad_norm": 0.3732445240020752, "learning_rate": 6.076794067381052e-05, "loss": 0.029, "step": 36750 }, { "grad_norm": 0.6374455094337463, "learning_rate": 6.074775636531905e-05, "loss": 0.0262, "step": 36760 }, { "grad_norm": 0.424908310174942, "learning_rate": 6.072757022032997e-05, "loss": 0.0305, "step": 36770 }, { "grad_norm": 0.22850307822227478, "learning_rate": 6.070738224229253e-05, "loss": 0.0272, "step": 36780 }, { "grad_norm": 0.260703444480896, "learning_rate": 6.0687192434656314e-05, "loss": 0.0282, "step": 36790 }, { "grad_norm": 0.262813001871109, "learning_rate": 6.066700080087121e-05, "loss": 0.0247, "step": 36800 }, { "grad_norm": 0.23624397814273834, "learning_rate": 6.0646807344387424e-05, "loss": 0.0216, "step": 36810 }, { "grad_norm": 0.35292553901672363, "learning_rate": 6.062661206865543e-05, "loss": 0.029, "step": 36820 }, { "grad_norm": 0.296561598777771, "learning_rate": 6.06064149771261e-05, "loss": 0.0265, "step": 36830 }, { "grad_norm": 0.30112606287002563, "learning_rate": 6.058621607325051e-05, "loss": 0.0258, "step": 36840 }, { "grad_norm": 0.34090757369995117, "learning_rate": 6.056601536048014e-05, "loss": 0.0247, "step": 36850 }, { "grad_norm": 0.3782539963722229, "learning_rate": 6.0545812842266725e-05, "loss": 0.0324, "step": 36860 }, { "grad_norm": 0.21864435076713562, "learning_rate": 6.052560852206232e-05, "loss": 0.024, "step": 36870 }, { "grad_norm": 0.34207671880722046, "learning_rate": 6.05054024033193e-05, "loss": 0.0384, "step": 36880 }, { "grad_norm": 0.33525609970092773, "learning_rate": 6.048519448949032e-05, "loss": 0.0243, "step": 36890 }, { "grad_norm": 0.35072410106658936, "learning_rate": 6.046498478402839e-05, "loss": 0.024, "step": 36900 }, { "grad_norm": 0.2378273755311966, "learning_rate": 6.044477329038677e-05, "loss": 0.0248, "step": 36910 }, { "grad_norm": 0.3112776577472687, "learning_rate": 6.042456001201906e-05, "loss": 0.0227, "step": 36920 }, { "grad_norm": 0.31948286294937134, "learning_rate": 6.040434495237917e-05, "loss": 0.0244, "step": 36930 }, { "grad_norm": 0.4263837933540344, "learning_rate": 6.0384128114921256e-05, "loss": 0.0246, "step": 36940 }, { "grad_norm": 0.2942236065864563, "learning_rate": 6.036390950309987e-05, "loss": 0.0225, "step": 36950 }, { "grad_norm": 0.3016320765018463, "learning_rate": 6.0343689120369805e-05, "loss": 0.0298, "step": 36960 }, { "grad_norm": 0.3237210512161255, "learning_rate": 6.032346697018616e-05, "loss": 0.0195, "step": 36970 }, { "grad_norm": 0.2963261306285858, "learning_rate": 6.0303243056004375e-05, "loss": 0.0399, "step": 36980 }, { "grad_norm": 0.25295111536979675, "learning_rate": 6.0283017381280136e-05, "loss": 0.023, "step": 36990 }, { "grad_norm": 0.3716724216938019, "learning_rate": 6.026278994946948e-05, "loss": 0.0217, "step": 37000 }, { "grad_norm": 0.4177488386631012, "learning_rate": 6.02425607640287e-05, "loss": 0.0284, "step": 37010 }, { "grad_norm": 0.38040295243263245, "learning_rate": 6.022232982841441e-05, "loss": 0.0243, "step": 37020 }, { "grad_norm": 0.26963701844215393, "learning_rate": 6.020209714608355e-05, "loss": 0.0248, "step": 37030 }, { "grad_norm": 0.39045780897140503, "learning_rate": 6.018186272049331e-05, "loss": 0.0276, "step": 37040 }, { "grad_norm": 0.2807769775390625, "learning_rate": 6.01616265551012e-05, "loss": 0.0242, "step": 37050 }, { "grad_norm": 0.25001826882362366, "learning_rate": 6.014138865336503e-05, "loss": 0.0216, "step": 37060 }, { "grad_norm": 0.25691768527030945, "learning_rate": 6.0121149018742905e-05, "loss": 0.0267, "step": 37070 }, { "grad_norm": 0.35854703187942505, "learning_rate": 6.010090765469325e-05, "loss": 0.0228, "step": 37080 }, { "grad_norm": 0.2983665466308594, "learning_rate": 6.008066456467473e-05, "loss": 0.0226, "step": 37090 }, { "grad_norm": 0.2778838276863098, "learning_rate": 6.0060419752146335e-05, "loss": 0.0232, "step": 37100 }, { "grad_norm": 0.29856622219085693, "learning_rate": 6.0040173220567353e-05, "loss": 0.0228, "step": 37110 }, { "grad_norm": 0.22912456095218658, "learning_rate": 6.001992497339737e-05, "loss": 0.0193, "step": 37120 }, { "grad_norm": 0.3080194890499115, "learning_rate": 5.999967501409626e-05, "loss": 0.0317, "step": 37130 }, { "grad_norm": 0.21051505208015442, "learning_rate": 5.997942334612418e-05, "loss": 0.0291, "step": 37140 }, { "grad_norm": 0.2103281170129776, "learning_rate": 5.995916997294158e-05, "loss": 0.0174, "step": 37150 }, { "grad_norm": 0.249118834733963, "learning_rate": 5.9938914898009214e-05, "loss": 0.0297, "step": 37160 }, { "grad_norm": 0.425952672958374, "learning_rate": 5.991865812478813e-05, "loss": 0.02, "step": 37170 }, { "grad_norm": 0.42602071166038513, "learning_rate": 5.989839965673964e-05, "loss": 0.0351, "step": 37180 }, { "grad_norm": 0.3374117910861969, "learning_rate": 5.987813949732539e-05, "loss": 0.0241, "step": 37190 }, { "grad_norm": 0.212883859872818, "learning_rate": 5.9857877650007255e-05, "loss": 0.0231, "step": 37200 }, { "grad_norm": 0.2410152554512024, "learning_rate": 5.983761411824744e-05, "loss": 0.0242, "step": 37210 }, { "grad_norm": 0.35189709067344666, "learning_rate": 5.981734890550844e-05, "loss": 0.0281, "step": 37220 }, { "grad_norm": 0.31702205538749695, "learning_rate": 5.979708201525301e-05, "loss": 0.0284, "step": 37230 }, { "grad_norm": 0.38243263959884644, "learning_rate": 5.977681345094422e-05, "loss": 0.0287, "step": 37240 }, { "grad_norm": 0.2907295227050781, "learning_rate": 5.97565432160454e-05, "loss": 0.024, "step": 37250 }, { "grad_norm": 0.2605346143245697, "learning_rate": 5.9736271314020186e-05, "loss": 0.0305, "step": 37260 }, { "grad_norm": 0.43596741557121277, "learning_rate": 5.971599774833251e-05, "loss": 0.0279, "step": 37270 }, { "grad_norm": 0.33457401394844055, "learning_rate": 5.9695722522446525e-05, "loss": 0.0265, "step": 37280 }, { "grad_norm": 0.3552413880825043, "learning_rate": 5.9675445639826765e-05, "loss": 0.0279, "step": 37290 }, { "grad_norm": 0.26952433586120605, "learning_rate": 5.965516710393796e-05, "loss": 0.0246, "step": 37300 }, { "grad_norm": 0.2814500629901886, "learning_rate": 5.963488691824516e-05, "loss": 0.0326, "step": 37310 }, { "grad_norm": 0.3988037705421448, "learning_rate": 5.96146050862137e-05, "loss": 0.0339, "step": 37320 }, { "grad_norm": 0.3591569662094116, "learning_rate": 5.959432161130919e-05, "loss": 0.0241, "step": 37330 }, { "grad_norm": 0.20065313577651978, "learning_rate": 5.9574036496997545e-05, "loss": 0.0288, "step": 37340 }, { "grad_norm": 0.27868983149528503, "learning_rate": 5.955374974674488e-05, "loss": 0.0247, "step": 37350 }, { "grad_norm": 0.21530456840991974, "learning_rate": 5.9533461364017696e-05, "loss": 0.0246, "step": 37360 }, { "grad_norm": 0.25289636850357056, "learning_rate": 5.9513171352282716e-05, "loss": 0.0236, "step": 37370 }, { "grad_norm": 0.22630885243415833, "learning_rate": 5.949287971500692e-05, "loss": 0.024, "step": 37380 }, { "grad_norm": 0.25792232155799866, "learning_rate": 5.947258645565762e-05, "loss": 0.0218, "step": 37390 }, { "grad_norm": 0.23890814185142517, "learning_rate": 5.945229157770237e-05, "loss": 0.0239, "step": 37400 }, { "grad_norm": 0.26701784133911133, "learning_rate": 5.9431995084609006e-05, "loss": 0.0252, "step": 37410 }, { "grad_norm": 0.3040284216403961, "learning_rate": 5.941169697984564e-05, "loss": 0.025, "step": 37420 }, { "grad_norm": 0.31965172290802, "learning_rate": 5.9391397266880675e-05, "loss": 0.0239, "step": 37430 }, { "grad_norm": 0.3803396224975586, "learning_rate": 5.937109594918279e-05, "loss": 0.0383, "step": 37440 }, { "grad_norm": 0.3746122419834137, "learning_rate": 5.9350793030220884e-05, "loss": 0.0446, "step": 37450 }, { "grad_norm": 0.314655601978302, "learning_rate": 5.933048851346421e-05, "loss": 0.025, "step": 37460 }, { "grad_norm": 0.38132068514823914, "learning_rate": 5.931018240238224e-05, "loss": 0.0351, "step": 37470 }, { "grad_norm": 0.2856760323047638, "learning_rate": 5.928987470044471e-05, "loss": 0.0263, "step": 37480 }, { "grad_norm": 0.3872112035751343, "learning_rate": 5.9269565411121695e-05, "loss": 0.024, "step": 37490 }, { "grad_norm": 0.26279279589653015, "learning_rate": 5.924925453788347e-05, "loss": 0.0304, "step": 37500 }, { "grad_norm": 0.294650137424469, "learning_rate": 5.92289420842006e-05, "loss": 0.0275, "step": 37510 }, { "grad_norm": 0.37593749165534973, "learning_rate": 5.9208628053543945e-05, "loss": 0.0301, "step": 37520 }, { "grad_norm": 0.41451284289360046, "learning_rate": 5.918831244938462e-05, "loss": 0.0273, "step": 37530 }, { "grad_norm": 0.33689188957214355, "learning_rate": 5.9167995275194e-05, "loss": 0.0235, "step": 37540 }, { "grad_norm": 0.421710729598999, "learning_rate": 5.914767653444373e-05, "loss": 0.029, "step": 37550 }, { "grad_norm": 0.5095646977424622, "learning_rate": 5.912735623060572e-05, "loss": 0.0365, "step": 37560 }, { "grad_norm": 0.38840851187705994, "learning_rate": 5.910703436715217e-05, "loss": 0.0302, "step": 37570 }, { "grad_norm": 0.22392185032367706, "learning_rate": 5.908671094755552e-05, "loss": 0.0225, "step": 37580 }, { "grad_norm": 0.3796482980251312, "learning_rate": 5.906638597528851e-05, "loss": 0.0232, "step": 37590 }, { "grad_norm": 0.3120202422142029, "learning_rate": 5.9046059453824076e-05, "loss": 0.0239, "step": 37600 }, { "grad_norm": 0.2940347194671631, "learning_rate": 5.9025731386635505e-05, "loss": 0.0228, "step": 37610 }, { "grad_norm": 0.34336578845977783, "learning_rate": 5.900540177719629e-05, "loss": 0.0209, "step": 37620 }, { "grad_norm": 0.33778414130210876, "learning_rate": 5.898507062898021e-05, "loss": 0.027, "step": 37630 }, { "grad_norm": 0.42138174176216125, "learning_rate": 5.8964737945461316e-05, "loss": 0.0217, "step": 37640 }, { "grad_norm": 0.19461701810359955, "learning_rate": 5.8944403730113885e-05, "loss": 0.0236, "step": 37650 }, { "grad_norm": 0.2994660437107086, "learning_rate": 5.892406798641248e-05, "loss": 0.0248, "step": 37660 }, { "grad_norm": 0.3911971151828766, "learning_rate": 5.890373071783193e-05, "loss": 0.0229, "step": 37670 }, { "grad_norm": 0.5264248847961426, "learning_rate": 5.888339192784732e-05, "loss": 0.0349, "step": 37680 }, { "grad_norm": 0.29040905833244324, "learning_rate": 5.8863051619934003e-05, "loss": 0.0234, "step": 37690 }, { "grad_norm": 0.26712745428085327, "learning_rate": 5.8842709797567554e-05, "loss": 0.024, "step": 37700 }, { "grad_norm": 0.3139951229095459, "learning_rate": 5.8822366464223855e-05, "loss": 0.0248, "step": 37710 }, { "grad_norm": 0.26927420496940613, "learning_rate": 5.880202162337901e-05, "loss": 0.0282, "step": 37720 }, { "grad_norm": 1.168687105178833, "learning_rate": 5.8781675278509405e-05, "loss": 0.027, "step": 37730 }, { "grad_norm": 0.22864378988742828, "learning_rate": 5.8761327433091696e-05, "loss": 0.02, "step": 37740 }, { "grad_norm": 0.3847953975200653, "learning_rate": 5.874097809060275e-05, "loss": 0.0272, "step": 37750 }, { "grad_norm": 0.24985894560813904, "learning_rate": 5.87206272545197e-05, "loss": 0.0409, "step": 37760 }, { "grad_norm": 0.2807951271533966, "learning_rate": 5.8700274928319955e-05, "loss": 0.0252, "step": 37770 }, { "grad_norm": 0.39183348417282104, "learning_rate": 5.867992111548118e-05, "loss": 0.0219, "step": 37780 }, { "grad_norm": 0.26922374963760376, "learning_rate": 5.865956581948131e-05, "loss": 0.0215, "step": 37790 }, { "grad_norm": 0.2447250634431839, "learning_rate": 5.863920904379845e-05, "loss": 0.0198, "step": 37800 }, { "grad_norm": 0.41969913244247437, "learning_rate": 5.8618850791911064e-05, "loss": 0.0312, "step": 37810 }, { "grad_norm": 0.27881261706352234, "learning_rate": 5.859849106729779e-05, "loss": 0.0254, "step": 37820 }, { "grad_norm": 0.3571089208126068, "learning_rate": 5.857812987343758e-05, "loss": 0.0236, "step": 37830 }, { "grad_norm": 0.9107143878936768, "learning_rate": 5.855776721380957e-05, "loss": 0.0279, "step": 37840 }, { "grad_norm": 0.32186025381088257, "learning_rate": 5.8537403091893217e-05, "loss": 0.0243, "step": 37850 }, { "grad_norm": 0.33319151401519775, "learning_rate": 5.851703751116816e-05, "loss": 0.0269, "step": 37860 }, { "grad_norm": 0.28864187002182007, "learning_rate": 5.8496670475114336e-05, "loss": 0.0227, "step": 37870 }, { "grad_norm": 0.37344932556152344, "learning_rate": 5.84763019872119e-05, "loss": 0.0339, "step": 37880 }, { "grad_norm": 0.33684104681015015, "learning_rate": 5.845593205094131e-05, "loss": 0.0234, "step": 37890 }, { "grad_norm": 0.21727325022220612, "learning_rate": 5.843556066978318e-05, "loss": 0.0207, "step": 37900 }, { "grad_norm": 0.26812922954559326, "learning_rate": 5.8415187847218455e-05, "loss": 0.0262, "step": 37910 }, { "grad_norm": 0.35331085324287415, "learning_rate": 5.839481358672827e-05, "loss": 0.0224, "step": 37920 }, { "grad_norm": 0.27370208501815796, "learning_rate": 5.837443789179407e-05, "loss": 0.0261, "step": 37930 }, { "grad_norm": 0.4295285940170288, "learning_rate": 5.8354060765897445e-05, "loss": 0.0253, "step": 37940 }, { "grad_norm": 0.39960572123527527, "learning_rate": 5.8333682212520334e-05, "loss": 0.0434, "step": 37950 }, { "grad_norm": 0.32377392053604126, "learning_rate": 5.831330223514486e-05, "loss": 0.0246, "step": 37960 }, { "grad_norm": 0.21919898688793182, "learning_rate": 5.8292920837253396e-05, "loss": 0.0213, "step": 37970 }, { "grad_norm": 0.34438201785087585, "learning_rate": 5.827253802232857e-05, "loss": 0.0216, "step": 37980 }, { "grad_norm": 0.49407830834388733, "learning_rate": 5.825215379385327e-05, "loss": 0.0225, "step": 37990 }, { "grad_norm": 0.21301983296871185, "learning_rate": 5.823176815531057e-05, "loss": 0.0343, "step": 38000 }, { "grad_norm": 0.22564643621444702, "learning_rate": 5.8211381110183826e-05, "loss": 0.0342, "step": 38010 }, { "grad_norm": 0.3049284517765045, "learning_rate": 5.8190992661956645e-05, "loss": 0.0232, "step": 38020 }, { "grad_norm": 0.3827219009399414, "learning_rate": 5.817060281411284e-05, "loss": 0.0271, "step": 38030 }, { "grad_norm": 0.4049873352050781, "learning_rate": 5.815021157013647e-05, "loss": 0.0236, "step": 38040 }, { "grad_norm": 0.3136849105358124, "learning_rate": 5.8129818933511856e-05, "loss": 0.0368, "step": 38050 }, { "grad_norm": 0.8695645928382874, "learning_rate": 5.8109424907723544e-05, "loss": 0.0306, "step": 38060 }, { "grad_norm": 0.23010142147541046, "learning_rate": 5.80890294962563e-05, "loss": 0.0187, "step": 38070 }, { "grad_norm": 0.35762572288513184, "learning_rate": 5.806863270259515e-05, "loss": 0.027, "step": 38080 }, { "grad_norm": 0.356060653924942, "learning_rate": 5.804823453022536e-05, "loss": 0.0253, "step": 38090 }, { "grad_norm": 0.31419554352760315, "learning_rate": 5.80278349826324e-05, "loss": 0.0216, "step": 38100 }, { "grad_norm": 0.23075959086418152, "learning_rate": 5.8007434063302014e-05, "loss": 0.026, "step": 38110 }, { "grad_norm": 0.31286922097206116, "learning_rate": 5.7987031775720136e-05, "loss": 0.0269, "step": 38120 }, { "grad_norm": 0.217466339468956, "learning_rate": 5.7966628123372976e-05, "loss": 0.0319, "step": 38130 }, { "grad_norm": 0.4375070631504059, "learning_rate": 5.7946223109746956e-05, "loss": 0.033, "step": 38140 }, { "grad_norm": 0.3387031853199005, "learning_rate": 5.7925816738328754e-05, "loss": 0.0173, "step": 38150 }, { "grad_norm": 0.46378105878829956, "learning_rate": 5.790540901260521e-05, "loss": 0.0256, "step": 38160 }, { "grad_norm": 0.38830864429473877, "learning_rate": 5.788499993606351e-05, "loss": 0.0215, "step": 38170 }, { "grad_norm": 0.47225794196128845, "learning_rate": 5.786458951219096e-05, "loss": 0.028, "step": 38180 }, { "grad_norm": 0.2548818588256836, "learning_rate": 5.784417774447517e-05, "loss": 0.0276, "step": 38190 }, { "grad_norm": 0.47285324335098267, "learning_rate": 5.782376463640393e-05, "loss": 0.0267, "step": 38200 }, { "grad_norm": 0.302734375, "learning_rate": 5.780335019146531e-05, "loss": 0.0202, "step": 38210 }, { "grad_norm": 0.3622147738933563, "learning_rate": 5.778293441314755e-05, "loss": 0.0273, "step": 38220 }, { "grad_norm": 0.34846702218055725, "learning_rate": 5.776251730493917e-05, "loss": 0.034, "step": 38230 }, { "grad_norm": 0.30997416377067566, "learning_rate": 5.774209887032887e-05, "loss": 0.0251, "step": 38240 }, { "grad_norm": 0.35962721705436707, "learning_rate": 5.772167911280565e-05, "loss": 0.0264, "step": 38250 }, { "grad_norm": 0.2858627736568451, "learning_rate": 5.770125803585864e-05, "loss": 0.021, "step": 38260 }, { "grad_norm": 0.3879155218601227, "learning_rate": 5.768083564297726e-05, "loss": 0.0279, "step": 38270 }, { "grad_norm": 0.43751898407936096, "learning_rate": 5.766041193765114e-05, "loss": 0.027, "step": 38280 }, { "grad_norm": 0.2607804834842682, "learning_rate": 5.763998692337015e-05, "loss": 0.0195, "step": 38290 }, { "grad_norm": 0.32124510407447815, "learning_rate": 5.761956060362433e-05, "loss": 0.027, "step": 38300 }, { "grad_norm": 0.2637123763561249, "learning_rate": 5.7599132981904e-05, "loss": 0.0251, "step": 38310 }, { "grad_norm": 0.26784929633140564, "learning_rate": 5.75787040616997e-05, "loss": 0.033, "step": 38320 }, { "grad_norm": 0.2870744466781616, "learning_rate": 5.755827384650212e-05, "loss": 0.0239, "step": 38330 }, { "grad_norm": 0.291047602891922, "learning_rate": 5.753784233980228e-05, "loss": 0.0223, "step": 38340 }, { "grad_norm": 0.30789467692375183, "learning_rate": 5.751740954509135e-05, "loss": 0.0231, "step": 38350 }, { "grad_norm": 0.45409974455833435, "learning_rate": 5.7496975465860715e-05, "loss": 0.02, "step": 38360 }, { "grad_norm": 0.20883576571941376, "learning_rate": 5.747654010560202e-05, "loss": 0.0181, "step": 38370 }, { "grad_norm": 0.5060750842094421, "learning_rate": 5.7456103467807097e-05, "loss": 0.0261, "step": 38380 }, { "grad_norm": 0.4167976677417755, "learning_rate": 5.7435665555968046e-05, "loss": 0.0227, "step": 38390 }, { "grad_norm": 0.21805492043495178, "learning_rate": 5.74152263735771e-05, "loss": 0.0293, "step": 38400 }, { "grad_norm": 0.3871384859085083, "learning_rate": 5.739478592412677e-05, "loss": 0.0276, "step": 38410 }, { "grad_norm": 0.5114097595214844, "learning_rate": 5.7374344211109766e-05, "loss": 0.0278, "step": 38420 }, { "grad_norm": 0.3132859170436859, "learning_rate": 5.735390123801904e-05, "loss": 0.0292, "step": 38430 }, { "grad_norm": 0.3424296975135803, "learning_rate": 5.7333457008347704e-05, "loss": 0.0347, "step": 38440 }, { "grad_norm": 0.3718424141407013, "learning_rate": 5.7313011525589156e-05, "loss": 0.0386, "step": 38450 }, { "grad_norm": 0.4062808156013489, "learning_rate": 5.729256479323694e-05, "loss": 0.0293, "step": 38460 }, { "grad_norm": 0.4551774859428406, "learning_rate": 5.727211681478485e-05, "loss": 0.0261, "step": 38470 }, { "grad_norm": 0.29390180110931396, "learning_rate": 5.7251667593726886e-05, "loss": 0.0297, "step": 38480 }, { "grad_norm": 0.3197035491466522, "learning_rate": 5.723121713355728e-05, "loss": 0.0371, "step": 38490 }, { "grad_norm": 0.3485197424888611, "learning_rate": 5.721076543777044e-05, "loss": 0.0239, "step": 38500 }, { "grad_norm": 0.277334988117218, "learning_rate": 5.7190312509860986e-05, "loss": 0.0244, "step": 38510 }, { "grad_norm": 1.1840559244155884, "learning_rate": 5.716985835332379e-05, "loss": 0.0286, "step": 38520 }, { "grad_norm": 0.2599651515483856, "learning_rate": 5.714940297165389e-05, "loss": 0.0294, "step": 38530 }, { "grad_norm": 0.38143301010131836, "learning_rate": 5.712894636834656e-05, "loss": 0.0261, "step": 38540 }, { "grad_norm": 0.3880692720413208, "learning_rate": 5.7108488546897276e-05, "loss": 0.0195, "step": 38550 }, { "grad_norm": 0.29781386256217957, "learning_rate": 5.708802951080172e-05, "loss": 0.0199, "step": 38560 }, { "grad_norm": 0.397193044424057, "learning_rate": 5.706756926355576e-05, "loss": 0.0323, "step": 38570 }, { "grad_norm": 0.28727924823760986, "learning_rate": 5.704710780865554e-05, "loss": 0.0224, "step": 38580 }, { "grad_norm": 0.4764528274536133, "learning_rate": 5.7026645149597325e-05, "loss": 0.028, "step": 38590 }, { "grad_norm": 0.27208393812179565, "learning_rate": 5.700618128987764e-05, "loss": 0.0246, "step": 38600 }, { "grad_norm": 0.25720351934432983, "learning_rate": 5.698571623299317e-05, "loss": 0.0228, "step": 38610 }, { "grad_norm": 0.3939795196056366, "learning_rate": 5.696524998244086e-05, "loss": 0.0196, "step": 38620 }, { "grad_norm": 0.37273144721984863, "learning_rate": 5.6944782541717836e-05, "loss": 0.0302, "step": 38630 }, { "grad_norm": 0.25449925661087036, "learning_rate": 5.69243139143214e-05, "loss": 0.0267, "step": 38640 }, { "grad_norm": 0.28229591250419617, "learning_rate": 5.6903844103749125e-05, "loss": 0.0251, "step": 38650 }, { "grad_norm": 0.20143263041973114, "learning_rate": 5.688337311349869e-05, "loss": 0.0226, "step": 38660 }, { "grad_norm": 0.3141612708568573, "learning_rate": 5.6862900947068074e-05, "loss": 0.0237, "step": 38670 }, { "grad_norm": 0.36683088541030884, "learning_rate": 5.6842427607955374e-05, "loss": 0.0302, "step": 38680 }, { "grad_norm": 0.30024453997612, "learning_rate": 5.682195309965893e-05, "loss": 0.021, "step": 38690 }, { "grad_norm": 0.40508896112442017, "learning_rate": 5.6801477425677294e-05, "loss": 0.0306, "step": 38700 }, { "grad_norm": 0.4819195568561554, "learning_rate": 5.678100058950917e-05, "loss": 0.0318, "step": 38710 }, { "grad_norm": 0.4287523329257965, "learning_rate": 5.676052259465352e-05, "loss": 0.0354, "step": 38720 }, { "grad_norm": 0.3781270980834961, "learning_rate": 5.674004344460945e-05, "loss": 0.0245, "step": 38730 }, { "grad_norm": 0.3974356949329376, "learning_rate": 5.6719563142876295e-05, "loss": 0.0247, "step": 38740 }, { "grad_norm": 0.3004567325115204, "learning_rate": 5.669908169295359e-05, "loss": 0.0217, "step": 38750 }, { "grad_norm": 0.28117236495018005, "learning_rate": 5.667859909834105e-05, "loss": 0.0266, "step": 38760 }, { "grad_norm": 0.24265754222869873, "learning_rate": 5.6658115362538565e-05, "loss": 0.0255, "step": 38770 }, { "grad_norm": 0.3183826506137848, "learning_rate": 5.663763048904628e-05, "loss": 0.0237, "step": 38780 }, { "grad_norm": 0.36322465538978577, "learning_rate": 5.661714448136447e-05, "loss": 0.0192, "step": 38790 }, { "grad_norm": 0.2705110013484955, "learning_rate": 5.659665734299366e-05, "loss": 0.0181, "step": 38800 }, { "grad_norm": 0.26436203718185425, "learning_rate": 5.6576169077434516e-05, "loss": 0.0259, "step": 38810 }, { "grad_norm": 0.3886537551879883, "learning_rate": 5.6555679688187944e-05, "loss": 0.0223, "step": 38820 }, { "grad_norm": 0.37682032585144043, "learning_rate": 5.6535189178755e-05, "loss": 0.0215, "step": 38830 }, { "grad_norm": 0.32751232385635376, "learning_rate": 5.651469755263698e-05, "loss": 0.0266, "step": 38840 }, { "grad_norm": 0.157822847366333, "learning_rate": 5.6494204813335316e-05, "loss": 0.0344, "step": 38850 }, { "grad_norm": 0.20422230660915375, "learning_rate": 5.647371096435168e-05, "loss": 0.0195, "step": 38860 }, { "grad_norm": 0.3529871702194214, "learning_rate": 5.645321600918788e-05, "loss": 0.0235, "step": 38870 }, { "grad_norm": 0.3313123285770416, "learning_rate": 5.643271995134597e-05, "loss": 0.025, "step": 38880 }, { "grad_norm": 0.30209407210350037, "learning_rate": 5.641222279432814e-05, "loss": 0.0243, "step": 38890 }, { "grad_norm": 0.47403812408447266, "learning_rate": 5.6391724541636834e-05, "loss": 0.023, "step": 38900 }, { "grad_norm": 0.3378881514072418, "learning_rate": 5.6371225196774605e-05, "loss": 0.0255, "step": 38910 }, { "grad_norm": 0.28426089882850647, "learning_rate": 5.635072476324423e-05, "loss": 0.027, "step": 38920 }, { "grad_norm": 0.3539223372936249, "learning_rate": 5.63302232445487e-05, "loss": 0.0192, "step": 38930 }, { "grad_norm": 0.21663597226142883, "learning_rate": 5.6309720644191144e-05, "loss": 0.0288, "step": 38940 }, { "grad_norm": 0.4928050935268402, "learning_rate": 5.628921696567491e-05, "loss": 0.0223, "step": 38950 }, { "grad_norm": 0.3760104775428772, "learning_rate": 5.62687122125035e-05, "loss": 0.0241, "step": 38960 }, { "grad_norm": 0.3705645203590393, "learning_rate": 5.624820638818062e-05, "loss": 0.0214, "step": 38970 }, { "grad_norm": 0.46468740701675415, "learning_rate": 5.6227699496210164e-05, "loss": 0.0368, "step": 38980 }, { "grad_norm": 0.22319476306438446, "learning_rate": 5.6207191540096195e-05, "loss": 0.0245, "step": 38990 }, { "grad_norm": 0.3724420368671417, "learning_rate": 5.618668252334296e-05, "loss": 0.0186, "step": 39000 }, { "grad_norm": 0.2702041566371918, "learning_rate": 5.616617244945488e-05, "loss": 0.0198, "step": 39010 }, { "grad_norm": 0.2849270701408386, "learning_rate": 5.614566132193656e-05, "loss": 0.0218, "step": 39020 }, { "grad_norm": 0.273060142993927, "learning_rate": 5.612514914429282e-05, "loss": 0.024, "step": 39030 }, { "grad_norm": 0.2745307385921478, "learning_rate": 5.610463592002863e-05, "loss": 0.0202, "step": 39040 }, { "grad_norm": 0.3348398208618164, "learning_rate": 5.608412165264909e-05, "loss": 0.0229, "step": 39050 }, { "grad_norm": 0.39640456438064575, "learning_rate": 5.606360634565959e-05, "loss": 0.021, "step": 39060 }, { "grad_norm": 0.1945483386516571, "learning_rate": 5.604309000256559e-05, "loss": 0.0242, "step": 39070 }, { "grad_norm": 0.40207722783088684, "learning_rate": 5.6022572626872785e-05, "loss": 0.0215, "step": 39080 }, { "grad_norm": 0.39556685090065, "learning_rate": 5.600205422208704e-05, "loss": 0.0268, "step": 39090 }, { "grad_norm": 0.25500819087028503, "learning_rate": 5.59815347917144e-05, "loss": 0.0268, "step": 39100 }, { "grad_norm": 0.30839428305625916, "learning_rate": 5.596101433926103e-05, "loss": 0.0296, "step": 39110 }, { "grad_norm": 0.2911731004714966, "learning_rate": 5.5940492868233364e-05, "loss": 0.0206, "step": 39120 }, { "grad_norm": 0.38085874915122986, "learning_rate": 5.591997038213793e-05, "loss": 0.0307, "step": 39130 }, { "grad_norm": 0.22275680303573608, "learning_rate": 5.5899446884481475e-05, "loss": 0.019, "step": 39140 }, { "grad_norm": 0.36944058537483215, "learning_rate": 5.5878922378770906e-05, "loss": 0.0316, "step": 39150 }, { "grad_norm": 0.24314865469932556, "learning_rate": 5.5858396868513285e-05, "loss": 0.0233, "step": 39160 }, { "grad_norm": 0.29096341133117676, "learning_rate": 5.583787035721586e-05, "loss": 0.0305, "step": 39170 }, { "grad_norm": 0.41336581110954285, "learning_rate": 5.581734284838606e-05, "loss": 0.0253, "step": 39180 }, { "grad_norm": 0.3346363306045532, "learning_rate": 5.579681434553147e-05, "loss": 0.0234, "step": 39190 }, { "grad_norm": 0.2847994863986969, "learning_rate": 5.5776284852159854e-05, "loss": 0.0216, "step": 39200 }, { "grad_norm": 0.3576805293560028, "learning_rate": 5.575575437177913e-05, "loss": 0.0306, "step": 39210 }, { "grad_norm": 0.3896729350090027, "learning_rate": 5.573522290789742e-05, "loss": 0.0212, "step": 39220 }, { "grad_norm": 0.2433193325996399, "learning_rate": 5.571469046402298e-05, "loss": 0.0258, "step": 39230 }, { "grad_norm": 0.2963736355304718, "learning_rate": 5.5694157043664205e-05, "loss": 0.0254, "step": 39240 }, { "grad_norm": 0.26798948645591736, "learning_rate": 5.567362265032975e-05, "loss": 0.0171, "step": 39250 }, { "grad_norm": 0.2996116578578949, "learning_rate": 5.565308728752836e-05, "loss": 0.0238, "step": 39260 }, { "grad_norm": 0.28775328397750854, "learning_rate": 5.5632550958768956e-05, "loss": 0.0182, "step": 39270 }, { "grad_norm": 0.3633127212524414, "learning_rate": 5.5612013667560636e-05, "loss": 0.0253, "step": 39280 }, { "grad_norm": 0.35850653052330017, "learning_rate": 5.5591475417412676e-05, "loss": 0.0377, "step": 39290 }, { "grad_norm": 0.338927686214447, "learning_rate": 5.557093621183451e-05, "loss": 0.0299, "step": 39300 }, { "grad_norm": 0.2851278483867645, "learning_rate": 5.55503960543357e-05, "loss": 0.0239, "step": 39310 }, { "grad_norm": 0.525019645690918, "learning_rate": 5.552985494842601e-05, "loss": 0.0287, "step": 39320 }, { "grad_norm": 0.20025838911533356, "learning_rate": 5.550931289761534e-05, "loss": 0.0202, "step": 39330 }, { "grad_norm": 0.23040388524532318, "learning_rate": 5.548876990541378e-05, "loss": 0.0227, "step": 39340 }, { "grad_norm": 0.44706249237060547, "learning_rate": 5.5468225975331564e-05, "loss": 0.0319, "step": 39350 }, { "grad_norm": 0.38284286856651306, "learning_rate": 5.544768111087909e-05, "loss": 0.0232, "step": 39360 }, { "grad_norm": 0.2569766342639923, "learning_rate": 5.54271353155669e-05, "loss": 0.0228, "step": 39370 }, { "grad_norm": 0.2991831600666046, "learning_rate": 5.5406588592905715e-05, "loss": 0.0236, "step": 39380 }, { "grad_norm": 0.22356566786766052, "learning_rate": 5.5386040946406416e-05, "loss": 0.0276, "step": 39390 }, { "grad_norm": 0.6999120712280273, "learning_rate": 5.536549237958004e-05, "loss": 0.0314, "step": 39400 }, { "grad_norm": 0.34885433316230774, "learning_rate": 5.5344942895937744e-05, "loss": 0.0386, "step": 39410 }, { "grad_norm": 0.24612736701965332, "learning_rate": 5.5324392498990904e-05, "loss": 0.0259, "step": 39420 }, { "grad_norm": 0.35840362310409546, "learning_rate": 5.5303841192251e-05, "loss": 0.0265, "step": 39430 }, { "grad_norm": 0.3284221291542053, "learning_rate": 5.52832889792297e-05, "loss": 0.0287, "step": 39440 }, { "grad_norm": 0.3228456676006317, "learning_rate": 5.526273586343881e-05, "loss": 0.0218, "step": 39450 }, { "grad_norm": 0.2611158788204193, "learning_rate": 5.5242181848390306e-05, "loss": 0.022, "step": 39460 }, { "grad_norm": 0.29872554540634155, "learning_rate": 5.5221626937596285e-05, "loss": 0.0222, "step": 39470 }, { "grad_norm": 0.2451634705066681, "learning_rate": 5.520107113456903e-05, "loss": 0.0326, "step": 39480 }, { "grad_norm": 0.40563470125198364, "learning_rate": 5.5180514442820974e-05, "loss": 0.0352, "step": 39490 }, { "grad_norm": 0.2504143714904785, "learning_rate": 5.515995686586469e-05, "loss": 0.0259, "step": 39500 }, { "grad_norm": 0.34490761160850525, "learning_rate": 5.5139398407212916e-05, "loss": 0.029, "step": 39510 }, { "grad_norm": 0.3580920398235321, "learning_rate": 5.511883907037849e-05, "loss": 0.0283, "step": 39520 }, { "grad_norm": 0.4494149386882782, "learning_rate": 5.509827885887449e-05, "loss": 0.0275, "step": 39530 }, { "grad_norm": 0.37754344940185547, "learning_rate": 5.507771777621406e-05, "loss": 0.0313, "step": 39540 }, { "grad_norm": 0.3108735680580139, "learning_rate": 5.505715582591052e-05, "loss": 0.0387, "step": 39550 }, { "grad_norm": 0.3761553466320038, "learning_rate": 5.50365930114774e-05, "loss": 0.0286, "step": 39560 }, { "grad_norm": 0.21781642735004425, "learning_rate": 5.5016029336428255e-05, "loss": 0.0284, "step": 39570 }, { "grad_norm": 0.5214876532554626, "learning_rate": 5.49954648042769e-05, "loss": 0.0256, "step": 39580 }, { "grad_norm": 0.24167674779891968, "learning_rate": 5.4974899418537226e-05, "loss": 0.0294, "step": 39590 }, { "grad_norm": 0.3260767161846161, "learning_rate": 5.4954333182723316e-05, "loss": 0.023, "step": 39600 }, { "grad_norm": 0.4112304747104645, "learning_rate": 5.493376610034937e-05, "loss": 0.0219, "step": 39610 }, { "grad_norm": 0.3463881313800812, "learning_rate": 5.4913198174929735e-05, "loss": 0.0259, "step": 39620 }, { "grad_norm": 0.19390930235385895, "learning_rate": 5.489262940997889e-05, "loss": 0.0278, "step": 39630 }, { "grad_norm": 0.2770695686340332, "learning_rate": 5.487205980901151e-05, "loss": 0.0295, "step": 39640 }, { "grad_norm": 0.3278077244758606, "learning_rate": 5.485148937554234e-05, "loss": 0.0434, "step": 39650 }, { "grad_norm": 0.24092596769332886, "learning_rate": 5.483091811308635e-05, "loss": 0.0196, "step": 39660 }, { "grad_norm": 0.3787707984447479, "learning_rate": 5.4810346025158564e-05, "loss": 0.0283, "step": 39670 }, { "grad_norm": 0.7465793490409851, "learning_rate": 5.478977311527421e-05, "loss": 0.0261, "step": 39680 }, { "grad_norm": 0.2597190737724304, "learning_rate": 5.476919938694863e-05, "loss": 0.0226, "step": 39690 }, { "grad_norm": 0.3197837769985199, "learning_rate": 5.474862484369733e-05, "loss": 0.0186, "step": 39700 }, { "grad_norm": 0.27872318029403687, "learning_rate": 5.472804948903589e-05, "loss": 0.022, "step": 39710 }, { "grad_norm": 0.27386927604675293, "learning_rate": 5.470747332648013e-05, "loss": 0.0163, "step": 39720 }, { "grad_norm": 0.4363768398761749, "learning_rate": 5.468689635954591e-05, "loss": 0.0368, "step": 39730 }, { "grad_norm": 0.23649820685386658, "learning_rate": 5.46663185917493e-05, "loss": 0.023, "step": 39740 }, { "grad_norm": 0.47208744287490845, "learning_rate": 5.464574002660645e-05, "loss": 0.0266, "step": 39750 }, { "grad_norm": 0.2506071627140045, "learning_rate": 5.4625160667633724e-05, "loss": 0.024, "step": 39760 }, { "grad_norm": 0.24382585287094116, "learning_rate": 5.4604580518347505e-05, "loss": 0.0275, "step": 39770 }, { "grad_norm": 0.3157714605331421, "learning_rate": 5.458399958226442e-05, "loss": 0.0281, "step": 39780 }, { "grad_norm": 0.7505044937133789, "learning_rate": 5.456341786290119e-05, "loss": 0.0268, "step": 39790 }, { "grad_norm": 0.492876261472702, "learning_rate": 5.454283536377465e-05, "loss": 0.0278, "step": 39800 }, { "grad_norm": 0.25613531470298767, "learning_rate": 5.452225208840179e-05, "loss": 0.0232, "step": 39810 }, { "grad_norm": 0.35727590322494507, "learning_rate": 5.450166804029975e-05, "loss": 0.0274, "step": 39820 }, { "grad_norm": 0.3363818824291229, "learning_rate": 5.448108322298574e-05, "loss": 0.0184, "step": 39830 }, { "grad_norm": 0.24885547161102295, "learning_rate": 5.446049763997717e-05, "loss": 0.0225, "step": 39840 }, { "grad_norm": 0.306081622838974, "learning_rate": 5.4439911294791546e-05, "loss": 0.0209, "step": 39850 }, { "grad_norm": 0.2889585494995117, "learning_rate": 5.441932419094652e-05, "loss": 0.02, "step": 39860 }, { "grad_norm": 0.3203999996185303, "learning_rate": 5.439873633195985e-05, "loss": 0.0256, "step": 39870 }, { "grad_norm": 0.23086582124233246, "learning_rate": 5.437814772134947e-05, "loss": 0.0291, "step": 39880 }, { "grad_norm": 0.25795507431030273, "learning_rate": 5.4357558362633366e-05, "loss": 0.0217, "step": 39890 }, { "grad_norm": 0.2461138665676117, "learning_rate": 5.4336968259329726e-05, "loss": 0.0154, "step": 39900 }, { "grad_norm": 0.23243895173072815, "learning_rate": 5.431637741495681e-05, "loss": 0.0226, "step": 39910 }, { "grad_norm": 0.2738039195537567, "learning_rate": 5.429578583303307e-05, "loss": 0.0272, "step": 39920 }, { "grad_norm": 0.39439868927001953, "learning_rate": 5.427519351707701e-05, "loss": 0.0258, "step": 39930 }, { "grad_norm": 0.3003493845462799, "learning_rate": 5.42546004706073e-05, "loss": 0.0237, "step": 39940 }, { "grad_norm": 0.3427765369415283, "learning_rate": 5.4234006697142735e-05, "loss": 0.0193, "step": 39950 }, { "grad_norm": 0.24473154544830322, "learning_rate": 5.421341220020224e-05, "loss": 0.025, "step": 39960 }, { "grad_norm": 0.2774721682071686, "learning_rate": 5.419281698330482e-05, "loss": 0.0272, "step": 39970 }, { "grad_norm": 0.37738072872161865, "learning_rate": 5.4172221049969665e-05, "loss": 0.027, "step": 39980 }, { "grad_norm": 0.23919788002967834, "learning_rate": 5.415162440371604e-05, "loss": 0.0198, "step": 39990 }, { "grad_norm": 0.611315131187439, "learning_rate": 5.413102704806334e-05, "loss": 0.0243, "step": 40000 }, { "grad_norm": 0.6045412421226501, "learning_rate": 5.41104289865311e-05, "loss": 0.0288, "step": 40010 }, { "grad_norm": 0.25267767906188965, "learning_rate": 5.408983022263898e-05, "loss": 0.0214, "step": 40020 }, { "grad_norm": 0.7023836970329285, "learning_rate": 5.406923075990673e-05, "loss": 0.0319, "step": 40030 }, { "grad_norm": 0.6157547831535339, "learning_rate": 5.404863060185423e-05, "loss": 0.024, "step": 40040 }, { "grad_norm": 0.39725103974342346, "learning_rate": 5.40280297520015e-05, "loss": 0.0275, "step": 40050 }, { "grad_norm": 0.256365031003952, "learning_rate": 5.400742821386865e-05, "loss": 0.0232, "step": 40060 }, { "grad_norm": 0.305960476398468, "learning_rate": 5.398682599097592e-05, "loss": 0.0258, "step": 40070 }, { "grad_norm": 0.2742340564727783, "learning_rate": 5.396622308684367e-05, "loss": 0.0167, "step": 40080 }, { "grad_norm": 0.4450824558734894, "learning_rate": 5.394561950499236e-05, "loss": 0.027, "step": 40090 }, { "grad_norm": 0.2096996307373047, "learning_rate": 5.39250152489426e-05, "loss": 0.0245, "step": 40100 }, { "grad_norm": 0.3437958359718323, "learning_rate": 5.390441032221507e-05, "loss": 0.0212, "step": 40110 }, { "grad_norm": 0.27832096815109253, "learning_rate": 5.388380472833062e-05, "loss": 0.0237, "step": 40120 }, { "grad_norm": 0.4791315197944641, "learning_rate": 5.386319847081016e-05, "loss": 0.0292, "step": 40130 }, { "grad_norm": 0.3804197311401367, "learning_rate": 5.384259155317473e-05, "loss": 0.0248, "step": 40140 }, { "grad_norm": 0.42050111293792725, "learning_rate": 5.38219839789455e-05, "loss": 0.0263, "step": 40150 }, { "grad_norm": 0.2216203361749649, "learning_rate": 5.380137575164376e-05, "loss": 0.0318, "step": 40160 }, { "grad_norm": 0.2347855120897293, "learning_rate": 5.378076687479085e-05, "loss": 0.0167, "step": 40170 }, { "grad_norm": 0.19567811489105225, "learning_rate": 5.3760157351908305e-05, "loss": 0.0239, "step": 40180 }, { "grad_norm": 0.17971712350845337, "learning_rate": 5.373954718651768e-05, "loss": 0.0196, "step": 40190 }, { "grad_norm": 0.19132834672927856, "learning_rate": 5.371893638214074e-05, "loss": 0.0186, "step": 40200 }, { "grad_norm": 0.29609215259552, "learning_rate": 5.369832494229927e-05, "loss": 0.0195, "step": 40210 }, { "grad_norm": 0.4007367491722107, "learning_rate": 5.367771287051524e-05, "loss": 0.0258, "step": 40220 }, { "grad_norm": 0.28760844469070435, "learning_rate": 5.365710017031065e-05, "loss": 0.0259, "step": 40230 }, { "grad_norm": 0.24656660854816437, "learning_rate": 5.3636486845207654e-05, "loss": 0.0208, "step": 40240 }, { "grad_norm": 0.25499287247657776, "learning_rate": 5.3615872898728526e-05, "loss": 0.0185, "step": 40250 }, { "grad_norm": 0.5075778365135193, "learning_rate": 5.3595258334395614e-05, "loss": 0.027, "step": 40260 }, { "grad_norm": 0.3152239918708801, "learning_rate": 5.3574643155731396e-05, "loss": 0.0286, "step": 40270 }, { "grad_norm": 0.4651948809623718, "learning_rate": 5.3554027366258405e-05, "loss": 0.0241, "step": 40280 }, { "grad_norm": 0.29603111743927, "learning_rate": 5.3533410969499355e-05, "loss": 0.0248, "step": 40290 }, { "grad_norm": 0.23817525804042816, "learning_rate": 5.3512793968977e-05, "loss": 0.0224, "step": 40300 }, { "grad_norm": 0.3736793100833893, "learning_rate": 5.349217636821423e-05, "loss": 0.0236, "step": 40310 }, { "grad_norm": 0.23119325935840607, "learning_rate": 5.347155817073405e-05, "loss": 0.0215, "step": 40320 }, { "grad_norm": 0.279739111661911, "learning_rate": 5.34509393800595e-05, "loss": 0.0246, "step": 40330 }, { "grad_norm": 0.27127230167388916, "learning_rate": 5.343031999971381e-05, "loss": 0.0241, "step": 40340 }, { "grad_norm": 0.3879010081291199, "learning_rate": 5.340970003322026e-05, "loss": 0.0213, "step": 40350 }, { "grad_norm": 0.5204145908355713, "learning_rate": 5.33890794841022e-05, "loss": 0.0223, "step": 40360 }, { "grad_norm": 0.26287224888801575, "learning_rate": 5.336845835588318e-05, "loss": 0.023, "step": 40370 }, { "grad_norm": 0.22661566734313965, "learning_rate": 5.334783665208674e-05, "loss": 0.0234, "step": 40380 }, { "grad_norm": 0.30805644392967224, "learning_rate": 5.332721437623657e-05, "loss": 0.0177, "step": 40390 }, { "grad_norm": 0.53074049949646, "learning_rate": 5.3306591531856464e-05, "loss": 0.03, "step": 40400 }, { "grad_norm": 0.32777947187423706, "learning_rate": 5.3285968122470295e-05, "loss": 0.0223, "step": 40410 }, { "grad_norm": 0.33320528268814087, "learning_rate": 5.326534415160207e-05, "loss": 0.0293, "step": 40420 }, { "grad_norm": 0.33434373140335083, "learning_rate": 5.324471962277582e-05, "loss": 0.0331, "step": 40430 }, { "grad_norm": 0.25568076968193054, "learning_rate": 5.3224094539515746e-05, "loss": 0.0202, "step": 40440 }, { "grad_norm": 0.32099640369415283, "learning_rate": 5.3203468905346076e-05, "loss": 0.0244, "step": 40450 }, { "grad_norm": 0.32071205973625183, "learning_rate": 5.3182842723791195e-05, "loss": 0.0285, "step": 40460 }, { "grad_norm": 0.46486788988113403, "learning_rate": 5.316221599837554e-05, "loss": 0.021, "step": 40470 }, { "grad_norm": 0.31742486357688904, "learning_rate": 5.314158873262366e-05, "loss": 0.0278, "step": 40480 }, { "grad_norm": 0.24966198205947876, "learning_rate": 5.312096093006018e-05, "loss": 0.018, "step": 40490 }, { "grad_norm": 0.5001295208930969, "learning_rate": 5.3100332594209846e-05, "loss": 0.0221, "step": 40500 }, { "grad_norm": 0.8316570520401001, "learning_rate": 5.3079703728597454e-05, "loss": 0.0276, "step": 40510 }, { "grad_norm": 0.25497034192085266, "learning_rate": 5.305907433674794e-05, "loss": 0.0211, "step": 40520 }, { "grad_norm": 0.3764949440956116, "learning_rate": 5.3038444422186284e-05, "loss": 0.0243, "step": 40530 }, { "grad_norm": 0.3441605567932129, "learning_rate": 5.301781398843757e-05, "loss": 0.027, "step": 40540 }, { "grad_norm": 0.7791882753372192, "learning_rate": 5.299718303902699e-05, "loss": 0.0253, "step": 40550 }, { "grad_norm": 0.2613503038883209, "learning_rate": 5.297655157747979e-05, "loss": 0.0215, "step": 40560 }, { "grad_norm": 0.3667272925376892, "learning_rate": 5.295591960732136e-05, "loss": 0.0247, "step": 40570 }, { "grad_norm": 0.35923653841018677, "learning_rate": 5.293528713207708e-05, "loss": 0.0301, "step": 40580 }, { "grad_norm": 0.20979170501232147, "learning_rate": 5.291465415527253e-05, "loss": 0.03, "step": 40590 }, { "grad_norm": 0.18563663959503174, "learning_rate": 5.2894020680433296e-05, "loss": 0.0217, "step": 40600 }, { "grad_norm": 0.3194112777709961, "learning_rate": 5.287338671108507e-05, "loss": 0.0307, "step": 40610 }, { "grad_norm": 0.7248928546905518, "learning_rate": 5.285275225075367e-05, "loss": 0.0253, "step": 40620 }, { "grad_norm": 0.4168004095554352, "learning_rate": 5.283211730296492e-05, "loss": 0.0243, "step": 40630 }, { "grad_norm": 0.33414870500564575, "learning_rate": 5.281148187124477e-05, "loss": 0.0247, "step": 40640 }, { "grad_norm": 0.8422065377235413, "learning_rate": 5.279084595911927e-05, "loss": 0.0266, "step": 40650 }, { "grad_norm": 0.3534501791000366, "learning_rate": 5.277020957011453e-05, "loss": 0.0331, "step": 40660 }, { "grad_norm": 0.4898775815963745, "learning_rate": 5.274957270775673e-05, "loss": 0.0309, "step": 40670 }, { "grad_norm": 0.2570348083972931, "learning_rate": 5.2728935375572164e-05, "loss": 0.0208, "step": 40680 }, { "grad_norm": 0.1986812949180603, "learning_rate": 5.2708297577087176e-05, "loss": 0.0204, "step": 40690 }, { "grad_norm": 0.3486802875995636, "learning_rate": 5.2687659315828196e-05, "loss": 0.0202, "step": 40700 }, { "grad_norm": 0.2732609808444977, "learning_rate": 5.266702059532175e-05, "loss": 0.0206, "step": 40710 }, { "grad_norm": 0.3376827836036682, "learning_rate": 5.264638141909444e-05, "loss": 0.0253, "step": 40720 }, { "grad_norm": 0.42256200313568115, "learning_rate": 5.2625741790672925e-05, "loss": 0.0227, "step": 40730 }, { "grad_norm": 0.22066345810890198, "learning_rate": 5.2605101713583935e-05, "loss": 0.0205, "step": 40740 }, { "grad_norm": 0.34701383113861084, "learning_rate": 5.2584461191354315e-05, "loss": 0.0277, "step": 40750 }, { "grad_norm": 0.2886185348033905, "learning_rate": 5.2563820227510964e-05, "loss": 0.0226, "step": 40760 }, { "grad_norm": 0.24187813699245453, "learning_rate": 5.2543178825580864e-05, "loss": 0.0233, "step": 40770 }, { "grad_norm": 0.34559395909309387, "learning_rate": 5.2522536989091054e-05, "loss": 0.0242, "step": 40780 }, { "grad_norm": 0.34771034121513367, "learning_rate": 5.2501894721568655e-05, "loss": 0.0224, "step": 40790 }, { "grad_norm": 0.3707639276981354, "learning_rate": 5.248125202654089e-05, "loss": 0.0238, "step": 40800 }, { "grad_norm": 0.41705572605133057, "learning_rate": 5.246060890753501e-05, "loss": 0.0371, "step": 40810 }, { "grad_norm": 0.21992383897304535, "learning_rate": 5.243996536807837e-05, "loss": 0.0273, "step": 40820 }, { "grad_norm": 0.2909299433231354, "learning_rate": 5.2419321411698384e-05, "loss": 0.019, "step": 40830 }, { "grad_norm": 0.2606687545776367, "learning_rate": 5.239867704192253e-05, "loss": 0.026, "step": 40840 }, { "grad_norm": 0.2043449431657791, "learning_rate": 5.237803226227838e-05, "loss": 0.0217, "step": 40850 }, { "grad_norm": 0.3698985278606415, "learning_rate": 5.235738707629354e-05, "loss": 0.0249, "step": 40860 }, { "grad_norm": 0.34530869126319885, "learning_rate": 5.233674148749575e-05, "loss": 0.028, "step": 40870 }, { "grad_norm": 0.4372304677963257, "learning_rate": 5.231609549941272e-05, "loss": 0.0282, "step": 40880 }, { "grad_norm": 0.23405303061008453, "learning_rate": 5.229544911557233e-05, "loss": 0.0289, "step": 40890 }, { "grad_norm": 0.171901136636734, "learning_rate": 5.227480233950246e-05, "loss": 0.0265, "step": 40900 }, { "grad_norm": 0.2959231734275818, "learning_rate": 5.22541551747311e-05, "loss": 0.0266, "step": 40910 }, { "grad_norm": 0.6619501709938049, "learning_rate": 5.223350762478626e-05, "loss": 0.0242, "step": 40920 }, { "grad_norm": 0.2714272439479828, "learning_rate": 5.221285969319606e-05, "loss": 0.0204, "step": 40930 }, { "grad_norm": 0.3624473214149475, "learning_rate": 5.219221138348865e-05, "loss": 0.0237, "step": 40940 }, { "grad_norm": 0.3198573589324951, "learning_rate": 5.217156269919228e-05, "loss": 0.0235, "step": 40950 }, { "grad_norm": 0.3616117537021637, "learning_rate": 5.215091364383523e-05, "loss": 0.0193, "step": 40960 }, { "grad_norm": 0.3859403729438782, "learning_rate": 5.213026422094588e-05, "loss": 0.0218, "step": 40970 }, { "grad_norm": 0.199065163731575, "learning_rate": 5.210961443405262e-05, "loss": 0.0194, "step": 40980 }, { "grad_norm": 0.2915172278881073, "learning_rate": 5.208896428668396e-05, "loss": 0.017, "step": 40990 }, { "grad_norm": 0.30915263295173645, "learning_rate": 5.206831378236845e-05, "loss": 0.0238, "step": 41000 }, { "grad_norm": 0.2894975543022156, "learning_rate": 5.2047662924634666e-05, "loss": 0.0182, "step": 41010 }, { "grad_norm": 0.28246009349823, "learning_rate": 5.20270117170113e-05, "loss": 0.0183, "step": 41020 }, { "grad_norm": 0.3396519422531128, "learning_rate": 5.200636016302707e-05, "loss": 0.0283, "step": 41030 }, { "grad_norm": 0.25813576579093933, "learning_rate": 5.1985708266210754e-05, "loss": 0.0265, "step": 41040 }, { "grad_norm": 0.211778461933136, "learning_rate": 5.1965056030091206e-05, "loss": 0.0224, "step": 41050 }, { "grad_norm": 0.23425757884979248, "learning_rate": 5.194440345819732e-05, "loss": 0.0209, "step": 41060 }, { "grad_norm": 0.2879831790924072, "learning_rate": 5.1923750554058084e-05, "loss": 0.0235, "step": 41070 }, { "grad_norm": 0.2662571966648102, "learning_rate": 5.1903097321202476e-05, "loss": 0.0208, "step": 41080 }, { "grad_norm": 0.41843611001968384, "learning_rate": 5.18824437631596e-05, "loss": 0.0325, "step": 41090 }, { "grad_norm": 0.269462913274765, "learning_rate": 5.186178988345856e-05, "loss": 0.024, "step": 41100 }, { "grad_norm": 0.4532000720500946, "learning_rate": 5.184113568562855e-05, "loss": 0.0241, "step": 41110 }, { "grad_norm": 0.27315494418144226, "learning_rate": 5.18204811731988e-05, "loss": 0.0268, "step": 41120 }, { "grad_norm": 0.38049450516700745, "learning_rate": 5.179982634969862e-05, "loss": 0.0224, "step": 41130 }, { "grad_norm": 0.42014428973197937, "learning_rate": 5.177917121865734e-05, "loss": 0.0315, "step": 41140 }, { "grad_norm": 0.3094930648803711, "learning_rate": 5.1758515783604346e-05, "loss": 0.0207, "step": 41150 }, { "grad_norm": 0.3953510820865631, "learning_rate": 5.173786004806911e-05, "loss": 0.0279, "step": 41160 }, { "grad_norm": 0.5635406374931335, "learning_rate": 5.1717204015581135e-05, "loss": 0.0254, "step": 41170 }, { "grad_norm": 0.3505273163318634, "learning_rate": 5.169654768966994e-05, "loss": 0.0213, "step": 41180 }, { "grad_norm": 0.4955894947052002, "learning_rate": 5.1675891073865156e-05, "loss": 0.0269, "step": 41190 }, { "grad_norm": 0.1996188908815384, "learning_rate": 5.1655234171696424e-05, "loss": 0.0163, "step": 41200 }, { "grad_norm": 0.23489812016487122, "learning_rate": 5.163457698669343e-05, "loss": 0.0183, "step": 41210 }, { "grad_norm": 0.27874132990837097, "learning_rate": 5.1613919522385936e-05, "loss": 0.0184, "step": 41220 }, { "grad_norm": 0.47734367847442627, "learning_rate": 5.1593261782303746e-05, "loss": 0.0284, "step": 41230 }, { "grad_norm": 0.41858235001564026, "learning_rate": 5.157260376997669e-05, "loss": 0.0216, "step": 41240 }, { "grad_norm": 0.35792824625968933, "learning_rate": 5.155194548893466e-05, "loss": 0.0252, "step": 41250 }, { "grad_norm": 0.2857693135738373, "learning_rate": 5.153128694270759e-05, "loss": 0.0301, "step": 41260 }, { "grad_norm": 0.4654758870601654, "learning_rate": 5.151062813482548e-05, "loss": 0.032, "step": 41270 }, { "grad_norm": 0.30819523334503174, "learning_rate": 5.148996906881832e-05, "loss": 0.0183, "step": 41280 }, { "grad_norm": 0.3504035174846649, "learning_rate": 5.1469309748216196e-05, "loss": 0.0236, "step": 41290 }, { "grad_norm": 0.21364249289035797, "learning_rate": 5.144865017654923e-05, "loss": 0.0174, "step": 41300 }, { "grad_norm": 0.3020979166030884, "learning_rate": 5.1427990357347556e-05, "loss": 0.0225, "step": 41310 }, { "grad_norm": 0.17767074704170227, "learning_rate": 5.140733029414139e-05, "loss": 0.0236, "step": 41320 }, { "grad_norm": 0.36539000272750854, "learning_rate": 5.138666999046098e-05, "loss": 0.0236, "step": 41330 }, { "grad_norm": 0.6671412587165833, "learning_rate": 5.136600944983658e-05, "loss": 0.0235, "step": 41340 }, { "grad_norm": 0.2937222123146057, "learning_rate": 5.134534867579853e-05, "loss": 0.0216, "step": 41350 }, { "grad_norm": 0.3977045714855194, "learning_rate": 5.1324687671877184e-05, "loss": 0.0202, "step": 41360 }, { "grad_norm": 0.29419758915901184, "learning_rate": 5.130402644160296e-05, "loss": 0.0271, "step": 41370 }, { "grad_norm": 0.22294485569000244, "learning_rate": 5.128336498850628e-05, "loss": 0.0181, "step": 41380 }, { "grad_norm": 0.38548722863197327, "learning_rate": 5.126270331611761e-05, "loss": 0.0228, "step": 41390 }, { "grad_norm": 0.3577396273612976, "learning_rate": 5.124204142796748e-05, "loss": 0.0209, "step": 41400 }, { "grad_norm": 0.3909716308116913, "learning_rate": 5.122137932758644e-05, "loss": 0.0207, "step": 41410 }, { "grad_norm": 0.3577578067779541, "learning_rate": 5.1200717018505074e-05, "loss": 0.0194, "step": 41420 }, { "grad_norm": 0.27379918098449707, "learning_rate": 5.118005450425403e-05, "loss": 0.0219, "step": 41430 }, { "grad_norm": 0.5454990267753601, "learning_rate": 5.115939178836392e-05, "loss": 0.0205, "step": 41440 }, { "grad_norm": 0.38664865493774414, "learning_rate": 5.113872887436547e-05, "loss": 0.0236, "step": 41450 }, { "grad_norm": 0.2597120702266693, "learning_rate": 5.11180657657894e-05, "loss": 0.025, "step": 41460 }, { "grad_norm": 0.44170060753822327, "learning_rate": 5.109740246616649e-05, "loss": 0.024, "step": 41470 }, { "grad_norm": 0.4550316631793976, "learning_rate": 5.10767389790275e-05, "loss": 0.0355, "step": 41480 }, { "grad_norm": 0.5639116168022156, "learning_rate": 5.1056075307903265e-05, "loss": 0.0231, "step": 41490 }, { "grad_norm": 0.2798197567462921, "learning_rate": 5.1035411456324634e-05, "loss": 0.0262, "step": 41500 }, { "grad_norm": 0.46132779121398926, "learning_rate": 5.1014747427822504e-05, "loss": 0.0258, "step": 41510 }, { "grad_norm": 0.40413370728492737, "learning_rate": 5.0994083225927804e-05, "loss": 0.0324, "step": 41520 }, { "grad_norm": 0.2502342760562897, "learning_rate": 5.0973418854171475e-05, "loss": 0.024, "step": 41530 }, { "grad_norm": 0.1793304979801178, "learning_rate": 5.095275431608447e-05, "loss": 0.0273, "step": 41540 }, { "grad_norm": 0.256462961435318, "learning_rate": 5.093208961519782e-05, "loss": 0.023, "step": 41550 }, { "grad_norm": 0.2879115343093872, "learning_rate": 5.091142475504255e-05, "loss": 0.0256, "step": 41560 }, { "grad_norm": 0.41048750281333923, "learning_rate": 5.089075973914971e-05, "loss": 0.024, "step": 41570 }, { "grad_norm": 0.34152668714523315, "learning_rate": 5.08700945710504e-05, "loss": 0.0287, "step": 41580 }, { "grad_norm": 0.25163161754608154, "learning_rate": 5.0849429254275714e-05, "loss": 0.0253, "step": 41590 }, { "grad_norm": 0.1797989457845688, "learning_rate": 5.0828763792356804e-05, "loss": 0.0195, "step": 41600 }, { "grad_norm": 0.4046301543712616, "learning_rate": 5.080809818882483e-05, "loss": 0.0248, "step": 41610 }, { "grad_norm": 0.30149373412132263, "learning_rate": 5.078743244721097e-05, "loss": 0.0244, "step": 41620 }, { "grad_norm": 0.24703989923000336, "learning_rate": 5.0766766571046455e-05, "loss": 0.0234, "step": 41630 }, { "grad_norm": 0.36401739716529846, "learning_rate": 5.07461005638625e-05, "loss": 0.0253, "step": 41640 }, { "grad_norm": 0.35099655389785767, "learning_rate": 5.072543442919037e-05, "loss": 0.0201, "step": 41650 }, { "grad_norm": 0.27208611369132996, "learning_rate": 5.070476817056132e-05, "loss": 0.0307, "step": 41660 }, { "grad_norm": 0.3403150737285614, "learning_rate": 5.068410179150668e-05, "loss": 0.0147, "step": 41670 }, { "grad_norm": 0.26197731494903564, "learning_rate": 5.066343529555775e-05, "loss": 0.0284, "step": 41680 }, { "grad_norm": 0.29574933648109436, "learning_rate": 5.064276868624587e-05, "loss": 0.0197, "step": 41690 }, { "grad_norm": 0.27073580026626587, "learning_rate": 5.0622101967102396e-05, "loss": 0.0168, "step": 41700 }, { "grad_norm": 0.31886860728263855, "learning_rate": 5.0601435141658714e-05, "loss": 0.0232, "step": 41710 }, { "grad_norm": 0.3290729820728302, "learning_rate": 5.058076821344622e-05, "loss": 0.0202, "step": 41720 }, { "grad_norm": 0.43500110507011414, "learning_rate": 5.056010118599632e-05, "loss": 0.0302, "step": 41730 }, { "grad_norm": 0.4068182706832886, "learning_rate": 5.053943406284044e-05, "loss": 0.0242, "step": 41740 }, { "grad_norm": 0.46719038486480713, "learning_rate": 5.051876684751006e-05, "loss": 0.0266, "step": 41750 }, { "grad_norm": 0.27772894501686096, "learning_rate": 5.0498099543536584e-05, "loss": 0.0216, "step": 41760 }, { "grad_norm": 0.2660539150238037, "learning_rate": 5.047743215445152e-05, "loss": 0.022, "step": 41770 }, { "grad_norm": 0.35621944069862366, "learning_rate": 5.045676468378637e-05, "loss": 0.0262, "step": 41780 }, { "grad_norm": 0.348299503326416, "learning_rate": 5.0436097135072626e-05, "loss": 0.0279, "step": 41790 }, { "grad_norm": 0.2866050899028778, "learning_rate": 5.041542951184181e-05, "loss": 0.0213, "step": 41800 }, { "grad_norm": 0.18757234513759613, "learning_rate": 5.039476181762545e-05, "loss": 0.0199, "step": 41810 }, { "grad_norm": 0.25282686948776245, "learning_rate": 5.037409405595508e-05, "loss": 0.0168, "step": 41820 }, { "grad_norm": 0.2780488431453705, "learning_rate": 5.035342623036229e-05, "loss": 0.0238, "step": 41830 }, { "grad_norm": 0.29938870668411255, "learning_rate": 5.033275834437862e-05, "loss": 0.025, "step": 41840 }, { "grad_norm": 0.3247593641281128, "learning_rate": 5.031209040153564e-05, "loss": 0.0194, "step": 41850 }, { "grad_norm": 0.2678529620170593, "learning_rate": 5.0291422405364955e-05, "loss": 0.0341, "step": 41860 }, { "grad_norm": 0.3852657675743103, "learning_rate": 5.0270754359398133e-05, "loss": 0.0201, "step": 41870 }, { "grad_norm": 0.4514133632183075, "learning_rate": 5.025008626716682e-05, "loss": 0.0238, "step": 41880 }, { "grad_norm": 0.2770475447177887, "learning_rate": 5.0229418132202585e-05, "loss": 0.0204, "step": 41890 }, { "grad_norm": 0.2908383905887604, "learning_rate": 5.020874995803707e-05, "loss": 0.0274, "step": 41900 }, { "grad_norm": 0.30525338649749756, "learning_rate": 5.01880817482019e-05, "loss": 0.0305, "step": 41910 }, { "grad_norm": 0.25613585114479065, "learning_rate": 5.01674135062287e-05, "loss": 0.0221, "step": 41920 }, { "grad_norm": 0.3295629322528839, "learning_rate": 5.0146745235649115e-05, "loss": 0.026, "step": 41930 }, { "grad_norm": 0.3662875294685364, "learning_rate": 5.0126076939994795e-05, "loss": 0.0237, "step": 41940 }, { "grad_norm": 0.2052205502986908, "learning_rate": 5.010540862279736e-05, "loss": 0.0247, "step": 41950 }, { "grad_norm": 0.3266448974609375, "learning_rate": 5.008474028758846e-05, "loss": 0.0225, "step": 41960 }, { "grad_norm": 0.3115103244781494, "learning_rate": 5.0064071937899784e-05, "loss": 0.0243, "step": 41970 }, { "grad_norm": 0.2934742569923401, "learning_rate": 5.004340357726296e-05, "loss": 0.0243, "step": 41980 }, { "grad_norm": 0.5375166535377502, "learning_rate": 5.002273520920965e-05, "loss": 0.0221, "step": 41990 }, { "grad_norm": 0.23020629584789276, "learning_rate": 5.000206683727151e-05, "loss": 0.0206, "step": 42000 }, { "grad_norm": 0.2860504388809204, "learning_rate": 4.998139846498021e-05, "loss": 0.0199, "step": 42010 }, { "grad_norm": 0.3305635154247284, "learning_rate": 4.9960730095867405e-05, "loss": 0.023, "step": 42020 }, { "grad_norm": 0.33351609110832214, "learning_rate": 4.9940061733464755e-05, "loss": 0.0268, "step": 42030 }, { "grad_norm": 0.5026934146881104, "learning_rate": 4.991939338130392e-05, "loss": 0.0315, "step": 42040 }, { "grad_norm": 0.4255141615867615, "learning_rate": 4.989872504291653e-05, "loss": 0.0325, "step": 42050 }, { "grad_norm": 0.32257506251335144, "learning_rate": 4.9878056721834273e-05, "loss": 0.0293, "step": 42060 }, { "grad_norm": 0.3112999498844147, "learning_rate": 4.98573884215888e-05, "loss": 0.0193, "step": 42070 }, { "grad_norm": 0.42352309823036194, "learning_rate": 4.9836720145711715e-05, "loss": 0.0265, "step": 42080 }, { "grad_norm": 0.3187059462070465, "learning_rate": 4.9816051897734725e-05, "loss": 0.0236, "step": 42090 }, { "grad_norm": 0.3871201276779175, "learning_rate": 4.979538368118941e-05, "loss": 0.027, "step": 42100 }, { "grad_norm": 0.1988004446029663, "learning_rate": 4.9774715499607446e-05, "loss": 0.0193, "step": 42110 }, { "grad_norm": 0.43766093254089355, "learning_rate": 4.9754047356520416e-05, "loss": 0.0309, "step": 42120 }, { "grad_norm": 0.39284759759902954, "learning_rate": 4.973337925545999e-05, "loss": 0.0231, "step": 42130 }, { "grad_norm": 0.21531711518764496, "learning_rate": 4.9712711199957746e-05, "loss": 0.0164, "step": 42140 }, { "grad_norm": 0.3493574261665344, "learning_rate": 4.969204319354531e-05, "loss": 0.0247, "step": 42150 }, { "grad_norm": 0.2662866711616516, "learning_rate": 4.9671375239754267e-05, "loss": 0.0201, "step": 42160 }, { "grad_norm": 0.21904736757278442, "learning_rate": 4.9650707342116195e-05, "loss": 0.0172, "step": 42170 }, { "grad_norm": 0.2287312000989914, "learning_rate": 4.96300395041627e-05, "loss": 0.0217, "step": 42180 }, { "grad_norm": 0.5719941854476929, "learning_rate": 4.960937172942532e-05, "loss": 0.0203, "step": 42190 }, { "grad_norm": 0.16952350735664368, "learning_rate": 4.958870402143565e-05, "loss": 0.0207, "step": 42200 }, { "grad_norm": 0.2562576234340668, "learning_rate": 4.9568036383725186e-05, "loss": 0.0244, "step": 42210 }, { "grad_norm": 0.35633254051208496, "learning_rate": 4.9547368819825516e-05, "loss": 0.0203, "step": 42220 }, { "grad_norm": 0.4068620800971985, "learning_rate": 4.952670133326812e-05, "loss": 0.0228, "step": 42230 }, { "grad_norm": 0.24227876961231232, "learning_rate": 4.950603392758453e-05, "loss": 0.0241, "step": 42240 }, { "grad_norm": 0.2947617173194885, "learning_rate": 4.948536660630621e-05, "loss": 0.0229, "step": 42250 }, { "grad_norm": 0.2611733376979828, "learning_rate": 4.9464699372964685e-05, "loss": 0.0226, "step": 42260 }, { "grad_norm": 0.2626269459724426, "learning_rate": 4.9444032231091395e-05, "loss": 0.022, "step": 42270 }, { "grad_norm": 0.27462488412857056, "learning_rate": 4.9423365184217776e-05, "loss": 0.0197, "step": 42280 }, { "grad_norm": 0.34203559160232544, "learning_rate": 4.940269823587529e-05, "loss": 0.0215, "step": 42290 }, { "grad_norm": 0.23755177855491638, "learning_rate": 4.938203138959533e-05, "loss": 0.0149, "step": 42300 }, { "grad_norm": 0.36070549488067627, "learning_rate": 4.9361364648909325e-05, "loss": 0.0201, "step": 42310 }, { "grad_norm": 0.21882064640522003, "learning_rate": 4.9340698017348605e-05, "loss": 0.0174, "step": 42320 }, { "grad_norm": 0.36416468024253845, "learning_rate": 4.932003149844458e-05, "loss": 0.0162, "step": 42330 }, { "grad_norm": 0.2588927745819092, "learning_rate": 4.929936509572857e-05, "loss": 0.0298, "step": 42340 }, { "grad_norm": 0.5626840591430664, "learning_rate": 4.927869881273191e-05, "loss": 0.0256, "step": 42350 }, { "grad_norm": 0.3506235182285309, "learning_rate": 4.9258032652985894e-05, "loss": 0.0243, "step": 42360 }, { "grad_norm": 0.2897201478481293, "learning_rate": 4.9237366620021786e-05, "loss": 0.0251, "step": 42370 }, { "grad_norm": 0.7603388428688049, "learning_rate": 4.921670071737089e-05, "loss": 0.0213, "step": 42380 }, { "grad_norm": 0.30639246106147766, "learning_rate": 4.91960349485644e-05, "loss": 0.0246, "step": 42390 }, { "grad_norm": 0.31535008549690247, "learning_rate": 4.9175369317133566e-05, "loss": 0.0222, "step": 42400 }, { "grad_norm": 0.18491153419017792, "learning_rate": 4.915470382660953e-05, "loss": 0.0219, "step": 42410 }, { "grad_norm": 0.25874534249305725, "learning_rate": 4.9134038480523524e-05, "loss": 0.02, "step": 42420 }, { "grad_norm": 0.23707979917526245, "learning_rate": 4.911337328240664e-05, "loss": 0.0216, "step": 42430 }, { "grad_norm": 0.1914302110671997, "learning_rate": 4.909270823579003e-05, "loss": 0.0219, "step": 42440 }, { "grad_norm": 0.325128972530365, "learning_rate": 4.907204334420476e-05, "loss": 0.0205, "step": 42450 }, { "grad_norm": 0.22505798935890198, "learning_rate": 4.9051378611181884e-05, "loss": 0.0216, "step": 42460 }, { "grad_norm": 0.2945755422115326, "learning_rate": 4.9030714040252486e-05, "loss": 0.0217, "step": 42470 }, { "grad_norm": 0.2382335662841797, "learning_rate": 4.901004963494752e-05, "loss": 0.023, "step": 42480 }, { "grad_norm": 0.38868287205696106, "learning_rate": 4.898938539879802e-05, "loss": 0.0305, "step": 42490 }, { "grad_norm": 0.3558856248855591, "learning_rate": 4.8968721335334904e-05, "loss": 0.0226, "step": 42500 }, { "grad_norm": 0.33327600359916687, "learning_rate": 4.894805744808912e-05, "loss": 0.0227, "step": 42510 }, { "grad_norm": 0.3171023428440094, "learning_rate": 4.8927393740591524e-05, "loss": 0.0213, "step": 42520 }, { "grad_norm": 0.42515045404434204, "learning_rate": 4.890673021637302e-05, "loss": 0.0198, "step": 42530 }, { "grad_norm": 0.28451284766197205, "learning_rate": 4.888606687896442e-05, "loss": 0.0275, "step": 42540 }, { "grad_norm": 0.25781771540641785, "learning_rate": 4.886540373189652e-05, "loss": 0.0227, "step": 42550 }, { "grad_norm": 0.3692363500595093, "learning_rate": 4.88447407787001e-05, "loss": 0.0226, "step": 42560 }, { "grad_norm": 0.27495628595352173, "learning_rate": 4.8824078022905864e-05, "loss": 0.0211, "step": 42570 }, { "grad_norm": 0.24900229275226593, "learning_rate": 4.880341546804455e-05, "loss": 0.0163, "step": 42580 }, { "grad_norm": 0.379098117351532, "learning_rate": 4.878275311764679e-05, "loss": 0.0238, "step": 42590 }, { "grad_norm": 0.38450485467910767, "learning_rate": 4.8762090975243254e-05, "loss": 0.0224, "step": 42600 }, { "grad_norm": 0.1828886717557907, "learning_rate": 4.874142904436448e-05, "loss": 0.0187, "step": 42610 }, { "grad_norm": 0.3817150294780731, "learning_rate": 4.8720767328541094e-05, "loss": 0.0274, "step": 42620 }, { "grad_norm": 0.37428709864616394, "learning_rate": 4.870010583130358e-05, "loss": 0.0196, "step": 42630 }, { "grad_norm": 0.2158813327550888, "learning_rate": 4.867944455618241e-05, "loss": 0.0256, "step": 42640 }, { "grad_norm": 0.2833746373653412, "learning_rate": 4.865878350670807e-05, "loss": 0.0238, "step": 42650 }, { "grad_norm": 0.27663514018058777, "learning_rate": 4.8638122686410914e-05, "loss": 0.021, "step": 42660 }, { "grad_norm": 0.24333776533603668, "learning_rate": 4.861746209882137e-05, "loss": 0.0152, "step": 42670 }, { "grad_norm": 0.2290867418050766, "learning_rate": 4.859680174746972e-05, "loss": 0.0207, "step": 42680 }, { "grad_norm": 0.24990063905715942, "learning_rate": 4.857614163588629e-05, "loss": 0.0278, "step": 42690 }, { "grad_norm": 0.2949153184890747, "learning_rate": 4.855548176760131e-05, "loss": 0.0189, "step": 42700 }, { "grad_norm": 0.21972200274467468, "learning_rate": 4.8534822146144986e-05, "loss": 0.0221, "step": 42710 }, { "grad_norm": 0.36125537753105164, "learning_rate": 4.851416277504749e-05, "loss": 0.0317, "step": 42720 }, { "grad_norm": 0.2823760211467743, "learning_rate": 4.8493503657838923e-05, "loss": 0.0198, "step": 42730 }, { "grad_norm": 0.19603806734085083, "learning_rate": 4.84728447980494e-05, "loss": 0.0175, "step": 42740 }, { "grad_norm": 0.5414766669273376, "learning_rate": 4.8452186199208914e-05, "loss": 0.0244, "step": 42750 }, { "grad_norm": 0.25218793749809265, "learning_rate": 4.843152786484749e-05, "loss": 0.0216, "step": 42760 }, { "grad_norm": 0.20472240447998047, "learning_rate": 4.8410869798495035e-05, "loss": 0.0178, "step": 42770 }, { "grad_norm": 0.2778458595275879, "learning_rate": 4.8390212003681486e-05, "loss": 0.0204, "step": 42780 }, { "grad_norm": 0.46683794260025024, "learning_rate": 4.836955448393667e-05, "loss": 0.0302, "step": 42790 }, { "grad_norm": 0.23543888330459595, "learning_rate": 4.8348897242790406e-05, "loss": 0.0189, "step": 42800 }, { "grad_norm": 0.24517440795898438, "learning_rate": 4.832824028377243e-05, "loss": 0.0198, "step": 42810 }, { "grad_norm": 0.3933175802230835, "learning_rate": 4.830758361041249e-05, "loss": 0.0196, "step": 42820 }, { "grad_norm": 0.28439801931381226, "learning_rate": 4.828692722624022e-05, "loss": 0.0242, "step": 42830 }, { "grad_norm": 0.15527699887752533, "learning_rate": 4.826627113478522e-05, "loss": 0.0391, "step": 42840 }, { "grad_norm": 0.19635455310344696, "learning_rate": 4.824561533957708e-05, "loss": 0.0201, "step": 42850 }, { "grad_norm": 0.3118305206298828, "learning_rate": 4.822495984414527e-05, "loss": 0.0224, "step": 42860 }, { "grad_norm": 0.3558688461780548, "learning_rate": 4.8204304652019304e-05, "loss": 0.019, "step": 42870 }, { "grad_norm": 0.43026265501976013, "learning_rate": 4.8183649766728537e-05, "loss": 0.0219, "step": 42880 }, { "grad_norm": 0.4432372450828552, "learning_rate": 4.816299519180237e-05, "loss": 0.0232, "step": 42890 }, { "grad_norm": 0.20056992769241333, "learning_rate": 4.8142340930770066e-05, "loss": 0.0188, "step": 42900 }, { "grad_norm": 0.22686786949634552, "learning_rate": 4.81216869871609e-05, "loss": 0.0178, "step": 42910 }, { "grad_norm": 0.3405471444129944, "learning_rate": 4.8101033364504064e-05, "loss": 0.0197, "step": 42920 }, { "grad_norm": 0.2828798294067383, "learning_rate": 4.808038006632866e-05, "loss": 0.0282, "step": 42930 }, { "grad_norm": 0.38353657722473145, "learning_rate": 4.805972709616383e-05, "loss": 0.0182, "step": 42940 }, { "grad_norm": 0.2121635377407074, "learning_rate": 4.8039074457538556e-05, "loss": 0.0166, "step": 42950 }, { "grad_norm": 0.19540086388587952, "learning_rate": 4.801842215398184e-05, "loss": 0.0217, "step": 42960 }, { "grad_norm": 0.3028792142868042, "learning_rate": 4.799777018902256e-05, "loss": 0.0225, "step": 42970 }, { "grad_norm": 0.26206183433532715, "learning_rate": 4.797711856618961e-05, "loss": 0.0167, "step": 42980 }, { "grad_norm": 0.27989116311073303, "learning_rate": 4.7956467289011765e-05, "loss": 0.0179, "step": 42990 }, { "grad_norm": 0.38995134830474854, "learning_rate": 4.793581636101778e-05, "loss": 0.0197, "step": 43000 }, { "grad_norm": 0.3244372308254242, "learning_rate": 4.7915165785736326e-05, "loss": 0.0202, "step": 43010 }, { "grad_norm": 0.3700791001319885, "learning_rate": 4.789451556669599e-05, "loss": 0.0189, "step": 43020 }, { "grad_norm": 0.2728840410709381, "learning_rate": 4.7873865707425384e-05, "loss": 0.0227, "step": 43030 }, { "grad_norm": 0.23324964940547943, "learning_rate": 4.7853216211452975e-05, "loss": 0.0179, "step": 43040 }, { "grad_norm": 0.3169993460178375, "learning_rate": 4.78325670823072e-05, "loss": 0.0225, "step": 43050 }, { "grad_norm": 0.22832819819450378, "learning_rate": 4.781191832351641e-05, "loss": 0.025, "step": 43060 }, { "grad_norm": 0.41039201617240906, "learning_rate": 4.7791269938608965e-05, "loss": 0.0225, "step": 43070 }, { "grad_norm": 0.34172624349594116, "learning_rate": 4.777062193111305e-05, "loss": 0.0217, "step": 43080 }, { "grad_norm": 0.20276693999767303, "learning_rate": 4.77499743045569e-05, "loss": 0.0218, "step": 43090 }, { "grad_norm": 0.3310873210430145, "learning_rate": 4.7729327062468596e-05, "loss": 0.0215, "step": 43100 }, { "grad_norm": 0.40806686878204346, "learning_rate": 4.770868020837619e-05, "loss": 0.0231, "step": 43110 }, { "grad_norm": 0.2240891009569168, "learning_rate": 4.768803374580768e-05, "loss": 0.0172, "step": 43120 }, { "grad_norm": 0.30635350942611694, "learning_rate": 4.766738767829094e-05, "loss": 0.0177, "step": 43130 }, { "grad_norm": 0.2368614375591278, "learning_rate": 4.764674200935388e-05, "loss": 0.0241, "step": 43140 }, { "grad_norm": 0.22779417037963867, "learning_rate": 4.762609674252424e-05, "loss": 0.019, "step": 43150 }, { "grad_norm": 0.274012953042984, "learning_rate": 4.760545188132974e-05, "loss": 0.0364, "step": 43160 }, { "grad_norm": 0.24072375893592834, "learning_rate": 4.758480742929801e-05, "loss": 0.0207, "step": 43170 }, { "grad_norm": 0.5729919075965881, "learning_rate": 4.756416338995664e-05, "loss": 0.0287, "step": 43180 }, { "grad_norm": 0.3760319650173187, "learning_rate": 4.7543519766833145e-05, "loss": 0.0171, "step": 43190 }, { "grad_norm": 0.21508169174194336, "learning_rate": 4.752287656345492e-05, "loss": 0.0209, "step": 43200 }, { "grad_norm": 0.2949329912662506, "learning_rate": 4.7502233783349355e-05, "loss": 0.0267, "step": 43210 }, { "grad_norm": 0.22826996445655823, "learning_rate": 4.7481591430043694e-05, "loss": 0.0166, "step": 43220 }, { "grad_norm": 0.22422610223293304, "learning_rate": 4.746094950706521e-05, "loss": 0.0227, "step": 43230 }, { "grad_norm": 0.31614989042282104, "learning_rate": 4.744030801794099e-05, "loss": 0.0218, "step": 43240 }, { "grad_norm": 0.2648484408855438, "learning_rate": 4.741966696619813e-05, "loss": 0.0236, "step": 43250 }, { "grad_norm": 0.21043819189071655, "learning_rate": 4.739902635536361e-05, "loss": 0.0287, "step": 43260 }, { "grad_norm": 0.2911488711833954, "learning_rate": 4.737838618896436e-05, "loss": 0.0209, "step": 43270 }, { "grad_norm": 0.3353821635246277, "learning_rate": 4.7357746470527203e-05, "loss": 0.0235, "step": 43280 }, { "grad_norm": 0.35369089245796204, "learning_rate": 4.73371072035789e-05, "loss": 0.028, "step": 43290 }, { "grad_norm": 0.5880051851272583, "learning_rate": 4.731646839164616e-05, "loss": 0.0243, "step": 43300 }, { "grad_norm": 0.3164820671081543, "learning_rate": 4.729583003825556e-05, "loss": 0.0262, "step": 43310 }, { "grad_norm": 0.2440141886472702, "learning_rate": 4.727519214693367e-05, "loss": 0.0238, "step": 43320 }, { "grad_norm": 0.26604485511779785, "learning_rate": 4.725455472120689e-05, "loss": 0.0231, "step": 43330 }, { "grad_norm": 0.33786216378211975, "learning_rate": 4.723391776460164e-05, "loss": 0.0258, "step": 43340 }, { "grad_norm": 0.4505155086517334, "learning_rate": 4.7213281280644186e-05, "loss": 0.0255, "step": 43350 }, { "grad_norm": 0.2451440989971161, "learning_rate": 4.719264527286075e-05, "loss": 0.0264, "step": 43360 }, { "grad_norm": 0.35219013690948486, "learning_rate": 4.717200974477744e-05, "loss": 0.0245, "step": 43370 }, { "grad_norm": 0.2616503834724426, "learning_rate": 4.715137469992034e-05, "loss": 0.03, "step": 43380 }, { "grad_norm": 0.4288248121738434, "learning_rate": 4.713074014181539e-05, "loss": 0.0219, "step": 43390 }, { "grad_norm": 0.23014359176158905, "learning_rate": 4.711010607398846e-05, "loss": 0.0181, "step": 43400 }, { "grad_norm": 0.366390198469162, "learning_rate": 4.708947249996538e-05, "loss": 0.0211, "step": 43410 }, { "grad_norm": 0.2738763689994812, "learning_rate": 4.706883942327183e-05, "loss": 0.0167, "step": 43420 }, { "grad_norm": 0.36525389552116394, "learning_rate": 4.704820684743347e-05, "loss": 0.0246, "step": 43430 }, { "grad_norm": 0.3001834452152252, "learning_rate": 4.702757477597581e-05, "loss": 0.0204, "step": 43440 }, { "grad_norm": 0.31026387214660645, "learning_rate": 4.700694321242435e-05, "loss": 0.0217, "step": 43450 }, { "grad_norm": 0.34850412607192993, "learning_rate": 4.698631216030441e-05, "loss": 0.0244, "step": 43460 }, { "grad_norm": 0.18162676692008972, "learning_rate": 4.6965681623141314e-05, "loss": 0.0174, "step": 43470 }, { "grad_norm": 0.3014936149120331, "learning_rate": 4.694505160446024e-05, "loss": 0.0163, "step": 43480 }, { "grad_norm": 0.21205884218215942, "learning_rate": 4.692442210778628e-05, "loss": 0.0214, "step": 43490 }, { "grad_norm": 0.38790929317474365, "learning_rate": 4.690379313664448e-05, "loss": 0.0219, "step": 43500 }, { "grad_norm": 0.3687449097633362, "learning_rate": 4.688316469455973e-05, "loss": 0.0216, "step": 43510 }, { "grad_norm": 0.26625609397888184, "learning_rate": 4.6862536785056914e-05, "loss": 0.021, "step": 43520 }, { "grad_norm": 0.522006094455719, "learning_rate": 4.684190941166073e-05, "loss": 0.0231, "step": 43530 }, { "grad_norm": 0.2291153073310852, "learning_rate": 4.682128257789587e-05, "loss": 0.0201, "step": 43540 }, { "grad_norm": 1.043678879737854, "learning_rate": 4.680065628728687e-05, "loss": 0.0213, "step": 43550 }, { "grad_norm": 0.28929436206817627, "learning_rate": 4.678003054335822e-05, "loss": 0.0234, "step": 43560 }, { "grad_norm": 0.36229637265205383, "learning_rate": 4.675940534963428e-05, "loss": 0.0266, "step": 43570 }, { "grad_norm": 0.21428024768829346, "learning_rate": 4.673878070963931e-05, "loss": 0.0217, "step": 43580 }, { "grad_norm": 0.5903191566467285, "learning_rate": 4.671815662689756e-05, "loss": 0.0172, "step": 43590 }, { "grad_norm": 0.2627982497215271, "learning_rate": 4.669753310493306e-05, "loss": 0.0199, "step": 43600 }, { "grad_norm": 0.21399006247520447, "learning_rate": 4.6676910147269845e-05, "loss": 0.0233, "step": 43610 }, { "grad_norm": 0.33591151237487793, "learning_rate": 4.665628775743177e-05, "loss": 0.0204, "step": 43620 }, { "grad_norm": 0.2326565533876419, "learning_rate": 4.66356659389427e-05, "loss": 0.02, "step": 43630 }, { "grad_norm": 0.30910319089889526, "learning_rate": 4.661504469532629e-05, "loss": 0.0183, "step": 43640 }, { "grad_norm": 0.28717586398124695, "learning_rate": 4.659442403010617e-05, "loss": 0.0178, "step": 43650 }, { "grad_norm": 0.31333181262016296, "learning_rate": 4.6573803946805845e-05, "loss": 0.0231, "step": 43660 }, { "grad_norm": 0.22428211569786072, "learning_rate": 4.6553184448948695e-05, "loss": 0.0221, "step": 43670 }, { "grad_norm": 0.24606014788150787, "learning_rate": 4.653256554005807e-05, "loss": 0.0172, "step": 43680 }, { "grad_norm": 0.1740453690290451, "learning_rate": 4.6511947223657145e-05, "loss": 0.0221, "step": 43690 }, { "grad_norm": 0.4393605887889862, "learning_rate": 4.649132950326906e-05, "loss": 0.0248, "step": 43700 }, { "grad_norm": 0.3786892890930176, "learning_rate": 4.647071238241679e-05, "loss": 0.0244, "step": 43710 }, { "grad_norm": 0.19057515263557434, "learning_rate": 4.6450095864623264e-05, "loss": 0.0169, "step": 43720 }, { "grad_norm": 0.31785842776298523, "learning_rate": 4.642947995341123e-05, "loss": 0.0152, "step": 43730 }, { "grad_norm": 0.18593133985996246, "learning_rate": 4.640886465230345e-05, "loss": 0.0219, "step": 43740 }, { "grad_norm": 0.30023273825645447, "learning_rate": 4.6388249964822485e-05, "loss": 0.0237, "step": 43750 }, { "grad_norm": 0.37744447588920593, "learning_rate": 4.63676358944908e-05, "loss": 0.0187, "step": 43760 }, { "grad_norm": 0.4326973259449005, "learning_rate": 4.634702244483081e-05, "loss": 0.0255, "step": 43770 }, { "grad_norm": 0.17305506765842438, "learning_rate": 4.632640961936475e-05, "loss": 0.0186, "step": 43780 }, { "grad_norm": 0.2331002801656723, "learning_rate": 4.6305797421614835e-05, "loss": 0.0184, "step": 43790 }, { "grad_norm": 0.33875352144241333, "learning_rate": 4.6285185855103105e-05, "loss": 0.0228, "step": 43800 }, { "grad_norm": 0.344493567943573, "learning_rate": 4.626457492335151e-05, "loss": 0.0202, "step": 43810 }, { "grad_norm": 0.24887274205684662, "learning_rate": 4.624396462988188e-05, "loss": 0.0174, "step": 43820 }, { "grad_norm": 0.368610143661499, "learning_rate": 4.6223354978215985e-05, "loss": 0.0218, "step": 43830 }, { "grad_norm": 0.30308374762535095, "learning_rate": 4.620274597187544e-05, "loss": 0.0256, "step": 43840 }, { "grad_norm": 0.2968035638332367, "learning_rate": 4.6182137614381736e-05, "loss": 0.0201, "step": 43850 }, { "grad_norm": 0.24736304581165314, "learning_rate": 4.616152990925631e-05, "loss": 0.019, "step": 43860 }, { "grad_norm": 0.22736182808876038, "learning_rate": 4.614092286002041e-05, "loss": 0.0255, "step": 43870 }, { "grad_norm": 0.2891647517681122, "learning_rate": 4.612031647019528e-05, "loss": 0.0304, "step": 43880 }, { "grad_norm": 0.6405381560325623, "learning_rate": 4.609971074330193e-05, "loss": 0.0291, "step": 43890 }, { "grad_norm": 1.9297974109649658, "learning_rate": 4.607910568286136e-05, "loss": 0.0231, "step": 43900 }, { "grad_norm": 0.5065783858299255, "learning_rate": 4.605850129239438e-05, "loss": 0.0283, "step": 43910 }, { "grad_norm": 0.30340656638145447, "learning_rate": 4.603789757542174e-05, "loss": 0.0218, "step": 43920 }, { "grad_norm": 0.3787758946418762, "learning_rate": 4.601729453546404e-05, "loss": 0.0233, "step": 43930 }, { "grad_norm": 0.295548677444458, "learning_rate": 4.599669217604177e-05, "loss": 0.0245, "step": 43940 }, { "grad_norm": 0.3847356140613556, "learning_rate": 4.597609050067532e-05, "loss": 0.0289, "step": 43950 }, { "grad_norm": 0.33070850372314453, "learning_rate": 4.595548951288495e-05, "loss": 0.0266, "step": 43960 }, { "grad_norm": 0.2507905960083008, "learning_rate": 4.593488921619081e-05, "loss": 0.0243, "step": 43970 }, { "grad_norm": 0.37177643179893494, "learning_rate": 4.591428961411289e-05, "loss": 0.0306, "step": 43980 }, { "grad_norm": 0.3119390308856964, "learning_rate": 4.589369071017117e-05, "loss": 0.0263, "step": 43990 }, { "grad_norm": 0.16262218356132507, "learning_rate": 4.587309250788538e-05, "loss": 0.023, "step": 44000 }, { "grad_norm": 0.21809469163417816, "learning_rate": 4.585249501077522e-05, "loss": 0.0188, "step": 44010 }, { "grad_norm": 0.25210145115852356, "learning_rate": 4.583189822236021e-05, "loss": 0.0215, "step": 44020 }, { "grad_norm": 0.5005665421485901, "learning_rate": 4.5811302146159816e-05, "loss": 0.0196, "step": 44030 }, { "grad_norm": 0.23151667416095734, "learning_rate": 4.579070678569332e-05, "loss": 0.0181, "step": 44040 }, { "grad_norm": 0.3119015693664551, "learning_rate": 4.5770112144479904e-05, "loss": 0.0228, "step": 44050 }, { "grad_norm": 0.4060675799846649, "learning_rate": 4.5749518226038645e-05, "loss": 0.0218, "step": 44060 }, { "grad_norm": 0.3893332779407501, "learning_rate": 4.572892503388845e-05, "loss": 0.02, "step": 44070 }, { "grad_norm": 0.22583557665348053, "learning_rate": 4.570833257154817e-05, "loss": 0.0224, "step": 44080 }, { "grad_norm": 0.2818431556224823, "learning_rate": 4.568774084253646e-05, "loss": 0.0275, "step": 44090 }, { "grad_norm": 0.26052162051200867, "learning_rate": 4.566714985037191e-05, "loss": 0.0283, "step": 44100 }, { "grad_norm": 0.386844664812088, "learning_rate": 4.564655959857295e-05, "loss": 0.0279, "step": 44110 }, { "grad_norm": 0.29816552996635437, "learning_rate": 4.5625970090657893e-05, "loss": 0.0224, "step": 44120 }, { "grad_norm": 0.20558400452136993, "learning_rate": 4.560538133014491e-05, "loss": 0.0188, "step": 44130 }, { "grad_norm": 0.3368612825870514, "learning_rate": 4.5584793320552055e-05, "loss": 0.0191, "step": 44140 }, { "grad_norm": 0.2262062430381775, "learning_rate": 4.556420606539728e-05, "loss": 0.0221, "step": 44150 }, { "grad_norm": 0.17952634394168854, "learning_rate": 4.554361956819836e-05, "loss": 0.0185, "step": 44160 }, { "grad_norm": 0.19924180209636688, "learning_rate": 4.552303383247299e-05, "loss": 0.0196, "step": 44170 }, { "grad_norm": 0.21057291328907013, "learning_rate": 4.5502448861738664e-05, "loss": 0.0177, "step": 44180 }, { "grad_norm": 0.27551278471946716, "learning_rate": 4.548186465951284e-05, "loss": 0.029, "step": 44190 }, { "grad_norm": 0.411012202501297, "learning_rate": 4.546128122931277e-05, "loss": 0.0263, "step": 44200 }, { "grad_norm": 0.3369060754776001, "learning_rate": 4.544069857465561e-05, "loss": 0.0257, "step": 44210 }, { "grad_norm": 0.6333129405975342, "learning_rate": 4.542011669905837e-05, "loss": 0.0199, "step": 44220 }, { "grad_norm": 0.38170769810676575, "learning_rate": 4.53995356060379e-05, "loss": 0.0173, "step": 44230 }, { "grad_norm": 0.3008088767528534, "learning_rate": 4.5378955299110994e-05, "loss": 0.0187, "step": 44240 }, { "grad_norm": 0.350247859954834, "learning_rate": 4.5358375781794233e-05, "loss": 0.0266, "step": 44250 }, { "grad_norm": 0.3309068977832794, "learning_rate": 4.533779705760411e-05, "loss": 0.0251, "step": 44260 }, { "grad_norm": 0.21941465139389038, "learning_rate": 4.5317219130056934e-05, "loss": 0.0178, "step": 44270 }, { "grad_norm": 0.26033708453178406, "learning_rate": 4.5296642002668964e-05, "loss": 0.0201, "step": 44280 }, { "grad_norm": 0.3939238488674164, "learning_rate": 4.5276065678956216e-05, "loss": 0.0244, "step": 44290 }, { "grad_norm": 0.27216869592666626, "learning_rate": 4.525549016243466e-05, "loss": 0.0228, "step": 44300 }, { "grad_norm": 0.26961788535118103, "learning_rate": 4.523491545662008e-05, "loss": 0.0206, "step": 44310 }, { "grad_norm": 0.25223761796951294, "learning_rate": 4.5214341565028104e-05, "loss": 0.0258, "step": 44320 }, { "grad_norm": 0.3517848253250122, "learning_rate": 4.519376849117428e-05, "loss": 0.0213, "step": 44330 }, { "grad_norm": 0.1737210750579834, "learning_rate": 4.517319623857395e-05, "loss": 0.0267, "step": 44340 }, { "grad_norm": 0.36735519766807556, "learning_rate": 4.5152624810742394e-05, "loss": 0.0224, "step": 44350 }, { "grad_norm": 0.2358761578798294, "learning_rate": 4.5132054211194664e-05, "loss": 0.0193, "step": 44360 }, { "grad_norm": 0.3094848394393921, "learning_rate": 4.511148444344574e-05, "loss": 0.018, "step": 44370 }, { "grad_norm": 0.30613985657691956, "learning_rate": 4.509091551101041e-05, "loss": 0.024, "step": 44380 }, { "grad_norm": 0.21519222855567932, "learning_rate": 4.507034741740338e-05, "loss": 0.0181, "step": 44390 }, { "grad_norm": 0.3242504596710205, "learning_rate": 4.5049780166139145e-05, "loss": 0.0239, "step": 44400 }, { "grad_norm": 0.3310684561729431, "learning_rate": 4.5029213760732075e-05, "loss": 0.0298, "step": 44410 }, { "grad_norm": 0.26412123441696167, "learning_rate": 4.5008648204696434e-05, "loss": 0.0282, "step": 44420 }, { "grad_norm": 0.2590215504169464, "learning_rate": 4.4988083501546284e-05, "loss": 0.0181, "step": 44430 }, { "grad_norm": 0.2507920265197754, "learning_rate": 4.4967519654795606e-05, "loss": 0.018, "step": 44440 }, { "grad_norm": 0.26044631004333496, "learning_rate": 4.494695666795816e-05, "loss": 0.0165, "step": 44450 }, { "grad_norm": 0.37702009081840515, "learning_rate": 4.4926394544547644e-05, "loss": 0.0207, "step": 44460 }, { "grad_norm": 0.2824760973453522, "learning_rate": 4.490583328807752e-05, "loss": 0.0237, "step": 44470 }, { "grad_norm": 0.2746288776397705, "learning_rate": 4.488527290206117e-05, "loss": 0.022, "step": 44480 }, { "grad_norm": 0.34350594878196716, "learning_rate": 4.48647133900118e-05, "loss": 0.0186, "step": 44490 }, { "grad_norm": 0.24849040806293488, "learning_rate": 4.484415475544244e-05, "loss": 0.024, "step": 44500 }, { "grad_norm": 0.23030690848827362, "learning_rate": 4.4823597001866035e-05, "loss": 0.0181, "step": 44510 }, { "grad_norm": 0.34061381220817566, "learning_rate": 4.480304013279532e-05, "loss": 0.0199, "step": 44520 }, { "grad_norm": 0.3935643136501312, "learning_rate": 4.478248415174292e-05, "loss": 0.0171, "step": 44530 }, { "grad_norm": 0.28429335355758667, "learning_rate": 4.476192906222126e-05, "loss": 0.0243, "step": 44540 }, { "grad_norm": 0.2894410192966461, "learning_rate": 4.474137486774268e-05, "loss": 0.0317, "step": 44550 }, { "grad_norm": 0.2923363745212555, "learning_rate": 4.4720821571819296e-05, "loss": 0.0217, "step": 44560 }, { "grad_norm": 0.25083470344543457, "learning_rate": 4.470026917796314e-05, "loss": 0.0268, "step": 44570 }, { "grad_norm": 0.30173584818840027, "learning_rate": 4.4679717689686005e-05, "loss": 0.0382, "step": 44580 }, { "grad_norm": 0.2736093997955322, "learning_rate": 4.465916711049963e-05, "loss": 0.0183, "step": 44590 }, { "grad_norm": 0.2706838548183441, "learning_rate": 4.4638617443915524e-05, "loss": 0.0218, "step": 44600 }, { "grad_norm": 0.3208436071872711, "learning_rate": 4.4618068693445055e-05, "loss": 0.0246, "step": 44610 }, { "grad_norm": 0.20507599413394928, "learning_rate": 4.459752086259946e-05, "loss": 0.0167, "step": 44620 }, { "grad_norm": 0.4374319314956665, "learning_rate": 4.457697395488977e-05, "loss": 0.0331, "step": 44630 }, { "grad_norm": 0.26838499307632446, "learning_rate": 4.455642797382693e-05, "loss": 0.0296, "step": 44640 }, { "grad_norm": 0.31834670901298523, "learning_rate": 4.4535882922921643e-05, "loss": 0.0294, "step": 44650 }, { "grad_norm": 0.30966392159461975, "learning_rate": 4.451533880568455e-05, "loss": 0.0192, "step": 44660 }, { "grad_norm": 0.19468989968299866, "learning_rate": 4.449479562562603e-05, "loss": 0.0267, "step": 44670 }, { "grad_norm": 0.2926475405693054, "learning_rate": 4.447425338625639e-05, "loss": 0.0245, "step": 44680 }, { "grad_norm": 0.28649646043777466, "learning_rate": 4.4453712091085705e-05, "loss": 0.0226, "step": 44690 }, { "grad_norm": 0.48792222142219543, "learning_rate": 4.443317174362392e-05, "loss": 0.0262, "step": 44700 }, { "grad_norm": 0.631357729434967, "learning_rate": 4.441263234738085e-05, "loss": 0.0233, "step": 44710 }, { "grad_norm": 0.30610227584838867, "learning_rate": 4.4392093905866086e-05, "loss": 0.0141, "step": 44720 }, { "grad_norm": 0.3662363588809967, "learning_rate": 4.437155642258911e-05, "loss": 0.0175, "step": 44730 }, { "grad_norm": 0.2021157443523407, "learning_rate": 4.4351019901059177e-05, "loss": 0.0264, "step": 44740 }, { "grad_norm": 0.28192493319511414, "learning_rate": 4.433048434478547e-05, "loss": 0.0326, "step": 44750 }, { "grad_norm": 0.3322913348674774, "learning_rate": 4.4309949757276906e-05, "loss": 0.0249, "step": 44760 }, { "grad_norm": 0.24528877437114716, "learning_rate": 4.428941614204233e-05, "loss": 0.0254, "step": 44770 }, { "grad_norm": 0.2466924786567688, "learning_rate": 4.426888350259034e-05, "loss": 0.0229, "step": 44780 }, { "grad_norm": 0.27947384119033813, "learning_rate": 4.424835184242939e-05, "loss": 0.0311, "step": 44790 }, { "grad_norm": 0.1768038421869278, "learning_rate": 4.422782116506784e-05, "loss": 0.0187, "step": 44800 }, { "grad_norm": 0.39571964740753174, "learning_rate": 4.4207291474013754e-05, "loss": 0.0243, "step": 44810 }, { "grad_norm": 0.28314849734306335, "learning_rate": 4.418676277277514e-05, "loss": 0.0275, "step": 44820 }, { "grad_norm": 0.25012239813804626, "learning_rate": 4.416623506485976e-05, "loss": 0.0285, "step": 44830 }, { "grad_norm": 0.22326254844665527, "learning_rate": 4.414570835377527e-05, "loss": 0.0263, "step": 44840 }, { "grad_norm": 0.7520406246185303, "learning_rate": 4.412518264302909e-05, "loss": 0.0219, "step": 44850 }, { "grad_norm": 0.32572901248931885, "learning_rate": 4.4104657936128535e-05, "loss": 0.0234, "step": 44860 }, { "grad_norm": 0.3756319582462311, "learning_rate": 4.408413423658071e-05, "loss": 0.0248, "step": 44870 }, { "grad_norm": 0.2541848421096802, "learning_rate": 4.406361154789253e-05, "loss": 0.0224, "step": 44880 }, { "grad_norm": 0.35046544671058655, "learning_rate": 4.4043089873570776e-05, "loss": 0.0225, "step": 44890 }, { "grad_norm": 0.3346552550792694, "learning_rate": 4.4022569217122025e-05, "loss": 0.0188, "step": 44900 }, { "grad_norm": 0.30166539549827576, "learning_rate": 4.400204958205274e-05, "loss": 0.0201, "step": 44910 }, { "grad_norm": 0.3711039423942566, "learning_rate": 4.3981530971869125e-05, "loss": 0.0278, "step": 44920 }, { "grad_norm": 0.2529226243495941, "learning_rate": 4.396101339007727e-05, "loss": 0.0153, "step": 44930 }, { "grad_norm": 0.6839801669120789, "learning_rate": 4.394049684018304e-05, "loss": 0.024, "step": 44940 }, { "grad_norm": 0.42333123087882996, "learning_rate": 4.3919981325692186e-05, "loss": 0.0227, "step": 44950 }, { "grad_norm": 0.1516130268573761, "learning_rate": 4.389946685011024e-05, "loss": 0.0197, "step": 44960 }, { "grad_norm": 0.20673221349716187, "learning_rate": 4.387895341694255e-05, "loss": 0.0259, "step": 44970 }, { "grad_norm": 0.18285098671913147, "learning_rate": 4.3858441029694324e-05, "loss": 0.0273, "step": 44980 }, { "grad_norm": 0.4187049865722656, "learning_rate": 4.3837929691870527e-05, "loss": 0.0211, "step": 44990 }, { "grad_norm": 0.513529896736145, "learning_rate": 4.381741940697604e-05, "loss": 0.0234, "step": 45000 }, { "grad_norm": 0.3029758632183075, "learning_rate": 4.379691017851547e-05, "loss": 0.0207, "step": 45010 }, { "grad_norm": 0.3835185468196869, "learning_rate": 4.3776402009993304e-05, "loss": 0.0225, "step": 45020 }, { "grad_norm": 0.6342633962631226, "learning_rate": 4.3755894904913794e-05, "loss": 0.0205, "step": 45030 }, { "grad_norm": 0.3827994465827942, "learning_rate": 4.373538886678109e-05, "loss": 0.0202, "step": 45040 }, { "grad_norm": 0.23980192840099335, "learning_rate": 4.371488389909909e-05, "loss": 0.0195, "step": 45050 }, { "grad_norm": 0.2666758894920349, "learning_rate": 4.3694380005371515e-05, "loss": 0.0234, "step": 45060 }, { "grad_norm": 0.2423291951417923, "learning_rate": 4.367387718910196e-05, "loss": 0.0204, "step": 45070 }, { "grad_norm": 0.28315338492393494, "learning_rate": 4.3653375453793764e-05, "loss": 0.0198, "step": 45080 }, { "grad_norm": 0.3872467577457428, "learning_rate": 4.3632874802950136e-05, "loss": 0.0209, "step": 45090 }, { "grad_norm": 0.45211514830589294, "learning_rate": 4.3612375240074034e-05, "loss": 0.0231, "step": 45100 }, { "grad_norm": 0.20721399784088135, "learning_rate": 4.3591876768668325e-05, "loss": 0.021, "step": 45110 }, { "grad_norm": 0.44013431668281555, "learning_rate": 4.3571379392235605e-05, "loss": 0.0266, "step": 45120 }, { "grad_norm": 0.3234816789627075, "learning_rate": 4.3550883114278335e-05, "loss": 0.0195, "step": 45130 }, { "grad_norm": 0.4361172318458557, "learning_rate": 4.353038793829876e-05, "loss": 0.0226, "step": 45140 }, { "grad_norm": 0.29840657114982605, "learning_rate": 4.350989386779891e-05, "loss": 0.017, "step": 45150 }, { "grad_norm": 0.3885846436023712, "learning_rate": 4.3489400906280724e-05, "loss": 0.0317, "step": 45160 }, { "grad_norm": 0.32311227917671204, "learning_rate": 4.3468909057245845e-05, "loss": 0.0234, "step": 45170 }, { "grad_norm": 0.1775510013103485, "learning_rate": 4.3448418324195794e-05, "loss": 0.0239, "step": 45180 }, { "grad_norm": 0.6165237426757812, "learning_rate": 4.342792871063184e-05, "loss": 0.0206, "step": 45190 }, { "grad_norm": 0.38026392459869385, "learning_rate": 4.3407440220055145e-05, "loss": 0.02, "step": 45200 }, { "grad_norm": 0.21583767235279083, "learning_rate": 4.33869528559666e-05, "loss": 0.0168, "step": 45210 }, { "grad_norm": 0.2730981409549713, "learning_rate": 4.336646662186696e-05, "loss": 0.0144, "step": 45220 }, { "grad_norm": 0.23339632153511047, "learning_rate": 4.334598152125672e-05, "loss": 0.0202, "step": 45230 }, { "grad_norm": 0.24992704391479492, "learning_rate": 4.3325497557636276e-05, "loss": 0.0235, "step": 45240 }, { "grad_norm": 0.18813863396644592, "learning_rate": 4.330501473450574e-05, "loss": 0.025, "step": 45250 }, { "grad_norm": 0.28041961789131165, "learning_rate": 4.328453305536507e-05, "loss": 0.0184, "step": 45260 }, { "grad_norm": 0.3921130895614624, "learning_rate": 4.326405252371404e-05, "loss": 0.0233, "step": 45270 }, { "grad_norm": 0.29124513268470764, "learning_rate": 4.324357314305221e-05, "loss": 0.018, "step": 45280 }, { "grad_norm": 0.24246929585933685, "learning_rate": 4.3223094916878945e-05, "loss": 0.0215, "step": 45290 }, { "grad_norm": 0.25663501024246216, "learning_rate": 4.320261784869338e-05, "loss": 0.0302, "step": 45300 }, { "grad_norm": 0.1685113161802292, "learning_rate": 4.318214194199455e-05, "loss": 0.0214, "step": 45310 }, { "grad_norm": 0.1341930478811264, "learning_rate": 4.316166720028118e-05, "loss": 0.015, "step": 45320 }, { "grad_norm": 0.28427496552467346, "learning_rate": 4.3141193627051864e-05, "loss": 0.0177, "step": 45330 }, { "grad_norm": 0.32723933458328247, "learning_rate": 4.312072122580496e-05, "loss": 0.0215, "step": 45340 }, { "grad_norm": 0.22410161793231964, "learning_rate": 4.3100250000038646e-05, "loss": 0.021, "step": 45350 }, { "grad_norm": 0.3605695068836212, "learning_rate": 4.307977995325091e-05, "loss": 0.0342, "step": 45360 }, { "grad_norm": 0.256998747587204, "learning_rate": 4.30593110889395e-05, "loss": 0.0227, "step": 45370 }, { "grad_norm": 0.5134300589561462, "learning_rate": 4.3038843410602016e-05, "loss": 0.0151, "step": 45380 }, { "grad_norm": 0.1829022318124771, "learning_rate": 4.3018376921735774e-05, "loss": 0.021, "step": 45390 }, { "grad_norm": 0.20113135874271393, "learning_rate": 4.299791162583799e-05, "loss": 0.0185, "step": 45400 }, { "grad_norm": 0.2119656503200531, "learning_rate": 4.29774475264056e-05, "loss": 0.0263, "step": 45410 }, { "grad_norm": 0.22402243316173553, "learning_rate": 4.2956984626935365e-05, "loss": 0.0188, "step": 45420 }, { "grad_norm": 0.29144975543022156, "learning_rate": 4.293652293092383e-05, "loss": 0.0243, "step": 45430 }, { "grad_norm": 0.27058377861976624, "learning_rate": 4.2916062441867324e-05, "loss": 0.0183, "step": 45440 }, { "grad_norm": 0.311392605304718, "learning_rate": 4.289560316326201e-05, "loss": 0.027, "step": 45450 }, { "grad_norm": 0.3297133445739746, "learning_rate": 4.28751450986038e-05, "loss": 0.0242, "step": 45460 }, { "grad_norm": 0.5274912714958191, "learning_rate": 4.2854688251388444e-05, "loss": 0.0261, "step": 45470 }, { "grad_norm": 0.23421694338321686, "learning_rate": 4.2834232625111425e-05, "loss": 0.0252, "step": 45480 }, { "grad_norm": 0.5761767029762268, "learning_rate": 4.2813778223268086e-05, "loss": 0.0203, "step": 45490 }, { "grad_norm": 0.4039790630340576, "learning_rate": 4.2793325049353477e-05, "loss": 0.0183, "step": 45500 }, { "grad_norm": 0.2867858111858368, "learning_rate": 4.2772873106862535e-05, "loss": 0.0234, "step": 45510 }, { "grad_norm": 0.22616046667099, "learning_rate": 4.275242239928993e-05, "loss": 0.0186, "step": 45520 }, { "grad_norm": 0.33657756447792053, "learning_rate": 4.273197293013009e-05, "loss": 0.0178, "step": 45530 }, { "grad_norm": 0.19489873945713043, "learning_rate": 4.271152470287731e-05, "loss": 0.0147, "step": 45540 }, { "grad_norm": 0.16755105555057526, "learning_rate": 4.2691077721025594e-05, "loss": 0.0192, "step": 45550 }, { "grad_norm": 0.36388126015663147, "learning_rate": 4.267063198806883e-05, "loss": 0.0199, "step": 45560 }, { "grad_norm": 0.26891082525253296, "learning_rate": 4.2650187507500574e-05, "loss": 0.0194, "step": 45570 }, { "grad_norm": 0.22225579619407654, "learning_rate": 4.2629744282814275e-05, "loss": 0.0133, "step": 45580 }, { "grad_norm": 0.21706078946590424, "learning_rate": 4.2609302317503074e-05, "loss": 0.0206, "step": 45590 }, { "grad_norm": 0.2519666254520416, "learning_rate": 4.258886161505999e-05, "loss": 0.0193, "step": 45600 }, { "grad_norm": 0.20146790146827698, "learning_rate": 4.2568422178977775e-05, "loss": 0.0198, "step": 45610 }, { "grad_norm": 0.26153838634490967, "learning_rate": 4.254798401274894e-05, "loss": 0.0266, "step": 45620 }, { "grad_norm": 0.36160555481910706, "learning_rate": 4.252754711986583e-05, "loss": 0.0225, "step": 45630 }, { "grad_norm": 0.20544469356536865, "learning_rate": 4.250711150382052e-05, "loss": 0.019, "step": 45640 }, { "grad_norm": 0.2998218238353729, "learning_rate": 4.248667716810495e-05, "loss": 0.0188, "step": 45650 }, { "grad_norm": 0.43014225363731384, "learning_rate": 4.246624411621074e-05, "loss": 0.0169, "step": 45660 }, { "grad_norm": 0.3809720575809479, "learning_rate": 4.244581235162938e-05, "loss": 0.0225, "step": 45670 }, { "grad_norm": 0.1487596482038498, "learning_rate": 4.2425381877852075e-05, "loss": 0.0227, "step": 45680 }, { "grad_norm": 0.37125295400619507, "learning_rate": 4.2404952698369856e-05, "loss": 0.0212, "step": 45690 }, { "grad_norm": 0.2709335386753082, "learning_rate": 4.238452481667349e-05, "loss": 0.0228, "step": 45700 }, { "grad_norm": 0.41702666878700256, "learning_rate": 4.2364098236253526e-05, "loss": 0.0267, "step": 45710 }, { "grad_norm": 0.25318431854248047, "learning_rate": 4.2343672960600356e-05, "loss": 0.0178, "step": 45720 }, { "grad_norm": 0.27784985303878784, "learning_rate": 4.232324899320406e-05, "loss": 0.0157, "step": 45730 }, { "grad_norm": 0.3097398281097412, "learning_rate": 4.230282633755457e-05, "loss": 0.0206, "step": 45740 }, { "grad_norm": 0.21556933224201202, "learning_rate": 4.2282404997141515e-05, "loss": 0.0167, "step": 45750 }, { "grad_norm": 0.22062833607196808, "learning_rate": 4.2261984975454397e-05, "loss": 0.0237, "step": 45760 }, { "grad_norm": 0.3011507987976074, "learning_rate": 4.224156627598239e-05, "loss": 0.0167, "step": 45770 }, { "grad_norm": 0.45595255494117737, "learning_rate": 4.222114890221453e-05, "loss": 0.022, "step": 45780 }, { "grad_norm": 0.4389505982398987, "learning_rate": 4.2200732857639546e-05, "loss": 0.0244, "step": 45790 }, { "grad_norm": 0.2006371021270752, "learning_rate": 4.2180318145746035e-05, "loss": 0.0212, "step": 45800 }, { "grad_norm": 0.34507670998573303, "learning_rate": 4.215990477002227e-05, "loss": 0.0248, "step": 45810 }, { "grad_norm": 0.23867227137088776, "learning_rate": 4.2139492733956356e-05, "loss": 0.0172, "step": 45820 }, { "grad_norm": 0.4074862003326416, "learning_rate": 4.211908204103615e-05, "loss": 0.0304, "step": 45830 }, { "grad_norm": 0.3418256342411041, "learning_rate": 4.2098672694749265e-05, "loss": 0.0333, "step": 45840 }, { "grad_norm": 0.32711997628211975, "learning_rate": 4.2078264698583133e-05, "loss": 0.0217, "step": 45850 }, { "grad_norm": 0.237388476729393, "learning_rate": 4.205785805602488e-05, "loss": 0.0208, "step": 45860 }, { "grad_norm": 0.33827126026153564, "learning_rate": 4.203745277056149e-05, "loss": 0.0229, "step": 45870 }, { "grad_norm": 0.19164833426475525, "learning_rate": 4.201704884567964e-05, "loss": 0.0228, "step": 45880 }, { "grad_norm": 0.23498877882957458, "learning_rate": 4.1996646284865816e-05, "loss": 0.0302, "step": 45890 }, { "grad_norm": 0.29326924681663513, "learning_rate": 4.197624509160625e-05, "loss": 0.018, "step": 45900 }, { "grad_norm": 0.1716953068971634, "learning_rate": 4.195584526938692e-05, "loss": 0.0157, "step": 45910 }, { "grad_norm": 0.3194812834262848, "learning_rate": 4.193544682169365e-05, "loss": 0.0219, "step": 45920 }, { "grad_norm": 0.2885892391204834, "learning_rate": 4.1915049752011946e-05, "loss": 0.0165, "step": 45930 }, { "grad_norm": 0.27906936407089233, "learning_rate": 4.189465406382712e-05, "loss": 0.0185, "step": 45940 }, { "grad_norm": 0.24802477657794952, "learning_rate": 4.187425976062422e-05, "loss": 0.0229, "step": 45950 }, { "grad_norm": 0.519863486289978, "learning_rate": 4.18538668458881e-05, "loss": 0.0219, "step": 45960 }, { "grad_norm": 0.24355237185955048, "learning_rate": 4.183347532310333e-05, "loss": 0.0187, "step": 45970 }, { "grad_norm": 0.1706441044807434, "learning_rate": 4.181308519575429e-05, "loss": 0.0181, "step": 45980 }, { "grad_norm": 0.2605689465999603, "learning_rate": 4.179269646732507e-05, "loss": 0.0297, "step": 45990 }, { "grad_norm": 0.37100133299827576, "learning_rate": 4.177230914129954e-05, "loss": 0.017, "step": 46000 }, { "grad_norm": 0.20804224908351898, "learning_rate": 4.175192322116136e-05, "loss": 0.0173, "step": 46010 }, { "grad_norm": 0.24754443764686584, "learning_rate": 4.173153871039391e-05, "loss": 0.0175, "step": 46020 }, { "grad_norm": 0.35128480195999146, "learning_rate": 4.171115561248036e-05, "loss": 0.0229, "step": 46030 }, { "grad_norm": 0.1516474336385727, "learning_rate": 4.16907739309036e-05, "loss": 0.0149, "step": 46040 }, { "grad_norm": 0.22916467487812042, "learning_rate": 4.167039366914633e-05, "loss": 0.018, "step": 46050 }, { "grad_norm": 0.23210081458091736, "learning_rate": 4.165001483069096e-05, "loss": 0.0224, "step": 46060 }, { "grad_norm": 0.257956326007843, "learning_rate": 4.16296374190197e-05, "loss": 0.0268, "step": 46070 }, { "grad_norm": 0.2067098766565323, "learning_rate": 4.1609261437614464e-05, "loss": 0.0201, "step": 46080 }, { "grad_norm": 0.49158817529678345, "learning_rate": 4.158888688995696e-05, "loss": 0.0268, "step": 46090 }, { "grad_norm": 0.40601328015327454, "learning_rate": 4.1568513779528645e-05, "loss": 0.0184, "step": 46100 }, { "grad_norm": 0.30218642950057983, "learning_rate": 4.1548142109810704e-05, "loss": 0.0274, "step": 46110 }, { "grad_norm": 0.15242835879325867, "learning_rate": 4.152777188428414e-05, "loss": 0.0177, "step": 46120 }, { "grad_norm": 0.24297083914279938, "learning_rate": 4.1507403106429646e-05, "loss": 0.0194, "step": 46130 }, { "grad_norm": 0.27515196800231934, "learning_rate": 4.148703577972768e-05, "loss": 0.0176, "step": 46140 }, { "grad_norm": 0.497987300157547, "learning_rate": 4.146666990765846e-05, "loss": 0.0216, "step": 46150 }, { "grad_norm": 0.23008202016353607, "learning_rate": 4.1446305493701986e-05, "loss": 0.0183, "step": 46160 }, { "grad_norm": 0.31021741032600403, "learning_rate": 4.142594254133796e-05, "loss": 0.0196, "step": 46170 }, { "grad_norm": 0.37429389357566833, "learning_rate": 4.1405581054045835e-05, "loss": 0.0221, "step": 46180 }, { "grad_norm": 0.3299601078033447, "learning_rate": 4.1385221035304864e-05, "loss": 0.0182, "step": 46190 }, { "grad_norm": 0.3431588113307953, "learning_rate": 4.136486248859398e-05, "loss": 0.0213, "step": 46200 }, { "grad_norm": 0.4653676450252533, "learning_rate": 4.1344505417391955e-05, "loss": 0.0194, "step": 46210 }, { "grad_norm": 0.2300666719675064, "learning_rate": 4.132414982517721e-05, "loss": 0.0195, "step": 46220 }, { "grad_norm": 0.18015724420547485, "learning_rate": 4.130379571542798e-05, "loss": 0.0205, "step": 46230 }, { "grad_norm": 0.19073215126991272, "learning_rate": 4.12834430916222e-05, "loss": 0.0192, "step": 46240 }, { "grad_norm": 0.3103136122226715, "learning_rate": 4.126309195723763e-05, "loss": 0.0175, "step": 46250 }, { "grad_norm": 0.2189653217792511, "learning_rate": 4.124274231575168e-05, "loss": 0.0222, "step": 46260 }, { "grad_norm": 0.22118723392486572, "learning_rate": 4.122239417064154e-05, "loss": 0.0149, "step": 46270 }, { "grad_norm": 0.1676354855298996, "learning_rate": 4.1202047525384184e-05, "loss": 0.0167, "step": 46280 }, { "grad_norm": 0.2674920856952667, "learning_rate": 4.118170238345627e-05, "loss": 0.0189, "step": 46290 }, { "grad_norm": 0.3032902479171753, "learning_rate": 4.1161358748334256e-05, "loss": 0.023, "step": 46300 }, { "grad_norm": 0.18369124829769135, "learning_rate": 4.1141016623494266e-05, "loss": 0.0206, "step": 46310 }, { "grad_norm": 0.3251674175262451, "learning_rate": 4.112067601241227e-05, "loss": 0.0181, "step": 46320 }, { "grad_norm": 0.21916943788528442, "learning_rate": 4.110033691856387e-05, "loss": 0.0201, "step": 46330 }, { "grad_norm": 0.46683764457702637, "learning_rate": 4.107999934542451e-05, "loss": 0.0305, "step": 46340 }, { "grad_norm": 0.25738248229026794, "learning_rate": 4.105966329646928e-05, "loss": 0.0233, "step": 46350 }, { "grad_norm": 0.2443055659532547, "learning_rate": 4.103932877517308e-05, "loss": 0.0239, "step": 46360 }, { "grad_norm": 0.19260504841804504, "learning_rate": 4.101899578501052e-05, "loss": 0.022, "step": 46370 }, { "grad_norm": 0.2585577070713043, "learning_rate": 4.099866432945595e-05, "loss": 0.0194, "step": 46380 }, { "grad_norm": 0.34250888228416443, "learning_rate": 4.097833441198346e-05, "loss": 0.0208, "step": 46390 }, { "grad_norm": 0.15688559412956238, "learning_rate": 4.095800603606685e-05, "loss": 0.0221, "step": 46400 }, { "grad_norm": 0.29173406958580017, "learning_rate": 4.093767920517975e-05, "loss": 0.0209, "step": 46410 }, { "grad_norm": 0.2880520224571228, "learning_rate": 4.091735392279539e-05, "loss": 0.0202, "step": 46420 }, { "grad_norm": 0.1935461312532425, "learning_rate": 4.089703019238685e-05, "loss": 0.0184, "step": 46430 }, { "grad_norm": 0.3274417817592621, "learning_rate": 4.0876708017426866e-05, "loss": 0.0164, "step": 46440 }, { "grad_norm": 0.24175526201725006, "learning_rate": 4.085638740138798e-05, "loss": 0.0205, "step": 46450 }, { "grad_norm": 0.28808513283729553, "learning_rate": 4.08360683477424e-05, "loss": 0.0196, "step": 46460 }, { "grad_norm": 0.30259042978286743, "learning_rate": 4.0815750859962085e-05, "loss": 0.0211, "step": 46470 }, { "grad_norm": 0.28609269857406616, "learning_rate": 4.079543494151879e-05, "loss": 0.0238, "step": 46480 }, { "grad_norm": 0.3322443664073944, "learning_rate": 4.07751205958839e-05, "loss": 0.0205, "step": 46490 }, { "grad_norm": 0.33777379989624023, "learning_rate": 4.0754807826528615e-05, "loss": 0.0175, "step": 46500 }, { "grad_norm": 0.2860097885131836, "learning_rate": 4.073449663692379e-05, "loss": 0.0237, "step": 46510 }, { "grad_norm": 0.2065737545490265, "learning_rate": 4.07141870305401e-05, "loss": 0.0164, "step": 46520 }, { "grad_norm": 0.2823200523853302, "learning_rate": 4.0693879010847866e-05, "loss": 0.0267, "step": 46530 }, { "grad_norm": 0.36399364471435547, "learning_rate": 4.06735725813172e-05, "loss": 0.0184, "step": 46540 }, { "grad_norm": 0.35633954405784607, "learning_rate": 4.0653267745417903e-05, "loss": 0.0159, "step": 46550 }, { "grad_norm": 0.23350931704044342, "learning_rate": 4.063296450661949e-05, "loss": 0.0204, "step": 46560 }, { "grad_norm": 0.15358856320381165, "learning_rate": 4.061266286839128e-05, "loss": 0.0193, "step": 46570 }, { "grad_norm": 0.2693282961845398, "learning_rate": 4.0592362834202225e-05, "loss": 0.0226, "step": 46580 }, { "grad_norm": 0.20840385556221008, "learning_rate": 4.057206440752107e-05, "loss": 0.0204, "step": 46590 }, { "grad_norm": 0.21775279939174652, "learning_rate": 4.0551767591816245e-05, "loss": 0.0174, "step": 46600 }, { "grad_norm": 0.24199910461902618, "learning_rate": 4.0531472390555935e-05, "loss": 0.0237, "step": 46610 }, { "grad_norm": 0.2643130421638489, "learning_rate": 4.051117880720802e-05, "loss": 0.0168, "step": 46620 }, { "grad_norm": 0.2131420522928238, "learning_rate": 4.049088684524015e-05, "loss": 0.0161, "step": 46630 }, { "grad_norm": 0.23159946501255035, "learning_rate": 4.0470596508119636e-05, "loss": 0.0183, "step": 46640 }, { "grad_norm": 0.2473633587360382, "learning_rate": 4.0450307799313524e-05, "loss": 0.0213, "step": 46650 }, { "grad_norm": 0.3235212564468384, "learning_rate": 4.0430020722288656e-05, "loss": 0.0237, "step": 46660 }, { "grad_norm": 0.26302996277809143, "learning_rate": 4.040973528051148e-05, "loss": 0.0243, "step": 46670 }, { "grad_norm": 0.2493060678243637, "learning_rate": 4.038945147744827e-05, "loss": 0.0232, "step": 46680 }, { "grad_norm": 0.19827009737491608, "learning_rate": 4.0369169316564945e-05, "loss": 0.0156, "step": 46690 }, { "grad_norm": 0.4372735023498535, "learning_rate": 4.034888880132718e-05, "loss": 0.023, "step": 46700 }, { "grad_norm": 0.8231554627418518, "learning_rate": 4.032860993520035e-05, "loss": 0.0166, "step": 46710 }, { "grad_norm": 0.2094946950674057, "learning_rate": 4.030833272164959e-05, "loss": 0.025, "step": 46720 }, { "grad_norm": 0.17747294902801514, "learning_rate": 4.0288057164139705e-05, "loss": 0.0234, "step": 46730 }, { "grad_norm": 0.1834513396024704, "learning_rate": 4.0267783266135205e-05, "loss": 0.0142, "step": 46740 }, { "grad_norm": 0.3489164113998413, "learning_rate": 4.024751103110039e-05, "loss": 0.0197, "step": 46750 }, { "grad_norm": 0.19631646573543549, "learning_rate": 4.0227240462499176e-05, "loss": 0.0195, "step": 46760 }, { "grad_norm": 0.16423030197620392, "learning_rate": 4.020697156379531e-05, "loss": 0.0193, "step": 46770 }, { "grad_norm": 0.28257521986961365, "learning_rate": 4.018670433845215e-05, "loss": 0.0221, "step": 46780 }, { "grad_norm": 0.1865534633398056, "learning_rate": 4.016643878993284e-05, "loss": 0.0176, "step": 46790 }, { "grad_norm": 0.3590641915798187, "learning_rate": 4.014617492170017e-05, "loss": 0.0228, "step": 46800 }, { "grad_norm": 0.2862389385700226, "learning_rate": 4.0125912737216726e-05, "loss": 0.0326, "step": 46810 }, { "grad_norm": 0.28663522005081177, "learning_rate": 4.0105652239944735e-05, "loss": 0.0215, "step": 46820 }, { "grad_norm": 0.25093626976013184, "learning_rate": 4.0085393433346144e-05, "loss": 0.0219, "step": 46830 }, { "grad_norm": 0.23056821525096893, "learning_rate": 4.006513632088268e-05, "loss": 0.0164, "step": 46840 }, { "grad_norm": 0.23267467319965363, "learning_rate": 4.004488090601567e-05, "loss": 0.0146, "step": 46850 }, { "grad_norm": 0.18617874383926392, "learning_rate": 4.002462719220626e-05, "loss": 0.0182, "step": 46860 }, { "grad_norm": 0.17689639329910278, "learning_rate": 4.000437518291522e-05, "loss": 0.0228, "step": 46870 }, { "grad_norm": 0.18415164947509766, "learning_rate": 3.9984124881603094e-05, "loss": 0.0197, "step": 46880 }, { "grad_norm": 0.31862884759902954, "learning_rate": 3.9963876291730086e-05, "loss": 0.0219, "step": 46890 }, { "grad_norm": 0.22341080009937286, "learning_rate": 3.994362941675614e-05, "loss": 0.019, "step": 46900 }, { "grad_norm": 0.3093947768211365, "learning_rate": 3.992338426014088e-05, "loss": 0.0223, "step": 46910 }, { "grad_norm": 0.5919180512428284, "learning_rate": 3.9903140825343636e-05, "loss": 0.0236, "step": 46920 }, { "grad_norm": 0.49334195256233215, "learning_rate": 3.98828991158235e-05, "loss": 0.0207, "step": 46930 }, { "grad_norm": 0.25943616032600403, "learning_rate": 3.9862659135039185e-05, "loss": 0.0235, "step": 46940 }, { "grad_norm": 0.20878857374191284, "learning_rate": 3.984242088644918e-05, "loss": 0.02, "step": 46950 }, { "grad_norm": 0.30998823046684265, "learning_rate": 3.9822184373511615e-05, "loss": 0.0207, "step": 46960 }, { "grad_norm": 0.22992998361587524, "learning_rate": 3.98019495996844e-05, "loss": 0.0256, "step": 46970 }, { "grad_norm": 0.15919454395771027, "learning_rate": 3.978171656842507e-05, "loss": 0.0197, "step": 46980 }, { "grad_norm": 0.2774263024330139, "learning_rate": 3.976148528319091e-05, "loss": 0.0182, "step": 46990 }, { "grad_norm": 0.26554733514785767, "learning_rate": 3.974125574743888e-05, "loss": 0.0199, "step": 47000 }, { "grad_norm": 0.41584116220474243, "learning_rate": 3.9721027964625686e-05, "loss": 0.0224, "step": 47010 }, { "grad_norm": 0.24803318083286285, "learning_rate": 3.9700801938207676e-05, "loss": 0.019, "step": 47020 }, { "grad_norm": 0.32759159803390503, "learning_rate": 3.9680577671640916e-05, "loss": 0.0253, "step": 47030 }, { "grad_norm": 0.21381200850009918, "learning_rate": 3.966035516838121e-05, "loss": 0.0287, "step": 47040 }, { "grad_norm": 0.3494032323360443, "learning_rate": 3.9640134431884014e-05, "loss": 0.0218, "step": 47050 }, { "grad_norm": 0.23360757529735565, "learning_rate": 3.961991546560451e-05, "loss": 0.021, "step": 47060 }, { "grad_norm": 0.2309335619211197, "learning_rate": 3.959969827299753e-05, "loss": 0.0272, "step": 47070 }, { "grad_norm": 0.3329469561576843, "learning_rate": 3.9579482857517684e-05, "loss": 0.0205, "step": 47080 }, { "grad_norm": 0.44861021637916565, "learning_rate": 3.955926922261921e-05, "loss": 0.0191, "step": 47090 }, { "grad_norm": 0.49628931283950806, "learning_rate": 3.9539057371756084e-05, "loss": 0.0226, "step": 47100 }, { "grad_norm": 0.2715901732444763, "learning_rate": 3.951884730838195e-05, "loss": 0.0142, "step": 47110 }, { "grad_norm": 0.4270801246166229, "learning_rate": 3.949863903595012e-05, "loss": 0.0175, "step": 47120 }, { "grad_norm": 0.2964761555194855, "learning_rate": 3.947843255791369e-05, "loss": 0.0189, "step": 47130 }, { "grad_norm": 0.20485414564609528, "learning_rate": 3.9458227877725364e-05, "loss": 0.0234, "step": 47140 }, { "grad_norm": 0.7964703440666199, "learning_rate": 3.943802499883758e-05, "loss": 0.0152, "step": 47150 }, { "grad_norm": 0.3415299952030182, "learning_rate": 3.9417823924702437e-05, "loss": 0.0221, "step": 47160 }, { "grad_norm": 0.3825627565383911, "learning_rate": 3.939762465877178e-05, "loss": 0.0217, "step": 47170 }, { "grad_norm": 0.2824743688106537, "learning_rate": 3.937742720449708e-05, "loss": 0.0215, "step": 47180 }, { "grad_norm": 0.2754804193973541, "learning_rate": 3.9357231565329563e-05, "loss": 0.0199, "step": 47190 }, { "grad_norm": 0.28526434302330017, "learning_rate": 3.933703774472008e-05, "loss": 0.0176, "step": 47200 }, { "grad_norm": 0.31319528818130493, "learning_rate": 3.93168457461192e-05, "loss": 0.0171, "step": 47210 }, { "grad_norm": 0.35447147488594055, "learning_rate": 3.9296655572977216e-05, "loss": 0.0209, "step": 47220 }, { "grad_norm": 0.34195056557655334, "learning_rate": 3.927646722874404e-05, "loss": 0.0225, "step": 47230 }, { "grad_norm": 0.19392970204353333, "learning_rate": 3.925628071686934e-05, "loss": 0.017, "step": 47240 }, { "grad_norm": 0.24040156602859497, "learning_rate": 3.9236096040802415e-05, "loss": 0.0208, "step": 47250 }, { "grad_norm": 0.5149526000022888, "learning_rate": 3.9215913203992294e-05, "loss": 0.028, "step": 47260 }, { "grad_norm": 0.2941279709339142, "learning_rate": 3.9195732209887645e-05, "loss": 0.0199, "step": 47270 }, { "grad_norm": 0.194162517786026, "learning_rate": 3.9175553061936875e-05, "loss": 0.0193, "step": 47280 }, { "grad_norm": 0.16740728914737701, "learning_rate": 3.9155375763588045e-05, "loss": 0.0229, "step": 47290 }, { "grad_norm": 0.22070394456386566, "learning_rate": 3.913520031828889e-05, "loss": 0.0182, "step": 47300 }, { "grad_norm": 0.29226362705230713, "learning_rate": 3.911502672948685e-05, "loss": 0.0179, "step": 47310 }, { "grad_norm": 0.29873326420783997, "learning_rate": 3.9094855000629014e-05, "loss": 0.0199, "step": 47320 }, { "grad_norm": 0.21962741017341614, "learning_rate": 3.907468513516223e-05, "loss": 0.0159, "step": 47330 }, { "grad_norm": 0.21988986432552338, "learning_rate": 3.905451713653294e-05, "loss": 0.0166, "step": 47340 }, { "grad_norm": 0.27829739451408386, "learning_rate": 3.903435100818731e-05, "loss": 0.0159, "step": 47350 }, { "grad_norm": 0.7232052087783813, "learning_rate": 3.901418675357117e-05, "loss": 0.0206, "step": 47360 }, { "grad_norm": 0.33206093311309814, "learning_rate": 3.8994024376130075e-05, "loss": 0.0277, "step": 47370 }, { "grad_norm": 0.2372271716594696, "learning_rate": 3.8973863879309194e-05, "loss": 0.0207, "step": 47380 }, { "grad_norm": 0.33479785919189453, "learning_rate": 3.8953705266553394e-05, "loss": 0.0194, "step": 47390 }, { "grad_norm": 0.3895012140274048, "learning_rate": 3.893354854130727e-05, "loss": 0.0227, "step": 47400 }, { "grad_norm": 0.3562539219856262, "learning_rate": 3.8913393707015006e-05, "loss": 0.0203, "step": 47410 }, { "grad_norm": 0.26660892367362976, "learning_rate": 3.889324076712056e-05, "loss": 0.0231, "step": 47420 }, { "grad_norm": 0.2293783724308014, "learning_rate": 3.8873089725067476e-05, "loss": 0.0175, "step": 47430 }, { "grad_norm": 0.3507387042045593, "learning_rate": 3.885294058429905e-05, "loss": 0.0257, "step": 47440 }, { "grad_norm": 0.17039279639720917, "learning_rate": 3.8832793348258206e-05, "loss": 0.0202, "step": 47450 }, { "grad_norm": 0.24923326075077057, "learning_rate": 3.881264802038756e-05, "loss": 0.0359, "step": 47460 }, { "grad_norm": 0.3263435363769531, "learning_rate": 3.879250460412939e-05, "loss": 0.0223, "step": 47470 }, { "grad_norm": 0.2476179152727127, "learning_rate": 3.8772363102925644e-05, "loss": 0.0206, "step": 47480 }, { "grad_norm": 0.4011671543121338, "learning_rate": 3.875222352021798e-05, "loss": 0.0176, "step": 47490 }, { "grad_norm": 0.2035774290561676, "learning_rate": 3.8732085859447686e-05, "loss": 0.0202, "step": 47500 }, { "grad_norm": 0.26364564895629883, "learning_rate": 3.871195012405575e-05, "loss": 0.019, "step": 47510 }, { "grad_norm": 0.30441632866859436, "learning_rate": 3.869181631748278e-05, "loss": 0.0228, "step": 47520 }, { "grad_norm": 0.4418538510799408, "learning_rate": 3.867168444316915e-05, "loss": 0.0184, "step": 47530 }, { "grad_norm": 0.1516626477241516, "learning_rate": 3.865155450455481e-05, "loss": 0.0154, "step": 47540 }, { "grad_norm": 0.24392041563987732, "learning_rate": 3.8631426505079426e-05, "loss": 0.0258, "step": 47550 }, { "grad_norm": 0.36824074387550354, "learning_rate": 3.8611300448182304e-05, "loss": 0.0222, "step": 47560 }, { "grad_norm": 0.23271416127681732, "learning_rate": 3.859117633730248e-05, "loss": 0.0153, "step": 47570 }, { "grad_norm": 0.37103283405303955, "learning_rate": 3.857105417587858e-05, "loss": 0.0195, "step": 47580 }, { "grad_norm": 0.2786635458469391, "learning_rate": 3.855093396734894e-05, "loss": 0.0155, "step": 47590 }, { "grad_norm": 0.2343485802412033, "learning_rate": 3.8530815715151545e-05, "loss": 0.0312, "step": 47600 }, { "grad_norm": 0.3015483617782593, "learning_rate": 3.851069942272405e-05, "loss": 0.0151, "step": 47610 }, { "grad_norm": 0.3612404465675354, "learning_rate": 3.849058509350382e-05, "loss": 0.0191, "step": 47620 }, { "grad_norm": 0.26804232597351074, "learning_rate": 3.8470472730927783e-05, "loss": 0.0232, "step": 47630 }, { "grad_norm": 0.4529186487197876, "learning_rate": 3.845036233843264e-05, "loss": 0.0175, "step": 47640 }, { "grad_norm": 0.19185282289981842, "learning_rate": 3.843025391945469e-05, "loss": 0.019, "step": 47650 }, { "grad_norm": 0.2627488672733307, "learning_rate": 3.841014747742992e-05, "loss": 0.0161, "step": 47660 }, { "grad_norm": 0.33148446679115295, "learning_rate": 3.839004301579397e-05, "loss": 0.0139, "step": 47670 }, { "grad_norm": 0.23478780686855316, "learning_rate": 3.8369940537982097e-05, "loss": 0.019, "step": 47680 }, { "grad_norm": 0.4202609956264496, "learning_rate": 3.834984004742933e-05, "loss": 0.0166, "step": 47690 }, { "grad_norm": 0.5107983946800232, "learning_rate": 3.832974154757026e-05, "loss": 0.0144, "step": 47700 }, { "grad_norm": 0.26500561833381653, "learning_rate": 3.830964504183919e-05, "loss": 0.0184, "step": 47710 }, { "grad_norm": 0.285684734582901, "learning_rate": 3.828955053367003e-05, "loss": 0.0179, "step": 47720 }, { "grad_norm": 0.609041690826416, "learning_rate": 3.826945802649642e-05, "loss": 0.0254, "step": 47730 }, { "grad_norm": 0.20483531057834625, "learning_rate": 3.824936752375159e-05, "loss": 0.0253, "step": 47740 }, { "grad_norm": 0.2886914014816284, "learning_rate": 3.822927902886848e-05, "loss": 0.0167, "step": 47750 }, { "grad_norm": 0.3493245840072632, "learning_rate": 3.8209192545279653e-05, "loss": 0.0201, "step": 47760 }, { "grad_norm": 0.2057337611913681, "learning_rate": 3.8189108076417326e-05, "loss": 0.0207, "step": 47770 }, { "grad_norm": 0.262077659368515, "learning_rate": 3.816902562571342e-05, "loss": 0.0282, "step": 47780 }, { "grad_norm": 0.2674368917942047, "learning_rate": 3.814894519659944e-05, "loss": 0.0173, "step": 47790 }, { "grad_norm": 0.1902201622724533, "learning_rate": 3.812886679250661e-05, "loss": 0.0159, "step": 47800 }, { "grad_norm": 0.17000547051429749, "learning_rate": 3.810879041686575e-05, "loss": 0.0178, "step": 47810 }, { "grad_norm": 0.44791826605796814, "learning_rate": 3.808871607310741e-05, "loss": 0.0272, "step": 47820 }, { "grad_norm": 0.27362263202667236, "learning_rate": 3.806864376466169e-05, "loss": 0.0243, "step": 47830 }, { "grad_norm": 0.2053656429052353, "learning_rate": 3.804857349495845e-05, "loss": 0.0158, "step": 47840 }, { "grad_norm": 0.4301356375217438, "learning_rate": 3.802850526742713e-05, "loss": 0.0376, "step": 47850 }, { "grad_norm": 0.17899847030639648, "learning_rate": 3.800843908549683e-05, "loss": 0.0143, "step": 47860 }, { "grad_norm": 0.21017400920391083, "learning_rate": 3.7988374952596325e-05, "loss": 0.0134, "step": 47870 }, { "grad_norm": 0.3297625184059143, "learning_rate": 3.7968312872154e-05, "loss": 0.0218, "step": 47880 }, { "grad_norm": 0.3249579071998596, "learning_rate": 3.7948252847597965e-05, "loss": 0.021, "step": 47890 }, { "grad_norm": 0.32824626564979553, "learning_rate": 3.7928194882355885e-05, "loss": 0.0228, "step": 47900 }, { "grad_norm": 0.20731666684150696, "learning_rate": 3.790813897985515e-05, "loss": 0.0162, "step": 47910 }, { "grad_norm": 0.40294450521469116, "learning_rate": 3.7888085143522726e-05, "loss": 0.0236, "step": 47920 }, { "grad_norm": 0.28537482023239136, "learning_rate": 3.7868033376785314e-05, "loss": 0.018, "step": 47930 }, { "grad_norm": 0.5865050554275513, "learning_rate": 3.784798368306919e-05, "loss": 0.018, "step": 47940 }, { "grad_norm": 0.3864937126636505, "learning_rate": 3.782793606580029e-05, "loss": 0.0171, "step": 47950 }, { "grad_norm": 0.20768246054649353, "learning_rate": 3.7807890528404205e-05, "loss": 0.0231, "step": 47960 }, { "grad_norm": 0.3685877025127411, "learning_rate": 3.778784707430616e-05, "loss": 0.0218, "step": 47970 }, { "grad_norm": 0.23742763698101044, "learning_rate": 3.776780570693107e-05, "loss": 0.0151, "step": 47980 }, { "grad_norm": 0.25043678283691406, "learning_rate": 3.774776642970342e-05, "loss": 0.0243, "step": 47990 }, { "grad_norm": 0.19088560342788696, "learning_rate": 3.77277292460474e-05, "loss": 0.0153, "step": 48000 }, { "grad_norm": 0.337083101272583, "learning_rate": 3.770769415938678e-05, "loss": 0.0186, "step": 48010 }, { "grad_norm": 0.42884311079978943, "learning_rate": 3.768766117314506e-05, "loss": 0.0249, "step": 48020 }, { "grad_norm": 0.56043541431427, "learning_rate": 3.76676302907453e-05, "loss": 0.025, "step": 48030 }, { "grad_norm": 0.8788143992424011, "learning_rate": 3.764760151561021e-05, "loss": 0.0157, "step": 48040 }, { "grad_norm": 0.24567173421382904, "learning_rate": 3.76275748511622e-05, "loss": 0.0144, "step": 48050 }, { "grad_norm": 0.3592418432235718, "learning_rate": 3.7607550300823255e-05, "loss": 0.0184, "step": 48060 }, { "grad_norm": 0.26193955540657043, "learning_rate": 3.7587527868015044e-05, "loss": 0.0261, "step": 48070 }, { "grad_norm": 0.24557313323020935, "learning_rate": 3.756750755615881e-05, "loss": 0.0198, "step": 48080 }, { "grad_norm": 0.16954511404037476, "learning_rate": 3.754748936867553e-05, "loss": 0.0189, "step": 48090 }, { "grad_norm": 0.1991887390613556, "learning_rate": 3.752747330898573e-05, "loss": 0.0184, "step": 48100 }, { "grad_norm": 0.19836953282356262, "learning_rate": 3.750745938050962e-05, "loss": 0.0187, "step": 48110 }, { "grad_norm": 0.271129310131073, "learning_rate": 3.7487447586667025e-05, "loss": 0.0197, "step": 48120 }, { "grad_norm": 0.24343064427375793, "learning_rate": 3.7467437930877426e-05, "loss": 0.0147, "step": 48130 }, { "grad_norm": 0.3107084035873413, "learning_rate": 3.744743041655992e-05, "loss": 0.0182, "step": 48140 }, { "grad_norm": 0.2056213915348053, "learning_rate": 3.742742504713324e-05, "loss": 0.0212, "step": 48150 }, { "grad_norm": 0.24897953867912292, "learning_rate": 3.740742182601576e-05, "loss": 0.0202, "step": 48160 }, { "grad_norm": 0.20741204917430878, "learning_rate": 3.7387420756625464e-05, "loss": 0.0201, "step": 48170 }, { "grad_norm": 0.3631853759288788, "learning_rate": 3.736742184238002e-05, "loss": 0.0193, "step": 48180 }, { "grad_norm": 0.22120758891105652, "learning_rate": 3.7347425086696684e-05, "loss": 0.0202, "step": 48190 }, { "grad_norm": 0.3598904311656952, "learning_rate": 3.732743049299235e-05, "loss": 0.0198, "step": 48200 }, { "grad_norm": 0.2228470742702484, "learning_rate": 3.730743806468354e-05, "loss": 0.0185, "step": 48210 }, { "grad_norm": 0.3085968494415283, "learning_rate": 3.7287447805186436e-05, "loss": 0.0175, "step": 48220 }, { "grad_norm": 0.2170453518629074, "learning_rate": 3.726745971791682e-05, "loss": 0.022, "step": 48230 }, { "grad_norm": 0.31032395362854004, "learning_rate": 3.724747380629008e-05, "loss": 0.0279, "step": 48240 }, { "grad_norm": 0.21461212635040283, "learning_rate": 3.72274900737213e-05, "loss": 0.0215, "step": 48250 }, { "grad_norm": 0.2012546807527542, "learning_rate": 3.7207508523625123e-05, "loss": 0.0163, "step": 48260 }, { "grad_norm": 0.23023511469364166, "learning_rate": 3.718752915941588e-05, "loss": 0.0181, "step": 48270 }, { "grad_norm": 0.3537990152835846, "learning_rate": 3.7167551984507464e-05, "loss": 0.0178, "step": 48280 }, { "grad_norm": 0.19969072937965393, "learning_rate": 3.714757700231346e-05, "loss": 0.0162, "step": 48290 }, { "grad_norm": 0.27574077248573303, "learning_rate": 3.712760421624703e-05, "loss": 0.0171, "step": 48300 }, { "grad_norm": 0.1868366152048111, "learning_rate": 3.710763362972099e-05, "loss": 0.017, "step": 48310 }, { "grad_norm": 0.3501412868499756, "learning_rate": 3.708766524614774e-05, "loss": 0.0241, "step": 48320 }, { "grad_norm": 0.37274110317230225, "learning_rate": 3.7067699068939335e-05, "loss": 0.0173, "step": 48330 }, { "grad_norm": 0.149724543094635, "learning_rate": 3.704773510150748e-05, "loss": 0.0199, "step": 48340 }, { "grad_norm": 0.3635808825492859, "learning_rate": 3.702777334726344e-05, "loss": 0.0226, "step": 48350 }, { "grad_norm": 0.27701595425605774, "learning_rate": 3.7007813809618164e-05, "loss": 0.0168, "step": 48360 }, { "grad_norm": 0.2858533263206482, "learning_rate": 3.6987856491982145e-05, "loss": 0.0234, "step": 48370 }, { "grad_norm": 0.2630637288093567, "learning_rate": 3.696790139776558e-05, "loss": 0.0207, "step": 48380 }, { "grad_norm": 0.25877252221107483, "learning_rate": 3.6947948530378235e-05, "loss": 0.0235, "step": 48390 }, { "grad_norm": 0.21581226587295532, "learning_rate": 3.6927997893229516e-05, "loss": 0.0183, "step": 48400 }, { "grad_norm": 0.1846637725830078, "learning_rate": 3.690804948972845e-05, "loss": 0.0145, "step": 48410 }, { "grad_norm": 0.20496532320976257, "learning_rate": 3.688810332328363e-05, "loss": 0.0177, "step": 48420 }, { "grad_norm": 0.2538420855998993, "learning_rate": 3.686815939730336e-05, "loss": 0.0207, "step": 48430 }, { "grad_norm": 0.26668277382850647, "learning_rate": 3.684821771519548e-05, "loss": 0.0186, "step": 48440 }, { "grad_norm": 0.21779587864875793, "learning_rate": 3.68282782803675e-05, "loss": 0.0209, "step": 48450 }, { "grad_norm": 0.3007507622241974, "learning_rate": 3.6808341096226504e-05, "loss": 0.0261, "step": 48460 }, { "grad_norm": 0.2811194360256195, "learning_rate": 3.678840616617924e-05, "loss": 0.0195, "step": 48470 }, { "grad_norm": 0.3260117769241333, "learning_rate": 3.6768473493632e-05, "loss": 0.0214, "step": 48480 }, { "grad_norm": 0.35613304376602173, "learning_rate": 3.6748543081990783e-05, "loss": 0.0213, "step": 48490 }, { "grad_norm": 0.36269593238830566, "learning_rate": 3.672861493466112e-05, "loss": 0.015, "step": 48500 }, { "grad_norm": 0.22231917083263397, "learning_rate": 3.670868905504818e-05, "loss": 0.0203, "step": 48510 }, { "grad_norm": 0.4179409146308899, "learning_rate": 3.6688765446556784e-05, "loss": 0.0169, "step": 48520 }, { "grad_norm": 0.29933610558509827, "learning_rate": 3.6668844112591276e-05, "loss": 0.0193, "step": 48530 }, { "grad_norm": 0.3891209363937378, "learning_rate": 3.664892505655573e-05, "loss": 0.0192, "step": 48540 }, { "grad_norm": 0.25143730640411377, "learning_rate": 3.662900828185373e-05, "loss": 0.0217, "step": 48550 }, { "grad_norm": 0.2068774700164795, "learning_rate": 3.6609093791888516e-05, "loss": 0.0182, "step": 48560 }, { "grad_norm": 0.2975204885005951, "learning_rate": 3.658918159006292e-05, "loss": 0.0295, "step": 48570 }, { "grad_norm": 0.3068350553512573, "learning_rate": 3.656927167977942e-05, "loss": 0.0256, "step": 48580 }, { "grad_norm": 0.4785834848880768, "learning_rate": 3.654936406444006e-05, "loss": 0.0181, "step": 48590 }, { "grad_norm": 0.35299384593963623, "learning_rate": 3.65294587474465e-05, "loss": 0.0186, "step": 48600 }, { "grad_norm": 0.24619802832603455, "learning_rate": 3.650955573220002e-05, "loss": 0.0477, "step": 48610 }, { "grad_norm": 0.3230055570602417, "learning_rate": 3.648965502210149e-05, "loss": 0.0177, "step": 48620 }, { "grad_norm": 0.3230606019496918, "learning_rate": 3.646975662055142e-05, "loss": 0.0192, "step": 48630 }, { "grad_norm": 0.3451356589794159, "learning_rate": 3.644986053094987e-05, "loss": 0.0183, "step": 48640 }, { "grad_norm": 0.25463831424713135, "learning_rate": 3.642996675669659e-05, "loss": 0.0206, "step": 48650 }, { "grad_norm": 0.2043651044368744, "learning_rate": 3.641007530119083e-05, "loss": 0.0156, "step": 48660 }, { "grad_norm": 0.2315162718296051, "learning_rate": 3.639018616783153e-05, "loss": 0.02, "step": 48670 }, { "grad_norm": 0.23108038306236267, "learning_rate": 3.637029936001719e-05, "loss": 0.0188, "step": 48680 }, { "grad_norm": 0.21324217319488525, "learning_rate": 3.6350414881145886e-05, "loss": 0.0223, "step": 48690 }, { "grad_norm": 0.22083376348018646, "learning_rate": 3.6330532734615386e-05, "loss": 0.023, "step": 48700 }, { "grad_norm": 0.21281097829341888, "learning_rate": 3.6310652923822975e-05, "loss": 0.0208, "step": 48710 }, { "grad_norm": 0.2547609210014343, "learning_rate": 3.629077545216558e-05, "loss": 0.0191, "step": 48720 }, { "grad_norm": 0.3782709538936615, "learning_rate": 3.627090032303969e-05, "loss": 0.0193, "step": 48730 }, { "grad_norm": 0.22198863327503204, "learning_rate": 3.625102753984146e-05, "loss": 0.0207, "step": 48740 }, { "grad_norm": 0.1899823397397995, "learning_rate": 3.623115710596659e-05, "loss": 0.0166, "step": 48750 }, { "grad_norm": 0.36149507761001587, "learning_rate": 3.6211289024810395e-05, "loss": 0.022, "step": 48760 }, { "grad_norm": 0.31051284074783325, "learning_rate": 3.619142329976777e-05, "loss": 0.0166, "step": 48770 }, { "grad_norm": 0.2628532946109772, "learning_rate": 3.6171559934233247e-05, "loss": 0.0161, "step": 48780 }, { "grad_norm": 0.22369353473186493, "learning_rate": 3.615169893160093e-05, "loss": 0.0147, "step": 48790 }, { "grad_norm": 0.3337535262107849, "learning_rate": 3.61318402952645e-05, "loss": 0.0216, "step": 48800 }, { "grad_norm": 0.40190041065216064, "learning_rate": 3.6111984028617285e-05, "loss": 0.0214, "step": 48810 }, { "grad_norm": 0.2521508038043976, "learning_rate": 3.6092130135052134e-05, "loss": 0.0244, "step": 48820 }, { "grad_norm": 0.21603676676750183, "learning_rate": 3.6072278617961584e-05, "loss": 0.0253, "step": 48830 }, { "grad_norm": 0.3025529682636261, "learning_rate": 3.605242948073767e-05, "loss": 0.0209, "step": 48840 }, { "grad_norm": 0.3296757936477661, "learning_rate": 3.603258272677212e-05, "loss": 0.018, "step": 48850 }, { "grad_norm": 0.25502699613571167, "learning_rate": 3.601273835945616e-05, "loss": 0.0193, "step": 48860 }, { "grad_norm": 0.2587636709213257, "learning_rate": 3.5992896382180664e-05, "loss": 0.019, "step": 48870 }, { "grad_norm": 0.20983156561851501, "learning_rate": 3.597305679833609e-05, "loss": 0.02, "step": 48880 }, { "grad_norm": 0.384685218334198, "learning_rate": 3.595321961131245e-05, "loss": 0.0223, "step": 48890 }, { "grad_norm": 0.28031548857688904, "learning_rate": 3.593338482449942e-05, "loss": 0.0276, "step": 48900 }, { "grad_norm": 0.27749574184417725, "learning_rate": 3.591355244128618e-05, "loss": 0.0294, "step": 48910 }, { "grad_norm": 0.27238309383392334, "learning_rate": 3.589372246506158e-05, "loss": 0.0275, "step": 48920 }, { "grad_norm": 0.24816787242889404, "learning_rate": 3.5873894899213984e-05, "loss": 0.0162, "step": 48930 }, { "grad_norm": 0.31271466612815857, "learning_rate": 3.5854069747131416e-05, "loss": 0.017, "step": 48940 }, { "grad_norm": 0.5237837433815002, "learning_rate": 3.583424701220143e-05, "loss": 0.0297, "step": 48950 }, { "grad_norm": 0.32629087567329407, "learning_rate": 3.581442669781121e-05, "loss": 0.0235, "step": 48960 }, { "grad_norm": 0.3680708408355713, "learning_rate": 3.579460880734749e-05, "loss": 0.0191, "step": 48970 }, { "grad_norm": 0.21198728680610657, "learning_rate": 3.577479334419657e-05, "loss": 0.0191, "step": 48980 }, { "grad_norm": 0.27548980712890625, "learning_rate": 3.575498031174444e-05, "loss": 0.0172, "step": 48990 }, { "grad_norm": 0.304288774728775, "learning_rate": 3.573516971337657e-05, "loss": 0.0195, "step": 49000 }, { "grad_norm": 0.3638042211532593, "learning_rate": 3.5715361552478046e-05, "loss": 0.0225, "step": 49010 }, { "grad_norm": 0.2962365448474884, "learning_rate": 3.5695555832433536e-05, "loss": 0.0166, "step": 49020 }, { "grad_norm": 0.36140522360801697, "learning_rate": 3.5675752556627325e-05, "loss": 0.02, "step": 49030 }, { "grad_norm": 0.3201123774051666, "learning_rate": 3.565595172844322e-05, "loss": 0.0157, "step": 49040 }, { "grad_norm": 0.26458674669265747, "learning_rate": 3.5636153351264666e-05, "loss": 0.0222, "step": 49050 }, { "grad_norm": 0.20890863239765167, "learning_rate": 3.5616357428474655e-05, "loss": 0.021, "step": 49060 }, { "grad_norm": 0.2794489860534668, "learning_rate": 3.559656396345575e-05, "loss": 0.0232, "step": 49070 }, { "grad_norm": 0.20491661131381989, "learning_rate": 3.5576772959590146e-05, "loss": 0.0222, "step": 49080 }, { "grad_norm": 0.23030175268650055, "learning_rate": 3.5556984420259545e-05, "loss": 0.0191, "step": 49090 }, { "grad_norm": 0.242388516664505, "learning_rate": 3.5537198348845305e-05, "loss": 0.0225, "step": 49100 }, { "grad_norm": 0.23172961175441742, "learning_rate": 3.551741474872831e-05, "loss": 0.0228, "step": 49110 }, { "grad_norm": 0.32881391048431396, "learning_rate": 3.549763362328903e-05, "loss": 0.0262, "step": 49120 }, { "grad_norm": 0.27388954162597656, "learning_rate": 3.5477854975907515e-05, "loss": 0.0169, "step": 49130 }, { "grad_norm": 0.18848252296447754, "learning_rate": 3.5458078809963416e-05, "loss": 0.0163, "step": 49140 }, { "grad_norm": 0.24901248514652252, "learning_rate": 3.543830512883594e-05, "loss": 0.0166, "step": 49150 }, { "grad_norm": 0.28908559679985046, "learning_rate": 3.5418533935903824e-05, "loss": 0.02, "step": 49160 }, { "grad_norm": 0.4053391218185425, "learning_rate": 3.539876523454547e-05, "loss": 0.0199, "step": 49170 }, { "grad_norm": 0.3077586889266968, "learning_rate": 3.537899902813878e-05, "loss": 0.0219, "step": 49180 }, { "grad_norm": 0.22510430216789246, "learning_rate": 3.5359235320061293e-05, "loss": 0.0167, "step": 49190 }, { "grad_norm": 0.30339813232421875, "learning_rate": 3.533947411369003e-05, "loss": 0.0161, "step": 49200 }, { "grad_norm": 0.20691923797130585, "learning_rate": 3.53197154124017e-05, "loss": 0.018, "step": 49210 }, { "grad_norm": 0.3934285342693329, "learning_rate": 3.52999592195725e-05, "loss": 0.0209, "step": 49220 }, { "grad_norm": 0.2507574260234833, "learning_rate": 3.5280205538578224e-05, "loss": 0.0185, "step": 49230 }, { "grad_norm": 0.28860050439834595, "learning_rate": 3.5260454372794236e-05, "loss": 0.0162, "step": 49240 }, { "grad_norm": 0.21281658113002777, "learning_rate": 3.524070572559545e-05, "loss": 0.0204, "step": 49250 }, { "grad_norm": 0.2593461871147156, "learning_rate": 3.5220959600356395e-05, "loss": 0.0151, "step": 49260 }, { "grad_norm": 0.2108980119228363, "learning_rate": 3.5201216000451145e-05, "loss": 0.0182, "step": 49270 }, { "grad_norm": 0.25518518686294556, "learning_rate": 3.5181474929253335e-05, "loss": 0.0155, "step": 49280 }, { "grad_norm": 0.2592945396900177, "learning_rate": 3.516173639013615e-05, "loss": 0.0195, "step": 49290 }, { "grad_norm": 0.1309375762939453, "learning_rate": 3.5142000386472406e-05, "loss": 0.0171, "step": 49300 }, { "grad_norm": 0.23451898992061615, "learning_rate": 3.5122266921634427e-05, "loss": 0.0149, "step": 49310 }, { "grad_norm": 0.24766896665096283, "learning_rate": 3.510253599899413e-05, "loss": 0.0168, "step": 49320 }, { "grad_norm": 0.15198656916618347, "learning_rate": 3.5082807621922965e-05, "loss": 0.0132, "step": 49330 }, { "grad_norm": 0.28120261430740356, "learning_rate": 3.506308179379201e-05, "loss": 0.0195, "step": 49340 }, { "grad_norm": 0.17402374744415283, "learning_rate": 3.5043358517971844e-05, "loss": 0.0184, "step": 49350 }, { "grad_norm": 0.22277025878429413, "learning_rate": 3.502363779783264e-05, "loss": 0.0215, "step": 49360 }, { "grad_norm": 0.28930026292800903, "learning_rate": 3.500391963674415e-05, "loss": 0.0163, "step": 49370 }, { "grad_norm": 0.2764851748943329, "learning_rate": 3.4984204038075615e-05, "loss": 0.0233, "step": 49380 }, { "grad_norm": 0.28570061922073364, "learning_rate": 3.496449100519595e-05, "loss": 0.023, "step": 49390 }, { "grad_norm": 0.2462293952703476, "learning_rate": 3.494478054147354e-05, "loss": 0.0173, "step": 49400 }, { "grad_norm": 0.22267603874206543, "learning_rate": 3.4925072650276395e-05, "loss": 0.0153, "step": 49410 }, { "grad_norm": 0.6558098196983337, "learning_rate": 3.4905367334972016e-05, "loss": 0.0169, "step": 49420 }, { "grad_norm": 0.37744250893592834, "learning_rate": 3.488566459892752e-05, "loss": 0.0155, "step": 49430 }, { "grad_norm": 0.4784802198410034, "learning_rate": 3.4865964445509585e-05, "loss": 0.0173, "step": 49440 }, { "grad_norm": 0.2315104603767395, "learning_rate": 3.484626687808438e-05, "loss": 0.0147, "step": 49450 }, { "grad_norm": 0.2276560217142105, "learning_rate": 3.4826571900017735e-05, "loss": 0.0118, "step": 49460 }, { "grad_norm": 0.3055298924446106, "learning_rate": 3.480687951467495e-05, "loss": 0.0149, "step": 49470 }, { "grad_norm": 0.31614354252815247, "learning_rate": 3.4787189725420925e-05, "loss": 0.0151, "step": 49480 }, { "grad_norm": 0.2166633903980255, "learning_rate": 3.4767502535620086e-05, "loss": 0.0184, "step": 49490 }, { "grad_norm": 0.3198692500591278, "learning_rate": 3.474781794863648e-05, "loss": 0.0225, "step": 49500 }, { "grad_norm": 0.3175573945045471, "learning_rate": 3.472813596783363e-05, "loss": 0.0247, "step": 49510 }, { "grad_norm": 0.26406756043434143, "learning_rate": 3.470845659657466e-05, "loss": 0.0185, "step": 49520 }, { "grad_norm": 0.23765763640403748, "learning_rate": 3.468877983822223e-05, "loss": 0.0163, "step": 49530 }, { "grad_norm": 0.22927092015743256, "learning_rate": 3.466910569613855e-05, "loss": 0.0157, "step": 49540 }, { "grad_norm": 0.3478211760520935, "learning_rate": 3.464943417368542e-05, "loss": 0.0203, "step": 49550 }, { "grad_norm": 0.34150710701942444, "learning_rate": 3.462976527422415e-05, "loss": 0.0181, "step": 49560 }, { "grad_norm": 0.2975916266441345, "learning_rate": 3.461009900111562e-05, "loss": 0.0203, "step": 49570 }, { "grad_norm": 0.37523677945137024, "learning_rate": 3.459043535772023e-05, "loss": 0.0192, "step": 49580 }, { "grad_norm": 0.37823930382728577, "learning_rate": 3.4570774347398014e-05, "loss": 0.0233, "step": 49590 }, { "grad_norm": 0.5282583832740784, "learning_rate": 3.4551115973508454e-05, "loss": 0.0249, "step": 49600 }, { "grad_norm": 0.36486712098121643, "learning_rate": 3.453146023941066e-05, "loss": 0.0287, "step": 49610 }, { "grad_norm": 0.2470591962337494, "learning_rate": 3.451180714846325e-05, "loss": 0.0184, "step": 49620 }, { "grad_norm": 0.9411562085151672, "learning_rate": 3.449215670402438e-05, "loss": 0.0236, "step": 49630 }, { "grad_norm": 0.5263985395431519, "learning_rate": 3.447250890945181e-05, "loss": 0.0204, "step": 49640 }, { "grad_norm": 0.29294708371162415, "learning_rate": 3.4452863768102754e-05, "loss": 0.0226, "step": 49650 }, { "grad_norm": 0.22144894301891327, "learning_rate": 3.443322128333409e-05, "loss": 0.0171, "step": 49660 }, { "grad_norm": 0.47862961888313293, "learning_rate": 3.441358145850215e-05, "loss": 0.0208, "step": 49670 }, { "grad_norm": 0.2917853593826294, "learning_rate": 3.439394429696286e-05, "loss": 0.0215, "step": 49680 }, { "grad_norm": 0.270155668258667, "learning_rate": 3.4374309802071644e-05, "loss": 0.0204, "step": 49690 }, { "grad_norm": 0.3524714410305023, "learning_rate": 3.435467797718353e-05, "loss": 0.0165, "step": 49700 }, { "grad_norm": 0.26535749435424805, "learning_rate": 3.433504882565306e-05, "loss": 0.0175, "step": 49710 }, { "grad_norm": 0.32782259583473206, "learning_rate": 3.43154223508343e-05, "loss": 0.0233, "step": 49720 }, { "grad_norm": 0.16748939454555511, "learning_rate": 3.429579855608089e-05, "loss": 0.0158, "step": 49730 }, { "grad_norm": 0.2708004415035248, "learning_rate": 3.427617744474597e-05, "loss": 0.0176, "step": 49740 }, { "grad_norm": 0.2531599998474121, "learning_rate": 3.425655902018231e-05, "loss": 0.0206, "step": 49750 }, { "grad_norm": 0.17703942954540253, "learning_rate": 3.423694328574211e-05, "loss": 0.0197, "step": 49760 }, { "grad_norm": 0.2650673985481262, "learning_rate": 3.42173302447772e-05, "loss": 0.0176, "step": 49770 }, { "grad_norm": 0.15153202414512634, "learning_rate": 3.419771990063886e-05, "loss": 0.0152, "step": 49780 }, { "grad_norm": 0.6797863245010376, "learning_rate": 3.417811225667803e-05, "loss": 0.0157, "step": 49790 }, { "grad_norm": 0.24920549988746643, "learning_rate": 3.415850731624508e-05, "loss": 0.0205, "step": 49800 }, { "grad_norm": 0.27126333117485046, "learning_rate": 3.4138905082689945e-05, "loss": 0.0246, "step": 49810 }, { "grad_norm": 0.296388179063797, "learning_rate": 3.4119305559362145e-05, "loss": 0.0177, "step": 49820 }, { "grad_norm": 0.223341166973114, "learning_rate": 3.4099708749610684e-05, "loss": 0.0151, "step": 49830 }, { "grad_norm": 0.32858356833457947, "learning_rate": 3.408011465678413e-05, "loss": 0.0257, "step": 49840 }, { "grad_norm": 0.21898789703845978, "learning_rate": 3.406052328423055e-05, "loss": 0.0198, "step": 49850 }, { "grad_norm": 0.19537581503391266, "learning_rate": 3.4040934635297615e-05, "loss": 0.0164, "step": 49860 }, { "grad_norm": 0.29519331455230713, "learning_rate": 3.4021348713332466e-05, "loss": 0.0146, "step": 49870 }, { "grad_norm": 0.34421205520629883, "learning_rate": 3.4001765521681807e-05, "loss": 0.0197, "step": 49880 }, { "grad_norm": 0.2675069272518158, "learning_rate": 3.398218506369188e-05, "loss": 0.0308, "step": 49890 }, { "grad_norm": 0.1553792655467987, "learning_rate": 3.3962607342708404e-05, "loss": 0.0152, "step": 49900 }, { "grad_norm": 0.2551923394203186, "learning_rate": 3.394303236207673e-05, "loss": 0.0133, "step": 49910 }, { "grad_norm": 0.2535931468009949, "learning_rate": 3.392346012514166e-05, "loss": 0.0191, "step": 49920 }, { "grad_norm": 0.3395741879940033, "learning_rate": 3.390389063524757e-05, "loss": 0.0174, "step": 49930 }, { "grad_norm": 0.2404567450284958, "learning_rate": 3.3884323895738324e-05, "loss": 0.016, "step": 49940 }, { "grad_norm": 0.25931358337402344, "learning_rate": 3.386475990995738e-05, "loss": 0.02, "step": 49950 }, { "grad_norm": 0.2513774037361145, "learning_rate": 3.384519868124765e-05, "loss": 0.0144, "step": 49960 }, { "grad_norm": 0.27246272563934326, "learning_rate": 3.3825640212951645e-05, "loss": 0.0178, "step": 49970 }, { "grad_norm": 0.43943652510643005, "learning_rate": 3.380608450841134e-05, "loss": 0.019, "step": 49980 }, { "grad_norm": 0.3065723478794098, "learning_rate": 3.3786531570968305e-05, "loss": 0.0253, "step": 49990 }, { "grad_norm": 0.2420298308134079, "learning_rate": 3.3766981403963584e-05, "loss": 0.0241, "step": 50000 }, { "grad_norm": 0.16999465227127075, "learning_rate": 3.374743401073775e-05, "loss": 0.0171, "step": 50010 }, { "grad_norm": 0.17763224244117737, "learning_rate": 3.372788939463095e-05, "loss": 0.0154, "step": 50020 }, { "grad_norm": 0.3021986484527588, "learning_rate": 3.370834755898281e-05, "loss": 0.0138, "step": 50030 }, { "grad_norm": 0.3065021336078644, "learning_rate": 3.3688808507132493e-05, "loss": 0.0203, "step": 50040 }, { "grad_norm": 0.21392923593521118, "learning_rate": 3.3669272242418685e-05, "loss": 0.0241, "step": 50050 }, { "grad_norm": 0.26934126019477844, "learning_rate": 3.364973876817961e-05, "loss": 0.0185, "step": 50060 }, { "grad_norm": 0.2570512890815735, "learning_rate": 3.363020808775299e-05, "loss": 0.0156, "step": 50070 }, { "grad_norm": 0.1515992283821106, "learning_rate": 3.361068020447611e-05, "loss": 0.0166, "step": 50080 }, { "grad_norm": 0.1949201375246048, "learning_rate": 3.3591155121685724e-05, "loss": 0.013, "step": 50090 }, { "grad_norm": 0.302429735660553, "learning_rate": 3.357163284271814e-05, "loss": 0.0187, "step": 50100 }, { "grad_norm": 0.27804964780807495, "learning_rate": 3.355211337090919e-05, "loss": 0.0148, "step": 50110 }, { "grad_norm": 0.2616806924343109, "learning_rate": 3.353259670959421e-05, "loss": 0.0219, "step": 50120 }, { "grad_norm": 0.23102451860904694, "learning_rate": 3.351308286210808e-05, "loss": 0.0185, "step": 50130 }, { "grad_norm": 0.9008920788764954, "learning_rate": 3.3493571831785156e-05, "loss": 0.0169, "step": 50140 }, { "grad_norm": 0.287807822227478, "learning_rate": 3.347406362195936e-05, "loss": 0.015, "step": 50150 }, { "grad_norm": 0.8294472098350525, "learning_rate": 3.345455823596411e-05, "loss": 0.0199, "step": 50160 }, { "grad_norm": 0.2039308398962021, "learning_rate": 3.3435055677132346e-05, "loss": 0.0191, "step": 50170 }, { "grad_norm": 0.35942015051841736, "learning_rate": 3.3415555948796505e-05, "loss": 0.0216, "step": 50180 }, { "grad_norm": 0.37040677666664124, "learning_rate": 3.3396059054288556e-05, "loss": 0.0199, "step": 50190 }, { "grad_norm": 0.3249152600765228, "learning_rate": 3.3376564996940015e-05, "loss": 0.0282, "step": 50200 }, { "grad_norm": 0.22891484200954437, "learning_rate": 3.3357073780081836e-05, "loss": 0.0139, "step": 50210 }, { "grad_norm": 0.19551682472229004, "learning_rate": 3.333758540704459e-05, "loss": 0.0147, "step": 50220 }, { "grad_norm": 0.21435512602329254, "learning_rate": 3.331809988115827e-05, "loss": 0.0141, "step": 50230 }, { "grad_norm": 0.2997051179409027, "learning_rate": 3.329861720575244e-05, "loss": 0.0186, "step": 50240 }, { "grad_norm": 0.213626891374588, "learning_rate": 3.3279137384156126e-05, "loss": 0.0175, "step": 50250 }, { "grad_norm": 0.28126606345176697, "learning_rate": 3.3259660419697934e-05, "loss": 0.0291, "step": 50260 }, { "grad_norm": 0.15581226348876953, "learning_rate": 3.3240186315705926e-05, "loss": 0.0249, "step": 50270 }, { "grad_norm": 0.40081652998924255, "learning_rate": 3.322071507550769e-05, "loss": 0.0277, "step": 50280 }, { "grad_norm": 0.27457234263420105, "learning_rate": 3.320124670243033e-05, "loss": 0.0149, "step": 50290 }, { "grad_norm": 0.19239620864391327, "learning_rate": 3.318178119980045e-05, "loss": 0.0155, "step": 50300 }, { "grad_norm": 0.26964935660362244, "learning_rate": 3.316231857094421e-05, "loss": 0.0151, "step": 50310 }, { "grad_norm": 0.21987703442573547, "learning_rate": 3.31428588191872e-05, "loss": 0.0193, "step": 50320 }, { "grad_norm": 0.3192971646785736, "learning_rate": 3.312340194785458e-05, "loss": 0.0198, "step": 50330 }, { "grad_norm": 0.2807474434375763, "learning_rate": 3.3103947960270975e-05, "loss": 0.0155, "step": 50340 }, { "grad_norm": 0.21626438200473785, "learning_rate": 3.308449685976058e-05, "loss": 0.0147, "step": 50350 }, { "grad_norm": 0.27798911929130554, "learning_rate": 3.3065048649647024e-05, "loss": 0.0191, "step": 50360 }, { "grad_norm": 0.2759127616882324, "learning_rate": 3.304560333325348e-05, "loss": 0.0151, "step": 50370 }, { "grad_norm": 0.34318414330482483, "learning_rate": 3.3026160913902624e-05, "loss": 0.0185, "step": 50380 }, { "grad_norm": 0.38315272331237793, "learning_rate": 3.300672139491662e-05, "loss": 0.0231, "step": 50390 }, { "grad_norm": 0.2247193306684494, "learning_rate": 3.298728477961717e-05, "loss": 0.0199, "step": 50400 }, { "grad_norm": 0.22089709341526031, "learning_rate": 3.296785107132544e-05, "loss": 0.0176, "step": 50410 }, { "grad_norm": 0.6749515533447266, "learning_rate": 3.2948420273362156e-05, "loss": 0.0152, "step": 50420 }, { "grad_norm": 0.17672765254974365, "learning_rate": 3.292899238904747e-05, "loss": 0.0174, "step": 50430 }, { "grad_norm": 0.3844781816005707, "learning_rate": 3.29095674217011e-05, "loss": 0.0236, "step": 50440 }, { "grad_norm": 0.24969857931137085, "learning_rate": 3.289014537464224e-05, "loss": 0.0198, "step": 50450 }, { "grad_norm": 0.2663617730140686, "learning_rate": 3.287072625118955e-05, "loss": 0.0184, "step": 50460 }, { "grad_norm": 0.23476535081863403, "learning_rate": 3.285131005466129e-05, "loss": 0.0179, "step": 50470 }, { "grad_norm": 0.2889226973056793, "learning_rate": 3.2831896788375105e-05, "loss": 0.0191, "step": 50480 }, { "grad_norm": 0.46901389956474304, "learning_rate": 3.281248645564822e-05, "loss": 0.0187, "step": 50490 }, { "grad_norm": 0.19785845279693604, "learning_rate": 3.2793079059797306e-05, "loss": 0.0244, "step": 50500 }, { "grad_norm": 0.2730732262134552, "learning_rate": 3.277367460413859e-05, "loss": 0.0176, "step": 50510 }, { "grad_norm": 0.1748238205909729, "learning_rate": 3.275427309198773e-05, "loss": 0.0219, "step": 50520 }, { "grad_norm": 0.22257867455482483, "learning_rate": 3.273487452665993e-05, "loss": 0.0151, "step": 50530 }, { "grad_norm": 0.3155366778373718, "learning_rate": 3.271547891146986e-05, "loss": 0.0133, "step": 50540 }, { "grad_norm": 0.28440022468566895, "learning_rate": 3.269608624973173e-05, "loss": 0.015, "step": 50550 }, { "grad_norm": 0.2534908354282379, "learning_rate": 3.26766965447592e-05, "loss": 0.0218, "step": 50560 }, { "grad_norm": 0.2836790978908539, "learning_rate": 3.2657309799865424e-05, "loss": 0.0237, "step": 50570 }, { "grad_norm": 0.1693061739206314, "learning_rate": 3.2637926018363084e-05, "loss": 0.0156, "step": 50580 }, { "grad_norm": 0.3541996479034424, "learning_rate": 3.261854520356432e-05, "loss": 0.0205, "step": 50590 }, { "grad_norm": 0.16730539500713348, "learning_rate": 3.2599167358780816e-05, "loss": 0.0216, "step": 50600 }, { "grad_norm": 0.4192500114440918, "learning_rate": 3.2579792487323676e-05, "loss": 0.017, "step": 50610 }, { "grad_norm": 0.19985255599021912, "learning_rate": 3.256042059250358e-05, "loss": 0.0231, "step": 50620 }, { "grad_norm": 0.5145930647850037, "learning_rate": 3.254105167763062e-05, "loss": 0.0202, "step": 50630 }, { "grad_norm": 0.20472273230552673, "learning_rate": 3.252168574601443e-05, "loss": 0.0153, "step": 50640 }, { "grad_norm": 0.2789583206176758, "learning_rate": 3.2502322800964115e-05, "loss": 0.0189, "step": 50650 }, { "grad_norm": 0.4231075346469879, "learning_rate": 3.248296284578825e-05, "loss": 0.0155, "step": 50660 }, { "grad_norm": 0.43391916155815125, "learning_rate": 3.246360588379497e-05, "loss": 0.0237, "step": 50670 }, { "grad_norm": 0.22080180048942566, "learning_rate": 3.24442519182918e-05, "loss": 0.0277, "step": 50680 }, { "grad_norm": 0.22134794294834137, "learning_rate": 3.2424900952585845e-05, "loss": 0.0159, "step": 50690 }, { "grad_norm": 0.30221959948539734, "learning_rate": 3.240555298998362e-05, "loss": 0.0203, "step": 50700 }, { "grad_norm": 0.26165467500686646, "learning_rate": 3.2386208033791194e-05, "loss": 0.0175, "step": 50710 }, { "grad_norm": 0.3417763113975525, "learning_rate": 3.2366866087314064e-05, "loss": 0.0196, "step": 50720 }, { "grad_norm": 0.31871142983436584, "learning_rate": 3.234752715385727e-05, "loss": 0.0251, "step": 50730 }, { "grad_norm": 0.2616446614265442, "learning_rate": 3.232819123672529e-05, "loss": 0.0224, "step": 50740 }, { "grad_norm": 0.4432082772254944, "learning_rate": 3.2308858339222076e-05, "loss": 0.0269, "step": 50750 }, { "grad_norm": 0.344460129737854, "learning_rate": 3.2289528464651144e-05, "loss": 0.0215, "step": 50760 }, { "grad_norm": 0.2271946370601654, "learning_rate": 3.22702016163154e-05, "loss": 0.0174, "step": 50770 }, { "grad_norm": 0.30828654766082764, "learning_rate": 3.2250877797517295e-05, "loss": 0.025, "step": 50780 }, { "grad_norm": 0.2185600847005844, "learning_rate": 3.223155701155872e-05, "loss": 0.0177, "step": 50790 }, { "grad_norm": 0.22466441988945007, "learning_rate": 3.2212239261741105e-05, "loss": 0.0205, "step": 50800 }, { "grad_norm": 0.18379154801368713, "learning_rate": 3.219292455136528e-05, "loss": 0.0163, "step": 50810 }, { "grad_norm": 0.3163118064403534, "learning_rate": 3.217361288373165e-05, "loss": 0.0174, "step": 50820 }, { "grad_norm": 0.1487226039171219, "learning_rate": 3.215430426214002e-05, "loss": 0.0191, "step": 50830 }, { "grad_norm": 0.33028027415275574, "learning_rate": 3.2134998689889705e-05, "loss": 0.0186, "step": 50840 }, { "grad_norm": 0.4651772975921631, "learning_rate": 3.211569617027952e-05, "loss": 0.0189, "step": 50850 }, { "grad_norm": 0.24338878691196442, "learning_rate": 3.209639670660769e-05, "loss": 0.0191, "step": 50860 }, { "grad_norm": 0.6757581830024719, "learning_rate": 3.207710030217202e-05, "loss": 0.0181, "step": 50870 }, { "grad_norm": 0.5519312024116516, "learning_rate": 3.205780696026972e-05, "loss": 0.0191, "step": 50880 }, { "grad_norm": 0.17519047856330872, "learning_rate": 3.203851668419749e-05, "loss": 0.015, "step": 50890 }, { "grad_norm": 0.23329481482505798, "learning_rate": 3.201922947725149e-05, "loss": 0.017, "step": 50900 }, { "grad_norm": 0.23075953125953674, "learning_rate": 3.199994534272742e-05, "loss": 0.0178, "step": 50910 }, { "grad_norm": 0.24090828001499176, "learning_rate": 3.1980664283920394e-05, "loss": 0.0211, "step": 50920 }, { "grad_norm": 0.23930829763412476, "learning_rate": 3.196138630412499e-05, "loss": 0.0124, "step": 50930 }, { "grad_norm": 0.32550328969955444, "learning_rate": 3.194211140663532e-05, "loss": 0.0182, "step": 50940 }, { "grad_norm": 0.3025106191635132, "learning_rate": 3.1922839594744914e-05, "loss": 0.023, "step": 50950 }, { "grad_norm": 0.31285443902015686, "learning_rate": 3.190357087174683e-05, "loss": 0.0204, "step": 50960 }, { "grad_norm": 0.12879759073257446, "learning_rate": 3.188430524093353e-05, "loss": 0.0135, "step": 50970 }, { "grad_norm": 0.4003813862800598, "learning_rate": 3.1865042705597014e-05, "loss": 0.0191, "step": 50980 }, { "grad_norm": 0.2712312936782837, "learning_rate": 3.184578326902868e-05, "loss": 0.0159, "step": 50990 }, { "grad_norm": 0.3197658956050873, "learning_rate": 3.182652693451949e-05, "loss": 0.0183, "step": 51000 }, { "grad_norm": 0.3331717252731323, "learning_rate": 3.180727370535981e-05, "loss": 0.0191, "step": 51010 }, { "grad_norm": 0.1557009518146515, "learning_rate": 3.178802358483945e-05, "loss": 0.0164, "step": 51020 }, { "grad_norm": 0.23802180588245392, "learning_rate": 3.17687765762478e-05, "loss": 0.0157, "step": 51030 }, { "grad_norm": 0.3378967344760895, "learning_rate": 3.174953268287357e-05, "loss": 0.0193, "step": 51040 }, { "grad_norm": 0.42354685068130493, "learning_rate": 3.173029190800508e-05, "loss": 0.0153, "step": 51050 }, { "grad_norm": 0.29957666993141174, "learning_rate": 3.171105425493e-05, "loss": 0.0174, "step": 51060 }, { "grad_norm": 0.18047894537448883, "learning_rate": 3.169181972693557e-05, "loss": 0.0128, "step": 51070 }, { "grad_norm": 0.4109176695346832, "learning_rate": 3.1672588327308404e-05, "loss": 0.0165, "step": 51080 }, { "grad_norm": 0.34661000967025757, "learning_rate": 3.1653360059334646e-05, "loss": 0.017, "step": 51090 }, { "grad_norm": 0.41963720321655273, "learning_rate": 3.163413492629985e-05, "loss": 0.0304, "step": 51100 }, { "grad_norm": 0.3616742193698883, "learning_rate": 3.161491293148912e-05, "loss": 0.0187, "step": 51110 }, { "grad_norm": 0.26457691192626953, "learning_rate": 3.1595694078186925e-05, "loss": 0.0186, "step": 51120 }, { "grad_norm": 0.2397484928369522, "learning_rate": 3.157647836967724e-05, "loss": 0.0165, "step": 51130 }, { "grad_norm": 0.22314417362213135, "learning_rate": 3.155726580924354e-05, "loss": 0.0184, "step": 51140 }, { "grad_norm": 0.21952153742313385, "learning_rate": 3.1538056400168676e-05, "loss": 0.0254, "step": 51150 }, { "grad_norm": 0.17362146079540253, "learning_rate": 3.151885014573506e-05, "loss": 0.02, "step": 51160 }, { "grad_norm": 0.2952524721622467, "learning_rate": 3.149964704922448e-05, "loss": 0.0232, "step": 51170 }, { "grad_norm": 0.235393688082695, "learning_rate": 3.148044711391825e-05, "loss": 0.0167, "step": 51180 }, { "grad_norm": 0.9229655265808105, "learning_rate": 3.146125034309707e-05, "loss": 0.0196, "step": 51190 }, { "grad_norm": 0.20861108601093292, "learning_rate": 3.1442056740041195e-05, "loss": 0.0206, "step": 51200 }, { "grad_norm": 0.19492153823375702, "learning_rate": 3.142286630803026e-05, "loss": 0.0304, "step": 51210 }, { "grad_norm": 0.2884725034236908, "learning_rate": 3.140367905034337e-05, "loss": 0.0161, "step": 51220 }, { "grad_norm": 0.23509037494659424, "learning_rate": 3.138449497025914e-05, "loss": 0.0209, "step": 51230 }, { "grad_norm": 0.35699862241744995, "learning_rate": 3.136531407105557e-05, "loss": 0.0197, "step": 51240 }, { "grad_norm": 0.2834831476211548, "learning_rate": 3.1346136356010184e-05, "loss": 0.0219, "step": 51250 }, { "grad_norm": 0.16705922782421112, "learning_rate": 3.132696182839988e-05, "loss": 0.0145, "step": 51260 }, { "grad_norm": 0.33643555641174316, "learning_rate": 3.130779049150111e-05, "loss": 0.0192, "step": 51270 }, { "grad_norm": 0.35384127497673035, "learning_rate": 3.128862234858971e-05, "loss": 0.0176, "step": 51280 }, { "grad_norm": 0.24135014414787292, "learning_rate": 3.1269457402941005e-05, "loss": 0.015, "step": 51290 }, { "grad_norm": 0.3175881803035736, "learning_rate": 3.125029565782974e-05, "loss": 0.0165, "step": 51300 }, { "grad_norm": 0.4071807861328125, "learning_rate": 3.123113711653013e-05, "loss": 0.0228, "step": 51310 }, { "grad_norm": 0.2542189061641693, "learning_rate": 3.121198178231587e-05, "loss": 0.0193, "step": 51320 }, { "grad_norm": 0.1531580537557602, "learning_rate": 3.119282965846006e-05, "loss": 0.0172, "step": 51330 }, { "grad_norm": 0.30299559235572815, "learning_rate": 3.1173680748235303e-05, "loss": 0.0209, "step": 51340 }, { "grad_norm": 0.29179665446281433, "learning_rate": 3.115453505491358e-05, "loss": 0.0213, "step": 51350 }, { "grad_norm": 0.33212709426879883, "learning_rate": 3.1135392581766404e-05, "loss": 0.0174, "step": 51360 }, { "grad_norm": 0.23023362457752228, "learning_rate": 3.111625333206467e-05, "loss": 0.0162, "step": 51370 }, { "grad_norm": 0.20827561616897583, "learning_rate": 3.109711730907881e-05, "loss": 0.0137, "step": 51380 }, { "grad_norm": 0.4515785872936249, "learning_rate": 3.107798451607859e-05, "loss": 0.0202, "step": 51390 }, { "grad_norm": 0.1902654767036438, "learning_rate": 3.105885495633328e-05, "loss": 0.0175, "step": 51400 }, { "grad_norm": 0.11808500438928604, "learning_rate": 3.103972863311163e-05, "loss": 0.016, "step": 51410 }, { "grad_norm": 0.31527137756347656, "learning_rate": 3.102060554968178e-05, "loss": 0.0205, "step": 51420 }, { "grad_norm": 0.24760973453521729, "learning_rate": 3.100148570931137e-05, "loss": 0.0136, "step": 51430 }, { "grad_norm": 0.38412076234817505, "learning_rate": 3.098236911526744e-05, "loss": 0.0173, "step": 51440 }, { "grad_norm": 0.27630338072776794, "learning_rate": 3.0963255770816495e-05, "loss": 0.0138, "step": 51450 }, { "grad_norm": 0.19134938716888428, "learning_rate": 3.094414567922447e-05, "loss": 0.0157, "step": 51460 }, { "grad_norm": 0.2116578370332718, "learning_rate": 3.092503884375678e-05, "loss": 0.018, "step": 51470 }, { "grad_norm": 0.3526932895183563, "learning_rate": 3.0905935267678255e-05, "loss": 0.0186, "step": 51480 }, { "grad_norm": 0.2568264901638031, "learning_rate": 3.0886834954253166e-05, "loss": 0.0146, "step": 51490 }, { "grad_norm": 0.27898022532463074, "learning_rate": 3.086773790674524e-05, "loss": 0.0173, "step": 51500 }, { "grad_norm": 0.6064465641975403, "learning_rate": 3.0848644128417617e-05, "loss": 0.0222, "step": 51510 }, { "grad_norm": 0.37526488304138184, "learning_rate": 3.082955362253294e-05, "loss": 0.0171, "step": 51520 }, { "grad_norm": 0.2936483919620514, "learning_rate": 3.081046639235323e-05, "loss": 0.0188, "step": 51530 }, { "grad_norm": 0.24063363671302795, "learning_rate": 3.0791382441139986e-05, "loss": 0.0134, "step": 51540 }, { "grad_norm": 0.1987522840499878, "learning_rate": 3.077230177215411e-05, "loss": 0.0198, "step": 51550 }, { "grad_norm": 0.149063840508461, "learning_rate": 3.0753224388656e-05, "loss": 0.0145, "step": 51560 }, { "grad_norm": 0.26423490047454834, "learning_rate": 3.073415029390544e-05, "loss": 0.0249, "step": 51570 }, { "grad_norm": 0.32683172821998596, "learning_rate": 3.071507949116166e-05, "loss": 0.0174, "step": 51580 }, { "grad_norm": 0.22226859629154205, "learning_rate": 3.069601198368337e-05, "loss": 0.0172, "step": 51590 }, { "grad_norm": 0.1707063466310501, "learning_rate": 3.067694777472864e-05, "loss": 0.0188, "step": 51600 }, { "grad_norm": 0.2465878576040268, "learning_rate": 3.065788686755508e-05, "loss": 0.0188, "step": 51610 }, { "grad_norm": 0.31142303347587585, "learning_rate": 3.063882926541961e-05, "loss": 0.0208, "step": 51620 }, { "grad_norm": 0.23813454806804657, "learning_rate": 3.061977497157872e-05, "loss": 0.0186, "step": 51630 }, { "grad_norm": 0.18932774662971497, "learning_rate": 3.0600723989288224e-05, "loss": 0.018, "step": 51640 }, { "grad_norm": 0.37649160623550415, "learning_rate": 3.058167632180343e-05, "loss": 0.0227, "step": 51650 }, { "grad_norm": 0.4300965368747711, "learning_rate": 3.0562631972379074e-05, "loss": 0.0236, "step": 51660 }, { "grad_norm": 0.220543771982193, "learning_rate": 3.0543590944269276e-05, "loss": 0.0162, "step": 51670 }, { "grad_norm": 0.3231789171695709, "learning_rate": 3.052455324072766e-05, "loss": 0.0212, "step": 51680 }, { "grad_norm": 0.32468053698539734, "learning_rate": 3.0505518865007244e-05, "loss": 0.0221, "step": 51690 }, { "grad_norm": 0.23732736706733704, "learning_rate": 3.048648782036048e-05, "loss": 0.0162, "step": 51700 }, { "grad_norm": 0.4898773729801178, "learning_rate": 3.0467460110039236e-05, "loss": 0.0186, "step": 51710 }, { "grad_norm": 0.3289518654346466, "learning_rate": 3.0448435737294866e-05, "loss": 0.0173, "step": 51720 }, { "grad_norm": 0.2965058386325836, "learning_rate": 3.042941470537808e-05, "loss": 0.0155, "step": 51730 }, { "grad_norm": 0.20500072836875916, "learning_rate": 3.0410397017539072e-05, "loss": 0.0153, "step": 51740 }, { "grad_norm": 0.41210341453552246, "learning_rate": 3.039138267702742e-05, "loss": 0.0175, "step": 51750 }, { "grad_norm": 0.14430224895477295, "learning_rate": 3.0372371687092193e-05, "loss": 0.0137, "step": 51760 }, { "grad_norm": 0.2711982727050781, "learning_rate": 3.0353364050981823e-05, "loss": 0.0182, "step": 51770 }, { "grad_norm": 0.3311534523963928, "learning_rate": 3.033435977194418e-05, "loss": 0.0179, "step": 51780 }, { "grad_norm": 0.3033488094806671, "learning_rate": 3.0315358853226618e-05, "loss": 0.0214, "step": 51790 }, { "grad_norm": 0.2626388370990753, "learning_rate": 3.0296361298075844e-05, "loss": 0.0173, "step": 51800 }, { "grad_norm": 0.29567545652389526, "learning_rate": 3.027736710973803e-05, "loss": 0.018, "step": 51810 }, { "grad_norm": 0.4825538694858551, "learning_rate": 3.025837629145875e-05, "loss": 0.0185, "step": 51820 }, { "grad_norm": 0.28309154510498047, "learning_rate": 3.0239388846483048e-05, "loss": 0.0181, "step": 51830 }, { "grad_norm": 0.217359721660614, "learning_rate": 3.022040477805532e-05, "loss": 0.0221, "step": 51840 }, { "grad_norm": 0.30577123165130615, "learning_rate": 3.020142408941946e-05, "loss": 0.0181, "step": 51850 }, { "grad_norm": 0.24993261694908142, "learning_rate": 3.018244678381873e-05, "loss": 0.0149, "step": 51860 }, { "grad_norm": 0.40320298075675964, "learning_rate": 3.0163472864495812e-05, "loss": 0.023, "step": 51870 }, { "grad_norm": 0.27493563294410706, "learning_rate": 3.014450233469287e-05, "loss": 0.0191, "step": 51880 }, { "grad_norm": 0.3321534991264343, "learning_rate": 3.0125535197651422e-05, "loss": 0.0257, "step": 51890 }, { "grad_norm": 0.23468884825706482, "learning_rate": 3.0106571456612448e-05, "loss": 0.0173, "step": 51900 }, { "grad_norm": 0.2211921364068985, "learning_rate": 3.0087611114816305e-05, "loss": 0.0206, "step": 51910 }, { "grad_norm": 0.1940089762210846, "learning_rate": 3.006865417550284e-05, "loss": 0.0163, "step": 51920 }, { "grad_norm": 0.2398482859134674, "learning_rate": 3.0049700641911242e-05, "loss": 0.0178, "step": 51930 }, { "grad_norm": 0.21873082220554352, "learning_rate": 3.0030750517280175e-05, "loss": 0.0159, "step": 51940 }, { "grad_norm": 0.24630235135555267, "learning_rate": 3.0011803804847682e-05, "loss": 0.0192, "step": 51950 }, { "grad_norm": 0.6336149573326111, "learning_rate": 2.9992860507851224e-05, "loss": 0.0183, "step": 51960 }, { "grad_norm": 0.2079552263021469, "learning_rate": 2.9973920629527723e-05, "loss": 0.0163, "step": 51970 }, { "grad_norm": 0.10881531983613968, "learning_rate": 2.9954984173113453e-05, "loss": 0.0127, "step": 51980 }, { "grad_norm": 0.2770560085773468, "learning_rate": 2.993605114184418e-05, "loss": 0.0197, "step": 51990 }, { "grad_norm": 0.30852001905441284, "learning_rate": 2.9917121538955005e-05, "loss": 0.0218, "step": 52000 }, { "grad_norm": 0.47016170620918274, "learning_rate": 2.98981953676805e-05, "loss": 0.023, "step": 52010 }, { "grad_norm": 0.22412526607513428, "learning_rate": 2.9879272631254594e-05, "loss": 0.0184, "step": 52020 }, { "grad_norm": 0.25883305072784424, "learning_rate": 2.986035333291073e-05, "loss": 0.0179, "step": 52030 }, { "grad_norm": 0.20627890527248383, "learning_rate": 2.9841437475881646e-05, "loss": 0.0217, "step": 52040 }, { "grad_norm": 0.37780648469924927, "learning_rate": 2.9822525063399555e-05, "loss": 0.0203, "step": 52050 }, { "grad_norm": 0.20733685791492462, "learning_rate": 2.9803616098696087e-05, "loss": 0.0221, "step": 52060 }, { "grad_norm": 0.2639886736869812, "learning_rate": 2.9784710585002228e-05, "loss": 0.0181, "step": 52070 }, { "grad_norm": 0.32039281725883484, "learning_rate": 2.9765808525548467e-05, "loss": 0.0204, "step": 52080 }, { "grad_norm": 0.2613570988178253, "learning_rate": 2.974690992356461e-05, "loss": 0.0176, "step": 52090 }, { "grad_norm": 0.2017713189125061, "learning_rate": 2.9728014782279934e-05, "loss": 0.0204, "step": 52100 }, { "grad_norm": 0.29092440009117126, "learning_rate": 2.970912310492307e-05, "loss": 0.0189, "step": 52110 }, { "grad_norm": 0.2296217679977417, "learning_rate": 2.9690234894722134e-05, "loss": 0.0142, "step": 52120 }, { "grad_norm": 0.24018532037734985, "learning_rate": 2.9671350154904577e-05, "loss": 0.0208, "step": 52130 }, { "grad_norm": 0.34333309531211853, "learning_rate": 2.9652468888697282e-05, "loss": 0.018, "step": 52140 }, { "grad_norm": 1.7212574481964111, "learning_rate": 2.9633591099326562e-05, "loss": 0.0175, "step": 52150 }, { "grad_norm": 0.24058763682842255, "learning_rate": 2.9614716790018072e-05, "loss": 0.0173, "step": 52160 }, { "grad_norm": 0.1862356811761856, "learning_rate": 2.959584596399697e-05, "loss": 0.012, "step": 52170 }, { "grad_norm": 0.30676692724227905, "learning_rate": 2.9576978624487717e-05, "loss": 0.0211, "step": 52180 }, { "grad_norm": 0.21403548121452332, "learning_rate": 2.9558114774714263e-05, "loss": 0.016, "step": 52190 }, { "grad_norm": 0.2961083650588989, "learning_rate": 2.9539254417899897e-05, "loss": 0.0197, "step": 52200 }, { "grad_norm": 0.1954983025789261, "learning_rate": 2.9520397557267365e-05, "loss": 0.0188, "step": 52210 }, { "grad_norm": 0.26638898253440857, "learning_rate": 2.9501544196038765e-05, "loss": 0.0227, "step": 52220 }, { "grad_norm": 0.19452688097953796, "learning_rate": 2.9482694337435614e-05, "loss": 0.0211, "step": 52230 }, { "grad_norm": 0.1935282200574875, "learning_rate": 2.946384798467887e-05, "loss": 0.0142, "step": 52240 }, { "grad_norm": 0.2935766577720642, "learning_rate": 2.9445005140988824e-05, "loss": 0.0139, "step": 52250 }, { "grad_norm": 0.3415764272212982, "learning_rate": 2.942616580958524e-05, "loss": 0.0193, "step": 52260 }, { "grad_norm": 0.8217490315437317, "learning_rate": 2.9407329993687193e-05, "loss": 0.018, "step": 52270 }, { "grad_norm": 0.30203676223754883, "learning_rate": 2.938849769651326e-05, "loss": 0.0161, "step": 52280 }, { "grad_norm": 0.29535144567489624, "learning_rate": 2.9369668921281345e-05, "loss": 0.0175, "step": 52290 }, { "grad_norm": 0.17434468865394592, "learning_rate": 2.9350843671208773e-05, "loss": 0.013, "step": 52300 }, { "grad_norm": 0.28634214401245117, "learning_rate": 2.9332021949512245e-05, "loss": 0.0163, "step": 52310 }, { "grad_norm": 0.39837828278541565, "learning_rate": 2.9313203759407908e-05, "loss": 0.0173, "step": 52320 }, { "grad_norm": 0.2694302201271057, "learning_rate": 2.929438910411127e-05, "loss": 0.0269, "step": 52330 }, { "grad_norm": 0.368338942527771, "learning_rate": 2.927557798683722e-05, "loss": 0.0176, "step": 52340 }, { "grad_norm": 0.3608972728252411, "learning_rate": 2.9256770410800095e-05, "loss": 0.0166, "step": 52350 }, { "grad_norm": 0.33202794194221497, "learning_rate": 2.9237966379213554e-05, "loss": 0.0189, "step": 52360 }, { "grad_norm": 0.2913803160190582, "learning_rate": 2.9219165895290736e-05, "loss": 0.0179, "step": 52370 }, { "grad_norm": 0.24790772795677185, "learning_rate": 2.9200368962244083e-05, "loss": 0.0195, "step": 52380 }, { "grad_norm": 0.3046967089176178, "learning_rate": 2.9181575583285536e-05, "loss": 0.0158, "step": 52390 }, { "grad_norm": 0.2752707004547119, "learning_rate": 2.916278576162632e-05, "loss": 0.0207, "step": 52400 }, { "grad_norm": 0.2527807950973511, "learning_rate": 2.9143999500477116e-05, "loss": 0.0154, "step": 52410 }, { "grad_norm": 0.2835181653499603, "learning_rate": 2.9125216803048004e-05, "loss": 0.0218, "step": 52420 }, { "grad_norm": 0.2229052633047104, "learning_rate": 2.9106437672548403e-05, "loss": 0.0172, "step": 52430 }, { "grad_norm": 0.392238587141037, "learning_rate": 2.9087662112187154e-05, "loss": 0.0126, "step": 52440 }, { "grad_norm": 0.2515636384487152, "learning_rate": 2.9068890125172498e-05, "loss": 0.014, "step": 52450 }, { "grad_norm": 0.3884345591068268, "learning_rate": 2.9050121714712058e-05, "loss": 0.0218, "step": 52460 }, { "grad_norm": 0.4710506498813629, "learning_rate": 2.9031356884012835e-05, "loss": 0.0203, "step": 52470 }, { "grad_norm": 0.21044157445430756, "learning_rate": 2.9012595636281247e-05, "loss": 0.0207, "step": 52480 }, { "grad_norm": 0.17331142723560333, "learning_rate": 2.8993837974723016e-05, "loss": 0.026, "step": 52490 }, { "grad_norm": 0.18129169940948486, "learning_rate": 2.8975083902543394e-05, "loss": 0.0144, "step": 52500 }, { "grad_norm": 0.2549022436141968, "learning_rate": 2.895633342294688e-05, "loss": 0.0191, "step": 52510 }, { "grad_norm": 0.2575293779373169, "learning_rate": 2.893758653913744e-05, "loss": 0.0161, "step": 52520 }, { "grad_norm": 0.22029832005500793, "learning_rate": 2.8918843254318423e-05, "loss": 0.0189, "step": 52530 }, { "grad_norm": 0.3108661472797394, "learning_rate": 2.8900103571692483e-05, "loss": 0.0187, "step": 52540 }, { "grad_norm": 0.2726788818836212, "learning_rate": 2.8881367494461796e-05, "loss": 0.0152, "step": 52550 }, { "grad_norm": 0.34253787994384766, "learning_rate": 2.886263502582779e-05, "loss": 0.0229, "step": 52560 }, { "grad_norm": 0.2790732979774475, "learning_rate": 2.8843906168991353e-05, "loss": 0.0182, "step": 52570 }, { "grad_norm": 0.23702996969223022, "learning_rate": 2.8825180927152728e-05, "loss": 0.0198, "step": 52580 }, { "grad_norm": 0.3464653491973877, "learning_rate": 2.8806459303511547e-05, "loss": 0.0152, "step": 52590 }, { "grad_norm": 0.35156184434890747, "learning_rate": 2.8787741301266852e-05, "loss": 0.018, "step": 52600 }, { "grad_norm": 0.25404092669487, "learning_rate": 2.876902692361697e-05, "loss": 0.0159, "step": 52610 }, { "grad_norm": 0.2831242084503174, "learning_rate": 2.875031617375976e-05, "loss": 0.0167, "step": 52620 }, { "grad_norm": 0.20508208870887756, "learning_rate": 2.8731609054892318e-05, "loss": 0.015, "step": 52630 }, { "grad_norm": 0.15475018322467804, "learning_rate": 2.871290557021119e-05, "loss": 0.0168, "step": 52640 }, { "grad_norm": 0.2447449266910553, "learning_rate": 2.8694205722912298e-05, "loss": 0.0178, "step": 52650 }, { "grad_norm": 0.49753719568252563, "learning_rate": 2.8675509516190936e-05, "loss": 0.0198, "step": 52660 }, { "grad_norm": 0.33429795503616333, "learning_rate": 2.8656816953241772e-05, "loss": 0.0247, "step": 52670 }, { "grad_norm": 0.2674863338470459, "learning_rate": 2.863812803725887e-05, "loss": 0.0154, "step": 52680 }, { "grad_norm": 0.1966005265712738, "learning_rate": 2.8619442771435623e-05, "loss": 0.0156, "step": 52690 }, { "grad_norm": 0.21913743019104004, "learning_rate": 2.860076115896484e-05, "loss": 0.0252, "step": 52700 }, { "grad_norm": 0.3120143711566925, "learning_rate": 2.8582083203038712e-05, "loss": 0.021, "step": 52710 }, { "grad_norm": 0.2452782392501831, "learning_rate": 2.8563408906848778e-05, "loss": 0.0147, "step": 52720 }, { "grad_norm": 0.2859804332256317, "learning_rate": 2.8544738273585993e-05, "loss": 0.0165, "step": 52730 }, { "grad_norm": 0.20549693703651428, "learning_rate": 2.8526071306440595e-05, "loss": 0.0152, "step": 52740 }, { "grad_norm": 0.251774400472641, "learning_rate": 2.850740800860233e-05, "loss": 0.0144, "step": 52750 }, { "grad_norm": 0.2192951887845993, "learning_rate": 2.8488748383260178e-05, "loss": 0.0144, "step": 52760 }, { "grad_norm": 0.4795328676700592, "learning_rate": 2.8470092433602635e-05, "loss": 0.02, "step": 52770 }, { "grad_norm": 0.3034661114215851, "learning_rate": 2.845144016281745e-05, "loss": 0.0162, "step": 52780 }, { "grad_norm": 0.19099010527133942, "learning_rate": 2.8432791574091744e-05, "loss": 0.0226, "step": 52790 }, { "grad_norm": 0.47367286682128906, "learning_rate": 2.8414146670612134e-05, "loss": 0.0177, "step": 52800 }, { "grad_norm": 0.1501426100730896, "learning_rate": 2.8395505455564446e-05, "loss": 0.0207, "step": 52810 }, { "grad_norm": 0.2409372329711914, "learning_rate": 2.837686793213403e-05, "loss": 0.0165, "step": 52820 }, { "grad_norm": 0.29111984372138977, "learning_rate": 2.8358234103505477e-05, "loss": 0.0168, "step": 52830 }, { "grad_norm": 0.3814500868320465, "learning_rate": 2.8339603972862806e-05, "loss": 0.0262, "step": 52840 }, { "grad_norm": 0.73445063829422, "learning_rate": 2.83209775433894e-05, "loss": 0.0228, "step": 52850 }, { "grad_norm": 0.21392126381397247, "learning_rate": 2.8302354818268013e-05, "loss": 0.0154, "step": 52860 }, { "grad_norm": 0.28459542989730835, "learning_rate": 2.8283735800680754e-05, "loss": 0.0224, "step": 52870 }, { "grad_norm": 0.18879160284996033, "learning_rate": 2.826512049380913e-05, "loss": 0.0159, "step": 52880 }, { "grad_norm": 0.271496057510376, "learning_rate": 2.8246508900833935e-05, "loss": 0.0171, "step": 52890 }, { "grad_norm": 0.2751138210296631, "learning_rate": 2.822790102493541e-05, "loss": 0.0196, "step": 52900 }, { "grad_norm": 0.1771927922964096, "learning_rate": 2.820929686929314e-05, "loss": 0.0202, "step": 52910 }, { "grad_norm": 0.1887824386358261, "learning_rate": 2.819069643708605e-05, "loss": 0.0122, "step": 52920 }, { "grad_norm": 0.27624863386154175, "learning_rate": 2.8172099731492484e-05, "loss": 0.0185, "step": 52930 }, { "grad_norm": 0.19959677755832672, "learning_rate": 2.815350675569003e-05, "loss": 0.0244, "step": 52940 }, { "grad_norm": 0.2643636465072632, "learning_rate": 2.8134917512855807e-05, "loss": 0.0233, "step": 52950 }, { "grad_norm": 0.19904428720474243, "learning_rate": 2.8116332006166134e-05, "loss": 0.0126, "step": 52960 }, { "grad_norm": 0.3302159309387207, "learning_rate": 2.809775023879685e-05, "loss": 0.0265, "step": 52970 }, { "grad_norm": 0.2690529227256775, "learning_rate": 2.807917221392299e-05, "loss": 0.0206, "step": 52980 }, { "grad_norm": 0.3745521605014801, "learning_rate": 2.8060597934719067e-05, "loss": 0.0156, "step": 52990 }, { "grad_norm": 0.18239165842533112, "learning_rate": 2.8042027404358927e-05, "loss": 0.0163, "step": 53000 }, { "grad_norm": 0.2255023568868637, "learning_rate": 2.802346062601572e-05, "loss": 0.0189, "step": 53010 }, { "grad_norm": 0.39874833822250366, "learning_rate": 2.800489760286207e-05, "loss": 0.0281, "step": 53020 }, { "grad_norm": 0.2661474943161011, "learning_rate": 2.7986338338069834e-05, "loss": 0.0176, "step": 53030 }, { "grad_norm": 0.3222648799419403, "learning_rate": 2.7967782834810297e-05, "loss": 0.0234, "step": 53040 }, { "grad_norm": 0.24780774116516113, "learning_rate": 2.794923109625409e-05, "loss": 0.0163, "step": 53050 }, { "grad_norm": 0.3787126839160919, "learning_rate": 2.7930683125571205e-05, "loss": 0.0164, "step": 53060 }, { "grad_norm": 0.4044104516506195, "learning_rate": 2.7912138925930997e-05, "loss": 0.0245, "step": 53070 }, { "grad_norm": 0.2636852264404297, "learning_rate": 2.7893598500502117e-05, "loss": 0.0278, "step": 53080 }, { "grad_norm": 0.4191303551197052, "learning_rate": 2.7875061852452644e-05, "loss": 0.0177, "step": 53090 }, { "grad_norm": 0.31696850061416626, "learning_rate": 2.7856528984949982e-05, "loss": 0.0204, "step": 53100 }, { "grad_norm": 0.1619156002998352, "learning_rate": 2.7837999901160888e-05, "loss": 0.0141, "step": 53110 }, { "grad_norm": 0.2867867648601532, "learning_rate": 2.7819474604251484e-05, "loss": 0.0197, "step": 53120 }, { "grad_norm": 0.31031298637390137, "learning_rate": 2.780095309738725e-05, "loss": 0.0182, "step": 53130 }, { "grad_norm": 0.2781525254249573, "learning_rate": 2.778243538373294e-05, "loss": 0.0185, "step": 53140 }, { "grad_norm": 0.16689439117908478, "learning_rate": 2.7763921466452826e-05, "loss": 0.0164, "step": 53150 }, { "grad_norm": 0.2924966812133789, "learning_rate": 2.7745411348710336e-05, "loss": 0.0245, "step": 53160 }, { "grad_norm": 0.3697795569896698, "learning_rate": 2.7726905033668395e-05, "loss": 0.0244, "step": 53170 }, { "grad_norm": 0.33090344071388245, "learning_rate": 2.7708402524489214e-05, "loss": 0.0231, "step": 53180 }, { "grad_norm": 0.3030230700969696, "learning_rate": 2.7689903824334364e-05, "loss": 0.0207, "step": 53190 }, { "grad_norm": 0.1946885883808136, "learning_rate": 2.7671408936364785e-05, "loss": 0.0131, "step": 53200 }, { "grad_norm": 0.27027416229248047, "learning_rate": 2.765291786374069e-05, "loss": 0.0183, "step": 53210 }, { "grad_norm": 0.4407604932785034, "learning_rate": 2.7634430609621775e-05, "loss": 0.0183, "step": 53220 }, { "grad_norm": 0.3992142677307129, "learning_rate": 2.7615947177166956e-05, "loss": 0.02, "step": 53230 }, { "grad_norm": 0.33001551032066345, "learning_rate": 2.7597467569534553e-05, "loss": 0.0126, "step": 53240 }, { "grad_norm": 0.2279125154018402, "learning_rate": 2.757899178988226e-05, "loss": 0.0166, "step": 53250 }, { "grad_norm": 0.3749818205833435, "learning_rate": 2.7560519841367005e-05, "loss": 0.0174, "step": 53260 }, { "grad_norm": 0.1615377515554428, "learning_rate": 2.7542051727145237e-05, "loss": 0.0169, "step": 53270 }, { "grad_norm": 0.27077674865722656, "learning_rate": 2.7523587450372578e-05, "loss": 0.023, "step": 53280 }, { "grad_norm": 0.42247456312179565, "learning_rate": 2.750512701420409e-05, "loss": 0.0171, "step": 53290 }, { "grad_norm": 0.16904239356517792, "learning_rate": 2.7486670421794158e-05, "loss": 0.014, "step": 53300 }, { "grad_norm": 0.232849583029747, "learning_rate": 2.7468217676296515e-05, "loss": 0.0258, "step": 53310 }, { "grad_norm": 0.32530781626701355, "learning_rate": 2.7449768780864226e-05, "loss": 0.0233, "step": 53320 }, { "grad_norm": 0.28045061230659485, "learning_rate": 2.7431323738649724e-05, "loss": 0.0154, "step": 53330 }, { "grad_norm": 0.38671740889549255, "learning_rate": 2.7412882552804713e-05, "loss": 0.0183, "step": 53340 }, { "grad_norm": 0.21556802093982697, "learning_rate": 2.739444522648032e-05, "loss": 0.0117, "step": 53350 }, { "grad_norm": 0.21977059543132782, "learning_rate": 2.737601176282697e-05, "loss": 0.0159, "step": 53360 }, { "grad_norm": 0.3064897656440735, "learning_rate": 2.735758216499445e-05, "loss": 0.0147, "step": 53370 }, { "grad_norm": 0.27769041061401367, "learning_rate": 2.7339156436131864e-05, "loss": 0.0134, "step": 53380 }, { "grad_norm": 0.30518656969070435, "learning_rate": 2.7320734579387663e-05, "loss": 0.0174, "step": 53390 }, { "grad_norm": 0.4427906572818756, "learning_rate": 2.730231659790966e-05, "loss": 0.0135, "step": 53400 }, { "grad_norm": 0.31253018975257874, "learning_rate": 2.728390249484494e-05, "loss": 0.0235, "step": 53410 }, { "grad_norm": 0.18721894919872284, "learning_rate": 2.7265492273340032e-05, "loss": 0.0155, "step": 53420 }, { "grad_norm": 0.6830272078514099, "learning_rate": 2.7247085936540678e-05, "loss": 0.014, "step": 53430 }, { "grad_norm": 0.15758638083934784, "learning_rate": 2.722868348759205e-05, "loss": 0.0187, "step": 53440 }, { "grad_norm": 0.2649015188217163, "learning_rate": 2.7210284929638635e-05, "loss": 0.0201, "step": 53450 }, { "grad_norm": 0.20540985465049744, "learning_rate": 2.7191890265824183e-05, "loss": 0.0169, "step": 53460 }, { "grad_norm": 0.209754079580307, "learning_rate": 2.7173499499291926e-05, "loss": 0.0152, "step": 53470 }, { "grad_norm": 0.41526514291763306, "learning_rate": 2.7155112633184277e-05, "loss": 0.0159, "step": 53480 }, { "grad_norm": 0.30691632628440857, "learning_rate": 2.713672967064307e-05, "loss": 0.0208, "step": 53490 }, { "grad_norm": 0.1561158150434494, "learning_rate": 2.711835061480945e-05, "loss": 0.0135, "step": 53500 }, { "grad_norm": 0.5517591238021851, "learning_rate": 2.7099975468823896e-05, "loss": 0.021, "step": 53510 }, { "grad_norm": 0.2586701214313507, "learning_rate": 2.708160423582622e-05, "loss": 0.0184, "step": 53520 }, { "grad_norm": 0.553787112236023, "learning_rate": 2.706323691895557e-05, "loss": 0.0207, "step": 53530 }, { "grad_norm": 0.3273690640926361, "learning_rate": 2.70448735213504e-05, "loss": 0.013, "step": 53540 }, { "grad_norm": 0.31167072057724, "learning_rate": 2.702651404614852e-05, "loss": 0.0193, "step": 53550 }, { "grad_norm": 0.39398324489593506, "learning_rate": 2.7008158496487056e-05, "loss": 0.016, "step": 53560 }, { "grad_norm": 0.21769575774669647, "learning_rate": 2.6989806875502487e-05, "loss": 0.0144, "step": 53570 }, { "grad_norm": 0.23277679085731506, "learning_rate": 2.6971459186330584e-05, "loss": 0.0129, "step": 53580 }, { "grad_norm": 0.34911367297172546, "learning_rate": 2.695311543210648e-05, "loss": 0.0171, "step": 53590 }, { "grad_norm": 0.2753715217113495, "learning_rate": 2.693477561596463e-05, "loss": 0.0148, "step": 53600 }, { "grad_norm": 0.30320557951927185, "learning_rate": 2.6916439741038756e-05, "loss": 0.0155, "step": 53610 }, { "grad_norm": 0.35293224453926086, "learning_rate": 2.689810781046203e-05, "loss": 0.0161, "step": 53620 }, { "grad_norm": 0.19897952675819397, "learning_rate": 2.6879779827366823e-05, "loss": 0.0167, "step": 53630 }, { "grad_norm": 0.16531768441200256, "learning_rate": 2.6861455794884904e-05, "loss": 0.0159, "step": 53640 }, { "grad_norm": 0.2579430639743805, "learning_rate": 2.6843135716147373e-05, "loss": 0.0162, "step": 53650 }, { "grad_norm": 0.2569361925125122, "learning_rate": 2.6824819594284556e-05, "loss": 0.0201, "step": 53660 }, { "grad_norm": 0.23989389836788177, "learning_rate": 2.6806507432426275e-05, "loss": 0.0179, "step": 53670 }, { "grad_norm": 0.47711408138275146, "learning_rate": 2.6788199233701512e-05, "loss": 0.0183, "step": 53680 }, { "grad_norm": 0.3026496171951294, "learning_rate": 2.6769895001238652e-05, "loss": 0.0223, "step": 53690 }, { "grad_norm": 0.30117011070251465, "learning_rate": 2.67515947381654e-05, "loss": 0.019, "step": 53700 }, { "grad_norm": 0.19525618851184845, "learning_rate": 2.6733298447608745e-05, "loss": 0.0139, "step": 53710 }, { "grad_norm": 0.45139825344085693, "learning_rate": 2.6715006132695074e-05, "loss": 0.0224, "step": 53720 }, { "grad_norm": 0.21680112183094025, "learning_rate": 2.6696717796549985e-05, "loss": 0.0233, "step": 53730 }, { "grad_norm": 0.3360690474510193, "learning_rate": 2.667843344229848e-05, "loss": 0.0158, "step": 53740 }, { "grad_norm": 0.39084991812705994, "learning_rate": 2.6660153073064846e-05, "loss": 0.0184, "step": 53750 }, { "grad_norm": 0.31240060925483704, "learning_rate": 2.664187669197271e-05, "loss": 0.0253, "step": 53760 }, { "grad_norm": 0.3766762614250183, "learning_rate": 2.6623604302145005e-05, "loss": 0.0171, "step": 53770 }, { "grad_norm": 0.2695196270942688, "learning_rate": 2.6605335906703975e-05, "loss": 0.0176, "step": 53780 }, { "grad_norm": 0.42289969325065613, "learning_rate": 2.65870715087712e-05, "loss": 0.022, "step": 53790 }, { "grad_norm": 0.251659095287323, "learning_rate": 2.6568811111467573e-05, "loss": 0.0168, "step": 53800 }, { "grad_norm": 0.2304128110408783, "learning_rate": 2.6550554717913258e-05, "loss": 0.0174, "step": 53810 }, { "grad_norm": 0.25772738456726074, "learning_rate": 2.6532302331227805e-05, "loss": 0.0169, "step": 53820 }, { "grad_norm": 0.3053140938282013, "learning_rate": 2.651405395453004e-05, "loss": 0.0201, "step": 53830 }, { "grad_norm": 0.2615862488746643, "learning_rate": 2.6495809590938115e-05, "loss": 0.0199, "step": 53840 }, { "grad_norm": 0.13949429988861084, "learning_rate": 2.64775692435695e-05, "loss": 0.0174, "step": 53850 }, { "grad_norm": 0.21261845529079437, "learning_rate": 2.6459332915540928e-05, "loss": 0.0156, "step": 53860 }, { "grad_norm": 0.3023392856121063, "learning_rate": 2.644110060996856e-05, "loss": 0.0221, "step": 53870 }, { "grad_norm": 0.21744921803474426, "learning_rate": 2.642287232996774e-05, "loss": 0.0181, "step": 53880 }, { "grad_norm": 0.25957849621772766, "learning_rate": 2.6404648078653205e-05, "loss": 0.0124, "step": 53890 }, { "grad_norm": 0.4229104518890381, "learning_rate": 2.6386427859139002e-05, "loss": 0.0196, "step": 53900 }, { "grad_norm": 0.326178640127182, "learning_rate": 2.63682116745384e-05, "loss": 0.0182, "step": 53910 }, { "grad_norm": 0.2964751422405243, "learning_rate": 2.6349999527964138e-05, "loss": 0.0145, "step": 53920 }, { "grad_norm": 0.230095773935318, "learning_rate": 2.63317914225281e-05, "loss": 0.0183, "step": 53930 }, { "grad_norm": 0.18083959817886353, "learning_rate": 2.6313587361341585e-05, "loss": 0.0176, "step": 53940 }, { "grad_norm": 0.5336701273918152, "learning_rate": 2.6295387347515165e-05, "loss": 0.0222, "step": 53950 }, { "grad_norm": 0.22241465747356415, "learning_rate": 2.6277191384158727e-05, "loss": 0.0245, "step": 53960 }, { "grad_norm": 0.21683593094348907, "learning_rate": 2.6258999474381457e-05, "loss": 0.0182, "step": 53970 }, { "grad_norm": 0.23132306337356567, "learning_rate": 2.624081162129186e-05, "loss": 0.0198, "step": 53980 }, { "grad_norm": 0.23759566247463226, "learning_rate": 2.6222627827997765e-05, "loss": 0.0273, "step": 53990 }, { "grad_norm": 0.24915026128292084, "learning_rate": 2.6204448097606236e-05, "loss": 0.0169, "step": 54000 }, { "grad_norm": 0.19893856346607208, "learning_rate": 2.6186272433223726e-05, "loss": 0.0201, "step": 54010 }, { "grad_norm": 0.17441298067569733, "learning_rate": 2.6168100837955943e-05, "loss": 0.0203, "step": 54020 }, { "grad_norm": 0.19285103678703308, "learning_rate": 2.6149933314907926e-05, "loss": 0.025, "step": 54030 }, { "grad_norm": 0.3498072028160095, "learning_rate": 2.6131769867184e-05, "loss": 0.0254, "step": 54040 }, { "grad_norm": 0.25691020488739014, "learning_rate": 2.611361049788783e-05, "loss": 0.0193, "step": 54050 }, { "grad_norm": 0.34269678592681885, "learning_rate": 2.6095455210122292e-05, "loss": 0.0181, "step": 54060 }, { "grad_norm": 0.190910205245018, "learning_rate": 2.6077304006989712e-05, "loss": 0.0112, "step": 54070 }, { "grad_norm": 0.3447502553462982, "learning_rate": 2.6059156891591562e-05, "loss": 0.0203, "step": 54080 }, { "grad_norm": 0.2684779763221741, "learning_rate": 2.6041013867028718e-05, "loss": 0.018, "step": 54090 }, { "grad_norm": 0.24036505818367004, "learning_rate": 2.6022874936401347e-05, "loss": 0.0143, "step": 54100 }, { "grad_norm": 0.1866522878408432, "learning_rate": 2.6004740102808832e-05, "loss": 0.0188, "step": 54110 }, { "grad_norm": 0.286321759223938, "learning_rate": 2.598660936935e-05, "loss": 0.0117, "step": 54120 }, { "grad_norm": 0.3633863925933838, "learning_rate": 2.5968482739122845e-05, "loss": 0.0189, "step": 54130 }, { "grad_norm": 0.27306661009788513, "learning_rate": 2.595036021522472e-05, "loss": 0.0156, "step": 54140 }, { "grad_norm": 0.4034426808357239, "learning_rate": 2.5932241800752278e-05, "loss": 0.0213, "step": 54150 }, { "grad_norm": 0.44443413615226746, "learning_rate": 2.5914127498801453e-05, "loss": 0.0305, "step": 54160 }, { "grad_norm": 0.26463356614112854, "learning_rate": 2.5896017312467497e-05, "loss": 0.0171, "step": 54170 }, { "grad_norm": 0.27237623929977417, "learning_rate": 2.587791124484493e-05, "loss": 0.0146, "step": 54180 }, { "grad_norm": 0.34215736389160156, "learning_rate": 2.5859809299027615e-05, "loss": 0.0168, "step": 54190 }, { "grad_norm": 0.2421867400407791, "learning_rate": 2.5841711478108632e-05, "loss": 0.027, "step": 54200 }, { "grad_norm": 0.30080944299697876, "learning_rate": 2.582361778518043e-05, "loss": 0.017, "step": 54210 }, { "grad_norm": 0.24635133147239685, "learning_rate": 2.580552822333472e-05, "loss": 0.019, "step": 54220 }, { "grad_norm": 0.5411547422409058, "learning_rate": 2.578744279566252e-05, "loss": 0.0228, "step": 54230 }, { "grad_norm": 0.19439847767353058, "learning_rate": 2.576936150525413e-05, "loss": 0.0176, "step": 54240 }, { "grad_norm": 0.3760696053504944, "learning_rate": 2.5751284355199168e-05, "loss": 0.018, "step": 54250 }, { "grad_norm": 0.13028302788734436, "learning_rate": 2.573321134858646e-05, "loss": 0.0108, "step": 54260 }, { "grad_norm": 0.16668391227722168, "learning_rate": 2.5715142488504286e-05, "loss": 0.0131, "step": 54270 }, { "grad_norm": 0.14651086926460266, "learning_rate": 2.5697077778040042e-05, "loss": 0.017, "step": 54280 }, { "grad_norm": 0.30514776706695557, "learning_rate": 2.5679017220280522e-05, "loss": 0.0328, "step": 54290 }, { "grad_norm": 0.20386049151420593, "learning_rate": 2.5660960818311796e-05, "loss": 0.0164, "step": 54300 }, { "grad_norm": 0.2730887532234192, "learning_rate": 2.564290857521915e-05, "loss": 0.0157, "step": 54310 }, { "grad_norm": 0.23265109956264496, "learning_rate": 2.5624860494087298e-05, "loss": 0.0174, "step": 54320 }, { "grad_norm": 0.13141217827796936, "learning_rate": 2.5606816578000115e-05, "loss": 0.0201, "step": 54330 }, { "grad_norm": 0.30072200298309326, "learning_rate": 2.558877683004082e-05, "loss": 0.0156, "step": 54340 }, { "grad_norm": 0.19828271865844727, "learning_rate": 2.557074125329192e-05, "loss": 0.0159, "step": 54350 }, { "grad_norm": 0.29498693346977234, "learning_rate": 2.5552709850835195e-05, "loss": 0.0156, "step": 54360 }, { "grad_norm": 0.21680057048797607, "learning_rate": 2.5534682625751738e-05, "loss": 0.0114, "step": 54370 }, { "grad_norm": 0.350604772567749, "learning_rate": 2.551665958112186e-05, "loss": 0.0217, "step": 54380 }, { "grad_norm": 0.1733824461698532, "learning_rate": 2.549864072002527e-05, "loss": 0.0117, "step": 54390 }, { "grad_norm": 0.2430744171142578, "learning_rate": 2.5480626045540858e-05, "loss": 0.0206, "step": 54400 }, { "grad_norm": 0.4481259882450104, "learning_rate": 2.546261556074684e-05, "loss": 0.016, "step": 54410 }, { "grad_norm": 0.25109678506851196, "learning_rate": 2.5444609268720726e-05, "loss": 0.0183, "step": 54420 }, { "grad_norm": 0.2548120617866516, "learning_rate": 2.5426607172539297e-05, "loss": 0.0143, "step": 54430 }, { "grad_norm": 0.2369641810655594, "learning_rate": 2.5408609275278617e-05, "loss": 0.0225, "step": 54440 }, { "grad_norm": 0.20029085874557495, "learning_rate": 2.5390615580014055e-05, "loss": 0.0175, "step": 54450 }, { "grad_norm": 0.2634066641330719, "learning_rate": 2.5372626089820207e-05, "loss": 0.0214, "step": 54460 }, { "grad_norm": 0.34070977568626404, "learning_rate": 2.5354640807770997e-05, "loss": 0.0162, "step": 54470 }, { "grad_norm": 0.34236785769462585, "learning_rate": 2.5336659736939622e-05, "loss": 0.0152, "step": 54480 }, { "grad_norm": 0.19596096873283386, "learning_rate": 2.5318682880398554e-05, "loss": 0.02, "step": 54490 }, { "grad_norm": 0.22711442410945892, "learning_rate": 2.530071024121956e-05, "loss": 0.0185, "step": 54500 }, { "grad_norm": 0.2410094141960144, "learning_rate": 2.5282741822473627e-05, "loss": 0.0138, "step": 54510 }, { "grad_norm": 0.30035725235939026, "learning_rate": 2.526477762723114e-05, "loss": 0.0169, "step": 54520 }, { "grad_norm": 0.2643774449825287, "learning_rate": 2.5246817658561618e-05, "loss": 0.0183, "step": 54530 }, { "grad_norm": 0.1583261340856552, "learning_rate": 2.5228861919533965e-05, "loss": 0.0168, "step": 54540 }, { "grad_norm": 0.26675111055374146, "learning_rate": 2.5210910413216326e-05, "loss": 0.0264, "step": 54550 }, { "grad_norm": 0.2295876145362854, "learning_rate": 2.5192963142676086e-05, "loss": 0.0129, "step": 54560 }, { "grad_norm": 0.2281540036201477, "learning_rate": 2.517502011098001e-05, "loss": 0.0142, "step": 54570 }, { "grad_norm": 0.3072354197502136, "learning_rate": 2.5157081321193987e-05, "loss": 0.0154, "step": 54580 }, { "grad_norm": 0.3030400574207306, "learning_rate": 2.5139146776383356e-05, "loss": 0.0194, "step": 54590 }, { "grad_norm": 0.17215441167354584, "learning_rate": 2.5121216479612575e-05, "loss": 0.0146, "step": 54600 }, { "grad_norm": 0.3075355887413025, "learning_rate": 2.510329043394546e-05, "loss": 0.0247, "step": 54610 }, { "grad_norm": 0.3187592923641205, "learning_rate": 2.508536864244508e-05, "loss": 0.0216, "step": 54620 }, { "grad_norm": 0.32992416620254517, "learning_rate": 2.5067451108173778e-05, "loss": 0.0166, "step": 54630 }, { "grad_norm": 0.2511064410209656, "learning_rate": 2.5049537834193204e-05, "loss": 0.0169, "step": 54640 }, { "grad_norm": 0.8806582689285278, "learning_rate": 2.5031628823564194e-05, "loss": 0.0166, "step": 54650 }, { "grad_norm": 0.2604846656322479, "learning_rate": 2.5013724079346933e-05, "loss": 0.0182, "step": 54660 }, { "grad_norm": 0.19239942729473114, "learning_rate": 2.4995823604600854e-05, "loss": 0.0133, "step": 54670 }, { "grad_norm": 0.1959029734134674, "learning_rate": 2.497792740238465e-05, "loss": 0.0153, "step": 54680 }, { "grad_norm": 0.23835328221321106, "learning_rate": 2.49600354757563e-05, "loss": 0.02, "step": 54690 }, { "grad_norm": 0.2589567005634308, "learning_rate": 2.494214782777306e-05, "loss": 0.0174, "step": 54700 }, { "grad_norm": 0.35627707839012146, "learning_rate": 2.4924264461491386e-05, "loss": 0.0186, "step": 54710 }, { "grad_norm": 0.282357394695282, "learning_rate": 2.4906385379967133e-05, "loss": 0.0185, "step": 54720 }, { "grad_norm": 0.27498456835746765, "learning_rate": 2.4888510586255285e-05, "loss": 0.0128, "step": 54730 }, { "grad_norm": 0.17834733426570892, "learning_rate": 2.487064008341018e-05, "loss": 0.015, "step": 54740 }, { "grad_norm": 0.38644400238990784, "learning_rate": 2.4852773874485407e-05, "loss": 0.0176, "step": 54750 }, { "grad_norm": 0.29017373919487, "learning_rate": 2.483491196253377e-05, "loss": 0.017, "step": 54760 }, { "grad_norm": 0.227016419172287, "learning_rate": 2.4817054350607443e-05, "loss": 0.018, "step": 54770 }, { "grad_norm": 0.42342525720596313, "learning_rate": 2.4799201041757742e-05, "loss": 0.0224, "step": 54780 }, { "grad_norm": 0.33253347873687744, "learning_rate": 2.4781352039035373e-05, "loss": 0.0164, "step": 54790 }, { "grad_norm": 0.17843864858150482, "learning_rate": 2.4763507345490194e-05, "loss": 0.0125, "step": 54800 }, { "grad_norm": 0.19179795682430267, "learning_rate": 2.4745666964171386e-05, "loss": 0.0165, "step": 54810 }, { "grad_norm": 0.9033106565475464, "learning_rate": 2.4727830898127387e-05, "loss": 0.0182, "step": 54820 }, { "grad_norm": 0.18379509449005127, "learning_rate": 2.4709999150405895e-05, "loss": 0.0233, "step": 54830 }, { "grad_norm": 0.28323638439178467, "learning_rate": 2.469217172405388e-05, "loss": 0.0139, "step": 54840 }, { "grad_norm": 0.20205412805080414, "learning_rate": 2.4674348622117527e-05, "loss": 0.0124, "step": 54850 }, { "grad_norm": 0.40089866518974304, "learning_rate": 2.465652984764234e-05, "loss": 0.0203, "step": 54860 }, { "grad_norm": 0.20899252593517303, "learning_rate": 2.4638715403673056e-05, "loss": 0.0138, "step": 54870 }, { "grad_norm": 0.44097504019737244, "learning_rate": 2.4620905293253676e-05, "loss": 0.0187, "step": 54880 }, { "grad_norm": 0.46150675415992737, "learning_rate": 2.4603099519427463e-05, "loss": 0.0207, "step": 54890 }, { "grad_norm": 0.7335881590843201, "learning_rate": 2.458529808523695e-05, "loss": 0.0166, "step": 54900 }, { "grad_norm": 0.22146937251091003, "learning_rate": 2.456750099372387e-05, "loss": 0.0164, "step": 54910 }, { "grad_norm": 0.22257930040359497, "learning_rate": 2.4549708247929327e-05, "loss": 0.018, "step": 54920 }, { "grad_norm": 0.23531173169612885, "learning_rate": 2.4531919850893554e-05, "loss": 0.0137, "step": 54930 }, { "grad_norm": 0.12459072470664978, "learning_rate": 2.4514135805656125e-05, "loss": 0.0135, "step": 54940 }, { "grad_norm": 0.27752426266670227, "learning_rate": 2.449635611525587e-05, "loss": 0.0172, "step": 54950 }, { "grad_norm": 0.24395416676998138, "learning_rate": 2.447858078273079e-05, "loss": 0.0174, "step": 54960 }, { "grad_norm": 0.22807525098323822, "learning_rate": 2.4460809811118275e-05, "loss": 0.023, "step": 54970 }, { "grad_norm": 0.19534869492053986, "learning_rate": 2.444304320345483e-05, "loss": 0.0137, "step": 54980 }, { "grad_norm": 0.18959034979343414, "learning_rate": 2.4425280962776354e-05, "loss": 0.0181, "step": 54990 }, { "grad_norm": 0.2889939546585083, "learning_rate": 2.4407523092117875e-05, "loss": 0.0178, "step": 55000 }, { "grad_norm": 0.45132648944854736, "learning_rate": 2.438976959451374e-05, "loss": 0.0143, "step": 55010 }, { "grad_norm": 0.2679935395717621, "learning_rate": 2.4372020472997565e-05, "loss": 0.0151, "step": 55020 }, { "grad_norm": 0.21535679697990417, "learning_rate": 2.4354275730602122e-05, "loss": 0.0153, "step": 55030 }, { "grad_norm": 0.2601882815361023, "learning_rate": 2.433653537035958e-05, "loss": 0.018, "step": 55040 }, { "grad_norm": 0.28208985924720764, "learning_rate": 2.431879939530123e-05, "loss": 0.0194, "step": 55050 }, { "grad_norm": 0.22858670353889465, "learning_rate": 2.4301067808457684e-05, "loss": 0.0132, "step": 55060 }, { "grad_norm": 0.24210382997989655, "learning_rate": 2.428334061285878e-05, "loss": 0.0124, "step": 55070 }, { "grad_norm": 0.3784042000770569, "learning_rate": 2.426561781153361e-05, "loss": 0.0171, "step": 55080 }, { "grad_norm": 0.14893320202827454, "learning_rate": 2.424789940751052e-05, "loss": 0.0186, "step": 55090 }, { "grad_norm": 0.21664796769618988, "learning_rate": 2.423018540381712e-05, "loss": 0.0182, "step": 55100 }, { "grad_norm": 0.14922219514846802, "learning_rate": 2.421247580348021e-05, "loss": 0.0177, "step": 55110 }, { "grad_norm": 0.2595577836036682, "learning_rate": 2.4194770609525897e-05, "loss": 0.0144, "step": 55120 }, { "grad_norm": 0.1658071130514145, "learning_rate": 2.417706982497951e-05, "loss": 0.0157, "step": 55130 }, { "grad_norm": 0.19603510200977325, "learning_rate": 2.415937345286563e-05, "loss": 0.0183, "step": 55140 }, { "grad_norm": 0.2826789915561676, "learning_rate": 2.4141681496208087e-05, "loss": 0.0166, "step": 55150 }, { "grad_norm": 0.21679256856441498, "learning_rate": 2.4123993958029946e-05, "loss": 0.0126, "step": 55160 }, { "grad_norm": 0.4054364264011383, "learning_rate": 2.4106310841353548e-05, "loss": 0.0195, "step": 55170 }, { "grad_norm": 0.1576727330684662, "learning_rate": 2.4088632149200398e-05, "loss": 0.0192, "step": 55180 }, { "grad_norm": 0.2718808054924011, "learning_rate": 2.4070957884591367e-05, "loss": 0.0148, "step": 55190 }, { "grad_norm": 0.45864003896713257, "learning_rate": 2.4053288050546464e-05, "loss": 0.0215, "step": 55200 }, { "grad_norm": 0.18810206651687622, "learning_rate": 2.403562265008498e-05, "loss": 0.0191, "step": 55210 }, { "grad_norm": 0.6273738145828247, "learning_rate": 2.4017961686225483e-05, "loss": 0.017, "step": 55220 }, { "grad_norm": 0.23557701706886292, "learning_rate": 2.400030516198568e-05, "loss": 0.024, "step": 55230 }, { "grad_norm": 0.38724079728126526, "learning_rate": 2.3982653080382673e-05, "loss": 0.0164, "step": 55240 }, { "grad_norm": 0.4257667362689972, "learning_rate": 2.396500544443266e-05, "loss": 0.0159, "step": 55250 }, { "grad_norm": 0.22272707521915436, "learning_rate": 2.3947362257151156e-05, "loss": 0.014, "step": 55260 }, { "grad_norm": 0.4576515555381775, "learning_rate": 2.39297235215529e-05, "loss": 0.0198, "step": 55270 }, { "grad_norm": 0.24398748576641083, "learning_rate": 2.3912089240651873e-05, "loss": 0.0195, "step": 55280 }, { "grad_norm": 0.15491986274719238, "learning_rate": 2.389445941746129e-05, "loss": 0.0244, "step": 55290 }, { "grad_norm": 0.28314295411109924, "learning_rate": 2.3876834054993625e-05, "loss": 0.0182, "step": 55300 }, { "grad_norm": 0.21039198338985443, "learning_rate": 2.3859213156260522e-05, "loss": 0.0145, "step": 55310 }, { "grad_norm": 0.21571941673755646, "learning_rate": 2.3841596724272948e-05, "loss": 0.0153, "step": 55320 }, { "grad_norm": 0.26802611351013184, "learning_rate": 2.382398476204106e-05, "loss": 0.0142, "step": 55330 }, { "grad_norm": 0.2387898713350296, "learning_rate": 2.3806377272574254e-05, "loss": 0.0148, "step": 55340 }, { "grad_norm": 0.29580169916152954, "learning_rate": 2.3788774258881174e-05, "loss": 0.014, "step": 55350 }, { "grad_norm": 0.5375280976295471, "learning_rate": 2.37711757239697e-05, "loss": 0.0199, "step": 55360 }, { "grad_norm": 0.2793704569339752, "learning_rate": 2.3753581670846954e-05, "loss": 0.0095, "step": 55370 }, { "grad_norm": 0.16119369864463806, "learning_rate": 2.3735992102519216e-05, "loss": 0.0174, "step": 55380 }, { "grad_norm": 0.2556630074977875, "learning_rate": 2.371840702199215e-05, "loss": 0.015, "step": 55390 }, { "grad_norm": 0.20155476033687592, "learning_rate": 2.3700826432270494e-05, "loss": 0.011, "step": 55400 }, { "grad_norm": 0.22519129514694214, "learning_rate": 2.3683250336358326e-05, "loss": 0.0182, "step": 55410 }, { "grad_norm": 0.16753263771533966, "learning_rate": 2.3665678737258923e-05, "loss": 0.0216, "step": 55420 }, { "grad_norm": 0.16232937574386597, "learning_rate": 2.3648111637974745e-05, "loss": 0.0153, "step": 55430 }, { "grad_norm": 0.20995382964611053, "learning_rate": 2.36305490415076e-05, "loss": 0.016, "step": 55440 }, { "grad_norm": 0.3377925157546997, "learning_rate": 2.3612990950858392e-05, "loss": 0.0188, "step": 55450 }, { "grad_norm": 0.1633714884519577, "learning_rate": 2.359543736902735e-05, "loss": 0.0152, "step": 55460 }, { "grad_norm": 0.18458405137062073, "learning_rate": 2.3577888299013896e-05, "loss": 0.013, "step": 55470 }, { "grad_norm": 0.2711886465549469, "learning_rate": 2.3560343743816683e-05, "loss": 0.0149, "step": 55480 }, { "grad_norm": 0.22662527859210968, "learning_rate": 2.354280370643362e-05, "loss": 0.0202, "step": 55490 }, { "grad_norm": 0.43228280544281006, "learning_rate": 2.3525268189861777e-05, "loss": 0.0155, "step": 55500 }, { "grad_norm": 0.3634880781173706, "learning_rate": 2.3507737197097513e-05, "loss": 0.0146, "step": 55510 }, { "grad_norm": 0.24931223690509796, "learning_rate": 2.34902107311364e-05, "loss": 0.0151, "step": 55520 }, { "grad_norm": 0.19587399065494537, "learning_rate": 2.3472688794973225e-05, "loss": 0.019, "step": 55530 }, { "grad_norm": 0.7257823944091797, "learning_rate": 2.3455171391602016e-05, "loss": 0.0194, "step": 55540 }, { "grad_norm": 0.3638611435890198, "learning_rate": 2.343765852401601e-05, "loss": 0.0143, "step": 55550 }, { "grad_norm": 0.19377514719963074, "learning_rate": 2.342015019520768e-05, "loss": 0.0183, "step": 55560 }, { "grad_norm": 0.2293856292963028, "learning_rate": 2.3402646408168742e-05, "loss": 0.0163, "step": 55570 }, { "grad_norm": 0.3791721761226654, "learning_rate": 2.3385147165890074e-05, "loss": 0.0223, "step": 55580 }, { "grad_norm": 0.18939612805843353, "learning_rate": 2.336765247136184e-05, "loss": 0.0148, "step": 55590 }, { "grad_norm": 0.3701583743095398, "learning_rate": 2.33501623275734e-05, "loss": 0.0243, "step": 55600 }, { "grad_norm": 0.30374082922935486, "learning_rate": 2.333267673751334e-05, "loss": 0.0144, "step": 55610 }, { "grad_norm": 0.20559830963611603, "learning_rate": 2.33151957041695e-05, "loss": 0.0112, "step": 55620 }, { "grad_norm": 0.27083247900009155, "learning_rate": 2.329771923052884e-05, "loss": 0.0165, "step": 55630 }, { "grad_norm": 0.21716058254241943, "learning_rate": 2.3280247319577697e-05, "loss": 0.0181, "step": 55640 }, { "grad_norm": 0.3704397678375244, "learning_rate": 2.3262779974301473e-05, "loss": 0.0234, "step": 55650 }, { "grad_norm": 0.28622967004776, "learning_rate": 2.3245317197684895e-05, "loss": 0.015, "step": 55660 }, { "grad_norm": 0.33532023429870605, "learning_rate": 2.322785899271188e-05, "loss": 0.016, "step": 55670 }, { "grad_norm": 0.29792463779449463, "learning_rate": 2.3210405362365507e-05, "loss": 0.0231, "step": 55680 }, { "grad_norm": 0.23960423469543457, "learning_rate": 2.31929563096282e-05, "loss": 0.019, "step": 55690 }, { "grad_norm": 0.22823438048362732, "learning_rate": 2.317551183748146e-05, "loss": 0.0149, "step": 55700 }, { "grad_norm": 0.19223259389400482, "learning_rate": 2.3158071948906103e-05, "loss": 0.0164, "step": 55710 }, { "grad_norm": 0.2740671932697296, "learning_rate": 2.314063664688212e-05, "loss": 0.0143, "step": 55720 }, { "grad_norm": 0.47056660056114197, "learning_rate": 2.3123205934388725e-05, "loss": 0.0121, "step": 55730 }, { "grad_norm": 0.1849863976240158, "learning_rate": 2.310577981440436e-05, "loss": 0.0191, "step": 55740 }, { "grad_norm": 0.3282267451286316, "learning_rate": 2.3088358289906665e-05, "loss": 0.0131, "step": 55750 }, { "grad_norm": 0.2847813069820404, "learning_rate": 2.307094136387252e-05, "loss": 0.0137, "step": 55760 }, { "grad_norm": 0.28800928592681885, "learning_rate": 2.305352903927796e-05, "loss": 0.022, "step": 55770 }, { "grad_norm": 0.3398512601852417, "learning_rate": 2.303612131909831e-05, "loss": 0.0152, "step": 55780 }, { "grad_norm": 0.3950486183166504, "learning_rate": 2.3018718206308054e-05, "loss": 0.0207, "step": 55790 }, { "grad_norm": 0.32451242208480835, "learning_rate": 2.3001319703880925e-05, "loss": 0.0178, "step": 55800 }, { "grad_norm": 0.3075406551361084, "learning_rate": 2.2983925814789835e-05, "loss": 0.0146, "step": 55810 }, { "grad_norm": 0.3081924021244049, "learning_rate": 2.2966536542006957e-05, "loss": 0.0199, "step": 55820 }, { "grad_norm": 0.18878887593746185, "learning_rate": 2.294915188850358e-05, "loss": 0.0185, "step": 55830 }, { "grad_norm": 0.14204417169094086, "learning_rate": 2.2931771857250333e-05, "loss": 0.0165, "step": 55840 }, { "grad_norm": 0.20721155405044556, "learning_rate": 2.2914396451216946e-05, "loss": 0.0158, "step": 55850 }, { "grad_norm": 0.27110555768013, "learning_rate": 2.2897025673372412e-05, "loss": 0.0147, "step": 55860 }, { "grad_norm": 0.21595831215381622, "learning_rate": 2.287965952668494e-05, "loss": 0.0142, "step": 55870 }, { "grad_norm": 0.24718844890594482, "learning_rate": 2.2862298014121873e-05, "loss": 0.0157, "step": 55880 }, { "grad_norm": 0.607753336429596, "learning_rate": 2.28449411386499e-05, "loss": 0.0153, "step": 55890 }, { "grad_norm": 0.21772029995918274, "learning_rate": 2.282758890323477e-05, "loss": 0.017, "step": 55900 }, { "grad_norm": 0.17243507504463196, "learning_rate": 2.2810241310841528e-05, "loss": 0.0156, "step": 55910 }, { "grad_norm": 0.23816488683223724, "learning_rate": 2.2792898364434412e-05, "loss": 0.0133, "step": 55920 }, { "grad_norm": 0.321359246969223, "learning_rate": 2.2775560066976843e-05, "loss": 0.0159, "step": 55930 }, { "grad_norm": 0.16018478572368622, "learning_rate": 2.275822642143147e-05, "loss": 0.014, "step": 55940 }, { "grad_norm": 0.2504613697528839, "learning_rate": 2.274089743076014e-05, "loss": 0.0157, "step": 55950 }, { "grad_norm": 0.32010602951049805, "learning_rate": 2.2723573097923917e-05, "loss": 0.0168, "step": 55960 }, { "grad_norm": 0.203527569770813, "learning_rate": 2.2706253425883024e-05, "loss": 0.0135, "step": 55970 }, { "grad_norm": 0.24250954389572144, "learning_rate": 2.2688938417596933e-05, "loss": 0.0177, "step": 55980 }, { "grad_norm": 0.12663283944129944, "learning_rate": 2.2671628076024305e-05, "loss": 0.0136, "step": 55990 }, { "grad_norm": 0.4416891634464264, "learning_rate": 2.2654322404123008e-05, "loss": 0.0124, "step": 56000 }, { "grad_norm": 0.16276811063289642, "learning_rate": 2.2637021404850105e-05, "loss": 0.0156, "step": 56010 }, { "grad_norm": 0.3126128613948822, "learning_rate": 2.2619725081161885e-05, "loss": 0.0193, "step": 56020 }, { "grad_norm": 0.20332081615924835, "learning_rate": 2.260243343601376e-05, "loss": 0.0177, "step": 56030 }, { "grad_norm": 0.1977594643831253, "learning_rate": 2.2585146472360473e-05, "loss": 0.017, "step": 56040 }, { "grad_norm": 0.31381791830062866, "learning_rate": 2.2567864193155834e-05, "loss": 0.0172, "step": 56050 }, { "grad_norm": 0.22577358782291412, "learning_rate": 2.255058660135294e-05, "loss": 0.0177, "step": 56060 }, { "grad_norm": 0.39673787355422974, "learning_rate": 2.253331369990407e-05, "loss": 0.0189, "step": 56070 }, { "grad_norm": 0.17178696393966675, "learning_rate": 2.251604549176063e-05, "loss": 0.0195, "step": 56080 }, { "grad_norm": 0.27096113562583923, "learning_rate": 2.249878197987337e-05, "loss": 0.017, "step": 56090 }, { "grad_norm": 0.5361501574516296, "learning_rate": 2.2481523167192087e-05, "loss": 0.0183, "step": 56100 }, { "grad_norm": 0.19183744490146637, "learning_rate": 2.2464269056665867e-05, "loss": 0.0156, "step": 56110 }, { "grad_norm": 0.20731496810913086, "learning_rate": 2.2447019651242958e-05, "loss": 0.0135, "step": 56120 }, { "grad_norm": 0.3037734627723694, "learning_rate": 2.2429774953870818e-05, "loss": 0.0143, "step": 56130 }, { "grad_norm": 0.36957603693008423, "learning_rate": 2.241253496749611e-05, "loss": 0.0184, "step": 56140 }, { "grad_norm": 0.3837926685810089, "learning_rate": 2.2395299695064614e-05, "loss": 0.0135, "step": 56150 }, { "grad_norm": 0.4909755289554596, "learning_rate": 2.237806913952145e-05, "loss": 0.0206, "step": 56160 }, { "grad_norm": 0.25970837473869324, "learning_rate": 2.2360843303810798e-05, "loss": 0.017, "step": 56170 }, { "grad_norm": 0.2554221749305725, "learning_rate": 2.2343622190876084e-05, "loss": 0.0177, "step": 56180 }, { "grad_norm": 0.2714121639728546, "learning_rate": 2.2326405803659935e-05, "loss": 0.0143, "step": 56190 }, { "grad_norm": 0.2941417992115021, "learning_rate": 2.230919414510416e-05, "loss": 0.0141, "step": 56200 }, { "grad_norm": 0.39464274048805237, "learning_rate": 2.229198721814976e-05, "loss": 0.0175, "step": 56210 }, { "grad_norm": 0.23621027171611786, "learning_rate": 2.2274785025736948e-05, "loss": 0.0153, "step": 56220 }, { "grad_norm": 0.31711751222610474, "learning_rate": 2.225758757080507e-05, "loss": 0.0174, "step": 56230 }, { "grad_norm": 0.35122790932655334, "learning_rate": 2.2240394856292723e-05, "loss": 0.0151, "step": 56240 }, { "grad_norm": 0.3842892050743103, "learning_rate": 2.2223206885137664e-05, "loss": 0.0182, "step": 56250 }, { "grad_norm": 0.4547453224658966, "learning_rate": 2.2206023660276853e-05, "loss": 0.0175, "step": 56260 }, { "grad_norm": 0.3642098903656006, "learning_rate": 2.218884518464645e-05, "loss": 0.0156, "step": 56270 }, { "grad_norm": 0.40233173966407776, "learning_rate": 2.2171671461181732e-05, "loss": 0.0176, "step": 56280 }, { "grad_norm": 0.29811617732048035, "learning_rate": 2.2154502492817292e-05, "loss": 0.016, "step": 56290 }, { "grad_norm": 0.22864364087581635, "learning_rate": 2.2137338282486782e-05, "loss": 0.011, "step": 56300 }, { "grad_norm": 0.08704877644777298, "learning_rate": 2.2120178833123113e-05, "loss": 0.0159, "step": 56310 }, { "grad_norm": 0.4297831058502197, "learning_rate": 2.210302414765838e-05, "loss": 0.0194, "step": 56320 }, { "grad_norm": 0.21287497878074646, "learning_rate": 2.2085874229023808e-05, "loss": 0.0118, "step": 56330 }, { "grad_norm": 0.16086743772029877, "learning_rate": 2.2068729080149907e-05, "loss": 0.0153, "step": 56340 }, { "grad_norm": 0.35295963287353516, "learning_rate": 2.205158870396625e-05, "loss": 0.0154, "step": 56350 }, { "grad_norm": 0.21572303771972656, "learning_rate": 2.2034453103401732e-05, "loss": 0.0187, "step": 56360 }, { "grad_norm": 0.11319134384393692, "learning_rate": 2.20173222813843e-05, "loss": 0.0102, "step": 56370 }, { "grad_norm": 0.22112421691417694, "learning_rate": 2.200019624084116e-05, "loss": 0.0169, "step": 56380 }, { "grad_norm": 0.1991216391324997, "learning_rate": 2.1983074984698687e-05, "loss": 0.0157, "step": 56390 }, { "grad_norm": 0.34056392312049866, "learning_rate": 2.1965958515882433e-05, "loss": 0.0169, "step": 56400 }, { "grad_norm": 0.2672194540500641, "learning_rate": 2.1948846837317162e-05, "loss": 0.0152, "step": 56410 }, { "grad_norm": 0.2265513688325882, "learning_rate": 2.1931739951926738e-05, "loss": 0.0107, "step": 56420 }, { "grad_norm": 0.1874879151582718, "learning_rate": 2.191463786263429e-05, "loss": 0.0115, "step": 56430 }, { "grad_norm": 0.37701287865638733, "learning_rate": 2.1897540572362095e-05, "loss": 0.0154, "step": 56440 }, { "grad_norm": 0.1804189682006836, "learning_rate": 2.1880448084031614e-05, "loss": 0.011, "step": 56450 }, { "grad_norm": 0.20130418241024017, "learning_rate": 2.1863360400563482e-05, "loss": 0.0155, "step": 56460 }, { "grad_norm": 0.40004003047943115, "learning_rate": 2.184627752487754e-05, "loss": 0.0164, "step": 56470 }, { "grad_norm": 0.2827085554599762, "learning_rate": 2.1829199459892725e-05, "loss": 0.017, "step": 56480 }, { "grad_norm": 0.18419241905212402, "learning_rate": 2.1812126208527282e-05, "loss": 0.0148, "step": 56490 }, { "grad_norm": 0.27533984184265137, "learning_rate": 2.179505777369852e-05, "loss": 0.0212, "step": 56500 }, { "grad_norm": 0.5408284664154053, "learning_rate": 2.1777994158322974e-05, "loss": 0.0169, "step": 56510 }, { "grad_norm": 0.36402490735054016, "learning_rate": 2.176093536531637e-05, "loss": 0.0164, "step": 56520 }, { "grad_norm": 0.22644972801208496, "learning_rate": 2.1743881397593537e-05, "loss": 0.0163, "step": 56530 }, { "grad_norm": 0.3958406150341034, "learning_rate": 2.1726832258068595e-05, "loss": 0.0242, "step": 56540 }, { "grad_norm": 0.3549968898296356, "learning_rate": 2.170978794965472e-05, "loss": 0.0154, "step": 56550 }, { "grad_norm": 0.9020429253578186, "learning_rate": 2.169274847526438e-05, "loss": 0.0148, "step": 56560 }, { "grad_norm": 0.4795321226119995, "learning_rate": 2.1675713837809103e-05, "loss": 0.0153, "step": 56570 }, { "grad_norm": 0.16098380088806152, "learning_rate": 2.1658684040199655e-05, "loss": 0.0112, "step": 56580 }, { "grad_norm": 0.1996147483587265, "learning_rate": 2.1641659085345974e-05, "loss": 0.0127, "step": 56590 }, { "grad_norm": 0.1763467788696289, "learning_rate": 2.1624638976157154e-05, "loss": 0.0196, "step": 56600 }, { "grad_norm": 0.28884363174438477, "learning_rate": 2.1607623715541476e-05, "loss": 0.0229, "step": 56610 }, { "grad_norm": 0.21573574841022491, "learning_rate": 2.159061330640636e-05, "loss": 0.013, "step": 56620 }, { "grad_norm": 0.24364762008190155, "learning_rate": 2.1573607751658425e-05, "loss": 0.0104, "step": 56630 }, { "grad_norm": 0.20344598591327667, "learning_rate": 2.155660705420347e-05, "loss": 0.0189, "step": 56640 }, { "grad_norm": 0.7197622656822205, "learning_rate": 2.153961121694644e-05, "loss": 0.0181, "step": 56650 }, { "grad_norm": 0.24018879234790802, "learning_rate": 2.152262024279145e-05, "loss": 0.0244, "step": 56660 }, { "grad_norm": 0.42392751574516296, "learning_rate": 2.150563413464183e-05, "loss": 0.018, "step": 56670 }, { "grad_norm": 0.22642700374126434, "learning_rate": 2.1488652895399974e-05, "loss": 0.0132, "step": 56680 }, { "grad_norm": 0.36640697717666626, "learning_rate": 2.1471676527967587e-05, "loss": 0.0161, "step": 56690 }, { "grad_norm": 0.2050611823797226, "learning_rate": 2.1454705035245414e-05, "loss": 0.0201, "step": 56700 }, { "grad_norm": 0.2599371671676636, "learning_rate": 2.143773842013343e-05, "loss": 0.0147, "step": 56710 }, { "grad_norm": 0.2919771671295166, "learning_rate": 2.1420776685530796e-05, "loss": 0.0184, "step": 56720 }, { "grad_norm": 0.2687605023384094, "learning_rate": 2.1403819834335742e-05, "loss": 0.0111, "step": 56730 }, { "grad_norm": 0.14594413340091705, "learning_rate": 2.1386867869445808e-05, "loss": 0.0148, "step": 56740 }, { "grad_norm": 0.17506200075149536, "learning_rate": 2.1369920793757548e-05, "loss": 0.0194, "step": 56750 }, { "grad_norm": 0.28688713908195496, "learning_rate": 2.1352978610166828e-05, "loss": 0.0133, "step": 56760 }, { "grad_norm": 0.3330082595348358, "learning_rate": 2.1336041321568546e-05, "loss": 0.0174, "step": 56770 }, { "grad_norm": 0.19795501232147217, "learning_rate": 2.131910893085684e-05, "loss": 0.0166, "step": 56780 }, { "grad_norm": 0.20725250244140625, "learning_rate": 2.1302181440925012e-05, "loss": 0.0167, "step": 56790 }, { "grad_norm": 0.3693224787712097, "learning_rate": 2.128525885466546e-05, "loss": 0.016, "step": 56800 }, { "grad_norm": 0.2508029639720917, "learning_rate": 2.1268341174969847e-05, "loss": 0.0177, "step": 56810 }, { "grad_norm": 0.2536512315273285, "learning_rate": 2.1251428404728894e-05, "loss": 0.0149, "step": 56820 }, { "grad_norm": 0.27765071392059326, "learning_rate": 2.123452054683256e-05, "loss": 0.0163, "step": 56830 }, { "grad_norm": 0.4217716157436371, "learning_rate": 2.1217617604169927e-05, "loss": 0.0188, "step": 56840 }, { "grad_norm": 0.27984389662742615, "learning_rate": 2.120071957962924e-05, "loss": 0.0174, "step": 56850 }, { "grad_norm": 0.2492569535970688, "learning_rate": 2.1183826476097917e-05, "loss": 0.0239, "step": 56860 }, { "grad_norm": 0.2599927484989166, "learning_rate": 2.116693829646254e-05, "loss": 0.0184, "step": 56870 }, { "grad_norm": 0.35226431488990784, "learning_rate": 2.1150055043608806e-05, "loss": 0.0144, "step": 56880 }, { "grad_norm": 0.3550354242324829, "learning_rate": 2.113317672042162e-05, "loss": 0.0222, "step": 56890 }, { "grad_norm": 0.43042629957199097, "learning_rate": 2.1116303329785025e-05, "loss": 0.0107, "step": 56900 }, { "grad_norm": 0.21450841426849365, "learning_rate": 2.109943487458222e-05, "loss": 0.0171, "step": 56910 }, { "grad_norm": 0.2578526437282562, "learning_rate": 2.1082571357695574e-05, "loss": 0.0131, "step": 56920 }, { "grad_norm": 0.20801661908626556, "learning_rate": 2.1065712782006557e-05, "loss": 0.0162, "step": 56930 }, { "grad_norm": 0.13206526637077332, "learning_rate": 2.104885915039591e-05, "loss": 0.0121, "step": 56940 }, { "grad_norm": 0.2105836123228073, "learning_rate": 2.103201046574338e-05, "loss": 0.0161, "step": 56950 }, { "grad_norm": 0.33339858055114746, "learning_rate": 2.1015166730928022e-05, "loss": 0.0152, "step": 56960 }, { "grad_norm": 0.20314882695674896, "learning_rate": 2.0998327948827912e-05, "loss": 0.0162, "step": 56970 }, { "grad_norm": 0.2478194385766983, "learning_rate": 2.0981494122320363e-05, "loss": 0.0179, "step": 56980 }, { "grad_norm": 0.3800753951072693, "learning_rate": 2.0964665254281822e-05, "loss": 0.018, "step": 56990 }, { "grad_norm": 0.2357318103313446, "learning_rate": 2.094784134758784e-05, "loss": 0.0142, "step": 57000 }, { "grad_norm": 0.2247644066810608, "learning_rate": 2.0931022405113226e-05, "loss": 0.0255, "step": 57010 }, { "grad_norm": 0.2833455502986908, "learning_rate": 2.091420842973183e-05, "loss": 0.0154, "step": 57020 }, { "grad_norm": 0.17468547821044922, "learning_rate": 2.0897399424316715e-05, "loss": 0.0142, "step": 57030 }, { "grad_norm": 0.2700914144515991, "learning_rate": 2.0880595391740078e-05, "loss": 0.0161, "step": 57040 }, { "grad_norm": 0.3156748414039612, "learning_rate": 2.0863796334873277e-05, "loss": 0.0162, "step": 57050 }, { "grad_norm": 0.2087637335062027, "learning_rate": 2.08470022565868e-05, "loss": 0.0165, "step": 57060 }, { "grad_norm": 0.20929645001888275, "learning_rate": 2.0830213159750317e-05, "loss": 0.0125, "step": 57070 }, { "grad_norm": 0.3715061545372009, "learning_rate": 2.0813429047232596e-05, "loss": 0.0135, "step": 57080 }, { "grad_norm": 0.15666165947914124, "learning_rate": 2.0796649921901594e-05, "loss": 0.0156, "step": 57090 }, { "grad_norm": 0.2193104773759842, "learning_rate": 2.077987578662441e-05, "loss": 0.0141, "step": 57100 }, { "grad_norm": 0.2598761320114136, "learning_rate": 2.0763106644267277e-05, "loss": 0.0172, "step": 57110 }, { "grad_norm": 0.21425552666187286, "learning_rate": 2.0746342497695607e-05, "loss": 0.0082, "step": 57120 }, { "grad_norm": 0.2248040735721588, "learning_rate": 2.0729583349773886e-05, "loss": 0.0114, "step": 57130 }, { "grad_norm": 0.40926483273506165, "learning_rate": 2.0712829203365853e-05, "loss": 0.014, "step": 57140 }, { "grad_norm": 0.28496620059013367, "learning_rate": 2.0696080061334267e-05, "loss": 0.016, "step": 57150 }, { "grad_norm": 0.33388590812683105, "learning_rate": 2.067933592654117e-05, "loss": 0.011, "step": 57160 }, { "grad_norm": 0.32836222648620605, "learning_rate": 2.066259680184763e-05, "loss": 0.0172, "step": 57170 }, { "grad_norm": 0.13494925200939178, "learning_rate": 2.0645862690113908e-05, "loss": 0.0113, "step": 57180 }, { "grad_norm": 0.2398405373096466, "learning_rate": 2.0629133594199436e-05, "loss": 0.0191, "step": 57190 }, { "grad_norm": 0.2015194594860077, "learning_rate": 2.0612409516962704e-05, "loss": 0.0177, "step": 57200 }, { "grad_norm": 0.34990882873535156, "learning_rate": 2.0595690461261467e-05, "loss": 0.0197, "step": 57210 }, { "grad_norm": 0.29839515686035156, "learning_rate": 2.0578976429952503e-05, "loss": 0.0143, "step": 57220 }, { "grad_norm": 0.4350932240486145, "learning_rate": 2.0562267425891802e-05, "loss": 0.0167, "step": 57230 }, { "grad_norm": 0.229599267244339, "learning_rate": 2.0545563451934467e-05, "loss": 0.0146, "step": 57240 }, { "grad_norm": 0.20469732582569122, "learning_rate": 2.0528864510934764e-05, "loss": 0.0137, "step": 57250 }, { "grad_norm": 0.23925164341926575, "learning_rate": 2.0512170605746096e-05, "loss": 0.0218, "step": 57260 }, { "grad_norm": 0.23509690165519714, "learning_rate": 2.0495481739220963e-05, "loss": 0.0167, "step": 57270 }, { "grad_norm": 0.2105579376220703, "learning_rate": 2.0478797914211045e-05, "loss": 0.0193, "step": 57280 }, { "grad_norm": 0.24726983904838562, "learning_rate": 2.046211913356716e-05, "loss": 0.015, "step": 57290 }, { "grad_norm": 0.20723682641983032, "learning_rate": 2.0445445400139247e-05, "loss": 0.0182, "step": 57300 }, { "grad_norm": 0.3213176429271698, "learning_rate": 2.0428776716776405e-05, "loss": 0.0162, "step": 57310 }, { "grad_norm": 0.18529094755649567, "learning_rate": 2.0412113086326856e-05, "loss": 0.0156, "step": 57320 }, { "grad_norm": 0.21048440039157867, "learning_rate": 2.0395454511637918e-05, "loss": 0.0122, "step": 57330 }, { "grad_norm": 0.32734280824661255, "learning_rate": 2.037880099555616e-05, "loss": 0.0147, "step": 57340 }, { "grad_norm": 0.22894048690795898, "learning_rate": 2.0362152540927144e-05, "loss": 0.013, "step": 57350 }, { "grad_norm": 0.20196281373500824, "learning_rate": 2.0345509150595666e-05, "loss": 0.0253, "step": 57360 }, { "grad_norm": 0.2608848810195923, "learning_rate": 2.0328870827405617e-05, "loss": 0.0186, "step": 57370 }, { "grad_norm": 0.2160242795944214, "learning_rate": 2.0312237574200043e-05, "loss": 0.0142, "step": 57380 }, { "grad_norm": 0.24527393281459808, "learning_rate": 2.029560939382112e-05, "loss": 0.0137, "step": 57390 }, { "grad_norm": 0.2668481469154358, "learning_rate": 2.0278986289110097e-05, "loss": 0.0137, "step": 57400 }, { "grad_norm": 0.1827363669872284, "learning_rate": 2.0262368262907484e-05, "loss": 0.0143, "step": 57410 }, { "grad_norm": 0.2926248610019684, "learning_rate": 2.024575531805279e-05, "loss": 0.0174, "step": 57420 }, { "grad_norm": 0.4359194338321686, "learning_rate": 2.0229147457384735e-05, "loss": 0.0115, "step": 57430 }, { "grad_norm": 0.3467879295349121, "learning_rate": 2.0212544683741157e-05, "loss": 0.0235, "step": 57440 }, { "grad_norm": 0.20694519579410553, "learning_rate": 2.0195946999958976e-05, "loss": 0.0259, "step": 57450 }, { "grad_norm": 0.19402572512626648, "learning_rate": 2.017935440887434e-05, "loss": 0.0197, "step": 57460 }, { "grad_norm": 0.17478278279304504, "learning_rate": 2.0162766913322423e-05, "loss": 0.0124, "step": 57470 }, { "grad_norm": 0.1578773558139801, "learning_rate": 2.0146184516137588e-05, "loss": 0.0183, "step": 57480 }, { "grad_norm": 0.2179117649793625, "learning_rate": 2.012960722015332e-05, "loss": 0.0155, "step": 57490 }, { "grad_norm": 0.25279346108436584, "learning_rate": 2.0113035028202214e-05, "loss": 0.0192, "step": 57500 }, { "grad_norm": 0.2732517421245575, "learning_rate": 2.009646794311602e-05, "loss": 0.0157, "step": 57510 }, { "grad_norm": 0.1542346030473709, "learning_rate": 2.007990596772559e-05, "loss": 0.0171, "step": 57520 }, { "grad_norm": 0.1805008053779602, "learning_rate": 2.0063349104860923e-05, "loss": 0.0164, "step": 57530 }, { "grad_norm": 0.2656106948852539, "learning_rate": 2.0046797357351116e-05, "loss": 0.0218, "step": 57540 }, { "grad_norm": 0.33503249287605286, "learning_rate": 2.0030250728024412e-05, "loss": 0.0141, "step": 57550 }, { "grad_norm": 0.5530653595924377, "learning_rate": 2.001370921970819e-05, "loss": 0.0118, "step": 57560 }, { "grad_norm": 0.16772893071174622, "learning_rate": 1.9997172835228932e-05, "loss": 0.0139, "step": 57570 }, { "grad_norm": 0.3381676971912384, "learning_rate": 1.9980641577412262e-05, "loss": 0.0175, "step": 57580 }, { "grad_norm": 0.2666695713996887, "learning_rate": 1.9964115449082925e-05, "loss": 0.0153, "step": 57590 }, { "grad_norm": 0.2813039720058441, "learning_rate": 1.9947594453064742e-05, "loss": 0.0165, "step": 57600 }, { "grad_norm": 0.22889375686645508, "learning_rate": 1.9931078592180774e-05, "loss": 0.0158, "step": 57610 }, { "grad_norm": 0.2512509524822235, "learning_rate": 1.9914567869253065e-05, "loss": 0.0179, "step": 57620 }, { "grad_norm": 0.12007320672273636, "learning_rate": 1.989806228710287e-05, "loss": 0.021, "step": 57630 }, { "grad_norm": 0.17515452206134796, "learning_rate": 1.9881561848550555e-05, "loss": 0.0103, "step": 57640 }, { "grad_norm": 0.3148823380470276, "learning_rate": 1.9865066556415544e-05, "loss": 0.0193, "step": 57650 }, { "grad_norm": 0.20787374675273895, "learning_rate": 1.98485764135165e-05, "loss": 0.0128, "step": 57660 }, { "grad_norm": 0.1712072342634201, "learning_rate": 1.983209142267109e-05, "loss": 0.0129, "step": 57670 }, { "grad_norm": 0.22194905579090118, "learning_rate": 1.9815611586696165e-05, "loss": 0.0134, "step": 57680 }, { "grad_norm": 0.2873072028160095, "learning_rate": 1.9799136908407667e-05, "loss": 0.0189, "step": 57690 }, { "grad_norm": 0.23114249110221863, "learning_rate": 1.9782667390620678e-05, "loss": 0.0182, "step": 57700 }, { "grad_norm": 0.21098266541957855, "learning_rate": 1.976620303614939e-05, "loss": 0.0147, "step": 57710 }, { "grad_norm": 0.3310333788394928, "learning_rate": 1.9749743847807108e-05, "loss": 0.0173, "step": 57720 }, { "grad_norm": 0.2890685498714447, "learning_rate": 1.9733289828406272e-05, "loss": 0.0134, "step": 57730 }, { "grad_norm": 0.3521648943424225, "learning_rate": 1.9716840980758382e-05, "loss": 0.0157, "step": 57740 }, { "grad_norm": 0.2702847123146057, "learning_rate": 1.9700397307674134e-05, "loss": 0.0168, "step": 57750 }, { "grad_norm": 0.21644669771194458, "learning_rate": 1.968395881196328e-05, "loss": 0.0125, "step": 57760 }, { "grad_norm": 0.5285356640815735, "learning_rate": 1.966752549643473e-05, "loss": 0.0173, "step": 57770 }, { "grad_norm": 0.19206936657428741, "learning_rate": 1.965109736389647e-05, "loss": 0.0166, "step": 57780 }, { "grad_norm": 0.27404335141181946, "learning_rate": 1.9634674417155645e-05, "loss": 0.0163, "step": 57790 }, { "grad_norm": 0.283046156167984, "learning_rate": 1.9618256659018434e-05, "loss": 0.0326, "step": 57800 }, { "grad_norm": 0.2849017381668091, "learning_rate": 1.9601844092290257e-05, "loss": 0.0124, "step": 57810 }, { "grad_norm": 0.3907921314239502, "learning_rate": 1.9585436719775512e-05, "loss": 0.0149, "step": 57820 }, { "grad_norm": 0.2221856266260147, "learning_rate": 1.9569034544277793e-05, "loss": 0.0144, "step": 57830 }, { "grad_norm": 0.28244420886039734, "learning_rate": 1.9552637568599798e-05, "loss": 0.0137, "step": 57840 }, { "grad_norm": 0.2707217037677765, "learning_rate": 1.953624579554327e-05, "loss": 0.0187, "step": 57850 }, { "grad_norm": 0.1622738242149353, "learning_rate": 1.951985922790918e-05, "loss": 0.0181, "step": 57860 }, { "grad_norm": 0.19950427114963531, "learning_rate": 1.9503477868497505e-05, "loss": 0.0103, "step": 57870 }, { "grad_norm": 0.35092097520828247, "learning_rate": 1.9487101720107375e-05, "loss": 0.0192, "step": 57880 }, { "grad_norm": 0.270573228597641, "learning_rate": 1.9470730785537032e-05, "loss": 0.0172, "step": 57890 }, { "grad_norm": 0.19351501762866974, "learning_rate": 1.9454365067583823e-05, "loss": 0.014, "step": 57900 }, { "grad_norm": 0.1572919338941574, "learning_rate": 1.9438004569044215e-05, "loss": 0.0135, "step": 57910 }, { "grad_norm": 0.3137320280075073, "learning_rate": 1.9421649292713724e-05, "loss": 0.0302, "step": 57920 }, { "grad_norm": 0.18045002222061157, "learning_rate": 1.9405299241387076e-05, "loss": 0.0142, "step": 57930 }, { "grad_norm": 0.18241117894649506, "learning_rate": 1.9388954417858007e-05, "loss": 0.0109, "step": 57940 }, { "grad_norm": 0.3160291314125061, "learning_rate": 1.9372614824919417e-05, "loss": 0.0137, "step": 57950 }, { "grad_norm": 0.2632976770401001, "learning_rate": 1.9356280465363284e-05, "loss": 0.0173, "step": 57960 }, { "grad_norm": 0.2074917107820511, "learning_rate": 1.9339951341980723e-05, "loss": 0.0131, "step": 57970 }, { "grad_norm": 0.3077009618282318, "learning_rate": 1.9323627457561916e-05, "loss": 0.0147, "step": 57980 }, { "grad_norm": 0.3695645332336426, "learning_rate": 1.9307308814896198e-05, "loss": 0.012, "step": 57990 }, { "grad_norm": 0.1965440809726715, "learning_rate": 1.9290995416771935e-05, "loss": 0.0132, "step": 58000 }, { "grad_norm": 0.18136373162269592, "learning_rate": 1.9274687265976665e-05, "loss": 0.0121, "step": 58010 }, { "grad_norm": 0.2897217571735382, "learning_rate": 1.9258384365297e-05, "loss": 0.0138, "step": 58020 }, { "grad_norm": 0.3010284900665283, "learning_rate": 1.924208671751866e-05, "loss": 0.0126, "step": 58030 }, { "grad_norm": 0.31089314818382263, "learning_rate": 1.9225794325426492e-05, "loss": 0.0121, "step": 58040 }, { "grad_norm": 0.16971606016159058, "learning_rate": 1.920950719180436e-05, "loss": 0.0099, "step": 58050 }, { "grad_norm": 0.27183106541633606, "learning_rate": 1.919322531943536e-05, "loss": 0.0139, "step": 58060 }, { "grad_norm": 0.23774223029613495, "learning_rate": 1.917694871110157e-05, "loss": 0.0137, "step": 58070 }, { "grad_norm": 0.4463646709918976, "learning_rate": 1.9160677369584234e-05, "loss": 0.0142, "step": 58080 }, { "grad_norm": 0.18483638763427734, "learning_rate": 1.9144411297663694e-05, "loss": 0.0145, "step": 58090 }, { "grad_norm": 1.1964261531829834, "learning_rate": 1.9128150498119328e-05, "loss": 0.0183, "step": 58100 }, { "grad_norm": 0.3507434129714966, "learning_rate": 1.9111894973729726e-05, "loss": 0.013, "step": 58110 }, { "grad_norm": 0.39268988370895386, "learning_rate": 1.9095644727272454e-05, "loss": 0.0187, "step": 58120 }, { "grad_norm": 0.36796802282333374, "learning_rate": 1.907939976152429e-05, "loss": 0.018, "step": 58130 }, { "grad_norm": 0.26421239972114563, "learning_rate": 1.906316007926101e-05, "loss": 0.0208, "step": 58140 }, { "grad_norm": 0.3879135251045227, "learning_rate": 1.904692568325755e-05, "loss": 0.0168, "step": 58150 }, { "grad_norm": 0.35114291310310364, "learning_rate": 1.9030696576287925e-05, "loss": 0.0159, "step": 58160 }, { "grad_norm": 0.30962812900543213, "learning_rate": 1.9014472761125242e-05, "loss": 0.0194, "step": 58170 }, { "grad_norm": 0.18334069848060608, "learning_rate": 1.899825424054172e-05, "loss": 0.0159, "step": 58180 }, { "grad_norm": 0.25852566957473755, "learning_rate": 1.898204101730863e-05, "loss": 0.016, "step": 58190 }, { "grad_norm": 0.28389084339141846, "learning_rate": 1.8965833094196394e-05, "loss": 0.0261, "step": 58200 }, { "grad_norm": 0.5174591541290283, "learning_rate": 1.8949630473974495e-05, "loss": 0.0134, "step": 58210 }, { "grad_norm": 0.23129884898662567, "learning_rate": 1.8933433159411517e-05, "loss": 0.0149, "step": 58220 }, { "grad_norm": 0.13030795753002167, "learning_rate": 1.891724115327514e-05, "loss": 0.0144, "step": 58230 }, { "grad_norm": 0.2202945500612259, "learning_rate": 1.8901054458332152e-05, "loss": 0.0187, "step": 58240 }, { "grad_norm": 0.21150435507297516, "learning_rate": 1.8884873077348364e-05, "loss": 0.0146, "step": 58250 }, { "grad_norm": 0.2504095733165741, "learning_rate": 1.88686970130888e-05, "loss": 0.0161, "step": 58260 }, { "grad_norm": 0.17640885710716248, "learning_rate": 1.8852526268317456e-05, "loss": 0.0162, "step": 58270 }, { "grad_norm": 0.2932445704936981, "learning_rate": 1.883636084579749e-05, "loss": 0.0151, "step": 58280 }, { "grad_norm": 0.25112614035606384, "learning_rate": 1.8820200748291145e-05, "loss": 0.0156, "step": 58290 }, { "grad_norm": 0.1923387050628662, "learning_rate": 1.8804045978559686e-05, "loss": 0.0151, "step": 58300 }, { "grad_norm": 0.23203326761722565, "learning_rate": 1.8787896539363594e-05, "loss": 0.0159, "step": 58310 }, { "grad_norm": 0.4442758858203888, "learning_rate": 1.8771752433462298e-05, "loss": 0.0147, "step": 58320 }, { "grad_norm": 0.21507813036441803, "learning_rate": 1.8755613663614447e-05, "loss": 0.0176, "step": 58330 }, { "grad_norm": 0.36264675855636597, "learning_rate": 1.873948023257767e-05, "loss": 0.0173, "step": 58340 }, { "grad_norm": 0.20710355043411255, "learning_rate": 1.8723352143108747e-05, "loss": 0.0111, "step": 58350 }, { "grad_norm": 0.2816162109375, "learning_rate": 1.870722939796352e-05, "loss": 0.019, "step": 58360 }, { "grad_norm": 0.21219289302825928, "learning_rate": 1.869111199989693e-05, "loss": 0.0151, "step": 58370 }, { "grad_norm": 0.2135143131017685, "learning_rate": 1.8674999951663018e-05, "loss": 0.0137, "step": 58380 }, { "grad_norm": 0.4972015917301178, "learning_rate": 1.8658893256014857e-05, "loss": 0.0166, "step": 58390 }, { "grad_norm": 0.16417650878429413, "learning_rate": 1.8642791915704655e-05, "loss": 0.0117, "step": 58400 }, { "grad_norm": 0.38125818967819214, "learning_rate": 1.8626695933483697e-05, "loss": 0.0137, "step": 58410 }, { "grad_norm": 0.2954283356666565, "learning_rate": 1.8610605312102335e-05, "loss": 0.0148, "step": 58420 }, { "grad_norm": 0.3516964018344879, "learning_rate": 1.8594520054310032e-05, "loss": 0.0177, "step": 58430 }, { "grad_norm": 0.22470848262310028, "learning_rate": 1.8578440162855326e-05, "loss": 0.0176, "step": 58440 }, { "grad_norm": 0.18889205157756805, "learning_rate": 1.8562365640485774e-05, "loss": 0.0152, "step": 58450 }, { "grad_norm": 0.198745459318161, "learning_rate": 1.854629648994815e-05, "loss": 0.0138, "step": 58460 }, { "grad_norm": 0.16525939106941223, "learning_rate": 1.8530232713988183e-05, "loss": 0.0147, "step": 58470 }, { "grad_norm": 0.3678939938545227, "learning_rate": 1.851417431535074e-05, "loss": 0.0175, "step": 58480 }, { "grad_norm": 0.20207731425762177, "learning_rate": 1.8498121296779785e-05, "loss": 0.0241, "step": 58490 }, { "grad_norm": 0.2752036452293396, "learning_rate": 1.848207366101829e-05, "loss": 0.0175, "step": 58500 }, { "grad_norm": 0.15329113602638245, "learning_rate": 1.8466031410808422e-05, "loss": 0.0169, "step": 58510 }, { "grad_norm": 0.3022865355014801, "learning_rate": 1.844999454889129e-05, "loss": 0.0154, "step": 58520 }, { "grad_norm": 0.239234060049057, "learning_rate": 1.8433963078007237e-05, "loss": 0.015, "step": 58530 }, { "grad_norm": 0.28593316674232483, "learning_rate": 1.841793700089554e-05, "loss": 0.017, "step": 58540 }, { "grad_norm": 0.26572364568710327, "learning_rate": 1.8401916320294644e-05, "loss": 0.0158, "step": 58550 }, { "grad_norm": 0.300114244222641, "learning_rate": 1.838590103894205e-05, "loss": 0.0157, "step": 58560 }, { "grad_norm": 0.11758382618427277, "learning_rate": 1.8369891159574292e-05, "loss": 0.0145, "step": 58570 }, { "grad_norm": 0.2636963725090027, "learning_rate": 1.835388668492708e-05, "loss": 0.0176, "step": 58580 }, { "grad_norm": 0.29554277658462524, "learning_rate": 1.8337887617735095e-05, "loss": 0.0137, "step": 58590 }, { "grad_norm": 0.14768972992897034, "learning_rate": 1.8321893960732157e-05, "loss": 0.0117, "step": 58600 }, { "grad_norm": 0.24450422823429108, "learning_rate": 1.8305905716651138e-05, "loss": 0.0166, "step": 58610 }, { "grad_norm": 0.5119179487228394, "learning_rate": 1.8289922888224e-05, "loss": 0.016, "step": 58620 }, { "grad_norm": 0.1765437126159668, "learning_rate": 1.8273945478181765e-05, "loss": 0.0153, "step": 58630 }, { "grad_norm": 0.2808144688606262, "learning_rate": 1.8257973489254558e-05, "loss": 0.0183, "step": 58640 }, { "grad_norm": 0.29349514842033386, "learning_rate": 1.824200692417152e-05, "loss": 0.0138, "step": 58650 }, { "grad_norm": 0.1731949895620346, "learning_rate": 1.8226045785660912e-05, "loss": 0.0126, "step": 58660 }, { "grad_norm": 0.43030232191085815, "learning_rate": 1.821009007645006e-05, "loss": 0.0154, "step": 58670 }, { "grad_norm": 0.3141326308250427, "learning_rate": 1.8194139799265357e-05, "loss": 0.0143, "step": 58680 }, { "grad_norm": 0.2981768846511841, "learning_rate": 1.8178194956832295e-05, "loss": 0.0167, "step": 58690 }, { "grad_norm": 0.15825216472148895, "learning_rate": 1.8162255551875346e-05, "loss": 0.0143, "step": 58700 }, { "grad_norm": 0.20233027637004852, "learning_rate": 1.8146321587118194e-05, "loss": 0.0193, "step": 58710 }, { "grad_norm": 0.29940930008888245, "learning_rate": 1.8130393065283448e-05, "loss": 0.0155, "step": 58720 }, { "grad_norm": 0.11624126881361008, "learning_rate": 1.8114469989092925e-05, "loss": 0.0124, "step": 58730 }, { "grad_norm": 0.24631863832473755, "learning_rate": 1.8098552361267397e-05, "loss": 0.0146, "step": 58740 }, { "grad_norm": 0.2009602189064026, "learning_rate": 1.8082640184526763e-05, "loss": 0.0168, "step": 58750 }, { "grad_norm": 0.28416064381599426, "learning_rate": 1.806673346158999e-05, "loss": 0.0195, "step": 58760 }, { "grad_norm": 0.24006348848342896, "learning_rate": 1.8050832195175067e-05, "loss": 0.0155, "step": 58770 }, { "grad_norm": 0.3380744159221649, "learning_rate": 1.8034936387999136e-05, "loss": 0.0173, "step": 58780 }, { "grad_norm": 0.2102806568145752, "learning_rate": 1.8019046042778315e-05, "loss": 0.0132, "step": 58790 }, { "grad_norm": 0.18192782998085022, "learning_rate": 1.800316116222785e-05, "loss": 0.0138, "step": 58800 }, { "grad_norm": 0.16445757448673248, "learning_rate": 1.7987281749062018e-05, "loss": 0.0337, "step": 58810 }, { "grad_norm": 0.2582811415195465, "learning_rate": 1.7971407805994195e-05, "loss": 0.0279, "step": 58820 }, { "grad_norm": 0.27296721935272217, "learning_rate": 1.7955539335736787e-05, "loss": 0.0147, "step": 58830 }, { "grad_norm": 0.1574074625968933, "learning_rate": 1.7939676341001304e-05, "loss": 0.0215, "step": 58840 }, { "grad_norm": 0.26377072930336, "learning_rate": 1.7923818824498275e-05, "loss": 0.0135, "step": 58850 }, { "grad_norm": 0.24826554954051971, "learning_rate": 1.7907966788937315e-05, "loss": 0.0156, "step": 58860 }, { "grad_norm": 0.18919187784194946, "learning_rate": 1.7892120237027116e-05, "loss": 0.0222, "step": 58870 }, { "grad_norm": 0.1804414540529251, "learning_rate": 1.7876279171475413e-05, "loss": 0.0138, "step": 58880 }, { "grad_norm": 0.16175906360149384, "learning_rate": 1.7860443594989028e-05, "loss": 0.0165, "step": 58890 }, { "grad_norm": 0.26109328866004944, "learning_rate": 1.784461351027379e-05, "loss": 0.0105, "step": 58900 }, { "grad_norm": 0.3547385632991791, "learning_rate": 1.7828788920034677e-05, "loss": 0.0228, "step": 58910 }, { "grad_norm": 0.31045249104499817, "learning_rate": 1.7812969826975623e-05, "loss": 0.0156, "step": 58920 }, { "grad_norm": 0.17454960942268372, "learning_rate": 1.7797156233799738e-05, "loss": 0.0144, "step": 58930 }, { "grad_norm": 0.17548628151416779, "learning_rate": 1.7781348143209094e-05, "loss": 0.0175, "step": 58940 }, { "grad_norm": 0.21509762108325958, "learning_rate": 1.7765545557904862e-05, "loss": 0.024, "step": 58950 }, { "grad_norm": 0.22300773859024048, "learning_rate": 1.7749748480587302e-05, "loss": 0.0194, "step": 58960 }, { "grad_norm": 0.17417378723621368, "learning_rate": 1.773395691395564e-05, "loss": 0.0138, "step": 58970 }, { "grad_norm": 0.15552064776420593, "learning_rate": 1.7718170860708305e-05, "loss": 0.0126, "step": 58980 }, { "grad_norm": 0.18401426076889038, "learning_rate": 1.770239032354264e-05, "loss": 0.0118, "step": 58990 }, { "grad_norm": 0.28059035539627075, "learning_rate": 1.7686615305155124e-05, "loss": 0.025, "step": 59000 }, { "grad_norm": 0.26247769594192505, "learning_rate": 1.767084580824128e-05, "loss": 0.013, "step": 59010 }, { "grad_norm": 0.44719958305358887, "learning_rate": 1.765508183549569e-05, "loss": 0.0171, "step": 59020 }, { "grad_norm": 0.21771062910556793, "learning_rate": 1.763932338961199e-05, "loss": 0.0152, "step": 59030 }, { "grad_norm": 0.3206771910190582, "learning_rate": 1.762357047328284e-05, "loss": 0.0105, "step": 59040 }, { "grad_norm": 0.22640343010425568, "learning_rate": 1.7607823089199997e-05, "loss": 0.0198, "step": 59050 }, { "grad_norm": 0.17663684487342834, "learning_rate": 1.7592081240054265e-05, "loss": 0.0195, "step": 59060 }, { "grad_norm": 0.26644742488861084, "learning_rate": 1.7576344928535488e-05, "loss": 0.0168, "step": 59070 }, { "grad_norm": 0.4158550500869751, "learning_rate": 1.7560614157332573e-05, "loss": 0.0196, "step": 59080 }, { "grad_norm": 0.25237104296684265, "learning_rate": 1.7544888929133495e-05, "loss": 0.0183, "step": 59090 }, { "grad_norm": 0.16233393549919128, "learning_rate": 1.752916924662522e-05, "loss": 0.0141, "step": 59100 }, { "grad_norm": 0.14305411279201508, "learning_rate": 1.751345511249387e-05, "loss": 0.0154, "step": 59110 }, { "grad_norm": 0.1508803516626358, "learning_rate": 1.749774652942452e-05, "loss": 0.0146, "step": 59120 }, { "grad_norm": 0.25831595063209534, "learning_rate": 1.748204350010135e-05, "loss": 0.0145, "step": 59130 }, { "grad_norm": 0.1422909051179886, "learning_rate": 1.7466346027207574e-05, "loss": 0.0144, "step": 59140 }, { "grad_norm": 0.1766703724861145, "learning_rate": 1.745065411342547e-05, "loss": 0.0144, "step": 59150 }, { "grad_norm": 0.24972373247146606, "learning_rate": 1.7434967761436366e-05, "loss": 0.0197, "step": 59160 }, { "grad_norm": 0.2895525395870209, "learning_rate": 1.741928697392058e-05, "loss": 0.0152, "step": 59170 }, { "grad_norm": 0.2400987446308136, "learning_rate": 1.7403611753557597e-05, "loss": 0.0132, "step": 59180 }, { "grad_norm": 0.1883925050497055, "learning_rate": 1.738794210302584e-05, "loss": 0.0116, "step": 59190 }, { "grad_norm": 0.12199573963880539, "learning_rate": 1.7372278025002837e-05, "loss": 0.0136, "step": 59200 }, { "grad_norm": 0.22333776950836182, "learning_rate": 1.7356619522165164e-05, "loss": 0.0181, "step": 59210 }, { "grad_norm": 0.151023268699646, "learning_rate": 1.7340966597188378e-05, "loss": 0.0113, "step": 59220 }, { "grad_norm": 1.0739330053329468, "learning_rate": 1.732531925274722e-05, "loss": 0.0147, "step": 59230 }, { "grad_norm": 0.24773339927196503, "learning_rate": 1.7309677491515318e-05, "loss": 0.0138, "step": 59240 }, { "grad_norm": 0.22709357738494873, "learning_rate": 1.7294041316165455e-05, "loss": 0.0125, "step": 59250 }, { "grad_norm": 0.27584120631217957, "learning_rate": 1.727841072936942e-05, "loss": 0.0176, "step": 59260 }, { "grad_norm": 0.3110584616661072, "learning_rate": 1.7262785733798058e-05, "loss": 0.0129, "step": 59270 }, { "grad_norm": 0.2929806113243103, "learning_rate": 1.7247166332121246e-05, "loss": 0.0156, "step": 59280 }, { "grad_norm": 0.3997170627117157, "learning_rate": 1.7231552527007933e-05, "loss": 0.0126, "step": 59290 }, { "grad_norm": 0.24349761009216309, "learning_rate": 1.721594432112606e-05, "loss": 0.013, "step": 59300 }, { "grad_norm": 0.24503877758979797, "learning_rate": 1.7200341717142655e-05, "loss": 0.0198, "step": 59310 }, { "grad_norm": 0.31639209389686584, "learning_rate": 1.718474471772377e-05, "loss": 0.0142, "step": 59320 }, { "grad_norm": 0.16918939352035522, "learning_rate": 1.7169153325534528e-05, "loss": 0.0132, "step": 59330 }, { "grad_norm": 0.24843516945838928, "learning_rate": 1.7153567543239045e-05, "loss": 0.0166, "step": 59340 }, { "grad_norm": 0.17130374908447266, "learning_rate": 1.7137987373500525e-05, "loss": 0.0114, "step": 59350 }, { "grad_norm": 0.1463262140750885, "learning_rate": 1.7122412818981198e-05, "loss": 0.0152, "step": 59360 }, { "grad_norm": 0.24340513348579407, "learning_rate": 1.7106843882342283e-05, "loss": 0.016, "step": 59370 }, { "grad_norm": 0.21743568778038025, "learning_rate": 1.709128056624415e-05, "loss": 0.0179, "step": 59380 }, { "grad_norm": 0.27672749757766724, "learning_rate": 1.70757228733461e-05, "loss": 0.0129, "step": 59390 }, { "grad_norm": 0.17529170215129852, "learning_rate": 1.706017080630653e-05, "loss": 0.0147, "step": 59400 }, { "grad_norm": 0.21185903251171112, "learning_rate": 1.7044624367782873e-05, "loss": 0.011, "step": 59410 }, { "grad_norm": 0.1946372240781784, "learning_rate": 1.7029083560431553e-05, "loss": 0.0239, "step": 59420 }, { "grad_norm": 0.2814110815525055, "learning_rate": 1.701354838690813e-05, "loss": 0.0169, "step": 59430 }, { "grad_norm": 0.28835731744766235, "learning_rate": 1.6998018849867087e-05, "loss": 0.0126, "step": 59440 }, { "grad_norm": 0.19247765839099884, "learning_rate": 1.698249495196202e-05, "loss": 0.0183, "step": 59450 }, { "grad_norm": 0.18705716729164124, "learning_rate": 1.6966976695845528e-05, "loss": 0.0138, "step": 59460 }, { "grad_norm": 0.1381300687789917, "learning_rate": 1.6951464084169268e-05, "loss": 0.0133, "step": 59470 }, { "grad_norm": 0.258078396320343, "learning_rate": 1.6935957119583916e-05, "loss": 0.0192, "step": 59480 }, { "grad_norm": 0.15836580097675323, "learning_rate": 1.6920455804739205e-05, "loss": 0.0134, "step": 59490 }, { "grad_norm": 0.1915072500705719, "learning_rate": 1.6904960142283855e-05, "loss": 0.0156, "step": 59500 }, { "grad_norm": 0.1830894500017166, "learning_rate": 1.6889470134865666e-05, "loss": 0.011, "step": 59510 }, { "grad_norm": 0.20248064398765564, "learning_rate": 1.687398578513145e-05, "loss": 0.0154, "step": 59520 }, { "grad_norm": 0.8387084007263184, "learning_rate": 1.6858507095727066e-05, "loss": 0.0173, "step": 59530 }, { "grad_norm": 0.22834432125091553, "learning_rate": 1.6843034069297403e-05, "loss": 0.0119, "step": 59540 }, { "grad_norm": 0.2636481821537018, "learning_rate": 1.682756670848637e-05, "loss": 0.0159, "step": 59550 }, { "grad_norm": 0.21626025438308716, "learning_rate": 1.6812105015936937e-05, "loss": 0.0113, "step": 59560 }, { "grad_norm": 0.435698002576828, "learning_rate": 1.6796648994291027e-05, "loss": 0.018, "step": 59570 }, { "grad_norm": 0.20777475833892822, "learning_rate": 1.678119864618973e-05, "loss": 0.0157, "step": 59580 }, { "grad_norm": 0.23509971797466278, "learning_rate": 1.6765753974273023e-05, "loss": 0.0151, "step": 59590 }, { "grad_norm": 0.3714315593242645, "learning_rate": 1.675031498118001e-05, "loss": 0.0115, "step": 59600 }, { "grad_norm": 0.2609054148197174, "learning_rate": 1.6734881669548804e-05, "loss": 0.0127, "step": 59610 }, { "grad_norm": 0.5662123560905457, "learning_rate": 1.671945404201648e-05, "loss": 0.0192, "step": 59620 }, { "grad_norm": 0.24439667165279388, "learning_rate": 1.670403210121927e-05, "loss": 0.0129, "step": 59630 }, { "grad_norm": 0.23992207646369934, "learning_rate": 1.6688615849792312e-05, "loss": 0.0108, "step": 59640 }, { "grad_norm": 0.3307933807373047, "learning_rate": 1.6673205290369832e-05, "loss": 0.0162, "step": 59650 }, { "grad_norm": 0.3474332392215729, "learning_rate": 1.6657800425585086e-05, "loss": 0.0175, "step": 59660 }, { "grad_norm": 0.2736380696296692, "learning_rate": 1.6642401258070327e-05, "loss": 0.0224, "step": 59670 }, { "grad_norm": 0.3746025860309601, "learning_rate": 1.6627007790456884e-05, "loss": 0.0152, "step": 59680 }, { "grad_norm": 0.15070568025112152, "learning_rate": 1.6611620025375037e-05, "loss": 0.0143, "step": 59690 }, { "grad_norm": 0.21679659187793732, "learning_rate": 1.6596237965454154e-05, "loss": 0.0171, "step": 59700 }, { "grad_norm": 0.3556162714958191, "learning_rate": 1.6580861613322607e-05, "loss": 0.0153, "step": 59710 }, { "grad_norm": 0.278968870639801, "learning_rate": 1.6565490971607796e-05, "loss": 0.0129, "step": 59720 }, { "grad_norm": 0.20517337322235107, "learning_rate": 1.655012604293614e-05, "loss": 0.0149, "step": 59730 }, { "grad_norm": 0.22305195033550262, "learning_rate": 1.653476682993309e-05, "loss": 0.0183, "step": 59740 }, { "grad_norm": 0.30310386419296265, "learning_rate": 1.651941333522311e-05, "loss": 0.0143, "step": 59750 }, { "grad_norm": 0.2726520597934723, "learning_rate": 1.6504065561429715e-05, "loss": 0.0141, "step": 59760 }, { "grad_norm": 0.23110368847846985, "learning_rate": 1.6488723511175385e-05, "loss": 0.0217, "step": 59770 }, { "grad_norm": 0.24004791676998138, "learning_rate": 1.6473387187081668e-05, "loss": 0.0173, "step": 59780 }, { "grad_norm": 0.21062734723091125, "learning_rate": 1.6458056591769123e-05, "loss": 0.0151, "step": 59790 }, { "grad_norm": 0.3172623813152313, "learning_rate": 1.6442731727857335e-05, "loss": 0.0162, "step": 59800 }, { "grad_norm": 0.21475572884082794, "learning_rate": 1.642741259796492e-05, "loss": 0.0118, "step": 59810 }, { "grad_norm": 0.2209998518228531, "learning_rate": 1.641209920470944e-05, "loss": 0.012, "step": 59820 }, { "grad_norm": 0.6189222931861877, "learning_rate": 1.639679155070762e-05, "loss": 0.014, "step": 59830 }, { "grad_norm": 0.25007790327072144, "learning_rate": 1.6381489638575048e-05, "loss": 0.0166, "step": 59840 }, { "grad_norm": 0.13653206825256348, "learning_rate": 1.636619347092643e-05, "loss": 0.0113, "step": 59850 }, { "grad_norm": 0.2887629568576813, "learning_rate": 1.6350903050375476e-05, "loss": 0.0166, "step": 59860 }, { "grad_norm": 0.30384889245033264, "learning_rate": 1.6335618379534856e-05, "loss": 0.0138, "step": 59870 }, { "grad_norm": 0.19401639699935913, "learning_rate": 1.6320339461016364e-05, "loss": 0.0122, "step": 59880 }, { "grad_norm": 0.31472495198249817, "learning_rate": 1.6305066297430687e-05, "loss": 0.0165, "step": 59890 }, { "grad_norm": 0.2561160624027252, "learning_rate": 1.6289798891387654e-05, "loss": 0.0197, "step": 59900 }, { "grad_norm": 0.2775033712387085, "learning_rate": 1.6274537245495995e-05, "loss": 0.013, "step": 59910 }, { "grad_norm": 0.2391381710767746, "learning_rate": 1.6259281362363527e-05, "loss": 0.0138, "step": 59920 }, { "grad_norm": 0.3508571982383728, "learning_rate": 1.6244031244597068e-05, "loss": 0.0219, "step": 59930 }, { "grad_norm": 0.2843906283378601, "learning_rate": 1.622878689480244e-05, "loss": 0.0129, "step": 59940 }, { "grad_norm": 0.22792750597000122, "learning_rate": 1.6213548315584498e-05, "loss": 0.0191, "step": 59950 }, { "grad_norm": 0.2628953754901886, "learning_rate": 1.6198315509547074e-05, "loss": 0.0158, "step": 59960 }, { "grad_norm": 0.14503376185894012, "learning_rate": 1.6183088479293056e-05, "loss": 0.0122, "step": 59970 }, { "grad_norm": 0.24918889999389648, "learning_rate": 1.6167867227424316e-05, "loss": 0.013, "step": 59980 }, { "grad_norm": 0.18500521779060364, "learning_rate": 1.615265175654176e-05, "loss": 0.0172, "step": 59990 }, { "grad_norm": 0.31257736682891846, "learning_rate": 1.613744206924529e-05, "loss": 0.0161, "step": 60000 }, { "grad_norm": 0.16967883706092834, "learning_rate": 1.6122238168133845e-05, "loss": 0.0163, "step": 60010 }, { "grad_norm": 0.16507647931575775, "learning_rate": 1.6107040055805305e-05, "loss": 0.0134, "step": 60020 }, { "grad_norm": 0.22796012461185455, "learning_rate": 1.6091847734856673e-05, "loss": 0.0126, "step": 60030 }, { "grad_norm": 0.2962973415851593, "learning_rate": 1.6076661207883865e-05, "loss": 0.0154, "step": 60040 }, { "grad_norm": 0.13604125380516052, "learning_rate": 1.6061480477481848e-05, "loss": 0.0129, "step": 60050 }, { "grad_norm": 0.2624184787273407, "learning_rate": 1.604630554624461e-05, "loss": 0.0158, "step": 60060 }, { "grad_norm": 0.324759840965271, "learning_rate": 1.603113641676509e-05, "loss": 0.0139, "step": 60070 }, { "grad_norm": 0.2519397735595703, "learning_rate": 1.6015973091635338e-05, "loss": 0.013, "step": 60080 }, { "grad_norm": 0.17977671325206757, "learning_rate": 1.6000815573446288e-05, "loss": 0.0167, "step": 60090 }, { "grad_norm": 0.35066112875938416, "learning_rate": 1.5985663864788002e-05, "loss": 0.0201, "step": 60100 }, { "grad_norm": 0.2865278720855713, "learning_rate": 1.597051796824946e-05, "loss": 0.0171, "step": 60110 }, { "grad_norm": 0.18694573640823364, "learning_rate": 1.5955377886418682e-05, "loss": 0.0134, "step": 60120 }, { "grad_norm": 0.24685382843017578, "learning_rate": 1.5940243621882704e-05, "loss": 0.0122, "step": 60130 }, { "grad_norm": 0.31479722261428833, "learning_rate": 1.5925115177227555e-05, "loss": 0.02, "step": 60140 }, { "grad_norm": 0.30666685104370117, "learning_rate": 1.5909992555038288e-05, "loss": 0.0171, "step": 60150 }, { "grad_norm": 0.20265795290470123, "learning_rate": 1.5894875757898912e-05, "loss": 0.022, "step": 60160 }, { "grad_norm": 0.18583691120147705, "learning_rate": 1.58797647883925e-05, "loss": 0.0152, "step": 60170 }, { "grad_norm": 0.3765733242034912, "learning_rate": 1.586465964910109e-05, "loss": 0.0157, "step": 60180 }, { "grad_norm": 0.23498867452144623, "learning_rate": 1.5849560342605734e-05, "loss": 0.0142, "step": 60190 }, { "grad_norm": 0.2828176021575928, "learning_rate": 1.583446687148651e-05, "loss": 0.0154, "step": 60200 }, { "grad_norm": 0.3001587688922882, "learning_rate": 1.581937923832248e-05, "loss": 0.0123, "step": 60210 }, { "grad_norm": 0.22199906408786774, "learning_rate": 1.5804297445691663e-05, "loss": 0.0153, "step": 60220 }, { "grad_norm": 0.22666455805301666, "learning_rate": 1.578922149617119e-05, "loss": 0.0149, "step": 60230 }, { "grad_norm": 0.18479007482528687, "learning_rate": 1.5774151392337084e-05, "loss": 0.0109, "step": 60240 }, { "grad_norm": 0.4121665358543396, "learning_rate": 1.575908713676442e-05, "loss": 0.014, "step": 60250 }, { "grad_norm": 0.4078825116157532, "learning_rate": 1.5744028732027293e-05, "loss": 0.0158, "step": 60260 }, { "grad_norm": 0.29301100969314575, "learning_rate": 1.5728976180698723e-05, "loss": 0.0172, "step": 60270 }, { "grad_norm": 0.16370655596256256, "learning_rate": 1.5713929485350836e-05, "loss": 0.017, "step": 60280 }, { "grad_norm": 0.17173080146312714, "learning_rate": 1.569888864855465e-05, "loss": 0.0125, "step": 60290 }, { "grad_norm": 0.2600894272327423, "learning_rate": 1.568385367288028e-05, "loss": 0.0112, "step": 60300 }, { "grad_norm": 0.26058194041252136, "learning_rate": 1.5668824560896755e-05, "loss": 0.0201, "step": 60310 }, { "grad_norm": 0.21045786142349243, "learning_rate": 1.5653801315172156e-05, "loss": 0.0135, "step": 60320 }, { "grad_norm": 0.24524542689323425, "learning_rate": 1.5638783938273554e-05, "loss": 0.0165, "step": 60330 }, { "grad_norm": 0.23048807680606842, "learning_rate": 1.5623772432766966e-05, "loss": 0.0189, "step": 60340 }, { "grad_norm": 0.300734281539917, "learning_rate": 1.5608766801217507e-05, "loss": 0.0202, "step": 60350 }, { "grad_norm": 0.11385609209537506, "learning_rate": 1.5593767046189184e-05, "loss": 0.0108, "step": 60360 }, { "grad_norm": 0.8098885416984558, "learning_rate": 1.5578773170245064e-05, "loss": 0.0158, "step": 60370 }, { "grad_norm": 0.3915046453475952, "learning_rate": 1.5563785175947182e-05, "loss": 0.0176, "step": 60380 }, { "grad_norm": 0.20671498775482178, "learning_rate": 1.5548803065856582e-05, "loss": 0.0136, "step": 60390 }, { "grad_norm": 0.38727936148643494, "learning_rate": 1.55338268425333e-05, "loss": 0.0148, "step": 60400 }, { "grad_norm": 0.2533993124961853, "learning_rate": 1.5518856508536373e-05, "loss": 0.0138, "step": 60410 }, { "grad_norm": 0.27365538477897644, "learning_rate": 1.5503892066423786e-05, "loss": 0.0159, "step": 60420 }, { "grad_norm": 0.38260453939437866, "learning_rate": 1.548893351875258e-05, "loss": 0.022, "step": 60430 }, { "grad_norm": 0.2652050256729126, "learning_rate": 1.547398086807875e-05, "loss": 0.0123, "step": 60440 }, { "grad_norm": 0.2724563181400299, "learning_rate": 1.5459034116957305e-05, "loss": 0.0221, "step": 60450 }, { "grad_norm": 0.2666417360305786, "learning_rate": 1.544409326794225e-05, "loss": 0.0255, "step": 60460 }, { "grad_norm": 0.25439462065696716, "learning_rate": 1.542915832358651e-05, "loss": 0.0161, "step": 60470 }, { "grad_norm": 0.1638629138469696, "learning_rate": 1.541422928644214e-05, "loss": 0.0117, "step": 60480 }, { "grad_norm": 0.2945355772972107, "learning_rate": 1.5399306159060024e-05, "loss": 0.0204, "step": 60490 }, { "grad_norm": 0.18763454258441925, "learning_rate": 1.5384388943990185e-05, "loss": 0.0123, "step": 60500 }, { "grad_norm": 0.2627585828304291, "learning_rate": 1.5369477643781526e-05, "loss": 0.012, "step": 60510 }, { "grad_norm": 0.2767098546028137, "learning_rate": 1.5354572260981985e-05, "loss": 0.0167, "step": 60520 }, { "grad_norm": 0.22460468113422394, "learning_rate": 1.533967279813851e-05, "loss": 0.0128, "step": 60530 }, { "grad_norm": 0.28007054328918457, "learning_rate": 1.5324779257796956e-05, "loss": 0.0131, "step": 60540 }, { "grad_norm": 0.2361796796321869, "learning_rate": 1.5309891642502293e-05, "loss": 0.0133, "step": 60550 }, { "grad_norm": 0.257339745759964, "learning_rate": 1.5295009954798357e-05, "loss": 0.0127, "step": 60560 }, { "grad_norm": 0.3212008774280548, "learning_rate": 1.5280134197228036e-05, "loss": 0.016, "step": 60570 }, { "grad_norm": 0.14551828801631927, "learning_rate": 1.526526437233319e-05, "loss": 0.017, "step": 60580 }, { "grad_norm": 0.12405551224946976, "learning_rate": 1.525040048265466e-05, "loss": 0.0133, "step": 60590 }, { "grad_norm": 0.3560001254081726, "learning_rate": 1.5235542530732288e-05, "loss": 0.0128, "step": 60600 }, { "grad_norm": 0.19682909548282623, "learning_rate": 1.5220690519104901e-05, "loss": 0.0151, "step": 60610 }, { "grad_norm": 0.18685372173786163, "learning_rate": 1.5205844450310275e-05, "loss": 0.0135, "step": 60620 }, { "grad_norm": 0.23109941184520721, "learning_rate": 1.519100432688521e-05, "loss": 0.014, "step": 60630 }, { "grad_norm": 0.2017749547958374, "learning_rate": 1.5176170151365476e-05, "loss": 0.0144, "step": 60640 }, { "grad_norm": 0.11371797323226929, "learning_rate": 1.5161341926285833e-05, "loss": 0.0107, "step": 60650 }, { "grad_norm": 0.21130183339118958, "learning_rate": 1.5146519654180025e-05, "loss": 0.0131, "step": 60660 }, { "grad_norm": 0.2924068570137024, "learning_rate": 1.5131703337580739e-05, "loss": 0.0188, "step": 60670 }, { "grad_norm": 0.1728462129831314, "learning_rate": 1.5116892979019731e-05, "loss": 0.017, "step": 60680 }, { "grad_norm": 0.15994660556316376, "learning_rate": 1.5102088581027623e-05, "loss": 0.0114, "step": 60690 }, { "grad_norm": 0.3956398069858551, "learning_rate": 1.5087290146134154e-05, "loss": 0.0126, "step": 60700 }, { "grad_norm": 0.20866724848747253, "learning_rate": 1.5072497676867915e-05, "loss": 0.0134, "step": 60710 }, { "grad_norm": 0.23229728639125824, "learning_rate": 1.505771117575655e-05, "loss": 0.0151, "step": 60720 }, { "grad_norm": 0.22442153096199036, "learning_rate": 1.5042930645326691e-05, "loss": 0.0127, "step": 60730 }, { "grad_norm": 0.2725278437137604, "learning_rate": 1.5028156088103879e-05, "loss": 0.0174, "step": 60740 }, { "grad_norm": 0.4431364834308624, "learning_rate": 1.5013387506612735e-05, "loss": 0.0191, "step": 60750 }, { "grad_norm": 0.19083812832832336, "learning_rate": 1.4998624903376767e-05, "loss": 0.015, "step": 60760 }, { "grad_norm": 0.6836031079292297, "learning_rate": 1.4983868280918517e-05, "loss": 0.0155, "step": 60770 }, { "grad_norm": 0.2588300108909607, "learning_rate": 1.4969117641759478e-05, "loss": 0.0185, "step": 60780 }, { "grad_norm": 0.21735996007919312, "learning_rate": 1.495437298842014e-05, "loss": 0.0162, "step": 60790 }, { "grad_norm": 0.2793368995189667, "learning_rate": 1.4939634323419976e-05, "loss": 0.0147, "step": 60800 }, { "grad_norm": 0.42748358845710754, "learning_rate": 1.4924901649277384e-05, "loss": 0.0196, "step": 60810 }, { "grad_norm": 0.44105958938598633, "learning_rate": 1.4910174968509793e-05, "loss": 0.0131, "step": 60820 }, { "grad_norm": 0.1788712441921234, "learning_rate": 1.4895454283633598e-05, "loss": 0.0126, "step": 60830 }, { "grad_norm": 0.4967018663883209, "learning_rate": 1.488073959716415e-05, "loss": 0.0167, "step": 60840 }, { "grad_norm": 0.26206424832344055, "learning_rate": 1.4866030911615791e-05, "loss": 0.0152, "step": 60850 }, { "grad_norm": 0.1565820425748825, "learning_rate": 1.4851328229501849e-05, "loss": 0.0124, "step": 60860 }, { "grad_norm": 0.28893306851387024, "learning_rate": 1.4836631553334562e-05, "loss": 0.0126, "step": 60870 }, { "grad_norm": 0.1816035360097885, "learning_rate": 1.4821940885625251e-05, "loss": 0.0124, "step": 60880 }, { "grad_norm": 0.46683329343795776, "learning_rate": 1.4807256228884109e-05, "loss": 0.0142, "step": 60890 }, { "grad_norm": 0.1424633264541626, "learning_rate": 1.4792577585620353e-05, "loss": 0.0124, "step": 60900 }, { "grad_norm": 0.2030237466096878, "learning_rate": 1.4777904958342164e-05, "loss": 0.0146, "step": 60910 }, { "grad_norm": 0.25106820464134216, "learning_rate": 1.4763238349556691e-05, "loss": 0.0099, "step": 60920 }, { "grad_norm": 0.1838584840297699, "learning_rate": 1.4748577761770072e-05, "loss": 0.0161, "step": 60930 }, { "grad_norm": 0.19258499145507812, "learning_rate": 1.4733923197487354e-05, "loss": 0.0131, "step": 60940 }, { "grad_norm": 0.23524954915046692, "learning_rate": 1.4719274659212662e-05, "loss": 0.0143, "step": 60950 }, { "grad_norm": 0.14615929126739502, "learning_rate": 1.4704632149448983e-05, "loss": 0.0148, "step": 60960 }, { "grad_norm": 0.17396770417690277, "learning_rate": 1.4689995670698343e-05, "loss": 0.0134, "step": 60970 }, { "grad_norm": 0.18624946475028992, "learning_rate": 1.4675365225461728e-05, "loss": 0.0149, "step": 60980 }, { "grad_norm": 0.36405253410339355, "learning_rate": 1.4660740816239032e-05, "loss": 0.011, "step": 60990 }, { "grad_norm": 0.288730651140213, "learning_rate": 1.464612244552923e-05, "loss": 0.0132, "step": 61000 }, { "grad_norm": 0.12728281319141388, "learning_rate": 1.4631510115830161e-05, "loss": 0.0114, "step": 61010 }, { "grad_norm": 0.2064526081085205, "learning_rate": 1.4616903829638679e-05, "loss": 0.0119, "step": 61020 }, { "grad_norm": 0.16380514204502106, "learning_rate": 1.46023035894506e-05, "loss": 0.0109, "step": 61030 }, { "grad_norm": 0.3452244997024536, "learning_rate": 1.4587709397760713e-05, "loss": 0.0158, "step": 61040 }, { "grad_norm": 0.24034298956394196, "learning_rate": 1.4573121257062755e-05, "loss": 0.0149, "step": 61050 }, { "grad_norm": 0.476027250289917, "learning_rate": 1.4558539169849472e-05, "loss": 0.0219, "step": 61060 }, { "grad_norm": 0.3169218599796295, "learning_rate": 1.4543963138612499e-05, "loss": 0.0135, "step": 61070 }, { "grad_norm": 0.19508050382137299, "learning_rate": 1.4529393165842498e-05, "loss": 0.0109, "step": 61080 }, { "grad_norm": 0.27181410789489746, "learning_rate": 1.4514829254029084e-05, "loss": 0.0123, "step": 61090 }, { "grad_norm": 0.1706039309501648, "learning_rate": 1.450027140566083e-05, "loss": 0.0122, "step": 61100 }, { "grad_norm": 0.29675060510635376, "learning_rate": 1.4485719623225268e-05, "loss": 0.0136, "step": 61110 }, { "grad_norm": 0.17195674777030945, "learning_rate": 1.4471173909208912e-05, "loss": 0.0141, "step": 61120 }, { "grad_norm": 0.34005206823349, "learning_rate": 1.4456634266097236e-05, "loss": 0.0199, "step": 61130 }, { "grad_norm": 0.3477203845977783, "learning_rate": 1.444210069637461e-05, "loss": 0.0187, "step": 61140 }, { "grad_norm": 0.6034073233604431, "learning_rate": 1.4427573202524502e-05, "loss": 0.0107, "step": 61150 }, { "grad_norm": 0.16268113255500793, "learning_rate": 1.4413051787029208e-05, "loss": 0.0194, "step": 61160 }, { "grad_norm": 0.18108820915222168, "learning_rate": 1.439853645237006e-05, "loss": 0.0173, "step": 61170 }, { "grad_norm": 0.1641942411661148, "learning_rate": 1.4384027201027344e-05, "loss": 0.0122, "step": 61180 }, { "grad_norm": 0.1963454633951187, "learning_rate": 1.4369524035480253e-05, "loss": 0.014, "step": 61190 }, { "grad_norm": 0.8139209747314453, "learning_rate": 1.4355026958207035e-05, "loss": 0.0132, "step": 61200 }, { "grad_norm": 0.24960267543792725, "learning_rate": 1.434053597168481e-05, "loss": 0.0137, "step": 61210 }, { "grad_norm": 0.3663712739944458, "learning_rate": 1.432605107838969e-05, "loss": 0.0177, "step": 61220 }, { "grad_norm": 0.2902981638908386, "learning_rate": 1.4311572280796759e-05, "loss": 0.0129, "step": 61230 }, { "grad_norm": 0.3140300214290619, "learning_rate": 1.4297099581380047e-05, "loss": 0.0218, "step": 61240 }, { "grad_norm": 0.3049159049987793, "learning_rate": 1.4282632982612538e-05, "loss": 0.0175, "step": 61250 }, { "grad_norm": 0.18393130600452423, "learning_rate": 1.4268172486966197e-05, "loss": 0.0189, "step": 61260 }, { "grad_norm": 0.1981292963027954, "learning_rate": 1.4253718096911894e-05, "loss": 0.0191, "step": 61270 }, { "grad_norm": 0.2737196385860443, "learning_rate": 1.4239269814919504e-05, "loss": 0.02, "step": 61280 }, { "grad_norm": 0.24036744236946106, "learning_rate": 1.4224827643457843e-05, "loss": 0.0177, "step": 61290 }, { "grad_norm": 0.21837031841278076, "learning_rate": 1.4210391584994686e-05, "loss": 0.013, "step": 61300 }, { "grad_norm": 0.2677915394306183, "learning_rate": 1.4195961641996763e-05, "loss": 0.0132, "step": 61310 }, { "grad_norm": 0.17745257914066315, "learning_rate": 1.4181537816929751e-05, "loss": 0.0134, "step": 61320 }, { "grad_norm": 0.33220794796943665, "learning_rate": 1.4167120112258303e-05, "loss": 0.0182, "step": 61330 }, { "grad_norm": 0.1342563033103943, "learning_rate": 1.4152708530445963e-05, "loss": 0.0136, "step": 61340 }, { "grad_norm": 0.20147167146205902, "learning_rate": 1.4138303073955344e-05, "loss": 0.0141, "step": 61350 }, { "grad_norm": 0.26225876808166504, "learning_rate": 1.4123903745247897e-05, "loss": 0.0141, "step": 61360 }, { "grad_norm": 0.12666839361190796, "learning_rate": 1.4109510546784082e-05, "loss": 0.0106, "step": 61370 }, { "grad_norm": 0.40027642250061035, "learning_rate": 1.4095123481023325e-05, "loss": 0.0212, "step": 61380 }, { "grad_norm": 0.17498256266117096, "learning_rate": 1.4080742550423936e-05, "loss": 0.0122, "step": 61390 }, { "grad_norm": 0.16302211582660675, "learning_rate": 1.4066367757443278e-05, "loss": 0.0146, "step": 61400 }, { "grad_norm": 0.40281152725219727, "learning_rate": 1.4051999104537572e-05, "loss": 0.0163, "step": 61410 }, { "grad_norm": 0.3744615316390991, "learning_rate": 1.4037636594162034e-05, "loss": 0.0151, "step": 61420 }, { "grad_norm": 0.3108070194721222, "learning_rate": 1.402328022877083e-05, "loss": 0.0093, "step": 61430 }, { "grad_norm": 0.22755227982997894, "learning_rate": 1.4008930010817073e-05, "loss": 0.0203, "step": 61440 }, { "grad_norm": 0.24139007925987244, "learning_rate": 1.3994585942752831e-05, "loss": 0.01, "step": 61450 }, { "grad_norm": 0.3332999348640442, "learning_rate": 1.398024802702909e-05, "loss": 0.0147, "step": 61460 }, { "grad_norm": 0.2850643992424011, "learning_rate": 1.3965916266095819e-05, "loss": 0.0155, "step": 61470 }, { "grad_norm": 0.3362695276737213, "learning_rate": 1.395159066240192e-05, "loss": 0.0128, "step": 61480 }, { "grad_norm": 0.18616025149822235, "learning_rate": 1.3937271218395248e-05, "loss": 0.0092, "step": 61490 }, { "grad_norm": 0.21920417249202728, "learning_rate": 1.3922957936522613e-05, "loss": 0.0177, "step": 61500 }, { "grad_norm": 0.3490198850631714, "learning_rate": 1.3908650819229758e-05, "loss": 0.0143, "step": 61510 }, { "grad_norm": 0.5404391884803772, "learning_rate": 1.389434986896137e-05, "loss": 0.0128, "step": 61520 }, { "grad_norm": 0.14325155317783356, "learning_rate": 1.388005508816112e-05, "loss": 0.0093, "step": 61530 }, { "grad_norm": 0.2461075782775879, "learning_rate": 1.3865766479271557e-05, "loss": 0.0172, "step": 61540 }, { "grad_norm": 0.1823623776435852, "learning_rate": 1.3851484044734225e-05, "loss": 0.0168, "step": 61550 }, { "grad_norm": 0.18155403435230255, "learning_rate": 1.3837207786989608e-05, "loss": 0.0239, "step": 61560 }, { "grad_norm": 0.22474126517772675, "learning_rate": 1.382293770847713e-05, "loss": 0.0152, "step": 61570 }, { "grad_norm": 0.22892341017723083, "learning_rate": 1.3808673811635159e-05, "loss": 0.0122, "step": 61580 }, { "grad_norm": 0.25498321652412415, "learning_rate": 1.3794416098900975e-05, "loss": 0.0159, "step": 61590 }, { "grad_norm": 0.2868068218231201, "learning_rate": 1.3780164572710886e-05, "loss": 0.0218, "step": 61600 }, { "grad_norm": 0.2037365585565567, "learning_rate": 1.3765919235500035e-05, "loss": 0.0181, "step": 61610 }, { "grad_norm": 0.19116684794425964, "learning_rate": 1.375168008970259e-05, "loss": 0.0122, "step": 61620 }, { "grad_norm": 0.2797824740409851, "learning_rate": 1.3737447137751635e-05, "loss": 0.0111, "step": 61630 }, { "grad_norm": 0.25058266520500183, "learning_rate": 1.372322038207915e-05, "loss": 0.0095, "step": 61640 }, { "grad_norm": 0.270185261964798, "learning_rate": 1.3708999825116159e-05, "loss": 0.0194, "step": 61650 }, { "grad_norm": 0.26840585470199585, "learning_rate": 1.3694785469292526e-05, "loss": 0.0207, "step": 61660 }, { "grad_norm": 0.30150172114372253, "learning_rate": 1.3680577317037101e-05, "loss": 0.0163, "step": 61670 }, { "grad_norm": 0.2311825156211853, "learning_rate": 1.3666375370777678e-05, "loss": 0.0126, "step": 61680 }, { "grad_norm": 0.2953733503818512, "learning_rate": 1.3652179632940981e-05, "loss": 0.0137, "step": 61690 }, { "grad_norm": 0.12413538247346878, "learning_rate": 1.3637990105952664e-05, "loss": 0.0147, "step": 61700 }, { "grad_norm": 0.21351510286331177, "learning_rate": 1.3623806792237337e-05, "loss": 0.0166, "step": 61710 }, { "grad_norm": 0.157749742269516, "learning_rate": 1.3609629694218551e-05, "loss": 0.0125, "step": 61720 }, { "grad_norm": 0.25384730100631714, "learning_rate": 1.3595458814318762e-05, "loss": 0.0113, "step": 61730 }, { "grad_norm": 0.38066715002059937, "learning_rate": 1.3581294154959389e-05, "loss": 0.0182, "step": 61740 }, { "grad_norm": 0.31726422905921936, "learning_rate": 1.3567135718560792e-05, "loss": 0.0132, "step": 61750 }, { "grad_norm": 0.14579327404499054, "learning_rate": 1.3552983507542261e-05, "loss": 0.0118, "step": 61760 }, { "grad_norm": 0.4875801205635071, "learning_rate": 1.3538837524322023e-05, "loss": 0.0187, "step": 61770 }, { "grad_norm": 0.10303390026092529, "learning_rate": 1.3524697771317251e-05, "loss": 0.0149, "step": 61780 }, { "grad_norm": 0.21640001237392426, "learning_rate": 1.3510564250943997e-05, "loss": 0.0144, "step": 61790 }, { "grad_norm": 0.15352784097194672, "learning_rate": 1.3496436965617353e-05, "loss": 0.0139, "step": 61800 }, { "grad_norm": 0.37558358907699585, "learning_rate": 1.3482315917751243e-05, "loss": 0.0179, "step": 61810 }, { "grad_norm": 0.31692907214164734, "learning_rate": 1.3468201109758582e-05, "loss": 0.0195, "step": 61820 }, { "grad_norm": 0.37862473726272583, "learning_rate": 1.3454092544051222e-05, "loss": 0.0133, "step": 61830 }, { "grad_norm": 0.2383601814508438, "learning_rate": 1.3439990223039878e-05, "loss": 0.0119, "step": 61840 }, { "grad_norm": 0.23337958753108978, "learning_rate": 1.3425894149134321e-05, "loss": 0.0126, "step": 61850 }, { "grad_norm": 0.31630101799964905, "learning_rate": 1.3411804324743132e-05, "loss": 0.0206, "step": 61860 }, { "grad_norm": 0.24873405694961548, "learning_rate": 1.3397720752273896e-05, "loss": 0.0135, "step": 61870 }, { "grad_norm": 0.5410347580909729, "learning_rate": 1.3383643434133108e-05, "loss": 0.0121, "step": 61880 }, { "grad_norm": 0.3010287284851074, "learning_rate": 1.3369572372726196e-05, "loss": 0.0151, "step": 61890 }, { "grad_norm": 0.21059659123420715, "learning_rate": 1.3355507570457525e-05, "loss": 0.011, "step": 61900 }, { "grad_norm": 0.21177054941654205, "learning_rate": 1.334144902973038e-05, "loss": 0.0118, "step": 61910 }, { "grad_norm": 0.2625563144683838, "learning_rate": 1.3327396752946997e-05, "loss": 0.0149, "step": 61920 }, { "grad_norm": 0.24339668452739716, "learning_rate": 1.3313350742508496e-05, "loss": 0.0165, "step": 61930 }, { "grad_norm": 0.1845693290233612, "learning_rate": 1.3299311000814973e-05, "loss": 0.0191, "step": 61940 }, { "grad_norm": 0.10530567914247513, "learning_rate": 1.3285277530265432e-05, "loss": 0.0184, "step": 61950 }, { "grad_norm": 0.5710243582725525, "learning_rate": 1.3271250333257813e-05, "loss": 0.0199, "step": 61960 }, { "grad_norm": 0.35665297508239746, "learning_rate": 1.3257229412188977e-05, "loss": 0.0121, "step": 61970 }, { "grad_norm": 0.2641836702823639, "learning_rate": 1.3243214769454726e-05, "loss": 0.0169, "step": 61980 }, { "grad_norm": 0.15460892021656036, "learning_rate": 1.3229206407449751e-05, "loss": 0.0121, "step": 61990 }, { "grad_norm": 0.19611862301826477, "learning_rate": 1.3215204328567742e-05, "loss": 0.0112, "step": 62000 }, { "grad_norm": 0.16762013733386993, "learning_rate": 1.3201208535201232e-05, "loss": 0.0119, "step": 62010 }, { "grad_norm": 0.17542435228824615, "learning_rate": 1.3187219029741737e-05, "loss": 0.0137, "step": 62020 }, { "grad_norm": 0.30926987528800964, "learning_rate": 1.3173235814579693e-05, "loss": 0.0113, "step": 62030 }, { "grad_norm": 0.22392931580543518, "learning_rate": 1.3159258892104398e-05, "loss": 0.0148, "step": 62040 }, { "grad_norm": 0.1955718696117401, "learning_rate": 1.3145288264704198e-05, "loss": 0.0159, "step": 62050 }, { "grad_norm": 0.3306247293949127, "learning_rate": 1.3131323934766237e-05, "loss": 0.0136, "step": 62060 }, { "grad_norm": 0.14690843224525452, "learning_rate": 1.311736590467666e-05, "loss": 0.0105, "step": 62070 }, { "grad_norm": 0.1719086617231369, "learning_rate": 1.3103414176820505e-05, "loss": 0.015, "step": 62080 }, { "grad_norm": 0.2334456592798233, "learning_rate": 1.308946875358174e-05, "loss": 0.0226, "step": 62090 }, { "grad_norm": 0.3222976326942444, "learning_rate": 1.3075529637343276e-05, "loss": 0.0192, "step": 62100 }, { "grad_norm": 0.25736117362976074, "learning_rate": 1.3061596830486883e-05, "loss": 0.0154, "step": 62110 }, { "grad_norm": 0.21123841404914856, "learning_rate": 1.3047670335393353e-05, "loss": 0.009, "step": 62120 }, { "grad_norm": 0.21168453991413116, "learning_rate": 1.3033750154442298e-05, "loss": 0.0151, "step": 62130 }, { "grad_norm": 0.15149568021297455, "learning_rate": 1.3019836290012316e-05, "loss": 0.011, "step": 62140 }, { "grad_norm": 0.2089943140745163, "learning_rate": 1.30059287444809e-05, "loss": 0.0198, "step": 62150 }, { "grad_norm": 0.2908168435096741, "learning_rate": 1.299202752022447e-05, "loss": 0.0234, "step": 62160 }, { "grad_norm": 0.18434837460517883, "learning_rate": 1.2978132619618371e-05, "loss": 0.0127, "step": 62170 }, { "grad_norm": 0.2162005752325058, "learning_rate": 1.2964244045036866e-05, "loss": 0.0195, "step": 62180 }, { "grad_norm": 0.15432319045066833, "learning_rate": 1.295036179885311e-05, "loss": 0.0105, "step": 62190 }, { "grad_norm": 0.11482543498277664, "learning_rate": 1.293648588343922e-05, "loss": 0.0142, "step": 62200 }, { "grad_norm": 0.3036695122718811, "learning_rate": 1.2922616301166196e-05, "loss": 0.0139, "step": 62210 }, { "grad_norm": 0.28735047578811646, "learning_rate": 1.2908753054403976e-05, "loss": 0.0114, "step": 62220 }, { "grad_norm": 0.2671095132827759, "learning_rate": 1.2894896145521429e-05, "loss": 0.0133, "step": 62230 }, { "grad_norm": 0.23387785255908966, "learning_rate": 1.2881045576886275e-05, "loss": 0.0141, "step": 62240 }, { "grad_norm": 0.1937682181596756, "learning_rate": 1.2867201350865254e-05, "loss": 0.0166, "step": 62250 }, { "grad_norm": 0.21643297374248505, "learning_rate": 1.2853363469823914e-05, "loss": 0.0131, "step": 62260 }, { "grad_norm": 0.2286914587020874, "learning_rate": 1.283953193612682e-05, "loss": 0.0118, "step": 62270 }, { "grad_norm": 0.19644272327423096, "learning_rate": 1.2825706752137372e-05, "loss": 0.0135, "step": 62280 }, { "grad_norm": 0.19585464894771576, "learning_rate": 1.2811887920217896e-05, "loss": 0.0155, "step": 62290 }, { "grad_norm": 0.1182856559753418, "learning_rate": 1.2798075442729707e-05, "loss": 0.0133, "step": 62300 }, { "grad_norm": 0.20255732536315918, "learning_rate": 1.2784269322032922e-05, "loss": 0.0138, "step": 62310 }, { "grad_norm": 0.23174598813056946, "learning_rate": 1.2770469560486686e-05, "loss": 0.0151, "step": 62320 }, { "grad_norm": 0.1337282359600067, "learning_rate": 1.2756676160448956e-05, "loss": 0.0149, "step": 62330 }, { "grad_norm": 0.2112371176481247, "learning_rate": 1.2742889124276663e-05, "loss": 0.017, "step": 62340 }, { "grad_norm": 0.2748241126537323, "learning_rate": 1.2729108454325639e-05, "loss": 0.0165, "step": 62350 }, { "grad_norm": 0.257996529340744, "learning_rate": 1.2715334152950614e-05, "loss": 0.0167, "step": 62360 }, { "grad_norm": 0.16266101598739624, "learning_rate": 1.2701566222505246e-05, "loss": 0.0109, "step": 62370 }, { "grad_norm": 0.31709596514701843, "learning_rate": 1.2687804665342107e-05, "loss": 0.0162, "step": 62380 }, { "grad_norm": 0.29359880089759827, "learning_rate": 1.267404948381265e-05, "loss": 0.0135, "step": 62390 }, { "grad_norm": 0.21717685461044312, "learning_rate": 1.2660300680267267e-05, "loss": 0.01, "step": 62400 }, { "grad_norm": 0.34094706177711487, "learning_rate": 1.2646558257055257e-05, "loss": 0.0141, "step": 62410 }, { "grad_norm": 0.13079951703548431, "learning_rate": 1.263282221652482e-05, "loss": 0.0083, "step": 62420 }, { "grad_norm": 0.15242283046245575, "learning_rate": 1.2619092561023088e-05, "loss": 0.0141, "step": 62430 }, { "grad_norm": 0.1581144779920578, "learning_rate": 1.2605369292896036e-05, "loss": 0.014, "step": 62440 }, { "grad_norm": 0.11445272713899612, "learning_rate": 1.2591652414488658e-05, "loss": 0.0143, "step": 62450 }, { "grad_norm": 0.22641055285930634, "learning_rate": 1.2577941928144732e-05, "loss": 0.0125, "step": 62460 }, { "grad_norm": 0.4110051989555359, "learning_rate": 1.256423783620706e-05, "loss": 0.0134, "step": 62470 }, { "grad_norm": 0.24873222410678864, "learning_rate": 1.2550540141017264e-05, "loss": 0.0142, "step": 62480 }, { "grad_norm": 0.21999526023864746, "learning_rate": 1.253684884491591e-05, "loss": 0.0103, "step": 62490 }, { "grad_norm": 0.25953835248947144, "learning_rate": 1.2523163950242483e-05, "loss": 0.013, "step": 62500 }, { "grad_norm": 0.33654290437698364, "learning_rate": 1.2509485459335313e-05, "loss": 0.0133, "step": 62510 }, { "grad_norm": 0.22079092264175415, "learning_rate": 1.2495813374531739e-05, "loss": 0.0161, "step": 62520 }, { "grad_norm": 0.3905636668205261, "learning_rate": 1.2482147698167907e-05, "loss": 0.0113, "step": 62530 }, { "grad_norm": 0.18375205993652344, "learning_rate": 1.2468488432578911e-05, "loss": 0.0117, "step": 62540 }, { "grad_norm": 0.46265938878059387, "learning_rate": 1.2454835580098761e-05, "loss": 0.0136, "step": 62550 }, { "grad_norm": 0.18900781869888306, "learning_rate": 1.2441189143060338e-05, "loss": 0.0128, "step": 62560 }, { "grad_norm": 0.3112533390522003, "learning_rate": 1.242754912379548e-05, "loss": 0.0131, "step": 62570 }, { "grad_norm": 0.2535235285758972, "learning_rate": 1.2413915524634844e-05, "loss": 0.0272, "step": 62580 }, { "grad_norm": 0.2336231917142868, "learning_rate": 1.2400288347908073e-05, "loss": 0.0173, "step": 62590 }, { "grad_norm": 0.3986072838306427, "learning_rate": 1.2386667595943663e-05, "loss": 0.0149, "step": 62600 }, { "grad_norm": 0.2362554967403412, "learning_rate": 1.2373053271069035e-05, "loss": 0.0122, "step": 62610 }, { "grad_norm": 0.22597040235996246, "learning_rate": 1.2359445375610501e-05, "loss": 0.012, "step": 62620 }, { "grad_norm": 0.19646713137626648, "learning_rate": 1.2345843911893301e-05, "loss": 0.0123, "step": 62630 }, { "grad_norm": 0.3248133659362793, "learning_rate": 1.2332248882241498e-05, "loss": 0.0237, "step": 62640 }, { "grad_norm": 0.26770180463790894, "learning_rate": 1.2318660288978178e-05, "loss": 0.013, "step": 62650 }, { "grad_norm": 0.2429155856370926, "learning_rate": 1.2305078134425213e-05, "loss": 0.0127, "step": 62660 }, { "grad_norm": 0.18691542744636536, "learning_rate": 1.2291502420903434e-05, "loss": 0.0147, "step": 62670 }, { "grad_norm": 0.23563817143440247, "learning_rate": 1.2277933150732567e-05, "loss": 0.0138, "step": 62680 }, { "grad_norm": 0.15093903243541718, "learning_rate": 1.2264370326231216e-05, "loss": 0.015, "step": 62690 }, { "grad_norm": 0.1642000824213028, "learning_rate": 1.2250813949716927e-05, "loss": 0.0153, "step": 62700 }, { "grad_norm": 0.30683499574661255, "learning_rate": 1.2237264023506063e-05, "loss": 0.015, "step": 62710 }, { "grad_norm": 0.11110567301511765, "learning_rate": 1.2223720549913987e-05, "loss": 0.0139, "step": 62720 }, { "grad_norm": 0.30144113302230835, "learning_rate": 1.221018353125487e-05, "loss": 0.0125, "step": 62730 }, { "grad_norm": 0.15386821329593658, "learning_rate": 1.2196652969841837e-05, "loss": 0.014, "step": 62740 }, { "grad_norm": 0.3522981107234955, "learning_rate": 1.2183128867986904e-05, "loss": 0.0157, "step": 62750 }, { "grad_norm": 0.16690382361412048, "learning_rate": 1.2169611228000927e-05, "loss": 0.012, "step": 62760 }, { "grad_norm": 0.1706807017326355, "learning_rate": 1.2156100052193752e-05, "loss": 0.0118, "step": 62770 }, { "grad_norm": 0.270387202501297, "learning_rate": 1.214259534287403e-05, "loss": 0.0104, "step": 62780 }, { "grad_norm": 0.1124551072716713, "learning_rate": 1.2129097102349363e-05, "loss": 0.0148, "step": 62790 }, { "grad_norm": 0.5783519744873047, "learning_rate": 1.2115605332926227e-05, "loss": 0.013, "step": 62800 }, { "grad_norm": 0.15035255253314972, "learning_rate": 1.2102120036910003e-05, "loss": 0.0218, "step": 62810 }, { "grad_norm": 0.21863846480846405, "learning_rate": 1.2088641216604956e-05, "loss": 0.0148, "step": 62820 }, { "grad_norm": 0.20896530151367188, "learning_rate": 1.2075168874314264e-05, "loss": 0.0126, "step": 62830 }, { "grad_norm": 0.17618554830551147, "learning_rate": 1.2061703012339942e-05, "loss": 0.0118, "step": 62840 }, { "grad_norm": 0.27782994508743286, "learning_rate": 1.204824363298297e-05, "loss": 0.0134, "step": 62850 }, { "grad_norm": 0.22001147270202637, "learning_rate": 1.2034790738543173e-05, "loss": 0.0139, "step": 62860 }, { "grad_norm": 0.168801948428154, "learning_rate": 1.2021344331319284e-05, "loss": 0.0123, "step": 62870 }, { "grad_norm": 0.16049887239933014, "learning_rate": 1.200790441360894e-05, "loss": 0.0141, "step": 62880 }, { "grad_norm": 0.13638579845428467, "learning_rate": 1.199447098770864e-05, "loss": 0.0121, "step": 62890 }, { "grad_norm": 0.19831182062625885, "learning_rate": 1.198104405591381e-05, "loss": 0.0223, "step": 62900 }, { "grad_norm": 0.1757093220949173, "learning_rate": 1.1967623620518697e-05, "loss": 0.011, "step": 62910 }, { "grad_norm": 0.3260781168937683, "learning_rate": 1.1954209683816554e-05, "loss": 0.0149, "step": 62920 }, { "grad_norm": 0.2010943740606308, "learning_rate": 1.1940802248099402e-05, "loss": 0.0127, "step": 62930 }, { "grad_norm": 0.27560698986053467, "learning_rate": 1.1927401315658232e-05, "loss": 0.0154, "step": 62940 }, { "grad_norm": 0.3979473114013672, "learning_rate": 1.1914006888782898e-05, "loss": 0.0158, "step": 62950 }, { "grad_norm": 0.23284663259983063, "learning_rate": 1.1900618969762106e-05, "loss": 0.0154, "step": 62960 }, { "grad_norm": 0.31168147921562195, "learning_rate": 1.188723756088354e-05, "loss": 0.0123, "step": 62970 }, { "grad_norm": 0.1821615993976593, "learning_rate": 1.1873862664433672e-05, "loss": 0.0107, "step": 62980 }, { "grad_norm": 0.2290107160806656, "learning_rate": 1.1860494282697927e-05, "loss": 0.0102, "step": 62990 }, { "grad_norm": 0.2378922402858734, "learning_rate": 1.1847132417960588e-05, "loss": 0.0107, "step": 63000 }, { "grad_norm": 0.2871287167072296, "learning_rate": 1.1833777072504832e-05, "loss": 0.0146, "step": 63010 }, { "grad_norm": 0.20944200456142426, "learning_rate": 1.1820428248612731e-05, "loss": 0.0117, "step": 63020 }, { "grad_norm": 0.3853001296520233, "learning_rate": 1.1807085948565245e-05, "loss": 0.0157, "step": 63030 }, { "grad_norm": 0.23767495155334473, "learning_rate": 1.1793750174642172e-05, "loss": 0.013, "step": 63040 }, { "grad_norm": 0.2505250573158264, "learning_rate": 1.1780420929122254e-05, "loss": 0.0162, "step": 63050 }, { "grad_norm": 0.205285906791687, "learning_rate": 1.176709821428309e-05, "loss": 0.0176, "step": 63060 }, { "grad_norm": 0.3247770667076111, "learning_rate": 1.1753782032401173e-05, "loss": 0.0191, "step": 63070 }, { "grad_norm": 0.3059360086917877, "learning_rate": 1.1740472385751866e-05, "loss": 0.016, "step": 63080 }, { "grad_norm": 0.19697003066539764, "learning_rate": 1.172716927660943e-05, "loss": 0.0106, "step": 63090 }, { "grad_norm": 0.31546831130981445, "learning_rate": 1.1713872707247015e-05, "loss": 0.018, "step": 63100 }, { "grad_norm": 0.16323895752429962, "learning_rate": 1.1700582679936595e-05, "loss": 0.0139, "step": 63110 }, { "grad_norm": 0.3000403940677643, "learning_rate": 1.1687299196949136e-05, "loss": 0.0204, "step": 63120 }, { "grad_norm": 0.42066702246665955, "learning_rate": 1.1674022260554374e-05, "loss": 0.0148, "step": 63130 }, { "grad_norm": 0.20233023166656494, "learning_rate": 1.1660751873020987e-05, "loss": 0.0154, "step": 63140 }, { "grad_norm": 0.4071878492832184, "learning_rate": 1.1647488036616538e-05, "loss": 0.0166, "step": 63150 }, { "grad_norm": 0.16514454782009125, "learning_rate": 1.1634230753607417e-05, "loss": 0.0183, "step": 63160 }, { "grad_norm": 0.17876040935516357, "learning_rate": 1.1620980026258982e-05, "loss": 0.0141, "step": 63170 }, { "grad_norm": 0.1631121188402176, "learning_rate": 1.1607735856835373e-05, "loss": 0.0107, "step": 63180 }, { "grad_norm": 0.1530112773180008, "learning_rate": 1.1594498247599677e-05, "loss": 0.0115, "step": 63190 }, { "grad_norm": 0.33543768525123596, "learning_rate": 1.158126720081384e-05, "loss": 0.0189, "step": 63200 }, { "grad_norm": 0.3603689968585968, "learning_rate": 1.156804271873868e-05, "loss": 0.0153, "step": 63210 }, { "grad_norm": 0.3191321790218353, "learning_rate": 1.1554824803633924e-05, "loss": 0.0158, "step": 63220 }, { "grad_norm": 0.3321809470653534, "learning_rate": 1.154161345775811e-05, "loss": 0.0096, "step": 63230 }, { "grad_norm": 0.16204650700092316, "learning_rate": 1.1528408683368724e-05, "loss": 0.0134, "step": 63240 }, { "grad_norm": 0.3197104036808014, "learning_rate": 1.1515210482722088e-05, "loss": 0.0158, "step": 63250 }, { "grad_norm": 0.10363064706325531, "learning_rate": 1.150201885807342e-05, "loss": 0.0193, "step": 63260 }, { "grad_norm": 0.2285495400428772, "learning_rate": 1.1488833811676807e-05, "loss": 0.0119, "step": 63270 }, { "grad_norm": 0.18665966391563416, "learning_rate": 1.1475655345785213e-05, "loss": 0.0156, "step": 63280 }, { "grad_norm": 0.19663989543914795, "learning_rate": 1.1462483462650481e-05, "loss": 0.0126, "step": 63290 }, { "grad_norm": 0.21509383618831635, "learning_rate": 1.1449318164523331e-05, "loss": 0.016, "step": 63300 }, { "grad_norm": 0.2879549264907837, "learning_rate": 1.1436159453653334e-05, "loss": 0.0115, "step": 63310 }, { "grad_norm": 0.2607254087924957, "learning_rate": 1.1423007332288955e-05, "loss": 0.0149, "step": 63320 }, { "grad_norm": 0.2816573977470398, "learning_rate": 1.1409861802677546e-05, "loss": 0.0124, "step": 63330 }, { "grad_norm": 0.19557879865169525, "learning_rate": 1.1396722867065313e-05, "loss": 0.0118, "step": 63340 }, { "grad_norm": 0.2160761058330536, "learning_rate": 1.1383590527697352e-05, "loss": 0.0155, "step": 63350 }, { "grad_norm": 0.2014501690864563, "learning_rate": 1.137046478681758e-05, "loss": 0.0109, "step": 63360 }, { "grad_norm": 0.2642011344432831, "learning_rate": 1.1357345646668888e-05, "loss": 0.0266, "step": 63370 }, { "grad_norm": 0.23725777864456177, "learning_rate": 1.1344233109492924e-05, "loss": 0.0185, "step": 63380 }, { "grad_norm": 0.22233226895332336, "learning_rate": 1.1331127177530292e-05, "loss": 0.0142, "step": 63390 }, { "grad_norm": 0.29090917110443115, "learning_rate": 1.1318027853020441e-05, "loss": 0.012, "step": 63400 }, { "grad_norm": 0.5111983418464661, "learning_rate": 1.1304935138201645e-05, "loss": 0.0133, "step": 63410 }, { "grad_norm": 0.2085438370704651, "learning_rate": 1.1291849035311153e-05, "loss": 0.0129, "step": 63420 }, { "grad_norm": 0.2644292712211609, "learning_rate": 1.1278769546584972e-05, "loss": 0.0136, "step": 63430 }, { "grad_norm": 0.26224756240844727, "learning_rate": 1.1265696674258052e-05, "loss": 0.0105, "step": 63440 }, { "grad_norm": 0.21250616014003754, "learning_rate": 1.1252630420564186e-05, "loss": 0.0149, "step": 63450 }, { "grad_norm": 0.37553834915161133, "learning_rate": 1.1239570787736036e-05, "loss": 0.0172, "step": 63460 }, { "grad_norm": 0.4771874248981476, "learning_rate": 1.1226517778005135e-05, "loss": 0.0212, "step": 63470 }, { "grad_norm": 0.3587695062160492, "learning_rate": 1.1213471393601893e-05, "loss": 0.0129, "step": 63480 }, { "grad_norm": 0.34104594588279724, "learning_rate": 1.1200431636755587e-05, "loss": 0.0212, "step": 63490 }, { "grad_norm": 0.31018897891044617, "learning_rate": 1.1187398509694336e-05, "loss": 0.0181, "step": 63500 }, { "grad_norm": 0.1657567173242569, "learning_rate": 1.1174372014645146e-05, "loss": 0.0117, "step": 63510 }, { "grad_norm": 0.1822197586297989, "learning_rate": 1.1161352153833899e-05, "loss": 0.0183, "step": 63520 }, { "grad_norm": 0.16281703114509583, "learning_rate": 1.1148338929485325e-05, "loss": 0.014, "step": 63530 }, { "grad_norm": 0.21412716805934906, "learning_rate": 1.113533234382304e-05, "loss": 0.0151, "step": 63540 }, { "grad_norm": 0.21755124628543854, "learning_rate": 1.1122332399069513e-05, "loss": 0.0128, "step": 63550 }, { "grad_norm": 0.23223146796226501, "learning_rate": 1.1109339097446047e-05, "loss": 0.0179, "step": 63560 }, { "grad_norm": 0.24836768209934235, "learning_rate": 1.1096352441172897e-05, "loss": 0.0157, "step": 63570 }, { "grad_norm": 0.2749805748462677, "learning_rate": 1.1083372432469086e-05, "loss": 0.0114, "step": 63580 }, { "grad_norm": 0.1954602748155594, "learning_rate": 1.107039907355255e-05, "loss": 0.0183, "step": 63590 }, { "grad_norm": 0.3237098753452301, "learning_rate": 1.1057432366640103e-05, "loss": 0.012, "step": 63600 }, { "grad_norm": 0.3193657696247101, "learning_rate": 1.1044472313947352e-05, "loss": 0.0177, "step": 63610 }, { "grad_norm": 0.27244868874549866, "learning_rate": 1.1031518917688877e-05, "loss": 0.017, "step": 63620 }, { "grad_norm": 0.34881526231765747, "learning_rate": 1.1018572180078007e-05, "loss": 0.0215, "step": 63630 }, { "grad_norm": 0.18093542754650116, "learning_rate": 1.1005632103327018e-05, "loss": 0.0118, "step": 63640 }, { "grad_norm": 0.1620943248271942, "learning_rate": 1.0992698689646996e-05, "loss": 0.0118, "step": 63650 }, { "grad_norm": 0.21549350023269653, "learning_rate": 1.0979771941247919e-05, "loss": 0.0171, "step": 63660 }, { "grad_norm": 0.19935283064842224, "learning_rate": 1.0966851860338611e-05, "loss": 0.0143, "step": 63670 }, { "grad_norm": 0.15596981346607208, "learning_rate": 1.0953938449126766e-05, "loss": 0.0133, "step": 63680 }, { "grad_norm": 0.41281312704086304, "learning_rate": 1.0941031709818933e-05, "loss": 0.0237, "step": 63690 }, { "grad_norm": 0.17532449960708618, "learning_rate": 1.0928131644620509e-05, "loss": 0.0233, "step": 63700 }, { "grad_norm": 0.2299247831106186, "learning_rate": 1.0915238255735766e-05, "loss": 0.0124, "step": 63710 }, { "grad_norm": 0.1572919636964798, "learning_rate": 1.0902351545367833e-05, "loss": 0.0138, "step": 63720 }, { "grad_norm": 0.28876498341560364, "learning_rate": 1.0889471515718702e-05, "loss": 0.0131, "step": 63730 }, { "grad_norm": 0.19287002086639404, "learning_rate": 1.087659816898921e-05, "loss": 0.012, "step": 63740 }, { "grad_norm": 0.22680975496768951, "learning_rate": 1.0863731507379082e-05, "loss": 0.0127, "step": 63750 }, { "grad_norm": 0.2568276524543762, "learning_rate": 1.0850871533086827e-05, "loss": 0.0138, "step": 63760 }, { "grad_norm": 0.28911465406417847, "learning_rate": 1.0838018248309927e-05, "loss": 0.0117, "step": 63770 }, { "grad_norm": 0.34733977913856506, "learning_rate": 1.0825171655244615e-05, "loss": 0.0171, "step": 63780 }, { "grad_norm": 0.22817592322826385, "learning_rate": 1.0812331756086025e-05, "loss": 0.0129, "step": 63790 }, { "grad_norm": 0.25618743896484375, "learning_rate": 1.079949855302817e-05, "loss": 0.0127, "step": 63800 }, { "grad_norm": 0.2718304395675659, "learning_rate": 1.0786672048263852e-05, "loss": 0.0139, "step": 63810 }, { "grad_norm": 0.3001168668270111, "learning_rate": 1.0773852243984817e-05, "loss": 0.0164, "step": 63820 }, { "grad_norm": 0.2288248986005783, "learning_rate": 1.0761039142381586e-05, "loss": 0.0151, "step": 63830 }, { "grad_norm": 0.1913755238056183, "learning_rate": 1.0748232745643577e-05, "loss": 0.0143, "step": 63840 }, { "grad_norm": 0.2556373178958893, "learning_rate": 1.0735433055959055e-05, "loss": 0.0151, "step": 63850 }, { "grad_norm": 0.23326542973518372, "learning_rate": 1.0722640075515133e-05, "loss": 0.0129, "step": 63860 }, { "grad_norm": 0.16678711771965027, "learning_rate": 1.0709853806497795e-05, "loss": 0.0117, "step": 63870 }, { "grad_norm": 0.34048113226890564, "learning_rate": 1.0697074251091831e-05, "loss": 0.0173, "step": 63880 }, { "grad_norm": 0.21887357532978058, "learning_rate": 1.0684301411480962e-05, "loss": 0.0147, "step": 63890 }, { "grad_norm": 0.2568177878856659, "learning_rate": 1.067153528984769e-05, "loss": 0.0142, "step": 63900 }, { "grad_norm": 0.20573893189430237, "learning_rate": 1.0658775888373395e-05, "loss": 0.0151, "step": 63910 }, { "grad_norm": 0.1589937061071396, "learning_rate": 1.0646023209238314e-05, "loss": 0.0144, "step": 63920 }, { "grad_norm": 0.14883139729499817, "learning_rate": 1.0633277254621537e-05, "loss": 0.0118, "step": 63930 }, { "grad_norm": 0.16406702995300293, "learning_rate": 1.0620538026700994e-05, "loss": 0.0137, "step": 63940 }, { "grad_norm": 0.20775535702705383, "learning_rate": 1.0607805527653486e-05, "loss": 0.0133, "step": 63950 }, { "grad_norm": 0.2742469310760498, "learning_rate": 1.059507975965462e-05, "loss": 0.0136, "step": 63960 }, { "grad_norm": 0.2035197913646698, "learning_rate": 1.0582360724878898e-05, "loss": 0.0116, "step": 63970 }, { "grad_norm": 0.26890820264816284, "learning_rate": 1.0569648425499651e-05, "loss": 0.0164, "step": 63980 }, { "grad_norm": 0.16372479498386383, "learning_rate": 1.0556942863689063e-05, "loss": 0.0128, "step": 63990 }, { "grad_norm": 0.327936589717865, "learning_rate": 1.054424404161819e-05, "loss": 0.0109, "step": 64000 }, { "grad_norm": 0.13981670141220093, "learning_rate": 1.053155196145686e-05, "loss": 0.0119, "step": 64010 }, { "grad_norm": 0.26990336179733276, "learning_rate": 1.0518866625373863e-05, "loss": 0.0116, "step": 64020 }, { "grad_norm": 0.11706584692001343, "learning_rate": 1.0506188035536735e-05, "loss": 0.0187, "step": 64030 }, { "grad_norm": 0.12888775765895844, "learning_rate": 1.0493516194111919e-05, "loss": 0.0163, "step": 64040 }, { "grad_norm": 0.3835656940937042, "learning_rate": 1.0480851103264688e-05, "loss": 0.0126, "step": 64050 }, { "grad_norm": 0.36640089750289917, "learning_rate": 1.046819276515913e-05, "loss": 0.014, "step": 64060 }, { "grad_norm": 0.28061333298683167, "learning_rate": 1.0455541181958256e-05, "loss": 0.0113, "step": 64070 }, { "grad_norm": 0.6023801565170288, "learning_rate": 1.0442896355823822e-05, "loss": 0.0162, "step": 64080 }, { "grad_norm": 0.21574585139751434, "learning_rate": 1.0430258288916539e-05, "loss": 0.0118, "step": 64090 }, { "grad_norm": 0.10302980244159698, "learning_rate": 1.0417626983395868e-05, "loss": 0.014, "step": 64100 }, { "grad_norm": 0.18068116903305054, "learning_rate": 1.0405002441420165e-05, "loss": 0.0136, "step": 64110 }, { "grad_norm": 0.19679801166057587, "learning_rate": 1.039238466514662e-05, "loss": 0.0127, "step": 64120 }, { "grad_norm": 0.2156357765197754, "learning_rate": 1.0379773656731262e-05, "loss": 0.0148, "step": 64130 }, { "grad_norm": 0.2877911329269409, "learning_rate": 1.0367169418328986e-05, "loss": 0.0128, "step": 64140 }, { "grad_norm": 0.4111930727958679, "learning_rate": 1.0354571952093484e-05, "loss": 0.0155, "step": 64150 }, { "grad_norm": 0.20941677689552307, "learning_rate": 1.034198126017733e-05, "loss": 0.0146, "step": 64160 }, { "grad_norm": 0.2414843589067459, "learning_rate": 1.0329397344731928e-05, "loss": 0.0134, "step": 64170 }, { "grad_norm": 0.23378804326057434, "learning_rate": 1.0316820207907524e-05, "loss": 0.0174, "step": 64180 }, { "grad_norm": 0.47789475321769714, "learning_rate": 1.0304249851853215e-05, "loss": 0.0133, "step": 64190 }, { "grad_norm": 0.25974342226982117, "learning_rate": 1.0291686278716933e-05, "loss": 0.011, "step": 64200 }, { "grad_norm": 0.16767925024032593, "learning_rate": 1.0279129490645418e-05, "loss": 0.0114, "step": 64210 }, { "grad_norm": 0.3001171350479126, "learning_rate": 1.0266579489784328e-05, "loss": 0.0142, "step": 64220 }, { "grad_norm": 0.22773167490959167, "learning_rate": 1.0254036278278084e-05, "loss": 0.0131, "step": 64230 }, { "grad_norm": 0.20450803637504578, "learning_rate": 1.0241499858269982e-05, "loss": 0.0096, "step": 64240 }, { "grad_norm": 0.39142876863479614, "learning_rate": 1.0228970231902169e-05, "loss": 0.0153, "step": 64250 }, { "grad_norm": 0.2784750163555145, "learning_rate": 1.0216447401315582e-05, "loss": 0.0132, "step": 64260 }, { "grad_norm": 0.2449691742658615, "learning_rate": 1.020393136865007e-05, "loss": 0.0112, "step": 64270 }, { "grad_norm": 0.38198333978652954, "learning_rate": 1.0191422136044242e-05, "loss": 0.0217, "step": 64280 }, { "grad_norm": 0.2260729968547821, "learning_rate": 1.017891970563563e-05, "loss": 0.0117, "step": 64290 }, { "grad_norm": 0.19897758960723877, "learning_rate": 1.0166424079560516e-05, "loss": 0.0156, "step": 64300 }, { "grad_norm": 0.20646817982196808, "learning_rate": 1.0153935259954078e-05, "loss": 0.0102, "step": 64310 }, { "grad_norm": 0.16288834810256958, "learning_rate": 1.0141453248950311e-05, "loss": 0.0149, "step": 64320 }, { "grad_norm": 0.19850555062294006, "learning_rate": 1.0128978048682054e-05, "loss": 0.0165, "step": 64330 }, { "grad_norm": 0.32609400153160095, "learning_rate": 1.0116509661280982e-05, "loss": 0.0121, "step": 64340 }, { "grad_norm": 0.19791951775550842, "learning_rate": 1.0104048088877576e-05, "loss": 0.0097, "step": 64350 }, { "grad_norm": 0.29476460814476013, "learning_rate": 1.0091593333601201e-05, "loss": 0.0125, "step": 64360 }, { "grad_norm": 0.3538481593132019, "learning_rate": 1.0079145397580031e-05, "loss": 0.0181, "step": 64370 }, { "grad_norm": 0.3797991871833801, "learning_rate": 1.006670428294107e-05, "loss": 0.0191, "step": 64380 }, { "grad_norm": 0.19441832602024078, "learning_rate": 1.0054269991810166e-05, "loss": 0.0146, "step": 64390 }, { "grad_norm": 0.17625446617603302, "learning_rate": 1.0041842526312024e-05, "loss": 0.0117, "step": 64400 }, { "grad_norm": 0.2185794860124588, "learning_rate": 1.0029421888570101e-05, "loss": 0.0113, "step": 64410 }, { "grad_norm": 0.29058122634887695, "learning_rate": 1.0017008080706813e-05, "loss": 0.022, "step": 64420 }, { "grad_norm": 0.5613530278205872, "learning_rate": 1.0004601104843287e-05, "loss": 0.0117, "step": 64430 }, { "grad_norm": 0.39191290736198425, "learning_rate": 9.992200963099562e-06, "loss": 0.0155, "step": 64440 }, { "grad_norm": 0.1327633261680603, "learning_rate": 9.979807657594486e-06, "loss": 0.0138, "step": 64450 }, { "grad_norm": 0.25837260484695435, "learning_rate": 9.967421190445703e-06, "loss": 0.0129, "step": 64460 }, { "grad_norm": 0.18993952870368958, "learning_rate": 9.955041563769769e-06, "loss": 0.0125, "step": 64470 }, { "grad_norm": 0.5233603119850159, "learning_rate": 9.942668779681974e-06, "loss": 0.0138, "step": 64480 }, { "grad_norm": 0.24455446004867554, "learning_rate": 9.930302840296541e-06, "loss": 0.014, "step": 64490 }, { "grad_norm": 0.15343579649925232, "learning_rate": 9.917943747726426e-06, "loss": 0.0157, "step": 64500 }, { "grad_norm": 0.5597707629203796, "learning_rate": 9.905591504083484e-06, "loss": 0.0153, "step": 64510 }, { "grad_norm": 0.2652439773082733, "learning_rate": 9.893246111478382e-06, "loss": 0.0144, "step": 64520 }, { "grad_norm": 0.42001044750213623, "learning_rate": 9.88090757202057e-06, "loss": 0.0126, "step": 64530 }, { "grad_norm": 0.2283959537744522, "learning_rate": 9.868575887818421e-06, "loss": 0.0202, "step": 64540 }, { "grad_norm": 0.5279597043991089, "learning_rate": 9.856251060979044e-06, "loss": 0.0132, "step": 64550 }, { "grad_norm": 0.21281881630420685, "learning_rate": 9.843933093608426e-06, "loss": 0.0157, "step": 64560 }, { "grad_norm": 0.13257308304309845, "learning_rate": 9.831621987811368e-06, "loss": 0.0153, "step": 64570 }, { "grad_norm": 0.4126738905906677, "learning_rate": 9.819317745691509e-06, "loss": 0.012, "step": 64580 }, { "grad_norm": 0.17498226463794708, "learning_rate": 9.8070203693513e-06, "loss": 0.011, "step": 64590 }, { "grad_norm": 0.25189873576164246, "learning_rate": 9.794729860892048e-06, "loss": 0.0133, "step": 64600 }, { "grad_norm": 0.18398258090019226, "learning_rate": 9.782446222413827e-06, "loss": 0.0175, "step": 64610 }, { "grad_norm": 0.40703195333480835, "learning_rate": 9.770169456015598e-06, "loss": 0.0167, "step": 64620 }, { "grad_norm": 0.09259866178035736, "learning_rate": 9.75789956379512e-06, "loss": 0.0108, "step": 64630 }, { "grad_norm": 0.23519684374332428, "learning_rate": 9.74563654784898e-06, "loss": 0.0121, "step": 64640 }, { "grad_norm": 0.1942194402217865, "learning_rate": 9.733380410272596e-06, "loss": 0.0124, "step": 64650 }, { "grad_norm": 0.2052263468503952, "learning_rate": 9.721131153160207e-06, "loss": 0.0125, "step": 64660 }, { "grad_norm": 0.14412173628807068, "learning_rate": 9.708888778604886e-06, "loss": 0.01, "step": 64670 }, { "grad_norm": 0.2004493921995163, "learning_rate": 9.696653288698488e-06, "loss": 0.01, "step": 64680 }, { "grad_norm": 0.20416627824306488, "learning_rate": 9.684424685531763e-06, "loss": 0.0142, "step": 64690 }, { "grad_norm": 0.23964889347553253, "learning_rate": 9.672202971194216e-06, "loss": 0.0196, "step": 64700 }, { "grad_norm": 0.17076300084590912, "learning_rate": 9.659988147774213e-06, "loss": 0.0163, "step": 64710 }, { "grad_norm": 0.21590928733348846, "learning_rate": 9.647780217358942e-06, "loss": 0.0126, "step": 64720 }, { "grad_norm": 0.1956978291273117, "learning_rate": 9.635579182034376e-06, "loss": 0.0098, "step": 64730 }, { "grad_norm": 0.36173215508461, "learning_rate": 9.623385043885386e-06, "loss": 0.0183, "step": 64740 }, { "grad_norm": 0.18482615053653717, "learning_rate": 9.61119780499557e-06, "loss": 0.0101, "step": 64750 }, { "grad_norm": 0.18094095587730408, "learning_rate": 9.599017467447418e-06, "loss": 0.0096, "step": 64760 }, { "grad_norm": 0.14242151379585266, "learning_rate": 9.586844033322206e-06, "loss": 0.0107, "step": 64770 }, { "grad_norm": 0.3561282753944397, "learning_rate": 9.574677504700052e-06, "loss": 0.0171, "step": 64780 }, { "grad_norm": 0.16963735222816467, "learning_rate": 9.562517883659877e-06, "loss": 0.0187, "step": 64790 }, { "grad_norm": 0.25343161821365356, "learning_rate": 9.55036517227944e-06, "loss": 0.0125, "step": 64800 }, { "grad_norm": 2.5417356491088867, "learning_rate": 9.538219372635282e-06, "loss": 0.0122, "step": 64810 }, { "grad_norm": 0.30991825461387634, "learning_rate": 9.526080486802802e-06, "loss": 0.0154, "step": 64820 }, { "grad_norm": 0.20226146280765533, "learning_rate": 9.513948516856203e-06, "loss": 0.0126, "step": 64830 }, { "grad_norm": 0.12444043904542923, "learning_rate": 9.501823464868504e-06, "loss": 0.0103, "step": 64840 }, { "grad_norm": 0.1404520571231842, "learning_rate": 9.489705332911547e-06, "loss": 0.0117, "step": 64850 }, { "grad_norm": 0.26798513531684875, "learning_rate": 9.477594123055994e-06, "loss": 0.0115, "step": 64860 }, { "grad_norm": 0.1787731647491455, "learning_rate": 9.465489837371321e-06, "loss": 0.018, "step": 64870 }, { "grad_norm": 0.22743314504623413, "learning_rate": 9.453392477925794e-06, "loss": 0.0122, "step": 64880 }, { "grad_norm": 0.18821322917938232, "learning_rate": 9.441302046786566e-06, "loss": 0.013, "step": 64890 }, { "grad_norm": 1.2804267406463623, "learning_rate": 9.429218546019519e-06, "loss": 0.0165, "step": 64900 }, { "grad_norm": 0.37917548418045044, "learning_rate": 9.41714197768941e-06, "loss": 0.0189, "step": 64910 }, { "grad_norm": 0.21444237232208252, "learning_rate": 9.405072343859805e-06, "loss": 0.0112, "step": 64920 }, { "grad_norm": 0.31122639775276184, "learning_rate": 9.393009646593043e-06, "loss": 0.013, "step": 64930 }, { "grad_norm": 0.22704216837882996, "learning_rate": 9.38095388795035e-06, "loss": 0.0126, "step": 64940 }, { "grad_norm": 0.3956202566623688, "learning_rate": 9.36890506999169e-06, "loss": 0.0127, "step": 64950 }, { "grad_norm": 0.14873506128787994, "learning_rate": 9.356863194775894e-06, "loss": 0.0135, "step": 64960 }, { "grad_norm": 0.29024985432624817, "learning_rate": 9.344828264360583e-06, "loss": 0.0153, "step": 64970 }, { "grad_norm": 0.27282625436782837, "learning_rate": 9.332800280802201e-06, "loss": 0.0137, "step": 64980 }, { "grad_norm": 0.14130899310112, "learning_rate": 9.32077924615602e-06, "loss": 0.0104, "step": 64990 }, { "grad_norm": 0.22580787539482117, "learning_rate": 9.308765162476063e-06, "loss": 0.0119, "step": 65000 }, { "grad_norm": 0.20055940747261047, "learning_rate": 9.296758031815239e-06, "loss": 0.0171, "step": 65010 }, { "grad_norm": 0.36211302876472473, "learning_rate": 9.284757856225229e-06, "loss": 0.0112, "step": 65020 }, { "grad_norm": 0.1411832571029663, "learning_rate": 9.272764637756538e-06, "loss": 0.0139, "step": 65030 }, { "grad_norm": 0.264912873506546, "learning_rate": 9.260778378458479e-06, "loss": 0.0158, "step": 65040 }, { "grad_norm": 0.14857472479343414, "learning_rate": 9.248799080379172e-06, "loss": 0.014, "step": 65050 }, { "grad_norm": 0.1843366175889969, "learning_rate": 9.236826745565558e-06, "loss": 0.0176, "step": 65060 }, { "grad_norm": 0.212405726313591, "learning_rate": 9.224861376063388e-06, "loss": 0.0172, "step": 65070 }, { "grad_norm": 0.2447606474161148, "learning_rate": 9.212902973917192e-06, "loss": 0.0109, "step": 65080 }, { "grad_norm": 0.15447169542312622, "learning_rate": 9.20095154117035e-06, "loss": 0.0122, "step": 65090 }, { "grad_norm": 0.35453733801841736, "learning_rate": 9.189007079865036e-06, "loss": 0.0148, "step": 65100 }, { "grad_norm": 0.22093971073627472, "learning_rate": 9.177069592042226e-06, "loss": 0.0138, "step": 65110 }, { "grad_norm": 0.3184112012386322, "learning_rate": 9.165139079741724e-06, "loss": 0.0178, "step": 65120 }, { "grad_norm": 0.2627743184566498, "learning_rate": 9.153215545002098e-06, "loss": 0.0179, "step": 65130 }, { "grad_norm": 0.1831582933664322, "learning_rate": 9.141298989860798e-06, "loss": 0.0168, "step": 65140 }, { "grad_norm": 0.06736632436513901, "learning_rate": 9.129389416353994e-06, "loss": 0.0102, "step": 65150 }, { "grad_norm": 0.24571509659290314, "learning_rate": 9.11748682651673e-06, "loss": 0.0121, "step": 65160 }, { "grad_norm": 0.177368625998497, "learning_rate": 9.105591222382837e-06, "loss": 0.012, "step": 65170 }, { "grad_norm": 0.1865001916885376, "learning_rate": 9.093702605984915e-06, "loss": 0.0158, "step": 65180 }, { "grad_norm": 0.2340279519557953, "learning_rate": 9.081820979354455e-06, "loss": 0.0115, "step": 65190 }, { "grad_norm": 0.27146634459495544, "learning_rate": 9.069946344521663e-06, "loss": 0.0116, "step": 65200 }, { "grad_norm": 0.22359587252140045, "learning_rate": 9.058078703515598e-06, "loss": 0.0114, "step": 65210 }, { "grad_norm": 0.28488826751708984, "learning_rate": 9.046218058364125e-06, "loss": 0.0209, "step": 65220 }, { "grad_norm": 0.42794811725616455, "learning_rate": 9.034364411093893e-06, "loss": 0.0128, "step": 65230 }, { "grad_norm": 0.13992495834827423, "learning_rate": 9.022517763730371e-06, "loss": 0.0147, "step": 65240 }, { "grad_norm": 0.40801799297332764, "learning_rate": 9.010678118297827e-06, "loss": 0.0153, "step": 65250 }, { "grad_norm": 0.2466808259487152, "learning_rate": 8.998845476819345e-06, "loss": 0.0142, "step": 65260 }, { "grad_norm": 0.2558814287185669, "learning_rate": 8.987019841316773e-06, "loss": 0.0116, "step": 65270 }, { "grad_norm": 0.23581285774707794, "learning_rate": 8.975201213810802e-06, "loss": 0.01, "step": 65280 }, { "grad_norm": 0.2829042673110962, "learning_rate": 8.963389596320915e-06, "loss": 0.0135, "step": 65290 }, { "grad_norm": 0.10302476584911346, "learning_rate": 8.951584990865391e-06, "loss": 0.0122, "step": 65300 }, { "grad_norm": 0.28990811109542847, "learning_rate": 8.939787399461319e-06, "loss": 0.0128, "step": 65310 }, { "grad_norm": 0.4752906262874603, "learning_rate": 8.927996824124591e-06, "loss": 0.0141, "step": 65320 }, { "grad_norm": 0.2979363799095154, "learning_rate": 8.916213266869854e-06, "loss": 0.0156, "step": 65330 }, { "grad_norm": 0.1933094710111618, "learning_rate": 8.904436729710658e-06, "loss": 0.016, "step": 65340 }, { "grad_norm": 0.4239846467971802, "learning_rate": 8.892667214659245e-06, "loss": 0.01, "step": 65350 }, { "grad_norm": 0.19375842809677124, "learning_rate": 8.880904723726713e-06, "loss": 0.0115, "step": 65360 }, { "grad_norm": 0.20488856732845306, "learning_rate": 8.869149258922971e-06, "loss": 0.0141, "step": 65370 }, { "grad_norm": 0.2620696425437927, "learning_rate": 8.857400822256662e-06, "loss": 0.011, "step": 65380 }, { "grad_norm": 0.10093910992145538, "learning_rate": 8.845659415735324e-06, "loss": 0.0178, "step": 65390 }, { "grad_norm": 0.18322089314460754, "learning_rate": 8.83392504136521e-06, "loss": 0.0149, "step": 65400 }, { "grad_norm": 0.09381907433271408, "learning_rate": 8.822197701151407e-06, "loss": 0.0121, "step": 65410 }, { "grad_norm": 0.17260242998600006, "learning_rate": 8.810477397097804e-06, "loss": 0.0093, "step": 65420 }, { "grad_norm": 0.1517702341079712, "learning_rate": 8.798764131207077e-06, "loss": 0.0147, "step": 65430 }, { "grad_norm": 0.18063709139823914, "learning_rate": 8.787057905480706e-06, "loss": 0.0133, "step": 65440 }, { "grad_norm": 0.2466394603252411, "learning_rate": 8.775358721918958e-06, "loss": 0.0108, "step": 65450 }, { "grad_norm": 0.28755471110343933, "learning_rate": 8.763666582520923e-06, "loss": 0.0142, "step": 65460 }, { "grad_norm": 0.26546886563301086, "learning_rate": 8.751981489284445e-06, "loss": 0.0126, "step": 65470 }, { "grad_norm": 0.23404106497764587, "learning_rate": 8.740303444206188e-06, "loss": 0.0134, "step": 65480 }, { "grad_norm": 0.1297542005777359, "learning_rate": 8.72863244928162e-06, "loss": 0.0114, "step": 65490 }, { "grad_norm": 0.159888356924057, "learning_rate": 8.716968506504991e-06, "loss": 0.0187, "step": 65500 }, { "grad_norm": 0.4322567582130432, "learning_rate": 8.70531161786935e-06, "loss": 0.0121, "step": 65510 }, { "grad_norm": 0.1531832218170166, "learning_rate": 8.693661785366558e-06, "loss": 0.0208, "step": 65520 }, { "grad_norm": 0.14930661022663116, "learning_rate": 8.682019010987208e-06, "loss": 0.0205, "step": 65530 }, { "grad_norm": 0.165874183177948, "learning_rate": 8.670383296720786e-06, "loss": 0.0096, "step": 65540 }, { "grad_norm": 0.2457515299320221, "learning_rate": 8.658754644555478e-06, "loss": 0.0095, "step": 65550 }, { "grad_norm": 0.2775149643421173, "learning_rate": 8.647133056478313e-06, "loss": 0.0177, "step": 65560 }, { "grad_norm": 0.3064843416213989, "learning_rate": 8.635518534475123e-06, "loss": 0.0128, "step": 65570 }, { "grad_norm": 0.14896558225154877, "learning_rate": 8.623911080530467e-06, "loss": 0.0148, "step": 65580 }, { "grad_norm": 0.22301046550273895, "learning_rate": 8.61231069662779e-06, "loss": 0.0118, "step": 65590 }, { "grad_norm": 0.21596969664096832, "learning_rate": 8.600717384749252e-06, "loss": 0.0108, "step": 65600 }, { "grad_norm": 0.48133039474487305, "learning_rate": 8.58913114687584e-06, "loss": 0.0127, "step": 65610 }, { "grad_norm": 0.3516840636730194, "learning_rate": 8.57755198498732e-06, "loss": 0.0156, "step": 65620 }, { "grad_norm": 0.2888239920139313, "learning_rate": 8.565979901062265e-06, "loss": 0.0212, "step": 65630 }, { "grad_norm": 0.15450236201286316, "learning_rate": 8.554414897078033e-06, "loss": 0.0146, "step": 65640 }, { "grad_norm": 0.11955814808607101, "learning_rate": 8.542856975010727e-06, "loss": 0.0114, "step": 65650 }, { "grad_norm": 0.15176230669021606, "learning_rate": 8.531306136835337e-06, "loss": 0.0098, "step": 65660 }, { "grad_norm": 0.28781601786613464, "learning_rate": 8.519762384525548e-06, "loss": 0.0136, "step": 65670 }, { "grad_norm": 0.19813615083694458, "learning_rate": 8.508225720053875e-06, "loss": 0.0144, "step": 65680 }, { "grad_norm": 0.25856947898864746, "learning_rate": 8.496696145391625e-06, "loss": 0.0111, "step": 65690 }, { "grad_norm": 0.2034502923488617, "learning_rate": 8.485173662508889e-06, "loss": 0.0108, "step": 65700 }, { "grad_norm": 0.1709357649087906, "learning_rate": 8.473658273374536e-06, "loss": 0.0114, "step": 65710 }, { "grad_norm": 0.20730869472026825, "learning_rate": 8.462149979956253e-06, "loss": 0.0097, "step": 65720 }, { "grad_norm": 0.2428853064775467, "learning_rate": 8.450648784220461e-06, "loss": 0.0154, "step": 65730 }, { "grad_norm": 0.21650640666484833, "learning_rate": 8.439154688132417e-06, "loss": 0.0148, "step": 65740 }, { "grad_norm": 0.23260003328323364, "learning_rate": 8.427667693656143e-06, "loss": 0.0172, "step": 65750 }, { "grad_norm": 0.2212064117193222, "learning_rate": 8.416187802754454e-06, "loss": 0.0134, "step": 65760 }, { "grad_norm": 0.27677351236343384, "learning_rate": 8.404715017388965e-06, "loss": 0.014, "step": 65770 }, { "grad_norm": 0.22064270079135895, "learning_rate": 8.393249339520015e-06, "loss": 0.0145, "step": 65780 }, { "grad_norm": 0.20029018819332123, "learning_rate": 8.381790771106834e-06, "loss": 0.0118, "step": 65790 }, { "grad_norm": 0.2561340928077698, "learning_rate": 8.370339314107339e-06, "loss": 0.0091, "step": 65800 }, { "grad_norm": 0.239988312125206, "learning_rate": 8.358894970478281e-06, "loss": 0.0136, "step": 65810 }, { "grad_norm": 0.201361745595932, "learning_rate": 8.347457742175196e-06, "loss": 0.0122, "step": 65820 }, { "grad_norm": 0.300674170255661, "learning_rate": 8.33602763115236e-06, "loss": 0.0113, "step": 65830 }, { "grad_norm": 0.3246892988681793, "learning_rate": 8.324604639362916e-06, "loss": 0.0179, "step": 65840 }, { "grad_norm": 0.21048234403133392, "learning_rate": 8.31318876875869e-06, "loss": 0.0106, "step": 65850 }, { "grad_norm": 0.21040448546409607, "learning_rate": 8.30178002129039e-06, "loss": 0.0095, "step": 65860 }, { "grad_norm": 0.23414850234985352, "learning_rate": 8.290378398907423e-06, "loss": 0.0112, "step": 65870 }, { "grad_norm": 0.2327592372894287, "learning_rate": 8.278983903558029e-06, "loss": 0.0167, "step": 65880 }, { "grad_norm": 0.2480584979057312, "learning_rate": 8.26759653718921e-06, "loss": 0.0106, "step": 65890 }, { "grad_norm": 0.1795441061258316, "learning_rate": 8.25621630174676e-06, "loss": 0.011, "step": 65900 }, { "grad_norm": 0.11866048723459244, "learning_rate": 8.244843199175261e-06, "loss": 0.0184, "step": 65910 }, { "grad_norm": 0.17509619891643524, "learning_rate": 8.233477231418046e-06, "loss": 0.0087, "step": 65920 }, { "grad_norm": 0.37239980697631836, "learning_rate": 8.22211840041725e-06, "loss": 0.0148, "step": 65930 }, { "grad_norm": 0.3932245373725891, "learning_rate": 8.210766708113792e-06, "loss": 0.0195, "step": 65940 }, { "grad_norm": 0.25406181812286377, "learning_rate": 8.199422156447367e-06, "loss": 0.0121, "step": 65950 }, { "grad_norm": 0.3119126260280609, "learning_rate": 8.188084747356451e-06, "loss": 0.0184, "step": 65960 }, { "grad_norm": 0.17651887238025665, "learning_rate": 8.176754482778299e-06, "loss": 0.0107, "step": 65970 }, { "grad_norm": 0.25466787815093994, "learning_rate": 8.165431364648918e-06, "loss": 0.0167, "step": 65980 }, { "grad_norm": 0.22835874557495117, "learning_rate": 8.154115394903162e-06, "loss": 0.0125, "step": 65990 }, { "grad_norm": 0.2227765917778015, "learning_rate": 8.142806575474582e-06, "loss": 0.0116, "step": 66000 }, { "grad_norm": 0.2582200765609741, "learning_rate": 8.131504908295562e-06, "loss": 0.0139, "step": 66010 }, { "grad_norm": 0.24703256785869598, "learning_rate": 8.120210395297262e-06, "loss": 0.0099, "step": 66020 }, { "grad_norm": 0.2363395243883133, "learning_rate": 8.108923038409565e-06, "loss": 0.0149, "step": 66030 }, { "grad_norm": 0.2717467248439789, "learning_rate": 8.097642839561226e-06, "loss": 0.0139, "step": 66040 }, { "grad_norm": 0.36799895763397217, "learning_rate": 8.08636980067966e-06, "loss": 0.0152, "step": 66050 }, { "grad_norm": 0.17946188151836395, "learning_rate": 8.075103923691186e-06, "loss": 0.0109, "step": 66060 }, { "grad_norm": 0.1300010234117508, "learning_rate": 8.063845210520793e-06, "loss": 0.0103, "step": 66070 }, { "grad_norm": 0.1810551881790161, "learning_rate": 8.052593663092295e-06, "loss": 0.0151, "step": 66080 }, { "grad_norm": 0.2398698627948761, "learning_rate": 8.04134928332827e-06, "loss": 0.0124, "step": 66090 }, { "grad_norm": 0.2655318081378937, "learning_rate": 8.030112073150086e-06, "loss": 0.0137, "step": 66100 }, { "grad_norm": 0.27665331959724426, "learning_rate": 8.018882034477881e-06, "loss": 0.0109, "step": 66110 }, { "grad_norm": 0.20453166961669922, "learning_rate": 8.007659169230541e-06, "loss": 0.0106, "step": 66120 }, { "grad_norm": 0.2499767690896988, "learning_rate": 7.996443479325755e-06, "loss": 0.0123, "step": 66130 }, { "grad_norm": 0.32639241218566895, "learning_rate": 7.985234966679977e-06, "loss": 0.0114, "step": 66140 }, { "grad_norm": 0.1634461134672165, "learning_rate": 7.974033633208438e-06, "loss": 0.0134, "step": 66150 }, { "grad_norm": 0.17463192343711853, "learning_rate": 7.962839480825135e-06, "loss": 0.0109, "step": 66160 }, { "grad_norm": 0.5078514218330383, "learning_rate": 7.951652511442858e-06, "loss": 0.015, "step": 66170 }, { "grad_norm": 0.3636202812194824, "learning_rate": 7.940472726973125e-06, "loss": 0.0129, "step": 66180 }, { "grad_norm": 0.22126638889312744, "learning_rate": 7.929300129326289e-06, "loss": 0.0119, "step": 66190 }, { "grad_norm": 0.2632155418395996, "learning_rate": 7.91813472041142e-06, "loss": 0.0123, "step": 66200 }, { "grad_norm": 0.13397255539894104, "learning_rate": 7.906976502136376e-06, "loss": 0.0121, "step": 66210 }, { "grad_norm": 0.16508612036705017, "learning_rate": 7.89582547640782e-06, "loss": 0.0145, "step": 66220 }, { "grad_norm": 0.2783474922180176, "learning_rate": 7.884681645131115e-06, "loss": 0.0146, "step": 66230 }, { "grad_norm": 0.2857446074485779, "learning_rate": 7.87354501021048e-06, "loss": 0.0116, "step": 66240 }, { "grad_norm": 0.210943803191185, "learning_rate": 7.86241557354882e-06, "loss": 0.0111, "step": 66250 }, { "grad_norm": 0.18118900060653687, "learning_rate": 7.8512933370479e-06, "loss": 0.012, "step": 66260 }, { "grad_norm": 0.20361486077308655, "learning_rate": 7.840178302608158e-06, "loss": 0.0112, "step": 66270 }, { "grad_norm": 0.13302506506443024, "learning_rate": 7.82907047212888e-06, "loss": 0.0125, "step": 66280 }, { "grad_norm": 0.17734865844249725, "learning_rate": 7.81796984750809e-06, "loss": 0.0127, "step": 66290 }, { "grad_norm": 0.1827525943517685, "learning_rate": 7.806876430642546e-06, "loss": 0.0123, "step": 66300 }, { "grad_norm": 0.4579582214355469, "learning_rate": 7.795790223427862e-06, "loss": 0.0144, "step": 66310 }, { "grad_norm": 0.3232204020023346, "learning_rate": 7.784711227758324e-06, "loss": 0.0142, "step": 66320 }, { "grad_norm": 0.22565104067325592, "learning_rate": 7.773639445527053e-06, "loss": 0.014, "step": 66330 }, { "grad_norm": 0.21374869346618652, "learning_rate": 7.762574878625905e-06, "loss": 0.0142, "step": 66340 }, { "grad_norm": 0.18602445721626282, "learning_rate": 7.751517528945513e-06, "loss": 0.0108, "step": 66350 }, { "grad_norm": 0.3542037606239319, "learning_rate": 7.740467398375278e-06, "loss": 0.0155, "step": 66360 }, { "grad_norm": 0.2913631796836853, "learning_rate": 7.729424488803378e-06, "loss": 0.0145, "step": 66370 }, { "grad_norm": 0.20183612406253815, "learning_rate": 7.71838880211671e-06, "loss": 0.0232, "step": 66380 }, { "grad_norm": 0.2066805064678192, "learning_rate": 7.707360340200997e-06, "loss": 0.0079, "step": 66390 }, { "grad_norm": 0.1454821527004242, "learning_rate": 7.696339104940697e-06, "loss": 0.0159, "step": 66400 }, { "grad_norm": 0.23117081820964813, "learning_rate": 7.685325098219038e-06, "loss": 0.0116, "step": 66410 }, { "grad_norm": 0.16719022393226624, "learning_rate": 7.674318321918017e-06, "loss": 0.0118, "step": 66420 }, { "grad_norm": 0.14229059219360352, "learning_rate": 7.663318777918366e-06, "loss": 0.0152, "step": 66430 }, { "grad_norm": 0.22130171954631805, "learning_rate": 7.652326468099647e-06, "loss": 0.0231, "step": 66440 }, { "grad_norm": 0.1765558272600174, "learning_rate": 7.641341394340096e-06, "loss": 0.0104, "step": 66450 }, { "grad_norm": 0.46568572521209717, "learning_rate": 7.630363558516818e-06, "loss": 0.0151, "step": 66460 }, { "grad_norm": 0.24834485352039337, "learning_rate": 7.619392962505578e-06, "loss": 0.015, "step": 66470 }, { "grad_norm": 0.22955074906349182, "learning_rate": 7.6084296081809725e-06, "loss": 0.0126, "step": 66480 }, { "grad_norm": 0.3561505377292633, "learning_rate": 7.597473497416347e-06, "loss": 0.0134, "step": 66490 }, { "grad_norm": 0.13386660814285278, "learning_rate": 7.586524632083764e-06, "loss": 0.0095, "step": 66500 }, { "grad_norm": 0.191047802567482, "learning_rate": 7.5755830140541326e-06, "loss": 0.013, "step": 66510 }, { "grad_norm": 0.175993874669075, "learning_rate": 7.5646486451970475e-06, "loss": 0.0124, "step": 66520 }, { "grad_norm": 0.11409193277359009, "learning_rate": 7.553721527380897e-06, "loss": 0.0134, "step": 66530 }, { "grad_norm": 0.22549106180667877, "learning_rate": 7.542801662472826e-06, "loss": 0.0142, "step": 66540 }, { "grad_norm": 0.5356887578964233, "learning_rate": 7.5318890523387475e-06, "loss": 0.0114, "step": 66550 }, { "grad_norm": 0.11537913978099823, "learning_rate": 7.52098369884332e-06, "loss": 0.0083, "step": 66560 }, { "grad_norm": 0.20927982032299042, "learning_rate": 7.510085603849992e-06, "loss": 0.0165, "step": 66570 }, { "grad_norm": 0.1702008694410324, "learning_rate": 7.499194769220918e-06, "loss": 0.0122, "step": 66580 }, { "grad_norm": 0.43630942702293396, "learning_rate": 7.48831119681706e-06, "loss": 0.021, "step": 66590 }, { "grad_norm": 0.49422064423561096, "learning_rate": 7.477434888498119e-06, "loss": 0.0125, "step": 66600 }, { "grad_norm": 0.2528023421764374, "learning_rate": 7.466565846122564e-06, "loss": 0.0183, "step": 66610 }, { "grad_norm": 0.21372458338737488, "learning_rate": 7.455704071547626e-06, "loss": 0.0147, "step": 66620 }, { "grad_norm": 0.20953109860420227, "learning_rate": 7.444849566629247e-06, "loss": 0.0162, "step": 66630 }, { "grad_norm": 0.21230344474315643, "learning_rate": 7.434002333222212e-06, "loss": 0.0094, "step": 66640 }, { "grad_norm": 0.42654913663864136, "learning_rate": 7.423162373179976e-06, "loss": 0.0206, "step": 66650 }, { "grad_norm": 0.2607711851596832, "learning_rate": 7.412329688354835e-06, "loss": 0.012, "step": 66660 }, { "grad_norm": 0.3754195272922516, "learning_rate": 7.40150428059776e-06, "loss": 0.0096, "step": 66670 }, { "grad_norm": 0.2954622507095337, "learning_rate": 7.3906861517585354e-06, "loss": 0.0109, "step": 66680 }, { "grad_norm": 0.20714512467384338, "learning_rate": 7.37987530368569e-06, "loss": 0.0133, "step": 66690 }, { "grad_norm": 0.1501983255147934, "learning_rate": 7.369071738226474e-06, "loss": 0.0107, "step": 66700 }, { "grad_norm": 0.16125977039337158, "learning_rate": 7.358275457226954e-06, "loss": 0.0108, "step": 66710 }, { "grad_norm": 0.24699565768241882, "learning_rate": 7.347486462531899e-06, "loss": 0.015, "step": 66720 }, { "grad_norm": 0.24851390719413757, "learning_rate": 7.336704755984858e-06, "loss": 0.0112, "step": 66730 }, { "grad_norm": 0.1405171900987625, "learning_rate": 7.325930339428133e-06, "loss": 0.014, "step": 66740 }, { "grad_norm": 0.372863233089447, "learning_rate": 7.315163214702769e-06, "loss": 0.0152, "step": 66750 }, { "grad_norm": 0.23660510778427124, "learning_rate": 7.304403383648595e-06, "loss": 0.0148, "step": 66760 }, { "grad_norm": 0.13305744528770447, "learning_rate": 7.293650848104139e-06, "loss": 0.0133, "step": 66770 }, { "grad_norm": 0.12790179252624512, "learning_rate": 7.2829056099067374e-06, "loss": 0.0148, "step": 66780 }, { "grad_norm": 0.38976842164993286, "learning_rate": 7.2721676708924494e-06, "loss": 0.0119, "step": 66790 }, { "grad_norm": 0.5979146957397461, "learning_rate": 7.261437032896096e-06, "loss": 0.0221, "step": 66800 }, { "grad_norm": 0.18372558057308197, "learning_rate": 7.250713697751255e-06, "loss": 0.0145, "step": 66810 }, { "grad_norm": 0.1434183269739151, "learning_rate": 7.239997667290255e-06, "loss": 0.0117, "step": 66820 }, { "grad_norm": 0.08961968123912811, "learning_rate": 7.2292889433441425e-06, "loss": 0.0118, "step": 66830 }, { "grad_norm": 0.25156277418136597, "learning_rate": 7.218587527742793e-06, "loss": 0.0117, "step": 66840 }, { "grad_norm": 0.28800252079963684, "learning_rate": 7.207893422314749e-06, "loss": 0.012, "step": 66850 }, { "grad_norm": 0.2079891562461853, "learning_rate": 7.1972066288873545e-06, "loss": 0.0104, "step": 66860 }, { "grad_norm": 0.6319926977157593, "learning_rate": 7.186527149286687e-06, "loss": 0.0227, "step": 66870 }, { "grad_norm": 0.32773876190185547, "learning_rate": 7.175854985337576e-06, "loss": 0.0109, "step": 66880 }, { "grad_norm": 0.1315547376871109, "learning_rate": 7.16519013886362e-06, "loss": 0.0125, "step": 66890 }, { "grad_norm": 0.1959647387266159, "learning_rate": 7.154532611687109e-06, "loss": 0.013, "step": 66900 }, { "grad_norm": 0.27229905128479004, "learning_rate": 7.143882405629177e-06, "loss": 0.014, "step": 66910 }, { "grad_norm": 0.13740716874599457, "learning_rate": 7.133239522509605e-06, "loss": 0.0103, "step": 66920 }, { "grad_norm": 0.15839774906635284, "learning_rate": 7.1226039641469956e-06, "loss": 0.0117, "step": 66930 }, { "grad_norm": 0.22491192817687988, "learning_rate": 7.111975732358678e-06, "loss": 0.0103, "step": 66940 }, { "grad_norm": 0.13172538578510284, "learning_rate": 7.101354828960693e-06, "loss": 0.0134, "step": 66950 }, { "grad_norm": 0.25551047921180725, "learning_rate": 7.090741255767918e-06, "loss": 0.0105, "step": 66960 }, { "grad_norm": 0.25947991013526917, "learning_rate": 7.080135014593875e-06, "loss": 0.0118, "step": 66970 }, { "grad_norm": 0.24312400817871094, "learning_rate": 7.069536107250896e-06, "loss": 0.0096, "step": 66980 }, { "grad_norm": 0.2148337960243225, "learning_rate": 7.058944535550049e-06, "loss": 0.0122, "step": 66990 }, { "grad_norm": 0.19793330132961273, "learning_rate": 7.048360301301138e-06, "loss": 0.012, "step": 67000 }, { "grad_norm": 0.21813353896141052, "learning_rate": 7.03778340631272e-06, "loss": 0.0131, "step": 67010 }, { "grad_norm": 0.27073702216148376, "learning_rate": 7.0272138523921e-06, "loss": 0.0117, "step": 67020 }, { "grad_norm": 0.23910939693450928, "learning_rate": 7.016651641345334e-06, "loss": 0.0163, "step": 67030 }, { "grad_norm": 0.2520272135734558, "learning_rate": 7.0060967749772e-06, "loss": 0.012, "step": 67040 }, { "grad_norm": 0.37878090143203735, "learning_rate": 6.995549255091238e-06, "loss": 0.0191, "step": 67050 }, { "grad_norm": 0.21827192604541779, "learning_rate": 6.98500908348973e-06, "loss": 0.0153, "step": 67060 }, { "grad_norm": 0.2688538432121277, "learning_rate": 6.974476261973711e-06, "loss": 0.0121, "step": 67070 }, { "grad_norm": 0.3323848843574524, "learning_rate": 6.963950792342949e-06, "loss": 0.0107, "step": 67080 }, { "grad_norm": 0.16983436048030853, "learning_rate": 6.9534326763959715e-06, "loss": 0.0079, "step": 67090 }, { "grad_norm": 0.23021388053894043, "learning_rate": 6.942921915929995e-06, "loss": 0.0108, "step": 67100 }, { "grad_norm": 0.20091286301612854, "learning_rate": 6.9324185127410735e-06, "loss": 0.0181, "step": 67110 }, { "grad_norm": 0.2807179391384125, "learning_rate": 6.921922468623921e-06, "loss": 0.0115, "step": 67120 }, { "grad_norm": 0.1858329027891159, "learning_rate": 6.911433785372023e-06, "loss": 0.0138, "step": 67130 }, { "grad_norm": 0.23712992668151855, "learning_rate": 6.900952464777632e-06, "loss": 0.0179, "step": 67140 }, { "grad_norm": 0.26402366161346436, "learning_rate": 6.8904785086316815e-06, "loss": 0.0111, "step": 67150 }, { "grad_norm": 0.2609748840332031, "learning_rate": 6.880011918723927e-06, "loss": 0.01, "step": 67160 }, { "grad_norm": 0.23352979123592377, "learning_rate": 6.86955269684279e-06, "loss": 0.0107, "step": 67170 }, { "grad_norm": 0.25268566608428955, "learning_rate": 6.859100844775473e-06, "loss": 0.0225, "step": 67180 }, { "grad_norm": 0.1578036993741989, "learning_rate": 6.84865636430792e-06, "loss": 0.0132, "step": 67190 }, { "grad_norm": 0.337615966796875, "learning_rate": 6.838219257224804e-06, "loss": 0.0134, "step": 67200 }, { "grad_norm": 0.24615958333015442, "learning_rate": 6.827789525309536e-06, "loss": 0.0145, "step": 67210 }, { "grad_norm": 0.3392684757709503, "learning_rate": 6.81736717034428e-06, "loss": 0.0127, "step": 67220 }, { "grad_norm": 0.24123995006084442, "learning_rate": 6.806952194109933e-06, "loss": 0.0203, "step": 67230 }, { "grad_norm": 0.11721707135438919, "learning_rate": 6.796544598386112e-06, "loss": 0.0091, "step": 67240 }, { "grad_norm": 0.16786743700504303, "learning_rate": 6.786144384951204e-06, "loss": 0.0158, "step": 67250 }, { "grad_norm": 0.16470184922218323, "learning_rate": 6.775751555582322e-06, "loss": 0.0108, "step": 67260 }, { "grad_norm": 0.39908915758132935, "learning_rate": 6.76536611205531e-06, "loss": 0.0141, "step": 67270 }, { "grad_norm": 0.25896769762039185, "learning_rate": 6.754988056144762e-06, "loss": 0.015, "step": 67280 }, { "grad_norm": 0.11066891252994537, "learning_rate": 6.744617389624014e-06, "loss": 0.0095, "step": 67290 }, { "grad_norm": 0.2515883445739746, "learning_rate": 6.734254114265087e-06, "loss": 0.0121, "step": 67300 }, { "grad_norm": 0.24917957186698914, "learning_rate": 6.723898231838843e-06, "loss": 0.0113, "step": 67310 }, { "grad_norm": 0.3100333511829376, "learning_rate": 6.713549744114766e-06, "loss": 0.0136, "step": 67320 }, { "grad_norm": 0.1593143343925476, "learning_rate": 6.703208652861159e-06, "loss": 0.013, "step": 67330 }, { "grad_norm": 0.17879121005535126, "learning_rate": 6.69287495984503e-06, "loss": 0.0144, "step": 67340 }, { "grad_norm": 0.2083233892917633, "learning_rate": 6.6825486668321e-06, "loss": 0.0118, "step": 67350 }, { "grad_norm": 0.12098205089569092, "learning_rate": 6.672229775586886e-06, "loss": 0.0141, "step": 67360 }, { "grad_norm": 0.23474496603012085, "learning_rate": 6.661918287872576e-06, "loss": 0.0098, "step": 67370 }, { "grad_norm": 0.15792299807071686, "learning_rate": 6.6516142054511346e-06, "loss": 0.0134, "step": 67380 }, { "grad_norm": 0.24917185306549072, "learning_rate": 6.641317530083241e-06, "loss": 0.0097, "step": 67390 }, { "grad_norm": 0.23256167769432068, "learning_rate": 6.631028263528322e-06, "loss": 0.0121, "step": 67400 }, { "grad_norm": 0.19766327738761902, "learning_rate": 6.620746407544537e-06, "loss": 0.0116, "step": 67410 }, { "grad_norm": 0.2043454796075821, "learning_rate": 6.610471963888742e-06, "loss": 0.0134, "step": 67420 }, { "grad_norm": 0.20085610449314117, "learning_rate": 6.600204934316606e-06, "loss": 0.0092, "step": 67430 }, { "grad_norm": 0.17098650336265564, "learning_rate": 6.589945320582452e-06, "loss": 0.0101, "step": 67440 }, { "grad_norm": 0.18229086697101593, "learning_rate": 6.579693124439374e-06, "loss": 0.016, "step": 67450 }, { "grad_norm": 0.19296763837337494, "learning_rate": 6.56944834763919e-06, "loss": 0.0122, "step": 67460 }, { "grad_norm": 0.2370760142803192, "learning_rate": 6.5592109919324575e-06, "loss": 0.0168, "step": 67470 }, { "grad_norm": 0.2288629561662674, "learning_rate": 6.54898105906846e-06, "loss": 0.0162, "step": 67480 }, { "grad_norm": 0.13984400033950806, "learning_rate": 6.5387585507952155e-06, "loss": 0.0189, "step": 67490 }, { "grad_norm": 0.2882106602191925, "learning_rate": 6.528543468859461e-06, "loss": 0.014, "step": 67500 }, { "grad_norm": 0.24287214875221252, "learning_rate": 6.518335815006682e-06, "loss": 0.0147, "step": 67510 }, { "grad_norm": 0.3676014542579651, "learning_rate": 6.5081355909810845e-06, "loss": 0.0124, "step": 67520 }, { "grad_norm": 0.25380972027778625, "learning_rate": 6.497942798525608e-06, "loss": 0.0159, "step": 67530 }, { "grad_norm": 0.5414338707923889, "learning_rate": 6.487757439381936e-06, "loss": 0.018, "step": 67540 }, { "grad_norm": 0.2724912166595459, "learning_rate": 6.477579515290433e-06, "loss": 0.0112, "step": 67550 }, { "grad_norm": 0.1548256129026413, "learning_rate": 6.467409027990273e-06, "loss": 0.015, "step": 67560 }, { "grad_norm": 0.2072252631187439, "learning_rate": 6.457245979219279e-06, "loss": 0.0152, "step": 67570 }, { "grad_norm": 0.30659571290016174, "learning_rate": 6.447090370714054e-06, "loss": 0.0158, "step": 67580 }, { "grad_norm": 0.1490667462348938, "learning_rate": 6.436942204209917e-06, "loss": 0.0175, "step": 67590 }, { "grad_norm": 0.14336776733398438, "learning_rate": 6.4268014814408804e-06, "loss": 0.0185, "step": 67600 }, { "grad_norm": 0.19294288754463196, "learning_rate": 6.41666820413977e-06, "loss": 0.0119, "step": 67610 }, { "grad_norm": 0.20719286799430847, "learning_rate": 6.406542374038033e-06, "loss": 0.0155, "step": 67620 }, { "grad_norm": 0.21384763717651367, "learning_rate": 6.396423992865935e-06, "loss": 0.022, "step": 67630 }, { "grad_norm": 0.6123155951499939, "learning_rate": 6.386313062352412e-06, "loss": 0.0127, "step": 67640 }, { "grad_norm": 0.29069381952285767, "learning_rate": 6.376209584225152e-06, "loss": 0.0103, "step": 67650 }, { "grad_norm": 0.2717827260494232, "learning_rate": 6.366113560210557e-06, "loss": 0.0098, "step": 67660 }, { "grad_norm": 0.45742708444595337, "learning_rate": 6.356024992033766e-06, "loss": 0.0223, "step": 67670 }, { "grad_norm": 0.21716570854187012, "learning_rate": 6.345943881418648e-06, "loss": 0.0108, "step": 67680 }, { "grad_norm": 0.17972898483276367, "learning_rate": 6.33587023008777e-06, "loss": 0.0117, "step": 67690 }, { "grad_norm": 0.2156384438276291, "learning_rate": 6.32580403976245e-06, "loss": 0.0143, "step": 67700 }, { "grad_norm": 0.2276863008737564, "learning_rate": 6.31574531216273e-06, "loss": 0.0134, "step": 67710 }, { "grad_norm": 0.33617502450942993, "learning_rate": 6.305694049007371e-06, "loss": 0.0179, "step": 67720 }, { "grad_norm": 0.2836352586746216, "learning_rate": 6.2956502520138575e-06, "loss": 0.0131, "step": 67730 }, { "grad_norm": 0.19734860956668854, "learning_rate": 6.285613922898409e-06, "loss": 0.0084, "step": 67740 }, { "grad_norm": 0.3650253415107727, "learning_rate": 6.275585063375927e-06, "loss": 0.0133, "step": 67750 }, { "grad_norm": 0.571176290512085, "learning_rate": 6.265563675160113e-06, "loss": 0.0134, "step": 67760 }, { "grad_norm": 0.13811594247817993, "learning_rate": 6.25554975996332e-06, "loss": 0.01, "step": 67770 }, { "grad_norm": 0.19638240337371826, "learning_rate": 6.245543319496661e-06, "loss": 0.0157, "step": 67780 }, { "grad_norm": 0.18474312126636505, "learning_rate": 6.2355443554699685e-06, "loss": 0.0134, "step": 67790 }, { "grad_norm": 0.14263390004634857, "learning_rate": 6.225552869591766e-06, "loss": 0.0113, "step": 67800 }, { "grad_norm": 0.16191141307353973, "learning_rate": 6.215568863569365e-06, "loss": 0.0151, "step": 67810 }, { "grad_norm": 0.4582670331001282, "learning_rate": 6.205592339108712e-06, "loss": 0.0191, "step": 67820 }, { "grad_norm": 0.2258705198764801, "learning_rate": 6.195623297914577e-06, "loss": 0.0134, "step": 67830 }, { "grad_norm": 0.29800161719322205, "learning_rate": 6.185661741690357e-06, "loss": 0.0133, "step": 67840 }, { "grad_norm": 0.31714215874671936, "learning_rate": 6.1757076721382145e-06, "loss": 0.0156, "step": 67850 }, { "grad_norm": 0.28064998984336853, "learning_rate": 6.165761090959038e-06, "loss": 0.0187, "step": 67860 }, { "grad_norm": 0.22667758166790009, "learning_rate": 6.155821999852424e-06, "loss": 0.0143, "step": 67870 }, { "grad_norm": 0.1835252195596695, "learning_rate": 6.145890400516696e-06, "loss": 0.0114, "step": 67880 }, { "grad_norm": 0.1932975947856903, "learning_rate": 6.1359662946488816e-06, "loss": 0.0112, "step": 67890 }, { "grad_norm": 0.24953995645046234, "learning_rate": 6.12604968394474e-06, "loss": 0.0142, "step": 67900 }, { "grad_norm": 0.1788681596517563, "learning_rate": 6.11614057009875e-06, "loss": 0.0154, "step": 67910 }, { "grad_norm": 0.19021126627922058, "learning_rate": 6.106238954804111e-06, "loss": 0.0125, "step": 67920 }, { "grad_norm": 0.2622101306915283, "learning_rate": 6.096344839752738e-06, "loss": 0.0121, "step": 67930 }, { "grad_norm": 0.09613176435232162, "learning_rate": 6.086458226635278e-06, "loss": 0.0128, "step": 67940 }, { "grad_norm": 0.23801597952842712, "learning_rate": 6.076579117141046e-06, "loss": 0.0198, "step": 67950 }, { "grad_norm": 0.3654000759124756, "learning_rate": 6.066707512958153e-06, "loss": 0.0161, "step": 67960 }, { "grad_norm": 0.37862735986709595, "learning_rate": 6.056843415773361e-06, "loss": 0.0147, "step": 67970 }, { "grad_norm": 0.18032753467559814, "learning_rate": 6.0469868272721776e-06, "loss": 0.0115, "step": 67980 }, { "grad_norm": 0.2003789246082306, "learning_rate": 6.037137749138844e-06, "loss": 0.0166, "step": 67990 }, { "grad_norm": 0.27799010276794434, "learning_rate": 6.027296183056252e-06, "loss": 0.0193, "step": 68000 }, { "grad_norm": 0.1645144373178482, "learning_rate": 6.017462130706114e-06, "loss": 0.0134, "step": 68010 }, { "grad_norm": 0.1556084156036377, "learning_rate": 6.0076355937687516e-06, "loss": 0.0147, "step": 68020 }, { "grad_norm": 0.18330608308315277, "learning_rate": 5.9978165739232925e-06, "loss": 0.0109, "step": 68030 }, { "grad_norm": 0.3471183776855469, "learning_rate": 5.988005072847508e-06, "loss": 0.0105, "step": 68040 }, { "grad_norm": 0.23773030936717987, "learning_rate": 5.978201092217928e-06, "loss": 0.0185, "step": 68050 }, { "grad_norm": 0.19241052865982056, "learning_rate": 5.9684046337097895e-06, "loss": 0.0083, "step": 68060 }, { "grad_norm": 0.1498246043920517, "learning_rate": 5.958615698997016e-06, "loss": 0.0123, "step": 68070 }, { "grad_norm": 0.28142619132995605, "learning_rate": 5.948834289752303e-06, "loss": 0.0114, "step": 68080 }, { "grad_norm": 0.2737812399864197, "learning_rate": 5.939060407646996e-06, "loss": 0.0174, "step": 68090 }, { "grad_norm": 0.23056168854236603, "learning_rate": 5.929294054351198e-06, "loss": 0.0161, "step": 68100 }, { "grad_norm": 0.33715584874153137, "learning_rate": 5.919535231533707e-06, "loss": 0.0135, "step": 68110 }, { "grad_norm": 0.21176496148109436, "learning_rate": 5.909783940862046e-06, "loss": 0.0119, "step": 68120 }, { "grad_norm": 0.2694878578186035, "learning_rate": 5.900040184002436e-06, "loss": 0.0139, "step": 68130 }, { "grad_norm": 0.23985229432582855, "learning_rate": 5.890303962619831e-06, "loss": 0.0102, "step": 68140 }, { "grad_norm": 0.32759591937065125, "learning_rate": 5.88057527837787e-06, "loss": 0.0173, "step": 68150 }, { "grad_norm": 0.1660887598991394, "learning_rate": 5.870854132938924e-06, "loss": 0.0141, "step": 68160 }, { "grad_norm": 0.3148997128009796, "learning_rate": 5.861140527964071e-06, "loss": 0.0156, "step": 68170 }, { "grad_norm": 0.14551150798797607, "learning_rate": 5.851434465113098e-06, "loss": 0.0095, "step": 68180 }, { "grad_norm": 0.16975563764572144, "learning_rate": 5.841735946044524e-06, "loss": 0.0121, "step": 68190 }, { "grad_norm": 0.2732149660587311, "learning_rate": 5.832044972415523e-06, "loss": 0.0158, "step": 68200 }, { "grad_norm": 0.36580827832221985, "learning_rate": 5.8223615458820655e-06, "loss": 0.012, "step": 68210 }, { "grad_norm": 0.12124752998352051, "learning_rate": 5.812685668098733e-06, "loss": 0.0123, "step": 68220 }, { "grad_norm": 0.17175498604774475, "learning_rate": 5.803017340718913e-06, "loss": 0.0139, "step": 68230 }, { "grad_norm": 0.16282011568546295, "learning_rate": 5.793356565394636e-06, "loss": 0.0131, "step": 68240 }, { "grad_norm": 0.25372281670570374, "learning_rate": 5.783703343776658e-06, "loss": 0.012, "step": 68250 }, { "grad_norm": 0.20598316192626953, "learning_rate": 5.774057677514477e-06, "loss": 0.0104, "step": 68260 }, { "grad_norm": 0.21960608661174774, "learning_rate": 5.764419568256235e-06, "loss": 0.0116, "step": 68270 }, { "grad_norm": 0.3253277540206909, "learning_rate": 5.754789017648865e-06, "loss": 0.0213, "step": 68280 }, { "grad_norm": 0.09327663481235504, "learning_rate": 5.745166027337934e-06, "loss": 0.0144, "step": 68290 }, { "grad_norm": 0.15110279619693756, "learning_rate": 5.735550598967754e-06, "loss": 0.0135, "step": 68300 }, { "grad_norm": 0.23321545124053955, "learning_rate": 5.725942734181339e-06, "loss": 0.0141, "step": 68310 }, { "grad_norm": 0.26324257254600525, "learning_rate": 5.7163424346204146e-06, "loss": 0.0132, "step": 68320 }, { "grad_norm": 0.22668184340000153, "learning_rate": 5.706749701925407e-06, "loss": 0.0109, "step": 68330 }, { "grad_norm": 0.17779213190078735, "learning_rate": 5.697164537735461e-06, "loss": 0.0109, "step": 68340 }, { "grad_norm": 0.4356062412261963, "learning_rate": 5.687586943688406e-06, "loss": 0.0128, "step": 68350 }, { "grad_norm": 0.14597846567630768, "learning_rate": 5.678016921420792e-06, "loss": 0.0107, "step": 68360 }, { "grad_norm": 0.19969919323921204, "learning_rate": 5.668454472567875e-06, "loss": 0.0177, "step": 68370 }, { "grad_norm": 0.3042212724685669, "learning_rate": 5.658899598763617e-06, "loss": 0.0131, "step": 68380 }, { "grad_norm": 0.23146021366119385, "learning_rate": 5.6493523016407035e-06, "loss": 0.0121, "step": 68390 }, { "grad_norm": 0.12283626198768616, "learning_rate": 5.6398125828304615e-06, "loss": 0.0122, "step": 68400 }, { "grad_norm": 0.15864883363246918, "learning_rate": 5.630280443963015e-06, "loss": 0.0163, "step": 68410 }, { "grad_norm": 0.21284841001033783, "learning_rate": 5.620755886667112e-06, "loss": 0.013, "step": 68420 }, { "grad_norm": 0.23578554391860962, "learning_rate": 5.611238912570271e-06, "loss": 0.0132, "step": 68430 }, { "grad_norm": 0.3397988975048065, "learning_rate": 5.601729523298649e-06, "loss": 0.012, "step": 68440 }, { "grad_norm": 0.188615083694458, "learning_rate": 5.592227720477161e-06, "loss": 0.0118, "step": 68450 }, { "grad_norm": 0.243108868598938, "learning_rate": 5.582733505729415e-06, "loss": 0.0125, "step": 68460 }, { "grad_norm": 0.5559942722320557, "learning_rate": 5.573246880677668e-06, "loss": 0.0171, "step": 68470 }, { "grad_norm": 0.28673747181892395, "learning_rate": 5.56376784694298e-06, "loss": 0.0168, "step": 68480 }, { "grad_norm": 0.27348095178604126, "learning_rate": 5.554296406145027e-06, "loss": 0.017, "step": 68490 }, { "grad_norm": 0.1770523488521576, "learning_rate": 5.544832559902219e-06, "loss": 0.0138, "step": 68500 }, { "grad_norm": 0.3490489721298218, "learning_rate": 5.53537630983168e-06, "loss": 0.015, "step": 68510 }, { "grad_norm": 0.16896739602088928, "learning_rate": 5.5259276575492125e-06, "loss": 0.0147, "step": 68520 }, { "grad_norm": 0.31292977929115295, "learning_rate": 5.516486604669357e-06, "loss": 0.0168, "step": 68530 }, { "grad_norm": 0.18456067144870758, "learning_rate": 5.507053152805303e-06, "loss": 0.0189, "step": 68540 }, { "grad_norm": 0.2383275181055069, "learning_rate": 5.497627303568975e-06, "loss": 0.0164, "step": 68550 }, { "grad_norm": 0.11703372746706009, "learning_rate": 5.488209058571003e-06, "loss": 0.0081, "step": 68560 }, { "grad_norm": 0.23032134771347046, "learning_rate": 5.4787984194207054e-06, "loss": 0.0111, "step": 68570 }, { "grad_norm": 0.755582869052887, "learning_rate": 5.469395387726095e-06, "loss": 0.0119, "step": 68580 }, { "grad_norm": 0.2924034893512726, "learning_rate": 5.45999996509391e-06, "loss": 0.0107, "step": 68590 }, { "grad_norm": 0.2131764441728592, "learning_rate": 5.450612153129536e-06, "loss": 0.0132, "step": 68600 }, { "grad_norm": 0.2029443234205246, "learning_rate": 5.4412319534371426e-06, "loss": 0.0136, "step": 68610 }, { "grad_norm": 0.24740396440029144, "learning_rate": 5.431859367619513e-06, "loss": 0.0103, "step": 68620 }, { "grad_norm": 0.364927738904953, "learning_rate": 5.422494397278172e-06, "loss": 0.0148, "step": 68630 }, { "grad_norm": 0.45789629220962524, "learning_rate": 5.413137044013344e-06, "loss": 0.0132, "step": 68640 }, { "grad_norm": 0.2343641221523285, "learning_rate": 5.403787309423941e-06, "loss": 0.0147, "step": 68650 }, { "grad_norm": 0.2658311724662781, "learning_rate": 5.3944451951075835e-06, "loss": 0.0155, "step": 68660 }, { "grad_norm": 0.2110290229320526, "learning_rate": 5.385110702660562e-06, "loss": 0.0111, "step": 68670 }, { "grad_norm": 0.13558021187782288, "learning_rate": 5.375783833677922e-06, "loss": 0.016, "step": 68680 }, { "grad_norm": 0.29166749119758606, "learning_rate": 5.36646458975334e-06, "loss": 0.0138, "step": 68690 }, { "grad_norm": 0.18521815538406372, "learning_rate": 5.3571529724792294e-06, "loss": 0.0092, "step": 68700 }, { "grad_norm": 0.34627383947372437, "learning_rate": 5.347848983446702e-06, "loss": 0.0112, "step": 68710 }, { "grad_norm": 0.2542897164821625, "learning_rate": 5.3385526242455185e-06, "loss": 0.0163, "step": 68720 }, { "grad_norm": 0.33707666397094727, "learning_rate": 5.329263896464226e-06, "loss": 0.0131, "step": 68730 }, { "grad_norm": 0.31646955013275146, "learning_rate": 5.3199828016899715e-06, "loss": 0.0137, "step": 68740 }, { "grad_norm": 0.25728845596313477, "learning_rate": 5.310709341508657e-06, "loss": 0.0157, "step": 68750 }, { "grad_norm": 0.1914583444595337, "learning_rate": 5.301443517504861e-06, "loss": 0.0095, "step": 68760 }, { "grad_norm": 0.18690387904644012, "learning_rate": 5.292185331261862e-06, "loss": 0.0103, "step": 68770 }, { "grad_norm": 0.22648440301418304, "learning_rate": 5.28293478436162e-06, "loss": 0.0096, "step": 68780 }, { "grad_norm": 0.172097846865654, "learning_rate": 5.273691878384829e-06, "loss": 0.0144, "step": 68790 }, { "grad_norm": 0.23233284056186676, "learning_rate": 5.264456614910812e-06, "loss": 0.0115, "step": 68800 }, { "grad_norm": 0.22115376591682434, "learning_rate": 5.255228995517647e-06, "loss": 0.0189, "step": 68810 }, { "grad_norm": 0.1423092931509018, "learning_rate": 5.246009021782067e-06, "loss": 0.0109, "step": 68820 }, { "grad_norm": 0.28027036786079407, "learning_rate": 5.2367966952795225e-06, "loss": 0.0189, "step": 68830 }, { "grad_norm": 0.19435253739356995, "learning_rate": 5.2275920175841485e-06, "loss": 0.013, "step": 68840 }, { "grad_norm": 0.14257346093654633, "learning_rate": 5.218394990268771e-06, "loss": 0.0139, "step": 68850 }, { "grad_norm": 0.32795625925064087, "learning_rate": 5.209205614904916e-06, "loss": 0.0147, "step": 68860 }, { "grad_norm": 0.30824992060661316, "learning_rate": 5.200023893062777e-06, "loss": 0.0113, "step": 68870 }, { "grad_norm": 0.17441794276237488, "learning_rate": 5.190849826311289e-06, "loss": 0.0111, "step": 68880 }, { "grad_norm": 0.3095221221446991, "learning_rate": 5.181683416218025e-06, "loss": 0.0138, "step": 68890 }, { "grad_norm": 0.16992424428462982, "learning_rate": 5.172524664349276e-06, "loss": 0.0084, "step": 68900 }, { "grad_norm": 0.2878931760787964, "learning_rate": 5.1633735722700416e-06, "loss": 0.01, "step": 68910 }, { "grad_norm": 0.12605075538158417, "learning_rate": 5.154230141543958e-06, "loss": 0.0187, "step": 68920 }, { "grad_norm": 0.7059679627418518, "learning_rate": 5.145094373733433e-06, "loss": 0.0122, "step": 68930 }, { "grad_norm": 0.14819122850894928, "learning_rate": 5.135966270399478e-06, "loss": 0.0109, "step": 68940 }, { "grad_norm": 0.5184137225151062, "learning_rate": 5.126845833101857e-06, "loss": 0.0153, "step": 68950 }, { "grad_norm": 0.18502429127693176, "learning_rate": 5.1177330633989994e-06, "loss": 0.011, "step": 68960 }, { "grad_norm": 0.40879544615745544, "learning_rate": 5.108627962848033e-06, "loss": 0.0265, "step": 68970 }, { "grad_norm": 0.241559699177742, "learning_rate": 5.099530533004759e-06, "loss": 0.0148, "step": 68980 }, { "grad_norm": 0.1456114798784256, "learning_rate": 5.090440775423699e-06, "loss": 0.0101, "step": 68990 }, { "grad_norm": 0.27535855770111084, "learning_rate": 5.081358691658022e-06, "loss": 0.013, "step": 69000 }, { "grad_norm": 0.23168067634105682, "learning_rate": 5.072284283259621e-06, "loss": 0.013, "step": 69010 }, { "grad_norm": 0.146340012550354, "learning_rate": 5.063217551779053e-06, "loss": 0.0111, "step": 69020 }, { "grad_norm": 0.21422281861305237, "learning_rate": 5.054158498765588e-06, "loss": 0.0092, "step": 69030 }, { "grad_norm": 0.27024054527282715, "learning_rate": 5.0451071257671625e-06, "loss": 0.0134, "step": 69040 }, { "grad_norm": 0.30916929244995117, "learning_rate": 5.036063434330407e-06, "loss": 0.0167, "step": 69050 }, { "grad_norm": 1.3305364847183228, "learning_rate": 5.027027426000652e-06, "loss": 0.0129, "step": 69060 }, { "grad_norm": 0.46273958683013916, "learning_rate": 5.017999102321886e-06, "loss": 0.0128, "step": 69070 }, { "grad_norm": 0.3552987575531006, "learning_rate": 5.0089784648368224e-06, "loss": 0.0132, "step": 69080 }, { "grad_norm": 0.5539349913597107, "learning_rate": 4.999965515086829e-06, "loss": 0.0108, "step": 69090 }, { "grad_norm": 0.23317837715148926, "learning_rate": 4.990960254611976e-06, "loss": 0.0098, "step": 69100 }, { "grad_norm": 0.47047168016433716, "learning_rate": 4.9819626849510194e-06, "loss": 0.0116, "step": 69110 }, { "grad_norm": 0.17584845423698425, "learning_rate": 4.97297280764138e-06, "loss": 0.0158, "step": 69120 }, { "grad_norm": 0.15511399507522583, "learning_rate": 4.963990624219211e-06, "loss": 0.0175, "step": 69130 }, { "grad_norm": 0.40963271260261536, "learning_rate": 4.955016136219298e-06, "loss": 0.0142, "step": 69140 }, { "grad_norm": 0.20300233364105225, "learning_rate": 4.946049345175146e-06, "loss": 0.0111, "step": 69150 }, { "grad_norm": 0.36606040596961975, "learning_rate": 4.937090252618937e-06, "loss": 0.0125, "step": 69160 }, { "grad_norm": 0.15228161215782166, "learning_rate": 4.928138860081521e-06, "loss": 0.0159, "step": 69170 }, { "grad_norm": 0.17747795581817627, "learning_rate": 4.919195169092472e-06, "loss": 0.0156, "step": 69180 }, { "grad_norm": 0.33606213331222534, "learning_rate": 4.910259181179994e-06, "loss": 0.013, "step": 69190 }, { "grad_norm": 0.22744493186473846, "learning_rate": 4.901330897871015e-06, "loss": 0.015, "step": 69200 }, { "grad_norm": 0.32366830110549927, "learning_rate": 4.8924103206911375e-06, "loss": 0.0103, "step": 69210 }, { "grad_norm": 0.16298584640026093, "learning_rate": 4.883497451164637e-06, "loss": 0.0134, "step": 69220 }, { "grad_norm": 0.28883859515190125, "learning_rate": 4.874592290814484e-06, "loss": 0.0128, "step": 69230 }, { "grad_norm": 0.5528842806816101, "learning_rate": 4.865694841162327e-06, "loss": 0.021, "step": 69240 }, { "grad_norm": 0.17444568872451782, "learning_rate": 4.856805103728496e-06, "loss": 0.0116, "step": 69250 }, { "grad_norm": 0.21371504664421082, "learning_rate": 4.847923080032007e-06, "loss": 0.011, "step": 69260 }, { "grad_norm": 0.2135511040687561, "learning_rate": 4.8390487715905486e-06, "loss": 0.0127, "step": 69270 }, { "grad_norm": 0.3800562024116516, "learning_rate": 4.8301821799205e-06, "loss": 0.0106, "step": 69280 }, { "grad_norm": 0.11351863294839859, "learning_rate": 4.821323306536918e-06, "loss": 0.0087, "step": 69290 }, { "grad_norm": 0.1443036049604416, "learning_rate": 4.8124721529535455e-06, "loss": 0.0137, "step": 69300 }, { "grad_norm": 0.2561165392398834, "learning_rate": 4.803628720682807e-06, "loss": 0.0184, "step": 69310 }, { "grad_norm": 0.19563600420951843, "learning_rate": 4.794793011235776e-06, "loss": 0.0122, "step": 69320 }, { "grad_norm": 0.30694204568862915, "learning_rate": 4.78596502612228e-06, "loss": 0.0186, "step": 69330 }, { "grad_norm": 0.27614450454711914, "learning_rate": 4.777144766850738e-06, "loss": 0.0151, "step": 69340 }, { "grad_norm": 0.18502762913703918, "learning_rate": 4.768332234928313e-06, "loss": 0.0186, "step": 69350 }, { "grad_norm": 0.10330826789140701, "learning_rate": 4.759527431860828e-06, "loss": 0.013, "step": 69360 }, { "grad_norm": 0.1968691349029541, "learning_rate": 4.750730359152755e-06, "loss": 0.0217, "step": 69370 }, { "grad_norm": 0.26877912878990173, "learning_rate": 4.741941018307311e-06, "loss": 0.0135, "step": 69380 }, { "grad_norm": 0.14386044442653656, "learning_rate": 4.733159410826321e-06, "loss": 0.0162, "step": 69390 }, { "grad_norm": 0.14122669398784637, "learning_rate": 4.724385538210357e-06, "loss": 0.011, "step": 69400 }, { "grad_norm": 0.21227607131004333, "learning_rate": 4.715619401958599e-06, "loss": 0.0125, "step": 69410 }, { "grad_norm": 0.2534257769584656, "learning_rate": 4.706861003568958e-06, "loss": 0.0149, "step": 69420 }, { "grad_norm": 0.16345134377479553, "learning_rate": 4.698110344538003e-06, "loss": 0.0114, "step": 69430 }, { "grad_norm": 0.3629626929759979, "learning_rate": 4.689367426360975e-06, "loss": 0.014, "step": 69440 }, { "grad_norm": 0.31542304158210754, "learning_rate": 4.680632250531819e-06, "loss": 0.0086, "step": 69450 }, { "grad_norm": 0.15217602252960205, "learning_rate": 4.671904818543115e-06, "loss": 0.0111, "step": 69460 }, { "grad_norm": 0.35169872641563416, "learning_rate": 4.66318513188615e-06, "loss": 0.0127, "step": 69470 }, { "grad_norm": 0.5121157169342041, "learning_rate": 4.654473192050884e-06, "loss": 0.0159, "step": 69480 }, { "grad_norm": 0.1759777069091797, "learning_rate": 4.64576900052594e-06, "loss": 0.0131, "step": 69490 }, { "grad_norm": 0.21514812111854553, "learning_rate": 4.637072558798638e-06, "loss": 0.0083, "step": 69500 }, { "grad_norm": 0.30635446310043335, "learning_rate": 4.628383868354969e-06, "loss": 0.0139, "step": 69510 }, { "grad_norm": 0.17952772974967957, "learning_rate": 4.6197029306795595e-06, "loss": 0.0119, "step": 69520 }, { "grad_norm": 0.11002679914236069, "learning_rate": 4.611029747255779e-06, "loss": 0.0106, "step": 69530 }, { "grad_norm": 0.26156941056251526, "learning_rate": 4.6023643195656164e-06, "loss": 0.0128, "step": 69540 }, { "grad_norm": 0.10371633619070053, "learning_rate": 4.593706649089768e-06, "loss": 0.0103, "step": 69550 }, { "grad_norm": 0.35363221168518066, "learning_rate": 4.585056737307597e-06, "loss": 0.0122, "step": 69560 }, { "grad_norm": 0.24487660825252533, "learning_rate": 4.576414585697103e-06, "loss": 0.0093, "step": 69570 }, { "grad_norm": 0.2718004584312439, "learning_rate": 4.567780195735044e-06, "loss": 0.0124, "step": 69580 }, { "grad_norm": 0.17032943665981293, "learning_rate": 4.559153568896757e-06, "loss": 0.0098, "step": 69590 }, { "grad_norm": 0.4898870289325714, "learning_rate": 4.550534706656329e-06, "loss": 0.0126, "step": 69600 }, { "grad_norm": 0.11067963391542435, "learning_rate": 4.541923610486465e-06, "loss": 0.0104, "step": 69610 }, { "grad_norm": 0.2593923807144165, "learning_rate": 4.533320281858578e-06, "loss": 0.0104, "step": 69620 }, { "grad_norm": 0.22684155404567719, "learning_rate": 4.52472472224274e-06, "loss": 0.0188, "step": 69630 }, { "grad_norm": 0.23022152483463287, "learning_rate": 4.51613693310769e-06, "loss": 0.0129, "step": 69640 }, { "grad_norm": 1.0020209550857544, "learning_rate": 4.507556915920868e-06, "loss": 0.0153, "step": 69650 }, { "grad_norm": 0.21099886298179626, "learning_rate": 4.498984672148332e-06, "loss": 0.0105, "step": 69660 }, { "grad_norm": 0.31264880299568176, "learning_rate": 4.490420203254864e-06, "loss": 0.017, "step": 69670 }, { "grad_norm": 0.31021663546562195, "learning_rate": 4.481863510703893e-06, "loss": 0.0137, "step": 69680 }, { "grad_norm": 0.11054731160402298, "learning_rate": 4.473314595957523e-06, "loss": 0.0102, "step": 69690 }, { "grad_norm": 0.19213956594467163, "learning_rate": 4.464773460476535e-06, "loss": 0.0147, "step": 69700 }, { "grad_norm": 0.13925401866436005, "learning_rate": 4.456240105720372e-06, "loss": 0.0112, "step": 69710 }, { "grad_norm": 0.13778173923492432, "learning_rate": 4.4477145331471405e-06, "loss": 0.0224, "step": 69720 }, { "grad_norm": 0.26967549324035645, "learning_rate": 4.439196744213653e-06, "loss": 0.0134, "step": 69730 }, { "grad_norm": 0.8360865712165833, "learning_rate": 4.430686740375339e-06, "loss": 0.0153, "step": 69740 }, { "grad_norm": 0.47941455245018005, "learning_rate": 4.422184523086342e-06, "loss": 0.0123, "step": 69750 }, { "grad_norm": 0.24597792327404022, "learning_rate": 4.41369009379946e-06, "loss": 0.0118, "step": 69760 }, { "grad_norm": 0.3535231351852417, "learning_rate": 4.405203453966139e-06, "loss": 0.0123, "step": 69770 }, { "grad_norm": 0.36670583486557007, "learning_rate": 4.396724605036539e-06, "loss": 0.0126, "step": 69780 }, { "grad_norm": 0.2858984172344208, "learning_rate": 4.388253548459437e-06, "loss": 0.0111, "step": 69790 }, { "grad_norm": 0.16789822280406952, "learning_rate": 4.3797902856823395e-06, "loss": 0.0112, "step": 69800 }, { "grad_norm": 0.10159020125865936, "learning_rate": 4.371334818151357e-06, "loss": 0.012, "step": 69810 }, { "grad_norm": 0.23108622431755066, "learning_rate": 4.362887147311306e-06, "loss": 0.0108, "step": 69820 }, { "grad_norm": 0.13703303039073944, "learning_rate": 4.354447274605672e-06, "loss": 0.0113, "step": 69830 }, { "grad_norm": 0.18759679794311523, "learning_rate": 4.346015201476572e-06, "loss": 0.0157, "step": 69840 }, { "grad_norm": 0.1541365683078766, "learning_rate": 4.337590929364855e-06, "loss": 0.0145, "step": 69850 }, { "grad_norm": 0.2104155272245407, "learning_rate": 4.329174459709973e-06, "loss": 0.0113, "step": 69860 }, { "grad_norm": 0.16148118674755096, "learning_rate": 4.320765793950071e-06, "loss": 0.0149, "step": 69870 }, { "grad_norm": 0.371367484331131, "learning_rate": 4.312364933521962e-06, "loss": 0.0134, "step": 69880 }, { "grad_norm": 0.1572556495666504, "learning_rate": 4.303971879861129e-06, "loss": 0.0146, "step": 69890 }, { "grad_norm": 0.28499075770378113, "learning_rate": 4.295586634401716e-06, "loss": 0.0127, "step": 69900 }, { "grad_norm": 0.21539117395877838, "learning_rate": 4.287209198576536e-06, "loss": 0.0095, "step": 69910 }, { "grad_norm": 0.1736128181219101, "learning_rate": 4.278839573817045e-06, "loss": 0.0138, "step": 69920 }, { "grad_norm": 0.3652603030204773, "learning_rate": 4.270477761553399e-06, "loss": 0.0141, "step": 69930 }, { "grad_norm": 0.1854444295167923, "learning_rate": 4.262123763214393e-06, "loss": 0.011, "step": 69940 }, { "grad_norm": 0.18273979425430298, "learning_rate": 4.253777580227508e-06, "loss": 0.0105, "step": 69950 }, { "grad_norm": 0.18425588309764862, "learning_rate": 4.2454392140188755e-06, "loss": 0.0137, "step": 69960 }, { "grad_norm": 0.1944046914577484, "learning_rate": 4.2371086660132785e-06, "loss": 0.0155, "step": 69970 }, { "grad_norm": 0.22331373393535614, "learning_rate": 4.228785937634205e-06, "loss": 0.0104, "step": 69980 }, { "grad_norm": 0.14613251388072968, "learning_rate": 4.220471030303758e-06, "loss": 0.0083, "step": 69990 }, { "grad_norm": 0.18946535885334015, "learning_rate": 4.212163945442754e-06, "loss": 0.0078, "step": 70000 }, { "grad_norm": 0.3308236002922058, "learning_rate": 4.203864684470621e-06, "loss": 0.0113, "step": 70010 }, { "grad_norm": 0.10055329650640488, "learning_rate": 4.19557324880549e-06, "loss": 0.0112, "step": 70020 }, { "grad_norm": 0.19028882682323456, "learning_rate": 4.187289639864145e-06, "loss": 0.0117, "step": 70030 }, { "grad_norm": 0.1776491105556488, "learning_rate": 4.1790138590619974e-06, "loss": 0.0152, "step": 70040 }, { "grad_norm": 0.20684027671813965, "learning_rate": 4.170745907813195e-06, "loss": 0.0162, "step": 70050 }, { "grad_norm": 0.28195029497146606, "learning_rate": 4.162485787530479e-06, "loss": 0.0128, "step": 70060 }, { "grad_norm": 0.18275201320648193, "learning_rate": 4.154233499625282e-06, "loss": 0.0109, "step": 70070 }, { "grad_norm": 0.12574352324008942, "learning_rate": 4.145989045507692e-06, "loss": 0.0103, "step": 70080 }, { "grad_norm": 0.14128266274929047, "learning_rate": 4.1377524265864666e-06, "loss": 0.0138, "step": 70090 }, { "grad_norm": 0.21257802844047546, "learning_rate": 4.1295236442690175e-06, "loss": 0.014, "step": 70100 }, { "grad_norm": 0.5870394110679626, "learning_rate": 4.121302699961421e-06, "loss": 0.0128, "step": 70110 }, { "grad_norm": 0.27567023038864136, "learning_rate": 4.113089595068403e-06, "loss": 0.0137, "step": 70120 }, { "grad_norm": 0.28124621510505676, "learning_rate": 4.104884330993364e-06, "loss": 0.0131, "step": 70130 }, { "grad_norm": 0.15336717665195465, "learning_rate": 4.0966869091383585e-06, "loss": 0.0152, "step": 70140 }, { "grad_norm": 0.19024808704853058, "learning_rate": 4.088497330904101e-06, "loss": 0.0144, "step": 70150 }, { "grad_norm": 0.3309122622013092, "learning_rate": 4.080315597689976e-06, "loss": 0.0105, "step": 70160 }, { "grad_norm": 0.19881576299667358, "learning_rate": 4.072141710893995e-06, "loss": 0.017, "step": 70170 }, { "grad_norm": 0.326415091753006, "learning_rate": 4.063975671912879e-06, "loss": 0.019, "step": 70180 }, { "grad_norm": 0.22174715995788574, "learning_rate": 4.055817482141949e-06, "loss": 0.0094, "step": 70190 }, { "grad_norm": 0.29723700881004333, "learning_rate": 4.047667142975259e-06, "loss": 0.0117, "step": 70200 }, { "grad_norm": 0.2931164503097534, "learning_rate": 4.039524655805443e-06, "loss": 0.016, "step": 70210 }, { "grad_norm": 0.1520141363143921, "learning_rate": 4.03139002202384e-06, "loss": 0.0078, "step": 70220 }, { "grad_norm": 0.13491420447826385, "learning_rate": 4.023263243020447e-06, "loss": 0.0124, "step": 70230 }, { "grad_norm": 0.11395935714244843, "learning_rate": 4.015144320183884e-06, "loss": 0.0161, "step": 70240 }, { "grad_norm": 0.22719784080982208, "learning_rate": 4.007033254901482e-06, "loss": 0.0103, "step": 70250 }, { "grad_norm": 0.4344649314880371, "learning_rate": 3.9989300485591795e-06, "loss": 0.0124, "step": 70260 }, { "grad_norm": 0.17875629663467407, "learning_rate": 3.990834702541601e-06, "loss": 0.0117, "step": 70270 }, { "grad_norm": 0.22403018176555634, "learning_rate": 3.982747218232019e-06, "loss": 0.0085, "step": 70280 }, { "grad_norm": 0.5068649649620056, "learning_rate": 3.97466759701236e-06, "loss": 0.0134, "step": 70290 }, { "grad_norm": 0.27094873785972595, "learning_rate": 3.966595840263226e-06, "loss": 0.0108, "step": 70300 }, { "grad_norm": 0.30639633536338806, "learning_rate": 3.958531949363836e-06, "loss": 0.0121, "step": 70310 }, { "grad_norm": 0.21867509186267853, "learning_rate": 3.950475925692098e-06, "loss": 0.0136, "step": 70320 }, { "grad_norm": 0.26279979944229126, "learning_rate": 3.9424277706245685e-06, "loss": 0.0121, "step": 70330 }, { "grad_norm": 0.20942455530166626, "learning_rate": 3.934387485536451e-06, "loss": 0.0131, "step": 70340 }, { "grad_norm": 0.22450122237205505, "learning_rate": 3.926355071801619e-06, "loss": 0.0126, "step": 70350 }, { "grad_norm": 0.28643348813056946, "learning_rate": 3.9183305307925965e-06, "loss": 0.0108, "step": 70360 }, { "grad_norm": 0.17806881666183472, "learning_rate": 3.910313863880533e-06, "loss": 0.0122, "step": 70370 }, { "grad_norm": 0.11824017763137817, "learning_rate": 3.902305072435292e-06, "loss": 0.0095, "step": 70380 }, { "grad_norm": 0.20813898742198944, "learning_rate": 3.894304157825329e-06, "loss": 0.0116, "step": 70390 }, { "grad_norm": 0.18970130383968353, "learning_rate": 3.886311121417791e-06, "loss": 0.01, "step": 70400 }, { "grad_norm": 0.11608945578336716, "learning_rate": 3.878325964578472e-06, "loss": 0.0111, "step": 70410 }, { "grad_norm": 0.10875550657510757, "learning_rate": 3.870348688671815e-06, "loss": 0.0088, "step": 70420 }, { "grad_norm": 0.2602100968360901, "learning_rate": 3.862379295060931e-06, "loss": 0.0102, "step": 70430 }, { "grad_norm": 0.22023169696331024, "learning_rate": 3.854417785107534e-06, "loss": 0.0161, "step": 70440 }, { "grad_norm": 0.4866692125797272, "learning_rate": 3.8464641601720755e-06, "loss": 0.0133, "step": 70450 }, { "grad_norm": 0.3387834131717682, "learning_rate": 3.838518421613579e-06, "loss": 0.0149, "step": 70460 }, { "grad_norm": 0.29302915930747986, "learning_rate": 3.830580570789766e-06, "loss": 0.0136, "step": 70470 }, { "grad_norm": 0.17496170103549957, "learning_rate": 3.822650609057005e-06, "loss": 0.0102, "step": 70480 }, { "grad_norm": 0.15383607149124146, "learning_rate": 3.814728537770285e-06, "loss": 0.0119, "step": 70490 }, { "grad_norm": 0.22314345836639404, "learning_rate": 3.806814358283306e-06, "loss": 0.0141, "step": 70500 }, { "grad_norm": 0.1992959827184677, "learning_rate": 3.7989080719483596e-06, "loss": 0.0082, "step": 70510 }, { "grad_norm": 0.18796800076961517, "learning_rate": 3.7910096801164143e-06, "loss": 0.0146, "step": 70520 }, { "grad_norm": 0.30478471517562866, "learning_rate": 3.7831191841371016e-06, "loss": 0.011, "step": 70530 }, { "grad_norm": 0.9606838226318359, "learning_rate": 3.775236585358688e-06, "loss": 0.0159, "step": 70540 }, { "grad_norm": 0.1568395346403122, "learning_rate": 3.7673618851280843e-06, "loss": 0.0131, "step": 70550 }, { "grad_norm": 0.25539737939834595, "learning_rate": 3.759495084790887e-06, "loss": 0.012, "step": 70560 }, { "grad_norm": 0.25110167264938354, "learning_rate": 3.751636185691282e-06, "loss": 0.0111, "step": 70570 }, { "grad_norm": 0.13591524958610535, "learning_rate": 3.7437851891721607e-06, "loss": 0.0108, "step": 70580 }, { "grad_norm": 0.5190805792808533, "learning_rate": 3.7359420965750404e-06, "loss": 0.0165, "step": 70590 }, { "grad_norm": 0.1600446105003357, "learning_rate": 3.7281069092400922e-06, "loss": 0.0133, "step": 70600 }, { "grad_norm": 0.2202637493610382, "learning_rate": 3.7202796285061348e-06, "loss": 0.0099, "step": 70610 }, { "grad_norm": 0.19586396217346191, "learning_rate": 3.712460255710637e-06, "loss": 0.0145, "step": 70620 }, { "grad_norm": 0.25375956296920776, "learning_rate": 3.704648792189719e-06, "loss": 0.0118, "step": 70630 }, { "grad_norm": 0.30326181650161743, "learning_rate": 3.696845239278124e-06, "loss": 0.0122, "step": 70640 }, { "grad_norm": 0.2430727332830429, "learning_rate": 3.689049598309302e-06, "loss": 0.0128, "step": 70650 }, { "grad_norm": 0.4071879982948303, "learning_rate": 3.681261870615288e-06, "loss": 0.0208, "step": 70660 }, { "grad_norm": 0.24967221915721893, "learning_rate": 3.6734820575268004e-06, "loss": 0.0216, "step": 70670 }, { "grad_norm": 0.09606766700744629, "learning_rate": 3.665710160373209e-06, "loss": 0.0155, "step": 70680 }, { "grad_norm": 0.1521749645471573, "learning_rate": 3.65794618048248e-06, "loss": 0.0096, "step": 70690 }, { "grad_norm": 0.1703500896692276, "learning_rate": 3.6501901191813125e-06, "loss": 0.0097, "step": 70700 }, { "grad_norm": 0.1055600717663765, "learning_rate": 3.642441977794975e-06, "loss": 0.0098, "step": 70710 }, { "grad_norm": 0.27832889556884766, "learning_rate": 3.634701757647424e-06, "loss": 0.0114, "step": 70720 }, { "grad_norm": 0.126303568482399, "learning_rate": 3.6269694600612468e-06, "loss": 0.0076, "step": 70730 }, { "grad_norm": 0.28819024562835693, "learning_rate": 3.619245086357681e-06, "loss": 0.019, "step": 70740 }, { "grad_norm": 0.42400678992271423, "learning_rate": 3.611528637856615e-06, "loss": 0.0119, "step": 70750 }, { "grad_norm": 0.301371306180954, "learning_rate": 3.6038201158765884e-06, "loss": 0.0116, "step": 70760 }, { "grad_norm": 0.6987459659576416, "learning_rate": 3.5961195217347534e-06, "loss": 0.0205, "step": 70770 }, { "grad_norm": 0.28664806485176086, "learning_rate": 3.588426856746946e-06, "loss": 0.0139, "step": 70780 }, { "grad_norm": 0.10049910843372345, "learning_rate": 3.5807421222276316e-06, "loss": 0.0197, "step": 70790 }, { "grad_norm": 0.23958894610404968, "learning_rate": 3.5730653194899156e-06, "loss": 0.0119, "step": 70800 }, { "grad_norm": 0.21149171888828278, "learning_rate": 3.565396449845554e-06, "loss": 0.0106, "step": 70810 }, { "grad_norm": 0.4492609202861786, "learning_rate": 3.557735514604954e-06, "loss": 0.0138, "step": 70820 }, { "grad_norm": 0.20492984354496002, "learning_rate": 3.5500825150771633e-06, "loss": 0.0113, "step": 70830 }, { "grad_norm": 0.15615177154541016, "learning_rate": 3.5424374525698466e-06, "loss": 0.0105, "step": 70840 }, { "grad_norm": 0.2316025048494339, "learning_rate": 3.5348003283893704e-06, "loss": 0.0113, "step": 70850 }, { "grad_norm": 0.082049660384655, "learning_rate": 3.527171143840685e-06, "loss": 0.0128, "step": 70860 }, { "grad_norm": 0.16628748178482056, "learning_rate": 3.519549900227409e-06, "loss": 0.0147, "step": 70870 }, { "grad_norm": 0.4030948877334595, "learning_rate": 3.511936598851828e-06, "loss": 0.0151, "step": 70880 }, { "grad_norm": 0.1527680903673172, "learning_rate": 3.504331241014813e-06, "loss": 0.0147, "step": 70890 }, { "grad_norm": 0.4956211447715759, "learning_rate": 3.4967338280159414e-06, "loss": 0.0104, "step": 70900 }, { "grad_norm": 0.13309964537620544, "learning_rate": 3.4891443611533846e-06, "loss": 0.0153, "step": 70910 }, { "grad_norm": 0.2136106789112091, "learning_rate": 3.481562841723984e-06, "loss": 0.016, "step": 70920 }, { "grad_norm": 0.24893540143966675, "learning_rate": 3.473989271023215e-06, "loss": 0.0103, "step": 70930 }, { "grad_norm": 0.24297967553138733, "learning_rate": 3.4664236503451854e-06, "loss": 0.0139, "step": 70940 }, { "grad_norm": 0.2137177288532257, "learning_rate": 3.4588659809826674e-06, "loss": 0.0096, "step": 70950 }, { "grad_norm": 0.328082799911499, "learning_rate": 3.451316264227039e-06, "loss": 0.0169, "step": 70960 }, { "grad_norm": 0.726862370967865, "learning_rate": 3.4437745013683574e-06, "loss": 0.0111, "step": 70970 }, { "grad_norm": 0.19685177505016327, "learning_rate": 3.4362406936952916e-06, "loss": 0.0124, "step": 70980 }, { "grad_norm": 0.1804562211036682, "learning_rate": 3.4287148424951786e-06, "loss": 0.0111, "step": 70990 }, { "grad_norm": 0.18917717039585114, "learning_rate": 3.4211969490539617e-06, "loss": 0.0136, "step": 71000 }, { "grad_norm": 0.3128080666065216, "learning_rate": 3.4136870146562584e-06, "loss": 0.0141, "step": 71010 }, { "grad_norm": 0.24106350541114807, "learning_rate": 3.406185040585308e-06, "loss": 0.0104, "step": 71020 }, { "grad_norm": 0.24665667116641998, "learning_rate": 3.3986910281229966e-06, "loss": 0.0143, "step": 71030 }, { "grad_norm": 0.23376847803592682, "learning_rate": 3.391204978549828e-06, "loss": 0.0128, "step": 71040 }, { "grad_norm": 0.41686171293258667, "learning_rate": 3.3837268931449785e-06, "loss": 0.0172, "step": 71050 }, { "grad_norm": 0.23656395077705383, "learning_rate": 3.376256773186248e-06, "loss": 0.0127, "step": 71060 }, { "grad_norm": 0.2593514919281006, "learning_rate": 3.3687946199500664e-06, "loss": 0.0165, "step": 71070 }, { "grad_norm": 0.180088609457016, "learning_rate": 3.3613404347115295e-06, "loss": 0.0103, "step": 71080 }, { "grad_norm": 0.16185788810253143, "learning_rate": 3.35389421874433e-06, "loss": 0.0104, "step": 71090 }, { "grad_norm": 0.1195165291428566, "learning_rate": 3.3464559733208446e-06, "loss": 0.0123, "step": 71100 }, { "grad_norm": 0.32383644580841064, "learning_rate": 3.3390256997120505e-06, "loss": 0.0158, "step": 71110 }, { "grad_norm": 0.2767280340194702, "learning_rate": 3.331603399187583e-06, "loss": 0.0138, "step": 71120 }, { "grad_norm": 0.08197338879108429, "learning_rate": 3.324189073015721e-06, "loss": 0.0101, "step": 71130 }, { "grad_norm": 0.2150905281305313, "learning_rate": 3.3167827224633465e-06, "loss": 0.0127, "step": 71140 }, { "grad_norm": 0.20537903904914856, "learning_rate": 3.309384348796024e-06, "loss": 0.0152, "step": 71150 }, { "grad_norm": 0.1656138151884079, "learning_rate": 3.301993953277921e-06, "loss": 0.0137, "step": 71160 }, { "grad_norm": 0.39676985144615173, "learning_rate": 3.29461153717186e-06, "loss": 0.0131, "step": 71170 }, { "grad_norm": 0.22789399325847626, "learning_rate": 3.287237101739293e-06, "loss": 0.0156, "step": 71180 }, { "grad_norm": 0.35451540350914, "learning_rate": 3.2798706482403075e-06, "loss": 0.0185, "step": 71190 }, { "grad_norm": 0.38175734877586365, "learning_rate": 3.2725121779336285e-06, "loss": 0.0124, "step": 71200 }, { "grad_norm": 0.3123573362827301, "learning_rate": 3.265161692076618e-06, "loss": 0.0166, "step": 71210 }, { "grad_norm": 0.09212899208068848, "learning_rate": 3.257819191925282e-06, "loss": 0.0088, "step": 71220 }, { "grad_norm": 0.23329102993011475, "learning_rate": 3.2504846787342392e-06, "loss": 0.0131, "step": 71230 }, { "grad_norm": 0.34950965642929077, "learning_rate": 3.243158153756759e-06, "loss": 0.0112, "step": 71240 }, { "grad_norm": 0.1270274668931961, "learning_rate": 3.2358396182447514e-06, "loss": 0.0081, "step": 71250 }, { "grad_norm": 0.18943150341510773, "learning_rate": 3.2285290734487496e-06, "loss": 0.0112, "step": 71260 }, { "grad_norm": 0.1142297014594078, "learning_rate": 3.221226520617926e-06, "loss": 0.0137, "step": 71270 }, { "grad_norm": 0.19680428504943848, "learning_rate": 3.213931961000094e-06, "loss": 0.0165, "step": 71280 }, { "grad_norm": 0.26676034927368164, "learning_rate": 3.2066453958416733e-06, "loss": 0.0081, "step": 71290 }, { "grad_norm": 0.2795449197292328, "learning_rate": 3.199366826387773e-06, "loss": 0.0129, "step": 71300 }, { "grad_norm": 0.21593450009822845, "learning_rate": 3.192096253882071e-06, "loss": 0.012, "step": 71310 }, { "grad_norm": 0.28072839975357056, "learning_rate": 3.1848336795669177e-06, "loss": 0.0131, "step": 71320 }, { "grad_norm": 0.11633823066949844, "learning_rate": 3.1775791046833035e-06, "loss": 0.0093, "step": 71330 }, { "grad_norm": 0.234633207321167, "learning_rate": 3.170332530470804e-06, "loss": 0.0112, "step": 71340 }, { "grad_norm": 0.5203905701637268, "learning_rate": 3.1630939581677012e-06, "loss": 0.0162, "step": 71350 }, { "grad_norm": 0.25503483414649963, "learning_rate": 3.155863389010838e-06, "loss": 0.0125, "step": 71360 }, { "grad_norm": 0.1960696280002594, "learning_rate": 3.1486408242357323e-06, "loss": 0.0136, "step": 71370 }, { "grad_norm": 0.4116365611553192, "learning_rate": 3.1414262650765248e-06, "loss": 0.0275, "step": 71380 }, { "grad_norm": 0.25317609310150146, "learning_rate": 3.134219712765979e-06, "loss": 0.0156, "step": 71390 }, { "grad_norm": 0.14501580595970154, "learning_rate": 3.1270211685355044e-06, "loss": 0.0144, "step": 71400 }, { "grad_norm": 0.2090539038181305, "learning_rate": 3.1198306336151338e-06, "loss": 0.0101, "step": 71410 }, { "grad_norm": 0.1891876459121704, "learning_rate": 3.11264810923354e-06, "loss": 0.0117, "step": 71420 }, { "grad_norm": 0.06700609624385834, "learning_rate": 3.105473596618008e-06, "loss": 0.0088, "step": 71430 }, { "grad_norm": 0.2099655270576477, "learning_rate": 3.0983070969944683e-06, "loss": 0.0119, "step": 71440 }, { "grad_norm": 0.5833874940872192, "learning_rate": 3.09114861158748e-06, "loss": 0.0131, "step": 71450 }, { "grad_norm": 0.10837140679359436, "learning_rate": 3.0839981416202314e-06, "loss": 0.0107, "step": 71460 }, { "grad_norm": 0.222650408744812, "learning_rate": 3.0768556883145507e-06, "loss": 0.0094, "step": 71470 }, { "grad_norm": 0.20529362559318542, "learning_rate": 3.0697212528908834e-06, "loss": 0.011, "step": 71480 }, { "grad_norm": 0.20885160565376282, "learning_rate": 3.0625948365683e-06, "loss": 0.0118, "step": 71490 }, { "grad_norm": 0.1327100694179535, "learning_rate": 3.055476440564531e-06, "loss": 0.0154, "step": 71500 }, { "grad_norm": 0.2156619131565094, "learning_rate": 3.0483660660958924e-06, "loss": 0.012, "step": 71510 }, { "grad_norm": 0.3723852336406708, "learning_rate": 3.0412637143773624e-06, "loss": 0.0199, "step": 71520 }, { "grad_norm": 0.2567701041698456, "learning_rate": 3.034169386622554e-06, "loss": 0.009, "step": 71530 }, { "grad_norm": 0.16345931589603424, "learning_rate": 3.027083084043658e-06, "loss": 0.0133, "step": 71540 }, { "grad_norm": 0.16053912043571472, "learning_rate": 3.0200048078515676e-06, "loss": 0.0119, "step": 71550 }, { "grad_norm": 0.2504352629184723, "learning_rate": 3.012934559255742e-06, "loss": 0.0123, "step": 71560 }, { "grad_norm": 0.2713932991027832, "learning_rate": 3.0058723394642994e-06, "loss": 0.0121, "step": 71570 }, { "grad_norm": 0.21975332498550415, "learning_rate": 2.998818149683985e-06, "loss": 0.0174, "step": 71580 }, { "grad_norm": 0.21527060866355896, "learning_rate": 2.9917719911201627e-06, "loss": 0.0125, "step": 71590 }, { "grad_norm": 0.2544984817504883, "learning_rate": 2.9847338649768352e-06, "loss": 0.0121, "step": 71600 }, { "grad_norm": 0.21102991700172424, "learning_rate": 2.977703772456608e-06, "loss": 0.0184, "step": 71610 }, { "grad_norm": 0.23388569056987762, "learning_rate": 2.9706817147607535e-06, "loss": 0.0109, "step": 71620 }, { "grad_norm": 0.2613407075405121, "learning_rate": 2.9636676930891393e-06, "loss": 0.0131, "step": 71630 }, { "grad_norm": 0.373918354511261, "learning_rate": 2.9566617086402625e-06, "loss": 0.0127, "step": 71640 }, { "grad_norm": 0.1897059679031372, "learning_rate": 2.9496637626112655e-06, "loss": 0.0113, "step": 71650 }, { "grad_norm": 0.27432066202163696, "learning_rate": 2.9426738561979027e-06, "loss": 0.0096, "step": 71660 }, { "grad_norm": 0.39727193117141724, "learning_rate": 2.9356919905945524e-06, "loss": 0.0266, "step": 71670 }, { "grad_norm": 0.1811302751302719, "learning_rate": 2.928718166994243e-06, "loss": 0.0119, "step": 71680 }, { "grad_norm": 0.3530515432357788, "learning_rate": 2.9217523865885833e-06, "loss": 0.0152, "step": 71690 }, { "grad_norm": 0.24688079953193665, "learning_rate": 2.9147946505678537e-06, "loss": 0.012, "step": 71700 }, { "grad_norm": 0.13138025999069214, "learning_rate": 2.9078449601209313e-06, "loss": 0.0108, "step": 71710 }, { "grad_norm": 0.4087546169757843, "learning_rate": 2.900903316435333e-06, "loss": 0.0132, "step": 71720 }, { "grad_norm": 0.2653743624687195, "learning_rate": 2.8939697206971983e-06, "loss": 0.0125, "step": 71730 }, { "grad_norm": 0.21900956332683563, "learning_rate": 2.8870441740912746e-06, "loss": 0.0115, "step": 71740 }, { "grad_norm": 0.21227610111236572, "learning_rate": 2.8801266778009762e-06, "loss": 0.0142, "step": 71750 }, { "grad_norm": 0.1609942615032196, "learning_rate": 2.8732172330082798e-06, "loss": 0.0157, "step": 71760 }, { "grad_norm": 0.1682349145412445, "learning_rate": 2.8663158408938517e-06, "loss": 0.0139, "step": 71770 }, { "grad_norm": 0.14635294675827026, "learning_rate": 2.859422502636938e-06, "loss": 0.0131, "step": 71780 }, { "grad_norm": 0.21470032632350922, "learning_rate": 2.8525372194154076e-06, "loss": 0.0111, "step": 71790 }, { "grad_norm": 0.1735251545906067, "learning_rate": 2.8456599924057913e-06, "loss": 0.0115, "step": 71800 }, { "grad_norm": 0.10447487980127335, "learning_rate": 2.8387908227831995e-06, "loss": 0.0122, "step": 71810 }, { "grad_norm": 0.1153714582324028, "learning_rate": 2.831929711721404e-06, "loss": 0.0113, "step": 71820 }, { "grad_norm": 0.2685800790786743, "learning_rate": 2.825076660392767e-06, "loss": 0.0124, "step": 71830 }, { "grad_norm": 0.1826363056898117, "learning_rate": 2.8182316699682908e-06, "loss": 0.0162, "step": 71840 }, { "grad_norm": 0.12377209961414337, "learning_rate": 2.811394741617601e-06, "loss": 0.0121, "step": 71850 }, { "grad_norm": 0.39323559403419495, "learning_rate": 2.8045658765089356e-06, "loss": 0.0142, "step": 71860 }, { "grad_norm": 0.19857637584209442, "learning_rate": 2.7977450758091605e-06, "loss": 0.0167, "step": 71870 }, { "grad_norm": 0.18427135050296783, "learning_rate": 2.790932340683783e-06, "loss": 0.0132, "step": 71880 }, { "grad_norm": 0.8563022017478943, "learning_rate": 2.7841276722968823e-06, "loss": 0.0117, "step": 71890 }, { "grad_norm": 0.3750643730163574, "learning_rate": 2.7773310718112067e-06, "loss": 0.012, "step": 71900 }, { "grad_norm": 0.264441579580307, "learning_rate": 2.7705425403881102e-06, "loss": 0.0137, "step": 71910 }, { "grad_norm": 0.5359944105148315, "learning_rate": 2.763762079187565e-06, "loss": 0.0161, "step": 71920 }, { "grad_norm": 0.1468634307384491, "learning_rate": 2.756989689368178e-06, "loss": 0.0127, "step": 71930 }, { "grad_norm": 0.23220118880271912, "learning_rate": 2.750225372087134e-06, "loss": 0.0136, "step": 71940 }, { "grad_norm": 0.3246726989746094, "learning_rate": 2.7434691285003033e-06, "loss": 0.0091, "step": 71950 }, { "grad_norm": 0.19425907731056213, "learning_rate": 2.7367209597621178e-06, "loss": 0.0135, "step": 71960 }, { "grad_norm": 0.2317807674407959, "learning_rate": 2.729980867025683e-06, "loss": 0.0133, "step": 71970 }, { "grad_norm": 0.26500821113586426, "learning_rate": 2.7232488514426724e-06, "loss": 0.0116, "step": 71980 }, { "grad_norm": 0.17867761850357056, "learning_rate": 2.7165249141634097e-06, "loss": 0.0093, "step": 71990 }, { "grad_norm": 0.24949932098388672, "learning_rate": 2.709809056336837e-06, "loss": 0.0126, "step": 72000 }, { "grad_norm": 0.2658992111682892, "learning_rate": 2.703101279110498e-06, "loss": 0.0121, "step": 72010 }, { "grad_norm": 0.28478696942329407, "learning_rate": 2.696401583630587e-06, "loss": 0.0113, "step": 72020 }, { "grad_norm": 0.18532703816890717, "learning_rate": 2.689709971041887e-06, "loss": 0.0101, "step": 72030 }, { "grad_norm": 0.24110645055770874, "learning_rate": 2.6830264424878126e-06, "loss": 0.0088, "step": 72040 }, { "grad_norm": 0.18314921855926514, "learning_rate": 2.676350999110394e-06, "loss": 0.0182, "step": 72050 }, { "grad_norm": 0.14251667261123657, "learning_rate": 2.6696836420502856e-06, "loss": 0.02, "step": 72060 }, { "grad_norm": 0.18374191224575043, "learning_rate": 2.663024372446765e-06, "loss": 0.0097, "step": 72070 }, { "grad_norm": 0.18528053164482117, "learning_rate": 2.6563731914377055e-06, "loss": 0.0089, "step": 72080 }, { "grad_norm": 0.20982488989830017, "learning_rate": 2.6497301001596087e-06, "loss": 0.0173, "step": 72090 }, { "grad_norm": 0.3494599163532257, "learning_rate": 2.643095099747611e-06, "loss": 0.0142, "step": 72100 }, { "grad_norm": 0.17118847370147705, "learning_rate": 2.636468191335445e-06, "loss": 0.0121, "step": 72110 }, { "grad_norm": 0.304232120513916, "learning_rate": 2.629849376055471e-06, "loss": 0.014, "step": 72120 }, { "grad_norm": 0.20196595788002014, "learning_rate": 2.6232386550386678e-06, "loss": 0.0115, "step": 72130 }, { "grad_norm": 0.18196812272071838, "learning_rate": 2.6166360294146097e-06, "loss": 0.0101, "step": 72140 }, { "grad_norm": 0.2184862494468689, "learning_rate": 2.6100415003115275e-06, "loss": 0.0176, "step": 72150 }, { "grad_norm": 0.2035028636455536, "learning_rate": 2.603455068856225e-06, "loss": 0.0157, "step": 72160 }, { "grad_norm": 0.20125527679920197, "learning_rate": 2.5968767361741584e-06, "loss": 0.0141, "step": 72170 }, { "grad_norm": 0.21671532094478607, "learning_rate": 2.590306503389378e-06, "loss": 0.0139, "step": 72180 }, { "grad_norm": 0.14824794232845306, "learning_rate": 2.583744371624558e-06, "loss": 0.0164, "step": 72190 }, { "grad_norm": 0.2692915201187134, "learning_rate": 2.57719034200099e-06, "loss": 0.0102, "step": 72200 }, { "grad_norm": 0.20750752091407776, "learning_rate": 2.570644415638568e-06, "loss": 0.0111, "step": 72210 }, { "grad_norm": 0.4341175854206085, "learning_rate": 2.5641065936558295e-06, "loss": 0.0124, "step": 72220 }, { "grad_norm": 0.08006125688552856, "learning_rate": 2.557576877169898e-06, "loss": 0.0104, "step": 72230 }, { "grad_norm": 0.2715272605419159, "learning_rate": 2.5510552672965205e-06, "loss": 0.009, "step": 72240 }, { "grad_norm": 0.1786516159772873, "learning_rate": 2.544541765150077e-06, "loss": 0.0158, "step": 72250 }, { "grad_norm": 0.3813740313053131, "learning_rate": 2.5380363718435165e-06, "loss": 0.0128, "step": 72260 }, { "grad_norm": 0.13487066328525543, "learning_rate": 2.5315390884884714e-06, "loss": 0.0098, "step": 72270 }, { "grad_norm": 0.22197632491588593, "learning_rate": 2.5250499161951214e-06, "loss": 0.0135, "step": 72280 }, { "grad_norm": 0.16218192875385284, "learning_rate": 2.518568856072301e-06, "loss": 0.0102, "step": 72290 }, { "grad_norm": 0.20893333852291107, "learning_rate": 2.512095909227441e-06, "loss": 0.0119, "step": 72300 }, { "grad_norm": 0.22956624627113342, "learning_rate": 2.50563107676659e-06, "loss": 0.0129, "step": 72310 }, { "grad_norm": 0.13632997870445251, "learning_rate": 2.4991743597944152e-06, "loss": 0.0138, "step": 72320 }, { "grad_norm": 0.11603113263845444, "learning_rate": 2.492725759414205e-06, "loss": 0.0126, "step": 72330 }, { "grad_norm": 0.29443395137786865, "learning_rate": 2.4862852767278234e-06, "loss": 0.0087, "step": 72340 }, { "grad_norm": 0.36711448431015015, "learning_rate": 2.479852912835784e-06, "loss": 0.0134, "step": 72350 }, { "grad_norm": 0.22706672549247742, "learning_rate": 2.4734286688372075e-06, "loss": 0.0129, "step": 72360 }, { "grad_norm": 0.7971838116645813, "learning_rate": 2.467012545829811e-06, "loss": 0.0145, "step": 72370 }, { "grad_norm": 0.2372976690530777, "learning_rate": 2.460604544909945e-06, "loss": 0.0096, "step": 72380 }, { "grad_norm": 0.2311311513185501, "learning_rate": 2.454204667172555e-06, "loss": 0.0167, "step": 72390 }, { "grad_norm": 0.2067732810974121, "learning_rate": 2.447812913711217e-06, "loss": 0.0105, "step": 72400 }, { "grad_norm": 0.2725408971309662, "learning_rate": 2.44142928561808e-06, "loss": 0.0147, "step": 72410 }, { "grad_norm": 0.18753817677497864, "learning_rate": 2.435053783983965e-06, "loss": 0.0112, "step": 72420 }, { "grad_norm": 0.28774169087409973, "learning_rate": 2.4286864098982453e-06, "loss": 0.014, "step": 72430 }, { "grad_norm": 0.10871447622776031, "learning_rate": 2.4223271644489397e-06, "loss": 0.0152, "step": 72440 }, { "grad_norm": 0.27759024500846863, "learning_rate": 2.415976048722679e-06, "loss": 0.0096, "step": 72450 }, { "grad_norm": 0.33734825253486633, "learning_rate": 2.4096330638046673e-06, "loss": 0.0078, "step": 72460 }, { "grad_norm": 0.19679826498031616, "learning_rate": 2.4032982107787816e-06, "loss": 0.0124, "step": 72470 }, { "grad_norm": 0.2617242932319641, "learning_rate": 2.396971490727451e-06, "loss": 0.0153, "step": 72480 }, { "grad_norm": 0.1935802549123764, "learning_rate": 2.3906529047317493e-06, "loss": 0.0101, "step": 72490 }, { "grad_norm": 0.18321573734283447, "learning_rate": 2.3843424538713465e-06, "loss": 0.0137, "step": 72500 }, { "grad_norm": 0.3274153470993042, "learning_rate": 2.3780401392245298e-06, "loss": 0.0118, "step": 72510 }, { "grad_norm": 0.23400789499282837, "learning_rate": 2.3717459618681883e-06, "loss": 0.013, "step": 72520 }, { "grad_norm": 0.19555887579917908, "learning_rate": 2.3654599228778274e-06, "loss": 0.0101, "step": 72530 }, { "grad_norm": 0.26144272089004517, "learning_rate": 2.3591820233275607e-06, "loss": 0.0137, "step": 72540 }, { "grad_norm": 0.21173378825187683, "learning_rate": 2.3529122642901024e-06, "loss": 0.0146, "step": 72550 }, { "grad_norm": 0.16517645120620728, "learning_rate": 2.3466506468367845e-06, "loss": 0.0188, "step": 72560 }, { "grad_norm": 0.39878034591674805, "learning_rate": 2.3403971720375506e-06, "loss": 0.0111, "step": 72570 }, { "grad_norm": 0.3460572063922882, "learning_rate": 2.334151840960952e-06, "loss": 0.0089, "step": 72580 }, { "grad_norm": 0.11964350938796997, "learning_rate": 2.3279146546741347e-06, "loss": 0.0099, "step": 72590 }, { "grad_norm": 0.128351092338562, "learning_rate": 2.32168561424288e-06, "loss": 0.0103, "step": 72600 }, { "grad_norm": 0.43417587876319885, "learning_rate": 2.3154647207315307e-06, "loss": 0.0153, "step": 72610 }, { "grad_norm": 0.17493993043899536, "learning_rate": 2.309251975203103e-06, "loss": 0.015, "step": 72620 }, { "grad_norm": 0.2825966477394104, "learning_rate": 2.303047378719159e-06, "loss": 0.0127, "step": 72630 }, { "grad_norm": 0.24209830164909363, "learning_rate": 2.2968509323399e-06, "loss": 0.013, "step": 72640 }, { "grad_norm": 0.15448226034641266, "learning_rate": 2.290662637124147e-06, "loss": 0.0108, "step": 72650 }, { "grad_norm": 0.11913652718067169, "learning_rate": 2.2844824941292807e-06, "loss": 0.0078, "step": 72660 }, { "grad_norm": 0.27292582392692566, "learning_rate": 2.2783105044113406e-06, "loss": 0.0124, "step": 72670 }, { "grad_norm": 0.30601146817207336, "learning_rate": 2.272146669024944e-06, "loss": 0.0095, "step": 72680 }, { "grad_norm": 0.14794385433197021, "learning_rate": 2.26599098902332e-06, "loss": 0.0118, "step": 72690 }, { "grad_norm": 0.21103732287883759, "learning_rate": 2.2598434654583113e-06, "loss": 0.012, "step": 72700 }, { "grad_norm": 0.3489603102207184, "learning_rate": 2.253704099380355e-06, "loss": 0.0184, "step": 72710 }, { "grad_norm": 0.19907909631729126, "learning_rate": 2.247572891838512e-06, "loss": 0.0122, "step": 72720 }, { "grad_norm": 0.2540438771247864, "learning_rate": 2.241449843880422e-06, "loss": 0.015, "step": 72730 }, { "grad_norm": 0.07080300152301788, "learning_rate": 2.235334956552354e-06, "loss": 0.0123, "step": 72740 }, { "grad_norm": 0.4291803240776062, "learning_rate": 2.2292282308991775e-06, "loss": 0.0163, "step": 72750 }, { "grad_norm": 0.19149525463581085, "learning_rate": 2.223129667964363e-06, "loss": 0.0119, "step": 72760 }, { "grad_norm": 0.2876875102519989, "learning_rate": 2.2170392687899834e-06, "loss": 0.0114, "step": 72770 }, { "grad_norm": 0.2027886062860489, "learning_rate": 2.210957034416733e-06, "loss": 0.0113, "step": 72780 }, { "grad_norm": 0.2498544156551361, "learning_rate": 2.2048829658838867e-06, "loss": 0.0152, "step": 72790 }, { "grad_norm": 0.28823569416999817, "learning_rate": 2.1988170642293525e-06, "loss": 0.0087, "step": 72800 }, { "grad_norm": 0.16215144097805023, "learning_rate": 2.1927593304896075e-06, "loss": 0.0091, "step": 72810 }, { "grad_norm": 0.14579641819000244, "learning_rate": 2.1867097656997626e-06, "loss": 0.0105, "step": 72820 }, { "grad_norm": 0.1680118590593338, "learning_rate": 2.180668370893524e-06, "loss": 0.0118, "step": 72830 }, { "grad_norm": 0.7580523490905762, "learning_rate": 2.174635147103199e-06, "loss": 0.0111, "step": 72840 }, { "grad_norm": 0.16513633728027344, "learning_rate": 2.1686100953597075e-06, "loss": 0.0108, "step": 72850 }, { "grad_norm": 0.2138228714466095, "learning_rate": 2.1625932166925433e-06, "loss": 0.0106, "step": 72860 }, { "grad_norm": 0.15294840931892395, "learning_rate": 2.1565845121298556e-06, "loss": 0.0121, "step": 72870 }, { "grad_norm": 0.3321465253829956, "learning_rate": 2.150583982698351e-06, "loss": 0.0139, "step": 72880 }, { "grad_norm": 0.11809127032756805, "learning_rate": 2.144591629423359e-06, "loss": 0.01, "step": 72890 }, { "grad_norm": 0.14062657952308655, "learning_rate": 2.138607453328817e-06, "loss": 0.0133, "step": 72900 }, { "grad_norm": 0.18757271766662598, "learning_rate": 2.132631455437234e-06, "loss": 0.011, "step": 72910 }, { "grad_norm": 0.21767109632492065, "learning_rate": 2.1266636367697714e-06, "loss": 0.0187, "step": 72920 }, { "grad_norm": 0.2737908661365509, "learning_rate": 2.120703998346152e-06, "loss": 0.0111, "step": 72930 }, { "grad_norm": 0.2379523664712906, "learning_rate": 2.1147525411847114e-06, "loss": 0.0148, "step": 72940 }, { "grad_norm": 0.20168785750865936, "learning_rate": 2.108809266302403e-06, "loss": 0.0113, "step": 72950 }, { "grad_norm": 0.2641315758228302, "learning_rate": 2.102874174714764e-06, "loss": 0.014, "step": 72960 }, { "grad_norm": 0.21050438284873962, "learning_rate": 2.0969472674359393e-06, "loss": 0.0098, "step": 72970 }, { "grad_norm": 0.10053959488868713, "learning_rate": 2.091028545478668e-06, "loss": 0.0121, "step": 72980 }, { "grad_norm": 0.1426243931055069, "learning_rate": 2.08511800985432e-06, "loss": 0.0093, "step": 72990 }, { "grad_norm": 0.18063805997371674, "learning_rate": 2.07921566157282e-06, "loss": 0.0178, "step": 73000 }, { "grad_norm": 0.1731051653623581, "learning_rate": 2.073321501642728e-06, "loss": 0.0134, "step": 73010 }, { "grad_norm": 0.11222461611032486, "learning_rate": 2.0674355310711937e-06, "loss": 0.0125, "step": 73020 }, { "grad_norm": 0.14246506989002228, "learning_rate": 2.0615577508639682e-06, "loss": 0.0078, "step": 73030 }, { "grad_norm": 0.08966214209794998, "learning_rate": 2.055688162025399e-06, "loss": 0.01, "step": 73040 }, { "grad_norm": 0.14990867674350739, "learning_rate": 2.0498267655584546e-06, "loss": 0.014, "step": 73050 }, { "grad_norm": 0.12798558175563812, "learning_rate": 2.0439735624646626e-06, "loss": 0.0118, "step": 73060 }, { "grad_norm": 0.07625731080770493, "learning_rate": 2.0381285537442008e-06, "loss": 0.0089, "step": 73070 }, { "grad_norm": 0.22310864925384521, "learning_rate": 2.032291740395803e-06, "loss": 0.0118, "step": 73080 }, { "grad_norm": 0.39653947949409485, "learning_rate": 2.0264631234168276e-06, "loss": 0.0134, "step": 73090 }, { "grad_norm": 0.42780762910842896, "learning_rate": 2.0206427038032326e-06, "loss": 0.016, "step": 73100 }, { "grad_norm": 0.22723498940467834, "learning_rate": 2.0148304825495457e-06, "loss": 0.0133, "step": 73110 }, { "grad_norm": 0.22366935014724731, "learning_rate": 2.0090264606489496e-06, "loss": 0.0117, "step": 73120 }, { "grad_norm": 0.27830320596694946, "learning_rate": 2.003230639093162e-06, "loss": 0.0123, "step": 73130 }, { "grad_norm": 0.38823413848876953, "learning_rate": 1.9974430188725525e-06, "loss": 0.0161, "step": 73140 }, { "grad_norm": 0.1972135603427887, "learning_rate": 1.9916636009760513e-06, "loss": 0.0137, "step": 73150 }, { "grad_norm": 0.11566104739904404, "learning_rate": 1.985892386391208e-06, "loss": 0.0142, "step": 73160 }, { "grad_norm": 0.16770584881305695, "learning_rate": 1.980129376104173e-06, "loss": 0.0105, "step": 73170 }, { "grad_norm": 0.19140486419200897, "learning_rate": 1.9743745710996796e-06, "loss": 0.0091, "step": 73180 }, { "grad_norm": 0.20937803387641907, "learning_rate": 1.9686279723610757e-06, "loss": 0.0125, "step": 73190 }, { "grad_norm": 0.18319225311279297, "learning_rate": 1.962889580870281e-06, "loss": 0.0106, "step": 73200 }, { "grad_norm": 0.1723104566335678, "learning_rate": 1.9571593976078385e-06, "loss": 0.0141, "step": 73210 }, { "grad_norm": 0.2806481719017029, "learning_rate": 1.9514374235528876e-06, "loss": 0.0116, "step": 73220 }, { "grad_norm": 0.3036612272262573, "learning_rate": 1.9457236596831408e-06, "loss": 0.0078, "step": 73230 }, { "grad_norm": 0.32590118050575256, "learning_rate": 1.940018106974939e-06, "loss": 0.0133, "step": 73240 }, { "grad_norm": 0.2228996902704239, "learning_rate": 1.9343207664032025e-06, "loss": 0.0111, "step": 73250 }, { "grad_norm": 0.15099112689495087, "learning_rate": 1.928631638941436e-06, "loss": 0.011, "step": 73260 }, { "grad_norm": 0.20707470178604126, "learning_rate": 1.922950725561784e-06, "loss": 0.0117, "step": 73270 }, { "grad_norm": 0.2551286220550537, "learning_rate": 1.9172780272349312e-06, "loss": 0.0089, "step": 73280 }, { "grad_norm": 0.17703671753406525, "learning_rate": 1.911613544930202e-06, "loss": 0.0136, "step": 73290 }, { "grad_norm": 0.2250799983739853, "learning_rate": 1.9059572796155e-06, "loss": 0.0121, "step": 73300 }, { "grad_norm": 0.15633946657180786, "learning_rate": 1.9003092322573192e-06, "loss": 0.0115, "step": 73310 }, { "grad_norm": 0.13180361688137054, "learning_rate": 1.8946694038207647e-06, "loss": 0.0119, "step": 73320 }, { "grad_norm": 0.3369571566581726, "learning_rate": 1.8890377952695215e-06, "loss": 0.0097, "step": 73330 }, { "grad_norm": 0.12118212133646011, "learning_rate": 1.8834144075658865e-06, "loss": 0.009, "step": 73340 }, { "grad_norm": 0.28325751423835754, "learning_rate": 1.8777992416707357e-06, "loss": 0.0116, "step": 73350 }, { "grad_norm": 0.17204707860946655, "learning_rate": 1.8721922985435458e-06, "loss": 0.009, "step": 73360 }, { "grad_norm": 0.3929659426212311, "learning_rate": 1.8665935791424006e-06, "loss": 0.0195, "step": 73370 }, { "grad_norm": 0.1664552241563797, "learning_rate": 1.8610030844239512e-06, "loss": 0.0099, "step": 73380 }, { "grad_norm": 0.18747572600841522, "learning_rate": 1.8554208153434838e-06, "loss": 0.0117, "step": 73390 }, { "grad_norm": 0.118034727871418, "learning_rate": 1.8498467728548296e-06, "loss": 0.0104, "step": 73400 }, { "grad_norm": 0.2205413430929184, "learning_rate": 1.8442809579104547e-06, "loss": 0.0098, "step": 73410 }, { "grad_norm": 0.3127245306968689, "learning_rate": 1.8387233714614039e-06, "loss": 0.0133, "step": 73420 }, { "grad_norm": 0.11896590888500214, "learning_rate": 1.8331740144573174e-06, "loss": 0.0201, "step": 73430 }, { "grad_norm": 0.32620686292648315, "learning_rate": 1.8276328878464255e-06, "loss": 0.0142, "step": 73440 }, { "grad_norm": 0.3770099878311157, "learning_rate": 1.82209999257556e-06, "loss": 0.0178, "step": 73450 }, { "grad_norm": 0.1506708264350891, "learning_rate": 1.8165753295901311e-06, "loss": 0.0096, "step": 73460 }, { "grad_norm": 0.19984830915927887, "learning_rate": 1.8110588998341616e-06, "loss": 0.0121, "step": 73470 }, { "grad_norm": 0.17101708054542542, "learning_rate": 1.805550704250253e-06, "loss": 0.0173, "step": 73480 }, { "grad_norm": 0.22447501122951508, "learning_rate": 1.8000507437796077e-06, "loss": 0.0112, "step": 73490 }, { "grad_norm": 0.18101146817207336, "learning_rate": 1.7945590193620242e-06, "loss": 0.0117, "step": 73500 }, { "grad_norm": 0.20167365670204163, "learning_rate": 1.7890755319358742e-06, "loss": 0.0107, "step": 73510 }, { "grad_norm": 0.2595217823982239, "learning_rate": 1.7836002824381525e-06, "loss": 0.0146, "step": 73520 }, { "grad_norm": 0.3048449158668518, "learning_rate": 1.7781332718044165e-06, "loss": 0.011, "step": 73530 }, { "grad_norm": 0.30162957310676575, "learning_rate": 1.772674500968835e-06, "loss": 0.0126, "step": 73540 }, { "grad_norm": 0.23611284792423248, "learning_rate": 1.7672239708641624e-06, "loss": 0.0124, "step": 73550 }, { "grad_norm": 0.16203424334526062, "learning_rate": 1.7617816824217371e-06, "loss": 0.0114, "step": 73560 }, { "grad_norm": 0.21837715804576874, "learning_rate": 1.7563476365715148e-06, "loss": 0.0174, "step": 73570 }, { "grad_norm": 0.12214773893356323, "learning_rate": 1.750921834242003e-06, "loss": 0.0148, "step": 73580 }, { "grad_norm": 0.4438193142414093, "learning_rate": 1.7455042763603436e-06, "loss": 0.0181, "step": 73590 }, { "grad_norm": 0.246726393699646, "learning_rate": 1.7400949638522345e-06, "loss": 0.0095, "step": 73600 }, { "grad_norm": 0.3704357445240021, "learning_rate": 1.734693897641987e-06, "loss": 0.0262, "step": 73610 }, { "grad_norm": 0.25729402899742126, "learning_rate": 1.7293010786524955e-06, "loss": 0.0131, "step": 73620 }, { "grad_norm": 0.22470013797283173, "learning_rate": 1.72391650780524e-06, "loss": 0.0122, "step": 73630 }, { "grad_norm": 0.3089427649974823, "learning_rate": 1.718540186020301e-06, "loss": 0.0145, "step": 73640 }, { "grad_norm": 0.14895689487457275, "learning_rate": 1.7131721142163437e-06, "loss": 0.0159, "step": 73650 }, { "grad_norm": 0.27363264560699463, "learning_rate": 1.7078122933106233e-06, "loss": 0.0144, "step": 73660 }, { "grad_norm": 0.12088077515363693, "learning_rate": 1.7024607242189905e-06, "loss": 0.0142, "step": 73670 }, { "grad_norm": 0.2870631515979767, "learning_rate": 1.6971174078558749e-06, "loss": 0.0085, "step": 73680 }, { "grad_norm": 0.08863799273967743, "learning_rate": 1.6917823451343073e-06, "loss": 0.0121, "step": 73690 }, { "grad_norm": 0.35506102442741394, "learning_rate": 1.6864555369659142e-06, "loss": 0.012, "step": 73700 }, { "grad_norm": 0.20239916443824768, "learning_rate": 1.681136984260878e-06, "loss": 0.0132, "step": 73710 }, { "grad_norm": 0.22089186310768127, "learning_rate": 1.6758266879280172e-06, "loss": 0.0153, "step": 73720 }, { "grad_norm": 0.266768217086792, "learning_rate": 1.6705246488746996e-06, "loss": 0.0125, "step": 73730 }, { "grad_norm": 0.22717119753360748, "learning_rate": 1.6652308680069062e-06, "loss": 0.0091, "step": 73740 }, { "grad_norm": 0.26835671067237854, "learning_rate": 1.6599453462292081e-06, "loss": 0.0091, "step": 73750 }, { "grad_norm": 0.5637785792350769, "learning_rate": 1.6546680844447326e-06, "loss": 0.0165, "step": 73760 }, { "grad_norm": 0.17052434384822845, "learning_rate": 1.6493990835552475e-06, "loss": 0.0107, "step": 73770 }, { "grad_norm": 0.1424383968114853, "learning_rate": 1.6441383444610547e-06, "loss": 0.0109, "step": 73780 }, { "grad_norm": 0.21772617101669312, "learning_rate": 1.6388858680610908e-06, "loss": 0.0131, "step": 73790 }, { "grad_norm": 0.19381596148014069, "learning_rate": 1.6336416552528543e-06, "loss": 0.0115, "step": 73800 }, { "grad_norm": 0.13090640306472778, "learning_rate": 1.6284057069324288e-06, "loss": 0.018, "step": 73810 }, { "grad_norm": 0.1264895349740982, "learning_rate": 1.623178023994504e-06, "loss": 0.0096, "step": 73820 }, { "grad_norm": 0.5165658593177795, "learning_rate": 1.6179586073323483e-06, "loss": 0.0128, "step": 73830 }, { "grad_norm": 0.2972957193851471, "learning_rate": 1.612747457837821e-06, "loss": 0.0135, "step": 73840 }, { "grad_norm": 0.2651349604129791, "learning_rate": 1.607544576401354e-06, "loss": 0.0099, "step": 73850 }, { "grad_norm": 0.40619856119155884, "learning_rate": 1.6023499639119754e-06, "loss": 0.0135, "step": 73860 }, { "grad_norm": 0.4999363422393799, "learning_rate": 1.5971636212573138e-06, "loss": 0.0101, "step": 73870 }, { "grad_norm": 0.2827552258968353, "learning_rate": 1.5919855493235714e-06, "loss": 0.0122, "step": 73880 }, { "grad_norm": 0.2287859320640564, "learning_rate": 1.586815748995535e-06, "loss": 0.0095, "step": 73890 }, { "grad_norm": 0.3292054235935211, "learning_rate": 1.5816542211565866e-06, "loss": 0.0122, "step": 73900 }, { "grad_norm": 0.19388723373413086, "learning_rate": 1.5765009666886765e-06, "loss": 0.0091, "step": 73910 }, { "grad_norm": 0.35102832317352295, "learning_rate": 1.5713559864723781e-06, "loss": 0.0104, "step": 73920 }, { "grad_norm": 0.21355551481246948, "learning_rate": 1.5662192813868104e-06, "loss": 0.0092, "step": 73930 }, { "grad_norm": 0.14587648212909698, "learning_rate": 1.561090852309699e-06, "loss": 0.0157, "step": 73940 }, { "grad_norm": 0.4774894714355469, "learning_rate": 1.5559707001173651e-06, "loss": 0.0171, "step": 73950 }, { "grad_norm": 0.28344088792800903, "learning_rate": 1.5508588256846757e-06, "loss": 0.0137, "step": 73960 }, { "grad_norm": 0.10996870696544647, "learning_rate": 1.5457552298851319e-06, "loss": 0.0095, "step": 73970 }, { "grad_norm": 0.23142701387405396, "learning_rate": 1.5406599135907918e-06, "loss": 0.0121, "step": 73980 }, { "grad_norm": 0.2440231889486313, "learning_rate": 1.5355728776723088e-06, "loss": 0.0105, "step": 73990 }, { "grad_norm": 0.155409574508667, "learning_rate": 1.5304941229989155e-06, "loss": 0.0134, "step": 74000 }, { "grad_norm": 0.5238902568817139, "learning_rate": 1.5254236504384345e-06, "loss": 0.0118, "step": 74010 }, { "grad_norm": 0.17793217301368713, "learning_rate": 1.5203614608572726e-06, "loss": 0.013, "step": 74020 }, { "grad_norm": 0.31706249713897705, "learning_rate": 1.5153075551204044e-06, "loss": 0.0097, "step": 74030 }, { "grad_norm": 0.2737722098827362, "learning_rate": 1.5102619340914225e-06, "loss": 0.0191, "step": 74040 }, { "grad_norm": 0.3411445617675781, "learning_rate": 1.505224598632482e-06, "loss": 0.0089, "step": 74050 }, { "grad_norm": 0.14370188117027283, "learning_rate": 1.5001955496043164e-06, "loss": 0.0142, "step": 74060 }, { "grad_norm": 0.3598209321498871, "learning_rate": 1.4951747878662604e-06, "loss": 0.0098, "step": 74070 }, { "grad_norm": 0.11304149031639099, "learning_rate": 1.4901623142762221e-06, "loss": 0.0093, "step": 74080 }, { "grad_norm": 0.16270340979099274, "learning_rate": 1.4851581296907001e-06, "loss": 0.0147, "step": 74090 }, { "grad_norm": 0.27262741327285767, "learning_rate": 1.4801622349647714e-06, "loss": 0.0118, "step": 74100 }, { "grad_norm": 0.32625553011894226, "learning_rate": 1.4751746309520976e-06, "loss": 0.0129, "step": 74110 }, { "grad_norm": 0.23628085851669312, "learning_rate": 1.4701953185049132e-06, "loss": 0.0148, "step": 74120 }, { "grad_norm": 0.13079141080379486, "learning_rate": 1.4652242984740661e-06, "loss": 0.0146, "step": 74130 }, { "grad_norm": 0.17122411727905273, "learning_rate": 1.4602615717089484e-06, "loss": 0.0199, "step": 74140 }, { "grad_norm": 0.2681158185005188, "learning_rate": 1.4553071390575657e-06, "loss": 0.0134, "step": 74150 }, { "grad_norm": 0.22386403381824493, "learning_rate": 1.450361001366496e-06, "loss": 0.017, "step": 74160 }, { "grad_norm": 0.19479432702064514, "learning_rate": 1.4454231594809021e-06, "loss": 0.0127, "step": 74170 }, { "grad_norm": 0.4800432622432709, "learning_rate": 1.4404936142445036e-06, "loss": 0.0143, "step": 74180 }, { "grad_norm": 0.2624521255493164, "learning_rate": 1.4355723664996546e-06, "loss": 0.0101, "step": 74190 }, { "grad_norm": 0.34877732396125793, "learning_rate": 1.4306594170872433e-06, "loss": 0.015, "step": 74200 }, { "grad_norm": 0.23058395087718964, "learning_rate": 1.4257547668467598e-06, "loss": 0.0156, "step": 74210 }, { "grad_norm": 0.21492110192775726, "learning_rate": 1.4208584166162886e-06, "loss": 0.0096, "step": 74220 }, { "grad_norm": 0.24000392854213715, "learning_rate": 1.4159703672324553e-06, "loss": 0.0124, "step": 74230 }, { "grad_norm": 0.2296416461467743, "learning_rate": 1.4110906195305252e-06, "loss": 0.0128, "step": 74240 }, { "grad_norm": 0.16580621898174286, "learning_rate": 1.406219174344292e-06, "loss": 0.0096, "step": 74250 }, { "grad_norm": 0.18615806102752686, "learning_rate": 1.401356032506157e-06, "loss": 0.0112, "step": 74260 }, { "grad_norm": 0.17162823677062988, "learning_rate": 1.3965011948471051e-06, "loss": 0.0143, "step": 74270 }, { "grad_norm": 0.2687269449234009, "learning_rate": 1.3916546621966842e-06, "loss": 0.0094, "step": 74280 }, { "grad_norm": 0.1429235190153122, "learning_rate": 1.3868164353830482e-06, "loss": 0.009, "step": 74290 }, { "grad_norm": 0.23573057353496552, "learning_rate": 1.3819865152329082e-06, "loss": 0.0101, "step": 74300 }, { "grad_norm": 0.19210459291934967, "learning_rate": 1.377164902571565e-06, "loss": 0.0129, "step": 74310 }, { "grad_norm": 0.883461058139801, "learning_rate": 1.372351598222904e-06, "loss": 0.0115, "step": 74320 }, { "grad_norm": 0.17499710619449615, "learning_rate": 1.3675466030093843e-06, "loss": 0.0128, "step": 74330 }, { "grad_norm": 0.504421591758728, "learning_rate": 1.3627499177520486e-06, "loss": 0.0133, "step": 74340 }, { "grad_norm": 0.18979056179523468, "learning_rate": 1.3579615432705196e-06, "loss": 0.0177, "step": 74350 }, { "grad_norm": 0.16991101205348969, "learning_rate": 1.3531814803829978e-06, "loss": 0.0118, "step": 74360 }, { "grad_norm": 0.1781783103942871, "learning_rate": 1.3484097299062803e-06, "loss": 0.0121, "step": 74370 }, { "grad_norm": 0.24408729374408722, "learning_rate": 1.343646292655698e-06, "loss": 0.013, "step": 74380 }, { "grad_norm": 0.21648316085338593, "learning_rate": 1.3388911694452277e-06, "loss": 0.0106, "step": 74390 }, { "grad_norm": 0.36061492562294006, "learning_rate": 1.3341443610873638e-06, "loss": 0.0141, "step": 74400 }, { "grad_norm": 0.22724859416484833, "learning_rate": 1.329405868393213e-06, "loss": 0.0144, "step": 74410 }, { "grad_norm": 0.218338742852211, "learning_rate": 1.3246756921724613e-06, "loss": 0.0133, "step": 74420 }, { "grad_norm": 0.22134540975093842, "learning_rate": 1.3199538332333506e-06, "loss": 0.014, "step": 74430 }, { "grad_norm": 0.11617959290742874, "learning_rate": 1.3152402923827411e-06, "loss": 0.013, "step": 74440 }, { "grad_norm": 0.19249197840690613, "learning_rate": 1.3105350704260277e-06, "loss": 0.0093, "step": 74450 }, { "grad_norm": 0.16204319894313812, "learning_rate": 1.305838168167206e-06, "loss": 0.0108, "step": 74460 }, { "grad_norm": 0.1123889908194542, "learning_rate": 1.301149586408862e-06, "loss": 0.0113, "step": 74470 }, { "grad_norm": 0.13417968153953552, "learning_rate": 1.2964693259521322e-06, "loss": 0.0149, "step": 74480 }, { "grad_norm": 0.19746319949626923, "learning_rate": 1.2917973875967548e-06, "loss": 0.01, "step": 74490 }, { "grad_norm": 0.42734983563423157, "learning_rate": 1.2871337721410249e-06, "loss": 0.0178, "step": 74500 }, { "grad_norm": 0.2410673052072525, "learning_rate": 1.2824784803818379e-06, "loss": 0.0144, "step": 74510 }, { "grad_norm": 0.2499757707118988, "learning_rate": 1.2778315131146467e-06, "loss": 0.0101, "step": 74520 }, { "grad_norm": 0.2609564960002899, "learning_rate": 1.2731928711334994e-06, "loss": 0.01, "step": 74530 }, { "grad_norm": 0.15943242609500885, "learning_rate": 1.2685625552310122e-06, "loss": 0.0093, "step": 74540 }, { "grad_norm": 0.18517085909843445, "learning_rate": 1.263940566198374e-06, "loss": 0.0109, "step": 74550 }, { "grad_norm": 0.10580436885356903, "learning_rate": 1.259326904825353e-06, "loss": 0.0112, "step": 74560 }, { "grad_norm": 0.24560391902923584, "learning_rate": 1.2547215719003137e-06, "loss": 0.0145, "step": 74570 }, { "grad_norm": 0.26918748021125793, "learning_rate": 1.2501245682101703e-06, "loss": 0.0136, "step": 74580 }, { "grad_norm": 1.1682652235031128, "learning_rate": 1.2455358945404171e-06, "loss": 0.0135, "step": 74590 }, { "grad_norm": 0.1549265831708908, "learning_rate": 1.2409555516751493e-06, "loss": 0.0121, "step": 74600 }, { "grad_norm": 0.252274751663208, "learning_rate": 1.2363835403970125e-06, "loss": 0.011, "step": 74610 }, { "grad_norm": 0.20015384256839752, "learning_rate": 1.2318198614872489e-06, "loss": 0.0112, "step": 74620 }, { "grad_norm": 0.22708174586296082, "learning_rate": 1.2272645157256457e-06, "loss": 0.0107, "step": 74630 }, { "grad_norm": 0.20657947659492493, "learning_rate": 1.222717503890608e-06, "loss": 0.015, "step": 74640 }, { "grad_norm": 0.4102552831172943, "learning_rate": 1.218178826759081e-06, "loss": 0.0174, "step": 74650 }, { "grad_norm": 0.3393413722515106, "learning_rate": 1.213648485106611e-06, "loss": 0.0141, "step": 74660 }, { "grad_norm": 0.6132303476333618, "learning_rate": 1.2091264797073066e-06, "loss": 0.018, "step": 74670 }, { "grad_norm": 0.2600077688694, "learning_rate": 1.2046128113338494e-06, "loss": 0.0228, "step": 74680 }, { "grad_norm": 0.27847862243652344, "learning_rate": 1.200107480757512e-06, "loss": 0.0146, "step": 74690 }, { "grad_norm": 0.14134091138839722, "learning_rate": 1.1956104887481168e-06, "loss": 0.0125, "step": 74700 }, { "grad_norm": 0.1663334220647812, "learning_rate": 1.191121836074094e-06, "loss": 0.0093, "step": 74710 }, { "grad_norm": 0.3965494930744171, "learning_rate": 1.1866415235024186e-06, "loss": 0.0086, "step": 74720 }, { "grad_norm": 0.1873675286769867, "learning_rate": 1.182169551798662e-06, "loss": 0.015, "step": 74730 }, { "grad_norm": 0.23843415081501007, "learning_rate": 1.177705921726957e-06, "loss": 0.0103, "step": 74740 }, { "grad_norm": 0.16705864667892456, "learning_rate": 1.173250634050016e-06, "loss": 0.0132, "step": 74750 }, { "grad_norm": 0.16757483780384064, "learning_rate": 1.1688036895291354e-06, "loss": 0.0099, "step": 74760 }, { "grad_norm": 0.21332775056362152, "learning_rate": 1.1643650889241574e-06, "loss": 0.0124, "step": 74770 }, { "grad_norm": 0.16831418871879578, "learning_rate": 1.1599348329935311e-06, "loss": 0.0116, "step": 74780 }, { "grad_norm": 0.19767625629901886, "learning_rate": 1.1555129224942673e-06, "loss": 0.0196, "step": 74790 }, { "grad_norm": 0.19779294729232788, "learning_rate": 1.1510993581819396e-06, "loss": 0.0119, "step": 74800 }, { "grad_norm": 0.22705799341201782, "learning_rate": 1.1466941408107112e-06, "loss": 0.0119, "step": 74810 }, { "grad_norm": 0.1282455325126648, "learning_rate": 1.1422972711333247e-06, "loss": 0.0107, "step": 74820 }, { "grad_norm": 0.18242375552654266, "learning_rate": 1.1379087499010565e-06, "loss": 0.0118, "step": 74830 }, { "grad_norm": 0.1731024533510208, "learning_rate": 1.1335285778638127e-06, "loss": 0.0092, "step": 74840 }, { "grad_norm": 0.32123416662216187, "learning_rate": 1.1291567557700333e-06, "loss": 0.012, "step": 74850 }, { "grad_norm": 0.23802943527698517, "learning_rate": 1.124793284366743e-06, "loss": 0.0107, "step": 74860 }, { "grad_norm": 0.2794607877731323, "learning_rate": 1.1204381643995455e-06, "loss": 0.0104, "step": 74870 }, { "grad_norm": 0.2414575070142746, "learning_rate": 1.116091396612595e-06, "loss": 0.0127, "step": 74880 }, { "grad_norm": 0.1800323873758316, "learning_rate": 1.1117529817486538e-06, "loss": 0.0086, "step": 74890 }, { "grad_norm": 0.17976945638656616, "learning_rate": 1.107422920549034e-06, "loss": 0.0128, "step": 74900 }, { "grad_norm": 0.31420695781707764, "learning_rate": 1.1031012137536157e-06, "loss": 0.0152, "step": 74910 }, { "grad_norm": 0.22600847482681274, "learning_rate": 1.0987878621008695e-06, "loss": 0.0103, "step": 74920 }, { "grad_norm": 0.32953354716300964, "learning_rate": 1.0944828663278284e-06, "loss": 0.0137, "step": 74930 }, { "grad_norm": 0.2187875658273697, "learning_rate": 1.0901862271700925e-06, "loss": 0.0142, "step": 74940 }, { "grad_norm": 0.21193927526474, "learning_rate": 1.0858979453618467e-06, "loss": 0.0153, "step": 74950 }, { "grad_norm": 0.36588042974472046, "learning_rate": 1.0816180216358441e-06, "loss": 0.0186, "step": 74960 }, { "grad_norm": 0.24829575419425964, "learning_rate": 1.0773464567233937e-06, "loss": 0.011, "step": 74970 }, { "grad_norm": 0.2279224991798401, "learning_rate": 1.0730832513543953e-06, "loss": 0.0115, "step": 74980 }, { "grad_norm": 0.1196766346693039, "learning_rate": 1.0688284062573218e-06, "loss": 0.0111, "step": 74990 }, { "grad_norm": 0.3604983687400818, "learning_rate": 1.0645819221591969e-06, "loss": 0.0102, "step": 75000 }, { "grad_norm": 0.22820121049880981, "learning_rate": 1.0603437997856347e-06, "loss": 0.0103, "step": 75010 }, { "grad_norm": 0.11846846342086792, "learning_rate": 1.0561140398608228e-06, "loss": 0.0102, "step": 75020 }, { "grad_norm": 0.10685727000236511, "learning_rate": 1.0518926431074937e-06, "loss": 0.0088, "step": 75030 }, { "grad_norm": 0.18774160742759705, "learning_rate": 1.0476796102469877e-06, "loss": 0.0125, "step": 75040 }, { "grad_norm": 0.2502007484436035, "learning_rate": 1.0434749419991786e-06, "loss": 0.0133, "step": 75050 }, { "grad_norm": 0.1629611998796463, "learning_rate": 1.0392786390825415e-06, "loss": 0.0103, "step": 75060 }, { "grad_norm": 0.3438652753829956, "learning_rate": 1.0350907022141087e-06, "loss": 0.0135, "step": 75070 }, { "grad_norm": 0.3420163691043854, "learning_rate": 1.0309111321094744e-06, "loss": 0.0143, "step": 75080 }, { "grad_norm": 0.19541282951831818, "learning_rate": 1.026739929482834e-06, "loss": 0.0102, "step": 75090 }, { "grad_norm": 0.3108680546283722, "learning_rate": 1.022577095046906e-06, "loss": 0.0162, "step": 75100 }, { "grad_norm": 0.26926475763320923, "learning_rate": 1.0184226295130217e-06, "loss": 0.0126, "step": 75110 }, { "grad_norm": 0.2806774079799652, "learning_rate": 1.0142765335910575e-06, "loss": 0.0133, "step": 75120 }, { "grad_norm": 0.42483705282211304, "learning_rate": 1.0101388079894746e-06, "loss": 0.013, "step": 75130 }, { "grad_norm": 0.12401572614908218, "learning_rate": 1.0060094534152908e-06, "loss": 0.0103, "step": 75140 }, { "grad_norm": 0.1409725397825241, "learning_rate": 1.0018884705741028e-06, "loss": 0.0112, "step": 75150 }, { "grad_norm": 0.2662583291530609, "learning_rate": 9.977758601700805e-07, "loss": 0.0155, "step": 75160 }, { "grad_norm": 0.23372596502304077, "learning_rate": 9.93671622905945e-07, "loss": 0.0178, "step": 75170 }, { "grad_norm": 0.16729478538036346, "learning_rate": 9.895757594830024e-07, "loss": 0.0109, "step": 75180 }, { "grad_norm": 0.09821132570505142, "learning_rate": 9.854882706011204e-07, "loss": 0.0117, "step": 75190 }, { "grad_norm": 0.16644379496574402, "learning_rate": 9.814091569587513e-07, "loss": 0.011, "step": 75200 }, { "grad_norm": 0.41191932559013367, "learning_rate": 9.77338419252888e-07, "loss": 0.0138, "step": 75210 }, { "grad_norm": 0.24078978598117828, "learning_rate": 9.732760581791289e-07, "loss": 0.0115, "step": 75220 }, { "grad_norm": 0.19900760054588318, "learning_rate": 9.692220744315972e-07, "loss": 0.0146, "step": 75230 }, { "grad_norm": 0.19237558543682098, "learning_rate": 9.651764687030162e-07, "loss": 0.0091, "step": 75240 }, { "grad_norm": 0.19848956167697906, "learning_rate": 9.611392416846776e-07, "loss": 0.0136, "step": 75250 }, { "grad_norm": 0.1966714859008789, "learning_rate": 9.571103940664238e-07, "loss": 0.0148, "step": 75260 }, { "grad_norm": 0.10926397889852524, "learning_rate": 9.530899265366822e-07, "loss": 0.0126, "step": 75270 }, { "grad_norm": 0.2173493504524231, "learning_rate": 9.490778397824307e-07, "loss": 0.0082, "step": 75280 }, { "grad_norm": 0.3515896499156952, "learning_rate": 9.450741344892322e-07, "loss": 0.0125, "step": 75290 }, { "grad_norm": 0.31628894805908203, "learning_rate": 9.410788113412116e-07, "loss": 0.0116, "step": 75300 }, { "grad_norm": 0.26128140091896057, "learning_rate": 9.370918710210563e-07, "loss": 0.0118, "step": 75310 }, { "grad_norm": 0.28547438979148865, "learning_rate": 9.331133142100268e-07, "loss": 0.0126, "step": 75320 }, { "grad_norm": 0.20347417891025543, "learning_rate": 9.291431415879459e-07, "loss": 0.0102, "step": 75330 }, { "grad_norm": 0.16007280349731445, "learning_rate": 9.251813538332155e-07, "loss": 0.014, "step": 75340 }, { "grad_norm": 0.18881437182426453, "learning_rate": 9.212279516227884e-07, "loss": 0.0125, "step": 75350 }, { "grad_norm": 0.12737560272216797, "learning_rate": 9.172829356322021e-07, "loss": 0.0098, "step": 75360 }, { "grad_norm": 0.11189355701208115, "learning_rate": 9.133463065355452e-07, "loss": 0.0083, "step": 75370 }, { "grad_norm": 0.3240272104740143, "learning_rate": 9.094180650054796e-07, "loss": 0.0107, "step": 75380 }, { "grad_norm": 0.39925047755241394, "learning_rate": 9.054982117132404e-07, "loss": 0.0128, "step": 75390 }, { "grad_norm": 0.2633154094219208, "learning_rate": 9.015867473286143e-07, "loss": 0.0114, "step": 75400 }, { "grad_norm": 0.330078125, "learning_rate": 8.976836725199778e-07, "loss": 0.0103, "step": 75410 }, { "grad_norm": 0.29415401816368103, "learning_rate": 8.937889879542416e-07, "loss": 0.0103, "step": 75420 }, { "grad_norm": 0.18304096162319183, "learning_rate": 8.899026942969068e-07, "loss": 0.0097, "step": 75430 }, { "grad_norm": 0.12668372690677643, "learning_rate": 8.860247922120424e-07, "loss": 0.0142, "step": 75440 }, { "grad_norm": 0.3189423084259033, "learning_rate": 8.821552823622737e-07, "loss": 0.0118, "step": 75450 }, { "grad_norm": 0.10679828375577927, "learning_rate": 8.782941654087884e-07, "loss": 0.0131, "step": 75460 }, { "grad_norm": 0.2727847099304199, "learning_rate": 8.744414420113478e-07, "loss": 0.0117, "step": 75470 }, { "grad_norm": 0.17683474719524384, "learning_rate": 8.705971128282753e-07, "loss": 0.0114, "step": 75480 }, { "grad_norm": 0.23375117778778076, "learning_rate": 8.66761178516473e-07, "loss": 0.0085, "step": 75490 }, { "grad_norm": 0.1806006282567978, "learning_rate": 8.629336397313781e-07, "loss": 0.0107, "step": 75500 }, { "grad_norm": 0.17484155297279358, "learning_rate": 8.591144971270282e-07, "loss": 0.0137, "step": 75510 }, { "grad_norm": 0.265155553817749, "learning_rate": 8.553037513560069e-07, "loss": 0.0157, "step": 75520 }, { "grad_norm": 0.17978829145431519, "learning_rate": 8.515014030694546e-07, "loss": 0.0112, "step": 75530 }, { "grad_norm": 0.1535443514585495, "learning_rate": 8.47707452917107e-07, "loss": 0.0111, "step": 75540 }, { "grad_norm": 0.16649149358272552, "learning_rate": 8.439219015472343e-07, "loss": 0.0127, "step": 75550 }, { "grad_norm": 0.26536548137664795, "learning_rate": 8.401447496066861e-07, "loss": 0.0079, "step": 75560 }, { "grad_norm": 0.21905378997325897, "learning_rate": 8.363759977408792e-07, "loss": 0.0088, "step": 75570 }, { "grad_norm": 0.22451524436473846, "learning_rate": 8.326156465937817e-07, "loss": 0.0128, "step": 75580 }, { "grad_norm": 0.5694459676742554, "learning_rate": 8.28863696807941e-07, "loss": 0.0204, "step": 75590 }, { "grad_norm": 0.36250045895576477, "learning_rate": 8.251201490244609e-07, "loss": 0.0112, "step": 75600 }, { "grad_norm": 0.18816633522510529, "learning_rate": 8.213850038830129e-07, "loss": 0.013, "step": 75610 }, { "grad_norm": 0.18804776668548584, "learning_rate": 8.176582620218309e-07, "loss": 0.0154, "step": 75620 }, { "grad_norm": 0.1896064579486847, "learning_rate": 8.139399240777057e-07, "loss": 0.0108, "step": 75630 }, { "grad_norm": 0.32092952728271484, "learning_rate": 8.102299906860122e-07, "loss": 0.0123, "step": 75640 }, { "grad_norm": 0.37744855880737305, "learning_rate": 8.065284624806657e-07, "loss": 0.014, "step": 75650 }, { "grad_norm": 0.3251887261867523, "learning_rate": 8.028353400941601e-07, "loss": 0.0124, "step": 75660 }, { "grad_norm": 0.09575869888067245, "learning_rate": 7.991506241575575e-07, "loss": 0.0117, "step": 75670 }, { "grad_norm": 0.3550485372543335, "learning_rate": 7.954743153004541e-07, "loss": 0.0181, "step": 75680 }, { "grad_norm": 0.3376442790031433, "learning_rate": 7.918064141510528e-07, "loss": 0.0158, "step": 75690 }, { "grad_norm": 0.17851042747497559, "learning_rate": 7.881469213360859e-07, "loss": 0.0163, "step": 75700 }, { "grad_norm": 0.15156660974025726, "learning_rate": 7.844958374808642e-07, "loss": 0.0097, "step": 75710 }, { "grad_norm": 0.26185473799705505, "learning_rate": 7.808531632092608e-07, "loss": 0.0102, "step": 75720 }, { "grad_norm": 0.5054251551628113, "learning_rate": 7.772188991436946e-07, "loss": 0.0129, "step": 75730 }, { "grad_norm": 0.13384002447128296, "learning_rate": 7.735930459051799e-07, "loss": 0.0109, "step": 75740 }, { "grad_norm": 0.25775840878486633, "learning_rate": 7.699756041132655e-07, "loss": 0.0113, "step": 75750 }, { "grad_norm": 0.1780657172203064, "learning_rate": 7.663665743860793e-07, "loss": 0.0097, "step": 75760 }, { "grad_norm": 0.2633357644081116, "learning_rate": 7.627659573403001e-07, "loss": 0.0119, "step": 75770 }, { "grad_norm": 0.3386204242706299, "learning_rate": 7.591737535911802e-07, "loss": 0.01, "step": 75780 }, { "grad_norm": 0.13569197058677673, "learning_rate": 7.555899637525288e-07, "loss": 0.0099, "step": 75790 }, { "grad_norm": 0.2844915986061096, "learning_rate": 7.520145884367058e-07, "loss": 0.0081, "step": 75800 }, { "grad_norm": 0.19349715113639832, "learning_rate": 7.484476282546615e-07, "loss": 0.011, "step": 75810 }, { "grad_norm": 0.33443722128868103, "learning_rate": 7.448890838158806e-07, "loss": 0.0159, "step": 75820 }, { "grad_norm": 0.33521410822868347, "learning_rate": 7.413389557284267e-07, "loss": 0.0128, "step": 75830 }, { "grad_norm": 0.15753835439682007, "learning_rate": 7.377972445989201e-07, "loss": 0.0114, "step": 75840 }, { "grad_norm": 0.3149203062057495, "learning_rate": 7.342639510325377e-07, "loss": 0.0137, "step": 75850 }, { "grad_norm": 0.37639832496643066, "learning_rate": 7.307390756330246e-07, "loss": 0.01, "step": 75860 }, { "grad_norm": 0.26630958914756775, "learning_rate": 7.272226190026876e-07, "loss": 0.0082, "step": 75870 }, { "grad_norm": 0.29755908250808716, "learning_rate": 7.237145817423907e-07, "loss": 0.0107, "step": 75880 }, { "grad_norm": 0.18643832206726074, "learning_rate": 7.202149644515654e-07, "loss": 0.0109, "step": 75890 }, { "grad_norm": 0.2853076756000519, "learning_rate": 7.167237677281946e-07, "loss": 0.0122, "step": 75900 }, { "grad_norm": 0.2929772734642029, "learning_rate": 7.132409921688288e-07, "loss": 0.0135, "step": 75910 }, { "grad_norm": 0.11163314431905746, "learning_rate": 7.09766638368592e-07, "loss": 0.009, "step": 75920 }, { "grad_norm": 0.24067500233650208, "learning_rate": 7.063007069211313e-07, "loss": 0.0114, "step": 75930 }, { "grad_norm": 0.2775317132472992, "learning_rate": 7.028431984187067e-07, "loss": 0.0115, "step": 75940 }, { "grad_norm": 0.16200540959835052, "learning_rate": 6.993941134520898e-07, "loss": 0.0144, "step": 75950 }, { "grad_norm": 0.2244184911251068, "learning_rate": 6.959534526106537e-07, "loss": 0.0091, "step": 75960 }, { "grad_norm": 0.17773018777370453, "learning_rate": 6.925212164822948e-07, "loss": 0.0082, "step": 75970 }, { "grad_norm": 0.29847702383995056, "learning_rate": 6.890974056535049e-07, "loss": 0.0142, "step": 75980 }, { "grad_norm": 0.28442150354385376, "learning_rate": 6.856820207093106e-07, "loss": 0.0089, "step": 75990 }, { "grad_norm": 0.1766383945941925, "learning_rate": 6.822750622333063e-07, "loss": 0.0101, "step": 76000 }, { "grad_norm": 0.2694431245326996, "learning_rate": 6.788765308076539e-07, "loss": 0.0113, "step": 76010 }, { "grad_norm": 0.4259417653083801, "learning_rate": 6.754864270130668e-07, "loss": 0.0107, "step": 76020 }, { "grad_norm": 0.2273525446653366, "learning_rate": 6.721047514288203e-07, "loss": 0.0123, "step": 76030 }, { "grad_norm": 0.35037246346473694, "learning_rate": 6.687315046327469e-07, "loss": 0.0114, "step": 76040 }, { "grad_norm": 0.4093415439128876, "learning_rate": 6.65366687201252e-07, "loss": 0.0163, "step": 76050 }, { "grad_norm": 0.27409690618515015, "learning_rate": 6.620102997092814e-07, "loss": 0.015, "step": 76060 }, { "grad_norm": 0.15669037401676178, "learning_rate": 6.586623427303596e-07, "loss": 0.0109, "step": 76070 }, { "grad_norm": 0.21618522703647614, "learning_rate": 6.553228168365455e-07, "loss": 0.0163, "step": 76080 }, { "grad_norm": 1.8274401426315308, "learning_rate": 6.519917225984884e-07, "loss": 0.0129, "step": 76090 }, { "grad_norm": 0.23422494530677795, "learning_rate": 6.486690605853718e-07, "loss": 0.012, "step": 76100 }, { "grad_norm": 0.34663036465644836, "learning_rate": 6.453548313649527e-07, "loss": 0.0129, "step": 76110 }, { "grad_norm": 0.3417895436286926, "learning_rate": 6.420490355035446e-07, "loss": 0.013, "step": 76120 }, { "grad_norm": 0.41388753056526184, "learning_rate": 6.387516735660071e-07, "loss": 0.0134, "step": 76130 }, { "grad_norm": 0.21451018750667572, "learning_rate": 6.35462746115778e-07, "loss": 0.0102, "step": 76140 }, { "grad_norm": 0.2685851454734802, "learning_rate": 6.321822537148358e-07, "loss": 0.0149, "step": 76150 }, { "grad_norm": 0.28144872188568115, "learning_rate": 6.289101969237432e-07, "loss": 0.0123, "step": 76160 }, { "grad_norm": 0.22779683768749237, "learning_rate": 6.256465763015918e-07, "loss": 0.0102, "step": 76170 }, { "grad_norm": 0.12507788836956024, "learning_rate": 6.22391392406052e-07, "loss": 0.011, "step": 76180 }, { "grad_norm": 0.16339479386806488, "learning_rate": 6.191446457933403e-07, "loss": 0.0089, "step": 76190 }, { "grad_norm": 0.08418549597263336, "learning_rate": 6.159063370182406e-07, "loss": 0.0086, "step": 76200 }, { "grad_norm": 0.22663402557373047, "learning_rate": 6.126764666340879e-07, "loss": 0.0097, "step": 76210 }, { "grad_norm": 0.12964670360088348, "learning_rate": 6.094550351927852e-07, "loss": 0.0113, "step": 76220 }, { "grad_norm": 0.2994533181190491, "learning_rate": 6.062420432447757e-07, "loss": 0.0127, "step": 76230 }, { "grad_norm": 0.14818690717220306, "learning_rate": 6.030374913390813e-07, "loss": 0.0105, "step": 76240 }, { "grad_norm": 0.3760454058647156, "learning_rate": 5.998413800232694e-07, "loss": 0.0107, "step": 76250 }, { "grad_norm": 0.4345239996910095, "learning_rate": 5.966537098434755e-07, "loss": 0.0104, "step": 76260 }, { "grad_norm": 3.631582021713257, "learning_rate": 5.934744813443694e-07, "loss": 0.0097, "step": 76270 }, { "grad_norm": 0.24509909749031067, "learning_rate": 5.903036950692054e-07, "loss": 0.012, "step": 76280 }, { "grad_norm": 0.1665489822626114, "learning_rate": 5.871413515597835e-07, "loss": 0.0119, "step": 76290 }, { "grad_norm": 0.2930546700954437, "learning_rate": 5.839874513564547e-07, "loss": 0.0084, "step": 76300 }, { "grad_norm": 0.44516730308532715, "learning_rate": 5.808419949981436e-07, "loss": 0.0156, "step": 76310 }, { "grad_norm": 0.17297250032424927, "learning_rate": 5.777049830223257e-07, "loss": 0.0152, "step": 76320 }, { "grad_norm": 0.20889975130558014, "learning_rate": 5.745764159650114e-07, "loss": 0.0127, "step": 76330 }, { "grad_norm": 0.2984192967414856, "learning_rate": 5.71456294360806e-07, "loss": 0.0088, "step": 76340 }, { "grad_norm": 0.2537604570388794, "learning_rate": 5.683446187428443e-07, "loss": 0.009, "step": 76350 }, { "grad_norm": 0.22537478804588318, "learning_rate": 5.652413896428288e-07, "loss": 0.0105, "step": 76360 }, { "grad_norm": 0.28069382905960083, "learning_rate": 5.621466075910131e-07, "loss": 0.0122, "step": 76370 }, { "grad_norm": 0.21834830939769745, "learning_rate": 5.590602731162187e-07, "loss": 0.0094, "step": 76380 }, { "grad_norm": 0.23845340311527252, "learning_rate": 5.559823867458125e-07, "loss": 0.0133, "step": 76390 }, { "grad_norm": 0.18098941445350647, "learning_rate": 5.529129490057128e-07, "loss": 0.0145, "step": 76400 }, { "grad_norm": 0.23907335102558136, "learning_rate": 5.498519604204167e-07, "loss": 0.0141, "step": 76410 }, { "grad_norm": 0.18015432357788086, "learning_rate": 5.467994215129501e-07, "loss": 0.0126, "step": 76420 }, { "grad_norm": 0.1935555338859558, "learning_rate": 5.437553328049183e-07, "loss": 0.0108, "step": 76430 }, { "grad_norm": 0.2225203961133957, "learning_rate": 5.407196948164661e-07, "loss": 0.0127, "step": 76440 }, { "grad_norm": 0.1580640822649002, "learning_rate": 5.376925080663009e-07, "loss": 0.0131, "step": 76450 }, { "grad_norm": 0.5702149868011475, "learning_rate": 5.346737730716977e-07, "loss": 0.0141, "step": 76460 }, { "grad_norm": 0.13823524117469788, "learning_rate": 5.316634903484607e-07, "loss": 0.0097, "step": 76470 }, { "grad_norm": 0.17967160046100616, "learning_rate": 5.286616604109728e-07, "loss": 0.0119, "step": 76480 }, { "grad_norm": 0.22140616178512573, "learning_rate": 5.256682837721627e-07, "loss": 0.017, "step": 76490 }, { "grad_norm": 0.3341751992702484, "learning_rate": 5.226833609435156e-07, "loss": 0.0106, "step": 76500 }, { "grad_norm": 0.1469019651412964, "learning_rate": 5.19706892435079e-07, "loss": 0.0093, "step": 76510 }, { "grad_norm": 0.2874971032142639, "learning_rate": 5.167388787554406e-07, "loss": 0.0121, "step": 76520 }, { "grad_norm": 0.2845056653022766, "learning_rate": 5.137793204117614e-07, "loss": 0.0116, "step": 76530 }, { "grad_norm": 0.37454378604888916, "learning_rate": 5.108282179097479e-07, "loss": 0.0115, "step": 76540 }, { "grad_norm": 0.2546420395374298, "learning_rate": 5.078855717536523e-07, "loss": 0.0126, "step": 76550 }, { "grad_norm": 0.514593780040741, "learning_rate": 5.049513824463059e-07, "loss": 0.0095, "step": 76560 }, { "grad_norm": 0.13145489990711212, "learning_rate": 5.020256504890742e-07, "loss": 0.0116, "step": 76570 }, { "grad_norm": 0.29616260528564453, "learning_rate": 4.991083763818849e-07, "loss": 0.0109, "step": 76580 }, { "grad_norm": 0.3359984755516052, "learning_rate": 4.961995606232228e-07, "loss": 0.014, "step": 76590 }, { "grad_norm": 0.18064914643764496, "learning_rate": 4.932992037101236e-07, "loss": 0.0107, "step": 76600 }, { "grad_norm": 0.21485641598701477, "learning_rate": 4.904073061381798e-07, "loss": 0.0149, "step": 76610 }, { "grad_norm": 0.32043442130088806, "learning_rate": 4.875238684015349e-07, "loss": 0.0143, "step": 76620 }, { "grad_norm": 0.26396745443344116, "learning_rate": 4.846488909928948e-07, "loss": 0.0079, "step": 76630 }, { "grad_norm": 0.12118850648403168, "learning_rate": 4.81782374403511e-07, "loss": 0.0152, "step": 76640 }, { "grad_norm": 0.13218775391578674, "learning_rate": 4.789243191231918e-07, "loss": 0.0079, "step": 76650 }, { "grad_norm": 0.20376814901828766, "learning_rate": 4.76074725640302e-07, "loss": 0.0086, "step": 76660 }, { "grad_norm": 0.23572269082069397, "learning_rate": 4.732335944417632e-07, "loss": 0.0197, "step": 76670 }, { "grad_norm": 0.2208375483751297, "learning_rate": 4.704009260130371e-07, "loss": 0.0151, "step": 76680 }, { "grad_norm": 0.12149504572153091, "learning_rate": 4.675767208381587e-07, "loss": 0.0106, "step": 76690 }, { "grad_norm": 0.435602605342865, "learning_rate": 4.647609793997032e-07, "loss": 0.0125, "step": 76700 }, { "grad_norm": 0.2642024755477905, "learning_rate": 4.619537021788023e-07, "loss": 0.0126, "step": 76710 }, { "grad_norm": 0.16005907952785492, "learning_rate": 4.5915488965515005e-07, "loss": 0.0133, "step": 76720 }, { "grad_norm": 0.17560088634490967, "learning_rate": 4.563645423069807e-07, "loss": 0.0135, "step": 76730 }, { "grad_norm": 0.289975106716156, "learning_rate": 4.535826606110849e-07, "loss": 0.0112, "step": 76740 }, { "grad_norm": 0.29253724217414856, "learning_rate": 4.508092450428214e-07, "loss": 0.0148, "step": 76750 }, { "grad_norm": 0.1394178867340088, "learning_rate": 4.480442960760778e-07, "loss": 0.0134, "step": 76760 }, { "grad_norm": 0.144084632396698, "learning_rate": 4.4528781418332053e-07, "loss": 0.0108, "step": 76770 }, { "grad_norm": 0.41720086336135864, "learning_rate": 4.4253979983555073e-07, "loss": 0.0171, "step": 76780 }, { "grad_norm": 0.13432817161083221, "learning_rate": 4.3980025350233154e-07, "loss": 0.0122, "step": 76790 }, { "grad_norm": 0.15096114575862885, "learning_rate": 4.370691756517664e-07, "loss": 0.0102, "step": 76800 }, { "grad_norm": 0.22000572085380554, "learning_rate": 4.3434656675053753e-07, "loss": 0.0083, "step": 76810 }, { "grad_norm": 0.21117763221263885, "learning_rate": 4.3163242726385613e-07, "loss": 0.0081, "step": 76820 }, { "grad_norm": 0.16517966985702515, "learning_rate": 4.289267576554956e-07, "loss": 0.0133, "step": 76830 }, { "grad_norm": 0.27640169858932495, "learning_rate": 4.262295583877807e-07, "loss": 0.0121, "step": 76840 }, { "grad_norm": 0.2522316575050354, "learning_rate": 4.235408299215815e-07, "loss": 0.0163, "step": 76850 }, { "grad_norm": 0.18149030208587646, "learning_rate": 4.2086057271634173e-07, "loss": 0.0089, "step": 76860 }, { "grad_norm": 0.20609843730926514, "learning_rate": 4.181887872300394e-07, "loss": 0.0118, "step": 76870 }, { "grad_norm": 0.1920863389968872, "learning_rate": 4.155254739192038e-07, "loss": 0.0097, "step": 76880 }, { "grad_norm": 0.18022900819778442, "learning_rate": 4.128706332389265e-07, "loss": 0.0154, "step": 76890 }, { "grad_norm": 0.2642173171043396, "learning_rate": 4.1022426564284453e-07, "loss": 0.01, "step": 76900 }, { "grad_norm": 0.15847879648208618, "learning_rate": 4.075863715831574e-07, "loss": 0.0116, "step": 76910 }, { "grad_norm": 0.3482627272605896, "learning_rate": 4.0495695151059887e-07, "loss": 0.0122, "step": 76920 }, { "grad_norm": 0.2632211446762085, "learning_rate": 4.023360058744763e-07, "loss": 0.0095, "step": 76930 }, { "grad_norm": 0.11987138539552689, "learning_rate": 3.997235351226314e-07, "loss": 0.0085, "step": 76940 }, { "grad_norm": 0.28523489832878113, "learning_rate": 3.971195397014571e-07, "loss": 0.0118, "step": 76950 }, { "grad_norm": 0.25411778688430786, "learning_rate": 3.9452402005591417e-07, "loss": 0.0104, "step": 76960 }, { "grad_norm": 0.21118731796741486, "learning_rate": 3.9193697662950334e-07, "loss": 0.0115, "step": 76970 }, { "grad_norm": 0.39585959911346436, "learning_rate": 3.893584098642822e-07, "loss": 0.0127, "step": 76980 }, { "grad_norm": 0.09624014049768448, "learning_rate": 3.867883202008538e-07, "loss": 0.0135, "step": 76990 }, { "grad_norm": 0.18185487389564514, "learning_rate": 3.842267080783779e-07, "loss": 0.0126, "step": 77000 }, { "grad_norm": 0.21957683563232422, "learning_rate": 3.816735739345656e-07, "loss": 0.0119, "step": 77010 }, { "grad_norm": 0.2419295758008957, "learning_rate": 3.791289182056679e-07, "loss": 0.0119, "step": 77020 }, { "grad_norm": 0.27031463384628296, "learning_rate": 3.7659274132650913e-07, "loss": 0.0121, "step": 77030 }, { "grad_norm": 0.24784329533576965, "learning_rate": 3.7406504373044826e-07, "loss": 0.0116, "step": 77040 }, { "grad_norm": 0.28248587250709534, "learning_rate": 3.7154582584939533e-07, "loss": 0.0136, "step": 77050 }, { "grad_norm": 0.2705613374710083, "learning_rate": 3.690350881138227e-07, "loss": 0.0124, "step": 77060 }, { "grad_norm": 0.24692599475383759, "learning_rate": 3.665328309527427e-07, "loss": 0.0108, "step": 77070 }, { "grad_norm": 0.22076375782489777, "learning_rate": 3.640390547937245e-07, "loss": 0.0099, "step": 77080 }, { "grad_norm": 0.14733286201953888, "learning_rate": 3.615537600628882e-07, "loss": 0.0129, "step": 77090 }, { "grad_norm": 0.18654946982860565, "learning_rate": 3.5907694718489427e-07, "loss": 0.012, "step": 77100 }, { "grad_norm": 0.21111218631267548, "learning_rate": 3.566086165829707e-07, "loss": 0.0172, "step": 77110 }, { "grad_norm": 0.4255448281764984, "learning_rate": 3.5414876867888024e-07, "loss": 0.015, "step": 77120 }, { "grad_norm": 0.17897728085517883, "learning_rate": 3.5169740389295326e-07, "loss": 0.0132, "step": 77130 }, { "grad_norm": 0.263163298368454, "learning_rate": 3.4925452264405466e-07, "loss": 0.0125, "step": 77140 }, { "grad_norm": 0.20597074925899506, "learning_rate": 3.468201253496062e-07, "loss": 0.0099, "step": 77150 }, { "grad_norm": 0.2224951684474945, "learning_rate": 3.443942124255861e-07, "loss": 0.0116, "step": 77160 }, { "grad_norm": 0.21883949637413025, "learning_rate": 3.4197678428650183e-07, "loss": 0.0149, "step": 77170 }, { "grad_norm": 0.14818352460861206, "learning_rate": 3.3956784134544504e-07, "loss": 0.0075, "step": 77180 }, { "grad_norm": 0.24261833727359772, "learning_rate": 3.3716738401402547e-07, "loss": 0.0137, "step": 77190 }, { "grad_norm": 0.1611536741256714, "learning_rate": 3.3477541270241495e-07, "loss": 0.0087, "step": 77200 }, { "grad_norm": 0.2032715082168579, "learning_rate": 3.3239192781934215e-07, "loss": 0.0107, "step": 77210 }, { "grad_norm": 0.5335571765899658, "learning_rate": 3.3001692977207563e-07, "loss": 0.0114, "step": 77220 }, { "grad_norm": 0.1754211187362671, "learning_rate": 3.276504189664409e-07, "loss": 0.0127, "step": 77230 }, { "grad_norm": 0.19730888307094574, "learning_rate": 3.252923958068088e-07, "loss": 0.0112, "step": 77240 }, { "grad_norm": 0.21379606425762177, "learning_rate": 3.2294286069609046e-07, "loss": 0.0104, "step": 77250 }, { "grad_norm": 0.14931604266166687, "learning_rate": 3.2060181403577583e-07, "loss": 0.0109, "step": 77260 }, { "grad_norm": 0.21891877055168152, "learning_rate": 3.1826925622587823e-07, "loss": 0.0093, "step": 77270 }, { "grad_norm": 0.15243074297904968, "learning_rate": 3.159451876649622e-07, "loss": 0.0119, "step": 77280 }, { "grad_norm": 0.25560352206230164, "learning_rate": 3.1362960875015444e-07, "loss": 0.0104, "step": 77290 }, { "grad_norm": 0.37143251299858093, "learning_rate": 3.1132251987711637e-07, "loss": 0.0139, "step": 77300 }, { "grad_norm": 0.19145745038986206, "learning_rate": 3.0902392144007695e-07, "loss": 0.0121, "step": 77310 }, { "grad_norm": 0.18185539543628693, "learning_rate": 3.0673381383179436e-07, "loss": 0.0125, "step": 77320 }, { "grad_norm": 0.3677922487258911, "learning_rate": 3.0445219744358875e-07, "loss": 0.0088, "step": 77330 }, { "grad_norm": 0.3294815123081207, "learning_rate": 3.021790726653262e-07, "loss": 0.0127, "step": 77340 }, { "grad_norm": 0.36556199193000793, "learning_rate": 2.9991443988542366e-07, "loss": 0.0103, "step": 77350 }, { "grad_norm": 0.12155543267726898, "learning_rate": 2.976582994908439e-07, "loss": 0.0118, "step": 77360 }, { "grad_norm": 0.18399372696876526, "learning_rate": 2.954106518671007e-07, "loss": 0.0103, "step": 77370 }, { "grad_norm": 0.2771947681903839, "learning_rate": 2.9317149739825356e-07, "loss": 0.0092, "step": 77380 }, { "grad_norm": 0.5053636431694031, "learning_rate": 2.909408364669075e-07, "loss": 0.0119, "step": 77390 }, { "grad_norm": 0.3286198675632477, "learning_rate": 2.887186694542299e-07, "loss": 0.0156, "step": 77400 }, { "grad_norm": 0.318441241979599, "learning_rate": 2.865049967399225e-07, "loss": 0.0123, "step": 77410 }, { "grad_norm": 0.2121318131685257, "learning_rate": 2.842998187022439e-07, "loss": 0.0112, "step": 77420 }, { "grad_norm": 0.2039528489112854, "learning_rate": 2.8210313571800374e-07, "loss": 0.013, "step": 77430 }, { "grad_norm": 0.3764019310474396, "learning_rate": 2.7991494816255184e-07, "loss": 0.0101, "step": 77440 }, { "grad_norm": 0.1768513321876526, "learning_rate": 2.777352564097779e-07, "loss": 0.0125, "step": 77450 }, { "grad_norm": 0.17271336913108826, "learning_rate": 2.755640608321508e-07, "loss": 0.0112, "step": 77460 }, { "grad_norm": 0.36869940161705017, "learning_rate": 2.73401361800657e-07, "loss": 0.0115, "step": 77470 }, { "grad_norm": 0.18043172359466553, "learning_rate": 2.7124715968484537e-07, "loss": 0.0108, "step": 77480 }, { "grad_norm": 0.2577446401119232, "learning_rate": 2.691014548528104e-07, "loss": 0.0128, "step": 77490 }, { "grad_norm": 0.17643137276172638, "learning_rate": 2.669642476711864e-07, "loss": 0.01, "step": 77500 }, { "grad_norm": 0.16047321259975433, "learning_rate": 2.648355385051815e-07, "loss": 0.012, "step": 77510 }, { "grad_norm": 0.24117986857891083, "learning_rate": 2.627153277185157e-07, "loss": 0.0097, "step": 77520 }, { "grad_norm": 0.2335115671157837, "learning_rate": 2.606036156734881e-07, "loss": 0.0079, "step": 77530 }, { "grad_norm": 0.17663228511810303, "learning_rate": 2.5850040273092127e-07, "loss": 0.0105, "step": 77540 }, { "grad_norm": 0.21435770392417908, "learning_rate": 2.5640568925020536e-07, "loss": 0.0177, "step": 77550 }, { "grad_norm": 0.22214917838573456, "learning_rate": 2.5431947558927083e-07, "loss": 0.0153, "step": 77560 }, { "grad_norm": 0.2512970566749573, "learning_rate": 2.522417621045825e-07, "loss": 0.0129, "step": 77570 }, { "grad_norm": 0.19356386363506317, "learning_rate": 2.501725491511786e-07, "loss": 0.0135, "step": 77580 }, { "grad_norm": 0.11953205615282059, "learning_rate": 2.481118370826263e-07, "loss": 0.0118, "step": 77590 }, { "grad_norm": 0.26061174273490906, "learning_rate": 2.4605962625104393e-07, "loss": 0.0095, "step": 77600 }, { "grad_norm": 0.26658040285110474, "learning_rate": 2.440159170070955e-07, "loss": 0.0182, "step": 77610 }, { "grad_norm": 0.3269061744213104, "learning_rate": 2.41980709699996e-07, "loss": 0.0109, "step": 77620 }, { "grad_norm": 0.4363299608230591, "learning_rate": 2.3995400467751174e-07, "loss": 0.0113, "step": 77630 }, { "grad_norm": 0.23232071101665497, "learning_rate": 2.3793580228594902e-07, "loss": 0.0188, "step": 77640 }, { "grad_norm": 0.21410341560840607, "learning_rate": 2.3592610287015982e-07, "loss": 0.0088, "step": 77650 }, { "grad_norm": 0.15700674057006836, "learning_rate": 2.3392490677354718e-07, "loss": 0.0125, "step": 77660 }, { "grad_norm": 0.19763053953647614, "learning_rate": 2.3193221433806533e-07, "loss": 0.0078, "step": 77670 }, { "grad_norm": 0.3950352966785431, "learning_rate": 2.299480259042086e-07, "loss": 0.0147, "step": 77680 }, { "grad_norm": 0.07201764732599258, "learning_rate": 2.2797234181102244e-07, "loss": 0.0096, "step": 77690 }, { "grad_norm": 0.24599574506282806, "learning_rate": 2.2600516239609238e-07, "loss": 0.0115, "step": 77700 }, { "grad_norm": 0.16878432035446167, "learning_rate": 2.240464879955606e-07, "loss": 0.0129, "step": 77710 }, { "grad_norm": 0.1874052882194519, "learning_rate": 2.2209631894410387e-07, "loss": 0.0118, "step": 77720 }, { "grad_norm": 0.23794959485530853, "learning_rate": 2.2015465557496117e-07, "loss": 0.01, "step": 77730 }, { "grad_norm": 0.19376859068870544, "learning_rate": 2.1822149821990602e-07, "loss": 0.01, "step": 77740 }, { "grad_norm": 0.4491516351699829, "learning_rate": 2.1629684720926303e-07, "loss": 0.0123, "step": 77750 }, { "grad_norm": 0.19311393797397614, "learning_rate": 2.1438070287189693e-07, "loss": 0.012, "step": 77760 }, { "grad_norm": 0.18226636946201324, "learning_rate": 2.1247306553523472e-07, "loss": 0.0157, "step": 77770 }, { "grad_norm": 0.6683747172355652, "learning_rate": 2.105739355252323e-07, "loss": 0.0133, "step": 77780 }, { "grad_norm": 0.13501864671707153, "learning_rate": 2.0868331316639678e-07, "loss": 0.0122, "step": 77790 }, { "grad_norm": 0.11664067953824997, "learning_rate": 2.0680119878179193e-07, "loss": 0.0108, "step": 77800 }, { "grad_norm": 0.19210371375083923, "learning_rate": 2.0492759269301054e-07, "loss": 0.0123, "step": 77810 }, { "grad_norm": 0.4032769501209259, "learning_rate": 2.0306249522021315e-07, "loss": 0.0137, "step": 77820 }, { "grad_norm": 0.17248201370239258, "learning_rate": 2.0120590668207816e-07, "loss": 0.0079, "step": 77830 }, { "grad_norm": 0.20369653403759003, "learning_rate": 1.993578273958574e-07, "loss": 0.009, "step": 77840 }, { "grad_norm": 0.19380885362625122, "learning_rate": 1.975182576773371e-07, "loss": 0.0109, "step": 77850 }, { "grad_norm": 0.421114444732666, "learning_rate": 1.9568719784083812e-07, "loss": 0.0126, "step": 77860 }, { "grad_norm": 0.1612517088651657, "learning_rate": 1.9386464819924898e-07, "loss": 0.0153, "step": 77870 }, { "grad_norm": 0.17089740931987762, "learning_rate": 1.9205060906399285e-07, "loss": 0.0161, "step": 77880 }, { "grad_norm": 0.3206695020198822, "learning_rate": 1.9024508074503845e-07, "loss": 0.0118, "step": 77890 }, { "grad_norm": 0.21841613948345184, "learning_rate": 1.8844806355089452e-07, "loss": 0.0177, "step": 77900 }, { "grad_norm": 0.17816178500652313, "learning_rate": 1.8665955778863208e-07, "loss": 0.0213, "step": 77910 }, { "grad_norm": 0.18875621259212494, "learning_rate": 1.848795637638512e-07, "loss": 0.0124, "step": 77920 }, { "grad_norm": 0.302034854888916, "learning_rate": 1.831080817807085e-07, "loss": 0.0138, "step": 77930 }, { "grad_norm": 0.18944431841373444, "learning_rate": 1.813451121418952e-07, "loss": 0.018, "step": 77940 }, { "grad_norm": 0.19033727049827576, "learning_rate": 1.795906551486648e-07, "loss": 0.009, "step": 77950 }, { "grad_norm": 0.18288905918598175, "learning_rate": 1.7784471110079414e-07, "loss": 0.0102, "step": 77960 }, { "grad_norm": 0.4406304657459259, "learning_rate": 1.7610728029662793e-07, "loss": 0.0143, "step": 77970 }, { "grad_norm": 0.11378033459186554, "learning_rate": 1.7437836303303979e-07, "loss": 0.0121, "step": 77980 }, { "grad_norm": 0.17576657235622406, "learning_rate": 1.726579596054545e-07, "loss": 0.01, "step": 77990 }, { "grad_norm": 0.22794823348522186, "learning_rate": 1.7094607030784803e-07, "loss": 0.0127, "step": 78000 }, { "grad_norm": 0.40344125032424927, "learning_rate": 1.6924269543272532e-07, "loss": 0.0139, "step": 78010 }, { "grad_norm": 0.24333840608596802, "learning_rate": 1.6754783527115348e-07, "loss": 0.0108, "step": 78020 }, { "grad_norm": 0.2371966689825058, "learning_rate": 1.6586149011273422e-07, "loss": 0.0109, "step": 78030 }, { "grad_norm": 0.268208771944046, "learning_rate": 1.6418366024562038e-07, "loss": 0.0193, "step": 78040 }, { "grad_norm": 0.24849286675453186, "learning_rate": 1.6251434595651037e-07, "loss": 0.0107, "step": 78050 }, { "grad_norm": 0.14188529551029205, "learning_rate": 1.608535475306372e-07, "loss": 0.009, "step": 78060 }, { "grad_norm": 0.3450813293457031, "learning_rate": 1.5920126525179048e-07, "loss": 0.0104, "step": 78070 }, { "grad_norm": 0.23711352050304413, "learning_rate": 1.5755749940229435e-07, "loss": 0.0106, "step": 78080 }, { "grad_norm": 0.2566092312335968, "learning_rate": 1.559222502630353e-07, "loss": 0.0122, "step": 78090 }, { "grad_norm": 0.39776673913002014, "learning_rate": 1.5429551811341757e-07, "loss": 0.0118, "step": 78100 }, { "grad_norm": 0.21759240329265594, "learning_rate": 1.5267730323141882e-07, "loss": 0.0141, "step": 78110 }, { "grad_norm": 0.17651624977588654, "learning_rate": 1.5106760589353454e-07, "loss": 0.0082, "step": 78120 }, { "grad_norm": 0.19030249118804932, "learning_rate": 1.4946642637483355e-07, "loss": 0.0092, "step": 78130 }, { "grad_norm": 0.22981253266334534, "learning_rate": 1.4787376494889703e-07, "loss": 0.0092, "step": 78140 }, { "grad_norm": 0.257168710231781, "learning_rate": 1.4628962188787955e-07, "loss": 0.0114, "step": 78150 }, { "grad_norm": 0.26076650619506836, "learning_rate": 1.4471399746247006e-07, "loss": 0.0143, "step": 78160 }, { "grad_norm": 0.18849177658557892, "learning_rate": 1.4314689194188103e-07, "loss": 0.0106, "step": 78170 }, { "grad_norm": 0.38752710819244385, "learning_rate": 1.4158830559390933e-07, "loss": 0.0163, "step": 78180 }, { "grad_norm": 0.21543283760547638, "learning_rate": 1.4003823868486422e-07, "loss": 0.0249, "step": 78190 }, { "grad_norm": 0.27033352851867676, "learning_rate": 1.3849669147960598e-07, "loss": 0.0128, "step": 78200 }, { "grad_norm": 0.11627281457185745, "learning_rate": 1.3696366424155726e-07, "loss": 0.0092, "step": 78210 }, { "grad_norm": 0.610464870929718, "learning_rate": 1.35439157232653e-07, "loss": 0.0162, "step": 78220 }, { "grad_norm": 0.24297362565994263, "learning_rate": 1.3392317071340144e-07, "loss": 0.0091, "step": 78230 }, { "grad_norm": 0.4235554337501526, "learning_rate": 1.3241570494283984e-07, "loss": 0.0122, "step": 78240 }, { "grad_norm": 0.35032007098197937, "learning_rate": 1.3091676017855103e-07, "loss": 0.0149, "step": 78250 }, { "grad_norm": 0.16339968144893646, "learning_rate": 1.2942633667666904e-07, "loss": 0.0128, "step": 78260 }, { "grad_norm": 0.18767458200454712, "learning_rate": 1.2794443469185679e-07, "loss": 0.0159, "step": 78270 }, { "grad_norm": 0.1964983195066452, "learning_rate": 1.2647105447734508e-07, "loss": 0.0109, "step": 78280 }, { "grad_norm": 0.15139204263687134, "learning_rate": 1.2500619628488254e-07, "loss": 0.0129, "step": 78290 }, { "grad_norm": 0.2059711068868637, "learning_rate": 1.2354986036477446e-07, "loss": 0.0151, "step": 78300 }, { "grad_norm": 0.15285828709602356, "learning_rate": 1.221020469658718e-07, "loss": 0.0104, "step": 78310 }, { "grad_norm": 0.2775760591030121, "learning_rate": 1.2066275633556556e-07, "loss": 0.0124, "step": 78320 }, { "grad_norm": 0.2084350287914276, "learning_rate": 1.192319887197979e-07, "loss": 0.0099, "step": 78330 }, { "grad_norm": 0.17961004376411438, "learning_rate": 1.1780974436303438e-07, "loss": 0.0091, "step": 78340 }, { "grad_norm": 0.13504058122634888, "learning_rate": 1.1639602350830836e-07, "loss": 0.0119, "step": 78350 }, { "grad_norm": 0.21673089265823364, "learning_rate": 1.1499082639718217e-07, "loss": 0.0122, "step": 78360 }, { "grad_norm": 0.1207801103591919, "learning_rate": 1.1359415326976374e-07, "loss": 0.01, "step": 78370 }, { "grad_norm": 0.13048696517944336, "learning_rate": 1.1220600436470663e-07, "loss": 0.0146, "step": 78380 }, { "grad_norm": 0.1685943901538849, "learning_rate": 1.1082637991920996e-07, "loss": 0.0131, "step": 78390 }, { "grad_norm": 0.2812291979789734, "learning_rate": 1.0945528016901851e-07, "loss": 0.0125, "step": 78400 }, { "grad_norm": 0.19992293417453766, "learning_rate": 1.0809270534840599e-07, "loss": 0.0105, "step": 78410 }, { "grad_norm": 0.133831188082695, "learning_rate": 1.067386556902028e-07, "loss": 0.0094, "step": 78420 }, { "grad_norm": 0.15672080218791962, "learning_rate": 1.05393131425785e-07, "loss": 0.0092, "step": 78430 }, { "grad_norm": 0.6508257985115051, "learning_rate": 1.0405613278505199e-07, "loss": 0.014, "step": 78440 }, { "grad_norm": 0.2686660587787628, "learning_rate": 1.027276599964766e-07, "loss": 0.0108, "step": 78450 }, { "grad_norm": 0.36325889825820923, "learning_rate": 1.0140771328704391e-07, "loss": 0.0089, "step": 78460 }, { "grad_norm": 0.13145673274993896, "learning_rate": 1.0009629288231237e-07, "loss": 0.0093, "step": 78470 }, { "grad_norm": 0.3517235815525055, "learning_rate": 9.879339900635276e-08, "loss": 0.0134, "step": 78480 }, { "grad_norm": 0.4194032549858093, "learning_rate": 9.74990318817981e-08, "loss": 0.0115, "step": 78490 }, { "grad_norm": 0.2871461808681488, "learning_rate": 9.621319172982701e-08, "loss": 0.0083, "step": 78500 }, { "grad_norm": 0.1384715437889099, "learning_rate": 9.493587877015264e-08, "loss": 0.0107, "step": 78510 }, { "grad_norm": 0.27436700463294983, "learning_rate": 9.366709322102263e-08, "loss": 0.0102, "step": 78520 }, { "grad_norm": 0.2748011350631714, "learning_rate": 9.240683529924688e-08, "loss": 0.0101, "step": 78530 }, { "grad_norm": 0.23206789791584015, "learning_rate": 9.115510522016979e-08, "loss": 0.0105, "step": 78540 }, { "grad_norm": 0.13282093405723572, "learning_rate": 8.991190319767583e-08, "loss": 0.0092, "step": 78550 }, { "grad_norm": 0.1634552776813507, "learning_rate": 8.867722944419509e-08, "loss": 0.0106, "step": 78560 }, { "grad_norm": 0.13863863050937653, "learning_rate": 8.745108417069214e-08, "loss": 0.0094, "step": 78570 }, { "grad_norm": 0.19157466292381287, "learning_rate": 8.623346758669381e-08, "loss": 0.0105, "step": 78580 }, { "grad_norm": 0.19432199001312256, "learning_rate": 8.502437990025037e-08, "loss": 0.0097, "step": 78590 }, { "grad_norm": 0.1569766253232956, "learning_rate": 8.382382131795764e-08, "loss": 0.0116, "step": 78600 }, { "grad_norm": 0.11797786504030228, "learning_rate": 8.263179204496818e-08, "loss": 0.0165, "step": 78610 }, { "grad_norm": 0.21847409009933472, "learning_rate": 8.144829228496354e-08, "loss": 0.0126, "step": 78620 }, { "grad_norm": 0.2588965594768524, "learning_rate": 8.027332224016526e-08, "loss": 0.0118, "step": 78630 }, { "grad_norm": 0.4614890515804291, "learning_rate": 7.910688211135164e-08, "loss": 0.0132, "step": 78640 }, { "grad_norm": 0.15746109187602997, "learning_rate": 7.794897209783546e-08, "loss": 0.015, "step": 78650 }, { "grad_norm": 0.1230878084897995, "learning_rate": 7.679959239746403e-08, "loss": 0.0126, "step": 78660 }, { "grad_norm": 0.2489539384841919, "learning_rate": 7.565874320664135e-08, "loss": 0.0106, "step": 78670 }, { "grad_norm": 0.2967480719089508, "learning_rate": 7.45264247203059e-08, "loss": 0.0118, "step": 78680 }, { "grad_norm": 0.5267738103866577, "learning_rate": 7.340263713194184e-08, "loss": 0.0098, "step": 78690 }, { "grad_norm": 0.27886274456977844, "learning_rate": 7.228738063356777e-08, "loss": 0.0103, "step": 78700 }, { "grad_norm": 0.3608345687389374, "learning_rate": 7.118065541575903e-08, "loss": 0.015, "step": 78710 }, { "grad_norm": 0.1943524181842804, "learning_rate": 7.008246166761435e-08, "loss": 0.0108, "step": 78720 }, { "grad_norm": 0.3755629360675812, "learning_rate": 6.899279957679472e-08, "loss": 0.0117, "step": 78730 }, { "grad_norm": 0.28314176201820374, "learning_rate": 6.791166932949011e-08, "loss": 0.0075, "step": 78740 }, { "grad_norm": 0.36767157912254333, "learning_rate": 6.683907111043608e-08, "loss": 0.012, "step": 78750 }, { "grad_norm": 0.07574048638343811, "learning_rate": 6.577500510290824e-08, "loss": 0.0117, "step": 78760 }, { "grad_norm": 0.26878422498703003, "learning_rate": 6.47194714887278e-08, "loss": 0.0238, "step": 78770 }, { "grad_norm": 0.1558740884065628, "learning_rate": 6.367247044825608e-08, "loss": 0.0112, "step": 78780 }, { "grad_norm": 0.14710654318332672, "learning_rate": 6.263400216039994e-08, "loss": 0.0125, "step": 78790 }, { "grad_norm": 0.27569735050201416, "learning_rate": 6.160406680260078e-08, "loss": 0.0201, "step": 78800 }, { "grad_norm": 0.07290557026863098, "learning_rate": 6.058266455084561e-08, "loss": 0.0114, "step": 78810 }, { "grad_norm": 0.1710890680551529, "learning_rate": 5.956979557967257e-08, "loss": 0.0108, "step": 78820 }, { "grad_norm": 0.28184741735458374, "learning_rate": 5.856546006214325e-08, "loss": 0.0134, "step": 78830 }, { "grad_norm": 0.27534258365631104, "learning_rate": 5.7569658169881466e-08, "loss": 0.0097, "step": 78840 }, { "grad_norm": 0.22377678751945496, "learning_rate": 5.658239007303445e-08, "loss": 0.0146, "step": 78850 }, { "grad_norm": 0.32134178280830383, "learning_rate": 5.560365594030059e-08, "loss": 0.0106, "step": 78860 }, { "grad_norm": 0.2177514284849167, "learning_rate": 5.463345593891833e-08, "loss": 0.0134, "step": 78870 }, { "grad_norm": 0.3241778016090393, "learning_rate": 5.367179023467173e-08, "loss": 0.0128, "step": 78880 }, { "grad_norm": 0.23110923171043396, "learning_rate": 5.271865899187378e-08, "loss": 0.0104, "step": 78890 }, { "grad_norm": 0.3110871911048889, "learning_rate": 5.177406237340532e-08, "loss": 0.0278, "step": 78900 }, { "grad_norm": 0.18772602081298828, "learning_rate": 5.083800054065946e-08, "loss": 0.0137, "step": 78910 }, { "grad_norm": 0.21031789481639862, "learning_rate": 4.991047365358603e-08, "loss": 0.0118, "step": 78920 }, { "grad_norm": 0.545742928981781, "learning_rate": 4.899148187067493e-08, "loss": 0.0205, "step": 78930 }, { "grad_norm": 0.10016516596078873, "learning_rate": 4.808102534895609e-08, "loss": 0.0097, "step": 78940 }, { "grad_norm": 0.1318632811307907, "learning_rate": 4.717910424400507e-08, "loss": 0.009, "step": 78950 }, { "grad_norm": 0.29457080364227295, "learning_rate": 4.628571870993193e-08, "loss": 0.0106, "step": 78960 }, { "grad_norm": 0.22078564763069153, "learning_rate": 4.5400868899392324e-08, "loss": 0.0145, "step": 78970 }, { "grad_norm": 0.3628736138343811, "learning_rate": 4.4524554963576436e-08, "loss": 0.0152, "step": 78980 }, { "grad_norm": 0.2644658386707306, "learning_rate": 4.36567770522367e-08, "loss": 0.0093, "step": 78990 }, { "grad_norm": 0.14777018129825592, "learning_rate": 4.279753531364339e-08, "loss": 0.0125, "step": 79000 }, { "grad_norm": 0.294202595949173, "learning_rate": 4.1946829894617955e-08, "loss": 0.0144, "step": 79010 }, { "grad_norm": 0.1458333134651184, "learning_rate": 4.11046609405219e-08, "loss": 0.0201, "step": 79020 }, { "grad_norm": 0.26821812987327576, "learning_rate": 4.027102859526233e-08, "loss": 0.0087, "step": 79030 }, { "grad_norm": 0.18708501756191254, "learning_rate": 3.9445933001280856e-08, "loss": 0.0132, "step": 79040 }, { "grad_norm": 0.16623149812221527, "learning_rate": 3.8629374299564704e-08, "loss": 0.0154, "step": 79050 }, { "grad_norm": 0.18213048577308655, "learning_rate": 3.78213526296467e-08, "loss": 0.0119, "step": 79060 }, { "grad_norm": 0.39959198236465454, "learning_rate": 3.702186812958308e-08, "loss": 0.014, "step": 79070 }, { "grad_norm": 0.2656205892562866, "learning_rate": 3.623092093599789e-08, "loss": 0.0166, "step": 79080 }, { "grad_norm": 0.3169934153556824, "learning_rate": 3.544851118403303e-08, "loss": 0.0128, "step": 79090 }, { "grad_norm": 0.23685899376869202, "learning_rate": 3.46746390073871e-08, "loss": 0.0134, "step": 79100 }, { "grad_norm": 0.2445598542690277, "learning_rate": 3.390930453829322e-08, "loss": 0.0161, "step": 79110 }, { "grad_norm": 0.21558581292629242, "learning_rate": 3.3152507907519005e-08, "loss": 0.0127, "step": 79120 }, { "grad_norm": 0.15141373872756958, "learning_rate": 3.240424924438323e-08, "loss": 0.0097, "step": 79130 }, { "grad_norm": 0.16534723341464996, "learning_rate": 3.1664528676750295e-08, "loss": 0.011, "step": 79140 }, { "grad_norm": 0.2117430567741394, "learning_rate": 3.093334633100797e-08, "loss": 0.0192, "step": 79150 }, { "grad_norm": 0.2553882896900177, "learning_rate": 3.021070233210077e-08, "loss": 0.0144, "step": 79160 }, { "grad_norm": 0.34270796179771423, "learning_rate": 2.9496596803507692e-08, "loss": 0.0106, "step": 79170 }, { "grad_norm": 0.2000289261341095, "learning_rate": 2.879102986725335e-08, "loss": 0.0106, "step": 79180 }, { "grad_norm": 0.19224874675273895, "learning_rate": 2.809400164389131e-08, "loss": 0.0142, "step": 79190 }, { "grad_norm": 0.34275561571121216, "learning_rate": 2.740551225253185e-08, "loss": 0.011, "step": 79200 }, { "grad_norm": 0.3139853775501251, "learning_rate": 2.6725561810819754e-08, "loss": 0.0228, "step": 79210 }, { "grad_norm": 0.4068152904510498, "learning_rate": 2.605415043493431e-08, "loss": 0.0116, "step": 79220 }, { "grad_norm": 0.19698075950145721, "learning_rate": 2.5391278239605965e-08, "loss": 0.0125, "step": 79230 }, { "grad_norm": 0.25831368565559387, "learning_rate": 2.4736945338094118e-08, "loss": 0.0107, "step": 79240 }, { "grad_norm": 0.263300359249115, "learning_rate": 2.4091151842214887e-08, "loss": 0.0105, "step": 79250 }, { "grad_norm": 0.2788829505443573, "learning_rate": 2.3453897862318884e-08, "loss": 0.0145, "step": 79260 }, { "grad_norm": 0.2488081157207489, "learning_rate": 2.2825183507285686e-08, "loss": 0.0092, "step": 79270 }, { "grad_norm": 0.46368497610092163, "learning_rate": 2.220500888455157e-08, "loss": 0.0152, "step": 79280 }, { "grad_norm": 0.1671268194913864, "learning_rate": 2.1593374100081777e-08, "loss": 0.0146, "step": 79290 }, { "grad_norm": 0.1668635755777359, "learning_rate": 2.09902792583927e-08, "loss": 0.0182, "step": 79300 }, { "grad_norm": 0.7633387446403503, "learning_rate": 2.0395724462540788e-08, "loss": 0.01, "step": 79310 }, { "grad_norm": 0.4003687798976898, "learning_rate": 1.9809709814111453e-08, "loss": 0.0144, "step": 79320 }, { "grad_norm": 0.1851893812417984, "learning_rate": 1.923223541324126e-08, "loss": 0.0167, "step": 79330 }, { "grad_norm": 0.21907110512256622, "learning_rate": 1.8663301358606833e-08, "loss": 0.0146, "step": 79340 }, { "grad_norm": 0.5862895846366882, "learning_rate": 1.8102907747419295e-08, "loss": 0.0135, "step": 79350 }, { "grad_norm": 0.3544244170188904, "learning_rate": 1.7551054675435385e-08, "loss": 0.0123, "step": 79360 }, { "grad_norm": 0.14762867987155914, "learning_rate": 1.7007742236957447e-08, "loss": 0.0117, "step": 79370 }, { "grad_norm": 0.22676216065883636, "learning_rate": 1.647297052481678e-08, "loss": 0.0124, "step": 79380 }, { "grad_norm": 0.1250983476638794, "learning_rate": 1.5946739630390285e-08, "loss": 0.0151, "step": 79390 }, { "grad_norm": 0.15625926852226257, "learning_rate": 1.5429049643606032e-08, "loss": 0.0171, "step": 79400 }, { "grad_norm": 0.17863605916500092, "learning_rate": 1.4919900652909935e-08, "loss": 0.0145, "step": 79410 }, { "grad_norm": 0.2197185605764389, "learning_rate": 1.4419292745310175e-08, "loss": 0.0154, "step": 79420 }, { "grad_norm": 0.25683537125587463, "learning_rate": 1.3927226006343885e-08, "loss": 0.0121, "step": 79430 }, { "grad_norm": 0.24860970675945282, "learning_rate": 1.3443700520093805e-08, "loss": 0.0144, "step": 79440 }, { "grad_norm": 0.16020198166370392, "learning_rate": 1.296871636917718e-08, "loss": 0.0117, "step": 79450 }, { "grad_norm": 0.17942117154598236, "learning_rate": 1.2502273634762419e-08, "loss": 0.0101, "step": 79460 }, { "grad_norm": 0.3173125684261322, "learning_rate": 1.2044372396546876e-08, "loss": 0.0151, "step": 79470 }, { "grad_norm": 0.2072560042142868, "learning_rate": 1.1595012732773524e-08, "loss": 0.0119, "step": 79480 }, { "grad_norm": 0.2646748125553131, "learning_rate": 1.1154194720225386e-08, "loss": 0.0152, "step": 79490 }, { "grad_norm": 0.2822803258895874, "learning_rate": 1.0721918434231093e-08, "loss": 0.0099, "step": 79500 }, { "grad_norm": 0.183882936835289, "learning_rate": 1.0298183948648232e-08, "loss": 0.0082, "step": 79510 }, { "grad_norm": 0.09816506505012512, "learning_rate": 9.88299133588555e-09, "loss": 0.0086, "step": 79520 }, { "grad_norm": 0.2610109746456146, "learning_rate": 9.476340666891847e-09, "loss": 0.0178, "step": 79530 }, { "grad_norm": 0.1341410130262375, "learning_rate": 9.078232011139332e-09, "loss": 0.0097, "step": 79540 }, { "grad_norm": 0.21867656707763672, "learning_rate": 8.68866543666802e-09, "loss": 0.0098, "step": 79550 }, { "grad_norm": 0.09893544018268585, "learning_rate": 8.307641010035782e-09, "loss": 0.0104, "step": 79560 }, { "grad_norm": 0.4964182674884796, "learning_rate": 7.93515879635165e-09, "loss": 0.0162, "step": 79570 }, { "grad_norm": 0.20345312356948853, "learning_rate": 7.571218859264706e-09, "loss": 0.0092, "step": 79580 }, { "grad_norm": 0.43799376487731934, "learning_rate": 7.215821260958544e-09, "loss": 0.0148, "step": 79590 }, { "grad_norm": 0.269979864358902, "learning_rate": 6.868966062162363e-09, "loss": 0.0115, "step": 79600 }, { "grad_norm": 0.29472848773002625, "learning_rate": 6.530653322145419e-09, "loss": 0.0094, "step": 79610 }, { "grad_norm": 0.1492508351802826, "learning_rate": 6.200883098717025e-09, "loss": 0.0087, "step": 79620 }, { "grad_norm": 0.40636733174324036, "learning_rate": 5.879655448226551e-09, "loss": 0.0138, "step": 79630 }, { "grad_norm": 0.13951733708381653, "learning_rate": 5.566970425557872e-09, "loss": 0.0104, "step": 79640 }, { "grad_norm": 0.33557137846946716, "learning_rate": 5.26282808414047e-09, "loss": 0.0115, "step": 79650 }, { "grad_norm": 0.23936845362186432, "learning_rate": 4.967228475949437e-09, "loss": 0.0134, "step": 79660 }, { "grad_norm": 0.27580344676971436, "learning_rate": 4.680171651494369e-09, "loss": 0.0096, "step": 79670 }, { "grad_norm": 0.28599467873573303, "learning_rate": 4.40165765981937e-09, "loss": 0.0082, "step": 79680 }, { "grad_norm": 0.3050116002559662, "learning_rate": 4.131686548519698e-09, "loss": 0.0099, "step": 79690 }, { "grad_norm": 0.2043679803609848, "learning_rate": 3.8702583637251214e-09, "loss": 0.0132, "step": 79700 }, { "grad_norm": 0.14481548964977264, "learning_rate": 3.617373150105463e-09, "loss": 0.0097, "step": 79710 }, { "grad_norm": 0.5462249517440796, "learning_rate": 3.3730309508706036e-09, "loss": 0.0109, "step": 79720 }, { "grad_norm": 0.2503841519355774, "learning_rate": 3.137231807781582e-09, "loss": 0.0113, "step": 79730 }, { "grad_norm": 0.48723146319389343, "learning_rate": 2.9099757611172894e-09, "loss": 0.0133, "step": 79740 }, { "grad_norm": 0.15710310637950897, "learning_rate": 2.6912628497133275e-09, "loss": 0.0118, "step": 79750 }, { "grad_norm": 0.2583891749382019, "learning_rate": 2.481093110945354e-09, "loss": 0.0134, "step": 79760 }, { "grad_norm": 0.19889004528522491, "learning_rate": 2.2794665807235327e-09, "loss": 0.0092, "step": 79770 }, { "grad_norm": 0.15039566159248352, "learning_rate": 2.0863832934980843e-09, "loss": 0.0081, "step": 79780 }, { "grad_norm": 0.18292726576328278, "learning_rate": 1.901843282264837e-09, "loss": 0.0124, "step": 79790 }, { "grad_norm": 0.12231525778770447, "learning_rate": 1.7258465785541245e-09, "loss": 0.0149, "step": 79800 }, { "grad_norm": 0.22091276943683624, "learning_rate": 1.5583932124418887e-09, "loss": 0.012, "step": 79810 }, { "grad_norm": 0.17091438174247742, "learning_rate": 1.3994832125441282e-09, "loss": 0.0123, "step": 79820 }, { "grad_norm": 0.3192465305328369, "learning_rate": 1.2491166060057958e-09, "loss": 0.0134, "step": 79830 }, { "grad_norm": 0.14741250872612, "learning_rate": 1.1072934185230032e-09, "loss": 0.0133, "step": 79840 }, { "grad_norm": 0.2149595469236374, "learning_rate": 9.740136743319195e-10, "loss": 0.0127, "step": 79850 }, { "grad_norm": 0.22203706204891205, "learning_rate": 8.492773962087696e-10, "loss": 0.0104, "step": 79860 }, { "grad_norm": 0.31613707542419434, "learning_rate": 7.330846054587337e-10, "loss": 0.014, "step": 79870 }, { "grad_norm": 0.28419795632362366, "learning_rate": 6.254353219492526e-10, "loss": 0.0133, "step": 79880 }, { "grad_norm": 0.20674501359462738, "learning_rate": 5.263295640600685e-10, "loss": 0.0101, "step": 79890 }, { "grad_norm": 0.14578399062156677, "learning_rate": 4.357673487387359e-10, "loss": 0.0123, "step": 79900 }, { "grad_norm": 0.41159188747406006, "learning_rate": 3.537486914506616e-10, "loss": 0.0125, "step": 79910 }, { "grad_norm": 0.22629214823246002, "learning_rate": 2.8027360621241117e-10, "loss": 0.0122, "step": 79920 }, { "grad_norm": 0.18904425203800201, "learning_rate": 2.153421055806071e-10, "loss": 0.0128, "step": 79930 }, { "grad_norm": 0.16524341702461243, "learning_rate": 1.5895420064637733e-10, "loss": 0.0115, "step": 79940 }, { "grad_norm": 0.22774125635623932, "learning_rate": 1.1110990105200891e-10, "loss": 0.012, "step": 79950 }, { "grad_norm": 0.2907968759536743, "learning_rate": 7.180921496874326e-11, "loss": 0.013, "step": 79960 }, { "grad_norm": 0.24471010267734528, "learning_rate": 4.10521491134297e-11, "loss": 0.012, "step": 79970 }, { "grad_norm": 0.2782478928565979, "learning_rate": 1.883870874297422e-11, "loss": 0.015, "step": 79980 }, { "grad_norm": 0.4603424370288849, "learning_rate": 5.1688976432373584e-12, "loss": 0.0134, "step": 79990 }, { "grad_norm": 0.17334619164466858, "learning_rate": 4.271816234080461e-14, "loss": 0.0183, "step": 80000 } ], "logging_steps": 10, "max_steps": 80000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 20000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 8, "trial_name": null, "trial_params": null }